Indexed by:
Abstract:
Conventional forward design strategies for broadband achromatic metalens usually suffer from exact physical knowledge and high fabrication requirements of complicated meta-atoms structures. Recently, the inverse design aims at automatically optimizing meta-optics design but has been currently limited by expensive brute-force numerical solvers and complex structures. Therefore, it is essential to develop a time-efficient design method that meets the requirements of high-performance metalens while maintaining structural simplicity. Meanwhile, there are relatively few researches in the field of mid-infrared meta-optics which possesses extensive application potential. In this work, we propose an end-to-end inverse design strategy for designing mid-infrared broadband achromatic polarization-insensitive metalenses utilizing elliptical meta-atoms. The framework comprises a high-performance forward prediction network as a fast approximate solver and a global multi-objective optimization model to achieve an end-to-end design. As a proof, we designed a broadband achromatic and polarizations-insensitive metalens operating at 4–5 μm using elliptical meta-atoms with only two design degrees of freedom. The average focusing efficiency is 61.37% with a maximum focal length shift of only 3.56%. The numerical aperture (NA) of the proposed broadband achromatic metalens is 0.178. It is worth mentioning that by using this framework, we can utilize elliptic meta-atom with structural asymmetry to achieve polarization-insensitive properties. The structure of meta-atoms is simple and easy to fabricate, achieving a balance between complex functions, high efficiency and simple fabrication process. Our framework provides new opportunities for avoiding complicated physical models and enabling the integration of complex functionalities into a single metalens, thereby driving intelligent design of efficient and emerging mid-infrared meta-devices. © 2025
Keyword:
Reprint 's Address:
Email:
Source :
Optics Communications
ISSN: 0030-4018
Year: 2025
Volume: 579
2 . 2 0 0
JCR@2023
Cited Count:
SCOPUS Cited Count:
ESI Highly Cited Papers on the List: 0 Unfold All
WanFang Cited Count:
Chinese Cited Count:
30 Days PV: 0
Affiliated Colleges: