Graph-based machine learning model for adsorption energy prediction of CO2 hydrogenation intermediates on metal oxide catalysts
4 affiliations
- 1Institute of Materials Chemistry
- 2TU Wien
- 3Departament de Ciència de Materials i Química Física & Institut de Química Teòrica i Computacional
- 4Universitat de Barcelona
Abstract
Mixed metal oxide catalysts are highly active for CO2 hydrogenation to methanol and typically more selective than classical Cu-based catalysts. Adsorption energies of key reaction intermediates can serve as activity descriptors for high-throughput screening of metal oxide-dopant combinations, but this requires numerous computationally demanding density functional theory (DFT) calculations. Machine-learning interatomic potentials (MLIPs) and adsorption energy prediction models offer faster alternatives, yet current architectures struggle with the structural and electronic complexity of mixed metal oxides introduced by dopants, oxygen vacancies, and diverse adsorption motifs. In this work, a graph-based Gaussian process regression (GPR) model is developed, combining graph representations of adsorption structures with node-level electronic descriptors. Trained on 638 adsorption structures of CO and all formate-pathway intermediates on pristine and Zn-doped ZrO2 and HfO2, as well as pristine In2O3 surfaces, the model achieves a test-set mean absolute error (MAE) of 0.16 eV. Leave-one-feature-out analysis identifies the d-band center, surface work function, and adsorbate LUMO energy as the dominant descriptors governing the prediction accuracy. Coupling the model with UMA small v1.2-D3 establishes an efficient workflow for rapid structural relaxation and accurate adsorption energy prediction, enabling high-throughput screening of mixed metal oxide catalysts for CO2 hydrogenation to methanol.
Related
- Engineering MoOₓ-based catalysts and reaction conditions to balance deoxygenation and hydrogenation in glycerol conversion for stable propylene production
- Tuning supported Fe-Ag catalysts for selective acetylene hydrogenation: A combined experimental and computational study
- A closer look into transfer hydrogenation between perhydro-dibenzyltoluene (H₁₈-DBT) and acetone for safe hydrogen release using Pt-based catalysts
- Case Study of Bayesian Batch Optimization Strategies for Heterogeneous Catalysts in CO2 Hydrogenation
- Adaptive Catalytic Synthesis of Aromatic and Aliphatic Amines via CO2-Induced Feedback Inhibition Using Standard Hydrogenation Catalysts.
BibTeX
@misc{risansyauqi2026graphbased,
title = {{Graph-based machine learning model for adsorption energy prediction of CO2 hydrogenation intermediates on metal oxide catalysts}},
author = {Muhammad Helmi Risansyauqi and Jean C. Villa-Arpi and Hector Prats and Aleix Comas-Vives},
howpublished = {ChemRxiv},
year = {2026},
month = oct,
doi = {10.26434/chemrxiv.15010002/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010002/v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Graph-based machine learning model for adsorption energy prediction of CO2 hydrogenation intermediates on metal oxide catalysts
AU - Muhammad Helmi Risansyauqi
AU - Jean C. Villa-Arpi
AU - Hector Prats
AU - Aleix Comas-Vives
PY - 2026
DA - 2026/10/06
PB - ChemRxiv
DO - 10.26434/chemrxiv.15010002/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010002/v1
AB - Mixed metal oxide catalysts are highly active for CO2 hydrogenation to methanol and typically more selective than classical Cu-based catalysts. Adsorption energies of key reaction intermediates can serve as activity descriptors for high-throughput screening of metal oxide-dopant combinations, but this requires numerous computationally demanding density functional theory (DFT) calculations. Machine-learning interatomic potentials (MLIPs) and adsorption energy prediction models offer faster alternatives, yet current architectures struggle with the structural and electronic complexity of mixed metal oxides introduced by dopants, oxygen vacancies, and diverse adsorption motifs. In this work, a graph-based Gaussian process regression (GPR) model is developed, combining graph representations of adsorption structures with node-level electronic descriptors. Trained on 638 adsorption structures of CO and all formate-pathway intermediates on pristine and Zn-doped ZrO2 and HfO2, as well as pristine In2O3 surfaces, the model achieves a test-set mean absolute error (MAE) of 0.16 eV. Leave-one-feature-out analysis identifies the d-band center, surface work function, and adsorbate LUMO energy as the dominant descriptors governing the prediction accuracy. Coupling the model with UMA small v1.2-D3 establishes an efficient workflow for rapid structural relaxation and accurate adsorption energy prediction, enabling high-throughput screening of mixed metal oxide catalysts for CO2 hydrogenation to methanol.
ER -