Numerical Simulation of Spatial Atomic Layer Deposition on Moving Substrates with Microgrooves and Porous Structures

Abstract

The need of ultrathin and highly conformal coatings on large-area microstructure substrates such as photovoltaic cells, flexible displays, and Li-ion battery (LIB) electrodes has driven the development of high-efficient ALD. Spatial ALD is a promising high-throughput technique capable of producing ultrathin films on large substrates, which can reach up to two orders of magnitude faster than temporal ALD. However, spatial ALD is a complex and strong-coupled process of fluid flow, heat and mass transfer, as well as chemical reactions. Compared to the flat substrate, the diffusion process in the microstructures takes longer, limiting the substrate velocity and the deposition rate. In this work, spatial ALD on moving microgroove substrates and porous LIB electrodes are quantitatively studied through the multiscale modelling method by coupling computational fluid dynamics (CFD) with chemical kinetics. Dynamic mesh method is implemented to simulate the ALD reaction with the in-line movement of the substrate. Results show that the vortices always exist in the micro-gap with microgroove substrates. Besides, the top of the microstructure on the first contact with the precursor is the highest, and the surface coverage at the corner on the other side of the microstructure is the lowest. For porous LIB electrodes with large specific surface area, the film coating depth is limited by both the low precursor diffusion rate and the amount of precursor supply. Increasing the carrier gas flow rate increases the film coating depth by increasing the precursor supply over the same time period. Besides, the gas byproduct of ALD remaining at the electrode bottom can be more effectively removed. Quantitative process optimizations of substrate velocity and precursor concentration for different electrode structures are also carried out. It is found that the electrode gradient design with higher porosity on the top is beneficial for obtaining deeper coating depth and higher precursor utilization compared to the inverse gradient design. The present multiscale modeling work provides a comprehensive understanding and important guidance for the industrial application of the spatial ALD process on microstructure substrates.