Lei Zhang

We propose a curvature-profile optimization method for designing compact, ultra-low-loss 90° silicon nitride (SiN) waveguide bends in the O-band. The curvature profiles of the inner and outer bend boundaries are independently parameterized using cubic spline interpolation with control points defined at fixed normalized arclengths. This approach enables simultaneous, flexible engineering of local curvature and waveguide width, eliminating the constraints of predefined analytical functions. By integrating particle swarm optimization with three-dimensional finite-difference time-domain simulations, we achieve insertion losses of 24.25, 3.31, 1.02, and 0.63 mdB per 90° bend for effective bending radii of 15, 20, 25, and 30 μm at 1310 nm, respectively. Broadband performance evaluations across 1260–1360 nm and fabrication tolerance analyses further demonstrate the practical feasibility of the optimized bends. This strategy provides a versatile framework for the design of compact, low‑loss bends for photonic integrated circuits.

Computational spectrometers have garnered significant interest owing to their robust tolerance to fabrication imperfections and potential for compact, cost-effective spectral sensing. However, most existing schemes rely on stochastic design strategies to produce random spectral responses, which often underutilize the device’s inherent performance limits. In this work, we propose an inverse design methodology for computational spectrometers leveraging a modified particle swarm optimization (PSO) algorithm. Using cascaded add-drop microring resonators (MRRs) as fundamental filtering units, we develop a global optimization framework to tailor the spectral responses of the complete MRR filter bank and suppress both auto-correlation and inter-channel cross-correlation. Compared with randomly generated designs, the proposed approach reduces the average auto-correlation HWHM from 0.32 nm to 0.08 nm, demonstrating enhanced spectral encoding capability. Fabricated on an 8-inch silicon nitride (SiN) platform via 180-nm photolithography, the spectrometer achieves a spectral resolution of 80 pm across the C-band, validating the efficacy of the proposed design framework.

Optical phased arrays (OPAs) have emerged as promising beam steering technologies for advanced spatial perception and free-space optical (FSO) communications. However, inherent fabrication errors inevitably introduce stochastic initial phase variations, which must be precisely calibrated to enable high-quality beam steering. In this study, we propose an improved stochastic parallel gradient descent (SPGD) algorithm, integrated with Nesterov-accelerated adaptive moment estimation (Nadam) and termed NASPGD, to achieve rapid and robust initial phase calibration. The proposed NASPGD algorithm was experimentally validated on a 32-element OPA chip. Comparative analysis against conventional SPGD, Adam-SPGD, and other modified SPGD variants demonstrates the superior performance of our approach. Notably, NASPGD effectively suppresses noise interference and escapes local optima, achieving a higher side-lobe suppression ratio with a significantly reduced number of iterations.

Fully integrated photonic continuous-variable quantum key distribution is a promising route toward compact and scalable secure optical links, yet the achievable transmission distance has been limited by the performance of integrated receiver chips and excess noise suppression methods. Here, we demonstrate a Gaussian-modulated continuous-variable quantum key distribution system with an integrated silicon photonic receiver, operating over 60 km. Enabled by a high-clearance, broadband silicon photonic receiver, in conjunction with a robust digital signal processing framework that utilizes a time-domain superposition algorithm and secure dynamic single-tap equalization, the system effectively achieves an asymptotic secret key rate of 1.68 Mbps, and a finite-size secret key rate of 0.80 Mbps at a data block length of 1.55×10 ⁹ . This work extends chip-based continuous-variable quantum key distribution to metropolitan-scale distances, confirming the viability of integrated receivers for large-scale quantum networks.

We demonstrate a 4-channel O-band (de)multiplexer based on silicon nitride (SiN) Mach–Zehnder interferometer (MZI) lattice filters. By utilizing geometry-dependent modal behavior to tailor the effective thermo-optic coefficient (d n _eff /d T ) of the SiN waveguides, near-athermal operation is achieved against global temperature fluctuations. The device architecture utilizes Hermite-curve-assisted folded arms to ensure a compact footprint while maintaining high sensitivity to local micro-heaters for power-efficient channel alignment. Fabricated on an 8-inch wafer with 300-nm-thick SiN, the device exhibits a significantly suppressed temperature sensitivity of ~2 pm/K. The integrated phase shifters achieve a low π-phase-shift power of ~30 mW. Comparative analysis with state-of-the-art SiN platforms highlights the superior performance of this design in both passive thermal stability and active tuning efficiency.

