Journal of Materials Science

, Volume 53, Issue 9, pp 6652–6664 | Cite as

Machine learning properties of binary wurtzite superlattices



The burgeoning paradigm of high-throughput computations and materials informatics brings new opportunities in terms of targeted materials design and discovery. The discovery process can be significantly accelerated and streamlined if one can learn effectively from available knowledge and past data to predict materials properties efficiently. Indeed, a very active area in materials science research is to develop machine learning based methods that can deliver automated and cross-validated predictive models using either already available materials data or new data generated in a targeted manner. In the present contribution, we show that fast and accurate predictions of a wide range of properties of binary wurtzite superlattices, formed by a diverse set of chemistries, can be made by employing state-of-the-art statistical learning methods trained on quantum mechanical computations in combination with a judiciously chosen numerical representation to encode materials’ similarity. These surrogate learning models then allow for efficient screening of vast chemical spaces by providing instant predictions of the targeted properties. Moreover, the models can be systematically improved in an adaptive manner, incorporate properties computed at different levels of fidelities and are naturally amenable to inverse materials design strategies. While the learning approach to make predictions for a wide range of properties (including structural, elastic and electronic properties) is demonstrated here for a specific example set containing more than 1200 binary wurtzite superlattices, the adopted framework is equally applicable to other classes of materials as well.



GP and X-YL acknowledge support from the Los Alamos National Laboratory’s ongoing MaRIE project and helpful discussions with Zhehui Wang. Computational support for this work was provided by LANL’s high performance computing clusters. Los Alamos National Laboratory is operated by Los Alamos National Security, LLC, for the National Nuclear Security Administration of the (US) Department of Energy under Contract DE-AC52-06NA25396.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary material

10853_2018_1987_MOESM1_ESM.pdf (2.4 mb)
Supplementary material 1 (pdf 2471 KB)


