Machine Learning Tailored Anodes for Efficient Hydrogen Energy Generation in Proton-Conducting Solid Oxide...

Area of Science:

• Materials Science
• Electrochemistry
• Energy Conversion

Background:

• Proton-conducting solid oxide electrolysis cells (P-SOECs) are crucial for efficient hydrogen energy production, but require high-performance anodes.
• Existing anodes face challenges in achieving both high catalytic activity and ionic conductivity for water electrolysis.

Purpose of the Study:

• To design and develop novel, high-performance anode materials for P-SOECs using a machine learning approach.
• To investigate the synergistic enhancement of water oxidation kinetics and proton conduction in tailored anode materials.

Main Methods:

• Machine learning model-assisted design of La_0.9Ba_0.1Co_0.7Ni_0.3O_3-δ (LBCN9173) and La_0.9Ca_0.1Co_0.7Ni_0.3O_3-δ (LCCN9173) perovskite oxides.
• Experimental validation combined with density functional theory (DFT) calculations.
• Analysis of distribution of relaxation time spectra and Gibbs energy for reaction mechanism elucidation.

Main Results:

• LBCN9173 and LCCN9173 oxides demonstrated synergistic enhancement of water oxidation reaction kinetics and proton conduction.
• The LBCN9173 anode facilitated H₂O dissociation, leading to superior performance.
• P-SOEC with LBCN9173 anode achieved a record current density of 2.45 A cm^-2 at 1.3 V and low polarization resistance (0.05 Ω cm²) at 650 °C.

Conclusions:

• Machine learning provides a robust framework for designing high-performance anodes for P-SOECs.
• The developed LBCN9173 anode significantly advances P-SOEC technology for efficient hydrogen production.
• This research paves the way for accelerated discovery of advanced materials for clean energy applications.