Lithium thioantimonate argyrodite solid electrolytes, Li 6+x M x Sb 1–x S 5 I (M=Si, Ge), are promising candidates for all-solid-state batteries due to their exceptional ionic conductivity. However, limited mechanistic understanding hinders the rational design of these materials. In this study, we systematically investigate the underlying Li-ion conduction mechanisms and propose a cation-disorder-driven design strategy using machine-learned interatomic potentials (MLIPs). While inter-cage migration via the Wyckoff 16e (T4) site remains significant, enhanced inter-cage migration through Wyckoff 48 h (T2) sites induced by Si and Ge dopants emerges as a critical factor for achieving high ionic conductivity. Additionally, Si and Ge exhibit distinct inductive effects: Si requires higher substitution to activate T2 pathways, while Ge achieves optimal conductivity at lower levels. Co-substitution of Si and Ge further increases cation disorder, yielding ionic conductivity up to ~50 mS/cm. This study demonstrates the effectiveness of MLIPs in elucidating conduction mechanisms and facilitating the rational design of advanced argyrodite electrolytes. • MLIPs are used to reveal Li-ion conduction in thioantimonate argyrodites. • Si and Ge doping activates inter-cage migration via T2 and T4 sites • T2-mediated migration plays the key role in enhancing conductivity. • Si-Ge co-doping boosts cation disorder, yielding 50 mS cm -1 conductivity.
Park et al. (Thu,) studied this question.