Acta Energiae Solaris Sinica•2026•DOI: 10.19912/j.0254-0096.tynxb.202608_9729
Addressing the scarcity of labeled data for training classification models in wind turbine planetary gearbox anomaly identification, this study proposes an unsupervised automated detection method. Log Mel-band energy features are extracted from raw vibration signals and fed into an unsupervised anomaly recognition model centered on a U-net autoencoder. A health-state threshold is established based on reconstruction error between model input and output, enabling anomaly identification. The method is validated using factory gearbox test data and operational data from a wind farm in Yangtouya, Shanxi. For factory gearboxes, dual validation is performed using a spectrum amplitude modulation-based signal processing method. Results demonstrate that the proposed method achieves 93.34% recognition accuracy on both factory and wind farm test sets, confirming its capability to automatically and correctly separate abnormal wind turbine gearboxes. The approach eliminates reliance on labeled fault data, offering a scalable solution for full-lifecycle health monitoring, from factory acceptance testing to in-service early anomaly detection, adaptable across different operating conditions and turbine models.
Acta Energiae Solaris Sinica•2026•DOI: 10.19912/j.0254-0096.tynxb.202608_9680
The boron-rich layer (BRL) and phosphorus-rich layer (PRL) formed during thermal diffusion in crystalline silicon solar cells are detrimental to carrier lifetime and conversion efficiency. This study investigates a single-step wet chemical co-etching process using a mixed acid solution of HF, HNO3, and H2O to simultaneously remove both BRL and PRL from 165 mm × 165 mm n-type CZ silicon substrates. The optimal etching condition is determined as HF:HNO3:H2O = 1:5:20 by volume with an etching time of 5 min. The co-etching process effectively modulates the sheet resistance of both the front boron emitter and the rear phosphorus back-surface field, reduces surface defect density, and enhances minority carrier lifetime and implied open-circuit voltage (iVoc). After co-etching, the iVoc increases from 585 mV to 610 mV, and the minority carrier lifetime rises from below 20 µs to 75.7 µs. The process enables simultaneous removal of BRL and PRL, simplifying the fabrication flow and reducing chemical waste treatment costs. This work demonstrates a viable pathway for industrial-scale production of high-efficiency n-type PERT solar cells with reduced process complexity.
SCIENCE CHINA Materials•2026•DOI: 10.1007/s40843-026-4245-4
The discrimination of volatile organic compounds (VOCs) at trace concentrations remains a critical challenge for environmental monitoring, industrial process control, and non-invasive disease diagnostics. Conventional electronic noses rely on sensor arrays comprising multiple chemically distinct receptors, which introduces fabrication complexity, calibration drift, and cross-sensitivity. Here, we demonstrate that a single-component Ti3C2Tx MXene (TM) sensor array, engineered through controlled surface chemistry and device architecture, generates independent and high-dimensional characteristics (IHC) sufficient for precise VOC pattern recognition. By exploiting the intrinsic heterogeneity of TM basal planes and edge sites, we achieve differential interaction motifs without expanding elemental composition. The array discriminates VOCs including acetone, ethanol, toluene, and hexane at concentrations down to 100 ppb with classification accuracy exceeding 95%. Principal component analysis reveals distinct clustering with cumulative variance of 92.3% captured by the first three principal components. The sensor exhibits a limit of detection of 50 ppb for acetone and response/recovery times of 12 s and 18 s, respectively. Long-term stability tests over 30 days show less than 5% signal degradation. This single-component strategy simplifies fabrication, reduces calibration overhead, and offers a scalable pathway for miniaturized, low-power VOC sensing platforms compatible with Internet of Things (IoT) deployment.
SCIENCE CHINA Materials•2026•DOI: 10.1007/s40843-025-3698-5
The oxygen evolution reaction (OER) is a critical bottleneck in next-generation sustainable energy systems due to its sluggish kinetics. Developing cost-effective, high-efficiency electrocatalysts requires understanding the dynamic structural evolution at electrode-electrolyte interfaces under operating conditions. In situ techniques are invaluable for identifying active centers and monitoring key intermediates. This review comprehensively summarizes recent advances in cutting-edge in situ methods for characterizing OER electrocatalyst structure evolution. It provides a brief overview of active motifs and robust structures using multiple in situ correlative techniques, establishing essential structure-performance relationships and updating mechanistic understanding at atomic scale under realistic conditions. Key challenges and perspectives are highlighted to promote rational design of promising electrocatalysts for efficient oxygen-associated electrocatalysis and electrosynthesis.
SCIENCE CHINA Materials•2026•DOI: 10.1007/s40843-026-4054-x
The sluggish kinetics and high onset potentials of the oxygen evolution reaction (OER) at the anode of alkaline water/seawater electrolyzers limit overall energy efficiency. Noble-metal oxides like RuO2 are active but suffer from high cost, agglomeration, and dissolution under oxidizing potentials, especially in chloride-rich electrolytes where competing chloride oxidation reaction (ClOR) occurs. Here, we report a mild two-step dry-wet milling strategy to achieve throughout lattice doping of F− into MnO2 (F-MnO2) and subsequent anchoring of atomically dispersed Ru via Ru–O/F hybrid bonds. The strengthened Mn 3d–O/F 2p hybridization and negative charge shielding of surface F− enhance OER activity/selectivity relative to ClOR and impart superior Cl− tolerance and corrosion resistance. The resulting F-Ru-MnO2-TH electrocatalyst exhibits overpotentials of 280 mV and 200 mV at 10 mA cm−2 in alkaline water and simulated seawater, respectively. It retains ~100% of initial activity after 200 h continuous operation in alkaline media and >95% after 300 h in simulated seawater, significantly outperforming Ru-MnO2 and commercial RuO2. This work provides a scalable route to durable, high-performance OER catalysts for seawater electrolysis.
