Professional Documents
Culture Documents
Abstract: This work comprises the use of a multi-analytical approach combined with microbiological studies to
characterize six paper samples, containing foxing stains, from the 20th century, regarding their cellulose matrix,
llers, and sizing materials, and to evaluate possible paper degradation that might have occurred during the foxing
stains. Photography under different illuminations and optical microscopy were used for morphological
characterization of the paper samples and foxing stains. Scanning electron microscopy coupled energy dispersive
spectroscopy (SEM-EDS) was of particular importance for dening the presence of ber disorder and disruption
on the surface of some of the stains, and localized accumulations of mineral-like particles on the surface of others.
SEM-EDS, attenuated total reection Fourier transform infrared spectroscopy (ATR-FT-IR), and energy
dispersive X-ray uorescence (EDXRF) were used for the identication of mineral llers, whereas sizing agents
were analyzed using ATR-FT-IR. EDXRF results showed that no differences, within the standard deviation, were
found in iron and copper contents between the foxed and unfoxed areas. Fungi belonging to the genus Penicillium
spp. were found in all the paper samples. Unfoxed areas presented lower contamination than the foxed areas.
Key words: foxing stains, unfoxed area, paper degradation, SEM-EDS, ATR-FT-IR
I NTRODUCTION
Paper has been used for centuries to record important
historical and cultural heritage events all over the world and
is present in thousands of private and public libraries,
museums, and archives (Goltz et al., 2010). Historical paper
differs in many ways from contemporary paper (Manso &
Carvalho, 2009). Before 1840, the raw materials used to
produce paper were cellulose bers obtained from rags and
ropes (Proniewicz et al., 2002). With the increasing scarcity
of rag bers, wood gradually became the main source of
cellulose. In the hand papermaking process, gelatin, a very
stable and impermeable lm-forming external sizing agent,
was used. As a consequence of industrialized machine
papermaking, gelatin was replaced by the acidic rosin/
aluminum internal sizing, which induced the formation of
sulfuric acid, promoting acid hydrolysis in paper (Erhardt &
Tumosa, 2005). In the 1980s this acidic sizing process was
replaced by alkaline neutral sizing process using alkyl ketene
dimer (AKD) (Espy, 1990; Pinzari et al., 2006, 2010; Area &
Cheradame, 2011; Song et al., 2011).
Fillers were used by the end of the 18th century, and
china clay (kaolin) was the most common ller used (Beazley,
1991) because of its good compatibility with the acid sizing
process. It was used until the second half of the 20th century,
being replaced by alkaline calcium carbonate (CaCO3),
Received April 13, 2014; accepted January 6, 2015
*Corresponding author. tasf@uevora.pt
64
Description
P1
P2
P3
P6
P7
P8
Papers were selected by visual observation of the different colors and morphologies of the foxing stains. Based on
these parameters, papers presenting only one typology of
foxing stain were selected. A minimum of ten points on each
paper, both on the stains and on the neighboring unfoxed
paper surface, was subjected to chemical and morphological
analysis. Samples from each paper containing foxing stains
were used in the microbiology studies. Each sample was
properly identied and stored, protected from light, at constant room temperature.
SEM-EDS
ATR-FT-IR
The ATR mode allowed a nondestructive analysis of the paper
samples. Infrared analysis was carried out at room temperature
and ambient humidity with a Brker Alpha spectrometer
(Billerica, Massachusetts) coupled with a single-reection diamond ATR module. All spectra were acquired in the absorbance mode, in the range from 4,000 to 375 cm 1, from a total
of 128 scans at 4 cm 1 resolution. Spectra were recorded and
analyzed using OPUS/Mentor software (version 6.5).
Before sample analysis, background measurement was
performed in order to reduce the effect of atmospheric carbon dioxide and water vapor. Paper samples were positioned
on the surface of the sample holder, in direct contact with the
diamond crystal, and were subjected to some pressure. The
identication of paper components and detection of microbial
presence were carried out by comparing the main features of
the obtained spectra with published data (Derrick et al., 1999;
Proniewicz et al., 2001, 2002; Zotti et al., 2008, 2011; Brandt
et al., 2009; Manente et al., 2012) or with the ATR-FT-IR
database (ATR-FTIR Library COMPLETE, 2009). The absorbance spectra were normalized for comparison purposes.
EDXRF
X-ray uorescence spectrometry is a nondestructive technique, which makes it completely appropriate for the elemental
analysis of items of cultural heritage, paper in particular.
Samples were analyzed directly without any preparation.
Qualitative data from paper artifacts, such as XRF peak
height or intensity, can be compared and used to identify
differences in relative concentrations of media between artifacts
or in areas in the same artifact for samples with similar composition and thickness. XRF analysis of paper is difcult due to
its minimal thickness, limited density, and low concentrations
of the elements of interest (Barrett et al., 2012).
