The rapid advancement of artificial intelligence (AI) imposes unprecedented speed and energy requirements on large-scale computation. Owing to its intrinsic high bandwidth, massive parallelism capacity and low energy consumption, optical computing is widely regarded as one of the most promising technologies to address AI computation demands. Here we offer a synthesis of photonic platforms that have already shown programmability, high integration density and stability, and that can support large-scale AI models, edge computing, and logic and scientific computing. We summarize the requirements that optical computing imposes on nanomaterials properties, architecture design, nanofabrication and large-scale integration. Finally, we analyse commercial systems and provide a perspective on how nanotechnology-informed solutions can improve performance in terms of computing speed, scalability and complexity in optical computing. This Review discusses photonic platforms that have already shown programmability, high integration density, and stability, and that can support large-scale AI models, edge computing, and logic and scientific computing.
Direct selective conversion of methane into high value-added chemical products has long attracted attention in both industry and academia. The production of C2 oxygenates is especially challenging owing to the requirement for both C-H activation and C-C coupling. Here we report a boron nanosheet-supported Cu single atom with Cu-B4 sites (Cu-SAs/B), which can effectively catalyse the conversion of methane to acetic acid with 97% selectivity and a high activity of 221.3 mmol gCu-1 h-1 without the addition of CO. In situ X-ray absorption fine structure analysis reveals that Cu single atoms undergo reversible 'switching' to Cu4 clusters under CH4 oxidation conditions, enabling efficient C-H activation and C-C coupling. The switching behaviour is triggered by the presence of H2O2, and the coupling occurs between CH3* and CHO* intermediates formed on Cu single atoms and Cu clusters, respectively, as confirmed by in situ spectroscopic techniques. These findings provide a proof of concept for designing highly effective methane oxidation catalysts based on switchable nanocatalysts.
Quantum condensed states in solids often reveal their fundamental nature via interactions with impurities, as epitomized by Yu-Shiba-Rusinov (YSR) bound states at magnetic impurities in superconductors. Although analogous bound states are predicted within quantum condensates of excitons, their existence has been elusive. Here we visualize in-gap electronic states bound to impurities inside an exciton condensate phase of a van der Waals crystal Ta2Pd3Te5 at 4.4 K, utilizing scanning tunnelling microscopy and spectroscopy. We find that the energies of in-gap states are correlated with the excitonic band gap, which is systematically tuned by local strain and carrier injection. Our theoretical analyses reveal that these in-gap states can be explained by charge dipoles associated with Pd atoms on Ta sites through a charge-exciton version of the YSR mechanism. Our findings introduce the analogue of YSR physics in exciton condensates and a microscopic tool to probe and control quantum properties in exciton condensates persisting up to room temperature.
Nonlinear frequency conversion underpins important technologies such as telecommunications and quantum computation; however, weak nonlinearities and architectures that resist miniaturization currently limit devices' efficiency and widespread adoption. Here we combine a band-structure-engineered GaAs/AlGaAs multi-quantum-well heterostructure with a high-quality-factor dielectric metasurface and symmetry-broken guided-mode field profiles to enhance the material nonlinear susceptibility. By engineering a resonant interband transition in the heterostructure, we realize a second-order nonlinear tensor element of 1.6 nm V-1 at 1.57 μm wavelength. We then make it free-space accessible and boost the effective nonlinearity to ~14 nm V-1 using a metasurface patterned on the material. Our proof-of-concept experiment establishes that combining interband-transition engineering and metasurfaces enables giant effective nonlinearities in the near-infrared to visible spectrum. This addresses material and device-level constraints in nonlinear photonics, providing a scalable route to compact, efficient devices.
