{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T11:06:32Z","timestamp":1780311992081,"version":"3.54.0"},"reference-count":48,"publisher":"AIP Publishing","issue":"21","funder":[{"DOI":"10.13039\/100000001","name":"National Science Foundation","doi-asserted-by":"publisher","award":["CHE-2102583"],"award-info":[{"award-number":["CHE-2102583"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["pubs.aip.org"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2026,6,7]]},"abstract":"<jats:p>Accurate wavefunction methods such as CCSD(T) can predict chemical and thermodynamic properties of small molecules with near-experimental precision. However, their steep computational cost limits their use for large systems or extensive datasets. In contrast, density functional theory (DFT) is faster and more practical for large molecules but often fails to accurately capture electronic changes with quantitative accuracy. To address these limitations, we have developed a machine learning model for the prediction of electron affinities with physics-based structural features (RDKit and SMARTS) and quantum chemistry-based electronic features (Mulliken charge analysis). These features, which capture both local bonding motifs and global electronic context, are used as input descriptors to an XGBR (eXtreme Gradient Boosting Regressor) model in a \u0394ML framework. By embedding QM-based electronic features to the \u0394ML model, our enhanced \u0394ML+ model attains a mean absolute error of 0.03\u00a0eV with respect to G4MP2 values, surpassing conventional chemical accuracy targets while exhibiting markedly reduced dependence on the underlying DFT functional. By analyzing both vertical and adiabatic EAs within the same framework, we highlight the crucial role of geometry relaxation in predictive modeling. Overall, our approach offers an efficient and transferable route to benchmark accurate electron affinity predictions, pointing toward next generation computational protocols that overcome the limitations of standalone DFT.<\/jats:p>","DOI":"10.1063\/5.0334579","type":"journal-article","created":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T10:31:22Z","timestamp":1780309882000},"update-policy":"https:\/\/doi.org\/10.1063\/aip-crossmark-policy-page","source":"Crossref","is-referenced-by-count":0,"title":["Physics-informed machine learning model for accurate prediction of electron affinities"],"prefix":"10.1063","volume":"164","author":[{"ORCID":"https:\/\/orcid.org\/0009-0000-3233-2382","authenticated-orcid":false,"given":"Debashis","family":"Swain","sequence":"first","affiliation":[{"name":"Department of Chemistry, Indiana University , Bloomington, Indiana 47405,","place":["USA"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8496-109X","authenticated-orcid":false,"given":"Surya Sekhar","family":"Manna","sequence":"additional","affiliation":[{"name":"Department of Chemistry, Indiana University , Bloomington, Indiana 47405,","place":["USA"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sarah","family":"Maier","sequence":"additional","affiliation":[{"name":"Department of Chemistry, Indiana University , Bloomington, Indiana 47405,","place":["USA"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3275-1426","authenticated-orcid":false,"given":"Krishnan","family":"Raghavachari","sequence":"additional","affiliation":[{"name":"Department of Chemistry, Indiana University , Bloomington, Indiana 47405,","place":["USA"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"317","published-online":{"date-parts":[[2026,6,1]]},"reference":[{"issue":"1","key":"2026060106311804400_c1","doi-asserted-by":"publisher","first-page":"4870","DOI":"10.1038\/s41467-021-24904-0","article-title":"Teaching a neural network to attach and detach electrons from molecules","volume":"12","year":"2021","journal-title":"Nat. Commun."},{"issue":"10","key":"2026060106311804400_c2","doi-asserted-by":"publisher","first-page":"3684","DOI":"10.1073\/pnas.0800093105","article-title":"DNA binding shifts the redox potential of the transcription factor SoxR","volume":"105","year":"2008","journal-title":"Proc. Natl. Acad. Sci. U. S. A."},{"issue":"12","key":"2026060106311804400_c3","doi-asserted-by":"publisher","first-page":"4701","DOI":"10.1021\/ja00038a037","article-title":"One-electron-reduction potentials of pyrimidine bases, nucleosides, and nucleotides in aqueous solution. Consequences for DNA redox chemistry","volume":"114","year":"1992","journal-title":"J.\u00a0Am. Chem. Soc."