Data-centric materials informatics has become a transformative paradigm for accelerating the discovery and design of superalloys, particularly by enabling efficient prediction of properties that are experimentally inaccessible or computationally intractable due to constraints in cost, time, or complexity. By harnessing the ability of machine learning (ML) to model complex, nonlinear, and high-dimensional relationships, this approach provides a compelling alternative to traditional trial-and-error and simulation-based strategies. This review presents a comprehensive and critical assessment of recent advances in ML for superalloys. We first delineate the essential workflow for ML-enabled superalloy design, encompassing foundational data resources, quantitative assessments of data quality, feature descriptors and feature-selection strategies, representative algorithms tailored to small and heterogeneous datasets, rigorous model-evaluation protocols, and model interpretation through explainable ML and symbolic regression. We then summarize state-of-the-art ML applications targeting specific high-temperature performance metrics, particularly γ' phase stability, creep behavior, fatigue life, and oxidation resistance, and highlight how approaches such as multi-fidelity learning, data augmentation, transfer learning, and optimization algorithms facilitate efficient exploration of vast composition-processing design spaces. Finally, we discuss persisting challenges and emerging opportunities, including data scarcity and reliability, model confidence and uncertainty quantification, cross-system generalizability across Co-, Ni-, and multi-principal superalloys, high-dimensional multi-objective optimization, and the integration of physics-informed models and large language models into materials-informatics workflows. By synthesizing these developments, this review outlines a strategic roadmap for harnessing ML to accelerate the discovery, performance optimization, and intelligent design of next-generation superalloys.

More than 2 billion tons of coal-based solid wastes (CBSW) are produced annually in China present not only significant environmental hazards, including air pollution from dust, soil degradation, and water contamination from heavy metals, but also direct safety risks such as spontaneous combustion and landslides. Currently, soil degradation is becoming an increasingly serious concern. Artificial soil is a crucial green construction material. However, the current resource utilization of CBSW in artificial soil is confronted with difficulties such as low efficiency, high ecological risks, and obstacles to industrialization. Therefore, there is an urgent requirement to develop a stable and eco-friendly approach for the construction of artificial soil. This paper reviews the physicochemical properties of CBSW and its adaptability to soil improvement. Considering the application directions of CBSW in ecological soil (such as remediating contaminated soil, improving poor soil quality, and promoting plant growth). It focuses on key methods for preparing artificial soil. These methods include pretreatment technology, optimizing the ratio of solid waste, additives, and soil, and evaluating ecological effects. This work provides insights into transforming coal waste into a valuable resource for ecological restoration.

Mineral processing wastewater poses severe environmental risks due to its complex composition (high suspended solids, residual reagents, heavy metals), making its treatment critical for sustainable mining. This review systematically summarizes mineral processing wastewater treatment technologies, including conventional methods and emerging approaches. Conventional physical-chemical methods are widely used but suffer from sludge production and limited resource recovery. Advanced oxidation processes (e.g., plasma oxidation, photo-Fenton) efficiently degrade refractory organics and novel adsorbents (MOFs, selective resins) enable targeted heavy metal recovery and deep purification. Artificial intelligence and digital twin further promote intelligent process control. Future directions focus on integrating multi-technologies into "classification treatment-quality-based reuse" systems to achieve comprehensive recovery of water, salts, and valuable metals, advancing mining towards a circular economy and near-zero discharge.

