Hieu Anh Doan¹, Garvit Agarwal¹, Rajeev Assary¹; ¹Argonne National Laboratory

A flow battery is an electrochemical system capable of providing reversible conversion between electrical and chemical energy. As the electro-active materials are stored externally in an electrolyte and only introduced into the device during operation, this separation of power and energy components allows for flexible design and optimization. However, these electro-active species are often susceptible to polymerization and become dysfunctional. Therefore, a strategy for improving performance of flow batteries is to screen for electro-active materials with motifs that enable regeneration via programmable decomposition. Due to the extremely large search space, we employed a computational workflow consisting of high-throughput quantum mechanical calculations and machine learning. Our results demonstrate the efficacy of this platform in facilitating the discovery of new and promising active materials for flow batteries.