Accelerated discovery of solid-state electrolytes using Bayesian optimisation

Sherif Tawfik,Julian Berk,Tiffany Walsh,Santu Rana,Svetha Venkatesh
DOI: https://doi.org/10.26434/chemrxiv-2024-0c6zk
2024-10-17
Abstract:Current lithium batteries do not fully meet the longevity and safety requirements of electric vehicles. Novel solid state lithium ion batteries could be a compelling solution to these problems. In this work we unravel some of these new materials with potentially high lithium conductivity by using a Bayesian optimization approach. This involves exploring the material space for new solid-state electrolyte materials with the objective of maximising - lithium diffusivity. The materials selected by the Bayesian optimisation algorithm are then examined using ab initio molecular dynamics to estimate their diffusion energy barrier. We establish that the materials are electronic insulators, a requirement in electrolyte materials, by computing the electronic bandgaps of each of the selected materials using a hybrid exchange method, and then examine the stability of the materials at the lithium metal anode interface by computing the crystal decomposition energies. Out of the selected materials, we find that Li3YBr6 has a reasonably low diffusion barrier, a high bandgap and is potentially the most stable material at the lithium metal interface. In addition to introducing stable and high-diffusivity solid-state electrolyte materials, our work presents a material discovery method that can be applied for a broad range of applications.
Chemistry
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: currently, lithium - ion batteries cannot fully meet the requirements of electric vehicles in terms of lifespan and safety, and new - type solid - state lithium batteries may be an effective solution to these problems. However, existing solid - state electrolyte materials have key limitations such as low lithium - ion conductivity and poor interface stability, which hinder their commercial applications. Therefore, this article aims to accelerate the discovery of new solid - state electrolyte materials with high lithium - ion diffusivity through the Bayesian Optimization (BO) method and evaluate the stability of these materials at the lithium - metal anode interface. Specifically, the research objectives include: 1. **Explore new materials**: Use the Bayesian optimization algorithm to search for new solid - state electrolyte materials with high lithium - ion diffusivity in the material space. 2. **Evaluate material performance**: Calculate the diffusion energy barrier of selected materials through ab initio molecular dynamics (AIMD) simulations to evaluate their lithium - ion diffusion ability. 3. **Verify electronic insulation**: Calculate the electronic bandgap of selected materials to ensure that they are electronic insulators, which is a basic requirement for electrolyte materials. 4. **Evaluate interface stability**: Calculate the crystal decomposition energy to evaluate the stability of materials at the lithium - metal anode interface and ensure their feasibility in practical applications. Through the above steps, researchers hope to find one or more new solid - state electrolyte materials with high lithium - ion diffusivity and stability at the lithium - metal anode interface, thereby promoting the commercialization process of solid - state lithium batteries.