Study of the oxygen reduction reaction by kinetic Monte Carlo simulations from first-principles
Typ
Examensarbete för masterexamen
Master's Thesis
Master's Thesis
Program
Övrigt, MSc
Physics (MPPHS), MSc
Physics (MPPHS), MSc
Publicerad
2024
Författare
Vanmoerkerke, Willem
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
While the oxygen reduction reaction is a technologically important and widely studied
reaction, the origin of the onset overpotential remains debated. This study
aims to elucidate the reaction mechanism and the rate-determining step by kinetic
Monte Carlo simulations based on first-principles calculations. It is found that the
overpotential is determined by a potential dependent coverage of oxygen species
(∗O, ∗OH, ∗H2O) that block O2 adsorption. The coverage is largely determined
by adsorbate-adsorbate interactions. Therefore, the origin of these interactions are
studied extensively. A model is proposed and implemented in the kinetic Monte
Carlo simulations based on d-band shifts of surface atoms and hydrogen bonding
interactions. Additionally, attention is given to the effect of an aqueous environment
on the adsorbent by employing ab-initio molecular dynamics simulations. The
effect of strain, particle size and shape is investigated and particles are optimized
using Bayesian optimization. It is shown that particles with sizes between 3 and
6 nm and a large proportion of (111) facets are optimal for the reaction. Furthermore,
compressive strains have a positive effect on the activity, while tensile strain
reduces activity. Effects of experimentally determined site-specific strain and grain
boundaries are explored, which further improve the activity of the catalyst. Subsequently,
the limits of the activity is explored by generating kinetic volcano plots
from the kinetic Monte Carlo simulations. Following that, the model is generalized
to metal alloys, showing how strong interactions between Ag clusters and Pt
clusters in AgxPt1–x binary surface alloys can improve catalyst activity. Finally, a
framework is presented to simulate linear sweep voltammetries, which can be readily
compared with experimental data. Overall, this thesis underscores the significance
of interaction effects in accurately describing experimental conditions and introduces
a first-principles method for catalyst exploration.
Beskrivning
Ämne/nyckelord
Oxygen reduction reaction , fuel cell , computational catalysis , electrochemistry , platinum nanoparticles , adsorbate-adsorbate interactions , kinetic Monte Carlo , Density-functional Theory , Bayesian optimization