Overview of accelerated materials design efforts in the Hacking Materials research group
1 like331 views
This document provides an overview of accelerated materials design efforts in the Hacking Materials research group. It discusses using high-throughput computing and simulations like density functional theory to generate large datasets for materials screening. Machine learning techniques like matminer are used to represent materials as feature vectors to enable predictive modeling. Text mining of scientific literature is also discussed as a way to automatically extract knowledge from millions of published articles to inform new materials discoveries. The goal is to develop automated methods that can suggest the next best computational experiments to optimize properties of interest.
![What constrains traditional approaches to materials design?
4
“[The Chevrel] discovery resulted from a lot of
unsuccessful experiments of Mg ions insertion
into well-known hosts for Li+ ions insertion, as
well as from the thorough literature analysis
concerning the possibility of divalent ions
intercalation into inorganic materials.”
-Aurbach group, on discovery of Chevrel cathode
for multivalent (e.g., Mg2+) batteries
Levi, Levi, Chasid, Aurbach
J. Electroceramics (2009)](https://image.slidesharecdn.com/nissanml-190401193838/85/Overview-of-accelerated-materials-design-efforts-in-the-Hacking-Materials-research-group-4-320.jpg)
![>60 featurizer classes can
generate thousands of potential
descriptors that are described in
the literature
34
Matminer contains a library of descriptors for various
materials science entities
feat = EwaldEnergy([options])
y = feat.featurize([input_data])
• compatible with scikit-
learn pipelining
• automatically deploy
multiprocessing to
parallelize over data
• include citations to
methodology papers](https://image.slidesharecdn.com/nissanml-190401193838/85/Overview-of-accelerated-materials-design-efforts-in-the-Hacking-Materials-research-group-34-320.jpg)
![• Thus far, 2 of our top 20 predictions made in
~August 2018 have already been reported in the
literature for the first time as thermoelectrics
– Li3Sb was the subject of a computational study
(predicted zT=2.42) in Oct 2018
– SnTe2 was experimentally found to be a moderately
good thermoelectric (expt zT=0.71) in Dec 2018
• We are currently trying to make some of the
other compounds on the list
– The full list will be published alongside a paper
(currently under review)
53
Next steps
[1] Yang et al. "Low lattice thermal conductivity and
excellent thermoelectric behavior in Li3Sb and Li3Bi."
Journal of Physics: Condensed Matter 30.42 (2018):
425401
[2] Wang et al. "Ultralow lattice thermal conductivity and
electronic properties of monolayer 1T phase semimetal
SiTe2 and SnTe2." Physica E: Low-dimensional Systems and
Nanostructures 108 (2019): 53-59](https://image.slidesharecdn.com/nissanml-190401193838/85/Overview-of-accelerated-materials-design-efforts-in-the-Hacking-Materials-research-group-53-320.jpg)
Overview of accelerated materials design efforts in the Hacking Materials research group
- 1. Overview of accelerated materials design efforts in the Hacking Materials research group Anubhav Jain Energy Technologies Area Lawrence Berkeley National Laboratory Berkeley, CA
- 2. 2 Materials and their properties determine what is technologically possible Electric vehicles and solar power are two technologies that have been dreamed about for many decades, but did not have much real impact for a long time … 1910 1956
- 3. 3 Materials and their properties determine what is technologically possible Today’s revolution in clean energy technologies are largely due to advancements in materials – science, engineering, and manufacturing. Much else might be possible with better materials … but, as past examples demonstrate, it can take a long time.
- 4. What constrains traditional approaches to materials design? 4 “[The Chevrel] discovery resulted from a lot of unsuccessful experiments of Mg ions insertion into well-known hosts for Li+ ions insertion, as well as from the thorough literature analysis concerning the possibility of divalent ions intercalation into inorganic materials.” -Aurbach group, on discovery of Chevrel cathode for multivalent (e.g., Mg2+) batteries Levi, Levi, Chasid, Aurbach J. Electroceramics (2009)
- 7. What is density functional theory (DFT)? 7 • 1920s: The Schrödinger equation essentially contains all of chemistry embedded within it • it is almost always too complicated to solve due to the numerous electron interactions and complexity of the wave function entity • 1960s: DFT is developed and reframes the problem for ground state properties of the system to be in terms of the charge density, not wavefunction • makes solutions tractable while in principle not sacrificing accuracy for the ground state! e– e– e– e– e– e–
- 8. How does one use DFT to design new materials? 8 A. Jain, Y. Shin, and K. A. Persson, Nat. Rev. Mater. 1, 15004 (2016).
