Non-aqueous Vanadium Redox Flow Battery Research Articles

Mixed Solvents to Optimize the Energy Density of Nonaqueous Vanadium Redox Flow Batteries

Nonaqueous redox flow batteries have the potential to be more energy-dense than their aqueous counterparts. The wider electrochemical stability windows of organic solvents enable higher-voltage redox reactions, which have already been demonstrated to improve molar energy density (Wh/molactives). Exceeding the theoretical maximum volumetric energy density (Wh/L) of state-of-the-art commercial active liquids (37.5 Wh/L for aqueous vanadium at 1.4 V and 2.0 M, a standard concentration [1]) requires increasing the active-species concentration past a threshold that is inversely proportional to the nonaqueous cell’s potential. Vanadium acetylacetonate (V(acac)3) is a potentially useful active species for nonaqueous flow batteries because its disproportionation delivers 2.2 V [2].As well as containing an electrochemically active species, flow-battery electrolytes must also contain a supporting salt, whose ions maintain electroneutrality during cell operation by travelling across the separator membrane. The minimum ionic strength needed in the electrolyte to maintain this charge balance up to the maximum theoretical state of charge depends on the number of charges received by the active species across all of its oxidation states. This minimum support concentration may be different from the optimal concentration of salt required to achieve sufficient conductivity.For a given nonaqueous electrolyte, optimization for a given active-species chemistry to achieve maximum volumetric energy density requires careful selection of solvent systems and supporting electrolyte, subject to the constraints of solubility limits (at all oxidation states of the active species), electroneutrality, and conductivity [3]–[5].In this work we use the well-established disproportionation chemistry of the V(acac)3 active species, with tetraethylammonium tetrafluoroborate (TEABF4) as supporting salt. The solubility limit of V(acac)3 in several solvents is determined densitometrically [4]. Promising binary mixtures of mutually miscible solvents are explored further, with the aim of choosing one solvent that preferentially solvates V(acac)3 and the other that solvates TEABF4. Optimization of the binary electrolyte identifies a solvent system of acetonitrile and dioxolane which offers a maximum V(acac)3 solubility of 0.95 M and a theoretical energy density of 29.3 Wh/L. The reversible electrochemistry of this mixture is confirmed with voltammetry, and stable cyclability is demonstrated in a benchtop flow cell [6]. Key performance metrics will be reported, including energy, voltaic, coulombic, and utilization efficiencies, electrolyte conductivity, and peak power density.While the results for the V(acac)3/ TEABF4 system are promising, this work provides a logical workflow and strategy for maximizing energy density that can be applied to other nonaqueous chemistries to achieve their true potential.[1] C. Minke and T. Turek, ‘Materials, system designs and modelling approaches in techno-economic assessment of all-vanadium redox flow batteries – A review’, J. Power Sources, vol. 376, pp. 66–81, Feb. 2018.[2] Q. Liu, A. E. S. Sleightholme, A. A. Shinkle, Y. Li, and L. T. Thompson, ‘Non-aqueous vanadium acetylacetonate electrolyte for redox flow batteries’, Electrochem. Commun., vol. 11, no. 12, pp. 2312–2315, 2009.[3] M. O. Bamgbopa, N. Pour, Y. Shao-Horn, and S. Almheiri, ‘Systematic selection of solvent mixtures for non-aqueous redox flow batteries – vanadium acetylacetonate as a model system’, Electrochimica Acta, vol. 223, pp. 115–123, 2017.[4] A. A. Shinkle, T. J. Pomaville, A. E. S. Sleightholme, L. T. Thompson, and C. W. Monroe, ‘Solvents and supporting electrolytes for vanadium acetylacetonate flow batteries’, J. Power Sources, vol. 248, no. January 2015, pp. 1299–1305, 2014.[5] T. Herr, J. Noack, P. Fischer, and J. Tübke, ‘1,3-Dioxolane, tetrahydrofuran, acetylacetone and dimethyl sulfoxide as solvents for non-aqueous vanadium acetylacetonate redox-flow-batteries’, Electrochimica Acta, vol. 113, pp. 127–133, Dec. 2013.[6] J. D. Saraidaridis and C. W. Monroe, ‘Nonaqueous vanadium disproportionation flow batteries with porous separators cycle stably and tolerate high current density’, J. Power Sources, vol. 412, pp. 384–390, Feb. 2019.

