Monte carlo- And simulated-annealing-based funneled approach for the prediction of cation ordering in mixed transition-metal oxide materials

The development of cathode materials with good electrochemical performance is vital to the effort to design rechargeable batteries for applications such as electric mobility. Li-intercalated mixed transition-metal layered oxides such as LiNi_xCo_yMn_(1-x-y)O₂ (NCM) and LiNi_xCo_yAl_(1-x-y)O₂ (NCA) have gained considerable attention due to their good electrochemical properties. The properties of these materials are strongly affected by the arrangement of ions in their crystal lattice, yet accurate determination of this arrangement using experiments and theory remains a challenge. Here, we present a hybrid approach based on Monte Carlo (MC) and simulated annealing (SA) in conjunction with an empirical potential to determine the most probable ionic arrangements from a large number of possibilities. The MCSA approach is followed by a density functional theory (DFT)-based rescoring to determine the energetically most favorable configurations at a higher level of theory. The MCSA + DFT approach is tested on several well-known NCM and NCA cathode materials. We also employed this approach to determine the energetically favored dopant sites and cation mixing configurations. The lowest-energy configurations are then validated by comparing calculated structural parameters, electronic structures and ⁷Li NMR chemical shifts with experiments. The MCSA + DFT approach reported here is a computationally cheap and efficient approach for structure prediction of mixed transition-metal-based materials.