Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 31 min ago
PreprintCC BYChemRxiv

BiFuN: Machine Learning-Guided Bifunctional Catalysis in Hydrogenation Reactions

  • Lucía Morán-González1
  • Hannes Kneiding1
  • Mohamed E. A. Safy1
  • Matthias Beller2
  • David Balcells1
  • Ainara Nova1,3
  1. 1Hylleraas Centre for Quantum Molecular Sciences, Department of Chemistry, University of Oslo, P.O. Box 1033, Oslo, Norway, 0315
  2. 2Leibniz-Institut für Katalyse e.V., Albert-Einstein-Straße 29a, Rostock, Germany, 18059
  3. 3Centre for Materials Science and Nanotechnology, Department of Chemistry, University of Oslo, P.O. Box 1033, Oslo, Norway, 0315
Open at sourcePDFDOICite

Abstract

Hydrogen activation and further hydrogenation of substrate cover chemical processes in synthetic and energetic-related technologies. In homogeneous catalysis, transition metal complexes with bifunctional nature offer a platform for the activation and later utilization of the hydrogen-source. However, the development of new catalysts is limited by long-term trial-and-error practices. Screening approaches offer the possibility of relying on key properties to identify promising candidates. Here, we present a data-driven workflow that integrates high-throughput chemical screening, quantum-mechanical (QM) calculations, and machine learning (ML) predictions to accelerate the rational discovery of promising bifunctional catalysts (BFC). Using the Ru-MACHO scaffold as a BFC reference, we systematically explore metal complex variations across four design axes: 11 metal centers, 2,538 pincer ligands, 74 transligands, and 3 axial ligands, generating ca. 1.5M candidate structures. QM-derived data and ML models predict catalyst hydrogenation thermodynamics, enabling rapid identification of top-performing candidates for further mechanistic validation in CO2 hydrogenation. This integrated approach delivers prioritized systems for experimental follow up, demonstrating how curated datasets, targeted QM calculations, and ML together streamline the discovery of next generation homogeneous catalysts for efficient hydrogenation/dehydrogenation reactions.

Related

  1. How accurate are foundational machine learning interatomic potentials for heterogeneous catalysis?
    Luuk H. E. Kempen, Raffaele Cheula, Mie Andersen
    The Journal of Chemical PhysicsPeer-reviewedLicence not statedFull text9 citedMaterials & polymersOpenAlex
  2. Simulating enzyme catalysis with electrostatically embedded machine learning potentials
    Valentin Gradisteanu, Elliot W Chan, Lester Hedges +7
    Chemical SciencePeer-reviewedCC BY 4.0Full text15 citedMaterials & polymersHAL
  3. Graph-based machine learning model for adsorption energy prediction of CO2 hydrogenation intermediates on metal oxide catalysts
    Muhammad Helmi Risansyauqi, Jean C. Villa-Arpi, Hector Prats +1
    ChemRxivPreprintCC BYMaterials & polymersChemRxiv
  4. In situ XAFS analysis system for high-pressure catalytic reactions and its application to CO2 hydrogenation over a Rh/Y-zeolite catalyst
    Kyoko K. Bando, Takeru Saito, Koichi Sato +5
    Journal of Synchrotron RadiationPeer-reviewedLicence not statedFull text19 citedCatalysisHAL
  5. Concerted, Facile CO2 Hydrogenation in Ni-MOF-74 Revealed by Machine Learning Interatomic Potential-Based Enhanced Sampling Simulations
    Dhananjay, Nishtha Sharma, Tarak Karmakar
    ChemRxivPreprintCC BY-NCInorganic & organometallic chemistryChemRxiv

Cite

BibTeXRIS
BibTeX
@misc{morangonzalez2026bifun,
  title = {{BiFuN: Machine Learning-Guided Bifunctional Catalysis in Hydrogenation Reactions}},
  author = {Lucía Morán-González and Hannes Kneiding and Mohamed E. A. Safy and Matthias Beller and David Balcells and Ainara Nova},
  howpublished = {ChemRxiv},
  year = {2026},
  month = sep,
  doi = {10.26434/chemrxiv.15008296/v1},
  url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15008296/v1},
  note = {CC BY},
}
RIS
TY  - UNPB
TI  - BiFuN: Machine Learning-Guided Bifunctional Catalysis in Hydrogenation Reactions
AU  - Lucía Morán-González
AU  - Hannes Kneiding
AU  - Mohamed E. A. Safy
AU  - Matthias Beller
AU  - David Balcells
AU  - Ainara Nova
PY  - 2026
DA  - 2026/09/04
PB  - ChemRxiv
DO  - 10.26434/chemrxiv.15008296/v1
UR  - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15008296/v1
AB  - Hydrogen activation and further hydrogenation of substrate cover chemical processes in synthetic and energetic-related technologies. In homogeneous catalysis, transition metal complexes with bifunctional nature offer a platform for the activation and later utilization of the hydrogen-source. However, the development of new catalysts is limited by long-term trial-and-error practices. Screening approaches offer the possibility of relying on key properties to identify promising candidates. Here, we present a data-driven workflow that integrates high-throughput chemical screening, quantum-mechanical (QM) calculations, and machine learning (ML) predictions to accelerate the rational discovery of promising bifunctional catalysts (BFC). Using the Ru-MACHO scaffold as a BFC reference, we systematically explore metal complex variations across four design axes: 11 metal centers, 2,538 pincer ligands, 74 transligands, and 3 axial ligands, generating ca. 1.5M candidate structures. QM-derived data and ML models predict catalyst hydrogenation thermodynamics, enabling rapid identification of top-performing candidates for further mechanistic validation in CO2 hydrogenation. This integrated approach delivers prioritized systems for experimental follow up, demonstrating how curated datasets, targeted QM calculations, and ML together streamline the discovery of next generation homogeneous catalysts for efficient hydrogenation/dehydrogenation reactions.
ER  -