Subscriber access provided by CNR | Consiglio Nazionale delle Ricerche
Article
The Degradation Process of Lead Chromate in paintings by
Vincent van Gogh studied by means of Spectromicroscopic
methods. Part III: Synthesis, characterization and detection
of different crystal forms of the chrome yellow pigment
Letizia Monico, Koen H. Janssens, Costanza Miliani, Brunetto Giovanni Brunetti, Manuela Vagnini, Frederik
Vanmeert, Gerald Falkenberg, Artem M. Abakumov, Yinggang Lu, He Tian, Johan Verbeeck, Marie Radepont,
Marine Cotte, Ella Hendriks, Muriel Geldof, Luuk Van der Loeff, Johanna Salvant, and Michel Menu
Anal. Chem., Just Accepted Manuscript • Publication Date (Web): 10 Oct 2012
Downloaded from http://pubs.acs.org on November 5, 2012
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.
Page 1 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 2 of 30
The Degradation Process of Lead Chromate in paintings
by Vincent van Gogh studied by means of
Spectromicroscopic methods. Part III: Synthesis,
characterization and detection of different crystal forms of
the chrome yellow pigment
Letizia Monico,a,b Koen Janssens,b,* Costanza Miliani,c Brunetto Giovanni Brunetti,a,c Manuela
Vagnini,d Frederik Vanmeert,b Gerald Falkenberg,e Artem Abakumov,f Yinggang Lu,f He Tian,f
Johan Verbeeck,f Marie Radepont,b,g Marine Cotte,g,h Ella Hendriks,i Muriel Geldof,l Luuk van
der Loeff,m Johanna Salvant,n Michel Menu.n
* Correspondence: Koen Janssens, University of Antwerp, Department of Chemistry,
Groenenborgerlaan 171, B-2020 Antwerp, Belgium, tel. +32 3 265 33 22, fax. +32 3 265 32 33,
koen.janssens@ua.ac.be
a
Centre SMAArt and Dipartimento di Chimica, Università degli Studi di Perugia, via Elce di
Sotto 8, I-06123 Perugia, Italy.
b
University of Antwerp, Department of Chemistry, Groenenborgerlaan 171, B-2020 Antwerp,
Belgium.
c
Istituto CNR di Scienze e Tecnologie Molecolari (CNR-ISTM), c/o Dipartimento di Chimica,
Università degli Studi di Perugia, via Elce di Sotto 8, I-06123 Perugia, Italy.
d
Associazione Laboratorio di Diagnostica per i Beni Culturali, piazza Campello 2, I-06049
Spoleto (Perugia), Italy.
1
ACS Paragon Plus Environment
Page 3 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
e
Hamburger Synchrotronstrahlungslabor HASYLAB at Deutsches Elektronensynchrotron
DESY, Notkestr. 85, D-22603 Hamburg, Germany.
f
University of Antwerp, Department of Physics (EMAT), Groenenborgerlaan 171, B-2020
Antwerp, Belgium.
g
Laboratoire d’Archéologie Moléculaire et Structurale, CNRS UMR8220, 3, rue Galilée, F-
94200 Ivry-Sur-Seine, France.
h
European Synchrotron Radiation Facility (ESRF), 6, rue Jules Horowitz, F-38000 Grenoble,
France.
i
Van Gogh Museum, Paulus Potterstraat 7, 1070 AJ Amsterdam, The Netherlands.
l
The Netherlands Cultural Heritage Agency (RCE), Movable Heritage Knowledge Sector,
Hobbemastraat 22, 1071 ZC Amsterdam, The Netherlands.
m
Conservation Department, Kröller-Müller Museum, Houtkampweg 6, NL-6731AW Otterlo,
The Netherlands.
n
Centre de Recherche et de Restauration des Musées de France (C2RMF), Palais du Louvre,
Porte
des
Lions,
14
Quai
François
Mitterrand,
F-75001
Paris,
France.
2
ACS Paragon Plus Environment
Page 4 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ABSTRACT
The painter Vincent van Gogh and some of his contemporaries frequently made use of the
pigment chrome yellow that is known to show a tendency towards darkening. This pigment may
correspond to various chemical compounds such as PbCrO4 and PbCr1-xSxO4, that each may be
present in various crystallographic forms, with different tendencies towards degradation.
Investigations by X-ray diffraction (XRD), mid-Fourier Transform infrared (FTIR) and Raman
instruments (bench-top and portable), synchrotron radiation-based micro (SR µ-) XRD and Xray absorption near edge structure (XANES) facilities, performed on oil paint models prepared
with in-house synthesized PbCrO4 and PbCr1-xSxO4, permitted us to characterize the
spectroscopic features of the various forms. On the basis of these results, an extended study has
been carried out on historic paint tubes and on embedded paint micro-samples taken from
yellow-orange/pale yellow areas of 12 Van Gogh paintings, demonstrating that Van Gogh
effectively made use of different chrome yellow types. This conclusion was also confirmed by in
situ mid-FTIR investigations on Van Gogh’s Portrait of Gauguin (Van Gogh Museum,
Amsterdam).
KEYWORDS: lead chromate, chrome yellow, co-precipitates, sulfates, paintings, Van Gogh,
pigments, spectroscopy, FTIR, Raman, XRD, non-invasive.
3
ACS Paragon Plus Environment
Page 5 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
1. INTRODUCTION
Historical and more recent documentations report that chrome yellow pigments were widely
used by Van Gogh1 and his contemporaries.2,3,4 They are characterized by different composition
[PbCrO4, PbCr1-xSxO4, (1-x)PbCrO4∙xPbO] and crystallographic forms and have been
extensively studied recently because of their limited stability (darkening) under the influence of
light and other environmental factors.3,4
Chrome yellows available to artists are nowadays known as Primrose Chrome (PbCr1-xSxO4 ,
0.45≤x≤0.55), Lemon Chrome (PbCr1-xSxO4, 0.2≤x≤0.4) and Middle Chrome (mainly pure
PbCrO4). The former two varieties show pale, greenish-yellow shades, while the latter has a
reddish-yellow hue.5,6 Van Gogh, in several paint orders of 1888-18907 already mentioned the
use of chrome yellow type 1, 2 and 3, corresponding to the “lemon”, “yellow” and “orange”
shades.
