Cover

title:
author:
publisher:
isbn10 | asin:
print isbn13:
ebook isbn13:
language:
subject
publication date:
lcc:
ddc:
subject:

Encyclopedia of PVC
Nass, Leonard I.
CRC Press
0824774272
9780824774271
9780585386447
English
Polyvinyl chloride.
1998
TP1180.V48E5 1998eb
668.4/237
Polyvinyl chloride.

Page i
Encyclopedia of PVC

Page ii

Page iii

Encyclopedia of PVC
Second Edition, Revised and Expanded
Volume 1: Resin Manufacture and Properties
edited by
LEONARD I. NASS L.I. Nass Company Warren, New Jersey
CHARLES A. HEIBERGER Plastics Industry Consultant Princeton, New Jersey
MARCEL DEKKER, INC. New York and Basel

Page iv
Library of Congress Cataloging in Publication Data
Main entry under title:
Encyclopedia of PVC.
Includes bibliographies and indexes.
Contents: v. 1. Resin manufacture and properties.
1. Polyvinyl chloride. I. Nass, Leonard I.
II. Heiberger, Charles A.
TP1180.V48E5 1985 668.4'237 8520753
ISBN 0-8247-7427-2 (v. 1)

COPYRIGHT 1986 MARCEL DEKKER, INC. ALL RIGHTS RESERVED

Neither this book nor any part may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, microfilming, and recording, or by any
information storage and retrieval system, without permission in writing from the publisher.
MARCEL DEKKER, INC.
270 Madison Avenue, New York, New York 10016
Current printing (last digit):
10 9 8 7 6 5 4 3 2 1
PRINTED IN THE UNITED STATES OF AMERICA

Page v

Charles Heiberger left us on September 13, 1985, just as this volume was being readied for
final printing and publication. However, his presence as my coeditor has insured that he will
always be with us, at least in memory. This volume, and the three remaining volumes to come,
are a reflection of his contributions to this series.
With Charles' death, I lost a collaborator who was very special. His many ideas were innovative
and stimulating and his presence was inspiring. The reorganization of topics into our new format
and the subsequent expansion of subject matter and of the series from three to four volumes
result mainly from Charles' incisive thinking and powers of persuasion. It is also through
Charles' efforts that we have so many new contributors working with us on this revised series.
It was Charles who first expounded the thought that the new Encyclopedia of PVC would not
necessarily make obsolete the original edition. What our new authors have done is to provide
additional insight and complementary viewpoints to topics that were originally well prepared and
documented some 10 years ago. Thus, both editions can rightfully stand together on one's
shelves, side by side today, and this too is because of Charles' organizational abilities.
It will not be easy finishing this work without you, Charles. You were not just a staunch
collaborator, but a good friend as well.
Leonard I. Nass

Page vi

Page vii

Foreword

Old soldiers never diethey just fade away. Old plastics never diethey hang on and stay.
There exist two outstanding examples that prove the validity of the metaphoric version of this
sentence: the thermoset phenolics (Bakelite) and the thermoplastic polyvinyl chlorides (PVC).
The former just celebrated the 75th anniversary of its appearance; the latter became
commercial about 50 years ago. There is, however, a significant difference between these two
classical synthetics: Bakelite has beenand still isessentially restricted to those applications
which need hard, thermostable materials; PVC, on the other hand, has beenand still is
expanding over all possible practical uses from soft rubbery shoe soles and garden hoses to
hard and tough window frames and to giant sewer pipes. In 1983 the worldwide production and
consumption of PVC was 12.5 million metric tons (27.5 billion pounds) and there is all reason to
expect it will be about 17 million tons (37.4 billion pounds) in 1990.
At this time, it is extremely desirable that a new comprehensive presentation of all the many
facets of this important material should be published in the form of four volumes of an
encyclopedia where each chapter is contributed by a qualified expert to give an account of the
development together with a detailed description of the present state of the art in synthesis,
characterization, structure, properties, processing, and applications.
Such encyclopedic presentations with fact and figure collections can easily become
monotonous and purely informativelike a telephone bookbut they also may, beyond that, be
educational and stimulating when the material is not presented as a dry enumeration of data
and events but has suggestive concepts and ideas appropriately sprinkled throughout the text.
The editors and authors have been remarkably successful in introducing into these volumes a
refreshing flavor of progressive and imaginative contribution, which make them a source not
only of reliable information but also of intellectual challenge. All chapters of the four volumes will
convince the reader that PVC is an unusual polymer and will remain an important leader in the
vast fields of its preferred uses for many more years.
PVC is technically not an organic resin because it contains normally 56.8 weight percent of
chlorine and, in its after-chlorinated types, as much

Page viii
as 65 weight percent. Essentially, it is an inorganic-organic hybrid polymer and owes many of
its interesting properties, such as flame retardancy, resistance to common solvents, polarity,
compatibility with additives, and high softening characteristics, to the presence of its inorganic
component. There exist several other inorganic-organic polymers: for example, polyvinylidene
chloride (Saran), polytetrafluoroethylene (Teflon), and the large families of silicones and
phosphazenes. In all these cases, it is the inorganic componentsilicon, phosphorus, flourine,
chlorinewhich endows these materials with such valuable properties that all the enumerated
polymers are commercially very successful even though in most cases the monomeric species
are rather expensive and the methods of polymerization are not simple. In all these cases the
inorganic component is a chemical constituent of each macromolecule. It is known, however,
that the mechanical blend of truly organic polymers with inorgano-organic species also offers
interesting new properties, and even the simple admixture of inorganic fillers to organic
polymers is a field of widespread, successful activities, including the loading of PVC with finely
powdered inorganics.
Any new large-scale demand for specific property combinations generates ideas for novel
compositions and processing techniques, and the users of this encyclopedia will find their
expectations for information, stimulations, and guidance fully and wholly satisfied.
Herman Mark Polytechnic Institute of New York
Brooklyn, New York

Page ix

Preface

The original edition of this work was published in 197677 as a three-volume comprehensive
reference work on PVC technology as it existed prior to about 1975. Since then it has become
increasingly evident that an updated edition is necessary to cover a number of significant
changes and advances that have occurred in PVC science and technology, among which may
be mentioned:
Up-to-date handling methods for VCM; compliance with the EPA, OSHA, and FDA regulatory
procedures in the United States
Computerized large reactor polymerization processes with improved resin quality and control,
reduced environmental hazards, and better economics
Newly developed compounding additives plus improved compounding processes demonstrating
better performance at lower cost, including fully automated continuous compounding methods
New processing techniques and equipment designs with computerized controls for faster
throughput cycles with better quality control
New chlorinated-PVC materials and compounds with broad application in extrusion and molding
markets
Renewed emphasis and attention to foamed, rigid PVC products in both Europe and the
Western Hemisphere
The new Encyclopedia of PVC has now been expanded to four volumes with individual chapters
grouped according to common interests: Volume 1, Resin Manufacture and Properties; Volume
2, Compound Design and Additives; Volume 3, Compounding Processes, Product Design, and
Specifications; and Volume 4, Conversion and Fabrication Processes.
Safety and environmental concerns are now covered in two separate chapters: in Volume 1 for
resins, and in Volume 3 for compounds and finished products. Likewise, the subjects of
stabilization and plasticization have been divided: in Volume 1 when relating to resin properties,
and in Volume 2 where individual stabilizer and plasticizer materials and formulations are more
appropriately grouped with other additive classes. The earlier chapter on chemical

modifications of PVC has been replaced by a more complete coverage of chlorinated PVC,
which lately has become an important engineering plastic.

Page x
New chapters covering the subjects of block and graft interpolymers, alloys, and polyblends;
PVC foams; and testing of flexible PVC products have been added, emphasizing recent
developments in these new or renewed-interest technologies.
Special efforts have been made to have a stronger international representation not only in the
selection of individual contributors, but also in the consideration and coverage of significant
foreign technology and methodology by all authors.
Each author is an expert in his field as well as knowledgeable in PVC technology as a whole. In
all cases, their individual contributions have been unified and enhanced by input from the editors
and from other authors. Hopefully, the readers of this new edition will benefit greatly from this
cross-fertilization of ideas.
Leonard I. NassCharles A. Heiberger

Page xi

Contributors to Volume 1

Lyle F. Albright, Purdue University, Lafayette, Indiana

John T. Barr, Air Products and Chemicals, Inc., Allentown, Pennsylvania
Eugen Bezdadea, Deutsches Kunststoff-Institut, Darmstadt, Federal Republic of Germany
Daniel J. Brandt, Occidental Chemical Corporation, Pottstown, Pennsylvania
Dietrich Braun, Deutsches Kunststoff-Institut, Darmstadt, Federal Republic of Germany
Joseph R. Darby, Darby Consulting, Inc., Webster Groves, Missouri
Richard S. Guise, Occidental Chemical Corporation, Pottstown, Pennsylvania
Michael Langsam, Air Products and Chemicals, Inc., Allentown, Pennsylvania
John A. Manson, Lehigh University, Bethlehem, Pennsylvania
Gilbert A. Martello, B. F. Goodrich Chemical Group, Cleveland, Ohio
Bernard A. McCoy, Occidental Chemical Corporation, Pottstown, Pennsylvania
Richard G. Parker, B. F. Goodrich Company, Brecksville, Ohio
J. Kern Sears, Monsanto Company, St. Louis, Missouri

Elliott L. Weinberg, Technical Information Exchange, East Brunswick, New Jersey

Donald E. Witenhafer, B. F. Goodrich Company, Brecksville, Ohio

Page xii

Page xiii

Contents of Volume 1

FOREWORD (Herman Mark)

PREFACE

v
vii

CONTRIBUTORS TO VOLUME 1

ix

CONTENTS OF THE OTHER VOLUMES

xv

1. PVC: PAST, PRESENT, AND FUTURE Elliott L. Weinberg

I. Historical Development of the PVC Industry

II. Current Commercial Significance

III. Future Trends

15

IV. Summary

24

References

24

Appendix: A History of Polyvinyl Chloride

25

2. VINYL CHLORIDE PRODUCTION Lyle F. Albright

27

I. Introduction
II. Major Process for Production of Vinyl Chloride
III. Hydrochlorination of Acetylene
IV. Technology of Possible Importance
V. Economic and Environmental Considerations
VI. Storage and Handling of Vinyl Chloride
VII. Summary

27
29
37
41
42
43
43

References
3. PVC PROCESSES AND MANUFACTURE Michael Langsam
I. Introduction

44
47
48

II. Polymerization Overview

48

III. Classical Free-Radical Polymerization

49

Page xiv
IV. Nonclassical Free-Radical Polymerization
V. Polymer Chain Structure

57
62

VI. Commercial Manufacturing Processes

69

VII. Suspension Polymerization

72

VIII. Bulk Polymerization

127

IX. Emulsion Polymerization

X. Future Developments

138
152

Symbols

153

References

154

4. COPOLYMER PROCESSES AND MANUFACTURE Michael Langsam

I. Introduction
II. Scope and Size of the Commercial Copolymer Market
III. Copolymerization Process

161
162
162
164

IV. Copolymer Structure and Properties

178

V. Overview of Commercial Processes

199

VI. Vinyl Acetate Copolymers

200

VII. Olefin Copolymers

VIII. Acrylic-Acrylonitrile Copolymers
IX. Future Developments
Symbols
References
5. SAFETY AND ENVIRONMENTAL CONCERNS IN RESIN MANUFACTURE
John T. Barr
I. Introduction

216
229
231
231
232
239
240

 II. Toxicity Considerations

III. Regulatory Status
IV. General Safety Procedures

240
260
268

V. Waste Streams

285

VI. Analytical Methods

287

VII. Vinyl Acetate Copolymers

289

Glossary of Acronyms
References

290
291

6. RESIN STRUCTURE AND PROPERTIES Donald E. Witenhafer

309

I. PVC Molecular Structure

309

II. Supermolecular Structures of PVC

III. Properties of PVC
References

320
332
342

Page xv
7. RESIN TESTING AND QUALITY ASSURANCE Daniel J. Brandt, Richard S.
Guise, and Bernard A. McCoy
I. Introduction

347
348

II. Sources of Test Methods

348

III. Analytical

350

IV. Physical Properties: General-Purpose Resins

362

V. Physical Properties: Dispersion Resins

377

VI. Physical Properties: Solution Resins

387

VII. ASTM Resin Classification

387

VIII. Quality Assurance

390

References

393

8. THEORY OF DEGRADATION AND STABILIZATION MECHANISMS Dietrich

Braun and Eugen Bezdadea

397

I. Introduction
II. Degradation and Stability of PVC
III. Thermal Degradation
IV. Photolytic Degradation
V. Stabilization
References
9. SOLVATION AND PLASTICIZATION J. Kern Sears and Joseph R. Darby
I. Introduction
II. Historical Perspective of Plasticization
III. Mechanism of Plasticizer Action
IV. Compatibility of Plasticizer with Resin
V. External Versus Internal Plasticization

397
398
411
413
416
429
435
436
437
440
448
477

 VI. Plasticization of Polyvinyl Chloride

478

VII. Effect on Physical Properties

490

VIII. Permanence Properties of Plasticized PVC

500

IX. Polymerizable Plasticizers

534

X. Plasticization by Fillers, Pigments, Salts, and Air

XI. Antiplasticizers and Antiplasticization
XII. Toxicity and Environmental Aspects

535
538
542

XIII. What Plasticizer to Use: The Desirability Function

544

References

547

10. PVC BLENDS, BLOCKS, GRAFTS, AND INTERPENETRATING POLYMER

NETWORKS John A. Manson

555

I. Introduction
II. Principles of Polymer-Polymer Systems

555
556

Page xvi
III. PVC Blends, Blocks, Grafts, and Interpenetrating Polymer
Networks
References

571
604

11. CHEMICAL MODIFICATIONS: CHLORINATED PVC Richard G. Parker and

Gilbert A. Martello

619

I. Introduction

619

II. Methods of Chlorination

620

III. The Structure of CPVCs and the Mechanism of Chlorination

623

IV. Thermal Stability and Stabilization

632

V. Properties

636

VI. Markets and Commercial Uses

648

References

651

AUTHOR INDEX

655

SUBJECT INDEX

693

Page xvii

Contents of the Other Volumes

VOLUME 2: Compound Design and Additives

1. Formulation Considerations and Design, Norman L. Perry
2. Stabilizers: Types, Properties, and Performance, Thomas Jennings and Charles Fletcher
3. Plasticizers: Types, Properties, and Performance, Leonard G. Krauskopf
4. Lubricants: Types, Properties, and Performance, Robert Lindner and Kurt Worschech
5. Polymeric Modifiers: Types, Properties, and Performance, Vivian E. Malpass, John T. Lutz,
and Robert P. Petrich
6. Fillers: Types, Properties, and Performance, Kris Mathur and Dennis Van der Heiden
7. Colorants: Types, Properties, and Performance, John Graff
8. Special Additives: Types, Properties, and Performance, Saul Gobstein
VOLUME 3: Compounding Processes, Product Design, and Specifications
1. Equipment for Mixing and Compounding, Fred Roesler and Ashok Shah
2. Process Control in Compounding, Robert Brookman
3. Solid Vinyl Compound Systems, Zenas Crocker, David V. Wise, and Ved P. Gupta
4. Liquid Vinyl Compound Systems, Russell Park
5. Testing of Rigid PVC Products: Significance of Test Results, I. Luis Gomez and Peter F. V.
Lloyd
6. Testing of Flexible PVC Products: Significance of Test Results, R. A. Ronczka
7. Specifications for Vinyl Products, Lewis B. Weisfeld

 8. Environmental and Occupational Safety/Health Considerations, John T. Barr

VOLUME 4: Conversion and Fabrication Processes
1. General Principles of Plant Operation: Operating a Plant for Profit, Norman L. Perry

Page xviii
2. Rheology and Processability, Pravin L. Shah
3. Extrusion, Martin Batiuk
4. Molding: Injection, Compression, Blow, B. C. Harris
5. Calendering and Calender Laminating, William M. Coaker
6. Spread Coating and Lamination, Arnold Werner and Charles Brower
7. Vinyl Foam Processes, Drago Sencar
8. Miscellaneous Molding/Casting and Coating Processes from Powders, Douglas Richart
9. Miscellaneous Molding/Casting and Coating Processes from Liquids, David Hurter
10. Postfabrication Processes: Decorating and Finishing, Thomas W. Lembo, Albert Seaman,
and Claude Henning

Page 1

1 PVC: Past, Present, and Future

ELLIOTT L. WEINBERG
Technical Information Exchange East Brunswick, New Jersey

I. HISTORICAL DEVELOPMENT OF THE PVC INDUSTRY

A. Introduction

B. Discovery and Growth

II. CURRENT COMMERCIAL SIGNIFICANCE

A. Production and Competition

B. Markets

C. Economic Impact of the Overall PVC Industry

13

D. Environmental and Safety Aspects

14

III. FUTURE TRENDS

15

A. The PVC Industry

15

B. Monomer
C. Polymerization
D. Compounding and Fabrication
IV. SUMMARY
REFERENCES
APPENDIX: A HISTORY OF POLYVINYL CHLORIDE

I.HISTORICAL DEVELOPMENT OF THE PVC INDUSTRY

15
16
16
24
24
25

A.Introduction
Polymerization of vinyl chloride, with or without other vinyl monomers, by free-radical initiation
generates polymers or copolymers which are converted by compounding and fabrication into
plastic articles and devices of great variety. Vinyl chloride polymers are uniquely responsive to
functional additives which permit the generation of rigid and flexible products, useful in novelties
at low cost and in designed engineering applications.
In contrast to the orderly scientific and technical development of the polyolefin polymer industry,
polyvinyl chloride (PVC) has developed into a worldwide multibillion-pound business in a very
unplanned fashion. Only in recent years has real scientific effort been applied to the
understanding of PVC and its proper formulation and fabrication. The inherently favorable

Page 2
economics, based on the high chlorine content and improved manufacturing techniques, point to
the continued high-volume use of vinyl chloride polymers using newly discovered technology for
compounding and fabrication. A continuing shakeout in the ownership of resin, additive, and
fabrication facilities is to be expected for the mature PVC industry. Other chapters in this
encyclopedia elaborate on much of the above and other comments that follow.
It is interesting to note that it took 100 years for the monomer, vinyl chloride, to be produced
and converted to polymer in multimillion-pound quantity. Real growth took place in the Plastic
Age following World War II. As with growth and maturation for all things, all sorts of problems
have beset the PVC industry, and the industry has responded well to protect and promote its
interest. The existence of the Vinyl Institute illustrates the success of the public relations effort.
B.Discovery and Growth
The monomer, vinyl chloride, was discovered by Justus von Liebig in 1835. This chlorinecontaining gas resulted from the reaction of ethylene dichloride with alcoholic potassium
hydroxide. On assignment from Liebig, Victor Regnault confirmed this reaction and was allowed
to publish alone and gain credit for this discovery. By 1838, Regnault observed what he thought
to be polymerized vinyl chloride, but study indicated that he had polyvinylidene chloride. The
first report of authentic vinyl halide polymer was made in 1860 by A. W. Hoffman. He noted the
change of vinyl bromide monomer to a white mass without compositional change. This polyvinyl
bromide resin was followed in 1872 by E. Baumann's discovery of polyvinyl chloride. Baumann
detailed the light-induced change of vinyl chloride monomer to solid products which he thought
were isomers of the monomer. The properties described by him are those we ascribe to
polyvinyl chloride (1).
The further development of polyvinyl chloride is a tale of two continents as well as different
reasons and objectives. Actually, it took 50 years for the issue of a German patent for the
manufacture of vinyl chloride by reaction of acetylene and hydrogen chloride. This work resulted
from the assignment to Fritz Klatte by his superiors at Chemische Fabrik Griesheim-Elektron of
the task of finding uses for acetylene available from the excess calcium carbide no longer used
in gas lighting. The monomer synthesis patents were followed by polymerization patents for the
same company; these initial polymerization processes cited light and/or heat as the initiators.
Free-radical polymerization initiators were patented in 1914. The same substances as those in
use today were listed: for example, organic peroxides, ozonides, and organic acid anhydrides in
combination with oxygen or oxygen-producing substances such as perborates and
percarbonates.
The practical interest in this polymer of Klatte and his co-workers was to find a replacement for
celluloid. By using the same processing techniques as those used for celluloid, polyvinyl chloride
resin was converted or proposed for conversion to useful plastics as substitutes for horn, film,
fibers, and lacquers of natural origin by hot water forming and casting from solution.
Plasticization was attempted with camphor. At about the same time, Ivan Ostromislensky in
Moscow studied polyvinyl halides in an attempt to generate rubberlike materials. This effort
failed and Ostromislensky did further work

Page 3
on vinyl and polyvinyl chloride years later in the United States. All this work came too early in
the development process and was not sufficiently backed by structural understanding such as
that suggested by Staudinger's macromolecular theory. More than a decade passed before
interest was aroused again, industrial laboratory efforts in PVC beginning in earnest in the
1930s [2]. Griesheim-Elektron allowed its patents to lapse and this may have opened the door
for other companies.
In the United States, Carbide and Carbon Chemical Corp., now Union Carbide Corp., had in this
period, an inordinate amount of by-product ethylene dichloride generated from its production of
ethylene chlorohydrin. Caustic dehydrochlorination produced vinyl chloride monomer, which was
polymerized in both solution and mass to obtain both homo- and copolymers. This effort
resulted in a 4-million-pound polymer plant at South Charleston, West Virginia, in 1936 [3]. At
about this time, Waldo Semon began to work with vinyl chloride and its polymers at B. F.
Goodrich in an attempt to make an adhesive from a simple synthetic organic polymer. He
reviewed the work of the pioneers mentioned above and prepared flexible and elastic materials
by dehydrochlorinating solutions of the polymer. He applied hot dichlorobenzene solutions in
coating insulation. Returning to his main problem of lining tanks, Semon solved the adhesion
problem by attaching metal gauze to the interior of a tank and coating the gauze with a paste of
PVC polymer and plasticizer such as tricresyl phosphate, fusing the mix to give a bonded,
chemically resistant PVC lining. This was the origin of plasticized PVC and of the plastisol
process. B. F. Goodrich contracted for Carbide and Carbon to produce high-molecular-weight
PVC after finding that Carbide and Carbon's vinyl chloride/vinyl acetate copolymers could not
make the flexible, rubberlike products that were desired for the Goodrich product development
effort. This contract is apparently the origin of the larger bulk polymer plant (10,000,000
pounds) at South Charleston, which supplied resin for Goodrich's trademarked plasticized vinyl
sheeting, Koroseal, and General Electric's wire insulation, Flamenol. The success of these
products put the Goodrich company into resin production. Product development work followed
Semon's work both in the United States and in Europe.
During World War II all resin production was commandeered for the military. The modern
history of PVC began immediately after the war. Five U.S. companies were producing PVC by
1950; the number of producers increased to 20 in the mid-1960s. During this period, much work
was carried out on polymerization processes, compounding techniques, additive development,
and processing techniques [4]. The extent of the growth of PVC into an international multibillionpound business is described in the next section. Figures 1 and 2 show the U.S. production of
vinyl polymers and their unit sales value during the period 19551978 [4]. The U.S. marketplace
has changed over the past years and the end-use pattern for PVC has gone from mainly
flexible products such as flooring, wall coverings, upholstery, shower curtains, and the like in the
early 1970s to mainly rigid products such as pipe, siding, and profiles in the early 1980s. The
ratios of flexible to rigid PVC have changed from 70:30 in the 1970s to under 40:60 in the
1980s [5].
A graphic presentation of the history of polyvinyl chloride was made available in an article on
vinyl additives [6] and is presented as Appendix I to give perspective. The historical growth

pattern of various PVC markets is shown in Figure 3 together with projected 1985 markets [7].

Page 4

FIGURE 1 U.S. production of PVC and copolymer, 19551983.

II.CURRENT COMMERCIAL SIGNIFICANCE

A.Production and Competition
The total PVC resin production in 1983 has been estimated to be 30.0 billion pounds worldwide.
This amount represents about 80% of the estimated capacity of 37.5 billion pounds (Table 1).
International consumption in 1983 is estimated at 25.5 billion pounds [10]. These figures
indicate that sufficient capacity exists until 1990. During this upcoming 5-year period, there will
be continuing rationalization. of the industry as has been taking place over the past 5 years.
Typical area capacity figures are shown in Table 1 covering Western Europe, the western
hemisphere, the Far East, the eastern bloc, and the rest of the world. Table 2 lists 1983
capacities for the major producing countries by area.
The cyclical nature of the PVC industry has been demonstrated by the turmoil and changes in
the major producing countries. Table 3 illustrates the down period (19811982) in which many
rationalization moves took place. Descriptions of some of the rationalization moves and
ownership changes are given in subsequent paragraphs.

Page 5

FIGURE 2 U.S. unit sales price of PVC, 19551983 (averaged).

It is of interest to compare the major thermoplastics and to place polyvinyl chloride and its
copolymers in proper economic status. The major thermoplastics are: polyvinyl chloride, lowdensity and linear low-density polyethylene, high-density polyethylene, polystyrene, and
polypropylene. In Table 4 is shown the U.S. annual production of these materials during the
period 19731983 [11]. The Western European thermoplastic resin consumption for a similar
period is displayed in Table 5. Upon examination of these tables, it is clear that PVC is either
first or second in these areas and is second in volume on an international basis.
Having looked at the PVC consumption in Western Europe, it is well to examine the per capita
consumption figures for this area since it is the highest

FIGURE 3 PVC demand by market (billion pounds). (From Ref. 7.)

Page 6
Western Europe
Western Hemisphere
Far East
Eastern Bloc
Rest of the world

12.0
11.0
8.0
6.0
0.5
37.5

Western Europe
Belgium
France
West Germany
Great Britain
Italy
Netherlands
Spain
Others
Western Hemisphere
United States
Canada
Mexico
Brazil
Others
Far East
Japan
China
South Korea
Taiwan
Others
Eastern Bloc
East Germany
Poland
USSR
Yugoslavia
Others

12.0
0.8
2.1
3.0
1.0
2.4
0.75
0.8
1.15
11.0
8.4
0.8
0.6
0.7
0.5
8.0
4.2
0.95
0.7
1.6
0.55
6.0
1.2
1.0
2.1
0.8
0.9

Page 7

Taiwan
Canada
Japan
United States
France
Italy
West Germany

Polyvinyl chloride
LDPE
HDPE
Polystyrene
Polypropylene

1.0

3.15

1.60
1.48
2.10

4.56
5.80
2.64
5.02
2.16

5.72
7.11
4.20
5.08
3.07

1.0
0.49
2.49
5.71
1.56
1.28
2.02

5.71
7.69
4.70
5.47
3.96

1.01
0.43
2.68
5.33
1.73
1.29
1.90

5.33
7.50
4.93
4.66
3.48

1.15
0.54
3.13
6.07
1.80
1.46
2.40

6.07
8.04
5.69
5.55
4.43

2.9
3.3
8.0
1.0
7.4

Source: Ref. 11.

Polyvinyl chloride
LDPE
HDPE
Polystyrene
Polypropylene
Source: Ref. 10.

7.43
6.68
2.23
2.62
1.23

8.68
8.82
3.26
2.82
2.78

7.16
7.49
3.00
2.40
2.95

7.77
7.71
3.04
2.42
3.06

Page 8

West Germany
France
Benelux
Italy
Great Britain
Scandinavia

11.8
7.4
7.5
6.2
5.7
7.9

17.5
12.9
11.8
10.0
6.5
10.7

15.0
12.2
10.8
9.6
7.0
12.5

Source: Ref. 9.
in the world. Per capita consumption figures for major Western European countries are given in
Table 6. Except for Scandinavia, consumption would appear to be stabilizing.
Over the period 19791984, U.S. polyvinyl chloride annual capacity increased from 6.66 billion
pounds to 8.4 billion pounds, with a concurrent utilization percentage decrease from 92% to
79%. During these years, Formosa Plastics entered the industry in the United States as other
foreign companies have done or have tried to do. Shintech is owned by Shinetsu of Japan;
Chemische Werke Hls of West Germany has made a number of attempts to acquire an
existing U.S. producing facility. The following companies left the industry in this period: Diamond
Shamrock Corp., Firestone Tire&Rubber, Great American Chemical, International Materials,
and Stauffer Chemical [12]. The current producers of PVC in the United States are listed in
Table 7, where rank and capacity changes are shown for the period 19791984. Some of the
changes merely represent new ownership, whereas others actually indicate new or expanded
production facilities. In any event, the net result is sufficient capacity to 1990, as stated earlier.
This capacity may change hands as a slower-paced rationalization effort takes place.
West European rationalization and concentration moves have lowered the number of PVC
producers from 24 to 17, compared to the U.S. drop from 20 to 16. In Italy and France, the
formation of Eni-Chimica and Chloe provides governmental control of major PVC manufacturing.
Eni-Chimica's capacity of 860,000 metric tons makes it the largest producer in Western Europe
and the world. Second place in Europe goes to the Belgian company, Solvay, with a capacity of
790,000 metric tons per year. Chloe of France is third in Europe, with about 600, 000 metric
tons per year. The next largest producer is ICI of Great Britain, with about 570,000 metric tons
per year of capacity. These four are followed by Royal Dutch/Shell, LVM (a joint venture of
French and Benelux interests of SAV and DSM), Chemische Werke Hls, and Wacker. Like the
industry changes in the United States, expansions and buyouts or exchanges have modified the
cast of industry participants. ICI has 115,000 tons coming on stream in Wilhelmshaven and has
taken over Lonza's PVC operation (Swiss), Norsk Hydro purchased both Vinatex and BIP Vinyls
of Great Britain. A portfolio exchange between ICI and BP further increased ICI's vinyl position
[9]. In Japan, the government imposed a 24% reduction in the allowed production of PVC by all
the producers in 1983. This international struggle for profitability is further complicated by the
barter payment

Page 9

1
2
3
4
5
6

1
2
8
5

7
8
9
10
11
12
12
14
15
16

9
13
16
11
3
12
16
10
18
20

Goodrich
Tenneco
Georgia Pacific
Borden
Formosa
Du-Pont Conoco
(Vista)
Shintech
Occidental
Air roducts
Certainteed
Diamond Shamrock
General Tire
Union Carbide
Goodyear
Pantasote
Keysor

1720
965
850
820
810
710

+545
+270
+500
+295
+810
+150

680
650
400
190
130
125
125
120
55
50

+350
+470
+250
0
-470
-60
-25
-150
-85
0

Source: SRI International.

with PVC resin by the East Bloc for purchased technology and by the contract sale of PVC
made in hydrocarbon-rich areas such as Saudia Arabia. in spite of the obvious immediate
oversupply, expansions such as Shintech's in the United States to 1 billion pounds continue to
take place, possibly encouraged by a technology advantage and the firming of resin prices. In
another resource-rich area, Canada, affiliates of the U.S. companies Goodrich, Exxon, and
Diamond Shamrock are generating resin supply which might compete with U.S. producers, in
contrast to the Saudi competition with Western European resin suppliers. Mexico has hopes of
selling resin in the U.S. market. Poland has a new 200,000-ton plant coming on stream and will
exert export pressure to Western Europe.
B.Markets
Having examined the competitive scene, it would do well to examine the markets that all these
existing and would-be producers are competing for. By referring to Tables 8 to 10, the breadth
of the markets affected by the PVC industry can be studied. The major use areas are in the
building and packaging in

Page 10

Building and Construction

Flooring
Paneling
Pool-pond liners
Roof membranes
Other building
Transportation Auto
upholstery/trim
Other upholstery trim
Auto tops
Packaging: sheet
Electrical: tapes
Consumer and institutional
Sporting/recreation
Toys
Baby pants
Footwear

Building and
construction
Building and
Pipe and conduit
7381construction
pressure
1114 Pipe fittings
48 55 Water
1514 Other building
3 5 Gas
1214Transportation: bumper
Irrigation
5 7
1 2 parts
Other
Electrical/electronics
Drain/waste/vent
3032
Plugs, connectors, etc. 24 30 Conduit
7 7 Appliances, bus.
Sewer/drain
22 27
4 5 machines
Other
3536Consumer and
Siding and
5 5 institutional
accessories
Footwear
19 18 Window profiles All109
Hospital and health
vinyl-windows
7 9
1414 care
Composite windows
3 2 Other injection
6 9 Mobile home skirt
1414Total injection molding 134160 Gutters/down-spouts
Foam moldings
Weatherstripping

383402
8 9
12180
21 23
129193
185193
180227
6 9
109159
17 20
36
10
5
17
17

45
11
7
20
20

Page 11

Handbags/cases
Luggage
Bookbinding
Tablecloths, mats
Hospital and health care
Credit cards
Decorative film (Adh.-back)
Stationery, novelties
Tapes, labels, etc.
Furniture/furnishings
Upholstery
Shower curtains
Window shades/blinds/awnings
Waterbed sheet
Wallcovering
Other calendering
Total calendering

11
9
3
7
10
10
5
3
8
33
4
5
4
14
5
365

Lighting
Transportation
9
Vehicle floor mats
9
Bumper strips
2
Packaging
7
Film
11
Sheet
9
Electrical:
5
wire and cable
2
Consumer and institutional
9
Garden hose
Medical tubing
34
Blood/solution bags
5
Stationery/novelties
5
Appliances
5
Other extrusion
14
Total extrusion
7
Blow molding: Bottles
382
Compression molding:
Sound records

9
5

9
5

125
15

134
16

159

187

20
19
25
5
14
57
1705
75

20
23
30
7
18
52
1928
95

36

32

Page 12

Building: flooring
Transportation
1518 Transportation
Packaging:
Auto upholstery/trim
closures
1516 Other upholstery/trim
Consumer and institutional
Anticorrosion coatings
Toys
3 5 Consumer and institutional
Sporting/recreation
7 7 Apparel/outerwear
Footwear
5 5 Luggage
Handles, grips
5 5 Tablecloths, mats
Appliances
5 7 Hospital/health care
Industrial: traffic cones
7 7 Furniture/furnishings
Adhesive, etc.
Upholstery
Adhesives
6 9 Window shades/
Sealants
5 7 blinds/awnings
Miscellaneous
5 5 Wallcoverings
Other dispersion
3 2 Carpet backing
Total dispersion molding 8193 Other
Total dispersion coating
Source: Ref. 8.

45 68
9
2
4

10
2
4

5
3
3
2

7
4
4
3

Adhesive
sealants
Export
Grand total

5 5
5 6
6 8
137141

20

27

243 155
27963013

Page 13

Rigid PVC
Pipes and fittings
Profiles
Film and sheet
Blow moldings
Records
Other
Plasticized PVC
Film and sheet
Synthetic leather
Floor coverings
Tube and fittings
Wire and cable insulation
Other
Total consumption

2001
750
478
340
243
65
125
1251
305
160
170
ca. 130
318
168
3252

2152.5
892.5
420.0
350.0
262.5
70.0
157.5
1347.5
315.0
157.5
175.0
140.0
350.0
210.0
3500.0

dustries. Both of these are expected to grow at above economy growth rates, with particular
growth in the pipe, profile, and container segments. Growth in the electrical /electronics area
will depend on population growth and continued electrification of our living style. A need to
improve insulation formulations is indicated in order to maintain market share in the face of
potential fluorocarbon competition. Very little growth is expected, and possible major declines,
in the use of vinyls in furniture and transportation areas. Again, it is well to remember that the
basic economics of PVC are very favorable for continued growth. Growth rates have been
projected for various markets as follows: pipe and fittings, 10% siding, 20%; and packaging
and containers, 10%. All of these estimates are per annum. The estimated growth rates for the
other use areas tend to be 4% per year. Some ambitious estimates have put U.S. demand at
10,000,000,000 pounds by 1990.
C.Economic Impact of the Overall PVC Industry
The manufacture of vinyl chloride monomer, polyvinyl chloride, and products made therefrom
provide employment for about 2 million persons in the United States in production, consumer,
and realted industries. Direct chemical production accounts for about 1500, workers in
monomer production, and about 5000 people are involved with polymer production. About 3/4
million people with various talents are responsible for converting polymer into finished plastics.
Monomer production takes place primarily on the Gulf coast; polymer plants are located in
various parts of the country, mainly in the Gulf coast area, the Middle Atlantic, and the midwest.
Fabrication plants tend to concentrate in the eastern United States; some west coast
fabricators are active in pipe and film/sheeting operations.
The favorable energy requirement comparison of PVC with other materials will continue to
support the interest of both producers and users. The pro-

Page 14

Appliances
Building
Flooring
Lighting fixtures
Panels and siding
Pipes, fittings, conduit
Profile (windows, spouts, etc.)
Vapor barriers (liners)
Wall coverings
Electrical /electronics
Furniture
Housewares
Packaging
Closures
Coatings
Containers and lids Blow-molded
Thermoformed
Film
Toys
Transportation

41

52

141
8
119
1081
92
27
19
159
31
21

141
9
170
1191
112
28
19
187
43
23

15
13
75
35
125
17
62

16
14
95
36
134
19
70

Source: Compiled from data in Ref. 8.

duction of aluminum requires more than three times the energy necessary to make the PVC
that makes the siding that cmpetes with aluminum. As hydrocarbon feedstocks increase in
price, PVC will be less affected, by reason of the high inorganic content.
D.Environmental and Safety Aspects
Vinyl chloride monomer is a nonproblem today as a result of the excellent effort on the part of
the industry to remove or lower substantially the residual level of the monomer from resin.
Some resin producers have decided to remove the hazard statement from bags, bins, trucks,
and railcars. Excellent analytical devices are available to detect in an unambiguous manner very
low levels of vinyl chloride monomer, in all media. In the area of polymer reactors and
processing equipment, monomer monitors are standard. All vestiges of monomer are scrubbed,
recovered, and recycled. The combustibility of PVC and plasticized PVC has been studied and
various myths have been shown to

Page 15
be inaccurate. Phosgene is not a factor in combustion gases. Hydrogen chloride is an irritant in
the combustion gas and has lower toxicity than carbon monoxide and acrolein, which are
present at higher levels in fires. With an estimated 4% maximum concentration of PVC waste in
municipal garbage, there are no hazards to be experienced in a properly designed garbage
combustion plant. The untoward attacks on PVC are being met by the considered actions of the
Vinyl Institute, an industry group charged with spreading the truth about monomer and polymer.
The solid waste problems associated with waste PVC can also be handled by recycling the
thermoplastic. France does an excellent job of recycling by grinding, shredding, and fusing
waste bottles and containers to make useful product.

III.FUTURE TRENDS
A.The PVC Industry
Throughout the world, there will be a slower but continuing consolidation and rationalization
movement. The commodity character of the resin production effort will be disturbed by the
activities of resource-rich and market-poor regions. New applications are projected and the
convertors are looking forward to better equipment, resins, and additives in order to meet these
new needs with innovation.
There will tend to generate in coming years a greater separation of resin, compounding, and
fabrication activities. This separation is in contrast with the great efforts to integrate forward
and backward that have dominated the industry in past years. Acquisitions and mergers will
solve some problems that companies face while seeking profitable operations. Analysis of the
world scene indicates that larger and fewer resin producers will dominate the PVC industry. As
stated previously, the major markets are construction, including pipe, fittings, and siding;
packaging; and containers. Blending, modifying, and reinforcing of base resin will provide new
product opportunities. Difficult though they may be, government regulations can be met with
fact and proper action on the part of all phases of the industry. The rules are a fact of life and
their cost is part of the cost of doing business.
B.Monomer
The key to the favorable economics of vinyl chloride is the cost of chlorine. There is a pressure
from caustic /chlroine producers to raise the chlorine price to make up for a lagging market in
caustic soda. Better use of electrolysis techniques may alleviate this pressure. Although the
hydrocarbon source for vinyl chloride has been ethylene in recent times, some have speculated
that acetylene derived from coal may again become the major hydrocarbon used. With the
great concern about transport accidents and the increasing physical distribution costs,
monomer plants will tend to be relatively close to polymer plants. Pipeline transport between
supplier and user will become more prevalent. In the meantime, nameplate capacity for vinyl
chloride will be more than sufficient until 1990. The retrofitted safety features in monomer
plants have generated more automation in larger plants. Monomer prices have been pressured
from 14 cents a pound to over 20 in the 19821983 period. Stabilization of this price will come
with process improvments and a balanced caustic /chlorine industry.

Page 16
C.Polymerization
The most significant development in PVC is the successful use of large reactors in long
production runs. Technology promoted by B. F. Goodrich and Shinetsu, involving surface
treatment of the reactor walls to prevent polymer buildup, requires less frequent cleanouts.
These over- 15, 000-gallon reactors provide economics in extended production runs.
Improvements in the polymerization process, involving suspending agents or coalescence
regulators, secondary suspending agents, improved agitation, and better control of the degree
of conversion, have made it possible to produce resins with more uniform porosity, higher bulk
density, and lower gel count, both in suspension and mass polymerization [13]. It is certain that
research into morphology, rheology, and other scientific disciplines for PVC and vinyl chloride
copolymers will lead to new product potentials. Work continues on the development of
continuous polymerization processes and novel polymerization techniques. New copolymers and
grafts of vinyl chloride are being investigated. It is important to recognize that a single ideal vinyl
polymer (PVC) for universal usage is not achievable; all resins are produced with compromised
properties. However, the tremendous variety of resin types permits the fabricator to carry out a
multitide of processes and to make an endless array of useful products.
D.Compounding and Fabrication
To use PVC and its copolymers, blends, and alloys properly, it is necessary to compound the
product as a dry blend or as a fused pellet or mass for calendering, molding, and extrusion
operations. Fortunately, a variety of mixing and fusing machines are available to satisfy the
fancy of the plastics engineer confronted with developing or manufacturing a product.
Continuous PVC compounding on the extruder with dedicated measuring and metering systems
has attracted the attention of equipment manufacturers, custom cornpounders, and fabricators
looking to improve quality and lower costs [14]. Using a new mixing turbine, Buss-Condux has
introduced a continuous dryblend feed to a melt compounder. This approach is economical of
space, as shown in Figure 4. Another example of innovative PVC compounding is the use of the
Gelimat, originally designed and manufactured by Draiswerke GmbH and modified by
Synergistics' members, Carlew and Cary Chemicals [16] (Figure 5). This machine permits very
short time flux cycles for flexible PVC. Full compounding systems are offered by all the extruder
manufacturers, each offering modifications suited to their comprehension of industry needs. Out
of the Baker Perkins organization has come an interesting centrifugal pelletizer which is energy
and space efficient. This pelletizer takes an atmospheric-pressure molten feed and gives
designed size and shape by controlling die speed and distance of the cutter blade from the die
rim [17]. The B. F. Goodrich Chemical Group has developed technology incorporating a new
group of rigid PVC compounds for profile extrusion coupled with a unique extruder screw,
designed to get the output increased illustrated in Figure 6 [18]. Regular reviews of the
equipment scene appear in the industry press and should be followed to permit one to keep up
to date. For compounding and processing, computer controls are bringing better quality and
lower costs to the fabricator. Newer independent compounding facilities are being established
to pick up the anticipated markets that are developing in the fastmoving pipe, profile, and
container areas (Table 11) [19]. Even as these independent compounders proliferate, more
fabricators are using compounding

Page 17

FIGURE 4 Continuous dry-blend compounder. One of the many advantages of Buss-Condux's

new mixing turbine is the small amount of space it requires. In the schematic at left, the device
is contrasted to a traditional hot/cold batch preblending equipment configuration. The schematic
cross section above illustrates the device's novel rotor/stator design. While solid components
flow through the unit, liquids are continuously metered directly onto the rotor blades. Individual
injection nozzle adjustments permit precise metering for consistent product quality. (From Ref.
15.)

FIGURE 5 The Gelimat mixing system, an ultrahigh-speed mixing process. (From Ref. 7.)

Page 18

FIGURE 6 Output rate vs. melt temperature for a typical low-temperature Geon rigid cube
profile compound.
systems of the types mentioned previously in an attempt to realize cost savings and take on
quality responsibility. This is the continuing dilemma of makeor-buy that is part of the industrial
scene.
Multiscrew extruders have had a profound effect on pipe production technology. Compounding
twin-screw extruders are used in series with singlescrew extruders. Melt pumps are being used
in many cases to increase production rates. Stretch blow molding of bottles and containers has
come on the scene rapidly because better PVC bottles are made. The oriented product has
higher tensile strength and impact strength, allowing thinner walls and giving better clarity.
Extruded foamed rigid PVC is under development as a fire-resistant replacement for wood.
Other foam applications are contemplated in plastic design where bulk without mass is needed.
Injection molding of rigid PVC dry blend has been proposed by a major U.S. resin supplier.
Structures and composites are being made with rigid PVC for light panels and corrugated
pipes. Cross-linking of plasticized PVC gives outstanding properties. Radiation cross-linking of
contained polyfunctional methacrylate monomer gives greatly improved properties, such as heat
distortion, abrasion resistance, cut-through resistance, and solvent resistance [20].

Page 19
a Supplier Plant locationsb Capacity (1000 metric tons)b,c Compound typed Specialties and
applications Remarks Abbey Plastics Hudson, MA 6 FX, RG EC, CS, EL, ME New: translucent
EC compound. Abcorn Mishawaka, IN NA' FX, RG CS, WC, EL, XP, IP, HF Rigids are new.
Alpha Chem. Newark, NJ 27 FX, RG TC, WC, EL, ME, AU, BO, CM, GP Capacity has been
doubled, with new equipment for rigids. B. F. Goodrich Avon Lake, OH 160 E FX, RG DB, TC,
EL, AU, CM, BO, FI, SH, PI, XP, IP, WC, ME, HE, CS, GP Makes resin. Proprietary and
custom compounds. New: HF, low-smoke, and more processable CPVC compounds,
respectively. Henry, IL Long Beach, CA Louisville, KY Fedricktown, NJ Deer Park, TX
Shawinigan, Que. Kitchener, Ont. B. P. Polymers Mansfield, MA 55 FX WC, FW, CY, AU, CM,
GP Formerly Blane Div. of Reichhold Corp. low-smoke and rigid specialties. Visalia, CA
Canada Colors Colborne, Ont. 5 FX CS, TC, WC, ME, GP Also makes color concentrates.
Canadian General Tower Oakville, Ont. 9 FX, RG AU, XP, IP, FC, EL, GP To add 3000 tons
and offer window compounds in 1984. Carlew, subs Synergistics St. Remi, Que. 90 FX, RG
WC, XP, IP, CM, GP Broadening range of window compounds. Orangeville, Ont. Calgary, Alta.

Page 20
b Capacity (1000 metric tons)b,c Compound typed Specialties and applications Remarks Cary,
subs Synergistics Farmingdale, NJ 55 FX WC, FW, GM, GP Farmingdale on stream early
1984. Edison capacity to be consolidated there. Colorite Div., Dart-Kraft Ridgefield, NJ Sparks,
NV 45 FX, RG ME, EL, CM, GP Some captive use. Added 9000 tons this year. New: radiationsterilizable compounds. Ethyl Delaware City, DE 90 RG TC, GA, BO, FI, SH, XP, HF, CM
Makes resin. Proprietary compounds. Georgia-Pacific has agreed to buy business. Tiptonville,
TN Gallman, MS Flex Products Baltic, OH 14 FX, RG WC, AU, FC, GP Some captive use.
Recently modernized, expanded. Franklin Plastics Kearny, NJ 45 FX CS, FW, ME, EL, GP New
compounding unit permits fast turnaround on samples. G. S. Plastics Cleveland, OH 7 FX CS,
AU, GP Some captive use. Gary Chemical Leominster, MA 34 FX CS, WC, FW, ME, AU
Recently added 16,000 tons. Another 10,000 likely in 1984. Higher Molding Jamestown, NC 9
FX, RG CS, TC, XP, CM Recently added 2000 tons. Hoffman Plastic Paramount, CA 20 FX,
RG TC, ME, WC, BO, HF, XP, IP, CM, GP Expansion of 4000 tons, due early 1984, enables
firm to supply rigids (BO, HF, XP, IP). Intamix Conroe, TX 90 RG DB, PI, XP, SH All customs,
no pellets. May add 45,000 tons for TC powder.

Page 21
Yardville, NJ
FX,WC, FW, FC, Yardville: is new. Another 800 tons there
19
Brooklyn, NY
RG AU, GP
by late 1984.
Saugus, CA
Makes resin. Tolls BO and CM compounds
Delaware
FX,RE, BO, HF,
Keysor-Century
72
for Formosa Plastics. Custom and
City,, DE
RG SH, CM
proprietary.
Ajax, Ont.
Pitman, NJ
FX,
Recent expansion at Pitman was for entry
Lenahan Assoc. Murfreesboro,16
RE, BO
RG
into BO.
TN
CS, FW, WC, Rebuilding (after 1981 fire) and expansion
Lynn Plastics
Lynn, MA
45FX
EL, AU, CM adding 34,000 tons by early 1984.
City of
Moore Chemical
12FX WC, CM
Custom compounds.
Industry, CA
Aberdeen,
WC, EL, AU,
North American MS
35FX
Jackson startup Jan. 1984.
CY, CM, GP
Jackson, MS
Novatec, subs. Eatontown,
TC, BO, SH, Some captive use. New firm with recently
11RG
PVC Container NJ
FI, GP
expanded plant.
30FX,GA, WC, ME,
Pantasote
Passaic, NJ
Makes resin. Some captive use.
E RG XP, FI, CM
Kendallville,
PVC Compounds
IN
11FX WC, EL
Some captive use.
Div., Lyall Electric
El Paso, TX
TC, GA, BO,
Makes resin. Phased expansion to add
Occidental
Burlington, NJ 45RG SH, CM, HG,
27,000 tons by 1986.
XP
Valleyfield,
Rochevert, subs Que.
FX,
Supplies BO test quantities to affilliate
35
TC, FW, BO
Synergistics
Lindsay, Ont.
RG
Cary Chemicals.
Hudson, NH
Kalflex Div., Kalex

Page 22
b Capacity (1000 metric tons)b,c Compound typed Specialties and applications Remarks
Roscom Trenton, NJ 15 FX, RG CS, TC, ME, AU, BO, FI, SH, XP, CM, GP Recently added
9000 tons. At work on window compounds. Rutland Plastics Pineville, NC 18 E FX, RG EC,
CS, DB, YC, EL, XP, IP, CM, GP To add 5000 tons in 1984. At work on PVC/TPE alloys and
EMI shielding compounds. Schulman, A. Akron, OH 36 FX, RG EC, GA, CS, AU, ME, WC, XP,
IP, CM Recently added 11,000 tons at Akron. Proprietary compounds. New: XP compounds
based on PVC/acrylic graft copolymer. Bellevue, OH Orange, TX St. Thomas, Ont. Stanchem
E. Berlin, CT g FX, RG FW, CM Two compounds: FW; specialty FX items like printing plates.
Sylvania Chem. Denver, PA 19 FX, RG DB, WC, AU, EL, PI, XP, IP, GP Expansion for rigids
nearly complete. Tecknor Apex Pawtucket, RI 45 E FX, RG TC, WC, ME, AU, CY, BO, CM,
GP New compounding unit for clears makes possible entry into BO markets. Brownsville, TN
Industry, CAh

Page 23
TC, GA, WC, FW, RE, XP,
Burlington, FX,
Tenneco
45
IP, PI, SH, FI, BO, CM, HF,
NJ
RG
GP
Grand
FX,
ViChem Rapids, 14
CS, AU, IP, XP,
RG
MI
Aberdeen, FX,
Vista
60
DB, PI, XP, WC, SH, GP
MS
RG

Makes resin. Developing food-grades

BO.
Has added 5000 tons. At work on
nonmigrating compound. Will enter rigids
in 1984.
Makes resin, plasticizer.

aList contains all firms cited as merchant-market suppliers by directories and industry sources.
No doubt it is incomplete. Excluded because of space are suppliers of dispersion, calendering,
casting, coating, laminating, or solution materials. Source: Telephone interviews with sources at
all firms listed.
bVinyl compounding operations only.
cBallpark estimates of nameplate annual capacity as of early 1984. Unless designated E (for
Modern Plastics estimate), figures come from suppliers themselves. Capacity varies with
production schedule and type of product.
dAbbreviations: FX, flexible or semirigid (containing plasticizer).; and RG, rigid.
eAbbreviations: AU, automotive; BO, bottles; CM, custom molding /extrusion; CS, color
specialties; CY, extrusion-coated yarn; DB, dryblend or powderblend; EC, electrically
conductive; EL, electrical/electronics (other than wire and cable); FC, fencecoating (extrusioncoated wire); FI, film; FW, footwear; GA, tough compounds based on graft copolymers or
alloys; GP, generalpurpose molding/extrusion; HF, high-flow injection molding; IP, interior profile;
ME, medical items; PI, pipe and fittings; RE, phonograph records; SH, sheet, TC, transparent
or clear compounds; WC, wire and cable; and XP, exterior profile.
fNot available.
gCapacity is less than 1000 tons.
hPlant operated by The Maclin Co., a subsidiary.
Source: Reprinted from Modern Plastics (Nov. 1983), copyright 1983 by McGraw-Hill, Inc. with
all rights reserved.

Page 24

IV.SUMMARY
The PVC Industry and its attendant support facilities are expected to continue to grow as a
mature industry. There are problems and challenges that call for the talented efforts of many
trained people to solve and meet. In the earlier sections of this chapter, a brush has been used
to paint a picture of the immensity and complexity of the PVC industry. In subsequent chapters
and volumes, experts will present expanded views of mentioned subjects and other pertinent
items. It is hoped that the perspective presented here excites the imagination of the reader to
contribute to the solution of problems and the meeting of challenges.

ACKNOWLEDGMENTS
The author wishes to thank many people, worldwide, who have contributed to my knowledge
and experience in the PVC industry. Some of these individuals and organizations are
acknowledged in the following references. Others are forever in my memory as gracious
contributors to my meaningful professional career.

REFERENCES
1. M. Kaufman, The Chemistry and Industrial Production of Polyvinyl Chloride: The History of
PVC, Gordon and Breach, New York, 1969.
2. M. Kaufman, Trans. J. Plast. Inst., 365371 (1967).
3. Personal communication, Roland Glenn, Combustion Processes, Inc., N. Y.
4. W. L. Semon and G. A. Stahl, J. Macromol. Sci. Chem., 15 (1981).
5. R. E. Swantkowski, Prep. SPI/SPE Meet., 255258 (1984).
6. S. Stinson, Chem. Eng. News, 38 (June 18, 1984).
7. E. A. Vogelfanger, Chem. Mark. Res. Assoc. Pap., May 1982.
8. Mod. Plast., 62(1), 6669 (Jan. 1985).
9. Kunststoffe, 73(9), 463465 (1983).
10. Author's estimates based on review of data from Plastics Technology, Modern Plastics,
Kunststoffe, publications of SPI, SPE, and SRI International, as well as various private
communications.
11. Chem. Eng. News, June 11, 1984.
12. A. D. Abshire, Chem. Mark. Res. Assoc. Pap., Feb. 1984.

13. Y. Fagnoni, Euretec-Society of Plastics Engineers, May 1982.

14. M. Weber, Ind. Prod. Eng., 3, 8693 (1980).
15. Plast. Technol., July 1983.
16. Plast. Compound., Sept. /Oct. 1982.
17. G-Force Centrifugal Pelletiser, Baker Perkins, Inc., Saginaw, Mich.
18. Hi-Pro, Super Cubes,, B. F. Goodrich Chemical Group, Cleveland, Ohio.
19. Mod. Plast., 5255 (Nov. 1983).
20. R. C. Stephenson, J. Vinyl Technol., 5(1), 1318 (1983).

Page 25

APPENDIX: A HISTORY OF POLYVINYL CHLORIDE*

1835. Justus von Liebig at the University of Giessen discovers that reaction of the oil of the
Dutch chemists (ethylene dichloride) with alcholic potash produces a chlorine-containing gas.
Liebig assigns conformation of the work to his student, Victor Regnault, and lets Regnault
publish as sole author. As a result, Liebig is forgotten as the discoverer of vinyl chloride.
1872. E. Baumann finds that exposure of vinyl chloride to sunlight in a sealed tube produces a
while solid that resists attack by potassium hydroxide or water and melts with decomposition
above 130C.
1912. Chemische Fabrik Griesheim-Electron gives chemist Fritz Klatte the task of finding uses
for acetylene, and he develops commerical synthesis of vinyl chloride by addition of hydrogen
chloride.
1915. Klatte finds that organic peroxides initiate polymerization to polyvinyl chloride.
1926. After years of effort to develop rigid PVC as a replacement for celluloid, German
workers have failed to build machinery that will process the resin and have not overcome the
problem of thermal instability. GriesheimElectron stops paying the fees that maintain its
patients. Meanwhile in the U.S., Union Carbide, looking for uses for the ethylene dichloride byproduct of chlorohydrin production, converts it to vinyl chloride with sodium hydroxide. PVC also
interest Waldo L. Semon at B. F. Goodrich, who is looking for an adhesive to bond rubber to
metal. Semon observes that boiling PVC in such liquids as tricresyl phosphate or dibutyl
phthalate renders it highly elastic, thus discovering plasticizers. In addition to molding parts from
plasticized PVC, Semon goes on to coat wire, bloth, and metal parts with solutions of PVC in
plasticizer or plasticized PVC in organic solvents, thereby inventing plastisol and organosol
processing techniques. He uses white lead and sodium silicate as heat stabilizers.
1928. Union Carbide and Du Pont each find that copolymers of vinyl chloride with vinyl acetate
have improved processibility.
1934. Frazier Groff of Union Carbide discovers that alkaline earth soaps stabilize PVC to heat.
1936. Union Carbide's PVC homopolymer plant comes on stream at 4 million lb per year.
Chemists there find that dialkyltin soaps are also heat stabilizers. At B. F. Goodrich, T. L.
Gresham tests thousands of compounds before setting on di-2-ethylhexyl phthalate as the best
plasticizer.
1941. During World War II, PVC is commandeered for military use, with the Navy in particular
using large amounts of phosphate-plasticized, fireretardant wire insulation developed by
General Electric.
1947. One or more workers now unknown hit upon synergistic combinations of barium,

cadmium, calcium, and zinc soaps as heat stabilizers.

* Source: Ref. 6.

Page 26
1952. Elliott Weinberg at Metal and Thermit Corp. invents the dibutyltin dimercaptide with 2ethylhexyl thioglycolate as a heat stabilizer for rigid PVC, and at the same time invents the
antimony trimercaptide low-cost stabilizer for pipe, which will be commercialized 25 years later
as tin prices skyrocket.
1960. U.S. production of PVC tops 1 billion lb.
1970. Pier Luigi Viola at Regina Elena Institute for Cancer Research in Rome finds that
exposure to very high levels of vinyl chloride vapor causes tumors in rats. U.S. production of
PVC tops 4 billion lb.
1973. Cesare Maltoni, working for a group of European firms at the Institute of Oncology in
Bologna, Italy, finds that inhalation of vinyl chloride vapors causes rare angiosarcoma liver
cancer in rats. The Manufacturing Chemists Association is allowed to see his work, but only
under a confidentiality agreement. Forbidden to communicate Maltoni's findings, MCA
commissions its own studies of rats and worker exposures. Late in the year physician John L.
Creech links deaths of two B.F. Goodrich workers from angiosarcoma to vinyl chloride
exposure in Louisville, Ky. Eventually, angiosarcoma deaths of three dozen people worldwide
are tied to vinyl chloride exposure.
1974. Goodrich tells the National Institute of Occupational Safety and Health of the cancer link.
The Occupational Safety and Health Administration reduces permissible vinyl chloride exposure
to 50 ppm from 500 ppm and proposes a zero detectable limit. Industry spokesmen say that
will force them to shut down. OSHA agrees to a phased reduction to 1 ppm. PVC makers
announce that they can meet this and alter their plants to do so. Later, activists tar industry for
saying first that the OSHA limit would shut them down and then agreeing that they can meet it.
1980. The National Cancer Institute finds evidence that di-2-ethylhexyl phthalate plasticizer
causes cancer in rats and mice.
1982. Stung by inroads of PVC into pipe, metal-pipe makers counterattack with a campaign to
warn public authorities of toxicity of gases emitted by PVC in fires.
1984. A recovering U.S. economy carries PVC use to a record 6.3 billion lb.

Page 27

2 Vinyl Chloride Production

LYLE F. ALBRIGHT
Purdue University Lafayette, Indiana

I. INTRODUCTION

27

II. MAJOR PROCESS FOR PRODUCTION OF VINYL CHLORIDE

29

A. Chlorination of Ethylene

30

B. Oxychlorination of Ethylene

31

C. Pyrolysis of 1, 2-Dichloroethane

34

III. HYDROCHLORINATION OF ACETYLENE

37

IV. TECHNOLOGY OF POSSIBLE IMPORTANCE

41

V. ECONOMIC AND ENVIRONMENTAL CONSIDERATIONS

42

VI. STORAGE AND HANDLING OF VINYL CHLORIDE

43

VII. SUMMARY

43

REFERENCES

44

I. INTRODUCTION
The production capacity of vinyl chloride (often designated as VCM, for vinyl chloride monomer)
has increased substantially in the United States in the last 10 years as well as in the rest of the
world [8, 10, 12, 22]. In 1970 in the United States, annual total capacity was about 4 billion
pounds (1.9 million metric tons), but it had increased to approximately 10 billion pounds (4.5
million metric tons) by the early 1980s. Worldwide capacity has been estimated to be 36 million
pounds (16.5 million metric tons) or perhaps slightly more by about 1980. During the economic
recession of the early 1980s, several smaller and/or less efficient vinyl chloride units were shut
down, either temporarily or permanently. Most units also operated at less than full capacity.

Annual production rates dropped in 1982 to a temporary low of 4. 9 billion pounds (2.23 million
metric tons), but with the improved economy in 1983 the rate increased to almost 7 billion
pounds (3.16 million metric tons) [8]. Production rates tend to correlate well with the health of
the economy.

Page 28
Almost all vinyl chloride is used in the production of high polymers and by far the largest amount
is used for production of polyvinyl chloride (PVC). Improvement in the methods of producing
vinyl chloride, of polymerizing it, and of stabilizing and fabricating the polymers have each
contributed significantly to the present large-scale importance of vinyl chloride and PVC
plastics. The 1970s were a traumatic period for both vinyl chloride and PVC production. In the
mid-1970s, vinyl chloride was found to be a carcinogen. Initially, it was feared that neither vinyl
chloride nor PVC plants could meet the required environmental standards necessary to protect
the workers. Cost-effective methods were, however, found to meet the standards and to
minimize vinyl chloride losses to the surroundings [7, 22], as will be discussed later. Then in the
late 1970s and early 1980s, both the building and transportation industries, which use large
amounts of PVC polymers, were badly hurt by the economic recession.
PVC competes to an appreciable extent for several markets or end uses with the other major
thermoplastics: polyethylene, polypropylene, and polystyrene. Hydrocarbon feedstocks used
for production of all four of these thermoplastics are liquefied paraffins recovered from natural
gas and various distillate fractions obtained from crude oil. Chlorine, produced mainly from
common salt (NaCl), is also needed for the production of vinyl chloride (and hence PVC); on a
weight basis, chlorine accounts for 56.8% of the total weight. PVC is hence less affected by
the costs of petroleum and natural gas than are the other three thermoplastics, and petroleum
and natural gas prices will probably rise in the future more rapidly than chlorine prices. Since
feedstock costs are by far the single most important cost in producing olefins or monomers,
PVC's prices will probably increase less rapidly than those of the other three thermoplastics.
Vinyl chloride has a normal boiling point of -13.8C and is normally stored or shipped as a
liquid, which is clear and colorless. At ambient temperatures, liquid vinyl chloride has vapor
pressures ranging from about 35 to 75 psig. Gaseous mixtures of vinyl chloride and air can be
explosive and accidents have occurred when such mixtures exploded causing serious damage,
Detailed properties are tabulated in Chapter 5, Table 1.
Impurity levels, expressed in ppm, that are presently accepted as the maximum permissible are
as follows [10]:
Acetylene
Acidity, as HCl by weight
Acetaldehyde
Alkalinity, as NaOH by weight
Butadiene
1-Butene
2-Butene
Ethylene
Propylene
Dichloroethane
Nonvolatiles
Water

2.5
0.5
0.0
0.3
6.0
3.0
0.5
4.0
8.0
10.0
150
200

Iron, by weight
0.25
Most vinyl chloride is now produced using ethylene, chlorine, and oxygen (or air) as feedstocks
[2, 10, 32]. Acetylene and anhydrous hydrogen chloride are used in other vinyl chloride
processes, but to a considerably

Page 29
smaller extent. Production capacities of vinyl chloride plants have, on the average, increased in
the last 15 years. The cost of producing a given amount of vinyl chloride in general decreases
as the size of the plant increases. In the United States in 1968, new plants generally had annual
capacities of at least 400 million pounds (180 thousand metric tons), but by the early 1980s,
several plants had capacities of at least 1.0 billion pounds (450,000 metric tons).

II. MAJOR PROCESS FOR PRODUCTION OF VINYL CHLORIDE

Ethylene, which is much cheaper than acetylene, is now the predominant hydrocarbon
feedstock for production of vinyl chloride. Three different types of reactors are employed in the
major process for the three main reactions (see Figure 1). About 90% of all vinyl chloride is
produced by this balanced process; a balanced process is defined here as one in which the
product of one reactor is used as feed to another reactor.
Ethylene is first chlorinated to produce 1, 2-dichloroethane, which is then pyrolyzed to produce
vinyl chloride and hydrogen chloride:
CH2=CH2 + Cl2 CH2Cl-CH2Cl
[1]
CH2Cl-CH2Cl CH2=CHCl + HCl
[2]
Then additional ethylene, the hydrogen chloride from reaction (2), and oxygen (or air) are
reacted to form more 1, 2-dichloroethane; this reaction is normally referred to as an
oxychlorination reaction.

Adding reactions (1), (2) (taken twice), and (3) yields the overall reaction for balanced process
shown in Figure 1:

The three major phases of the process are described in detail next.

FIGURE 1 Balanced process using separate oxychlorination step. (From Ref. 1.)

Page 30
A. Chlorination of Ethylene
Ethylene is chlorinated commercially by both liquid-phase [2, 10, 14, 18, 19, 32] and gas-phase
reactions [2, 5, 35]. Liquid-phase chlorination involves using the product, 1, 2-dichloroethane, as
the solvent for chlorine and ethylene. These reactants are dissolved by bubbling them up
through the liquid. Mechanical agitation is sometimes used to promote solubility. In the Kureha
process [14], the reactor is operated at a pressure of 4 to 5 atm and a temperature of 50 to
70C. The molar ratio of the entering chlorine to ethylene varies between 0.96 and 0.98. Other
liquid-phase reactors employ somewhat different ratios that may even introduce a slight excess
of chlorine. Conversion of the limiting reactant is almost 100% and selectivities to 1, 2dichloroethane are often greater than 99% [22]. The main by-product is 1, 1, 2-trichloroethane.
Processing of the effluent stream from the reactor depends on the ratio of reactants and on the
amount of the unreacted components. Iron chlorides are either deliberately added or are
formed in situ as chlorine reacts to at least a limited extent with the metal surfaces of the
reactor. Ferric chloride catalyzes the chlorination as indicated by the following reactions [2]:

Two modifications of the liquid-phase reactor have been used [10, 21, 22]. 1, 2-Dichloroethane
can be removed from the reactor either as a liquid or a gas. As a liquid, the stream is rather
impure 1, 2-dichloroethane containing by-products and dissolved ferric chloride. The ferric
chloride is removed by either water washings or by adsorption. Distillation is then used to purify
the liquid stream. The heat of chlorination when the product is withdrawn as a liquid is
transferred to either water or air; the heat is then unavailable for operation of the distillation
column. In the last several years, boiling-type reactors have been developed in which the heat
of chorination vaporizes part of the 1, 2-dichloroethane in the reactor and provides sufficient
vapors for the operation of the distillation unit used for purifying 1, 2-dichloroethane. Large
savings in steam are as a result realized. In the column developed by Stauffer Chemical Co.
[21], the top product is light ends, the side stream is relatively pure 1, 2-chloroethane, and the
bottom product is heavy ends. Ferric chloride concentrates in the heavy ends, most of which
are recirculated to the reactor; the remainder of the heavy ends is removed from the system.
Stauffer employs a slight excess of ethylene in their reactor. Boiling-type reactors appear to be
the preferred type for the future.
In the vapor-phase processes, chlorine and ethylene react at temperatures of about 90 to
130C and at pressures of approximately 7 to 10 atm. [1, 2]. The reaction is highly exothermic,
and excellent temperature control must be maintained to prevent runaway or even explosive
reactions. A shell-and-tube heat exchanger is frequently used as a reactor, and the gases pass
in such cases through cooled tubes.

Page 31
Other provisions often used in the vapor-phase process to maintain good temperature control
include (1) presence of an inert gas [5, 35] or excess ethylene in the feed stream to the
reactor, (2) rapid mixing of the chlorine and ethylene streams as they are combined, and (3)
granular solids as packing in the tubes of the reactor.
The walls or surfaces in the reactor often have catalytic effects on the reaction. The chlorine
gas probably actually reacts at least to a limited extent with the metal walls to produce metal
chlorides such as ferrous or ferric chlorides. The chemical role of the surface is not yet well
understood, but at least some initiation of the free-radical reactions probably occurs there.
Conversions of the controlling reactant, generally chlorine, are usually allowed to go essentially
to completion in the vapor-phase process. Some polychlorinated hydrocarbons are produced in
small quantities, depending somewhat on the reaction conditions used. The effluent from the
reactor is cooled with water, and at the pressures used almost all the dichloroethane is
condensed. It is later purified by distillation. The small amount of unreacted chlorine present is
removed by scrubbing first with water and then with a caustic solution. Unreacted ethylene is
recovered and recycled if it is present in significant quantities.
The chlorination reactions are rapid with few by-products, and the reactor design for especially
the liquid-phase process is quite simple. The cost of operating the chlorination section is the
lowest in the balanced process shown in Figure 1. The liquid-phase processes have several
important advantages compared with the gas-phase processes: better temperature control is
obtained because of the large amount of liquid present, and an excess of ethylene or the use of
inert gases is not as necessary for safety reasons. When a boiling-type liquid-phase reactor is
used, the heat of chlorination can be effectively used.
B. Oxychlorination of Ethylene
The latest to be perfected and hence the key development of the balanced process of Figure 1
was oxychlorination [reaction (3)]. This step in the overall process has been highly successful,
and investigators have publicized several versions of it [2, 6, 10, 17, 19, 22, 32, 37]. In all
cases, the catalysts used are copper chlorides that are mixed frequently with potassium
chloride or other alkali metal chlorides. The basic chemistry in the oxychlorination reactions that
occur in the range 250 to 350C are probably as follows:

At reaction conditions, the copper chlorides are a mixture of cuprous and cupric chlorides. The
potassium chloride is added since it reduces the melting point of the salt mixture to fairly low
temperaturesin the range 150 to 200C depending on the amounts added. It also decreases
the vapor pressure of the copper chloride melts.

Page 32
Both conversions per pass and overall yields of 1, 2-dichloroethane are as high as 97 to 99% in
at least some units. By-products obtained are 1, 1, 2-trichloroethane, ethyl chloride, and
polychloroethane, including hexachloroethane. Some carbon dioxide is also produced as a
result of undesired oxidation reactions. The amount of by-products obtained varies with the
catalyst and the operating conditions used. In a well-operated balanced process using
oxychlorination (see Figure 1) about 0.48 to 0.50 lb of ethylene and 0.67 to 0.70 lb of chlorine
are required to produce a pound of vinyl chloride [2].
The ethylene used in the oxychlorination step must, in general, be considerably purer than the
ethylene required in the chlorination step. The actual specifications for ethylene in the
oxychlorination reaction vary depending on the specific process conditions and the catalyst
used. One specification for the ethylene indicates that water, olefins heavier than ethylene, and
oxygenated compounds (carbonyl compounds, esters, and alcohols) should each be maintained
at less than 50 to 100 ppm. The sulfur content should be no greater than 1 to 2 ppm. Generally,
ethylene with a purity of at least 99.5% is used, but nitrogen, hydrogen, nonolefinic
hydrocarbons, and carbon oxides totaling several percent can sometimes be tolerated.
Solid catalysts are employed in all commercial oxychlorination units. Alumina, silica, and other
porous solid materials are used as supports for the chloride salt catalyst. These solids
apparently complex or act as absorbants for the salts, which are otherwise quite volatile. Even
so, loss of salts and hence catalyst activity can occur slowly at reaction temperatures. The
porosity and size of particles used are important relative to the overall performance of the
catalyst, but information is limited concerning these factors. Transfer of the reactants and
products to and from the catalyst surface and in the pores is obviously important in addition to
the adsorption (or absorption) of the reactants, especially oxygen [1, 2]. Temperature has a
complex effect on the process since it affects (1) the kinetics of the various chemical steps, (2)
the melting points and viscosities of the molten salts, and (3) the solubility or adsorption of the
reactants on the surface.
Temperature control of this highly exothermic reaction is important, and two types of reactors
are involved when solid catalysts are used. In one type, the catalyst is packed inside the
tubes of what is basically a shell-and-tube heat exchanger [10, 22, 37]. Steam is generated in
these heat exchangers when water is used as the coolant. Tubes with internal diameters of 2
in. or less have been used. Corrosion may sometimes be a problem in such a reactor since
steam, hydrogen chloride, and perhaps even molten salts are present. Nickel and stainless
steel tubes are frequently mentioned in the patent literature, but carbon steel is indicated for
one commercial unit.
A fluidized or expanded bed of solid catalyst is used in the second type of reactor system [6,
10, 19, 22, 32]. The solid catalyst is fluidized by the reactant gases, and essentially complete
reaction occurs as the gases pass through the fluidized bed within several seconds' residence
time. It is claimed that the fluidized-bed reactor has a marked reduction in capital and operating
costs compared with the fixed-bed reactor. Factors that must be considered in designing the
fluidized-bed reactors include minimizing formation of catalyst fines, selecting materials of

construction to minimize erosion or corrosion of the reactor walls, and maintaining uniform flows
throughout the reactor. In the last few years, there seems to have been increased use of
fluidizedbed reactors compared to fixed-bed reactors.

Page 33
The operating conditions used in oxychlorination reactors vary to an appreciable extent
depending on the catalyst used and on whether air or oxygen is used as the oxidant.
1. Temperatures of 200 to 350C have been reported. Temperatures of approximately 200 to
250C are presumably used with new or very active catalysts. Catalyst life and yields of 1, 2dichloroethane are increased if low temperatures can be maintained.
2. Pressures in the range 2 to 10 atm have been reported in the literature. When a fluidized-bed
reactor is employed, lower pressures would seem to be preferred to allow the large volumes of
gas necessary for fluidization. Pressure also affects the density of the gases, and hence
indirectly the problems associated with temperature control. Higher pressures, however,
simplify condensation and separation of the product stream.
The feedstock ratios or compositions for an oxychlorination process can be varied significantly.
First, either air or oxygen can be used [27, 38]. Air was used almost exclusively until perhaps
10 years ago, but oxygen use has increased steadily since then. When oxygen is used, most of
the effluent gas from the oxychlorination reactor is recycled with the result that a large
concentration of argon (an impurity in the oxygen) and carbon dioxide builds up. Such a recycle
permits temperature control of the reactor (and reactants) that is comparable to the control
possible with air. There are two major advantages to the use of oxygen. First, there is easier
recovery of 1, 2-dichloroethane and less loss of vinyl chloride (an impurity in the exhaust
gases). Second, there is a drastic reduction in the volume of vent gas discharged. These
factors are of major importance in meeting environmental standards. Oxygen is, of course,
more expensive than air.
Attempts have been made to use aqueous solutions of hydrogen chloride in addition to
anhydrous hydrogen chloride. If aqueous solutions could be used, slight excesses of hydrogen
chloride could be fed to the reactor with the result that aqueous solutions would be produced.
Furthermore, aqueous solutions can sometimes be purchased cheaply or are attainable from
other processes. Such solutions can be used in the M.W. Kellogg process [13, 17] to be
discussed later. Occasionally, a plant may have a need for producing aqueous solutions of
hydrogen chloride for use in other processes, in which case excess hydrogen chloride may be
used. Unreacted ethylene that is present in the effluent stream of an oxychlorination reactor can
be converted (and hence recovered) to 1, 2-dichloroethane by a chlorination step [30].
The M.W. Kellogg oxychlorination process [13, 17], although never used commercially, is a
most interesting alternative that may eventually find industrial use. Mixtures of ethylene,
hydrogen chloride, and oxygen are bubbled up through an aqueous solution of cupric chloride at
about 170 to 185C and at pressures of 16 to 20 atm. This process has several important
advantages:

1. Excellent temperature control is possible, and high yields of the desired product are claimed.
2. Aqueous solutions of hydrogen chloride that are frequently quite cheap can be used as a
feedstock.
3. Chlorination as well as oxychlorination of ethylene can be practiced simultaneously in this
reactor system. A balanced process using such a reactor is shown in Figure 2.

Page 34

FIGURE 2 Balanced process with combined oxychlorination and chlorination steps. (From Ref.
1.)
C. Pyrolysis of 1, 2-Dichloroethane
Dehydrochlorination of 1, 2-dichloroethane [reaction (3)] is industrially accomplished mainly if
not exclusively by noncatalytic thermal reactions [1, 2, 5, 10, 22]. The dehydrochlorination
occurs primarily by means of the following two free-radical reactions:

The Cl radical to initiate the chain mechanism above can be formed by thermal decomposition
of either 1, 2-dichloroethane or elemental chlorine deliberately added to the feed stream as
indicated by the following two equations:

Noncatalytic dehydrochlorination of 1, 2-dichloroethane occurs readily in a matter of seconds at

470 to 540C. Vinyl chloride yields based on the 1, 2-dichloroethane feed are in range 97 to
99% when 50 to 70% of the 1, 2-dichloroethane is converted per pass. Yields in general
decrease as the conversions increase. Conversions of 60 to 70% are thought to be common in
most current vinyl chloride units. At least two pressure ranges are employed. B. F. Goodrich
Chemical Co. employs inlet pressures of about 24 to 26 atm [15, 20]. Some companies prefer
much lower pressures, about 7 to 10 atm. Both Imperial Chemical Industries (ICI) and Solvay
et Cie indicate that they prefer the higher pressures.
The Goodrich high-pressure pyrolysis process is described by at least two patents [15, 20] and
Figure 3 shows a tube (or coil) arrangement in the box-type furnace employed. Several tube
sizes were reported in the patent, and they vary in inside diameter from 1.0 to 3.35 in. and in
length from 335 to 970 ft. The larger tubes are probably used in the bigger units, and stain-

Page 35

FIGURE 3 Box-type furnace for high-pressure pyrolysis of dichloroethane. (From Processes for
Major Addition-Type Plastics and Their Monomers, by Lyle F. Albright. Copyright 1974 by
McGraw-Hill, Inc. Used with permission of McGraw-Hill Book Company.)
less steels rich in both nickel and chromium are required to obtain several years of service.
Schmillmoyer [29] has discussed steels suitable for vinyl chloride units. Alloys 600 and 800
have both been successfully used.
The first two U sections of the tubes are provided for vaporizing and for preheating the entering
1, 2-dichloroethane that is at optimum pressure of 24 to 25 atm. These sections are in the
cooler, convection portion of the furnace. The actual pyrolysis occurs primarily in the last three
U sections of the tubes located in the radiant section of the furnace. Temperatures in these
sections of the tube range from about 470 to 540C. Gas burners are provided for heating the
tubes.
Information in the patents was used to calculate various factors concerning operation of the
pyrolysis units. Average residence times of the gases in the commercial tubular reactors vary
and are about 9 to 20 see. Average rates of heat transfer in the tubes vary and are about
10,000 to 30,000 Btu/ hr per square foot of heat transfer surface. Actual rates in the reaction
zone are probably even higher. Relatively large temperature differences obviously occur in the
boundary layer of gases on the inside walls of the tubes.
When the larger tubes are used, 70 to 130 million pounds of vinyl chloride are produced in a
tube each year. Overall yields of vinyl chloride based on dichloroethane reacting are reported to
be as high as 99%. Slightly higher yields than these are probably obtained commercially. Byproducts obtained in small quantities include acetylene (obtained by pyrolysis of vinyl chloride),
butadiene, benzene, and methyl chloride. Some tar and cokes apparently form on the tube
walls, but the buildup is very slow. The tubes have to be cleaned by burning them out with air
about every 6 months.
Figure 4 is a flowsheet of the pyrolysis furnace plus auxiliary equipment for it. Liquid
dichloroethane is pumped from a storage tank to the pyrolysis furnace. The exhaust gases from
the furnace are cooled in a quench column wherein a stream of a cooling liquid, which is the

crude condensation product

Page 36

FIGURE 4 High-pressure pyrolysis unit for dichloroethane. (From Processes for Major AdditionType Plastics and Their Monomers, by Lyle F. Albright. Copyright 1974 by McGraw-Hill, Inc.
Used with permission of McGraw-Hill Book Company.)
of pyrolysis, is continuously sprayed and then is recycled. This column is operated to remove
only the superheat of the vapors, and the gases leave at their dew point.
A heat exchanger shown in Figure 4 is used to condense the product from the quench column.
The liquid at a pressure of about 12 atm is then fed to the first distillation column, and highpurity hydrogen chloride is obtained as a top product of this column. Vinyl chloride is the
product of the second column, which is operated at about 4.8 atm. The third column separates
lowboiling by-products, and the fourth column separates the high-boiling byproducts from the
dichloroethane, which is recirculated to the pyrolysis furnace.
For the low-pressure dehydrochlorination processes, the energy that needs to be transferred
per unit volume of reacting gas is considerably less than that transferred in higher-pressure
processes. At lower pressures, the densities of the reacting gases are less, and there is hence
less mass of reactants per unit volume, and lower rates of heat transfer are needed for the
endothermic reactions. As a result, larger-diameter tubes (that have lower surface-to-volume
ratios) can be employed. Tubes of 5 or 6 in. I.D. are frequently used in the low-pressure units.
Condensation and recovery of vinyl chloride is more difficult in these low-pressure processes
since lower temperatures are needed to obtain the desired condensation of product stream.
Solid catalysts [4, 5, 11, 34, 35, 37] and inert solids [9, 26] have been packed in tubular
reactors used for dehydrochlorination of 1, 2-dichloroethane. Lower operating temperatures
and/or higher vinyl chloride yields have been claimed. Coke will, however, collect on the
granular solids so that eventually the unit needs to be decoked or the catalyst or inert solids
need to be replaced. No significant advantages for such packed reactors are apparent,
however, compared to a well-operated noncatalytic unit. All American units now in operation are
thought to be unpacked.

Page 37
More information is still needed concerning techniques for minimizing coke formation and
concerning the role of surface reactions during the dehydrochlorination reaction. Purification of
the feed 1, 2-dichloroethane is important. The presence of oxygen [39] , chloroprene [33],
chloral [22], or trichloroethylene [36] lead to undesired side reactions and to lower vinyl chloride
yields. Acetylene, butadiene, and benzene are all known coke precursors [3, 16]; all three
compounds are formed in small amounts during the dehydrochlorination reactions. Surface
reactions are involved in the mechanism of at least part of the coke since the coke contains
finely divided metal granules. Chlorine and hydrogen chloride react with the metal surfaces to
produce ferric chloride, which is volatile and is transferred through the tubes.
Several companies are attempting to develop transfer line exchangers to be used instead of the
quench tower in Figure 4. Such exchangers would not only cool the product stream from the
dehydrochlorination furnace but would also produce high-pressure steam. Although such
exchangers are widely used in pyrolysis units of ethylene plants for recovery and reuse of
energy, significant modifications will probably be needed before they can be adapted to vinyl
chloride units. Formation or collection of coke in the exchangers is a somewhat different
phenomena in vinyl chloride units, and problems experienced have yet to be solved sufficiently.

III. HYDROCHLORINATION OF ACETYLENE

The predominant method of producing vinyl chloride until the early 1960s always involved
hydrochlorination of acetylene using a mercuric chloride catalyst:

In some plants, balanced processes using other reactions were used, as will be discussed
later, but in many plants, reaction (15) was the only reaction and only one type of reactor was
needed. Some of the plants using hydrochlorination reactions are still in operation, particularly in
Europe.
Yields and conversions for reaction (15) are high, in the range 95 to 99%, especially when
highly purified reactants are used. Figure 5 indicates a flowsheet for the production of vinyl
chloride from acetylene and hydrogen chloride by reaction (15) [1, 2]. The acetylene used by
one industrial company is analyzed as follows on a water-free basis:
Acetylene
98.82 wt %
Total sulfur
0.0011
Hydrogen sulfide
0.0001
Organic sulfur
0.0010
Phosphorus
0.0025
Phosphine
0.0302
Ammonia
0.0052
Carbon dioxide
0.0037
The sulfur content of the acetylene must be kept low since it poisons the mercuric chloride

catalyst. Other impurities in the acetylene that may occur in small amounts include arsenic and
silicon compounds. Until the last few years, acetylene was produced commerically primarily by
causing calcium

Page 38

FIGURE 5 Hydrochlorination of pure acetylene to produce vinyl chloride. (From Ref. 1.)
carbide and water to react. Considerable amounts of acetylene are now produced, however,
from various hydrocarbons by high-temperature pyrolysis or partial oxidation processes.
The process for the production of vinyl chloride from acetylene and hydrogen chloride is
relatively simple with fairly low capital (and erection) cost. Furthermore, the hydrochlorination
process can be operated successfully at low production rates. Although some of the units built
are still used, it is doubtful that any new facilities will be built in this country unless much
cheaper acetylene can be obtained.
The entering acetylene, as shown in Figure 5, is compressed to about 2 atm (absolute) and is
then dried, with sulfuric acid used as a scrubbing agent. Higher pressures are not used, since
acetylene can then become dangerously explosive. The dried acetylene is passed through a
bed of activated carbon to remove catalyst poisons, and it is then mixed with purified hydrogen
chloride in approximately a 1:1 molar ratio.
The gas mixture at about 1.5 to 1.6 atm pressure is fed to the reactors, often arranged in
series or in parallel or both. The reactors are shell-and-tube heat exchangers with solid catalyst
particles packed in the tubes. Activated-carbon pellets impregnated with mercuric chloride are
effective catalysts. A heat transfer fluid on the shell side removes the heat of reaction,

Page 39
which is about 24.5 kcal/g mol of vinyl chloride produced. Temperatures of about 90 to 140C
are maintained in the catalyst bed. The lower temperatures are used with new catalyst, and the
life of the catalyst is probably at least half a year when the catalyst poisons are effectively
removed from the reactants.
About 98 to 99% conversion of each reactant occurs in the reactor system. By-products
obtained in trace amounts include trichloroethylene, dichloroethylene, and aldehydes. The
product stream from the reactor is scrubbed first with water to remove most unreacted
hydrogen chloride. Caustic scrubbing removes the remainder of the hydrogen chloride. After
entrained water is removed, the gases are compressed to about 7 atm (absolute) and cooled.
Most of the vinyl chloride, chlorine, and hydrocarbons is condensed. A combination of stripping,
absorption, and distillation columns is used to separate the vinyl chloride, chlorinated byproducts, and unreacted acetylene, which is recycled.
For 1183 lb of hydrogen chloride (32.5 lb mol) and 877 lb of acetylene (33. 7 lb mol), the
following yields have been obtained commercially [1, 2]: 2000 lb of vinyl chloride (32 lb mol) and
23 lb of chlorinated by-products (0.18 lb mol if pure trichloroethylene is assumed).
The balanced process shown in Figure 6 was popular before the development of the
oxychlorination step and the balanced process shown in Figure 1. Both acetylene and ethylene
were used as hydrocarbon feedstocks; chlorine was the other feedstock. The process
described in Figure 6 employs three reactors for reactions (1), (2), and (15); the overall
reaction is basically as follows:
Approximately 50% of the vinyl chloride is produced from acetylene, but this process is in
general no longer competitive with the one of Figure 1.
Another balanced process using both ethylene and acetylene as hydrocarbon feedstocks is
shown in Figure 7 [2, 5, 14]. In the latter sequence, a hydrocarbon stream such as naphtha,
gas oil, or even crude oil is pyrolyzed to produce a mixture containing essentially equal moles of
acetylene and ethylene. The C2 fraction is separated and is combined first with hydrogen
chloride. The acetylene reacts [reaction (15)] almost quantitatively to produce vinyl chloride,
which is then rather easily separated. The remaining gas stream is then combined with chlorine.
The ethylene then reacts [reaction (1)] to produce 1, 2-dichloroethane, which is recovered and
then pyrolyzed [reaction (2)] to produce hydrogen chloride and more vinyl chloride. The overall
chemistry for this process is the chemistry of reaction (16). The

FIGURE 6 Balanced process using concentrated acetylene, ethylene, and chlorine. (From Ref.

1.)

Page 40

FIGURE 7 Balanced process using mixture of acetylene and ethylene with chlorine (From Ref.
1.)

Page 41
advantage of the balanced process of Figure 7 over the process of Figure 6 is that the
expensive separation of ethylene and acetylene is replaced by the easier separations of vinyl
chloride and 1, 2-dichloroethane from the gas stream. This particular balanced process will be
of considerable commercial interest if some of the newer techniques for pyrolysis of heavy
hydrocarbons are commercialized; some of the new techniques being considered produce
considerable amounts of acetylene as well as ethylene.

IV. TECHNOLOGY OF POSSIBLE IMPORTANCE

The Deacon's process for production of chlorine from hydrogen chloride was used several
years ago in at least one vinyl chloride plant [18]. The basic chemistry is as follows:

A balanced process in which equation (1) (taken twice), equation (2) (taken twice), and
equation (17) are combined is shown in Figure 8. The overall chemistry is basically the same as
that of equation (4). Currently, no Deacon's process or modification is known that would make
balanced processes as described by Figure 8 competitive with processes as described by
Figure 1.
In the Transcat process [25, 28], which has not yet been commercialized, ethane, chlorine, and
copper oxychloride (dissolved in a molten salt) are contacted and caused to react by a
complicated series of steps. Reactions that occur include at least the following: chlorine and
ethane form ethyl chloride and hydrogen chloride; ethyl chloride decomposes, forming ethylene
and more hydrogen chloride; hydrogen chloride with copper oxychloride produces chlorine,
cuprous chloride, and water; chlorine and ethylene produce 1, 2-dichloroethane; and the latter
compound pyrolyzes, producing vinyl chloride and hydrogen chloride. In a second reactor, the
molten salt is contacted with air, causing the cuprous chloride to be reoxidized to copper
oxychloride. The overall chemistry for the process is essentially as follows:

FIGURE 8 Balanced process using oxidation of HCl. (From Process for Major Addition-Type
Plastics and Their Monomers, by Lyle F. Albright. Copyright 1974 by McGraw-Hill, Inc. Used
with permission of McGraw-Hill Book Company.)

Page 42

Efforts have been made to produce vinyl chloride from ethylene by substitutive chlorination as
follows:
However, yields are apparently always low.

V. ECONOMIC AND ENVIRONMENTAL CONSIDERATIONS

The manufacturing costs of vinyl chloride have increased severalfold within the last 10 to 15
years due to the following factors: increased cost of reactants, inflation, increased capital
costs, and increased need to minimize inadvertent losses of vinyl chloride to the environment.
McPherson et al. [22] have estimated costs for a typical American vinyl chloride plant with
balanced process of Figure 1. Operating costs when divided as indicated below and when
reported on a percentage basis are as follows: ethylene, 38.8; chlorine, 16.2; catalyst and
other chemicals, 1.8; utilities, 6.5; labor, 5.4; miscellaneous, 4.5; and capital charges, including
depreciation, 26.8. As indicated, reactant costs account for over 50% of the manufacturing
costs. For processes using acetylene as a reactant, reactant costs would be significantly higher
on both a relative and an absolute basis compared to the above-mentioned units that employ
only ethylene.
In 1981, a vinyl chloride plant with an annual capacity of 0.7 billion pounds (320,000 metric
tons) was estimated to cost $140 million if constructed in the Gulf coast area of the United
States [22]. Capital costs of such a plant have increased by a factor of almost 10 in the last 10
to 15 years. Part of the increase is caused by the need to provide additional equipment in order
to minimize vinyl chloride losses to the environment. Current regulations of the Environmental
Protection Agency (EPA) require vinyl chloride concentrations of 10 ppm or less emissions from
all point sources except the oxychlorination unit (where it is 0.02 part by weight per 100 parts of
1, 2-dichloroethane product). These regulations are under review and may be lowered in the
future. Plant modifications made in the last 10 years have frequently included installation of all
of the following units [22, 23]:
1. Primary and redundant incineration facilities for all vinyl chloride point sources and collected
fugitive emissions.
2. HCl scrubbing and neutralization or recovery units.
3. Closed process sewers, collection systems, and larger (or redundant) wastewater strippers.
4. Double seals on pumps and agitators.
5. Leak detection systems and portable monitors.

6. Enclosed sampling and analytical systems.

7. Vapor recovery systems for loading or transferring steps.
The capital cost required for the equipment listed above is not cheap, and the plant costs have
been estimated at $1 million to $15 million, depending on who makes the estimate.

Page 43
The balanced process of Figure 7 is claimed to be most attractive when a relatively cheap
source of naphtha is available, as it is in some places in Europe or Asia. In such a case, a
relatively cheap mixture of ethylene and acetylene can be obtained. In the United States,
however, naphthas (and heavier hydrocarbon feedstocks) have become less attractive within
the last few years. Ethane and propane have instead become relatively more plentiful and
cheaper; both of these feedstocks are widely used for production of ethylene. The relatively
cheap ethane now available may revitalize interest in the Transcat process [25, 28].
Balanced processes involving simultaneous oxychlorination and chlorination of ethylene in the
same reactor (Fig. 2) or a balanced process using a Deacon process (Figure 8) may eventually
be preferred to the current balanced process using an oxychlorination step (Figure 1). As
already indicated, the process of Figure 2 seems to offer several advantages compared with
the process of Figure 1. If a cheap Deacon process is developed, the balanced process of
Figure 8 would become more attractive, since the chlorination step to produce 1, 2dichloroethane is significantly cheaper to operate than the oxychlorination step to produce the
same compound.
For a balanced process in which three chemical steps are occurring, and in which each step
affects the other two, special attention must be taken to provide smooth and continuous
operation. Methods of controlling the overall process must be carefully chosen. Computercontrolled units are now common and have resulted in much improved operation.
Several companies offer vinyl chloride processes for license [12, 22]. Final selection of a
process requires careful consideration of numerous variables.

VI. STORAGE AND HANDLING OF VINYL CHLORIDE

In the past, vinyl chloride was stored or shipped with a small amount of phenol inhibitor to
prevent premature polymerization [2]. The phenol was removed before the vinyl chloride was
used for production of PVC resins by a caustic wash followed by a water wash. The vinyl
chloride produced by most manufacturers is at present of sufficient purity so that such an
inhibitor is no longer needed.
Contamination of the vinyl chloride with rust must be prevented. Water contamination in the vinyl
chloride may produce some hydrogen chloride that may be corrosive. Care must then be taken
to prevent such contamination during handling and shipping. When the vinyl chloride is dry, as it
should be after its manufacture, it can be shipped in plain carbon steel tanks or vessels.
Railroad tank cars of such steel are frequently employed for shipping, and as already
discussed, precautions must be taken to minimize vinyl chloride losses during shipment and
transfer. Some details on procedures to be followed during storage and transfer have been
reported [24, 31]; see also Chapter 5, Section IV.A.

VII. SUMMARY

Important process improvements have occurred in the last few years for the production of vinyl
chloride. The most modern American plants use balanced processes with ethylene and chlorine
as feedstock.

Page 44

REFERENCES
1. Albright, L. F., Chem. Eng., 123130 (Mar. 27, 1967); 219226 (Apr. 10, 1967).
2. Albright, L. F., Processes for Major Addition-Type Plastics and Their Monomers,
McGraw-Hill, New York, 1974, pp. 173211, 228230; 2nd ed., Krieger, Melbourne, Fla., 1985,
Chap 5.
3. Albright, L. F. and Tsai T. C. H., in Pyrolysis: Theory and Industrial Practice, L. F. Albright,
B. L. Crynes, and W. H. Corcoran, eds., Academic Press, New York, 1983, Chap. 10.
4. Braconier, F. F. and Godart, J. A., U.S. Pat. 2, 779, 804 (Jan. 29, 1957).
5. Braconier, F. F., Hydrocarbon Process., 140, (Nov. 1964).
6. Braconier, F. F., Br. Chem. Eng. Process Suppl., 79 (Nov. 1967).
7. Braconier, F. F., Chem. Eng. News, 2728 (Apr. 7, 1980).
8. Braconier, F. F., Chem. Eng. News, 3240 (June 11, 1984).
9. Cheney, H. A., U.S. Pat. 2, 569, 923 (Oct. 2, 1951).
10. Cowfer, J. A. and Magistro, A. J., in Kirk-Othmer Encyclopedia of Chemical Technology,
3rd ed., Vol. 23, M. Grayson, ed., Wiley, New York, 1983, pp. 865885.
11. Eberly, K. C., U.S. Pat. 2, 875, 255 (Feb. 24, 1959).
12. Fallwell, W. F. and Greek, B. F., Chem. Eng. News, 69 (Oct. 10, 1983).
13. Friend, L., Wender, L., and Yarze, J. C., Div. Pet. Chem., Am. Chem. Soc. Meet., New
York. Sept. 1116, 1966.
14. Gomi, S., Hydrocarbon Process., 165 (Nov. 1964).
15. B.F. Goodrich Co., Br. Pat. 938, 824 (Oct. 9, 1963).
16. Graff, M. J., and Albright, L. F., Carbon, 20, 319 (1982).
17. Heinemann, H., Miller, K. D., and Spector, M. L., Div. Pet. Chem., Am. Chem. Soc. Meet.,
New York, Sept. 1116, 1966.
18. Heinemann, H., Miller, K. D. and Spector, M. L., Hydrocarbon Process., 44(11), 198
(1965).
19. Heinemann, H., Miller, K. D. and Spector, M. L., Hydrocarbon Process., 60(11), 235237

(1981); 62(11), 155157 (1983).

20. Krehler, H. (to Farbwerke Hoechet), U.S. Pat. 2, 724, 006 (Nov. 15, 1955).
21. McNaughton, K. J., Chem. Eng. 5458 (Dec. 12 1983).
22. McPherson, R. W., Starks, C. M., and Fryar, G. J., Hydrocarbon Process., 7578 (March
1979).
23. Mears, C. L., Chem. Eng. Prog., (Oct. 1318 1979).
24. Mukerji, A., Chem. Eng. 155 (Sept. 12 1977).
25. Nislick, A. S., Riegel, R., and Abrahams, M. A., 79th Nat. Meet. Am. Inst. Chem. Eng.,
Houston, Mar. 1975.
26. N. V. deBataafsche Petroleum Maataschappij, Br. Pat. 663, 221 (Dec. 12, 1949).
27. Reich, P., Hydrocarbon Process., 8589 (Mar. 1976).
28. Riegel, H., Schindler. H. D., and Sze, M. C., Chem. Eng. Prog., 89 (Oct. 1973).
29. Schmillmoyer, C. M., Hydrocarbon Process., 8993 (Mar. 1979).
30. Severino, F. T. (to Stauffer Chemical Co.). U.S. Pat. 4, 046, 822 (Sept. 6, 1977).
31. Shelley, P. G., and Sills, E. J., Chem. Eng. Prog., 65(4), 29 (1969).

Page 45
32. Sitting, M., Vinyl Chloride and PVC Manufacture, Process, and Environmental Aspects,
Noyes Data Corporation Park Ridge, N.J., 1978.
33. Smalley, E. W., Kurtz, B. E., and Bandyopadhyany (to Allied Chemical. Corp), U.S. Pat. 4,
060, 460 (Nov. 29, 1977).
34. Societe Belge de l'Azote, Fr. Pat. 1, 290, 953 (Mar. 12, 1962).
35. Societe Beige de l'Azote, Br. Pat. 954, 791 (Apr. 8, 1964).
36. Strini, J. C., and Costes, J. R. (to Rhone-Prigil), U.S. Pat. 3, 935, 286 (Jan. 27, 1976).
37. Wacker-Chemie GmbH, Br. Pat. 979,309 (Jan. 1, 1965).
38. Wimer, W. E. and Feathers, R. E., Hydrocarbon Process., 8184 (Mar. 1976).
39. Young, D. P. (to British Petroleum Chemicals), U.S. Pat. 3, 896, 182 (July 22, 1975).

Page 46

Page 47

3 PVC Processes and Manufacture

MICHAEL LANGSAM
Air Products and Chemicals, Inc. Allentown, Pennsylvania

I. INTRODUCTION

48

II. POLYMERIZATION OVERVIEW

48

III. CLASSICAL FREE-RADICAL POLYMERIZATION

49

A. Initiation

49

B. Propagation

51

C. Other Reactions of the Free-Radical Chain End

52

D. Overall Rate Expressions

54

IV. NONCLASSICAL FREE-RADICAL POLYMERIZATION

57

A. Diffusion-Dependent Rate Constants

57

B. Multiphase-Dependent Rate Constants

58

C. Nonclassical Kinetic Model

V. POLYMER CHAIN STRUCTURE
A. Thermal Stability and Chain Defects
B. Tacticity
C. Glass Transition Temperature
D. Crystallinity and Aggregation
VI. COMMERCIAL MANUFACTURING PROCESSES

59
62
63
66
67
68
69

 VII. SUSPENSION POLYMERIZATION

A. Suspension Polymerization Overview
B. Resin Productivity

72
72
76

C. Resin Quality

103

VIII. BULK POLYMERIZATION

127

A. Bulk Polymerization Overview

129

B. Resin Productivity
C. Resin Quality

131
133

IX. EMULSION POLYMERIZATION

138

A. Emulsion Polymerization Overview

139

B. Emulsion Polymerization Kinetics

C. Resin Productivity
D. Resin Quality

143
150
151

Page 48
X. FUTURE DEVELOPMENTS
SYMBOLS
REFERENCES

152
153
154

I.INTRODUCTION
Polyvinyl chloride (PVC) has been manufactured commercially for the past 50 years [1] and is,
for calender year 1985, the second largest volume thermoplastic manufactured in the world [2].
Polyvinyl chloride is used in a wide range of applications because of its combined physical
properties of:
Compounding versatility as rigids or flexibles
High modulus (unplasticized)
Ease of fabrication
Low flammability
Low cost
Polyvinyl chloride is sensitive to environmental attack, and vigorous programs to develop costeffective ultraviolet stabilizers are an ongoing area of research.
Despite its commercial maturity, there are vigorous research and development programs
currently under way to increase understanding of the polymerization and properties of polyvinyl
chloride. These efforts have been directed at polymer kinetics, large-reactor technology,
particle morphology, and polymer structure. The major developments in these areas are
discussed in this chapter.

II.POLYMERIZATION OVERVIEW
Vinyl chloride monomer is polymerized through a free-radical mechanism. There are no known
cationic or anionic systems which are effective for the homopolymerization of vinyl chloride.
Polyvinyl chloride is insoluble in vinyl chloride monomer and the polymer precipitates from the
monomer phase at 0.1% conversion [3, 3a]. This phase separation of polymer from monomer is
also found for acrylonitrile and vinylidene chloride [4]. The phenomenon of phase separation has
been quantified by Flory and Huggins [5] in a thermodynamic analysis of solvent-solute
interactions. The phase separation of the polymer from the monomer affects both the
polymerization kinetics and particle morphology of the polyvinyl chloride resin. Classical solution
free-radical kinetic mechanisms have been modified by Talamini [6] and Hamielec [7] to account
for this separation of polymer phase and monomer phase. The development of classical free-

radical polymerization kinetics is detailed in Section III, and the consequences of the phase
separation phenomenon are described in Section IV.

Page 49

III.CLASSICAL FREE-RADICAL POLYMERIZATION

A.Initiation
The polymerization process begins with the generation of a high-energy species which is
capable of interacting with the olefin double bond of the vinyl chloride monomer. The source of
this high-energy species is a molecule called an initiator which can be represented by the form
I-I'
[1]
The initiator may or may not be symmetrical (therefore, I = I). Throughout this chapter initiator
will be represented by I. The molar concentration of the initiator will be represented by [I]. The
only requirements for an initiator are that the chemical bond between I and I must have a low
bond energy and that the bond breaking occurs in a homolytic fashion. The consequences of
these requirements have been to limit the initiator types which are useful commercially in the
polymerization of vinyl chloride to those based on structures such as
R1-O-O-R2
[2]
R3-N = N-R4
[3]
Homolytic bond scission occurs at the O-O bond for the peroxide and with the release of N2 for
the azo compound. The bond energy for these homolytic scissions are approximately 25 to 40
kcal/mol at 25C [8]. The formation of the free-radical species by thermally induced homolytic
bond breaking can be represented as
[4]
If we make the following assumptions and definitions:
1. I . and I' . are indistinguishable as far as the kinetic mechanism is concerned.
2. Kd is the homolytic decomposition constant for reaction (4).
then the rate of formation of free-radical species will be equal to the rate of decomposition of II. The rate expression for the formation of free-radical species is

The numerical value 2 that appears in the Ri expression as well as subsequently in Rd and Rt
follows the convention established by Flory and others to indicate either the formation or
destruction of two species. Rearranging (5) and (6) as

Page 50

and then integrating to determine the effects of time on the level of initiation, we have

where [I]0 is the concentration of initiator at time 0 and [I]t is the concentration of initiator at
time t. Equation (8) can be converted to the exponential form as
[I]t = [I]0 exp(-Kdt)
[9]
The decomposition of initiator follows first-order kinetics. This permits us to relate the
decomposition constant, Kd, to the half-life (t1/2) of the initiator in the following manner:

The half-life of an initiator can be determined at several different temperatures. An Arrhenius

plot of In t1/2 versus 1/T can then be used to determine t1/2 at different temperatures. The
activation energy for initiator decomposition (Ed) can be determined from a plot of In kd versus
1/T, where the slope of the line is -Ed/R. The half-life temperature for initiator is used to
compare different types of initiators in Section VII.B.1.a.
These kinetic equations indicate that the rate of formation of free-radical species by the thermal
decomposition of initiator will follow an exponential decay curve of the form
Ri = 2Kd[I]0 exp(-Kdt)
[11]
A premise of this analysis is that the decomposition of initiator is completely effective;
therefore, equation (4) always generates two active free-radical species which will each initiate
polymerization. In actuality there are numerous side reactions which can occur to the free
radicals before they diffuse apart [9]. The consequences of these side reactions are to reduce
the apparent level of initiator. This reduction can be treated mathematically by including an
efficiency factor, f, in equation (11). The value of f will vary for different vinyl chloride-initiator
systems and will usually fall in the range 0.5 < f < 1.0. The final form for the isothermal rate of
formation of free radicals is
Ri = 2Kdf[1]0 exp(-Kdt)
[12]
The constants Kd and f are difficult to quantify individually. In most kinetic models the two
constants are considered inseparable. Initiator decomposition studies described in Section
VII.B.1.a detail the effects of the solvents on the value for Kd.

Page 51
B.Propagation
The propagation step in a free-radical polymerization is a series of sequential additions of
monomer units in a fashion depicted in Figure 1. Each propagation step has a unique kinetic
rate constant: Kp1 for the first addition of monomer, Kp2 for the second addition of monomer,
and Kpi for the addition of the ith monomer unit. The overall kinetic expression for propagation
can be considered the summation of each of the ith steps. The rigorous use of this summation
approach can make the kinetic expression very difficult to handle. Flory [10] has made the
approximation that above a certain chain length the rate constant for polymerization of the ith
sequence will be equal to the i + 1 sequence. In that way the value for Kpi will suffice for the
entire series. The effect of chain length on Kp approaches an asymptotic value at chain lengths
of about 10 repeat units. Since the chain length of PVC is typically of the order of about 700
repeat units, the effect of variations in Kpi for the first 10 sequential units can be neglected in
the overall rate of propagation. The kinetic expression for propagation can then be simplified as

where [Mi] is the concentration of active polymer chain ends, [M] is the concentration of
monomer, Kpi is the average rate constant for polymerization, and Rp is the rate of
polymerization.
The polymerization expression defined in equation (13) was developed for isothermal
conditions. The effects of variations in polymerization temperature on the propagation step can
be quantified through an Arrhenius relationship as

FIGURE 1 Kinetic sequence for propagation.

Page 52
where A is a constant, Epi the activation energy of polymerization, R the universal gas constant,
and T the absolute temperature. Raising the temperature increases Kpi and therefore increases
Rp. Throughout the rest of this chapter the propagation rate constant for chain length of greater
than 10 units will be represented by Kp.
C.Other Reactions of the Free-Radical Chain End
The growing polymer chain end is an active free-radical species. This chain end will add
monomer in a polymerization mode but will also react in a manner that will either terminate the
active species or transfer it to another site. The kinetic rate constants for these reactions will
determine both the molecular weight and molecular weight distribution of the polymer.
1. Termination Reactions: The termination steps in a free-radical polymerization are all the
reactions that results in a total loss of active chain ends. In free-radical reactions this can occur
either through the use of a chain terminating reagent (also called a short-stop agent) or by the
mutual reactions of two active chain ends. Termination reactions can be of two forms:
disproportionation reactions or combination reactions.
In disproportionation two active ends interact [in reaction (15)] as

In combination reactions the two active ends combine to form

In both cases the rate expression for termination will be in the form
Rt = 2Kt [Mi ] [Mi ]
For this discussion the termination rate constant for disproportionation and combination are
indistinguishable. In the case of vinyl chloride polymerization, the termination occurs primarily by
disproportionation. Green and Paisley [11] have examined the molecular weight distribution of
polyvinyl chloride using a Schultz [11a] distribution procedure and found that the resin has a
molecular weight distribution of 2.0 (MwMn). If the termination process followed a coupling
procedure, the molecular weight distribution would have been 1.5. These results support the
termination by a disproportionation mechanism.

Page 53
2. Chain Transfer Reactions: The propagation step described in Section III.B indicates that the
active chain end grows in a stepwise manner, adding one monomer unit at a time. The
implication of this mechanism (see Figure 1) is that chain growth occurs continually along a
given chain. In actuality, the active species can transfer from the growing chain end to a new
and independent site, as shown in reactions (17) and (18).

The material AB represents a molecular species which will homolytically dissociate at the
bond between A and B. The result is the termination of the polymerizing chain [in reaction (17)]
and the initiation of a new chain [(in reaction (18)]. In the polymerization of vinyl chloride in the
presence of AB there will be the two competitive reactions represented by (13) and (19) as

Combining (13) and (19) defines a quantity called the chain transfer coefficient [12] (CTr), which
is the ratio of rate constant for transfer divided by rate constant for propagation.

In free-radical polymerization there can be chain transfer to monomer, solvent/chain transfer

agent, initiator, and polymer. The effects of chain transfer reactions on the number-average
molecular weight of polymers have been discussed by Walling [13]. The specific case for
polyvinyl chloride is detailed in Section III.D.2.

Page 54
3. Polymerization Temperature Effects: The active chain end in polyvinyl chloride will rapidly
chain transfer to free monomer during the polymerization. This chain transfer reaction at the
polymerization temperature forms the upper boundary on the molecular weight of the polymer.
Barr [14] has documented the relationship between polymerization temperature and relative
viscosity measurements. This can be recast into a graphic form where both number-average
molecular weight and inherent viscosity can be related to polymerization temperature. This is
depicted in Figure 2. The relationship spans the temperature range 45 to 80C and indicates
the reduction in both inherent viscosity and number-average molecular weight with increases in
polymerization temperature. The commercially useful range for polyvinyl chloride is in the
inherent viscosity range 0.50 to 1.10. The mechanism for sthe chain transfer of the active chain
end from the polymeric vinyl chloride to vinyl chloride monomer is discussed in Section V.
D.Overall Rate Expressions
Thus far in Section III the individual kinetic expressions for initiation, propagation, termination,
and transfer have been presented. The expression

FIGURE 2 Inherent viscosity and number-average molecular weight vs. polymerization

temperature.

Page 55
[Mi ] has appeared in these rate relationships. Quantification of the rate expressions with [Mi ]
can be done if the Flory steady-state approximation [15] is used. The steady-state
approximation holds that the rate of formation of active chain ends will be equal to the rate of
disappearance of active chain ends at steady state.

Therefore, the concentrations of active chain ends [M i] will be

This expression for the concentration of active ends will enable us to calculate two very
important quantities for a polymerization:
1. How fast is the polymer being formed?
2. What is the molecular weight of the polymer?
1. Propagation Rate: Substituting into equation (24), the propagation rate expression (13)
indicates that

Further rearrangement can be made to the propagation rate expression rate expression so that
it is in the form

In this form the propagation rate expression accounts for the time dependent decrease in
initiator concentration and has the propagation and termination constants in the form
,
which is more useful for determining kinetic expressions. A further modification of the
propagation expression is to convert it from the disappearance of monomer (-d[M] /dt) to the
formation of polymer.
Since the sum of the weight fractions of polymer and monomer is equal to 1, [M] = 1 - [P], or 1
- x, where x = conversion. Then equation (26) can be restructured to be

Page 56
A term V is included to account for the volume contraction of the system as monomer is
converted into polymer. The kinetic approach indicates that the rate of polymerization will be a
function of the square root of the initiator concentration.
The propagation expression considers only the reaction of active chain end to monomer. In
Section III.C.2 it was shown that the active chain end [M ] could react with other materials in a
chain transfer reaction. The chain transfer reactions will not affect the overall propagation
expression provided that the concentration of chain transfer species is low relative to monomer
and the chain transfer constant (KTr) is of the same magnitude as the propagation constant
(Kp).
2. Number-Average Molecular Weight: The kinetic rate relationship developed in Section III.D.1
can be rearranged to the kinetic chain length, , which is the average number of monomer
repeat units reacting with the average active center [Mi ] from initiation to termination.

The expression for kinetic chain length in (28) does not account for effects of chain transfer
reactions. When those chain transfer reactions are considered, the expression for kinetic chain
length becomes

This expression can be expanded as follows for number-average molecular weight in its inverse
form:

where KTrS, KTrM, and KTrI are the kinetic rate constants for transfer to solvent, monomer,
and initiator, respectively.
Equation (30) can be rearranged to the following form:

The kinetic chain length (), or as it is also defined, the number-average molecular weight, is
inversely proportional to the square root of initiator concentration [for the first term in equation
(31)] and inversely proportional to initiator concentration [for the fourth term in equation (31)].
Since CI is usually very small, the first term is dominant in equation (31).
Classical kinetic equations have not been effective in predicting the rate of polymerization or the
molecular weight of polyvinyl chloride. The causes for the collapse of classical kinetics in the
case of polyvinyl chloride and the development of predictive equations for rate of polymerization
and molecular weight are discussed in Section IV.

Page 57

IV.NONCLASSICAL FREE-RADICAL POLYMERIZATION

The kinetic relationships for classical free-radical polymerizations were developed in Section III.
The assumptions used in that development were that the rate constants were not diffusion
dependent and that there was single phase for the reaction site.
The classical picture of a growing radical chain is one in which the active end is in an open
environment and is able to react with monomer radicals or with other radical chain ends. At
relatively high conversions the active radical end is in a restricted environment in which the chain
segments of the polymer surround the growing radical end. The size of this restricted
environment depends on the number of monomer units added to the active end and the
difference between the interaction of polymer-polymer segments and polymer-monomer units.
When either the number of chain segments is low (therefore, low molecular weight) or the
polymer-monomer interaction is high, the active chain end is in an open and accessible
environment. By comparison, if the number of chain segments is high (therefore, high molecular
weight) or polymer-monomer interaction is low, the active chain end is in a restricted
environment.
In the homopolymerization of vinyl chloride there is an effect of diffusion dependence for some
of the rate constants and there is a multiphased system effect during the major portion of the
polymerization. The effects of these conditions on the polymerization kinetics are discussed in
Sections IV.A and IV.B.
A.Diffusion-Dependent Rate Constants
Hayden and Melville [16] studied the polymerization of methyl methacrylate in a bulk system.
Their experimental results (Figure 3) indicate that both Kt and Kp decrease with polymer
conversion and a limiting conversion of ~81% is the maximum conversion at the end of the
reaction. Hamielec and co-workers investigated this phenomenon for polymethyl methacrylate,
polystyrene, polyacrylonitrile, and polyvinyl chloride [17, 18, 19]. Their work indicated that these
monomers exhibit limits to the conversion which are dependent on the polymerization
temperatures. This is shown in Figure 4. This phenomenon of conversion limit occurs in both
homogeneous polymerizations (polymethyl methacrylate and polystryene) and in multiphase
polymeri-zation (polyvinyl chloride and polyacrylonitrile). In these individual polymer/monomer
systems the presence of monomer decreases the glass transition temperature of the polymer.
If the polymerization temperature is below the glass transition temperature of the
monomer/polymer mixture, the diffusion of monomer to the active site approaches zero and the
rate of polymerization falls to a very low level. The data presented in Figure 4 indicate that for
PVC polymerized at ~55C there is a diffusion-controlled limit of 96% conversion.
Soong and coworkers [19a] have recently described the effect of diffusion-controlled kinetics
for the homogeneous free-radical polymerization of polymethyl methacrylate. Their particular
model can predict the molecular weight distribution of the polymer as well as the rate of
polymer formation. This approach to kinetic modeling of homogeneous polymerization system
forms an excellent comparison to the heterogeneous models described by Hamielec.

Page 58

FIGURE 3 Effect of conversion on the propagation and termination rate constants and on the
rate of polymerization for the bulk polymerization of methyl methacrylate. (From Ref. 16.)
B.Multiphase-Dependent Rate Constants
In the polymerization of vinyl chloride the polymer phase separates from the monomer at a
conversion of 0.1% and the polymerization occurs simultaneously in both the polymer phase
and the monomer phase. This phase separation is the dominant factor which distinguishes the
kinetic modeling of polyvinyl chloride from all homogeneous kinetic models. Over the past 15
years various approaches have been presented by Talamini et al. [6], Ugelstad et al. [20],
Kuchanov [21], and Olaj [22]. The common feature in all these models is that there is an
assumption of polymer formation in both the monomer and polymerphases. Talamini's model
assumes that the polymerization proceeds in both the monomer and polymer phases quite
independently of each other. Radicals and polymer formed in each phase grow and terminate in
that phase without and transfer of active radicals between phases. By comparison, Ugelstad's
model accounts for a transfer of radical species between the two phases by a
sorption/desorption phenomenon. The Talamini and Ugelstad model's give identical rate
expressions at very low conversions. Olaj has further modified the Talamini model by assuming
that radicals formed in the monomer phase can grow only by addition of monomer to a tightly
coiled structure. This tightly coiled structure with its active radical end is quickly absorbed into
the polymer phase because it is thermodynamically unstable in the monomer phase. All these
models have differences that affect the kinetic modeling of the suspension polymerization
process at different conversions. Abdel-Alim and Hamielec [7] have modified the Talamini
approach by including the change in

Page 59

FIGURE 4 Polymerization temperature vs. limiting conversion for different monomer-polymer

systems. (From Ref. 18.)
volume that occurs as a function of conversion and changes in initiator as a function of
conversion. This model has successfully predicted the rate of polymerization for vinyl chloride
and has been used in a number of industrial applications.
C.Nonclassical Kinetic Model
To calculate the kinetic expressions for the polymerization of vinyl chloride, it is necessary to
know the composition of the two phases at all times during the polymerization. The solubility of
vinyl chloride in polyvinyl chloride depends on the thermodynamics of the system. The
equilibrium concentration of monomer dissolved in polymer will be a function of the free energy
of mixing and the free energy for the surface area. The loss in free energy by mixing needs to
be balanced by the gain in free energy that the polymer particles experience by increasing their
surface area by swelling with monomer. Abdel-Alim and Hamielec [7] have examined the
composition of the phases in a vinyl chloride polymerization. Their experimental results are
illustrated in Figure 5. They demonstrated in this study that phase separation occurs during the
polymerization, with one phase being rich in monomer and the other rich in polymer. Figure 5
can be considered as an equilibrium diagram for the dual mixture of monomer and polymer. The
monomer phase has only trace amounts of polymer, probably less than 0.1% and may
therefore be considered pure monomer. The polymer-rich phase is at equilibrium with the
monomer

Page 60

FIGURE 5 Two-phase model for vinyl chloride/polyvinyl chloride mixture as a function of

temperature.
as long as there is a free monomer phase. The concentration of vinyl chloride monomer in
equilibrium with polyvinyl chloride is defined to be 1 - xc and is approximately 23% by weight at
50C.
Figure 5 demonstrates that there are three distinct stages in the polymerization of vinyl chloride.
1. 0 to 0.1% conversion: Polymer soluble in monomerhomogeneous kinetics are obeyed.
2. 0.1 to ~77% conversion: Polymer phase separates from monomer and forms a series of
agglomerated particles. The polymer phase has about 23% monomer dissolved in it as well as
initiator. The monomer phase shrinks as the polymer phase grows and absorbs monomer.
When the total conversion of the system reaches a value of xc, there is no longer a free
monomer phase present in the reaction mixture. The disappearance of the free monomer phase
is associated with the pressure break in the reaction.
3. >77% conversion. At the pressure break the system returns to a single phase of polymer
swollen with monomer. There no longer is a free monomer phase. It should be noted that the
disappearance of the free monomer phase is defined as xc and is temperature dependent. The
typical value for xc varies with polymerization temperature as

p (C) xc 30 0.80 50 0.77 70 0.72

Page 61
Abdel-Alim and Hamielec [7] have estimated the situation of polymerizdtion in both phases and
they have derived both a general set equations and limiting case equations specific for vinyl
chloride polymerization. Their derivation is presented in lecture notes prepared by Hamielec
[23] and Abdel-Alim [24] and is

where the term 1 - Bx corrects for changes in volume as a function of conversion, B is defined
as dp - dm/dm, and P is an adjustable parameter.
This relationship of monomer conversion as a function of time has successfully predicted the
conversion of vinyl chloride under a wide range of temperatures. This is illustrated in Figure 6.
The effect of temperature on the kinetic parameters is tabulated in Table 1. These relationships
have been developed for azobisisobutynonitrile (AIBN) initiator. Any charge to another initiator
would require recalculation of kinetic parameters. Despite the use of these adjustable
parameters, this model has successfully been used to quantify the kinetics of
homopolymerization. It has not been used for copolymer systems because of the as-yetunaccounted effect of comonomer partitioning between the phases.
Figure 6 illustrates the heat kick phenomenon for vinyl chloride polymerizations. At about 60 to
70% conversion the termination reaction becomes diffusion limited and the rate of
polymerization increases. Multiple initiator systems are described in Section VII.B.1.b which
reduce the effects of this diffusion-controlled heat kick.
This kinetic model has also been used to predict the molecular weight distribution for an
isothermal polymerization successfully. This is illustrated in Figure 7. This model predicts that
the concentration of initiator will have a minimal effect on the molecular weight of polyvinyl
chloride. This has been

FIGURE 6 Conversion as a function of time ([initiator]1/2). , Model; --, experimental data.

Page 62
TABLE 1 Kinetic Constants for Polymerization of Vinyl Chloride for AIBN Initiator

Source: Ref. 23.

confirmed by laboratory studies [25]. This has also been reported by Barr [14].

V.POLYMER CHAIN STRUCTURE

The properties of polyvinyl chloride resins have been examined by a variety of analytical
methods to determine the effects of polymerization conditions on

FIGURE 7 Comparison between theoretical and experimental MWD values in vinyl chloride
polymerization at 70C. MN, Number-average molecular weight; MW, weight-average molecular
weight. (From Ref. 7.)

Page 63
polymer chain structure. The polymerization of vinyl chloride has been depicted as following a
regular pattern, resulting in a polymer of highly ordered repeat structure as

This representation implies that all polymer chains are initiated by active species I, the repeat
monomer units are placed in a preferred head-to-tail arrangement, and all polymer chains are
terminated by a stopping species X.
This representation, while useful for kinetic studies, does not represent the actual situation in
vinyl chloride polymerizations. During the past decade extensive efforts have been directed
toward an understanding of both the chemical structure of the polymer chains and the spatial
and physical relationships between inter- and intrachain bonding. Studies in the areas of thermal
stability, stereoregularity, glass transition temperature, and crystallinity aggregation have shed
new light on the structure of polyvinyl chloride.
A.Thermal Stability and Chain Defects
Strong evidence that polyvinyl chloride did not have the regular structure represented by (33)
was noted in studies of the dehydrochlorination of the homologous series of chloroalkanes [26]:

The model compounds all demonstrated a resistance to dehydrochlorination well above 200C.
Under the same conditions polyvinyl chloride degraded rapidly at 150C. The primary causes
for the low thermal stability of polyvinyl chloride are rearrangement effects in the polymer
chains, unsaturated structures in the polymer chains, and the tertiary chlorine structures which
are formed by chain transfer to polymer.
1. Chain Transfer Mechanisms and Unsaturated Structures: Chain transfer from active polymer
end to monomer is a major factor in controlling the molecular weight in polyvinyl chloride resin
and introduces structures other than those depicted by (33). A mechanism for chain transfer
has been proposed by Breitenbach and coworkers [26a] in a reaction scheme where vinyl
chloride donates to the active chain end as

Page 64
The reaction scheme depicted above represents the transfer of either the -hydrogen (33a-1),
-hydrogen (33a-2), or chlorine (33a-3) to terminate the growing chain and initiate a new
polymer chain. Nuclear magnetic resonance (NMR) studies have not shown any unsaturated
end groups of the type in any of the polymers:

A chain transfer mechanism which invokes the interaction between active chain end and
monomer will not fit the present experimental evidence. Recent assumption as to the
mechanism for chain transfer are based on the head-to-head addition suggested by Rigo and
coworkers [27].
The addition of vinyl chloride monomer in an abnormal head-to-head sequence and the
subsequent rearrangement is as follows:

The head-to-head addition in equation (34) results in a 1, 2-dichloro grouping, while the
rearrangement in equation (35) results in a chloromethyl grouping. Park et al [28] have agreed
with Rigo's work and they have calculated that there will be one head-to-head addition in each
75 monomer units and that half of the head-to-head groups would rearrange to chloromethyl
groups [reaction (35)].
An alternative mechanism for the formation of the chloromethyl groups was proposed by Abbas
et al [29], who suggested that a hydrogen shift such as

could account for the presence of the chloromethyl group. The Abbas mechanism cannot
account for any 1, 2-dichloro groupings which have been reported by Mitani et al [30] by
iodometric procedure. NMR studies by Starnes [et al. [31] indicate that there are no 1, 2dichloro groups. The current experimental

Page 65
evidence is not able to differentiate between the Rigo or Abbas mechanism. The formation of
unsaturate groups by disproportionation termination has not been quantified in comparison to
head-to-head rearrangements.
The formation of the chlorobutyl group can be viewed as a 1, 5 hydrogen shift:

These butyl side groups are present at levels of about 2 to 3 per 1000 monomer units [32] and
contribute to the thermal instability of the polyvinyl chloride through the tertiary chlorine on the
backbone.
Recent work [32a] has shown that the structure of polyvinyl chloride can be examined by a
combination of pyrolysis, hydrogenation, and 13C Fourier transform NMR techniques. The
evidence for short-chain branches and thermally labile tertiary chlorine is confirmed by this
work.
The rearrangement mechanisms associated with the head-to-head addition of vinyl chloride to
the active chain end can account for both the unsaturated bonds in the resin and for the chain
transfer mechanism. The double-bond content of polyvinyl chloride has been found to be from
1.5 to 3.0 per 1000 monomer units [33]. The structure of these unsaturated groups has been
examined by NMR techniques [3436] and an overall mechanism for unsaturated groups was
proposed by Caraculacu [36] (Figure 8). The generation of active double-bond structures with
the polymer chains have been measured by ozonation procedures [37] and by chemical
reactions of allylic chlorines with phenol [38]. These unsaturated structures are accompanied by
the generation of primary chlorine or hydrogen free radicals.

FIGURE 8 Formation of unsaturated structures in polyvinyl chloride. (From Ref. 36.)

Page 66
In the commercial manufacture of polyvinyl chloride it is difficult to eliminate all traces of oxygen
from the reaction medium. Oxygen interacts with the growing polymer radical as indicated by
Bauer and Sabel [39]:

Subsequently, it was shown [40] that the CO could copolymerize with the growing chain to form
internal carbonyl structures.
The nature of the carbonyl structure has been open to speculation during the past ten years.
Kawai and coworkers [40a] initially suggested a conventional linear adduct which would result in
a carbonyl in the main chain as

Ratti and coworkers [40b] proposed a rather interesting 1, 2 chloro shift after the initial CO
addition to form a chlorocarbony site group as

Starnes and coworkers [40c] have shown, by 13C NMR spectroscopy, that the carbon
monoxide enters the polymer as the chlorocarbonyl side group [mechanism (38B)]. Small levels
of main-chain carbonyl were found [mechanism (38A)], but they were not accounted for and are
as yet mechanistically undefined.
B.Tacticity
The structure depicted for the polyvinyl chloride chains in reaction (33) is a planar
representation of a three-dimensional structure. The addition of each monomer occurs in a
manner such that there is free rotation around the terminal carbon-carbon bond. Looking at the
last two monomer units, therefore, the last four carbon atoms in the chain, there are several
possible configurations. These are depicted in Figure 9.
The term isotactic is used for the situation where the chloride substituents are on the same side
of the main chains and syndiotactic is used for the form where the chlorines are on opposite
sides of the main chain. The isotactic form can exist in both the eclipsed and skewed form. The
skewed form will be the dominant form.
The tacticity of polyvinyl chloride has been extensively studied by NMR as well as by infrared
procedures. A review of the analytical procedures used to study the tacticity of polyvinyl
chloride [41] was presented and the author concluded that NMR spectroscopy was better than
infrared spectroscopy for

Page 67

FIGURE 9 Stereospecific grouping in polyvinyl chloride.

determining tacticity. The effects of polymerization temperature on the tacticity of polyvinyl
chloride [42] indicated that there is a linear relationship between the level of syndiotactic diads
(pairs) and polymerization temperature. This is depicted in Figure 10. This relationship was
found to be independent of the polymerization process and was therefore only a function of the
energy barriers for addition of each monomer. Over the range of commercial manufacture for
polyvinyl chloride (45 to 75C) there is little variation in the level of syndiotactic diads.
C.Glass Transition Temperature
The chain segments in a polymer backbone are in a state of kinetic equilibrium; there are
interchain bonding forces holding the segments into a solid

FIGURE 10 Syndiotactic index vs. polymerization temperature. , See Ref. 184; , see Ref.
185; , see Ref. 186.

Page 68
structure while the thermal energy supplied by the environmental provides force for chain
segment motion. The temperature at which, on the atomic scale, the motion of chain segments
of about seven carbon atoms can exhibit unhampered rotation has been defined as the glass
transition temperature. At that temperature, on the macroscale, there are second-order
changes in the properties, such as thermal expansion coefficient, heat capacity, or dielectric
constant as a function of temperature. Below the glass transition temperature the polymer is a
glassy solid. Above the glass transition the polymer will behave as either a rubber or a viscous
liquid, depending on the polymer molecular weight and structure.
Polyvinyl chloride has a glass transition temperature of about 75C [43] with a linear
dependence of glass transition temperature of the polymer on the polymerization temperature.
This is shown in Figure 11. For polyvinyl chloride prepared over the temperature range 45 to
75C, there is little variation in the glass transition temperature.
D.Crystallinity and Aggregation
In either solution or in a solid state, the concentration and/or orientation of the polymeric
segment will be in a random and therefore equivalent state. The polymer system can then be
said to be amorphous. If there is an anisotropic effect on a molecular level where polymer
order or chain segment concentration is not equal, the polymer system will exhibit unusual
solution and solid-state effects.
Polyvinyl chloride has a low level of crystallinity, about 10% for polymers prepared in the range
45 to 75C [44]. As the polymerization temperature is decreased, the percent crystallinity
increases and this increase in crystallinity has been related to the increase in syndiotactic diads
in the polyvinyl chloride [45]. The phenomenon of aggregation of the polymer chains has been
identified for polyvinyl chloride with the presence of supermolecular structures approximately
1000 to 5000 in diameter. These molecular aggregates were first observed by Doty et al.
[46]. The aggregation of the polyvinyl chloride chains in solution can be reversed by heating the
polymer

FIGURE 11 Glass transition temperature vs. polymerization temperature. (From Ref. 187.)

Page 69
solution or by ultrasonic energy. The intent of these treatments are to disrupt the polymerpolymer interactions and replace them with polymer-solvent interactions. The presence of these
aggregates will affect all solution properties and will lead to an overestimation of the molecular
weight of the polymer [47]. This topic is discussed in greater detail in Chapter 7.

VI.COMMERCIAL MANUFACTURING PROCESSES

The purpose of a commercial manufacturing process is to take a raw material and to convert it
into an end product in a manner that ensures an overall profit for the manufacturer. In the
production of polyvinyl chloride the profitability of a commercial manufacturing process is
related to the ability to manufacture at the lowest cost while maintaining the product quality. The
balance of these two goals requires that the successful manufacturer be aware not only of the
current markets but also the potential future markets. The sales of polyvinyl chloride are highly
cyclical, depending in large part on the home construction industry for outlets for various
finished products. During 1983 the total U.S. sales of PVC and various copolymers reached
6.07 billion pounds, a 14% increase over 1982 sales. This substantial 1-year increase needs to
be compared, though, against a rather shallow upward trend of only 2.9% during 19731983,
[48]. The estimated annual production capacity was 8.1 billion pounds in June 1983 [48a]. At
this utilization ratio of only 75% operational utilization, only the most efficient producers can
hope to maintain a strong status in the market. A list of U.S. producers and their capacity is
given in Table 2.
There are four commercial processes for the manufacture of polyvinyl chloride. Their share in
the U.S. market is estimated at
Suspension
75%
Bulk
10%
Emulsion
12%
Microsuspension
3%
The resins from the suspension and bulk processes are in direct competition for the same
market segments: extruded pipe and profile for medium-molecular-weight resin; blown bottles
and injection molding applications for low-molecular-weight resin. The suspension process has
captured the largest fraction of both the U.S. and worldwide market because of the variety of
products that can be produced by this process and by the extensive technical experience which
has developed in the suspension process. The bulk process produces a cleaner resin and
initially (prior to 1974) had the potential for lower-cost manufacturing economics compared to
the suspension process. The need for resin with low levels of residual vinyl chloride has forced
process modifications to the bulk process and has narrowed the gap in process economics
between suspension and bulk processes.
The emulsion and microsuspension processes produce resins which are used in plastisol
applications: coated fabric, roto-molding, and slush molding. These applications tend to put a
premium on small particle size (0.5 to 20 m diameter) and high rates of plasticizer adsorption.

The commercial development of polyvinyl chloride production began with the emulsion process
in Germany in the 1930s. In 1937, I.G. Farben Industries began full-scale production of
emulsion-grade polyvinyl chloride. The suspension process was examined extensively by ICI as
a means of producing electrical cable insulation material during the early 1940s. By the early
1950s, the suspension process achieved commercial dominance throughout the world.

Page 70

Air Products, Pensacola, Fla.a

Borden, Illiopolis, Ill.
Borden, Leominster, Mass.b
Certain-Teed, Lake Charles, La.
Diamond Shamrock, Deer Park, Tex.
Ethyl, Baton Rouge, La.
Ethyl, Delaware City, Del.
Formosa, Delaware City, Del.
Formosa, Point Comfort, Tex.
General Tire, Ashtabula, Ohio
Georgia-Gulf, Plaquemine, La.
Goodrich, Avon Lake, Ohio
Goodrich, Dear Park, Tex.
Goodrich, Henry, Ill.
Goodrich, Long Beach, Calif.
Goodrich, Louisville, Ky.
Goodrich, Pedricktown, N.J.

150
340
185
225
130
180
150
310
528
120
700
400
130
200
150
375
400

150

1.9

525
225
130

6.5
2.8
1.6

330

4.0

838
120
700

10.3
1.5
8.6

1845

22.8

Page 71
Goodrich, Plaquemine, La.
Goodyear, Niagara Falls, N.Y.
Keysor, Saugus, Calif.b
Occidental, Baton Rouge, La.
Occidental, Burlington, N.J.
Occidental, Pottstown, Pa.b
Pantasote, Passaic, N.J.
Shintech, Freeport, Tex.
Tenneco, Burlington, N.J.b
Tenneco, Flemington, N.J.
Tenneco, Pasadena, Tex.
Union Carbide, South Charleston, W. Va.b
Union Carbide, Texas City, Tex.b
Vista, Aberdeen, Miss.
Vista, Oklahoma City, Okla.

190
70
50
230
180
180
50
660
160
100
700
50
100
455
260
_____
8108

70
50

.9
.6

590

7.3

50
660

.6
8.1

960

11.8

150

1.9

715
_____
8108

8.8
_____
100

aAir Products annualized capacity increased to 440 million pounds per year (MMppy) by the
startup of the Calvert City plant (250 MMppy) and debottlenecking of the Pensacola plant (190
MMppy) (July 1984).
bIncludes a small quantity of copolymer. Resin demand: 1982, 5.32 billion pounds; 1983, 6.07
billion pounds.
Source: Ref. 48a.

Page 72
The suspension process constitutes the largest segment of the polyvinyl chloride manufacturing
process and will be discussed in the greatest detail. The difference between the suspension
process and the bulk, emulsion, and microsuspension process will be indicated in each of their
respective sections. The principal concerns in the remainder of this chapter will be the
productivity and quality of the resin. The topics discussed will center on the manufacturing
control parameters: how they can be controlled to increase productivity and maintain the resin
quality within set bounds.

VII.SUSPENSION POLYMERIZATION
The suspension polymerization process has undergone a technological revolution during the
past decade. Resin porosity properties have been increased by as much as 300%, while the
residual vinyl chloride levels have been reduced from about 1000 ppm to less than 10 ppm in
the finished resin. At the same time, reactor configurations and complexity have been increased
from 4000-gal glass-lined reactors to 25,000- to 50,000-gal stainless steel reactors. The
concept of closed reactor operation has led to major technical modifications resulting in the
clean reactor process. These changes, barely conceived of in 1970, became rapidly maturing
technologies by 1980. These changes have tended to be interactive and have resulted in a
significant increase in the fundamental knowledge in polymer science. This concept of
equipment and process interactions will be illustrated in this section. An introduction to the
recipe and equipment will set the stage for the remainder of this section.
A.Suspension Polymerization Overview
1. Polymerization Recipe: The suspension polymerization of vinyl chloride is practiced in the
United States by eight major companies. Each of these companies has what they believe to be,
for their equipment and customers, the optimum process for the manufacture of polyvinyl
chloride. The recipe illustrated in Table 3 represents a typical homopolymer formulation for a
U.S. manufacturer using large reactors of 10,000- to 50,000-gal capacity. The manner in which
the recipe components are added to the reactor, as well as the time and temperature during
the addition of the individual components, are all closely guarded secrets for most
manufacturers. A procedure detailed by Johnston [48b] gives an overview of the reactor
charging process. Johnston's procedure includes a purge step for the removal of oxygen from
the reactor which probably is not needed in a closed reactor process.
The formulation will result in a resin with an inherent viscosity of 0.92 and a complete
polymerization cycle of about 7 to 11 hr. The total cycle is broken down to its individual
components in Table 3. To achieve this cycle time requires well-designed and integrated
equipment both upstream and downstream from the reactor. This is described in Section
VII.A.2.
2. Process Equipment: A process block flow diagram for a suspension polymerization facility is
depicted in Figure 12a [49]. The equipment flow diagrams for this process are depicted in
Figure 12bd. Figure 12b is the polymerization and stripping section. The hypothetical plant
design would have a 45,000-gal reactor for each 50,000-gal blowdown tank and stripping

column. Figure 12c is the dewatering and drying section. Figure 12d is the monomer recovery
system. The battery limits for this presentation do not in-

Page 73

Water
Vinyl chloride
Suspending agent(s)
Initiator(s)
Buffer salts
Temperature

Charge and heat up

Polymerization
Drop and vent
Wash and prorated antifouling treatments

24,000 gal
146,000 lb
60 lb/30 lb
80 lb
60 lb
56C

137
100
0.040/0.020
0.051
0.041

50100
280400
4070
50100
_______
420670

clude the environmental treatment facilities or utility distribution facilities, which will be similar for
most chemical manufacturing sites. The site depicted in Figure 12bd includes the following
typical equipment:

Reactor
Blowdown vessel
Stripping tower
Blend tank
Centrifuge
Fluid-bed dryer

45,000 gal
50,000 gal
700,000 lb/day
40,000 gal
700,000 lb/day
700,000 lb/day

This equipment configuration will produce, depending on product mix and the frequency of
product changes, up to 250 million pounds of product per year. This review will discuss only
polymerization and monomer stripping of a commercial operation.

2
2
2
5
2
2

Page 74

FIGURE 12 Polyvinyl chloride by suspension process: (a) block flow diagram; (b)
polymerization and stripping; (c) dewatering and drying; (d) monomer recovery.

Page 75

FIGURE 12 (Continued)

Page 76
B.Resin Productivity
The polymerization of vinyl chloride is an exothermic reaction, releasing 25 kcal per gram mole
of polymer formed. The reaction depicted in (34) is considered the model for the
polymerization. All other reactions are considered to be minor components in the total
exothermic process. A commercial process similar to that represented in Table 3 will release
approximately 90 million Btu of heat at 86% conversion.
To keep the contents of the reactor at isothermal conditions the internal cooling surfaces of the
reactor must be able to remove the heat of polymerization as fast as it is produced. The sizing
of the reactor cooling surfaces as well as the average temperature of the cooling medium will
affect the capital cost of the manufacturing facilities. The ideal situation would be to have a
constant rate of polymerization and therefore a constant heat load to the reactor cooling
surfaces. The usual practice results in a nonideal situation with a nonuniform rate of
polymerization and a heat peak at the end of the polymerization. These two cases are depicted
in Figure 13. In both cases the same quantity of heat is released; the difference in time and the
peak heat load causes the nonideal system to be a more expensive process than the ideal
system.
The techniques for achieving a uniform rate of polymerization and therefore uniform heat
generation are considered in Section VII.B.1. The techniques for heat removal are discussed in
Section VII.B.2.
1. Heat Generation Mechanism: The model for the polymerization of vinyl chloride has been
described in Section IV.B. This model indicates that the rate of polymerization is related to the
rate of formation of free-radical species at the site of polymerization up to the limit of the
presence of a free monomer phase (xc). Further modifications have been made to this model in
the region where there is no free monomer phase (x > xc). Chan et al. [50]

FIGURE 13 Heat load vs. time. , Ideal constant polymerization system; ---, nonideal.

Page 77
have further modified this model to account for the monomer in the vapor phase and its effect
on the heat generation mechanism.
In the second stage of the polymerization the concentration of monomer dissolved in the
polymer remains constant (Section IV.B.1). To achieve a uniform polymerization rate, the
concentration of free-radical species should also be constant. Since the initiators decompose in
an exponential manner [equation (11)], a combination of initiators such that Ri is constant in the
polymer particles during the second stage of polymerization (~0.1 to 77%) will be necessary for
an ideal constant rate of polymerization system. The identification of the optimum combination
of initiators that will give a uniform rate of polymerization depends on understanding the firstorder decomposition of the initiators under different process conditions.
a. Initiator Structure and Radical Formation: The kinetic aspects of thermal decomposition of
the initiator into free radicals were discussed in Section III.A. The nature of the groups (R1 and
R2) will have a major impact on effectiveness of the initiator in the areas of half-life
temperature, water solubility, and resistance to hydrolysis.
Half-life temperature: Pauling [8] has indicated that the bond energy for homolytic scission of a
peroxy bond is 25 to 40 kcal/mol. During the past two decades commercial manufacturers of
initiators have examined both peroxy and azo compounds as initiators for polyvinyl chloride
manufacture. In the United States the peroxy compounds have tended to dominate the market
because of the variety of structures and therefore activity of the peroxy initiator which are
available. In addition, the peroxy initiators do not have the toxicological problems associated
with the azonitriles. Azobisisobutyronitrile has been reported to form tetramethylsuccinonitrile
[51]. The resins prepared from azobisisobutyronitrile initiators also appear to have a tendency
to exhibit a color development under resin conversion procedures and there has been an
indication that the reactor wall fouling is higher with the AIBN initiator. These factors have
tended to limit the use of azo initiation for suspension processes.
Private communications with practioners of Shinetsu suspension technology indicates that the
use of azo initiators are not detrimental to the process or product. These findings have not been
quantified in the open literature.
The peroxy compounds used in the manufacture of polyvinyl chloride fall into the following
classifications:

Page 78
TABLE 4 Effect of Substituent in Peroxydicarbonate Initiator

Page 79
Other initiators, such as -chloroacyl peroxides [52], peroxyester carbonates [53], and
acetylsulfonyl peroxides [54] have some use in the manufacture of suspension-grade polyvinyl
chloride.
In Section III.A the relationship of decomposition constant (Kd) and half-life (t1/2) was
developed for first-order kinetics of initiator decomposition. Initiator decomposition is usually
tabulated by 1-hr and 10-hr half-life temperatures, that is, the temperature at which half of the
initiator will be consumed in either 1 hr or 10 hr. The effects of the R1 and R2 groups on 1-hr
half-life temperature (t11/2) and 10-hr half-life temperature (t101/2) have been determined by
Kamath and coworkers [55] and are detailed in Tables 4 to 6. Their results indicate the
following:
1. For peroxy dicarbonates
(Table 4) the presence of either one or two alkyl groups
on the -carbon of the R group has no effect on activity. All peroxydicarbonates with two alpha
substituents could be expected to have the same activity. At the present time there are no
known trisubstituted peroxydicarbonates.
TABLE 5 Effect of Substituent in Diacyl Peroxide Initiator

Page 80
2. For diacyl peroxides
(Table 5) the unsubstituted compounds and mono-substituted
compounds on the -carbon of the R group have about the same activity as initiators. The
disubstituted form of the diacyl peroxide is extremely active as an initiator. The trisubstituted
form has not been prepared for commercial use.

3. For peroxy esters

(Table 6) trisubstitution on the -carbons of R1 and R2 can be
controlled. The steric effects of the R1 and R2 groups can be used directly to affect the activity
of this class of initiators.
The activity of the peroxy esters and the diacyl peroxides are affected by the number of
substitutions at the -carbon of the R group. By comparison the peroxy dicarbonates do not
exhibit this steric effect because the oxygen-oxygen bond is farther away from the -substituted
carbon.
Verhelst [56] has examined the effects of substitution on the -carbon of the R group of diacyl
peroxides and found that increasing the number of methyl substituents affected the half-life at
60C in the following order:

The solvent used in the decomposition studies can affect the decomposition constant. The data
reported in Tables 4 to 6 use benzene and trichloro-

Page 81
TABLE 6 Effect of Substituent on Perester Initiator

ethylene. The results are consistent within each group, but it is difficult to compare initiator
decomposition data with two different solvents. The solvent can affect decomposition of the
peroxy bond by lowering the energy needed to break the oxygen-oxygen bond through radicalinduced decomposition. This peffect of solvent type on the 10-hr half-life temperature of
diisopropyl peroxydicarbonate (IPP) is detailed in Table 7. The trichloroethylene acts as a
radical scavenger and the induced decomposition of the IPP is largely liminated. This results in
a higher temperature and illustrates the solvent effect on t1/2.
Water solubility and resistance to hydrolysis: The effectiveness of the initiator depends on its
ability to generate active radicals at the site of polymerization (i.e., the swollen vinyl chloride
monomer phase). Any side reaction that removes the initiator from this site will reduce the
initiator's effectiveness and will lead to undesired side reactions, such as increased fouling; this
will be discussed in Section VII.B.2.b. Verhelst [56] has measured the water solubility of a
group of peroxy initiators and these data are detailed in Table 8. They indicate that generally,
the bulkier the R group, the lower the water solubility of the initiator.

Page 82

In a separate study [55] the hydrolytic stability of different initiators was examined under
polymerization conditions. The peroxydicarbonate compounds are sensitive to base hydrolysis.
A di(sec-butyl) peroxydicarbonate initiator system showed an increase of 60 min (from 330 to
390 min) in polymerization time when the pH of the aqueous phase was increased from 4.0 to
10.0. Increasing the bulk of the alkyl group appears to reduce this problem. The di(2-ethyhexyl)
peroxydicarbonate under the same pH range shows an increase in polymerization time from
280 min to 300 min. The peroxy esters are unaffected by base hydrolysis in the pH range 7 to
10.
b. Initiator Systems for Uniform Heat Generation: The advantages of a uniform heat generation
polymerization system (UHGPS) are shorter polymerization time coupled with reduced capital
cost for cooling capacity. No single initiator has been identified which is effective as a UHGPS
initiator over the commercial range 45 to 75C. The patent literature indicates that there are
combinations of initiators that will be effective as UHGPS over narrow temperature ranges. A
segment of this patent literature is noted in Table

Dilauroyl peroxide
t-Butyl peroxypivalate
t-Butyl peroxyneodecanoate
Dicetyl peroxydicarbonate
Bis(4-t-butylcyclohexyl) peroxydicarbonate
Bis(2-ethylhexyl) peroxydicarbonate

<10
2000
140
350
25
1000

Page 83
9. In addition, several U.S. initiator manufacturers recommend multiple initiator systems for
UHGPS [57]. This open literature offers few insights as to the rationale for selection of initiator
combinations for UHGPS.
Many of the patents in this area of UHGPS also claim such additional features as improved
thermal stability and/or reduced wall fouling. These claims may be due, in part, to other
features, such as suspension agents and/or buffer salt systems used in the polymerization.
There is no easy way to determine the interactive effects in these patents.
The initiators in these patents and commercial publications appear to be combinations of at
least a pair of initiators, one with a t1/2 range 1 to 2 hr and one with a t1/2 of 4 to 6 hr at the
polymerization temperature. The available commercial data for initiator decomposition is highly
dependent on the solvent used in the study. In addition, the commercial data have been
developed in a relatively concentrated solution compared to a commercial polymerization. There
are two approaches available for developing initiator combinations which will give uniform
polymerization rates:
1. Examine multiple combinations of initiators using an evolutionary operations procedure in
production-scale reactors.
2. Determine the Kd and Ed for several initiators in laboratory-scale polymerization at several
different temperatures and fit the data to a kinetic curve similar to equation (32).
The advantages of the second approach are that with accurate Kd and Ed data, computer
simulations of the effects of variations in initiator ratio as well as variations in polymerization
temperatures can be predicted. From these predictions it is possible to develop useful UHGPS
initiator combinations. In addition, the effects of temperature staging can also be analyzed using
the Kd and Ed data.
c. In Situ Initiators: The initiators described thus far are manufactured commercially by a
Schotten-Baumann procedure [58] and are used by polyvinyl chloride manufacturers as either
pure materials or diluted in the range 30 to 75% with an organic solvent. An alternative
approach is to prepare the initiator species in the polymerization reactor by an in situ process.
The advantages of the in situ process are that it eliminates the need to store and handle
peroxides and it offers initiator structures which are not available commercially. The
disadvantages of the in situ process are that it introduces other toxicity hazards and may affect
particle morphology.
Interest in situ initiators was very high in the early 1960s. At that time the active initiators were
limited to the lower alkyl peroxydicarbonates. Diisopropyl peroxydicarbonate (IPP) initiator, with
a self-accelerated decomposition temperature (SADT) of 30F (in 1-gal containers), was
difficult to handle [59]. This low SADT meant that the IPP had to be stored and transferred at
temperatures no higher than 30F. An alternative was to prepare the IPP initiator by an in situ
process. In the in situ process, isopropylchloroformate was dissolved in the monomer phase

and hydrogen peroxide and sodium hydroxide were dissolved in the aqueous phase. Mixing the
two phases resulted in the formation of the IPP initiator in the monomer phase. Combinations of
acid halides, alkylchloroformates, and alkylsulfonyl chlorides could

Page 84
TABLE 9 Initiator Systems for Uniform Heat Generation Polymerization Systems

Page 85

Page 86
also be used, and this would result in a rather wide range of mixed products. This can be
illustrated by the process in Figure 14. Mixing an acid chloride and a chloroformate in the
monomer phase would result in a potential initiator mixture composed of reaction products 1a1,
1a2, and 2a2. The exact amounts would depend on the acid chloride/chloroformate ratio and
the reactivity in a Schotten-Baumann reaction. The wide variety of products makes this in situ
process quite flexible in the type of initiator species that can be formed.
Manner [60, 61] and Barter [62, 63] of PPG have synthesized the mixed peroxides of isobutryl
peracetate and isobutryl peroxyethyl carbonate, and the symmetrical diisobutryl peroxide.
Langsam [6466] of APCI has prepared a wide range of mixed sulfonyl peroxides. These in situ
initiators have exhibited a tendency to increased fouling in laboratory-scale reactors [63, 65, 66]
and exhibit a much broader particle-size distribution with a cellulose suspension system.
The handling problems associated with low-SADT initiators such as IPP have been eliminated
by the synthesis of higher-molecular-weight peroxydicarbonates such as di(4t-butyl cyclohexyl)
peroxydicarbonate, which is a crystalline solid with an SADT > 75F, and with the recent advent
of frozen emulsions of peroxydicarbonates [67]. The use of these higher-preformed peroxy
dicarbonates is both safe and efficient. The major advantages of the in situ initiator system is
the synthetic flexibility in preparing mixed initiator structures and the control this technique could
have over the rate of generation of initiator species in the reactor. A process whereby all the
organohalo moiety is added at once to the monomer phase and the hydrogen peroxide/sodium
hydroxide is added in a controlled monomer over the polymerization would result

FIGURE 14 In situ initiator process.

Page 87
in a controlled generation of the initiator, which, in turn, would affect the polymerization rate.
The accurate determination of polymerization kinetics depends on the ability to measure a
change in the reaction mixture and to relate this change to the conversion of monomer to
polymer. In a homogeneous system it is possible to measure such parameters as solution
viscosity or volume contraction and to relate these to polymer formation. The heterogeneous
nature of commercial polyvinyl chloride manufacture precludes these methods from normal use.
Total heat evolution from a well-insulated reactor has been the major technique used to
determine the polymerization rate from laboratory-scale studies. This procedure is difficult to
run in small reactors because of the heat losses from uninsulated reactor surface/interface
(reactor agitator shaft, support brackets, etc.). Gas chromatography procedures offer an
alternative method for measuring polymerization kinetics [67a,b]. The gas chromatography
procedure was described by Kissin and Beach [67a] as an experimental method to measure
the reactivity ratio for Ziegler-type copolymerization of ethylene propylene. In this gas
chromatography method, the change in the vapor-phase monomer composition is related to the
rate of monomer consumption and therefore to the reactivity ratios. This gas chromatography
procedure can be used to measure the rate of vinyl chloride disappearance in the vapor phase
of a reactor provided that the vinyl chloride can be compared to an internal standard material.
One method that can be used is to add a small quantity of a nonreacting substance with similar
properties to vinyl chloride and to measure the change in the vapor-phase ratio of vinyl chloride
to the nonreacting substance. The change in the vapor-phase ratio of monomer/internal
standard is a measure of the rate of disappearance of vinyl chloride in the liquid phase and the
rate of polymer formation. The advantage to this gas chromatography procedure is that heat
balance for the reaction vessel need not be measured and therefore no elaborate insulation
procedures are needed. Langsam [67b] has reported that the gas chromatography technique
using n-butane/vinyl chloride is accurate to within 1% conversion compared to a weight
recovery technique. The procedure is easily applied to larger reactors and provides a means of
determining the conversion without resorting to heat balance measurements.
2. Heat Removal Process for Suspension Polymerization: The formation of polyvinyl chloride is
an exothermic reaction. One of the functions of the reactor vessel is to provide the cooling
surfaces for heat removal. The reactor size is limited by the ratio of reactor volume to cooling
surface and the agitation system needed to maintain the monomer-water mixture at the proper
droplet size. Commercial reactors of 25,000- to 50,000-gal capacity are standard for the
suspension polymerization industry. The reactor productivity will depend on the heat transfer
capacity of the vessel: The higher the heat transfer capacity, the more polymer that can be
made per unit of time. The heat transfer capacity of a reaction vessel can be calculated using
the following relationship:
Q = UA(T)
[39]
where Q is the total heat transfer capacity in Btu/hr, A the area of cooling surface in ft2, (T)
the temperature differential between the reactor contents and the fluid in the jacket in F, and U
the heat transfer coefficient of the cooling surface in Btu/ft2-F-hr.

Page 88
Increasing any of these factors will increase the Q for the reaction vessel. Efforts to maximize
the heat transfer capacity of the reactor vessel have been directed at reactor configurations,
reactor wall fouling reduction, reactor cleaning procedures, and temperature-step processes.
a. Reactor Configuration: During the past 10 years large stainless steel reactors equipped with
reflux condensers and multiple turbine agitators have been developed for the manufacture of
polyvinyl chloride resin. This evolution in reactor configuration is depicted in Figure 15. The
efforts to increase the Q for large reactors have focused on the following factors: cooling
surface area/volume ratio, materials of construction for the wall, reflux condenser utilization,
and temperature of the cooling fluid in the jacket.
Cooling surface area/volume ratio: As the size of a commercial reactor is increased, the
surface area/volume ratio decreases. Albright and coworkers

FIGURE 15 Polyvinyl chloride by suspension process: reactor types.

Page 89
[6870], in a computer simulation study, have related the factor of reactor volume,
polymerization cycle time, monomer loading, and heat transfer coefficient to the cooling-water
temperature in the reactor jacket. Based on their computer simulation, they have identified a
safe operating envelope for the polymerization of vinyl chloride. Their data were based on a
single initiator system (IPP) and indicated that at the following process conditions:
Reactor size
20,000 gal
Polymerization time
5 hr
Heat transfer coefficient
75 Btu/hr-ft2-F
Water/monomer ratio
1.5:1
the polymerization could not be completed because of inadequate heat transfer capacity [70].
The use of reflux condensers and UHGPS initiators have extended the safe operating envelope
[69]. Burgess [71] has calculated the effect of increasing reactor volume on the surface area
available for cooling. This is detailed in Table 10. His calculation is based on a reactor with a
2:1 length/diameter ratio. The surface area increases by a factor of 20 as the volume increases
by a factor of 100. In addition, the wall thickness increases by a factor of 2 for the reactor as
the volume increases from 2500 gal to 25,000 gal.
In actual practice, the scale-up of reactor vessels uses the length/diameter ratio as a variable.
This is illustrated in Table 11. The reactors in this analysis vary in length/diameter ratio and the
agitation system is designed so that the agitator tip speed falls within a narrow range. The
reactors detailed in Table 11 will produce a resin with the same particle properties using the
same process formulation.
In a recent patent Wacker Chemie [71a] described the internal configurations of large reactors
(13,000 to 50,000 gal). The patent indicated that it was not necessary to have a reflux
condenser for short polymerization times. The examples given with the patent were not broad
enough to examine their claims fully.
Materials of construction for the wall: The use of stainless steel clad over carbon steel has
become an accepted construction mode for the wall of a reactor [72]. The stainless steel
(either as 316 or 304 type) provides the corrosion resistance together with a high heat transfer
coefficient. Typical values for such a wall configuration are 75 to 105 Btu/ft2-F-hr. By
comparison, a glass-coated carbon steel reactor wall has a heat transfer coefficient of 40 to
50. The film-side heat transfer coefficient of the reactor walls
2) Surface area/ 1000 gal 250 5.2 2.6 42.4 170 2,500 11.2 5.6 197.0 79 25,000 24 12.0 904.3
36

Page 90

Length (ft)
Diameter (ft)
L/D ratio cooling
Surface area (ft2)
Jacket
Reflux condenser
Agitator tip speed (ft/sec)

10
8
1.25:1

13.3
9
1.48:1

27.9
13
2.15:1

200
28.3

300
32.1

900
900
33.3

changes with conversion [73] as the viscosity of the polymerization system increases.
Reductions as high as 50% have been reported for the heat transfer coefficient of the walls.
Reflux condenser utilization: A reflux condenser functions by condensing vapor to liquid and
removing the latent heat of vaporization from the system. In large reactors, a reflux condenser
is the most effective way to increase the available cooling surface without increasing the
length/diameter ratio of the reaction vessel. The use of a reflux condenser to provide additional
cooling capacity in vinyl chloride polymerization reactors was discussed by Terwiesch [74] for
reactors of 50,000-gal capacity. A Conoco patent [75] describes the installation and use of a
reflux condenser. In a recent patent, Shinetsu [76] indicates that the reflux condenser should be
used only after 5% of the monomer is converted to polymer, because utilization of the reflux
condenser before 5% conversion results in extensive reflux condenser fouling and a coarser
resin.
The reflux condenser operation envelope (amount of reflux, startup time for reflux) will depend
on the ability of the reactor agitation system to reincorporate the refluxed monomer effectively
into the polymerizing mass, and on the foaming tendency of the polymerization recipe. Venting
procedures have been described [76a] to remove the noncondensable gases from the vapor
phase of the reactor and reflux condenser. The level of noncondensable gases in the reactor
vapor phase will be affected by the quality of the monomer, how well the reactor has been
evacuated prior to polymerization, and whether the polymerization process generates an inert
gas. The use of azo initiators will result in the formation of nitrogen as a by-product of the
initiator decomposition. In addition, a carbonate buffer system will produce CO2 gas if the
aqueous phase becomes acidic. Both situations will result in the generation of inert,
noncondensable gases which will reduce the efficiency of the reflux condenser.
Temperature of the cooling fluid: Increasing the temperature differential between the fluid in the
reactor jacket and the slurry will increase Q, the heat transfer capacity, for the reactor vessel.
The fluid in the reactor jacket

Page 91
can be either all cooling tower water or all refrigerated water. If the following assumptions are
made as to the temperatures of the materials:
Reactor slurry temperature
136F
Cooling-tower-water temperature
86F
Refrigerated-water temperature
66F
then the change in the T factor for these two cases will result in a 40%* increase in Q for the
refrigerated-water case. The cost associated with refrigerated machines compared to cooling
towers needs to be determined for each specific manufacturing situation.
b. Reactor Wall Fouling Reduction: A characteristic of most polymerization processes is the
development of reactor wall fouling on the inner surfaces of the reaction vessel. The wall fouling
reduces the heat transfer coefficient for the cooling surfaces and, in the form of large chunks of
polymer, causes difficulties in the material handling of the resin in the process drying train. In
addition, the presence of wall fouling reduces the product quality of the resin.
The present state of the art in suspension polyvinyl chloride manufacture has not reached a
point where all fouling can be eliminated from all reactors. The current understanding of the wall
fouling mechanism is that wall fouling is due to an instability in the formation and polymerization
of the monomer droplets. Some causes for this instability are:
Surface phenomena of the droplet
Agitation and cooling-system operation
Surface phenomena at the reactor walls
Aqueous-phase polymerization
A review of the patent literature indicates that there is a rather wide range of proposed
remedies for fouling in suspension PVC reactors. An analysis of these patent citations indicates
that the deformation of the polymerizing monomer droplet against the reactor wall or other
physical obstructions leads to an adhesion of the droplet to the surface. The subsequent
adsorption of additional monomer and initiator causes the development of wall fouling. The
technical approaches to preventing wall fouling are: prevent the deformation of the monomer
droplet; prevent the adhesion of the polymer to a reactor surface; prevent the adsorption of
monomer to a reactor surface; and prevent the free-radical polymerization either at the wall or
in the aqueous solution. The mechanism for wall fouling prevention is principally concerned with
both surface phenomena and reactor operations.
Surface phenomena of the monomer droplet: The monomer droplet is coated with a suspending
agent that functions to prevent the agglomeration of the droplets and to aid in the formation of
the internal pore structure of the resin particle. The strength of the monomer-water interface,
which can be measured by either surface tension [77] or the gel strength of the interface [78],
will affect the level of wall fouling. As a general trend it has been observed that suspension

systems with higher interfacial tensions and higher

*

Page 92
gel strengths tend to give lower levels of fouling. The trend to using suspension systems with
high interfacial tension has to be balanced against the effect of the suspension system on the
internal porosity of the resin; higher interfacial tension usually results in lower porosity.
Agitation system operation: The suspension droplet is an easily deformed elastic body in
constant motion within the reactor. The agitation system, which is composed of rotating agitator
and fixed baffles, serves to ensure steady heat transfer through mixing and to develop and
maintain the size distribution of the monomer droplet in the water. As a trend it has been
observed that structures within the reactor which lead to fluid dynamic flow turbulence will
increase reactor wall fouling. Therefore, baffle support braces need to be designed for the
optimum fluid dynamic stability.
Cooling system operation: The reactor jacket serves the dual purpose of first heating the
reaction mixture to polymerization temperature by passing a high-temperature fluid through the
jacket and then maintaining polymer slurry at a fixed temperature by passing a low-temperature
fluid through the jacket. In large reactors additional cooling is available through the use of a
reflux condenser. In a batch polymerization process the reaction mass is heated as quickly as
possible to the polymerization temperature so as to increase the productivity of the reactor.
This requires that steam-heated water at about 180 to 190F be circulated in the jacket. It has
been suspected that this type of heat-up process leads to increased wall fouling. An alternative
approach would be to use preheated process water [79]. A second alternative would be to
sparge steam directly into the reactor [80]. Direct sparging into a reactor has many problems
associated with product contamination and is not widely practiced. A third approach is to heat
the reactor adiabatically by means of a very active initiator system [81, 82]. The advantages of
adiabatic processes will be discussed in Section VII.B.2.d.
The operation of a reflux condenser within certain process envelopes can lead to increased
fouling. Recent work by Shinetsu has indicated that the operation of a reflux condenser during
the particle formation period (0 to 5% conversion) can lead to a significantly coarser resin [76].
This has been confirmed by APCI laboratory-scale reactors using internal vapor-phase coils. By
contrast, Conoco [75] has claimed that the use of a reflux condenser for all of the heat removal
during polymerization has no effect on either the resin particle properties or reactor fouling. The
apparent conflict in the results reported by Shinetsu and Conoco may be due to differences in
the design and operation of the agitator system or due to differences in the suspension system.
There is no easy way to identify which is the cause for these differences.
Surface phenomena of the reactor walls: The droplets of monomer in a reaction mixture are in
an equilibrium between the vapor phase, the aqueous liquid phase, and the wall of the reactor.
Electron photomicrographs of a sample of 316 stainless steel show numerous microcracks in
the metal surface (Figure 16). In addition, the metal surface is not monolithic, but is made up of
numerous grain boundaries where there are discontinuities in the metal surface. The
microcracks in the surface of the metal can act as adsorption sites for immobilized free radicals
or other high-energy species. Similarly, the boundary between the crystal groups in the metal
surface can act as an electrochemical discontinuity. Therefore, process changes that reduce

the adsorption of monomer on the walls or deactivate immobilized high-energy species on the
walls will reduce the formation of fouling. The adsorption of

Page 93

FIGURE 16 Scanning electron photomicrograph of 316 stainless steel at 2000 magnification.

vinyl chloride onto the reactor wall can be retarded by treating the reactor wall with a highly
polar, hydrophilic material. This approach has been used by Air Products with a caustic wash
system [83, 84], and by Dow Chemical [85] with a sulfonation process. The process practiced
by Air Product has achieved a level of 45 batches between solvent cleaning in a 24,000-gal
reactor. Both approaches apparently cause the formation of a highly polar hydrophilic surface
on the reactor walls. The use of a caustic wash system may also have added benefits in
improving the thermal stability of the resin during the stripping and drying process. A summary
of these technical approaches is detailed in Table 12. Reactor wall defects can act as a site for
immobilized high-energy species. This can include immobilized free radicals or surface
discontinuities which lead to microscale electrochemical potentials. Hydroxy functional aromatic
compounds which are known free-radical scavengers can be used to destroy these high-energy
species. The technical challenge is to immobilize these free-radical scavengers on the walls of
the reactor and to maintain the radical activity in the swollen polymer droplets. An approach
that has been successful has been to coat the reactor walls with a solution of a base-catalyzed
condensation polymer of a substituted aromatic phenol. Polymers of this type are coated on the
walls of a reactor at high pH, and are then precipitated on the walls by a water wash at pH of
about 7 to 8. The coating remains on the reactor wall as long as the aqueous pH remains below
pH 10. Another approach is to coat the walls of a reactor with a mixture of a metal salt which
will complex with a cross-linkable copolymer. A system of this type has been described in a
B.F. Goodrich patent [86]. Shinetsu [87] has a series of patents based on combinations of
organic dyes which are fixed on the

Page 94

Air
May
Products 1984
Air
Sept.
U.S. 4,224,424
Products 1980
Air
Oct.
U.S. 4,359,560
Products 1982
Dow
Nov.
G.B. 1,534,245
Chemical 1978
Jap.Kokai 54
Feb.
Kanegafuchi
21, 484
1979
U.S. 4,451,625

Use of NaOH solution and additives and a bake cycle at

6080C for 1030 min
Use of Na2CO3 or NaHCO3 solution and a bake cycle of
80C for 5 min
Use of NAOH solution and a bake cycle of 6080 for 10
30 min
Sulfonation of the reactor surfaces with anhydrous SO3
Coating the reactor walls with a sulfonic acid polymer

Page 95
walls of a reactor and inhibit the formation of wall fouling. Table 13 contains a summary of
patents in this technology.
The interaction between agitation system, suspension system, and the wall coating system
described in Tables 12 and 13 are difficult to quantify. There is probably s synergism implied
but not quantified in each of the patents which would make the transfer of a particular wall
treatment system into a different system of agitation and suspension agents very difficult.
Aqueous-phase polymerization: The solubility of vinyl chloride monomer in the aqueous phase of
a suspension polymerization system is about 0.4 wt% at 60C and 150 psi pressure. This
monomer can be polymerized in the aqueous phase by any free radical generated in the
aqueous phase. The polymer produced in the aqueous phase will not have a protective colloid
coating and could precipitate on the walls of the reactor. A technical approach to reducing this
mode of wall fouling is to use an aqueous-phase free-radical scavenger or a chelating agent to
eliminate any redox reaction between initiator in the aqueous-phase and transition metal ions. A
summary of the patent citations in this area is detailed in Table 14.
c. Wall Fouling Removal: The antifouling systems described in Section VII.B.2.b have not
reached the technical maturity where all wall fouling can be eliminated in polymerization
reactors. Some form of wall fouling removal is needed on a periodic basis. The frequency of
reactor cleaning will have an effect on the overall process economics. This is illustrated in
Figure 17, where the prorated reactor cleaning time is plotted against number of batches
between reactor cleaning. When the antifouling system is of low efficiency, there is a need to
clean the reactors frequently and the costs for wall fouling removal, in the form of contribution
to the reactor polymerization cycle, are high. As the technology for wall fouling prevention
improves, the number of

FIGURE 17 Clean time/batch vs. number of batches/cleaning cycle.

Page 96

Ger. Offen.
2,440,434
U.S.
4,024,330
U.S.
4,080,173
G.B.
1,439,339

B.F.
Mar.
Goodrich1975
B.F.
May
Goodrich1977
B.F.
Mar.
Goodrich1978
June
ICI
1976

A cross-linked polymer composed of polyethylene and

formaldehyde coated on the reactor walls with SnCl2
Polyaromatic amine cross-linked with m-phenylene diamine to
form a coating on the reactor
Resorcinol self-cross-linked with NaOH to MW 1000 and coated
on the reactor walls
Reaction of polyethyelene imine and formaldehyde in situ on the
reactor surface

Page 97
U.S.
ICIJan.
4,181,787Austrialia1980
Jap.
Mitsui Sept.
Kokai
Toatsu 1980
U.S.
Aug.
Shinetsu
4,105,838
1980
G.B.
Jan.
Shinetsu
1,559,334
1980
U.S.
Oct.
Stauffer
4,294,945
1981

Reaction of polyethylene imine and hydroquinone in caustic system

Condensation of humic acid with aromatic aldehyde compound in
caustic system
Coating the reactor with an organic dye and then reacting the dye with
a metal salt and heating at 40 for 2 hr
Combination of two dyes: (1) nitrogen-containing base dye and (2)
sulfonic acid-containing dye, coated and reacted on the walls
Dioctyltin mercaptoacetic acid

Page 98

Ger. Offen.
Oct. Aminocarboxylic acids and polyamino/polycarboxylic acids added to the
BASF
2,810,050
1979aqueous phase of the reactionlevels of 0.10.3 phm
U.S.
Dec. Salts of oxalic acid at levels of 1 x 10-3 to 1 x 10-1 phm suppress
Chisso
3,997,707
1976fouling for up to five batches
Jap. Kokai
Combinations of oxalic acid salt and metal nitrites at levels of 2 x 10-3
Oct.
52128,
Chisso
-10-1 for oxalate and 10-510-3 for nitrite suppress fouling for >10
1977
986
batches

Page 99
U.S. 3,962,196
U.S. 3,925,910
Jap. Kokai 5176,
379
Jap. Kokai 52
980080
Jap. Kokai 5238,
320
Jap. Kakai 528089

June
1976
Dec.
Goodyear
1975
Conoco

Mixtures of sodium salt of EDTA and benzotriazole

NaNO2 at 10-210-1 phm

KanegafuchiJuly 1976 Dithiocarbonic acids at 10-610-2 phm

Aug.
Kanegafuchi
1977

Alkali metal salts of citrazinic acid at 10-410-2

phm

Shinetsu

Jan. 1977Alkali metal bisulfite at 0.1 phm to reduce O2 level

Shinetsu

Jan. 1977Vitamins A2, B1, B2, K1 at 10-410-2 phm

Page 100
batches between reactor cleaning operations increase. An asymptotic value is approached
which represents the technical stage at which additional efforts in developing improvements in
wall fouling prevention technology are no longer cost-effective.
The technology used to remove wall fouling has changed during the past decade because of
the increased size of the polymerization reactors. Although it was possible to clean a 4000-gal
reactor manually with some internal scaffolding, manual cleaning of a 24,000-gal reactor would
require major scaffolding in the reactor and the associated problems of scratching the reactor
surface. The alternatives to manual cleaning are high-pressure water spray cleaning and
solvent cleaning.
High-pressure water spray: Water at pressures of 3000 to 14,000 psi can be used to dislodge
the wall polymer on the inner surface of the reactor [74]. The high-pressure water physically
breaks up the polymer film into pieces small enough to be flushed out through the bottom valve.
The effectiveness of a water spray system depends on the internal configuration in the reactor,
and the number and placement of the baffles and agitator blade. Reactor 2 (in Figure 15) has a
single marine impeller and no baffles. This reactor would be quite adaptable to a high-pressure
water spray system. Reactor 3 (in Figure 15) has internal baffles and multiple impellers which
require a highly articulated spray nozzle and boom. The effectiveness of a spray nozzle
decreases sharply as the distance from the nozzle to the sprayed surface increases; therefore,
a 25,000-gal reactor with a diameter of 12 ft will require a different high-pressure spray system
than a 2500-gal reactor with a diameter of 5.6 ft. High-pressure spray systems have been
developed by Artisan Industries and by Sugimo Machine of Japan for reactor cleaning.
Solvent cleaning: The use of solvent to clean a reaction vessel is commonly used. The optimum
solvent will have the characteristics of solubility coefficient identical to polyvinyl chloride [88],
high solubility for polyvinyl chloride, low vapor pressure, low toxicity, and stability to oxygen and
thermal-hydrolytic decomposition.
The solubility coefficient for polyvinyl chloride is reported to be in the range 8 to 12 [89]. The
solvents with similar solubility coefficient are described in Table 15 together with their physical
properties [90]. Solvents such as p-dioxane and tetrahydrofuran are prone to peroxide
formation and are hazardous to use in a plant operation. Ethylene dichloride and methyl ethyl
ketone are low-boiling solvents which tend to form vapor pockets at high temperatures (>85C).
Solvents such as cyclohexanone, dimethylacetamide, dimethylformamide, dimethyl sulfoxide,
and N-methyl pyrrolidone are, in overall characteristics, good choices to remove fouling.
In a solvent cleaning operation the water in the reactor should be reduced to as low a level as
possible by a combination of heating and vacuum. The solvent, preheated to about 175F, is
pumped into the reactor and circulated for 8 to 12 hr. The reactor is them emptied and flushed
sequentially with clean solvent and water. The solvent efficiency at a given process temperature
depends on the amount of dissolved polymer and water in the solvent. The typical upper limits
for dissolved polymer and water in dimethyl formamide solvent are about 5% polymer and
about 2.5% water. Above these levels the solvent is inefficient and needs to be purified.

Air Products [91] has developed an efficient dimethyl formamide recovery system which has
been used in a production system since 1977. A solvent system based on dimethylformamidetoluene has been cited [92] as being

Page 101
a (C) Solubility coefficient Cyclohexanone 98.15 155.6 2.3 8.0 147 9.9 Dimethyl acetamide
87.12 166.1 171 10.8 Dimethyl formamide 73.10 153.0 153 12.1 Dimethyl sulfoxide 78.1 189.0
25.3 180 12.0 p-Dioxane 88.1 101.3 36 10.0 Ethylene dichloride 99.0 83.4 0.81 0.15 15 9.8
Methyl ethyl ketone 71.2 79.6 24 10 0 9.3 N-methyl pyrrolidone 99.1 202 B - 180 11.3
Tetrahydrofuran 72.1 66 B - -10 9.1 aTag close cup.

Page 102
particularly effective for removing wall fouling. The choice of reactor cleaning solvent will
depend on the economics which are unique for each resin manufacturer. Those companies with
a captive source of ethylene dichloride from a vinyl chloride manufacturing process may find the
internal cost structure favorable for the use of this solvent. Similar situations may exist for other
solvent materials.
Temperature-Step Process: The polymerization processes described thus far in Section VII.B
have been isothermal with respect to the slurry temperature. To maintain an isothermal slurry
temperature the heat of polymerization cannot exceed the heat removal capacity of the reactor
vessel. An approach to increasing the total heat removal capacity of the reactor vessel is to
increase the temperature differential between the slurry and the jacket fluid [T term in
equation (39)]. In Section VII.B.2.a the case of using refrigerated water instead of coolingtower water to increase the T terms was discussed. An alternative approach to increasing the
T term is to increase the slurry temperature in a controlled manner. This technique has been
cited in a recent patent application to Kureha [81]. Hamielec et al [82] have extensively
reviewed the temperature-step process with regard to the effect of variation in slurry
temperature on the molecular weight of the resin, the amount of residual initiator at any time
(and therefore conversion), and the instantaneous rate of polymerization.
In Sections IV.A and IV.B the effects of diffusion control on both the polymerization rate
constant (Kp) and the termination rate constant (Kt) were discussed. The effects of
approaching a glassy state were described and it was noted that any increase in the
polymerization temperature would increase Kp (see Section III.B). In a simulation that used a
binary initiator and a heat-up driven by the heat of polymerization rather than high-temperature
fluid in the reactor jacket, it was shown that increasing the polymerization temperature after the
free monomer phase has disappeared (x > xc) would increase the polymerization rate (Rp) and
also increase the final conversion to ~91% and not affect the cumulative molecular weight of the
resin. The process characteristics for this simulation are detailed in Figure 18. Figure 18a
represents the slurry temperature as a function of time. The reaction requires about 1 hr to
heat from 32C to 52C. It is then held at isothermal conditions for 4 hr and then is heated to
66C. The polymerization is then held between 61 and 63C for about 1 hr and is then
terminated. The effects of this temperature profile on rate of polymerization and conversion as
a function of time (Figure 18b), residual initiator as a function of time (Figure 18c), and weightaverage molecular weight as a function of time (Figure 18d) are depicted in the indicated
illustrations.
Hjerberg and Sorvik [92a] have shown that polymerizations carried out at subambient pressure
(U polymerization) result in a polymer with increased thermal instability. Laboratory-scale
polymerizations have confirmed this in isothermal polymerizations [92b]. The depletion of the
vinyl chloride monomer in the temperature-step polymerization coupled with the increase in
polymerization temperature may result in a degradation in the thermal properties of the finished
resin.
The temperature-step process can occur in an unplanned fashion when, at the end of a

polymerization, the heat generated by the reaction exceeds the cooling capacity of the reactor.
If there is no pressure excursion (no free monomer phase, x > xc), temperature increases of
about 10 to 15C can be tolerated. The effects of these temperature excusions on the particle
morphology of the resin are not well quantified but are likely to be extremely

Page 103
bad, and therefore the temperature-step process should be carefully monitored in any
commercial operation.
A temperature excursion associated with a pressure excursion (x < xc, free monomer phase
present) has the potential for a runaway reaction and a release of monomer. This type of
polymerization must be quickly terminated with an effective short-stop reagent. A good
terminating agent will have the following characteristics: very soluble in the swollen polymer
phase, rapid reaction with the active chain end, and easy separation from the recovered
monomer.
The chain terminating agent will trap the free radical in a low-energy state where the free
radical is no longer able to initiate a new polymer chain. Aromatic hydroxy compounds are
known chain terminating agents. The effectiveness of the aromatic hydroxy compounds is due
to the fact that a free radical can be delocalized across several sites on the molecule and is
therefore at a lower energy state. -Methyl styrene and hydroquinone are some of the
materials that have been used as chain terminating agents in production facilities.
C.Resin Quality
The manufacturing facilities described in Section VII.A.2 will produce approximately 650,000 lb
of resin per day using the technology described in Section VII.B. This high productivity must be
coupled with resin of consistent and controlled quality for the operation to be successful
economically.
The quality of the resin produced in the process is a function of the process conditions and the
quality of the raw materials used in the polymerization. An understanding of the interaction
between process conditions and raw material quality is absolutely essential in producing a resin
of consistent quality. An efficient approach to controlling the product quality of a resin is to
develop a series of process control equations which relate the input parameters of process
variables and raw material quality to the output parameter of resin quality. Several statistical
procedures [93, 94] can be used to develop a data base and to reduce the data to a
mathematical equation of the form
y = f(x1,x2,x3, . . ., xn)
where y is the output parameter and x1 to xn are the process and raw material input
parameters. The choice of experimental design will determine whether the mathematical form
of the control equation is linear in structure [93] or whether interactive and quadratic terms are
also generated [94] in the control equation. A review of the applied experimental statistics
literature [95, 96] is recommended for addition details.
Commercial grades of polyvinyl chloride resin are classified by the following manufacturing
specifications:
1. Molecular weight

2. Particle morphology
3. Diffusion processes
a. Adsorption of plasticizer
b. Residual level of monomer
4. Bulk density

Page 104

FIGURE 18 (a) Slurry temperature vs. time; (b) conversion and instantaneous heat generation
as a function of time; (c) residual level of initiator as a function of time [Lucidol 223M di-2ethyl hexyl peroxydicarbonate; Vazo 33 2, 2-azobis(2, 4-dimethyl-4-methoxy valeronitrile)];
(d) molecular weight of resin prepared by adiabatic process.
Control of the resin quality depends on the identification and control of the process input
parameters. Broadly speaking, the molecular weight of the resin is controlled by the chain
transfer effects of the growing chain end, while the other manufacturing specifications are
related to the particle morphology of the resin grains. The remainder of this section is devoted
to a review of the known process conditions that control the manufacturing specifications for
polyvinyl chloride resin.

