Historically, materials discovery has been driven by a laborious trial-and-error process. The growth of materials databases and emerging informatics approaches finally offer the opportunity to transform this practice into data- and knowledge-driven rational design. By using data from the AFLOW repository for high-throughput ab-initio calculations, we have generated Quantitative Materials Structure-Property Relationship (QMSPR) models to predict eight critical electronic and thermomechanical materials properties, such as the metal/insulator classification, band gap energy, bulk and shear moduli, Debye temperature, and heat capacity. The prediction accuracy obtained with these QMSPR models approaches training data for virtually any stoichiometric inorganic crystalline material. The success and universality of these models is attributed to the construction of new materials descriptors---referred to as the universal Property-Labeled Materials Fragments (PLMF). The representation requires only minimal structural input and affords straightforward model interpretation in terms of simple heuristic design rules that guide rational materials design. This study demonstrates the power of materials informatics to dramatically accelerate the search for new materials.
The Gibbs energy, G, determines the equilibrium conditions of chemical reactions and materials stability. Despite this fundamental and ubiquitous role, G has been tabulated for only a small fraction of known inorganic compounds, impeding a comprehensive perspective on the effects of temperature and composition on materials stability and synthesizability. Here, we use the SISSO (sure independence screening and sparsifying operator) approach to identify a simple and accurate descriptor to predict G for stoichiometric inorganic compounds with ~50 meV/atom (~1 kcal/mol) resolution, and with minima…
Solid-state electrolytes (SSE) with high ion conductivity are pivotal for the development and large-scale adoption of green-energy conversion and storage technologies such as fuel cells, electrocatalysts and solid-state batteries. Yet, SSE are extremely complex materials for which general rational design principles remain indeterminate. Here, we unite first-principles materials modelling, computational power and modern data analysis techniques to advance towards the solution of such a fundamental and technologically pressing problem. Our data-driven survey reveals that the correlations between…
The unprecedented structural flexibility and diversity of inorganic frameworks of layered hybrid halide perovskites (LHHPs) rise up a wide range of useful optoelectronic properties thus predetermining the extraordinary high interest to this family of materials. Nevertheless, the influence of different types of distortions of their inorganic framework on key physical properties such as band gap has not yet been quantitatively identified. We provided a systematic study of the relationships between LHHPs' band gaps and six main structural descriptors of inorganic framework, including interlayer d…
Structures and properties of many inorganic compounds have been collected historically. However, it only covers a very small portion of possible inorganic crystals, which implies the presence of numerous currently unknown compounds. A powerful machine-learning strategy is mandatory to discover new inorganic compounds from all chemical combinations. Herein we propose a descriptor-based recommender-system approach to estimate the relevance of chemical compositions where stable crystals can be formed [i.e., chemically relevant compositions (CRCs)]. As well as data-driven compositional similarity …
Hybrid organic-inorganic perovskites are exceedingly interesting candidates for new solar energy technologies, for both ground-based and space applications. However, their large-scale production is hampered by the lack of long-term stability, mostly associated to ion migration. The specific role of non-covalent bonds in contributing to the stability remains elusive, and in certain cases controversial. Here, we perform an investigation on a large perovskite chemical space via a combination of first-principles calculations for the bond strengths and the recently developed Sure Independent Screen…
The expansion of programmatically-accessible materials data has cultivated opportunities for data-driven approaches. Highly-automated frameworks like AFLOW not only manage the generation, storage, and dissemination of materials data, but also leverage the information for thermodynamic formability modeling, such as the prediction of phase diagrams and properties of disordered materials. In combination with standardized parameter sets, the wealth of data is ideal for training machine learning algorithms, which have already been employed for property prediction, descriptor development, design rul…
One-dimensional inorganic nanotubes hold promise for technological applications due to their distinct physical/chemical properties, but so far advancements have been hampered by difficulties in producing single-wall nanotubes with a well-defined radius. In this work we investigate, based on Density Functional Theory (DFT), the formation mechanism of 135 different inorganic nanotubes formed by the intrinsic self-rolling driving force found in asymmetric 2D Janus sheets. We show that for isovalent Janus sheets, the lattice mismatch between inner and outer atomic layers is the driving force behin…
Harnessing the recent advance in data science and materials science, it is feasible today to build predictive models for materials properties. In this study, we employ the data of high-throughput quantum mechanics calculations based on 170,714 inorganic crystalline compounds to train a machine learning model for formation energy prediction. Different from the previous work, our model reaches a fairly good predictive ability (R2=0.982 and MAE=0.07 eVatom-1, DenseNet model) and meanwhile can be universally applied to the large phase space of inorganic materials. The improvement comes from severa…
Generative models based on generative adversarial networks (GANs) and variational autoencoders (VAEs) have been widely studied in the fields of image generation, speech generation, and drug discovery, but, only a few studies have focused on the generation of inorganic materials. Such studies use the crystal structures of materials, but material researchers rarely store this information. Thus, we generate chemical compositions without using crystal information. We use a conditional VAE (CondVAE) and a conditional GAN (CondGAN) and show that CondGAN using the bag-of-atom representation with phys…
