Computational Discovery of Conduction Mechanisms in Lithium-Ion Solid Electrolytes

Abstract

As society attempts to combat climate change and reduce reliance on fossil fuels, energy storage technologies, including batteries, are of increasing importance and so improvement of their performance is essential. Electrolytes are vital in battery performance; electrolytes in commercial batteries are normally liquids, but the use of solid-state electrolytes could improve properties such as thermal stability, safety and energy density. However, most solid-state electrolytes have far lower ionic conductivities than their liquid counterparts and atomic-level insights are needed to develop "design rules" for new, improved solid electrolytes with higher conductivities. In this thesis we used machine-learned interatomic potentials (MLIPs) and molecular dynamics to probe the ion dynamics of three potential solid-state electrolytes - Li3 PS4 , Ag3 SI and Li6+x PS5+y Cl1 – y and investigate which factors affect the cation conductivity in each. In Li3 PS4 , we ran MLMD simulations on cells consisting of varying proportions of the low temperature phase, γ-Li3 PS4 , and β-Li3 PS4 , the higher temperature phase. We found that increased disorder in the anion orientations increased ionic conductivity, and that the "paddlewheel mechanism" mentioned in literature was actually reorientations caused by a phase transition. For Ag3 SI, we investigated how the quenched high-temperature phase, α* -Ag3 SI, displayed higher conductivities than the lower temperature β-phase. We quantified the anion disorder by defining an order parameter, before running MD simulations on a series of cells with varying numbers of anion swaps. We found that greatest effect on the conductivity occurred within a few anion swaps, and this increase was caused by sulphide ions forming a facile pathway for Ag-ion diffusion. For our investigation on Li6+x PS5+y Cl1 – y , we altered both the Li+ and Cl – stoichiometry of the cell. We found that increased Li+ ions promoted concerted diffusion through increased occupation of the interstitial sites involved in the long-range diffusion pathway, and increased Cl – stoichiometry resulted in an increased cell volume, which opened up diffusion bottlenecks. We determined that the disorder in the anion framework was the more influential factor, but that both utilised together could result in a promising material that exhibited conductivities of up to 0.88 S cm−1 at in Li7 PS5 Cl at 400 K, and that aliovalent substitution of the phosphorous could be a way to achieve a charge-balanced version of this stoichiometry. It is hoped that the factors found in this thesis can be used to create higher-performing solid-state electrolytes in the future, with the goal of making commercial solid-state batteries a reality.

Date of Award	24 Jun 2026
Original language	English
Awarding Institution	University of Bath
Supervisor	Benjamin Morgan (Supervisor), Steve Parker (Supervisor) & M. Saiful Islam (Supervisor)