Calphad Modeling of LRO and SRO Using ab initio Data

Abstract

Results from DFT calculations are in many cases equivalent to experimental data. They describe a set of properties of a phase at a well-defined composition and temperature, T, most often at 0 K. In order to be practically useful in materials design, such data must be fitted to a thermodynamic model for the phase to allow interpolations and extrapolations. The intention of this paper is to give a summary of the state of the art by using the Calphad technique to model thermodynamic properties and calculate phase diagrams, including some models that should be avoided. Calphad models can decribe long range ordering (LRO) using sublattices and there are model parameters that can approximate short range ordering (SRO) within the experimental uncertainty. In addition to the DFT data, there is a need for experimental data, in particular, for the phase diagram, to determine the model parameters. Very small differences in Gibbs energy of the phases, far smaller than the uncertainties in the DFT calculations, determine the set of stable phases at varying composition and T. Thus, adjustment of the DFT results is often needed in order to obtain the correct set of stable phases.