Data Driven Discovery of MOFs for Hydrogen Gas Adsorption

Area of Science:

• Materials Science
• Computational Chemistry
• Energy Storage

Background:

• Hydrogen gas (H2) is a promising clean energy carrier, but efficient storage remains a significant challenge.
• Metal-organic frameworks (MOFs) offer tunable porosity for gas storage but designing optimal structures is complex.
• The vast design space of MOFs hinders the selection of materials with high H2 storage capacity.

Purpose of the Study:

• To develop a data-driven computational framework for designing novel functionalized MOFs with enhanced H2 storage capacity.
• To identify key MOF structural and chemical properties influencing hydrogen adsorption.
• To demonstrate the application of machine learning and deep learning in predicting H2 adsorption in MOFs.

Main Methods:

• Hybrid particle swarm optimization integrated genetic algorithm for MOF design.
• Grand Canonical Monte Carlo (GCMC) simulations for adsorption analysis.
• In-house MOF structure generation code and machine learning/deep learning models.

Main Results:

• A functionalized IRMOF-10 exhibited a ~6-fold increase in H2 adsorption at 1 bar and 77 K compared to the pristine material.
• The computational framework successfully identified design rules for enhancing H2 adsorption in MOFs.
• Machine learning models demonstrated predictive capability for H2 adsorption based on MOF properties.

Conclusions:

• The developed data-driven framework effectively designs functionalized MOFs for superior H2 storage.
• This approach accelerates the discovery of advanced materials for clean energy applications.
• Machine learning and deep learning are powerful tools for understanding and optimizing MOF-based H2 storage.