Page 105

FIGURE 18 (Continued)
1. Molecular Weight: The molecular weight of commercial grades of polyvinyl chloride will affect
both the fabrication technology and the mechanical properties of the finished resin. The end-use
applications require that a resin with a consistent molecular weight range be used in a particular
fabrication process. The molecular weight of the resin is usually reported as a dilute solution
viscosity measurement [97], as either inherent or reduced viscosity. The ability of a dilute
solution viscosity measurement to represent the resin molecular weight depends on the fact
that the molecular weight distribution for polyvinyl chloride is rather insensitive to the
polymerization conditions. Several excellent reviews of the relationship between dilute solution
viscosity and polymer structure are recommended for an in-depth understanding of this subject
[98, 99]. It is sufficient for this discussion to indicate that the dilute solution viscosity of the resin
can be related to the viscosity-average molecular weight by the graph in Figure 2 or by the
Mark-Houwink equation [99]. The constants for the Mark-Houwink equation will depend on the
solvent and temperature used in the determination. The Mark-Houwink constants for polyvinyl
chloride are detailed in Table 16. The molecular weight distribution of the resin can be
measured by gel permeation chromatography (GPC) procedures. The inherent viscosity and
GPC procedures should be calibrated

Page 106
[] = KMa
+1 (d1/g) a Molecular weight x 10-4 Cyclohexanone 20 11.6 0.85 2 20 13.7 1.0 7 20 112.5
0.63 9 25 12.3 0.82 2 25 24.0 0.77 3 25 204 0.56 2 25 174 0.55 6 25 8.5 0.75 4 25 13.8 0.78
1 30 16.3 0.77 3 Tetrahydrofuran 20 3.6 0.92 2 25 15.0 0.77 1 25 16.3 0.77 2 25 49.8 0.69 4
30 63.8 0.65 3 30 83.3 0.83 3 30 219 0.54 5 Source: Ref. 100.
against well-defined samples of polyvinyl chloride resin. The Mark-Houwink constants then can
be calculated from experimental data and compared to the literature values.
a. Control of Molecular Weight: The molecular weight of polyvinyl chloride resin is controlled by
the competitive reaction between the addition of monomer to the active chain end and the
transfer of the active chain end to a new site and the deactivation of the growing chain. These
reactions were detailed in Section III.D.2. The expression for number-average molecular weight
() [equation (31)] can be restructured in the form

and with the following assumptions:

Page 107
1. In a bulk Polymerization CM is a constant quantity.
2. 0 is the combination of the polymer and monomer transfer constant at a given temperature.
3. CI and [I] are very small and can be neglected.
Therefore, the approximation

can be made. A plot of 1/ versus [S] at fixed [M] is a means of determining the chain transfer
coefficient for a given material. The intercept 1/ is the limiting molecular weight that can be
achieved at a given temperature.
In commercial polymerization reactions, the pressure limitation of the reactor establishes the
maximum polymerization temperature and therefore the minimum molecular weight that can be
made in the reactor without special procedures. The typical maximum operating pressure of
reactors described in Section VII.A.2 is 170 psi, which translates to a working temperature of
65C. These manufacturing conditions will produce a resin with an inherent viscosity in the
range 0.78 to 0.75 dl/g. A chain transfer agent is needed to prepare a resin with a lower
inherent viscosity.
The factors that control the inherent viscosity of polyvinyl chloride resins are polymerization
temperature and the type and level of chain transfer agent. A control equation that predicts the
inherent viscosity for polyvinyl chloride resin will be in the functional form
A linear equation can be developed from the work of Ravey and Waterman [100a] to relate
inherent viscosity to the process parameters of temperature and chain transfer agent:

where Tabs is the absolute temperature and [S] is the weight percent of chain transfer agent.
The constants K1, K2, and K3 need to be determined for each chain transfer agent.
b. Chain Transfer Agent: A wide variety of materials have been reported to be effective chain
transfer agents for polyvinyl chloride. The choice of a chain transfer agent for a particular
process will be dictated, in part, by its efficiency in reducing the polymer chain length and its
effect on resin thermal stability.
Efficiency in reducing the polymer chain length: The kinetic aspects of chain transfer were
discussed in Section III.C.2 where the ratio of KTr/Kp was defined as the chain transfer
coefficient (CTC). The literature values for CTC are not consistent in either the experimental
procedures or in trends [101]. Experience has shown that the CTC for each chain transfer
agent needs to be determined in the specific polymerization system for which it is used.
Differences in the water/monomer ratio have been observed to affect the CTC values. The

pseudo-chain-transfer coefficients for alkylated

Page 108
olefins and trichloroethylene were measured at 65C to determine the effects of the allylic
hydrogen content of the alkylated olefins in reducing the molecular weight of polyvinyl chloride
[25]. The study indicated that under the particular experimental conditions trimethylethylene had
the same pseudo-chain-transfer coefficient as trichloroethylene. This method of calculating the
chain transfer coefficient is not the same as Mayo's procedure [102] since the conversion is
carried to about 80% and the ratio of chain transfer agent to monomer changes with
conversion. The advantage of the pseudo-chain transfer procedure [25] is that it represents the
actual process in the commercial manufacture of polyvinyl chloride. This procedure should be
applicable to all of the chain transfer agents used in polyvinyl chloride synthesis. As a general
trend it has been observed that the efficiency of a chain transfer agent is related to the bond
strength for the reaction

The compounds with structures such as

have rather weak bonds [103] and are efficient chain transfer agents for polyvinyl chloride.
Fischer [101] has indicated that with a highly active chain transfer agent it is possible to alter
the monomer/chain transfer agent ratio with conversion and to broaden the molecular weight
distribution of the resin. This effect was observed for the copolymerization of vinyl chloride and
propylene [104]. In that study the propylene, which acts as both a chain transfer agent and a
comonomer, increases in concentrations relative to vinyl chloride during the polymerization. As a
consequence, the molecular weight of the resin decreases with conversion, going from an
inherent viscosity of 0.73 to 0.49.
Effect on resin thermal stability: The chain transfer reaction was depicted mechanistically in
equation (18). The termination of an active polymer and the initiation of a new polymer chain
introduces unusual structures into the end of the polymer chain which can affect the thermal
stability of the resin. These are detailed in Figure 19. Recent work on aldehyde chain transfer
agents by Marshall et al. of Goodyear [105] and Bauman of Air Products [106] indicates that
the use of substituted aldehydes such as isobutyraldehyde or 2-ethylhexylaldehyde eliminated
the problem of thermal instability for aldehyde chain transfer agents. The probable reason for
the improvement in thermal stability is the preferred decarbonylation of the substituted aldehyde
radical before it adds to a monomer unit.

Page 109

FIGURE 19 Effect of chain transfer reactions on the end group structure.

The presence of carbon monoxide from the decarbonylation may enter the growing polymer
chain. The presence of such structure does not affect the overall thermal stability of the resin
when compared to a material prepared with either trichloroethylene (TCE) or a primary
aldehyde.
The thermal instability of resin made with trichloroethylene is due to the allylic chlorine sites
which are prone to rapid dehydrochlorination. The brominated compounds introduce a labile CBr bond into the polymer which, based on laboratory experience, is very sensitive to thermal
degredation. The substituted aldehydes appear to have the best combination of efficiency and
thermal stability.

Page 110
The color and clarity properties of a polyvinyl chloride resin are virtually impossible to determine
in the unstabilized form; some level of thermal stabilizer is needed to perform the initial fluxing of
the resin into a molded piece. Experience has shown that the type and level of the suspension
system, the type of buffer system, and the choice of chain transfer agent will affect the color
and clarity of the resin.
Amide-containing suspension systems tend to impact an early thermal degradation to the resin.
In particular when barium-cadmium stabilizers are used. High levels of buffer salts will affect
both the clarity and the electrical properties of the resin. The choice of suspending agent, buffer
system, and chain transfer agent (when needed) will represent a judicious balance between the
efficiency for the polymerization process and the quality of the finished resin. Many of these
choices will depend on competitive technology and customer (end user) desires. The domain
size of the suspension system in a fused resin and the difference in refractive index will be the
major factors in determining whether the resin in the pressed state is optically clear or hazy.
2. Particle Morphology: The phase separation of polyvinyl chloride from vinyl chloride monomer
[3, 3a] and the agglomeration of the polyvinyl chloride resin within the monomer phase establish
the morphological character of the resin. Eliassaf [107109] demonstrated in laboratory-scale
equipment that the interaction between the agitation system and the suspension system
controlled the morphology of the resin. There have been no comprehensive studies published in
the open literature detailing the interactive effects on a commercial scale of agitation and
suspending agent structure on resin morphology; the technology has been held as proprietary
secrets by the resin manufacturers. In this section the individual effects of suspension agent
and agitation on particle morphology are discussed.
a. Particle Growth Through Aggregation: The aggregation process within the monomer droplet
has been recognized as being very important in establishing particle morphology of polyvinyl
chloride. Following the Second International Symposium on PVC (1976, Lyon), Geil [110]
published a unifying nomenclature to describe the particle regions which were being reported in
the open literature. A summary is detailed in Table 17. This nomenclature will be used
throughout the section. A schematic representation for the agglomeration of the resin is
depicted in Figure 20. The assumptions in Figure 20 are that the monomer droplets are 40 m
in diameter and there is sufficient agitation and protective colloid present to maintain the
monomer drop size. The monomer droplets are converted to polymer in the following manner
within the 40-m droplet.
The polymerization is initiated by a radical species. When the polymer radical gets to a chain
length of about 10 repeat units, it separates as a coiled macroradical in the monomer phase.
Approximately 50 of these coiled macroradicals aggregate to form a microdomain. This occurs
at less than 1% conversion. The microdomains are unstable and aggregation occurs again with
about 1000 microdomains merging to form each primary particle nucleus. This occurs at 1 to
2% conversion. The primary particles grow from about 0.1 to 0.2 m at 2% conversion to 0.6
to 0.8 m at ~85% conversion. The primary particles go through a third aggregation to form the
agglomerate at about 4 to 10% conversion. The agglomerate is initially about 1 m in diameter

and grows to about 8 m diameter at about 85% conversion. Within the monomer droplet, after
about 10% conversion there are only agglomerates swollen with monomer (at a ratio of 77:23
polymer/monomer) and the free monomer

Page 111

50
Visible constituent of free-flowing powders, made up of more than one
130
250
monomer droplet.
10
Subgrain
40 Polymerized monomer droplet.
150
Formed during early stage of polymerization by coalescence of primary
Agglomerate110 5
particles (12 mm). Grows with conversion to size shown.
Primary
0.6
Grows from domain. Formed at low conversion (less than 2%) by
0.7
particle
0.8
coalescence of microdomain; grows with conversion to size shown.
Primary particle nucleus. Contains about 103 microdomains. Only
0.1
observed at low conversion (less than 2%) or after mechanical working.
Domain
0.2
0.2
Term used only to describe 0.1-mm species; becomes primary particle
as soon as growth starts.
0.01
Smallest species so far identified. Aggregate of polymer chains
Microdomain
0.02
0.02
probably about 50 in number.
Grain

Source: Ref. 111.

Page 112

FIGURE 20 Agglomeration scheme for suspension-grade polyvinyl chloride. (From Ref. 111.)
phase. Initiator molecules are distributed throughout both phases. There is no evidence that
indicates a preference for the initiator in either phase. The aggregates grow, consuming the
free monomer phase in the droplet. At about 77% conversion the free monomer phase in the
droplet is gone and polymerization continues in the aggregates until the monomer droplet has
been converted into a subgrain. Depending on the agitation system and suspension system, the
subgrains will agglomerate to form a grain about 130 m in diameter. Recent work by Zichy
and coworkers at ICI [112115] and Davidson et al. at B.F. Goodrich [116] examined the
formation and packing mode for agglomerates in quiescent polymer droplets. They concluded
that there was an electrical charge on the surface of the agglomerates which served to
stabilized their structure and prevented the fusion of the individual agglomerate. This electrical
charge may have a reduced effect in well-agitated system where the monomer droplets are
easily disrupted.
b. Agitation Effects: The agitation system provides the mechanical energy to disperse the
monomer into droplets about 40 m in diameter. Without a

Page 113
continuous input of this mechanical energy, the polymerization mixture would separate into a
water phase and into a monomer swollen polymer phase. The agglomeration steps described in
Section VII.C.2.a will be affected by the agitation. Hopff and coworkers [117119] examined
the effects of the agitator configuration, agitator speed, phase ratio of water to monomer, and
the type of suspension agent on the particle properties of styrene, methyl methacrylate, and
vinyl acetate polymerizations. All of their work was performed in laboratory-scale reactors of
0.5- to 4.0-liter capacity. They found that the factors controlling particle size were agitator
speed, reactor volume, and the water-phase viscosity. Belopolisky et al. [120] have reported on
the polymerization of vinyl chloride in pilot-plant and production-scale reactors (250, 3500,
7500, and 10,000 gal). This group found that the particle size and particle-size distribution of
the resin were related to the agitator power/ volume and the suspension system concentration.
Recent work by Johnson et al. [121, 122] on polyvinyl chloride systems has related the
agitation parameters in both pilot-plant reactors [121] and production- and laboratory-scale
reactors [122] to the particle properties of the resin. To determine the effects of variation in
agitation on resin particle properties, Johnson varied the width, diameter, and speed of the
agitator in a 1250-gal pilot-plant reactor. The agitator variables were grouped together using
the impeller Weber number (We):

where t is the density of the total system, N the impeller speed, D the impeller diameter, the
interfacial tension, and g the gravitational constant. Johnson's results indicated that over the
range of all agitator designs used, the median particle size of the resin goes through a minimum
at a We of ~106. This is illustrated in Figure 21. At low agitation levels (~105We) the monomer
droplets are rather large, and as a result a coarse resin is formed. As the agitation level is
increased (105 to 106 We), the droplet size decreases and the particles formed decreased in
size to 150-m diameter. As the agitation is increased (106 to 2 x 106 We), the particles
coarsen again. The phenomenon occurring in the region >106We is an increase in the total
surface area for droplets which exceeds the available suspending agent to protect it from
agglomeration. The individual particles begin to agglomerate and the total surface area is
reduced. At some point an equilibrium between the total surface area and available suspending
agent is reestablished. It is possible at very high We for the entire suspension to agglomerate
into a single particle.
Johnson [121] studied the bulk density and dry-blend time for these resins. He found that over
the middle range of Weber numbers (5 x 105 to 106We) both the bulk density and dry-blend
time were only minimally affected by the agitation systems (Figure 22). Lewis and Johnson
[122] extended this work to both laboratory-scale and production-scale reactors. They
observed (Figure 23) that there is a parabolic curve for median particle size as a function of
Weber number for the laboratory-scale reactor and for the production-scale reactor.
The width of the particle-size distribution, as determined by the coefficient of variation, was
analyzed for all three types of reactors (Figure 24). The results indicated two rather interesting
points:

Page 114

FIGURE 21 Mean particle size vs. Weber number. (From Ref. 116.)
1. The width of the particle-size distribution approached a minimum value at the minimum
medium particle size (compared to Figure 23).
2. The width of the distribution decreases as the reactor size increases.
An analysis of the bulk density of the resins (Figure 25) and the dry-blend times of the resins
(Figure 26) as a function of agitator Weber number indicates that at the optimum Weber
number for each reactor, determined by the minimum with medium particle-size curve (Figure
23), the bulk density for the resin is apparently independent of reactor volume and there is a
slight decrease in dry-blend time. These results indicate that the agitation in a reactor can be
optimized to achieve the minimum particle size but that at this optimum agitation there is no way
to vary either the bulk density or dry-blend time. Bulk density and dry-blend time will be a
function of the manner in which the primary particles pack in the resin grains. This packing can
be controlled by the suspension system.
c. Suspension Systems Effects: The function of the suspension system is to control the
interparticle and intraparticle agglomeration in the polymer grains. The interparticle
agglomeration will determine the particle-size distribution for the resin and the intraparticle
agglomeration will control the bulk density and porosity of the resin. Johnson reported [122]
that variations in the agitation parameters of a laboratory-scale reactor affect the median
particle size of the resin but had little effect on the bulk density and dry-blend time of the resin.
Cheng [123] has confirmed these findings, expanded the work to several types of suspending
systems at different suspending system/monomer ratios.
He found that the parabolic curve of median particle size as a function of agitation was shifted
as the level of a hydroxypropylmethyl cellulose (HPMC)

Page 115

FIGURE 22 (a) Bulk density vs. Weber number; (b) dry-blend vs. Weber number.

FIGURE 23 Mean particle size vs. Weber number for three different size reactors.

Page 116

FIGURE 24 Coefficient of variation vs. Weber number.

FIGURE 25 Bulk density vs. Weber number.

FIGURE 26 Dry-blend time vs. Weber number. , Bench scale; , pilot plant; , commercial
production.

Page 117

FIGURE 27 Resin median particle size vs. agitator speed. (From Ref. 123.)
suspending agent was increased. This is seen in Figure 27. The resins prepared at the
minimum points of the parabolic curves (Figure 27) were tested for both bulk density and
porosity by the ASTM D3367 procedure. The bulk density for the resins was about the same
(30 to 32 lb/ft3), but the porosity decreased slightly from 24 to 21 of di-2-ethylhexyl phthalate
(DOP) per 100 g as the level of suspending agent was increased. These results indicate that
there is an interactive effect between the agitation system and the suspension system and that
variations in the concentration of a HPMC-type suspension have very minor effects on the
porosity of the resin.
The patent literature for polyvinyl chloride has numerous citations of the use of multicomponent
suspending agents to improve the particle properties of the resins. A sampling of the literature
is noted in Table 18. An examination of this literature indicates that these suspension systems
are combinations of at least two materials:
1. Primary suspending agent: a material capable of maintaining the particle size of the resin
2. Secondary suspending agent: a material needed to control the agglomeration of the primary
particles
The primary suspension agents are typically the cellulose ethers such as HPMC or the polyvinyl
alcohols of 75 to 90% hydrolysis. The secondary suspension agents are typically the
surfactants of hydrophile-lipophile balance

Page 118

Jap. Kokai
5236, 179
Ger. Offen.
2,451,695
U.S.
4,143,224
U.S.
3,995,696
Jap. Kokai
5221, 670
Jap. Kokai
53136, 089

Mar.
1977
May
D. N.
1976
Mar.
Hoechst
1979
Nov.
Huls
1976
July
Kureha
1976
Nippon Nov.
Synthetic 1978
Sigma
Apr.
GB 1,543,431
Italiana 1979
Denki

A mixture of HPMC (0.010.06), PVOH (0.0250.12) 80%

hydrolysis, and sorbitan monolaurate (0.20.4)
A mixture of HPMC (0.14) and sorbitan monolaurate (0.11)
A mixture of PVOH (0.039) 7595% hydrolysis, PVOH (0.070)
65% hydrolsis, and HEC (0.10)
A mixture of HPMC (0.1), HPC (0.05), and sorbitan monolaurate
(0.03)
A mixture of PVOH (0.050.20) 6080% hydrolysis and MC
(0.050.5)
A mixture of PVOH (0.05) 7080% hydrolysis and PVOH (0.05)
4050% hydrolysis
A mixture of HPMC (0.07) and PVOH (0.05) 29.538.5%
hydrolysis

Page 119
(HLB) range 4 to 10 [124] or polyvinyl alcohols of 25 to 40% hydrolysis [125]. Partially
hydrolyzed polyvinyl alcohol (hydrolysis level 60 to 80 mol %) offer several distinct advantages
over the substituted cellulosic materials as primary suspending agents. The molecular weight,
molecular weight distribution level of hydrolysis, and sequence block length can be varied so as
to offer a broad range of properties and phenomena at the water/vinyl chloride interface.
Recent work by Bugada and Rudin [125a] has described analytical procedures that can be
used to determine both the block length and blockiness of the substituents. Their
characterization of several commercial polyvinyl alcohols is summarized in Table 19. With the
exception of recent publications of Cheng et al. [126, 127], there have been no reports on the
relationship between the physical properties and chemical structures of the suspension system
and the particle morphology of the resin.
The particle morphology of polyvinyl chloride resin is not uniform across all the particles in a
large population sample. It has been estimated that for a typical resin sample approximately 1
particle in about 105 particles will have a low porosity and a slow rate of plasticizer adsorption.
These low plasticizer adsorbing particles will result in the phenomenon called fisheyes in a
flexible film. The nature of a fisheye will be discussed in this section as it applies to particle
morphology and in Section VII.C.3 as it applies to monomer removal and plasticizer adsorption.
Kolinsky and coworkers [128a] have examined the atypical resin particles from a suspension
process. Their analysis indicates that the precursors of fisheye particles (PFE) have the same
molecular structure as the typical or normal particles as determined by light scattering,
solubility, and x-ray diffraction. The PFE particles do have a different particle morphology which
Kolinsky et al. [128a] feels is the major cause of fisheyes in flexible films.
Cheng et al. [126, 127] have determined the partition coefficient of the suspending agent
between vinyl chloride/water and the interfacial tension of a water/monomer interface for a
series of HPMC materials of increasing molecular weight. They found that the porosity of a
resin increased as the partition coefficient of the suspension agent increased. This is seen in
Figure 28.
a OH (mol %) OH OAc Number of branches/ 1000 main chains Goshenol 79.3 11.8 3.1 0
Goshenol 78.2 11.1 3.1 65 Revertex 75.0 9.1 3.0 10 Revertex 78.4 10.1 2.8 0 Revertex 60.0
3.5 2.3 60 Sigma 58.1 4.0 2.9 100 aGiven without code number. Source: Summarized from
Ref. 125a.

Page 120

FIGURE 28 Porosity vs. partition coefficient for suspending agent.

In addition, they observed that partition coefficient was very sensitive to temperature,
increasing the temperature increased the solubility of the HPMC in the monomer phase. The
interfacial tension for a vinyl chloride/ water was measured [77] for HPMC suspending agents
of different substitution levels. The results indicated that there was a 2- to 3-dyn/cm difference
between HPMC material of 28% methoxy and 21% methoxy at the same hydroxy propoxy level
(8.5%) and the same molecular weight. The HPMC with the lower interfacial tension
consistantly gave resins with improved porosity.
All the previous examples demonstrated the use of a polymeric primary suspending agent with
a monomer-soluble secondary suspending agent. Daniels and coworkers [128] have shown that
a primary suspending agent based on the in situ formation of a calcium dodecylbenzene
sulfonate salt from the reaction of a monobasic calcium phosphate and sodium dodecylbenzene
sulfonate would stabilize the particle agglomeration to a grain size of about 250 m. By adding
a vinyl chloride copolymer latex the internal particle structure of the resin grains was improved;
the porosity of the resin was increased from 22 to 35. The structure of the primary suspending
agentwhether it is the monododecylbenzene salt or the didoceylbenzene saltwas not
determined by Daniels.
It can be speculated from these patent and literature citations that the internal porosity of the
resin can be increased by using suspension systems that have a high monomer solubility. These
agents may act to control the form and nature of the aggregates. Conversely, a high-bulk
density resin can be formed by reducing the monomer solubility of the suspension systems.

Page 121
3. Diffusion Processes: Diffusion processes are important in the manufacture and fabrication of
polyvinyl chloride resins. The removal of vinyl chloride monomer at the end of the polymerization
process and the absorption of plasticizer during the fabrication process are generally believed
to occur by a diffusion mechanism [129]. These diffusion processes are controlled by the
particle morphology of the resin.
a. Diffusion of Vinyl Chloride in Polyvinyl Chloride: In a commercial operation it is not possible to
polymerize completely all of the vinyl chloride monomer. The diffusion-controlled kinetics
described in Section IV place an upper limit on the percent monomer that can be polymerized.
The process economics dictate that substantial levels of monomer, typically 10 to 15%, are left
unreacted in the reaction mixture. Prior to 1972 it was not uncommon for commercial resin to
be shipped to customers with vinyl chloride monomer levels of 1000 to 2000 ppm. In 1982 the
level of unreacted monomer in extrusion-grade resin was typically less than 10 ppm. This 100fold reduction in residual vinyl chloride level in the resin has been achieved by understanding the
diffusion of monomer in the different phases of the polymerization mixture and scaling up this
technology to systems capable of handling up to 15,000 lb of polymer per hour.
The removal of vinyl chloride from polyvinyl chloride is a diffusion-controlled process. Stern and
Frisch [130] have reviewed the general aspects of diffusion theory and their article is
recommended for an overview of the subject. The diffusion mechanisms for vinyl chloride in
polyvinyl chloride resin fall into two regimes; above and below the glass transition temperature
(Tg) for the resin. At temperatures above Tg, the resin is in a rubbery state and the diffusion
mechanism conforms to a simple Fickian relationship. Below Tg the resin is in a glassy state
and the diffusion mechanism conforms to a nonclassical Fickian diffusion [130] mechanism. The
implications of this nonclassical mode of diffusion on the long-term migrations of monomer, as it
applies to packaging and water pipe, have been studied [131, 132] but will not be discussed in
this chapter. The topic of diffusion from resin at temperatures above Tg will be discussed in
some detail.
Berens has examined the relationship between monomer diffusion and the process parameters
of particle morphology and temperature [133135]. Using polyvinyl chloride in the form of wellcharacterized suspension and emulsion particles, Berens was able to measure the diffusion
coefficient for vinyl chloride monomer in a polyvinyl chloride matrix. He found that the
mathematical approximation developed by Crank [136] for uniform spheres,

where
D
= diffusion coefficient for vinyl chloride, cm2/sec
d
= diameter of the particle, cm
t1/2
= time for half saturation, sec
could predict the diffusion of vinyl chloride out of polyvinyl chloride. The smaller the diameter of
the particle, the more rapidly the vinyl chloride will diffuse out of the particle. The presence of
agglomerated or nonporous (glassy) particles of large diameter will significantly affect the rate

at which

Page 122

FIGURE 29 Vinyl chloride monomer desorption studies at 90C: effect of particle size of the
resin. A, 20 M diameter, 100%; B, 1.9 M/20 M mixture 90/10 w/w(%); C, 1.9 M diameter,
100%.
vinyl chloride will diffuse out of a polymer matrix. This is illustrated in Figure 29, where the rate
of monomer desorption from particles of 1.9 and 20-pm diameter were examined in different
weight ratios at temperature of 90C. The presence of the large-particle-size resin at levels as
low as 10% will result in a high residual vinyl chloride level after 5 min of heating. Variations in
resin molecular weight, over the range 40,000 to 80,000 (number average), does not appear to
affect the diffusion coefficient [135]. Increasing the temperature will increase the diffusion
coefficient in the manner shown in Figure 30. Changes in diffusion coefficient have been
examined by Chan and coworkers [137, 138] in a study of the effects of increased temperature
on the rate of diffusion of vinyl chloride from polyvinyl chloride. Their results are illustrated in
Figure 31. At temperatures of 105C the vinyl chloride level in the polymer is reduced to very
low levels within 5 min. The curvature seen in these results may indicate a mass transfer limit in
the water phase or an agglomeration of the primary particles as a consequence of heating.
These experimental results indicate that the two principal process factors that control the rate
at which vinyl chloride monomer can be removed from polyvinyl chloride resin are the particle
morphology of the resin and the temperature. Good agitation in the slurry is necessary to
reduce the external constraints to monomer transport. To achieve a rapid reduction in the level
of residual monomer in the resin, it is very important to reduce the agglomerated, nonporous
glassy particles to a minimum level.
b. Commercial-Scale Monomer Removal: The objective of a commercial monomer stripping
operation is to reduce the monomer content of the resin to a level that makes the subsequent
fabrication and utilization of polyvinyl chloride articles safe within current understanding. Other
objectives are to recover the monomer in a manner that protects the environment and maintains
the physical properties of the resin.

Page 123

FIGURE 30 Arrhenius plot of log diffusion coefficient vs. reciprocal absolute temperature.
All of the foregoing objectives need to be attained within a reasonable cost. The translation of
the diffusion studies described in Section VII.C.3.a to commercial-scale operations has been
quite successful. Starting with a resin of fixed particle morphology, the temperature and time
are the process variables that determine the level of residual monomer in the resin. In a
commercial stripping operation the resin, usually in the presence of water, is quickly heated to a
high temperature (in the range 90 to 105C) and a fluid in the form of a gas or liquid sweeps
over the resin to carry away the vinyl chloride monomer. The purpose of the sweep fluid is to
ensure that all the

FIGURE 31 Effect of temperature on vinyl chloride removal from polyvinyl chloride. =

normalized concentration of monomer at time = 0; CS = normalized concentration of monomer
at time = t.

Page 124
resin is heated uniformly. The residence time for polyvinyl chloride resin at high temperatures is
very critical: too long a residence time and the resin will begin to dehydrochlorinate; too short a
residence time will result in a resin with high levels of residual vinyl chloride.
Several different approaches have been used to heat the resin; among these have been
microwave irradiation, hot gases, and steam. The microwave irradiation technique has been
reported by Air Products [139], British Petroleum [140], and Kanegafuchi [141]. In these
processes the resin and about 10 to 30% (by weight) of water are irradiated at 2450 MHz for
about 6 to 8 min. The microwave radiation is absorbed by the water molecules and raises the
temperature of the entire mass. As the water molecules boil they act as an internal sweep fluid
to carry the vinyl chloride vapor away from the polymer. The water acts as a sink to absorb the
microwave radiation and is critical to this type of process.
Another stripping process uses a nitrogen sparge through the slurried resin [142]. The sparge
system forces the vinyl chloride monomer out of the aqueous phase and into vapor phase
above the slurry. The disadvantage of this system is that the monomer recovery system has to
handle a mixed gas system composed, in part, of a noncondensable component.
Countercurrent steam stripping columns have been very effective in reducing the monomer level
in resin. Several Japanese companies [143] and B.F. Goodrich [146] have received patents on
the design and operation of countercurrent monomer stripping columns. A typical stripping
column is illustrated in Figure 12a. The usual operation of a stripping column can be described
as follows. Resin slurry at 25,000 to 50,000 ppm of monomer and a temperature of about 50C
flows through a heat exchanger, which raises the slurry temperature to about 75C. The slurry
can enter the stripping column at a number of different entry points; the higher up on the
column, the longer the slurry residence time. The slurry then descends through the column,
flowing over a series of trays whose purpose is to provide a large surface area for contact
between the ascending steam and descending slurry. The slurry temperature profile increases
from 75C to 90C at the top of the column to 95 to 110C at the bottom of the column. The
residence time in the column can vary from 3 to 8 min. The slurry exits the column at a residual
monomer level of 200 to 20 ppm. The slurry then flows through a heat exchanger, which lowers
its temperature about 25C.
The design of the trays is an extremely important factor in the efficient operation of the column.
The tray design will affect the utilization of steam and therefore the process economics and
formation of burned particles. The wier height will affect the residence time on each tray and
the time-temperature profile of the column. Using the existing literature citations, it should be
possible to model the mass transfer equations for monomer in the liquid, solid, and vapor
phases for each tray on a monomer stripping column. This model should permit the most
efficient design and operation of a stripping column.
c. Plasticizer Absorption: The plasticization of polyvinyl chloride resin represents a gradation of
increasing interactions between the liquid plasticizer and the solid resin. The first step in the
process is the mixing of the liquid plasticizer and the solid resin into what appears to be a lumpy

mixture. The plasticizer, in this state, coats the outer surface of the resin grain and the mixture
appears wet to the touch. The mixture is usually stirred in either a ribbon blender or high-speed
intensive mixer to aid in the uniform absorption of the liquid plasticizer into the resin particles.
Davidson [147]

Page 125
has examined plasticizer absorption in both a Brabender Plasticorder and a high-speed mixer.
He found that the temperature of the mass was the most important factor in determining the
rate of plasticizer take-up. Schwab and Wingrave [148] studied the effect of conversion on the
Bradender dry-blend time for resin and the rate of plasticizer take-up. They found that as the
polymer conversion was increased, dry-blend time was extended and the rate of plasticizer
take-up was slowed. They also reported on the use of an undefined additive which improved
the plasticizer absorption effect by reducing the dry-blend time and increasing the rate of
absorption. The additive did not increase the total capacity of the resin for plasticizer. Quy [149]
extended these studies on additives and found that the size of the agglomerate particles
increase with conversion. He also found that this increase in agglomerate diameter (d) with
conversion was related to the dry-blend time. This is illustrated in Figure 32a. Quy indicated
that additives such as cetyl alcohol and lauryl alcohol could reduce the diameter of the
agglomerates by possibly stabilizing the primary particles against flocculation and thereby
ensuring more agglomerates, of smaller average diameter.
The results indicate that the diffusion of plasticizer into the agglomerate particles is the factor
that controls both dry-blend time and the rate of plasticizer take-up. Additives that will reduce
the agglomerate diameter will reduce the dry-blend time and increase the rate of plasticizer
take-up.
These effects of the diffusion of plasticizer into a resin grain are related to the fisheye level of a
resin. There are numerous industrial test procedures used to determine the rate of plasticizer
uptake and the uniformity of plasticizer uptake for a polyvinyl chloride resin. Balakirskaya and
Shtarkman [149a] have developed a solvent density procedure using methanol/carbon
tetrachloride to determine the porosity distribution within a resin. These tests are based on
answering two practical questions:
1. How long will it take to make a uniform mass?
2. How high a temperature will be needed to obtain a totally uniform mass?
The efficient use of a compound plant demands that the smallest/lowest-energy mixing
equipment be used in the shortest possible mixing cycles. Given a long mixing time and high
temperatures, all the resin would adsorb plasticizer and the resin would form a uniform mass on
fusion. This is shown in Figure 32b where gels as a function of mill time are illustrated for three
lots of production resin.
It has been found that the suspension system (primary and secondary components) has a
major impact on the level of fisheyes in a resin. It is reasonable to assume that the agitation
system and the reactor configuration will also affect the fisheye level. Since Berens [133135]
has already demonstrated the effects of particle diameter on vinyl chloride diffusion in and out
of a particle and Quy [149] has examined effect of secondary suspension components on
plasticizer adsorption, the next step will be to couple the two techniques to determine those
components that will permit the uniform and rapid adsorption of plasticizer at low temperature.

These materials will be able to control the agglomeration and growth characteristics of the
primary particles. In the discussion of bulk polymerization processes (Section VIII.C.2.a), the
effects of additives on particle morphology are reviewed. The technology that is developing for
the bulk polymerization process will have an impact on the suspension process.

Page 126

FIGURE 32 (a) Dry-blend time vs. diameter of agglomerates; (b) gel count vs. total time on
mill: 0.025 in. gap, 310 F, 34/24 rpm, initial mixing only, carbon black.

Page 127

VIII.BULK POLYMERIZATION
The bulk polymerization process can be viewed as a suspension process without water. As
such, the bulk process has several features which simplify the overall manufacture of the resin.
This can be illustrated by comparing the block flow diagrams for the suspension system (Figure
12a) and the bulk system (Figure 33a). There is no water in the polymerization reactor, and as
a consequence the reactor can be smaller and the cooling and heating system for temperature
control in the reactor can be smaller. Second, there is no water to be removed from the resin,
and as a consequence there is no need for either centrifuge or dryer capacity. Also, there is no
need for pretreatment or posttreatment of the process water.
The development of a commercial bulk process began in the 1930s with work at Produits
Chemiques St. Gobain. By 1939 a bulk process had been scaled up to a 200-gal reactor and
was reported to produce, for that period, commercial-quality resin. However, problems in scaleup beyond 200 gal prevented any further development in the bulk process until the 1950s.
During the early 1950s, problems with particle-size control, heat removal, and wall fouling were
studied. By 1956 a new process, code-named L-51, had been developed to overcome these
problems. The L-51 process consisted of a 3000-gal reactor mounted horizontally, rotating
along its long axis. Agitation of the slurry was assisted in the L-51 process by a number of
stainless steel balls. The reactor equipment could be viewed as a ball mill with the rotation
providing the circular motion and the balls providing a grinding action.
The process was modified in the late 1950s to a two-stage process in which the polymerization
was performed in two separate reactors. The first reactor was in a vertical configuration and
was equipped with a high-speed agitator. The particle-size distribution for the product was
established in the first 7 to 12% conversion in this first-stage reactor. The second-stage reactor
was mounted horizontally, as in the L-51 process, but was agitated by a low-speed ribbon
blender type of impeller. The polymerization was carried to a final conversion of 65 to 80% in
the second-stage reactor. Reflux condensers were added to the first- and second-stage
reactors to improve the heat transfer, and in 1962 this process was put into commercial
production as the M-66 process. A merger of St. Gobain and Pechiney occurred at this time
and the M-66 process became known as the two-stage PSG bulk process.
In the late 1960s additional process modifications were developed for the two-stage process
which increased the reactor productivity. The first modification was the use of the first-stage
reactor to feed several second-stage reactors. The second modification was to increase the
capacity of the reactors in the first and second stages. By the mid-1970s a single first-stage
reactor of 6000-gal capacity was used with five second-stage reactors of 12,000-gal capacity.
This process became known as the N-8 modification and was commercialized in 1975.
Industrial mergers in the French chemical industry took place at the same time and the N-8
process became the property of Rhone-Poulenc.
In the mid-1970s experimental studies on agitation indicated that a vertical second-stage
reactor would have certain process advantages, and in 1978 the V-50 version of the bulk

process was commercialized. The V-50 process has a single first-stage reactor of 8000-gal
capacity feeding five second-stage reactors of 13,200-gal capacity [149b].

Page 128

FIGURE 33 (a) Polyvinyl chloride by mass process block flow diagram; (b) polyvinyl chloride by
mass process polymerization and monomer recovery; (c) polyvinyl chloride by mass process
product classification; (d) second-stage reactor in the V-50 bulk process. (From Ref. 151.)

Page 129

FIGURE 33 (Continued)
A.Bulk Polymerization Overview
The two reactors in the bulk polymerization have two different functions: process conditions in
the first reactor determine the particle-size distribution and have an effect on the particle
morphology; process conditions in the second reactor determine the molecular weight of the
resin and also affect the particle morphology of the resin. The differences in the process
conditions for the first- and second-stage reactors are illustrated in Sections VIII.A.1 and
VIII.A.2.
1. Polymerization Recipe: The bulk process recipe is much simpler than the suspension process
in that the chemical and physical effects of the water phase and suspension system are not
present. The agitation system in the first-stage reactor and the additives control the particle
morphology of the bulk process resin. The most recent development in the bulk process, the V50 modification [150, 151], will be used for this discussion. In a similar fashion as the
suspension system, the process will be illustrated with a medium-molecular-weight resin with an
inherent viscosity of 0.92 0.02. The process recipe and cycle time are detailed in Table 20.

Page 130
a. Recipe
a Reactor 1: 8000 gal Vinyl chloride 41,000 lb 53.2 ACSPb 3.1 lb 0.0040 DEPDCc 1.7 lb
0.0022 Particle control agents ~255 lb ~0.32 Polymerization temperature 69C Agitator speed
15 ft/sec The polymerization is carried to about 10% conversion, and the contents of the first
reactor are added to the second reactor, which already contains: Reactor 2: 13,200 gal Vinyl
chloride 36,000 lb 46.8 ACSPb 3.3 lb 0.004 DEPDCc 4.4 lb 0.006 Polymerization temperature
55C Polymerization time 255 min Agitator speedd Helical 525 rpm Anchor aBased on 77,000
lb of monomer. bAcetycyclohexyl sulfonyl peroxide. cDiethyl peroxydicarbonate. dSee Figure
33.
b. Cycle time (min)

Charge and heat-up

Polymerization
Venting and stripping
Drop
Wash and prorated antifouling system application

aDrop from reactor 1 to reactor 2.

bDrop from reactor 2 to bulk handling system.

30
20
30a
5
___
85

75
255
100
45b
30
___
505

Page 131
In the bulk process about half the total monomer charge is polymerized in the first-stage
reactor to about 10% conversion at a rather high temperature in a short period of time. The
content of the first-stage reactor is then pumped to the second-stage reactor together with the
second half of the monomer and additional initiator. The type of pumping between the first and
second reactors can affect the particle character of the resin, and this part of the process is
handled very carefully [152]. The polymerization process is then completed in the second-stage
reactor. The V-50 process is designed for a single first-stage reactor to provide the polymer
feed for a process train of five second-stage reactors. The cycle times for first- and secondstage reactors are balanced for this process.
2. Process Equipment: A process flow diagram for the V-50 bulk polymerization process is
depicted in Figure 33b and c. Figure 33b represents an overview of the polymerization process.
Compared to the suspension process flow diagram (Figure 12bd), the bulk process has
eliminated all the water-related process equipment upstream from the reactors: water
deaerator, demineralized water tank, suspension makeup, and buffer makeup. In addition, the
stripping column is not used because there is no water in the reaction mixture. Downstream
from the reactor the bulk process does not use the water-related blending system and water
removal systems depicted in Figure 12b. The monomer recovery system for the two processes
will be about the same in overall design (Figure 12d). The bulk process does have a more
extensive dry-resin-handling system. This can be seen in a comparison of Figure 12c and 33c.
The resin is screened for both undersized and oversized resin and is then transferred to a
series of blending bins and products silos. Each of the intermediate dry-powder-handling
vessels can be used for vinyl chloride removal.
The major equipment development of the V-50 modification is the vertical second-stage recator.
This second-stage reactor, depicted in Figure 33d [151], has a dual agitiation system consisting
of a helical agitator for the turnover of the dry powder and an anchor-type agitator to sweep
the reactor walls. These two agitators are independently driven and usually run at different
agitator speeds. The vertical configuration for the second-stage reactor facilitates the rapid and
efficient discharge of the resin at the end of the polymerization cycle. The equipment depicted
in Figure 33b includes:

First-stage reactor
Second-stage reactor

8,000
13,200

1
5

Depending on product mix and scheduling, the product facility depicted in Figure 33b and c will
produce up to 225 million pounds of product per year.
B.Resin Productivity
Resin productivity in the bulk polymerization system is controlled by the same process factors
that were discussed in connection with the suspension polymerization system: heat generation
and heat transfer processes. Many

Page 132
of the issues discussed for the suspension polymerization system, such as overall initiator
kinetics and initiator structure-property relationships, are identical for the bulk polymerization
process and will not be reiterated. On the other hand, the two-stage bulk process, because of
the different process characteristics of the first- and second-stage reactors, does have a
number of different factors that will affect the resin productivity. The effects of these different
process characteristics on heat generation and heat removal will be discussed in detail.
1. Heat Generation and Removal in the First-Stage Reactor: The heat generation in the firststage reactor occurs within a rather narrow process envelope. Initiation and polymerization are
performed in the temperature range 63 to 75C and in a polymerization time of about 20 min.
The initiators used in the first-stage reactor are added to the rapidly stirring monomer at the
polymerization temperature; adding the initiator at temperatures below 62C would result in a
significant quantity of resin being formed at polymerization temperatures outside the optimum
process conditions. These optimum process factors are detailed in Section VIII.3. The
polymerization in the first-stage reactor generates an extremely high polymerization heat load of
about 2.95 million Btu in 20 min.* This heat is removed from the first-stage reactor through the
conventional approach of reflux condensers and refrigerated water in the reactor jacket and the
process may also use a nonconventional approach of temperature-step processing and deadstop polymerization.
The use of both reflux condensers and refrigerated water as heat removal surfaces was
discussed for suspension polymerization in Section VII.B.2.a. These factors are the same for
the first-stage reactor in a bulk process and therefore need no additional discussion. A major
difference between the suspension polymerization process and the first-stage reactor in the
bulk process is the ratio of heat removal between the jacket and reflux condenser. In the
suspension process the reflux condenser is not used during the initial particle formation period
of < 5% conversion. In the bulk process about 60% [153] of the total heat load is removed
through reflux condenser operations in the first-stage reactor. If the heat removal by reflux
condenser and refrigerated water is not able to maintain an isothermal process, the
polymerization temperature may be increased in a controlled temperature-step process similar
to that described in Section VII.B.2.d. The upper limit for the polymerization temperature by the
step process will be about 75C. This is the usual upper limit on the working pressure of the
first-stage reactor and on the process conditions for bulk polymerization.
A fourth technique for controlling the heat generation process is dead-stop polymerization [154].
In dead-stop polymerization conditions, the rate of polymerization drops to a very low level
because the initiator is depleted in the system. It has been shown that the rate of
polymerization is proportional to the square root of available initiator [equation (32), Section
IV.C). ACSP initiation has a 5-min half life at 69C, the operating temperature of the first-stage
reactor. Therefore, the ACSP initiator in the first-stage reactor will be reduced to about 7% of
the original level after 20 min at 69C.

* 41,000 lb x 0.10 conversion x 720 Btu/lb = 2,952,000 Btu.

Page 133
Any additional increase in the polymerization temperature in the first-stage reactor above 69C
will further deplete the available initiator. There are several advantages to the dead-stop
polymerization process: The reaction is self-limiting in time and therefore in conversion, and the
resin from a first-stage reactor can be held at a fixed conversion without using a short stop until
a second-stage reactor is available. This dead-stop technique can be used with other initiators
and is a key process control parameter in the bulk polymerization process.
2. Heat Generation and Heat Removal in the Second-Stage Reactor: The heat generation
mechanism in the second stage of a bulk process is the conversion of monomer to polymer. As
in the suspension process, a multicomponent initiator system is used so that there is a uniform
heat release process in the second-stage reactor. The approach to identifying an initiator
system that will give a uniform heat release process is the same as described in Section
VII.B.1.b. In the bulk process there is a wider choice of initiators since there is no problem with
aqueous hydrolysis of initiator. The productivity for the second stage depends on how rapidly
the monomer can be converted to polymer. The amount of monomer converted to polymer in
the bulk process varys between 65 and 80%. Therefore, in the V-50 process the total heat load
to the second-stage reactor will be between 33.1 million and 41.4 million Btu. The heat removal
process in the second-stage reactor is considerably different from the suspension process.
There is no continuous water phase to act as either a heat transfer medium between the
polymerizing mass and the reactor walls or as a heat ballast to moderate any localized
overheating. The major process concerns for heat removal will be providing adequate heat
transfer capacity and keeping the heat transfer surfaces clean of wall fouling. The antifouling
coatings described in Section VII for the suspension process can be used in the bulk process.
In addition, there is no aqueous-phase polymerization to contend with in the bulk process and
therefore no need for aqueous-phase polymerization inhibitors. The bottom entering anchor
agitator provides a scouring action to the lower half of the inner wall of the reactor.
The total cooling surfaces of the second-stage reactor in contact with the dry powder is about
900 ft2. This surface area can provide only about half the heat removal capacity needed for the
temperature control in the second-stage reactor. The reflux condensers are needed to provide
the additional heat removal capacity. The bulk polymerization process is apparently less
sensitive to high levels of reflux than is the suspension process. In the suspension process the
reincorporation of refluxed monomer into the droplets will affect the particle morphology. The
effectiveness of the reflux condenser to act as heat transfer surfaces is dependent on the vapor
pressure of the vinyl chloride monomer. At the pressure break (xc) the vapor pressure of the
monomer begins to decrease and therefore the effectiveness of the reflux condenser is
reduced. This reduction in total heat removal capacity may require the termination of the bulk
process at conversion levels lower than that of the suspension process.
C.Resin Quality
1. Molecular Weight: The molecular weight of a bulk polymerized resin is controlled by the
same factors that control a suspension process material, namely, the competitive reactions of
chain addition and chain transfer. In the bulk process the small quantity of resin produced in the
first-stage reactor has a minor effect on the overall molecular weight of the finished resin.

Page 134
The temperature of the second-stage reactor determines the molecular weight of the finished
resin. The relationship between polymerization temperature and inherent viscosity for the bulkpolymerized resin is the same as that defined in Figure 2. The upper pressure limit of the
reactor determines the maximum polymerization temperature and therefore minimum inherent
viscosity that can be achieved in the second-stage reactor. As with the suspension reactors, a
maximum polymerization temperature of 65C will give a resin with an inherent viscosity of
about 0.75 dl/g. Chain transfer agents are needed to prepare resins with lower inherent
viscosity. The general trends concerning types of chain transfer agents described in Section
VII.C.1.a also apply to the bulk polymerization process.
2. Particle Morphology: The particle morphology for resin prepared in the bulk process is
determined by the manner in which the primary particles form and agglomerate. The phase
separation of the polyvinyl chloride from the vinyl chloride monomer in the bulk process
develops in the same manner as in the suspension process, through to the formation of the
primary particles (stage C, Figure 20). In the suspension process the agitation system and
primary suspending agent determine the particle size of the resin, while the particle morphology
is determined by the interaction between the primary and secondary suspending agents. In the
bulk polymerization process, the resin morphology is controlled by the temperature and
agitation in the reactors, and by the effects of additives in the first-stage reactor.
a. Effects in the First-Stage Reactor: The agitation effects in the first-stage reactor have been
documented by Chatelain [156] and Marks [157]. The median particle size (MPS) is controlled
by the agitator tip speed in the first-stage reactor (Figure 34). Variations in the diameter of the
agitator, from 25 to 35 cm, do not affect the MPS of the resin. The results are similar to the
effect reported by Johnson and Lewis [121, 122] using the Weber number for suspension
polymerizations. The relationship of MPS to agitator tip speed is noted in Figure 34. The key to
the particle formation in the bulk process is the rate and the uniformity of the formation of the
primary particles. The high rate of polymerization and high agitation in the first stage reactor
are the process parameters that control the median particle size.
The primary particles aggregate into the agglomerate stage at a conversion of 4 to 10% (stage
E, Figure 20). This aggregation is controlled by the

FIGURE 34 Median particle size vs. agitator tip speed.

Page 135
temperature in the first stage reactor. The higher the temperature, the greater the adhesion
between the aggregates and the lower the porosity in the finished resin. Conversely, as the
temperature of the first stage is lowered, the porosity of the resin is increased, while the
adhesion between the aggregates is decreased. The need for stable aggregates and
reasonable porosity in the resin sets the boundary conditions for the temperature in the first
stage reactor. This is illustrated in Figure 35 [56]. The lower limit on temperature is about 62C.
Below this the agglomerates are not sufficiently stable to be transferred from the first to the
second stage reactor. Above 73C the porosity of the finished resin is reduced to a low level.
The porosity is adjusted by process conditions in the second stage reactor. This is discusssed
in Section VIII.C.2.b.
The effects of monomer-soluble materials on the particle morphology of suspension polyvinyl
chloride have been discussed by Cheng and coworkers [126, 127] and Daniels and coworkers
[128]. These authors indicate that the packing of the polymer in the aggregate state can be
controlled by monomer-soluble polymers and surfactants. This type of approach has also been
used to control the particle morphology of bulk polyvinyl chloride. Hooker Chemical has
demonstrated that the addition of a wide variety of additives to the first-stage reactor will
reduce the particle-size distribution of the resin and in some cases will affect the processing
characteristics of the resin. The effects of the additives are detailed in Table 21. The additives
may reduce the particle size of the bulk resin by either maintaining the electric charge on the
surface of the primary particles in a manner described by Zichy et al. [112115] and Davidson
and coworkers [116] or by affecting the monomer-polymer interfacial tension as described by
Cheng and coworkers [126, 127]. The additives described in Table 17 reduce the median
particle size of the bulk resin from about 140 m to as low as 30 m. Tornell and Uustalu [157a]
have examined the effects of additives on the formation and agglomeration of primary particles.
Sorbitan monolaurate decreased the size of the primary particles. Polymeric materials such as
ethylene/vinyl acetate copolymer or polymethyl methacrylate were very effective in stabilizing
the agglomeration of the primary particles.
b. Effects in the Second-Stage Reactor: In the bulk polymerization process, the process
variables of temperature and conversion can be used to con-

FIGURE 35 Porosity vs. polymerization temperature in the first-stage reactor. Second-stage

reactor temperature, 55C; total conversion, 80%.

Page 136

G.B.
June
1,396,3481975
G.B.
June
1,436,1651976
U.S.
June
3,816,5351974
G.B.
May
1,436,1641976
U.S.
July
4,163,0331979

Small-particle-size silica made from dichlorodimethyl Particle size is reduced to

silane (0.15 phm) and Triton surfactant
2050 m

N-Vinylpyrrolidone
SB elastomer 80,000200,000 molecular weight at
0.050.5 phm
Ethylene-propylene copolymer at 3 wt %

Particle size is reduced to

2050 m
Particle size and impact
strength of the resin
Particle-size reduction
and improved impact

Page 137
trol the resin porosity (Figure 35). The resin slurry from the first-stage reactor is transferred to
the second-stage reactor, where additional vinly chloride monomer and initiator are added to
the recipe (see Table 20). The resin slurry is agitated by a combined helical screw and anchor
system, each turning independently.
The porosity of the bulk resin can be controlled by varying the polymerization temperature, as
illustrated in Figure 36 [155]. Lowering the polymerization temperature reduces the adhesion
between the agglomerate particles and increased the porosity of the resin. Since the molecular
weight of the resin is controlled by the polymerization temperature (Section VIII.C.1), this
process variable will need to be fixed for a particular grade of resinindependent of any
porosity need. The porosity of the resin can be increased by decreasing the conversion in the
second-stage reactor. This is illustrated in Figure 37. A linearized equation relating the process
control variables of polymerization temperature in the first-stage reactor (T1) and conversion in
the second-stage reactor (C2) to resin porosity (P) can be formulated as
P = K1 + K2 (T1) + K3 (C2)
[45]
The constant K2 and K3 are related to the slope of the curves in Figures 35 and 37,
respectively. Other factors, such as particle-size distribution, will also affect the porosity of the
resin, but there are no available data in the open literature to quantify this factor.
3. Commercial-Scale Monomer Removal: The mechanism of monomer diffusion out of
suspension-grade resin described in Section VII.C.3.a will generally apply to the bulk process
except that there is no water phase to act as a heat transfer medium. Fischer [157b] has
monomer removal characteris-

FIGURE 36 Porosity vs. polymerization temperature in second-stage reactor. First-stage

reactor temperature, 70C; total conversion, 80%.

Page 138

FIGURE 37 Porosity vs. conversion in the second-stage reactor. First-stage reactor

temperature, 70C; second-stage reactor temperature, 55C.
tics for bulk resin, with particular emphasis on the interaction between particle morphology and
monomer removal rate. The bulk resin is stripped of the unreacted monomer by heating the
resin to temperatures of 90 to 100C under alternate purges with nitrogen and vacuum in the
second-stage reactor. There is no separate stripping vessel for the V-50 process and therefore
the use of the second-stage reactor in a nonpolymerization process represents a definite
limitation to the bulk process. Problems of localized overheating have been overcome by
injecting steam or water into the second-stage reactor during the stripping process [158].
Water levels of up to 0.5 wt % (on polymer) can be used to act as both a heat transfer medium
and a sweep gas to purge the residual monomer out of the polymer. The use of both water and
nitrogen does introduce problems with corrosion in the downstream equipment and excessive
noncondensable gases in the monomer compressors and inert vent streams from the
condensers. These operating procedures will make the recovery system for the bulk process
larger and less monomer efficient than the monomer recovery system for a suspension facility
of the same annual capacity.

IX.EMULSION POLYMERIZATION
The emulsion polymerization process, as a commercial approach to preparing polymers, dates
back to the early twentieth century with the work of Hofman and Delbruch of Farbinfabriken
Bayer A.G. At that time, a major commercial goal for Bayer and others was the development of
a synthetic route to polyisoprene that would circumvent the Brasilian monopoly on natural
rubber. It seemed quite logical to many workers in the field that since naturally occurring rubber
is formed in a latex state in the Havea tree, the soundest approach would be to mimic that
biological process. The fact that the natural synthesis of polyisoprene is via an adenosine
triphosphate (ATP)-aided condensation route was of no historical consequences. The emulsion
free-radical polymerization process was developed and is a feasible process for the
preparation of a wide range of polymer products.
The development of the emulsion process in the late 1930s and early 1940s centered on the
manufacture of synthetic elastomers and in particular the large-scale production of styrenebutadiene rubber (SBR). The commercial development of the SBR process proceeded in a
pragmatic fashion without the

Page 139
benefit of a coherent theory to explain the process. The early work of Kolthoff, Price, and
Melville in the area of emulsions theory has been reviewed by Blackley [159] and is a good
exposition of the developments that led up to the work of Harkins and the Smith-Ewart theory.
Gardon [160] has also reviewed the early technical literature and his interpretation of the
development of the Smith-Ewart theory offers an opportunity for an alternative mechanistic
development. Ray and Min have reviewed the different models for emulsion polymerization and
their predictive abilities were compared to experimental data [160a]. More recently, these
authors have proposed an all-inclusive model equation set which combines both material and
energy balance for the emulsion process [160b].
The Smith-Ewart theory forms the cornerstone for classical free-radical emulsion polyerization
and will serve as the basis for comparison. first to the previously discussed suspension and
bulk process, then to the recent work in the area of emulsion polymerization of vinyl chloride
and the effects of phase separation and chain transfer to monomer.
A.Emulsion Polymerization Overview
The emulsion polymerization process represents a major conceptual departure from the
suspension and bulk process. In the suspension and bulk processes both the initiator and the
monomer components are uniformly distributed throughout the polymerization mass. The
particle structure for the suspension and bulk polyvinyl chloride resin is determined by an
aggregation process which results in a polymer with a median particle size of approximately
150 m. The internal porosities of these resins are developed through the void distribution
during various stages of the particle formation. In the emulsion polymerization process, the
initiator is segregated from the monomer phase by differences in solubility, for the initiator is
soluble only in the water phase, with thermally induced radicals entering the active monomer
phase in an individual fashion. This initiation mechanism for a water-soluble initiator such as
potassium persulfate can be pictured as follows:

The initiation of monomer addition to the sulfate anion radical in the water. phase and the
diffusion of that radical-tailed sulfate anion into the monomer-swollen micelle structures is a key
portion of the emulsion process. The polymerization particles in the emulsion process are
formed from monomer-swollen surfactant micelles. The number and size distribution of the
surfactant micelles are generally independent of any agitation within the reactors. The size of
the micelles are on the order of 150 to 300 and depend on the nature of the surfactant. The
resin formed in an emulsion polymerization is approximately 0.1 to 1 m in diameter, although
particle agglomeration in the drying stage can result in resin of about 20 to 50 m diameter.
The differences in the emulsion process are also reflected in the recipe and process equipment.
The kinetic features of the emulsion polymerization process have been described by Smith and
Ewart and with be examined in Section IX.B.3. The special case of the emulsion polymerization
of vinyl chloride will be examined

Page 140

Page 141

FIGURE 38 Polyvinyl chloride by emulsion process: (a) block flow diagram; (b) polymerization
and stripping; (c) blending and drying.
in the subsequent analysis of the special forms of the Smith-Ewart theory, case 1 and case 3.
1. Polymerization Recipe: A polymerization recipe for a 0.92 inherent viscosity resin is detailed
in Table 22 together with a typical polymerization cycle. The polymerization is typically by a
batch free-radical process.
2. Process Equipment: The product of the emulsion polymerization process can be either a dry
resin or a latex. In case of polyvinyl chloride, the dry resin is the preferred form. In the case of
the copolymers of vinyl chloride and, let us say ethylene, the latex form is preferred. A block
flow diagram of the emulsion process is depicted in Figure 38a. Compared to the suspension
(Figure 12a) and bulk (Figure 33a) processes, the process has the same unit operations as the
suspension process in that there is a need to prepare aqueous solution, separate the product
from the water, and clean up the wastewater stream. If the product were used as an emulsion,
the drying operation would be eliminated and the wastewater treatment would be greatly
reduced.

Page 142
a. Recipe

Water
Vinyl chloride
Sodium lauryl sulfate
Potassium persulfate
Buffers
Polymerization temperature

Charge
Polymerization
Vent
Drop
Prorate cleaning

48,000
40,000
180
400
400
56C

120
100
0.45
1.0
1.0

0.30.4
5.57.0
0.30.4
0.50.6
0.60.8
7.29.2

The polymerization and vinyl chloride stripping operations are illustrated in Figure 38b. Each
reactor, typically of 12,000-gal capacity, is associated with a batch stripping vessel of similar
capacity. The latex product from the reactor is degassed and heated with a purge gas. The
latex particles tend to be easily agglomerated, so countercurrent column stripping is not usually
used in the emulsion process. The residual vinyl chloride level in the emulsion product is
reduced further in the blending and drying process, depicted in Figure 38c.
The emulsion latex, at about 45% solids from the batch stripper, is blended with several
different batches to achieve uniformity in physical properties and is then processed through a
thin-film evaporator to increase the latex solids to about 60 to 65% by reducing the water
content of the latex. The use of membranes processes to reduce the water content of a latex
has also been reported. The process is also used to reduce the residual vinyl chloride level in
the latex. The latex is then pumped into a spray drier, where the latex is sprayed in a fine mist
into a heated atmosphere where the remaining water and vinyl chloride are removed. The spray
drying process tends to aggregate the dry latex particles from their original size of 0.1 to 1.0
m to about 20 to 50 m. The size and size distribution of these aggregates can be controlled
by the spray drier process. The downstream processing of the residual vinyl chloride monomer
in the monomer recovery unit case be similar to the process detailed in Figure 12d.

Page 143
An emulsion polymerization process as described in this section with an annual capacity of 225
million pounds would have the following major equipment:

Polymerization reactor
Batch stripper
Blend tanks
Thin-film evaporator
Spray drier

12,000 gal
14,000 gal
16,000 gal
25,000 lb/hr
25,000 lb/hr

4
4
4
2
2

B.Emulsion Polymerization Kinetics

Emulsion polymerization and suspension polymerization are distinguished by two major
differences: the presence of micelle structures in the aqueous phase of the emulsion process
and the deviation from classical free-radical kinetics regarding polymerization rate and polymer
molecular weight.
1. Micelle Formation: The property of micelle formation by soluble fatty acid salts has been
known for many years [161]. The orientation of the fatty acid, first at the air-water interface and
then at the monomer-water interface, is responsible for the formation of micelle particles. This
has been described by Blackley [162]. The formation of the micelles can be followed by
measuring the change in surface free energy of an aqueous solution of soluble fatty acid salt.
This is illustrated in Figure 39. The critical micelle concentration (CMC) is the point at which
stable aggregates of surface-active

FIGURE 39 Surface free energy vs. surfactant concentration.

Page 144

FIGURE 40 Stages in the emulsion polymerization process: (a) prior to initiation; (b) less than
5% conversion; (c) free surfactant phase gone; (d) free monomer phase gone; (e) all monomer
gone. , surfactant molecules; , monomer; R, initiator radical.
molecules form in the aqueous phase. Emulsion polymerizations are usually run with a
surfactant level higher than the CMC, so that there can be micelle particles in the aqueous
phase.
2. Harkins Qualitative Approach: The emulsion polymerization process exhibits distinct
differences in overall kinetics when compared to suspension polymerization. In the suspension
process the polymerization rate is proportional to the square root of initiator concentration,
while the molecular weight of the resin is inversely proportional to the square root of initiator
concentration. In the emulsion polymerization process it is possible to achieve both

Page 145

FIGURE 40 (Continued)
high rates of polymerization and high-molecular-weight resin [163]. This apparent violation of
classical free-radical kinetics was correctly described by Harkins [164166] as being due to the
formation of surfactant micelles and the subsequent polymerization of the monomers in the
micelles. The Harkins description of the emulsion polymerization process can be summarized by
the five stages in Figure 40.
Stage 1 (Figure 40a): Before the addition of initiator the monomer and fatty acid surfactant are
in equilibrium and are distributed in the aqueous phase as indicated. The micelles are swollen
with monomer and the monomer droplets are coated with surfactant.
Stage 2 (Figure 40b): The initiator (I-I) dissociates in the aqueous phase to two free radicals
which can enter either a monomer droplet or a micelle. The total surface area of all the micelles
is much greater

Page 146
than the monomer droplets, so initiation occurs preferentially in the monomer-swollen micelle
particles. As the polymerization occurs in in the micelles, there is a depletion of monomer in the
micelles, and a net movement of monomer is set up to move from the droplets, through the
water phase, and into the micelles to maintain a fixed ratio of monomer to polymer. For vinyl
chloride/poly vinyl chloride this is 23: 77.
Stage 3 (Figure 40c): As the polymer-swollen micelle grows, there is an increase in the surface
area of these particles, and as a result the free fatty acid surfactant in the aqueous phase is
adsorbed onto the surface of the swollen micelles. At the same time the monomer droplets are
shrinking in size and the fatty acid surfactant is being depleted from the monomer droplet
surface.
Stage 4 (Figure 40d): The monomer droplets have disappeared and the micelles contain all the
polymer and unreacted monomer. Initiator radicals are still being generated in the aqueous
phase and polymerization formation continues in the monomer-swollen, polymer-filled particles.
Stage 5 (Figure 40e): All the monomer is polymerized and the polymer particles are covered
with fatty acid surfactant.
The major features of the Harkins description of the emulsion process are the following:
1. The main site for polymerization are the monomer-swollen micelle particles; the monomer
droplets serve only as a reservoir for monomer.
2. The growth of the swollen micelles leads to a depletion of the fatty acid surfactant in both the
aqueous phase and on the surface of the monomer droplets. There can be no additional
formation of micelle structures since the fatty acid surfactant is below the CMC level in the
aqueous phase.
3. The initiation takes place by the penetration of a free radical into the monomer-swollen
micelle.
3. The Smith-Ewart Quantitative Theory: Smith and Ewart [167] examined the experimental
evidence for the emulsion polymerization of styrene and developed a quantitative mathematical
model that could predict the polymerization rate and the number-average molecular weight of
the polystyrene. In the development of this quantitative theory, Smith and Ewart considered
three general cases:
Case 1: radical concentration << 1 in the micelle particles
Case 2: radical concentration = 0.5 in the micelle particles
Case 3: radical concentration >> 1 in the micelle particles
Case 1 represents a situation that describes vinyl chloride emulsion polymerization. This is

discussed in detail in Section IX.B.3.b. Case 3 represents a situation that described vinyl
chloride dispersion polymerization. This is discussed in detail in Section IX.B.3.c. The general
case 2 approach has been extremely successful in predicting the polymerization parameters
and resin properties for the emulsion polymerization of water-insoluble monomers such as
styrene. The development of the Smith-Ewart theory is best illustrated by examining the
assumptions and final equations for the case 2 system.

Page 147
a. Case 2: General-Emulsion Polymerization: The case 2 modification of the Smith-Ewart
theory makes the assumption that the radicals formed in the aqueous phase are absorbed into
the micelles swollen with monomer and that there is no mechanism for a radical to desorb from
a micelle. As a consequence of the presence of a radical within a monomer-swollen micelle, the
polymerization of the monomer begins. The entry of a second radical will terminate the growing
polymer chain through a mutual termination. The entry of a radical into a monomer-swollen
micelle is totally random, but at any given time exactly half of the existing micelles will have a
growing radical chain while the other half of the micelles will be inactive. Based on this
proposed mechanism, the overall polymerization rate of the system will be

where
KPM
polymerization rate constant in the micelles
[M]M
concentration of monomer in the micelles
N
number of micelles per unit of volume
The important point to note here is that Rp is independent of the rate at which radicals enter the
micelle. The rate of radical entry will affect only the number-average molecular weight of the
resin. The molecular weight of a polymer formed under the reaction conditions will be related to
the average number of monomer units that can add to a growing center before a second radical
can enter the particle and terminate the active chain end. For polystyrene, molecular weights
six to eight times those expected in a bulk or suspension process are found with case 2 SmithEwart process. The situation for vinyl chloride polymerization deviates from this behavior
because chain transfer to monomer is the dominant factor in controlling molecular weight. This
is discussed further in Section IX.B.3.b.
Smith and Ewart examined the experimehtal evidence associated with the generation of
micelles in an emulsion and bracketed the actual value by an overestimation and
underestimation calculations. The results of these studies were that the number of micelles per
unit of volume could be calculated as
N = B[I]2/5[S]3/5
where
[I]
= initial concentration of initiator
[S]
= initial concentration of micelles
B
= constant with upper and lower limits from 0.37 to 0.53
Therefore, the final form for the rate equation of the case 2 Smith-Ewart system will be

[47]

The case 2 form of the Smith-Ewart theory has been examined using several different
approaches by O'Toole [168] and by Ugelstad and coworkers [169] and most recently by
Gardon [170]. The overall corrections of the case 2 assumptions and predictions have been
confirmed by these works. The polymerization of vinyl chloride deviates from the case 2 form of
the Smith-Ewart theory and conforms to either case 1 or case 3. These will be considered

below in further detail.

Page 148
b. Case 1: Vinyl Chloride Emulsion Polymerization: The emulsion of polymerization of vinyl
chloride does not follow the kinetic scheme defined by the case 2 form of the Smith-Ewart
theory. The assumption of the Smith-Ewart theory is that the radical species is immobilized in
the polymer particle. Ugelstad and coworkers [171] found that the combination of high rates of
chain transfer from active polymer end to monomer and high frequency of radical desorption
resulted in a radical concentration in the average active particle of about 10-1 to 5 x 10-4. They
formulated a polymerization rate expression to account for this desorption, related to the
number of polymer particles as
[49]
where
[M]p
concentration of monomer in the polymer particles
N
number of polymer particles per liter
Kdes
rate of radical desorption
Friis and Hamielec [172] have confirmed the validity of this rate expression through
experimental studies. They also have shown that the molecular weight and molecular weight
distribution for polyvinyl chloride prepared in an emulsion process is identical to that of the bulk
and suspension process because of the dominating effect of chain transfer from polymer to
monomer. Therefore, the emulsion process offers no advantage in increased molecular weight
compared to either the bulk or suspension process. The rate expression for the emulsion
process does depend on the number of polymer particles to the 1/6 power [equation (49)] and
as a consequence, an increase in the overall rate compared to either suspension or bulk can be
achieved. Marten

FIGURE 41 Conversion as a function of particle size. Experimental conditions: I0 = 1.72 x 10-3

g; temperature = 50C. [() Data from Ref. 171, (x) data from Ref. 172, () data from Ref.
173.]

Page 149
and Lyngaae-Jorgensen [173] have examined the emulsion rate expression and have
formulated general relationships between polymer conversion and emulsion process variables
of time, initiator level, and number of particles. This is illustrated in Figure 41. Using emulsifiers
with a low CMC level results in a higher level of micelles and therefore a higher level of polymer
particles. At a micelle concentration of ~2 x 10+18 an emulsion process will have about a 10fold increase in overall rate of polymerization compared to a bulk or suspension process.
Therefore, both the number of micelles and the initiator can be used to control the
polymerization rate in an emulsion process.
c. Case 3: Vinyl Chloride Dispersion Polymerization: Vinyl chloride can be polymerized in
emulsion system where the initiator is monomer soluble. This process accounts for
approximately 3% of the total polyvinyl chloride market and although rather small ( 270 million
pounds in 1981), it does illustrate an interesting technology. In the typical dispersion system,
the initiator and monomer are mixed and then emulsified with the water-emulsifier in a highintensity mixer. Since the initiator is monomer soluble, the generation of initiator species by
thermal decomposition will result in a pair of initiator radicals within the monomer droplet. The
Smith-Ewart theory described in Sec tion IX.B.1 predicts that a mutual termination reaction will
occur within a droplet of emulsion dimensions (0.1 to 0.3 m). To avoid this mutual termination
and therefore reduced polymerization rates, there can either be an active radical desorption
mechanism operative as in the case 1 polymerization or the particle size of the monomerswollen droplets must be sufficiently sized to prevent mutual termination from being favored.
Vander Hoff [174] described a subdivision factor by which particle diameter and rate of
polymerization could be examined for the situation where radicals are acquired singly or in
pairs. This is illustrated in Figure 42. The subdivision factor is a function of the average number
of radicals per particle, the kinetics of the process, and the volume of the average droplet. The
Z and a quantities have been specifically defined as

where
n
= average number of radicals per particle
Ri
= rate of initiation
Kt
= rate of termination
NA
= Avogadro's constant

= volume of the particles

The three regions in Figure 42 represent cases 2 and 3 for emulsion polymerization and the
general case for bulk polymerization. In region 1 the volume of the particles are large and
therefore the parameter a is large. Normal free-radical polymerization kinetics are followed in
this region. As the size of the particle decreases and smaller values of a are approached, the
polymerization can proceed along two different reaction pathways, with Z increasing or
decreasing. The case for Z increasing with decreasing a (path 2) occurs for the emulsion
polymerization with the average radical content of

Page 150

FIGURE 42 Subdivision factor Z = /(a/4). (From Ref. 174.)

the particle less than 1. By comparison the case for Z decreasing with decreasing a (path 3)
occurs for the average radical content of the particle greater than 1. A typical size regime for
the merger point of these three regions is about 2 to 5 m.
C.Resin Productivity
The productivity of the emulsion process will depend on how rapidly the heat of polymerization
can be removed from the typical polymerization recipe. The process conditions described for
the suspension and bulk processes (Sections VII and VIII) can be applied to the emulsion
process.
The heat generation process depends on how rapidly the initiator radicals are formed and the
efficiency with which these initiator radicals are converted into polymer radicals. The case 1
and case 3 systems represent two different approaches to controlling the polymerization
kinetics. In case 1 the number of micelles will affect the N/NO ratio and the polymerization can
be speeded up by using a surfactant with a low CMC value or a high concentration of
surfactant. Since the thermal decomposition characteristics of the persulfate initiator are not
easily varied, it may be necessary to feed persulfate incrementally to the reaction mixture to
maintain a constant polymerization rate. The dispersion system operates with organic-soluble
initiators in the particles-size regime of the bulk polymerization process ( >5 m). As such, the
kinetics of polymerization will obey the same general rules for UHGPS as described in Section
VII.B.1.b. Each system represents different process envelopes and offers different
benefit/liability considerations to the manufacturer.
The heat removal process depends on how efficiently the reactor cooling surface can be kept
clean, both in the vapor phase and in the liquid phase. The use of reflux condensers for vaporphase condensation, antifouling agents to supress the formation of wall fouling, and chilled
water in the reactor jacket to increase the total heat removal capacity are techniques practiced
in emulsion and dispersion polymerization systems.

Page 151
D.Resin Quality
The physical and chemical properties of molecular weight, median particle size and particle-size
distribution, and sorption-desorption are important qualities for an emulsion resin. These resins
are used in plastisol applications, where the viscosity of a resin-plasticizer system has to follow
a well-defined time-temperature relationship and in calendered sheet where the fusion and
coalescence of the resin into a uniform film requires a minimum of heat and shear rate.
1. Molecular Weight: The molecular weight of resin prepared in the emulsion process is
controlled by the chain transfer of active polymer to monomer (Cm). This is identical to the
suspension and bulk process, and therefore the molecular weight of the resin is established by
the polymerization temperature. The relationship of polymerization temperature to inherent
viscosity and number-average molecular weight described in Figure 2 can be applied to the
emulsion process.
2. Median Particle Size and Particle-Size Distribution: The emulsion-grade polyvinyl chloride
resin is dried to a powder prior to its fabrication either as a plastisol or in calendered
applications. The median particle size (MPS) and particle-size distribution (PSD) of the resin will
be affected by the PSD of the emulsion and the drying conditions used in the finishing step of
the process.
a. Emulsion Particle-Size Distribution: The effect of the type and level of surfactant on the
emulsion particle properties have been recognized for more than 40 years. Early
experimentalists in the synthetic elastomers field qualitatively judged the particle size of an
emulsion by the blueness of the latex. The bluer the emulsion, the smaller the MPS of the
droplets. This qualitative approach to determining the particle-size properties of an emulsion
has been replaced by a series of instrumental techniques which accurately measure both the
PSD and MPS of an emulsion polymer. These techniques have been reviewed by Davidson and
coworkers [175, 176]. The authors point out that the measuring techniques used to determine
the PSD of an emulsion can introduce a systematic error, and as a consequence they
recommended using several techniques to ensure both precision and accuracy in any
experimental study. The structure of the emulsifier will affect the CMC level for the emulsion
system. Shinoda [177] has shown that the CMC for alkyl carboxylate salts decreases as the
linear alkyl chain is increased from C7 to C13. He also indicated that the pKa value for an
anionic emulsifier would affect the CMC; the alkyl carboxylates have a higher CMC than the
alkyl sulfonates. Rangnes and Palmgren [178] have determined the effect of both emulsifier
type and level on the MPS of the latex particle. They found that the length of the alkyl chain in a
sodium alkylsulfate has an effect on the particle properties of the emulsion. These results follow
the same trend reported by Shinoda [177]. The particle properties of the emulsion can be
controlled by an incremental feed of emulsifier during the polymerization [179] or by the use of
a seed latex [180]. The use of a seed latex and a continuous addition of surfactant can be
employed to prepare emulsions with a narrow PSD. This is achieved by keeping the surfactant
concentration below the CMC level and ensuring that the unreacted monomer is absorbed into
the seed latex particles. The growth of the polymer particles may be controlled in the seed
latex so that the particles make the size transition from the emulsion process to a suspension

process. The advantage of such a process would be that it would combine the

Page 152
kinetic advantages of the emulsion process with the handling characteristics of the suspension
process.
b. Resin Particle-Size Distribution: The emulsion form of a polyvinyl chloride resin is converted
into a solid resin by a careful drying procedure where the solids content of the emulsion is
raised by an evaporation step in a thin film evaporator and the high solids emulsion (60 to 65%)
is then spray dried to a powder. These dewatering steps have an effect on the PSD of the dry
resin and the subsequent end-use application of the resin.
The emulsion particle is making a transition from colloidally dispersed particles of about 0.5 m
diameter to resin agglomerates of about 20 m diameter in a process train that is composed of
thin film evaporators and spray driers. The agglomeration has to occur in a controlled manner
so that the process lines and pumps between the equipment are not plugged. The process flow
diagram in Figure 38c indicates that additional surfactant is added during the dewatering
process. Palmgren [182] has reviewed the process aspects that affect the mechanical stability
of polyvinyl chloride emulsions and their agglomeration. At a given surface concentration of
emulsifier, the stability of the emulsion to mechanical shear agglomeration decreases with
increasing particle size. The stability of the emulsion through the thin-film evaporation and
pumping steps will depend on the added surfactant. There is an upper limit to the total amount
of added surfactant that can be used in this portion of the emulsion process. The additional
surfactant will affect the optical characteristics of the fabricated plastisol or calendered coating;
therefore, the end-use application needs dictate the amount of surfactant that can be used in
the process.
Hoffmann and Saffron [183] have examined the effects of the spray-drying parameters on the
PSD of the dried resin. The process variables they examined were the atomizer tip speed and
the temperature at both the inlet and outlet to the drier. Their results showed that the
agglomerate size of the primary emulsion particles in the dry resin increased with both atomizer
tip speed and drier temperature.
The emulsion resin is formed by a spray-dry agglomeration of a latex. An alternative approach
is to prepare a small-particle-size suspension system. The use of increasing levels of a
cellulosic suspension system is limited because of the relatively high levels of suspending agent
that would be needed. Using the data from Figure 27, it is evident that about 1 to 2% levels of
cellulosic would be needed to form an emulsion-type resin. The in situ preparation of a
surfactant system by the polymerization of vinyl chloride and acrylic acid has been observed
[184]. In a suspension copolymerization of vinyl chloride and acrylic acid (99.5:0.5 by weight)
the copolymer resin had a median particle size of 18 m with 100% of the resin below 37 m
size. The experimental evidence indicates that the initial copolymer has a high acrylic acid
content and this material acts as an anionic surfactant. The first portion of the reaction is an
emulsion, which then collapses in a controlled manner into a small-particle-suspension resin.
The low level of carboxylic acid functions on the polymer chain serve to stabilize both the
emulsion stage and the suspension stage of this process.

X.FUTURE DEVELOPMENTS
During the past decade there has been a wide range of developments in the area of
polymerization kinetics and surface and colloid chemistry for polyvinyl chloride. These
developments are leading to a better understanding of the

Page 153
particle internal morphology and the formation of wall fouling in production-scale reactors. The
future developments in the production of polyvinyl chloride will continue these trends and should
result in processes where the particle properties of the resins, such as porosity and particlesize distribution, can be controlled to a high degree. Further advances in the area of wall fouling
prevention will make a continuous polymerization process a technical possibility. The bulk
polymerization procedure probably will be the area where this will first be practiced, because of
the rapid development in the use of additives as modifiers of particle morphology and the
understanding of the interaction between particle agglomeration and agitation.

SYMBOLS
Kinetic Terms
Kd
homolytic decomposition rate constant for initiator
Kp
polymerization rate constant
Kt
termination rate constant
KTr
transfer rate constant
RI
rate of initiation Rp rate of polymerization
Rt
rate of termination
RTR
rate of transfer
Ed
activation energy for initiator decomposition
Ep
activation energy for polymerization
t
time
T
temperature, absolute
I
initiator
M
monomer
P.
active chain end
P
polymer
t1/2
half-life temperature
f
initiator efficiency factor

kinetic chain length

[]
concentration
Subscripts
0
initial state
t
state at time t
1
first unit
i
ith unit
x
polymer conversion, (M0Mt)/M0
xc
polymer conversion at which free monomer phase disappears
MN
number-average molecular weight
MW
weight-average molecular weight
KH-H rate constant for head-to-head addition
Krr
rate constant for chloro shift in head-to-head form
Krr
rate constant for hydrogen shift

SADT
Q
u
A

self-accelerating decomposition temperature

heat transfer
heat transfer coefficient
area (of cooling jacket)

Page 154

D
N

g
We
d
D

density
diameter (of impellar)
agitator speed
interfacial tension
universal gravitational constant
Weber number
diameter (of particles)
diffusion constant

REFERENCES
1. Sittig M., Vinyl Chloride and PVC Manufacture, Noyes Data Corporation, Park Ridge,
N.J., 1978, p. 2.
2. Chem. Eng. News, 62(26), 15 (1984).
3. Bort, N. D., Marinin, V. G., Kalinin, A. Y. A., and Kargin, V. A., Vysokomol. Soedin., 10,
2574 (1968).
3a. Cotman, J. D., Gonzales, M. F., and Clever, G. C., J. Polym. Sci., Part A-1, 5, 1137
(1967).
4. Wessling R., and Edwards, F. G., in Kirk-Othmer Encyclopedia of Chemical Technology,
2nd ed., Vol. 21, Wiley, New York, 1970, p. 278.
5. Flory, P. J., Principles of Polymer Chemistry, Cornell University Press, Ithaca, N.Y.,
1953, p. 541.
6. Talamini, G., and Gasparini, P., Makromol. Chem., 117, 140 (1968).
7. Abdel-Alim, A. H., and Hamielec, A. E., J. Appl. Polym. Sci., 16, 782 (1973).
8. Pauling, L., The Nature of the Chemical Bond, 3rd ed. Cornell University Press, Ithaca,
N.Y., 1960.
9. Walling, C., Free Radicals in Solution, Wiley, New York, 1957, p. 74.
10. Flory, P. J., Principles of Polymer Chemistry, Cornell University Press, Ithaca, N.Y., p.
75.0
11. Green, J. H. S., and Paisley, H. M., Research (London), 14, 170 (1961).
11a. Schulz, G. V., Z. Phys. Chem., B30, 379 (1935).
12. Flory, P. J., J. Am. Chem. Soc., 59, 241 (1937).

 13. Walling, C., Free Radicals in Solution, Wiley, New York, 1957, p. 159.
14. Barr, J. T., The Manufacture of Plastics, W. M. Smith, ed., Reinhold, New York, 1968,
Chap. 7, p. 315.
15. Flory, P. J., Principles of Polymer Chemistry, Cornell University Press, Ithaca, N.Y.,
1953, p. 112.
16. Hayden, P., and Melville, H., J. Polym. Sci., 43, 201 (1960).
17. Friis, N., and Hamielec, A. E., J. Polym. Sci., 11, 3321 (1973).
18. Friis, N., and Hamielec, A. E., J. Polym. Sci., 12, 251 (1974).
19. Friis, N., and Hamielec, A. E., ACS Poly. Symp., 24, 82 (1976).
19a. Chiu, W. Y., Carratt, G. M., and Soong, D. S., Macromolecules, 16, 348 (1983).
20. Ugelstad, J., Flogstad, H., Hertzberg, T., and Sund, E., Makromol. Chem., 164, 171
(1973).
21. Kuchanov, S. J., and Bort, D. N., Polym. Sci. USSR, 15, 2712 (1973).
22. Olaj, O. F., Angew. Makromol. Chem 1, 47 (1975).

Page 155
23. Hamielec, A. E., and Friis, N., in Polymer Reaction Engineering, Course notes edited
by A. E. Hamielec, June 1977.
24. Abdel-Alim, A., in Polymer Reaction Engineering, Course notes edited by Hamielec, A.
E., May 1982.
25. Langsam, M., J. Appl. Polym. Sci., 23, 867 (1979).
26. Mayer, Z., J. Macromol. Sci.Rev. Macromol., C10, 263 (1974).
26a. Breitenbach, J. W., Olaj, O. F., Reefl, H., and Schindler, A.,Makromol. Chem., 122, 51
(1969).
27. Rigo, A., Palma, G., and Talamini, G., Makromol. Chem., 153, 219 (1972).
28. Park, G. S., and Salleem, M., Polym. Bull., 1, 409 (1979).
29. Abbas, K. B., Bovey, F. A., and Scheilling, F. C., Macromol. Chem. Suppl., 1, 227
(1975).
30. Mitani, K., Ogata, T., Awaya, H., and Tomari, U. J., J. Polym. Sci., Chem., A13, 2813
(1975).
31. Starnes, W. H., Shilling, F. C., Abbas, K. B., Cals, R. E., and Bovey, F. A.,
Macromolecules, 12, 556 (1979).
32. Bovey, F. A., Schilling, F. C., and Starnes, W. H., Polym. Prepr., Am. Chem. Soc. Div.
Polym. Chem., 20, 160 (1979).
32a. Liebman, S. A., D. H. Ahlstrom, Starnes, W. H., and Schilling, F. C., J. Macromol. Sci.
Chem, A17, 935 (1982).
33. Boissel, J., J. Appl. Polym. Sci., 21, 855 (1977).
34. Petieud E., and Quand-Tho-Pham, Makromol. Chem., 178, 741 (1977).
35. Schwenk, U., and Konig, J., Angew. Makromol. Chem., 83, 183 (1979).
36. Caraculacu, A., J. Polym. Sci., Polym. Chem. Ed., 16, 2741 (1978).
37. Michel, A., Schmidt, G., and Guyot, A., Angew. Makromol. Chem., 47, 61 (1975).
38. Michel. A., Burille, P., and Guyot, A., Prepr. MakroMaintz, 603 (1979).
39. Bauer, J., and Sabel, A., Angew. Makromol. Chem., 47, 15 (1975).

 40. Braun, D., and Wolf, M., Angew. Makromol. Chem., 70, 71 (1978).
40a. Kawai, W., and Ichihashi, T., J. Poly Sci., Part, A-1, 10, 1709 (1972).
40b. Ratti, L., Visani, F., and Ragazzini, M., Eur. Polym. J., 9, 429 (1973).
40c. Wescott, L. D., Villacorta, G. M., Schilling, F. C., Plitz, I. M., and Starnes, W. H.,
Polym. Prepr., 50, 290 (1984).
41. Sorvik, E. M., J. Appl. Polym. Sci., 21, 2769 (1977).
42. Abdel-Alim, A. H., and Hamielec, A. E., J. Polym. Sci., 12, 483 (1974).
43. Reding, F. P., Walter, E. R., and Welch, F. J., J. Polym. Sci., 50, 225 (1962).
44. Nakajima, A., Hamada, H., and Hayaski, S., Makromol. Chem., 75, 40 (1966).
45. Abdel-Alim, A. H., and Hamielec. A. E., J. Appl. Polym. Sci., 16, 1093 (1972).
46. Doty, P., Wagner, H., and Singer, S., J. Phys. Chem., 51, 32 (1964).
47. Hlousek, M., and Lanikova, J., J. Polym. Sci., C16, 935 (1967).
48. Chem. Eng. News, 62(24), 39 (1984).
48a. Chem. Mark. Rep., 223(36), 54 (June 27, 1983).
48b. Johnston, C. W., in Encyclopedia of PVC, L. I. Nass, ed., Vol. 1, Marcel Dekker, New
York, 1976, Chap. 3, p. 76.

Page 156
49. Friis, N., AIChE Northeast Reg. Meet., May 1980.
50. Chan, R. K. S., Langsam, M., and Hamielec, A. E., J. Macromol. Sci.Chem., A17, 969
(1982).
51. Hammond, G. S., Sen, J. N., and Boozer, C. E., J. Am. Chem. Soc., 77, 3244 (1955).
52. U. S. Pat. 3,652,681 to Argus Chemical (Mar. 1973).
53. Br. Pat. 1,243,679 to Argus Chemical (Aug. 1971).
54. Br. Pat. 1,304,553 to PennWalt (Jan. 1973).
55. Kamath, V. R., Mod. Plast., 54 (Feb. 1981).
56. Verhelst, W. F., Initiator Seminar, New York, 1980.
57. Tech. Publ. 30.90, Lucidol Division, PennWalt Corporation, Buffalo, N.Y.
58. Fieser, L. F., and Fieser, M., Advanced Organic Chemistry, Reinhold, New York, 1961,
p. 290.
59. Lucidol Brochure 3777-1M.
60. U.S. Pat. 3,997,366 to PPG (Oct. 1976).
61. U.S. Pat. 3,985,720 to PPG (Sept. 1976).
62. U.S. Pat. 3,923,766 to PPG (Dec. 1975).
63. Barter, J. A., and Kellar, D. E., J. Polym. Sci., Chem., 15, 2545 (1977).
64. U.S. Pat. 3,797,716 to Air Products (Mar. 1974).
65. U.S. Pat. 3,801,556 to Air Products (Apr. 1974).
66. U.S. Pat. 3,801,557 to Air Products (Apr. 1974).
67. U.S. Pat. 3,988,261 to PPG (Oct. 1976).
67a. Kissin, Y. V., and Beach, D. L., J. Polym. Sci., Chem., A21, 1065 (1983).
67b. Langsam, M., J. Polym. Sci., Polym. Lett., 22, 549 (1984).
68. Soni, Y., and Albright, L. F., J. Appl. Polym. Sci., Appl. Polym. Symp., 36, 113 (1981).
69. Albright, L. F., and Soni, Y., J. Macromol. Sci.Chem., A17, 1065 (1982).

 70. Albright, L. F., and Bild, C. G., Chem. Eng., 121 (Sept. 15, 1975).
71. Burgess, R. H., in The Manufacture and Processing of PVC, Macmillan, New York, 1982,
p. 28.
71a. U.S. Pat. 4,298,576 to Wacker-Chemie (Nov. 1981).
72. Markovitz, R. E., Hydrocarbon Process., Aug. 1973.
73. Cameron, J. B., Lundeen, A. J., and McCulley, Jr., J. H., Hydrocarbon Process., 39
(Mar. 1980).
74. Terwiesch, B., Hydrocarbon Process., 117 (Nov. 1976).
75. U.S. Pat. 3,980,628 to Conoco (Sept. 1976).
76. U.S. Pat. 4,136,242 to Shinetsu (Jan. 1979).
76a. Beckmann, B., Chemtech, 304 (May 1973).
77. Cazias, J. L., Schechter, R. S., and Wade, W. H., ACS Symp. Ser., 8, 234 (1975).
78. Sarkar, N., annual meeting, Institute of Food Technologists, Philadelphia, June 5, 1977.
79. Jap. Kokai 54-50059 to Kanegafuchi Chemical (Apr. 1979).
80. Br. Pat. 1,424,645 to British Petroleum (Nov. 1975).
81. Br. Pat. Appl. 2,074,589A to Kureha (Nov. 1981).
82. Hamielec, A. E., Gomez-Vaillard, R., and Marten, F. L., J. Macromol. Sci.Chem., A17,
1005 (1982).
83. U.S. Pat. 4,451,625 to Air Products (May 1984).
84. U.S. Pat. 4,224,424 to Air Products (Sept. 1980).

Page 157
85. Br. Pat. 1,534,245 to Dow Chemical (Nov. 1978).
86. Ger. Pat. 2,440,434 to B.F. Goodrich (Mar. 1975).
87. U.S. Pat. 4,105,838 to Shinetsu (Aug. 1980).
88. Hanson, C. M., J. Paint Technol., 39, 104 (1967).
89. Burrell, H., Polymer Handbook, 2nd ed., Brandrup J., and Immergut, E. H., eds., Wiley,
New York, 1975.
90. Riddick, J. A., and Bunger, W. A., Organic Solvents, 3rd ed., Wiley, New York, 1970.
91. U.S. Pat. 4,009,048 to Air Products (Feb. 1977).
92. U.S. Pat. 3,998,665 to SIR (Dec. 1976).
92a. Hjerberg T., and Sorvik, E., Polym. Prepr., 25(1), 69 (1984).
92b. Langsam, M., unreported results.
93. Plackett, R. L., and Burman, J. P., Biometrika, 33, 305 (1946).
94. Box, G. E. P., and Behnken, D. W., Technometrics, 2, 447 (1960).
95. The Design and Analyses of Industrial Experiments, 2nd ed., O. L. Davies, ed.,
Longman, London, 1978.
96. Box, G. E. P., Hunter, W. G., Hunter, J. S., Statistics for Experimenters, Wiley, New
York, 1978.
97. ASTM Procedure D-1243-79, 1980 Annual Book of ASTM Standards American Society
for Testing and Materials, Philadelphia, 1979, p. 489.
98. Tompa, H., in Polymer Solutions, pp. 263291.
99. Miller, M. L., in The Structure of Polymers, Reinhold, New York, 1966, pp. 181236.
100. Kurata, M., Tsunashiama, Y., Iwama, M., and Kamada, K., in The Polymer Handbook,
2nd ed., J. Brandrup and E. H. Immergut, eds., Wiley, New York, 1975, p. IV14.
100a. Ravey, M., and Waterman, J. A., J. Polym. Sci., Chem., A13, 1475 (1975).
101. Young, L. J., and Fischer, J. P., in The Polymer Handbook, 2nd ed., J. Brandrup and E.
H. Immergut, eds., Wiley, New York, 1975, p. 1157.

102. Mayo, F. R., J. Am. Chem. Soc., 65, 2324 (1943).

103. Walling, C., Free Radicals in Solution, Wiley, New York, 1957, p. 50.
104. Langsam, M., J. Appl. Polym. Sci., 21, 1057 (1977).
105. U.S. Pat. 4,187,363 to Goodyear (Feb. 1980).
106. Br. Pat. Appl. 2,001,661A to Air Products (Feb. 1979).
107. Eliassaf, J., Chem. Ind. Dev., Aug. 1973, p. 24.
108. Eliassaf, J., J. Macromol. Sci.Chem., A8, 459 (1974).
109. Eliassaf, J., J. Appl. Polym. Sci., 17, 1625 (1973).
110. Geil, P. H., J. Macromol. Sci.Phys., B14, 171 (1977).
111. Allsopp, M. W., Pure Appl. Chem., 53, 449 (1981).
112. Zichy, E. L., J. Macromol. Sci.Chem., A11, 1205 (1977).
113. Wilson, J. C., and Zichy, E. L., Polymer, 20, 264 (1979).
114. Cooper, W. D., Speirs, R. M., Wilson, J. C., and Zichy, E. L., Polymer, 20, 265 (1970).
115. Rance D. G., and Zichy, E. L., Polymer, 20, 266 (1979).
116. Davidson, J. A., and Witenhafer, D. E., J. Polym. Sci. Phys., 18, 51 (1980).
117. Hopff, H., Lussi, H., and Gerspacher, P., Makromol. Chem., 78, 24, 37 (1964).
118. Hopff, H., Lussi, H., and Hammer, E., Makromol. Chem., 82, 175, & 184, (1965).

Page 158
119. Hopff, H., Lussi, H., and Hammer, E., Makromol. Chem., 84, 274, 282, 286 (1965).
120. Belopolisky, A. O., Konkhin, Yu. A., Mukhanov, A. A., Smirnov, Yu. A., and Planovsky, A.
N., Khim. Neft. Mashinostr., 2, 22 (1975).
121. Johnson, G. R., J. Vinyl. Technol., 2, 138 (1980).
122. Lewis, M. H., and Johnson, G. R., J. Vinyl Technol., 3, 102 (1981).
123. Cheng, J. T., APCI unpublished results.
124. Testa F., and Vianello, G., J. Polym. Sci., C27, 69 (1969).
125. Br. Pat. 1,543,341 to Sigman Italiana (Apr. 1979).
125a. Bugada, D. C., and Rudin, A., Polym. Mater. Prepr., 50, 280 (1984).
126. Cheng, J. T., J. Macromol Sci.Phys., B20, 365 (1981).
127. Cheng, J. T., and Langsam, M., J. Macromol Sci.Chem., A21, 395 (1984).
128. U.S. Pat. 3,882,195 to Air Products (May 1975).
128a. Lukas, R., Tyrackova, V., and Kolinsky, M., J. Appl. Polym. Sci., 29, 901 (1984).
129. Hopfenberg, H. B., Permeability of Plastic Films and Coatings to Gases, Vapors and
Liquids, Plenum Press, New York, 1974.
130. Stern S. A., and Frisch, H. L., Annu. Rev. Mater. Sci., 11, 523 (1981).
131. Daniels G. A., and Procter, D. E., Mod. Pack., 45 (Apr. 1975).
132. Berens, A. R., and Hopfenberg, H. B., in Recent Developments in Separation Science, N.
Li, ed., CRC Press, Cleveland, Ohio, 1979, p. 293.
133. Berens, A. R., Polym. Prepr., 15, 203 (1974).
134. Berens, A. R., Polymer., 18, 697 (1977).
135. Berens, A. R., J. Vinyl Technol., 1, 8 (1979).
136. Crank, J., The Mathematics of Diffusion, Clarendon Press, Oxford, 1956.
137. Patel, C. B., Guptar, R., and Chan, R. K. S., J. Polym. Sci., Chem., A17, 3775 (1979).
138. Patel, C. B., Grandin, R. E., Guptar, R., Reynolds, C., and Chan, R. K. S., Polym. J., 11,
43 (1979).

139. U.S. Pat. 4,117,220 to Air Products (Sept. 1979).

140. Ger. Pat. 2,612,572 to British Petroleum (Oct. 1976).
141. Jap. Kokai 51-86586 to Kanegafuchi (July 1976).
142. Br. Pat. 1,487,820 to Conoco (Oct. 1979).
143. Ger. Pat. 2,855,146 to Chisso (June 1979).
144. Br. Pat. 1,547,051 to Shinetsu (June 1979).
145. Jap. Kokai 52105,106 to Toa Gosei (Sept. 1977).
146. Br. Pat. 1,551,758 to B.F. Goodrich (Aug. 1979).
147. Davidson, J. A., J. Appl. Polym. Sci., 27, 3219 (1982).
148. Schwab, P. A., and Wingrave, J. A., J. Macromol. Sci.Phys., B20, 429 (1981).
149. Quy, R. B., J. Macromol Sci.Phys., B20, 235 (1981).
149a. Balakirskaya, V. L., and Shtarkman, B. P., Colloid J. USSR, 27, 252 (1965).
149b. U.S. Pat. 4,198,376 to Rhone-Progil (Apr. 1980).
150. Elastomerics, 26 (Mar. 1980).
151. U.S. Pat. 4,277,585 to Rhone-Progil (July 1981).

Page 159
152. Br. Pat. 1,334,489 to Rhone-Progil (Oct. 1973).
153. Fischer, N., and Goiran, L., Hydrocarbon Process., 143 (May 1981).
154. Tobolsky, A. V., J. Am. Chem. Soc., 80, 5927 (1958).
155. Noury Chemical, Initiator Half-Life Table.
156. Chatelain, J., Br. Polym. J., 5, 457 (1972).
157. Marks, G. C., in Developments in PVC Technology, J. H. L. Henson and A. Whelan, eds.,
Applied Science Publishers, London, 1973, p. 17.
157a. Tornell, B. E., and Uustalu, J. M., J. Vinyl Technol., 4, 53 (1982).
157b. Fischer, M., J. Vinyl Technol., 6, 35 (1984).
158. U.S. Pat. 4,271,444 to Rhone-Poulenc (Aug. 1980).
159. Blackley, D. C.., in Emulsion Polymerization, Applied Science Publishers, London, 1975,
Chap. 3.
160. Gardon, J. L., in Polymer Processes, C. E. Schildknecht and I. Skeist, eds., Wiley, New
York, 1977, Chap. 6.
160a. Min, K. W., and Ray, W. H., J. Macromol. Sci.Rev. Macromol. Chem., C11, 177
(1974).
160b. Min, K. W., and Ray, W. H., J. Appl. Polym. Sci., 22, 89 (1978).
161. G. S. Harley, Aqueous Solutions of Paraffin Chain Salts, Hermann, Paris, 1936.
162. Blackely, D. C., in Emulsion Polymerization, Applied Science Publishers, London, 1975,
Chap. 1.
163. Mark, H., and Tobolsky, A. V., in Physical Chemistry of High Polymeric Systems,
Interscience, New York, 1950, p. 451.
164. Harkins, W. D., J. Chem. Phys., 13, 381 (1945).
165. Harkins, W. D., J. Am. Chem. Soc., 69, 1428 (1947).
166. Harkins, W. D., J. Polym. Sci., 5, 217 (1950).
167. Smith, W. V., and Ewart, R. H., J. Chem. Phys., 16, 592 (1948).

168. O'Toole, J. T., J. Appl. Polym. Sci., 9, 1291 (1965).

169. Ugelstad, J., Mork, P. C., and Aasen, J. O., J. Polym. Sci., A5, 2281 (1967).
170. Gardon, J. L., Rubber Chem. Technol., 43, 74 (1970).
171. Ugelstad, J., Mork, P. C., Dahl, P., and Rangnes, P., J. Polym. Sci., C27, 49 (1969).
172. Peggion, E., Testa, F., and Talamini, G., Macromol. Chem., 71, 173 (1964).
173. Marten, F. L., and Lyngaae-Jorgensen, J., Polym. Colloids Prepr., NATO Advanced
Study Institute, June 1975.
174. VanderHoff, B. M. E., in Polymerization and Polycondensation Process, Adv. Chem. Ser.
34, R. F. Gould, ed., American Chemical Society, Washington D.C., 1962, pp. 631.
175. Collins, E. A., Davidson, J. A., and Daniels, C. A., J. Paint Technol., 47, 35 (May 1975).
176. Daniels, C. A., McDonald, S. A., and Davidson, J. A., in Emulsions, Latices and
Dispersions, P. Becker and M. H. Yudenfreund, eds., Marcel Dekker, New York, 1978, pp.
175193.
177. Shinoda, K., in Colloidal Surfactant: Some Physio Chemical Properties, K. Shinoda, T.
Nakagawa, B. Tamamushi and T. Isemura, eds., Academic Press, New York, 1963, pp. 196.
178. Rangnes, P., and Palmgren, O., J. Polym. Sci., C-33, 181 (1971).

Page 160
180. Gardon, D. L., and Weidner, K. R., in Emulsion Polymers and Emulsion Polymerization,
D. R. Bassett and A. E. Hamielec, eds., ACS Symp. Ser. 165, American Chemical Society,
Washington, D.C., pp. 515532.
181. Ugelstad, J., Longstad, H. F., Hansen, F. K., and Ellingsen, T., J. Polym. Sci., 12, 473
(1973).
182. Palmgren, D., in Emulsion Polymerization, I. Piirma and Gardon, J. L., eds., ACS Symp.
Ser. 24, 1976, American Chemical Society, Washington, D.C., pp. 258275.
183. Hoffman, D., and Saffron, P. M., in Emulsion Polymers and Emulsion Polymerization, D.
R. Bassett and A. E. Hamielec, eds., ACS Symp. Ser. 165, American Chemical Society,
Washington, D.C., 1981, pp. 209224.
184. Langsam, M., unpublished results.
185. Talamini, G., and Vindetto, G., Makromol. Chem., 100, 48 (1967).
186. Bovey, F. A., Hood, F. R., Anderson, E. W., and Kornegay, R. I., J. Phys. Chem., 71,
312 (1967).
187. Cavalli, L., and Corraro, G., IIT Newslett., Nov. 1967.
188. Garbuglio, G., Rodella, A., Borsini, G. C., and Gallinella, E., Chim. Ind. (Milan) 46, 166
(1964).

Page 161

4 Copolymer Processes and Manufacture

MICHAEL LANGSAM
Air Products and Chemicals, Inc. Allentown, Pennsylvania
I. INTRODUCTION

162

II. SCOPE AND SIZE OF THE COMMERCIAL COPOLYMER MARKET

162

III. COPOLYMERIZATION PROCESS

164

A. Copolymerization Theory

164

B. Kinetic Aspects of Copolymerization

178

IV. COPOLYMER STRUCTURE AND PROPERTIES

178

A. Rheology

178

B. Thermal Stability and Processability

195

C. Glass Transition Temperature

195

V. OVERVIEW OF COMMERCIAL PROCESSES

VI. VINYL ACETATE COPOLYMERS
A. Suspension Polymerization
B. Solution Polymerization
C. Emulsion Polymerization
VII. OLEFIN COPOLYMERS
A. Overview
B. Propylene Copolymers
C. Ethylene Copolymers
VIII. ACRYLIC-ACRYLONITRILE COPOLYMERS
A. Acrylate Esters
B. Acrylonitrile

199
200
200
213
215
216
216
216
223
229
229

230
IX. FUTURE DEVELOPMENTS
SYMBOLS
REFERENCES

231
231
232

Page 162

I. INTRODUCTION
Despite the fact that polyvinyl chloride (PVC) was first prepared in 1835 [1], it was not until the
late 1920s that any significant work was done on the melt fabrication of this resin. In retrospect
we can view this time lag as being due, in part, to the inherent thermal instability of the polyvinyl
chloride resin [2]. Much of this early work in the processing of the resin was done by rubber
fabricators, who used their existing laboratory equipment for processing studies. This
equipment would prove to be inadequate and combined with the lack of proper thermal
stabilizers, would result in many years of delay in the commercialization of polyvinyl chloride
resin. A major breakthrough in the processing of the homopolymer resin was developed by
Waldo Semon of B.F. Goodrich [3], who discovered the effects of external plasticizers on melt
rheology and processibility of the resin. At the same time the general approach of
copolymerization was being explored as a means of modifying the physical properties of
different resins and elucidating the mechanism of polymerization. Copolymerization of vinyl
chloride with certain monomers became an alternative way to reduce the severity of the
processing conditions needed to fabricate the resin into finished articles.
The copolymerization of vinyl chloride monomer with other reactive monomers and polymers to
form new products has been a technical challenge to polymer scientists for the past 50 years.
In the 1930s and 1940s the field of copolymerization became an organized technology and
several distinct classes of materials were identified as typical of the copolymerization process.
These included random copolymerization, block copolymerization, and graft copolymerization. A
comparison of the structures of these different materials has been reviewed in many
introductory texts [4, 5] and needs no further illustration. In the recently issued Tenth Collective
Index [6] of Chemical Abstracts there are listed no fewer than 130 different types of
copolymers of vinyl chloride and a second reactive species and more than 350 different ter-orgreater polymers with vinyl chloride. A review of this recent body of knowledge would be a very
difficult task to complete in a text of this nature. It will be the goal of this chapter to review the
recent developments in the areas of commercial copolymer products where vinyl chloride is the
major constituent in the resin and the material is formed from a random copolymerization
process. The copolymers formed by graft and block processes are discussed in Chap. 10.
This chapter discusses the physical and chemical properties of the commercial copolymers and
how they exhibit different properties compared to the homopolymer resin. In addition, the
manufacturing aspects of the commercial copolymer resins are compared to homopolymer
processes.

II. SCOPE AND SIZE OF THE COMMERCIAL COPOLYMER MARKET

During the 25-year period from 1935 to 1960 the principal goal of commercial manufacturers of
polyvinyl chloride was to develop a resin that combined the properties of good thermal stability
during melt processing, together with chemical resistance and physical toughness in the
fabricated article. All these goals were to be achieved at a cost-effective price compared to

other materials of construction. The relatively poor melt properties of vinyl chloride
homopolymer essentially prevented the large-scale commercial application of this resin during
the 1940s and 1950s. In 1962 the commercial sales of vinyl chlo-

Page 163
ride/vinyl acetate resins reached a level of 295 million pounds [7]. In 1974 the market volume
for vinyl chloride copolymers expanded and reached a peak of approximately 608 million
pounds per year. At the same time renewed efforts were directed at improving the melt
rheological properties of the homopolymer resin and investigating new market applications for
the homopolymer resin. Technical advances in the areas of resin molecular weight control,
thermal stabilizers, and process aids as well as processing equipment resulted in a rapid
growth for the homopolymer resin. From 1966 to 1970 the commercial sales of homopolymer
resin quadrupled from 120 million pounds to 475 million pounds. The homopolymer resins found
new major applications in the siding and piping markets, while the copolymers were used mainly
for phonograph records and asbestos floor tile, both old established markets.
In the 1970s the U.S. market for vinyl chloride homopolymer resins grew very rapidly and
reached a total sales of 5.4 billion pounds by 1982 [8]. By contrast, the market for all vinyl
chloride copolymers declined during the decade and several classes of vinyl chloride
copolymers were withdrawn from commercial production. The suspension-grade vinyl
chloride/vinyl acetate copolymers decreased in total U.S. sales from about 500 million pounds
in 1972 to 190 million pounds in 1982. The reduction in sales of suspension-grade vinyl
chloride/vinyl acetate resin has been very dramatic during the past 4 years. This is illustrated in
Table 1. This decline has represented an average 7% reduction in sales each year for the past
decade. A market of about 150 million pounds is predicted for these copolymers in the mid1980s based on the future of the phonograph record and floor tile markets. Any increase in
market demand requires the identification and development of a new market for copolymer
resin. This reduction in copolymer demand forced Air Products to mothball its Calvert City vinyl
chloride/vinyl acetate production facilities and left Borden and Tenneco as major producers and
Keysor-Century, Formosa Plastics, and General Tire as minor producers of suspension-grade
vinyl chloride/vinyl acetate resin in the United States. The solution-grade Vinylite resins
produced by Union Carbide are the only domestically produced solution-grade vinyl
chloride/vinyl acetate resins. Their application is primarily in the can-coating market. There have
been no plans announced to expand this solution-grade resin market.
Vinyl chloride/acrylonitrile polymers typified by DuPont's Orlon FLR are sold in volumes of about
11 million pounds per year into the fiber market. Vinyl chloride/propylene copolymers were
produced by Air Products through

Year

Sales volume
(million pounds)

1979
1980
1981
1982

375
285
235
190

Page 164
1978 but were withdrawn from commercial sales in that year. Vinyl chloride/ ethylene
suspension copolymers of Union Carbide were also removed from commercial sale. In 1982 the
commercial vinyl chloride copolymer market in the United States represented only about 5% of
the total vinyl chloride resin market, with the vinyl chloride/vinyl acetate resin representing the
major portion of this copolymer market.
The principal thrust of this chapter is to review the structure-property developments during the
past 10 years for both vinyl chloride/vinyl acetate resins and for vinyl chloride/olefin resins,
indicating the effects of the comonomer on the properties of the copolymer resins and on the
manufacturing technology and the process economic aspects for these copolymers.

III. COPOLYMERIZATION PROCESS

The copolymerization of vinyl chloride with other monomers was one of the approaches used in
the 1930s and 1940s to develop commercial products and to elucidate the polymerization
mechanism. Copolymer processes and products are more complex to analyze than a
corresponding homopolymer process or product, but the copolymer process and product offer
a wider range of materials for potential commercial application, and the presence of at least
two reacting comonomers offers a means of following the process. The subject of vinyl chloride
copolymerization will be examined from the following two aspects: copolymer formation and the
effects of comonomers on the manufacturing process.
A. Copolymerization Theory
1. Historical Background: One of the technical objectives of a copolymerization process is to
modify the physical or chemical properties of a homopolymer so as to improve the product for a
particular application. The modification of the chemical and physical properties of a metal by
combining or alloying was recognized by early technical cultures; the formation of hard alloys
from soft metals such as copper and tin to form bronze. Another is the low-melting alloy called
Wood's metal made from cadmium, tin, antimony, and bismuth. In the nineteenth century these
alloying techniques were used to manufacture Celluloid [9] from cellulose nitrate and camphor.
Simple mechanical mixing of the low-molecular-weight camphor with the polymeric cellulose
nitrate was sufficient to form a uniform, more resilient, and tougher product for film and billiard
balls.
Early attempts at mixing different combinations of high-molecular-weight polymers were not
successful because of the incompatibility of the polymer blends. The phenomena associated
with compatible polymer mixing were subsequently described by Flory and coworkers [10]. This
topic has recently been reviewed by Obalisi and coauthors [11] and Paul and Newman [12].
Compatibility between two polymers does not usually occur because the free energy of mixing
is greater than zero.
Fmix = Hmix - T Smix 0
[1]
Thermodynamically, solubility occurs only when the Fmix is negative. This occurs when the
enthalpy of mixing term (Hmix) is more negative than the entropy of mixing term (T Smix). If

the total free energy of mixing is posi-

Page 165
tive, the polymers will not form a homogeneous mixture at both the visible and microscopic
levels. Notable exceptions are the cases where there are interactions between two polymers.
Examples of this are the hydrogen bonding in mixtures of polyvinyl chloride with
polycaprolactone [13] and the blends of polyacrylates which are used as impact modifiers in
polyvinyl chloride compounds [14]. It should be noted that while these mixtures of polymers are
visibly compatible (i.e., clarity in film form), at a domain level there is segregation between the
polymers. The question of polymer-polymer compatibility should be defined in a relative
framework, at either a macroscopic or microscopic range. For this work we consider
macroscopic (visible light) compatibility as a reference point.
An alternative to mixing individual polymers is to combine different monomers and to polymerize
the mixture into a random copolymer. This technique was practiced in a pragmatic fashion
during the late 1930s and the random copolymers of many different compositions were
prepared and examined. The relationship between the initial monomer composition and the
instantaneous copolymer composition was first described successfully by Wall [15]. He showed
that the relationship of initial copolymer composition (A/B) to initial comonomer composition
(a/b) was
A/B = constant (a/b)
[2]
The constant in Wall's relationship was related to the relative rate of addition of monomer a and
b to the growing chain. This relationship was expanded by Mayo and Lewis [16] and Alfrey and
Goldfinger [17], who recognized that in a copolymerization of monomers M1 and M2 there
would be four distinctly different propagation steps. These were described as

with the propagation rate constants of K11, K21, K22, and K12, respectively. The rate of
polymerization for monomer 1 and monomer 2 could then be represented as

The relative rate of consumption of monomer 1 or monomer 2 could now be determined by a

simple division of equation (7) by equation (8) to yield

Page 166
The concentration of active ends, [M1]. and [M2], can be reduced to a single term by
assuming that the rate of formation of M1 and the rate of its disappearance must be equal,
leading to the following steady-state condition:
Theil [18] has recently reexamined the steady-state assumptions in the copolymer equation
using a statistical probability approach and has shown quite succinctly the correctness of the
original derivation.
Equation (9) can now be rearranged as

The factors K11/K12 and K22/K21 are the reactivity ratios for the monomers and are
designated as R1 and R2, respectively. The monomer reactivity ratios for the binary system of
vinyl chloride/X are listed in Table 2. This analysis of the copolymerization reaction was based
on the interaction between the terminal group of the polymer chain and the two monomers in
the reaction mixture. The interaction between the next nearest neighbor (penultimate group) and
the terminal group may also affect the manner in which the monomers are added to the
polymer chain. Mertz and coworkers [19] developed a mathematical relationship for a
propagation system where the next nearest neighbor was considered as being effective in the
copolymerization reaction. This procedure increased the number of propagation reactions for a
two-monomer system from four to eight. The effect of this penultimate group was verified
experimentally by Ham [20] in a series of studies on the highly polar copolymerization of the
styrene-fumaronitrile system. This penultimate system has not been needed in the analysis of
the vinyl chloride copolymer system.
2. Reactivity Ratios and Copolymer Composition: The composition of a copolymer formed at
any instant of time in a copolymerization process is determined by the molar concentration of
the monomers at the reaction site and the reactivity ratios of the two comonomers. There are
three major types of random copolymers formed when the reactivity ratios of the two
comonomers are different. These can be illustrated using vinyl chloride as monomer 1 (with
reactivity ratio R1) as follows.
a. R1 > 1 and R2 < 1: This can be illustrated by vinyl chloride/vinyl acetate (where R1 = 1.68
and R2 = 0.23) [21]. In this pair of monomers the vinyl chloride comonomer polymerizes
preferentially compared to the vinyl acetate comonomer. As a consequence, the initially formed
copolymer is rather low in vinyl acetate relative to the comonomer composition. This condition
shifts during the copolymerization and the copolymer formed at any instant has an everincreasing amount of vinyl acetate. This effect of shifting composition was described
qualitatively by Thomas and Hinds [22], who reported that for an initial comonomer ratio of vinyl
chloride to vinyl acetate of 85:15 (weight ratio) the composition of the copolymer varied from
8% bound vinyl acetate formed during the first 5% conversion to 23% formed in the ninetieth
percentile conversion. This shift in composition is illustrated in Figure la for a 95:5 initial

comonomer system. Both the instantaneous copolymer composition and the cumulative
copolymer composition are plotted against conver-

Page 167
1) and Other Monomers R1 M2 R2 T(C) Referencea 0.13 Abityl acrylate 4.2 60 1 0.001
Acenaphthylene 64 2 0 Acrylamide 19.6 40 3 0.02 Acrylonitrile 3.28 60 4 0.04 Acrylonitrile 2.8
40 5 0.074 Acrylonitrile 3.7 50 6 0.052 Acrylonitrile 3.6 50 7 0.04 Acrylonitrile 2.7 60 8 1.2 Allyl
acetate 40 9 1.16 Allyl acetate 0 40 10 2.0 Allyl triethoxysilane 0 50 11 0.74
Bicyclo(2.2.1)-2,5-heptadiene 0.35 50 12 1.41 Bicyclo(2.2.1)-2,5-heptadiene 0.08 50 13 0.06
Bornyl methacrylate 12.5 60 14 0.035 Butadiene 8.8 50 15 3.4 Butene 0.21 60 16 0.4 2Butenyl triethoxysilane 0 50 11 0.07 Butyl acrylate 4.4 45 17 5 t-Butylethylene 0 70 18 0.05
Butyl methacrylate 13.5 45 19 0.35 Butyl vinyl sulfonate 0.30 70 20 0.70 2-Chloroallyl acetate 0
100 21 1.13 1-Chloro-1-propene 0.24 50 15 0.75 2-Chloro-1-propene 0.58 50 15 2.53
Chlorotrifluoroethylene 0.01 60 22 0.2 -Chlorovinyltriethoxysilane 0 50 11 8.8 cis-Butene-2
0.001 16 0.4 3-Cyclohexenyltriethoxy-silane 0 50 11 1.4 Dubutyl maleate 0.0 40 23 5.39 1, 2Dichloroethylene 0.072 50 24

Page 168
R1

M2

R2

T(C)

Referencea

0.65
0.12
0.42
0.06
0.8
0.77
0.9
0.06
5.74
0.053
0.06
0.5
1.93
3.60
1.85
0.04
0.14
0.12
2.05
1.3
4.3
2.2
0.296
0.31
0.06
0
0.12
0.083
0
2.96
7.54
0

Diisobutyl maleate
Diethyl fumarate
Di-2-ethylhexyl maleate
Diethyl itaconate
Diethyl maleate
Diethyl maleate
Diethyl maleate
Diisopropyl itaconate
Diketene
Dimethyl itaconate
Dioctyl itaconate
Dioctyl maleate
n-Dodecyl vinyl ether
Ethylene
Ethylene
Glycidyl mathacrylate
Hydroxypropyl acrylate
Isobornyl methacrylate
Isobutylene
Isobutylene
Isobutylene
Isopropenyl acetate
Maleic anhydride
Methallyl chloride
Methyl acrylate
Methyl acrylate
Methyl acrylate
Methyl acrylate
N-Methylol acrylamide
3-Methyl-1-butene
2-Methyl-2-butene
Methyl methacrylate

0.1
0.47
0
5.65
0
0.009
0
6.0
0.2
5.0
7.0
0
0.15
0.24
0.20
8.84
4.3
10.0
0.08
0.03
0
0.25
0.008
0
4
5
4.4
9.0
23.5
0.03
0
12.5

40
60
68
50
40
60
70
50
60
50
60
68
50
90
70

25
26
27
28
29
26
18
28
30
31
28
27
32
33
34
25
1
14
4
35
36
37
38
10
17
9
1
6
39
40
40
9

60
60
60
0
65
65
75
45
45
50
50
50
60

60

Page 169
R1

M2

R2

T(C) Referencea

0.1
0.02
0.22
0.10
2.10
0.12
1.90
0.04
0.43
0.5
8.0
2.35
0.035
0.02
0.067
0.077

0.045
1.6
7.3
0.15
0.90
2.1
1.8
1.68
1.35
3.74
1.7
0.72
1.35
2.0

Methyl methacrylate
Methyl methacrylate
2-Methyl-1-vinyl imidazole
Methyl vinyl ketone
n-Octadecyl vinyl ether
Octyl acrylate
n-Octyl vinyl ether
Octyl methacrylate
Pentachlorostyrene
1-Pentene
Propenyl triethoxysilane
Propylene
Styrene
Styrene
Styrene
Styrene
Styrene
Styrene
6-Triethoxysilyl bicyclo-(2.2.1)-2-heptene
trans-Butene-2
trans-Trimethyl aconitate
Bis(trimethylsiloxy)vinyl methyl silane
Vinyl acetate
Vinyl acetate
Vinyl acetate
Vinyl acetate
Vinyl acetate
Vinyl benzoate
Vinyl benzoate
Vinyl butanoate
Vinyl isobutyl ether

10
15
2.13
8.3
0.1
4.8
0.1
14.0
5.3

0
0.24
5.7
17
35
35

12.4
0
0.2
0
0.50
0.3
0.6
0.23
0.65
0.033
0.5
0.28
0.65
0.02

68
45
40
70
50
45
50
45

68
50

75
60
50
50

50

60
50
68
40
60
40
-32
40
45
40
50

21
19
41
42
32
17
32
19
43
27
11
44
45
46
47
6
48
43
11
16
49
50
27
51
52
53
54
53
19
55
56

Page 170
R1

M2

R2

T(C)

Referencea

0.17
4.7
1.8
1.35
3.2
5.2
1.0
0.3
0.5
0.23
0.14
0.2
0.2
0.017
0.0093
2.0
0.874
0.963
7.4
1.40
1.0
1.16
0.745
0.84
0.60
0.7
1.458
1.35
0.02
0.53
0.38
0.9

N-Vinylcarbazole
Vinyl caprate
Vinyl caproate
Vinyl caproate
Vinyl caprylate
Vinylene carbonate
Vinyl ether diethoxysilane
Vinylidene chloride
Vinylidene chloride
Vinylidene chloride
Vinylidene chloride
Vinylidene chloride
Vinylidene chloride
Vinylidene cyanide
Vinylidene cyanide
Vinyl isobutyl ether
Vinyl 4-ketostearate
Vinyl 12-ketostearate
Vinyl laurate
Vinyl levulinate
Vinyl methyl diethoxysilane
Vinyl nonanoate
Vinyl octadecanoate
N-Vinyl oxazolidinone
N-Vinyl oxazolidinone
Vinyl phenyl diethoxysilane
Vinyl pinonate
Vinyl propinonate
4-Vinyl pyridine
N-Vinyl pyrrolidone
N-Vinyl pyrrolidone
Vinyl triethoxysilane

4.77
0.2
0.1
0.65
0.2
0.09
0
3.2
7.5
3.15
Large
4.5
1.8
0.54
0.72
0.02
0.320
0.248
0.2
0.419
0
0.282
0.290
0.35
1.50
0
0.446
0.65
23.4
0.38
0.53

40
40
40
86
80
50
60
47
55
68
50
45
50

60
60

60
50
60
60
50

50
60
40
60

50

2
57
57
57
57
58
11
9
59
60
27
15
55
61
62
47
63
63
57
64
11
64
64
65
66
11
64
57
58
67
67
11

Page 171
R1 M2

R2

T(C)

Referencea

0.8 Vinyl triisopropoxysilane

0.8 Vinyl trimethoxysilane
1.06Vinyl undecanoate

0
0
0.358

50
50
60

11
11
64

a1. Marvel, C. S., and Schwen, R., J. Am. Chem. Soc., 79, 6003 (1957).
2. Imoto, M., Kobunski Kagaku, 18, 747 (1961).
3. Matsuoka, M., Takemoto, K., and Imoto, M., Kogyo Kagaku Zaszhi, 69, 134 (1960).
Lewis, F. M., Walling, C., Cummings, W., E. R. Briggs, and W. J. Wenisch, J. Am. Chem.
4.
Soc., 70, 1527 (1948).
5. U.S. Pat. 2,511,593 to U.S. Rubber (June 1950).
6. Chaplin, E. C., Ham, G. E., and Fordyce, R. G., J. Am. Chem. Soc., 70, 538 (1948).
7. Thompson, B. R., and Raines, R. H., J. Polym. Sci., 41, 265 (1959).
8. Ashakari, N., and Nishimura, A., J. Polym. Sci., 31, 250 (1958).
9. Mayo, F. R., Lewis, F. M., and Walling, C., J. Am. Chem. Soc., 70, 1529 (1948).
10. U.S. Pat. 2, 356, 871 to PPG Industries (Aug. 1944).
11. Thompson, R. B., J. Polym. Sci., 19, 373 (1966).
12. Zutty, N. L., J. Polym. Sci., Part A-1, 2231 (1963).
Zabolotskaya, E. V., Khodzhemijov, V. A., Gantmakher, A. R., and Mednedev, S. S., Bull.
13.
Acad. Sci. USSR, 140, 964 (1961).
14. Imoto, M., Otsu, T., Tserda, H., and Ito, T., J. Polym. Sci., Part A-2, 1407 (1964).
Tkachenoko, G. V., Khomikovskii, P. H., Abkin, A. D., and Medvedev, S. S., Zh. Fiz. Khim.,
15.
31, 242 (1957).
Takemoto, K., Matsuoka, K., Dtsuka, M., and Imoto, M., Kogyo Kagaku Zaschi, 69, 2026
16.
(1966).
Tkachenko, G. V., Stupen, L. V., Kofman, L. P., and Frolova, L. Z., Zh. Fiz. Khim., 31, 2676
17.
(1957).
18. Alfrey, T., Bohrer, J., Haas, H., and Lewis, C., J. Polym. Sci., 5, 719 (1950).
Tkachenko, G. V., Stupen, L. V., Kofman, L. P., and Raracheva, L. A., Zh. Fiz. Khim., 32,
19.
2492 (1958).
20. Overberger, C. G., Baldwin, P. E., and Gregor, H. P., J. Am. Chem. Soc., 72, 4864 (1950).
21. U.S. Pat. 2, 419, 221 to W. O. Kengen (Apr. 1947).
22. Klewan, N., and Lazar, M., Chem. Prum., 9, 668 (1959).
23. Kimura, T., and Yoshida, K., Kagaku To Kogyo (Osaka), 28, 158 (1954).
24. Kitamura, R., Vrata, M., and Takeuchi, T., J. Appl. Polym. Sci., 9, 166 (1965).
25. Kimura, T., and Yoshida, K., Kagaku To Kogyo (Osaka), 29, 43 (1955).
26. Lewis, F. M., and Mayo, F. R., J. Am. Chem. Soc., 70, 1533 (1948).
27. Agron, P., Alfrey, T., Bohrer, J., Haas, H., and Wechsler, H., J. Polym. Sci., 3, 157 (1948).
28. Nagai, S., and Yoshida, K., Kobunshi Kagako, 17, 77 (1960).

Page 172
29.Kimura, T., and Yoshida, K., Kagaku To Kogyo (Osaka), 20, 288 (1955).
30.Matsuoka, M., Takemoto, K., and Imoto, M., Kogyo Kagako Zaschi, 69, 134 (1960).
31.Nagai, S., Uno, T., and Yoshida, K., Mobunski Kagaka, 15, 550 (1958).
32.Kazome, G. A., Sakai, S., Choshi, Y., and Murai, K., Kobunshi Kagaku, 17, 478 (1960).
33.Burkhart, R. D., and Zutty, N. L., J. Polym. Sci., Part A-1, 1137 (1963).
34.Erussalimsky, B., Duntoff, F., and Tumarkin, N., Makromol. Chem., 66, 205 (1963).
Sheinker, A. P., Yakovlena, M. K., Kristalyni, B. V., and Abkin, A. D., Dokl. Akad. Nauk.
35.
SSSR, 124, 632 (1959).
36.U.S. Pat. 2, 379, 292 to Standard Oil Development Co. (June 1945).
37.Petrova, G. A., Shtraikhman, G. A., and Vansheidt, A. A., Zh. Fiz. Khim., 33, 1246 (1959).
38.deWilde, M. C., and Smets, G., J. Polym. Sci., 5, 253 (1950).
39.Matsuoka, M., Takemoto, K., and Imoto, M., Kogyo Kagaku Zaschi, 69, 142 (1966).
40.Otsuke, M., Takemoto, K., and Imato, M., Osaka Univ. Rep., 8 (1966).
41.Matsuoka, M., Takemoto, K., and Imato, M., Kogyo Kagaku Zaschi, 69, 137 (1966).
Alfrey, T., Arond, L., and Overberger, C. G., Copolymerization. Vol. 8 of The High Polymer
42.
Series, Wiley-interscience, New York, 1952.
Tkachenko, G. V., Etlis, V. S., Stupen, L. V., and Kafman, L. P., Zh. Fiz. Khim., 33, 25
43.
(1959).
44.Yamazaki, N., Aritoni, M., and Kaubara, S., Jap. Chem. Soc. Annu. Meet. 47103 (1966).
45.Hart, R., Makromol. Chem., 47, 143 (1961).
46.Doak, K. W., J. Am. Chem. Soc., 70, 1525 (1948).
47.Platzer, N., in Copolymerization, G. E. Ham, ed., Interscience, New York, 1952, p. 803.
48.Ham, G. E., J. Polym. Sci., 38, 543 (1959).
Marvel, C. S., Johnson, T. W., Economy, J. P., Scott, J. P., Taft, W. H., and Lable, B. G., J.
49.
Polym. Sci., 20, 437 (1956).
50.Pike, R. M., and Bailey, P. L., J. Polym. Sci., 22, 55 (1956).
Marvel, C. S., Jones, G. P., Mastin, T. W., and Schertz, G. L., J. Am. Chem. Soc., 64, 2356
51.
(1942).
Mayo, F. R., Walling, C., Lewis, F. M., and Halse, W. F., J. Am. Chem. Soc., 70, 1523
52.
(1948).
53.Kimura, T., and Yoshida, K., Kagaku To Kogyo (Osaka), 32, 269 (1958).
54.B Baker, W. P. J. Polym. Sci., 42, 578 (1960).
55.Poltz, H., Ind. Eng. Chem., 47, 472 (1955).
56.Chapin, E. C., Ham, G. E., and Mills. C. L., J. Polym. Sci., 4, 597 (1949).
57.Kimura, T., and Yoshida, K., Kagaku To Kogyo (Osaka), 32, 223 (1958).

Page 173
58.Hayashi, K., and Smets, G., J. Polym. Sci., 27, 275 (1958).
59.Bengough, W. L., and Norrish, R. G. W., Proc. R. Soc., Lond., A 218, 155 (1953).
60.Embree, W. H., Mitchell, M. J., and Williams, H. L., Can. J. Chem., 29, 253 (1951).
Gilbert, H., Miller, F. F., Averill, S. J., Carlson, E. J., Folt, V. L., Heller, H. J., Stewart, F. D.,
61.
Schmidt, R. F., and Trumball, H. L., J. Am. Chem. Soc., 78, 1669 (1956).
62.Can. Pat. 510, 359 to V. L. Folt (1955).
63.Marvel, C. S., Dykstra, T. H., and Mogne, F. C., J. Polym. Sci., 62, 369 (1962).
64.Marvel, C. S., and DePierre, W. G., J. Polym. Sci., 27, 39 (1958).
65.Bork, J. F., and Coleman, L. E., J. Polym. Sci., 43, 413 (1960).
66.Hart, R., and Timmerman, D., Makromol. Chem., 31, 223 (1959).
67.Beitenbach, J. W., and Edelhauser, H., Ric. Sci., 25, 242 (1956).
sion. The wide range in the instantaneous vinyl acetate composition is an indication of the rather
broad composition distribution of this copolymer system.
b. R1 < 1 and R2 > 1: This system can be illustrated by vinyl chloride/ methyl methacrylate (R1
= 0.1, R2 = 10) [23]. The methyl methacrylate monomer will polymerize during early stages of
the polymerization and the unreacted monomer mixture will very rapidly become almost
exclusively vinyl chloride after 45% conversion. Attempts to modify the properties of the vinyl
chloride resin will be unsuccessful since the copolymer will be essentially a block of methyl
methacrylate and a block of vinyl chloride. This system is illustrated in Figure 1b. Approaches
that use an incremental feed system of methyl methacrylate to even out the copolymer
composition have been employed commercially.
c. R1 < 1 and R2 < 1: This system can be illustrated by vinyl chloride/N-vinyl pyrrolidone (R1 =
0.53, R2 = 0.38) [24]. In this system the active chain end will add the other comonomer
preferentially, so that an alternating copolymer of vinyl chloride-N-vinyl pyrrolidone will form as
long as the unreacted monomer mixture has both comonomers. This system is illustrated in
Figure 1c.
3. Determination of Reactivity Ratios for Comonomers: The experimental accuracy of the
reactivity ratio value depends on measuring the following factors:
1. Composition of the initial comonomer system at the polymerization site
2. Analysis of the copolymer composition
3. Conversion
and then using some form of the copolymer equation (11) developed in Section III.A.1 to
measure R1 and R2. Before detailing the recent developments in

Page 174

Page 175

FIGURE 1 (a) Level of bound vinyl acetate vs. polymer conversion; (b) level of bound methyl
methacrylate vs. polymer conversion; (c) level of bound N-vinyl pyrrolidone vs. polymer
conversion.
the determination of the reactivity ratio, it would be appropriate to indicate the experimental
aspects of the measurements.
a. Comonomer Composition: In vinyl chloride suspension polymerization, the polymer phase
separates from the monomer phase after about 0.1% conversion. This phase separation also
occurs in copolymerization in which vinyl chloride is a major component. The partitioning of the
comonomer system into the polymer phase can lead to major deviations in the comonomer ratio
at the polymerization site. Similar problems have been encountered with comonomer
suspension systems where one monomer is very soluble in the aqueous phase, as in the case
of acrylic acid at aqueous pH levels above 8.0. One method of determining if comonomer
partitioning is occurring in a polyvinyl chloride copolymerization would be to compare the
copolymer composition for a heterogeneous and a homogeneous system. If there are no
differences in copolymer compositions, there is not a significant comonomer partitioning effect.
If there is a major difference in copolymer composition, the comonomers are partitioning
between the monomer and polymer phase and the magnitude of the partition coefficient needs
to be determined.
b. Polymer Composition: The copolymer composition of a material can be determined by a
number of analytical procedures. The chloride analysis, while convenient and familiar, needs to
be replicated several times to ensure accuracy in the results. An absolute error of 0.57% in the
chloride analysis will result in a 1% error in the weight percent of the comonomer.
Spectroscopic

Page 176
adsorption provides a good approach to measuring the comonomer content of appropriate
copolymers. The infrared carbonyl adsorption at 5.74 cm-1 has proven very useful in wellcalibrated systems for measuring the vinyl acetate content of copolymers. Mori [24a] has
reported an analytical system for vinyl acetate/vinyl chloride systems which simultaneously
measure both molecular weight by gel permeation chromatography (GPC) procedures and vinyl
acetate content by infrared adsorption. In principle, a dual adsorption-chromatography system
can be used on a wide variety of copolymers to measure simultaneously molecular weight
distribution and composition.
c. Conversion: In any study of reactivity ratios it is important to keep the conversion to a low
level, typically less than 5%. This will prevent any significant drift in the composition of the
unreacted comonomer. Several techniques, such as short-stopping at low conversions or rapid
venting of unreacted monomers, have been used. If a total material balance in a system is
needed, these procedures tend to introduce experimental errors that are difficult to quantify.
The dead-stop polymerization system as described by Tobolsky [24b,c] is a method for
preparing low-conversion copolymers in a fashion which does not disturb the phase equilibrium
of the system.
The first approach to determining the reactivity ratios for comonomers was the graphical
technique described by Mayo and Lewis [21]. This graphical technique was then modified by
Fineman and Ross [25] to the linear, but nonsymmetrical with respect to R1 and R2, form

where f1 is the mole fraction of component 1 in the monomer and F1 is the mole fraction of
component 1 in the polymer. This linear equation [as represented in equation (12)] leads to R1
as the slope and R2 as the intercept. In this nonsymmetrical form the replacement of monomer
1 for monomer 2 can result in a second and different set of reactivity ratios. Yezrielev and
coworkers [26] developed a linear equation that overcame the lack of symmetry in the earlier
work [25]. A nonlinear approach to the reactivity ratio equation was proposed by Tidwell and
Mortimer [27] which reduced the errors associated with using the data with the largest value of
f2/F.
Several excellent review articles [28, 29, 30] have recently been published comparing these
different approaches to measuring quantitatively the reactivity ratios for different comonomer
systems. The symmetrical linear technique of Yezrielev [26] was recommended in all the review
articles as being superior to the Fineman-Ross [25] technique. The nonlinear procedure of
Tidwell is reputed to be the most accurate [28] for matching copolymer composition to reactivity
ratio despite its more complex nature.
4. The Q-e Scheme: The reactivity of an active chain end with a monomer such as vinyl chloride
can be depicted by the following reaction:

Page 177
The convention used is that the polymer main chain is in the plane of the paper and that the
substituents add in a head-to-tail manner with the interaction between the free radical at the
end of the polymer chain and the orbitals of the monomer. The overall reaction of adding a
single monomer unit to an active chain end involves resonance, polar, and steric interactions
between the active chain end and the monomer. The reaction is represented as reversible: KPwill be favored under certain conditions, notably in conditions where the monomer/active chain
end is above the ceiling temperature; KP+ will be favored when the total free energy of the
system is reduced by polymerization. In this illustration we are not considering the placement of
the chlorines into any atactic, syndiotactic, or isotactic preference. The illustration is merely a
depiction of the addition of a monomer unit to an active chain end.
In the mid-1940s numerous attempts were made to relate the physical properties of the active
chain end and the monomer to the measured reactivity ratio of the copolymers. Following the
pioneer work of Hammett [31] on the hydrolysis of substituted aromatic esters, various workers
[32, 33] examined the electron donating and electron accepting properties of the monomers
and the transition state. Alfrey and Price [34] extended these earlier efforts to a combined
approach which unified both the resonance stabilization and the polarity of the monomer and the
active chain end into a semiempirical model called the Q-e scheme. The main assumption of
this approach is that the rate constant for the addition of monomer 1 to active chain end 2 will
be
K21 = P2Q1 exp(-e2e1)
[14]
where
P2 = characteristic of the chain end 2
Q1 = reactivity of monomer 1
e1, e2 = measures of the polarity of the monomer 1 and active chain end 2
When comparing the relative rate constants for two monomers with the same active chain end,
the P component is eliminated as

Although the Q-e scheme is generally regarded as an empirical method of correlating the
structure/reactivity ratio, several other workers have extended this concept. Wall [35] has
separated the polarity of the monomer and adduct so that the expression for R1 becomes

which can improve the fitting of the reactivity ratio. Levinson [36] has used linear combinations
of atomic orbital techniques to determine the electron affinity and relate it to the e term and to
the localization energy and then to relate it to the Q term. One of the major advantages of the
Q-e scheme is that these parameters are a measure of a monomer's activity, not strictly a
monomer pair parameter. Knowing the Q-e values for different monomers

Page 178
would permit the calculation of reactivity ratios for all the copolymer pairs without
experimentation. Recent work by Greenley allows the calculation of Q-e for a wide range of
materials with different monomers [37]. This work has been extended to telogens [38] and
would be very effective in a study of monomer impurities.
B. Kinetic Aspects of Copolymerization
The polymerization of vinyl chloride was described in Section IV Chapter 3 as being a
heterophase system with the polymer phase separating from monomer at about 0.1%
conversion. This phase separation results in a monomer-rich phase and a polymer-rich phase
(see Figure 5, Chapter 3). The development of a predictive model that defines the rate of
polymerization for a vinyl chloride copolymer system using parameters such as monomerpolymer partition coefficients [see equation (32), Chapter 3] has not been reported. Except for
the rare case of a comonomer azeotrope situation, the composition of the comonomer phase
changes with conversion for copolymer processes. The change in the composition of the
comonomer phase results in the potential for a wide range of partition coefficient values over
the conversion range. Due to the rather low commercial interest in the vinyl chloride copolymer
market, there has been little reported work on predictive copolymer kinetic models. A first step
in developing a predictive copolymer rate expression would be quantification of the partition
coefficients for copolymer-comonomer mixtures over the range of polymerization temperatures.
Since the intent of any commercial polymerization process is to maximize reactor productivity,
the optimization of the copolymer kinetics has been done using an evolutionary operations
procedure with different initiator combinations in a well-instrumented reactor (see Section
VII.B.1, Chapter 3). In Section V of this chapter the effects of increased levels of vinyl acetate
monomer and propylene monomer are discussed in detail.

IV. COPOLYMER STRUCTURE AND PROPERTIES

A. Rheology
Rheology is the study of flow properties of materials and the relationship of these flow
properties to the molecular structure of the materials. A major aspect in the commercial
feasibility of polyvinyl chloride resins is the ability to convert the particulate resin into a finished
commercial object. During the past 30 years a wide range of commercial fabricating processes
have been developed for homopolymer and copolymer resins. Menges and Berndtsen [39] have
reviewed these techniques for both rigid and plasticized formulations and their results are
summarized in Figure 2. The efficient conversion of a particulate resin into a finished
commercial object requires an understanding of the rheological aspects of the resin.
The topic of polymer rheology has been reviewed in a number of recent books and articles [40
43], and these are recommended for background information. The flow characteristics of a
polymeric material depend, in part, on such aspects as the intrinsic molecular properties of the
resin and the extrinsic system properties. These properties can be summarized as:

Page 179

FIGURE 2 (a) Processing in a rigid formulation; (b) processing in a flexible formulation. (From
Ref. 39.)
1. Intrinsic molecular properties
a. Composition
b. Molecular weight
c. Molecular weight distribution
d. Branching
e. Crystallinity
2. Extrinsic system properties
a. Temperature
b. Pressure
c. Compounding ingredients

d. Design of the processing equipment

Page 180
The polymer structure will have a profound effect on the melt-processing characteristics of a
resin. The tacticity of the vinyl chloride homopolymer prepared over the polymerization range 25
to 60C varies in a very small degree [44], and there are about 10 short branches for every
1000 repeat units [45] with few long-chain branches. The crystallinity of the homopolymer has
been measured and found to be about 5 to 15% for a wide range of normal polymerization
conditions [46]. The copolymerization of vinyl chloride with different comonomers can result in
large changes in the polymer structure, and these changes will result in significant changes in
the rheological properties of the resin. Two copolymer structures that result in major changes in
the rheological properties of these resins are the long-chain branching present in vinyl acetate
copolymers [47] and the reduction of chain-chain polar interactions present in propylene
copolymers [48]. The branched polymers will affect the melt elasticity of the polymer [49], for
example to modify the groove reproducibility of phonograph records. The propylene copolymers
will reduce the viscous component of the polymer melt [50]. The effects of copolymers on these
properties compared to PVC homopolymer are the topic of this section.
1. Polyvinyl Chloride Homopolymer. The rheology or flow properties of polyvinyl chloride resin
tend to be more complex than those of other commercial thermoplastics, such as polystyrene
or polyethylene, because of the following factors:
1. Particulate or supermolecular flow effects
2. Thermal dehydrochlorination effects
3. Viscoelastic flow effects
The dehydrochlorination of polyvinyl chloride resin is due, in large part, to the branched
structure. This topic was discussed in detail in Section V.A of Chapter 3, which should be
referred to for details. The supermolecular flow properties and viscoelastic flow properties of
polyvinyl chloride resin both affect the melt processing of the resin and can be used as an
indication of the nature of the polymer structure. Since the aim of a fabrication process is to
maximize the economic conversion of resin into finished product, an understanding of these
effects is vitally important.
a. Particulate Flow: The particulate flow properties of polyvinyl chloride were first described by
Berens and Folt [5153]. They observed that the particle structure of an emulsion-polymerized
polyvinyl chloride was discernible in a fabricated article after certain extrusion conditions. This
observation of particle structure in fabricated articles was examined by Collins and coworkers
[54, 55] as a means of understanding the unusual melt viscosity/temperature relationship for
polyvinyl chloride resins. Melt viscosity of a resin should be related to temperature by an
Arrhenius-type relationship as
[17]

where
= melt viscosity
A = constant
EA = activation energy
R = gas constant
T = absolute temperature

Page 181
This linearity had not been found by earlier workers [56, 57]. Collins and coworkers [54] found,
by extremely careful experimentation, that there were two flow regimes for polyvinyl chloride.
This is illustrated in Figure 3. At a constant shear rate, there are distinctly different slopes to
the log versus 1/T curve. The activation energy, in kcal/K mol for this polymer, was found to
decrease at higher shear rates and, by extrapolation, would probably coincide at shear rates of
>1000 see-1. This is shown in Figure 4. Collins and coworkers [55] also examined the effects
of both molecular weight and chain-chain association on melt flow properties of polyvinyl
chloride. They found that the temperature of polymerization affected the syndiotacticity of the
resin and at the same molecular weight affected the melt viscosity of the polyvinyl chloride
resin. This is illustrated in Figure 5. These effects were summarized by Sieglaff [50] and
Lyngaae-Jorgensen [59]. Recently, Lyngaae-Jorgensen [60] has extended this study of the melt
viscosity of polyvinyl chloride resin to temperature ranges where the chain-chain interactions
were reduced to minimum levels. He found that above 240C the melt behaved as a
monomolecular system.
b. Viscoelastic Flow: The processing characteristics for polyvinyl chloride resin and compounds
have been examined by the viscous flow properties of the material. In Section IV.A.1.a, the
particulate flow properties of the resin were identified by viscosity versus shear rate
relationships. In many cases this viscous flow versus shear rate is adequate to characterize a
resin or to compare small changes in resin properties between two different lots of the same
materials. The melt viscosity can easily be determined in a capillary rheometer and the appeal
of this type of measurement is the ease with which it can be measured and the straightforward
nature of the interpretation of the data: Increased melt viscosity predicts a more difficult
processing situation.

FIGURE 3 Melt viscosity vs. reciprocal temperature at constant shear rate. (From Ref. 54.)

Page 182

FIGURE 4 Shear rate vs. activation energy. (From Ref. 54.)

FIGURE 5 Melt viscosity vs. shear rate: flow curves for three PVCs of the same molecular
weight polymerized at 5, 40, and 70C. (From Ref. 55.)

Page 183
The problem with using melt viscosity as the processing index for a resin is that it neglects the
elastic component in a molten resin and the effect that the elastic component will have on the
melt rheology. The melt elasticity of a polymer is evident from the die swell phenomenon or
Barus effect [61]. Racin and Bogue [61a] have examined the effects of molecular weight and
molecular weight distribution for polystyrene resins. They report that the low-molecular-weight
fraction in the polystyrene resin is responsible for the die swell effects. Benbow and Lamb
[61b] have examined the various theoretical approaches to explaining extrusion die swell for a
wide range of polymers. Sieglaff [61c] and Paradis [61d] have focused this on polyvinyl chloride
compounds.
The melt elasticity becomes very important when comparing different types of polymers and, in
particular, polymers with different types of branching. The melt elasticity is more sensitive to
branching and changes in the molecular weight distribution than is the melt viscosity.
The elastic and viscous components of a polymer can be compared to the mechanical analogs
of a spring and a dashpot. These concepts can therefore be used to relate the stress () to the
strain () or rate of strain (d/dt). The elastic component for the stress (e) of a melt will
therefore be analogous to a spring and will be

This analysis indicates the e will be in phase with strain and will be 90 out of phase with
strain.
In a substance with viscoelastic characteristics, both viscous stress and elastic stress are
present when the material is strained. The quantitative relationship of e and can be
measured by a variety of commercial equipment. Two particular manufacturers are
Rheometrics, which manufactures the mechanical spectrometer, and the Seiscor Division of
Seismograph Services Corp. of Tulsa, which offers two different on-line rheometers, one de-

Page 184
veloped by C. D. Han and the other developed by A. S. Lodge [62]. The advantages to these
on-line instruments are that they can follow the effects of variations in processing conditions or
compound additives on the melt rheology, and therefore more easily optimize the thermal
processing of resin to finished product. A general review of laboratory and commercial
rheometers was prepared by VanWeizer [62a]. This should also be examined for additional
details on this topic.
Rheological studies on the flow characteristics of the homopolymer have been used to analyze
the effects of blending [63] and thermal stability [64, 65]. 65]. Park and Riley [63] examined the
viscoelastic properties of homopolymers of 56,000 and 123,000 weight-average molecular
weight. They found that the high-molecular-weight portion of the blend affected both the viscous
and elastic characteristics of the melt. These effects will be discussed in the section on blends
of homopolymer and vinyl acetate/vinyl chloride copolymers for compression-molded
phonograph records.
Thermal stability properties of polyvinyl chloride can be predicted in a Brabender Plasticorder
torque rheometer [64] and in a slit die extruder [65]. Using these thermal stability and rheology
techniques, it is possible to maximize the output of an extruder and/or minimize the raw
materials cost of a compound.
The melt rheology of vinyl chloride copolymers is more difficult to interpret because of the
influence of the comonomer in the polymer main chain. Undoubtedly, the comonomer can affect
the rheology through changes in such factors as crystallinity, chain branching, chain-chain
interaction, and chain flexibility. The thrust of any of this work would be to develop a copolymer
in which melt processability was improved without affecting the mechanical or thermal stability
properties of the resin. This task has been very difficult. The goal of using polyvinyl chloride
copolymers to replace plasticized homopolymer has not succeeded. The viscoelastic behavior
of plasticized homopolymer has not been duplicated by copolymers with immobilized
plasticizer chains. The economic factors of price and performance for the copolymer would also
be important. Although a large number of copolymers have been prepared over the past
decade, extensive studies on the relationship between polymer structure and melt rheology
have tended to be directed at copolymers of vinyl acetate and copolymers of 1-olefins.
2. Vinyl Acetate Copolymers: Some of the commercially available vinyl acetate copolymers are
tabulated in Table 3, where the inherent viscosity (IV) and bound vinyl acetate (BVAc) levels are
noted. The resin characteristics of inherent viscosity and bound vinyl acetate level, when
graphically displayed (Figure 6), indicate that these commercial materials fall into three product
types:
Type IV/BVAc

Application

1
2

Primary resin for audio records Asbestos filled floor tile

Calendered film and sheet stock

0.450.50/1213
0.600.70/910

0.750.80/56

Secondary resin for audio records

Page 185
a Manufacturer Grade Percent vinyl acetateb Inherent viscosityc Tenneco 315R 12.513.0
0.47049 316R 12.4 0.46 318R 12.013.0 0.470.48 389 11.513.0 0.460.55 382 8.09.0
0.670.70 305 5.9 0.75 400 8.5 0.62 Borden 126 10.010.6 0.570.59 113R 14.414.6 0.49
0.51 171C 10.1 0.68 Keysor KR-1 4.6 0.77 KR-2 5.6 0.76 Occidental 4301R 13.6 0.46 aAll
analyses performed by Air Products, Calvert City quality control laboratory. Ranges represent
limits measured on multiple samples of the same grade received in the period 19781980.
bpercent vinyl acetate determined by infrared adsorption band at 5.74 m. cInherent viscosity

by ASTM method D1243-60-A.

FIGURE 6 Inherent viscosity vs. bound vinyl acetate for commercial vinyl chloride/vinyl acetate
copolymers.

Page 186
The rheology properties of these copolymers have been studied by Rangnes [66] over a limited
range of 12 to 15% BVAc, and Mondval and coworkers [67] have developed master curve
relationships of reduced viscosity to shear rate for several commercial copolymers. Powell and
Khanna [68a,b] extended this work to cover a wider range of BVAc (6 to 17%) and also
examined the effects of molecular weight and copolymer sequence on rheological properties at
several different processing temperatures. The resins used by Powell and Khanna are listed in
Table 4.
Comparison of copolymer resins 1 to 6 from this study at different processing temperatures
illustrates the combined effects of molecular weight and vinyl acetate content on the apparent
viscosity at a shear rate of 1 sec-1. This is illustrated in Figure 7. As the processing
temperature and vinyl acetate content of the resin are increased, the apparent viscosity
decreases. A comparison of copolymer resins 7 and 8 demonstrates the effects of long blocks
of vinyl acetate from normal batch technique versus random placement of vinyl acetate units
from delayed addition technique on melt rheology. The random placement of the vinyl acetate
groups results in a lower apparent melt viscosity, and Powell and Khanna ascribe this to a
higher level of branching. This is illustrated in Figure 8. This supposition could be tested by a
measurement of the elastic shear stress (e), but as yet this has not been reported in the open
literature. They do report a higher die swell ratio for the delay addition copolymer (resin 8)
which does support their conclusion of higher elasticity for the delay addition material.
A comparison of resins 9, 10, and 11 illustrates the effects of decreased polymerization
temperature (or increased chain transfer agent). The resin prepared at 70C has a lower
viscosity over the entire shear rate range. This is illustrated in Figure 9. The reduction in melt
viscosity over this 10C range of polymerization may be due to increased branching in the
higher-temperature polymerization systems or to a small but subtle difference in crystallinity of
these three resins.
This work of Powell and Khanna indicates that the elasticity component of melt rheology for the
vinyl acetate copolymers may be controlled by varying the branching component in the resin. A
systematic study of the effects of both polymerization temperature and vinyl acetate monomer
feed could represent a novel rheological control scheme for these resins. In a later work
Khanna [69] briefly reviewed the effects of mixtures of resins on the melt viscosity/shear rate
relationships and the experimental procedures needed to study audio record manufacturing. He
indicated that a successful molding operation required a correct balance of melt viscosity and
melt elasticity in a record compound.
Langsam and Cheng [70] have examined the effects of both polymerization conditions and
copolymer structure on the melt rheological properties of vinyl chloride/vinyl acetate
copolymers. They have developed empirical equations which predict successfully both the melt
index of a copolymer from Tinius-Olsen measurements and the viscous modulus/elastic
modulus/complex viscosity from mechanical spectrometer measurements for a wide range of
commercial copolymers. Their results are based either on the copolymer physical properties of

inherent viscosity and bound vinyl acetate or on the copolymerization process parameters of
polymerization temperature/charged vinyl acetate/chain transfer agent level. They have also
demonstrated that copolymer resins with the same melt index can have different complex
viscosity values at different shear rates.

Page 187

Copolymer

Vinyl acetate
(wt %)

spa

Polymerization
temperature (C)

1
2
3
4
5
6

6.5
12.5
17.2
6.2
12.2
17.2

0.24165
0.24165
0.24165
0.27265
0.27265
0.27365

22
22
22
22
22
22

14.0

0.25765

22

14.0

0.25765

22

9
10
11

12.7
12.4
12.5

0.27360
0.27265
0.27370

22
22
22

aSpecific viscosity in nitrobenzene at 25C.

Source: Ref. 68.

Dumping pressure
Remark
(psig)

Normal batch
technique
Delayed addition
technique

Page 188

FIGURE 7 Apparent viscosity vs. vinyl acetate content. (From Ref. 68.)

FIGURE 8 Apparent viscosity vs. temperature for copolymers 7 and 8 showing the difference
between delayed monomer addition () and normal batch-polymerized copolymer ( ). (From
Ref. 68.)

Page 189

FIGURE 9 Apparent viscosity vs. shear rate at 151C for copolymers 9, 10, and 11
polymerized at different temperatures: 0, 60C; , 65C; , 70C. (From Ref. 68.)
Their experimental work [70] was based on a 4 x 4 matrix which spanned the inherent viscosity
range from 0.4 to 0.7 dl/g and the bound vinyl acetate range from 4 to 16%. The
copolymerization conditions and copolymer physical properties are detailed in Table 3A together
with the melt index values and mechanical spectrometer results. The melt index for the
copolymers could be predicted from an empirical equation based either on the copolymer
physical properties [equation (22a)] or on the copolymerization processes [equation (22b)].
They are noted below:

The constants in empirical equations (22a) and (22b) are different and can be used only within
the process and product envelope defined by Table 5. The success of predictive equations
(22a) and (22b) can be demonstrated by a comparison of the actual and predicted melt index
values detailed in Table 6. Melt index contours for these copolymers can be constructed using
the relationships defined in equations (22a) and (22b). Figure 10a demonstrates the effects of
copolymer physical properties on the melt index values and Figure 10b demonstrates the
effects copolymerization conditions on melt index values. In a similar manner the values for
viscous modulus, elastic modulus, and complex viscosity can also be predicted from model
equations using either the copolymer properties or the copolymerization process condition.

Page 190

Experiment

VAcC
(%)

Tp TCE
(C) (phm)

MI
VAcB
(g/10) (%)

IV
Yield
(dl/g) (%)

62.5 0

0.6

4.4

0.725 87.2

70.0 0

2.6

3.8

0.587 84.0

73

0.88

29.0

3.8

0.501 86.5

73

2.20

109.0 3.8

0.414 87.9

10

60

3.4

8.2

0.707 87.6

10

67.5 0

14.0

8.0

0.609 87.3

10

73

0.43

65.0

8.1

0.495 84.4

10

73

1.71

210.0 8.1

0.419 83.5

1.92 x.
106
7.80 x
105
1.17 x
105
4.03 x
104
1.34 x
106
5.66 x
105
1.81 x
105
1.48 x
104

1.10 x
106
9.01 x
105
2.41 x
105
1.05 x
105
1.12 x
106
6.72 x
105
3.57 x
105
7.89 x
104

Tan
*c
b
0.57
1.15
2.05
2.61
0.84
1.19
1.97
5.29

2.21 x
105
2.21 x
105
3.03 x
104
1.13 x
104
1.74 x
105
8.78 x
104
4.00 x
104
8.03 x
103

Page 191
9 15
10 15
11 16.2
12 16.2
13 20
14 20
15 20
16 20

57.5
65.0
73
73
55
62.5
73
73

0
0
0.24
1.43
0
0
0.14
1.37

8.0
26.0
118.0
303
110.6
36.0
151.0
387.0

12.4
12.1
12.9
12.9
17.4
16.7
16.5
16.8

0.716
0.616
0.492
0.410
0.730
0.617
0.489
0.401

87.7
86.6
87.3
82.9
91.8
89.0
84.3
85.2

8.09 x 105
4.16 x 105
7.14 x 104
1.71 x 104
3.57 x 105
1.05 x 105
2.02 x 104
3.40 x 103

8.84 x 105
6.51 x 105
2.29 x 105
9.62 x 104
5.84 x 105
2.75 x 105
1.13 x 105
3.11 x 104

1.09
1.57
3.21
5.62
1.64
2.62
5.50
9.73

1.20 x 105
7.72 x 104
2.40 x 104
9.77 x 103
6.85 x 104
2.94 x 104
1.15 x 104
3.13 x 103

aVAcC, charged vinyl acetate level; Tp, polymerization temperature; TCE, trichloroethylene
level; at MI, melt index, ASTM D1238-57-T, at 175C; VAcB, bound vinyl acetate level; IV,
inherent viscosity, ASTM D1243-60-A, cyclohexanone at 30C.
G, elastic modulus, dyn/cm2; G, viscous modulus, dyn/cm2; at dynamic oscillation frequency
of 10 rad/sec at 175C. bTan = G/G = loss factor.
c*, complex viscosity (poise).
Source: Ref. 70.

Page 192
Resin Actual Predictive equation (22a)

Residual Predictive equation (22b)

Residual

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

-0.018
-1.16
7.3
-1
0.62
0.90
3.9
-13
0.46
-1.3
-22
-52
-0.9
1.2
4.0
63.0

-0.208
0.24
1.7
3.0
-0.14
3.5
8.2
-17.0
-0.34
0.8
-25.0
-51.0
0.8
2.9
-28.0
64.0

0.6
2.6
29.0
109.0
3.4
14.0
65.0
210.0
8.0
26.0
118.0
303.0
11.6
36.0
151.0
388.0

0.618
3.76
21.7
108
2.78
13.1
61.1
223
7.54
27.3
140
355
12.7
34.8
147
325

0.808
2.36
27.3
106
3.54
10.5
56.8
227
8.34
25.2
145
354
10.8
33.1
179
324

Page 193

FIGURE 10 (a) Melt index contour for vinyl chloride/vinyl acetate copolymers. Melt index
(g/10): A, 5; B, 10; C, 25; D, 50; E, 100; F, 200; G, 300. (b) Melt index contours for vinyl
acetate copolymer polymerized at 65C. Melt index (g/10): A, 5; B, 10; C, 25; D, 50; E, 100.
The effects of copolymer structure on melt rheological properties of elastic and viscous moduli
and on complex viscosity have not been related to either temperature or to compounding
ingredients. Further work in this area may indicate particular product areas where copolymers
can offer unusual performance advantages over homopolymers of vinyl chloride.
3. Propylene Copolymers: The copolymerization of propylene into the main chain of vinyl
chloride reduces the melt viscosity of the copolymer. When reported in terms of melt flow
properties, the vinyl chloride/propylene

Page 194
copolymer with 3 to 6% propylene has been shown to have up to several orders of magnitude
higher melt flow properties than the corresponding homopolymer of the same molecular weight
[71]. This is seen in Figure 11. The scatter for the melt flow properties of the vinyl
chloride/propylene copolymers is probably due to variations in the propylene content over the
range 3.5 to 5.0 wt %. The melt flow properties of these copolymers are greatly infuenced by
both the initial propylene level and the copolymer conversion. Preparing industrial quantities of
these copolymers requires precise control over the initial comonomer composition, the
conversion, and the polymerization temperatures.

FIGURE 11 Melt flow vs. inherent viscosity.

Page 195
B. Thermal Stability and Processability
The thermal stability of polyvinyl chloride has been examined rigorously by a number of authors
[74, 75] in review articles. Several different techniques are described for determining the
thermal degradation process, including both solution and solid-state techniques as well as static
and dynamic procedures. The reader should refer to these review articles as well as several
recent articles [7679] which discuss specific topics in the thermal stability of polyvinyl chloride.
While a Brabender shearing process is sufficient for a quality control test on an established
resin, a new process or new resin should be examined with several different thermal
degradation tests to characterize fully the frabrication and end-use envelope.
Heiberger and coworkers [50] reported on the thermal stability of polyvinyl chloride
homopolymers and copolymers using a test method that simulates a fully loaded fabrication
device. They reasoned that under commercial manufacturing conditions for rigid compounds, a
fabricator will load machinery to the power limit using whatever minimum temperature is needed
at that particular power consumption to achieve compound output. They used an empirical value
of the stability time at 1500 m-g (t1500) of torque and compared this to the intrinsic viscosity of
the resin. Their experimental results in a tinstabilized rigid formulation are summarized in Table
7. The t1500 values for the vinyl chloride homopolymers were compared to the vinyl acetate
copolymers, ethylene copolymers, and propylene copolymers. Using the homopolymers as a
reference point of 1.0 in relative stability, these results indicate that the propylene copolymers
are, on average, 2.2 times more stable than the homopolymers (tre1= 2.2), while the vinyl
acetate copolymers have about half the stability of the homopolymer resin. The presence of
ethylene at about 2.5% has minimal effect on the copolymer thermal stability. The effects of
different thermal stabilizer systems were tested for selected propylene copolymer resins and
vinyl acetate resins. The relative stability times (versus homopolymer as 1.0) have remained the
same. This is summarized in Table 8.
C. Glass Transition Temperature
On a microscopic level a polymeric material can be viewed as an array of long-chain molecules
which are held together by some combination of chemical and/ or physical forces. At a
sufficiently low temperature, the polymer behaves as a hard solid. As the temperature is
increased, the polymer will exhibit the expected volume expansion for solids. At a higher
temperature, characterized by the composition and structure of the polymeric material [71a],
the polymer will exhibit a change in the slope of the thermal expansion coefficient versus
temperature relationship. This second-order transition is called the glass transition and is
depicted in Figure 12. This second-order transition represents the change, at lower
temperatures, from a plasticlike material which is hard to a rubberylike material, at higher
temperatures, which is flexible. This macroscopic observation is related to the freedom of
motion, on the microscopic level, for segmental portions of the polymer chain. In a material
such as linear polyethylene there should be a single rather sharp transition for the -CH2- groups
of the backbone. As the polymer becomes more complex there should be multiple glass
transition phenomena which are characteristic of the polymer structure.
The glass transition temperature is related to the polymer melting temperature [Tg/Tm (K)] by a

factor of 2/3 for a wide range of polymers [72].

Page 196
Resin Wt % Mol % Intrinsic viscosity, [] (dl/g)

t1500(exptl.)(min)

t1500(calc.)a (min)

trel

Vinyl chloride/propylene
1
4.6 6.6
1.04
2
4.3 6.1
1.02
3
2.5 3.7
0.92
4
2.5 3.7
0.86
5
2.6 3.8
0.81
6
2.0 2.9
0.76
7
3.3 4.9
0.75
8
6.5 9.3
0.76
9
2.2 3.1
0.67
10
4.7 6.9
0.66
11
7.3 10.5 0.68
12
7.8 11.2 0.64
13b 2.0 2.9
0.62
14
5.1 7.2
0.58
15
4.8 7.0
0.52
16
8.3 11.8 0.50

15
19
25
32
50
45
69
155
87
142
190
300
112
220
295
575

18
19
22
32
43
51
71
120
92
152
216
287
121
248
348
645

2.6
2.9
2.0
1.7
2.0
1.3
1.8
4.3
1.4
2.1
3.2
3.9
1.3
2.0
1.9
3.0

Page 197
17
6.8
9.7
18
10.0
14.0
Vinyl chloride/ethylene
19
2.7
5.5
20
2.5
5.2
21
2.7
5.5
Vinyl chlorideinyl acetate
22
10
7.4
23
13
9.8
Polyvinyl chloride
24
25
26
27
28c
29
30
-

0.45
0.45

670
1090

692
1050

2.6
4.2

0.85
0.76
0.67

26
42
63

1.3
0.95
1.0

0.73
0.49

23
120

0.55
0.6

1.15
1.01
0.93
0.73
0.67
0.62
0.52

2.8
6.8
11
46
56
92
150

2.8
6.9
11.5
42
62
86
164

1.0
1.0
1.0
1.0
1.0
1.0
1.0

alog t1500 = 3.672.8 [] for homopolymers, log t1500 = 3.67 (1 + p + p[]2)-2.8 [] for
propylene copolymers, where p = mole fraction of propylene.
bExperimental bulk-polymerized copolymer.
cCommercial bulk-polymerized PVC resin.
Source: Ref. 50.

Page 198
trela
Resin

Sn

Ca-Zn

Pb

Ba-Cd

1
4
7
10
13
14
16
22
23

2.6
1.7
1.8
2.1
1.3
2.0
3.0
0.5
0.6

3.9
1.7
1.9
2.2
1.5
1.7
3.5
0.6
0.6

4.1
2.0
2.1
2.3
1.3
2.0
4.5
0.4
0.6

2.2
1.6
1.5
1.8
1.2
1.5
4.1
0.5
0.5

aSymbols and metal stabilizer types are Sn, Tin; Ca-Zn, calcium-zinc; Pb, lead; Ba-Cd, bariumcadmium.
Source: Ref. 50.
The glass transition temperature represents a processing and end-use reference temperature
at which a polymer becomes soft and can be masticated with compounding ingredients. As
noted in Chapter 3 (Section V.C), the glass transition temperature (78 to 80C) for polyvinyl
chloride is rather insensitive to the manufacturing conditions and the molecular weight of the
homo-

FIGURE 12 Volume-temperature characteristics of high polymers. (From Ref. 166.)

Page 199
polymer resin over the practical range of molecular weight (1.1 to 0.5 dl/g for inherent
viscosity). The use of copolymers to reduce melt temperature and glass transition temperature
became the primary drive in practical copolymer studies of polyvinyl chloride resins.
The comonomers will affect the glass transition temperature in an additive manner and then will
take the form
[23]
where
Tg1
= glass transition temperature (K) for component 1
W1
= weight of fraction of component 1
W2, Tg1
= correspond to component 2
Tg
= glass transition temperature (K) for the copolymer
This relationship has been examined by Wood [73] and his review should be consulted for
additional details.
The choice of comonomers to copolymerize with vinyl chloride so as to reduce the melt and
glass transition temperature becomes a balance between the manufacturing costs and the
other end-use properties. Ethylene segments in a polyvinyl chloride chain represent one of the
best ways to reduce the glass transition temperature. The manufacturing problems associated
with handling a high-pressure polymerization and the problems associated with the recovery
system have made this technical approach more expensive and the market size for vinyl
chloride/ethylene copolymers is rather small. Another approach is to use esters such as n-butyl
acrylate and diethyl fumarate as comonomers. These comonomers also lower the glass
transition temperature, but the combination of raw material costs and capital costs to ensure a
random distribution of the comonomer into the polymer main chain has also resulted in this
approach being impractical.
The use of vinyl acetate as a comonomer to reduce the glass transition temperature combines
rather straightforward copolymerization techniques with inexpensive comonomer costs. At
approximately 13 wt% of vinyl acetate comonomer the glass transition temperature of the
copolymer is reduced from about 75C to 68C.

V. OVERVIEW OF COMMERCIAL PROCESSES

The processes used to prepare commercial quantities of vinyl chloride copolymers cover a wide
range of free-radical polymerization procedures. The success of any commercial process is
based on manufacturing a resin with unique properties for a particular market niche within a
cost-effect envelope. During the past 10 years many of the rheological and end-use properties
which made the vinyl chloride copolymer resins feasible commercial materials have been
superseded by the cheaper homopolymer resin and changes in market demand. The major
copolymers and processes used in the manufacture of vinyl chloride copolymers are detailed in

Table 9 along with the market volume for the resin.

Page 200
Polymerization process
Copolymer type

Suspension

Emulsion

Solution

Bulk

Vinyl acetate
1-Olefins
Ethylene
Propylene
Acrylates
Acrylonitrile

190

20

75

<1a

10b
50c
-

<10d

aAvailable from Mitsui Toatsu in limited quantities together with higher 1-olefins (1-dodecene).
bAvailable from Air Products in ethylene content of 15 to 30 wt %.
cAvailable from B.F. Goodrich as part of the Geon latex product line; sales volume an estimate.
dMade by Monsanto; very small market.

VI. VINYL ACETATE COPOLYMERS

A. Suspension Polymerization
1. Process and Recipe: The suspension polymerization of vinyl chloride/ vinyl acetate
comonomer systems is the most important commercial copolymer process. The current U.S.
production of 190 million pounds (1982) far exceeds all other vinyl chloride copolymers. The
typical manufacturing system for these copolymers depends, in part, on the level of vinyl
acetate and the molecular weight of the copolymer resin that is to be produced. Generally, the
higher the level of bound vinyl acetate in the copolymer and the lower the molecular weight, the
more difficult will be the manufacturing process. These difficulties will be centered on the
stability of the suspension during the polymerization and the recovery of the unreacted
comonomers at the end of the polymerization.
Typical manufacturing recipes for the preparation of a range of vinyl acetate copolymers are
detailed in Table 10. The polymerization temperature and chain transfer agent level can be
varied over a wide range to achieve a particular molecular weight; increasing the polymerization
temperature permits a reduction in the chain transfer agent level. The choice of manufacturing
proceduresmolecular weight control through high levels of chain transfer agent or high
polymerization temperaturedepends on several factors: suspension system stability, recovery
system operation, initiator system and heat transfer capacity of the reaction, and the pressure
rating of the reactors. These are discussed in this section.
The process flow diagrams (depicted in Figure 13) for a suspension vinyl acetate/vinyl chloride
copolymerization are similar to that of the homopolymerization with the following exceptions:

Page 201
1. First-stage recovery system condenser will separate a vinyl acetate-vinyl chloride mixture
(80: 20 to 90: 10, depending on operating conditions).
2. Nash compressor seal water will be contaminated with vinyl acetate and will therefore need
a stream of free-radical scavenger (such as hydroquinone) to prevent pluggage by polymer
formation.
3. Vinyl acetate storage and recovered vinyl acetate storage will be needed.
4. Reflux condensor operation will be affected by the comonomer shift during the polymerization
(drop in internal pressure); the use of refrigerated water in either the reactor jacket or reflux
condensor may be needed.
5. Recovered vinyl chloride monomer stream will be contaminated with vinyl acetate and cannot
be used directly in a homopolymer process.
A typical polymerization cycle time for a type 1 primary record resin would be:
Time (min)
Charge and heat-up
Polymerization
Drop and vent
Wash and prorated cleaning
Antifouling application

3040
265
40
6080

395425
The polymerization and downstream equipment for this facility would include the following major
equipment:
Equipment

Capacity

1 polymerization reactor
2 blow-down vessels
1 stripping column
2 blend tanks
1 centrifuge
1 fluid-bed drier

50,000 gal
40,000 gal
20,000 lb/hr
30,000 gal
20,000 lb/hr
20,000 lb/hr

A plant with this configuration could produce up to 130 million pounds of resin per year

depending on the frequency of product changes.

Page 202

FIGURE 13 (a) Polyvinyl acetate/polyvinyl chloride by suspension process block flow diagram;

(b) polyvinyl acetate/polyvinyl chloride by suspension process polymerization and stripping; (c)
polyvinyl acetate/polyvinyl chloride by suspension process monomer recovery; (d) polyvinyl
acetate/polyvinyl chloride by suspension process dewatering and drying.

Page 203

FIGURE 13 (Continued)

Page 204
Secondary record resin

Water (gal)
Vinyl chloride (lb)
Vinyl acetate (lb)
Suspending agent (lb)
Initiator (lb)
Buffer (lb)
Chain transfer agent (lb)
Polymerization temperature (C)
Inherent viscosity (dl/g)
Bound vinyl acetate
Water/monomer ratio

Film and sheet resin Primary record resin

As is

PHM

As is

PHM

As is

PHM

22,000
123,800
8,200
100
50
200
0
60
0.75
5.0
1.39

139
93.8
6.2
0.076
0.038
0.15
0

22,000
117,00
15,000
120
50
200
0
65
0.65
9.1
1.39

139
88.6
11.4
0.09
0.038
0.15
0

22,000
109,000
23,000
160
70
324
330
75
0.48
14.0
1.39

139
82.6
17.4
0.12
0.053
0.20
0.50

Page 205
2. Resin Productivity: The rate at which copolymer can be produced in a manufacturing facility
is largely dependent on the heat transfer capacity of the reactor portion of the process train
and the copolymerization kinetics.
a. Initiation Systems: The effects of initiator structure on polymerization kinetics were reviewed
in Section VII.B.1 of Chapter 3. Many of the effects discussed for homopolymerization will
apply to these copolymer systems with the exceptions noted below.
To date there have been no kinetic models discussed in the open literature which can be used
as predictive tools to determine the rate of copolymerization for vinyl chloride. Several factors
that are unique to the copolymerization systems are difficult to quantify because of the shift in
the composition of the comonomer as a function of conversion:
1. Distribution of initiator between the polymer phase and monomer phase
2. Partition of the comonomer between the polymer and monomer phase
3. Change in initiator decomposition constant, Kd
The copolymer kinetics for the vinyl acetate/vinyl chloride system have been developed in an
evolutionary operational procedure (EVOP) wherein small changes in the initiator levels and
types have been examined under different operating conditions. The initiator mixtures used in
these copolymerizations still remain closely guarded industrial secrets.
It is reasonable to assume that a starting point for developing a uniform heat release
polymerization system would be a pair of initiators, one of which had a 10-hr half-life about
10C lower than the polymerization temperature and the second which had a 10-hr half-life
about 20C lower than the polymerization temperature. Using that criteria the initiator systems
which could be used for the resins in Table 10 can be estimated to be those described in Table
11.
b. Heat Transfer Systems: The resin productivity of a polymerization reactor is directly related
to its heat transfer capability. In Section VII.B.2 of Chapter 3 the effects of reactor
configuration, wall fouling, and temperature step polymerization were discussed for
homopolymer systems. These proces-

Polymerization temperature (C)

75

Initiator
1

Initiator
2

Type

T1/2Type

T1/2

Lauroyl peroxide

64 t-Butyl perpivalate

57

65
60

t-Butyl perpivalate
57 t-Amyl perneodecanoate 46
t-Butyl perneodecanoate 49 -Cumyl perneodecanoate 38

aEvolutionary operations would be required to determine both the level and the specific
ratios of these suggested initiator systems.

Page 206
ses will apply to the copolymer system also, but with certain modifications that should be
considered.
Reflux condenser operation: The change in comonomer composition as a function of conversion
leads to a change in the composition of the unreacted comonomer in the liquid phase, which in
turn changes the composition and therefore the pressure in the vapor phase of the reflux
condenser tubes. Typically, the pressure in the vapor phase will decrease from 170 psi to 60
psi (for the primary record resin). The effect of this decrease in pressure will be to reduce the
total heat transfer capacity of the reflux condenser. These pressure effects are discussed in
Perry's handbook [80], which should be consulted for details. The effects of differences in heat
capacity and film viscosity for the mixed components in the vapor phase have been quantified,
but in general tend to be less than the vapor-pressure effect. This decrease in heat transfer
capacity for the reflux condenser may be compensated for by switching from cooling-tower
water to refrigerated water in the shell of the reflux condenser.
Temperature-step operation: The change in comonomer composition and decrease in the
reactor vapor pressure does permit the polymerization temperature to be increased to
compensate for the decrease in the heat transfer capacity of the reflux condenser. The
increase in polymerization temperature will increase the temperature differential across the
shell of reactor and will increase the vapor pressure in the reflux condenser tube. This process
can only be performed under controlled conditions where the time-temperature profile for the
reaction is reproducible from batch to batch and where the resin properties of molecular
weight, molecular weight distribution, particle-size distribution, and particle morphology are not
degraded. Provided that these criteria can be met, the temperature-step operation offers a
method for higher overall reactor productivity.
Antifouling technology: A key aspect in maintaining the heat transfer capacity of the reactor
vessel is the prevention of polymer buildup. As with the homopolymer antifouling technology
(Chapter 3, Section VII B.2), the copolymer technology is described within the context of a
particular recipe (initiator/suspension system/buffer) and reactor configuration (L/D ratio/ baffle
system/agitation system). The recipe and reactor configuration will affect the efficiency of a
particular antifouling system. This will make the transfer of an antifouling system difficult to
quantify, but as with homopolymer systems, general trends can be identified. The patent
literature indicates at least four general approaches to reducing the buildup of polymer on the
walls of a reactor:
1. Increase the hydrophilic nature of the wall
2. Reduce water-phase polymerization
3. Reduce the electrolytic potential of the wall
4. Immobilize radical scavengers

The technical details associated with polymer buildup on a reactor were discussed in Section
VII.B.2.b of Chapter 3.
1. Hydrolytic nature of the walls: Several techniques using either very strong acid or base have
been reported in the patent literature. Dow Chemical [81, 82] has developed a procedure
whereby SO3 gas is used to treat the walls of a reactor. The process has been reported for

Page 207
both glass and stainless steel reactors. A possible mechanism that may be working for this
technology is the sulfonation of a thin skin of resin on the reactor surface and the formation of a
water-swollen polymer film which will not adsorb either monomer or initiator. Shinetsu [83] has
reported the use of sodium hydroxide solutions (pH > 12.0) as a precoat for reactor walls. Air
Products [84] has identified a highly alkaline system based on carbonates and bicarbonates as
effective. The addition of a surfactant has also been found beneficial [84a].
2. Reduce water-phase polymerization: A wide variety of water-soluble additives have been
reported to be effective in reducing the fouling in a reactor. Kanegafuchi has reported the use of
dithiocarboxylic acids [85] and dimethyl dithiocarbonates [86] as effective antifouling agents.
Mitsubishi Monsanto [87] indicates that a polysulfide used at 1 to 1000 ppm in the aqueous
phase is effective, while Shinetsu [88] uses propargyl alcohol at 300 ppm.
These patents, together with those reported in Table 11 of Chapter 3, need to be viewed within
the context of the initiator system reported and the partition coefficient for the initiator between
the monomer phase and the water phase. Modification of the initiator system may be
impossible with these antifouling technologies.
3. Electrolytic wall potential: Various inorganic heavy metal salts have been used as either
precoatings on the reactor walls or as aqueousphase additives to reduce the electrolytic
potential between grain boundaries on the wall. Mitsubishi Monsanto [89a, b, c] has reported
the use of vanadium salts in the +4 stage at levels of 1 to 1000 ppm, copper in the +2 stage at
0.2 ppm, and manganese in the +2 stage at 0.01 ppm. Chisso [90] also reports the use of
heavy metals, such as tin in the +2 state and silver and copper in the +1 state.
All these heavy metals have to be used with a great deal of caution: Their presence in the
finished resin can lead to accelerated thermal dehydrochlorination; in the effluent water these
heavy metals can require additional water purification processes, which will increase the
process costs.
An alternative electrolytic technique has been reported by Hoechst [91] for reducing reactor
fouling by imposing low amperage on the walls of a reactor, making the walls the anode and a
separate plate in the reactor as the cathode. At current levels of 200 to 500 mA (dc) the fouling
in the reactor is eliminated.
4. Immobilized radical scavengers: Wall coating systems that include free-radical scavengers
have been reported to be effective antifouling agents. Shinetsu has used nigrosine dye alone
[92] and with sodium silicate [93] and oxidized aromatic amine compounds [94]. Kanegafuchi
[95] has used a mixture of catechol and benzaldehydesalicylic acid condensation product.
3. Resin Quality: The copolymerization of vinyl chloride/vinyl acetate by a batch suspension
process results in a polymer with a wide range of composition and molecular weight.

a. Compositional Control of Vinyl Acetate: The shift in comonomer composition will also affect
the particle properties of the resin. Despite the commercial significance of this material, it has
only been within the last 10 years that any published work has appeared on compositional
control.

Page 208
Batch feed: Chen and Blanchard [96] reported the fractionation of copolymer and found that Mn
varied from 102,000 to 6200 for a series of resin fractions where the original resin had a Mn of
28, 000 and a bound vinyl acetate level (BVAc) of 15%. Janca and coworkers [97100] have
analyzed the molecular weight distribution and composition distribution for commercial
copolymers. They found the same molecular weight spread as that found by Chen and Blachard
[96]. Mori [24a] has developed a procedure for determining the molecular weight distribution
and the bound vinyl acetate content using a coupled infrared and refractive index detector. His
results, illustrated in Figure 14, demonstrate the spread in molecular weight and bound vinyl
acetate. As could be anticipated, the higher molecular weight has the lowest bound vinyl
acetate content, while the lowest molecular weight has the highest bound vinyl acetate content.
This shift in composition is a conse-

FIGURE 14 (a) Inherent viscosity contours for 85:15 vinyl chloride/vinyl acetate copolymers:
polymerization temperature and trichloroethylene process variables. Inherent viscosity (dl/g): A,
0.75; B, 0.70; C, 0.65; D, 0.60; E, 0. 55; F, 0. 50; G, 0. 45; H, 0.40. (b) Weight percent
polymer and bound vinyl acetate content as a function of molecular weight. -, Composition,
and -, molecular weight distribution, of PVC/PVAc). (From Ref. 24a.)

