Optimization of Electrochemical Devices for Sustainable Chemical Manufacturing.

다국어 초록 (Multilingual Abstract)

Our increased efforts to curb global warming have led to a drastic surge in the deployment of renewable electricity sources, such as wind and solar power. However, as these sources form a larger fraction of the energy in the grid, their intermittency has started to cause supply instability and large fluctuations in energy prices. Electrochemical energy storage devices have started to enter the utility-scale energy storage market to address this need, but high costs associated with manufacturing and maintain these devices have limited their impact, especially for long-duration energy storage. As an alternative to using battery systems for storage, electrochemically produced fuels from water and/or CO2 have been proposed as viable storage opportunities, especially for seasonal energy storage requirements. H2 has gained significant attention as a promising energy vector for a renewable-rich energy future given its high gravimetric energy density that makes it desirable for both stationary and mobile applications. Despite the fact that clean H2 can be produced electrochemically through water electrolysis, the high cost of this method has limited its deployment, mainly due to the electricity costs. In the same way, CO2 has the ability to be reduced to many useful products such as carbon monoxide and ethylene, but the selectivity and energy efficiency of these devices need to be improved. The work presented in this dissertation aims at providing technological solutions to these challenges that preclude the deployment of electrochemical technologies at scale. In particular, I describe the development and implementation of Bayesian learning methods to optimize electrochemical devices that go beyond the more common Edisonian approaches used in the development of these devices. Edisonian search approaches are widely used in the chemical sciences to discover reactions, process conditions, material compositions, or product formulations with optimal performance for their intended application. These experimental design methods rely on the generation of grids of variables where experimentally accessible conditions are systematically and/or combinatorically explored. While these methods are simple to implement, they often evaluate a suboptimal parameter space where the quality of information derived depends on the numbers of combinations of variables explored, slowing and sometimes preventing the identification of optimal conditions. These shortcomings represent significant impediments for expensive experimental campaigns or those with large design spaces that can only afford the implementation of coarse experimental grids, underscoring the need for more efficient experimental optimization methods. In order to implement the electrochemical devices described above, it is important to be able to improve the speed of experimental campaigns to ultimately implement these systems when they can make the most impact.The following chapters cover the characterization and optimization of two renewable fuel systems, as well as a physics-based Bayesian optimization method for improved optimization of these electrochemical systems. In Chapter 2, a cerium-mediated energy storage and hydrogen production system is introduced and characterized. In Chapter 3, a technoeconomic analysis is performed that optimized the operation schedule and sizing of the system to minimize hydrogen production cost. In Chapter 4, a study and optimization of potential pulses on the cerium(III) oxidation reaction is presented with the goal of improving the efficiency of this reaction. In Chapter 5, the optimization of potential pulses on a CO2 electroreduction device is described. Finally, in Chapter 6 a chemically-informed Bayesian optimization algorithm is introduced.