EUVIMEDEUVIMED entdecken
Crossref

Machine learning prediction of organic compound melting points informed by condensed-phase and electronic descriptors

Frank T Mtetwa iD, Neil F Giles iD, W Vincent Wilding iD, Thomas A Knotts iD

The Journal of Chemical Physics · 21.08.2026 · 165 (7)

DOI: 10.1063/5.0342619 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The Journal of Chemical Physics, online erschienen am 17.08.2026. Verfasst von Frank T Mtetwa und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Chemical and Biological Engineering, Brigham Young University, Provo, Utah 84602, USA. Die Arbeit nennt 68 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Melting point (MP) is an important thermophysical property for the chemical process industry, yet accurate prediction of MP for organic compounds in the absence of experimental data remains challenging due to the complex interplay between molecular packing, intermolecular interactions, and electronic structure. Traditional group contribution and quantitative structure–property relationship models, which rely primarily on static molecular descriptors, often fail to capture these critical condensed-phase effects. In this study, we present a hybrid machine learning framework that integrates cheminformatics descriptors with quantum chemical features and dynamic condensed-phase descriptors derived from molecular dynamics (MD) simulations. Using a curated subset of the DIPPR 801 database, multiple machine learning architectures, including light gradient boosting machine (LightGBM) and graph convolutional networks, were evaluated with feature sets of increasing physical fidelity. The best-performing model, based on LightGBM trained on Dragon descriptors augmented with MD and quantum chemical features, achieves a mean absolute error of 22.5 K, outperforming descriptor-only models and structure-based deep learning baselines. Shapley additive explanations interpretability analysis reveals that melting behavior is governed primarily by molecular topology, surface-area-weighted electronic descriptors, and condensed-phase interaction properties. In contrast, many isolated functional group and single molecule electronic descriptors contribute negligibly once these effects are accounted for. These results demonstrate that incorporating physics-informed, multi-scale descriptors enables more accurate and physically interpretable MP predictions.

Autor:innen und Institutionen

  • Department of Chemical and Biological Engineering, Brigham Young University, Provo, Utah 84602, USA.

ORCID-Profile: Frank T Mtetwa, Neil F Giles, W Vincent Wilding, Thomas A Knotts

Zitieren

APA

Mtetwa, F. T., Giles, N. F., Wilding, W. V., & Knotts, T. A. (2026). Machine learning prediction of organic compound melting points informed by condensed-phase and electronic descriptors. The Journal of Chemical Physics, 165(7). https://doi.org/10.1063/5.0342619

Vancouver

Mtetwa FT, Giles NF, Wilding WV, Knotts TA. Machine learning prediction of organic compound melting points informed by condensed-phase and electronic descriptors. The Journal of Chemical Physics. 2026;165(7). doi:10.1063/5.0342619

BibTeX
@article{Mtetwa2026,
  author = {Mtetwa, Frank T and Giles, Neil F and Wilding, W Vincent and Knotts, Thomas A},
  title = {Machine learning prediction of organic compound melting points informed by condensed-phase and electronic descriptors},
  journal = {The Journal of Chemical Physics},
  year = {2026},
  volume = {165},
  number = {7},
  doi = {10.1063/5.0342619}
}
Kostenlos starten →