  1. 1.
    Agrawal A, Choudhary A (2016) Materials informatics and big data: realization of the ‘fourth paradigm’ of science in materials science. APL Mater 4:53208CrossRefGoogle Scholar
  2. 2.
    Materials Genome Initiative.
  3. 3.
    Curtarolo S et al (2013) The high-throughput highway to computational materials design. Nat Mater 12:191CrossRefGoogle Scholar
  4. 4.
    Jain A, Hautier G, Ong SP, Persson K (2016) New opportunities for materials informatics: Resources and data mining techniques for uncovering hidden relationships. J Mater Res 31:977–994CrossRefGoogle Scholar
  5. 5.
    Rajan K (2005) Materials informatics. Mater Today 8:38CrossRefGoogle Scholar
  6. 6.
    LeSar R (2009) Materials informatics: an emerging technology for materials development. Stat Anal Data Min 1:372CrossRefGoogle Scholar
  7. 7.
    Ramprasad R, Batra R, Pilania G, Mannodi-Kanakkithodi A, Kim C (2017) Machine learning and materials informatics: recent applications and prospects. ArXiv Prepr arXiv:170707294
  8. 8.
    Montavon G et al (2013) Machine learning of molecular electronic properties in chemical compound space. New J Phys 15:95003CrossRefGoogle Scholar
  9. 9.
    Pilania G, Wang C, Jiang X, Rajasekaran S, Ramprasad R (2013) Accelerating materials property predictions using machine learning. Sci Rep 3:2810CrossRefGoogle Scholar
  10. 10.
    Meredig B et al (2014) Combinatorial screening for new materials in unconstrained composition space with machine learning. Phys Rev B 89:094104CrossRefGoogle Scholar
  11. 11.
    Deml AM, O’Hayre R, Wolverton C, Stevanovic V (2016) Predicting density functional theory total energies and enthalpies of formation of metal-nonmetal compounds by linear regression. Phys Rev B 93:085142CrossRefGoogle Scholar
  12. 12.
    Legrain F, Carrete J, van Roekeghem A, Curtarolo S, Mingo N (2017) How the chemical composition alone can predict vibrational free energies and entropies of solids. arXiv preprint arXiv:1703.02309
  13. 13.
    Medasani B et al (2016) Predicting defect behavior in B2 intermetallics by merging ab initio modeling and machine learning. NPJ Comput Mater 2:1CrossRefGoogle Scholar
  14. 14.
    Seko A, Maekawa T, Tsuda K, Tanaka I (2014) Machine learning with systematic density functional theory calculations: Application to melting temperatures of single- and binary component solids. Phys Rev B 89:054303CrossRefGoogle Scholar
  15. 15.
    Pilania G, Gubernatis JE, Lookman T (2015) Structure classification and melting temperature prediction in octet AB solids via machine learning. Phys Rev B 91:214302CrossRefGoogle Scholar
  16. 16.
    Aryal S, Sakidja R, Barsoum MW, Ching W-Y (2014) A genomic approach to the stability, elastic, and electronic properties of the MAX phases. Phys Status Solidi 251:1480–1497CrossRefGoogle Scholar
  17. 17.
    De Jong M et al (2016) A statistical learning framework for materials science: Application to elastic moduli of k-nary inorganic polycrystalline compounds. Sci Rep 6:34256CrossRefGoogle Scholar
  18. 18.
    Pilania G et al (2016) Machine learning bandgaps of double perovskites. Sci Rep 6:19375CrossRefGoogle Scholar
  19. 19.
    Lee J, Seko A, Shitara K, Nakayama K, Tanaka I (2016) Prediction model of band gap for inorganic compounds by combination of density functional theory calculations and machine learning techniques. Phys Rev B 93:115104CrossRefGoogle Scholar
  20. 20.
    Pilania G, Gubernatis JE, Lookman T (2017) Multi-fidelity machine learning models for accurate bandgap predictions of solids. Comput Mater Sci 129:156–163CrossRefGoogle Scholar
  21. 21.
    Weston L, Stampfl C (2017) Machine learning the band gap properties of kesterite \(\text{I}_2\)–II–IV–\(\text{V}_4\) quaternary compounds for photovoltaics applications. ArXiv Prepr arXiv:170808530
  22. 22.
    Seko A et al (2015) Prediction of low-thermal-conductivity compounds with first-principles anharmonic lattice-dynamics calculations and bayesian optimization. Phys Rev Lett 115:205901CrossRefGoogle Scholar
  23. 23.
    Kim C, Pilania G, Ramprasad R (2016) From organized high-throughput data to phenomenological theory using machine learning: the example of dielectric breakdown. Chem Mater 28:1304–1311CrossRefGoogle Scholar
  24. 24.
    Kim C, Pilania G, Ramprasad R (2016) Machine learning assisted predictions of intrinsic dielectric breakdown strength of \(\text{ABX}_3\) perovskites. J Phys Chem C 120:14575–14580CrossRefGoogle Scholar
  25. 25.
    Hong WT, Welsch RE, Shao-Horn Y (2016) Descriptors of Oxygen-Evolution activity for oxides: A statistical evaluation. J Phys Chem C 120:78–86CrossRefGoogle Scholar
  26. 26.
    Li Z, Ma X, Xin H (2017) Feature engineering of machine-learning chemisorption models for catalyst design. Catal Today 280:232–238CrossRefGoogle Scholar
