Date of Award


Document Type

Open Access Dissertation


Chemical Engineering

First Advisor

Ralph E White


Li-ion batteries are promising candidates as power sources for hybrid electric/electric vehicles, as well as storage devices for renewable energies (wind, solar). Longer life batteries are more desirable for large-scale application, which would help lower the capital cost ($/KW) and improve the system stability. However, the aging problem of Li-ion batteries obstructs their fast penetration into these markets. The cell life can be improved based on an in-depth understanding of the fade mechanisms. The purpose of this dissertation is to explore, through a mathematical modeling approach, failure mechanisms of Li-ion batteries.

To study the capacity fade of a LiMn2O4 (LMO) electrode, a pseudo-2 dimensional (P2D) model based on porous electrode theory is first developed. This model takes into account the loss of LMO due to acid attack and the breakdown of the Li ion diffusion pathway due to the formation of the solid electrolyte interphase (SEI) film. The acid stems from the decomposition of the LiPF6 salt and the organic solvent. The decrease of the Li-ion diffusion coefficient is implemented as an empirical function of the loss of LMO. Good agreement is achieved between our simulation results and the experimental data reported in literature. Next, we provide a mathematical model to study the generation of mechanical stress in LMO particles, which are mixed with LiNi0.8Co0.15Al0.05O2 (NCA) as a battery cathode. The mechanical equations which capture the stress buildup in the LMO particle due to Li insertion/extraction are incorporated into the P2D model. The predictions obtained from our blended cathode (LMO and NCA) model show that the stress generated in the LMO particles is reduced at the end of discharge due to adding NCA particles in the electrode. This detailed model can help elucidate the effect of adding NCA particles on the improvement of the LMO electrode performance. Finally, a two-dimensional model is developed for large-format LMO/carbon cells to understand inhomogeneous degradation. The model considers the non-uniform porous electrode properties and the electrode mismatch. The simulation results show that when the anode edge is extended over the cathode edge, the LMO particles near the edge will suffer larger potential drop, larger charge/discharge depth, and higher diffusion-induced stress. Therefore, the loss of LMO is more pronounced near the electrode edge in agreement with experimental observations.


© 2013, Yiling Dai