Osteoarthritis (OA) is a whole-joint disease characterized by progressive structural degeneration and chronic low-grade inflammation affecting the cartilage, synovium, subchondral bone, and immune compartments. It is a complex degenerative disorder associated with substantial morbidity, heterogeneous clinical trajectories, and limited disease-modifying treatment options. Extracellular vesicles (EVs) have emerged as important mediators of intercellular communication within the OA joint microenvironment, and are implicated in pathophysiological responses to mechanical stress, inflammatory cues, and metabolic dysfunction. Through the transfer of context-dependent nucleic acid, protein, and lipid cargoes, EVs can amplify pathogenic processes in OA such as synovitis, cartilage catabolism, and cellular senescence, while also supporting reparative pathways. Understanding the mechanisms governing EV biogenesis, cargo selection, tissue targeting, and functional heterogeneity offers opportunities to identify mechanistically informed biomarkers and therapeutic strategies. This review discusses emerging concepts in EV-mediated joint communication, highlights translational potential and limitations, and outlines key priorities for advancing EV-based diagnostics and therapies in OA. However, EV-based strategies remain largely at the preclinical or experimental stage, and rigorous validation is required before they can enter routine clinical practice.

Materials databases are increasingly the backbone of data-driven discovery for energy materials. In this Perspective, we map the ecosystem of computational and experimental databases, and argue that database architecture, which covers ingestion, curation, metadata, provenance, and access interfaces, strongly influences the performance and trustworthiness of modern AI models. We classify computational repositories into bulk-property and surface/interface resources, and summarize representative experimental databases spanning crystal structures, catalysis, energy storage, and characterization. Beyond single-modality repositories, we highlight integrated platforms that connect computed descriptors with context-rich experimental evidence and tool interfaces, enabling iterative hypothesis testing and closed-loop validation. Building on these examples, we propose a database-model-experiment roadmap for training and deploying graph neural networks, machine learning interatomic potentials, and large language model-based AI Agents. Finally, we outline key bottlenecks that must be addressed for reliable autonomous discovery, including FAIR (Findable, Accessible, Interoperable, Reusable)-aligned standardization, bias and missing negative results, and cross-code reproducibility.

Electrocatalytic reduction of nitrate for the synthesis of high-value chemical hydroxylamine (NH2OH) represents a sustainable pathway that combines the dual functions of green synthesis and wastewater treatment. However, the selective control of the process is extremely challenging, mainly due to the susceptibility of nitrate to overreduction during the reaction, leading to uncontrolled conversion of the target product NH2OH to the thermodynamically stable product ammonia (NH3). In this paper, the Cu-PDA MOF precursor was synthesized using the emulsion method, and ZIF-8 was introduced to form the structure of Cu-PDA@ZIF-8. Subsequently, the nitrogencarbon-doped Cu-ZnO/N-C catalysts were formed by calcination under N2 atmosphere. The synthesized catalyst effective improved NH2OH selectivity during nitrate reduction. The NH2OH yield and FE efficiency of Cu-ZnO/NC reached 564.97 mu mol h- 1 mgcat-1 and 46.64 %, respectively, under neutral electrolyte conditions. Cu, as the main component of the catalyst, plays a dominant role in hydroxylamine by electrocatalytic nitrate reduction to promote the reaction in the stage of NO3-->*NH2OH. Zn, on the other hand, plays a key role in the phase *NH2OH -> NH2OH, selectively generating the product NH2OH and blocking the excessive reduction of NH2OH to NH3. Additionally, this study achieved tandem catalytic selective generation of NH2OH at copper and zinc heterogeneous interfaces and offers a potential approach and reference for applications involving the nitrate reduction reaction.

Platinum‐based catalysts are highly effective for the oxygen reduction reaction (ORR), but their prohibitive cost and insufficient activity impede large‐scale commercialization. Herein, we report a Pt@Fe‐NC electrocatalyst synthesized by depositing uniform platinum nanoparticles onto an iron–nitrogen–carbon (Fe–NC) support via an ethylene glycol reduction method. The Fe–NC support, prepared from an iron (II)‐1,10‐phenanthroline complex precursor to ensure high iron utilization, modulates the electronic structure of the Pt nanoparticles, thereby enhancing both catalytic activity and stability. The optimized Pt@Fe‐NC catalyst (13.54 wt% Pt) exhibits exceptional ORR performance in acidic media, with a half‐wave potential of 0.852 V and notable stability. When integrated into a zinc‐air battery, the catalyst delivered a high specific capacity of 652.66 mAh g Zn −1 . Furthermore, a proton exchange membrane fuel cell (PEMFC) employing this catalyst achieved a high open‐circuit voltage (OCV) of 0.964 V and a peak power density of 1.722 W cm −2 , outperforming most previously reported Pt‐based catalysts. This study highlights a synergistic strategy between Pt nanoparticles and metal‐nitrogen‐carbon (M–N–C) supports to boost ORR performance, presenting a viable path toward advanced, cost‐effective catalysts for energy conversion devices like PEMFCs and zinc‐air batteries.

