(623h) Multiscale Computational Fluid Dynamics Modeling of Spatial Atomic Layer Etching for Aluminum Oxide Thin Films

With the global demand for semiconductor devices skyrocketing, current semiconductor fabrication methods have been unable to produce an abundant supply to satisfy these demands. Atomic layer etching (ALE) and deposition (ALD) are time-consuming processes in the production of these devices where sequential doses of precursors are used to produce highly conformal thin films in a batch reactor system for nanochip devices. With the majority of the time associated with the purging cycles during these processes, spatial atomic layer etching (SALE) and deposition (SALD) processes have emerged to reduce the overall process time while maintaining the conformity of thin films by functioning continuously as opposed to conventional ALD and ALE processes [1,2]. Despite the emergence of SALE and SALD processes, spatial reactors have not been fully investigated, resulting in skepticism in industry.

This research constructs a multiscale computational fluid dynamics (CFD) model using a dynamic mesh to characterize the SALE of aluminum oxide thin films and determine optimal conditions to better enhance the performance of the SALE process. First, a microscopic scale model [3] based on a kinetic Monte Carlo algorithm for the kinetic surface domain is adopted from previous work. Next, a macroscopic-scale model for the gas-phase domain is established by combining the CFD simulation with the microscopic model, leading to a multiscale mathematical model. Finally, the multiscale CFD model is utilized to optimize the operation of spatial reactors and analyze the effects of process parameters on the quality of the thin film. This research will provide guidance for expanding the research of spatial reactor design and operation.