ARTICLE https://doi.org/10.1038/s41467-019-12363-7 OPEN Catalytic production of impurity-free V3.5+ electrolyte for vanadium redox flow batteries Jiyun Heo1, Jae-Yun Han2, Soohyun Kim1, Seongmin Yuk1, Chanyong Choi1, Riyul Kim1, Ju-Hyuk Lee1, Andy Klassen3, Shin-Kun Ryi2* & Hee-Tak Kim 1,4* The vanadium redox flow battery is considered one of the most promising candidates for use in large-scale energy storage systems. However, its commercialization has been hindered due 1234567890():,; to the high manufacturing cost of the vanadium electrolyte, which is currently prepared using a costly electrolysis method with limited productivity. In this work, we present a simpler method for chemical production of impurity-free V3.5+ electrolyte by utilizing formic acid as a reducing agent and Pt/C as a catalyst. With the catalytic reduction of V4+ electrolyte, a high quality V3.5+ electrolyte was successfully produced and excellent cell performance was achieved. Based on the result, a prototype catalytic reactor employing Pt/C-decorated carbon felt was designed, and high-speed, continuous production of V3.5+ electrolyte in this manner was demonstrated with the reactor. This invention offers a simple but practical strategy to reduce the production cost of V3.5+ electrolyte while retaining quality that is adequate for high-performance operations. 1 Department of Chemical and Biomolecular Engineering, Korea Advanced Institute of Science and Technology, 291, Daehak-ro, Yuseong-gu, Daejeon, Republic of Korea. 2 Advanced Materials and Devices Laboratory, Korea Institute of Energy Research (KIER), 152 Gajeong-ro, Yuseong-gu, Daejeon, Republic of Korea. 3 Avalon Battery, 3070 Osgood Ct Fremont, Fremont, CA 94539, USA. 4 Advanced Battery Center, KAIST Institute for the NanoCentury, KAIST, 291, Daehak-ro, Yuseong-gu, Daejeon, Republic of Korea. *email: [email protected]; [email protected] NATURE COMMUNICATIONS | (2019) 10:4412 | https://doi.org/10.1038/s41467-019-12363-7 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12363-7 nvironmental concerns are driving the development and use potential, which are strong reducing agents, can reduce V4+ to of renewable energy sources with better quality and greater V3+, although with release of some ion impurities24. However, the E 1,2 quantity . However, due to the unsteady nature of ions negatively influence VRFB performance by metal deposition renewable energy sources, electrochemical energy storage (EES) is and consequently accelerate hydrogen evolution13. For these rea- in high demand to make practical use of renewable energy in grid sons, inventing a new greener and simpler method to replace applications3–5. The redox flow battery is a very promising system electrolysis when producing impurity-free vanadium electrolyte among the candidates for the use in EES due to its large energy promises to offer a significant advance in VRFB technology. capability, high safety, and flexible control of the energy-to-power Here, we present a large-scalable method for producing ratio6,7. Notably, an all-vanadium redox flow battery (VRFB) has impurity-free V3.5+ electrolyte using catalyzed oxidation of received much attention because it has high efficiency and long ORAs. The use of ORAs is beneficial for obtaining impurity-free 8,9 life time without concern about cross-contamination . Yet, the electrolyte because only water and easily removable CO2 are commercialization of VRFB is still hindered due to the expensive released through the reaction; however, the chemical reduction of cell components despite such advantages10. In particular, vana- V4+ to V3+ by the oxidation of ORAs remains unachieved due to dium electrolyte accounts for a large portion of VRFB costs the low reactivity of ORAs. To accelerate the reaction, we adopted because of the need for expensive vanadium precursor materials a catalytic reaction inspired by the oxidation of organic fuels in and the high cost of electrolyte production. For example, for direct organic fuel cells. Pt-based catalysts used for these fuel cells systems of 10 kW/120 kWh, the cost for vanadium and electrolyte lower the activation energy needed for the oxidation of organic production cost account for 40 and 41%, respectively, of the total fuels in acidic media. Similar to the fuel cell reactions, Pt catalysts energy cost11. Furthermore, the portion of the electrolyte cost in can provide facile decomposition of ORA and electron transfer the total VRFB cost increases with increasing energy capacity of a from ORA to V4+. Consequently, by adjusting the amount of system12,13. Therefore, cost-effective production of VRFB elec- ORA, V3.5+ electrolyte can be produced from V4+ electrolyte by trolyte must be developed to achieve broader acceptance of the catalytic reaction. As described in Fig. 1, the V4+ electrolyte VRFB13,14. injected into a catalytic reactor incorporating