Hydrogen is regarded as an attractive secondary energy carrier because of its high gravimetric energy density and environmentally benign conversion products. However, the practical application of hydrogen energy is still strongly limited by the lack of safe and efficient hydrogen storage materials. Among various solid-state hydrogen storage candidates, MgH2 has attracted extensive attention due to its high theoretical hydrogen capacity, natural abundance, and low cost. Nevertheless, its practical use is hindered by high thermodynamic stability and sluggish hydrogen absorption/desorption kinetics. In this work, a rod-like Fe3O4/FeNi3 heterointerface catalyst regulated by multivalent Mo and V species was synthesized via a one-step hydrothermal route followed by calcination, and its catalytic effect on the hydrogen storage behavior of MgH2 was systematically investigated. The results show that the addition of the catalyst markedly improves the hydrogen sorption performance of MgH2. In particular, the MgH2 + 5 wt% catalyst composite absorbs 5.70 wt% H2 within 1 min and desorbs 6.52 wt% H2 within 10 min at 593 K, indicating a substantial enhancement in sorption kinetics. Meanwhile, the onset dehydrogenation temperature is significantly reduced from 570 K for pristine MgH2 to 463 K for the catalyst-containing sample, and the apparent dehydrogenation activation energy decreases from 132.66 to 69.85 kJ mol-1 H2. Structural and kinetic analyses suggest that the enhanced performance mainly originates from the synergistic effect of the Fe3O4/FeNi3 heterointerface, which facilitates H2 dissociation, interfacial hydrogen transfer, and Mg-H bond activation. The Mo/V species are considered to act primarily as electronic regulators at the surface/interface region rather than as separate bulk crystalline phases. This work provides useful insight into oxide-metal heterointerface engineering for improving the hydrogen storage properties of MgH2.

Improving the overall kinetics of the alkaline hydrogen evolution reaction (HER) is crucial for practical applications such as anion exchange membrane water electrolysis (AEMWE). However, the overall catalytic efficiency remains limited because most existing strategies focus only one elementary step, either water dissociation or hydrogen adsorption. Herein, we propose an auxiliary-driving strategy by incorporating VO2 around Ru active sites to consecutively optimize Volmer and Heyrovsky steps. The formation of V-O-Ru conjugated pi-bonds promotes water dissociation by dynamically modulating the electronic structure. Meanwhile, reversible hydrogen spillover optimizes the hydrogen adsorption free energy (Delta G H*), eventually positioning the catalyst within the optimal region of the two-dimensional microkinetic volcano model. This approach delivers an overpotential of 12 mV at 10 mA cm-2 and a high turnover frequency (TOF) of 12.2 s-1, with improved HER activity relative to Ru/C and Pt/C under identical testing conditions. Furthermore, the practicality of this effect is demonstrated by the distribution of relaxation time (DRT) analysis in an AEMWE device, providing mechanistic guidance for designing electrocatalysts for alkaline HER.

Methane pyrolysis via molten catalysts offers a transformative route for coke-free hydrogen production and high-value carbon capture. However, the development of molten catalysts is hindered by a vast compositional space and the disordered atomic structure of the molten state, which makes traditional trial-and-error experimentation inefficient. Here, we introduce an artificial intelligence-empowered digital catalysis platform (DigMethpy) to accelerate the development of molten catalysts. This platform integrates experimental and computational data with machine learning models, literature-based knowledge bases, and large language models, forming a closed-loop workflow of “data → model → prediction → validation”. It provides a data-centric framework for intelligent catalyst design by iteratively refining prediction models and intelligent agents through data feedback. The platform is poised to evolve from a single-agent workflow toward multi-agent collaboration and a self-driving system, offering a scalable digital infrastructure to connect the research community and accelerate the industrialization of methane pyrolysis.

