Rapid discovery of short-wave infrared (SWIR) detection materials requires efficient strategies to identify candidates with suitable bandgaps, favorable carrier transport properties, and structural stability. Here, we propose a high-throughput screening (HTS) framework that integrates machine learning (ML) models with density functional theory (DFT) calculations to accelerate the prediction and validation of infrared-detection materials [see Graphical Abstract]. Using a curated dataset of 1327 I-X-VI chalcogenide compounds retrieved from the Materials Project database, we trained five regression models-random forest, gradient boosting, support vector regression, extreme gradient boosting, and decision tree-to predict electronic bandgaps with high accuracy and computational efficiency. The optimized extreme gradient boosting regression (XGBR) model delivers a test-set coefficient of determination (R²) of 0.945, a mean absolute error (MAE) of 0.150 eV, and a mean squared error (MSE) of 0.056 eV, with a 5-fold cross-validation (R²) of 0.927, verifying its robust prediction performance and generalization ability. This ML-guided screening highlights five promising chalcogenides: KGaSe₂, KGaTe₂, KInSe₂, KInTe₂, and CsInTe₂. These candidates were further evaluated using first-principles DFT calculations to assess their band structures, density of states, and carrier effective masses. Among them, KGaSe₂ exhibits a direct bandgap of ~0.8 eV, low effective mass, and excellent thermodynamic stability, making it a highly attractive candidate for SWIR detection. This work demonstrates the power of combining ML and DFT in accelerating the discovery of infrared (IR) optoelectronic materials and provides a scalable, generalizable approach for next-generation photodetector design.

Sustainable hydrogen energy offers a promising solution to the growing global energy demand associated with fossil fuel consumption. The development of efficient electrocatalysts for the hydrogen evolution reaction (HER) is important, yet the high computational cost of density functional theory (DFT) limits the rapid screening of candidate materials. In this work, a machine learning-assisted framework integrated with DFT calculations is proposed to systematically investigate the HER performance of carbon nanotube (CNT)-supported single-atom catalysts (SACs). A dataset consisting of Gibbs free energy of hydrogen adsorption (ΔG_H*) was constructed from DFT calculations, including 84 M-N₄-CNT(n, n) models involving 28 transition-metal centers anchored on CNTs with three different chirality indices. Based on selected intrinsic transition-metal features and the CNT chirality index, a random forest regression (RFR) model was identified as the optimal model after comparison with multiple machine learning algorithms for predicting ΔG_H*. The RFR model exhibited excellent predictive accuracy, achieving a coefficient of determination (R²) of 0.98 on the test set. Notably, when applied to previously unseen M-N₄-CNT(7, 7) structures, the model maintained high reliability (R² = 0.96), demonstrating strong generalization capability. Machine learning identified Fe-N₄-CNT(7, 7) as a highly promising HER electrocatalyst, with further DFT-based kinetic analysis showing that it follows a Volmer-Tafel reaction pathway. In addition, the SISSO algorithm was employed to derive an interpretable descriptor for ΔG_H* based on elemental properties, achieving high fitting accuracy across different chirality indices. This descriptor provides an efficient tool for rapid catalyst screening while offering mechanistic insights into the key factors governing HER activity in M-N₄-CNT systems.

Lithium-ion batteries are extensively utilized in applications like new energy vehicles and aerospace owing to their high energy density and safety, but their service life diminishes due to irreversible capacity degradation from repeated charge-discharge cycles, making accurate remaining useful life (RUL) prediction critical for reliability and operational safety. Current data-driven methods often struggle with long-range dependencies, noise from capacity regeneration, and efficient data utilization. To address these challenges, this study introduces a novel hybrid neural network architecture that integrates complementary ensemble empirical mode decomposition with adaptive noise (CEEMDAN) for denoising preprocessing with a Transformer-convolutional neural network (CNN)-bidirectional gated recurrent unit (BiGRU) model. The framework employs health indicators extracted from voltage and current profiles as inputs, where CEEMDAN mitigates interference effects, the Transformer captures global degradation trends via self-attention mechanisms, the CNN extracts localized short-term features, and the bidirectional GRU models temporal dependencies bidirectionally. Experimental validation on National Aeronautics and Space Administration (NASA) and our own test datasets demonstrates that the proposed approach significantly outperforms other models in key metrics such as mean absolute error (MAE), root mean square error (RMSE) and mean absolute percentage error (MAPE), achieving high accuracy even with minimal training data (e.g., only 40% of cycles). Furthermore, cross-dataset validation demonstrates robust generalization of our model, achieving MAPE below 3.5% when transferring between NASA and our battery data without retraining. This hybrid model offers a robust, data-efficient solution for enhancing RUL prediction in practical battery management systems, with strong generalization across diverse battery types.

