The influence of twin boundary on character and motion of dislocations in silicon

A high density of growth twins is commonly observed within individual Si nanowires as well as Si nanofibers in rapidly solidified Al-Si eutectic. Using atomistic simulations and topological model for pure Si, we reveal two types of twinning dislocations, Shockley partial b p = 1 6 〈 11 ‾ 2 〉 and synchroshear dislocation b t = 1 3 〈 11 2 ¯ 〉 , and confirmed that synchroshear dislocation is more mobile than Shockley partial although synchroshear dislocation has larger magnitude Burgers vector. Both synchroshear dislocations and Shockley partials have higher kinetic barriers for glide on glide-set plane in perfect Si crystal than on CTB in twinned crystal. Regarding slip transmission across CTBs, a mixed full dislocation is blocked by the CTB, while slip transmission does occur via cross-slip mechanism when a screw full dislocation interacts with CTBs. When an edge synchroshear dislocation interacts with an CTB, slip transmission takes place via emitting a screw full dislocation nucleates on the SS while an edge Shockley partial is left on the CTB. Most surprising, CTB-assisted full dislocations gliding has been observed in simulating dislocation loop contraction. This is attributed to two changes in dislocation behavior. Full edge dislocation that is sessile on shuffle-set in perfect crystal becomes significantly glissile on CTB due to a wider core. For full screw dislocation, cross-slip and core transition are suppressed and planar glide mobility on CTB is enhanced as compared to a full screw dislocation on shuffle-set in perfect crystals. The implication of these simulation results on recent experimental observations of room temperature dislocation glide in twinned Si nanofibers are discussed.