A guideline for searching for stable grain boundary structure by utilizing crystallographic information: A method based on non-identical termination

Abstract

Determining the stable atomic structure of grain boundaries (GBs) is one of the most important tasks for computational materials science to comprehensively understand the structure-property relationship of GBs in real polycrystalline materials. In this paper, we revisit a proposed method (PM) that was thought to be ineffective in determining the stable GB structures based on sampling non-identical terminating crystal planes of the two adjoining crystals forming the GB. In contrast to the previous conclusion, we found that the PM is effective regarding both the reproducibility of stable structures and calculation cost. This method could determine the most stable GB structures and energies of not only elemental materials of face-centered-cubic and body-centered-cubic metals, including Cu and Mo, but also a binary ceramic material, MgO. We show that the PM can greatly reduce the number of necessary candidate structures calculated to determine the stable GB structures compared with other brute-force methods sampling the rigid body translation between the two adjoining crystals. We demonstrated that sampling GB structures made of non-identical termination combinations is a useful and effective method to determine the stable GB structures of various materials.