OpenLB
Open Library of Bioscience
Advanced Search
Nature computational science
Jul, 2024;4 (7): 522-531.

Virtual node graph neural network for full phonon prediction.

Ryotaro Okabe, Abhijatmedhi Chotrattanapituk, Artittaya Boonkird, Nina Andrejevic, Xiang Fu, Tommi S Jaakkola, Qichen Song, Thanh Nguyen, Nathan Drucker, Sai Mu, Yao Wang, Bolin Liao, Yongqiang Cheng, Mingda Li
Author Information
  1. Ryotaro Okabe: Quantum Measurement Group, Massachusetts Institute of Technology, Cambridge, MA, USA. rokabe@mit.edu. ORCID
  2. Abhijatmedhi Chotrattanapituk: Quantum Measurement Group, Massachusetts Institute of Technology, Cambridge, MA, USA. ORCID
  3. Artittaya Boonkird: Quantum Measurement Group, Massachusetts Institute of Technology, Cambridge, MA, USA.
  4. Nina Andrejevic: Argonne National Laboratory, Lemont, IL, USA.
  5. Xiang Fu: Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, USA. ORCID
  6. Tommi S Jaakkola: Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, USA.
  7. Qichen Song: Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA, USA. ORCID
  8. Thanh Nguyen: Quantum Measurement Group, Massachusetts Institute of Technology, Cambridge, MA, USA.
  9. Nathan Drucker: Quantum Measurement Group, Massachusetts Institute of Technology, Cambridge, MA, USA. ORCID
  10. Sai Mu: SmartState Center for Experimental Nanoscale Physics, Department of Physics and Astronomy, University of South Carolina, Columbia, SC, USA.
  11. Yao Wang: Department of Chemistry, Emory University, Atlanta, GA, USA. ORCID
  12. Bolin Liao: Department of Materials, University of California, Santa Barbara, Santa Barbara, CA, USA. ORCID
  13. Yongqiang Cheng: Chemical Spectroscopy Group, Spectroscopy Section, Neutron Scattering Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA. chengy@ornl.gov. ORCID
  14. Mingda Li: Quantum Measurement Group, Massachusetts Institute of Technology, Cambridge, MA, USA. mingda@mit.edu. ORCID
PMID: 38997585 DOI: 10.1038/s43588-024-00661-0

Abstract

Understanding the structure-property relationship is crucial for designing materials with desired properties. The past few years have witnessed remarkable progress in machine-learning methods for this connection. However, substantial challenges remain, including the generalizability of models and prediction of properties with materials-dependent output dimensions. Here we present the virtual node graph neural network to address the challenges. By developing three virtual node approaches, we achieve Γ-phonon spectra and full phonon dispersion prediction from atomic coordinates. We show that, compared with the machine-learning interatomic potentials, our approach achieves orders-of-magnitude-higher efficiency with comparable to better accuracy. This allows us to generate databases for Γ-phonon containing over 146,000 materials and phonon band structures of zeolites. Our work provides an avenue for rapid and high-quality prediction of phonon band structures enabling materials design with desired phonon properties. The virtual node method also provides a generic method for machine-learning design with a high level of flexibility.

