Two low-expansion Li-ion cathode materials with promising multi-property performance

We have identified CrOF4 and NbFe3(PO4)6 as candidate cathodes with density functional theory (DFT) calculations which suggest a useful balance of low chemical expansion and gravimetric energy density. Low chemical expansion is a likely requirement for high cycle batteries and the employment of solid electrolytes. CrOF4 and NbFe3(PO4)6 have predicted average voltages of 5.1 and 4.0 V and chemical expansion less than 3% and 1% within the stoichiometric range of 0 to 1 Li per formula unit, respectively, significantly outperforming commonly used cathode materials. While practical energy densities can be challenging to estimate with DFT calculations, DFT suggests that these exhibit gravimetric capacity densities in excess of 200 and 100 Ah/kg and gravimetric energy density in excess of 1020 and 400 Wh/kg, respectively, depending on irreversible processes during cycling. These were identified by screening approximately 38,000 compounds using statistical models trained on available data and physically motivated descriptors. Renewable power sources such as solar and wind energy require stable, long lasting grid energy-storage systems that can hold and distribute energy when the sun is set. Battery cathodes with high mechanical durability are required for high cycle applications approaching 1000 or more cycles, including grid and electric vehicle applications. However, useful battery cathodes must also satisfy a number of other constraints, including high energy density, high rate capability, and electrolyte compatibility in addition to sustainability requirements such as elimination of cobalt and other elements that are expensive or are acquired at the expense of ethical issues. Historical approaches to the identification of new cathodes have focused largely on the optimization of a single of these properties rather than a holistic optimization. Here we identify two new cathode candidates that satisfy a useful balance of the spectrum of required cathode properties, including minimal chemical expansion for maximization of cycle life, smaller than commonly employed cathodes. These new cathodes were identified using a data driven approach that identifies promising multi-property materials through statistical models. This approach represents a new paradigm for materials search that directly addresses the largest bottleneck for materials deployment: multi-property optimization.