Role of Electrode Microstructure in Performance of Electrochemical Energy Storage Devices
[Thesis]
Sadeghi, Mohammad Amin
Servio, Phillip
McGill University (Canada)
2019
132
Ph.D.
McGill University (Canada)
2019
Renewable energy is being increasingly deployed at both small and large scales. At large scale, the intermittent nature of renewable energy sources prohibits a full integration with the current power grid. Among others, redox flow battery is a promising energy storage technology that can be used to mitigate this issue. At smaller scale, specifically the transportation sector, the industry is already undergoing a change toward electric vehicles. Fuel cell and plug-in electric vehicles have been commercially deployed in the past few years and are projected to replace the traditional heat engine vehicles. Both the redox flow battery and fuel cell technologies are still expensive. Maximizing the energy and power density of these devices is one way to reduce their cost. This thesis is directed toward developing the required numerical framework through which the multiphysics occurring within the microstructure of electrochemical devices could be better understood, and eventually the impact of their internal structure on their overall performance could be predicted. Reactive transport was studied in two different systems of diffusion-dominated and combined advection-diffusion, which are relevant in the context of fuel cells and redox flow batteries. The pore network approach was employed for modeling transport in these systems. This approach enabled the study of meaningfully large domains that otherwise would be infeasible using typical pore-scale modeling approaches, usually referred as direct numerical simulation (DNS). As for diffusion-dominated systems, reactive transport in a hierarchically porous particle was considered, and the effect of the internal structure was studied. It was found that using a hierarchy of porosity, a 350% increase in power density could be achieved, merely by manipulating the internal structure. It was also found that using smaller-sized templates for creating macropores within a nanoporous particle leads to much larger performance gains when increasing the volume fraction of macropores. For energy storage devices in which mass transfer is a result of both diffusion and advection, it was found that the most common approach for modeling dispersion in pore networks is too crude an approximation. Therefore, in a comprehensive study, this discrepancy was demonstrated via comparison with DNS as ground truth and was followed by introducing a novel approach adapted from the CFD literature. The developed pore-scale model was shown to agree remarkably well with the results obtained from DNS with a maximum relative error of 0.5%. Finally, the developed pore-scale model was used in a multiphysics pore-network study of redox flow batteries, demonstrated in the context of a hydrogen bromine system. The effect of porosity at constant fiber diameter and fiber alignment on the overall performance of the battery was studied, both of which can be manipulated during electrode manufacture. It was shown that despite the decrease in available reactive surface area, the battery generally performed better at higher porosities. Furthermore, it was shown that aligning the fibers along the flow direction, while initially helping the electrode performance, leads to diminishing returns beyond slight alignment. This phenomenon was shown to correlate with the diminishing return in permeability of the electrode as alignment increased. This observation, along with the trend in performance against porosity shows the significance of permeability as the dominant factor in performance of such systems. This thesis highlights pore-network modeling as a practical way to model transport in energy storage devices at the pore-scale with minimal computing requirements. Developing such a modeling framework paves the way for better understanding how the internal structure of these devices affects their performance, serving as a guideline for making better prototypes with higher energy densities. The power of this framework was demonstrated in the context of two promising energy storage devices, but we expect that the different studies presented in this thesis be the cornerstone for many future studies, through which a more complete understanding of multiphysics at pore-scale will be achieved.