Polymer electrolyte fuel cells (PEFCs) are already put into the market as the power source for hydrogen fuel cell vehicles (FCVs), but further technical advancement is necessary to achieve both higher cell performance and lower cost required for wider distribution. Regarding the latter requirement, a precious metal-free cathode is an attractive alternative and various candidates have been examined such as nonprecious metal-containing or metal-free carbon-based catalysts, metal oxides, oxynitrides, and metal chalcogenides [1]. In spite of extensive researches, however, the cell performances and stabilities using those materials are still significant lower than conventional Pt-based cells. Another approach for a precious metal-free cathode is to flow a catholyte containing oxidized-state redox mediator through porous carbon electrode. The system is similar to a redox flow battery, and the reduced-state mediator is oxidized with air in another compartment (we call this system as "redox flow fuel cell"; RFFC). ACAL Energy Inc. has reported comparable cell performance to Pt-based one by this concept and also excellent durability in a potential cycle test [2]. In this paper, a molybdovanado phosphoric heteropoly acid, a kind of polyoxometalate (POM) is mentioned as a mediator; however, the characteristics of the catholyte were not described in detail.

For better system efficiency, high cell performance must be achieved at high catholyte utilization, which is required to minimize the external pump energy consumption. High catholyte utilization, however, means low mediator oxidation states (i.e. state of charge; SOC) at the outlet. Therefore, the high cell performance even with low SOC catholyte is necessary. In this study, we examined the cell performances of RFFCs as functions of SOC using a molybdovanado phosphoric heteropoly acid as the mediator, and conducted ex-situ catholyte analysis using XAS, ³¹P-NMR, ⁵¹V-NMR and pH measurements to clarify the reactions in the catholyte of RFFCs.

Membrane electrode assemblies (MEAs) were constructed using NR212 perfluorinated membrane (DuPont), a precious metal-free carbon cloth for the cathode and a Pt/C catalyst layer coated MPL/GDL for the anode. The electrode area was 1cm² and the interdigitated carbon flow field was used for the cathode. The catholyte, 140 cc of 0.3 M aqueous solution of H₆PMo₉V₃O₄₀ (V3-POM) powder purchased from Nippon Inorganic Colour & Chemical Co., was circulated to the cathode with a tube pump at the flow rate of 14 cc/min. Fully humidified hydrogen gas was supplied to the anode at the flow rate of 100 cc/min. The cell performance was measured at the temperature maintained at 40˚C. The SOC of the catholyte was controlled by the time of period of constant current (0.3 A/cm²) discharge. Small amount of catholyte (~1 cc) was sampled at each SOC and used for analysis, XAS, ³¹P-NMR, ⁵¹V-NMR and pH measurements.

The obtained cell performance was shown in below. OCV exceeded 1.0V at SOC=100%, and iR-corrected cell voltage at 0.3A/cm² was about 0.86V. This performance is excellent as a precious metal-free cathode cell and comparable to a typical conventional Pt-based cathode cell. This result indicates that the RFFC has a good potential for a high energy conversion FC system. The ohmic resistance was found to be about 500mΩcm² by high frequency impedance measurements, five times as large as a typical conventional PEFC with a similar structure. The post analysis of MEA by cross-sectional EPMA revealed that membrane contained vanadium, but no molybdenum nor phosphor, suggesting decomposition of V3-POM. The dependence of OCV on SOC was found to be significantly larger than expected from the Nernst equation (E=E₀+RT/nF ln(x_ox/(1-x_ox); where x_oxis the concentration of the oxidized mediator). This discrepancy indicates that the redox reaction of the catholyte is not simple and that V3-POM is not suitable for a high utilization operation because the OCV at low SOC is too low to obtain a high cell voltage. The redox species was confirmed as vanadium by ex-situ XAS measurements conducted at BL33XU in SPring-8, Japan. The detail of reaction mechanism of V3-POM revealed by NMR and pH measurements will be discussed in the presentation.

Reference:

[1] M. Shao, Q. Chang, J. Dodelet, and R. Chenitz; Chem. Rev., (2016), 116, 3594–3657.

[2] A. Creeth; Fuel Cells Bulletin (2011) 12–15.