Search Menu
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information
Search Scholarly Publications

Title: Fast predictions of liquid-phase acid-catalyzed reaction rates using molecular dynamics simulations and convolutional neural networks

Abstract

The rates of liquid-phase, acid-catalyzed reactions relevant to the upgrading of biomass into high-value chemicals are highly sensitive to solvent composition and identifying suitable solvent mixtures is theoretically and experimentally challenging. We show that the complex atomistic configurations of reactant–solvent environments generated by classical molecular dynamics simulations can be exploited by 3D convolutional neural networks to enable accurate predictions of Brønsted acid-catalyzed reaction rates for model biomass compounds. We develop a 3D convolutional neural network, which we call SolventNet, and train it to predict acid-catalyzed reaction rates using experimental reaction data and corresponding molecular dynamics simulation data for seven biomass-derived oxygenates in water–cosolvent mixtures. We show that SolventNet can predict reaction rates for additional reactants and solvent systems an order of magnitude faster than prior simulation methods. This combination of machine learning with molecular dynamics enables the rapid, high-throughput screening of solvent systems and identification of improved biomass conversion conditions.

Authors:
ORCiD logo [1]; ORCiD logo [2]; ORCiD logo [2]; ORCiD logo [2]; ORCiD logo [1]
+ Show Author Affiliations
  1. Department of Chemical and Biological Engineering, University of Wisconsin-Madison, Madison, USA, DOE Great Lakes Bioenergy Research Center
  2. Department of Chemical and Biological Engineering, University of Wisconsin-Madison, Madison, USA
Publication Date:
Research Org.:
Great Lakes Bioenergy Research Center (GLBRC), Madison, WI (United States)
Sponsoring Org.:
USDOE Office of Science (SC), Biological and Environmental Research (BER); National Science Foundation (NSF)
Contributing Org.:
UW-Madison Center for High Throughput Computing (CHTC)
OSTI Identifier:
1686180
Alternate Identifier(s):
OSTI ID: 1684636
Grant/Contract Number:  
SC0018409; ACI-1549562
Resource Type:
Published Article
Journal Name:
Chemical Science
Additional Journal Information:
Journal Name: Chemical Science Journal Volume: 11 Journal Issue: 46; Journal ID: ISSN 2041-6520
Publisher:
Royal Society of Chemistry
Country of Publication:
United Kingdom
Language:
English
Subject:
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CHEMISTRY

Citation Formats

Chew, Alex K., Jiang, Shengli, Zhang, Weiqi, Zavala, Victor M., and Van Lehn, Reid C. Fast predictions of liquid-phase acid-catalyzed reaction rates using molecular dynamics simulations and convolutional neural networks. United Kingdom: N. p., 2020. Web. doi:10.1039/D0SC03261A.
Copy to clipboard
Chew, Alex K., Jiang, Shengli, Zhang, Weiqi, Zavala, Victor M., & Van Lehn, Reid C. Fast predictions of liquid-phase acid-catalyzed reaction rates using molecular dynamics simulations and convolutional neural networks. United Kingdom. https://doi.org/10.1039/D0SC03261A
Copy to clipboard
Chew, Alex K., Jiang, Shengli, Zhang, Weiqi, Zavala, Victor M., and Van Lehn, Reid C. Mon . "Fast predictions of liquid-phase acid-catalyzed reaction rates using molecular dynamics simulations and convolutional neural networks". United Kingdom. https://doi.org/10.1039/D0SC03261A.
Copy to clipboard
@article{osti_1686180,
title = {Fast predictions of liquid-phase acid-catalyzed reaction rates using molecular dynamics simulations and convolutional neural networks},
author = {Chew, Alex K. and Jiang, Shengli and Zhang, Weiqi and Zavala, Victor M. and Van Lehn, Reid C.},
abstractNote = {The rates of liquid-phase, acid-catalyzed reactions relevant to the upgrading of biomass into high-value chemicals are highly sensitive to solvent composition and identifying suitable solvent mixtures is theoretically and experimentally challenging. We show that the complex atomistic configurations of reactant–solvent environments generated by classical molecular dynamics simulations can be exploited by 3D convolutional neural networks to enable accurate predictions of Brønsted acid-catalyzed reaction rates for model biomass compounds. We develop a 3D convolutional neural network, which we call SolventNet, and train it to predict acid-catalyzed reaction rates using experimental reaction data and corresponding molecular dynamics simulation data for seven biomass-derived oxygenates in water–cosolvent mixtures. We show that SolventNet can predict reaction rates for additional reactants and solvent systems an order of magnitude faster than prior simulation methods. This combination of machine learning with molecular dynamics enables the rapid, high-throughput screening of solvent systems and identification of improved biomass conversion conditions.},
doi = {10.1039/D0SC03261A},
journal = {Chemical Science},
number = 46,
volume = 11,
place = {United Kingdom},
year = {Mon Oct 19 00:00:00 EDT 2020},
month = {Mon Oct 19 00:00:00 EDT 2020}
}
Copy to clipboard
Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record
https://doi.org/10.1039/D0SC03261A
Copyright Statement
Other availability
Search WorldCat Search WorldCat to find libraries that may hold this journal

