Development of a virtual metrology method using plasma harmonics analysis

A virtual metrology technique based on plasma harmonics is developed for predicting semiconductor processes. From a plasma process performed by 300 mm photoresist stripper equipment, a strong correlation is found between optical plasma harmonics intensities and the process results, such as the photoresist strip rate and strip non-uniformity. Based on this finding, a general process prediction model is developed. The developed virtual metrology model shows that the R-squared (R2) values between the measured and predicted process results are 95% and 64% for the photoresist strip rate and photoresist strip non-uniformity, respectively. This is the first research on process prediction based on optical plasma harmonics analysis, and the results can be applied to semiconductor processes such as dry etching and plasma enhanced chemical vapor deposition.