Li 5 NCl 2 : A Fully-Reduced, Highly-Disordered Nitride-Halide Electrolyte for Solid-State Batteries with Lithium-Metal Anodes.

Most highly Li-conducting solid electrolytes (σ RT > 10 -3 S cm -1 ) are unstable against lithium-metal and suffer from detrimental solid-electrolyte decomposition at the lithium-metal/solid-electrolyte interface. Solid electrolytes that are stable against lithium metal thus offer a direct route to stabilize lithium-metal/solid-electrolyte interfaces, which is crucial for realizing all-solid-state batteries that outperform conventional lithium-ion batteries. In this study, we investigate Li 5 NCl 2 (LNCl), a fully-reduced solid electrolyte that is thermodynamically stable against lithium metal. Combining experiments and simulations, we investigate the lithium diffusion mechanism, different synthetic routes, and the electrochemical stability window of LNCl. Li nuclear magnetic resonance (NMR) experiments suggest fast Li motion in LNCl, which is however locally confined and not accessible in macroscopic LNCl pellets via electrochemical impedance spectroscopy (EIS). With ab-initio calculations, we develop an in-depth understanding of Li diffusion in LNCl, which features a disorder-induced variety of different lithium jumps. We identify diffusion-limiting jumps providing an explanation for the high local diffusivity from NMR and the lower macroscopic conductivity from EIS. The fundamental understanding of the diffusion mechanism we develop herein will guide future conductivity optimizations for LNCl and may be applied to other highly-disordered fully-reduced electrolytes. We further show experimentally that the previously reported anodic limit (>2 V vs Li + /Li) is an overestimate and find the true anodic limit at 0.6 V, which is in close agreement with our first-principles calculations. Because of LNCl's stability against lithium-metal, we identify LNCl as a prospective artificial protection layer between highly-conducting solid electrolytes and strongly-reducing lithium-metal anodes and thus provide a computational investigation of the chemical compatibility of LNCl with common highly-conducting solid electrolytes (Li 6 PS 5 Cl, Li 3 YCl 6 , ...). Our results set a framework to better understand and improve highly-disordered fully-reduced electrolytes and highlight their potential in enabling lithium-metal solid-state batteries.