Oxygen Reduction Activities of Strained Platinum Core-Shell Electrocatalysts Predicted by Machine Learning

Area of Science:

• Materials Science
• Nanotechnology
• Catalysis

Background:

• Core-shell nanocatalyst performance is governed by material composition, shell thickness, and particle size.
• Understanding the role of strain in nanocatalysts is crucial for optimizing their activity.

Purpose of the Study:

• To develop a machine learning framework for predicting strain effects on platinum (Pt) core-shell nanocatalysts.
• To rationalize how strain influences oxygen reduction reaction (ORR) activities.

Main Methods:

• Utilized a machine learning framework to predict strain with site-specific precision.
• Investigated Pt@Cu, Pt@Ni, Pt@Au, and Pt@Ag core-shell nanocatalyst systems.

Main Results:

• Large compressive strain on Pt@Cu and Pt@Ni at 1.9 nm nanoparticle size optimizes mass activities.
• Pt@Au and Pt@Ag exhibit optimal mass activities at 2.8 nm with weak compressive strain.
• Optimal strain is size-dependent, allowing further activity enhancement by tuning strain levels.

Conclusions:

• Strain is a critical factor in tuning the catalytic activity of Pt core-shell nanocatalysts for ORR.
• The developed machine learning framework provides precise predictions for optimizing nanocatalyst design.
• Tailoring strain in conjunction with nanoparticle size and composition offers a pathway to enhanced catalytic performance.