Notice of Retraction Failure Mode Identification Using Thermal Data Based on Mixed Effects Model in Chemical Mechanical Planarization Process

Tao Liu,Xi Zhang
DOI: https://doi.org/10.1109/qr2mse.2013.6625700
2013-01-01
Abstract:Process variables in chemical mechanical planarization (CMP) processes such as pad temperature and coefficient of friction (COF) usually contain rich information and are highly related to process conditions and presented in a functional form, and may characterize the polishing process condition changes. In this paper, we proposed a physically motivated mixed effects model based on these process variables for process failure mode identification by separating this dynamic process into two components: heat transmission model compliant with physical laws and random effects model by latent factors. This model has been demonstrated the efficiency in distinguishing different polishing conditions through our real experimental validation.
What problem does this paper attempt to address?