Ashing is a photoresist-stripping process using oxygen or hydrogen radicals and is one of key process step in the semiconductor manufacturing processes. Uniform and fast stripping is the key factor in ashing. In this study, a computational fluid dynamics simulation was applied to find conditions for uniform molecular flux over the wafer surface and to optimize the ashing chamber geometry. In particular, the distance between the gas inlet baffle and wafer stage in the 450 mm wafer chamber was determined through inductive inference statistics. To improve the reliability of this simulation, the correlations between the calculated molecular flux distribution and the measured ashing rate distribution over 300 mm wafers were sought first. Effects of the distance between the baffle and wafer stage, wafer stage temperature, and gas flow rate on distributions of molecule flux and velocity, temperature and gas molecule density were calculated. The simulation showed that the density distribution over 450 mm wafer surface was more uniform when the distance between gas inlet baffle and wafer stage was between 35 mm and 60 mm, and that the reactant flux distribution was more uniform when the distance was between 60 mm and 80 mm. Therefore, the distance between the gas inlet baffle and wafer stage was chosen to be 60 mm.
Computational fluid dynamics simulation and performance optimization of an electrical vehicle Air-conditioning system
