Optimal High Efficiency 3D Plasmonic Metasurface Elements Revealed by Lazy Ants

Abstract

Recent transmissive optical metamaterials that leverage a generalized form of Snell's law to induce an anomalous refraction of light have garnered considerable interest in both optical and materials communities. However, most of these designs have primarily centered around parametric studies of planar canonical structures for their low profile and relative ease of manufacturing. In many of these cases, all-dielectric designs are preferred over metallodielectrics due to their low loss characteristics. Moreover, considering modern advances in nanofabrication techniques, these canonical structures represent only a small portion of the design space that is explorable. In this work, we exploit a generalized Multi-Objective Lazy Ant Colony Optimization (MOLACO) algorithm and a modified Pareto locus search mechanism to optimize arbitrary three-dimensional metamaterial unit cells in the optical regime based on the Membrane Projection Lithography technique. Our exploration has revealed unintuitive metallodielectric structures for phase-gradient metasurface applications in the midwave infrared (MWIR) regime that achieve transmission magnitudes comparable to the highest-performance all-dielectric designs found in the literature. As a proof-of-concept, a beam-steering metasurface is synthesized using these unintuitive unit cell geometries and is shown to achieve over 84% diffraction efficiency, which is among the highest performing metallodielectric metasurfaces in the MWIR reported to date.

Numerical Optimization Methods for Metasurfaces

“…Various devices have been recently optimized using GA or using some advanced evolutionary strategies as shown in refs. [53][54][55][56]61]. To highlight the basic concepts of the numerical optimization methods and mention briefly some possible applications, the readers can refer to recent review works about metasurfaces and nanophotonics (refs.…”

“…[64] Recently, some advanced evolutionary strategies have also been extended to design multifunctional metasurfaces. [53,65,66] For the GA approaches presented here, all require fullwave calculations of Maxwell's equations for all devices in each generation, which adds up to at least a few thousand of simulations. Therefore, this technique is computationally expensive and has to be used with efficient fullwave solvers, especially when considering the optimization of 3D structures.…”

“…[21–23] for a broad overview of optimized configurations based on GA or, more generally, on advanced evolutionary strategies as in refs. [53–57]. a) An all‐dielectric metasurface configuration made of Si nanodisks.…”

“…[ 64 ] Recently, some advanced evolutionary strategies have also been extended to design multifunctional metasurfaces. [ 53,65,66 ]…”

Design for quality: reconfigurable flat optics based on active metasurfaces

“…In the optical and IR regimes, these nanoantennas are usually limited to simple canonical shapes such as triangles, rectangles, rings, cylinders, and their combinations. However, it has been shown that nanoantennas based on unintuitive shapes (i.e., free-form geometries) can significantly outperform more traditional geometries in various metrics, such as maximum transmission or reflection magnitudes, phase coverage, field of view, and spectral bandwidth [283,284]. When paired with active materials, metasurfaces comprised of unintuitive nanoantenna shapes offer the potential to achieve major At .-μm wavelength with   cm − hole concentration.…”