Development of integrated process control system utilizing neural network for plasma etching

Abstract

The purpose of this study is to provide the integrated process control system utilizing neural network modeling, to search for the appropriate choice of control input, and to keep the process output within the desired range in the real etch process. Variations in the process output are classified as the drift and the shift. The drift is caused by a natural noise that changes over a period of time slowly and steadily partly due to the aging of equipment. Although the drift moves the process output away from the target value, its variation is infinitesimal. On the other hand, the shift results in a larger variation due to the various causes and its width is normally greater than that of the drift. Without appropriate procedures, the process output will move away from the target value greatly. Therefore the control strategy is to minimize the process shifts, in which process outputs are measured by monitoring wafers periodically.