# Fileset

[ACS AEM-2020_3_4377-4383.pdf](https://mdr.nims.go.jp/filesets/45933153-ed1f-4e7f-9e95-38b7b5d68de2/download)

## Creator

Akihiro Ohira, Takashi Funaki, Erika Ishida, [Je-Deok Kim](https://orcid.org/0000-0003-4301-1044), Yukari Sato

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Redox-Flow Battery Operating in Neutral and Acidic Environments with Multielectron-Transfer-Type Viologen Molecular Assembly](https://mdr.nims.go.jp/datasets/3d1645c1-f65a-4ff3-97e3-0e6151479265)

## Fulltext

ae0c00067 1..7Redox-Flow Battery Operating in Neutral and Acidic Environmentswith Multielectron-Transfer-Type Viologen Molecular AssemblyAkihiro Ohira,* Takashi Funaki,* Erika Ishida, Je-Deok Kim, and Yukari SatoCite This: https://dx.doi.org/10.1021/acsaem.0c00067 Read OnlineACCESS Metrics & More Article Recommendations *sı Supporting InformationABSTRACT: Recently, redox flow batteries (RFBs) haveattracted attention as a large-scale energy storage technology. Toimprove their energy density, we investigated organic-based activematerials with high water solubility, synthesized regular dendriticstructures comprising viologen molecular assemblies, and appliedthem to RFBs. The compounds containing 3, 5, and 13 viologenmolecular units showed electrolysis activity corresponding to thenumber of units, and it was found that the redox reactionprogressed quantitatively. Furthermore, the charge and dischargecharacteristics confirmed that the energy efficiency was improvedcompared to methyl viologen batteries. These compounds havehigh solubility, and the viologen derivatives can function in acidicconditions, in which methyl viologen does not work as a redox active material, due to reduced interaction between molecules and thecation exchange membrane. Finally, molecules having 5 and 13 viologen units can be prepared to concentrations of 1 and 0.5 M,theoretically reaching a capacity of 134 and 174 Ah/L, respectively.KEYWORDS: redox flow battery, organic molecules, aqueous solution, molecular assembly, dendritic structure, viologen■ INTRODUCTIONThe introduction of power generation systems using naturalenergy, such as solar and wind power, into power grids is beingpromoted in various countries around the world. Althoughthey do not generate greenhouse gases, such as carbon dioxide,compared to power sources like thermal and hydroelectricpower, these renewable energy sources provide an unstablepower supply that is affected by various conditions, such aslocation and climate. To use renewable energy efficiently, thegenerated electricity must be stored and discharged as needed;thus, a storage battery (secondary battery) that can adjust theconstantly fluctuating output over short and long cycles isgreatly needed.Large-capacity storage batteries include lithium-ion batteries,sodium−sulfur batteries, lead-acid batteries, and redox flowbatteries (RFBs),1−3 and these storage batteries have beenused for distributed power systems in various places dependingon the size and application. Above all, although RFBs4,5 havetechnical issues, including energy density lower than that ofother storage batteries, and their vanadium content makesthem expensive, they have long life and high design flexibility.In addition, they operate at normal temperature and pose nodanger of thermal runaway or explosion. Therefore, theadvantages of RFBs have been attracting attention as oneway to achieve power leveling for renewable energy.RFBs were successfully developed by NASA around 1974.6At the beginning of development, the Fe/Cr system was theleading technology,7 but it was not put into practical usebecause mixing of the positive and negative electrolytesthrough the membrane decreases the capacity of the battery.Around 1984, Skyllas-Kazacos et al. developed a vanadium-based RFB,8 and since that time, the awareness of RFBs hasincreased and research and development has been accelerating.A RFB mainly comprises an electrode, an electrolytemembrane, and an electrolyte. A carbon felt or carbon paperis used as the electrode, and to improve the reactivity with theelectrolytic solution, metal particles can be incorporated intothe carbon as a catalyst. The electrolyte membrane preventsmixing of the active materials of the positive and negativeelectrodes, and it is responsible for the ion transport andcharge compensation associated with charge and discharge.Vanadium RFBs mainly use a proton-conducting membrane,such as Nafion. In recent years, materials research onelectrolyte membranes, electrodes, and diffusion layers, whichwere cultivated during research and development of polymerelectrolyte fuel cells, has improved cell design technology andbattery performance.9Received: January 12, 2020Accepted: April 16, 2020Published: April 16, 2020Articlewww.acsaem.org© XXXX American Chemical SocietyAhttps://dx.doi.org/10.1021/acsaem.0c00067ACS Appl. Energy Mater. XXXX, XXX, XXX−XXXDownloaded via NATL INST ADVNCD IND SCI & TECHLGY on April 29, 2020 at 23:45:27 (UTC).See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.https://pubs.acs.org/action/doSearch?field1=Contrib&text1="Akihiro+Ohira"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Takashi+Funaki"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Erika+Ishida"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Je-Deok+Kim"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Yukari+Sato"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/showCitFormats?doi=10.1021/acsaem.0c00067&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?goto=articleMetrics&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?goto=recommendations&?ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?goto=supporting-info&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=agr1&ref=pdfwww.acsaem.org?ref=pdfhttps://pubs.acs.org?ref=pdfhttps://pubs.acs.org?ref=pdfhttps://dx.doi.org/10.1021/acsaem.0c00067?ref=pdfhttps://www.acsaem.org?ref=pdfhttps://www.acsaem.org?ref=pdfMore recently, research on alternative technology, such asorganic, Ti−Mn,10 or hybrid RFBs,11−13 has become active,and in terms of performance, some alternative RFBs, especiallyorganic RFBs, have become comparable to high-cost vanadiumRFBs. One of the greatest features of organic RFBs is that it ispossible to increase the energy density by controlling thesolubility and redox potential based on molecular design, andpioneering and unique research has been reported so far.14−20The electron transfer reaction rate of the active material in anorganic RFB is larger than that in a vanadium RFB, and thereactivity with the carbon electrode is relatively good. For thisreason, continued development of RFBs is expected to providea new power storage technology that can flexibly cope withbeing combined with other secondary batteries and hydrogenproduction technologies.Viologen and its derivatives have been widely studied asmodel compounds for photoelectrochemistry,21−25 and inrecent years have also been studied as organic active materialsin RFBs.25−32 Promising characteristics of viologen com-pounds for RFBs include that they can be applied to bothaqueous and nonaqueous systems by means of counterionspecies, have relatively high solubility, and have high redoxstability. Janoschka et al. designed an active material group inwhich viologen molecules are introduced into a polymerbackbone, and they reported the possibility of improved energydensity.26 Beh et al. designed highly water-soluble viologen andferrocene derivatives as anolytes and catholytes, respectively,for aqueous RFBs. They also reported that organic RFBs showexcellent charge−discharge cycling performance comparable tothat of vanadium RFBs.27 Another study reported that chargeand discharge cycles are stabilized by changing the functionalgroup bonded to the viologen molecule.28 Burgess et al.synthesized a double-headed molecule containing a viologenunit and discussed the effect of the backbone structure on theelectrochemical performance.33Currently, we are focusing on viologen units and have newlysynthesized an assemblage of viologen molecules withrelatively high symmetry and a regular structure. A regularFigure 1. Chemical structures of four generations of viologen molecular assemblies.ACS Applied Energy Materials www.acsaem.org Articlehttps://dx.doi.org/10.1021/acsaem.0c00067ACS Appl. Energy Mater. XXXX, XXX, XXX−XXXBhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig1&ref=pdfwww.acsaem.org?ref=pdfhttps://dx.doi.org/10.1021/acsaem.0c00067?ref=pdfstructure, such as a highly symmetrical dendrimer, can increasethe solubility and viscosity of a solution, accompanied by anincrease in molecular weight.34 We are adopting that conceptfor active material design and aim at improving RFBperformance by increasing both the solubility of the designedviologen assembly and the redox response of the individualintroduced viologen molecules. Here, we report the applicationof this idea to aqueous RFBs using newly synthesized viologenmolecular units for an anolyte with a relatively high symmetryand regular structure.Here, we report the use of methyl viologen (MV2+) as thefirst generation (G1) and synthesis of compounds to which 3(G2), 5 (G3), and 13 (G4) viologen molecular units werebound (Figure 1). These compounds, which have a chloridecounterion and relatively high water solubility, were used forelectrochemical measurements and charge−discharge tests.The solubility of those compounds in 1 M NaCl aqueoussolution is 0.5 M for G2, 1 M for G3, and 0.5 M for G4. In thecases of G3 and G4, solubilities of 1 and 0.5 M are equivalentto 5 and 6.5 M when converted to per-viologen-unit values,respectively, and those values correspond to theoreticalcapacities of 134 and 174 Wh/L as dication states per unit,respectively. These results clearly indicate that viologen unitsdo not aggregate when they are assembled in a regularframework with relatively high symmetry, which suppresses thedecrease of molecular solubility.