Broadband tunable and low-noise integrated lasers are crucial components in modern optical communication and sensing systems. However, conventional Vernier structure-based external-cavity lasers suffer from a fundamental trade-off between Vernier free spectral range (FSR) and insertion loss. This constrains the laser performance, making it difficult to simultaneously achieve an enlarged tuning range and narrower linewidth. To address this limitation, we propose the inverse Vernier structure that resolves this compromise by superimposing a periodic suppression envelope onto the transmission response of the conventional Vernier structure while preserving the high transmission of the primary mode, effectively expanding the FSR of the overall system without introducing significant additional losses. As a proof of concept, we have experimentally validated a hybrid III–V/Si3N4 laser incorporating the proposed inverse Vernier structure. The device achieves a fourfold expanded Vernier FSR of 15.58 THz (140.7 nm), facilitating a wavelength tunability of 9.4 THz (82.3 nm) across the L- and U-bands and a narrow intrinsic linewidth of 88.8 Hz. In particular, the proposed inverse Vernier structure is flexibly compatible with diverse material platforms and configurations and poised to become a key enabler for widely tunable and narrow-linewidth hybrid integrated lasers. This work opens new avenues for applications requiring broadband tunability and high coherence, including high-speed coherent optical communication, high-resolution fiber sensing, and quantum communication technology.
Narrow-linewidth, frequency-modulated continuous-wave (FMCW) lasers are indispensable components in modern sensing and measurement systems. Toward their integrated photonic implementations, self-injection-locked lasers stand out for the ultralow noise and operational simplicity. However, their tuning range remains limited, primarily due to the restricted locking range imposed by the long-delay external cavities. In this work, we address this limitation by demonstrating a widely tunable FMCW integrated laser through self-injection locking to a strongly reflective multimode micro-ring resonator (MRR). This MRR adopts a tailored over-coupled design with negligible intermodal crosstalk, unifying large free spectral range and high-Q strong reflection, substantially extending the laser self-injection locking range. Experimental results reveal that the multimode MRR presents a high intrinsic Q value of 1.45×107 together with a reflectivity of −0.92 dB. The demonstrated hybrid integrated laser leverages a combination of exhibits, a static mode-hop-free tuning range up to 37.84 GHz, a side-mode suppression ratio exceeding 65 dB, and an intrinsic linewidth of 1.55 kHz. Furthermore, it achieves FMCW generation with a record-high chirp bandwidth of 25.42 GHz at a 100 Hz repetition rate. These results highlight the promising application prospects and practical value of the proposed laser in cutting-edge fields, including light detection and ranging, microwave photonics, and optical fiber sensing.
Multimode nonlinear systems offer a fertile platform for optical functionalities that are fundamentally inaccessible in linear or single-mode regimes, a potential revitalized in recent years by the framework of optical thermodynamics. In this work, we reveal a previously unexplored class of wave phenomena in optical fibers driven by the interplay between beam expansion and nonlinear mixing. We predict that, at high powers, tightly focused beams can undergo an irreversible expansion, which self-organizes their intensity and phase into the fundamental mode of an arbitrary guiding landscape, with conversion efficiencies exceeding 90%. Viewed through the optical thermodynamic framework, these states correspond to near-zero-temperature thermodynamic equilibria, formed through an optical analog of Joule–Thomson expansion. We show that expansion-driven effects can exceed fundamental limits of linear-optics in graded-index and step-index fibers, revealing a new approach to autonomous beam shaping and beam capture without switching or adaptive control.
Optical data centers have become the technology enablers for numerous emerging bandwidth-hungry applications and services, such as artificial intelligence, 4K/8K video streaming, and social networking. This is resulting in an exponential increase in data center traffic. Similar to all the previously released Ethernet standards, the future 1.6/3.2 TbE standards are likely to adopt wavelength division multiplexed transmission in the O-band, utilizing both an increased number of wavelength channels and data rate per channel. This will increase the overall energy consumption of data center networks due to the elevated number of high-power lasers and high-sensitivity wideband transceivers within the data center. The semiconductor optical amplifier (SOA) is a promising candidate for enabling a transition toward high-capacity and energy-efficient data center networks due to their small footprint and high integrability with optical transceivers. In recent years, SOAs based on quantum dot (QD) technologies have shown great potential for WDM signal amplification using a single SOA with reduced nonlinear signal distortions and energy consumption. In this paper, we discuss various SOA technologies, with a special focus on QD-SOAs, and several applications of SOAs to develop future data center network architectures, systems, and transceiver concepts with the goal of delivering new energy-efficient physical layer technologies for optical data center interconnects.
