Transition metal diborides (TMB₂s) are renowned for their high melting point and exceptional wear, corrosion, and erosion resistance, making them promising candidate materials for applications in extreme environments. As such, there is an urgent need for reliable material design tools for TMB₂s to improve efficiency in developing new materials. To address this need, we have developed a domain-specific medium-scale interatomic potential model for TMB₂s that encompasses elements Ti, Zr, Hf, V, Nb, Ta, Cr, Mo, W, and B. The prediction errors in energy and force of our model are 8.8 meV/atom and 387 meV/Å, respectively. Furthermore, the model demonstrates high accuracy in predicting various material properties, including lattice parameters, elastic constants, equations of states, and melting points, as well as grain boundary segregations. By providing a reliable and efficient tool for material design, this model will play a crucial role in the development of new, high-performance TMB₂s for use in extreme environments.

This study addresses the challenge of predicting the tensile stress of natural rubber with limited molecular dynamics simulation data, which is a crucial mechanical property for this material. Molecular dynamics (MD) simulations are limited by their scale and computational cost, making it difficult to obtain sufficient data to train machine learning algorithms. To overcome this limitation, we propose a machine learning framework involving three stages: (1) utilizing a Variational Autoencoder (VAE) to rapidly expand the data diversity; (2) employing Ordinary Kriging (OK) to label the VAE-generated virtual samples; and (3) training gradient enhanced regression [Gradient Boosting Regression (GBR)] models by using relevant data on tensile stress in natural rubber. The results demonstrate that the generated data exhibits enhanced rationality, significantly improving the accuracy and reliability of various regression models. This approach provides an effective solution to the problem of data scarcity in MD simulations.

There is growing interest in novel MAX phase materials for various applications ranging from aircraft/spacecraft and defense to energy and electronics due to their unique combination of metallic and ceramic properties. Traditional materials discovery has mostly relied on human intuition coupled with rigorous experiments; however, this approach has been time-consuming and inefficient. Over the last few decades, advances in fundamental and data-driven approaches such as first-principles modeling, materials informatics, machine learning and optimization, coupled with an exponential rise in computational power, have enabled faster and more efficient materials discovery. Here, we present an exploration of high elastic modulus novel boride-based M₂AX phase materials using a combination of the aforementioned methods. Specifically, an ensemble of gradient boosted machine learning models was developed to predict the elastic modulus from informatics-based structural features by leveraging a dataset of Density Functional Theory (DFT)-predicted elastic moduli for 223 M₂AX phase materials (carbides and nitrides). Using Bayesian optimization, inverse modeling was carried out to maximize the model-predicted elastic modulus by identifying the optimal features. Finally, model predictions for 1,035 candidate M₂AX materials were generated to compare their features with the optimal features to identify potential novel promising materials. We found that Ta₂PB, Nb₂PB, and V₂PB have similar high elastic moduli (371.7, 351.5, and 347.4 GPa) to their carbide counterparts (364.7, 357.7, and 373.5 GPa), and our results support the possibility that borides can be a viable tertiary element for M₂AX phases.

In this study, we employ data-driven and first-principles methods (machine learning, density-functional theory and language model) to comprehensively explore crystal structures, electronic properties and applications of an emerging perovskite material, gadolinium scandate (GdScO₃), which is an intriguing material that demonstrates potentials in electronics and optics. Using advanced machine learning algorithms based on genetic programming, we have discovered new crystal structures of GdScO₃ that have not been previously reported, which are further examined via density functional theory (DFT) calculations and language models to provide detailed insights into their electronic and optical properties and potential applications. Our findings reveal novel new stable phases of GdScO₃ and highlight the intricate influence of structural variations on its electronic band structures and light absorption properties. A subsequent domain-specific language model analysis indicates its possibility for photovoltaics pending further efforts to engineer defects revealed in the first-principles calculations. The integration of machine learning with first-principles calculations demonstrates a feasible approach for accelerating the exploration and analysis of materials. This work enriches the understanding of GdScO₃ and establishes a robust framework for exploration and ontological analysis of new functional materials combining diverse data-driven techniques (e.g., language model and genetic programming) and first-principles methods.

As a generalized method of mathematical statistics, machine learning (ML) is playing an increasingly significant role in the realm of materials design. More sophisticated methodologies for in-depth understandings and wide applications have been developed from initially simple data relation mappings. The present work first summarizes the basic technical issues of ML and then systematically reviews the main implementation strategies for ML methods in accelerating materials research and development process in recent years, encompassing three primary aspects. Firstly, it is necessary to establish the relationship between the key characteristic parameters and properties in any given materials system for a better prediction and exploration of new materials. Then, the computational algorithms in materials science need to be optimized to replace complex calculations with model-predicted data. Finally, the ML methods are applied to summarize the one-dimensional property data and two-dimensional microstructural images of materials to establish standardized analysis methods. During this process, the domain knowledge in a specific system is of great significance to improving the prediction accuracy and efficiency of ML methods, whether pre-processing experimental or computational databases. The powerful capability of ML methods to handle high-dimensional data will enable researchers to make more effective decisions in materials design. In the future, the relationship between the microstructure and mechanical properties, which is necessary to establish a more effective search engine for alloys with targeted mechanical properties, will be the focus of ML mechanical properties of alloy materials.

Crystal structure prediction (CSP) plays a crucial role in condensed matter physics and materials science, with its importance evident not only in theoretical research but also in the discovery of new materials and the advancement of novel technologies. However, due to the diversity and complexity of crystal structures, trial-and-error experimental synthesis is time-consuming, labor-intensive, and insufficient to meet the increasing demand for new materials. In recent years, machine learning (ML) methods have significantly boosted CSP. In this review, we present a comprehensive review of the ML models applied in CSP. We first introduce the general steps for CSP and highlight the bottlenecks in conventional CSP methods. We further discuss the representation of crystal structures and illustrate how ML-assisted CSP works. In particular, we review the applications of graph neural networks (GNNs) and ML force fields in CSP, which have been demonstrated to significantly speed up structure search and optimization. In addition, we provide an overview of advanced generative models in CSP, including variational autoencoders (VAEs), generative adversarial networks (GANs), and diffusion models. Finally, we discuss the remaining challenges in ML-assisted CSP.