Electronic charge effects on dislocation cores in silicon

Abstract

Using first-principles calculations, we investigate electronic charge effects on the structural stability of partial dislocations in silicon. For the 30° partial dislocation, we find that the unreconstructed core sustains all possible charge states associated with the dislocation-related electronic bands, as the Fermi level (μe) sweeps the electronic band gap, while the reconstructed core remains neutral for p-type doping and intrinsic regimes. Both core configurations become negatively charged for n-type doping. In the case of the 90° partial dislocation, the three known core configurations (namely, the single-period and double-period reconstructed cores and the unreconstructed one) remain neutral in the p-type and intrinsic regimes, but the negatively charged states become stable in the n-type region, for all three geometries. More important, we find that the relative stability between the three structures is strongly charge-state dependent, with the unreconstructed core becoming energetically favorable in the n-type regime. Our results provide elements for understanding the role of doping on dislocation mobility in semiconductors.