Page 209
quence of the shift in comonomer composition. The cumulative composition of these copolymer
resins varies with conversion. There has been no report in the literature concerning shift in the
reactivity ratio over the temperature range used for commercial copolymerizations, and it is
reasonable to assume that the temperature effect would be small over the range 55 to 75C.
Therefore, the average level of vinyl acetate in a copolymer should be predictable using a linear
equation of the following form:
% BVAc = K1 + K2[CVAc] + K3[CON]
[24]
where
CVAc = percent vinyl acetate initially added
CON = conversion
If the polymer conversion can be kept within narrow bounds, the K3[CON] term will be a
constant and can be merged into K1 and equation (24) can be reduced to
% BVAc = K1 + K2[CVAc]
[25]
Langsam and Cheng [70] have used the relationship defined in equation (25) to predict the
bound vinyl acetate content for copolymer resins of Table 5. Using values for K1 and K2 of
K1 = -0.39563
K2 = 0.85094
their comparison of the actual and predicted bound vinyl acetate levels for the 16 copolymer
resins are detailed in Table 12. The wide range of polymerization temperatures and
trichloroethylene level did not have an effect on the level of bound vinyl acetate over the
conversion range.
Incremental feed: The compositional variation observed for the batch copolymerization system
can be eliminated by an incremental feed system where the comonomer ratio is kept constant.
For the vinyl chloride/vinyl acetate system this can be done by charging all the vinyl acetate and
a portion of the vinyl chloride and, after initiating the polymerization, feeding the remainder of
the vinyl chloride so as to keep a constant vapor pressure in the reactor. This procedure was
described by Thomas and Hinds [22] and has been shown to form a copolymer that does not
vary with conversion [101]. Recent work by Kronman and coworkers [102] has examined the
physical properties of uniform composition copolymers, rates of solubilization in solvent
systems, thermoforming properties, and viscosity of concentrated solutions. In all cases
Kronman found that the uniform copolymer exhibited overall superior properties. These results
confirm other unpublished data on physical properties of these uniform copolymers [101].
These uniform compositions do offer several manufacturing difficulties:
1. Particle morphology is difficult to control because of the increase in the monomer
(polymer)/water ratio during the polymerization. High levels of wall fouling are very common for
these systems.

2. Monomer/initiator ratios shift during the comonomer feed and polymerization rates are
affected.

Page 210
Observation
Actual
Predicted
Residual
1
4.4
3.859
+0.541
2
3.8
3.859
-0.059
3
3.8
3.859
-0.059
4
3.8
3.859
-0.059
5
8.2
8.114
+0.086
6
8.0
8.114
-0.114
7
8.1
8.114
-0.014
8
8.1
8.114
-0.014
9
12.4
12.369
+0.031
10
12.1
12.369
-0.269
11
12.9
13.390
-0.490
12
12.9
13.390
-0.490
13
17.4
16.624
+0.776
14
16.7
16.624
+0.076
15
16.5
16.624
-0.124
16
16.8
16.624
+0.176
3. Productivity (pounds of resin per reactor-hour) tends to be much lower than for batch feed
systems. A continuous polymerization scheme may offer manufacturing procedures that can
circumvent these problems.
Peigjaner-Corbella and coworkers [102a] have developed a mathematical technique for
determining the optimum rate of addition of the more reactive comonomer in the preparation of
a uniform composition copolymer. Specific cases for vinyl chloride/vinyl acetate and vinyl
chloride/propylene are detailed. Changes in vapor pressure of the unreacted comonomer
mixture are used as the indication of the comonomer composition.
b. Molecular Weight: The molecular weight of these copolymers is a major factor in the
rheological end use applications for these copolymers and therefore needs to be controlled. As
with the homopolymer system, chain transfer between the active chain end and the monomer is
the major mechanism for control of molecular weight. For the copolymer system there is the
added effect of the vinyl acetate monomer and the increasing level of vinyl acetate as a function
of conversion. The presence of vinyl acetate in the copolymerization will reduce the molecular
weight of the resin due to a higher polymer end-to-monomer chain transfer coefficient.
The patents literature cites a different type of chain transfer agent for these copolymers.
Tenneco has disclosed the use of bromo compounds [103a] such as 1, 1, 2, 3, 4, 4hexabromo-2-butene, mercapto compounds [103b] such as

Page 211
ethylene bis(mercaptoacetate), and linear and branched aldehydes of C6-C15 carbon length
[103c]. Shinetsu has examined the mercapto compounds [104a, b, c] as chain transfer agents.
The use of a particular chain transfer agent needs to be balanced against the total process
economics, which includes the needs for equipment to handle the residual chain transfer agent
in the recovered monomer stream and the effect of the chain transfer agent on compound
properties.
The molecular weight of these copolymers depends on the following factors:
1. The initial vinyl acetate in the comonomer mixture, [CVAc]
2. The type and initial level of chain transfer agent used, [S]
3. The polymerization temperature, Tp
4. The level of conversion, x
Kronman and coworkers [105] reported that the viscosity of a copolymer resin depended on the
bound vinyl acetate level and the polymerization temperature. Combining this information with
the relationship developed for homopolymers [see Section VII.C.1.a, Chapter 3 equation (43)]
leads to the following linear equation for resin inherent viscosity:

The bound vinyl acetate term (BVAc) can be replaced by a charged vinyl acetate term [see
equation (25)], so that a reasonable working relationship for resin inherent viscosity would be

Langsam and Cheng [70] have used the relationship defined by equation (27) to develop
predictive equations for the inherent viscosity of copolymer resins. Using the 16 copolymer
resins defined in Table 5, they identified the values for the constants in equation (27) as
K1 = -3.868
K2 = -4.873 x 10-3
K3 = -6.632 x 10-2
K4 = +1.540 x 10+3
Using these constants in equation (27), they compared the actual and predicted values for
inherent viscosity. This is detailed in Table 13. A graphical representation of the effect of
polymerization temperature and trichloroethylene content on the inherent viscosity contours for
an 85:15 comonomer mixture is depicted in Figure 14a. The contours demonstrate the trade-off
between polymerization temperature and trichloroethylene at a fixed inherent viscosity. For
example, a 0.55 IV resin can be prepared at 70C with no chain transfer agent or at 60C with

approximately 2.2 wt % trichloroethylene. The

Page 212
Observation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

Observed value
0.725
0.587
0.501
0.414
0.707
0.609
0.495
0.419
0.716
0.616
0.492
0.410
0.730
0.617
0.489
0.401

Predicted value
0.700
0.600
0.503
0.414
0.709
0.608
0.507
0.422
0.720
0.617
0.489
0.410
0.731
0.627
0.477
0.395

Residual
0.025
-0.013
-0.002
0.000
-0.002
0.001
-0.012
-0.003
-0.004
-0.001
0.003
0.000
-0.001
-0.010
0.012
0.006

choice of reaction conditions will depend, in part, on the heat transfer capacity of the reactor
and on the effects of suspension agent stability at the polymerization temperature.
c. Particle Properties: The particle properties of the vinyl chloride/vinyl acetate copolymers
depend on the interactions between the reactor agitation system and the suspension system
used in the polymerization. These are described in Chapter 3 (Section VII.C.2). The vinyl
chloride/vinyl acetate copolymers are formed over a very wide range of manufacturing
conditions (see Table 9) and as a result, a single type of suspension system has not been found
which will be equally effective for all types of copolymer resins. Type 1 copolymer resin is
prepared at high temperature and high vinyl acetate monomer levels. The resin is used in
flooring and audio record applications, where high bulk density and low porosity are needed.
For this system a polyvinyl pyrrolidone suspension system [106] is adequate. By comparison, a
type 3 resin made at lower temperature and vinyl acetate content can be prepared with a
hydroxypropyl methyl cellulose suspension system. The choice and level of suspension system
used will be determined by the agitation system of the reactor and the desired particle
properties of the copolymer resin. There are no well-established procedures for studying the
effects of suspension agents on the particle morphology of copolymer resins. Ueda and
coworkers [107] developed an optical microscopy procedure whereby the particle structure of
a suspension system could be examined during the course of the polymerization. Their system
has not been reported in any

Page 213
subsequent work. The effect of suspending agent on the surface tension [108] of a water-air
interface and interfacial tension [109, 110] of a water-monomer mixture has been used as the
criterion for a suspension system's ability to stabilize resin particles. The advantage of the
interfacial method is that the effects of changes in the comonomer composition as a function of
time can be related to the interfacial tension. Also, the conversion of monomer to polymer can
be followed in a modified spinning drop interfacial tensiometer.
The particle morphology for the vinyl chloride/vinyl acetate copolymers can be controlled
through the addition of monomer-soluble agents. Shinetsu [111a, b] has disclosed additives
such as propylene glycol adipate ester [111a] or butyl acrylate/methyl methacrylate copolymer
[111b] which improve the porosity or particle-size control of the copolymers. A recent
Czechoslovakian patent [112] describes the use of oleyl alcohol as the monomer-soluble
component at 0.10 parts per 100 of monomer and Methocel HG65 at 0.15 part per 100 of
monomer. This Czech patent is probably for low-vinyl-acetate resins. The effect of these
additives will be influenced by the primary suspending agent, the level of vinyl acetate in the
initial monomer charge, and the polymerization temperature. These additives will tend to be less
effective at higher vinyl acetate levels and higher polymerization temperatures. There are no
established relationships between the additive structure and improved porosity. The trend, as
indicated by Cheng and coworkers [109, 110], is to relate the porosity to the partition
coefficient of the additive between the monomer-water phase and the interfacial tension of the
water-monomer interface. Additional work on intraparticle structure by scanning electron
microscopy and BET adsorption isotherms would also quantify the internal particle morphology.
d. Residual Monomer: Prior to 1973, vinyl chloride resins were sold commercially with rather
high levels of residual monomer. Levels of 1000 ppm of vinyl chloride in the homopolymer resin
were typically found. Efforts to reduce these levels of residual monomer were described in
Chapter 3 (see Section VII.C.3). The copolymers, depending on their type, tend to be far more
difficult for removal of residual monomer. The presence of residual vinyl acetate monomer
plasticizes the copolymer resin and increases the solubility of vinyl chloride in the resin. The
copolymer resins tend to have an agglomerated internal pore structure, with the agglomeration
being more severe for the type 1 (primary audio resin) than for the type 3 (secondary audio
resin) copolymer. Steam stripping column techniques have been described by Shinetsu [113a,
b] for reducing the vinyl chloride content in these resins. This steam stripping procedure works
on a countercurrent flow of slurry entering at the top of the column and descending over a
series of plates as the temperature and pressure are raised. Residence times of 1 to 90 min
are claimed. Analytical procedures for determining the residual vinyl chloride content have
progressed to the point where the presence of the vinyl acetate and chain transfer agents will
not interfere with the test at the parts per billion level [114].
B. Solution Polymerization
The solution polymerization of vinyl chloride/vinyl acetate is practiced commercially in the U.S.
only by Union Carbide. The process is rather old [115, 116] and the market has stabilized at
approximately 75 million pounds per year. There have been recent presentations in the
environmental /safety news which discuss the process [117119] and the solution products. For

the purposes of

Page 214
this discussion, emphasis will be placed on the VYHH product (14% bound vinyl acetate and
0.48 dl/g inherent viscosity).
1. Process and Recipe. The information available concerning the commercial solution
polymerization as practiced by Union Carbide is limited; patent citations indicate [115, 116] nbutane as the typical solvent. This would not result in a true polymer solution but rather in a
phase separation type of polymerization. Other sources of information indicate that the process
is run as a true solution and that the solution copolymers are made in a continuous manner
using a solvent in which the copolymer is completely soluble. There is no readily available
information source which can be used to define the present Union Carbide process. If the
process is carried out as a true solution polymerization, the process can be carried out in a
reactor train depicted in Figure 15. The process can be run in a single reactor or in a multiplereactor system. In a single-reactor mode, the solvent, comonomers, and initiator(s) are added
continuously in an overflow technique. Residence time is set to form a copolymer syrup with
approximately a 500-cP viscosity at polymerization temperature. Typical conversions of 40 to
60% of comonomer to copolymer are achieved in this mode. In a multiple-reactor mode, the
vinyl chloride monomer and solvent can be fed into the reactor system in an incremental
fashion. This type of operation will limit the copolymer composition distribution and produce a
resin that would resemble the technology defined by Thomas and Hinds [22].
The use of a solvent will reduce the operating pressure in the reactor train. The typical solvents
for this process can be aliphatic and alicyclic ketones as well as halogentaed organic materials.
Chain transfer to the sol-

FIGURE 15 Solution polymerization flow diagram.

Page 215
vent will be a part of this process. The recovery system operation is a major aspect of the
process economics. The process needs to be run at the highest loading of polymer to solvent
and at the highest conversion of comonomer to polymer that is possible within the viscosity
limits of the pumping capacity of the equipment.
Solvent and comonomer recovery will be a large portion of a solution process, and peroxide
formation from either atmospheric-oxygen or residual free radicals can be a problem. A typical
recipe for the solution process in a 3750-gal reactor system is as follows:
PHM

Actual

Solvent(s)
150
2000 gal
Vinyl chloride
76
7600 lb
Vinyl acetate
24
2400 lb
Initiator(s)
0.05
5 lb
The polymerization temperature will depend on the chain transfer coefficient for the solvent(s)
used. At about 40% conversion (4000 lb polymer, 6000 lb unreacted monomer, 2000 gal of
solvent), the reaction mixture is transferred to the solvent/comonomer recovery system.
2. Resin Quality: Assuming the solution process to be a true polymer solution, the resin
molecular weight and copolymerization kinetics would be defined by the solution process
described by Mickley and coworkers [119a]. The bound vinyl acetate will be dependent on the
charged vinyl acetate level and the conversion range.
The particle morphology for the solution copolymers is controlled by the precipitation process.
There is little information published about the approach used: vapor phase or liquid phase. The
polymer syrup could be atomized into a heated atmosphere above an aqueous phase. The
droplet size of the syrup would be controlled by the atomizer head and the temperature and
pressure in the vapor phase. Recent review articles on polymer phase separation [120, 121]
are instructive in developing process conditions for particle formation.
C. Emulsion Polymerization
The emulsion polymerization of vinyl chloride/vinyl acetate is used primarily for the preparation
of dispersion-grade copolymer resin. The polymerization process is well described in Section IX
of Chapter 3. In the preparation of a dispersion resin (also called a microsuspension resin) the
initiator is usually a monomer-soluble species [as contrasted to the aqueous soluble initiator
described in Table 22 (Section IX) of Chapter 3]. The final product has a particle size of 0.1 to
3 m and is conveniently spray dried into a free-flowing powder for end-use applications.

Page 216

VII. OLEFIN COPOLYMERS

A. Overview
Copolymerization of vinyl chloride with 1-olefins is an effective way to reduce the glass
transition temperature of the copolymer with low levels of comonomer. Glass transition
temperatures for typical 1-olefin polymers, noted in Table 14, indicate that ethylene is the most
effective comonomer in reducing the glass transition temperature. Recent Monsanto patents
[123a, b] describe the emulsion polymerization of vinyl chloride/ethylene systems to form
flexible coatings with 6 to 30 wt % ethylene. Low levels of ethylene (2 to 8%) in the copolymer
are effective for blood plasma storage bags [124]. Propylene copolymers have technical
advantages in improving the rheology and processability of PVC, but the price/performance
ratio was not sufficient in the face of competitive resins and compounds. Higher 1-olefin
copolymers have had little commercial success. Mitsui Toatsu has sampled small quantities of
what analyze, by 13C NMR procedures, to be a dodecene/vinyl chloride copolymer, suggested
for use in injection molding applications. The commercial success of these 1-olefin copolymers
depends on identifying a market where the price/performance ratio of the copolymer will
support the total manufacturing cost. The technical properties of the propylene and ethylene
copolymers illustrate this. Ethylene is unique among the olefins described in Table 14 in that it
does not have any allylic hydrogens and will not exhibit any degredative chain transfer effects
during the copolymerization.
B. Propylene Copolymers
1. Process and Recipe: The propylene copolymers were manufactured commercially by AIRCO
and Air Products from 1965 to 1978 using a suspension polymerization procedure. Flow
diagrams for the suspension homopolymer system (Chapter 3, Figure 12) will generally suffice
for an overview of this copolymer process. The recovered comonomer composed of about a
3:1 weight ratio of vinyl chloride/propylene requires careful analysis since the charged
propylene level will have a major impact on both the polymerization kinetics and the melt
rheology of the resin. The presence of propylene in the recovered comonomer mixture will
increase the vent emissions and the load on

Name
Ethylene
Propylene
1-Hexane
1-Octene
1-Decene
1-Dodecene
aAs a polymer; from Ref.
122.

Molecular
weight

Glass transition temperaturea

(C)

Boiling point
(K)

28.05
42.08
84.16
112.21
144.29
168.31

-120
-10
-50
-65
-36
-30

169
225
336
394
443
486

Page 217
the downstream vent incinerator unless an oversized Freon vent condensor is used.
A typical vinyl chloride/propylene copolymer recipe and process cycle time is noted in Table 15.
High levels of initiator(s) are needed for this process because of the degradative chain transfer
effects of the propylene allylic hydrogens [125, 125a]. Production equipment for this process
includes the following:
Equipment

Quantity

Reactor
3750 gal
6
Blowdown vessel
6000 gal
1
Stripping tower
6000 lb/hr
1
Blend tank
10,000 gal
4
Centrifuge
6000 lb/hr
1
Dryer
6000 lb/hr
1
This equipment configuration will produce, depending on the product mix and frequency of
product change, up to 40 million pounds of product per year.
2. Resin Productivity: Polymerization of propylene/vinyl chloride is complicated by the
comonomer shift during the reaction, which leads to an enrichment of propylene in the
unreacted comonomer. This increasing level of propylene as a function of conversion (Figure
16) results in a rapid decay in the polymerization kinetics. Mrazek and coworkers [126, 127]
have used this increase in the propylene content to follow the polymerization rate. Gas
chromatography of the gas phase was related to copolymer formation. This gas
chromatography technique can also be used for homopolymer with n-butane as a nonreactive
reference [128]. The decay in the polymerization kinetics can be controlled by an incremental
feed of initiator into the reaction mixture. This incremental feed is the key to high-polymerization
kinetics. Initiator is metered into the reactor to maintain a uniform temperature differential
between the slurry and the jacket. This technique has been used in production reactors and the
polymerization detailed in Table 15 can be completed in 8 hr.
Temperature-step techniques cannot be used because the reactor pressure increases as the
propylene content of the unreacted comonomer increases. The other heat removal procedure
of reflux condenser operation and antifouling technology which have been discussed for
homopolymers can be applied to propylene copolymers.
3. Resin Quality: The composition of a copolymer can be determined by the concentrations of
the comonomers at the polymerization site and the reactivity ratios of the comonomers.
a. Composition Control: The reactivity ratios for the vinyl chloride-propylene system have been
measured by a number of groups [129a-e]. Values of R1(VCM) = 2.90R2 (propylene) = 0.0
have been used to model compositional shifts in both the unreacted monomer and copolymer
[125, 48]. Various analy-

Page 218
a. Recipea
Components
Water
Vinyl chloride
Propylene
Suspending agent(s)
Initiator
Buffers
Polymerization temperature

1840 gal
10,340 lb
660 lb
11.0 lb
55 lb
5.5 lb
55C

140
94
6
0.10
0.50
0.05

aThis polymerization, when carried to 75% total monomer conversion, will result in a copolymer
resin with a melt flow of 10 per 10 min (ASTM D1238-condition F) and an inherent viscosity of
0.73 dl/g.
b. Cycle
Time (min)
Charge and heat-up
Polymerization
Drop/vent
Wash

3060
350450
30
30
________
440570

tical techniques have been used to determine the propylene content, including NMR
spectroscopy [130], chlorine content of the resin, and density [131, 132]. The propylene
content of these copolymers was examined at both low conversion [125] (20 to 30%) and high
conversion [48] (70 to 80%) using constant-composition copolymers as models.
The composition of these copolymers, when taken to a narrow conversion range, will be a
function of the charged propylene level. A mathematical relationship similar to that of the vinyl
acetate system [equation (25)] will describe the system:
% Propylene in copolymer = K1 + K2 [propylene charged]
[28]
The compositional charge in the polymer formed in 1% increments of conversion is depicted in
Table 16.
b. Molecular Weight Control: The molecular weight of the copolymers is controlled by the chain
transfer from active chain end to both vinyl chloride and propylene. The chain transfer
coefficient for propylene is much higher than for vinyl chloride [133], and since the propylene
content increases with

Page 219

FIGURE 16 Residual monomer composition as a function of conversion: , calculated for r1 =

2. 9, r2 = 0. , Experimental results at 50C; , experimental results at 30C. (From Ref. 125.)
conversion, the molecular weight of the copolymer will decrease with conversion. Ravey and
coworkers [125] have shown that the reciprocal of the copolymer inherent viscosity can be
related to the initial propylene content (Figure 17). Langsam [48] showed that the inherent
viscosity for these copolymers goes through a maximum point at about 50% conversion (Figure
18). This is due to the competitive reactions between chain transfer from the ever-increasing
level of propylene in the unreacted monomer (decreasing the molecular weight) and chain
growth [134].
An empirical relationship that can be used to determine the inherent viscosity of these resins at
a narrow conversion range is

where Pc is the propylene charged level.

c. Melt Flow Properties: The melt flow properties of the vinyl chloride/ propylene copolymers
are controlled by both the propylene content and the molecular weight of the resin [48]. The
melt flow changes with conversion, rising rapidly after 50% conversion (Figure 19). The
cumulative melt flow of these resins has been modeled with the instantaneous propylene
content as the independent variable and the instantaneous melt flow and inherent viscosity as
the dependent variables. The predicted melt flow and experimental melt flow (Figure 19) are in
good agreement. For polymerization at 50C the bound propylene in the copolymer affects the
inherent viscosity and melt flow as depicted in Table 17.

Page 220
a INSTANTANEOUS COPOLYMER COMPOSITION OF A BINARY SYSTEM CON W1 W2
CON W1 W2 % % % % % % 1 97.56 2.43 2 97.54 2.45 3 97.52 2.47 4 97.51 2.48 5 97.49
2.50 6 97.47 2.52 7 97.46 2.53 8 97.44 2.55 9 97.42 2.57 10 97.40 2.59 11 97.38 2.61 12
97.36 2.63 13 97.34 2.65 14 97.32 2.67 15 97.30 2.69 16 97.28 2.71 17 97.26 2.73 18 97.24
2.75 19 97.22 2.77 20 97.19 2.80 21 97.17 2.82 22 97.15 2.84 23 97.12 2.87 24 97.10 2.89
25 97.07 2.92 26 97.05 2.94 27 97.02 2.97 28 96.99 3.00 29 96.97 3.02 30 96.94 3.05 31
96.91 3.08 32 96.88 3.11 33 96.85 3.14 34 96.82 3.17 35 96.79 3.20 36 96.75 3.24 37 96.72
3.27 38 96.69 3.30 39 96.65 3.34 40 96.62 3.37 41 96.58 3.41 42 96.54 3.45 43 96.50 3.49
44 96.46 3.53 45 96.42 3.57 46 96.38 3.61 47 96.33 3.66 48 96.29 3.70 49 96.24 3.75 50
96.19 3.80 51 96.14 3.85 52 96.09 3.90 53 96.03 3.96 54 95.98 4.01 55 95.92 4.07 56 95.86
4.13 57 95.80 4.19 58 95.74 4.25 59 95.67 4.32 60 95.60 4.39

Page 221
INSTANTANEOUS COPOLYMER COMPOSITION OF A BINARY SYSTEM
CON
W1
W2
CON
W1
%
%
%
%
%
61
95.53
4.46
62
95.45
63
95.37
4.62
64
95.29
65
95.20
4.79
66
95.11
67
95.02
4.97
68
94.92
69
94.81
5.18
70
94.70
71
94.58
5.41
72
94.46
73
94.33
5.66
74
94.19
75
94.04
5.95
76
93.88
77
93.72
6.27
78
93.53
79
93.34
6.65
80
93.13
MEAN VALUE OF W1 = 96.24 W2 = 3.75
aBased on an initial comonomer ratio of VCM/P of 93:7 and R1/R2 of 2.90:0.0.
Source: Ref. 48.

W2
%
4.54
4.70
4.88
5.07
5.29
5.53
5.80
6.11
6.46
6.86

FIGURE 17 Reciprocal intrinsic viscosity of copolymer as a function of propylene concentration

in the feed. , polymerized at 60C, , polymerized at 50C, , polymerized at 30C.(From Ref.
125.)

Page 222

FIGURE 18 Inherent viscosity of batch-copolymerized VCM-P as a function of conversion. ,

Predicts inherent viscosity; , experimental points 7% propylene; , experimental points 8.33%
propylene. (From Ref. 48.)

FIGURE 19 Melt flow of batch-copolymerized VCM-P as a function of conversion. (From Ref.

48.)

Page 223
Bound propylene (%)

Inherent
Melt flow
viscosity (dl/g) (g/10 min)

2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0

0.775
0.737
0.700
0.660
0.620
0.580
0.545
0.507

2
4
8
16
32
64
126
282

aAdditional experimental data would be needed to construct melt

flow contour plots similar to Figure 10.
Source: Ref. 48.
C. Ethylene Copolymers
Copolymerization of ethylene with vinyl chloride offers two product advantages over the
homopolymer:
1. Reduction in the film-forming temperature in latex form
2. Increased flexibility for nonrigid applications
The disadvantages of very high comonomer vapor pressure and monomer mass transfer from
the vapor phase have been overcome by increased reactor wall thickness, internal cooling coils,
and modified agitation techniques.
The manufacture of vinyl chloride/ethylene resins at 15 to 30% bound ethylene in emulsion
systems represents a growing market for can coating and nonwoven binder applications. In
many cases a small quantity of a cross-linking termonomer is used to increase either the
adhesion or mechanical properties of these resins. These termonomers can be based on
carboxylic groups, methylol groups, or hydroxy groups. These termonomers provide a crosslinking site which during the application of the copolymer will either cross-link the material into a
three-dimensional network or will bond to the substrate. The chemistry of these functional
termonomers is beyond the scope of this chapter. Recent review articles [135] and publications
[136] have been devoted to this topic and should be consulted for further detail.
1. Process and Recipe: The manufacture of vinyl chloride/ethylene emulsion copolymers where
the ethylene composition can be as high as 30 wt % requires significantly higher pressures than
in homopolymer emulsion processes. In Chapter 3 the emulsion homopolymers were
manufactured at operating pressures of 150 to 175 psig. The ethylene copolymers are
manufactured

Page 224
at pressures up to 2000 psig. This higher comonomer vapor pressure means that the reactor
walls need to be much thicker. Heat transfer from the reactor through the walls is greatly
diminished and internal coils are used for heating and cooling the reaction mixture. The mass
transfer of ethylene into the reaction mixture is facilitated by subsurface sparging of ethylene
into the liquid phase. Approaches that use ring spargers in proximity to the agitator blades are
far more effective than adding the ethylene in the vapor phase. The subsurface spargers are
subject to liquid-phase contamination and blockage, and must be used with some form of trap
system to prevent contamination of the ethylene feed system.
Aside from the heavy-walled reactor vessel, the downstream process equipment is similar to
that described for the homopolymer system in Chapter 3 (Figure 38a and b). The ethylene/vinyl
chloride lattices are not usually dried but rather, are used in their emulsion form. The thin-film
evaporators and driers (Figure 38c) are not generally needed and therefore simplify the overall
process.
Ethylene monomer storage and delivery systems need to be refrigerated below 9.6C (critical
temperature for ethylene), or supercritical delivery systems need to be used for addition of
ethylene to the reactor. Alternatively, the ethylene can be added through a dome port into the
vapor phase and the charging of ethylene is controlled by the comonomer vapor pressure in the
reactor head space. This procedure depends on the ethylene mass transfer from the gas
phase to the liquid monomer/polymer phase being reproducible from batch to batch during the
entire polymerization. The ethylene composition in the copolymer will be variable from batch to
batch if the agitation parameters are not held constant and if the total surface area of the latex
is not held constant. The comonomer recovery system can be simplified to an ethylene vapor
flare and liquid vinyl chloride recovery. The flare ethylene stream will be contaminated with vinyl
chloride and therefore will require an incineration process compatible with hydrogen chloride
contamination.
A polymerization recipe for an emulsion process in a 5000-gal reactor would be:

Water
Vinyl chloride
Ethylene
Sodium lauryl sulfate
Ferrous chloride
t-Butyl hydroperoxide
Sodium hydrosulfite
Polymerization temperature
Inherent viscosity
Bound ethylene
Polymer total solids

Actual

PHM

3000 gal
7600 lb
1900 lb
475 lb
0.66 lb
14.2 lb
104.5 lb
55C
0.35 dl/g
12%
53%

150
80
20
5
0.007
0.15
1.1

Glass transition temperature

10%

Page 225
A typical polymerization cycle time would be:
Time (min)
Charge and heat-up
Polymerization
Drop and vent
Wash and prorated cleaning

45
300
50
50
________
445

Several modifications of this batch approach have been discussed in the patent literature. Union
Carbide [137, 137a] has reported that the ethylene and vinyl chloride monomers are premixed
and are added to the reaction mixture in a monomer-starved condition. The advantage to this is
that the comonomers do not exist as a separate liquid phase, and therefore the vapor pressure
is lower than equilibrium vapor pressure. The patents claim that composition and kinetic
controls of the reaction are superior to those of the saturated comonomer technique and that
the overall problem of mass transfer of ethylene into the reaction particles is decreased. The
overall productivity of the reactor (pounds of polymer per annual gallon) is not discussed. This
may be a problem with subsaturated pressure techniques.
Bayer [138] has described a continuous emulsion polymerization process for the manufacture of
high-ethylene copolymers. Using a continuous feed, detailed in Table 18, a latex was produced
with the following characteristics:
Yield per hour (latex)
4770 g
Latex total solids
52.5%
Ethylene content
21.6%
AMPS content
2.1%
Vinyl chloride content
76.3%
This reaction was run for 1000 hr in a 40-liter stainless steel reactor. The residual monomers,
ethylene at 250 g/hr and vinyl chloride at 280 g/hr, were vented from the product latex stream.
These continuous emulsion polymerization processes represent a highly productive approach to
the formation of the vinyl chloride copolymers.
The polymerization kinetics for these emulsion copolymers are similar to the process described
in Chapter 3 (Section IX). There are no established kinetic relationships for the vinyl chloride
copolymers and as a result, evolutionary operation procedures are needed to establish the
optimum initiator levels and feeds for each polymerization recipe. The heat removal
mechanisms will also be similar to the technology discussed earlier in this chapter and in
Chapter 3.

Page 226
a Water 2300 AMPSb 53 C12-C18 alkane sulfonate (sodium salt) 36 K2S2O8 20 4% solution
hydrolyzed PVAc (88% hydrolysis) 25 NaHCO3 20 Vinyl chloride 2200 Ethylene 800
aPolymerization temperature, 75C; polymerization reactor pressure, 90 atm. bSodium-2-Nacrylamido-2-methylpropane sulfonate: Source: Ref. 138.
2. Resin Quality: The latex application of the ethylene/vinyl chloride copolymers places a
premium on the film-forming characteristics of the latex, mechanical properties, and the ability
of the film to resist chemical and solvent attack. A wide range of elastomeric additives are also
needed to the comonomer mixture to ensure resistance to impact damage. There are no wellestablished citations in the open literature which quantify the end-use application properties to
either the polymer properties or to the process variables. An approach to analyzing the
relationship between the input process variables or polymer properties on the output product
specifications is needed if the product is to be made and solid in a commercially feasible
manner.
In Chapter 3 (Section VII) the concept of the statistical relationship between input variables and
output variables was discussed. The intent was to develop a process control equation relating
the interaction between identified input control variables and output product specification
variables. In the emulsion copolymerization, it is possible to identify a series of input process
variables, intermediate polymer variables, and output product quality variables; this is illustrated
in Figure 20. The subgroups tabulated for the

Page 227

FIGURE 20 Process control variables for emulsion copolymerization of vinyl chloride/ethylene.

Page 228
process, polymer, and fabrication variables are far from exhaustive but rather are listed to
indicate the extent to which a product can be affected by a manufacturing or fabrication
change.
An approach to understanding the factors controlling the product quality of these copolymers is
to understand fully interactions 1, 2, 3, and 4 in Figure 20. This would be the ideal situation of
open interactions among research, manufacturing, and applications functions. In many cases
these interactions are retarded because they cross intra- or intercompany barriers. Then only
paths 1 and 5 can be used to understand the relationship between polymer process and
product quality.
a. Minimum Film Formation Temperature: The film-forming characteristics of lattices have been
reviewed by Ottewill [139] and Vanderhoff and coworkers [140a-c]. The physical mechanism by
which an aqueous dispersion of latex droplets loses water and is converted into a close-packed
array of particles is determined largely by the physical characteristics of the latex: latex particle
size and distribution and type of emulsifier. Narrow-particle-size lattices tend to have gaps in
the film because of the void area in the close-packing array. A broader distribution of, let us
say, 0.1- to 2.0- m particles will enhance the packing into the void spaces with the smaller
particles. The compatibility of the emulsifier can be very important on the coalescence of the
latex particles into a film. Incompatible emulsifiers will retard the coalesence and will lead to an
exudation problem.
At a point-to-point contact situation, the coalescence of the particles depends on the glass
transition temperature (Tg) of the copolymer and the presence of any coalescence agents and
the ambient temperature. Assuming a Tg value of the polymer below ambient temperature, the
polymer latex particles will fuse and interdiffuse into continuous film. The rate at which this
interdiffusion-coalescence occurs will be a function of the molecular weight and the molecular
weight distribution of the copolymer as well as the level and distribution of the ethylene in the
copolymer.
A reactive third monomer in the chain will have little effect on the coalescence at low
concentrations. The purpose of the functional monomer is to cross-link the polymer mass
and/or bind to a substrate. This is discussed in subsequent sections.
b. Mechanical Properties: The ethylene/vinyl chloride copolymer in film form is usually of low
molecular weight, typically 30, 000 to 50, 000 for weight-average molecular weight, and does
not exhibit good tensile properties. The presence of a reactive termonomer in the polymer main
chain which will either self-cross-link or react with an additive or substrate significantly
increases the polymer molecular weight and therefore the mechanical properties of the film. In
a cellulosic appplication, N-methylolacrylamide will react with the hydroxy function of the
cellulose main chain. Carboxylic functional termonomers are effective in metal coating, with or
without an epoxide additive.
The cross-link density of the termonomer will not affect the Tg of the polymer mass above a

cross-link density of about 3000. The viscoelasticity properties of an infinite, three-dimensional

network will be difficult to predict. They will depend on the uniformity of the cross-link density
and the distance between cross-links.
c. Water-Solvent Resistance: The solvent resistance of the ethylene/vinyl chloride copolymers
will be poor, becoming extremely bad with a poorly coalescent film. The presence of
water/solvent-sensitive surfactants could lead to water blushing and avenues for solvent attack.
The functional third mono-

Page 229
mer in cross-linked form enhances the solvent resistance by creating an infinite network
structure. The technology for the effective use of a functional third monomer is described in the
patent literature but does not teach those skilled in the polymerization art the theoretical basis
for polymerization structure and mechanism.

VIII. ACRYLIC-ACRYLONITRILE COPOLYMERS

The use of the acrylic monomers and acrylonitrile open many alternative structural paths for
modifying the properties and end-use applications for vinyl chloride polymers. Broadly, these
comonomers offer the following diverse properties:
1. With long-chain acrylate esters:
a. Lowering of Tg for the vinyl chloride polymer
b. Improved adhesion properties and flexibility
c. Control of particle morphology for bulk and suspension resin
2. With acrylonitrile: Provide melt rheological and end-use properties for fiber properties and
film properties
A. Acrylate Esters
The commercial application of acrylate esters as comonomer with vinyl chloride depends on a
balance in the cost and performance characteristics of the comonomer. A goal of reduced Tg
needs to be balanced against the cost increment for the comonomer. The most reasonable
balance of cost and performance is found with the butyl acrylate and 2-ethylhexyl acrylate
monomers. The higher and lower members of the commercially available materials tend to be
inferior in cost/performance ratios. Because of the drastic difference in reactivity ratio values,
the method of copolymerization will result in either a uniform copolymer (if the acrylate
comonomer is fed incrementally into the reaction mixture) or a very heterogeneous copolymer
(if the acrylate comonomer is all present at the start of the copolymerization).
The vinyl chloride/acrylate copolymers find applications in low-heat-distortion applications [146]
and in applications where high impact is needed [147a, b]. Derl and Riczko [148] have
measured the storage modulus and loss modulus for the acrylate copolymers and have
developed relationships that relate these rheological properties to the copolymer composition.
Smets and coworkers [149a, b] have examined the solid-state morphology of acrylate-vinyl
chloride copolymers, demonstrating the phase separatuon of acrylate segments in certain types
of copolymers. An unusual application of these acrylate copolymers is the control of particle
morphology which has been used in both suspension polymerization by Daniels and coworkers
[150] and in bulk polymerization [151]. In both cases the copolymer is thought to reduce the

final particle-size distribution either by changing the internal viscosity of the monomer droplet or
by altering the adhesion of the primary grains of the forming resin particles.
Recently, pyrolysis studies [152a, b] have been used to analyze the composition of the acrylate
copolymers. The ease of lactonization of the acrylate esters has been used to prepare reactive
sites on these copolymers [153] and to cross-link during fabrication [154].

Page 230
B. Acrylonitrile
The copolymerization of vinyl chloride and acrylonitrile has been practiced for many years.
Smith [155] has summarized the commercial fibers available from the copolymerization of
acrylonitrile with other comonomers. The commercial applications of these various copolymers
in the fiber and fabric areas are based on their cost and performance characteristics, of which
a detailed description can be found in standard textile textbooks [156]. The vinyl
chloride/acrylonitric copolymer with vinyl chloride as the major constituent in the initial
comonomer will be the particular subject of this section. The market for this resin is very small,
being used in synthetic fibers for wigs and outerwear synthetic furs as well as in flameproof
children's sleepwear. The name dynel is no longer a registered trademark. As a fiber the
composition is typically 60:40 vinyl chloride/acrylonitrile [155]. The material can be prepared in
a number of ways: solutions, solution-precipitation, suspension, or emulsion. Literature citations
are not readily available as to the nature of the commercial process. A recent patent to Bayer
A. G. [157] describes a continuous emulsion process for the preparation of a 60% vinyl
chloride-containing polymer with a k value of 76.
One of the major applications for these copolymers has been in the area of synthetic fibers and
in the manufacture of wigs. The flame retardancy imparted by the chlorine content as well as
the close simulation to human hair are the major factors affecting this application. A summary of
some recent patents in this area is presented in Table 19.
Some of the more exciting new applications for these copolymers are in the area of membrane
separation technology. Because of the chemical resistance and ease of formation of
microporous structures, these copolymers are being studied in a wide range of applications.
Kedem and coworkers [158] have used these membranes as sites for the immobilization of
urease in an enzymic reactor. Desalination characteristics have also been reported [159]. A
number of applications in the area of electrolytic barriers in either fuel cells or electrical storage
batteries are summarized in Table 20.
Patent

Assignee (date)

Jap. Kokai 83 Kanegafuchi (June

28, 367
1983)
Jap Koaki 8282, Kanegafuchi (May
516
1982)
Ger. Offen. 2,
Bayer A.G. (Dec.
524, 125
1976)
Jap. Kokai 77 Mitsubishi Rayon
53, 065
(Apr. 1977)
Jap. Kokai 75 Ashai Chemical (Mar.
24, 573
1975)

Description
Hairlike synthetic fibers manufactured by wet spinning
Fire-resistant/transparent fibers by wet spinning
Improved coloristic properties and dimensional stability
by melt spinning
Fiber blends of copolymer and cellulosic for flame
resistance
Acrylic nonwoven textiles

Page 231
Patent

Assignee (date)

Description

Belg. BE 892418
Jap. Kokai 82165, 965
Ger. Offen. 2, 816, 088
U.S. 4, 096, 317

Dorr-Oliver (July 1982)

Matsushita (Oct. 1982)
Rohm & Haas (Nov. 1978)
ESD (June 1978)

Solid polymeric electrolyte

Cadmium anodes for alkaline batteries
Ion-repellent membrane
Separator

The structure of vinyl chloride/acrylonitrile copolymers has been examined by a variety of

analytical procedures. Koenig and Wendisch [160] have determined the diad distribution using
13C NMR spectroscopy. They found that the experimental diad values agreed with the
predicted values based on currently accepted reactivity ratio values. Pyrolytic techniques have
been used to determine sequence distribution for these copolymers [161, 162] and to identify
the combustion product distribution. The thermal stability of these copolymers as well as
controlled cyclization pathways were also examined [163, 164]. The thermal stability was found
to decrease as the acrylonitrile content increased from 23% to 75% [165]. These copolymers
can be modified by the Arbuzov reaction to form diethyl phosphine substituents [165]. This
modification should improve the flame resistance of the polymer and provide reactive sites for
subsequent substitutions.

IX. FUTURE DEVELOPMENTS

The vinyl chloride copolymer market currently is a small fraction of the overall homopolymer
market and small when compared to other commodity polymeric materials. Any substantial
growth in the use of these copolymers will require a significant growth in the understanding of
polymer process-structure and end-use applications. Figure 20 indicates the extent to which
these interactions need to be quantified. Furture developments in these copolymers need to
accentuate the applications where cost, flame retardancy, and melt viscoelasticity are superior
to other products.

SYMBOLS
M
M
K
R
fi
Fi
Pi
Qi

active chain end

monomer unit
rate constant
reactivity ratio
mole fraction of monomer i in the total monomer mix
mole fraction of polymer i in the total polymer mix
characteristic of chain end i
reactivity of monomer 1

Page 232
ei

k
t

Tg
Wi

polarity of monomer I
stress
strain
rate of strain
viscosity constant
elastic constant
time
frequency of displacement
glass transition temperature
weight fraction of polymer i

REFERENCES
1. Regnault, V., Ann., 14, 22 (1835).
2. Nietsche, R., and Wolf, K. A., Kuntststoff. Handbuch, Vol. II, Springer-Verlag, Berlin,
1962, p. 891.
3. U.S. Pat. 1,929,453 to B.F. Goodrich (Oct. 1933).
4. Billmeyer, F. W., Jr., Textbook of Polymer Science, Wiley, New York, 1962.
5. Elias, H.-G., Macromolecules, Part 2, Plenum Press, New York, 1977.
6. Tenth Collective Index, Chemical Abstract Service, Vol. 10CS13, pp. 21466CS-21486CS.
7. Gabbett, J. F., and Smith, W. M., in Copolymerization, G. E. Han, ed., Interscience, New
York, 1968, p. 595.
8. Chem. Mark. Rep., 223(36), 54 (June 27, 1983).
9. U.S. Pat. 105,338 to Celluloid Manufacturing (July 1870).
10. Flory, P. J., Principles of Polymer Chemistry, Cornell University Press, Ithaca, N.Y., 1953,
p. 541590.
11. Obalisi, O., Robeson, L. M., and Shaw, M. T., in PolymerPolymer Miscibility, Academic
Press, New York, 1979.
12. Paul, D. R., and Newman, J., in Polymer Blends, Academic Press, New York, 1978.
13. Bohn, L., Rubber Chem. Technol., 41, 495 (1968).
14. Friday, A., Cooper, D. R., and Booth, C., Polymer, 18, 164 (1977).

15. Wall, F. T., J. Am. Chem. Soc., 63, 1862 (1941).

16. Mayo, F. R., and Lewis, F. M., J. Am. Chem. Soc., 66, 1594 (1944).
17. Alfrey, T., and Goldfinger, G., J. Chem. Phys., 12, 205 (1944).
18. Theil, M. H., J. Polym. Sci., Chem., 21, 633 (1983).
19. Mertz, E., Alfrey, T., and Goldfinger, G., J. Polym. Sci., 1, 75 (1946).
20. Ham, G. E., J. Polym. Sci., 45, 160, 177, 182 (1960).
21. Mayo, F. R., Walling, C., Lewis, F. M., and Hulse, W. F., J. Am. Chem. Soc., 70, 1532
(1948).
22. Thomas, C. M., and Hinds, J. R., Br. Plast., 522 (Dec. 1958).
23. Agron, P., Alfrey, T., Bohrer, J., Haas, H., and Wechsler, H., J. Polym. Sci., 3, 157
(1948).
24. Breitenbach, J. W., and Edelhauser, H., Ric. Sci., 25, 242 (1955).
24a. Mori, S., J. Chromatogr., 157, 75 (1978).
24b. Tobolsky, A. V., J. Am. Chem. Soc., 80, 5927 (1958).
24c. Tobolsky, A. V., Rogers, C. E., and Brinkman, R. D., J. Am. Chem. Soc., 82, 1277
(1960).

Page 233
25. Fineman, M., and Ross, S. D., J. Polym. Sci., 5, 269 (1950).
26. Yezrielev, A. J., Brokina, E. L., and Roskin, Y. S., Vysokomol. Soedin., Ser., A-11, 1670
(1969).
27. Tidwell, P. W., and Mortimer, G. A., J. Polym. Sci., A3, 369 (1965).
28. Joshi, R. M., J. Macromol. Sci. Chem., A7, 1231 (1973).
29. VanderMeer, R., Linssen, H. N., and German, A. L., J. Polym. Sci., Chem., 16, 2915
(1978).
30. McFarlane, R. C., Reilly, P. M., and O'Driscoll, K. F., J. Polym. Sci., Chem., 18, 251
(1980).
31. Johnson, C. D., The Hammett Equation, Cambridge University Press, Cambridge, 1973.
32. Lewis, F. M., Mayo, F. R., and Hulse, W. F., J. Am. Chem. Soc., 67, 1701 (1945).
33. Bartlett, P. D., and Nozaki, K., J. Am. Chem. Soc., 68, 1495 (1946).
34. Alfrey, T., and Price, C. C., J. Polym. Sci., 2, 101 (1947).
35. Wall, L. A., J. Polym. Sci., 2, 542 (1947).
36. Levinson, G. S., J. Polym. Sci., 60, 43 (1962).
37. Greenley, R. Z., J. Macromol. Sci. Chem., A9, 505 (1975).
38. Greenley, R. Z., J. Macromol. Sci. Chem., A11, 933 (1977).
39. Menges, G., and Berndtsen, N., Pure Appl. Chem., 99, 597 (1977).
40. Berry, G. C., and Fox, T. G., Adv. Polym. Sci., 5, 261 (1968).
41. Kumar, N. G., J. Polym. Sci., Macromol Rev., 15, 225 (1980).
42. Nielsen, L. E., Polymer Rheology, Marcel Dekker, New York, 1977.
43. Han, C. D., Rheology in Polymer Processing, Academic Press, New York, 1976.
44. Pezzin, G., Plast. Polym., 37, 295 (1969).
45. Lyngaae-Jorgensen, J., J. Chromatogr. Sci., 9, 331 (1971).
46. Talamini, G., and Vidotto, G., Makromol. Chem., 100, 48 (1967).

47. Graessley, W. W., and Prentice, J. S., J. Polym. Sci., Part A2 6, 1881 (1968).
48. Langsam, M., J. Appl. Polym. Sci., 21, 1057 (1977).
49. Combs, R. L., Slonaker, D. F., and Coover, H. W., J. Appl. Polym. Sci., 13, 519 (1969).
50. Heiberger, C. A., Phillips, R., and Cantow, M. J. R., Polym. Eng. Sci., 9, 445 (1969).
51. Berens, A. R., and Folt, V. L., Trans. Soc. Rheol., 11(1), 95 (1968).
52. Berens, A. R. and Folt, V. L., Polym. Eng. Sci., 8, 5 (1968).
53. Berens, A. R., and Folt, V. L., Polym. Eng. Sci., 9, 27 (1969).
54. Collins, E. A., and Metzger, A. P., Polym. Eng. Sci., 10, 57 (1970).
55. Collins, E. A., and Daniels, C. A., Polym. Eng. Sci., 14, 357 (1974).
56. Sieglaff, C. L., SPE Trans., 4, 129 (1964).
57. Mayerich, R. C., and Sieglaff, C. L., J. Appl. Polym. Sci., 9, 3917 (1965).
58. Sieglaff. C. L., Pure Appl. Chem., 53, 509 (1981).
59. Lyngaae-Jorgensen, J., Pure Appl. Chem., 53, 533 (1981).
60. Lyngaae-Jorgensen, J., J. Appl. Polym. Sci., 20, 2497 (1976).
61. Barus, C., Am. J. Sci., 45, 87 (1893).
61a. Racin, R., and Bogue, D. C., J. Rheol., 23, 263 (1979).
61b. Benbow, J. J., and Lamb. P., SPE Trans., 3, 1 (1963).
61c. Sieglaff. C. L., Polym. Eng. Sci., 9, 1 (1969).
61d. Paradis, R. A., Chem. Eng. Prog., 62, 12 (1966).

Page 234
62. Stinson, S. T., Chem. Eng. News, 26 (June 11, 1979).
62a. VanWeizer, J. R., Viscosity and Flow Measurements, Interscience, New York, 1962.
63. Park. I. K., and Riley, D. W., J. Vinyl Technol., 2, 160 (1980).
64. Collins, E. A., Metzger, A. P., and Furgason, R. R., Polym. Eng. Sci., 16, 240 (1976).
65. Chauffoureauex, J. C., Dehennau, C., and Van Rijckevorsel, J., J. Rheol., 23, 1 (1979).
66. Rangnes, P., Kunstoffe, 51, 428 (1961).
67. Mondvai, I., Halasz, L., and Molnar, I., Period. Polytech., 3, 203 (1971).
68a. Powell, E., and Khanna, S. K., J. Appl. Polym. Sci., 16, 2013 (1972).
68b. Khanna, S. K., and Powell, E., SPE J., 27, 33 (1971).
69. Khanna, S. K., J. Audio Eng. Soc., 24, 464 (1976).
70. Langsam, M., and Cheng, J. T., J. Appl. Polym, Sci. 30, 1285 (1985).
71. Technical data, Flowel Resin, Plastics Division, Air Products and Chemicals, Trexlertown,
Pa.
71a. Boyer, R. F., Rubber Rev., 36, 1303 (1963).
72. Lee, W. A., and Knight, G. J., Br. Polym. J., 2, 73 (1970).
73. Wood, L. A., J. Polym. Sci., 28, 319 (1958).
74. Geddes, W. C., Rubber Chem. Technol., 40, (1967).
75. Onozuka, M., and Asahina, M., J. Macromol. Sci. Rev. Chem., C3, 235 (1969).
76. Gupta, V. P., and St. Pierre, L. E., J. Polym. Sci., Chem., 17, 797 (1979).
77. Sheppard, M. J., and Gilbert, J., Eur. Polym. J., 17, 285 (1981).
78. Nagy, T. T., Kelen, T., and Tuchsanyi, B., Polym. Bull., 2, 749 (1980).
79. Millan, J., Martinez, G., Mijangos, C., J. Polym. Sci., Chem., 18, 505 (1980).
80. Perry's Chemical Engineering Handbook, 4th ed., R. H. Perry, C. Chilton, and S.
Kirkpatrick, eds., McGraw-Hill, New York, 1963, Sec. 10.
81. U.S. Pat. 4, 105, 841 to Dow Chemical (Aug. 1976).

82. Ger. Offen. 2, 719, 919 to Dow Chemical (Nov. 1977).

83. Rom. Pat. 62,747 to Shinetsu (Nov. 1977).
84. U.S. Pat. 4,451,625 to Air Products (May 1984).
84a. Eur. Pat. 83107572.6 to Air Products (July 1982).
85. Bel. Pat. 837.070 to Kanegafuchi (June, 1976).
86. Jap. Kokai 7785, 281 to Kanegafuchi (July 1977).
87. Jap. Kokai 78114, 892 to Mitsubishi Monsanto (Oct. 1978).
88. Jap. Kokai 7969, 195 to Shinetsu (June 1979).
89a. Jap. Kokai 7760, 884 to Mitsubishi Monsanto (May 1977).
89b. Jap. Kokai 7772, 789 to Mitsubishi Monsanto (May 1977).
89c. Jap. Kokai 7776, 390 to Mitsubishi Monsanto (June 1977).
90. Ger. Offen. 2, 637, 582 to Chisso (Mar. 1977).
91. Ger. Offen. 2, 636, 653 to Hoechst (Mar. 1977).
92. Ger. Offen. 2, 632, 468 to Shinetsu (Feb. 1977).
93. Ger. Offen. 2, 632, 469 to Shinetsu (Feb. 1977).
94. Jap. Kokai 7813, 689 to Shinetsu (Feb. 1978).
95. Jap. Kokai 7936, 389 to Kanegafuchi (Apr. 1979).
96. Chen, H. R., and Blanchard, L. P., J. Appl. Polym. Sci., 16, 603 (1972).

Page 235
97. Janca, J., and Kolinsky, M., J. Appl. Polym. Sci., 21, 83 (1977).
98. Janca, J., Mrkvickova, L., Kolinsky, M., and Brar, A. S., J. Appl. Polym. Sci., 22, 2661
(1978).
99. Janca, J., Pokorny, S., and Kolinsky, M., J. Appl. Polym. Sci., 23, 1811 (1979).
100. Janca, J., Kroupova, J., Pokorny, S., Stehlicek, J., Kalal, J., Gankina, E. S., Kever, E.
E., and Belenkii, B. G., J. Chromatogr., 171, 81 (1979).
101. Langsam, M., unpublished data.
102. Kronman, A. G., Fedoseev, B. I., Chekushina, M. A., Arknipova, L. I., and Lepaev, A. F.,
Plast. Massy, 7, 48 (1976).
102a. Peigjaner-Corbella, L., Baxaris, F. R., and Jene, M. O., Ing. Quim. (Madrid), 13, 175
(1981).
103a. U.S. Pat. 4, 052, 546 to Tenneco (Oct. 1977).
103b. Ger. Offen. 2, 748, 469 to Tenneco (May 1978).
103c. Ger. Offen. 2, 840, 334 to Tenneco (Mar. 1979).
104a. Ger. Offen. 2, 807, 180 to Shinetsu (Sept. 1976).
104b. Jap. Kokai 7910, 392 to Shinetsu (Mar. 1979).
104c. Jap. Kokai 7910, 393 to Shinetsu (Mar. 1979).
105. Kronman, A. G., and Fedoseev, B. I., Plast. Massy, 11, 58 (1980).
106. U.S. Pat. 3,459,719 to Tenneco (Aug. 1969).
107. Ueda, T., Takeuchi, K., and Kato, M., J. Polym. Sci., Chem., 10, 2841 (1972).
108. Clark, M., in Particulate Nature of PVC, G. Butter, ed., Applied Science Publishers,
London, 1982, pp. 14, 15.
109. Cheng, J. T., J. Macromol. Sci. Phys., B20, 365 (1981).
110. Cheng, J. T., and Langsam, M., J. Macromol. Sci. Chem., A21, 395 (1984).
111a. Jap. Kokai 80160, 006 to Shinetsu (May 1980).
111b. Ger. Offen. 2, 942, 343 to Shinetsu (Apr. 1980).

112. Czech. Pat. 162, 899 to J. Porubsky et al. (Mar. 1976).

113a. Ger. Offen. 2, 621, 987 to Shinetsu (May 1975).
113b. Ger. Offen. 2,855,146 to Shinetsu (June 1980).
114. Gibb, T. B., and Wolf, P. H., J. Chromatogr. Sci., 20, 471 (1982).
115. U.S. Pat. 2,064,565 to Union Carbide (Dec. 1936).
116. U.S. Pat. 2,075,429 to Union Carbide (Mar. 1937).
117. Wheeler, Jr., R. N., Environ. Health Perspect., 41, 123 (1981).
118. Wheeler, Jr., R. N., Conf. to Reevaluate the Toxicity of Vinyl Chloride, Polyvinyl Chloride
and Analog. National Institutes of Health, Bethesda, Md., Mar. 20, 1980, p. 245254.
119. Vinyl Chloride: A Review of National Emission Standards, EPA-450/ 382003, Feb.
1982, p. 329/332.
119a. Mickley, H. S., Michaels, A. S., and Moore, A. L., J. Polym. Sci., 60, 121 (1962).
120. Cohen, C., Tanny, G. B., and Prager, S., J. Polym. Sci., Phys., 17, 477 (1979).
121. Strathmann H., and Kock, K., Desalination, 21, 241 (1977).
122. Nielsen, L. E., Mechanical Properties of Polymers, Reinhold, New York, 1962, pp. 16,
17.
123a. U.S. Pat. 3, 827, 997 to Monsanto (Aug. 1974).
123b. U.S. Pat. 3, 925, 287 to Monsanto (Dec. 1975).
124. U.S. Pat. 3, 892, 692 to Air Products (Sept. 1984).
125. Ravey, M., Waterman, J. A., Shorr, L. M., and Kramer, M., J. Polym. Sci., Chem., 14,
1609 (1976).

Page 236
125a. Ulbricht, J., and Wolf, K. F., Plaste Kautsch., 22, 471 (1975).
126. Mrazek, Z., Jungwirt, A., and Kolinsky, M., J. Appl. Polym. Sci., 27, 1513 (1982).
127. Mrazek, Z., Jungwirt, A., and Kolinsky, M., J. Appl. Polym. Sci., 27, 2079 (1982).
128. Langsam, M., J. Polym. Sci. Polym. Lett., 22, 549 (1984).
129a. Cantow, M. J. R., Cline, C. W., Heiberger, C. A., Huibers, D. T., and Phillips, R., Mod.
Plast., 46, 128 (1969).
129b. Takemoto, K., Kunstoffe, 57, 697 (1967).
129c. Osugi, J., and Hamanone, K., Polym. Rep., 135, 31 (1969).
129d. Cain, A. R., 159th ACS Meet., Houston, Feb. 23, 1970.
129e. Heiberger, C. A., and Phillips, R., Rubber Plast. Age, 48, 636 (1967).
130. Randall, J. C., Polymer Sequence Determination, Academic Press, New York, 1977.
131. Chan, R. K. S., and Worman, C. H., Polym. Eng. Sci., 14, 389 (1974).
132. Ravey, M., J. Polym. Sci., Chem., 13, 2635 (1975).
133. Langsam, M., J. Appl. Polym. Sci., 23, 867 (1979).
134. Cotman, J. D., Gonzales, M. F., and Clever, C. G., J. Polym. Sci., Part A-1, 5, 1137
(1967).
135. Bufkin, B. G., and Grave, J. R., J. Coat. Technol. : Part I, 50(641), 41 (1978); Part II,
50(643), 67 (1978); Part III, 50(644), 83 (1978); Part IV, 50(645), 70 (1978); Part V, 50(647),
65 (1978); Part VI, 51(649), 34 (1979).
136. Functional Monomers, Vols. 1 and 2, R. H. Yocum and E. B. Nyquist, eds., Marcel
Dekker, New York, 1974.
137. Eur. Pat. Appl. 0026490 to Union Carbide (Apr. 1981).
137a. U.S. Pat. 4, 331, 577 to Union Carbide (May 1982).
138. U.S. Pat. 3, 929, 706 to Bayer (Dec. 1975).
139. Ottewill, R. H., in Science and Technology of Polymer Colloids, G. W. Poehlein, R. G.
Ottewill, and J. W. Goodwin, eds., NATO ASI Series, Martinus Nijhoff Publishers, The Hague,
1983, pp. 503522.

140a. Vanderhoff, J. W., Tarkowsk, H. L., Jenkins, M. C., and Bradford, E. B., J. Macromol.
Chem., 1, 361 (1966).
140b. Bradford, E. B., and Vanderhoof, J. W., J. Macromol. Sci. Phys. B6, 671 (1972).
140c. Vanderhoff, J. W., Bradford, E. B., and Carrington, W. K., J. Polym. Sci., Polym. Symp.,
41, 155 (1973).
141. G. B., Pat. 1, 277, 289 to Farbewerke Hoeschst (June 1972).
142. U.S. Pat. 3, 668, 194 to Borden (June 1972).
143. Ger. Offen. 2, 735, 022 to Hooker (Feb. 1978).
144. Florjanczyk, Z., Kuran, W., Langswald, N., and Sitkowska, J., Makromol. Chem., 184
2457 (1983).
145. Logothetis, A. L. and McKenna, J. M., J. Polym. Sci., Polym. Lett., 12, 131 (1974).
146. Jap. Kokai 82200435 to Nippon Zeon (Dec. 1982).
147a. Ger. Offen. 2, 251, 515 to Czech. Academy Ved (Apr. 1973).
147b. Jap. Kokai 724937 to Japanese Carbide (Feb. 1972).
148. Deri, M., and Riczko, A., Muanyag Gumi, 19(6), 173 (1982).
149a. Overberg, N., Weinand, G., and Smets, G., J. Polym., Sci., Chem, 16, 3107 (1978).
149b. Weinand, G., and Smets, G., J. Polym. Sci., Chem., 16, 3091 (1978).

Page 237
150. Ger. Offen. 2, 413, 151 to Air Products (Dec. 1974).
151. Ger. Offen. 2, 735, 022 to Hooker (Feb. 1978).
152a. Komkova, G. S., and Selivokhin, P. I., Kozh.-Obuvn., 22(2), 35 (1980).
152b. Komkova, G. S., and Selivokhin, P. I., Kozh.-Obuvn., 23(1), 27 (1981).
153. Johnson, N. W., Polym. Prep., 13(2), 1065 (1972).
154. Ger. Offen. 2, 108, 372 to VEB Leuna-Werke Walter Ulbrich (Sept. 1971).
155. Smith, W. M., in Vinyl Resins, Reinhold, New York, 1958, p. 52.
156. Hill, R., ed., Fibres from Synthetic Polymers, Elsevier, Amsterdam, 1953.
157. Ger. Offen. 2, 644, 592 to Bayer A.G. (Feb. 1978).
158. Tanny, G. B., Kedem, O., and Bohak, Z., J. Membr. Sci., 4, 363 (1979).
159. Freilich, D., and Tanny, G. B., Desalination, 27, 233 (1978).
160. Koenig, J., and Wendisch, D., Angew. Makromol. Chem., 92, 191 (1980).
161. Tanaka, M., Nishimura, F., and Shono, T., Anal. Chim. Acta, 74, 119 (1975).
162. Iglauer, N., and Bentley, F. F., J. Chromatogr. Sci., 12, 23 (1974).
163. Joesten, B. L., and Johnston, N. W., J. Macromol. Sci. Chem., 8, 83 (1974).
164. Henrici-Olive, G., and Olive, S., Polym. Bull. (Berlin) 5, 457 (1981).
165. Borisenko, Z. V., Sorokin, D. S., Bondareva, N. S., Romovskii, E. S., and Roskin, E. N.,
Vysokomol. Soedin., Ser. B, 17, 707 (1975).
166. Bekkedahl, N., J. Res. Natl. Bur. Stand., 13, 411 (1934).

Page 238

Page 239

5 Safety and Environmental Concerns in Resin Manufacture

JOHN T. BARR
Air Products and Chemicals, Inc. Allentown, Pennsylvania
I. INTRODUCTION
II. TOXICITY CONSIDERATIONS
A. Vinyl Chloride
B. Vinyl Acetate
C. Trichloroethylene
D. 1, 2-Dichloroethane
E. Polyvinyl Chloride
III. REGULATORY STATUS
A. The Food and Drug Administration
B. The Occupational Safety and Health Administration
C. The Environmental Protection Agency
IV. GENERAL SAFETY PROCEDURES
A. Raw Material Handling and Storage
B. Monomer Production
C. Polymerization
D. Stripping
E. Downstream Operations
V. WASTE STREAMS
A. Water
B. Liquids and Gases
C. Solids
VI. ANALYTICAL METHODS
VII. VINYL ACETATE COPOLYMERS
GLOSSARY OF ACRONYMS
REFERENCES

240
240
240
254
255
258
259
260
261
262
263
268
271
275
275
282
284
285
285
286
287
287
289
290
291

Page 240

I.INTRODUCTION
Significant changes have occurred in polyvinyl chloride (PVC) manufacture in recent years
because of developments in safety and environmental concerns in general and the concerns for
vinyl chloride specifically. For many years the primary safety hazards of VC were thought to be
its flammability and anesthetic properties, plus the need to prevent overpressuring of equipment
by uncontrolled polymerization. Data became available in the 1960s which gave reason for
concern at exposures below the anesthetic/explosive range. Vinyl chloride was shown to be
toxic to the liver in animals; to cause acroosteolysis (AOL), a degenerative disease of the bone
tufts, in humans; and by 1973 it was found to cause angiosarcoma of the liver (ASL), a rare
and usually fatal liver cancer, in both animals and humans. There are lesser health concerns for
the other components of the resin manufacturing process, although each has been found to
have some undesirable effects. A summary of the toxicity data for VC and some of the more
important constituents in PVC is presented in Section II.
These findings have resulted in regulatory action by several governmental agencies, including
the Food and Drug Administration (FDA), the Consumers Product Safety Commission (CPSC),
the Occupational Safety and Health Administration (OSHA), and the Environmental Protection
Agency (EPA), which are discussed in Section III. Similar regulatory actions have been enacted
by many foreign governments. Extensive changes have been made in manufacturing practices
and procedures to meet the challenge to reduce worker and environmental exposure, as well
as because of technological advances. Some of these are presented in Section IV. Section V
describes methods for safe disposal of the waste streams encountered in polymer
manufacture, and Part VI discusses some analytical procedures specifically applicable to safety
and health matters. Some problems relating to copolymer manufacture are discussed briefly in
Section VII. A glossary of common acronyms is given in Section IX.

II.TOXICITY CONSIDERATIONS
This section is intended to provide a general perspective on vinyl chloride and the materials
most closely associated with it in PVC manufacture, from which the specific safety hazards and
work paractices can be reviewed in the later sections.
A.Vinyl Chloride
See Table 1 for a listing of some physical properties of VC; Table 2 contains data on selected
properties of other substances often associated with VC polymerization processes.
1. Acute Toxicity: The anesthetic property of VC was recognized in the early 1930s [1] and has
been investigated by several workers [25]. Death occurs rapidly in animals at concentrations
much above the anesthetic levels of 8 to 12% [1, 6, 7], but no histological damage was
reported at 5% exposure for 100 days [8].
A review of the toxicity of VC in 1943 concluded [9] that vinyl chloride is one of the least
dangerous of the chlorinated hydrocarbons. Our present knowledge of the toxicity of this class

of compounds does not let us derive as much comfort from that statement now as may have
been felt then.

Page 241
Formula weight
Heat of formation, 25C, gas (kcal/mol)
Free energy of formation (Btu/lb)
Density, liquid (g/ml)
32F, 0C
50F, 10C
68F, 20C
86F, 30C
104F, 40C
Refractive index, d15
Freezing point (C/F)
Boiling point, 760 mm (C/F)
Liquid viscosity, absolute, CP
32F
50F
68F
86F
Heat of fusion (cal/g)
Heat of vaporization at 57F (Btu/lb)
Specific heat
Liquid, 25C (kcal/kg)
Vapor, 25C, constant pressure (kcal/kg mol)
Vapor, constant volume
Heat of polymerization (Btu/lb)
Explosive limits in air
Lower
Wt %
Vol %
Upper
Wt %
Vol %
Minimum oxygen content for ignition (%)
Flash point, open cup (C)
Autoignition temperature (C)
Critical temperature (K)
Critical pressure (atm)
Critical density (g/cm3)
Vapor cloud explosion yield
(lb to yield the equivalent of 1 ton of TNT)
Vapor pressure (psia)
-10C
0C

62.50
7.5
-3310
0.9471
0.9293
0.9109
0.8918
0.8721
1.398
-153.7/-244.7
-13.37/7.9
0.225
0.207
0.193
0.181
18.14
158.4
0.38
12.83
10.84
660

8.3
3.5
51.4
33
12
-78
472
431.4
52.7
0.370
24,305

18
26

10C
30C
50C
70C

35
48
115
180

Page 242
Heat of combustion (kcal/mol)
Latent heat (Btu/lb)
0C
50C
Solubility in water, 30C (wt %)
Partial pressure
0.5 atm
1.0 atm
Autogenous
Solubility of water in VC (%)

2826
147
126

0.5
1
2
0.11

However, careful examination of liver effects in animals at 100 to 500 ppm for 4 to 6 months led
investigators at Dow [10] to recommend a human exposure limit of 50 ppm time-weighted
average (TWA). The American Conference of Governmental and Industrial Hygienists (ACGIH)
accepted the 500 ppm recommendation of Yale investigators [11] instead, and published this
first as a TWA, and then as a ceiling value [12]. This was the value set by OSHA in 1971 during
its mass adoption of voluntary consensus standards as regulations, and remained in effect until
1974. The Dow workers later stated [13]: Had our recommendations based upon relatively
simple toxicology been followed then, the difficulties of today may never have occurred.
Human response to acute exposures is very close to that of animals. The human narcotic range
is given variously as 7 to 10%, with 12% being dangerous [4, 11]. Deaths have been reported
of workers exposed to high but unknown concentrations [14] and there are several anecdotal
reports of workers losing consciousness temporarily [1517, 59]. The odor threshold has been
reported as varying over a very wide range [17], but the most reasonable figure seems to be
either 1200 to 2000 ppm [18] or 500 to 1000 ppm [19]. Some sensations from exposure are
reported as confusion, intoxication, burning of the soles of the feet, and subsequent headaches
[20]. There are no immediate effects noticed at 50 to 500 ppm [21].
Vinyl chloride had been considered as a potential dental anesthetic, but the finding of serious
cardiac arrhythmias in dogs and the development of sensitization [2, 5, 22] discouraged this
application. Vinyl chloride is autocryogenic and can cause frostbite if the liquid contacts the
skin. Adequate protective clothing should be worn to avoid this contact.
2. Metabolism: Vinyl chloride is metabolized by the mixed-function oxidative action in the
cytochrome P-450 component of cells [23, 23a]. This pathway is saturable [24, 25] and if an
excess of vinyl chloride is inhaled, the liver capacity is overwhelmed and metabolism to the
carcinogenic intermediate then occurs in other organs of the body, allowing tumor formation to
occur there also. This result has been seen in animals [26], and Bartsch [26a] has shown that
rat and mouse lungs can metabolize VC at 10 to 20% of the rate of their livers. Epidemiological
data [2732] do not show it to occur in humans at either the ambient or occupational
concentrations now experienced.

Humans and animals have similar metabolic routes [33], but that in humans is much slower. This
has been used to adjust risk estimates based on

Page 243
Vinyl chloride Dichloroethane Trichloroethylene Vinyl acetate
Odor threshold (ppm in air)
Explosive limits in air (vol %)
Flash point (F), COC
LD50, rat (mg/kg)
LC50, rat (ppm)

1,000
3.633
108
500
>50,000

Aquatic toxicity (ppm)

>1,000
Vapor pressure, 20C (mmHg)
1,600
Solubility in water, 20C (g/100 g) 1 (1 atm)

50
6.215.9
56
770

1001,000
62
0.8

2080
12.590
None
4,920
8,000
(4 hr)
1001,000
47
0.1

0.4
2.613.4
18
2,920
4,000
(4 hr)
10100
88
2.4

animal data to give figures which are much closer to actual human experience [3435]. The
rhesus monkey appears to resemble humans much better than do rodents [3637].
An equilibrium is established quickly between the ambient and blood concentrations of vinyl
chloride [3, 33, 36] and the blood level decreases just as rapidly upon cessation of exposure.
Thus analysis of breath samples can be used as a rough indication of recent exposure levels. It
has been estimated that consumption of 20 ppm VC in all fluid intake is equivalent to 2 ppm
exposure by inhalation for 24 hr [36a].
Administration of cytochrome P-450 inhibitor blocks the respiratory uptake [38, 39]. Skin
absorption by monkeys of the vapor is only 0.1% as rapid as absorption through the lungs [24].
However, there are anecdotal reports of deep anesthesia in humans exposed to extensive skin
exposure.
3. Chronic Toxicity: There are few animal studies extending past 6 months except for
carcinogenic bioassays. Viola [40] attempted to reproduce AOL in rats by exposing them to 3%
VC for 4 hr/day, 5 days/week for 12 months. He reported that the animals were slightly
soporific and began to show a decrease in weight and reaction to external stimuli. Half of the
animals died of cardiorespiratory complications and two of hematoperitoneum. Most showed
pathological involvement of the brain, liver, kidney, and thyroid. Six showed pathological
alterations of the skeleton, bone metaplasia, and

Page 244
changes in the cartilage. The latter effect may have been the rat equivalent of AOL. There
were, in addition, tumors at various sites. Feron and Krees [41] exposed rats to 5000 ppm, 7
hr/day, 5 days/week for up to 1 year and found tubular nephrosis, focal degeneration of the
myocardium, and spleen damage, in addition to various primary tumors.
Several articles appeared before 1974 describing what has come to be called VC poisoning
or VC disease, although the latter term has now become more closely associated with AOL
than with gastro/neural problems. Many of these reports are not particularly useful because
there are no exposure data and there often is known exposure to other recognized toxic
materials. It does appear, however, in light of subsequent information, that the exposures must
have been quite high for these symptoms to have appeared so quickly. Some of these reports
are listed briefly below.
One article that has been cited frequently as supplying an early warning of the toxicity of VC is
that by Tribukh in 1949, which discusses health conditions in a PVC processing plant in the
USSR [42]. The author actually does not ascribe the health problems to any specific material,
but mentions diphenyl chloride, hydrogen chloride, and other toxic materials as being present.
No measurements were made for VC, but it is unlikely that any significant quantities could have
been present in the workplace because of the type of PVC being used at that plant.
Other early papers reported various gastro/neural symptoms: spastic angioneurosis [43], a
decrease in catalase and an increase in peroxidase activities and glutathione levels [44], a
decrease in albumin and an increase in beta and gamma globulins [45, 46], cardiac
disturbances [47], and lowered thyroid activity and production of 17-ketosteroids [48, 49].
Kramer and Mutchler [50] made a statistical analysis of the difference between a group that
had been exposed to VC for up to 25 years of work history at up to 300 ppm versus other
chemical workers, and found minor changes in certain blood chemistry and liver functions.
More recent articles have examined workers from cohorts that include AOL or ASL disease.
They find portal fibrosis and portal hypertension, thrombocytopenia, esophageal varices, and
abnormal sinusoidal lining cell development [49, 5158]. It has been postulated that these are
early stages of ASL, but there have not been enough observations to confirm this hypothesis
generally. It may well be correct for certain fibrotic conditions [58a].
The other major area of concern, AOL, was described earlier as a degenerative disease of the
bone tufts. It usually is accompanied by Raynaud's disease, and frequently also by scleradoma.
Suciu [48] first reported this disease, then Cordier [59]. These were followed by Harris and
Adams [60], Wilson et al. [61], and Basalaev [62]. One industry-sponsored survey [63]
identified 25 definitive cases and 16 suspect cases in the United States. No certain etiological
agent was found, but the cases were clearly associated with hand cleaning of reactors [64],
where there is a combination of physical joint insult and VC exposure. The disease is most
often seen in the hands and fingers, but occasionally in the feet or back [60]. Dodson [65] could
find no obvious medical reasons for predeliction to the disease in the four cases that he

studied. It appears to be reversible after cessation of exposure [66]. A total of 126 confirmed
cases had been identified worldwide by 1979 [17a].
Maricq [67] found a strong association of capillary abnormalities in the hands with workers
suffering from AOL. Lillis [58] reported that an abnormal Allen test for circulatory efficiency was
found in many affected workers,

Page 245
as well as many other organic symptoms related to the liver and circulatory systems.
Bertozzi et al. [68] studied the status as of 1975 of a group of 4777 workers, some of whom
had been employed since 1952 in VC/PVC production facilities. No control or comparison data
are given, and many different laboratories performed the analyses, so only relative trends
within the cohort can be identified. They stated that the highest exposures were above 800
ppm. Confirmed and suspected cases of AOL increased with the degree of exposure and the
age of the worker, but not with the length of exposure. Abnormal liver results increased with
length of exposure but not the degree. Heavy drinking appeared to act synergistically with
duration of exposure in affecting hepatomegaly and elevated GGT.
Grainger et al. [66] reviewed the literature on symptoms associated with VC exposure and
discussed the symptoms of 88 workers from a factory, 9 of whom were stated to have definite
VC disease. They report a gradation of findings from those with the symptoms to those without,
but do not make comparisons with unexposed controls. They postulate that vascular and/or
immunological changes are responsible for the effects observed and state that they expect no
new cases to develop at current exposures below 5 ppm.
Knowledge of the exact exposures of these cases would be of great assistance in evaluating
the concern for exposures experienced at present, but no definitive estimates have been made.
Suciu [69] reported clinical symptoms associated with exposures that appear to be far too low
in light of industry experiences since 1974. OSHA [70] estimated that reactor cleaners had
been exposed to 1600 ppm in their work. A CEFIC publication [71] has estimated the average
exposure for all European PVC workers in the 19451960 era as up to and beyond 1000
ppm, and there is no reason to believe that U.S. conditions were much different, but even this
is an average for all workers, and the symptoms of AOL, chronic liver damage, and ASL are
more closely associated with reactor entry and cleaning than with other jobs. The National
Toxicology Program [72] quotes IARC data, which also cites very high potential exposures, and
Fishbein [73] quotes several other sources that report high values.
The EPA requires [74] all PVC processors to displace the vapor from reactors with water
before opening for entry, or to employ a procedure of equivalent efficiency. This is based on a
study [75, p. 471] which showed that this reduced the residual content of the reactor vapor to
8000 ppm. This is a new procedure which had not been in general use before 1975. It had
been the practice of some companies to force air through an opened reactor before entry, but
this was generally an unmonitored procedure. Cook et al. [64] reported that unventilated
reactors were often over 3000 ppm, and Filatova and Gronsberg [76] stated that excursions
were seen up to 34,000 ppm. These comments, coupled with the anecdotal reports of
anesthesia of workers [1517, 59], support the conclusion that reactor cleaners certainly were
exposed to recurring concentrations in the range of several thousand ppm. This fact must be
considered in any attempt to evaluate the hazard to workers at the present time, or to the
population at large.
4. Carcinogenicity: The first report of carcinogenicity in animals came as the result of the

attempt by Viola to reproduce AOL-like symptoms in rats [40]. This work, sponsored by a
group of European VC/PVC producers, used a 3% (30,000 ppm) concentration for 12 months.
The attempt to cause AOL was not totally successful; instead, tumors developed at many sites,

Page 246
especially in conjunction with the Zymbal gland, and in the lungs, skin, and bones. The Zymbal
gland is an organ near the ear which secretes the oil used by the rat to groom itself, and is not
found in humans. Most of the other tumors were thought to be metastasized from the Zymbals,
not primary tumors. Perhaps because the exposures were so high and the tumors were not
associated with human organs, this report attracted little response from regulatory agencies.
The sponsors then undertook a much larger study at lower concentrations under the direction of
Maltoni. Partial reports from this study began appearing by 1973 [76a], but final results were
not available until 1979 [26].
In summary, it was found that tumors appeared in rats at several sites, depending on the
concentration used. The lowest doses at which statistically significant elevations of various
tumors were seen were reported as:
Forestomach papillomas
30,000 ppm
Neuroblastomas
10,000 ppm
Zymbal gland carcinomas
10,000 ppm
Nephrolastomas
250 ppm
ASL
Male
250 ppm,
50 mg/kg
Female
50 ppm,
16.7 mg/kg
Mammary adenocarcinoma
5 ppm
The reported finding of an increase in mammary adenoma at very low exposures led to concern
for female workers, particularly when a study of fabricator employees found an excess of
breast cancer among females [30]. However, a case-control follow-up [31] found no
relationship to VC exposure in the cases seen in those workers. In any event, the very high and
variable incidence of such tumors in the controls, about which Maltoni has often commented in
his oral presentations, makes it very difficult to support a conclusion that the test animals
actually did respond at such doses.
Schaeffer and coworkers [76a] found that the Maltoni data fit an exponential expression which
predicts a no-effect level for liver cancer in rats at about 3 ppm, and predicted an average
latency period greater than normal lifetime at about 29 ppm. Many attempts have been made to
estimate human risk from the animal data, but none of these has been successful unless proper
biotransformation factors were applied [265].
A number of other bioassays were conducted by various industrial and governmental groups
[41, 7783]. In general, these confirmed the findings by Maltoni. It was determined that mice
are the most sensitive species, followed by rats. Wistar rats appear more resistant than the
Sprague-Dawley strain. Rabbits and hamsters are much more resistant. Either ingestion or
inhalation produces tumors, with the latter route tending to give tumors at more diverse sites.
This result is consistent with the saturable metabolic process discussed earlier. Both very
young and older rats appear more sensitive than juveniles. This may be explainable by the
differing repair and detoxification capacities at various ages. There is a regular decrease in the

latency period as the dose increases, with liver tumors appearing only at the end of the lifetime
at 50 ppm or less [26] in rats. Mice show a dose response to below 10 ppm for pulmonary
tumors after exposure for 4 weeks [83a].
Drew and coworkers published data [88a] which they believe contradict the statements by
Maltoni et al. [26] and Groth et al. [88b] that older animals

Page 247
appear more susceptible. They found that withholding exposures until later in life produced
fewer tumors in the animals. They did not take into consideration that the low doses used (50
ppm for mice, 100 ppm for rats) yield long latency times, and thus the animals died of other
causes before the VC-induced tumors could kill them. It is true that fewer deaths occur if
exposure is delayed, but that is not necessarily because of increased resistance. In fact, the
time-to-tumor increased in Drew's experiments, as the age at exposure decreased. Thus any
rigorous examination of such data must include a proper consideration of the time factor.
The first connection between VC and cancer in humans was made in 1973 when physicians at
the Louisville, Kentucky, plant of B.F. Goodrich, Inc. recognized the association between three
deaths of workers from ASL [84]. A review of company records [85] revealed several other
cases at that plant.
This tumor is rare. A review [86] of all cases reported in the United States for the period 1964
1974 revealed 167 cases, of which 19 were ascribed at that time to occupational VC exposure,
26 to Thorotrast given medically, and 9 to arsenic in Fowler's solution, also used medically. The
remainder were of unknown etiology, with no connection to VC. There are references in the
literature to one case [87] thought to be associated with hair spray use. The high level of
interest in this specific tumor is such that any subsequent cases associated with environmental
exposure to VC would most certainly have been reported, and none have. For a time, NIOSH
published a summary of VC-related cases, but this task was taken over by John Stafford of
ICI, England [88]. His most recent compilation shows a total of 34 cases in the United States
and 111 worldwide. A summary of the number of cases by countries and of the U.S. cases by
company and by date of death as given in Stafford's reports are shown in Tables 3 to 5,
respectively.
The average latency period in the United States has been 25 years, but with a mode of about
22 years. The latency period in Europe, particularly in Germany, has been somewhat shorter.
There is an unusual clustering of cases in relatively few plants (Table 4). All the U.S.
occupational cases, and almost all of the cases in the rest of the world, are closely associated
with the job of reactor cleaning, which was once done manually at the end of the polymerization
cycle. It may be speculated that differing work programs and job progressions have had some
effect on the rates at various plants.
Ten cases of ASL have been reported in one plant in Canada, the last in 1976, with no new
cases since that time [89, 90]. These cases are completely typical, both as to the clustering
and the medical symptoms. It is also worthy of note that there is, at most, one case of both
AOL and ASL in the same person [88], although both of these diseases are associated with VC
exposure as a reactor cleaner.
An industry-sponsored epidemiological survey of workers in the VC/PVC industry covered 8384
men with at least 1 year of exposure before 1973 [91]. The expected excess of ASL was
found. There were also suggestions of an excess of cancers in the brain, respiratory, and at
unknown sites, and of lymphoma. This study was expanded to 10, 173 workers [27, 92], where

the excess of brain and respiratory cancers continued to be seen without, however, an
association between the brain cancer and exposure. In addition, most of the lung cancer cases
come from the same facility, with many plants having no cases. A follow-up study of this cohort
to determine the status of the workers as of the end of 1980 is underway.

Page 248
a United States 34 West Germany 21 France 17 Canada 10 United Kingdom 9 Sweden 5
Yugoslavia 4 Italy 4 Czechoslovakia 2 Japan 2 Belgium 2 Norway 1 ___ 111 aIncludes 2 U.S.
cases still alive. Source: Adapted from Ref. 88.
Marsh [92a] followed 2490 workers who had been exposed for at least 1 year between 1949
and 1966 in a plastics-producing plant at which included PVC production. Vital status was
determined for 99.7% of these as of the end of 1976. Death records for 98% of the 603
deceased workers showed a slight, but not statistically significant, excess of genitourinary
system cancer (SMR = 154). A follow-up case-control study did not relate these findings to
either type or length of occupational exposure. One case of ASL occurred in this cohort in 1979
[88].
Goodrich,
Louisville, Ky.
Union Carbide,
South Charleston, W. Va.
Goodyear,
Niagara, N.Y.
Seven others

14
9
4
7
__
34

Page 249
a Year Deaths Year of first exposure 1961 1 1946 1962 0 1963 0 1964 1 1944 1965 0 1966 0
1967 0 1968 3 1944, 1951, 1952 1969 2 1949, 1950 1970 1 1946 1971 1 1955 1972 0 1973 3
1945, 1948, 1958 1974 1 1942 1975 4 1945, 1947, 1954, 1962 1976 3 1943, 1947, 1955
1977 1 1946 1978 2 1941, 1944 1979 0 1980 4 1942, 1951, 1955, 1964 1981 0 1982 1 1946
1983 3 1944, 1955, 1963 1984 1 1953 __ 32 aTwo cases living (first exposures 1954 and
1956). Source: Data from Ref. 88.
Several studies have been conducted on smaller groups of workers which are also subjects of
the larger study discussed above. Monson et al. [93] found an excess of brain and lung cancers
in the plant, which developed the most ASL cases in the United States. Waxweiler et al. [94]
studied 1151 workers who had at least 5 years of exposure in four older PVC plants, and
reported an excess of brain, respiratory, and lymphatic cancer, as well as

Page 250
the known cases of ASL. A later study [95, 96] expressed the opinion that it was not VC
exposure that was responsible for the excess of respiratory cancer, and speculated that it may
be due to PVC dust. However, preliminary results on a study of subsequent lung cancer cases
in that same plant [97] do not show an association with PVC dust. Theriault and Allard [90] and
Falk and Waxweiler [98] also state that it is unlikely that PVC exposure is responsible. The
method used in the follow-up study was an elaborate method for determination of exposure
indices for workers potentially exposed to several chemicals [99] which is useful in identifying
which exposure may be most closely associated with the cases.
Beaumont and Breslow [100 evaluated the statistical power of nine epidemiological studies
dealing with possible lung cancer from VC exposure and concluded that the lack of a general
trend in these results indicated that VC is not a human lung carcinogen. This was supported by
the negative results in the two studies with the highest statistical power [27, 29]. They
concluded that the reported studies were compatible with a relationship between VC and brain
cancer. This was based, however, on the assumption that the result of the EEH study [27] was
positive, a conclusion that is not altogether clear.
Tamburro and his associates at the University of Louisville have followed closely the histories of
the ASL cases at the Goodrich plant. This work has been summarized by Dannaher et al. [101].
Diagnostic methods, treatment, and survival are described. The progression of the disease
from the initial focal nodular hyperplasia through fibrosis to necrosis is described in more detail
by Tamburro [102]. An extensive medical regimen for VC-exposed workers was proposed by
Tamburro et al. [103]. A liver scan appears to be the most effective diagnostic tool [103, 103a].
Of the various tests required by the OSHA medical program, the GGTP test provided the
highest positive predictive value but the least specificity, and the ICG clearance test was
recommended as the preferred screen [104], although none is very effective.
It has been suggested that human data show the transport of some metabolic intermediate
from the hepatocyte to the adjacent sinusoidal lining to initiate the first stage of tumor
development [87, 105106]. Ottenwalder and Bolt [107] came to the same conclusion from
animal studies, and this mechanism is supported by other work at Louisville [108].
An extensive multiyear research program at the University of Louisville was sponsored by the
Chemical Manufacturing Association. Much unpublished material on the subjects of metabolism,
immune response, and ASL detection and surveillance methods is contained in the final report
on this project [108a]. In addition, more than 40 papers and talks have resulted from this effort.
Many of these articles have been cited in the preceding paragraphs.
Duck et al. [109] found no excess of mortality, including cancer, in British workers for 1948
1973, while following 2120 workers. Wagoner et al. [110] criticized the mathematical treatment
of the data and stated that there was an excess mortality in the longer-exposed group. Duck
and Carter [111] then made corrections to the numerical results, but did not change the
conclusion. Berry and Rossiter [112] criticized both the original calculations and the changes
proposed by Wagoner and Infante, as did Fox [113], but neither found any evidence of excess

mortality in the group. Fox and Collier [29] studied 7000 men who had worked with VC in Great
Britain between 1940 and 1974 and found no evidence that cancers other than that of the liver
are associated with VC exposure.

Page 251
Frentzel-Beyme et al. [32] reported on 1618 VC-PVC workers in Germany, and could not
confirm the U.S. reports that tumors at other sites than the liver were in excess, and suggested
that this may be because of the consistently low exposures at the plant which they studied. A
paper by Reinl et al. [114] reported excess deaths in German workers, but the authors have
since found calculation errors in the processing of the data. A later summary of this study [115]
found an elevation of lymphatic tumors in addition to the expected ASL cases, but no elevations
of lung or brain tumors in PVC production workers. PVC processing employees did have a
small elevation of brain tumors. Molina et al. [116] found that the Swedish work group had an
elevated heart disease rate, but no tumors other than ASL. A follow-up study of Texas chemical
workers found no relationship between vinyl chloride exposure and brain tumors [116a]. A casecontrol study of 7736 Japanese beauticians who may have used hair spray containing a VC
propellant showed a slight elevation for stomach cancer, but not for liver, lung, or brain [116b].
Workers who fabricated PVC were of interest as a group whose exposure to VC was
significantly less than the workers in the VC/PVC industry [117], but much higher than any
expected exposure to the general population. Chiazze et al. [30] studied 4341 deaths of
employees of 17 PVC fabricators, and found no ASL. There was an excess of deaths from
intestinal cancer in both sexes, and breast and urinary cancer in females, using proportionate
mortality ratios based on an external standard. A case-study follow-up on the breast cancer
deaths showed [31] no relationship to VC exposure. Baxter and Fox [118] found very similar
results in a study of 107 deaths of male fabrication workers in Great Britain. There was no
excess of lung or brain cancer in either cohort.
There is no consistent trend in these occupational studies for an excess of tumors other than
ASL, and it appears that the occasional report of elevated incidence at some other site is only
a quirk of statistics because of the many site/incidence ratios being evaluated.
On an overall basis, about 0.1% of the estimated working population in VC/PVC plants has
been affected by ASL. All of these have been in the most exposed group of reactor cleaners or
associated duties, and in this group the incidence is about 1%. It has been suggested that
some genetic difference, such as metabolic or repair rates, distinguish this susceptible fraction
from the 99% who have not developed the disease from similar exposures. However, it is
difficult to see how that explanation is compatible with the geographic clustering that is
observed. Those who did develop ASL probably inhaled more than 25 kg of VC during their
work exposure [17a].
Several studies have been made of the general population using ASL as the marker disease in
an effort to detect an association with possible environmental exposure to VC. There was no
association with living near a VC handling plant in the general U.S. survey conducted by the
Center for Disease Control [86]. Brady et al. [119] surveyed 26 ASL deaths in New York State
between 1970 and 1975, and found five who lived nearer VC handling plants than did their
matched controls, but could not establish a direct connection with the disease to exposure. Ten
cases of ASL in Wisconsin were examined for possible connection with VC exposure, and none
was found [120]. Baxter et al. [121] found no relationship between distance of residence from

VC emitters and the 47 cases of ASL in the general population of Great Britain reported in
19631973. A later update [122] found one case who had lived the last 6 years of his life near
a PVC plant and three cases where the men had worked in the plastics fabricating industry but
for whom there were

Page 252
no records to indicate exposure to VC. The lack of relationship between residence near vinyl
chloride operations and cases of unknown etiology was confirmed. Saric et al. [123] studied the
deaths during the years 19681971 in an area surrounding a PVC plant that had been in
operation since 1949 and in which three workers had died of ASL. No relationship was found
for liver or lung/bronchial cancer and place of residence for the general population. A similar
study for communities near a Swedish plant that had operated since 1945 and had found four
ASL cases showed [124] no unexpected elevation of fetal mortality, deaths from all cancers, or
cancer of the liver or lungs during the years 19611974. Pancreatic cancer in males was
elevated in the age group over 60. All ASL cases in Holland since 1950 (27 cases) were
studied, and none had any traceable contact with VC [125]. Iturra [126] observed an excess of
cancer deaths in a city in Canada with a PVC plant compared to a similar nearby city. This
difference was principally found in males aged 20 to 64, which is not indicative of a general
pollution effect. The author drew no conclusion as to why the condition existed.
Representatives of the Environmental Protection Agency have stated [127128] that it has been
unable to establish a link between living near VC handling plants and ASL. It awarded a
contract in 1978 (Contract 68022986 to Science Application Incorporated) to examine the
present health of a cohort that was presumed to have been exposed to VC as children, but this
project was not completed.
The disease ASL is difficult to diagnose [129130], is almost invariably fatal within a short time,
and presents a variety of symptoms, including portal fibrosis and hypertension with
splenomegaly and varices, proliferation of the sinusoidal lining, megalocytosia, and
thrombocytopenia [54, 131]. Metastasis is frequently involved. These symptoms are very
similar to those seen in the mouse [132] and rat [41] and the pathology also is similar [133]. No
really adequate early warning tests have been devised [103104], although the gammaglutamyl
transpepsidase test is promising, together with ICG clearance and SGOT. Radiographic liver
scans and tomography and sonography [134] are said to be useful confirmatory tests.
In summary, VC is a classical procarcinogen, and is clearly a human carcinogen, causing ASL in
a small percentage of highly exposed workers. There is suggestive evidence that it may be a
weak general carcinogen at high concentrations, perhaps through an immunosuppressive
mechanism, but more data are required to confirm this suspicion. Several studies of large
populations have not shown a connection between general ambient exposure and an increased
incidence of cancer.
5. Reproductive Effects: Testing of vinyl chloride for mutagenicity has given mixed results,
possibly because of the need for metabolic activation to an active species and because of its
volatility. However, it is clear that it is a mutagen to several strains of bacteria and yeasts and in
fruit flies under proper conditions [135].
Chromosome damage has been reported in workers with VC illness [136], but the changes do
not appear to be permanent, and are repaired after exposure is stopped [137138]. Picciano
[139] concluded that any cytogenic observations were probably related to length and degree of

exposure, and that any genetic risks were avoidable by adequate control of exposure. Basler
and Rohrborn [140] found that this was true for the bone marrow cells of Chinese hamsters
exposed to high levels of VC in vivo. Concurrent exposure to alcohol enhances the changes in
rat mitochondria on exposure to VC [140a].

Page 253
A test was made of the significance of the chromosomal damage to possible genetic risks by
performing a dominant lethal study in male mice, which were mated with two untreated females
for 8 successive weeks after exposure to 3000 to 30,000 ppm of VC for 6 hr/day for 5 days.
There was no increase in the number of early deaths per implantation, and it was concluded
that any expression of harm to the chromosomes of somatic cells was not carried over to stem
cells [141]. Short et al. [142] performed a similar experiment with longer exposures to lower
concentrations, and also found no effect on reproduction or survival. Himeno and coworkers
also confirmed the lack of effect on male mice at 5000 and 10,000 ppm, and reported that
there was no change in sperm shape or mobility [141a].
Hehir et al. [83] included a three-generation study in their program in which parent rats were
exposed to 50 or 500 ppm. VC 1 hr/day, 5 days/week for 10 weeks before mating and the
subsequent three generations were examined for litter size, percent stillborn, growth, viability,
and reproductive anomalies. No effect on the parents or the offspring from VC exposure was
seen.
Studies by Schwetz et al. [143] and by John et al. [144] found no excess fetal wastage in mice,
rats, or rabbits at VC exposures sufficient to cause maternal toxicity. The authors also found
that VC, either alone or in combination with ethanol, was not teratogenic when dams were
exposed on days 6 to 15 at 50 to 2500 ppm VC. The combination of alcohol and VC did cause
delayed development and a higher incidence of some skeletal variations. Rice [145] concluded
that there is no evidence that exposure to VC has produced increased tumors in the offspring of
the animal tests.
Infante [146] has reported finding an excess of congeneital birth malformations in three
communities in Ohio that are near VC processing plants. However, the Center for Disease
Control (CDC) performed a follow-up study and stated [147] that it could not establish any
association between cases and vinyl chloride exposure. Edmonds [148] has discussed the
methodology of the follow-up study, which was of the case-control type, and stated that no
relationship was found between the cases and their parents' employment or place of residence
relative to the VC plants.
The CDC performed two other birth defect studies in areas possibly associated with vinyl
chloride. In one [149] the hospital records for a city in Pennsylvania where a PVC plant is
located were reviewed, and no increase in birth defect was seen. In another [148], hospital
records for Kanawha County, West Virginia, were reviewed for 19701974 and all cases of
birth defects were compared for residence and employment by case-control methodology. The
study concluded that no relationship between infants with malformations and parents' exposure
to VC could be established.
Theriault and Goulet [150] reported a comparison of two cities in Canada and found an
increase in birth defects in the city, which contained a VC processing plant. The increase was
spread over a wide variety of types of defects, and only raw statistics were used. There was
no attempt to compare exposures of the parents, nor were there controls for any other

environmental factors. Thus the significance of this finding cannot be evaluated, and the authors
were careful not to ascribe excess significance to their data. A more detailed study of birth
defects in Shawinigan, Canada, led to the conclusion that stillbirths were not in excess. There
was no relationship between the cases of defects and the parental occupation or residence,
nor were the cases confined to any particular body system [151]. Ambient concentrations up to
45 ppb were reported in this study.

Page 254
Infante et al. [152] have reported an increase in fetal wastage among the wives of workers in a
PVC plant. This study has been criticized by Paddle [153], MacMahon [154], Downs et al.
[155], and by Monson [156] on the grounds of improper data-gathering techniques, incorrect
statistical treatment, and incomplete reporting. Many of the reported incidents occurred prior to
the date of employment at that plant [157]. In addition, the statistical significance of the
reported excess of fetal wastage of exposed workers' wives disappears if those women
subject to chronic spontaneous abortion are omitted. Hass and Schottenfeld [158] and
Clemmesen [159] concluded that the inferences by Infante could not be sustained by the data.
Hatch et al. [160] explored the statistical power of the various studies on reproductive effects.
They found that the Ohio birth defect study [146] was deficient in power, but that the negative
CDC recheck [147148] of this report had adequate power to detect a significant effect, as did
the CDC [149] study in West Virginia, which also was negative. Similarly, the worker study
[152] on abortions and miscarriages had design deficiencies that prevented its results from
being accurate. Their conclusion was that there are no data which point unambiguously to a
relation between VC and reproductive outcome.
In summary, VC does not appear to be teratogenic or to cause excess fetal wastage in animals
or humans. It can cause reversible chromosome damage in somatic cells, but apparently not in
stem cells, and thus does not present a risk of reproductive effects.
B.Vinyl Acetate
Vinyl acetate (VAc) hydrolyzes readily in body fluids to acetic acid and acetaldehyde, both of
which are normal metabolic products [161162]. The acetaldehyde is converted rapidly to
acetic acid, also. Thus there are only minor effects on mammals from moderate exposure to
VAc. Both the acute and chronic toxicity of VAc have been reviewed extensively recently [162],
and only the highlights will be summarized here. The EPA is in the process of developing a
chemical information hazard profile (CHIP) for VAc and has requested that unpublished health
studies be submitted for inclusion.
1. Acute Toxicity: The LD50 doses in rodents by ingestion are reported as 2920 mg/kg for rats,
500 for guinea pigs, and 1613 for mice. The equivalent inhalation dose in 4 hr for rats is 4000
ppm [163]. Humans experience eye irritation at 22 ppm. Rabbits suffer eye irritation from a
500-mg dose and skin irritation after 24 hr from 10 mg. Toxic effects in aquatic life are seen at
10 to 50 ppm in the standard 48 to 96 hr tests [163].
The current TLV is 10 ppm based on human eye irritation. NIOSH has recommended a
maximum exposure of 250 mg/m3, about 83 ppm, and states: The irritations reported have all
been reversible, and there are no known residual systemic effects [162]. Exposure above the
TLV causes throat and bronchial irritation. The olfactory threshold is well below 1 ppm in air
[162a] and about 0.25 ppm in water solution. It is probable that the ready hydrolysis to
acetaldehyde responsible for this low threshold is responsible for the sharp odor around most
acetate copolymer plants. This, together with the irritant effect, serves as a sensitive warning
to potentially harmful exposures. However, olfactory fatigue can occur on prolonged exposures.

A subchronic test [161] found that there was an 8% reduction in body weight in rats, but not
mice, dosed with 5000 ppm in the drinking water for 3 months. This result was not seen at
1000 ppm or lower, and there were

Page 255
no other hematological or histopathological effects. Inhalation exposure at 1000 ppm for 3
months caused decreased weight gain in both species, and in mice at 200 ppm. Irritation in the
lungs also occurred, and mice developed hyperplasia and metaplasia in the bronchi at the
higher dose. Similar effects were seen from 1500 ppm for 4 weeks.
The was no evidence in the same study for teratological effects in rats from up to 5000 ppm in
the water during days 6 to 15 of gestation. Inhalation of 1000 ppm was slightly fetotoxic; lower
levels were not. Metabolic conversion was found to be rapid, with most being expired as
carbon dioxide within 2 hr of exposure. There was no evidence of significant binding to tissues.
Vinyl acetate shares the property of most organic liquids of being an irritant to the skin by its
defatting properties as a solvent for skin oils. Proper precautions should be taken to avoid
direct skin contact during handling [164].
The sulfhydryl group appears to be involved in the detoxification of metabolized VAc, much as it
is in VC. Sharply lower free nonprotein thiol levels are found in rodents after vinyl acetate
treatment [164a].
2. Chronic Toxicity: In the first chronic toxicity study reported, Maltoni exposed rats to 2500
ppm VAc in air for 1 year. Survival was only 50%, and the results have not been reported in
detail, but no neoplasms were found [165]. It is understood, but not confirmed, that similar
negative results and low survival was seen at 1000 ppm in a later study.
A recent small lifetime feeding test of vinyl acetate in water [166a] resulted in an increase in
neoplasms of the thyroid and uterus in female rats at the high dose rate (2500 mg/liter, about a
100-g/kg lifetime dose) but no increase in neoplasms in the male high dose rats or in rats of
either sex at 1000 mg/liter. The vinyl acetate solution was prepared twice a week, so the actual
applied dose was smaller than indicated, and considerable acetaldehyde and acetic acid were
ingested. The authors described the results as not negative and recommended a study with
larger groups and fresh solutions.
This report has prompted a group of producers and users to sponsor through the Society of the
Plastics Industry a large-scale bioassay program and to conduct an epidemiological survey of
industry workers. These studies should be complete in 1985.
The ACGIH quotes data showing no pathological effects at exposures as high as 630 ppm, or
to repeated doses at 100 ppm in rats [166]. Study of a worker cohort with a mean service of
15 years at average exposures of 5 to 10 ppm, with excursions in the range 50 to 300 ppm,
revealed no evidence to suggest chronic effects or serious residual injury from the excursions if
treated promptly [164]. One recent report states that there is a slight elevation of abnormal
chromosomes in exposed workers after 3 years as compared to controls [167]. Only negative
mutagenicity tests have been reported for VAc [168169] with or without activation. References
to teratology studies other than the report cited above [167] were not found.

C.Trichloroethylene
Trichloroethylene (TCE) or other highly chlorinated aliphatic compounds are used as chain
transfer agents in the manufacture of low-molecular-weight polymers, particularly the
copolymers with vinyl acetate. This class of substances shares many toxicological features, so
this discussion will focus on TCE as the prototype for the class.

Page 256
1. Toxicity: Exposure to TCE in the manufacture of PVC will be largely by inhalation, although
skin absorption can occur also. The end effects, except for the skin irritation from contact with
the liquid, are generally the same by either route, and occur primarily in the central nervous,
cardiovascular, and billary systems. The symptoms are related and thus are usually seen
together.
Some critical toxicity data for TCE are summarized in Table 2. More extensive information is
available in references 178 and 179. Death has been reported in humans from inhalation of
2900 ppm. The LD50 in rats is 4920 mg/kg. Oral doses of 6 to 7 g/kg cause death in humans
and rodents, but death from injection occurs at doses as low as 0.1 g/kg.
TCE was used as an anesthetic for many years because of its strong narcotic effect on the
central nervous system. Several patients developed trigeminal palsies following such treatment,
possibly due to reaction products formed with the soda lime in the closed-circuit anesthesia
apparatus [170]. Rapid, shallow breathing is a typical symptom. Industrial overexposures have
been reported to cause headache, dizziness, nausea, and occasionally permanent nerve
deficiencies [171174]. Death often involves ventricular fibrilation. Deliberate misuse, as in glue
sniffing, results in respiratory and cardiac failure in extreme cases and in liver and kidney
damage in less extensive use [175176].
Chronic low exposures (up to 200 or 300 ppm) cause tremors, loss of motor function, insomnia,
and cardiac disturbances [177178]. In general, these same symptoms have been reproduced
in rodents.
Liver damage occurs infrequently in humans, but usually is massive and fatal, suggesting that
some other complication has contributed to the event [179]. The purity of the TCE may be
important also [180], because of the various toxic corrosion inhibitors that may be present.
Mice are more sensitive than are rats to liver damage from exposure to TCE at low doses
[181]. Increased liver size, cell damage, and cell death are seen. This probably results from the
higher ratio of metabolism in mice compared to rats.
Alcohol has long been recognized as intensifying the toxicity of TCE [182] through a competition
for the metabolism detoxification steps. The combination of TCE exposure and alcohol intake
can result in development of red splotches on the skin that have been called degreaser's
blush.
TCE is metabolized by the same general process as is vinyl chloride; however, there is a major
difference in the rates and in the detoxification stage. Little or no TCE is bound to the DNA, but
is excreted as small metabolized molecules [183]. The putative metabolites do not cause skin
cancer in mice [183a] or bind to glutathione or to DNA directly [183b]. Theoretical
considerations suggest that the chlorine-containing epoxide metabolites should decrease in
reactivity as the chlorine content increases, and this has been confirmed experimentally [183c].

The principal metabolic product in humans is trichloroacetic acid, which can be detected in the
urine and used as an exposure monitor. Soleo and coworkers recently suggested that
exposures in the range 15 to 20 ppm produce urine concentrations of this metabolic product
which they consider safe [183d]. An equilibrium is established between the blood and expired
breath contents of TCE, and breath analysis can be used, together with urine analysis, to
estimate exposures [183e].
2. Carcinogenicity: The carcinogenicity of TCE is a very controversial subject. The regulatory
agencies follow a general rule that any positive mu-

Page 257
tagenicity or bioassay tests requires classification of a substance as a potential human
carcinogen, and thus TCE often is referred to in the public literature as a carcinogen. However,
the data are less dogmatic.
Mutagenicity test results have been mixed, with both positive and negative reports. Often, the
positive results are from technical-grade material containing a few percent of an inhibitor, and
the pure material is very weakly positive, so it is not clear what substance is the cause of the
results [179a]. The putative metabolic intermediates are not mutagenic [184] and, as discussed
above, there is considerable evidence that they do not bind to DNA or possess any of the
expected properties of a carcinogen.
Similar problems exist with the carcinogenic bioassays. Positive results have been obtained
with mice at doses that clearly were toxic, using technical material [185]. Rats and hamsters
have been negative consistently, as have other tests with mice, using purified material [185
188]. It is of special interest that inhalation, as compared to gavage, produced a negative
bioassay [188a].
Recent press reports [189] have stated that preliminary evaluation of a repeat bioassay by the
NCI using pure TCE has shown elevated tumor incidence. This study, as were several previous
ones, was conducted at doses causing extensive systemic toxicity. This latest series of studies
used four different strains of male rats at the NTP, and preliminary results indicate that only one
of these four has given a positive response. However, that bioassay program is under review
because of procedural difficulties discussed during a quality audit [189a]. Several authors [23a,
183c, 189b] have developed data which indicate that the intermediate oxide product from
metabolism is not a carcinogen, as is the case for vinyl chloride. Thus the relationship of these
results to the hazards to humans at current ambient level is not clear. Epidemiological studies
have been uniformly negative, and place an upper limit on any risk that TCE may present to
humans [190194].
The EPA has performed an in-depth review of the health effects of TCE [179] and has
concluded that long-term exposure of humans to environmental (ambient) levels of (TCE) is not
likely to represent a health concern-signs of liver dysfunction have been observed only in
experimental animals during exposure to excessively high levels (>1,000 ppm). In regard to
human carcinogenicity it was stated that the more conservative scientific sentiment would
regard (TCE) as a probable human carcinogen, but there is considerable scientific sentiment for
regarding (TCE) as an agent that cannot be classified as to its carcinogenicity for humans. No
evidence of dominant lethal mutations was seen at 450 ppm, nor was there any loss of fertility
or fetal development [194a]. Teratology studies, although unrelated directly to carcinogenicity,
have also been negative in mice, rats, and rabbits, further reducing the concern for harm from
exposure to TCE [195]. See reference 179 for a review of several other earlier reports.
TCE appears to belong to that class of materials which do not cause direct harm to the genetic
DNA, but may, if given in sufficient doses, produce tumor formation in animals by severe organ
damage. This class of substances has been termed nongenetic or epigenetic carcinogens and it

seems probable that they are not actually carcinogenic at doses that do not produce permanent
organ damage [196]. A Committee of the National Academy of Science concluded that the low
carcinogenic potency of TCE requires no special precautions beyond normal good industrial
hygiene practices [197]. Another NAS review group stated [198] that additional long-term
studies . . . should be conducted with purified TCE in order to determine if TCE is a toxicant,

Page 258
mutagen, or carcinogen, and the minimum times and doses that are required to produce
adverse effects. Thus TCE should be handled with respect, but it appears that it can be used
with safety under proper conditions. Meanwhile, its human carcinogenicity remains controversial
[197a]. IARC places it in category 3: cannot be classified as to its carcinogenicity to humans
[197b].
D.1, 2-Dichloroethane
Many of the toxic properties of 1, 2-dichloroethane (EDC or DCE) are very similar to those of
TCE, and the same general precautions should be taken for both substances.
1. Acute Toxicity: The LD50 in rats for a single oral dose has been reported as 680 mg/kg [199]
and 0.77 ml/kg [200]. Deaths in humans have resulted from doses estimated to be in the range
of 20 to 50 ml, and the ability of dogs and humans to regurgitate, which rodents do not have,
appears to permit them to survive higher ingested doses than rodents [201202]. Skin
absorption occurs readily and gives the same symptoms as inhalation or injection, in addition to
the irritation effect of the skin by defatting [178]. Severe pain and irritation results from eye
contact, and foxes and dogs, but not other species, develop an irreversible clouding of the
cornea, apparently from the production of a secondary metabolic product [203].
Inhalation produces the typical halogenated solvent symptoms of drowsiness, nausea,
dizziness, and other signs of central nervous system depression. Liver and kidney damage may
also occur [201202]. Rats survive 200 to 300 ppm for 7 hr, but only 1 hr at 3000 ppm and 12
min at 20,000 ppm [204]. No LC50 data as such have been reported for inhalation exposures,
but can be inferred from other data as about 2000 ppm. The slightly sweet, typical chlorinated
solvent odor at 100 to 200 ppm becomes unpleasant for most persons at 1000 to 2000 ppm
and can cause drowsiness at 2000 ppm in as little as 5 min. Heparin has been used
successfully in treatment of acute poisoning [205].
2. Chronic Toxicity: Animal studies [204, 207209] show little effect on health at prolonged
exposures of 100 to 200 ppm, but levels of 400 to 500 ppm or higher resulted in liver damage,
and there was pulmonary congestion, kidney damage, and deaths at 1000 ppm.
Review of human cases [210211] shows kidney and liver damage. Many of the fatal cases
were from accidental ingestion. Exposure data are uncertain for occupational inhalation. Studies
collected by NIOSH [201] suggest toxic effects at exposures as low as 10 to 15 ppm, but this
is not confirmed.
3. Carcinogenicity: The same controversy exists as to the carcinogenicity of DCE as for TCE.
One NCI bioassay [212] gave increased tumor formation. The initial dose was strongly toxic
and had to be reduced during the experiment to maintain the animals alive. Other studies [186,
209, 213] found no such effects. A National Academy of Science review group concluded that
further tests are needed to settle the issue [198]. Meanwhile, the regulatory agencies
sometimes list DCE as a potential human carcinogen, but it is not included on the NTP list of
carcinogens.

The mutagenicity data to support this conclusion are mixed. Several tests are available which
report both positive and negative results. There is concern that the putative metabolites
chloroethanol and chloroacetaldehyde

Page 259
may be the active species, but these also show mixed results. See references 198 and 205 for
a review of these data.
A multigeneration reproductive study at doses up to 50 mg/kg/day showed no significant
dominant lethal or teratogenetic effects on mice in either of the two generations of offspring,
nor on survival or weight gain [206]. EDC also has been found not to be teratogenic in rats,
chickens, or rabbits, nor does it affect reproductive capacity at doses high enough to show
severe maternal toxicity [208, 209, 214215].
Thus DCE appears to have well-established no-effect levels for its toxi-cological effects, and if
it is a carcinogen, must act through a nongenetic process such as that discussed earlier for
TCE. See references 178, 198, 201, 202, and 215 for more detailed discussion of these points.
E.Polyvinyl Chloride
PVC is an inert, indigestible material with no known direct toxic effects. There has been
concern for problems that may be associated with residual monomers or polymerization
adjuncts, but there are none associated with the polymer itself [216]. The principal health
concern for PVC is from inhalation. PVC is regulated as an inert or nuisance dust by OSHA
(29 CFR 1910.1000, table Z-3), which sets exposure limits at 5 mg/m3 for the respirable
portion and 15 mg/m3 for total dust. The ACGIH recommendation is 5 and 10 mg/m3,
respectively.
Various reports have been issued concerning the effects of PVC dusts on animals and humans,
most of them originating in Europe. Miller and co-workers [217] found diminished pulmonary
function in long-term workers exposed to VC and PVC. Arnaud et al. reported [218] that a
bagger with 23 years' experience had PVC entrapped by the microphages of his lungs.
Waxwieler et al. speculated that his finding of no association between VC exposure and lung
cancer in a plant cohort could throw suspicion on the PVC as a causative factor [95]. However,
this was shown later not to be the case [90, 97, 98]. Mastrangelo et al. studied 20 workers
with high PVC dust exposure and found observable x-ray abnormalities and some respiratory
impairment [219]. Cordasco et al. [220] found similar symptoms in three patients. These
reports gave very little detail on the degree of exposure or the type of resin or other materials
that may have been present.
Two more detailed studies of workers at ICI plants in England have been reported. Chivers et
al. [221] examined the respiratory function of 509 workers, including 112 controls, and
concluded that PVC dust has not produced deleterious effects on ventilatory functions.
A more extensive study of 818 workers, many of whom had worked in older plants
manufacturing plastisol resins, included x-rays and several lung function tests [222]. Their
conclusion was that there was evidence of a slight, nonspecific respiratory effect that was
difficult to distinguish from the effects of aging and smoking. A follow-up study performed about
a year later included the original 818 workers plus others in that plant, for a total of 1047
persons, and a group of 127 workers from a second plant which produced only suspension

resins. The results of the first study were confirmed, with the exception that the men in the
second plant appeared to have slightly less response than those in the first [222a].
OSHA issued a call for information on the occupational effects of PVC, and a symposium was
held in March 1980 [223]. The effects discussed above were reviewed, but no additional data
were presented. The proceedings of

Page 260
this symposium were published as Vol. 41 of Environmental Health Perspective, Dec. 1981.
Similar effects have been observed in animals after exposure to PVC dust as have been
reported for humans [224227]. An unpublished NIOSH study exposed rats (12 months),
monkeys (22 months), and guinea pigs (12 months) to more than 10 mg/m3 of respirable PVC
dust (plastisol grade) for 6 hr/day, 5 days/week. Extensive biochemical, pathological, and
respiratory tests (monkeys only) were performed on the animals. It was concluded that no liver
damage was seen, some accumulation of dust in the microphages occurred, and for the
monkeys there was no impairment of respiratory function [228]. In general, these authors
remark on the mild effect of PVC dust, and compare it to other nontoxic substances.
Some PVC dusts show greater in vitro cytotoxicity and fibrogenicity than do others [229], and
these two properties run in parallel. Richards and coworkers found [230231] that the biological
activity was due to the surface-active agents present on plastisol resins, and that sodium
dodecylbenzene sulfonate was the most active of those tested. Washing with ethanol or water
reduced the activity greatly. Wheeler has emphasized the need to distinguish between plastisol
and suspension resins when considering health effects [223].
Chronic feeding studies of PVC, copolymers, or extracts of these to rats and dogs has not
shown any serious effect [232234]. These resins are prior sanctioned by the FDA for food and
cosmetic applications (21 CFR 121.106).
PVC is like other solids that it will induce local sarcomas when the proper-size pieces are
implanted in rodents. These results have been ' reviewed in several places [162, 235236] and
are not considered relevant to risks for humans [195].
The difficulty of ignition of PVC dust is a function of its particle size. Very little yield is obtained
from material 100 pm or larger, while 10-m material is about as explosive as baking flour
[237]. The presence of small amounts of flammable gas increases the hazard. Solid PVC is
nonsupportive of combustion, and most fabricated products earn the Underwriters' Laboratories
rating of SE-0 unless sufficient modifier or plasticizer is added to offset the lack of flammability
[238].
It is well known that PVC will produce hydrogen chloride upon heating, and this is the basis of
the need for stabilizers during processing. This hydrogen chloride is the principal toxic hazard
during fires that are large enough to force continued burning of PVC articles [239240],
together with the carbon monoxide that may be present in any combustion gases. This aspect
of polymer toxicity is discussed in more detail in Chapter 26.
Some evolution of HCl occurs at any elevated temperature, but studies have shown [241243]
that this is not a significant risk to those who handle the hot materials, such as meat wrappers,
when adequate ventilation is provided. There is no depolymerization of PVC to the monomer,
and the amount of vinyl chloride found in decomposition gases at processing temperatures is
only that expected to be present as residual monomer [244].

III.REGULATORY STATUS
The current regulatory status of the materials discussed in Section II is summarized in Table 6,
where an X indicates that there is a document or rule

Page 261
VC
OSHA
TLV ppm
Standard
ACGIH
NIOSH
Crit. Doc.
EPA
Standard
Priority pollutant
Hazardous substance
RCRA
CERCLA (Superfund)
FIFRA
FDA
BATF
DOT
CPSC

1/5
x
5

VAC

TCE

EDC

50/200

10/20

100/300
(50/150)
100/150

x
x
x

x
x
x
x

x
x
x
x

x
x
x
x
x
x
x
x
x

10/15

by that agency for the substance listed in the column heading. These rules are discussed in
more detail in the following pages.
A.The Food and Drug Administration
The first specific regulatory initiative toward vinyl chloride was in 1973, after it was found that
up to 20 ppm of vinyl chloride could migrate into the contents of miniature liquor bottles. The
Bureau of Alcohol, Tax, Firearms of the Department of the Treasury (BATF) proposed
(38FR12931, Sept. 1973) to withdraw the prior sanction status of PVC for use in these bottles.
The proposal was based on adulteration of the bottle contents, not on any specific health issue
at that time. Bottlers stopped this use of PVC voluntarily, and no further action has been taken
on this proposal.
The BATF has indicated to the FDA that it is willing to reconsider the use of PVC for liquor
bottles if the FDA would clarify the status of PVC for that use. The FDA replied in a letter of
January 13, 1981, from the Deputy Director of Foods to the Assistant Director of BATF that the
FDA was still considering its policy in regard to indirect food additives. Two proposals
concerning that policy were published for public comment at 47FR4972, February 2, and
47FR14464, April 2, 1982. Thus this issue still is not resolved. The BATF has removed formal
barriers to the manufacture of liquor bottles from plastics in general (47FR43944, October 5,
1982) but has not approved PVC specifically.
Earlier, the FDA published in September 7, 1975, at 40FR40529, notice of intent to withdraw its
sanction for the use of rigid and semirigid PVC as

Page 262
food packaging material, while allowing the use of flexible materials to continue. This was
based on the finding that residual vinyl chloride in flexible film was undetectable but could be
found to be present in rigid sheet, and therefore was presumed to migrate into foods. No action
has been taken on this proposal, either. One of the reasons has been the strong activity by
members of the Society of the Plastics Industry (SPI), who presented data to the FDA that the
current very low residual vinyl chloride in PVC does not result in detectable quantities of vinyl
chloride in packaged food [245]. In addition, the District of Columbia Circuit Court of Appeals
ruled in Monsanto v. Kennedy [631F 2nd 947 (D.C. Circuit 1979)] that risks cannot be inferred
or assumed, but must be found by a reliable scientific process. The Second Annual Report on
carcinogens [246] states that the FDA is reconsidering this proposal, and may withdraw it.
Meanwhile, the polymers and copolymers of vinyl chloride continue to be prior sanctioned for
food and cosmetic packaging. Concern over possible action by the FDA has significantly
affected this application, however. A recent industry survey by the SPI found that about 7.8% of
the nations' food supply is now packaged in PVC, and estimated that this could rise to about
11% if the proposal were withdrawn [246a]. No significant exposure to ingested VC can be
expected at this level of use and the current low residual monomer levels.
Vinyl chloride was used as a propellant in a variety of pesticides and cosmetics up to 1973,
when this use was withdrawn voluntarily. When the first reports of ASL issued, the three
agencies having jurisdiction over these uses promulgated bans, the FDA at 39FR30830, the
EPA at 39FR30112, and the Consumers Product Safety Commission (CPSC) at 39FR30112.
The CPSC action was upset on precedural grounds and was later reinstated (43FR12308)
without a by then useless recall provision. The FDA noted (39FR14215) in its proposal that a
generally effective voluntary recall of unused packages had occurred in early 1974.
B.The Occupational Safety and Health Administration
OSHA adopted the recommendations of the American Conference of Governmental Industrial
Hygienists (ACGIH) for occupational exposure to various substances as a part of its rulemaking
by reference. These appear at 29CFR1910.100, tables Z-1 and Z-2. There is no listing for vinyl
acetate. OSHA has not kept up with the subsequent revisions of this list by ACGIH, so that
current ACGIH recommendations (shown in parentheses in Table 6) differ from the official
OSHA exposure limits. The ACGIH recommendation at the time of adoption by OSHA was 500
ppm. See the discussion of this point in Section II.A.1.
The allowable occupational exposure for vinyl chloride of 1 ppm is set by the OSHA workplace
standard at 29CFR1910.1017. This was adopted in 1974, after extensive public hearings, and
became effective in April 1975. OSHA first set an emergency temporary standard of 50 ppm
and proposed a permanent limit of nondetectable exposure by a test sensitive to 1 ppm.
Economic impact studies sponsored by both OSHA and industry showed that such a limit was
not feasible [245a], and OSHA then promulgated a final standard of an 8-hr time-weighted
average (TWA) of 1 ppm, and a 15-min ceiling of 5 ppm, without regard to respirators.
In brief, the regulation sets:

1. An action level of 0.5 ppm below which no response is required.

Page 263
This generally exempts most fabrication plants and laboratories and many monomer plants.
2. A regulated area where exposures are above 0.5 ppm, which restricts entry to authorized
persons.
3. Medical examination schedules and exposure record retention for authorized employees.
4. A list of acceptable respirators for use at exposures over 1 ppm.
5. Monitoring and alarm systems for the workplace, and routine measurement of worker
exposure.
6. Labeling and signs for regulated areas and containers of vinyl chloride and PVC.
7. Work procedures for hazardous operations.
8. Training programs for employees.
OSHA described this as a feasible standard, and states that it was not derived from health
considerations, therefore should not be considered a safe exposure limit.
It is difficult to determine the cost to the industry for compliance with this standard because of
the work going on at the same time to respond to the expected EPA emission standard. In
addition, some steps had been undertaken earlier by industry in response to the concern for
AOL [63]. The immediate effect was a sharp drop in productivity as new work practices were
instituted and equipment was installed. Some of this productivity loss was recovered later, but
there is a general industry consensus that there has been at least a permanent 5 to 10% loss
[247, 248]. A few of the smaller older plants shut down, amounting to about a 5% loss of
capacity and jobs [248a]. Trade association estimates prepared a few years later indicate an
overall capital expenditure of about $200 million that can be ascribed directly to OSHA, and
added annual expenses of about $25 million.
Several authors have attempted to estimate the cost-effectiveness of this standard on a costper-life-saved basis. This is an especially difficult approach, because of the uncertainty of what
exposures may have been in the absence of a standard, the inaccuracy of the risk
assessments which were made from animal data, and the reluctance of most persons to
accept a monetary value for a life. Nevertheless, Graham and Vaupel [249] estimated that the
OSHA rule cost $7.5 million per life saved and $490,000 per life-year saved over the option of
leaving the limit at 50 ppm. Luken and Miller [250] arrive at an imputed value of $4 million per
life, while Morrell [251] derived the higher cost of $200 million per life, assuming that the
residual incidence rate would have been 0.1 case per year without the standard. Northrup [252]
calculated that the cost was $9 million per life, or $450,000 per year of life saved, based on no
voluntary action by industry. There are no specific OSHA standards for the other materials in
this group except for the 8-hr exposure limits of 40CFR1910.1000 that are shown in Table 6.

C.The Environmental Protection Agency

1. Air Emissions: There are no recognized natural sources for the materials discussed in this
chapter. Vinyl chloride is suspected to be formed by the photochemical decomposition of other
halogenated materials, but this has not been confirmed. It is formed by the biological
degradation of other chloroolefins (see below). Hoffman reported [253] that vinyl chloride was

Page 264
found in tobacco smoke, and speculated that it might be present in combustion gases from all
chloride-contaminated organic materials, and therefore may be ubiquitous.
Grinard calculated [254] that the steady-state, worldwide ambient concentration in 1973 was
about 1.4 ppt based on emissions at that time. Vinyl chloride is active photochemically with a
half-life in sunlight of about 5 or 6 hr [255258]. The reaction rates are slightly less than those
of ethylene in the reaction with NO, and considerably less in the reactions with ozone. The
residence time in the atmosphere was estimated as 1.8 days by Singh et al. [259], with a 43%
loss per 12 hr of sunlight, based on the rate of reaction with hydroxyl radical only. Smog
chamber data were used to predict a consumption rate of 12% per hour, and the study showed
a rate of about 18% [259a].
The EPA has conducted three ambient monitoring programs around VC-handling plants [117,
255]. The first, in 1974, found measurable quantities at distances up to 0.5 km from a PVC
plant. The third program failed to find significant quantities at the fence line of five large
fabricating plants. The results of the second program have not been released formally, but an
analysis of the data has shown that the average concentration in early 1975 at the plant tested
was about 40 ppb at 500 m from the plant center, 10 ppb at 1 km, and 2 ppb at 2 km. The EPA
had calculated [75] an average exposure of 17 ppb to persons residing within 5 miles of a
typical PVC plant, using emission data and modeling techniques which were strongly disputed
by industry. The estimated 95% reduction of emissions by the current standard presumably
reduces the current exposure to those within 5 miles of 0.4 ppb, by the EPA assumptions. The
generally accepted field monitoring method for VC has a lower sensitivity of 10 ppb, so that
these estimates cannot be verified broadly, but recent tests in one plant suggest that the actual
values are only 10 to 25% of this estimate [248].
Vinyl chloride, TCE, and DCE currently are found in the ppb range in the air around industrial
locations, but generally not in detectable quantities at rural sites [179, 259262]. There are
diurnal and seasonal variations which link the emissions to human activity. A summary of an
extensive EPA-sponsored survey is available in paper copy or on computer tape [262a].
The EPA formed a study group early in 1974 while OSHA was conducting its rulemaking.
Several publications discussed the environmental aspects of VC [75, 255, 263264] and the
EPA ultimately concluded that there were nearly 5 million residents within 5 miles of VC/PVC
production facilities that had been exposed to an average annual concentration of 17 ppb for 30
to 40 years, and that this led to the possibility of 10 to 20 deaths per year [127]. This risk
estimate has several serious flaws [265] and a more realistic estimate is several orders of
magnitude lower. In any event, there have not been any cases of ASL which have been
attributed to general ambient exposure. See the discussion in Section II.A.4. The final EPA
standard was calculated by the EPA to reduce environmental exposure by 95%, and thus the
projected death rate to less than 1 per year [266].
The final EPA standard [74] establishes the following conditions:

1. Fugitive emissions controls by leak patrols and design standards for pump and compressor
seals, agitators, and loading devices
2. Work practices for vessel openings and sampling
3. Stripping requirements for residual monomer in resins and waste-water

Page 265
4. Abatement of all point-source emissions to 10 ppm
5. Prohibition of relief valve discharges, except for nonpreventable emergencies
6. Extensive monitoring, reporting, and recordkeeping requirements
7. Specific analytical procedures
The agency prepared a model plant by choosing among the best control devices existent in the
industry [75] and it now judges the compliance of a particular plant by the projected emissions
from that model.
As stated earlier, it is difficult to separate the compliance costs for this standard from those of
the OSHA standard. An agency report [267] estimated that the cost of compliance was $296
million through July 1981, and that an additional $470 million would be spent in the next 5 years,
all calculated in 1977 dollars. A more recent update of this report prepared for the agency put
the costs for 19811990 at $981 million. If the agency estimate of 20 deaths per year is
accepted, this would be a cost of about $5 million per life saved. However, there is no evidence
that any lives have been saved by this standard and we do know that at least two workmen
were killed during the compiance efforts of the industry. The industry position has been that the
OSHA standard had provided adequate protection to the environment, and that the EPAgenerated costs were unnecessary.
Environmental groups challenged the standard as being too lax, and as part of a settlement
agreement EPA proposed additional restrictions on emissions (42FR28154, June 2, 1977). That
proposal has now been withdrawn as part of a proposed revision of the standard which would
set numerical limits on emergency relief device emissions rather than try to define
nonpreventable (50 CFR 1182, January 9, 1985).
Meanwhile, the U.S. District Court for the Middle District of Louisiana dismissed cases brought
by the EPA against Ethyl Corporation and Occidental Chemical Corporation for violations of the
emergency discharge rule (40 CFR 61.65(a)) on the grounds tht the standard is a work
practice standard and not an emission standard, and as such was not authorized by the Clean
Air Act as it existed in 1976 [267a]. Work practices were authorized by 1977 and 1978
amendments to Sec. 112 of the Clean Air Act, but this was after promulgation of the standard,
and thus the court noted that the present standard was subject to the Supreme Court ruling in
Adamo Wrecking Co. v. United States [334 US 275 (1978)], which held that work practices
were not authorized at the time of promulgation.
At the same time, the District Court for Massachusetts has denied the petition by Borden, Inc.,
for dismissal of a similar suit. It ruled that the Adamo decision did not apply, and that Borden
was barred from challenging the standard at this time.
The agency has filed notice of appeal in the Louisiana rulings, and has continued to file

additional cases against other facilities on an identical basis. It appears that this matter may be
in litigation for some time.
Industry argued the same point that was made by the Louisiana court during the rulemaking
procedures, but was not successful at that time. After promulgation, attempts to obtain a
clarifying guideline from the Standards Setting Group in the Air Office were unproductive, and
the Enforcement Office issued a series of memoranda which took a literal view of an
emergency being little other than an Act of God. The effort to attain relief here was
complicated by the outstanding proposed amendments to the standard of June 1977, published
as the result of the negotiated agreement with the National Resources Defense Council after
their petition for review of the original stand-

Page 266
ard, which would have made the standard more stringent. Neither these proposed amendments
nor the requests for clarification of the definition of emergency were acted on.
The other materials discussed in this section are not regulated by specific emission standards,
but all of them, including VC, are affected by general rules on volatile organic emissions, and
thus come under the State Implementation Plans and New Source Performance Standards,
which often require stricter control measures than does this standard. The oxychlorination
reactor in particular has been subjected to additional controls (48FR40278, September 6,
1983), as have storage tanks. Most new volatile organic chemical manufacturing facilities are
subject to a leak detection requirement similar to that for VC (48FR48328, October 18, 1983).
Vinyl chloride, TCE, and DCE are listed as priority pollutants under Sec. 307(a) of the Clean
Water Act, and Water Quality Criteria Documents have been prepared for them [274]. This
subjects these materials to special considerations when discharge permits are issued. A Health
Assessment Document has been prepared for TCE [179]. These documents summarize the
evaluation by the agency of the health risks from these substances, and give some data on
methods of waste treatment. Other sources of agency data on VC, TCE, and EDC can be
found in the Treatability Manual [262], Multimedia Environmental Goals for Environmental
Assessment [273], and Fate of Priority Pollutants in Publicly Owned Treatment Works [273a].
Vinyl acetate, DCE, and TCE also are subject to reporting requirements for spills as hazardous
substances under Sec. 311(b) of the Water Pollution Control Act and Sec. 102 of CERCLA.
See 40CFR116, 117, and 302.4. Reportable quantities (RQ) of spills are 5000 lb for DCE, and
1000 lb for vinyl acetate and TCE. Vinyl chloride has a temporary 1-lb reportable quantity under
CERCLA, pending designation of a permanent figure by the EPA. Reports are not necessary
for releases at federally permitted facilities. The EPA has established a new RQ for vinyl
acetate of 5000 lb (50FR13456, April 4, 1985). The EPA issued a rule at 49FR5308 under
which the distillation residues from the preparation of EDC and VC are to be listed specifically
as hazardous wastes, rather than as the result of broader RCRA rules. This ensures that all
such wastes are to be disposed of only by RCRA-approved procedures.
2. Water: All the substances discussed in the preceding section are also controlled under
general discharge provisions to the extent that they produce conventional pollutant properties,
such as biological oxygen demand (BOD). As would be expected, vinyl chloride has low water
solubility and is easily lost to the atmosphere from streams and discharges [268, 269]. Under
one set of experimental conditions, the evaporative half-lives of VC, TCE, and EDC are less
than hr [269]. In another experiment, a stirred breaker lost 96% of its original 16 ppm of VC
in 2 hr, while an unstirred beaker lost 25%, at 22C. Plots of log concentration versus time gave
straight lines, indicating volatility to be the only important loss mechanism. There was no
difference in loss rates between distilled water, river water, or industrial effluent. It does not
appear to be absorbed by microorganisms, as shown by tests with five mixed bacteria
populations, three mixed fungal populations, two axenic bacterial cultures, and one algae. The
mixed bacteria did not degrade the VC, nor was it toxic to the bacteria at concentrations up to
900 mg/liter [268]. It does not bioaccumulate in the food chain [270].

The EPA has reported [255] finding VC in the water supplies of some cities in the ppb range.
The concentration was higher in the finished than in the raw water, indicating that it may be
produced in the chlorination

Page 267
step. Dressman and McFarren [271] have found VC in the ppb range in water from distribution
systems using PVC pipe. It is also present in the discharge of some VC-handling plants in the
low-ppm range [263]. Banzer [272] states that no extraction of VC occurs from pipe containing
less than 1 ppm residual by a test sensitive to 2 ppb.
The EPA and the FDA have entered into a memorandum of understanding which assigns the
EPA the responsibility for regulating exposure for VC extracted from plastic water distribution
pipe [274a]. No specific action has been taken in this regard, possibly because of the rapid
decline in detectable amounts in drinking water after new piping systems are put into use
[274b]. There is in place a voluntary industry standard limiting residual VC in the finished pipe to
10 ppmw.
This program appears to have been successful in eliminating the leaching of VC into water.
Compliance is being monitored by the National Sanitation Foundation (NSF), whose stamp of
approval is required by most codes before pipe can be sold or installed for potable water use.
A recent report [274c] described a test for chlorinated organics in water extracts of plastic pipe
which is sensitive to 2 ppb, and the president of NSF has stated [274d] that VC is not found in
the extract at this limit of detection.
Similarly, TCE and DCE are found in streams and water supplies in low concentrations [179,
262, 273, 274; see also 47FR9350]. These seem to be present because of a combination of
waste discharges, formation during chlorination of water supplies, and in the case of TCE, use
by individuals as a degreasing treatment for home sewer systems. Some state legislatures are
considering a prohibition of this last use.
A relationship between vinyl chloride and chlorinated ethylenes in ground-water has been
established by recent work of Florida International University for the EPA [274e]. Analyses in
the outer perimeter of tri- or tetrachloroethylene-contaminated groundwater showed a larger
than expected decrease in the original contaminants and the presence of vinyl and vinylidene
chloride and 1, 2-dichloroethylene. Experiments with anaerobic bacteria from Florida muck
demonstrated that several classes of anaerobic bacteria, including the ubiquitous Escherichia
coli, can biodegrade the more highly chlorinated ethylenes to the mono- and dichlorinated
derivatives.
Their work also confirmed the resistance of VC to further biodegradation, and reported a halflife of greater than 60 days under their experimental conditions, as compared to 43 and 34
days, respectively, for the tri-and tetra-substituted ethylenes. They developed a spray aeration
system which attained greater than 95% efficiency for removal of VC per stage. Vinyl acetate
would not be expected to be persistent in water because of its rapid hydrolysis, which is
reported to be 8.5% per day at room temperature and 4% per day at 4C in tap water [166a].
The EPA has in progress two rulemakings which will regulate the allowable emission of VC,
TCE, and EDC beyond that of the general emission standard. On March 4, 1982, the EPA
published an advance notice of proposed rulemaking (47FR9350), requesting public input on

work under way to limit volatile synthetic organic chemicals under authority of the Safe Drinking
Water Act. It stated that it was considering establishing Maximum Concentration Limits of 1 to
100 ppb for VC and EDC, and 5 to 500 ppb for TCE. The Health Assessment Documents
prepared as part of this initiative still are undergoing scientific review.
On March 21, 1983, the EPA published a proposed regulation (48FR11828) containing effluent
limitation guidelines for the organic chemicals and plastic industries which would place a 50 ppb
maximum for the VC content of plant

Page 268
effluents. The limit for TCE is proposed as 75 ppb maximum for any one day, and 50 ppb for
the average of any four consecutive days. The EDC limits are 150 and 100 ppb, respectively.
The standard for EDC would apply to both direct discharges and those whose effluent is
treated further at public treatment works; those for TCE and VC apply only to direct
discharges.
3. Solid Wastes: Vinyl chloride is listed as a hazardous waste under the Resource Conservation
and Recovery Act (RCRA), as toxic and ignitable. Vinyl acetate and EDC are ignitable wastes,
and TCE is a toxic waste under these rules (see 40CFR261). Any disposal of these substances
is subject to permits under RCRA. The EPA recently included the distillation wastes from the
production of VC and EDC as hazardous wastes, as noted earlier.
4. New Product Manufacture: The EPA also administers the Toxic Substances Control Act
(TSCA), which establishes health and environmental regulations for both new and existing
substances. No one may manufacture or use a substance which is not on the agency's official
inventory unless the EPA has accepted a Premanufacturing Notice (PMN). Polymers containing
more than 2% of an incorporated substance are required to be on this inventory, in contrast to
European rules, which do not cover polymers. Incorporation of less than 2% of a substance is
not deemed to be a new polymer, and a PMN is not required for manufacture. Producers have
the choice of making these resins of low comonomer composition without a PMN, in which case
they are not placed on the inventory and future variations are limited to less than 2%, or of
filling a PMN, in which case future variations may contain any desired amount, but full disclosure
of all processing ingredients and conditions are required. See 40CFR720, 48FR21722, May 13,
and 41132, September 13, 1983, for more details on this regulation.
The EPA provided a very limited exemption for certain classes of polymers under the PMN rule,
but the conditions for this exemption are so rigid that industry has opposed the proposal and
asked for a blanket exemption for all structural polymers similar to the European rules
(49FR46066, November 21, 1984).

IV.GENERAL SAFETY PROCEDURES

Extensive changes have occurred in the work practices and procedures in monomer and
polymer plants in the last few years as the result of concern for the health effects of VC. These
have been amplified by efforts to comply with the OSHA and EPA standards on vinyl chloride
[248, 252, 275, 276], and the broader volatile organic emission rules. There is no doubt that
these compliance efforts have made the industry a safer place to work.
The VC/PVC industry always has been a relatively safe industry from the standpoint of major
disasters and fatalities due to the inherent hazards of the processes, even when the ASL cases
are considered. Nevertheless, there have been several major incidents that resulted in fire and
loss of life [276a] one of which ranked among the 100 largest losses of the last 30 years
[276b]. Table 8 contains a list of major incidents in the past few years. Thus the basic
considerations of the flammability and the need for careful control of polymerization still remain

the controlling factors in efforts to protect life and property.

Efforts to improve plant safety and to develop more effective means of compliance with health
regulations have been aided significantly by voluntary participation in trade associations. The
Vinyl Chloride Safety Associa-

Page 269
tion has been especially productive, but substantial developments have come from the
Compressed Gas Association, the Chemical Manufacturers Association, and the Vinyl Institute
and its predecessor organization, which was an arm of the Society of the Plastics Industry. The
industry has been very generous in exchanging nonproprietary information relating to safety and
health.
A formal Process Hazard Review (PHR) is used by many organizations to identify, evaluate,
and plan remedial actions for potentially hazardous situations that may arise in plant operations.
The PHR is in many ways an outgrowth of accidental investigation, the difference being that the
study is made before an event, with the intention of preventing it [287a, 290e]. Emphasis
usually centers on the reactor and its attendant systems.
A team with broad experience conducts the PHR. The skills included should cover plant
operations, safety engineering, environmental engineering, and experts in the process involved,
and also any particular special problems posed by the process. This usually involves four to
seven persons for all of the study, and others for portions. The initial phase may occupy 3 to 10
days at the plant sites, after gathering the initial background documents and data which will
help identify the concerns to be addressed.
Much of this first phase consists of a line-by-line examination of the process flow diagrams, but
field inspections and interviews are important also. If the review is for an operating plant, plant
operating and support personnel should be interviewed. If the plant is in the design state, the
designers and the future operators should be included.
Each section or unit operation is discussed by someone familiar with the design concept and
operating philosophy. A person familiar with the fault tree concept then leads in an examination
of each major piece of equipment for potential failure modes, determining (or speculating on)
conditions that could cause, prevent, or modify such failures.
The results can be depicted in a fault tree, which show those conditions necessary to cause
particular result as an and branch, and gathers those various groups that may cause the
same result into or groups. Figure 1 is a generalized fault tree for the failure under pressure
of a monomer storage tank, such as a recovered monomer vessel, and Figure 2 is a more
detailed analysis of one of the branches of Figure 1, the probability of an external fire in the
process area. The result is a description of the events necessary and sufficient to cause the
potential incident [234].
A second phase of the PHR is to attempt to quantify the probability of each contributing event,
and of the postulated end result. This is usually done by a specialist in risk assessment, who
often is the leader of the team. Where available, actual experience for failure rates should be
used, but publications giving predicted rates or rates for similar siruations are available [290f,
g]. In this way those conditions or combinations of conditions which are unlikely to occur or to
lead to a dangerous result can be identified and more attention paid to more probable
scenarios.

The third phase consists of preparation of the PHR report, which contains specific
recommendations by the whole team, for preventative measures for the serious consequence
conditions, with appropriate priorities and the assignment of responsibility for implementation of
each action.
The final phase is a progress review of the implementation steps. Further change or revisions
may be required as the result of new data or cost estimates, or further examination of the
interaction of the changes made. For new plants, the Operational Readiness Inspection (ORI)
before startup should include a further review of compliance with the PHR report.

Page 270

FIGURE 1 Fault tree for pressure failure of recovered monomer tank.

FIGURE 2 Fault tree for external fire in the process area.

Page 271
A major uncertainty in the quantification analysis is the weight to be ascribed to operator error.
This factor is often larger than the other factors that errors in estimation of its value override
the other uncertainties in the estimate [290g]. This emphasizes the value of the thorough
operator training in the prevention of accidents.
Another fact apparent from such an experience is that an accident seldom results from a single
event, but that a combination of failures and errors is necessary in most cases. This is because
of the redundancy and diversity of design safety built in to most facilities. The use of the PHR to
identify likely fault paths influences both design and operating philosophies so as to reduce the
probability of undesirable combinations.
A.Raw Material Handling and Storage
1. Vinyl Chloride: The shipment of vinyl chloride is regulated by the Coast Guard (46FR40.151)
and the Department of Transportation (DOT) (49CFR172,-173). In addition, OSHA has specific
rules for in-plant labeling of vinyl chloride containers (29CFR1910.1017) and requires that all
shipping packages or containers of PVC carry a warning that PVC contains VC.
The principal method of shipment for vinyl chloride is by railcar, in type 105A and 112J tankers,
with capacity of up to 30,000 gal. A few users are near enough to monomer suppliers to use
pipeline transfer. Two suppliers can furnish barge loads of monomer, and one of these
maintains a terminal in the Northeast from which tank truck or rail transhipments can be made.
International shipments are made by specially equipped tankers.
Concern for the safe disposal of the contents of derailed tankers has led to the development of
a technique to pierce the car shell with shaped explosive charges [276c, 290h]. The escaping
material is then ignited. Leaking cars not already on fire are also ignited. This is a general rule
for leaking vessels in which the leak cannot be stopped. The danger of an explosive vapor cloud
is far greater than that of the combustion products, and point-source flames should never be
extinguished unless the source can be stopped. Reports of some recent transportation
accidents involving VC can be found in the references listed under [276c].
The Compressed Gas Association, Arlington, VA 22202, has developed a field repair kit that
can be applied to leaks that have developed in the loading dome in valves or at welds. Two
sizes are available, which will fit most of the cars now in service.
The Department of Transportation has required that all new cars put in vinyl chloride service be
equipped with spade couplers, head shields, and has set minimum body insulation and relief
valve specifications. A schedule has been set for retrofitting of all existing cars (49FR3468). An
analysis of rail accidents since this change has been required indicates that it has sharply
reduced the severity of such incidents [276d].
The Society of the Plastics Industry, New York, has organized a mutual assistance program
between VC producers and users that can supply an Emergency Response Team for

assistance from nearby participants in case of a transportation emergency. This is coordinated

through the CHEMTREC emergency number, (800)-4249300, which should be the first contact
point.
A principal source of information on the safe handling of VC is in the data manuals provided by
the manufacturers. The National Fire Protection Association (NFPA) codes specify the electrical
and fire protection standards which are recommended [277]. Insurance carrier requirements
and local build-

Page 272
ing codes vary, but generally require as a minimum the standards of the American Society of
Mechanical Engineers (ASME) for pressure vessel construction. Many of these factors were
summarized [278] a few years ago, and the following discussion is drawn from this source.
Vinyl chloride is stable in the absence of oxygen, water, and light and may by handled safely in
iron, steel, or stainless steel if those substances are excluded. The use of phenol or other
inhibitors to stabilize against spontaneous polymerization of the pure monomer was stopped
more than 15 years ago. Inhibitors are still used in recovery systems to prevent polymerization
of the recycle monomer under conditions that favor peroxide formation.
Oxygen reacts readily with VC to form a variety of cyclic and linear peroxides by simple
addition [279283]. These peroxides are shock sensitive, decompose violently on heating, and
can initiate polymerization in either the monomer or the water phase [268]. They can be a major
safety hazard in any manufacturing system which involves water and has the possibility of the
entrance of air, and have caused serious damage in industrial accidents [283284]. The
peroxide formation is accelerated by the presence of acids and aldehydes, but can be
prevented by the presence of a base [234, 283]. Careful warming with 5 to 10% sodium
hydroxide in water, or preferably methanol if polymer residues are present, is adequate for
removal from process vessels.
The EPA standard establishes specific rules for vinyl chloride unloading and handling
(40CFR61.60). These are in addition to the general unloading rules prescribed by the
Department of Transportation in 49CFR174.67(i). Unloading systems must be designed to
minimize the release of vinyl chloride when the lines are disconnected, and slip gauges no
longer are allowed. Pumps must be equipped with double mechanical seals, or the equivalent.
Rupture disks are required under the relief valves on storage tanks to prevent leakage through
the valve.
Storage tanks should be provided with adequate deluge systems to protect against local or
adjacent fires. Efficient deluge systems which provide an effective water curtain sometimes can
help prevent the spread of a vapor cloud, or the flame from its ignition. Saturation of the air with
moisture also reduces the ignitability of the vapor. Drainage area within a dike is recommended
which is sufficient to allow any spillage to run away from under the tank. The relatively high heat
of vaporization of VC will cause large liquid spills to pool for several hours, thus increasing the
potential for heat damage from fire. Fireproofing of supporting steel is required by most
authorities. Critical valves in liquid service should have firesafe seats and automatic closing
devices. Relief valves should be designed for at least the capacity required by a pool fire.
Redundant level control systems are desirable to prevent overfilling. Insulation and/or reflective
coatings are recommended to reduce heat input. All piping sections that can be isolated by
valves should be provided with protection from thermal expansion damage.
Earlier prohibitions against the use of brass or copper instruments or tubing were due to the
concern for formation of copper acetylide from the acetylenic impurities in monomer made by
the addition of hydrogen chloride to acetylene. This does not seem to be a problem with VC

from the oxychlorination process. Consideration of the ignition temperature and energy of vinyl
chloride suggests that spark-proof tools are not necessary, and this practice no longer is
customary [284a].
Small fires can be extinguished with carbon dioxide or carbon dioxide-generating solid
extinguishers, but care should be taken for reignition of

Page 273
the flame if vapor remains in the area of hot surfaces. Larger flames should be allowed to burn
if the source cannot be closed off. Water is not effective on liquid fires because it is heavier
than VC, and of course, freezes on contact with the liquid, serving only to vaporize the liquid
faster. Air-supplied respirators should be used by firefighters that will protect against both the
carbon monoxide and hydrogen chloride that are produced. Water spray or curtains should be
used to protect adjacent equipment, reduce the spread of the fire, and to some degree, absorb
the combustion gases [278a].
Conventional gas dispersion models of the type used by the EPA for studying diffusion effects
[278b] can be helpful in preparing rough estimates of the size and concentration of vapor clouds
for relatively small VC releases, particularly when no pool of liquid is involved. These small
releases do not cause significant changes in the density of the gas volume. However, large
releases, especially those which involve evaporation from liquid pools, should be estimated by
use of dispersion models designed especially for heavy gases. It is necessary to consider the
effects of changes in gas density and cooling which result from such large releases [278c].
All equipment should be properly bonded into a common grounding system to prevent static
accumulation. This is especially important in flanged equipment, rail or truck unloading systems,
and polymer air conveying systems. Continuity of the grounding system should be checked on a
regular basis.
It should be noted that there are no cartridge-type respirators available for VC service which
have effective end-of-service indicators. Therefore, as a general rule such respirators should
not be reused and should be reserved for short-term service where the VC concentration is
known to be low. Airline pressure-type respirators are used more widely than other types for
this reason. All respirator programs must comply with OSHA regulation 29CFR1910.134.
2. Vinyl Acetate: Stainless steel construction should be used for storage and handling of vinyl
acetate because of corrosion from the acetic acid potentially present from hydrolysis. Inhibition
with a few ppm of hydroquinone or an equivalent is still practiced. The flammability of the
vapors requires exclusion of air in the headspace, and nitrogen padding is a usual practice. In
addition, fill and recycle lines should return to the bottom of the tank and not generate spray in
the vessel. Hydrocarbon emission rules vary with the locality, but some type of emission control
is required by most authorities. Coast Guard and DOT rules apply to shipping, and EPA spill
control rules apply to storage.
3. DCE and TCE: Carbon steel equipment is satisfactory for dry pure material; stainless steel is
needed for wet or recycle streams. The same general type of rules for vinyl acetate applies to
transportation, emission controls, and storage. Aluminum is not a satisfactory material of
construction.
4. Initiators: The search for greater productivity in the polymerization cycle has led to the use of
more reactive peroxide initiators. Most of those in use now require storage well below ambient
temperature. The oxygen content of these materials is such that they often need little external

oxygen for combustion. This reactivity has led to the classification of initiators by their selfaccelerating decomposition temperature (SADT), the temperature at which the decomposition
is self-sustaining and becomes violent, usually with self-ignition [285]. Table 7 presents the
SADT and recommended storage temperatures for several commonly used materials [286
287]. Dilution

Page 274
Name

SADT (F) Recommended storage temperature (F)

t-Butyl peroxypivalate
t-Butyl peroxyneodecanate(pure)
75% solution
Bis(2-ethylhexyl) peroxydicarbonate
Di-n-propyl peroxydicarbonate
Diisopropyl peroxydicarbonate
-Cumyl peroxyneodecanoate, 75%
Azobisisobutyronitrile
Lauroyl peroxide

80
65
75
34
20
30
59
>70
123

40
0
32
0
-10
0
0
60
80

with inert solvents or making a suspension of the initiator in water decreases the danger of
handling these materials, while storing in a reduced oxygen atmosphere lessens the danger of
fire.
The most commonly used storage method is in commercial top-opening freezers which have
been adapted to reduce the ignition hazard from the light and the thermostat. These are usually
arranged in an open shed with fire-resistant partitions between each unit and remote internal
temperature alarms.
General safety rules for the use of initiators include:
1. Mark each type of initiator with a distinctive label and store only one type in each location.
2. Remove only enough for immediate use from storage and keep it at the proper temperature
on the operating floor.
3. Try to arrange the packaging so that only full shipping containers are used. Use extreme
care in dividing packages, especially those that are solid at storage temperature.
4. Use only clean, dedicated containers if an intermediate container is necessary.
5. Never return material to a shipping container.
6. Remove empty containers promptly and dispose of separately, not with other trash.
7. Use proper protective clothing.
The most critical period in initiator handling is when the proper charge for a batch has been
weighed and is placed into the charging device. Any delay in the batch, or leakage of vinyl
chloride or other materials into the charge device, can initiate a violent reaction. Cooling of the

charge device

Page 275
with a jacket or by addition of cold water can be of some assistance, but only if the SADT is
above 32F.
Decomposition is catalyzed by any organic reducing substance and many metal ions such as
iron. Thus dust, rust, concrete, and many such common contaminants have caused problems.
Leaking containers may be the single most prevalent problem, and care should be taken to see
that all containers are stored upright and that the caps are tight. Any spilled material should be
absorbed in an inert solid such as vermiculite and destroyed immediately.
Manufacturers provide proper disposal directions for each product [287], but in the absence of
specific instructions the material can be added cautiously to glowing charcoal embers in a ditch
or hole at a safe distance from any flammable material.
B.Monomer Production
In addition to the general hazards of handling, storing, and shipping large quantities of VC, the
production plant must deal with a potential explosive mixture in the oxychlorination reactor. Flow
control failure for any of the three major streams-ethylene, hydrogen chloride, or air (oxygen)can produce a reaction mixture in the explosive range [287a]. In addition to being a necessary
reactant, the hydrogen chloride serves to dilute the oxygen content of the mixture below the
explosive range. Therefore, considerable effort must be made to assure that flammable
concentrations cannot occur.
Another hazard is the danger of overheating the reaction vessel by loss of the cooling system
flow or excessive rates of reaction. This is particularly true when fresh, active catalyst is in use,
and is a greater problem for fixed-tube reactors than for fluid beds.
C.Polymerization
The major safety and health hazards in the polymerization section have been from the danger of
uncontrolled reactions and the exposure to vinyl chloride, especially from reactor cleaning.
Methods to control these hazards and others are discussed in this section. An EPA report is
available which presents the results of a brief industry survey by a contractor concerning some
of the causes of emergency releases [248], one of the contributors to safety problems in this
area.
1. Reactor Control: The heat of polymerization of vinyl chloride is about 660 Btu/lb (Table 1),
and this heat is not released uniformly during the batch unless special initiator blends are used.
The heat release tends to be less at the beginning of the cycle, and accelerates until just before
the pressure drop begins in the case of homopolymers [282]. There is a tendency to design the
reactor batch charge so that there is a small but adequate margin of control left at the time of
the peak exotherm. However, because all of the batch is charged at the beginning, there is no
easy method for making adjustments during the cycle for unforeseen conditions.
The standard method of temperature control is to add cooled water to the jacket circulating
system. The heat transfer values of a clean system (Q value) range from 50 to 110, with the

lower figure obtained with glassed steel, and the higher with polished stainless steel. The value
decreases during the cycle as the viscosity of the batch increases [234`, and also can be
decreased drastically by fouling on either the water or process sides. If that should happen
unexpectedly during a batch, or if the cooling water supply or the agitation is lost, a runaway
batch can result.

Page 276
Control of a runaway batch depends on the successful removal of the heat of polymerization.
This can be done by several means, such as additional cooling, addition of a chemical to stop
the polymerization, removal of the unreacted monomer, or in the last resort, relief of the
pressure by manual venting or by activation of the pressure relief devices.
Most facilities use mechanical pressure relief valves on reactors and VC storage and handling
vessels, preceded by a rupture disk, as is required by the EPA standard. A few facilities,
however, use only rupture disks, usually two in series, as the protective device on reactors.
Mass polymerization systems must use rupture disks rather than relief valves.
There are no really satisfactory means of calculating the required venting area of a polymerizing
system. Theoretical calculations have been made [288290], but the situation is complicated
severely by the mixed flow regime which occurs, the plugging tendency of the spongy, partially
polymerized mass, and the fact that the rate of increase of the reaction with temperature is a
function of the slope of the initiator half-life. Neither the amount of initiator left at a particular
point nor the slope of the decomposition rate at very high temperatures is known with any
degree of certainty.
This has resulted in an empirical approach to relief valve sizing. Nozzles on which the relief
valves are placed range from 4 in. on the smaller vessels up to about 4000 gal to as much as
16 to 18 in. on 25,000-gal reactors. Relief valve sizes range from a single 4 x 6 in. valve on the
smaller reactors to multiple 8 in. x 12 in. valves on larger sizes.
This appears to have been a satisfactory approach, for there are no known reports of reactors
having exploded because of relief capacity design. One case is known where a reactor
exploded after an enormous overcharge of initiator, perhaps as much as 30 times normal, was
pumped in as the result of an instrument failure, but this can hardly be ascribed to the relief
valve design. Also, there are anecdotal reports of one glass-lined reactor having been stretched
sufficiently to spall off much of the lining, but not failing. The size of the relief system for that
reactor is not reported.
The American Institute of Chemical Engineers has formed a Design Institute for Emergency
Relief Systems which has been working for several years to improve the design basis for relief
systems. The results of this effort are expected to be released in 1985 [290a].
Installation of a rupture disk below a relief valve requires reduction of the valve capacity to 80%
of its original rating, unless that specific combination has been tested for capacity. There must
be a pressure gauge, vent, or other suitable telltale device between the valve and disk to
indicate if the disk has leaked pressure into the volume under the valve [290b]. Many other
factors must also be considered in rupture disk design, and installation, and final choices of
materials and type should be made only after consultation with suppliers [290c].
Both redundancy and diversity in control instruments are used to assure proper information on
the internal temperature of the reactor. Dual temperature probes plus at least one pressure

check are the most common forms. Reflux condensers can be used on some types of
processes. These can offer a significant reserve cooling capacity, and often are connected to
the emergency cooling circuit that is activated by high-temperature or high-pressure alarms on
the reactor.
One company has described an emergency monomer removal-cooling system which connects
the reactor to a large external condenser and tank. In the case of an emergency the monomer
is then condensed outside the reactor,

Page 277
both cooling the reactor and removing the source of any further heat. This has been used
successfully on reactors up to 50,000-gal capacity [290d].
Agitation failure usually results in a relief valve discharge because of the reduced heat transfer
in an unagitated vessel. Current flow or torque monitors can be used to confirm proper agitator
action, but greater assurance results from detectors which measure the rotation of the agitator
shaft itself. Routine vibration measurements and visual inspection of the drive motor, gear box,
coupling, and seal are important preventative maintenance items.
A polymerization inhibitor (short-stop) is effective in controlling or slowing overheated batches if
enough is added before the reactor is truly out of control and there is adequate agitation to
assure good mixing. Organic monomers which do not copolymerize well with vinyl chloride, such
as butadiene, styrene, or a-methyl styrene, have been used, as well as straight inhibitors such
as phenol, t-butyl catechol, or similar substances. Sodium nitrite is effective if the pH is low
enough to assure formation of nitrogen oxides from decomposition of the nitrous acid. This
requires a pH of about 5 or below. Nitrogen or steam sparging through the bottom valve can
sometimes supply the needed mixing action if the agitator is not operating and if sufficient
monomer venting rates cannot be achieved.
Short-stop addition systems should have a means of injection that is independent of the plant
utility systems. A dedicated nitrogen pressure system or manual charge pot systems can be
used. All nozzles on the reactor head should be inspected frequently to be sure that they are
not plugged with polymer. This is especially true, of course, for those leading to the emergency
relief and short-stop systems. Where there is room in the reactor, such as near the end of a
batch, the injection of cold water can sometimes provide sufficient time for other systems to
overcome an incipient runaway reaction.
The EPA has accepted manual venting to the atmosphere as a last resort step to avoid the
usually irreversible action of the relief valve system [40CFR61.64(a)(3)]. Venting to the
recovery system, flare, or a gasholder may be helpful in marginal cases, but great care should
be taken that foam and polymer are not carried over and plug the system.
It is not feasible to build a gasholder large enough to hold all the vapor from a large reactor or
more than one small reactor. Limitations on the rate of movement of the piston control the
speed with which vapor can be added. Thus gasholders are not to be considered emergency
devices except to a limited extent. Improper attempts to control or delay emergency releases
may cause more total release than if not used, if there is a general plantwide emergency.
Relief valve assemblies should be anchored firmly to resist the thrust that develops on
activation. Tail pipes should be short, and if curved upward for better dispersal, a weep hole or
easily removable plastic cap should be used to prevent ice blockage in the winter. An effective
rain shield can be made from a short length of larger size pipe supported by standoffs on the
tailpipe. Insurance provisions and local codes establish relief valve inspection and test
schedules, which should be no longer than 1 year, and after every activation.

Instrument failure during charging can result in overfilling of the reactor (or other vessels) and
cause hydrostatic pressure on the rupture disk as the temperature increases. Level control
instruments inside the reactor have not proven effective because of the fouling problems, and
external devices such as radiation meters are not sufficiently sensitive or reliable. Dili-

Page 278
Year Place

Cause

Result

Broken gauge glass on a storage tank, ignited

Major plant damage.
by nearby boiler.
Attempts to tighten a reactor sight glass while
Seven killed, 22 injured.
1964 Connecticut
under pressure failed. Ignited by nearby
Plant destroyed.
extruder operation.
Operator opened wrong reactor bottom valve,
One killed. Plant
1966 New Jersey discharging contents when handle failed. Ignited
destroyed.
by static or other source.
1967 Louisiana
Pump failure. Ignition course unknown.
$830,000 damage.
Extensive reactor
1968 Rhode Island Manway gasket failed. Ignited by static?
damage from falling roof
members.
Four killed, 8 injured in
1970?Japan
Discharged contents of wrong reactor.
plant, 2 outside. Major
structural damage.
One killed. Instrument
Head gasket failed. Ignited by nearby gas-fired
1970 Delaware
and control systems
drier.
destroyed by fire.
Five killed outside
Thermowell failure in bulk reactor. Ignited in the
1973 Germany
structure by flame front.
recovery section.
Plant badly damaged.
1973 Japan
Broken valve yoke.
Unknown.
Several weeks'
Manway not secured properly; ignition from
1974 New Jersey
production lost from
static discharge.
wiring damage.
Initiator overcharge to bulk reactor due to
One killed, 1 injured.
1977 Texas
instrument failure.
Plant destroyed.
1955 Massachusetts

Page 279
Year Place

Cause

Result

Workman serviced valve on storage tank

One killed, 3 injured,
1977Mexico
improperly, discharged contents, ignited at
major damage.
adjacent plant.
Buildup of peroxides in recovery system exploded Major equipment
1978Germany
during steam purging.
damage.
Operator opened wrong hottom valve, discharged
Two injured, damage
1980Massachusettsfresh batch. Vapor cloud ignited above ground
over $1 million.
level.
Use of improper valve type allowed a bottom
1980California
valve to remain partially open. Ignition at nearby Major damage.
switch box.
Five injured. Destroyed
Buildup of vapor in a sewer line entered
1981Canada
laboratory and control
laboratory building.
room.
gent instrument maintenance, redundant metering, and visual inspection are the most reliable
means of prevention for overfilling.
Premature rupture disk failure has been a problem in compliance with EPA rules [248]. Fatigue
from pressure-vacuum cycling during the batch or from vibration or swaying of the vent system,
mechanical damage during installation, corrosion, and distortion by polymer formation are
among the most frequent causes of this failure. Careful installation, frequent inspection, and
routine replacement are necessary.
Insurance and corporate safety codes usually forbid manifolding of reactor relief valve
discharge systems. On the few occasions where this has been tried, it was abandoned quickly
because of the nearly instantaneous plugging of the system. The greatest safety in the long run
is obtained by rapid dispersion of any vapors released by a discharge. Other relief valves not in
polymer service may, however, be manifolded to flares or other abatement devices [248].
A survey of the major accidents involving the production and polymerization of VC is presented
in Table 8. This table does not include transportation accidents, for none of those have been
reported to have caused any loss of life or major property damage beyond the accident scene.
Also not included are VC releases which did not result in serious injury or major damage. Two
recent events of major proportion which did not cause loss of life are described in references
290h and 290i.

Page 280
2. Worker Exposure: Plant design, process procedures, work practices, surveillance
equipment, and personal protective devices all play a part in reducing worker exposures to
potentially harmful concentrations of vapors. These methods are all interwoven with the efforts
to protect workers and equipment from fire and explosion.
The production and polymerization of vinyl chloride are, of course, closed processes. Outdoortype construction is used to the greatest extent possible, but the severe winters of 1979, 1980,
and 1983 illustrated the limits to this feature, with even facilities in the Gulf region suffering
damage and production interruptions. Area and local ventilation are used for those parts of the
plant that must be enclosed.
Reactor cleaning operations were associated with most of the AOL and ASL cases, and thus
this procedure has been limited sharply. Additional impetus came from the introduction of large
polymerization vessels, which are not practical to clean by hand. Antifouling treatments [291,
292], solvent cleaning procedures [293], and improved suspension recipes have allowed closed
reactor operation for many polymerization cycles. High-pressure water cleaning is a useful
supplementary tool, and massive buildups can be loosened by dynamiting rather than by hand
cutting as in the past.
The development of effective spray rinse valves has assisted in reducing worker exposure by
reducing the frequency of reactor opening for inspection and cleaning [293a]. These valves can
be used to improve the efficiency of application of antifouling agents or rinsing solutions, and for
better distribution of short-stop solutions.
When vessel entry is necessary, careful adherence to detailed vessel entry and lockout
procedures, the use of mechanical or human standby systems, the wearing of a proper safety
harness, forced ventilation of the vessel, and proper monitoring of vapors and oxygen
concentrations can help ensure the safety of the worker. The OSHA regulation for VC requires
the use of respirators and protective clothing and prescribes the type to be worn under various
circumstances such as monomer loading/unloading or mechanical repairs, as well as vessel
entry.
Extensive stripping of the unreacted monomer from the polymer slurry has made a major
contribution to reduced worker exposure. Emissions from the slurry vessels, centrifuge
raffinate, and dryer outlets are controlled to less than 10 ppm by the EPA standard, either by
direct emission controls or by stripping of the slurry before transfer to these systems. This is an
example where an expenditure for EPA rules has assisted in meeting the OSHA standard.
Another example is the requirement for pressurized double mechanical seals on pumps,
compressors, and agitators.
The OSHA requirement for an area monitoring system to warn operators of concentrations
requiring the use of respirators was adopted by the EPA as a leak detection device. Overall
correlation between area concentrations and personnel exposure can be shown if adequate
attention is put on a time-motion study, but short-term conformance is poor [234, 293b]. It has

been observed that mechanics and senior operators/foremen sometimes experience higher
exposures than those performing routine tasks because they are more often at the site of
unusual occurrences.
Personal monitoring using devices such as portable pumps with either gas collection bags or
absorption tubes, or the more recent passive monitoring badges, is required by OSHA on a
periodic basis. These devices were valuable in the regulatory learning process in assisting
workers in determining which of their work habits were more likely to result in exposure,

Page 281
and in convincing them that the area monitoring reports were accurate, even at those low
levels.
The efforts to reduce worker and environmental exposure have been successful. Workplace
concentration had been reduced to below 5 ppm by mid-1975 [294], and by 1976 to less than 1
ppm. The general consensus among PVC producers is that compliance with the OSHA
standard is greater than 95%, with the exceptions being due to malfunctions or process upsets.
Vinyl chloride producers usually operate as deregulated areas except during maintenance
turnarounds.
In the long term, however, the conscientious worker makes the greatest contribution to overall
safety, including reduction of exposure. The employer's training program must be designed to
motivate workers to safe work practices by helping them understand the need for such
precautions, and by teaching safe habits and sound work practices. Firm enforcement of safety
rules and ongoing retraining are fundamental to a successful program.
3. Other Considerations: The emphasis on reduction of worker and environmental exposure has
decreased the probability of major accidents from the release of flammable vapors, but that
possibility should not be forgotten. Electrical construction should be grounded properly and in
conformance with NFPA Class I Group D specifications [277]. Structural steel and tank and
reactor supports should have adequate fire protection. Major valves should be fire-safe and failsafe on loss of power. Damage to instrument and power lines often is a major cause of
production loss from fires, and adequate protection should be provided for those services.
Walk surfaces should be of a type which provide sure footing when wet or covered with PVC
powder. There should be adequate access platforms for elevated areas such as around relief
valves, where prompt and frequent service may be required.
Many employers have utilized the action-level concept to establish a policy for casual visitors
and others who are not authorized employees under the OSHA rule. As an example of its
application, visitors are allowed to be inside regulated areas for no longer than 45 min per day
as long as the area monitors indicate that the ambient levels are below 5.0 ppm, or 3.5 hr at 1
ppm. It is impossible to accumulate more than 4 ppm-hr under this policy, and thus the OSHA
program is not invoked.
Some employers have taken conservative positions regarding the potential teratogenicity or
transplacental carcinogenicity of VC, and have established rules that do not permit women with
childbearing capacity to work in regulated areas. At least one such policy has withstood legal
challenge [294a].
Sight glasses are useful devices on reactors and other vessels. They allow monitoring of the
contents to assure that reactors actually are emptied between batches, and serve as a check
on the charging system to prevent overfilling, or excessive foaming during venting or evacuation.
They also allow some degree of evaluation of the degree of fouling on the walls and thus help

prevent the loss of temperature control of the batch.

After the 1964 explosion and fire in Connecticut, manufacturers reevaluated their use of sight
glasses. Many decided to stop their use of them on reactors, and others changed to a different
type. The glasses in most general use before 1964 were simple disks of thick glass, sealed in
place by gaskets above and below, and tightened by a conventional flange. The type in most
common use today consists of a laminated disk sealed by compression on the circumference
from pressure by a lantern ring or packing. Even if cracked,

Page 282
these will not fail catastrophically, and major releases will not occur. It still remains good
practice, however, not to attempt to adjust these devices under operating conditions.
Care should be taken to install an air gap between the sight glass mount and any permanent
light source so that excessive heat buildup will not occur. See the discussion on chlorideinduced stress corrosion in the following section. Localized high temperatures can accelerate
the decomposition of any polymer deposits, and thus the corrosion rates also.
Similarly, gauge glasses are useful manual checks on level devices for storage or process
tanks. These are subject to fouling, either internally at the connection ports or from sunlight
through the glass. Adequate attention to inhibitor level and an aggressive preventative
maintenance program can reduce these problems. A special glass is now available that
transmits fewer ultraviolet rays and thus reduces the glass coating by polymer. Magnetic
indicator devices are available that eliminate the glass sections altogether. Any glass units that
are used should have excess flow systems in the isolation valves.
A major reason for the trend toward more computer-controlled large reactor operations is the
increased safety that comes from fewer units, and thus fewer connections and fewer systems
to control. This improvement has been realized, but brings with it some hazards of its own. It
has added another level of interface, and it requires a higher level of technical sophistication for
maintenance. In addition, if total reliance is placed on electronic systems, manual recovery from
disaster conditions is lost. It is normal to install sufficient analog systems to allow at least an
orderly shutdown in case of necessity. Tunkel [322] has reviewed the design criteria to be
considered in protecting vital equipment from blast damage.
Computers with dual, automatic switchover processing units have demonstrated on-line service
of well over 99%. However, operators have reported rare, unexplainable gremlins that either
cause loss of control or issue random uncontrollable signals. Thus special attention should be
paid to the computer installation and its maintenance program [323].
Careful attention should be paid to the failure mode of critical valves not only for loss of
operating power, but also for loss of operating signal. In addition to emergency analog control,
the most critical valves should have an independent hardwired signal to the control room.
D.Stripping
Prior to 1974 unreacted monomer was recovered from the PVC batch on an optimized
economical basis. The reaction mass was transferred from the reactor to a blowdown tank, or
the pressure was reduced on the reactor by venting, at the point where conversion rates no
longer justified utilization of reactor time, and further conversion was likely to reduce desirable
properties of the resin such as porosity [282, 294b]. The slurry was then subjected to a short
vacuum exposure (30 min to 1 hr) at temperatures of 160 to 195F, depending on the residual
heat stability of the product. This procedure left upward of 2% by weight of vinyl chloride
dissolved in the resin. Much of this was lost during subsequent transfer and drying operations,
but the finished product usually contained 1000 to 2000 ppm of VC at the time of shipment

[294c].
The EPA standard set a limit of 400 ppm VC in suspension resin (2000 ppm for emulsion
products) at the time the slurry was released from a closed

Page 283
system as an alternative to placing abatement controls on the VC in the dryer discharge air
stream. The time to achieve this level by the conventional means is excessive, and puts the
resin through a harmful heat history, resulting in yellowing.
Most producers of suspension resins have adopted some variation of a continuous stripping
system in which the slurry is passed down a tray tower against a countercurrent stream of
steam. Short residence time at elevated temperature and good agitation results in lower final
VC content and less heat stress than did the older method [294d].
General-purpose resins of average porosity generally exit the column at a few ppm residual
VC, which is reduced to well under 5 ppm by time of shipment. Low-molecular-weight resins,
which generally have lower porosity, give somewhat higher figures, and very low molecular
weight resins (bottle grade) and copolymers have difficulty meeting this standard on a 100%
basis, although the long-term average is well within the requirement [248].
Emulsion resins are more difficult to strip because of their strong foaming tendency and
sensitivity to coagulation with heat. A variety of falling film or spray devices have been
developed which allow meeting the 2000 ppm standard [294e]. These and other less useful
devices are discussed by Burgess [294f].
The theory of VC migration in PVC has been developed by several workers. The monomer is
quite soluble in the polymer, although the contrary is not true. The final solubility depends on the
pressure and the temperature, and the rate of equilibration is a function of temperature, particle
size, and morphology [294g]. The rate is diffusion controlled and can be described by classical
thermodynamic equations [294c, h]. Diffusion rate is more important than temperature at higher
concentrations, but temperature is the controlling parameter for the final interphase partition
coefficient [294i]. There is a discontinuity in the controlling constants at the glass transition
temperature (Tg) of the resin [294c, j], and above this temperature the rate of diffusion
increases sharply.
In practical terms, these basic data show that stripping consists of movement of the monomer
molecule through the body of the resin and across the solid-liquid or solid-gas interface, through
the pores of the resin into the larger body of the suspending water, through the water to the
liquid-gas interface, and eventually out of the vessel. Any condition that can shorten or speed
this movement assists in the stripping rate. Smaller particle size, greater porosity, absence of a
pericellular membrane on the resin, good agitation, and temperatures above Tg all assist the
progress of the monomer. Lack of porosity and especially the presence of glassy beads or gels
hinder the rate greatly, and even a small amount of such particles can prevent proper stripping.
Low-molecular-weight resins tend to be less porous, and this offsets any advantage of the
greater mobility that might be expected within shorter molecules.
Application of these concepts has resulted in successful equipment design for the stripping
step, and allows prediction of the migration of monomer under many conditions [294k, 1]. The
latter reference contains a procedure for estimating the potential exposure to workers in

warehouses and other storage areas, for example.

Even further reduction in residual monomer can be achieved by applying these principles in the
processing and compounding steps also [294m], and levels can be obtained that approach the
limits of detection. It has been proposed on both theoretical and experimental bases that a
residual level can be reached beyond which no further diffusion will occur [294n, o].

Page 284
Each mechanical stage of air transfer, unloading, or processing represents a disturbance of the
established equilibrium between the dissolved VC in the polymer and its surroundings, and thus
will cause some release of VC, and the beginning of a faster rate of release of the monomer
until equilibrium is reestablished. Therefore, caution should be exercised in opening and entering
railcars, storage silos, or other closed storage areas unless it has been established that the
free space is below the allowable concentration. Even moderate ventilation will assure that this
has been achieved.
E.Downstream Operations
The PVC process changes from a pressurized batch operation to a generally open, continuous
system after the blowdown/stripping step, and the safety hazards change to those related to
material handling procedures. The exception to this generalization is in the monomer recovery
section, which is of necessity closed, and is pressurized after the compressors. Partial vacuum
can occur upstream of the compressors in the blowdown/stripping section, and provides the
opportunity for air to enter the system. The oxygen can react with vinyl chloride and other
olefins at the temperatures encountered in the compressor to produce peroxides, as discussed
above. These peroxides can cause extensive fouling and plugging of the recovery system and
are shock sensitive. Dilute caustic can be used to control the formation or to remove these
products. However, careful maintenance to prevent leaks in vacuum lines, together with proper
inhibitors and pH control, can prevent their formation.
Entry and cleaning of blowdown tanks, stripper vessels or towers, and recovered monomer
tanks present the same potential hazards as does reactor cleaning and the same precautions
should be observed. There are anecdotal reports of workers being asphyxiated by the inert
atmospheres that may be maintained in these vessels. Thorough work entry procedures,
including oxygen monitoring, and the use of color-coded air and nitrogen hoses with
noninterchangeable fittings can reduce these dangers.
The more thorough stripping required by the EPA standard and by commercial considerations
has reduced greatly the exposure to VC that occurred in the past in this section of the plant.
Many of the manufacturers have found that the drying and shipping areas can be deregulated
under the OSHA standard. Fabricators and processors generally find their operations are below
the action level for the OSHA standard if they have even rudimentary ventilation systems.
The half-life of residual VC in bagged resin is about 1 week, so there is some possibility of
exposure in large unventilated warehouses for bagged resin. Bulk storage presents a higher
probability of monomer accumulation in the air space, so that silos and bulk cars should be
ventilated and tested before entry.
This extensive stripping has caused an additional hazard, however. The higher operating
temperature at longer times has accelerated the formation of hydrogen chloride in the slurry,
which has resulted in accelerated chloride-induced stress corrosion near the welds in stainless
steel equipment. This appears to be caused by chromium depletion in areas near carbide
precipitates, and is especially pronounced at pH below 5 [294p]. Cracking has been seen in

reactors, especially those also used for stripping, and particularly near nozzles which may have
some polymer deposition, in blow-down tanks, in continuous stripping towers on the trays and
support rings, and occasionally on centrifuge scrolls.

Page 285
Routine removal of polymer deposits, additional buffering of the treated slurry to avoid the
autoacceleration of decomposition by acid, and avoidance of excessive localized temperatures
can help alleviate this situation. Clad vessels will be less prone to catastrophic disintegration
than will solid stainless vessels. The use of proper grades of construction materials, such as
low-carbon 316 stainless steel or high-nickel alloys [295], and careful adherence to good
welding practices also are of assistance. Dye tests of suspect areas can sometimes reveal the
problem before it becomes a serious threat to safety.
Dust exposure is a common problem in the bagging and shipping areas, more so with the fine
emulsion and dispersion resins than with the coarser suspension resins [296]. Also, the problem
exists more with the total dust levels than with the respirable fraction. Operators are reluctant
to wear respirators or masks in these work areas because of the greater exertion necessary
and the higher ambient temperatures often encountered, so careful equipment design and good
local ventilation are required.
Slips and falls from bulk cars and trucks can be a hazard, especially in wet or icy weather.
Safety harnesses have been designed for this purpose, and access platforms can be provided
to avoid much of the climbing that would otherwise be necessary.
Strong static charges can develop during the air conveying of PVC, especially during dry, cool
weather. Static discharge in a conveying system is suspected as being the source of ignition for
at least one major explosion following a vinyl chloride release. Various methods have been tried
to prevent or dissipate this static buildup, but the most effective seems to be controlled
humidification of the conveying air.

V.WASTE STREAMS
A.Water
The EPA standard requires that water streams which have been in contact with vinyl chloride be
stripped to below 10 ppm before they are released or mixed with other streams. This is
because vinyl chloride degasses readily from water at atmospheric pressure and would
therefore become an air contaminant. It has been observed in the past that the centrifuge
waste streams provided a significant contribution to worker exposure if they were transported
in open trenches. This is no longer true for slurries which are stripped in accordance with EPA
rules.
All the substances discussed here undergo biodegradation or removal in biological effluent
treatment systems [262, 297301]. Somewhat surprisingly, one EPA report states that EDC is
said to be removed more readily in some cases by air stripping than by biodegradation in
comparison to the other substances [301a].
Another EPA report [301b] states that EDC is relatively difficult to air strip, and that
performance can be predicted by the Henry's law constant of the substances. The values
stated are 180 for VC, 0.5 for TCE, and less than 0.1 for EDC. The data of Dilling and of

Roberts and Dandliker [269] suggest that all these substances should air strip easily. However,
all these substances showed ready biodegradation in activated sludges, as they did in 24-hr
batch tests [301c].
Air stripping and steam stripping are reported to be economically feasible alternatives to
biodegradation. The EPA has calculated that six theoretical trays are required to reduce EDC to
50 ppb from a saturated feed when

Page 286
using reflux, while only four are needed for TCE or VC [301a]. Without reflux eight trays gave
the same results for EDC and required 5 g steam per kilogram of feed. Under the same
conditions, TCE needed seven trays and 3 g/kg of steam, and VC only six trays and 2 g/kg, to
obtain 100% efficiency. This same document gave examples of activated carbon removal
efficiencies of 99% for TCE from very dilute streams, but did not show examples of the
application of this technology to the other substances. The Dade County water system found air
stripping to be effective in removing VC from contaminated groundwater [274g].
A recent paper by Zhu et al. [301d], describes a prevaporation technique for removal of EDC
and other chlorinated hydrocarbons from dilute aqueous solution by permeation through a
polymeric membrane against a vacuum. The organic solutes permeate preferentially and may
be collected in a concentrated form.
Some success has been seen in reuse of wastewater, especially as makeup to cooling towers,
provided that adequate filtration of residual solids is performed. There has been limited success
in reuse of concentrate in the polymerization batch because of potential cross-contamination by
residual suspending agents. It has been shown to be feasible in pilot runs, however, and could
be possible in a single-product plant. Very thorough filtration is necessary, and a blowdown
stream may be required to purge dissolved salts [234, 298].
B.Liquids and Gases
There are few nonaqueous liquid streams from polymer production, but still bottoms and byproducts are formed in monomer production. These can be incinerated, provided that the
chlorine formation is minimized by careful combustion air control and the hydrogen chloride is
removed from the stack gas. Processes have been developed for redistillation or catalytic
decomposition of these substances which permit substantial recovery of the chlorine values
[302304]. The EPA has listed the liquid wastes from the production of VC and EDC as
hazardous wastes, which will restrict the disposal of these by-products to RCRA-permitted
facilities.
The EPA also classifies spent TCE from degreasing operations, and bottoms from TCE
manufacture, as hazardous wastes (40CFR261.31 and .32). The commercial substances vinyl
chloride, EDC, and TCE are themselves hazardous wastes (40CFR261.33) and may not be
discarded without following RCRA regulations.
Incineration has become the process of choice for abatement of the collected vent and purge
gases from polymerization operations. Carbon absorption processes have been developed
which recapture the vinyl chloride, but various operating problems have prevented widespread
adoption of this process. One problem with copolymer operations is the difficulty experienced
with desorbing the vinyl acetate. In addition, there is a need for a purge stream to remove the
nonreactive methyl chloride from the recycle stream [304a]. This occurs in monomer at 25 to 75
ppm, but as it is concentrated about 10-fold each cycle, it can soon build up to an unacceptable
level as an inert diluent. It is necessary to have a small incinerator to destroy the purge stream
containing this contaminant, and economics often do not justify any additional equipment.