■ ELECTROCHEMICAL PROPERTIES OF VIOLOGENDERIVATIVESCyclic voltammetry using a glassy carbon electrode as theworking electrode revealed two peaks on each of the oxidationand reduction sides of each sample. The peak on the positiveside indicates a one-electron redox reaction, and the peak onthe negative side indicates a two-electron redox reaction(Figure S13). As the two-electron reaction progresses, theglassy carbon surface developed a film-like appearance. Thisindicates that the loss of charge desorbed Cl− counterions,resulting in a decrease in solubility and aggregation on theelectrode surface.Table 1 shows the viscosity of the solutions having relativelylow and high concentrations of G1, G2, G3, and G4 in D2O,and the cell resistance of the sample solutions prepared so thatthe viologen unit was equivalent to 0.1 M. Since the cellresistance is the same for the type of the carbon electrode andthe membrane used, these comparisons reflect the resistance ofthe sample solution.At relatively low concentrations (0.5 M), the increase inviscosity of G3 and G4 is small relative to the increase inmolecular weight, even when compared to G1. At higherconcentrations, the increase in viscosity tends to be greaterthan at lower concentrations, but interestingly, there is nosignificant difference in viscosity between G3 and G4. Thisindicates that the introduction of the dendritic-type regularstructure might be effective in controlling the viscosity of theelectrolyte, which is one of the causes of loss of systemefficiency in the flow battery.To conduct electrochemical measurements closer to RFBoperating conditions, a carbon felt (diameter 0.5 cm) andelectrolyte membrane (Selemion DSV or Nafion NR212) wereplaced in a small test cell, and the redox property of activematerials as catholytes and anolytes was evaluated (FigureS15). Figure 2 shows the current potential curves of the foursynthesized viologen molecular assemblies. The value of E1/2 isaround −0.64 V (vs Ag/AgCl) for G1, whereas G2, having 3viologen units, is around −0.49 V and G3, having 5 units, isaround −0.47 V. Moreover, G4, with 13 units, is shifted to thenoble potential side, around −0.44 V. Because the viologenunit in the framework has an ethyl group and a benzyl groupbonded together, the redox potential is thought to have shiftedto the positive side compared to G1.The amount of charge from the current−potential curve inFigure 2 was estimated to obtain the number of reactionelectrons from the ratio of the actual amount of electrolysisand the active material concentration present in the electrode(Figure S16). As a result, the ratio of the actual reactant andactive material concentration in the electrode was 0.98 for G1,2.8 for G2, 4.1 for G3, and 14 for G4. If one electron isinvolved in the redox reaction in each unit, almost all of theintroduced viologen units contributed to the one-electronredox reaction. These results indicate that a multielectronreaction system corresponding to the number of introducedmolecules can be realized.The absorption coefficients obtained from the UV−visabsorption spectra of the G3 and G4 aqueous solutions (FigureS17) are 21 100 M−1 cm−1 and 20 900 M−1 cm−1 whenconverted to (viologen) unit equivalents. These values areclose to those reported for MV2+.35 This indicates that theviologen unit introduced into the molecular assembly does notaggregate within the framework, and each behaves in isolation,and this also supports the results that all of the introducedviologens can contribute to the redox reaction.Heinen et al. reported a spectroelectrochemical analysis ofreaction mechanisms in a dendrimer structure linked with aviologen unit.36,37 They synthesized nonaqueous compoundsTable 1. Viscosity and Cell Resistance of G1, G2, G3, andG4 SolutionaG1 G2 G3 G4viscosity at 0.5 M (mPa·s) 1.6 2.1 2.3 3.0viscosity at 2.0 M 6.3 34 28cell resistance (Ω·cm2) at 0.1 M 1.5 1.5 2.8 2.5aThe sample concentrations shown here are equivalent to viologenunits. 2 M equivalent of G2 cannot be prepared due to low solubility.Figure 2. Current−potential (I−E) curves of viologen molecularassemblies. Scan rate: 1 mV/s. Concentration of active material: 10mM G1, 3.5 mM G2, 2 mM G3, and 0.77 mM G4. Supportingelectrolyte: 1 M NaCl. Counter electrode: Carbon felt (AAF304ZS).Reference electrode: Ag/AgCl.ACS Applied Energy Materials www.acsaem.org Articlehttps://dx.doi.org/10.1021/acsaem.0c00067ACS Appl. Energy Mater. XXXX, XXX, XXX−XXXChttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig2&ref=pdfwww.acsaem.org?ref=pdfhttps://dx.doi.org/10.1021/acsaem.0c00067?ref=pdfwith a PF6− counterion, which have a structure similar to ourG2, G3, and G4 molecular structures. According to theirreport, G3 and G4 had a fast reduction reaction, and theoxidation reaction occurs in stages in the core phenyl and outerbenzyl viologen (BV2+) units. In the G3 and G4 current−voltage measurements in Figure 2, the reduction reaction is asingle peak, whereas the oxidation reaction is split, as reportedby Heinen et al.37 Similar to nonaqueous systems, the reactionrate in an aqueous solution system would also be differentbetween the reduction and oxidation side.Interestingly, G2, G3, and G4 were found to show reversibleredox reactions even in acidic solutions with Nafion as theelectrolyte membrane (Figure S18). In the case of G1, only areduction reaction was observed. To investigate the cause ofthe difference in redox behavior in acidic solutions, a crossovertest between G1 and G2 was