Abstract Evanescent-field imaging enables selective probing of near-interface regions and has been widely used to visualize cell membrane dynamics, cell–substrate interfaces, nanoparticles, and surface-associated biological structures. Although conventional plasmonic illumination enables sensitive detection of subtle changes in the local dielectric environment through resonant field enhancement, its effective probing range is typically confined to within 100–200 nm of the interface. Here, we present long-range plasmonic illumination microscopy (LRPIM) based on a metal–dielectric multilayer chip. At the sample-facing surface, the chip generates an enhanced evanescent field that penetrates deeply into the sample while preserving low-background near-field excitation. Using individual microspheres as scattering probes, we established a calibration relationship between the evanescent-field penetration depth and the full width at half-maximum (FWHM) of the focused hotspot formed at the exit surface of each microsphere. Based on this calibration, the chip exhibited an electric-field penetration depth approximately 4.3 times that obtained with plasmonic scattering microscopy (PSM), enabling near-surface illumination over a micrometer-scale axial range. Furthermore, LRPIM supports fluorescence imaging, enhancing fluorescence signals from propidium iodide (PI)-stained cells and revealing localized nucleic acid enrichment within their nuclei. We anticipate that LRPIM will provide a straightforward chip-based approach to evanescent-field scattering and fluorescence microscopy, enabling structures to be probed over a micrometer-scale axial range near interfaces.
Abstract Conventional neuromorphic vision systems suffer from architectural fragmentation, wherein sensing, memory, and processing are distributed across discrete modules, causing high latency, energy waste, and poor dynamic adaptability. Here, we present a wavelength-driven trimodal field-effect phototransistor (TriM-FEP) based on layered HfS2, which offers a new solution to this challenge. This single device integrates polarity-switchable photoresponses and synaptic dynamics without external bias or complex circuitry. By simply tuning the incident wavelength, the TriM-FEP delivers three distinct, dynamically switchable photoresponses: a synaptic-like photoresponse under solar-blind ultraviolet, fast negative photocurrent in the visible, and fast positive photocurrent in the near-infrared. Notably, this unique polarity reversal and speed variation are dominated by the photogating effect and the bolometric effect at different wavelengths, respectively. Beyond fundamental physics, the TriM-FEP enables real-time multichannel encrypted optical communication and in-sensor image preprocessing, eliminating the need for separate computing units. This work transcends the limitations of multicomponent cooperation, establishing a compact, energy-efficient platform that unifies sensing, memory, and processing at the device level.
Abstract Solid-state quantum emitters (QEs) in two-dimensional semiconductors offer compact, chip-compatible sources for quantum photonics. In transition-metal dichalcogenides (TMDs), nanopillars are widely used to induce localized emission, yet the underlying confinement mechanism and the relative roles of strain versus dielectric environment remain unclear. Here we combine hyperspectral superlocalization of photoluminescence with coregistered AFM topography and phase to map localized states (LS) positions in MoSe2 on GaN nanopillars and correlate them with bending strain and the local environment. Contrary to a purely strain-driven picture, LS cluster at suspended–supported interfaces around the pillar apex and span a broad strain range with no clear threshold, while being scarce along high-strain ripples. We discuss the candidate mechanisms, including the local dielectric environment, and suggest coengineering of strain gradients and nanoscale interface heterogeneity for deterministic emitter positioning in TMDs.
Abstract Quantum phenomena in two-dimensional electron gases (2DEGs) provide new opportunities for high-responsivity far-infrared detection. A prime example is photoelectric tunable-step (PETS) terahertz (THz) detectors, which use the in-plane photoelectric (IPPE) effect as a sensitive photonic detection mechanism. Here, we show that the terahertz sensitivity of this quantum photodetection mechanism can be dynamically tuned by using infrared illumination. We leverage the persistent photoconductivity (PPC) effect in AlGaAs/GaAs-based heterostructures to tune the device performance by adjusting the electron density and mobility of the channel. Our approach enables optimization of the responsivity for different operating modes and dynamic control over the output impedance of PETS terahertz detectors of approximately one order of magnitude. We provide an explanation for the underlying physics that advances the understanding of quantum photodetection phenomena in 2DEGs. Our findings pave the way to in situ tunable optoelectronics combining concepts from the infrared and terahertz spectral regions.
