Feature scale modeling of pulsed plasma-enhanced chemical vapor deposition

Abstract

A two dimensional feature scale model has been developed to predict film profile evolution during pulsed plasma enhanced chemical vapor deposition (PECVD). The model employs a diffusion reaction framework with a spatially variable Knudsen diffusion coefficient to account for molecular transport in evolving feature geometries. The transient model accounts for the alternating atomic layer deposition (ALD) and PECVD growth modes that contribute to deposition. A scaling analysis is used to account for diffusion limitations that affect the ALD component in high aspect ratios due to insufficient precursor exposure. The model predictions have been validated using cross section micrographs of nanostructures produced by pulsed PECVD of aluminum oxide on patterned substrates. The model provides a tool to design the fabrication of nanostructures with digital precision. By varying the relative contributions of the two modes pulsed PECVD provides an opportunity engineer the degree of conformality to produce unique morphologies that are distinct from either steady state vapor deposition processes or conventional ALD.