National Science Review最新文献

Targeted modulation of the splenic nerve offers a selective alternative to vagus nerve stimulation, avoiding off-target activation of mixed vagal fibers. However, the small diameter and anatomical complexity of the splenic nerve pose great challenges for stable neural interfacing. Here, we report a splenic nerve wireless stimulator (SpNWS) built from stretchable and highly conductive hydrogel for chronic electroceutical immunomodulation therapy. The hydrogel-fiber neural electrodes conform to the geometries of splenic neurovascular bundles (SNVBs) and are secured using a bioadhesive approach, without inducing fibrosis or immune activation in SNVBs. SpNWSs deliver wireless stimulation using subcutaneously implanted hydrogel thin film as battery-free powering electrodes with a capacitive-coupling effect. In a chronic inflammatory bowel disease rat model, splenic nerve stimulation by an SpNWS significantly reduced colitis severity and restored intestinal immune balance by suppressing T_H1/T_H17 responses and enhancing T_H2/T_reg activity. This work establishes a bioelectronic platform that combines material compliance, interface adaptability and wireless power delivery to enable organ-specific neuromodulation for electroceutical treatment of refractory diseases.

With the increasing demand for high-performance computing, 2T0C DRAM has been extensively studied for its high integration density, low power consumption and non-destructive readout. Two-dimensional (2D) semiconductors, with ultra-low leakage current, improve the retention characteristics but face limitations in conventional 2D-based 2T0C cells: the subthreshold operation of positive-threshold transistors at low write voltages reduces read current, introduces nonlinearity, and degrades robustness, and thus requires higher write voltages and increased power consumption. To address this, we propose a hybrid-gate MoS₂ 2T0C DRAM, where a low-leakage Au-gate transistor serves as the write node and a depletion-mode Al-gate transistor functions as the readout node. The device achieves >100 s retention time and reduces the minimum write voltage to 0.2 V, enabling distinguishable 3-bit storage. Furthermore, a 32 × 32 MoS₂ 2T0C DRAM circuit demonstrates image storage and readout capabilities with <5% bit error rate after 600 s, highlighting its potential for future high-density, low-power memory applications.

Photodynamic therapy (PDT), which relies on the activation of photosensitizers by specific wavelengths of light to generate reactive oxygen species (ROS) for targeted pathogen or diseased tissue eradication, offers substantial promise for clinical wound management. However, its application in diabetic wound management remains constrained by suboptimal therapeutic efficacy, recurrent infections, treatment-associated pain, scar formation, and dependence on costly specialized equipment. Here, we present a sunlight-activated nanospray formulation comprising chitosan oligosaccharide-coated nanoparticles (SPS), engineered for the rapid and effective management of diabetic wounds in outdoor and resource-limited settings. SPS enables efficient ROS generation under ambient natural light, thereby reducing treatment-associated discomfort while maintaining effective photodynamic antimicrobial action. Additionally, the chitosan oligosaccharide coating confers intrinsic hemostatic, antibacterial, and antioxidative properties that synergistically accelerate wound closure and reduce the risk of scarring. The topical spray delivery circumvents systemic phototoxicity, improving patient compliance and broadening the accessibility of PDT. This strategy effectively overcomes the inherent limitations of conventional PDT by leveraging ambient light for activation and providing a cost-effective, non-invasive, and patient-friendly platform for diabetic wound care, thereby expanding the clinical utility of photodynamic interventions for managing chronic and complex wounds.

Zeolites have undergone a significant development, leading to their widespread application in catalytic processes. China plays a pivotal role in the rapid growth of pioneering zeolite research. The National Natural Science Foundation of China (NSFC) has facilitated the transition of Chinese scientists from a position of research followers to a leading role on the global stage. The future development of zeolite science and technology in China depends on the sustained provision of innovations, advanced techniques and effective scientific funding from the Chinese government, including the NSFC. This paper firstly provides a comprehensive review of the funding projects in the field of zeolite research from the NSFC. Subsequently, a spotlight on new zeolite synthesis, novel characterization techniques and zeolite applications is presented to reveal the significance and effectiveness of these fundings. Finally, the future prospects of the NSFC supports in zeolites are envisaged.

