ACS Photonics最新文献

Designing free-form photonic devices is fundamentally challenging due to the vast number of possible geometries and the complex requirements of fabrication constraints. Traditional inverse-design approaches─whether driven by human intuition, global optimization, or adjoint-based gradient methods─often involve intricate binarization and filtering steps, while recent deep-learning strategies demand prohibitively large numbers of simulations (10⁵–10⁶). To overcome these limitations, we present AdjointDiffusion, a physics-guided framework that integrates adjoint sensitivity gradients into the sampling process of diffusion models. AdjointDiffusion begins by training a diffusion network on a synthetic, fabrication-aware dataset of binary masks. During inference, we compute the adjoint gradient of a candidate structure and inject this physics-based guidance at each denoising step, steering the generative process toward high-Figure of Merit (FoM) solutions without requiring meticulous binarization or filtering. We show that our method achieves approximately 15% higher FoM at equal simulation cost compared to state-of-the-art nonlinear optimizers (e.g., Method of Moving Asymptotes (MMA), Sequential Least-Squares Quadratic Programming (SLSQP)), or requires about 3× fewer simulations to reach the same FoM, all while ensuring fabrication-aware manufacturability. Compared to pure deep-learning approaches, our method requires ∼10³× fewer simulations. By eliminating complex binarization schedules and minimizing simulation overhead, AdjointDiffusion offers a simulation-efficient and fabrication-aware inverse-design algorithm with the nonconvex optimization capabilities of deep learning. Our open-source implementation is available at https://github.com/dongjin-seo2020/AdjointDiffusion.

Photoacoustic (PA) and ultrasound (US) dual-modality imaging, combining the high resolution of PA imaging and the high penetration depth of US imaging, has long been envisioned as an edge tool. This complementary duality holds transformative potential for applications ranging from lymph node characterization to intravascular diagnostics. Yet, conventional dual-modality imaging systems remain constrained by the limited integration complexity. This study pioneers a paradigm-shifting imaging system with a highly integrated single-fiber-based PA-US dual-modality transducer. The core of the system lies in a nontoxic ultraviolet glue-dye (UV-dye) transducer design, which exhibits wavelength-selective optical properties, enabling efficient US generation and PA excitation simultaneously. This transducer architecture achieves an ultrasound generation efficiency of 0.04 MPa mJ^–1 cm² with a 37.1 MHz bandwidth excited by a 532 nm laser pulse, which are 67% and 106% higher than the existing transducer, respectively. Experimental validation is conducted on the established integrated PA-US dual-modality system, in which the lateral resolutions of US and PA imaging modalities are calibrated to be as high as 90 and 125 μm, while axial resolutions are, respectively, verified as 60 and 55 μm at a depth of 2.5 mm, effectively equipped as an intelligent microscope for biological tissue imaging. Further, the system successfully reconstructs dual-modal images of ex vivo tissues that transcend traditional single-mode limitations, revealing detailed structural and chromophore information on the biological tissue. The proposed all-fiber dual-modality imaging system owns the capability of high resolution and sufficient tissue details, demonstrating its broad application prospect in angiography and oncology.

Dissipative Kerr soliton (DKS) microcombs generated in microresonators feature octave-spanning bandwidths and gigahertz-to-terahertz repetition rates, underpinning applications from coherent communications to precision metrology. Reliable deployment, however, requires not only thermally robust soliton generation in chip-scale resonators but also real-time, noise-tolerant recognition of distinct soliton states beyond the limitations of heuristic methods and conventional Fourier analysis. Here, we present a hybrid deep-learning framework that integrates convolutional neural networks (CNNs) and Transformer-based multihead attention (TMHA), enabling precise and automated classification of measured soliton spectra. Stable DKS generation with 250 nm optical bandwidth is demonstrated in high-Q silicon nitride (Si₃N₄) microring resonators (MRRs), aided by phase-modulated sideband stabilization to suppress thermal instabilities. Single-soliton, multi-soliton, and perfect soliton crystal (PSC) states are consistently produced and quantitatively characterized via an improved fast Fourier transform–genetic algorithm (FFT-GA) routine. The hybrid CNN-TMHA model achieves a classification accuracy of 98.70%, with the shortest average inference time (6.51 ms) and highest throughput (153.58 samples/s) among all evaluated models. The combination of thermally stabilized soliton generation and high-performance state recognition enables field-ready, on-device recognition that supports closed-loop control, thereby moving microcombs from laboratory demonstrations to deployable microcomb-enabled photonic systems.

As machine learning scales, its computational and energy demands increase rapidly. Reservoir computing (RC), owing to its easy training and low hardware overhead, may offer viable solutions to the growing energy costs of machine learning. However, it faces significant bottlenecks in speed and energy efficiency when handling complex tasks. Inspired by biological vision, where graded neurons achieve high sensory precision with low energy, we develop a graded-like spiking reservoir architecture leveraging a semiconductor laser with controllable carrier dynamics. This hybrid approach is well-suited for high-speed and energy-efficient neuromorphic and photonic computing. By employing an electrical injection method, a solitary laser (implemented as one element of an integrated laser array) can demonstrate graded-like dynamics, bypassing the pulse rate limitations imposed by the refractory period or feedback loop, thereby enabling high-speed processing. The laser neuron operates without external perturbations or auxiliary components, forming a simple and energy-efficient core. Based on this, we construct an RC system that experimentally achieves 95.8% accuracy in MNIST digit classification and 91.8% in discrete-time bifurcation identification. Importantly, we numerically demonstrate that input encoding strategies can be seamlessly integrated into a graded-like spiking RC framework, significantly enhancing the computational performance without added hardware complexity. Furthermore, by incorporating the quasi-convolutional encoding algorithm, the normalized mean square error on the Mackey-Glass time series prediction task is experimentally reduced from 0.0114 to 0.0063. The numerical results show good qualitative agreement with the experiment. This work presents a generalizable and scalable framework for high-speed photonic neural computation.