- 9. 9 Examples of experimentally-confirmed materials designed with DFT (1) Jain, A., Shin, Y., Persson, K.A., 2016. Computational predictions of energy materials using density functional theory. Nature Reviews Materials 1, 15004.
- 10. 10 Examples of experimentally-confirmed materials designed with DFT (2) Jain, A., Shin, Y., Persson, K.A., 2016. Computational predictions of energy materials using density functional theory. Nature Reviews Materials 1, 15004.
- 11. High-throughput DFT is useful for generating large data sets, e.g., for materials screening 11 M. De Jong, W. Chen, T. Angsten, A. Jain, R. Notestine, A. Gamst, M. Sluiter, C. K. Ande, S. Van Der Zwaag, J. J. Plata, C. Toher, S. Curtarolo, G. Ceder, K. a Persson, and M. Asta, Sci. Data, 2015, 2, 150009. >10,000 elastic tensors >48000 Seebeck coefficients + cRTA transport Ricci, Chen, Aydemir, Snyder, Rignanese, Jain, & Hautier, Sci Data 2017, 4, 170085.
- 12. High-throughput DFT is useful for generating large data sets, e.g., for materials screening 12 M. De Jong, W. Chen, T. Angsten, A. Jain, R. Notestine, A. Gamst, M. Sluiter, C. K. Ande, S. Van Der Zwaag, J. J. Plata, C. Toher, S. Curtarolo, G. Ceder, K. a Persson, and M. Asta, Sci. Data, 2015, 2, 150009. >10,000 elastic tensors >48000 Seebeck coefficients + cRTA transport Ricci, Chen, Aydemir, Snyder, Rignanese, Jain, & Hautier, Sci Data 2017, 4, 170085. Atomate’s goal: make high-throughput easy and scalable for everyone
- 13. A “black-box” view of performing a calculation 13 “something” Results! researcher What is the GGA-PBE elastic tensor of GaAs?
- 14. Unfortunately, the inside of the “black box” is usually tedious and “low-level” 14 lots of tedious, low-level work… Results! researcher What is the GGA-PBE elastic tensor of GaAs? Input file flags SLURM format how to fix ZPOTRF? q set up the structure coordinates q write input files, double-check all the flags q copy to supercomputer q submit job to queue q deal with supercomputer headaches q monitor job q fix error jobs, resubmit to queue, wait again q repeat process for subsequent calculations in workflow q parse output files to obtain results q copy and organize results, e.g., into Excel
- 15. What would be a better way? 15 “something” Results! researcher What is the GGA-PBE elastic tensor of GaAs?
- 16. What would be a better way? 16 Results! researcher What is the GGA-PBE elastic tensor of GaAs? Workflows to run q band structure q surface energies ü elastic tensor q Raman spectrum q QH thermal expansion
- 17. Ideally the method should scale to millions of calculations 17 Results! researcher Start with all binary oxides, replace O->S, run several different properties Workflows to run ü band structure ü surface energies ü elastic tensor q Raman spectrum q QH thermal expansion q spin-orbit coupling
- 18. Atomate tries make it easy, automatic, and flexible to generate data with existing simulation packages 18 Results! researcher Run many different properties of many different materials!
- 19. Each simulation procedure translates high-level instructions into a series of low-level tasks 19 quickly and automatically translate high-level (minimal) specifications into well-defined FireWorks workflows What is the GGA-PBE elastic tensor of GaAs? M. De Jong, W. Chen, T. Angsten, A. Jain, R. Notestine, A. Gamst, et al., Charting the complete elastic properties of inorganic crystalline compounds, Sci. Data. 2 (2015).
- 20. Atomate contains a library of simulation procedures 20 VASP-based • band structure • spin-orbit coupling • hybrid functional calcs • elastic tensor • piezoelectric tensor • Raman spectra • NEB • GIBBS method • QH thermal expansion • AIMD • ferroelectric • surface adsorption • work functions • NMR spectra* • Bader charges* • Magnetic orderings* • SCAN functionals* Other • BoltzTraP • FEFF method • Q-Chem* *=added / major updates in past year Mathew, K. et al Atomate: A high-level interface to generate, execute, and analyze computational materials science workflows, Comput. Mater. Sci. 139 (2017) 140–152.
