Terahertz technology holds immense potential for next-generation high-speed wireless communications; however, challenges it faces, such as incompatibility with complex surfaces, restrict its range of applications. Conformal flexible metasurfaces offer a promising solution for dynamic wavefront control in complex environments. Herein, we propose a multifunctional terahertz wavefront manipulation device based on frequency-polarization multiplexing using cascaded conformal flexible metasurfaces. The device consists of a frequency multiplexing layer with triple ring resonators, a polarization multiplexing layer with concentric cross-ring resonators, and an intermediate metal layer forming a Fabry–Perot cavity. By leveraging optical path difference (OPD) compensation and conformal coordinate transformation (CCT) for phase design, the device achieves independent wavefront modulation in four channels under different conformal states: two frequency channels and two orthogonal polarization channels, with respective bandwidths of 0.11 THz, 0.15 THz, 0.14 THz, and 0.17 THz. This work offers a feasible approach to broadening application scope of flexible metasurfaces.
Adaptive optics (AO) performance is limited by interaction matrix mismatch induced by a lateral shift between the deformable mirror (DM) and wavefront sensor (WFS). We propose a phase-domain interpolation-based interaction matrix calibration method that directly maps shift-induced phase variations into a slope-domain matrix calibration form without full remeasurement. The proposed phase-domain interpolation preserves sub-aperture-scale spatial continuity and numerical accuracy. For lateral shifts up to 2.3 sub-apertures, numerical results show reductions of 61.5% in mean position error and 87.6% in normalized root-mean-square error. Simulations with different DM-WFS configurations and closed-loop experiments on a real adaptive optics system confirm stable closed-loop performance. The method offers an efficient phase-domain-guided interaction matrix calibration approach under lateral shifts, providing a potential basis for rapid matrix calibration in more complex AO configurations.
Brillouin dynamic grating (BDG) optical time-domain reflectometry (OTDR) with a chirped probe pulse enables high-speed dynamic birefringence mapping in polarization-maintaining fibers. However, the underlying interaction between a chirped pulse and a BDG cannot be well interpreted within the conventional spectral interpretation developed from conventional four-wave mixing under monochromatic wave assumption. Here we show that the idler wave arises from Fresnel-type coherent accumulation of distributed reflections carrying a chirp-induced quadratic phase, leading to an Euler-spiral phasor trajectory in the complex plane. This framework yields a chirp-rate-dependent effective interaction length L eff ∝1/ C that governs sub-pulse spatial response, together with an oscillatory idler-intensity response versus pulse duration. Both features are verified experimentally. Coherent detection recovers the complex idler field, allowing direct extraction of the local idler frequency set by the local BDG resonance condition rather than the instantaneous probe frequency. The model is experimentally verified over 800 m of fiber through single-shot retrieval of the absolute birefringence profile, with sub-pulse spatial resolution and a measured variation of 5.8×10 −6 . These results establish a general physical picture for chirped-pulse interrogation of dynamic gratings and open the path to real-time monitoring of polarization-maintaining fibers in gyroscopes, structural health, and quantum networks.
Underwater optical communication is essential for high-efficiency ocean information exchange and coordinated operation of marine systems. In this work, we propose a high-capacity underwater communication technique based on photon-level single-pixel image-free sensing. A dual-layer physical encryption framework integrating spatial encryption coding and dual-channel signal superposition is constructed, achieving synergistic optimization of environmental robustness and physical-layer security. To address the information loss and latency inherent in the reconstruction-then-recognition pipeline, an image-free 2D-code direct extraction network is designed, effectively simplifying the decoding chain. Meanwhile, to suppress the strong image fluctuations caused by photon statistics, physics-inspired branches are designed to guide spatial energy concentration and structural constraints, enhancing the signal-to-noise ratio of the 2D-codes. The system can maintain a low bit error rate and a stable communication performance even under extremely low compression ratios and photon-counting conditions. Experimental results demonstrate that the proposed method achieves high recognition accuracy and robustness under low-photon conditions or through dynamically turbid underwater environments. At a transmission distance of 50 m, a communication bit rate on the order of 10 4 bit /s and a bit error rate on the order of 10 −5 are achieved, with a preprocessing and decoding time per 2D-code below 0.4533 ms. The results demonstrate a synergistic improvement in photon-level communication capacity and real-time performance, providing a promising solution for underwater optical communication systems.
