CEMENT and CONCRETE RESEARCH, Vol. 20, pp. 910-918, 1990. Printed in the USA.

0008-8846/90. $3.00+00. Copyright (c) 1990 Pergamon Press plc.

R e a c t i o n M e c h a n i s m s of C o n c r e t e A d m i x t u r e s

P. Paulini
Institut f~ir Baustofflehre und Materialpri.ifung
Universit~it lnnsbruck, A-6020 Innsbruck, Austria

(Communicated by F.W. [.,ocher)

(Received May 21, 1990)

ABSTRACT
Concrete admixtures influence the kinetic of cement hydration
mainly during the d o r m a n t period. The d o m i n a n t influence of admixtures seems to lie in different bound forces between dissociated ions
in the pore water solution. Repulsive forces characterize the solvation process while attractive forces d o m i n a t e during crystallization.
These changes of ion bound forces lead twice to volume changes
during phase transitions of hydration. Volume changes measured
with an immersion weighing setup show clearly the effect of concrete
admixtures on cement r e a c u o n . Retarder agents produce a volume
swelling while accelerators force an i m m e d i a t e shrinkage behaviour.
A m e c h a n i s m as introduced by Le Chatelier involving a solution-crystallization step seems to describe the hydration process most adequately. As long as repulsive forces dominate, a volume swelling
occurs and no strength gain can take place. The d o r m a n t period is
defined by the length of the swelling process. Hardening and strength
growth start at the point at which volume shrinkage appears.

Concrete admixtures are used to influence concrete hardening mainly in the early
tPrhase of hydration. By adding very small amounts of admixtures, properties of
esh conc~:ete can be influenced in a wide range. The hydraulic reaction can be
accelerated or r e t a r d e d using appropriate agents and workability can be improved
using superplasticizers. Many mechanisms have been suggested to explain all
these effects [1],[2]. There is still no full u n d e r s t a n d i n g of hydration mechanisms
including all these p h e n o m e n a at once. In this contribution emphasis shall be put
at energy transformations occuring during physico-chemical processes of c e m e n t
hydration.
910

Vol. 20, No. 6

REACTION MECHANISMS, ADMIXTURES, VOLUME CHANGE

9t I

Hydration Mechanisms
Cement hydration evolves in a way which includes two phase transitions. [n contact with water, metastable solid cement phases dissolve into an aqueous solution.
This dissociation leads to an increase of CaO concentration up to 20-40 mmol/I in
the first hours of C3S hydration [3]. Reaching a state of supersaturation in the
pore water solution, crystallization into stable and solid CSH-products can follow.
In both phase transitions a sudden volume change occurs, resulting from changes
in bound forces. Figure l shows this phenomenon in a plot according to the Vander-Waals theory.

iitlransition
Area

m
u)

Figure l

a=

Phase Transition ace.

Van-der-Waals-Theory

Volume

I ~solid =J)TAtI--

fluid

-J-I

During the solvationprocess a solid-fluid phase transition leads to an increase in

volume (swelling). Crystallization involves a volume contraction (shrinkage)
which is higher in absolute terms than the previous volume swelling. A shrinkage
volume remains after fluid-solid transition and can be measured volumetrically or
gravimetrically. Both processes interact simultaneously, but with different velocities. Hydration can be seen as a solvation-crystallization step and was first described by Le Chatelier [4].
An other ph.ysico-chemical phenom, enon can be noted during hydration, gamely
an exothermtc heat productton during the reactton. Ftgure 2 shows the Ca z + ion
concentration and the reaction heat rate of a C3S hydration according to Vernet

[5].

HYDRATION

VARIATIONS

OF

ALITE : SEQUENCE

Figure 2

Hydration of C3S,
ace.

[5]

TIME :
PERIOD :

H.

-I

HYDRATES - -

CSH

10 H.

1)'1.
II

--III-. . . . .

--

--

CH

100 H.
I

CSH

. . . . . .

