Abstract
We summarise our ab initio calculations on diamond to date and their consequences for diamond film growth. Local spin density calculations on 50 and 62 atom clusters have given good estimates for hydrogenation reactions including activation energies and surface reconstructions. Larger calculations on the icosahedral, tetrahedrally bonded C100 molecule reveal it should spontaneously decompose into two concentric fullerenes (C20 and C80). It is possible that a similar mechanism might cause graphitisation at a diamond asperity if hydrogen were not present, at least where a twin boundary meets the surface. However, the saturated, tetrahedrally bonded C100H60 molecule was found to be stable, elegantly illustrating one of the much discussed roles of hydrogen in diamond growth. We have tested the graphite-diamond interface model of Lambrecht et al. (Nature, 364 (1993) 607–610) and find the interface to be stable, but less so than their interatomic potential calculations suggested. We find some evidence for bond reconstruction and basal dislocation formation.