Reducing the power consumption of on-chip systems is of critical importance for large-scale photonic integration. Thermo-optic phase shifters, with advantages such as broadband response, CMOS compatibility, and fabrication simplicity, remain widely used in integrated photonics, yet their relatively high power consumption continues to be a major limitation for large-scale applications. Here, we introduce and experimentally demonstrate a mode-cycling topology that enhances thermo-optic modulation efficiency by sequentially converting the optical field among different guided modes, which forces the light to traverse the phase-shifting region multiple times under the same heater. This cycling in the mode domain effectively multiplies the light–matter interaction without enlarging the device footprint. The scheme employs an inverse-designed bi-directional (de)multiplexer and a mode exchanger, which provide compactness, low loss, and improved robustness compared to conventional asymmetric directional couplers. As a proof of concept, we fabricated a 2 × 2 mode-cycling thermo-optic enhanced switch. The measured results show the average insertion losses at about 2.5 dB with a dynamic extinction ratio over 32.7 dB. The device exhibits a half-wave power of approximately 11.9 mW, nearly halving the power requirement compared to a conventional reference switch of similar size. These results confirm the validity of the mode-cycling topology and highlight its potential as a scalable route toward energy-efficient and compact photonic integrated systems.

The silicon nitride (SiN) platform offers a low-loss solution for photonic integrated circuits (PICs). The bent waveguide represents one of the primary sources of propagation loss in such integrated systems. In this work, we present a comprehensive review of recent advances in SiN bent waveguide technologies. We propose a low-loss SiN bent waveguide design integrated with modified Hermite curves. We adopt a waveguide geometry of 800 nm × 300 nm in the simulations, with the material refractive index derived from the LPCVD process. Simulations conducted at a wavelength of 1311 nm reveal that the proposed bends exhibit bending losses of 0.076 dB, 0.021 dB, 0.0055 dB, and 0.0012 dB per 90° bend, corresponding to bending radii of 15 μm, 20 μm, 25 μm, and 30 μm, respectively. Furthermore, the fabrication tolerance and wavelength dependence of the proposed design are systematically investigated to verify its practical applicability.

Vortex electromagnetic waves carrying orbital angular momentum (OAM) enable differentiated scaling of multiple targets within the beam coverage, granting them superior theoretical performance in multi-dimensional multi-target staring imaging. Owing to the approximate dual relationship between OAM modes and target azimuth angles, multi-mode vortex scanning can rapidly and accurately determine spatial azimuth information of targets, significantly enhancing radar imaging resolution and speed in the azimuthal dimension. Recently, multiple factors, including insufficient OAM mode diversity, complex noise environments and low mode purity, collectively hinder the practical implementation of existing vortex imaging algorithms for high-resolution OAM based radar systems. To address these challenges, this article proposes a high-resolution RELAX algorithm for vortex imaging technology based on precise analysis of echo signal components. The method treats echo signals as superpositions of multiple complex sinusoidal signals and achieves accurate extraction of target positional information through iterative refinement. A comparative analysis of spectral estimation methods, sparse recovery methods and the RELAX method is presented. Experimental results demonstrate that compared to conventional imaging algorithms, the RELAX algorithm exhibits superior target recovery capability, effectively suppresses spurious targets, mitigates noise interference, and significantly improves image resolution, which is particularly pronounced under limited OAM mode diversity. The experimental findings reveal that under with the OAM modes of [-7,7], the RELAX algorithm achieves a limiting resolution of 6.7cm×19°, with the normalized mean-square error of 0.4557 and the image correlation value of 0.6917. This approach substantially alleviates the inherent dependence on OAM mode quantity in vortex radar systems, providing a robust resolution for practical scenarios with constrained electromagnetic (EM) resources. The proposed method notably expands the application scope of vortex EM waves in radar imaging architectures, establishes a solid new technical pathway for vortex radar signal processing, and facilitates the practical deployment of vortex radar.

The increasing demand for miniaturized and portable spectrometers across diverse applications drives the development of computational spectrometers based on algorithmic reconstruction. This work presents an integrated computational spectrometer fabricated on a silicon-on-insulator (SOI) substrate. The device utilizes six cascaded tunable add-drop micro-rings, each with two tuning states, providing 128 distinct filtering channels within a compact footprint of 0.62 mm². Experimental results demonstrate successful spectral reconstruction within the 1525–1565 nm band, achieving a resolution of 40 pm. The average power consumption of the device is 140 mW. © 2026 Chinese Optics Letters

This Letter proposes a hybrid digital-analog microwave-photonics beamforming architecture, which is highly suitable for space-ground integrated network communications due to its high scanning speed and precise pointing capabilities. The frontend digital module generates phase-coded microwave waveform at different slopes, while the backend analog module aligns the beamforming using fixed optical true-time delays to achieve beam pointing in various directions. Beam switching is controlled solely by digital encoding, enabling an arbitrary angle switching time of less than 5 ns. Experiments demonstrate that the beam pointing error is less than 0.70° for four phase slopes.