Fabrication of a composite anion exchange membrane with aligned ion channels for a high-performance non-aqueous vanadium redox flow battery.

Fabrication of high-conductivity ion exchange membranes (IEMs) is crucial to improve the performance of non-aqueous vanadium redox flow batteries (NAVRFBs). In the present work, anion exchange membranes with high-conductivity were fabricated by aligning ion channels of the polymer electrolyte impregnated in porous polytetrafluoroethylene (PTFE) under electric fields. It was observed that the ion channels of the polymer electrolyte were uniformly orientated in the atomic-force microscopy image. Its morphological change could minimize detouring of the transport of BF4− ions. The results showed through-plane conductivity was improved from 12.7 to 33.1 mS cm−1. The dimensional properties of the fabricated membranes were also enhanced compared with its cast membrane owing to the reinforcing effect of the substrate. Especially, the NAVRFB assembled with the optimized membrane showed increased capacities, with a 97% coulombic efficiency and 70% energy efficiency at 80 mA cm−2. Furthermore, the optimized membrane made it possible to operate the NAVRFB at 120 mA cm−2. Its operating current density was 120 times higher than that of a frequently used AHA membrane for RFBs.

Highly selective composite membranes using ladder-like structured polysilsesquioxane for a non-aqueous redox flow battery

Nanoporous composite membranes coated on the microporous support with a ladder-like ionic polymer were prepared and the structural selectivity effect of the membrane on a non-aqueous vanadium redox flow battery (RFB) was investigated. Crosslinkable ladder-like ionic polysilsesquioxane (L-PSQ) was synthesized and a Celgard 2400 membrane was used for the support. An L-PSQ/Celgard composite membrane was prepared by the thermal crosslinking reaction of crosslinkable L-PSQ after the dip-coating process. The permeability of ionic L-PSQ/Celgard composite membranes decreased by two orders of magnitude upon thin coating while maintaining its high ion conductivity of the porous support, which provides higher charge/discharge capacity and cell efficiencies of RFB at all current densities compared to those of the RFB with the porous Celgard 2400 membrane. Due to the reduction of the crossover, the coulombic and energy efficiencies of non-aqueous vanadium RFB increase and reached 62.7% and 43.7%, respectively, for the RFB with the membrane containing 5.4 wt% of L-PSQ. These values are higher than those obtained for the Celgard membrane and the commercial Neosepta AHA membrane. This result implies that structure control of the membrane is a practical solution for the adaptation of the non-aqueous RFB for the future energy storage systems.

A model study on effects of vanadium ion diffusion through ion exchange membranes in a non-aqueous redox flow battery

A simple charge-discharge model of a nonaqueous vanadium redox flow battery (VRFB) was introduced in this study to analyze the diffusion coefficients of active materials permeating an electrolyte membrane. Using this model, the effects of diffusion coefficients on the long-term performance, efficiency, and capacity of the system have been studied. Based on a literature review, the diffusion coefficients of approximately 10−7–10−6 cm2 s−1 were applied and the results were analyzed to determine relevant parameters for this study. Through this study, it has been confirmed that the capacity of the system was maintained at 100% under ideal conditions in the long-term operation, whereas that of the system reduced by more than half when only one vanadium species moved and is 10 mAh when all vanadium species moved. In addition, the simulation clarified the effect of diffusion coefficients on coulombic efficiency. In most conditions of the various diffusion coefficients, coulombic efficiency enhanced through cycle had been operated. In the case of voltage efficiency, the value was not largely changed regardless of diffusion coefficients since it was affected more by other factors. One important factor revealed through this model study is that electrolyte membrane performance of the VRFB system was not only affected by the diffusion coefficient of a specific ion, but also by the relationship between the diffusion coefficients of all active ions involved in the reaction. This study provides guidance for the development of an electrolyte membrane to improve the performance of VRFB systems.

