Seeing Is Believing: Experimental Spin States from Machine Learning Model Structure Predictions
Determination of ground-state spins of open-shell transition-metal complexes is critical to understanding catalytic and materials properties but also challenging with approximate electronic structure methods. As an alternative approach, we demonstrate how structure alone can be used to guide assignm...
| Main Authors: | , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
American Chemical Society (ACS)
2020
|
| Online Access: | https://hdl.handle.net/1721.1/125686 |
| _version_ | 1826212731647164416 |
|---|---|
| author | Taylor, Michael D. Yang, Tzuhsiung Lin, Sean Nandy, Aditya Janet, Jon Paul Duan, Chenru Kulik, Heather Janine |
| author2 | Massachusetts Institute of Technology. Department of Chemical Engineering |
| author_facet | Massachusetts Institute of Technology. Department of Chemical Engineering Taylor, Michael D. Yang, Tzuhsiung Lin, Sean Nandy, Aditya Janet, Jon Paul Duan, Chenru Kulik, Heather Janine |
| author_sort | Taylor, Michael D. |
| collection | MIT |
| description | Determination of ground-state spins of open-shell transition-metal complexes is critical to understanding catalytic and materials properties but also challenging with approximate electronic structure methods. As an alternative approach, we demonstrate how structure alone can be used to guide assignment of ground-state spin from experimentally determined crystal structures of transition-metal complexes. We first identify the limits of distance-based heuristics from distributions of metal-ligand bond lengths of over 2000 unique mononuclear Fe(II)/Fe(III) transition-metal complexes. To overcome these limits, we employ artificial neural networks (ANNs) to predict spin-state-dependent metal-ligand bond lengths and classify experimental ground-state spins based on agreement of experimental structures with the ANN predictions. Although the ANN is trained on hybrid density functional theory data, we exploit the method-insensitivity of geometric properties to enable assignment of ground states for the majority (ca. 80-90%) of structures. We demonstrate the utility of the ANN by data-mining the literature for spin-crossover (SCO) complexes, which have experimentally observed temperature-dependent geometric structure changes, by correctly assigning almost all (>95%) spin states in the 46 Fe(II) SCO complex set. This approach represents a promising complement to more conventional energy-based spin-state assignment from electronic structure theory at the low cost of a machine learning model. |
| first_indexed | 2024-09-23T15:36:32Z |
| format | Article |
| id | mit-1721.1/125686 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T15:36:32Z |
| publishDate | 2020 |
| publisher | American Chemical Society (ACS) |
| record_format | dspace |
| spelling | mit-1721.1/1256862022-09-29T15:02:12Z Seeing Is Believing: Experimental Spin States from Machine Learning Model Structure Predictions Taylor, Michael D. Yang, Tzuhsiung Lin, Sean Nandy, Aditya Janet, Jon Paul Duan, Chenru Kulik, Heather Janine Massachusetts Institute of Technology. Department of Chemical Engineering Massachusetts Institute of Technology. Department of Materials Science and Engineering Massachusetts Institute of Technology. Department of Chemistry Determination of ground-state spins of open-shell transition-metal complexes is critical to understanding catalytic and materials properties but also challenging with approximate electronic structure methods. As an alternative approach, we demonstrate how structure alone can be used to guide assignment of ground-state spin from experimentally determined crystal structures of transition-metal complexes. We first identify the limits of distance-based heuristics from distributions of metal-ligand bond lengths of over 2000 unique mononuclear Fe(II)/Fe(III) transition-metal complexes. To overcome these limits, we employ artificial neural networks (ANNs) to predict spin-state-dependent metal-ligand bond lengths and classify experimental ground-state spins based on agreement of experimental structures with the ANN predictions. Although the ANN is trained on hybrid density functional theory data, we exploit the method-insensitivity of geometric properties to enable assignment of ground states for the majority (ca. 80-90%) of structures. We demonstrate the utility of the ANN by data-mining the literature for spin-crossover (SCO) complexes, which have experimentally observed temperature-dependent geometric structure changes, by correctly assigning almost all (>95%) spin states in the 46 Fe(II) SCO complex set. This approach represents a promising complement to more conventional energy-based spin-state assignment from electronic structure theory at the low cost of a machine learning model. United States. Defense Advanced Research Projects Agency (Grant D18AP00039) United States. Office of Naval Research (Grant N00014-17-1-2956) United States. Office of Naval Research (Grant N00014-18-1-2434) United States. Department of Energy (Grant DE-SC0018096) National Science Foundation (U.S.) (Grant CBET-1846426) 2020-06-05T15:08:19Z 2020-06-05T15:08:19Z 2020-03 2020-05-18T16:51:50Z Article http://purl.org/eprint/type/JournalArticle 1089-5639 https://hdl.handle.net/1721.1/125686 Taylor, Michael G. et al. “Seeing Is Believing: Experimental Spin States from Machine Learning Model Structure Predictions” The journal of physical chemistry. A, vol. 124, no. 16, 2020, pp. 3286-3299 © 2020 The Author(s) en https://dx.doi.org/10.1021/acs.jpca.0c01458 The journal of physical chemistry. A Creative Commons Attribution 4.0 International license https://creativecommons.org/licenses/by/4.0/ application/pdf American Chemical Society (ACS) ACS |
| spellingShingle | Taylor, Michael D. Yang, Tzuhsiung Lin, Sean Nandy, Aditya Janet, Jon Paul Duan, Chenru Kulik, Heather Janine Seeing Is Believing: Experimental Spin States from Machine Learning Model Structure Predictions |
| title | Seeing Is Believing: Experimental Spin States from Machine Learning Model Structure Predictions |
| title_full | Seeing Is Believing: Experimental Spin States from Machine Learning Model Structure Predictions |
| title_fullStr | Seeing Is Believing: Experimental Spin States from Machine Learning Model Structure Predictions |
| title_full_unstemmed | Seeing Is Believing: Experimental Spin States from Machine Learning Model Structure Predictions |
| title_short | Seeing Is Believing: Experimental Spin States from Machine Learning Model Structure Predictions |
| title_sort | seeing is believing experimental spin states from machine learning model structure predictions |
| url | https://hdl.handle.net/1721.1/125686 |
| work_keys_str_mv | AT taylormichaeld seeingisbelievingexperimentalspinstatesfrommachinelearningmodelstructurepredictions AT yangtzuhsiung seeingisbelievingexperimentalspinstatesfrommachinelearningmodelstructurepredictions AT linsean seeingisbelievingexperimentalspinstatesfrommachinelearningmodelstructurepredictions AT nandyaditya seeingisbelievingexperimentalspinstatesfrommachinelearningmodelstructurepredictions AT janetjonpaul seeingisbelievingexperimentalspinstatesfrommachinelearningmodelstructurepredictions AT duanchenru seeingisbelievingexperimentalspinstatesfrommachinelearningmodelstructurepredictions AT kulikheatherjanine seeingisbelievingexperimentalspinstatesfrommachinelearningmodelstructurepredictions |