Materials Genome Initiative Research Articles

NJmat: Data-Driven Machine Learning Interface to Accelerate Material Design.

Machine learning techniques have significantly transformed the way materials scientists conduct research. However, the widespread deployment of machine learning software in daily experimental and simulation research for materials and chemical design has been limited. This is partly due to the substantial time investment and learning curve associated with mastering the necessary codes and computational environments. In this paper, we introduce a user-friendly, data-driven machine learning interface featuring multiple "button-clicking" functionalities to streamline the design of materials and chemicals. This interface automates the processes of transforming materials and molecules, performing feature selection, constructing machine learning models, making virtual predictions, and visualizing results. Such automation accelerates materials prediction and analysis in the inverse design process, aligning with the time criteria outlined by the Materials Genome Initiative. With simple button clicks, researchers can build machine learning models and predict new materials once they have gathered experimental or simulation data. Beyond the ease of use, NJmat offers three additional features for data-driven materials design: (1) automatic feature generation for both inorganic materials (from chemical formulas) and organic molecules (from SMILES), (2) automatic generation of Shapley plots, and (3) automatic construction of "white-box" genetic models and decision trees to provide scientific insights. We present case studies on surface design for halide perovskite materials encompassing both inorganic and organic species. These case studies illustrate general machine learning models for virtual predictions as well as the automatic featurization and Shapley/genetic model construction capabilities. We anticipate that this software tool will expedite materials and molecular design within the scope of the Materials Genome Initiative, particularly benefiting experimentalists.

Thermodynamic modeling of Bi2Te3 in the defect energy formalism

Bi2Te3 is a promising thermoelectric material that is often touted as one of the best-performing low-temperature thermoelectric materials. As a result, it has been widely used commercially, both for clean energy generation and in cooling devices. Like many other thermoelectric materials, defects play a key role in the performance of Bi2Te3. As a result, numerous studies have attempted to experimentally and computationally map out the dominant defects in the phase, these include efforts to determine the dominant defect, estimate defect energies, and predict their concentration. The computer coupling of phase diagrams and thermochemistry (CALPHAD) is one of many tools under the auspices of the materials genome initiative (MGI) that enables the rapid design of new functional materials with improved properties. The Defect energy formalism (DEF) with a charged sublattice, an offshoot of the Compound energy formalism (CEF), provides a way to directly include first-principle defect energy calculations into CALPHAD descriptions of solid phases. The introduction of the charge sublattice enables the estimation of the free carrier concentrations in the phase. Here we apply the DEF to the Bi2Te3 system, emphasizing the robustness of the DEF in describing meaningful endmembers and the elimination of fitting parameters. Unlike previous assessments using the Wagner–Schottky defect model, we include the description of the charged defects in our assessment. The DEF with a charged sublattice provides a good prediction of the non-stoichiometry of the phase when compared with experimental data and also predicts a thermodynamic defect concentration at low temperature that is physically reasonable.

Novel Composites for Lightweighting in Deep Space Applications

Abstract The evolution of carbon nanotube composites, as inspired by the Materials Genome Initiative, unlocks new potential for materials sustainability in challenging deep space applications. This article describes research and development work at the Institute for Ultra-Strong Composites by Computational Design (US-COMP), a NASA Space Technology Research Institute consortium. US-COMP has developed a novel composite material with mechanical properties per unit mass that exceed state-of-the-art structural composites.

LUNAR: Automated Input Generation and Analysis for Reactive LAMMPS Simulations.

Generating simulation-ready molecular models for the LAMMPS molecular dynamics (MD) simulation software package is a difficult task and impedes the more widespread and efficient use of MD in materials design and development. Fixed-bond force fields generally require manual assignment of atom types, bonded interactions, charges, and simulation domain sizes. A new LAMMPS pre- and postprocessing toolkit (LUNAR) is presented that efficiently builds molecular systems for LAMMPS. LUNAR automatically assigns atom types, generates bonded interactions, assigns charges, and provides initial configuration methods to generate large molecular systems. LUNAR can also incorporate chemical reactivity into simulations by facilitating the use of the REACTER protocol. Additionally, LUNAR provides postprocessing for free volume calculations, cure characterization calculations, and property predictions from LAMMPS thermodynamic outputs. LUNAR has been validated via building and simulation of pure epoxy and cyanate ester polymer systems with a comparison of the corresponding predicted structures and properties to benchmark values, including experimental results from the literature. LUNAR provides the tools for the computationally driven development of next-generation composite materials in the Integrated Computational Materials Engineering (ICME) and Materials Genome Initiative (MGI) frameworks. LUNAR is written in Python with the usage of NumPy and can be used via a graphical user interface, a command line interface, or an integrated design environment. LUNAR is freely available via GitHub.