912

P. Paulini

Vol. 20, No. 6

This plot shows s o m e very interesting facts

1) In period [ - the initial p e r i o d of hydration - a high increase of Ca 2~ ion
c o n c e n t r a t i o n is a c c o m p a n i e d by a ~evere drop ot e x o t h e r m i c heat rate.
A c c o r d i n g to the A r r h e n i u s law. a higher t e m p e r a t u r e should a c c e l e r a t e the
reaction. This is obviously not valid during period I. The solvation process
requires an activation energy, which is t a k e n as heat energy from the system,
resulting in an e n d o t h e r m i c process with d e c r e a s i n g heat rates.
2) I n p e r i o ~ t l I - t h e d o r m a n t p e r i o d - t h e s o l v a t i o n .
proceeds., still, producino_oa
gain of C a - + c o n c e n t r a t i o n . During this perxod, the e x o t h e r m i c heat rate (if
p r e s e n t ? ) is low.
3) p e r i o d III - the a c c e l e r a t i o n period - is c h a r a c t e r i z e d bv a sudden drop of
Ca z + ion c o n c e n t r a t i o n . The starting point of period Ill is" d e f i n e d by a state
of s u p e r s a t u r a t i o n in the p o r e w a t e r solution. O n c e nucleation is initiated in
the area of s u p e r s a t u r a t i o n , the s u b s e q u e n t reaction kinetic is d e t e r m i n e d by
a s i m u l t a n e o u s process of solvation - t r a n s p o r t - and crystallization. During
this period the exothermic heat rate r e a c h e s a peak, after which a continuously d e c r e a s i n g reaction rate follows.
C o n c r e t e admixtures are of interest within p e r i o d s I and II. Once the a c c e l e r a t e d
period is reached, no strong effects of admixtures on hydration kinetics can be
e x p e c t e d . The main purpose of c o n c r e t e admixtures lies t h e r e f o r e in influencing
the solvation process of c e m e n t hydration.
T h e r m o d y n a m i c s of H y d r a t i o n
H y d r a u l i c r e a c t i o n s are irreversible p r o c e s s e s evolvin~ from an ener~,etic state of
non.eqmlibrmm towards an e q m h b r i u m . During the r e a c n o n an e n e r ~ transformatron occurs from a high free energy level towards a lower one. This results in
two types of o p p o s i t e directed energies, a r e a c t i o n heat and a v o l u m e work ( b o u n d
energy). In o r d e r to start the hydration process, m e t a s t a b l e crystalline CS-phases
need a certain a m o u n t of activation energy. The r e a c t i o n path may be d e s c r i b e d
either by entropy changes or by free energy changes. Figure .3 shows a schematic
plot of a reaction path which r e q u i r e s an activation energy G ~:

t
o

E
c
i.i

la_

AG

Go
m

ZXG

Time
Figure3

A c t i v a t e d reaction path in free energies

Vol. 20, No. 6

REACTION MECHANISMS, ADMIXTURES.VOLUMECHANGE

913

A c c o r d i n g the 2nd law of t h e r m o d y n a m i c s , e n t r o p y changes must be positive "

dS : dS i + dS e >- 0
By d e c o m p o s i n g e n t r o p y p r o d u c t i o n into an inner term dS i remaining in the
system, and a term dS e, which reacts with the e n v i r o n m e n t , it can be shown [6]:
dS i -> - d S e -> 0
Both e n t r o p y p r o d u c t i o n terms are c o u n t e r a c t i n g terms, w h e r e b y the remaining
term dS i d o m i n a t e s the flux term dS e.
A chemical reaction can equally be d e s c r i b e d in terms of free energies. H e l m hohz free energy and Gibbs free enthalpy must decline continuously during a
chemical reaction.
dF = -19.dV - S.dT _< 0
dG
V.dp S.dT_< 0
The chemical p o t e n t i a l , u is normally d e f i n e d as the d i f f e r e n c e b e t w e e n the starting and end point of Gibbs free enthalpy. We have seen that CS-phases need a
certain activation energy G * in o r d e r to start the reaction. This energy is supplied
as heat energy to the reacting agents and is derived from the mLx c o m p o n e n t s and
the surrounding. The total chemical potential is t h e r e f o r e
]tto t = (G 0 - G e q ) + G *
H y d r a u l i c r e a c t i o n s don't show any global electrical or magnetical effects. We can
t h e r e f o r e m e a s u r e the total energy t r a n s f o r m a t i o n during hydration using a v e c t o r
(p,T,V). We consider an a d i a b a t i c state f u n c t i o n as shown on Figure 4 .
Pe

Figure 4
A d i a b a t i c State F u n c t i o n

P(
VI

v,

V,

In this case we u n d e r s t a n d u n d e r the p r e s s u r e term PB an inner b o u n d pressure.

and not as usual an exterior p r e s s u r e working on the system. The v o l u m e chan~e
b e t w e e n original state 0 and end state 1 r e p r e s e n t s a shrinkage v o l u m e V s resulting f r o m increasing b o u n d stresses.
By following an a d i a b a t i c reversible e q u i l i b r i u m p r o c e s s along the path 0-I one
finds, that reversible v o l u m e w o r k and reversible heat are o p p o s i t e energies of an
e q u a l a b s o l u t e a m o u n t . This b e h a v i o u r of c o u n t e r a c t i n g energies is known as the
Le C h a t e l i e r principle.

