Size mismatch effects in oxide solid solutions using Monte Carlo and configurational averaging

Abstract

Local minima configurational averaging (CA) and Monte Carlo (MC) simulations are used to examine in detail the variation of thermodynamic and structural properties of binary oxide solid solutions with the volume mismatch between the end members. The maximum volume mismatch studied corresponds to that in the CaO–MgO solid solution, a prototype example of a strongly non-ideal system with large miscibility gap. In addition, solid solutions of CaO–HypO using designed hypothetical atoms (Hyp) with atomic radii between those of Ca²⁺ and Mg²⁺ have been considered. Calculations on the hypothetical systems allow not only the systematic investigation of size mismatch, but also the detailed examination and comparison of the CA and MC methods. A particularly efficient implementation of the CA method is via the rapid calculation of the radial distribution function (RDF) for all possible arrangements obtained by distributing the different ions on their respective crystallographic sites followed by full structural optimisation of just one configuration from each group with the same RDF. Comparison of results from CA, using optimisations in the static limit, and MC indicates the importance of cell-size and vibrational effects, which can be particularly important for the largest size mismatches. The enthalpies, excess configurational entropies, vibrational entropies and volumes of mixing scale roughly quadratically for all but the largest volume mismatches. Equally sized atoms cluster together in the first coordination shell for all volume mismatches studied.