STUDY OF SLURRY SELECTIVITY AND END POINT DETECTIONS IN Cu- CMP PROCESS

摘要

The chemical mechanical polishing (CMP) process end point detection is one of the key aspects of effective process characterization. Due to the inherently deficient knowledge of the exact working of CMP process, experimental methods have been widely used for CMP process characterization. The CMP bench top tribometer is used as metrology and characterization tool to optimize the Cu CMP process. Here three different candidate materials for Cu damascene process namely, Cu (metal), SiLK~(TM) (dielectric) and Ta (barrier) are polished in three different slurries, where one is Cu selective, the other one is Ta selective and third one is non-selective. The high frequency acoustic emission (AE) signal and the dynamic coefficient of friction (COF) during the CMP process are monitored in situ. The change in the tribological properties of the candidate materials can be used for effective end point detection, while the difference in the material removal rate of the testing specimens in different slurries is used to estimate the slurry selectivity.