Abstract

Chemical mechanical polishing (CMP) is a widely utilized technique for achieving ultra-smooth surfaces, yet a comprehensive understanding of the surface microtopography evolution remains elusive due to the intricate chemical and mechanical interactions inherent in chemical mechanical polishing. Accordingly, this study explores the evolution law of different spatial frequency features of surface microtopography of monocrystalline silicon by analyzing the changes in the power spectral density (PSD) curve. More importantly, the evolution mechanism of the chemical mechanical polishing surface micromorphology is revealed through the analysis of the power spectral density under different mechanical and chemical conditions. The results emphasize the noticeable impact of both mechanical and chemical actions on the evolution of different spatial frequency features in surface microtopography. The mechanical action of a large particle, less likely to contact the valley region with a small curvature radius, enhances the smoothing speed of high-spatial frequency roughness. Moreover, the mechanical action of polishing pad demonstrates a greater efficiency in smoothing mid-spatial frequency roughness when utilizing a hard pad. This is attributed to the limited impact of pad hardness on the removal depth of single particles within the reaction layer, rendering it less significant for high-spatial frequency roughness. Upon elevating the chemical reaction rate, a notable acceleration in the reduction of roughness in both mid- and high-spatial frequency regions is observed. Furthermore, insights derived from molecular dynamics (MD) simulations of monocrystalline silicon and quartz glass show that the manner of surface atom removal, whether through single atom removal or cluster removal, exerts a significant effect on the final high-spatial frequency components of the power spectral density. This study contributes valuable insights into the intricate formation mechanism of a smooth surface in chemical mechanical polishing.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call