Multiscale Computational Fluid Dynamics Modeling of Thermal Atomic Layer Etching

Book excerpt: In recent years, increasing demand for microchips have been fueled by the rise of nano scale technologies that are difficult to fabricate. Atomic layer etching (ALE), an emerging etching process, is a promising method that can overcome the challenges that are encountered during the assembly of these nano-scale devices. Experiments may be conducted to investigate chamber configuration designs and optimal operating conditions; however, these experiments can be costly andtime-consuming. Therefore, this work aims to develop a multiscale computational fluid dynamics (CFD) modeling framework to simulate thermal ALE of aluminum oxide thin films to investigate several reactor chamber designs and evaluate their performance under a range of operating conditions. First, a macroscopic reactor model for each reactor design (typical, multi-inlet, shower-head, and plate) is constructed through Ansys software. Next, the macroscopic model is combined with a previously developed microscopic model of the etching process, which is based on a kinetic Monte Carlo (kMC) algorithm. The multiscale CFD model is used to determine the ideal reactor configuration for achieving film etching uniformity while minimizing process operating time and reagent consumption. It is ultimately determined that the reactor with the inclined plate can produce the most conformal and ultra-thin films in the fastest time while also being economical and unwasteful.