From the crystallographic point of view, PbCrO4 and PbSO4 show monoclinic8,9 and
orthorhombic10 structures, respectively. In the solid solution PbCr1-xSxO4, when x exceeds 0.4, a
change from a monoclinic to an orthorhombic structure is observed.11,12,13 Under specific
experimental conditions, the less stable orthorhombic PbCrO4 can be also synthesized.14,15
Our previous investigations on both artificially aged model samples of commercial chrome
yellow pigments3 and two paint micro-samples taken from paintings by Van Gogh4 demonstrated
that the chrome yellow alteration is caused by the reduction of original Cr(VI) to Cr(III), as
highlighted by a spatial correlation between the brown alteration and the local Cr(III)concentration.
4
ACS Paragon Plus Environment
Page 6 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Moreover, we found that only the sulfate-rich historic paint model featured a significant
darkening after photochemical ageing, while in original samples the alteration [Cr(III)-species]
was especially encountered in areas rich in S, Ba and/or Al/Si.
On the basis of these results,3,4 the available literature about the UV ageing9 and synthesis16 of
chrome yellow pigments based on 19th Century recipes and the analysis of chromate materials
taken from historic paintings and paint tubes,17 this work is aimed to extend and deepen the
insights on sulfates may influence the stability of this class of pigments.
For this purpose we have synthesized and characterized different crystal forms of PbCrO4 and
PbCr1-xSxO4 (0.1≤x≤0.75) by employing XRD, mid-FTIR, Raman as well as SR µ-XRD and S
K-edge µ-XANES. Complementary information was collected by scanning transmission electron
microscopy (STEM) equipped energy dispersive X-ray (EDX) spectrometry. An extended study
has been then carried out on pigments from historic paint tubes and on micro-samples from
paintings by Van Gogh and contemporaries. The results demonstrate that it is possible to
distinguish among the various orthorhombic and monoclinic forms of PbCrO4 or PbCr1-xSxO4,
and that Van Gogh and contemporaries effectively made use of these different chrome yellow
pigments.
In the following part IV,18 by the artificially ageing of the aforementioned chrome yellow-based
model paints, we demonstrate that different forms of chrome yellows show a strongly different
tendency towards darkening. To be able to distinguish among these different varieties is
therefore very relevant, since it may open the possibility to investigate whether there is an
effective correlation between the chrome yellow composition/crystalline structure and its
preservation state in original paintings.
.
5
ACS Paragon Plus Environment
Page 7 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
2. EXPERIMENTAL SECTION
2.1 Synthesis of PbCrO4 and PbCr1-xSxO4 and preparation of paint models
The synthesis of monoclinic PbCrO4 powder (S*1mono;* denotes pure inorganic powder without
organic binder added) and of several PbCr1-xSxO4 solid solutions with increasing x values, (S*3A,
S*3B, S*3C, S*3D) was performed following Crane et al.,13 while the preparation of orthorhombic
PbCrO4 (S*1ortho) was accomplished according to Xiang et al.15 (SI for further details about the
synthesis). Table I shows the properties of the synthesized lead chromate-based compounds.
Paint models (S1- S3D) were prepared by mixing the powders with linseed oil in a weight ratio
4:1 and applying the mixture on polycarbonate microscopy slides. Employing the same
procedure, two further paints were prepared: the first, hereby indicated as D1, by mixing oil and a
commercial pigment (CIBA, BASF), while the other, D2, was made of oil, PbCrO4 and PbSO4
powders (both Aldrich), mixed in a 1:2 molar ratio.
6
ACS Paragon Plus Environment
Page 8 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table I. Composition of in-house synthesized/commercial lead chromate-based powders and historic chrome yellow
paint A. Starting CrO42-:SO42- molar ratio employed for synthesizing the powders/samples’ description and quantitative
powder XRD results obtained via Rietveld analysis.
Bench-top powder XRD(a)
Sample(b)
Starting
CrO42-:SO42- molar ratio/
description
S*1mono
1:0
S*3A
0.9:0.1
S*3B
0.75:0.25
S*3C(c)
0.5:0.5
S*3D(c)
0.25:0.75
D*1
Commercial PbCr1-xSxO4
(CIBA, BASF).
(c)
A
Phases
Space group
Mass fraction
(%)
PbCrO4
PbCrO4
monoclinic P21/n
orthorhombic Pnma
98.82(5)
1.18(6)
PbCr0.89S0.11O4 monoclinic P21/n
PbCr0.76S0.24O4 monoclinic P21/n
PbCr0.54S0.46O4 monoclinic P21/n
PbCr0.09S0.91O4 orthorhombic Pnma
PbCrO4
orthorhombic Pnma
PbCr0.6S0.4O4 monoclinic P21/n
PbCr0.1S0.9O4 orthorhombic Pnma
PbCrO4
orthorhombic Pnma
PbCr0.52S0.48O4 monoclinic P21/n
PbSO4
100
100
60.0(2)
31.1(1)
9.10(8)
11.5(3)
74.7(3)
13.8(2)
75.0(1)
orthorhombic Pnma
25.0(1)
historic oil paint tube belonging PbCr0.8S0.2O4 orthorhombic Pnma
to the Flemish Fauvist Rik
PbCr0.1S0.9O4 orthorhombic Pnma
Wouters (1882-1913).
PbCr0.6S0.4O4 monoclinic P21/n
41.0(1)
58.0(1)
1.0 (1.0)
(a)
For cell parameters see SI, Table S-1.
(b)
The XRD pattern of S*1ortho (SI, Fig. S-1) resembles that reported in literature.13,15
(c)
Sulfur and chromium abundances were calculated as weighted average for the mass fraction of
orthorhombic and monoclinic phases. The chemical composition for S3C and A was estimated as PbCr0.4
S0.6O4; for S3D this was PbCr0.2S0.8O4.
7
ACS Paragon Plus Environment
Page 9 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
2.2 Original samples
Historic oil paint tubes. In addition to the already investigated historic chrome yellow paints
A, B1 and B2 belonging to late 19th century artists,3 two other paint samples (denoted below as
DG1 and DG2), supplied by the Musée d’Orsay (M’O, Paris, FR), were taken from two oil paint
tubes originally belonging to the Dr. Paul Gachet collection, assumed to have been used by Van
Gogh.19 Some of their properties are reported in Table II and in the SI.
