Atomistic Models and Molecular Dynamics

Here we show how atomistic computer simulation can help experiment unravel the rich structuralcomplexity of oxide nanomaterials and, ultimately, aid the fabrication of nanomaterials withimproved, tuneable or indeed new properties. We first explore the simulation methodologies:energy minimisation, monte-carlo, genetic algorithms and molecular dynamics together with thepotential models used to describe the interactions between metal and oxide ions. These tools can beused to generate realistic structures that include all the essential microstructural features observedexperimentally, such as surface structure (morphology, surface energy, faceting, surface steps,corners and edges), grain-boundaries and dislocations, intrinsic and extrinsic point defects andepitaxy. We show how the theoretician is able to capture all these (experimentally observed)structural details by attempting to simulate crystallisation. Equipped with realistic models,important properties can be calculated, including: electronic, chemical (catalytic activity, ionicdiffusion and conductivity) and mechanical (hardness, elastic constants). This is illustrated bycalculating the ease of oxygen extraction from the surface of a CeO2 nanocrystal compared with thebulk parent material with implications for oxidative catalysis. Throughout this chapter weemphasise the importance of molecular graphics - a much maligned and underrated tool - butwithout which, the generation of much of the simulation and experimental data would not havebeen possible.