performed using Nafion. Themobility of G1 and G2 between acidic solutions at the sameconcentration on both sides of the Nafion membrane wascompared. There was a tendency for MV2+ to move to theopposite side over time compared to G2 (Figure S19). Incontrast, in the case of G2, it was found that movement to thesulfuric acid solution on the opposite side seemed to besuppressed.When Nafion is hydrated, the cluster structure in whichperfluorosulfonic acid groups in Nafion are assembled becomeslarger due to the inclusion of water molecules, and the waterchannel structure is 3.6 nm in diameter from the small-angle X-ray scattering measurements (Figure S20), which is consistentwith previously reported results.38 When the diffusioncoefficient was estimated from the Levich plot obtained fromhydrodynamic voltammetry (Figure S21), the diffusioncoefficient tended to decrease as the number of moleculesincreased. In addition, there is a linear relationship between thenumber of units and the diffusion coefficient, indicating thatthe introduced molecules do not aggregate but rather enter theframework in an orderly manner (Figure S22). The hydro-dynamic radius calculated by Stokes−Einstein equation fromthe diffusion coefficient was 1.16 nm for G2, 1.42 nm for G3,and 2.72 nm for G4. Although the molecular size of G2 issmaller than 3.6 nm (1.8 nm radius), when considered as thesize of the hydration structure including counterions, it mightbe difficult for G2 with high symmetry to diffuse through the 4nm channel in Nafion. In contrast, MV2+ has a smallermolecular size than G2 and is likely to diffuse in the channel,so a stable redox reaction on the electrode surface is notexpected. In fact, reversible one-electron redox reactions ofMV2+ were observed by ordinary three-electrode cyclicvoltammetry in acidic solutions using a glassy carbon electrode(Figure S14). Therefore, it is thought that these differences inredox behavior are caused by diffusion of molecules throughthe Nafion membrane. For details, structural and electronicstate analysis, including the hydration structure, is necessary.■ CHARGE−DISCHARGE CHARACTERISTICS INNEUTRAL CONDITIONFigure 3 compares the charge and discharge characteristicswhen the catholyte was ferrocene trimethylammoniumchloride (Fe-TMACl) dissolved in NaCl aqueous solutionand the anolyte was G1 (Figure 3A,B) and G3 (Figure 3C,D)dissolved in 1 M NaCl aqueous solution. Selemion-DSV wasused for the anion exchange membrane. For G1, when thecatholyte concentration was 0.15 M and the anolyte was 0.1 M,the Coulomb efficiency and the voltage efficiency were 99%and 91%, respectively. The energy density was 0.69 Wh/L. Incontrast, for G3, when the catholyte and anolyte concen-trations were 0.1 and 0.04 M, the Coulomb efficiency andvoltage efficiency were 99% and 83%, respectively, and theenergy density was 0.91 Wh/L. Compared with G1, G3 has ahigher energy density despite the lower active materialFigure 3. Galvanostatic charge−discharge performance of Fe-TMACl−G1 flow battery (A, B) and Fe-TMACl−G3 flow battery (C, D) in neutralmedia. (A, C) Relationship between cell voltage and capacity in second cycle. (B, D) Change in cell voltage and positive and negative electrodepotential vs Ag/AgCl in three charge−discharge cycles. The current efficiency, voltage efficiency, and the energy density were estimated from thecharge−discharge curve of the second cycle. Applied current density: 10 mA/cm2. Concentration: (A) 0.15 M Fe-TMACl−0.1 M G1 in 1 M NaCl,(C) 0.1 M Fe-TMACl−0.04 M G3 in H2O. Volume of electrolyte solution: (A) positive 10 mL, negative 5 mL, (C) positive 7 mL, negative 7 mL.Flow rate (A, B): 20 mL/min. Membrane: Selemion DSV. Electrode: AAF304ZS (active area of 5 cm2).ACS Applied Energy Materials www.acsaem.org Articlehttps://dx.doi.org/10.1021/acsaem.0c00067ACS Appl. Energy Mater. XXXX, XXX, XXX−XXXDhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig3&ref=pdfwww.acsaem.org?ref=pdfhttps://dx.doi.org/10.1021/acsaem.0c00067?ref=pdfconcentration, again indicating that each intramolecular unitcontributes to the redox reaction.We have estimated the utilization of the theoretical capacityfor each generation: G1 is 84%; G3 is 43%. The origin of theselow utilization might be due to local inhomogeneity ofintramolecular or intermolecular reactions in flow batterysystem. Although the electrochemical measurement allows amultielectron-transfer reaction by one-electron reaction of eachviologen unit, these results clearly show that there is room forfurther improvement from the viewpoint of application as aflow battery.When the catholyte concentration was increased to 1 M andthe anolyte concentration with G3 was increased to 0.2 M, theCoulomb efficiency and voltage efficiency were 98% and 76%,respectively, and the energy density was 7.8 Wh/L (FigureS23).The anolyte was dissolved by about 1 M in the case of G3 inan uncharged state, that is, when the charge state in the unitwas a dication. This is a concentration corresponding to about5 M in MV2+. However, in the cation radical state generated bythe one-electron reduction reaction, the counterions aredecreased as compared