Data-mining fluoride-based solid-state electrolytes for monovalent metal batteries

Abstract

Developing a high-performance solid-state electrolyte (SSE) for Li and Na metal anodes in high-energy-density batteries involves several challenges, including the need for a material with high ionic conductivity, good mechanical properties, and good compatibility with the anode material. To address these challenges this data-mined study systematically screens and analyzes SSE materials for Li and Na metal batteries (LMBs and NMBs), with a focus on their mechanical deformability – a pivotal yet often-overlooked aspect in high-throughput research. The research employs data-driven materials screening using the Materials Project database, utilizing filters to identify suitable SSEs based on key material-focused requirements, such as low electronic conductivity, chemical and electrochemical (interface) stability, and good mechanical deformability. Our findings reveal that ternary metal fluorides emerge as optimal SSEs for both LMB and NMB systems, exhibiting exceptional mechanical deformability and interfacial stability with the electrodes compared to conventional SSEs. Through machine learning-accelerated molecular dynamics simulations, we efficiently predict and determine the Li and Na ionic diffusivity for the selected fluorides. While most compounds in the refined list have been experimentally synthesized and reported in the literature, the majority of them have not been tested as potential Li or Na-ion conductors. Noteworthy SSE candidates identified include Li₃ScF₆, NaPrF₄, and Na₃HfF₇, and these hold great promise for LMB and NMB applications.