Multimode fibers (MMFs) support high-dimensional optical transport within a compact form factor, with growing applications in biomedical imaging, telecommunications, and quantum engineering. Realizing this potential requires accurate knowledge of the fiber transmission matrix (TM); yet experimentally measured TMs are often corrupted by system-level distortions that obscure the underlying modal physics and limit wavefront control. Here, we show that the reciprocity of optical waves imposes symmetry constraints that provide self-consistency conditions for identifying and correcting such distortions. Enforcing these constraints enables improved recovery of the transported fields and allows the computational synthesis of confocal images from reflection measurements through the fiber. We further introduce matrix normality as a physically meaningful metric that quantifies the orthogonality of the propagating modes encoded in the TM. Using controlled static compression of the fiber, we demonstrate experimentally that departures from normality at a single wavelength predict a monotonic reduction in the spectral bandwidth of the principal mode basis. These results reveal a direct connection between spatial modal structure and spectral stability in MMF transport. More broadly, they establish a unified framework for understanding and controlling broadband light propagation in MMFs and complex waveguides.
Silicon photonics has emerged as a key platform for next-generation communication and sensing technologies, enabling compact, scalable, and energy-efficient optical systems compatible with large-volume semiconductor manufacturing. When combined with highly nonlinear organic electro-optic materials, silicon photonics gains functionalities unavailable in silicon alone, opening new opportunities for ultrafast modulation and for direct transduction between high-frequency wireless electromagnetic fields and guided optical signals. In this work, we investigate hybrid silicon–organic devices based on patch-array antennas integrated with silicon waveguides for the detection of free-space terahertz waves. A telecom probe beam is routed through the silicon circuitry and converted into a plasmonic slot mode that interacts with the terahertz field accumulated in the nanoscale gap of the antenna. The latter is filled with a poled organic material providing the Pockels nonlinearity required for electro-optic terahertz modulation. We design various patch-antenna geometries and compare them to more conventional bowtie antennas. We show that patch-array antennas provide enhanced sensitivity at selected frequencies through a stronger resonant field concentration, in contrast to bowtie layouts that instead exhibit broadband responses owing to their tapered dipolar design. We further analyze how the detected signal depends on the relative position of the terahertz beam spot with respect to these two types of antenna on the chip, thus reconstructing the spatial coupling behavior of the two configurations. These results clarify the trade-offs between bandwidth, sensitivity, and spatial tolerance in planar antenna-coupled plasmonic terahertz detectors and provide practical design guidelines for integrated optical sampling of terahertz signals.
Paraxial optical Skyrmions attract a lot of research interest recently due to their topologically protected spin textures. However, combining nonlinear propagation dynamics with complex spin textures remains largely unexplored. We establish a theoretical framework for studying ultrafast laser pulses with nontrivial spin textures in a gas-filled hollow-core fiber (HCF) that naturally bridges these two regimes. Using a complete basis of linearly polarized modes derived for the HCF, we first analytically reveal a new kind of fractional optical Skyrmions, which are different from previous ones studied in free space. Then, a simplified nonlinear model is established to show the key influence of Kerr nonlinearity on the spin-texture evolution and identifies an optimal energy ratio between orthogonal polarizations, which enables balanced spectral broadening while preserving the spin evolution pattern. As an application of the theory, rigorous simulations of nonlinear pulse compression are conducted under realistic conditions, revealing the characteristics of the underlying dynamics that are not measurable in experiments. The output pulse is compressed to a few optical cycles covering near 200 nm-width spectrum, with each wavelength having a nontrivial spin texture. A method to characterize the temporal polarization distribution of this unique source is proposed. By connecting strong-field nonlinear propagation and topological light manipulation, this work lays a unified foundation for an emerging discipline at the interface of ultrafast nonlinear science and topological optics.
Third harmonic generation (THG) microscopy provides label-free structural contrast of biological tissues. While the THG process can be enhanced by electronic transitions in absorbing molecules, knowledge of the third-order nonlinear properties of biological pigments is limited, and the potential of THG imaging for chemically selective imaging remains underutilized. In this study, we investigated resonant THG in pigments by performing nonlinear microspectroscopy in situ on three representative absorbers: hemoglobin in red blood cells, pteridines in xanthophores, and melanin in human hair. We measured THG spectra in the 1120–1300 nm excitation range. We observed a wavelength-dependent signal enhancement ranging from 10 to 100 times in these pigmented structures, depending on their respective absorption properties. We developed a numerical model of THG from interfaces, including resonant and non-resonant contributions, to interpret these observations. Our calculations reproduce key experimental trends, including resonance-induced enhancement factors and spectral properties. Finally, we demonstrate that third-order sum-frequency generation (TSFG) imaging provides spectroscopic contrast capable of distinguishing red blood cells from xanthophores in live zebrafish larvae. These results suggest that THG/TSFG is a promising approach to label-free, pigment-specific multiphoton imaging.
