Multiscale computational fluid dynamics modelling of spatial ALD on porous li-ion battery electrodes

The self-limiting surface reaction characteristic of atomic layer deposition (ALD) makes it ideal for the surface modification of electrode materials for lithium-ion batteries (LIBs). Spatial ALD shows promise as a scalable method for the coating on pre-fabricated electrode sheets. As a strong-coupled multiscale process, various process conditions and microstructure parameters have great influences on the macroscale fluid dynamics and the pore-scale diffusion–reaction process, thus affecting the coating efficiency. This study presents a multiscale numerical model that combines computational fluid dynamics (CFD) with multilevel pore-scale diffusion–reaction kinetics to explore the spatial ALD process on porous LIB electrodes. The dynamic mesh method is utilized to simulate electrode movement. The considerable active surface-to-volume ratio of the porous structure limits the precursor infiltration depth due to the low diffusion rate and inadequate precursor supply. As the electrode velocity increases, an asymmetric distribution of precursor concentration under the injector is observed with a rapid decrease. Elevating both the precursor concentration and inlet gas velocity augments the coating depth by enhancing the supply of the precursor. The experimental data aligns well with our numerical results, verifying the accuracy of the multiscale CFD model. Our observations reveal that a relatively lower operating pressure, around 0.1 atm, compared to 0.01 atm and 1 atm, optimizes the deposition rate along the electrode depth during the half-ALD cycle, especially when the pore size is larger. Electrode porosity of about 0.4 notably improves coating uniformity by elevating the precursor diffusion rate. Predictions show that with a substrate velocity of 0.2 m/s, the coating depth on an electrode having higher porosity at the top compared to the bottom via atmospheric spatial ALD could reach a depth of 38 μm with a precursor utilization of 78 %.