Ion implant data log analysis for process control and fault detection

Data mining techniques have been introduced to the semiconductor industry in recent years. In this paper, we report on progress in developing a network based implant data log (IDL) data analysis system which can be used to access data from multiple tools and multiple recipes. The system that is currently under development can be used to generate individual control charts from any of over 100 process variables for a user selectable process recipe or from all implants in the database. Multi-variable models are being developed to compare relationships among process variables. These models calculate predicted values of process variables, and the differences between the model and the actual variables are used as indicators of process drift, hardware malfunctions, recipe integrity, and in some cases mis-processing.