Relationship between activity and structure of carbon materials for Br₂/Br⁻ in zinc bromine flow batteries

Abstract

Zinc bromine flow battery (ZBFB) is one of the highly efficient and low cost energy storage devices. However, the low operating current density hinders its progress. Developing high activity cathode materials is an efficient way to reduce cell electrochemical polarization and improve the operating current density. Thus, it is essential to study the relationship between the activity and structure of carbon materials to optimize the performance of ZBFB. The pore parameters and phase structure of four commercialized carbon materials were investigated by an N₂ sorption isotherm experiment and X-ray diffraction (XRD), respectively. The electrochemical property of the four carbon materials was systematically studied by cyclic voltammetry (CV), linear sweep voltammetry (LSV) and electrochemical impedance spectroscopy (EIS) and the kinetic parameters and diffusion coefficients were calculated. The results indicate that specific surface area, pore size distribution and electrical conductivity are the main factors affecting the electrochemical activity of carbon materials. The carbon material with high surface area, suitable pore size distribution and excellent electrical conductivity shows high activity to the Br₂/Br⁻ redox couple in ZBFB. This study lays foundations for developing cathode materials of excellent activity for ZBFB, which can efficiently improve the power density, reduce the stack size of the ZBFB and boost its potential for commercial application.

---

Introduction

Energy consumption has increased with the advance of science and technology resulting in the excessive consumption of fossil fuels such as oil and coal. The consequences are a series of environmental issues, for example global warming and pollution haze.¹ Thus, increasing the percentage of renewable energy, such as solar, wind and tidal, is an effective way to alleviate the environmental issues.² However, the non-sequential and unstable nature of renewable energy indicates that it cannot be employed as a total replacement. Large scale energy storage devices therefore play an important role in smoothing energy output and work as a buffer to store surplus energy in abundant time or supply energy in shortage time.³ Flow battery is a promising battery system in large-scale energy storage applications. It has the advantages of independent design of power and energy, high safety, long cycle life and low maintenance costs.^4,5 Among the various flow batteries, the zinc bromine flow battery (ZBFB) is one of the efficient and promising energy storage devices. Br₂/Br⁻ and Zn²⁺/Zn are the positive and negative redox couples of ZBFB, respectively. Fig. 1 shows the ZBFB schematic and the cell reactions are described as follows:

Fig. 1 Schematic of zinc bromine flow batteries.

The abundant and low-cost positive and negative redox couple materials contribute to the economic and competitive nature of the ZBFB system. Another outstanding advantage of the ZBFB is the high theoretical energy density (440 W h kg⁻¹), which results from the high cell potential window (1.83 vs. SHE) and the high solubility of active materials.⁶ Therefore, the ZBFB is one of the most promising battery systems for commercial realization. However, the low operating current density, mainly resulting from the mismatching of positive and negative reaction rates (the reaction rate of Br₂/Br⁻ is slower than that of Zn²⁺/Zn) largely hampers the widespread application of the ZBFB.⁷ Hence, it is both important and urgent to develop high activity cathode materials to accelerate the Br₂/Br⁻ reaction rate, reduce cell electrochemical polarization and increase the operating current density, which is an efficient way to increase power density and further reduce the battery stack size and cost.

