UCLA

Electrochemical technologies have the potential to dramatically reshape the global energy landscape, but implementation at commercial scales has been slow. Technologies like fuel cells and electrolyzers are recently gaining more widespread traction and are being demonstrated at large scale. These developments are being fueled by the push to convert to energy carriers that can be produced without greenhouse gas emissions. However, technologies like electrochemical CO2 reduction, that involve more complicated chemistry and multi-step reaction mechanisms, are still far from industrial applications despite having been under development for decades. Progress in these more complicated systems is slow due to the multitude of variables that affect the reaction. Complex interactions between mass, charge, and heat transport, along with reaction kinetics are often convoluted and difficult to distinguish using traditional research methods. These methods rely on relatively small data sets and often use poorly characterized experimental systems that provide inconsistent results. For example, in electrochemical CO2 reduction research, mass transport effects have generally been overlooked in the study of different catalysts over the past decade. However, recent work has shown that what is measured in these electrochemical cells is not the intrinsic reaction kinetics, but rather the reactor kinetics that are dependent on the reactor geometry and its influence on the many factors involved in the reaction of interest. Designing experimental systems that can isolate specific phenomena will be extremely difficult, if not impossible. Thus, new methods must be developed to aid in the extraction of fundamental relationships governing the processes occurring in these experimental systems.

This thesis explores how experimental electrochemistry can utilize methods inspired by Smart Manufacturing (SM) to accelerate the extraction of underlying relationships in complex electrochemical systems. In contrast to traditional experimental methods, SM techniques rely on large data sets to extract useful information from complex processes and optimize control. With large data sets of high quality data, modern computing capabilities can take advantage of computational methods in machine learning and multi-scale modeling to quickly draw hidden foundational relationships out of complex systems. Additionally, by utilizing sensing, data handling, and modeling techniques similar to those that will be used when an experimental technology is commercialized, this SM inspired approach can help to ensure that appropriate techniques are used when the technology is scaled up.

First, the framework for the application of SM techniques to electrochemical research will be discussed, and brief demonstrations of its utility will be presented. The following chapters will provide more detailed discussion of computational fluid dynamics modeling efforts in the application of this framework to electrochemical CO2 reduction research and characterization of our experimental system. Finally, a second case where the framework has been applied to the development of an experimental system for testing electrochemical steam methane reforming will be discussed along with motivation for the development of this system. Together, these cases demonstrate how SM techniques can be applied in the lab to accelerate discoveries and provide an understanding of how they might be applied in other areas of research.