Learning from combined XRD and DFT data unveils critical structural details in NaxCrO2 electrochemical evolution

Abstract

Layered sodium transition metal oxides are typical electrode materials for Na-ion batteries. However, they often can only cycle within certain limited voltage ranges, beyond which the capacity decays quickly. The evolution of structural details approaching the reversible limit thus forms one crucial aspect to understand the decay mechanism for the design of enhanced structural stability of batteries over a potentially wider Na composition and voltage range for higher energy densities. In this paper, we investigate a model system of Na_xCrO₂ through the analysis of both in situ X-ray diffraction data and density functional theory simulations, unveiling for the first time a unique pattern of large nano-stripes formed by an accumulation of Na vacancies embedded in the background of a Na ion density wave, which is critical to the wider Na composition range that can be cycled in NaCrO₂, in comparison with other similar materials, e.g., NaTiO₂, which do not exhibit such a structural detail of large vacancy nano-stripes. Our analysis method of XRD that is grounded in physical principles shares the same optimization philosophy as machine learning methods, highlighting the potential of incorporating machine learning into crystallographic analysis for large and complex structures in the future.