Adjoint Method Supported Topology Optimization for Electromagnetic/Photonic Inverse Design

Abstract

Topology optimization is a powerful computational approach for designing photonic and electromagnetic structures by optimizing material distribution within a given domain. In this study, we explore novel applications of adjoint-based topology optimization, including enhanced substrate design for two-dimensional materials, improved quantum efficiency in photodetectors, and a photonic medium for real-time object classification. We demonstrate a substrate optimization method that increases the absorption of broadband excitations in monolayer MoS2 by 72% and boosts quantum efficiency by 141%. Additionally, we introduce an optimized photonic medium that classifies objects at the speed of light by manipulating wavefront propagation.