Degradation process of lead chromate in paintings by Vincent van Gogh studied by means of spectromicroscopic methods. 3. Synthesis, characterization, and detection of different crystal forms of the chrome yellow pigment

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. 622% 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