In the era of materials genome engineering, data-driven machine learning has become a powerful tool for accelerating the research and development of metallic materials. However, the predictive accuracy and generalization ability of traditional machine learning models are often limited by the scarcity and heterogeneity of available data, especially in small-sample scenarios. To address these challenges, transfer learning has emerged as an effective strategy to leverage knowledge from related domains, thereby enhancing model performance with limited target data. This review systematically summarizes the fundamental concepts, methodologies, and representative applications of transfer learning in the prediction of metallic materials’ properties. Transfer learning can be categorized into feature-based, instance-based, parameter-based, and knowledge-based methods. This work discusses their respective mechanisms, advantages, and limitations. Case studies demonstrate that transfer learning can significantly improve prediction accuracy, data efficiency, and model interpretability in tasks such as mechanical property prediction and alloy design. Furthermore, this work highlights emerging trends including hybrid, multi-task, meta, and adaptive transfer learning, which further expand the applicability of these techniques. Finally, this work outlines future research directions, emphasizing the need for data standardization, algorithmic innovation, multimodal data fusion, and the integration of physical principles to achieve robust, interpretable, and generalizable models. The perspectives presented aim to advance the intelligent design and discovery of metallic materials, promoting efficient knowledge transfer and collaborative innovation in materials science.

The rapid development of electronic devices and communication technologies has resulted in increasingly severe electromagnetic-wave (EW) pollution. Efficient EW absorption (EWA) materials are essential to mitigate their impact and ensure human safety in modern society. Fe-based EWA materials have garnered significant attention owing to their cost-effectiveness, high saturation magnetization, and superior magnetic loss capabilities. This review begins with an introduction to Fe-based EWA materials, followed by a brief description of their EWA mechanisms. Various pristine Fe-based absorbers, such as carbonyl iron powder, ferrite-based materials, Fe-based alloys, Fe-based high-entropy alloys (HEAs), and Fe-based layered ternary transition-metal borides, have been systematically reviewed. Key strategies to enhance the performance of Fe-based composite absorbers, including doping, in-situ oxidation, porous structuring, and composite construction, are critically discussed. Finally, the review presents a summary and future perspectives in this field, highlighting the synergy between Fe-based and high-entropy materials in advancing next-generation EWA for applications in stealth technology, wearable electronics, and harsh environments.

Cement paste backfill (CPB) technology is a key method for mine waste treatment, and pipeline transport is critical for safe and efficient waste transfer. Variations in raw material properties can cause slurry segregation, increase pipeline wear and resistance, raise the risk of blockages or bursts, and disrupt operations. To study CPB slurry segregation during transport, CPB was prepared using cement as the cementitious material and unclassified tailings as inert materials. A small annular-tube device using an electrical resistance tomography system was developed to analyze its flow characteristics, and quantitative segregation assessment methods were developed. The results indicated that CPB conductivity increases with transport time but decreases with higher solid mass content, with the latter having a greater impact. At a low solid content, solid particles migrated toward the bottom of the pipe as the flow time increased, and the migratory behavior of the particles diminished as the solid content increased. At a flow rate of 1.25 m/s, the heterogeneity index for CPB with 58wt% solid content increased by 1.24 in 20 min, whereas that for CPB with 62wt% solid content increased by 2.17. Higher solid mass content amplifies the effect of conveying time on segregation, emphasizing the need to balance these factors for minimizing segregation. These insights can guide the optimization of mine pipeline transport systems.

