Li(Ni,Co,Al)O2 Cathode Delithiation: A Combination of Topological Analysis, Density Functional Theory, Neutron Diffraction, and Machine Learning Techniques

Here we have combined topological analysis, density functional theory (DFT) modeling, operando neutron diffraction, and machine learning algorithms within the comparative analysis of the known widely LiNiO 2 (LNO) and LiNi 0.8 Co 0.15 Al 0.05 O 2 (NCA) cathode materials. Full configurational spaces of the mentioned materials during delithiation were set using the topological approach starting from the 2 × 2 × 1 supercell (12 formula units in total) of the LNO structure (space group R 3̅ m ). Several types of the DFT models were applied for the structural relaxation of entries of the LNO configurational space (87 configurations) demonstrating a strong dependence of the results of optimization on the initial structure guess (at the latter delithiation stages) and on the Hubbard correction application (for the whole range of delithiation). Within the computationally easiest model considered for LNO, subsequent modeling of the NCA configurational space (20760 configurations) results in structural changes of the model cell that are well-consistent (relative errors <1.5% with respect to the lattice parameter values) with data of operando neutron diffraction experiments during charge–discharge cycling. In the scope of the machine learning approach, topology of Li layers and relative disposition of Li and Al in NCA structure are found to be the most important descriptors during the energy balance estimations.