- 21. 21 Full operation diagram job 1 job 2 job 3 job 4 structure workflow database of all workflows automatically submit + executeoutput files + database
- 22. 22 A web-based interface is in progress to give atomate users a “personal Materials Project” of their own calculations
- 23. Atomate now powers the Materials Project • Online resource of density functional theory simulation data for ~85,000 inorganic materials • Includes band structures, elastic tensors, piezoelectric tensors, battery properties and more • >75,000 registered users • Free • www.materialsproject.org 23 Jain et al. Commentary: The Materials Project: A materials genome approach to accelerating materials innovation. APL Mater. 1, 11002 (2013).
- 24. 24 Getting started with atomate Mathew, K. et al. Atomate: A high- level interface to generate, execute, and analyze computational materials science workflows. Comput. Mater. Sci. 139, 140–152 (2017). hackingmaterials.github.io/ atomate https://groups.google.com/ forum/#!forum/atomate Paper Docs Support
- 26. • With atomate/FireWorks, the user must decide which calculations to perform – E.g., which materials to calculate • Rocketsled is an extension to FireWorks that lets the computer decide what the next best calculation is based on the results of previous calculations • Works for materials design or any other “inverse computational problem” 26 Rocketsled uses adaptive design to suggest the best computations to optimize some metric
- 27. 27 Given a search domain, Rocketsled uses an optimization engine to select calculations and submit to supercomputers Optimization engine includes 4 built-in regressors (e.g., RandomForest, Gaussian Process) and 5 acquisition functions (e.g., Expected Improvement). Can bootstrap uncertainty estimates. Or use your own!
- 28. 28 Results of using optimization can be dramatic! In the problem of finding materials with high K and high G for superhard materials (7394 possibilities), Rocketsled finds solutions ~30-60X faster than randomly computing the space. Can use pure ML approaches or use matminer featurizations for materials science (latter helps give such good performance)
- 29. 29 Results of using optimization can be dramatic! In the problem of finding materials with high K and high G for superhard materials (7394 possibilities), Rocketsled finds solutions ~30-60X faster than randomly computing the space. Even after just 200 calculations of the 7394 possibilities, all solutions are almost certain to be found with Rocketsled. Can use pure ML approaches or use matminer featurizations for materials science (latter helps give such good performance)
- 30. 30 Getting started with rocketsled Dunn, A.R., et al. Rocketsled: a software library for optimizing high-throughput computational searches. J. Phys. Mater. https://doi.org/10.1088/2515- 7639/ab0c3d hackingmaterials.github.io/ rocketsled https://groups.google.com/for um/#!forum/fireworkflows Paper Docs Support
- 32. 32 What is needed to do machine learning on materials? How can we represent chemistry and structure as vectors? How do we get enough output data for training?
- 33. Matminer connects materials data with data mining algorithms and data visualization libraries 33 Ward, L. et al. Matminer: An open source toolkit for materials data mining. Comput. Mater. Sci. 152, 60–69 (2018).
- 34. >60 featurizer classes can generate thousands of potential descriptors that are described in the literature 34 Matminer contains a library of descriptors for various materials science entities feat = EwaldEnergy([options]) y = feat.featurize([input_data]) • compatible with scikit- learn pipelining • automatically deploy multiprocessing to parallelize over data • include citations to methodology papers
- 35. 35 Interactive Jupyter notebooks demonstrate use cases https://github.com/hackingmaterials/matminer_examples Many examples available: • Retrieving data from various databases • Predicting bulk / shear modulus • Predicting formation energies: • from composition alone • with Voronoi-based structure features included • with Coulomb matrix and Orbital Field matrix descriptors (reproducing previous studies in the literature) • Making interactive visualizations • Creating an ML pipeline
- 36. 36 Getting started with matminer Ward et al. Matminer : An open source toolkit for materials data mining. Computational Materials Science, 152, 60–69 (2018). Paper Docs Support hackingmaterials.github.io /matminer https://groups.google.com/ forum/#!forum/matminer
- 38. 38 Typically several steps of machine learning are performed by a human researcher – can these be automated? Descriptors developed and chosen by a researcher ML model developed and chosen by a researcher Why can’t we just give the computer some raw input data (compositions, crystal structures) and output properties and get back an ML model?
