PhD Research

If you want to know more about my PhD research, check out my thesis!

Summary

Advanced electrochemical systems, such as redox flow batteries, rely on porous electrodes, which determine the system performance and costs. Conventional porous electrodes currently used in convection-enhanced technologies are fibrous carbonaceous materials developed for low-temperature fuel cells; as such, their microstructure and surface chemistry are not suited for redox flow batteries. Thus, there is a need for targeted synthesis and engineering of porous electrodes with tailored properties to meet the requirements of liquid-phase electrochemistry. Specifically, the three-dimensional structure of the electrode is critical as it determines the available surface area for electrochemical reactions, electrolyte transport, fluid pressure drop, and electronic and thermal conductivity. However, the role of the electrode microstructure and, consequently, the optimal design for specific flow battery chemistries remains unknown. Hence, a multi-variable optimization problem at different length scales must be solved with highly coupled transport phenomena and kinetics. In this doctoral dissertation, a novel framework is developed to design and synthesize electrode structures from the bottom up, tailored to emerging electrochemical technologies with a focus on redox flow batteries. In this thesis, fundamental structure-function-performance relationships are elucidated by imaging and modeling commercial electrodes, which are used to design hierarchically organized porous electrodes through topology optimization and to manufacture model grid structures with 3D printing.

In Chapter 1, redox flow batteries are introduced as a technology that is promising for large-scale energy storage to bridge the temporal and geographical gaps between energy demand and supply. Thereafter, the transport phenomena and cell overpotentials in redox flow batteries are discussed in detail. Furthermore, experimental diagnostics, electrode manufacturing, multiphysics simulations, and computational electrode optimization strategies are introduced to aid the theoretical understanding and design of advanced electrode structures, as the empirical design process alone is time- and resource-intensive, limiting exploration of the wider design space. Finally, the scope of the doctoral dissertation is presented based on four objectives.

In Chapter 2, the developed pore network model is explained, which is a simulation framework for flow batteries that is microstructure-informed and electrolyte-agnostic, constructed using an open-access platform (OpenPNM), and validated with experimental data. The model utilizes a network-in-series approach to account for species depletion over the entire length of the electrode, enabling the simulation of large electrode sizes. To validate the robustness of the modeling framework, single-electrolyte flow cell experiments were performed for two distinct electrolytes – aqueous and non-aqueous – and two types of porous electrodes – carbon paper and -cloth, extracted using X-ray computed tomography and converted into a pore network. The electrochemical model solves the electrolyte fluid transport and couples both half-cells by iteratively solving the species and charge transport at a low computational cost. The electrochemical performance of the non-aqueous electrolyte is well captured by the model without fitting parameters, allowing rapid benchmarking of porous electrode microstructures. For the aqueous electrolyte, it is found that incomplete wetting of the electrode results in overprediction of the electrochemical performance. Finally, a parametric sweep is discussed for the identification of operation envelopes.

In Chapter 3, the pore network model of Chapter 2 is used, together with experimental techniques, to investigate the effect of stacking electrode layers for two prevailing commercial electrodes and flow fields. To this end, the pore network model is extended to simulate an interdigitated flow field design, and thickness-structure-performance relationships are obtained for specific electrode-flow field configurations. By stacking commercial electrodes, for example, two carbon paper electrodes with a flow-through configuration, the overall reactor efficiency can be enhanced. Furthermore, both the electrochemical power and pressure drop are improved, providing a facile strategy to enhance the performance of flow batteries.

In Chapters 4 and 5, a genetic algorithm is developed and coupled with the pore network model of Chapter 2 for the bottom-up design of electrode structures. Using this approach, the electrode microstructure evolves driven by a fitness function that minimizes the pumping power and maximizes the electrochemical power, requiring only the electrolyte chemistry, initial electrode morphology, and flow field geometry as inputs. Thus, making it a versatile framework that can be applied to other electrochemical systems. In Chapter 4, the principle of the genetic algorithm is introduced, where a flow-through cubic lattice structure with fixed pore positions is analyzed and shows significant improvement in the fitness function over 1000 generations. The evolution results in a structure with a bimodal pore size distribution containing longitudinal electrolyte flow pathways of large pores, significantly reducing the pumping requirements. Additionally, an increase in surface area at the membrane-electrode interface is found, resulting in an enhancement of the electrochemical performance.

In Chapter 5, the genetic algorithm is extended to allow for more evolutionary freedom during the optimization process by evolving structures beyond fixed flow-through cubic lattice designs. To this end, pore merging and splitting are incorporated to allow larger geometrical flexibility. In addition, the optimization of complex structures is investigated by the implementation of Voronoi networks and X-ray tomography extracted off-the-shelf fibrous electrodes as starting networks. Subsequently, the effect of operational conditions is analyzed by evaluating two chemistries and two prevailing flow field geometries. Furthermore, optimization definitions (fitness function and geometrical definitions) are elaborated upon to inform about their importance and to show the flexibility of the algorithm.

In Chapter 6, a new neutron radiography approach is introduced to quantify concentration distributions in operando redox flow cells, providing a new diagnostic tool to better understand reactive transport phenomena in electrochemical reactors. The presented approach is developed for a non-aqueous model redox system where the attenuation comes from the hydrogen or boron atoms that compose redox active molecules or supporting ions in non-aqueous electrolytes. Concentration profiles are resolved across the electrode thickness by employing in-plane imaging and are correlated to the cell performance with polarization measurements under various operating conditions. Two neutron imaging methods are used, where first the combined attenuation of the active species and supporting salt is examined over the electrode thickness. Thereafter, a time-of-flight neutron imaging approach is used to analyze deconvoluted active species and supporting ion transport over time. With this approach, the transport of species in the reactor under a voltage bias is revealed, gaining insights into reactive transport phenomena within an operating flow cell.

In Chapter 7, the neutron radiography approach of Chapter 6 is utilized to study local transport properties and concentration distributions in the porous electrode for three distinct electrode structures and with two flow field designs. By tracking the cumulative active species movement, it is found that, for electrolytes with facile kinetics and low ionic conductivity, an electrode structure with a bimodal pore size distribution with large through-plane voids is favorable combined with parallel flow fields because of the high through-plane hydraulic conductance and effective diffusivity, enhancing the current output. Comparably, interdigitated flow fields feature higher reaction rates and current output compared to parallel designs because of forced convection. Moreover, neutron radiography is proven useful in the detection of system secondary phenomena including salt precipitation and underutilization of flow field channels.

In Chapter 8, the manufacturing of porous electrodes using 3D printing is presented. Model grid structures were manufactured with stereolithography followed by carbonization to tune the physicochemical properties of electrodes. A suite of microscopy, tomography, spectroscopy, fluid dynamics, and electrochemical diagnostics was employed to evaluate the thermal behavior, manufacturing fidelity, and fluid and mass transport performance of ordered lattice structures in non-aqueous redox flow cells. The influence of the pillar geometry, printing orientation regarding the printing platform, and flow field design on the electrode performance is investigated. It is found that although commercial electrodes feature a greater internal surface area and therefore better performance, the area-normalized mass transfer coefficient is improved and the pressure drop is reduced by utilizing 3D-printed electrodes.

Finally, the main findings of this work are summarized in Chapter 9, and future research directions to accelerate and broaden the design and fabrication processes of advanced electrode structures are provided.