The Al₈Cu₄X phase has emerged as a promising heat-resistant strengthening candidate for Al-Cu alloys, mitigating the instability of Ω/θ′ precipitates at elevated temperatures. In this work, we employ high-throughput first-principles calculations to systematically investigate 57 Al₈Cu₄X compounds, focusing on their thermodynamic stability and phase trans-formation behavior. Density functional theory calculations reveal negative formation energies for 53 compounds, and those containing rare earth elements, Ca, Sr, and Y are identified as favorable candidates for forming microscale phases at high temperatures. Phase transformation energies exhibit a pronounced periodic trend, with 33 compounds showing negative values, supporting the feasibility of forming nanoscale strengthening precipitates via high-temperature phase transformation. Symbolic regression analysis further identifies atomic volume as the primary descriptor governing the phase transformation energies, while bond order analysis demonstrates that the enhanced stability originates from a strengthened Al-Al bonding network and newly introduced Al-X bonds within the Al₈Cu₄X structures. Overall, this work provides a theoretical foundation for the future design and application of heat-resistant Al₈Cu₄X phases in aluminum alloys.

Under extreme service conditions, adiabatic shear banding critically limits the performance of titanium alloys in warhead applications, creating an urgent demand for strategies to achieve strength-ductility synergy. In this work, a knowledge-enabled data-driven multi-objective optimization framework is proposed to investigate the composition of near-α titanium alloys under high strain rates. By integrating domain knowledge with twelve machine learning models, key performance parameters (KPPs) governing strength are identified through feature engineering, including strain rate, Fermi energy, and phase formation parameters, while ductility is controlled by the KPPs of strain rate, bulk/shear modulus (B/G) ratio, and mixing enthalpy. Using a gradient boosting regression tree model for strength prediction [test the coefficient of determination (R²) = 0.91] and a random forest model for ductility prediction (test R² = 0.82), the nondominated sorting genetic algorithm II (NSGA-II) is integrated to identify 14 Pareto-optimal alloys from a pool of 200,000 candidate compositions of near-α titanium alloys (Ti-Al-V-Mo-Zr-Sn system). A breakthrough combination of 1,600 MPa dynamic compressive strength and 26% ductility at a strain rate of 3,000 s^-1 is achieved, which is superior to that of the TA15 alloy, as confirmed by a constitutive equation model. This framework successfully designs novel near α-titanium alloys with strength-ductility synergy through knowledge-driven feature engineering and multi-objective optimization algorithms, establishing a new paradigm for the intelligent design of titanium alloys under extreme conditions.

Functional oxide films offer precise control over diverse properties through tunable physical characteristics and interface effects, with their functionality primarily determined by morphology. However, conventional methods are incapable of obtaining large-scale morphological data and face significant challenges in data identification and classification, which fundamentally limit the rapid assessment of thin film properties and functional screening. Herein, we establish a comprehensive morphological database of oxide films utilizing high-throughput experimental methods and develop a machine learning framework for automated identification and classification of atomic force microscopy data. Using gradient-thickness SrRuO₃ films as a representative example, this framework achieves enhanced performance through hyperparameter optimization and strategic adjustments, ultimately reaching a classification accuracy of 86.67% in independent tests, demonstrating its effectiveness in morphology analysis of functional oxide films. Furthermore, this approach shows significant potential for automated microstructure analysis of complex oxides and is expected to accelerate research on structure-property correlations in functional oxide films.