Carbon materials possess the advantages of low cost, excellent electrical conductivity, good resistance to acid and oxidation, simple synthetic methods and controllable fabrication of the structure and surface properties. They show excellent potential to act as the cathode material in ZBFB. The Br₂/Br⁻ kinetics on vitreous carbon, glass carbon and graphite electrodes and the reaction mechanism have been inspected.^8,9 This research provides some essential information on the Br₂/Br⁻ reaction for subsequent investigations. The activity of multi-walled carbon nanotubes (MWCNTs), singe-wall carbon nanotubes (SWCNTs) and oxygen-containing functionalized singe-wall carbon nanotubes (FSWCNTs) toward Br₂/Br⁻ reaction was investigated.^10–12 It was concluded that the oxygen-containing functional groups and many more active sites are conductive to good activity towards Br₂/Br⁻. Our previous study demonstrates that active carbon shows good activity towards Br₂/Br⁻ and can improve the energy efficiency (EE) from 68% to 75% at 40 mA cm⁻².¹³ Various carbon materials have already been investigated as cathode electrode materials. However, the performances of these carbon materials are quite different from each other and the correlations between their properties and activities for Br₂/Br⁻ are still unclear. Thus, a systematic study on the factors affecting activities of carbons is needed to form a theory that can guide researchers to rationally design and synthesize specific materials with high activity cathode materials for Br₂/Br⁻. This is extremely significant for reducing battery polarization, improving battery efficiency and current density and further reducing stack size and cost and effectively accelerating commercial progress.

Herein, four commercialized carbon materials: acetylene black (AB), expanded graphite (EG), carbon nanotube (CNT) and BP2000 (BP), with unique properties were selected for the investigations on the relationships between the activity for the Br₂/Br⁻ electrochemical reaction and the structures of the carbon material. The structures of the four carbons were observed and the electrochemical activity of the four carbons for Br₂/Br⁻ was compared. The characteristics of the inner connections between the carbons properties and activity to Br₂/Br⁻ were revealed through analyzing the results. This lays the foundations to controllably fabricate a high activity cathode material for the ZBFB system.

Experimental section

Materials characterization

X-ray diffraction (XRD) patterns were recorded on a DX-2700 X diffractometer (Angstrom Advanced Inc., USA) in the scan range of 10–80° using Cu-K_α1 radiation (λ = 1.54056 Å). An N₂ sorption isotherm experiment was conducted in a gas adsorption analyzer to investigate the pore structures of the carbon materials (ASAP-2010 Micromeritics Instrument) after degassing at 473 K and 7 μTorr for 6 h. The specific surface areas and pore size distributions were calculated from the adsorption branch by Brunauer–Emmett–Teller (BET) and Barrerr–Joyner–Halenda (BJH) methods, respectively.

Electrochemical measurement

The electrochemical properties of the materials were investigated in a three-electrode system, including a working electrode, a counter electrode and a reference electrode, using a Gamry multichannel system installation (Reference 3000). Carbon-decorated glass carbon electrode (GC, d = 5 mm), graphite rod, and Ag/AgCl electrode (0.198 V vs. SHE) acted as the working, counter and reference electrodes, respectively. The working electrode was prepared by dripping 10 μL of carbon material ink onto the surface of the GC electrode that had been polished and washed with deionized water in advance. The carbon material ink was made by dispersing 2.5 mg carbon material and 10 μL of 0.05 wt% Nafion (to act as a binder) in 0.5 mL isopropanol under strong ultrasound treatment. Electrochemical impedance spectra (EIS) were obtained in the frequency range of 0.01 to 10⁵ Hz.

ZBFB single cell test

The single cell was assembled by sandwiching two membranes (Daramic® HP) and a CCM (Catalyst Coating Membrane) with two carbon felts (electrode) and clamping by two polar plates. All these components were fixed between two stainless steel plates. The blank cell was assembled as described above without the CCM. The CCM was prepared by spraying carbon ink on the surface of a 200 μm thick membrane (Daramic® HP) with a loading of ∼3 mg cm⁻². The well-distributed carbon ink was made by dispersing carbon materials and 0.05 wt% Nafion into isopropanol via strong sonication. After drying at 60 °C for several hours, the CCM was assembled into a single cell with an effective electrode area of 9 cm² for the charge–discharge test. A 2 M ZnBr₂ solution served as electrolyte, whereas 3 M KCl and 0.05 M N-methylethylpyrrolidinium bromide (MEPB) were used as the supporting electrolyte and bromine complex. The cell was charged and discharged in a constant current density of 20 mA cm⁻². After charging for 50 min, the cell begins to discharge until the cell voltage below 0.5 V.