Incinerators must be equipped with scrubbers to meet local and federal limits on hydrogen
chloride and particulates (47FR27520). Incinerator design

Page 287
is based on the requirement for less than 10 ppm in the stack gas; actual performance is much
better, with concentrations usually below 1 ppm. Vinyl chloride is readily combustible, and the
calculated combustion chamber temperature for 99.99% destruction at 1 sec residence time is
1371F [305]. Vinyl acetate requires a temperature of 1223F under the same conditions.
Studies have shown [305a, b] that flares and industrial boilers can give greater than 99%
destruction of organic wastes, but the EPA will not permit these to be used for routine removal
of hazardous air pollutants without further demonstration of their efficiency.
C.Solids
Polyvinyl chloride is biologically inert and is suitable for disposal in any properly maintained
landfill. There was a statement by the EPA (45FR33118, May 19, 1980) that it intended to
promulgate RCRA rules for batch residues from the batch polymerization of chlorinated
polymers, but no further action has been taken. Presumably, this was to have been done
because of reports that vinyl chloride had been detected in very low concentration around some
pre-1975 landfills. Current operating procedures preclude the probability that any significant
amount of vinyl chloride will find its way to a landfill.
Water from the centrifuge in suspension processes contains a small amount of fine polymer.
This settles rapidly in clarifiers or sedimentation ponds. In fact, it is a good substrate on which
other suspended materials gather. Together with the larger particles resulting from the cleanup
of spills and from washdown operations, this material, or sludges in which it may be present,
may be disposed of in any convenient fashion.
Solids recovered from solvent cleaning operations are free of vinyl chloride and thus also are
not restricted for disposal. The only problem concerns large unstripped particles, such as those
from BB batches, filters ahead of the stripper, and equipment cleaning. These are not now
regulated by RCRA, but prudence would dictate that the residual monomer content should be
reduced to prevent either employee or environmental exposure. This can best be done by
weathering in some isolated locale.
Solid PVC wastes can be incinerated if mixed with an adequate quantity of a combustible
material. There is little reason to do this, except for the disposal of used consumer items. The
generation of hydrogen chloride from household trash containing PVC articles has been a
matter of controversy in the past, but several studies have shown that the present consumption
rates of PVC in consumer goods add little to the normal chloride content of wastes [306307].

VI.ANALYTICAL METHODS
The EPA prescribes the analytical methods 106 and 107 of 40CFR61, Appendix B, for the
analysis of gases and of water or solids, respectively. Method 106 uses a 2-m Chromosorb
102 chromatographic column followed by a flame ionizing detector to analyze an integrated gas
bag sample. A secondary column of Chromosorb B is required if acetaldehyde is present. The
method is said to have an absolute sensitivity of 1 to 4 x 10-7 mg of vinyl chloride. Collaborative
tests indicate that the repeatability is about 1 ppm at 10 ppm concentrations and 10 ppm at

50 ppm [308].

Page 288
Method 107 uses the headspace method [309], in which an equilibrium is established in the free
space above the sample in a vial, and an aliquot is injected onto a 2-m column of Carbowax
1500 on Carbopak A. Poropak Q is used if methanol or acetaldehyde is present. This is stated
to have the same absolute sensitivity as method 106. Equipment is available which performs
the entire analysis automatically once the sample vials are filled [309a]. Revisions to test
methods 106 and 107 were published at 47FR39168, September 7, 1982, and 47FR39485,
September 8, 1982, which permit alternate columns to be used, and impose certain quality
assurance requirements.
OSHA requires that an analytical method be used for personal monitoring that has a 95%
confidence level of 35% at 1 ppm of VC. A procedure generally based on NIOSH Methods
127 and 178 has come into broad use for this application. Air is drawn over granular carbon at
a known rate for a known time. The carbon is extracted with carbon disulfide and an aliquot is
analyzed by gas chromatography [310]. A variation on this procedure involves the use of other
commercial absorbents (Tenax or Spherocarb) and/or desorption by heat rather than by a
solvent [311]. These procedures have been developed so that under ideal conditions they are
capable of detecting as little as 0.2 ppb [312] but under average field conditions are reliable at
about 10 ppb (41FR46560).
Instrumental procedures such as infrared or ultraviolet absorption, or decomposition of the vinyl
chloride followed by measurement of water conductance caused by those products, have been
used in the past [276]. These are limited generally to the ppm range and are not as versatile or
portable as the carbon tube or bag collection methods.
The absorption/desorption method also is applicable to a wide range of substances and can be
used to determine the concentration of many different substances from the same sample.
Unless the constituents are well known, it is necessary to use a combined gas
chromatograph/mass spectrometer to identify the peaks with certainty. This combination has
been used to measure the ambient concentrations of vinyl chloride, trichloroethylene, ethylene
dichloride, and many other substances in several areas of the country [260261]. If this method
is used for vinyl acetate analyses, special precautions must be taken to avoid hydrolysis or
polymerization of the absorbed acetate. Kimble [313] has described a procedure which
answers these requirements.
The EPA has developed purge-and-trap methods for determining trace constituents in water
which are useful to 0.2 g/liter and claims a detection limit of 0.01 g/liter. The same
problems of interference exist in this system or any chromatographic method [313a].
Several helpful manuals have been published describing practical application of the analytical
procedure described above [314317]. Reference 317 contains an extended discussion of
interferences and alternate column packing.
Bromination of VC in water samples, followed by extraction into hexane and analysis with an
electron capture chromatograph, is reported to have a limit of detection of 0.3 g/liter, or 0.3

ppb. This method has been used to analyze surface and drinking water supplies [317a].
Portable vapor detectors have come into general use as leak detectors. They are used in
conjunction with the fixed-point or area detectors for vinyl chloride, which are required by the
OSHA standard, to locate the sources of excursions, and they are used for patrolling areas
outside the area detector coverage. The EPA has promulgated a requirement that all new
volatile

Page 289
organics processing facilities use these as a part of an emission control program (40CFR60,
Appendix A, Method 21), thus extending their application to all the substances discussed in this
chapter. These may have short chromatographic columns attached to provide some selectivity
for various materials, but most often they are used in the nonselective mode as simple
combustible vapor detectors. They have a wide range of sensitivity for different substances and
usually come calibrated for methane. The EPA has published tables of response factors for
other compounds [318]. Numerous problems have been described in the actual field use of
these instruments [319] and they are not suitable for precise work, but they are a useful adjunct
to a leak detection and preventative maintenance program.
One additional type of device has come into limited use in Great Britain as a fence-line monitor
[320]. Based on the reaction of vinyl chloride with potassium permanganate-impregnated paper,
it is not sensitive to concentrations much below 1 ppm and is subject to a number of
interferences. Similar limitations apply to the Drager tube procedure. For these reasons, these
devices are not particularly valuable at present conditions.
Passive personal monitors have come into wider use because of the bulk of the electric pumps,
and the cost of maintaining them in operation. These have exposed absorbant cartridges which
can be developed and analyzed by a variety of means and can sometimes be regenerated for
reuse. The accuracy of such devices is adequate to meet regulatory needs [320a] for long-term
samples, but generally do not have sufficiently rapid response times for 15-min tests.
Experience has shown that process sampling and analysis of VC can result in high personnel
exposure if adequate precautions are not taken [117]. One of the European ASL cases is
reported to have received his primary exposure as the result of sampling activity, by what must
have been a very unsatisfactory procedure. Laboratory analysts must guard against inadvertent
direct or indirect exposure, especially when conducting some of the evaporative tests on the
monomer.
Sampling systems have been developed which use essentially closed piping loops [320b]. A
double-ended sample tank is placed in a bypass system at the sample point and a flow-through
process allows purging and filling the container with a minimum of release. This procedure also
disposes of the unused sample safely. The EPA requires that all VC samples be taken with a
system equivalent to the one described.

VII.VINYL ACETATE COPOLYMERS

The manufacture of VC/VAc copolymers is more difficult than that of homo-polymers or of some
other copolymers. This increased difficulty results in additional hazards not seen in other
processes. One spectacular hazard is that of agglomeration of the reactor charge, or a setup.
This results from failure of the suspending system, and its effects are enhanced by the
plasticizihg effect of the unreacted VAc. The reacting mixture shrinks in volume (increases in
density) as the polymerization progresses, and at the same time the unpolymerized monomers
are enriched in VAc because of the relative reactivity ratios of the two monomers [321]. The

unreacted VAc swells and softens the precipitated PVC in each droplet, increasing the
tendency toward agglomeration if the suspending system is not performing correctly. Setups
occur very rarely in homopolymer systems, and usually occur in copolymer

Page 290
batches after the density of the organic phase has increased to more than 1.0. This is well into
the polymerization cycle, at about 60 to 75% conversion. Severe mechanical damage can occur
to the agitator, shaft, baffle, and drive units. Release of VC is not usual, because of the
relatively low amount of unreacted monomers at that stage.
The primary hazards result from the difficulty of removing the rubbery mass from the reactor in
the presence of VC and VAc. The most satisfactory method appears to be to cook the mass
under vacuum to remove as much monomer as possible, then reduce the mass to workable
size by many small explosive charges [234]. This has been found to be both quicker and safer
than manual removal, and results in less damage to the equipment.
The residual unreacted VAc also causes problems in the stripping step because of its
plasticizing action. Copolymer is inherently less heat stable than homopolymer, and the greater
tendency to adhere to vessel walls adds to the probability of producing burned resin. This
results in more potential worker exposure from cleaning operations.
The presence of VAc in the recovered monomer stream presents several potential hazards.
Acetaldehyde, which is formed readily by hydrolysis, accelerates the formation of VC peroxides
[234, 282] and polymeric sludges in the recovery equipment. It is usually necessary to add
more inhibitor in the recovery systems for copolymers and to maintain better control over the
pH of the system in an attempt to control both the hydrolysis and peroxidation reactions and to
reduce corrosion of the equipment. Limitation of the oxygen content of the recovery streams
also becomes more important. Careful design is necessary to avoid polymer buildup on
instruments, relief valves, and outlet lines.
Liquid and solid waste problems increase with copolymer production. There will be increased
BOD loads on the waste treatment system from the vinyl acetate and its hydrolysis products.
Solid wastes increase and are more likely to have entrapped organic materials.
The recycle streams are more corrosive than those in homopolymer systems because of the
presence of higher chloride levels and acetic acid. Type 316L stainless steel is recommended
for use here, and attention should be given to prevention of corrosion at welds.
Thus considerably more attention in design and proper operation of the reactor and recovery
system, in particular, are necessary in the manufacture of copolymers in order to offset the
additional hazards present.

GLOSSARY OF ACRONYMS
ACIGH
AOL
ASL
ASME
BATF

American Conference of Government and Industrial Hygienists, Cincinnati, OH 45211

acroosteolysis
angiosarcoma of the liver
American Society of Mechanical Engineers, New York
Bureau of Alcohol, Tax, and Firearms, a section of the Treasury Department

Center for Disease Control, a part of the Health and Human Services Department,
Atlanta, Ga.
CEFIC European Council of Chemical Manufacturers' Federations
Comprehensive Environmental Responses, Compensation, and Liability Act of 1980
CERCLA
(Superfund)
CDC

Page 291
Code of Federal Regulations, a compilation of promulgated rules; OSHA rules are in
Chapter 29, those for the EPA in Chapter 40
COC Cleveland Open Cup, one method of testing for flammability
CPSC Consumer Product Safety Commission
DNA deoxyribonucleic acid, the constituent of chromosomes
DOT Department of Transportation
EDC 1, 2-dichloroethane
EPA Environmental Protection Agency
FDA Food and Drug Administration
Federal Register, the official daily publication of the federal government; the number
FR
before the letters gives the volume, the following numbers are the page; Volume 48 was
published in 1983
GGT same as GGTP
GGTP gamma glutanyl transpepsidase, a liver enzyme
IARC International Agency for Research on Cancer, Lyon, France
IGC Indocyanine Green Clearance, a test of liver function
LD50 lethal dose for 50% death of the experimental animals within 14 days
NAS National Academy of Science
NCI National Cancer Institute
NFPA National Fire Protection Association, Quincy, Massachusetts
NIOSHNational Institute of Occupational Safety and Health
NSF National Sanitation Foundation, Ann Arbor, Michigan
NTIS National Technical Information Service, Springfield, VA 22161
NTP National Toxicology Program
OSHA Occupational Safety and Health Administration
PB
number document identification number used in ordering from NTIS
PHR Process Hazard Review
ppb parts per billion; units are per volume for gases, by weight for solids or liquids
ppm parts per million; see ppb for units
ppt
parts per trillion
PVC polyvinyl chloride, homo- or copolymer
RCRA Resource Conservation and Recovery Act
self-accelerating decomposition temperature, at which peroxygen compounds
SADT
decompose violently
TCE Trichloroethylene
TLV threshold limit value, a guide to allowable exposure set by the ACGIH
TWA Time-weighted average of exposure to substances in air, usually for 8 hr
VAc vinyl acetate
VC
vinyl chloride
CFR

REFERENCES
1. Patty, F. A., Yant, W. P., and Waite, C. F., Public Health Rep., 45, 1963 (1930).

 2. Peoples, S. A., and Leake, S. D., J. Pharmacol. Exp. Theory, 48, 284 (1933).

Page 292
3. Schaumann, O., Med. Chem., 2, 132 (1934).
4. Schaumann, O., Arch. Exp. Pathol., 181, 144 (1936).
5. Oster, R. H., Carr, C. J., Krantz, J. C., and Sauerwald, M. J., Anesthesiology, 8, 359
(1947).
6. Mastromatteo, E., Fisher, A. M., Christie, H., and Danziger, H., Am. Ind. Hyg. Assoc. J.,
21, 394 (1960).
7. Prodan, L., Ann. N.Y. Acad. Sci., 246, 154 (1975).
8. Kuebler, H., Aerosol Age, 9(14), 44 (1964).
9. Lehman, L. B., and Flury, F., Toxicology and Hygiene of Industrial Solvents, trans. by E.
King and H. F. Smyth, Jr., Williams & Wilkins, Baltimore, 1943.
10. Torkelson, T. R., Oyen, F., and Rowe, V. K., Am. Ind. Hyg. Assoc. J., 22, 354 (1961).
See also T. R. Torkelson, Statement before the Senate Subcommittee on Environment, Aug.
21, 1974. Committee on Commerce, Serial No. 93110.
11. Lester, D., Greenberg, L. A., and Adams, W. R., Am. Ind. Hyg. Assoc. J. , 24, 265
(1963).
12. Documentation of the Threshold Limit Value, Am. Conf. Gov. Ind. Hygienists, 1963.
13. Rowe, V. K., and Torkelson, T. R., Am. Ind. Hyg. Assoc. J., 38, A25 (1977).
14. Danziger, H., Can. Med. Assoc. J., 82, 828 (1968).
15. Spiritas, R., McMichael, A. J., Gamble, J., and Van Ert, M., Am. Ind. Hyg. J. , 36, 729
(1975).
16. Cole, J., Inside Story on Vinyl Chloride at B.F. Goodrich in Avon Lake, The Journal,
Lorain, Ohio, Mar. 26, 1975.
17. Klein, J., The Plastic Coffin of Charlie Arthur, Rolling Stone, Jan. 15, 1976.
17a. Lelbach, W. K., and Marsteller, H. J., in Advances in Internal Medicine and Pediatrics,
Vol. 47, Springer-Verlag, New York, 1981.
18. Union Carbide, unpublished data, 1974.
19. Hori, M., Kobayash, Y., and Ota, Y., Plast. Ind. News, 18(11), 164 (1972).
20. Dublin and Vane, 1935, cited in Lehman, K. B., and Flury, F., Toxikol. u. Hyg. d.

techniochem Losungs mittel, J. Springer, Berlin, 1938, pp. 130143.

21. Baretta, E. D., Stewart, R. D. , and Mutchler, J. E., Am. Ind. Hyg. Assoc. J. , 30, 537
(1969).
22. Schaumann, O., quoted in Ref. 9.
23. Guengerich, F. P., and Strickland, T. W., Mol. Pharmacol., 13, 993 (1977).
23a. Laib, R. J., in Reviews on Drug Metabolism and Drug Interactions, Vol. 4, No. 1, A. H.
Beckett and J. W. Gottod, eds., Freund Publishing House, London, 1982, p. 1; L. M. Gwinner,
R. J. Laib, J. G. Filser, and H. M. Bolt, Carcinogenesis, 4, 1483 (1983).
24. Hefner, R. E., Jr., Watanabe, P. G., and Gehring, P. J., Ann. N. Y. Acad. Sci. , 246, 135
(1975).
25. Gehring, P. J., Watanabe, P. G., and Young, J. D., in Origins of Human Cancer, Vol. A.
H. H. Hiatt, J. D. Watson, and J. A. Winsten, eds., Cold Spring Harbor Laboratory, Cold Spring
Harbor, N.Y., 1977, p. 187.

Page 293
26. Maltoni, C., Lefemine, G., Ciliberti, A., Cotti, G., and Carretti, D., paper presented at Le
Club de Cancerogenese Chimique, Institue Curie, Paris, Nov. 10, 1979. See also Environ.
Health Perspect., 41, 3 (1981).
26a. Bartsch, H., Malaveille, C., and Camus, A. M., in Organ and Species Specificity in
Chemical Carcinogenesis, R. Langenbach, S. Nesnow, and J. M. Rice, eds., EPA-600/983
008, June 1983, PB 83220137.
27. Equitable Environmental Health, Inc., Epidemiological Study of Vinyl Chloride Workers,
Final Report, prepared for Manufacturing Chemists Assoc., Washington, D.C., Jan. 1978.
28. Ott, M. G., Langner, R. R., and Holder, B. B., Arch. Environ. Health, 30, 333 (1975).
29. Fox, A. J., and Collier, P. F., Br. J. Ind. Med., 34, 1 (1977).
30. Chiazze, L., Jr., Nichols, W. E., and Wong, O., J. Occup. Med., 19, 623 (1977).
31. Chiazze, L., Wong, O., Nichols, W., and Ference, L., J. Occup. Med., 22(10), 677
(1980); Environ. Health Perspect., 41, 137 (1981).
32. Frentzel-Beyme, R., Schmitz, T., and Thiess, A. M., Arbeitsmed. Sozialmed.
Praeventivmed., 13, 218 (1978).
33. Buchter, A., Bolt, H. M., Filser, J., Goergens, H. W., Laib, R. J., and Bolt, W., Verh.
Dtsch. Ges. Arbeitsmed., 18, 111 (1978).
34. Gehring, P. J., Watanabe, P. G., and Park, C. N., Toxicol. Appl. Pharmacol., 49, 15
(1979).
35. Anderson, M. W., Hoel, D G., and Kaplan, N. L., Toxicol. Appl. Pharmacol., 55, 154
(1980).
36. Bolt, H. M., Filser, J. G., and Buchter, A., Arch. Toxicol., 48, 213 (1981).
36a. Withey, J. R., and Collins, B. T., J. Toxicol. Environ. Health, 2, 311 (1976).
37. Buchter, A., Filser, J. G., Peter, H., and Bolt, H. M., Toxicol. Lett., 6, 33 (1980).
38. Jaeger, R., J., Reynolds, E. S., Conolly, R. B., Moslen, M. T., Szabo, S., and Murphy, S.
D., Nature, 252, 724 (1974).
39. Bolt, H. M., Kappus, H., Buchter, A., and Bolt, W., Arch. Toxicol., 35, 153 (1976).
40. Viola, P. L., Med. Lav., 61, 174 (1970); P. L. Viola, A. Bigotti, and A. Caputo, Cancer
Res., 31, 516 (1971).

 41. Feron, V. J., and Krees, R., Toxicology, 13, 131 (1979).
42. Torkelson, T. R. , Oyen, F., and Rowe, V. K., Am. Ind. Hyg. Assoc. J., 22, 354 (1961).
43. Filatova. V. S., and Gronberg, E. S., Gig. Sanit., 22, 38 (1957).
44. Gabor, S., Lecca-Radu, M., and Manta, I., Prom. Toksikol. Klin. Prof. Zabole Khim.
Etiol., 221 (1962).
45. Gabor, S., Radu, M., Preda, S., Abrudean, S., Ivanof, L., Anea, Z., and Valaczkay, C.,
Igiena (Bucharest), 13, 409 (1964).
46. Grigorescu, I., and Toba, Gh., Rev. Chim. (Bucharest), 17, 499 (1966).
47. Kudryavyseva, D. G., Gig. Tr. Prof. Zabol., 14, 54 (1970).
48. Suciu, J., Drejman, J., and Valaskai, M., Med. Int., 15, 867 (1963).
49. Lange, C. E., Schwinger, E., and Veltman, G., Dtsch. Ges. Arbeitsmed. Munich (1975).

Page 294
50. Kramer, C. G., and Mutchler, J. C., Am. Ind. Hyg. Assoc. J., 331(1), 19 (1972).
51. Suciu, F., Drejman, J., and Valaskai, M., Med. Lav., 58, 261 (1967).
52. Marsteller, H. J. . Dtsch. Med. Wochenschr., 98, 2311 (1973).
53. Falk, H., Creech, J. L., Jr., Heath, C. W., Johnson, M. N., and Key, M. M., JAMA, 230,
59 (1974).
54. Thomas, L. B., and Popper, H., Ann. N.Y. Acad. Sci., 246 (1975).
55. Lange, C. E., Juhe, S., Stein, G., and Veltman, G., Int. Arch. Arbeitsmed., 32, 1 (1974).
56. Waxweiler, R. J., Falk, H., McMichael, A., and Mallone, J. S., A Cross-sectional
Epidemiologic Survey of Vinyl Chloride Workers, Center for Disease Control, Division of
Surveillance, Cincinnati, Ohio, Apr. 1977.
57. Veltman, G., Lange, C. E., Jube, S., Stein, G., and Bachner, U., Ann. N. Y. Acad. Sci.,
246, 6 (1975).
58. Lillis, R., Anderson, H., Nicholson, W. J., Daum, S., Fishbein, A. S., and Selikoff, I. J.,
Am. N. Y. Acad. Sci., 246, 22 (1975).
58a. Jones, D. B., and Smith, P. M. . Br. J. Ind. Med., 39, 306 (1982).
59. Cordier, J. M., Fievez, C., Lefevre, M. J., and Sevru, A., Cahiers Med. Trav., 4(143)
(1966).
60. Harris, D. K., and Adams. W. G. F., Br. Med. J., 5567, 712 (1967).
61. Wilson, R. H., McCormick, W. E., Tatum, C. F., and Creech, J. L., JAMA, 201, 577
(1967).
62. Basalaev, A. V., Gig. Tr. Prof. Zabol., 14, 34 (1970).
63. Dinman, B. D., Cook, W. A., Whitehouse, W. M., Magnuson, H. J., and Ditcheck, T.,
Arch. Environ. Health, 22, 61 (1971).
64. Cook, W. A., Giever, P. M., Dinman, B. D., and Magneson, H. J., Arch. Environ. Health,
22, 74 (1971).
65. Dodson, V. N., Dinman, B. D., Whitehouse, W. M., Naso, A. N. M., and Magneson, H. J.,
Arch. Environ. Health, 22, 83 (1971).
66. Grainger, R. G., Walker, A. E., and Ward, A. M. , in Induced Disease; Drug; Irradiation;
Occupation, L. Preger, ed., Grune and Stratton, London, 1980.

 67. Maricq, H. R., Johnson, M. N., Whetstone, C. L., and LeRoy, E. C., JAMA, 236, 1368
(1976).
68. Bertozzi, P. A., Arh. Hig. Rada Toksikol., 30(suppl.), 379 (1979).
69. Suciu, I., Prodan, L., Ilea, E., Paduraru, A., and Poscu, L., Ann. N.Y. Acad. Sci., 246, 53
(1975).
70. Occupational Safety and Health Administration, Vinyl Chloride, Job Health Hazard Series,
OSHA, Rep. 2225, June 1975.
71. CEFIC, Vinyl Chloride Toxicity and the Use of PVC for Packaging Foodstuff, Conseil
Europeen des Federations de l'Industrie Chimique, Brussels, February 1976.
72. National Toxicology Program, First Animal Report on Carcinogens, Vol. II, Dept. of
Health and Human Services, Washington, D.C., July 1980, p. 190.
73. Fishbein, L., Potential Industrial Carcinogens and Mutagens, Elsevier, New York, 1979.
74. Environmental Protection Agency, Standard for Vinyl Chloride, 41FR46560, Oct. 21,
1976, 40CFR61.
75. Environmental Protection Agency, Standard Support Document and Environmental
Impact Statement: Emission Standard for Vinyl Chloride, EPA 450/275/009, Sept. 1975.

Page 295
76. Filatova, V. S., and Gronsberg, E. Sh., Gig. Tr. Prof. Zabol., 15, 32 (1971).
76a. Barnes, A. W., Chem. Eng. News, 52(27), 21 (1974); B.F. Goodrich, Job Saf. Health,
1(2), 20 (1977).
77. Caputo, A., Viola, P. L., and Bigotti, A., IRCS, 2, 1582 (1974); J. Int. Res. Commun., 21,
1582 (1974).
78. Keplinger, M. L., Goode, S. W., Gordon, D. E., and Calandra, J. C., Ann. N.Y. Acad.
Sci., 246, 219 (1975).
79. Radike, M. J., Stemmer, K. L., and Bigham, E., Environ. Health Perspect., 41, 59
(1981).
80. Feron, V. J., Hendriksen, C. F. M., Speek, A. J., Til, H. P., and Spit, Ing B. J., Food
Cosmet. Toxicol., 19, 317 (1981).
81. Lee, C. C., Bhandari, J. C., Winston, J. M., House, W. B., Dixon, R. L., and Woods, J.
S., J. Toxicol. Environ. Health, 4, 15 (1978); Environ. Health Perspect., 21, 25 (1977).
82. Hong, C. B., et al., Additional Evaluation of the Environmental Toxicants VC and
Vinylidene Chloride. Final Report, Midwest Research Institute under NIH-NIGHS contract 1-ES22084, 1979; see also J. Toxicol. Environ. Health, 7, 909 (1981).
83. Hehir, R. M., McNamara, B. P., McLaughlin, J., Jr., Willigan, D. A., Bierbower, G., and
Hardisty, J. F., Toxicology, Carcinogenicity, and Reproductive Effects of Single and Multiple
Exposures to Vinyl Chloride in Rats and Mice, Pre-publication draft, U.S. Consumer Product
Safety Commission, Washington, D.C., Feb. 6, 1980; see also Environ. Health Perspect., 41,
63 (1981).
83a. Suzuki, Y., Environ. Res., 32, 91 (1983).
84. Creech, J. L., and Johnson, M. N., J. Occup. Med., 16, 150 (1974).
85. Block, J. B., JAMA, 229, 53 (1974).
86. Popper, H. Thomas, L. B., Telles, N. C., Falk, H., and Selikoff, I. J., Am. J. Pathol., 92,
349 (1978); H. Falk, J. Herbert, S. Crowley, K. G. Ishak, L. B. Thomas, H. Popper, and G. G.
Caldwell, Environ. Health Perspect., 41, 107 (1981).
87. Fortwengler, H. P., Jones, D., Espinosa, E., and Tamburro, H., Gastroenterology. 80,
1415 (1981).
88. Stafford, J., private communication, Jan. 1985. Since August 1983 this compilation has
been continued by Brian Bennett. Data reported here include his January 1985 report.

 88a. Drew, R. T., Boorman, G. A., Haseman, J. K., McConnell, E. E., Busey, W. M., and
Moore, U. A., Toxicol. Appl. Pharmacol., 68, 120 (1983).
88b. Groth, D. H., Coate, W. B., Ulland, B. M., and Hornung, R. W., Environ. Health
Perspect., 41, 53 (1981).
89. Delorme, F., and Theriault, G., J. Occup. Med., 20, 338 (1978).
90. Theriault, G., and Allard, P., J. Occup. Med., 23, 671 (1981).
91. Tabershaw, I. R., and Gaffey, W. R., J. Occup. Med., 16, 509 (1974).
92. Cooper,C., Environ. Health perspect., 41, 101 (1981).
92a. Cooper, C., Environ. Health Perspect., 41, 101 (1981). 92a. Marsh, G. M., J. Occup.
Med., 25, 219 (1983).
93. Monson, R. R., Peters, J. M., and Johnson, M. N., Lancet, ii, 397 (1974).
94. Waxweiler, R., Falk, H., McMichael, A., Mallow, J. S., Am. J. Epidemiol., 104, 347
(1976).
95. Waxweiler, R., Smith, A. H., Tyroler, H. A., and Falk, H., 19th Int. Cong. Occupational
Health, Dubrovnik, Yugoslavia, Sept. 1978.

Page 296
96. Waxweiler, R., Smith, A. H., Falk, H., and Tyroler, H. A., Environ. Health Perspect., 41,
159 (1981).
97. Greenburg, R., discussion of paper by R. Waxweiler in Ref. 223, transcript pages 337
342.
98. Falk, H., and Waxweiler, R. J., Proc. R. Soc. Med., 69, 303 (1976).
99. Greenberg., R. A., and Tamburro, C. H., J. Occup. Med. . 23(5), 353 (1981).
100. Beaumont, J. J., and Breslow, N. E., Am. J. Epidemiol., 114, 725 (1981).
101. Dannaher, C. L., Tamburro, C. H., and Yam, L. T., Cancer, 47, 466 (1981).
102. Tamburro, C. H., Med. Clin. N. Am. 63(3), 545 (1979).
103. Whelan, J. G., Creech, J. L., and Tamburro, C. H., Radiology, 118(3), 549 (1976).
103a. Langbein, G., Permanetter, W., and Dietz, A., Dtsch. Med. Wochenschr., 108, 741
(1983).
104. Tamburro, C. H., and Greenberg, R., Environ. Health Perspect., 41, 117 (1981).
105. Tamburro, C. H., Yale J. Biol. Med., 51, 67 (1978).
106. Eckardt, F., Muliwan, H., DeRuiter, N., and Kappus, H., Mutat. Res., 91, 381 (1981).
107. Ottenwalder, H., and Bolt, H. M., J. Environ. Pathol. Toxicol., 4, 411 (1980).
108. Du, J. T., Sandoz, J. P., Tseng, M. T., and Tamburro, C. H., J. Toxicol. Environ. Health, 5,
1119 (1979); Toxicol. Appl. Pharmacol., 62, 1 (1982).
108a. University of Louisville, Report on Research Techniques and Methods for the Detection
and Prevention of Carcinogenesis in the Industrial Worker, Louisville, Ky., 1982.
109. Duck, B. W., Carter, J. T., and Coombes, E. J., Lancet, ii, 1197 (1975).
110. Wagoner, J. K., Infante, P. F., and Saracci, R., Lancet, i, 194 (1976).
111. Duck, B. W., and Carter, J. T., Lancet, i, 195 (1976).
112. Berry, G., and Rossiter, C. E., Lancet, ii, 416 (1976).
113. Fox, A. J. , Lancet, ii, 416 (1976).
114. Reinl, W., Weber, H., and Greiser, E., 19th Int. Conf. Occup. Health, Dubrovnik,

Yugoslavia, Sept. 19, 1978.

115. Weber, H., Reinl, W., and Greiser, E., Environ. Health Perspect., 41, 95 (1981); E.
Grieser, W. Reinl, and H. Weber, Zentralbl. Arbeitsmed. Arbeitsschutz, Prophyl. Ergon., 32, 44
(1982).
116. Molina, G., Holmberg, B., Elofsson, S., Holmlund., L., Moosing, R., and Westerholm, P.,
Environ. Health Perspect., 41, 145 (1981).
116a. Austin, S. G. , and Schnetter, A. R., J. Occup. Med., 25, 313 (1983).
116b. Kono, S., Tokudome, S., Ikeda, M., Yoshimera, T., and Kuratsune, M., J. Natl. Cancer
Inst., 70, 443 (1983).
117. Jones, J. H., Worker Exposure to Vinyl Chloride in VC and PVC Production and
Fabrication, Aug. 1977, PB 83116053; see also W. F. Dimmick, Environ. Health Perspect., 41,
203 (1981).
118. Baxter, P. J., and Fox, A. J., Lancet, i, 245 (1976).
119. Brady, J., Liberatore, F., Harper, P., Greenwald, P., Burnett, W., Davies, J. N. P., Bishop,
M., Polan, A., and Vianna, N., J. Natl. Cancer Inst., 59, 1383 (1977).

Page 297
120. Fiechtner, J., Reyes, C., Rentmesster, K., and Skinner, H. G., Morbid. Mortal. Weekly
Rep., (Center for Disease Control), 25, 57 (1976).
121. Baxter, P. J., Anthony, P. P., McSween, R. N. M., and Scheuer, P. J., Br. Med. J., 11,
919 (1977).
122. B Baxter, P. J., Anthony, P. P., McSween, R. N. M., and Scheuer, P. J., Br. J. Ind. Med.,
37, 213 (1980).
123b. Saric, M., Kulcar, Z., Zorica, M., and Gelic, J., Environ. Health Perspect., 17, 189
(1976).
124. Elinder, C. G., and Pershagen, G., Pilot Study Concerning the Mortality in Njurunda
Community, Swedish Nature Conservancy Board, Apr. 1978.
125. Dalderup, L. M., Freni, S. C., Bras, G., and Bronckhorst, F. B., Lancet, i, 246 (1976); J.
Occup. Med., 17, 285 (1975).
126. Iturra, H., Proc. Air Environ. Specialty Conf., Pittsburgh, Pa, 1976, p. 96.
127. Kuzmack, A. M., and McCaughy, R. E., Quantitative Risk Assessment for Community
Exposure to Vinyl Chloride, U.S. EPA, Washington, D.C., Dec. 5, 1975.
128. Marcus, W., Comments during Hearing on the Vinyl Chloride Standard, EPA, Washington,
D.C., Feb. 3, 1976, Transcript, p. 43.
129. Block, J. B., J. Ky. Med. Assoc., 72(9), 483 (1974).
130. Heath, C. W., Flak, H., and Creech, S. L. Jr., Ann. N.Y. Acad. Sci., 246, 231 (1975); see
also Environ. Res., 14, 68 (1977).
131. Gedigk, P., Muller, R., and Bechtelsheimer, H., Am. N. Y. Acad. Sci., 246, 278 (1975).
132. Schaffner, F., Falk Symp., 25, 1978, p. 189.
133. Gordon, D. E., Thomas, L. B., Calandra, J. C., Popper, H., and Kent, G., Int. Acad.
Pathol. Meet., New Orleans, Mar. 5, 1975; abstracted in Lab. Invest., 32(3), 8 (1975).
134. Koischwitz, D., Marsteller, H. J., Lackner, K., Brecht, G., and Brecht, T., Fortschr.
Rontgenstr., 134(3), 283 (1981).
135. Hopkins, J., Food Cosmet. Toxicol., 17, 542 (1979).
136. Fleig, J., and Thiess, A. M., ASP, 9, 282 (1974); Abstr. 2nd Int. Conf. Environ.
Mutagens, 1977, p. 219.

137. Hansteen, I. L., Hillestad, I. -L. L., Thiis-Evensen, E., and Heldas, S. S., Mutat. Res., 78,
211 (1978).
138. Anderson, D., Richardson, C. R., and Purchase, I. F. H., Mutat. Res., 83, 137 (1981).
139. Picciano, D. J., Flake, R. F., Gay, P. C., and Kilian, D. J., J. Occup. Med., 19, 527
(1977).
140. Basler, A., and Rohrborn, G., Arch. Toxicol., 45, 1 (1980).
140a. Miller, M. L., Radike, M. J., Andruiga, A., and Bingham, E., Environ. Res., 29, 272
(1982).
141. Anderson, D., Richardson, C. R., Purchase, I. F. H., Evans, H. J., and O'Riordan, M. L.,
Mutat. Res., 40, 359 (1976).
141a. Himeno, S., Okuda, H., and Suzuki, T., Toxicol. Lett., 16, 47 (1983).
142. Short, R. D., Minor, J. L., Winston,. J. M., and Lee, C. C., J. Toxicol. Environ. Health, 3,
965 (1977).
143. Schwetz, B. A., Leong, B. K., and Smith, F. A., Toxicol. Appl. Pharmacol., 33, 134
(1975).
144. John, J. A., Smith, F. A., and Schwetz, B. A., Toxicol. Appl. Pharmacol., 39, 497 (1977);
Environ. Health Perspect., 41, 171 (1981).

Page 298
145. Rice, J. M., Environ. Health Perspect., 41, 179 (1981).
146. Infante, P. F., Ann. N. Y. Acad. Sci., 271, 49 (1976); see also Mutat. Res., 41, 131
(1976).
147. Morbid. Mortal. Weekly Rep., (Center for Disease Control), 24(29), 245 (July 19, 1975).
148. Edmonds, L., Proc. Conf. Women and the Workplace, Washington, D.C., 1976; see also
Teratology, 17, 137 (1978).
149. Edmonds, L. D., Falk, H., and Nissim, J. E., Lancet, ii, 1098 (1975).
150. Theriault, G. P., and Goulet, L., Annu. Meet. Am. Public Health Assoc., 1977.
151. Theriault, G. P., Iturra, H., and Gringas, S., Association Between Birth Defects and
Exposure to Ambient Vinyl Chloride, EPA-600/181/057, Sept. 1981, PB 81238883, NTIS,
Springfield, Va.
152. Infante, P. F., Wagoner, J. K., McMichael, A. J., Waxweiler, R. J., and Falk, H., Lancet,
734 (1976).
153. Paddle, G. M., Lancet, 1079 (1976).
154. MacMahon, B., Vinyl Cloride and Human Reproduction, submitted by the Society of the
Plastics Industries, Inc., in comments on the Proposed Amendment to the EPA Vinyl Chloride
Standard, 1977.
155. Downs, T. D., Stallones, R. A., Frankowski, R. F., and Labarthe, D. R., Vinyl Chloride,
Birth Defects, and Fetal Wastage-A Critical Review, prepared for the Society of the Plastics
Industries, Inc., by Research Statistics Inc., Houston, Sept. 16, 1977.
156. Monson, R. R., Occupational Epidemiology, CRC Press, Boca Raton, Fla., 1980, p. 190.
157. Buffler, P. A., and Aase, J. M., J. Occup. Med, 24, 305 (1982).
158. Hass, J. F., and Schottenfeld, D., J. Occup. Med., 21, 607 (1979).
159. Clemmesen, J. Mutat. Res., 98, 97 (1982).
160. Hatch, M., Kline, J., and Stein, Z., Environ. Health Perspect., 41, 195 (1981).
161. Society of the Plastics Industries, Inc., Vinyl Acetate Task Force Steering Committee,
Report on the Toxicity Studies of Vinyl Acetate Performed by Hazelton Laboratories Europe
Ltd., New York, June 19, 1980.
162. National Institute of Occupational Safety and Health, Criteria for a Recommended

Standard-Occupational Exposure to Vinyl Acetate, DHEW (NIOSH) Publ. 78205, 1978.

162a. Hellman, T. M., and Small, F. H., Chem. Eng. Prog., 69(9), 75 (1973).
163. National Institute of Occupational Safety and Health, Registry of Toxic Effects of
Chemical Substances, DHHS (NIOSH) Publ. 81116, Cincinnati, Ohio, 1982.
164. Deese, D. E., and Joynes, R. E., Am. Ind. Hyg. Assoc. J., 30, 449 (1969).
164a. Holbub, I., and Tarkowski, S., Int. Arch. Environ. Health, 51, 185 (1982).
165. Maltoni, C., in Origins of Human Cancer, Book A, H. H. Hiatt, J. D. Watson, and J. A.
Weinstein, eds., Cold Spring Harbor Laboratories, Cold Spring Harbor, N.Y., 1977, p. 119.
166. American Conference of Governmental Industrial Hygienists, Documentation of the
Threshold Limit Values, 3rd ed., Cincinnati, Ohio, 1971.
166a. Lijinsky, W., and Reuber, M. D., Toxicol. Appl. Pharmacol., 68, 43 (1983).

Page 299
167. Shirinyan, G. S., and Arutynenyan, R. M., Biol. Zh. Arch., 33, 748 (1980).
168. International Agency for Research on Cancer, Screening Tests in Chemical
Carcinogenesis, Sci. Publ. 12, WHO-IARC, Lyon, France, 1976.
169. Lijinsky, W., and Andrews, A. W., Teratog. Carcinog. Mutagen., 1, 289 (1980).
170. Boulton, T. B., and Sweet, R. B., J. Mich. St. Med. Soc., 59, 270 (1960).
171. Kleinfeld, M., and Tabershaw, I. R., AMA Arch. Ind. Hyg. Occup. Med., 10, 134 (1954).
172. St. Hill, C. A., Trans. Soc. Occup. Med., 16, 6 (1966).
173. Tomasini, M., and Sartorelli, E., Med. Lav., 62, 277 (1971).
174. National Institute for Occupational Safety and Health, Special Occupational Hazard
Review of Trichloroethylene, DHEW (NIOSH) Publ. 78130, 1978, PB 81226987.
175. James, W. R. L., Br. J. Ind. Med., 20, 47 (1963).
176. Clearfield, H., Dig. Dis., 15, 851 (1970).
177. Lillis, R., Stanescu, P., and Roventa, A., Med. Lav., 60, 595 (1969).
178. Torkelson, R. Tr., and Rowe, K. V. Patty's Industrial Hygiene and Toxicology, 3rd rev.
ed., Vol. IIB, G. D. and F. E. Clayton, eds., Wiley-Interscience, New York, 1981, Chap. 48.
179. Environmental Protection Agency, Health Assessment Document for Trichloroethylene,
EPA-600/882006B, Washington, D.C., 1983.
179a. Crebelli, R., Bignami, M., Conti, L., and Carene, A., Ann. Ist. Super. Sanita, 18, 117
(1982).
180. Secchi, G. C., Chiappino, G., Lotto, A., and Zurlo, N., Med. Lav., 59, 486 (1967).
181. Chemical Manufacturers Assoc., The Pharmacokinetics and Macromolecular Interaction
of Trichloroethylene as Related to Oncogenicity, prepared by G. S. Stott et al., of Dow
Chemical Co., Washington, D.C., 1981.
182. Steward, R. D., Hake, C. L., and Peterson, J. E., Arch. Environ. Health, 29, 1,6 (1974).
183. Stott, W. T., Reitz, R. H., Schumann, A. M., and Watanabe, P. G., Food Cosmet.
Toxicol., 19, 567 (1981).
183a. Van Duuren, B. L., Kline, S. A., Melchionne, S., and Seidman, J., Can. Res., 43, 159
(1983).

183b. Miller, R. E., and Guengerich, F. P., Can. Res., 43, 1145 (1983).
183c. Politzer, P., and Hedges, W. L., Int. J. Quantum Chem.: Quantum Biol. Symp., 9, 307
(1982); P. Politzer et al., Ann. N.Y. Acad. Sci., 367, 478 (1981).
183d. Soleo, L., Elia, G. E., and Cassano, F., Riv. Med. Lav. Ig. Ind 3, 127 (1979).
183e. Stewart, R. D., Hake, C. L., LeBrun, A. J., Pitenson, J. E., and Foster, H. U., Biologic
Standards for the Industrial Worker by Breath Analysis: Trichloroethylene, PB 83175844,
NTIS, Springfield, Va.
184. Klein, S. A., McCoy, E. C., Rosenkranz, H. S., and Van Duuren, B. L., Mutat. Res., 101,
115 (1982).
185. National Cancer Institute, Carcinogenesis Bioassay of Trichloroethylene, HEW (NIH)
Publ. 76802, 1976.
186. Maltoni, C., and Lefemine, G., Banbury Rep., 5, 3 (1980).
187. Henschler, D., Eder, E., Neudecker, T., and Metzler, H., Arch. Toxicol., 37, (1977).

Page 300
188. Van Duuren, B. L., Goldschmidt, B. M., Loewengart, G., Smith, A. C., Melchionne, S.,
Seldman, I., and Roth, D., J. Natl. Cancer Inst., 63, 1433 (1979).
188a. Henschler, D., Forschungsber.-Bundesminist. Forsch. Technol., Hum. Arbeitslebens,
BMFT-FB-HA 82007, 1982.
189. NTP Reluctantly Accepts Report Linking TCE to Cancer, Pestic. Toxic Chem. News, 14
(June 23, 1982).
189a. Rall Says Faulty NTP Studies May Be Salvaged by Reviewers, Food Chem. News, 1
(Nov. 14, 1983).
189b. Jones, R. B., and Mackrodt, W. C., Biochem Pharmacol., 31, 3710 (1982); 32, 2359
(1983).
190. Axelson, O., Andersson, K., Hogstedt, C., Holmberg, B., Molina, G., and deVerdier, A.,
J. Occup. Med., 20, 194 (1978).
191. Tola, J., J. Occup. Med., 22, 737 (1980).
192. Blair, A., Decaufle, P., and Grouman, D., Am. J. Public Health, 69, 508 (1979).
193. Lin, R. S., and Kessler, I. I., JAMA, 245, 147 (1981). 194. Katz, R. M., and Jowett, D.,
Am. J. Public Health, 71, 306 (1981).
194. Katz, R.M. , and Jowett, D., Am. J. Public Health, 71, 305(1981).
194a. Slacik-Erben, R., Roll, R., Franke, G., and Uehleke, H., Arch. Toxicol., 45, 37 (1980).
195. Healey, T. E. J., Poole, T. R., and Hooper, T. R., Br. J. Anaesth., 5, 4337 (1982); R. P.
Beliles, PB 82185075, 1982.
196. Squire, R. A., Science, 214, 877 (1981); G. M. Williams, Food Cosmet. Toxicol., 19, 577
(1981); H. Greim, U. Andrae, W. Goggelmann, S. Hesse, L. R. Schwarz, and K. H. Summer,
Prog. Mutat. Res., 2, 129 (1981); H. F. Kraybill, J. Environ. Sci. Health, C1(2), 175 (1983).
197. National Research Council Committee on Hazardous Substances in the Laboratory,
Prudent Practices for Handling Hazardous Chemicals in Laboratories, National Academy Press,
Washington, D.C., 1981.
197a. United Kingdom Health and Safety Executive, Toxicity Review-Tri-chloroethylene. HMSO,
London, 1982.
197b. International Agency for Research on Cancer, Chemicals, Industrial Processes and
Industries Associated with Cancer in Humans, IARC Monographs, Vols. 1 to 29, Suppl. 4, Lyon,
France, 1982.

198. National Research Council Safe Drinking Water Committee, Drinking Water and Health,
Vol. 3, National Academy Press, Washington, D.C., 1970.
199. McCollister, D. D., AMA Arch. Ind. Health, 13, 1 (1956).
200. Smyth, H. F., Jr., and Wiel, C. S., Am. Ind. Hyg. Assoc. J., 30, 470 (1969).
201. National Institute for Occupational Safety and Health, Criteria for a Recommended
Standard-Occupational Exposure to Ethylene Dichloride, DHEW (NIOSH) Publ. 76139, 1976,
Washington, D.C.
202. World Health Organization, Toxicological Evaluation of Some Extraction Solvents and
Certain Other Substances, FAO Nutr. Meet. Rep. Ser 48A WHO/FAO ADD/70.39, 1970.
203. Heppel, L. P., Neal, P. A., Endicott, K. M., and Porterfield, V. T., AMA Arch. Optothalmol.
32, 391 (1944).
204. Spencer, G., Rowe, V. K., Adams, E. M., McCallister, D. D., and Irish, D. D., AMA Arch.
Ind. Hyg. Occup. Med., 4, 482 (1951).
205. Prezezdziak, J., and Bakula, S., Wiad. Lek., 28, 983 (1975).
206. Lane, R. W., Riddle, B. L., and Borzelleca, J. F., Toxicol. Appl. Pharmacol., 63, 409
(1982).

Page 301
207. Hofonau, H. T., Birnstial, H., and Johst, P., Arch. Toxicol., 27, 244 (1971).
208. Heppel, L. A., Neal, P. A., Perric, T. L., Endicott, K. M., and Porterfield, N. T., J. Ind.
Hyg. Toxicol., 28, 113 (1946).
209. Alumont, E., Nachtomi, E., Mandel, E., Holstein, P., Bandi, A., and Hergberg, M., Food
Cosmet. Toxicol., 14, 105 (1976).
210. Menschick, H., Arch, Gewerbepathol. Gewerbehyg., 15, 241 (1957).
211. von Oettingen, W. F., The Halogenated Aliphatic, Cyclic Aromatic, and Aliphatic-Aromatic
Hydrocarbon, U.S. Public Health Serv. Publ. 414, 1955.
212. National Cancer Institute, Bioassay of 1, 2-Dichloroethane for Possible Carcinogenicity,
DHEW (NIH) Publ. 781361 (1978).
213. Maltoni, C., Valgmigh, L., and Scarnato, C., Cold Spring Harbor Symp., Nov. 1519,
1979, Banbury Rep., 5, 3 (1980).
214. Ralo, K. S., Murray, J. S., Deacon, M. M., John, J. A., Calhoun, L. L., and Young, J. T.,
in Banbury Rep., 5, Cold Spring Harbor Laboratory, 1980.
215. Riddle, B. L., Carchman, R. A., and Borzellua, J. F., Toxicologist, 1, 26 (1981).
216. Chastain, C. E., SPE Tech. Pap., 18, Part 1, 202 (1972).
217. Miller, A., Teirstein, A. S., Chuang, M., Sellikoff, I. J., and Warsaw, R., Ann. N.Y. Acad.
Sci., 246, 42 (1975).
218. Axnaud, A., Pommier, P., de Sante, Gaebe, L., Payan, H., and Charpin, J., Thorax, 33,
19 (1978).
219. Mastrangelo, G., Manno, M., Marcer, G., Bartolucci, G. B., Gemignani, C., Saladino, G.,
Simonato, L., and Saia, B., J. Occup. Med., 21, 541 (1979).
220. Cordasco, E. M., Demeter, S. L., Kerkay, J., Van Ordstrand, H. S., Lucas, E. V., Chen,
T., and Golish, J. A., Chest, 78, 828 (1980).
221. Chivers, C. P., Lawrence-Jones, C., and Paddle, G. M., Br. J. Ind. Med., 37, 147
(1980).
222. Soutar, C. A., Copland, L. H., Thornley, P. E., Hurley, J. F., Ottery, J., Adams, W. G. F.,
and Bennett, B., Thorax, 35, 644 (1980).
222a. Soutar, C. A., Gauld, S., Lloyd, M., Copeland, L. H., and Hurley, J. F., Epidemiological
and Clinical Studies of Polyvinylchloride Workers, Rep. TM/81/8, Institute of Occupational

Medicine, Edinburgh, July 1981, PB 84110402.

223. National Institute of Health, Conf. to Reevaluate the Toxicity of Vinyl Chloride Monomer,
Poly(vinyl chloride) and Structural Analogs, Bethesda, Md., Mar, 2021, 1980.
224. Frangia, N., Spinozzala, A., and Bucanelli, A., Med. Lav., 65, 321 (1974).
225. Agarwal, D. K., Environ. Res., 16, 333 (1978).
226. Tetley, T. D., Rose, F. A., and Richards, R. J., Inflammation (New York), 5, 137 (1931).
227. Ventkin, Yu. I., and Nikonov, A. A., Gig. Tr. Prof. Zobol., No. 7, 48 (1981).
228. NIOSH unpublished report, PVC Chronic Inhalation Toxicology Study, 1977.
229. Styles, J. A., and Wilson, J., Ann. Occup. Hyg., 16, 241 (1973).
230. Richards, R. J., Desai, R., Hext, P. M., and Rose, F. A., Nature, 260, 664 (1975).

Page 302
231. Richards, R. J., Desai, R., and Rose, F. A., Nature, 260, 55 (1976).
232. Smyth, H. F., Jr., and Weil, C. S., Toxicol. Appl. Pharmacol., 9, 501 (1966).
233. Johnson, W. S., and Scmid, R. E., Am. J. Vet. Res., 38, 1891 (1977).
234. Air Products and Chemicals, Inc., unpublished data.
235. International Agency for Research on Cancer, Vinyl Chloride, PVC and VC/VAC
Copolymers, Sci. Publ. 19, Lyon, France, 1979.
236. Ziche, M., and Gullino, P. M., Cancer Res., 41, 5060 (1981).
237. Bartnecht, W., Explosions, Springer-Verlag, New York, 1981.
238. Modern Plastics Encyclopedia, Vol. 58(10A), McGraw-Hill, New York, 1981, p. 632.
239. Hilado, C. J., Mod. Plast., 54(7), 64 (1977).
240. Hilado, C. J., J. Consum. Prod. Flamm., 4, 244 (1977).
241. Cook, W. A., Am. Ind. Hyg. Assoc. J., 41, 508 (1980).
242. Boettner, E. A., and Ball, G. L., Am. Ind. Hyg. Assoc. J., 41, 513 (1980).
243. Sangha, G. K., Matijik, M., and Alarie, Y., Am. Ind. Hyg. Assoc. J., 42, 481 (1981).
244. Wakeman, I. B., and Johnson, H. P., Polym. Eng. Sci., 18, 404 (1978).
245. See the comments filed by the SPI and others in Docket 75N-0190 at the FDA, in 1975
1976.
245a. Showdown on Vinyl Plant Rule Pressages Shutdown, Chem. Week, Sept. 25, 1974, p.
15.
246. National Toxicology Program, Second Annual Report on Carcinogens, Dept. of Health
and Human Services, Washington, D.C., 1981.
246a. Society of the Plastics Industry, Report on Food Contact Usage of PVC Plastics, SPI,
New York, Aug. 1983.
247. Chemical Week, Mar. 28, 36 (1979).
248. Environmental Protection Agency, Vinyl Chloride. A Review of National Emission
Standards, EPA-450/382003, Feb. 1982, PB 84114354.

248a. Perry, C. R., Toxicol. Sub. J., 2, 215 (1981).

249. Graham, J. D., and Vaupel, J. W., Risk Anal., 1, 89 (1981).
250. Luken, R. H., and Miller, S. G., J. Air Pollut. Control Assoc., 31, 1204 (1981).
251. Morrell, J. F., from a study for the Office of Management and Budget, as quoted in
Pestic. Toxic Chem. News, 14 (Jan. 13, 1982).
252. Northrup, H. R., et al., The Impact of OSHA, Labor Relations and Public Policy Series
17, The Wharton School, University of Pennsylvania, Philadelphia, 1978.
253. Hoffman, D., Patrianokos, D., Brunnerman, C., and Goni, K. G., Anal. Chem., 48, 47
(1976).
254. Grinard, M., Taft, P., and Wiberg, G. S., Report of Task Force on Vinyl Chloride,
Environmental Health Directorage, Ottawa, Canada, June 1976.
255. Environmental Protection Agency, Scientific and Technical Assessment Report on Vinyl
Chloride and Polyvinyl Chloride, EPA-600/675004, June 1975.
256. Dilling, W. L., Bredeweg, C. J., and Tefertiller, N. B., Environ. Sci. Technol., 10, 351
(1976).
257. Cox, R. A., Eggelton, A. F. J., and Sandalls, F. J., Photochemical Reactivity of Vinyl
Chloride, U.K. at Energy Res. Establishment, Publ. AGRE-R-7820, Sept. 1974.
258. Gay, B. W., Jr., Noonan, R. C., and Bufalini, J. J., Environ. Sci. Technol., 10, 58 (1976).

Page 303
259. Singh, H. B., Salas, L. J., and Stiles, R. E., Environ. Sci. Technol., 16, 872 (1982); also
EPA-600/383002, Mar. 1983.
259a. Dimethriades, B., Gay, B. W., Jr., and Seila, R. L., J. Air Pollut. Control Assoc., 33, 575
(1983).
260. Lillian, D., Singh, H. B., Appleby, A., Lobban, L., Arnts, R., Gumpert, R., Hogue, R.,
Toomey, J., Kozozis, J., Antell, M., Hansen, D., and Scott, B., Environ. Sci. Technol., 9, 1042
(1975).
261. Bozzelli, J. W., Ward, J., and Kebbekus, B. B., Analysis of Selected Toxic and
Carcinogenic Substances in Ambient Air in New Jersey, N.J. Dept. of Environmental Protection,
May 1980, and Paper 821.5, 75th Natl. Meet. Air Pollut. Control Assoc., New Orleans, June
20, 1982. See also P. J. Liay and J. M. Daisey, J. Air Pollut. Control Assoc., 33, 649 (1983).
262. Environmental Protection Agency, Treatability Manual, rev., Sept. 1981, EPA-600/282
001. 26. 2a. Bradzivsky, R., and Singh, H. B., Volatile Organic Chemicals in the Atmosphere:
An Assessment of Available Data. EPA-600/383027A, Apr. 1983, PB 83195503 (paper or
microfiche) and PB 83195511 (computer tape), NTIS, Springfield, Va.
262a. Bradzivsky, R., and Singh, H. B., Volatile Organic Chemicals in the Atmosphere: An
Assessment of Available Data. EPA-600/383027A, Apr. 1983. PB 83195503 (paper or
microfiche) and PB 83195511 (computer tape), NTIS, Springfield, Va.
263. Environmental Protection Agency, Preliminary Assessment of the Environmental
Problems Associated with Vinyl Chloride and Polyvinyl Chloride, EPA-500/474001, 1974.
264. Environmental Protection Agency, Ambient Water Quality Criteria for Vinyl Chloride, EPA440/580078, Oct. 1980.
265. Barr, J. T., Paper 829.2, 75th Annu. Meet. Air Pollut. Control Assoc., New Orleans,
June 28, 1982.
266. Albert, R. E., letter to R. S. Naveen, EPA, Comparison of Vinyl Chloride Carcinogenic
Risks with Risk from Other Pollutants, U.S. EPA, Washington, D.C., June 16, 1978.
267. Environmental Protection Agency, The Cost of Clean Air and Clean Water, Annual Report
to the Congress, Dec. 1979. Senate Document 9638, U.S. Government Printing Office,
Washington, D.C.
267a. United States of America v. Ethyl Corp., Civil action 830120. A, Middle District of
Louisiana, Minute Entry of July 1, 1983, Judge C. V. Parker.
268. Hill, J., IV, Kollig, H., Ravis, D., Wolfe, N., and Zepp, R., Dynamic Behavior of Vinyl
Chloride in Aquatic and Cosystems, EPA-600/376001, Jan. 1976.

269. Dilling, W. L., Environ. Sci. Technol., 9, 833 (1975). See also P. V. Roberts and P. G.
Dandliker, Environ. Sci. Technol. 17, 484 (1983).
270. Lu, P. Y., et al. Arch. Environ. Contam. Toxicol., 6, 129 (1977).
271. Dressman, R. C., and McFarren, G. F., Am. Water Works Assoc. J., 70, 29 (1978).
272. Banzer, J. D., J. Vinyl Technol., 1, 164 (1979).
273. Environmental Protection Agency, Multimedia Environmental Goals for Environmental
Assessment, EPA-600/779176, Vol. III, Aug. 1979.
273a. Environmental Protection Agency, Fate of Priority Pollutants in Publicly Owned Treatment
Works, EPA-440/182303, Sept. 1982, PB 83122788.
274. Environmental Protection Agency, Water Quality Criteria Documents. Dichloroethane, PB
81117400; TCE, PB 81117871; Vinyl Chloride, PB 81117889. See 45FR79318 for a
summary of these.

Page 304
274a. Memorandum of Understanding between the EPA and the FDA. Signed June 12, 1979 by
the EPA, June 22, 1979 by the FDA.
274b. National Research Council, Drinking Water and Health, Vol. 4, National Academy Press,
Washington, D.C., 1982.
274c. National Sanitation Foundation, Proposed Organohalide Leachate Testing Protocol for
Plastic Piping, Ann Arbor, Mich., Apr. 1983.
274d. McClelland, N., as quoted in Food Chem. News, June 27, 1983, p. 9.
274e. Wood, P. R., Parsons, F. Z., Lang, R. F., and Payan, I. L., American Water Works
Assoc. Annu. Conf. Expo., St. Louis, Mo., June 711, 1981; F. Parsons, P. R. Wood, and J.
DeMarco, J. Am. Water Works Assoc., 76 56 (1984); J. C. Baker, An Investigation into the
Source of Vinyl Chloride Detected at the Preston and Hialeah Water Treatment Plants, Dept. of
Environmental Resource Management, Dade County, Fla., 1983.
275. NIOSH, Engineering Control Technology Assessment for the Plastics and Resins
Industry, DHEW (NIOSH) Publ. 78159, Mar. 1978.
276. Proc. NIOSH-Sponsored Symp.: Control Technology in the Plastics and Their Industry,
Atlanta, Feb. 2728, 1979, prepared by Enviro. Controls, Inc., Rockville, Md.
276a. Veralin, C. N., Hydrocarbon Process., 78(2), 183 (1978).
276b. Chem. Eng. News, 58(31), 7 (1980).
276c. M & M Protection Consultants, One Hundred Largest Losses-A 30-year Review of
Property Damage Losses in the Hydrocarbon-Chemical Industry, Mrash & McLennan, New
York, 1981. See also Chem. Eng. News, 58(31), 7 (1980), 60(44), 28 (1982); and Chem.
Week, 128(3), 21 (1981).
276d. Phillips, E. A., and Role, H., Phase 02 Report on Effectiveness of Shelf Couplers. Head
Shields and Thermal Shields, Rep. RA-02344, Association of American Railroads, Chicago,
May 1981.
277. NFPA Fire Protection Guide on Hazardous Materials, 5th ed., Boston, 1973.
278. Wheeler, R. N., Barr, J. T., Laundry, R. W., and Snyder, P. J., Properties and Essential
Information for Safe Handling and Use of Vinyl Chloride. Appendix A-II of Ref. 276.
278a. McQuaid, J., and Fitzpatrick, R. D., J. Occup. Accid., 5, 121 (1983); J. McQuaid, J.
Occup. Accid., 5, 135 (1983).
278b. Snyder, W. H., Guideline for Fluid Modeling of Atmospheric Diffusion, EPA-450/479

016, U.S. EPA, Office of Air Quality Planning and Standards, Research Triangle Park, N.C.,
1979.
278c. Hartwig, C., ed., Heavy Gas Risk Assessment, 2 (Proceeding of the Second Symposium,
1982), Reidel, Dordnecht, The Netherlands, 1983.
279. Boggus, J. D., and Adams, N. B., Anal. Chem., 30, 1471 (1958).
280. Lederer, M., Angew. Chem., 71, 162 (1959).
281. Rozuvalv, G. G., and Minsker, K. S., Zh. Obsch. Khim., 27, 2875 (1955); 28, 933 (1958).
282. Barr. J. T., Manufacture of Plastics, Vol. I, W. M. Smith, ed., Reinhold, New York, 1964.
283. Terwiesch, C., Chemiche Werke Huls, Marl, Germany, private communication.
284. Chemical Manufacturers Association, Case Histories of Accidents in the Chemical
Industry, Vol. 3, Washington, D.C., 1970, p. 142 (Case History 1551).

Page 305
284a. Spark Ignition, Properties of Hand Tools, ADI PSD 2214, American Petroleum Institute,
Washington, D.C., Oct. 1980.
285. Noller, D. C., Mozurowski, S. J., Linden, G. F., DeLeeuw, F. J. G., and Mageli, O. I., Ind.
Eng. Chem., 56(12), 18 (1964); J. B. Armitage and H. W. Strauss, Ind. Eng. Chem., 56(12), 28
(1964).
286. Verhelst, W. F., From VCM to PVC: By Which Initiator, Noury Chemical Corp., Burt, N.Y.,
1980.
287. Ludicol Div., Pennwalt Corp., Peroxyesters, Buffalo, N.Y., n.d.; D. H. Walrod and R. G.
Gimbarski, Plast. Compd., Jan./Feb. 1979; Noury Chemical Corp., Organic Peroxide Safety
and Handling, Burt, N.Y., 1978.
287a. Illidge, J. T., and Wolstenholme, J., Inst. Chem. Eng. Loss Prevention Symp., Atlanta,
Feb. 28-Mar. 2, 1978.
288. Boyle, W. J., Jr., Chem. Eng. Prog., 63(8), 61 (1967).
289. Huff, J. E., Loss Prevention (CEP Tech. Manual), 7, 45 (1973).
290. Duxbury, H. A., 81st Natl. Meet. Am. Inst. Chem. Eng., Kansas City, Mo., Apr. 1976;
Chem. Eng., 31 (1980).
290a. Kemp, H. S., Chem. Eng. Prog., 79(6), 9 (June 1983).
290b. ASME Sec. VIII, Div. I, Para. UG-127, American Society for Mechanical Engineering,
New York.
290c. Harris, L. R., Hydrocarbon Process., 83(5), 75 (1983).
290d. Chemische Werke Huls, U.S. Pat. 4,004,880 (Jan. 25, 1977).
290e. Fressell, J. B., Synthetic Tree Model-A Formal Methodology for Fault-Tree Construction,
Aerojet-Nuclear Report ANCR-1098, Mar. 1973, NTIS, Springfield, Va.; G. J. Powers and F. C.
Tompkins, AICLE J., 20 376 (1974); V. Platz, Hydrocarbon Process., 80(5), 275 (1980).
290f. Bush, S. H., Trans. AMSE, 54 (1975); L. C. Doelp and P. L. T. Brian, Ind. Eng. Chem.
Fundam., 21, 101 (1982); T. Kletz, Process Technol. Int., 18, 111 (1973).
290g. Lees, F. P., Loss Prevention in the Process Industries, Butterworth, Boston, 1980; A. P.
Swain and H. E. Guttman, Handbook of Human Reliability Analysis with Emphasis on Nuclear
Power Plant Applications, NUREG/CR-1271, U.S. Nuclear Regulatory Commission,
Washington, D.C., Oct. 1980.
290h. National Transportation Safety Board, Railroad Accident Report-Derailment of Illinois

Central Gulf Railroad Freight Train Extra 9629 East (GS-228) and Release of Hazardous
Materials at Livingston, Louisiana, September 28, 1982, Rep. NTSB/RAR-83/05, Aug. 10,
1983, PB 83916305.
290i. Kilmartin, J., Fire J., 78(1), 62 (1984).
291. Cameron, J. B., Lunden, A. J., and McCultey, J. H., Jr., Hydrocarbon Process., 57(3),
39 (1980).
292. Krause, F. E., Prevention of PVC Reactor Fouling, in Ref. 262.
293. Barr, J. T., Solvent Cleaning of PVC Reactors, in Ref. 262. See also U.S. Pat.
4,009,048 to Air Products (Feb. 22, 1977).
293a. Fetterolf Corp., Tech Sheet Ram-Seal Spray-Rinse Valves Operation and Maintenance
Manual, Skippack, Pa., 19474.
293b. Jones, B., and Harris, R. L., Am. Ind. Hyg. Assoc. J., 44, 795 (1983).
294. National Institute of Occupational Safety and Health, Industrial Hygiene Survey, PB 81
225344 and PB 82109976, NTIS, Springfield, Va.
294a. J. Steele v. B. F. Goodrich Chemical Co. State of Illinois Human Rights Commission,
Charge No. 1980 CF0952, decision by J. Gerl, entered Nov. 17, 1982.

Page 306
294b. Cameron, J. B., Lundren, A. J., and McCulley, J. H., Jr., Hydrocarbon Process. , 80(3),
39 (Mar. 1980).
294c. Mantell, G. J., Barr, J. T., and Chan, R. K. S., Chem. Eng. Prog., 71(9), 54 (1975).
294d. Krause, F. E., Vinyl Chloride Monomer Stripping of PVC Resins, in Ref. 276. See also
B.F. Goodrich, Belgian Pat. 843, 624 (1977).
294e. Davis, R. J., Berens, A. R., Huddleston, G. R., Jr., and Witenhafer, D. E., (B.F.
Goodrich). Ger. Pat. 2,628,700 (Jan. 20, 1977); K. Kidoh, H. Wakamori, and K. Jimura
(Kureha). Ger. Pat. 2, 581, 429 (July 19, 1974) (see also U.S. Pat. 4032497); F. Boetsch, C.
D. I. Heinze, and H. Wolff (Hoechst A.G.), Ger. Pat. 2,435,704 (Feb. 5, 1976) (see also U.S.
Pat. 4,158,092).
294f. Burgess, R. H., ed., Manufacture and Processing of PVC, Macmillan, New York, 1981.
294g. Berens, A. R., Polym. Prepr., 15(2), 203 (1974).
294h. Berens, A. R., Polymer, 18, 697 (1977).
294i. Patel, C. B., Grandin, R. E., Gupta, R., Phillips, E. M., Reynolds, C. E., and Chan, R. K.
S., Polym. J., 11, 43 (1979).
294j. Berens, A. R., Polym. Eng. Sci., 20, 95 (1980).
294k. Berens, A. R., and Daniels, C. A., Polym. Eng. Sci., 16, 552 (1976).
294l. Crider, L. B., O'Mara, M. M., and Bowles, R. L., J. Vinyl Technol., 1, 168 (1979).
294m. Mack, W. A., Chem. Eng. Prog., 71(9), 41 (1975).
294n. Gilbert, S. G., and Giacin, J. R., Saf. Health Plastic, Natl. Tech. Conf. Soc. Plast. Eng.,
35 (1977); J. Miltz, J. Food Process Preserv., 4(12), 27 (1980).
294o. Haefner, A. J., and Hughmark, G. A., J. Vinyl Technol., 1, 5 (1979).
294p. Wallace, J. R., 33rd Can. Chem. Eng. Conf., Toronto, Oct. 25. 1983.
295. Schillmoller, C. M., Hydrocarbon Process., 56(3), 89 (1979); F. G. Hodge, Ind. Res.
Dev., 25(7), 82 (1983).
296. Wheeler, R. N., Jr., Environ. Health Perspect., 41, 123 (1981).
297. Minot, J. D., O'Mara, M. M., and Bowles, R. L., Chem. Eng. Prog., 69(8), 71 (1973).
298. Crider, L. B., Encyclopedia of PVC, Vol. 3, L. I. Nass, ed., Marcel Decker, New York,

1977, Chap. 32.

299. Engineering Science, Inc. CMA/EPA Five-Plant Study, Report for the Chemical
Manufacturers Association, Washington, D.C., Apr. 1982.
300. Tabak, H. H., Quave, S. A., Mashini, C. I., and Barth, E. F., J. Water Pollut. Control
Fed., 53, 1504 (1981).
301. Patterson, J. W., and Kodulka, P. S., Chem. Eng. Prog., 77(4), 48 (1981).
301a. Environmental Protection Agency, Development Document for Effluent Guidelines and
Standards for the Organic Chemicals and Plastics and Synthetic Fibers, EPA 440/183009,
Vols. I-III, Feb. 1983.
301b. Environmental Protection Agency, Removal of Volatile Organic Contaminants from
Ground Water, EPA 600/D-83011, PB 83168617, NTIS, Springfield, Va., Feb. 1983.
301c. Environmental Protection Agency, Treatment Compatability of Municipal Waste and
Biologically Hazardous Industrial Compounds, EPA 600/282075, Nov. 1982. See also
Treatment of Volatile Organic Compounds in Drinking Water, EPA-600/883019, PB 83
239434, May 1983.

Page 307
301d. Zhu, C. L., Yuang, C.-W., Fried, J. R., and Greenberg, D. B., Environ. Prog., 2, 132
(1983).
302. Scharein, G., Hydrocarbon Process., 58(9), 193 (1981).
303. McPherson, R. W., Starks, C. M., and Fryar, G. J., Hydrocarbon Process., 56(3), 75
(1979).
304. Stauffer Chemical Co., Hydrocarbon Process., 58(11), 237 (1981).
304a. Keck, D. R., Paper 68f, 83rd Natl. Meet. Am. Inst. Chem. Eng., Mar. 24, 1977.
305. Lee, K. C., Morgan, N., Hansen, J. L., and Whipple, G. M., Paper 825.3, 75th Meet.
Air Pollut. Control Assoc., New Orleans, June 2025, 1982.
305a. Environmental Protection Agency, VOC Fugitive Emissions in Synthetic Organic
Chemicals Manufacturing Industry. EPA-450/380033b, June 1982. See also 48FR57549.
305b. Waterland, L. B., Pilot-Scale Investigation of Surrogate Means of Determining POUC
Destruction, prepared for the Chemical Manufacturers Association by Acurey Corp., July 1983.
306. Kaiser, E. R., and Carotti, A. A., Report to the Society of the Plastics Industries, Inc.,
New York, June 30, 1971.
307. DeBell and Richardson, Inc., Plastics Waste Management, Report for the Chemical
Manufacturers Association, Washington, D.C., Oct. 1976.
308. Environmental Protection Agency, Collaborative Testing of EPA Method 106, EPA-600/4
78058, Oct. 1978, PB 298775.
309. Berens, A. R., Angew. Makromol. Chem., 47, 97 (1981); Polym. Sci. Eng., 20, 95
(1980).
309a. Olsen, R., Anal. Chem., 53, 929 (1981).
310. Severs, L. W., and Scary, L. K., Am. Ind. Hyg. Assoc. J., 36, 669 (1975).
311. Myers, S. A., Quinn, H. J., and Zook, W. C., Am. Ind. Hyg. Assoc. J., 36, 333 (1975).
312. Snell, J., Pollut. Eng., 14(6), 36 (1982).
313. Kimble, H. J., Ketcham, N. H., Kuryla, W. C., Neff, J. E., and Patel, M. A., Am. Ind. Hyg.
Assoc. J., 43, 137 (1982).
313a. Environmental Protection Agency, Method 502.1, Determination of Halogenated
Chemicals in Water by the Purge and Trap Method, Environmental Monitoring and Support

Laboratory, Cincinnati, Ohio, Apr. 1981.

314. Environmental Protection Agency, Ambient Air Carcinogenic Vapors, Improved Sampling
and Analytical Techniques and Field Studies, EPA-600/279081, May 1979, PB 297932.
315. International Agency for Research on Cancer, Environmental Carcinogens, Selected
Methods of Analysis, Vol. 2: Vinyl Chloride, IARC Scientific Publ. 22, Lyon, France, 1978.
316. Chemical Industries Association, The Determination of Vinyl Chloride-A Plant Manual,
CIA, London, 1977.
317. Wilson, J. L., Determination of Volatile Organics of Industrial and Municipal
Wastewaters, EPA-600/481071, Aug. 1981, PB 82119090.
317a. Gregorzik, H., and Bauer, U., Wasser, 60, 15 (1983).
318. Environmental Protection Agency, Response Factors of VOC Analyzers Calibrated with
Methane for Selected Organic Chemicals, EPA-600/281002, May 1981; Response Factors
of VOC Analyzers at a Meter Reading of 10, 000 PPMV for Selected Organic Compounds,
EPA-600/281051, Sept. 1981.

Page 308
319. Radian Corp., Emission Factors and Frequency of Leak Occurrence for Filterings in
Refinery Process Units, Report to the EPA under contract 62022665, 1979.
320. Burgess, R. H., ed., Manufacture and Processing of PVC, Macmillan, New York, 1981.
320a. Coutant, R. W., Laboratory Evaluation of Commercially Available Passive Organic
Personal Monitors, EPA-600/482031, Apr. 1982, PB 82234261; L. A. Wallace and W. R.
Ott, J. Air Pollut. Control Assoc., 32, 601 (1982); J. B. Perkins, N. H. Price, L. Eggenberger,
and J. A. Burkart, Evaluation of Passive Organic Vapor Monitors, Dec. 1981, PB 83221028.
320b. Bell, Z. G., Jr., Lafleur, J. C., Lynch, R. P., and Work, G. A., Chem. Eng. Prog., 71(9),
45 (1975).
321. Mayo, F. R., Walling, C., Lewis, F. M., and Hulse, W. F., J. Am. Chem. Soc. , 70, 1523
(1948).
322. Tunkel, S., J., Chem. Eng. Prog., 79(9), 50 (1983).
323. Klondike, D. W., Prof. Safety, 28(4), 17 (1983).

Page 309

6 Resin Structure and Properties

DONALD E. WITENHAFER
B. F. Goodrich Company Brecksville, Ohio
I. PVC MOLECULAR STRUCTURE

309

A. Tacticity

309

B. Molecular Weight

312

C. Chloromethyl Branches, Head-to-Head Addition, and Chain End Groups

318

D. Branching and Tertiary Chlorine

319

E. Defect Structures Related to Double Bonds

320

II. SUPERMOLECULAR STRUCTURES OF PVC

320

A. Crystallinity

320

B. State of the Glass

324

C. Particulate Structure
III. PROPERTIES OF PVC
A. Thermal Stability
B. Gelation or Fusion of PVC
C. Melt Viscosity
D. Mechanical Properties
REFERENCES

324
332
332
336
338
339
342

I.PVC MOLECULAR STRUCTURE

A.Tacticity
The chemical structure of polyvinyl chloride (PVC) can be represented as shown in Figure 1.
Because the molecular chain will never be exactly identical in both directions, each of the
carbon atoms with a hydrogen and a chlorine atom attached has four distinctly different groups
bonded to that carbon atom. Whenever this condition of four distinctly different groups bonded

to a single carbon exists, stereoisomerization will result for each of the appropriate asymmetric
carbon atoms. A convenient way to visualize this relative stereoisomerism is to represent the
carbons in the PVC chain as a planar zigzag, as shown in Figure 2. In this representation the
carbons containing the two hydrogens

Page 310

FIGURE 1 PVC chemical repeating unit.

can be thought of as being down below the plane of the paper and the asymmetric carbons
containing the hydrogen and chlorine atoms, designated with an *, are above the plane of the
paper. In this representation, chlorine (1) and (2) are on the same side of the chain and are
said to be meso (m) to each other. Chlorines (2) and (3) are on opposite sides of the chain and
said to be racemic (r) to each other. An alternative term that is sometimes used is that chlorine
(1) and (2) are isotactic to each other and (2) and (3) are syndiotactic to each other. These
representations are called diads since they involve two asymmetric carbons. The chain
configuration can also be represented by higher-order structures that involve more than two
asymmetric centers, such as triads, tetrads, and pentads, depending on whether three, four, or
five asymmetric centers are used in the representation. Obviously, as the number of
asymmetric centers increases, the number of possible representations (i.e., possible
premutations and combinations) increases rapidly. This principle can be illustrated by
considering the longer chain represented in Figure 3. Carbons (1), (2) and (3) represent an
isotactic triad (mm) (three asymmetric centers). Carbons (2), (3), and (4) represent a
heterotactic triad (mr). Carbons (3), (4), and (5) represent a syndiotactic triad (rr). Tetrads and
higher have so many combinations that they are usually represented by the m and r notation.
Thus carbons (1), (2), (3), and (4) represent an mmr tetrad. Carbons (2), (3), (4), and (5)
represent an mrr tetrad. Carbons (1), (2), (3), (4), and (5) represent an mmrr pentad. Notice
that pentads require 4 m's and r's, for representations. Another feature about these
representations is that they are independent of the chain direction. Thus an mmrr pentad is
exactly equivalent to rrmm, depending on whether you are going left to right or right to left. That
is why the triads can be represented by the three names: syndiotactic (rr), isotactic (mm), or
heterotactic (mr or rm). If the tactic placements were totally random with equal probability of an
m or r placement, we would have 50% syndiotactic (r) diads and 50% isotactic (m) diads.
However, since mr is equivalent to rm, the random, 50%, polymer would consist of 25%
syndiotactic triads, 25% isotactic triads, and 50% heterotactic triads.

FIGURE 2 Portion of a PVC molecular chain in planar zigzag conformation showing three
asymmetric carbon atoms designated with asterisks.

Page 311

FIGURE 3 Portion of a hypothetical PVC chain showing the stereochemical placement of five
asymmetric carbon atoms.
The latter concept naturally leads into a discussion of properties of certain kinds of placements
in real growing polymer chains. Every time a vinyl chloride monomer unit is added to a growing
polymer chain, it can add in either the syndiotactic (r) configuration or the isotactic (m)
configuration. If each and every addition takes place with the same probability of being r (Pr),
the addition is said to follow Bernoullian propagation statistics even if the probability of an r
placement is greater than 0.50. Remember that Pr + Pm = 1. If, on the other hand, the
probability of adding in the r configuration depends on whether the preceding addition has been
r or m, the addition is said to follow Markoffian statistics. Markoffian statistics now require two
independent probability parameters Prm and Prr (i.e., the probability of adding r to an m final
placement and the probability of adding r to an r final placement). Remember that Pmm + Prm
= 1 and Prr + Pmr = 1. In order to determine whether a polymer follows Bernoullian statistics or
some other model such as Markoffian, it is necessary to have triad, tetrad, or pentad
information in addition to the sample diad information. The diad information can be used to
calculate triad and higher percentage placements assuming Bernoullian or other model
statistics, and the calculated results can be compared with experimental findings in order to
prove or disprove the model.
Before the appearance of NMR (nuclear magnetic resonance), the tacticity of PVC was
generally determined by infrared spectroscopy using the 1428-cm1 and 1434-cm1 bands
[13]. These studies showed that PVC was predominantly syndiotactic with the probability of
syndiotactic diads varying between 0.51 at 90C polymerization temperature and about 0.80 at
-75C polymerization temperature. Today NMR has totally replaced infrared spectroscopy as
the most acceptable method for determining tactic structure because it is self-calibrating (signal
area is proportional to the number of nucleii in that state) and because triad, tetrad, and even
pentad fine structures can sometimes be determined. This permits testing of Bernoullian and
other statistical models. Initially, the NMR work focused on proton NMR [46]. Although this
method gave further information about the tactic structure of PVC, the signals were somewhat
overlapping and required curve resolving and computer fitting to resolve signals. Results were
somewhat speculative.
The difficulties associated with proton NMR have now all but been overcome with the use of
13C NMR. Carman et al. [79] have published high-resolution, proton-decoupled 13C NMR
results showing fine structure up to pentads. A PVC polymerized by free radicals at 50C was
shown to follow Bernoullian statistics with about 0.55 probability of syndiotactic placement.
Pham et al. [10] used 13C NMR to study the effect of polymerization temperature on

syndioactivity, with the following results:

Page 312
Polymerization temperature (C)

Syndiotacticit fraction

55
25
0
-30
-50
-76

0.55
0.57
0.60
0.64
0.66
0.68

The results of Pham et al. [10] are in good agreement at high temperatures with the infrared
results of Germer et al. [2] and Nakajima et al. [3] reported earlier but show less syndioactivity
than the infrared method at low polymerization temperatures. One possible explanation for this
discrepancy might be found in the findings of Wilkes [11], who showed that low-temperature
polymerized PVC tends to crystallize or aggregate in solution and that this can lead to not
observing the more syndiotactic portion of the sample in NMR measurements. Another possible
explanation has been offered by Maddams and Tooke [12], who showed that the 1430-cm1
infrared doublet actually consists of three absorptions and not two, thus rendering the infrared
method questionable.
From the published work to date we can conclude that normal commercial suspension and
mass PVC (40 to 70C) is slightly syndiotactic (0.52 to 0.56) and follows random (Bernoullian)
statistics. Lower polymerization temperatures increase the syndiotactic content substantially.
13C NMR is the currently preferred method for measuring tacticity.
B.Molecular Weight
As a PVC molecule is growing via the free-radical addition process with normally used
polymerization rates, the molecular weight is almost totally dependent on the polymerization
temperature. This phenomenon can be explained by considering the three processes that
control the molecular weight of free-radical polymerization: chain propagation, chain transfer to
monomer, and chain termination by disproportionation or by coupling. These processes may be
schematically represented as:

Page 313

In the case of PVC, k5[R]2 is small and gets large enough to become important only if an
unusually high level of initiator is used in order to increase [R]. For normal initiator levels and
polymerization rates, only chain transfer and propagation need be considered. The molecular
weight is then dependent on the relative rates of these processes:

where DP is the average degree of polymerization.

Thus the molecular weight will depend only on polymerization temperature, because k1 and k2
are functions of temperature. Assuming that k1 and k2 behave in the normal activation eneryArrhenius manner, Kuchanov and Bort [13] obtained a value of 9.2 x 103 exp(7400/RT) for
k1/k2. Figure 4 is a plot of the degree of polymerization expressed as number-average
molecular weight versus polymerization temperature in the normal PVC commercial range 50 to
70C [14].
The PVC polymer will not consist of chains of only one length but will contain a distribution of
chain lengths depending on the actual individual random events for each growing chain. The
molecular weight distribution can be represented by different kinds of averages. The commonly
used averages are the number-average molecular weight and the weight-average molecular
weight:

Page 314

where Ni is the number of molecules of molecular weight Mi.

Number-average molecular weight can in theory be measured using any colligative property of
a polymer solution, such as vapor pressure reduction, boiling point elevation, or freezing point
depression, but for high-molecular-weight polymers, only osmotic pressure through a
differentially permeable membrane produces an effect strong enough to be measured
accurately. Similarly, the weight-average molecular weight can be measured using light
scattering. A Zimm [15] plot, which involves a double extrapolation of the scattering data to zero
concentration and zero scattering angle, gives a value from which the weight-average molecular
weight can be calculated.
A simpler and much more convenient method to get information about molecular weights and
distributions is gel permeation chromatography (GPC). With this method, a solution of the
polymer is eluted through a packed column of a material having a network structure of size
similar to but larger than the molecular sizes of the eluted polymer. The colum material is
usually swollen, cross-linked, polystyrene gel of different cross-link density, depending on the
molecular dimensions of the PVC to be analyzed. The high-molecular-weight PVC cannot get as
close to the gel network as the low-molecular-weight PVC because of its larger size. The highmolecular-weight fraction is therefore swept through the gel faster because the solvent velocity
is greater away from the fixed molecular gel structure. The PVC concentration eluting from the
column is usually detected using refractive index. A plot of concentration versus elution volume
represents the molecular weight distribution. It should be pointed out that it is necessary to
calibrate this technique using

FIGURE 4 Effect of polymerization temperature on the number-average molecular weight of

PVC in the normal commercial polymerization temperature range of 50 to 70C.

Page 315
osmotic pressure and light scattering as the primary standards. GPC has been shown to be
able to compute both Mn and Mw when properly calibrated [1619]. Normal PVC has been
shown to have a Mw/Mn of 2.0 to 2.5, so GPC is rarely run as a quality control technique.
Dilute solution viscosity has been shown to be related to the molecular weight of PVC and is
used because of its simplicity and accuracy. These measurements involve determining the
viscosity increase caused by the addition of small amounts of polymer to solvent. The method
has become very popular and has resulted in many different test procedures and data
representations.
Table 1 shows some definitions. The ASTM standard uses 0.2% solution in cyclohexanone and
expresses the result as inherent viscosity [20]. The ISO standard uses 0.5% solution and
reports the result as viscosity number [21]. Many European manufacturers prefer to express
results in terms of K value. Table 2, based on results published by Matthews and Pearson [22],
compares the various tests and methods of expression.
One caution needs to be expressed on all molecular weight determinations. A substantial
amount of evidence has been published showing that PVC tends to aggregate in solution. The
exact nature of this aggregation is not
TABLE 1 Commonly Used Dilute Solution Viscosity Terms and Definitions

Page 316
Viscosity
number
ISO/ R1741961(E)
1 50
2 52
3 54
4 57
5 59
6 61
7 64
8 67
9 70
1073
1177
1280
1383
1487
1590
1694
1798
18102
19105
20109
21113
22117
23121
24125
25130
26134
27138
28142
29145
30149
31153
32157
33161
34165
35169
36173
37177

Inherent
viscosity ASTM
D1243- 58-T
(Method A)

Specific
viscosity ASTM
D-1243- 58-T
(Method B)

Relative viscosity
K value 0.5 g/
0.5g/ 100ml
K value 1% 100 ml
ethylene dichloride cyclohexanonecyclohexanone
at 25C
at 25C

0.42
0.44
0.47
0.49
0.52
0.55
0.57
0.60
0.62
0.65
0.67
0.70
0.73
0.75
0.78
0.80
0.83
0.85
0.88
0.91
0.92
0.95
0.98
1.01
1.03
1.06
1.08
1.11
1.13
1.16
1.18
1.21
1.23
1.26
1.28
1.30
1.33

0.155
0.165
0.175
0.185
0.195
0.206
0.217
0.228
0.239
0.25
0.264
0.275
0.285
0.3
0.31
0.32
0.33
0.34
0.36
0.37
0.38
0.39
0.40
0.41
0.43
0.44
0.45
0.46
0.47
0.49
0.50
0.51
0.53
0.54
0.56
0.57
0.58

1.216
1.227
1.237
1.247
1.258
1.269
1.280
1.292
1.304
1.316
1.329
1.342
1.355
1.369
1.383
1.397
1.412
1.427
1.443
1.458
1.474
1.491
1.508
1.525
1.543
1.562
1.581
1.60
1.62
1.64
1.661
1.682
1.704
1.726
1.749
1.772
1.796

47
49
51
52
53
54
55
57
58
59
60
61
62
63
64
65
66
67
68
69
70
70.5
71
72
73
74

45
46
47
48
49.3
50.5
51.5
52.7
53.9
55
56.1
57.2
58.3
59.5
60.6
61.9
62.9
64
65.2
66.3
67.4
68.5
69.7
70.8
72
73.1
74.2
75.3
76.5
77.7
79
80.3
81.7
83.1
84.8
86.3
88

38181
39185

1.35
1.38

0.6
0.61

1.821
1.847

89.8
91.8

Page 317
Specific
viscosity
1% cyclohexa-none
at 25C
1
2
3
4 0.56
5 0.59
6 0.62
7 0.65
8 0.68
9 0.70
100.74
110.77
120.80
130.83
140.86
150.89
160.92
170.95
180.98
191.01
201.04
211.07
221.1
231.13
241.16
251.19
261.22
271.25
281.28
291.31
30
31
32
33
34
35
36
37

Specific
Specific
viscosity
viscosity 0.2
0.4%
gl 100 ml
nitrobenzene nitrobenzene
at 25C
at 20C
0.23

0.24

0.25

0.26
0.097
0.27
0.099
0.29
0.1
0.3
0.105
0.31
0.11
0.32
0.115
0.33
0.12
0.34
0.125
0.35
0.132
0.36
0.140
0.37
0.145
0.38
0.15
0.39
0.155
0.41
0.162
0.43
0.168
0.44
0.175
0.46
0.18
0.48
0.185
0.49
0.19
0.5
0.195
0.51
0.202
0.52
0.21
0.53
0.215
0.54
0.22
0.55
0.225
0.56
0.23

Specific
ApproximateApproximate
viscosity 0.2 gl Polymerizationweightnumber100 ml
degree JIS average
average
cyclohexanone K6721
molecular molecular
at 30C
weight
weight
0.08
275
40,000
20,000
0.09
310

0.10
350

0.105
380
54,000
26,000
0.11
415

0.115
450

30,000
0.120
495

0.125
525

0.13
560
70,000
36,000
0.14
600

0.145
640

40,000
0.15
680

0.155
720

0.16
760
100,000
45,500
0.17
800

0.175
840

50,000
0.18
885

0.19
930

0.195
975
140,000
55,000
0.20
1025

0.205
1070

0.21
1120

60,000
0.22
1175

0.225
1230
200,000
64,000
0.23
1300

0.235
1350

0.24
1420

70,000
0.25
1490

0.255
1570
260,000
73,000
0.26
1650

0.27
1720

0.275
1810

0.28
1900

80,000
0.29
1980
340,000
82,000
0.295
2070

0.30
2170

0.31
2260

38
39

0.315
0.32

2360
2460

480,000

90,000
91,500

Page 318
totally known, but it does seem to be more predominant with polymers prepared at low
temperature and therefore would seem to be associated with some kind of precrystallization of
syndiotactic sequences. For example, Daniels and Collins [23] have shown abnormal (curved)
Zimm light-scattering plots for polymers polymerized below 5C but normal plots for polymers
polymerized above 20C. Abdel-Alim and Hamielec [24] have used GPC to study aggregation
of low-temperature-polymerized PVC. They found that heating PVC solutions at 200C
disintegrated the aggregates. Thus heating PVC solutions prior to molecular weight
determination and appropriate care, especially with higher-syndiotactic-content PVCs, is
essential for correct results.
C.Chloromethyl Branches, Head-to-Head Addition, and Chain End Groups
The first question that might be asked is: Why group these seemingly unrelated structures
together in one section for discussion? The answer is that one unified series of chemical
reactions seems to be responsible for the formation of all these structures. A great amount of
evidence about the defect structures of PVC has been accumulated recently as the result of
studies involving the reduction of PVC with tributyltin hydride or tributyltin deuteride, followed by
13C NMR analysis of the resulting polyethylene or polyethylene deuteride depending on
whether the hydride or dueteride is used [25, 26]. Chloromethyl branches in the range 2 to 4
per 1000 monomer units have been reported [2729]. This feature is of the form

The most frequent saturated end group has been shown to be [2732]

The content is about 0.8 to 0.9 per molecule [29].

The most common unsaturated end group is [2931]
CH2 CH=CH CH2Cl
The content is about 0.7 per molecule [29]. A unifying scheme has been proposed to explain
these end groups and the formation of the chloromethyl branches [30, 33, 34]. The process
begins with a head-to-head addition of vinyl chloride followed by chlorine migration:

Page 319
If species (1) adds a monomer unit in the normal manner, a head-to-head defect in the chain
will occur. Head-to-head defects have not been found, so species (1) must rapidly rearrange to

If species (2) adds a monomer unit, a chloromethyl branch results:

This reaction must take place fairly regularly because of the relatively high concentration of
chloromethyl branches observed in PVC. If species (2) does not add a monomer rapidly
enough, it rearranges to species (4) and (5).

The chlorine free radical can now react with monomer to start a new chain with the other
experimentally established end group:

The last two steps show the detailed chemistry associated with the chain transfer-to-monomer
step that was shown to control the molecular weight of PVC in the preceding section.
D.Branching and Tertiary Chlorine
Just as reduction with tributyltin hydride and tributyltin deuteride and 13C NMR analysis of the
product has been useful in identifying chloromethyl branches, it is even more useful for
identifying longer branches in PVC.

Page 320
Branching in PVC has now been identified as consisting of butyl branches and long-chain
branches. Starnes et al. [35] have reported that a PVC polymerized at 100C contains 1.8 butyl
branches per 1000 monomer units and 0.6 long-chain branches per 1000 monomer units.
Hjertberg and Sorvik [29] report total branch content (butyl and long chain) in a 55Cpolymerized PVC at about 0.9 per 1000 monomer units. They did not have enough signal to
separate butyl and long-chain branches successfully.
The question of whether the branching points have tertiary chlorine or tertiary hydrogen has
been the subject of some controversy in the past but now seems to be resolved in favor of
tertiary chlorine. If the butyl branches result from a back-biting mechanism as proposed by
Bovey and Tiers [36], the fifth carbon back from the growing chain end is CHCI, and since the
butyl branch does not terminate with a -CHCl2 unit, the presence of tertiary chlorine associated
with butyl branches is indicated. Early work by Caraculacu et al. [37, 38] showed that tertiary
chlorine in copolymers with 2, 4-dichloropentene-1 could be readily detected by phenolysis.
However, no tertiary chlorine was found in PVC. Recently, Caraculacu et al. [39] have used a
more sensitive ultraviolet analysis of the phenolysis reaction product rather than the less
sensitive infrared method of the earlier work. They now report labile chlorine in the range 0.4 to
1.9 per 1000 monomer units. Hjertberg and Sorvik [29] show that the butyl branches contain
tertiary chlorine based on their reductions with tributyltin deuteride [29]. They also suggest that
the major portion of the long-chain branches contain tertiary chlorine and result from abstraction
of hydrogen from -CHCl- units of an existing chain by a growing macroradical.
E.Defect Structures Related to Double Bonds
In an earlier section, a discussion of unsaturated end groups was presented. Unsaturation can
also occur along the chain. Boissel [40] has used bromination to measure total unsaturation in
PVC. Michel et al. [41, 42] and Hjertberg and Sorvik [43] have used ozonolysis to cleave the
double bonds in PVC. The molecular weight reduction can then be related to the internal double
bonds.
These studies show that PVC contains about 0.1 to 0.2 internal double bonds per 1000
monomer units and about 10 times that amount of total double bonds depending on molecular
weight. Double bonds have an allylic chlorine associated with them.

II.SUPERMOLECULAR STRUCTURES OF PVC

A.Crystallinity
Several techniques have been used to study the crystalline nature of polymers. These
techniques include electron and x-ray diffraction, vibrational spectroscopy (infrared and
Raman), thermal analysis such as calorimetry or differential thermal analysis (DTA), and density
determination methods. It should be remembered that, in general, each method examines
somewhat different aspects of the crystalline nature of the polymer. Each method has its
individual strengths and weaknesses and because the methods look at different aspects of the
crystallinity, each method might tend to call different levels of order crystalline. In other words,
in a molecular system with varying degrees of order, each technique will draw the line between

crystalline and noncrystal-line at a different position on the order spectrum. It is therefore

important

Page 321
to remember what each technique is actually measuring. Also, the level of order in any given
sample is a very strong function of thermal history. When this is understood, it is not too
surprising that the amount of PVC crystallinity has been reported to vary widely from polymer to
polymer as well as between measurement methods.
X-ray and electron diffraction require that a high level of long-range order be present so that
large (on a molecular scale) planes of atoms are positioned correctly to give positive
reinforcement of the scattered radiation by the well-known Bragg diffraction theory. Thus
diffraction will draw the line between crystalline and noncrystalline at a relatively high level of
order and we would expect lower values of crystallinity to be associated with this technique
than with any of the other methods.
Vibrational spectroscopy can be used to assess crystallinity-related information [44]. In some
polymers, such as polyethylene, there are infrared absorptions at 720 to 730 cm1 that result
from crystal field splitting. These bands result from specific interactions between chains in a
crystal lattice. Purists would argue, with some justification, that only these bands should be
used to determine crystallinity. Unfortunately, crystal field splitting is rare and other types of
absorption bands can still be useful for understanding solid-state structure. These bands are
sensitive to the conformation of the chains. As a chain crystallizes, its conformation becomes
more regular and results in increased absorption for the crystalline conformation. Long-range
order might not be required with this technique, so it would be expected to draw the line
between crystalline and amorphous at a lower degree of order than that drawn by diffraction
methods.
Calorimetry or DTA analysis looks at the total enthalpy of the system. Ordering is a lowerenergy state and the total heat associated with breaking up any chain-chain associations or
creating higher-energy chain conformations (random order) is measured with this technique.
The technique is therefore sensitive to the total spectrum of order in the polymer and would be
expected to draw the crystalline-noncrystalline line even further than vibrational spectroscopy
from the highly ordered state.
Density measurements also draw the line fairly far from the ordered state. The basic idea here
is that order is more compact or higher density than disorder and the density generally
increases with crystallization. Changes in density can be related to order development.
Each of the crystallinity methods suffers from several drawbacks. Each must assume some
stadard state for zero crystallinity. In the case of x-ray diffraction, the scattering pattern of the
noncrystalline background must be assumed or determined and then it is assumed that it does
not change. The vibrational spectroscopy techniques have unknown extinction coefficients and
require some assumption about the relative amorphous conformations for an accurate
percentage crystallinity determination. The thermal methods require knowing the heat of fusion
for the crystalline state. Density methods require knowledge of the crystal density and the
amorphous density. Perhaps our best resolution of this situation is to realize that the basic
model of a mixture of pure highly ordered crystals dispersed in a matrix of totally unstructured

amorphous polymer is starting to be shown to be a fairly poor model for many polymers.
X-ray diffraction studies of Natta and Corradini [45] showed that the crystal structure of PVC is
orthorhombic (with a, b, and c dimensions of 10.6, 5.4, and 5.1, respectively). There are two
chains per unit cell and the chains are syndiotactic in the planar zigzag conformation. The
background

Page 322
or amorphous x-ray scattering of PVC is fairly complex in nature [46]. Typical values for x-ray
crystallinity are [47]:
Polymerized at 90C
67%
Commercial polymers
810%
Polymerized at 5C
15%
Polymerized at 75C
30%
Urea canal PVC
65%
Vibrational spectroscopy has been used to characterize the crystallinity in PVC. Early workers
[4850] ratioed either the 638-cm1 or the 601-cm1 infrared band to the 690-cm1
infrared band as a crystallinity index. The 638-cm1 and 601-cm1 bands represented long,
planar zigzag syndiotactic conformations. Witenhafer [51] and Koenig and Druesdow [52]
reported full spectral results on the infrared and Raman spectrum of PVC as related to
crystallinity. Vibrational spectroscopy generally supports the x-ray findings.
Early thermal analyses of PVC crystallinity were reported by Maron and Filisko [53] and Illers
[54]. Maron and Filisko used microcalorimetry to measure heats of solution. and heats of
dilution for PVC dissolved in cyclohexanone and tetrahydrofuran. They concluded that PVC is
about 20% crystalline. Illers also used normal DTA-type measurements to verify the presence
of crystallinity in PVC. Chartoff et al. [55] have interrelated infrared spectroscopy, x-ray
diffraction, and density data on the same samples.
In 1960, White [56] reported that vinyl chloride could be polymerized in the solid state inside
urea canal complexes at 78C using gamma radiation. This produced a high-syndiotactic,
high-crystallinity PVC that has subsequently been very useful as a model compound to better
understand PVC. This urea canal PVC was not soluble in known solvents and did not melt even
at 400C. In our laboratory we have verified that urea canal PVC is not soluble even in
superheated (200C) cyclohexanone. In fact, it is not even swollen. Thus the melting point of
large perfect PVC crystals must be very high indeed, well above decomposition temperatures.
Smith and Wilkes [57] have reported that very low molecular weight PVC with a degree of
polymerization of about 40 can be grown into single crystals from 2% chlorobenzene solution.
Figure 5 shows an electron micrograph of PVC single crystals. Two types are shown. The
electron diffraction pattern is from the flat platelet crystal. The other crystal is nucleated in a
manner to produce a spiral growth. The crystal structure of these crystals was subsequently
verified [58] by x-ray diffraction to be close to the structure of Natta and Corradini except that
the a dimension was somewhat reduced, 10.24 . Presumably the very low molecualr weight of
the PVC allows a fractional separation of the higher syndiotactic chains into the single crystals.
If the available information about the crystalline nature of PVC is summarized, it makes an
interesting picture. The PVC industry is very fortunate that normal polymerizations of PVC are
only slightly syndiotactic. Otherwise, PVC would resemble urea canal complex PVC and would
be almost totally un-processable. We are fortunate indeed that the isotactic placements break
up the crystallinity and keeps it to a low level. Some low level of crystallinity in PVC is essential

for plasticized applications. The crystallinity makes plasticized PVC behave as if it had crosslinks and thereby provides elasticity to the material. Without this elasticity, a plasticized PVC
raincoat would slowly sag and stretch to the floor when hung on a hook a useless material
indeed. The crystal structure of commercial PVC has been shown by both infra-

Page 323

FIGURE 5 Electron micrograph of PVC single crystals with the electron diffraction pattern from
the indicated flat platelet crystal. Also shown is a spiral crystal. (Courtesy of R. W. Smith, B.F.
Goodrich; reprinted by permission of John Wiley & Sons, Inc.)

Page 324
red [51] and thermal (DTA) analysis [54] to melt over a wide temperature range, from about
105C to 210C. Crystals can have a wide melting range because of different degrees of
perfection (low perfection-low melting) or because the crystals are small and of varying size
(small size-low melting). There is substantial evidence that PVC melts over a wide range for
both reasons. An analysis by Juijn and coworkers [59] shows that normal PVC with Bernoullian
arrangement of the syndiotactic placements could not be as crystalline, as has been
determined without incorporating a substantial amount of isotactic material into the lattice. Also,
x-ray line broadening [46, 60, 61] and small-angle x-ray diffraction [61, 62] seem to indicate
small crystallite sizes. Wenig [61] proposed a lamellar crystal structure, and Blundell [62]
proposed a more spherical structure, which is in closer agreement with electron microscopy
results [63, 64]. Crystals with varying degrees of order are also consistent with the x-ray
diffraction results of Mammi and Nardi [65] and others [6668]. The large difference in melting
temperature between urea canal PVC and normal commercial PVC is consistent with a high
level of crystalline disorder in normal PVC. It has also been shown [51, 54] that annealing PVC
in the melting range 105 to 210C produces more perfect or larger crystals that subsequently
melt somewhat above the annealing temperature. Either a melting and recrystallization or some
other annealing process must be occurring at these temperatures.
B.State of the Glass
In addition to the crystalline nature of PVC, we are beginning to realize that additional molecular
structure associated with the nature of the glassy state is important. Commercial PVC has its
glass transition temperature at 80 to 85C. The glass transition temperature increases as the
polymerization temperature decreases [69]. As the PVC cools through this temperature, the
molecular motion is reduced to a rate where the molecules can no longer change conformation
rapidly enough to follow the equilibrium conformation that the drop in temperature would dictate.
The glass at room temperature will then contain nonequilibrium conformational structure
associated with a higher temperature. The rate at which the material is cooled through the
glass transition will establish the glassy state of the PVC. Illers [54] observed this phenomenon
in his pioneering work with DTA in 1969. He found that annealing below the glass transition
produced measurable changes in the thermogram, which he interpreted as changes in the hole
concentration in the glassy state. In a later work he expanded on these concepts [70]. Also,
Maron and Filisko [53] measured a heat of the glass associated with glassy enthalpy when they
did their microcalorimetry heat of solution and dilution experiments. Berens [71] proposed that
vapor solubility measurements could be used to measure the free-volume state of glassy
polymers. Recent work in this area has been published by Berens and Hodge [72, 73]. This
area of polymer research offers a great deal of potential for further understanding.
C.Particulate Structure
The next larger level of structure in PVC is the structure of the as-polymerized PVC particle
(either mass PVC or suspension PVC). Knowledge of the particle structure of PVC is important
to both flexible (plasticized) and rigid PVC applications. In the rigid PVC case, we need to
remember that we are generally processing PVC at temperatures between 160 and 200C.
PVC is

Page 325
thus processed within its crystalline melting range and the presence of unmelted crystallinity at
these temperatures can and does cause vestiges of the original particle structure to persist
right through into the finished product. In the flexible PVC case, the particles must be capable
of picking up larger quantities of plasticizer rapidly during powder mixing, and the particulate
structure plays a major role in this plasticizer pickup. Flexible PVC is also processed in the
crystalline melting range and the same basic considerations apply as in the rigid PVC case. It is
beyond the scope of this chapter to go into detail about the influence on particulate structure.
These topics are covered in Chapter 3, which describes the various processes for
manufacturing PVC. The purpose of this section is to introduce the reader to an overview of the
particulate structure of PVC so that it can be related to properties in the next section.
1. Description and Origin of Particulate Structure: PVC resins are normally manufactured so
that their average grain size is between 90 and 200 m in diameter. This size is a most
convenient compromise between the dusting hazards and handling difficulties associated with
resins having a small particle size and the polymerization difficulty of preventing colloidal
instability and coarse and solid batches associated with large particle sizes. Figure 6 shows a
scanning electron micrograph of a typical PVC particle. If the par-

FIGURE 6 Scanning electron micrograph of a single-suspension PVC particle polymerized from

one drop of vinyl chloride.

Page 326
ticles are sliced open as shown in Figure 7, the interior is seen to be composed of
agglomerates of many small 1-m particles called primary particles. It is the volume between
the primary particles that gives PVC resin its porosity, which in turn allows PVC to be easily
compounded with lubricants, plasticizers, and stabilizers. The reason primary particles form is
that PVC is insoluble in its own monomer. As the polymerization proceeds, PVC precipitates
and flocculates to form the primary particles. There is substantial evidence that the primary
particles are in turn composed of aggregates of even smaller particles [7478]. The nature of
these subparticles has been more difficult to establish because of the difficulty of observation
as the size gets smaller. Even the terminology used in the technical literature is confusing since
authors often use their own terminology to describe particulate structure. During the Second
International Symposium on PVC in 1976 at Lyon, a discussion of terminology was held and the
agreed-upon terminology was subsequently published by Geil [79]:

Term

Approximate
Origin or description
size

Grain
100
Agglomerate10
Primary
1
particle

m
m

Free flowing at room temperature

Formed during polymerization by merging of primary particles

Formed from single polymerization site at conversions of 1050%

Domain

nm

100

Microdomain10

nm

Presence not clearly proven, possibly formed by mechanical working

within or from primary particles
Crystallite or nodule

Based on this terminology, the subprimary particles are called domains or microdomains,
depending on their size. Allsopp [80] believes that domains grow into primary particles during
polymerization and therefore exist only at low conversion rates. However, Geil and associates
have identified domain-size structures in plasticized PVC [63, 81]. 8 is an ultrathin-section
transmission electron micrograph showing PVC primary particles composed of microdomains.
In addition to this internal structure, which seems to be common to both mass and suspension
resins, suspension resins have been shown to contain a thin 1-m pericellular membrane [82
84] surrounding the grain as shown in Figure 9. We now believe that this membrane is formed
from early PVC particles that migrate to the vinyl chloride-water interface. These particles can
originate from both the vinyl chloride phase as well as from the water phase, where vinyl
chloride can graft onto dispersant to form particles. Another feature of suspension resins is that
under specific agitation and surfactant recipes, individual grains can consist of subgrains [80].
Subgrains are almost

Page 327

FIGURE 7 Close-up scanning electron micrograph of a sliced-open PVC particle similar to that
in Figure 6. The outer pericellular membrane as well as the internal porosity and primary
particles may be seen.
certainly early droplets of vinyl chloride that aggregate during the polymerization after the
pericellular membrane has been formed. Figure 10 shows a particle composed of subgrains.
Davidson and Witenhafer [84] showed that some suspension recipes produce a multiple
emulsion in which droplets of water are dispersed in the vinyl chloride and result in particles
containing a large void and a membrane inside the PVC particle as shown in Figure 11. The
foregoing features are not found in mass resins, but some mass resins have been shown to
have a porosity gradient from the interior (high) to the exterior (low) of the particle [80]. It is
appropriate that we now discuss some of the tests used to characterize PVC resin particulate
structure.
2. Particle Size and Distribution Measurement Methods: There are many methods that could be
used to determine average grain size and particle-to-

Page 328

FIGURE 8 Ultrathin transmission electron micrograph of primary particles of PVC showing

microdomains. (Courtesy of M. W. Allsopp, Imperial Chemical Industries Petrochemicals and
Plastics Division; reprinted by permission of Applied Science Publishers Ltd.)
particle size distribution. Microscopy, followed by individual particle-size measurement and
counting, is one possibility, but this is very time consuming because of the large number of
particles that must be counted. New electronic image analysis and counting techniques do offer
hope to make this method reasonable. Various screening or sieving methods can be used
ranging from wet to dry and from hand shaken to sonically vibrated. Potential problems with
these methods are static electricity buildup, particles not wet in wet screening, and screen
blinding in which the screen holes get plugged with particles wedged in place which do not allow
smaller particles to pass. The screening methods are, however, very fast and economical in
addition to being fairly reproducible. Various instrumental particle counting methods can also be
used. These methods all rely on being able to pass one particle at a time through a sizing and
counting zone. Examples of this type of instrument is a Coulter Counter [85], in which a dilute
suspension of the particles in electrolyte is passed between closely spaced electrodes and the
diameter (signal intensity) is stored in the appropriate computer memory register. Other
measuring techniques use laser diffraction from individual particles, as with the Leeds and
Northrup or Malvern particle sizer, or they use light-blockage techniques, such as with the HIAC
Corporation instrument.

Page 329

FIGURE 9 Close-up view of the outside of a sliced-open PVC particle similar to Figure 7 but
showing details of the structure of the pericellular membrane.
3. Porosity Measurement: The pore volume inside PVC particles is usually expressed in cm3/g
and can generally be measured in two ways. Mercury intrusion porosimeters [86] determine the
pressure necessary to force mercury (a nonwetting fluid) into the pores of the resin. A plot
similar to Figure 12 results in which the logarithm of pressure is plotted against the volume of
mercury that penetrates a previously evacuated PVC. It is normally assumed that the lowpressure break upward in the curve is associated with that point where the voids between
particles have been filled and the mercury starts to penetrate the much smaller pores. When
the high-pressure break occurs, the pores are assumed to be full. Some people relate the
penetrant pressure to a pore size and distribution function. Recent work by Davidson has
brought some of these assumptions into question. Davidson [87] studied the intrusion of lowmelting alloys into PVC. He could stop the intrusion process at different pressures by freezing
the alloy and then sectioning the solid alloy to see where the liquid metal (now solid) was to be
found. He found that with suspension resins, the pericellular membrane offered the limiting
resistance to penetration. An individual particle was either filled with metal or it was empty.
Thus the penetrant pressure had nothing to do with pore-

Page 330

FIGURE 10 Scanning electron micrograph of an agglomerated PVC suspension particle

showing subgrains that originally were individual vinyl chloride droplets.
size distribution. Davidson [88] also found that the voids between particles are often filled
before the upward (low-pressure) break in the volume pressure curve. The conclusion from this
work is that accurate pore volumes (porosity) can be determined for PVC by independently
determining where the pores start to be filled.
The second method for determining resin porosity involves filling the pores with a high-boiling
liquid that wets PVC, such as di-2-ethylhexyl phthalate (DOP) [89]. The excess DOP is
removed by centrifugation through a screen and the weight gain is related to the total porosity.
The potential difficulty with this method occurs when particles have a pericellular membrane so
complete that the wetting liquid fails to penetrate the particle even though it has porosity. This
situation would not interfere with mercury penetration because the high-pressure mercury would
eventually rupture the membrane.
4. Surface Area Measurement: Although the surface area is obviously highly related to the
porosity of the PVC, it is still desirable to be able to measure this resin property. Normally, gas
absorption using the Brunauer-Emmett-Teller (BET) theory [90] is used. Recently, several
researchers have

Page 331

FIGURE 11 Optical micrograph of a thin cross section of a multiple-emulsion PVC particle that
was polymerized with a drop of water inside the vinyl chloride droplet.

FIGURE 12 Typical mercury intrusion plot for PVC resins. The low-pressure penetration is
associated primarily with the filling of voids between particles, whereas the high-pressure
penetration is associated with filling the pores between primary particles inside each grain.

Page 332
proposed a simplified surface area measurement using the PVC as a gas chromatographic
column packing and determining the relative retardation time for a gas such as octane that
absorbs on PVC but does not substantially penetrate the solid polymer [9092].
5. Primary Particle-Size Measurement: If the diffusion coefficient for a PVC swelling agent such
as vinyl chloride is known and PVC particles are exposed to a known concentration of vinyl
chloride at a temperature above the glass transition temperature, so as to avoid state of the
glass complications, the rate of weight loss upon application of a vacuum to the powder can be
related to the distribution of diffusion distances in the PVC. This technique has been expanded
by Berens [93] with a gas chromatographic method and later refined by Daniels and Longeway
[94] to measure the average primary particle size of PVC resins. They observed a small weight
fraction of large glassy PVC to be present.
6. Measures of Particle-to-Particle Structural Differences: Obviously, in an ideal PVC, each
resin particle is exactly like every other resin particle. Each particle has the same porosity, the
same size, the same degree of completeness of the pericellular membrane, and so on.
Unfortunately, we do not live in an ideal world and the foregoing statement is not true. Particles
have different structures and therefore pick up plasticizers, lubricants, and stabilizers
nonuniformly. Several tests have been proposed that relate to this phenomenon. Schwaegerle
[95] screened PVC resin and ran porosity on the coarse and fine fractions. The ratio of these
porosities is a measure of porosity uniformity. Most resins have reduced porosity in the smallparticle-size fraction. Krzewki and Sieglaff [96] used a density flotation method to assess the
density distribution from particle to particle. Resins with broad density distributions were shown
to have inaccessible pores, resulting in less uniform absorption of plasticizer and less
satisfactory processing. Tregan and Bonnemayre [97] have proposed that PVC be examined
under a microscope in an excess of penetrating liquid, such as DOP, which has a refractive
index close to that of PVC. The particles or portions of particles (i.e., subgrains) will then
appear in one of three ways. If the liquid does not penetrate the particles, perhaps because the
pericellular membrane is too complete, but the particle does have porosity, then the large
refractive index difference between air and the PVC primary particles will highly scatter the light
and the particle will look black. If the portion of the particle is penetrated by the liquid, the
relatively small scattering caused by a much smaller refractive index difference will make the
particle appear opaque. If the particle has little or no porosity, there will be very little if any
penetration and it will appear clear. Thus resins can be roughly rated for uniformity using this
technique.

III.PROPERTIES OF PVC
A.Thermal Stability
The thermal degradation of PVC is considered by most experts to be a major disadvantage of
this material. Upon exposure to temperatures as low as 100 to 120C [98, 99] or upon
exposure to ultraviolet radiation [100102], PVC undergoes a degradation reaction that
releases hydrochloric acid and forms long polyene sequences of conjugated double bonds. The

dehydrochlorination reaction can be represented as

Page 333

The longer polyene sequences are intensely colored and only a small amount of them can make
the PVC appear black. Since these polyenes appear at very low levels of dehydrochlorination
[103, 104], they must result from a zipper mechanism in which the first double bond activates
the formation of the second conjugated double bond, and so on, until long polyenes are formed.
Many mechanisms that can account for this zipper reaction have been proposed, but a freeradical mechanism [105, 106] and an ion-pair mechanism [107109] seem to be preferred. The
free-radical mechanism can be represented as follows:

The main argument against the free-radical mechanism is based on the idea that the Cl free
radical would diffuse away from the growing polyene and stop the propagation [110]. Also,
model (low-molecular-weight) compound studies have shown that the allylic hydrogens at the
end of the polyenes are only slightly more reactive than normal hydrogens in PVC to chlorine
free-radical abstraction [111]. The observation that hydrochloric acid is an extremely powerful
catalyst for the further release of hydrochloric acid [112,

Page 334
113] is probably more consistent with the ionic mechanism also. Thus currently, the ionic
mechanism seems to be preferred.
1. Heat Stability Tests: It seems that there are almost as many tests for heat stability as there
are workers in the field. This section describes them in general and discusses the advantage
and disadvantage of each type of test.
Probably the most commonly used tests in scientific publications are tests measuring the
amount of HCl evolved from a known amount of PVC as a function of time at a fixed
temperature. These HCl evolution tests have the advantage of being a quantitative measure of
the rate of the degradation reaction and therefore have the possibility of being related back to
the basic chemical mechanism. The major disadvantage of this kind of test is that the HCl
released is an extremely powerful catalyst for the degradation itself. Diffusion of the HCl away
from the sample can be an important complication. Thus, if resin is being tested, the
morphology of the resin can play an important role. If the degradation is done in solution, the
solvent, impurities, and HCl buildup in the solution are additional possible complications. Many
different equipment arrangements, sample sizes, gas flow rates, and HCl detection schemes
are found in the literature.
The other class of tests generally involve the heating of the sample followed by noting the time
until some degradation event appears. The noted degradation event could be as simple as
black color detected by eye to a very complex examination of the polyene nature of the PVC
using such possible techniques as ultraviolet spectroscopy or, more recently, resonance Raman
methods [114]. Sometimes stabilizers or other compounding ingredients are used. The heating
can be static, or shearing mixing can be used. The advantage of these kinds of tests is that
they can be quite simple and relate directly to color development, which is the property of the
PVC that is of concern in the marketplace.
The disadvantage of most of these tests is that different stabilizers and compounding
ingredients sometimes have unique and specific side reactions that cloud the basic issue. Also,
the blending of non-PVC soluble stabilizers with PVC can be highly dependent on the PVC
morphology. Shear heating in the dynamic (mixing) tests is a substantial complication. The PVC
molecular weight in dynamic tests is often more important than the inherent heat stability of the
PVC itself because of shear heating. It is this author's opinion that tests to measure a PVC
resin's heat stability should be as simple as possible. A simple blend of PVC and soluble
stabilizer in a fixed geometry without mixing is recommended. If knowledge of HCl evolution is
desired, we need to find a way to convert the PVC into a small-particle-size standard
morphology. If HCl evolution from solution is to be used, standard solvents and solvent quality
need to be established. Laboratory-to-laboratory agreement on these important points has not
yet been achieved. Fortunately, however, it is still possible to draw conclusions about the
relative effects of changes in PVC molecular structure and their effect on heat stability, even
though many authors use their own standard tests. Tests for the stability of a PVC compound
can be either static or dynamic, but the PVC morphology through complex lubrication and
stabilization mechanisms and the effect of molecular weight on shear heating can often interfere

with a clear understanding of the relationship between PVC molecular structure and the
dynamic compound stability.
2. Role of Defect Structures: Low-molecular-weight model compounds for PVC have been
shown to dehydrochlorinate substantially slower than PVC

Page 335
[98, 99, 101, 103, 109] and this observation suggests that some kind of chain defect structure,
such as perhaps a tertiary chlorine, chloromethyl branches, or unsaturation (both internal and
chain end), should play a major role in initiating the zipper reaction. Recent work by Starnes
and Plitz [115, 116] supports this idea. They reacted PVC with Bu2Sn(SnC12H25)2 and
carefully purified the PVC. They showed a dramatic reduction in dehydrochlorination rate with
only a minor amount of sulfur content in the PVC. Thus the reactive sites (defects) were
presumed to be removed by this pretreatment.
The most recently published work seems to indicate that tertiary chlorine located at chain
branch points is highly suspect as the most important defect. Berens [117] reported the
dehydrochlorination rates of a number of vinyl chloride/ 2-chloroprene copolymers containing
known amounts of tertiary chlorine from the 2-chloroprene. He concluded that the
dehydrochlorination rate of normal PVC can be accounted for with 0.5 to 1.0 tertiary chlorine
per 1000 monomer units. Recently, Hjertberg and Sorvik [29] have reported a dramatic
increase in dehydrochlorination with increasing branching for a series of PVCs containing
unusually high levels of branching. They made these PVC by running the polymerization under
vinyl chloride subsaturation conditions. These subsaturation conditions caused the propagation
reaction to be slowed while not appreciably altering the chain branching reaction. These authors
could not relate thermal stability to end-group content. This conclusion is also supported by
recent work by VanDenHeuvel and Weber [118]. Thus unsaturated end groups do not seem to
be of major importance. The role of internal chain unsaturation and related allylic chlorine is not
clear, perhaps because the amount of this defect is generally quite low, as discussed earlier.
3. Other Dehydrochlorination Initiation Mechanisms: The work of Hjert-berg and Sorvik [29]
indicates that if the level of branching (tertiary chlorine) is extrapolated to zero branches, the
PVC would have a dehydrochlorination rate about one-half to one-fourth that of normal PVC.
Starnes and Plitz [115, 116] also showed that chemical removal of the defects would greatly
reduce the dehydrochlorination rate but that a measurable rate remained even after extensive
reaction. Both findings suggest that the zipper reaction can be initiated without a defect.
Starnes and coworkers [119, 120] have proposed that the nondefect initiation takes place by an
ion-pair mechanism:

A substantial amount of chemistry in support of this mechanism has been established [119,
120].
4. Termination of the Growing Polyene: Since the zipper mechanism seems to take place quite
rapidly when PVC is heated, it is quite appropriate to ask what reactions might take place to
terminate the process. Haddon and Starnes [121, 122] have proposed that the positive charge
on the chain is delocalized over the length of the polyene sequence and the chloride ion is

therefore attracted away from the growth site at the end of the polyene se-

Page 336
quence, thus slowing down the degradation process as the sequence lengthens. In other
words, it is a self-limiting process. Other possible termination reactions are intermolecular
Diels-Adler processes leading to cross-linking [123] or intermolecular ring formation by similar
chemistry:

This ring structure is later capable of splitting off aromatic species [124]. The fact that this is
primarily an intramolecular process has been established by degrading mixtures of deuterated
PVC with normal PVC. The benzene-type products contain either all hydrogens or all
deuteriums [125].
5. Other Considerations: The preceding sections applied almost exclusively to the nonoxidative,
nondynamic, nonphoto degradation of PVC. The majority of work has been done in this area
primarily because an understanding without the foregoing complications is fundamental to a
total understanding. The reader is reminded that the process can be substantially more
complicated. Structural defects caused by mechanical processing have been shown to be
significant by Scott and coworkers [126129]. Air or other oxidation of PVC itself and the
polyenes after formation has been shown to be important [98, 101103, 107, 112]. The role of
ultraviolet radiation is far from being well established. Perhaps the best summary is that it is an
extremely complicated subject but we are now starting to make headway toward better
understanding the basic mechanisms.
B.Gelation or Fusion of PVC
At this point in a chapter on the structure and properties of a more normal material than PVC,
it would be time to discuss the relationship of molecular structure to melt flow (processability)
and to physical properties such as tensile strength and impact strength. Two basic properties of
PVC prevent us from doing this, First, PVC is thermally unstable, as discussed in the preceding
section, and second, PVC is crystalline with a broad melting range (also discussed). Thus PVC
must usually be formed into the finished article (processed) while still in the partially crystalline
state. Remnants of the as-polymerized morphology can remain in the melt as well as in the
finished article to be tested. The process whereby the PVC agglomerates, primary particles,
domains, and microdomains are attached together during processing is referred to as gelation
(in Europe) or fusion (in the United States). To develop useful properties, this knitting together
must take place, and a substantial amount of research on this aspect of PVC technology has
been published recently. Before a meaningful understanding of melt rheology or physical
properties can be discussed, the fusion process must be understood.
The early work of Berens and Folt [130, 131] paved the way for a basic understanding of the
fusion process and its role in rheology. These workers studied the extrusion of compacted

emulsion PVC with a capillary rheometer

Page 337
at various temperatures. At the lower extrusion temperatures the extrudate was found to
contain particles whose size could be related to the particle size of the emulsion particles in the
original samples. As the extrusion temperature was increased, the extrudate lost its particulate
nature and the melt viscosity increased dramatically. The transition temperature between the
two flow regimes was about 180 to 200C. With normal suspension resins, the 1m primary
particle was found to be the flow unit. They also found that the extrudates containing particulate
residue had good physical properties and therefore postulated a process of molecular diffusion
between the particles as being responsible for the interparticle knitting. Summers et al. [132,
133] have shown that this fusion process takes place in plasticized PVC as well as in rigid PVC.
Samples with various levels of plasticizer were extruded over a temperature range, and were
then swollen with acetone and sheared between microscope slides. At low temperatures, clear
evidence of residual primary particles could be found, but their presence disappeared at higher
temperatures. The phenomenon was independent of plasticizer level except that the fusion
temperature decreased with increasing plasticizer level. The most probable mechanism for this
interparticle binding mechanism may well be the interparticle diffusion of molecular chain ends,
followed by subsequent recrystallization inside the other primary particle or perhaps on the
border between the two particles as the material is cooled. Thus the degree of diffusion is
associated with the degree of network formation [134]. Processing aids and lubricants are
known to affect the fusion process dramatically [135, 136].
The 100m grain structure in PVC loses its identity quite rapidly during processing. Recently,
the mechanism for this disappearance has been studied extensively. Parey and Menges [137]
report that the grains are broken down by shear and then the primary particles are fused.
Allsopp [138] has also studied this process extensively. He concludes that the breakdown or
comminution mechanism applies only to high-shear equipment such as Banbury mixers (lab and
plant scale) and laboratory Brabender mixers. Low-shear devices, such as extruders, mills, and
compression molders, work by compacting and fusing the grain internal structure (primary
particles), followed by elongation of the grains until the original grain structure is no longer
discernible.
Many types of tests are available to rank the extent of fusion in PVC. Perhaps the simplest is
the practice of pipe extrusion operators to grab some hot melt out of the die and pull it. If the
melt has elasticity (associated with a network), the fusion is probably good. If the melt breaks
or fractures, the fusion is not good. Acetone swelling tests [139] on finished pipe can roughly
indicate the level of fusion. Well-fused pipe will swell and get rubbery, whereas poorly fused
pipe will fall apart as swollen powder when immersed in acetone. Recently, there has been a
high level of interest in a capillary rheometry test developed by Gonze [140] and Lamberty
[141]. In this test a short capillary die is used, which maximizes the elastic die entrance effects
and minimizes the viscous flow effects. The more network structure in the sample, the greater
the amount of elasticity. In this test the powder compound to be studied is first fused at
different temperatures and then each sample is extruded using the die entry test. This results in
an S-shaped plot of extrusion pressure versus fusion temperature, shown schematically in
Figure 13. The sample of unknown degree of fusion is then extruded with the die entry test and

the percentage fusion is determined from the percentage distance up the pressure curve.

Page 338

FIGURE 13 Schematic plot showing a typical standard fusion curve for a PVC formulation. The
percent fusion of an unknown PVC sample with the same formulation and processing equipment
can be determined by noting its percentage distance up the force curve.
Gilbert and Vyvoda [142] have reported a thermal analysis technique that can be related to
fusion in PVC. Like Illers [54, 70], they found that an endothermic peak developed in
compounds that was a function of prior heat history. The endothermic peak showed up on the
second heating somewhat above the annealing temperature. The thermograms could thus be
related to the previous processing. temperatures and thereby can also be related to the fusion
of the compound.
Some evidence that the fusion process may be reversible has started to appear. Portingell
[143] was able to show that a sample processed at 180C in a Brabender Plasticorder could
be reprocessed in the Brabender at 135C and showed similar fusion as measured by the short
land die extrusion test as a. sample originally processed to 130C only. Electron microscopy of
fracture surfaces also supported the reversibility of fusion. It does seem entirely possible that
until the primary particles are almost totally fused, the original particle boundaries are weak
spots that can be broken down upon subsequent shearing. An alternative explanation is that
PVC melts have a natural flow unit size that is a function of molecular structure and
temperature. Only further research can resolve these questions. In the author's opinion, there is
still a substantial amount of research to be done in this area.
C.Melt Viscosity
From the preceding sections of this chapter, the reader can start to appreciate the potentially
complex nature of PVC melt viscosity measurements. At low melt temperatures (probably
below about 200C but possibly much higher for syndiotactic polymers) the very model of a
melt is questionable. Perhaps a model of tiny elastic balls that might be lubricated to flow past
one another if lubricant is present in the compound might be a more accurate flow model than a
melt. At higher melt temperatures and low syndiotacticity, and without

Page 339
lubricants, processing aids, and so on, the PVC melt might be more like a true melt. In the lowtemperature, lubricated case, the melt viscosity would be a complex function of the starting
particle morphology; type and amounts of lubricants, processing aids, plasticizers, stabilizers,
and so on; degree of fusion; and melt temperature, in addition to the normal molecular
parameters that are usually associated with melt viscosity, such as molecular weight, degree of
long-chain branching, or tacticity. A complete discussion of all these aspects is far beyond the
scope of this chapter. The reader is referred to Volume 4, Chapter 2, for further details.
The most comprehensive results of melt viscosity determinations on simple PVC formulations
were reported by Collins and Daniels [144]. They looked at the simple mix of PVC plus 2.5
parts of soluble tin stabilizer. Melt temperatures at 210C and 220C were studied. They
prepared a series of PVCs of different molecular weights using polymerization chain transfer
agents at polymerization temperatures between 5 and 70C. Thus they were able to separate
molecular weight effects from polymerization temperature effects (i.e., tacticity). They also
studied the effect of shear rates on melt viscosity.
Most high polymer melt viscosities depend on molecular weight through the 3.4 power rule
[145, 146]:

Collins and Daniels found this to be the case only at low shear rates (2.9 sec1) for PVC
polymerized above 40C. They investigated Mw between 40,000 and 300,000. The value of K
was 9.33 x 1013 (poise). As the shear rate was increased, the exponent decreased regularly
to about 0.75 at 3000 sec1, with much lower viscosities. Thus molecular weight and shear
rate both can change the apparent melt viscosity by a factor of 50. They found the melt
viscosity of the PVC polymerized at 5C to be 3 to 10 times more viscous than equal-molecularweight PVC polymerized at 40C or 70C, depending on shear rate (10 times higher at low
shear). The 40C and 70C polymers had superimposable viscosity/shear rate curves. Collins
and Daniels interpreted this polymerization-temperature effect by suggesting that the more
syndiotactic (5C) polymer had more order (crystallinity) in the melt.
Plasticizers probably play the most dramatic role in PVC melt viscosity. The melt viscosity can
be greatly reduced by plasticizers, but the mechanical properties are also greatly altered.
Plasticization is so important that there would probably be no PVC industry today except for the
plasticizer technology developed in the late 1930s that enabled this intractable, unprocessable
PVC to be made into useful items of commerce.
D.Mechanical Properties
Just as the melt flow of PVC is a complex situation, the mechanical properties situation is even
more complex perhaps because the mechanical success or failure of any given part in
commerce depends on more than one independent measurable mechanical property and each
mechanical property in turn is dependent on a large number of variables. To try to give a
complete picture for each mechanical property would be an impossible task even if the

knowledge were available (which it is not). Rather, this section will deal with some of the
general principles associated with mechanical properties and then try to

Page 340

FIGURE 14 Schematic curve showing the effect of molecular weight on most mechanical
properties of PVC. The exact nature of the plot will depend on the mechanical property
selected.
highlight some of the surprising new emerging technology that seems to be developing in
relation to mechanical properties of PVC. Specifically, the role of molecular weight, plasticizers,
degree of fusion, and effects of annealing both above and below the glass transition
temperature will be discussed.
The general relationship of physical properties to molecular weight can be represented as in
Figure 14. As the molecular weight is increased, the mechanical properties (strength) increase.
The effect is generally most pronounced for low molecular weight, where properties such as
tensile strength, elastic modulus, and impact strength are all low or perhaps even essentially
nonexistant. As molecular weight increases, almost all useful mechanical properties are
increased but with sequentially dimishing returns. The only useful property lost with increasing
molecular weight is melt flow, as described in the preceding section. Thus chosing the right
molecular weight is an optimization between mechanical properties and processability.
Plasticizers dissolve into the PVC and lower the glass transition temperature. Plasticizers can
change rigid, hard PVC with high tensile strength and low elongation into a soft, rubbery
material having reduced tensile strength but very high elongation. This ability to be plasticized
and produce a wide variety of mechanical properties is why PVC is the large-volume plastic
that it is today. Because plasticization increases melt flow, the high-molecular-weight PVCs can
be used and still maintain acceptable processability.
Although the degree of fusion is important in both plasticized (flexible) and nonplasticized (rigid)
PVC mechanical properties, it has been most extensively studied in the rigid case. Gotham and
Hitch [147] studied the fatigue resistance of a well-fused PVC relative to a poorly fused PVC.
They found that a substantially improved load could be supported under cyclic loading if the
PVC was well fused. Benjamin [148, 149] has probably reported the most information about the
role of fusion on mechnical properties. He made PVC pipe with different extrusion conditions in

order to produce different

Page 341
levels of fusion as measured by the short-land-die extrusion test. Benjamin found that those
properties associated with stiffness increase monotonically with increasing fusion, whereas
those properties associated with brittle/ductile transition behavior showed a maximum versus
the degree of fusion. He thus concludes that 60% fusion gives the best balance of properties.
Benjamin's indicated maximum in properties was not supported by the work of Lam-berty [141].
Perhaps Summers and coworkers [150] have offered a possible explanation of the differences.
They also observed a maximum in drop dart impact strength with extrusion temperature in a
twin-screw extruder, but attributed the drop in impact to lubricant failure and melt fracture
rather than overfusion. A maximum in impact resistance was reported by Menges and
Berndsten [151] for a polyethylene/polyvinyl acetate (EVA) impact-modified formulation. They
found that the notched impact strength of a PVC compound processed at 187C was four times
greater than that of the same compound processed at either 180 or 210C. Microscopic
examination of the 187C specimen showed primary particles surrounded by impact modifier
(EVA). At high temperatures the fusion was complete and the EVA existed as discrete globules.
At 180C, generally poor fusion was found. Thus the detailed morphology is also highly
important in understanding mechanical properties. This is an area where much more research
will be appearing shortly and the benefits to the PVC industry should be substantial.
Other investigators have studied the role of annealing both above and below the glass transition
temperature on the mechanical properties of PVC. Annealing above the glass transition
temperature (about 85C) is designed to change the crystallinity of the PVC, whereas annealing
quenced PVC below the glass trnaisiton temperature is designed to relax away excess free
volume that was frozen in place by the quenching process as described earlier.
Shinozaki et al. [152] studied the effect of annealing PVC and stabilizer only at 110C. They
report a 15% increase in modulus and a 40% decrease in elongation at break for samples
annealed about 20 min, thus proving the importance of crystallinity on mechanical properties.
They do not indicate enough details about sample preparation to estimate the level of fusion
unfortunately. It might be speculated to be quite high since the samples were injection molded,
a process that usually requires high temperatures.
Even more interesting are the results on the role of the state of the glass on the mechanical
properties of PVC [70, 153, 154]. Illers [70] reports a nearly 400% increase in Young's modulus
for a quenched sample subsequently annealed at 60C for less than 2 days. Retting [153]
showed substantial changes in both elongation and tensile at break for different annealing times
below the glass transition temperature. Stuick [154] showed that this annealing relaxation can
take place even at 20C (but at a much slower rate). He studied the effect of aging (annealing)
quenched PVC up to about 3 years at 20C on creep properties. Substantial changes were
reported.
From the results reported above, it is quite clear that a great many variables ranging from the
degree of crystallinity and the state of the glass to the detailed morphology of the specimen, as
determined by the compounding ingredients, the starting particle morphology, and the degree of
fusion (processing conditions), can all play an extremely important role. Considering all these

results and the tremendous amount of knowledge still to be uncovered, there can be no
question that for now at least, quality control must and will be the watchword of our industry.
Until we sort out all these complications, we must do our polymerizations, compounding, and
processing exactly the same each and every time. Small changes can and do make substantial
and sometimes quite surprising differences in the outcome.

Page 342

REFERENCES
1. S. Krimm, A. R. Berens, V. L. Folt, and J. J. Shipman, Chem. Ind., 1512 (1958).
2. H. Germer, K.-H. Hellwege, and U. Johnsen, Makromol. Chem., 60, 106 (1963).
3. A. Nakajima, H. Hamada, and S. Hayashi, Makromol. Chem., 95, 40 (1966).
4. M. E. R. Robinson, D. I. Bower, W. F. Maddams, and H. Pyszora, Makromol . Chem.,
179, 2895 (1978).
5. A. H. Abdel-Alim, J. Appl. Polym. Sci. . 19, 1227 (1975).
6. F. Heatley and F. A. Bovey, Macromolecules, 2, 241 (1969).
7. C. J. Carman, A. R. Tarpley, and J. H. Goldstein, J. Am. Chem. Soc., 93, 2864 (1971).
8. C. J. Carman, A. R. Tarpley, and J. H. Goldstein, Macromolecules, 4, 445 (1971).
9. C. J. Carman, Macromolecules, 6, 725 (1973).
10. Q. T. Pham, J. Millan, and E. L. Madruga, Makromol. Chem., 175, 945 (1974).
11. C. E. Wilkes, Macromolecules, 4, 443 (1971).
12. W. F. Maddams and P. B. Tooke, Third Int. Symp. PVC, Cleveland, Ohio, 1980.
13. S. J. Kuchanov and D. N. Bort, Polym. Sci. USSR, 15, 2712 (1974).
14. R. H. Burgess, Manufacture and Processing of PVC, R. H. Burgess, ed., Applied
Science Publishers, London, 1982, p. 8.
15. B. H. J. Zimm, J. Chem. Phys., 16,. 1099 (1948).
16. R. K. S. Chan, Polym. Eng. Sci., 11(2), 152 (1971).
17. S. Grubisia, P. Rempp, and H. J. Benoit, J. Polym. Sci., B(5), 573 (1967).
18. M. Kolinsky and J. Janca, J. Polym. Sci., Chem., 12, 1181 (1974).
19. E. M. Sorvik, J. Appl. Polym. Sci., 21, 2769 (1977).
20. ASTM D1243-66, American Society for Testing and Materials, Philadelphia, 1972.
21. ISO 174-1974, International Standards Organization, Geneva, 1974.

 22. G. A. R. Matthews and R. B. Pearson, Plastics, 28(307), 98 (1963).

23. C. A. Daniels and E. A. Collins, J. Macromol. Sci.-Phys., B10, 287 (1974).
24. A. H. Abdel-Alim and A. E. Hamielec, J. Appl. Polym. Sci., 17, 3033 (1973).
25. W. H. Starnes, R. L. Hartless, F. C. Schilling, and F. A. Bovey, Adv. Chem, Ser., 169,
324 (1978).
26. T. Hjertberg and A. Wendel, Polymer, 23, 1641 (1982).
27. F. A. Bovey, K. E. Abbas, F. C. Schilling, and W. H. Starnes, Macromolecules 8, 437
(1975).
28. F. A. Bovey, F. C. Schilling, and W. H. Starnes, Polym. Prepr. Am. Chem. Soc. Div.
Polym. Chem., 20, 160 (1979).
29. T. Hjertberg and E. M. Sorvik, Polymer, 24, 673 (1983).
30. T. Hjertberg and E. M. Sorvik, J. Macromol. Sci., A17, 983 (1982).
31. R. Petiaud and Q. T. Pham, Makromol. Chem., 178, 741 (1977).
32. U. Schwenk, I. Konig, F. Cavagna, and B. Wrackmeyer, Angew. Makromol. Chem., 83,
183 (1979).
33. W. H. Starnes, F. C. Schilling, K. B. Abbas, R. E. Cais, and F. A. Bovey,
Macromolecules, 12, 556 (1979).

Page 343
34. A. Caraculacu, E. Buruiana, and G. J. Robila, J. Polym. Sci., Polym. Symp. Ed., 64, 189
(1978).
35. W. H. Starnes, F. C. Schilling, I. M. Plitz, R. E. Cais, and F. A. Bovey, Polym. Bull., 4, 55
(1981).
36. F. A. Bovey and G. V. D. Tiers, Chem. Ind. (London), 1826 (1962).
37. A. A. Caraculacu, O. Wichterle, and B. J. Schneider, Polym. Sci., C16, 495 (1967).
38. A. A. Caraculacu, E. C. Bezdadea, and G. Istrate, J. Polym. Sci., A1 (8), 1239 (1970).
39. G. Robila., E. C. Burviana, and A. A. Caraculacu, Eur. Polym. J., 13, 22 (1977).
40. J. J. Boissel, Appl. Polym. Sci., 25, 855 (1977).
41. A. Michel, G. Schmidt, E. Castaneda, and A. Guyot, Angew. Makromol. Chem., 47, 61
(1975).
42. A. Michel, G. Schmidt, and A. Guyot, Polym. Prepr., Am. Chem. Soc. Div. Poly. Chem.,
14, 665 (1973).
43. T. Hiertberg and E. M. Sorvik, Polymer, 24, 685 (1983).
44. G. Zerbi, F. Ciampelli, and V. Zamboni, J. Polym. Sci., C7, 3 (1964).
45. G. Natta and P. Corradini, J. Polym. Sci., 20, 251 (1956).
46. C. Baker, W. F. Maddams, and J. E. Preedy, J. Polym. Sci., Phys., 15, 1041 (1977).
47. W. F. Maddams, Particulate Nature of PVC-Formation, Structure and Processing, G.
Butters, ed. Applied Science Publishers, London, 1982, p. 125.
48. O. C. Bockman, J. Polym. Sci., A3, 3399 (1965).
49. A. Crugnola, M. Pegoraro, and F. Danusso, J. Appl. Polym. Sci., 6, 1705 (1968).
50. A. Crugnola and F. J. Danuso, J. Polym. Sci., B6, 535 (1968).
51. D. E. Witenhafer, J. Macromol. Sci.-Phys., B4, 915 (1970).
52. J. L. Koenig and D. Druesdow, J. Polym. Sci., A2(7), 1075 (1969).
53. S. H. Maron and F. E. Filisko, J. Macromol. Sci.-Phys., B6(2), 413 (1972).
54. K. H. Illers, Makromol. Chem., 127, 1 (1969).

 55. R. P. Chartoff, T. S. K. Lo, E. R. Harrell, and R. J. Roe, J. Macromol. Sci.-Phys., B

20(3), 287 (1981).
56. D. M. White, J. Am. Chem. Soc., 82, 5678 (1960).
57. R. W. Smith and C. E. Wilkes, Polym. Lett., 5, 433 (1967).
58. C. E. Wilkes, V. L. Folt, and S. Krimm, Macromolecules, 6(2), 235 (1973).
59. J. A. Juijn, J. H. Gisolf, and W. A. deJong, Kolloid-Z. Z. Polym., 251(7), 456 (1973).
60. A. Michel and A. Guyot, J. Polym. Sci., C33, 75 (1971).
61. W. J. Wenig, J. Polym. Sci., Phys., 16, 1635 (1978).
62. D. J. Blundell, Polymer, 20, 934 (1979).
63. D. M. Gezovich and P. H. Geil, Int. J. Polym. Mater., 1, 3 (1971).
64. C. J. Singleton, T. Stephenson, J. Isner, P. H. Geil, and E. A. Collins, J. Macromol. Sci.,
B14, 29 (1977).
65. M. Mammi and V. Nardi, Nature, 199, 247 (1963).
66. V. P. Lebedev, N. A. Okladnov, and M. N. Shlykova, Polym. Sci. USSR, 9, 553 (1967).
67. M. Gilbert and J. C. Vyvoda, Polymer, 19, 862 (1978).
68. J. C. Vyvoda, M. Gilbert, and D. A. Hemsley, Polymer, 21, 109 (1980).
69. C. A. Daniels, J. Vinyl Technol., 1(4), 212 (1979).

Page 344
70. K. H. Illers, J. Macromol. Sci.-Phys., B14(4), 471 (1977).
71. A. R. Berens, Angew. Makromol. Chem., 47, 97 (1975).
72. A. R. Berens and I. M. Hodge, Macromolecules, 15, 756 (1982).
73. I. M. Hodge and A. R. Berens, Macromolecules, 15, 762 (1982).
74. H. S. Mickley, A. S. Michaels, and A. C. Moore, J. Polym. Sci., 60, 121 (1962).
75. D. N. Bort, V. G. Marinin, A. I. Kalinin, and V. A. Kargin, Vysokomol. Soedin., A 10(11),
2574 (1968).
76. J. D. Cotman, M. F. Gonzalez, and G. C. Clover, J. Polym. Sci., A1 (5), 1137 (1967).
77. J. Boissel and N. J. Fisher, J. Macromol. Sci.-Chem., A 11(7), 1249 (1977).
78. G. Palma, G. Talamini, M. Tavan, and M. Carehza,. J. Polym. Sci., Phys., 15, 1537
(1977).
79. P. H. Geil, J. Macromol. Sci.-Phys., B14(1), 171 (1977).
80. M. W. Allsopp, Manufacture and Processing of PVC, R. H. Burgess, ed., Macmillan,
New York, 1982, Chap. 7.
81. C. Singleton, J. Isner, D. Gezovich, P. Tsou, P. Geil, and E. Collins, Polym. Eng. Sci.,
14(5), 371 (1974).
82. L. M. Barclay, Angew. Makromol. Chem., 52, 1 (1976).
83. J. E. Glass and J. W. Fields, J. Appl. Polym. Sci., 16, 2269 (1972).
84. J. A. Davidson and D. E. Witenhafer, J. Polym. Sci., Phys., 18, 51 (1980).
85. W. H. Coulter, Proc. Natl. Electron. Conf., 12, 1034 (1956).
86. ASTM D2873-70 (reapproved 1976), American Society for Testing and Materials,
Philadelphia.
87. J. A. Davidson, Powder Technol., 23, 233 (1979).
88. J. A. Davidson, Powder Technol., 23, 239 (1979).
89. ASTM D3367-75, American Society for Testing and Materials, Philadelphia; ISO 4608
1977, International Standards Organization, Geneva.

 90. J. B. Pausch, C. A. McKalen, and A. R. Berens, J. Macromol. Sci.-Phys., B20(2), 207

(1981).
91. R. Solovey, R. Cortellucci, and A. Roaldi, Polym. Eng. Sci., 14(2), 120 (1974).
92. E. Sund, E. Haanes, O. Smidrsd, and J. Ugelstad, J. Appl. Polym. Sci., 16, 1869 (1972).
93. A. R. Berens, Polym. Eng. Sci., 18, 864 (1978).
94. C. A. Daniels and G. D. Longeway, Polym. Eng. Sci., 19, 181 (1979).
95. P. R. Schwaegerle, Plast. Eng., 37(1), 42 (1981).
96. R. T. Krzewki and C. L. Sieglaff, Polym. Eng. Sci., 18(15), 1174 (1978).
97. R. Tregan and A. Bonnemayre, Plast. Mod. Elast., 23(7), 220 (1971).
98. Z. Mayer, J. Macromol. Sci.-Rev. Macromol. Chem., C10, 263 (1964).
99. G. Ayrey, B. Head, and R. Poller, J. Polym. Sci., Macromol. Rev., 8, 1 (1974).
100. H. O. Wirth and H. Andreas, Pure Appl. Chem., 49, 627 (1977).
101. E. D. Owen, ACS Symp. Ser., 25, 208 (1976).
102. W. H. Starnes, Am. Chem. Soc., Div. Org. Coat. Plast. Chem., 42, 716 (1980).
103. C. David, Comp. Chem. Kinet., 14, 78 (1975).
104. F. Tudos, T. Kelen, T. Nagy, and B. Turcsanyi, Pure Appl. Chem., 38, 201 (1974).
105. D. E. Winkler, J. Polym. Sci., 35, 3 (1959).

Page 345
106. R. Stromberg, S. Straus, and B. Achhammer, J. Polym. Sci., 35, 355 (1959).
107. W. H. Starnes, Developments in Polymer Degradation-3, N. Grassie, ed., Applied
Science Publishers, London, 1981, p. 135.
108. G. C. Marks, J. L. Benton, and C. M. Thomas, SCI Monogr., 26, 204 (1967).
109. D. Braun, in Degradation and Stabilisation of Polymers (Proc. Plenary Main Lect. Int.
Symp., 1974), G. Geuskens, ed., Wiley, New York, 1975, p. 23.
110. W. H. Starnes and D. Edelson, Macromolecules, 12, 797, 1237 (1979).
111. D. G. Hendry, T. Mill, L. Pisziewicz, J. Howard, and H. Eigenmann, J. Phys. Chem. Ref.
Data, 3, 937 (1974).
112. F. Tudos, T. Kelen, and T. Nagy, Developments in Polymer Degradation-2, N. Grassie,
ed., Applied Science Publishers, London, 1979, p. 187.
113. B. Troitskii and L. Troitskaya, Polym. Sci. USSR, 20, 1621 (1978).
114. D. L. Gerrard and W. F. Maddams, Macromolecules, 14, 1356 (1981) ibid 16: 578
(1983).
115. W. H. Starnes and I. M. Plitz, Polym. Prepr., 16, 500 (1975).
116. W. H. Starnes and I. M. Plitz, Macromolecules, 11, 373 (1978).
117. A. R. Berens, Polym. Eng. Sci., 14: 318 (1974).
118. C. J. M. VanDenHeuvel and A. J. M. Weber, Makromol. Chem., 184, 2261 (1983).
119. W. H. Starnes, R. C. Haddon, D. C. Hische, I. M. Plitz, C. L. Schosser, F. C. Shilling, and
D. J. Freed, Polym. Prepr., 21(2), 176 (1980).
120. W. H. Starnes, R. C. Haddon, D. C. Hische, I. M. Plitz, C. L. Schosser, F. C. Schilling,
and D. J. Freed, ACS Polym. Prepr., Polym. Chem. Div., 21 (2), 138 (1980).
121. R. C. Haddon and W. H. Starnes, Polym. Prepr., 18, 505 (1977).
122. R. C. Haddon and W. H. Starnes, ACS Adv. Chem. Ser., 169, 333 (1978).
123. T. Kelen, B. Ivan, T. Nagy, B. Turcsani, F. Tudos, and J. Kennedy, Polym. Bull. (Berlin), 1,
79 (1978).
124. S. A. Liebman, D. H. Ahlstrom, and C. R. Foltz, J. Polym. Sci., Polym. Chem. Ed., 16,
3139 (1978).

125. M. M. O'Mara, Pure Appl. Chem., 49, 649 (1977).

126. G. Scott, Developments in Polymer Degradation-1, N. Grassie, ed., Applied Science
Publishers, London, 1977, p. 205.
127. G. Scott, Adv. Chem. Ser., 169, 30 (1978).
128. G. Scott, M. Tahan, and J. Vyvoda, Eur. Polym. J., 14, 377 (1978).
129. G. Scott, Polym.-Plast. Technol. Eng., 11, 1 (1978).
130. A. R. Berens and V. L. Folt, Polym. Eng. Sci., 8(1), 5 (1968).
131. A. R. Berens and V. L. Folt, Polym. Eng. Sci., 9(1), 27 (1969).
132. J. W. Summers, J. D. Isner, and E. B. Rabinovitch, 36th SPE Antec, 757 (1978).
133. J. W. Summers, J. D. Isner, and T. E. Chapman, SPE Technical Papers XXV, Soc.
Plastics Engineers, Brookfield Center, Conn., (1979).
134. G. C. Portingell, Particulate Nature of PVC-Formation, Structure, and Processing, G.
Butters, ed., Applied Science Publishers, London, 1982, p. 147.
135. G. C. Portingell, Particulate Nature of PVC-Formation, Structure, and Processing, G.
Butters, ed., Applied Science Publishers, London, 1982, pp. 212, 175.

Page 346
136. D. A. Tester, Manufacture and Processing of PVC, R. H. Burgess, ed., Macmillan, New
York, 1982, p. 234.
137. J. Parey and G. Menges, J. Vinyl Technol., 3, 3, 152 (1981).
138. M. W. Allsopp, Manufacture and Processing of PVC, R. H. Burgess, ed., Macmillan, New
York, 1982, Chap. 8.
139. ASTM D2152-67, reapproved 1972, American Society for Testing and Materials,
Philadelphia.
140. A, Gonze, Plastics, 24(2), 49 (1971).
141. M. Lamberty, Plast. Mod. Elast., 82 (Dec. 1974).
142. M. Gilbert and J. Vyvoda, Polymer, 22, 1134 (1981).
143. G. C. Portingell, Particulate Nature of PVC-Formation, Structure, and Processing, G.
Butters, ed., Applied Science Publishers, London, 1982, p. 168.
144. E. A. Collins and C. A. Daniels, Polym. Eng. Sci., 14(5), 357 (1974).
145. F. Bueche, J. Chem. Phys., 20, 1959 (1952); 25, 599 (1956).
146. T. G. Fox and S. Loshaek, J. Appl. Phys., 26, 1080 (1955).
147. K. V. Gotham and M. J. Hitch, Br. Polym. J., 10, 47 (1978).
148. P. Benjamin, Int. Conf. PVC Process., Plastics and Rubber Institute, London, 1978.
149. P. Benjamin, J. Vinyl Technol., 2(4), 254 (1980).
150. J. W. Summers, E. B. Rabinovitch, and J. G. Quisenberry, J. Vinyl Technol., 4(2), 67
(1982).
151. G. Menges and N. Berndsten, Extrusion of PVCThe Consequences of Particle
Structure, PRI Int. Conf., PVC Processing, 1978.
152. D. M. Shinozaki, K. W. J. Vlachopoulos, and A. Hamielec, J. Appl. Polym. Sci., 21, 3345
(1977).
153. W. Retting, Angew. Makromol. Chem., 8, 87 (1969).
154. L. C. E. Stuick, Polym. Prepr. (ACS), 17, 142 (1976).

Page 347

7 Resin Testing and Quality Assurance

DANIEL J. BRANDT,
RICHARD S. GUISE, and
BERNARD A. McCOY
Occidental Chemical Corporation Pottstown, Pennsylvania
I. INTRODUCTION

348

II. SOURCES OF TEST METHODS

348

A. Company Standards

348

B. American Society for Testing and Materials

348

C. International Organization for Standardization

349

III. ANALYTICAL

350

A. Chlorine Analysis

350

B. Infrared Analysis
C. Molecular Weight
D. Molecular Weight Distribution
E. Thermal Analysis
F. Microscopy
IV. PHYSICAL PROPERTIES: GENERAL-PURPOSE RESINS
A. Apparent (Bulk) Density
B. Particle Size
C. Gels
D. Contamination
E. Volatiles
F. Plasticizer Absorption /Porosity

350
354
354
359
362
362
362
363
365
365
366

366
G. Thermal Stability
H. Melt Viscosity
I. Electricals

368
371
376

V. PHYSICAL PROPERTIES: DISPERSION RESINS

377

A. Viscosity of Plastisols at Low Shear Rates

378

B. Viscosity of Plastisols at High Shear Rates

379

C. Particle Size

379

D. Air Release

380

E. Volatiles

383

F. Contamination

383

G. Gelation Temperature

384

H. Gelation-Fusion Characteristics

384

Page 348
VI. PHYSICAL PROPERTIES: SOLUTION RESINS
A. Viscosity of Solution Resins at Low Shear Rates

387
387

B. Insoluble Particles

387

C. Color and Clarity

387

VII. ASTM RESIN CLASSIFICATION

387

A. Polyvinyl Chloride Resins

B. Vinyl Chloride Copolymers

387
390

VIII. QUALITY ASSURANCE

390

REFERENCES

393

I.INTRODUCTION
Testing and characterizing polyvinyl chloride (PVC) resins have different meanings and degrees
of complexity to different people. To the research chemist it may deal with relatively complex
subjects such as infrared analysis or scanning electron microscopy. To the processor
purchasing PVC resin, it may simply be checking the bulk density of the resin. PVC has proven
to be very versatile polymer because it is normally compounded with other ingredients, which
may affect the physical properties of the finished product as much or more than the PVC resin
itself. This versatility sometimes complicates the issue of resin testing and characterization. A
test or property that is important in a clear plasticized application may not be relevant in a rigid
pigmented one. This is where experience becomes important to determine what resin property
and test is important for a specific application.
The purpose of this chapter is to discuss some of the important resin tests and show how they
relate to resin properties. Whenever possible, physical properties and test procedures of
uncompounded PVC resins are discussed. However, PVC resin is not commercially used
without certain additives such as stabilizers, lubricants, and plasticizers (for flexible
applications). Therefore, some important properties of the basic resin require preparation of a
compound in order to prepare and test the resin sample properly. This chapter will concentrate
on basic resin properties; the testing of compounds is covered in Volume 3, Chapters 5 (rigids)
and 6 (flexibles), and compounding of PVC resin is covered in other chapters.

II.SOURCES OF TEST METHODS

A.Company Standards
Most test methods are initially developed by a company to cheek resin properties that are

important to their processing requirements. These may be considered proprietary due to the
efforts involved to develop the test, or because of the formulation or equipment used in the test.
Since quality is a paramount issue between buyer and seller, agreement on the assessment of
quality can be very difficult when a mutual standardized test is not available.
B.American Society for Testing and Materials
The ASTM was developed to fill this void and provide an open exchange of ideas between
producers and users to develop standardized test procedures.

Page 349
ASTM now has 140 committees with over 28, 000 members from 92 countries around the
world and controls 7450 standards. ASTM Committee D20, Plastics, consists of 670 members
and has 322 active standards [1].
C.International Organization for Standardization
The ISO is a worldwide federation of national standards institutes (ISO member bodies).
Member countries develop standards for use as International Standards by the ISO Council.
ASTM Committee D20 works with Committee ISO/TC61, Plastics. In this chapter ASTM test
procedures have been referenced whenever possible, since they are the prime source of
published standards in this country. ISO test procedures have also been cross-referenced
(Table 1).
Property

ASTM method

ISO method

Molecular weight
Apparent (bulk) density
Compacted
Sieve analysis
Dry
In water
Plasticizer absorption, centrifuge
Room temperature
Hot
Plasticizer absorption, spatula, at room temperature
Thermal stability
Resin
Fused film
Volatiles
Contamination
PVC resin designation
Homopolymer
Copolymer
Paste viscosity
Extrusion viscometer
Rotating viscometer

D1243
D1895
-

174
60
1068

D1921
D1705

4610
1624

D3367
D1755

4608
4574
-

D4202
D2115
D3030
-

R182
305
1269
1265

D1755
D2474

1060
2798

D1823
D1824

4575
2555

Page 350

III.ANALYTICAL
A.Chlorine Analysis
In the testing of polyvinyl chloride resins and copolymers, the quantitative determination of the
percent chlorine in the material can provide very useful information relating to chemical
composition. There are several acceptable procedures for chlorine-content determination, all of
them requiring careful technique for accurate results. The two most widely used procedures
involve chlorine-to-chloride conversion using either the sodium peroxide bomb fusion or the
oxygen flask combustion techniques. Details of the sodium peroxide fusion method may be
found in ASTM method D130355 [1]. The oxygen combustion flask involves combusting the
sample in a closed oxygen atmosphere system and absorbing the generated hydrogen chloride
in sodium hydroxide solution. The oxygen combustion technique, in one of its many
modifications, is the procedure of choice due to its high accuracy and short analysis time.
To begin the oxygen combustion flask test procedure a weighed sample of PVC (approximately
20 to 30 mg) is placed in a chlorine-free black paper wrapper. This preparation is then placed
in an oxygen-purged combustion flask containing sodium hydroxide absorption solution and is
then ignited most conveniently using a Thomas-Ogg Infrared Igniter available from Arthur H.
Thomas Co. The solution prepared from the contents of the ignition flask is then analyzed for
chloride content using an automatic titrator such as the Buchler-Cotlove Chloridometer, which
titrates the chloride ion coulo-metrically and detects the end point amperometrically. More
detailed information on oxygen combustion flask techniques may be found in works by Haslam
et al. [24].
Before a chlorine analysis is performed, PVC resins should be purified by Soxlet extraction with
methanol to remove polymerization additives which are present at levels varying from 0.1% for
suspension resins to greater than 3% for dispersion resins. With a theoretical chlorine level of
56.8% for homopolymer, most copolymers would be expected to yield a lower chlorine
percentage and chlorinated PVC to yield a greater value. The major limitation of this technique
is the inability to provide useful data on copolymers of vinyl chloride with other chlorinated
monomers such as vinylidene chloride.
B.Infrared Analysis
Infrared spectroscopy is without a doubt the most widely used technique for the identification of
organic compounds. Its usefulness results from its ability to identify molecular functional groups
and group interactions by the absorption of selective frequencies in the infrared region of the
electromagnetic spectrum. The most widely used infrared region is between 4000 and 600
wave numbers (cm-1). A spectrometer designed to operate in this region is used to measure
sample absorption at selected frequencies. Commercially available instruments (Figure 1)
provide a plot of frequency versus absorption or transmission, and it is this instrument read out
that provides the information necessary to identify functional groups from their characteristic
absorption bands. The most common method employed for the infrared identification of an
unknown is, however, matching the total unknown spectrum with the spectrum of a reference

material. There is much basic information available in the literature on the theory and application
of infrared (IR) spectroscopy [57] and will, therefore, not be discussed in any further detail.

Page 351

FIGURE 1 Infrared spectrophotometer and data station. Perkin-Elmer model 1320 and model
3600 data station. (Courtesy of Perkin-Elmer.)
The application of IR spectroscopy to PVC and its related copolymers is relatively
straightforward, and a great deal of information about the sample can be obtained in a few
minutes with a very small amount of sample. As with many other techniques, sample
preparation is of major importance in infrared spectroscopy. PVC, like many other polymers,
absorbs very strongly in the infrared spectrum and therefore requires either thin-film sections or
low-concentration solutions for proper examination. Thin films (0.01 to 0.10 mm) can be
obtained by casting the film from solution onto a watch glass or directly onto a NaCl plate.
Tetrahydrofuran is a very suitable solvent for this procedure. Care must be taken to remove all
residual solvent so that spurious absorption bands will not be present. Drying the cast film at
steam bath temperatures for 2 hr is usually sufficient for solvent removal. For certain
applications it may be necessary to examine a polymer solution using an infrared (IR) liquid cell.
The solubility of PVC limits solution makeup to just a few solvents: tetrahydrofuran,
cyclohexanone, o-dichlorobenzene, methyl ethyl ketone, and ethylene dichloride are some that
can be used with most homopolymer and copolymer resins.

Page 352
In the testing of polyvinyl chloride resins as incoming or competitive raw materials, the most
important information to be gleaned from the infrared spectra is whether or not we are dealing
with a homopolymer or copolymer and the possible types of comonomers present. Figure 2
shows a typical infrared absorption curve for polyvinyl chloride with labeling of various
characteristic functional group frequencies. In general, comparing infrared spectra of
commercial PVC resins will not generate any useful information. The only value is being able to
say that the sample is either homopolymer or copolymer. When the spectrum of a PVC resin is
compared to an unknown vinyl chloride resin and additional absorption bands are observed, a
copolymer resin is indicated. Figure 3 shows a typical curve of a vinyl chloride/ vinyl acetate
copolymer. This is a good copolymer example to use because the vinyl acetate comonomer
generates three readily discernible absorption bands. This type of vinyl chloride copolymer is
also the most widely used in the industry. The appearance of an ester carbonyl (-C=O) band at
1950 to 1600 cm-1 indicates the presence of comonomers such as vinyl acetate, di-butyl
maleate, and methyl acrylate. This absorption is due to the stretching of the carbon-to-oxygen
double bond. The appearance of a doublet absorption at approximately 1040 cm-1 and 1060
cm-1 indicates the presence of vinylidene chloride comonomer. Numerous books are available
containing spectral libraries of polymers and related materials [4, 811].
Infrared spectroscopy is not an extremely powerful method in quantitative analysis but does
lend itself quite well to certain quantitations in the polymer field. This is due mainly to its specific
nature and to the lack of better, easily performed methods. Probably the most widely used
quantitative procedure for vinyl polymers is the determination of the vinyl acetate

FIGURE 2 Infrared spectra: polyvinyl chloride.

Page 353

FIGURE 3 Infrared spectra: vinyl chloride/vinyl acetate copolymer.

content of vinyl chloride/vinyl acetate copolymers. The procedure consists of dissolving the
sample in a suitable solvent (tetrahydrofuran or ethylene dichloride) and comparing the sample
carbonyl absorption at 1740 cm-1 to that of standards of known vinyl acetate content. This
technique works well for most ester copolymers and usually gives adequately accurate results.
With the proper choice of absorption bands, non-carbonyl-containing comonomers can be
quantitated. Standard procedures for quantitative infrared spectroscopy are given in ASTM
E16867 [1]. Most books on infrared spectroscopy also have sections on quantitative analysis
[57].
With the advent of less expensive small computer systems, two other areas of infrared analysis
instrumentation have become available. These are the field of computer-assisted IR and Fourier
transform infrared spectroscopy (FT-IR). Computer-aided infrared spectroscopy is the coupling
of relatively low cost data processing hardware to a microprocessor-controlled
spectrophotometer. Software is then utilized to provide all the desired functions, such as
spectral storage, manipulation, and quantitative analysis. The advantages of such a system are
increased precision and accuracy of analysis with reduced time and effort [12]. The FT-IR
technique provides us with even greater accuracy and speed and can be used to aid in the
elucidation of the fine structure of PVC resins [13]. Books of Bell and Griffiths discuss in detail
the theory and chemical application of FT-IR [14, 15].
There are numerous other techniques which may aid in the identification of a polymer
composition. Pyrolysis gas chromatography can be used to get a generally characteristic
fingerprint of the polymer decomposition products. When this technique is coupled with mass
spectroscopy, actual identification of the separated products can be made. Nuclear magnetic
resonance spectro-

Page 354
scopy (NMR) can give information on the chemical nature and bonding situation of the nuclei
within a molecule and is useful for copolymer analysis (e.g., direct measure of propylene).
Several studies on PVC branch content and unsaturation have been done using the NMR
technique [1618]. At this time, of the many spectroscopic analytical techniques that are
available, only IR lends itself to providing useful quality control information on PVC resins.
C.Molecular Weight
The standard techniques that are available for determining the molecular weight of high
polymers can be used for analysis of PVC polymers. All PVC resins and almost all polymers in
general consist of molecules of various sizes. The process of polymerization is responsible for
this statistical effect. All commercially available PVC resins have a wide molecular weight
distribution, consequently, measurement of molecular weight provides only average values. The
two most commonly used molecular weight expressions are weight average molecular weight
(Mw) and number-average molecular weight (Mn). Number-average molecular weights are
frequently measured by making osmotic pressure measurements on solutions of the polymer in
an appropriate solvent. The measurement of light scattering of polymer solutions is the most
widely used and accepted technique for the determination of weight-average molecular
weights.
Since the subject of molecular weight and its determination has been covered in Chapter 6,
Section I. B, it will not be considered further here. Instead we shall consider the effects of
molecular weight distribution.
D.Molecular Weight Distribution
In the study of molecular weight more and more importance is being assigned to the distribution
of molecular weights within a polymer sample. It has been determined through studies of resin
processing and physical properties that two resins could have the same weight-average
molecular weight or solution viscosity and yet have dramatic differences in processing and
physical characteristics, such as tensile, flex life, brittleness, and flow properties.
The molecular weight distribution (MWD) for PVC is defined during the polymerization process
and is controlled by such variables as temperature, molecular weight modifying additives, and
percent conversion. Polymerization temperature is the main factor controlling molecular weight
on most commercial PVC resins. In general, the lower the polymerization temperature, the
higher the molecular weight of the resin. The addition of molecular-weight-modifying agents at
any temperature can act to either raise or lower molecular weight. Additives to lower molecular
weight are very commonly used in vinyl chloride/vinyl acetate copolymers. The MWD for PVC
homopolymers is normally near-Gaussian and does not deviate from one manufacturer to
another within a specific polymerization technique (e.g., suspension polymerization). The MWD
can be skewed to the high or low region by the addition of cross-linking or chain transfer
agents. The distribution can also be adjusted through polymerization technique (e.g., continuous
emulsion polymerization).

Page 355
Today with modern analytical equipment employing the technique of size exclusion
chromatography the MWD, together with Mw, Mn, Mv, Mz, and [], can be determined in less
than 30 min. This is in contrast to the many hours of tedious work required to determine MWD
by the older, commonly used method of fractional precipitation. This fractionation technique
involves adding a nonsolvent to a polymer solution in increments to precipitate fractions of
progressively lower molecular weight. To define the MWD, a plot is made of the polymer weight
fraction remaining in solution versus the nonsolvent volume percentage.
The instrumental technique in widespread use today is a specialized area of the field of liquid
chromatography. The area is called gel permeation chromatography (GPC) or size exclusion
chromatography (SEC). The procedure is performed by an instrument specifically equipped for
this purpose (Figure 4). The instrument consists of three main sections:
1. Solvent delivery system
2. Fractionation columns
3. Detector
The solvent delivery system is usually a high-pressure, pulse-dampened precision pump. The
most commonly used detector is the differential refractometer, which gives a continuous
readout of the difference in refractuve index between the polymer solution and the pure solvent.
A change in refractive index is directly proportional to concentration in dilute polymer solutions
and is usually not dependent on molecular weight when above a molecular weight of a few
hundred. The fractionation columns are a series of columns packed with a fine particulate
material of controlled diameter pore size. The most commonly used packing material for
polymer work in organic solvents is highly cross-linked styrene/divinyl benzene copolymer.
Conditions are controlled during the polymerization of this packing to control pore size
accurately in the individual particles.
The fractionation that is performed on the polymer molecules as they are carried through the
packing material by the carrier liquid is dependent on the increase in residence time of the
progressively smaller molecules as they migrate in and out of the packing material pores. The
larger molecules, which spend little time in the packing pores, are eluted first from the column.
A set of columns with pore sizes of 106, 105, 104, and 103 is a common arrangement for
work with PVC. The refractive index detector then gives us a continuous polymer concentration
readout. If we can relate retention time to molecular weight, we can then calculate molecular
weight averages. The calibration for molecular weight is typically performed using commercially
available polystyrene standards with a Mw/Mn ratio of approximately 1.0. A plot is then made
of retention time or elution volume versus molecular weight on multicycle log paper. Figure 5
shows a typical calibration curve. Calibration plots are generally linear between molecular
weights of 103 and 106. The signal output of the detector, for purposes of rapid data reduction

and analysis, can be input into a calculating intergrator designed for GPC analysis. Figure 6 is a
GPC curve for a PVC resin. The data-reduction device

Page 356

FIGURE 4 Gel permeation chromatographs. Waters model GPC I and GPC II. (Courtesy of
Waters Associates.)

Page 357

FIGURE 5 Typical GPC calibration plot using polystyrene standards.

Page 358

FIGURE 6 Typical gel permeation chromatogram for polyvinyl chloride.

splits the GPC curve into slices and intergrates the area of each slice. Through the use of the
calibration plot and the slice retention time, the average molecular weight can be calculated.
This slice information is then used to calculate the various molecular weight averages for the
polymer sample [25]. ASTM standard test method D3593, Molecular Weight Averages and
Molecular Weight Distribution of Certain Polymers by Liquid Size-Exclusion Chromatography
(Gel Permeation ChromatographyGPC) Using Universal Calibration, covers all the areas of
GPC from instrumentation through calculations in a very thorough fashion [1]. This method is
applicable only to linear polymers soluble in tetrahydrofuran. The molecular weights calculated
from a polystyrene calibration curve are not absolute but are relative to polystyrene. The Qfactor method is one way to correct the curve to represent PVC more accurately. In this
method, ratios of chain length and weight are made for the polymers involved [25]. Other
correction methods involve the use of a single or several broad molecular weight standards of
PVC for which molecular weight averages have previously been determined by absolute
methods [26].
A term commonly encountered when working with molecular weights is an expression of
Mw/Mn, known as polydispersity or heterogeneity index. This expression gives some indication
of the broadness of a molecular weight distribution. If all the polymer molecules are the same
size, the polydispersity would equal 1. The polydispersity for normal production PVC is
generally between 2.0 and 2.5. Great care must be exercised when comparing the polydispersity of resin samples due to the inherent inaccuracies in the size exclusion
chromatographic technique. The reproducibility for Mw is generally 2.5%, but the
reproducibility for Mn is much poorer at 7.5% without utilizing special precautions while
performing the analysis. Because of the in-

Page 359
accuracies noted above, the application of SEC to resin quality control can lead the analyst to
incorrect conclusions if proper consideration is not given to the limitations of the procedure. For
most laboratories, the measurement of resin solution viscosity is a far more accurate and useful
indicator of molecular weight variation than is the development of molecular weight distribution
curves.
E.Thermal Analysis
In the study of the thermal properties of polymers, it becomes immediately apparent that there
are many polymers which are mixtures of crystalline and amorphous regions. Polyvinyl chloride
homopolymer and copolymers fall into this catagory. X-ray diffraction measurements on
commercial PVC indicate the polymer to be largely amorphous, with between 5 and 10% of
small crystallite regions. PVC shows a characteristic change in properties as its temperature is
raised. At room temperature PVC is in a glassy state and when heated passes through a welldefined glass transition temperature, Tg. This transition is characterized by a change in
modulus as well as heat capacity and expansion coefficient. This transition is generally
associated with the onset of free segmental motion in the polymer molecules.
The Tg of PVC varies with the temperature of polymerization of the resin with practically all
commercially available resins having a Tg between 80 and 85C. Since polymerization
temperature is the main controlling factor for molecular weight, Tg is directly related to
molecular weight. For many polymers this is a linear relationship. In practice the properties of
many polymers are determined by the position of the glass transition; consequently, the
measurement of this parameter can be of some interest in the examination of resin quality.
As a quality control technique, the examination of the Tg of vinyl chloride copolymer can be very
useful. With a proper calibration curve of Tg versus percent polymer composition, it would be a
simple, rapid matter to measure the Tg of a sample and read the percent composition from the
calibration curve. Tg responds to polymer-plasticizer compositions in a fashion similar to that of
copolymers. An equation that is widely used to relate the glass transition temperature of a
copolymer or resin-plasticizer mixture to the Tg of its components is as follows:

where
Tg =
glass transition temperature of mixtures, T(K)
W1, W2, . . . , WN =
weight fraction individual components
Tg1, Tg2, . . . , TgN =
Tg of individual components, T(K)
The measurement technique for Tg, which is commonly used, is based on the change in heat
capacity in the region of the glass transition. There are several commercial instruments
available which are suitable for the measurement of Tg. Figure 7 illustrates one such instrument
equipped with a computer data analysis station. These instruments either measure temperature
differences or heat flow rate differences between a sample and a reference material, and are
consequently known as differential thermal analyzers. A plot of these temperature or heat flow
rate differences versus temperature is

Page 360

FIGURE 7 Perkin-Elmer DSC-4 with 3600 data station. (Courtesy of Perkin-Elmer Corp.)

Page 361

FIGURE 8 Glass transition thermogram of polyvinyl chloride.

known as a thermogram. Figure 8 is an illustration of a thermogram of a typical commercial
PVC resin, showing the glass transition temperature at 81.6C. The test is usually conducted on
a sample weight of 10 to 20 mg, with scan rates of 10 to 20C/min. Samples are sealed in
small aluminum pans with a blank pan used as a reference.
A few factors must be considered when comparing Tg results between samples and between
laboratories. The scan rate of the analysis can cause shifts in the Tg value. The method used
graphically to determine the Tg from the thermogram can be another source of discrepancy.
The convention in thermal analysis is to use the midpoint of the transition, while some analysts
use the initial onset of change as the glass transition temperature. Instrument manufacturers
have much useful literature available on thermal analysis techniques and application to various
systems [1921]. Other thermal analysis techniques and instruments are available for use with
small samples; among these are thermogravimetric analysis (TGA) and thermomechanical
analysis (TMA). TGA is a very sensitive weight loss versus temperature or time technique and
generally is not very useful as a resin quality control technique, but can be useful to detect
stability differences between PVC homopolymers and copolymers. TMA, on the other hand,
can provide useful information on mechanical softening point, coefficient of linear expansion,
and heat distortion temperature. Some applications of the various thermal analysis techniques
and their relationship to ASTM testing standards are described in a work by Brennan [22].

Page 362
F.Microscopy
The application of the light microscope in the examination of resin particle size and morphology
is fast giving ground to the scanning electron microscope (SEM), due to the recent availability
of less expensive instruments and the high-quality information which they provide. The modern
SEM is easy to operate and maintain and is one of the best tools for the observation and
measurement of particle size of dispersion resins. Photographs taken at 10 to 10,000 are
considered routine. Slicing suspension resin particles followed by SEM examination is an
excellent procedure to relate internal particle morphology to resin dry blending characteristics.
Care must always be taken to ensure that statistically significant sampling has been done. The
application of too few observations in a microscopic examination can lead the analyst to serious
error. The stereo light microscope is still a useful tool for the examination of suspension resin
particles. The observation of plasticizer uptake by resin particles has been an often used
procedure to compare resin porosity. This procedure is used more as a spot test and not as a
strict quality control procedure.