Embedded micro-paint samples. Table II shows some properties and the results obtained from
the investigation of fifteen micro-paint samples from chromium-based yellow areas of twelve
paintings by Van Gogh and one by Gauguin and two other samples from the palettes of Cézanne
and Van Gogh, that were supplied by the Kröller-Müller Museum (KMM, Otterlo, NL), the M’O
and the Van Gogh Museum (VGM, Amsterdam, NL). For sake of brevity, in this paper only the
results acquired from micro-samples of the following five Van Gogh paintings [Still life with
grapes (1887, VGM, F603/3), View of Arles with Irises (1888, VGM, F409/1), Falling leaves
(“Les Alyscamps”) (1888, KMM, 224/1), The Dance Hall at Arles (“Ball in Arles”) (1888,
M’O, 10872), Roots and tree trunks (1890, VGM, F816/3)] are discussed in detail. Results
collected from a sample taken from Van Gogh’s palette that he used in Auvers-sur-Oise (1890,
M’O, 10455)19 are also presented.
Non-invasive in situ investigations. In addition to studies on paint samples, reflection midFTIR measurements20 were directly performed on Van Gogh’s Portrait of Gauguin [VGM, F546
s 257 v/1962, painted on December 1888].
8
ACS Paragon Plus Environment
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 10 of 30
2.3 Analytical methods
The following methods were used to investigate both paint models and original samples: bench
top, portable and SR µ-XRD, high angle annular dark field (HAADF) STEM-EDX; S K-edge SR
µ-XANES; bench top and portable mid-FTIR; bench top and portable Raman.
Details about the instruments and the experimental conditions are described in the SI.
3. RESULTS AND DISCUSSION
3.1 Characterization of paint models
XRD and HAADF/STEM-EDX.
A combination of XRD and HAADF/STEM-EDX mapping were used to determine the
morphology, the S and Cr local distributions and the phase composition of in-house synthesized
and commercial PbCr1-xSxO4.
Diffractograms (Fig. 1A, top; SI, Fig. S-1) of powders recorded by the bench-top equipment
(black color) are similar to those collected from the corresponding paint models at the PETRAIII SR-facility (blue color). Rietveld refinement (Table I) indicate that powders S*1mono, S*3A and
S*3B are composed of a single monoclinic phase (a minor amount of orthorhombic PbCrO4 is
present in S*1mono). As in Crane et al.13 the lattice parameters (SI, Table S-1) of each PbCr1-xSxO4
phase decrease with increasing S-content. This is observable on the recorded patterns (Fig. 1A,
bottom) by a progressive shift of the diffraction peaks towards higher Q values [see e.g. (111)
and (020) peaks of S*1mono-S*3D monoclinic phase].
While the fraction of monoclinic PbCr1-xSxO4 decreases to ca. 60 wt.% in S*3C and 11.5 wt.%
in S*3D, the orthorhombic equivalents (such as PbCr0.1S0.9O4) become more prevalent (ca. 30
wt.% in S*3C and ca. 75% in S*3D), A contribution of 9-14 wt.% of orthorhombic PbCrO4 is also
present in these samples. In powder D*1, although revealing an elemental composition similar to
9
ACS Paragon Plus Environment
Page 11 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
that of S*3C, monoclinic PbCr1-xSxO4 is the main constituent; the presence of some PbSO4 was
also observed.
The combined use of HAADF-STEM (SI, Fig. S-2) and STEM-EDX investigation of S3B and
S3D (Fig. 1B; SI, Fig. S-2C for S3C results) revealed the presence of nano-crystals of two different
shapes and elemental composition: Cr-rich elongated rods of variable size (ca. 200-500 nm) and
S-rich globular particles (ca. 50 nm diameter). Consistent with the average XRD results (Table
I), literature data15 and STEM-EDX quantitative analysis (Fig. 1B), the rods correspond to
monoclinic phases and the globular particles to orthorhombic ones.
pH measurements of 10 mL of water equilibrated with 1-2 mg of the self-synthesized powders
showed that orthorhombic PbCrO4 (S*1ortho) and the monoclinic co-precipitates (S*3A, S*3B)
yielded a slightly acidic pH value (5.7±0.1), while the PbCr1-xSxO4 materials composed of
monoclinic and orthorhombic forms (S*3C and S*3D) and PbSO4 yielded a significantly lower pH
(4.5±0.1). The monoclinic PbCrO4 (S*1mono) featured a slightly higher pH value (6.1±0.1). The
pH value obtained for S*1ortho is consistent with Crane et al.13 and can be indirectly related to the
higher solubility of orthorhombic PbCrO4 (Kps=10-10.71, ΔfG0orthorhombic~813.2 kJ/mol) compared
to that of the thermodynamically more stable monoclinic form (Kps=10-12.60, ΔfG0monoclinic~824
kJ/mol). Similar conclusions can be drawn for the solubility of orthorhombic PbCr1-xSxO4
relative to their monoclinic equivalents.
10
ACS Paragon Plus Environment
Page 12 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
S K-edge µ-XANES.
S-K edge µ-XANES spectra of PbCr1-xSxO4 paint models (Fig. 1C) are generally similar to that
of the PbSO4 reference compound, featuring a prominent peak at around 2.482 keV, specific for
sulfate-species.21,22 However, there are subtle differences when the Cr-content decreases in
PbCr1-xSxO4: an additional pre-edge feature gradually appears on the left side of the S(VI)-peak
around 2.481 keV, while several post-edge features become more clearly defined. Figueiredo et
al.23 observed analogous features in various Fe(II) and Fe(III) sulfate minerals that were related
to differences in the symmetry and nature of the S-binding site. The Fig. 1C spectra are
consistent with the view that while in PbSO4, all sulfate oxygen atoms are directly bound to Pbatoms, giving rise to a simple S(VI) XANES pattern, in the more Cr-rich co-precipitate
materials, the sulfate groups are more isolated, implying the disappearance of the pre- and postedge features.
