The photoreduction of environmental contaminants such as nitrate (NO₃⁻) and carbon dioxide (CO₂) into clean and renewable fuels has emerged as a key strategy for mitigating global environmental challenges, in which perovskite photocatalysts offer a promising, cost-effective, and sustainable solution. In the current research, a novel CuBi₂S₄/Al₂WO₆/Ti₃C₂ MXene Schottky/Z-scheme ternary heterojunction photocatalyst was synthesized and developed for the efficient photoreduction of nitrate and carbon dioxide, as well as photocatalytic water splitting under visible-light irradiation. The nanocomposite integrates three distinct components: (i) zero-dimensional (0D) CuBi₂S₄ quantum dot (QDs) nanoparticles (acting as a metal-assisted sulfide perovskite photocatalyst), (ii) three-dimensional (3D) aluminum tungstate (Al₂WO₆) double perovskite (serving as the central oxide perovskite photocatalyst), and (iii) two-dimensional (2D) Ti₃C₂ MXene (functioning as a non-metallic co-catalyst facilitating interfacial charge transfer). A comprehensive assessment of operating factors revealed their significant influence on the photocatalytic behavior of the CuBi₂S₄/Al₂WO₆/Ti₃C₂ ternary photocatalyst. The CuBi₂S₄/Al₂WO₆/Ti₃C₂ photocatalyst achieved a nitrate reduction efficiency of 80%, with nitrogen gas (N₂) identified as the predominant reduction product (55% selectivity). The same catalyst also exhibited a CO₂ photoreduction efficiency of 70%, in which methane (CH₄) displayed the highest generation rate (13.87 mL∙g⁻¹∙h⁻¹; 619 μmol∙g⁻¹∙h⁻¹) corresponding to a 50% selectivity. Moreover, the composite demonstrated an impressive hydrogen evolution rate of 16 mL∙g⁻¹∙h⁻¹ (714 μmol∙g⁻¹∙h⁻¹) during photocatalytic water splitting with an efficiency of 60%. Furthermore, the ternary heterojunction photocatalyst exhibited excellent reusability and structural stability, retaining its photocatalytic performance over five consecutive cycles.

The start-up performance of proton exchange membrane fuel cells in low-temperature environments directly affects their service life and market promotion prospects. However, it is still challenging to fully understand how different operating parameters synergistically intensify the cold startup efficiency of proton exchange membrane fuel cells. In this study, the cold-start performance of proton exchange membrane fuel cells is optimized via cathode catalytic H₂-O₂ reaction heating, integrated with machine learning for key indicator prediction and multi-objective optimization for operating parameter screening. The proposed strategy achieves a temperature rise exceeding 30 °C without external load at –20 °C, suppressing the peak ice volume fraction in the cathode catalyst layer to 3.28 vol % and ensuring post-start stability. Machine learning models can predict key cold-start indicators with high precision. SHapley Additive exPlanations analysis further reveals the complex nonlinear interactions between parameters and clarifies the key factors affecting cold-start performance. Non-dominated Sorting Genetic Algorithm-II optimization identifies Pareto-optimal solutions, demonstrating enhanced cold-start efficiency via synergistic regulation of reactant supply, temperature elevation, controlled anode back pressure, and coolant flow. These findings provide guidance for the engineering design and parameter regulation of proton exchange membrane fuel cells in cold-climate applications.

This study develops a two-dimensional fluid model for atmospheric pressure non-equilibrium CO₂-H₂O plasma needle-plate configuration, incorporating a comprehensive set of plasma chemical reactions and photoionization effects. It focuses on investigating the influence of the CO₂/H₂O concentration ratio and quenching pressure on plasma streamer initiation and propagation dynamics. Numerical simulations show that increasing initial water vapor content significantly reduces electron energy and density, causing the discharge channel to contract when the reduced electric field is below 200 Td, due to strong dissociative adsorption reactions between electrons and water molecules. At higher reduced electric fields (above 200 Td), variations in water vapor content have minimal impact on primary electron transport parameters, likely because dissociative and ionizing collisions between electrons and CO₂/H₂O molecules become dominant. Increasing the quenching pressure enhances photoionization, but plasma discharge remains primarily sustained by direct electron-impact ionization. Low initial water vapor content and elevated quenching pressure both accelerate streamer propagation, with the concentration ratio exerting a more significant effect. Finally, the primary reaction pathways for key products (CO, OH, and electrons) are analyzed. These findings contribute to a better understanding of how the reactant concentration ratio and quenching pressure regulate the discharge reaction mechanism in atmospheric pressure non-equilibrium CO₂-H₂O plasma.

