Explore the vast range of titles available.

Innovative approaches and tools play an important role in shaping design, characterization and optimization for the field of photonics. As a subset of machine learning that learns multilevel abstraction of data using hierarchically structured layers, deep learning offers an efficient means to design photonic structures, spawning data-driven approaches complementary to conventional physics- and rule-based methods. Here, we review recent progress in deep-learning-based photonic design by providing the historical background, algorithm fundamentals and key applications, with the emphasis on various model architectures for specific photonic tasks. We also comment on the challenges and perspectives of this emerging research direction.The application of deep learning to the design of photonic structures and devices is reviewed, including algorithm fundamentals.

In organic photovoltaics, morphological control of donor and acceptor domains on the nanoscale is the key for enabling efficient exciton diffusion and dissociation, carrier transport and suppression of recombination losses. To realize this, here, we demonstrated a double-fibril network based on a ternary donor–acceptor morphology with multi-length scales constructed by combining ancillary conjugated polymer crystallizers and a non-fullerene acceptor filament assembly. Using this approach, we achieved an average power conversion efficiency of 19.3% (certified 19.2%). The success lies in the good match between the photoelectric parameters and the morphological characteristic lengths, which utilizes the excitons and free charges efficiently. This strategy leads to an enhanced exciton diffusion length and a reduced recombination rate, hence minimizing photon-to-electron losses in the ternary devices as compared to their binary counterparts. The double-fibril network morphology strategy minimizes losses and maximizes the power output, offering the possibility of 20% power conversion efficiencies in single-junction organic photovoltaics. The morphology of donor–acceptor blends in organic photovoltaics dictates the efficiency of the exciton dissociation and charge diffusion, and thus the final device performance. Here, the authors show that filament assembly helps to maximize the output, further enabling a power conversion efficiency greater than 19%.

Photodetectors based on two-dimensional (2D) materials have been the focus of intensive research and development over the past decade. However, a gap has long persisted between fundamental research and mature applications. One of the main reasons behind this gap has been the lack of a practical and unified approach for the characterization of their figures of merit, which should be compatible with the traditional performance evaluation system of photodetectors. This is essential to determine the degree of compatibility of laboratory prototypes with industrial technologies. Here we propose general guidelines for the characterization of the figures of merit of 2D photodetectors and analyze common situations when the specific detectivity, responsivity, dark current, and speed can be misestimated. Our guidelines should help improve the standardization and industrial compatibility of 2D photodetectors. The lack of a standardized approach for the characterization of the performance of 2D photodetectors represents an important obstacle towards their industrialization. Here, the authors propose practical guidelines to characterize their figures of merit and analyse common situations where their performance can be misestimated.

Complex-valued neural networks have many advantages over their real-valued counterparts. Conventional digital electronic computing platforms are incapable of executing truly complex-valued representations and operations. In contrast, optical computing platforms that encode information in both phase and magnitude can execute complex arithmetic by optical interference, offering significantly enhanced computational speed and energy efficiency. However, to date, most demonstrations of optical neural networks still only utilize conventional real-valued frameworks that are designed for digital computers, forfeiting many of the advantages of optical computing such as efficient complex-valued operations. In this article, we highlight an optical neural chip (ONC) that implements truly complex-valued neural networks. We benchmark the performance of our complex-valued ONC in four settings: simple Boolean tasks, species classification of an Iris dataset, classifying nonlinear datasets (Circle and Spiral), and handwriting recognition. Strong learning capabilities (i.e., high accuracy, fast convergence and the capability to construct nonlinear decision boundaries) are achieved by our complex-valued ONC compared to its real-valued counterpart. Most demonstrations of optical neural networks for computing have been so far limited to real-valued frameworks. Here, the authors implement complex-valued operations in an optical neural chip that integrates input preparation, weight multiplication and output generation within a single device.

The realization of integrated, low-cost and efficient solutions for high-speed, on-chip communication requires terahertz-frequency waveguides and has great potential for information and communication technologies, including sixth-generation (6G) wireless communication, terahertz integrated circuits, and interconnects for intrachip and interchip communication. However, conventional approaches to terahertz waveguiding suffer from sensitivity to defects and sharp bends. Here, building on the topological phase of light, we experimentally demonstrate robust terahertz topological valley transport through several sharp bends on the all-silicon chip. The valley kink states are excellent information carriers owing to their robustness, single-mode propagation and linear dispersion. By leveraging such states, we demonstrate error-free communication through a highly twisted domain wall at an unprecedented data transfer rate (exceeding ten gigabits per second) that enables real-time transmission of uncompressed 4K high-definition video (that is, with a horizontal display resolution of approximately 4,000 pixels). Terahertz communication with topological devices opens a route towards terabit-per-second datalinks that could enable artificial intelligence and cloud-based technologies, including autonomous driving, healthcare, precision manufacturing and holographic communication.Robust terahertz wave transport is demonstrated on a silicon chip using the valley Hall topological phase. Error-free communication is achieved at a data rate of 11 Gbit s−1, enabling real-time transmission of uncompressed 4K high-definition video.