11
ACS Paragon Plus Environment
Page 13 of 30
(B)
Intensity
P1
S
Cr
S3D
P2
100 nm
x3
S1mono
P1: Cr/S = 0; P2: Cr/S = (3±2)
12.0 13.5 15.0 17.5
-1
20.0
Q (nm )
P1
orthorhombic PbCrO4
monoclinic PbCrO4
PbSO4
monoclinic PbCr1-xSxO4
orthorhombic PbCr1-xSxO4
(111)
P2
P3
(201)
S1ortho
(111)
S3D
S3C
S3B
S3A
(A)
(C)
Normalized fluorescence
D1
S1mono
200 nm
P1: Cr/S = (0.34±0.07);
P2:Cr/S =(1.1±0.2); P3:Cr/S =0
(111)
(120)
20)
Intensity
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
2.481
S 3A
S 3B
S 3c
S 3D
PbSO 4
(111) (020)
16.20
16.74 -1 17.28
.28
Q (nm )
2.48 2.49 2.50
Energy (KeV)
Figure 1. (A) (top) XRD patternss of S1mono (monoclinic PbCrO4) and S3D
(PbCr0.2S8O4) obtained by (blue) SR (PETRA-III), (black) bench-top and (red)
portable instrumentation; (bottom) detail
de
of SR µ-XRD patterns of PbCrO4
(S1mono, S1ortho) and PbCr1-xSxO4 (S3A-S
S3D, D1). (B) Composite S/Cr EDX map of
(top) S3B and (bottom) S3D. Labels “Pi” indicate the positions where quantitative
analyses were performed. (C) S K-edg
dge XANES spectra of in-house synthesized
PbCr1-xSxO4 and commercial PbSO4 paint
pa models. [SI, Figs. S-1/S-2, for details
on (A) and (B)].
12
ACS Paragon Plus Environment
Page 14 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Infrared spectroscopy.
(a) Transmission mid-FTIR.
Spectra collected from paint models (Fig. 2A) feature some changes when the SO42-content
increases going from S3A to S3D; additionally, other differences are visible in the spectra of the
two polymorph forms of PbCrO4 (S1mono, S1ortho).
In the ν3 sulfate asymmetric stretching region, the monoclinic S3A and S3B show a weak signal
around 1102 cm-1; this band becomes stronger and moves towards higher wavenumbers (1115
cm-1) in S3C and S3D where orthorhombic phases are present. When going from S3A to S3D, two
additional signals around 1165 and 1047 cm-1 [ν3(SO42-)] become progressively visible, as well
as the IR-forbidden sulfate symmetric stretching mode (ν1 at 966 cm-1).
The ν4 sulfate asymmetric bending region is characterized by the presence of two signals at
627 and 597 cm-1. With increasing sulfate amount, a shift of the former band to 620-618 cm-1 is
observed, while also an increase of the relative intensity of the band at 597 cm-1 takes place.
In the ν3 chromate asymmetric stretching region (930-800 cm-1) both a band broadening and a
shift towards higher wavenumbers is detected, when the amount of chromate decreases.
Consistent with the XRD data (Table I) and studies on other MCr1-xSxO4 (M=Ba, Ca, Sr, Pb and
Na),24,25,26 this is justified by changes in the crystalline structure.
Two bands at 852 and 832 cm-1 are present in the spectra of S1mono and S3A; their position shifts
towards highest energies for S3B (855, 832 cm-1) and S3C (859, 837 cm-1). For these two latter
samples an additional signal around 885 cm-1 is visible. The spectrum of S3D features a broad
band around 865 cm-1, that shifts up to 905 cm-1 for S1ortho. Consistent with the XRD results
(Table I), the spectral features of D1 resemble those of S3C and D2 (the latter showing band
features of both monoclinic PbCrO4 and of orthorhombic PbSO4).
13
ACS Paragon Plus Environment
Page 15 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
(b) Reflection mid-FTIR.
In order to establish whether or not it is possible to detect the differences between the various
chrome yellow forms in a non-invasive manner on paintings, all paint models were examined
also by reflection mid-FTIR. Comparison of Fig. 2B to 2A illustrates the presence of spectral
distortions. These anomalies depend on the band strength, the concentration and particle size
distribution as well as on the set-up geometry.27 The strong ν3(SO42-) and ν3(CrO42-) absorption
bands are inverted by the reststrahlen effect, with minima at about 1030 cm−1 and 822 cm−1,
respectively, while the weaker ν4(SO42-) bands show a derivative-shape. Consistent with the
transmission mid-FTIR results, the relative intensity of the two ν4(SO42-) signals changes and the
splitting of the ν3(CrO42-) (860-800 cm-1) disappears when the sulfate amount increases.
14
ACS Paragon Plus Environment
Page 16 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Raman spectroscopy.
Fig. 2C shows Raman spectra of the model paints obtained by the bench-top device (black lines).
Details about the mathematical treatment of spectra are reported in Fig. S-3 (SI). (Data obtained
from D1 and D2 gives results similar to those of S3C and S1mono therefore are not shown).
When the sulfate amount increases, the wavenumber of the ν1(CrO42-) stretching mode28,29
monotonically increases from 841 cm-1 for PbCrO4 (S1mono, S1ortho) to 844 cm-1 for S3C and S3D
(Figs. 2C, S-3A). The full width at half maximum (FWHM) of this component also increases.
These observations can be explained by the lattice compression effect and the prevention of
intermolecular coupling occurring when sulfate replace chromate anions inside the structure.24,25
With increasing sulfate amount, a progressive shift of the position towards highest energy,
increase of the FWHM and a change of the relative intensities are observed also for the
components describing the chromate bending modes (400-280 cm-1) (Fig. S-3B). Consistent with
the literature,28,29 for S1mono, the ν4(CrO42-) modes are located at 400, 376 and 357 cm-1, while
those at 336 and 323 cm-1 are attributable to the ν2(CrO42-) vibration. The band between 339-342
cm-1 is characteristic for the presence of orthorhombic compounds, becoming clearly visible for
S3D and S1ortho (Figs. 2C, S-3B). Similar to the chromate vibrations, the ν1(SO42-) position shifts
from 971 cm-1 (S3A) to 980 cm-1 (D2, the pure PbSO4) (Fig. 2C).
As done for FTIR spectroscopy, equivalent data were collected by a portable Raman
spectrometer suitable for in situ analyses. Fig. 2C (red lines) illustrates that, despite the lower
instrumental spectral resolution, analogous systematic differences can be observed as a function
of the CrO42-:SO42- ratio.