Community-Scale Problem-Solving: Reflections on a Decade of Infrastructure Development in the MGI

This paper describes the need for a unified representation of Materials Genome Initiative (MGI) problem-solving processes. An approach for solving this problem in the form of pattern-languages is proposed as well as a means for realizing them through existing infrastructure and community development activities. As part of this strategy, key lifecycle questions related to the importance of community-based modeling and sustainability are considered. Central to the entire paper are two observations: a) that MGI itself, as well as its domains, communities, and infrastructure, are complex systems; and b) that scalability of these complex systems is maximized through robust modeling. These ideas are discussed as a strategy toward community-centered growth, unification, and sustainability in MGI ecosystem-scale problem-solving and infrastructure development.

A machine learning‐based crystal graph network and its application in development of functional materials

AbstractAn active area of MGI (Materials Genome Initiative)/MGE (Materials Genome Engineering) is to accelerate the development of new materials by means of active learning and “digital trial‐error” using a prediction model of material property. Machine learning methods have widely been employed for predicting crystalline materials properties with crystal graph neural networks (CGNN). The prediction accuracy of the state‐of‐the‐art (SOTA) CGNN models based on big models and big data is generally higher. However, for the development of some classes of materials, the datasets obtained by experiments are usually lacking due to costly experiments and measurement costs. The lack of datasets will impact the accuracy of CGNN models and may result in overfitting during training models. This paper proposes a simplified crystal graph convolutional neural network (S‐CGCNN) which possesses higher prediction accuracy while reducing the vast amount of train datasets and computation costs. The S‐CGCNN model has successfully predicted properties of crystalline materials, such as piezoelectric materials and dielectric materials, and increased the prediction accuracy up to 12%–20% than existing SOTA CGNN models. Furthermore, the distribution map between properties and compositions of materials has been built to screen the latent space of candidate materials efficiently by principal component analysis.

Simulation‐Guided Design of Gradient Multilayer Microwave Absorber with Tailored Absorption Performance

AbstractFlexible microwave absorber (MAR), vital in advanced applications such as wearable electronics and precision devices, are highly valued for their lightweight, exceptional electromagnetic waves (EWs), and ease of fabrication. However, optimizing the electromagnetic parameters of microwave absorption materials (MAMs) to enhance absorption ability and expand effective absorption broadband (EAB, reflection loss (RL) <−10 dB) is a considerable challenge. Herein, a permittivity‐attenuation evaluation diagram (PAED) is constructed using parameter scanning based on the Materials Genome Initiative to determine the ideal electromagnetic parameters and thickness, optimize absorption efficiency, and obtain highly efficient absorbers. Guided by the PAED, a multilayer MAR consisting of a “matching‐absorption‐reflection layer” and a dielectric loss gradient aligned with the direction of EWs propagation is developed. This design significantly enhances the EWs penetration and ensures effective absorption, attributed to the well‐matched impedance and attenuation characteristics. As anticipated, the microwave absorption of the absorber (density = 0.063 g cm−3) is optimized, with an RL of −34 dB at d = 4 mm and an EAB covering the entire X‐band (8.2–12.4 GHz). This study presents a novel approach for establishing a material database for MAMs and developing high‐performance absorbers characterized by thinness, lightness, broad operational frequency range, and robust absorption capacity.

State of the Art Review for Titanium Fluorine Glasses and Glass Ceramics.

