We report the first measurement of backward stimulated Brillouin scattering (SBS) in generic InP waveguides, supporting enhanced Brillouin gain of g_B/Q_m = 3.5 ± 0.6 W^-1m^-1 mediated by weakly-guided pressure waves in InGaAsP. This leads to SBS gain coefficients as high as 737 ± 54 W^-1m^-1, observed in a mature, foundry-accessible photonic integration platform.

In this work, we report on linewidth improvements of $\mathbf{7 5 0} \boldsymbol{-} \boldsymbol{\mu}$ m-long distributed-feedback (DFB) lasers with embedded thermal shunts on the membrane-on-silicon (IMOS) technology platform. Linewidth improvements from 8 MHz to 152 kHz were measured, which opens up the route towards dense photonic circuits for low-linewidth applications.

Integrated photonics offers a promising route toward low-latency and energy-efficient neural computing. However, the inherently linear nature of most photonic devices makes the realization of compact and efficient optical nonlinear activation functions a key obstacle for scalable photonic neural networks. Moreover, large-scale architectures require monolithically integrated nonlinearities for dense cascading without the losses and complexity associated with hybrid or heterogeneous integration. Here, we experimentally demonstrate monolithically integrated all-optical nonlinear activation units based on quantum-well saturable absorbers (SAs) on a generic InP platform, co-integrated with passive waveguides and electro-optic components. Three nonlinear activation units are realized: a standalone SA, a Mach-Zehnder-interferometer-integrated SA, and a microring-resonator-enhanced SA. By exploiting saturable absorption, resonance, and interference, these devices provide programmable nonlinear transfer functions, including softplus-, sigmoid-, ReLU-, and RBF-like responses. Experimentally measured transfer functions are incorporated into convolutional neural networks for CIFAR-10 and Fashion-MNIST classifications, achieving 89% and 92% accuracies, respectively.

Comprehensive characterisation of geometric design variants for co-planar stripline Mach-Zehnder modulators with consistent electro-optic (EO) bandwidths of >100 GHz across multiple device configurations. Physics-based modelling of the EO elucidates bandwidth trends across the design space.

Recent advancements in silicon photonics have enabled the integration of diverse optical devices. However, efficiently bridging guided modes with free-space radiation remains challenging, as current grating solutions often necessitate intricate 3D structures to enable subtle and independent control of radiation channels in multiple directions. Bound States in the Continuum (BIC) offer a promising solution due to their ability to remain localized within the radiation continuum. In this work, we propose and experimentally demonstrate, for the first time, a method that harnesses quasi-BIC (qBIC) in the evanescent field to build a toolset for guided-radiation interface, enabling controlled switching of radiation channels. To validate the method's capabilities, we implement it in a grating structure and showcase its potential through a 128-channel optical phased array (OPA), which achieves enhanced diffraction control and improved far-field beam quality. The proposed method significantly improves upon traditional grating technologies, offering a promising approach for advanced photonic integration.

Bound states in the continuum (BICs) enable counterintuitive light confinement without radiation loss, providing a powerful foundation for integrated photonic waveguides. However, existing BIC waveguides are predominantly realized through geometry-dependent designs, where the BIC condition is restricted to narrowly defined structural parameters, limiting design flexibility and practical applicability. Artificial optical anisotropy is introduced as a new design paradigm for BIC waveguides. Implemented using subwavelength-grating (SWG) metamaterials, continuously tailorable anisotropy provides an independent degree of freedom for deterministically reshaping the radiative continuum, enabling flexible formation and systematic control of BIC waveguides over a broad design space. Anisotropy-engineered symmetry breaking further enables controllable asymmetric radiation and precisely tailored field leakage. This paradigm transforms BIC waveguides from geometry-constrained structures into an anisotropy-engineered platform, establishing a general framework for programmable radiation engineering and next-generation integrated photonic devices.

We present the design and characterization of co-planar stripline Mach-Zehnder modulators on an InP platform. The co-planar design exhibited 50 Ω impedance with velocity-matched optical and electrical signals. We investigated devices with a range of design parameters to identify optimal configurations for high bandwidths (≈80 GHz) and state-of-the-art data transmission rates (320 Gbit/s). An equivalent circuit model that enables fast and holistic design space exploration is developed and experimentally verified. The model predicts ∼120 GHz bandwidth for optimized modulator dimensions.