with the dication, which likely lowersthe solubility. It is emphasized here that such a molecularassembly has a higher solubility than MV2+, and it can be acandidate structure that potentially contributes to an improve-ment in energy density, provided that a suitable catholyte isalso designed.The cycle test at high concentration, which is the one of themost important index of the performance for the flow battery,was also performed as shown in Figure 4 (catholyte, 1 M Fe-TMACl in 1 M NaCl; anolyte, 77 mM G4 in 1 M NaCl). Wehave found that the capacity drastically decreased after 10cycles. When the cell after the test was disassembled, thesurface of the membrane which was in contact with thepositive electrode was discolored and some precipitation wasalso observed. The results of FT-IR and UV−vis measurement(Figures S24 and S25) revealed that the precipitation wasderived from Fe-TMACl. In addition, Fe-TMACl adsorbedinside the membrane cannot be removed by washing. Thisphenomenon was also reproduced in the flow battery test atrelatively low concentrations (Figure S26), suggesting thatstrong interaction existed between the membrane interface andFe-TMACl. Although Liu et al. have achieved high-cyclingstability with the Fe-TMACl and MV2+flow battery system,30the membranes they used differed from our experimentalsystems, so further optimization is needed in our cases.For the anolyte, CV measurement of G4 before and aftercycling test was performed to confirm whether there was anychange. As a result, a decrease in the redox current and a peakshift have been observed (Figure S27). This suggests that someof the viologen units may have changed during the cycling.Since the CV profile is similar, it is considered that themolecular framework might be still maintained.Although in the electrochemical measurement without flow,those materials that can perform a multielectron-transferreaction due to a one-electron reduction reaction, in a batterytest, the reaction distribution is highly likely to occur, and atwo-electron reduction reaction may be induced during thecycling test. As a result, side reactions such as dimerizationbetween molecules might happen resulting in structuralchange. Therefore, the main possible causes of the capacitydecrease are assumed as follows:(1) Overvoltage rise due to adsorption of positive electrodeactive material on membrane surface and inside.(2) Partial insolubilization of viologen unit by progress ofpartial two-electron reduction.■ CHARGE−DISCHARGE CHARACTERISTICS INACIDIC CONDITIONBasically, viologen cannot function as a flow battery in acidicenvironment using Nafion. This is because the sulfonic acidgroup of Nafion strongly interacts with viologen, and in orderto avoid this strong interaction, a molecule containing thesame sulfonic acid group as the side chain functional group ofNafion must be introduced into the viologen skeleton.32 Wepreviously discussed how the viologen molecular assemblyshows a reversible redox reaction in acid using Nafion protonexchange membrane. This indicates that charging anddischarging are possible using an acidic electrolyte and aproton exchange membrane.Figure 5A,B shows the results of a charge−discharge testwith a Nafion electrolyte membrane, aqueous vanadium oxidesulfate as the catholyte, and G3 aqueous solution as theanolyte. For comparison, a charge−discharge test was alsoconducted using a sulfuric acid aqueous solution of BV2+ as theanolyte (Figure 5C,D). When a molecule has a viologen unit,the redox reaction does not depend on pH, unlike redox with aquinone molecule. Therefore, a molecule having a positiveredox potential can be selected for the catholyte, as comparedwith the neutral solution, to improve the energy density.In fact, the energy density of the acidic solution was higherthan that of the neutral solution. In G2, the Coulombicefficiency and voltage efficiency were 93% and 86%,respectively, and the energy density was 1.12 Wh/L. In G3,the Coulombic efficiency and voltage efficiency were 91% and90%, respectively, and the energy density is 1.07 Wh/L, whichis higher than that for the neutral solution. This is because theelectromotive force is increased by using vanadium as thecatholyte. For G4, the viologen unit equivalent concentrationwas about half of that of G2 and G3; the Coulomb efficiencyand voltage efficiency were 94% and 81%, respectively; and theenergy density was 0.42 Wh/L. For comparison, the energydensity of a 0.5 M BV2+−0.5 M VOSO4−1.5 M H2SO4 flowbattery was 0.22 Wh/L, which is lower than those of G2, G3,Figure 4. Charge−discharge profiles of selected cycles of Fe-TMACl−G4 battery at 13 mA/cm2. Concentration: 1 M Fe-TMACl in 1 M NaCl aq and 0.077 M G4 in 1 M NaCl aq. Due toextremely low capacity, the data after 30 cycles were omitted.ACS Applied Energy Materials www.acsaem.org Articlehttps://dx.doi.org/10.1021/acsaem.0c00067ACS Appl. Energy Mater. XXXX, XXX, XXX−XXXEhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig4&ref=pdfwww.acsaem.org?ref=pdfhttps://dx.doi.org/10.1021/acsaem.0c00067?ref=pdfand G4, despite the relatively high acid concentration. Inaddition, the Coulomb efficiency is as low as about 66%, andinteraction with the Nafion membrane is thought to reducereactivity with the electrode. It is emphasized here that ourmolecules can be operated in acidic conditions where MV2+does not work, and the origin is control of the interactionbetween the membrane surfaces and molecules by adjustingthe molecular size. Thus, the cation exchange membrane otherthan Nafion can be used to operate the RFB in the neutralregion. This has the advantage of increasing the choice ofactive materials on the positive electrode side.