The wide-field spiral phase contrast microscope (SPCM) is an all-optical edge-enhanced imaging technique that performs Fourier filtering based on a 4f system. It allows large phase gradients to be highlighted through convolution of a phase object with the point spread function (PSF) given by the Fourier transform of the coherent transfer function (CTF). In the case of pupil mismatch between the 4f system and the spiral phase filter (SPF) placed at the Fourier plane, the physical edge diffraction of the SPF is non-negligible and will cause strong sidelobes in the PSF, leading to low image resolution and background noise after convolution. Herein, we propose a scanning SPCM, which uses an Airy spot to sample the phase object laterally point by point and detects the on-axis intensities at the image plane accordingly. This structural change transforms the SPCM from Fourier filtering to match filtering. Match filtering allows the SPCM to optically compute the complex weight that measures the matching degree between the Fourier transform of the local sampled field of the phase object and the joint CTF, which is the convolution of the illumination pupil and the imaging pupil. The edge enhanced image is directly given by the complex weights at the sampling coordinates. The scanning SPCM takes the advantage of the joint CTF, which not only shows a much higher cutoff spatial frequency than the wide-field SPCM, achieving high-resolution imaging, but also shows soft edges, achieving suppressions of physical edge diffraction of the SPF and sidelobes in the joint PSF. The imaging results of phase objects verified that the scanning SPCM achieves twofold resolution improvement. The scanning SPCM breaks through the inherent barrier of low image resolution in the wide-field SPCM by transforming Fourier filtering to match filtering, and it possesses great advantages in all-optical high-resolution edge enhancement.
We demonstrate a monolithic extended-cavity vertical-cavity surface-emitting laser (VCSEL) achieving a linewidth of 890 kHz, which, to our knowledge, ranks among the narrowest reported for this class of lasers. The device incorporates an extended cavity and an intermediate reflector monolithically integrated into the epitaxial structure, forming a coupled configuration that significantly increases photon circulation path and enhances mode selectivity. The device has an emitting wavelength of 795 nm, with a side-mode suppression ratio of 40 dB. This work validates the feasibility of using monolithic extended cavities to achieve sub-megahertz linewidths in VCSELs without external feedback or complex packaging, offering a promising pathway for applications in coherent optical communications, laser spectroscopy, and quantum sensing.
The mid-infrared (MIR) spectral region is an attractive band for major applications, such as spectroscopy owing to the presence of many molecular absorption signatures. Fiber lasers operating in this spectral region are highly sought after because of their compactness and ease of integration into detection systems. Considerable effort has been invested in advancing the ultrafast operation of such lasers, particularly in the 3 μm spectral range where erbium ions exhibit broadband gain under 976 nm pumping. However, achieving passively mode-locked femtosecond operation with erbium fiber oscillators remains challenging due to the complexities involved in designing saturable absorbers, particularly with adapted carrier recovery time. In this study, we present a compact setup for the sub-picosecond operation of an erbium fluoride fiber laser. The passive mode-locking operation is enabled by a graphene film on a gold mirror fabricated in-lab using a wet chemical transfer process. The laser produced a record pulse duration of 656 fs at 2.8 μm and delivered energies of up to 4.93 nJ. This work explores the soliton dynamic operation of the laser and paves the way for the further development of such MIR femtosecond fiber oscillators.
The mid-infrared (MIR) spectral region is an attractive band for major applications, such as spectroscopy owing to the presence of many molecular absorption signatures. Fiber lasers operating in this spectral region are highly sought after because of their compactness and ease of integration into detection systems. Considerable effort has been invested in advancing the ultrafast operation of such lasers, particularly in the 3 μm spectral range where erbium ions exhibit broadband gain under 976 nm pumping. However, achieving passively mode-locked femtosecond operation with erbium fiber oscillators remains challenging due to the complexities involved in designing saturable absorbers, particularly with adapted carrier recovery time. In this study, we present a compact setup for the sub-picosecond operation of an erbium fluoride fiber laser. The passive mode-locking operation is enabled by a graphene film on a gold mirror fabricated in-lab using a wet chemical transfer process. The laser produced a record pulse duration of 656 fs at 2.8 μm and delivered energies of up to 4.93 nJ. This work explores the soliton dynamic operation of the laser and paves the way for the further development of such MIR femtosecond fiber oscillators.