Structural optimization is a fundamental step in density functional theory (DFT) calculations, typically driven by the Broyden-Fletcher-Goldfarb-Shanno (BFGS) optimizer. However, the standard BFGS algorithm relies on a local quadratic approximation of the potential energy surface (PES), which frequently breaks down in highly non-quadratic regimes typical of complex surface adsorption systems and defective bulk materials. This breakdown leads to "Hessian pollution", a phenomenon where higher-order anharmonicities introduce spurious off-diagonal inter-atomic couplings that distort curvature estimates and significantly stall convergence. Herein, we propose a physics-inspired algorithmic intervention to the BFGS method that systematically suppresses this pollution. Once the maximum residual force drops below a specific activation threshold (e.g., 0.5 or 0.1 eV/& Aring;), our approach conditionally resets all off-diagonal Hessian blocks, and introduces an isotropic background stiffness strategy where these blocks can be repopulated with a small positive constant rather than zeroed completely. This balances the robust stability of diagonal dominance with accelerated convergence speed. Implemented as an add-on to the Atomic Simulation Environment (ASE) Library, the method is lightweight, transferable, and compatible with standard DFT codes. Tests across diverse chemical systems, including atomic and molecular adsorbates (O*, H*, CO*) on Pt(111) surfaces and defective bulk oxides (WO3-x), demonstrate substantial reductions in the number of required force calls without biasing the final optimized geometry. It offers a practical tool for high-throughput DFT workflows that eliminates the need for domain-specific training. This method is available via our open-source package, Hessian-Engineered Relaxation Optimizer (HERO).

Solid electrolytes (SEs) are central to next-generation metal batteries, yet their discovery remains constrained by fragmented data, limited transferability of simulations, and slow experimental iteration. Unlike catalysis, where surface reactivity dominates, SEs require simultaneous optimization of bulk ion transport, defect chemistry, mechanical integrity, and interfacial stability. Here, we outline a framework for autonomous SE discovery enabled by large artificial intelligence (AI) models, including machine learning interatomic potentials (MLIPs) and large language models (LLMs). We discuss the evolution from static materials databases to dynamic, self-updating knowledge systems, the role of MLIPs in bridging density functional theory (DFT) and long-timescale ion migration, and the emergence of LLMs as engines for literature mining, hypothesis generation, and scientific reasoning. We further describe a closed-loop architecture integrating AI-driven candidate design, multiscale simulation, uncertainty-aware selection, and experimental validation. Such systems shift SE research from intuition-guided exploration to data-informed, self-improving cycles. We conclude by highlighting challenges in data standardization, interfacial complexity, and reproducibility, and we propose design principles for building autonomous laboratories for solid-state battery materials.

Solid-state batteries are limited by an interfacial trade-off: interfacial reactions are necessary to establish ion-transport pathways, yet excessive reactions lead to passivation and polarization. Here, we establish a descriptor-guided framework for heterogeneous metal anodes by integrating interfacial energy, phase-to-phase Volta potential difference, and microstructural percolation. This framework identifies Mg2Sn as the optimal secondary phase, with an interfacial formation energy of -0.55 J m-2 and a moderately positive phase-to-phase Volta potential difference that together enable controlled interfacial reactivity. A percolated Mg2Sn network together with continuous alpha-Mg pathways delivers a stripping current more than 400 times that of pure Mg and stable stripping/plating for over 1300 h at 0.1 mA cm-2, representing a record cycling performance for solid-state Mg metal anodes. This descriptor set provides a general framework for secondary-phase engineering in solid-state batteries.

ABSTRACT Nitrate (NO 3 − ) pollution threatens aquatic ecosystems and drinking water safety, while electrocatalytic nitrate reduction to ammonia (NO 3 RR) offers a route for pollutant removal and nitrogen recovery. Here, density functional theory calculations were used to screen 100 graphene‐supported asymmetric M1N 3 ─M2N 4 dual‐atom catalysts (M1, M2 = 3 d transition metals) by considering stability, pristine‐site availability, reaction pathways, and ammonia desorption. Single‐H adsorption free energy alone was insufficient to describe competition with the hydrogen evolution reaction in dual‐site systems; *2H and *OH surface states should also be considered when determining catalyst‐specific electrochemical potential windows. NO 3 − mainly adopted side‐on@bridge and side‐on@M1 configurations, leading to site‐dependent pathway branching. Zero or near‐zero U L values often coincide with strong NH 3 binding. NiN 3 ─ZnN 4 ‐P1 and CuN 3 ─CoN 4 ‐P2 displayed balanced profiles, with limiting potentials of −0.29 and −0.23 V and NH 3 desorption free energies of 0.12 and 0.35 eV, respectively. These results provide theoretical guidance for evaluating and designing asymmetric dual‐site NO 3 RR catalysts by jointly considering site availability, pathway thermodynamics, and product release.

