Guiding maps of solvents for lithium-sulfur batteries via a computational data-driven approach

Abstract

Practical realization of lithium-sulfur batteries requires designing optimal electrolytes with controlled dissolution of polysulfides, high ionic conductivity, and low viscosity. Computational chemistry techniques enable tuning atomistic interactions to discover electrolytes with targeted properties. Here, we introduce ComBat (Computational Database for Lithium-Sulfur Batteries), a public database of ∼2,000 quantum-chemical and molecular dynamics properties for lithium-sulfur electrolytes composed of solvents spanning 16 chemical classes. We discuss the microscopic origins of polysulfide clustering and the diffusion mechanism of electrolyte components. Our findings reveal that polysulfide solubility cannot be determined by a single solvent property like dielectric constant. Rather, observed trends result from the synergistic effect of multiple factors, including solvent C/O ratio, fluorination degree, and steric hindrance effects. We propose binding energy as a proxy for Li+ dissociation, which is a property that impacts the ionic conductivity. The insights obtained in this work can serve as guiding maps to design optimal lithium-sulfur electrolyte compositions.