An Atomistic Interpretation of the Oxygen K-Edge X-Ray Absorption Spectra of Layered Li-Ion Battery Cathode Materials

The Li-ion battery has been instrumental in the development of many important consumer electronics, and is supporting the trend towards increased use of renewable energy sources. The energy density and cost of modern Li-ion batteries is currently limited by the available cathode materials, which are typically based on layered metal-oxides that can readily intercalate Li (e.g. LiCoO₂). Current efforts to increase accessible cathode specific capacity and reduce cost have focused on replacing scarce and hazardous elements (e.g. Co) with less expensive elements (e.g. Mn and Ni), giving rise to resulted lithium nickel-manganese-cobalt oxides (NMC)[1]. The charge compensation mechanism, in which both the transition metals (TMs) and the oxygen atoms can participate, needs to better understood to optimise the structure and composition of NMC materials. Understanding the role of oxygen in the charge compensation mechanism is thought to be particularly crucial due to its possible role in increasing the achievable voltage and therefore accessible capacity, whilst also providing a pathway to mitigate potentially being involved in certain degradation mechanisms[2]. X-ray absorption spectroscopy (XAS) of the O K-edge is an experimental probe of the oxygen environment, and can also be atomistically interpreted through first-principles simulation methods based on density functional theory (DFT).<br/><br/>Here, we will introduce our recent work on systematically interpreting the O K-edge of layered lithium transition-metal (TM) oxides from first principles using DFT[3]. Our benchmark spectra show that the semi-local meta-GGA functional rSCAN provides a better match to experiment of the excitation energies of spectral features compared to the GGA functional PBE, or PBE with a Hubbard U correction, especially at energies close to the main edge. Using rSCAN, DFT modelling of the O K-edge XAS of LiNiO₂ and a simulation cell that includes a Jahn-Teller distortions, a closer match to the experimental spectra is achieved. This reveals that the pre-edge contains information about not only the chemical species, but also geometric distortion. Atomistic interpretation of the O K-edge XAS of layered Li TM oxides is also shown to be sensitive to other changes in the octahedral environment, including changes in the chemical identity and the magnetic configuration of coordinating species.<br/><br/>The direct comparison of theoretical spectra arising from simple structural models with experimental data has highlighted the heterogeneity present even in nominally pristine materials. Understanding such heterogeneous structural and electronic environments in pristine materials and beyond, lies in the ability to create fingerprint spectra from more complex models of materials[4]. Thus, preliminary results of clustering projected density of states that arise from molecular dynamics (MD) trajectory of LiNiO₂ will also be discussed. The use of smooth overlap of atomic positions (SOAP) descriptors to connect changes in the partial density-of-states (pDOS) to the local environment is<br/>shown, along with the use of a regression model to learn the electronic structure of the trajectory. This methodology could be readily extended to observe the impact on the spectral shape arising from the material in different states of charge, or the effect of a change in the composition, to help optimize NMC materials and design novel ones.<br/><br/>Thus, the work shown in this presentation, and the future directions that it opens up, highlight the power of the atomistic tool of DFT to interpret the oxygen environments of layered Li-ion battery cathode materials, and helps bridge the gap between theory and experiment.<br/><br/>[1] de Biasi, L. <i>et al. </i>Adv. Mater. 2019, 24.<br/>[2] Assat, G. <i>et al. </i>Nat. Energy 2018, 3, 373–386.<br/>[3] Ramesh, N. <i>et al. </i>“An atomistic interpretation of the oxygen K-edge X-ray absorption spectra of layered Li-ion battery cathode materials”, submitted.<br/>[4] Aarva, A. <i>et al. </i>Chem. Mater. 2019, 31, 9243–9255.