Pt-based catalyst is The V3.5+ electrolyte, which is an equimolar mixture of V4+ simply converted to V3.5+ electrolyte. This contrasts with the (VO2+) and V3+ electrolyte, is especially preferred in industry as conventional electrolysis method in that electric energy is not both positive and negative electrolytes because VRFB can be consumed and surplus vanadium electrolyte is not generated. operated without initial re-balancing of its positive and negative This emphasizes the merit of the catalytic production of V3.5+ capacity. A full charging of VRFB with the use of the same V3.5+ electrolyte. In this work, a concept for catalytic electrolyte pro- electrolyte for positive and negative electrodes results in V5+ and duction is proposed with rational material selections, and the V2+ electrolyte at the positive and negative electrodes, respec- feasibility of large-scale, continuous electrolyte production is fl tively. In most cases, V2O5 is commonly used as a vanadium proved with a prototype ow reactor. It was found the catalytic source for preparing V3.5+ electrolyte because of its low cost, production of V3.5+ electrolyte can reduce the production cost by compared with other vanadium precursors. The conventional 40% compared to conventional electrolysis method due to the 3.5+ route for preparing V electrolyte from V2O5 includes the process simplicity. 5+ + 4+ chemical reduction of V (VO2 )toV with a reducing agent and the electrolysis of V4+ electrolyte to produce V3+ electro- lyte15. The reduction of V5+ to V4+ can easily be achieved Results with a residue-free organic reducing agent (ORA) such as oxalic Catalytic production of V3.5+ electrolyte. For use as a reducing – + + acid16 18. However, the reduction of V4 to V3 with ORA is agent for V4+ solution, ORA should have a lower redox potential quite sluggish, which presents a major challenge for achieving than that of V4+/V3+ (0.34 V vs. standard hydrogen electrode + practical chemical production of V3.5 electrolyte. Therefore, (SHE)), no remaining residue after the reduction, and low cost. + instead of chemical reduction, electrolysis of V4 electrolyte has Considering the requirements, methanol, formic acid, and oxalic been employed using a VRFB stack17,19,20. As indicated by the acid are typically selected because their standard redox potentials + inventions from Skyllas–Kazacos’s group, the reduction of V4 are 0.0225, −0.03, and −0.43 V, respectively, and their oxidation electrolyte at the negative electrode can be coupled with either products are water and easily removable CO . As a proof of + 2 oxidation of V4 at the positive electrode or water splitting concept, the catalytic reduction of V4+ was first demonstrated reaction16,21. Preparing V3.5+ electrolyte by using V4+ electrolyte in the positive electrode during electrolysis ensures impurity-free production, but an additional reduction process for the surplus V4+ + 3+ 4.5 e– V V from the anode is required. On the other hand, the elec- CO2 + H O trolysis method based on water splitting reaction can prevent the 2 4.5+ generation of surplus V electrolyte. However, advanced HCOOH engineering may be needed to address the vanadium ion cross- over to oxygen evolution reaction (OER) electrode and the carbon Catalytic 22 production 3.5+ corrosion at OER electrode . Catalytic V 4+ reactor + H O In search of simpler production, a few chemical processes have V CO2 2 been attempted. The use of V2O3 enables the production of impurity-free V3+ electrolyte;18 however, the slow dissolution rate 23 and high cost of V2O3 restrict its application. Tanaka et al. invented a method by which to prepare a V3+ electrolyte by Electrolysis 3.5+ Low production V mixing V2O5 and sulfur, followed by calcination process to form a VRFB + V4+ stack 3 4.5+ Surplus soluble V compound. Although this method enables the che- V 3.5+ 4+ mical production of V electrolyte from V2O5, its complexity, V high temperature processing conditions (200–300 °C), and the 3.5+ possibility of toxic SO2 gas generation inhibit its application. Aside Fig. 1 Schematics for producing the V electrolyte. The catalytic reaction from the ORAs, alkali or transition metals with low redox (top) and conventional electrolysis (bottom) 2 NATURE COMMUNICATIONS | (2019) 10:4412 | https://doi.org/10.1038/s41467-019-12363-7 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-12363-7 ARTICLE ab2.0 c500 +4 +3.5 100 ) –1 450 1.5 80 +1 min min 50 °C +5 min 60 –1 400 1.0 +15 min 60 °C 70 °C +60 min 40 350 80 °C 0.5 Conversion (%) Without Pt/C Absorbance (a.u.) 20 300 TOF (mmol g 0 0.0 250 400 600 800 1000 0 102030405060 40 50 60 70 80 90 Wavelength (nm) Time (min) Temperature (°C) Fig. 2 Catalytic reduction of V4+ to V3.5+. a Ultraviolet–visible (UV–Vis) spectra of the reactant electrolyte at different reaction times at 50 °C (inset: color changes of the reactant with the catalytic reaction).