A sequential experimentation strategy and response surface methodologies for process optimization

Abstract

The effective implementation of a three-level sequential experimentation strategy has been illustrated for systematically and efficiently optimizing a novel dual-tone lithographic process. Starting from an initial group of seven factors, the design strategy converged on reversal bake temperature, softbake temperature, and develop strength as significantly dominant factors. Through the use of a transformation of variable technique on the final RSM (response surface model) design, three potential process alternatives were located, based on model predictions that simultaneously minimize photoresist loss, optimize resolution, and maximize process latitude. The model validity of the three process alternatives was experimentally verified.<>