We demonstrate temporal pattern formation in a coherently driven fiber ring cavity whose effective finesse is continuously reconfigured using distributed Raman amplification. We achieve an effective finesse of up to Feff≈800, corresponding to a linewidth of ∼725 Hz (Qeff ≈ 2.7 × 1011) at 1555 nm. By exploiting the resulting increase in effective photon lifetime, we excite stable temporal cavity solitons and generate a low-repetition-rate frequency comb with a spacing of 580 kHz. Finally, we analyze the impact of the Raman loss-compensation mechanism, particularly its associated noise, and show that a trade-off exists between soliton excitation threshold and stability.
Functional near-infrared spectroscopy (fNIRS) enables portable, non-invasive monitoring of cerebral oxygenation, yet its quantitative accuracy in continuous-wave optical topography (CW-fNIRS-OT) is often constrained by conventional methods that rely on an empirical differential path length factor (DPF). The empirical DPF typically assumes homogeneous tissue and ignores the layered head structure as well as inter-subject anatomical variability. In this study, we overcome the DPF-induced quantitative limitations in CW-fNIRS-OT by introducing a time-shift inversion strategy that, in tandem with layer-sensitive time-gated windows, achieves an efficient optical properties inversion while effectively decoupling and compensating for path length-dependent distortions. Leveraging a statistical brain model derived from anatomical atlases, our approach synthesizes subject-specific head geometry from simple anthropometric data, explicitly eliminating MRI dependency. Photon-transport simulations are employed to predefine time gates with preferential sensitivity to superficial and deep layers, while a dynamic time-shift correction compensates for residual inter-subject geometric mismatches. Ultimately, this design permits the precise inversion of layer-specific optical properties via a computationally lightweight single-layer diffusion equation model. Validation through numerical simulations, phantom experiments, and in vivo breath-holding tasks demonstrates that the proposed method significantly outperforms conventional empirical-DPF approaches. It corrects systematic path length errors, achieving a quantitativeness ratio of ∼0.95 vs ∼0.85 for conventional methods, and improves spatial fidelity, contrast-to-noise ratio, and task-state discriminability in SVM classification. Affording personalized, high-accuracy DPF inversion with computational efficiency, this framework provides a clinically viable route to individualized cerebral oxygenation assessment and is poised to catalyze the broader adoption of fNIRS in both neuroscientific inquiry and diagnostic practice.
Polarization controller (PC) serves as a critical component for manipulating the polarization state of light and holds great significance in both classical and quantum communication systems. Anisotropic electro-optic (EO) modulation in thin-film lithium niobate provides a new avenue for realizing high-speed PCs. This paper presents a high-performance on-chip PC with a simple architecture, assembled from linear tapers, Y-branch splitters, and EO phase shifters. The proposed PC has an ultra-broad working bandwidth that covers the entire optical communication spectrum. It enables arbitrary polarization manipulation. Endless polarization generation is realized at typical wavelengths of 1550 and 1310 nm, while exhibiting a record-high dynamic polarization extinction ratio >43 dB. High-speed random four-polarization encoding for quantum key distribution at a switching speed of 20 MHz is demonstrated. Enabling polarization manipulation with high speed, high PER and broad bandwidth, the PC offers huge potential for performance enhancement in optical systems.
High-sensitivity, bias-selective dual-band infrared detection is essential for two-color absolute thermometry of mid- to long-wavelength infrared sources, including the picowatt-class signal regime relevant to cryogenic near-field thermal measurements. Here, we report a triple-quantum-well charge-sensitive infrared phototransistor in which two spatially isolated floating gates provide electrically separated photoresponse pathways, allowing the 10.1 and 16.4 μm channels to be addressed individually or jointly by gate bias. The cross-hole plasmonic grating supports two geometrically distinguishable resonances with different dominant tuning sensitivities: a period-dominated Rayleigh anomaly–surface plasmon polaritons (SPP) mode and an arm-length-dominated localized shape plasmon–SPP hybrid mode. This partially mode-decoupled response reduces spectral entanglement and enables practical co-design of the two infrared bands. At an optimized well doping of 8.0 × 1017 cm−3, the device delivers an integration responsivity of 6.17 × 105 A W−1 at VSD = 10 mV under a 4.24 pW, 300 K background; the band-resolved photocurrent responsivities at VSD = 120 mV are 1.35 × 103 A W−1 at 10.1 μm and 6.07 × 103 A W−1 at 16.4 μm. Quantum efficiencies of 19.4% (10.1 μm) and 16.0% (16.4 μm) yield specific detectivities of 8.5 × 1011 and 1.1 × 1012 cm Hz1/2 W−1, respectively, with a balanced (∼1:1) dual-band spectral response suitable for two-color pyrometry. The architecture establishes a route to high-gain dual-band mid- to long-wavelength infrared sensing for cryogenic two-color thermometry.
