Composition-Dependent Phase Concentrations from First Principles: Simulating Combinatorial Libraries of Transition Metal Oxides*

To identify material phases in experimental combinatorial libraries, we develop a theoretical model as a complementary approach to accelerate phase identification. In this approach, samples in a combinatorial library are simulated as mixtures in chemical equilibria. Each of these mixtures contains all the solid-state phases, which can possibly exist in the library. Using the total energies of these phases obtained in first-principle calculations, we calculate the Gibbs free energy changes in the corresponding chemical reactions, and subsequently evaluate the equilibrium concentrations of the phases in every sample according to the law of mass action. Furthermore, to test this approach, we simulate pseudobinary libraries MnxV1-xOy and CuxV1-xOy. Interestingly, we find that the composition-dependent phase concentrations calculated within our approach agree well with the experimental results measured with XRD spectroscopy.