15
ACS Paragon Plus Environment
Page 17 of 30
885
859
837
2-
2(CrO4 )
976
S3C
S3D
x15
973
S3C
843
S3B
S3C
S3B
852
S3B
S3A
971
1102
855
832
2-
340
342
4(CrO4 )
Intensity
A'=Log(1/R)
618
2(CrO4 )
2-
1(CrO4 )
S3D
627
885
859
837
S3D
844
x3
D1
865
D1
841
S1ortho
D2
1115
1047
1102
D2
2-
2-
2-
626
597
852
832
SO4 )
4(CrO4 )
938
SO4 )
979
S1ortho
2-
1165
1117
1058
966
SO4 )
(C)
2-
S1ortho
2-
CrO4 )
450 (SO2-)
4
438 2
(B)
626 SO2-)
4
597
2-
905
875
CrO4 )
(A)
Absorbance
1(SO4 )
S3A
841
S3A
S1mono
S1mono
1050 875 700 420
S1mono
1400 1200 1000 800 600
357
336 323
2-
400
376
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
350
-1
Raman shift (cm )
1400 1200 1000 800 600
-1
wavenumber (cm )
-1
wavenumber (cm )
Figure 2. (A) Transmission and (B) reflection mid-FTIR spectra of paint models of PbCrO4 (S1mono, S1ortho), PbCr1-xSxO4 (S3A-S3D, D1)
and mixture (D2) of monoclinic PbCrO4 and PbSO4 (1:2 molar ratio). Spectra collected by bench-top and portable instrument are
illustrated in (A) and (B) respectively. (C) Raman spectra of paint models S1mono, S1ortho and S3A-S3D acquired by (black) bench-top and
(red) portable device [SI, Fig. S-3 for details on (C)].
16
ACS Paragon Plus Environment
Page 18 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3.2 Identification of different form of chrome yellow pigments in historic paint tube
samples and micro-samples of original paintings
Table II reports a list of twenty-two paint samples obtained from a series of twelve paintings
by Van Gogh, one Gauguin’s painting, the palette by Van Gogh and Cézanne and five historic oil
paint tubes. In twenty cases, XRD, reflection mid-FTIR and/or Raman spectroscopy provided
appropriate information to characterize the type of chrome yellow. Among these, six showed the
presence of monoclinic PbCrO4, while in the others chrome yellow was found to be present in
co-precipitated form. In this latter group, six of the twelve samples showed features most similar
to those of S3B (monoclinic PbCr1-xSxO4, x~0.25), while other six revealed characteristics close
to those of S3C (mixture of monoclinic and orthorhombic PbCr1-xSxO4, x~0.50). In one sample
(F485/4) two chrome yellow forms (monoclinic PbCrO4 and PbCr1-xSxO4) were found, while in
another one (sample A) only orthorhombic PbCr1-xSxO4 was identified. For two other samples
(F469/2, F383/4) neither Raman nor mid-FTIR spectroscopy allowed us to identify the exact
nature of the chrome yellow present. The detection of the Raman signal at ca. 841-845 cm-1
(chromate stretching mode) indicates the presence of a generic lead chromate-based compound.
For F469/2, an orthorhombic lead chromate-based compound was identified by SR µ -XRD,
while no information was obtained for F383/4.
Consistent with the composition, the chrome yellow hue qualitatively range from the orangeyellow for those sample containing PbCrO4 and S-poor PbCr1-xSxO4 compounds to the paleyellow for those materials made of a S-rich PbCr1-xSxO4. Orange and yellow-greenish shades are
conveyed to those samples in which chrome orange (DG2, B2, F482/7) and zinc yellow (F469/2)
were identified.
17
ACS Paragon Plus Environment
Page 19 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
Table II. List of the original embedded paint micro-samples and historic chrome yellow pigments investigated: origin, sample number/name,
composition, hue and analytical techniques that allowed among different chrome yellow forms to be distinguished.
Origin of paint sample
historic oil paint tube belonging
to the Dr. Gachet collection (ca.
1890) (Berthaut, Paris) (M’O)
Rik Wouters’s oil paint tube
(1882-1913) (Mommen & Cie,
Brussels) (Fine Arts Royal
Museum, Antwerp )
historic oil paint tube (end 19th
C) (Elsens, Brussels) (Fine Arts
Royal Academy, Antwerp)
Banks of the Seine, 1887 (VGM)
Still life with grapes, 1887
(VGM)
Sunflowers gone to seed, 1887
(VGM)
Self-portrait, 1887 (VGM)
White grapes, apples, pears,
lemons and an orange, 18871888 (VGM)
View of Arles with Irises, 1888
(VGM)
Sample’s
number/name
DG1 -
Chrome yellow composition
PbCrO4
PbCr1-xSxO4
monoclinic
DG2(c)
(a)
-
-
Techniques for
detecting the chrome
yellow type
XRD,(b) Raman, FTIR
(b)
(109C)
monoclinic
XRD, Raman, FTIR
(130C)
A
orthorhombic
XRD, Raman. FTIR
(3945C)
B1.
monoclinic
XRD,(b) Raman, FTIR
(108C)
B2(c).