Microplastics (MPs) are infiltrating global food systems, where they disseminate systemically, posing unmet health risks as current strategies fail to prevent bioaccumulation. We have engineered an oral alginate-chitin micro-nanofibre framework (Alg-Ch) as a pH-responsive scavenger, which was formed via lyophilization-induced hydrogen bonding and physical entanglement between alginate microfibres and chitin nanofibrous sheets (1:10 ratio). Alg-Ch captured MPs predominantly through two mechanisms: electrostatic adsorption onto a protonated chitin nanonetwork in gastric acid, and physical entrapment by swollen alginate at intestinal pH. It captured 500-nm spherical MPs of varying surface chemistry (polystyrene (PS), PS-COOH, PS-NH2) and composition (polyethylene terephthalate, polymethyl methacrylate), and irregular fragments including PS fibres, polypropylene, and polyethylene, achieving capacities of 816.6 mg g-1 (stomach) and 1114.5 mg g-1 (intestine), and retained >47% efficacy with food. In mice, Alg-Ch reduced colonic MP fluorescence by ∼50% within 2 h and accelerated faecal elimination. A 13-week Alg-Ch intervention restored tight-junction proteins ZO-1, occludin and claudin-5, decreased serum levels of interleukin-6, lipopolysaccharide, tumour necrosis factor and interleukin-1β, and promoted recovery of short-chain fatty acid-producing genera, with no evidence of body-weight loss, organ toxicity or histopathological lesions. This biocompatible platform unifies mechanical sequestration, barrier repair and microbiome rehabilitation, offering a scalable strategy to mitigate the risks of ingested MPs and the global health burden of plastic pollution.
Converting continuous spin textures into discrete hidden valleys via local symmetry breaking unlocks long-lived, room-temperature spin control in two-dimensional perovskites.
Catecholamines, including dopamine, noradrenaline (also known as norepinephrine) and adrenaline (also known as epinephrine), are essential regulators of neural and endocrine function. Although they are synthesized through a branched phenylalanine metabolic pathway, direct and dynamic monitoring of the enzymatic processes governing their production has remained challenging. Here we reconstitute the complete catecholamine biosynthetic pathway from phenylalanine to adrenaline in vitro using purified enzymes and cofactors, and couple this system with nanopore-based single-molecule sensing. By integrating two orthogonal molecular recognition modalities within a single nanopore platform, we achieve highly specific detection of all metabolic intermediates and end products throughout the pathway. Real-time monitoring enables direct observation of pathway progression and regulatory perturbations at single-molecule resolution. Using this approach, we show that 6-methylisothiocyanate disrupts catecholamine biosynthesis, providing mechanistic insight into how iodotyrosine dehalogenase 1 (DEHAL1) deficiency may compromise catecholamine production. Together, this integrated enzymology and label-free nanopore sensing platform enables time-resolved dissection of neurotransmitter metabolism and regulation, and offers a general framework for studying complex biochemical pathways with single-molecule resolution.
The acidic oxygen-evolution reaction is intrinsically sluggish and requires large overpotentials, creating a key bottleneck for proton-exchange membrane water electrolysis technology. Here we show an edge-sharing single-layer oxide, 1T-phase ruthenium oxide (1T-RuO2). The edge-sharing configuration enables parallel alignment of ruthenium 4d orbitals across adjacent RuO6 octahedral clusters, facilitating intersite electron transport, in contrast to conventional rutile-type RuO2 with corner-/edge-sharing structures. 1T-RuO2 exhibits high acidic oxygen-evolution reaction activity with a low overpotential of 77 mV at 10 mA cm-2. It also delivers a mass activity of 3 , 743 . 43 A g Ru - 1 and a turnover frequency of 23.99 s-1 at 1.50 V versus the reversible hydrogen electrode, exceeding those of rutile-RuO2 and showing highly competitive performance under the described experimental framework. In addition, 1T-RuO2 maintains a current density of ∼2.9 A cm-2 at a cell voltage of 1.70 V for over 1,100 h in a proton-exchange membrane water electrolyser.
α-Helical nanopores are attractive molecular sensors, yet their rational design and assembly remain challenging. Here we show that the single peptide pPorA, derived from porin PorACj, self-assembles into flexible α-helical nanopores, inserts in lipid membranes and exists in distinct small- and large-conductance states. By strategically incorporating unnatural amino acids, we engineered small- and large-diameter pores exhibiting single-channel conductances of 2.4 nS and 3.5 nS in 1 M KCl, respectively, while retaining a common octameric architecture. These nanopores enabled the detection of sugars, peptide enantiomers and intrinsically disordered disease proteins that form dynamic, heterogeneous assemblies. The large pores detected multiple α-synuclein (α-syn) variants, including a pathogenic Parkinson's disease-associated C-terminal deletion mutant with nanomolar affinity (KD ≈ 20 nM). Selective electrostatic trapping of the α-syn N-terminus enabled charge-resolved identification of individual α-syn species within heterogeneous mixtures. The nanopores further resolved time-dependent and inhibitor-modulated α-syn aggregation pathways from monomers to toxic oligomers and fibrils. The small pores detected humanin and superoxide dismutase peptides associated with apoptosis and amyotrophic lateral sclerosis, demonstrating tunable sensing through pore-size control. These conformationally programmable α-helical nanopores provide a versatile platform for ultrasensitive profiling of disease biomarkers.
