Multiobjective Statistical Learning Optimization Of Rgb Metalens

Modeling of multiwavelength metasurfaces relies on adjusting the phase of individual nanoresonators at several wavelengths. The traditional procedure neglects the near-field coupling between the nanoresonators, which dramatically reduces the overall diffraction efficiency, bandwidth, numerical aperture, and device diameter. Another alternative design strategy is to combine a numerical optimization technique with full-wave simulations to mitigate this problem and optimize the entire metasurface at once. Here, we present a global multiobjective optimization technique that utilizes a statistical learning method to optimize RGB spherical metalenses at visible wavelengths. The optimization procedure, coupled to a high-order full-wave solver, accounts for the near-field coupling between the resonators. High-numerical-aperture RGB lenses (NA = 0.47 and 0.56) of 8 and 10 mu m diameters are optimized with numerical average focusing efficiencies of 55% and 45%, respectively, with an average focusing error of less than 6% for the RGB colors. The fabricated and experimentally characterized devices present 44.16% and 31.5% respective efficiencies. The reported performances represent the highest focusing efficiencies for high NA > 0.5 RGB metalenses obtained so far. The integration of multiwavelength metasurfaces in portable and wearable electronic devices requires high performances to offer a variety of applications ranging from classical imaging to virtual and augmented reality.