DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Combining machine learning and high-throughput experimentation to discover photocatalytically active organic molecules

Abstract

Light-absorbing organic molecules are useful components in photocatalysts, but it is difficult to formulate reliable structure–property design rules. More than 100 million unique chemical compounds are documented in the PubChem database, and a significant sub-set of these are π-conjugated, light-absorbing molecules that might in principle act as photocatalysts. Nature has used natural selection to evolve photosynthetic assemblies; by contrast, our ability to navigate the enormous potential search space of organic photocatalysts in the laboratory is limited. Here, we integrate experiment, computation, and machine learning to address this challenge. A library of 572 aromatic organic molecules was assembled with diverse compositions and structures, selected on the basis of availability in our laboratory, rather than more sophisticated criteria. This training library was then assessed experimentally for sacrificial photocatalytic hydrogen evolution using a high-throughput, automated method. Quantum chemical calculations and machine learning were used to visualise, interpret, and ultimately to predict the photocatalytic activities of these molecules, covering a much broader chemical space than for previous polymer photocatalyst libraries. By applying unsupervised learning to the molecular structures, we identified structural features that were common in molecules with high catalytic activity. Further analysis using calculated molecular descriptors within a suite of supervised classification algorithmsmore » revealed that light absorption, exciton electron affinity, electron affinity, exciton binding energy, and singlet–triplet energy gap had correlations with the photocatalytic performance. These trained predictive models can be used in future studies as filters to deprioritise or discard would-be low-activity candidate molecules from experiments, and to prioritize more favourable candidates. As a demonstration, we used virtual in silico experiments to show that it was possible to halve the experimental cost of finding 50% of the most active photocatalysts by using the machine learning model as an experimental advisor. We further showed that the ML advisor trained on the 572-molecule library could be used to make predictions for an unseen set of 96 molecules, achieving equivalent predictive accuracies to those in the initial training set. This marks a step toward the machine-learning assisted discovery of molecular organic photocatalysts and the approach might also be applied to problems beyond photocatalytic hydrogen evolution, such as CO2 reduction and photoredox chemistry.« less

Authors:
ORCiD logo [1];  [2]; ORCiD logo [3];  [1]; ORCiD logo [3]; ORCiD logo [3]
  1. Department of Chemistry & Materials Innovation Factory, University of Liverpool, 51 Oxford Street, Liverpool L7 3NY, UK
  2. Department of Chemistry & Materials Innovation Factory, University of Liverpool, 51 Oxford Street, Liverpool L7 3NY, UK, National Synchrotron Light Source II, Brookhaven National Laboratory, Upton, New York 11973, USA
  3. Department of Chemistry & Materials Innovation Factory, University of Liverpool, 51 Oxford Street, Liverpool L7 3NY, UK, Leverhulme Research Centre for Functional Materials Design, Materials Innovation Factory and Department of Chemistry, University of Liverpool, 51 Oxford Street, Liverpool L7 3NY, UK
Publication Date:
Research Org.:
Brookhaven National Lab. (BNL), Upton, NY (United States)
Sponsoring Org.:
USDOE Office of Science (SC), Basic Energy Sciences (BES); USDOE Laboratory Directed Research and Development (LDRD) Program; Engineering and Physical Science Research Council (EPSRC)
OSTI Identifier:
1798927
Alternate Identifier(s):
OSTI ID: 1809061
Report Number(s):
BNL-221836-2021-JAAM
Journal ID: ISSN 2041-6520; CSHCBM
Grant/Contract Number:  
SC0012704; EP/N004884/1
Resource Type:
Published Article
Journal Name:
Chemical Science
Additional Journal Information:
Journal Name: Chemical Science Journal Volume: 12 Journal Issue: 32; Journal ID: ISSN 2041-6520
Publisher:
Royal Society of Chemistry (RSC)
Country of Publication:
United Kingdom
Language:
English
Subject:
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY

