A novel method for statistical process control of gate oxide and front-end cleans monitoring in a manufacturing environment

A novel approach to monitoring the integrity of front-end cleans and gate oxidation is presented. This approach utilizes difference statistics of the 10-, 50-, and 90-percentile values in the charge-to-breakdown (QBD) distributions. Difference statistics are utilized to achieve improvements in the signal-to-noise ratio of potential issues and to minimize the number of wafers used in the monitoring matrix. Implementation of the analyses in this statistical process control (SPC) methodology enables the origin of issues to be brought to the attention of sustaining personnel without the need for in-depth understanding of gate oxide integrity (GOI) testing, thus making it 100% applicable to a manufacturing environment.