Tantalum (Ta) and niobium (Nb) are key strategic metals used in the aerospace, steel, and chemical industries. Columbite–tantalite is the primary Ta- and Nb-containing mineral. Flotation is an effective and practical approach for preconcentrating columbite–tantalite. However, the inevitable introduction of Ca, Mg, and other ions from process water and mineral dissolution during beneficiation can significantly affect the flotation performance of columbite–tantalite. This study systematically investigated the effects of Ca²⁺ and Mg²⁺ on columbite–tantalite flotation in a sodium oleate (NaOL) system. Flotation experiments revealed that, at pH = 10, the addition of Ca²⁺ and Mg²⁺ markedly suppressed the flotation of columbite–tantalite, reducing the recovery by 94.86% and 92.55%, respectively. Characterization revealed that NaOL forms a hexagonal ring structure with Mn sites on the columbite–tantalite (100) crystal surface. However, Ca²⁺ and Mg²⁺ ions interfere with the chemical adsorption of NaOL by reacting with it to form oleate precipitates, which subsequently cover the mineral surface. Therefore, excess NaOL did not facilitate the effective flotation of columbite–tantalite. Furthermore, NaOL, as compared with the columbite–tantalite surface, tended to interact with Ca²⁺ and Mg²⁺ to deactivate the collector. This paper elucidates the inhibitory effects of Ca²⁺ and Mg²⁺ on the flotation of columbite–tantalite. Consequently, the selective removal of metal ions, such as Ca²⁺ and Mg²⁺, from the slurry is essential to improve both the flotation efficiency and recovery of columbite–tantalite, particularly when processing ores with high water hardness or containing easily leachable metal ions.

This study addresses the challenge of predicting zinc (Zn) recovery from carbonate ores via sodium hydroxide (NaOH) leaching. This complex process influenced by variable ore composition, surface passivation effects, and nonlinear reaction dynamics, which complicate reagent optimization and process control in hydrometallurgical operations. To tackle this, a dataset containing 422 experimental observations was compiled from previous studies, incorporating ore composition and process parameters, such as NaOH concentration, leaching time, temperature, and solid-to-liquid ratio. Four regression models (decision tree, neural network, generalized additive model, and random forest) were trained and evaluated using performance metrics, such as coefficient of determination (R²), root mean squared error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and symmetrical mean absolute percentage error (SMAPE). Among these, the random forest model achieved the best predictive accuracy, with R² value of 0.8541 on the test set and the lowest error rates, demonstrating its effectiveness in capturing the complex relationships between input variables and Zn recovery. Explainable artificial intelligence, particularly SHapley additive exPlanations (SHAP) analysis, revealed that NaOH concentration, leaching time, and solid-to-liquid ratio had the most positive influence on Zn recovery, whereas elements such as Ca, Fe, and Pb had inhibitory effects. These findings align with known geochemical behavior and provide valuable insights for reagent optimization and process efficiency in leaching processes. This study demonstrates the practical potential of machine learning in mineral processing, offering a scalable framework for optimizing Zn recovery from non-sulfide ores and a data-driven approach to enhance decision-making in hydrometallurgical applications.

SiO₂–CaO–Al₂O₃ ternary inclusions are among the most common complex oxide inclusions in steel. Nevertheless, the chemical and physical properties of these composite inclusions, particularly with detailed composition changes, have not been sufficiently investigated. In this study, first-principles density functional theory calculations were used to determine the electronic, mechanical, and thermodynamic properties of two stable phases in the SiO₂–CaO–Al₂O₃ ternary inclusion system: anorthite (CaAl₂Si₂O₈) and gehlenite (Ca₂Al₂SiO₇). Based on the electronic density of states analysis and band structure calculations, oxygen atoms play important roles in the electron reactivity of both phases. Young’s modulus and Poisson’s ratios were calculated and compared with those of the SiO₂–CaO inclusions. The Young’s moduli of CaAl₂Si₂O₈ (101.32 GPa) and Ca₂Al₂SiO₇ (131.43 GPa) were close to the maximum and minimum Young’s moduli of the binary oxide inclusions, respectively. With increasing temperature, the Young’s moduli of CaAl₂Si₂O₈ and Ca₂Al₂SiO₇ showed slight increasing and decreasing trends, respectively, whereas the Poisson’s ratio decreased. Furthermore, the thermodynamic properties, particularly temperature-related thermal expansion coefficients, were also deeply investigated. The thermal expansion coefficients of both CaAl₂Si₂O₈ and Ca₂Al₂SiO₇ increased rapidly with increasing temperature in the low-temperature regime above 300 K. As the temperature increased, the increasing trend slowed. When the temperature reached 2000 K, the thermal expansion coefficients of CaAl₂Si₂O₈ and Ca₂Al₂SiO₇ respectively were 12 × 10⁻⁶ and 8.5 × 10⁻⁶ K⁻¹. These findings enhance the understanding of the physical nature of ternary inclusions in steels and provide a scientific foundation for analyzing their effects on steel performance using a more comprehensive inclusion database, thereby contributing to inclusion engineering in the development of materials with superior mechanical integrity.