This work proposes a unified neuromorphic spike-based large-language-model (NSLLM) framework to simultaneously address the challenges of high energy consumption and low interpretability in LLMs. Our framework transforms LLMs into efficient NSLLMs by converting their behaviors into neural dynamics-such as spike trains-through rigorous mathematical modeling and complemented by advanced techniques including quantization and sparsification. This transformation also enables the analysis of information encoding processes using computational neuroscience tools, thereby offering a novel neuroscientific perspective that conceptualizes LLMs as neural populations to enhance their interpretability. Leveraging a hardware-algorithm co-design paradigm, an NSLLM can completely eliminate matrix multiplication (MatMul) while maintaining high performance. We designed a custom MatMul-free hardware core on the VCK190 field-programmable gate array to validate the 1.5-billion-parameter NSLLM, achieving a dynamic power consumption of only 13.849 W and an inference throughput of 161.8 tokens per second. Compared with the A800 GPU, this implementation improves energy efficiency, memory usage and inference throughput by 19.8[Formula: see text], 21.3[Formula: see text] and 2.2[Formula: see text], respectively. This work provides a novel perspective within a unified framework to enhance both the energy efficiency and interpretability of LLMs, offering valuable insights for future neuromorphic chip designs tailored for large models.

Colloidal quantum dots (CQDs) are promising materials for constructing 'second-window' near-infrared (1000-1700 nm) light-emitting diodes (NIR-II LEDs), but their practical application has been hampered by low film external quantum efficiency (EQE). Here, we report a chemical strategy that incorporates photoactive fluorophores-spanning fluorescence, phosphorescence and thermally activated delayed fluorescence-into CQD films to boost NIR-II emission. Energy transfer from fluorophores (via both singlet and triplet pathways) raises the photoluminescence quantum efficiency of CQD to 85% beyond 1000 nm. As a result, these composite films power NIR-II LEDs with a record EQE of 25.3% for emission of >1000 nm, the highest among all LEDs with emission of >1000 nm. We further demonstrate the scalability of the approach by fabricating large-area (30 mm × 30 mm) NIR-II LEDs with uniform high performance.

With the advent of the Internet of Things generation, functional electronics have rapidly emerged as a focus in materials science and electronic engineering. Ionogels have gained significant attention due to their unique physicochemical properties-such as non-volatility, excellent thermal and electrochemical stability, mechanical properties and ionic conductivity-making them crucial materials in flexible electronics. This work focuses on reviewing research progress in the design of ionogels with diverse functionalities and their applications in flexible electronics (primarily over the past five years), compared with previously published reviews. In this review, we comprehensively introduce the fundamental composition and structure of ionogels based on their structure-property-application relationships. Further discussion highlights innovative applications of ionogels in sensors, energy-harvesting devices, storage devices and smart devices, emphasizing their broad potential in flexible electronics. Finally, we propose future directions for ionogels, including multifunctional integration, long-term stability and self-healing capabilities, intended to provide guidance and insights for researchers in flexible electronics.

The increasing imperative to mitigate greenhouse gas emissions and foster the transition to a low-carbon bioeconomy has intensified interest in methane bioconversion as a sustainable approach for transforming methane into valuable bioproduction. Although advancements have been made in optimizing methanotrophic pathways to improve bioproduction, significant challenges persist, including methane solubility, bioavailability, and metabolic flexibility, limiting the efficiency of methane bioconversion. This review provides a comprehensive overview of the initiatives aimed at developing next-generation methanotrophic cell factories by overcoming the physiological limitations of natural methanotrophs. We first analyze the metabolic characteristics of methanotrophs for assimilating methane into cellular building blocks. Then, we discuss methane assimilation pathways and their unique characteristics in matter and energy transmission for facilitating the integration of methane into central carbon metabolism. Further, we propose a systematic framework for designing methane-based biomanufacturing to enable low-carbon bioproduction by integrating synthetic biology, metabolic engineering, and systems biology, thereby developing efficient methane assimilation cell factories for producing high-value bioproducts. Finally, we prospect the potential for valorizing methane derived from anthropogenic emissions and renewable sources, while identifying the key challenges and future research directions necessary for advancing a sustainable, low-carbon bioeconomy.

As the cornerstone of structural chemistry, the elemental compositions and spatial arrangements of atoms determine the functionalities of compounds. This principle is fully epitomized by 'magic' angle materials, where the lattice twisting extends the periodicity of the Moiré superlattice, revealing many unexpected properties. Here, we investigate how the extended lattice periodicity affects the properties of lattice dynamics, with a primary focus on thermal conductivity. Through the modulation of bond length and angle, the lattice periodicities of binary iodides CsI, BaI₂, BiI₃ and TeI₄ are extended as the cationic valences increase from monovalent to tetravalent states, leading to a substantial decrease in thermal conductivity. It is revealed that even in a simple binary compound like TeI₄, an extremely low thermal conductivity of 0.17 W m^-1 K^-1 at room temperature can be achieved. Compared to CsI, BaI₂ and BiI₃, the superior heat insulation of TeI₄ is found to stem from the large extended periodicity of the atomic arrangement enabled by having nearly an order of magnitude more atoms in the primitive cell.