Citation Formats

Li, Xiaobo, Maffettone, Phillip M., Che, Yu, Liu, Tao, Chen, Linjiang, and Cooper, Andrew I. Combining machine learning and high-throughput experimentation to discover photocatalytically active organic molecules. United Kingdom: N. p., 2021. Web. doi:10.1039/D1SC02150H.
Li, Xiaobo, Maffettone, Phillip M., Che, Yu, Liu, Tao, Chen, Linjiang, & Cooper, Andrew I. Combining machine learning and high-throughput experimentation to discover photocatalytically active organic molecules. United Kingdom. https://doi.org/10.1039/D1SC02150H
Li, Xiaobo, Maffettone, Phillip M., Che, Yu, Liu, Tao, Chen, Linjiang, and Cooper, Andrew I. Wed . "Combining machine learning and high-throughput experimentation to discover photocatalytically active organic molecules". United Kingdom. https://doi.org/10.1039/D1SC02150H.
@article{osti_1798927,
title = {Combining machine learning and high-throughput experimentation to discover photocatalytically active organic molecules},
author = {Li, Xiaobo and Maffettone, Phillip M. and Che, Yu and Liu, Tao and Chen, Linjiang and Cooper, Andrew I.},
abstractNote = {Light-absorbing organic molecules are useful components in photocatalysts, but it is difficult to formulate reliable structure–property design rules. More than 100 million unique chemical compounds are documented in the PubChem database, and a significant sub-set of these are π-conjugated, light-absorbing molecules that might in principle act as photocatalysts. Nature has used natural selection to evolve photosynthetic assemblies; by contrast, our ability to navigate the enormous potential search space of organic photocatalysts in the laboratory is limited. Here, we integrate experiment, computation, and machine learning to address this challenge. A library of 572 aromatic organic molecules was assembled with diverse compositions and structures, selected on the basis of availability in our laboratory, rather than more sophisticated criteria. This training library was then assessed experimentally for sacrificial photocatalytic hydrogen evolution using a high-throughput, automated method. Quantum chemical calculations and machine learning were used to visualise, interpret, and ultimately to predict the photocatalytic activities of these molecules, covering a much broader chemical space than for previous polymer photocatalyst libraries. By applying unsupervised learning to the molecular structures, we identified structural features that were common in molecules with high catalytic activity. Further analysis using calculated molecular descriptors within a suite of supervised classification algorithms revealed that light absorption, exciton electron affinity, electron affinity, exciton binding energy, and singlet–triplet energy gap had correlations with the photocatalytic performance. These trained predictive models can be used in future studies as filters to deprioritise or discard would-be low-activity candidate molecules from experiments, and to prioritize more favourable candidates. As a demonstration, we used virtual in silico experiments to show that it was possible to halve the experimental cost of finding 50% of the most active photocatalysts by using the machine learning model as an experimental advisor. We further showed that the ML advisor trained on the 572-molecule library could be used to make predictions for an unseen set of 96 molecules, achieving equivalent predictive accuracies to those in the initial training set. This marks a step toward the machine-learning assisted discovery of molecular organic photocatalysts and the approach might also be applied to problems beyond photocatalytic hydrogen evolution, such as CO2 reduction and photoredox chemistry.},
doi = {10.1039/D1SC02150H},
journal = {Chemical Science},
number = 32,
volume = 12,
place = {United Kingdom},
year = {Wed Aug 18 00:00:00 EDT 2021},
month = {Wed Aug 18 00:00:00 EDT 2021}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record
https://doi.org/10.1039/D1SC02150H

Save / Share:

Works referenced in this record:

Photosensitized singlet oxygen and its applications
journal, November 2002


Organic Photoredox Catalysis
journal, June 2016


A Laser Flash Photolysis Study of Pyrene-1-Aldehyde. Intersystem Crossing Efficiency, Photoreactivity and Triplet State Properties in Various Solvents
journal, August 1983


Photochemical Mechanisms Responsible for the Versatile Application of Naphthalimides and Naphthaldiimides in Biological Systems
journal, December 1997

  • Aveline, Béatrice M.; Matsugo, Seiichi; Redmond, Robert W.
  • Journal of the American Chemical Society, Vol. 119, Issue 49
  • DOI: 10.1021/ja971993c

Shining Light on Carbon Nitrides: Leveraging Temperature To Understand Optical Gap Variations
journal, June 2018


Prediction of higher-selectivity catalysts by computer-driven workflow and machine learning
journal, January 2019


Proflavin Catalyzed Photoproduction of Hydrogen from Organic Compounds
journal, February 1979


On representing chemical environments
journal, May 2013


Quantum Yield of Singlet Oxygen Production by Xanthene Derivatives
journal, March 1983


Current understanding and challenges of solar-driven hydrogen generation using polymeric photocatalysts
journal, September 2019


QSAR Modeling: Where Have You Been? Where Are You Going To?
journal, January 2014

  • Cherkasov, Artem; Muratov, Eugene N.; Fourches, Denis
  • Journal of Medicinal Chemistry, Vol. 57, Issue 12
  • DOI: 10.1021/jm4004285

Design of BODIPY Dyes as Photosensitisers in Multicomponent Catalyst Systems for Light-Driven Hydrogen Production
journal, August 2015

  • Dura, Laura; Ahrens, Johannes; Pohl, Marga-Martina
  • Chemistry - A European Journal, Vol. 21, Issue 39
  • DOI: 10.1002/chem.201501637