We propose a low-threshold silicon photonic diode based on nanocylinder-loaded silicon microring. Benefiting from the small mode volume, high Q factor, and manipulable chirality, the proposed diode achieves a low threshold of similar to -7.83 dBm.

Brain-inspired, neuromorphic devices implemented in integrated photonic hardware have attracted significant interest as part of efforts towards non-von Neumann computing paradigms that use the low-loss, high-speed, and parallel operations in optics. We present here the design and measurements of an all-optical spiking laser neuron that was realized for the first time on a generic InP photonic foundry platform, which may be a practical alternative to other semi-integrated photonic and electronic-based spiking neuron implementations. The measured device demonstrates excitability, nanosecond refractory period, and self-pulsating capabilities. We show that the device offers sufficient optical gain for incoming spikes, an essential feature for directly cascading neurons. In addition, we demonstrate multi-wavelength injection and excitability and discuss the possibilities of combining this device with other on-chip photonic devices. The reported characteristics are an improvement over the state-of-the-art, because the proposed device allows for direct on-chip cascading of neurons and weights, which is essential for future fully connected, multi-wavelength, all-optical photonic spiking neural networks.

This roadmap consolidates recent advances while exploring emerging applications, reflecting the remarkable diversity of hardware platforms, neuromorphic concepts, and implementation philosophies reported in the field. It emphasizes the critical role of cross-disciplinary collaboration in this rapidly evolving field.

This publisher's note contains a correction to Opt. Lett.47, 5397 (2022)10.1364/OL.470365.

We demonstrate programmable all-optical nonlinear functions using a microring resonator integrated with a saturable absorber on a generic InP platform. The device exhibits programmable behavior resembling radial basis, clamped ReLU, sigmoid, and softplus-like functions. Simulations on the MNIST classification using measured responses achieve over 94% accuracy.

We experimentally demonstrate all-optical nonlinear activation functions using standalone saturable absorbers, as well as saturable absorbers integrated into microring resonators and Mach-Zehnder interferometer structures on a monolithic InP platform. These configurations enable programmable nonlinear responses and support both passive and active operation modes. The measured activation characteristics exhibit softplus- and ReLU-like behaviors, suitable for neuromorphic computing. When incorporated into a convolutional neural network for MNIST digit classification, the system achieves over 97% accuracy, highlighting the potential of SA-based photonic components for compact, low-power, and scalable optical neural networks.

In this paper, we present a method that utilizes quasi bound states in the continuum within the evanescent field to manipulate the interface between guided and radiation modes, enabling controlled switching of radiation channels.

Artificial neural networks (ANNs) have become ubiquitous in high-performance information processing. However, conventional electronic hardware, based on the sequential Von Neumann architecture, struggles to efficiently support ANN computations due to their inherently massive parallelism. Additionally, electrical parasitics further limit energy efficiency and processing speed, pushing traditional architectures toward their fundamental constraints. To overcome these limitations, researchers are exploring integrated photonics, leveraging the inherent parallelism of optical devices for more efficient computation. Despite these efforts, most existing optical computing schemes encounter scalability challenges, given that the number of optical elements typically grows quadratically with the computational matrix size. In this work, a compact programmable multimode interference (MMI) coupler on an indium phosphide membrane platform is proposed for realizing a photonic feedforward neural network. MMIs present a unique opportunity to accelerate matrix multiplication processes by exploiting the interference properties of light modes, promising advancements in both speed and energy efficiency. The programmable MMI coupler, comprising four input and three output (4 × 3 MMI) InP waveguides, makes use of hybrid integration of liquid crystals as cladding material, which offers reconfigurability to the MMI structure. Three electrically tunable sections are made to perform parallel multiplication operations. A novel modeling technique is introduced to facilitate effective training and inference operations. Finite-Difference Time-Domain (FDTD) simulations are employed for calculating the optical mode propagation process within the programmable MMI structure. Based on the FDTD results, a compact optical neural network is implemented and assessed on the Iris flower dataset, demonstrating a testing accuracy of 86.67%. This novel MMI device concept offers a promising pathway toward energy-efficient, scalable optical computing systems, contributing to the advancement of next-generation artificial intelligence hardware.

