Discovering unusual structures from exception using big data and machine learning techniques.

Recently, machine learning (ML) has become a widely used technique in materials science study. Most work focuses on predicting the rule and overall trend by building a machine learning model. However, new insights are often learnt from exceptions against the overall trend. In this work, we demonstrate that how unusual structures are discovered from exceptions when machine learning is used to get the relationship between atomic and electronic structures based on big data from high-throughput calculation database. For example, after training an ML model for the relationship between atomic and electronic structures of crystals, we find AgO2F, an unusual structure with both Ag3+ and O22−, from structures whose band gap deviates much from the prediction made by our model. A further investigation on this structure might shed light into the research on anionic redox in transition metal oxides of Li-ion batteries.

[1]  F. Pan,et al.  Spectroscopic Signature of Oxidized Oxygen States in Peroxides. , 2018, The journal of physical chemistry letters.

[2]  Yue Liu,et al.  Materials discovery and design using machine learning , 2017 .

[3]  Weile Jia,et al.  Fast plane wave density functional theory molecular dynamics calculations on multi-GPU machines , 2013, J. Comput. Phys..

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

[5]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[6]  Michael I. Jordan,et al.  Machine learning: Trends, perspectives, and prospects , 2015, Science.

[7]  Paul Raccuglia,et al.  Machine-learning-assisted materials discovery using failed experiments , 2016, Nature.

[8]  R. Ramprasad,et al.  Machine Learning in Materials Science , 2016 .

[9]  James E. Gubernatis,et al.  Multi-fidelity machine learning models for accurate bandgap predictions of solids , 2017 .

[10]  Stefan Goedecker,et al.  Identification of Novel Cu, Ag, and Au Ternary Oxides from Global Structural Prediction , 2015, 1503.07327.

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

[12]  F. Illas,et al.  An Empirical, yet Practical Way To Predict the Band Gap in Solids by Using Density Functional Band Structure Calculations , 2017 .

[13]  G. Shao.  Melting of metallic and intermetallic solids: An energetic view from DFT calculated potential wells , 2008 .

[14]  K. Fujimura,et al.  Accelerated Materials Design of Lithium Superionic Conductors Based on First‐Principles Calculations and Machine Learning Algorithms , 2013 .

[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]  Beatriz Cordero,et al.  Covalent radii revisited. , 2008, Dalton transactions.

[17]  Francois Gygi,et al.  Optimization algorithm for the generation of ONCV pseudopotentials , 2015, Comput. Phys. Commun..

[18]  F. Pan,et al.  Mechanism of Exact Transition between Cationic and Anionic Redox Activities in Cathode Material Li2FeSiO4. , 2018, The journal of physical chemistry letters.

[19]  W. M. Haynes.  CRC Handbook of Chemistry and Physics , 1990 .

[20]  Siqi Shi,et al.  Multi-scale computation methods: Their applications in lithium-ion battery research and development , 2016 .

[21]  M. Rupp,et al.  Machine learning of molecular electronic properties in chemical compound space , 2013, 1305.7074.

[22]  Siqi Shi,et al.  The onset temperature (Tg) of AsxSe1−x glasses transition prediction: A comparison of topological and regression analysis methods , 2017 .

[23]  Lin Lin,et al.  Adaptively Compressed Exchange Operator. , 2016, Journal of chemical theory and computation.

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

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

[26]  Tie-Yan Liu,et al.  LightGBM: A Highly Efficient Gradient Boosting Decision Tree , 2017, NIPS.

[27]  O. A. von Lilienfeld,et al.  Electronic spectra from TDDFT and machine learning in chemical space. , 2015, The Journal of chemical physics.

[28]  Weile Jia,et al.  The analysis of a plane wave pseudopotential density functional theory code on a GPU machine , 2013, Comput. Phys. Commun..

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

[30]  D. Hamann.  Optimized norm-conserving Vanderbilt pseudopotentials , 2013, 1306.4707.

[31]  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.

[32]  F. D. Boer.  Cohesion in Metals: Transition Metal Alloys , 1989 .

[33]  Tim Mueller,et al.  Machine Learning in Materials Science , 2016 .