Ab Initio Study on Surface Segregation and Anisotropy of Ni-Rich LiNi_1–2yCo_yMn_yO₂ (NCM) (y ≤ 0.1) Cathodes

Abstract

Advances in ex situ and in situ (operando) characteristic techniques have unraveled unprecedented atomic details in the electrochemical reaction of Li-ion batteries. To bridge the gap between emerging evidences and practical material development, an elaborate understanding on the electrochemical properties of cathode materials on the atomic scale is urgently needed. In this work, we perform comprehensive first-principle calculations within the density functional theory + U framework on the surface stability, morphology, and elastic anisotropy of Ni-rich LiNiCoMnO (NCM) (y ≤ 0.1) cathode materials, which are strongly related to the emerging evidence in the degradation of Li-ion batteries. On the basis of the surface stability results, the equilibrium particle morphology is obtained, which is mainly determined by the oxygen chemical potential. Ni-rich NCM particles are terminated mostly by the (012) and (001) surfaces for oxygen-poor conditions, whereas the termination corresponds to the (104) and (001) surfaces for oxygen-rich conditions. Besides, Ni surface segregation predominantly occurs on the (100), (110), and (104) nonpolar surfaces, showing a tendency to form a rocksalt NiO domain on the surface because of severe Li-Ni exchange. The observed elastic anisotropy reveals that an uneven deformation is more likely to be formed in the particles synthesized under poor-oxygen conditions, leading to crack generation and propagation. Our findings provide a deep understanding of the surface properties and degradation of Ni-rich NCM particles, thereby proposing possible solution mechanisms to the factors affecting degradation, such as synthesis conditions, coating, or novel nanostructures.

Kinetic Stability of Bulk LiNiO2 and Surface Degradation by Oxygen Evolution in LiNiO2‐Based Cathode Materials

“…The (104) facet is the predominant surface for as‐synthesized LiNiO 2 at high temperature in O‐rich environment, and thus it has been adopted as our surface model . This confirmed by an energy comparison for different surface facets as a function of the oxygen chemical potential is also given in Figure S6 (Supporting Information), as well as in our recently published work . As shown in Figure 5 a, only Li diffusions are observed in the bulk phase during 5 ps of simulation time, and there is no evidence for oxygen migration.…”

“…[37,38] This observation indicates a possibility that oxygen evolution can happen on crystal regions with more open space with reduced migration barrier for the oxygen atom or dimer. [40] As shown in Figure 5a, only Li diffusions are observed in the bulk phase during 5 ps of simulation time, and there is no evidence for oxygen migration. In order to investigate the effect of the available space on oxygen migration, we have compared atomic kinetics in the oxide bulk and surface regions.…”

Review of Computational Studies of NCM Cathode Materials for Li‐ion Batteries

“…Further, there are studies that show existence of additional structural phases, such as rock‐salt, which coexist with the layered structure at the surface of layered materials . It has been proposed that such rock‐salt phases form in concert with release of oxygen .…”

Single-crystal based studies for correlating the properties and high-voltage performance of Li[NixMnyCo1−x−y]O2 cathodes

“…Previous studies using density functional theory (DFT) calculations revealed the importance of the primary particle surface on cathode reactivity and cycling stability. [26][27][28][29] For example, the calculation carried out on LiCoO 2 and NMC111 determined that their equilibrium particle shapes consist of (104), (001) and (012) facets, 26,27 and the (104) family facets are more stable than the (012) facets during electrochemical cycling. Experimentally, oxygen release and thermal decomposition of LiCoO 2 were found to be facet-dependent, and structural reconstruction from the layered to spinel and rock-salt phases occurred preferably on (012) and (104) family facets, as opposed to the (001) facets.…”