Low-k dielectric material removal rate, which is significantly affected by process factors such as polishing load, wafer carrier rotation, platen rotation speed and pad age, is one of the critical issues in CMP planarization of a dielectric film when concerning productivity, throughputs and stabilization of the process, especially when trying to achieve a target polishing thickness. Scratching is another critical issue in low-k dielectric filmi CMP planarization due to the lower hardness relative to silicon dioxide. This research relates to a methodology for in-situ monitoring of the low-k dielectric material CMP planarization process, specifically for monitoring material removal rate and scratch occurrence, using acoustic emission (AE) sensing technology.
Systematic investigations of CMP process variables on AE signals were carried out in this research. The sensitivity of AE to polishing load, polishing speed, wafer surface roughness (wafer pattern density) and pad roughness were verified. The results showed that, under steady state, the AE rms signal increases with increasing polishing load, polishing speed, slurry particle size, wafer surface roughness and pad roughness.
Based on the research in tribology and other application fields of loose abrasive machining such as lapping and polishing, scratching was known to be caused by the presence of particles which are much larger than average slurry particles. It has been proven that scratching can be avoided or reduced by timely cleaning the slurry supply system. Therefore, to avoid scratching, one strategy is to develop an in-situ method for detecting larger particles involved in CMP process. In this paper, the high sensitivity of AE signals to the presence of larger particles during CMP was experimentally verified.