Multi-component carbide ceramics have garnered significant attention as ultra-high-temperature structural materials due to their exceptionally high melting points and excellent mechanical properties. In this work, we systematically investigate the synergistic effects of C vacancies and Ti alloying on the thermodynamic stability and elastic behavior of (Zr, Ti)C_x carbides using first-principles calculations. Specific cluster structural models of [C-M₆](C,□)₅ (M = Zr/Ti, □ = vacancy) were constructed by considering the local chemical short-range orders of elemental distribution and the ordering of vacancies on C sublattice, which were then employed as inputs for first-principles calculations. The results reveal that the introduction of C vacancies decreases the free energy at high temperatures and enhances the thermodynamic stability, whereas Ti substitution for Zr tends to reduce stability. Notably, the ternary carbide Zr₅Ti₁C₅ ([C-Zr₅Ti₁](C,□)₅) with an equimolar ratio of Ti-to-vacancy exhibits superior high-temperature thermodynamic stability. Analysis of entropy contributions indicates that both vacancies and Ti addition primarily alter the free energy by modifying the lattice vibration modes, an effect dominated by the vibrational entropy. These two types of defects weaken the M-C bond strength, resulting in reduced binding energy and Young’s modulus. Furthermore, this synergistic effect considerably lowers the critical temperature required to stabilize the single-phase solid solution structure in multi-component carbides, which is attributed to a decrease in mixing enthalpy and an increase in configurational entropy caused by vacancies. The cluster-model-embedded first-principles approach offers valuable insight for designing high-performance carbides in complex ceramic systems.

Carbon segregation is a persistent defect in continuous casting of large-diameter steel billets, leading to deteriorated mechanical properties and compromised service reliability. Conventional empirical or machine learning models generally estimate segregation indices but cannot resolve local variations of carbon distribution across billet sections. In this work, a microstructure-informed convolutional neural network (CNN) framework is proposed to predict and map carbon segregation in 600 mm round 42CrMo steel billets. A comprehensive dataset comprising microstructural images and corresponding carbon content measurements was established. The customized CNN achieved a testing accuracy of 81.3% with a mean absolute error of 0.012 wt.% and showed good robustness in out-of-sample validation. Compared with transfer learning models (VGG16, VGG19, etc.), the customized architecture exhibited superior generalization on this domain-specific dataset. Contrast-enhanced imaging significantly improved predictive performance, while Gradient-weighted Class Activation Mapping visualizations highlighted key microstructural regions correlated with carbon distribution, providing interpretability. This study demonstrates a proof-of-concept methodology to achieve quantitative mapping of segregation patterns in large-diameter 42CrMo billets, offering a complementary tool to traditional metallurgical analysis and providing a workflow that may support future data-driven research on segregation formation mechanisms and process optimization in steel casting when extended to additional steels and casting conditions.

In this study, glass fiber-reinforced polyamide 6 (PA6-GF) was selected as a representative system to establish a predictive framework for thermal-oxidative aging behavior. Through nonlinear fitting of traditional empirical models followed by the evaluation of twelve machine learning algorithms, it was found that the conventional models exhibited limited predictive accuracy and generalization capability, whereas the machine learning approaches were able to more effectively capture the complex nonlinear interactions among temperature, oxygen partial pressure, specimen thickness, and aging time. To further elucidate the underlying mechanisms, SHapley Additive exPlanations (SHAP) analysis was employed, highlighting the distinct roles and relative contributions of aging time, oxygen partial pressure, temperature, and thickness in governing the thermal-oxidative aging process. These findings enhance the understanding of multi-factor aging mechanisms and provide practical guidance for improving the long-term durability and reliability of engineering components operating under complex service conditions.

High-entropy alloys (HEAs) have attracted extensive attention due to their exceptional mechanical, physical, and chemical properties, making them promising candidates for extreme environments. Understanding the complex structure-property relationships in these multi-principal element systems is crucial for discovering and designing high-performance HEAs. However, their vast compositional space and high-dimensional chemical complexity pose major challenges to traditional trial-and-error design. Machine learning (ML) offers a transformative strategy to overcome these barriers by enabling data-driven exploration. This perspective first reviews the critical challenges currently limiting HEA development, then summarizes recent ML breakthroughs in phase formation prediction, multi-objective optimization, and accelerated atomistic simulations. Finally, we discuss ongoing challenges and propose future opportunities for integrating ML with experimental and computational methods to create more interpretable, data-efficient, and autonomous ML-driven HEA design frameworks.