Abstract Tailoring light-matter interaction using photonic modes depends on both the spectral and spatial overlap between the modes and quantum emitters, which is critical for optimizing light-matter coupling and enabling efficient, controllable emission. Here, we demonstrate deterministic control of emission from colloidal quantum dots (CQDs) integrated into dielectric light-emitting metasurfaces (LEMs) by precisely localizing the emitters at the field maxima of the photonic modes across the entire metasurface. Numerical simulations show that deterministic localization enhances the emission, yielding a higher emission enhancement factor than that of LEMs with randomly distributed emitters. Experimentally, such global spatial overlap is achieved using a single-template lift-off process to localize the CQDs at the desired positions. Optical characterization confirms efficient coupling between CQD emission and the photonic modes, resulting in a narrowed spectral linewidth and highly directional out-of-plane emission. Deterministic emission control in metasurfaces adds a new degree of freedom to LEM design, paving the way for an efficient integrated light source for classical and quantum photonic devices.
Abstract Topologically structured optical fields exhibit rich physical properties and offer new opportunities for information encoding and light-matter interaction. In this work, we propose a general and scalable framework for generating and reconfiguring topological light using coherent beam combination (CBC). By treating the CBC system as an electronically addressable optical phased array, we demonstrate that the topology of the optical field becomes a programmable parameter. As a representative example, we focus on optical skyrmions and present an experimental demonstration of deterministic and dynamic switching between Bloch-type, Néel-type, and antiskyrmion textures using a 7-channel Yb-doped fiber master oscillator power amplifier CBC system. This approach establishes a scalable and programmable route toward high-power topological vector fields and reconfigurable structured light.
Abstract Metal halide perovskites are attractive gain media for coherent lasers due to their high optical gain, strong photoluminescence and tunable bandgap. Mixed-cation design can optimize lattice structures and mitigate nonradiative recombination, but mismatched crystallization and incomplete precursor conversion may introduce grain defects and metallic Pb0 species. Pb0-related deep traps exacerbate carrier losses and deteriorate overall lasing performance. This work presents a bifunctional modification strategy using biuret to suppress Pb0 precipitation and passivate defects in mixed-cation perovskites. Owing to carbonyl and amino groups, biuret simultaneously restrains Pb0 formation via Lewis acid–base coordination and passivates defects through hydrogen bonding with bromide ions. The modified perovskite films possess a material gain of 3754 cm–1, an extended gain lifetime of 148 ps, a low amplified spontaneous emission (ASE) threshold of 2.60 μJ cm–2, and improved long-term ASE stability. The constructed single-mode vertical-cavity surface-emitting laser (VCSEL) delivers outstanding lasing performance with a narrow emission line width of 0.068 nm and a high Q factor of 7848. This study reveals the relationship between Pb0 inhibition, defect passivation and optical gain, and offers a practical route for high-performance perovskite lasers.
Abstract In optical microscopy, high spatial resolution comes at the cost of a short depth of field. This trade-off prevents the formation of sharp images of three-dimensional objects or objects moving in and out of focus. Moreover, it is often difficult to know the extent to which the object is out of focus, which makes it challenging to determine the point spread function that describes the blurring. This hinders the ability to restore the blurred image using digital postprocessing. To resolve these issues, we design a phase mask that, when inserted into the microscope, extends the depth of field, making the point spread function insensitive to the location of the object. We leverage end-to-end machine learning tools to design this phase mask together with a Richardson-Lucy-type deconvolution algorithm to remove image blurring. The phase mask is then manufactured with a commercial 3D nanoprinter and used in a microscope to demonstrate defocus-insensitive imaging of microfabricated objects. The experiments successfully verify the operation of both the phase mask and the image restoration algorithm.
Abstract We report the first demonstration of monolithically integrated amorphous indium–gallium–zinc oxide (IGZO) synaptic transistors on a silicon nitride (Si3N4) waveguide, enabling direct on-chip optical programming by using guided 532 nm light. In this architecture, the evanescent field of the guided mode modulates the IGZO channel conductance, while the three-terminal transistor configuration enables selective electrical access and readout of the programmed synaptic states. Unlike previously reported IGZO optoelectronic synapses that rely on externally aligned free-space illumination, the waveguide serves as a lithographically defined optical programming path through which a single guided optical signal interacts with multiple synaptic transistors distributed along the same waveguide. The monolithically integrated guided optical input and electrical output configuration therefore provides a hybrid interface between integrated photonic signal delivery and transistor-based neuromorphic hardware. The waveguide geometry was designed to maintain evanescent coupling to the transistor channel while limiting propagation loss, and guided-light-induced modulation was experimentally observed in five transistors integrated along one waveguide. The IGZO TFT additionally exhibited photoresponse under free-space illumination from 405 to 635 nm, confirming the visible range photosensitivity of the active layer. Under pulsed guided-light excitation, the devices exhibited persistent photoconductivity, paired-pulse facilitation, and tunable short- and long-term synaptic plasticity. An ANN simulation using experimentally extracted conductance update characteristics achieved a handwritten digit classification accuracy of 98.02%, demonstrating the applicability of the measured synaptic response to neuromorphic learning tasks. These results establish a monolithically integrated waveguide–IGZO synaptic platform that combines guided optical programming, persistent electronic weight storage, and device-selective electrical readout for optoelectronic neuromorphic computing.
