Proposal of hybrid deep learning systems for process and material design in thermal nanoimprint lithography

Abstract

A hybrid smart process and material designing system for nanoimprint is proposed based on combination of deep learning system and numerical simulations. The system consists of deep learning system based on experimental data base and simulated table data base. The system complements the missing information with each other, and suggests better processes condition and/or materials. We demonstrate process and material design for low temperature nanoimprinting process using glycol contained polyvinyl alcohol.