The present study investigates the performance and mechanism of nitrogen-doped carbon-based catalysts in selective catalytic reduction (SCR) reactions for removing nitrogen oxides (NO_x) through a combination of experiments and density functional theory (DFT) calculations. A series of catalysts with a gradient distribution of nitrogen content were prepared, and the types, contents, and structural characteristics of their nitrogen-containing functional groups were characterised. The experimental findings demonstrated that with an increase in nitrogen content, there was an initial rise and subsequent decrease in NO conversion among the catalysts. The AC-N-3 catalyst exhibited the highest NO conversion, with an observed value of 83.0 %. DFT calculations revealed that nitrogen doping enhanced the adsorption capacity of the catalysts for NO and O₂ through the introduction of functional groups. The active centre is located at the nitrogen functional group and its adjacent carbon atom. The centre of the molecule is responsible for driving the charge migration process, which in turn causes a stretching of the bond length of the reactants. This effect leads to the efficient pre-activation of the reactants, thereby significantly enhancing their catalytic activity. Through the analysis of the NH₃-SCR reaction pathway, the fundamental steps of the reaction were presented in a comprehensible manner.

In this paper, atomic simulation of the thermomechanical fatigue (TMF) behavior of Ni-based single crystal superalloys has been achieved, and the effect of Rhenium (Re) on the TMF properties are studied by molecular dynamics (MD) simulation. The reasons why 3%Re improving TMF properties of superalloys are explained from an atomic perspective. The results show that adding 3%Re to the superalloys can increase the cyclic stress amplitude and plastic deformation resistance, reduce the dislocation density and plastic strain energy density, and thereby improve the fatigue life of superalloys. The microstructure evolution reveals that the improvement of TMF properties in superalloys mainly depends on the pinning and dragging effects of Re on dislocation motion. Due to the pinning and dragging effects of Re, the stability of microstructure is significantly enhanced, leading to a reduction in plastic deformation and thus improving the TMF mechanical properties and fatigue life of superalloys. The research results will contribute to a deeper understanding of the TMF mechanisms and Re effects of superalloys.

tensile twins were introduced by pre-compressing the rolled AZ31 Mg alloy sheet along the transverse direction (TD) with a strain of 3 %, aiming to investigate the effect of pre-existing twins on its bending deformation behavior. For the AZ31 Mg alloy, the pre-existing tensile twins significantly improved the mechanical properties, the tension-compression yield asymmetry coefficient (0.57 vs. 0.35), and the bending property (bend angle: 97° vs. 65°). The pre-existing twins led to the deflection of the c-axis of the grains, thus modifying the strong (0001) basal texture, which improved the tension-compression yield asymmetry, making the strain distribution during the bending process in each region of the specimen more uniform. The basal slip caused by grain deflection on the rolling direction (RD)-normal direction (ND) plane increased the thickness-direction strain of the specimen during the bending deformation process. Moreover, the introduction of a large number of twin lamellae effectively subdivided and refined the grains, enhancing the plastic deformation ability of the specimen. In summary, these factors led to a significant improvement in the bending formability of the AZ31 Mg alloy.

Nitrate () is a pervasive pollutant in industrial water treatment due to its widespread presence in industrial processes. Quaternary ammonium (R₄N⁺) groups are key for removal, but their efficiency varies with water quality. This study developed a mixed-grafted ion exchange resin, CMIET-A, by grafting trimethylamine and triethylamine onto a poly (methyl methacrylate) backbone, with Ferromagnetic γ-Fe₂O₃ nanoparticles to enhance separability and recyclability. Experiments and DFT calculations showed that CMIET-A effectively removed across a broad pH range (4.0-10.0), with a maximum adsorption capacity of 82.79 mg/g. The adsorption behavior fit the Freundlich isotherm model, and the process followed the pseudo-second-order kinetic model. After 20 cycles, the resin maintained a removal rate over 70 %. Both experiments on the influence of external ions and Molecular dynamics simulations indicated higher binding energy and diffusion coefficients for CMIET-A with , enhancing performance even in the presence of Cl^-. Characterization revealed that ion exchange, pore filling, electrostatic attraction, hydrogen bonding, and metal bridging collectively drove adsorption. Overall, this novel resin offers an efficient solution for removal in industrial settings.