Figures / Tables:

Fig. 1 Fig. 1: Overview of solvent effects on acid-catalyzed reactions and model systems. (a) Two example acid-catalyzed reactions: xylitol (XYL) dehydration and levoglucosan (LGA) hydrolysis. (b) Hypothesized effect of mixed-solvent environments on the free energy landscape of acid-catalyzed reactions. The schematic illustrates the formation of a local solvent domain (within themore » circular dashed line) around the reactant in a mixed-solvent environment that modifies the reaction free energy landscape, thus affecting reaction kinetics. (c) Organic, polar aprotic cosolvents modeled in this study, including dioxane (DIO), $γ$-valerolactone (GVL), tetrahydrofuran (THF), dimethyl sulfoxide (DMSO), acetonitrile (MeCN), and acetone (ACE). Molecules drawn in black were included in the training set. Molecules drawn in gray were included in the test set. (d) Biomass-derived model reactants modeled in this study, including ethyl tert-butyl ether (ETBE), tert-butanol (TBA), cellobiose (CEL), glucose (GLU), LGA, 1,2-propanediol (PDO), fructose (FRU), and XYL. The color scheme follows part (c), except TBA, PDO, and FRU were included as part of some of the reactant–solvent combinations in both training and test sets.« less
All figures and tables (7 total)
Save / Share:
Export Metadata
Save to My Library
You must Sign In or Create an Account in order to save documents to your library.

Works referenced in this record:

Organic Solvent Effects in Biomass Conversion Reactions
journal, December 2015

Tackling Exascale Software Challenges in Molecular Dynamics Simulations with GROMACS
book, January 2015

Solvent Effects in Acid-Catalyzed Biomass Conversion Reactions
journal, September 2014

  • Mellmer, Max A.; Sener, Canan; Gallo, Jean Marcel R.
  • Angewandte Chemie International Edition, Vol. 53, Issue 44
  • DOI: 10.1002/anie.201408359

High precision protein functional site detection using 3D convolutional neural networks
journal, September 2018

LINCS: A linear constraint solver for molecular simulations
journal, September 1997

A review of semantic segmentation using deep neural networks
journal, November 2017

  • Guo, Yanming; Liu, Yu; Georgiou, Theodoros
  • International Journal of Multimedia Information Retrieval, Vol. 7, Issue 2
  • DOI: 10.1007/s13735-017-0141-z

Understanding solvent effects in the selective conversion of fructose to 5-hydroxymethyl-furfural: a molecular dynamics investigation
journal, January 2012

