Modeling atomic layer deposition process parameters to achieve dense nanocrystal-based nanocomposites

Atomic layer deposition (ALD) is a technique capable of depositing conformal coatings in highly tortuous 3D nanostructures. One configuration that has attracted attention is nanocrystal (NC) based nanocomposite films, whereby a 3D network of randomly packed nanocrystals is infilled via ALD to yield a dense nanocomposite. In this work, we demonstrate criteria for predicting three important thermal ALD process parameters necessary to completely infill 3D NC networks: cycle number, precursor pulse time, and purge time. A description of representative pore geometry is developed using parameters of the film comprised of nanocrystals before infill, specifically NC diameter, NC volume fraction, and film thickness. This geometric description allowed for prediction of required precursor pulse times to saturate the NC film surface. A finite-difference model of water vapor transport during purging revealed that desorption kinetics can be used to predict purge times required to achieve complete infill. The model predictions show good agreement with experiments carried out by infilling films comprised of GaN NCs with ZnO by the diethylzinc/water process and films comprised of Al2O3 NCs with Al2O3 by the trimethylaluminum/water process.