Process engineering and theoretical chemistry for ALD

This topic will treat aspects like fundamental selection and evaluation of ALD metal organic precursors as well as the numerical modeling of their behavior in ALD conditions. Numerical modeling could be used to assess their thermal stability at the scale of the molecule/surface interaction or their transport and kinetic properties at reactor scale.

ALD is by essence driven by a surface limited adsorption-reaction scheme. The molecule scale is key to understand the surface-molecule interaction, the adsorption kinetic and the breakup of chemical bonds that leads to solid surface growth. Two different ways of understanding the deposition story are 1) in-situ characterization of surface during growth or 2) in-silico assessment of molecule interaction with surface (ab-initio calculations).

On other the other side of the topic are the technical developments of original ALD reactors and precursors. Reactors depends strongly on the properties of precursors so new reactors and new delivery systems should be developed. For example, DLI ALD (Direct liquid injection) could be used for low vapor pressure reactors, ALE (Atomic Layer Etching) could be used to reverse the process and allow a top down approach and finally spatial ALD could replace the time phasing of recipe by a space phasing. All these approaches could benefit from CFD (computational fluid dynamics) simulation to speed up the physical understanding, the technical development and the industrial up scaling of ALD process.

In consequence, this topic will treat from various aspects like multi-scale numerical modeling (ab-initio, phase field, kinetic Monte Carlo, multi-physics CFD modeling) and like reactor development by the mean of numerical tool or empirical approach. The idea is to link the communities so that theoretical chemistry and chemical engineering work closely to improve current ALD reactor and precursors to converge to good practices in ALD. These good practices must be taught to the whole ALD community through the GdR.

Speaking about good practices in ALD, a final focus will be done on the good empirical methods like design of experiments (DOE). Indeed, numerical simulations cannot solve everything and among many growth parameters in ALD, certain are impossible to control or just assess. Complementarities between DOE and simulation will be emphasized.

Example of optical port protection for in-situ characterization by the mean of the argon gaz curtain : results from numerical modeling (CFD).