First-principles thermodynamic modeling of atomic ordering in yttria-stabilized zirconia

Yttria-stabilized zirconia (YSZ) is modeled using a cluster expansion statistical thermodynamics method built upon a density-functional theory database. The reliability of cluster expansions in predicting atomic ordering is explored by comparing with the extensive experimental database. The cluster expansion of YSZ is utilized in lattice Monte Carlo simulations to compute the ordering of dopant and oxygen vacancies as a function of concentration. Cation dopants show a strong tendency to aggregate and vacate significantly sized domains below 9 mol % Y2O3, which is likely important for YSZ aging processes in ionic conductivity. Evolution of vibrational and underlying electronic properties as a function of Y doping is explored.