Structural phase transition path-following and stable phase scouting through a coupled DFT-BFB algorithm

This work formulates a new computational scheme to efficiently explore the configuration space of materials and to identify a material's stable equilibrium structures. This computational tool is obtained by coupling quantum-based density functional theory (DFT) calculations (employing periodic boundary conditions) with branch-following and bifurcation (BFB) techniques. BFB is used to map equilibrium paths (stable and unstable) on the DFT energy landscape as a function of the applied load and ultimately creates 'bifurcation maps' that identify the material's stable structures and connections between them, including: soft deformation directions, transition states, transformation mechanisms, etc. This new approach has been used to study structural transitions in Si and Fe under pressure loading. The results obtained so far indicate that the new DFT-BFB methodology has the potential to provide a significant new insight into the mechanisms that drive structural phase transitions in a wide range of technologically important materials.