The counterfeiting of works of art is not a new phenomenon, but its scale is now encouraging new thinking that can incorporate scientific analysis in a more systematic way. This lecture has sought to explain a few notions relating to the use of material characterization to detect fraud.
One way is to study the cracks in painted film. Their characteristics can be considered as imprints left by the mechanical stresses accumulated over time in the paint layer, and dissipated during their formation. The paint is weakened as the binder dries, and then as it ages. When subjected to mechanical stress resulting from stretching the canvas on the frame, shocks or storage of the work as a roller, cracks form in a wide variety of patterns, depending on the composition of the layer, its thickness, adhesion to supports, etc. Forger Hans van Meegeren, who skilfully invented works by Johannes Vermeer in the 1930s-1940s, was able to make his paintings look old by adding a phenol-formaldehyde resin to his pictorial material, which rapidly hardened it and induced these cracking phenomena. The cracks were then darkened by redrawing them with ink.
Another way is to check the nature of the pigments present in a work of art. The many colorants invented since the 19thcentury were often used very quickly by artists, and their presence or absence tells us when a work was created. Titanium dioxide is one of them, and the history of its many variants punctuates the 20thcentury . Actual production of titanium white began in Norway in 1918. The first such whites were composite anatases, i.e. mixed with other products such as barium sulfate or calcium sulfate. In 1923-24, the first pure titanium white pigments with an anatase structure appeared in France, where they were marketed in 1925. Rutile-structured titanium white, prepared by a different process, was launched in the USA in 1941 and in Europe in 1946. To improve the pigmentary qualities of rutile, principally the reduction of chalking, treated rutiles (coated with alumina or silica) were manufactured in the USA as early as 1950, and in Europe in 1956. The pigments produced by the different processes are quite different, and therefore have characteristics that have been used to demonstrate falsification, notably in the case of the Vinland map (Yale University, USA) or paintings by various20th-century artists (e.g. in the Wolfgang Beltracchi case).