Abstract The resolving ability of widefield fluorescence microscopy is fundamentally limited by out-of-focus background owing to its low axial resolution, particularly for densely labeled biological samples. Although total internal reflection fluorescence (TIRF) microscopy provides strong near-surface sectioning, it is intrinsically restricted to shallow imaging depths. Here we present computational TIRF (cTIRF), a deep learning-based imaging modality that generates TIRF-like sectioned images directly from conventional widefield epifluorescence measurements without any optical modification. By integrating a physics-informed forward model into network training, cTIRF achieves effective background suppression and axial-resolution enhancement while maintaining consistency with the measured widefield data. We demonstrate that cTIRF recovers near-surface structures with performance comparable to experimental TIRF, and further enables both single-frame and volumetric sectioned reconstruction in densely labeled samples where conventional TIRF fails. This work establishes cTIRF as a practical and deployable alternative to hardware-based optical sectioning in fluorescence microscopy, enabled by rapid adaptation to new imaging systems with minimal calibration data.
Abstract Snapshot ultrafast optical imaging (SUOI) is critical for real-time visualization of transient physical processes. Compressed ultrafast photography (CUP), a leading SUOI technique, enables capture of nonrepeatable events at ultrahigh frame rates. However, conventional CUP relies primarily on intensity-based detection, restricting its ability to resolve transparent phase objects and transient scenes characterized by weak boundaries or subtle refractive-index variations. To address this limitation, we present vortex-filtering-enabled edge-enhanced CUP (VE-CUP), which incorporates a spiral phase contrast module in front of a CUP system. By converting weak-phase or amplitude fluctuations into edge-enhanced intensity modulations prior to spatiotemporal encoding and reconstruction, VE-CUP efficiently suppresses low-frequency background noise, accentuates boundary and gradient features, and thereby enhances the visibility and structural clarity of low-contrast transient scenes. Single-shot imaging of femtosecond-laser-induced plasma evolution in air and shock-wave propagation in MgO crystals demonstrates that VE-CUP can effectively visualize and quantitatively characterize representative ultrafast dynamics across picosecond to nanosecond time scales. Our findings establish VE-CUP as a powerful extension of CUP for probing weak-phase, low-contrast, and boundary-dominated ultrafast phenomena.
Abstract Photonic crystal and whispering-gallery-mode resonators are the leading platforms for high-performance microcavity lasers. However, both face inherent trade-offs: photonic crystal cavities achieve strong optical confinement at the expense of large footprints due to extensive surrounding mirrors, whereas whispering-gallery-mode cavities offer high quality factors but suffer from large mode volumes and extreme sensitivity to sidewall roughness. Here, we demonstrate monolithic photonic crystal ring lasers on silicon-on-insulator to combine the advantages of photonic crystal and whispering-gallery-mode cavities. The photonic crystal ring lasers were fabricated from III-V membranes grown on silicon-on-insulator via selective lateral heteroepitaxy and exhibit a compact footprint of 5.51 μm2 (9.50 × 0.58 μm2) and a mode volume of 0.381 (λ/n)3. The devices achieved continuous-wave lasing with a threshold power of 0.27 μW at room temperature and maintained operation with a sub-μW threshold up to 100 °C. Our approach offers a compelling route toward high-density, low-power on-chip light sources for silicon photonics.