Highly dispersive transmission gratings for hard X-rays can be used for spectrometers, interferometers, and pulse compressors. Such instruments require both high dispersion and high efficiency. While crystals, with periods set by atomic spacings comparable to the X-ray wavelength, offer the highest dispersion, these have a narrow angular acceptance (and relative bandwidth of acceptance) of less than 10 −3 that can be limiting. Particularly for use with undulator radiation, there is a need for high-efficiency transmission gratings of periods of nanometers and with relative bandwidths of about 1%. Here, we present an experimental characterization of multilayer Laue gratings made to meet this need. These transmission gratings were prepared by first depositing a thick periodic multilayer on a flat substrate and then sectioning it. With complete control of the deposition process, the functionality and application can be tailored with the choice of materials and the variation of period with layer height. Gratings used in these studies were prepared from WC/SiC multilayers with periods of 2.5 nm, 5 nm, and 10 nm and grown as high as 50 μm. The volume transmission gratings presented here show close to theoretical diffraction efficiencies of 57% at a photon energy of 17.5 keV and above 80% at 60 keV photon energy, and operate over relative bandwidth larger than 25% depending on the choice of materials and the period. A qualitative model describing layer imperfections and their effect on grating properties is presented and discussed.
The reliability of photoelectronic imaging sensors is frequently compromised in extreme environments by large-scale clustered physical damage, leading to irreversible information loss that exceeds the restorative capabilities of traditional interpolation-based algorithms. Here, we propose a physics-driven end-to-end framework designed to achieve high-resilience and high-fidelity imaging through the synergistic optimization of an optical wavefront coding frontend and a neural reconstruction backend. Grounded in statistical optics, we derive a spectral orthogonality constraint to minimize the overlap between the power spectral density of the system’s point spread function and the bad-pixel mask, guiding the frontend to implement active spatial redundancy encoding. The backend, a cascaded U-Net architecture, performs joint information interpolation and deconvolution under non-blind priors. Simulations and experiments demonstrate exceptional resilience against 100 random 40×40 pixels damage clusters, accurately restoring complex semantic features and supporting high-fidelity full-color reconstruction using a single physical mask. This paradigm shifts from passive hardware dependency to active physical redundancy, providing a robust foundation for imaging in extreme conditions.
Micro-nanophotonic devices have significantly propelled sensitive biosensing. Yet the intrinsic trade-off between quality ( Q )-factor and sensitivity restricts the limit of detection (LOD), as the simultaneous optimization of resonance certainty and response intensity remains constrained. Here, we integrate a plasmonic metasurface within a vertical microcavity, facilitating the out-of-plane nonlocal coupled mode between microscale Fabry–Perot (F–P) and nanoscale plasmonic resonances. Leveraging the off-resonant coupling and constructing the biointerface intracavity, the coupled mode inherits the robust high- Q and bulk sensitivity from the F–P resonance, and the high surface sensitivity from the nonlocal plasmonic mode. Such cross-scale sensing capability is further demonstrated by a coupled mode sensing model. Consequently, the experimental figure-of-merit (FOM) reaches 165 RIU −1 , representing state-of-the-art performance among currently reported plasmonic nanosensors. Meanwhile, the off-resonant coupled mode achieves notable experimental Q -factors (up to 541) at approximately 850 nm and exhibits remarkable robustness, with experimental coefficients-of-variation remaining approximately 10% even under 33% in-plane structural variations. Benefiting from the robust high- Q and sensitivity, the coupled mode possesses a significantly improved theoretical LOD across several to hundreds of nanometers compared to bare lattice resonance or F–P modes, which is corroborated by the label-free detection of pseudovirus particles and small viral proteins.
Developing high-power femtosecond laser oscillators characterized by structural simplicity and cost-efficiency is a long-standing pursuit in ultrafast laser technology. In this work, we demonstrate a high-power SESAM mode-locked Yb:CALGO bulk oscillator by implementing synergistic mechanical and optical thermal management. The system directly generates a record average power of 40.1 W with 340-fs pulses at 74.9 MHz, maintaining near-diffraction-limited beam quality with M 2 <1.27. Furthermore, by optimizing the resonator configuration alongside tailored dispersion management, we achieved a high-peak-power regime delivering pulses as short as 73.3 fs with an average power of 23.8 W ( M 2 <1.12), corresponding to a peak power up to 3.91 MW. To the best of our knowledge, these results represent the highest average output powers ever reported for femtosecond bulk oscillators. This work further demonstrates the exceptional potential of Yb:CALGO crystals for achieving megawatt-level peak power and multi-ten-watt femtosecond ultrafast lasers directly from a compact laser oscillator.
