Diffusion of a MgO dimer on a MgO(100) surface was investigated by using both density functional theory and empirical ionic potentials. Barriers for diffusion via hopping and exchange mechanisms were calculated. A qualitative difference was found between density functional theory and the empirical potential for the oxide exchange barrier. Density functional theory predicted a saddle-point for the process, with a barrier of 0.88eV. The empirical potential of Lewis and Catlow, with a formal charge of ±2.0e, predicted this structure to be a stable intermediate minimum with an energy of 0.19eV, relative to the most stable ad-dimer structure. The empirical potential predicted that the oxide hopping and exchange mechanisms were equally likely, whereas density functional theory showed that the oxide ad-ion hopping mechanism had a lower energy barrier. A Bader population analysis of the density functional theory charge density indicated that the Mg and oxide ions had partial charges of magnitude ±1.7e. Using an empirical potential with this partial charge, the local minimum in the O exchange process became a saddle at 0.62eV, which was in better agreement with density functional theory. The standard deviation between the energy of the density functional theory minima and the saddle-points with those of the empirical potential was reduced from 0.32eV when using the formal charge parameters of Lewis and Catlow to 0.15eV using partial charges. The qualitative agreement found for each diffusion barrier using the partial charge model suggested that a Bader analysis could be used to obtain suitable partial charges for constructing empirical potentials.

MgO Addimer Diffusion on MgO(100) - a Comparison of ab initio and Empirical Models. G.Henkelman, B.P.Uberuaga, D.J.Harris, J.H.Harding, N.L.Allan: Physical Review B, 2005, 72[11], 115437 (8pp)