monoclinic
F293/3
monoclinic
F603/3
F377/2
F469/2
monoclinic
(b)
XRD, Raman, FTIR
XRD,
XRD,
monoclinic
(e)
orthorhombic
F383/4(e)
(d)
(d)
(109C)
Raman,FTIR
(108C)
XRD,
difference not detectable
(130C)
Raman
Raman
(d)
Raman
(109C)
(387C)
Raman
(3945C)
F409/1
monoclinic
XRD,(f) Raman, FTIR
(109C)
F482/7(c)
monoclinic
XRD,(d) Raman ,FTIR
(1375C)
The bedroom, 1888, (VGM)
The Dance Hall at Arles (“Ball
in Arles”),1888 (M’O)
Falling leaves (“Les
Alyscamps”), 1888 (KMM)
Portrait of Gauguin, 1888
(VGM)
PANTONE
Hue
(d)
F482/8
monoclinic
XRD,
Raman ,FTIR
(3945C)
10872
monoclinic
XRD,(f) Raman, FTIR
(109C)
XRD,(d) Raman, FTIR
(3945C)
XRD,(d) Raman
(3945C)
monoclinic and
possible orthorhombic
monoclinic and
possible orthorhombic
monoclinic and
possible orthorhombic
224/1 (F486)
X448_2
(F456)
X484_3
(F458)
XRD,(d) Raman
(393C)
monoclinic
XRD,(d) Raman, FTIR
(109C)/
(3945C)
F816/3
monoclinic and
possible orthorhombic
XRD,(d) Raman, FTIR
(108C)
2751
monoclinic and
possible orthorhombic
XRD,(f) Raman, FITR
(3945C)
Van Gogh’s palette, 1890 (M’O)
10455
monoclinic
XRD,(f) Raman, FTIR
(116C)/
(393C)
Cézanne’s palette (M’O)
10426
monoclinic and
possible orthorhombic
XRD,(f) Raman, FTIR
(108C)
Vase with sunflowers, 1889
(VGM)
Roots and tree trunks,1890
(VGM)
Be mysterious (“Soyez
mystérieuses”), P. Gauguin,
1890 (M’O)
F458/4
monoclinic
(a)
“monoclinic”: PbCr1-xSxO4 more similar to the reference S3B; “monoclinic and possible orthorhombic”: PbCr1-xSxO4 more similar to the
reference S3C (see text for details).
(
(b)
XRD performed by employing only the portable instrumentation. An indirect semi-quantitative estimation of the S amount was also performed
by SEM-EDX (Fig. S-4).
(c)
Mixture of chrome yellow and orange [phoenicochroite – (1-x)PbCrO4∙xPbO].
(d)
SR µ -XRD, DESY/PETRA III–beamline P06 (Hamburg, GE).
(e)
Mixture of chrome and zinc yellow.
(f)
SR µ -XRD, DESY/DORIS III-beamline L (Hamburg, GE)
18
ACS Paragon Plus Environment
Page 20 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Historical chrome yellow paints.
In order to characterize the historic chrome yellow paints A, B1, B2, DG1 and DG2, first the
sulfur amount was semi-quantitatively determined via SEM-EDX (SI, Fig. S-4), then XRD, midFTIR and Raman spectrometry were used to obtain information on the compounds present.
As Fig. S-4 illustrates, in sample A ca. 622% of the anions are sulfate. Paints B1, B2 and DG2
contain less (around 33-35%), while DG1 does not contain any measurable amount of sulfate.
The high S-abundance in sample A is consistent with quantitative XRD data (Table I), that reveal
the presence of two orthorhombic phases: PbCr0.8S0.2O4 and PbCr0.1S0.9O4. A small amount of
monoclinic PbCr1-xSxO4 is also likely to be present. For this paint, Fig. 3A illustrates that
comparable XRD patterns were obtained by SR-based, bench-top and, despite the lower spectral
resolution, portable instrumentation.
HAADF/STEM-EDX and S K-edge XANES analysis of paint A (SI, Fig. S-5A/B) confirm the
presence of the orthorhombic co-precipitate as in XRD.
Transmission mid-FTIR (Fig. 3B) and Raman spectra (Fig. 3C) of sample A show features that
are mostly similar to the compounds having an orthorhombic phase (S3D/D2, S1ortho), while the
spectra of the other (less S-rich) paint samples B1, B2 and DG2 resemble that of monoclinic coprecipitate S3B; for DG1 the vibrational spectral features are analogous to that of monoclinic
PbCrO4 (cfr. Table I and Figs. 2A, 2C). These observations are confirmed by XRD (Fig. 3A).
In the FTIR spectra of B1 and B2, the ν3(CrO42-) and ν4(SO42-) band-shapes show differences
due to the presence of MgCO3, that contributes with signals around 800 and 597 cm-1,3 while for
DG2 and B2 (dark chrome yellow), additional Raman bands at 826, 376, 340 and 323 cm-1 are
ascribable to phoenicochroite, a compound also identified by XRD.
19
ACS Paragon Plus Environment
Page 21 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
The above-mentioned results point to the fact that the historic paint A, a material that proved
itself to be the very susceptible to darkening due to UVA-Visible irradiation,3 contains
orthorhombic PbCr1-xSxO4 phases, while the materials that proved to be significantly less prone
to darkening (B1 and B2)3 contain the monoclinic ones.
20
ACS Paragon Plus Environment
DG2
S1mono
15
342
DG1
DG1
18
21
24
27
-1
Q (nm )
30
33
322
359
357
400
376
DG1
1160
337
323
357
monoclinic PbCrO4
DG2
841
S3B
(A)
411
386
358
826
P P
P
832
P
B2
B1
854
P
DG2
B1
833
B1
Intensity
B2
405
376
357
340
323
x15
826
843
A
973
1163
1102
1054
P Px3
P
P
Absorbance
B2
A
834
855
monoclinic PbCr1-x-SxO4
P
844
939
x3
977
627 618
597
Page 22 of 30
966
A
870
orthorhombic PbCr1-x-SxO4
Intensity
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1163
1115
1060
Analytical Chemistry
870
580
-1
(B) wavenumber (cm )
1075 860 645 425
340
-1
(C) Raman shift (cm )
Figure 3. (A) XRD, (B) transmission mid-FTIR and (C) Raman spectral data of the historic chrome yellow paints DG1, DG2, B1, B2 and A.
In (A) the XRD pattern of sample A collected by the bench-top (black), portable (red) and SR-based (blue) device are shown. In grey,
spectra of reference S1mono and S3B; “P” labels indicate the peaks of phoenicochroite.
21
ACS Paragon Plus Environment
Page 23 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
Embedded paint micro-samples and non-invasive in situ investigations.