Titanium (Ti) and fluorine (F) have the potential to provide a variety of desirable physical, chemical, mechanical, and biological properties applicable to a broad range of indications. Consequently, Ti- and F-containing glasses and glass ceramics are currently under investigation for use in nuclear, optical, electrochemical, dental, and industrial fields. Accordingly, significant interest exists with respect to understanding the individual and interaction effects that these elements have on material structure and properties to support the accelerated design, development, and deployment of these materials. This review aims to serve as a foundational reference across multiple disciplines, highlighting the fundamental properties and versatility of Ti- and F-containing glasses and glass ceramics. By consolidating our current knowledge of these materials, this broad overview will identify areas in which we can further our understanding to support the a priori prediction and effective design of these systems. Finally, this paper will introduce the potential to improve material design by integrating experimentation, modelling, and computational approaches in a manner commensurate with the principles of the Materials Genome Initiative.

Cutting Edge High‐Throughput Synthesis and Characterization Techniques in Combinatorial Materials Science

AbstractThe Materials Genome Initiative is expected to accelerate the materials discovery and design by fundamentally changing the trial‐and‐error research paradigm. However, mass data from high‐throughput experiments is still essential for the revelation of rules and verification of theories. In fact, the development of combinatorial materials science is always on the strength of the upgrade and evolution of high‐throughput techniques in each stage, especially high‐throughput materials synthesis and characterization. Herein, this review summarizes the high‐throughput synthesis methods for combinatorial materials libraries, especially the co‐deposition and masking techniques of thin‐film fabrication; and details the high‐throughput characterization methods for specific material properties and typical material categories, which comes down to the spectroscopy and microscopy techniques. It is considered that high‐throughput concepts will be the predominant lab experimentation in the future, along with advanced experimental techniques and convenient data processing procedures. Before that, more cooperation between multiple researchers from different fields should be conducted to complete the combinatorial materials research, since the high‐throughput technology covers multiple disciplines with a huge span.

The MatHub‐3d first‐principles repository and the applications on thermoelectrics

AbstractFollowing the Materials Genome Initiative project, materials research has embarked a new research paradigm centered around material repositories, significantly accelerating the discovery of novel materials, such as thermoelectrics. Thermoelectric materials, capable of directly converting heat into electricity, are garnering increasing attention in applications like waste heat recovery and refrigeration. To facilitate research in this emerging paradigm, we have established the Materials Hub with Three‐Dimensional Structures (MatHub‐3d) repository, which serves as the foundation for high‐throughput (HTP) calculations, property analysis, and the design of thermoelectric materials. In this review, we summarize recent advancements in thermoelectric materials powered by the MatHub‐3d, specifically HTP calculations of transport properties and material design on key factors. For HTP calculations, we develop the electrical transport package for HTP purpose, and utilize it for materials screening. In some works, we investigate the relationship between transport properties and chemical bonds for particular types of thermoelectric compounds based on HTP results, enhancing the fundamental understanding about interested compounds. In our work associated with material design, we primarily utilize key factors beyond transport properties to further expedite materials screening and speedily identify specific materials for further theoretical/experimental analyses. Finally, we discuss the future developments of the MatHub‐3d and the evolving directions of database‐driven thermoelectric research.

Scientists’ disciplinary characteristics and collaboration behaviour under the convergence paradigm: A multilevel network perspective

The convergence paradigm underlines the importance of integrating multiple disciplines through collaboration. However, the crucial question of how scientists' disciplinary characteristics influence scientific collaboration remains unresolved. Using an exponential random graph model for multilevel networks, this study provides insights into the impact of scientists' disciplinary characteristics on their collaborative behaviour based on data from the Materials Genome Initiative, a typical convergence field. These results show that: under the convergence paradigm, scientists with a greater number of affiliated disciplines or with greater disparities in knowledge systems among their affiliated disciplines are less active in collaboration, whereas scientists with more balanced competence across their affiliated disciplines are more active. Scientists are more likely to collaborate with people who have a similar ability to integrate multidisciplinary knowledge. Scientists with a focus on applied disciplines are more likely to collaborate than are those with a preference for basic disciplines. Scientists who focus more on peripheral and external disciplines are more active in collaboration than scientists who focus on core and internal disciplines. Scientists collaborate based on shared disciplines and utilise the unique disciplines of their collaborators to advance knowledge and thus expand their own research space. This study provides evidence for the selection of partners based on scientists' disciplinary characteristics and highlights its importance for interdisciplinary teams and project management.

The Case for a Defect Genome Initiative.