Genome-Guided Interpretable Screening of Phase-Stable, Lead-Free Double Perovskite Absorbers for All-Inorganic Semiconductors, Sensors, and Photovoltaics with DFT-Validated Design Rules
Research and Analysis Institute, RAI Initiative Ltd（拉伊倡议有限公司研究与分析研究所）；Bangladesh University of Textiles（孟加拉国纺织大学）；Universiti Putra Malaysia（马来西亚博特拉大学）；Institute of Science Tokyo（东京科学大学）；University of Évora（埃武拉大学）；Daffodil International University（达富迪尔国际大学）；University of Idaho（爱达荷大学）；North Carolina A&T State University（北卡罗来纳州立农业技术大学）
NEP89: Universal neuroevolution potential for inorganic and organic materials across 89 elements
Bohai University（渤海大学）；The Chinese University of Hong Kong（香港中文大学）；Chalmers University of Technology（查尔姆斯理工大学）；Shenzhen Technology University（深圳职业技术大学）；Shenzhen University（深圳大学）；Xinyu University（新余学院）；University of Science and Technology Beijing（北京科技大学）
Accelerating Discovery of Extreme Lattice Thermal Conductivity by Crystal Attention Graph Neural Network (CATGNN) Using Chemical Bonding Intuitive Descriptors
The high-throughput screening of periodic inorganic solids using machine learning methods requires atomic positions to encode structural and compositional details into appropriate material descriptors. These atomic positions are not available {\it a priori} for new materials which severely limits exploration of novel materials. We overcome this limitation by using only crystallographic symmetry information in the structural description of materials. We show that for materials with identical structural symmetry, machine learning is trivial and accuracies similar to that of density functional th…
Finetuning-Free Diffusion Model with Adaptive Constraint Guidance for Inorganic Crystal Structure Generation
CNRS-Saint-Gobain-NIMS（法国国家科学研究中心-圣戈班-日本纳米科学研究所）；Laboratory for Innovative Key Materials and Structures (LINK)（创新关键材料与结构实验室）；Laboratory of Computational, Quantitative, and Synthetic Biology (CQSB)（计算、定量与合成生物学实验室）；Data-driven Materials Design Group（数据驱动材料设计组）；Center for Basic Research on Materials（材料基础研究中心）；LTCI, Télécom Paris, Institut Polytechnique de Paris（LTCI，巴黎电信，巴黎理工学院）
Traditional materials discovery approaches - relying primarily on laborious experiments - have controlled the pace of technology. Instead, computational approaches offer an accelerated path: high-throughput exploration and characterization of virtual structures. These ventures, performed by automated ab-initio frameworks, have rapidly expanded the volume of programmatically-accessible data, cultivating opportunities for data-driven approaches. Herein, a collection of robust characterization methods are presented, implemented within the Automatic Flow Framework for Materials Discovery (AFLOW), …
Dynamics of protein self-assembly on the inorganic surface and the resultant geometric patterns are visualized using high-speed atomic force microscopy. The time dynamics of the classical macroscopic descriptors such as 2D Fast Fourier Transforms (FFT), correlation and pair distribution function are explored using the unsupervised linear unmixing, demonstrating the presence of static ordered and dynamic disordered phases and establishing their time dynamics. The deep learning (DL)-based workflow is developed to analyze detailed particle dynamics on the particle-by-particle level. Beyond the ma…
Understanding material composition-structure-function relationships is of critical importance for the design and discovery of novel functional materials. While most such studies focus on individual materials, we conducted a global mapping study of all known materials deposited in the Material Project database to investigate their distributions in the space of a set of seven compositional, structural, physical, and neural latent descriptors. These two-dimensional materials maps along with their density maps allow us to illustrate the distribution of the patterns and clusters of different shapes…
We present a benchmark test suite and an automated machine learning procedure for evaluating supervised machine learning (ML) models for predicting properties of inorganic bulk materials. The test suite, Matbench, is a set of 13 ML tasks that range in size from 312 to 132k samples and contain data from 10 density functional theory-derived and experimental sources. Tasks include predicting optical, thermal, electronic, thermodynamic, tensile, and elastic properties given a materials composition and/or crystal structure. The reference algorithm, Automatminer, is a highly-extensible, fully-automa…
Photovoltaic cells (PVc), as an energy provider to the next generation and the biggest source of renewable energy. Since the last decade improving efficiency and reducing the cost of PVc has been a subject of active research among scientists. Promising progress in the field of material science and manufacturing process at Nano-level played a big role. Still, at present there are many challenges before photovoltaics for efficient and economic energy. However, Photovoltaics cell based on p-n type homojunction semiconductors with different organic and inorganic materials reported thus for general…
Prediction of Large Magnetic Moment Materials With Graph Neural Networks and Random Forests
Mila - Quebec Artificial Intelligence Institute（米拉-魁北克人工智能研究所）；IVADO - Institut de Valorisation des Données（IVADO-数据价值化研究所）；Université de Sherbrooke（舍布鲁克大学）；Institut Quantique（量子研究所）
The large-scale search for high-performing candidate 2D materials is limited to calculating a few simple descriptors, usually with first-principles density functional theory calculations. In this work, we alleviate this issue by extending and generalizing crystal graph convolutional neural networks to systems with planar periodicity, and train an ensemble of models to predict thermodynamic, mechanical, and electronic properties. To demonstrate the utility of this approach, we carry out a screening of nearly 45,000 structures for two largely disjoint applications: namely, mechanically robust co…
Indian Institute of Science（印度科学学院）；Shell Technology Centre Bangalore（壳牌班加罗尔技术中心）
Developing accurate, transferable and computationally inexpensive machine learning models can rapidly accelerate the discovery and development of new materials. Some of the major challenges involved in developing such models are, (i) limited availability of materials data as compared to other fields, (ii) lack of universal descriptor of materials to predict its various properties. The limited availability of materials data can be addressed through transfer learning, while the generic representation was recently addressed by Xie and Grossman [1], where they developed a crystal graph convolution…
