Impact of extended defects on ion diffusion and reactivity in binary oxides : assessed by atomistic simulations

Abstract

Extended defects, such as dislocations, grain boundaries and surface step edges, are ubiquitous in nanomaterials that function in electrochemical devices and heterogeneous catalysts. The objective of this thesis is to advance the understanding of the influence of extended defects on ion diffusion and surface reactivity. The thesis investigates the interplay between extended defects and point defects and how this interplay alters ion diffusion and surface reactivity of binary oxides using multiple advanced computational modeling methods. The findings provide physics based insights into how to design the microstructure of materials in order to enhance the reactivity of materials in solid oxide fuel/electrolysis cells and catalysis. The work brings together computational techniques to address the chosen problems at suitable size and time scales.

In the first two studies, the goal is to resolve the distribution and transport kinetics of charged point defects, including both anions and cations, the latter being much slower than the former. The approach that is adapted for this problem area in this thesis is hybrid Monte Carlo and Molecular Dynamics simulations. The first study models an edge dislocation in reduced and doped ceria. The results showed that the dislocation redistributes point defects via elastic energy minimization and charged defect electrostatic association. The defect redistribution slows down oxide ion transport in the dislocation, contrary to the role of dislocations in accelerating atom migration in metals. The finding indicates that dislocations are detrimental for binary oxide solid electrolytes used for solid fuel cells, electrolyzers and membranes. In the second study, the interface between CeO₂ and Y₂O₃ was investigated.

The interface structural vacancies stabilizes and attract Ce³⁺ ions and oxygen vacancies. This leads to 1-2 orders of magnitude higher oxygen non-stoichiometry at the interface at low temperature and oxygen rich envirionments compared with that in bulk CeO₂. The interfacial enhanced oxygen capacity and the enhancement of electron polarons can be beneficial for catalyzing oxidation and reduction reactions and electron transport on ceria-based heterogeneous catalysts. The third study focuses on resolving whether dislocations could be sites to stabilize single atom catalysts, that are attractive for catalysis reactions but difficult to keep stable as single atoms at elevated temperatures. The approach here is density functional theory to compute defect formation energy at the core of an edge dislocation in Cu-CeO₂ in comparison to that in the bulk.

The result shows that dislocations can enrich Cu defects in an atomically-sized area and also stabilizes catalytically active cation species Cu¹⁺ and Ce³⁺. This result indicates that dislocations in Cu-CeO₂ have a great potential in anchoring single atom catalysts and enhance local reactivity. Lastly, a coupled quantum mechanics and molecular mechanics algorithm (QM/MM) is established and used to quantify the reactivity for oxygen evolution reaction (OER) at corners and edges of realistic nanoparticles. A 9.5-nm sized anatase titania nanoparticle was simulated by this approach. The reaction energies of oxygen evolution at the corners and edges are found 0.1 - 0.5 eV lower than on the facets since the former have a stronger adsorption of hydroxyl. However, some of the active sites are also prone to form electron polarons which can recombine with holes and compromise the high OER activity.

By considering both factors, the most active structures are found to be the (101) facets and the edges between {101} facets. This work provides insights for designing nanoparticle shapes for better photocatalysts and also demonstrates the potential of using the QM/MM method to simulate nanoparticles with realistic sizes.