One of the central objectives of multiscale materials modeling is to develop and apply simulation techniques to uncover atomic-level mechanisms without the significant limitations on system size and simulation time inherent to purely atomistic methods. In this context, I have made two contributions. I developed an interatomic potential finite element method (IPFEM) to study homogeneous dislocation nucleation by nano-indentation. The implementation of IPFEM facilitates simulations at length scales that are large compared to atomic dimensions, while remaining faithful to the nonlinear interatomic interactions. Aided by a shear localization criterion, which was also calculated from the interatomic potential, I was able to provide atomically accurate predictions about when, where and how a dislocation nucleates beneath a nanoindenter. My second contribution was to extend the time-scale of atomistic simulation of fracture by adopting several reaction pathway sampling schemes. I studied the thermally activated processes at a crack tip that control the brittle to ductile transitions in solids. Using the sampling scheme of the nudged elastic band method, atomistic pathways were identified that characterize dislocation loop emission in Cu, cleavage crack extension in Si, and water-assisted bond ruptures in a silica nanorod. The associated energetics and atomistic geometries were quantified, thus making contact with previous continuum analyses and experimental observations.