THE DEVELOPMENT OF CHARACTERIZATION PROFILES AND A SPECTRAL LIBRARY FOR PIGMENTS

Abstract

Our understanding of artists and how they work can be greatly enhanced by a knowledge of the materials used and the method of application. Such information not only underpins our understanding within an art historical context, it also informs our approach to conservation and can be critical in cases of challenged authenticity. Most methods of pigment analysis require the removal of a sample from the artwork, for example the examination of crosssectional paint samples or the use of the polarizing microscope. This procedure is not always feasible when paint films are thin, as is the case with watercolour paintings, or even desirable when the painting is in otherwise good condition. In such cases the artwork is rarely able to contribute any detailed information regarding the materials used and researchers must rely entirely on textual sources such as diaries, letters and studio records, where they exist. If a non-sampling, non-destructive method of pigment analysis could be developed, it would provide an invaluable tool in developing our understanding of the materials used by a wider range of artists. Since many individual methods of analysis are often inconclusive, the aim of the research was to develop a holistic, non-sampling approach to pigment analysis, which would develop distinct pigment profiles. These profiles will be made readily available on our website and eventually contribute to a much broader 'Characterisation Database'. Each profile will be supported by detailed guidelines on the methods of analysis, a directory of institutions or consultants offering the relevant analytical services, access to our spectral libraries, a dynamic archive of materials and techniques, together with links to other relevant websites. The initial research has focused on a range of blue pigments from contemporary pigment manufacturers, together with historic samples from the Roberson Archive held by the Hamilton Kerr Institute, University of Cambridge. This archive includes a wealth of written documentation regarding the company as well as original paint-boxes, together with paint recipes and painted-out sample cards. It also holds a complete record of all the account-holders with the firm as well as their original correspondence. It has been known for some time that certain materials have specific reflectance or absorption characteristics when viewed under ultraviolet light: metal-based pigments such as lead white appear black whereas animal adhesives fluoresce lemon-yellow. To date, no comprehensive study of these characteristics has been published. A 1941 publication by Sheldon Keck [1] drew attention to the fact that certain pigments appear a different colour when viewed under infrared light. This characteristic colour change has become known as the 'false colour' of the pigment and this technique of analysis as 'false colour infrared' [2]. Although small pockets of research have looked at these techniques [3] it has not been possible until comparatively recently to obtain a recording of the colour changes since photographic processing to the required specification was not available. This situation changed during the 1980s when Kodak developed a high specification film which could record accurately in the infrared end of the spectrum and which did not require expensive processing [4, 5]. Our conservation department has for some time used both ultraviolet and false colour infrared as investigative tools. Although the tool was a useful comparative indicator, its application was limited due to the variations in colour which could result from variables in photography and developing techniques. The inclusion of a colour chart of known paint samples within the infrared image, as recommended by Clarke and Meijers, has helped develop the technique as a method that can be used to make a more positive identification of certain pigments [6]. The contemporary and historic blue pigments which have been investigated include ultramarine, cobalt blue, indigo, Prussian blue and azurite. Each pigment has been photographically recorded using visible, ultraviolet and infrared light, followed by recording the spectra produced by each sample using spectrophotometry, and crystalline analysis using Raman microscopy. Instrumental analysis of each sample was also carried out, where appropriate, using Fourier transform infrared spectroscopy, scanning electron microscopy/energy dispersive X-ray analysis and gas chromatography/mass spectrometry, for a better understanding of any deviations from the expected photographic and spectral results. This has enabled us to produce characteristic profiles for each of the pigments, which will contribute towards the development of the Characterisation Database. It is intended to extend this work into a wider range of pigments and pigment mixtures and, most importantly, to compare the findings with similar analyses of other historic and contemporary pigment samples. Jean Brown is senior lecturer, Colin Liddie is a conservation intern and Anne Bacon is course director for the MA Conservation of Fine Art, School of Humanities, Burt Hall, University of Northumbria, Newcastle upon Tyne NEl 8ST, UK.