Related Experiment Video
Updated: May 16, 2026
10:52
Multiscale Sampling of a Heterogeneous Water/Metal Catalyst Interface using Density Functional Theory and Force-Field Molecular Dynamics
Published on: April 12, 2019
Towards accurate and scalable high-throughput MOF adsorption screening: merging classical force fields and universal
Satyanarayana Bonakala1, Mohammad Wahiduzzaman1, Taku Watanabe2
1ICGM, Univ. Montpellier, CNRS, ENSCM 34293 Montpellier France guillaume.maurin1@umontpellier.fr.
Chemical Science
|May 15, 2026
Summary
We developed a hybrid computational screening method combining classical force fields and machine learning potentials to accurately predict gas adsorption in metal-organic frameworks (MOFs). This approach enhances material discovery for applications like food packaging.
Area of Science:
- Computational materials science
- Adsorption science
- Chemical engineering
Background:
- High-throughput computational screening (HTCS) for gas adsorption in metal-organic frameworks (MOFs) traditionally uses generic force fields, limiting accuracy for complex host-guest interactions.
- Universal machine learned interatomic potentials (u-MLIPs) provide near-quantum accuracy at a lower computational cost than density-functional theory (DFT), but their large-scale use in adsorption screening is not yet established.
Purpose of the Study:
- To introduce an efficient hybrid screening workflow integrating classical force fields and u-MLIPs for accurate MOF adsorption performance assessment.
- To apply this workflow to identify MOFs for selective ethylene capture under humid conditions, crucial for food preservation packaging.
Main Methods:
- Developed a hybrid screening workflow combining classical generic force fields and u-MLIPs within a Monte Carlo simulation scheme.
- Applied the workflow to a large database of MOF structures.
- Validated the workflow's ability to accurately assess adsorption performance and framework flexibility.
Main Results:
- The hybrid workflow successfully screened a large MOF database for selective ethylene capture.
- Accurate treatment of host-guest energetics and framework flexibility using u-MLIPs was shown to be essential for reliable performance rankings.
- Identified top-performing MOFs for ethylene capture under humid conditions relevant to food packaging.
Conclusions:
- The developed hybrid HTCS workflow offers an efficient and accurate method for evaluating gas adsorption in MOFs.
- Integrating u-MLIPs is critical for capturing nuanced host-guest interactions and framework dynamics, leading to more reliable material discovery.
- This approach significantly advances the potential for designing MOFs for specific gas separation applications, such as in food preservation.