References

  1. Oganov, A. R. & Glass, C. W. Crystal structure prediction using ab initio evolutionary techniques: principles and applications. J. Chem. Phys. 124, 244704 (2006). [DOI: 10.1063/1.2210932]
  2. Le, T., Epa, V. C., Burden, F. R. & Winkler, D. A. Quantitative structure–property relationship modeling of diverse materials properties. Chem. Rev. 112, 2889–2919 (2012). [DOI: 10.1021/cr200066h]
  3. Cheng, Y. & Ma, E. Atomic-level structure and structure–property relationship in metallic glasses. Progress Mater. Sci. 56, 379–473 (2011). [DOI: 10.1016/j.pmatsci.2010.12.002]
  4. Mishra, A., Fischer, M. K. & Bäuerle, P. Metal-free organic dyes for dye-sensitized solar cells: from structure:property relationships to design rules. Angew. Chem. Int. Ed. 48, 2474–2499 (2009). [DOI: 10.1002/anie.200804709]
  5. Dresselhaus, M. S. et al. New directions for low-dimensional thermoelectric materials. Adv. Mater. 19, 1043–1053 (2007). [DOI: 10.1002/adma.200600527]
  6. Liu, Z. et al. Antiferroelectrics for energy storage applications: a review. Adv. Mater. Technol. 3, 1800111 (2018). [DOI: 10.1002/admt.201800111]
  7. Zheng, W. & Lee, L. Y. S. Metal–organic frameworks for electrocatalysis: catalyst or precatalyst? ACS Energy Lett. 6, 2838–2843 (2021). [DOI: 10.1021/acsenergylett.1c01350]
  8. Jancar, J. et al. Current issues in research on structure–property relationships in polymer nanocomposites. Polymer 51, 3321–3343 (2010). [DOI: 10.1016/j.polymer.2010.04.074]
  9. Kumar, N., Guin, S. N., Manna, K., Shekhar, C. & Felser, C. Topological quantum materials from the viewpoint of chemistry. Chem. Rev. 121, 2780–2815 (2020). [DOI: 10.1021/acs.chemrev.0c00732]
  10. Oganov, A. R., Pickard, C. J., Zhu, Q. & Needs, R. J. Structure prediction drives materials discovery. Nat. Rev. Mater. 4, 331–348 (2019). [DOI: 10.1038/s41578-019-0101-8]
  11. Zhu, T. et al. Charting lattice thermal conductivity for inorganic crystals and discovering rare earth chalcogenides for thermoelectrics. Energy Environ. Sci. 14, 3559–3566 (2021). [DOI: 10.1039/D1EE00442E]
  12. Dunn, A., Wang, Q., Ganose, A., Dopp, D. & Jain, A. Benchmarking materials property prediction methods: the matbench test set and automatminer reference algorithm. npj Comput. Mater. 6, 1–10 (2020).
  13. Peng, J. et al. Human and machine-centred designs of molecules and materials for sustainability and decarbonization. Nat. Rev. Mater. https://doi.org/10.1038/s41578-022-00466-5 (2022).
  14. Altintas, C., Altundal, O. F., Keskin, S. & Yildirim, R. Machine learning meets with metal organic frameworks for gas storage and separation. J. Chem. Inform. Model. 61, 2131–2146 (2021). [DOI: 10.1021/acs.jcim.1c00191]
  15. Schwalbe-Koda, D. et al. A priori control of zeolite phase competition and intergrowth with high-throughput simulations. Science 374, 308–315 (2021). [DOI: 10.1126/science.abh3350]
  16. Yao, Y. et al. High-entropy nanoparticles: synthesis–structure–property relationships and data-driven discovery. Science 376, eabn3103 (2022). [DOI: 10.1126/science.abn3103]
  17. Hart, G. L., Mueller, T., Toher, C. & Curtarolo, S. Machine learning for alloys. Nat. Rev. Mater. 6, 730–755 (2021). [DOI: 10.1038/s41578-021-00340-w]
  18. Wagih, M., Larsen, P. M. & Schuh, C. A. Learning grain boundary segregation energy spectra in polycrystals. Nat. Commun. 11, 1–9 (2020). [DOI: 10.1038/s41467-020-20083-6]
  19. Guo, K., Yang, Z., Yu, C.-H. & Buehler, M. J. Artificial intelligence and machine learning in design of mechanical materials. Mater. Horizons 8, 1153–1172 (2021). [DOI: 10.1039/D0MH01451F]
  20. Stanev, V., Choudhary, K., Kusne, A. G., Paglione, J. & Takeuchi, I. Artificial intelligence for search and discovery of quantum materials. Commun. Mater. 2, 1–11 (2021). [DOI: 10.1038/s43246-021-00209-z]
  21. Fung, V., Zhang, J., Juarez, E. & Sumpter, B. G. Benchmarking graph neural networks for materials chemistry. npj Comput. Mater. 7, 1–8 (2021). [DOI: 10.1038/s41524-021-00554-0]
  22. Xie, T. & Grossman, J. C. Crystal graph convolutional neural networks for an accurate and interpretable prediction of material properties. Phys. Rev. Lett. 120, 145301 (2018). [DOI: 10.1103/PhysRevLett.120.145301]
  23. Geiger, M. and Smidt, T. e3nn: Euclidean neural networks. Preprint at https://arxiv.org/abs/2207.09453v1 (2022).
  24. Thomas, N. et al. Tensor field networks: rotation- and translation-equivariant neural networks for 3D point clouds. Preprint at https://arxiv.org/abs/1802.08219v3 (2018).
  25. Delaire, O. et al. Phonon density of states and heat capacity of LaTe. Phys. Rev. B 80, 184302 (2009). [DOI: 10.1103/PhysRevB.80.184302]
  26. Baroni, S., Gironcoli, S. & Corso, A. Phonons and related crystal properties from density-functional perturbation theory. Rev. Mod. Phys. 73, 515 (2001). [DOI: 10.1103/RevModPhys.73.515]
  27. Kong, L. T. Phonon dispersion measured directly from molecular dynamics simulations. Comput. Phys. Commun. 182, 2201–2207 (2011). [DOI: 10.1016/j.cpc.2011.04.019]
  28. Petretto, G. et al. High-throughput density-functional perturbation theory phonons for inorganic materials. Sci. Data 5, 1–12 (2018). [DOI: 10.1038/sdata.2018.65]
  29. Togo, A. Phonon database at Kyoto University. Kyoto University https://github.com/atztogo/phonondb/tree/main (2015).
  30. Sham, L. Electronic contribution to lattice dynamics in insulating crystals. Phys. Rev. 188, 1431 (1969). [DOI: 10.1103/PhysRev.188.1431]
  31. Chen, C. & Ong, S. P. A universal graph deep learning interatomic potential for the periodic table. Nat. Comput. Sci. 2, 718–728 (2022). [DOI: 10.1038/s43588-022-00349-3]
  32. Batatia, I., Kovacs, D. P., Simm, G., Ortner, C. & Csanyi, G. MACE: higher order equivariant message passing neural networks for fast and accurate force fields. In 36th Conference on Neural Information Processing Systems (eds Koyejo, S. et al.) 11423–11436 (Curran Associates, 2022).
  33. Choudhary, K. & DeCost, B. Atomistic Line Graph Neural Network for improved materials property predictions. npj Comput. Mater. 7, 185 (2021). [DOI: 10.1038/s41524-021-00650-1]
  34. Deng, B. et al. CHGNet as a pretrained universal neural network potential for charge-informed atomistic modelling. Nat. Mach. Intell. 5, 1031–1041 (2023). [DOI: 10.1038/s42256-023-00716-3]
  35. Yu, H., Giantomassi, M., Materzanini, G., Wang, J. & Rignanese, G. M. Systematic assessment of various universal machine-learning interatomic potentials. Preprint at https://arxiv.org/abs/2403.05729v2 (2024).
  36. Kong, S. et al. Density of states prediction for materials discovery via contrastive learning from probabilistic embeddings. Nat. Commun. 13, 949 (2022). [DOI: 10.1038/s41467-022-28543-x]
  37. Chen, Z. et al. Direct prediction of phonon density of states with Euclidean neural networks. Adv. Sci. 8, 2004214 (2021). [DOI: 10.1002/advs.202004214]
  38. Jain, A. et al. Commentary: The Materials Project: a materials genome approach to accelerating materials innovation. APL Mater. 1, 011002 (2013). [DOI: 10.1063/1.4812323]
  39. Hinuma, Y., Pizzi, G., Kumagai, Y., Oba, F. & Tanaka, I. Band structure diagram paths based on crystallography. Comput. Mater. Sci. 128, 140–184 (2017). [DOI: 10.1016/j.commatsci.2016.10.015]
  40. Togo, A. & Tanaka, I. spglib: a software library for crystal symmetry search. Preprint at https://arxiv.org/abs/1808.01590v2 (2018).
  41. Togo, A. & Tanaka, I. First principles phonon calculations in materials science. Scripta Mater. 108, 1–5 (2015). [DOI: 10.1016/j.scriptamat.2015.07.021]
  42. Ong, S. P. et al. Python materials genomics (pymatgen): a robust, open-source Python library for materials analysis. Comput. Mater. Sci. 68, 314–319 (2013). [DOI: 10.1016/j.commatsci.2012.10.028]
  43. Mortensen, J. et al. The atomic simulation environment—a Python library for working with atoms. J. Phys. Condens. Matter 29, 273002–273002 (2017). [DOI: 10.1088/1361-648X/aa680e]
  44. Paszke, A. et al. PyTorch: an imperative style, high-performance deep learning library. Adv. Neural Inf. Process. Syst. 32, 8024–8035 (2019).
  45. Miller, B. K., Geiger, M., Smidt, T. E. & Noé, F. Relevance of rotationally equivariant convolutions for predicting molecular properties. Preprint at arXiv https://arxiv.org/abs/2008.08461v4 (2020).
  46. Okabe, R. & Chotrattanapituk, A. Virtual node graph neural network for full phonon prediction. OSF https://doi.org/10.17605/OSF.IO/K5UTB (2024).
  47. Okabe, R. & Chotrattanapituk, A. Virtual node graph neural network for full phonon prediction. Zenodo https://doi.org/10.5281/zenodo.8028365 (2024).