Multispectral compatible stealth of high-value targets has intrigued long-standing interest in response to the rapid development of multispectral detection technologies, and a series of ingenious metasurfaces profiting from the exotic electromagnetic (EM) property has provided exceptional platforms for realizing multispectral compatible stealth. Nevertheless, most existing multispectral compatible stealth metasurfaces still suffer from the drawback of immutable stealth performance, which tremendously hinders their practical application in various complex scenarios. Herein, an inspiring strategy of the programmable coding metasurface (PCM) with a thickness of approximately 0.1 λ 0 is proposed to fulfill dynamic microwave manipulation, low infrared radiation, and high optical transparency. Owing to continuous amplitude dynamic modulation and 1-bit phase dynamic modulation implemented by adjusting the PIN diodes of well-designed meta-atoms, the proposed PCM is capable of independently and dynamically controlling the absorption intensity and scattering direction of EM waves in broad bandwidth. Simultaneously, the average infrared emissivity of the PCM can be reduced to about 0.25 from 3 to 14 μm, attributed to the low infrared radiation of the surface indium-tin-oxide (ITO) structures, and the optical transparency can reach 65.9% at 565 nm due to the design of the copper mesh structure and the selection of the transparent dielectric substrate. Multitudinous simulations and experiments of the proof-of-concept prototype are in accordance with theoretical predictions and corroborate the effectiveness of our methodology. This remarkable paradigm of the PCM shows unprecedented intelligence and integration in multispectral compatible stealth and may also find potential applications in communication, imaging, and other intelligent metadevices.
Second-harmonic spectroscopy is a powerful tool for imaging, sensing, and in situ monitoring, with a broad range of applications in the characterization and analysis of nanomaterials. However, its sensitivity is limited by the inherently low nonlinear conversion efficiency of nanomaterials. This study demonstrates a dual-comb second-harmonic generation (SHG) platform that achieves unprecedented sensitivity through the integration of femtosecond dual-comb spectroscopy with plasmonic array-enhanced SHG from ZnO nanocrystal films. The system enables single-spectrum acquisition within 2.16 μs. When combined with microscopy, the platform performs SHG distribution imaging across 16×81 spatial points within 2.8 ms of sampling time, resolving polarization-dependent features at 1.0 μm spatial resolution. This high-speed, ultrasensitive spectroscopic approach enhances the detection of nanoscale optoelectronic properties.
Conventional optical computing architectures often rely on static physical layers or bulky optical systems, limiting their practicality for compact, cost-sensitive edge applications. We introduce a reconfigurable hybrid opto-electronic processing architecture in which a monolithic device performs voltage-programmable feature encoding. Specifically, we realize a physically reconfigurable optical encoder using an electrically tunable liquid crystal-polymer composite (LCPC). By exploiting the volumetric reorientation of liquid crystal domains via a single scalar voltage control, we instantiate unique random scattering kernels that map the same input to statistically distinct output speckle fields. In a hybrid opto-electronic prototype, we demonstrate that this single optical frontend executes multiple, statistically independent encoding operations, enabling hardware-defined context switching. A unified neural network recovers task labels with similar to 90% accuracy under matched voltages, while crosstalk under mismatched voltages is effectively suppressed, demonstrating native physical functional specificity. The final output of this hybrid system is a class decision (semantic label) rather than a reconstructed image. The output patterns exhibit near-maximal entropy (similar to 7.5/8 bits) and resilience under coarse spatial sampling due to the holographic broadcasting nature of the scattering, enabling privacy-preserving, bandwidth-efficient processing suitable for resource-constrained nodes. We further simulate a diffractive metasurface that optically implements the inverse transformation, demonstrating a route toward an active-passive hybrid pipeline with power-efficient inference. This architecture outlines a route to versatile, low-latency, and physics-native edge computing with potential in real-time photonic co-processors and autonomous systems. (c) 2026 Chinese Laser Press
Laser interferometers used as ultrasound receivers provide an unprecedented technique for ultrasound imaging (USI), eliminating the need for mechanical coupling and freeing the system from environmental constraints. Despite these advantages, detecting low-frequency ultrasound waves remains a significant challenge in such systems due to the existence of intrinsic noise, which limits their applications in industrial inspection and marine surveying. Here, we develop an end-to-end noise suppression scheme that effectively extends the Fourier frequency range of ultrasound detection to approximately 35 kHz. Furthermore, we achieve quantum-enhanced USI with a 4 dB improvement in the signal-to-noise ratio and a 1.37-fold enhancement in imaging contrast compared with the case without squeezed vacuum states at the same power level. The expansion of the ultrasound frequency band and the improvement in detection performance are of great significance for the practical application of USI. In addition, this noncontact detection method can be applied to photoacoustic imaging, magnetoacoustic imaging, and other applications that employ interferometers as ultrasound receivers.