We demonstrate a high-brightness on-chip photon pair source based on a high-Q racetrack-shaped silicon microresonator, achieving a quality factor of approximately 1.75 x 105 and an on- chip photon pair generation brightness of 9.8 MHz center dot mW(-1)center dot nm(-1)

This paper presents a high-sensitivity bimodal waveguide sensor based on a dispersion turning point (DTP), realized through a fishbone-like bimodal waveguide (BiMW) design. Experimental validation confirms a maximum sensitivity of 10666.7 nm/RIU, indicating strong promise for integrated optical sensing applications.

High-speed and energy-efficient optical interconnects critically rely on electro-optical (EO) modulators, whose performance metrics struggle to meet the exponentially increasing demands of the near future. Silicon-organic hybrid (SOH) modulators present a promising solution due to the favorable electro-optic coefficients and fast response times of EO organic materials. However, the waveguide's nature limits the effective interaction between photons and EO materials. Although this interaction can be enhanced by utilizing advanced structures such as slot waveguides and slow-light techniques, new challenges arise, including strong dispersion that compromises bandwidth. In this paper, we propose a novel low-dispersion, slow-light waveguide structure based on a coupled onedimensional photonic crystal slot resonator waveguide (coupled 1D PC SROW). By cascading multiple coupled resonators, the structure creates a low-dispersion, slow-light region within the photonic bandgap. Combining the strong optical field confinement of the slot with the slow-light enhancement in the time domain, modulation efficiency, quantified by VπL, can be significantly improved. As an example, we demonstrate that a VπL of 0.57 Vmm can be achieved for a low-dispersion wavelength range of 2.55 nm. The improvement in modulation efficiency allows for a reduction in the phase shifter length to 119 μm, overcoming the bandwidth limitations imposed by spatial walk-off between the electrical and optical waves and enabling a bandwidth of 108 GHz, a value challenging for conventional approaches. This study presents a viable alternative for realizing compact, ultra-broadband, and energy-efficient optical modulators.

This paper presents a high-precision temperature sensing system for in-situ monitoring of the deep-sea environments. The system utilizes a silicon-based Fabry-Perot interferometer as its sensing probe and employs a current-modulated distributed feedback laser to generate a linearly periodic wavelength-varying light source. The absorption line of (HCN)-C-13-N-14 gas cell is used as a reference for real-time laser wavelength calibration. Laboratory test results indicate that the sensor has a sensitivity of 75 pm/degrees C and a resolution of 6.2 x 10(-5 )degrees C with a dynamic range of 0 to 35 degrees C. A watertight and pressure-resistant packaging structure was designed and fabricated, capable of enduring hydrostatic pressures up to 115 MPa, as verified through laboratory-simulated pressure tests, thus guaranteeing the system's adaptability to deep-sea environments. Field tests were conducted in the South China Sea, encompassing both short-term profile tests and long-term bottom-sitting tests, with comparative experiments performed against an SBE37 Conductivity-Temperature-Depth (CTD) sensor. The test results demonstrated consistent temperature measurement details between the sensor and the CTD, exhibiting excellent long-term stability and reliability. The system features a simple structure, high resolution, rendering it a promising tool for oceanographic research, resource exploration, and deep-sea environmental monitoring.

Balanced photodetector (BPD) is an important component for high-speed coherent receiver. Optimization strategy of waveguide-based multi-quantum well (MQW) BPDs, operating at 1550 nm is demonstrated on generic InP platform. Design parameters of BPD are optimized towards achieving the highest bandwidth for a responsivity through an algorithm based on Particle Swarm Optimization (PSO). We do so by establishing an equivalent circuit model of BPD and analyzing its opto-electronic transfer function through numerical modelling. We address the major bottlenecks of high-speed BPDs: transit time of generated carriers and RC loading in our model. The algorithm is able to provide multiple combinations of design parameters with the same output characteristics. Design methodology to integrate laser with optimized BPD is presented to successfully implement coherent receiver.