  27. 27.
    Pilania G et al (2017) Using machine learning to identify factors that govern amorphization of irradiated pyrochlores. Chem Mater 29:2574–2583CrossRefGoogle Scholar
  28. 28.
    Xue D et al (2016) Accelerated search for materials with targeted properties by adaptive design. Nat Commun 7:11241CrossRefGoogle Scholar
  29. 29.
    Xue D et al (2016) Accelerated search for \(\text{ BaTiO }_3\)-based piezoelectrics with vertical morphotropic phase boundary using bayesian learning. Proc Natl Acad Sci 113:13301–13306CrossRefGoogle Scholar
  30. 30.
    Ashton M, Hennig RG, Broderick SR, Rajan K, Sinnott SB (2016) Computational discovery of stable \(\text{ M }_2\text{ AX }\) phases. Phys Rev B 94:20CrossRefGoogle Scholar
  31. 31.
    Pilania G, Balachandran PV, Kim C, Lookman T (2016) Finding new perovskite halides via machine learning. Front Mater 3:19CrossRefGoogle Scholar
  32. 32.
    Faber FA, Lindmaa A, von Lilienfeld OA, Armiento R (2016) Machine learning energies of 2 million elpasolite (\(\text{ABC}_2\text{D}_6\)) crystals. Phys Rev Lett 117:135502CrossRefGoogle Scholar
  33. 33.
    Fernandez M, Boyd PG, Da TD, Aghaji MZ, Woo TK (2014) Rapid and accurate machine learning recognition of high performing metal organic frameworks for \(\text{CO}_2\) capture. J Phys Chem Lett 5:3056–3060CrossRefGoogle Scholar
  34. 34.
    Emery AA, Saal JE, Kirklin S, Hegde VI, Wolverton C (2016) High-Throughput computational screening of perovskites for thermochemical water splitting applications. Chem Mater 28:5621–5634CrossRefGoogle Scholar
  35. 35.
    Pilania G, Mannodi-Kanakkithodi A (2017) A First-principles identification of novel double perovskites for water-splitting applications. J Mater Sci 52:8518–8525. CrossRefGoogle Scholar
  36. 36.
    Ghiringhelli LM, Vybiral J, Levchenko SV, Draxl C, Scheffler M (2015) Big data of materials science: critical role of the descriptor. Phys Rev Lett 114:105503CrossRefGoogle Scholar
  37. 37.
    Mueller T, Kusne AG, Ramprasad R (2016) Machine learning in materials science. In: Parril AL, Lipkowitz KB (eds) Reviews in computational chemistry. Wiley, Hoboken, pp 186–273Google Scholar
  38. 38.
    Bartok AP, Kondor R, Csanyi G (2013) On representing chemical environments. Phys Rev B 87:184115CrossRefGoogle Scholar
  39. 39.
    Behler J (2014) Representing potential energy surfaces by high-dimensional neural network potentials. J Phys Condens Matter 26:183001CrossRefGoogle Scholar
  40. 40.
    Li Z, Kermode JR, De Vita A (2015) Molecular dynamics with on-the-fly machine learning of quantum-mechanical forces. Phys Rev Lett 114:096405CrossRefGoogle Scholar
  41. 41.
    Botu V, Ramprasad R (2015) Learning scheme to predict atomic forces and accelerate materials simulations. Phys Rev B 92:094306CrossRefGoogle Scholar
  42. 42.
    Botu V, Ramprasad R (2015) Adaptive machine learning framework to accelerate ab initio molecular dynamics. Int J Quantum Chem 115:1074–1083CrossRefGoogle Scholar
  43. 43.
    Yu ET, Chow DH, McGill TC (1988) Commutativity of the GaAs/AlAs (100) band offset. Phys Rev B 38:12764CrossRefGoogle Scholar
  44. 44.
    Larsson MW, Wagner JB, Wallin M, Hakansson P, Fröberg LE, Samuelson L, Wallenberg LR (2006) Strain mapping in free-standing heterostructured wurtzite InAs/InP nanowires. Nanotechnology 18:015504CrossRefGoogle Scholar
  45. 45.
    Chang YM, Liou SC, Chen CH, Lee HM, Gwo S (2010) The electrostatic coupling of longitudinal optical phonon and plasmon in wurtzite InN thin films. Appl Phys Lett 96:041908CrossRefGoogle Scholar
  46. 46.
    Koguchi M, Kakibayashi H, Yazawa M, Hiruma K, Katsuyama T (1992) Crystal structure change of GaAs and InAs whiskers from zinc-blende to wurtzite type. Jpn J Appl Phys 31:2061CrossRefGoogle Scholar
  47. 47.
    Ma C, Moore D, Li J, Wang ZL (2003) Nanobelts, nanocombs, and nanowindmills of wurtzite ZnS. Adv Mater 15:228–231CrossRefGoogle Scholar
  48. 48.
    Kohn W (1999) Nobel Lecture: Electronic structure of matter–wave functions and density functionals. Rev Mod Phys 71:1253CrossRefGoogle Scholar
  49. 49.
    Martin R (2004) Electronic structure: basic theory and practical methods. Cambridge University Press, New YorkCrossRefGoogle Scholar
  50. 50.
    Kresse G, Furthmuller J (1996) Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys Rev B 54:11169–11186CrossRefGoogle Scholar
  51. 51.
    Materials Project.
  52. 52.
    Perdew JP, Wang Y (1992) Accurate and simple analytic representation of the electron-gas correlation energy. Phys Rev B 45:13244CrossRefGoogle Scholar
  53. 53.
    Blöchl P (1994) Projector augmented-wave method. Phys Rev B 50:17953–17979CrossRefGoogle Scholar
  54. 54.
