Predicting Lithium Transport Mechanisms Using Solvent Metrics With Experimentally-Validated Classical Molecular Dynamics Simulations

Lithium-oxygen batteries have higher energy densities than traditional lithium-ion batteries but are not yet commercially viable due to poor efficiency, high charging voltages, and low cycle lifetimes. These issues could be addressed with a deeper fundamental understanding of the atomistic behavior of these batteries, especially how different factors impact lithium transport behavior. We have used classical molecular dynamics (MD) simulations to examine lithium transport behavior in different solvents. We have validated our classical force fields by experimentally measuring the densities, viscosities, and ionic conductivities of our solvent–LiTFSI systems for comparison. However, our classical MD simulations have allowed us to investigate properties that are more difficult to access experimentally, such as the residence time of solvent molecules in the lithium solvation shell. Our atomistic simulations also allow us to examine whether a vehicular or solvent exchange mechanism is the dominant lithium transport mechanism in different solvents. Our goal in this work is to develop solvent metrics based on easily measured solvent properties that can be used to predict the microscopic lithium transport behavior in the solvent.