The effect of manganese sulfide (MnS) inclusions and gadolinium–sulfide (Gd–S) inclusions on the deformation behavior of steel matrix at different stages was studied by in-situ tensile experiments using a scanning electron microscopy (SEM) at room temperature. Two in-situ tensile experiments of tensile force along the elongation direction of inclusions and perpendicular to the elongation direction were conducted. The hole-induced nucleation mechanism of different tensile directions and inclusion types during the tensile deformation process was revealed. When the tensile direction of the steel without Gd was parallel to the forging elongation direction, the tensile strength was 454 MPa. Meanwhile, long strip MnS inclusions were broken and shed, forming long strip holes perpendicular to the fracture direction. When the tensile direction was perpendicular to the forging elongation direction, the gap between long strip MnS inclusions and the steel matrix was expanded into a long strip hole parallel to the fracture direction, and the tensile strength was 402 MPa. Anisotropy of the steel was induced by long strip MnS inclusions. In the steel with a total gadolinium (T.Gd) content of 730 ppm, the tensile strength was 468 MPa when the tensile direction was parallel to the forging elongation direction. The tensile strength of the steel was 446 MPa when the tensile direction was perpendicular to the forging elongation direction. The addition of Gd in the steel was beneficial to improve the tensile properties of the steel and reduce the anisotropy of the steel.

Liquid core reduction (LCR) technology, originally developed for continuous thin-slab casting, allows space for a submerged entry nozzle in a mold while improving production efficiency. Recent experimental attempts explore the implementation of LCR in regular slab casting processes. However, regular slabs (2–3 times thicker than thin slabs) face critical challenges in terms of excessive deformation and stress concentration under external forces, which induce intermediate cracks and thus hinder successful LCR adoption in regular slab production. This study evaluates the feasibility of LCR for producing regular slabs and identifies optimal reduction parameters to prevent crack initiation. A three-dimensional thermal-mechanical coupled model is proposed using the finite element method (FEM), integrated with the equivalent replacement liquid steel (ERLS) method and the normalized Cockcroft–Latham damage model, to achieve quantitative prediction of intermediate crack risk during the LCR process. The ERLS model simulates the extrusion flow and expulsion behavior of the liquid core, and its accuracy is validated against actual production measurements. To identify the critical damage value leading to intermediate crack initiation, this study conducts a consistency analysis between high-temperature tensile tests and FEM-based simulations using damage models. Based on this value, crack prediction is performed for Q355 slabs with cross-sectional dimensions of 170 mm × 1450 mm. Using the prediction results, an optimal reduction scheme is determined, wherein the second segment accounts for 50% of the total reduction, the third segment for 32.5%, and the fourth segment for 17.5%, with the theoretical value of maximum reduction being 34 mm. These results provide actionable guidelines for the potential implementation of LCR in regular slab-casting systems.

Driven by efforts toward carbon-neutral steelmaking, increased scrap usage elevates Sn content in steels. While the general effects of Sn on steel have been studied, its specific influence on resistance spot welding (RSW) remains unclear. This study investigates Sn’s impact on the mechanical properties of RSW joint of 460 MPa HSLA steel. Cross-tension tests reveal that both the RSW joint without Sn and the RSW joint·containing 0.09wt% Sn exhibit pull-out failure. The RSW joint containing 0.09wt% Sn showing higher peak load and energy absorption attributed to Sn’s solid–solution strengthening. Conversely, the RSW joint containing 0.52wt% Sn exhibited the partial interface failure mode, significantly reducing the peak load and energy absorption. The primary reason is the segregation of Sn in the interdendritic regions of the fusion zone, which weakens atomic cohesion and reduces fracture toughness. Such severe segregation arises from RSW’s high cooling rates, which shift the primary solidification phase from δ-ferrite to austenite. Fortunately, double-pulse RSW mitigates Sn segregation, restoring failure mode and mechanical performance. This study assesses the impact of Sn on RSW joint properties, and these findings highlight the broader significance of understanding scrap-related residual element effects in sustainable steel production.