    Monkhorst HJ, Pack JD (1976) Special points for Brillouin-zone integrations. Phys Rev B 13:5188CrossRefGoogle Scholar
  55. 55.
    Heyd J, Scuseria GE, Ernzerhof E (2006) Hybrid functionals based on a screened Coulomb potential. J Chem Phys 124:219906CrossRefGoogle Scholar
  56. 56.
    Born M, Huang K (1956) Dynamical theory of crystal lattices. Clarendon, OxfordGoogle Scholar
  57. 57.
    Huan TD, Mannodi-Kanakkithodi A, Ramprasad R (2015) Accelerated materials property predictions and design using motif-based fingerprints. Phys Rev B 92:014106CrossRefGoogle Scholar
  58. 58.
    Botu V, Batra R, Chapman J, Ramprasad R (2016) Machine learning force fields: construction, validation, and outlook. J Phys Chem C 121:511–522CrossRefGoogle Scholar
  59. 59.
    Jasrasaria D, Pyzer-Knapp EO, Rappoport D, Aspuru-Guzik A (2016) Space-filling curves as a novel crystal structure representation for machine learning models. arXiv preprint arXiv:1608.05747
  60. 60.
    Huang B, von Lilienfeld OA (2016) Communication: understanding molecular representations in machine learning: the role of uniqueness and target similarity. J Chem Phys 145:161102CrossRefGoogle Scholar
  61. 61.
    Seko A, Togo A,Tanaka I (2017) Descriptors for machine learning of materials data. arXiv preprint arXiv:1709.01666
  62. 62.
    Cubuk ED, Malone BD, Onat B, Waterland A, Kaxiras E (2017) Representations in neural network based empirical potentials. J Chem Phys 147:024104CrossRefGoogle Scholar
  63. 63.
    Artrith N, Urban A, Ceder G (2017) Efficient and accurate machine-learning interpolation of atomic energies in compositions with many species. Phys Rev B 96:014112CrossRefGoogle Scholar
  64. 64.
    Huo H, Rupp M (2017) Unified representation for machine learning of molecules and crystals. arXiv preprint arXiv:1704.06439
  65. 65.
    Mannodi-Kanakkithodi A, Pilania G, Huan TD, Lookman T, Ramprasad R (2016) Machine learning strategy for accelerated design of polymer dielectrics. Sci Rep 6:20952CrossRefGoogle Scholar
  66. 66.
    Hastie T, Tibshirani R, Friedman J (2009) The elements of statistical learning: data mining, inference, and prediction. Springer, New YorkCrossRefGoogle Scholar
  67. 67.
    Vu K et al (2015) Understanding kernel ridge regression: common behaviors from simple functions to density functionals. Int J Quantum Chem 115:1115–1128CrossRefGoogle Scholar
  68. 68.
    Rupp M (2015) Machine learning for quantum mechanics in a nutshell. Int J Quantum Chem 115:1058–1073CrossRefGoogle Scholar
  69. 69.
    Muller K-R, Mika S, Ratsch G, Tsuda K, Scholkopf B (2001) An introduction to kernel-based learning algorithms. IEEE Trans Neural Netw 12:181CrossRefGoogle Scholar
  70. 70.
    Sze SM, Ng KK (2006) Physics of semiconductor devices. Wiley, HobokenCrossRefGoogle Scholar
  71. 71.
    Weber MJ (2002) Handbook of optical materials. CRC Press, Boca RatonCrossRefGoogle Scholar
  72. 72.
    Madelung O (2012) Semiconductors: data handbook. Springer, New YorkGoogle Scholar
  73. 73.
    Sham LJ, Schlüter M (1983) Density-functional theory of the energy gap. Phys Rev Lett 51:1888CrossRefGoogle Scholar
  74. 74.
    Cohen AJ, Mori-Sánchez P, Yang W (2008) Fractional charge perspective on the bandgap in density-functional theory. Phys Rev B 77:115123CrossRefGoogle Scholar
  75. 75.
    Mori-Sánchez P, Cohen AJ, Yang W (2008) Localization and delocalization errors in density functional theory and implications for band-gap prediction. Phys Rev Lett 100:146401CrossRefGoogle Scholar
  76. 76.
    Jain A, Shin Y, Persson KA (2016) Computational predictions of energy materials using density functional theory. Nat Rev Mater 1:15004CrossRefGoogle Scholar
  77. 77.
    Olivares-Amaya R et al (2011) Accelerated computational discovery of high-performance materials for organic photovoltaics by means of cheminformatics. Energy Environ Sci 4:4849CrossRefGoogle Scholar
  78. 78.
    Nilsson A, Pettersson LG, Nørskov J (2011) Chemical bonding at surfaces and interfaces. Elsevier, AmsterdamGoogle Scholar
  79. 79.
    Dey P, Bible J, Datta S, Broderick S, Jasinski J, Sunkara M, Menon M, Rajan K (2014) Informatics-aided bandgap engineering for solar materials. Comput Mater Sci 83:185CrossRefGoogle Scholar
  80. 80.
    Castelli IE, Garca-Lastra JM, Hüser F, Thygesen KS, Jacobsen KW (2013) Stability and bandgaps of layered perovskites for one-and two-photon water splitting. New J Phys 15:105026CrossRefGoogle Scholar
  81. 81.
    Rasmussen F, Thygesen KS (2015) Computational 2D materials database: electronic structure of transition-metal dichalcogenides and oxides. Phys Chem C 119:13169–13183CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Materials Science and Technology DivisionLos Alamos National LaboratoryLos AlamosUSA

Personalised recommendations