Achieving simultaneous high selectivity and permeability for monovalent cation separation remains challenging due to energy-intensive dehydration requirements in current rigid nanopores. Here we develop a hydration-layer-mediated sieving strategy by anchoring hydrated ions at the pore rim of a covalent organic framework monolayer to form dynamic angstrom-scale hydrapores. Ion transport is regulated via attraction and repulsion between bound and migrating ions, arising from the merging and squeezing of hydration layers, enabling precise discrimination without full dehydration. Under a concentration gradient (∆C = 1 M), the membrane exhibits selectivity of 148 for K+/Li+ and 42 for Na+/Li+ with a K+ permeance of 2 × 104 mol m-2 h-1, three orders of magnitude higher than those of state-of-the-art membranes. A low activation energy of 5.5 kcal mol-1 indicates near-frictionless transport, offering a new paradigm for monovalent cation separation.
Implantable bioelectronics are typically inaccessible once implanted, and therefore structural damage, degradation and functional loss often go undetected until complications arise. Reliance on batteries further exacerbates these risks by imposing limited lifetimes, leakage hazards and non-resorbable components that may require surgical removal. Here we report a battery-free, bioresorbable triboelectric implant that enables externally readable, on-demand visualization of device status while generating electrical output. A data-driven materials-to-device workflow combines machine-learning-assisted photophysical screening with molecular-dipole-moment-based selection to identify nanoscale iridium(III) complexes optimized for both optical reporting and triboelectric charge generation. The resulting devices integrate a transcutaneous phosphorescent readout with ultrasound-driven energy harvesting, producing outlines visible to the unaided eye under handheld illumination and generating up to 3.6 Vpp at 0.5 W cm-2 of ultrasound power. In vivo studies in mice demonstrate that optical readouts track implant position, morphology and damage, correlate structural defects with loss of electrical output, and monitor integrity and bioresorption over 38 weeks. This work establishes a transient in vivo power platform whose morphology reports on structural integrity and energy-harvesting function, offering a route to observable bioelectronic implants and supporting timely intervention.
Atomically thin semiconductor junctions offer a platform for probing optoelectronic processes beyond the continuum limit, where reduced screening and strong exciton binding make local fields especially important for charge separation. Yet it remains unclear how individual dopants contribute to the photovoltaic response when the junction thickness becomes comparable with the atomic length scales and smaller than conventional depletion or diffusion lengths. Here we demonstrate this concept by probing the microscopic photoresponse of a van der Waals semiconductor homobilayer containing ionizing acceptors. Using photoconductive atomic force microscopy on vanadium-doped WSe2 (V:WSe2) bilayers, we directly visualize nanometre-scale photocurrent hotspots centred on single dopants, which have opposite current polarities for dopants in the top and bottom layers. Vertical WSe2/V:WSe2 homobilayer devices show that the macroscopic photocurrent scales linearly with dopant concentration and exhibits a compensation voltage that is independent of illumination power and dopant density, in contrast to bulk homojunction devices. Photocurrent spectroscopy and quasi-classical modelling indicate that charged dopants locally convert tightly bound intralayer excitons into charge-separated interlayer states, thereby enabling efficient exciton dissociation within a region of about 1 nm. These results establish dopant-defined point-like junctions as the elementary photovoltaic units in atomically thin homobilayers.
Curvilinear magnetism has emerged as a powerful approach to create chiral and anisotropic responses at the nanoscale, using the effects of geometric curvature and topology. This concept complements traditional material screening when tailoring material properties for specific applications. Current research in curvilinear magnetism largely focuses on mean-field micromagnetics-a computational approach to predict magnetic microstructures. The roles of inhomogeneous strains and curvature-induced magnetoelectric coupling, mediated by spin textures, have gained attention only recently. In this Review, we discuss these novel phenomena and their relevance for designing 'metageometric' materials. Advances in geometrically engineered magnetic nanostructures could lead to exciting opportunities for energy-efficient, scalable nanoelectronic devices.
