Developing an improved crystal graph convolutional neural network framework for accelerated materials discovery

The recently proposed crystal graph convolutional neural network (CGCNN) offers a highly versatile and accurate machine learning (ML) framework by learning material properties directly from graph-like representations of crystal structures ("crystal graphs"). Here, we develop an improved variant of the CGCNN model (iCGCNN) that outperforms the original by incorporating information of the Voronoi tessellated crystal structure, explicit 3-body correlations of neighboring constituent atoms, and an optimized chemical representation of interatomic bonds in the crystal graphs. We demonstrate the accuracy of the improved framework in two distinct illustrations: First, when trained/validated on 180,000/20,000 density functional theory (DFT) calculated thermodynamic stability entries taken from the Open Quantum Materials Database (OQMD) and evaluated on a separate test set of 230,000 entries, iCGCNN achieves a predictive accuracy that is significantly improved, i.e., 20% higher than that of the original CGCNN. Second, when used to assist high-throughput search for materials in the ThCr2Si2 structure-type, iCGCNN exhibited a success rate of 31% which is 310 times higher than an undirected high-throughput search and 2.4 times higher than that of the original CGCNN. Using both CGCNN and iCGCNN, we screened 132,600 compounds with elemental decorations of the ThCr2Si2 prototype crystal structure and identified a total of 97 new unique stable compounds by performing 757 DFT calculations, accelerating the computational time of the high-throughput search by a factor of 130. Our results suggest that the iCGCNN can be used to accelerate high-throughput discoveries of new materials by quickly and accurately identifying crystalline compounds with properties of interest.

[1]  Chi Chen,et al.  Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals , 2018, Chemistry of Materials.

[2]  Ankit Agrawal,et al.  Machine-learning-accelerated high-throughput materials screening: Discovery of novel quaternary Heusler compounds , 2018, Physical Review Materials.

[3]  Wei-keng Liao,et al.  ElemNet: Deep Learning the Chemistry of Materials From Only Elemental Composition , 2018, Scientific Reports.

[4]  Tian Xie,et al.  Hierarchical Visualization of Materials Space with Graph Convolutional Neural Networks , 2018, The Journal of chemical physics.

[5]  K-R Müller,et al.  SchNet - A deep learning architecture for molecules and materials. , 2017, The Journal of chemical physics.

[6]  Jeffrey C Grossman,et al.  Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties. , 2017, Physical review letters.

[7]  Alok Choudhary,et al.  Including crystal structure attributes in machine learning models of formation energies via Voronoi tessellations , 2017 .

[8]  Miguel A. L. Marques,et al.  Predicting the Thermodynamic Stability of Solids Combining Density Functional Theory and Machine Learning , 2017 .

[9]  Bryce Meredig,et al.  Robust FCC solute diffusion predictions from ab-initio machine learning methods , 2017, 1705.08798.

[10]  Samuel S. Schoenholz,et al.  Neural Message Passing for Quantum Chemistry , 2017, ICML.

[11]  Atsuto Seko,et al.  Representation of compounds for machine-learning prediction of physical properties , 2016, 1611.08645.

[12]  Alexandre Tkatchenko,et al.  Quantum-chemical insights from deep tensor neural networks , 2016, Nature Communications.

[13]  Cormac Toher,et al.  Universal fragment descriptors for predicting properties of inorganic crystals , 2016, Nature Communications.

[14]  Taylor D. Sparks,et al.  High-Throughput Machine-Learning-Driven Synthesis of Full-Heusler Compounds , 2016 .

[15]  Wei Chen,et al.  A Statistical Learning Framework for Materials Science: Application to Elastic Moduli of k-nary Inorganic Polycrystalline Compounds , 2016, Scientific Reports.

[16]  Ankit Agrawal,et al.  Predictive analytics for crystalline materials: bulk modulus , 2016 .

[17]  Olexandr Isayev,et al.  Material informatics driven design and experimental validation of lead titanate as an aqueous solar photocathode , 2016 .

[18]  Logan T. Ward,et al.  A General-Purpose Machine Learning Framework for Predicting Properties of Inorganic Materials , 2016, 1606.09551.

[19]  Chiho Kim,et al.  Finding New Perovskite Halides via Machine Learning , 2016, Front. Mater..

