Structural characterization of deformed crystals by analysis of common atomic neighborhood

Simulations of crystal deformation and structural transformation may generate complex datasets involving networks with million to billion chemical bonds which makes local structure analysis a challenge. An ideal analysis method must recognize perfect crystal structures, such as face-centered cubic, body-centered cubic and hexagonal close packed, and differentiate structural defects such as dislocations, stacking faults, grain boundaries, cracks and surfaces. Currently a few methods are used for this purpose, e.g., the Common Neighbor Analysis (CNA) and the Centrosymmetry Parameter (CSP). This paper proposes an alternative method based on the calculation of a single parameter that depends on the common atomic neighborhood. We validate the method characterizing local structures in complex molecular-dynamics datasets, clarifying its advantages over the CNA and the CSP methods.