Abstract Silicon/germanium (Si/Ge) core–shell nanostructures remain a critical platform for CMOS-compatible short-wave infrared (SWIR) photodetection. To address the intrinsic weak absorption of silicon and germanium, the use of subwavelength nanostructures as resonators has emerged as a key strategy. Despite the widespread adoption of subwavelength nanoresonators to enhance light–matter interaction, the critical problem of “Absorption-Responsivity mismatch” persists, in which optical absorption does not inherently translate into efficient carrier collection. This study investigates the relationship between absorption and responsivity by controlling the cross-sectional evolution of nanoresonators from polygons (triangles, squares) to circular geometries. Finite-difference time-domain (FDTD) simulations confirm that not all absorption enhancement peaks efficiently translate into device high responsivity: this depends on whether the optical field is effectively localized within the “Active area” of nanoresonators rather than leaking into the air. More importantly, we observed that even when absorption peaks originate from the same “active area,” changes in the geometric cross section still induce significant differences in responsivity. Further analysis reveals that this difference arises because of competitive dynamics between radiative loss dissipation and effective photocarrier absorption. Specifically, electromagnetic energy in low-symmetry resonators is frequently dissipated via radiative leakage channels instead of being converted into effective photocarriers. Radiative loss is progressively suppressed as the cross section evolves from polygons to circular geometry, thereby confining optical energy in active Mie resonance modes that spatially overlap the nanoresonators. The circular Si/Ge nanoresonator array consequently achieves the highest responsivity of 8.5 A/W at a light intensity of 0.1 mW/cm2. This study on “absorption-responsivity mismatch” mechanisms provides useful insights for resolving absorption-carrier collection mismatch and for constructing high-performance silicon-based optoelectronic devices.
Abstract Near-infrared (NIR) photodetectors based on lead sulfide (PbS) quantum dots (QDs) are attractive for low-cost and large-area optoelectronics, but their performance is often limited by inefficient charge separation, trap-assisted recombination, and dark current. Here, we report an all-solution-processed PbS QD photodetector based on ligand engineering and an inkjet-printed Ag top electrode. Tetrabutylammonium iodide (TBAI) and 1,2-ethanedithiol (EDT) treatments were used to form ligand-modified PbS QD layers with distinct energy-level alignments, enabling a bilayer junction that promotes charge separation and suppresses dark current. The optimized PbS-TBAI/PbS-EDT device exhibited a dark current of 2.71 × 10–8 A cm–2, a responsivity of 0.390 A W–1, and an estimated shot-noise-limited specific detectivity of 4.18 × 1012 Jones under 905 nm illumination at −1 V. The device also showed subsecond photoresponse, preliminary storage stability, and good device-to-device reproducibility. A 5 × 7 array was further fabricated to demonstrate proof-of-concept near-infrared pattern imaging. These results indicate that ligand-engineered PbS QD bilayers are promising for scalable, solution-processed NIR sensing and imaging platforms.
Abstract Two-dimensional materials hold immense application potential in the long-wave infrared (LWIR) domain. However, surmounting the bandgap limitation to boost their photon absorption capacity within this spectral range remains a formidable challenge. In this work, a high-performance photodetector Ag/[Si-NWs/rGO]/rGO/PtTe2/Au, in which graphene oxide is first spin-coated into the gaps of a silicon nanowires (Si-NWs) array, followed by high-temperature annealing at 900 °C to convert it into reduced graphene oxide (rGO), and then a crystalline PtTe2 layer on the rGO layer, is presented. By constructing a light-trapping heterojunction with Si-NWs/rGO, the absorption range of rGO is extended to the LWIR region (8–14 μm), and the natural passivation properties of PtTe2 enable the formation of a defect-free rGO/PtTe2 van der Waals heterojunction. Thus, a rapid carrier separation is achieved by leveraging the unique tilted Dirac cones of multilayer PtTe2 and the built-in electric field formed with rGO. As a result, a specific detectivity of 1.59 × 1011 Jones under 0.25 mW/cm2 11.5 μm illumination is obtained at 0 V and at room temperature, demonstrating excellent LWIR detection performance. This dual progressive heterojunction design can be extended to other two-dimensional materials, utilizing the unique interlayer coupling capabilities of van der Waals heterojunctions for enhancing LWIR response.
