The high-throughput highway to computational materials design

High-throughput computational materials design provides one efficient solution to accelerate the discovery and development of functional materials. Its core concept is to build a large quantum materials repository and to search for target materials with desired properties via appropriate materials descriptors in a high-throughput fashion, which shares the same idea with the materials genome approach. This article reviews recent progress of discovering and developing new functional materials using high-throughput computational materials design approach. Emphasis is placed on the rational design of high-throughput screening procedure and the development of appropriate materials descriptors, concentrating on the electronic and magnetic properties of functional materials for various types of industrial applications in nanoelectronics.

Interscalable material microstructure organization in performance-based computational design

Computational material design (CMD) employs quantum mechanical simulations, density functional theory, and machine learning techniques to correlate electronic structural attributes with physical and chemical properties of materials. Over the last decade, CMD has proven to be critical to the advancement of materials science and a variety of engineering fields. This contribution provides an overview of CMD’s success in driving materials discovery for catalysis in the context of sustainable energy applications. Specifically, we discuss how CMD has enabled the development of catalysts for three electrochemical processes that are critical to sustainable energy applications, oxygen reduction reactions, water oxidation reactions, and CO2 reduction reactions. We illustrate how CMD provides a powerful and efficient method for understanding underlying reaction mechanisms as well as predicting and optimizing catalyst properties. Furthermore, we demonstrate how this strategic approach has enabled researchers to effectively navigate the vast chemical space of potential catalysts and rapidly identify novel materials possessing desirable electronic structures and catalytic activity.

Maximally-localised Wannier functions (MLWFs) are routinely used to compute from first-principles advanced materials properties that require very dense Brillouin zone integration and to build accurate tight-binding models for scale-bridging simulations. At the same time, high-throughput (HT) computational materials design is an emergent field that promises to accelerate reliable and cost-effective design and optimisation of new materials with target properties. The use of MLWFs in HT workflows has been hampered by the fact that generating MLWFs automatically and robustly without any user intervention and for arbitrary materials is, in general, very challenging. We address this problem directly by proposing a procedure for automatically generating MLWFs for HT frameworks. Our approach is based on the selected columns of the density matrix method and we present the details of its implementation in an AiiDA workflow. We apply our approach to a dataset of 200 bulk crystalline materials that span a wide structural and chemical space. We assess the quality of our MLWFs in terms of the accuracy of the band-structure interpolation that they provide as compared to the band-structure obtained via full first-principles calculations. Finally, we provide a downloadable virtual machine that can be used to reproduce the results of this paper, including all first-principles and atomistic simulations as well as the computational workflows.

<sec><p indent="0mm">Two-dimensional (2D) magnetic materials have emerged as promising candidates for next-generation spintronic devices, owing to their distinct characteristics and potential for efficient low-power information processing. Since the successful exfoliation of graphene, extensive efforts have focused on realizing intrinsic magnetism in 2D materials. The experimental breakthrough occurred in 2017 with the discovery of intrinsic magnetism in monolayer CrI₃ and bilayer Cr₂Ge₂Te₆, leading to the synthesis of diverse 2D van der Waals (vdW) magnetic materials. These materials consist of atomically thin layers stacked together by weak vdW interactions. The weak interlayer coupling enables precise control over the relative crystal alignment between layers, resulting in diverse stacking configurations. Recent advances in synthesis methods and polymer-assisted transfer techniques have made on-demand stacking of vdW layers feasible, revealing novel physical phenomena such as stacking-dependent magnetic ordering and non-collinear spin textures in moiré superlattices. Moreover, sliding ferroelectricity, arising from polar stacking of nonpolar monolayers, provides alternative pathways for realizing magnetoelectric coupling, leading to novel correlated physics and device applications. These discoveries demonstrate that stacking order significantly influences the charge and spin redistribution between adjacent layers, thus affecting the topological properties, electronic correlations, and spin behaviors. Recent studies have focused on modulating the stacking order of 2D vdW magnetic materials, particularly through the manipulation of structural degrees of freedom such as interlayer sliding and twisting. These tunable variables provide a comprehensive platform for exploring and engineering novel quantum states in bilayer magnetic systems. </sec><sec> First-principles calculations based on fundamental quantum mechanical laws play a crucial role in materials simulation, particularly in predicting stacking configurations of bilayer magnetic materials through interlayer coupling analysis. The emergence of data-driven paradigms has transformed materials science research through the integration of high-throughput calculations and deep learning methods. High-throughput computational materials design enables efficient generation and analysis of extensive materials databases, overcoming traditional computational limitations. Simultaneously, advances in deep learning have enabled artificial neural networks to efficiently process large-scale datasets and establish structure-property relationships without requiring first-principles calculations. This approach, integrating first-principles calculations, high-throughput computation, and deep learning, significantly accelerates the discovery of novel bilayer magnetic materials with desired properties. </sec><sec> In this review, we first introduce the intriguing physical phenomena observed in bilayer vdW magnetic materials controlled through interlayer sliding and twisting, including stacking-dependent magnetism, magnetoelectric coupling, and the emergence of noncollinear spins in moiré superlattices. We then present recent progress in high-throughput computational design and deep learning applied to bilayer magnetic materials. Through these studies, we systematically describe the fundamental mechanisms of interlayer exchange coupling, stacking order modulated electronic and magnetic properties, and data-driven research in magnetic materials. Finally, we provide insights into future directions for material design and artificial intelligence-enabled research in 2D magnetic materials. </sec>