914

P. Paulini

~Wrev

= - j

PB

dV

Vol. 20, No. 6

- ~Qrev

= Seq

(Tl

- TO)

Focusing now on an irreversible process taking place b e t w e e n states 0-2-t we see.

that the material u n d e r g o e s not only reversible, but also irreversible volume work.
The total volume work done on the haaterial is
~W

~Wrev

~Wirr

~ PB

Vs

S t r u c t u r e f o r m a t i o n and strength growth of a material d e p e n d only on the irreversible part of volume work.
~Wirr

&PB

Vs

+ ~

PB

dV

S i m u l t a n e o u s to the irreversible work, an o p p o s i t e directed heat of r e a c t i o n / , O R

o c c u r s . Bv applying the 2nd law it can be shown [7], that for e x o t h e r m i c processes.
the react{on heat is smaller in a b s o l u t e terms than irreversible volume work
remaining in the system. The contrary, is true for e n d o t h e r m i c reactions.
T.dS e -< 0 -< T.dS i
exothermicreactions

dQ R

-< 0 <-- dWir r

e n d o t h e r m i c reactions

dWir r <_ 0 _ dQ R

That means, that exothermic reactions are a c c o m p a n i e d by a v o l u m e shrinkage

and increasing bound forces. E n d o t h e r m i c reactions show a v o l u m e swelling and
d e c r e a s i n g bounds. Chemical hardening is t h e r e f o r e always of exothermic nature.
T h e s e t h e r m o d y n a m i c relations hold even for hydration processes. We have seen
that e x o t h e r m i c hydration heat and irreversible v o l u m e work are o p p o s i t e energy
terms. They can be regarded as energies, r e s p o n s i b l e for both terms of c o u n t e r a c t ing e n t r o p y production. Only the part of irreversible volume work remaining in
the system can influence a hydraulic stiffening and hardening process. In this
sense a main interest has to be put on the irreversible volume work occuring
during hydration.

Kinetics of Hvdration
Kinetic m e a s u r e m e n t s of chemical reactions can be p e r f o r m e d based on the above
explanations recording either heat energies or v o l u m e work. Until now mostly
c a l o r i m e t r i c recordings have been used i n t e r p r e t i n g hydraulic reactions.
In o r d e r to d e s c r i b e the irreversible v o l u m e work remaining in the material we
would have to record both, b o u n d stresses and shrinkage volumes during a reaction. A direct m e a s u r e m e n t of intermo-}ecular bound stresses is a rather difficult
task. It is much easier to record volume changes during a reaction and to calculate b o u n d pressures. This can be done using
dPB = - K . d V s / V 0
w h e r e K is the bulk modulus, V s the shrinkage v o l u m e and Vf~ the initial volume.
We have t h e r e f o r e used a setup b a s e d on the i m m e r s i o n weighinz m e t h o d (Fig. 5).
In this way v o l u m e shrinkages during c e m e n t reaction can be effsily and a c c u r a t e
r e c o r d e d [8].

Vol. 20, No. 6

REACTION MECHANISMS, ADMIXTURES, VOLUME CHANGE

915

Figure5

(~ Balance
(~) DataLoggingUnit
(~) TherrnmCmt
(~ Sample

Immersion Weighing Setup

Reaction rates recorded by volume shrinkage show the same characteristics as

recorded bv calorimetric methods. The total shrinkage volume for cement hydration dep,en~ls on the fractions of the hydraulic phases and lies in the range of 0.060,07 cm-3/g cement. All the followin~ recordings have been taken from cement
paste reaciions with a w/c-ratio of 0.-40, and if not otherwise mentioned with a
tsothermal t e m p e r a t u r e of 20 C.
The early phase of cement hydration can present a complete different behaviour
as known~from calorimetric records. Early volume changes of cement hydration
can produce an increase in volume (swelling process) which turns into a shrinkage
process after the dormant period. Figure 6 shows the wide range of early cement
reactions without use of any admixture.
Shrinkage Volume ( c m $ / 9 )
0,02

