Battery performance in part relies on the microstructure of each electrode, which is determined by the electrode composition and fabrication process. Accurate and direct experimental measurement of electrode properties is challenging due to the thickness of electrode films (30-80 µm), yet those properties are needed to generate cell-level performance models and to improve manufacturing processes. Therefore, a realistic microstructure prediction model is necessary for designing and manufacturing better batteries. In our model, we consider all the fabrication steps that would determine the final electrode microstructure in a predictive fashion.

We are not aware of any prior slurry models that can simulate the multi-phase nature of the Li-ion electrode fabrication process. We have been developing a series of particle-based microstructure models using LAMMPS (a particle simulator) [1]. In the most recent model, the governing equations are based on the smooth particle hydrodynamics (SPH) method. The major benefit of SPH is its Lagrangian formulation of the Navier-Stokes equations that would eliminate the need for mesh in our simulation.

We propose a model that imitates the microstructure of a commercial electrode like Toda NMC 523 in terms of composition and active material particle shapes (based on SEM images). We consider each component as a clump of a certain number of particles with a certain size; we also assign density, smoothing length and viscosity as the factors that would affect the interaction between particles.

In this presentation we will discuss the new model, including validation efforts. We use fluid viscosity, interactions between particles, and interfacial behavior as some indicators of our model accuracy. We also will show how the model predicts coating behavior in a slot die coating process. Our hope is to also use the model later to predict behavior during the drying and calendering steps, leading to the final microstructure of the film. The completed model will be used to explore the factors that create such microstructure and how we can control them to manufacture better batteries.

Figure 1: Simulation box model of slurry containing Toda 523; carbon domain is red, active material is yellow, and solvent particles are blue.

Reference

[1]	M. Forouzan et al., J. Power Sources 312, 172-183 (2016).