Despite the abundance of renewable energy resources, a lack of economically feasible storage solutions for addressing intermittency remains a barrier to advancing their widespread adoption. Reversible solid oxide cells, which can store hydrogen during periods of renewable energy overproduction, have demonstrated potential for grid stabilization applications given their high potential efficiencies and power densities. However, to become economically competitive, improvements to reversible solid oxide cell performance stability and lifetimes are required. This research focuses on understanding the relationship between microstructure and performance in solid oxide cells and explores avenues for mitigating electrode polarization and degradation. Connections between microstructure and performance were first considered in Ni/YSZ symmetric cells, where the relationship between reaction site density and performance was quantified in nanocatalyst-infiltrated cells using EIS, SEM and FIB/SEM 3-D reconstruction. In Ni-infiltrated electrodes, results showed that both increased triple phase boundary density and decreased reaction rate constants contribute to lowering electrode polarization at intermediate temperatures. In electrodes infiltrated with GDC, a mixed ionic/electronic conducting material, reactions can take place on the GDC surface, greatly decreasing electrode polarization. Calculations considering the performance of baseline and GDC-infiltrated electrodes indicated that reactions take place up to 84nm from triple phase boundaries on nickel scaffold particle surfaces. Microstructure/performance relationships were also examined in full cells tested for 500h under electrolysis or reversible conditions; the fuel and oxygen electrodes were characterized with methods including low-voltage SEM, FIB/SEM 3-D reconstruction, and TEM. In both scenarios, the oxygen electrode was shown to contribute minimally to cell degradation. In the fuel electrode, degradation was mainly precipitated by Ni coarsening and loss of active sites; however, these were mitigated during reversible testing by 9% and 8% respectively compared to electrolysis-tested cells. Finally, strategies are discussed for mitigating long-term degradation. To further stabilize reversible full cells, GDC infiltration into the fuel electrode and adjustments to oxygen electrode phase compositions to prevent long-term decomposition are suggested. On the SOC system level, ALD spinel coatings for interconnect materials are considered. To this end, a successful ALD coating process for manganese oxide on stainless steel is discussed.

Thermodynamic analysis of solid oxide cells Petr Vágner The thesis deals with continuum thermodynamic modeling and analysis of phe- nomena in solid oxide electrochemical cells. A general description of the evo- lution of charged mixtures using partial mass densities, momentum density, entropy density, electric induction, magnetic field, polarization, and magnetiza- tion based on the GENERIC framework is formulated. The formulation is used to recover the Landau-Lifshitz magnetization relaxation model, the Single Re- laxation Time model for dielectrics, and the generalized Poisson-Nernst-Planck model. The latter model is consequently linked to the second part, where a novel double layer model of an yttria-stabilized zirconia interface is formulated within non-equilibrium thermodynamics. The model is solved for numerically in the time domain, and cyclic voltammetry of the system is analyzed. The last part of the thesis demonstrates the limits of Exergy Analysis on a simple solid oxide hydrogen fuel cell model with non-isothermal boundary. It is demon- strated that the minimization of entropy production does not necessarily lead to the maximization of the electric power for certain optimization scenarios. The thesis consists of a compilation of published and unpublished results of the author.

Ce travail porte sur la compréhension de l’impact de la microstructure sur les performances des Cellules à Oxyde Solide (SOC), avec une illustration sur l’électrode à oxygène en LSCF-CGO. Une approche couplant de la modélisation géométrique et électrochimique a été adoptée pour cet effet. Le modèle des champs aléatoires plurigaussiens et un autre basé sur des empilements de sphères ont été développés et adaptés pour les microstructures des SOCs. Ces modèles 3D de géométrie stochastique ont été ensuite validés sur différentes électrodes reconstruites par nano-holotomographie aux rayons X au synchrotron ou par tomographie avec un microscope électronique à balayage couplé à une sonde ionique focalisée. Ensuite, des corrélations semi-analytiques ont été proposées et validées sur une large base de microstructures synthétiques. Ces relations permettent de relier les paramètres ‘primaires’ de l’électrode (la composition, la porosité et les diamètres des phases) aux paramètres qui pilotent les réactions électrochimiques (la densité de points triples, les surfaces spécifiques interphases) et sont particulièrement pertinents pour les équipes de mise-en-forme des électrodes qui ont plus de contrôle sur ce premier ensemble de paramètres. Concernant la partie portant sur l’électrochimie, des tests sur une cellule symétrique en LSCF-CGO ont permis de valider un modèle déjà développé au sein du laboratoire, et qui permet de simuler la réponse électrochimique d’une électrode à oxygène à partir des données thermodynamiques et de microstructure. Finalement, le couplage des deux modèles validés a permis d’étudier l’impact de la composition des électrodes, leur porosité ou encore taille des grains sur leurs performances. Ces résultats pourront guider les équipes de mise-en-forme des électrodes vers des électrodes plus optimisées. / This work aims at better understanding the impact of Solid Oxide Cells (SOC) microstructure on their performance, with an illustration on an LSCF-CGO electrode. A coupled 3D stochastic geometrical and electrochemical modeling approach has been adopted. In this frame, a plurigaussian random field model and an in-house sphere packing algorithm have been adapted to simulate the microstructure of SOCs. The geometrical models have been validated on different electrodes reconstructed by synchrotron X-ray nano-holotomography or focused ion-beam tomography. Afterwards, semi-analytical microstructural correlations have been proposed and validated on a large dataset of representative synthetic microstructures. These relationships allow establishing the link between the electrode ‘basic’ parameters (composition, porosity and grain size), to the ‘key’ electrochemical parameters (Triple Phase Boundary length density and Specific surface areas), and are particularly useful for cell manufacturers who can easily control the first set of parameters. Concerning the electrochemical part, a reference symmetrical cell made of LSCF-CGO has been tested in a three-electrode setup. This enabled the validation of an oxygen electrode model that links the electrode morphological parameters to its polarization resistance, taking into account the thermodynamic data. Finally, the coupling of the validated models has enabled the investigation of the impact of electrode composition, porosity and grain size on the cell electrochemical performance, and thus providing useful insights to cell manufacturers.