Abstract

The optimization of high-performance meta-devices is an essential prerequisite to developing meta-optics which are competitive with traditional optical design along standard performance metrics (MTF, WFE, etc.). In this paper we share recent contributions to supercell fabrication robustness and topology optimization efficiency which aid in this goal. Minimum feature size enforcement of arbitrary binary masks allows for global optimization strategies to meet the strict feature size constraints of state-of-the-art nanofabrication processes. Additionally, the introduction of multiple-gradient descent into topology optimization enhances its convergence for planar metasurface design by at least a factor of 2.

© 2023 The Author(s)