Understanding the mechanism of nonlinear second harmonic generation (SHG) in 2D magnetic materials is of great physics interest, with profound implications for both fundamental research and technological applications. Here, we report the first observation of abnormal thickness dependence of SHG in centrosymmetric 2D antiferromagnets: clear SHG in ultrathin 2D Cobalt Monoxide (CoO) nanoflakes and a systematic decrease of SHG with increasing thickness. As centrosymmetric crystals generally have no SHG signal, the observed SHG in ultrathin CoO nanoflakes originates from the antiferromagnetic spin orders that break the time-reversal symmetry and from defect states that break the spatial inversion symmetry. Temperature-dependent and polarization-resolved SHG studies reveal that the abnormal SHG in ultrathin CoO nanoflakes mainly originates from surface antiferromagnetic spin-ordered surface magnetic-dipole-contributed SHG, stronger than bulk magnetic-dipole-contributed SHG and surface electric-dipole-contributed SHG. Our work further suggests that polarization-resolved SHG provides a sensitive probe for studying surface magnetic properties in 2D magnetic materials.

Small-molecule donor:polymer acceptor (SMD:PA) organic solar cells have garnered attention due to their excellent active layer stability, yet their efficiency remains significantly lower than other OSC types. This study addresses the challenge of morphology control in SMD:PA systems via a layer-by-layer (LBL) process to optimize the donor–acceptor interpenetrating network. Using small-molecule donor B1 and polymer acceptor PY-IT with chloroform as a universal solvent, we systematically investigated the impact of LBL processing on the active layer morphology and device performance. The inverted LBL device (ITO/ZnO/PY-IT/B1/MoO₃/Ag) achieved a power conversion efficiency of 8.6%, significantly outperforming the bulk heterojunction devices (inverted 2.91% and normal 6.11%) and previously reported LBL SMD:PA cells (1.12%). Static and femtosecond transient absorption spectra, time-resolved photoluminescence, and grazing incidence X-ray diffraction analyses revealed that the LBL and nonorthogonal solvent strategy facilitated effective B1 infiltration into the PY-IT layer, forming an optimized active layer with refined phase separation and improved donor/acceptor interfaces, thus resulting in enhanced exciton dissociation and charge transport while reducing recombination losses. This work validates the feasibility of LBL processing for high-efficiency SMD:PA OSCs, offering a novel strategy to overcome the efficiency limitations of this class of OSCs.

Edge detection is a fundamental operation for data compression, feature recognition, and structural analysis, underpinning a wide range of scientific and technological applications. Despite recent advances, most optical analogue edge detection methods based on compact metalenses suffer from a lack of tunable directional selectivity, posing challenges for their deployment in real-world scenarios. Here, we present a compact vector vortex metalens composed of a single-layer silicon carbide metasurface for real-time, broadband, direction-selective edge detection. By engineering the superposition of spin-dependent vortex and antivortex beams, the metalens generates a point spread function with radially varying polarization states. Directional edge features are selectively extracted by introducing a linear analyzer after the metalens without requiring external Fourier optics or computational reconstruction. This directional selectivity offers the key advantage of effectively eliminating directional defects in the observed objects, which allows the contours of the objects to be better identified. We experimentally demonstrate high-resolution edge detection across a broadband spectrum for both amplitude-type and phase-type objects such as biological samples. This approach offers an ultrathin and integrable solution for next-generation optical systems that demand real-time orientation-dependent feature analysis within a minimal footprint.

Color centers in hexagonal boron nitride (hBN) emerge as promising quantum light sources at room temperature, with potential applications in quantum communications, among others. The temporal coherence of emitted photons (i.e., their capacity to interfere and distribute photonic entanglement) is essential for many of these applications. Hence, it is crucial to study and determine the temporal coherence of this emission under different experimental conditions. In this work, we report the coherence time of the single photons emitted by an hBN defect in a nanocrystal at room temperature, measured via Michelson interferometry. The visibility of this interference vanishes when the temporal delay between the interferometer arms is a few hundred femtoseconds, highlighting that the phonon dephasing processes are 4 orders of magnitude faster than the spontaneous decay time of the emitter. We also analyze the single photon characteristics of the emission via correlation measurements, defect blinking dynamics, and its Debye–Waller factor. Our room temperature results highlight the presence of a strong electron–phonon coupling, suggesting the need to work at cryogenic temperatures to enable quantum photonic applications based on photon interference.

We propose a compact strategy for the experimental synthesis of randomly structured sources with robust higher-order Poincaré polarization states by manipulating the second-order spatial coherence structure through nonuniform optical coherence engineering. This approach employs a pseudomode representation of the cross-spectral density matrix and utilizes a single amplitude-only digital micromirror device, combined with a common-path interferometer system, to construct patterns, enabling the near-real-time synthesis of sources. Experimental results demonstrate that nonuniform optical coherence engineering significantly enhances the robustness of higher-order Poincaré sphere beams, allowing them to preserve polarization integrity throughout propagation and overcoming the limitations of conventional scalar and vector uniform optical coherence engineering approaches. Furthermore, the synthesized beams exhibit strong resilience to turbulence, with the encoded polarization states preserved with high quality in the far field (focal plane). We believe that optical coherence engineering has the potential to extend the capabilities of existing resilient optical systems, offering a promising solution for compensation-free optical communication systems.