Metal halide perovskites (e.g., CsPbBr₃) have emerged as highly promising scintillator materials owing to their superior optoelectronic characteristics. However, their practical deployment is constrained by intrinsic structural instability and nonradiative recombination-induced energy loss. Herein, they show that Ni²⁺ doping constitutes a potent regulatory strategy for synergistically improving the scintillation performance and environmental robustness of CsPbBr₃ perovskite. Various characterizations combined with threotecial calculation indicate that the incorporation of Ni²⁺ dopants triggers lattice contraction, thereby enhancing the resistance to environmental perturbations. Ni-doping also passivates intrinsic defects and traps, leading to marked suppression of nonradiative recombination pathways. This synergy in as-prepared Ni-doped CsPbBr₃ leads to an 11-fold enhancement in photoluminescence intensity and a substantial increase in photoluminescence quantum yield from 56.0% to 93.7%. Notably, it delivers an exceptional light yield of 38428.5 photons MeV⁻¹, a low detection limit of 64.9 nGyₐᵢᵣ s⁻¹, and superior radiation tolerance. Furthermore, a flexible scintillation screen containing Ni-doped CsPbBr₃ enables high-resolution X-ray imaging with a spatial resolution of 16.6 lp mm⁻¹, notably surpassing that of the majority of reported perovskite-based scintillators. This study provides profound insights into the pivotal role of metal doping into halide perovskites for enhancing environmental stability in radiation detection technologies.

We demonstrated a facile FJH strategy to achieve the controlled synthesis of atomically dispersed Cu species via ultrafast heating-cooling process. By systematically regulating the pulse parameters, we successfully anchored isolated Cu species with a well-defined local coordination environment. The optimized photocatalyst exhibits an approximately 10-fold enhancement in photocatalytic CO evolution under simulated solar irradiation, offering a promising new pathway for clean fuel production and carbon capture

A team from the College of Chemistry and Chemical Engineering of Shaanxi University of Science and Technology and Functional Inorganic Materials Energy Conversion Laboratory started from “yarn” as a continuously processable and weave-integrable material unit. They designed a dynamic electrostatic cladding-spinning strategy that integrates a high-strength mechanical skeleton, a photothermal conversion shell, and thermoelectric waste-heat recovery into a single yarn platform.

From drug synthesis planning and production scheduling to facility location problem, combinatorial optimization (CO) is a backbone of efficient operation in science, engineering, and society. Yet building effective solvers still often requires expert modeling and extensive manual tuning. A new review surveys rapid progress in using large language models (LLMs) across three major paradigms — LLM-assisted modeling and solver collaboration, LLM-assisted heuristics, and LLM-based end-to-end solvers — and analyze their workflows, strengths, and limitations. The paper also summarizes application areas such as planning and scheduling, wireless networks and engineering design, and outlines research directions for making LLM-assisted CO more reliable.

A research paper by scientists at Fudan University presented BioPINN-LM, a hybrid framework that combines physics-informed neural networks and multimodal large language models to rapidly predict aortic aneurysm wall stress and generate interpretable biomechanical risk assessments.