The Fe(III)|Fe(IV) redox couple in iron-containing Na layered oxides enables high-capacity, cost-effective positive electrodes. However, although a high iron content (when the Fe concentration exceeds 33 at.% on transition metal layers) leads to rapid capacity decay during battery cycling, the underlying mechanism of this detrimental behaviour remains unclear. Here we report that the electrochemomechanical failure mechanism in Fe-rich Na layered oxides is related to the stability of the Fe octahedral coordination environment at the nanoscale. Fe-ion migration and dissolution govern the formation of intragranular microcracking in the positive electrode active material particles, accompanied by dislocations and an uneven distribution of mechanical stress. Driven by the non-uniform strain field, microcracks proliferate and planar gliding occurs, resulting in a stepped surface. By nanoscale doping with Al(III) (1 at.%), Y(III) (1 at.%) and Co(III) (3 at.%), we inhibit the Fe-ion migration and dissolution, thereby reducing cracks and planar gliding. Using the multi-element nanoscale-doped iron-rich sodium layered oxide at the positive electrode and a hard-carbon-based negative electrode, we assembled and tested 2.7-Ah Na-ion pouch cells showing an initial specific energy of 121 Wh kg-1 (based on the total mass of the cell) at 26 mA g-1, and a discharge capacity retention of 83.4% after 2,000 cycles at 130 mA g-1 at 25 °C.
Bacterial contractile injection systems provide a model for membrane penetration and targeted delivery of molecular cargo through mechanical actuation. Replicating these features in synthetic nanoscale systems remains challenging, particularly with respect to coupling structural organization with dynamic actuation, reversibility and spatiotemporal regulation. Here we report a DNA origami nanosyringe that integrates these capabilities to enable programmable membrane translocation. The DNA origami nanosyringe comprises two ~70-nm DNA origami bundles crosslinked by a ~10-nm gold nanoparticle. One bundle forms a cholesterol-functionalized membrane-anchoring base, whereas the other serves as a DNA fuel-driven sliding needle. After binding to supported lipid bilayers or vesicle membranes, DNA-fuel actuation drives the needle downwards in ~14-nm steps, thereby enabling membrane penetration, while reverse actuation retracts the needle and promotes membrane resealing. We show that this device provides controllable delivery of cargo tethered to the needle tip into lipid-bounded compartments and can regulate biochemical processes within cell-sized environments, including membrane-localized hybridization chain reactions, RNA transcription and catalytic RNA cleavage. These results demonstrate a strategy for constructing dynamic DNA devices that operate at membrane interfaces and coordinate mechanical actuation with biochemical function.
A nanodevice inspired by bacterial contractile injection systems enables reversible, fuel-actuated membrane penetration and spatio-temporally controlled cargo delivery for programmable biochemical regulation in synthetic cells.
3,4-dihydroxybenzoic acid (DHB), a molecule produced by the gut microbiota from dietary fibre, promotes T cell stemness and boosts the anti-tumour efficacy of immunotherapies in multiple mouse tumour models.
Gut microbial metabolites play crucial roles in regulating systemic immunity, but their mechanisms and limited drug-like properties remain unresolved. Here we report an oral nano-formulation that leverages gut microbial metabolites to modulate T cell metabolism and amplify antitumour immunity. Through an in vitro screening of gut microbial metabolites, we identified 3,4-dihydroxybenzoic acid that improved adoptive T cell therapy and enhanced CD8+ T cell stemness by suppressing glycolysis and regulating the Akt-mTORC1-Myc pathway. To harness the potency of 3,4-dihydroxybenzoic acid for systemic cancer immunotherapy, we engineered a 3,4-dihydroxybenzoic acid prodrug nano-emulsion, significantly increasing its oral absorption and half-life. In multiple murine tumour models, the oral nano-emulsion enhanced the expansion of antigen-specific, stem-like CD8+ T cells, sensitizing tumours to anti-PD-1 blockade and exerting robust antitumour efficacy. By integrating nanotechnology with microbial-metabolite-based immunotherapy, this study establishes a mechanistic link between the gut microbiota and T cell immunity, offering a promising approach for cancer immunotherapy.