Investigations of the yellow areas of embedded paint micro-samples F409/1, 10872, F603/3,
10455, 224/1 and F816/3 (Fig. 4A) clearly demonstrate the presence of different chrome yellow
forms (see also, Table 2). FTIR (Fig. 4B) and Raman (Fig. 4C) spectra, compared to those of the
reference paints S1mono, S3B and S3C show that the yellow-orange paint layer of F409/1 and 10872
is composed of monoclinic PbCrO4, while the lighter-yellow regions of the other ones are
formed by PbCr1-xSxO4. More in detail it appears that for F603/3 and 10455 the PbCr1-xSxO4
composition is close to that of the S3B model, in which the monoclinic structure dominates, while
for F816/3 and 224/1 the spectral features resemble those of S3C, containing both the monoclinic
and the orthorhombic form. These vibrational spectroscopic results were also confirmed by SR
µ-XRD measurements (not shown in Fig. 4; cfr. Fig. 1A). Although the micro-FTIR, Raman and
SR µ-XRD analyses clearly demonstrated only the presence of a S-rich monoclinic PbCr1-xSxO4
in F816/3 and 224/1, the additional presence of very low quantities of the orthorhombic phase
cannot be excluded for these samples since a co-existence of both monoclinic and orthorhombic
PbCr1-xSxO4 can be observed starting from SO42- molar amount around 40%.13
Finally, in order to ascertain whether or not it is possible to distinguish among different forms of
PbCr1-xSxO4 via non-invasive in situ measurements, we have examined Van Gogh’s Portrait of
Gauguin, (Fig. 5A) at the Van Gogh Museum, using reflection mid-FTIR spectroscopy. Fig. 5A
illustrates the locations in the yellow areas where FTIR spectra of Fig. 5B were acquired. The
resulting spectral data (four are shown as examples) show the presence of two inverted
reststrahlen bands around 1039 cm-1 [ν3(SO42-)] and 822 cm-1 [ν3(CrO42-)], and two derivativeshape signals at 626 and 597 cm-1 [ν4(SO42-)]. These spectra show features in between of those of
the S3B and S3C model materials. The identification of a chrome yellow form more similar to S3C
22
ACS Paragon Plus Environment
Page 24 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
is confirmed by Raman data obtained from a paint micro-sample (X448_2; Figs. 5A, C) coming
from the painting itself. Likely due to the low abundance of the yellow pigment, no information
about its presence was demonstrated by mid-FTIR analysis.
From a stratigraphic point of view, the presence of bright yellow lacunae is visible on both the
yellow and orange painted areas of the painting background.
The identification of zinc white (inverted band at ca. 387 cm-1, not shown in Fig. 5B) and lead
white (inverted band at ca. 1385 cm-1 and derivative signal at 687 cm-1) in the outer yellow areas
(M_22, M_24), and that of barium sulfate (SO42- combination bands,27 not shown in Fig. 5B) in
the lacunae (M_04, M_15) explains the subtle differences that are visible among the spectra of
Fig. 5B.
23
ACS Paragon Plus Environment
224/1
360
x3
339
323
CrO4 )
2-
F816/3
407
378
2-
845
SO4 )
(C)
977
(B)
2-
F816/3
SO4 )
(A)
F816/3
224/1
224/1
10455
F603/3
843
404
376
10455
F603/3
S3B
357
400
376
336
323
S3B
841
10872
x15
973
Intensity
F603/3
S3C
359
338
323
S3C
10455
A'=Log(1/R)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
627
Page 25 of 30
F409/1
10872
10872
F409/1
F409/1
S1mono
1140
S1mono
950
760
1075 860
60 645 425
-1
wavenum
number (cm )
340
-1
Ram
aman shift (cm )
Figure 4. From top to botto
ttom: (A) optical microscope images, (B) reflect
ection mid-FTIR and (C) Raman spectra collected
ted from yellow areas of original
embedded paint micro-sam
mples F816/3, 224/1, 10455, F603/3, 10872 and
nd F409/1 taken from different Van Gogh’s paint
intings (SI, Table S-2 for further
details). In (B) and (C) spec
pectra of paints S1mono, S3B and S3C are illustratedd iin red color.
24
ACS Paragon Plus Environment
Analytical Chemistry
(A)
M_22
M_24
M_06
M_03
M_04
M_01
M_16
M_15
SO4 )
SO4 )
2-
A'=Log(1/R)
M_24
CrO4 )
2-
626
597
2-
(B)
M_22
M_15
LW
M_04
S3C
S3B
1150
920
690
-1
975
Intensity
360
S3C
406
(C)
845
wavenumber (cm )
x448_2
LW
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 26 of 30
1080
810
420
350
-1
Raman shift (cm )
Figure 5. (A) Photographs of Portrait of Gauguin (37x33 cm, F546 s 257
v/1962) by V. van Gogh (1888, Van Gogh Museum, Amsterdam, NL)
and related micro-sample X448_2. (B) Reflection mid-FTIR spectra
recorded from (blue) the yellow lacunae and (black) the outer pale
yellow areas; (C) Raman spectrum (black) obtained from the yellow
region of X448_2. In (A) labels indicate areas where FTIR data were
acquired, while the dotted yellow rectangle points out the sampling
location; “LW” shows the CO32- (B) bending and (C) stretching mode of
lead white. In (B-C) (red) spectra of paints S3B and S3C.
25
ACS Paragon Plus Environment
Page 27 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
4. CONCLUSION
The combined use of analytical techniques, such as XRD, FTIR and Raman spectroscopy, and
SR-based methods, such as SR µ-XRD and S K-edge µ-XANES, permitted us to identify specific
spectral features to distinguish among the different types of chrome yellow pigments in use at the
end of 19th century. The information acquired was used to investigate original chrome yellow
samples taken from historic oil paint tubes and from paintings by Van Gogh and contemporaries.
The study revealed that indeed different lead chromate-based pigments were used by the
painters. The extended use by Van Gogh of monoclinic PbCrO4, monoclinic PbCr1-xSxO4 and
mixtures of monoclinic and orthorhombic PbCr1-xSxO4 was demonstrated. A co-precipitated
PbCr1-xSxO4 form was also found in one painting by Gauguin and in one sample from a palette
used by Cézanne.
A relevant result is that the characterization of different chrome yellow forms is possible by
using portable instrumentations. Preliminary in situ reflection mid-FTIR investigations
performed on yellow areas of the Portrait of Gauguin by Van Gogh, allowed us to identify the
presence of a PbCr1-xSxO4 chrome yellow, composed by a mixture of monoclinic and
orthorhombic phases. This results was confirmed by laboratory measurements on a related
micro-sample.
As described in the following part IV,18 the relation between the sulfate content, the crystal form
and the susceptibility to darkening during photochemical ageing of different chrome yellowbased model paints has been investigated. The results demonstrate that the exact nature of the
chrome yellow type strongly influences its long-term stability.
