Application of Multivariate Data Analysis Techniques in Modeling Structure–Property Relationships of Some Superconductive Cuprates

Multivariate analysis methods are used for the examination of the fine structure and superconductivity properties of the 1212type superconductive copper oxides with the stoichiometry of MA2QCu2O6+z (M=Cu, Hg, Tl/Pb; A=Ba, Sr; Q=rare-earth element, Ca; z=0–1). PCA (principal component analysis) is used for the qualitative evaluation of relations between structural variables such as the cation–oxygen bond lengths, bond angles, and oxygen content z, and the superconductivity transition temperature TC, adopted from a number of neutron diffraction studies published for MA2QCu2O6+z samples with various compositions. The different ways of doping positive holes in the superconductive CuO2 plane are discussed on the basis of the PCA results. Quantitative modeling of the value of TC in the CuA2QCu2O6+z system is successfully performed by PLS (projections to latent structures by means of partial least squares), resulting in a model with predictive power of ~93%. The present study demonstrates the potential of multivariate analysis methods in studying structure–property relations of inorganic materials with ionic structure.