Summary: | 博士 === 國立清華大學 === 動力機械工程學系 === 91 === As the number of metal levels and the wafer size increase, the need for global planarity across the wafer becomes more crucial. Chemical mechanical planarization (CMP) is considered the key technique to achieve this goal. Accurate in-situ endpoint detection and monitoring method can reduce the product variance, improve the yield and throughput. During CMP process, the wafer is brought downward against the polishing pad completely. The monitoring of the wafer polishing in such a configuration becomes a difficult task. Many methods have been proposed in the past years, including optical, electrical, acoustical/ vibrational, thermal, frictional, chemical/electrochemical methods and others. Many of them require the rearrangement of the machine configuration and wiring. Some can only be implanted on certain types of machines. An accurate and easy in-situ monitoring and endpoint detection method is desired.
An endpoint in a metal CMP process is considered as the time when the metal film above the dielectric layer is fully removed, only the metal filled in the vias is remained. In this study, a pad thermal analysis model is derived according to the kinematic between the wafer and the pad. The model is capable of predicting the thermal distribution on the pad, and the polishing condition on wafer surface can be better understood. The acoustic emission signal during polishing process is also analyzed. One or two of these methods can detect the endpoint of the metal CMP process.
|