■ CONCLUSIONWe applied a newly designed viologen-assembled regularstructure as an anolyte to create a high-performance RFB.Here, by introducing a dendrimer-type molecular structure,we succeeded in developing a highly water-soluble activematerial and demonstrated that we can operate as a flowbattery in both neutral and acidic environments. However,significant issue with cycling stability has also been found, andfurther study will be needed to search for a more stablestructure that would not cause side reactions such asdimerization.Currently, we have developed only the anolyte, but ourconcept can also be applied to the catholyte. Because radical-based redox reactions with viologen derivatives do not dependon pH, these materials are promising for improving RFBenergy density.■ ASSOCIATED CONTENT*sı Supporting InformationThe Supporting Information is available free of charge athttps://pubs.acs.org/doi/10.1021/acsaem.0c00067.Synthesis details, NMR spectra, and additional exper-imental details (PDF)■ AUTHOR INFORMATIONCorresponding AuthorsAkihiro Ohira − Research Institute for Energy Conservation,National Institute of Advanced Industrial Science andTechnology (AIST), Tsukuba, Ibaraki 305-8565, Japan;orcid.org/0000-0002-5748-4830; Email: a-oohira@aist.go.jpTakashi Funaki − Research Center for Photovoltaics, AISTTsukuba Central 5, Tsukuba, Ibaraki 305-8565, Japan;Email: takasi-funaki@aist.go.jpAuthorsErika Ishida − Research Institute for Energy Conservation,National Institute of Advanced Industrial Science andTechnology (AIST), Tsukuba, Ibaraki 305-8565, JapanJe-Deok Kim − Hydrogen Production Materials Group, Centerfor Green Research on Energy and Environmental Materials,National Institute for Materials Science, Tsukuba, Ibaraki 305-0044, JapanYukari Sato − Research Institute for Energy Conservation,National Institute of Advanced Industrial Science andTechnology (AIST), Tsukuba, Ibaraki 305-8565, JapanComplete contact information is available at:https://pubs.acs.org/10.1021/acsaem.0c00067NotesThe authors declare no competing financial interest.■ ACKNOWLEDGMENTSA.O and T.F. thank Department of Energy and Environment ofAIST for funding support of seeds research. The small-angle X-ray scattering (SAXS) was measured by using the beam timeallotted from Proposal 2017G687 at Photon Factory of KEK.The authors thank Hideaki Takagi and Nobutaka Shimizu ofKEK-PF for their help with the SAXS measurements.Figure 5. Galvanostatic charge−discharge performance of vanadium−G3 flow battery (A, B) and vanadium−BV2+flow battery (C, D) in acidicmedia. (A, C) Relationship between cell voltage and capacity in second cycle. (B, D) Change in cell voltage and positive and negative electrodepotential vs Ag/AgCl in three charge−discharge cycles. The current efficiency, voltage efficiency, and the energy density were estimated from thecharge−discharge curve of the second cycle. Applied current density: 7 mA/cm2. Concentration: (A) 0.1 M VOSO4 in 0.3 M H2SO4−0.1 M G3 inH2O, (B) 0.5 M VOSO4 in 1.5 M H2SO4−0.5 M BV2+ in 1.5 M H2SO4. Volume of electrolyte solution: (A) positive 7 mL, negative 7 mL, (C)positive 6 mL, negative 6 mL. Flow rate: (A) 20 mL/min, (C) 3 mL/min. Membrane: Nafion NR212. Electrode: AAF304ZS (active area of 5 cm2).ACS Applied Energy Materials www.acsaem.org Articlehttps://dx.doi.org/10.1021/acsaem.0c00067ACS Appl. Energy Mater. XXXX, XXX, XXX−XXXFhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?goto=supporting-infohttp://pubs.acs.org/doi/suppl/10.1021/acsaem.0c00067/suppl_file/ae0c00067_si_001.pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Akihiro+Ohira"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttp://orcid.org/0000-0002-5748-4830http://orcid.org/0000-0002-5748-4830mailto:a-oohira@aist.go.jpmailto:a-oohira@aist.go.jphttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Takashi+Funaki"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfmailto:takasi-funaki@aist.go.jphttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Erika+Ishida"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Je-Deok+Kim"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Yukari+Sato"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig5&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig5&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig5&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaem.0c00067?fig=fig5&ref=pdfwww.acsaem.org?ref=pdfhttps://dx.doi.org/10.1021/acsaem.0c00067?ref=pdf■ REFERENCES(1) Zakeri, B.; Syri, S. Electrical energy storage systems: Acomparative life cycle cost analysis. Renewable Sustainable EnergyRev. 