Accurate energy barriers for catalytic reaction pathways: an automatic training protocol for machine learning force fields
Abstract We introduce a training protocol for developing machine learning force fields (MLFFs), capable of accurately determining energy barriers in catalytic reaction pathways. The protocol is validated on the extensively explored hydrogenation of carbon dioxide to methanol over indium oxide. With...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-10-01
|
| Series: | npj Computational Materials |
| Online Access: | https://doi.org/10.1038/s41524-023-01124-2 |
| _version_ | 1797452079188410368 |
|---|---|
| author | Lars L. Schaaf Edvin Fako Sandip De Ansgar Schäfer Gábor Csányi |
| author_facet | Lars L. Schaaf Edvin Fako Sandip De Ansgar Schäfer Gábor Csányi |
| author_sort | Lars L. Schaaf |
| collection | DOAJ |
| description | Abstract We introduce a training protocol for developing machine learning force fields (MLFFs), capable of accurately determining energy barriers in catalytic reaction pathways. The protocol is validated on the extensively explored hydrogenation of carbon dioxide to methanol over indium oxide. With the help of active learning, the final force field obtains energy barriers within 0.05 eV of Density Functional Theory. Thanks to the computational speedup, not only do we reduce the cost of routine in-silico catalytic tasks, but also find an alternative path for the previously established rate-limiting step, with a 40% reduction in activation energy. Furthermore, we illustrate the importance of finite temperature effects and compute free energy barriers. The transferability of the protocol is demonstrated on the experimentally relevant, yet unexplored, top-layer reduced indium oxide surface. The ability of MLFFs to enhance our understanding of extensively studied catalysts underscores the need for fast and accurate alternatives to direct ab-initio simulations. |
| first_indexed | 2024-03-09T15:03:35Z |
| format | Article |
| id | doaj.art-063fc02329424cd6a8bc5ef70ff95b77 |
| institution | Directory Open Access Journal |
| issn | 2057-3960 |
| language | English |
| last_indexed | 2024-03-09T15:03:35Z |
| publishDate | 2023-10-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | npj Computational Materials |
| spelling | doaj.art-063fc02329424cd6a8bc5ef70ff95b772023-11-26T13:47:04ZengNature Portfolionpj Computational Materials2057-39602023-10-019111010.1038/s41524-023-01124-2Accurate energy barriers for catalytic reaction pathways: an automatic training protocol for machine learning force fieldsLars L. Schaaf0Edvin Fako1Sandip De2Ansgar Schäfer3Gábor Csányi4Engineering Laboratory, University of CambridgeBASF SEBASF SEBASF SEEngineering Laboratory, University of CambridgeAbstract We introduce a training protocol for developing machine learning force fields (MLFFs), capable of accurately determining energy barriers in catalytic reaction pathways. The protocol is validated on the extensively explored hydrogenation of carbon dioxide to methanol over indium oxide. With the help of active learning, the final force field obtains energy barriers within 0.05 eV of Density Functional Theory. Thanks to the computational speedup, not only do we reduce the cost of routine in-silico catalytic tasks, but also find an alternative path for the previously established rate-limiting step, with a 40% reduction in activation energy. Furthermore, we illustrate the importance of finite temperature effects and compute free energy barriers. The transferability of the protocol is demonstrated on the experimentally relevant, yet unexplored, top-layer reduced indium oxide surface. The ability of MLFFs to enhance our understanding of extensively studied catalysts underscores the need for fast and accurate alternatives to direct ab-initio simulations.https://doi.org/10.1038/s41524-023-01124-2 |
| spellingShingle | Lars L. Schaaf Edvin Fako Sandip De Ansgar Schäfer Gábor Csányi Accurate energy barriers for catalytic reaction pathways: an automatic training protocol for machine learning force fields npj Computational Materials |
| title | Accurate energy barriers for catalytic reaction pathways: an automatic training protocol for machine learning force fields |
| title_full | Accurate energy barriers for catalytic reaction pathways: an automatic training protocol for machine learning force fields |
| title_fullStr | Accurate energy barriers for catalytic reaction pathways: an automatic training protocol for machine learning force fields |
| title_full_unstemmed | Accurate energy barriers for catalytic reaction pathways: an automatic training protocol for machine learning force fields |
| title_short | Accurate energy barriers for catalytic reaction pathways: an automatic training protocol for machine learning force fields |
| title_sort | accurate energy barriers for catalytic reaction pathways an automatic training protocol for machine learning force fields |
| url | https://doi.org/10.1038/s41524-023-01124-2 |
| work_keys_str_mv | AT larslschaaf accurateenergybarriersforcatalyticreactionpathwaysanautomatictrainingprotocolformachinelearningforcefields AT edvinfako accurateenergybarriersforcatalyticreactionpathwaysanautomatictrainingprotocolformachinelearningforcefields AT sandipde accurateenergybarriersforcatalyticreactionpathwaysanautomatictrainingprotocolformachinelearningforcefields AT ansgarschafer accurateenergybarriersforcatalyticreactionpathwaysanautomatictrainingprotocolformachinelearningforcefields AT gaborcsanyi accurateenergybarriersforcatalyticreactionpathwaysanautomatictrainingprotocolformachinelearningforcefields |