General-Purpose Algorithm for Two-Material Minimum Feature Size Enforcement of Freeform Nanophotonic Devices

A new iterative minimum feature size enforcement algorithm is presented for directly coercing two-material (i.e., fill and void regions) minimum length scales on freeform binary masks. The method is flexible, allowing for independent control over minimum widths, gaps, areas, and enclosed areas. Curvature constraints are automatically enforced in the process. A supercell optimization study is conducted using the method to demonstrate the relevance of the algorithm to state-of-the-art optical meta-lens design and fabrication. The proposed method can be used to characterize how freeform metasurface performance upper-bounds will decrease as minimum feature size enforcement increases. We show how this process is a critical prerequisite for realizing high-performance wafer-scale optical metasurfaces. Moreover, because this method can be employed to directly modify a binary mask, it is suitable for use with global and multiobjective optimization methods, making it more general purpose than existing methods used in topology optimization.