Computer simulation of ionic solids of technological interest

In the present study we have applied Quantum Mechanical (QM) and Molecular Mechanics (MM) computational methods to solid state materials of interest, specifically: metal hydrides, olivines, ceria and tin oxide crystals.

A new empirical potential set has been derived for the following metal hydrides; NaH, LiH, MgH2, CaH2 and BaH2. Studies of atomic diffusion have been performed in LiH and CaH2 and agree well with the available experimental data. MM simulations of cation diffusion in olivines have been performed. Diffusion has been found to come mainly from Mi-Mi nearest neighbour jumps with a smaller contribution by site exchange, MrM2 and M2-M1; jumps. The spatial exploration of non-linear diffusion paths has been shown crucial in order to yield correct estimations of the activation energy of diffusion. MM calculations of solution and clustering of Cu2+ in ceria show that, on the contrary to previous work, it is quite likely that the dopant enters in interstitials sites and, consequently all clusters observed experimentally may be formed by Cu2+ interstitials.

QM cluster and periodic slab simulations have been performed on the CO and C02 adsorption on the (110) perfect surface of Sn02 with HF and DFT methods. These calculations show that, for CO, the molecule adsorbs on pentacoordinated cations on, ideal, truncated bulk, and relaxed surfaces. The interaction with the surface is mostly electrostatic and well described by uncorrelated methods. In the C02 case, the molecule adsorbs perpendicularly to the surface on the same site, but the main bonding contribution is the polarization of the C02 and DFT methods are necessary to properly describe the binding. Carbonates have been shown to be formed only as metastable species on the perfect (100) Sn02 surface.