The analyses were carried out with a Brker TRACER IIISD (Billerica, Massachusetts) handheld portable spectrometer
with a rhodium tube. The S1 TRACER was used in a stationary,
bench top conguration, which is more adequate for paper
analysis, with a yellow lter (304.8 m Al + 25.4 m Ti) that
removes low-energy X-rays coming from the tube. This
reduced background levels, and therefore improved the detection limits (Barrett et al., 2012). Use of the yellow lter was
adequate for analysis of metals (Ti to Ag and W to Bi), but there
was little sensitivity to elements below Ca. Operating conditions
of 40 kV, 12.5 A and an acquisition time of 300 s were used for
65
all the analyses. The characteristic radiation emitted by elements present in the samples was collected by a silicon drift
detector that permitted an energy resolution of 149.68 eV for
Mn K radiation at a count rate of 100 kcps. Three analyses in
the foxed areas and in the unfoxed areas were carried out for
each sample. An XRF cup containing water and covered with a
Proleene XRF lm was positioned behind the paper samples
being analyzed, and together these were positioned against the
nose of the instrument in order to supply a consistent light
pressure that kept samples in their nal position to be analyzed
and diminished the distance between the sample and the
detector. A beam spot of (3 4) mm2 was used. The equipment
also had a camera that allowed visualization of the analyzed
area, and captured the image and spot of analysis. S1 PXRF
software (v. 3.8.30) was used to record the spectra, and ARTAX
software (v. 5.3.0.0) was used for spectra evaluation. Net counts
of the elements were also normalized to the Compton scattered
peak by dividing the net counts for each analyzed element,
obtained from spectra deconvolution, by the net counts of the
Compton scattered peak from the same spectrum (obtained by
evaluation of the region of interest).
RESULTS
AND
DISCUSSION
66
Table 2.
Standard Light
Raking Light
UV Radiation
P1
P2
P3
P6
P7
P8
67
6.3
40
Tone
P1
Off-white
toned paper
P2
Cream toned
paper
P3
Cream toned
paper
P6
Cream toned
paper with
brown margins
P7
Off-white
paper
P8
Tanned cream
paper with
oxidized brown
surface middle
Minute foxing
areas that are usually larger than the visibly stained areas,
which seems to be the case on sample P3. This was the only
paper that showed uorescence.
For a more detailed and an amplied view of the
degraded areas by foxing, OM was used (Table 3).
Manso et al. (2009) analyzed 14 Portuguese drawings
from the 19th century and reported that from a
68
Table 4.
Scanning Electron Microscopy Micrographs of Unfoxed and Foxed Areas in Paper Samples P1P8.
Sample
Unfoxed Area
Foxed Area
P1
P2
P3
P6
P7
P8
SEM-EDS Analyses
Analysis by SEM-EDS permitted examination of the morphology of the different paper samples and foxing stains and
69
70
Figure 2. Energy dispersive X-ray spectra of samples (a) P6, (b) P2, (c) P1, and (d) P3 (unfoxed areas).
ATR-FT-IR Study
ATR-FT-IR analysis was carried out for the evaluation of
llers and sizing materials. In addition, it was used to compare unfoxed paper surfaces with foxed areas. It was also a
suitable technique for evaluating the presence of fungi.
Composition of Paper
Figure 4 shows the spectra corresponding to paper samples
P1 to P8 obtained for unfoxed and foxed areas. The most
important peak wave numbers and their interpretation are
shown in Table 5.
Cellulose peaks in all spectra were well recognized, especially in the ngerprint region at the range of 8501,500 cm 1
(Manente et al., 2012). The vibrational pattern of cellulose is
very complex in this range, where stretching and deformation
modes of various groups couple. The range between 900 and
1,200 cm 1 covers the CO and CC stretching, antisymmetric in-phase ring stretching, anti-symmetric bridge
71
EDXRF
EDXRF analysis is a nondestructive technique sensitive to
a wide range of elements. It was, therefore, used to obtain
information on elemental composition of the papers
and foxing stains (Bicchieri et al., 2002; Manso et al., 2011).
72
Figure 4. Attenuated total reection Fourier transform infrared spectra of unfoxed (black line) and foxed areas (gray
line) of paper samples: (a) P1, (b) P2, (c) P3, (d) P6, (e) P7, and (f) P8.
661
592
535
472
661
592
523
467
663
595
535
468
1,105
1,029
877
813
1,456
P6
663
592
520
459
1,447
1,427
1,105
1,029
879
813
1,647
1,559
1,636
3,333
2,895
P7
536
469
1,454
1,420
1,104
1,027
874
1,647
1,651
1,558
1,560
3,321
2,897
P8
808
665
595
522
469
1,508
1,447
1,424
1,106
1,029
876
1,624
1,559
1,669
1,647
3,330
2,896
Tentative Assignment
SiOAl stretching
SiOSi bending
OH stretching
CH stretching
C=O stretching
C=O stretching
OH bending
C=O stretching
C=O stretching
CN stretching
CNH bending
Aryl ring asymmetric stretching
NH bending; COH bending
CO32 stretching
SO42 stretching; COC symmetric stretching
COC asymmetric stretching
OCO bending; skeletal vibration
CO stretching
CH deformation out of plane, aromatic ring
SO42 bending
Interpretation
Rows in gray refer to the foxed area spectra. Other results were obtained from the unfoxed area spectra (Derrick et al., 1999; Bke et al., 2004; Saikia & Parthasarathy, 2010).