We report the first demonstration of 1.3 μm optical coherence tomography (OCT) at over 100 MHz A-scan rate for highly scattering biomedical imaging. To realize high-speed imaging, a 160 MHz broadband swept source centered at 1.3 μm is constructed by time-stretching a high-coherence supercontinuum source with smooth broadband spectral output, the highest sweeping rate ever recorded for the wavelength range to our best knowledge. Three different stretcher configurations are tested: without amplifier, with a booster semiconductor optical amplifier, and with a bismuth-doped fiber amplifier, each employing suitable single-mode fibers for pulse stretching. To enhance OCT performance, we propose a linear wavenumber space resampling method based on unwrapped phases of interference fringes and perform dispersion compensation via the fractional Fourier transform. The unamplified swept-source OCT system exhibits low noise characteristics intrinsic to the supercontinuum source and a high axial resolution of 5.31 μm, though the sensitivity remains at 61.20 dB. In contrast, both amplified configurations provide output powers exceeding 130 mW, enabling higher sensitivity for cross-sectional imaging of highly scattering samples. The bismuth-doped fiber amplifier provides low-noise amplification while avoiding pronounced sidelobes, leading to 160 MHz swept-source OCT with 9.83 μm axial resolution and 76.64 dB sensitivity, as well as higher imaging quality.
We present a combined experimental and theoretical study of collinear four-wave mixing (FWM) in helium driven by an XUV attosecond pulse and two femtosecond near-infrared (NIR) pulses. The first NIR pulse is synchronized with XUV, enabling the simultaneous excitation of both bright and dark states from the ground state. A second, time-delayed NIR pulse then induces resonant coupling between states of opposite parity, generating full-NIR-cycle oscillations in the transient absorption spectrum. By comparing experimental measurements with simulations based on solving the time-dependent Schrödinger equation and a simplified multi-level model, we systematically disentangle the individual coupling pathways involved in helium. We find that highly excited bright states and light-induced states exhibit long-lived oscillations because the participating states are real bound states with long lifetimes. In contrast, oscillations associated with the 1s4p state decay rapidly due to the absence of an indirect coupling channel involving real states. Light-induced states near 1s2p exhibit both low- and high-frequency oscillations originating from two competing indirect pathways, and the relative amplitudes of these components can be actively tuned by adjusting the NIR pulse intensity. This collinear FWM scheme provides access to more complete information on XUV-driven electron dynamics and can be readily extended to more complex atomic and molecular systems.
Photonic frequency interleaving (PFI) has emerged as a key enabling technique for ultra-broadband signal acquisition, with the potential to overcome channel mismatch and clock jitter that fundamentally limit time-interleaved analog-to-digital converters (ADCs). However, existing PFI schemes face trade-offs among acquisition bandwidth, spectral slicing density, and inter-channel isolation, hindering the concurrent achievement of wide bandwidth and high fidelity. In this work, we propose and experimentally demonstrate a dense photonic-frequency-interleaved (DPFI) ADC architecture featuring ultra-broadband operation and high spurious-free dynamic range (SFDR). By incorporating wavelength-division multiplexing, the proposed scheme greatly increases spectral slicing density and the channel count while maintaining high inter-channel isolation. Restricting each sub-ADC to a sub-2-GHz sampling bandwidth markedly improves SFDR and relaxes the stringent jitter requirement of the sampling clock. In a proof-of-concept experiment, a 24-channel DPFI-ADC achieves a total acquisition bandwidth of 72 GHz and an SFDR of 62.9 dB at 68.38 GHz. The broadband acquisition of an 11.8-GHz linear-frequency-modulated waveform and a 1024QAM signal further validates the application potential of the proposed scheme. To the best of our knowledge, our work demonstrates the first frequency-interleaved-ADC with such massive spectral interleaving and record-high SFDR performance, which state-of-the-art electronic ADCs can only achieve at bandwidths below 10 GHz.