Sodium iron pyrophosphate (NFPP) is a cost-effective and safe cathode material for sodium-ion batteries (NIBs). However, NFPP suffers from low electron conductivity and sluggish Na+ diffusion kinetics, resulting in a poor rate capability. Here, we show that partial co-substitution of V3+ and Al3+ for Fe2+ synergistically enhances the rate capability of NFPP with improved Na+ storage capacity. The V3+/Al3+ co-substitution narrows the band gap energy of NFPP from 2.99 to 1.53 eV, thereby increasing electron conductivity. It also reduces the Na+ diffusion energy barrier from 0.529 eV to 0.398 eV, enhancing Na+ diffusivity, leads to the formation of Na vacancies to further facilitate Na+ transport, and improves Na+ storage capacity due to the enhanced electroactivity of the Fe2+/Fe3+ couple. At 0.1, 5 and 10C, the Na+ storage capabilities of the V3+/Al3+ co-substituted NFPP cathode are 120, 85 and 72 mAh g-1, respectively. At 5C, this electrode maintains 81% capacity after 1000 cycles. A full cell fabricated with the co-substituted NFPP as the cathode and a commercial hard carbon (HC) as the anode delivers a Na+ storage capacity of 87 mAh g-1 with 77% capacity retention after 100 cycles at 1C, outperforming a counterpart full cell fabricated with NFPP as the cathode under the same anode and electrolyte conditions.

Magnesium hydride (MgH2) offers high gravimetric hydrogen storage capacity and good reversibility, but its high dehydrogenation temperature and sluggish kinetics limit practical applications. In this study, an MgH2 (110)/γ-graphdiyne heterojunction model was developed using density functional theory (DFT), with Pd, Pt, Ru, Rh, Ir, and Os atoms intercalated into the interfacial gap to evaluate the dehydrogenation pathway and kinetics. Results show that the γ-graphdiyne-induced interfacial charge redistribution weakens the Mg-H interaction, lowering the dehydrogenation barrier from 2.54 eV to 2.12 eV. Noble metal intercalation further reduces the barrier, with the Rh-intercalated system showing the lowest value of 0.63 eV and a decreased reaction energy. Interestingly, the dehydrogenation barriers correlate well with reaction energies, following the Bell-Evans-Polanyi relationship (R2 = 0.93). Furthermore, the noble metals' d-orbital electron count exhibits a segmented linear correlation with the barrier, providing a screening descriptor and predicting a barrier of approximately 1.32 eV for Ag- and Au-intercalated systems. This study offers insights into interfacial design and metal screening for catalytic dehydrogenation systems based on MgH2.

ABSTRACT Controlling reaction pathways in electrocatalytic biomass upgrading remains challenging because mass transport, substrate adsorption, and elementary kinetics are intrinsically coupled within catalyst architectures. Here, we report a ligand‐intercalation strategy that enables selective reaction‐pathway engineering in layered metal–organic frameworks (MOFs) by decoupling effects of steric and electronic microenvironments. Aromatic dicarboxylate ligands with systematically varied length and π‐electron density are intercalated into NiCo‐based MOFs to create tunable interlayer nanochannels that independently regulate molecular diffusion and substrate–catalyst interactions. Expanded interlayer spacing enhances alcohol oxidation by improving mass transport and active‐site accessibility, whereas π‐electron‐rich ligands selectively promote aldehyde oxidation through strengthened π–π interactions and accelerated hydrogen atom transfer (HAT), resulting in a shift of the rate‐determining step (RDS) from a chemical to an electrochemical step. These orthogonal effects are quantitatively correlated with kinetic analysis, impedance spectroscopy, adsorption measurements, in situ spectroscopy, and density functional theory calculations. As a result, the optimized MOFs deliver low onset potentials, current densities up to 200 mA cm −2 , and near‐quantitative Faradaic efficiencies and product yields in the selective oxidation of representative biomass substrates, 5‐hydroxymethylfurfural and 2,5‐diformylfuran. This work establishes ligand‐intercalated MOFs as a versatile platform for microenvironment‐driven reaction‐pathway control in electrocatalytic biomass valorization.

