Structure, Thermodynamics, Chemomechanics, and Kinetics: The Hidden Complexities of LiNiO2

Abstract

After being abandoned due to mechanical and thermodynamic instabilities, layered transition metal oxides with ultra-high Ni content currently see a revival as promising cathode active material in the effort of further pushing the energy density of lithium ion batteries. Because such materials are in many aspects closely related to the pure LiNiO2, understanding the properties and mechanisms of this model system on an atomic level is vital for the development of more complex materials. Therefore, we have analyzed the structure, thermodynamics, chemomechanics and kinetics of LiNiO2 and its delithiated states including the presence of native defects by making use of atomistic simulations to grasp the peculiarities of this material.What makes the structure and related properties of LiNiO2 especially interesting and challenging is the presence of dynamic Jahn-Teller distortions of the NiO6 octahedra. This leads to local monoclinic distortions and our simulations revealed why the material still appears to be of rhombohedral symmetry on a global scale.[1]Upon delithiation and the concurrent oxidation to Ni4+, the Jahn-Teller distortions start to vanish and particularly stable Li orderings may be observed. In this context, we used a cluster expansion approach and were able to reproduce the experimental phase diagram. Moreover, we included native defects in the analysis because LiNiO2 can hardly be prepared stoichiometric. Instead, many synthesis approaches lead to additional Ni ions occupying the Li layers (NiLi). The introduction of such defects as well as other substitutionals in a surrogate model shows that single phase regions tend to be suppressed and solid-solution behavior is favored. This smooths the voltage curves and gives good agreement with experiments.[2]We then focused on more extended defects in layered transition metal oxides: Stacking faults (observed at low Li content and responsible for stacking changes from O3 to O1-type layering) and dislocations (needed to drive stacking changes through the material). We found that the energetics and gliding barriers of stacking faults in stoichiometric LiNiO2 are similar to LiCoO2 for all degrees of lithiation. However, by acknowledging the presence of NiLi we quantified for the first time how this defect hinders layer gliding: In the fully delithiated material, 2% of NiLi in the Li layer has a similar effect as 11%-15% Li, effectively increasing the gliding barriers. In terms of dislocations, their excess energies in LiNiO2 are much lower than in LiCoO2, which we mostly attribute to the Jahn-Teller distortions that are able to collinearly orient in the vicinity of a dislocation in LiNiO2, effectively taking a part of the strain. Moreover, dislocations are able to attract both Li and O vacancies, potentially affecting the kinetics of the material as well as acting as a precursor for further degradation mechanisms.[3]Lastly, we mention our recent analysis again focusing on NiLi defects. Our simulations show that such Ni ions remain in a +2 oxidation state independent on the degree of lithiation. This contradicts the broad opinion of an oxidation state change to +3 at low Li content, which is believed to lead to a local constriction of the layer that can hardly be relithiated again. Nevertheless, we show that NiLi indeed leads to an attraction of Li vacancies within two lattice sites and the potential to split fast divacancies into two considerably slower single vacancies. These observations together with the fact that any NiLi depicts an obstacle for diffusing Li ions or Li vacancies shed light on the kinetics of non-stoichiometric LiNiO2 and support the development of further optimization schemes.[4][1] Chem. Mater. 2020, 32, 23, 10096–10103[2] J. Mater. Chem. A , 2021,9, 14928-14940[3] Chem. Mater. 2023, 35, 2, 584–594[4] https://doi.org/10.26434/chemrxiv-2023-hkmsn-v3