Investigation of Non-Aqueous Electrolyte and Crossover of Perfluorsulfonate Ion Exchange Membrane in a Non-Aqueous Flow Battery

From aqueous to non-aqueous redox flow batteries, the general membrane properties are unchanged: electronically isolating anolyte and catholyte and allowing conductance through the membrane to keep charge balance. However, on top of the high cost(1), existing membranes for non-aqueous flow battery inherit the natural disadvantage of non-hydrogen-bonding transport compared to all-vanadium redox system while leaving the door wide open for redox species to pass through. Crossover-induced capacity fade remains one of the most challenging problems for non-aqueous flow battery(2). Efforts have been made to address the issue. For example, composite membranes(3), crosslinking(4) and size exclusion approaches (5) have all been used. Herein, we use Perfluorsulfonate Ion Exchange Membrane (PFSA) as a benchmark to study the crossover behavior in non-aqueous system. For membranes swollen with non-aqueous solvents, tetraalkylammonium cations in the membrane have outperform lithium ion in terms of ionic conductivity(6). However, a large volume of swelling is observed in our study and also in literature(3). In fact, our previous vibrational data suggested that only a small portion of the organic solvent in the membrane participates in solvation of cations, leaving a substantial portion of the solvent molecules as potential carriers to dissolve and transport redox species across the membrane(7). The solvent mediated process, crossover of redox species through PFSA will be presented from a fundamental, interaction-oriented perspective. In-depth understanding of interactions among redox species, solvent and membrane are obtained by combining various spectroscopic measurements with transport measurements. This understanding is crucially necessary to shed light on future membrane design exclusive for non-aqueous flow batteries. Acknowledgement We gratefully acknowledge the support of this work by the U.S. Department of Energy, Office of Electricity Delivery and Energy Reliability (Dr. Imre Gyuk). We also thank 3M for providing membranes. References Darling RM, Gallagher KG, Kowalski JA, Ha S, Brushett FR. Pathways to low-cost electrochemical energy storage: a comparison of aqueous and nonaqueous flow batteries. Energy Environ Sci. 2014;7(11):3459–77.Su L, Darling RM, Gallagher KG, Xie W, Thelen JL, Badel AF, et al. An Investigation of the Ionic Conductivity and Species Crossover of Lithiated Nafion 117 in Nonaqueous Electrolytes. J Electrochem Soc. 2016;163(1):A5253–62.Bamgbopa MO, Almheiri S. Influence of solvents on species crossover and capacity decay in non-aqueous vanadium redox flow batteries: Characterization of acetonitrile and 1, 3 dioxolane solvent mixture. J Power Sources. 2017 Feb 28;342:371–81.Li Y, Sniekers J, Malaquias JC, Van Goethem C, Binnemans K, Fransaer J, et al. Crosslinked anion exchange membranes prepared from poly(phenylene oxide) (PPO) for non-aqueous redox flow batteries. J Power Sources. 2018;378:338–44.Hendriks KH, Robinson SG, Braten MN, Sevov CS, Helms BA, Sigman MS, et al. High-Performance Oligomeric Catholytes for Effective Macromolecular Separation in Nonaqueous Redox Flow Batteries. ACS Cent Sci. 2018;4(2):189–96.Escalante-García IL, Wainright JS, Thompson LT, Savinell RF. Performance of a Non-Aqueous Vanadium Acetylacetonate Prototype Redox Flow Battery: Examination of Separators and Capacity Decay. J Electrochem Soc. 2015;162(3):A363–72.Lou K, Peng J, Tang Z, Zawodzinski TA. Solvation of Perfluorsulfonate Ion Exchange Membrane in Non-Aqueous Solvents. Meet Abstr. 2018;MA2018-01(30):1760–1760.

