Efficient Machine Learning Prediction of Solvent‐Dependent 1H NMR Chemical Shifts in Zinc Complexes
Journal of Computational Chemistry · 15.04.2026 · 47 (10)
DOI: 10.1002/jcc.70368 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Computational Chemistry, online erschienen am 12.04.2026. Verfasst von Jyothika R Pillay und 5 weiteren Autor:innen. Beteiligte Einrichtung: Institute of Physical Chemistry (IPC), Friedrich Schiller University Jena, Jena, Germany. Die Arbeit nennt 102 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Deutsche Forschungsgemeinschaft, Thüringer Aufbaubank.
Abstract (Originalsprache)
Accurate prediction of NMR chemical shifts in transition metal complexes remains challenging due to the wide range of coordination environments and complex electronic structures of these systems. In this work, we present a machine learning approach (ML) for rapid and accurate prediction of NMR shifts in zinc complexes across multiple solvent environments. We systematically selected a diverse set of zinc complexes from the transition metal quantum mechanics (tmQM) database using K‐means clustering on SOAP descriptors, and performed DFT NMR calculations across five solvents to generate training data. We combine smooth overlap of atomic positions (SOAP) descriptors with tree‐based ensemble methods to predict proton chemical shifts. Among several ML algorithms evaluated, LightGBM achieved the best performance on held‐out test complexes (MAE = 0.016 ppm, RMSE = 0.028 ppm, = 0.99), demonstrating excellent generalization to unseen molecular structures. External validation against experimental NMR data across multiple solvents revealed strong predictive performance ( = 0.90, MAE = 0.56 ppm), with exceptional accuracy in methanol ( = 0.96) and acetonitrile ( = 0.91). Notably, the model demonstrated robust transferability to acetonitrile despite this solvent not being included in the training set. This approach provides a computationally efficient alternative to expensive quantum chemical calculations for predicting NMR shifts in transition metal complexes, offering prediction times that are orders of magnitude faster while maintaining accuracy comparable to DFT methods, potentially accelerating the characterization and design of organometallic compounds.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Institute of Physical Chemistry (IPC), Friedrich Schiller University Jena, Jena, Germany.
- Institute for Organic and Macromolecular Chemistry (IOMC), Friedrich Schiller University Jena, Jena, Germany.
- Jena Center for Soft Matter (JCSM), Friedrich Schiller University Jena, Jena, Germany.
- Helmholtz Institute for Polymers in Energy Applications (HIPOLE Jena), Jena, Germany.
- Helmholtz-Zentrum Berlin Für Materialien und Energie (HZB), Berlin, Germany.
- Center for Energy and Environmental Chemistry Jena (CEEC Jena), Friedrich Schiller University Jena, Jena, Germany.
ORCID-Profile: Michael Ringleb, Stefanie Gräfe
Förderung
- Deutsche Forschungsgemeinschaft
- Thüringer Aufbaubank
Zitieren
APA
Pillay, J. R., Ringleb, M., Croy, A., Zechel, S., Schubert, U. S., & Gräfe, S. (2026). Efficient Machine Learning Prediction of Solvent‐Dependent 1H NMR Chemical Shifts in Zinc Complexes. Journal of Computational Chemistry, 47(10). https://doi.org/10.1002/jcc.70368
Vancouver
Pillay JR, Ringleb M, Croy A, Zechel S, Schubert US, Gräfe S. Efficient Machine Learning Prediction of Solvent‐Dependent 1H NMR Chemical Shifts in Zinc Complexes. Journal of Computational Chemistry. 2026;47(10). doi:10.1002/jcc.70368
BibTeX
@article{Pillay2026,
author = {Pillay, Jyothika R and Ringleb, Michael and Croy, Alexander and Zechel, Stefan and Schubert, Ulrich S and Gräfe, Stefanie},
title = {Efficient Machine Learning Prediction of Solvent‐Dependent 1H NMR Chemical Shifts in Zinc Complexes},
journal = {Journal of Computational Chemistry},
year = {2026},
volume = {47},
number = {10},
doi = {10.1002/jcc.70368}
}
EUVIMED