Materials process informatics-assisted precise particle size control of metal-organic frameworks

Area of Science:

• Materials Science
• Chemical Engineering
• Data Science

Background:

• Precise control of metal-organic framework (MOF) particle size is crucial for applications in catalysis, separation, and drug delivery.
• Current synthesis methods rely on empirical trial-and-error, lacking predictive power for optimizing MOF particle size.
• Understanding the kinetics of nucleation and growth is key to controlling MOF particle formation.

Purpose of the Study:

• To establish a materials process informatics framework for the predictive control of MOF particle size.
• To utilize zeolitic imidazolate framework-8 (ZIF-8) as a model system for developing and validating the predictive framework.
• To enable data-driven design for MOF synthesis, moving beyond empirical methods.

Main Methods:

• Constructed a comprehensive database by curating literature data on MOF synthesis.
• Employed seven process descriptors as input features for machine learning models.
• Benchmarked multiple machine learning algorithms, including Categorical Boosting (CB), and utilized SHapley Additive exPlanations (SHAP) for parameter analysis.

Main Results:

• The Categorical Boosting (CB) model achieved a high predictive performance with R² = 0.90 on the test set.
• SHAP analysis identified precursor concentration ratio and reaction time as the most influential parameters for MOF particle size.
• Experimental validation using an automated synthesis platform confirmed excellent agreement between predicted and measured particle sizes.

Conclusions:

• The developed materials process informatics framework enables predictive control over MOF particle size.
• The data-driven approach facilitates intelligent synthesis optimization and targeted experimental design.
• This work provides a broadly applicable strategy for controllable MOF synthesis and advanced materials development.