  • Mushrif, Samir H.; Caratzoulas, Stavros; Vlachos, Dionisios G.
  • Physical Chemistry Chemical Physics, Vol. 14, Issue 8
  • DOI: 10.1039/c2cp22694d

Effects of chloride ions in acid-catalyzed biomass dehydration reactions in polar aprotic solvents
journal, March 2019

  • Mellmer, Max A.; Sanpitakseree, Chotitath; Demir, Benginur
  • Nature Communications, Vol. 10, Issue 1
  • DOI: 10.1038/s41467-019-09090-4

Origin of 5-Hydroxymethylfurfural Stability in Water/Dimethyl Sulfoxide Mixtures
journal, January 2014

  • Tsilomelekis, George; Josephson, Tyler R.; Nikolakis, Vladimiros
  • ChemSusChem, Vol. 7, Issue 1
  • DOI: 10.1002/cssc.201300786

Converting fructose to 5-hydroxymethylfurfural: a quantum mechanics/molecular mechanics study of the mechanism and energetics
journal, April 2011

Increasing the revenue from lignocellulosic biomass: Maximizing feedstock utilization
journal, May 2017

  • Alonso, David Martin; Hakim, Sikander H.; Zhou, Shengfei
  • Science Advances, Vol. 3, Issue 5
  • DOI: 10.1126/sciadv.1603301

Optimization of the Additive CHARMM All-Atom Protein Force Field Targeting Improved Sampling of the Backbone ϕ, ψ and Side-Chain χ 1 and χ 2 Dihedral Angles
journal, August 2012

  • Best, Robert B.; Zhu, Xiao; Shim, Jihyun
  • Journal of Chemical Theory and Computation, Vol. 8, Issue 9
  • DOI: 10.1021/ct300400x

K DEEP : Protein–Ligand Absolute Binding Affinity Prediction via 3D-Convolutional Neural Networks
journal, January 2018

  • Jiménez, José; Škalič, Miha; Martínez-Rosell, Gerard
  • Journal of Chemical Information and Modeling, Vol. 58, Issue 2
  • DOI: 10.1021/acs.jcim.7b00650

Effect of Mixed-Solvent Environments on the Selectivity of Acid-Catalyzed Dehydration Reactions
journal, December 2019

Deep convolutional networks for quality assessment of protein folds
journal, June 2018

ImageNet Large Scale Visual Recognition Challenge
journal, April 2015

  • Russakovsky, Olga; Deng, Jia; Su, Hao
  • International Journal of Computer Vision, Vol. 115, Issue 3
  • DOI: 10.1007/s11263-015-0816-y

Solvent system for effective near-term production of hydroxymethylfurfural (HMF) with potential for long-term process improvement
journal, January 2019

  • Motagamwala, Ali Hussain; Huang, Kefeng; Maravelias, Christos T.
  • Energy & Environmental Science, Vol. 12, Issue 7
  • DOI: 10.1039/C9EE00447E

Systematic parameterization of lignin for the CHARMM force field
journal, January 2019

  • Vermaas, Josh V.; Petridis, Loukas; Ralph, John
  • Green Chemistry, Vol. 21, Issue 1
  • DOI: 10.1039/C8GC03209B

Production of levoglucosenone and 5-hydroxymethylfurfural from cellulose in polar aprotic solvent–water mixtures
journal, January 2017

  • He, Jiayue; Liu, Mingjie; Huang, Kefeng
  • Green Chemistry, Vol. 19, Issue 15
  • DOI: 10.1039/C7GC01688C

A graph-convolutional neural network model for the prediction of chemical reactivity
journal, January 2019

  • Coley, Connor W.; Jin, Wengong; Rogers, Luke
  • Chemical Science, Vol. 10, Issue 2
  • DOI: 10.1039/C8SC04228D

CHARMM general force field: A force field for drug-like molecules compatible with the CHARMM all-atom additive biological force fields
journal, January 2009

  • Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.
  • Journal of Computational Chemistry
  • DOI: 10.1002/jcc.21367

