Testing with Thin Lithium Anode and Practical Capacities for Fast Evaluation of Polymer Electrolytes for Solid-State Batteries

Abstract

Lithium metal solid-state batteries can reach greater than 400 Wh/kg energy density compared to ~300 Wh/kg of conventional lithium-ion batteries. However, the amount of inactive material needs to be significantly reduced, and active material loading maximized. For example, the lithium metal anode needs to be <10 µm, the solid electrolyte separator layer needs to be <20 µm, and the cathode capacity >4 mAh/cm2. Several new inorganic and polymeric electrolytes have been discovered in the past decade. Nevertheless, the integration and testing of these novel candidate materials into functioning cells with the electrode loadings and dimensions required to meet the cell-level energy density goals have been limited. For example, typically, hundreds of microns thick lithium foils are used, which can hide any lithium losses at the anode due to reactions or physical isolation and lead to artificially long cycle life. The amount of lithium cycled is also typically too low (<1 mAh/cm2) compared to the desired 4 mAh/cm2 in both symmetric and full-cell tests. Such testing makes it difficult to gauge the progress of the field in reaching practical energy density goals. We argue that testing new materials solutions at full-cell level with practical electrode loading/dimensions can help understand their limitations quicker and progress faster towards reaching the cell-level energy density goal of >400 Wh/kg.In this work, using two polymer electrolytes (a “dry” and a gel-type), we present several examples that demonstrate the impact on solid-state battery performance as electrode thicknesses/loadings and cycling capacities are shifted towards practical values. First, using thermally-evaporated ultra-thin (1-10 µm) lithium metal anodes, we quantified the lithium losses during cycling at the lithium/polymer-electrolyte interface. Our results show severe lithium losses or ultra-low coulombic efficiency of a common gel-polymer electrolyte leading to complete consumption of the lithium reservoir within a few cycles. Second, we present examples that highlight the challenges of increasing cycling capacities towards practical values. These examples underscore the need to test novel materials solutions directly with such practical capacities, either in full-cell or symmetric-cell formats, to accelerate the materials development endeavors.