A Materials Genome Approach for Discovery of Novel Li-Ion Cathodes

Abstract

With the advent of solid-state electrolytes enabling the safer use of Li-metal anodes, Li battery cathodes can operate from a non-Li-containing state-of-charge. This opens up the possibility for new families of cathode materials. First-principles predictions of cathode material performance (in terms of voltage and capacity) is well established [1] and requires calculating the energies of structures with varying amounts of Li in a given framework material. We have developed a robust method of identifying the stable and metastable Li-insertion sites based on analyzing the charge density. This allows us to successively populate any crystalline framework with Li atoms to obtain the voltage and capacity. Furthermore, the identification of all stable/metastable sites in the host material enables us to map-out possible Li diffusion pathways through the lattice and analyze the migration along those pathways. We demonstrate the approach on a test set of more than 10,000 compounds and show the reliability in comparing to known intercalation systems as well as systems that do not intercalate Li.