},{"issue":"39","key":"2026060106311804400_c4","doi-asserted-by":"publisher","first-page":"25062","DOI":"10.1074\/jbc.273.39.25062","article-title":"Redox potential controls the structure and DNA binding activity of the paired domain","volume":"273","year":"1998","journal-title":"J.\u00a0Biol. Chem."},{"issue":"8\u20139","key":"2026060106311804400_c5","doi-asserted-by":"publisher","first-page":"223","DOI":"10.2166\/wst.1993.0622","article-title":"Binding and mobilization of heavy metals in contaminated sediments affected by pH and redox potential","volume":"28","year":"1993","journal-title":"Water Sci. Technol."},{"key":"2026060106311804400_c6","volume-title":"Environmental Chemistry of Soils","year":"1994"},{"issue":"7","key":"2026060106311804400_c7","doi-asserted-by":"publisher","first-page":"221C","DOI":"10.1149\/1.2115949","article-title":"Redox pathways: Applications in catalysis","volume":"131","year":"1984","journal-title":"J.\u00a0Electrochem. Soc."},{"issue":"22","key":"2026060106311804400_c8","doi-asserted-by":"publisher","first-page":"7083","DOI":"10.1021\/ic049479c","article-title":"Tuning of redox potentials for the design of ruthenium anticancer drugs\u2014An electrochemical study of [trans-RuCl4L(DMSO)]- and [trans-RuCl4L2]- complexes, where L = imidazole, 1,2,4-Triazole, indazole","volume":"43","year":"2004","journal-title":"Inorg. Chem."},{"issue":"9","key":"2026060106311804400_c9","doi-asserted-by":"publisher","first-page":"4430","DOI":"10.1039\/c6ta09760j","article-title":"Designing strategies to tune reduction potential of organic molecules for sustainable high capacity battery application","volume":"5","year":"2017","journal-title":"J.\u00a0Mater. Chem. A"},{"issue":"3","key":"2026060106311804400_c10","doi-asserted-by":"publisher","first-page":"731","DOI":"10.1063\/1.555735","article-title":"Binding energies in atomic negative ions: II","volume":"14","year":"1985","journal-title":"J.\u00a0Phys. Chem. Ref. Data"},{"issue":"4","key":"2026060106311804400_c11","doi-asserted-by":"publisher","first-page":"361","DOI":"10.1255\/ejms.556","article-title":"The gas-phase basicity and proton affinity of 1,3,5-cycloheptatriene\u2014Energetics, structure and interconversion of dihydrotropylium ions","volume":"9","year":"2003","journal-title":"Eur. J. Mass Spectrom."},{"issue":"2","key":"2026060106311804400_c12","doi-asserted-by":"publisher","first-page":"391","DOI":"10.1021\/cr990081l","article-title":"Experimental techniques in gas-phase ion thermochemistry","volume":"101","year":"2001","journal-title":"Chem. Rev."},{"key":"2026060106311804400_c13","volume-title":"Redox: Fundamentals, Processes and Applications","year":"1999"},{"issue":"16","key":"2026060106311804400_c14","doi-asserted-by":"publisher","first-page":"4170","DOI":"10.1002\/anie.201709943","article-title":"Computational chemistry: The fate of current methods and future challenges","volume":"57","year":"2018","journal-title":"Angew. Chem., Int. Ed."},{"issue":"19","key":"2026060106311804400_c15","doi-asserted-by":"publisher","first-page":"2315","DOI":"10.1080\/00268976.2017.1333644","article-title":"Thirty years of density functional theory in computational chemistry: An overview and extensive assessment of 200 density functionals","volume":"115","year":"2017","journal-title":"Mol. Phys."},{"issue":"9","key":"2026060106311804400_c16","doi-asserted-by":"publisher","first-page":"4129","DOI":"10.1063\/1.1638736","article-title":"W3 theory: Robust computational thermochemistry in the kJ\/mol accuracy range","volume":"120","year":"2004","journal-title":"J.\u00a0Chem. Phys."},{"issue":"1-3","key":"2026060106311804400_c17","doi-asserted-by":"publisher","first-page":"106","DOI":"10.1007\/s00214-005-0037-5","article-title":"Coupled cluster theory with emphasis on selected new developments","volume":"116","year":"2006","journal-title":"Theor. Chem. Acc."},{"issue":"6","key":"2026060106311804400_c18","doi-asserted-by":"publisher","first-page":"479","DOI":"10.1016\/s0009-2614(89)87395-6","article-title":"A fifth-order perturbation comparison of electron correlation theories","volume":"157","year":"1989","journal-title":"Chem. Phys. Lett."},{"issue":"3","key":"2026060106311804400_c19","doi-asserted-by":"publisher","first-page":"712","DOI":"10.1021\/acs.jcim.3c01923","article-title":"Leveraging DFT and molecular fragmentation for chemically accurate pKa prediction using machine learning","volume":"64","year":"2024","journal-title":"J.\u00a0Chem. Inf. Model."},{"issue":"1","key":"2026060106311804400_c20","doi-asserted-by":"publisher","first-page":"19375","DOI":"10.1038\/srep19375","article-title":"Machine learning bandgaps of double perovskites","volume":"6","year":"2016","journal-title":"Sci. Rep."