The Materials Genome Initiative (MGI) has streamlined the materials discovery effort by leveraging generic traits of materials, with focus largely on perfect solids. Defects such as impurities and perturbations, however, drive many attractive functional properties of materials. The rich tapestry of charge, spin, and bonding states hosted by defects are not accessible to elements and perfect crystals, and defects can thus be viewed as another class of "elements" that lie beyond the periodic table. Accordingly, a Defect Genome Initiative (DGI) to accelerate functional defect discovery for energy, quantum information, and other applications is proposed. First, major advances made under the MGI are highlighted, followed by a delineation of pathways for accelerating the discovery and design of functional defects under the DGI. Near-term goals for the DGI are suggested. The construction of open defect platforms and design of data-driven functional defects, along with approaches for fabrication and characterization of defects, are discussed. The associated challenges and opportunities are considered and recent advances towards controlled introduction of functional defects at the atomic scale are reviewed. It is hoped this perspective will spur a community-wide interest in undertaking a DGI effort in recognition of the importance of defects in enabling unique functionalities in materials.

Ab initio artificial intelligence: Future research of Materials Genome Initiative

AbstractThe marriage of artificial intelligence (AI) and Materials Genome Initiative (MGI) could profoundly change the landscape of modern materials research, leading to a new paradigm of data‐driven and AI‐driven materials discovery. In this perspective, we will give an overview on the central role of AI in the MGI research. In particular, an emerging research field of ab initio AI, which applies state‐of‐the‐art AI techniques to help solve bottleneck problems of ab initio computation, will be introduced. The development of ab initio AI will greatly accelerate high‐throughput computation, promote the construction of large materials database, and open new opportunities for future research of MGI.

OCPMDM 2.0: An intelligent solution for materials data mining

Machine learning methods have played a significant role in materials design. With the rapid development of the Materials Genome Initiative (MGI) and data science, material researchers are confronted with the requirements to conduct sophisticated data analytics in modeling the property of materials. To make it more convenient for material researchers to design new materials using machine learning methods, an intelligent web platform called the second version of online computation platform for material data mining (OCPMDM 2.0) has been updated from the previous computation platform in our lab. Besides the various data mining algorithms developed in OCPMDM 1.0, the new platform tries to provide an intelligent solution for materials datamining, including descriptors filling, virtual screening of candidate materials, connection with materials databases and model sharing. It is convenient for materials researchers to obtain the machine learning model and the result of applying model by only submitting materials data on OCPMDM 2.0. To demonstrate the applications of the platform, two data sets of different kinds of materials were automatically processed to obtain the best models in the platform. The models are applied to screen out candidates with better properties than those in the training dataset. Material data mining process can be implemented via the platform, which provides convenient ways for material researchers in materials design and optimization. The URL of the platform is http://materials-data-mining.com/ocpmdm/.

Genomic materials design: CALculation of PHAse Dynamics

The CALPHAD system of fundamental phase-level databases, now known as the Materials Genome, has enabled a mature technology of computational materials design and qualification that has already met the acceleration goals of the national Materials Genome Initiative. As first commercialized by QuesTek Innovations, the methodology combines efficient genomic-level parametric design of new material composition and process specifications with multidisciplinary simulation-based forecasting of manufacturing variation, integrating efficient uncertainty management. Recent projects demonstrated under the multi-institutional CHiMaD Design Center notably include novel alloys designed specifically for additive manufacturing. With the proven success of the CALPHAD-based Materials Genome technology, current university research emphasizes new methodologies for affordable accelerated expansion of more accurate CALPHAD databases. Rapid adoption of these new capabilities by US apex corporations has compressed the materials design and development cycle to under 2 years, enabling a new “materials concurrency” integrated into a new level of concurrent engineering supporting an unprecedented level of manufacturing innovation.

Implement the Materials Genome Initiative: Machine Learning Assisted Fluorescent Probe Design for Cellular Substructure Staining

AbstractThe Materials Genome Initiative (MGI) is accelerating the pace of advanced materials development by integrating high‐throughput experimentation, database construction, and intelligence computation. Live‐cell imaging agents, such as fluorescent dyes, are exemplary candidates for MGI applications for two reasons: i) they are essential in visualizing cellular structures and functional processes, and ii) the unclear relationship between the chemical structure of fluorescent dyes and their live‐cell imaging properties severely restricts the current trial‐and‐error dye development. Herein, the MGI is followed to present an intelligent combinatorial methodology for predicting the staining cell ability of dyes utilizing machine learning (ML) driven by a structurally diverse combinatorial library. This study demonstrates how to high‐throughput synthesize 1,536 dyes and evaluate their imaging properties to establish a feature dataset for ML. A set of high‐precision ML‐predictors is then successfully modeled for assisting live‐cell staining and endoplasmic reticulum judgment. This approach is believed to bridge the gap between dye structure and corresponding staining behavior, and can accelerate the discovery of novel organelle‐specific stains.

