Simulation analysis of resist flow in contact hole shrinkage and its impact on block copolymers

초록

A small size and tight pitch of contact holes (C/Hs) are crucial for achieving high device density and reducing manufacturing costs. Therefore, the increasing cost and limited resolution of C/Hs in extreme ultraviolet lithography (EUV) make shrinking C/Hs using the resist flow process (RFP) a promising technology. In this study, the Surface Evolver method, finite-element method (FEM), machine learning, and deep learning were applied to RFP to develop a physically accurate RFP model. Deep learning and machine learning proved effective for regression and classification in physical optimization problems. Additionally, self-consistent field theory (SCFT) was used to describe the self-assembly of cylinder-forming block copolymers (BCPs) confined in RFP C/Hs to achieve smaller C/H dimensions. A convolutional neural network (CNN) predicted RFP and BCP outcomes with an error margin of less than 5%, making it suitable for practical applications. This research paves the way for improved RFP shrinkage modeling of random C/Hs and the fabrication of smaller C/Hs.

키워드