Enhanced Ionic Conductivity at the Solid Electrolyte Interphase of Oxygen-Doped Li₆PS₅Cl

Understanding the formation and transport properties of the solid electrolyte interphase (SEI) at the Li metal | solid state electrolyte (SSE) interface remains a central challenge for all-solid-state batteries, as this buried interphase governs interfacial resistance yet is largely inaccessible to direct experimental characterization. In this work, we combine machine-learned interatomic potential molecular dynamics (MLIP-MD) simulations with a machine-learning phase identification framework to resolve kinetically formed SEI phases at the Li₆PS_5-xClO_x | Li interface and to elucidate the role of oxygen doping. Our simulations reveal that the dominant SEI formed at both the undoped and oxygen-doped interfaces is a Li₂S_1-x-yP_0.5xCl_0.5xO_y phase, corresponding to an anion-substituted Li₂S structure. Moderate oxygen incorporation into the SEI enhances ionic conductivity, whereas excessive oxygen doping degrades both bulk electrolyte conductivity and SEI thermodynamic stability. This trade-off provides an atomistic explanation for the experimentally observed non-monotonic dependence of interfacial performance on oxygen content.