Shrinkage
0,01

. :

Typ

TypN

-0,01

-0,02

Sw~'~

-0,05

-O,04
0

Figure 6

TypS
i
6

Time(h)
6

I
10

'12

14

Range of early Cement Reactions

Type L cements have a low initial reaction and were found by finer ground cement
types. Type N cements show no clear initial reaction and were found in normal
portland cements and sulhpate resistant cements. Swelling processes of type S
cements were observed especially in slag cements, old cements and fresh ground
cements.
In o r d e r to prove that the time during which swelling occurs can be defined as
d o r m a n t pertod, accelerated cement reactions with different isothermal temperatures are represented in Figure 7. Leaving out the swelling phase, Figure 7 shows

916

P. PafiLi.ni

Vol. 20. No. 6

only volume contractions, starting at the time where chemical shrinkage appears.
We note the well known behaviour of temperature-accelerated cement reactions
which have shorter dormant periods, react faster at the beginning, and reach
lower degrees of hvdration at the end.
0,00

~ m k , ~ l e Volume (cm,3/,~

0.03

0.04
0,03
0.O2
0.01

20

40

Figure7

60
T i m e (h)

80

100

120

Isothermal Cement Reactions

The following plots give an insight into the working mechanisms of concrete
admixtures. Adding accelerator agents leads to an increase of early volume shrinkage. This is illustrated in Figure 8 using 5% accelerators of different brands and
a rather fast reacting CE II S cement. Figure 9 shows the same result with a slowreacting sulphate-resistant portland cement, Both, dry and fluid agents have been
used.

0,01

vca (cm3/g)

0,03

v~e (crib/g)

0,008
0.02 ~
O,OOe

0.01

0,004
Q
0,002
oo

CE !I-

Ace.
-0,02

0,002
0.5

t,3

2,5

3,11

4.5

II

5.1l

T i m e (h)

Figure 8 Cement Reactionw. 5% acc.

Control --- CE II S Cement

i
2

J
4

I
6

L
8

I
10

L
12

~
14

h
le

L
18

Time (h)

F i g u r e 9 Cement reactionw. 5% acc.

Control = SRPC Cement

On the other hand, the use of retarders leads [o an increase of volume swelling.
Figure 10 shows this, using 0.4% retarder with an ordinary CE I cement and a CE
II S slag cement. Increasing the dosage of retarder leads to a higher propagation
of swelling volumes. This is shown in Figure 11 with an ordinary PC 3 5 F c e m e n t
using a retarder based on hydroycarboxilic acid.

,
20

VoI. 20, No. 6

v~

0,03

REACTION MECHANISMS, ADMIXTURES, VOLUME CHANGE

(cm3/;d

...........

0.02

veo

0.06

917

(m3/0)
, , . . - - - ' - " - - - " ~ n~roCo
n t rol

0.04

0.5%
J

0.03

Control CE II S

0.01

1.0%

O.02
O.01

0
k ~ " ~" C
~

II S * 0.4% Ret.

-0.01 [

*0,01

"0.021
0

10

20
Time (h)

30

40

Retarder

PC 3 5 F

-0.02

10

F i g u r e 10 C e m e n t Reaction with
0.4% Retarder

20

Figurell

30
Time (h)

40

OO

so

C e m e n t Reaction with
incr. Retarder Dosage

Portland cement consists of metastable crystal modifications which are influenced

bv many factors, like burning-cooling processes, chemical composition, type of
~'inding and state of prehydration. Metastable modifications are to be found in a
trozen state of high free e n e r ~ and tend to r e o r d e r themselves into stable crystals
even in solid state reactions. ~ h e influence of a concrete admixture on the c e m e n t
reaction can vary widely. The specific action of expected admixture effects should
therefore be proved in every different cement. An example is given in the two
following figures.

0.04

0,06

v=~ (om,,~o)

~ t o ~

0,03
0.02

_
"

Yea

(m3/g|

0.04

2.5%
o

Control
0.5%

0.03

0,01
0

0.02

-0.01

0.01

-0.02

0.04

PC 3 5 F + CaCI 2

CaCJ2

o0,03
I

10

20

30

40

50

60

70

Time (h)

F i g u r e 12 CaCI9 infh, ence on

PC rI s slag c e m e n t

..+.

