Atomic layer deposition process optimization by computational fluid dynamics

This paper presents a computational fluid dynamics model to optimize atomic layer deposition (ALD) process, in which the temperature, precursor mass fraction, mass flow and pressure have been quantitatively analyzed by combining surface chemical reactions with species transport. Simulation results show that the higher temperature increases the growth rate and accelerates the surface deposition process, yet has little impact on precursor distribution in the entire chamber. Both computational and experimental results reveal that precursor concentration is the critical parameter which affects the cycle time and the precursor mass. The gas velocity, depended by the mass flow rate and chamber pressure, is the determinant factor for minimizing the cycle time. Moreover, quicker diffusion and homogeneous distribution resulted from low pressure and high mass flow rate facilitate the optimization of the ALD process. This quantitative model has been verified by experiments under different fluid conditions, which could provide instructive guidance to optimize deposition process in a large pressure range.