Results and discussion

Electrochemical performance

The electrochemical activity of the four carbons was evaluated by cyclic voltammetry (CV). All curves show the similar process: the potential gradually increases to a certain potential called the onset potential where the oxidation reaction of Br⁻ begins. The anodic current increases without a peak with the positive shift of the potential, whereas the cathodic peak appears in the reverse process (Fig. 2). This phenomenon indicates that mass transfer is quick enough in the oxidation process for there to be no anodic peak, whereas the mass transfer is insufficient at the negative potential in the reduction process. Comparing the oxidation current at 1 V and the cathodic peak current (I_p) of the four carbons, it is obvious that the activity increases in the order AB, CNT, EG and BP. The activity of BP is at least double the other carbons. The high activity is attributed to the large specific surface area supplying lots of active sites to the Br₂/Br⁻ reaction and suitable pore size distribution facilitating mass transfer, which is confirmed by the results of the N₂ sorption experiment, linear sweep voltammetry (LSV) and electrochemical impedance spectra (EIS).

Fig. 2 CV curves of AB, CNT, EG and BP at the scan rate of 10 mV s⁻¹ in electrolyte of 0.005 M ZnBr₂ + 0.005 M Br₂.

The pore structures of the four carbons were investigated by an N₂ sorption experiment. The N₂ sorption isotherms of CNT, EG and BP show the similar pseudo-type I curve with H1 hysteresis loops at high relative pressure (P/P₀ = 0.9–1.0) and an indistinct capillary condensation step at low relative pressure (P/P₀ = 0.20–0.40), which are ascribed to small and large mesopores, respectively, in the interparticle texture (Fig. 3). The pore parameters of the four carbons are listed in Table 1. The specific surface areas of AB, CNT, EG and BP are 60.3, 256.4, 785.4, and 1354.7 m² g⁻¹, respectively. In order to verify the presumption that specific surface areas significantly impact the activity for Br₂/Br⁻, the I_p–S_BET curve is pictured (Fig. 4). The activity of four carbons is nearly proportional to its S_BET, except CNT which possesses a higher activity than expected (red dot in Fig. 4). The carbons' activity increases with increase in its S_BET, which indicates that the larger specific surface area offers more active sites to the electrochemical reaction of Br₂/Br⁻. The unexpected high activity of CNT may owe to its excellent electrical conductivity, which is verified by the XRD results.

Fig. 3 (a) N₂ sorption isotherms and (b) pores size distribution curves of AB, CNT, EG and BP.

Table 1 Pore parameters of carbon materials^a

Samples	AB	CNT	EG	BP
a SBET specific surface area; SExternal external surface area; SMicro micropore area; VPore pore volume.
SBET/m^2 g	60.3	256.4	785.4	1354.7
SMicro/m^2 g	6.9	15.1	122.8	429.7
SExternal/m^2 g	53.4	241.3	662.6	925.0
VPore/cm^3 g	0.23	1.67	1.11	3.3

---

Fig. 4 Relationship between I_p and S_BET.

The XRD results reveal the phase structures of the four carbons. The peak shapes and peak positions of AB, CNT, EG and BP are similar: two shoulder peaks around 25° and 43°, which correspond to the C (002) and C (101) crystal planes, respectively¹⁴ (Fig. 5). The broad peaks in the patterns show that AB, CNT, EG and BP are amorphous. The peaks of CNT are much sharper than the other carbons, which indicates a higher degree of crystallization in CNT than in the other carbons. The higher crystallization degree of CNT leads to a better electrical conductivity, which contributes to the reduction of charge transfer resistance and the unexpected high activity.¹⁵

Fig. 5 XRD patterns of AB, CNT, EG and BP.