80

o0.01
0

to

t4

20

30 36
Time (hi

25

40

411 50

54

F i g u r e 13 CaCI9 influence on
PC 35F reaction

Calciumchlorides are known as strong accelerators. Figure 12 shows the influence

of different dosages on the reaction of a slag cement. A low dosage of 0.5% leads
to a stronger solvation and swellin~ process compared to the control reaction.
Using 1% leads to an immediate imtial reaction while higher percentages show
again a short swelling period.

Picture 13 shows the reaction influence of CaCI 2 on a fast reacting PC 35F. Increasing the dosage up to 2.5% leads to declining shrinkage volumes and to a
swelling period. A dosage of 5% shows again a faster reaction.

60

9].8

P. Paulini

Vol. 20. No. 6

Kinetic reactions recorded by the immersion weighing m e t h o d give a clear insight

in the working mechanisms of cement hydration. They are d e t e r m i n e d by a double
phase transition of the solvation-crysrallisation step of cement hydration. During
solvation a solid-fluid transition takes place in which repulsive forces are released
and ions from solid state crystals are converted to an aqueous solution. This
process leads to lower ion bounds and higher distances between ions resulting in a
v o l u m e swelling. After reaching a state of supersaturation in the pore water solution the crystallization step to CSH-phases, including a fluid-solid transition, can
follow. Here attractive ion forces d o m i n a t e the proces~ - ion distances become
smaller - and an total shrinkage volume of 0,06-0,07 cm /g cement remains after
c o m p l e t e hydration. This Le Chatelier T h r o u g h Solution model can be described
by a consecutive reaction
A

kx

>

kz

>

with constant velocities kx and k z. Such a model, applied to c e m e n t reactions, was

p r e s e n t e d in [7].
The addition of concrete admixtures influence these mechanisms in a clear way.
The usage of r e t a r d e r agents increases repulsive forces of the solvation process
and leads to high swelling volumes in the first hours after mixing. As long as repulswe forces prevail over attractive forces of the crystallization process, no strength
gain can occur. The d o r m a n t period is defined by the length of the swelling period.
Superplasticizers accelerate the solution process and lead as well to swelling
volumes during the early reaction. The following CSH-crystallization runs faster
because of higher CS-ion concentrations in the pore water solution. This is a well
known behaviour of many superplasticizing agents. Accelerators show an immediate shrinkage volume just atter mixing. This can be the result of a higher chemical potential, so that just a low activation e n e r ~ is n e e d e d in order to start the
reaction. An other possibility is to increase the number of crystalline nuclei using
stone or .quartz powders. All these mechanisms can be shown with a sensitive
v o l u m e t r i c m e a s u r e m e n t setup.

ILd.c.r..c.ggt~
1
2
3
4
5
6
7
8

R a m a c h a n d r a n V.S., Concrete A d m i x t u r e s H a n d b o o k , Noyes Publ., New

Jersey, 1984
R/xom M.R., Mailvaganam N.P., Chemical Admixtures for Concrete, 2nd ed.,
Spon LTD., L o n d o n , 1986
R I L E M Comm. 68-MMH, Task G r o u p 3, The Hydration of Tricalciumsilicate, Materials and Structures, Vol.17, (1984), pp.457-468
Le Chatelier H., R e c h e r c h e s experimentales sur la constitution des mortiers
hydrauliques, These, Paris, 1887
ernet C., Flux t h e r m i q u e et cinetique d'hydratation, Proc.8th.lnt. Congr.
nemistry of C e m e n t , Rio de Janeiro, 1986, Vol.III, pp.64-69
Glansdorff P., Prigogine I., T h e r m o d y n a m i c Theory of Structure, Stability
and Fluctuations, Wiley Interscience, L o n d o n , 1971
Paulini P., Kinetik der Z e m e n t h y d r a t a t i o n und ihr Einfluss auf die Festigkeit
von Beton im j u n g e n Alter, Dissertation, Universitiit Innsbruck, 1988
.Paulini.P., Kinetic reaction m e a s u r e m e n t s in c e m e n t hydration with the
Immerslon weighing method, Z e m e n t - K a l k - G i p s International, 41.Jg, 10,
(1988), pp.525-531, 12,(1988), pp.297-300