A Precise Photolithography Process Control Method using Virtual Metrology

Abstract

Virtual metrology, which enables both precise controllability and economic efficiency for semiconductor manufacturing processes, has recently attracted interest. The current virtual metrology model is mainly based on hypothesis verification methods that depend on engineers' skills and updation of the model in accordance with situation changes in tools and processes. We propose a precise photolithography process control method using virtual metrology. In our method, we extract the current virtual metrology model by data mining, thus making updation unnecessary.