Twinning-induced plasticity (TWIP) steel was processed using electrically assisted friction stir welding (EFSW). The microstructure, mechanical properties, and deformation behavior of the welded joints were systematically investigated. The results show that the average grain size was refined from 3.67 µm in the base material (BM) to 1.39 µm in the stir zone (SZ), while it increased to 4.19 µm in the heat-affected zone (HAZ). The fraction of twin boundaries (TBs) decreased from 20.7% in the BM to 6.9% in the SZ and increased to 24.5% in the HAZ. The ultimate tensile strength, yield strength, and elongation of the BM were 1021 MPa, 505 MPa, and 65.8%, respectively. In comparison, the EFSW joint exhibited values of 1055 MPa, 561 MPa, and 60.8%, corresponding to 103.3%, 111.1%, and 92.4% of those of the BM, respectively. During tensile testing, plastic deformation was primarily concentrated in the BM, although both the SZ and HAZ also exhibited notable plastic deformation. Fracture ultimately occurred in the BM.

High-entropy magnetocaloric alloys offer exceptional compositional flexibility and stability for magnetic refrigeration. However, enhancing their magnetic entropy change, working temperature range, and refrigeration capacity remains challenging. In this study, we demonstrate that microalloying GdTbDyHo with only 0.4at% nonmagnetic Y effectively addresses this limitation. Our analysis indicates that Y uniformly dissolves into the hexagonal matrix lattice, disrupting the 4f–4f exchange interactions and inducing a local short-range order. This weakens the antiferromagnetic coupling, accelerates the antiferromagnetic–ferromagnetic transition, and broadens its range. Consequently, the peak magnetic entropy change increases from 8.2 to 8.7 J·kg⁻¹·K⁻¹, the working temperature range expands from 77 to 89 K, and the refrigeration capacity improves by 23%, reaching 774 J·kg⁻¹ (5 T) relative to the Y-free alloy, while the Néel temperature remains constant (∼195 K). This study establishes nonmagnetic microalloying as a cost-effective and scalable strategy for designing high-performance magnetocaloric materials.

Enhancing the oxidation resistance of Co-based superalloys by adding a high content of Cr, while simultaneously ensuring the stability of the γ/γ′ phases, presents a significant challenge. This study evaluated the alloying potential of Co–30Ni–10Al–5V–4Ta using the CALPHAD method, revealing promising characteristics. The developed Co–30Ni–10Al–5V–4Ta–12Cr alloy characterized by high Cr content and γ/γ′ two-phase structure, demonstrating high γ′ solvus temperature of 1139°C, low density of 8.48 g/cm³, minimal γ/γ′ lattice misfit of +0.28%, high compressive yield strength of 651 MPa at 800°C, and excellent oxidation resistance with a weight gain of 6.5 mg/cm³ after 200 h at 1000°C. Examination of the oxidation behavior at 1000°C revealed an oxide layer consisting of a porous outer CoO, NiO, and V₃O₄ (CNV) oxide and a denser inner mixed oxide layer comprising CoO, NiO, and V₃O₄ (CNV) oxide, Al₂O₃, Cr₂O₃, CoO, and NiO (CNAC) oxide, and TaO₂, CoO, and NiO (CNT) oxide.

The use of Al–V alloys as intermediate additives is pivotal for producing high-performance Ti alloys. Traditionally, the synthesis of these alloys relies on high-purity V₂O₅, with sodium metavanadate as an essential intermediate in V₂O₅ production. This study explores an alternative approach utilizing sodium metavanadate directly, offering an aluminothermic process to alleviate the environmental impact and reduce the time required for V₂O₅ preparation. Al–V alloys are synthesized using sodium metavanadate derived from a shale V-rich solution, and the impurity-migration behaviors are comprehensively analyzed, specifically focusing on Fe, Al, and Na. The results reveal that Al interacts with CaO to form a slag phase that is different from the alloy, whereas Na undergoes a sequence of reductions (NaVO₃ → Na₂V₂O₅ → NaVO₂ → Na) and volatilizes at 25–1200°C, thereby avoiding incorporation into the alloy. Fe, reduced by Al, enriches the alloy phase and induces a phase transition (Al–V → Al–Fe → Fe–V) in the presence of excess Fe. Sodium metavanadate (Fe ≤ 0.05wt%) derived from the shale V-rich solution enables the production of a uniform AlV65 alloy with 66.56wt% V, 33.14wt% Al, 0.08wt% Fe, 0.07wt% C, 0.02wt% N, and 0.12wt% O. These results establish a streamlined, efficient framework for the future preparation of Al–V alloys from shale V-rich solutions.

