A Caltech Library Service

Discovery of complex oxides via automated experiments and data science

Yang, Lusann and Haber, Joel A. and Armstrong, Zan and Yang, Samuel J. and Kan, Kevin and Zhou, Lan and Richter, Matthias H. and Roat, Christopher and Wagner, Nicholas and Coram, Marc and Berndl, Marc and Riley, Patrick and Gregoire, John M. (2021) Discovery of complex oxides via automated experiments and data science. Proceedings of the National Academy of Sciences, 118 (37). Art. No. e2106042118. ISSN 0027-8424. PMCID PMC8449358. doi:10.1073/pnas.2106042118.

[img] PDF - Published Version
Creative Commons Attribution Non-commercial No Derivatives.

[img] PDF - Supplemental Material
Creative Commons Attribution Non-commercial No Derivatives.


Use this Persistent URL to link to this item:


The quest to identify materials with tailored properties is increasingly expanding into high-order composition spaces, with a corresponding combinatorial explosion in the number of candidate materials. A key challenge is to discover regions in composition space where materials have novel properties. Traditional predictive models for material properties are not accurate enough to guide the search. Herein, we use high-throughput measurements of optical properties to identify novel regions in three-cation metal oxide composition spaces by identifying compositions whose optical trends cannot be explained by simple phase mixtures. We screen 376,752 distinct compositions from 108 three-cation oxide systems based on the cation elements Mg, Fe, Co, Ni, Cu, Y, In, Sn, Ce, and Ta. Data models for candidate phase diagrams and three-cation compositions with emergent optical properties guide the discovery of materials with complex phase-dependent properties, as demonstrated by the discovery of a Co-Ta-Sn substitutional alloy oxide with tunable transparency, catalytic activity, and stability in strong acid electrolytes. These results required close coupling of data validation to experiment design to generate a reliable end-to-end high-throughput workflow for accelerating scientific discovery.

Item Type:Article
Related URLs:
URLURL TypeDescription Information CentralArticle ItemData
Yang, Lusann0000-0001-5071-2936
Haber, Joel A.0000-0001-7847-5506
Yang, Samuel J.0000-0003-2460-6456
Richter, Matthias H.0000-0003-0091-2045
Berndl, Marc0000-0001-6750-5861
Gregoire, John M.0000-0002-2863-5265
Additional Information:© 2021 the Author(s). Published by PNAS. This open access article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND). Edited by Alexis T. Bell, University of California, Berkeley, CA, and approved August 2, 2021 (received for review April 16, 2021). This material is based on work performed by the Joint Center for Artificial Photosynthesis, a Department of Energy Energy Innovation Hub, supported through the US Department of Energy Office of Basic Energy Sciences under Award DE-SC0004993, which supported materials synthesis and characterization experiments. Google Applied Science supported the development and execution of the computational workflow as well as procurement of the hyperspectral microscope. Structural characterization and follow-up validation experiments were supported by the US Department of Energy Office of Basic Energy Sciences under Award DE-SC0020383. We are grateful for helpful discussions and guidance in the development of the computational workflow from Muskaan Goyal, Eric Christiansen, Edward A. Baltz, Derek Leong, Austin Blanco, and Mike Ando (Google Applied Science). We also appreciate the support of the experiment workflow from Edwin Soedarmadji (Caltech). We additionally appreciate helpful suggestions by David Fork and Michael Brenner (Google Applied Science). Data Availability: The optical absorption spectra, fitted phase diagrams, and mixture probabilities have been deposited in Google Cloud Storage (; see SI Appendix for documentation and access instructions). Author contributions: L.Y., J.A.H., M.B., P.R., and J.M.G. designed research; L.Y., J.A.H., K.K., L.Z., and M.H.R. performed research; L.Y., J.A.H., Z.A., S.J.Y., C.R., M.C., M.B., P.R., and J.M.G. contributed new reagents/analytic tools; L.Y., J.A.H., N.W., and J.M.G. analyzed data; and L.Y., J.A.H., and J.M.G. wrote the paper. Competing interest statement: As listed in the affiliations, several authors are current or former employees of Google, a technology company that sells machine learning services as part of its business. This article is a PNAS Direct Submission. This article contains supporting information online at
Funding AgencyGrant Number
Department of Energy (DOE)DE-SC0004993
Google Applied ScienceUNSPECIFIED
Department of Energy (DOE)DE-SC0020383
Subject Keywords:data science; materials discovery; complex oxides; optical absorption; oxygen evolution electrocatalysis
Issue or Number:37
PubMed Central ID:PMC8449358
Record Number:CaltechAUTHORS:20210914-182227400
Persistent URL:
Official Citation:Discovery of complex oxides via automated experiments and data science. Lusann Yang, Joel A. Haber, Zan Armstrong, Samuel J. Yang, Kevin Kan, Lan Zhou, Matthias H. Richter, Christopher Roat, Nicholas Wagner, Marc Coram, Marc Berndl, Patrick Riley, John M. Gregoire. Proceedings of the National Academy of Sciences Sep 2021, 118 (37) e2106042118; DOI: 10.1073/pnas.2106042118
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:110848
Deposited By: Tony Diaz
Deposited On:14 Sep 2021 20:05
Last Modified:05 Oct 2021 18:27

Repository Staff Only: item control page