The commentary discusses a recent report of metasurface-enhanced infrared photothermal (MEIP) microscopy, which simultaneously overcomes long-standing sensitivity and spatial resolution barriers in midinfrared spectroscopy, advancing detection sensitivity to the nanomolar level.

In the past two decades, optogenetic technology has developed to be the most accurate method for investigating or treating neural correlated diseases. Currently, the applications of optogenetic technology have been expanded from the initial central nervous system to the peripheral nervous system, circulatory system, locomotor system, alimentary system, urinary system, and so on. We summarize the recent progress of optogenetic technology in biomedical applications through two categories: activation or inhibition of neural impulses. The involved diseases include Alzheimer’s disease, ischemic stroke, Parkinson’s disease, epilepsy, spinal cord injury, cardiac arrhythmias, and chronic kidney disease. Furthermore, basic and clinical research in optogenetic technology for visual restoration is highlighted, and the challenges of optogenetic technology for clinical applications are discussed.

Computational imaging (CI) leverages the joint optimization of optical system design and reconstruction algorithms, enabling superior performance in terms of dimensionality, resolution, efficiency, and hardware complexity. It has found widespread applications in medical diagnosis and astronomy, among others. Recently, deep learning (DL) has changed the paradigm of CI by harnessing learned priors from data through trained neural network models. However, widely used data-driven DL-based CI methods encounter difficulties related to training data acquisition, computation requirements, generalization, and interpretability. Recent studies have indicated that integrating the physics prior of the CI system into various components of DL pipelines (including training data, network design, and loss functions) holds promise for alleviating these challenges. To provide readers with a better understanding of the current research status and ideas, we present an overview of the state-of-the-art in DL-based CI. We begin by briefly introducing the concepts of CI and DL, followed by a comprehensive review of how DL addresses inverse problems in CI. Particularly, we focus on the emerging physics-enhanced approaches. We highlight the perspectives of future research directions and the transfer to real-world applications.

Quantum theory of photons based on the first quantization technique, similar to that used by Schrödinger in the formulation of quantum mechanics, is considered. First, scalar quantum mechanics of photons operating with the photon wave functions is discussed. Using the first quantization, the wave equation, the Schrödinger-like equations, and the Dirac equation for photons are derived. Then, vector quantum mechanics of photons is introduced, which defines the electromagnetic vector fields. Using the first quantization, the Maxwell equations for photons in a magneto-dielectric medium are obtained. Because the photon’s electric and magnetic fields satisfy the Maxwell equations, all that is known about the classical optical fields can be directly transferred to photons demonstrating their quantum diversity. Relationships between the scalar and vector quantum mechanics of photons and between the Dirac and Maxwell equations are analyzed. To describe the propagation of photons in dispersive media, modified Maxwell equations are introduced.

Infrared spectroscopy with its rich vibrational information plays a crucial role in biochemical sensing. Metasurface-enhanced infrared spectroscopy amplifies the detection sensitivity by enhancing local electromagnetic fields. However, conventional far-field techniques lack the spatial resolution to image the optical response of a single plasmonic nanostructure. In this work, we introduce mid-infrared (mid-IR) photothermal microscopy to map the hot spot distribution and thermal response of a plasmonic metasurface resonant in the mid-IR spectral range. We demonstrate infrared photothermal detection of proteins and drug molecules around a single nanoantenna. Our metasurface-enhanced infrared photothermal microscope achieves a detection limit as low as 0.24 monolayer surface coverage of bovine serum albumin, paving the way for high-throughput, highly sensitive mid-IR analysis of low-abundance molecules.

Visible optical frequency combs are essential to optical atomic clocks, astronomical spectrograph calibration, and biological imaging. However, due to the limitations of the dispersion properties of available materials and the Q-factors of the optical microresonators working at visible wavelengths, the generation of the visible soliton microcomb remains highly challenging. Here, we demonstrate a chip-based visible single-soliton microcomb spanning two-thirds of an octave (ranging from 632.5 to 950.1 nm) by precisely engineering the dispersion of a silica microdisk resonator. Two dispersion waves at the spectrum edges are created by simultaneously employing the higher order dispersion and mode interaction in a microresonator. In particular, we achieve a high Q-factor with small mode volume at the short wavelength, which facilitates the generation of the visible soliton microcombs with 1.1 mW pump power. Moreover, based on the soliton self-frequency shift, we implement a precise adjustment of the dispersive wave, which makes the highest power tooth within the dispersive wave access the transition of Sr88 + for the application in optical atomic clocks.

Quantum-confined Stark effects (QCSEs), where external or built-in electric fields modify optical transition energies, have garnered significant interest due to their potential for tuning emission energies to couple with quantum dots, metasurfaces, cavities, etc. However, only external electric-field-enabled QCSEs in 2D semiconductors have been reported so far, owing to the challenges posed by small and uncontrollable built-in electric fields, as well as charge modulation effects. We report the first observation of giant built-in electric field-enabled QCSEs in 1L WSe2 / 1L graphene heterostructure (HS) with an air-gap structure that suppresses graphene screening and bandgap renormalization. Electrical control of QCSEs demonstrates a maximum Stark shift of ∼56.97 meV. This significant shift is attributed to enhanced built-in electric fields resulting from the doping-induced increase of chemical potential difference. While increasing optical doping or reducing the interlayer distance, QCSEs weaken due to reduced built-in electric fields. By leveraging efficient exciton dissociations from built-in electric fields, the responsivity (R) and response speed of HS photodetectors increase by 3 orders of magnitude and threefold, respectively, compared with 1L WSe2. Our results offer a new avenue for enhancing exciton tunability and exploring device applications of 2D materials in photodetectors, polariton transistors, and quantum light sources.