ATR-FT-IR, attenuated total reection Fourier transform infrared spectroscopy.
1,447
1,427
1,105
1,032
876
814
1,653
1,558
1,653
1,558
1,456
1,427
1,105
1,029
875
1,618
1,636
1,636
1,646
1,651
1,558
1,558
P3
3,334
2,897
1,720
3,333
2,897
1,716
P2
Peak Wave Numbers (cm 1), Tentative Assignment, and Interpretation of ATR-FT-IR Spectra of Samples P1P8.
3,335
2,899
P1
Table 5.
73
74
Table 6. Net Counts Obtained from the Deconvolution of Experimental Spectra of the Foxed and Unfoxed Areas, Divided by the Net
Counts of the Region of Interest (ROI) of the Compton Scattered Peak, Multiplied by 10,000.
Sample
Name
P1
Unfoxed
Foxed
Unfoxed
Foxed
Unfoxed
Foxed
Unfoxed
Foxed
Unfoxed
Foxed
Unfoxed
Foxed
P2
P3
P6
P7
P8
Ti
Cr
677.8 5.6
676.4 2.2
34.7 0.7
33.6 0.7
34.5 0.6
35.2 0.9
34.4 2.3
37.5 1.4
97.8 0.4
96.6 1.2
36.9 0.7
38.7 0.5
66.4 2.2
67.9 1.6
7.5 0.8
7.0 0.4
5.4 0.1
4.3 1.3
11.1 1.6
10.3 0.6
4.4 0.3
5.5 0.8
6.0 0.6
5.2 0.6
Mn
13.0 0.7
14.8 5.4
7.5 0.5
9.7 0.7
6.2 0.3
5.2 0.8
Fe
Ni
Cu
Zn
Pb
817.0 10.6
809.3 18.4
207.2 2.0
253.2 98.2
39.5 1.1
41.7 1.0
180.6 4.3
184.7 4.4
132.4 1.9
128.6 4.2
124.3 1.9
118.4 4.4
8.4 0.4
5.7 0.5
10.4 0.4
13.3 4.4
1.4 1.1
2.5 0.8
22.9 1.1
24.7 0.8
1.7 0.1
2.4 0.5
2.4 0.2
3.1 0.4
410.4 5.6
392.9 5.8
60.1 0.5
59.2 2.2
44.7 0.2
45.4 0.5
54.8 0.5
56.1 1.9
68.1 0.4
67.6 1.5
65.8 1.8
67.7 2.6
280.2 2.1
271.5 6.3
18.6 0.6
19.1 3.6
8.9 0.1
9.1 0.7
13.0 1.1
17.9 2.3
19.4 0.1
19.8 1.6
17.6 2.3
27.3 3.4
2,682.6 32.2
2,623.6 34.7
75
Microscopic Features of Isolated Fungal Strains from Foxed Areas of Samples P1P8 and their Identication.
Microscopic features
Identication
Samples
Penicillium spp. 1
P1, P2,
P3, P6,
P7, P8
Penicillium spp. 2
Penicillium spp. 3
P1, P2,
P3, P8
Penicillium spp. 4
CONCLUSIONS
This work consisted of the study of six papers from the 20th
century containing foxing stains. The materials used in the
production of papers and evaluation of a possible degradation
process of the substrate as a result of the foxing stains were
evaluated. Inorganic components used as llers were studied
using SEM-EDS, ATR-FT-IR, and EDXRF, whereas sizing
76
ACKNOWLEDGMENTS
The authors acknowledge Vanda Amaral and Nuno Carrio
for image processing. They also acknowledge Ana Margarida
Cardoso, Catarina Miguel, Jos Miro, and Massimo
Beltrame for the suggestions and fruitful discussions.
REFERENCES
ABDEL-MAKSOUD, G. (2011). Analytical techniques used for the
evaluation of a 19th century Quranic manuscript conditions.
Measurement 44, 16061617.
AREA, M.C. & CHERADAME, H. (2011). Paper aging and degradation:
Recent ndings and research methods. Bioresources 6(4), 53075337.
ATR-FTIR LIBRARY COMPLETE (2009). Vol. 2 2009 ST Japan, Inc.
BARRETT, T., ROBERT, S. & WADE, J. (2012). XRF analysis of historical
paper in open books. In Studies in Archaeological Sciences,
Shugar, A.N. & Mass, J.L. (Eds.), pp. 191214. Leuven, Brussels:
Leuven University Press.
77