},{"issue":"7715","key":"2026060106311804400_c21","doi-asserted-by":"publisher","first-page":"547","DOI":"10.1038\/s41586-018-0337-2","article-title":"Machine learning for molecular and materials science","volume":"559","year":"2018","journal-title":"Nature"},{"issue":"24","key":"2026060106311804400_c22","doi-asserted-by":"publisher","first-page":"241722","DOI":"10.1063\/1.5019779","article-title":"SchNet\u2014A deep learning architecture for molecules and materials","volume":"148","year":"2018","journal-title":"J.\u00a0Chem. Phys."},{"issue":"8","key":"2026060106311804400_c23","doi-asserted-by":"publisher","first-page":"eaav6490","DOI":"10.1126\/sciadv.aav6490","article-title":"Accurate and transferable multitask prediction of chemical properties with an atoms-in-molecules neural network","volume":"5","year":"2019","journal-title":"Sci. Adv."},{"issue":"1","key":"2026060106311804400_c24","doi-asserted-by":"publisher","first-page":"4895","DOI":"10.1038\/s41467-020-18556-9","article-title":"Retrospective on a decade of machine learning for chemical discovery","volume":"11","year":"2020","journal-title":"Nat. Commun."},{"issue":"32","key":"2026060106311804400_c25","doi-asserted-by":"publisher","first-page":"7761","DOI":"10.1021\/acs.jpclett.1c02042","article-title":"Machine-learned energy functionals for multiconfigurational wave functions","volume":"12","year":"2021","journal-title":"J.\u00a0Phys. Chem. Lett."},{"issue":"5","key":"2026060106311804400_c26","doi-asserted-by":"publisher","first-page":"2087","DOI":"10.1021\/acs.jctc.5b00099","article-title":"Big data meets quantum chemistry approximations: The \u0394-machine learning approach","volume":"11","year":"2015","journal-title":"J.\u00a0Chem. Theory Comput."},{"issue":"8","key":"2026060106311804400_c27","doi-asserted-by":"publisher","first-page":"4846","DOI":"10.1021\/acs.jctc.2c00501","article-title":"Machine learning of coupled cluster (T)-energy corrections via delta (\u0394)-learning","volume":"18","year":"2022","journal-title":"J.\u00a0Chem. Theory Comput."},{"issue":"10","key":"2026060106311804400_c28","doi-asserted-by":"publisher","first-page":"5614","DOI":"10.1021\/acs.jctc.9b00465","article-title":"Toward fast and reliable potential energy surfaces for metallic Pt clusters by hierarchical delta neural networks","volume":"15","year":"2019","journal-title":"J.\u00a0Chem. Theory Comput."},{"issue":"3","key":"2026060106311804400_c29","doi-asserted-by":"publisher","first-page":"1546","DOI":"10.1021\/acs.jctc.8b00832","article-title":"Boosting quantum machine learning models with a multilevel combination technique: Pople diagrams revisited","volume":"15","year":"2019","journal-title":"J.\u00a0Chem. Theory Comput."},{"issue":"19","key":"2026060106311804400_c30","doi-asserted-by":"publisher","first-page":"6648","DOI":"10.1073\/pnas.0408036102","article-title":"Ab initio quantum chemistry: Methodology and applications","volume":"102","year":"2005","journal-title":"Proc. Natl. Acad. Sci. U. S. A."},{"issue":"37","key":"2026060106311804400_c31","doi-asserted-by":"publisher","first-page":"11011","DOI":"10.1002\/anie.201611308","article-title":"Benchmarking quantum chemical methods: Are we heading in the right direction?","volume":"56","year":"2017","journal-title":"Angew. Chem., Int. Ed."},{"issue":"31","key":"2026060106311804400_c32","doi-asserted-by":"publisher","first-page":"6872","DOI":"10.1021\/acs.jpca.1c06152","article-title":"A fragmentation-based graph embedding framework for QM\/ML","volume":"125","year":"2021","journal-title":"J.\u00a0Phys. Chem. A"},{"issue":"15","key":"2026060106311804400_c33","doi-asserted-by":"publisher","first-page":"3472","DOI":"10.1021\/acs.jpca.2c08821","article-title":"Quantitative prediction of vertical ionization potentials from DFT via a graph-network-based Delta machine learning model incorporating electronic descriptors","volume":"127","year":"2023","journal-title":"J.\u00a0Phys. Chem. A"},{"issue":"11","key":"2026060106311804400_c34","doi-asserted-by":"publisher","first-page":"1699","DOI":"10.1515\/pac-2025-0574","article-title":"Combining molecular fragmentation and machine learning for accurate prediction of adiabatic ionization potentials","volume":"97","year":"2025","journal-title":"Pure Appl. Chem."},{"issue":"4","key":"2026060106311804400_c35","doi-asserted-by":"publisher","first-page":"2109","DOI":"10.1021\/acs.jctc.9b00875","article-title":"Accurate quantum chemical calculation of ionization potentials: Validation of the DFT-LOC approach via a large data set obtained from experiments and benchmark quantum chemical calculations","volume":"16","year":"2020","journal-title":"J.