The potential of multivariate analysis to phase identification based on X-ray diffraction patterns

• A method of chemical compounds identification using multivariate statistical data analysis based on X-ray diffractograms is proposed.
• The PCA biplot was utilised for a purpose of comparing the experimental diffraction fingerprints of the samples.
• The method is proven to be an efficient tool in the analysis of a multicomponet system.

The possibility of using multivariate data analysis in the X-ray diffraction technique for chemical compound identification is presented. Two data sets consisting of X-ray patterns of the Au–Sn thin film system and laser-induced coloured oxide films on a titanium substrate (Ti–O) are analysed as examples. The proposed approach is based on the assumption that data has a bilinear mathematical structure and, therefore, it can be subjected to principal component analysis and a biplot can be created. The key information was gained when the PCA model that we have developed was employed to reference a database of X-ray patterns. The obtained biplot, in a simple way, allowed for a visual appraisal of the chemical compounds that constituted the analysed samples. Chemical information about the composition of the two sets of targets, Au–Sn and Ti–O, was found to be consistent with that reported in the literature and determined using other methods. Moreover, in the case of Ti–O, which is known as a very complex system, multivariate data analysis enabled the determination of new supplementary information concerning the role of the individual chemical compounds in terms of laser-induced colouring of the titanium surface.