Topological insulators are a new class of materials that exhibit robust and scatter-free transport along their edges — independently of the fine details of the system and of the edge — due to topological protection. To classify the topological character of two-dimensional systems without additional symmetries, one commonly uses Chern numbers, as their sum computed from all bands below a specific bandgap is equal to the net number of chiral edge modes traversing this gap. However, this is strictly valid only in settings with static Hamiltonians. The Chern numbers do not give a full characterization of the topological properties of periodically driven systems. In our work, we implement a system where chiral edge modes exist although the Chern numbers of all bands are zero. We employ periodically driven photonic waveguide lattices and demonstrate topologically protected scatter-free edge transport in such anomalous Floquet topological insulators. Vanishing Chern numbers usually mean that a system is topologically trivial, but this rule may be violated for periodically driven systems. Here, Maczewsky et al. report topologically protected edge modes in a periodically driven photonic lattice with all bands of zero Chern number.

Next-generation light-emitting displays on skin should be soft, stretchable and bright 1 – 7 . Previously reported stretchable light-emitting devices were mostly based on inorganic nanomaterials, such as light-emitting capacitors, quantum dots or perovskites 6 – 11 . They either require high operating voltage or have limited stretchability and brightness, resolution or robustness under strain. On the other hand, intrinsically stretchable polymer materials hold the promise of good strain tolerance 12 , 13 . However, realizing high brightness remains a grand challenge for intrinsically stretchable light-emitting diodes. Here we report a material design strategy and fabrication processes to achieve stretchable all-polymer-based light-emitting diodes with high brightness (about 7,450 candela per square metre), current efficiency (about 5.3 candela per ampere) and stretchability (about 100 per cent strain). We fabricate stretchable all-polymer light-emitting diodes coloured red, green and blue, achieving both on-skin wireless powering and real-time displaying of pulse signals. This work signifies a considerable advancement towards high-performance stretchable displays. A material design strategy and fabrication process is described to produce all-polymer light-emitting diodes with high brightness, current efficiency and good mechanical stability, with applications in skin electronics and human–machine interfaces.

Exceptional points (EPs) arising in non-Hermitian systems have led to a variety of intriguing wave phenomena, and have been attracting increased interest in various physical platforms. In this Review, we highlight the latest fundamental advances in the context of EPs in various nanoscale systems, and overview the theoretical progress related to EPs, including higher-order EPs, bulk Fermi arcs and Weyl exceptional rings. We peek into EP-associated emerging technologies, in particular focusing on the influence of noise for sensing near EPs, improving the efficiency in asymmetric transmission based on EPs, optical isolators in nonlinear EP systems and novel concepts to implement EPs in topological photonics. We also discuss the constraints and limitations of the applications relying on EPs, and offer parting thoughts about promising ways to tackle them for advanced nanophotonic applications. This Review discusses the latest theoretical progress related to exceptional points in non-Hermitian physics and the associated implications for emerging technologies in nanophotonics.

Although three-dimensional metal halide perovskite (ABX3) single crystals are promising next-generation materials for radiation detection, state-of-the-art perovskite X-ray detectors include methylammonium as A-site cations, limiting the operational stability. Previous efforts to improve the stability using formamidinium–caesium-alloyed A-site cations usually sacrifice the detection performance because of high trap densities. Here we successfully solve this trade-off between stability and detection performance by synergistic composition engineering, where we include A-site alloys to decrease the trap density and B-site dopants to release the microstrain induced by A-site alloying. As such, we develop high-performance perovskite X-ray detectors with excellent stability. Our X-ray detectors exhibit high sensitivity of (2.6 ± 0.1) × 104 μC Gyair−1 cm−2 under 1 V cm−1 and ultralow limit of detection of 7.09 nGyair s−1. In addition, they feature long-term operational stability over half a year and impressive thermal stability up to 125 °C. We further demonstrate the promise of our perovskite X-ray detectors for low-bias portable applications with high-quality X-ray imaging and monitoring prototypes.X-ray detectors based on dual-site-doped perovskite single crystals exhibit excellent sensitivity of 2.6 × 104 μC Gyair−1 cm–2 under a low field of 1 V cm–1. The detectable dose rate is as low as 7.09 nGyair s–1. The operational stability is beyond half a year.

Blue/deep-blue emission is crucial for organic optoelectronics but remains a formidable challenge in organic afterglow due to the difficulties in populating and stabilizing the high-energy triplet excited states. Here, a facile strategy to realize the efficient deep-blue organic afterglow is proposed via host molecules to sensitize the triplet exciton population of guest and water implement to suppress the non-radiative decays by matrices rigidification. A series of highly luminescent deep-blue (405–428 nm) organic afterglow materials with lifetimes up to 1.67 s and quantum yields of 46.1% are developed. With these high-performance water-responsive materials, lifetime-encrypted rewritable paper has been constructed for water-jet printing of high-resolution anti-counterfeiting patterns that can retain for a long time (>1 month) and be erased by dimethyl sulfoxide vapor in 15 min with high reversibility for many write/erase cycles. These results provide a foundation for the design of high-efficient blue/deep-blue organic afterglow and stimuli-responsive materials with remarkable applications. Though realizing organic materials with deep blue emission is attractive for next-generation display technologies, achieving this emission in afterglow molecules remains a challenge. Here, the authors report blue organic afterglow via a strategy involving guest sensitization and matrix rigidification.