2015, 42 (2015), 569−596.(2) Shi, F. Reactor and Process Design in Sustainable EnergyTechnology; Elsevier: Amsterdam, 2016.(3) Dunn, B.; Kamath, H.; Tarascon, J.-M. Electrical Energy Storagefor the Grid: A Battery of Choices. Science 2011, 334, 928−935.(4) Noack, J.; Roznyatovskaya, N.; Herr, T.; Fischer, P. TheChemistry of Redox-Flow Batteries. Angew. Chem., Int. Ed. 2015, 54,9776−9809.(5) Soloveichik, G. L. Flow Batteries: Current Status and Trends.Chem. Rev. 2015, 115, 11533−11558.(6) Theller, L. H. Electrically Rechargeable Redox Flow Cells; NASATM X-71540; National Aeronautics and Space Administration:Washington, DC, 1974; pp 1−5.(7) Gahn, R. F.; Hagedorn, N. Negative-Electrode Catalysts for Fe/Cr Redox Cells. NASA Tech Briefs; NASA, February 1987; pp 42−43.(8) Skyllas-Kazacos, M. S.; Rychcik, M.; Robins, R. G.; Fane, A. G.;Green, M. A. New All-Vanadium Redox Flow Cell. J. Electrochem. Soc.1986, 133, 1057−1058.(9) Dennison, C. R.; Agar, E.; Akuzum, B.; Kumbur, E. C.Enhancing Mass Transport in Redox Flow Batteries by Tailoring FlowField and Electrode Design. J. Electrochem. Soc. 2016, 163, A5163−A5169.(10) Dong, Y. R.; Kaku, H.; Hanafusa, K.; Moriuchi, K.; Shigematsu,T. A Novel Titanium/Manganese Redox Flow Battery. ECS Trans.2015, 69, 59−67.(11) Xu, Y.; Wen, Y.; Cheng, J.; Cao, G.; Yang, Y. Study on a singleflow acid Cd−chloranil battery. Electrochem. Commun. 2009, 11,1422−1424.(12) Wei, X.; Xu, W.; Vijayakumar, M.; Cosimbescu, L.; Liu, T.;Sprenkle, V.; Wang, W. TEMPO-Based Catholyte for High-EnergyDensity Nonaqueous Redox Flow Batteries. Adv. Mater. 2014, 26,7649−7653.(13) Huskinson, B.; Marshak, M. P.; Suh, C.; Er, S.; Gerhardt, M. R.;Galvin, C. J.; Chen, X.; Aspuru-Guzik, A.; Gordon, R. G.; Aziz, M. J. Ametal-free organic−inorganic aqueous flow battery. Nature 2014, 505,195−198.(14) Yang, B.; Hoober-Burkhardt, L.; Wang, F.; Prakash, G. K. S.;Narayanan, S. R. An Inexpensive Aqueous Flow Battery for Large-Scale Electrical Energy Storage Based on Water-Soluble OrganicRedox Couples. J. Electrochem. Soc. 2014, 161, A1371−A1380.(15) Winsberg, J.; Stolze, C.; Muench, S.; Liedl, F.; Hager, M. D.;Schubert, U. S. TEMPO/Phenazine Combi-Molecule: A Redox-Active Material for Symmetric Aqueous Redox-Flow Batteries. ACSEnergy Lett. 2016, 1, 976−980.(16) Lin, K.; Goḿez-Bombarelli, R.; Beh, E. S.; Tong, L.; Chen, Q.;Valle, A.; Aspuru-Guzik, A.; Aziz, M. J.; Gordon, R. G. A redox-flowbattery with an alloxazine-based organic electrolyte. Nat. Energy 2016,1, 16102.(17) Hollas, A.; Wei, X.; Murugesan, V.; Nie, Z.; Li, B.; Reed, D.;Liu, J.; Sprenkle, V.; Wang, W. A biomimetic high-capacity phenazine-based anolyte for aqueous organic redox flow batteries. Nat. Energy2018, 3, 508−514.(18) Wang, C.; Yang, Z.; Wang, Y.; Zhao, P.; Yan, W.; Zhu, G.; Ma,L.; Yu, B.; Wang, L.; Li, G.; Liu, J.; Jin, Z. High-Performance AlkalineOrganic Redox Flow Batteries Based on 2-Hydroxy-3-carboxy-1,4-naphthoquinone. ACS Energy Lett. 2018, 3, 2404−2409.(19) Wang, C.; Li, X.; Yu, B.; Wang, Y.; Yang, Z.; Wang, H.; Lin, H.;Ma, J.; Li, G.; Jin, Z. Molecular Design of Fused-Ring PhenazineDerivatives for Long-Cycling Alkaline Redox Flow Batteries. ACSEnergy Lett. 2020, 5, 411−417.(20) Yan, W.; Wang, C.; Tian, J.; Zhu, G.; Ma, L.; Wang, Y.; Chen,R.; Hu, Y.; Wang, L.; Chen, T.; Ma, J.; Jin, Z. All-polymer particulateslurry batteries. Nat. Commun. 2019, 10, 2513.(21) Ebbesen, T. W.; Ferraudi, G. Photochemistry of MethylViologen in Aqueous and Methanolic Solutions. J. Phys. Chem. 1983,87, 3717−3721.(22) Nanasawa, M.; Matsukawa, Y.; Jin, J. J.; Haramoto, Y. Redoxphotochromism of viologen in organized solid state. J. Photochem.Photobiol., A 1997, 109, 35−38.(23) Bird, C. L.; Kuhn, A. T. Electrochemistry of the viologens.Chem. Soc. Rev. 1981, 10, 49−82.(24) Striepe, L.; Baumgartner, T. Viologens and Their Application asFunctional Materials. Chem. - Eur. J. 2017, 23, 16924−16940.(25) Ding, J.; Zheng, C.; Wang, L.; Lu, C.; Zhang, B.; Chen, Y.; Li,M.; Zhai, G.; Zhuang, X. Viologen-inspired functional materials:synthetic strategies and applications. J. Mater. Chem. A 2019, 7,23337−23360.(26) Janoschka, T.; Martin, N.; Martin, U.; Friebe, C.; Morgenstern,S.; Hiller, H.; Hager, M. D.; Schubert, U. S. An aqueous, polymer-based redox-flow battery using non-corrosive, safe, and low-costmaterials. Nature 2015, 527, 78−81.(27) Beh, E. S.; De Porcellinis, D.; Gracia, R. L.; Xia, K. T.; Gordon,R. G.; Aziz, M. J. A Neutral pH Aqueous Organic−OrganometallicRedox Flow Battery with Extremely High Capacity Retention. ACSEnergy Lett. 2017, 2, 639−644.(28) Hu, B.; Tang, Y.; Luo, J.; Grove, G.; Guo, Y.; Liu, T. L.Improved radical stability of viologen anolytes in aqueous organicredox flow batteries. Chem. Commun. 2018, 54, 6871−6874.(29) Hagemann, T.; Winsberg, J.; Grube, M.; Nischang, I.;Janoschka, T.; Martin, N.; Hager, M. D.; Schubert, U. S. An aqueousall-organic redox-flow battery employing a (2,2,6,6-tetramethylpiper-idin-1-yl)oxyl-containing polymer as catholyte and dimethyl viologendichloride as anolyte. J. Power Sources 2018, 378, 546−554.