Exploring the effects of defect concentrations and distribution on Li diffusion in Li3OBr solid-state electrolyte using a deep potential model

Abstract

Anti-perovskite solid-state electrolytes with high ionic conductivity have been reported to have the potential to replace conventional liquid electrolytes. However, the ionic conductivity values of some solid-state electrolytes obtained from ab initio molecular dynamics (AIMD) or classical molecular dynamics (MD) simulations deviated significantly from experimental measurements. Herein, based on density functional theory (DFT) calculation and a deep potential model, the mechanical and electronic properties of Li₃OBr solid-state electrolyte and the dynamic diffusion behavior of Li ions were studied. Firstly, we calculated the band gap structures, elastic modulus, defect formation energy, and Li⁺ migration energy of the Li₃OBr material using the first-principles DFT method, and the results show that the Li₃OBr material exhibits good chemical and mechanical stability. Then we carried out deep potential molecular dynamics (DPMD) simulation to investigate the dynamic migration behavior of Li ions and explore in depth the effect of defect concentrations and distribution on it. The results indicate that Li ion diffusion ability was linearly dependent on the defect concentration, and the influence of defect distribution on the ionic conductivity is very small. The calculated ionic conductivity value of the Li₃OBr material at a defect concentration of 0.7% was in good agreement with the previous experimental value. When the concentration of LiBr-Schottky defects is increased to 4.2%, the ionic conductivity increases to 5.71 × 10⁻⁴ S cm⁻¹ at room temperature. Our work demonstrated that DPMD presents high precision and efficiency in studying the properties of solid-state electrolytes.