Grants

  1. DE-SC0020148/U.S. Department of Energy (DOE)
  2. DE-SC0021940/U.S. Department of Energy (DOE)
  3. DE-AC05-00OR22725/U.S. Department of Energy (DOE)
  4. DMR-2118448/NSF | Directorate for Mathematical & Physical Sciences | Division of Materials Research (DMR)
  5. DMR-2118523/NSF | Directorate for Mathematical & Physical Sciences | Division of Materials Research (DMR)
  6. NSF ITE-2345084/National Science Foundation (NSF)
Journal Article
Article has an altmetric score of 144

Word Cloud

Created with Highcharts 10.0.0phononpredictionnodematerialspropertiesmachine-learningvirtualdesiredchallengesgraphneuralnetworkΓ-phononfullbandstructuresprovidesdesignmethodUnderstandingstructure-propertyrelationshipcrucialdesigningpastyearswitnessedremarkableprogressmethodsconnectionHoweversubstantialremainincludinggeneralizabilitymodelsmaterials-dependentoutputdimensionspresentaddressdevelopingthreeapproachesachievespectradispersionatomiccoordinatesshowcomparedinteratomicpotentialsapproachachievesorders-of-magnitude-higherefficiencycomparablebetteraccuracyallowsusgeneratedatabasescontaining146000zeolitesworkavenuerapidhigh-qualityenablingalsogenerichighlevelflexibilityVirtual

Similar Articles

Cited By