Modeling and experimental demonstration of high-throughput flow-through spatial atomic layer deposition of Al2O3 coatings on textiles at atmospheric pressure

Abstract

Atomic layer deposition (ALD) shows promise for forming thin films on temperature-sensitive materials, such as polymers, for applications in filtration, sensing, etc. However, traditional batch ALD generally proceeds slowly and requires controlled, low-pressure equipment. One approach to combat this limitation is spatial ALD, which uses moving substrates through zones of reactant exposure. To date, studies of spatial ALD have primarily explored growth on planar and nonporous substrates. Here, the authors demonstrate a proof-of-concept atmospheric pressure flow-through spatial ALD reactor specifically designed for through-porous substrates, such as fiber webs. This paper describes detailed gas flow modeling and experimental analysis of their prototype reactor. Model results identify precursor gas flow rates, channel spacing, and the distance between the substrate and reactor surfaces as key factors to achieve uniform deposition. Using a previously developed surface wetting protocol, the authors experimentally verify operating conditions for uniform ALD alumina on polypropylene as a model fiber substrate. Under good ALD conditions, the spatial ALD reactor can complete ∼60 cycles/min over a large substrate area, which is 60 times faster than batch ALD. The authors quantify growth saturation conditions and find that under reduced gas flow rates or slow fiber translation speeds, a transition from ALD to chemical vapor deposition-like growth can be induced. Additionally, the authors demonstrate that fiber mat properties such as mat density and air permeability play important roles in the penetration depth of the precursors and, therefore, the conditions needed to achieve ALD. Overall, this work demonstrates a proof-of-concept reactor for high throughput ALD on porous substrates, and identifies important design challenges and considerations for future high-throughput ALD.