IV.PHYSICAL PROPERTIES: GENERAL-PURPOSE RESINS

A.Apparent (Bulk) Density
ASTM D1895 defines apparent density as the weight per unit volume of a material, including
voids inherent in the material as tested. The term bulk density is commonly used for materials
such as molding powder [27]. Thus in practice the terms apparent or bulk density are often
used interchangeably to describe a property of PVC resin (or compound).
Density measurements are very useful in determining storage requirements, mixer or equipment
charge size, and handling characteristics. Density can also affect processing parameters, such
as fusion and output in an extrusion operation. Bulk density is affected by the density of the
individual resin particles as well as particle-size distribution, which can allow combinations of
different-size particles to pack together more than uniform-size ones.
ASTM D1895 offers Method A for fine granules and powders that can be poured easily through
a funnel having a 9.5-mm-diameter hole in the bottom and is often used for more finely divided
powders such as vinyl resins [27]
Pourability is defined in ASTM as a measure of the time required for a standard quantity of
material to flow through a funnel of specified dimensions and is a measure of the readiness
with which such materials will flow through hoppers and feed tubes and deliver uniform weights
of material [27].
Flow properties are often difficult to measure consistently and correlate with handling in
processing, because of variables such as static, vibration, funnel or hopper hole size, and
handling of the resin before the test [28]. Care must be taken in running the relatively simple
apparent density and/or pourability tests to control these variables as much as possible. Resin
should be conditioned and loaded in the funnel as consistently as possible. Antistat, such as

carbon black mentioned in ASTM D1895, may be added to cut down on static, or samples may
be conditioned at higher relative humidity to alleviate that problem.
Apparent (bulk) density is measured according to ASTM D1895, Method B, as follows:

Page 363
1. A standard funnel having a 25.4-mm-diameter opening at the bottom is mounted 38 mm
above a 400-cm3 measuring cup.
2. While closing the small end of the funnel with a suitable flat strip, pour a 500 + 20-cm3
sample into the funnel.
3. Open the bottom of the funnel quickly and allow the material to flow freely into the cup.
4. When all the material has passed through the funnel, scrape off the excess on top of the cup
with a straightedge without shaking the cup.
5. Weigh the material in the cup to the nearest 0.1 g and calculate the density in grams per
cubic centimeter (g/cm3). To convert grams per cubic centimeter to pounds per cubic foot,
multiply by 62.43.
ISO 60 also contains a test procedure for apparent bulk density, while ISO 1068 cheeks
compacted apparent bulk density. In this procedure 100-g of PVC resin is shaken by a
mechanical apparatus in 250-cycle increments until the measured volume after two successive
shaking cycles is nearly constant. At that point the smallest compacted volume is used to
calculate the compacted bulk density of the resin.
B.Particle Size
Particle size is one of the most obvious ways of distinguishing between the different types of
PVC resin: dispersion resin, dispersion-blending resin, or general-purpose resin. Table 2 shows
a general classification of these resin types.
Since dispersion resins are so small in particle size, it has been difficult to develop a
standardized means of analyzing particle size. One method to characterize resin before it is
blended with plasticizer is the Coulter principle [29]. Microscopy such as transmission electron
microscope (TEM) or scanning electron microscope (SEM) can also be used but are more
research tools than quality control tests. Since dispersion resin typically is used in paste form,
various techniques have been investigated to separate and measure plasticized particles, but
no standardized test has evolved. Generally, these tests use centrifugal force or gravity to
separate particles by size.
The dispersion-blending resins can be evaluated by ASTM D1705 utilizing a wet sieve
technique. Woven-wire screens for wet screening are available as small as 38 m in the
following standard sizes [30]:
US standard mesh

Micrometers

230

63

270
325
400

53
45
38

Page 364

Dispersion
Dispersion-blending
General purpose

Particle size (m)

Typical test method

1
20
120

Microscopy, centrifugal
Microscopy, wet sieving
Dry sieving

In ASTM D1705, wetting agent is added to a resin sample and the mixture is poured on top of
the assembled screens. The resin is gradually washed through the successive screens with a
gentle stream of water. The screens are then dried and percent retention is calculated. ISO
1624 also describes a similar wet-screening test procedure. Different techniques have been
tried to prevent blinding of the screens, but they ultimately make use of rinsing the resin,
vibrating or moving the screens, vaccum or ultrasonic techniques, or combinations of these to
assist the resin in passing through the screens.
General-purpose resins are usually tested by a dry sieve analysis such as the method
described in ASTM D1921. A mechanical sieve-shaking device is used to rotate and tap the
collection of sieves with a pan and cover. The amount of sample retained on each sieve is then
weighed and reported as percent retention. Since static can cause fine powders to clog
screens, antistatic agents such as carbon black are normally added to the resin sample for
more accurate results. It is also important to use the recommended amount of resin for the
diameter of the screen being used, since too much sample may not allow proper shaking and
give false results.
Most dry screening of general-purpose resin uses the following range of screens [30]:
US standard mesh

Micrometers

20
40
60
80
100
120
140
170
200
230

850
425
250
180
180
125
106
90
75
63

Page 365
ISO 4610 conducts a sieve analysis using an air-jet sieve apparatus. This equipment uses air
jets to aid the movement of resin particles through the screens. Other systems combine sonic
vibrations, tapping of the screens, and air to assist resin movement and prevent blinding of the
screens.
Particle size of general-purpose resins can affect bulk-handling properties and processing
characteristics. An excessive amount of fine particles (usually considered smaller than 200mesh screen retention) can cause dusting and packing problems when handling, and can also
contribute to blending problems when adding plasticizer. Coarse particles (usually considered
40 mesh and larger in size) may cause gel problems in processing due to their poorer
absorption and fusion characteristics.
C.Gels
Gels, or fisheyes, are not a simple physical property to measure like particle sizeor bulk
density. They are resin particles which have not fused under a given set of processing (or
testing) conditions. They often appear as translucent particles in a finished product, because
they have not accepted compound ingredients such as pigment (and plasticizer if used) during
blending and processing. Gels in plasticized applications are usually caused by oversized or
nonporous particles formed during polymerization or nonuniform blending. Contamination with
nonhomogeneous resin (usually higher molecular weight) is the primary cause of gels in rigid
formulations. The number of gels is generally related to the amount of shear and/or heat that a
compound is subjected to in its processing. Since the observed gel properties of a given
sample of resin may depend on the formula, the processing temperature, the processing speed
(shear rate), and the resin itself, a universal test is very difficult to develop.
ASTM D3596 provides a general flexible gel (fisheye) test to try to determine the propensity for
a resin sample to produce gels under controlled conditions. In this test, a sample of resin is
blended with calcium stearate and di-2-ethylhexyl phthalate plasticizer (DOP) in a laboratory
mixer. After mixing, the dry blend is milled on a two-roll mill, with the addition of carbon black
during the milling. After a specified milling time, the sheet is removed and press-polished. Gels
are then counted in a specified area using a back-lighted viewing box and a comparator. Great
care must be taken to ensure consistent time, temperature, and working of the stock on the mill
rolls during the test.
As a quality control test, processors often develop their own procedures for checking gels using
their own formula and either a mill or scaled-down version of their processing equipment.
Correlation between a test laboratory and commercial equipment can still be difficult, because
the apparent number of gels is so sensitive to the amount of heat and shear developed in the
fusion process. Gel tests run for extended times reduce the gel count to a minimum value
associated with infusible glassy particles. This is a low number but can be important in critical
applications (e.g., sound records).
D.Contamination
There are always many potential sources of contamination during the manufacture, shipping,

blending, and processing of PVC resin. The type of processing, formulation, appearance
requirements, and end use of the finished product usually determine how critical contamination
is to a proces-

Page 366
sor. Tests to detect a potential contamination problem before it reaches production generally
involve looking visually at resin or fusing the material and looking at the product, such as a mill
sheet.
1. Resin: Looking for contamination can be as simple as spreading a measured amount of resin
on a laboratory bench and counting the number of off-color dark particles. These can also be
checked to see if they are metallic in nature, which may help to determine the source of the
contamination. Another method involves making a slurry of resin in water by adding a surfactant
to wet out a measured amount of resin in a flask. After shaking vigorously, counting the number
of dark particles that have settled to the bottom of the flask gives an indication of
contamination.
ISO 1265 cheeks PVC resin for impurities of foreign particles by observing a certain amount of
resin which has been flattened between glass plates. A glass plate is inscribed with a grid of
100 squares each 30 mm 30 mm. The glass plate is placed over 200 cm3 of resin and moved
slightly to spread the resin over at least 25 squares. Particles are then counted twice in 25
squares and reported as number of specks per 100 squares of the grid.
A mechanical means such as a Syntron Vibra-flow Feeder can also be used to sift resin while
looking for visual contamination. A measured amount of PVC resin gradually flows down a
vibrator chute while the operator visually counts the particles of contamination. Even though
attempts can be made to measure size of contamination, any test is quite subjective in nature
due to the operator (observer) factor.
2. Mill Sheet: A common method of preparing fused samples of material for contamination
cheek is by a two-roll mill. A simple formula containing resin, calcium stearate, DOP, and
titanium dioxide (white pigment helps to make contamination stand out) may be used. The blend
is milled and the mill sheet examined for contamination in a specified area. A processor may
also use his commercial formula and therefore cheek all raw materials for contamination by
checking the appearance of a finished sheet.
E.Volatiles
Volatile contents of PVC resin are typically moisture, but can also be trace amounts of additives
from the polymerization process. Large amounts of volatiles can create problems in handling
and blending of resin, as well as cause porosity in the fused product.
ASTM D3030 covers the quantitative determination of the volatile matter (including water)
present in vinyl chloride resins [31]. In this test a 10-g sample of resin is heated at 110-C in a
forced ventilation oven until the resin reaches constant mass. The volatile matter is then
reported as percent of the original mass. Volatile analyzers are available which are a
combination constant-temperature oven and balance, thus allowing the weight loss to be
checked without removing the sample from the oven. Many quality control tests modify the
ASTM procedure by specifying a constant time in order to shorten the length of the test. ISO
1269 also contains a nearly identical volatile test performed on 5 g of resin.

F.Plasticizer Absorption /Porosity

The ability of a PVC resin to absorb plasticizer has long been recognized as an important resin
property. It may determine the speed at which a blend dries up after adding plasticizer to the
resin. It may be a means of measur-

Page 367
ing porosity of the resin, which, in turn, can affect fusion characteristics. It may even be an
indication of a resin's ability to absorb other ingredients in a rigid formula where no plasticizer is
present. Thus many tests have been developed over the years to predict a resin's ability to
absorb other ingredients, both at room temperature and at elevated temperature experienced in
blending.
1. Plasticizer Sorption (Spatula Test): A relatively simple room temperature test is described in
ASTM D1755. Plasticizer sorption is a measurement of the amount of plasticizer that a resin
can absorb at the standard laboratory temperature and is one of the parameters for judging
the dry blending properties of a resin [32]. DOP is added incrementally to 5 g of resin and
mixed with a spatula on a glass plate. The blend is then shaped in a mold to match the spatula
shape and placed on the spatula. The end point is deter-mined when the blend first slides off
the blade when the spatula is rotated from a horizontal to vertical position. The amount of
plasticizer added to reach that end point is measured and the plasticizer sorption is calculated
and expressed in terms of parts of DOP per 100 parts of resin. The test is used to categorize
resins by cell classification to reflect its general room-temperature sorption characteristics.
2. Centrifuge Plasticizer Sorption: ASTM D3367 describes a newer room-temperature
plasticizer sorption test which removes some of the human subjectivity of the spatula test. A
sample of PVC resin is saturated in a screening tube with an amount of DOP plasticizer
approximately twice the weight of the PVC. The mixture is centrifuged for 40 min to remove all
excess plasticizer that is not absorbed by the PVC particles. The weight of plasticizer absorbed
by the PVC is then reported as a percent of the resin weight.
ISO 4608 also describes a similar centrifuge plasticizer sorption test at room temperature,
while ISO 4574 determines hot plasticizer sorption of PVC resins.
3. Mercury Intrusion Porosimetry: ASTM D2873 describes a procedure for measuring the
interior pore volume of porous PVC resins by forcing mercury under pressure into the pores of
resin samples. The method is intended to compare differences in the total interior pore volume
of porous vinyl resins. This should give general indications of comparisons in porosity and dry
blending properties of the resins. Since the test forces mercury into pores that may not
normally be filled by simple absorption of plasticizer, correlation with actual blending
characteristics is not always possible.
4. Torque Rheometer Powder Mix: To test resin absorption characteristics under conditions
more similar to those encountered in a commercial blender, a torque rheometer can be used to
heat and mix the resin while blending the plasticizer (and other ingredients as required).
ASTM D2396 describes a Powder-Mix Text of Poly(vinyl chloride) (PVC) Resins using a Torque
Rheometer [33]. In this particular test a sample of resin is mixed with clay and basic lead
carbonate in a torque rheometer using a sigma mixing head. After 5 min of mixing and preheat,
DOP plasticizer is added. A wet lumpy mix occurs immediately after plasticizer addition which
changes to a free-flowing powder as mixing continues. The transition from the wet lumpy stage

to the dry powder stage is accompanied by a rather abrupt change in the torque value [33]. As
shown in Figure 9, the time from the plasticizer addition to the dry point can be measured and
reported as the powder-mix time or dryness time.

Page 368

FIGURE 9 PVC powder mix test. A, plasticizer addition; B, dry point. (Courtesy of Haake
Buchler Instruments, Inc.)
The speed with which the resin absorbs the plasticizer is related to the rate of heating of the
resin-plasticizer mix, the porosity of the particles, the surface area, the particle size and
distribution, molecular weight of the resin, and the physical properties of the plasticizer, and
other ingredients present [33]. This procedure is often modified to use the torque rheometer
with a formula and temperature similar to that used in a processor's plant.
G.Thermal Stability
Thermal stability of a PVC compound is of obvious concern to a processor to ensure that no
degradation occurs during the heat history imposed by processing the material. Two
approaches are taken to thermal stability: testing the PVC resin alone or testing the resin in a
compound.
1. Thermal Stability of PVC Resin: ASTM D4202 describes a new test method to determine the
thermal stability of PVC resins. A sample of PVC resin (no additives) is placed in a test tube
which is stoppered after inserting a small glass tube with Congo red paper on the end. The
tube is immersed in a hot-oil bath until the Congo red paper turns blue from the evolution of
hydrogen chloride. The results are reported as the time from immersion in the oil until
appearance of the blue color. The ASTM procedure has not been validated for PVC
compounds. ASTM D4202 states As each type and quantity of stabilizer will have a different
effect on thermal stability of various resins, this method is not intended to predict the thermal
stability of a given resin in a given compound [34]. Since PVC is commercially compounded
with other ingredients, which will affect the processing thermal stability, other test procedures
conducted with the compounds are usually considered more reliable.
Resin stability can also be tested by TGA when run at various temperatures and times in air or
inert gas; weight loss versus time curves can be significant to compare PVC homopolymers and
copolymers [35].

Page 369
2. Static Heat Stability: A type of heat stability more similar to some commercial uses of PVC is
an oven heat stability test such as that described in ASTM D2115. Test specimens of a
formulated compound (not less than 1 in. 1 in.) are typically cut from a mill sheet prepared on
a two-roll mill. Specimen supports (such as foil) are prepared which will hold one specimen of
each of the resins or compounds under test, for each time interval of test. One specimen of
each resin or compound under test is placed on each foil support so that when removed from
an air circulating oven a visual comparison of all samples for a given time and temperature can
be observed.
This practice is particularly applicable for determining gross differences in the heat stabilities of
poly(vinyl chloride) compositions that are detectable as a color change. It is not intended to
measure absolute thermal stability. Although the observed color changes may be evidence of
degradation, molecular degradation phenomena such as chain-scission or cross-linking may not
be identifiable [36].
ISO 305 describes a heat stability test on a thin sheet of fused PVC compound which is heated
in a hot-oil bath at a specified temperature between 120C and 200C. Samples are removed
at intervals until complete blackening occurs.
3. Dynamic Heat Stability: ASTM D2538 describes a fusion test for PVC powder mix using a
torque rheometer. Many laboratories expand the use of the torque rheometer by extending such
a fusion test until the PVC compound begins degrading, thus measuring a time until failure or a
dynamic stability time.
A torque rheometer with a fusion head (Figure 10) is used for the fusion-stability test. The bowl
or mixing chamber temperature can be controlled either electrically or by oil to a specified
temperature. Speed of the rotors can also be controlled and is a convenient way of controlling
shear rate and time of the test. Resin can be compounded in a very simple formula with
stabilizer and lubricant, or a commercial formula of the type being used in a plant. After
carefully preheating and checking the temperature of the bowl, the sample is loaded.
Consistent loading technique is very important to reproducibility of the test. Sample size is also
important and should be consistent with recommendations by the equipment manufacturers.
Figure 11 shows a typical torque curve from a torque rheometer stability test. The torque rises
sharply as the compound fuses and then drops off as the fused compound decreases in melt
viscosity. After reaching an equilibrium torque, the torque will again increase as the PVC begins
to degrade and cross-link. The end point can then be determined from this point of torque
increase, and the stability time calculated from fusion until degradation. Generally, if ultimate
stability is the object of a torque rheometer test, relatively high shear rates and temperatures
are used in order to see the test results rather quickly.
When conducting a Brabender fusion test at a fixed temperature, usually the resins
(compounds) develop different torques (melt viscosity) and shear rates. However, in a
commercial operation (such as extrusion) a die and related equipment would be designed for

use at a fixed melt viscosity and shear rate. Machine temperature (and speed) would then be
adjusted to obtain a melt viscosity and shear rate compatible with the die. Therefore, another
approach on a Brabender has been to develop stability data at a fixed torque (or melt viscosity)
on a Brabender which would more closely relate to the manufacturing parameters. At a
selected fixed torque, stability times define processability differences between rigid vinyl
compounds [37]. Bra-

Page 370

FIGURE 10 Torque rheometer. (Courtesy of C. W. Brabender Instruments, Inc.)

Page 371

FIGURE 11 Torque rheometer stability curve. (Courtesy of C. W. Brabender Instruments, Inc.)

bender stability time and torque data are measured for a given compound at several test
temperatures. Stability time is then plotted versus torque in order to determine graphically the
stability time at the fixed torque (e.g., 1500 m-g). Then, after determining this stability time at
the fixed torque for various resins (compounds), a comparison can be made of the
processability (stability) of the resin (compounds) [37].
H.Melt Viscosity
Since most commercial processes involve the flow of molten PVC through a die or mold, melt
viscosity of the resin (compound) is a very important property. Basic resin properties such as
molecular weight have been used to predict melt viscosity, but a laboratory test measuring
actual viscosity can better predict flow characteristics in processing.
Laboratory tests studying the flow of molten PVC are based on the standard principles of
rheology, which is the study of flow. Rheology deals with the comparison of rate of stress
(force per unit area) on a material and rate of shear (rate of deformation) of the material.
Viscosity is then defined as the ratio of the shear stress to shear rate. Figure 12 shows typical
shear stress versus shear rate curves for several types of materials, while Figure 13 shows the
corresponding viscosity versus shear rate curves.
The curves show that Newtonian fluids have a viscosity that is independent of rate of shear.
Since viscosity of such a liquid remains constant, measuring its viscosity at a given shear rate
would suffice to characterize the viscosity at any shear rate (at a particular temperature).
However, plastics such as PVC are typically non-Newtonian in nature, and most such

Page 372

FIGURE 12 Shear stress/shear rate relationship.

Page 373

FIGURE 13 Viscosity/shear rate relationship.

Page 374
polymers are pseudoplastic (i.e., their viscosity decreases with increasing shear rate). Dilatant
materials are those in which viscosity increases with increasing shear rate.
Most laboratory rheometers consist of a basic capillary die through which the molten plastic is
forced. They differ widely in the manner in which the plastic is melted and forced through the
capillary, but they all utilize the same fundamental concepts of rheology. By measuring the
capillary radius and length, the pressure applied to the melt, and the output rate of the
extrudate, the shear stress and shear rate can be calculated from the following equations
[38]:

Due to the many uses of PVC, viscosity must be measured over a wide range of shear rates in
order to predict the flow for various types of processing. Typical shear rates encountered in
processing PVC are as follows:
Compression molding
110 sec-1
Calendering
10102 sec-1
Extrusion
102103 sec-1
Injection molding
103104 sec-1
Figure 14 shows typical pseudoplastic behavior for two PVC samples over a wide range of
shear rates.
The curves show that a laboratory test measuring viscosities of the two compounds at a low
shear rate of 102 sec-1 would give a misleading prediction of comparative processing at a
higher shear rate of 103 see-1. For this reason it is important that the laboratory test measures
the shear rate appropriate for the type of processing planned for the materials. Capillary
rheometers which measure viscosity over a wide range of shear rates have become popular
because test data generated can be applied to various types of processing.
ASTM D3835, Standard Test Method for Rheological Properties of Thermoplastics with a
Capillary Rheometer, includes definitions of rheology terms and procedures for measuring
rheology. When measuring rheology over a

Page 375

FIGURE 14 Shear stress comparison.

range of shear rates, visual observation is also important to detect if there is a critical shear
rate where melt fracture begins to occur. Also, general surface gloss and color observations
may provide worthwhile comparisons between samples [39].
ASTM D1238 describes the measurement of rheology by means of an extrusion plastometer or
melt indexer. Thus a relatively simple instrument measures melt flow at a constant stress by a
dead-weight piston apparatus at a controlled temperature. It measures viscosities only at low
shear rates, which do not represent some of the predominant PVC manufacturing processes.
Other types of capillary rheometers offer the advantage of measuring viscosities over a wider
range of shear rates. Capillary dies can be used in conjunction with an extruder attachment on
the torque rheometer (Figure 10). An Instron tensile tester can be modified with an extrusion
assembly and temperature controls to produce a rheometer utilizing the electromechanical drive
of the machine to control the applied force.
Several popular capillary rheometers use gas pressure to regulate the shear rate of the
material through the capillary. The Sieglaff-McKelvey rheometer uses clean air or nitrogen to
force the heated sample through the capillary. The applied stress is measured by a load cell
and plotted by a recorder. The velocity of the ram, which is proportional to the volumetric flow
rate through the capillary, is also recorded, and viscosity calculations are made from these
data. The Monsanto rheometer also uses air or nitrogen pressure to drive a piston. Plunger
force is measured and a timer measures the time required for the plunger to move through a
certain distance. Viscosities can then be calculated from the data obtained.
Detailed studies of rheology often apply correction factors for test error when measuring the
flow of diverse materials. However, when dealing with similar materials (such as similar PVC
resins in the same formulation), apparent viscosity under identical conditions will generally give
valid comparative data. Therefore, simple comparative testing can be a useful tool in predicting
large-scale processing from small-scale laboratory testing.

Page 376

Desig nation
order

1
2
3

4
5
6

Cell limits
Property

Dilute solution (inherent) viscosity

0.70
Unspecified<0.70
0.79

Sieve analysis, percent through No. 200 (75Unspecified09 1019

m) sieve
144
Unspecified<144
232
Apparent (bulk) density. g/1000 cm3 lb/ft3
9.00
Unspecified<9.00
14.51
Plasticizer sorption, parts DOP phra
Unspecified<50 5074
Dry flow (s/400 cm3)
Unspecified
Conductivity, max. (S/ emg)
Unspecified

3
0.80
0.89
2029
233328
14.51
20.50
7599
<2.0
<6

aphr, per 100 parts of resin.

I.Electricals
Some major application areas for PVC take advantage of its insulative properties.
Measurements of electrical properties has great importance in fields such as wire and cable.
Several types of test are available, ranging from a relatively simple resin test to coating wire
with a PVC compound and measuring the insulation resistance of the jacket.
1. Resin Conductivity: ASTM D1755 lists conductivity as a resin property in cell 6 of its
classification system (Table 3). This test is intended to distinguish between electrical and
nonelectrical grades of unprocessed re-sins. In general, the test will not detect relatively small
differences among different lots of electrical grade resin [40]. Isopropyl alcohol and high-purity
water are added to resin in an Erlenmeyer flask and boiled gently for 5 min. The conductivity of
the water extract is then measured to detect ionic impurities in the resin that would adversely
affect its use for electrical insulation [41].
2. Volume Resistivity: ASTM D257 contains a test procedure for measuring volume resistance
of a specimen and for calculating its volume resistivity. This type of test is often used as a
quality control test for checking raw ma-

Page 377
Cell limits
4

0.900.99
3039
329425
20.5126.50
100124
2.03.9
6

1.001.09
4049
426520
26.5132.50
125149
4.05.9

1.101.19
5059
521616
32.5138.50
150174
6.07.9

1.201.29
6079
617712
38.5144.50
175199
8.09.9

1.301.39
8099
713808
44.5150.50
200244
10

1.40
100
809
50.51
225

terials in an electrical compound. Like conductivity, it can be used to cheek gross differences
between resins (in a duplicate electrical formulation), but may not show minor differences
between electrical-grade resins.
3. Insulation Resistance: The ultimate measurement of electrical properties of resin-compound
is determined by coating wire and then measuring insulation resistance. This is not a simple
procedure, since it involves compounding resin into an electrical formulation, extruding the
compound, coating wire, and finally measuring the insulation resistance of the coated wire.
Typical test methods are those described by regulating agencies such as Underwriters'
Laboratories (UL), Canadian Standards Association (CSA), or National Electrical Manufacturers
Association (NEMA). These tests are conducted on specific temperature-rated compounds.
Each agency publishes standards that outline test procedures and requirements for the specific
wire constructions and temperature ratings.

V.PHYSICAL PROPERTIES: DISPERSION RESINS

Dispersion resins are used in combination with plasticizers and possibly other additives, such as
stabilizers, colorants, fillers, and surfactants. These com-

Page 378
pounds are always of a liquid consistency, which can range from a light syrup to a system as
thick as mayonnaise. It is therefore quite important to characterize this class of resins by their
theological behavior when compounded with plasticizers. In fact, dispersion resins are selected
for an application based to a large extent on their theological characteristics. Other important
characteristics are particle size, air release, volatile content, contamination, and gelation and
fusion.
A.Viscosity of Plastisols at Low Shear Rates
This method of determining the viscosity of a compounded dispersion resin (plastisol) is the
most widely used method of characterizing dispersion resins. The test method is designated by
ASTM as D1824. The instrument used to determine viscosity is the Brookfield[42] viscometer,
which operates in a shear-rate range of only a few reciprocal seconds. An example of a
Brookfield viscometer is illustrated in Figure 15. This low shear rate would translate to
applications involving very slow flow rates of the plastisol during a coating process.

FIGURE 15 Brookfield viscometer, model RVF. (Courtesy of Brookfield Engineering

Laboratories.)

Page 379
A plastisol (or organosol) is prepared for the test. Typically, the resin is compounded in a 100:
60 resin/DOP weight ratio, with or without stabilizer, in a planetary mixer for a predetermined
time. Then it may or may not be deaerated. It is then conditioned for 2 hr at 23C.
The plastisol sample would then be tested for apparent viscosity with the Brookfield viscometer.
There are a number of Brookfield models, such as the RVF and LW (each having four rotational
speeds) and RVT and LW (each having eight rotational speeds). The model RVF is a very
popular viscometer for dispersion resins.
The current ASTM revision suggests obtaining a reading after 2 min of rotation of the sensing
spindle, at each rotational speed tested. This interval permits stabilization of the instrument
mechanism and reading.
Probably the two values of most interest would be the 2-rpm viscosity value and the 20-rpm
value, when using the model RVF viscometer. Resin suppliers will frequently categorize their
resins by the 20-rpm Brookfield viscosity of a 100:60 resin/DOP paste made from the resins.
The 2 rpm/20 rpm viscosity ratio gives an indication of how a resin might behave during dip
coating or spread coating. For example, the higher the ratio, the greater will be the tendency of
the plastisol to not move after it is applied to the substrate (no drain-off or sag).
B.Viscosity of Plastisols at High Shear Rates
This method of determining the viscosity of a compounded dispersion resin (plastisol) is
frequently used to characterize dispersion resins. The test method is designated by ASTM as
D-1823. It is an extremely important test for anyone who is compounding dispersion resin to be
used for a high-speed coating operation or who will be spraying a coating utilizing dispersion
resin. The instrument used for this high shear viscosity determination is a Castor-Severs[43]
viscometer. An example of this instrument is illustrated in Figure 16. This viscometer consists of
a vertical tube approximately 1.5 in. in inside diameter and 12 in. in length as a reservoir for the
plastisol (or organosol). At the lower end of the tube is mounted an orifice of known diameter.
The orifice sizes most typically encountered are 0.15, 0.32, or 0.63 cm inside diameter and are
measured accurately to four decimal places.
The upper end of the tube has a cap containing a fitting to permit nitrogen pressure to be
applied to the upper surface of the plastisol sample in the tube. The nitrogen pressure is
controlled by appropriate regulators, and a three-way valve controls application of the pressure
to the contents to be extruded from the tube.
The test is conducted by extruding a conditioned plastisol or organosol through the calibrated
orifice at a regulated pressure for a specific time. Using the weight of extrudate, the time of
extrusion, the pressure of extrusion, and density of the sample, together with the orifice radius
and orifice length in the appropriate formula to convert to proper dimensional units, the
apparent viscosity can be calculated for applied pressures up to 100 psi. This instrument is
capable of determining viscosities at shear rates in excess of 5000 reciprocal seconds.

C.Particle Size
This method of determining the size of the particles of resin in a PVC dispersion is designated
by the Society of the Plastics Industry, Inc., [44] Vinyl

Page 380

FIGURE 16 Castor-Severs viscometer. (Courtesy of Burrell Corporation.)

Dispersion Division, as SM-VD-T10. However, with only minor modifications, the general testing
procedure is used for determining grinds of paints, lacquers, inks, and resin solutions as well.
The main difference for all these tests lies in the calibration of the gauge that is used as the test
instrument. An example of a Hegman [45] gauge is illustrated in Figure 17. Basically, all the
gauges consist of at least one inclined channel (but may contain more) which gradually vary(ies)
in depth from as much as 5 mils (250 m) to 0 mils. This slanted depth channel is calibrated
along its length.
Sufficient PVC dispersion is placed in the channel (which can range from 1 to 3 in. in width) so
that it would more than fill the wedge that was machined from the block of steel to form the
gauge. A scraper (held vertical to the gauge block) is drawn over the surface of the gauge from
the deep end of the gauge to the shallow end at a uniform rate.
The gauge is immediately viewed to the side such that the line of vision is at right angles to the
length of the channel. Angle of sight to the surface of the gauge should be in the range 20 to
30 to the surface of the gauge.
The grind reading is determined where the first significant number of particles appear on the
surface of paste in the channel. Isolated specks farther up the channel at the deep end may be
ignored.
If the material being checked contains solvent (especially a volatile one), speed of reading the
test is very important, since the surface of the material will be changing rapidly, giving
erroneous results.
D.Air Release
There are three methods of determining the air release property of plastisols that will be
discussed. All are described in the SPI-VD-T20 Statement on Vinyl Dispersion Air Release.
Since dispersion resins are compounded with

Page 381

FIGURE 17 Hegman gauge. (Courtesy of Precision Gauge and Tool Co.)

equipment utilizing varying degrees of shearing action, depending on the formulation used, and
its resultant theological characteristics, it is inevitable that air will be introduced into the
compound.
One method used to determine how this introduced air will be released is called the vacuum
rise method. The test is conducted by placing 5 ml of paste into a 25-ml graduated cylinder.
The cylinder is placed in a vacuum chamber where it can be observed under vacuum. A vacuum
is then pulled on the chamber and the paste in the cylinder. The final volume is noted before the
foam collapses. The percent volume increase is calculated and reported.
A second method of determining air release of a plastisol that may have air introduced during
processing after the compounding step is syringe air release. The test is conducted on a
deaerated paste. The paste is carefully placed in a 4- to 6-oz coated-paper hot cup. Then the
paste is stirred lightly. A syringe is inserted through the side of the cup 1 in. below paste
surface and parallel to the surface. Then 0.2 ml of air is injected into the paste and the syringe
is immediately withdrawn. As soon as the bubble appears at the surface of the paste, a timer is
started, and as soon as the bubble breaks, the timer is stopped. The quicker the bubble
breaks, the better would be the air release of the paste.
A third method of determining air release of a plastisol used for certain molding operations
using predominantly plastisols of lower viscosity ranges is the wet film strength method. This
test involves a specialized piece of equipment involving a steel band 0.75 in. 0.03125 in.
having a series of holes of 3/8, 5/16, 1/4, 3/16, 5/32, 1/8, 3/32, 1/16, 3/64, and 1/32 in. in
diameter. This band is made into a circular configuration 1.875 in. in diameter and a handle is
attached by means of a semicircular band to the cylindrical band containing the holes. Details of
this tester are shown in Figure 18a, and an overall view in Figure 18b.
The test is conducted on a deaerated paste. Again, the paste is stirred lightly before each test.
An appropriate hole is chosen to determine the film break time (one that will give a break time
not excessive in duration). The

Page 382

FIGURE 18a Wet film strength tester. (Courtesy of the Society of Plastics Industry.)

Page 383

FIGURE 18b Tester overall view.

tester is immersed so as to cover the band containing the holes and also the semicircular band
to which the handle is attached. The tester is rotated three times in the paste and then
withdrawn. A timer is started the instant the ring containing the holes emerges from the paste,
and the timer is stopped the instant the film over the chosen test hole breaks.
With all these air release tests, it should be kept in mind that some of the factors affecting air
release are the resin, compound viscosity, surfactants in the formulation, and stabilizers. If the
paste is kept to a simple 100:60 resin/DOP formulation, the tests can be used to evaluate
inherent air release of dispersion resins.
E.Volatiles
Determination of volatile content of dispersion resins may be conducted by ASTM D3030,
Physical PropertiesGeneral Purpose Resins, described earlier in the section.
One variation of this test procedure involves a Rapid Moisture/Volatiles Tester marketed by C.
W. Brabender Instruments, Inc. [43]. The big advantage to using this instrument is that a direct
reading in percent volatiles can be obtained, thus making it very attractive for production quality
control testing.
F.Contamination
Determination of contamination by the mill sheet method may be conducted on dispersion resins
the same as on general-purpose resins, and is as effective a method of determining
contamination as is available.

Page 384
G.Gelation Temperature
This method of determining the gelation temperature of a plastisol is SPI-VD-T18. When
conducted on a paste of 100: 60 resin/DOP, it can be used to characterize gel properties of the
resins. The hot bench consists of a plate approximately 6 in. 18 in. that is heated by electrical
heaters inserted in holes in one end while the other end is being cooled by water circulating
through a hole drilled through its width. This effectively gives a thermal gradient of 125 to
275F, depending on rheostat settings of the heaters. Figure 19 represents a hot bench.
The test is run by casting a film of 6 mils from the cold end to the hot end of the bench. After 60
see a strip of aluminum foil (carefully flattened prior to application) is laid completely over the
film. Moderate pressure must be applied to the foil over its entire area to make certain that the
foil is in contact with the whole film. T18 suggests an aluminum rod 2 in. in diameter and 4 in. in
length that is rolled across the foil over the plates length. The foil is immediately stripped off the
film, going from the cold end to the hot end. The gel point is the point at which no plastisol
adheres to the foil. The temperature can be determined by checking the bench (plate) at the
exact location using a pyrometer. Some pieces of equipment have a series of thermo-couples
located along the gradient to determine the gel temperature.
H.Gelation-Fusion Characteristics
A method of determining the gelation and fusion characteristics of dispersion resins in a
plastisol over a variable temperature range was developed as SPI-VD-T23.
The specific piece of equipment cited in this test reference is the C. W. Brabender [46] torque
rheometer. Originally, the heat transfer medium was hot oil, but now there are electrically
heated mixing heads with air-cooling capabilities available. This test involves an oil-heated head.
Figure 20

FIGURE 19 Gelation hot bench.

Page 385

FIGURE 20 Torque rheometer. (Courtesy of C. W. Brabender Instruments, Inc.)

represents one model of a Brabender torque rheometer. The mixing head consists of two
rollers that are roughly triangular in cross section with flights directing the stock toward the
center of the length of each roller. The rollers are about 1/4 in. from each other and
counterrotate such that the stock is moved from top to bottom of the chamber, passing
between them in the process. The rollers rotate at a ratio of 3/2 rpm, increasing the amount of
mixing action in the process. Another part of the torque rheometer is a controller programmed
to increase the head temperature at a rate of approximately 3C/ min.
A dynamometer is attached to the mixing head to record the resistance of the stock to the
mixing action as torque in meter-kilograms. As the stock is heated, it initially goes to a solid gel,
and then into a fused mass with further heating.
The test is started by preheating the oil to a constant 50C in the oil circulator. Then the mixing
head is turned on and 50 ml of plastisol paste containing adequate heat stabilizer is injected into
the head. In rapid sequence the oil is turned onto the head and torque and temperature charts
are started. At the moment the oil temperature reaches 190C, the programmer is stopped, so
as to continue the test at a constant oil temperature of 190C. The whole test is terminated
after 70 min from the moment the heating had started.
From the torque and temperature charts, a plot of torque versus temperature of stock can be
made. Points of interest from this plot are the gel temperature (point at which torque first
increases), gel rate (slope of torque curve at initial increase), and hot melt viscosity at constant
temperature over last 10 min of test. Polymerization technology used to make any dispersion
resin will give it a characteristic plot for this gelation-fusion plot. Factors affecting this plot are
molecular weight and particle size of resin and their distributions, using a standard plastisol
formulation of 100: 60: 1 resin/DOP Thermolite 20 (or other suitable stabilizer). Hot-melt
viscosity from this tech-

Page 386

FIGURE 21 Torque-temperature curves of plastisols. (Courtesy of Occidental Chemical Corp.)

Page 387
nique is probably more reliable than dilute solution viscosity as an indicator of molecular weight
because occasionally, dispersion resins might contain cross-linked polymer. This would create
solubility problems and subsequent false high readings from the capillary viscometer.
Figure 21 illustrates differences in torque-temperature profiles of resins of similar molecular
weight. Resin A has a greater number of fine particles than does resin B, permitting a faster
rate of gelation.

VI.PHYSICAL PROPERTIES: SOLUTION RESINS

Solution resins, as the name implies, are a group of vinyl resins which are dissolved in ketone
or ether solvents having moderate to high levels of solvency. They are characterized by
molecular weight, bulk density, and volatiles in the same way as general-purpose resins.
A.Viscosity of Solution Resins at Low Shear Rates
This method of determining the viscosity of a solution resin is conducted in the same way as
ASTM D1824. A resin solution is first made up in an appropriate solvent (or blend of solvents).
It would again be brought to a temperature of 23C and permitted to stabilize at this
temperature. Then the actual Brookfield viscosity test would be conducted in the same way as
detailed for dispersion resin testing.
B.Insoluble Particles
Again, the method used for determining insoluble particles using the Hegman gauge would
follow the procedure outlined in SPI-VD-T10. The main difference is that the industry primarily
involved with solution resins is the paint industry. People in this industry almost invariably use a
gauge calibrated in units from 0 (4 mils) to 8 (0 mils). This system is referred to as North
Standard. The gauge design is basically the same; only the numerical system is different.
C.Color and Clarity
In this test, the subject solution is normally compared to some minimum acceptable standard.
For color, the sample should be no darker (usually yellow) than a standard solution. For clarity,
the sample tested should be no more hazy or cloudy than a standard solution.
In the coating industry, other tests for color or clarity are sometimes used. References for
these include ASTM D1209, D1544, and D2090.

VII.ASTM RESIN CLASSIFICATION

A.Polyvinyl Chloride Resins
ASTM D1755, Standard Specification for Poly(Vinyl Chloride) Resins, is commonly used to
establish requirements for homopolymers of PVC powder. It covers general-purpose (GP) and
dispersion (D) resins. The specification provides for as many grades of resin as may be
selected from the possible combinations of requirements [47] (Tables 3 and 4). A grade is
designated by first indicating the type (GP or D), followed by cell numbers for each

Page 388
Cell limits
Desig nation order
Property
1
2
3

Prop erty
order

Dilute solution (inherent) viscosity Unspecified <0.90 0.900.99 1.001.09

Brookfield viscosity (RVF) (poise) Unspecified 024 2549
5074
Severs viscosity (poise)
Unspecified 049 5099
100149

Cell limits
Property

Type of copolymer

Vinyl
Vinylidene
Unspecified
aceate chloride

Concentration of comonomer percent by

Unspecified05
weight of resin

Dilute solution (Inherent) viscosity

4
Bulk density g/liter lb/3
5

Specific gravity, 23/23C

Specific gravity

610

3
Maleic
ester
1115

0.50
0.59
Unspecified<144
144232
233328
14.51
Unspecified<9.00 9.0014.50
20.50
1.360
Unspecified<1.350 1.3501.359
1.369
1.430
1.450
Unspecified
1.4401.449
1.439
1.459
Unspecified<0.40

0.400.49

Page 389
Cell limits
4

1.101.19
7599
150199

1.201.29
100124
200299

1.301.39
125149
300499

1.401.49
150174
500999

1.501.59
175199
10001499

1.60
>199
>1499

Cell limits
4

Aerylonitrile
1620
0.600.60
329425
20.5126.50
1.3701.379
1.4601.469

Polycomponent systems
2125
2630
3135
3640
4145
0.700.79
0.800.89
0.900.99
1.001.10
>1.10
426520
521616
617712
713808
809
26.5132.50 32.5138.50 38.5144.50 44.5150.50 50.51
1.3801.389 1.3901.399 1.4001.409 1.4101.419 1.4201.429
1.4701.479 1.4801.489 1.4901.499 1.5001.509 >1.508

Page 390
Cell code

Property

Limits

Test method

5
1
5
4
4
3

Inherent viscosity
Percent through No. 200 sieve
Apparent (bulk) density (g/1000 cm3)
Plasticizer sorption (phr)
Dry flow (S/400 cm3)
Conductivity (S/cmg)

1.001.09
09
426520
100124
2.03.9
<6

D1243
D1921
D1895
D1755
D1755
D1755

property in the order in which they are listed in the tables. When there is no interest in a
property, a 0 is entered in place of a cell number. The cell number of the first property (inherent
viscosity) is separated from the following numbers by a dash.
Table 3 shows the cell classification system for GP resin. Table 6 shows an example of a GP
resin with a cell classification of GP515443. Table 4 shows a similar system for classifying
dispersion resins by inherent viscosity, Brookfield viscosity, and Severs viscosity. ISO 1060,
Plastics Polyvinyl Chloride Resins Designation, uses the same concept to establish cell
classifications using ISO test methods and requirements.
B.Vinyl Chloride Copolymers
ASTM D2474, Standard Specification for Vinyl Chloride Copolymer Resins, establishes a cell
classification system for copolymer of vinyl chloride. It covers vinyl chloride/vinyl acetate, vinyl
chloride/vinylidene chloride, vinyl chloride/maleic ester, and vinyl chloride/acrvlonitrile
copolymers. Table 5 shows the cell classes which give the type of copolymer, the amount of
comonomer, inherent viscosity, bulk density, and specific gravity. Table 7 shows an example of
a resin C123340.

VIII.QUALITY ASSURANCE
Testing some or all of the resin properties that have been discussed cheeks the consistency of
the product as well as the acceptability of a given segment of production to a certain customers
needs. Table 1 showed a summary of ASTM-ISO methods used to cheek the quality of PVC
resin. Table 8 shows

Page 391
Cell code Property

Limits

1
2
3
3
4
0

Vinyl chloride/vinyl acetate

610
0.500.59
233328
1.3701.379
None

Type of copolymer
Concentration of comonomer (% by weight of resin)
Dilute solution (inherent) viscosity
Bulk density (g/liter)
Specific gravity, 23/23C
Specific gravity, 23/23C

Resin type
Lot number
Inherent viscosity
Bulk density
Screen analysis

Gels, 340 mill

Contamination, 50 g,
Syntron
Plasticizer absorption
Volatiles

ASTM D1243
ASTM D1895, B
ASTM D1921
40 mesh
60 mesh
80 mesh
100 mesh
140 mesh
200 mesh
Pan
3 min
4 min
5 min

ASTM D3367
ASTM D3030

PVC XYZ
84256
0.90%
0.54 g/cm3
0
0.1
12.0
50.4
34.6
2.6
0.3
2
1
0
5
17.2%
0.9%

Page 392
an example of a quality report that a PVC producer might send to a customer, utilizing some of
these test methods. Reference to specific test methods may not always appear on such a
sheet, but test methods should always be clearly understood by the supplier and purchaser.
Even when both parties are using the same test method to cheek a property, discrepancies can
arise due to differences in techniques. If quality reports are to be used as a meaningful tool to
convey information, suppliers and customers should compare test methods, techniques, and
equipment to prevent problems due to test discrepancies.
After establishing what properties are important and which quality control tests are to be
performed, statistics can be applied to a quality assurance program. By properly sampling and
testing resin from a repetitive process such as that in a manufacturing plant, data can be
obtained which should have a statistical pattern of variation. Such data commonly fall into a
normal or bell-shaped distribution curve, as shown in Figure 22.
This distribution can be determined by calculating the arithmetic mean x and the standard
deviation , which is the spread of data around the average. Standard deviation is calculated
using the equation

FIGURE 22 Normal distribution curve.

Page 393
After determining these values, the distribution of test values around the mean, when data fall
into a normal distribution, can be shown by the following relationship:
x 1
68.3% of the data
x 2
95.5% of the data
x 3
99.7% of the data
After determining these values and ranges for a specific property of a resin (such as inherent
viscosity), each new test result can be compared to these values to determine if it is an
acceptable variation or an abnormal occurrence. Typically, x 3 is used to define an unusual
occurrence. This means that there are approximately 3 chances in 1000 that a point located
outside these limits is due to natural variation. Therefore, if a test result is outside these limits,
it is probably due to the production process being out of control and serves as a signal to look
for the cause.
By checking the physical properties of resin as soon as it is manufactured and applying proper
analysis of the test results, the manufacturer can decide if the material is normal and should
be processed normally and shipped to a customer.

REFERENCES
1. American Society for Testing and Materials, Book of ASTM Standards, ASTM,
Philadelphia, 1983.
2. J. Haslam, J. B. Hamilton, and D. C. Squirrell, Analyst (London), 85, 556 (1960).
3. J. Haslam, J. B. Hamilton, and D. C. Squirrell, J. Appl. Chem., 10, 97 (1960).
4. J. Haslam, H. A. Willis, and D. C. Squirrell, Identification and Analysis of Plastics, Ilifec
Books, London, 1972.
5. W. Bauman, An Introduction to Infrared Spectroscopy, Methuen, London, 1962.
6. R. P. Bauman, Absorption Spectroscopy, Wiley, New York, 1962.
7. J. H. Van der Maas, Basic Infrared Spectroscopy, Heyden, London, 1969.
8. Aldrich Chemical Co., C. J. Pochert, The Aldrich Library of Infrared Spectra, Milwaukee,
Wis., 1981.
9. D. O. Hummel, Infrared Analysis of Polymers, Resins and Additives, An Atlas; Vol. 1, Part
2, Wiley-Interscience, New York, 1969.
10. Chicago Society for Coatings Technology, Infrared Spectroscopy Committee, Federation
of Societies for Coatings Technology, An Infrared Spectroscopy Atlas for the Coatings Industry,
Philadelphia, 1980.

11. The Dow Chemical Company, R. A. Nyquist, Infrared Spectra of Plastics and Resins,
Midland, Mich., 1961.
12. E. G. Bartick, J. C. Corbett, and G. L. McClure, Comput. Appl. Appl. Polym. Sci., 185
196 (1982).
13. P. Simak, J. Macromol. Sci. Chem., A17(6), 923933 (1982).
14. R. J. Bell, Introductory Fourier Transform Spectroscopy, Academic Press, New York,
1972.

Page 394
15. P. R. Griffiths, Chemical Infrared Fourier Transform Spectroscopy, Wiley-Interscience,
New York, 1975.
16. E. Von Bezdadeas, D. Braun, E. Burviana, A. Caraculacu, and G. Istrate-Robila, Angew.
Makromol. Chem., 37, 35 (1974).
17. A. Caraculacu, and E. Bezdadeas, J. Polym. Sci., Polym. Chem. Ed., 15, 611 (1977).
18. F. A. Bovey, K. B. Abbas, F. C. Schilling, and W. H. Starnes, Jr., Macromolecules, 8, 437
(1975).
19. Perkin-Elmer Corporation, What Is a Tg? A Review of the Scanning Calorimetry of the
Glass Transition, Thermal Analysis Application Study 7, Norwalk, Conn., 1973.
20. Perkin-Elmer Corporation, Tg and Plasticizers, Thermal Analysis Application Study 11,
Norwalk, Conn., 1973.
21. Perkin-Elmer Corporation, Tg as a Function of Molecular Weight, Thermal Analysis
Application Study 8, Norwalk, Conn., 1973.
22. W. P. Brennan, Thermochim. Acta, 18, 101111 (1977).
23. Encyclopedia of Polymer Science and Technology, Wiley-Interscience, New York, 1964.
24. M. L. Miller, The Structure of Polymers, Reinhold, New York, 1966.
25. Waters Associates, Data Module Model 730 Operators Manual, Milford, Mass., 1981,
App. E.
26. P. Szewezyk, J. Appl. Polym. Sci., 26, 27272741 (1981).
27. ASTM D189569, American Society for Testing and Materials, Philadelphia.
28. P. E. Schwaegerle, SPE Annu. Tech. Papers XXIX, 1983.
29. T. Allen, Particle Size Measurements, Chapman & Hall, London, 1975.
30. ASTM E1181, American Society for Testing and Materials, Philadelphia.
31. ASTM D303079, American Society for Testing and Materials, Philadelphia.
32. ASTM D175581, American Society for Testing and Materials, Philadelphia.
33. ASTM D239679, American Society for Testing and Materials, Philadelphia.
34. ASTM D420282, American Society for Testing and Materials, Philadelphia.

35. L. Weintraub, J. Zufall, and C. A. Heiberger, Polym. Eng. Sci., 8(1), 64 (1968).
36. ASTM D211567, American Society for Testing and Materials, Philadelphia.
37. C. A. Helberger, R. Phillips, and M. J. R. Canton, Polym. Eng. Sci., 9(1) (1969).
38. J. M. Dealy, Rheometers for Molten Plastics, Van Nostrand Reinhold, New York, 1982.
39. ASTM D383579, American Society for Testing and Materials, Philadelphia.
40. ASTM D175581, American Society for Testing and Materials, Philadelphia.
41. J. B. DeCoste, and B. A. Stiratelli, Rubber Age, 87, 279285 (1960).

Page 395
42. Brookfield Engineering Laboratories, Inc., 240 Cushing Street, Stoughton, MA 02072.
43. Burrell Corporation, 2223 Fifth Avenue, Pittsburgh, PA 15219.
44. The Society of the Plastics Industry, Inc., Vinyl Dispersion Division, 250 Park Avenue, New
York, NY 10177.
45. Precision Gauge & Tool Co., 320 East Third Street, Dayton, OH 45402.
46. C. W. Brabender Instruments, Inc., 50 East Wesley Street, South Hackensack, NJ 07606.
47. ASTM D175581, American Society for Testing and Materials, Philadelphia.

Page 396

Page 397

8 Theory of Degradation and Stabilization Mechanisms

DIETRICH BRAUN and EUGEN BEZDADEA
Deutsches Kunststoff-Institut Darmstadt, Federal Republic of Germany

I. INTRODUCTION

397

II. DEGRADATION AND STABILITY OF PVC

398

A. Structural Irregularities

399

B. Dehydrochlorination and Subsequent Reactions

405

III. THERMAL DEGRADATION

411

IV. PHOTOLYTIC DEGRADATION

413

V. STABILIZATION

416

A. Role of Stabilizers

417

B. Stabilization Mechanisms

423

REFERENCES

429

I. INTRODUCTION
Polyvinyl chloride (PVC) is one of the oldest plastics. As a result of the low production costs
and its wide range of applications it is manufactured in enormous quantities. As is known, this
polymer cannot be processed and put to its final application without the use of additivesstabilizers. It is also known that heat and light can lead to a breakdown of this polymer (loss of
mechanical, electrical, and optical properties). Because of the fact that pure low-molecular
structures such as 2, 4-dichloropentane or 2, 4, 6-trichloroheptane are relatively stable organic
compounds (HCl cleavage taking place only slowly at temperatures about 300C) [1, 2], many
studies have been centered around investigating the causes of PVC degradation. A superficial
consideration of the structure of this polymer shows a picture similar to that of its lowermolecular relatives. It consists of 1-chloroethylene units linked head to tail to monomeric vinyl

chloride in a radical polymerization.

Page 398
As a consequence, the polymer should be more stable than it actually is. But as studies on
structural conformation show, there are other conformations to be found in PVC alongside the
normal trans-trans linkages (t-t) [3, 4]. Mitani and Ogata [4] found in nuclear magnetic
resonance (NMR) investigations numerous trans-gauche (t-g) and gauche-trans (g-t) linkages.
More recent investigations [5] of the low-molecular components of a number of industrially
manufactured PVCs showed in oligomers with a molecular weight of 1000 mol/g that the growth
of the PVC macromolecule at the start of the polymerization reaction follows a diadenic
syndiotactic pattern along the first five or six monomer units. In many of these low-molecular
oligomers the isotactic and heterotactic stereochemical possibilities are present to only a very
slight extent (as shown by NMR measurements). During further growth of the molecule, the
state of the high-molecular atactic polymer is reached which represents the various sites of
deviation from the ideal structure and hence the cause of the lower stability.
Apart from the various configurations and conformations, it is possible to detect along the PVC
chain various alien structures such as head-to-head structures, double bonds, branchings, and
groups containing oxygengroups which are totally foreign to PVC (e.g., initiator residuals,
carbonyl or hydroxyl groups). All these alien structures are potential disturbing factors for the
most important breakdown reaction of PVC, the dehydrochlorination (HCl cleavage). In Section
II the structural defects and their origin, together with the influence of HCl cleavage and its
consequences for PVC, are discussed in more detail.

II. DEGRADATION AND STABILITY OF PVC

If unstabilized PVC is exposed to heat (temperatures above 100C), ultraviolet light, or gamma
radiation, then depending on the intensity of the exposure and the type of PVC, a cleavage of
HCl accompanied by polyene sequence formation and cross-linkage along the chain or
intermolecularly can occur [611, 65]. As a result, the polymer discolors from yellow to dark
brown, the solubility in various solvents decreases, and the original mechanical and electrical
properties of the PVC deteriorate rapidly. All these effects are intensified in the presence of
oxygen, but the color of degraded polymer may become less intense in the presence of
oxygen. Considering the fact that the processing of PVC requires temperatures between 170
and 210C and takes place in the presence of oxygen, it becomes understandable why various
stabilizers such as thermostabilizers, UV stabilizers and antioxidants are necessary. These
compounds should generally be compatible with the PVC as well as with many other additives,
present as a result of the numerous applications of PVC in various batch mixtures (e.g.,
softeners, plasticizers, lubricants, colorants, impact modifiers, blowing agents, fillers, etc.). The
selection of a stabilizer for a particular mixture is often made purely by empirical means and is
usually recommended to the processor of the PVC by the manufacturer. Whereas the effect of
a particular stabilizer in a mixture is known, the mechanisms of stabilization, even for the best
known and most established stabilizers, have not yet been fully clarified. An attempt is made
here to discuss the role of structural defects on the aging process of PVC and to interpret in
mechanistic terms the action of the stabilizers in combating aging.

Page 399
A. Structural Irregularities
The most important structural defects in PVC can be divided into two groups according to their
origins:
1. Defects occurring as a result of the polymerization reaction of the monomer. These include
branchings, double bonds, and long-chain ends.
2. Groups containing oxygen resulting from oxidative reactions during both the polymerization
reaction and the manufacture of the crude polymers (predominantly during drying and storage).
These defects include carbonyl, carboxyl, hydroperoxide, and peroxide groups.
Numerous authors [1214] have worked on the determination of the degree of branching of
PVC using infrared (IR) spectroscopy, although the imp14ortant role of this type of structure in
PVC has not always been recognized. Although it has been possible to demonstrate using
macromolecular models [15 16] that the presence of such structures in PVC has a considerable
effect on the breakdown of the polymer, the IR determination of the extent of branching in PVC
has remained quantitatively unsatisfactory [17, 18]. One of the causes for the inadequate
quantitative information on the degree of branching of this polymer was the incomplete reaction
of PVC with LiAlH4 to form a polyethylenelike material [12, 13]. The reduction with tri-n-butyltin
hydride (Bu3SnH) has been shown to be more complete than that with LiAlH4 [19]. This reagent
undergoes a radical reaction with the polymer in two stages [20]:

Using this reduction reaction together with Bu3SnH reduction, the following branchings could be
identified using 13C NMR on the reduced PVC samples [21].

Page 400
Whereas the short chloromethyl branches (1) and the 2-chloroethyl branches (3) were present
in all types of polymer [2023], the long-chain branchings were found only by Starnes et al. [21,
22, 24] in bulk PVC (manufactured at 100C). The short chloromethyl branchings were formed
in the PVC as a result of an intramolecular shift of a head-to-head sequence [20, 25, 26]:

It is to be assumed that such tertiary hydrogen atoms (a) cannot play a role in the thermal
degradation of PVC. Copolymers of vinyl chloride and 3-chloro-penta-l-ene [16] do not show
any difference in degradation behavior in comparison with an ordinary PVC. Hjertberg and
Srvik [27] as well as Braun and Holzer [28] found that in both PVC manufactured under
reduced VC pressure (U-PVC) and suspension PVC (S-PVC), the majority of the branchings in
the PVC are chloromethyl and butyl groups. In ordinary PVC, CH2Cl and CH2CHCl
CH2CH2Cl branches were first shown to predominate by other workers [29, 30]. The
formation of butyl side chains in PVC can be the result of a back-biting mechanism [31]:

whereby (b) with a new monomer unit (propagation reaction to the monomer) secures further
growth of the polymer chain:

Starnes et al. [19, 32] found long branches with over four carbon atoms bearing tertiary
chlorine atoms at the C branching.
All these new investigations with 13C NMR, in part with deuterated polymers, were able to
show the presence of tertiary chlorine atoms in the PVC independent of the degree of
branching reported (on average, one to five side chains per 1000 monomer units). These Cl
atoms, which can be present at a concentration of 0.5 to 1 per 1000 monomer units in the PVC

Page 401
[5], are in all cases probably at least partially responsible for the thermal degradation [21, 27,
30].
Many authors working on the thermal stability of PVC place greatest importance on the allyl
chloride groups in the PVC for the degradation of these polymers [33, 34]. The search for
these groups, known from low-molecular organic compounds to be labile, has made use of
chemical and spectroscopic means. The quantitative determination by chemical means of the
chloro-allyl groups in PVC is based on several reactions:
1. Substitution reaction of the allyl chlorine atom with a specific substitution agent, such as
cadmium acetate [35, 36] or phenol [37, 38].
2. Additive reaction, which can be used to determine the total number of double bonds in PVC.
Here there is the possibility of bromination [39], microhalogenation [40], or reaction with iodine
chloride [41].
3. Oxidation reaction with potassium permanganate (KMn04) [42] or ozone [4345].
The most important spectroscopic methods (both qualitative and quantitative) for determining
these groups are 1H NMR and 13C NMR spectroscopy [4649]. Using these methods, two
types of allyl chloride groups were identified in PVC:
(I) ~ CH2CH=CHCH2Cl
(end double bond)
(II) ~ CH2CH=CHCHCl~
(internal double bond)
These chloroallyl groups were found in practically all types of PVC. The number of these
groups varies with the nature (temperature and method of manufacture) of the polymer under
investigation [5].
The contents of these groups increase more in the low-molecular parts of the polymer. A strong
increase in the allyl chlorine atoms could be shown in U-PVC [46, 50]. This fact explains why at
the end of the polymerization reaction in an environment of reduced monomer concentration
(monomer starvation) the number of transfer reactions to the monomer increases immensely. In
one U-PVC that was investigated, the allyl isomer of type II was found [45]:

However, this type of grouping could not be identified in polymers manufactured in the
conventional manner. The chloroallyl group found most (highest concentration) is type I [5, 21,
46, 45], which is present in PVC at a concentration of around 1 per macromolecule on average.
The internal double bond of type II is present at a concentration of one-fifth to one-third of that
of type I [5, 45]. As a result of insight into the known dehydrochlorination mechanism of PVC
(zip reaction),

Page 402
the chloroallylic grouping of type I is considered not to play a decisive role in the degradation
process of PVC. It is, however, perfectly possiblealthough very difficult to provethat
intermolecular HCl elimination can lead to cross-linking within the polymer by these groups. The
structure of type II can follow a zip reaction leading to a breakdown of the polymer, but from
the NMR tracings it is unfortunately not possible to determine how far from the chain end
internal chloroallyl groups are. Hence it is not possible to determine, to any degree of accuracy,
the level of participation of this grouping in the degradation process. By ozonolysis an internal
double-bond content of about 0.2 C=C per 1000 VC was found [51]a result which again tells
us nothing about the labile allyl chlorine atoms.
The formation of the predominant allyl structure of type I in the PVC is based on a shift of the
higher-energy head-to-head sequence [20, 29]:

According to Hjertberg and Srvik [46], such chlorine atoms liberated during the polymerization
reaction can assist in the formation of internal chloroallyl groups:

The presence of the head-to-head sequence in PVC is disputed considering the molecular
transfer reactions and the formation of double bonds, despite the fact that Mitani et al. [52] did
show their presence in the extent of 0.6% (referred to the C atoms within the chain) using
iodometric analysis [24, 30]. The investigations on macromolecular models (chlorinated
polybutadiene) [53] showed that the head-to-head sequence can considerably lower the
thermal stability of the polymer. Braun [54] was able to give a detailed analysis of the most
important investigations on the head-to-head sequence. From this it can be inferred that the
role of these structures in the degradation process of PVC is of no significance.
A more important group of structural irregularities in PVC are the structures containing oxygen,
which can be formed during the polymerization [55], manufacture (drying), and storage of the
polymers. More recent investigations by Braun and Sonderhof [56, 57] showed that in the
suspension polymerization of vinyl chloride with oxygen, both HCl and formaldehyde can be
formed (Figure 1). The same authors have also copolymerized vinyl chloride with carbon
monoxide (suspension polymerization) to ascertain the role of the carbonyl groups built into
PVC during the degradation process. From Table 1 it can be seen that with increasing 02
concentration in the monomer, there is an increase in the carbonyl and peroxide groups
detected by IR spectroscopy as well as in the number of internal double bonds detect-

Page 403

FIGURE 1 Chloride ion and formaldehyde formation in the aqueous phase during vinyl
chloride/oxygen suspension polymerization (see Table 1).
ed by ozonolysis, while simultaneously there is a drop in the molecular weight (Mn). As the
carbon monoxide concentration increases, so does the number of CO groups in the polymer
(Table 2). However, a drop in thermal stability of the VC/CO copolymer was not ascertained;
rather, the reverse was found (Figure 2). The VC/CO copolymers showed with increasing
peroxide concentration a drop, as expected, of the thermal stability of the samples being tested
[56]. Hjertberg and Srvik [46] using NMR spectroscopy were not able to confirm the presence
of conjugated ketoallyl groups of the type
2 in feed (mol %)
in polymer (mol %) Mn Internal double bonds (mol %) 12 0 0 53,100
0.008 13 0.057 0.048 44,900 0.0069 14 0.155 0.080 43,400 0.0086 15 0.660 0.093 42,200
0.0210 16 1.770 0.228 36,400 0.0400 17 3.430 0.230 40,000 0.0430 18 4.750 0.260 35,450
0.0510

Page 404
co in feed (atm) CO dissolved in VC, 50C (mol %)
in polymer (mol %) Mn Internal double
bonds (mol %) 1 0 0 0 49,900 0.011 2 1.0 0.3 0.12 51,200 0.014 3 2.0 0.6 0.25 46,800 0.027
4 2.4 0.68 0.3 48,500 0.021 5 3.3 1.0 0.46 48,700 0.012 6 4.2 1.2 0.5 49,900 0.033 7 6.1
1.75 0.74 44,100 0.013 8 7.9 2.25 0.94 26,000 0.046 9 11.5 3.22 1.33 43,600 0.015 10 15.8
4.5 1.7 43,100 0.024 11 19.3 5.5 1.97 44,600 0.018
~COCH=CHCHClCH2~
which Minsker, Svetly et al. [58, 59] hold responsible for the thermal degradation of PVC. The
carbonyl groups are apparently present in PVC as both keto groups [60, 61] and
chlorocarbonyl groups [62]. This was confirmed

FIGURE 2 Thermal dehydrochlorination of vinyl chloride/CO copolymers (see Table 2).

Page 405
by Wescott et al. [63] in their investigations in which they reduced sequentially VC/CO
copolymers (prepared under various CO concentrations) with LiAlH4 and Bu3SnH to a
polyethylenelike material and demonstrated by 13C NMR spectroscopy the
groups in the reduced PVC.
From this it can be deduced that the mechanism suggested by Ratti et al. [62] for the formation
of chlorocarbonyl groups in PVC is correct:

As already stated, Braun and Sonderhof [57] were able to verify on the basis of these
copolymers with incorporated chlorocarbonyl groups that the presence of such structures does
not contribute to thermal instability of the polymer.
B. Dehydrochlorination and Subsequent Reactions
The dehydrochlorination of PVC is a process which continues as long as the source of
degradation (heat, light, gamma radiation) is present. As a rule, dehydrochlorination reaction
investigations are carried out on unstabilized polymers using conductometric analyses of the
HCl solution formed during the degradation. The conductivity is a function of the HCl
concentration of the solution measured (Figure 3).
The most common method for investigating the stability of a PVC sample is the method of
Braun and Thallmaier [64], which is easy to perform and permits rapid determination of the rate
of degradation of a stabilized or unstabilized sample at different temperatures of degradation.
Using this method it was possible to determine exactly the course of degradation of unstabilized

FIGURE 3 Conductivity measurement of splitting of HCl during thermal degradation.

Page 406

FIGURE 4 Dependence of dehydrochlorination on the weight of the disk (total HCl elimination =
66.6 x 10-5 g).
PVC both thermally and photolytically in nitrogen, air or N2/O2 mixtures [6567]. The thermal
degradation takes place throughout the entire PVC sample. In N2 both PVC disks and foils
cleave HCl at a constant rate independent of the mass. The degradation follows a hyperbolic
course in this case (Figure 4). In air the thermal degradation is a nonlinear process with
increasing rate of degradation with time (Figure 5). Here it could be ascertained that HCl
elimination and oxidation (atmospheric oxygen) are not competing processes.
Photolytic degradation takes place fundamentally at the surface of the sample. In this case only
those layers are degraded which are in direct contact with the irradiation (photons) impacting
on the surface. The intermolecular cross-linkages directly formed here are much more
numerous in terms of the eliminated HCl as compared to those formed during the thermal
degradation. The rate of degradation of PVC samples exposed to light increases

FIGURE 5 Rate of HCl elimination (wt % HCl /hr) as a function of the degradation time during
thermal degradation at 180C. , 19.8 x 10-3 g-foil, under N2; , 15.4 x 10-3 g-disk, under N2;
o, 21.8 x 10-3 g-disk, under air; , 17.3 x 10-3 g-foil, under air.

Page 407
rapidly both in N2 and in air (or in other N2/02 mixtures). Depending on the source of light used,
it could be shown [67] that not only were the chlorine atoms present at defect sites in the
surface layer eliminated, but also that other chlorine atoms could be cleaved in this way. Both
an increase in temperature and the presence of air produce an increase in HCl production
compared to that at room temperature in a nitrogen atmosphere. As soon as the surface has
been degraded and cross-linked, it forms a protective layer so that after a few hours the
degradation ceases entirely (Figure 6). The degraded surface of the sample is brittle and
disintegrates on trying to remove the foil, producing a fibrous material, which as a consequence
of the extensive cross-linking is insoluble in any solvent. An insoluble portion is formed after only
short exposure to UV light (0.1 to 0.2% HCl loss).
As a result of the dehydrochlorination in the absence of air, polyene sequences occur in the
polymer, immaterial of the mechanism according to which the HCl release takes place (ionic or
free radical):

The longer polyene sequences have a basic character and can bind a portion of the cleaved
HCl [68]. This type of ionic complex (initially probably a type -p) has the properties of an
onium salt and contributes to the intensification of the color of the degraded polymer
(discoloration):

FIGURE 6 Decrease of dehydrochlorination rate with photodegradation time and temperature

both under air and under N2. , 20C, under N2; , 90C, under N2; 20C, under air; , 90C,
under air.

Page 408
This fact must be (and is) taken into consideration for the stabilization (depending on the
formulation and the usage of the end product).
In earlier years the autocatalytic effect of the cleaved HCl had been discovered. The
investigations by Druesedow [69] showed that the remaining HCl in the sample can in the
presence of atmospheric oxygen lead to an intensification of the degradation. More recent
investigations [65, 67] have shown that in all the samples degraded (both thermally and
photolytically), even after continuous removal of the HCl using N2 or argon, there is still a
residue of HCl in the degraded sample, which can be shown by a change in conductivity. As
discussed earlier, the hydroperoxides and peroxides contained in the polymers contribute to the
thermal degradation of the polymer. Hence antioxidation is an important factor in preventing
thermal degradation of the polymer [7072]. If the oxygen attacks the methylene groups,
hydroperoxides are formed which react further to cleave more HCl or to form keto groups [73,
74]. The oxygen reacts upon formation of a free radical (e.g., as a result of a transfer reaction
to the polymer):

Although the resulting ketone, as already discussed, plays no role in thermal degradation, it
could play an important role as energy carrier in the photolytic degradation [75]. The resulting
chlorine radical can in any case contribute to HCl elimination from the polymer. Decker [76], on
the other hand, provided strong evidence to indicate that olefinic unsaturation is the principal
chromophoric involved in the photoinitiation process. This mechanism has admittedly not been
directly proven to date; however, there are authors who have discussed the possibility of the
oxygen attacking the polyene sequences formed [75, 77, 78]. In this case both hydroperoxides
(from the longer polyene sequences) and cyclic peroxides are postulated by some authors:

For this reason the polymers are protected with antioxidants, which can react with O2 during
manufacture and processing or can neutralize the resultant radicals.
During processing, mechanical exposure can produce fragments (radicals), which have been
discussed by Scott [79]. These radicals, in turn, contain labile chlorine atoms at the sites of
mechanical damage and can thus lead to direct dehydrochlorination or to the effects to oxygen,
as discussed.

Page 409
A large number of mechanisms for dehydrochlorination have been described in the literature. All
free-radical reactions are more or less variants of the degradation mechanism of Winkler [80].
Various chemical reactions, predominantly the electron spin resonance measurements [81, 82]
have supported to some extent this free-radical mechanism. A series of facts, such as the
ineffectiveness of radical acceptors during degradation, the action of numerous ionic stabilizers,
or the acid-base catalysis of the dehydrochlorination, have not been able to support this
mechanism. Predominantly, however, because no other degradation products except HCl are
present, and in view of the autocatalytic effect of the HCl already cleaved off on the further
elimination of HCl, a doubt does arise that a free-radical mechanism applies here. An ionic
degradation mechanism is then considered [21]. The difficulties in explaining the
dehydrochlorination reaction occur mainly because there are few proofs of the initiation
processes (the starting step of the dehydrochlorination reaction).
The initiation reaction of the dehydrochlorination can be caused, as described above, by the
labile sites in the PVC. A starting reaction through homolytic cleavage of the secondary C-Cl
bond (the majority of the bonds in the chain) would in energy terms be possible only if the
degradation temperature were to rise above 200C (the bond strength of secondary C-Cl is
about 300 kJ/mol) [83]. However, it would appear for statistical reasons (the number of these
bonds is extremely high) to be possible that such weak secondary C-Cl bonds do exist in PVC,
and evidence for this has been presented by Starnes et al. [8486]. Based on mechanisms
from Rieche et al. [87] and Marks [88], it can be assumed that the first steps of
dehydrochlorination are of an ionic nature. The investigations using model substances in solution
and in substance support the ionic mechanism, which lies in the nature of the polymer chain
itself [89] (the distribution of the bonding electrons polarizes the chain sequentially) or is to be
found in a statistical HCl elimination as a result of a unimolecular reaction which proceeds via an
activated complex [90]:

(1) can naturally secure the continuation of the dehydrochlorination via the chloroallyl groups,
whereas (2) was not found in PVC.
The cleaved HCl contributes to a stronger polarization of the C-Cl bond and thus permits further
ionization, which results in the formation of stronger mesomeric allyl cations:

Page 410

The resulting polyene-HCl complex, in which the HCl2 occurs as a stabilizing anion,
disintegrates when the polyene growth has come to a stop through stabilization of the
resonance. In this way the polyene cation is neutralized. More recent investigations by Haddon
and Starnes [91, 92] support this mechanism. The authors calculated the geometry, conjugation
energy, and -orbital atomic charge density of the polyene sequences with 11 carbon atoms in
the conjugated system, and as a result of these calculations, found the formation of such ion
pairs to be possible:

Starnes and Edelson [93] were able to explain the formation of aromatic compounds during the
thermal degradation of PVC (benzene, toluene):

The cyclization of the polyenes can reduce the length of the polyene sequences, and when
cyclizations occur at the ends of the polyene sequences, they can contribute to an inhibition of
the dehydrochlorination [93].
One degradation mechanism in three stages was suggested by Shapiro et al. [94, 95]. After
unimolecular HCl elimination in the first stage, a 1, 4-cis-HCl elimination from a six-component
intermediary complex can ensure in the second stage to form conjugated double bonds. In the
final stage the HCl already cleaved off catalyzes further conversion of the resultant conjugated
diene into two isolated double bonds which through further HCl cleavage (repeat of stage 2)
can lead to the chromophoric polyenes. This mechanism, which provides a good explanation for
the discoloration of the degraded PVC, does contradict the proven isomerization of the
chloroallyl structures, which leads to an increase in stability [96] and not (as according to this
mechanism) to an elimination of HCl.
The termination reactions in the degradation of PVC have not been investigated to a great
extent. Generally speaking, the discontinuation of the growth of the polyene sequences as a
result of the stabilization through the resonance already described here is the most important
intramolecular discontinuation process. The radical transfer between chain radical and C=C

Page 411
double bond cannot be regarded as a discontinuation reaction of the degradation since it
probably takes place during the entire degradation process. Apart from the discolorations
(formation of polyene sequences), the cross-linkings occurring in the degraded polymers are
the most important consequences of the degradation reaction. The cross-linkings are formed in
the polymers both as a result of intermolecular radical recombination as well as through the
intermolecular elimination of HCl. This type of reaction has not been thoroughly investigated. It
is entirely possible under the conditions of degradation [5] that it plays an important role in the
degradation process of PVC.
The chloroallyl structures
with a terminal allyl chlorine atom in the PVC are unlikely
to be able to start an intramolecular dehydro-chlorination, although they represent the main
quantity of double bonds in the PVC. For this reason they are not considered in more recent
publications [46] to be the most important cause of the degradation process among the labile
sites in the polymer. These structures are at least theoretically capable of starting an
intermolecular dehydrochlorination. This hypothesis can be supported by the fact that, as
reported, the cross-linkages have been found in the degraded polymer directly after the start of
the dehydrochlorination and not only at a more advanced stage of degradation [67].
Other means of formation of cross-linkages are the Diels-Alder reactions, which are feasible
both intramolecularly and intermolecularly between neighboring unsaturated PVC sequences.
These were suggested long ago by Druesedow [97] as a possibility for cross-linkage. This
mechanism is supported by the finding that the double-bond concentration increases more
slowly during the degradation process than the eliminated HCl that is measured [98].

III. THERMAL DEGRADATION

In the absence of atmospheric oxygen a nonoxidative degradation takes place during which the
polymer loses HCl and undergoes cross-linking and polyene sequence formation. This was
discussed in Section II under dehydrochlorination and following reactions. It has already been
mentioned that in the presence of air the dehydrochlorination is a nonlinear process with a
tendency to increase in rate in the course of the reaction, and that the polyene sequences can
react with 02 to produce other structures that favor degradation.
For this reason degradation in the presence of heat and air is termed thermo-oxidative
degradation. The most important features were summarized by Reicherdt [99] using results
from the literature:
1. In air the HCl loss is double (in comparison with the HCl loss in N2), whereas in pure oxygen
at 190C it is four times higher.
2. Oxygen accelerates the dehydrochlorination by autocatalysis.
3. The activation energy during HCl loss in air is reduced by about one-third from the value in

N2 or in vacuo.
4. The discoloration of the polymer degraded in the presence of air is less distinct than in the
presence of N2 or in vacuo.
5. The chain cleavage reaction competes with the cross-linking at the

Page 412
beginning of the degradation; in the further course of the degradation the cross-linking
increases also in the presence of air.
6. In the polymer, groups containing 02 increase (e.g., hydroperoxides, peroxides, ketones, and
carboxyl groups).
The mechanism of thermo-oxidative degradation of PVC is easily explained in terms of the
general oxidation of saturated hydrocarbons [100]. The oxygen molecule, which is
paramagnetic (two unpaired electrons of parallel spin), behaves like a diradical with a triplet
structure. This has a strong affinity for radicals and can thus bind all the radicals necessary for
continuation of the reaction, according to
The peroxy radical thus formed can initiate autoxidation reactions:
This general reaction scheme explains many of the oxygen-containing groups in polymers:

Although the thermo-oxidative reaction mechanism of PVC has not yet been clarified
completely, it is possible to state on the basis of numerous observations that the first step is
the dehydrochlorination of the polymer [101103]. The more polyene sequences are formed,
the higher is the measured HCl loss [101]. This means that oxidation of the polyene sequences
is accompanied by a higher HCl loss. The dehydrochlorination increases with the increase in
partial O2 pressure [104]. The oxidation of the polyenes thus formed can lead to five- or sixmembered cyclic peroxides:

Tds et al. [104] have described the thermo-oxidative degradation process as follows:

Page 413
These authors describe the further HCl elimination in kinetic terms according to the following
equation:

where is the conversion measured by HCl loss, t the time, and Ks the steady-state rate
extrapolated to the beginning of the degradation. K and are the formal constants.
Using the Charlesby-Pinner plot [105] it is possible to work out a measurement of the outcome
of thermo-oxidative degradation in PVC, of the cross-linking, and of the cleavages in the chain.
There was found [104] in pure O2 a rate of cleavage of 1.4 to 13 x 10-6 min-1 at 180C and a
cross-linking rate of 3 to 20 x 10-6 min-1. This result shows that both reactions ensue in
parallel. Despite oxidation, which causes cleavage of the chain, the cross-linkage remains the
more important reaction, as was the case for degradation in N2 [106].
The big question still to be answered on thermo-oxidative degradation of PVC is how to obtain
a quantitative relationship between the parameters of thermo-oxidative degradation
(temperature and oxygen) and the loss of the mechanical, electrical, and optical properties of
the polymer. It has to be assumed that there is a similarity in terms of mechanism of formation
of O2-containing groups between thermo-oxidative degradation of PVC and degradation of
PVC under gamma radiation. Braun [54] found the mechanisms of Gupta and St. Pierre [71]
(thermo-oxidative degradation) and Decker [107] (gamma radiation) to be similar both in the
following products and in the radical character.

IV. PHOTOLYTIC DEGRADATION

The course of dehydrochlorination under ultraviolet (UV) light of unstabilized PVC samples in the
form of disks and foils has already been described. It is important to note here that the
description of the photolysis mechanism of PVC can also be valid for the radiolysis of the
polymer (gamma radiation) since parallel occurrence of both types of degradation reaction has
been described [107109]. Degradation under ionizing irradiation takes place much quicker than
photolytic degradation. Golub and Parker [109] reported that the radiolysis is some 13 times
faster than photolysis in terms of dehydrochlorination. Reported here are the mechanism and
the following reactions of photolysis, not only because of the similarity of the mechanisms of
both types of degradation, but also because of the greater practical significance in terms of
stabilization of photolytic degradation.
Photolysis in the absence of air leads to formation of long polyene sequences and to very
extensive cross-linking within the polymer [65, 110]. Also, breaks in the chain have been
reported which do, however, depend on the wavelength of the UV source [10]. Compared to
photo-oxidation (photolytic degradation in the presence of O2), the dehydrochlorination at the
beginning of the photolysis in N2 is very extensive [65]. The rate of degradation then declines.
The formation of long polyene sequences blocks a further HCl elimination by increasing
resonance stabilization. The cross-linking occurring with the simultaneous discoloration of the

surface (at room tempera-

Page 414
ture under a UV mercury vapor lamp the PVC foil becomes dark brown after only 15 min in a
N2 atmosphere) absorbs to an increasing extent the energy of the radiation. A protective layer
forms on the surface, leaving the underlying layers unaffected. The material soluble in
tetrahydrofuran (THF) after degradation has the same molecular weight and the same structure
as the starting polymer. An increase in temperature from 20C to 90C increases the quantity
of HCl detectable by conductometry by about 50% (photolytic degradation in N2). If photolytic
degradation takes place in the presence of oxygen, the HCl elimination is slower at the
beginning of the process than in N2 and is significantly slower at room temperature than at
elevated temperature. It would appear that in the first 2 hr of degradation under the UV
mercury lamp, the air protects the PVC foil from degradation. This illusion stems from the
amount of HCl measured and from the weak yellow discoloration of degraded foil. The
insolubles from the foil examined after this degradation time show, however, that despite the
slight dehydrochlorination, the chemical changes of the material are more prominent in air than
in nitrogen [65, 67]. As the degradation time increases (Figure 7), the rate of the HCl
elimination in air increases more than in nitrogen. This fact lends proof to oxidations occurring
parallel to the elimination of HCl, especially on the polyene sequences formed during
degradation, which are of the same nature as those produced during thermo-oxidative
degradation. The cleavage of HCl increases both with an increase in reaction temperature and
with an increase in the O2 concentration. More oxygen and more heat cause, as expected, a
more rapid oxidation reaction and a rapid deterioration of the original material (the amount of
substances insoluble in THF also increases) [65, 111].

FIGURE 7 Photodegradation of bulk-PVC foils at room temperature under (O)N2. , Air; , O2;
, 50% O2/50% N2.

Page 415

FIGURE 8 UV spectra of bulk-PVC foils photodegraded under (a) 100% N2; (b) 95% N2/5%
O2; (c) air; (d) 70% N2/30% O2; (e) 50% N2/50% O2; (f) 100% O2.
The UV spectra of the degraded surfaces of the PVC samples show that the longer polyene
sequences are to be found in the polymers irradiated in nitrogen (Figure 8) [65]. The total
number of double bonds in the polyene sequences is probably higher in samples degraded in
oxygen than in nitrogen. The weaker absorption of UV light by the PVC in oxygen is attributed
to the formation of short , -unsaturated ketones
, which have a lower extinction
coefficient (absorptivity) than the polyenes and can overlap the UV maxima of the polyenes in
the UV spectrum.
In substances insoluble in THF (prepared by degradation in various N2/O2 mixtures under UV
radiation), it was possible by IR spectroscopic analysis to detect carboxyl, ketone, and
hydroxyl groups. The carbonyl groups increased with temperature and O2 concentration (Figure
9) [65]. The light discoloration of the polymer degraded in oxygen compared to the dark color
on irradiation in nitrogen can be explained by the cross-linking of the polymer sequences formed
in the reaction under the action of oxygen together with simultaneous formation of other groups
;

, as a result of which further elimination of HCl is avoided.

The material soluble in THF produced by photolytic degradation of various PVC samples in
various N2/O2 gas mixtures shows no change in the original molecular weight, but shows
formation of carbonyl groups by oxidation of CHCl groups as demonstrated by Kwei [112]
and recently by Davidson [113]. This shows that the O2 penetrates and reacts with the material
not attacked by the UV radiation. A general scheme for the radiation process is as follows
[113]:

Page 416

FIGURE 9 Dependence of carbonyl concentration (IR absorption at 1725 cm-1) formed in the
photodegraded bulk-PVC foils upon the irradiation time and temperature. , O2; , 50%
N2/50% O2; , air; , 70% N2/30% O2; , 90C; -, 40C.

V. STABILIZATION
The structural defects of PVC previously discussed and the most important degradation
mechanisms are not sufficient to clarify the role of the stabilizers being used. An important
aspect in the stabilization of PVC is also the action of the contaminants of the PVC picked up
during the polymerization process.

Page 417
These include polymerization material and material introduced during subsequent processing
(iron, iron chlorides, emulsifiers, etc.). The dehydrochlorination and cross-linking effect of FeCl3
on polyene sequences in PVC, which incorporates an autocatalytic effect, has also been
reported [106]:

Such effects must also be considered in the stabilization. It is thus necessary in some cases to
test the polymer for traces of FeCl3 using x-ray fluorometry or another rapid analytical
procedure. The use of stabilizers is considered an absolute necessity for the guaranteed
durability of a manufactured product as well as for the actual manufacturing process, especially
under consideration of the dehydrochlorination reactions described here.
A. Role of Stabilizers
Depending on the desired effect, PVC thermostabilizers have two basic functions to fulfill [114,
117]: (1) a preventive function and (2) a curative function.
la. The most familiar preventive function of a stabilizer is the binding of HCl split off in whatever
way. Cadmium stearate, for example, can bind HCl as follows:

This function alone is not sufficient for stabilization, since then many compounds would have the
character of a stabilizer, which is not the case: for example, calcium carbonate:
1b. The second preventive function of a stabilizer is the exclusion of labile starting sites for
dehydrochlorination in the polymer. For example, the organotin mercaptides can remove the
labile allyl chlorine atoms by substitution:

Page 418

In this case, too, as already described under the determination of such labile sites, not every
compound that reacts with such sites can be used as a stabilizer, for example, phenol:

1c. Finally, the stabilizer must be capable of suppressing auto-oxidation. An antioxidant is a

special additive, generally in the class of sterically blocked phenols or thiophenols, (e.g., 4, 4dimethyl-6, 6-di-tert-butyl- 2, 2-thio-bisphenol). This compound is known to be a good
antioxidant but has, however, no effect as a thermostabilizer and can protect
dehydrochlorination only above a certain concentration. The organotin compounds are capable
of functioning as an antioxidant as the reaction with organic hydroperoxides has proven [115,
127]. As a result of the reaction organotin oxides are formed, as are alcohols and organic
disulfides, for example:

2. The curative function of a stabilizer protects the polymer against thermal damage which has
already taken place for some reason (e.g., during processing). The better the curative action of
the stabilizer, the longer a product will retain its original properties. The curative actions of a
stabilizer are:
2a. The addition to already formed polyene sequences. In this way the stabilizer interrupts the
growth of a polyene sequence and/or contributes to the shortening of these polyene
sequences. This ensures a retention of the original color of the product. The best stabilizers in
this respect are the tin stabilizers. Troitskaya and Troitski [116] have demonstrated the DielsAlder reaction of polyene sequences of the polymer with dialkyltin maleate:

Page 419

2b. The reaction of the stabilizer with the already formed onium salts is the second curative
function. As described under dehydrochlorination, a polyene sequence binds HCl by strong
complex formation. The conversion of the onium salt by binding the HCl to the stabilizer and
reconstituting the polyene sequences can be carried out by all stabilizers, since the binding of
HCl is the basic condition needed to be able to stipulate whether a compound is suitable as a
stabilizer. Whereas some classes of stabilizers (e. g., organotin mercaptides) possess all
described functions, other stabilizers fulfill only a certain number of these functions; for this
reason, depending on final use and cost-effectiveness, stabilizer systems that consist of several
components find use [117].
In practice, right at the polymerization stage various possibilities are utilized to save a
preventive and also a curative function of the stabilizers later to be used (thus allowing a drop in
the quantity of stabilizer) and in this way to give the polymer itself a certain stabilizing function.
These possibilities include:
1. The use of chemically pure monomers and/or other chemicals which come into contact with
the polymer, such as water, suspension additives, solvents, emulsifiers, initiators, and chain
transfer agents. Here it is recommended to utilize radiation initiation instead of organic initiators
[114].
2. Another way of carrying out preventive stabilization without stabilizers is to add small
amounts of ethylene or propylene (1 to 3%) [114] for copolymerization with VC, with the
purpose of interrupting HCl elimination by the small comonomer blocks formed. In this way
better stabilization of the polymer is achieved from the start.
Such practices have also been investigated with a view to achieving a curative stabilization. The
most important possibility here is the polymerization of vinyl chloride with small amounts of
comonomers with stabilizer properties with the intention of building a stabilizer effect into the
chain. This kind of attempt at copolymerization with lead carboxylates of acrylic and undecylenic acids [118], salicylate groups [119, 120], and so on, has shown a certain
improvement in thermal and UV stability over homopolymers. Despite all these possibilities, it
has not yet been feasible to process mixtures with homopolymers and copolymers of vinyl
chloride without stabilizers being added. The choice of a stabilizer for a particular mixture is
generally dependent on two factors:
1. Influence of the stabilizer on the processing properties of the compound; the mixture should
on economic reasons alone be easy to process at the required temperature and for the
required throughput.

Page 420
2. The formulation of the PVC mixture; the stabilizer should be compatible with all other
components of the mixture, so that subsequently no leaching from the finished product takes
place from the stabilizer itself or other components, nor should plate-out occur. On the other
hand, the intensity of the action and the quantity of the stabilizer should be matched to the
processing conditions (temperature): A plasticized mixture requires less stabilizer than a rigid
PVC mixture.
The most important classes of thermostabilizers can be briefly summarized as follows:
1. Organotin stabilizers, the most effective in their action and at the same time the most
effective stabilizer systems used in practice, are subdivided into:
a. Sulfur-containing tin stabilizers (organotin mercaptides and sulfides) with the general formula:

Because of their excellent efficiency, they are used in situations where temperatures of 200C
and above occur. They can be used in all formulations, but because of the absence of a
lubricative function, they have to be used with lubricants. They cannot be used with cadmium or
lead stabilizers or pigments because of the danger of formation of cadmium or lead sulfides,
which can lead to discoloration of the product. Because of the minimal migration ascertained,
they can be used in mixtures for foodstuff packaging and for pipes for drinking water. In the
manufacture of high-transparency foils, this stabilizer class can be used in practice only when
UV absorbers are included.
b. Sulfur-free tin stabilizers (organotin carboxylates or salts of alkyltin oxide with maleic acid or
partially esterified maleic acid) with the general formula [117]

In contrast to the sulfurous tin compounds, the tin carboxylates have to be used with
antioxidants. The advantage of the use of these stabilizers over those containing sulfur is the
good photostability and the lack of odor. These stabilizers, too, need additional lubricants, since
they tend to stick to the metal parts of the calenders and the extruders.
2. Barium-cadmium stabilizers with the general structural formula [117]

Page 421
The two compounds are used because of the synergism (interplay by combination of the
individual actions) in the various mixture ratios; generally, costabilizers (phosphites),
antioxidants (biphenols), and polyols of these mixtures are used to counteract the destabilizing
effect of the cadmium chloride that results from the subsequent stabilization reaction. The
greatest disadvantage of this stabilizer system is the toxicity, predominantly of the Cd part,
which has in recent times led to a regional ban on the stabilizer in many fields.
3. Barium-zinc stabilizers with the general formula (RCOO)2Ba and (RCOO)2Zn. These are
used in various mixtures because of the same synergistic effect as that of the Ba-Cd
stabilizers. The conversion of Zn carboxylate into ZnCl2 during the stabilization action produces
an even stronger destabilization effect than that of CdCl2 (ZnCl2 is a much stronger Lewis acid
than CdCl2) [115]. Hence the comixture with phosphites, epoxidized fatty acid esters, and
polyols in compounds is indispensable. This comixture neutralizes the ZnCl2 in its action.
Recently, because of the toxicological action of cadmium compounds, these stabilizers have
found increasing use, predominantly in the field of plastisols [117].
4. Ca-Zn stabilizers with a structural formula similar to that of the Ba-Cd or Ba-Zn stabilizers
have recently been finding many applications, despite their very much weaker effects, because
of the two nontoxic components: calcium and zinc stearate. They are always used in
combination with costabilizers, epoxy plasticizers, and antioxidants.
5. The lead stabilizers are inorganic compounds which can be used with lead stearate in various
combinations: sulfates, phosphites, or carbonates. The most important advantage of the
stabilizing systems based on lead compounds is the formation of lead chloride (PbCl2), which
has no destabilizing effect like that of ZnCl2 or CdCl2 and does not increase the conductivity
properties of PVC [117]. For the latter reason they are used in cable insulation. One further
advantage of the use, of the lead stabilizer system is that the finished products have a high
Vicat index and hence good mechanical properties [117].
6. Metal-free stabilizers are being increasingly discovered and investigated. The oldest of these
are diphenylthiourea (I) and -phenylindole (II):

Of some new organic stabilizers, -aminocrotonic acid and its esters should be mentioned
[114], compounds which have found use for emulsion PVC (E-PVC):

The diester of -aminocrotonic acid with thiodiethylene glycol has been used in formulations for
suspension PVC but is also effective in formulations for emulsion PVC [114].

Page 422

Wirth [115] found this compound to have a good costabilizing action, and hence it is used in
combination with Ca-Zn stabilizers to produce a long-term action.
Because of the inadequate action in the stabilization with the abovementioned stabilizers, many
formulations also use costabilizers which improve either the starting color or the antioxidant
properties of the stabilizer being used. A few stabilizer systems contain costabilizers already in
the commercial form; for example, the most commonly used metal soaps (Cd-Ba, Ca-Zn, BaZn carboxylate) are mixed in the commercial form with organic phosphites P(OR)3 in quantities
of 0.2 to 0.5 wt %. The phosphites and the epoxy compounds (epoxidized esters of oleic acid)
or the polyols are capable of some stabilizing functions such as (a) the binding of HCl:

or (b) the exclusion of labile chlorine atoms in PVC in, the presence of metal cations of the main
stabilizers by substitution [88, 20]:

7. Other important additives to improve the curative action of the stabilizer are antioxidants,
mostly phenolic compounds such as Bisphenol A (I) or 2, 6-di-tert-butyl-p-cresol (II) [107]:

Like the costabilizers, these are generally contained in a stabilizer package.

Page 423
B. Stabilization Mechanisms
The foregoing brief list of stabilizer classes was presented to facilitate the further description of
the individual stabilization mechanisms. The predominant theory on the stabilization of PVC is
the Frye-Horst theory [122124], which is based on the reaction of stabilizers with the labile
sites in PVC. In this way the zip reaction in PVC can be suppressed. According to this theory,
following treatment of a PVC sample with a stabilizer the number of stabilizer residues bound in
the polymer must be equal to or at least proportional to the number of defect sites in the
original sample. On the other hand, a direct correlation between cleaved HCl and stabilizer
residues chemically bound to the degraded sample should be detectable by measurement of
residual stability. This was investigated and discussed by Starnes and Plitz [125, 126], and the
result conclusively verified the Frye-Horst theory. In the pretreatment of a sample of PVC with
dibutyldithiododecyltin [Bu2Sn(SC12H25)2] and subsequent heat treatment, the authors were
able to show a correlation between the amount of sulfur bound in the PVC and the HCl
elimination. Thus the Frye-Horst theory remains the basis for showing the efficacy of
stabilization of PVC by most organometallic stabilizers.
1. Reaction of Stabilizers with Hydrogen Chloride: As already mentioned earlier, all stabilizers
are capable of reaction with HCl present in the PVC or already cleaved off, and then of binding
the acid. The compounds used must have a strong acid-receptor character and must not be
substances of an alkaline nature (NaOH rapidly breaks down PVC). Other compounds with a
weak-base character (e.g., all lead stabilizers, and also CaCO3, Na2CO3, amines, Na3PO3,
etc.) can be used as stabilization additives.
The organotin mercaptides and carboxylates bind HCl by substitution in two stages [117]:

Bellenger et al. [127] were also able to follow the course of the reaction of a dibutyltin
thioglycolate ester during extrusion, using IR spectroscopy.
In the case of tin carboxylates the corresponding organic acids are formed. Also by substitution
(mostly in two stages) the metal soaps react with HCl, for example, Pb stearate:

In the case of mixed metal soaps, the mechanism of HCl binding is more complex, due to the
desired synergism. In the case of Ba-Cd carboxylates, the barium carboxylate reacts
preferentially with HCl [117] and also ensures reconversion of CdCl2 (formed as a result of
reaction with initiation sites of dehydrochlorination) into Cd carboxylate which is reusable:

Page 424

This type of synergism was observed with all other mixtures of metal soaps (e.g., with the CaZn carboxylates) [128]. The metal-free stabilizers can also bind HCl. The -aminocrotonic acid
esters show a distinct induction period in their dehydrochlorination curve, which indicates HCl
binding [117], probably through coordination on the NH2 group:

2. Reaction Mechanisms of the Allylic and Tertiary Chlorine Atoms with Stabilizers: The tin
stabilizers are eminently suited to blocking the initiation sites of dehydrochlorination of PVC.
They react by substitution of an organic ligand of the stabilizer with the labile chlorine atoms of
the defect sites [21, 127, 129134].
Michel and Guyot [129] investigated the reaction of tin thioglycolate on models for the internal
chloroallylic group in PVC (a mixture of 4-chloro-2-hexene and its allylic isomer, 2-chloro-3hexene). They showed, using IR spectroscopy, that the reaction proceeds lonically via allylic
carboxylation:

The tin chloride formed can enhance the catalysis of the dehydrochlorination by coordination of
the labile allylic chlorine atoms according to

Page 425

This type of dehydrochlorination can also be caused by metal-free stabilizers such as aminocrotonic acid esters, particularly under the influence of ZnCl2 [135]:

In the same way, the tertiary chlorine atoms react with organotin compounds [114]:

As shown by Onozuka [136], the tin stabilizers not only substitute the labile chlorine atoms, they
can also isomerize the allylic chlorine atom of a growing polyene chain with the blocking of a
further dehydrochlorination (retardation of HCl elimination):

Because of their synergism the mixed metal soaps (Ba-Cd, Ba-Zn, and Ca-Zn carboxylates,
etc.) react with initiation sites of dehydrochlorination, also by substitution. Here, too, the
mechanism is of an ionic nature: the formation of a carbocation with subsequent attack on the
carboxyl group anion [121, 137, 138]. With Ba-Cd carboxylates the Cd carboxylate shows
preference for substitution of the labile chlorine atom because of the complexochemical
properties of the cadmium ion [116]. It regenerates itself by reaction with Ba carboxylate [114,
116]:

Page 426

If this synergistic action does not take place, there is the possibility of HCl elimination by
catalysis of the metal salts. The metal carboxylates (e.g., Fe stearate) are perfectly capable of
promoting elimination of HCl [99]. The stabilization reaction (substitution of labile chlorine
atoms) is a competing reaction to dehydrochlorination [114]:

The optimum synergistic effect is achieved when, for example, in the case of Ca-Zn
carboxylates at a ratio of 65 mol % Ca carboxylate and 35 mol % Zn carboxylate, both
components form a complex with a polymer melt so that the dehydrochlorination is suppressed
[136].
The synergistic effect increases in the case of combination of Cd-Ba carboxylates with
epoxidized fatty acid esters [114]:

In investigations on model compounds of PVC such as 4-chloro-2-hexene, a synergistic effect

was also noted in the case of reaction of dialkytin chloride and epoxy compounds [139]; the
route was explained via an ionic mechanism. Another way of blocking out starting sites in PVC
is the Diels-Alder addition reaction. Only the dialkytin maleates can undergo this reaction [114,
117]:

Page 427

In this way the polyene sequence is interrupted and the discoloration suppressed.
The stabilization with lead salts that contain no organic ligands, for example, the dibasic lead
carbonate (2PbO.PbCO3) or tribasic lead sulfate (3PbO.PbSO4.H2O) remains unexplained.
These are some of the oldest stabilizers for PVC. Although according to the investigations of
Graedel et al. [140] it is certain that these compounds can suppress the splitting of HCl for a
long time, it has not been clarified exactly how the reaction of the blocking of the starting sites
proceeds.
3. Inactivation of Impurities: As described earlier, the impurities that may occur in the
polymerization and subsequent work can be a factor in the thermal lability of PVC. Few
mechanisms are known that describe the action of stabilizers on these impurities. Jasching
[141] discussed the action of organic phosphites on metallic chlorides from the viewpoint of the
formation of chelate structures. As mentioned, epoxy compounds and phosphites are capable
of blocking out the labile chlorine atoms in PVC by substitution. The latter react with these
atoms as a result of Arbuzov rearrangement [114]:

It is still completely unknown how the iron chloride important for thermal destabilization in the
PVC can be blocked using stabilizers. This Fe chloride is a strong Lewis acid which accelerates
HCl elimination and the cross-linkage reaction.
4. Inactivation of UV Radiation: Many thermostabilizers are capable to a certain extent of
absorbing light and thus protecting the polymer against radiation. The most common are alkyltin
carboxylates, which in contrast to the sulfur-containing tin-based stabilizers, have a good
photostabilizing effect [117]. When the PVC is used outside (e.g., in window frames, foils,
sheeting, roofing, etc.), UV absorbers are generally incorporated. Other additives also prevent
damage by sunlight. Davidson [113] investigated photolytic de-gradation of PVC with and
without titanium dioxide and found that incorporated TiO2 can effectively protect PVC against
UV light. TiO2 is a pigment that cannot be used in all formulations, predominantly not for
compounds designed for transparent end products. Special additives, the UV absorbers, are
employed in this case. Nass [114] defined the properties of a UV absorber as follows:
1. Good absorption in the range 2900 to 3900
2. The ability to turn UV radiation into heat without discoloring the polymers
3. Intrinsically photostable and must not convert to hazardous products on absorption of UV
energy

Page 428
4. Compatible with all components of the compound
5. Nonvolatile and does not leach out (migration out of compound)
A large number of inorganic and organic salts which are not true absorbers of UV act as such
and are added to compounds (mostly in plastisols or flexible foils and sheeting). These include
salts of phosphoric acid, dibasic lead phosphite, barium phosphite, barium lauryl polyphosphate,
and sodium propyl phosphate. The mechanism of their action is as yet unknown. On the other
hand, there are many investigations on genuine UV absorbers such as 2hydroxyphenylbenzotriazole derivatives, 2-hydroxybenzophenone derivatives, and salicyclic acid
esters [142147].
The way radiation energy is converted chemically can be shown in the case of 2hydroxybenzophenone derivatives as follows [144]:

The keto form (I) absorbs the energy and converts to the enol form (II), which on giving heat to
the surroundings converts back to the keto form (I). It is important to note that 2hydroxybenzophenones must not be used with compounds of high polarity since the keto-enol
form described above (intra-molecular hydrogen bridges) prevents stabilization as these
complexes are unstable to light and can even act as sensitizers [145].
Similarly, the compounds of the second most important class of absorbers, the 2-(2hydroxyphenyl) benzotriazoles, also act to transpose absorbed energy into harmless heat
[148]:

In the salicyclic acid esters (the third most important UV absorber class for PVC) the
absorption of the energy of the light also causes an intramolecular reaction:

Page 429
if R is an aliphatic radical (e.g., dipropylene glycol salicylate). When R is an aromatic radical
(e.g., meta-hydroxyphenylsalicylate), a photo-Fries transposition takes place [149]:

The UV absorber concentration which may be used in PVC formulations is between 0.2 and
0.5% by weight and can in some cases (transparent sheeting for outside use or transparent
roofing) be increased to a maximum of 1% by weight under due consideration of the
compatibility with other components of the mixture.
The rheological effects of different stabilizer components are not discussed here (see Vol. 2,
Chapter 2). Some (such as the alkyltin mercaptides) are strong viscosity depressants; they
may also act as plasticizers. In so doing, they promote flow, retard frictional heat development,
and thus further contribute to stabilizer efficacy.
Others, such as inorganic lead salts, are relatively inert and innocuous with respect to rheology.
Their effect on melt rheology is similar to that of particulate mineral fillers (clay, talc, or ground
limestone); the effect will vary depending on the particle size of the lead stabilizer and the
concentration.
Finally, some stabilizer materials (barium octoate, calcium stearate, etc.) are strong gelling
agents. These salts are used in the lubrication industry to convert oils to greases. In this
respect, hot, molten PVC may be considered as a chlorine-substituted paraffin. The inclusion of
a long-chain barium or calcium carboxylate will exert a strong viscosity-building, or stiffening,
effect. The PVC oil is converted to a grease. During melt mixing and processing, extrusion
screws and the like must work harder to break down this extra viscosity and move these melts.
Accordingly, frictional heat is generated which then accelerates thermal degradation by various
chemical means, as discussed more fully in Vol. 2, Chapter 2.

REFERENCES
1. M. Ashina and M. Onozuka, J. Polym. Sci., A2, 3505 (1964).
2. M. Asahina and M. Onozuka, J. Polym. Sci., A2, 3515 (1964).
3. T. M. Birshtein and O. B. Ptitsyn, Conformations of Macromolecules, Wiley-Interscience,
1966.
4. K. Mitani and T. Ogata, J. Appl. Polym. Sci., 18, 3205 (1974).
5. D. Braun and E. Bezdadea, unpublished data.
6. K. Thinius, Stabilisierung und Alterung von Plastwerkstoffen, Verlag Chemie, Berlin,

1969.
7. F. Chevassus and R. de Broutelles, The Stabilization of PVC, Edward Arnold, London,
1963.
8. D. Braun and M. Thallmaier, J. Polym. Sci., C 16, 2351 (1967).

Page 430
9. D. Braun, Pure Appl. Chem., 26, 173 (1971).
10. W. C. Geddes, Rubber Chem. Technol., 40, 177 (1967).
11. J. Brandrup and E. H. Immergut, Polymer Handbook, Wiley-Interscience, 1975, See. II,
p. 475.
12. J. D. Cotman, Ann. N.Y. Acad. Sci., 57, 417 (1953).
13. S. Boccato, A. Rigo, S. Talamini, and F. Zilio-Grandi, Makromol. Chem., 108, 218
(1967).
14. D. Braun and W. Schurek, Angew. Makromol. Chem., 7, 121 (1969).
15. A. A. Caraculacu, E. Bezdadea, and G. Robila, J. Polym. Sci., A-1(8), 1239 (1970).
16. E. Bezdadea, D. Braun, E. Buruiana, A. Caraculacu, and G. Istrate-Robila, Angew.
Makromol. Chem., 37, 35 (1974).
17. A. Nakajima, H. Hammad, and S. Hayoshi, Makromol. Chem., 95, 40 (1966).
18. G. Bier and H. Krmer, Makromol. Chem., 18, 151 (1956).
19. W. H. Starnes, Jr., R. L. Hartless, F. C. Schilling, and F. A. Bovey, Adv. Chem. Ser., 169,
324 (1978).
20. W. H. Starnes, Jr., F. C. Schilling, K. B. Abbas, R. E. Cais, and F. A. Bovey,
Macromolecules, 12, 556 (1979).
21. W. H. Starnes, Jr., Dev. Polym. Degrad., 3, 135 (1981).
22. F. A. Bovey, K. B. Abbas, F. C. Schilling, and W. H. Starnes, Jr., Macromolecules, 8, 437
(1975).
23. G. Robila, E. C. Buruiana, and A. A. Caraculacu, Eur. Polym. J., 13, 21 (1977).
24. T. Hjertberg, E. Srvik, and A. Wendel, Makromol. Chem., Rapid Commun., 4, 175
(1983).
25. A. Rigo, G. Palma, and G. Talamini, Makromol. Chem., 153, 219 (1972).
26. G. S. Park and M. Saleem, Polym. Bull., 1, 409 (1979).
27. T. Hjertberg and E. M. Srvik, Polymer, 24, 673 (1983).
28. D. Braun and G. Holzer, Angew. Makromol. Chem., 116, 51 (1983).

 29. W. H. Starnes, Jr., F. C. Schilling, I. M. Plitz, R. E. Cais, and F. A. Bovey, Polym. Bull., 4,
555 (1981).
30. W. H. Starnes, Jr., F. C. Schilling, J. M. Plitz, R. E. Cais, D. J. Freed, R. L. Hartless, and
F. A. Bovey, Macromolecules, 16, 790 (1983).
31. G. V. D. Tiers and F. A. Bovey, Chem. Ind. (London), 1826 (1962).
32. W. H. Starnes, Jr., F. C. Schilling, K. B. Abbas, I. M. Plitz, R. L. Hartless, and F. A.
Bovey, Macromolecules, 12, 13 (1979).
33. D. Braun and F. Weiss, Angew. Makromol. Chem., 13, 55, 67 (1970).
34. T. Suzuki and M. Nakamura, Jpn. Plast., 4, 16 (1970).
35. A. H. Frye and R. W. Horst, J. Polym. Sci., 40, 419 (1959).
36. W. J. Bengough and M. Onozuka, Polymer, 6, 625 (1965).
37. A. A. Caraculacu, J. Polym. Sci., A-1(4), 1839 (1966).
38. E. C. Buruiana, V. J. Barbinta, and A. A. Caraculacu, Eur. Polym. J., 13, 311 (1977).
39. J. Boissel, J. Appl. Polym. Sci., 21, 855 (1977).
40. W. Merz and K. Mller, Microchim. Acta, 6, 60 (1974).
41. J. J. A. Wiys, Ber. Dtsch. Chem. Ges., 31, 750 (1898).
42. D. Braun and W. Quarg, Angew. Makromol. Chem., 29/30, 163 (1973).
43. A Michel, E. Castaneda, and A. Guyot, J. Macromol. Sci., A12, 227 (1978).
44. M. J. Abdullin, R. F. Gataullin. K. S. Minsker, A. A. Kefeli, S. D. Razumovskii, and G. E.
Zaikov, Eur. Polym. J., 14, 811 (1978).
45. D. Braun, E. Bezdadea, G. Holzer, and D. Sonderhof. Angew. Makro-tool. Chem., 104,
145 (1982).

Page 431
46. T. Hjertberg and E. M. Srvik, Polymer, 24, 685 (1983).
47. A. A. Caraculaeu and E. Bezdadea, J. Polym. Sci., 15, 611 (1977).
48. R. Petiaud and Q. T. Pham, Makromol. Chem., 178, 741 (1977).
49. U. Schwenk, I. Knig, F. Cavaga, and B. Wrackmeyer, Angew. Makromol. Chem., 83,
183 (1979).
50. D. Braun, G. Holzer, and T. Hjertberg, Polym. Bull., 5, 367 (1981).
51. D. Braun and D. Sonderhof, 3rd Int. Symp. Poly(vinylchloride), Cleveland, Ohio, Aug.
1980. Preprints, p. 14.
52. K. Mitani, T. Ogata, H. Awaya, and Y. Tomari, J. Polym. Sci., A1(13), 2813 (1975).
53. S. Crawley and J. C. MeNeill, J. Polym. Sci., Polym. Chem. Ed., 16, 2593 (1978).
54. D. Braun, Dev. Polym. Degrad., 3, 101 (1981).
55. J. Bauer and A. Sabel, Angew. Makromol. Chem., 47, 15 (1975).
56. D. Braun and D. Sonderhof, Eur. Polym. J., 18, 141 (1982).
57. D. Braun and D. Sonderhof, to be published.
58. K. S. Minsker, A. A. Berlin, V. V. Lisickij, S. V. Kolesov, and R. S. Korneva, Plast. Massy,
10, 69 (1976).
59. J. Svetl, R. Luc , J. Michalcov, and M. Kolinsk, Makromol. Chem., Rapid Commun.,
1, 247 (1980).
60. W. Kawai and T. Ichihashi, J. Polym. Sci., A-1(10), 1709 (1972).
61. D. Braun and M. Wolf, Angew. Makromol. Chem., 70, 71 (1978).
62. L. Ratti, M. Visani, and M. Ragazzini, Eur. Polym. J., 9, 429 (1973).
63. L. D. Wescott, Jr., M. Villacorta, F. C. Schilling, I. M. Plitz, and W. H. Starnes, Jr., ACS
Org. Coat. Appl. Polym. Sci. Proc., to be published.
64. D. Braun and M. Thallmaier, Kunststoffe, 56, 80 (1966).
65. D. Braun and E. Bezdadea, Angew. Makromol. Chem., 99, 55 (1981).
66. D. Braun and S. Kull, Angew. Makromol. Chem., 106, 165 (1982).

 67. D. Braun, E. Bezdadea, and G. Holzer, Angew. Makromol. Chem., 106, 47 (1982).
68. E. D. Owen and I. Pasha, J. Polym. Sci., Polym. Lett. Ed., 16, 429 (1978).
69. D. Druesedow and C. F. Gibbs, Natl. Bur. Stand. Circ., 525, 69 (1953).
70. T. T. Nagy, B. Turcsnyi, T. Kelen, and F. Tds, React. Kinet. Catal. Lett., 8, 7 (1978).
71. V. P. Gupta and L. E. St. Pierre. J. Polym. Sci., Polym. Chem. Ed., 17, 797 (1979).
72. T. T. Nagy, T. Kelen, B. Turcsnyi, and F. Tds, Magy. Kem. Poly., 83, 486 (1977).
73. B. B. Cooray and G. Scott, Eur. Polym. J., 16, 169 (1980).
74. K. Figge and W. Findei, Angew. Makromol. Chem., 47, 141 (1975).
75. D. Braun and S. Kull, Angew. Makromol. Chem., 86, 171 (1980).
76. M. Balandier and C. Decker, Eur. Polym. J., 14, 995 (1978).
77. T. T. Nagy, T. Kelen, B. Turcsnyi, and F. Tds, J. Polym. Sci., A-1(15), 853 (1977).
78. R. M. Aseeva, Y. S. Aseev, A. A. Berlin, and V. J. Kasatochkin, Zh. Strukt. Chim., 6, 47
(1965).
79. G. Scott, M. Tahan, and J. Vyvoda, Eur. Polym. J., 14, 377 (1978).
80. D. E. Winkler, J. Polym. Sci., 35, 3 (1959).
81. V. P. Gupta and L. E. St. Pierre, J. Polym. Sci., A1(8), 37 (1970).

Page 432
82. A. Guyot, M. Bert, A. Michel, and R. Spitz, J. Polym. Sci., A 1(8), 1596 (1970).
83. L. Pauling, The Nature of the Chemical Bond, 3rd ed., Cornell University Press, Ithica,
N.Y., 1960.
84. W. H. Starnes, Jr., Polym. Prepr., Am. Chem. Soc. Div. Polym. Chem., 18(1), 493
(1977).
85. W. H. Starnes, Jr., Adv. Chem. Ser., 169, 309 (1978).
86. E. D. Owen, ACS Symp. Ser., 25, 208 (1976).
87. A. Rieche, A. Grimm, and H. Mcke, Kunststoffe, 52, 265 (1962).
88. G. C. Marks, J. L. Benton, and C. M. Thomas, Soc. Chem. Ind. (London), Monogr., 26,
204 (1967).
89. M. Imoto and T. Nakaya, Kogyo Kagaku Zasshi, 68, 2285 (1965).
90. F. Tds and T. Klen, in Macromolecular Chemistry-8, K. Saarela, ed., Butterworth,
London, 1973.
91. R. C. Haddon and W. H. Starnes, Jr., Polym. Prepr., 18, 505 (1977).
92. R. C. Haddon and W. H. Starnes, Jr., ACS Adv. Chem. Ser., 169, 333 (1978).
93. W. H. Starnes, Jr., and D. Edelson, Macromolecules, 12, 797, 1237 (1979).
94. J. S. Shapiro and K. P. Nolan, J. Polym. Sci., Polym. Symp., 55, 201 (1976).
95. A. R. Amer and J. S. Shapiro, J. Macromol. Sci.Chem., A 14(2), 185 (1980).
96. H. O. Wirth, H. A. Mller, and W. Wehner, 2nd Int. Symp. Degrad. Stab. Polym.,
Dubrovnik, Oct. 46, 1978.
97. D. Druesedow and C. F. Gibbs, Mod. Plast., 30, 123 (1953).
98. T. Kelen, B. Ivan, T. T. Nagy, G. Turcsnyi, F. Tds, and F. Kennedy, Polym. Bull., 1, 79
(1978).
99. W. Reicherdt, ZFJ-Mitteilungen Nr. 63, Akademie der Wissenschaften der DDR, Leipzig,
1982.
100. J. Ernest, Bindung, Struktur und Reaktionsmechanismen in der organischen Chemie,
Springer-Verlag, Vienna, 1972.

101. G. Talamini and G. Pezzin, Makromol. Chem., 39, 26 (1960).

102. T. T. Nagy, T. Kelen, B. Turcsnyi, and F. Tds, Plaste Kautsch., 23, 894 (1976).
103. K. B. Abbas and E. M. Sorvik, J. Appl. Polym. Sci., 17, 3577 (1973).
104. F. Tds, T. Klen, and T. T. Nagy, Dev. Polym. Degrad., 2, 187 (1979).
105. A. Charlesby and S. H. Pinner, Proc. R. Soc., A249, 367 (1959).
106. E. J. Arlman, J. Polym. Sci., 12, 547 (1954).
107. C. Decker, J. Appl. Polym. Sci., 20, 3321 (1976).
108. W. W. Kolesnikova, O. W. Kolnikow, V. K. Milincuk, and S. J. Psezeckij, Vysokomol.
Soedin., A18, 11, 2431 (1976).
109. M. A. Golub and A. Parker, Makromol. Chem., 85, 6 (1965).
110. D. Braun and S. Kull, Kunstst., Fortschrittsber. 5, 90 (1980).
111. C. Decker and M. Balandier, Eur. Polym. J., 18, 1085 (1982).
112. K. P. Kwei, J. Polym. Sci., A17, 237 (1969).
113. R. S. Davidson, Polym. Photochem., 2, 1 (1982).
114. L. I. Nass, Encyclopedia of PVC, Vol. 1, Marcel Dekker, New York, 1976, p. 271.
115. H. O. Wirth and H. Andreas, Pure Appl. Chem., 49, 627 (1977).
116. L. S. Troitskaya and B. B. Troitski, Plast. Massy, 12 (1968).

Page 433
117. H. Andreas, in Gchter/Mller, Kunststoff-Additive, Carl Hanser Verlag, Munich, 1983, p.
199.
118. D. Braun, in Stabilization of Polymers and Stabilizer Processes, Adv. Chem. Ser., 85, R.
F. Gould, ed., American Chemical Society, Washington, D.C., 1968, p. 65.
119. J. Fertig, A. I. Goldberg, and M. Skoultchi, J. Appl. Polym. Sci., 9, 903 (1965).
120. J. Fertig, A. I. Goldberg, and M. Skoultchi, J. Appl. Polym. Sci., 10, 663 (1966).
121. D. F. Anderson and D. A. McKenzie, J. Polym. Sci., 8, 2905 (1970).
122. A. H. Frye and R. W. Horst, J. Polym. Sci., 45, 1 (1960).
123. A. H. Frye, R. W. Horst, and M. A. Paliobagis, J. Polym. Sci., A-2, 1765 (1964).
124. A. H. Frye, R. W. Horst, and M. A. Paliobagis, J. Polym. Sci., A-2, 1801 (1964).
125. W. H. Starnes, Jr. and I. M. Plitz, Polym. Prepr., 16, 500 (1975).
126. W. H. Starnes, Jr. and I. M. Plitz, Macromolecules, 9, 633, 878 (1976).
127. V. Bellenger, J. Verdu, and L. B. Carette, Polym. Degrad. Stab., 3, 3 (1981).
128. T. V. Hoang, A. Michel, and A. Guyot, Polym. Degrad. Stab., 1, 237 (1979).
129. A. Michel and A. Guyot, Polym. Degrad. Stab., 2, 277 (1980).
130. A. Michel and A. Guyot, Polym. Degrad. Stab., 3, 137 (1980).
131. A. Michel and A. Guyot, Polym. Degrad. Stab., 4, 213 (1982).
132. V. Bellenger, J. Verdu, L. B. Carette, Z. Vymalalova, and Z. Vymazal, Polym. Degrad.
Stab., 4, 303 (1982).
133. V. Bellenger, J. Verdu, L. B. Carette, Z. Vymalalova, and Z. Vymazal, Polym. Degrad.
Stab., 4, 365 (1982).
134. B. B. Cooray and G. Scott, Polym. Degrad. Stab., 2, 35 (1980).
135. A. Michel, T. V. Hoang, and A. Guyot, J. Macromol. Sci.Chem., A12, 411 (1978).
136. M. Onozuka, J. Polym. Sci., A-1(5), 2229 (1967).
137. M. Onozuka and M. Asahina, J. Macromol. Sci.Rev. Macromol. Chem., C3, 235
(1960).

138. G. Briggs and N. G. Wood, J. Appl. Polym. Sci., 15, 25 (1971).

139. T. Van Hoang, A. Michel, and A. Guyot, Polym. Degrad. Stab., 4, 427 (1982).
140. E. Graedel, J. P. Franey, W. H. Starnes, Jr., D. C. Hische, and P. C. Warren, J. Appl.
Polym. Sci., 23, 1769 (1979).
141. W. Jasching, Kunststoffe, 52, 458 (1962).
142. O. Cicchetti, M. Dubini, P. Parrini, G. P. Vicario, and E. Bua, Eur. Polym. J., 4, 419
(1968).
143. W. Klpffer, J. Polym. Sci., Polym. Symp., 57, 205 (1916).
144. J. G. Calvert and J. N. Pitts, Jr., Photochemistry, Wiley, New York, 1967, p. 534.
145. A. A. Lamola and L. J. Sharp, J. Phys. Chem., 70, 2634 (1966).
146. H. Gysling and H. J. Heller, Kunststoffe, 51, 13 (1961).
147. D. J. Carlsson, T. Suprunchuk, and D. M. Wiles, J. Appl. Polym. Sci., 16, 615 (1972).
148. H. J. Heller and H. R. Blattman, Pure Appl. Chem., 36, 141 (1973).
149. D. Bellus, Adv. Photochem., 8, 109 (1971).

Page 434

Page 435

9 Solvadon and Plastidzation

J. KERN SEARS
Monsanto Company
St. Louis, Missouri
JOSEPH R. DARBY
Darby Consulting, Inc.
Webster Groves, Missouri

I. INTRODUCTION

436

II. HISTORICAL PERSPECTIVE OF PLASTICIZATION

437

A. Early Technology Leading to Plasticized PVC

437

B. General-Purpose and Specialty Plasticizers

438

III. MECHANISM OF PLASTICIZER ACTION

440

A. Plasticizers Versus Processing Aids

440

B. General Theories: Lubricity, Gel, Mechanistic, and Free Volume

C. Plasticization Step by Step
IV. COMPATIBILITY OF PLASTICIZER WITH RESIN
A. Theories
B. Primary and Secondary Plasticizers
C. Temperature Dependence of Compatibility
V. EXTERNAL VERSUS INTERNAL PLASTICIZATION
VI. PLASTICIZATION OF POLYVINYL CHLORIDE
A. Methods of Compounding Plasticizers and Resin

442
444
448
448
474
476
477
478
479

B. Ease of Fusion
VII. EFFECT ON PHYSICAL PROPERTIES
A. Mechanical, Thermal, and Electrical Properties

486
490
490

B. Temperature Effects

494

C. Efficiency

497

D. Blending to Obtain Desired Efficiency

500

VIII. PERMANENCE PROPERTIES OF PLASTICIZED PVC

A. Volatility

500
500

B. Exudation Under Pressure

501

C. Humidity Compatibility: Environmental Compatibility

510

D. Extraction Resistance
E. Plasticizer Migration to Other Materials

510
513

Page 436
F. Migration of Contaminants from Other Materials
G. Controlled-Release Agents

513
515

H. Corrosion Resistance and Permeability

518

I. Heat Stability Effects

519

J. Light Stability and Weathering

522

K. Radiation Stability
L. Sterilization of Plasticized PVC

528
530

M. Fungal and Bacterial Resistance

530

N. Attack by Insects, Termites, and Rodents

534

IX. POLYMERIZABLE PLASTICIZERS

X. PLASTICIZATION BY FILLERS, PIGMENTS, SALTS, AND AIR
XI. ANTIPLASTICIZERS AND ANTIPLASTICIZATION
A. Crystallinity Theory

534
535
538
538

B. Reduced Free Volume Theory

539

C. Hindered Molecular Motion Theory

540

D. Uses of Antiplasticized Polymers

542

XII. TOXICITY AND ENVIRONMENTAL ASPECTS

XIII. WHAT PLASTICIZER TO USE: THE DESIRABILITY FUNCTION
REFERENCES

542
544
547

I. INTRODUCTION
Plasticization is traditionally compared to the formation of a solution [178]. In a binary solution
the solvent is the dispersing medium; the solute is the substance that goes into solution. Such
distinctions are useful in dealing with dilute solutions, but they lose precision when, for example,
a 50:50 alcohol-water mixture is considered. They have at times impeded progress in the study
of concentrated solutions when molecular interactions are strong [97]. In polymer solutions, and

particularly in plasticized resins or polymers, this is especially true.

A plasticizer is a material, frequently solventlike, incorporated in a plastic to increase its
workability and its flexibility or distensibility. Adding the plasticizer may lower the melt viscosity,
the temperature of the second-order transition, or the elastic modulus of the plastic [7].
Plasticizers have been in common use from prehistoric times. Water for clay to make pottery
and lime for easier working of cements are but two ancient examples from inorganic
technology. Gelatin with incorporated water for marshmallows and desserts is a plasticized
resin. Solution terminology is commonly used; the water is considered to solvate the resin.
For plasticization of a resin, resin and plasticizer must be intimately mixed. This is most
commonly done by heating until the resin has dissolved into the plasticizer or the plasticizer has
dissolved into the resin. The material is then formed and cooled. Alternatively, the resin and
plasticizer can be mixed by dissolution in a common solvent, and the solvent then removed by
evaporation. In an industrial example plasticizer effects can be dramatic-

Page 437
ally shown in several ways when a high-molecular-weight polyvinyl chloride homopolymer is
fluxed on a hot roll mill. At 1600C, the resin particles pour unchanged, like sand, between the
rolls. If the temperature is raised, the resin softens and eventually forms a tough sheet around
one of the rolls. While it is softening and sheeting out on the mill, it is also decomposing from
excessive heat. When the sheet is removed and cooled, it is brown to black, horny to brittle,
corrosive from liberated hydrogen chloride, and worthless. If, on the other hand, a small amount
of a plasticizer such as di-2-ethylhexyl phthalate (DOP) is used to wet the resin particles, the
resin softens and fuses when it is poured onto the roll mill at 160C; the mass becomes
homogeneous and forms a sheet around one roll. When the sheet is removed and cooled it is
clear, colorless to faintly yellow, noncorrosive, and semirigid to very flexible, depending on the
amount of plasticizer used. This is a simplified version of the more sophisticated formulations
entering the multimillion-dollar market of plasticized polymers.
Although many types of high-molecular-weight materials, organic and inorganic, are benefited
by plasticizers, also organic and inorganic, the main emphasis is on organic plasticizers for
synthetic organic polymers, with polyvinyl chloride (PVC) applications requiring about 80% of
plasticizer production.
Organic plasticizers are usually liquids of moderately high molecular weight, or they are
occasionally low-melting solids. Most commonly, they are esters of carboxylic acids or
phosphoric acid. Other types also used include hydrocarbons, halogenated hydrocarbons,
ethers, polyglycols, and sulfonamides.

II. HISTORICAL PERSPECTIVE OF PLASTICIZATION

A. Early Technology Leading to Plasticized PVC
Plasticizer technology today is a hybrid of art and science, the art having its roots in the
discoveries of many primative people. The most direct evolutionary line of today's plasticizer
industry is in surface coatings. Search for superior vehicles over many years led to the blending
of many natural resins, gums, and drying oils. When a resin used in a spirit varnish left a film
that was too brittle and inelastic, it was modified with a drying oil or other material that would
toughen. it by giving it just the right amount of plasticity [10,71,99,132]. The gap between
surface coatings and massive plastic articles was frequently bridged; an example is the ancient
Chinese lacquer vases formed by many layers of tung oil, dried, and carved. The gap was
bridged permanently when shellac, a surface-coating material, was used with suitable fillers in
making picture albums and handles of looking glasses about the time of the Civil War [69].
Shortly before this, Pelouze in 1838 and Schonbein in 1846 discovered and developed an
adequate process for preparing cellulose nitrate. This first clearly defined resin deliberately
created was of little value. It took the incorporation of a plasticizer to make it a commercial
success. For surface coatings the unmodified resin was too brittle except in very thin films used
for lacquering silver. For other uses it was softened by gums. Pellen [151] added castor oil to
give lacquer distensibility enough for coating aeronautical ballons to make them impermeable to
hydrogen [216].

John and Isaiah Hyatt, interested in massive articles, performed their classical experiments and
filed for patent rights to what they modestly called

Page 438
Improvement in Treating and Molding Pyroxyline [102]. They had modified the new resin that
could not be molded by adding approximately 50 phr (parts per hundred of resin) of camphor
and molding with heat and pressure to obtain a substitute for horn, ivory, bone, etc.
The camphor, when molten, had plasticized the cellulose nitrate without the auxiliary solvents
used for lacquers. It permitted the plastic mass to be formed, while hot, into any desired
shape. It permitted the molded material to cool without cracking; and it gave the final objects a
toughness to withstand abuse that the resin alone did not have. This material, celluloid,
remained the principal thermoplastic material until shortly before World War II. Between 1870
and 1930, objections to odor, volatility, and the high price of Japanese camphor motivated
research for other plasticizers for cellulose nitrate. In 1912, triphenyl phosphate became a
reasonably successful replacement for camphor, to be followed by liquid phosphates such as
tricresyl phosphate. The use of phthalates as plasticizers began in the early 1920s [68], and di2-ethylhexyl phthalate (DOP) was patented in 1933 [122]. Other thermoplastic resins also
appeared polyvinyl acetate (1925), cellulose acetate (1927) which benefited from at least
small amounts of plasticizers. The slow development of celluloid and related products was the
induction period for the plasticizer and thermoplastic resin industry. Its history was repeated
with almost explosive rapidity in PVC. Although PVC had been in the literature for many years,
it was a most unpromising resin, even in 1931, when first produced on a small commercial
scale. It was insoluble in common solvents, could not be molded without thermal decomposition,
and turned black on a few days' exposure to sunlight; but it was unusually tough, and its
inertness to solvents was challenging. It took plasticization to make it a success. The critical
point was passed when Waldo Semon in 1933 found that PVC would form rubber-like
compositions when it was dissolved hot in such solvents as o-nitrodiphenyl ether,
chloronaphthalene, dibutyl phthalate, or tricresyl phosphate, and was allowed to gel on cooling.
His vision was broad enough to foresee many of the current uses of plasticized PVC, such as
coated cloth, convertible automobile tops, waterproof boots and shoes, electrical insulation,
floor tile, industrial hoses, and tank liners [183]. With the shift in emphasis through the years
from plasticizers for cellulose nitrate to plasticizers for polyvinyl chloride, the search for new
plasticizers increased.
B. General-Purpose and Specialty Plasticizers
1. General-Purpose Plasticizers
Out of the thousands of plasticizers that have been patented and several score that are
commercially available, there are a few that have good all-around physical properties and price
that enable them to plasticize PVC for most of its uses. These are all moderate-molecularweight dialkyl phthalates. Di-2-ethylhexyl phthalate, dioetyl phthalate (or DOP) was long the
unquestioned industry standard for a general-purpose plasticizer for PVC, the one to which all
others were compared. For many years it accounted for a fourth or more of the total plasticizer
production. Other, more highly branched chain alkyl phthalates, such as diisooctyl and
diisodecyl phthalates, are considered general-purpose plasticizers even though their extra
branching makes them less efficient, especially at low temperatures, and less stable toward
oxidation. Phthalates of unbranched or rarely branched alcohols, the so-called linear

phthalates, are also general-purpose plasticizers. Although their linearity makes them much
more efficient at low temperatures, less volatile, and more

Page 439
stable toward oxidation than branched phthalates of equal molecular weight, it also makes
fusion slower. The di(heptyl,nonyl,undecyl) phthalate from mixed C7, C9, and C11 alcohols has
become a second industry standard general-purpose plasticizer with improved performance at
slightly higher price. Alkyl benzyl phthalates, with very fast fusion properties, excellent
compatibility, and fair low-temperature efficiency and resistance to oxidation, are used so
widely that they are at times considered among the general-purpose plasticizers; more often
they are considered specialty plasticizers [37].
2. Specialty Plasticizers
Many uses for plasticized PVC demand properties and performance not adequately supplied by
the general-purpose plasticizers. Other (specialty) plasticizers alone or more frequently blended
with a general-purpose plasticizer are used to achieve these properties. The industry has
become adept at constructing new plasticizers to do specific jobs, and even more adept at
using the strong points of any plasticizer.
Use of flexible vinyl in housing and transportation equipment, coupled with greater emphasis on
safety, demands plasticizers with flame-retardant chemical structures that also maintain good
compatibility. Phosphate esters, and to a limited extent, highly chlorinated paraffins, are used
for flame retardance. Tricresyl phosphate was for 40 years the industry standard in this field.
Later isopropylphenyl and t-butylphenyl diphenyl phosphates and alkyl diphenyl phosphates
became more important.
The need for very good flexibility at low temperatures is met in many cases by linear phthalates
and especially by diundecyl phthalate, or by C9, C10,C11 phthalate. When exceptionally low
temperatures are required, aliphatic esters of adipic, azelaic, and sebacic acids are used and
to a limited extent tri-2-ethylhexyl phosphate. Epoxidized alkyl oleates and tallates (from tall oil
fatty acids) are frequently used in small amounts.
The initial use of plasticized vinyl in automobile upholstery, crash pads, and door panels was
accompanied by fogging. Part of the problem was lack of permanence, as plasticizers
distilled from the PVC are condensed on windshields when cars are parked in the sun. High
standards for nonfogging plasticizers put emphasis on products that are exceptionally low in
volatile components. Electrical insulation for wire and cable also demands plasticizers with very
low volatility. In these areas diundecyl, C9,C10,C11 and ditridecyl phthalate, various trialkyl
trimellitates, and linear polyesters from glycols and dicarboxylic acids, especially adipic acid,
are most commonly used. When resistance to extraction by fats, oils, and gasoline is required,
reliance is placed on alkyl benzyl phthalates, linear polyesters, and to some extent on blends of
PVC with low-molecular-weight nitrile rubbers, polyurethanes, and certain other semiflexibilizing
polymers.
For food packaging films and plasticized PVC items that contact food animals, the plasticizers
must be approved by the U.S. Food and Drug Administration or the U.S. Department of
Agriculture or other national government agencies. Such plasticizers with extremely low toxicity
ratings include a few of the general-purpose and specialty plasticizers mentioned above, such

as DOP, butyl benzyl phthalate, various adipates, and others, including diethyl phthalate,
glyceryl triacetate, certain trialkyl and acetyl citrates, and epoxidized soya oil. Plasticizer
suppliers and government regulating agencies should be contacted for precise information
[41,42].
For vinyl floor coverings, tile, and sheet goods, plasticizers are desired that give adequate
flexibility for installation and yet withstand traffic stain-

Page 440
ing. The stain problem varies with such factors as contaminants tracked in on shoe soles,
tanning agents in shoe leather, and antioxidants in rubber heels, as well as environmental
conditions such as humidity. Satisfactory performance is a direct function of the plasticizer
used, its concentration, its compatibility with PVC, and its affinity for the stainant. Butyl benzyl
phthalate and other plasticizers, such as 2,2,4-trimethyl-1,3-pentane-diol monoisobutyrate
benzoate give improved vinyl flooring.
In these, and all other applications, association between the plasticizer and PVC is extremely
important in controlling the final performance properties of the compound.

III. MECHANISM OF PLASTICIZER ACTION

A. Plasticizers Versus Processing Aids
Plasticizers, as defined earlier, are used at times to make processing of the resin or fabrication
of the final article easier without seriously changing the properties of the original resin. Usually,
the plasticizer is used in larger quantities and substantially modifies the properties of the resin
as it appears in the final article (e.g., a soft fleshlike doll of plasticized PVC versus rigid vinyl
pipe).
As thermoplastic materials with higher heat-distortion temperatures (engineering
thermoplastics) invade the field of thermosets and even replace metals, and as rigid PVC
assumes greater economic significance, more effort is being put on the first use of plasticizers
as aids in processing. Processing can be made easier by better internal lubricity, reduced
sticking of the molten resin to the metal equipment, lower hot melt viscosity, or improved resin
stability. At times, proper choice of a plasticizer can aid processing in all these ways
simultaneously.
Figure 1 shows the idealized response of a resin to plasticizers as processing aids and as enduse plasticizers. In this case, the processing aid is visualized as lowering the processing
temperature without changing the soften-

FIGURE 1 Idealized response of resins to processing aids and end-use plasticizers. (From Ref.
179.)

Page 441
ing temperature or upper limit of usefulness of the final article. In contrast with this, the end-use
plasticizer must lower the softening point, (glass transition temperature, Tg), which is now the
lower limit of usefulness, and leave the flow temperature unchanged. This represents the
broadening of the elastomeric range desired from the plasticizer. The plasticizer must therefore
leave some of the strong intermolecular forces intact.
Frequently, rubbery polymeric materials are added to give just enough elongation or enough
softening to the resin, at least in highly dispersed micro-regions, that it will withstand more
mechanical abuse; these impact modifiers are occasionally classified as plasticizers.
The ideal results described in Figure 1 are never completely obtained. Figure 2 is more
realistic. When DOP is added to PVC, the elastomeric range between the two lines becomes
broader, but at the expense of some noticeable drop-off of the upper limit. Polyvinyl chloride is
unusually responsive in this manner; most resins show no more than a slight broadening of the
elastomeric range. This broadening of the elastomeric range is seen in dynamic mechanical
testing as a shift to lower temperature and a widening of the -transition region peak. This has
been explained to result from the multiplicity of molecular conformations that plasticizer and
resin molecules can form. There are different ways in which each plasticizer molecule,
depending in part on its shape, can separate PVC chains from each other. PVC is itself very
heterogeneous. There are density fluctuations in the amorphous regions as a result of variations
in chain packing. There are various degrees of crystallinity with regions of unmilled primary
particles. There are, therefore, regions of microheterogeneous plasticization, some regions
more plasticized than others. This, in theory, results in the unusual broadening of the
elastomeric region. By contrast, diphenyl phthalate, as a processing aid in a styreneacrylonitrile copolymer (SAN), shows the desired lowering of the flow temperature with only
moderate lowering of the softening temperature. Compounds such as diphenyl phthalate might
owe some of this ability to their tendency to crystallize on slow cooling. In the crystalline form
they act as fillers. When a resin containing such a processing aid is quench-cooled to avoid
crystallization, the mixture is frequently clearer but more brittle, in keeping with the common
behavior of small amounts of liquid plasticizers.

FIGURE 2 Response of PVC to di-2-ethylhexyl phthalate as an end-use plasticizer compared

with the response of SAN to diphenyl phthalate as a processing aid. (From Ref. 179.)

Page 442
Volatile plasticizers may also be used as processing aids. The major portion is flashed off
during processing, but at least traces remain to give some plasticization. In general, when a
plasticizer is used to aid only in processing, the best results are obtained when the plasticizer is
then removed from the point of action by volatilization, crystallization, or chemical changes
such as polymerization.
B. General Theories: Lubricity, Gel, Mechanistic, and Free Volume
Four principal theories have been proposed to account for the main effects that plasticizers
produce. The lubricity, gel, and mechanistic theories are the products of many workers in the
first half of this century [67,68]. The lubricity theory views the resistance of a resin to
deformation (or its rigidity) as the result of intermolecular friction. The plasticizer acts as a
lubricant to facilitate movement of the resin macromolecules over each other, or segments of
the macromolecules, even if other parts of the macromolecules are associated in some kind of
gel network [38]. The plasticizer gives internal lubricity [21].
The gel theory, devised for amorphous polymers, considers their resistance to deformation to
result from an internal, three-dimensional honeycomb structure or gel. This gel is formed by
more-or-less loose attachments that occur at intervals along the polymer chains. The
dimensions of the cells in a stiff or brittle resin are small because the points of attachment are
close together. Any attempt to deform the specimen cannot be readily accommodated by
movement within the mass. Its limit of elasticity is low. A thermoplastic or even a thermosetting
resin with points of attachment between macromolecules widely separated is flexible without
plasticization. Natural rubbers, unvulcanized and lightly vulcanized, are examples of such
materials. For elasticity, it is not necessary to have two completely separate phases; it is
necessary, rather, to have bonding forces of two different orders of magnitude [28].
The action of a plasticizer on a resin having many points of attachment along the polymer
chains is to break the attachment and mask the centers of force that have held these polymer
chains together by selectively solvating the polymers at these points. This produces much the
same result as if fewer points for attachment had been provided on the macromolecules in the
first place.
According to the mechanistic theory it is assumed that solvents or plasticizers of different
classes are attracted to the resin macromolecules by forces of different magnitude and that
none of them are bound permanently when they are attached. Instead, there is a continuous
exchange whereby one plasticizer molecule becomes attached to a given active group or force
center only to be dislodged and replaced by another. This results in a dynamic equilibrium
between solvation and desolvation in which a certain fraction of the force centers of the
polymer chains are masked by solvent or plasticizer under a given set of conditions such as
plasticizer concentration, temperature, and pressure. For partially crystalline resins,
plasticization affects primarily the amorphous regions or the areas of crystal imperfections.
Above the glass transition temperature, Tg, of PVC plasticized with dibutyl phthalate, all the
DBP is at least partially bound to the PVC, with the lifetime of the PVC-DBP solvates calculated
to be 10-11 to 10-5 sec [129].

Page 443
The free volume theory assumes that there is nothing but free space between molecules. With
an increase in temperature a polymer and a plasticizer, like most matter, increase in volume.
Since no mass has been added, the increase is the result of an increase in free volume. Free
volume increases with increased molecular motion, and there is an exceptionally large amount
of free volume associated with chain end groups in a polymer system. Increasing the
concentration of chain ends increases the free volume (see Figure 3a). Therefore, decreasing
the molecular weight of a polymer increases free volume, which is end-group plasticization.
Addition of flexible side chains (part b) produces the same result. A crankshaft free volume in
some polymers (part c) is believed to help impact resistance. All three of these effects are
achieved by internal plasticization, and free volume is geographically fixed with regard to the
polymer molecule. Addition of a small molecule with flexible ends affects a large free volume at
its location, and its location can shift. This is external plasticization (see Figure 3d).
Enough free volume permits freedom of movement as if there were a hole for a nearby atom,
molecule, or chain segment to move into. The normal result of adding more free volume to a
polymer is that it is plasticized. Its Tg becomes lower; the modulus and tensile strength
decreases, but elongation

FIGURE 3 Sources of free volume for plasticization: (a) chain end motion; (b) side-chain
motion; (c) main-chain crankshaft motion; (d) external plasticizer motion. (From Ref. 176.)

Page 444
and impact strength increase. However, the incorporation of a small amount of plasticizer into a
resin can give enough free volume that the resin molecules have the freedom to move closer to
each other, if they are strongly attracted to each other, and polymer-polymer association can
increase in many areas in the system. The gel structure becomes tight, and the resin becomes
more rigid even though the Tg is lower and the free volume is probably higher in areas from
which plasticizer has not been excluded by resin-resin associations or has not been bound
tightly by polar or hydrogen bonds (see Section XI).
C. Plasticization Step by Step
Several different steps during plasticization have been described depending on what physical
properties were measured and how the measurements were made. It is difficult to correlate
most of these phenomena from one observer to another [48]. The step-by-step process is of
concern in PVC plastisols and dry blends, but the changes occurring are not unique to these
systems; they are simply more important mechanically. A plastisol of PVC (i.e., a pourable
dispersion of a finely divided PVC in a liquid plasticizer) on warming or on long standing at room
temperature will gel as the plasticizer solvates the resin, but the gel has little physical integrity.
If the mixture is heated further, it gradually develops enough integrity to hold itself together, and
finally, at about 1600C, it fuses. When cooled, it has its maximum physical properties: tensile
strength, modulus, elongation, and so on. But what happens on a molecular scale is less well
known. Some of the thermal changes occurring may be seen in Figure 4. The first five traces
show behavior of plastisols with different plasticizers and different stabilizer systems, including
the stabilizing plasticizer epoxidized soya oil. These may be compared with trace 6, where the
plasticizer has been replaced with a silicone, Dow-Corning Fluid 550 (Dow-Corning Corp.), that
has little or no solvating effect on the resin. The first break in trace 6 at 81C, corresponds well
with the known glass transition temperature of PVC.
In the plastisols, the changes that occur are most pronounced in the fast-fusing butyl benzyl
phthalate (BBP) and least pronounced in the slow-fusing diisodecyl adipate (DIDA). With butyl
benzyl phthalate, there is first, a small exotherm at the proper temperature for gelation, after
which the changes become endothermic. The endothermic portion of the curve shows that
fusion occurs in several steps and is not a simple change, such as the melting of ice, as it had
commonly been thought. The breaks in the thermograms are probably related to changes in
crystallinity and perhaps other degrees of order in the polymer.
Hoy [100] showed that the surface gloss of plastisols during fusion changes in a similar
stepwise fashion (Figure 5). He defined the gel point as the temperature at which the gloss
level just begins to decrease. McKinney [134, 135], in studying the dry-blending process and
subsequent fusion of the dry blend, detected changes in crystallinity as well as second-order
transitions.
An overall picture involving four to six steps has been proposed [178, 179] based on the work
of many investigators. First, there is a rapid, irreversible plasticizer uptake as plasticizer
molecules work their way into porous areas of the resin and adhere there. Next, there is an
induction period of slow penetration of the surface. Then there comes an absorption step during

which the total volume of resin-plasticizer may decrease, although the resin particles swell
slowly on the outside, setting up strong internal strains. Rate studies during early particle
swelling indicate low to moderate activation ener-

Page 445

FIGURE 4 Differential thermal analysis of plastisol fusion. The rate of heating, 5 hr for each
thermogram, was slow compared with the usual laboratory or plant fusion practice of 1 to 30
min but fast compared with shelf-life studies. Much faster heating obscured some of the breaks
in the thermograms. (From Ref. 179.)
gies (5 to 50 kcal) and half-lives directly proportional to the vapor pressure of the swelling
agent or plasticizer [31]. This indicates a simple diffusion process. There follows a fourth step
with higher activation energies (68 to 111 kcal) [50,51], during which more severe changes take
place inside the particle, yet with little or no overall volume change. Dry blends increase in
volume slightly as the resin molecules relax [135]. This appears to be a diffusion process with
half-lives much longer than the first but also proportional to the vapor pressure of the
plasticizers. (Complete absorption to obtain a dry blend at room temperature is very slow; fastfusing butyl benzyl phthalate itself is calculated to take 4 years, 4 months) [135]. From the
order of magnitude of the activation energies, it appears that polymer molecules are being
disentangled and separated. The higher values indicate that the energies involved are enough,
at times, to break primary carbon-carbon bonds. That bond rupture actually takes place in
many high-molecular weight polymers on swelling, even without mechanical mixing, has been
demonstrated [3335].

Page 446

FIGURE 5 Effect of gelation and fusion on reflectance from the surface of a DOP plastisol. The
gel point is the point at which reflectance first drops. This point rises with increased plasticizer
content. (From Ref. 100.)
At the start of this fourth step, most or all of the plasticizer has been incorporated in the resin
and the slight volume changes have taken place, but plasticization is not complete. The
plasticizer is present as clusters of molecules in between clusters or bundles of polymer
segments or molecules. At this point, a mixture of PVC and plasticizer may be formed into a
finished article that will be hard and have high tensile strength and poor elongation compared
with a properly fused piece of identical composition [160]. As energy is applied to the system
during this step, a marked change occurs. The volume remains the same, but the dielectric
constant rises significantly, showing that the polymer molecules are no longer rigidly held
together but are free to move as if the polymer were in its rubbery state rather than its, glassy
state [50,51]. The plasticizer molecules have now penetrated the clusters or bundles of polymer
molecules and plasticization is essentially complete. Grotz [90] points out that the energy
barrier for dipole orientation found by Dannis [50,51] is higher than the energy barrier for
movement of polymer segments to accommodate the diffusing plasticizer. But for an elastomer
such as plasticized PVC, with maximum physical properties to be produced, this energy barrier
must be overcome.
When plasticization occurs by heating until the mixture is fluid, step 5, some toughening
structure must be formed on cooling, step 6. This may result from van der Waals forces or
hydrogen bonding of polymer segments, either directly or through the plasticizer, or it may
occur from entanglement or crystallization of the polymer segments. This toughening is shown
in plasticized PVC in Figure 6. The activation energy for thermal rupture of a microdumbbell.
above about 165C is low; it is of the order of magnitude of strong hydrogen bonding or van der
Waals forces or of the activation energy of easily diffusible molecules. Below 165C, the
activation energies are higher, probably signifying the necessity for overcoming serious
molecular entanglement or for destroying some low order of crystallinity. That this trans-

Page 447

FIGURE 6 Arrhenius plots of thermal dissociation of microdumbbells in the zero-strength-time

(ZST) test. Specimens were of PVC plasticized with butyl benzyl phthalate (BBP) and
diisodecyl phthalate (DIDP). (From Ref. 179.)
ition should occur at about the temperature at which plasticized PVC is milled and molded is not
surprising. Neither is it surprising that the more solvating butyl benzyl phthalate should give
material with a lower activation energy than the less solvating diisodecyl phthalate.
The six steps of plasticization may not all occur for every polymer-plasticizer combination. The
steps as described also appear at times to overlap somewhat. The fact that plasticization of
PVC with the usual moderate-molecular-weight plasticizers can proceed at room temperature
only through the first few steps, but is thermodynamically prohibited from proceeding further
until some threshold temperature is passed [5,87,100], shows that the steps at other times are
quite distinct.
In some systems, there may be additional substeps. Gelation in the plastisols of Figure 4 may
be such. Gelation occurs when resin particles have swollen until they touch [100], but this alone
would not account for an exotherm. it is possible to conceive of polymer segments on the
surface of each particle that becomes free to move with initial swelling of the polymer. These
segments from one particle may then become associated with those of an adjacent particle in
some kind of order, perhaps crystalline, thus leading to gelation and a slight exotherm.
McKinney observed a similar exotherm in dry blends that might have resulted from improved
thermal conductivity on particle swelling [135] but might also have been caused by slight
crystallization as polymer segments became free to move.

Page 448

IV. COMPATIBILITY OF PLASTICIZER WITH RESIN

A. Theories
To those concerned with development and production of plastics materials such as resins,
solvents, and plasticizers, compatibility has long meant the ability of two or more substances to
mix with each other to form a homogeneous composition of useful plastic properties [187]. To
some concerned with using finished plastic articles in specific environments it may mean exactly
the opposite. Thus a resin that can be used for gasketing in a hydraulic system without being
swollen or shrunken by the hydraulic fluid is considered to be compatible with the fluid, since
they may be used together [13,14]. For the rest of this discussion compatibility will be used in
the first sense, indicating that the materials may be intimately blended. Most of the theoretical
studies have been made with solvents, and the results have been applied with some misgivings
to plasticizers. Hansen's data show that with the tougher resins, including PVC, plasticizers are
rarely very good solvents [93].
The reasons one solvent will dissolve a given resin, another only swell it, and a third leave the
resin untouched have prompted a prodigious amount of empirical trial-and-error study and more
than the usual amount of theorizing. Since many polymers during their processing or use, or in
determination of their structure, are dissolved in or swollen by liquids, the subject is vital to all
concerned with high polymers, whether they deal with structural engineering polymers, surface
coatings, adhesives, rubbers, fibers, or molecular biology.
1. Like Dissolves Like
In a less sophisticated age, when less precision was demanded, the common view of
compatibility was that there are many useful pointers. The polarity of the plastic base and the
plasticizer are most useful in assisting the correct choice of prospective plasticizers. It is a case
of like dissolving like [109, 116]. Like dissolves like has been remarkably successful through
the years but has also afforded enough exceptions to cast doubt on the whole concept [26].
For example, cellulose is not soluble in water; polystyrene is soluble in butyl acetate but not in
methyl cyclohexane; and polymers are frequently incompatible with their monomers or
oligomers. Like is never clearly defined and is always essentially intuitive, so exceptions are
subconsciously held in abeyance until research explains why they are or are not alike. Further
research, especially application of thermodynamics, has helped clear up many compatibility
problems, but the basic concept lingers on in the recognition that cohesive energy densities
must be alike.
2. Thermodynamic Theories
Thermodynamic properties of polymer solutions depend greatly on the molecular weight of the
polymer, and deviations from ideality in polymer solutions are much greater than deviations
encountered in low-molecular-weight systems.
According to thermodynamic theory, two substances will be miscible when the free energy of
mixing, as shown in equation (1), is negative:

G = H - T S

(1)

Page 449
where
G =
free energy of mixing
H =
heat of mixing
S =
entropy of mixing
T =
temperature
Application of the theory to real solutions has led to the Hildebrand solubility parameter , the
Flory-Huggins interaction parameter , and the Heil-Prausnitz segment interaction parameters
G1 and G2. The Hildebrand theory [97] was designed for solutions in which solvent and solute
molecules are of about the same size, so entropy is ideal. Solubility is controlled primarily by
H, and according to theory,

The solubility parameter therefore affords one measure of approximating H. Although the
value for a given solvent is characteristic of the solvent and independent of the solute, it must
be used with the value for the desired solute, since the difference between them is what is
significant. When 1 - 2 approaches zero, H approaches zero and G becomes negative for
low-molecular-weight solutes. For high-molecular-weight solutes some empirical adjustment is
made to account for the nonideal entropy change.
The Flory-Huggins [77] interaction parameter is a means of correcting for nonideality of both
heat and entropy of mixing. The exact value of depends on the concentration at which it is
determined, because a polymer molecule responds not as an equal number of monomer units
nor as a large, completely unified molecule, but somewhere in between, since it is strongly
influenced by molecular associations. The Heil-Prausnitz [95,96] segment interaction
parameters G1 and G2 generalize the Flory-Huggins theory to apply at all concentrations.
3. Predictors of Compatibility
Predictors of compatibility are physical property measurements or thermodynamic properties of
a plasticizer or of a polymer which may be measured or calculated and which give an indication
that the plasticizer will or will not be adequately compatible with the resin to be useful with it.
This is done without even putting the plasticizer into contact with the polymer. Several
predictors of varying type and sensitivity are frequently used.
a. Solubility Parameter and Its Modifications. The cohesive-energy density (CED) is one
measure of the intensity of intermolecular interactions in a pure liquid or solid. The strength of
the solvent-solvent bonds and polymer-polymer bonds is related to it. It is numerically equal to
the potential

Page 450
energy of 1 cm3 of material but opposite in sign. The solubility parameter , as a
thermodynamic quantity, is represented by
(3)
where
E =
molar energy of vaporization
H =
heat of vaporization per mole
V =
molar volume, or molecular weight/density, d
R =
gas constant
T =
temperature
With a number of assumptions detailed in the literature, which are seldom or never fulfilled in
real polymer solutions (e.g., ideal entropy when entropy is never ideal), the theory requires that
a polymer will be soluble in a solvent when the two have solubility parameters, that are the
same or that do not differ by more than about 1.5 (cal/cm3)1/2. Heats of vaporation from
calorimetric data or from vapor pressures are not always known for solvents. For plasticizers
that are usually very low in volatility and are rarely one isomerically pure chemical species, such
data are almost nonexistent. A reasonable approximation of H can be obtained for many of
the lower-molecular-weight plasticizers for PVC by the Hildebrand rule [97]. In theory, the heat
of vaporization must be related to the normal boiling point. The empirical relation of Figure 7
(extended to higher temperatures than Hildebrand was concerned with and modified for direct
estimation of E) works well for relatively non-polar solvents and plasticizers when the normal
boiling point is known. From the estimated E, the solubility parameter can be calculated by
equation (3).
The solubility parameter is closely related to other physical measurements that reflect the
forces holding the molecules to each other. It can be estimated from internal pressure
measurements, from the van der Waals correction to the gas law, and from coefficients of
thermal expansion and compressibility, critical pressure, and so on, most of which are applied
with great difficulty to liquids as high boiling as plasticizers. The estimation from surface tension
is much easier to apply:
(4)
where
= surface tension, dyn/cm
V = molar volume, cm3
k = 4.1 at 20C
Application to slightly polar compounds and approximations in the equations result, however, in
slightly different values for , making it difficult to combine values from boiling points and values
from surface tension in estimating plasticizer-resin compatibility.
For many plasticizers, particularly new ones for which adequate physical data are lacking, the
method of estimation of that Small [188] devised has been very useful. For higher-molecularweight plasticizers and polymers, it has been by far the most frequently used approximation of

Page 451

FIGURE 7 Energy of vaporization by the Hildebrand rule. The portion of the graph above 180C
is an extrapolation beyond his data. (From Refs. 24 and 97.)
As with many other thermodynamic quantities, it is possible to find a set of additive constants
for the more common groups in organic molecules that allow calculating the (EV)1/2 and
therefore . The additive constants are called molar-attraction constants F in the equations.

Small also supplied most of the molar-attraction constants of Table 1. To estimate solubility
parameter, it is only necessary to add the proper molar-attraction constants and divide by the
molar volume. The method gives a good first approximation but lacks precision for more polar
plasticizers. For PVC itself, the estimated value of 9.66 agrees well with the value 9.48
obtained by swelling PVC in solvents of known cohesive energy densities.

Page 452
TABLE 1 Group Molar-Attraction Constants for Determination of Solubility Parameters

Page 453

Page 454
The solubility parameter theory is quite successful in application to polymer solutions even
though it was devised for low-molecular-weight solutes [12]. Its application to plasticizers was
slower because determination of 6 for plasticizers (which are high boiling) is more difficult than
for solvents (which are low boiling), and because the theory does not work as well for PVC
without modification as it does for some other resins, notably acrylics and rubbers. The
solubility parameter concept is frequently more valuable after a semiempirical adjustment.
b. Adaptations and Adjustments of . Burrell [2427] applied the cohesive-energy density
concept in the surface coating field after adjustments for poorly, moderately, and strongly
hydrogen bonded solvents. He found that the solubility parameters calculated for esters,
ketones, alcohols, and so on, were too low when calculated from the Hildebrand rule by means
of Figure 7 compared to those determined from heats of vaporization. He recommended as
corrections:
1. For alcohols, add 1.4 to calculated .
2. For esters, add 0.6 to calculated .
3. For ketones, add 0.5 to calculated if the boiling point is below 100C; otherwise, add
nothing.
It later appears that no correction need be made for high-molecular-weight esters [178].
c. Three-Parameter Systems. Gardon [81] analyzed the effect of polarity separately from the
effect of hydrogen bonding and obtained improved fit between predicted and observed solubility
by a second parameter called partial polarity. Crowley et al. [49] recognized that the attractive
forces between organic molecules are of four distinguishable types: London forces (dispersion),
Keesom forces (dipole-dipole), Debye forces (dipole-induced dipole), and Lewis forces (acidbase forces, of which the most,important is hydrogen bonding). The solubility parameter is a
gross measure of all these forces, and because these different types of forces do not always
interact to promote solution, a similarity of solubility parameters does not always assure
miscibility. Adapting the spectroscopic methods of Gordy [85] and of Bondi and Simkin [20] to
determine the amount of hydrogen bonding, Crowley et al. could give a numerical value for this
parameter. This plus dipole moments (gas phase) and the solubility parameter give them a
three-dimensional system that defines the compatibility of solvents. Figure 8 shows the theory
applied to cellulose acetate. Solvents with dipole moments, hydrogen-bonding tendencies, and
solubility parameters that would put them inside the solid diagram would dissolve the resin.
Their system could probably be applied to PVC: the solid would have a different location and
shape.
Hansen [93], like many able investigators, at first underestimated the effect of chemical
affinities and specific interactions in the dissolution and plasticization of polymers. Later, he

developed a three-dimensional system for predicting solubilities similar to that of Crowley et al.
However, he broke Hildebrand's solubility parameter into three components, d, p, and h,
representing the contributions of the dispersion forces (d) polarity (p), and hydrogen bonding
(h) to the total value of . Calculated values plus some empirical trial-and-error adjustments
led to graphical solid volumes similar to Figure 8 that defined with unusual precision solubilities
of 33 polymers in 90 solvents and 10 plasticizers. One of the resins is PVC.

Page 455

FIGURE 8 The three-parameter solventability system applied to cellulose acetate butyrate. A

similar volume should exist for PVC. (From Ref. 49.)
d. Solubility Parameters and Dielectric Constant . The early work on solubility parameters
was oriented toward solvents. Later, became used in another scheme together with dielectric
constant for predicting compatibility of plasticizers for PVC [58]. None of the parameters so far
mentioned is truly a measure of compatibility, but compatibility of a plasticizer with any resin is
a function of its chemical makeup. The number, kind, and arrangement of the atoms will govern
the forces involved. These will all influence the energy of vaporization and . They will also
influence the dipole moment, hydrogen bonding, viscosity, and so on, but not in the same
proportions. Measuring any one of these properties gives one view of the plasticizer molecule,
which can then be related to its compatibility with any given resin (Tables 2 to 4). For example,
if its value lies between 7.8 and 13.4 (cal/cm3)1/2, it will be compatible with polyvinyl acetate
if it is moderately, but only moderately, hydrogen bonded. Measuring a second property gives a
second view of the plasticizer molecule, which may also be related to its compatibility. Two or
more views from wisely chosen properties can theoretically pinpoint the compatibility behavior
of a plasticizer with any given resin. Dielectric constant is a measure of these intermolecular
forces and their effect on the molecule's ability to adapt itself to the electrical forces of its
environment. Solubility parameter and dielectric constant can then be combined as predictors of
plasticizer compatibility (Figure 9). Classical plasticizers for PVC have values between about
8.4 and 11.4 and values between about 4 and 8 to 10 [58,113]. These predictions can be
extended to other resins [180].
Dielectric constant is strongly influenced by dipole moment and hydrogen bonding. Inspection of
the curves of Figure 9 indicates that the slopes and intercepts are governed to a large extent
by these parameters. High dipole moments (e.g., 3.9 for alkyl nitriles versus 0 for paraffins)
appears to lead to low slope and high intercept, but strong hydrogen bonding (e.g., 18 to 20 for
alcohols versus 0 for paraffins) accompanies high slope and low intercept. Hydrogen bonding is
opposite to polarity, as Gardon [81] has observed.

Page 456
Solubility parametera Dielectric constantb (at I kHz)
Polyvinyl chloride, rigid
Polyvinyl chloride, plasticized
Polyvinyl chloride/acetate
Polyvinyl chloride, chlorinated
68.9% Cl
Polyvinylidene chloride
Polyethylene
Polypropylene, d = 0.906
Ethylene/vinyl acetate
Polymethyl methacrylate
Polyvinyl formal
Polyvinyl butyral
Polyvinyl butyral (flexible)
Polyvinylidene fluoride
Polytetrafluoroethylene
Nylon 66
Cellulose acetate (sheet and molding)
Cellulose acetate /butyrate (molding)
Cellulose nitrate, 1/2 sec
Ethyl cellulose (sheet and molding)

9.489.7
10.4

3.03.3
4.08.0
4.08.0

9.77
12.212.6
7.9
8.07
8.618.65
9.25
ave. 11.0
ave. 11.4
7.11
6.2
13.6
10.9

3.55.0
2.252.35
2.22.6
2.602.98
33.5
3.6
3.5
5.10
7.7
<2.1
3.94.5
3.47.0
3.46.4
7.0
3.04.1

11.5
10.3

aFrom Refs. 2427, 184, and 178.

bFrom Refs. 141 and 178.
Salomon and Van Amerongen [172] considered hydrogen bonding to overshadow polarity. It
tends to neutralize polarity.
The two horizontal dashed lines of Figure 9, at values of 8.4 and 11.4, span the value of 9.6
reported for PVC. This span of 3.0 hildebrands (cal/cm3)1/2, found by direct observation of
compatible plasticizers, agrees with the span Gardon calculated from theory (2.7 to 3.6) [81].
The two vertical dashed lines are drawn at values of 4 and 8, the limits Kaufman mentioned
for compatibility with PVC [113]. Experience. shows that the upper line might be drawn as high
as = 10 for some types of plasticizers. The small rectangle

Page 457
bounded by the four dashed lines includes all the classical plasticizers for PVC. The predictions
based on and agree in this region.
From Figure 9 it appears that solvents (e.g., cyclohexanone) may lie well outside the rectangle.
There may be several factors combined to cause this, but at least one is molecular size. Figure
10 shows this effect, in stylized form, as seen in actual compatibility studies. The shape of the
curve at very high molecular weight is unknown for two reasons. When polymeric materials are
blended with another polymer such as PVC, thermodynamic solubility can be achieved at mill
temperature, but a metastable condition is produced on cooling. Although exudation
theoretically should occur, the high molecular weights slow down the process, so what has
been designated operational compatibility is observed [92]. The second problem in
determining the upper part of the curve in Figure 10 is that additives of very high molecular
weight usually produce rigid polyblends, and polarity requirements probably differ for rigid and
flexible vinyl.
The theoretical basis for the narrow solubility limits at higher molecular weight has been pointed
out at times [93,97]; it uses the equation
H = Vm(I - 2)2VI V2
(7)
where
H = heat of mixing
Vm = average molar volume of the solvent
1 = solubility parameter of the solvent
2 = solubility parameter of the polymer
V1 = volume fraction of the solvent
V2 = volume fraction of the polymer
For the free energy G of equation (1) to be negative (indicating solubility), a low value of H is
desirable. Since H is proportional to the molar volume of the solvent in equation (7), it is seen
that as the molar volume increases from that of ordinary solvents to that of monomeric
plasticizers on up to polymeric plasticizers, the value of H tends to increase. To reduce the
value of H, the solubility parameters of polymer and solvent (or plasticizer) must become more
and more alike. The dielectric constant limits of about 4 to 810 are at present empirical. That
they agree closely with the lower and upper limits observed for PVC itself [161] at
temperatures above its most rigidly ordered state may help understand them. In other words,
the dielectric constant of the plasticizer must be similar to the dielectric constant of
unplasticized PVC in its rubbery-to-fluid state rather than its rigid state [178].
From the thermodynamic point of view, perhaps the reason any of the modifications of the
solubility parameter by polarity, dipole moment, and so on, have merit is that in some way they
adjust for the nonideality of entropy of mixing.
4. Interaction Parameters
The theory of Flory and Huggins is based on a statistical mechanical treatment of a lattice
model of polymer solutions. According to theory, the free energy of mixing of a polymer solution

is given by

Page 458
a Material trade namea Solubility parameter, [hildebrands (cal/cm3)] Dielectric constant (at
10 kHz)b Dipole moment,c Hydrogen-bonding,d Flory-Huggins interaction parameter,
Compatibility solubility Mill or loope Lit.f C Acetates Ethyl 9.1 25 6.02 1.8 8.4 I n-Propyl
8.75 19 5.69 1.9 8.6 Isopropyl 8.6 1.9 8.6 Butyl 8.55 20 5.01 1.9 8.8 0.44 I Isobutyl 8.3 20
5.29 1.9 8.8 I Amyl 8.45 20 4.75 1.8 9.0 Ethyl acetoacetate 22 15.7 0.56 Acids Formic 12.2 16
58. 1.52 I Acetic 12.6 20 6.15 1.74 20.0 I n-Butyric 11.5 20 2.97 1.65 I n-Valeric 10.8 2.6
Acrylates and methacrylates Butyl acrylate 8.9 25 5.05 2-Ethylhexyl acrylate 7.8 25 4.32
Lauryl. methacrylate 25 3.53 Cyclohexyl methacrylate (11) 25 5.00 Adipates Dimethyl (5) 9.64
24 7.45 Diethyl 9.19 25 6.15

Page 459
Diisohexyl (11)
Dicyclohexyl
Di-2-ethyhexyl (DOA) (10)
Heptyl nonyl
Diisodecyl (DIDA) (10)
Didodecyl
Dioctadecyl
Benzyl 2-ethylhexyl (11)
Diallyl (11)
Alcohols
Water
Methanol
Ethanol
n-Propanol
n-Butanol
Isobutyl alcohol
n-Amyl alcohol
n-Hexyl alcohol
Cyclohexanol
n-Heptyl alcohol
n-Octyl alcohol
2-Ethylhexanol
Benzyl alcohol
Glycol
Glycerol
Aldehydes
Acetaldehyde
n-Heptaldehyde
Benzaldehyde

8.46
8.5
8.25

23.4
14.5
12.7
11.9
11.4
10.5
10.9
10.7
11.4
10.0
10.3
9.5
11.97
14.2
16.7
9.8
9.7
10.4

25
35
25
25
25
25
25
25
25

4.40
4.82
4.13
4.15
3.74
2.59
2.47
4.50
5.46

78.54
32.6
24.3
20.3
17.8
17.9
13.9
13.3
15.0
12.1
10.3
90 4.41
13.1
37.
42.5
21.1
9.07
17.8

0.23
0.28

C
C
S-M
H-S-C
H

1.85
1.7
1.7
1.7
1.7
1.8
1.7
1.7
1.7
1.7
1.7
1.7
1.7
2.3
~2.3

39.0
18.7
18.7
18.7
18.7

18.7
18.7
18.7
18.7
18.7
18.7
20.6

3.15

I
I
I
I
I

I
I

2.51
2.77

11.7
8.4

Sol

Page 460
a Solubility parameter, [hildebrands (cal/cm3)] Dielectric constant (at 10 kHz)b Dipole
moment,c Hydrogen-bonding,d Flory-Huggins interaction parameter, Compatibility solubility Mill
or loope Lit.f C Amides Acetamide 18.7 74. ~3.4 Dibutyl. stearamide 9.0 N,NDimethylformamide 12.1 25 37.2 2.0 11.7 V.S1 N,N-Dimethylacetamide 10.8 25 37.8 2.0 12.3
Azelates Diethyl 30 5.13 Di-2-ethylhexyl (DOZ) (6) 8.44 25 4.04 C Benzoates Methyl 10.5 20
6.59 1.9 6.3 Ethyl 9.7 20 6.02 -2.0 6.3 Isobutyl 20 5.38 Amyl 20 5.00 Benzyl 10.06 20 5.18 6.3
Dipropylene glycol di- (15) 9.61 25 7.44 C Diethylene glycol di- (15) 10.10 25 7.12 C
Nonylphenyl (11) 25 3.76 C Pentaerythritol tetra (11) (extrapolated) 25 4.95 2,2,4-Trimethyl1,3-pentanediol monoisobubyrate nbenzoate (6,17) 8.82 25 5.08 C

Page 461
Butyrates
Ethyl
n-Butyl
Isobutyl
Isoamyl
2,2,4-Trimethyl-1, 3-pentanediol diisobutyrate (6)
Sucrose acetate isobutyrate
Citrates
Tributyl (4)
Acetyl tributyl (4)
Acetyl tri-2-ethylhexyl (4)
Acetyl tri-n-octyl, decyl (4)
Epoxides
Ethylene oxide
Epoxidized soya oil (13)
Ethers
Di-n-butyl
Dioxane
Anisol
Furan
Tetrahydrofuran
Fumarates
Diethyl
Dibutyl
Dihexyl (11)
Diisooctyl (11)
1,3-Butanediol bis (2-ethylhexyl-) (11)

8.15
8.0
8.32
8.5

185.10 1.74
1.9 8.8
204.1
204.0
254.78
253.56

C
NSM

9.04
9.02
8.60

257.62
266.02
254.83
254.42

C
C
C
vs

11.1
~8.9

-1 13.9 1.9 10.0

235.47

7.1
10.0
9.9
9.1
9.52

203.09 1.2 11.0

252.21 0
9.9 0.47
254.33 1.4 7.0 0.51
252.94 0.71
257.58 1.6 11.5 0.21

8.9-9.4 236.56
8.94
254.68
254.16
255.26

I
Sol

0.38 NSM
0.29
C
0.65

C
S

Page 462
a Solubility parameter, [hildebrands (cal/cm3)] Dielectric constant (at 10 kHz)b Dipole
moment,c Hydrogen-bonding,d Flory-Huggins interaction parameter, Compatibility solubility Mill
or loope Lit.f C Glutarates Diethyl 30 6.66 Dinonyl 20 4.0 Diisodecyl (7) 25 3.85 H
Glycolates Butyl phthalyl butyl (10) 9.6 25 6.80 C Ethyl phthalyl ethyl (10) 10.2 25 8.00 0.56 MH Methyl phthalyl ethyl (10) 10.5 25 8.39 0.75 M-H Dibenzyl diglycolate (11) Supercooled liquid
25 8.24+ Solid mass 25 3.14 Halogenated hydrocarbons Chloroform 9.3 25 4.608 1.2 1.5 ~I
Chlorobenzene 9.5 25 5.62 1.6 1.5 0.52 ~I 1,1,2,2-Tetrachloroethane 9.7 20 8.2 1.36 1.5
Ethylene chloride 9.8 25 10.36 1.86 1.5 0.51 o-Dichlorobenzene 10.0 25 9.93 2.27 1.5 ~I mDichlorobenzene 25 5.04 1.72 p-Dichlorobenzene 95 2.41 0.0 Chlorinated biphenyls 32% Cl
10.3 25 5.7 C

Page 463
42% Cl
48% Cl
54% Cl
68% Cl
Chlorinated terphenyl 60% Cl
Hydrocarbons
n-Hexane
Cyclohexane
Octane
Kerosene
Benzene
Toluene
Ethylbenzene
Styrene
n-Propylbenzene
Mesitylene
Isobutylbenzene
t-Butylbenzene
Biphenyl (liquid)
Isopropylbiphenyl
Isopropylbicyclohexyl
o-Terphenyl
m-Terphenyl
p-Terphenyl
1-Methylnaphthalene (5)
o-Xylene
m-Xylene
p-Xylene
2-Methylnaphthalene (5) (extra-polated)
Tetralin

10.6
10.8
11.2
12.2
11.6

25
25
25
25
25

5.8
5.6
5.1
2.5
2.5

7.3
8.2
7.55
7.2
9.15
8.9
8.8
9.3
8.65
8.9

25
25
25
25
25
25
20
25
20
20
17
20
75
23

1.88
2.01
1.95
2.09
2.23
2.38
2.41
2.43
2.36
2.28
2.35
2.38
2.53
2.62

25
25
25
25
20
20
20
25
20

2.8
2.86
2.95
2.84
2.57
2.37
2.27
2.90
2.8

9.2
8.2
10.03
10.16
10.26
9.0
8.9
8.8
9.5

C
C
C

1.4

0
0
0

0
0

0
0.4
0.6
0

0
4.5
1.5
1.5

0
0.36
0.36

I
I

0.85
0.80
1.11
0.66
1.46
1.32

~I
~I
I
I

2.24
0.57

2.4
0.59

0.45
0.30
0.02
0.6

I
I
I
0.31
0.54 I

Page 464
a Solubility parameter, [hildebrands (cal/cm3)] Dielectric constant (at 10 kHz)b Dipole
moment,c Hydrogen-bonding,d Flory-Huggins interaction parameter, Compatibility solubility Mill
or loope Lit.f C Partially hydrogenated terphenyl mixture (10) 9.02 25 2.53 M
Polyalkylated naphthalene (2) 25 2.51 NSM Sovaloid C (14) 25 2.89 C Ketones Acetone 10.0
25 20.7 2.9 9.7 0.62 I Sol 2-Butanone (MEK) 9.3 20 18.5 2.7 7.7 0.42 Sl Sol 2-Heptanone 8.5
20 11.9 2.7 7.7 0.2 Cyclohexanone 9.9 20 18.3 2.7 11.7 0.23 Sol 4-Methylpentanone-2 (MIBK)
8.6 20 13.1 2.7 7.7 0.29 Sl Sol Benzophenone (liquid) 7.5 50 11.4 C d-Camphor 9.0 20 11.35
Lactates Ethyl 10.0 25 15.4 1.9 7.0 Isoamyl 25 9.60 Maleates Diethyl 9.9 23 8.58 2.54 0.64
Di-2-ethylhexyl (10) 25 5.71 0.36 C 1,3-Butyleneglycol bis(2-ethyl-hexyl-) (11) 25 5.75

Page 465
Malonates
Dimethyl
Diethyl
Diethyl benzal
Nitriles
Acetonitrile
Propionitrile
Butyronitrile
Valeronitrile
Isocapronitrile
Malononitrile
Benzonitrile
Fatty acid nitrile
Nitro compounds
1-Nitropropane
2-Nitropropane
Nitrobenzene
o-Nitrotoluene
o-Nitrobiphenyl ONB
o-Nitrodiphenyl ether
Oleates
Butyl
Oxalates
Diethyl
Palmitates
Ethyl
n-Butyl

11.0
10.3

20 10.3
25 8.03
20 7.6

11.9
10.7
10.5
9.6
10.2
15.1
8.4

20
20
21
21
22
36
25
23

37.5
27.2
20.3
17.4
15.5
46.
25.2
8.06

3.9
4.0
4.0
3.57
3.53

6.3
7.7
7.7
8.4

4.4

7.7

10.3
9.9
10.0
10.4
10
10.16

30
30
25
20

23.2
25.5
34.82
21.6

3.7
3.7
4.3

8.39

25 4.0

8.6

21 8.1

8.28

20 3.20

2.54

I
I

C
2.5 0.56
2.8 0.29
4.9

3.79

1
2.49

6.3

1.7

C
C
C
C
NSM

~I
Mod

Page 466
a Solubility parameter, [hildebrands (cal/cm3)] Dielectric constant (at 10 kHz)b Dipole
moment,c Hydrogen-bonding,d Flory-Huggins interaction parameter, Compatibility solubility Mill
or loope Lit.f C Phenols Phenol 14.4 60 9.78 7.0 m-Cresol 10.2 25 11.8 1.54 7.0 I pCresol 58 9.9 1.54 Nonylphenol 9.4 Phosphates Tributyl 8.56 25 8.05 3.07 11.5 -0.53 Tri-2ethylhexyl (16) 8.23 25 4.90 -0.30 C Triphenyl 10.10 9.6 C Tricresyl (TCP) (10) 8.4-9.9 25 7.20
C Cresyl diphenyl (10) 10.1 25 7.60 C t-Butylphenyl diphenyl (10) 10.18 23 7.33 2-Ethylhexyl
diphenyl (10) 7.8 25 7.33 -0.35 C Isodecyl diphenyl (10) 9.03 24 6.67 Tributoxyethyl (10)
(extrapolated) 8.57 23 8.7 -0.16 Exud. Tris (tetrahydrofurfuryl) (8) (extrapolated) 25 10.7 C
Dibutyl phenyl 8.7 C ortho-Phthalates Dimethyl (10) 10.5 25 8.52 0.52 C Diethyl (10) 9.9 25
7.56 0.34 P Dibutyl (10) 9.4 25 6.42 5.7 -0.05 C

Page 467
Dihexyl
Diisoheptyl
Di-2-ethylhexyl (DOP) (10)
Diisoctyl (DIOP) (10)
Dicapryl (DCP) (9)
N-Octyl N-decyl (DNODP-6,10) (10)
Di(hepyl nonyl) (10)
Diisononyl (DINP)
Di-n-decyl (DNDP)
Diisodecyl (DIDP-E) (10)
Di(heptyl nonyl undecyl) (10)
Diundecyl (DUP) (10)
Ditridecyl (DTDP) (10)
N-Butyl cyclohexyl (1)
Dicyclohexyl (10) supercooled liquid
Butyl benzyl (10)
Isohexyl benzyl (11)
Alkyl benzyl (10)
Isodecyl benzyl (11)
7-(2,6,6,8-Tetramethyl-4-oxa-3-oxononyl) benzyl (10)
Dibenzyl (11)
Supercooled liquid
Solid mass
Dibutyl. Cellosolve
Diphenyl (extrapolated) (10)
Diallyl (8)
High-molecular-weight phthalate(12)
Phthalic acid polyester (3)
Thiodiethyleneglycol bis(benzylphthalate) (11)
Ditetrahydrofurfuryl (16)

9.02 23.5 5.69

-0.09 C
~9.0 23 5.25
C
8.8 25 5.18 2.84
C
25 4.90
C
8.7 25 5.00
C
7.36 25 4.70
C
25 5.04
C
24 4.67
5.8
C
0.19 S-M
8.56 25 4.46
C
8.92 23 4.76
5.6
C
8.77 23 4.43
5.8
M
8.5 25 4.05
S-H
9.53 25 6.44
-0.25 C
8.10 25 6.42
C
9.88 25 6.41
0.10 C
25 5.97
C
9.55 23 5.70
C
25 5.17
C
9.35 25 5.86
5.7
C
25
25
9.21
10.3 25
9.75 25
25
25
25
9.63 25

6.32
3.06

C
Bloom
0.43

6.80
6.68
7.41
8.9
7.77
9.71

C
0.17 C
C
C
-0.13 C

Page 468
a Solubility parameter, [hildebrands (cal/cm3)] Dielectric constant (at 10 kHz)b Dipole
moment,c Hydrogen-bonding,d Flory-Huggins interaction parameter, Compatibility solubility Mill
or loope Lit.f C Hydrophthalates Di-2-ethylhexyl tetrahydro-phthalate 20 4.5 Di-2ethyhexyl hexahydrophthal-ate (16) 8.58 4.56 C Isophthalates Dibutyl (11) 25 5.58 -0.01 C
Diisohexyl (11) 25 4.88 C Di-2-ethylhexyl (16) 8.87 25 4.21 0.31 M Diisooetyl (11) 25 4.33 M
Djisodecyl (11) 25 3.99 Exud. Dinonylphenyl (11) -9.9 25 3.29 Exud. Terephthalates Dibutyl (5)
23 5.59 0.01 C Di-2-ethylhexyl (DOTP) 23 4.48 5.1 0.46 H Polyesters and permanent
plasticizers Santicizer 405 (10) 9.50 25 8.9 C Santicizer 409 (10) 9.35 25 6.03 0.23 C
Santicizer 411 (10) 9.34 25 6.18 C Santicizer 412 (10) 9.34 24 5.45 C

Page 469
Santicizer 429 (10)
Paraplex G-30 (13)
Flexol R-2-H (16)
Kodaflex: NP-10 (6) (extrapolated)
Polymeric flow aid Modaflow (10)
Propionates
Ethyl
Propyl
Butyl utyl
Isoamyl
Pyromellitates
Tetra-2-ethylhexyl (11)
Tetralsooctyl (11)
Tetradecyl (11)
Sebacates
Dimethyl (5) (supercooled liquid)
Diethyl
Dibutyl
Di-2-ethylhexyl (DOS) (13)
Didodocyl
Dioctadecyl
Didocosyl
Dibenzyl
Stearates
Ethyl
Butyl

9.20

8.55
8.4
8.5
8.6
8.4

9.05
8.68
8.45

9.44

24
25
23
23
19
20

6.03
6.37
5.34
5.3
4.36

C
C
-0.18 C
C

20

5.65 1.74
4.7 1.9 7.0
1.9 8.4
4.2

25
25
25

5.03
4.77
4.31

25
25
26
25
25
25
25
0.54

5.37
4.79
4.73
3.88
2.64
2.43
2.34

0.24
0.16

0.53

3.1
3.11 1.88

1.6
>1

NSM

10.17 28
8.21 25

VS-M
M
Exud.

C
C

Page 470
a Solubility parameter, [hildebrands (cal/cm3)] Dielectric constant (at 10 kHz)b Dipole
moment,c Hydrogen-bonding,d Flory-Huggins interaction parameter, Compatibility solubility Mill
or loope Lit.f C Succinates Diethyl 30 6.64 2.64 Di-n-butyl (5) 25 5.02 Di-n-octyl 0.46
Tetrabutyl thiodi- (16) 25 5.68 Sulfonamides N-Ethyl o,p-toluenesulfonamide(extrapolated) (10)
14.39 25 36.5 (N)8. 5 0.63 Exud. (O) 23.4 o,p-Toluenesulfonamide (solid) (10) 16.22 25 3.76
Sulfonamide-formaldehyde resin (10) 14.9 23 -24.0 C Tartrates Dibutyl 9.24 41 9.4 S-M

Page 471
Trimellitates
Trihexyl (11)
Tri-isodecyl (4)
Tri-2-ethylhexyl (TOTM)
Tri(heptyl, nonyl) (10)
Valerates
Methyl

9.00
8.75

25
25
23
23

5.30
4.19
4.70
4.81

7.9

19

4.3

5.4

C
M
M
M

aManufacturers of plasticizers used at the time of testing: (1) Allied Chemical Corp.; (2) Amoco
Chemicals Corp.; (3) Ashland Chemical Co., Div. Ashland Oil, Inc.; (4) Pfizer, Inc.; (5)
Distillation Products Industries; (6) Eastman Chemical Products; (7) E.I. dupont de Nemours
and Co.; (8) F.M.C. Corp.; (9) Hatco Chemical Div., W. R. Grace & Co.; (10) Monsanto
Company; (11) Monsanto Company (experimental product); (12) Reichhold Chemicals, Inc.;
(13) Rohm & Haas Co.; (14) Mobil Chemical Co., Inc.; (15) Velsicol Chemical Corp.; (16) Union
Carbide Corp.; (17) Tenneco Chemicals, Inc.
bC = temperature at which was measured; , dielectric constant (the values are normally
those determined at 10 kHz; some old literature values are for much lower or higher
frequencies [58]).
cAll x 10-18 esu; from Ref. 49.
dFrom Refs. 49 and 178.
eNSM, no sheet on roll mill incompatible; Exud., exudation either from milled or molded sheet
or occasionally in the loop; C, compatible no exudation, even in the loop; V, very; S, slight
exudation in the loop; M, moderate exudation in the loop; H, heavy exudation in the loop.
fFrom the literature: C, compatible, according to Modern Plastics Encyclopedia; P, partially
compatible, according to Modern Plastics Encyclopedia; I, insoluble, according to Hansen [93];
Sol, soluble, according to Hansen [93] ; Sol, slightly soluble, according to Hansen [93].

Page 472
a a a (25C) Cyclohexanone 1.33 Best 9.7 0.18 18.3 Tetrahydrofuran 1.22 9.1 0.21 7.39
Cyclopentanone 1.20 10.45 0.19 13.45 (20C) Morpholine 1.17 10.2 0.24 7.46 Nitrobenzene
1.03 11.0 0.30 38.8 Tetrahydropyran 1.00 Poorest 8.7 0.355 aData from Ref. 143.

FIGURE 9 Classical plasticizers for PVC have solubility parameters between 8.4 and 11.4 and
dielectric constants between about 4 and 810. Curves: 1, carboxylic acids; 2, hydrocarbons;
3, chlorinated hydrocarbons; 4, esters; 5, ethers; 6, alcohols; 7, alkyl nitriles; 8, aldehydes; 9,
ketones; and 10, nitro compounds. (From Ref. 58.)