Holography plays a crucial role in optics, yet traditional methods require complex setups and bulky devices, being unfavourable for optical integration. Although metasurface-based holograms can be ultra-compact, holographic images generated by previously realized metadevices were mostly scalar ones, with a few vectorial holograms realized so far suffering from restrictions on efficiency, incident polarization, and resolution. We propose and experimentally demonstrate an efficient meta-platform to generate vectorial holographic images with high resolutions under arbitrary incident polarizations. Combining Gerchberg–Saxton algorithm and the wave-decomposition technique, we establish a generic strategy to retrieve the optical properties (e.g., reflection phases and polarization-conversion capabilities) of meta-atoms required to construct a metasurface for generating a predesigned vectorial holographic image under a predesigned incident polarization. We next design a series of high-efficiency and deep-subwavelength single-structure meta-atoms exhibiting tailored reflection phases and polarization-conversion capabilities governed by both structural resonances and the Pancharatnam–Berry effect, and experimentally characterize their optical scattering properties. We finally construct a series of ultra-thin metadevices with these meta-atoms and experimentally demonstrate that they can generate pre-designed vectorial holographic images under illuminations of circularly polarized light at 1064 nm. We provide a highly efficient and ultra-thin platform to generate predesigned vectorial holographic images under illuminations of light with arbitrary given polarization, which can inspire numerous future applications in on-chip photonics.

Robust three-dimensional (3D) recognition across different viewing angles is crucial for dynamic applications such as autonomous navigation and augmented reality; however, the application of the technology remains challenging owing to factors such as orientation, deformation, and noise. Wave-based analogous computing, particularly diffraction neural networks (DNNs), constitutes a scan-free, energy-efficient means of mitigating these issues with strong resilience to environmental disturbances. Herein, we present a real-time all-directional 3D object recognition and distortion correction system using a deep knowledge prior DNN. Our approach effectively addressed complex two-dimensional (2D) and 3D distortions by optimizing the metasurface parameters with minimal training data and refining them using DNNs. Experimental results demonstrate that the system can effectively rectify distortions and recognize objects in real time, even under varying perspectives and multiple complex distortions. In 3D recognition, the prior DNN reliably identifies both dynamic and static objects, maintaining stable performance despite arbitrary orientation changes, highlighting its adaptability to complex and dynamic environments. Our system can function either as a preprocessing tool for imaging platforms or as a stand-alone solution, facilitating 3D recognition tasks such as motion sensing and facial recognition. It offers a scalable solution for high-speed recognition tasks in dynamic and resource-constrained applications.

All-inorganic CsPbBr3 perovskite polycrystalline films, renowned for their remarkable optoelectronic properties, solution processability, and enhanced stability over organic–inorganic counterparts, are emerging as next-generation gain media for high-performance lasers. However, due to the limited understanding of how to realize population inversion under slow carrier injection, and a lack of convenient strategies to suppress Auger recombination while retaining low optical loss, achieving high-performance quasi-continuous-wave (quasi-CW) or CW lasing based on CsPbBr3 films at ambient temperature is still challenging. We devised a phase reconstruction strategy employing volatile ammonium, which achieves substantial suppression of Auger recombination through elimination of low-dimensional phase impurities and remains low optical loss via precisely controlled film crystallization dynamics. Importantly, this strategy emphasizes the critical role of Auger recombination suppression for high-performance lasing under slower carrier injection. Ultimately, an ultralow amplified spontaneous emission threshold of 9.6 μJ cm - 2 was achieved under quasi-continuous nanosecond-pulsed excitation, which facilitated the realization of a single-mode vertical-cavity surface-emitting laser with a threshold of 17.3 μJ cm - 2 and a quality factor of 3850 under quasi-CW pumping. We represent the exceptional performance of quasi-CW lasing to date, offering valuable insights for future advancements in high-performance CW lasing and even electrically driven lasers.

Convolutional operations are computationally intensive in artificial intelligence (AI) services, and their overhead in electronic hardware limits machine learning scaling. Here, we introduce a photonic joint transform correlator (pJTC) using a near-energy-free on-chip Fourier transformation to accelerate convolution operations. The pJTC reduces computational complexity for both convolution and cross-correlation from O ( N4 ) to O ( N2 ) , where N2 is the input data size. Demonstrating functional Fourier transforms and convolution, this pJTC achieves 98.0% accuracy on an exemplary Modified National Institute of Standards and Technology inference task. Furthermore, a wavelength-multiplexed pJTC architecture shows potential for high throughput and energy efficiency, reaching 305 TOPS / W and 40.2 TOPS / mm2, based on currently available foundry processes. An efficient, compact, and low-latency convolution accelerator promises to advance next-generation AI capabilities across edge demands, high-performance computing, and cloud services.