ABSTRACT The selective oxidation of n‑butane to maleic anhydride (MA) over vanadium phosphorus oxide (VPO) catalysts is critically dependent on the synergy between V 4+ ‑containing phases (VO) 2 P 2 O 7 and V 5+ ‑containing phases VOPO 4 . This reaction has long been hindered by a trade‑off between conversion and selectivity. In this study, ultrathin VOPO 4 nanosheets with varying mass fractions (0.1–0.7 wt%) were attached to the VOHPO 4 ·0.5H 2 O precursor, enabling systematic control over the surface V 5+ /V 4+ ratio and lattice oxygen content in the V 5+ phase. An optimal VOPO 4 loading of 0.3 wt% increased the lattice oxygen content in the V 5+ phase from 3.85 to 6.76 mmol/g, raising n‑butane conversion from 85.6% to 95.0% and achieving a MA yield as high as 56.0%. Structure–activity relationship studies indicate that the amount of lattice oxygen in the V 5+ phase is the primary factor regulating n‐butane conversion, and its effect on selectivity depends on whether the lattice oxygen in the V 4+ phase sites is over‐reduced. The balance between these two factors is fundamental to achieving high performance. However, excessive VOPO 4 loading (≥ 0.5 wt%) leads to over‑reduction of V 4+ ‑associated lattice oxygen, thereby disrupting the selective oxidation pathway. This study provides theoretical guidance for optimizing VPO catalysts through precise regulation of the distribution state of V 5+ species.

Microfracture (MF) is widely used for cartilage repair, but it often yields limited clinical benefit because it predominantly induces fibrocartilage formation. However, the mechanisms underlying this fibrocartilaginous repair remain insufficiently defined. In this study, temporal histopathological profiling and integrative bioinformatic analyses of post-microfracture specimens revealed that early extracellular iron accumulation was positively associated with ferroptosis severity. Single-cell RNA sequencing further delineated a bifurcating differentiation trajectory of bone marrow mesenchymal stem cells (BMSCs) toward either hyaline-like chondrocytes or fibrocartilaginous chondrocytes, with ferroptosis acting as a critical regulator at the branch point. Given the antioxidant and ferroptosis-modulating activity of curcumin-derived components, we hypothesized that curcuma-derived extracellular vesicles (CDEVs) could suppress ferroptosis and bias the differentiation of BMSCs toward hyaline cartilage. Mechanistically, in vitro assays identified Pvu-miR-159 in CDEVs as a key functional cargo that attenuates ferroptosis via the PTPN12-ERK1/2-ATF4-GPX4 pathway. To enhance its translational potential, we developed an injectable reactive oxygen species (ROS)-responsive hydrogel enabling sustained CDEVs delivery, which effectively reduced ferroptosis and promoted cartilage regeneration in vivo. Together, these findings uncover a ferroptosis-driven mechanism contributing to suboptimal microfracture repair and support a plant-vesicle-based, ROS-responsive delivery strategy to reprogram BMSCs toward improved regenerative outcomes.

Hydrogen storage remains a central bottleneck for scalable hydrogen energy systems due to the multiscale and coupled nature of the thermodynamics, kinetics, and microstructural evolution of hydrogen storage materials (HSMs). Although artificial intelligence (AI) has accelerated materials discovery, current approaches remain constrained by fragmented data, limited physical consistency, and weak integration with experimental validation. Here, we propose a unified framework that integrates coherent data infrastructure, physics-grounded modeling, and AI-driven inverse design within a closed-loop discovery paradigm. By embedding physical constraints and experimental feedback, this approach enables adaptive, physically consistent optimization, thereby establishing a pathway toward autonomous, digital-twin-enabled discovery of HSMs.

The design of a low-Ir-loading anode catalyst with high activity and stability is crucial for the proton exchange membrane water electrolysis (PEMWE), yet it remains a formidable challenge. Herein, an ordered Ba2EuIrO6 double perovskite is demonstrated as a promising anode material for catalyzing oxygen evolution reaction (OER) in acid electrolyte. The Ba2EuIrO6 achieves a low overpotential of 250 mV at 10 mA cm-2 and high mass activity with 1.39 A mg-1 toward OER, outperforming BaIrO3 and commercial IrO2 catalysts. It is discovered that the oxygen bridged Ir─Obri─Eu unit in Ba2EuIrO6 plays a critical role as the catalytically active center. In situ spectroscopic studies, isotope labeling measurements and theoretical calculations reveal that the Ir─Obri─Eu units possess strong proton affinity for proton capture from OOH* and OH*, triggering the bridging oxygen-mediated deprotonation mechanism to break traditional scaling relationships during the OER. Furthermore, the incorporation of Eu modulates the Ir dz2 orbital to increase the spin density of adsorbed oxygen, accelerating ─OH attack and reducing the energy barrier for OOH* formation. The Ba2EuIrO6-loading PEMWE delivers over 1.0 A cm-2 at only 1.67 V and operates stably for 350 h at 1.0 A cm-2, demonstrating its good potential for practical applications.