Atomistic insights into the chemical stability and ionic transport at Li-metal/Li-argyrodite interfaces

Abstract

All solid-state batteries (ASSBs) based on solid-state electrolytes (SSEs) are a novel Li-ion battery technology with the potential of enhanced safety, longer lifetimes, and increased energy density when coupled with the Li-metal anode. Li-argyrodite (Li₆PS₅Cl) is a promising SSE with high ionic conductivity, produced using cheap and sustainable precursors, and therefore of interest to both academia and industry. Like many other sulfide-based SSEs, it is however unstable against Li-metal. Using ab initio and machine-learning methods, we simulate three representative Li-metal/Li-argyrodite interface models to investigate whether the exact surface termination affects the chemical stability and ion transport capability. We present a systematic approach to create low-energy interfaces by screening 28 low Miller-index surface terminations of Li-argyrodite and coupling them with Li-metal. Custom-made machine-learned interatomic potentials trained on ab initio data enable the simulation of large interface models with over 2000 atoms for 5 ns. We find that all three interfaces decompose into an amorphous solid-electrolyte interphase (SEI) layer, consisting of Li₃P, Li₂S and LiCl, which then crystallises into an antifluorite phase Li₂S_1−x−yP_xCl_y; {x, y = 0.14–0.15}. Calculating the flux of Li-ions across a fixed membrane, we show that the crystalline SEI layer is a poor ion conductor, similar to Li₂S. While all three interfaces form the same crystalline SEI layer, the exact rates of the decomposition and crystallisation depend on the actual surface composition. These atomic-level insights will be practical to control the SEI formation in sulphide-based SSEs and others.