- Multiobjective statistical learning optimization of RGB metalens hal link

Auteur(s): Elsawy Mahmoud, Gourdin A., Binois Mickaël, Duvigneau Régis, Felbacq D., Khadir Samira, Genevet Patrice, Lanteri Stéphane

(Document sans référence bibliographique)
Texte intégral en Openaccess : fichier pdf

Ref HAL: hal-03212349_v2
Exporter : BibTex | endNote

Modelling of multi-wavelength metasurfaces relies on adjusting the phase of indi-vidual nanoresonators at several wavelengths.The traditional procedure neglects thenear-field coupling between the nanoresonators, which dramatically reduces the over-all diffraction efficiency, bandwidth, numerical aperture and device diameter.Anotheralternative 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 statisticallearning method to optimize RGB spherical metalenses at the visible wavelengths. Theoptimization procedure, coupled to a high-order full-wave solver, accounts for the nearfield coupling between the resonators. High numerical aperture RGB lenses(NA= 0.47and NA= 0.56) of 8μm and 10μm diameters are optimized with numerical average1focusing efficiencies of 55% and 45%, with an average focusing error smaller than 6%for the RGB colors. The fabricated and experimentally characterized devices present44.16% and 31.5% respective efficiencies. The reported performances represent thehighest focusing efficiencies for highNA >0.5 RGB metalenses obtained so far. Theintegration of multi-wavelength metasurfaces in portable and wearable electronic de-vices requires high performances to offer a variety of applications ranging from classicalimaging to virtual and augmented reality.