A Modeling Approach for Optimization of Printed NMC622 Cathode for Capacity Density Improvement under Fast Charging Condition- 3D Simulation and Experimental Validation

This paper aims to investigate the impact of channels in positive electrode on gravimetric capacity density of LIBs (lithium ion batteries) under different charging current rates (C_rate). Typically, the energy/capacity density of LIBs get reduced significantly at fast charging rates due to high internal resistance and tortuosity, and this limits the diffusion/movement of lithium ions across the electrode. Implementation of channels along the thickness of the electrodes improves the diffusivity of lithium ions at fast charging rates. In order to optimize the LIBs performance for fast charging applications, knowledge on the effect of various design factors such as electrode thickness, porosity, width of the channels and distance between the channels is required. Therefore, a 3D model with and without channels was developed in COMSOL Multiphysics software using LIB module to clearly understand the effect of channels in the cathode half-cell. In addition, as a first step towards electrode optimization, the effect of electrode thickness at different charging rates was also simulated. The results indicated that adding channels along the thickness of the electrode improves the capacity density, when compared to the electrode without channels, particularly at extremely fast charging conditions of 4C. In addition, it was found that the increase in electrode thickness increases the impact of channels on improving capacity density and therefore thicker electrodes are getting more benefits from the included channels. Finally, the cathodes with and without channels were fabricated by depositing LiNi
0.6
Mn
0.2
Co
0.2
O
2
(NMC 622) on aluminum substrate using screen printing and bar coating techniques, respectively. The fabricated cathodes were subjected to battery performance tests at different C-rates and a good agreement between the simulated and experimental data was obtained, thus validating both models.