Selection of Electrolyte Additive Quantities for Lithium-Ion Batteries Using Bayesian Optimization

Abstract

The composition of the liquid electrolyte is a key factor in lifetime performance of lithium-ion batteries. The selection and quantification of additives to the electrolyte is an active field of research. This study focuses on finding the optimal additive combination of fluoroethylene carbonate (FEC) and vinylene carbonate (VC) for NMC622-Graphite cells. The central goal of this work is to accelerate the experimental search in a large search area by using a Bayesian-optimization algorithm to guide the search. Different measurements are used as target variable such as open-circuit voltage gradient and coulombic efficiency. Consequentially, the capability of these measurements for accelerated lifetime prediction compared to conventional ageing tests by cycling is investigated. The search gathered and confirmed additive combinations with excellent performance after four iterations with a total of 15 additive combinations analyzed. The results of this study give insights into the interaction of VC and FEC with regard to ageing.