These days, in order to stably utilize renewable energy, there are increasing demands for large-capacity power storage devices with safe and inexpensive. An aqueous redox flow battery (RFB) has low risks to ignite and can get a greater capacity easily with increasing the number of electrolyte tanks. The most famous RFB is a vanadium-type one (V-RFB). It has high efficiency and long life durability [1-3]. However, an energy density of V-RFB is relatively low and a cost is high [4, 5]. An aqueous zinc/iodine RFB (Zn/I-RFB) can overcome those limitation because zinc compounds and iodine compounds have high solubilities, and zinc and iodine are inexpensive elements [6]. Here we report a single electrolyte-type Zn/I-RFB, which can reduce its cost owing to the elimination of an expensive ion-exchange membrane. The use of an inexpensive additive enables the separation of cathode charged products through the formation of polyiodide complexes. We will introduce a specific reaction, between polyiodide and additives, and also cycling performance of a single electrolyte-type Zn/I-RFB.

Electrolytes for evaluating cycle properties were prepared by dissolving 1.5 M NaI, 0.5 M ZnCl₂, and some inexpensive additives in ultrapure water. Figure 1 shows a scheme of a single electrolyte-type Zn/I-RFB. A Ti mesh and a carbon felt were used as a positive electrode, and a Zn plating Fe mesh was served as a negative one. The cell was designed to insert the porous membrane, between the positive electrode and the negative one, and had a flow channel at the side of the negative. Cycle properties were measured by galvanostatic charge-discharge tests at a current density of 20 mA/cm².

Figure 2 shows pictures of charging reactions on a positive electrode; (a) without and (b) with an additive. A hydrophobic liquid was observed in Figure 2 (b) as distorted black hemispheres.

The hydrophobic liquid was formed by the combination effects between cathode charged products and specific inexpensive additives. In charging, I⁻ in electrolytes is oxidized to triiodide ions (I₃⁻) and partially to iodine molecules (I₂) on a positive electrode [7, 8]. I₂ reacts chemically with I⁻ to form I₃⁻ which is to diffuse from the surface of the electrode (Figure 2 (a)). However, when the rate of I₂ electrochemical formation exceeds that of the chemical reaction between I₂ and I⁻, it is considered that I₂-films are formed on an electrode. These I₂-films were considered to increase a resistance and to impede the charging reactions. An additive employed in this study was to accelerate the removal of the I₂-films and formed hydrophobic liquid (Figure 2 (b)). The removal of the I₂-films from the surface of an electrode could increase the charging currents by over 50%. The hydrophobic liquid could be also reduced because the hydrophobic liquid could contain charged products. That is, charge and discharge reactions in the hydrophobic liquid could take place.

In addition, the hydrophobic liquid sank and accumulated to the bottom of an electrolytic cell. This feature is of great importance from a viewpoint of separating the charged products between a positive electrode and a negative one. When a positive electrode is allocated at the bottom side of a cell and a negative one is the top side, charged products of both the anode and the cathode can be spatially separated without an ion-exchange membrane (Figure 1). The anode charged product is zinc metal which stays on the negative electrode, and cathode charged products sunk to the bottom side don’t contact the zinc.

This finding described above was verified by galvanostatic charge-discharge tests. Figure 3 indicates the result of charge-discharge tests with a prototype cell designed on the concept here. The inserted figure in Figure 3 shows the three kinds of efficiencies. The stable cyclability of the charge-discharge tests was achieved over the course of 50 cycles. Through almost all of 50 cycles, over 90 % of coulombic efficiencies were achieved without an ion-exchange membrane. This result suggests that this single electrolyte-type Zn/I-RFB is considered to be a promising candidate for large-capacity power storage devices.

[1] M. Skyllas-Kazacos et al., J. Electrochem. Soc., 133, 1057 (1986).

[2] M. Rychcik et al., J. Power Sources, 22, 59 (1988).

[3] M. Skyllas-Kazacos et al., J. Electrochem. Soc., 158, R55 (2011).

[4] V. Viswanathan et al., J. Power Sources, 247, 1040 (2014).

[5] M. Zhang et al., J. Electrochem. Soc., 159, A1183 (2012).

[6] B. Li et al., Nature Commun., 6, 6303 (2015).

[7] T. Bejerano et al., J. Electrochem. Soc., 124, 1720 (1977).

[8] L. Ma et al., J. Electrochem. Soc., 146, 4152 (1999).