The ability to manipulate ultrafast pulses has been at the heart of major technological advances, including chirp pulse amplification, pulse compression, and frequency conversion. At terahertz (THz) frequencies, precise control of pulse dispersion and frequency is key to next-generation high-speed wireless communication and sensing applications, yet such control remains difficult owing to inefficient nonlinear materials and the fundamental constraints of passive devices. Here, to overcome these limitations, we introduce a time-varying semiconductor waveguide platform with a cascade of temporal and spatial boundaries and demonstrate programmable control of THz pulse waveforms with efficient frequency conversion. By precisely timing optical pulses at multiple temporal boundaries to control the THz intra-pulse frequency shifts, we generate tailored positive, negative, and non-monotonic chirp pulses, as well as tunable frequency upconversion. These results open a way toward programmable THz pulse manipulation for time-varying photonics.
Emerging applications of integrated optics such as free-space beam projection, atomic spectroscopy, and interfacing with ultra-stable laser resonators often require beam sizes of the order of 100 μm or even higher, i.e., a hundred times larger than the mode field of common integration platforms such as silicon-on-insulator. Adiabatic tapers become prohibitively large for such extreme mode expansions, while other alternatives such as evanescent couplers, Bragg deflectors, and integrated lenses often suffer from narrow bandwidths or have not yet been proved beyond beam sizes of 40 μm. This study introduces an on-chip spot-size converter based on the Cassegrain dual-reflector geometry, adapted for the first time to the planar constraints of integrated optics platforms, alongside a novel, fully analytical design methodology. We experimentally demonstrate a device that expands the fundamental TE mode from a 0.5 μm waveguide to a 116 μm-wide beam, with an insertion loss of 0.6 dB at λ0 = 1.55 μm and a 1-dB bandwidth exceeding 200 nm (1.45–1.65μm. This design also offers a reduced area compared to its single parabolic analog. We, furthermore, use the device to feed a large-aperture optical antenna, producing a collimated free-space beam with a beam size of 116 × 235 μm2.
The implementation of reconfigurable nonlinear activation functions (NLAFs) is a key challenge in advancing photonic neural networks to model complex relationships. This study addresses this issue by introducing a device that exploits the phase-dependent optical properties of a Ge2Sb2Te5 superlattice (GST-SL) integrated on a silicon microring resonator. By manipulating the distinct absorption losses and refractive indices of the GST-SL across its crystallization states, we experimentally demonstrate five types of NLAFs: half-sigmoid, ELU, softplus, ReLU, and radial basis. The characterized functions exhibit a low activation power threshold of 1.76 mW and a high operational speed of up to 1 MHz. In a system-level benchmark using a three-convolutional-block neural network for MNIST handwritten digit classification, our activation functions boost the classification accuracy from 95.35% to 99.24%. These results highlight the promising performance and application potential of our device for future on-chip photonic neural networks and advanced optical computing.
This letter presents a novel nonlinear optical diagnostic technique to simultaneously investigate flow phenomena over molecular and fluid timescales, using a single femtosecond laser source. We combine hybrid femtosecond/picosecond coherent anti-Stokes Raman spectroscopy (fs/ps CARS) and femtosecond laser-induced grating spectroscopy (fs LIGS), adopting a new planar phase-matching geometry, whereby 1D imaging spectroscopy is realized over a shared probe volume. This dual diagnostic approach uniquely bridges physical scales: fs/ps CARS provides molecular-level snapshots, while fs LIGS probes macroscopic phenomena, such as wave propagation in compressible gases. We demonstrate this technique in an underexpanded supersonic air jet, resolving the temperature and speed of sound fields across the Mach disk. Hybrid fs/ps CARS probes the rotational energy distribution of the Raman-active molecules, providing the local temperature and composition fields with a 20 μm resolution. Through multi-photon ionization, we generate a volumetric plasma Bragg grating, which enables thermal fs LIGS imaging in virtually any gas mixture, from which we extract the acoustic response of the fluid medium and, thus, determine the local speed of sound. Overall, this work establishes a general framework for high-precision, time- and space-resolved characterization of complex flows where microscopic inter-molecular interactions determine the macroscopic fluid dynamics, with significant implications for high-enthalpy aerodynamics and supercritical fluid technologies.
