Mechanics of Chemical-Mechanical Polishing for Microelectronic Materials

Iowa State University, Ames IA

Investigators

Abstract

This grant provides funding for the development of fundamental understanding of the mechanics of material removal during Chemical Mechanical Polishing (CMP), which is a necessary process step in fabrication of multilevel metallizations integrated circuits (IC) with sub 0.5 micron line dimensions. The project will utilize a combination of experimentations and analytical efforts. Nanoscale indentation and single grit scratch experiments will be conducted at the laboratory scale using the interference of an atomic force microscope (AFM) tip with the workpiece. The observations from these AFM tests will be synthesized in a mechanistic model of the CMP process. The model will be validated against experimental observations from a full scale CMP operation conducted in collaboration with industrial partners. Both uniform as well as patterned wafers will be tested, and the model will be further refined in view of the feature scale and wafer scale measurements taken during these CMP tests. Detailed parametric study will be conducted, and the validated model will then be utilized to seek new design avenues for further improving the CMP technique. In particular, the effects of various process parameters on the degree of nonuniformity (die scale and wafer scale) of local material removal rate will be investigated. If successful, the research will facilitate CMP technology development in three different avenues: (1) It will aid in identifying the dominant causal relationships between the material and process parameters, and the effectiveness (measured by quality and integrity of the finished surface) and efficiency (measured by material removal rate) of a CMP process. Such a mechanistic description will provide a fundamental understanding of the material removal process in CMP, and facilitate process optimization. (2) The model, through its mechanistic description, will aid in alleviating impediments and will facilitate a much more versatile use of experimentally gathered data for realistic CMP processes. (3) In addition to providing a more fundamental understanding of the CMP process, the value of the experimental and modeling efforts lie in its enhanced capability for exploration of the "design space". Currently, many process design options (e.g. optimum selection of slurry and pad properties) remain a trial and error procedure due to lack of reliable models depicting those effects. The model will aid in identifying such unexplored process parameters, and guide toward new and novel avenues for designing CMP processes. This capability to explore new design avenues may have the potential to provide a new impetus to effective and efficient designs of CMP processes and related equipment.