ABSTRACT As electromagnetic detection systems evolve toward multispectral integration and 0°–60° wide‐angle coverage, conventional multispectral stealth technologies suffer severe performance degradation at oblique incidence. Specifically, the 10.6 μm laser absorption peak shifts drastically at 30°, and X‐band microwave reflectivity exceeds 50% at 45°, failing stringent engineering demands. This work presents a wide‐angle‐compatible, flexible, multispectral stealth metasurface for visible light, laser, and microwave. It comprises a visible light camouflage layer (VCL), wide‐angle laser stealth layer (WLSL), and oblique‐incidence microwave absorption layer (OMAL). Large wavelength disparity across the three bands enables complete functional decoupling among layers. A resonance‐pre‐biased asymmetric Fabry–Pérot (F‐P) cavity maintains laser reflectivity below 20% at 60° incidence. A high‐duty‐cycle frequency selective surface (FSS) with double‐layer indium tin oxide (ITO) square rings eliminates the microwave transmission barrier from the WLSL, achieving > 85% absorptivity at 60°. Thin‐film interference modulates optical properties to match natural backgrounds for visible camouflage. Fabricated on PI/PC/PET flexible substrates via electron beam evaporation and laser etching, it offers conformal application potential and technical feasibility for batch manufacturing and cost optimization. This work breaks the angle‐sensitivity limitation of conventional designs and provides a feasible solution for wide‐angle multispectral stealth.
ABSTRACT Practical high‐Q metasurface sensors face two bottlenecks: maximizing analyte near‐field coupling while suppressing radiative loss and ensuring compatibility with low‐cost large‐area fabrication. Here, we present a nanoimprinted tetramer architecture (NTA) sensor fabricated via two‐step nanoimprint lithography, consisting of polystyrene nanopillar tetramer arrays atop a PS waveguide and gold mirror, supporting a prominent resonance mode at 810 nm (D 1 mode) exhibiting a pronounced Fano line shape with a high Q of 3124. In contrast, the experimentally measured resonance of the fabricated device exhibits a FWHM of approximately 12 nm, corresponding to a Q factor of about 70. Multipole decomposition confirms that the D 1 mode is governed by toroidal dipoles, with suppressed conventional electric/magnetic multipoles, thereby achieving radiative cancellation via multipole manipulation. Measured in aqueous glycerol solutions, the sensor achieves a refractive index sensitivity of 501 nm/RIU and a figure of merit of 42 RIU −1 . Label‐free hemoglobin biosensing yields a sensitivity of 354.7 nm·(μg/L) −1 . Benefiting from tetramer symmetry, the resonance response remains stable across 0°–90° polarization angles. This work develops a scalable low‐cost strategy for toroidal‐dipole‐enhanced high‐Q biosensors, promising high‐sensitivity detection for medical diagnosis and environmental monitoring.
ABSTRACT Tailoring light fields at remote or confined locations is a key requirement for many emerging photonic technologies. Here, a multicore fiber functionalized with a 3D‐printed hologram for the reconfigurable generation of structured light distributions is demonstrated. A computationally optimized 3D‐printed phase‐only hologram maps the fields emitted by the individual fiber cores onto prescribed output field distributions. By exciting different combinations of cores, these distributions can be combined into various reconfigurable light patterns. The concept is demonstrated by generating seven‐segment digits, which clearly illustrate the possibility of programmable pattern formation. The hologram is fabricated by high‐resolution 3D nanoprinting on a spacer integrated with the fiber facet, enabling beam expansion before wavefront shaping. Systematic optical characterization confirms reliable image formation and shows that the generated patterns are robust to variations in the relative phases between cores. The approach provides a compact and remotely addressable route to structured illumination, with potential applications in life sciences, telecommunications, and integrated photonics.
ABSTRACT Integrated 90° optical hybrids are essential passive components in coherent receivers, yet their quadrature accuracy and power balance are highly susceptible to fabrication‐induced deviations. Here, we demonstrate postfabrication phase trimming of silicon‐on‐insulator 4 × 4 multimode‐interference‐based 90° optical hybrids using localized femtosecond‐laser modification. The hybrid is implemented with a shallow‐etched rib multimode section and characterized by mapping the relative output phase onto wavelength‐dependent interference spectra. As‐fabricated devices show phase deviations exceeding the ± 5° target and exhibit device‐to‐device performance fluctuations among nominally identical structures. Raman spectroscopy and high‐resolution transmission electron microscopy confirm that femtosecond‐laser irradiation induces a localized crystalline‐to‐amorphous transformation in silicon, introducing a permanent refractive‐index perturbation. By selectively modifying the MMI region, the modal phase evolution and output phase balance can be adjusted after fabrication. After trimming, the phase deviations are reduced to within ± 5° across the full C‐band, accompanied by recovery of the quadrature response, while maintaining common‐mode rejection ratio above 20 dB. These results demonstrate a permanent and power‐free calibration route for coherent silicon photonic circuits.
ABSTRACT Volumetric meta‐optics have emerged as a compact high‐performance platform for advanced imaging, color routing in image sensors, and on‐chip wavefront control. Realizing such devices depends on solving a high‐dimensional nonconvex inverse‐design problem governed by Maxwell's equations. Adjoint optimization is the standard route to gradients, yet broadband three‐dimensional (3D) design typically requires thousands of full‐wave forward/adjoint simulations, making industrial‐scale optimization prohibitively slow. To address this bottleneck, a neural adjoint method is introduced, in which a solver‐supervised Fourier neural operator (FNO) directly predicts dense, per‐voxel 3D adjoint gradient fields from a voxelized permittivity volume. To preserve the sharp, resonance‐driven sensitivity peaks that single‐pass operator networks tend to oversmooth, a stage‐wise FNO is proposed, in which successive stages refine residual errors with progressively higher‐frequency emphasis. A domain‐specific dataset is curated from paired forward/adjoint finite‐difference time‐domain (FDTD) simulations spanning three volumetric tasks: spectral sorting (color routers), achromatic focusing (metalenses), and waveguide mode conversion. Across all benchmarks, the surrogate‐driven optimization attains figures of merit within 86%–99% of rigorous FDTD‐based adjoint optimization whilst reducing end‐to‐end design time from hours to seconds, with speedups exceeding . These results establish a pathway toward fast large‐scale volumetric meta‐optical design enabled by neural‐operator surrogates.
