Deep learning of experimental electrochemistry for battery cathodes across diverse compositions

Artificial intelligence (AI) has emerged as a tool for discovering and optimizing novel battery materials. However, the adoption of AI in battery cathode representation and discovery is still limited due to the complexity of optimizing multiple performance properties and the scarcity of high-fidelity data. We present a machine learning model (DRXNet) for battery informatics and demonstrate the application in the discovery and optimization of disordered rocksalt (DRX) cathode materials. We have compiled the electrochemistry data of DRX cathodes over the past 5 years, resulting in a dataset of more than 19,000 discharge voltage profiles on diverse chemistries spanning 14 different metal species. Learning from this extensive dataset, our DRXNet model can capture critical features in the cycling curves of DRX cathodes under various conditions. Our approach offers a data-driven solution to facilitate the rapid identification of novel cathode materials, accelerating the development of next-generation batteries for carbon neutralization. [Display omitted] •Machine learning battery informatics across diverse compositions•Representation learning of cathode electrochemistry from experiments•An open-source dataset of discharge voltage profiles Artificial intelligence (AI) has emerged as a tool for discovering and optimizing novel battery materials. Most machine learning approaches predicting battery performance have been focused on predictions for a specific chemistry or limited chemical space of commercialized cathodes due to the complexity of optimizing multiple performance properties simultaneously. Recent studies in computational material science have demonstrated the feasibility of building universal models for atomistic modeling by harnessing more than 10 years of ab initio calculations spanning the periodic table. It becomes a logical extension to envision universal models for the experimental discovery of battery materials. We present a machine learning model that uses an end-to-end training pipeline to encode and learn the (electro)chemical information from experimental voltage profiles. Our approach offers a data-driven solution to facilitate the rapid identification of novel cathode materials. AI has shown promise in battery informatics, but its application to cathodes is limited by complexity and data scarcity. A machine learning model, DRXNet, is developed by training on an extensive experimental electrochemistry dataset encompassing diverse compositions