Coupling of modern computational techniques to study the oxygen reduction reaction

Abstract

The oxygen reduction reaction (ORR) at the fuel cell cathode is currently one of the bottlenecks for successful commercialisation of Proton Exchange Membrane Fuel Cells. The multifaceted nature of the ORR, involving multiple reaction steps, various intermediates, and complex interfacial phenomena, demands increasingly sophisticated computational approaches. The interplay between catalyst structure and oxidation, solvation effects, and electric fields at the interface further complicates the modelling process. In this thesis, we aim to develop a comprehensive understanding of the ORR on platinum catalysts by integrating advanced computational techniques with fundamental electrochemical principles. We begin by evaluating current forcefields and theories, finding that machine learning methods improve the accuracy of thermodynamical prediction on idealised systems when compared to classical methods. To account for electrolyte and solvent effects, we explore the recent reference interaction site model revealing its strengths in predicting electrochemical properties but limitations for thermodynamic predictions. To further bridge the gap between idealised models and realistic conditions, we perform grand-canonical Monte Carlo simulations on a realistic platinum nanoparticle. These simulations serve as an additional comparison for various forcefields, and provide insights into oxidation behaviour across a range of pressures. Finally, we build upon the previous findings, and provide a unified computational model to predict the overpotential of the ORR. Our findings suggest that the overpotential is likely due to a place-exchange mechanism near 1.05 V, resulting in oxide formation that hinders both reaction kinetics and thermodynamics. This approach, which unifies both thermodynamic and kinetic aspects, has been made possible through the combination of advanced computational methods and fundamental electrochemical principles. In this framework we develop a kinetic model, which includes properties and their variations directly computed with modelling techniques. Both the framework and the methodology can be further improved and open the way to future catalyst design, paving the way for fuel cell technology optimisation; potentially leading to more efficient and cost-effective energy conversion devices.

More information

Identifiers

ORCID for Tom Demeyere: orcid.org/0000-0002-5023-6156

ORCID for Chris Skylaris: orcid.org/0000-0003-0258-3433

Catalogue record

Export record