An effective SPC approach to monitoring semiconductor manufacturing processes with multiple variation sources

In this research, we develop an integrated sampling and statistical process control (SPC) strategy for semiconductor processes with multiple variation sources. We first construct a process model to characterize the complex nature of semiconductor processes. Three types of variations: among-site, among-zone and among-batch variations, are considered in the model. Based on this process model and rational sub-grouping techniques, multivariate. T/sup 2/ control charts are then proposed to monitor the process variations. It is shown that the proposed control charts are more effective than conventional charting techniques in detecting various types of process excursions.