This study investigates the fabrication and characterization of Al alloy matrix composites reinforced with graphene oxide (GO) using accumulative roll bonding (ARB). The annealed Al 6061 sheets were processed through 5-pass ARB with GO reinforcement applied during the initial passes. Scanning electron microscopy revealed effective mitigation of GO agglomeration and improved interface bonding due to microscale material mixing. Raman spectroscopy confirmed the strong interaction between GO and the Al alloy matrix, as evidenced by the increased D band intensities and enhanced 2D band symmetry. Mechanical testing indicated an approximately 338.37% increase in yield strength (YS) and 86.42% improvement in hardness for the ARB-processed (ARBed) Al 6061/GO composite (0.2wt%) compared with annealed Al 6061 and an approximately 14.15% increase in YS and 17.23% improvement in hardness for the ARBed Al/GO composite (0.2wt%) compared with unreinforced ARBed Al 6061 specimens after five passes. X-ray diffraction analysis indicated an increased dislocation density, corroborating the observed enhancements in mechanical properties. Fracture surface analysis revealed reduced elongation with deep dimples, highlighting the tradeoff between strength and ductility. These results demonstrate the effectiveness of ARB for integrating GO into the Al 6061 matrix to improve the mechanical performance and interfacial bonding and underscore its potential for advanced composite materials.

The ultrasonic energy field (UEF)-induced grain refinement mechanisms in laser powder direct energy deposition-manufactured Ti5321G alloys were systematically investigated in this study. This study focused on the interplay between recrystallization in the high-temperature solid deposition layers and the ultrasonic cavitation-acoustic streaming effects during molten pool solidification. A novel experimental design was developed to decouple these mechanisms by creating four distinct UEF action zones (without UEF-N, with UEF-S, with UEF-L, and with UEF-S + L) within a single-pass multilayer sample. This approach enabled the dual effects of UEF (recrystallization in solidified layers and ultrasonic cavitation-acoustic streaming effects in liquid pools) to be directly compared. The UEF significantly refined the microstructures, reducing the average grain size by 64.2% (from (399.6 ± 28.6) to (143.1 ± 16.1) µm) in the with UEF-S + L zone, while promoting columnar-to-equiaxed transition, with the equiaxed grain probability increasing from 11.1% (without UEF) to 53.8%. The texture intensity was reduced by approximately 52.4% and the mechanical properties were enhanced, achieving a 6.2% increase in yield strength ((702.0 ± 10.6) MPa) and 31.7% improvement in elongation. Crucially, this study revealed the synergistic effect of the dual-action mechanisms of UEF, where recrystallization and cavitation-acoustic streaming collectively enabled non-linear grain refinement. This study provides a strategy for microstructural control in additive manufacturing, eliminating the need for complex post-processing and thereby advancing the industrial application of high-performance titanium components.

This study investigates the effect of Ce content on the hydrogen storage properties of Ti_0.98Zr_0.02Mn_1.5Cr_0.05V_0.43Fe_0.09Ce_x (x = 0, 0.02, 0.04, and 0.06, at%) alloys. Microstructural analysis of these alloys revealed dendritic microstructures without the segregation of chemical elements, with the C14 Laves phase identified as the dominant phase. After two activation cycles at 4 MPa and 293 K, the alloys exhibited excellent hydrogen absorption properties. The addition of Ce significantly improved the kinetics of the alloys. At x = 0.02, the hydrogen absorption capacity reached 90% of its maximum within 137 s at 293 K. Pressure–composition–temperature curves indicated that hydrogen absorption capacity initially increased first and then decreased with increasing Ce content, reaching a maximum value of 1.85wt% at x = 0.04. Thermodynamic results demonstrated that the enthalpy and entropy of hydrogen absorption followed a similar trend, which was consistent with the variation in hydrogen storage capacity. Thus, the improvement in hydrogen absorption capacity due to the addition of Ce is attributed to the increase in enthalpy. The increase of the lattice constant in the C14 Laves phase and the deoxidization effect of Ce are expected to be beneficial for the improvement of hydrogen absorption kinetics.

