An improved method for identifying red lakes on art and historical artifacts.

A s Leona notes in a recent issue of PNAS (1), art and objects of cultural heritage hold information about the technology, trade, and aesthetics of a people. This knowledge can be accessed through chemical and material analysis combined with an understanding of the history of technology, science, and art. The information learned is of great value, especially for ancient objects that are so old we have no chance of finding written documentation such as bills of lading, inventories, or manuals and treatises that we may consult for descriptions of the materials used in their creation. The retrieval of this information requires detailed and accurate analysis of the materials and methods involved in the making of art and artifacts. For more than a century, museum scientists studying artworks have been able to identify single particles of many minerals and other inorganic pigments by using optical and chemical microscopy; however, the characterization of organic pigments and dyes from such small samples has been quite difficult. The ancient artists used many earths and minerals and likely just as many organic extracts to expand the range of colors. Unfortunately, many of these organic colorants have disappeared over time. However, some, in particular the red pigments formed by making insoluble complexes by reaction with aluminum salts or adsorption onto chalk or alumina, remain in good condition, especially if they have been protected from light. A fine example is the deep, rich red in the Virgin’s robe painted by Giorgione (Figs. 1 and 2). For cultural heritage studies, results that have value for elucidating our material heritage require stringent demands on the characterization of molecules. Many historical red dyes are anthraquinones. An extract from the root of madder (Rubia tinctorum) was widely used for both dyes and pigments. Three or four related anthraquinones may be present in the natural product, including the well-studied molecules alizarin and purpurin. They are similar to the colorants obtained from animal sources, kermes (also called grain), carmine, and lac. Kermes (kermesic acid; 9,10-dihydro-3,5,6,8-tetrahydroxy-1-methyl9,10-dioxo-2-anthracenecarboxylic acid) is distinguished from cochineal (carminic acid) only by a glucoside unit at the 7 carbon and is also similar to laccaic acid A, the principal component of the red pigment lac, which contains an amide functional group on a substituent at the 7 carbon. Culturally and geographically, however, these compounds have quite separate histories of use. Cochineal was introduced to Europe a couple of decades after the Spanish conquest of Mexico (1521), whereas kermes (and ‘‘Polish’’ cochineal that does in fact contain kermesic and carminic acid) and lac were cultivated and imported into Europe long before the 16th century, perhaps since Biblical times (2). Thus, accurate identification and the ability to distinguish among related molecules is a prerequisite for meaningful conclusions regarding the use and trade of colorants. HPLC was the method of choice for museum scientists to identify natural organic red pigments. Although HPLC requires relatively large samples, the slightly different molecules from different sources can be distinguished. Improvements in these methods are still needed to reduce the sample size that is acceptable for textile fibers but remains rather large for paintings and drawings. Coupling liquid chromatography with new mass spectrometric methods for dye analysis is under investigation (3). The use of in situ UV-visible and fluorescence spectroscopy to identify organic red pigments has been investigated (4). Although this methodology does not require sampling, closely related anthraquinone pigments are not readily distinguishable in paints, and fluorescence of binding media and self-absorption effects can make acquisition of precise data problematic and the analysis of specific pigments in artifacts difficult to perform even when samples are removed from an artwork (5). It would be a great achievement