The vanadium redox flow battery (VRFB) is one of the most promising energy storage technologies for large scale commercialization. The vanadium electrolyte is the main component, determining the VRFB’s energy density and capacity. The quality of the vanadium electrolyte is key to VRFB’s operation, as presence of soluble impurities will affect VRFB’s performance and durability. Understanding the impact of these impurities may ultimately lead to a specification for the impurity levels a VRFB can tolerate without compromising its performance and durability. Fe is an impurity element typically found in the vanadium ores, which is also present in the commercially available vanadium electrolytes. It was reported that Fe in positive electrolyte in a narrow concentration range (< 0.0286 M or 0.12 wt. %) could slightly affected the VRFB performance [1]. At high concentration (1.0 - 1.4 M Fe), Fe was reported to stabilize the positive electrolyte at high temperature (50oC) [2]. However, systematic studies on the effect of Fe on vanadium redox reactions, over a wider range of concentrations, are needed in order to achieve a good understanding of how it affects the VRFB performance. This work reports on the effect of Fe on vanadium redox reactions, with concentrations ranging from 0.05 wt. % or 0.012 M to 2 wt. % or 0.48 M, investigated by cyclic voltammetry and VRFB single cell cycling. In the present study, CV response of vanadium electrolyte on glassy carbon, Pt disk and graphite electrode was compared. On glassy carbon and Pt disk electrodes, vanadium redox reaction has shown irreversible or partially reversible CV response, while on graphite rod electrode, reversible redox reaction response was observed. Thus, a graphite rod was selected as working electrode for the subsequent investigations. The cyclic voltammogram on graphite rod electrode is dependent on electrode pre-treatment. Redox peak current, and peak separation are different on freshly polished graphite electrode compared to an electrode that has been used for a certain period of time (e.g. half hour). The difference is dependent on Fe concentration. In 0.05 wt. % (or 0.012 M) Fe electrolyte, electrode passivation was observed (Fig. 1), where the freshly polished electrode shows higher redox peak current than the one that has been used for CV for a certain time. However, in electrolytes with higher Fe content (e.g. 0.5 wt. % or 0.12 M), the freshly polished electrode presents lower redox peak current, indicating that electrode activation can occur during CV testing (Fig 2). VRFB performance was further evaluated, and it was found that the effect of Fe on VRFB capacity, capacity change profile with cycling, and efficiency is dependent on Fe concentration. Low Fe concentration affected more on the efficiency, while higher Fe concentration shows significant effect on capacity change during cycling. AC impedance and vanadium crossover were used to diagnosis the VRFB performance and degradation, and it was found that Fe concentration affects VRFB degradation and water transfer. The tolerance level of Fe in vanadium electrolyte can be deduced from this study, which may provide guidance on the design of low purity vanadium electrolyte. References: M. Ding, T. Liu, Y. Zhang, Z. Cai, Y. Yang, Y. Yuan, Effect of Fe(III) on the positive electrolyte for vanadium redox flow battery. R. Soc. Open Sci. 6 (2019) 181309. Z. Li, Y. Lin, L. Wan, B. Wang, Stable positive electrolyte containing high-concentration Fe2(SO4)3 for vanadium flow battery at 50 oC, Electrochim. Acta 309 (2019) 148 – 156 Figure captions: Fig 1 Cyclic voltammogram of graphite rod electrode (Red curve: electrode surface was polished before the CV measurement, Blue curve: electrode surface was polished then measure CV at a variety of scan rates, no polish was performed before each scan rate) in 1.6 M VOSO4 in 2 M H2SO4 containing 0.05 wt. % Fe (0.012 M). Scan rate: 10 mV/s Fig 2 Cyclic voltammogram of graphite rod electrode (Red curve: electrode surface was polished before the CV measurement, Blue curve: electrode surface was polished then measure CV at a variety of scan rates, no polish was performed before each scan rate) in 1.6 M VOSO4 in 2 M H2SO4 containing 0.5 wt. % Fe (0.12 M). Scan rate: 10 mV/s. Figure 1