Asymmetric Domain Nucleation from Dislocation Core in Barium Titanate: Molecular Dynamics Simulation using Machine‐Learning Potential through Active Learning

Barium titanate (BaTiO 3 ) is a ferroelectric material without toxic elements, whose properties such as permittivity, coercive field, and spontaneous polarization are affected by the nucleation of domains reversed motion domain walls. Dislocations can act obstacles to domain‐wall migration or active sites for nucleation. Thus, studies have been conducted on utilization dislocations improve BaTiO . However, atomistic mechanism around dislocation core still unclear. In this paper, machine learning (ML) potential developed study influence The trained using an active‐learning approach ensure accuracy in bulk paraelectric phases, well structures Molecular dynamics simulations ML show that reversal depends directional relationship between external electric field dislocation. Furthermore, strong local polarizations exist surrounding core, owing vacancies core. These both when ordered along line. This article protected copyright. All rights reserved.