Extension of the CHARMM general force field to sulfonyl-containing compounds and its utility in biomolecular simulations
journal, July 2012

  • Yu, Wenbo; He, Xibing; Vanommeslaeghe, Kenno
  • Journal of Computational Chemistry, Vol. 33, Issue 31
  • DOI: 10.1002/jcc.23067

Electronic structure at coarse-grained resolutions from supervised machine learning
journal, March 2019

  • Jackson, Nicholas E.; Bowen, Alec S.; Antony, Lucas W.
  • Science Advances, Vol. 5, Issue 3
  • DOI: 10.1126/sciadv.aav1190

Deep Convolutional Neural Networks for Image Classification: A Comprehensive Review
journal, September 2017

3D convolutional neural network for object recognition: a review
journal, December 2018

  • Singh, Rahul Dev; Mittal, Ajay; Bhatia, Rajesh K.
  • Multimedia Tools and Applications, Vol. 78, Issue 12
  • DOI: 10.1007/s11042-018-6912-6

Solvation dynamics and energetics of intramolecular hydride transfer reactions in biomass conversion
journal, January 2015

  • Mushrif, Samir H.; Varghese, Jithin J.; Krishnamurthy, Chethana B.
  • Physical Chemistry Chemical Physics, Vol. 17, Issue 7
  • DOI: 10.1039/C4CP05063K

Mapping membrane activity in undiscovered peptide sequence space using machine learning
journal, November 2016

  • Lee, Ernest Y.; Fulan, Benjamin M.; Wong, Gerard C. L.
  • Proceedings of the National Academy of Sciences, Vol. 113, Issue 48
  • DOI: 10.1073/pnas.1609893113

Machine learning of accurate energy-conserving molecular force fields
journal, May 2017

  • Chmiela, Stefan; Tkatchenko, Alexandre; Sauceda, Huziel E.
  • Science Advances, Vol. 3, Issue 5
  • DOI: 10.1126/sciadv.1603015

Quantifying the Stability of the Hydronium Ion in Organic Solvents With Molecular Dynamics Simulations
journal, June 2019

Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules
journal, January 2018

  • Gómez-Bombarelli, Rafael; Wei, Jennifer N.; Duvenaud, David
  • ACS Central Science, Vol. 4, Issue 2
  • DOI: 10.1021/acscentsci.7b00572

A Multifunctional Cosolvent Pair Reveals Molecular Principles of Biomass Deconstruction
journal, July 2019

  • Patri, Abhishek S.; Mostofian, Barmak; Pu, Yunqiao
  • Journal of the American Chemical Society, Vol. 141, Issue 32
  • DOI: 10.1021/jacs.8b10242

The missing term in effective pair potentials
journal, November 1987

  • Berendsen, H. J. C.; Grigera, J. R.; Straatsma, T. P.
  • The Journal of Physical Chemistry, Vol. 91, Issue 24
  • DOI: 10.1021/j100308a038

Solvent-enabled control of reactivity for liquid-phase reactions of biomass-derived compounds
journal, February 2018

  • Mellmer, Max A.; Sanpitakseree, Chotitath; Demir, Benginur
  • Nature Catalysis, Vol. 1, Issue 3
  • DOI: 10.1038/s41929-018-0027-3

DeepSite: protein-binding site predictor using 3D-convolutional neural networks
journal, May 2017

    Sort by:
    [ × clear filter / sort ]

    Figures / Tables found in this record:

    Fig. 1 (p. 2)figure
    Fig. 2 (p. 4)figure
    Fig. 3 (p. 5)figure
    Fig. 4 (p. 6)figure
    Fig. 5 (p. 7)figure
    Fig. 6 (p. 8)figure
    Fig. 7 (p. 9)figure
      [ × clear filter / sort ]
      Figures/Tables have been extracted from DOE-funded journal article accepted manuscripts.

      Similar Records in DOE PAGES and OSTI.GOV collections: