Dynamic Polaronic Control of Metal Cluster Adaptability on Reducible Oxides

Area of Science:

• Materials Science
• Catalysis
• Computational Chemistry

Background:

• Metal-oxide interactions are critical in catalysis, with defect chemistry influencing metal nanoparticle morphology.
• Strong metal-support interactions (SMSI) are complex and not fully understood.
• Understanding these interactions is key for designing efficient catalysts.

Purpose of the Study:

• To develop a physics-guided machine learning framework to elucidate metal-oxide interactions.
• To investigate the impact of oxygen vacancy concentration on platinum cluster morphology on ceria supports.
• To establish quantitative design principles for catalyst optimization.

Main Methods:

• Utilized a physics-guided machine learning framework.
• Simulated platinum clusters (Pt7 and Pt13) on ceria (CeO2-x) with varying oxygen vacancy concentrations (0-12.5%).
• Analyzed 528 configurations to model cluster shape and charge.

Main Results:

• Developed predictive models with R^2 > 0.97.
• Identified polaron swarms as the dominant factor controlling cluster shape and charge, overriding defect concentration.
• Demonstrated size-dependent pathways governing these interactions.

Conclusions:

• Polaron swarms are key to understanding metal cluster morphology in metal-support interactions.
• The developed framework provides a general methodology for studying metal-support interactions.
• Offers quantitative design principles for optimizing defect-driven catalysts.