3D Micro-Structure Resolved Simulations of Thick Li-Ion Batteries

Abstract

Li-Ion batteries are commonly used in portable electronic devices due to their outstanding energy and power density. A remaining issue which hinders the breakthrough e.g. in the automotive sector is the high production cost. Recently, new battery concepts were presented to resolve this issue1. For low power applications, such as stationary storage, batteries with thicker electrodes (>300 µm) were suggested. High energy densities can be attained with only a few electrode layers which reduces production time and cost2,3. However, mass and charge transport limitations can be severe at already small C-rates due to long transport pathways. This could be a trigger for degradation effects, such as Li plating at the graphite anode, and reduces the lifetime of the battery. A thorough understanding of relevant processes within the electrodes is urgently needed to avoid these problems. In our contribution we present 3D micro-structure resolved simulations of thick (electrodes > 300µm) Graphite-NMC batteries based on our thermodynamically consistent modeling framework BEST4. The simulations are performed on electrode micro-structures which are either taken from tomography data provided in the literature (NMC)5 or virtual reconstructions of SEM images (graphite). A reliable parameterization of the model is absolutely mandatory to ensure the predictability of simulation results. However, the determination of parameters for 3D models is challenging due to the high computational load and a direct fit to experimental data is practically impossible. We propose a systematic approach based on half-cell measurements and supplemental 1+1D simulations. First, thermodynamic and transport parameters are extracted from dedicated OCV and conductivity measurements. In a second step an estimate of the kinetic parameters is obtained by a fit of the 1+1D model to discharge curves of thin electrodes (70 µm). The results of our 3D full cell simulations agree favorably with the experimental data3 at low C-rates. At high currents the experimental capacity is considerably smaller than the simulated one. Our detailed 3D studies allow important insights on cell operation and indicate that an inhomogeneous distribution of the conductive soot in the NMC electrode contributes to the loss in capacity. Moreover, we investigate the possibility of Li plating during battery charge. Based on our simulations we are able to predict an upper limit for the charging current. Therefore, our model provides a tool to avoid critical operating conditions affecting the lifetime of the battery.