Abstract
To achieve carbon neutrality by mid-21st century, rapid deployment of large-scale electrical energy storage options is required to integrate intermittent renewable energy resources such as solar and wind, and catalyze smart-grid infrastructure. Electrochemical energy storage technologies such as Li-ion batteries (LIBs) and redox flow batteries (RFBs) offer the modularity and scalability required for large-scale energy storage. RFBs operate by spatial separation of electrode and electrolyte, providing energy and power decoupling. This gives them an advantage over conventional LIBs, in terms of long life cycles (>10,000 cycle, 10-20 years) and inherent safety. However, the volumetric energy density of RFBs (<100 Wh l-1) is far lower than standard LIBs (600-650 Wh l-1). A hybrid electrochemical system called redox-targeting flow batteries (RTFBs) was suggested in 2013 that uses chemical redox of charge dense Li-ion battery materials through redox-targeting reactions, and has a theoretical capacity of > 500 Wh l-1. RTFBs however are in embryonic stages of development. The design principles of electroactive materials and the chemical redox reaction mechanism would need to be established to realize full-scale deployment of RTFBs.
To address these challenges, chemical redox kinetics and RTFB design were investigated in this dissertation. In Chapter Ⅱ, chemical redox kinetics were compared to the well-studied electrochemical kinetics of LFP. Chemical redox of LFP by ferrocene-based redox shuttles was evaluated using in-operando UV-Vis spectroscopy, whereas LFP half-cells were potentiostatically charged/discharged. The kinetics and phase transformations of the two redox routes were estimated using Johnson-Mehl-Avrami-Erofeyev-Kolomogorov model (JMAEK). It was revealed that the reaction pathway and phase transformations are different for chemical and electrochemical redox. The first-order rate constants were consistently lower for chemical redox than electrochemical redox. Additionally, apparent activation energy calculations suggested that the low redox shuttle concentrations (<1×10-3 mol l-1) could be the cause behind poor kinetics. This issue was mitigated in Chapter Ⅲ by increasing the redox shuttle concentration by a hundredfold. Furthermore, six derivatives of ferrocene-based redox shuttles were characterized using cyclic voltammetry for the redox targeting kinetics with LFP. The reactions were analyzed using two first-order kinetic models – JMAEK and 1-D diffusion model. Chemical redox kinetics of LFP with the shuttles showed no significant dependence on the redox potential and diffusion coefficient of the shuttles. This suggests specific redox shuttle-particle interactions influence the reaction rates. Moreover, robust experimental system was designed to compare redox shuttles for redox-targeting applications.
In Chapter Ⅳ, the design of LFP packed bed reactor (PBR) as the chemical reservoir is explored by changing four design variables: packed-bed height, concentration of redox shuttle, flow rate of electrolyte, and operating temperature. It was found that concentration of redox shuttle is the leading factor affecting the rate of chemical oxidation of LFP PBR. Additionally, under the reaction conditions tested, the rate-limiting step was observed to be dependent upon operating conditions. To further explore the reaction mechanism in LFP PBR, spatiotemporal progression of chemical oxidation was mapped using X-ray and neutron tomography in Chapter Ⅴ. Reaction heterogeneities observed using neutron tomography were found to be correlated to the LFP particle size and random distribution of LFP aggregates in the PBR.
