An investigation of the feasibility of applying Raman microscopy for exploring stained glass.

Raman microscopy (RM) is widely used in archaeometrical studies of pigments, geomaterials and biomaterials in the Cultural Heritage, but one domain has received relatively less attention: the colouring of stained glass. This feasibility study investigates the advantages and disadvantages of employing RM alone in this field by means of a study of modern commercial glasses, modern commercial pigments, and a few archaeological stained glasses, but especially by an experimental project whereby the authors created stained glass. The different kinds of possible unreacted or reacted material are rigorously established. The distinction between Na, K, Ca glasses was explored, as well as the red colouring of an industrial glass which was proved to be due to the presence of (Zn, Cd)S(x)Se(1-x). Yellow, green, blue and maroon pigments were studied before and after an initial firing and then after heating on glass. The quality of the Raman spectra varied enormously and was sometimes disappointing. Nevertheless RM successfully identified various coloured products such as bindheimite, crocoite, cobalt aluminate, haematite; relict reactants such as corundum, eskolaite and oxides of Co or Pb; and provided indications of other phases such as maghemite or Co-olivine. One conclusion is that the amount of chemical reaction between the pigments and the glass is small compared to the amount in between the pigments. Comments are made on the potential for dating archaeological glass from the known age of synthesis of the pigments, and of the dangers of this approach. Overall it has been shown that RM can be useful for studying stained glass, especially for remote in situ analytical operations with mobile RM, but one must expect some problems either with fluorescence or weak spectra.