Page 473

FIGURE 10 Influence of molecular size on the permissible dielectric constant range. (From Ref.
58.)
G
= RT(n1 In v1 + n2 In v2 + n1 v2
(8)
where
G
= free energy of mixing
R
= gas constant
T
= absolute temperature
n1
= number of moles of solvent
n2
= number of moles of solute
v1
= volume fraction of solvent
v2
= volume fraction of solute

= interaction parameter
If G is negative, a polymer and solvent will form a solution, or presumably a polymer and a
plasticizer will be compatible. Since some experimentation is required to measure the effect of
the solute and the solvent on each other, and the measurements are then used for calculation of
, this interaction parameter is a semiempirical constant that adjusts for nonideality of both the
heat and the entropy of mixing. In reality, it is not truly a constant, for it can vary significantly
with, for example, concentration [9,81,87, 95]. In spite of this, the concept has proved
sufficiently useful for many equations involving to have been developed and for various types
of measurements to be correlated with polymer-solvent (or plasticizer) compatibility.
Sheehan and Bisio [184] give some of these collected from the literature. Among them are
equations for calculation of (old ) from the stress-strain behavior of swollen polymers,
osmotic pressure, equilibrium swelling, vapor pressures, viscosities, and depression of melting
or freezing points. The last method has given good results in extensive studies of swelling and
dissolution of various resins. For this, a convenient test devised by Anagnostopoulos et al. [46]
made use of the lowering of the melting point as observed

Page 474
in a particle of resin immersed in a droplet of plasticizer on a microscope hot stage. Brooks and
Lawrence extended the test with success to cellulose nitrate [22]. In these cases, the more
general equation

is simplified by the observation that the volume fraction of the solvent v1 approaches unity at
the apparent melting temperature. The equation becomes

Thus, for any given polymer, when the melting temperature T , the heat of fusion Hu, and the
molar volume of the polymer repeating unit Vu are known, X values for any number of
plasticizers can readily be determined from their molar volumes V1 and the depressed melting
points Tm. This simplified equation compresses the value of to an upper limit of unity. Since
according to the theory, must be 0.5 or lower to indicate compatibility for high-molecularweight polymers with low-molecular-weight plasticizers, this compression of high values is of
little practical concern. As the molecular weight of the plasticizer increases, the permissible
values of increase to a maximum of 2 when the plasticizer and resin have the same molar
volume [81].
The graphical relation between values and dielectric constants is shown in Figure 11 for
several types of plasticizers. The dip in each ester curve shows the acceptably low values of
(>0.5) right in the 4 to 810 range of dielectric constant. Table 3 lists values for the
interactions of plasticizers with PVC.
Heil and Prausnitz [95,96], aware that the one-parameter system of predicting solubility has
serious limitations, proposed a two-parameter system that visualized the polymer more nearly
as segments interacting with the solvent. It is reported to give excellent results regardless of
the nature of molecular interactions, molecular weights, multiple component solvent systems,
concentrations of resin or solvent, and so on. It is based on the activity coefficients of resin and
solvents as determined through vapor pressure measurements. Although earlier theories might
predict compatibility or incompatibility with or without actual measurements, they could give, at
best, only a hazy idea of how much resin would dissolve in a solvent or how much plasticizer
would be compatible with a resin. Activity data used in the Heil-Prausnitz manner have
delineated closely the solubility boundaries for such systems as polystyrene in methanol,
benzene, and acetone toluene mixtures. Doolittle suggests that application of the theory of
plasticized resins should allow prediction of composition at the point of phase separation, and
so designate the limits of compatibility [65].
B. Primary and Secondary Plasticizers
A working knowledge of the limits of compatibility, vague and variable as these limits are now
known to be, is vital to the plasticizer industry. There is a common idea among those who work
primarily with polymer solutions that

Page 475

FIGURE 11 Compatibility of various homologous series: the relation between and . (From
Ref. 58.)
a given resin is either entirely soluble or completely insoluble in a given liquid and that there will
be few liquids . within which solubility is considerable but incomplete [97,142,143]. It is not
at all uncommon, however, to deal with plasticizers that have considerable but incomplete
compatibility with a given resin. Solvents with limited solvent power to dissolve a given resin and
plasticizers with limited compatibility are found to be near the permissible limits in solubility
parameter, polarity, or hydrogen-bonding ability [93].
A primary plasticizer for a given polymer or resin is one which is compatible enough with that
polymer that it does not separate from the polymer into a discrete phase when used at high
enough concentration to achieve the desired degree of plasticization. A secondary plasticizer is
one that is not compatible enough to achieve the desired degree of plasticization without
separation of polymer and plasticizer into more or less discrete phases, although the plasticizer
might be completely compatible at some lower concentration. This separation may be as
exudation, which is the separation into liquid

Page 476
droplets or a liquid surface film and is at times called sweat out. It may occur as bloom,
which is the development of tiny crystals on the surface or slightly under the surface. In
borderline cases, it may occur only as a hazy or milky plastic sheet rather than a transparent
one. Secondary plasticizers are used in combination with primary plasticizers to achieve the
needed compatibility. Sometimes they are used to lower cost, but they are also used to achieve
some physical or chemical property that cannot readily be developed without them. The
distinction between primary and secondary plasticizers is vague, depending partly on resin
used, concentration desired, and environment and conditions of end use. In PVC, phthalates
such as DOP and butyl benzyl phthalate are sufficiently compatible in almost all environments to
be primary plasticizers, but alkylated aromatic hydrocarbons with adequate dispersion forces
and suitable solubility parameters are too low in polarity and perhaps in hydrogen-bonding
ability to be primary plasticizers. This is in spite of their excellent polarizability [58]. The
compatibility of a plasticizer with a given resin may be influenced by factors such as pressure,
temperature, humidity, and sunlight.
C. Temperature Dependence of Compatibility
The effect of temperature as a puzzling part of the mechanism of solvation and plasticization
has been observed through the years [172] and was partly explained by Doolittle in his
mechanistic theory [67]. With polymer-polymer forces tending toward aggregation of polymer
macromolecules and resultant syneresis or exudation of plasticizer, and polymer-plasticizer
forces tending toward disgregation and solvation, the system is in dynamic equilibrium at a
given temperature. As the temperature is changed, the magnitudes of the effects of these
forces are changed but not to the same degree. With more energy present (higher
temperature), it becomes easier in many cases to separate polymer-plasticizer bonds (favoring
aggregation). This is the Doolittle effect. For lower temperatures, the reverse is often true.
Depending on the relative magnitude of these changes, a given liquid may be a better solvent
for a polymer when it is hot than when it is cold, although another liquid may be a better solvent
cold than hot. Doolittle, in his studies of solvents for cellulosics, examined them by size and
structure. He has shown that in many homologous series the low-molecular-weight members
have a positive temperature dependence and are normal, being better solvents hot than cold,
while higher-molecular-weight members have a negative temperature dependence, being better
solvents cold than hot. The latter give solutions that gel on heating. Similarly, he found that
molecules that are more compact have a positive temperature dependence and that the more
extended molecules have a negative dependence. Also, when the active groups are buried in
the solvent molecule rather than located near the ends, the temperature dependence become
negative.
Most plasticizer molecules tend to be high-molecular-weight extended molecules with buried
active groups, all factors that tend toward abnormal temperature dependence. It is possible,
therefore, to find plasticizers with borderline compatibility that are perfectly compatible at room
temperature but exude at 100C, such as diisodecyl adipate in PVC. Upper- and lowertemperature regions of solvent-resin incompatibility are discussed by Delmas and Patterson
[63].

Since a plasticized resin resembles a gel in its internal structure, a plasticizer that tends to
liquefy a gel on cooling tends to counteract to some extent the natural tendency of the plastic to
stiffen and become more brittle on

Page 477
cooling. Thus adipates in plasticized PVC give products that remain rubbery over a wide
temperature range.
The many structural intricacies observed by Doolittle are summed up in the stylized
thermodynamic graph (Figure 12). The solubility parameters of resins and plasticizers both drop
with increasing temperature, but the drop for plasticizers (liquid) is steeper than for polymers
(solids). If a plasticizer has a solubility parameter lower than that of the resin and just barely
compatible at room temperature, it will exude on heating. If the plasticizer is barely compatible
warm, with a solubility parameter higher than the resin, it will exude on cooling. This is assuming
no change in crystallinity of the resin. The melting of crystallites produces more amorphous
polymer to accept more plasticizer; conversely, an increase in crystallinity can exclude
plasticizer and cause exudation. Such changes are independent of the Doolittle effect.

V. EXTERNAL VERSUS INTERNAL PLASTICIZATION

The solvation-desolvation equilibrium is characteristic of all externally plasticized resins. Since
the plasticizer molecules are not attached to the resin by primary bonds, plasticizer can be lost
by evaporation, migration, or extraction. When side chains of about the same size as the
plasticizer molecules are attached to the resin, an internally plasticized resin results that does
not have the deficiency of plasticizer loss. Many such products have been developed usually
to be modified by external plasticizers themselves. The copolymers of vinyl chloride and vinyl
stearate are of this kind. Even the nitration or acetylation of cellulose is internal plasticization.
Generally, resins with much internal plasticization have neither adequate strength at room

FIGURE 12 Stylized representation of change of solubility parameters of resins and plasticizers

with temperature to indicate exudation at elevated temperatures. (From Ref. 180.)

Page 478
temperature nor satisfactory flexibility at low temperature. Permanently bound plasticizer chains
do not have the freedom to solvate and desolvate the center of force on the resin molecules.
Several advantages can be attributed to the dynamic equilibrium that external plasticizer
molecules establish with the resin macromolecules. It has been demonstrated that a portion of
the plasticizer can be displaced under concentrated pressure loads. Ordinarily, this might be
considered a limitation, but it may become an advantage under certain circumstances. Thus the
composition remaining after some plasticizer has been squeezed out by pressure will be more
firmly held together because local desolvation permits more resin-resin unions in that area.
Such reasoning explains in part why the tear strength of externally plasticized vinyl chloride
resins is greater than the tear strength of internally plasticized resins of a similar type. A larger
factor in tear strength, however, is that of crystallinity of the resin. A relatively small amount of
internal plasticization permanently prohibits much of the crystallinity of PVC segments, whereas
an external plasticizer is mobile enough to permit, even promote, crystallization. The abnormal
temperature response is another advantage that many external plasticizers hold. Since
increasing the solvent strength is equivalent to increasing the concentration of plasticizer,
cooling a plasticized composition has the effect, to a limited degree, of adding more plasticizer.
More precisely, when the temperature of a plasticized composition is lowered, the natural
stiffening tendency caused by the close juxtaposition of the molecules and by the decreased
mobility of the plasticizer is to some extent counteracted by the improvement in solventability
gained by the plasticizer. Since the dependence on temperature of both viscosity and solvent
strength varies, plasticizers differ greatly in low-temperature behavior. Generally, external
plasticization achieves better performance (but not necessarily better permanence) than internal
plasticization.

VI. PLASTICIZATION OF POLYVINYL CHLORIDE

Modification of PVC by plasticizers can be taken today as the baseline for plasticization, with
which plasticization of all other resins can be compared. This is because of its nearly ideal
acceptance of proper plasticizers, not its similarity to other resins. Polyvinyl chloride is unique in
its acceptance of large amounts of plasticizers with gradual change in physical properties from
rigid solid to soft gel or viscous liquid. Most plasticized resins undergo this transformation
rapidly on either a slight increase of plasticizer content or a slight increase in temperature.
Barshtein and Kotlyarevskii [11] calculate that PVC has a helical structure with repeating units
of C28H42Cl14. Applying Zhurkov's law that there must be one molecule of plasticizer to
block each polar group on the polymer chain to the helical picture of PVC shows the need of
45 parts of DOP to 100 parts of resin to complete plasticization instead of the totally unrealistic
620 phr (parts per hundred parts of resin) calculated on the basis of C2H3Cl. This helical
structure, together with the resin's ability to maintain some kind of molecular order, as shown
by disappearance and reappearance of crystallites in PVC solutions [199], must in part account
for such unique response [3].
In general, plasticizers for PVC and polyvinyl chloride-acetate resins are esters of aliphatic and
aromatic di- and tricarboxylic acids, and of phosphoric acid. Secondary plasticizers include

high-molecular-weight alkyl aromatic hydrocarbons and chlorinated aliphatic hydrocarbons.

Nitrile rubbers (NBR) with high nitrile content to afford compatibility, certain polyurethanes,

Page 479
and ethylene/vinyl acetate/terpolymers with carbon monoxide to introduce carbonyl units for
compatibility with PVC are polymeric plasticizers.
A. Methods of Compounding Plasticizers and Resin
There are four main ways in which plasticizers are incorporated into resins such as PVC [152]:
hot compounding, dry blending, plastisol casting, and solvent casting. The first three methods
require an energy input by heating, frequently accompanied by vigorous mixing. In the fourth
method, the free energy of the solvent-resin system and the small molecular size of the solvent
permit intimate mixing of plasticizer and resin without heating.
1. Hot Compounding
Hot compounding involves mixing the plasticizer and resin powder (with stabilizers, fillers,
pigments, and lubricants, as desired) and fluxing together in an intensive mixer such as a batchtype Banbury mixer or any of several continuous mixers or on a two-roll rubber mill at about 150
to 170C. The exact temperature depends not only on the resin (molecular weight, monomer
ratio if a copolymer, etc.) but on the plasticizer also. Although rapid heating and mechanical
mixing tend to equalize the behavior of plasticizers that differ in structure and molar volume,
differences among the plasticizers still show up strongly in the ease of fusion [58] and in the
hot-melt rheology of plasticized PVC [138].
During hot compounding, the resin-plasticizer blend initially offers very little resistance to mixing.
A slight increase may be noticed as dry blending takes place. As the resin-plasticizer blend
begins to fuse and plasticization occurs on the surface, resistance to mixing increases rapidly.
This is detected as a sudden increase in torque when fusion is done in a laboratory torque
rheometer [137, 197]; or it may be seen as an increase in electrical power required to operate
a Banbury mixer [16]. Rapid penetrating and solvating plasticizers such as butyl benzyl
phthalate and dihexyl phthalate are fast fusing and produce a high torque followed almost
instantly by the lower torque of the completely fused mass. The time for fusion is shortened, so
that the total power required is less than for slow-fusing plasticizers such as DOP and higher
dialkyl phthalates and adipates. A blend of fast and slow-fusing plasticizers achieves a
compromise of moderate torque and moderately fast time. This is seen in easier milling to
obtain superior tensile properties [40], in ribbon blending, Henschel mixing and calendering to
give sheets of improved tear strength [149], and in dry blend extrusion to obtain greater output
with higher gloss [29, 150].
These effects on torque are seen in Figure 13 for butyl benzyl phthalate. The design,
construction, and method of operation of the equipment influence fusion time and temperature
drastically, so that with increasing plasticizer content, the fusion temperature may be either
lowered [134] or raised [16], depending on the friction between resin particles. For direct
comparison of plasticizers, as in Tables 5 and 6, it is necessary to have a uniform test method
[16].
The decrease in hot-melt viscosity with increasing temperature is shown for butyl benzyl
phthalate with PVC homopolymer and 87:13 PVC/VAc copolymer in Figure 14. Again, for

comparison, McKinney showed that for resins of equal molecular weight, 20 phr of external
plasticizer plus 13% of internal plasticizer is equivalent to 50 phr of external plasticizer alone. Or
for equal vinyl chloride content, 15 phr of vinyl acetate copolymerized is equal to 30

Page 480

FIGURE 13 Fusion of PVC with different concentrations of butyl benzyl phthalate as shown by
the sudden increase in torque in a Plasticorder. The temperature at 500 m-g is arbitrarily
designated the initial fusion temperature. (From Ref. 134.)
phr of external plasticizer in reduction of viscosity of resins of equal molecular weight. Internal
plasticization by reducing the ability of PVC to form crystallites reduces viscosity of the melt
and toughness of the final product.
From the Banbury mixer or roll mill, the hot plasticized PVC may go to a calender, an extruder,
or a press to be fabricated into film, sheeting, or massive articles; or it may be pelletized for
future conversion to finished goods.
a. Dry Blending. Dry blending involves incorporation of the plasticizer into the PVC particles by
stirring at 50 to 110C to yield a dry powder barely distinguishable in outward appearance from
the original resin [31]. For this, certain grades of resins with a high degree of porosity in the
resin particles are preferred for their absorption power. The method of polymerization and kind
and amount of surface contaminants will influence the rate of plasticizer uptake [185]. Dry
blending takes place most readily when the mixture is heated above the resin's glass transition
temperature Tg (about 80C for homopolymers) but well below the fusion temperature (Tables
5 and 6) under conditions free from pressure on the particles [134]. There appears to be no
true dry-blending temperature threshold in the way that there is a temperature threshold for
fusion. Plasticizer diffusion into the amorphous regions of the particles above or even below Tg
is the rate-controlling step. Below Tg, years may be required for diffusion of plasticizers of
usual molecular volume [135].

Page 481
a Relative fusion temperatureb (C) Minimum fluxing temperaturec (C) Clear pointd (C)
Adipates Dimethyl 104 Diisobutyl 58 Dibenzyl 114 Dibutoxyethyl 135 Di-2-ethylhexyl (DOA) 111
120 144 Diisooctyl, (DIOA) 114 Diisononyl (DINA) 138 Di-n-octyl, n-decyl (DNODA) 117
Diisodecyl (DIDA) 133 148 156 Di-n-decyl 158 Azelates Di-2-ethyhexyl (DOZ) 118 145 149
Benzoates Diethylene glycol di- 72 3-Methyl-1, 5-pentanedioldi- 72 Dipropylene glycol di- 78
108 Polyethylene glycol di- 82 Citrates Acetyl tributyl 102 Acetyl tri-2-ethylhexyl 165 Epoxy
stearates Butyl 98 2-Ethyhexyl 125 Epoxy tetrahydrophthalates Di-2-ethylhexyl 100 Diisodecyl
120 Glycolates Butyl phthalyl butyl glycolate 75 Glycol esters (see also Benzoates) Triethylene
glycol dihexoate 110 Triethylene glycol di-2-ethylhexoate 132 Polyethylene glycol di-2ethylhexoate 132 Diethylene glycol dipelargonate 136 Hexahydro isophthalates Di-2-ethylhexyl
123 Hexahydrophthalates Di-2-ethylhexyl 112 130 Diisodecyl 145

Page 482
a Relative fusion temperatureb (C) Minimum fluxing temperaturec (C) Clear pointd (C)
Hydrocarbon HB-40 142 Isophthalates Di-2-ethylhexyl 123 129 Diisodecyl 154 Isosebacate Di2-ethylhexyl 155 Palmitate Octyl 165 Phosphates Triethyl 90 Dibutyl phenyl 90 Tricresyl (TCP)
79 78 92 Triphenyl (TPPa) 96 Cresyl diphenyl 74 78 100 Isooctyl diphenyl 101 Tributyl 48 103
2-Ethylhexyl diphenyl 69 71 105 Isodecyl diphenyl 114 Isodecyl dicresyl 117 Didecyl cresyl 107
Tri-n-hexyl 121 Di-n-octyl phenyl 128 Di-2-ethylhexyl phenyl 86 128 Tri-2-ethylhexyl (TOF) 77
110 142 Phthalates Dibenzyl 80 Phenyl cresyl 81 Diethyl (DEP) 58 54 Dimethyl (DMP) 59 88
Phenyl benzyl 88 Butyl benzyl 69 75 89 Butyl cyclohexyl 68 92 Dibutyl (DBP) 62 68 97 Octyl
cresyl 104 Modified benzyl phthalate Santicizer 213 105 Diphenyl (DPP) 108 Di-n-hexyl (DHP)
84 75 113 Diisohexyl 82 Butyl octyl (BOP) 119

Page 483
a Relative fusion temperatureb (C) Minimum fluxing temperaturec (C) Clear pointd (C)
Phthalates (Contd) Butyl decyl 122 Diisooctyl (DIOP) 85 126 Di-2-ethylhexyl (DOP) 84 105 127
Di-n-octyl (DNOP) 127 Dicapryl (DCP) 88 106 137 Diisononyl (DINP) 122 Diisodecyl (DIDP) 94
124 138 Di-n-nonyl 139 Di-n-decyl 107 143 Didodecyl 123 Ditridecyl (DTDP) 130 155 155
Polyesters Plastolein 9720 132 Admex 760 135 Kodaflex NP-10 135 Paraplex G-54 138
Santicizer 409 142 153 Paraplex G-53 143 Paraplex G-30 93 Flexol R-2-H 147 Morflex P-50
153 Sebacates Dibutyl 102 Dibenzyl 103 Di-2-ethylhexyl (DOS) 125 152 155 Stearate Butyl
156 Terephthalate Di-2-ethylhexyl (DOTP) 135 140 aTrademarks: HB-40 and Santicizer,
Monsanto Co.; Plastolein, Emery Industries, Inc.; Admex, Archer, Daniels, Midland Co.,
Ashland Chemicals, Sherex Chemical Co.; Kodaflex, Eastman Chemical Products; Paraplex,
Rohm & Haas Co., C. P. Hall Co.; Flexol, Union Carbide Corp.; Morflex, Chas. Pfizer and Co.
Inc. bThe temperature at which torque in a Plasticorder torque rheometer begins to increase
rapidly at the onset of fusion [197]. cThe lowest temperature at which a plastisol develops
sufficient physical integrity to permit being lifted from the fusion plate [79]. dThe temperature at
which a plastisol becomes clear [87].

Page 484
a Dry-Blend in Plasticorderb Time (seconds) in Henschel mixerc Di-n-octyl, n-decyl adipate 174
Butyl phthalyl butyl glycolate 120 2-Ethylhexyl diphenyl phosphate 126 Tricresyl phosphate 138
Dibutyl phthalate 69 Butyl benzyl phthalate 87 <240 Alkyl benzyl phthalate, Santicizer 261 287
Modified benzyl phthalate, Santicizer 213 126 Butyl benzyl phthalate/di(heptyl,nonyl, undecyl)
phthalate (30/70) 385 Di-2-ethylhexyl phthalate 186 462 Butyl benzyl phthalate/di(heptyl,nonyl,
undecyl) phthalate (15/85) 484 Di(heptyl,nonyl,undecyl) phthalate 499 Diisononyl phthalate 555
Diisodecyl phthalate 390 653 Polyester, Paraplex G-30 255 Polyester, Santicizer 409 472
aTrademarks at the time of this work: Santicizer, Monsanto Co.; Paraplex, Rohm & Haas Co.
bTime to dry-blend 380.25 g of mix (Opalon 650, 100 parts; plasticizer 52 phr, Eagle-Picher
Dry White Lead A 10 phr; no 33 clay 7. 0 phr) in a Plasticorder at stock temperature 88C and
mixer speed 63 rpm. cFormula: Geon 102EP 100 parts; plasticizer 72 phr, Mark WS stabilizer
2 phr, in a Henschel mixer, 80C, 2 min mixing at 1800 rpm followed by 3800 rpm until
amperage curve had peaked and plateaued for 40 sec.
The dry blend may in some cases have plasticizer held loosely on the surface [185] and if need
be, can be made more free flowing by addition of traces of a very fine secondary additive such
as calcium carbonate or various silicates or pigments [44]. If the plasticizer-resin blend is
prepared in a torque rheometer with gradual heating, a slight increase in torque occurs as the
dry blend is formed at the fluff point [133]. The time required to obtain a dry blend is very
dependent on the temperature as well as on the plasticizers. Plasticizers high in molecular
volume or very viscous require a longer time or a higher temperature, or both, for dry blending.
Diethyl and dibutyl phthalates are small enough to dry blend quite well below the Tg of the
resin.

Page 485

FIGURE 14 Melt viscosity (torque) temperature relations for PVC and PVC/ VAc (87:13) at
different levels of butyl benzyl phthalate. Intrinsic viscosities: homopolymer 0.72; copolymer
0.50. (From Ref. 134.)
Times to dry blend under standard conditions at 88C and some at 80C in different equipment
are shown in Table 6.
It appears that in both dry blending and fusion, the rate of plasticizer uptake is strongly
influenced by molecular size, but with its thermodynamic solvent power as shown by solubility
parameter, polarity, and so on, having much influence. The latter properties, however, have
even more influence on the peak torque at either dry-blend fluff point or at fusion [16, 58, 133].
The material made by dry blending only is not yet plasticized [160]. It may be stored indefinitely
as is, or it may be milled, molded, calendered, extruded, or used for fluid-bed coatings as
desired, since plasticization does take place on heating briefly above the fusion temperature
about 160C for the homopolymer.
2. Plastisol Casting and Related Techniques
Plastisol casting and related techniques require certain PVC resins, paste resins, designed so
that the plasticizer will wet the resin particle at room temperature but only very slowly penetrate
and solvate it. These stable, creamy dispersions of resin powders in plasticizers are called
plastisols. Not every plasticizer will produce a plastisol with suitable viscosity and fusion
properties. Plastisols are very rarely Newtonian but are most often shear thinning or shear
thickening, and at times strongly so. Plastisol viscosity is related to plasticizer-resin interactions
directly and indirectly in several ways. Plastisols containing strongly solvating plasticizers tend
to have higher viscosities and to be more shear thickening than plastisols containing less

Page 486
solvating, less polar or polarizable plasticizers of essentially the same molecular size. The
plastisol will gel more quickly on standing and will fuse at a lower temperature (Table 5). An
increase in molecular size, however, tends toward higher viscosity but slower fusion.
Plastisols are used for coating cloth such as tarpaulins, tents, and upholstery [165] and metals
such as house siding and dishwasher liners. Dolls and gloves are plastisol cast, or slush
molded. The blending of plasticizers to control fusion, such as an alkyl benzyl phthalate added
to a dialkyl phthalate, results in a higher gloss on dip-molded articles [167] and improved
blowing of foam [164, 166] and plastisol lamanating adhesives [110] but can cause shorter gel
time than some processors need. Variations on this method involve adding a small amount of
diluent (soluble in the plasticizer but nonsolvating toward the resin), such as hexane, mineral
spirits, or toluene. The diluent permits more precise control of the viscosity and viscosity
stability by decreased solvent power during storage, but it flashes off on heating, so does not
seriously retard fusion. Such mixtures are organosols. When certain plasticizers, frequently of
high viscosity and high molecular weight, are used with modified paste resins or with silica gel,
the plastisol may become a gelatinous mass, reversibly gelled. It may then be used as a
plastigel in producing certain types of floor coverings with chips or other materials suspended in
a clear matrix.
3. Solvent Casting
Solvent casting requires the PVC to be dissolved in a suitable solvent such as cyclohexanone,
tetrahydrofuran, or mixtures of ketones and hydrocarbons such as a 1:2:4 mixture of methyl
isobutyl ketone, methyl ethyl ketone, and xylene. The solution can then be mixed with a
plasticizer that will make further adjustments in solventability as shown by solubility parameter,
polarity, and hydrogen bonding.
In solvents of high solvent power, the PVC molecules will uncoil, be less associated with
themselves, and be more associated with the solvent, all of which will result in a relatively
viscous solution. If solvent ability is lowered but left adequate, as by adding hydrocarbon or a
plasticizer with low solvent power, the polymer molecules tend to coil, becoming more nearly
spherical and less associated with the solvent [103]. This results in lower viscosity (Table 4)
and frequently gives easier solvent casting, film formation, or paint flowout.
In very good solvents such as cyclohexanone, cyclopentanone, and tetrahydrofuran, viscosity is
less sensitive to changes in resin concentration than it is in moderately good solvents such as
nitrobenzene, tetrahydropyran, and morpholine [143].
Copolymers of vinyl chloride are more soluble than the homopolymer. The 85:15 copolymer of
vinyl chloride and vinyl acetate is particularly well suited for use in vinyl lacquers. Properties and
prices of these lacquers have restricted their use to high-quality industrial and marine protective
coatings.
B. Ease of Fusion
The ease of fusion of the resin-plasticizer blend is technically important in determining the best

processing conditions and is also of economic concern (Tables 5, 7, and 8). In a dynamic fusion
test in a torque rheometer (Plasticorder, C. W. Brabender Instruments, Inc.), with a heating
bath temperature of 67C, a DOP/PVC blend required 19.4 min to fuse, whereas a compar-

Page 487

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.

Add single epoxy oxygen to aliphatic monoester.

Add second ester linkage in aliphatic monoester.
Replace chain with aromatic ring.
Add amide linkage in aliphatic diester.
Replace chain with cyclohexane ring.
Add third ester linkage in aliphatic diester.
Add epoxy oxygen to cyclohexane ring.
Add fourth ester linkage to aliphatic triester.
Add etheric oxygen into chain.
Add CH2 group to chain.

-58
-36
-34
-30
-27
-25
-12
+5
+6
+5 to +8

Source: Ref. 79.

able blend with butyl benzyl phthalate required only 3.8 min. When the Plasticorder is heated
during processing, the relative fusion temperatures (which show the onset of fusion) of
plasticizers may be determined. Table 5 gives values for a number of plasticizers. That the
values range from 58 to 133C emphasizes the economic aspect, since less heat and
mechanical energy must be put into fusion of PVC with DOP or TCP than with diisodecyl
adipate, and even less is required with butyl benzyl phthalate. This may be translated into the
power required and therefore cost of processing, as shown by similar studies in plant-type
equipment (Table 8) [16, 178]. Through proper selection of a plasticizer, or plasticizer blends,
optimum fusion times and power requirements can be developed.
Fusion points of plastisols prepared from various plasticizers furnish a basis for ranking the
plastisols according to ease of fusion and can aid in choice of plasticizer for each industrial
need. Depending on the method used to determine fusion, slightly different values will be given.
In Table 5, fusion temperatures are given as the temperature at which the plastisol became
clear (i.e., clear point) [87] and the temperature below which the plastisol would not quite hold
together, or the minimum fluxing temperature [79]. This temperature does not develop full
physical properties. For example, the minimum fluxing temperature for diisodecyl phthalate
(DIDP) is 124C, but as shown in Figure 15, a plastisol of DIDP held at the even higher
temperature of 150C will never develop more than half its potential distensibility. At this
temperature, however, butyl benzyl phthalate will develop essentially maximum tensile
properties in 35 min.
With notable exceptions, a general order of decreasing ease of fusion is phosphates (fastest),
phthalates, isophthalates, terephthalates, adipates, azelates, and sebacates (slowest) [16, 87].
Frissell [79] reported the effects

Page 488
a Conc. (phr) Fusion time (sec) Power peak (hp) Av. hp from 160 sec after fusion Fusion
temperature (C) Time to reach uniform temperature (sec) Uniform temperature (av.) (C) Di-2ethylhexyl phthalate (DOP) 40 44 6.4 5.4 116 210 158 50 71 5.2 4.9 118 265 155 60 92 4.4
4.1 121 281 146 Di-2-ethylhexyl adipate (DOA) 50 106 4.4 4.3 127 375 152 Butyl benzyl
phthalate (BBP) 40 18 8.4 6.4 104 150 159 50 47 6.8 5.4 116 190 154 60 56 5.6 4.6 121 200
150 Tricresyl phosphate (TCP) 40 32 8.2 6.5 110 180 159 50 56 6.6 5.6 116 220 156 60 67
5.6 4.8 121 207 157 Cresyl diphenyl phosphate 40 24 7.6 5.8 113 165 160 50 39 6.8 5.6 113
169 155 60 50 5.9 4.7 118 180 152

Page 489
2-Ethylhexyl diphenyl phosphate

Paraplex G-50
DOP/2-Ethylhexyl diphenyl phosphate

DOP /polyethylene glycol di(2-ethylhexoate) Flexol 4GO

Butyl benzyl phthalate/Flexol 4GO
DOP /tetrahydrofurfuryl oleate (THFO)
Butyl benzyl phthalate/THFO
DOP/HB-40
2-Ethylhexyl diphenyl phosphate/HB-40

40
25 8.0 4.0 99 169 159
50
49 6.6 5.5 116 195 153
60
55 5.7 4.6 113 188 149
50
283 4.2 4.2 137 520 154
60
274 3.4 3.4 127 430 149
20/20 37.5 7.2 5.8 110 180 158
25/25 60 6.2 5.3 116 240 152
30/30 62 5.4 4.6 118 210 150
35/15 65 4.8 4.4 118 295 158
35/15 44 6.0 5.2 110 195 154
35/15 65 4.8 4.2 221 295 157
35/15 43 6.4 5.3 110 195 155
25/25 69 5.0 4.6 116 235 155
25/25 43 6.0 5.1 107 210 154

aTrademarks: Monsanto Co.; HB-40; Paraplex, Rohm Haas Co. Flexol, Union Carbide Corp.
Source: From Ref. 16.

Page 490

FIGURE 15 Development of physical properties of PVC plastisols with fusion time. Maximum
obtainable elongation is the elongation that can be obtained by fusing each plastisol at its
optimum temperature, usually greater than 150C. Formulations: PVC 100, Plasticizer 65,
epoxidized soya oil 3, Ba-Cd stabilizer 2, triphenyl phosphite 0. 5. Plasticizers: 1, butyl benzyl
phthalate; 2, 2-ethylhexyl diphenyl phosphate; 3, DOP + partially hydrogenated terphenyls; 4,
DOP; 5, DIDP; 6, DIDA. (From Ref. 87.)
of various functional groups in the plasticizer molecule on fusion temperature; these are given in
Table 7. Fusion tends to become easier with increasing polarity up to a dielectric constant of
about 8; it becomes more difficult thereafter. When no shear is involved in mixing, the optimum
dielectric constant appears to be higher, coinciding with smaller molecular volumes in common
plasticizers [58, 179].
One value of plastisols is that the whole formulation can be stored as liquid until needed. If the
plasticizer solvates the resin too rapidly, shelf life will be poor. When the plastisol must be
stored in a hot warehouse, a slow-solvating plasticizer such as a linear phthalate or DIDP
should be selected even if higher temperatures and longer times are required for subsequent
fusion. If storage stability is not a factor, faster-fusing plasticizers can be used to advantage.
Figure 16 and Table 9 show variation of plastisol viscosity with shelf life for selected plasticizers
[54]. Blending of plasticizers to get a more suitable balance of properties is common practice.
Adding hydrocarbon secondary plasticizers can lower the plastisol viscosity and lengthen shelf
life effectively without seriously increasing fusion time and temperature.

VII. EFFECT ON PHYSICAL PROPERTIES

A. Mechanical, Thermal, and Electrical Properties
Plasticizers are compounded with PVC primarily to develop the desired physical properties.
Two variables of concern are the kind of plasticizer used and its concentration. The
temperature range in which the final product will be used, and other environmental factors, will
influence the choice and con-

Page 491

FIGURE 16 Viscosity buildup in plastisols as it is influenced by plasticizer choice and storage

temperatures. Plasticizer concentration 80 phr. (From Ref. 54.)
centration. In the usual range of concentration, above 20 phr, adding plasticizer makes the resin
softer and more flexible, reduces the modulus and tensile strength, and gives greater
elongation. The reverse effects, observed at low concentrations, were largely unknown before
1956 [98, 204] (see Section XI).
Rigid PVC had few uses; adding only small amounts of plasticizer left it more rigid, so enough
plasticizer was added to give flexibility and the low, useless range was unexplored. For most
resin-plasticizer combinations at room temperature, this plasticizer threshold concentration must
be passed before the normal plasticizer effects are observed on physical properties. Figures
17 and 18 show the plasticizing action of several monomeric and polymeric plasticizers. The
overall downward slope of tensile strength and modu-

Page 492
b (poise) 25C 40C 50C Plasticizera (80 phr) Initial 1 day 7 day 14 day 28 days Initial 1 day 7
day 14 days 28 days Initial 1 day 7 days 14 days 28 days Dioctyl phthalate (DOP) 36 65 97
116 132 37 84 215 305 426 66 Gel Diisodecyl phthalate (DIDP) 45 75 98 86 80 31 50 80 110
120 50 104 186 230 300 Di-(n-octyl,n-decyl) phthalate 22 32 38 42 46 25 44 70 89 112 105
458 Gel Diisooctyl phthalate (DIOP) 38 90 147 146 158 32 86 204 223 316 122 Gel Santicizer
160 35 50 120 118 164 37 163 Gel Gel Flexol CC-55 32 48 80 110 120 37 70 152 228 250 86
260 700 Gel Santicizer 140 31 45 52 42 56 32 65 80 123 220 285 Gel Santicizer 141 36 66
184 255 450 166 Gel Gel Tricresyl phosphate (TCP) 45 58 89 68 91 60 108 190 266 412 296
Gel Flexol TOF 16 24 30 38 38 16 45 85 100 150 61 226 Gel Dioctyl adipate (DOA) 14 36 46
52 65 19 70 107 118 90 41 192 Gel Diisodecyl adipate (DIDA)c 12 12 22 24 23 12 18 23 26
18 16 28 36 38 31 Santicizer B-16 55 75 172 182 162 64 315 Gel Gel DIDP/S-160 (3:1) 38 52
133 122 204 32 62 178 272 322 46 150 Gel DIDP/S-160 (1:1) 46 60 133 128 217 59 130 375
950 Gel 900 Gel DIDP/S-160 (1:3) 45 65 164 268 250 52 180 980 Gel 162 Gel DIDP/S-141
(3:1) 32 45 56 94 128 40 58 108 122 179 128 240 Gel DIDP/S-141 (1:1) 34 147 178 42
143 295 750 Gel 780 Gel DIDP/S-141 (1:3) 38 43 84 160 218 86 263 Gel Gel DOP/HB-40
(3:1) 40 54 61 58 79 34 85 86 92 146 81 950 Gel DOP/Sovaloid C (3:1) 32 68 57 51
109 112 85 400 >1000 aTrademarks: Monsanto Company; Santicizer, HB-40; Union
Carbide Corporation, Flexol; Mobil Chemical Company, Sovaloid. bMeasured with Brookfield
LVF viscometer, No. 4 Spindle at 12 rpm. Initial viscosity taken 4 hr after preparation. cPhase
separation was noted after 80 days of storage at 25C. Source: Ref. 54.

Page 493

FIGURE 17 Mechanical properties of PVC plasticized with various concentrations of DOP.

(From Ref. 82.)
lus with increased concentration is quite gentle with the low-polarity linear ester DOA. The
slope increases with increased polarity and molecular compactness. In general, then, the
system is much more sensitive to incremental additions of polar, compact, globular shaped
plasticizers than to plasticizers that are more linear and less polar, while the plasticization
threshold is at a lower concentration for the latter materials.
Other physical properties vary with choice and concentration of plasticizer. Taylor and Tobolsky
[196] show the effect of DOP in leveling out the stress relaxation time of PVC. Thermal
properties vary with increasing concentration of plasticizer; near room temperature the specific
heat and thermal conductivity increase while the thermal diffusivity decreases (Figure 19) [191].
Electrical properties, vital to good electrical insulation, and useful for better understanding of
the physical nature of the resin system, have been studied extensively over the last 30 years.
The dissipation factor is comparable to mechanical damping and is sensitive to choice and
concentration of plasticizer [103, 136, 159, 202]. The dielectric constant increases from about
3.0 to 3.6 for unplasticized PVC up to 4.0 to 8.0 on plasticization. Dielectric strength and
volume resistivity decrease. Conductivity depends on polarity and efficiency of the plasticizer,
and increases with plasticizer concentration, as shown in Figure 20. Actual conductance values
of plasticized PVC depend very little on plasticizer purity.
Plasticizers in PVC broaden the temperature range between the glass transition and the flow
temperature by lowering the glass transition tempera-

Page 494

FIGURE 18 Effect of concentrations of various plasticizers on tensile strength and elongation of

PVC. Plasticizers: 1, dicyclohexyl phthalate; 2, phthalate polyester; 3, adipic polyester; 4, TCP;
5, liquid butadiene-acrylonitrile copolymer; 6, DOP; 7, DOA. (From Ref. 82.)
ture more than they lower the flow temperature (Figure 2). They affect the amorphous region,
enlarging the cells in the gel structure and permitting increased flexibility at lower temperatures,
but they leave some cross-linking through residual crystallinity to maintain physical integrity
below the slightly depressed melting point or flow temperature. The amount of lowering of the
glass transition temperature is almost a linear function of plasticizer concentration; different
plasticizers differ in lowering ability (Figures 21 and 46). The less polar, more linear ones are
more efficient in this respect when they are also highly compatible. As they approach borderline
incompatibility, because of low polarity, they lose efficiency.
B. Temperature Effects
Plasticizer choice and concentration are of prime importance in developing a plasticized PVC
that will respond to temperature changes in a suitable way. For many uses it is vital to have
good low-temperature flexibility. Figures 22 to 24 show major portions of modulus versus
temperature curves with varying amounts of different plasticizers. Adding plasticizer shifts the
curves

Page 495

FIGURE 19 Thermal properties of PVC plasticized with butyl benzyl phthalate. (From Ref. 191.)
toward lower temperatures. Immediately after the threshold concentration, incremental addition
of more plasticizer gives a more gentle slope or a more gradual change in modulus with change
in temperature. Further addition of plasticizer eventually results in steeper slope or more rapid
change in stiffness with temperature. With the more polar, more compact plasticizers high in
solvent power [198], such as TCP, the change in slope is small. With more linear, less polar
plasticizers, low in solvent power, the change in slope is substantial. Trioctyl phosphate,
adipates, and sebacates, used to impart low-temperature properties to PVC, greatly extend
the stifflex range, the range between the glass transition and the flow temperature.
As slightly plasticized, semirigid PVC becomes more important, more emphasis is given-to
measurement of physical properties at higher temperatures. Figures 25 and 26 show the effect
of small amounts of plasticizer in increasing, elongation and reducing tensile strength at
temperatures up to 160C.

Page 496

FIGURE 20 Effect of plasticizer content on conductivity of plasticized PVC-VAc (95:5). Samples

contained 7 x 10-5 mol/liter of tetraethylammonium pic-rate as ionic contaminant. (From Ref.
203.)

FIGURE 21 Effect of plasticizer content on low-temperature flex of plasticized PVC.

Plasticizers: 1, dicyclohexyl phthalate; 2, TCP; 3, DOP; 4, 2-ethyl-hexyl diphenyl phosphate; 5,
DOA; 6, DOS. (From Ref. 82.)

Page 497

FIGURE 22 Modulus-temperature curves for PVC plasticized with TCP. (From Ref. 39.)
C. Efficiency
Curves such as those shown in Figures 25 and 26 help comparison of one plasticizer with
another. For spot-check comparisons, the temperature at which a formulation has a modulus of
1.35 105 psi (the flex temperature or Tf) is very close to the glass transition temperature
Tg. The lower the Tf for a given concentration, the more efficient the plasticizer is in this very
important theoretical and practical aspect.
The efficiency of a plasticizer refers to its ability to do the job desired. The greater the amount
of plasticizer required to produce the effect wanted, the lower the efficiency of the plasticizer.
Many attempts have been made to compare one plasticizer with another in efficiency by a onenumber efficiency factor. Frequently, plasticizers are referred to DOP as a standard with
efficiency of 1.00. Results are highly dependent on the test method and end-use property of
interest. For example, if a room temperature stiffness of about 750 psi is desired, inspection of
Figures 22 to 24 shows that this can be achieved with 41% TCP, 40% DOP, or 38% TOF. They
are almost alike at that point. However, a Tf of -39C requires over 50% TCP, 40% DOP, and
less than 30% TOF. Comparing the nearly linear curves for Tf versus concentration with the
sigmoidal curves for modulus or stiffness versus concentration and the sigmoidal curves of
modulus versus temperature shows the impossibility of obtaining an absolute efficiency factor.
The matter is further complicated by the plasticization threshold observed for most physical
properties but not for lowering of Tg. In spite of this, efficiency numbers developed by any
processor for comparing plasticizers for a specific end use are of great value [106,152].

Page 498

FIGURE 23 Modulus-temperature curves for PVC plasticized with DOP. Solid lines are from
Ref. 39; dashed lines are replotted from Ref. 184.

FIGURE 24 Modulus-temperature curves for PVC plasticized with tri-2-ethyl-hexyl phosphate

(TOF). (From Ref. 39.)

Page 499

FIGURE 25 Effect of temperature on tensile strength and elongation of PVC containing small
amounts of dialphanyl (C7-C9 mixed alcohol) phthalate. (From Ref. 154.)

FIGURE 26 Effect of temperature on tensile strength and elongation of PVC plasticized with 30
phr of (1) DOA, (2) dibutyl phthalate, and (3) triisooctyl phosphate. (From Ref. 154.)

Page 500
It has long been recognized as a rule of thumb that viscous plasticizers are less efficient than
less viscous ones [126]. Exceptions promote caution in predicting final properties from this
alone. In general, linear-shaped plasticizer molecules are lower in viscosity and yield more
flexible PVC compounds at low temperature, although not necessarily so at room temperature;
very compact molecules are low in efficiency, not being sufficiently self-plasticized by rotating
side groups [198]. A high proportion of compatibilizing groups in the molecule lowers efficiency,
and high molecular weight leads to low efficiency except when offset by a suitable linear
structure.
D. Blending to Obtain Desired Efficiency
Even with the many plasticizers now available, it is not always possible to find one with the best
possible combination of properties for a given job. By blending two or three, or occasionally
more, the possible choice is extended greatly. In modification of physical properties, a blend will
give effects that will be essentially the weighted average of the effects produced by the
individual plasticizers. To predict efficiency most accurately, the blend should be prepared on a
relative volume basis and not on a weight basis. This averaging effect holds true only for
plasticizers that are miscible in the ratio used and only when each is reasonably compatible with
the resin. In contrast to the nearly linear effect of plasticizer blends on physical properties, the
effect on chemical properties such as heat and light stability is frequently far from linear.

VIII. PERMANENCE PROPERTIES OF PLASTICIZED PVC

A. Volatility
Early plasticizers for PVC, such as dibutyl phthalate, were so high in vapor pressure and
volatility that permanence was poor (Figure 27); shrinkage and stiffening of finished vinyl
articles were common problems [157]. Figure 27 illustrates that if 10% loss, is failure, the life
expectancy of these 4-rail film in hours at 98C is equal to 0.08/vapor pressure of the
plasticizer. Adjustments can be made for other thicknesses and temperatures showing
realistically long service lives of many decades for good plasticizers, since use of highermolecular-weight phthalates relieved most of these problems. Yet plasticizers with even lower
vapor pressure are needed. For permanence, the volatility of plasticizers is so important that it
is usually determined during screening (Table 10). Verhoek and Marshal [200] very early found
that the vapor pressure of tri-m-cresyl phosphate as a plasticizer in PVC was independent of
concentration above 15% (17 phr) and was essentially the same as the vapor pressure for the
pure plasticizer. They concluded there was little attachment between the plasticizer and resin
and that they were detecting volatilization from a surface film of plasticizer. At low concentration
the vapor pressure, even of volatile solvents, becomes very low since the last traces are nearly
impossible to remove by vaporization [38]. Decomposition of the plasticized PVC may be
accompanied by a vapor pressure higher than the vapor pressure of the plasticizer [200].

Page 501

FIGURE 27 Relation between vapor pressure of plasticizers at 98C and loss from 0.004-in.thick PVC films at 98C. Films were plasticized with different plasticizers varying from 40 to 82
phr. The one point that is out of line is for polyethylene glycol di(2-ethylhexoate), which should
be very easily oxidized. (From Ref. 157.)
B. Exudation Under Pressure
A plasticizer that is compatible with PVC under ordinary conditions may exude under slight
pressure. Polyvinyl chloride plasticized with 80 phr DOP, normally considered compatible in all
proportions, lost about 30% of its plasticizer when subjected to 0.1 kg/cm2 of pressure in an
experiment especially designed to accept and remove the plasticizer (Figure 28). Although this
pressure

FIGURE 28 Plasticizer exudation under pressure. (From Ref. 78.)

Page 502
b Volatility plasticizer Low-temp. flex. Water extraction Kerosene extraction Hardness, Shore A
Plasticizera loss (%) Tf(C) (% loss) (% loss) Adipic acid derivatives Diisobutyl adipate 63 -55
3.54 35 60 Di-2-ethylhexyl adipate (DOA) 13.9 67.8 0.10 75 65 Heptyl nonyl adipate Santicizer
97 8.4 -71 0.15 74 66 Octyl decyl adipate (n-octyl, n-decyl; DNODA-6,10) 8.1 -66 0.08 72 68
Diisodecyl adipate (DIDA) 3.4 -63 0.07 >70 74 Dibutoxyethyl adipate 27.3 -67 1.30 32 71
Di(methyleyclohexyl) adipate 15.8 -27.6 0.07 4.5 64 Bis(2,2,4-trimethyl-1, 3-pentane-diolmono
isobutyrate) adipatec 3.8c -11c (6.95c,d) 10.7c 84c Azelaic acid derivatives Di-n-hexyl azelate
11.0 -71.9 0.08 74 64 Di-2-ethylbutyl azelate 25.8 -65 0.05 - 72 Di-2-ethylhexyl azelate (DOZ)
5.7 -67.2 0.06 73 70 Benzoic acid derivatives Ethylene glycol dibenzoate 11.5 -21.9 0.06 5.0
75 Diethylene glycol dibenzoate 5.5 -20 0.75 3.4 69 Triethylene glycol dibenzoate 1.2 -25.6
0.03 5.5 80 Polyethylene glycol (200) dibenzoate 23.8 -12.9 0.04 8.6 80

Page 503
Polythylene glycol (600) dibenzoate
Dipropylene glycol dibenzoate
2,2,4-Trimethyl-1, 3-pentanediol monoisobutyrate benzoate
Citric acid derivatives
Tri-n-butyl citrate
Acetyl tri-n-butyl citrate
Acetyl tri(n-octyl,n-decyl) citrate
Epoxy derivatives
Epoxidized soybean oil
2-Ethylhexyl epoxy tallate
Bisphenol A diglycidyl ether
Ethers
o-Nitrodiphenyl ether
Formal
Dibutoxyethoxyethyl formal
Glycol derivatives
Diethylene glycol dipelargonate
Triethylene glycol dipelargonate
Triethylene glycol di-2-ethylbutyrate
Triethylene glycol di-2-ethylhexoate
Polyethylene glycol (200) di-2-ethyl-hexoate
Glycolates
Methyl phthalyl ethyl glycolate
Ethyl phthalyl ethyl glycolate
Butyl phthalyl butyl glycolate

Immediate exudation
7.9 -12
0.45 2.9
24.8 -23.6 0.48 4.0

71
71

16.5 -38.8
17.8 -36
0.2 -53.8

0.43 8
0.09 0.13 90

66
73
80

0.6
1.5
-

-29
-58
+12.5

0.07 3.1
0.13 81
0.25 -

70
64
97

44

-18

3.8

72

13

Decomposed on molding
10.1
10
36.9
7.5
14.9

-65
-64
-61
-57.3
-54.9

26.9 -20.5
28.7 -23.4
12.0 -35

0.46
1.1
1.22
1.35
2.13

60.1
45
26
45
26

4.9 2.7
2.6 2.1
0.12 3.8

67
74
65
71
74
67
66
68

Page 504
b Volatility plasticizer Low-temp. flex. Water extraction Kerosene extraction Hardness, Shore A
Plasticizera loss (%) Tf(C) (% loss) (% loss) Petroleum derivatives Hydrogenated terphenyls(
(HB-40) 1:1 with DOP 27.6 -30.4 0.35 57 73 1: 2 with DOP 21.3 -30.5 0.16 44 70 Polyalkyl
naphthalenes, 1:2 with DOP 19.5 -31.7 0.39 65 70 Highly refined, white paraffin oil, 1:9 with
Dope 5.4 -41.4 0.4 51 76 Isobutyric acid derivatives 2,2,4-Trimethyl-1,3-pentanediol
diisobutyrate 23.7 33.1 2.83 5.2 76 Isophthalic acid derivatives Di-2-ethylhexyl isophthalate 6.5
-41.7 0.03 66 74 Isosebacic acid derivatives Di-n-butyl isosebacate 48.9 -65.5 0.51 34 67 Di-2ethylhexyl isosebacate 5.1 -67.4 0.05 82 73 Oleic acid derivatives Butyl oleate No sheet on mill
1:1 with DOP 17.9 -66.8 0.18 80.4 76

Page 505
Tetrahydrofurfuryl oleate
1:1 with DOP
Paraffin derivatives
Chlorinated paraffin, 42% chlorine
1:1 with DOP
1:2 with DOP
Chlorinated paraffin, 52% chlorine
1:3 with DOP
Pentaerythritol derivative
Pentaerythritol, fatty acid ester
Phosphoric acid derivatives
Tributyl phosphate
Tri-2-ethylhexyl phosphate (TOF)
Tributoxyethyl phosphate
Cresyl diphenyl phosphate (CDP)
Tricresyl phosphate (TCP)
Tridimethylphenyl phosphate
Triphenyl phosphate, 2:3 with 2-ethylhexyl diphenyl phosphate
Isopropylphenyl diphenyl phosphate (approx.)
t-Butylphenyl diphenyl phosphatee
2-Ethylhexyl diphenyl phosphate
Isodecyl diphenyl phosphate
Tris(2-chloroethyl) phosphate
1: 3 with DOP
Phthalic acid derivatives
Dibutyl phthalate
Diisohexyl phthalate

6.5 -68
6.5 -55

0.18 0.24 62

74
66

2.8
3.9
6.1
4.5

0.03
0.01
0.01
0.02

82
79
79
75

-28.8
-34.0
-14.2
-34.6

22
31
1.7
37.5

1.8 -33.7 0.10 28

74

58 -56 2.0 11.0 -70.0 0.21 74

6.2 -55.8 2.39 17.2
3.6 -18 0.04 2.1
1.0 -14 0.02 2.4
1.7 -4
0.03 6.2
9.2 -30.5 0.10 4.7
3.1 -14 0.7 1.8
3.1 -9
0.07 1.3
7.4 -39 0.06 7.3
2.8 -34 0.06 6.4
Immediate exudation
20.2 -37.6 7.3 6.7

71
75
70
72
72
67
71
75
80
68
69

45.4 -40.4 0.25 9.1

14.8 -39.8 0.01 6.9

62
67

69

Page 506
b Volatility plasticizer Low-temp. flex. Water extraction Kerosene extraction Hardness, Shore A
Plasticizera loss (%) Tf(C) (% loss) (% loss) Phthalic acid derivatives (Contd)
Dimethylisobutylcarbinyl phthalate 22.8 -17.7 0.15 20 73 Butyl octyl phthalate 9.5 -41 0.04 27
68 Butyl isodecyl phthalate 11.5 -40 0.08 45 69 Di-n-octyl phthalate 2.5 -46 0.03 46.6 70 Di-2ethylhexyl phthalate (DOP) 4.5 -39 0.01 44 69 Diisooctyl phthalate (DIOP) 4.3 -37 0.03 42 71
Dicapryl phthalate 4.6 -37 0.03 54 71 Di(heptyl,nonyl) phthalatee 2.5 -45 0.02 44 73
Di(heptyl,nonyl,undecyl) phthalate 1.7 -48 0.02 70 71 Diisononyl phthalate 2.1 -42.8 0.07 77 73
N-Octyl n-decyl phthalate (DNODP-6,10) 2.1 -50 0.0 64 71 Diisodecyl phthalate (DIDP) 1.8 -37
0.03 74 71 Diundecyl phthalate (DUP) 0.9 -52 0.05 >70 77 Ditridecyl phthalate (DTDP) 0.8
-37.4 0.07 88 83 Diallyl phthalate 51.2 -41.7 1.39 18.4 61 Dihydroabietyl phthalate 0.4 +40 - 100 Butyl cyclohexyl phthalate 18.8 -17.1 0.09 2.7 66 Butyl benzyl phthalate Santicizer 160 7.7
-24 0.07 3.4 68 Alkyl benzyl phthalate Santicizer 261 2.1 -26 0.07 3.8 71 Dicyclohexyl phthalate
(DCHP)f 3.3 +16.7 0.03 0.3 99 7.1 -21 0.40 3.5 78

Page 507
Diphenyl phthalate (DPP)f
Alkyl aryl modified phthalate Santicizer 213 (case hardening)
Dimethoxyethyl phthalate
Diethoxyethoxyethyl phthalate
Dibutoxyethyl phthalate
Polyesters
Adipic acid polyester, MW 6000
Paraplex G-40
Adipic acid polyester, MW 2200
Paraplex G-50
Adipic acid polyester, MW 3300
Paraplex G-54
Azelaic acid polyester, MW 2200
Plastolein 9750
Sebacic acid polyester
Paraplex G-25
Ricinoleic acid derivatives
Methyl ricinoleate
Methyl. acetyl ricinoleate
n-Butyl acetyl ricinoleate
Methoxyethyl acetyl ricinoleate
Sebacic acid derivatives (see also polyesters)
Di-2-ethylhexyl sebacate (DOS)
Stearic acid derivatives
n-Butyl stearate
1: 1 with DOP
Butyl acetoxy stearate

2.0
18
16.7
5.0
5.8

+14.2 0.05
-21 0.56
-28.5 1.72
-32 8.8
-40 0.45

0.1 97
6.3 76
6.5 73
5.4 66
9.6 61

1.6,
6 days
3.1,
6 days
2.2,
6 days
3.2,
6 days
1.2,
6 days

-9.6

5.1,d
4 days
13.1,d
4 days
10.0,d
4 days
7.1,d
4 days
0.4,d
4 days

84

74

75

78

77

12.1
16
8.1
4.7

-54
-61
-64.8
-54

0.45
0.56
0.53
0.32

58
77
-

73
68
75
73

4.2

-69.1 0.02

70 73

No sheet on mill
20.1 -38.7 0.29
5.4
-52 0.55

81 78
80 79

-19.4
-17.4
-18.7
-16.0

Page 508
b Volatility plasticizer Low-temp. flex. Water extraction Kerosene extraction Hardness, Shore A
Plasticizera loss (%) Tf(C) (% loss) (% loss) Stearic acid derivatives (Contd) Methoxye.thy!
acetoxy stearate 4.4 -46.7 0.02 64.7 74 Methyl pentachlorostearate 2.29 -16.5 0.02 - 84
Styrene derivatives Poly-a-methyl styrene (MW-325) No sheet on mill 1:1 with DOP 1.38 -17.0
0.49 83.7 86 Sulfonamides and sulfonates n-ethyl. o,p-toluene sulfonamide 33.2 -14.4 0.14 5.8
86 1:9 with di(heptyl,nonyl,undecyl) phthalate 10 -40 2.3 40 69 n-cyclohexyl p-toluene
sulfonamide 6.2 +18.8 0.60 0.3 100 Alkyl sulfonic acid ester of phenol and cresol 5.3 -31.4
0.03 4.8 72

Page 509
Tartaric acid derivatives
Dibutyl tartrate
-32.3
14.0 19.4 79
Terephthalates
Di-2-ethylhexyl terephthalatee
1.9 -41.8
0.09 70.8 76
Trimellitates
Tri-2-ethylhexyl trimellitate (TOTM)
0.5 -32.7
0.0
75
72
Tri(heptyl,nonyl)trimellitate
0.4 -39
0.07 >70
76
n-Octyl, n-decyl trimellitate
0.8 -45.6
0.05 85
82
Triisodecyl trimellitate
0.7 -37.4
0.02 87
84
aTrademarks: HB-40 and Santicizer, Monsanto Co.; Paraplex, Rohm & Haas Co. , C. P. Hall
Co. Plastolein, Emery Industries, Inc.
bperformance data are for 40% (67 phr) concentration in Opalon 650 (Monsanto Co. (PVC)],
except as noted. Volatility, 24 hr at 87C over activated carbon; low-temp. flex., Clash and Berg
test; water extraction, 24 hr at 50C; kerosene extraction, 24 hr at 23C.
cplasticizer concentration was 50 phr.
dTest was run in 1% Ivory soap solution instead of pure water.
eResin used was Geon 102 EP, B. F. Goodrich.
fCrystalline solid. Its performance as the sole plasticizer in PVC will vary with its degree of
crystallinity in the resin.

Page 510
seems very low, Salmon and Van Amerongen also found that rubber gels were very sensitive
to minute energetic changes, moderate distortion of the rubber resulting in exclusion or
inclusion of 20 to 50% of the swelling solvent [172]. A similar situation apparently exists in PVC,
and increased pressure does not cause further plasticizer loss, perhaps signifying that a part of
the plasticizer strongly solvates the resin while the rest is held in with plasticizer-plasticizer
bonds. Exudation under pressure may occur in finished goods, when they have not been
properly compounded for the given end use, as when PVC sheet is rolled tight or vinyl tile is
stacked or vinyl-covered wire is wrapped around an insulator. Laboratory testing for this is
frequently done by bending a plasticized vinyl sheet into a tight loop and noting any exudation
inside the loop. Usually, when excessive pressure occurs, the plasticizer simply migrates from
high-pressure to low-pressure areas. When pressure is released, the plasticizer readjusts itself.
In the unusual occasion when it is exuded, it is usually reabsorbed in time. Frequently, addition
of a small amount of a more compatible plasticizer can eliminate any such problem. The
presence of a filler can also alleviate it.
C. Humidity Compatibility: Environmental Compatibility
Some plasticizers that are compatible under normal conditions exude in an environment where
the humidity is high, particularly where the temperature and humidity are both high. It may be
observed as stickiness and increased dirt pickup on floor coverings or automobile crash pads in
the tropics, or clamminess and premature stiffening of a shower curtain. Exudation is not
usually a result of hydrolysis of plasticizer but can be accounted for by formation of a waterplasticizer blend that is not compatible, or by water solvating some of the centers of force on
the PVC, thus excluding plasticizer.
The effect is usually more pronounced with a polymeric plasticizer, perhaps because once the
large molecule reaches the surface, its massive size prevents it from returning into the PVC.
The sensitivity of polymerics may also result from the narrower limits of compatibility (shown by
solubility parameter and dielectric constant) with increased ease of going beyond these limits
by absorption of traces of water. This is coupled with the need for increased functionality, a
higher concentration of polar groups, in a high-molecular-weight plasticizer to achieve
compatibility in the first place. Increased functionality increases water sensitivity.
D. Extraction Resistance
Because the plasticizer is not held permanently to the resin, it may be partly extracted by oils,
waxes, water, soapy water, and other agents [163]. Resistance to extraction by any
environmental agent can be varied over a wide range by plasticizer choice but not without
accompanying changes in other properties. For example, it is not yet possible to get very good
low-temperature flex and excellent resistance to extraction by kerosene at the same time with
any one plasticizer since the linear, aliphatic plasticizers that give good low-temperature flex are
low in polarity and quite soluble in hydrocarbons. High molecular weight is conducive to good
extraction resistance but frequently gives poor low-temperature properties.
In theory, extraction occurs when the extractant diffuses into the PVC, dissolves the plasticizer,
and the two diffuse out together. This is the major

Page 511
mechanism for extraction with hexane, kerosene, and many other solvents. As the molecular
size of the extractant is increased, penetration of the PVC becomes more difficult until at quite
high molecular weight the extraction is more nearly migration (see the next section) as the
plasticizer molecules diffuse out alone to become dissolved in the so-called extractant. Note in
Table 10 the difference in performance between, for example, di(heptyl, non-yl, undecyl)
phthalate, DHNUP, and butyl benzyl phthalate (BBP) in kerosene extraction, 70% versus 3.4%,
respectively, in 24 hr at 23C. When the extractant is the high-molecular-weight ASTM No. 2 oil
(Figure 29), the amounts extracted from 67-phr sheets in 96 hr are 9% and 2.5% of the original
plasticizer content, respectively.
Figure 29 shows the effect of different plasticizer concentrations. Twenty-five phr, where
extraction began in this test, is just past the antiplasticization range, and the gel structure is
very tight. Most of the plasticizer molecules are intimately associated with PVC molecules. At
higher concentration the cells of the gel structure are larger, holding more plasticizer but more
loosely, so that extraction is faster and more extensive.
To achieve a reasonable degree of kerosene extraction with good low-temperature
performance, blends of a monomeric and a polymeric plasticizer are used. Such blends give
extraction results that do not lie on a straight line between the high extraction of, for example,
di(heptyl, nonyl, undecyl) phthalate (70%) and the low extraction of a polyester (1%) but on a
sigmoidal curve, since low concentrations of the polymeric are relatively ineffective.

FIGURE 29 Extraction of plasticized PVC by ASTM No. 2 oil as a function of plasticizer

concentration in PVC. Percent of total plasticizer extracted in 96 hr at 23C; 40-mil sheets.
(From Ref. 30.)

Page 512

FIGURE 30 Kerosene extraction of plasticized PVC reduced by polar plasticizers. (From Ref.
181.)
Blends with high-polarity, extraction-resistant monomerics are more efficient. Results lie near
the hyperbola as in Figure 30, apparently because of greater resistance to penetration of the
surface caused by polar molecules.
Screening tests usually include extraction by both water or soapy water (polar) and kerosene
or hexane (nonpolar). Water absorption into the plasticized PVC is also measured. Early
thinking that water extraction should parallel the solubility of the plasticizer in water proved
wrong, as static tests showed much more of certain plasticizers extracted than would dissolve
in the water; plasticizer would even float on top or settle to the bottom as a separate phase.
Any near mismatch in polarity or hydrogen-bonding ability of the plasticizer and PVC is
conducive to water extraction with, first, a swelling of the plasticized PVC by water, then
breaking of plasticizer-resin bonds, and establishment of plasticizer-water bonds or PVC-water
bonds with exclusion of plasticizer.
Swelling in water can be detrimental, for example, in vinyl floor tile, where buckling may occur
after the tile is laid. Water absorbed into some plasticized PVC may cause blushing or
whitening, which is disadvantageous in greenhouse film or raincoats. Extraction results in
shrinkage and stiffening. All of these effects can be minimized by proper choice of plasticizer
and stabilizer system. Extraction by nonpolar solvents has so far foiled the goal of a completely
satisfactory dry-cleanable vinyl. Uses around gasoline, in food wrap and in hand-held electrical
insulators such as electric shaver cords require proper choice of plasticizer.

Page 513
E. Plasticizer Migration to Other Materials
If plasticized PVC is placed in contact with other polymeric materials, the plasticizer may
migrate from the PVC to the other polymer if the resistance at the interface is not too high and
if the plasticizer is compatible with the second material. Plasticizers such as adipates and some
phthalates may migrate from vinyl upholstery, handbags, or coats to nitrocellulose or acrylic
lacquers on furniture or automobiles. This can soften, or in severe cases, destroy the surface
coating [61,162,163]. Migration to rubber causes swelling, with different kinds of elastomers
affected by different plasticizers.
Migration occurs molecule by molecule just as does diffusion and evaporation. Molecular size
and shape are important in the rate of diffusion. Compatibility with the PVC, which it is leaving,
and compatibility with and efficiency of plasticization of any polymer to which it is going are
important considerations. Figure 31 and Table 11 point up several of these factors well because
of the complex system involved. PVC plasticized with 67-phr plasticizer was covered with a
leather cover about 40 mils thick which had been coated with an acrylic latex and top-coated
with a PVC lacquer. Plasticizers migrated from the PVC through the leather and the acrylic coat
to soften the PVC top coat, causing it to wrinkle and sluff away from the acrylic base coat. The
figure and table show that a minimum molecular weight between 419 and 475 is necessary for
a phthalate to achieve little or no marring in the 1-week test. Volatility must be low, no more
than 1% per day in routine screening tests at 87C. A dielectric constant below about 4.5,
signifying difficulty of dissolving in the PVC top coat, helped prevent marring by higher
phthalates and even by a low-molecular-weight, easily diffusing adipate. The changes between
16 hr and 7 days show that caution is needed in the one-point-in-time testing that is commonly
done when kinetics controls the results. Low-molecular-weight contaminants are suspect in the
polyester and in didodecyl phthalate in Figure 31.
Migration to adhesives can destroy the adhesive bonds; for example, DOP migrates readily
from floor tile to asphalt adhesive, weakening the forces of adhesion and causing the tile to
come up. Butyl benzyl phthalate is more resistant to migration to asphalt, so does not present
the problem. Migration to rigid polymers may embrittle them and cause stress cracking.
Polystyrene refrigerator door liners with plasticized PVC gaskets are thus subject to cracking
unless a nonmigrating, usually polymeric, plasticizer is used. Plasticizers may migrate into foods
or beverages. Even when this involves no more than ultramicro quantities, it may affect odor
and taste.
Although the primary concern is usually the change occurring in the material to which
plasticizers migrate, the service life (permanence) of the PVC article is controlled by this also. If
migration is extensive, it may involve loss of some portion of the plasticizer and resultant
stiffening of the PVC.
F. Migration of Contaminants from Other Materials
Migration of trace amounts of other plasticizers from surface coatings or adhesives usually
makes no noticeable change in plasticized PVC. Problems arise when the material migrating
into the plasticizer of the PVC composition has some obvious, undesirable property. Staining is

such [178]. When plasticized PVC is used in shoe innersoles, tanning agents from the adjacent
leather may migrate into the PVC. The more mobile the plasticizer, the more severe the
problem [101]. High humidity frequently promotes staining by further swelling of the PVC,
reduced resin-plasticizer attraction, and increased mobil-

Page 514

FIGURE 31 Marring of PVC lacquer top coat on leather by plasticizers of different molecular
structure from plasticized PVC substrate through leather to top coat at 101C. Identification of
plasticizers in Table 11. Mar ratings: 1, no significant change, very slight distortion by knife while
hot but no scuffing; 2, slight distortion by knife, hot, but no scuffing; 3, slight scuffing, hot, by
knife with difficulty, but no wrinkling of surface by swelling; 4, easily skinned by knife, hot, but
no wrinkling; 5, very slight wrinkling of surface; 6, slight wrinkling of surface, skinned with
difficulty cold; 7, slight-moderate wrinkling, skinned moderately cold; 8, moderate wrinkling,
skinned easily cold; 9, severe wrinkling; 10, very severe wrinkling. (From Ref. 181.)
ity of the plasticizer. Traffic staining poses a severe problem to light-colored floor coverings.
Staining agents include asphalt tracked in from the street, leather tannins, dyes from adjacent
rugs, and rubber dust from automobile tires. The severity of the stain increases with increased
plasticizer concentration. Staining is usually at a minimum when plasticizer and PVC most
closely match in solubility parameter [153]. The plasticizer tends to exclude other materials in
these cases. Resistance to staining by all types of stainants cannot be hoped for in any one
plasticizer, but exceptionally good results are achieved, if other properties are acceptable, by
butyl benzyl phthalate and 2,2,4-trimethyl-1,3-pentanediol isobutyrate benzoate. Odors may
migrate into plasticized PVC and linger there. Absorption and retention of odor will vary with the
plasticizer used. If staining or odor absorption are severe enough to necessitate replacement of
a vinyl floor, discarding of baby

Page 515
a Volatility (%loss,b day at 87C) Butyl benzyl phthalate 312.4 6.41 10.0 Dihexyl phthalate
334 5.69 11.3 Diheptyl phthalate 362 4.9 7.9 t-Butylphenyl diphenyl phosphate 368 7.33 3.3
Alkyl benzyl phthalate 368 5.70 2.0 Heptyl nonyl adipate 370 4.15 8.0 Isodecyl diphenyl
phosphate 390 6.67 3.2 Di-2-ethylhexyl phthalate 390.6 5.19 4.0
Di(heptyl,nonyl,undecyl)phthalate 414 4.76 2.7 Diisononyl phthalate 418.6 4.67 1.7 Dinonyl
phthalate 418.6 4.67 1.1 Diundecyl phthalate 475 4.43 0.7 Didodecyl phthalate 503 4.03 1.0
Tri(heptyl,nonyl)trimellitate 540 4.81 0.9 Triisononyl trimellitate 580 4.39 0.5 Polyadipate
polyester 2200 6.03 Very slight a at 25C, 10 kHz. bplasticizer loss from PVC plasticized with
67 phr of the given plasticizer.
pants, or reupholstering of furniture, the permanence of the plasticized PVC has been
adversely affected even if its mechanical properties are still good.
G. Controlled-Release Agents
Although loss of plasticizer by volatility, exudation, bloom, extraction, or migration is not
normally desired, all of these effects have become commercially important as plasticizer
techniques have been employed in formulating pesticides and medicines. Many biologically
active materials are compounded into PVC plastisols or compositions designed for molding or
extrusion. As these slowly evaporate, exude as a liquid, or bloom as a crystalline solid from the
PVC composition, the effective life of the active material is greatly extended. Use of many
persistent pesticides has been discontinued because of broad-spectrum toxicity and
environmental persistance. Newer pesticides designed with specific effect targets and less
persistance can be effectively used in these controlled-release compositions. The active
ingredient is released continuously at a relatively constant rate for a predetermined time so that
the target is primarily attacked [148].
Table 12 shows formulations for two flea and tick collars for dogs and cats. In both cases, one
dry blended and extruded and one plastisol cast, the primary plasticizer was di-2-ethylhexyl
phthalate. The solid carbamate insecticide Sevin melts and dissolves in the PVC compound
during extrusion and stays in solution. But when it diffuses to the surface, as it can readily do in
this soft composition containing 70 phr of DOP, it crystallizes as a fine

Page 516
a extrusion Plastisolb casting PVC, general purpose 100 PVC, dispersion resin 67.4 PVC,
extender resin 32.6 DOP 70 30.8 Epoxidized soya oil 9.5 Epoxidized octyl tallate 4.5 Ca/Zn
stearates (50:50) 1.8 Dimethyl 1,2-dibromo-2,2-dichloroethyl phosphatec 37.5 N-methyl 2isopropoxyphenylearbamated 20 N-methyl 1-naphthylcarbamatee 8.5
Bromodichloroacetaldehydef 4.3 Amorphous silica 1.0 Stearic acid 0.5 aFrom Ref. 91. bFrom
Ref. 89. cVolatile insecticide, naled. dEssentially nonvolatile insecticide, Sendran. eEssentially
nonvolatile insecticide, Sevin. fporosity control agent.
white powder. Crystalline solids that develop this way on the surface, or sometimes under the
surface, of plasticized PVC are designated bloom as opposed to liquid exudation or
sweatout. (Excessive bloom of the biocide copper 8-quinolinate from plasticized PVC was a
problem in early technology, eventually controlled by the compatibilizing action of small amounts
of sulfonamide plasticizers [53a].) The crystalline powder is brushed off and distributed over the
body of the dog or cat by the normal animal motions and is replaced in time as more
carbamate blooms to the surface of the collar.
The more complex formula in Table 12 contain a similar carbamate insecticide, Sendran. With
only 30.8 phr of DOP, the composition would be fairly hard, and migration would be slow except
that 37.5 phr of a low-molecular-weight phosphate insecticide, naled (dimethyl 1,2-dibromo-2,2dichloroethyl phosphate), is also present. The formula also contains a halogenated acetaldehyde that vaporizes during fusion of the plastisol to produce microcratering on the surface
and some porosity to speed evaporation of the naled. Figure 32 shows the loss of naled over
20 weeks in curve 1 when naled is the only one of the three special additives used. A second
curve shows the effect of much greater surface area produced by the aldehyde. The fact that
curve 3 shows even faster loss of naled indicates that the carbamate which dissolves in the
clear PVC composition is itself a plasticizer giving a softer, more open gel structure through
which both the phosphate can migrate and evaporate and the carbamate can migrate and
bloom if the concentration is high enough. Without this technique of controlled release, naled
would be too volatile to last long enough to be effective. Many other pesticides are too unstable
in the environment to last long but may be effectively administered this way; dimethyl
dichlorovinyl phosphate is one such [73].

Page 517

FIGURE 32 Controlled release of insecticides from flea and tick collars. (From Ref. 89.)
Biologically active compounds, insecticides, and so on, that bloom from a polymer matrix such
as an insecticidal lacquer usually form much smaller crystals than does the same compound
when it crystallizes from a liquid solvent. Such tiny crystals are more effective than larger ones;
they more readily adhere to the insect foot or leg [156]. The three curves in Figure 32 show
that the rate of loss of naled is controlled by the rate of diffusion from within the PVC to the
surface. It is wiped off, or it evaporates as fast as it can reach the surface. The rate of loss
decreases with time, but the loss is linear if plotted against the square root of time according to
the equation [215]
Mt = kt0.5
(11)
When Mt is the amount of active material lost in a given time, t is time and k is a constant. After
about 60% of an active material has been lost, the rate of release drops exponentially
according to the equation [215]

In some special cases the loss is controlled by the area of the surface. Either the active agent
can diffuse to the surface more rapidly than it can evaporate or can be rubbed off, or the active
agent in the PVC is replenished from a reservoir (as in a capsule) so that its concentration in
the polymer does not decrease. In these cases loss is linear with time according to [215]
Mt = kt

(13)

Page 518
The rate of diffusion is a function of the molecular size and shape of the diffusing molecule. The
release rate is proportional to the square root of the diffusion coefficient. When an excess of an
active agent is available, the release rate is proportional to the square root of the solubility in
the polymer, so is related to solubility parameters of the solute and polymer [215]. The solute
may be an insecticide, herbicide, fungicide, bactericide, insect pheromone, medicinal drug,
fragrance, and the like [207]. The active compound may be an insecticide in PVC wire and
cable insulation to protect it against ants and termites [108]. It may be an anthelmintic
compound in PVC fed to poultry and livestock to control parasitic worms safely [137]. It may be
a fungicide, insecticide or acaricide incorporated into a latex sprayed on plants for long term,
low level usage [76].
H. Corrosion Resistance and Permeability
The excellent corrosion resistance of rigid PVC was one of its most interesting properties to
early investigators. Plasticized polyvinyl chloride maintenance paints have withstood weather,
seawater, and chemical plant atmospheres for up to 20 years with little degradation of the paint
film and no corrosion of the steel or concrete substrates [66].
As plasticizer content is increased in a film, water permeability increases. Table 13 [128] shows
the increase caused by several plasticizers from the very low 0.5 x 10-8 g/hrcm2(mmHg/cm)
for unplasticized PVC to values 6 to 24 times higher for different plasticizers at 12 mol % (about
85 phr for DOP). This also shows that at the same molar concentration of plasticizer, those
plasticizers that are more efficient in lowering Tg give the higher per8 (g/hrcm2mmHg/cm) for a content in mole percent of plasticizer of: Plasticizer 0 4 6 8 10 12
Tricresyl phosphate 0.50 0.55 0.65 0.92 1.74 2.96 Diethyl phthalate 0.50 0.58 0.86 1.18 2.00
3.38 Dibutyl phthalate 0.50 0.60 0.94 1.33 2.46 4.02 Dioctyl phthalatea 0.50 0.64 1.16 1.98
3.05 4.97 Dibutyl adipate 0.50 0.96 1.72 2.67 4.08 5.89 Dioctyl adipate 0.50 1.45 3.00 4.62
6.81 10.95 Dibutyl sebacate 0.50 1.09 2.13 3.30 5.00 8.64 Dioctyl sebacate 0.50 1.64 3.20
5.32 8.03 12.05 aFor comparison of mole percent, DOP concentrations in phr are: 0 26 40 54
69 86 Source: Ref. 128.

Page 519
meability. Di-2-ethylhexyl phthalate and especially TCP, with high aromaticity, give low
permeability.
If all other factors were equal, lower permeability would signify improved corrosion resistance.
In general, this is true, but crossovers do occur. Tests on plasticized PVC free film exposed to
12 environments, ranging in corrosiveness from distilled water through solutions of various
acids, bases, and salts to ferric chloride (most corrosive), showed that the phthalates were
least attacked; next in order were phosphates, adipates, and polymerics. Secondaries were at
times beneficial [88]. Degradation of the plasticized PVC itself, rather than of a metallic
substrate, was observed as either stiffening or exudation. The exact mechanism of penetration
and deterioration of film is not known, but Zitnansky and Lehka [217], in a study with radioactive
sulfuric acid solutions, showed that penetration can occur in at least four steps: an induction
period as the solution attacks the surface, then a rapid penetration as the corrosive medium fills
the fine pores, followed by a slow diffusion, and ending with another unexplained rapid
diffusion.
I. Heat Stability Effects
1. Effect of Plasticizer on PVC Stability
Early workers discovered that incorporating plasticizers into PVC tended to inhibit thermal
degradation during processing and offset darkening and stiffening of the plastic [189]. Several
factors contribute to this. Increased lubricity, decreased hot-melt viscosity, separation of the
PVC polymer chains from each other are physical factors [193]. Stepek et al. [192] observed
that the activation energy of dehydrohalogenation was unchanged when most plasticizers were
added, and reasoned that the plasticizers improved PVC heat stability by dilution. However,
they did find two adipic acid polyesters that appeared to inhibit by chemical means. In all their
work, they merely stirred resin and plasticizer together, let the mixture stand, and then heated it
to eliminate HCl. They probably had residual unplasticized resin in each test. A later study [179]
showed that PVC-DOP blended as above would liberate HCI earlier than a similar blend milled
to complete plasticization at 165C and then heated as above.
a. Effects of Specific Plasticizer Types. Plasticizers containing the epoxy (oxirane) group

were among the earliest stabilizers used in PVC and related resins [140]. They were added to
react with the liberated HCl and retard its catalytic effect on degradation. They were not
commercially important until the early 1950s, when synergism was discovered between them
and the cadmium and zinc stabilizers (Figure 33) [46,52,53,211]. Modern practice frequently
requires 2 to 10 phr of an epoxy stabilizing plasticizer in flexible vinyl (about 4 to 25% of the
plasticizer). In rigid vinyl, 2 to 3 phr can be used without serious modification of physical
properties. Epoxy plasticizers most commonly used are the epoxidized vegetable oils, although
simple esters such as epoxidized octyl oleate improve low-temperature flex, and epoxy resins
such as bisphenol A diglycidyl ether improve dirt resistance in addition to heat stab-

Page 520

FIGURE 33 Development of the synergistic stabilizing system involving phosphite and epoxy
plasticizer in stabilizer proportions. (From Ref. 53.)
ility. Early problems of epoxy-plasticizer compatibility have been largely overcome by careful
control during epoxidation to avoid producing incompatible ethers, alcohols, and acids [74], and
to avoid carbon-carbon double bonds, which led to oxidation aggravated by the presence of
epoxy groups [178]. Limiting the epoxy plasticizer concentration to low levels helped.
Organic phosphites, evaluated as plasticizers years ago but rarely considered as such now,
have a marked plasticizing action even at the low concentrations used [133]. For improved heat
stability and film clarity, they have been incorporated into most flexible PVC since about 1951
[127,130]. Triphenyl phosphite was the most common one in early packaged stabilizer systems
partly because of the reasonably good compatibility with PVC afforded by the three benzene
rings. The polarity of the phosphite group is too low to aid compatibility significantly. Other
phosphites have been developed with improved stabilizer activity and hydrolysis resistance.
Their value in maintaining good color is symbolized in Figure 33.
Other plasticizers can aggravate thermal degradation of the plasticized PVC; phosphates,
sulfonamides, and tartrates are among those that may require special care in stabilization when
used. The butadiene-acrylonitrile rubbers used early as plasticizers in PVC and now used in
more complex polyblends, and for impact resistance usually lower the heat stability; however, if
they are used in very small amounts, they will act as heat stabilizers [147].
2. Oxidative Stability
During processing, conditions are favorable for initiating auto-oxidation [144]. Plasticizers having
structures particularly susceptible, such as tertiary hydrogens, may at times contribute to early
color development and other untoward effects. This was noticed particularly when the esters of
oxoalcohols first began to be used. Fischer and Vanderbilt found an antioxidant such as
bisphenol A to be helpful [75].

Page 521
a. Compatibility. Compatibility can be changed. Oxidation may first produce hydroperoxides,
which then break down to alcohols, ketones, and aldehydes, some with C-C unsaturation.
These may undergo further oxidation to acids or polymerization to resinous materials. Some of
the products of oxidation are strongly hydrogen bonding or highly polar and exude making the
surface sticky. The oxidation products, if compatible, are less efficient than the starting
plasticizer, so the resin becomes stiff. While these changes in the plasticizers are occurring,
chain scission, cross-linking, and dehydro-halogenation of the resin may also take place.
If a plasticizer is low in polarity and solubility parameter, as DIDA is, the somewhat reverse
effect, antioxidant-permitted exudation, may occur to cause permanence problems [59].
Without an antioxidant present, a small amount of polar oxidation products produced may blend
with the low-polarity DIDA to give a compatible mixture. With antioxidants present, an exudate
may appear if the vinyl sheet is warmed, since DIDA is more compatible at room temperature
than at somewhat higher temperatures. This is common with plasticizers whose solubility
parameters are lower than that of PVC itself [178].
b. Oxidative Fragmentation. On oven aging, oxidative fragmentation occurs as the
hydroperoxides break down to smaller molecules, such as valer-aldehyde. These lowmolecular-weight fragments are lost, giving volatility figures higher than might be otherwise
expected. Even polymeric plasticizers, which were originally called permanent plasticizers, are
subject to oxidative fragmentation. An antioxidant can inhibit this pseudovolatility effect and give
more nearly normal volatility data either in the plasticized PVC or in the plasticizer per se.
Deterioration of physical properties on oven aging results, therefore, partly from volatility and
oxidative fragmentation and partly because the remaining plasticizer is more viscous and less
efficient. With proper formulation, these problems may be overcome [59].
c. Latent Degradation. Latent degradation is an even more subtle problem. Sometimes the
unwanted effects do not show up during or immediately after processing but are found after
further standing, particularly after storage in a hot warehouse [175]. Since the hydroperoxides
produced by oxidizing plasticizers during processing are colorless and odorless, no harm is
observed until they begin to decompose and oxidize further. The aldehydes, ketones, free
acids, and resinous products may have an odor of burned soap or one reminiscent of Limburger
cheese or of a herd of goats. The product may be covered with brownish splotches or may be
sticky from exuding plasticizer. Such delayed effects are called latent degradation, since the
damage has already been done during processing. Figure 34 shows these effects as related to
oxidation of plasticizers.
Since many of the oxidized fragments of the plasticizer molecule are of low molecular weight
and incompatible, they may spread the degradation to similar but unoxidized PVC products in
the same package by migration or even volatilization, causing them to begin to degrade. Such a
transfer of degradation is called envenomation or leprousation [175]. Adequate stabilization and
use of antioxidants can eliminate these problems. Oxidation problems may be aggravated,
however, by using certain pigments and stabilizers, since an HCl scavening stabilizer for the
resin may be a promoter of oxidation of the resin or plasticizer [177].

Page 522

FIGURE 34 A stylized version of latent degradation: peroxide development and decay with
undesirable color and odor effects. (From Ref. 178; reprinted with permission of John Wiley &
Sons, Inc.)
d. Effects of Plasticizer Blends. It is difficult to predict the effects of plasticizer blends. A small
amount of an epoxy plasticizer can act as an oxidation promoter when mixed with phthalate or
adipate plasticizers [175, 178]. Figure 35 shows that in 50:50 mixtures of an epoxidized soya oil
with plasticizers having varying amounts of olefinic unsaturation, the rate of oxidation is higher in
the mixtures than in the individual plasticizers. In thermal oxidation, the increase in rate is most
marked with saturated plasticizers that are slow to oxidize alone. The beneficial effect of epoxy
plasticizers outweighs these detrimental effects in a well-stabilized composition. The value of
increased epoxy plasticizer content with other plasticizers such as phosphates is well known.
Small amounts of hydrocarbon secondary plasticizers will give autoinhibiting mixtures with many
primary plasticizers [139,177]. Theory for such autoinhibition has been proposed [178]. The
effect in DOP is shown by Figure 36. This beneficial effect is at times offset or hidden by other
components, stabilizers, antioxidants, and pigments.
J. Light Stability and Weathering
1. Effect of General-Purpose Plasticizers on PVC Stability
Plasticizers have commonly improved the light stability of typical commercial PVC when used in
plasticizing amounts. The more common carboxylic acid ester plasticizers vary in their light
stability from fair to good. A clear, 4-mil film well stabilized with a good barium, cadmium, zinc,
phosphite epoxy system but containing no ultraviolet light (UV) screener, exposed to direct
weathering in Florida, will last about 10 to 24 months, depending on the plasticizer. A similar
film of rigid PVC (little plasticizer) may begin to darken in 2 to 4 months. In certain simple
formulations with different stabilizer systems, this difference between plasticized and
unplasticized PVC has appeared more complex; unplasticized PVC degraded more rapidly than
plasticized PVC

Page 523

FIGURE 35 Thermal oxidation promoted by epoxy plasticizers. Rate of oxygen uptake as

shown by pressure drop in a closed system is shown for three polyester plasticizers (Pe) with
varying degrees of unsaturation as shown by iodine number, and phenyl oleate per se, and in
admixture with 50% epoxidized soy oil (ESO): la, Pe-1, 12 No. 0.0; 1b, Pe-1 + ESO 50:50; 2a,
Pe-2, 12 No. 1.1; 2b, Pe-2 + ESO 50:50; 3a, ESO, 12 No. 1.7; 4a, Pe-3, 12 No. 6.7; 4b, Pe 3
+ ESO 50:50; sa, Phenyl oleate 12 No. 68.6; 5b, Phenyl oleate + ESO 50: 50.
when exposed to far UV (<290 nm) but degraded more slowly than plasticized PVC in the near
UV (>290 nm). Phthalates and trimellitates act as UV ab-sorbers to retard the very rapid
degradation of PVC common in the far UV but also absorb in the near UV, where a pure PVC
would not, to accelerate the otherwise slow degradation [15]. Far UV, when present, is very
damaging to plasticized PVC, but it is filtered out of sunlight by the atmosphere, so very little
occurs at sea level. Near UV is less energetic but reaches the earth in sufficient quantity to be a
major factor in weathering.
Low-molecular-weight plasticizers such as dibutyl phthalate are lost by volatilization rather than
by photode gradation; the plasticized film retains good color but stiffens. Higher-molecularweight phthalates are retained until the resin system, if clear, begins to develop brown spots or
turn brown all over from UV degradation and photo-oxidation. Even DOP, as high in molecular
weight as it is, suffers from volatile loss in long-term exposure. In compositions pigmented by
carbon black, and thus protected from UV, volatile loss is the major reason for failure. In those
pigmented by TiO2, photo-oxidation is likely to be the major reason for failure, at times slowed,
at times promoted by the pigment [178].
In dialkyl phthalates the more branching that exists in the alkyl groups, the more tertiary
hydrogens are available for easy oxidation [145,178]. Figure 37 shows this trend in 0.1-mm (4mil)-thick clear PVC sheets plasticized

Page 524

FIGURE 36 Thermal oxidation inhibited by hydrocarbon secondary plasticizer, partially

hydrogenated terphenyl (PH Ter). Concentrations at each curve are percentage of hydrocarbon
in a DOP-hydrocarbon blend. Solid lines show peroxide buildup; dashed lines show color
development. (From Ref. 177.)
with 50 phr of the individual plasticizers and stabilized with a Ba, Cd, Zn, ESO, UV absorber
system. Branching increases from the di(heptyl, nonyl, undecyl) phthalate to DOP, DINP, and
DIDP with the average number of branches per molecule being about 0.6, 2, 4, and 6,
respectively. The DIDP composition has severely failed in 2 years; DINP has become brown but
net so brittle in 2 years. Both DOP and the C7-C11 phthalates are good after 3 years of
exposure [178]. Brittleness testing on these films and on 0.25-mm (10-mil) and 0.5-mm (20 mil)
sheets of the same compositions show the DOP composition losing its low-temperature
properties faster than the C7-C11 phthalate, presumably by volatility.
For high-molecular-weight dialkyl phthalates with linear alkyl chains, stabilizer systems must be
selected which have good compatibility with the whole composition. Many that are adequately
compatible with DOP are not compatible enough with linear phthalates. This leads to early
deterioration at the surface.
Even strictly linear alkyl phthalate esters may also be oxidized, apparently losing two or four
carbons at a time from the chain to give lower homologs. When oxidation becomes severe,
much of the plasticizer becomes permanently trapped in a cross-linked PVC gel [208].

Page 525

FIGURE 37 Natural weathering of PVC plasticized with general-purpose phthalate plasticizers:

Effect of branching. Plasticizers: di(heptyl, nonyl, undecyl) phthalate (Di C7-C9P); di-2ethylhexyl phthalate; diisononyl phthalate; diisodecyl phthalate. Exposure: Miami, Florida, 45
from horizontal, due south, direct to sun, July 1975 to July 1978. (From Refs. 145 and 178;
reprinted by permission of John Wiley & Sons.)
2. Effects of Specialty Plasticizers
Each type of specialty plasticizer has its own unique behavior on weathering of plasticized PVC.
Rarely are the specialty plasticizers used as the sole plasticizer; most frequently they are used
in blends with a dialkyl phthalate. The effect on weathering of plasticizer blends is often very
unpredictable from their behavior as individual plasticizers.
a. Alkyl Benzyl Phthalates. Because of their high aromaticity and refractive index the alkyl
benzyl phthalates give improved surface gloss and cleanness or low dirt pickup. They are less
resistant to UV than are the dialkyl phthalates, but their good compatibility, their ability to aid
processing, and their surface effects make them serviceable in some outdoor uses.
b. Adipates, Azelates, and Sebacates. The dialkyl esters of these aliphatic dicarboxyl acids do
not absorb in the near UV or that portion of the UV spectrum that reaches the earth's surface
near sea level. It might be expected their compositions would have excellent weather
resistance. Because

Page 526
of relatively low compatibility within the resin, plasticizer, stabilizer system when these are used
as the sole plasticizer, their compositions are short lived because of early surface degradation.
However, when used in blends with dialkyl phthalates, they can improve light stability slightly.
Adipates are frequently used in exterior applications, especially in Europe. Volatility and water
sensitivity are problems in hot, moist climates. The higher-molecular-weight azelates and
sebecates are less volatile and so fit special needs. Fungus growth is a problem (see later).
c. Polyesters. Esters of glycols and adipic, azelaic, or sebacic acid capped off with either
fatty acids or long-chain alcohols show the same response to UV and for the same reason that
the monomeric adipates do. Once UV degradation begins, the polyester is easily oxidized,
and the sheet becomes brittle earlier than originally predicted for these so called permanent
plasticizers. When solvent resistance or very low volatility is demanded, a properly stabilized
composition will give reasonable service life outdoors for use in signs and automobile decals
and the like.
d. Trimellitates. Trimellitate esters absorb UV farther down into the near UV than the phthalates
do, and consequently are more difficult to stabilize against UV degradation [15,178]. Their low
volatility is the major factor that promotes their use in special weathering situations.
e. Epoxy Plasticizers. Plasticizers containing epoxy groups help the stability of plasticized PVC
formulations where temperatures are abnormally high, as in the southern desert areas of
Arizona. Weather resistance is enhanced, since heat stability is improved by the epoxy
compounds. In milder-temperature exposures the value of the epoxy is primarily to provide
good stability through processing so that high-quality products are exposed. There are smaller
but significant benefits during exposure in thick sheets (see also Section VIII. J. 3).
f. Phosphates. As the sole plasticizer in flexible PVC, phosphates do not last as long as
phthalates. 2-Ethylhexyl diphenyl phosphate is perhaps the best [57,62,145]. In blends with
dialkyl phthalates 2-ethylhexyl diphenyl phosphate can add significantly to outdoor durability. For
example, 4-mil film with no UV screener that is plasticized with 2-ethylhexyl diphenyl phosphate
will last 1 year in Florida; plasticized with DOP, it will last 1-1/2 years, but if 10% of the DOP is
replaced with 2-ethylhexyl diphenyl phosphate, the film will last about 2-1/2 years [55]. All the
reasons for this synergism are not clear, but phosphates improve compatibility of the total
system. They sequester certain pro-oxidant metal ions and show antioxidant effects. This
phosphate effect is produced by other alkyl aryl phosphates and triaryl phosphates. It is of
special value in agricultural film, where the use of UV absorbers would hamper proper growth
of many plants under the protection of the film.
g. Phosphites. Organic phosphites find wide use in PVC and many other resins because of the
improved light stability achieved with very small amounts. Triphenyl phosphite [55] at 0.5 phr
can more than double the outdoor durability of plasticized PVC. Other phosphites, diphosphites,
and poly-phosphites have been developed to give even longer product life. They act as
antioxidants, chelators of metal pro-oxidants, passivators of metal equipment surfaces, and so
on. They are extremely valuable in conjunction with UV absorbers.

Page 527
h. Ultraviolet-Absorbing Plasticizers. Early searches for different plasticizers led to esters of
salicylic acid and aromatic esters of aromatic acids. Some of these gave improved light stability
[32,70]. Many years passed before it was found that these esters rearranged in the presence
of UV light to give 2-hydroxybenzophenones, which were by that time used as UV screeners.
Their rearrangement has even been used as a method of preparation of various
benzophenones [117]. Although in proper formulation they can give unusually long life to
unpigmented PVC sheets, an amber color, by-product of the rearrangement, detracts from their
usefulness [56, 178].
3. Exudation Promoted by Light
For many years, in attempting to commercialize certain types of plasticizers, the industry was
confronted with compatibility-instability problems. The plasticizer would be completely
compatible until the final article was put into use; then the surface would become gummy as
plasticizer exuded. Migration carries plasticizer molecules to the surface and back again, but if
the plasticizer contains carbon-carbon unsaturation, it may photo-oxidize and polymerize at the
surface to form compounds that are too polar and of too high molecular weight to return into
the plastic. Photo-oxidation is aggravated by the pres-

FIGURE 38 Photo-oxidation promoted by epoxy plasticizers. Rate of oxidation is shown by

increase in peroxide content as plasticizers sat in loosely covered flasks in a window.
Plasticizers and number designations are the same as those in Figure 35.

Page 528
ence of epoxy plasticizers (Figure 38). Compatibility-instability is usually circumvented by
avoiding unsaturated plasticizers.
4. Plasticizer Concentration Effects
Weather resistance varies with plasticizer concentration. At very low concentrations (5 to 12
phr), DOP and similar products are reported to be harmful, whereas tricresyl phosphate (TCP)
improves light stability [64]. As plasticizer content is increased past the plasticization threshold,
these effects are reversed and the normal beneficial effect of DOP and similar plasticizers is
seen; but TCP as the sole plasticizer is poor. It undergoes chemical change, becoming dark
colored and less efficient [62]. With DOP, maximum stability occurs at a concentration of about
35 phr [55]. With other dialkyl phthalates, a maximum in stability occurs at about the same
plasticizer concentration when common Ba, Cd, Zn stabilizer systems are used. Some evidence
exists that more highly branched chain phthalates may not go through this maximum [209].
K. Radiation Stability
Irradiation by ionizing radiations is used constructively to cure inks and top coats on
plasticized PVC or, if desired, to initiate cross-linking in the PVC itself. Sterilization by radiation
of medical materials packaged in PVC is growing industrially. If radiation is too intense, it will
rapidly degrade the PVC and plasticizer as well.
Ionizing radiations are those which are energetic enough to produce ions either directly or
indirectly in a medium such as air or water. They may be charged particles such as rays and
rays (electrons), or they may be uncharged particles (neutrons) or electromagnetic radiation
rays or x rays. There is no sharp cutoff point between x rays and UV light. Radiation sources
are most commonly cobalt 60 ( and rays) and electron generators and accelerators. The
latter may vary from low power electron beam equipment of about 150 to 500 kV to powerful
billion-electron-volt (BeV) generators [210].
Polyvinyl chloride homopolymer is moderately resistant to gamma rays. It is almost identical to
polyethylene but quite inferior to polystyrene. Its liberation of HCl during irradiation makes it
unsuitable for many uses. Internal plasticization with vinyl acetate lowers its stability [115]. The
comparison with polyethylene and polystyrene illustrates the value of aromatic structures with
high resonance energy which absorb the radiant energy and dissipate it slowly and with
relatively little harm. Even when converted to free radicals, they result in slow reaction and
short kinetic chain reactions because of the resonance stabilization.
The extreme sensitivity of the overall system to plasticizer choice is suggested by the amount of
gel formed by cross-linking PVC in solution in various solvents. Figure 39 shows the degree of
cross-linking under the same irradiation (30 Mrad of gamma rays) of PVC with various solvents
and swelling agents.
Wells and Williams [205] found that PVC plasticized with 40-phr tritolyl phosphate retained its
mechanical properties better than when plasticized with DOP, which in turn was better than the
completely aliphatic ester DOZ (Table 14). The two polyesters were also rather good in

retention of mechanical properties. However, the phosphate and especially the polyester
became very dark in color. The moderate resistance of DOP to mechanical change and its

Page 529

FIGURE 39 Radiation cross-linking of PVC in solution or swollen by solvents. Concentration: 20

g of PVC in 80 g of solvent. Irradiation: 60Co, 30 Mrad. (From Refs. 36 and 212.)

Tritolyl phosphate
Di-2-ethylhexyl phthalate
Di-2-ethylhexyl sebacate
Polypropylene sebacate
Polypropylene sebacate modified
Source: Ref. 205.

2140
1910
1310
2050
1960

3.18
3.64
3.18
3.42
2.34

1590
670
185
1440
1620

2.08
1.20
0.44
1.58
1.94

Page 530
superior color stability resulted in its being the product of choice for PVC packaging designed
for radiation sterilization (see below).
In most irradiation studies concerning plasticizers and related compounds, they have been
irradiated alone. Polyvinyl chloride can withstand about 1.9 x 109 ergs/g (C) (ergs per gram
referred to carbon as an energy-absorbing control) [115]. By comparison, aromatic and alkyl
aromatic hydrocarbons, secondary plasticizers for PVC, are usually quite stable, having
damage thresholds varying from about 1 x 1010 ergs/g (C) for some alkylated aromatics to
about 4 x 1012 ergs/g (C) for the terphenyls. Of organic compounds, terphenyls are among the
most stable toward high-energy irradiation [45,213]. In alkyl side chains, both branching and
lengthening tend toward reduced stability. With esters, increases in viscosity and acidity are
slight to serious at 1010 ergs/g (C). Aromatic esters such as phthalates are more stable than
aliphatic esters such as adipates, sebacates, and azelates. Lengthening or branching the
carbon chains reduces stability here also. Phosphate esters are generally very sensitive to
irradiation, with large increases in viscosity and acidity. Because of resonance energy in the
aromatic rings, the aryl phosphates are much more stable than the aliphatic ones [213].
L. Sterilization of Plasticized PVC
Biological sterilization of plasticized PVC is most frequently done during and because of
sterilization of medical devices packaged in PVC. Available techniques include dry heat; steam;
use of fumigants, especially ethylene oxide (EO); and radiation. The last two are most suitable
[72,123]. Ethylene oxide can readily diffuse into and through plasticized PVC for sterilization of
packaged goods. it is to be expected that the higher the plasticizer content sand the more
efficient the plasticizer, the faster will be the permeation. The biocidal action results from
ethoxylation of amino and hydroxyl groups on the organism. The residual EO must be allowed
to diffuse from the package or the same reactions can burn the skin of the person who opens
it.
Irradiation by cobalt 60 or by electron beam is increasingly used to sterilize articles packaged in
PVC [186]. Plasticized PVC began to be used in significant amounts when it was realized that
the tinge of brown developed on irradiation did not hurt throwaway package materials. DOP
and linear dialkyl phthalate plasticizers are most commonly used. Dialkyl adipates are used to
some extent, although they are less stable [123]. Typical PVC stabilizers are used. Nontoxic or
low-toxicity stabilizers of barium, calcium, and zinc or the tin maleates are most favored [123].
Typical UV absorbers such as 2-hydroxy-4-octoxybenzophenone inhibit color formation fairly
well [1]. Superior proprietary stabilizer systems are also in use [123].
M. Fungal and Bacterial Resistance
With the large numbers of destructive microbiological agents and the adaptability of each, it is
highly probable that there are no plasticizers that are completely free from fungal or bacterial
attack. Some are readily used by microorganism as sources of carbon. Others are sufficiently
resistant that they are useful even in warm, moist, or wet environments. A study of 24 species
of fungi showed that all would attack at least some of 90 ester plasticizers, some vigorously
[18]. As the fungus grows, secreting digestive enzymes, film degradation precedes the

advancing mycelial growth. Destruction of the plasticizer results in tack and exudation,
embrittlement, weight loss,

Page 531

FIGURE 40 Ability of various aliphatic dicarboxylic acids to support fungus growth. Bars
indicate average growth of 24 different fungi after 2-week incubation. (From Ref. 18.)
and discoloration. In general, diesters of saturated, aliphatic dibasic acids can be used by fungi
if they contain 12 or more carbon atoms. Hydrolysis of smaller molecules is so rapid that the
liberated alcohols inhibit fungus growth. Attack on the dibasic acids is least with oxalic and
adipic, as shown in Figures 40 and 41. Alkyl phthalates and phosphates are resistant to fungi
[18,206]. Phthalate-phosphate blends of about 85:15 ratio, which give the best resistance to
UV degradation, also show the most resistance to fungus growth [57].
Polyhydric alcohols can be used by fungi if the hydroxyl groups are on adjacent or end carbon
atoms. Compounds, including stabilizers, containing fatty acids residues are especially
susceptible. This includes acetates, butyrates, laurates, oleates, stearates, and oils such as
tall, tung, linseed, soya, cottonseed, castor, and dehydrated castor oil. Polyesters from
polyhydric alcohols, dibasic acids and alcohol, or fatty acid terminating groups vary from easily
attacked to reasonably inert, depending on the selection of components. Epoxidized oil and
ester plasticizer-stabilizers are especially susceptible to fungus growth (Figure 42). They furnish
most of the food for fungus growth in the majority of PVC compositions. Epoxy resin stabilizers
derived from glycidyl ethers of Bisphenol A are not so attacked.
Fungus growth on plasticized PVC may be luxuriant at times. At other times, it is detected
mainly by pigments or dyes produced by the organism that dissolve into the plasticizer, spread,
and even migrate to other materials. The actinomycete Streptomycetes rubrireticuli produces
pinking on plasticized PVC under acid conditions [214]. A similar effect is produced by at least
one species of Penicillium and an orange stain by a Fusorium [83,174]. Other fungi produce
different colors under different conditions of acidity. Aspergillus fansecaeus normally produces
a brown pigment, but on ultraviolet radia-

Page 532

FIGURE 41 Ability of various dialkyl adipates to support fungus growth. Bars give average
growth of 24 fungi after 2
tion, it mutates and produces the yellow pigment fonsecin, which turns red in a basic medium
[80]. Proper choice of plasticizer and if need be a fungicide or fungistat in the formulation can
control fungal attack for some time under most conditions. Control is most difficult in direct
sunlight. Most infestation occurs on plasticizer, stabilizer, or lubricant which is low enough in
compatibility with PVC to migrate to the surface and adhere there [194] Many biocides are
available and recommended for use in plasticized PVC. Those most commonly used are 10,10'oxybisphenoxarsine, 2-n-octyl-4-iso-thiazolin-3-one, and N-(trichloromethylthio) phthalimide
[2,112,194]. Internally plasticized resins are less susceptible to attack than are externally
plasticized ones [190].
Bacterial attack is similar to fungal attack, less obvious but geographically more widely
distributed, since some bacteria may live in anaerobic environments deep in the soil or in
deoxygenated water. Polyvinyl chloride film exposed to Pseudomonas aeruginoso lost up to
23% in weight and shrank 13% in 14 days at 28C when diisooctyl sebacate was the
plasticizer. Phthalates and triaryl phosphates are quite resistant under the same conditions,
showing only slight weight loss, negligible shrinkage, and slight evolution of gases [23]. In
addition to undergoing shrinkage, stiffening, and discoloration, as obvious signs of degradation
from bacteria, PVC insulation may lose its electrical resistivity when buried undergrown; vinyl
pipe or tubing or vinyl tank liners may develop sufficient bacterial slime in a matter of a few
hours to stop the flow of water or cause mechanical problems in equipment

Page 533

FIGURE 42 Fungal attack with various plasticizers. The relative inertness of DOP and epoxy
resin is shown, and also the susceptibility of epoxidized soya oil. (From Ref. 53.)

Page 534
[124]. Selecting resistant plasticizers or including a bacteriostat can minimize these effects.
N. Attack by Insects, Termites, and Rodents
These agents scarcely bother rigid PVC except when it stands in their way to food.
Plasticization increases the probability of attack, particularly by termites, ants, and rodents.
Susceptibility increases as plasticizer level increases. Attack seems to be slight when TCP is
used as the plasticizer [206]. Incorporating insecticides and rodenticides with the plasticizer
shows some promise [108,156].

IX. POLYMERIZABLE PLASTICIZERS

The concept of a plasticizer that would remain fluid, act as a plasticizer for thermoplastic resins
during early fabrication steps, and then polymerize to a rigid resin has been of interest in PVC
technology to make rigid products by plastisol techniques. Although the basic concept is not
new, since polymerizable alkyds have long been used as plasticizers with many different
polymers, the adaptation to PVC is difficult. Polyglycol dimethacrylates such as dipropylene
glycol dimethacrylate were among the first to appear promising, followed by 1,3-butylene glycol
dimethacrylate and trimethylolpropane trimethacrylate, and other convertible plasticizers.
All of these can be used with paste PVC resins to give satisfactory plastisols. Formulation and
processing of PVC with polymerizable plasticizers has demanded unusual expertise to develop
the best possible properties. This has seriously restricted their use. A peroxide such as di-tbutyl peroxide as a free-radical initiator is required, at about 1 phr, to polymerize the plasticizer
as the resin fuses. The resin requires stabilization, and the usual stabilizers designed to inhibit
free-radical degradation processes are antagonistic in varying degrees to the polymerization
initiator. Improper processing can result in severe odor and color problems as well as gassing
and poor hardness or low impact strength. With suitable care, polymerizable plasticizers permit
production of semirigid products by the versatile plastisol techniques. Rotational casting has
produced hobbyhorses and mannequins; compression molding has. given appliance handles
and steering wheels; slush molding and dip coating have led to large dolls, advertising objects,
refrigerator trays, and dish drainer baskets. Polymerizable plasticizers have special merit in
coatings, where their use permits production of 100% solids coating materials for application to
steel, cloth, wood, or paper. As a top coat for vinyl tile, it can improve stain and abrasion
resistance [173].
Plastisol coatings with good resistance to solvent swelling and extraction can be made with
polymerization partially initiated by gamma radiation [114]. Trimethylolpropane trimethacrylate
at about 10 phr with dicumyl peroxide is incorporated in plastisol laminating adhesives used for
bonding PVC sheet to polyester cloth and other fabrics. Replacement of 10 phr of DOP with
the methacrylate can double the peel strength. Even more effective compositions also contain a
carboxyl-modified PVC as part of the resin and a moderate amount of a fast-fusing benzyl
phthalate [110].
Tear strength, abrasion resistance, resistance to swelling, and permeation of solvent all

increase with increasing methacrylate content to a point, then all become poorer with further
increase of methacrylate [47,114]. This

Page 535
probably signifies two phases. Compatibility of the methacrylate monomer with PVC is rather
easy to achieve, but compatibility tends to decrease with increasing molecular weight during
polymerization.

X. PLASTICIZATION BY FILLERS, PIGMENTS, SALTS, AND AIR

That part of the usual definition of plasticizers that describes them as liquids or low-melting
solids would exclude many materials of theoretical interest and at times of industrial
importance. In a resin, nonreinforcing fillers such as precipitated calcium carbonate, chalk,
kaolin, silica, and the like are usually wet by the resin molecules, and secondary bonds are
formed between them much like the resin-resin and resin-plasticizer bonds. The normal effect
of filler is to stiffen the resin system; modulus and hardness increase. With increasing filler
concentration, the tensile strength and ultimate elongation will drop as the binder (resinplasticizer) becomes severely diluted with solid filler particles. On the other hand, small
amounts of filler may behave abnormally to give a softer product with properties improved as
if by a plasticizer. Only after additional filler is added to immobilize the polymer chains is the
normal effect of the filler seen. This filler threshold is less frequently documented in the
literature than the plasticization threshold. It is not yet clear how broad it is with regard to
resins and filler types. It was reported by Sabia and Eirich, who found a decrease in modulus
and an increase in elongation at break below 10 vol % CaC03 filler [170]. Figure 43 shows the
decrease in hardness of PVC floor tile compositions when small amounts of calcium carbonate
are incorporated [179]. It is possible that this effect results from some disruption of the
crystallite structure of the PVC, but the work of Kumins and his coworkers [118,119,121] shows
that this cannot be all. With large amounts of filler, polymer segments appear to be immobilized

FIGURE 43 MeBurney hardness (indentation in mils) of solid vinyl floor tile having various
amounts of calcium-carbonate filler; all tile contained 26 phr of butyl benzyl phthalate. (From
Ref. 170.)

Page 536
by absorption at the surface, and the immobilization extends for some distance (~103 ) from
the filler particle, essentially throughout the polymer. The glass transition temperature for at
least several polymers is raised, the resin system is harder, and gas diffusion is reduced, so
that polymer near the solid surface has properties analogous to that of a glassy polymer. At
very low concentrations of TiO2 (rutile) in internally plasticized PVC-VAc (87/13), obvious
plasticization occurred as the upper Tg (PVC backbone transition) dropped until it merged
with the lower Tg (acetate side-chain transition). On further addition, Tg began to rise in the
ormal manner and then leveled off. The authors explain this plasticization by assuming
disruption of polymer-polymer bonds (disgregation) that had been formed between polymer
chains by means of the acetate groups, and the absorption of acetate onto the TiO2. With
occasional cross-links broken, the PVC backbone chains are more free to move. Yet as TiO2
concentration increases, eventually immobilization by absorption onto TiO2 dominates. It is
pertinent to point out that the plasticization observed [118] coincides with a decrease in density
(0 to 13% TiO2) and an increase in diffusion rates found later [119] (Figure 44). The additional
free space introduced is perhaps the immediate cause of plasticization.
Micaceous talc fillers in PVC polyblended with acrylonitrile rubber (NBR) show the abnormal
effect, since small amounts improve elongation, energy to break, and fatigue resistance (Figure
45).

FIGURE 44 Plasticization of PVC/VAc by TiO2 as shown by lowering of Tg at concentrations

that result in volume expansion and increased ease of diffusion of water vapor. (Data from
Refs. 118 and 119.)

Page 537

FIGURE 45 Plasticization of PVC/NBR blends with small amounts of talc filler. Curves 1, 2, and
3 were obtained with PVC/NBR 88:12; curve 4 with PVC/NBR 70:30. (From Ref. 125.)
Conventional plasticizers in a filled resin system tend to cause dewetting of the filler by the
resin, as plasticizer competes with the resin to form bonds with the filler. This contributes to
early whitening or blush during tensile tests as resin is pulled away from filler, leaving vacuoles.
Similarly, in creep tests, strain rate and set increase with plasticizer concentration at constant
filler content [170].
When air or other gas is deliberately or inadvertently incorporated into resin systems, a
noticeable plasticizing action can be produced. Perhaps the most important early commercial
example of this action was in vinyl asbestos floor tile. Moderate amounts of air included with the
large amounts of asbestos filler help produce tile with sufficient softness and flexibility for
installation at reduced plasticizer content. When the compound was reprocessed so that air
was worked out, it becomes more dense, much harder, and too brittle for installation. Very
slight foaming in various products such as electrical insulation, shoe soles, and so on, achieves
improved flexibility. On the macro scale this is similar to the free volume theory of viscous flow
and plasticization [120].

Page 538

XI. ANTI PLASTICIZERS AND ANTI PLASTICIZATION

When plasticizers are added to polymers, they normally lower the modulus and tensile strength
and increase elongation. But at times the exact opposite occurs. This is antiplasticization. It
results when relatively small amounts of a plasticizer are used. Whereas conventional
plasticizers at low concentrations are effective processing aids for rigid PVC [60], the final
product is unacceptably brittle [19]. The effects on tensile strength, modulus, and elongation
are shown in Figures 17 and 18.
Several factors are involved in antiplasticization. Without being bound by any theory, we
suggest that any change in configuration or arrangement of the polymer molecules and their
association with the plasticizer (or antiplasticizer) molecules that result in increased modulus
and lower elongation is antiplasticization. At least three theories have been proposed to explain
this phenomenon: increased crystallinity of the resin, decreased free volume, and increased
bulkiness of polymer and additive associated by polar bonds resulting in restricted movement of
polymer segments or main chains.
A. Crystallinity Theory
Horsley showed by x-ray diffraction increased order in an embrittled system which he attributed
to PVC crystallization permitted by the increased freedom of motion temporarily permitted
when a small amount of plasticizer was present (Figure 46). After larger amounts of plasticizer
were added, the x-ray diffraction intensity decreased, crystalline order was less, and
plasticization was observed. The peak in crystallinity occurred at about 10 phr, while that in
tensile strength was at about 5 phr in the Geon 202 (B.F. Goodrich) copolymer of vinyl chloride
/vinylidene chloride Horsley used. Anti-plasticization peaks need not coincide precisely because
of different methods and rates of testing. When the PVC-plasticizer blend was quench cooled
to avoid crystallization, no antiplasticization was present (Figure 46). Annealing gave time and
proper temperature for more ordered structures to be formed; crystallinity redeveloped, and
antiplasticization occurred.
The increase in refractive index (Figure 46) indicates a more compact structure. The addition of
plasticizer or antiplasticizer progressively lowers the glass transition temperature, with the first
few phr most effective (Figure 46). Both effects are in keeping with the later-developed free
volume theory of plasticization.
Although increased segment mobility can favor crystallization in many resins, and crystallinity
can increase tensile strength and modulus, this cannot be the total cause of the antiplasticizing
effect. The effect is greatest with polar plasticizers, which are less conducive to resin
crystallization than are less polar ones. Important changes occur in the amorphous regions of
the resin as well. Furthermore, crystallinity cannot be essential in antiplasticization, since a
number of noncrystalline polymers, including polymethyl methacrylate, can be antiplasticized.
Sefcek and coworkers found antiplasticization in a branched homopolymer PVC but detected
little or no crystallinity [182]. Razinskaya found that PVC prepared in different ways responded
quite differently foward low concentrations of plasticizers, with emulsion-polymerized PVC

showing little or no antiplasticization [158].

Page 539

FIGURE 46 Some effects of antiplasticization and plasticization of PVC. Resin Geon 202.
(From Ref. 98.)
B. Reduced Free Volume Theory
If the free volume theory is productive in understanding plasticization, an extension of it should
be applicable to the antiplasticization phenomenon. Robeson proposed that as a plasticizer is
added in small amounts, it fills the free volume in the resin before it begins to separate the
polymer molecules to create more free volume [168]. This hampers the local movement of
polymer segments. Robeson showed that it also results in an increase in density of various
polymer systems and an increase in the apparent density of the antiplasticizer. This decreases
the permeability of polymer to gases like C02 and water vapor. The decrease results mainly
from a decrease in the diffusion coefficient, supposedly from loss of free volume, but partly also
from loss of solubility in the polymer as potential solubility sites for water are taken up by the
antiplasticizer. The last point indicates more than mere filling of free volume is involved, since
some type of bonding appears to be involved. This bonding also immobilizes the polymer chain
so that the mechanical loss transition is eliminated and the resin is antiplasticized and brittle
[168].

Page 540
C. Hindered Molecular Motion Theory
While Robeson showed decreased permeability as small amounts of plasticizer are added to a
resin and the system enters the antiplasticized region [168], Sefcik and coworkers showed that
permeation again increases as more plasticizer (TCP) is added and plasticization occurs (Table
15). They also note that calculations of free volume indicate that free volume decreases
monatonically over the entire range, so does not increase as plasticization occurs. This is in
conflict with the concept that increased plasticizer content increases plasticization by increasing
free volume [182].
Quite early Ghersa had proposed as an explanation of antiplasticization that plasticizer
molecules become attached to PVC through their polar groups, and at low concentrations they
act as cross-links through increased steric hindrance [82]. Later it was observed that some of
the secondary transitions, particularly the transition, which occurs at a temperature lower
than the glass transition Tg (or transition), are associated with ductility and good impact
strength. If such a transition does not occur, the resin is brittle at room temperature. Small
amounts of plasticizers in PVC reduce and even eliminate its ~ transition, as shown in the work
of Sundgren and coworkers with PVC (anti) plasticized with polycaprolactone (PCL) (Figure
47). Impact energies of these PVC compositions decreased with increasing PCL content to
about 12% and then improved or rose with further increase in plasticizer content.
The reasons for this transition and why it is inhibited by small amounts of plasticizer are still
debated. The slight peaks in tan indicate some freedom of movement which is lost on cooling
below this transition temperature. One theory says the movement is of local chain segments.
More recent theory explains it as a cooperative motion of main chains [195]. From recent
theories of diffusivity of gases, it appears that a gas molecule like CO can slide, along
channels formed from, say, four polymer molecules lying parallel to each other until a region of
high density stops it. It can jump at right angles to
10 [cm3 (STP)cm/cm2cm Hg]a TCP (wt %) H2 CO 0 2.38 0.0248 5 1.91 0.0154 10.2 1.77
0.0178 15 1.69 0.0180 20.1 2.16 0.0274 30.8 2.86 0.105 40.0 3.61 0.374 aAt 270 cm Hg and
27C. Source: Data from Ref. 182.

Page 541

FIGURE 47 Loss tangent as a function of temperature for PVC/polycaprolactone (PCL) blends.

Dynamic mechanical measurements at 110 Hz. (From Ref. 195.)
these polymer chains only when two adjacent chains undergo sufficient molecular motion to
cause separation between them greater than the molecular dimensions of the diffusant, as
when two bars of a cage are bent apart simultaneously. The activation energy of diffusion is
viewed as the minimum energy to separate the molecules this distance, and the transition
reflects the loss of this freedom below the transition temperature [195,146,185].
This freedom of cooperative motion permits not only diffusion of gases but creep of the
polymer, both of which are greatly inhibited by antiplasticization. In keeping with Gersa's
concept of polar cross-links as the cause of antiplasticization [82], recent theory views the
effect of TCP, for example, as increasing the cohesion so that the polymer chains are harder to
separate. The energy of activation of diffusion is increased, and the modulus and tensile
strength are increased. But the loss of this freedom of motion causes the polymer to become
brittle [182]. As Sefcik and coworkers explained at high additive concentrations the physical
separation of the polymer chains by the diluent (plasticizer or antiplasticizer) must result in a
reduction of the energy necessary to separate the polymer chains, thereby plasticizing the
polymer and overcoming any antiplasticization effect of low concentrations of additives [182].
We have therefore come full cycle, since this is, with some clarification, the lubricity theory,
which was the first theory of plasticization of synthetic polymers.
If the polar groups of plasticizers and PVC are tied up in sluggishly movable structures, the
dielectric constant should drop at low plasticizer concentrations, which it does. In PVC/TCP
compositions it reaches a minimum at about 10 phr TCP [178]. Also the type and shape of the
antiplasticizing molecules with two polar groups should be reflected in their behavior. When di2-ethylhexyl sebacate (DOS) and di-2-ethylhexyl adipate (DOA) are used as

Page 542
antiplasticizers, it requires only five ester groups of DOS and 5.5 of DOA per 100 repeat units
of PVC to suppress completely the transition of PVC, but 10 ester groups of phthalates such
as DOP [195]. The ester groups of DOS and DOA are separated by long alkyl chains, so are
free to act independently, as if each molecule were two molecules. When these alkyl esters are
used, therefore, antiplasticization reaches a peak at about 2 phr and is bypassed by unusually
small amounts of plasticizer. Bohn speaks of this differently. He theorizes that commencement
and termination of brittleness as plasticizer concentration is increased must be treated as two
separate phenomena [19]. Brittleness is caused by the suppression of the low-temperature
relaxation process and is relatively independent of plasticizer type. The termination of
brittleness, on the other hand, is strongly influenced by plasticizer type. Thus the peak occurs at
lower concentrations with the more efficient plasticizers.
Several workers [17,195] have carried this study into the area of impact modifiers, the socalled particulate plasticizers [178]. Using nitrile-butadiene rubber (NBR), they found that with
an acrylonitrile content of 21.7% the NBR was semicompatible with the PVC and showed
plasticizer behavior by an increase in creep compliance or lower modulus even at 4% in PVC.
But an NBR with 40.6% acrylonitrile was much more compatible with PVC and showed
antiplasticization by decreased creep compliance or increased modulus.
Since polar bonding only retards molecular motion to give antiplasticization rather than stopping
it absolutely, antiplasticization effects are subject to the time-temperature superposition
principle. Mascia has shown that PVC containing 10 phr of TCP is antiplasticized below 38C
but is plasticized above that temperature. Below 38C it is rigid, not showing creep in the first
hour or two. But in longer test times, it shows creep with increased compliance (reciprocal of
modulus), as does a plasticized PVC [131].
D. Uses of Antiplasticized Polymers
The increased tensile strength and modulus have enticed many workers to try to find uses for
antiplasticized resins, including PVC. The increased brittleness has almost always precluded
commercial applications. Jackson and Caldwell made very thin strong film of antiplasticized
polycarbonate for special photographic purposes [104,106]. Others have hinted at minor use
for antiplasticized PVC [158,201]. Antiplasticized fibers have been produced commercially for
specialized uses. Apparently, antiplasticized surface coatings are in use for increased mar
resistance. The major interest in antiplasticization is still for a more fundamental understanding
of resin-additive behavior directed toward improvement in engineering thermoplastics, improved
impact modifiers, and improved control of diffusion through plastic membranes.

XII. TOXICITY AND ENVIRONMENTAL ASPECTS

PVC itself is relatively inert to most biological systems, since the molecular size is too large for
it to enter biological membranes. Plasticizers are lower in molecular weight and have biological
activity. Since they are not combined chemically with PVC, they are accessible to individual
organisms and to the environment as a whole as they leach from plasticized polymers.

Any plasticizer for use in PVC for possible food contact in processing or packaging in the United
States, is regulated by the Federal Food, Drug

Page 543
and Cosmetic Act of 1958 and associated amendments [41]. Among plasticizers that have been
approved for some uses in packaging are DOP, diethyl phthalate, butyl benzyl phthalate,
dicyclohexyl phthalate, bis-2-ethylhexyl adipate, di(heptyl, nonyl) adipate, diisononyl adipate, din-hexyl azelate, tributyl citrate, tributyl acetyl citrate, glyceryl triacetate, epoxy soya oil, n-ethyl
o,p-toluene sulfonamide, methyl acetylricinoleate, hydrogenated terphenyl hydrocarbons, and
certain polyesters [41]. The list is quite restricted for plasticizers for coatings or film that come
directly into contact with foods. A small number of plasticizers are exempt from requirements of
a tolerance by the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) for proper use
on raw agricultural commodities, including food animals, cattle, chickens, and so on [42].
No plasticizer is approved for any of these uses without extensive testing for short- and longterm physiological effects. Plasticizers for any use in PVC are characterized in short-term
(acute) tests to be nontoxic and to have minimal potential for skin sensitization, eye irritation,
and other untoward effects. Regulatory concern was heightened when the National Toxicology
Program (NTP) reported that DOP (di-2-ethylhexyl phthalate, DEHP) and DOA (di-2-ethylhexyl
adipate, DEHA) caused increased liver tumors in rodents. Results for DOA are considered to
be less significant since tumors were observed in only one species. Studies are continuing at
the NTP on butyl benzyl phthalate.
Phthalate esters have been manufactured and used safely in U.S. commerce for more than 40
years. Studies of workers exposed to phthalates have shown no evidence of unusual health
hazard. During this time, hundreds of health and ecological tests have been conducted to
confirm the safety of these products. Many of the test results were reported in the scientific
literature. In addition, the phthalate producers supplied results from about 400 health and safety
studies to EPA in response to reporting rules issued in 1978 and 1982. Hence the overall safety
of phthalate esters has been well documented.
Based on this documentation and on petitions by the producers, FDA issued regulations
permitting the use of several phthalate esters as indirect additives for food packaging
applications. These sanctions still exist. To the best of our knowledge, no evidence of harm has
been observed from this food packaging use.
There is a growing body of evidence to suggest that studies on rats, such as the DOP and
DOA studies mentioned earlier, may not be relevant to human risk assessment for phthalates.
Recent studies at the National Institute of Environmental Health Sciences and studies
sponsored by industry show that rodents metabolize phthalate esters by a different mechanism
than primates, presumably including humans. This suggests that for phthalates, the rat is not an
acceptable model for making direct extrapolations to probable effects on humans. It was also
observed that the very high dose levels used in the. bioassays inhibit the animals' normal repair
processes, thus causing effects which may be entirely absent at normal exposure levels.
Very early, certain chlorinated and nitrated plasticizers and o-cresyl phosphates were removed
from the market because of their irritating and toxic effects. Dibutyl phthalate can induce
allergic reactions in people who respond to bacterial allergens or to dust allergens. Certain

higher dialkyl phthalates and butyl benzyl phthalate appear to be essentially free of such skin
irritations and allergic effects [8]. Metabolism of ingested dialkyl phthalates occurs in a few
minutes in the small intestine of rats, yielding monoalkyl phthal-

Page 544
ates, which are excreted in the bile. Any toxic effects observed appear to result from these
half-esters, particularly from the lower phthalates [111]. Studies of aromatic phosphate ester
structures and neurotoxicity have shown that the toxic species are those with alkyl groups in the
ortho position and with at least one hydrogen on the carbon attached to the benzene ring. By
proper choice of structure such as that of p-t-butylphenyl diphenyl phosphate, a triaryl
phosphate plasticizer is produced that is not characterized by delayed neurotoxicity [107].
Similar concerns for the environment as a whole have increased following rapid industrialization
after World War II. Early monitoring studies for plasticizers were plagued by contamination
from plasticized laboratory and sampling equipment. Development of very sensitive analytical
tools and responsive test methods for environmental effects have led to the ability to monitor
individual compounds in the environment. Recent studies [140a] confirm that benzyl phthalates
and linear alkyl phthalates are essentially nondetectable in the environment. Much of the
physiological and environmental aspects of a plasticizer, in addition to its own physiological
properties; is controlled by its physical properties, its rate of volatile loss from PVC, and its
extract-ability by water, fats, blood, and so on [86]. Biodegradation and bioconcentration of
plasticizers have been studied. Phthalates are degraded quickly in the river die-away test and
the semicontinuous activated sludge test, although some not as quickly as the standard control,
a biodegradable sodium alkylbenzenesulfonate [171]. Methods for evaluation of environmental
persistence, bioconcentration, and aquatic toxicity are becoming more advanced. Phthalates,
for example, tend to partition into sediments in aqueous environments and appear not to
bioconcentrate to a significant degree. Chronic and acute toxicity studies with various algae,
invertebrates such as Daphnia, and to fish such as fathead minnows and bluegill confirm that
there is an adequate margin of safety in the environment (>100) [84].

XIII. WHAT PLASTICIZER TO USE: THE DESIRABILITY FUNCTION

Choosing a specific plasticizer for a given use is not easy when so many are available. Using
any plasticizer requires a compromise of desirable properties in every case. If it is desirable,
for example, to have high tensile strength, tear strength, and abrasion resistance, but good
flexibility at low temperatures, very low oil and water extraction, good cleanability, and low
gloss in a vinyl product, the optimum values of all properties cannot be achieved in one
formulation. Some of the desired properties will always be in conflict with others. In a
plasticizer, what leads to good low-temperature flex leads to high oil extraction; high tensile
strength is accompanied by high modulus and poor flexibility. When on top of this there must be
low odor, light stability, efficiency, processability, and always cost, selecting the plasticizer with
the best balance of properties at an acceptable price is difficult. With some experience, good
intuitive judgments may be made, especially with a checklist such as the one in Figure 48.
Errors in judgment are frequent and costly, commonly occurring because some property
overlooked or considered of little significance (color, fungal resistance, humidity compatibility,
etc.) meets serious consumer objections. To minimize the chance of poor judgment, Harrington
has proposed the desirability function, which-is one number calculated to express the overall
desirability obtained by consideration of individual properties, both objective and subjective [94].

On the desirability

Page 545

FIGURE 48 Plasticizer evaluation checklist.

curve (Figure 49), d represents the desirability of any property ranging from 0.00, completely
unacceptable, to 1.00, maximum desirability, as shown in Table 16. Measured values for each
property are made to conform to the Y scale, in such a way that the minimum (or maximum)
acceptable value corresponds to Y = 0. This is usually a product specification value. A property
value considered to be very good, beyond which improvement would be useless, is set at Y =
5. For example, if a low-temperature flex of -20C is ab-

Page 546

FIGURE 49 Desirability curve and its application to three properties. (From Ref. 94.)

1.00
0.80-1.00
0.63-0.80
0.40-0.63
0.30-0.40
0.0-0.30
0.00

Maximum desirability
Acceptable and excellent
Acceptable and good
Acceptable but poor
Borderline
Unacceptable
Completely unacceptable

Page 547
solutely required and -45C is as good as could be desired, -20C would be written at Y = 0
and -45C at Y = 5. Intermediate and higher temperatures would be filled in accordingly. When
all properties, including subjective ones such as odor, have been placed along the Y scale, the
measured properties of each plasticizer can be located and the corresponding desirability d
determined for each property from the curve. The overall desirability D is the geometric mean
of all the small d values. By this procedure, if any property is totally unacceptable and is given a
value d = 0.00, the overall desirability is zero no matter how fine other properties of the
plasticizer may be. Furthermore, D is strongly weighed by the smaller d values, just as a
consumer decides to accept or reject a product by its weaker points. With an overall
desirability D of 0.65 to 0.90, a high-quality product with good consumer acceptance is virtually
assured.

REFERENCES
1. R. V. Albarino and E. P. Otocka, J. Appl. Polym. Sci., 16, 6168 (1972).
2. N. Allbee, Plast. Compd., 5, 4146 (Mar./Apr. 1982).
3. T. Alfrey, N. Wiederhorn, R. Stein, and A. Tobolsky, Ind. Eng. Chem., 41, 701 (1949).
4. C. E. Anagnostopoulos and A. Y. Coran, J. Polym. Sci., 57, 1 (1962).
5. C. E. Anagnostopoulos, A. Y. Coran, and H. R. Gamrath, J. Appl. Polym. Sci., 4, 181
(1960).
6. C. E. Anagnostopoulos, A. Y. Coran, and H. R. Gamrath, Mod. Plast., 43(2), 141
(1965).
7. ASTM D-883 Plastics Nomenclature, 1966 Book of ASTM Standards, Part 27,
Philadelphia, 1966, p. 381.
8. N. A. Baida, D. P. Kachalai, and S. I. Yalkut, Allergiya, 1, 42 (1980).
9. C. H. Baker, W. B. Brown, G. Gee, J. R. Rowlinson, D. Stubley, and R. E. Yeadon,
Polymer, 3, 215 (1962).
10. T. H. Barry, Natural Varnish Resins, Ernest Bonn, London, 1932.
11. R. S. Barshtein and G. A. Kotlyarevskii, Sov. Plast. (Eng. trans.), 7, 18 (1966).
12. A. F. M. Barton, CRC Handbook of Solubility Parameter and Other Cohesion
Parameters, CRC Press, Boca Raton, Fla., 1983.
13. A. Beerbower, L. A. Kaye, and D. A. Pattison, Chem, Eng., 74(26), 118 (1967).

 14. A. Beerbower, D. A. Pattison, and G. D. Staffin, ASLE Trans., 6, 80 (1963).

15. I. S. Beggins, D. L. Gerrard, and G. E. Williams, J. Vinyl Technol., 4(4), 150156 (Dec.
1982).
16. H. S. Bergen and J. R. Darby, Ind. Eng. Chem., 43, 2404 (1951).
17. G. Bergman, H. Bertilsson, and Y. J. Schur, J. Appl. Polym, Sci., 21, 29532961 (1977).
18. S. Berk, H. Ebert, and L. Teitell, Ind. Eng. Chem., 49, 1115 (1957).
19. L. Bohn, Kunststoffe, 53, 826 (1963). Good bibliography.
20. A. Bondi and D. J. Simkin, AIChE J., 3, 473 (1957).
21. D. Bower, Plast. Compd., 2(1), 64 (1979).
22. R. L. Brooks and A. R. Lawrence, J. Appl. Polym. Sci., 9, 707 (1965).
23. R. Burgess and A. E. Darby, Br. Plast., 37, 32 (1964).
24. H. Burrell, Interchem. Rev., 14, 31 (1955).

Page 548
25. H. Burrell, FATIPEC (Fed. Assoc. Techniciens Ind. Peintures, Vernis, Emaux Encres
Imprimerie Europe Continental), Wiesbaden, 1962, pp. 2130.
26. H. Burrell, Off. Dig. Fed. Soc. Paint Technol., 27, 726 (1955).
27. H. Burrell, Off. Dig. Fed. Soc. Paint Technol., 29, 1159 (1957).
28. W. F. Busse, J. Phys. Chem., 36, 2862 (1932).
29. J. A. Cannon, D. E. Emgee, and M. T. Payne, J. Vinyl Technol., 2(4), 250253 (1980).
30. J. A. Cannon and G. Mahnken, unpublished work, Monsanto Co.
31. L. T. Carleton and E. Mishuck, J. Appl. Polym. Sci., 8, 1221 (1964).
32. T. F. Carruthers and C. M. Blair (to Carbide and Carbon Chemicals Corp.), U.S. Pat.
2,157,068 (1939).
33. R. J. Ceresa, Block and Graft Copolymerization, Butterworths, Washington, D.C., 1962,
p. 97.
34. R. J. Ceresa, Polymer, 1, 72 (1960).
35. R. J. Ceresa, Polymer, 2, 213 (1961).
36. A. Chapiro, Radiation Chemistry of Polymeric Systems, Wiley-Inter-science, New York,
1962, p. 585.
37. Chemicals and Additives for Plastics Hit New Peaks in Sales, Mod. Plast., 44(1), 94
(Sept. 1966).
38. F. W. Clark, Chem. Ind., 10, 225 (1941).
39. R. F. Clash and R. M. Berg, Mod. Plast., 21(11), 119124, 160 (July, 1944).
40. A. W. M. Coaker and C. D. Bias, 25th Annu. Tech. Conf., Society of Plastics Engineers,
New York, May 610, 1968, SPE Tech. Pap., 14, 9097 (1968).
41. Code of Federal Regulations, Title 21, Food and Drugs, Office of the Federal Register,
U.S. Government Printing Office, Washington, D.C., revised annually, Apr. 1; Fed. Reg., Part
121, Food Additives.
42. Code of Federal Regulations, Title 40, Protection of Environment, Office of the Federal
Register, U.S. Government Printing Office, Washington, D.C., revised annually, July 1.
43. A. Y. Coran and C. E. Anagnostopoulos, J. Polym. Sci., 57, 13 (1962).

 44. A. Corry, D. Renner, and A. Shiffler, SPE J., 22, 107 (1966).
45. S. L. Cosgrove, in Effects of Radiation on Materials and Components, J. F. Kircher and
R. E. Bowman, eds., Reinhold, New York, 1964, Chap. 4.
46. E. E. Cowell and J. R. Darby (to Monsanto Chemical Co.), U.S. Pat. 2,671,064 (Mar. 2,
1954).
47. G. P. Cowperthwaite, SPE J., 29(2), 56 (1973).
48. S. W. Critchley, A. Hill, and C. Paton, in Plasticization and Plasticizer Processes, R. F.
Gould, ed., Adv. Chem. Ser., 48, American Chemical Society, Washington, D.C., 1965, Chap.
14.
49. J. D. Crowley, G. S. Teague, and J. W. Lowe, J. Paint Technol., 38, 269 (1966).
50. M. L. Dannis, J. Appl. Phys., 21, 505 (1950).
51. M. L. Dannis, J. Appl. Phys., 21, 510 (1950).
52. J. R. Darby (to Monsanto Chemical Co.), U.S. Pat. 2,669,549 (Feb. 16, 1954).
53. J. R. Darby, SPE J., 20(8), 2 (1964).
53a. J. R. Darby and E. E. Cowell (to Monsanto Chemical Co.), U.S. Pats. 2,689,837 and
2,689,838 (Sept. 21, 1954).
54. J. R. Darby and P. R. Graham, Mod. Plast., 32(10), 148 (June 1955).
55. J. R. Darby and P. R. Graham, Mod. Plast., 39(5), 148 (1962).

Page 549
56. J. R. Darby and J. K. Sears, in Applied Polymer Sciences, J. K. Craver and R. W. Tess,
eds., American Chemical Society, Division of Organic Coatings and Plastics Chemistry,
Washington, D.C., 1975, Chap. 42, pp. 610631.
57. J. R. Darby, J. K. Sears, and N. W. Touchette, SPE J., 27(2), 3236 (1971); 270), 74
79 (1971). Corrected papers available from authors.
58. J. R. Darby, N. W. Touchette, and.J. K. Sears, Polym. Eng. Sci., 7, 295 (1967).
59. J. R. Darby, N. W. Touchette, and J. K. Sears, J. Appl. Polym. Sci., 14, 5364 (1970).
60. J. B. DeCoste, SPE J., 20, 764 (1965).
61. J. B. DeCoste, Simulated Service Testing in the Plastics Industry, ASTM Spec. Tech.
Publ., 375, Symposium, Chicago, 1965.
62. J. B. DeCoste, J. B. Howard, and V. T. Wallder, Chem. Eng. Data Ser., 3, 131 (1958).
63. G. Delmas and D. D. Patterson, Off. Dig. Fed. Soc. Paint Technol., 34, 677 (1962).
64. B. Dolezel, Chem. Prum., 13/38(3), 160 (1963), trans. M. Mark.
65. A. K. Doolittle, Prepr., Am. Chem. Soc. Div. Org. Coat. Plast. Chem., 26(2), 248 (Sept.
1966).
66. A. K. Doolittle, Mater. Prot. Mag., 2(11), 32 (Nov. 1963); reprinted with much additional
data, ER-50 Amercoat Corp.
67. A. K. Doolittle, in Plasticizer Technology, P. F. Bruins, ed., Vol. 1, Reinhold, New York,
1965, Chap. 1.
68. A. K. Doolittle, The Technology of Solvents and Plasticizers, Wiley, New York, 1954,
Chap. 15.
69. J. H. DuBois, Plast. World, 19(6), 22 (June 1961).
70. F. W. Duggan (to Union Carbide and Carbon Corp.), U.S. Pat. 2,169,179 (1938).
71. C. Ellis, The Chemistry of Synthetic Resins, Reinhold, New York, 1935, p. 18.
72. R. R. Ernst, in Industrial Sterilization, G. B. Phillips and W. S. Miller, eds., Duke
University Press, Durham, N.C., 1972, Chap. 12, pp. 181208.
73. Exuding Plasticizer Kills Insects, Mod. Plast., 42(5), 168 (Jan. 1965).
74. J. Fath, Mod. Plast., 37, 8, 135 (1960).

 75. W. Fischer and B. M. Vanderbilt, Mod. Plast., 33(9), 164 (1956).

76. M. Fitzgibbon (to Lunvale Products), Br. Pat. 831,790 (Mar. 30, 1960).
77. P. J. Flory, Principles of Polymer Chemistry, Cornell University Press, Ithaca, N.Y., 1953.
78. H. E. Frey, Kunststoffe, 46, 81 (1956).
79. W. J. Frissell, Mod. Plast., 38(9), 232 (May 1961).
80. O. L. Galmarini and F. H. Stodola, J. Org. Chem., 30, 112 (1965).
81. J. L. Gardon, J. Paint Technol., 38, 43 (1966).
81a. F. J. Gay and A. H. Wetherly, Laboratory Studies of Termite Resistance, IV: The
Termite Resistance of Plastics, Div. Entomol. Tech. Pap. 5, Commonwealth Scientific and
Industrial Research Organization, Melbourne, Australia, 1962.
82. P. Ghersa, Mod. Plast., 36, 2, 135 (Oct. 1958).
83. T. A. Girard and C. F. Koda, Mod. Plast., 36, 10, 148 (1959).

Page 550
84. W. E. Gledhill, R. G. Kaley, W. T. Adams, O. Hicks, P. R. Michael, and V. W. Saeger,
Environ. Sci. Technol., 14, 301305, (1980).
85. W. Gordy, J. Chem. Phys., 7, 167 (1939).
86. P. R. Graham, Annu. Tech. Conf., Society of Plastics Engineers, Atlanta, May 58, 1975,
SPE Tech. Pap., 21, 275279 (1975).
87. P. R. Graham and J. R. Darby, SPE J., 17, 91 (1961).
88. P. R. Graham and J. R. Darby, Plast. Technol., 8(4), 33 (Apr. 1962).
89. J. Greenberg and G. D. Cloyd, U.S. Pat. 4,158,051 (June 12, 1979). See also U.S. Pat.
4,134,977 (Jan. 16, 1979) and U.S. Pat. 3,918,407 (Nov. 11, 1975).
90. L. C. Grotz, J. Appl. Polym. Sci., 9, 207 (1965).
91. L. M. Grub and J. K. Baxter, U. S. Pat. 3,852,416 (Dec. 3, 1974).
92. V. E. Gul, E. A. Penskaya, and V. N. Kulenev, Colloid J. (USSR) (English trans.), 27, 283
(1965).
93. C. M. Hansen, J. Paint Technol., 39, 104 (1967).
94. E. C. Harrington, Ind. Qual. Control, 21, 494 (1965).
95. J. F. Heil and J. M. Prausnitz, AIChE J., 12, 678 (1966).
96. J. F. Heil, Phase Equilibria in Multi-Component Polymer Solutions, Dissertation, University
of California, University Microfilms, Ann Arbor, Mich., 1965.
97. J. H. Hildebrand and R. L. Scott, The Solubility of Non-electrolytes, 3rd ed., Reinhold,
New York, 1950.
98. R. A. Horsley, in Progress in Plastics, P. Morgan, ed., Iliffe, London, 1957, p. 77.
99. F. N. Howes, Vegetable Gums and Resins, Chronica Botanica Co., Waltham, Mass.,
1949.
100. K. L. Hoy, J. Appl. Polym. Sci., 10, 1871 (1966).
101. Z. Hrabal and K. Kotik, Kozzarsti, 6, 85 (1956).
102. J. W. Hyatt and I. S. Hyatt (to the Celluloid Manufacturing Co.), U.S. Pat. 105,338 (July
12, 1870).

103. E. H. Immergut and H. F. Mark, in Plasticization and Plasticizer Processes, R. F. Gould,

ed., Adv. Chem. Ser. 48, American Chemical Society, Washington, D.C., 1965, Chap. 1.
104. W. J. Jackson and J. R. Caldwell, in Plasticization and Plasticizer Processes, R. F.
Gould, ed., Adv. Chem. Ser. 48, American Chemical Society, Washington, D.C., 1965, Chap.
17.
105. W. J. Jackson and J. R. Caldwell, J. Appl. Polym. Sci., 11, 211 (1967).
106. W. J. Jackson and J. R. Caldwell, J. Appl. Polym. Sci., 11, 227 (1967).
107. F. R. Johannsen, P. L. Wright, D. E. Gordon, G. L. Levinskas, R. W. Radue, and P. R.
Graham, Toxicol. Appl. Pharmacol., 41(2), 291304 (Aug. 1977).
108. O. S. Johnson, Br. Pat. 853,612 (Nov. 9, 1960).
109. H. Jones, in discussion following F. W. Clark, Ref. 38.
110. S. F. Joyce and J. T. Renshaw, J. Vinyl Technol., 2(2), 106109 (June 1980).
111. H. Kaneshima, T. Yamaguchi, and K. Itoh, Bull. Environ. Contam. Toxicol., 20, 725
(1978).
112. A. M. Kaplan, M. Greenburger, and T. M. Wendt, Polym. Eng. Sci., 10(4), 241246
(1970).
113. M. Kaufman, Advances in PVC Compounding and Processing, MacLaren and Sons,
London, 1962.

Page 551
114. F. F. Khodzhevanov, M. A. Riskina, Yu. M. Al'ter, G. I. Chekanova, S. P. Novikava, and
G. R. Spelmanes, Plast. Massy, 5(12), 28 (1978); Int. Polym. Sci. Technol., 6(4), T/5 (1979).
115. R. W. King, N. J. Broadway, R. A. Mayer, and S. Palinchak, in Effects of Radiation on
Materials and Components, J. F. Kircher and R. E. Bowman, eds., Reinhold, New York, 1964,
Chap. 3.
116. A. Kirkpatrick, J. Appl. Phys., 11, 255 (1940).
117. H. Kobsa, J. Org. Chem., 27, 2293 (1962).
118. C. A. Kumins and J. Roteman, J. Polym. Sci., Part A-1, 527 (1963).
119. C. A. Kumins, J. Roteman, and C. J. Rolle, J. Polym. Sci., Part A-1, 541 (1963).
120. S. S. Kurtz, Jr., J. S. Sweely, and W. J. Stout, in Plasticizer Technology, P. F. Bruins,
ed., Reinhold, New York, 1965, Chap. 2.
121. T. K. Kwei and C. A. Kumins, J. Appl. Polym. Sci., 8, 1483 (1964).
122. L. P. Kyrides (to Monsanto Chemical Co.), U.S. Pat. 1,923,938 (1933).
123. H. Landfield and R. F. Johnson, Sterilization Technology: Medical Polymers and
Applications, Seminar, Tufts New England Medical Center, Boston, June 67, 1983.
124. S. J. Lederer, Today's Chem. Mater., 5 (Sept. 1966).
125. C. C. Lee, W. Rovatti, S. M. Skinner, and E. G. Bobalek, J. Appl. Polym. Sci., 9, 2047
(1965).
126. V. K. Leilich, Kolloid-Z., 99, 107 (1942).
127. W. F. Leistner, A. C. Hecker, and O. H. Knoepke (to Argus Chemical Laboratory), U.S.
Pat. 2,564,646 (1951).
128. Sh. L. Lelchuk and V. I. Sedlis, J. Appl. Chem. USSR (Eng. trans.), 30, 1106 (1957).
129. A. I. Maklakov, A. A. Maklakov, A. N. Temnikov, and B. F. Teplov, Polym. Sci. USSR, 20,
1492 (1979).
130. P. E. Marling (to Monsanto Chemical Co.), U.S. Pat. 2,572,571 (1951).
131. L. Mascia, Polymer, 19, 325329 (Mar. 1978).
132. J. J. Mattiello, ed., Protective and Decorative Coatings, Wiley, New York, 1941, Chap. 1.

133. R. P. Mazzur, SPE Tech. Pap., 13, 177 (May 1967).

134. P. V. McKinney, J. Appl. Polym. Sci., 9, 3359 (1965).
135. P. V. McKinney, J. Appl. Polym. Sci., 11, 193 (1967).
136. D. J. Mead, R. M. Tichenor, and R. M. Fuoss, J. Am. Chem. Soc., 64, 283 (1942). See
also earlier and later papers in this series.
137. J. J. Menn, F. B. Folkemer, and A. Miller, U.S. Pat. 3,166,472 (Jan. 1965).
138. E. H. Merz and R. E. Colwell, ASTM Bull., No. 232, p. 63 (1958).
139. H. E. Mertwoy and H. Gisser, Ind. Eng. Chem. Prod. Res. Dev., 3, 180 (1964).
140. G. Meyer, Br. Pat. 418,230 (1934).
140a. P. R. Michael, W. J. Adams, A. F. Werner, and O. Hicks, Environ. Toxicol Chem., 3,
377389 (1984).
141. Modern Plastics Encyclopedia 1968, McGraw-Hill, New York, Sept. 1967. Published
annually.
142. W. R. Moore, An Introduction of Polymer Chemistry, University of London Press, London,
1963, p. 104.
143. W. R. Moore and R. J. Hutchinson, J. Appl. Polym. Sci., 8, 2619 (1964).
144. H. S. Olcott, J. Am. Oil Chemist Soc., 35, 597 (1958).
145. J. H. Orem and J. K. Sears, J. Vinyl Technol., 1(2), 7982 (June 1979).

Page 552
146. R. J. Pace and A. Datner, J. Polym: Sci., Polym. Phys., 17, 437451 (1979).
147. T. C. Patton and L. J. Jubanowsky (to the Baker Castor Oil Co.), U.S. Pat. 2,580,460
(1952).
148. S. A. Patwardham and K. C. Das, in Controlled-Release Technology: Bioengineering
Aspects, K. G. Das, ed., Wiley, New York, 1983, pp. 214.
149. D. H. Paul, 37th Annu. Tech. Conf., Society of Plastics Engineers, New Orleans, May 7
10, 1979; SPE Tech. Pap., 25, 413417 (1979).
150. M. T. Payne and J. A. Cannon, J. Vinyl Technol., 4(3), 105109 (1982).
151. M. Pellen, Br. Pat. 2,256 (Mar. 10, 1856).
152. W. S. Penn, PVC Technology, MacLaren and Sons, London, 1962.
153. S. H. Pinner and B. H. Massy, Br. Plast., 574 (1963).
154. S. H. Pinner, Plastics, 30, 82 (May 1965).
155. S. H. Pinner, Plastics, 30, 220 (June 1965).
156. M. D. Price, Pest Control, 4758 (Oct. 1960).
157. H. M. Quackenbos, Jr., Ind. Eng. Chem., 46, 1335 (1954).
158. I. N. Razinskaya, Vysokomol. Soedin., 6, 427 (1964).
159. I. N. Razinskaya, P. V. Kozlov, and B. P. Shtarkman, Polym. Sci. USSR (English trans.),
6, 570 (1964).
160. I. N. Razinskaya and B. P. Shtarkman, Polym. Sci. USSR (English trans.), 5. 393 (1963).
161. W. Reddish, J. Polym. Sci., Part C., 14, 123 (1966).
162. M. C. Reed and J. Harding, Ind. Eng. Chem., 41, 675 (1949).
163. M. C. Reed, H. F. Klemm, and E. F. Schulz, Ind. Eng. Chem., 46, 1344 (1954).
164. J. T. Renshaw, Polym. Plast. Technol. Eng., 6(2), 137156 (1976).
165. J. T. Renshaw, Coated Fabrics, 10, 151166 (1980).
166. J. T. Renshaw, J. A. Cannon, and J. D. Hinchen, SPE J., 26(8), 4751 (1970).
167. J. T. Renshaw and J. D. Gabbard, Polym. Plast. Technol. Eng., 10(2), 131144 (1978).

168. L. M. Robeson, Polym. Eng. Sci., 9(4), 277281 (1969).

169. D. V. Rosato, Plast. World, 24(8), 38 (Aug. 1966).
170. R. Sabia and F. R. Eirich, J. Polym. Sci., Part A-2, 1909 (1964).
171. V. W. Saeger and E. S. Tucker, Plast. Eng., 29, 4649 (Aug. 1973).
172. G. Salmon and G. J. Van Amerongen, J. Polym. Sci., 2, 355 (1947).
173. Santoset 1 in Rigid and Flexible Vinyl Products, Monsanto Company, St. Louis, 1966.
174. J. P. Scullin, T. A. Girard, and C. F. Koda, Rubber Plast. Age, 46, 267 (1965).
175. J. K. Sears, Prepr., Am. Chem. Soc. Div. Paint Plast. Print. Ink Chem., 18(1), 256
(1958).
176. J. K. Sears, Mid-Ohio Valley Section of the Society of Plastics Engineers, Marietta,
Ohio, Apr. 12, 1977.
177. J. K. Sears and J. R. Darby, SPE J., 18, 671 (1962).
178. J. K. Sears and J. R. Darby, The Technology of Plasticizers, Wiley-Interscience, New
York, 1982.
179. J. K. Sears, J. R. Darby, and N. W. Touchette, 12th Annu. Joint Tech. Meet. Sabine Area
AIChE and the Texas-Louisiana Gulf ACS, Orange, Tex., Mar. 12, 1965.

Page 553
180. J. K. Sears, J. T. Renshaw, and T, C. Mathis, Proc, TAPPI Pap. Synthetics Conf.,
Washington, D.C. 9. Sept. 2426, 1979, pp. 189203.
181. J. K. Sears and R. L. Zacher, unpublished Work, Monsanto Co., 1982.
182. M. D. Sefeik, J.Schaefer, F. L. May, D. Raucher and S. M. Dubb, J. Polym. Sci., Polym.
Phys., 21, 10411054 (1983).
183. W. L. Semon (to B. F. Goodrich Co.), U.S. Pat. 1, 929,453 (Oct. 10, 1933).
184. C. J. Sheehan and A. L. Bisio, Rubber Chem. Technol., 39, 149 (1966).
185. A. S. Shevlyakov and M. T. Bryk, Sov. Plast. (English trans.), 15, 45 (1967).
186. W. C. Simms, Mod. Pack., May 1979.
187. H. R. Simonds and C. Ellis (and M. H. Bigelow), Handbook of Plastics, D. Van Nostrand,
New York, 1943, p. 251.
188. P. A. Small, J. Appl. Chem., 3, 71 (1953).
189. E. H. Sorg (to Glenn L. Martin Co.). U.S. Pat. 2,414,399 (Jan. 14, 1947).
190. W. H. Stahl and H. Pessen, Mod. Plast., 31(11), 111 (July 1954).
191. R. C. Steer, J. Appl. Polym. Sci., 10, 1673 (1966).
192. J. Stepek, A. Vymazal and B. Dolezel, Mod. Plast., 40(10), 146 (June 1963).
193. J. Stepek and Z. Vymazal, Mod. Plast., 41(7), 166 (Mar. 1964).
194. F. Stuhlen and E. H. Pommer, Kunststoffe, 73, 3235 (1983).
195. N. Sundgren, G. Bergman and Y. J. Shur, J. Polym. Sci., 22, 12551265 (1978).
196. R. B. Taylor and A. V. Tobolsky, J. Appl. Polym. Sci., 8, 1563 (1964).
197. N. W. Touchette, H. J. Seppala and J. R. Darby, Plast. Technol., 10(7), 33 (1964).
198. K. Ueberreiter, in Plasticization and Plasticizer Processes, R. F. Gould, ed., Adv. Chem.
Ser. 48, American Chemical Society, Washington, D.C., 1965, Chap. 3.
199. A. Utsuo and R. S. Stein, Polym. Lett., 3, 49 (1965).
200. F. H. Verhoek and A. L. Marshal, J. Am. Chem. Soc., 61, 2737 (1939).
201. V. A. Voskresenskii and S. S. Shakirzyanova, Zh. Prikl. Khim., 35, 1145 (1962).

202. A. J. Warner, ASTM Bull., No. 163, 53 (Apr. 1950).

203. L. H. Wartman, SPE J., 20, 254 (1964).
204. L. H. P. Weldon, Plast. Inst. (London) Trans. J., 24, 303 (1956).
205. H. Wells and I. Williamson, Br. Plast., 28, 310 (July 1958).
206. C. J. Wessel, SPE Trans., 4, 193 (1964).
207. R. M. Wilkins, in Controlled Release Technology: Bioengineering Aspects, K. G. Das,
ed., Wiley, New York, 1983, pp. 143178.
208. G. E. Williams, J. Polym. Sci., Polym. Chem., 21, 14911504 (1983).
209. G. E. Williams, British Steel Corp., Port Talbot, West Glamorgan, private communication,
Dec. 15, 1983.
210. J. E. Wilson, Radiation Chemistry of Monomers, Polymers and Plastics, Marcel Dekker,
New York, 1974.
211. De L. E. Winkler (to Shell Development Co.), U.S. Pat. 2,590,059 (Mar. 18, 1952).
212. C. Wippler, J. Polym. Sci., 29, 585594 (1958).

Page 554
213. R. E. Wyant, in Effects of Radiation an Materials and Components, J. F. Kircher and R.
E. Bowman, eds., Reinhold, New York, 1964, Chap. 5.
214. C. C. Yeager, Plast. World, 20, 14 (Dec. 1962).
215. L. T. Zeoli and A. F. Kydonieus, in Controlled Release Technology: Bioengineering
Aspects, K. G. Das, ed., Wiley, New York, 1983, pp. 61120.
216. F. Zimmer (trans. by H. K. Cameron), Nitrocellulose Ester Lacquers, Chapman & Hall,
London, 1934, p. 38.
217. B. Zitnansky and N. Lehka, Chem. Prum., 14(6), 313 (1964).

Page 555

10 PVC Blends, Blocks, Grafts, and Interpenetrating Polymer Networks

JOHN A. MANSON
Lehigh University Bethlehem, Pennsylvania

I. INTRODUCTION

555

II. PRINCIPLES OF POLYMER-POLYMER SYSTEMS

556

A. Structural Types

556

B. Miscibility in Polymers

557

C. Effects of Miscibility and Composition on Properties

561

III. PVC BLENDS, BLOCKS, GRAFTS, AND INTERPENETRATING POLYMER

571
NETWORKS
A. PVC Blends

571

B. Block Copolymers

586

C. Graft Copolymers

588

D. Interpenetrating Polymer Networks

601

REFERENCES

604

1. INTRODUCTION
The art of combining different materials in order to improve properties constitutes one of the
most ancient technologies known to civilization [1]. By the middle of this century, soon after the
development of synthetic polymers, this old art had been applied to the preparation of
copolymers, graft copolymers, and polyblends [2, 3]. Because of its low cost and versatility,
polyvinyl chloride (PVC) has received much of this attention as improvements in processability,
heat distortion temperature, impact strength, and service life have been sought. Fortuitously,
the structure and reactivity of the monomer and polymer are such that many modifications by
copolymerization or blending are facile. Indeed, at present, the emphasis of polymer science
and technology is tending to move away from wholly new monomers (except for specialized,

low-volumes uses) and toward modification of existing ones. Thus in the next decade, even
greater attention to both the basic and applied aspects of PVC may be expected.
The purpose of this chapter is threefold: (1) to outline the principles and major techniques for
making blends, blocks, grafts, and interpenetrating

Page 556
polymer networks (IPNs), with emphasis on PVC; (2) to discuss the implications of miscibility of
PVC-based systems with respect to properties and behavior; and (3) to provide examples of
the range of products that are now available either commercially or in the laboratory. Thus
parts of earlier discussions [4, 5] are updated, and a conceptual framework developed to aid in
the assessment of existing or future systems.
For the sake of completeness, some material will inevitably overlap somewhat with that of other
chapters. However, major attention is given to PVC-based blocks, grafts, and interpenetrating
polymer networks. In the case of PVC blends, emphasis is on the basic aspects of miscibility
as it relates to performance, and on less conventional modifiers; more detailed discussion of
conventional copolymers and modifiers at concentrations up to about 25% will be found in other
chapters and particularly in Chapter 4, and in Volume 2, Chapter 5.

II. PRINCIPLES OF POLYMER-POLYMER SYSTEMS

A. Structural Types
The principal methods of mixing two types of polymer together include blending (or
polyblending), graft copolymerization, block copolymerization, and synthesis to achieve the
interpenetration of networks [613]. In many cases, the products are linear or branched; in
others, cross-linked structures may be involved. Figure 1 illustrates limiting cases of these
principal types; variations will be indicated in the appropriate sections. It is true that formal
distinctions sometimes become ambiguous. For example, some blends and some
interpenetrating polymer networks (IPNs) may contain deliberate or adventitious grafting, and
some graft copolymers may exhibit network interpenetration. Also, the products of a standard
copolymerization may constitute a type of block copolymer if the sequences of one type or the
other are sufficiently long. For the purpose of this chapter, differentiation is based on the
apparent intent of the process involved; however, where appropriate, ambiguities in
classification will be pointed out.
While the properties and applications of all such systems must depend on the composition,
other factors are equally important: the state of molecular dispersion, and hence the degree of
phase separation, the morphology, and the adhesion between phases. In fact, it is useful to
consider phase-separated

FIGURE 1 Schema illustrating simple combinations of two polymers (--- and ) (a) polymer
blend without grafting or cross-linking; (b) simple graft copolymer; (c) block copolymer; (d)
classical IPN without significant phase separation.

Page 557
systems as composites [6, Chap. 1; 7]. Where two phases exist, emphasis is given to random
arrangements, although ordered systems may form spontaneously in some cases.
Because of the major role played by morphology, an understanding of the implications of the
thermodynamics and kinetics of miscibility (mutual solubility) is essential from both fundamental
and technological points of view. In this regard, a semantic question must be dealt with: The
term compatibility is still often used to mean miscibility. As applied to materials technology,
compatibility is a more general term best used to describe whether or not two materials can
be combined together advantageously [5, p. 2]. Strictly speaking, then, two miscible polymers
might in fact be incompatible with respect to a given use; in this chapter the term miscibility
will therefore be used to imply mixing and mutual solubility.
B. Miscibility in Polymers
1. General Aspects: Until recently, the literature on polymer blends was dominated by the
expectation that in general polymers tend to be immiscible (or incompatible) because of the
low entropy change (ASM) associated with the mixing of high-molecular-weight polymers
coupled with an often zero or endothermic heat of mixing (Hm).
Gm = Hm-TSm
(1)
Thus it seemed unlikely to realize the negative free energy of mixing (Gm) required for
miscibility or for phase stability [6, 11, 12]. The idea that like dissolves like, exemplified and
quantified by the solubility parameter concept, has also been widely accepted [6, 12]. Indeed,
many polymers are in fact immiscible on the molecular scale (say, a scale of about 10 nm).
Unfortunately, phase separation acquired a negative connotation (perhaps due in part to use of
the term incompatibility), and the development of useful polymer-polymer systems may well
have been thereby impeded. Whereas many two-phase systems do in fact exhibit poor
properties, especially if interfacial adhesion is low, many phase-separated or immiscible
combinations are extremely useful because of, not in spite of, their two-phase nature [11, 12].
For example, the existence of crystallinity in one phase, or the presence of a rubber phase in a
glassy matrix, may lead to useful properties.
It is now clear that one must deal with a spectrum of miscibility ranging between the extreme
cases of a single-phase, homogeneous mixture and a heterogeneous, immiscible system, and
that useful properties may be obtained deliberately at various points along the scale of
miscibility. In fact, partial miscibility (e.g., in the cases of one component in a block or graft
copolymer, or of a degree of interfacial mixing in a two-phase system) may be highly desirable
[68, 11, 12]. Interestingly, PVC can exhibit a very wide range of miscibility, depending on the
nature of the second component, and, of course, on whether or not true phase equilibrium has
been reached (a function of process conditions).
2. Thermodynamic Factors: Thus in designing any polymer-polymer system, the degree of
miscibility must be deliberately controlled. It is now recognized that the early pessimistic
conclusion about the general impossibility of obtaining mutual solubility in high polymers has an

increasing number of exceptions. An increasing body of literature has also demonstrated the
limita-

Page 558
tions of the solubility parameter approach to miscibility [11, 12], and the importance of
considering specific interactions (between unlike polymers differing in their solubility
parameters) as a route to negative (endothermic) values of Hm that can make Gm favorable
to mixing. Specifically, it has been shown [14] that such interactions [e.g., hydrogen-bonding or
more generalized electron donor/electron acceptor (Lewis acid-base) interactions] generally
dominate the mixing of two polar species, and that dipole-dipole interactions typically are
negligible. Interaction due to dispersion forces exists, but is not usually strong.
With PVC, several opportunities exist for specific interactions. Perhaps the most obvious is the
ability of the a-hydrogen in PVC (an H-bond donor) to interact with a carbonyl unit in an ester or
amide group [11, 1518]:

Possible interactions involving -hydrogens and non-H-bonded interactions between the C-Cl
and carbonyl groups have also been proposed, and the sometimes conflicting evidence
discussed [11, 12, 1520]. At the same time, steric effects can also be important; thus a typical
PVC is more miscible with syndiotactic than with isotactic polymethyl methacrylate [21].
3. Other Factors: Since the sum of effects of interactive pairs depends on their concentration,
miscibility must depend on the overall ratio of components, and if a graft or block copolymer is
involved, on the composition of that component itself. For example, the glassy components of
acrylonitrile-butadiene-styrene (ABS) and methacrylate-butadiene-styrene (MBS) terpolymers,
as well as butadiene-acrylonitrile copolymers, tend to be miscible with PVC. With a given
composition, the higher the molecular weight, the less favorable is mixing. Paradoxically,
random copolymers often form miscible blends with other polymers even when the
corresponding homopolymers do not; examples include blends of PVC with ethylene-vinyl
acetate and butadiene-acrylonitrile copolymers. It has been suggested that intramolecular
interactions are involved in this kind of behavior [12].
The effects of temperature are complex, and both lower and upper critical solution
temperatures (LCST and UCST, respectively) may be observed (Fig-ure 2). UCST behavior, in
which a homogeneous blend at a given temperature undergoes phase separation on cooling, is
characteristic of endothermic mixing and a positive entropy of mixing. Such behavior is common
in mixtures of low-molecular-weight materials and polymer solutions, and is predicted by
traditional thermodynamic theories [11]. In contrast, TCST behavior, in which components of a
mixture undergo phase separation on heating, is characteristic of negative heats of mixing and a
negative excess entropy [11, 12]. This type of behavior is being seen to an increasing extent in
polymer blends, and is predicted by newer theories [11]; in practice, however, phase
separation may not occur below the decomposition temperature. Variations from the idealized
behavior of Figure 2, e.g., when a crystalline polymer is involved, are reviewed elsewhere [11,
12].
In general, the phase behavior of PVC blends, blocks, grafts, and IPNs will be expected to

reflect the factors discussed above. However, the kinetics of phase separation must also be
considered; depending on process history, a given system may or may not be close to
thermodynamic equilibrium. Also, phase characteristics depend on factors such as molecular
weight and

Page 559

FIGURE 2 Schema showing upper and lower critical solution temperatures in polymer mixtures
(a and b, respectively).
whether or not unlike components are chemically bonded together. Thus with blocks and grafts
of immiscible components, some mixing at the interface is favored [6], while an IPN structure
forces a degree of miscibility [8].
4. Characterization of Miscibility: Many experimental methods have been used to establish
criteria for miscibility. These include evaluation of density, viscosity, mechanical, dilatometric or
dielectric transitions, standard and differential scanning calorimetry, thermo-optical analysis,
visible and electron microscopy, sonic velocity scattering techniques (light, neutrons, and x
rays), ternary solution methods, inverse gas chromatography, and various kinds of
spectroscopy [radioluminescence, nuclear magnetic resonance (NMR), infrared (IR) and
ultraviolet (UV)]. In practice, the most common techniques involve visual observations of clarity
or the determination of transition behavior. Criteria are somewhat arbitrary because each type
of test reflects a. characteristic scale of mixing on the molecular level [11]. For detailed
reviews, see references 6 and 1113; see also Section III.
5. Miscibility of PVC: As mentioned, PVC behaves generally as an electron acceptor, and thus
should interact strongly with electron donors. In fact, it is said that of all polymers studied to
date, PVC is the most miscible with structurally different homopolymers and copolymers [22].
Table 1, based on recent compilations [11, 22, 23], lists some of the many miscible and partially
miscible blends of PVC with a wide range of amorphous and semicrystalline polymers. It should
be noted that observations of miscibility behavior as a function of composition may vary
depending on the type of test and preparative techniques; for more details see reference 22
and recent papers cited below. The prevalence of electron-donating groups such as carbonyl,
ether, nitrile, and sulfonyl groups in the second polymer is obvious; the phenyl ring typically acts
in a similar manner, albeit more weakly. Of course, in the case of a semicrystalline polymer
such as PVC, crystallinity tends to be retained or to develop with time; hence only the
amorphous phase contains a homogeneous mixture of the two polymers. The realization that
such specific interactions are a key to controlling miscibility has led to a high level of current
research interest in the miscibility of PVC with other polymers.

Page 560
Poly (acrylonitrile-co-butadieneAcrylonitrile-styrene component miscible
costyrene) (ABS)
Poly(1, 3-butadiene-coMiscible at 2345 wt % acrylonitrile
acrylonitrile).
Poly(1, 3-butadiene-co-acrylonitrile) Blend with PVC and poly(vinyl chloride-co-vinylidene
/ polyvinylidene chloride
chloride), miscible and semimiscible, respectively [33]
Polybutylene terephthalate-polyAt least partial miscibility
tetrahydrofuran block copolymer
Poly--caprolactone and other polySeveral miscible [15, 3032]
lactones
Miscible to 65 wt % chlorination; multi-phase at 68 wt %
Chlorinated PVC
[33a, b]
Chlorinated poly (ethylene-co-vinyl
Chlorination improves miscibility at 45% vinyl acetate
acetate) (EVA)
Poly (ethylene-co-vinyl acetate)
Miscible at6570 wt % vinyl acetate [2629]
Polyethylene-co-n-butyl urethane Uncertain miscibility
Poly (ethylene-co-vinyl acetate-coSO2improves miscibility
sulfur dioxide)
Poly (ethylene-co-vinyl acetate-coCO improves miscibility
carbon monoxide)
Poly (ethylene-co-ethyl acrylate-coCO improves miscibility
carbon monoxide)
Polyethylene-co-2-ethylhexyl
CO improves miscibility
acrylate-co-carbon monoxide)
Poly (ethylene-co-N, N-dimethyl
Miscible at 7075 wt % PVC
acrylamide)
Partial miscibility; syndiotactic miscible up to 1: 1 molar
Polymethyl methacrylate
ratio
Poly(methyl methacrylate-coMethacrylate-styrene component miscible
butadiene-co-styrene) (MBS)
Poly(methyl methacrylate-co-ethyl
Miscible at 5 wt % ethyl acrylate
acrylate)
Poly(methyl methacrylate-co-butyl
Miscible at 10 wt % acrylate
acrylate)
Poly(-methylstyrene-co-acryloMiscible when terpolymer composition is 34: 8: 58
nitrile-co-methyl methacrylate)

Page 561
Poly (- methylstyrene-co-methacrylo-nitrile) Miscible at 57 wt % methacrylonitrile
Poly (-methylstyrene-co-methacrylo-nitrileMiscible at wt ratio 58: 40: 2
co-ethyl acrylate)
Polyurethanes
[24, 25]
Miscible at 72:28 wt ratio of styrene; poly
Poly (styrene-co-acrylonitrile)
(acrylonitrile) miscible
Poly(vinyl chloride-g-styrene)
Miscible (microheterogeneous)
Poly(vinyl chloride-co-vinyl acetate) nitrile
Miscible
rubber
Source: Refs. 11, 22, and 23; for more complete details, see references cited therein, in the
table, and below.
Recent reports describe fundamental studies of miscibility behavior and properties of blends
with such polymers as EVA copolymers [2629]; a highimpact EVA copolymer [34]; PMMA [35,
36]; acrylonitrile-butadiene copolymers [3739] ; polyvinyl acetate [40, 41]; various rubbers,
including chlorinated types and their grafts [4244]; various polymeric lactones [15, 3032],
with and without other plasticizers [45]; polyester-polyether block copolymers [4648];
polyurethanes [24, 25, 49, 50]; polyadipates [51]; ethylene-propylene-diene terpolymer [52];
ABS [53]; chlorinated PVC [54]; polyethers [24]; copolymers of styrene with maleic anhydride
[55]; and terpolymers of ethylene, vinyl acetate or acrylates, and SO2 or CO [11, 20, 50]. With
several blends, evidence for LCST behavior was obtained [24, 35]. Attention has also been
given to ternary blends of PVC with, for example, polyester-polyether block copolymers with
ABS [46] or polymethyl acrylate [44, 45]; and butadieneacrylonitrile copolymers with
polyvinylidene chloride and vinyl chloride/vinylidene chloride copolymers [33]. In such studies,
many of the techniques discussed above were used to elucidate the degree of miscibility,
including electron microscopy [44, 46, 47], dynamic mechanical spectroscopy [4447], thermal
analysis [40], sonic velocity [41, 5658], creep and stress relaxation [32, 45], infrared
spectroscopy [15, 24, 3032], neutron scattering [43], dielectric [32] and NMR relaxation [34],
viscosity [42], calorimetry [58], and radioluminescence [39].
A number of blends have been of commercial interest for some time, either in their own right or
as additives to improve properties such as processing behavior, and new blends are being
studied and proposed for application. In all such cases, an understanding of the effects of
miscibility as properties is important; major implications are discussed in the following section.
C. Effects of Miscibility and Composition on Properties
1. Morphology: For typical immiscible two-component systems in general, one of the
components is dispersed in a continuous matrix of the other. The

Page 562
dispersed phase may exhibit several shapes: from spheres to rods or fibrils to lamellae,
depending on the rheological conditions, interfacial tension, and molecular weight of the
components in a polymer blend [7, 59, 60]. A similar range of morphology is seen in block
copolymers, depending on the composition and mode of preparation.
Usually, the major component (in terms of volume fraction) forms the continuous phase. An
example is given in Figure 3, where it is seen that the minor-phase rubber component of a PVCpolybutadiene (PBD) blend (immiscible) exists as irregularly shaped dark particles dispersed in
the PVC matrix [61]; PVC/MBS blends behave in a similar fashion. However, if acrylonitrile (AN)
is incorporated into the PBD (yielding NBR) the morphology changes dramatically. At an AN
content of 20%, which corresponds to partial miscibility, the PVC phase is dispersed in a
continuous matrix of rubbery copolymer, even though the latter is still the minor constituent
(Figure 4). At an AN content of 40%, the phase structure is nearly destroyed (Figure 5); while
two phases still exist (on a scale of about 10 nm), they may well be co-continuous.
To a first approximation, the effects of composition and relative viscosity on phase continuity
have been proposed to follow qualitatively the trend in Figure 6; analogous behavior has been
suggested for IPNs [62]. To tailor the nature and scale of phase domains by sophisticated
blend compounding requires the careful use of intensive mixing devices (such as twin-screw
units) as well as attention to the melt viscosities of the components [33a, b; 54, 59, 60, 62].
Indeed, the subject of mixing and flow behavior in polymer blends requires more fundamental
research.
Interfacial characteristics are important both in processing and in the final product. If the blend
components are immiscible, interfacial tensions in the melt will be high and the ease and degree
of dispersion relatively low. Concomitantly, low interfacial adhesion in such a case usually
implies poor physical and mechanical properties; for an exception with PVC, see Section II.C.3.
If, on the other hand, some miscibility is possible at the interface, dispersion can be achieved
more readily and on a finer scale (see Figures 3 to 5).
When the final properties desired in a PVC blend require a two-phase system, partial miscibility
can enhance processability and final properties [6, Chap. 1]. This goal can also be achieved by
judicious selection of a standard copolymer (such as those discussed above) or a graft or block
copolymer to couple (or compatibilize) the two otherwise immiscible phases together [6,
Chap. 12]. A good example is the use of an MBS terpolymer (a graft polymer of MMA and
styrene on a styrene-butadiene copolymer substrate) as an impact modifier for PVC. With a
sufficient degree of grafting and an appropriate molecular weight, melt mixing is facilitated and
good impact strengtha property that requires strong interfacial adhesion between a
dispersed rubber and the matrixis obtained [63]. ABS resins serve a similar function.
2. Physical and Mechanical Properties: In a multicomponent or composite system, a property
PC depends on an average of the properties of the constituents, usually weighted in terms of
volume fraction [7, 22, 64, 65]:

The precise form of the equation depends on the system, on the interactions between the
components, and on the miscibility. This section discusses typical effects of concentration on
physical and mechanical behavior of miscible, semi-

Page 563

FIGURE 3 Electron micrograph of a thin section of a PVC/PBD blend (100: 15) showing the
dark rubber phase dispersed in the light PVC matrix. (From Ref. 61.)

FIGURE 4 Electron micrograph of a thin section of a PVC/NBR blend (100: 15); the NBR
contains 20% acrylonitrile. Note how the minor NBR constituent (dark phase) in this
semimiscible system appears to form a continuous network around the PVC. (From Ref. 61.)

Page 564

FIGURE 5 Electron micrograph of a thin section of a PVC/NBR blend (100/ 15); the NBR
contains 40% acrylonitrile. Note the fine dispersion and the nearly co-continuous networks in
this microheterogeneous semimiscible system. (From Ref. 61.)

FIGURE 6 Schematic representation of effect of composition and viscosity ratio on phase

continuity in a polymer blend. (After Ref. 12.)

Page 565
miscible, and immiscible polymer-polymer systems with emphasis on PVC as one component.
Transport behavior is discussed in Section III.A.5.
Miscible Mixtures: In this case A and B give miscible systems that are homogeneous on a fine
scale (say, about 10 nm). With some exceptions (Section II.3) the corresponding
homopolymers in a copolymer must themselves be miscible, or, if not, the reactivity ratios and
copolymerization conditions must be such that long sequences of one monomer unit will be
avoided. Usually, such mixtures will be transparent to the eye if A and B are essentially
amorphous, only one phase will be seen in the electron microscope (at a resolution of about 10
nm), and only one sharp Tg will be observed [6, 11, 21].
With such homogeneous systems equation (1) can be written as the following semiempirical
equation [22, 64]:
where K is an interaction parameter. When K is positive or negative, the general curves for PC
as a function of composition will in principle fall above and below, respectively, the linear rules
of mixtures corresponding to K = 0 (Figure 7). Positive and negative values of K correspond to
synergistic and nonsynergistic (or antagonistic) behavior, respectively. Of course, in practice
shapes may vary somewhat from the idealized ones shown in Figure 7.
The linear rule of mixtures applies to permeability in miscible PVC blends (Section III.B.5).
However, if the interphase attraction is strong, the density and modulus may pass through a
maximum as in curve (a) of Figure 7 [22]; a similar situation may arise if a monomer fills up the
free volume of a polymer, as in some grafts and IPNs (see Section III.C.2 and III.D.2). As
shown in Figure 8, the Tg will be intermediate between the Tg values of the constituents, and
will often follow one of the standard equations for mixtures [22], for example:
(For special cases, see Section III.A.1.)
Deviations from the linear rule of mixtures have been observed for the melt viscosity of a series
of blends of PVC and chlorinated PVC with ABS and with styrene (S)/acrylonitrile (AN)/-methyl
styrene (AMS) terpolymers [66], and the deviation from linearity has been related to the degree
of miscibility. As shown in Figure 9, the melt viscosity followed the type (c) behavior of Figure 7;
the greater the miscibility, the greater the deviation on the horizontal scale. It should be noted
that miscibility of the blends was determined by both thermodynamic and kinetic factors. Thus a
mismatch in melt viscosity may result in partial mixing even with inherently miscible pairs.
However, from a practical point of view, the effect of a given variation in miscibility on properties
is the same regardless of whether the variation arises from thermodynamic or kinetic factors.
Measurements of solution viscosity have been used in a similar manner to assess miscibility
[41].
Maxima in tensile strength have been observed with several systems: a blend with a copolymer

of AMS with methacrylonitrile [67] (Figure 10), and blends with the AMS terpolymers mentioned
(Figure 11). Evidently, the higher inherent ductility of the PVC (with respect to the AMS
copolymer) is manifested by an increase in the strain-to-break of the blend, thus delaying
fracture until higher-than-expected stresses are reached. Figure 11 also reveals

Page 566

FIGURE 7 Idealized curves for a property PC, of a blend of two miscible polymers as a
function of composition. Curves (a) and (c) represent synergistic and nonsynergistic
(antagonistic) behavior, respectively; curve (b) corresponds to a linear rule of mixtures.
significant differences in the deviations from linear behavior. Even though all systems were
nominally miscible, some were more miscible than others; these differences undoubtedly reflect
differences in the intensity of specific interactions. Thus ultimate mechanical properties, in which
large deformations are involved, can probably be used to detect subtle differences in such
intensities that might not show up in other tests. In contrast to the case of tensile strength, a
maximum impact strength appears to require limited miscibility (see below).
Immiscible Mixtures: Still more complex behavior is exhibited by wholly immiscible systems,
depending on the property concerned. If interfacial adhesion is weak, mechanical properties will
generally be poor [12]. An exception may exist if the voiding that takes place on deformation is
accompanied by significant plastic deformation [67]; for example, 5% polyethylene dispersed in
PVC can increase the impact strength [68]. If interfacial adhesion is strong,

FIGURE 8 Generalized behavior of the dynamic mechanical properties of a miscible blend (a)
and a two-phase immiscible blend (b). -----, Pure components; -----;, mixture. (After Ref. 11.)

Page 567

FIGURE 9 Melt viscosity vs. composition for (a) a miscible blend of chlorinated PVC with a
styrene/acrylonitrile/-methyl styrene terpolymer, and (b) a less miscible blend of PVC with
chlorinated PVC. (From Ref. 66.)
the modulus of a composite system will fall between upper and lower bounds, depending on
which phase is continuous (Figure 12). Note that interfacial adhesion in polymer-polymer blends
usually requires a degree of inherent miscibility at the interface, or the use of a block or graft
copolymer. Thus, although systems in which rubbery phases are dispersed in a glassy matrix of
PVC (or other polymers) often show peaks in impact strength at intermediate compositions,
such combinations are best discussed as semimiscible systems (see below).
Many of the equations (e.g., equations of Kerner [69] and Takayanagi [70]) proposed to
account for such behavior can be expressed in the following form [7, 22, 71:

Page 568

FIGURE 10 Typical stress-strain behavior for a miscible mixture of a brittle polymer (such as a
copolymer of -methyl styrene with methacrylonitrile) [67] (A) with a ductile polymer (B) (such
as PVC). (After Ref. 11.)

