Traditionally X-ray diffraction (XRD) has not been significantly used by the cultural heritage sector, as the most commonly used technique, X-ray powder diffraction (XRPD), requires sample removal from the artwork or artefact, followed by grinding of the sample to a fine powder. With the development of micro-X-ray diffraction (micro-XRD) and other associated advances, non-destructive, in situ analysis of crystalline pigments on micro-areas of paintings and artefacts has become possible. This is illustrated with the application of this technique to the in situ analysis of propriety paints, paint cross-sections and small paintings. However, workers have noted that XRD can be insufficiently sensitive with pigment mixtures, due to matrix effects. This study obtained X-ray diffractograms of 17 single pigment preparations and determined relative amounts of crystalline content. The purity of single pigments was compared with Joint Committee on Powder Diffraction Standards (JCPDS) database diffractograms._x000D_
The impact of pigments with high X-ray absorption coefficients on the detection limits of pigments with low X-ray absorption coefficients was further investigated by building on work done with the chromium oxide-zinc oxide series with the preparation of a second series of binary mixtures of zinc oxide (which has a high absorption coefficient and is almost fully crystalline) and ultramarine blue, which has a lower absorption coefficient but which is more crystalline. Semi-quantitative analysis, where baseline and absorption factor corrections are made, confirm detection limits as low as five per cent for chromium oxide and ultramarine blue when in a binary matrix with zinc oxide._x000D_
However, this type of analysis is very time consuming. So in order to theoretically simulate issues that will be encountered with binary pigment mixtures, the single phase diffractograms were used to theoretically investigate the matrix effects of the three white pigments with the various coloured pigments. This illustrated that detection limits can be as low as five per cent or in non-ideal conditions, such as with partial peak clashes or where there are large differences in relative peak intensity, as high as 50 per cent._x000D_
These results have significant implications for the analysis of painted surfaces on artworks or artefacts. It has been demonstrated that ideally this method is more suited to the investigation of surfaces where there is minimal mixing of pigments or where there is not a significant difference between the absorption coefficients of the mixed pigments.