Leveraging Machine Learning for Screening Metal-Organic Frameworks with Selective CO₂ Recognition for Early Thermal...

Area of Science:

• Materials Science
• Computational Chemistry
• Chemical Engineering

Background:

• Lithium-ion batteries pose safety risks due to thermal runaway, necessitating early failure detection.
• Carbon dioxide (CO2) release during battery decomposition offers a critical window for early warning.
• Developing sensors with high selectivity for CO2 is crucial for reliable monitoring.

Purpose of the Study:

• To systematically screen a large database of metal-organic frameworks (MOFs) for selective CO2 recognition.
• To identify optimal MOF structural and chemical properties for CO2 sensing in a competitive gas environment.
• To accelerate the discovery of novel MOF-based materials for enhanced lithium-ion battery safety.

Main Methods:

• Integration of Grand Canonical Monte Carlo (GCMC) simulations with Random Forest (RF) machine learning models.
• Screening of 1470 MOFs from the CoRE-MOF 2019 database for CO2, C2H4, and O2 adsorption.
• Performance evaluation based on working capacity, selectivity, and the trade-off metric (TSN); SHAP analysis for feature importance.

Main Results:

• The RF model achieved high predictive accuracy (R² > 0.92), identifying key performance drivers like CO2 binding strength and interactions with C2H4.
• Optimal MOFs possess specific characteristics: hard Lewis acid centers, polar clusters, moderate surface areas (965-1975 m²/g), narrow pore windows (4-7 Å), high void fractions (>0.6), and low densities (<1.3 g/cm³).
• AJOTEY was identified as the top-performing MOF, exhibiting a TSN of 6.43 mol/kg with a working capacity of 4.57 mol/kg and selectivity of 25.52.

Conclusions:

• The computational screening approach effectively identifies promising MOFs for CO2 sensing applications.
• Material properties influencing CO2 selectivity are linked to interactions with competing gases like ethylene (C2H4).
• The identified optimal MOF characteristics and the top candidate AJOTEY provide a practical foundation for developing advanced CO2 sensors to improve lithium-ion battery safety.