Enhanced photocatalytic hydrogen evolution from organic semiconductor heterojunction nanoparticles
journal, February 2020


Structure–Property Relationships for Tailoring Phenoxazines as Reducing Photoredox Catalysts
journal, March 2018

  • McCarthy, Blaine G.; Pearson, Ryan M.; Lim, Chern-Hooi
  • Journal of the American Chemical Society, Vol. 140, Issue 15
  • DOI: 10.1021/jacs.7b12074

Triplet photosensitizers: from molecular design to applications
journal, January 2013

  • Zhao, Jianzhang; Wu, Wanhua; Sun, Jifu
  • Chemical Society Reviews, Vol. 42, Issue 12
  • DOI: 10.1039/c3cs35531d

Room-temperature phosphorescence from organic aggregates
journal, August 2020


Xanthene dyes as Sensitizers for the Photoreduction of Water
journal, May 1985


A metal-free polymeric photocatalyst for hydrogen production from water under visible light
journal, November 2008

  • Wang, Xinchen; Maeda, Kazuhiko; Thomas, Arne
  • Nature Materials, Vol. 8, Issue 1
  • DOI: 10.1038/nmat2317

Making Hydrogen from Water Using a Homogeneous System Without Noble Metals
journal, July 2009

  • Lazarides, Theodore; McCormick, Theresa; Du, Pingwu
  • Journal of the American Chemical Society, Vol. 131, Issue 26
  • DOI: 10.1021/ja903044n

Hydrogen Production by Molecular Photocatalysis
journal, October 2007

  • Esswein, Arthur J.; Nocera, Daniel G.
  • Chemical Reviews, Vol. 107, Issue 10
  • DOI: 10.1021/cr050193e

Virtual Excited State Reference for the Discovery of Electronic Materials Database: An Open-Access Resource for Ground and Excited State Properties of Organic Molecules
journal, October 2019

  • Abreha, Biruk G.; Agarwal, Snigdha; Foster, Ian
  • The Journal of Physical Chemistry Letters, Vol. 10, Issue 21
  • DOI: 10.1021/acs.jpclett.9b02577

Predicting reaction performance in C–N cross-coupling using machine learning
journal, February 2018


Machine Learning for Catalysis Informatics: Recent Applications and Prospects
journal, December 2019


Efficient Photocatalytic Hydrogen Evolution from Water without an Electron Mediator over Pt−Rose Bengal Catalysts
journal, January 2009

  • Zhang, Xiaojie; Jin, Zhiliang; Li, Yuexiang
  • The Journal of Physical Chemistry C, Vol. 113, Issue 6
  • DOI: 10.1021/jp8085717

Photoredox Catalysis in Organic Chemistry
journal, June 2016

  • Shaw, Megan H.; Twilton, Jack; MacMillan, David W. C.
  • The Journal of Organic Chemistry, Vol. 81, Issue 16
  • DOI: 10.1021/acs.joc.6b01449

Predicting the state of charge and health of batteries using data-driven machine learning
journal, March 2020


Comparing molecules and solids across structural and alchemical space
journal, January 2016

  • De, Sandip; Bartók, Albert P.; Csányi, Gábor
  • Physical Chemistry Chemical Physics, Vol. 18, Issue 20
  • DOI: 10.1039/C6CP00415F

Discovery and characterization of an acridine radical photoreductant
journal, April 2020


Carbon dots as photosensitisers for solar-driven catalysis
journal, January 2017

  • Hutton, Georgina A. M.; Martindale, Benjamin C. M.; Reisner, Erwin
  • Chemical Society Reviews, Vol. 46, Issue 20
  • DOI: 10.1039/C7CS00235A

Highly efficient organic light-emitting diodes from delayed fluorescence
journal, December 2012

  • Uoyama, Hiroki; Goushi, Kenichi; Shizu, Katsuyuki
  • Nature, Vol. 492, Issue 7428
  • DOI: 10.1038/nature11687

A mobile robotic chemist
journal, July 2020


Multiwfn: A multifunctional wavefunction analyzer
journal, December 2011

  • Lu, Tian; Chen, Feiwu
  • Journal of Computational Chemistry, Vol. 33, Issue 5
  • DOI: 10.1002/jcc.22885

Accelerated Discovery of Organic Polymer Photocatalysts for Hydrogen Evolution from Water through the Integration of Experiment and Theory
journal, May 2019

  • Bai, Yang; Wilbraham, Liam; Slater, Benjamin J.
  • Journal of the American Chemical Society, Vol. 141, Issue 22
  • DOI: 10.1021/jacs.9b03591

