Researchers have developed a high‑accuracy Al‑O‑H machine learning interatomic potential combined with global optimization to explore the structural landscape of γ‑Al₂O₃ and its influence on propane dehydrogenation over single‑atom Pt catalysts. Two energetically favorable structures, γ‑NAV and γ‑AV, were identified with energies lower than the conventional model by up to 34.44 meV·(f.u.)⁻¹. These optimized structures expose abundant penta‑coordinated Al³⁺ sites on the (100) surface, serving as preferred anchoring sites for Pt atoms. Catalytic calculations demonstrate that γ‑NAV and γ‑AV significantly reduce propane dehydrogenation activation barriers through enhanced metal‑support interactions.

Scientists from Trinity College Dublin have developed a powerful way to unravel the changing weather patterns on distant worlds. Using this approach, they discovered that the weather on a well-studied brown dwarf, “SIMP 0136”, previously linked with Northern Lights-like phenomena, is largely shaped by just two dominant processes: changes in temperature and the vertical structure of its clouds.

High‑throughput workflows have revolutionized experimental heterogeneous catalysis by enabling rapid, parallel synthesis and screening of catalyst libraries. This review traces the evolution of high‑throughput technology from manual experiments in the early 1900s to today‘s autonomous, artificial intelligence‑driven platforms that integrate robotics, machine learning, and self‑driving laboratories. The state‑of‑the‑art mode encompasses five key steps: experiment design, catalyst synthesis, catalyst testing and screening, catalyst characterizations, and data handling and analysis. Despite its transformative potential, high‑throughput workflows face critical pitfalls—including gaps between microscale reactor results and industrial performance, pressure limitations, and chemical information disparity in funnel‑shaped screening—that must be addressed to fully realize its promise for rational catalyst discovery.

Structures critically control hydrothermal gold mineralization and orebodies distribution. Using borehole data, deep geophysical surveys, and 3D geological models from typical Jiaodong gold deposits, this study reveals the ore-controlling role of fault-dip variations. Gold orebodies preferentially occur in gently dipping fault segments and at steep-to-gentle transitions, forming a step-like mineralization pattern. Comparison with gold deposits across China identifies three step-control styles: (1) gently dipping segments and steep-to-gentle bends of a single fault; (2) subsidiary gently dipping faults linked to a steep master fault; (3) gently dipping interlayer structures bounded by steep faults. Mechanical analysis shows that gently dipping faults concentrate stress, provide good sealing, and have low wall-rock shear strength, which together promote fluid focusing and gold precipitation. These findings refine the theory of hydrothermal gold mineralization and offer valuable guidance for deep gold exploration.

 Researchers have developed an integrated analytical framework combining multi‑physics mechanism simulation and active learning‑based intelligent optimization to address the challenge of synergistic optimization of gas diffusion layer physical properties in anion exchange membrane water electrolyzers. The active learning‑based weighted ensemble surrogate model achieved a coefficient of determination of 0.9994 on the test set with a root mean square error of 5.0240 and a mean absolute error of 4.0694. The optimal parameter combination identified by the framework was a permeability of 1.18 × 10⁻¹⁰ m², an electrical conductivity of 2208.86 S·m⁻¹, and a porosity of 0.605, which effectively suppressed local heat accumulation and eliminated mass transport bottlenecks.

Researchers have developed an integrated analytical framework combining multi‑physics mechanism simulation and active learning‑based intelligent optimization to address the challenge of synergistic optimization of gas diffusion layer physical properties in anion exchange membrane water electrolyzers. The active learning‑based weighted ensemble surrogate model achieved a coefficient of determination of 0.9994 on the test set with a root mean square error of 5.0240 and a mean absolute error of 4.0694. The optimal parameter combination identified by the framework was a permeability of 1.18 × 10⁻¹⁰ m², an electrical conductivity of 2208.86 S·m⁻¹, and a porosity of 0.605, which effectively suppressed local heat accumulation and eliminated mass transport bottlenecks.