A theoretical study of diamond/silicon (001) interface structures with and without graphitic interlayers using a density-functional based tight-binding molecular-dynamics method is presented. The study is motivated by recent progress towards diamond heteroepitaxy on Si (001) using the bias technique. The proposal by Robertson that an initial graphitic deposit with the graphite planes normal to the Si surface (resulting from carbon subplantation) provides a well matched compliant buffer layer and hence may lower the interface energy is examined. It is found here that the models with ``graphitic'' or sp² bonded interlayers have, in fact, higher energy than those without and that the energy increases with the number of graphitic layers. Residual strain in the graphite planes is found to be one of the reasons for this energetic disadvantage. Si surface dimerization was assumed in the initial structures. The diamond surface layer is found to dimerize spontaneously in both cases (with and without graphite interlayers). In models with an odd number of sp² layers, the last sp² bonded atoms next to diamond become the diamond dimerized termination layer. Because of the dimerization, compatibility is required between the 3-on-2 epitaxy and the (2×1) dimerizations of both the Si and C surfaces joined together at the interface. The various ways in which the diamond and Si dimer arrays can be juxtaposed are analyzed and structural relaxation calculations are carried out for each to determine the optimum energy interface structure. The relative energies of the various structures are tentatively explained in terms of favorable and unfavorable local structural units. The optimum structure found has a stepped diamond surface with the lowermost carbon layer dimers orthogonal to the Si dimers.
No takes yet. Share an insight, caveat, or question.
Sternberg et al. (1997) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: