Scratch generation probability in chemical mechanical polishing (CMP) process

Abstract

Chemical mechanical polishing (CMP) process is a critical new technology to planarize the wafer surface aided by the combined force of the chemical etching and mechanical polishing. In this process, defectivity due to scratch generation has become more and more important as it influences the wafer product quality, throughput and cost significantly with the technology developments into sub 100 nanometer technology nodes. The overall objective of this research is to investigate the effects of particle agglomeration on scratch generation and predict the scratch generation probability in CMP process. A mechanical model considering smooth wafer surface and varying height pad asperities for CMP process will be selected. The indentation depth is taken as a measurement of the scratch depth. The probability of such indentation depth of the scratch is simulated with evolving pad surface topography and changing slurry particle size distribution. Diffusion Limited Agglomeration (DLA) and Reaction Limited Agglomeration (RLA) model will be developed to describe the slurry particle size distribution evolution. Predictions for scratch generation probability are compared with those two different agglomeration models.