[20]  Vijay S. Pande,et al.  Molecular graph convolutions: moving beyond fingerprints , 2016, Journal of Computer-Aided Molecular Design.

[21]  Ryan O'Hayre,et al.  Predicting density functional theory total energies and enthalpies of formation of metal-nonmetal compounds by linear regression , 2016 .

[22]  G. Pilania,et al.  Machine learning bandgaps of double perovskites , 2016, Scientific Reports.

[23]  Engineering,et al.  Prediction model of band gap for inorganic compounds by combination of density functional theory calculations and machine learning techniques , 2015, 1509.00973.

[24]  Felix A Faber,et al.  Machine Learning Energies of 2 Million Elpasolite (ABC_{2}D_{6}) Crystals. , 2015, Physical review letters.

[25]  Muratahan Aykol,et al.  The Open Quantum Materials Database (OQMD): assessing the accuracy of DFT formation energies , 2015 .

[26]  Edward O. Pyzer-Knapp,et al.  A Bayesian Approach to Calibrating High-Throughput Virtual Screening Results and Application to Organic Photovoltaic Materials , 2015, 1510.00388.

[27]  Alán Aspuru-Guzik,et al.  Convolutional Networks on Graphs for Learning Molecular Fingerprints , 2015, NIPS.

[28]  Atsuto Seko,et al.  Prediction of Low-Thermal-Conductivity Compounds with First-Principles Anharmonic Lattice-Dynamics Calculations and Bayesian Optimization. , 2015, Physical review letters.

[29]  James E. Gubernatis,et al.  Structure classification and melting temperature prediction in octet AB solids via machine learning , 2015 .

[30]  W. Steurer,et al.  More statistics on intermetallic compounds - ternary phases. , 2015, Acta crystallographica. Section A, Foundations and advances.

[31]  Felix A Faber,et al.  Crystal structure representations for machine learning models of formation energies , 2015, 1503.07406.

[32]  Krishna Rajan,et al.  Mining for elastic constants of intermetallics from the charge density landscape , 2015 .

[33]  J. Vybíral,et al.  Big data of materials science: critical role of the descriptor. , 2014, Physical review letters.

[34]  Alok Choudhary,et al.  Combinatorial screening for new materials in unconstrained composition space with machine learning , 2014 .

[35]  Somnath Datta,et al.  Informatics-aided bandgap engineering for solar materials , 2014 .

[36]  Atsuto Seko,et al.  Machine learning with systematic density-functional theory calculations: Application to melting temperatures of single- and binary-component solids , 2013, 1310.1546.

[37]  Muratahan Aykol,et al.  Materials Design and Discovery with High-Throughput Density Functional Theory: The Open Quantum Materials Database (OQMD) , 2013 .

[38]  Kristin A. Persson,et al.  Commentary: The Materials Project: A materials genome approach to accelerating materials innovation , 2013 .

[39]  Marco Buongiorno Nardelli,et al.  The high-throughput highway to computational materials design. , 2013, Nature materials.

[40]  Anubhav Jain,et al.  Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysis , 2012 .

[41]  Marco Buongiorno Nardelli,et al.  AFLOWLIB.ORG: A distributed materials properties repository from high-throughput ab initio calculations , 2012 .

[42]  Alán Aspuru-Guzik,et al.  The Harvard Clean Energy Project: Large-Scale Computational Screening and Design of Organic Photovoltaics on the World Community Grid , 2011 .

[43]  W. B. Pearson,et al.  Pearson's crystal data : crystal structure database for inorganic compounds , 2007 .

[44]  Kristin A. Persson,et al.  Predicting crystal structures with data mining of quantum calculations. , 2003, Physical review letters.

[45]  P. Luksch,et al.  New developments in the Inorganic Crystal Structure Database (ICSD): accessibility in support of materials research and design. , 2002, Acta crystallographica. Section B, Structural science.

[46]  G. Kresse,et al.  From ultrasoft pseudopotentials to the projector augmented-wave method , 1999 .

[47]  Hafner,et al.  Ab initio molecular dynamics for liquid metals. , 1995, Physical review. B, Condensed matter.

[48]  C. Zheng,et al.  Making and breaking bonds in the solid state: the ThCr2Si2 structure , 1985 .

[49]  I. D. Brown,et al.  The inorganic crystal structure data base , 1983, J. Chem. Inf. Comput. Sci..