- 39. 39 Automatminer develops an ML model automatically given raw data (structures or compositions plus output properties) Featurizer MagPie SOAP Sine Coulomb Matrix + many, many more • Missing value imputation • Scaling • One-hot encoding • PCA-based • Correlation • Relief-based (MultiSURF) Uses genetic algorithms to find the best machine learning model + hyperparameters
- 40. 40 Automatminer can be used as a black box
- 41. 41 We are benchmarking automatminer vs current state of the art against 11 problems intended to be a standard test set Dataset Target(s) Samples Elastic Tensor KVRH (GPa), GVRH (GPa) 10,987 Dielectric Tensor Refractive index 4,765 JARVIS 2D Exfoliation energy (meV/atom) 636 Materials Project phonons Highest LO Phonon Frequency (Last PhDOS peak) 1,265 Materials Project (stable) Band gap (eV), Is metallic? (classification) 106,113 Perovskites Formation energy (eV/atom) 18,928 Experimental Band Gaps Is metallic? (classification) 6,354 Experimental Metallic Glasses Glass forms? (classification) 7,190 Materials Project (all) Formation energy (eV/atom) 132,752
- 42. 42 Usually, automatminer does very well Usually, automatminer outperforms both state-of-the-art graph based models AND human-generated models! But …
- 43. 43 Graph-based approaches work better in some problems Hypothesis – automatminer approaches are better for smaller data sets, graph-based approaches are better for larger data sets Unfortunately, it can be difficult to train some of the graph models on large data sets, particularly without GPUs, so the results are not in yet!
- 44. 44 Getting started with automatminer Paper Docs Support hackingmaterials.github.io /automatminer https://groups.google.com/ forum/#!forum/matminer In preparation …
- 46. We have extracted ~3 million abstracts of scientific articles We will use natural language processing algorithms to try to extract knowledge from all this data 46 Goal: collect knowledge embedded in the scientific literature
- 47. 47 An engine to label the content of scientific abstracts Collect, clean, and extract information from millions of published materials science journal abstracts
- 48. 48 Application: a revised materials search engine Auto-generated summaries of materials based on text mining
- 49. 49 Application: materials compositions of interest … A search for thermoelectrics that do not have Pb or Bi
- 50. • We use the word2vec algorithm (Google) to turn each unique word in our corpus into a 200- dimensional vector • These vectors encode the meaning of each word meaning based on trying to predict context words around the target 50 Using the word2vec algorithm to extract knowledge and make predictions
- 51. • Dot product of a composition word with the word “thermoelectric” essentially predicts how likely that word is to appear in an abstract with the word thermoelectric • Compositions with high dot products are typically known thermoelectrics • Sometimes, compositions have a high dot product with “thermoelectric” but have never been studied as a thermoelectric • These compositions usually have high computed power factors! (BoltzTraP) 51 Vector dot products measure similarity
- 52. • We can test to see if our method can predict new compositions • For every year in the past ~2 decades (e.g. 2001), train word embeddings only until that point in time • Make predictions of what materials are the most promising thermoelectrics • See if those materials have thus far been actually studied as thermoelectrics 52 Can we predict future thermoelectrics discoveries with this method?
- 53. • Thus far, 2 of our top 20 predictions made in ~August 2018 have already been reported in the literature for the first time as thermoelectrics – Li3Sb was the subject of a computational study (predicted zT=2.42) in Oct 2018 – SnTe2 was experimentally found to be a moderately good thermoelectric (expt zT=0.71) in Dec 2018 • We are currently trying to make some of the other compounds on the list – The full list will be published alongside a paper (currently under review) 53 Next steps [1] Yang et al. "Low lattice thermal conductivity and excellent thermoelectric behavior in Li3Sb and Li3Bi." Journal of Physics: Condensed Matter 30.42 (2018): 425401 [2] Wang et al. "Ultralow lattice thermal conductivity and electronic properties of monolayer 1T phase semimetal SiTe2 and SnTe2." Physica E: Low-dimensional Systems and Nanostructures 108 (2019): 53-59
- 55. • Lead developers: – Atomate: Kiran Mathew – Rocketsled: Alex Dunn – Matminer: Logan Ward – Automatminer: Alex Dunn – Matscholar: Vahe Tshitoyan, John Dagdelen, Leigh Weston • And the dozens of other developers who have contributed to these packages or reported issues! • Funding: U.S. Department of Energy, Basic Energy Sciences, Early Career Award • Additional funding from the DOE-funded Materials Project 55 Acknowledgements