16 July 2025 in Nature Communications
Silicon photonics is a promising platform for the extensive deployment of optical interconnections, with the feasibility of low-cost and large-scale production at the wafer level. However, the intrinsic efficiency-bandwidth trade-off and nonlinear distortions of pure silicon modulators result in the transmission limits, which raises concerns about the prospects of silicon photonics for ultrahigh-speed scenarios. Here, we propose an artificial intelligence (AI)-accelerated silicon photonic slow-light technology to explore 400 Gbps/λ and beyond transmission. By utilizing the artificial neural network, we achieve a data capacity of 3.2 Tbps based on an 8-channel wavelength-division-multiplexed silicon slow-light modulator chip with a thermal-insensitive structure, leading to an on-chip data-rate density of 1.6 Tb/s/mm2. The demonstration of single-lane 400 Gbps PAM-4 transmission reveals the great potential of standard silicon photonic platforms for next-generation optical interfaces. Our approach increases the transmission rate of silicon photonics significantly and is expected to construct a self-optimizing positive feedback loop with computing centers through AI technology.
20 February 2024 in Light: Science & Applications
Harnessing optical supermode interaction to construct artificial photonic molecules has uncovered a series of fundamental optical phenomena analogous to atomic physics. Previously, the distinct energy levels and interactions in such two-level systems were provided by coupled microresonators. The reconfigurability is limited, as they often require delicate external field stimuli or mechanically altering the geometric factors. These highly specific approaches also limit potential applications. Here, we propose a versatile on-chip photonic molecule in a multimode microring, utilizing a flexible regulation methodology to dynamically control the existence and interaction strength of spatial modes. The transition between single/multi-mode states enables the “switched-off/on” functionality of the photonic molecule, supporting wider generalized applications scenarios. In particular, “switched-on” state shows flexible and multidimensional mode splitting control in aspects of both coupling strength and phase difference, equivalent to the a.c. and d.c. Stark effect. “Switched-off” state allows for perfect low-loss single-mode transition (Qi~ 10 million) under an ultra-compact bend size (FSR ~ 115 GHz) in a foundry-based silicon microring. It breaks the stereotyped image of the FSR-Q factor trade-off, enabling ultra-wideband and high-resolution millimeter-wave photonic operations. Our demonstration provides a flexible and portable solution for the integrated photonic molecule system, extending its research scope from fundamental physics to real-world applications such as nonlinear optical signal processing and sixth-generation wireless communication.
1 May 2024 in Photonics Research
Self-injection locking has emerged as a crucial technique for coherent optical sources, spanning from narrow linewidth lasers to the generation of localized microcombs. This technique involves key components, namely a laser diode and a high-quality cavity that induces narrow-band reflection back into the laser diode. However, in prior studies, the reflection mainly relied on the random intracavity Rayleigh backscattering, rendering it unpredictable and unsuitable for large-scale production and wide-band operation. In this work, we present a simple approach to achieve reliable intracavity reflection for self-injection locking to address this challenge by introducing a Sagnac loop into the cavity. This method guarantees robust reflection for every resonance within a wide operational band without compromising the quality factor or adding complexity to the fabrication process. As a proof of concept, we showcase the robust generation of narrow linewidth lasers and localized microcombs locked to different resonances within a normal-dispersion microcavity. Furthermore, the existence and generation of localized patterns in a normal-dispersion cavity with broadband forward–backward field coupling is first proved, as far as we know, both in simulation and in experiment. Our research offers a transformative approach to self-injection locking and holds great potential for large-scale production.
6 July 2018 in IEEE Transactions on Biomedical Engineering
Objective: As a novel noninvasive modality of oncotherapy or stroke treatment, high-intensity focused ultrasound (HIFU) has drawn more and more attention in the past decades. Whereas, real-time temperature monitoring and treatment-efficacy evaluation are still the key issues for HIFU therapy. Methods: Based on the temperature–conductivity relation of tissues with a sharp conductivity variation of irreversible thermocoagulation at 69 °C, a noninvasive method of treatment-efficacy evaluation for HIFU ablation using the magneto-acousto-electrical tomography (MAET) technology is theoretically studied. By applying the nonlinear Khokhlov–Zabolotskaya–Kuznetsov equation and Pennes equation, a cylindrical model is established to simulate the distributions of pressure, temperature, and conductivity with the consideration of harmonic components. Results: The MAET signals are simulated to analyze the characteristics of the peak amplitude and the axial interval of the two clusters generated by the conductivity boundary of HIFU ablation. Conclusion: The axial interval can be used as the indictor to evaluate the size of HIFU ablation with the minimum axial width of one wavelength. Significance: The favorable results demonstrate the feasibility of real-time treatment-efficacy evaluation for HIFU therapy using the MAET technology and suggest potential applications in clinical practice.
23 December 2025 in Nature Communications
The human eye achieves high resolution and energy efficiency through adaptive focusing on pertinent details. This natural capability sets a benchmark that current machine vision systems struggle to emulate due to rigid parallel-sensing mechanisms. Efforts to achieve retinal-level resolution through brute-force scaling of channels can lead to prohibitive cost, complexity, and power consumption. Here, we address this bottleneck with a retina-inspired chip-scale LiDAR architecture featuring dynamic gazing capability. By synergizing an agile external cavity laser with reconfigurable electro-optic frequency combs, our integrated photonic platform enables coherent LiDAR with parallel, reconfigurable channels, allocating sensing resources to critical regions without global oversampling. This hardware-efficient design enables real-time comb-based 4D imaging at 0.012° beyond-retinal resolution and supports 4D-plus imaging through cooperative sensing with a camera. With its beyond-retinal resolution, dynamic gazing capability, and scalable chip-level manufacturability, this technology offers a path toward lightweight, high-performance perception across aerial, terrestrial, and marine autonomous systems.
1 November 2023 in Measurement