For this reason, future works will be dedicated to a detailed study by SR µ-XANES and µ-XRF
investigations of a selection of the micro-samples of Table II with the aim of exploring whether
26
ACS Paragon Plus Environment
Page 28 of 30
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
or not it is possible to establish an effective correlation between the chrome yellow
composition/crystalline structure and the state of preservation of the pigment in original
paintings. On the other hand we also will also seek to document more systematically the relation
between the occurrence of the different types of chrome yellow and their exact
hue/context/function in the Van Gogh’s work.
27
ACS Paragon Plus Environment
Page 29 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analytical Chemistry
ASSOCIATED CONTENT
Supporting information. Additional figures and one table. This material is available free of
charge via the internet at http://pubs.acs.org.
ACKNOWLEDGMENTS
This research was supported by the Interuniversity Attraction Poles Programme - Belgian
Science Policy (S2-ART) and also presents results from GOA “XANES meets ELNES”
(Research Fund University of Antwerp, Belgium) and FWO (Brussels, Belgium) projects no.
G.0704.08 and G.01769.09. The analysis of Portrait of Gauguin was performed within the
MOLAB access activity of the EU FP7 programme CHARISMA (Grant Agreement 228330).
MIUR (PRIN08, Materiali e sistemi innovativi per la conservazione dell'arte contemporanea
2008 FFXXN9), DESY/PETRA III (beamline P06) and ESRF (beamline ID21) are also
acknowledged. Thanks are expressed to the staff of the Museé d’Orsay, the Van Gogh Museum
and the Kröller-Müller Museum for agreeable cooperation.
28
ACS Paragon Plus Environment
Analytical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 30 of 30
References
1
Hendriks, E. Van Gogh’s Working Practice: A Technical Study. In Vincent Van Gogh
Paintings 2: Antwerp & Paris,1885-1888, Hendriks, E.; Van Tilborgh, L.; Waanders Publishers,
2011, 90-143 and references therein.
2
Bomford, D.; Kirby, J.; Leighton, J.; Roy, A. Art in the Making: Impressionism; London:
National Gallery Publications, 1990, p. 158.
3
Monico, L.; Van der Snickt, G.; Janssens, K.; De Nolf, W.; Miliani, C.;Verbeeck, J.; Tian, H.;
Tan, H.; Dik, J.; Radepont, M.; Cotte, M. Anal. Chem. 2011, 83, 1214-1223 and references
therein.
4
Monico, L.; Van der Snickt, G.; Janssens, K.; De Nolf, W.; Miliani, C.; Dik, J.; Radepont, M.;
Hendriks, E.; Geldof, M.; Cotte, M. Anal. Chem. 2011, 83, 1224-1231 and references therein.
5
Kühn, H.; Curran, M. Chrome Yellow and other Chromate Pigments. In Artists’ Pigments: a
handbook of their history and characteristics; Feller R. L., Ed.; Cambridge University Press,
1986, Vol.1, pp. 187-200.
6
Eastaugh, N.; Walsh, V.; Chaplin, T.; Siddall, R. The Pigment Compendium [CD-ROM];
Elsevier, 2004.
7
Jansen, L.; Luijten, H.; Bakker, N. Vincent Van Gogh – The Letters; Thames & Hudson Ltd.,
2009.
8
Effenberger, H.; Pertlik, F. Zeitschrift f ür Kristallographie 1986, 176, 75 – 83.
9
Korenberg, C. British Museum Technical Research Bulletin 2008, 2, 49 – 57.
10
James, R. M.; Wood, W. A. Proc. R. Soc. Lond. A 1925, 109, 598-620.
11
Cole, R.J. Paint Res. Ass. Tech. Pap. 1955, 199, 1-62.
12
Watson, V.; Clay, H.F. J. Oil Colour Chem. Assoc. 1955, 38, 167-177.
13
Crane, M. J.; Leverett P.; Shaddick L. R.; Williams P. A.; Kloprogge J. T.; Frost R. L. N. Jb.
Miner. Mh. 2001, 11, 505-519.
14
Colotti, G.; Conti, I.; Zocchi, M. Acta Cryst. 1959, 12, 416.
15
Xiang, J.; Yu, S.; Xu, Z. Crystal Growth & Design 2004, 4, 1311-1315.
16
Otero, V.; Carlyle, L.; Vilarigues, M.; Melo, M. J. RSC Adv. 2012, 2, 1798–1805.
17
Burnstock, A.R.; Jones, C.G.; Cressey, G. Z. Kunst technol. Konserv. 2003, 17, 74-84.
18
Monico, L.; Janssens, K.; Miliani, C.; Van der Snickt, G.; Brunetti, B, G.; Cestelli Guidi, M.;
Radepont, M.; Cotte, M. Anal. Chem. 2012, submitted.
19
Rioux, J.P. Dossier de l’art 1999, 55H, 66–69.
20
Miliani, C.; Rosi, F.; Brunetti, B. G.; Sgamellotti, A. Acc. Chem. Res. 2010, 43, 728-738.
21
Cotte, M.; Susini, J.; Metrich, N.; Moscato, A.; Gratziu, C.; Bertagnini, A.; Pagano, M. Anal.
Chem. 2006, 78, 7484-7492.
22
Vairavamurthy, A. Spectrochimica Acta Part A 1998, 54, 2009–2017.
23
Figueiredo, M. O.; Pereira da Silva, T. Eur. J. Mineral. 2009, 21, 79–83
24
Alía, J.M.; Edwards, H.G.M; Fernández, A.; Prieto, M. J. Raman Spectrosc. 1999, 30, 105–
114.
25
Doyen, L.; Frech, R. J. Chem. Phys. 1996, 104, 7847-7853.
26
Stoilova, D.; Georgiev, M.; Marinova D. J. Mol. Struct. 2005, 738, 211–215.
27
Miliani, C.; Rosi,·F.; Daveri, A.; Brunetti, B.G. Appl. Phys. A: Mater. Sci. Process. 2012, 106,
295-307.
28
Wilkins, R. W. T. Mineral. Mag. 1971; 38, 249.
29
ACS Paragon Plus Environment
Page 31 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
29
Analytical Chemistry
Frost, R. L. J. Raman Spectrosc. 2004; 35, 153–158.
SYNOPSIS TOC
For TOC only
30
ACS Paragon Plus Environment