There has been a renewed interest in flow batteries, particularly those employing highly soluble aqueous multi-electron oxidants, characterized by auto-catalytic accelerated mechanism. This enables, in theory, very high capacity, in the range of 1000 Ah/kg, which is not only ideal for application in the energy storage sector, but also holds promise for the transport industry, where, in contrast to commercially available Li-ion batteries currently used in fully-electric vehicles, it can provide a comparable driving range of 500 km, but in less than half the weight of the battery system under use presently (209 kg compared to 450 kg as used in a Tesla Roadster) [1]. The system under consideration is the bromate solution, where salt of bromate acts as the electrolyte, wherein bromine undergoes reduction at the cathode (carbon electrode), which is produced via comproportionation of bromate forming bromide in the process, resulting in a Br2/Br- redox couple near the electrode vicinity, exchanging 6 electrons per bromate ion present in the solution. There has been some interesting research by Y. V. Tolmachev, highlighting an anomalous nature of the halogen redox couple characterized by the mechanism mentioned above, leading to decrease in limiting current with increase in the diffusion layer thickness, demonstrated by RDE tests [1]. This comprised of the EC’’ mechanism which explained the distinct behavior of the system under varying condition of the dimensionless coordinate, defined as the ratio of diffusion to kinetic layer thickness, categorized into three domains, within each of which a different behavior is seen [2-4]. As the analysis of the system was done primarily on a half cell, within the RDE setup, our objective is to extend the same to a cell-based comprehensive study, with H2 as the fuel in the anode side. The cell-based experimentation on the system was done to primarily demonstrate the feasibility of the new system, which consisted of both polarization tests and discharge tests at the various C rates. Currents as high as almost 1 A/cm2 could be withdrawn from the system as shown in Figure 1, and the utilization were around 95% for various discharge C rates. Subsequently, a modeling aspect was added to the system, primarily to map the anomalous behavior of the system in a full cell setup. The results obtained can help bridge the gap between the RDE model and the cell-based system. Furthermore, this enables us to analyze losses associated with charge transfer, ohmic, and mass transfer, which will provide guideline to decide optimum operating conditions and design parameters of the full cell.
Read full abstract