Realizing the potentials of density functional theory (DFT) and of the materials genome initiative (MGI)

From 1964 and 1965 to present, the wide spread utilization of an incomplete density functional theory (DFT) has led to mixed results: The second theorem of the theory asserts that the energy functional reaches its minimum if the calculation employs the ground state charge density—without providing a mechanism for finding this density. Calculations purporting to employ DFT have mostly assumed that results obtained with a judiciously selected basis set, following self-consistent iterations, are those of the ground state. The state obtained with a single basis set is a stationary one, among an infinite number of such states, with no proven relation to the actual ground state of the material. Most failures or limitations of the incomplete DFT can be traced to this error. We present results from calculations using the completed DFT. They are in excellent agreement with experiment and portend the realization of the Materials Genome Initiative.

High-Throughput Scanning Second Harmonic Generation Microscopy for Polar Materials.

The Materials Genome Initiative aims to discover, develop, manufacture, and deploy advanced materials at twice the speed of conventional approaches. To achieve this, high-throughput characterization is essential for the rapid screening of candidate materials. In this study, we developed a high-throughput scanning second harmonic generation (HTS-SHG) microscope with automatic partitioning, accurate positioning, and fast scanning that can rapidly probe and screen polar materials. Using this technique, we first investigated typical ferroelectrics, including periodically poled lithium niobate crystals and PbZr0.2 Ti0.8 O3 (PZT) thin films, whereby the microscopic domain structures were clearly revealed. We then applied this technique to a compositional-gradient (100-x)%BaTiO3 -x%SrTiO3 film and a thickness-gradient PZT film to demonstrate its high-throughput capabilities. Since SHG signal is correlated with the macroscopic remnant polarization over the probed region determined by the laser spot, it is free of artifacts arising from leakage current and electrostatic interference, while materials symmetries and domain structures must be carefully considered in the data analysis. We believe this work can help promote the high-throughput development of polar materials, and contribute to Materials Genome Initiative. This article is protected by copyright. All rights reserved.

CALPHAD and the materials genome A 10 year anniversary

The materials genome initiative was announced by US President Obama 2011. Its actual meaning was initially rather unclear and it seemed as it could contain everything. Nevertheless, in their Scripta paper 2014 Kaufman and Ågren argued that the materials genome, analogously with biological genomes, should be defined as a set of information (databases) allowing prediction of materials structure as well as its response to processing and usage conditions. It thus seemed natural that CALPHAD thermodynamics and kinetics should play an essential role.In the present paper the role of CALPHAD as a way of processing different kind of materials data will be emphasized. The extension from thermochemistry to other properties, industrial applications and extrapolation far from equilibrium will be reviewed.

Data Augmentation of Micrographs and Prediction of Impact Toughness for Cast Austenitic Steel by Machine Learning

The Material Genome Initiative has been driven by high-throughput calculations, experiments, characterizations, and machine learning, which has accelerated the efficiency of the discovery of novel materials. However, the precise quantification of the material microstructure features and the construction of microstructure–property models are still challenging in optimizing the performance of materials. In this study, we proposed a new model based on machine learning to enhance the power of the data augmentation of the micrographs and construct a microstructure–property linkage for cast austenitic steels. The developed model consists of two modules: the data augmentation module and microstructure–property linkage module. The data augmentation module used a multi-layer convolution neural network architecture with diverse size filter to extract the microstructure features from irregular micrographs and generate new augmented microstructure images. The microstructure–property linkage module used a modified VGG model to establish the relationship between the microstructure and material property. Taking cast austenitic stainless steels after solution treating in different temperatures as an example, the results showed that the prediction accuracy of the developed machine learning model had been improved. The coefficient R2 of the model was 0.965, and the medians were only ±2 J different with the measured impact toughness.