With the exponential growth of electricity generation using intermittent renewable power sources, there is a need for affordable energy storage. For grid level storage in past years, redox flow batteries (RFB) have drawn considerable attention. Most successful RFBs till now are all-vanadium RFB and all-iron RFB. But these RFBs have a few drawbacks associated with them like; First RFB developed by NASA ^[1], Iron-chromium RFB, had problems of ions crossing over; all-iron RFB ^[2] have a problem of hydrogen evolution reaction (HER) at the negative electrode which causes capacity decay; and the commercialized all-vanadium RFB's ^[3] is expensive and has limited vanadium reserves.

In an effort to address these problems, we present here the initial development of an all-tungsten RFB. Its key advantage is no capacity fading with cycles due to ions crossing over. The tungsten salt used was phosphotungstic acid (PTA). As shown in cyclic voltammograms (CV) of PTA in Figure 1(a), it is clear that PTA ions exist in four different oxidation states (PTA^3-, PTA^4-, PTA^5- and PTA^6-). For all-PTA redox flow battery experiments, we used the same PTA^5- active species for both negative as well as positive electrolyte. Prior to the all-PTA RFB tests, PTA^3- was reduced to PTA^5-. Since PTA^3- is the only salt available, it was reduced to PTA^5- by using it as a negative electrolyte for the iron-tungsten RFB using a procedure developed earlier ^[4].

As shown in reaction 1, during charging on the negative side, PTA^5- gains 2 electrons and reduces to PTA^6- oxidation state. Simultaneously, on the positive side PTA^5- loses 2 electrons and gets oxidized to PTA^3-. Moreover, the maximum potential difference that can be achieved by using the highest reduced form of phosphotungstic acid as a single cell is +0.611 V vs. RHE (Eq. 4). Although we are getting half the potential when compared to an all-vanadium or an all-iron battery. But due to its extremely fast kinetics and large discharge currents makes all-PTA RFB a promising alternative. For the large capacity of the battery, one of the deciding factors is the solubility of its active species salts and the maximum solubility of PTA^3- is 0.7 M.

Flow cell experiments were conducted using an in-house constructed standard fuel cell setup ^[4]. Meanwhile, CV experiments were conducted in 3 electrode setup using glassy carbon as working electrode, platinum mesh as counter electrode and mercury/mercurous sulphate (Hg/Hg₂SO₄) as reference electrode. Hg/Hg₂SO₄ was calibrated in 0.5 M H₂SO₄ using an in house RHE. Also, all the potentials mentioned here are with respect to RHE. A CV (see Figure 1(a)) for phosphotungstic acid was obtained which showed 3 distinct peaks corresponding to the 3 reactions. Reactions 2 and 3 involve transfer of 1 electron each, while reaction 1 requires 2 electrons as proven from the current peaks of the CVs as well as limiting currents obtained from their RDEs ^[5]. The involvement of parasitic HER is clear from the 3^rd reduction peak. Kinetic parameters such as rate constant (k⁰), Tafel slope, diffusion coefficients (D_o) and transfer coefficient (α) were determined using RDE studies. From RDE analysis at different RPM, rate constant for PTA^3- was found to be 4.49 x 10^-3 cm s^-1.

Meanwhile, for the flow cell experiments we have taken electrolyte solution concentrations similar to that for the CVs. The flow rate (120 mL min^-1) and volume (40 mL) for both the electrolyte were kept same. As shown in Figure 1(b), the all-PTA battery was charged and discharged using chronopotentiometry at a current density of 50 mA cm^-2. Furthermore, to obtain the i-V performance curve the cell was discharged from 2 mA to 1000 mA for 10s. The potential at the end of 10^th second was reported to eliminate the pseudo steady state ^[6]. i-V and power density performance curve was plotted from these current and potential values (see Figure 1(c)).

From just 40mM concentration of PTA^5- the maximum volumetric capacity we got is around 200 mAh L^-1, the maximum power density obtained was 10 mW cm^-2 and the maximum current that can be drawn was 85 mA cm^-2. The above results strongly suggest that the all-PTA RFB could indeed be a promising alternative.