CV curves at different scan rates were recorded to investigate the feature of the Br₂/Br⁻ electrochemical reaction on the four carbons (Fig. 6). It can be seen that the CV curve changes in the four carbons is different with the same increase of scan rates. With the increase of scan rate, the electrical double layer capacitances of the four carbons increase to different degrees. Among the carbons, the electrical double layer capacitances of EG and BP increase more obviously than the others. The electrical double layer capacitance is in proportion to the specific surface areas,¹⁶ thus the small specific surface areas of AB and CNT lead to a smaller rise of electrical double layer capacitance than in EG and BP with the same change of scan rates. In addition, the negative shift of the cathodic peak potential resulting from the polarization enlargement with the increase of scan rate indicates that the electrochemical redox reaction of Br₂/Br⁻ is a quasi-reversible reaction.¹⁷ The anodic current and cathodic peak current of AB are almost unchanged, which is caused by the small specific surface area offering fewer active sites to the Br₂/Br⁻ reaction.

Fig. 6 CV curves of the selected carbons: (a) AB, (b) CNT, (c) EG and (d) BP at different scan rates in an electrolyte of 0.005 M ZnBr₂ + 0.005 M Br₂.

The relationship between cathodic peak current (I_p) and square root of scan rate (ν^1/2) of the four carbons was displayed in Fig. 7. The I_p of BP is proportional to ν^1/2, which indicates that the Br₂/Br⁻ reaction on BP is mainly controlled by mass transfer.¹⁸ The I_p of other carbons is not proportional to ν^1/2 or ν, indicating the mixed control of charge transfer process and mass diffusion, which is different from the case of BP. Thus, the diffusion coefficients of Br₂ were calculated according to the eqn (4) (ref. 19) and listed in Table 2.

where I_p is peak current, n is the number of transferred electrons, A is the electrode surface, D₀ is the diffusion coefficient, C₀ is the reactant concentration and ν is the scan rate.

Fig. 7 Relationship between cathodic peak current (I_p) and square root of scan rate (ν^1/2) for the four carbons.

Table 2 Kinetics parameters of Br₂/Br⁻ on various electrodes

Samples	AB	CNT	EG	BP
Rct/Ω cm^2	593.1	123.2	96.4	58.9
i0/10^−5 A cm^−2	2.13	10.3	13.1	21.4
k0/10^−5 cm s^−1	2.2	10.6	13.6	22.2
D0/10^−7 cm^2 s^−1	—	6.6	1.7	25.2

---

The diffusion coefficients of Br₂ on BP and CNT are larger than on AB and EG, particularly BP whose diffusion coefficient is twenty times that of EG. The excellent mass transfer capacity is due to the large pore size (Fig. 3(b)). CNT, EG and BP possess two types pore size: one small mesopore of around two nanometers and one large mesopore of tens to hundreds of nanometers. The abundant large mesopore on BP contributes to the mass diffusion. In addition, the larger size of the small mesopore and the much larger large mesopore on CNT lead to better mass transfer capacity than for EG. It is concluded that the property of pore size distribution is significantly important to mass diffusion.

In order to investigate the electrochemical essentials of the Br₂/Br⁻ reaction, the kinetic parameters were obtained by LSV (Fig. 8). According to steady state polarization theory, current density and overvoltage tend to a linear relationship at low overvoltage (<25/n mV).²⁰ Thus, the kinetic parameters were calculated according to eqn (5–7) (ref. 21) and are listed in Table 2.

where R_ct is the charge transfer resistance, η is the overvoltage, j is the current density, i₀ is the exchange current density, R is the universal gas constant, T is the temperature in Kelvin, n is the number of electrons transferred, F is the Faraday constant, k₀ is the reaction rate constant and C₀ is the reactant concentration.

Fig. 8 LSV curves of four carbons at the scan rate of 1 mV s⁻¹ in the electrolyte 0.005 M ZnBr₂ + 0.005 M Br₂.

The charge transfer resistance of the four carbons decreases in the order AB, CNT, EG and BP, which is the same order as their activity. The values of i₀ and k₀ of the four carbons decrease in exactly the reverse sequence. The results indicate that the kinetics on BP are faster than on the other carbons, which corresponds to the results of the CV and confirm that the larger specific surface areas supply more active sites.

