Dislocation nucleation is essential to our understanding of plastic deformation, ductility, and mechanical strength of crystalline materials. Molecular dynamics simulation has played an important role in uncovering the fundamental mechanisms of dislocation nucleation, but its limited timescale remains a significant challenge for studying nucleation at experimentally relevant conditions. Here we show that dislocation nucleation rates can be accurately predicted over a wide range of conditions by determining the activation free energy from umbrella sampling. Our data reveal very large activation entropies, which contribute a multiplicative factor of many orders of magnitude to the nucleation rate. The activation entropy at constant strain is caused by thermal expansion, with negligible contribution from the vibrational entropy. The activation entropy at constant stress is significantly larger than that at constant strain, as a result of thermal softening. The large activation entropies are caused by anharmonic effects, showing the limitations of the harmonic approximation widely used for rate estimation in solids. Similar behaviors are expected to occur in other nucleation processes in solids.

Fracture of silicon and germanium nanowires in tension at room temperature is studied by molecular dynamics simulations using several interatomic potential models. While some potentials predict brittle fracture initiated by crack nucleation from the surface, most potentials predict ductile fracture initiated by dislocation nucleation and slip. A simple parameter based on the ratio between the ideal tensile strength and the ideal shear strength is found to correlate very well with the observed brittle versus ductile behaviours for all the potentials used in this study. This parameter is then computed by ab initio methods, which predict brittle fracture at room temperature. A brittle-to-ductile transition (BDT) is observed in MD simulations at higher temperature. The BDT mechanism in semiconductor nanowires is different from that in the bulk, due to the lack of a pre-existing macrocrack that is always assumed in bulk BDT models.