Fast-charging sodium-ion batteries are severely constrained by sluggish Na⁺ diffusion, structural instability, and rapid capacity fading in layered anodes, representing a major challenge for high-power energy storage applications. Here, a Co and Se co-doping strategy is implemented on MoS₂ (Co-MoS_1.8Se_0.2/C) to stabilize the metallic 1T-like phase, expand interlayer spacing, and introduce abundant defect sites, generating additional redox-active centers that facilitate rapid and reversible Na⁺ insertion and extraction. Cobalt doping serves as a catalytic regulator, promoting uniform SEI formation and enhancing interfacial stability, whereas selenium doping reduces Na⁺ diffusion barriers and alleviates strain induced by volumetric changes. A conductive carbon framework supports the nanosheet structure, prevents restacking, and buffers mechanical stress, ensuring structural integrity during extreme-rate cycling. The Co-MoS_1.8Se_0.2/C electrode achieves a reversible capacity of 250 mAh g^-1 at 20 A g^-1, corresponding to full charge/discharge in approximately 15 s, and maintains long-term cycling stability over 1400 cycles at 5 A g^-1. Structural analyses reveal partial electron transfer from Co and Se to Mo upon intercalation, triggering reorganization of Mo 4d orbitals and inducing a spontaneous 2H-to-1T phase transition, which enhances electrical conductivity. Reversible layered-to-metallic transformation occurs alongside the formation of a stable SEI layer, further promoting electrochemical kinetics and interfacial stability. The synergistic integration of dual-element doping and carbon framework design significantly improves structural robustness and sodium storage performance.

Arsenic pollution in water poses a significant environmental challenge due to its high toxicity and non-degradability. In this study, FeMn-layered double hydroxide (FeMn-LDH) was synthesized using hydrothermal and coprecipitation methods with different precursors for electrochemical arsenic remediation. The crystallinity of FeMn-LDH was enhanced with nitrate precursor compared to chloride precursor. The corresponding calcined layered double oxide obtained through the coprecipitation method (FMO-NO₃-Co) exhibits a significantly increased specific surface area and an optimal average pore size, facilitating efficient ion transport, and enhanced oxidation state of Mn, increasing arsenic removal efficiency. Electrochemical tests indicate that FMH-NO₃-Co exhibits relatively high specific capacitance and excellent electrochemical performance. Notably, the FMO-NO₃-Co achieves an electrosorption capacity of 55.5 mg g^-1 with 56.6 % of As(Ⅲ) being electrochemically oxidized, demonstrating superior electrocatalytic activity for the oxidation of As(Ⅲ) and high-performance electrosorption of As(V). The arsenic removal mechanism was comprehensively analyzed, revealing that Mn²⁺/Mn³⁺ redox cycling played a key role in As(Ⅲ) oxidation, while Fe-based coordination sites contributed to As(V) adsorption. Furthermore, the enhanced porosity and conductivity of the calcined LDH materials significantly improved charge transfer efficiency, thereby accelerating the arsenic removal process. Overall, this study provides valuable insights into the potential application of FeMn-LDH in electrochemical arsenic remediation.

Proton exchange membrane water electrolysis (PEMWE) has long been regarded as a promising technology for hydrogen production due to its high electrolytic efficiency, reliability, and rapid response to renewable energy sources. Currently, noble metals and their oxides—such as Pt, IrO₂, and RuO₂—remain the most widely used and high active electrocatalysts in acidic media to accelerate the water electrolysis processes. However, their large-scale pratical application is severely hindered by the factors such as scarcity and the trade-off between activity and stability. Recently, the integration of artificial intelligence (AI) with high-throughput synthesis technology has demonstrated an increasingly vital role in material screening, enabling the design of highly efficient and cost-effective catalysts. This paper first reviews the fundamental catalytic mechanisms of the hydrogen evolution reaction (HER) and the oxygen evolution reaction (OER) in acidic media. Then, it summarizes the design strategies and prevailing challenges for noble metal catalysts in acidic water electrolysis. Finally, it presents several data-driven, synergistic approaches enabled by AI in noble metal catalyst research and development (R&D), along with the latest progress, current challenges, and future prospects.