FIGURE 11 Tensile strength vs. composition for (a) a miscible blend of chlorinated PVC with a
styrene/ acrylonitrile / -methylstyrene terpolymer, (b) a miscible blend of PVC with the same
terpolymer, and (c) a less miscible blend of PVC with the same terpolymer incorporating 5%
butadiene. (From Ref. 66.)

Page 569

FIGURE 12 Relative modulus EPER of glassy polymer(P)/rubber(R) blends vs. composition.

Curves A and B correspond to Kerner's equation, with the glassy and rubber phases,
respectively, continuous. Curve C represents a series of homogeneous copolymers
corresponding in composition to the heterogeneous blends. Curve D represents the behavior of
a system in which A and B form two continuous phases. (After Ref. 71.)
where PA is the property of the continuous matrix, A (>0 and <) depends on the shape,
orientation, and interfacial character of the dispersed phase, B depends on the ratio PA/PB and
on A, and is a reduced concentration term that is a function of the maximum packing fraction
of the dispersed phase. This semiempirical mixing rule is obeyed by many physical, but not
ultimate properties; a schematic diagram is given in Figure 12 for a blend of a glassy with a
rubbery polymer. Intermediate behavior results if, as often happens, phase inversion takes
place as the composition changes, or if the phases exhibit dual continuity. The latter
phenomenon may occur in polymer blends that exhibit partial miscibility (see Figure 5) or in
IPNs (see Section IV.D); as seen in Figure 12, the behavior is nearly equivalent to that of a
homogeneous copolymer having corresponding compositions.
Whereas the curves of Figure 12 are all continuous, discontinuities occur in immiscible blends
when the complex modulus is determined as a function of temperature (Figure 13). Regardless
of the degree of interfacial adhesion, each component exhibits its characteristic Tg, as reflected
in a significant decrease in storage modulus (G' or E') or a peak in loss tangent (tan ) or loss
modulus (G or E).
Semimiscible Systems: Most polymer-polymer systems of technological interest are
semimiscible in nature. As mentioned above, semimiscibility may arise with inherently miscible
pairs (at equilibrium) for kinetic reasons, that is, if mixing has not been sufficiently intensive.
Alternative sources of semimiscibility include marginal miscibility per se, or the presence of a
suitable IPN system or block or graft copolymer.
With such mixtures, a range of viscoelastic behavior may be encountered, depending on the
scale of the heterogeneity. This is so not only with polymer blends but also with plasticized
polymers and copolymers in which long sequences of one component are immiscible with
sequences of the other. Many plasticized polymers have a single Tg as with a single-phase
system, but the transition is broadened due to a broadening of the whole relaxation spectrum;

Page 570

FIGURE 13 Typical curves of storage shear modulus (G) and loss tangent (tan ) for polymer
blends. A, Miscible; B, partially miscible, with separate phases readily detectable; C, partially
miscible (microheterogeneous), with separate phases not large enough to exhibit distinctive
behavior; D, immiscible. (After Ref. 11.)
heterogeneous copolymers of PVC with acrylates behave in a similar way [65]. Toward the
other end of the heterogeneity scale (e.g., in a partially miscible polymer blend) two situations
may exist. If the phase sizes are small but phases clearly evident, the behavior may be
intermediate between that of a miscible and an immiscible blend (Figure 13, curve B); two Tg
values may be detected, but they are usually closer together and often broadened and
suppressed in intensity. On the other hand, if the two phases are sufficiently small to be below
the limit of sensitivity for the test method used, only a very broad peak is observed in the loss
parameter and no discontinuities appear in the storage modulus. Such a system is said to be
microheterogeneous (cf. the case of the heterogeneous copolymer mentioned earlier). All this
behavior has been noted in PVC/NBR blends [72]. Thus semimiscible systems are
characterized by a continuum of viscoelastic response from a single broad transition to
transitions that are shifted and broadened on the temperature scale but still more or less
discrete.
Many of the common additives for PVC are semimiscible, at least in certain ranges of
composition, including the following [11] (see Section II.B.5): ABS, acrylonitrile-butadiene
copolymers (NBR), methacrylate-butadiene-styrene terpolymers (MBS), ethylene/vinyl acetate
copolymers (EVA), EVA/ VC graft copolymers, vinyl acetate/VC, some polyurethanes, and
chlorinated polyethylene (CPE). Requirements for particular purposes are discussed in Section
III.A.
A classic example of the desirability of partial miscibility is the effect of composition of NBR on
impact strength (Figure 14). As the concentration of AN was increased, the miscibility increased
(as judged by the transition behavior), and impact strength increased until for concentrations of
AN somewhat above and below 20% no break was obtained. In this range, the rubbery phase
was discrete and well bonded to the PVC phase (see Figure 4). However, at higher
concentrations, and correspondingly higher degrees of miscibility, the phase sizes decreased
until a microheterogeneous product was obtained at 40% AN content (see Figure 5). At this
composition, the rubbery phase was evidently less effective in generating enough energydissipating processes to inhibit fracture (see Section III.A.4).

Page 571

FIGURE 14 Effect of AN content in NBR on Charpy impact strength of PVC/ NBR blends.
Degree of miscibility increases with AN content; at 40% AN, system is microheterogeneous and
nearly miscible. No break was obtained at intermediate compositions. (After Ref. 61.)

III. PVC BLENDS, BLOCKS, GRAFTS, AND INTERPENETRATING POLYMER

NETWORKS
A. PVC Blends
Blending of PVC with other polymers is done for two reasons: to improve the properties of the
PVC, and, less often, to improve the properties of the other polymer. Usually, intensive melt
mixing is used, although the blending of solutions or of latexes is also practiced, especially in
laboratory studies. Master hatching is commonly practiced, in order to maintain consistent
properties. If blending is achieved on a sufficiently intimate scale, a type of interpenetrating
polymer network (IPN) may be obtained. Some occasional grafting or cross-linking may also
take place. In all cases, the actual state of miscibility obtained is a function of both inherent
thermodynamic miscibility and the kinetics of mixing.
For reasons discussed, most PVC blends of technological interest are semimiscible or miscible.
Whereas miscible blends are of most interest as polymeric plasticizers or for raising the heat
distortion temperature (HDT), partial miscibility is required for most other purposes. It should
also be noted that a blend component may be a homopolymer, copolymer (standard, graft, or
block), or a blend itself.
While the more common PVC blends (with concentrations of the modifier up to 25%) are
discussed in more detail elsewhere in these volumes (see

Page 572
a Impact strength or toughness 510 EVA, PBA (each may be grafted) 515 AMBS, MBS,
ABS, MBA 1020 CPE, NBR EPDM (may be grafted), EPR, HIPMMA, ABS + plasticizer,
PIB Permanent plasticization Often 4050% 32% DOP EVA, NBR, PCL, PBT-g-PTHF, PU,
PES, PCI-g-PS, EVA-CO Processability 110 PMMA, AMA, AMS/AN, MMA/EA, SAN To 30
VC/VA (thermoforming) PBA (for de-adhesion) (may be grafted) 712 SAN (foam extrusion)
To raise heat distortion temperature To 50 SAN, SAMB Chemical resistance 3050 NBR
Weathering 520 EVA (sometimes grafted), CPE, PA aABS, acrylonitrile-butadiene-styrene;
AMA, acrylate-methacrylate; AMS, -methyl styrene; AMBS, acrylonitrile/MBS; CPE,
chlorinated polyethylene; EPR, ethylene-propylene rubber; EVA, ethylene/vinyl acetate
copolymer; EVA-CO, ethylene/vinyl acetate/CO terpolymer; EPDM, ethylene-propylene-diene
terpolymer; HI-PMMAI, rubber-toughened PMMA; PBA, polybutyl acrylate; MBA,
methacrylate/butyl acrylate; MBS, methacrylate-butadiene-styrene; MMA/EA, MMA/ethyl
acrylate copolymer; NBR, acrylonitrile-butadiene copolymer; PBT-b-PTHF, block copolymer of
polybutylene terephthalate with polytetrahydrofuran; PCS, PCL-g-PTHF, PCL-g-PS, and PCLPU, poly--capro-lactone, graft with polytetrahydrofuran, graft with polystyrene, and PCLbased polyurethane; PES, polyester; PMMA, polymethyl methacrylate; SAMB, styreneacrylonitrile-methacrylate-butadiene copolymer; SAN, styreneacrylonitrile copolymer; VC/VA,
vinyl chloride/vinyl acetate copolymer. Source: Refs. 11, 12, and 7378.
Vol. 2, Chapter 5), it is useful to consider the fundamental requirements and phenomena
involved in various applications in the light of the principles of miscibility, and to provide
examples of other, less established blends, and examples of behavior of blends throughout the
whole range of composition.
Typical examples of modifiers for PVC of current technological interest are given in Table 2 [11,
12, 7378]. Many of these are available commercially, either as modifiers per se or as
compounds. Examples of PVC as a modifier for other polymers are also given.
The principles underlying these applications are summarized below; the general requirements
will of course also hold for other PVC systems, such as block and graft copolymers that may
be used in nonblended form.

Page 573
1. Plasticization: The external and internal plasticization of PVC has long been practiced (see
Chapters 4 and 9, and Vol. 2, Chapter 3). As discussed above, when relatively large
proportions of an immiscible comonomer (i.e., immiscible once polymerized) are used, the Tg is
broadened if the copolymerization rate constants are such that a broad range of compositions
(and hence sequence lengths of the second monomer unit) is obtained. However,
copolymerization is not as effective in reducing Tg as the introduction of an equivalent volume
fraction of a monomeric plasticizer, for the Tg of the latter is typically very much lower than that
of the polymer corresponding to a comonomer. Unfortunately, monomeric plasticizers tend to
migrate out of PVC, a problem resulting in deleterious effects, such as embrittlement or
contamination of a foodstuff. Thus much attention has been given to the development of
polymeric plasticizers; these are polymers that have low Tg values and that are miscible with
PVC: in practice, small proportions of a low-molecular-weight or monomeric plasticizer may
also be added to enhance blending. (For reviews, see references 11, 76, and 78; see also
Chapter 9, and Vol. 2, Chapter 3.)
Major features of the blending of plasticizing polymers with PVC to lower Tg and reduce
stiffness to yield flexible materials will now be summarized. The use of small proportions to
improve processability without significant loss in rigidity is also discussed below.
Clearly, the prediction of Tg as a function of composition is important. For this purpose, a
special case (79) of equation (4) has been shown to be useful for several miscible blends
including blends with PVC (80, 81):

where w is the weight fraction and Cp is the specific heat increment at Tg. If TgB/TgA is small
and CpATgA CpBTgB, equation (5) reduces to the well-known Fox equation [82]:

Equations (6) and (7) both describe the experimental results of blends of PVC with polystyrene
and polyphenylene oxide [80]; in this case, the values of Tg and Cp were sufficiently similar
that the limiting equation (7) sufficed. It was also suggested, based on studies with lowmolecular-weight plasticizers, that deviations from equation (6) may well reflect subtle
deviations from complete miscibility [82].
Perhaps the earliest example of a polymeric plasticizer for PVC is acrylonitrile-butadiene
copolymer (NBR) [11, 22]. At AN concentrations between about 10 and 25%, the blend is
semimiscible, while at higher concentrations miscibility increases; the blend is said to be
miscible for PVC/NBR ratios from 4:1 to 1:4 if the AN content is 30% [23]. At high
concentrations (>50%), the blend constitutes a plasticized PVC, while at lower concentrations
the PVC improves the properties of the rubber. Additional plasticizers may also be added.
Extensive studies have been made of the mechanical and viscoelastic behavior of such blends
with PVC as a function of composition and molecular weight [83, 8486]; blends of NBR with

vinyl chloride/vinyl stearate copolymers have also been studied [8486]. As well as greater
freedom

Page 574
a (1b/in.) Brittle point (C) 100:0:80b 95 1625 1300 300 380 -38 100:40:0 93 2800 1750 345
450 -27 100:27:39c 94 2825 1350 385 370 -29 100:13:54b 93 2000 1400 355 480 -32
aGraves test, with grain. bHycar 1452P-50 (powdered). cHycar 1422. Source: Ref. 89.

Page 575
from migration and loss of NBR than with low-molecular-weight plasticizers, the retention of
good abrasion resistance has been cited [87]. Both liquid [88] and powdered NBR have been
used [89, 90]; the use of free-flowing dry blends is said to minimize problems with processing
commonly encountered with polymeric plasticizers. Table 3 gives typical data for several
PVC/NBR blends, both with and without dioctyl phthalate as an additional plasticizer. Such
blends are processed conventionally, and have a long history of diverse applications (e.g., in
shoe components, adhesives, tubing, and films).
Oligomeric polyesters (e.g., polyethylene adipate) have also been used extensively for some
time. Their higher-molecular-weight analogs are reported to be effective as plasticizers, and
hence may well be miscible too; the higher molecular weight aids in reducing extractability.
Recently, much attention has been given to ethylene copolymers and other polyesters. Because
of the ability of the -hydrogen in PVC to form a hydrogen bond with the proton-accepting C=O
group in an ester, miscibility of ester groups with PVC is quite favorable (see Section II.B.2).
At vinyl acetate concentrations between 65 and 70 wt %, ethylene/vinyl acetate (EVA)
copolymers are quite miscible with PVC [77, 91]. To evade some processing problems with the
EVA compositions desired, the incorporation of carbon monoxide or sulfur dioxide (each of
which offers proton-accepting C=O groups) has been suggested [11, 20, 66, 9193]. If carbon
monoxide is present, vinyl acetate can also be replaced with ethyl acrylate; also, the ethylene
content can be increased. At least one of these copolymers is known to be offered
commercially [11]. A comparison of the ability to lower Tg is given in Figure 15.
As shown in Figure 16a, the Tg of PVC is reduced about 60C by the incorporation of a 50:50
blend with an ethylene/ethyl acrylate/carbon monoxide terpolymer (71.8:10.5:17.7). The single
sharp peak in tan indicates a high degree of miscibility; interestingly, the secondary ()
transition of PVC is

FIGURE 15 Effect on Tg of adding (a) EVA copolymers [91], (b) poly--caprolactone [94], and
(c) dioctyl phthalate. (From Ref. 78.)

Page 576

FIGURE 16 Shear modulus and loss tangent vs. temperature for PVC and PVC plasticized with
several polymers: (a) 72:10:18 E/EA/CO terpolymer (---), 50:50 blend of PVC with the
E/EA/CO terpolymer (---), and PVC (. . .); (b) 82:18 ethylene/vinyl/acetate copolymer (---),
50:50 blend of PVC with the EVA copolymer (---), and PVC (. . .). (From Ref. 92.)

Page 577
essentially suppressed. Generally, similar behavior was noted for terpolymers based on vinyl
acetate. In contrast (Figure 16b), blends with ethylene/ethyl acrylate copolymer (82/28)
exhibited separate peaks for both the PVC and the copolymer, as is expected for a two-phase
system.
Poly--caprolactone also serves as a miscible polymeric plasticizer for PVC as well as for other
polymers, although poly--propiolactone is immiscible [30, 94, 95]. However, slow
crystallization is a complicating factor [95, 96]; a small amount of low-molecular-weight
plasticizer may be used to minimize this phenomenon [78]. Ethylene / -caprolactone
copolymers are stated [30] to be useful permanent plasticizers for PVC. In any case, the
relative effectiveness of the plasticizers is in the order of the Tg values for the three groups
presented in Figure 15. Chlorinated polyethylene (% chlorine = 65.2%) and chlorinated EVA (%
Cl = 38%) are also miscible plasticizers [11].
Commercially available polymeric plasticizers have included the following: ethylene/vinyl
acetate/CO terpolymers, poly--caprolactone (also as a graft with polystyrene), NBR,
chlorinated polyethylene, polyurethanes, and polyesters. Chlorinated polyurethane is stated to
be less efficient than typical ethylene copolymers [11] (for discussion of its miscibility, see
references 54 and 97). Thermoplastic polyurethanes based on poly--caprolactone are miscible
with PVC [25], but their ability to lower Tg is limited by the hard urethane blocks. A similar
limitation exists with polybutylene terephthalate/polytetrahydrofuran block copolymers [11, 98,
99]; although excellent properties have been reported for blends with PVC [11, 100], the
economics are less favorable than for other modifiers.
2. Heat Distortion Temperature (HDT): While the Tg of PVC can be increased by
copolymerization with a higher-Tg monomer or by postchlorination, much attention has been
given to the blending of PVC with an essentially miscible higher-Tg polymer. Although the Tg
could also be raised by blending with an immiscible high-Tg polymer, efficiency would require
that the higher-Tg polymer be the continuous phase (see above). Some interfacial miscibility or
compatibilization with a block or graft polymer would also be required to avoid the deleterious
effect of poor interfacial bonding. Two other characteristics are also desirable: similar flow
characteristics during compounding, and freedom from adverse effects on properties such as
toughness, color, and stability. In all cases, the mixing rules given in the preceding section hold,
and the blends are transparent if truly miscible.
Several polymers having Tg values higher than that of PVC are known or believed to be
miscible with PVC. Styrene-acrylonitrile (72:28) copolymers (HDT 95C) are evidently miscible,
and are listed as HDT improvers at concentrations up to 50% [73]. Polymethyl methacrylate
(PMMA) is an interesting example of the importance of tacticity in determining miscibility. Thus,
while isotactic PMMA is quite immiscible with PVC at all concentrations [50], syndiotactic
PMMA (Tg = 115C) is miscible with PVC up to a 1:1 mole ratio. Ordinary (atactic) PMMA
appears to be reasonably miscible over a wide range of compositions at room temperature [23,
101]. Polycarbonate (PC), although not wholly miscible with PVC, does raise the Tg [102].
However, since the Tg values of SAN and PMMA are not much higher than that of PVC, only

small increases in Tg of the blend are possible (about 7C for the case of a 50:50 blend with
SAN). Chlorinated PVC has also been described as a miscible HDT improver [73].
Thus there has been much more interest in polymers based on styrene derivatives. While t-butyl
styrene/acrylonitrile copolymers [103] have been considered, most attention appears to have
been given to copolymers of -

Page 578
methyl styrene (AMS), whose homopolymer has a Tg of 170C, and especially to copolymers
with acrylonitrile (69:31) and acrylonitrile/MMA (70:20:20) [11, 53], which are probably miscible
with PVC [11]. Miscibility of AMS/methacrylonitrile/ethyl acrylate terpolymers has been
reported, and a single Tg of 100C reported for a blend with PVC [104]. The grafting of AMS to
PVC has also been reported [105] (see also Section III.C). In any case, significant increases in
Tg are possible with modest proportions of AMS. Recently, an interesting new blend based on
styrene/maleic anhydride copolymers has been described [106]. Good HDT characteristics
were reported, as well as good impact strengths when an impact modifier was also used.
3. Processing Aids: Problems with the melt processing of PVC include high melt viscosity, high
pseudoplasticity, and a tendency to fracture under stress. The plastification process itself
poses difficulties, for PVC cannot be truly melted and mixed on a molecular scale without
degradation. A domain-type microstructure persists in the melt; particle boundaries can serve
as easy paths for fracture. A diverse group of phenomena result from these tendencies [107
109]. For example, the occurrence of melt fracture limits the draw ratio in thermoforming and
extrusion blow molding, and causes breakup during calendering; in general, the slow relaxation
of the melt leads to a variety of surface defects such as matte and sharkskin effects. For a
more detailed discussion, see Chapter 7, and Vol. 2, Chapter 5.
Recent research [107109] has much improved our knowledge of the mechanisms involved [4].
Evidently, the action of processing aids involves a competition between limiting softening or
plasticization and melt elasticity. The modifier appears to serve as a binder for the PVC
particles which acts as a heat flux (reducing gelation time), smooths out stress concentrations
(homogenizing flow), and contributes extensibility due to its ductility. Whereas the softening
effect is increased by high segmental mobility and decreased by a high molecular weight, the
melt elasticity is increased (and improved) by a high molecular weight. For improved
plastification, a molecular weight of a methyl methacrylate/ethyl acrylate copolymer 50 times
that of the PVC con-cerned has been shown to be optimum [107]. In each case, partial
miscibility is required; the preferred degree of miscibility is higher when melt flow and elasticity
are concerned, and lower when sticking to metal is to be avoided (as in calendering). Thus by
adjusting the degree of miscibility, the range of processing conditions can be extended and
better properties obtained [108, 109].
A variety of polymers have been used as processing aids (Table 2); most are of the same
chemical types as some of the permanent plasticizers, impact modifiers, or HDT improvers, but
are usually used at relatively low concentrations, except in thermoforming and foam extrusion
[73]. The goal is usually to achieve the desired effect without changing the Tg to an
unacceptable degree. Most of the modifiers are based on copolymers of methyl methacrylate
(especially those containing minor proportions of ethyl acrylate), or styrene (e.g., ABS [10],
SAN [11], and MBS [63]); NBR [110], polymethyl methacrylate [73] and copolymers of -methyl
styrene with acrylonitrile [73] can also be used. An advantage of acrylic modifiers is said to be
the ability to confer good gloss in the products [111]. Advantages have also been claimed for
the use of multilayered acrylic graft copolymers in both rigid and flexible formulations [108].
Polyacrylates and vinyl chloride/vinyl acetate copolymers have been useful in reducing adhesion

to calender rolls and enhancing thermoformability, respectively [73].

Page 579
4. Mechanical Properties: Blends of PVC with nitrile rubber appear to constitute the earliest
commercial example of miscible polymer-polymer systems [11]. Both PVC-rich and NBR-rich
blends, the latter often involving plasticized PVC, have been important. With NBR-rich blends
(20 to 40% plasticized PVC) it was reported 40 years ago [112] that many properties could be
improved relative to those of the NBR, including resistance to ultraviolet rays and ozone, and
resistance to tearing, flexure, aging, and solvents. Although these advantages were reported to
be accompanied by some sacrifice in abrasion resistance, compression set, and tensile
strength, the blends have been useful in many applications, such as wire and cable insulation,
gaskets, and printing roll covers. The ozone resistance can be increased still further by the
incorporation of acrylates in the NBR, while the abrasion resistance can be increased by
increasing the molecular weight or by decreasing the nitrile content of the rubber (from 28% to
11%) [113]. Properties such as strength and elongation are also sensitive to the type of PVC
[113] and to the method of blending [114]; contradictory effects of PVC molecular weight have
been reported [113, 114]. Recent blends containing low-to-medium concentrations of PVC have
involved the use of hydroxyl-containing NBR [115] and other soft polymers or rubbers such as
EVA copolymers (both plain [116] and grafted with PVC [117, 118]), blends of ABS with
chlorinated PVC [119], EPR [117], and EPDM [120], and blends of EPDM/PVC grafts with NBR
[121, 122]. Blends with high concentrations of rigid polymers for HDT improvement have been
discussed above; PVC has also been used to toughen phenolics 122a].
At higher concentrations of PVC (50 to 85%), the blends with NBR become essentially
plasticized PVC (see above), while at still higher concentrations (85%) the blends constitute
one class of impact-resistant PVC, at least for AN contents that correspond to semimiscibility
[3, 7, 11, 22, 75, 122, 123]. In sufficient concentration, other rubbers serve as impact modifiers
as well (Table 2), provided that the Tg is sufficiently low that the rubber remains in the rubbery
state under the high strain rate experienced during impact loading, and that good interfacial
adhesion exists. As discussed below, the rubber must be sufficiently miscible to provide high
interfacial strength, but sufficiently immiscible to ensure the existence of a discrete rubber
phase after processing.
Nonrubbery materials such as polyethylene can also increase impact strength in an inherently
ductile but notch-sensitive polymer such as PVC [67, 68]. Also, blends of PVC with
polyethylene yield interesting carbonaceous products when pyrolyzed [124]. Blends of a-methyl
styrene copolymers with ABS are reported to have good HDT behavior as well as impact
strength [124a].
Several studies [125130a] address phase morphology. With EVA and CPE modifiers (but not
polyacrylates, ABS, or MBS), a desirable honeycomb structure can exist in which primary
particles of PVC are bonded together with an interpenetrating modifier phase [129130a]. In
contrast to the case of NBR modifiers, such phase continuity is possible with EVA even at high
concentrations of rubber (50%) [61]. In any case, the microstructure of PVC/rubber blends
does depend on process history [125, 131]; again, grafting may provide a means for balancing
properties [73, 132]. Many PVC blends (e.g., with polyacrylates, chlorinated polyethylene, and
EVA copolymers) can be formulated to achieve generally similar mechanical properties,

whether grafted or not [77]. Unless the economics of a particular application vary due to
variations in the efficiency of a modifier, selection can often be

Page 580
based on specific, and often minor, differences in, for example, processability, transparency,
surface appearance, and weatherability. In this respect, grafting may well be advantageous in
some cases [73, 133], and in general saturated rubbers tend to give better weatherability.
The improvement of response in notched or flawed specimens of PVC to impact loading is
usually accompanied by a lowering of the tensile modulus [equation (3)], and by increased
elongation at break and increased energy-to-break [134]. (For a discussion of the effects of
miscibility on tensile strength, see Section II.C.2.)
With graft-type additives such as MBS and ABS, the glassy grafted portion is essentially
miscible with PVC, leaving the rubbery phase dispersed but adherent to the matrix. However, in
contrast to the case of polystyrene, PVC is sufficiently miscible with many other polymers that
grafting may not always be needed to ensure good interfacial bonding [77, 123], although
judicious grafting can be used, as mentioned, to improve other properties, such as
transparency.
So far, the blends discussed are normally made by melt processing; blending of latex-form
modifiers with suspension PVC may not offer much advantage [77]. Nevertheless, blending has
been reported to be useful with plastisols. For example, advantages of blending vinyl
chloride/vinyl acetate copolymers in plastisol form with vinyl chloride/EVA grafts [135] or with
PVC [136] have been cited.
Mechanisms of Toughening: The most common mechanism of toughening by a discrete rubbery
phase is based on the ability of stress concentrations at the interface to induce a combination
of crazing and shear yielding in the matrix [3, 7, 122]. In this way, strain energy that could
otherwise drive a macroscopic crack is dissipated throughout a large volume of material; also in
some cases, shear bands may hinder the development of cracks. Cavitation at the interface, as
in a polyethylene/PVC blend [68] and in rubber-toughened epoxies [122], or within the rubber
particles, as in MBS-toughened PVC [136], provides another mode of energy dissipation. In the
latter, the rubber effectively shifts the ductile-brittle transition for notched PVC to below room
temperature [134].
Whereas crazing appears to predominate during the deformation of a polyblend having a
relatively brittle matrix such as polystyrene [3] at both slow and fast rates, shear-type
responses (e.g., cold drawing) predominate in rubber-modified PVC [3, 123, 134]. For
example, most of the creep in a 95:5 PVC/ABS blend involves changes in extension (due to
shear) rather than changes in volume (due to cavitation during crazing of the matrix). With this
blend, only about 10% of the creep was associated with crazing, in comparison with 95% for
high-impact polystyrene tested under similar loads. It was suggested that shear bands
generated in the PVC could control the growth of any crazes that might develop. Interestingly,
multiple crazes around rubber particles have been observed in deformed but unbroken rubbermodified PVC [137], but apparently not often in fractured PVC blends [134]. The stress
whitening observed in highly deformed PVC blends must therefore be attributed to another
cause such as mismatches in refractive index between the matrix and rubber that could arise

due to drawing of the PVC [134] or cavitation in the rubber [136].

Effect of Modifier Concentration and Characteristics: The degree of toughening depends on the
concentration of the rubbery phase, other parameters being equivalent. Normally, a minimum
proportion of modifier is needed

Page 581

FIGURE 17 Envelope showing impact strength range for four ABS modifiers in PVC as a
function of concentration. (After Refs. 75 and 138.)
to achieve an increased impact strength. Although most reports are restricted to concentrations
of modifier <50%, results of a few studies over the whole concentration range from 0 to 100%
PVC provide excellent examples of synergism in blends. In such cases [134, 138], peaks are
observed in impact strength at an intermediate concentration; the corresponding blend is
tougher than either constituent (Figure 17). Similar behavior has been reported for a series of
acrylic modifiers in PVC [75, 139]; with both ABS and acrylic modifiers, the position of the
curves depends on the grade, but the shape is similar. As modifier is added to the notchsensitive PVC, the yield stress is lowered with respect to the fracture stress, and the impact
strength rises to very high values in the case of a ductile polymer such as PVC. However,
beyond a certain modifier content, the role of the PVC is progressively diminished as the
composition approaches that of the rubber, which is inherently not as tough as PVC at its best.
Thus there is competition between strengthening of the rubber by PVC and weakening of the
blend at high concentrations of rubber (which probably correspond to a continuous rubber
phase).
As with polystyrene-based systems, the size, nature, and degree of dispersion of the rubbery
phase are important [3]. With MBS-modified PVC [140], a sixfold increase in rubber particle
size resulted in a fourfold increase in impact strength; results of tests with thick versus thin
specimens were consistent with a dominant shear response on impact. In general, however, the
average particle size required for toughening a ductile matrix like PVC is lower than that
required for a brittle matrix [3]. The microstructure of the particles may also play a major role;
thus porous particles of a EVA/PVC (6:92) graft copolymer have been shown to yield improved
processability and greater toughening efficiency [141].
Modifiers differ widely in their ability to increase toughness, in part because of variations in
effective rubber content; maximum overall contents usually range between 10 and 20%. For
example, with the several grades of ABS (at 15% concentration) included in Figure 17, impact
strengths varied approximately threefold. With three polyacrylates (at 10% concentration)
[139], the range was about twofold; in general, less polyacrylate than ABS

Page 582
was required to achieve a given impact strength. A comparison between EPDM and ABS is
also instructive. For an impact strength of 7.5 ft-lb/in., 25% of one of the ABS grades of Figure
17 was required, but only 9.2% of a grafted EPDM [141a]; however, the rubber contents were
5 and 5.3%, respectively.
In making comparisons, two other factors must be considered. First, the Tg of the rubber is
important; the lower the Tg, the greater the effectiveness [7]. Second, the toughness (fracture
energy) of a polymer is higher, the higher its molecular weight (M), although the rate of
increase of toughness with M decreases when M is greater than about 105 [122] because of
the more effective entanglement networks formed. Unfortunately, in a toughened polyblend, the
M of the matrix is often decreased when rubber is incorporated, to compensate for the
increase in viscosity due to the rubber [3]. Thus the effects of M and rubber content may be
confounded, unless comparisons at constant M are possible. In one study of MB S-modified
PVC [142], both Izod impact strength and the fatigue behavior were determined for a series of
molecular weights, with the MBS content varied (for each M) from 0 to 14 phr. With respect to
impact strength, it was shown that maximum toughening required the molecular weight of the
matrix to exceed some minimum value, the critical value being lower the higher the rubber
content. With only 6-phr MBS, the transition from low to high values of impact strength had not
yet occurred at Mw 105.
Fatigue resistance is also vital in applications such as pipes; the overall fatigue life comprises
time spent in the initiation and propagation of a crack [122, 143]. In the study just discussed
[142], MBS increased the resistance to fatigue crack growth (in specimens notched to simulate
a flaw) at lower concentrations than those required for significant improvement in impact
strength. Even at a value of M low enough to give little improvement in impact strength (Figure
18) the fatigue crack growth rate at a constant load range was decreased by an order of
magnitude. The relative effect of modifier was greatest at 6 phr, while the relative effect of M
was greatest at the lowest M; these effects may be related to softening due to the rubber per
se or to adiabiatic heating. Indeed, a high-M MBS-modified PVC (14-phr MBS) was shown to
be superior in fatigue crack growth behavior to a commercial polycarbonate.
Impact and fatigue data have also been reported for blends of PVC with composite acrylic/PVC
latex particles made in a two-stage process by polymerizing vinyl chloride onto preformed
acrylic particles [144]. While the improvement in Charpy impact strength was at most threefold,
much greater improvements in FCP resistance were noted. However, improvements in fatigue
crack resistance may not necessarily be reflected in the overall fatigue life; recent preliminary
research on unnotched specimens of MBS-modified PVC indicates that the rubber lowers the
number of cycles spent in initiation, perhaps because of the lower yield stress [145]. Another
core-shell system that exhibits excellent fatigue crack resistance has been based on the
polymerization of vinyl chloride onto PVC seed particles [146]. When the core had a higher M
than the shell, the core retained some integrity after processing, and the material exhibited
fatigue resistance (although not impact strength) resembling that of the acrylate-modified PVCs
of reference 146. In contrast, the inverse morphology, in which the high-M shell was
homogenized during processing, was ineffective in toughening.

The effects of physical aging at elevated temperatures below and above the Tg were also
studied with PVC and MBS-modified PVC [147, 148]. When enthalpy relaxation was greatest,
only slight decreases in fatigue crack growth resistance were observed, apparently because
the cycling tended to erase the effects of the conformational changes involved.

Page 583

FIGURE 18 Effects of M and MBS content in PVC blends on fatigue crack growth rate per
cycle (10 Hz). The stress intensity factor K is proportional to the range in applied stress ;
numbers on curves represent M x 10-5. Full lines correspond to 6 phr MBS; dashed and dotted
lines correspond to PVC (M = 7 x 104) with and without 14 phr MBS, respectively. (After Ref.
142.)
5. Permeability: Although not often reported for PVC blends, permeability to gases is important
for both practical and fundamental reasons. In fact, gas molecules can serve as very small
molecular probes to elucidate the phase relationships in polymer blends [11]. One useful
equation for weighting the contributions of each component has been proposed [149]:

In equation (8), X2 and Xb represent the fractional contributions of each component to the
continuous phase, and P1 and P2 are the permeabilities of the two phases.
Interestingly, the most thorough study of gas permeation in polymer blends has been made
using blends with PVC [150], and the findings agree with those of other techniques, such as
dynamic mechanical spectroscopy. The first studies involved blends with EVA containing either
45 or 65% vinyl acetate (VA), and that were immiscible and miscible, respectively. At a high
processing temperature (185C) the 65% VA blend gave linear behavior over the whole
concentration range, as expected for a miscible blend. In contrast, at a lower processing
temperature (160C), sigmoidal curves were obtained (Figure 19) for this blend. The curves
resembled those obtained for the immiscible 45% VA blend, with a phase inversion at about
7.5% EVA. This low value is believed to result from incomplete fusion of the PVC particles, with
the EVA phase yielding a continuous phase at higher concentrations [151;

Page 584

FIGURE 19 Permeability coefficients vs. composition for blends of PVC with EVA copolymers.
PVC/EVA (45% VA), milled at 160C: , PVC/EVA (65% VA), milled at 160C: , PVC/EVA
(65% VA), milled at 185C: , . Open and closed symbols for O2 and N2, respectively.
[Reprinted with permission from B. G. Rnby, J. Polym. Sci., Polym. Symp., 51, 89 (1975).
Copyright by John Wiley & Sons, Inc. I.
see also Section III.A.4]. It was also noted that the apparent activation energies for diffusion
passed through a maximum at about 30 wt % EVA and were higher, the higher the miscibility;
also, the densities of the miscible blends were higher than predicted based on additivity. These
observations are consistent with an increased interaction between the components.
Studies were also made of miscible blends with NBR containing from 22 to 42% acrylonitrile
[152]. Linear curves were obtained with O2 and N2, although some evidence for phase
inversion was seen with He and CO2 for the lowest concentrations of acrylonitrile. Again, the
apparent energies of activation paralleled the degree of miscibility, and the densities were
higher than calculated. Hence permeability reflects not only composition, but also morphology
and the thermodynamics of phase interaction.
6. Aging and Weathering: As with other polymers, changes in mechanical and other behavior
occur on the aging of PVC, especially due to physical relaxation, oxidative degradation, and in
the case of PVC, dehydrochlorination. Although the effects of purely physical aging in PVC resin
have been studied

Page 585
extensively [153], physical aging in PVC blends has received little attention; So far, it seems
that at least blends with ABS relax in a manner generally similar to that of the matrix resin [147,
148]. With respect to resistance to high temperatures, some lubricants have beneficial effects
[109], while some stabilizers, such as epoxidized oils, can also be lubricants. The incorporation
of polymers in a blend can also improve thermal stability, by physical protection or by a
scavenging action. As discussed in Section IV.C.4, graft copolymers are often far more stable
than corresponding physical blends in this respect.
Thus polyacrylates and EVA copolymers have long been known to be protective for PVC [154].
A recent study describes the effect of adding ABS and copolymers of methyl methacrylate with
methyl acrylate, butadiene, styrene, and vinyl acetate; ABS was said to be effective in
increasing resistance to UV radiation [155]. Both synergistic and antagonistic effects have been
reported in an extensive study of blends with polyacrylonitrile, polyacrylamide, and several
polyacrylates, and specific interactions described [156]; the nitrile group considerably
accelerated decomposition. The weatherability of PVC is improved by polystyrene, while PVC
improves the weatherability of polybutadiene [156a] and ABS [156b].
The more complex process of weathering also degrades the appearance, strength, and
toughness of PVC and its blends. With blends containing unsaturated groups, ultraviolet light
initiates photooxidation, which results in embrittlement; if the blending polymer is a rubber, it
becomes ineffective as a toughening agent [3, Chap. 10]. With ABS, effects are said to be
more severe than with MBS modifiers [154]. Although embrittlement occurs on the surface, the
cracks formed in the now brittle surface layer propagate readily into the interior [3, Chap. 10].
Such deleterious effects may be masked in impact tests at room temperature, and may
become evident first at low temperatures.
The usual approaches to improving weatherability involve serious disadvantages: carbon black
protects against ultraviolet light but decreases the impact strength, and antioxidants tend to
have limited usefulness over the long term [3, Chap. 10]. For this reason, saturated elastomers
such as polyacrylates have long been used as relatively age resistant toughening agents [3, 4,
73, 157]. In the case of polyacrylates, grafting with PVC may be used to improve miscibility and
transparency [133]. Even these, however, are not as resistant as might be desired; also, the Tg
of a typical acrylic rubber (-24 to -55C) is not as low as that of polybutadiene (-110C) (a lowas-possible Tg being desirable for the highest possible impact strength [3]). Chlorinated
polyethylene (percent chlorine from about 30 to 40%) is also used as an ageresistant modifier
with some sacrifice in Tg and a higher die swell during processing [73]. Ethylene/vinyl acetate
(typically 55:45 to 60:40) copolymers are also frequently used, again with some loss in Tg.
Following extensive, worldwide studies of white-pigmented EVA copolymers exposed to the
environment, functional performance was reported to be maintained even after 5 years of the
most severe exposure; in all locations, performance was better than that of controls [158].
Good weathering behavior (at least in laboratory tests) has also been claimed for other
blending polymers such as SAN/acrylic rubber blends [159]. For recent studies of durability in
blends of PVC with polyurethanes and vinyl chloride copolymers, see references 160 and 161;
with polyester-polyurethane block copolymers, weathering behavior is improved by substituting

hexamethylene diisocyanate for tolylene diisocyanate [11]. A more detailed discussion of

weathering is given in Vol. 2, Chapter 2, and Vol. 3, Chapter 5.

Page 586
7. Applications: Rubbery blends containing <50% PVC are used mainly in wire and cable
insulation, tubing, gaskets, and other applications for rubbers that require resistance to ozone
and hydrocarbons; the presence of PVC also contributes to flame retardance. Many blends
with PVC concentrations between about 50 and 85% are with plasticizing polymers, and are
used wherever a high degree of permanence is required. Plastisol blends can be used for the
manufacture of films and decorative facings. Impact-resistant blends are used in such
applications as films, bottles, electronic and electrical components, appliances, housings and
cabinets, toys, sporting goods, and building products. Again, the PVC can contribute fire
retardance to the blend that is lacking in the other component. With ABS, for example, an
excellent balance of strength, thermal stability, flame retardance and impact strength, HDT, and
chemical resistance can be obtained. Transparent blends with ABS, MBS, or AMBS can be
used indoors, while outdoor applications such as in window frames require the use of saturated
polymers such as EVA copolymers, chlorinated polyethylene, and polyacrylates, which, except
for the case of low concentrations of polyacrylates, are translucent.
B. Block Copolymers
Block copolymers consist of molecules composed of chemically different segments linked
together at the segment ends [7, 9, 162]. The arrangements of the segments, which in turn
constitute sequences of monomer repeat units, can be classified in terms of the following types:
(1) A-B systems, containing only one sequence of each segment type; (2) A-B-A systems,
containing two segments of one type linked to a central. segment of the other; and (3)
systems, containing many alternating segments of A and B.

Clearly, the properties of block copolymers must depend not only on the type involved but also
on the distribution of sequence lengths. Depending on the mode of synthesis, actual products
range from well to poorly defined structures. The best defined structures are obtained by the
use of highly selective polymerization: typically either sequential ring-opening additions based
on anionic (living polymer) or ring-opening methods, or the linkage of segments by classical
step-growth polymerization [9], although some cationic approaches have been used.
Although in principle step-type interchange reactions could be used to synthesize block
systems, in practice a mixture of homopolymers, random copolymers, and poorly defined block
copolymers is obtained. In any case, one may expect the cost of producing controlled block
structures to be generally higher than that for other types of multicomponent polymer systems.
In general, free-radical methods have proved to be more appropriate to the synthesis of graft
rather than block copolymers, mainly because relatively nonselective free radicals tend to graft
to the various sequences and to form significant amounts of homopolymer. It may be noted that
a free-radical copolymerization of monomers having different reactivity ratios can yield
heterogeneous copolymers whose molecules are composed of short blocks that vary

Page 587
considerably in length. Also, although mastication techniques can yield some block copolymers
due to recombination of fragments from two polymers, products are ill defined, contaminated
with homopolymer, and seldom characterized.
Unfortunately, vinyl chloride yields useful polymers only with free-radical polymerization, and
means of incorporating appropriately reactive end groups in PVC molecules are not as obvious
as with step-growth polymers. Thus the synthesis of block copolymers of PVC poses a difficult
challenge; indeed, the literature on this subject is sparse. Nevertheless, several ingenious
approaches have been demonstrated, and products with interesting properties obtained. A
major goal appears to be the combination of PVC with flexibilizing blocks in order to obtain the
benefits of permanent plasticization.
An early approach was based on the thermal decomposition of macroazonitriles to yield
diradicals, which then reacted with vinyl chloride to form A-B-A block copolymers [163]. Thus
isocyanate-terminated oligomeric polyurethanes were reacted with a bishydroxynitrile to form a
polymer containing several azo groups. On heating in the presence of vinyl chloride with the
evolution of nitrogen, the polyurethane (PU) diradicals reacted as follows:

In this way, low-molecular-weight block copolymers containing up to 22% PVC were formed;
the block character was confirmed by the fact that two glass transitions were observed.
A generally similar approach was used later to prepare A-B-A block copolymers in emulsion or
solution; the flexibilizing segment was based on isocyanate-terminated oligomeric
polyphenylene oxide or polytetramethylene oxide [164166]. However the maximum polyether
content (15%) was insufficient to decrease the flex temperature below 20C, and attempts to
use higher-molecular-weight oligomers resulted in excessive grafting on the backbone.
A second approach, also based on the preparation of a prepolymer containing internal freeradical-forming groups, involved the ozonization of polypropylene together with polystyrene or
an acrylic monomer [167169] to form peroxidic species that were able to generate diradicals
in the presence of vinyl chloride. In this case,
block copolymers were formed, although
the blocks were rather poorly defined in length. By varying the acrylic monomer, a range of
miscibility was obtained. Thus, with ethyl and butyl acrylates, the Tg values of the blocks were
shifted inward from those of the homopolymers (see above), in a manner suggesting partial
miscibility. With ethylhexyl and methyl acrylate, values of Tg indicated a high degree of phase
interpenetration, also confirmed by electron microscopy. Not surprisingly, stress-strain behavior
depended on both the morphology and the solvent used for casting the specimen. The
generation of active end groups by the peroxidation of PVC terminated with groups such as
aldehyde or carboxyl, and subsequent polymerization of styrene has been claimed, but yields
were low [170]. Metal complexes of, for example, copper have been reported to initiate the
block polymerization of PVC telomers terminated with CCl4 [171], and living polystyrene
terminated with peroxides is said to be effective in polymerizing vinyl chloride to obtain an AB

block [172].
Attention has also been given to step-growth techniques, especially to obtain
which in principle permit one to obtain higher con-

systems,

Page 588
centrations of the flexibilizing segment. These techniques have been based on the ozonolysis
and oxidation of PVC to obtain low-molecular-weight species with reactive functional groups on
each end [173176]. The carboxy groups thus formed can also be transformed to more
reactive acyl chloride groups, and the difunctional acyl chlorides combined with a difunctional
diol or diamine to yield the desired block [175, 176].
Higher molecular weights have been reported when the precursor diacids are reduced to diols
using, preferably, lithium aluminum hydride [177]; the reduction is also said to yield a more
stable product due to deactivation of residual peroxide groups. The polymeric diols can be
reacted with isocyanateterminated prepolymers (polytetramethylene oxide, poly-caprolactone, and polytetramethylene adipate) to obtain
systems. For 41 to 44% polyol,
flex temperatures of -20, -10, and 22C, respectively, were obtained, corresponding to the
effect of up to 40 phr of dioctyl phthalate; the more rigid the isocyanate used, the higher the
flex temperature. Although the use of poly--caprolactone gave the highest tensile strength (15
to 25 MPa, with an ultimate elongation of 400 to 600%), crystallization of the polylactone
resulted in hardening on storage.
Thus block copolymers of vinyl chloride are certainly technically feasible, although their costeffectiveness as permanently plasticized resins has not been demonstrated. For a review
specifically on PVC, see reference 178.
C. Graft Copolymers
1. General: Graft copolymers are prepared by the free-radical, ionic, cationic, or ring-opening
addition polymerization onto a preexistent polymer molecule. The latter must be activated in one
of several ways: (1) by chain transfer with a free radical or ionic species; (2) by direct
activation of the polymer molecule, (e.g., by irradiation); or (3) by provision with reactive
functional groups that can themselves react with the grafting monomer. A voluminous technical
and patent literature on grafting has developed since discovery of the process about four
decades ago. Several general reviews exist [7, 162, 178180], as well as specific reviews on
PVC graft copolymers [181, 182]. A major goal is to improve properties, especially miscibility,
thermal stability, and toughness.
2. Free-Radical Grafting to PVC:
Mechanochemical Method: As mentioned, a certain degree of mechanochemical grafting may
occur during mastication of two polymers together, due to the formation of free radicals and
subsequent transfer [183]. An example is the hot mixing of PVC with polyethylene [184, 185]
and with polyvinyl acetate and a vinyl chloride/vinyl acetate copolymer [184]; greater thermal
stability was indicated for the polyethylene blend [185]. However, although some grafting
(perhaps enough to modify properties) undoubtedly may occur adventitiously during polymer
blending, the products have usually not been well characterized, and this technique is not
common as a deliberate approach to grafting. Of course, if oxygen is present during mixing,
polymeric peroxides formed can serve as additional sources of free radicals.

Chain Transfer Method: With this approach, conventional free-radical initiators such as
peroxides [186188] or redox systems [189] are used. Grafting from the polymer chain is
believed to occur at radical sites generated when an initiator or other radical abstracts a labile
atom, especially an allylic or tertiary chlorine.

Page 589

Other additives (e.g., TiCl3) have been claimed to cleave C-Cl bonds and thus develop
additional free-radical sites [189a], although a cationic mechanism seems more likely, and the
presence of oxygen can serve as the source of peroxides that can initiate the polymerization of
styrene in the absence of other initiators [190]. Reactions may be conducted in solution, in
suspension, or in emulsion. Irradiation techniques are discussed below, as are polyfunctional
grafting monomers.
Heterogeneous systems are especially convenient, for the products of more or less
conventional suspension or emulsion polymerization can be used directly; alternatively,
preformed resin powders can be suspended in water [191]. Although monomers are generally
used in the liquid state, gaseous monomers have also been successfully grafted to PVC after
swelling the resin [192]; swelling solvents can also facilitate grafting [191]. To enlarge the
surface area available for contact with the monomer, the use of modified polymerizations that
yield porous resins [141] or of an emulsion of PVC can be advantageous. In all cases of
heterogeneous grafting in which a significant degree of grafting is involved, the possibility of
forming an IPN-type grafted structure cannot be excluded (see Section III.D).
Many different vinyl monomers have been grafted to PVC using the techniques discussed,
including butadiene [192], methyl methacrylate [193, 194] butyl acrylate [195],
dimethylaminoethyl acrylate [196], styrene [197], acrylamide [198], and butadiene monoxide
[186]. In at least some cases, the products have been characterized in detail and the presence
of a graft copolymer demonstrated explicitly [193]. The use of comonomers to yield copolymer
side chains on PVC has also been common. Examples include the following mixtures: methyl
methacrylate [187] with ethyl or butyl acrylate [199, 200], and with styrene, acrylonitrile, or
butadiene (added in a second step) [188]; butadiene with acrylonitrile [189] or butyl acrylate
[201, 202]; and acrylonitrile or methacrylonitrile with a vinyl ester of a fatty acid (or an acrylate,
fumarate, or itaconate) and a diolefin [203].
Dehydrochlorination of PVC prior to grafting has received considerable attention as a possible
means of increasing the number of grafting sites by increasing the number of allylic chlorines.
The desired degree of dehydrochlorination can be obtained by carefully heating PVC (e.g., in a
salt solution [204] or in an organic base such as pyridine [205]). Such prior treatment was
reported to be useful in the grafting of styrene, methacrylate, and vinyl acetate to PVC [206],
as well as butyl acrylate [204]. The dehydrochlorinated products have been well characterized
in terms of molecular weight distribution and composition, and other properties, such as thermal
stability [205]. On polymerization of styrene (in solution or bulk) in the presence of the
dehydrochlorinated PVC, mixtures of the expected graft copolymer were obtained, as well as
homopolymer [207]. Dehydrochlorination has also been used with radiolytic and ionic grafting
systems.
The introduction of other specific functional groups that can serve as active sites has been the

subject of several claims. Thus there are several reports of the grafting of a wide range of vinyl
monomers to PVC after treatment of the PVC with nitric acid or P2O5 [208, 209], with
halogens, hydrogen halides, and other halogen compounds [210], or with a peracid [211, 212].
Ozonolysis has also been claimed as a route to grafting (e.g., in suspension

Page 590
[213, 214]); the PVC used may be dehydrochlorinated beforehand [215]. However, the
ozonolysis process is complex, and typically associated with chain scission [216]; also,
ozonolysis has been used deliberately as a route to block copolymers of PVC (see Section
III.B). Hence an ill-defined product seems likely.
Radiation Grafting to PVC: As mentioned above, when a polymer is subjected to UV or ionizing
radiation (e.g., , , or rays), free radicals are generated by the breakage of bonds (C--C,
and, in the case of PVC, C--H, or C--Cl) [217, 218]. For thick specimens, or radiation is
suitable, while use of the less penetrating radiation (as from an electron beam) is limited to
relatively thin specimens such as coatings. In the latter case, however, continuous operation is
feasible (see Section III.C.5). Three general techniques have been used to utilize the polymeric
free radicals as grafting sites [217, 218]: (1) direct irradiation of PVC containing an imbibed
monomer or immersed in a monomer; (2) preirradiation of PVC under vacuum or an inert gas,
followed by immersion in a monomer; and (3) preirradiation of PVC in the presence of air or
oxygen, followed by immersion in a monomer. Although ionic polymerization mechanisms are
possible with some monomers at low temperatures, with typical process conditions and
monomers used with PVC, free-radical mechanisms are generally expected.
With the direct irradiation method, free radicals can be formed on both polymer and monomer,
so that, in general, a mixture of graft copolymer and homopolymer may be expected. Of
course, homopolymer can also be formed by chain transfer of a growing polymer chain with
monomer. Unfortunately, the proportion of homopolymer has seldom been determined.
Since PVC is very susceptible to radiation-induced bond rupture, grafting to PVC should be
relatively favorable, especially in the case of monomers having low sensitivities to radiation. In
any case, the phase relationships of the PVC and the monomer must be important to both the
kinetics of polymerization and the nature of the product; preswelling with monomer gives quite
different results than those with simple immersion in monomer. It should be noted also that the
preswelling technique may, in fact, lead to a kind of interpenetrating polymer network (see
Section III.D).
The second technique depends on the generation of free radicals that remain trapped in the
rigid matrix until the diffusion of monomer permits the initiation of polymerization. With the third
technique, polymeric peroxides formed can be activated by heat or catalysts to form free
radicals capable of initiating polymerization, as indicated below [218]:

Page 591
where X is a labile atom.
In early experiments with radiation grafting to PVC, acrylonitrile was found to work well as a
swelling monomer [219], especially if several successive polymerizations were effected with a
given specimen [219a]. In the latter case, up to 200% monomer (based on PVC) was
incorporated in the polymer, although the proportion of homopolymer was not known. The use
of other monomers was soon reported; polyfunctional esters [220], styrene as an immersing
monomer [221,222], styrene as a swelling monomer [223], and mixtures of styrene with
acrylonitrile [224, 225], vinyl acetate [226], and isobutylene [227]. (Examples of polyfunctional
monomers are discussed separately below.) Radiation grafting (probably resulting in intercrosslinking) was claimed for mixtures of PVC with a butadiene-acrylonitrile copolymer [228]. The
need to characterize the products of such systems was recognized, although difficulties with
extraction techniques were noted [226].
Techniques have also been varied to include the use of monomers in solution [229, 230], the
preirradiation of PVC [230], or the use of peroxydizing conditions [230232] followed by the
polymerization of methyl methacrylate, acrylonitrile, or butadiene (i.e., the use of methods 2
and 3) to yield up to 70% conversion of monomer [230]. Although -radiation grafting of the
larger-volume monomers has not become a commercial success, interest continues in the B
irradiation of films or fibers [232a] and in specialized systems. A recent study describes the
kinetics of grafting with acrylamide [233] and reports that phase separation occurs on heating;
low yields and the involvement of relatively few PVC molecules have also been noted with such
grafting [234]. With thin-layer acrylamide grafts, squalene can be used to control the
penetration of ions [235]. To obtain supports for catalytic metal complexes, PVC has been
grafted with several polymerizable organophosphines [236], while ionic grafts with an
acrylamidoalkyl sulfonic acid have been suggested as ion exchange resins [237]. Vinyl fluoride,
which is said to be relatively miscible with PVC (up to 24% sorption), has also recently been
grafted to PVC using radiation and the effect of temperature studied [238]. In some cases,
photopolymerization using UV light has provided an alternative to or radiation (e.g., with thin
layers [236, 239]); in the latter case, an acrylate epoxy was used as a top coat to protect PVC
[239].
Kinetic studies present several interesting features. In particular, the role of diffusion is
important, and also relevant to other grafting techniques. With unswollen films, the rate of
diffusion of styrene into PVC at room temperature was found to be the rate-determining step,
whereas the more miscible methyl methacrylate systems followed normal free-radical kinetics,
as did the styrene system at higher temperatures [238]. Significant solvent effects were also
noted. Gelation prior to complete conversion was shown to inhibit the polymerization of styrene;
full conversion required higher radiation doses or heating to above Tg after irradiation [223].
Evidence for trapped radicals in incompletely converted systems was also obtained, as well as
for parasitic termination reactions at a high monomer concentration (23%).
3. Free-Radical Grafting of Vinyl Chloride: There has been strong interest in the grafting of vinyl
chloride to a wide variety of polymers. As with the grafting of flexible monomers to rigid PVC, a

major aim has been the attainment of high impact strength in the graft copolymer per se or in
blends of the graft copolymer with PVC. The process may be effected in solution [240, 241],
emulsion [242, 243, 243a], suspension of the finely divided substrate [244246], or an existing
suspension of the substrate [247249]. A two-stage

Page 592
bulk polymerization has also been used in which the substrate is added last [250]. While
polymerization in suspension is usually effected with vinyl chloride in the liquid phase, a partial
pressure technique is sometimes preferred [246, 251]. With this method a mixture of monomer
and initiator is injected into a fluidized bed of porous particles of the substrate (e.g., EVA or
EPR); vaporized monomer serves as the fluidizing gas, and as the reaction proceeds, more
monomer is added, keeping the pressure always below saturation. Freeradical polymerization
using conventional initiators is most common (vinyl chloride being unresponsive to ionic or
coordination-type initiators); however, initiation by radiation is possible, although apparently not
common. It should be noted that the nominally grafted products of such polymerizations are
actually blends of graft copolymer with homopolymer.
Because vinyl chloride is a very active monomer with respect to its ability to react with existing
polymers by chain transfer, a wide range of substrates has been studied, and several products
have been of commercial interest. By grafting vinyl chloride with such polymers, a wide range of
properties can be obtained (e.g., varying degrees of miscibility with PVC). Much effort has
been devoted to substrates that are rubbery and hence candidates as impact modifiers.
Grafting to polyolefins such as polyethylene was reported long ago [240, 241]; more recently,
in the case of polypropylene [252], a grafted product containing about 30% PVC gave much
improved tensile and impact strengths in comparison to the corresponding blend. Because of
the versatility of chlorinated polyolefins, vinyl chloride has also been grafted to chlorinated
polyethylene [253255], polypropylene [253-255], and butyl rubber [256, 257], in addition to
chlorosulfonated polyethylene [253, 259]. By grafting vinyl chloride to chlorinated polyethylene
(CPE) it is possible to obtain properties equivalent to those of a blend with CPE containing
35% Cl, but with a CPE containing only about 24% Cl.
Hydrocarbon rubbers have also been of much interest. Thus the grafting of vinyl chloride to
several such polymers has been described and discussed recently, including EPR [141a, 250,
258, 259] and EPDM [141a, 250, 258, 260, 261]. One process for grafting to ethylenepropylene copolymers claims the use of soluble by-product resin as the substrate [262].
Usually, the ratio of monomer to substrate is fairly high (e.g., 80:20 [260]); at least with EPDM,
only about one graft per molecule has been reported [141a], and the chain length is said to be
less than obtained with methyl methacrylate [258]. With respect to a comparison between EPR
and EPDM, less than 60% monomer was found to be grafted to the former, and more than
60% to the latter [263]. Grafting to butyl rubber [258] and to partially oxidized polybutadiene
[264] has been described; again vinyl chloride was found to give shorter side chains than methyl
methacrylate. Grafts to EPR and EPDM can be used by themselves, or can be covulcanized
with EPR and EPDM to improve resistance to oils [265].
One of the most common grafting substrates is constituted by copolymers of ethylene and vinyl
acetate [249, 259, 266271]. A voluminous patent literature has developed over the years (with
examples often differing in subtle details), and the products have been commercialized. While
polymerization may be conducted with a suspension of EVA dissolved in monomer [247], use of
an inverse system, in which the monomer is swollen into finely divided EVA at a pressure below

saturation (see above), has also been reported [246, 251]; alternatively, the EVA may be used
in the form of a latex [243a]. Such systems are expected to yield IPNs (see Section III.D); by
use of a swelling tech-

Page 593
nique, it has been possible to prepare useful impact modifiers from EVA containing only about
25% vinyl acetate instead of about 45% [132].
In any case, grafting of vinyl chloride to EVA provides a good way to obtain the often-desired
PVC/EVA ratio of 50:50, a composition not readily obtainable by direct blending. Typically,
grafts containing 50% EVA are prepared as master batches, and later diluted with PVC to
yield the desired copolymer content [77]. Many variations are possible depending on the
properties desired, which depend in turn on the composition of the copolymer and on the final
PVC content. Vinyl acetate contents commonly range between about 5 and 80%. At moderate
PVC contents (say, about 50%), the properties are tough and flexible [268], and exhibit useful
properties as a processing aid, while at higher PVC content (say, about 90%) high-impact
PVCs result [5]. Grafting to blends of EVA with PVC and PMMA has also been claimed [272],
and EVA itself may be partially hydrolyzed to permit subsequent solution and cross-linking with
diisocyanates [273]. The stability of aggregates of EVA grafts has been examined by NMR
techniques [274]; for recent studies correlating impact strength with mechanical properties, see
reference 275.
Polyacrylates are also very common substrates [77] and the resulting grafts can be used
instead of blends of an acrylic latex with suspension PVC; as with grafted EVA, some of the
products are available commercially. An example is the polymerization of vinyl chloride in the
presence of poly-n-butyl acrylate or poly(n-butyl acrylate)-g-PMMA [77, 276, 277]. The butyl
acrylate component may be partially cross-linked and interpolymers (made by polymerizing vinyl
chloride alone or with another monomer in the presence of polymethyl methacrylate and
presumably somewhat grafted) are known [77, 277a]. For a review of the effects of processing
on polyacrylate grafts, see reference 125.
Although grafts to PVC modifiers or related materials are the most common, other substrates
have been studied as well. These include vinyl polymers such as ethyl acrylate/ethylene
copolymers [278], and various impactresistant materials [279]. Condensation or step-growth
substrates are less common, but recent reports have involved polycarbonate [280],
polyurdthane [281], polyethylene adipate [249], polyethylene terephthalate (by irradiation)
[282], and starch [283]. Whereas the extent of grafting is often limited, the incorporation of azocontaining side chains in the polycarbonate helped give high grafting yields [280].
Numerous examples of comonomers have been cited in the patent literature. Examples include
monomers such as styrene [280], acrylonitrile [284], vinyl acetate [285], a mixture of vinyl
acetate with n-butyl acrylate [285], vinylidene chloride [286], a mixture of hydroxypropyl
methacrylate with a maleic ester [287], and isobutyl vinyl ether [277a]. Cross-linking monomers
can also be included [288]. By judicious selection of comonomers, solvent applications are
possible [273, 286, 287].
4. Ionic Grafting: In contrast to free-radical techniques, ionic grafting tends to give better
defined copolymer structures [9, 289]. Thus an increasing number of scientific studies have
addressed the ionic grafting of various species onto PVC backbones, although little commercial

interest has appeared.

Anionic Grafting: Much of the research has been stimulated by the development of living
polymer procedures in which a polymeric anion is reacted with PVC through the Cl groups. The
most common technique involves the generation of polystyryl anions (with a counterion derived
from the organometallic initiator used) in a solvent for PVC (e.g., tetrahydrofuran):

Page 594

Both lithium and sodium compounds (e.g., phenyllithium and butylsodium) have been used
successfully. To avoid undesired termination of activity, air and polar compounds such as
moisture must be excluded. Substitution on the Cl atoms is accompanied by
dehydrohalogenation as the principal side reaction [289], which is more favored the greater the
polarity of the solvent; chain rupture can also occur.
Thus extensive studies have been conducted to graft polystyrene side chains onto PVC [289
300], to elucidate details of the mechanism, and to characterize the products (e.g., by gel
permeation chromatography [291, 292, 297]). Grafting efficiency has been noted to vary
inversely with both the molecular weight of the PVC and the temperature [296]. Homopolymer
was not found in an early study [292], but was detected in later studies [296, 297].
Polybutadiene can be grafted by a similar technique, but the polybutadienyl carbanion is less
efficient than the polystyryl carbanion, at least at 30C [289]; however, EPDM is more
susceptible than PVC [294]. Other variations include the grafting of long-chain thiolates [301],
allyl xanthate [302], and epoxy silanes [303]; with the appropriate choice of a long-chain
thiolate, an inherently plasticized PVC is obtained. Electrolytic polymerization is also claimed as
a technique [304].
Cationic Grafting: Cationic grafting to PVC was first studied about 25 years ago using
monomers responsive to Lewis acids, such as AlCl3 or TiCl4 [305]. Styrene and indene were
used as monomers, and more recently, n-vinyl carbazole [306]. However, these initiators tend
to induce dehydrohalogenation and cross-linking, as well as homopolymerization.
A preferred general route to the cationic grafting of several vinyl monomers to PVC has been
based on the use of trialkyl aluminum compounds or alkyl aluminum halides [307]. In effect,
combinations of PVC with these compounds constitute effective coinitiation systems for vinyl
polymerization that tends to be more specific than the Lewis acids used earlier. Under optimum
conditions, the grafted products can be obtained with little degradation of the PVC, and
relatively free from homopolymer, although transfer to monomer with consequent
homopolymerization is a common side reaction, and protonic impurities can also initiate
polymerization [308].
The general mechanism can be written

Page 595
Extensive studies have been conducted with isobutylene as the grafting monomer [308312],
especially with trimethyl or triethyl aluminum, and diethylaluminum chloride as coinitiators.
Copolymerizations can be conducted in suspension [310] or, more commonly, in solution.
Grafting efficiencies of >90, 53, and 5%, respectively, were reported for the three coinitiators
mentioned [311], and up to complete conversion of monomer without significant
homopolymerization is possible [310].
Although about 50% isobutylene has been incorporated into the grafted product, the number of
grafting sites (allylic chlorides) per molecule of a typical commercial PVC is lower than the
minimum number (two to four) ideally preferred for synthesis of an impact-resistant, modified
PVC [313]. To increase the concentration of allylic chlorides, two approaches were used: use
of a vinyl chloride/2-chloropropene copolymer, and controlled dehydrohalogenation of PVC.
With the latter system, BCl3 was used as coinitiator, and the tertiary chlorines that terminated
the polyisobutylene branches were cyclopentadienylated to yield thermally reversible networks.
However, the number of grafts per molecule was low, the average only slightly greater than
one.
Other monomers used in similar polymerizations using alkyl aluminum compounds include
butadiene [314316], ethylene oxide [317], and styrene [318320]; TiC13 [189a] and PCl3
[321] have also been used with styrene. In contrast to the general practice with free-radicalgrafted PVC, many of the studies mentioned developed relationships between synthetic
procedures, reaction mechanism, product composition, and physical properties. In some cases,
cationic grafting can be effected using irradiation (e.g., with vinyl n-butyl ether [322, 323] and methyl styrene [322]). Ziegler-Natta catalysts (although not, strictly speaking, cationic) have
been used in the grafting of propylene [324].
5. Grafting of Polyfunctional Monomers: The cross-linking of linear polymers by thermocatalytic
or irradiation techniques has long been of interest as a means for improving properties such as
dimensional stability and abrasion resistance. At the same time, as discussed above, the
polymerization of a monomer in contact with a polymer has long been a standard technique for
the preparation of graft copolymers. An interesting class of polymer-polymer systems has
resulted from attempts to enhance cross-linking and grafting efficiency by combining the two
approaches, (i.e., by polymerizing polyfunctional monomers that have been compounded with
polymers and other additives such as fillers or plasticizers) [7, 325363]. Indeed, several such
systems are commercially important, notably formulations for wire and cable [325330] and
ion-exchange membranes [331], and a foam based on a cross-linked blend [NBR/PVC/poly(VAco-VC)] has been claimed [332].
PVC has been of particular interest because of the desire to increase the heat distortion
temperature and creep resistance in the rigid and plasticized states, respectively, as well as the
resistance to thermal degradation generally. Although thermocatalytic crosslinking is feasible,
much attention has been given to irradiation techniques; with the advent of electron beam units,
continuous commercial operation is feasible and economical. Whereas irradiation of PVC by
itself is impractical due to excessive dehydrochlorination and low cross-linking efficiency, many

polyfunctional monomers can serve as excellent cross-linking agents, and as temporary

plasticizers as well. Thus the ability of polyfunctional monomers imbibed in PVC to cross-link
and graft rapidly to the substrate, and to enhance the thermal stability of the product, was
recognized very early [333]. Such monomers are also well known to sensitize the grafting of
monovinyl monomers as well [331, 339, 340]. Polyfunctional

Page 596
monomers that have been used successfully as cross-linking sensitizers include diallyl esters
[337, 341], triallyl cyanurate [338, 342, 343] and isocyanurate [344], divinylbenzene [331, 335,
336, 339], diacrylates [335, 336, 341, 345], triacrylates [335, 336, 346], dimethacrylates [334,
335, 336, 346353]; trimethacrylates [339, 347, 348, 354360], and a tetrafunctional
methacrylate [347, 348]. Of these, the acrylic monomers appear to be the most sensitive to
irradiationinduced cross-linking. Related systems have been based on other substrates, such
as an EVA/PVC graft copolymer [359], and a mixture of acrylonitrile and butyl acrylate has
been grafted to a copolymer of diallyl phthalate and vinyl chloride [361].
In general, the thermal stability is improved, or at least not diminished, by the treatment; the
higher the functionality, the higher the level of mechanical properties [338, 348, 362]. For
commercial applications in wire and cable insulation [326330], other additives are used,
including additional stabilizers, nonreactive plasticizers [347, 348, 358], and fillers [335, 336,
353]. Other recent applications are for clear top coats for plastics that can be cured by either
UV light or electron beams [360], and for ion-exchange resins [331]. The products are typically
characterized by good strength, low dielectric constants, and resistance to heat and abrasion.
Whereas acrylic monomers are preferred, the use of blends with silane-grafted rubbery
polymers has also been claimed [363]. For ion-exchange membranes the monomer used is a
mixture of styrene and divinylbenzene (see also Section III.D).
As the complex combination of polymerization, grafting, and cross-linking takes place, the
mixture is gradually transformed from a highly plasticized to a less flexible state. Gelation
usually occurs at relatively low doses and may even begin during blending if the monomer is
unstabilized [350]. Usually, the initial rate is high, and at doses in the range of 1 Mrad, a major
fraction of the ultimate gelation has occurred [355358]. The formation of gel is paralleled by
increases in modulus and strength [336, 348, 355], whereas little change is noted in control
specimens subjected to similar irradiation. At the same time, with a PVC/trimethacrylate system
[355], elongation at break was reduced as gelation proceeded, presumably due to a
combination of increased cross-linking and continued reaction of plasticizing residual monomer,
while a maximum in the energy-to-rupture appeared at 30% gel.
The effects of other processing variables on mechanical behavior have also been studied. In
several systems, the percent gel at a given radiation dose is greater, the higher the functionality
of the monomer [338, 348, 362], and the rate of gelation and maximum gel content attainable at
a reasonable dose is proportional to the monomer concentration [336, 356]. A high
concentration of plasticizer (>10%) reduces the strength and modulus of the systems and
generally increases the ultimate elongation. Complex effects of temperature on mechanical
properties have been observed and related to the nature of the three-dimensional network
evolved [355].
Recent studies have elucidated the sequence of events involving grafting, homopolymerization,
and the cross-linking of the grafts [334, 350, 352, 355358], and essentially confirmed earlier
suggestions [346, 351, 352]. In some cases, results have been correlated with a two-phase
morphology as seen by electron microscopy [351] and manifested in the occurrence of two Tg

values [351, 352]. All the results indicated a state of partial miscibility and probable
interpenetration of the phases. The earlier studies of kinetics and mechanisms have also been
extended recently to include the effects of a nonreactive plasticizer [357, 358].

Page 597
In summary, with the irradiation of polyfunctional monomer/PVC systems, the combination of
grafting to PVC, cross-linking through the grafts, and homopolymerization can yield useful
products whose properties can be varied by varying process variables and by incorporating a
plasticizer or filler. Although the general features are complex, they have been elucidated, at
least for a few model systems, and probably hold with minor modification for related systems,
such as those containing monovinyl diluents or solvents [334].
6. Properties and Applications: As with other multicomponent polymer systems, the properties
and behavior depend on the composition and morphology. Of course, if the raw grafted product
is characterized, the presence of homopolymer must be taken into account. In some cases
homopolymer has been separated from what is presumed to be the true graft copolymer by
extraction, fractional precipitation, or gel permeation chromatography [193, 207, 291, 292,
297]. Unfortunately, separations may not be clean cut if cross-linking occurs or if the product
undergoes phase separation in which the dispersed phase occludes polymer of the continuous
phase to yield a salami morphology, or a phase within a phase [7, pp. 82, 212]. (In fact, the
finely dispersed, occluded homopolymer may play an important role in toughening.) With such a
morphology, well known, for example, in ABS graft terpolymers, conventional extraction may
not remove the occluded polymer; little attention has been given to this possibility in PVC grafts.
Nevertheless, with this reservation in mind, it is useful to consider the characteristics of the
nominal grafted copolymer.
In any case, such experiments have clearly revealed that true grafting has occurred, although
the degree of grafting may thus be less than is sometimes supposed. The percent grafting has
been shown to vary inversely with monomer concentration [207, 319]. This would certainly be
expected for both ionic and free-radical systems due to the increased effect of transfer to
monomer at high monomer concentrations. A similar inverse dependence on temperature has
also been observed [207, 296] and, again, is to be expected. Also, an increase in the
concentration of allylic sites tends to result in an increase in percent grafting [207], although the
number of graft sites per molecule is not often determined, and may in fact be not much more
than one per PVC molecule [313].
As the extent of grafting is increased, the intrinsic viscosity ([]) exhibits interesting behavior,
typically first increasing to a maximum and then decreasing [207, 319, 364]. At least for PVC-gPS, the values of Huggins' k continues to increase throughout, from about 0.3 or 0.4 for pure
PS and PVC, respectively, to about 0.9 for a graft containing 47% PS [207]. An increase in k
implies a decrease in polymer-solvent interaction or an increase in polymer-polymer
interactions. Thus it seems likely that as grafting proceeds, the increased size of the grafted
molecule in solution increases, resulting in a higher value of [] and k (k is known to increase
with increasing molecular weight). However, beyond the peak concentration of grafted polymer,
the molecules can no longer expand in solution as if they were random coils, and may in fact
assume a tighter conformation and smaller size, thus resulting in a lower [] but still higher M. In
fact, branched molecules are known to exhibit similar behavior, and to form microgels [365]. A
reduction in melt viscosity, and hence an increase in processability, is also noted for PVC-gpolyisobutylene [366].

This behavior seems quite consistent with microheterogeneity (see Section II.C.2). Indeed, only
one broad Tg has been reported for PVC-g-PS [364],

Page 598
the breadth of the Tg increased with the percent grafting. At the same time, the solutions were
clear, whereas those of a physical blend were immiscible. Hence grafting can induce a measure
of miscibility between two otherwise immiscible polymers.
Thermal Stability: Increases in thermal stability have often been cited for grafted PVC materials
[201, 207, 295, 317, 364]; indeed, the substitution of phenyl groups on tertiary or allylic chlorine
in PVC is well known to improve the thermal response [316]. Also, in a raw graft, effects due to
the presence of homopolymer must be considered [156]. Whatever the cause, the resistance to
degradation can readily be characterized by direct observation of dehydrohalogenation or color
development, by a variety of thermoanalytical techniques such as thermogravimetric (TGA) or
differential thermal analysis (DTA).
Thus greater resistance to dehydrohalogenation has been observed; for example, for isolated
grafts of PVC with butadiene [315, 317], styrene [207, 319, 364], or isobutylene [317, 319].
The reduced evolution of HCl at a given time may reflect a combination of an induction time (not
always seen) with a reduced rate per se [317] (Figure 20). Although data are not strictly
comparable, styrene and isobutylene appear to be more effective than butadiene (BD). Also,
there is some evidence that PVC-g-PBD grafts prepared by cationic techniques in the absence
of CoCl2 as coinitiator may exhibit autocatalytic degradation as time increases, and may exhibit
poorer color stability. While the latter has been attributed to longer diene sequences associated
with less cross-linking [317], a general reduction in the number of single and conjugated double
bonds is otherwise seen, at least with styrene-grafted PVC [319].
Thermogravimetric analysis indicates a two-stage degradation for both PVC and grafted PVC
[207, 319]: dehydrohalogenation beginning somewhat above 200C, and general degradation
beginning at about 450C. Grafting

FIGURE 20 Dehydrochlorination conversion (x) of PVC as a function of time at 190C (in N2).
Samples A and C were unmodified commercial resins; samples B and D were prepared with
and without, respectively, a cobalt compound as co-catalyst with AIR3Cl; sample E was a
PVC/PIB graft. (After Ref. 317.)

Page 599
tends to increase the initial degradation temperature (IDT), the temperature for 50%
decomposition, and the integrated procedural decomposition temperature (calculated over the
temperature range from 100 to 500C), in comparison to controls [207, 319] and to
corresponding blends [207]. With PVC-g-PIB and PVC-g-PS, DSC studies revealed an increase
in the ultimate decomposition temperature [315, 319]. In some cases (e.g., PVC-g-PBD and
PVC-g-PIB [316]), TGA studies have confirmed a significant increase in IDT; with PVC-g-PS,
the evidence from both TGA and DSC results is conflicting [207, 319]. Enhanced thermal
stability may also be accompanied by greater resistance to radiation damage, as in a
PVC/PMMA graft [366].
Rheological and Mechahical Properties: Significant effects of grafting on stress-strain behavior
have been observed, depending on the nature of the copolymer. Not surprisingly, as the
concentration of a flexible chain such as polyisobutylene is increased to about 50%, the
modulus, yield stress, and tensile strength are reduced by about two orders of magnitude, and
by factors of 6 and 5, respectively. Also consistent with the increasing dominance of the
rubbery component, the ultimate elongation increases by a factor of about 30 [312]. (The
effects cited are for a raw graft containing homopolymer PS.) The melt viscosity is also
decreased, and hence the processability is increased [367]. Blends of the isolated graft with
PVC were found to exhibit a pronounced yield point, followed by stress softening, high
elongation (up to eight times), and stress hardening (Figure 21); possible the PIB phase
delayed the onset of final rupture. For PIB contents below 22%, PVC-g-PIB grafts somewhat
resemble a triblock thermoplastic elastomer (S-BD-S), but exhibit a higher tensile strength
[368],

FIGURE 21 Effect of composition on stress-strain behavior of blends of PVC with

polyisobutylene/PVC grafts. The numbers in parentheses indicate the synthesis temperature
(OC)/rubber content (%). (From Ref. 366.)

Page 600
while at higher PIB contents (>40%) the tensile behavior resembles that of a cured natural
rubber [369].
In contrast, grafted products containing a rigid side chain such as PS have been reported to
exhibit lower ultimate elongations (the value varying inversely with PS content), and less strain
hardening [287319]. This is to be expected, for the PS essentially dilutes the ductile PVC, or
to put it another way, the ductile PVC improves the brittle PS. On the other hand, a synergistic
effect of grafting with styrene has been reported for irradiated compositions containing about
10% polystyrene [223]; the ultimate elongation and flexural strength were increased.
Impact resistance is also of particular interest. The grafting of PS side chains was reported to
reduce the impact strength in proportion to the polystyrene content (e.g., from 1 to 0.6 ft-lb/in.
of notch as the percent polystyrene increases to 23 [364]). However, with some irradiated
polystyrene grafts, improved impact strength in such grafts has been claimed [370]. Of course,
with a glassy side chain, no toughening mechanism exists even if there is phase separation; the
cavitation mechanism discussed in Section III.A.4 is excluded due to the chemical bonds
between the components. In contrast, with plastic-rubber combinations improvements in impact
strength can be achieved, both with PVC as the substrate and with vinyl chloride as the grafting
monomer. If the PVC content is high, the graft itself may constitute a high-impact PVC,
whereas if the PVC content is lower, the graft is usually of interest as a blending resin.
Improvements in impact strength have been claimed for a variety of monomers; for example,
acrylates (alone or with other monomers) having at least four carbons in the side chain [199,
201, 204]. Unfortunately, sometimes, as with polyisobutylene grafts, a significant increase in
impact strength may be obtained only with concentrations of the grafted polymer high enough to
make the material rubbery [371] (Table 4). In such a case the graft must be considered as a
potential modifier. In this respect the grafting of butadiene and acrylonitrile [372] results in a
modifier that requires less rubber for a given level of improvement than is the case with
isobutylene. Indeed, grafts with NBR have been used commercially. For example, good results
with the grafting of butadiene and isoprene in conjunction with methyl methacrylate, a-methyl
styrene, or styrene have been claimed [373] (Table 4).
High impact strengths can also be obtained by the grafting of vinyl chloride to a rubber, such as
EVA [268, 271] EPDM [263], or EPR [262]. While some of the high-PVC-content grafts (say,
90% PVC) have been promoted commercially (e.g., EVA grafts), many such products have
lower PVC contents and are used as impact modifiers or processing aids. Grafts to EVA [246,
247], chlorinated polyethylene [253], and polyacrylates [274, 275] are common commercial
products for such uses.
Other Properties: The improvement of other properties in grafted PVC cable compositions has
been discussed above. Claims are common for the improvement of properties by the grafting of
diverse monomers to PVC for several applications, including antistatic surfaces [374], surface
coatings [235,239], adhesives [375], catalyst supports [236], and antithrombogenic materials
[196].

Other graft copolymers can also be of use in the bonding of PVC in laminates. A commercial
adhesive for the bonding of PVC to NBR is a graft copolymer of methyl methacrylate on natural
rubber. Superior peel strengths are obtained with a 50:50 graft, the methacrylate and the
rubber portions

Page 601
a PHR rubber Percent rubber Relative impact strength 1. PVC-g-P(MMA/B/AMS) 5 10 PVCg-P(MMA/IB/AMS) 5 6.8 PVC-g-P(MMA/AN/B) 5 9.2 2. PVC-g-NBR/PVC 5 2.4 1.0 10
4.5 2.0 20 8.3 4.8 3. PVC-g-polyacrylate/PVC 12 30 4. PVC-g-PEVA 14.6 5. PVC-gPIB 10 1.5 33 1.8 52 No break 6. PVC-g-PIB/PVC 5 0.8 10 1.5 aB, butadiene; IB,
isobutylene; AN, acrylonitrile; PEVA, ethylene/ vinyl acetate copolymer; NBR, acrylonitrilebutadiene copolymer. Source: Refs. 373, 372, 5, 5, 371, and 371 for systems 1 to 6,
respectively.
being miscible with the PVC and the natural rubber substrates, respectively [376].
Grafts of vinyl chloride (alone or combined with other monomers) with a variety of substrates
have also been proposed for a variety of applications, including biochemical materials [249,
270], conductive polymers, [277a], and cable components [255].
D. Interpenetrating Polymer Networks
1. General: Within the last decade commercidl and academic interest in interpenetrating
polymer networks (IPNs) has grown rapidly and new products have appeared on the market. In
this section the deliberate synthesis of PVC-based IPNs is discussed. However, although the
concept of an IPN is relatively new, many old and new laboratory and technological practices
are believed to involve the formation of IPNs, whether recognized

Page 602
or not. Some of these are based on PVC. Such techniques were identified and discussed
separately in the relevant sections above.
Broadly defined, an interpenetrating polymer network is a material consisting of two polymers,
each in the form of a network [7, 8] in which precursor molecules of at least one polymer are
linked together by chemical cross-links or by physical cross-links in the form of crystallites.
Although the networks may in fact be themselves joined together (e.g., by grafting), both can
be visualized as being continuous and interpenetrating on at least some scale. In the presence
of grafting, the IPN topology is said to exist when the deliberately introduced cross-link sites
outnumber the accidentally introduced graft sites. With two inherently miscible polymers, the
interpenetration is considered to occur at the molecular level, whereas if phase separation
occurs, the phases themselves are considered to interpenetrate each other, with molecular
interpenetration occurring at the phase boundaries. For recent reviews, see references 8, 62,
and 377.
2. Synthesis and Structure: Several classes of IPN may be distinguished depending on the
network characteristics (Figure 1). When both polymer I and polymer II are cross-linked, a full
SIN is said to exist, whereas if only one of the two is cross-linked, the product is called a semiIPN. A related type of material is the AB-cross-linked system, in which a single network is
formed from two polymers. Other convenient distinctions may be made based on the method of
synthesis. Thus a sequential IPN is formed if monomer II (including its initiator and cross-linker)
is swollen into polymer I and polymerized in situ. With a segmental synthesis of a semi-IPN, a
semi-IPN of the first or second kind is said to result when the first or second polymer,
respectively, is cross-linked. If, on the other hand, a mutual solution of each monomer is
polymerized by two noninterfering mechanisms, in one overall process step, a simultaneous
interpenetrating network (SIN) is formed. In a third mode of synthesis, two latexes of linear
polymers are mixed, coagulated, and cross-linked to form an interpenetrating elastomeric
network (IEN), while latex IPNs are synthesized by polymerizing a monomer II that has been
swelled into polymer I in the form of emulsion particles. A variation of the latter can be
constituted by the use of a suspension of polymer.
As with other multicomponent polymer systems, the properties of IPNs and related materials
depend on the composition and on the detailed morphology. With a combination df an
elastomeric and a rigid polymer, products typically range from reinforced elastomers to leathery
materials to toughened plastic, depending on which phase is dominant at the test temperature
involved. In a sequential IPN, polymer I tends to be dominant, and its degree of cross-linking
controls the size of the second phase; with an SIN, dominance is determined by the
compositional ratio and rheological conditions.
While many properties resemble those of corresponding blends, blocks, and grafts, a
significant difference is conferred by the mutual interpenetration: a suppression of creep and
flow. Also, if chemical cross-linking is involved, swelling in solvents is possible, but not
dissolution; on the other hand, with a thermoplastic IPN, a network can be dissolved selectively,
leaving behind the second polymer in porous form.

3. PVC-Based Systems: As mentioned above, IPNs of various kinds have undoubtedly been
formed adventitiously in many multicomponent polymer systems [7, 8]. In the case of PVC,
semi-IPNs are possible; the crystal-

Page 603
lites serve as effective cross-links. Examples include cross-linking by polyfunctional monomers,
and any of the heterogeneous grafting techniques involving swelling of a substrate by a
monomer (see Section III.C). However, there have been few explicit and deliberate attempts to
prepare PVC-based IPNs.
The first example appears to have been the synthesis and characterization of several PVC/NBR
latex-based IPNs formed by means of a two-stage emulsion polymerization [377]. After a seed
latex of cross-linked PVC (polymer I) had been prepared, the monomer mixture (to give
polymer II), containing an initiator and a cross-linker, was added and polymerized. The resultant
degree of cross-linking was low enough to permit the direct casting of films from the latex or
compression molding. Most of the NBR was polymerized on or within the seed particles yielding
a typical core-shell morphology, with the NBR content increasing from the center outward; on
film forming or molding, the PVC was observed as a discrete phase dispersed in a matrix of
rubber. Dynamic spectra for a 50:(25:25) system revealed a single broad transition indicative of
partial miscibility or microheterogeneity. The stress-strain behavior generally resembled that of
a reinforced elastomer, although in some cases yielding typical of a toughened polymer was
seen. Thus the typical morphology was deduced to be that of PVC/NBR IPN particles bonded
together with NBR; in the latter case, some fusion of the core material evidently occurred,
providing some gross continuity of the PVC component. The use of shear processing to permit
phase inversion would be of interest in such systems; the PVC phase could then become
continuous.
Although not synthesized in these studies, the reverse core-shell system (i.e., monomer
polymerized on an elastomeric seed latex) was predicted to yield improved impact strength [7,
p. 268]. In fact, evidence indicates that the polymerization of vinyl chloride on acrylic latex
particles involves penetration of the monomer within the particles [145]; when used as an
additive for PVC, the products were shown to improve the impact strength and resistance to
fatigue crack propagation [145, 146]. Some of the heterogeneous grafts on PVC (Section
III.C.3) are probably similar. Also, an interesting EVA-g-PVC copolymer has been prepared
deliberately under IPN-forming conditions using an EVA containing 25 wt % vinyl acetate (a
composition that ordinarily yields poor impact strength when grafted conventionally) [132].
Instead of dissolving the EVA in vinyl chloride, the vinyl chloride was swollen into the EVA prior
to polymerization; with low initial monomer/polymer ratios, monomer was added during
polymerization in order to obtain overall PVC contents up to about 95%. In effect, a semi-IPN
was evidently formed, with the PVC serving as one physically cross-linked phase. The initial
value of the monomer/polymer ratio was found to determine the appearance, morphology, and
gel fraction of the grafted material, and the processability and impact strength of blends with
PVC made to contain 5 wt % EVA. The lower the ratio for a given PVC content, the finer the
scale of dispersion of the EVA, and the greater the gel fraction and transparency. Low ratios
also favored long gelation times and high impact strengths as long as the blending temperature
was selected for optimum dispersion of the EVA component. Best values of impact strength
were noted when the initial ratio was less than 2, and the overall PVC content in the graft was
between about 40 and 70%. Thus the use of the modified IPN technique made it possible to

control both the dispersion of the rubbery phase without sacrificing control of the overall graft
composition.

Page 604
A third type of PVC-based IPN has also been described [378]. In this case, during research on
the recycling of plastic wastes, blends of polystyrene (PS) with PVC (with PS/PVC ratios from
100:30 to 100:50) were combined with either a cross-linked PS or an elastomeric polyacrylate
network (30 to 80 parts per 100 parts of blend). Although incorporation of the PS network did
not significantly improve the impact or ultimate tensile strengths of the blend, incorporation of
the acrylic polymer resulted in considerable improvement in both properties, especially when
the polyacrylate was cross-linked to an optimum extent. In addition, it was noted that the
processing temperature required (120C) was much lower than usual for thermoplastics (180
200C), and hence energy conserving.
It has also been suggested that thermoplastic IPNs may be formed by means of melt blending,
presumably under conditions yielding a very fine phase dispersion [8, Chap. 8]. One example is
a blend of PVC with a crystalline polymer such as a segmental polyester (see Section III.A and
reference 379); the existence of an interpenetrating network component in blends of a phenolic
resin with PVC has also been inferred from electron micrographs [122a].
The possibility of interpenetration in the cross-linking of PVC with polyfunctional monomers has
already been recognized [339, 351], and, in another case involving commercial ion-exchange
membranes, apparently expected [331]. With the in situ polymerization of butyl acrylate imbibed
in PVC, control of miscibility has been demonstrated using a reswelling technique and
knowledge of the PVC-PBA-monomer phase diagram [195].
In any case, the crystallites in PVC (or in a second polymer) can serve as the physical
equivalent of chemical cross-links, so that semi- or full IPNs may be expected, the type
depending on whether or not the other polymer involved forms a physically or chemically crosslinked system. The additional restraints on segmental mobility thus introduced may be expected
to have the beneficial effect of reducing creep and stabilizing the morphology. Further
consideration of IPNs as a route to the variation of properties and of the concept as a means to
understanding the behavior of adventitious IPNs is certainly to be expected.

REFERENCES
1. J. E. Gordon, The New Science of Strong Materials, 2nd ed., Pelican Books, New York,
1976.
2. R. J. Ceresa, in Block and Graft Copolymerization, R. J. Ceresa, ed., Vol. 1, Wiley, New
York, 1973, p. 48.
3. C. B. Bucknall, Toughened Plastics, Applied Science Publishers, London, 1977.
4. C. F. Ryan and R. L. Jalbert, in Encyclopedia of PVC, L. I. Nass, ed., Vol. 2, Marcel
Dekker, New York, 1977, Chap. 12.
5. M. W. Kline and E. N. Skiest, in Encyclopedia of PVC, L. I. Nass, ed., Vol. 1, Marcel

Dekker, New York, 1977, Chap. 4.

6. D. R. Paul and S. Newman, Polymer Blends, 2 vols., Academic Press, New York, 1978.
7. J. A. Manson and L. H. Sperling, Polymer Blends and Composites, Plenum Press, New
York, 1976.
8. L. H. Sperling, Interpenetrating Polymer Networks and Related Materials, Plenum Press,
New York, 1981.

Page 605
9. A. Noshay and J. E. McGrath, Block Copolymers, Academic Press, New York, 1977.
10. D. R. Paul and J. W. Barlow, J. Macromol. Sci.Rev. Macromol. Sci., C18, 109 (1980).
11. O. Olabisi, L. M. Robeson, and M. T. Shaw, Polymer-Polymer Miscibility, Academic
Press, New York, 1979.
12. D. R. Paul, Polym. Mater. Sci. Eng. , 50, 1 (1984).
13. K. Solc, ed., Polymer Compatibility and Incompatibility: Principles and Practices, MMI
Symposium Ser., Vol. 2, Harwood, N.Y., 1982.
14. F. M. Fowkes, ACS Symp. Ser., 99, 69 (1982).
15. D. F. Varnell, E. J. Moskala, P. C. Painter, and M. M. Coleman, Polym. Eng. Sci., 23,
658 (1983).
16. D. Allard and R. E. Prud'homme, J. Appl. Polym. Sci., 27, 559 (1982).
17. D. F. Varnell and M. M. Coleman, Polymer, 22, 1324 (1981).
18. F. Canelosi and M. T. Shaw, Polym. Eng. Sci., 23, 669 (1983).
19. A. Garton, Polym. Eng. Sci., 23, 663 (1983).
20. M. Iskander, C. Tran, L. M. Robeson, and J. E. McGrath, Polym. Eng. Sci., 23, 682
(1983).
21. J. W. Schurer, A. deBoer, and G. Challa, Polymer, 16, 201 (1975).
22. O. Olabisi, in Kirk-Othmer Encyclopedia of Chemical Technology, M. Grayson and D.
Eckroth, eds., Vol. 18, Wiley-Interscience, New York, 1982, p. 443.
23. S. Krause, in Polymer Blends, D. R. Paul and S. Newman, eds., Vol. 1, Academic Press,
New York, 1978, Chap. 2.
24. P. K. Bandyopadhyay and M. T. Shaw, J. Appl. Polym. Sci., 27, 4323 (1982).
25. H. W. Bonk, A. A. Sardonopoli, H. Ulrich, and A. A. R. Sayigh, J. Elastoplast., 3, 157
(1971).
26. M. M. Coleman, E. J. Moskala, P. C. Painter, D. J. Walsh, and S. Rostami, Polymer, 24,
1410 (1983).
27. C. Elmquist and S. E. Svanson, Eur. Polym. J., 12, 559 (1976).

 28. D. Feldman and M. Rusu, Eur. Polym. J., 10, 627 (1974).
29. E. Nolley, D. R. Paul, and J. W. Barlow, J. Appl. Polym. Sci., 23, 623 (1979).
30. J. V. Koleske, in Polymer Blends, D. R. Paul and S. Newman, eds., Vol. 2, Academic
Press, New York, 1978, Chap. 22.
31. R. E. Prud'homme, Polym. Eng. Sci., 22, 1138 (1982).
32. P. Malik and R. E. Prud'homme, Polym. Eng. Sci., 24, 144 (1984).
33. Y.-Y. Wang and S.-A. Chen, Polym. Eng. Sci., 21, 47 (1981).
33a. B. Carmoin, G. Villoutreux, and R. Berlot, J. Macromol. Sci. Phys., B14, 307 (1977).
33b. G. Ajroldi, G. Gatta, P. D. Gugelmeto, R. Rettore, and G. P. Talamini, in
Multicomponent Polymer Systems, R. F. Gould, ed., Adv. Chem. Ser. 99, American Chemical
Society, Washington, D.C., 1971.
34. S. E. Svanson, Eur. Polym. J., 19, 673 (1983).
35. E. J. Vorenkamp and G. Challa, Polym. Commun. 24, 290 (1983).
36. M. Sabliovschi and G. Grigoriu, J. Therm. Anal. 26, 23 (1983).
37. P. D. Trivedi and P. Chaterjee, Proc. 28th Macromol. Symp., 767 (1982).
38. W. Rovatti and E. G. Bobalek, J. Appl. Polym. Sci., 19, 2143 (1975).

Page 606
39. T. Inoue, T. Kobayashi, T. Hashimoto, T. Tanigami, and K. Miyasaka, Polym. Commun.,
25, 148 (1984).
40. S. G. Fischer and J. Chiu, Thermochim. Acta, 65, 9 (1983).
41. Y. P. Singh and R. P. Singh, Eur. Polym. J., 19, 535 (1983).
42. Y. P. Singh and R. P. Singh, Eur. Polym. J., 20, 201 (1984).
43. J. Jelenic, R. G. Kirste, R. C. Oberthur, and S. Schmitt-Strecker, Makromol. Chem.,
185, 129 (1984).
44. N. K. Karfoglou, Angew. Makromol. Chem., 118, 19 (1983).
45. N. Sundgren, G. Bergman, and Y. J. Shur, J. Appl. Polym. Sci., 22, 1255 (1978).
46. D. J. Hourston and I. D. Hughes, J. Appl. Polym. Sci., 26, 3467 (1981).
47. D. J. Hourston and I. D. Hughes, J. Appl. Polym. Sci., 26, 3487 (1981).
48. S. K. Khanna and W. I. Congdon, Polym. Eng. Sci., 23, 627 (1983).
49. J. Piglowski, T. Skrowronski, and B. Masiulanis, Angew. Chem., 84, 129 (1980).
50. J. Piglowski and W. Laskawski, Angew. Chem., 84, 163 (1980).
51. S. H. Goh, D. R. Paul, and J. W. Barlow, J. Appl. Polym. Sci., 27, 1091 (1982).
52. A. Pavan, T. Ricco, and M. Rink, Mater. Sci. Eng., 48, 9 (1981).
53. E. P. Chang, R. Kirsten, and R. Salovey, J. Appl. Polym. Sci., 24, 827 (1979).
54. D. D. Clark, E. A. Collins, and L. W. Kleiner, Polym. Eng. Sci., 22, 698 (1982).
55. L. G. Bourland and A. D. Wambach, J. Vinyl Technol., 5(3), 121 (1983).
56. L. M. Robeson and. J. E. McGrath, Polym. Eng. Sci., 17, 300 (1977).
57. D. J. Hourston and I. D. Hughes, Polymer, 19, 1181 (1978).
58. B. O. Schneier, J. Appl. Polym. Sci., 17, 3175 (1973).
59. H. Van Oene, J. Coll. Interface Sci., 40, 448 (1972); Polymer Blends, D. R. Paul and S.
Newman, eds., Academic Press, New York, 1978, Chap. 7.
60. C. D. Han, Multiphase Flow in Polymer Processing, Academic Press, New York, 1981.

 61. M. Matsuo, Jpn. Plast., 2, 6 (July 1968).

62. L. H. Sperling, Polym. Mater. Sci. Eng., 50, 19 (1984).
63. T. O. Purcell, Jr., in Encyclopedia of Polymer Science and Technology, H. F. Mark and N.
Bikales, eds., Vol. 14, Wiley-Interscience, New York, 1978, p. 319.
64. L. E. Nielsen, Predicting the Properties of Mixtures, Marcel Dekker, New York, 1978.
65. L. E. Nielsen, Mechanical Properties of Polymers and Composites, 2 vols., Marcel
Dekker, New York, 1974.
66. L. W. Kleiner, Abstr. Proc. 3rd Int. Conf. PVC, IUPAC, Cleveland, Ohio, 1980, p. 169.
67. A. F. Yee and R. P. Kambour, in Toughening of Plastics, Plastics and Rubber Institute,
London, 1978, p. 20.1.
68. C. A. Brighton, in Encyclopedia of Polymer Science and Technology, H. F. Mark and N.
Bikales, eds., Vol. 14, Wiley-Interscience, New York, 1971, p. 305.
69. E. H. Kerner, Proc. Phys. Soc., London, 69B, 808 (1956).

Page 607
70. M. Takayanagi, Mem. Fac. Eng., Kyushu Univ. 23, 11 (1963).
71. R. A. Dickie, in Polymer Blends, D. R. Paul and S. Newman, eds., Academic Press, New
York, 1978, Chap. 8.
72. M. Matsuo, Polym. Eng. Sci., 9, 206 (1969); in Encyclopedia of Polymer Science and
Technology, H. F. Mark and N. M. Bikales, eds., Suppl. Vol. 2, 1977, p. 402.
73. P. Frster and M. Herner, Kunstoffe, 69, 146 (1979).
74. J. A. Manson, Pure Appl. Chem., 53, 471 (1981).
75. R. D. Deanin, in Encyclopedia of Polymer Science and Technology, H. F. Mark and N.
Bikales, eds., Wiley-Interscience, New York, 1982, p. 458.
76. J. P. Tordella, J. T. Hyde, B. S. Gorton, and C. F. Hammer, Proc. 33rd Soc. Plast. Eng.
Annu. Tech. Conf., May 58, 1975, p. 135.
77. R. H. Burgess, in Manufacturing and Processing of PVC, R. H. Burgess, ed., Macmillan,
New York, 1982, p. 83.
78. C. F. Hammer, in Polymer Blends, D. R. Paul and S. Newman, eds., Academic Press,
New York, 1978, Chap. 17.
79. P. R. Couchman, Macromolecules, 11, 1156 (1978).
80. D. M. Leisz, L. W. Kleiner, and P. G. Gertenbach, Thermochim. Acta, 35, 51 (1980).
81. J. R. Fried, S.-Y. Lai, L. W. Kleiner, and M. E. Wheeler, Abstr., Proc. 3rd Int. Conf. PVC,
Cleveland, Ohio, 1980, p. 186.
82. T. G. Fox, Bull. Am. Phys. Soc., 1, 123 (1956).
83. H. F. Schwarz and W. S. Edwards, Appl. Polym. Symp., 25, 243 (1974).
84. E. F. Jordan, Jr., B. Artymyshyn, G. R. Riser, and A. N. Wrigley, J. Appl. Polym. Sci.,
20, 2715 (1976).
85. E. F. Jordan, Jr., B. Artymyshyn, G. R. Riser, and A. N. Wrigley, J. Appl. Polym. Sci.,
20, 2737 (1976).
86. R. F. Jordan, Jr., B. Artymyshyn, and G. R. Riser, J. Appl. Polym. Sci., 20, 2757 (1976).
87. R. D. Deanin, R. O. Normandin, and I. T. Patel, Prepr., Am. Chem. Soc. Div. Org. Coat.
Plast. Chem., 36(1), 304 (1976).

 88. V. R. Landi, Appl. Polym. Symp., 25, 223 (1974).

89. M. E. Woods and D. G. Frazer, SPE Tech. Pap., 32, 426 (1974).
90. R. D. DeMarco, M. E. Woods, and L. F. Arnold, Rubber Chem. Technol., 45, 1111
(1972).
91. C. F. Hammer, Macromolecules, 4, 69 (1971).
92. L. M. Robeson and J. E. McGrath, Polym. Eng. Sci., 17, 300 (1977).
93. J. J. Hickman and R. M. Ikeda, J. Polym. Sci., Polym. Phys. Ed. 11, 1713 (1973).
94. J. V. Koleske and R. D. Lundberg, J. Polym. Sci., Part A-2, 7, 795 (1969).
95. L. D. Robeson, J. Appl. Polym. Sci., 17, 3607 (1973).
96. D. J. Hourston and I. D. Hughes, Polymer, 20, 823 (1979).
97. C. P. Doube and D. J. Walsh, Polymer, 20, 1115 (1979).
98. T. Nishi and T. K. Kwei, J. Appl. Polym. Sci., 20, 1331 (1976).
99. D., J. Hourston and I. D. Hughes, J. Appl. Polym. Sci., 21, 3099 (1977).

Page 608
100. R. P. Kane, J. Elast. Plast., 9, 415 (1977).
101. A. E. Chalykh and I. N. Sapozhnikova, Abstr. Proc. 3rd Int. Conf. PVC, IUPAC,
Cleveland, Ohio, 1980, p. 195.
102. J. T. Cheng and G. J. Mantell, J. Appl. Polym. Sci., 23, 1733 (1978).
103. R. H. Hall, J. J. Lamson, A. J. Sikkema, and C. W. Roberts (to Dow Chemical Co.), U.S.
Pat. 3,424,823 (1969).
104. J. F. Kenney, in Recent Advances in Polymer Blends, Blocks, and Grafts, L. H. Sperling,
ed., Plenum Press, New York, 1974, p. 117.
105. P. D. Trivedi, R. J. Ambrose, and A. E. Altenau, Abstr. Proc. 3rd Int. Conf. PVC, IUPAC,
Cleveland, Ohio, 1980, p. 173.
106. L. G. Bourland and A. D. Wambach, J. Vinyl Technol., 5(3), 121 (1983); Plast. Eng.,
39(5), 23 (1983).
107. R. W. Gould and J. M. Player, Kunstoffe, 69(7), 10, 393 (1979).
108. F. Ide, Abstr. Proc. 3rd Int. Conf. PVC, IUPAC, Cleveland, Ohio, 1980, p. 143.
109. D. A. Tester, in Manufacture and Processing of PVC, Macmillan, New York, 1982, p.
215.
110. C. Vasiliu-Oprea and M. Popa, Colloid Polym. Sci., 260, 570 (1982).
111. J. T. Lutz, Org. Coat. Appl. Polym. Sci. Proc., 46, 635 (1981).
112. R. A. Emmett, Ind. Eng. Chem., 36, 730 (1944).
113. A. H. Jorgensen and D. G. Frazer, Appl. Polym. Symp., No. 7, 83 (1968).
114. J. W. Horvath, W. A. Wilson, H. S. Lundstrom, and J. R. Purdon, Appl. Polym. Symp.,
No. 7, 95 (1968).
115. Hitachi Cable Co., Jap. Pat. 58109, 548 (June 29, 1983).
116. Toyoda Gosei Co., Toyota Motor Co. Ltd., and Nippon Zeon Co. Ltd., Jap. Pat. 38118,
248 (July 14, 1983).
117. Showa Electric Wire and Cable Co. Ltd., and Riken Vinyl Industry Co. Ltd., Jap. Pat. 58
17, 145 (Feb. 1, 1983).
118. Tokuyama Sekisui Industry Co. Ltd., Jap. Pat. 5849, 764 (Mar. 24, 1983).

119. Pioneer Electric Co., Jap. Pat. 5871, 938 (Apr. 28, 1983).
120. K. Umeya, Jap. Pat. 5736, 472 (Aug. 4. 1982).
121. A. Yamamoto, M. Shiraishi, M. Isozumi, and K. Oka (to Japan EP Rubber Co. Ltd.), Jap.
Pat. 7867, 754 (June 16, 1978).
122. Toyoda Gosei Co. Ltd., Jap. Pat. 57126, 662 (Aug. 6, 1982).
122a. A. J. Kinloch and R. J. Young, Fracture Behavior of Polymers, Applied Science
Publishers, London, 1983.
122b. L. G. Bosveliev, I. D. Denchev, and A. M. Atanasov, Abstr. Proc. 3rd Int. Conf. PVC,
IUPAC, Cleveland, Ohio, 1980, p. 190.
123. S. Newman, in Polymer Blends, D. R. Paul and S. Newman, eds., Academic Press, New
York, 1978, Chap. 13.
124. M. Inagaki, K. Kuroda, and M. Sakai, Carbon, 21, 231 (1983).
124a. Denki Kagaku Kogyo K. K., Jap. Pat. 5813, 650 (Jan. 26, 1983).
125. E. Rhrl, Kunstoffe, 70, 40 (1980).
126. D. Fleischer, J. Brandrup, and O. Heinzmann, Kunstoffe, 67, 312316 (1977).
127. G. Menzel, Angew. Makromol. Chem., 47, 181199 (1975).
128. G. Menzel, G. Hundertnark, and A. Polte, Kunstoffe, 67, 339344 (1977).

Page 609
128a. K.-H. Michl, Kunstoffe, 70, 591 (1984).
129. G. Leps, J. Sachse, J. Bohse, and R. Stephan, Prakt. Metallogr., 20, 297 (1983).
130. G. Leps, J. Bohse, R. Stephan, and J. Sachse, Plast. Kautsch., 30(8), 458 (1983).
130a. A. Siegmann and A. Hiltner, Polym. Eng. Sci., 24, 867, 869 (1984).
131. W. J. McGill and T. Wittstock, Plast. Rubber Process Appl., 3, 77 (1983).
132. H. Tachibana, M. Takada, I. Takeshita, Z. Ogoshi, T. Shikano, and M. Kobayashi, Proc.
3rd Int. Conf. PVC, IUPAC, Cleveland, Ohio, 1980, p. 154.
133. M. Herner and H. Dikow, Proc. 3rd Int. Conf. PVC, IUPAC, Cleveland, Ohio, 1980, p.
152.
134. R. P. Petrich, Polym. Eng. Sci., 13, 248 (1973).
135. J. Weinlich, O. Plewan, and H. J. Leugering (to Hoechst A.-G.), Ger. Pat. 3,018,940
(Nov. 26, 1981).
136. H. Breuer, F. Haaf, and J. Stabenow, J. Macromol. Sci.Phys., B14, 387 (1977).
137. M. Matsuo, Polym. Eng. Sci., 9, 206 (1969).
138. R. D. Deanin and C. Moshar, ACS Polym. Prepr., 15(1), 403 (1974).
139. R. D. Deanin and H. R. Vyas, Prepr., ACS Div. Org. Coat. Plast. Chem., 34(1), 630
(1974).
140. T. O. Purcell, ACS Polym. Prepr., 13(1), 699704 (1972).
141. W. Frey, H. Hefner, P. Kunkel, and K.-H. Michl, Proc. 3rd Int. Conf. PVC, IUPAC,
Cleveland, Ohio, 1980, p. 160.
141a. L. E. Walker and Eugene L. Slagowski, Abstr. Proc. 37th SPE Annu. Tech. Conf., No.
399 (1979).
142. M. D. Skibo, J. A. Manson, S. M. Webler, and R. W. Hertzberg, ACS Symp. Ser., 95,
311 (1979).
143. R. W. Hertzberg and J. A. Manson, Fatigue in Engineering Plastics, Academic Press,
New York, 1980.
144. J. Turkanis, R. W. Hertzberg, and J. A. Manson, unpublished work, Lehigh University,
1984.

145. J. F. Ligeois, E. Wauters, J. A. Manson, J. Michel, and R. W. Hertzberg, 3rd Int. Conf.
PVC, IUPAC, Cleveland, Ohio, 1980.
146. J. C. Michel, J. A. Manson, R. W. Hertzberg, and J. F. Liegeois, in Polyblends, Plastics
and Rubber Institute, London, 1981, p. 15.1.
147. S. M. Webler, Ph.D. dissertation, Lehigh University, 1983.
148. C. M. Rimnac, Ph.D. dissertation, Lehigh University, 1983.
149. L. M. Robeson, A. Noshay, M. Matzner, and C. N. Merriam, Angew. Makromol. Chem.,
29/30, 47 (1973).
150. Y. J. Shur and B. G. Rnby, J. Appl. Polym. Sci., 19, 1337 (1975); 19, 2143 (1975); 20,
3105 (1976); 20, 3121 (1976).
151. H. Storstrom and B. Rnby, Adv. Chem. Ser., 99, 107 (1971).
152. B. G. Rnby, J. Polym. Sci., Polym. Symp., 51, 89 (1975).
153. L. C. E. Struik, Physical Aging in Polymers and Other Materials, Elsevier, New York,
1978.
154. J. A. Brydson, Plastics Materials, 3rd ed., Butterworth, London, 1975.
155. A. Kaminska and H. Koczmarek, Angew. Makromol. Chem., 119, 29 (1983).
156. I. C. McNeill, N. Grassie, J. N. R. Sampson, A. Jamieson, and T. Straiton, J. Macromol.
Sci.Chem., A12, 503 (1978).

Page 610
156a. E. G. Kolawole and M. O. Agboola, J. Appl. Polym. Sci., 27, 2317 (1982).
156b. E. G. Kolawole, G. E. Mbamali, and J. Y. Olayemi, J. Appl. Polym. Sci., 25, 2133
(1980).
157. M. Herner and H. Dikow, Proc. 3rd Int. Conf. PVC, IUPAC, Cleveland, Ohio, 1980, p.
152.
158. G. Menzel, Kunstoffe, 69, 480482 (1979).
159. W. J. Hornibrook, Proc. 3rd Int. Conf. PVC, IUPAC, Cleveland, Ohio, 1980, p. 149.
160. D. Feldman, Proc. 28th IUPAC Macromol. Symp., 671 (1982).
161. D. Feldman, J. Appl. Polym. Sci., 24, 3493 (1981).
162. J. E. McGrath, Pure Appl. Chem., 55, 1573 (1983).
163. J. Furukawa, S. Takamori, and S. Yamashita, Angew. Makromol. Chem., 1, 92 (1967).
164. J. T. Laverty and Z. G. Gardlund, ACS Polymer Prepr., 15(2), 306 (1974).
165. General Motors Corporation, U.S. Pat. 3,949,015 (1976).
166. J. T. Laverty and Z. G. Gardlund, J. Polym. Sci., Polym. Chem. Ed., 15, 2001 (1977).
167. G. Smets, G. Weinand, and S. Deguchi, J. Polym. Sci., Polym. Chem. Ed., 16, 3077
(1978).
168. G. Weinand and G. Smets, J. Polym. Sci., Polym. Chem. Ed., 16, 3091 (1978).
169. N. Overbergh, G. Weinand, and G. Smets, J. Polym. Sci., Polym. Chem. Ed., 10, 3107
(1978).
170. P. Lalet, H. Fassi, and A. Miletto, (to Aquitaine Total Organico), Fr. Pat. 2,213,293 (Aug.
2, 1973).
171. A. Guyot, M. Ceysson, A. Michel, and A. Revillon, Inf. Chim., 116, 127 (1973).
172. G. Riess and F. Palacin (to Agence Nationale de Valorisation de la Recherche), Fr. Pat.
2,123,590 (Oct. 20, 1972).
173. A. Michel, Ger. Pat. 2,428,542 (1975).
174. A. Michel, G. Schmidt, E. Castaneda, and A. Guyot, Angew. Makromol. Chem., 47, 61
(1975).

175. A. Michel, E. Castaneda, and A. Guyot, J. Macromol. Sci.Chem., A12, 227 (1978).
176. A. Michel, E. Castaneda, and A. Guyot, Bur. Polym. J., 15, 935 (1979).
177. J. J. Laverty and Z. G. Gardlund, J. Appl. Polym. Sci., 24, 3657 (1981).
178. R. J. Ceresa, ed., Block and Graft Polymerization, 2 vols., Wiley, New York, 1978, 1976.
179. V. M. Matzner, L. M. Robeson, A. Noshay, and J. E. McGrath, in Encyclopedia of
Polymer Science and Technology, H. F. Mark and N. Bikales, eds., Suppl. Vol. 2, WileyInterscience, New York, 1977, p. 129.
180. V. Stannett, J. Macromol. Sci.Chem., 4, 1177 (1970).
181. R. J. Ceresa, in Block and Graft Polymerization, R. J. Ceresa, ed., Vol. 2, Wiley, New
York, 1976, p. 273.
182. A. K. Mukherjee and A. Gupta, J. Macromol. Sci.Rev. Macromol. Sci., 19, 293 (1981).
183. A. Casale and R. S. Porter, Polymer Stress Reactions, 2 vols., Academic Press, New
York, 1978.

Page 611
184. A. Kishimoto, S. Hirata, and H. Ueno, Br. Pat. 1,254,354 (1968).
185. A. Ghaffar, C. Sadrmohaghegh, and G. Scott, Polym. Degrad. Stab., 3, 341 (1981).
186. A. Ravve and J. T. Khamis, U.S. Pat. 3,100,758 (1959).
187. A. C. Baskett and R. B. MacMillan (to Imperial Chemical Industries), Br. Pat. 1,062,308
(1965).
188. A. R. Nelson (to Dow Chemical Co.), U.S. Pat. 3,240,843 (1962).
189. J. M. Michel, Fr. Pat. 1,351,387 (1962).
189a. N. Moszner and M. Hartmann, Z. Chem., 18(11), 407 (1978).
190. A. Cooper, Br. Pat. 902,354 (1958).
191. General Tire & Rubber Co., Br. Pat. 994,342 (June 2, 1965).
192. Kanegafuchi Chemical Industries, Ltd., Jap. Pat. 57174,309 (Oct. 27, 1982).
193. S. P. Rao and M. Santappa, J. Polym. Sci., Part A-1, 6, 95 (1968).
194. G. D. Williams, Fr. Pat. 1,584,916 (1967).
195. D. J. Walsh and C. K. Sham, Polymer, 25, 1023 (1984).
196. Toray Industries Ltd., Jap. Pat. 82119,756 (1982).
197. I. B. Kotlyar, B. M. Sharetskii, I. I. Gorbachevakaya, and E. P. Schvarev, Plast. Massy,
8, 46 (1977).
198. D. A. Kritskaya, A. D. Pomogailo, A. N. Ponomarev, and F. S. Dyaschkovskii, J. Appl.
Polym. Sci., 25, 349 (1980).
199. M. Kozlowski, J. Piglowski, and T. S. Skowronski, Przem. Chem., 61, 177, 312 (1982).
200. T. Gaal, J. Timar, V. Kiss, and I. Lantos, Hung. Pat. 22,184 (Apr. 24, 1982).
201. D. Wulff, J. Gerecke, K. Thiele, A. Kuenzel, R. Wintzer, D. Loef-fler, and L. Haeussler (to
VEB Chemische Werke Buna), Ger. (East) Pat. 149,074 (June 24, 1981).
202. Mitsui Toatsu Chemicals, Inc., Jap. Pat. 57117,554 (July 22, 1982).
203. Solvay et Cie. , Br. Pat. 1, 093, 974 (1965).
204. E. G. Melby, H. W. Cocain, and H. J. Fabris (to General Tire and Rubber Co.), U.S. Pat.

4,423,193 (July 27, 1981).

205. A. K. Mukherjee and A. Gupta, J. Macromol. Sci.Chem., 15, 783, 1161, 1323 (1981);
J. Macromol. Sci.-Rev. Macromol. Chem., C20, 309 (1981).
206. B.F. Goodrich, Br. Pat. 1,121,977 (1966).
207. A. K. Mukherjee and A. Gupta, J. Appl. Polym. Sci., 28, 1245, 1283 (1983).
208. Solvay et Cie, Br. Pat. 899,993 (1960).
209. Solvay et Cie, Belg. Pat. 586,455 (1960).
210. H. H. Friederich and F. Gundel, Ger. Pat. 1,105,169 (1961).
211. Solvay et Cie, Belg. Pat. 618,194 (1962).
212. Solvay et Cie, Br. Pat. 967,288 (1962).
213. Y. Landler and P. Lebel, J. Polym. Sci., 48, 477 (1960).
214. Y. Landler and P. Lebel, Grafting on Polyvinylc ori e by Pre-ozonization, Int. Polym.
Conf., Moscow, 1960.
215. S. A. Solvic, Br. Pat. 881, 503 (1958).
216. A. Michel, E. Castaneda, and A. Guyot, J. Macromol. Sci.Chem., A12, 227 (1978).
217. A. Chapiro, Radiation Chemistry of Polymer Systems, High Polymer Series, Vol. xv,
Interscience, New York, 1962.

Page 612
218. A. Chapiro, in Irradiation of Polymers, Adv. Chem. Ser. 66, R. F. Gould, ed., American
Chemical Society, Washington, D.C., 1967, p. 22.
219. A. Chapiro, M. Magat, and J. Sebban (to Centre National de la Recherche Scientifique),
Fr. Pat. 1,130,099 (1956).
219a. J. Gabilly and M. Jobard (to Soc. Saint-Gobain), Fr. Pat. 1,147,722 (1957).
220. S. H. Pinner (to T.I. Ltd.), Fr. Pat. 1,184,174 (1959).
221. G. Hardy and J. Dobo, Chem. Prum., 9, 215 (1959).
222. A. Chapiro and A. Matsumoto, J. Polym. Sci., 57, 743 (1962).
223. D. Harmer, in Irradiation of Polymers, Adv. Chem. Series 66, R. F. Gould, ed., American
Chemical Society, Washington, D.C., 1957.
224. A. Chapiro and A. M. Jendrvchowska-Bonamour (to Centre National de la Recherche
Scientifique), Fr. Pat. 1,479,353 (May 5, 1967).
225. The Dow Chemical Co., Br. Pat. 829,512 (1960).
226. S. Danon, M. Jobard, M. Lautout, M. Magat, M. Michel, N. Riou, and C. Wippler, J.
Polym. Sci., 35, 517 (1959).
227. A. Chapiro and A. M. Jendrychowska-Bonamour, Fr. Pat. 1,533,942 (July 26, 1968).
228. C. Wippler (to Socit Anonyme des Manufactures des Glaces et Produits Chimiques de
Saint-Gobain), Fr. Pat. 1,166,793 (July 26, 1968).
229. A. Chapiro and G. Palma, Eur. Polym. J., 3, 151 (1967).
230. K. Sugimoto, T. Yoshitake, and-K. Imai (to Kurashiki Rayon Co.), Jap. Pat, 6902,230
(Jan. 30, 1969).
231. A. Chapiro, M. Magat, and J. Sebban, (to Centre National de la Recherche Scientifique),
Fr. Pat. 1,125,537 and 1,130,100 (1956).
232. A. Chapiro and Z. Mankowski, Eur. Polym. J., 2, 163 (1966).
232a. Japanese Atomic Energy Research Inst., Jap. Pat. 8032, 806 (Aug. 27, 1980).
233. R. Gouloubandi and A. Chapiro, Eur. Polym. J., 14, 957 (1980).
234. G. Gasparrini, M. Carenza, and G. Palma, J. Polym. Sci., Polym. Lett. Ed., 18, 29
(1980).

235. K. G. Mayhan and W. J. Bertrand (to American Hospital Supply Corp.), U.S. Pat.
4,311,573 (Jan. 19, 1982).
236. C. H. Ang, M. A. Long, J. L. Garnett, R. Levot, N. T. Yen, and K. J. Nicol, Stud. Surf.
Sci. Catal., Part B, 7, 953 (1981).
237. Y. Nakase and N. Akasaka, Rep. JAERI-M-82-059, 1982 (Japan; available from INIS),
from INIS Atomindex, 14(15), 1983, Abstr. 771099.
238. A. A. Zufarov, D. P. Kirywkhin, M. K. Asamov, and I. M. Barkalov, Uzb. Khim. Zh., 6, 43
(1983).
239. C. Decker, J. Appl. Polym. Sci., 28, 97 (1983).
240. F. F. Rugg and T. E. Potts (to Union Carbide Corp.), U.S. Pat. 2,947,719 (Aug. 2, 1960).
241. Union Carbide Corp., Br. Pat. 814,393 (June 3, 1959).
242. Y. Hoski and A. Kaneko (to Kureka Chemical Industry Co. Ltd.), Jap. Pat. 15,042 (Sept.
25, 1962).
243. R. T. Reid and L. E. Best (to the Firestone Tire & Rubber Co.), U.S. Pat. 3,079,208
(Jan. 30, 1960).
243a. C. A. Heilberger, private communication, 1984.

Page 613
244. Monsanto Co., Br. Pat. 1,029,634 (May 18, 1966).
245. Y. Landler and P. Lebel (to Polyplastic Co.), Fr. Pat. 1,166,698 (Nov. 13, 1958).
246. VEB Chemische Werke Buna, East Ger. Pat. 134,232 (Feb. 14, 1979).
247. D. Hardt and H. Bartl (to Farbenfabriken Bayer A. G.), Br. Pat. 1,076,930 (July 26,
1967).
248. H. Bartl and D. Hardt, Adv. Chem. Ser., 91, 477 (1969).
249. A. Kolinsky, V. Macho, J. Porubsky, V. Kuska, J. Manas, R. Lukas, and S. Sykora, Fr.
Pat. 2,466,480 (Apr. 10, 1981).
250. A. Takahashi (to Hooker Chemical Co.), U.S. Pat. 4,0171,582 (Jan. 31, 1978).
251. M. W. Allsop, in Manufacture and Processing of PVC, R. J. Burgess, ed., Macmillan,
New York, 1982, p. 59.
252. Tokuyamia Soda Co. Ltd., Jap. Pat. 8203,308 (Jan. 9, 1982).
253. Monsanto Co., Br. Pat. 1,095,831 (Dec. 20, 1967).
254. Monsanto Co., Br. Pat. 1,095,831 (Dec. 20, 1967).
255. Hitachi Cable Co. Ltd., Jap. Pat. 5750, 325 (Jan. 26, 1977).
256. Produits Chimiques Pechiney-Saint-Gobain, Br. Pat. 1,076,898 (July 26, 1967).
257. Produits Chimiques Pechiney-Saint-Gobain, Br. Pat. 1,076,898 (July 26, 1967).
258. M. Pegoraro and E. Beati, Proc. 28th IUPAC Macromol. Symp., 649 (1982).
259. V. Kuhlwilm, H. Reuter, J. Stude, and H. J. Wolf, Ger. (East) Pat. 129,088 (Dec. 28,
1977).
260. Ryo-Nichi Co. Ltd., Jap. Pat. 57131, 241 (Aug. 14, 1982).
261. W. L. Schall (to Hooker Chemicals and Plastics Corp.), Eur. Pat. Appl. 49,595 (Apr. 14,
1982).
262. Chisso Corp., Jap. Pat. 57185, 342 (Nov. 15, 1982).
263. A. Takahashi, Polym. Eng. Sci., 22, 54 (1982).
264. Kanegafuchi Chemical Industry Co., Jap. Pat. 8181, 325 (July 3, 1981).

265. Japan Synthetic Rubber Co., Jap. Pat. 5898, 348 (Dec. 8, 1981).
266. D. Hardt and H. Bartl (to Farbenfabriken Bayer A.G.), Br. Pat. 1,076,930 (July 26,
1967).
267. D. Hardt, W. Gobel, H. Bartl, and H. Scheurlen (to Farbenfabriken Bayer, A.G.), Br. Pat.
1,027,710 (Apr. 27, 1966).
268. L. Haeussier, J. Gerecke, W. Schneider, F. Zacheaus, H. Brueckmann, R. Wintzer, D.
Weiwad, and K. P. Geissler [to VEB Chem. Werke (Buna)], Ger. (East) Pat. 158,908 (Feb. 9.
1983).
269. Sekisui Chemical Co. Ltd., and Tokuyama Sekisui Industry Co. Ltd., Jap. Pat. 5857,
412 (Apr. 5, 1983).
270. Sekisui Chemical Co. Ltd. and Tokuyama Sekisui Industry Co. Ltd., Jap. Pat. 82133,
143 (Aug. 17, 1982).
271. P. Lamberty, Eur. Pat. Appl. EP 74, 139 (Mar. 16, 1983).
272. Toyo Soda Manufacturing Co. Ltd., Jap. Pat. 8270, 149 (Oct. 17, 1980).
273. D. Hardt, H. Bartl, and R. Wust (to Farbenfabriken Bayer A.G.), U.S. Pat. 3,355,516
(Nov. 28, 1967).
274. A. Lonnstrom, A. Litja, and V. Tammela, Kem.-Kemi, 9, 409 (1982).
275. J. Spevacek, Makromol. Chem., Rapid Commun., 3, 629 (1982).
276. Dynamit Nobel A.-G., Ger. Pat. 1,595,515 (Oct. 23, 1968).
277. Solvay et Cie., Br. Pat. 1, 530,854 (Nov. 24, 1979).

Page 614
277a. Nippon Zeon Co. Ltd., Jap. Pat. 57164, 746 (Oct. 8, 1982).
278. Denki Kagaku, Jap. Pat. 58136, 608 (Aug. 13, 1983).
279. W. Frey and W. Kleine (to Wacker-Chemie GmbH), Ger. Pat. 3,142,420 (May 5, 1983).
280. O. Nuyken, J. Gerum, and R. Kerber, Angew. Makromol. Chem., 91, 143 (1980).
281. Y. Fujiji, H. Hayashi, T. Suganuma, and Yo. Sakamoto (to Gosei Chemical Industry Co.
Ltd.), Belg. Pat. 894,300 (Jan. 3, 1983).
282. M. Onetiu, Rom. Pat. 66,502 (Sept. 25, 1980).
283. S. Kimura, T. Takitani, and M. Imoto, Bull. Soc. Chim. Jpn., 35, 2012 (1962).
284. D. Hardt, H. Bartl, W. Gobel, and K. Denges (to Farbenfabriken Bayer A.G.), Br. Pat.
1,078,095 (Aug. 2, 1967).
285. D. Hardt and H. Bartl (to Farbenfabriken Bayer A.G.), Br. Pat. 1,075,643 (July 12,
1967).
286. D. Hardt and H. Bartl (to Farbenfabriken Bayer A.G.), Br. Pat. 1,075,642 (July 12,
1967).
287. D. Hardt and H. Bartl (to Farbenfabriken Bayer A.G.), Br. Pat. 1,076,930 (July 26,
1967).
288. D. Hardt and H. Bartl (to Farbenfabriken Bayer A.G.), Br. Pat. 1,028,883 (May 11,
1966).
289. G. Lechermeier, A. Revillon, and C. Pillot, J. Macromol. Sci.-Chem., A12, 285 (1978).
290. F. Rempp, J. Parrod, G. Laurent, and Y. Gallot, Compt. Rend., 260, 903 (1965).
291. G. Finaz, Y. Gallot, F. Rempp, and J. Parrod, J. Polym. Sci., 58, 1363 (1963).
292. Y. Gallot, P. Remp, and J. Parrod, J. Polym. Sci., B1, 329 (1963).
293. Y. Minoura and K. Shiina, J. Polym. Sci., Al, (4), 1069 (1966).
294. Y. Minoura, H. Hironaka, T. Kasabo, and Y. Ueno, J. Polym. Sci., Al, (6), 2773 (1968).
295. J. D. Nichols, J. Polym. Sci., Lett. Ed., 11, 705 (1973).
296. R. C. Gregorian and P. J. Conway, U.S. Pat. 3,135,717 (1960).

297. S. P. Mitsengendler, K. I. Sokolova, C. A. Andreeva, A. A. Korotov, T. Kadyrov, S. I.

Klenin, and S. Ya. Magarik, Vysokomol. Soedin., 9A, 1133 (1967).
298. G. Lechermeier, C. Pillot, J. Gole, and A. Revillon, J. Appl. Polym. Sci., 19, 1979 (1975).
299. G. Lechermeier, A. Revillon, and C. Pillot, J. Appl. Polym. Sci., 19, 1989 (1975).
300. M. Kucera and M. Kunz, Czech. Pat. 209, 559 (Feb. 15, 1982).
301. S. Marian and G. Levin, J. Appl. Polym. Sci., 26, 3295 (1981).
302. S. P. Mitzengendler, C. A. Andreeva, K. J. Sokolova, and A. A. Korotkov, Vysokomol.
Soedin., 4, 1366 (1962).
303. A. Crespy and M. J. M. Abadie, Eur. Polym. J., 17, 221 (1981).
304. Inoue-Japan Research Inc., Jap. Pat. 8100,091 (Jan. 6. 1981).
305. P. H. Plesch, Br. Pat. 817,684 (1959).
306. O. F. Solomon, M. Dimonie, and C. Ciuciu, J. Polym. Sci., Part A-1, 8, 777 (1970).
307. J. P. Kennedy and F. P. Baldwin (to Esso Research and Engineering Co.), U.S. Pat.
3,904,708 (Sept. 9, 1975).
308. J. P. Kennedy, ed., Cationic Graft Copolymerization, Appl. Polym. Symp., No. 30, 1977.

Page 615
309. J. P. Kennedy and R. R. Smith, Polym. Propr., 13(2), 710 (1972).
310. J. P. Kennedy, J. J. Charles, and D. L. Davidson, Polym. Prepr., 14(2), 974 (1973).
311. J. P. Kennedy and D. L. Davidson, Polym. Propr., 15(2), 209 (1974).
312. J. P. Kennedy and D. L. Davidson, Appl. Polym. Symp., 30, 13, 51 (1977).
313. B. Ivan, J. P. Kennedy, T. Kelen, and F. Tudos, J. Macromol. Sci.Chem., A-16, 1473
(1981).
314. N. G. Gaylord and A. Takahashi, Belg. Pat. 729, 627 (1969).
315. N. G. Gaylord and A. Takahashi, J. Polym. Sci., Part B, 8, 349, 361 (1970).
316. N. G. Thame, R. D. Lundberg, and J. P. Kennedy, J. Polym. Sci., Part A-1, 10, 2507
(1972).
317. K. B. Abbaos and N. G. Thame, J. Polym. Sci., Polym. Chem. Ed., 13, 59 (1975).
318. Chisso Corp., Jap. Pat. 5869, 226 (Apr. 25, 1983).
319. K. K. Sharma and I. K. Varma, J. Appl. Polym. Sci., 26, 1079, 1087 (1980).
320. C. D. Doyle, Anal. Chem., 33, 77 (1961).
321. T. Taminaka, K. Iwamoto, and Y. Minoura, J. Polym. Sci., Polym. Chem. Ed., 15, 2219
(1977).
322. A. Chapiro and J. M. Jendrychowska-Bonamour, Br. Pat. 1,225,318 (1967).
323. V. Kudryatsev and V. Y. Kabanov, Vysokomol. Soedin., Ser. A, 24, 401 (1982).
324. Kanefaguchi Chemical Industry Co. Ltd., Jap. Pat. 57179, 214 (Nov. 4, 1982).
325. G. G. A. Bohm, F. Oliver, and D. S. Pearson, Soc. Plast. Eng. J., 27(7), 21 (1970).
326. R. A. Bellino, Wire Journal, 75 (Oct. 75).
327. D. C. Alexander, Plast. Polym., 195 (1975).
328. L. D. Loan, Radiat. Phys. Chem., 9, 253 (1977).
329. L. Saito, Radiat. Phys. Chem., 9, 675 (1977).
330. E. Scalo and W. F. Moore, Proc. Int. Conf. Radiat. Process. Plast. Rubbers, Brighton,
England, 1981, p. 34.1.

331. T. Hayashi, J. Ito, K. Mitani, and Y. Mitzutani, J. Appl. Polym. Sci., 28, 2867 (1983).
332. D. M. Florence and W. E. Smith (Armstrong World Industries Ltd.), Ger. Pat. 3,300,040
(July 14, 1983).
333. A. Chapiro, Radiation Chemistry of Polymeric Systems, Interscience, New York, 1962, p.
377.
334. G. Palma, M. Carenza, and C. Pollini, Eur. Polym. J., 16, 333 (1980).
335. A. A. Miller, Ind. Eng. Chem., 31, 1271 (1959).
336. A. A. Miller, J. Phys. Chem., 63, 1755 (1959).
337. S. H. Pinner, Plastics, 25, 35 (1960).
338. S. H. Pinner, Nature, 183, 1108 (1959).
339. C. H. Ang, J. L. Garnett, R. Levot, and M. A. Long, J. Appl. Polym. Sci., 27 4893
(1982).
340. H. Mitsui, F. Hosoi, and T. Kagiya, Polym. J., 6, 20 (1974).
341. H. Koozu, K. Oodan, and S. Sacki, Shikizai Kyokaishi, 362 (1963).
342. A. Zyball, Kunstoffe, 72, 487 (1982).

Page 616
343. Nisshin Electric Co. Ltd., Jap. Pat. 58208, 367 (Dec. 5, 1983).
344. V. 1. Dakin, Z. S. Egorova, and V. L. Karpow, Plast. Massy, 40, 2 (1979).
345. Nisshin Electric Co. Ltd., Jap. Pat. 58208, 366 (Dec. 5, 1983).
346. W. A. Salmon and L. D. Loan, J. Appl. Polym. Sci., 16, 671 (1972).
347. T. J. Szymczak, M.S. thesis, Lehigh University, June 1970.
348. T. J. Szymczak and J. A. Manson, Mod. Plast., 51(8), 66 (1974).
349. E. Shito and K. Mogi (to Furukawa Electric Co. Ltd.), Jap. Pat. 78138,450 (Dec. 2,
1978).
350. J. Dobo, Pure Appl. Chem., 46, 1 (1976).
351. H. E. Bair, M. Matsuo, W. A. Salmon, and T. K. Kwei, Macromolecules, 5, 114 (1972).
352. D. D. Davis and W. P. Slichter, Macromolecules, 6, 728 (1973).
353. V. Gorshkov, V. I. Sheikin, and A. M. Petrukhov, USSR Pat. 817,035 (June 28, 1979).
354. F. T. White and R. Mann (to Assoc. Electrical Industries Ltd.), Br. Pat. 1,066,610 (1967).
355. T. N. Bowmer, D. B. Davis, T. K. Kwei, and W. I. Vroom, J. Appl. Polym. Sci., 26, 3669
(1981).
356. T. N. Bowmer, M. Y. Hellman, T. K. Kwei, and W. I. Vroom, J. Appl. Polym. Sci., 28,
2083 (1983).
357. M. Y. Hellman, T. N. Bowmer, and G. N. Taylor, Prepr., Am. Chem. Soc. Div. Plast.
Chem., 45, 126 (1981).
358. T. M. Bowmer and W. I. Vroom, J. Appl. Polym. Sci., 28, 3527 (1983).
359. Hitachi Cable Co., Jap. Pat. 8145, 940 (Apr. 25, 1981).
360. L. Kushner and R. S. Tu, Mod. Plast., 60(4), 87 (1983).
361. Mitsui Toatsu Chemicals, Inc., Jap. Pat. 81,163,108 (May 20, 1980).
362. Izumi, M., Suzuki, F., and Yuri, T., Sumimoto Denki, 90, 44 (Oct. 1965).
363. E. Saito, M. Nishimura, and H. Fukazawa (to Furukawa Electric Co. Ltd.), Jap. Pat. 76
132,261 (Nov. 17, 1976).

364. J. P. Kennedy and M. Nakao, J. Macromol. Sci.Chem. A12(2), 197 (1978).

365. L. H. Cragg and J. A. Manson, J. Polym. Sci., 9, 265 (1952).
366. Y. V. Ovchinnikov, S. V. Zhukhova, O. M. Sharetskii, and A. G. Kronman, Khim. Vys.
Energ., 16, 326 (1982).
367. J. P. Kennedy, J. Macromol. Sci.Chem., A12(2), 301 (1978).
368. G. Kraus and H. E. Railsback, in Recent Advances in Polymer Blends, Grafts and
Blocks, L. H. Sperling, ed., Plenum Press, New York, 1974, p. 257.
369. R. D. Deanin, A. A. Deanin, and T. Sjoblom, in Ref. 368, p. 71.
370. L. C. Rubens, D. A. Harmer, J. D. Griffin, M. L. Leslie, and H. G. Hammon, Belg. Pat.
634, 955 (1963).
371. J. P. Kennedy and D. L. Davidson, J. Appl. Polym. Sci., Appl. Polym. Symp., 30, 51
(1977).
372. B. X. Plastics, Fr. Pat. 1,307,117 (Oct. 19, 1962).
373. Solvay et Cie, Belg. Pat. 673,317 (1965).
374. K. Hosoda, Ger. Pat. 1,946,347 (Apr. 16, 1970).
375. Toyoda Gosei Co. Ltd., Jap. Pat. 81136,866 (Oct. .26, 1981).

Page 617
376. T. D. Pendle, in Block and Graft Polymerization, R. J. Ceresa, ed., Vol. 1, Wiley, New
York, 1973.
377. J. Sionakidis, L. H. Sperling, and D. A. Thomas, J. Appl. Polym. Sci., 24, 1179 (1979).
378. H. Djomo, R. Colmenares, and G. C. Meyer, Eur. Polym. J., 31, 521 (1981).
379. T. Nishi, T. K. Kwei, and T. T. Wang, J. Appl. Phys., 46, 4157 (1975).

Page 618

Page 619

11 Chemical Modifications: Chlorinated PVC

RICHARD G. PARKER
B. F. Goodrich Company Brecksville, Ohio
GILBERT A. MARTELLO
B. F. Goodrich Chemical Group Cleveland, Ohio

I. INTRODUCTION

619

II. METHODS OF CHLORINATION

620

A. Solution Processes

621

B. Water-Slurry Processes

621

C. Fluid-Bed Processes

622

D. Fixed-Bed Processes

622

E. Liquid Chlorine Processes

622

III. THE STRUCTURE OF CPVCs AND THE MECHANISM OF CHLORINATION 623

IV. THERMAL STABILITY AND STABILIZATION
V. PROPERTIES
A. Chemical - Resistant Properties
B. Thermal Properties
C. Rheological Properties
D. Mechanical Properties
E. Dielectric Properties
F. Thermal Decomposition Characteristics
VI. MARKETS AND COMMERCIAL USES

632
636
636
638
640
642
644
646
648

A. Introduction
B. Market Considerations
C. Traditional Uses of CPVC

648
648
649

D. New Uses for CPVC

649

E. Summary

650

REFERENCES

651

1. INTRODUCTION
In the chapter Chemically Modified Polyvinyl Chloride in the first edition of this work, Dannis
and Ramp [1] discussed the various ways in which PVC may be chemically modified, and
concluded that except for chlorinated PVC, none

Page 620
of these materials has been shown to possess unique and desirable properties. This remains
true today.
Chlorinated polyvinyl chloride (CPVC) has become an important specialty polymer due to its
relatively low cost; high glass transition temperature; high heat distortion temperature;
outstanding mechanical, dielectric, and name and smoke properties; chemical inertness; and
low sensitivity to hydrocarbon costs.
The chlorination of PVC is conceptually simple. PVC (56.7% chlorine) and chlorine are allowed
to react, under a wide variety of conditions, in a typical free-radical chain reaction (Figure 1) to
form a polymer with increased chlorine content. A recent patent literature survey [2] revealed
that at least 55 companies worldwide hold some 125 patents related to the production of
CPVC. These patents describe a myriad of production methods and variations.
The free-radical chlorination of PVC produces a polymer for which no structural repeat unit can
be written. Since the substitution of chlorine into PVC occurs in a gradual manner, a family of
polymers, whose individual properties are dependent on the extent, and method, of chlorination,
can be produced. Thus the term CPVC cannot be applied to an individual polymer without
reference to its chlorine content and method of production.
Commercially, there are two major types of CPVC resins produced and sold. These are usually
described by their weight percent chlorine content, one type containing 63 to 64% chlorine, with
the second type containing 67 to 68% chlorine.
In the following discussion, the methods of production, the structure and mechanism of
formation, stability and stabilization, properties, and commercial uses of CPVCs will be
summarized [36].

II. METHODS OF CHLORINATION

CPVCs can be produced in a wide variety of processes, the majority of which involve the
formation of chlorine radicals. Chlorinations in the presence of Lewis acids, presumably ionic in
nature, have also been reported [7].
Five types of chlorination processes have received patent coverage over the years. The five
processes can be referred to generically as the (1) solution, (2) water-slurry, (3) fluid-bed, (4)
fixed-bed, and (5) liquid chlorine processes. An important consideration in all the processes is
the material of construction of the process equipment. Due to constant exposure to chlorine
(wet and dry) and hydrogen chloride (wet and dry), most CPVC process

FIGURE 1. Generalized mechanism for the free-radical chlorination of hydrocarbons.

Page 621
equipment tends to be specialized (titanium, tantalum), and expensive to fabricate and maintain.
A. Solution Processes
This is the oldest method of chlorinating PVC [4]. A solution of PVC, about 10% by weight in a
suitable solvent, usually a chlorinated solvent, is chlorinated under selected conditions of
temperature, pressure, initiation, and chlorine feed [8]. The polymer becomes more soluble as
the chlorination proceeds, so precipitation during the reaction is not a problem. At the end of
the reaction, the polymer can be recovered by precipitation with a nonsolvent [9] or by solvent
removal. In some applications the chlorine-free solution may be used directly.
The advantage of solution chlorination is that a product often reported [10] to contain a more
uniform distribution of chlorine than that produced in nonsolution processes is obtained. The
disadvantages are the large volumes of solvent needed, the cost of solvent recovery, and the
cost of residual solvent removal from the polymer.
B. Water-Slurry Processes
The most commonly practiced commercial processes are the water-slurry methods. These are
based primarily on the patent of Dannis and Ramp [11]. In the original process, PVC was
suspended in water together with a swelling agent such as chloroform. The resultant slurry
was then degassed under vacuum to remove oxygen, and heated to the desired temperature
(50C). Chlorine was added to a set pressure and ultraviolet lights were used to initiate the
chlorination. As the reaction proceeded, additional chlorine was added to maintain the set
pressure. The hydrogen chloride formed dissolved in the water to yield a slurry of increasing
acidity. The polymer was recovered by removal of the aqueous acid. At this point, a reslurry
with fresh water and a neutralization step, using a suitable inorganic base, occurred to ensure
removal of the aqueous acid. The wet polymer was then dried before processing or storage.
Many variations of the basic Dannis-Ramp process have appeared since 1961. Among these
are higher temperatures [12], saturation of the gaseous chlorine with the swelling agent as an
alternative to mixing it with the water [13]; removal of the final traces of chlorine with reducing
agents [14]; the use of chemical catalysts [15]; the deliberate introduction of controlled amounts
of oxygen to act as the catalyst [16]; the use of special PVCs polymerized in the presence of,
and still containing, the swelling agents [17]; the use of PVC preswollen with gaseous swelling
agents [18]; the replacement of water by concentrated hydrochloric acid [18]; and careful
monitoring of the amount of ultraviolet (UV) radiation used [19]. A very important finding was
that swelling agents are not necessary, and as a result, the use of swelling agents has now
been discontinued in the United States [20].
The main advantages of the modern water-slurry processes, compared with the solution
processes, are cost and the lack of need to remove and recover the residual swelling agents,
usually chlorinated hydrocarbons [21]. Because chlorine (or chlorine radicals) must diffuse from
the water phase into the PVC particles, water-slurry chlorinations are generally diffusion
controlled. As a result, more chlorination tends to occur near the outer surface of the particles
than near the center [22]. Compared with solution chlorina-

Page 622
tion, water-slurry polymers tend to be less homogeneously chlorinated. Under selected
conditions, it is possible to produce a polymer exhibiting two glass transition temperatures; one
similar to PVC, the other characteristic of some CPVC.
One important area of research concerns the role of aqueous chlorine in the chlorination
chemistry. At least six chlorine-containing species, in addition to molecular chlorine, are believed
to exist in aqueous solution [23].
C. Fluid-Bed Processes
Fluid-bed chlorination [24] of PVC is potentially the least expensive process because the
polymer can be recovered directly. PVC is fluidized in a suitable reactor using chlorine, or more
commonly, chlorine diluted with an inert gas. The use of swelling agents has also been reported
[24b]. The chlorination is initiated with heat or light, or a combination. The hydrogen chloride
produced is carried away to a scrubber system. When the desired chlorine content is reached,
the residual chlorine and hydrogen chloride are removed by entrainment, and the polymer is
recovered directly, ready for processing or storage.
However, like the water-slurry processes, fluid-bed processes are also diffusion controlled.
Furthermore, the activity (partial pressure) of chlorine is usually rather low, resulting in long
diffusion times for interior-particle chlorination. Thus fluid-bed polymer particles also tend to be
more highly chlorinated near the surface than water-slurry polymers. In addition, heat transfer
from the solid phase to the gas phase is rather inefficient, resulting in potential localized
overheating, due to the heat of reaction, of the PVC/ CPVC particles. This could result in
degradation of the polymer. Chain scission has also been noted in some experimental polymers
prepared using fluid-bed conditions at 160C [25].
D. Fixed-Bed Processes
These processes are very similar to the fluid-bed processes except that the reactions are
usually carried out using a rotating tube [26], conveyor belt [27], and so on. The characteristics
of the polymer produced should be very similar to that produced in the fluid-bed process. One
variation calls for the chlorination to be begun at a temperature below Tg with UV initiation, and
then finished above Tg without UV [15].
E. Liquid Chlorine Processes
Two recent patents [28,29] have been granted describing the use of liquid chlorine (LCl2, b.p.
-34.6C) as both the reactant and the reaction medium for photochlorinating PVCs. At one
preferred LCl2/PVC ratio of about 3 [29], the combination appears to be a free-flowing powder
which remains in that state throughout the low-temperature chlorination. The hydrogen chloride
(b.p. -84.9C) formed escapes into a scrubber system. The product is recovered directly after
evaporation of the chlorine. At LCl2/PVC ratios of 8 [28] or greater, a slurry results. As the
chlorination proceeds, the polymer slowly dissolves in the LCl2, until at about the 65% chlorine
level, a solution is obtained. The polymer is recovered by allowing the chlorine to evaporate, or
by precipitation with a nonreactive nonsolvent [28, 30].

The advantages of these processes appear to be the attainment of a chlorine distribution at

least as, or more, homogeneous than solution chlorina-

Page 623
tion. At similar chlorine levels, the relative thermal stability of the liquid chlorine CPVCs is
somewhat greater than that of CPVCs prepared by other processes [31]. The disadvantage of
the processes is the relative complexity of the required process equipment.

III. THE STRUCTURE OF CPVCs AND THE MECHANISM OF CHLORINATION

The structure of CPVC, and the mechanism of its formation, have been studied for many years,
often with conflicting results. Indeed, these two topics are generally inseparable in literature
reports.
In the early 1940s, Ohe [32] published a series of papers related to various aspects of CPVC.
On the basis of a reaction of his CPVC with aniline, and the lack of reaction with potassium
iodide, Ohe wrote the structural unit of CPVC as -CH2-CCl2-. Seipold [33] reached the same
conclusion from x-ray diffraction studies.
Infrared spectroscopy has been widely used to study CPVC structure [3439]. These studies
generally conclude that the CH2 carbons in PVC are exclusively chlorinated during the early
stages of chlorination and that some CHCl carbons are chlorinated at later stages to form CH2-CCl2- units. An important assumption leading to these conclusions was that not more than
one chlorine can substitute into each original -CH2-CHCl- unit.
In 1967, Petersen and Ranby [40] were the first to report the application of nuclear magnetic
resonance (NMR) spectroscopy to CPVC structural and mechanistic problems. They
interpreted their 60-MHz 1H NMR spectra in the following way. In the early stages of
chlorination, only the CH2 carbons of PVC are substituted to form -CHCl-CHCl- units. As the
chlorination proceeds, some substitution at the CHCl carbons of PVC occurs, leading to -CH2CCl2- units, which are never found to be adjacent. Svegliado and Grandi [41] also used 60-MHz
1H NMR to argue that while both CH2 and CHCl carbons in PVC are chlorinated, there is a
strong preference for chlorination of the CH2. Their measured ratio of CH2/CHCl chlorination
varied between 2.0 and 8.7 depending on the degree of chlorination. They further concluded
that the -CHCl-CHCl- units are only found between two unchlorinated units, or at the end of a
chlorinated sequence. Again, the assumption that only one chlorine can enter an original -CH2CHCl- unit was made in these NMR studies. Several other 1H NMR reports have appeared,
generally supporting the earlier conclusions [8, 42, 43].
At this time the composition of CPVC was usually described in terms of the structural units CH2-CHCl-, -CHCl-CHCl-, and -CH2-CCl2-. Compositional tables were published in the leading
articles, and physical properties were addressed from the terpolymer point of view. Dannis and
Ramp [1] attempted to explain the variation in the glass transition temperature (Tg) of CPVCs
using a ternary composition diagram with Tg values of the homopolymers (PVC, polyvinylidene
chloride, and poly-1, 2-dichloroethylene) at the corners. But as they pointed out, the
preparation of an authentic sample of poly -1, 2-dichloroethylene is open to question, as are
estimates of its properties.

Tsuge et al. [44,45] were among the first to suggest that more than one chlorine could be
substituted into an original -CH2-CHCl- unit. They proposed the presence of a fourth unit, CHCl-CCl2-, in CPVCs containing at least 65% chlorine. Their conclusion was based on a
series of pyrolysis-

Page 624

FIGURE 2 Proposed [47] elimination-addition mechanism for the formation of CCl2 carbons.
gas chromatography experiments in which the content of polychlorinated aromatics, resulting
from chain cyclization, was measured in the pyrolyzate.
An extensive series of papers by Kolinsky and coworkers [46] appeared during the 1970s
addressing the problems of structure and mechanism through the chlorination of deuterated
PVCs (PVC--d and PVC-, , -d2). Infrared, 1H NMR, and mass spectroscopy were used in
the studies. In a 1981 summary of this work [47], they concluded that (1) both CH2 and CHCl
carbons in PVC participate in chlorination but not necessarily through direct substitution; (2)
CCl2 carbons in CPVC arise through an elimination-addition mechanism (Figure 2). and are of
the -CHCl-CCl2- type; and (3) formation of these CCl2 carbons occurs only during the later
stages of chlorination. Conclusion 2 precludes the direct substitution of chlorine on a CHCl
carbon and the existence of isolated -CH2-CCl2- units. Kolinsky's results thus support the idea
of more than one chlorine per original -CH2-CHCl- unit, but not by direct substitution. The
elimination-addition mechanism has also been broadened recently [48] to include the formation
of -CHCl-CHCl- units via chlorine addition to double bonds (Figure 3).
While the infrared and 1H NMR-based research on unraveling CPVC structure and mechanism
served as a guide for workers in the area, much of it must be reexamined in light of the results
obtained using high-field 13C NMR spectroscopy.
In the mid-1970s, Keller and coworkers [49] began to report on the use of 22.6-MHz 13C NMR
to examine the microstructure of chlorinated polymers. The obvious advantage of 13C NMR
over 1H NMR, and other techniques, is that all three types of carbons (CH2, CHCl, CCl2) may
be observed directly. In suitable NMR experiments, their concentrations may be measured
quantitatively. In addition, the sensitivity of the 13C chemical shift to substituent effects, both
short and long range, is well documented [50]. Keller changed the perception of CPVC
structure by replacing the classical structural unit representations with five-carbon sequences,
the 13C chemical shift being that

FIGURE 3 Proposed (48] elimination-addition mechanism for the formation of -CHCl-CHClunits.

Page 625

FIGURE 4 Numerical designations of the five-carbon sequences in PVC.

of the center carbon atom. In doing so, he made extensive use of a numerical shorthand
system based on the number of chlorines substituted on the various carbons (0 = CH2, 1 =
CHCl, 2 = CCl2). Thus the CH2 carbons in PVC would center [01010] sequences (Figure 4),
whereas the CHCl carbons would center [10101] sequences. Configurational isomerism in the
sequences was ignored.
Through the use of model compounds and copolymers, Keller developed a set of 13C chemical
shift parameters which were used to predict the chemical shifts of many of the five-carbon
sequences he considered. Unfortunately, the structural complexity of the polymers, and the
unfavorable NMR properties of CCl2 carbons (low nuclear Overhauser enhancement, long spinlattice relaxation times), made the acquisition of high signal-to-noise spectra very difficult with
the 13C NMR equipment available prior to the introduction of wide-bore, cryomagnet
spectrometers.
The first study of CPVC structure using a cryomagnet-based, wide-bore (20-mm) 13C NMR
system at 50.3 MHz was reported by Komoroski et al. [51]. Based on the spectra of model
compounds, and the spectrum of a 60.1% chlorine vinyl chloride -1, 2-dichloroethylene
copolymer, they showed that the first observable chlorinated sequences in a 57.7% chlorine,
solution CPVC are the three configurational isomers of [01110]. The three configurational
isomers (Figure 5) were present in the ratio predicted based on random chlorination.

FIGURE 5 Formation of the configurational isomers of the sequence [01110].

Page 626
This fact indicates that initially the meso and racemic diads are being chlorinated at the same
rate. The first observable CCl2-centered sequence was detected in a 58.7% chlorine CPVC of
the same series. By comparing the chemical shift of the initial CCl2-centered sequence with
those of well-NMR-characterized vinyl chloride-vinylidene chloride copolymers [52], they
concluded that the first CCl2 carbon observed occurred in the sequence [10201]. This
sequence must be formed by direct substitution at a CHCl carbon. No direct spectroscopic
evidence was observed for the presence of double bonds at any chlorine content. Thus the
earlier work depicting CCl2 formation to be occurring only during the later stages of
chlorination, and the work of Kolinsky [47] in regard to CCl2 formation via an eliminationaddition mechanism, yielding only -CHCl-CCl2- units, must be reexamined.
The 50.3-MHz 13C NMR spectrum of a commercial CPVC (66% chlorine), produced by a
water-slurry process, is shown in Figure 6. Considerably more detail is apparent than in
previously reported spectra [49]. Peak assignments for the five-carbon sequences are also
indicated in Figure 6.
At least five types of CCl2-centered sequences are present in this commercial CPVC (Figure
6a). The sequences [10201] and [01210] are most prominent. Also present is the sequence
[10202], assigned based on the spectra of vinyl chloride-vinylidene chloride copolymers [52].
The presence of this sequence (i.e., adjacent -CH2-CCl2- units) was not considered possible in
earlier studies [40, 47]. While sequences of the type [01210], [11210], and [11201] could arise
from the elimination-addition mechanism proposed by Kolinsky [47], the presence of isolated
CCl2 carbons ([10201], [10202]) is difficult to rationalize except through a direct substitution
mechanism. In addition, sequences containing the [210], [211], [120], or [121] segments, all of
which can be thought of as representing the substitution of more than one chlorine into an
original -CH2-CHCl- unit, are observed at about 60% chlorine. This chlorine level is well below
that previously suggested [44, 45, 47] for the occurrence of these segments.
The results for low-chlorine CPVC and the vinyl chloride -1, 2-dichloro-ethylene copolymer
indicate that configurational isomerism of the chlorinated segments plays a major role in both
the CHCl (Figure 6b) and CH2 (Figure 6c) spectral regions. Isomers of the [XlllX] type appear
to contribute to every peak between 58 and 70 ppm. More highly chlorinated CCl2-containing
sequences also appear in this region of the spectrum, and it is difficult to determine where, and
to what extent, these structures contribute to the spectrum. Sequence assignments [51] in
Figure 6b and c were made based on the spectra of low-chlorine CPVC, predicted chemical
shifts, model copolymers, and the known configurational influences. The configurational
influence is especially strong in the case of CH2-centered sequences. Regional assignments
appear to be the best approach at this time.
Of particular interest in these high-field 13C NMR spectra is the presence of residual PVC,
observable as syndiotactic (s), heterotactic (h), and isotactic (i) triads of -CH2-CHCl- units in
Figure 6b. These triads can be estimated quantitatively, although as the chlorine content of the
CPVC increases, this becomes more difficult. From these estimates it was determined that this
water-slurry CPVC has a higher relative syndiotactic content than its parent PVC.

Mechanistically, this observation indicates that the crystalline regions of the PVC, containing a
high proportion of syndiotactic sequences, are less easily penetrated by chlorine (or chlorine
radicals). In solution chlorination, by contrast, it might be expected that all triads would be
chlorinated at much the same rate since all segments of the chain are thought

Page 627

FIGURE 6 13C NMR spectral regions of Geon 603 x 560 CPVC (66% chlorine): (a) CCl2
region, (b) CHCl region, (c) CH2 region.

Page 628

FIGURE 7 Syndiotactic PVC content of CPVCs as a function of weight percent chlorine.

to be equally accessible. In Figure 7, the relative syndiotactic content, s/ (s + h + i), as
determined by 13C NMR, is shown as a function of chlorine content for two series of samples
obtained from water-slurry and solution chlorinations. It should be pointed out that the accuracy
of these measurements is directly related to the amount of residual PVC in each sample.
Since there is more residual PVC in the water-slurry samples, the syndiotactic content in these
samples can be measured more accurately than in the solution samples. But as can be seen
from the data and trend lines in Figure 7, the relative syndiotacticity of the CPVCs from each
process eventually increases with increasing chlorine content. However, the syndiotactic content
of the solution samples remains relatively constant at chlorine contents of less than 62 to 63%.
Thus, in solution, at chlorine contents of less than 62 to 63%, all triads appear to be chlorinated
at about the same rate. This observation is not in agreement with the conclusions of Millan [53],
who reported that even in solution there is a tendency for chlorination to occur in the
heterotactic and isotactic triads due to conformational effects.
Although the microstructure of CPVC can be depicted in terms of five-carbon sequences
(Figure 6), it is extremely difficult to measure the concentration of the sequences quantitatively.
However, since many of the sequences have identical three-carbon centers, Komoroski et al.
[54] have developed a quantitative analysis for CPVC based on three-carbon sequences.
Ignoring configurational isomers, there are 15 possible three-carbon sequences (Table 1).
Only three CH2-centered sequences can occur in CPVC since adjacent methylenes do not
occur in PVC (neglecting any small amount of head-to-head

Page 629
2-centered CHCl-centered CCl2-centered [101] [010] [020] [201] [110] [120] [202] [111] [121]
[210] [220] [211] [221] [212] [222]
polymer postulated to be present). Of the six possible CHCl-centered sequences, all have been
observed except the highly chlorinated [212]. This sequence is not considered in the analysis.
Although structures with adjacent CCl2 carbons are possible, resonances assignable to such
sequences are not observed in commercial CPVCs and are likewise not considered in the
analysis. Thus the sequence distribution analysis describes CPVC in terms of 11 three-carbon
sequences. In this analysis it is also possible to calculate the average sequence lengths of
unchlorinated -CH2-CHCl- units (and thereby estimate the amount of residual PVC occurring in
triads), and to estimate the average sequence length of CHCl carbons.
The origin of each three-carbon sequence is shown in Figure 8. In Figure 8, the solid lines
indicate sequence formation pathways originating at the CHCl carbons of PVC [010]), while the
dashed lines indicate pathways originating at the CH2 carbons of PVC ([101]). It should be
noted that two of the CCl2-centered sequences, [020] and [120], must originate from [010] in
PVC, while the third, [121], can arise from either starting point. This scheme assumes that
substitution is the dominate mechanism in CPVC formation.
As indicated in the earlier discussion, a central question in past CPVC structure and mechanism
studies has been the relative rates of chlorination of CH2 and CHCl carbons in PVC. Since it
has been shown that both types of carbons are substituted by chlorine, measurement of the
ratio of the concentration of [111] to that of [020], at low overall conversion, should provide an
excellent estimate of the initial relative reactivity of chlorine radicals toward CH2 and CHCl
carbons in PVC. The ratio [111] / [020] was found to be 9.8 for solution chlorination (Figure 9,
least-squares fit) at 80C [51]. By contrast, the ratio for the same series of water-slurry
CPVCs shown in Figure 7 was found to be 5.3 (Figure 9, least-squares fit) at 80C [51]. By
where all carbons can be conceived to be equally available for substitution, the CH2 carbon in
[101] is chlorinated about 10 times faster than the CHCl carbon in [010] under the specified
conditions [51]. In the water-slurry system, where chlorine (or chlorine radicals) must diffuse
into solid PVC particles, the selectivity is reduced so that the CH2 carbon is chlorinated only
five times faster than the CHCl carbon. This reactivity ratio is representative only as long as the
reacting polymer continues to resemble PVC. As the chlorine content increases, the effect of
increasing numbers of neighboring chlorines will undoubtedly modify the relative rates at which
each type of carbon in each sequence can be substituted [41, 55]. As a result of this

Page 630

FIGURE 8 Pathways for the formation of the three-carbon structural sequences in CPVCs.
*Usually not observed.
reactivity ratio difference, it would be expected that, at the same degree of chlorination, CPVC
produced in a heterogeneous process would contain a higher proportion of CCl2-type
sequences than a CPVC produced in a homogeneous process (Figure 10). By the same
reasoning, a CPVC produced in a heterogeneous process would be expected to contain more
residual PVC of greater syndiotactic content (Figure 11). Gianelos and Grulke [22] have shown,
using x-ray photoelectron spectroscopy, that CPVC prepared in a heterogeneous process is
surface chlorinated to within 55% of its final value, while the overall chlorination is only about
10% complete.
In Section II, several different processes for preparing CPVCs were described. In the present
section, the use of 13C NMR as the method of choice for determining the structure of CPVCs
has been discussed. In Table 2 are presented the comparative 13C NMR analyses [54] for five
CPVCs prepared by five different experimental processes at B. F. Goodrich. The samples were
chlorinated to the same level (about 66% chlorine) to attempt to eliminate structural differences
due to overall chlorine content.
Terms that appear frequently in CPVC literature are homogeneity and heterogeneity, with
this chapter being no exception. These terms are never defined in regard to the chlorination of
PVC except in terms of phrases, such as homogeneous distribution of chlorines or
heterogeneously chlorinated. The probable intent of these phrases is to cause the reader to
imagine a polymer with one chlorine on each carbon, or, for example, a situation where the

Page 631

FIGURE 9 Reactivity ratios of CH2 vs. CHCl at low chlorine conversion based on the
concentrations of the sequences [111] and [020].
chlorine distribution within a given segment of 10 carbons is identical with every other 10-carbon
segment. Any other situation would thus be imagined as being heterogeneous. Examination of
these concepts in light of the available sequence analyses and recent mechanistic discussions
[51] reveals that defining homogeneity or heterogeneity in CPVC is not a simple matter.
Chlorination of solid PVC particles, as in the water-slurry or fluid-bed processes, can yield a
product in which all polymer chains have not been chlorinated to the same degree. It is
generally assumed that chlorination in solution is sufficiently random so that all chains are
chlorinated to the same degree. On this basis, one criterion of homogeneity when comparing
samples at the same degree of chlorination might be the amount of residual PVC in each
sample. The lower the residual PVC, the more homogeneous the chlorination. Another criterion
might be the average sequence length of vinyl chloride units, n10. A third might be the total
amount of CCl2 carbons since heterogeneous processes would be expected to generate more
CCl2 carbons near the surface of the particles. Examination of the data in Table 2 reveals that
the process LCl2-2 yields a more homogeneous CPVC than the solution process based on
the residual PVC criterion. On the other hand, the process LCl2-1 yields a more homogeneous
polymer on the basis of the n10 criterion. The fluid-bed process gives the most heterogeneous
CPVC, based on its CCl2 content.

Page 632

FIGURE 10 Concentration of CCl2 carbons as a function of weight percent chlorine.

IV. THERMAL STABILITY AND STABILIZATION

The thermal stability of CPVC has been reported to be both superior [56, 57] and inferior [57,
58] to that of PVC. Berticat [56] compared the percent weight loss from PVC with that of a
high-chlorine CPVC (73 to 74% chlorine) using thermogravimetric analysis in an inert
atmosphere. At 160C, no weight loss from the CPVC was observed. At higher temperatures,
190C and 215C, the ratios of PVC weight loss to CPVC weight loss were 47 (14:0.3) and 17
(31: 1.8), respectively.
Minsker and coworkers [57] studied two series of CPVCs, one prepared in water-slurry and the
second prepared in solution. They concluded that the heat stability (measured by the overall
rate of dehydrochlorination at constant temperature) of CPVCs is not determined by molecular
weight or total double-bond content, nor can it be related to the ratios of the commonly
considered structural units. Time of chlorination and chlorination conditions were determined to
be the major factors related to the measured heat stability. At low chlorine conversion and short
times, the CPVC formed was found to be unstable relative to PVC. This instability was
attributed to the formation of double bonds. At longer chlorination times the CPVCs formed
were more stable than PVC, but at high chlorine conversion, the polymers were slightly less
stable than PVC, regardless of the chlorination time. A logarithmic relationship between rate of
dehydrochlorination and the number of internal double bonds was proposed. They further
proposed that a CPVC with en-

Page 633

FIGURE 11 Residual PVC in CPVCs as a function of weight percent chlorine.

hanced stability can be produced if the polymer does not undergo molecular weight reduction,
has a low concentration of internal double bonds, and has a high degree of chlorination.
Millan [58] determined that the thermal stability (degradation by heating at 15C per minute to
500 to 600C) of CPVCs was related to structural heterogeneity, which can be influenced by
the degree of chlorination and the tacticity of the parent PVC. Of the samples studied, the
CPVCs prepared from the less syndiotactic PVC (51% syndiotactic) had the best thermal
stabilities. Higher syndiotactic PVCs were believed to yield heterogeneous, and less stable,
CPVCs. The resistance of the syndiotactic segments to chlorine penetration is undoubtedly a
major factor in these experiments.
The photodegradation of a CPVC (64.8% chlorine) has been extensively studied by Decker and
Balandier [59], both in solution and in the solid state (pressed films). Upon ultraviolet exposure,
the CPVC film undergoes a rapid dehydrochlorination reaction which develops 10 times more
efficiently than in PVC. Long polyene sequences are formed, and simultaneous chain scission
and cross-linking are observed, independent of either the presence or absence of oxygen. They
concluded that CPVC is much less stable than PVC to ultraviolet radiation.
The intrinsic thermal stability of chlorinated polymers cannot easily be defined nor measured.
Although, based on the preceding discussion, we perhaps have a vague notion of what is meant
by intrinsic thermal stability, an exact definition remains elusive. Part of the difficulty is that
thermal stability must be quantified using some type of physical test; it is not easy to simply
draw the chemical structures of chlorinated polymers, and then

Page 634
13C NMR Analyses of CPVCs Prepared by Various Experimental Processes Parameter
Solution LCl2-1 LCl2-2 H2O-1 H2O-2 Fluid bed Processa a b c d e f % Cl 65.97 65.18 66.96
66.20 66.50 66.37 Mol % CH2 32.8 33.3 29.8 32.2 33.1 37.2 Mol % CHCl 61.7 63.0 64.6 62.2
59.4 51.7 Mol % CCl2 5.5 3.7 5.6 5.7 7.4 11.1 % PVC (triads) 19.5 22.5 18.9 25.4 27.5 28.5
% Syndiotactic PVC 40.0 34.1 38.0 39.7 42.6 44.0 n10 2.2 2.1 2.2 2.5 2.6 2.6 aProcesses: a.
Tetrachloroethane, 80C, 0.25 9 of Cl2 per minute, UV initiation. b. By a process similar to that
described in reference 28. c. By a process similar to that described in reference 29. d. By a
process similar to that described in reference 11. e. By a process similar to that described in
reference 20. f. Chlorine concentration gradually increased (2 to 100%) over 360 min at 40C,
UV initiation.

Page 635
claim that one is more intrinsically stable than another. What test should be used? Perhaps a
high-temperature, static test, such as the rate of hydrogen chloride elimination [57], can be
used to relate stability to chemical structure. Perhaps a small-scale thermomechanical test,
such as available on instruments like the Brabender Plasti-corder [62], give the best indication
of polymer stability and processability. In this type of testing, stability and processability
become increasingly difficult to separate. Ultimately, a large-scale test on commercial
processing equipment will tell a polymer producer whether or not a product is salable.
Unfortunately, most early-stage experimental materials cannot be produced in the quantities
necessary to perform these large-scale tests. So we resort to the use of small-scale tests and
the concept of relative thermal stability/processability wherein experimental polymers are
compared with existing commercial polymers of known performance.
The measurement of CPVC relative thermal stability is particularly vexing compared with the
measurement of PVC stability. Each type of PVC is a nearly pure homopolymer whose
measured property values, such as Tg, melting point, viscosity, and melt flow, generally occur
within narrow ranges. Therefore, it is relatively straightforward to arrive at a standard set of
testing conditions. On the other hand, CPVC properties vary with chlorine content and
chlorination method. For example, two CPVCs of the same chlorine content, made by different
processes, may have different Tg values and melting points. How can thermomechanical testing
of these two materials at a fixed temperature be meaningfully related to thermal stability?
Again, how is the relative thermal stability of a series of CPVCs of differing chlorine content,
made by the same process, measured when each member of the series has a different T and
melting point? As an example, measurement [60] of the dehydrochlorination curves at 170C for
a series of water-slurry CPVCs yields the set of curves shown in Figure 12. On this basis it
could be concluded that the higher the chlorine content, the greater the relative thermal stability
of the CPVC (in this test). Occasionally, dehydrochlorination curves from low-chlorine CPVCs
are observed [60] that support Minsker's [57] report that these kinds of polymers are less
stable than PVC.
The choice of test temperature is critical. PVC is usually processed at temperatures at least
80C greater than its Tg thus the curves in Figure 12 were determined at 170C. If the
dehydrochlorination curve for each CPVC in the series is remeasured at its Tg plus 80C, the
set of curves in Figure 13 is obtained. If this were the only set of data available, it would be just
as easy to conclude that the relative thermal stability decreases with chlorine content. These
results point out that the relative thermal stability testing of CPVC resins and compounds must
be approached in a very careful and concise manner.
The primary commercial measure of the relative thermal stability and processability of CPVC
compounds (not resins) is the dynamic thermal stability (DTS) test. This test is designed to
measure the time-torque relationship (Figure 14) at selected temperatures using an instrument
such as the Brabender Plasti-corder [62]. The test value generally reported, and used for
comparison, is the DTS time. DTS time is usually defined as the time (tm) required for the
instrument torque to fall to its minimum value, with the polymer compound in the melted state,
before beginning to increase, presumably due to cross-linking. Alternatively, DTS times based

on tm2, tm+100, or td may be used in some applications. DTS time is dependent not only on
polymer properties, but also on temperature, sample size, stabilizers, lubricants, instrument
operating conditions, degree of instrument maintenance, and other

Page 636

FIGURE 12 Dehydrochlorination curves for a series of water-slurry CPVCs determined at

170C. (Courtesy of G. S. Huvard, B. F. Goodrich Co.)
factors. All of these variables must be closely controlled to maintain reproducibility.
The stabilization of CPVC compounds finds its roots, quite naturally, in PVC stabilization
technology. There have been very few literature reports dealing directly with the stabilization of
CPVCs. Most of the CPVC-compounding technology appearing in the patent literature is quite
similar to that known for the same type of PVC application. As with PVC compounds, organotin
stabilizers are widely used in CPVC compounds. A series of patents by Hall [61] has recently
disclosed the use of metal phosphates, in combination with a primary stabilizer, to give CPVC
compounds of enhanced stability.

V. PROPERTIES
A. Chemical-Resistant Properties
Although the most important reason for chlorinating PVC is to provide a polymer with an
increased glass transition temperature, the excellent chemical resistant properties of CPVCs
cannot be overlooked. Chemical resistance is one of the most important properties of CPVC,
especially CPVC-based pipe. Corrosive fluids which soften or attack a thermoplastic material
also reduce the long-term strength of that material.
Because of the high cost and difficulty in obtaining actual in-use data with corrosive fluids, an
immersion screening test has historically been used.

Page 637

FIGURE 13 Dehydrochlorination curves for a series of water-slurry CPVCs determined at

CPVC-Tg + 80C. (Courtesy of G. S. Huvard, B. F. Goodrich Co.)
This test determines the response of the CPVC to the test fluid by measuring some simple
properties, such as percent change in weight and/or volume. Typical 28-day fluid immersion test
values for a CPVC pipe extrusion compound are presented in Table 3.
In Table 3, changes between 0 and 1.5% generally indicate satisfactory chemical resistance.
Changes greater than 4% are generally interpreted as unacceptable. Changes between 1.5
and 4% suggest a need for caution and further testing.
Three generalizations can be drawn from these immersion tests. First, CPVC is generally
resistant to most mineral acids, bases, salts, and paraffinic hydrocarbons. Second, CPVC is
not recommended for use with most polar organic materials, including some solvents
(chlorinated hydrocarbons, aromatic hydrocarbons, esters, and ketones), organic acids, and
higher alcohols. Third, the resistance of CPVC to certain fluid mixtures, such as fuel oils with
aromatic content, cannot be determined on the basis of immersion testing alone. For these
classes of materials, actual use data must be obtained.
It should also be noted that in addition to temperature and test fluid concentration, other
factors, such as stress level and finished product quality, can also affect the measured
chemical resistance of a plastic material. Because of this, the final determination of suitability
must depend on some in-service testing, which must be carried out with care and safety.

Page 638

FIGURE 14 Idealized dynamic thermal stability (DTS) curve for CPVCs. (Courtesy of G. T.
Dalal, B. F. Goodrich Co.)
B. Thermal Properties
The thermal properties of CPVCs, such as glass transition temperature (Tg) heat distortion
temperature, and softening point, generally increase with increasing chlorine content [1].
At the same degree of chlorination, Trautvetter [10] observed that a CPVC, prepared by the
water-slurry process from a PVC containing a greater than normal amount of crystallinity,
exhibited a higher Vicat [64] softening point than did a CPVC similarly prepared from a
commercial PVC of lower crystallinity. However, when the two PVCs were chlorinated in
solution, both CPVCs exhibited the same Vicat softening point. Trautvetter's interpretation of
these results was that in solution all portions of the PVC were accessible to chlorination,
whereas in the water-slurry process, chlorination occurred primarily in the amorphous regions
because chlorine could not penetrate the crystallites in the solid particles. Thermal properties of
CPVCs were thus related to the method of chlorination.
Beginning with the classical differential scanning calorimetry study on PVC by Illers [65], the
thermal properties of PVC have been studied extensively [66]. Similar studies on CPVC have
received much less attention until recently. The specific heat capacity of a CPVC containing
66.8% chlorine was measured up to 150C [67], with the conclusion that it contained no
crystallinity. On the other hand, residual crystallinity in annealed commercial samples (68%
chlorine) made by fluid-bed chlorination has been reported [68]. Correlation of the glass
transition temperature (Tg) with degree of chlorination

Page 639

b % Wc % V % W Distilled water +1.12 +0.65 +3.82 +2.58 Acetic acid (20%) +0.81 +0.55
+3.21 +2.13 Acetic acid (glacial) +6.00 +4.61 +54.1 +38.2 Ammonium hydroxide (conc.) +17.0
+9.86 *d * Ammonium persulfate (sat.) +0.34 +0.20 * * Aqua regia +2.25 +1.78 -0.07 +0.07
Chromic acid (50%) +0.28 +0.08 -1.09 -1.86 Hydrochloric acid (37%) +0.56 +0.42 +1.98 +1.59
Hydrofluoric acid (10%) +3.34 +1.89 * * Hydrofluoric acid (40%) +7.95 +5.10 * * Nitric acid
(70%) +0.36 +0.52 +0.72 +1.19 Phosphoric acid (70%) +0.04 -0.01 +0.11 +0.02 Potassium
hydroxide (sat.) +0.12 -0.03 * * Sodium chloride (sat.) +0.38 +0.14 +0.72 +0.51 Sodium
chlorite (sat.) +1.09 +0.65 * * Sodium hypochlorite +0.92 +0.39 * * Sodium hydroxide (50%)
-0.06 -0.02 +0.04 +0.14 aTempRite 3007, B.F. Goodrich. bpercent change in volume.
cpercent change in weight. dAsterisks indicate not tested, not recommended. Source: Ref. 63.
for both water-slurry and solution-chlorinated polymers has also been reported [69].
An extensive study of the thermal properties of a series of solution CPVCs has been performed
by Lehr [70] at B. F. Goodrich. Lehr found that the Tg increases nearly linearly with chlorine
content (Figure 15). He also observed that the measured heats of melting of a series of
prefused samples [71] decrease very rapidly between 57 and 62% chlorine ( Hm2 in Figure
15). At this point, the original crystallinity had been reduced about 95%. Further increases in
chlorine content (67.5% chlorine) reduced the measurable crystallinity to near zero. This
reduction in crystallinity was further confirmed by the absence of detectable syndiotactic triads
in the 13C NMR spectra of the high-chlorine samples.
Lehr also noted that chlorination of PVC caused a significant reduction in the heat capacity
change at Tg when the Tg exceeded 100C. To interpret the effects in terms of molecular
behavior, he recalculated the data in terms of molar heat capacity per bead [72], with the aid of
13C NMR data. The bead molar heat capacity changed from 13.1 J/mol K for 100%
amorphous PVC to 7.7 J/mol K for a 69.6% chlorine CPVC, a decrease of 41%. This is not in
conformity with the constant-heat-capacity rule [73]. The major effect in the decrease is
believed to be the change in conformational heat capacity. This seems reasonable in view of
the greater steric hinderance expected upon sub-

Page 640

FIGURE 15 Glass transition temperature (Tg) and heat of melting ( Hm2) of CPVCs as a
function of weight percent chlorine.
stitution of a chlorine atom for a hydrogen atom. Chain stiffness would also increase.
C. Rheological Properties
Among the first published studies of CPVC rheology were those performed by Thamby and
coworkers [74] on a series, of CPVCs prepared by the water-slurry-chloroform process [11].
They measured the flow behavior of a series of lightly stabilized CPVC melts in a capillary
rheometer using dies with L/D ratios of up to 10, and determined the melt viscosities and
activation energies in the temperature range 190 to 210C. These parameters were related to
the cohesive energy density of the CPVCs. For CPVCs with chlorine contents of greater than
67%, melt fracture of the extrudate occurred, even at low shear rates. This phenomenon was
attributed, in part, to melt elasticity resulting from cross-linking during the thermal
dehydrochlorination reaction typical of most chlorinated polymers. Melt fracture of PVC, at low
shear rates, has also been observed in many investigations. For CPVCs with chlorine contents
in the range 63 to 65%, melt fracture at low shear rates is somewhat diminished [75].
Bonnebat and DeVries [68] determined the real and imaginary components of the complex
shear viscosity as a function of strain amplitude, frequency, and temperature for a series of
fluid-bed CPVCs (61.8 to 68.3% chlorine). These parameters were obtained using a
mechanical spectrometer with an eccentric rotating disk. They noted the importance of
temperature and thermal

Page 641
history on the samples used to obtain their data. Their results were interpreted based on the
presence of microcrystallites in the samples, an interpretation consistent with differential
scanning calorimetry results [68, 70].
The temperature-dependent viscoelastic properties of a commercial CPVC (67% chlorine)
prepared by a water-slurry process have been investigated by Harrell and Kumler [76] at B. F.
Goodrich. They have identified three distinct regions of behavior in relating the storage modulus
(GI), loss modulus (G), and complex viscosity (| n* |) to temperature (Figure 16). As the
temperature decreases from 150C to 135C, both G and G increase rapidly in magnitude,
and G tends to converge with G. This type of behavior is characteristic of glass transition
behavior where chain mobility becomes restricted. The large increase in G with decreasing
temperature is the result of vibrational resonance within the molecular segments, which
produces more efficient energy-loss mechanisms. A second type of behavior is observed over
the temperature range 160 to 200C. The G response decreases slowly with increasing
temperature, whereas G slightly increases. This behavior is characteristic of the rubbery
plateau over which G is the dominant response component. Within this region of behavior,
molecular flow is retarded by the cross-links provided by residual crystallinity and chain
entanglements. The third type of behavior is observed at temperatures in excess of 210C. All
three parameters decrease rapidly with temperature. Additionally, the G response is similar to,
or even larger than G, indicating that viscous flow is occurring on the molecular level. At these
temperatures, crystalline domains have been eliminated, or if still present, are ineffective crosslink sites.

FIGURE 16 Dynamic mechanical properties of CPVCs as a function of temperature at fixed

frequency. (Courtesy of E. R. Harrell, B. F. Goodrich Co.)

Page 642
D. Mechanical Properties [1]
Plastics have value primarily for the mechanical properties developed in the finished product.
Since many engineering properties of CPVC products are related to the particular compound
being used, a few underlying principles are needed to illustrate a wide range of data. These
concepts are best presented by discussing the stress-strain properties of compounds made
using CPVC.
Typical stress-strain curves for PVC and CPVC compounds are shown in Figure 17. The curve
for PVC can be divided into three distinct regions of behavior. In the AB region, the behavior is
almost that of a linear solid, showing a Young's modulus relation almost independent of the rate
of extension. This is the region that defines the normal use of the material. Point C is the yield
point of the material, approached through the transition region BC. This definition of yield is
related to the convenience of ordinary tensile test machines. The yield point is the maximum in
the curve. Extension after yield continues until the final failure occurs at D along CD. The point
D is not a single-valued definite extension that can be determined with a high degree of
accuracy, but is expressed by its own distribution function. This is because tensile failure is
related to the probability of encountering an internal flaw or defect in the test piece. While
extension at break is an important parameter of measurement, it should be noted that practical
failure has already occurred at the yield point C. This is the point at which the material changes
dimensions so drastically that sizing has been lost.

FIGURE 17 Typical stress-strain curves for PVC and CPVC compounds.

