Intense sand mining is putting the largest freshwater lake in Southeast Asia at risk of collapse with catastrophic consequences, a new study has found.

X-ray absorption spectroscopy (XAS) provides valuable information about a material’s properties and electronic states. However, it requires extensive expertise and manual effort for conventional analysis. Now, researchers from Japan have developed a novel artificial intelligence-based approach for analyzing XAS data that can enable rapid, autonomous, and object material identification. This novel approach outperforms the previous studies in terms of higher accuracy, accelerating the development of new materials.

Crystalline-amorphous composites comprise crystalline grains separated by amorphous boundaries. The combined role of grain size (D) and amorphous boundary thickness (l) on material properties has not been explored. Now, writing in National Science Review, a team from the Hong Kong University of Science and Technology reports simulation results of mechanical properties across the (D, l) parameter space. They identify optimal (D, l) values that provide maximum strength while also enhancing ductility, successfully circumventing the classic strength-ductility tradeoff.

Scientists have designed a gradient sodium-tin alloy/sodium bilayer anode that solves the two problems of dendrite growth and sodium loss in sodium batteries. This innovative structure features an upper "ion-buffering" layer that guides sodium ions for dendrite-free deposition and a bottom reservoir that dynamically compensates for lost sodium. The resulting batteries achieve an unprecedented energy density and ultralong cyclability in lab tests, paving the way for more powerful and durable energy storage.

Nanyang Technological University, Singapore (NTU Singapore) and Zero Gravity (0G), a decentralised AI infrastructure company, have announced a S$5 million partnership to establish a joint research hub advancing blockchain-based artificial intelligence (AI) technologies that will be more accessible and accountable. This marks 0G’s first university collaboration globally and will fund multiple projects exploring decentralised AI training, blockchain-integrated model alignment, and proof-of-useful-work consensus mechanisms.