Data-driven design of metal–organic frameworks for wet flue gas CO2 capture
journal, December 2019


Recent Advances in Conjugated Polymers for Visible‐Light‐Driven Water Splitting
journal, June 2020


Molecular systems for light driven hydrogen production
journal, January 2012

  • Eckenhoff, William T.; Eisenberg, Richard
  • Dalton Transactions, Vol. 41, Issue 42
  • DOI: 10.1039/c2dt30823a

Selective and Efficient Photocatalytic CO 2 Reduction to CO Using Visible Light and an Iron-Based Homogeneous Catalyst
journal, November 2014

  • Bonin, Julien; Robert, Marc; Routier, Mathilde
  • Journal of the American Chemical Society, Vol. 136, Issue 48
  • DOI: 10.1021/ja510290t

Water Oxidation with Cobalt‐Loaded Linear Conjugated Polymer Photocatalysts
journal, August 2020

  • Sprick, Reiner Sebastian; Chen, Zheng; Cowan, Alexander J.
  • Angewandte Chemie International Edition, Vol. 59, Issue 42
  • DOI: 10.1002/anie.202008000

Machine Learning for Accelerated Discovery of Solar Photocatalysts
journal, October 2019


Metal-free photocatalysts for hydrogen evolution
journal, January 2020

  • Rahman, Mohammad Ziaur; Kibria, Md Golam; Mullins, Charles Buddie
  • Chemical Society Reviews, Vol. 49, Issue 6
  • DOI: 10.1039/C9CS00313D

Autonomous discovery in the chemical sciences part II: Outlook
journal, September 2019

  • Coley, Connor W.; Eyke, Natalie S.; Jensen, Klavs F.
  • Angewandte Chemie International Edition
  • DOI: 10.1002/anie.201909989

Production of Hydrogen Peroxide by Photocatalytic Processes
journal, May 2020

  • Hou, Huilin; Zeng, Xiangkang; Zhang, Xiwang
  • Angewandte Chemie International Edition, Vol. 59, Issue 40
  • DOI: 10.1002/anie.201911609

Light-induced polymer and polymerization reactions XXXIII: direct photoinitiation of methyl methacrylate polymerization by excited states of ketones
journal, February 1989

  • Timpe, H. -J.; Kronfeld, K. -P.
  • Journal of Photochemistry and Photobiology A: Chemistry, Vol. 46, Issue 2
  • DOI: 10.1016/1010-6030(89)80012-7

Autonomous discovery in the chemical sciences part I: Progress
journal, September 2019

  • Jensen, Klavs F.; Coley, Connor W.; Eyke, Natalie S.
  • Angewandte Chemie International Edition
  • DOI: 10.1002/anie.201909987

Big-Data Science in Porous Materials: Materials Genomics and Machine Learning
journal, June 2020


Thermally activated delayed fluorescence (TADF) dyes as efficient organic photosensitizers for photocatalytic water reduction
journal, January 2019


Recent advances in organic thermally activated delayed fluorescence materials
journal, January 2017

  • Yang, Zhiyong; Mao, Zhu; Xie, Zongliang
  • Chemical Society Reviews, Vol. 46, Issue 3
  • DOI: 10.1039/C6CS00368K

Combining theory and experiment in electrocatalysis: Insights into materials design
journal, January 2017


Hydrogen evolution from the photolysis of alcoholic benzophenone solutions via redox catalysis
journal, December 1979

  • Graetzel, Carole K.; Graetzel, Michael
  • Journal of the American Chemical Society, Vol. 101, Issue 26
  • DOI: 10.1021/ja00520a032

Porous, Fluorescent, Covalent Triazine-Based Frameworks Via Room-Temperature and Microwave-Assisted Synthesis
journal, April 2012

  • Ren, Shijie; Bojdys, Michael J.; Dawson, Robert
  • Advanced Materials, Vol. 24, Issue 17
  • DOI: 10.1002/adma.201200751

Quantum chemistry structures and properties of 134 kilo molecules
journal, August 2014

  • Ramakrishnan, Raghunathan; Dral, Pavlo O.; Rupp, Matthias
  • Scientific Data, Vol. 1, Issue 1
  • DOI: 10.1038/sdata.2014.22

Machine learning for molecular and materials science
journal, July 2018


Robust and synthesizable photocatalysts for CO2 reduction: a data-driven materials discovery
journal, January 2019

  • Singh, Arunima K.; Montoya, Joseph H.; Gregoire, John M.
  • Nature Communications, Vol. 10, Issue 1
  • DOI: 10.1038/s41467-019-08356-1