The electrochemical activity of the four carbons was further studied by electrochemical impedance spectroscopy (EIS). All Nyquist plots consist of a semicircle at high frequencies and a linear part at low frequencies (Fig. 9). The diameter of the semicircle represents charge transfer resistance (R_ct), whereas the linear part represents the diffusion process.^20,22 It is obvious that the semicircle diameter of the four carbons increase in the order AB, CNT, EG and BP, which indicates that charge transfer resistance decreases in the same order. The results demonstrate the excellent performance of BP, which is due to the large specific surface area providing many more active sites for the Br₂/Br⁻ reaction and abundant large mesopores favoring mass transfer. The results of CV, LSV and EIS are highly consistent. The high frequency intersection of the semicircle with the real axis represents the ohmic resistance. The smallest value of intersection is achieved by CNT suggesting its good electrical conductivity. This verifies that CNT has the higher activity, which is considered to be consistent with the XRD results.

Fig. 9 Nyquist plots of the four carbons (inset: enlargement of the small red square).

ZBFB cell test performance

In order to verify the electrochemical activity of BP possessing the best activity for Br₂/Br⁻ among the four carbons, the four carbons were used as cathode materials to test their potential application in a ZBFB. Fig. 10(a) shows the charge–discharge curves of a blank cell and a cell with a BP CCM. The charge voltage plateau is reduced by about 100 mV and the discharge plateau is increased by about 120 mV using the BP CCM. This results in an increase of voltage efficiency (Fig. 10(a) and (b)). It may be ascribed to the high specific surface areas of BP providing more active sites for the Br₂/Br⁻ reaction and accelerating its kinetics. In addition, the efficiency of the blank cell and cell with different CCM was displayed in Fig. 10(b). The coulombic efficiency (CE) is almost the same, whereas the voltage efficiency (VE) increases by using the CCM (Fig. 10(b)). Among the four carbons, the cell with the BP CCM achieves the highest EE (84.4%). The cell efficiency difference of the four carbons is not as big as the difference in CV, LSV and EIS, which may be due to different responses to the cell environment. However, the trend of the four cells' performance is consistent with results of CV, LSV and EIS, which demonstrates the validity of this inference. The performance of the cell with CCM may further improve by optimizing the parameters of the CCM such as carbon load, thickness and preparation method. The cycling performance of a ZBFB was investigated and the results were shown in Fig. 10(c) and (d). After charge–discharge for 200 cycles, the efficiency of the blank cell and a cell with BP CCM keeps the same value as the original without obvious decay, which indicates good stability of the ZBFB system. The results suggest that BP, with high electrochemical activity towards Br₂/Br⁻, shows good potential in the ZBFB application and embodies the essence of this study to investigate the relationship between the structure and activity of the carbon materials.

Fig. 10 (a) Charge–discharge curves of blank cell and the cell with a BP CCM; (b) cell efficiency of a blank cell and a cell with different CCM operating at 20 mA cm⁻²; (c) cycle performance of the blank cell; (d) cycle performance of the cell with a BP CCM.

Conclusions

The relationship between activities and structures of carbon materials for the Br₂/Br⁻ reaction couple was investigated by studying the performance of four commercialized carbons with different structures to help guide the design of an excellent cathode material for the ZBFB. The highest specific surface areas, providing more active sites for the electrochemical reaction of Br₂/Br⁻, contribute to the highest activity of BP among the selected carbon materials. In addition, the appropriate pore size distribution with abundant large mesopores and numerous small mesopores efficiently facilitates mass transfer. Moreover, the excellent electrical conductivity deriving from a higher degree of graphitization largely reduces the ohmic resistance and contributes to a better performance. However, too high graphitization degree may reduce the specific active sites and reduce the reaction kinetic. Thus, it is concluded that the carbon material with high surface area, appropriate pore size distribution and degree of graphitization is a promising cathode material for the ZBFB. This work lays a foundation for the development of highly efficient and low cost ZBFBs and boosts the commercialization of ZBFBs.

