Show simple item record

dc.contributor.authorSchran, Christoph
dc.contributor.authorThiemann, Fabian L
dc.contributor.authorRowe, Patrick
dc.contributor.authorMüller, Erich A
dc.contributor.authorMarsalek, Ondrej
dc.contributor.authorMichaelides, Angelos
dc.date.accessioned2021-09-16T23:31:12Z
dc.date.available2021-09-16T23:31:12Z
dc.date.issued2021-09-21
dc.identifier.issn0027-8424
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/328158
dc.description.abstract<jats:p>Simulation techniques based on accurate and efficient representations of potential energy surfaces are urgently needed for the understanding of complex systems such as solid–liquid interfaces. Here we present a machine learning framework that enables the efficient development and validation of models for complex aqueous systems. Instead of trying to deliver a globally optimal machine learning potential, we propose to develop models applicable to specific thermodynamic state points in a simple and user-friendly process. After an initial ab initio simulation, a machine learning potential is constructed with minimum human effort through a data-driven active learning protocol. Such models can afterward be applied in exhaustive simulations to provide reliable answers for the scientific question at hand or to systematically explore the thermal performance of ab initio methods. We showcase this methodology on a diverse set of aqueous systems comprising bulk water with different ions in solution, water on a titanium dioxide surface, and water confined in nanotubes and between molybdenum disulfide sheets. Highlighting the accuracy of our approach with respect to the underlying ab initio reference, the resulting models are evaluated in detail with an automated validation protocol that includes structural and dynamical properties and the precision of the force prediction of the models. Finally, we demonstrate the capabilities of our approach for the description of water on the rutile titanium dioxide (110) surface to analyze the structure and mobility of water on this surface. Such machine learning models provide a straightforward and uncomplicated but accurate extension of simulation time and length scales for complex systems.</jats:p>
dc.languageen
dc.publisherProceedings of the National Academy of Sciences
dc.rightsAttribution-NonCommercial-NoDerivatives 4.0 International
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/4.0/
dc.titleMachine learning potentials for complex aqueous systems made simple
dc.typeArticle
prism.endingPagee2110077118
prism.issueIdentifier38
prism.publicationDate2021
prism.publicationNameProceedings of the National Academy of Sciences
prism.startingPagee2110077118
prism.volume118
dc.identifier.doi10.17863/CAM.75613
dcterms.dateAccepted2021-07-27
rioxxterms.versionofrecord10.1073/pnas.2110077118
rioxxterms.versionVoR
rioxxterms.licenseref.urihttp://www.rioxx.net/licenses/all-rights-reserved
rioxxterms.licenseref.startdate2021-09-21
dc.contributor.orcidSchran, Christoph [0000-0003-4595-5073]
dc.contributor.orcidThiemann, Fabian L [0000-0003-2951-6740]
dc.contributor.orcidMüller, Erich A [0000-0002-1513-6686]
dc.contributor.orcidMarsalek, Ondrej [0000-0002-8624-8837]
dc.contributor.orcidMichaelides, Angelos [0000-0002-9169-169X]
dc.identifier.eissn1091-6490
rioxxterms.typeJournal Article/Review
cam.issuedOnline2021-09-13


Files in this item

Thumbnail
Thumbnail
Thumbnail
Thumbnail
Thumbnail
Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record

Attribution-NonCommercial-NoDerivatives 4.0 International
Except where otherwise noted, this item's licence is described as Attribution-NonCommercial-NoDerivatives 4.0 International