Stress-Controlled Three-Dimensional Phase Evolution in Lithium Iron Phosphate

Abstract

Many battery compounds undergo first-order phase transformations induced by ion (de)intercalation during battery operation, and their rate capability is often controlled by the kinetics of two-phase boundary movement. LiFePO4, a commercial cathode materials for Li-ion batteries, is a model system for studying the unique aspects of phase transitions in battery intercalation compounds. In a recent in-situ scanning transmission X-ray microscopy (STXM) study [Ohmer et al. Nature Communication 6, 6045 (2015)], Li deintercalation and phase evolution in microsized LiFePO4 single crystal upon delithiation are found to be highly non-uniform. FePO4 phase was observed to form [010]-aligned filaments growing into the LiFePO4 crystal, which is postulated to be dominated by elastic effects. In this work, we performed three-dimensional phase-field simulation to examine in the details phase evolution in LiFePO4 and the effect of misfit strain energy arising from the lattice match between FePO4 and LiFePO4. Our simulation reproduces the filamentary growth morphology of FePO4 phase (Figure 1) as observed in experiment, and confirms the dominating role of misfit stress in causing such behavior. We show that phase evolution in LiFePO4 goes through two steps during delithiation. In the first step, phase separation occurs in LiFePO4 via surface-mode coherent spinodal decomposition and results in the formation of individual FePO4 domains on particle surface. Next, FePO4 grows into the LiFePO4 matrix as wedge shaped domains that coarsen at the same time. The filamentary morphology of FP phase within (001) plane is mainly a result of the anisotropic misfit strain between LiFePO4 and FePO4, which favors the elongation of FP domains along the [010] axis to reduce elastic strain energy. Simulation reveals that Li diffusion anisotropy also significantly influences the phase growth morphology. The existence of three-dimensional Li diffusion in LiFePO4, which results from the presence of antisite defects, is found to be essential for the development of filamentary FePO4. The excellent agreement between the phase-field simulation and in-situ STXM experiment suggests that misfit stress plays an important role in triggering non-uniform Li (de)intercalation especially in large electrode particles, which is detrimental to battery performance and life and deserves close attention. Figure 1