Atomic perspective on plasticity mechanism of ionic-covalent systems from machine learning molecular dynamics simulations

Abstract

Due to the diversity of atomic bonding, good plasticity is often considered a hallmark characteristic of metals. Novel plastic inorganic semiconductors like α-Ag2S have challenged this conventional thinking, but relevant first-principles calculations still lack an intuitive and comprehensive understanding of the underlying plasticity mechanisms. From the perspective of machine learning molecular dynamics that can describe the microstructure evolution aptly, this work reveals the plasticity mechanism of the ionic-covalent system α-Ag2S. Shear bands or kink bands originating from random and local micro-kinks signify the plastic features, and the subsequent amorphization enables sustained deformation under high strains. Different from features in metals, the oppositely signed dislocation pairs in α-Ag2S can achieve nucleation and motion through coordinated lattice expansion and contraction, while the twining-like kink triggered in a staggered manner allows the material to accommodate large shear strains. The established idealized models capture the unconventional dislocation pair and pseudo-twinning kink, narrowing the blind area in our understanding of plasticity mechanisms within similar systems. The summarized structural and deformation features provide clear clues for identifying other plastic ionic-covalent crystals in superionic conductors.