Researchers have developed a dynamic interfacial polymerization strategy that fabricates TpPa‑1 covalent organic framework (COF) membranes on polyacrylonitrile substrates in just 3 minutes – drastically reducing the conventional synthesis time from several days. The resulting membrane exhibits a permeance of 865 L·m⁻²·h⁻¹·MPa⁻¹ and a separation factor of 212 for methyl blue/Na₂SO₄ nanofiltration, outperforming most reported COF membranes. The membrane also demonstrates excellent stability, maintaining dye rejection above 99 % over 40 hours and showing good acid resistance.

Most research studies analyzing the effects of species loss on the health and productivity of grasslands do not reflect the conditions of mature or natural grasslands. In this study, a research team analyzed how changing the order of species removal in a grassland community changed the pathways linking species loss to community stability, which is more reflective of species loss scenarios in nature.

An international team, including the University of Tokyo, has created a sensor inspired by the lateral line in fish – their “sixth sense” organ – which measures the pulse of lab-grown 3D heart tissue (cardiac organoids). The device, called a biomechanical well plate, looks like a small white box containing four liquid-filled wells. When an organoid is placed in a well, each beat causes the liquid to bulge into an air cavity below, changing the air pressure. This change bends a cantilever sensor, which sends live data wirelessly to an app. The device is precise, reusable and less labor-intensive than some other methods. Due to its scalability, hundreds of tests could potentially be monitored simultaneously, with advantages for drug screening and future personalized medicine. 

A bio-inspired robotic bird capable of mimicking the key movements of kestrels is helping researchers unravel the mysteries behind the species’ exceptional hovering capabilities.

A research team from the Institute of Advanced Materials (INAM) at the Universitat Jaume I, comprising members of the Electrocatysis and Energy Group (ECATEN) and the Hybrid Catalytic Materials Group (HyCaM), has developed a new methodology for producing advanced catalytic materials for oxidation processes applicable to the chemical and petrochemical industries, as well as the energy sector.

The innovation lies in the modification of conductive supports to significantly improve their catalytic activity and efficiency. The invention has been validated at laboratory scale and in a 100 cm² demonstration prototype. Tests have shown optimal performance in the oxidation of various organic compounds and biomass-derived products, “confirming the material’s stability and its ability to operate efficiently under real-use conditions”, according to the research team.

Researchers have reviewed recent progress in ammonia synthesis via thermal catalysis, electrocatalysis, photocatalysis, and plasma catalysis under mild conditions, highlighting novel pathways that use nitrogen oxides (NOₓ) as the nitrogen source. The two‑step method (N₂ oxidation followed by NOₓ reduction) shows promise for green ammonia production using air and water with renewable electricity. However, challenges remain in conversion rates, energy efficiency, and detection accuracy. The review also emphasizes that rigorous ¹⁵N₂ isotope verification is essential to avoid false positives, calling for unified experimental standards to advance the field toward industrial application.

Researchers have systematically compared the combustion characteristics and oxidation kinetics of tar-rich coal and tar-inclusive coal using thermogravimetry and differential scanning calorimetry at multiple heating rates. The results show that tar-rich coal exhibits lower characteristic temperatures (by 5 to 17 °C), enabling it to reach ignition more easily. During the pyrolysis stage, the average activation energy of tar-rich coal exceeds that of tar-inclusive coal by more than 22.2 kJ·mol⁻¹, while in the combustion stage it is 5 kJ·mol⁻¹ lower. These differences arise from the hydrogen-rich molecular structure of tar-rich coal, which has important implications for both its fire risk and its in-situ thermal exploitation.

Researchers have developed AdapGNN, a novel model-agnostic framework that addresses the oversmoothing problem in graph neural networks (GNNs) for molecular property prediction while significantly improving model explainability. By integrating original atom features during message passing and introducing a weight projection module, AdapGNN preserves node uniqueness and highlights chemically important atoms. To enable quantitative evaluation of explainability, the team also created MolExplain, a new benchmark dataset with ground-truth atom importance scores. Experimental results show that AdapGNN outperforms baseline models (AFP and GCN) on 10 out of 12 tasks, with ROC-AUC for atom importance reaching up to 99.5 %.

Electric vehicles (EVs) are crucial for sustainable transport, but their environmental benefit depends heavily on when they charge. Charging during periods when the grid relies on fossil fuels can significantly increase their carbon footprint. New research from Multimedia University reveals that "carbon-aware" charging strategies can unlock substantial emission reductions simply by optimizing charging times.

The study, led by Md Sabbir Hossen and Gobbi Ramasamy from Multimedia University, quantifies the real-world carbon emission reduction potential of smart EV charging. Researchers reconstructed hourly residential charging data and combined it with time-varying grid carbon intensity from the Malaysian Peninsula power system. They then evaluated a reinforcement learning-based carbon-aware smart charging strategy, comparing its performance against both uncontrolled charging and an intelligent but carbon-unaware charging approach, all while delivering identical total energy.

The Power of Carbon Awareness

The findings are compelling: while intelligent charging without carbon awareness failed to reduce total carbon emissions, the carbon-aware strategy achieved a statistically significant reduction of approximately 1.62%. This crucial difference comes from proactively shifting charging demand away from periods when grid electricity is most carbon-intensive. Specifically, evening peak hours were found to be approximately 18% more carbon-intensive than off-peak periods, making targeted load shifting highly effective.