To improve the solid–solid interface performance of all solid-state lithium batteries (ASSLBs), a novel sandwich-structured solid electrolyte (SSE, total thickness of 0.7 mm) was investigated. It comprises a central layer of perovskite-type Li_0.37Sr_0.44Zr_0.25Ta_0.75O₃ (LSZT) electrolyte (thickness of 0.5 mm) sandwiched between two layers of composite solid polymer electrolyte (CSPE, each with a thickness of 0.1 mm). The thin CSPE interlayer not only effectively reduces interfacial resistance between LSZT and electrodes, but also suppresses Li-induced reduction degradation of LSZT while ensuring uniform current density distribution across the interface. The SSE demonstrates an ionic conductivity of 8.76 × 10⁻⁵ S·cm⁻¹ at 30°C, increasing to 1.13 × 10⁻³ S·cm⁻¹ at 100°C, with an activation energy of 0.36 eV. In addition, SSE is stable for Li metal and achieves electrochemical stability up to 4.58 V vs. Li⁺/Li. SSE shows outstanding electrode/electrolyte interfacial compatibility and significant suppression of the growth of Li dendrite. Ascribing to these merits, Li ∣ SSE ∣ Li symmetric cell maintained stable operation for 500 h at a current density of 0.3 mA·cm⁻² without short circuit, confirming robust inter-facial compatibility between SSE and Li electrode. The all-solid-state LiFePO₄ ∣ Li battery with SSE has an initial reversible discharge capacity of 109.8 mAh·g⁻¹ and a reversible capacity of 118.1 mAh·g⁻¹ after 50 cycles at a charge/discharge rate of 0.1C (30°C), demonstrating good cycling performance.

Controllable synthesis of one-dimensional nanowires (NWs) is crucial for their large-scale applications, but it usually requires complicated catalyst designs with multiple compositions and careful tuning of synthesis parameters. In this study, we performed a systematic investigation into the impact of the shape of Au particles on the geometry and composition of the obtained NWs. We discovered that octahedral, dodecahedral, and cubic Au particles selectively catalyze the growth of Ga, GaAs, and Ga/GaAs heterojunction NWs, respectively. The mechanism stems from the difference in the solubility of Ga in Au catalysts with distinct shapes (i.e., curvatures) due to the Gibbs–Thomson (G–T) effect: Au octahedrons (7.42 nm), featuring smaller curvature radii, enhance the solubility of Ga precursors, enabling efficient diffusion and faster growth of Ga NWs; Au dodecahedrons (11.22 nm) with larger curvature radii exhibit moderate Ga solubility, favoring the growth of GaAs NWs; Au cubes (10.51 nm) with intermediate Ga solubility, yield Ga/GaAs heterojunction NWs. Finally, we fabricated NW field effect transistors (FETs) and revealed that the Ga NWs exhibited promising electrical characteristics with a resistivity of 2.54 × 10⁻⁴ Ω·m, and GaAs NWs showed p-type characteristics. All these results illustrate the promising potential for tuning the geometry and composition of NWs by a single parameter, i.e., merely changing the shape of a single Au particle.

To enhance the visible light response of titanium dioxide (TiO₂), titanium carbide (TiC) nanoparticles (NPs) were thermally treated in carbon powder, effectively overcoming the challenges associated with conventional doping methods. During the treatment, a TiO₂ thin shell with oxygen vacancies (OVs) formed around the TiC NPs, creating a shell–core structure S-scheme photocatalyst. Transmission electron microscopy (TEM) and ultraviolet-visible (UV–vis) spectroscopy confirmed the successful formation of the TiO₂ shell. By optimizing the shell thickness, the TiO₂–TiC shell–core structure achieved an ideal shell–core ratio, resulting in strong visible light absorption (400–800 nm), and the degradation rate constant of Rhodamine B (RhB) of sample cHT500 reached 0.0687 min⁻¹, which is 20.8 times higher than that of pristine TiO₂ (0.0033 min⁻¹) under visible-light irradiation. In addition, cytocompatibility tests showed that sample cHT500 exhibits favorable cell viability, which is comparable to that of TiO₂ nanoparticles, and thus remarkably mitigates the poor biocompatibility inherent to TiC, making them promising candidates for biomedical and photocatalytic applications.