Journal of Fuel Chemistry and Technology•2026•DOI: 10.1016/S1872-5813(26)60701-3
Hydrodenitrogenation (HDN) is an effective method for removing nitrogen-containing heteroatom compounds from inferior feedstocks, with the core challenge being the development of catalysts that combine low cost and high performance. In this study, a FeZn-supported catalyst was modified by introducing six different metal promoters (La, Ti, Ce, Mn, Mg, and Cr). It was found that Cr exhibited a pronounced promotional effect on HDN performance. The promoting effect of Cr on the FeZn catalyst's activity originates from its electronic interaction with sulfided Fe species, rather than functioning as an independent active site. Specifically, Cr and Zn species act synergistically as electron donors, transferring electron density to the sulfided Fe species, thereby modulating the electronic structure of Fe to render it in an electron-rich state. This increased electronic density weakens the Fe–S bonds in the active phase, promoting their cleavage and facilitating the formation of hydrogenation active sites known as coordinated unsaturated sulfur vacancies (CUS). After introducing 3% Cr, under conditions of 340–380 °C, 4 MPa pressure, and a high weight hourly space velocity (WHSV) of 8.7 h−1, the catalyst's HDN conversion rate for the basic nitrogen compound quinoline increased by 14.5%–19.7% compared to the unmodified catalyst, reaching 81.9% at 380 °C. Furthermore, Cr introduction increased the number of medium-strength Lewis acid sites, which work synergistically with the increased CUS sites to enhance overall hydrogenation activity. Cr addition effectively governs the selectivity of the HDN pathway, with the reaction rate constant for the deep hydrogenation pathway over the FeZn3Cr@GA catalyst reaching 3.2 times that of the unmodified FeZn@GA catalyst. In summary, using Fe as the primary active metal component and regulating its electronic structure through promoters represents an effective approach for designing low-cost, high-performance HDN catalysts.
Chinese Journal of Environmental Engineering•2026•DOI: 10.12030/j.cjee.202510072
Porous structures are widely used in hydraulic and pneumatic systems for flow rectification and throttling, reducing velocity, regulating pressure, and improving flow stability. Numerical simulation is a common approach to study such flows, yet the lack of standardized parameter settings often leads to user-dependent errors. This study, based on the CFD software Fluent, systematically analyzes nine key parameters across three core stages: modeling, mesh generation, and solver settings. Under both quasi-2D and 3D configurations, the influence and underlying mechanisms of each parameter on simulation results are revealed, and a reference parameter-setting method is proposed. The method is validated against wind tunnel experiments, showing that the simulated average velocity reduction ratio γS deviates from experimental values by less than 6%, confirming its reliability and applicability. This work provides a basis for standardized parameter settings in numerical simulations of porous structures, enhancing consistency and predictive accuracy.
SCIENCE CHINA Materials•2025•DOI: 10.1007/s40843-025-3429-x
Organic electrochemical transistors (OECTs) offer high transconductance and biocompatibility for wearable biosensing, yet their deployment in conformal, long-term electrophysiological monitoring is constrained by the mechanical mismatch and leakage of liquid electrolytes. This work introduces a double-network stretchable gel electrolyte that simultaneously achieves a Young’s modulus of 114 kPa and an elongation at break of 640%, matching soft biological tissues while enabling photopatterning for high-density device arrays. Integrating this electrolyte with a stretchable PEDOT:PSS channel yields solid-state OECTs with a volumetric capacitance–mobility product ([μC*]) of 317.71 ± 11.61 F cm⁻¹ V⁻¹ s⁻¹ and an average transconductance of 7.89 mS across uniform arrays. Under 50% tensile strain, the devices maintain stable electrical performance and acquire electrocardiogram signals with a signal-to-noise ratio of approximately 30 dB. The fabrication route is low-cost and compatible with solution processing, addressing the trade-off between ionic conductivity and mechanical robustness that has hindered previous gel electrolytes. These results demonstrate a viable pathway for stretchable, solid-state OECTs in ambulatory cardiac monitoring and high-resolution biointerfaces, where mechanical compliance and signal fidelity are paramount.
SCIENCE CHINA Materials•2025•DOI: 10.1007/s40843-025-3641-y
Organic field-effect transistor (OFET)-based optoelectronic synapses are pivotal for neuromorphic vision, yet polycrystalline/amorphous films suffer from grain-boundary carrier scattering and threshold instability, limiting spatiotemporal fidelity. This work employs large-area C8-BTBT single crystals to fabricate a low-voltage (1 V) optoelectronic synaptic array with a coefficient of variation of 8% in synaptic weight modulation. The grain-boundary-free structure mitigates interfacial defects, ensuring device-to-device uniformity. The array emulates human visual processing under distinct cognitive states: dispersed-attention mode (V_GS = 0.5 V) yields rapid response and short-term plasticity, while focused-attention mode (V_GS = 1.5 V) enables noise suppression and long-term potentiation via polarity-dependent carrier trapping. At 9.6 μW cm⁻² illumination, the device replicates essential synaptic functions, including learning and memory. Pattern recognition tests with six grayscale intensities demonstrate that the concentration state enhances photoresponse sensitivity and contrast discrimination, resolving fine details such as speckle patterns on bird plumage, whereas the moderate attention state fails to resolve such features. This platform advances hardware-level perception-computation integration for biomimetic vision chips, offering a pathway to energy-efficient, context-aware neuromorphic systems.