\u00a0Chem. Theory Comput."},{"issue":"9","key":"2026060106311804400_c36","doi-asserted-by":"publisher","first-page":"095003","DOI":"10.1088\/1367-2630\/15\/9\/095003","article-title":"Machine learning of molecular electronic properties in chemical compound space","volume":"15","year":"2013","journal-title":"New J. Phys."},{"issue":"16","key":"2026060106311804400_c37","doi-asserted-by":"publisher","first-page":"e202300036","DOI":"10.1002\/slct.202300036","article-title":"Machine learning for the prediction of ionization potential and electron affinity energies obtained by density functional theory","volume":"8","year":"2023","journal-title":"ChemistrySelect"},{"issue":"5","key":"2026060106311804400_c38","doi-asserted-by":"publisher","first-page":"810","DOI":"10.1002\/wcms.59","article-title":"Gn theory","volume":"1","year":"2011","journal-title":"WIREs Comput. Mol. Sci."},{"issue":"25","key":"2026060106311804400_c39","doi-asserted-by":"publisher","first-page":"8732","DOI":"10.1021\/ja902302h","article-title":"970 million druglike small molecules for virtual screening in the chemical universe database GDB-13","volume":"131","year":"2009","journal-title":"J.\u00a0Am. Chem. Soc."},{"key":"2026060106311804400_c40","volume-title":"Gaussview 5.0","author":"Gaussian 16, Revision B.01","year":"2016"},{"issue":"8","key":"2026060106311804400_c41","doi-asserted-by":"publisher","first-page":"084108","DOI":"10.1063\/1.2436888","article-title":"Gaussian-4 theory","volume":"126","year":"2007","journal-title":"J.\u00a0Chem. Phys."},{"issue":"15","key":"2026060106311804400_c42","doi-asserted-by":"publisher","first-page":"154104","DOI":"10.1063\/1.3382344","article-title":"A consistent and accurate ab\u00a0initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu","volume":"132","year":"2010","journal-title":"J.\u00a0Chem. Phys."},{"issue":"7","key":"2026060106311804400_c43","doi-asserted-by":"publisher","first-page":"1456","DOI":"10.1002\/jcc.21759","article-title":"Effect of the damping function in dispersion corrected density functional theory","volume":"32","year":"2011","journal-title":"J.\u00a0Comput. Chem."},{"issue":"2","key":"2026060106311804400_c44","doi-asserted-by":"publisher","first-page":"785","DOI":"10.1103\/physrevb.37.785","article-title":"Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density","volume":"37","year":"1988","journal-title":"Phys. Rev. B"},{"issue":"41","key":"2026060106311804400_c45","doi-asserted-by":"publisher","first-page":"10257","DOI":"10.1073\/pnas.1810421115","article-title":"Revised M06 density functional for main-group and transition-metal chemistry","volume":"115","year":"2018","journal-title":"Proc. Natl. Acad. Sci. U. S. A."},{"key":"2026060106311804400_c46","first-page":"785","article-title":"XGBoost: A scalable tree boosting system","year":"2016"},{"issue":"32","key":"2026060106311804400_c47","doi-asserted-by":"publisher","first-page":"42138","DOI":"10.1021\/acsami.4c06243","article-title":"Integrated supervised and unsupervised machine learning approach to map the electrochemical windows over 4500 solvents for battery applications","volume":"16","year":"2024","journal-title":"ACS Appl. Mater. Interfaces"},{"issue":"7","key":"2026060106311804400_c48","doi-asserted-by":"publisher","first-page":"3191","DOI":"10.1021\/acs.chemmater.3c02905","article-title":"Machine learning-driven ionic liquids as electrolytes for the advancement of high-voltage dual-ion battery","volume":"36","year":"2024","journal-title":"Chem. Mater."}],"container-title":["The Journal of Chemical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pubs.aip.org\/aip\/jcp\/article-pdf\/doi\/10.1063\/5.0334579\/21021677\/214108_1_5.0334579.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/pubs.aip.org\/aip\/jcp\/article-pdf\/doi\/10.1063\/5.0334579\/21021677\/214108_1_5.0334579.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T10:31:24Z","timestamp":1780309884000},"score":1,"resource":{"primary":{"URL":"https:\/\/pubs.aip.org\/jcp\/article\/164\/21\/214108\/3393116\/Physics-informed-machine-learning-model-for"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,1]]},"references-count":48,"journal-issue":{"issue":"21","published-print":{"date-parts":[[2026,6,7]]}},"URL":"https:\/\/doi.org\/10.1063\/5.0334579","relation":{},"ISSN":["0021-9606","1089-7690"],"issn-type":[{"value":"0021-9606","type":"print"},{"value":"1089-7690","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2026,6,7]]},"published":{"date-parts":[[2026,6,1]]},"article-number":"214108"}}