Proceedings of the Second Workshop on Theory meets Industry (Erwin-Schrödinger-Institute (ESI), Vienna, Austria, 12–14 June 2007)

In a normal study, if the structure or composition of a material is given, it is possible to study the properties and functions of the material. This is a direct challenge. As solution techniques have developed considerably, they have become quite sophisticated, both experimentally and computationally. The study area known as computational materials design uses computer modeling to find a solution to this “inverse problem.” Using computer simulations, particularly those based on quantum theory, is a process known as computational materials design. The authors would like to introduce the current status and recent research on reactions at interfaces. The authors will explain the method of predicting the magnetism of a system called the dilute magnetic semiconductor, which has been attracting attention as the base material for semiconductor spintronics, based on a first-principles calculation. After a brief introduction to semiconductor spintronics, he will explain the method of handling the regularity peculiar to dilute magnetic semiconductors, using the band calculation method. New materials derived from carbon, such as carbon nanotubes, fullerenes, and grapheme, have been added to the traditional functional carbon materials such as diamond, graphite, and activated carbon. The tools used for materials informatics. With the aim of accelerating materials science research, data-driven materials research, that is, materials informatics research is becoming more active. One is the crystal structure prediction tool, CrySPY, and the other is the descriptor generation tool, LIDG, used for machine learning. as supercomputers, the knowledge and techniques required to master them are also changing, becoming more complex and sophisticated each year. The various techniques will be introduced in relation to the features of current computers in terms of single-CPU performance optimization and high-parallelism performance optimization.

High-throughput computational materials design is an emerging area in materials science, which is based on the fast evaluation of physical-related properties. The lattice thermal conductivity (κ) is a key property of materials for enormous implications. However, the high-throughput evaluation of κ remains a challenge due to the large resources costs and time-consuming procedures. In this paper, we propose a concise strategy to efficiently accelerate the evaluation process of obtaining accurate and converged κ. The strategy is in the framework of phonon Boltzmann transport equation (BTE) coupled with first-principles calculations. Based on the analysis of harmonic interatomic force constants (IFCs), the large enough cutoff radius (rcutoff), a critical parameter involved in calculating the anharmonic IFCs, can be directly determined to get satisfactory results. Moreover, we find a simple way to largely (~10 times) accelerate the computations by fast reconstructing the anharmonic IFCs in the convergence test of κ with respect to the rcutof, which finally confirms the chosen rcutoff is appropriate. Two-dimensional graphene and phosphorene along with bulk SnSe are presented to validate our approach, and the long-debate divergence problem of thermal conductivity in low-dimensional systems is studied. The quantitative strategy proposed herein can be a good candidate for fast evaluating the reliable κ and thus provides useful tool for high-throughput materials screening and design with targeted thermal transport properties.

Rare earth permanent magnets play a key role in view of the increasing market for green energy applications and new materials which can compete with Nd2Fe14B but are environmentally less problematic are highly sought after. Fe-based ThMn12 phases contain less RE material than the commercially used compounds and have attracted quite some interest. An efficient and resource saving way to identify out of this class new phases which have the potential to become permanent magnets is by computational materials design. However, predicting the magnetic properties of 4f systems is challenging because of the localization of the 4f electrons which determines to a large extend the level of theory needed for a reliable description and prediction.We have shown that the 4f electrons in Sm 1:12 can be viewed as fully localized and can be safely treated as core electrons while for CeFe11Ti the correlation effects are smaller and the 4f states can still be viewed as valence states [1,2]. In the case of NdFe11Ti the 4f electrons are partially localized as can be seen from the hybridization function in Fig. 1. The hybridization function serves as a qualitative measure for the interaction of the 4f electrons with the valence electrons in a system [3].As a consequence of the partial localization, the cone type magnetocrystalline anisotropy (MCA) of the latter system is only observed in a DFT+U description with an intermediate Hubbard U value, see Fig. 2. Assuming full localization (4f in core) results in a uniaxial MCA which contradicts the experimental findings at low temperatures. Plain DFT also fails. It is also discussed in how far the strong dependence of the MCA and related magnetic properties of NdFe11Ti on the theoretical description influences the prediction of new phases. As test case the quaternary system Nd1-xYxFe12-yTiy was chosen.Supported by NOVAMAG (EU686056), the Swedish Foundation for Strategic Research (EM16-0039), and SNIC(Swedish National Infrastructure for Computing).   New potential materials for rare earth lean permanent magnets from computational design and the challenge of the 4f electrons