(30) Hu, B.; DeBruler, C.; Rhodes, Z.; Liu, T. L. Long-CyclingAqueous Organic Redox Flow Battery (AORFB) toward Sustainableand Safe Energy Storage. J. Am. Chem. Soc. 2017, 139, 1207−1214.(31) Luo, J.; Hu, B.; Debruler, C.; Liu, T. L. A π-ConjugationExtended Viologen as a Two-Electron Storage Anolyte for TotalOrganic Aqueous Redox Flow Batteries. Angew. Chem., Int. Ed. 2018,57, 231−235.(32) DeBruler, C.; Hu, B.; Moss, J.; Luo, J.; Liu, T. L. A Sulfonate-Functionalized Viologen Enabling Neutral Cation Exchange, AqueousOrganic Redox Flow Batteries toward Renewable Energy Storage.ACS Energy Lett. 2018, 3, 663−668.(33) Burgess, M.; Cheńard, E.; Hernańdez-Burgos, K.; Nagarjuna,G.; Assary, R. S.; Hui, J.; Moore, J. S.; Rodríguez-Loṕez, J. Impact ofBackbone Tether Length and Structure on the ElectrochemicalPerformance of Viologen Redox Active Polymers. Chem. Mater. 2016,28, 7362−7374.(34) Dong, R.; Zhou, Y.; Zhu, X. Supramolecular DendriticPolymers: From Synthesis to Applications. Acc. Chem. Res. 2014,47, 2006−2016.(35) Watanabe, T.; Honda, K. Measurement of the ExtinctionCoefficient of the Methyl Viologen Cation Radical and the Efficiencyof Its Formation by Semiconductor Photocatalysis. J. Phys. Chem.1982, 86, 2617−2619.(36) Heinen, S.; Walder, L. Generation-Dependent IntramolecularCT Complexation in a Dendrimer Electron Sponge Consisting of aViologen Skeleton. Angew. Chem., Int. Ed. 2000, 39, 806−809.(37) Heinen, S.; Meyer, W.; Walder, L. Charge trapping indendrimers with a viologen skeleton and a radial redox gradient. J.Electroanal. Chem. 2001, 498, 34−43.(38) Mauritz, K. A.; Moore, R. B. State of Understanding of Nafion.Chem. Rev. 2004, 104, 4535−4585.ACS Applied Energy Materials www.acsaem.org Articlehttps://dx.doi.org/10.1021/acsaem.0c00067ACS Appl. Energy Mater. XXXX, XXX, XXX−XXXGhttps://dx.doi.org/10.1016/j.rser.2014.10.011https://dx.doi.org/10.1016/j.rser.2014.10.011https://dx.doi.org/10.1126/science.1212741https://dx.doi.org/10.1126/science.1212741https://dx.doi.org/10.1002/anie.201410823https://dx.doi.org/10.1002/anie.201410823https://dx.doi.org/10.1021/cr500720thttps://dx.doi.org/10.1149/1.2108706https://dx.doi.org/10.1149/2.0231601jeshttps://dx.doi.org/10.1149/2.0231601jeshttps://dx.doi.org/10.1149/06918.0059ecsthttps://dx.doi.org/10.1016/j.elecom.2009.05.021https://dx.doi.org/10.1016/j.elecom.2009.05.021https://dx.doi.org/10.1002/adma.201403746https://dx.doi.org/10.1002/adma.201403746https://dx.doi.org/10.1038/nature12909https://dx.doi.org/10.1038/nature12909https://dx.doi.org/10.1149/2.1001409jeshttps://dx.doi.org/10.1149/2.1001409jeshttps://dx.doi.org/10.1149/2.1001409jeshttps://dx.doi.org/10.1021/acsenergylett.6b00413https://dx.doi.org/10.1021/acsenergylett.6b00413https://dx.doi.org/10.1038/nenergy.2016.102https://dx.doi.org/10.1038/nenergy.2016.102https://dx.doi.org/10.1038/s41560-018-0167-3https://dx.doi.org/10.1038/s41560-018-0167-3https://dx.doi.org/10.1021/acsenergylett.8b01296https://dx.doi.org/10.1021/acsenergylett.8b01296https://dx.doi.org/10.1021/acsenergylett.8b01296https://dx.doi.org/10.1021/acsenergylett.9b02676https://dx.doi.org/10.1021/acsenergylett.9b02676https://dx.doi.org/10.1038/s41467-019-10607-0https://dx.doi.org/10.1038/s41467-019-10607-0https://dx.doi.org/10.1021/j100242a028https://dx.doi.org/10.1021/j100242a028https://dx.doi.org/10.1016/S1010-6030(97)00105-6https://dx.doi.org/10.1016/S1010-6030(97)00105-6https://dx.doi.org/10.1039/cs9811000049https://dx.doi.org/10.1002/chem.201703348https://dx.doi.org/10.1002/chem.201703348https://dx.doi.org/10.1039/C9TA01724Khttps://dx.doi.org/10.1039/C9TA01724Khttps://dx.doi.org/10.1038/nature15746https://dx.doi.org/10.1038/nature15746https://dx.doi.org/10.1038/nature15746https://dx.doi.org/10.1021/acsenergylett.7b00019https://dx.doi.org/10.1021/acsenergylett.7b00019https://dx.doi.org/10.1039/C8CC02336Khttps://dx.doi.org/10.1039/C8CC02336Khttps://dx.doi.org/10.1016/j.jpowsour.2017.09.007https://dx.doi.org/10.1016/j.jpowsour.2017.09.007https://dx.doi.org/10.1016/j.jpowsour.2017.09.007https://dx.doi.org/10.1016/j.jpowsour.2017.09.007https://dx.doi.org/10.1021/jacs.6b10984https://dx.doi.org/10.1021/jacs.6b10984https://dx.doi.org/10.1021/jacs.6b10984https://dx.doi.org/10.1002/anie.201710517https://dx.doi.org/10.1002/anie.201710517https://dx.doi.org/10.1002/anie.201710517https://dx.doi.org/10.1021/acsenergylett.7b01302https://dx.doi.org/10.1021/acsenergylett.7b01302https://dx.doi.org/10.1021/acsenergylett.7b01302https://dx.doi.org/10.1021/acs.chemmater.6b02825https://dx.doi.org/10.1021/acs.chemmater.6b02825https://dx.doi.org/10.1021/acs.chemmater.6b02825https://dx.doi.org/10.1021/ar500057ehttps://dx.doi.org/10.1021/ar500057ehttps://dx.doi.org/10.1021/j100211a014https://dx.doi.org/10.1021/j100211a014https://dx.doi.org/10.1021/j100211a014https://dx.doi.org/10.1002/(SICI)1521-3773(20000218)39:4<806::AID-ANIE806>3.0.CO;2-Ihttps://dx.doi.org/10.1002/(SICI)1521-3773(20000218)39:4<806::AID-ANIE806>3.0.CO;2-Ihttps://dx.doi.org/10.1002/(SICI)1521-3773(20000218)39:4<806::AID-ANIE806>3.0.CO;2-Ihttps://dx.doi.org/10.1016/S0022-0728(00)00215-1https://dx.doi.org/10.1016/S0022-0728(00)00215-1https://dx.doi.org/10.1021/cr0207123www.acsaem.org?ref=pdfhttps://dx.doi.org/10.1021/acsaem.0c00067?ref=pdf