Historically accurate reconstructions of Amadeo’s chrome yellows: an integrated study of their manufacture and stability

Abstract

The subject of this PhD is the study of 19th century yellow chromate pigments, in particular chrome yellow (lead chromate), which has been frequently found in the paintings by the Portuguese Modern painter Amadeo de Souza-Cardoso (1887-1918) and by Vincent Van Gogh (1853-1890). Detailed knowledge of the materials used by artists is fundamental for the understanding of their technique and to place their works in context as well as to determine the state of preservation and authenticity of their work. For this thesis new knowledge was gained through a multidisciplinary approach which combines art technological and conservation science research. This is the first doctoral work that fully explores the Winsor & Newton (W&N) 19th Century Archive Database and provides a unique insight into the company’s choices and workshop practices. W&N had a special concern for the quality and durability of their yellow chromate pigments. A complete evaluation of their 286 production records for yellow chromate pigments was undertaken. The majority of the production records (64%) pertain to different hues of lead chromate. W&N produced essentially three lead chromate pigment types: Lemon/Pale based on mixed crystals of lead chromate and lead sulfate [PbCr1-xSxO4] where x ≤ 0.4; Middle based on pure monoclinic lead chromate [PbCrO4]; and Deep that contains the latter admixed with basic lead chromate [Pb2CrO5]. Production records for the manufacture of barium chromate, BaCrO4 (25%), zinc potassium chromate, 4ZnCrO4·K2O·3H2O (9%), and strontium chromate, SrCrO4 (2%) were also found. Despite the high number of production records, each chromate pigment type is characterised by only one or two main synthetic pathways. W&N’s addition of extenders in their lead chromate pigment formulations, indicates that they were mainly used to adjust pigment properties and not as a means to decrease costs. The reduced number of records found for the production of the light-sensitive Primrose type of lead chromate (a formulation composed of sulfur-rich PbCr1-xSxO4), and strontium chromate, suggests W&N was not manufacturing or selling these pigment formulations on a large scale; more evidence that during the 19th century W&N was committed to primarily selling high quality durable chromate pigments. Pigment reconstructions following the main methods of synthesis were produced for this research and characterised by complementary analytical techniques: Colourimetry, Fibre Optic Reflectance Spectroscopy (FORS), micro-Energy Dispersive X-Ray Fluorescence Spectrometry (μ-EDXRF), X-Ray Diffraction (XRD), Scanning Electron Microscopy with X-ray microanalysis (SEM-EDS), micro-Raman (μ-Raman) and micro-Fourier Transform Infrared (μ-FTIR) spectroscopies. A very good correlation was found between the chemical composition of the pigment reconstructions and historic paint samples from 19th century oil paint tubes and Amadeo’s paintings. This validates their high degree of historical accuracy and attests their use as reference materials for further chemical studies. The lead chromate pigment formulations were made into oil and poly(vinyl acetate) (PVAc) paints and irradiated with a Xenon lamp (λirr > 300 nm) to assess their photostability. The degradation process was followed by the above mentioned techniques and was further studied by Synchrotron Radiation based techniques (μ-XRF, μ-FTIR and μ-XRD). By means of micro X-ray Absorption Near-Edge Structure (μ-XANES), an innovative Light Susceptibility Index (LSI) was developed based on the ratio of the Cr K pre-edge intensity after/before irradiation with a Xenon lamp, which decreases as the quantity of Cr3+ degradation species increases. This index will enable the prediction of the stability of lead chromate pigments in works of art. Three degrees of lightfastness were identified: 1) PbCr1-xSxO4 with 0 ≤ x ≤ 0.4/0.5, 2) PbCr1-xSxO4 with x ≥ 0.4/0.5 and 3) PbCrO4 admixed with a high quantity of the extenders chalk (CaCO3) and gypsum (CaSO4.2H2O). This work demonstrates that the manufacturing conditions, in particular the pH of the production of PbCr1-xSxO4 with 0.4 ≤ x ≤ 0.5, influences their photostability. Most importantly, this is the first work to formulate integrated mechanisms that account for the main causes of colour alteration, and which identifies the crucial role played by the pigment and binder formulations, including their additives. The combined use of the Synchrotron Radiation based techniques allowed access at the micro-scale, to the spatial distribution of the degradation products and intermediaries. This led to the discovery that the reduction of Cr6+ from lead chromate to Cr3+ species is driven by the presence of oxalic acid and/or oxalate compounds. It is proposed that these compounds result from the decomposition of carbonate compounds (present as additives) in acidic media or via decarboxylation of carboxylic acids. These degradation pathways are triggered by the photodegradation of the binder and the degradation rate is deeply dependent on the paint formulation. The presence of calcium oxalate as a degradation intermediary and/or protective patina from light absorption is also discussed in relation to these findings. For the first time, it was possible to identify a Cr3+ degradation compound, a chromium potassium sulfate, CrK(SO4)2·12H2O, by infrared spectroscopy. Finally, it was also possible to determine that no degradation resulting from the reduction of Cr6+ to Cr3+ is currently occurring in Amadeo’s paintings which contrasts with literature reports on the degradation of lead chromate pigments in the work of Van Gogh.