Discovery of High-Performance Ni_0.62Fe_0.23Ce_0.15O Electrocatalyst for Oxygen Evolution Reaction via Active Machine...

Area of Science:

• Materials Science
• Electrochemistry
• Computational Chemistry

Background:

• Heteroatom-doped transition metal oxides (H-TMOs) are promising electrocatalysts for the oxygen evolution reaction (OER).
• Optimizing H-TMOs' structure and composition for OER is challenging and time-consuming.
• Efficient OER electrocatalysts are crucial for renewable energy technologies.

Purpose of the Study:

• To develop an effective strategy integrating machine learning (ML) and genetic algorithms (GA) for forecasting OER electrocatalyst performance.
• To identify optimal NiO-based electrocatalysts with heteroatom doping for enhanced OER activity.
• To provide a paradigm for accelerating the rational design of novel electrocatalysts.

Main Methods:

• Developed an ML model, specifically Random Forest Regression (RFR), for predicting OER overpotentials.
• Integrated RFR with a genetic algorithm (GA) for efficient screening and optimization of catalyst compositions.
• Performed experimental validation and Density Functional Theory (DFT) calculations to confirm predictions and understand mechanisms.

Main Results:

• RFR model achieved high accuracy (RMSE of 4.73 mV) in predicting overpotentials.
• Predicted NiFeCeO electrocatalysts with specific Ce and Fe mole fractions exhibit lower overpotentials.
• Identified Ni _0.62Fe _0.23Ce _0.15O as the most promising catalyst, exhibiting an overpotential of 260 mV at 10 mA/cm ².
• DFT calculations revealed that Fe and Ce doping reduce the bandgap and improve electronic conductivity and kinetics, lowering the OER energy barrier.

Conclusions:

• The synergistic strategy of ML-guided screening, experimental validation, and DFT analysis accelerates catalyst discovery.
• The developed RFR-GA approach is effective for rational design of H-TMOs for OER.
• NiFeCeO demonstrates excellent OER performance, highlighting the potential of optimized heteroatom doping.