Li Transport in Defected LiNiO2 from First Principles: Diffusion, Nucleation, and Charge Transfer

Abstract

Electrochemical intercalation of Li in battery electrodes involves a few types of kinetic processes, including diffusion in inhomogeneous media, phase nucleation/transition, and surface charge transfer. They span several length and time scales and are often coupled with each other, which makes it challenging to identify the rate-limiting step and extract kinetic parameters by fitting an empirical model. Using layered cathode LiNiO2 as an example, I will present how all these processes can be incorporated in one atomistic model without empirical inputs, which reals the kinetic competition between different processes. In this scheme, the energetics of various local environments were calculated by the density functional theory (DFT). And the obtained data were used to train a surrogate Hamiltonian with the cluster expansion method. Then the kinetic Monte Carlo (KMC) simulations were conducted with ion hopping barriers updated on-the-fly based on the Brønsted−Evans−Polanyi (BEP) relation. The charge transfer barriers at the electrode-electrolyte interface were calculated separately using the constant-voltage eNEB method1.The simulation results shown below largely reproduce the experimental voltage profile of LiNiO2 2. Both the hysteresis and the irreversible capacity loss (IRC) at the first discharge are captured. The rate limiting step switches as the Li concentration changes during cycling. Moreover, the layered LiNiO2 is never perfect and there are always some excess Ni in the Li layer, which significantly affects the electrochemical performances. My simulations peek into the atomistic origins of these effects. The excess Ni formed in the bulk during synthesis increase IRC by impeding Li diffusion3; the excess Ni near the surface formed during cycling due to oxygen loss traps Li in the fatigued particles by suppressing Li-poor phase nucleation below certain Li concentration4.