Page 643
Commercial compounds made from CPVC vary widely from one another. In addition, changes
in chlorine content and the chemical structure can introduce further variables in compounding.
Each of these variables affects the different regions of the general stress-strain curve in a
different way.
The linear portion, AB, is affected by both compounding additives and the structure of the
polymer. As the softening point increases, Young's modulus also increases. The increase is
about 30 to 40% for commercial CPVCs. Much of this increase in modulus is canceled by the
compounding ingredients used in commercial compounds.
The yield stress of CPVC resins changes greatly with chlorination, as does the work to yield,
the area under the stress-strain curve up to the yield point. The net effect is about a 50-psi
increase in yield strength per degree of Tg increase [70, 77]. Some typical property values for
CPVC compounds are presented in Table 4.
The elongation-to-break, the CD region of the curve, normally decreases with increasing
chlorine content. This is consistent with the observation that the transition in PVC (about
-50C), as observed in dynamic mechanical testing, is systematically shifted to higher
temperature with increasing chlorine content [78]. The ductility of PVC has been attributed to its
low-temperature transition. Thus, when the transition is shifted to higher temperatures by
chlorination, some loss of ductility is expected [79].
a Profile extrusionb Wt % Cl Schoninger 67.0 63.5 Inherent viscosity ASTM D1243 1.0 Specific
gravity ASTM D792 73F g/cm3 1.55 1.44 Vicat softening point ASTM D1525 C 148 Bulk
density (approx) lb/ft3 28 Rockwell hardness ASTM D785 73F R 120 115 Heat distortion
ASTM D648 264 psi F 221 185 Tensile strength ASTM D638 73F psi 8,250 7,250 Tensile
modulus ASTM D638 73F psi 410,000 345,000 Flexural strength ASTM D790 73F psi 15,800
13,575 Flexural modulus ASTM D790 73F psi 420,000 366,700 Izod impact ASTM D256 73F
ft-1b/in. 1.8 7.0 Coefficient of thermal expansion ASTM D696 10-5 in. /in.-F 3.4 4.1 Dielectric
strength ASTM D149 V/mil 1,170 1,300 Dielectric constant ASTM D150 1000 Hz 3.25 3.2
Power factor ASTM D150 1000 Hz 0.007 0.014 Flammability UL-94 0.062 in. V-0 V-0
aTempRite 3101, B. F. Goodrich. bTempRite 3302, B. F. Goodrich.

Page 644
Compounding additives, such as impact modifiers, can be used to increase the path length, CD.
The result is tougher compounds. The area under the stress-strain curve is the work to
break, that is, the energy required to rupture the sample. Those materials having a high workto-break ratio will be tougher in service. However, since the performance of plastics is time
dependent, predictions of impact performance from laboratory tests can be misleading.
Comparisons of the relative performance of several materials can be made at any test rate, but
predictions of performance in service must be made at rates simulating performance conditions.
E. Dielectric Properties
The dielectric properties of CPVC are of interest from two different points of view. First, its
electrical insulation properties are used daily in many applications. Second, its low-frequency
dielectric constant can aid in characterizing molecular structure.
The effect of chlorine substitution in simple molecules can be predicted from the theories
presented in several classical textbooks [80]. Changes in the dipole moment of the molecule
change its behavior in an alternating field. However, the effect of polar groups in a polymer is
more complicated than in simple molecules. Since segments of polymers are held together with
primary valence bonds, these cannot rotate individually to align with an applied field [81, 82].
The dielectric properties of polymers determine energy storage related to the dielectric
constant (E) and energy loss related to the loss factor (E). Energy storage is related to the
number of dipoles that can be aligned with a steady applied electric field. Energy loss is related
to the work required to move the dipoles in an alternating field against the attractive forces of
their neighbors. The typical dielectric behavior of unplasticized PVC is shown in Figure 18. The
real part of the dielectric curve E has a low value, about 3 at room temperature. It then rises in
the second-order transition (Tg) region to a peak value of about 12, then decreases at higher
temperatures. The associated loss curve (E) goes through a maximum at about 104C at 1000
Hz. The curves are also frequency dependent (the dielectric curve and loss peak shift to higher
temperatures at higher measurement frequencies). The important features of these curves are
that the temperature of the loss peak is characteristic for PVC, and the dielectric constant is a
measure of the effective dipole moment.
Since CPVC is known to contain CCl2 carbons, studies [1] on the dielectric properties of vinyl
chloride-vinylidene chloride copolymers have been used in comparative discussions. Small
amounts of vinylidene chloride lower the temperature at which the loss peak occurs, and
increase its intensity. This is opposite to what might be expected. Based on examples such as
1, 4-dichlorobenzene, introduction of a second chlorine on a carbon might be expected to
cancel the dipole moment of the first. However, since the polymer chain is a three-dimensional
structure, the second chlorine generates a new dipole in a different direction, and the total
dipole moment is a vector sum whose magnitude exceeds the individual moments. Thus the
maximum real and loss parts of the dielectric constant are both larger in the copolymers than in
PVC. From the known melting point of polyvinylidene chloride (180C), It might also be
expected that the copolymers would have higher transition temperatures than PVC. However,
the dominant dielectric effect comes from the amorphous phase, and the lower Tg of

polyvinylidene chloride

Page 645

FIGURE 18 Dielectric behavior of unplasticized PVC. (From Refs. 83 and 84.)

(-20C) [85], lowers the Tg and the temperature of the loss peak in the copolymers.
In contrast, the effect of chlorination on PVC is to decrease the dielectric constant (Figure 19)
and loss (Figure 20) and increase the temperature of the loss peak [1, 86]. Increasing amounts
of chlorine result in a lower effective dipole and a lower loss peak at higher temperatures. The
lower high-temperature dielectric constant in this case may come in part from dipole
cancellation due to chlorine distributions along the chain which differ from the vinylidene chloride
copolymers. However, the principal effect probably comes from increased chain stiffness, which
also increases Tg and the temperature of the loss peak. Increased chain stiffness would also
be consistent with a smaller heat capacity change at Tg [70] with increasing chlorine content
(the heat capacity is proportional to fluctuations in entropy [87], which are smaller for stiffer
chains).
The dielectric properties shown in Figure 19 and 20 are for a group of CPVCs prepared under
similar conditions. CPVCs prepared by different processes, for example, in solution at high
termperature or in liquid chlorine

Page 646

FIGURE 19 Dielectric constant of PVC and CPVCs at 1 kHz as a function of temperature.

(Courtesy of I. M. Hodge, B. F. Goodrich Co.)
at low temperature, may have similar chlorine contents, but they differ in structure and
homogeneity and have different dielectric properties [88].
F. Thermal Decomposition Characteristics
The thermal decomposition of CPVCs has been the subject of several studies. Berticat [56, 89]
reported that the principal volatile pyrolyzate from CPVCs is hydrogen chloride, and the CPVCs
exhibit two temperature-dependent modes of decomposition. The first mode, which occurs in
the temperature range 270 to 300C, is characterized by hydrogen chloride evolution. The
second mode, occurring at temperatures greater than 300C, is characterized by the evolution
of chlorinated hydrocarbons.
At 460C, the volatile pyrolyzates from a series of CPVCs have been reported [44, 45] to
consist of benzene, toluene, naphthalene, chlorobenzene, and various isomers of di-, tri-, and
tetrachlorobenzenes. The chlorobenzene yield reached a maximum when the CPVC contained
about 64% chlorine, and then decreased at higher chlorine levels. At the higher chlorine levels,
di-

Page 647

FIGURE 20 Dielectric loss of PVC and CPVCs at 1 kHz as a function of temperature. (Courtesy
of I. M. Hodge, B. F. Goodrich Co.)
and tlrichlorobenzenes dominated. As pointed out earlier, this study was among the first to
provide evidence for the existence of the structural sequence -CHCl-CCl2- in CPVCs.
Lattimer and coworkers [90] have used pattern-recognition techniques to analyze the pyrolysismass spectroscopy data obtained from a series of CPVCs. At 600C they found that
chloroaromatic compounds increase in both relative and absolute abundance as the chlorine
content of the CPVCs increases. Aromatics, and aromatics with Cl or C2 side chains, also
increase in relative abundance with increasing chlorine level, but actually decrease in absolute
abundance [91]. In addition, they observed that aliphatics, and aromatics with more aliphatic
character, decrease in both relative and absolute abundance with increasing chlorine level.
Also, a linear correlation was noted between the yield of chlorobenzene and the sum of the
three-carbon sequences [111] and [020] for chlorine levels in the range examined (57 to 61%
chlorine). This observation is consistent with the established decomposition pathways [92].
The role of chlorine-containing vinyl polymers in fires is currently a topic of interest since they
have inherent flame-retardant properties not found in other plastics or in wood. A summary of
the current technical knowledge of real fire behavior, with an emphasis on PVC, has recently
been published by Dickens [93].

Page 648
Liebman and coworkers [94] have reported that the smoke evolution from burning CPVC is
reduced compared with PVC, especially at chlorine levels greater than 64%. The reduction in
smoke was attributed to the formation of increased amounts of chlorinated aromatics, which do
not produce as much smoke as benzene. At lower chlorine levels, greater amounts of smokeproducing benzene are generated. In addition, the total amount of volatile pyrolyzates was
lower at higher chlorine levels. They also observed that chlorinated polyethylenes produced
more smoke than CPVCs at the same degree of chlorination (no difference was found when the
chlorine level reached 70%). This difference in smoke levels was attributed to the structural
differences between the polymers, the structures becoming nearly identical at 70% chlorine.
In a preliminary study of the combustion products from chlorinated polymers, O'Mara [95, 96]
found that while a CPVC (68% chlorine) showed low smoke characteristics, it exhibited an
enhanced rate of carbon monoxide formation compared with woods in the NBS smoldering
mode test. When PVC is pyrolyzed at high temperatures in an inert atmosphere, about 4 to
5% carbonaceous char remains. Pyrolysis of CPVC under the same conditions yields about
25% char (accounting for about 90% of the original carbon present). This increased char
formation, resulting in a decreased organic fuel availability, is believed to be a major factor in
the low-smoke characteristics of CPVCs. Since one of the primary mechanisms of carbon
monoxide formation is through oxidation of porous chars, the increased amounts of char
observed for CPVCs may explain the carbon monoxide generation rates [96]. The chemistry
and mechanisms involved in the use of metal smoke retarders, with emphasis on PVC, has
been reviewed by Kroenke [97].

VI. MARKETS AND COMMERCIAL USES

A. Introduction
In the first edition of this work, Dannis and Ramp [1] correctly foresaw that the commercial
uses for CPVC would be in hot and cold water distribution, chemical fluid handling, and in fireresistive applications. For many years, extruded pipes and simple injection-molded fittings were
the only shapes possible. However, advances in CPVC technology have opened new market
opportunities for CPVC.
B. Market Considerations
In the past 20 years, customers have demanded more performance from thermoplastic
polymers, with the result that the markets for plastics having engineering properties have
grown steadily. Compounded annual growth rates for engineering thermoplastics have ranged
from 13 to 21%, compared to the 4 to 5% [98] for the volume thermoplastics. Typically,
customers require better performance parameters to justify replacement of other materials of
construction. The performance parameters demanded are heat resistance (usually described
by heat distortion under load [HDTUL]), creep resistance, chemical resistance, rigidity, and
strength. The engineering plastics compete for market share and price return based on having
price/property balances of these performance parameters to a greater or lesser degree than
their competitors. Advanced formulations of CPVC can compete in this arena.

Page 649
C. Traditional Uses of CPVC
Understanding customer needs is critical to providing products for a particular market. CPVC
compounds for pipe for hot and cold water distribution make an exact fit to the market
requirements (Table 4). Hot water is distributed through buildings at 140 to 180F; CPVC has a
continuous-use rating of 100 psi at 180F. CPVC pipe is readily joinable to a complete line of
molded fittings via simple solvent welding techniques, eliminating the need for torches, solder
pots, heavy tool kits, and so on. CPVC pipe resists water it will not rust or corrode; there is
no scale buildup; it is energy efficient due to its low thermal conductivity; and it does not
sweat. CPVC pipe is approved for potable water handling by the National Sanitation
Foundation. All major plumbing codes recognize its acceptability for hot and cold water
distribution. CPVC compounds formulated for this purpose will meet the requirements of ASTM
D2846 [99]. This standard covers water service up to and including 180F (82C). It is
estimated that over 3 million dwellings have been plumbed with CPVC pipe, and that the use of
CPVC pipe can save 15 to 25% of the total installed cost versus traditional materials [100].
In iron pipe size schedule 80 and 40 sizes, CPVC meets the needs for corrosion resistance for
chemical fluid handling. It has outstanding resistance to acids and bases and is used in process
piping for the chemical, pulp and paper, and electroplating industries. Additionally, valves,
fittings, and tanks have been fabricated from CPVC. Piping for residential low-cost, fastresponse fire sprinkler systems has recently been successfully tested in a full-scale fire test.
D. New Uses for CPVC
For many years, CPVC was unable to approach the engineering injection molding markets
because of its limited processability compared with modified PPO, flame-retarded ABS, or
polycarbonate/ABS alloys occupying that market niche. Technical efforts to broaden the
processability of CPVC have been successful, and highly injection-moldable 185-to-205F
HDTUL products have been introduced [101]. These products have found application in
appliances, business machine components, large filter frames, and whirlpool-bath pump
housings. Since commercial CPVCs contain more than 60% chlorine by weight, they have
definite economic advantages over their all-hydrocarbon competitors.
In addition to the potential for replacing hydrocarbon-based engineering plastic materials,
CPVC has found application among PVC users who require improved resistance to creep,
either at room temperature or elevated temperatures. Inasmuch as the CPVC compounds
designed for advanced engineering uses are new, and the measurement of creep resistance
long and arduous, the studies are not complete at this time. Preliminary results indicate that the
creep resistance of these new CPVC compounds is significantly better than for PVC. At 200F,
a selected CPVC molding formulation retains a tensile modulus of about 300,000 psi while
exhibiting a tensile strength of nearly 3500 psi.
A growing demand for higher heat resistance in thermoplastic window lineals made by profile
extrusion has resulted in the development of CPVC profile extrusion compounds which serve as
a substrate under PVC weatherable capstock. Such profiles are made by coextrusion
processes, testifying to the improved processability of CPVC.

Page 650
Further testimony to the improved processability of CPVC compounds has been the commercial
extrusion of 60- and 72-in., 0.125-in. thick sheet on 4.5-in. extrusion lines at competitive output
rates. Vacuum formability of the resultant sheet product has been characterized as approaching
that of general-purpose ABS. Also, CPVC has been successfully calendered for cooling-tower
fill applications.
Dannis and Ramp [1] referred to the outstanding fire-resistive properties of CPVC. Coupled
with the processability improvements and condensed phase additive technology [97] for smoke
suppression, CPVC compounds have been introduced that meet various fire codes [101].
For example, building codes for certain occupancies require Class I performance for interior
trim and finish as tested by ASTM E84 (NFPA255, UL 723). Values of flame spread of 25 or
less, and smoke development of 450 or less define Class 1 [102]. These flame spread and
smoke values are indices based on asbestos cement as a zero reference, and red oak as a
reference at 100. CPVC performs well in this test protocol, having a flame spread of less than
15 and a smoke development of 160.
Transit authorities are required (e.g., by the U.S. Department of Transportation [103]
requirement for mass-transit rail cars [UMTA] [104]) to provide materials for the interiors of
vehicles which do not produce quantities of smoke that obscure vision to such an extent that
one could not escape from a burning vehicle. ASTM E662 (NFPA 258) is used to test materials
for conveyance interiors. The minimum values of specific optical density (Ds) specified are 100
(maximum) at 90 sec, and 200 (maximum) at 4 min. CPVC products have been designed that
meet these requirements [101].
Resistance to flame spread is also required by transit authorities. This is measured by ASTM
E162, where flame spread and heat evolution are quantified. Authorities require that the
product (Is) of flame spread and the heat evolved not exceed 35 [103, 104]. Typically, CPVC
exhibits an Is value of 4 or less.
Fire performance of CPVC has been described in a series of large-room fire experiments [105].
CPVC performance in these large-scale fires is excellent and confirms the good fire-resistive
properties measured by small-scale tests. In the large-scale test, flame spread along 4 ft x 8 ft
sheets of compounded CPVC is minimal, smoke is low, and the room cannot be brought to
flashover conditions.
E. Summary
In summary, the applications for CPVC predicted by Dannis and Ramp [1] have been achieved.
Over 3 million homes in the United States have been piped with CPVC hot and cold water
distribution systems, and CPVC has found its way into a myriad of chemical processing
applications (valves, pumps, tanks, and piping). With CPVCs recently improved processability,
wide-sheet extrusion has been successfully accomplished. Thermoformed wide CPVC sheet is
finding application in passenger transit rail cars for seat backs, window reveals, and kick
panels. Calendered formulations find use in cooling towers. Injection-molded formulations have

been promoted for appliance parts, business machines, and electrical apparatus. Profile
extrusions for window and skylight frames, glazing beads, and electrical raceways are
commonplace today.

Page 651

ACKNOWLEDGMENTS
We wish to thank the following colleagues at B. F. Goodrich for their assistance, cooperation,
and encouragement: C. N. Bush, G. T. Dalal, E. J. DeWitt, E. D. Dickens, Jr., E. R. Harrell, I.
M. Hodge, G. S. Huvard, T. A. King, R. A. Komoroski, R. P. Lattimer, B. -L. Lee, M. H. Lehr, R.
C. Neuman, A. J. Olson, J. B. Pausch, R. K. Schlatzer, J. P. Shockcor, G. F. Smith, R. G.
Vielhaber, and J. A. Weaver. Special thanks to Ruth A. Spindler for assistance in literature
searching and information gathering; to JoAnn I. Barker for secretarial assistance; and to Arthur
R. Canterbury for the figures.

REFERENCES
1. M. L. Dannis and F. L. Ramp, in Encyclopedia of PVC, L. I. Nass, ed., Marcel Dekker,
New York, 1976, Chap. 6.
2. R. G. Vielhaber, B. F. Goodrich Co., unpublished results.
3. G. H. Arnold, Plast. Polym., 38, 21 (1970).
4. C. A. Brighton, in Encyclopedia of Polymer Science and Technology, N. M. Bikales, ed.,
Vol. 14, Wiley-Interscience, New York, 1971, pp. 460468.
5. T. Suzuki, Pure Appl. Chem., 49, 539 (1977).
6. E. Marechal, in IUPAC Macromolecules, H. Benoit and P. Rempp, eds., Pergamon
Press, Elmsford, N.Y., 1982, pp. 8597.
7. Z. Csuros, M. Groszmann, and B. Zsuffa, Magy. Kim. Foly., 64, 168 (1958).
8. P.-Q.-Tho and P. Berticat, Eur. Polym. J., 4, 265 (1968).
9. R. T. Foster and J. H. Taylor, Imperial Chemical Industries Ltd., U.S. Pat. 3,440,236
(Apr. 22, 1969).
10. W. Trautvetter, Kunstst. Plast., 13, 54 (1966).
11. M. L. Dannis and F. L. Ramp, B. F. Goodrich Co., U.S. Pat. 2,996,489 (Aug. 15, 1961).
12. (a) J. C. Shockney, B. F. Goodrich Co., U.S. Pat. 3,100,762 (Aug. 13, 1963). (b) W. L.
Young, J. S. Kennedy, N. F. Carnahan, and R. R. Blanchard, Dow Chemical Co., U.S. Pat.
3,632,848 (Jan. 4, 1972).
13. G. Bier and H.-E. Konermann, Dynamit Nobel Aktiengesellschaft, U.S. Pat. 3,362,896
(Jan. 9, 1968).

 14. S. Yonezu, Y. Nojima, and T. Motaki, Kanegafuchi Chemical Industries Co. Ltd., U.S. Pat.
3,551,400 (Dec. 29, 1970).
15. L. A. Beer, Monsanto Company, U.S. Pat. 3,328,371 (June 6, 1967).
16. T. Makino, S. Taiana, and S. Nozaki, Nippon Carbide Kogyo Kabushiki Kaisha, U.S. Pat.
3,506,637 (Apr. 14, 1970).
17. T. Wakabayashi and Y. Kobayashi, Nippon Carbide Kogyo Kabushiki Kaisha, U.S. Pat.
3,534,013 (Oct. 13, 1970).
18. G. Steinbach-Van Gaver, Produits Chimiques Pechiney-Saint-Gobain, U.S. Pat.
3,591,571 (July 6, 1971).
19. T. Adachi and T. Oda, Tokuyama Sekisui Kogyo Kabushiki Kaisha, U.S. Pat. 4,049,517
(Sept. 20, 1977).
20. A. J. Olson and R. G. Vielhaber, B. F. Goodrich Co., U.S. Pat. 4,412,898 (Nov. 1, 1983).

Page 652
21. R. J. Davis and T. W. Boyer, B. F. Goodrich Co., U.S. Pat. 4,147,859 (Apr. 3, 1979).
22. J. Gianelos and E. A. Grulke, Adv. X-Ray Anal., 22, 473 (1979).
23. A. J. Downs and C. J. Adams, in Comprehensive Inorganic Chemistry, J. C. Bailar, H. J.
Emeleus, R. Nyholm, and A. F. Trotman-Dickenson, eds., Vol. 2, Pergamon Press, Oxford, p.
1188.
24. (a) J.-P. Weben, M. Assenat, and C. Vrillon, Produits Chimiques Pechiney-Saint-Gobain,
U.S. Pat. 3,532,612 (Oct. 6. 1970). (b) M. DeVita, P. Pece, A. Palvarini, and P. Vaccari, Pozzi
Ferrandina S.P.A., U.S. Pat. 3,597,342 (Aug. 3, 1971).
25. E. J. DeWitt, B. F. Goodrich Co., unpublished results.
26. G. Zeitler and H.-G. Trieschmann, BASF Aktiengesellschaft, U.S. Pat. 3,663,392 (May
16, 1972).
27. M. Balandier, J. Favre, and C. Decker, Chloe Chimie, G. B. Pat. 2,085,897 (Apr. 23,
1982).
28. R. G. Parker, B. F. Goodrich Co., U. S. Pat. 4,350,798 (Sept. 21, 1982).
29. R. G. Parker, B. F. Goodrich Co., U.S. Pat. 4,377,459 (Mar. 22, 1983).
30. A. J. Olson and R. G. Vielhaber, B. F. Goodrich Co., U.S. Pat. 4,373,093 (Feb. 8,
1983).
31. R. G. Parker, B. F. Goodrich Co., unpublished results.
32. H. Ohe, J. Soc. Chem. Ind. Jpn., 48, 89 (1945).
33. O. Seipold, Plast. Kautschuk, 1, 11 (1954).
34. W. Fuchs and D. Louis, Makromol. Chem., 22, 1 (1957).
35. H. Germar, Makromol. Chem., 86, 89 (1965).
36. O. Fredricksen and J. A. Crowo, Makromol. Chem., 100, 231 (1967).
37. H. Kaltwasser and W. Klose, Plast. Kautschuk, 13, 515 (1966); 13, 583 (1966).
38. K. Fukawa, S. Takadono, M. Matsui, and H. Diamon, Kogyo Kogaku Zasshi, 65, 284
(1962).
39. V. R. Allen and R. D. Young, J. Polym. Sci., Part A-1, 8, 3123 (1970).

 40. J. Petersen and B. Ranby, Makromol. Chem., 102, 83 (1967).

41. G. Svegliado and F. Z. Grandi, J. Appl. Polym. Sci., 13, 1113 (1969).
42. S. Sobjima, N. Takagi, and H. Watase, J. Polym. Sci., Part A-2, 6, 223 (1968).
43. C. Decker, M. Balandier, and J. Favre, Abstr., 27th Int. Symp. Macromol., 1, 445
(1981).
44. S. Tsuge, T. Okumoto, and T. Takeuchi, Macromolecules, 2, 277 (1969).
45. T. Okumoto, H. Ito, S. Tsuge, and T. Takeuchi, Makromol. Chem., 151, 285 (1972).
46. R. Lukas, V. Paleckova, J. Svelty, M. Kolinsky, and D. Doskocilova, J. Polym. Sci.,
Polym. Chem. Ed., 17, 2263 (1979), and references therein.
47. R. Lukas, J. Svelty, and M. Kolinsky, J. Polym. Sci., Polym. Chem. Ed., 19, 295 (1981).
48. A. Caraculacu, G. Robila, E. Buruiana, and V. Barbinta, Polym. Bull., 5, 167 (1981).
49. F. Keller and B. Hosselbarth, Faserforsch. Textiltech., 29, 152 (1978), and references
therein.

Page 653
50. G. C. Levy, R. L. Lichter, and G. L. Nelson, Carbon-13 Nuclear Magnetic Resonance
Spectroscopy, 2nd ed., Wiley-Interscience, New York, 1980.
51. R. A. Komoroski, R. G. Parker, and M. H. Lehr, Macromolecules, 15, 844 (1982).
52. (a) C. J. Carman, ACS Symp. Ser., 142, 81 (1980). (b) R. A. Komoroski and J. P.
Shockcor, Macromolecules, 16, 1539 (1983).
53. J. Millan, J. Macromol. Sci. Chem., A12, 315 (1978).
54. (a) For a preliminary report, see R. A. Komoroski, R. G. Parker, J. P. Shockcor, and M.
H. Lehr, Polym. Prepr., Am. Chem. Soc. Div. Polym. Chem., 25, 336 (1984). (b) R. A.
Komoroski, R. G. Parker, and J. P. Shockcor, Macromolecules, 18, 1257 (1985).
55. H. C. Brown and A. B. Ash, J. Am. Chem. Soc., 77, 4019 (1955), and references
therein.
56. P. Berticat, Rev. Gen. Caout. Plast. , 48, 1361 (1971).
57. V. V. Lisitskii, L. S. Berdinskaya, I. K. Pakhomova, L. I. Srkhipova, V. P. Lebedev, A. P.
Savel'ev, E. V. Sonin, N. N. Mel'nikova, and K. S. Minsker, Plast. Massy, No. 3, 36 (1981).
58. J. Millan and J. De la Pena, Rev. Plast. Mod., 34, 525 (1977).
59. C. Decker and M. Balandier, Makromol. Chem., Rapid Commun., 1, 389 (1980);
Makromol. Chem., 183, 1263 (1982).
60. G. S. Huvard, B. F. Goodrich Co., unpublished results.
61. D. R. Hall, B. F. Goodrich Co., U.S. Pat. 4,331,775 (May 25, 1982); U.S. Pat. 4,345,040
(Aug. 17, 1982); U.S. Pat. 4,374,205 (Feb. 15, 1983).
62. (a) C. W. Brabender Instruments, Inc., 50 East Wesley St., South Hackensack, NJ
07606. (b) L. L. Blyler and J. H. Daane, Polym. Eng. Sci., 7, 178 (1967).
63. Technical Service Report 75-03, B. F. Goodrich Co., Chemical Group, 6100 Oak Tree
Blvd., Cleveland, OH 44131.
64. ASTM D1525, 1980 Annual Book of ASTM Standards, Part 35, American Society for
Testing and Materials, Philadelphia, 1980, p. 537.
65. K. H. Illers, Makromol. Chem., 127, 1 (1969).
66. I. M. Hodge and A. R. Berens, Macromolecules, 14, 1598 (1981), and references
therein.

 67. H. Wilski, Kolloid Z. Z. Polym., 238, 426 (1970).

68. C. Bonnebat and A. J. DeVries, Polym. Eng. Sci., 18, 824 (1978).
69. B. Hosselbarth, F. Keller, H. Rehor, and W. Reicherdt, Plast. Kautschuk, 24, 32 (1977).
70. M. H. Lehr, B. F. Goodrich Co., unpublished results.
71. M. H. Lehr, R. A. Komoroski, and R. G. Parker, Macromolecules, 18, 1265 (1985).
72. B. Wunderlich, J. Phys. Chem., 64, 1052 (1964).
73. B. Wunderlich and L. D. Jones, J. Macromol. Sci. Phys., B3, 67 (1969).
74. (a) M. Krishnan, V. B. Gupta, R. T. Thampy, and A. V. Kothari, J. Polym. Sci., C33, 171
(1971). (b) M. Krishnan, P. Singh, and R. T. Thampy, J. Polym. Sci., Polym. Chem. Ed., 12,
255 (1974).
75. Biing-Lin Lee, B. F. Goodrich Co., unpublished results.
76. E. R. Harrell and P. R. Kumler, B. F. Goodrich Co., unpublished results.

Page 654
77. S. Krozer and B. Schneier, J. Appl. Polym. Sci., 15, 1777 (1971).
78. J. Petersen and B. Ranby, Makromol. Chem., 133, 263 (1970).
79. H. Oberst and W. Retting, J. Macromol. Sci. Phys., B5, 559 (1971).
80. (a) P. E. Debye, Polar Molecules, Chemical Catalog Co., New York, 1929. (b) C. P.
Smyth, Dielectric Constant and Molecular Structure, Chemical Catalog Co., New York, 1931.
81. F. Bueche, J. Chem. Phys., 20, 1959 (1952).
82. N. Parkman, in Physics of Plastics, D. Van Nostrand Co., Princeton, N.J., 1965.
83. J. M. Davies, R. F. Miller, and W. F. Busse, J. Am. Chem. Soc., 63, 361 (1941).
84. R. Fuoss, J. Am. Chem. Soc., 61, 2334 (1939); 63, 2401, 2410, 2832 (1941); 64, 283
(1942).
85. R. F. Boyer and R. S. Spencer, in Advances in Colloid Science, Vol. 2, Interscience, New
York, 1946.
86. W. Reddish, J. Polym. Sci., C14, 123 (1966).
87. L. D. Landau and E. M. Lifshitz, Statistical Physics, Pergamon Press, London, 1958, p.
353.
88. I. M. Hodge, B. F. Goodrich Co., unpublished results.
89. P. Berticat, J. Chim. Phys. , 64, 887 (1967).
90. R. P. Lattimer, J. B. Pausch, and H. L. C. Meuzelaar, Macromolecules, 16, 1896 (1983).
91. R. P. Lattimer, B. F. Goodrich Co., unpublished results.
92. (a) M. M. O'Mara, Pure Appl. Chem., 49, 649 (1977). (b) W. H. Starnes, Jr. and D.
Edelson, Macromolecules, 12, 797 (1979). (c) R. P. Lattimer and W. J. Kroenke, J. Appl.
Polym. Sci., 27, 1355 (1982).
93. E. D. Dickens, Jr., J. Vinyl Technol., 5, 150 (1983).
94. E. J. Quinn, D. H. Ahlstrom, and S. A. Liebman, Polym. Prepr., Am. Chem. Soc. Div.
Polym. Chem., 14, 1022 (1973).
95. M. M. O'Mara, J. Polym. Sci., Part A-1, 9, 1387 (1971).
96. M. M. O'Mara, Polym. Prepr., Am. Chem. Soc. Div. Polym. Chem., 14, 1028 (1973).

 97. W. J. Kroenke, J. Appl. Polym. Sci., 26, 1167 (1981).

98. H. Kibbel and A. J. Evans, Soc. Plast. Ind., September 1980, pp. 6, 55.
99. ASTM D2846, 1982 Annual Book of ASTM Standards, Part 35, American Society for
Testing and Materials, Philadelphia, p. 1.
100. Product Brochure, Five Fitting Reasons for Using CPVC, The B. F. Goodrich Chemical
Group, 6100 Oak Tree Blvd., Cleveland, OH 44131.
101. K. M. Rockwell, J. Comm., May 5, 1982, p. 3.
102. For example, BOCA Basic Building Code, 1981, See. 1404.2, 1421.3, Building Officials
and Code Administrators International, Homewood, Ill.
103. Code of Federal Regulations, Chapter 17, Section 25.853.
104. ASTM E162, 1980 Annual Book of ASTM Standards, Part 18, American Society for
Testing and Materials, Philadelphia, p. 902.
105. E. D. Dickens, Jr. and G. F. Smith, Proc. 8th Int. Conf. Fire Saf., San Francisco, January
1983.

Page 655

Author Index
Italic numbers give the page on which the complete reference is cited.

Aase, J. M., 254(157), 298

Aasen, J. O., 147(169), 159
Abadie, M. J. M., 594(303), 600(303), 614
Abbas, K. B., 64(29, 31), 155, 318(27, 33), 342, 353(18), 394, 399(20), 400(20, 22, 32),
402(20), 412(103), 422(20), 430, 432, 595(317), 598(317), 600(317), 615
Abdel-Alim, A. H., 48(7), 50(7), 56(29), 58(7), 59(7), 61(7, 24), 62(7), 64(29), 66(42), 68(45),
154, 155, 311(5), 318(24), 342
Abdullin, M. J., 401(44), 430
Abkin, A. D., 165(15), 167(15), 168(35), 170(15), 171, 172
Abrahams, M. A., 41(25), 43(25), 44
Abrudean, S., 244(45), 293
Abshire, A. D., 8(12), 24
Achhammer, B., 333(106), 345
Adachi, T., 621(19), 651
Adams, C. J., 621(23), 652
Adams, E. M., 258(204), 300
Adams, N. B., 272(279), 304
Adams, W. G. F., 244(60), 259(222), 294, 301
Adams, W. J., 544(140a), 551
Adams, W. R., 242(11), 292
Adams, W. T., 544(84), 549

Agarwal, D. K., 260(225), 301

Agboola, M. O., 610
Agron, P., 168(27), 169(27), 170
[Agron, P., ] (27), 171, 173(23), 232
Ahlstrom, D. H., 65(32a), 155, 336(124), 345, 648(94), 654
Ajroldi, G., 560(33b), 562(33b), 605
Akasaka, N., 591(237), 612
Alarie, Y., 260(243), 302
Albarino, R. V., 530(1), 547
Albert, R. E., 264(266), 303
Albright, L. F., 28(2), 29(1), 30(1, 2), 31(2), 32(1, 2), 34(1, 2), 37(1, 2, 3, 16), 38(1), 39(1, 2),
40(1), 43(2), 44, 89(68, 69, 70), 156
Alexander, D. C., 595(327), 596(327), 615
Alfrey, M. H., 232
Alfrey, T., 166(18), 167(18), 168(18, 27), 169(27, 42), 170(27), 171, 172, 173(23), 177(34),
232, 233, 478(3), 547
Allard, D., 558(16), 566(16), 605
Allard, P., 247(90), 250(90), 259(90), 295
Allbee, N., 532(2), 547
Allen, T., 363(29), 394
Allen, V. R., 623(39), 652
Allsopp, M. W., 111(111), 112(111), 157, 326(80), 327(80, 138), 344, 346, 592(251), 613
Altenau, A. E., 578(105), 608
Al'ter, Yu. M., 534(114), 551

Page 656
Alumont, E., 258(209), 259(209), 301
Ambrose, R. J., 578(105), 608
Amer, A. R., 410(95), 432
Anagnostopoulos, C. E., 447(5), 473(4, 5, 6), 547
Anderson, D., 252(138), 253(141), 297, 425(121), 433
Anderson, E. W., 67(185), 160
Anderson, H., 244(58), 294
Anderson, K., 257(190), 300
Anderson, M. W., 243(35), 293
Andreas, H., 332(100), 344, 417(117), 418(115), 419(117), 420(117), 421(115, 117),
422(115), 423(117), 424(117), 426(117), 427(117), 432
Anea, Z., 244(45), 293
Andreeva, C. A., 594(297, 302), 597(297), 600(297, 302), 614
Andrews, A. W., 255(169), 299
Andruiga, A., 252(140a), 297
Ang, C. H., 591(236), 595(339), 596(339), 600(236), 604(339), 612, 615
Antell, M., 264(260), 288(260), 303
Anthony, P. P., 251(121, 122), 297
Antonyuzhenko, V. A., 245(76), 295
Appleby, A., 264(260), 288(260), 303
Aritoni, M., 169(44), 172
Arknipova, L. I., 209(102), 235
Arlman, E. J., 413(106), 417(106), 432
Arnaud, A., 259(218), 301

Arnold, G. H., 620(3), 651

Arnold, L. F., 575(90), 607
Arnts, 264(260), 288(260), 303
Arond, L., 169(42), 172
Arutynenyan, R. M., 255(167), 299
Asahina, M., 195(75), 234, 425(137), 433
Asamov, M. K., 591(238), 612
Aseev, Y. S., 408(78), 431
Aseeva, R. M., 408(78), 431
Ash, A. B., 629(55), 653
Ashakari, N., 163(8), 171
Ashina, M., 397(1, 2), 429
Assenat, M., 622(24), 652
Atanasov, A. M., 579(122a), 604(122a), 608
Austin, S. G., 251(116a), 296
Averill, S. J., 170(61), 173
Awaya, H., 64(30), 155, 402(52), 431
Axelson, O., 257(190), 300
Ayrey, G., 332(99), 335(99), 344
Bachner, U., 244(57), 294
Baida, N. A., 543(8), 547
Bailey, P. L., 169(50), 172
Bair, H. E., 595(351), 596(351), 604(351), 616
Baker, C. H., 322(46), 324(46), 343, 473(9), 547
Baker, W. P., 169(54), 172

Bakula, S., 258(205), 300

Balakirskaya, V. L., 125(149a), 158
Balandier, M., 408(76), 414(111), 431, 432, 622(27), 623(43), 633(59), 652, 653
Baldwin, F. P., 594(307), 600(307), 614
Baldwin, P. E., 166(20), 167(20), 171
Ball, G. L., 260(242), 302
Bandi, A., 258(209), 259(209), 301
Banyopadhyay, P. K., 561(24), 605
Banzer, J. D., 267(272), 303
Barbinta, V. J., 401(38), 430, 624(48), 652
Barclay, L. M., 326(82), 344
Baretta, E. D., 242(21), 292
Barkalov, I. M., 591(238), 612
Barlow, J. W., 556(10), 561(29, 51), 578(10), 605, 606
Barnes, A. W., 246(76a), 295
Barr, J. T., 54(14), 62(14), 154,

Page 657
[Barr, J. T., ] 246(265), 264(265), 272(278, 282), 275(282), 280(293), 282(282, 294c),
283(294c), 290(282), 303, 304, 305, 306
Barry, T. H., 437(10), 547
Barshtein, R. S., 478(11), 547
Barter, J. A., 86(63), 156
Barth, E. F., 285(300), 306
Bartick, E. G., 353(12), 394
Bartl, H., 591(248), 592(266, 267), 593(273, 284, 285, 286, 287, 288), 600(287, 288), 613
614
Bartlett, P. D., 177(33), 233
Bartnecht, W., 260(237), 302
Bartolucci, G. B., 259(219), 301
Barton, A. F. M., 454(12), 547
Bartsch, H., 242(26a), 293
Barus, C., 183(61), 233
Basalaev, A. V., 244(62), 294
Baskett, A. C., 588(187), 589(187), 611
Basler, A., 252(140), 297
Bauer, J., 66(39), 155, 402(55), 431
Bauer, U., 288(317a), 307
Bauman, R. P., 350(6), 353(6) 394
Bauman, W., 350(5), 353(5), 394
Baxaris, L., 210(102a), 235
Baxter, J. K., 516(91), 550
Baxter, P. J., 251(118, 121, 122), 296, 297

Beach, D. L., 87(67a), 156

Beati, E., 592(258), 613
Bechtelsheimer, H., 252(131), 297
Beckmann, B., 90(76a), 156
Beer, L. A., 621(15), 622(15), 651
Beerbower, A., 448(13, 14), 547
Beggins, I. S., 523(15), 526(15), 547
Behnken, D. W., 103(94), 157
Beitenbach, J. W., 170(67), 173
Bekkedahl, N., 198(166), 237
Belenkii, B. G., 208(100), 235
Bell, R. J., 353(14), 394
Bellenger, V., 418(127), 423(127), 424(127, 132, 133), 433
Bellino, R. A., 595(326), 596(326), 615
Bellus, D., 429(149), 433
Belopolisky, A. O., 113(120), 157
Benbow, J. J., 183(61b), 233
Bengough, W. L., 170(59), 173, 401(36), 430
Benjamin, P., 340(148, 149), 346
Bennett, B., 259(222), 301
Benoit, H. J., 315(17), 342
Bentley, F. F., 231(162), 237
Benton, J. L., 333(108), 345, 409(88), 422(88), 432
Berdinskaya, L. S., 632(57), 635(57), 653
Berens, A. R., 121(132, 133, 134, 135), 122(135), 125(133, 134, 135), 158, 180(51, 52, 53),

233, 283(294e, 294g, 294h, 294k), 288(309), 306, 307, 311(1), 324(71, 72, 73), 330(90),
332(90, 93), 335(117), 336(130, 131), 342, 344, 345, 638(66), 653
Berg, R. M., 497(39), 498(39), 548
Bergen, H. S., 479(16), 485(16), 487(16), 489(16), 547
Bergman, G., 540(195), 541(195), 542(195), 553, 561(45), 606
Bergman, G., 542(17), 547
Berk, S., 530(18), 531(18), 532(18), 547
Berlin, A. A., 404(58), 408(78), 431
Berlot, R., 560(33a), 562(33a), 605
Berndtsen, N., 178(39), 179(39), 233, 341(151), 346
Berry, G. C., 178(40), 233, 250(112), 296
Bert, M., 409(82), 432
Berticat, P., 621(8), 623(8), 632(56), 646(56, 89), 651, 653 654
Bertilsson., H., 542(17), 547
Bertozzi, P. A., 245(68), 294
Bertrand, W. J., 591(235), 600(235), 612

Page 658
Best, L. E., 591(243), 612
Bezdadea, E. C., 320(38), 343 353(17), 394, 398(5, 65), 399(15), 401(5, 45, 47), 406(65, 67),
407(65), 408(65, 67), 411(5, 67), 413(65), 414(65, 67), 415(16), 429, 430, 431
Bhandari, J. C., 246(81), 295
Bias, C. D., 479(40), 548
Bier, G., 399(18), 430, 621(13), 651
Bierbower, G., 246(83), 253(83), 295
Bigham, E., 246(79), 295
Bignami, M., 257(179a), 299
Bigotti, A., 246(77), 295
Biing-Lin Lee, 640(75), 653
Bild, C. G., 89(70), 156
Billmeyer, F. W. Jr., 232
Bingham, E., 252(140a), 297
Birnstial, H., 258(207), 301
Birshtein, T. M., 398(3), 429
Bishop, M., 251(119), 296
Bisio, A. L., 456(184), 473(184), 498(184), 553
Blackely, D. C., 138(159), 143(162), 159
Blair, A., 257(192), 300
Blair, C. M., 527(32), 548
Blanchard, L. P., 208(96), 234
Blanchard, R. R., 621(12), 651
Blattman, H. R., 428(148), 433

Block, J. B., 247(85), 252(129), 295, 297

Blundell, D. J., 324(62), 343
Blyler, L. L., 635(62), 653
Bobalek, E. G., 537(125), 551, 605
Boccato, S., 399(48), 430
Bockman, O. C., 322(48), 343
Boettner, E. A., 260(242), 302
Boggus, J. D., 272(279), 304
Bogue, D. C., 183(61a), 233
Bohak, Z., 230(158), 237
Bohm, G. G. A., 595(325), 615
Bohn, L., 232, 538(19), 542(19), 547
Bohrer, J., 166(18), 167(18), 168(18, 27), 169(27), 170(27), 171, 173(23), 232
Bohse, J., 579(129, 130), 609
Boissel, J. J., 65(33), 155, 320
[Boissel, J. J., ] (40), 326(77), 343, 344, 401(39), 430
Bolt, H. M., 242(33), 243(33, 36, 37, 39), 250(107), 293, 296
Bolt, W., 242(33), 243(33, 39), 293
Bondareva, N. S., 231(265), 237
Bondi, A., 454(20), 547
Bonk, H. W., 561(25), 577(25), 605
Bonnemayre, A., 332(97), 344
Boonebat, C., 638(68), 640(68), 641(68), 653
Boorman, G. A., 246(88a), 295
Boozer, C. E., 77(51), 155

Borisenko, Z. V., 231(165), 237

Bork, J. F., 170(65), 173
Borsini, G. C., 68(188), 160
Bort, D. N., 48(3), 58(21), 110(3), 154, 326(75), 344
Borzelleca, J. F., 256(206, 215), 300, 301
Bosveliev, L. G., 579(122a), 604(122a), 608
Boulton. T. B., 256(170), 299
Bourland, L. G., 561(55), 578(106), 606, 608
Bovey, F. A., 64(29, 31), 65(32), 67(185), 155, 160, 311(6), 318(25, 27, 28, 33), 320(35, 36),
342, 343, 353(18), 394 399(19, 20), 400(19, 20, 22, 29, 30, 31, 32), 401(30), 402(20, 29),
422(20), 430
Bower, D., 311(4), 342, 442(21), 547
Bowles, R. L., 282(2941), 285(297), 306
Bowmer, T. N., 595(355, 356, 357, 358, 596(355, 356), 357, 358), 616
Box, G. P., 103(94, 96), 157
Boyer, R. F., 195(71a), 234, 645(85), 654
Boyer, T. W., 621(21), 652
Boyle, W. J., 276(288), 305
Bozzelli, J. W., 264(261), 288(261), 303
Braconier, F. F., 27(8), 28(7), 30(5), 31(5, 6, 8, 9), 32(6), 34(5), 36(4, 5), 39(5), 44

Page 659
Bradford, E. B., 228(140a, 140b, 140c), 236
Brady, J., 251(119), 296
Bradzivsky, R., 264(262a), 303
Brandrup, J., 398(11), 430, 579(126), 608
Brar, A. S., 208(98), 235
Bras, G., 252(125), 297
Braun, D., 66(40), 155, 333(109), 345 353(16), 394 398(5, 8, 9, 65), 399(14, 16), 400(16, 28),
401(5, 33, 42, 45, 50), 402(51, 54, 56, 57), 403(56), 404(61), 405(57, 64), 406(65, 66, 67),
407(67), 408(65, 67, 75), 411(5, 67), 413(54, 65, 110), 414(65, 67), 415(65), 419(118), 429,
430, 431, 432 433
Brecht, G., 252(134), 297
Brecht, T., 252(134), 297
Breitenbach, J. W., 63(26a), 154, 173(24), 232
Brennan, W. P., 361(22), 394
Brighton, C. A., 579(68), 580(68), 606
Breuer, H., 580(136), 609
Briggs, E. R., 162(4), 167(4), 168(4), 171
Briggs, G., 425(138), 426(139), 433
Brighton, C. A., 579(68), 580(68), 606, 620(4), 621(4)
Brinkman, R. D., 176(24c), 232
Broadway, N. J., 528(115), 530(115), 551
Brokina, E. L., 176(26), 233
Bronckhorst, F. B., 252(125), 297
Brooks, R. L., 474(22), 547
Brown. H. C., 629(55), 653

Brown, W. B., 473(9), 547

Brueckmann, H. B., 592(268), 593(268), 600(268), 613
Brunnerman, C., 263(253), 302
Brydson, J. A., 585(145), 609
Bryk, M. T., 480(185), 484(185), 541(185), 553
Bua, E., 428(142), 433
Bucanelli, A., 260(224), 301
Buchter, A., 242(33), 243(33, 36, 37, 39), 293
Bucknall, C. B., 555(3), 579(3), 580(3), 581(3), 585(3), 604
Bueche, F., 644(81), 654
Bufalini, J. J., 264(258), 302
Buffler, P. A., 254(157), 298
Bufkin, B. G., 223(135), 236
Bugada, D. C., 119(125a), 158
Bunger, W. A., 100(90), 156
Burgess, R. H., 89(71), 156 283(294f), 289(320), 306, 308, 313(14), 339(145), 342, 346,
532(23), 547, 572(77), 575(77), 579(77), 580(77), 593(77), 607
Burille, P., 65(38), 155
Burkhart, R. D., 168(33), 172
Burman, J. P., 103(93), 157
Burnett, W., 251(119), 296
Burrell. H., 100(89), 156, 448(26), 451(24), 454(24, 25, 26 27), 456(24, 25, 26, 27), 547 548
Burviana, E. C., 318(34), 320(39), 343, 353(16), 394, 399(16), 400(16, 23), 401(38), 430,
624(48), 652
Busey, W. M., 246(88a), 295
Bush, S. H., 269(290f), 305

Busse, W. F., 442(28), 548, 645(83), 654

Cain, A. R., 217(129d), 236
Cais, R. E., 318(33), 320(35), 342, 343, 399(20), 400(20, 29, 30), 401(30), 402(20, 29, 30),
422(20), 430
Calandra, J., 246(78), 252(133), 295, 297
Caldwell, J. R., 497(106), 542(104, 106), 550
Calhoun, L. L., 259(214), 301
Cals, R. E., 64(31), 155
Calvert, J. G., 428(144), 433
Cameron, J. B., 90(73), 156 280(291), 282(294b), 305
Camus, A. M., 242(26a), 293
Canelosi, F., 558(18), 605
Cannon, J. A., 479(29, 150),

Page 660
[Cannon, J. A.] 486(166), 511(30), 552 548
Cantow, M. J. R., 180(50), 181(50), 195(50), 197(50), 198(50), 217(129a), 233, 236, 369(36),
371(36), 394
Caputo, A., 246(77), 295
Caraculacu, A. A., 65(36), 155, 318(34), 320(37, 38, 39), 343, 353(16, 17), 394 399(15, 16),
400(16, 23), 401(37, 38, 47), 430, 431, 624(48), 652
Carehza, M., 326(78), 344
Carene, A., 257(179a), 299
Carenza, M., 591(234), 595(334), 596(334), 597(334), 612, 615
Carette, L. B., 418(127), 423(127), 424(127, 132, 133), 433
Carchman, R. A., 259(215), 301
Carleton, L. T., 445(31), 480(31), 548
Carlson, E. J., 170(61), 173
Carlsson, D. J., 428(147), 433
Carman, C. J., 311(7, 8, 9), 342, 626(52), 653
Carmoin, B., 560(33a), 605
Carnahan, N. F., 621(12), 651
Carotti, A. A., 287(306), 307
Carr, C. J., 240(5), 242(5), 292
Carratt, G. M., 57(19a), 154
Carretti, D., 242(26), 246(26), 293
Carrington, W. K., 228(140c), 236
Carruthers, T. F., 527(32), 548
Carter, J. T., 250(109, 111), 296
Casale, A., 588(183), 610

Cassano, F., 256(183d), 299

Castaneda, E., 320(41), 343, 401(43), 430, 588(174, 175, 176), 590(216), 610, 611
Cavagna, F., 318(32), 342 401(49), 431
Cavalli, L., 67(187), 160
Cazias, J. L., 91(77), 120(77), 156
Ceresa, R. J., 555(2), 604
Ceresa, R. J., 445(33, 34, 35), 548, 588(178, 181), 610
Ceysson, M., 587(171), 610
Challa, G., 558(21), 561(35), 565(21), 605
Chalykh, A. E., 577(101), 608
Chan, R. K. S., 76(50), 122(137, 138), 155, 158, 218(131), 236, 282(294c), 283(294c, 294i),
306, 315(16), 342
Chang, E. P., 561(53), 578(53), 606
Chapiro, A., 592(36), 548, 590(217, 218), 591(219, 222, 224, 227, 229, 231, 232, 233),
595(322, 333), 611, 612, 615
Chaplin, E. C., 162(6), 167(6), 168(6), 169(6, 56), 171
Chapman, T. E., 337(133), 345
Charles, J. J., 595(310), 615
Charlesby, A., 413(105), 432
Charpin, J., 259(218), 301
Chartoff, R. P., 322(55), 343
Chastain, C. E., 259(216), 301
Chatelain, J., 134(156), 158
Chaterjee, P., 605
Chauffoureauex, J. C., 184(65), 234
Chekanova, G. I., 534(114), 551

Chekushina, M. A., 209(102), 235

Chen, H. R., 208(96), 234
Chen S. A., 560(33), 561(33), 605
Chen, T., 259(220), 301
Cheney, H. A., 36(9), 44
Cheng, J. T., 114(123), 117(123), 119(126, 127), 135(126, 127), 158, 186(70), 189(90),
191(70), 209(70), 211(70), 213(109, 110), 234, 235, 577(102), 608
Chevassus, F., 398(7), 429
Chiappino, G., 256(180), 299
Chiazze, L. Jr., 242(30, 31), 246(30, 31), 251(30, 31), 293
Chillmoller, C. M., 285(295), 306
Chiu, J., 561(40), 606

Page 661
Chivers, C. P., 259(221), 301
Choshi, Y., 168(32), 169(32), 172
Christie, H., 240(6), 292
Chuang, M., 259(217), 301
Ciampelli, F., 321(44), 343
Cicchetti, O., 428(142), 433
Ciliberti, A., 242(26), 246(26), 293
Ciuciu, C., 594(306), 600(306), 614
Clark, D. D., 561(54), 562(54), 606
Clark, F. W., 442(38), 500(38), 548
Clark, M., 213(108), 235
Clash, R. F., 497(39), 498(39), 548
Clearfield, H., 259(176), 299
Clemmesen, J., 254(159), 298
Clever, G. C., 48(39), 110(39), 154, 219(134), 236
Cline, C. W., 217(129a), 236
Clover, G. C., 326(76), 344
Cloyd, G. D., 516(89), 517(89), 550
Coaker, A. W. M., 479(40), 548
Coate, W. B., 246(88b), 295
Cocain, H. W., 589(204), 600(204), 611
Cohen, C., 215(120), 235
Cole, J., 242(16), 245(16), 292
Coleman, L. E., 170(65), 173

Coleman, M. M., 558(15, 17), 561(26), 605

Collier, P. F., 242(29), 250(29), 293
Collins, B. T., 243(36a), 293
Collins, E. A., 151(175), 159, 180(54, 55), 181(54, 55), 182(54, 55), 184(64), 233 234,
318(23), 324(64), 326(81), 339(114), 342, 343, 344, 346, 561(54), 562(54), 606
Colmenares, R., 604(379), 617
Colwell, R. E., 479(138), 551
Combs, R. L., 180(49), 233
Conolly, R. B., 243(38), 293
Conti, L., 257(179a), 299
Conway, P. J., 594(296), 597
[Coway, P. J.] (296), 600(296), 614
Cook, W. A., 244(63), 260(241), 263(63), 294, 302
Coombes, E. J., 250(109), 296
Cooper, A., 589(190), 611
Cooper, C., 247(92), 295
Cooper, W. D., 112(114), 135(114), 157
Cooray, B. B., 408(73), 424(134), 431, 433
Coover, H. W., 180(49), 233
Copeland, L. H., 259(222.222a), 301
Coran, A. Y., 447(5), 473(4, 5, 6), 547
Corbett, J. C., 353(12), 394
Cordasco, E. M., 259(220), 301
Cordier, J. M., 242(59), 244(59), 245(59), 294
Corradini, P., 321(45), 343

Corraro, G., 67(187), 160

Corry, A., 484(44), 548
Cortellucci, R., 332(91), 344
Cosgrove, S. L., 530(45), 548
Costes, J. R., 37(36), 45
Cotman, J. D., 48(39), 110(39), 154, 219(134), 236, 326(76), 344 399(12), 430
Cotti, G., 242(26), 246(26), 293
Couchman, P. R., 607
Coulter, W. H., 328(85), 344
Coutant, R. W., 289(320a), 308
Cowell, E. E., 516(53a), 519(46), 548
Cowfer, J. A., 27(10), 28(10), 30(10), 31(10), 32(10), 34(10), 44
Cowperthwaite, G. P., 534(47), 548
Cox, R. A., 264(256), 302
Cragg, L. H., 597(365), 616
Crank, J., 121(136), 158
Crawley, S., 402(53), 431
Crebelli, R., 257(179a), 299
Creech, J. L. Jr., 244(53, 61), 247(84), 250(103), 252(103, 130), 294, 295, 296, 297
Crespy, A., 594(303), 600(303), 614

Page 662
Crider, L. B., 283(2941), 285(298), 286(298), 306
Critchley, S. W., 444(48), 548
Crowley, J. D., 454(49), 455(49), 471(49), 548
Crowo, J. A., 623(36), 652
Crugnola, A., 322(49, 50), 343
Csuros, Z., 620(7), 651
Cummings, W., 162(4), 167(4), 168(4), 171
Daane, J. H., 635(62), 653
Dahl, P., 147(171), 148(171), 159
Dakin, V. I., 595(344), 596(344), 616
Dalderup, L. M., 252(125), 297
Daniels, G. A., 121(131), 151(175, 176), 158, 159, 180(55), 181(55), 182(55), 233,
283(294k), 306, 318(23), 324(69), 332(94), 339(114), 342, 343, 344, 346
Dannis, M. L., 445(50, 51), 446(50, 51), 548, 619(1), 621(11), 623(1), 638(1), 640(1), 644(1),
645(1), 648(1), 650(1), 651
Danon, S., 591(226), 612
Danusso, F. J., 322(49, 50), 343
Danziger, H., 240(6), 242(14), 292
Darby, A. E., 532(23), 547
Darby, J. R., 447(87), 455(58), 471(58), 473(58, 87), 475(58), 476(58), 479(16, 58, 97),
483(87, 197), 485(16, 58), 487(16, 87), 489(16), 490(54, 58, 87), 491(54), 492(54), 516(53a),
519 46, 52, 53, 88), 520(53), 521(58, 175, 177), 522(175, 177), 523(145), 524(145, 147),
525(145), 526(55, 57, 145), 527(56), 528(55), 531(57), 533(53), 547, 548, 549, 550, 551, 553
Das, K. C., 515(148), 552
Datner, A., 541(146), 552
Daum, S., 244(58), 294

David, C., 333(103), 335(103), 336(103), 344

Davidson, D. L., 592(311), 595(310, 311, 312), 600(371), 601(371), 615, 616
Davidson, J. A., 112(116), 114(116), 124(147), 135(116), 151(175, 176), 157, 158, 159
326(84), 327(84), 329(87), 330(88), 344
Davidson, R. S., 415(113), 427(113), 432
Davies, J. M., 645(83), 654
Davies, W., 251(119), 296
Davis, D. D., 595(352, 355), 596(352, 355), 616
Davis, R. J., 283(294e), 306, 621(21), 652
Deacon, M. M., 259(214), 301
Dealy, J. M., 374(38), 394
Deanin, A. A., 600(369), 616
Deanin, R. D., 572(75), 575(87), 579(75), 581(75, 138, 139), 600(369), 607, 609, 616
deBoer, A., 558(21), 565(21), 605
deBroutelles, R., 398(7), 429
Debye, P. E., 644(80), 654
Decaufle, P., 257(192), 300
Decker, C., 408(76), 413(107), 414(111), 422(107), 431, 432, 591(239), 600(239), 612,
622(27), 623(43), 633(59), 652, 653
DeCoste, J. B., 376(41), 395 513(61), 526(62), 528(62), 538(60), 549
Deese, D. E., 255(164), 298
Deguchi, S., 587(167), 610
Dehennau, C., 184(65), 234
deJong, W. A., 324(59), 343
DelaPena, J., 632(58), 633(58), 653
DeLeeuw, F. J. G., 273(285), 305

Delmas, G., 476(63), 549

Delorme, F., 247(89), 295
DeMarco, R. D., 575(90), 607
Demeter, S. L., 259(220), 301
Denchev, I. D., 579(122a), 604(122a), 608
Denges, K., 593(284), 614
Denki, Kagaku, Kogyo, K. K.,

Page 663
[DenkiKagaku Kogyo, K. K.] 579(124a), 608
DePierre, W. G., 170(64), 171(64), 173
Deri, M., 229(148), 236
DeRuiter, N., 250(106), 296
Desai, R., 260(230, 231), 301, 302
deSante, Gaebe, L., 259(218), 301
deVerdier, A., 257(190), 300
Devita, M., 622(246), 652
DeVries, A. J., 638(68), 640(68), 641(68), 653
deWilde, M. C., 168(38), 172
Dewitt, E. J., 622(25), 652
Diamon, H., 623(38), 652
Dickens, E. D., Jr., 647(93), 650(105), 654
Dickie, R. A., 567(71), 607
Dietz, A., 250(103a), 296
Dikow, H. 580(133), 585(133, 158), 609, 610
Dilling, W. L., 264(256), 266
(269), 285(269), 302, 303
Dimethriades, B., 264(259a), 303
Dimonie, M., 594(306), 600(306), 614
Dinman, B. D., 244(63, 64, 65), 245(63, 64), 294
Ditcheck, T., 244(63), 263(63), 294
Dixon, R. L., 246(81), 295
Djomo, H., 604(379), 617

Doak, K. W., 169(46), 172

Dobo, J., 596(350), 616
Dodson, V. N., 244(65), 294
Dolezel, B., 519(192), 528(64), 549, 553
Doolittle, A. K., 438(68), 442(67, 68), 474(65), 476(67), 518(66), 549
Doskocilva, D., 624(46), 652
Doty, P., 68(46), 155
Doube, C. P., 607
Downs, A. J., 621(23), 652
Downs, T. D., 254(155), 298
Doyle, C. D., 595(320), 615
Drejman, J., 244(48, 51), 293 294
Dressman, R. C., 267(271), 303
Drew, R. T., 246(88a), 295
Druesedow, D., 322(52), 324(52), 343, 408(69), 411(97), 431, 432
Dtsuka, M., 165(16), 167(16), 169(16), 171
Du, J. T., 250(108), 296
Dubb, S. M., 538(182), 540(182), 541(182), 553
Dubini, M., 428(142), 433
DuBois, J. H., 437(69), 549
Duck, B. W., 250(110, 111), 296
Duggan, F. W., 527(70), 549
Duntoff, F., 168(34), 172
Duxbury, H. A., 276(290), 305
Dyaschkovskii, F. S., 589(198), 611

Dykstra, T. H., 170(63), 173

Eberly, K. C., 36(11), 44
Ebert, H., 530(18), 531(18), 532(18), 547
Eckardt, F., 250(106), 296
Economy, J. P., 169(48), 172
Edelhauser, H., 170(67), 172,
Edelson, D., 333(110), 345
Edelson, D., 410(93), 432, 647(92), 654
Eder, E., 257(187), 299
Edmonds, L. D., 253(148, 149), 254(148, 149), 298
Edwards, F. G., 48(4), 154
Edwards, W. S., 573(83), 607
Eggelton, A. F. J., 264(257), 302
Egorova, Z. S., 595(344), 596(344), 616
Eigenmann, H., 333(111), 345
Eirich, F. R., 535(170), 537(170), 552
Elia, G. E., 256(183d), 299
Elias, H. G., 232
Eliassaf, J., 110(107, 108, 109), 157
Elinder, C. G., 252(124), 297
Ellis, C., 437(71), 448(187), 549, 553
Elmquist, C., 561(27), 605
Elofsson, S., 251(116), 296
Embree, W. H., 170(60), 173
Emett, R. A., 579(112), 608

Emgee, D. E., 479(29), 548

Endicott, K. M., 258(203, 208), 259(208), 300, 301
Ernest, J., 412(100), 432

Page 664
Ernest, R. R., 530(72), 549
Erussalimsky, B., 168(34), 172
Espinosa, E., 247(87), 250(87), 295
Etlis, V. S., 169(43), 172
Evans, A. J., 648(98), 654
Evans, J. J., 253(141), 297
Ewart, R. H., 146(167), 159
Fabris, H. J., 589(204), 600(204), 611
Fagnoni, Y., 16(13), 24
Falk, H., 244(53, 56), 247(86), 249(94), 250(95, 96, 98), 251(86), 253(149, 152), 254(149),
259(95, 98), 294, 295, 296, 298
Fallwell, W. F., 27(12), 43(12), 44
Fassi, H., 587(170), 610
Fath. J., 520(74), 549
Favre, J., 622(27), 623(43), 652
Feathers, R. E., 33(38), 45
Fedoseev. B. I., 209(102), 211(105), 235
Feldman, D., 561(28), 605, 610
Ference, L., 242(31), 246(31), 251(31), 293
Feron, V. J., 244(41), 246(41, 80), 252(41), 293, 295
Fertig, J., 419(119, 120), 433
Fiechtner, J., 251(120), 297
Fields, J. W., 326(83), 344
Fieser, L. F., 83(58), 156

Fieser, M., 83(58), 156

Fievez, C., 242(59), 244(59), 245(59), 294
Figge. K., 408(74), 431
Filatova, V. S., 244(43), 245(76), 293, 295
Filisko, F. E., 322(53), 324(53), 343
Filser, J. G., 242(33), 243(33, 36, 37), 293
Finaz, G., 594(291), 597(291), 600(291), 614
Findeib. W., 408(74), 431
Fineman, M., 176(25), 219(25), 233
Fischer, J. P., 107(101), 157
Fischer, M., 137(157b), 159
Fischer, N., 132(153), 158
Fischer, S. G., 561(40), 606
Fischer, W., 520(75), 549
Fishbein, A. S., 244(58), 294
Fishbein, L., 245(73), 294
Fisher, A. M., 240(6), 292
Fisher, N. J., 326(77), 344
Fitzgibbon, M., 518(76), 549
Fitzpatrick, R. D., 273(278a), 304
Flak, H., 252(130), 297
Flake, R. F., 252(139), 297
Flelcher, D., 579(126), 608
Fleig, J., 252(136), 297
Flogstad, H., 58(20), 154

Florence, D. M., 595(332), 615

Florjanezyk, A. L., 236
Flory, P. J., 48(5), 51(10), 55(15), 154, 232, 449(77), 549
Flury, F., 240(9), 242(20), 292
Folkemer, F. B., 479(137), 518(137), 551
Folt, V. L., 170(61), 173, 180(51, 52, 53), 233, 311(1), 322(58), 336(130, 131), 342, 343, 345
Foltz, C. R., 336(124), 345
Fordyee, R. G., 162(6), 167(6), 168(6), 169(6), 171
Forster, P., 572(73), 577(73), 578(73), 579(73), 580(73), 585(73), 607
Fortwengler, H. P., 247(87), 250(87), 295
Foster, H. U., 256(183e), 299
Foster, R. T., 621(9), 651
Fowkes, F. M., 558(14), 605
Fox, A. J., 242(29), 250(29, 113), 251(118), 293, 296
Fox, T. G., 178(40), 233, 339(146), 346, 573(82), 607
Franey, J. P., 427(140), 433
Frangia, N., 260(224), 301
Franke, G., 257(194a), 300
Frankowski, R. F., 254(155), 298
Frazer, D. G., 574(89), 575(89), 579(113), 607, 608
Fredricksen, O., 623(36), 652
Freed, D. J., 335(119, 120), 345

Page 665
[Freed, D. J.] 400(30), 401(30), 402(30), 430
Freilich, D., 230(159), 237
Freni, S. C., 252(125), 297
Frentzel-Beyme, R., 242(32), 251(32), 293
Fressell, J. B., 269(290e), 305
Frey, H. E., 501(78), 549
Frey, W., 581(141), 589(141), 593(279), 609, 614
Friday, A., 232
Fried, J. R., 607
Fried, J. R., 286(301d), 306
Friederich, H. H., 589(210), 611
Friend, L., 33(13), 44
Friis, N., 52(16), 57(16, 17, 18,, 19), 59(18), 61(23), 62(23), 72(49), 76(50), 148(172),
149(50), 154, 155, 159
Frisch, H. L., 121(130), 158
Frissell, W. J., 483(79), 487(79), 549
Frolova, L. Z., 165(17), 167(17), 168(17), 169(17), 171
Fryar, G. J., 27(22), 28(22), 30(22), 31(22), 32(22), 34(22), 37(22), 42(22), 43(22), 44
286(303), 307
Frye, A. H., 401(35), 423(122, 123, 124), 430, 433
Fuchs, W., 623(34), 652
Fujiji, Y., 593(281), 614
Fukawa, K., 623(38), 652
Fukazawa, H., 595(363), 596(363), 616
Fuoss, R. M., 493(136), 551 645(84), 654

Furgason, R. R., 184(64), 234

Furukawa, J., 587(163), 610
Gaal, T., 589(200), 611
Gabbard, J. D., 486(167), 552
Gabbett, J. F., 232
Gabilly, J., 591(219a), 612
Gabor, S., 244(44, 45), 293
Gaffey, W. R., 295
Gallinella, E., 68(188), 160
Gallot, Y., 594(290, 291), 597 (291), 600(290, 291), 614
Galmarini, O. L., 532(80), 549
Gamble, J., 242(15), 245(15), 292
Gamrath, H. R., 447(5), 473(5, 6), 547
Gankina, E. S., 208(100), 235
Gantmakher, A. R., 165(13), 167(13), 171
Garbuglio, G., 68(188), 160
Gardlund, Z. G., 587(164, 166), 588(177), 610
Gardon, J. L., 138(160), 147(170), 151(180), 159, 454(81), 456(81), 473(81) 474(81), 549
Garnett, J. L., 591(236), 595(339), 596(339), 600(236), 604(339), 612, 615
Garton, A., 605
Gasparini, P., 48(6), 49(6) 58(6), 154
Gasparrini, G., 591(234), 612
Gataullin, R. F., 401(44), 430
Gatta, G., 560(33b), 562(33b), 605
Gauld, S., 259(222a), 301

Gay, B. W. Jr., 264(258, 259a), 302, 303

Gay, P. C., 252(139), 297
Gaylord, N. G., 595(314, 315), 598(315), 599(315), 600(314, 315), 615
Geddes, W. C., 195(74), 234 398(10), 413(10), 430
Gedigk, P., 252(131), 297
Gee, G., 473(9), 547
Gehring, P. J., 242(24, 25), 243(24, 34), 292, 293
Gell, P. H., 110(110), 157, 324(63, 64), 326(63, 79, 81), 343, 344
Geissler, K. P., 592(268), 593(268), 600(268), 613
Gelic, J., 252(123), 297
Gemignani, C., 259(219), 301
Gerecke, J., 589(201), 592(268), 593(268), 598(201), 600(201, 268), 611, 613
German, A. L., 176(29), 233
Germer, H., 311(2), 312(2), 342, 623(35), 652
Gerrard, D. L., 334(114), 345 523(15), 526(15), 547
Gerspacher, 113(117), 157

Page 666
Gertenbach, P. G., 573(80), 607
Gerum, J., 593(280), 614
Gezovich, D. M., 324(63), 326(63, 81), 343, 344
Ghaffar, A., 611
Ghersa, P., 493(82), 494(82), 496(82), 540(82), 541(82), 549
Giacin, J. R., 283(294n), 306
Gianelos, J., 621(22), 652
Gibb, T. B., 213(114), 235
Gibbs, C. F., 408(69), 411(97), 431, 432
Giever, P. M., 244(64), 245(64), 294
Gilbert, H., 170(61), 173
Gilbert, J., 195(77), 234
Gilbert, M., 324(67, 68), 343
Gilbert, S. G., 283(294n), 306
Girard, T. A., 531(83, 174), 549, 552
Gisolf, J. H., 324(59), 343
Gisser, H., 522(139), 551
Glass, J. E., 326(82), 344
Gledhill, W. E., 544(84), 549
Gobel, W., 592(267), 593(283), 613, 614
Godart, J. A., 36(4), 44
Goergens, H. W., 242(33), 243(33), 293
Goh, S. H., 561(51), 606
Goiran, L., 132(153), 158

Goldberg, A. I., 419(119, 120), 433

Goldfinger, G., 232
Goldschmidt, B. M., 257(188), 300
Goldstein, J. H., 311(7, 8), 342
Gole, J., 594(298), 600(298), 614
Golish, J. A., 259(220), 301
Golub, M. A., 413(109), 432
Gomez-Vaillard, R., 92(82), 102(82), 156
Gomi, S., 30(14), 39(14), 44
Goni, K. G., 263(253), 302
Gonzalez, M. F., 48(39), 110
(39), 154, 219(134), 236 326(76), 344
Gonze, A., 337(140), 346
Goode, S. W., 246(78), 295
Gorbachevakaya, I. I., 589
[Gorbachevakaya, I. I.] (197), 611
Gordon, D. E., 246(78), 252(133), 295, 297, 544(107), 550
Gordon, J. E., 555(1), 604
Gordy, W., 454(85), 550
Gorshkov, V., 595(353), 596(353), 616
Gorton, B. S., 572(76), 607
Gotham, K. V., 340(147), 346
Gould, R. W., 578(107), 608
Goulet, L., 253(150), 298
Gouloubandi, R., 591(233), 612

Graedel, E., 427(140), 433

Graessley, W. W., 180(47), 233
Graff, M. J., 37(16), 44
Graham, J. D., 263(249), 302
Graham, P. R., 447(87), 473(87), 483(87), 487(87), 490(54, 87), 491(54), 492(54), 519(88),
526(55), 528(55), 544(86, 107), 548, 550
Grainger, R. G., 244(66), 245, 294
Grandi, F. Z., 623(41), 629(41), 652
Grandin, R. E., 122(138), 158 283(294i), 306
Grassie, N., 585(156), 598(156), 609
Grave, J. R., 223(135), 236
Greek, B. F., 27(12), 43(12), 44
Green, J. H. S., 50(11), 52(11), 154
Greenberg, D. B., 286(301d), 306
Greenberg, J., 516(89), 517(89), 550
Greenberg, L. A., 242(11), 292
Greenberg, R. A., 250(99, 104), 252(104), 296
Greenberg, R., 250(97), 259(97), 296
Greenburger, M., 532(112), 550
Greenley, R. Z., 178(37, 38), 233
Greenwald, P., 251(119), 296
Gregor, H. P., 166(20), 167(20), 171
Gregorian, R. C., 594(296), 597(296), 600(296), 614
Gregorzik, H., 288(317a), 307
Greiser, E., 251(114, 115), 296

Page 667
Griffin, J. D., 600(370), 616
Griffiths, P. R., 353(15), 394
Grigoriu, G., 605
Grigororescu, I., 244(46), 293
Grimm, A., 409(87), 432
Gringas, S., 253(151), 298
Grinard, M., 264(254), 302
Gronberg, E. S., 244(43), 293
Groszmann, M., 620(7), 651
Groth, D. H., 246(88b), 295
Grotz, L. C., 446(90), 550
Grouman, D., 257(192), 300
Grub, L. M., 516(91), 550
Grubisia, S., 315(17), 342
Grulke, E. A., 621(22), 630(22), 652
Guengerich, F. P., 242(23), 256(183b), 292, 299
Gugelmeto, P. D., 560(33b), 562(33b), 605
Gul, V. E., 457(92), 550
Gullino, P. M., 260(236), 302
Gumpert, R., 264(260), 288(260), 303
Gundel, F., 589(210), 611
Gupta, A., 588(182), 589(205, 207), 597(207), 598(207), 599(207), 610, 611
Gupta, R., 283(294i), 306
Gupta, V. B., 640(74), 653

Gupta, V. P., 195(76), 234, 408(71), 409(81), 413(71), 431

Gupta, R., 122(137, 138), 158
Guyot, A., 65(37), 66(38), 155 320(41, 42), 324(60), 343 401(43), 409(82), 424(128, 129,
130, 131), 425(135), 430, 432, 433, 587(171), 588(174, 175, 176), 590(216) 610, 611
Gysling, H., 428(146), 433
Haaf, F., 580(136), 609
Haanes, E., 332(92), 344
Haas, H., 166(18), 167(18), 168(18, 27), 169(27), 170(27), 171, 173(23), 232
Haddon, R. C., 335(119, 120, 121, 122), 345, 410(91, 92), 432
Haefner, A. J., 283(294o), 306
Haeussler, L., 589(201), 592(268), 593(268), 598(201), 600(201), 268), 611, 613
Hake, C. L., 256(182, 183e), 299
Halasz, L., 186(67), 234
Hall, D. R., 636(61), 653
Hall, R. H., 577(103), 608
Halsam, J., 350(2, 3, 4), 352(4), 393
Halse, W. F., 169(52), 172
Ham, G. E., 162((6), 167(6), 168(6), 169(6, 48, 56), 171, 172, 232
Hamada, H., 68(44), 155, 311(3) 312(3), 342
Hamanone, K., 217(129d), 236
Hamielec, A. E., 48(7), 50(7), 52(16), 57(16, 17, 18, 19), 58(7), 59(7, 18), 61(7, 23), 62(7,
23), 66(42), 68(45), 76(50), 92(82), 102(82), 148(172), 154, 155, 156, 159, 318(24),
341(152), 342, 346
Hamilton, J. B., 350(2, 3), 393
Hammad, H., 399(17), 430
Hammer, C. F., 572(76, 78), 575(78, 91), 577(78), 607

Hammer, E., 113(118, 119), 157

Hammon, H. G., 600(370), 616
Hammond, G. S., 77(51), 155
Han, C. D., 178(43), 233, 562(60), 606
Hansen, C. M., 448(93), 454(93), 457(93), 471(93), 475(93), 550
Hansen, D., 264(260), 288(260), 303
Hansen, J. L., 287(305), 307
Hanson, C. M., 100(88), 156
Hansteen, I. L., 252(137), 297
Harding, J., 513(162), 552
Hardisty, J. F., 246(83), 253(83), 295
Hardt, D., 591(248), 592(266, 267), 593(273, 284, 285, 286, 287, 288), 600(287, 288), 613,
614
Hardy, G., 591(221), 612
Harkins, W. D., 143(164, 165, 166), 159
Harley, G. S., 143(161), 159
Harmer, D. A., 591(223), 600(223, 370), 612, 616

Page 668
Harper, P., 251(119), 296
Harrell, E. R., 322(55), 343, 641(76), 653
Harrington, E. C., 544(94), 546(94), 550
Harris, D. K., 244(60), 294
Harris, L. R., 276(290c), 305
Harris, R. L., 280(293b), 305
Hart, R., 169(45), 170(66), 172, 173
Hartless, R. L., 318(25), 342, 399(19), 400(19, 30, 32), 401(30), 402(30), 430
Hartmann, M., 589(189a), 595(189a), 611
Hartwig, 273(278c), 304
Haseman, J. K., 246(88a), 295
Hashimoto, T., 561(39), 606
Hass, J. F., 254(158), 298
Hatashi, S., 311(3), 312(3), 342
Hatch, M., 254(160), 298
Hayashi, H., 593(281), 614
Hayashi, K., 170(58), 173
Hayashi, T., 595(331), 596(331), 604(331), 615
Hayaski, S., 68(44), 155
Hayden, 57(16), 154
Hayoshi, S., 399(17), 430
Head, B., 332(99), 335(99), 344
Healey, T. E. J., 257(195), 260(195), 300
Heath, C. W., 244(53), 252(130), 294, 297

Heatley, F., 311(6), 342

Hecker, A. C., 520(127), 551
Hedges, W. L., 256(183c), 257(183c), 299
Hefner, H., 581(141), 589(141), 609
Hefner, R. E. Jr., 242(24), 243(24), 292
Hehir, R. M., 246(83), 253(83), 295
Heiberger, C. A., 180(50), 181(50), 195(50), 197(50), 198(50), 217(129a, 129e, 233, 236,
368(35), 369(36), 371(36), 394, 591(24, 39), 592(243a), 612
Hell, J. F., 449(95, 96), 473(95), 474(95, 96), 550
Heinemann, H., 30(18, 19), 31(17), 32(19), 33(17), 41(18), 44
Heinzmann, O., 579(126), 608
Heldas, S. S., 252(137), 297
Heller, H. J., 170(61), 173, 428(146, 148), 433
Hellman, M. Y., 595(356, 357), 596(356, 357), 616
Hellman, T. M., 254(162a), 298
Hellwege, K. H., 311(2), 312(2), 342
Hemsley, D. A., 324(68), 343
Hendriksen, C. F. M., 246(80), 295
Hendry, D. G., 333(111), 345
Henrici-Olive, G., 231(164), 237
Henschler, D., 257(187, 188a), 299, 300
Heppel, L. A., 258(203, 208), 259(208), 301
Hergberg, M., 258(209), 259(209), 301
Herner, M., 572(73), 577(73), 578(73), 579(73), 580(735, 133), 585(73, 133, 158), 607, 609,
610
Hertzberg, R. W., 582(142, 143, 144, 145, 146), 583(142), 603(145, 146), 609

Hertzberg, T., 58(20), 154

Hext, P. M., 260(230), 301
Hickman, J. J., 575(93), 607
Hicks, O., 544(84, 140a), 549, 551
Hilado, C. J., 260(239, 240), 302
Hilderebrand, J. H., 436(97), 449(97), 450(97), 451(97), 457(97), 475(97), 550
Hill, A., 444(48), 548
Hill, J., 266(268), 272(268), 303
Hill, R., 230(156), 237
Hillestad, I., 252(137), 297
Himeno, S., 253(141a), 297
Hinds, J. R., 232
Hirata, S., 611
Hironaka, H., 594(294), 600(294), 614
Hische, D. C., 335(119, 120), 345
Hitch, M. J., 340(147), 346

Page 669
Hjertberg, T., 102(92a), 157 318(26, 29, 30), 320(29, 43), 335(29), 342, 343, 400(24, 27),
401(27, 46, 50), 402(24, 46), 411(46), 430, 431
Hlousek, M., 69(47), 155
Hoang, T. V., 424(128), 425(135), 433
Hodge, I. M., 324(72, 73), 344, 638(66), 646(88), 653, 654
Hoel, D. G., 243(35), 293
Hoffman, D., 152(183), 159, 263(253), 302
Hofonau, H. T., 258(207), 301
Hogstedt, C., 257(190), 300
Hogue, R., 264(260), 288(260), 303
Holbub, I., 255(164a), 298
Holder, B. B., 242(28), 293
Holmberg, B., 251(116), 257(190), 296, 300
Hohmlund, L., 251(116), 296
Holstein, P., 258(209), 259(209), 301
Holzer, G., 400(28), 401(45, 50), 406(67), 407(67), 408(67), 411(67), 414(67), 430, 431
Hong, C. B., 246(82), 295
Hood, F. R., 67(185), 160
Hooper, T. R., 257(195), 260(195), 300
Hopfenberg, H. B., 121(129, 132), 158
Hopff, H., 113(117, 118, 119), 157
Hopkins, J., 252(135), 297
Hori, M., 242(19), 292
Hornibrook, W., J., 585(159), 610

Hornung, R. W., 246(88b), 295

Horsley, R. A., 491(98), 539(98), 550
Horst, R. W., 401(35), 423(122, 123, 124), 430, 433
Horvath, J. W., 579(114), 608
Hoski, Y., 591(242), 612
Hosoda, K., 600(374), 616
Hosoi, F., 595(340), 615
Hosselbarth, B., 624(49), 626(49), 639(69), 652, 653
Hourston, D. J., 561(46, 47, 57), 577(96, 99), 606, 607
House, W. B., 246(81), 295
Howard, J. B., 333(111), 345
[Howard, J. B.] 526(62), 528(62), 549
Howes, F. N., 437(99), 550
Hoy, K. L., 444(100), 446(100), 447(100), 550
Hrabal, Z., 513(101), 550
Huddleston, G. R. Jr., 283(294e), 306
Huff, J. E., 276(289), 305
Hughes, I. D., 561(46, 47, 57), 557(96, 99), 606, 607
Hughmark, G. A., 283(294o), 306
Huibers, D. T., 217(129a), 236
Hulse, W. F., 177(32), 178(32), 232, 233, 289(321), 308
Hummel, D. O., 352(9), 394
Hundertnark, G., 579(128), 608
Hunter, J. S., 103(96), 157
Hunter, W. G., 103(96), 157

Hurley, J. F., 259(222, 222a), 301

Hutchinson, R. J., 472(143), 475(143), 486(143), 551
Huvard, G. S., 635(60), 653
Hyatt, I. S., 438(102), 550
Hyatt, J. W., 438(102), 550
Hyde, J. T., 572(76), 607
Ichihashi, T., 66(40a), 155, 404(60), 431
Ide, F., 578(108), 608
Iglauer, N., 231(162), 237
Ikeda, M., 251(116b), 296
Ikeda, R. M., 575(93), 607
Ilea, E., 245(69), 294
Illers, K. H., 322(54), 324(54, 70), 338(54, 70), 341(70), 343, 344, 638(65), 653
Illidge, J. T., 269(287a), 275(287a), 305
Imai, K., 591(230), 612
Immergut, E, H., 398(11), 430 486(103), 493(103), 550
Imoto, M., 162(2, 3), 165(14, 16), 167(2, 3, 14, 16), 168(14, 30, 39, 40), 169(16, 41), 170(2)
171, 172, 409(89), 432, 593(283), 614
Inagaki, M., 579(124), 608
Infante, P. F., 250(110), 253

Page 670
[Infante, P. F.] (146, 152), 254(146), 296 298
Inoue, T., 561(39), 606
Irish, D. D., 258(204), 300
Iskander, M., 558(20), 561(20), 575(20), 605
Isner, J. D., 324(64), 326(81), 337(132, 133), 343, 344, 345
Isozumi, M., 579(121), 608
Istrate, R. G., 320(38), 353(16), 343, 394, 399(16), 400(16), 430
Ito, H., 623(45), 626(45), 646(45), 652
Ito, J., 595(331), 596(331), 604(331), 615
Ito, T., 165(14), 167(14), 168(14), 171
Itoh, K., 544(111), 550
Iturra, H., 252(126), 253(151), 297, 298
Ivan, B., 336(123), 345, 411(98), 432, 595(313), 597(313), 600(313), 615
Ivanof, L., 244(45), 293
Iwama, M., 157
Iwamoto, K., 595(321), 615
Izumi, F., 595(362), 596(362), 616
Jackson, W. J., 497(106), 542(104, 106), 550
Jaeger, R. J., 243(38), 293
Jalbert, R. L., 556(4), 578(4), 585(4), 604
James, W. R. L., 256(175), 299
Jamieson, A., 585(156), 598(156), 609
Janca, J., 208(97, 98, 99, 100), 235, 315(18), 342
Jasching, W., 427(141), 433

Jelenic, J., 561(43), 606

Jendrychowska-Bonamour, A. M., 591(224, 227), 595(322), 612, 615
Jene, M. O., 210(102a), 235
Jenkins, M. C., 228(140a), 236
Jobard, M., 591(219a, 226), 612
Joesten, B. L., 231(163), 237
Johannsen, F. R., 544(107), 550
John, J. A., 259(214), 301
John, J. A., 253(144), 259(214), 297, 301
Johnsen, U., 311(2), 312(2), 342
Johnson, C. D., 177(31), 233
Johnson, C. W., 72(48b), 155
Johnson, G. R., 113(121, 122), 114(122), 134(121, 122), 157
Johnson, H. P., 260(244), 302
Johnson, N. W., 229(153), 237
Johnson, M. N., 244(53, 67), 247(84), 249(93), 294, 295
Johnson, O. S., 518(108), 534(108), 550
Johnson, R. F., 530(123), 551
Johnson, T. W., 169(49), 172
Johnson, W. S., 260(233), 302
Johnston, N. W., 231(163), 237
Johst, P., 258(207), 301
Jones, B., 280(293b), 305
Jones, D. B., 244(58a), 247(87), 250(87), 294, 295
Jones, G. P., 169(51), 172

Jones, H., 448(109), 550

Jones, J. H., 251(117), 264(117), 296
Jones, L. D., 639(73), 653
Jones, R. B., 257(189b), 300
Jordan, E. F., Jr., 573(84, 85, 86), 607
Jorgensen, A. H., 579(113), 608
Joshi, R. M., 176(28), 233
Jowett, D., 257(194), 300
Joyce, S. F., 486(110), 534(110), 550
Joynes, R. E., 255(164), 298
Jubanowsky, L. J., 520(147), 552
Juhe, S., 244(55, 57), 294
Juijn, J. A., 324(59), 343
Jungwirt, A., 217(126, 127), 236
Kabanov, V. Y., 595(323), 615
Kachalai, D. P., 543(8), 547
Kadyrov, T., 594(297), 597(297), 600(297), 614
Kafman, L. P., 169(43), 172
Kagiya, T., 595(340), 615

Page 671
Kaiser, E. R., 287(306), 307
Kalal, J., 208(100), 235
Kaley, R. G., 544(84), 549
Kalinin, A. I., 326(75), 344
Kalinin, A. Y. A., 48(3), 110(3), 154
Kaltwasser, H., 623(37), 652
Kamada, K., 157
Kamath, V. R., 78(55), 80(55), 82(55), 155
Kambour, R. P., 565(67), 566(67), 568(67), 579(67), 606
Kaminska, A., 585(155), 609
Kane, R. P., 577(100), 608
Kaneko, A., 591(242), 612
Kaneshima, H., 544(111), 550
Kaplan, A. M., 532(112), 550
Kaplan, N. L., 243(35), 293
Kappus, H., 243(39), 293, 250(106), 293, 296
Karfoglou, N. K., 561(44), 606
Kargin, V. A., 48(3), 110(3), 154, 326(75), 344
Karpow, V. L., 595(344), 596(344), 616
Kasabo, T., 594(294), 600(294), 614
Kasatochkin, V. J., 408(78), 431
Kato, M., 212(107), 235
Katz, R. M., 257(194), 300
Kaubara, S., 169(44), 172

Kaufman, M., 2(1), 3(2), 24, 455(113), 457(113), 550

Kaye, L. A., 448(13), 547
Kawai, W., 66(40a), 155, 404(60), 431
Kazome, G. A., 168(32), 169(32), 172
Kebbekus, B. B., 264(261), 288(261), 303
Keck, D. R., 286(304a), 307
Kedem, O., 230(158), 237
Kefeli, A. A., 401(44), 430
Kelen, T., 195(78), 234, 408(70, 72, 77), 409(90), 411(98), 412(102, 104), 413(104), 431,
432, 333(104, 112), 336(112, 123), 344, 345, 595(313), 597(313), 600(313), 615
Kellar, D. E., 86(63), 156
Keller, F., 624(49), 626(49),
[Keller, F.] 639(69), 652, 653
Kemp, H. S., 276(290a), 305
Kennedy, J., 336(123), 345
Kennedy, J. F., 578(104), 608
Kennedy, J. P., 592(311), 594(307, 308), 595(308, 309, 310, 311, 312, 313, 316), 597(364,
366), 598(316, 364), 599(366), 600(307, 308, 309, 310, 311, 312, 313, 316, 364, 371),
601(371), 614, 615, 616
Kennedy, J. S., 621(12), 651
Kent, G., 252(133), 297
Keplinger, M. L., 246(78), 295
Kerber, R., 593(280), 614
Kerkay, J., 259(220), 301
Kerner, E. H., 567(69), 606
Kessler, I. I., 257(193), 300

Ketcham, N. H., 288(313), 307

Kever, E. E., 208(100), 235
Key, M. M., 244(53), 294
Khamis, J. T., 588(186), 589(186), 611
Khanna, S. K., 186(68a, 68b, 69), 234, 561(48), 606
Khodzhemijov, V. A., 165(13), 167(13), 171
Khodzhevanov, F. F., 534(114), 551
Khomikovskii, P. H., 165(15), 167(15), 170(15), 171
Kibbel, H., 648(98), 654
Kilian, D. J., 252(139), 297
Kilmartin, J., 305
Kimble, H. J., 288(313), 307
Kimura, S., 593(283), 614
Kimura, T., 167(23), 168(25, 29), 169(53), 170(57), 171, 172
King, R. W., 528(115), 530(115), 551
Kinloch, A. J., 580(122), 582(122), 608
Kirkpatrick, A., 448(116), 551
Kirste, R. G., 561(43), 606
Kirsten, R., 561(53), 578(53), 606
Kirywkhin, D. P., 591(238), 612
Kishimoto, A., 611
Kiss, V., 589(200), 611
Kissin, Y. V., 87(67a), 156

Page 672
Kitamura, R., 167(24), 173(24), 171
Klein, J., 242(17), 245(17), 292
Klein, S. A., 257(184), 299
Kleine, W., 593(279), 614
Kleiner, L. W., 561(54), 562(54), 565(66), 567(66), 568(66), 575(66), 575(66), 606, 607
Kleinfeld, M., 256(171), 299
Klemm, H. F., 510(163), 513(163), 552
Klenin, S. I., 594(297), 597(297), 600(297), 614
Klewan, N., 166(22), 167(22), 171
Kline, J., 254(160), 298
Kline, M. W., 556(5), 593(5), 601(5), 604
Kline, S. A., 256(183a), 299
Klondike, D. W., 282(323), 308
Klopffer, W., 428(143), 433
Klose, W., 623(37), 652
Knight, G. J., 195(72), 234
Knoepke, O. H., 520(127), 551
Kobayashi, M., 579(132), 592(132), 603(132), 609
Kobayashi, T., 561(39), 606
Kobayashi, Y., 242(19), 292, 621(17), 651
Kobsa, H., 527(117), 551
Kock, K., 215(121), 235
Koda, C. F., 531(83, 174), 549, 552
Kodulka, P. S., 285(301), 306

Koenig, J. L., 231(160), 237 322(52), 324(52), 343

Kofman, L. P., 165(17), 166(19), 167(17, 19), 168(17), 169(17, 19), 171
Koischwitz, D., 252(134), 297
Kolawole, E. G., 610
Koleske, J. V., 560(30), 561(30), 575(94), 577(30, 95), 605, 607
Kolesnikova, W. W., 413(108), 432
Kolesov, S. V., 404(58), 431
Kolinsky, A., 591(249), 592(249), 593(249), 601(249), 613
Kolinsky, M., 119(128a), 158, 208(97, 98, 99), 217(126, 127),
[Kolinsky, M.] 235, 236, 315(18), 342, 404(59), 431, 624(46, 47), 626(47), 652
Kollig, H., 266(268), 272(268), 303
Kolnikow, O. W., 413(108), 432
Komkova, G. S., 229(152a, 152b), 237
Komoroski, R. A., 625(51), 626(51, 52), 628(54), 629(51), 630(54), 631(51), 639(71), 653
Konermann, H. E., 621(13), 651
Konig, I., 318(32), 342, 401(49), 431
Konig, J., 65(35), 155
Konkhin, Yu. A., 113(120), 157
Kono, S., 251(116b), 296
Koozu, H., 595(341), 596(341), 615
Kornegay, R., 67(185), 160
Korneva, R. S., 404(58), 431
Korotov, A. A., 594(297, 302), 597(297), 600(297, 302), 614
Kothari, A. V., 640(74), 653
Kotik, K., 513(101), 550

Kotlyar, I. B., 589(197), 611

Kotlyarevskii, G. A., 478(11), 547
Kozlov, P. V., 493(159), 552
Kozlowski, M., 589(199), 600(199), 611
Kozmarek, H., 585(155), 609
Kozozis, J., 264(260), 288(260), 303
Kramer, H., 399(18), 430
Kramer, M., 217(125), 218(125), 219(125), 221(125), 235
Kranmer, C. G., 244(50), 294
Krantz, J. C., 240(5), 242(5), 292
Kraus, G., 599(368), 616
Krause, F. E., 280(292), 305, 283(294d), 306
Krause, S., 559(23), 573(23), 577(23), 599(23), 605
Krees, R., 244(41), 246(41), 252(41), 293
Krehler, H., 34(20), 44
Krimm, S., 311(1), 322(58), 342, 343

Page 673
Krishnan, M., 640(74), 653
Kristalyni, B. V., 168(35), 172
Kritskaya, D. A., 589(198), 611
Kroenke, W. J., 647(92), 648(97), 650(97), 654
Kronman, A. G., 209(102), 211(105), 235
Kroupova, J., 208(100), 235
Krozer, S., 643(77), 653
Krzewki, R. T., 332(96), 344
Kucera, M., 594(300), 600(300), 614
Kuchanov, S. J., 58(21), 154, 313(13), 342
Kudryatsev, V., 595(323), 615
Kudryavyseva, D. G., 244(47), 293
Kuebler, H., 240(8), 292
Kuenzel, A., 589(201), 598(201), 600(201), 611
Kuhlwilm, V., 592(259), 613
Kulcar, Z., 252(123), 297
Kulenev, V. N., 457(92), 550
Kull, S., 406(66), 408(75), 413(110), 431, 432
Kumar, N. G., 178(41), 233
Kumins, C. A., 535(118, 119, 121), 536(118, 119), 551
Kumler, P. R., 641(76), 653
Kunkel, P., 581(141), 589(141), 609
Kunz, M., 594(300), 600(300), 614
Kuran, W., 236

Kurata, M., 157

Kuratsune, M., 251(116b), 296
Kuroda, K., 579(124), 608
Kurtz, B. E., 37(33), 45
Kurtz, S. S., Jr., 537(120), 551
Kuryla, W. C., 288(313), 307
Kushner, L., 595(360), 596(360), 616
Kuska, V., 591(249), 592(249), 593(249), 601(249), 613
Kumack, A. M., 252(127), 264(127), 297
Kwei, K. P., 415(112), 432
Kwei, T. K., 535(121), 551, 577(98), 595(351, 355, 356), 596(351, 355, 356), 604(351), 607,
616, 617
Kydonieus, A. F., 517(215), 518
[Kydonieus, A. F.(215), 554
Kyrides, L. P., 438(122), 551
Labarthe, D. R., 254(155), 298
Lable, B. G., 169(49), 172
Lackner, K., 252(134), 297
Lafleur, J. C., 289(320b), 308
Laib, R. J., 242(23a, 33), 243(33), 257(23a), 292, 293
Lalet, P., 587(170), 610
Lamb, P., 183(61b), 233
Lamberty, M., 337(141), 341(141), 346
Lamberty, P., 592(271), 600(271), 613
Lamola, A. A., 428(145), 433

Lamson, J. J., 577(103), 608

Landau, L. D., 645(87), 654
Landfield, H., 530(123), 551
Landi, V. R., 575(88), 607
Landler, Y., 590(213, 214), 591(245), 611, 613
Lane, R. W., 259(206), 300
Lang, R. F., 267(274e), 304
Langbein, G., 250(103a), 296
Lange, C. E., 244(49, 55, 57), 294
Langner, R. R., 242(28), 293
Langsam, M., 62(25), 76(50), 87(67b), 102(92b), 108(25, 104), 119(127), 135(127), 152(184),
154, 155, 156, 157, 160, 180(48), 186(70), 189(70), 191(70), 209(70, 101), 211(70),
213(110), 217(48, 128), 218(48, 133), 219(48), 221(48), 222(48), 223(48), 233, 234, 235, 236
Langswald, N., 236
Lanikova, J., 69(47), 155
Lantos, I., 589(200), 611
Laskawski, W., 561(50), 577(50), 606
Lattimer, R. P., 647(90, 91, 92), 654
Laundry, R. W., 272(278), 304
Laurent, G., 594(290), 600(290), 614
Lautout, M., 591(226), 612
Laverty, J. J., 587(164, 166),

Page 674
[Laverty, J. J.] 587(164, 166), 588(177), 610
Lawrence, A. R., 474(22), 547
Lawrence-Jones, C., 259(221), 301
Lazar, M., 166(22), 167(22), 171
Leake, S. D., 240(2), 242(2), 292
Lebedev, V. P., 324(66), 343, 632(57), 635(57), 653
Lebel, P., 590(213, 214), 591(245), 611, 613
LeBrun, A. J., 256(183e), 299
Lecca-Radu, M., 244(44), 293
Lechermeier, G., 593(289), 594(289, 298, 299), 600(289, 298, 299), 614
Lederer, M., 272(280), 304
Lederer, S. J., 534(124), 551
Lee, C. C., 246(81), 253(142), 295, 297, 537(125), 551
Lee, K. C., 287(305), 307
Lee, W. A., 195(72), 234
Lees, F. P., 269(290g), 271(290g), 305
Lefemine, G., 242(26), 246(26), 257(186), 258(186), 293, 299
Lefevre, M. J., 242(59), 244(59), 245(59), 294
Lehka, N., 519(217), 554
Lehman, K. B., 242(20), 292
Lehman, L. B., 240(9), 292
Lehr, M. H., 625(51), 626(51), 628(54), 629(51), 630(54), 631(51), 639(70, 71), 641(70),
643(70), 645(70), 653
Leilich, V. K., 500(126), 551

Leistner, W. F., 520(127), 551

Leisz, D. M., 573(80), 607
Lelbach, W. K., 244(17a), 251(17a), 292
Lelchuk, Sh. L., 518(128), 551
Leong, B. K., 253(143), 297
Lepaev, A. F., 209(102), 235
Leps, G., 579(129, 130), 609
LeRoy, E., 244(67), 294
Leslie, M. L., 600(370), 616
Lester, D., 242(11), 292
Leugering, H. J., 580(135), 609
Levin, G., 594(301), 600(301), 614
Levinskas, G. L., 544(107), 550
Levinson, G. S., 177(36), 233
Levot, R., 591(236), 595(339), 596(339), 600(236), 604(339), 612, 615
Levy, G. C., 624(50), 652
Lewis, C., 166(18), 167(18), 168(18), 171
Lewis, F. M., 162(4), 164(9), 167(4, 9), 168(4, 9, 26), 169(52), 170(9), 171, 172, 177(32),
178(32), 233, 289(321), 308
Lewis, M. H., 113(122), 114(122), 134(122), 157
Liberatore, F., 251(119), 296
Lichter, R. L., 624(50), 652
Liebman, S. A., 65(32a), 155, 336(124), 345, 648(94), 654
Liegeois, J. F., 582(145, 146), 603(145, 146), 609
Lifshitz, E. M., 645(87), 654

Lijinsky, W., 255(166a, 169), 267(166a), 298

Lillian, D., 264(260), 288(260), 303
Lillis, R., 244(58), 256(177), 294, 299
Lin, R. S., 257(193), 300
Linden, G. F., 273(285), 305
Linssen, H. N., 176(29), 233
Lisickij, V. V., 404(58), 431
Lisitskii, V. V., 632(57), 635(57), 653
Litja, A., 593(274), 600(274), 613
Lloyd, M., 259(222a), 301
Lo, T. S. K., 322(55), 343
Loan, L. D., 595(328, 346), 596(328, 346), 615, 616
Lobban, L., 264(260), 288(260), 303
Loeffler, D., 589(201), 598(201), 600(201), 611
Loewengart, G., 257(188), 300
Logothetis, A. L., 236
Long, M. A., 591(236), 595(339), 596(339), 600(236), 604(339), 612, 615

Page 675
Longeway, G. D., 332(94), 344
Lonnstrom, A., 593(274), 600(274), 613
Loshaek, S., 339(146), 346
Lotto, A., 256(180), 299
Louis, D., 623(34), 652
Lowe, J. W., 454(49), 455(49), 471(49), 548
Lu, P. L., 266(270), 303
Lucas, E. V., 259(220), 301
Lucas, R., 404(59), 431
Lukas, R., 119(128a), 158, 591(249), 592(249), 593(249), 601(249), 613, 624(46, 47), 652
Luken, R. H., 263(250), 302
Lundberg, R. D., 575(94), 595(316), 598(316), 600(316), 607, 615
Lunden, A. J., 90(73), 156, 280(291), 282(294b), 305
Lundstrom, H. S., 579(114), 608
Lussi, H., 113(117, 118, 119), 157
Lutz, J. T., 578(111), 608
Lynch, R. P., 289(320b), 308
Lyngaae-Jorgensen, J., 148(173), 159, 180(45), 181(59, 60), 233
Macho, V., 591(249), 592(249), 593(249), 601(249), 613
Mack, W. A., 283(294m), 306
Mackrodt, W. C., 257(189b), 300
MacMahon, B., 254(154), 298
MacMillan, R. B., 588(187), 589(187)
Maddams, W. F., 311(4), 312(12), 322(46, 47), 324(46), 334(114), 342, 343, 345

Madruga, E. L., 311(10), 312(10), 342

Magarik, S., Ya., 594(297), 597(297), 600(297), 614
Magat, M., 591(219, 226, 231), 612
Mageli, O. I., 273(285), 305
Magistro, A. J., 27(10), 28(10), 30(10), 31(10), 32(10), 34
[Magistro, A. J.] (10), 44
Magnuson, H. J., 244(63, 64, 65), 245(64), 263(63), 294
Mahnken, G., 511(30), 548
Makino, T., 621(16), 651
Maklakov, A. A., 442(129), 551
Maklakov, A. I., 442(129), 551
Malaveille, C., 242(26a), 293
Malik, P., 560(32), 561(32), 605
Mallow, J. S., 249(94), 295
Maltoni, C., 242(26), 246(26), 255(165), 257(186), 258(186, 213), 293, 298, 299, 301
Mammi, M., 324(65), 343
Manas, J., 591(249), 592(249), 593(249), 601(249), 613
Mandel, E., 258(209), 259(209), 301
Mankowski, Z., 591(232), 612
Mann, R., 595(354), 596(354), 616
Manno, M., 259(219), 301
Manson, J. A., 556(7), 557(7), 562(7), 567(7), 572(74), 579(7), 580(7), 582(7, 142, 143, 144,
145, 146), 583(142), 586(7), 595(348), 596(348), 597(365), 602(7), 604, 607, 609, 616
Manta, I., 244(44), 293
Mantell, G. J., 282(294c), 283(294c), 306, 577(102), 608

Marcer, G., 259(219), 301

Marcus, W., 252(128), 297
Marechal, E., 620(6), 651
Marian, S., 594(301), 600(301), 614
Maricq, H. R., 244(67), 294
Marinin, V. G., 48(3), 110(3), 154, 326(75), 344
Mark, H. F., 143(163), 159, 486(103), 493(103), 550
Markovitz, R. E., 89(72), 156
Marks, G. C., 134(157), 158, 333(108), 345, 409(88), 422(88), 432
Marling, P. E., 520(130), 551
Maron, S. H., 322(53), 343
Marsh, G. M., 248(92a), 295
Marshal, A. L., 500(200), 553
Marstellar, H. J., 244(17a, 52), 251(17a), 252(134), 292,

Page 676
[Marstellar, H. J.] 294, 297
Marten, F. L., 92(82), 102(82), 148(173), 156, 159
Martinez, G., 195(79), 234
Marvel, C. S., 162(1), 167(1), 168(1), 169(49, 51), 170(63, 64), 171(64), 171, 172, 173
Mascia, L., 542(131), 551
Mashini, C. I., 285(300), 306
Masiulanis, B., 561(49), 606
Mastin, T., 169(51), 172
Mastrangelo, G., 259(219), 301
Mastromatteo, E., 240(6), 292
Mathis, T. C., 455(180), 477(180), 553
Matijik, M., 260(243), 302
Matsui, M., 623(38), 652
Matsumoto, A., 591(222), 612
Matsuo, M., 562(61), 563(61) 564(61), 570(72), 571(61), 579(61), 580(137), 595(351),
596(351), 604(351), 606, 607, 609, 616
Matsuoka, M., 162(3), 165(16), 167(3, 16), 168(30, 39), 169(16, 41), 171, 172
Matthews, G. A. R., 315(22), 342
Mattiello, J. J., 437(132), 551
Matzner, M., 583(149), 609
Matzner, V. M., 588(179), 610
May, F. L., 538(182), 540(182), 541(182), 553
Mayer, R. A., 528(115), 530(115), 551
Mayer, Z., 62(25), 108(25), 154, 332(98), 335(98), 336(98), 344

Mayhan, K. G., 591(235), 600(235), 612

Mayo, F. R., 108(102), 157, 164(9), 167(9), 168(9, 26), 169(52), 170(9), 177(32), 178(32),
171, 172, 232, 233, 289(321), 308
Mazzur, R. P., 484(133), 485(133), 520(133), 551
Mbamali, G. E., 610
McCallister, D. D., 258(204), 300
McCaughy, R. E., 252(127),
[McCaughy, R. E.] 264(127), 297
McClelland, N., 267(274d), 304
McClure, G. L., 353(12), 394
McCollister, D. D., 258(199), 300
McConnell, E. E., 246(88a), 295
McCormick, W. E., 244(61), 294
McCoy, E. C., 257(185), 299
McCulley, J. H., Jr., 90(73), 156, 280(291), 282(294b), 305
McDonald, S. A., 151(176), 159
McFarlane, R. C., 176(30), 233
McFarren, G. F., 267(271), 303
McGill, W. J., 579(131), 609
McGrath, J. E., 556(9), 558(20), 561(20, 56), 575(20, 92), 576(92), 586(9, 162), 588(162,
179), 593(9), 605, 606, 607, 610
McKalen, C. A., 330(90), 332(90), 344
McKenna, J. M., 236
McKenzie, D. A., 425(121), 433
McKinney, P. V., 444(134, 135), 445(135), 447(135), 479(135), 480(134, 135), 485(134), 551
McLaughlin, J., Jr., 246(83), 253(83), 295

McMichael, A., 242(15), 245(15), 249(94), 253(152), 292, 295, 298

McNamara, B. P., 246(83), 253(83), 295
McNaughton, K. J., 30(21), 44
McNeill, I. C., 585(156), 598(156), 609
McNeill, J. C., 402(53), 431
McPherson, R. W., 27(22), 28(22), 30(22), 31(22), 32(22), 34(22), 37(22), 42(22), 43(22), 44,
286(303), 307
McQuald, J., 273(278a), 304
McSween, R. N. W., 251(121, 122), 297
Mead, D. J., 493(136), 551
Mears, C. L., 42(23), 44
Mednedev, S. S., 165(13, 15), 167(13, 15), 170(15), 171
Melby, E. G., 589(204), 600(204), 611

Page 677
Melchionne, S., 256(183a), 257(188), 299, 300
Mel'nikova, N. N., 632(57), 635(57), 653
Melville, H., 52(16), 154
Menges, G., 178(39), 179(39), 233, 337(137), 341(151), 346
Menn, J. J., 479(137), 518(137), 551
Menschick, H., 258(210), 301
Menzel, G., 579(127, 128), 585(157), 608, 610
Merriam, C. N., 583(149), 609
Mertwoy, H. E., 522(139), 551
Mertz, E., 232
Merz, E. H., 479(138), 551
Merz, W., 401(40), 430
Metzger, A. P., 180(54), 181(54), 182(54), 184(64), 233, 234
Metzler, H., 257(187), 299
Meuzelaar, H. L. C., 647(90), 654
Meyer, G., 519(140), 551
Meyer, G. C., 604(379), 617
Michael, P. R., 544(84, 140a), 549, 551
Michaels, A. S., 215(119a), 235, 326(74), 344
Michalcova, J., 404(59), 431
Michel, A., 65(37), 66(38), 155, 320(41, 42), 324(60), 343, 401(43), 409(82), 424(128, 129,
130, 131), 425(135), 430, 432, 433, 587(171), 588(173, 174, 175, 176), 590(216), 610, 611
Michel, J. M., 582(145, 146), 588(189), 589(189), 603(145, 146), 609, 611
Michel, M., 591(226), 612

Michl, K. H., 579(128a), 581(141), 589(141), 609

Mickley, H. S., 215(119a), 235, 326(74), 344
Mijangos, C., 195(79), 234
Miletto, A., 587(170), 610
Milincuk, V. K., 413(108), 432
Mill, T., 333(111), 345
Millan, J., 195(79), 234, 311(10), 312(10), 342, 628(53), 632(58), 633(58), 653
Miller, A., 259(217), 301, 479
[Miller, A.] (137), 518(137), 551, 595(335, 336), 596(335, 336), 615
Miller, F. F., 170(61), 173
Miller, H., 30(18, 19), 31(17), 32(19), 33(17), 41(18), 44
Miller, M. L., 105(99), 157, 394
Miller, R. E., 256(183b), 299
Miller, R. F., 645(83), 654
Miller, S. G., 263(250), 302
Mills, C. L., 169(56), 172
Min, K. W., 138(160a, 160b), 159
Minor, J. L., 253(142), 297
Minot, J. D., 285(297), 306
Minoura, Y., 594(293, 294), 595(321), 600(293, 294), 614, 615
Minsker, K. S., 272(281), 304, 401(44), 404(58), 430, 431, 632(57), 635(57), 653
Mishuck, E., 445(31), 480(31), 548
Mitani, K., 64(30), 155, 398(4), 402(52), 429, 431, 595(331), 596(331), 604(331), 615
Mitchell, M. J., 170(60), 173
Mitsengendler, S. P., 594(297, 302), 597(297), 600(297, 302), 614

Mitsui, H., 595(340), 615

Mitzutani, Y., 595(331), 596(331), 604(331), 615
Miyasaka, K., 561(39), 606
Mogi, K., 595(349), 616
Mogne, F. C., 170(63), 173
Molina, G., 251(116), 257(190), 296, 300
Molnar, I., 186(67), 234
Mondvai, I., 186(67), 234
Monson, R. R., 249(93), 254(156), 295, 298
Moore, A. C., 326(74), 344
Moore, A. L., 215(119a), 235
Moore, U. A., 246(88a), 295
Moore, W. F., 595(330), 596(330), 615
Moore, W. R., 472(143), 475(143), 486(143), 551
Moosing, R., 251(116), 296
Morgan, N., 287(305), 307
Mori, S., 176(24a), 208(24a), 232

Page 678
Mork, P. C., 147(169, 171), 148(171), 159
Morrall, J. F., 263(251), 302
Mortimer, G. A., 176(27), 233
Moshar, C., 581(138), 609
Moskala, E. J., 558(15), 560(15), 561(15, 26), 605
Moslen, M. T., 243(38), 293
Motaki, T., 621(14), 651
Mozurowski, S. J., 273(285), 305
Moszner, N., 589(189a), 595(189a), 611
Mrazek, Z., 217(126, 127), 236
Mrkvickova, L., 208(98), 235
Mucke, H., 409(87), 432
Mukerji, A., 43(24), 44
Mukhanov, A. A., 113(120), 157
Mukherjee, A. K., 588(182), 589(205, 207), 597(207), 598(207), 599(207), 610, 611
Muliwan, H., 250(106), 296
Muller, H. A., 410(96), 432
Muller, K., 401(40), 430
Muller, R., 252(131), 297
Murai, K., 168(32), 169(32), 172
Murphy, S. D., 243(38), 293
Murray, J. S., 259(214), 301
Mutchler, J. C., 244(50), 294
Mutchler, J. E., 242(21), 292

Myers, S. A., 288(311), 307

Nachtomi, E., 258(209), 259(209), 301
Nagai, S., 168(28, 31), 171, 172
Nagy, T. T., 195(78), 234, 333(104, 112), 336(112, 123), 344, 345, 408(70, 72, 77), 411(98),
412(102, 104), 413(104), 431, 432
Nakajima, A., 68(44), 155, 399(17), 430
Nakajima, H., 311(3), 312(3), 342
Nakamura, M., 401(34), 430
Nakao, M., 597(364), 598(364), 600(364), 616
Nakaya, T., 409(89), 432
Nakase, Y., 591(237), 612
Nardi, V., 324(65), 343
Naso, A. N. M., 244(65), 294
Nass, L. I., 417(114), 419(114), 421(114), 425(114), 426(114), 427(114), 432
Natta, G., 321(45), 343
Neal, P. A., 258(203, 208), 259(208), 300, 301
Neff, J. E., 288(313), 307
Nelson, A. R., 588(188), 589(188), 611
Nelson, G. L., 624(50), 652
Neudecker, T., 257(187), 299
Newman, J., 232
Newman, S., 556(6), 557(6), 559(6), 565(6), 579(123), 580(123), 604, 608
Nicol, K. J., 591(236), 600(236), 612
Nichols, J. D., 594(295), 597(295), 600(295), 614
Nichols, W. E., 242(30, 31), 246(30, 31), 251(30, 31), 293

Nicholson, W. J., 244(58), 294

Nielsen, L. E., 178(42), 216(122), 233, 235, 565(64), 570(65), 606
Nietsche, R., 232
Nikonov, A. A., 260(227), 301
Nishi, T., 577(98), 607, 617
Nishimura, A., 163(8), 171, 595(363), 596(363), 616
Nishimura, F., 237
Nislick, A. S., 41(25), 43(25), 44
Nissim, J. E., 253(149), 254(149), 298
Nojima, Y., 621(14), 651
Nolan, K. P., 410(94), 432
Noller, D. C., 273(285), 305
Nolley, E., 561(29), 605
Noonan, R. C., 264(258), 302
Normandin, R. O., 575(87), 607
Norrish, R. G. W., 170(59), 173
Northrup, H. R., 263(252), 268(252), 302
Noshay, A., 556(9), 583(149), 586(9), 588(179), 593(9), 605, 609, 610
Novikava, S. P., 534(114), 551

Page 679
Nozaki, K., 177(33), 233
Nozaki, S., 621(16), 651
Nuyken, O., 593(280), 614
Obalisi, O., 232
Oberst, H., 643(79), 653
Oberthur, R. C., 561(43), 606
Oda, T., 621(19), 651
O'Driscoll, K. F., 176(30), 233
Ogata, T., 64(30), 155, 398(4), 402(52), 429, 431
Ogoshi, Z., 579(132), 592(132), 603(132), 609
Ohe, H., 623(32), 652
Oka, K., 579(121), 608
Oklandnov, N. A., 324(66), 343
Okuda, H., 253(141a), 297
Okumoto, T., 623(44, 45), 626(44, 45), 646(44, 45), 652
Olabisi, O., 556(11), 557(11), 558(11), 559(11, 22), 561(11, 22), 562(22), 565(11, 22),
566(11), 567(22), 568(11), 570(11), 572(11), 573(11, 22), 575(11), 577(11), 578(11), 579(11,
22), 583(11), 585(11), 605
Olaj, O. F., 58(22), 63(26a), 154
Olayemi, Y. P., 610
Olive, S., 231(164), 237
Oliver, F., 595(325), 615
Olsen, R., 288(309a), 307
Olson, A. J., 621(20), 622(30), 651, 652
O'Mara, M. M., 283(2941), 285(297), 306, 336(125), 345, 647(92), 648(95, 96), 654

Onozuka, M., 195(75), 234, 397(1, 2), 401(36), 425(136, 137), 426(136), 429, 430, 433
Oodan, K., 595(341), 596(341), 615
Orem, J. H., 523(145), 525(145), 526(145), 551
O'Riordan. M. L., 253(141a), 297
Oster, R. H., 240(5), 242(5), 292
Osugi, J., 217(129c), 236
Ota, Y., 242(19), 292
Otocka, E. P., 530(1), 547
O'Toole, J. T., 147(168), 159
Otsu, T., 165(14), 167(14), 168(14), 171
Otsuke, M., 168(40), 172
Ott, M. G., 242(28), 293
Ottenwalder, H., 250(107), 296
Ottery, J., 259(222), 301
Ottewill, R. H., 228(139), 236
Ovehinnikov, Y. V., 616
Overberg, N., 229(149a), 236, 587(169), 610
Overberger, C. G., 166(20), 167(20), 169(42), 171, 172
Owen, E. D., 332(101), 335(101), 336(101), 344, 407(68), 409(86), 431, 432
Oyen, F., 242(10), 244(42), 292, 293
Pace, R. J., 541(146), 551
Paddle, G. M., 254(153), 259(221), 298, 301
Paduraru, A., 245(69), 294
Painter, P. C., 558(15), 560(15), 561(15, 26), 605
Paisley, H. M., 50(11), 52(11), 154

Pakhomova, I. K., 632(57), 635(57), 653

Palacin, F., 587(172), 610
Paleckova, V., 624(46), 652
Palinchak, S., 528(115), 530(115), 551
Paliobagis, M. A., 423(123, 124), 433
Palma, G., 64(27), 154, 326(78), 344, 400(25), 430, 591(229, 234), 595(334), 596(334),
597(334), 612, 615
Palmgren, D., 152(182), 159
Palmgren, O., 151(178), 159
Palvarini, A., 622(24b), 652
Paradis, R. A., 183(61d), 233
Parey, J., 337(137), 346
Park, C. N., 243(34), 293
Park, G. S., 64(28), 155, 400(26), 430
Park, I. K., 184(63), 234

Page 680
Parker, A., 413(109), 432
Parker, R. G., 622(28, 29), 623(31), 625(51), 626(51), 628(54), 629(51), 630(54), 631(51),
639(71), 652, 653
Parkman, N., 644(82), 654
Parrini, P., 428(142), 433
Parrod, J., 594(290, 291), 597(291), 600(290, 291), 614
Parsons, F. Z., 267(274e), 304
Pasha, I., 407(68), 431
Patel, C. B., 122(137, 138), 158
Patel, I. T., 575(87), 607
Patel, M. A., 283(294i), 288(313), 306, 307
Paton, C., 444(48), 548
Patrianokos, D., 263(253), 302
Patterson, D. D., 476(63), 549
Patterson, J. W., 285(301), 306
Pattison, D. A., 448(13, 14), 547
Patton, T. C., 520(147), 552
Patty, F. A., 240(1), 292
Patwardham, S. A., 515(148), 552
Paul, D. R., 232, 479(149), 552, 556(6, 10, 12), 557(6, 12), 558(12), 559(6), 561(29, 51),
564(12), 565(6), 566(12), 572(12), 578(10), 604, 605, 606
Pauling, L., 49(8), 50(8), 77(8), 154, 409(83), 432
Pausch, J. B., 330(90), 332(90), 344, 647(90), 654
Pavan, A., 561(52), 606
Payan, H., 259(218), 301

Payan, I. L., 267(274e), 304

Payne, M. T., 479(29, 150), 548, 552
Pearson, D. S., 595(325), 615
Pearson, R. B., 315(22), 342
Pece, P., 622(24b), 652
Pegoraro, M., 322(49), 343, 592(258), 613
Peigjaner-Corbella, L., 210(102a), 235
Pellen, M., 437(151), 552
Pendle, T. D., 601(376), 617
Penn, W. S., 479(152), 497(152), 552
Penskaya, E. A., 457(92), 550
Peoples, S. A., 240(2), 242(2),
[Peoples, S. A.] 292
Permanetter, W., 250(103a), 296
Perric, T. L., 258(208), 259(208), 301
Perry, C. R., 263(248a), 302
Pershagen, G., 252(124), 297
Pessen, H., 532(190), 553
Peter, H., 243(37), 293
Peters, J. M., 249(93), 295
Peterson, J. E., 256(182), 299, 623(40), 626(40), 643(78), 652, 653
Petiaud, R., 318(31), 342, 401(48), 431
Petieud, E., 65(34), 155
Petrich, R. P., 580(134), 581(134), 609
Petrova, G. A., 168(37), 172

Petrukhov, A. M., 595(353), 596(353), 616

Pezzin, G., 180(44), 233, 412(101), 432
Pham, Q. T., 311(10), 312(10), 318(31), 342, 401(48), 431
Phillips, E. A., 271(276d), 304
Phillips, E. M., 283(294i), 306
Phillips, S. R., 180(50), 181(50), 195(50), 197(50), 198(50), 217(129a, 129e), 233, 236,
369(36), 371(36), 394
Picciano, D. J., 252(139), 297
Piglowski, J., 561(49, 50), 577(50), 589(199), 600(199), 606, 611
Pike, R. M., 169(50), 172
Pillot, C., 593(289), 594(289, 298, 299), 600(289, 298, 299), 614
Pinner, S. H., 413(105), 432, 499(154), 514(153), 552, 591(220), 595(337, 338), 596,(337,
338), 612, 615
Piszlewicz, L., 333(111), 345
Pitenson, J. E., 256(183e), 299
Pitts, J. N., Jr., 428(144), 433
Plackett, R. L., 103(93), 157
Planovsky, A. N., 113(120), 157
Platzer, N., 169(47), 170(47), 172
Player, J. M., 578(107), 608
Plesch, P. H., 594(305), 600(305),

Page 681
[Plesch, P. H.] 614
Plewan, O., 580(135), 609
Plitz, I. M., 66(40c), 155, 335(115, 116, 119, 120), 345, 400(29, 30, 32), 401(30), 402(29, 30),
405(63), 423(125, 126), 430, 431, 433
Pokorny, S., 208(99, 100), 235
Polan, A., 251(119), 296
Politzer, P., 256(183c), 257(183c), 299
Poller, R., 332(99), 335(99), 344
Pollini, C., 595(334), 596(334), 597(334), 615
Polte, A., 579(128), 608
Poltz, H., 169(55), 170(55), 172
Pommer, E. H., 532(194), 553
Pommier, P., 259(218), 301
Pomogallo, A. D., 589(198), 611
Ponomarev, A. N., 589(198), 611
Poole, T. R., 257(195), 260(195), 300
Popa, M., 578(110), 608
Popper, H., 244(54), 247(86), 251(86), 252(54, 133), 294, 295, 297
Porter, R. S., 588(183), 610
Porterfield, N. T., 258(208), 259(208), 301
Porterfield, V. T., 258(203), 300
Portingell, G. C., 337(134, 135), 338(143), 345, 346
Porubsky, J., 591(249), 592(249), 593(249), 601(249), 613
Poseu, L., 245(69), 294

Potts, T. E., 591(240), 592(240), 612

Powell, E., 186(68a, 68b), 234
Prager. S., 215(120), 235
Prausnitz, J. M., 449(95), 473(95), 474(95), 550
Preda, S., 244(45), 293
Preedy, J. E., 322(46), 324(46), 343
Pretice, J. S., 180(47), 233
Prezedziak, J., 258(205), 300
Price, C. C., 177(34), 233
Price, M. D., 517(156), 528(156), 534(156), 552
Procter, D. E., 121(131), 158
Prodan, L., 240(7), 245(69), 292, 294
Prud'homme, R. E., 558(16), 560(31, 32), 561(31, 32), 566(16), 605
Psezeckij, S. J., 413(108), 432
Ptitsyn, O. B., 398(3), 429
Purcell, T. O., Jr., 562(63), 578(63), 581(140), 606, 609
Purchase, I. F. H., 252(138), 253(141), 297
Purdon, J. R., 579(114), 608
Pyszora, H., 311(4), 342
Quackenbos, H. M., Jr., 500(157), 501(157), 552
Quand-Tho-Pham, 65(34), 155
Quarg, W., 401(42), 430
Quave, S. A., 285(300), 306
Quinn, E. J., 648(94), 654
Quinn, H. J., 288(311), 307

Quisenberry, J. G., 341(150), 346

Quy, R. B., 125(149), 158
Rabinovitch, E. B., 337(132), 341(150), 345, 346
Racin, R., 183(61a), 233
Radike, M. J., 246(79), 252(140a), 295, 297
Radu, M., 244(45), 293
Radue, R. W., 544(107), 550
Ragazzini, M., 155, 404(62), 431
Railsback, H. E., 599(368), 616
Raines, R. H., 163(7), 167(7), 171
Ralo, K. S., 259(214), 301
Ramp, F. L., 619(1), 621(11), 623(1), 638(1), 640(11), 644(1), 645(1), 648(1), 650(1), 651
Ranby, B. G., 584(152), 609, 623(40), 626(40), 643(78), 652, 653

Page 682
Rance, D. G., 112(115), 135(115), 157
Randall, J. C., 218(130), 236
Rangnes, P., 147(171), 148(171), 151(178), 159, 186(66), 234
Rao, S. P., 589(193), 597(193), 611
Raracheva, L. A., 166(19), 167(19), 169(19), 171
Ratti, L., 155, 404(62), 431
Raucher, D., 538(182), 540(182), 541(182), 553
Ravey, M., 107(100a), 157, 217(125), 218(125, 132), 219(125), 221(125), 235
Ravis, D., 266(268), 272(268), 303
Ravve, A., 588(186), 589(186), 611
Ray, W. H., 138(160a, 160b), 159
Razinskaya, I. N., 446(160), 485(160), 493(159), 538(158), 542(158), 552
Razumovskii, S. D., 401(44), 430
Reddish, W., 457(161), 552, 645(86), 654
Reding, F. P., 68(43), 155
Reed, M. C., 510(163), 513(162, 163), 552
Reefl, H., 63(26a), 154
Regnautt, V., 232
Rehor, H., 639(69), 653
Reich, P., 33(27), 44
Reicherdt, W., 411(99), 426(99), 432, 639(69), 653
Reid, R. T., 591(243), 612
Reilly, P. M., 176(30), 233
Reinl, W., 251(114, 115), 296

Reitz, R. H., 256(183), 299

Remp, P., 594(292), 597(292), 600(292), 614
Rempp, F., 594(290, 291), 597(291), 600(290, 291), 614
Rempp, P., 315(17), 342
Renner, D., 484(44), 548
Renshaw, J. T., 455(180), 477(180), 486(110, 164, 165, 166, 167), 534(110), 550, 552, 553
Retting, W., 341(153), 346,
[Retting, W.] 643(79), 653
Rettore, R., 560(33b), 562(33b), 605
Reuber, M. D., 255(166a), 267(166a), 298
Reuter, H., 592(259), 613
Revillon, A., 587(171), 593(289), 594(289, 298, 299), 600(289, 298, 299), 610, 614
Reynolds, C. E., 122(138), 158, 283(294i) , 306
Reynolds, E. S., 243(38), 293
Ricco, T., 561(52), 606
Rice, J. M., 253(145), 298
Richards, R. J., 260(230, 231), 301, 302
Richardson, C. R., 252(139), 253(141), 297
Riezko, A., 229(148), 236
Riddick, J. A., 100(90), 156
Riddle, B. L., 259(206, 215), 300, 301
Rieche, A., 409(87), 432
Riegel, H., 41(28), 43(28), 44
Riegel, R., 41(25), 43(25), 44
Riess, G., 587(172), 610

Rigo, A., 64(27), 154, 399(13), 400(25), 430

Riley, D. W., 184(63), 234
Rimnac, C. M., 582(148), 585(148), 609
Rink, M., 561(52), 606
Riou, N., 591(226), 612
Riskina, M. A., 534(114), 551
Roaldi, A., 332(91), 344
Roberts, C. W., 577(103), 608
Robeson, L. M., 232, 539(168), 540(168), 552, 556(11), 557(11), 558(11, 20), 559(11),
561(11, 20, 56), 565(11), 566(11), 568(11), 570(11), 572(11), 573(11), 575(11, 20, 92),
576(92), 577(11), 578(11), 579(11), 583(11, 149), 585(11), 588(179), 605, 606, 607, 609, 610
Robila, G., 318(34), 320(39), 343, 399(15), 400(23), 430, 624(48), 652
Robinson, M. E. R., 311(4), 342
Rockwell, K. M., 649(101), 650(101), 654

Page 683
Rodella, A., 68(188), 160
Roe, R. J., 322(55), 343
Rogers, C. E., 176(24c), 232
Rohrl, E., 579(125), 608
Role, H., 271(276d), 304
Roll, R., 257(194a), 300
Rolle, C. J., 535(119), 536(119), 551
Romovskii, E. S., 231(165), 237
Ronby, B. G., 583(150, 151), 609
Rose, F. A., 260(230, 231), 301, 302
Rosenkranz, H. S., 257(184), 299
Roskin, E. N., 231(165), 237
Roskin, Y. S., 176(26), 233
Ross, S. D., 176(25), 219(25), 233
Rossiter, C. E., 250(112), 296
Rostami, S., 561(26), 605
Roteman, J., 535(118, 119), 536(118, 119), 551
Roth, D., 257(188), 300
Rovatti, W., 537(125), 551, 605
Roventa, A., 256(177), 299
Rowe, V. K., 242(10, 13), 244(42), 256(178), 258(178, 204), 292, 293, 299, 300
Rowlinson, J. R., 473(9), 547
Rozuvalv, G. G., 272(281), 304
Rubens, L. C., 600(370), 616

Rudin, A., 119(125a), 158

Rugg, F. F., 591(240), 592(240), 612
Rusu, M., 561(28), 605
Ryan, C. F., 556(4), 578(4), 585(4), 604
Sabel, A., 66(39), 155, 402(55), 431
Sabia, R., 535(170), 537(170), 552
Sabliovschi, M., 605
Sachse, J., 579(129, 130), 609
Sacki, S., 595(341), 596(341), 615
Sadrmohaghegh, C., 611
Saeger, V. W., 544(84, 171), 549, 552
Saffron, P. M., 152(183), 159
Sala, B., 259(219), 301
Salto, E., 595(363, 349), 596(363), 616
Saito, L., 595(329), 596(329), 615
Sakai, M., 579(124), 608
Sakai, S., 168(32), 169(32), 172
Sakamoto, Y., 593(281), 614
Saladino, G., 259(219), 301
Salas, L. J., 264(259), 303
Salleem, M., 64(28), 155, 400(26), 430
Salmon, G., 456(172), 476(172), 510(172), 552
Salmon, W. A., 595(346, 351), 596(346, 351), 604(351), 616
Salovey, R., 561(53), 578(53), 606
Sampson, J. N. R., 585(156), 598(156), 609

Sandalls, F. J., 264(257), 302

Sandoz, J. P., 250(108), 296
Sangha, G. K., 260(243), 302
Santappa, M., 589(193), 597(193), 611
Sapozhnikova, I. N., 577(101), 608
Saracci, R., 250(110), 296
Sardonopoli, A. A., 561(25), 577(25), 605
Saric, M., 252(123), 297
Sarkar, N., 91(78), 156
Sartorelli, E., 256(173), 299
Sauerwald, M. J., 240(5), 242(5), 292
Savel'ev, A. P., 632(57), 635(57), 653
Sayigh, A. A. R., 561(25), 577(25), 605
Sebban, J., 591(219, 231), 612
Secchi, G. C., 256(180), 299
Sefcik, M. D., 538(182), 540(182), 541(182), 553
Seidman, J., 256(183a), 299
Seila, R. L., 264(259a), 303
Seipold, O., 623(33), 652
Seldman, I., 257(188), 300
Selikoff, I. J., 244(58), 247(86), 251(86), 259(217), 294, 295, 301

Page 684
Selivokhin, P. I., 229(152a, 152b), 237
Semon, W. L., 3(4), 24, 438(183), 553
Sen, J. N., 77(51), 155
Seppala, H. J., 479(197), 483(197), 553
Severino, F. T., 33(30), 44
Severs, L. W., 288(310), 307
Sevru, A. , 242(59), 244(59), 245(59), 294
Scalo, E., 595(330), 596(330), 615
Scarnato, C., 258(213), 301
Scary, L. K., 288(310), 307
Schaefer, J., 538(182), 540(182), 541(182), 553
Schaffner, F., 252(132), 297
Schall, W. L., 592(261), 613
Scharein, G., 286(302), 307
Schaumann, O., 240(3, 4), 242(4, 22), 243(3), 292
Schechter, R. S., 91(77), 120(77), 156
Schertz, G. L., 169(51), 172
Scheuer, P. J., 251(121, 122), 297
Scheurlen, H., 592(267), 613
Schilling, F. C., 64(29, 31), 65(32, 32a), 66(40c), 155, 318(25, 27, 28, 33), 320(35), 335(119,
120), 342, 343, 345, 353(18), 394, 399(19, 20), 400(19, 20, 22, 29, 30, 32), 401(30), 402(20,
29, 30), 405(63), 422(20), 430, 431
Schillmoyer, C. M., 35(29), 44
Schindler, A., 63(26a), 154
Schindler, H. D., 41(28), 43(28), 44

Schmidt, G., 65(37), 155, 320(41, 42), 343, 588(174), 610

Schmidt, R. F., 170(61), 173
Schmitt-Strecker, S., 561(43), 606
Schmitz, T., 242(32), 251(32), 293
Schneider, B. J., 320(37), 343
Schneider, W., 592(268), 593(268), 600(268), 613
Schneier, B. O., 561(58), 606, 643(77), 653
Schnetter, A. R., 251(116a), 296
Schosser, C. L., 335(119, 120), 345
Schottenfeld, D., 254(158), 298
Schulz, E. F., 510(163), 513(163), 552
Schulz, G. V., 52(11a), 154
Schumann, A. M., 256(183), 299
Schur, Y. J., 542(17), 547
Schurek, W., 399(14), 430
Schurer, J. W., 558(21), 565(21), 605
Schvarev, E. P., 589(197), 611
Schwab, P. A., 125(148), 158
Schwaegerle, P. E., 362(28), 394
Schwaegerle, P. R., 332(95), 344
Schwarz, H. F., 573(83), 607
Schwen, R., 162(1), 167(1), 168(1), 171
Schwenk, U., 65(35), 155, 318(32), 342, 401(49), 431
Schwetz, B. A., 253(143, 144), 297
Schwinger, E., 244(49), 294

Scmid, R. E., 260(233), 302

Scott, B., 264(260), 288(260), 303
Scott, G., 336(126, 127, 128, 129), 345, 408(73, 79), 424(134), 431, 433, 611
Scott, J. P., 169(49), 172
Scott, R. L., 436(97), 449(97), 450(97), 451(97), 457(97), 475(97), 550
Scullin, J. P., 531(174), 552
Sears, J. K., 436(178), 440(179), 441(179), 445(179), 447(179), 453(178), 454(178), 455(58,
180), 456(178), 457(178), 471(58, 178), 473(58), 475(58), 476(58), 477(180), 479(58),
485(58), 487(178), 490(58, 179), 512(181), 514(181), 519(179), 520(178), 521(58, 175, 177,
178), 522(178), 523(145, 178), 524(177, 178), 525(145, 178), 526(57, 145, 178), 527(56,
178), 531(57), 535(179), 541(178), 542(178), 549, 551, 552, 553

Page 685
Shakirzyanova, S. S., 542(201), 553
Sham, C. K., 589(195), 604(195), 611
Shapiro, J. S., 410(94, 95), 432
Sharetskii, B. M., 589(197), 611, 616
Sharma, K. K., 595(319), 597(319), 598(319), 599(319), 600(319), 615
Sharp, L. J., 428(145), 433
Shaw, M. T., 232, 556(11), 557(11), 558(11, 18), 559(11), 561(11, 24), 565(11), 566(11),
568(11), 570(11), 572(11), 573(11), 575(11), 577(11), 578(11), 579(11), 583(11), 585(11), 605
Sheehan, C. J., 456(184), 473(184), 498(184), 553
Sheikin, V. I., 595(353), 596(353), 616
Sheinker, A. P., 168(35), 172
Shelley, P. G., 43(31), 44
Sheppard, M. J., 195(77), 234
Shevlyakov, A. S., 480(185), 484(185), 541(185), 553
Shiffler, A., 484(44), 548
Shiina, K., 594(293), 600(293), 614
Shikano, T., 579(132), 592(132), 603(132), 609
Shinoda, K., 151(177), 159
Shinozaki, D. M., 341(152), 346
Shipman, J. J., 311(1), 342
Shiraishi, M., 579(121), 608
Shirinyan, G. S., 255(167), 299
Shlykova, M. N., 324(66), 343
Shockcor, J. P., 626(52), 628(54), 630(54), 653

Shockney, J. C., 621(12), 651

Shono, T., 237
Shorr, L. M., 217(125), 218(125), 219(125), 221(125), 235
Short, R. D., 253(142), 297
Shtarkman. B. P., 125(149a), 158, 446(160), 485(160), 493(159), 552
Shtraikhman, G. A., 168(37), 172
Shur, Y. J., 540(195), 541(195), 542(195), 553, 561(45), 583(150), 606, 609
Sieglaff, C. L., 181(56, 57, 58), 183(61c), 233, 332(96), 344
Sikkema, A. J., 577(103), 608
Sills, E. J., 43(31), 44
Simak. P., 353(13), 394
Simkin, D. J., 454(20), 547
Simms, W. C., 530(186), 553
Simonato, L., 259(219), 301
Simonds, H. R., 448(187), 553
Singer, S., 68(46), 155
Singh, H. B., 264(259, 260, 262a), 288(260), 303
Singh, P., 640(74), 653
Singh, R. P., 561(41, 42), 565(4), 606
Singh, Y. P., 561(41, 42), 565(41), 606
Singleton, C. J., 324(64), 326(81), 343, 344
Sionakidis, J., 603(378), 617
Sitkowska, J., 236
Sitting, M., 28(32), 30(32), 31(32), 32(32), 45, 48(1), 154
Sjoblom, T., 600(369), 616

Skibo, M. D., 582(142), 583(142), 609

Skiest, E. N., 556(4), 578(4), 585(4), 604
Skinner, S. M., 537(125), 551
Skoultchi, M., 419(119, 120), 433
Skowronski, T. S., 561(49), 589(199), 600(199), 606, 611
Slacik-Erben, R., 257(194a), 300
Slagowski, E. L., 582(141a), 592(141a), 609
Slichter, W. P., 595(352), 596(352), 616
Slonaker, D. F., 180(49), 233
Small, F. H., 254(162a), 298
Small, P. A., 450(188), 553
Smalley, E. W., 37(33), 45
Smets, G., 168(38), 170(58), 172, 173, 229(149a, 149b), 236, 587(167, 168, 169), 610
Smidrsd, O., 332(92), 344
Smirnov, Yu. A., 113(120), 157

Page 686
Smith, A. C., 257(188), 300
Smith, A. H., 250(95, 96), 259(95), 295, 296
Smith, F. A., 253(143, 144), 297
Smith, G. F., 650(105), 654
Smith, M. W., 230(155), 237
Smith, P. M., 244(58a), 294
Smith, R. R., 595(309), 615
Smith, R. W., 322(57), 343
Smith, W. E., 595(332), 615
Smith, W. M., 232
Smith, W. V., 146(167), 159
Smyth, C. P., 644(80), 654
Smyth, H. F., Jr., 258(200), 260(232), 300, 302
Snell, J., 288(312), 307
Snyder, P. J., 272(278), 304
Snyder, W. H., 273(278b), 304
Sobjima, S., 623(42), 652
Sokolova, K. I., 594(297, 302), 597(297), 600(297, 302), 614
Solc, K., 556(13), 605
Soleo, L., 256(183d), 299
Solomon, O. F., 594(306), 600(306), 614
Solovey, R., 332(91), 344
Solvic, S. A., 590(215), 611
Sonderhof, D., 401(45), 402(51, 56, 57), 403(56), 405(57), 430, 431

Soni, Y., 89(68, 69), 156

Sonin, E. V., 632(57), 635(57), 653
Soong, D. S., 57(19a), 154
Sorg, E. H., 519(189), 553
Sorokin, D. S., 231(165), 237
Sorvik, E. M., 66(41), 102(92a), 155, 157, 315(19), 318(29, 30), 320(29, 43), 335(29), 342,
343, 400(24), 402(24), 412(103), 430, 432
Soutar, C. A., 259(222, 222a), 301
Spector, M. L., 30(18, 19), 31(17), 32(19), 33(17), 41(18), 44
Speek, A. J., 246(80), 295
Speirs, R. M., 112(114), 135(114), 157
Spelmanes, G. R., 534(114), 551
Spencer, G., 258(204), 300
Spencer, R. S., 645(85), 654
Sperling, L. H., 556(7, 8), 557(7, 8), 559(8), 562(7, 62), 567(7), 579(7), 580(7), 582(7),
586(7), 580(7), 582(7), 586(7), 602(7, 8), 603(378), 604, 606, 617
Spevacek, J., 600(275), 613
Spinozzala, A., 260(224), 301
Spiritas, R., 242(15), 245(15), 292
Spit, Ing B. J., 246(81), 295
Spitz, R., 409(82), 432
Squire, R. A., 257(196), 300
Squirrell, D. C., 350(2, 3, 4), 393
Srkhipova, L. I., 632(57), 635(57), 653
Stabenow, J., 580(136), 609
Staffin, G. D., 448(14), 547

Stafford, J., 247(88), 248(88), 295

Stahl, G. A., 3(4), 24
Stahl, W. H., 532(190), 553
Stallones, R. A., 254(155), 298
Stanescu, P., 256(177), 299
Stannett, V., 588(180), 610
Starks, C. M., 27(22), 28(22), 30(22), 31(22), 32(22), 34(22), 37(22), 42(22), 43(22), 44,
286(303), 307
Starnes, W. H., Jr., 64(31), 65(32, 32a), 66(40c), 155, 318(25, 27, 28, 33), 332(102),
333(107, 110), 335(115, 116, 119, 120, 121, 122), 336(102, 107), 342, 343, 344, 345,
353(18), 394, 399(19, 20, 21, 22), 400(19, 20, 21, 22, 29, 30, 32), 401(21, 30), 402(20, 29,
30), 405(63), 409(21, 84, 85), 410(91, 92, 93), 422(20), 423(125, 126), 424(21), 427(140),
430, 431, 432, 433, 647(92), 654
Steer, R. C., 493(191), 495(191), 553
Stehlicek, J., 208(100), 235
Stein, G., 244(55, 57), 294
Stein, R. S., 478(3, 199), 547, 553
Stein, Z., 254(160), 298
Steinbach-VanGaver, G., 621(18), 651
Stemmer, K. L., 246(79), 295
Stepek, J., 519(192), 553

Page 687
Stephan, R., 579(129, 130), 609
Stephenson, R. C., 18(20), 24
Stephenson, T., 324(64), 343
Stern, S. A., 121(130), 158
Steward, R. D., 256(182), 299
Stewart, F. D., 170(61), 173
Stewart, R. D., 242(21), 256(183e), 292, 299
St. Hill, C. A., 256(172), 299
Stiles, R. E., 264(259), 303
Stinson, S., 3(6), 24
Stinson, S. T., 184(62), 234
Stiratelli, B. A., 376(41), 395
Stodola, F. H., 532(80), 549
Stott, W. T., 256(183), 299
Stout, W. J., 537(120), 551
St. Pierre, L. E., 195(76), 234, 408(71), 409(81), 413(71), 431
Straiton, T., 585(156), 598(156), 609
Strathmann, H., 215(121), 235
Straus, S., 333(106), 345
Strickland, T. W., 242(23), 292
Strini, J. C., 37(36), 45
Stromberg, R., 333(106), 345
Struik, L. C. E., 585(153), 609
Stubley, D., 473(9), 547

Stude, J., 592(259), 613

Stuhlen, F., 532(194), 553
Stuick, L. C. E., 341(154), 346
Stupen, L. V., 165(17), 166(19), 167(17, 19), 168(17), 169(17, 19, 43), 171
Styles, J. A., 260(229), 301
Suciu, F., 244(51), 294
Suciu, I., 245(69), 294
Suciu, J., 244(48), 293
Suganuma, T., 593(281), 614
Sugimoto, K., 591(230), 612
Summers, J. W., 337(132, 133), 341(150), 345, 346
Sund, E., 58(20), 154, 332(91), 344
Sundgren, N., 540(195), 541(195), 542(195), 533, 561(45), 606
Suprunchuk, T., 428(147), 433
Suzuki, T., 253(141a), 297, 401(34), 430, 595(362),
[Suzuki, T.] 596(362), 616, 620(5), 651
Suzuki, Y., 246(83a), 295
Svegliado, G., 623(41), 629(41), 652
Svanson, S. E., 561(27), 605
Svetly, J., 404(59), 431, 624(46, 47), 626(47), 652
Swanson, S. W., 561(27, 34), 605
Swantkowski, R. E., 3(5), 24
Sweely, J. S., 537(120), 551
Sweet, R. B., 256(170), 299
Sykora, S., 591(249), 592(249), 593(249), 601(249), 613

Szabo, S., 243(38), 293

Sze, M. C., 41(28), 43(28), 44
Szewczyk, P., 358(26), 394
Szymczak, T. J., 595(347, 348), 596(347, 348), 616
Tabak, H. H., 285(300), 306
Tabershaw, I. R., 256(171), 295, 299
Tachibana, H., 579(132), 592(132), 603(132), 609
Taft, P., 264(254), 302
Taft, W. H., 169(49), 172
Tahan, M., 408(79), 431
Taiana, S., 621(16), 651
Takada, M., 579(132), 592(132), 603(132), 609
Takadono, S., 623(38), 652
Takagi, N., 623(42), 652
Takahashi, A., 592(250, 263), 595(314, 315), 598(315), 599(315), 600(263, 314, 315), 613,
615
Takamori, S., 587(163), 610
Takayanagi, M., 567(70), 607
Takemoto, K., 162(3), 165(16), 167(3, 16), 168(30, 39, 40), 169(16, 41), 171, 172, 217(129b),
236
Takeshita, I., 579(132), 592(132), 603(132), 609
Takeuchi, K., 212(107), 235
Takeuchi, T., 167(24), 173(24), 171, 623(44, 45), 626(44, 45), 646(44, 45), 652
Takitani, T., 593(281), 614
Talamini, G., 48(6), 49(6), 58(6),

Page 688
[Talamini, G.] 64(27), 67(184), 154, 160, 180(46), 233, 326(78), 344, 399(13), 400(25),
412(101), 430, 432, 560(33b), 562(33b), 605
Tamburro, C. H., 250(99, 101, 102, 103, 104, 105, 108), 252(103, 104), 296
Tamburro, H., 247(87), 250(87), 295
Taminaka, T., 595(321), 615
Tammela, V., 593(274), 600(274), 613
Tanaka, M., 237
Tanigami, T., 561(39), 606
Tanny, G. B., 215(120), 230(158, 159), 235, 237
Tarkowsk, H. L., 228(140a), 236
Tarkowski, S., 255(164a), 298
Tarpley, A. R., 311(7, 8), 342
Tatum, C. F., 244(61), 294
Tavan, M., 326(78), 344
Taylor, G. N., 595(357), 596(357), 616
Taylor, J. H., 621(9), 651
Taylor, R. B., 493(196), 553
Teague, G. S., 454(49), 455(49), 471(49), 548
Teirstein, A. S., 259(217), 301
Teitell, L., 530(18), 531(18), 532(18), 547
Telles, N. C., 247(86), 251(86), 295
Temnikov, A. N., 442(129), 551
Teplov, B. F., 442(129), 551
Terwiesch, B., 90(74), 100(74), 156

Terwiesch, C., 272(283), 304

Testa, F., 117(124), 158
Tester., D. A., 337(136), 346, 578(109), 585(109), 608
Tetley, T. D., 260(226), 301
Thallmaier, M., 398(8), 405(64), 429, 431
Thame, N. G., 595(316, 317), 598(316, 317), 600(316, 317), 615
Thampy, R. T., 604(74), 653
Theil, M. H., 232
Theriault, G. P., 247(89, 90), 250(90), 253(150, 151), 259
[Theriault, G. P.] (90), 295, 298
Thiele, K., 589(201), 598(201), 600(201), 611
Thiess, A. M., 242(32), 251(32), 252(136), 293, 297
Thiis-Evensen, E., 252(137), 297
Thinius., K., 398(6), 429
Tho, P. Q., 621(8), 623(8), 651
Thomas, C. M., 232, 333(108), 345, 409(88), 422(88), 432
Thomas, D. A., 603(378), 617
Thomas, L. B., 244(54), 247(86), 251(86), 252(54, 133), 294, 295, 297
Thompson, B. R., 163(7), 167(7), 171
Thompson, R. B., 164(11), 167(11), 169(11), 170(11), 171(11), 171
Thornley, P. E., 259(222), 301
Tichenor, R. M., 493(136), 551
Tidwell, P. W., 176(27), 233
Tiers, G. V. D., 320(36), 343, 400(31), 430
Til, H. P., 246(80), 295

Timar., J., 589(200), 611

Timmerman. D., 170(66), 173
Tkachenko, G. V., 165(15, 17), 166(19), 167(15, 17, 19), 168(17), 169(17, 19, 43), 170(15),
171, 172
Toba, Gh., 244(46), 293
Tobolsky, A. V., 132(154), 143(163), 158, 159, 176(24b, 24c), 232, 478(3), 493(196), 547,
553
Tokudome, S., 251(116b), 296
Tola, J., 257(191), 300
Tomari, U. J., 64(30), 155
Tomari, Y., 402(52), 431
Tomasini, M., 256(173), 299
Tompa. H., 105(98), 157
Tooke, P. B., 312(12), 342
Toomey, J., 264(260), 288(260), 303
Tordella, J. P., 572(76), 607
Torkelson, T. R., 242(10, 13), 244(42), 256(178), 258(178), 292, 293, 299
Tornell, B. E., 135(157a), 159
Touchette, N. W., 440(179), 441(179), 447(179), 455(179),

Page 689
[Touchette, N. W.] 471(58), 473(58), 474(58), 475(58), 476(58), 479(58, 197), 483(197),
485(58), 490(58, 179), 519(179), 526(27), 531(57), 535(179), 549, 552, 553
Tran, C., 558(20), 561(20), 575(20), 605
Trautvetter, W., 621(10), 638(10), 651
Tregan, R., 332(97), 344
Trieschmann, H. G., 622(26), 652
Trivedi, P. D., 578(105), 605, 608
Troitskaya, L. S., 334(113) 345, 418(116), 425(116), 432
Troitskii, B. B., 334(113), 345, 418(116), 425(116), 432
Trumball, H. L., 170(61), 173
Tsai, T. C. H., 37(3), 44
Tseng, M. T., 250(108), 296
Tserda, H., 165(14), 167(14), 168(14), 171
Tsou, P., 326(81), 344
Tsuge, S., 623(44, 45), 652
Tsunashiami, Y., 157
Tu, R. S., 595(360), 596(360), 616
Tuchsanyi, B., 195(78), 234, 333(104), 336(123), 344, 345, 408(70, 72, 76), 411(98),
412(102), 431, 432
Tucker, E. S., 544(171), 552
Tudos, F., 333(104, 112), 336(112, 123), 344, 345, 408(70, 72, 77), 409(90), 411(98),
412(102, 104), 413(104), 431, 432, 595(313), 597(313), 600(313), 615
Tumarkin, N., 168(34), 172
Tunkel, S. J., 282(322), 308
Turkanis, J., 582(144), 609

Tyrackova, V., 119(128a), 158

Tyroler, H. A., 250(95, 96), 259(95), 295, 296
Ueberreiter, K., 495(198), 500(198), 553
Ueda, T., 212(107), 235
Uehleke, H., 257(194a), 300
Ueno, H., 611
Ueno, Y., 594(294), 600(294), 614
Ugelstad, J., 58(20), 147(169, 171), 148(171), 154, 159, 332(92), 344
Ulbricht, J., 217(125a), 236
Ulland, B. M., 246(88b), 295
Ulrich, H., 561(25), 577(25), 605
Umeya, K., 579(120), 608
Uno, T., 168(31), 172
Utsuo, A., 478(199), 553
Uustalu, J. M., 135(157a), 159
Vaccari, P., 622(24b), 652
Valaczkay, C., 244(45), 293
Valaskai, M., 244(48, 51), 293, 294
Valgmigh, L., 258(213), 301
Van Amerongen, G. J., 456(172), 476(172), 510(172), 534(172), 552
VanDenHeuvel, C. J. M., 335(118), 345
Vanderbilt, B. M., 520(75), 549
Vander Hoff, B. M. E., 149(174), 159
Vanderhoff, J. W., 228(140a, 140b, 140c), 236
VanderMaas, J. H., 350(7), 353(7), 394

VanderMeer, R., 176(29), 233

VanDuuren, B. L., 256(183a), 257(184, 188), 299, 300
VanErt, M., 242(15), 245(15), 292
VanOene, H., 606
VanOrdstrand, H. S., 259(220), 301
VanRijckavorsel, J., 184(65), 234
Vansheidt, A. A., 168(37), 172
VanWeizer, J. R., 184(62a), 234
Varma, I. K., 595(319), 597(319), 598(319), 599(319), 600(319), 615
Varnell, D. F., 558(15, 17), 560(15), 561(15), 605

Page 690
Vasiliu-Oprea, C., 578(110), 608
Vaupel, J. W., 263(249), 302
Veltman, G., 244(49, 55, 57), 294
Ventkin, Yu. I., 260(227), 301
Veralin, C. N., 268(276a), 304
Verdu, J., 418(127), 423(127), 424(127, 132, 133), 433
Verhelst, W. F., 79(56), 80(56), 82(56), 156, 273(286), 305
Verhoek, F. H., 500(200), 553
Vianello, G., 117(124), 124
Vianna, N., 251(119), 296
Vicario, G. P., 428(142), 433
Vidotto, G., 180(46), 233
Vielhaber, R. G., 620(2), 621(20), 622(30), 651, 652
Villacorta, G. M., 66(40c), 155
Villacorta, M., 405(63), 431
Villoutreux, G., 560(33a), 562(33a), 605
Vindetto, G., 67(184), 160
Viola, P. L., 243(40), 245(40), 246(77), 293, 295
Visani, F., 155
Visani, M., 404(62), 431
Vlachopoulos, K. W. J., 341(152), 346
Vogelfanger, E. A., 3(7), 5(7), 17(7), 26
Von Bezdadeas, E., 353(16), 394
von Oettingen, W. F., 258(211), 301

Vorenkamp, E. J., 561(35), 605

Voskresenskii, V. A., 542(201), 553
Vrata, M., 167(24), 173(24), 171
Vrillon, C., 622(24), 652
Vroom, W. I., 595(335, 356, 358) 596(355, 356, 358), 616
Vyas, H. R., 581(139), 609
Vymalalova, Z., 424(132, 133), 433
Vymazal, A., 519(192), 553
Vymazal, Z., 424(132, 133), 433
Vyvoda, J. C., 324(67, 68), 338(142), 343, 346, 408(79), 431
Wade, W. H., 91(77), 120(77), 156
Wagner, H., 68(46), 155
Wagoner, J. K., 250(110), 253(152), 296, 298
Waite, C. F., 240(1), 292
Wakabayashi, T., 621(17), 651
Wakeman, I. B., 260(244), 302
Wallace, J. R., 284(294p), 306
Walling, C., 289(321), 308
Walker, A. E., 244(66), 245(66), 294
Walker, L. E., 582(141a), 592(141a), 609
Wall, F. T., 232
Wall, L. A., 177(35), 233
Wallder, V. T., 526(62), 528(62), 549
Walling, C., 50(9), 51(13), 53(13), 108(103), 154, 157, 162(4), 164(9), 167(4, 9), 168(4, 9),
169(52), 170(9), 171, 172, 232

Walsh, D. J., 561(26), 589(195), 604(195), 605, 607, 611

Walter, E. R., 68(43), 155
Wambach, A. D., 561(55), 578(106), 606, 608
Wang, T. T., 617
Wang, Y. Y., 560(33), 561(33), 605
Ward, A. M., 244(66), 245(66), 294
Ward, J., 264(261), 288(261), 303
Warner, A. J., 493(202), 553
Warsaw, R., 259(217), 301
Wartman, L. H., 496(203), 553
Watanabe, P. G., 242(24, 25), 243(24, 34), 256(183), 292, 293, 299
Watase, H., 623(42), 652
Waterland, L. B., 287(305b), 307
Waterman, J. A., 107(100a), 157, 217(125), 218(125), 219(125), 221(125), 235
Wauters, E., 582(145), 603(145), 609
Waxmeiler, R., 244(56), 249(94), 250(95, 96, 98), 253(152), 259(95, 98), 294, 295, 296, 298
Weben, J. P., 622(24), 652

Page 691
Weber, A. J. M., 335(118), 345
Weber, H., 251(114, 115), 296
Weber, M., 16(14), 24
Webler, S. M., 582(142, 147), 583(142), 585(147), 609
Wechsler, H., 168(27), 169(27), 170(27), 171, 173(23), 232
Wehner, W., 410(96), 432
Weidner, K. R., 151(180), 159
Weil, C. S., 260(232), 302
Weinand, G., 229(149a, 149b), 236, 587(167, 168, 169), 610
Weinlich, J., 580(135), 609
Weintraub, L., 368(35), 394
Weiss, F., 401(33), 430
Weiwad, D., 592(268), 593(268), 600(268), 613
Welch, F. J., 68(43), 155
Weldon, L. H. P., 491(204), 553
Wells, H., 529(205), 553
Wendel, A., 318(26), 342
Wender, L., 33(13), 44
Wendisch, D., 231(160), 237
Wendt, T. M., 532(112), 550
Wenig, W. J., 324(61), 343
Wenisch, W. J., 162(4), 167(4), 168(4), 171
Werner, A. F., 544(140a), 551
Wescott, L. D., 66(40c), 155, 405(63), 431

Wessel, C. J., 531(206), 534(206), 553

Wessling, R., 48(4), 154
Westerholm, P., 251(116), 296
Wheeler, R. N., Jr., 213(117, 118), 235, 272(278), 285(296), 304, 306
Whelan, J. G., 250(103), 252(103), 296
Whetstone, C. L., 244(67), 294
Whipple, G. M., 287(305), 307
White, D. M., 322(56), 343
White, F. T., 595(354), 596(354), 616
Whitehouse, W. M., 244(63, 65), 263(63), 294
Wiberg, G. S., 264(254), 302
Wichterle, O., 320(37), 343
Wiederhorn, N., 478(3), 547
Wiel, C. S., 258(200), 300
Wiles, D. M., 428(147), 433
Wilkes, C. E., 312(11), 322(57, 58), 342, 343
Wilkins, R. M., 518(207), 553
Williagan, D. A., 246(83), 253(83), 295
Williams, G. D., 589(194), 611
Williams, G. E., 523(15), 524(208), 526(15), 528(209), 547, 553
Williams, H. L., 170(60), 173
Williamson, I., 529(205), 553
Willis, H. A., 350(4), 352(4), 393
Wilski, H., 638(67), 653
Wilson, J., 260(229), 301

Wilson, J. C., 112(113, 114), 135(113, 114), 157

Wilson, J. E., 528(210), 553
Wilson, J. L., 288(317), 307
Wilson, R. H., 244(61), 294
Wilson, W. A., 579(114), 608
Wimer, W. E., 33(38), 45
Wingrave, J. A., 125(148), 158
Winkler, D. E., 333(105), 344, 409(80), 431, 519(211), 553
Winston, J. M., 246(81), 253(142), 295, 297
Wintzer, R., 589(201), 592(268), 593(268), 598(201), 600(201, 268), 611, 613
Wippler, C., 529(212), 553, 591(226, 228), 612
Wirth, H. O., 332(100), 344, 410(96), 418(115), 421(115), 422(115), 432
Witenhafer, D. E., 112(116), 114(116), 135(116), 157, 283(294e), 306, 322(51), 324(51),
326(84), 327(84), 343, 344
Withey, J. R., 243(36a), 293
Wittstock, T., 579(131), 609
Wiys, J. J. A., 401(41), 430
Wolf, H. J., 592(259), 613
Wolf, K. A., 232
Wolf, K. F., 217(125a), 236
Wolf, M., 66(40), 155, 404(61), 431
Wolf, P. H., 213(114), 235
Wolfe, N., 266(268), 272(268), 303
Wolstenholme, J., 269(287a), 275(287a), 305

Page 692
Wong, O., 242(30, 31), 246(30, 31), 251(30, 31), 293
Wood, L. A., 199(73), 234
Wood, N. G., 425(138), 426(139), 433
Wood, P. R., 267(274e), 304
Woods, J. S., 246(81), 295
Woods, M. E., 574(89), 575(89, 90), 607
Work, G. A., 289(320b), 308
Worman, C. H., 218(131), 236
Wrackmeyer, B., 318(32), 342, 401(49), 431
Wright, P. L., 544(107), 550
Wulff, D., 589(201), 598(201), 600(201), 611
Wunderlich, B., 639(72, 73), 653
Wyant, R. E., 530(213), 554
Yakovlena, M. K., 168(35), 172
Yalkut, S. I., 543(8), 547
Yamaguchi, T., 544(111), 550
Yamamoto, A., 579(121), 608
Yamashita, S., 587(163), 610
Yamazaki, N., 169(44), 172
Yant, W. P., 240(1), 292
Yarze, J. C., 33(13), 44
Yeaden, R. E., 473(9), 547
Yeager, C. C., 531(214), 554
Yee, A. F., 565(67), 566(67), 568(67), 579(67), 606

Yen, N. T., 591(236), 600(236), 612

Yezrielev, A. J., 176(26), 233
Yonezu, S., 621(14), 651
Yoshida, K., 167(23), 168(25, 28, 29, 30), 169(53), 170(57), 171, 172
Yoshimera, T., 251(116b), 296
Yoshitake, T., 591(230), 612
Young, D. P., 37(39), 45
Young, J. D., 242(25), 293
Young, J. T., 259(214), 301
Young, L. J., 107(101), 157
Young, R. D., 623(39), 652
Young, R. J., 580(122), 582(122), 608
Young, W. L., 621(12), 651
Yuang, C. W., 286(301d), 306
Yuri, T., 595(362), 596(362), 616
Zabolotskaya, E. V., 165(13), 167(13), 171
Zacheaus, F., 592(268), 593(268), 600(268), 613
Zacher, R. L., 512(181), 514(181), 553
Zaikov, G. E., 401(44), 430
Zamboni, V., 321(44), 343
Zeitler, G., 622(26), 652
Zeoli, L. T., 517(215), 518(215), 554
Zepp, R., 266(268), 272(268), 303
Zerbi, G., 321(44), 343
Zhu, C. L., 286(301d), 306

Zhurhova, S. V., 616

Ziche, M., 260(236), 302
Zichy, E. L., 112(112, 113, 114, 115), 135(112, 113, 114, 115), 157
Zilio-Grandi, F., 399(13), 430
Zimm, B. H. J., 314(15), 342
Zitnansky, B., 519(217), 554
Zook, W. C., 288(311), 307
Zorica, M., 252(123), 297
Zsuffa, B., 620(7), 651
Zufall, J., 368(35), 394
Zufarov, A. A., 591(238), 612
Zurlo, N., 256(180), 299
Zutty, N., 164(12), 167(12)5 171, 168(33), 172
Zyball, A., 595(342), 596(342), 615

Page 693

Subject Index

Acetates, solubility and compatibility with PVC of, 458

Acetylene, hydrochlorination of, for vinyl chloride production, 3741
Acids, solubility and compatibility with PVC of, 458
Acrylate ester copolymers, 229
Acrylates, solubility and compatibility with PVC of, 458
Acrylic-acrylonitrile copolymers, 229231
Acrylonitrile, 230231
Adipates
ease of fusion with PVC of, 481
performance in PVC of, 502
solubility and compatibility with PVC of, 458459
Aging of PVC blends, 584585
Air, plasticization by, 535537
Air release property of dispersion resins, 380383
Alcohols, solubility and compatibility with PVC of, 459
Aldehydes, solubility and compatibility with PVC of, 459
American Society for Testing and Materials (ASTM)
resin classifications of, 387391
test methods for resins of, 348349
Amides, solubility and compatibility with PVC of, 459
Analytical testing of PVC resins, 350362

chlorine analysis, 350

infrared analysis, 350354
microscopy, 362
molecular weight, 354
molecular weight distribution, 354359
thermal analysis, 359361
Antiplasticization, 538542
crystallinity theory of, 538539
hindered molecular theory of, 540542
reduced free volume theory of, 539
Antiplasticizers, 538542
Apparent (bulk) density of general-purpose resins, 362363
Azelates
ease of fusion of PVC with, 481
performance in PVC of, 502
solubility and compatibility with PVC of, 460
Bacterial resistance of plasticized PVC, 530534
Barium-cadmium stabilizers, 420421
Barium-zinc stabilizers, 421
Belgium, PVC production in (1983), 6
Benzoates
ease of fusion of PVC with, 481

Page 694
[Benzoates]
solubility and compatibility with PVC of, 460
Blending of PVC, 571586
Block copolymers, 586588
Brazil, PVC production in (1983), 6
Bulk polymerization, 127138
resin productivity, 131133
resin quality, 133138
by V-50 modification, 130
Butyrates, solubility and compatibility with PVC of, 460461
Calcium-zinc stabilizers, 421
Canada, PVC production in (19801983), 6, 7
Carcinogenicity
of 1, 2-dichloroethane, 258259
of trichloroethylene, 256258
of vinyl chloride, 245252
Cationic grafting of copolymers to PVC, 594595
Celluloid, 438
Cellulose nitrate, 437439
Chemical resistance of CPVC, 636637
China, PVC production in (1983), 6
Chlorinated PVC (CPVC), 619654
market and commerical uses for, 648650
methods of chlorination, 620623

properties of, 636648

chemical resistance, 636637
dielectric properties, 644646
mechanical properties, 642644
rheological properties, 640641
thermal properties, 638640
structure of CPVCs and mechanism of chlorination, 623631
thermal decomposition characteristics of, 646648
thermal stability and stabilization, 632636
Chlorine analysis of PVC resins, 350
Citrates
ease of fusion of PVC with, 381
performance in PVC of, 503
solubility and compatibility with PVC of, 461
Compounding of PVC, future trends in, 1618
Contamination
of dispersion resins, 383
of general-purpose resins, 365366
Continuous dry-blend compounder, 17
Copolymer processes and manufacture, 161237
acrylic-acrylonitrile copolymers 229231
acrylate ester copolymers, 229
acrylonitrile, 230231
commercial processes used, 199200
copolymerization process, 164178

future developments, 231

market scope and size, 162164
olefin copolymers, 216229
ethylene copolymers, 223229
propylene copolymers, 216223
structure and properties, 178199
glass transition temperature, 195199
rheology, 178194
thermal stability, 195, 196197
vinyl acetate copolymers, 200215
emulsion polymerization, 215
solution polymerization, 213215
suspension polymerization, 200213
Corrosion resistance of plasticized PVC, 518519
Crystallinity of PVC, 320324
Crystallinity theory of antiplasticization, 538539
Degradation of PVC, 398416
dehydrochlorination, 405411
photolytic degradation, 413416
structural irregularities, 399405
thermal degradation, 411413
Dehydrochlorination of PVC, 405411

Page 695
Dibenzoates, performance in PVC of, 502503
Dichloroethane, physical properties of, 243
1, 2-Dichloroethane
carcinogenicity of, 258259
pyrolysis of, 3437
toxicity of, 258
Dielectric properties
of CPVC, 644646
of PVC resins, 456
Dispersion resins, physical properties of, 377387
air release, 380383
contamination, 383
gelation-fusion characteristics, 384387
gelation temperature, 384
particle size, 379380
viscosity, 378379
volatiles, 383
East Germany, PVC production in (1983), 6
Economic impact of the PVC industry, 1314
Electrical properties
of general-purpose resins, 376377
plasticization of PVC and, 490491
Emulsion polymerization, 138152
kinetics of, 143150

resin productivity, 150

resin quality, 150152
of vinyl acetate copolymers, 215
Environmental and safety aspects of plasticized PVC, 10, 542544
of PVC, 1415
see also Safety and environmental concerns in resin manufacture
Environmental Protection Agency (EPA), resin manufacture and, 263268
Epoxides
performance in PVC of, 503
solubility and compatibility with PVC of, 461
Epoxy stearates, ease of fusion of PVC with, 481
Epoxy tetrahydrophthalates, ease of fusion of PVC with, 481
Ethers
performance in PVC of, 503
solubility and compatibility with PVC of, 461
Ethylene
chlorination of, 3031
oxychlorination of, 3134
Ethylene copolymers, 223229
External plasticization, internal plasticization versus, 477478
Extraction resistance of plasticized PVC, 510512
Exudation under pressure of plasticized PVC, 501510
Fabrication of PVC, future trends for, 1618
Fillers, plasticization by, 535537
Fixed-bed chlorination of PVC, 622

Fluid-bed chlorination of PVC, 622

Food and Drug Administration (FDA), resin manufacture and, 261262
France, PVC production in (19801983), 6, 7
Free-radical grafting
of copolymers to PVC, 588591
of vinyl chloride, 591593
Free-radical polymerization, 4962
Free volume theory of plasticization, 443444
Fumarates, solubility and compatibility with PVC of, 461
Fungal resistance of plasticized PVC, 530534
Future trends in the PVC industry, 1518
Gases, safety procedures for handling vinyl chloride wastes in, 286287
Gelation-fusion properties
of dispersion resins, 384387
of PVC, 336338
Gel tests for general-purpose resins, 365
Gel theory of plasticization, 442
General-purpose plasticizers, 438439

Page 696
General-purpose resins, physical properties of, 362377
apparent (bulk) density, 362363
contamination, 365366
electrical properties, 376377
gels, 365
melt viscosity, 371376
particle size, 363365
porosity, 366368
thermal stability, 368371
volatiles, 366
Glass transition temperature of PVC, 195199
Glutarates, solubility and compatibility with PVC of, 462
Glycolates
ease of fusion of PVC with, 481
performance in PVC of, 503
solubility and compatibility with PVC of, 462
Glycol derivaties, performance in PVC of, 503
Glycol esters, ease of fusion of PVC with, 481
Graft copolymers, 588601
free-radical grafting to PVC, 588591
free-radical grafting to vinyl chloride, 591593
grafting of polyfunctional monomers, 595597
ionic grafting, 593595
Great Britain, PVC production in (1983), 6

Halogenated hydrocarbons, solubility and compatibility with PVC of, 462463

Handling and storage of vinyl chloride, 43, 271275
Heat distortion temperature (HDT), PVC blends and, 577578
Heat stability effects on plasticized PVC, 519522
Hexahydro isophthalates, ease of fusion with PVC of, 481
Hexahydrophthalates, ease of fusion of PVC with, 481
Hindered molecular motion theory of antiplasticization, 540542
Historical development of PVC, 12, 2526
Hot compounding of plasticizers, 479485
Humidity compatibility of plasticized PVC, 510
Hydrocarbons
ease of fusion of PVC with, 482
solubility and compatibility with PVC of, 462464
Hydrochlorination of acetylene for production of vinyl chloride, 3741
Hydrophthalates, solubility and compatibility with PVC of, 468
Infrared analysis of PVC resins, 350354
Internal plasticization, external plasticization versus, 477478
International Organization for, Standardizaton (ISO), test methods for PVC resins of, 349
Interpenetrating polymer networks (IPNs), 601604
Ionic grafting of copolymers to PVC, 593595
Isobutyric acid derivatives, performance in PVC of, 504
Isophthalates
ease of fusion of PVC with, 482
performance in PVC of, 504
solubility and compatibility with PVC of, 468

Isosebacates
ease of fusion of PVC with, 482
performance in PVC of, 504
Italy, PVC production in (19801983), 6, 7
Japan, PVC production in (19801983), 6, 7
Ketones, solubility and compatibility with PVC of, 464

Page 697
Lactates, solubility and compatibility with PVC of, 464
Light stability of plasticized PVC, 522528
Liquid chlorine processes for chlorinating PVC, 622623
Liquids, safety procedures for handling vinyl chloride wastes in, 286287
Lubricity theory of plasticization, 442
Maleates, solubility and compatibility with PVC of, 464
Malonates, solubility and compatibility with PVC of, 464465
Markets for the PVC industry, 113, 14
Mechanical properties
of CPVC, 642644
of PVC, 339341
effect of plasticization on, 490491
of PVC blends, 577582
Mechanistic theory of plasticization, 442443
Melt viscosity
of general-purpose reisns, 371376
of PVC, 338339
Metabolism of vinyl chloride, 242243
Metal-free stabilizers, 421422
Methacrylates, solubility and compatibility with PVC of, 458
Mexico, PVC production in (1983), 6
Miscibility of polymers, 557571
Molecular structure of PVC, 309320
branching and tertiary chlorine, 319320

chain end groups, 318319

chloromethyl branches, 318319
defect structures related to double bonds, 320
head-to-head addition, 318319
molecular weight, 312318
tacticity, 309312
Molecular weight of PVC resins, 354
Molecular weight distribution of PVC resins, 354359
Netherlands, PVC production in (1983), 6
Nitriles, solubility and compatibility with PVC of, 465
Nitro compounds, solubility and compatibility with PVC of, 465
Nomenclature for PVC, 111
Occupational Safety and Health Administration (OSHA), resin manufacture and, 262263
Oleates
performance in PVC of, 504
solubility and compatibility with PVC of, 465
Olefin copolymer production, 216229
ethylene copolymers, 223229
propylene copolymers, 216223
Organotin stabilizers, 420
Oxalates, solubility and compatibility with PVC of, 465
Palmitates
ease of fusion of PVC with, 482
solubility and compatibility with PVC of, 465
Paraffin derivatives, performance in PVC of, 504

Particle size
of dispersion resins, 379383
of general-purpose resins, 363365
Pentaerythritol derivative, performance in PVC of, 505
Permanent plasticizers, solubility and compatibility with PVC of, 468469
Permeability
of plasticized PVC, 518519
of PVC blends, 583584
Petroleum derivatives, performance in PVC of, 503504
Phenols, solubility and compatibility

Page 698
[Phenols] with PVC of, 465
Phosphates
ease of fusion of PVC with, 482
performance in PVC of, 505
solubility and compatibility with PVC of, 466
Photolytic degradation of PVC, 413416
Phthalates
ease of fusion of PVC with, 482483
performance in PVC of, 505507
ortho-Phthalates, solubility and compatibility with PVC of, 466467
Physical properties
of dispersion resins, 377387
of general-purpose resins, 361377
Pigments, plasticization by, 535537
Plasticization, 435553
antiplasticizers and antiplasticization, 539542
crystallinity theory, 538539
hindered molecular motion theory, 540542
reduced free volume theory, 539
uses of antiplasticized polymers, 542
choice of plasticizer, 544547
compatibility of plasticizer with resin, 448477
primary and secondary plasticizers, 474476
temperature dependence and compatibility, 476477

theories, 448474
effect on physical properties, 490500
efficiency, 497500
mechanical, thermal, and electrical properties, 490494
temperature effects, 494495
external versus internal plasticization, 477478
historical perspective of, 437440
early technology leading to plasticized PVC, 437438
[Plasticization]
general-purpose and specialty plasticizers, 438440
mechanisms of plasticizer action, 440447
general theories, 442444
plasticization step by step, 444447
plasticizers versus processing aids, 440442
permanence properties of plasticized PVC, 500534
attack by insects, termites, and rodents, 534
controlled-release agents, 515518
corrosion resistance and permeability, 518519
extraction resistance, 510512
exudation under pressure, 501510
fungal and bacterial resistance, 530534
heat stability effects, 519522
humidity compatibility, 510
light stability and weathering, 522528
migration of contaminants, 513515

plasticizer migration, 513

radiation stability, 528530
sterilization, 530
volatility, 500501
plasticization by fillers, pigments, salts, and air, 535537
polymerizable plasticizers, 534535
of PVC, 478490
compounding of plasticizer and resin, 479486
ease of fusion, 481483, 486490
PVC blends and, 573577
toxicity and environmental aspects, 542544
Plasticizer evaluation checklist, 545
Plastisol casting, 485486
Poland, PVC production in (1983) 6
Polyesters
ease of fusion of PVC with, 483
performance in PVC of, 507
solubility and compatibility with

Page 699
[Polyesters] PVC of, 468469
Polyfunctional monomers, grafting of, 595597
Polymer chain structure, 6269
Polymerizable plasticizers, 534535
Polymer networks, interpenetrating, 601604
Polyvinyl chloride (PVC) processes and manufacture, 47159
bulk polymerization, 127138
resin productivity, 131133
resin quality, 133138
by V-50 modification, 130
commercial manufacturing processes, 6972
emulsion polymerization, 138152
kinetics of, 143150
resin productivity, 150
resin quality, 150152
free-radical polymerization, 4962
future developments for, 152153
polymer chain structure, 6269
polymerization overview, 48
suspension polymerization, 72125
resin productivity, 76103
resin quality, 103125
Porosity of general-purpose resins, 366368
Primary plasticizers, compatibility with resins of, 474476

Processing aids for PVC blends, 578

Propionates, solubility and compatibility with PVC of, 469
Propylene copolyemrs, 193194, 216223
Pyrolysis of 1, 2-dichloroethane, 3437
Pyromellitates, solubility and compatibility with PVC of, 469
Quality assurance of PVC resins, 391393
Radiation grafting of copolymers to PVC, 590591
Radiation stability of plasticized PVC, 528530
Reduced free volume theory of antiplasticization, 539
Reproductive effects of vinyl chloride, 252254
Resin manufacture, safety and environmental concerns in, 1415, 239308
analytical methods, 287289
general safety procedures, 268285
downstream operations, 284285
major accidents, 278279
monomer production, 275
polymerization, 275282
raw material handling and storage, 271275
stripping, 282284
regulatory status, 260268
Environmental Protection Agency, 263268
Food and Drug Administration, 261262
Occupational Safety and Health Administration, 262263
toxicity considerations, 240260
1, 2-dichloroethane, 258259

polyvinyl chloride, 259260

trichloroethylene, 255258
vinyl acetate, 254255
vinyl chloride, 240254
vinyl acetate copolymers, 289290
waste streams, 285287
liquids and gases, 286287
solids, 287
water, 285286
Resins, compatibility of plasticizers with, 448477
primary and secondary plasticizers, 474476
temperature dependence and compatibility, 476477
theories, 448474
Resin structure and properties, 309346
PVC molecular structure, 309320

Page 700
[Resin structure and properties]
branching and tertiary chlorine, 319220
chain end groups, 318319
chloromethyl branches, 318319
defect structures related to double bonds, 320
head-to-head addition, 318319
molecular weight, 312318
tacticity, 309312
PVC properties, 332341
gelation or fusion, 336338
mechanical properties, 339341
melt viscosity, 338339
thermal stability, 332336
PVC supermolecular structure, 320332
crystallinity, 320324
particulate structure, 324332
state of the glass, 324
Resin testing and quality assurance, 347395
analytical testing, 350362
chlorine analysis, 350
infrared analysis, 350354
microscopy, 362
molecular weight, 354
molecular weight distribution, 354359

thermal analysis, 359361

ASTM resin classification, 387391
PVC resins, 387390
vinyl chloride polymers, 390391
physical properties of dispersion resins, 377387
air release, 380383
contamination, 383
gelation-fusion characteristics, 384387
gelation temperature, 384
particle size, 379380
viscosity, 378379
volatiles, 383
physical properties of generalpurpose resins, 362377
apparent density, 362363
contamination, 365366
[Resin testing and quality assurance]
electricals, 376377
gels, 365
melt viscosity, 371376
particle size, 363365
porosity, 366368
thermal stability, 368371
volatiles, 366
physical properties of solution resins, 387
quality assurance, 391393

test methods, 348349

Rheological properties
of copolymers, 178194
of CPVC, 640641
Ricinoleic acid derivatives, performance in PVC of, 507
Safety and environmental concerns in resin manufacture, 1415, 239308
analytical methods, 287289
general safety procedures, 268285
downstream operations, 284285
major accidents, 278279
monomer production, 275
polymerizations, 275282
raw material handling and storage, 271275
stripping, 282284
regulatory status, 260268
Environmental Protection Agency, 263268
Food and Drug Administration, 261262
Occupational Safety and Health Administration, 262263
toxicity considerations, 240260
1, 2-dichloroethane, 258259
polyvinyl chloride, 259260
trichloroethylene, 255258
vinyl acetate, 254255
vinyl chloride, 240254
vinyl acetate copolymers, 289290

waste streams, 285287

liquids and gases, 286287
solids, 287
water, 285286

Page 701
Salts, plasticization by, 535537
Sebacates
ease of fusion of PVC with, 483
performance in PVC of, 507
solubility and compatibility with PVC of, 469
Secondary plasticizers, compatibility with resins of, 474476
Solid wastes, safety procedures for handling, 287
Solubility parameters of resins, 456
Solution polymerization of vinyl acetate copolymers, 213215
Solution processes for chlorinating PVC, 621
Solution resins, physical properties of, 387
Solvation (see Plasticization)
Solvent casting, 486
South Korea, PVC production in (1983), 6
Spain, PVC production in (1983), 6
Specialty plasticizers, 439440
Stabilization
of CPVC, 632636
of PVC, 416429
role of stabilizers in, 417422
Stearates
ease of fusion of PVC with, 483
performance in PVC of, 507
solubility and compatibility with PVC of, 469

Sterilization of plasticized PVC, 530

Storage and handling of vinyl chloride, 43, 271275
Styrene derivatives, performance in PVC of, 507
Succinates, solubility and compatibility with PVC of, 469
Sulfonamides
performance in PVC of, 508
solubility and compatibility with PVC of, 470
Sulfonates, performance in PVC of, 508
Sulfur-containing tin stabilizers, 420
Sulfur-free tin stabilizers, 420
Supermolecular structure of PVC, 320332
[Supermolecular structure of PVC]
crystallinity, 320324
particulate structure, 324332
state of the glass, 324
Suspension polymerization, 72125
resin productivity, 76103
resin quality, 103125
of vinyl acetate copolymers, 200213
Tacticity of polymer chain structure, 6667
Taiwan, PVC production in (19801983), 6, 7
Tartrates
performance in PVC of, 508
solubility and compatibility with PVC of, 470
Terephthalates

ease of fusion of PVC with, 483

performance in PVC of, 509
solubility and compatibility with PVC of, 468
Thermal analysis of PVC resins, 359361
Thermal decomposition of CPVCs, 646648
Thermal degradation of PVC, 411413
Thermal properties
of CPVC, 638640
effect of plasticization on, 490491
Thermal stability
of CPVC, 632636
of general-purpose resins, 368371
polymer chain structures and, 6366
of PVC, 195, 196197, 332336
Toxicity
of 1, 2-dichloroethane, 258
of plasticized PVC, 542544
of PVC, 259260
of trichloroethylene, 255256
of vinyl acetate, 254255
of vinyl chloride, 240254
Trichloroethylene (TCE)
carcinogenicity of, 256258
physical properties of, 243
toxicity of, 255256

Trimellitates