The physics that associated with the performance of lithium secondary batteries (LSB) are reviewed. The key physical problems in LSB include the electronic conduction mechanism, kinetics and thermodynamics of lithium ion migration, electrode/electrolyte surface/interface, structural (phase) and thermodynamics stability of the electrode materials, physics of intercalation and deintercalation. The relationship between the physical/chemical nature of the LSB materials and the batteries performance is summarized and discussed. A general thread of computational materials design for LSB materials is emphasized concerning all the discussed physics problems. In order to fasten the progress of the new materials discovery and design for the next generation LSB, the Materials Genome Initiative (MGI) for LSB materials is a promising strategy and the related requirements are highlighted.

High-throughput computational and experimental design of materials aided by machine learning have become an increasingly important field in material science. This area of research has emerged in leaps and bounds in the thermal sciences, in part due to the advances in computational and experimental methods in obtaining thermal properties of materials. In this paper, we provide a current overview of some of the recent work and highlight the challenges and opportunities that are ahead of us in this field. In particular, we focus on the use of machine learning and high-throughput methods for screening of thermal conductivity for compounds, composites and alloys as well as interfacial thermal conductance. These new tools have brought about a feedback mechanism for understanding new correlations and identifying new descriptors, speeding up the discovery of novel thermal functional materials.

Predicting unknown inorganic compounds and their crystal structure is a critical step of high-throughput computational materials design and discovery. One way to achieve efficient compound prediction is to use data mining or machine learning methods. In this chapter we present a few algorithms for data mining compound prediction and their applications to different materials discovery problems. In particular, the patterns or correlations governing phase stability for experimental or computational inorganic compound databases are statistically learned and used to build probabilistic or regression models to identify novel compounds and their crystal structures. The stability of those compound candidates is then assessed using ab initio techniques. Finally, we report a few cases where data mining driven computational predictions were experimentally confirmed through inorganic synthesis.

Recently, a series of double-perovskite halide compounds such as Cs2AgBiCl6 and Cs2AgBiBr6 have attracted intensive interest as promising alternatives to the solar absorber material CH3NH3PbI3 because they are Pb-free and may exhibit enhanced stability. The thermodynamic stability of a number of double-perovskite halides has been predicted based on density functional theory (DFT) calculations of compound formation energies. In this paper, we found that the stability prediction can be dependent on the approximations used for the exchange-correlation functionals, e.g., the DFT calculations using the widely used Perdew, Burke, Ernzerhof (PBE) functional predict that Cs2AgBiBr6 is thermodynamically unstable against phase-separation into the competing phases such as AgBr, Cs2AgBr3, Cs3Bi2Br9, etc., obviously inconsistent with the good stability observed experimentally. The incorrect prediction by the PBE calculation results from its failure to predict the correct ground-state structures of AgBr, AgCl, and CsCl. By contrast, the DFT calculations based on local density approximation, optB86b-vdW, and optB88-vdW functionals predict the ground-state structures of these binary halides correctly. Furthermore, the optB88-vdW functional is found to give the most accurate description of the lattice constants of the double-perovskite halides and their competing phases. Given these two aspects, we suggest that the optB88-vdW functional should be used for predicting thermodynamic stability in the future high-throughput computational material design or the construction of the Materials Genome database for new double-perovskite halides. Using different exchange-correlation functionals has little influence on the dispersion of the conduction and the valence bands near the electronic bandgap; however, the calculated bandgap can be affected indirectly by the optimized lattice constant, which varies for different functionals.

Moving the boundaries of knowledge, scientific research and technologies forward through the field of computational materials science requires a combination of fundamental understanding of materials properties, an appreciation of the limitations with our current approaches, and the ability to embrace new and emergent technologies. This short Essay highlights the works of the authors in the special issue “Computational Materials Design” published in Advanced Theory and Simulations. It emphasises the role that computational materials design plays in rationalizing and guiding experimental efforts in in the fields of catalysis, semiconductors, hydrogels, and solid‐state electrolytes. Increases in computational power together with accurate hybrid functionals within density functional theory are enabling more reliable and trustworthy descriptions of solids for various electronic and optoelectronic applications. Additionally, high‐throughput screening and machine learning are rapidly becoming indispensable tools in computational materials science across diverse applications such as engineering and predicting new catalysts. Such advances are setting the pace for our discovery of new and novel materials of the future.

Computational materials design aims at designing, simulating and predicting innovative materials. This paper reviews a few open challenges in the domain of computational materials design related to the colour appearance of materials. These relatively recent fields of research necessitate revisiting the fundamentals of colorimetry, such as reflectance models. First, we address the virtual design and the simulation of innovative materials, which requires modelling their appearance and fundamental properties. Then we discuss the latest advancements in the machine‐learning domain that have highly revolutionised computational and data‐minded methodologies, which are used for the design innovation, discovery and optimisation of materials.