Insights on the Electrochemical Activity of Porous Carbonaceous Electrodes in Non-Aqueous Vanadium Redox Flow Batteries

Porous carbonaceous electrodes have the potential utility as electrodes for non-aqueous VRFBs due to being inexpensive, chemically stable, and electrochemically active. In this work, we provide insights on the electrochemical kinetics of [V(acac)3]0/[V(acac)3]+ and [V(acac)3]0/[V(acac)3]− redox reactions at freestanding sheets of multiwalled carbon nanotubes (MWCNTs), or buckypapers (BPs), and graphitized carbon fiber-based nonwovens papers (CPs). The use of freestanding sheets eliminates the need for a metal substrate; thus, the true electrochemical response at these electrodes is measured. Even though porous carbonaceous electrodes provide high surface area for the electrochemical reaction, their porous nature poses a fundamental challenge caused by thin-film-like diffusion process; leading to a marked ohmic resistance within the pores of the electrodes. Higher current densities are obtained with the BP electrode due to their larger surface area in comparison to the CP electrode; however, the BP electrode show less efficient kinetic facility for the investigated redox couples as evident by electrochemical impedance spectroscopy and polarization curves results. Charge–discharge cycling demonstrates that porous carbonaceous electrodes may require various treatments to modify the surface chemistries to obtain improved cycling efficiencies.

Influence of solvents on species crossover and capacity decay in non-aqueous vanadium redox flow batteries: Characterization of acetonitrile and 1, 3 dioxolane solvent mixture

Abstract The importance of the choice of solvent in a non-aqueous redox flow battery (NARFB) cannot be overemphasized. Several studies demonstrated the influence of the solvent on electrolyte performance in terms of reaction rates, energy/power densities, and efficiencies. In this work, we investigate capacity decay as a direct consequence of varying reactant crossover rates through membranes in different solvent environments. Specifically, we demonstrate the superiority of an 84/16 vol% acetonitrile/1,3 dioxolane solvent mixture over pure acetonitrile in terms of energy efficiency (up to 89%) and capacity retention for vanadium NARFBs - while incorporating a Nafion 115 membrane. The permeability of Nafion to the vanadium acetylacetonate active species is an order of magnitude lower when pure acetonitrile is replaced by the solvent mixture. A method to estimate relative membrane permeability is formulated from numerical analysis of self-discharge experimental data. Furthermore, tests on a modified Nafion/SiO2 membrane, which generally offered low species permeability, also show that different solvents alter membrane permeability. Elemental and morphological analyses of cycled Nafion and NafionSi membranes in different solvent environments indicate that different crossover rates induced by the choice of solvent during cycling are due to changes in the membrane microstructure, intrinsic permeability, swelling rates, and chemical stability.

Surface-modified porous membranes with electrospun Nafion/PVA fibres for non-aqueous redox flow battery

Selective layered membranes were fabricated using electrospun Nafion/(polyvinyl alcohol (PVA) fibres on porous membranes for non-aqueous vanadium redox flow battery applications (RFB). Porous Celgard 2400 polypropylene membranes were chosen as a porous support because they have a high ionic conductivity with excellent mechanical and dimensional stability in acetonitrile. The layers composed of the electrospun polyelectrolyte (Nafion) fibres on the porous support provide selectivity while maintaining the ion conductivity through the porous membrane. The membranes showed low vanadium acetylacetonate (V(acac)3) permeability in acetonitrile compared to that of the pristine Celgard porous membrane. The performance of a non-aqueous RFB based on V(acac)3 with electrospun Nafion/PVA fibres surface-modified porous membranes was evaluated and compared to that of Nafion 212 and Neosepta AHA membranes. Low coulombic and energy efficiencies of 49% and 32%, respectively, were obtained for pristine Celgard porous membranes, which are comparable to the values of Neosepta AHA membranes, and this low efficiency is due to the non-selectivity of the pore in the Celgard membrane. The values increase with the accumulated thickness of the electrospun Nafion/PVA fibres, with the coulombic and energy efficiencies reaching 69% and 50%, respectively, which are even higher than those of the Nafion 212 dense membranes. It is considered that porous supports contribute to the mechanical stability, and electrospun polyelectrolyte fibres accumulated on the surface reduce the crossover of the redox active species, resulting in the high performance of RFBs. This result implies that the electrospun modification of the mechanically stable porous membrane may enable the successful use of a non-aqueous RFB for future energy storage systems.

