The Role of Anisotropy on the Chemo-Mechanical Failure of Secondary NMC Particles Embedded in a Sulfide-Based Solid Electrolyte

Lithium-Nickel-Manganese-Cobalt-Oxide (NMC) is being extensively applied as the cathode for solid-state batteries to match the high energy density of metallic anodes. In practice, NMC particles are embedded in the solid electrolyte (SE), forming a composite structure. One of the fundamental challenges for characterization is the co-existence of multiple failure mechanisms. Cracks can initiate and propagate along the grain boundaries between primary particles, through the primary particles, through the SE, or debonding the interface between particles and SE. Therefore, a thermodynamically consistent multi-physics modeling framework was developed to capture all these mechanisms. In order to consider the initiation and propagation of fracture, a regular phase-field fracture variable was employed. A diffused phase-field parameter was introduced to define the transition of chemo-mechanical properties between the grains, grain boundaries, electrolyte, and electrolyte-electrode interfaces. Inspired by the recent experimental evidence that clearly showed that morphological regulation can help mitigate the internal cracks, our model considered two sources of anisotropy, i.e. morphological orientation and crystal orientation. In addition, the concentration-dependent transport and mechanical properties were considered. The present model was implemented in the open-source FEM package MOOSE for solving three state variables: concentration, displacement, and phase-field damage parameter. Various parametric studies were performed to explore the effect of grain morphology and crystal orientation on the chemical-mechanical response of electrode material. The simulation results of this parametric study agreed well with experimental observations that regulated NMC811 particles have significantly fewer cracks than the randomly-oriented NMC532 particles. Ultimately, we applied this model of a polycrystalline particle embedded in a sulfide-based SE to study the pressure effect on the crack formation and particle fragmentation. Bayesian optimization was used to find the optimal distribution of external pressure on the SE to study three different objectives under the same current density: minimizing the total amount of cracks, minimizing the volume of primary particles that are isolated from Li pathways, or maximizing the coulombic efficiency. Results demonstrated that controlling external pressure can change the electrochemical failure patterns.