Acknowledgements

The authors greatly acknowledge the financial support from the China Natural Science Foundation (Grant No. 21476224, 21406219), the Outstanding Young Scientist Foundation, the Chinese Academy of Sciences (CAS), supported by the Key Research Program of the Chinese Academy of Sciences (KG2D-EW-602-2), and the Dalian Municipal Outstanding Young Talent Foundation (2014J11JH131).

References

1. Z. Yang, J. Zhang, M. C. W. Kintner-Meyer, X. Lu, D. Choi, J. P. Lemmon and J. Liu, Chem. Rev., 2011, 111, 3577–3613.
2. B. Dunn, H. Kamath and J.-M. Tarascon, Science, 2011, 334, 928–935.
3. A. Poullikkas, Renewable Sustainable Energy Rev., 2013, 27, 778–788.
4. J. Noack, N. Roznyatovskaya, T. Herr and P. Fischer, Angew. Chem., Int. Ed., 2015, 54, 9776–9809.
5. W. Wang, Q. Luo, B. Li, X. Wei, L. Li and Z. Yang, Adv. Funct. Mater., 2013, 23, 970–986.
6. F. Beck and P. Rüetschi, Electrochim. Acta, 2000, 45, 2467–2482.
7. P. Leung, X. Li, C. Ponce de Leon, L. Berlouis, C. T. J. Low and F. C. Walsh, RSC Adv., 2012, 2, 10125–10156.
8. M. Mastragostino and C. Gramellini, Electrochim. Acta, 1985, 30, 373–380.
9. K. T. Cho, P. Ridgway, A. Z. Weber, S. Haussener, V. Battaglia and V. Srinivasan, J. Electrochem. Soc., 2012, 159, A1806–A1815.
10. Y. Munaiah, S. Dheenadayalan, P. Ragupathy and V. K. Pillai, ECS J. Solid State Sci. Technol., 2013, 2, M3182–M3186.
11. Y. Munaiah, S. Suresh, S. Dheenadayalan, V. K. Pillai and P. Ragupathy, J. Phys. Chem. C, 2014, 118, 14795–14804.
12. X. Rui, A. Parasuraman, W. Liu, D. H. Sim, H. H. Hng, Q. Yan, T. M. Lim and M. Skyllas-Kazacos, Carbon, 2013, 64, 464–471.
13. L. Zhang, H. Zhang, Q. Lai, X. Li and Y. Cheng, J. Power Sources, 2013, 227, 41–47.
14. S. Liu, J. Zhu, Y. Mastai, I. Felner and A. Gedanken, Chem. Mater., 2000, 12, 2205–2211.
15. Z. Liu, G. Zhang, Z. Lu, X. Jin, Z. Chang and X. Sun, Nano Res., 2013, 6, 293–301.
16. E. Frackowiak, Phys. Chem. Chem. Phys., 2007, 9, 1774–1785.
17. M. V. Mirkin and A. J. Bard, Anal. Chem., 1992, 64, 2293–2302.
18. V. A. Ettel, B. V. Tilak, A. S. Gendron, V. A. Ettel, B. V. Tilak and A. S. Gendron, J. Electrochem. Soc., 1974, 121, 867–872.
19. Q. Liu, A. A. Shinkle, Y. Li, C. W. Monroe, L. T. Thompson and A. E. S. Sleightholme, Electrochem. Commun., 2010, 12, 1634–1637.
20. F. Rahman and M. Skyllas-Kazacos, J. Power Sources, 2009, 189, 1212–1219.
21. X. Wu, S. Liu, N. Wang, S. Peng and Z. He, Electrochim. Acta, 2012, 78, 475–482.
22. Y. Zhao, Y. Ding, J. Song, L. Peng, J. B. Goodenough and G. Yu, Energy Environ. Sci., 2014, 7, 1990–1995.

---