Counter anion effects in anion exchange membrane-fabricated non-aqueous vanadium redox flow battery

In order to understand the counter anionic effects in a non-aqueous vanadium redox flow battery (VRFB), we synthesized four types of electrolyte salts (1-ethyltriethamine tertafluoroborate, [E-TEDA]+[BF₄]?, 1-ethyltriethamine hexafluorophosphate, [E-TEDA]+[PF 6 ]?, 1-butyltriethylamine tertafluoroborate, [B-TEDA]+[BF₄]?, and 1-buthyltriethamine hexafluorophosphate [B-TEDA]+[PF 6 ]?) by counter anion exchange reaction after the SN₂ reaction. We confirmed the successful synthesis of the electrolyte salts [E-TEDA]+[Br]? and [B-TEDA]+[Br]? via ¹H-NMR spectroscopy and GC-mass analysis before the counter anion exchange reaction. The electric potential of the vanadium acetylacetonate, V(acac)3, as an energy storage chemical was shown to be 2.2 V in the acetonitrile solvent with each of the [E-TEDA]+[BF₄]?, [E-TEDA]+[PF 6 ]?, [B-TEDA]+[BF₄]?, and [B-TEDA]+[PF 6 ]? electrolyte salts. In a non-aqueous VRFB with a commercial Neosepta AFN membrane, the maximum voltages reached 1.0 V and 1.5 V under a fixed current value of 0.1 mA in acetonitrile with the [E-TEDA]+[BF₄]?and [E-TEDA]+[PF 6 ]? electrolyte salts, respectively. The maximum voltage was 0.8 V and 1.1 V under a fixed current value of 0.1 mA in acetonitrile with the [B-TEDA]+[BF₄]? and [B-TEDA]+[PF 6 ]? electrolyte salts, respectively. From these results, we concluded that in the non-aqueous VRFB more of the [PF 6 ]? counter anion than the [BF₄]? counter anion was transported onto the commercial Neosepta AFN anion exchange membrane.

Performance Evaluation of the Non-Aqueous Vanadium Redox Flow Battery with Different Separators

Non-aqueous redox flow battery (RFB) systems are being considered for grid-level energy-storage applications. Non-aqueous electrolytes exhibit wide electrochemical potential windows, generally > 2 V, which has a direct impact on the system energy and power density prospects. The increased cell potential is considered the primary advantage of non-aqueous RFBs over aqueous RFBs, which are limited by the electrochemical stability of water. Nevertheless, the development of non-aqueous RFBs is still at an early stage and thus significant research efforts are needed to achieve similar performance as in aqueous RFBs, especially, at the cell component level. The membrane or separator is one component that to a great extent affects the overall performance of RFB systems for practical applications. However, few reports have been found in the literature on the development and evaluation of separator materials for non-aqueous RFBs to date.In this work, a small-scale redox flow battery prototype was designed to simulate energy storage and to evaluate the performance parameters of non-aqueous redox electrolytes. Cation exchanged Nafion® ionomer and microporous Daramic separators were prepared and investigated. A non-aqueous, single-metal redox flow battery based on vanadium (III) acetylacetonate (V(acac)3) with a thermodynamic potential of 2.2 V was considered. At least a 70-fold increase in charge/discharge current density was achieved by using modified Nafion® or a microporous separator compared with current literature. Coulombic and round-trip energy efficiencies of 70-90% and of 65-80%, respectively, were obtained depending on the separator. Both the modified Nafion® and microporous separators are promising options for non-aqueous redox electrolyte systems.

Increasing the energy density of the non-aqueous vanadium redox flow battery with the acetonitrile-1,3-dioxolane–dimethyl sulfoxide solvent mixture

Different solvent mixtures were investigated for non-aqueous vanadium acetylacetonate (V(acac)3) redox flow batteries with tetrabutylammonium hexafluorophosphate as the supporting electrolyte. The aim of this study was to increase the energy density of the non-aqueous redox flow battery. A mixture of acetonitrile, dimethyl sulfoxide and 1-3-dioxolane nearly doubles the solubility of the active species. The proposed electrolyte system was characterized by Raman and FT-IR spectroscopy, cyclic voltammetry, electrochemical impedance spectroscopy and charge–discharge set-up. Spectroscopic methods were applied to understand the interactions between the solvents used and their impact on the solubility. The potential difference between oxidation and reduction of V(acac)3 measured by cyclic voltammetry was about 2.2 V. Impedance spectroscopy showed an electrolyte resistance of about 2400 Ω cm2. Experiments in a charge–discharge test cell achieved coulombic and energy efficiencies of ∼95% and ∼27% respectively. The highest discharge power density was 0.25 mW cm−2.

