Precise syngas ratio control is crucial for efficiently producing high-quality clean fuels such as naphtha, kerosene, and diesel. In light of the rising demand for energy-efficient hydrogen separation, this study investigates the H₂/CO separation performance of novel benzimidazole-linked polymer (BILP-101x) and poly(p-phenylene benzobisimidazole) (PBDI) membranes fabricated via a simple interfacial polymerization process. The effects of temperature, pressure, and H₂ molar fraction on the membranes' separation performance were comprehensively evaluated. Both membranes displayed high H2 permeance (BILP-101x∼136 GPU; PBDI∼76 GPU) and selectivity (BILP-101x∼78; PBDI∼50) for H2/CO separation at 150 °C. The superior H₂ permeance of BILP-101x was attributed to its higher fractional free volume and diffusion coefficients compared to PBDI, as confirmed by molecular simulations. Notably, both membranes demonstrated remarkable stability during long-term testing under simulated syngas conditions (50/25/25H₂/CO₂/CO, 100 °C for 120 h), outperforming most polymeric membranes reported in literature for H2/CO separation. The superior H2/CO separation performance coupled with the excellent stability (over 120 h) endows BILP-101x and PBDI membranes with an attractive application prospect for industrial syngas ratio adjustment.@en

Separation of hydrogen from carbon dioxide for sustainable H₂ production and CO₂ capture still faces great challenges due to the smaller size of H₂ and higher condensability of CO₂. Herein, a high-performance benzimidazole-linked polymer (BILP) membrane for hydrogen separation was directly prepared on α-Al₂O₃ substrate through a facile interfacial polymerization approach at room temperature. The separation performance of the BILP membrane were regulated by controlling the reaction time and the microstructure was systematically characterized. Molecular simulations were performed to deep understand the separation mechanism in the BILP membrane. The best performance membrane displays an extraordinary mixed gas selectivity of 40 for H₂/CO₂ together with outstanding H₂ permeance of 250 gas permeation units (GPU) at 473 K, far exceeding the Robeson's upper bound. Besides, the membrane can withstand high temperature and pressure, and also shows good H₂/N₂ and H₂/CH₄ selectivity. The excellent separation performance, coupled with high temperature and pressure resistance and easy preparation, render BILP membranes great potential for economic H₂ purification, H₂ recovery, and natural gas treatment.

@en

Deep learning models are widely used in traffic forecasting and have achieved state-of-the-art prediction accuracy. However, the black-box nature of those models makes the results difficult to interpret by users. This study aims to leverage an Explainable AI approach, counterfactual explanations, to enhance the explainability and usability of deep learning-based traffic forecasting models. Specifically, the goal is to elucidate relationships between various input contextual features and their corresponding predictions. We present a comprehensive framework that generates counterfactual explanations for traffic forecasting and provides usable insights through the proposed scenario-driven counterfactual explanations. The study first implements a deep learning model to predict traffic speed based on historical traffic data and contextual variables. Counterfactual explanations are then used to illuminate how alterations in these input variables affect predicted outcomes, thereby enhancing the transparency of the deep learning model. We investigated the impact of contextual features on traffic speed prediction under varying spatial and temporal conditions. The scenario-driven counterfactual explanations integrate two types of user-defined constraints, directional and weighting constraints, to tailor the search for counterfactual explanations to specific use cases. These tailored explanations benefit machine learning practitioners who aim to understand the model's learning mechanisms and domain experts who seek insights for real-world applications. The results showcase the effectiveness of counterfactual explanations in revealing traffic patterns learned by deep learning models, showing its potential for interpreting black-box deep learning models used for spatiotemporal predictions in general. @en

Ceramic membranes have been increasingly employed in water treatment owing to their merits such as high-stability, anti-oxidation, long lifespan and environmental friendliness. The application of ceramic membranes mainly focuses on microfiltration and ultrafiltration processes, and some precise separation can be achieved by introducing novel porous materials with superior selectivity. Recently, metal–organic frameworks (MOFs) have developed a wide spectrum of applications in the fields of the environment, energy, water treatment and gas separation due to the diversity and tunable advantages of metal clusters and organic ligands. Although the issue of water stability in MOF materials inhibits the development of MOF membranes in water treatment, researchers still overcome many obstacles to advance the application of MOF membranes in water treatment processes. To the best of our knowledge, there is still a lack of a reviews on the development process and prospects of ceramic-based MOF membranes for water treatment. Therefore, in this review, we mainly summarize the fabrication method for ceramic-based MOF membranes and their application in water treatment, such as water/salt separation, pollutant separation, heavy metal separation, etc. Following this, based on the high structural, thermal and chemical stability of ceramic substrates, and the high controllability of MOF materials, the superiority and insufficient use of ceramic-based MOF membranes in the field of water treatment are critically discussed.@en

High permeable and selective membranes are expected to be a promising alternative for low cost and energy efficient CO₂ separation. However, the preparation and development of highly permeable and selective membranes are limited by the trade-off effect. Here, we prepared a stabilized high permeable and selective mixed matrix membrane based on the structural regulation of the NH₂-functionalized Co/ZIF-8 heterometallic nanoparticles (NH₂-Co/ZIF-8). The NH₂-Co/ZIF-8 can be homogeneously dispersed in the polymer via chelation of the ester group and the metal node in the methacrylate-terminated PEO and the acrylate-terminated PEO. The amphiphilic interaction between the polymer and NH₂-Co/ZIF-8 significantly increased the CO₂ solubility resulting in the boost of both CO₂ permeability and CO₂/N₂ selectivity. The best performance membrane exhibits a CO₂ permeability of 2916 Barrer and a CO₂/N₂ selectivity of 47, surpassing the Robeson's upper bond (2008) by a wide margin.

@en