(Sc₂O₃)_0.1(CeO₂)_0.01(ZrO₂)_0.89 possesses excellent ionic conductivity among various stabilized ZrO₂ electrolyte materials for solid oxide fuel cells. However, its practical application is limited by susceptibility to phase transition and the high cost of Sc₂O₃ raw material. Herein, we address these challenges by partially replacing Sc₂O₃ in (Sc₂O₃)_0.1(CeO₂)_0.01(ZrO₂)_0.89 with low-cost Yb₂O₃. Quaternary (Yb₂O₃)_x(Sc₂O₃)_0.10‒x(CeO₂)_0.01(ZrO₂)_0.89 (x = 0.04−0.10) electrolyte discs are fabricated by coupling tape casting and in situ solid-state reaction. All Yb₂O₃ doped electrolytes exhibit a single cubic phase structure. With increasing in Yb₂O₃ amount, the grain boundary resistance decreases, leading to improved conductivity at low temperatures. (Yb₂O₃)_0.06(Sc₂O₃)_0.04(CeO₂)_0.01(ZrO₂)_0.89 exhibits the ionic conductivity of 0.088 and 0.0020 S∙cm^‒1 at 800 and 500 °C, respectively. In addition, both the thermal expansion coefficient and three-point bending strength of the electrolytes increase with higher Yb₂O₃ amount, satisfying the criteria for advanced electrolyte materials in solid oxide fuel cells. A single cell configuration comprising a Ni-Gd_0.2Ce_0.8O_1.9 anode|200 μm thick (Yb₂O₃)_0.06(Sc₂O₃)_0.04(CeO₂)_0.01(ZrO₂)_0.89|La_0.6Sr_0.4Co_0.2Fe_0.8O₃ cathode achieves a peak power density of 0.65 W∙cm^‒2 at 800 °C and operates stably for 100 h without noticeable degradation. The present findings provide a new approach for the development of cost-effective and highly conductive ZrO₂-based electrolyte for efficient and durable solid oxide fuel cells.

Traditional quantitative structure-property relationship (QSPR) methods rely on molecular descriptors to quantify molecular structures and establish correlations with physical properties. In this study, we propose an approach that incorporates complete molecular structures to refine traditional QSPR methods and improve predictive accuracy. The supercritical properties used for modeling are collected from the literature. Molecular structures are optimized using density functional theory, from which molecular descriptors are derived. Both the structures and descriptors serve as inputs to the models developed in this work. Three models are constructed: a traditional artificial neural network model, a ResNet model, and a convolutional neural network (CNN)-enhanced model. Comparison with the JOBACK method shows that the CNN-enhanced model achieves higher predictive accuracy, whereas the ResNet model, which relies solely on molecular structures, suffers from pronounced overfitting.

Real-time monitoring of gas-liquid sulfonation in microchannel reactors is challenging due to complex internal spatiotemporal dynamics and limited data availability, despite the reactors’ excellent heat and mass transfer properties. Therefore, this study proposes a deep learning-based measurement method that directly extracts key spatiotemporal information from reaction image sequences within microchannels, enabling accurate prediction of the yield level of sodium α-olefin sulfonate products. The core of the framework is a convolutional long short-term memory network and combines a TimeDistributed module to efficiently capture and analyze dynamic visual features. To address the issue of data sparsity in experimental studies, we developed a novel frame sampling temporal image augmentation strategy that significantly improves the temporal learning efficiency of the model by mining microscopic dynamic changes under macroscopic stable conditions. On the experimental data set, the augmented convolutional long short-term memory network model achieved an average accuracy of up to 97.44%, outperforming the model without augmentation by 19.66% and a traditional convolutional neural network by 9.94%. These results demonstrate that the proposed method is a robust and effective tool for monitoring microchannel gas-liquid sulfonation, paving the way for intelligent, data-driven control of complex micro-chemical processes.