Anion exchange membrane prepared from simultaneous polymerization and quaternization of 4-vinyl pyridine for non-aqueous vanadium redox flow battery applications

A simple, single step and environmentally friendly process is developed for the synthesis of anion exchange membrane (AEM) by simultaneous polymerization and quaternization, unlike the conventional membrane synthesis which consists of separate polymerization and quaternization step. The membrane synthesis is carried out by dissolving polyvinyl chloride (PVC) in cyclohexanone along with 4-vinyl pyridine (4VP) and 1,4-dibromobutane (DBB) in the presence of thermal initiator benzoyl peroxide, followed by film casting to get thin and flexible AEMs. The membrane properties such as ion exchange capacity, ionic conductivity and swelling behaviour are tuned by varying the degree of crosslinking. These AEMs exhibit low vanadium permeability, while retaining good dimensional and chemical stability in an electrolyte solution, making them appropriate candidates for non-aqueous vanadium acetylacetonate redox flow battery (VRFB) applications. The optimized membrane displays ion exchange capacity and ionic conductivity of 2.0 mequiv g−1 and 0.105 mS cm−1, respectively, whereas the efficiency of 91.7%, 95.7% and 87.7% for coulombic, voltage and energy parameter in non-aqueous VRFB, respectively. This study reveals that the non-aqueous VRFB performance is greatly influenced by membrane properties; therefore the optimal control over the membrane properties is advantageous for the improved performance.

Stability of composite anion exchange membranes with various functional groups and their performance for energy conversion

In this study, we report the relative stabilities of anion exchange membranes (AEMs) comprised of different functional groups. The base membrane was synthesized from vinylbenzyl chloride (VBC) that was cross-linked by divinylbenzene (DVB) using a porous polyethylene (PE) substrate, and subsequently quaternized by functional groups including ammonium, diammonium, and phosphonium types. Every synthetic process was confirmed by FTIR spectra, and hydroxide ion conductivity, ion-exchange capacity, and water uptake. For the various functional groups, the membrane stabilities were examined under alkaline and accelerated oxidative conditions. The membranes were found to be reasonably stable under alkaline conditions, whereas the oxidative stability was significantly dependent on the structures of the functional groups. Moreover, an alkaline fuel cell test was performed at 60°C for selected membranes. Membranes quaternized with trimethylamine (PE-TMA) and triethylamine (PE-TEA) showed the highest power density, while PE-TEA exhibited a higher oxidative stability. Also, the PE-TEA composite membrane was successfully tested for a non-aqueous vanadium redox flow battery.

Tetrabutylammonium hexafluorophosphate and 1-ethyl-3-methyl imidazolium hexafluorophosphate ionic liquids as supporting electrolytes for non-aqueous vanadium redox flow batteries

Tetrabutylammonium hexafluorophosphate (TEAPF 6) and 1-ethyl-3-methyl imidazolium hexafluorophosphate (EMIPF 6) are synthesized and used as the supporting electrolytes of a non-aqueous redox flow battery using vanadium acetylacetonate (V(acac) 3) as the active species. The conductivity and cyclic voltammograms of the electrolytes of the two ionic liquids are measured. The cyclic voltammograms show that both of them are stable in an operating potential range (−2.5–1.5 V). The diffusion coefficients of V(acac) 3 in the electrolytes are determined as 0.92–1.47 × 10 −6 cm 2 s −1 in 0.5 mol l −1 TEAPF 6 and 2.35–3.79 × 10 −6 cm 2 s −1 in 0.5 mol EMIPF 6, respectively. The charge–discharge performances of the two non-aqueous V(acac) 3 containing electrolytes with TEAPF 6 and EMIPF 6 as the supporting electrolytes are evaluated in an H-type glass cell. The coulombic efficiencies are measured as in the range of 53.31–57.44% at 50% state of charge with an electrolyte containing 0.2 mol l −1 TEAPF 6, and as in the range of 46.04–43.46% for the electrolyte containing EMIPF 6 with the same concentrations of the components.