Abstract Optical skyrmions provide a versatile platform for structuring the topology of light through spatially varying polarization textures. In most optical skyrmion textures, however, the azimuthal texture evolution remains linear and uniform, restricting the programmable control of local polarization gradients. Here, we introduce power-exponent phase modulation for generating ring-shaped optical skyrmions with accelerated azimuthal texture evolution. The nonlinear phase modulation reshapes the polarization topology while preserving the annular intensity distribution and produces a spatially nonuniform and enhanced azimuthal polarization gradient. We experimentally realize conventional and power-exponent-accelerated Néel-type, Bloch-type, and Anti ring-shaped skyrmion textures and further demonstrate higher-order accelerated textures with effective skyrmion numbers up to 20. Free-space propagation measurements confirm the preservation of both the topological texture and the enhanced polarization-gradient profile. Our results establish a programmable route to intensity-invariant skyrmion-texture engineering, with potential applications in polarization-gradient optical manipulation, vectorial optical tweezers, and topological photonic encoding.
Abstract Vectorial optical vortices provide a versatile route to multidimensional light control by combining states of polarization (SOPs) with orbital angular momentum (OAM). Conventional implementations, however, typically rely on spatially interleaved metasurfaces and therefore suffer from inherent trade-offs between information capacity and spatial resolution. Here we show that vectorial interference holography implemented on a noninterleaved, spin-multiplexed all-dielectric metasurface provides an effective route to overcoming these limitations. As a proof of concept, we demonstrate a coaxial trifocal vectorial optical vortex metasurface (VOVM) that generates three channels with customizable topological charges and polarization states. We further develop an eight-channel full-polarization VOVM that establishes a one-to-one correspondence between arbitrary SOPs on the Poincaré sphere and distinct topological charges, while maintaining polarization extinction ratios above 50. Leveraging this correspondence, we implement an optical encryption clock, where 64 physically addressable SOP-OAM states are transformed through an encoding protocol into a logical space containing 256 distinct encoding states. These results establish a compact platform for structured-light engineering, high-security optical encryption and integrated photonic information processing.
Abstract Digital SERS (dSERS) improves analytical sensitivity by replacing intensity-based measurements with statistical event counting, yet substrate-level constraints continue to limit detection in the ultralow-concentration regime relevant to early-stage disease biomarkers. Here, we present a synergistic, modular platform that achieves a 9-fold enhancement in dSERS sensitivity by integrating a nanoimprinted piezoelectric polymer matrix with a gold-coated architecture. The system leverages nanoparticle-on-mirror gap modes, photonic light trapping, and active piezoelectric field modulation to increase hot-spot density without requiring redesign of the underlying nanoparticle–analyte assembly. Quantification via generalized linear models and artificial neural networks confirmed an order-of-magnitude improvement in the limit of detection relative to a standard dSERS baseline, driven by an increased density of plasmonic hot spots rather than a change in average signal intensity. Demonstrated here with R6G as a model analyte, this assay-compatible platform provides a general strategy for lowering dSERS detection limits, establishing a basis for future extension to specific disease biomarkers, where absolute detection limits will be analyte-dependent.
Abstract Spectral-selective photodetection across the visible to short-wave infrared (VIS–SWIR) range is highly desirable for advanced imaging and sensing applications. Here, we report a bias-switchable colloidal quantum-dot (CQD) n–p–n photodetector based on vertically stacked PbS and PbSe CQD heterojunctions with oppositely oriented built-in electric fields. This architecture allows electric-field redistribution under external bias, enabling controllable carrier transport and wavelength-dependent response. A ZnTe interlayer is introduced to suppress dark current and enhance bias-polarity-dependent spectral selectivity. By reversing the bias polarity, the device electrically switches between VIS + near-infrared (NIR) and NIR + SWIR detection modes within a single pixel. The photodetector exhibits low dark current densities and high detectivities of 4.3 × 1011, 2.6 × 1011, and 1.6 × 1011 Jones at 600, 1000, and 1550 nm, respectively. Integration with a silicon thin-film transistor readout circuit enables multimodal VIS–NIR–SWIR imaging, and multilayer perceptron–assisted learning further enables high-quality spectral reconstruction. These results highlight the potential of this CQD n–p–n architecture for multispectral imaging and compact computational spectrometer applications.
Abstract Advanced manufacturing processes have progressed rapidly in recent years in their ability to fabricate complex 3D structures. Most notably, light-based 3D printing processes have particularly stood out due to the various processes through which light-matter interactions can be spatially defined, a necessary criterion for high-resolution 3D printing. Here, we offer a perspective on the current state of light-based 3D printing at the micro- and nanoscale, highlighting recent works that exemplify the advances in terms of light delivery methods, processable material diversity, and cutting-edge applications. We follow this with a discussion of potential directions for furthering the potential of light-based 3D printing, pointing out what we see as largely unexplored avenues open for exploration in the field.
