Poly(phenyl sulfone) anion exchange membranes with pyridinium groups for vanadium redox flow battery applications

Journal of Power Sources 282 (2015) 328e334

Contents lists available at ScienceDirect

Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour

Poly(phenyl sulfone) anion exchange membranes with pyridinium groups for vanadium redox flow battery applications Bengui Zhang*, Enlei Zhang, Guosheng Wang, Ping Yu, Qiuxia Zhao, Fangbo Yao College of Chemical Engineering, Shenyang University of Chemical Technology, Shenyang 110124, PR China

h i g h l i g h t s

g r a p h i c a l a b s t r a c t


PyPPSU membranes are first fabricated and investigated for vanadium redox flow battery application.
PyPPSU membranes show significantly low vanadium ions permeability.
High VRFB cell efficiencies are obtained by using PyPPSU membranes.
PyPPSU membrane shows stable performance in VRFB cycling test.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 December 2014 Received in revised form 10 February 2015 Accepted 13 February 2015 Available online 14 February 2015

To develop high performance and cost-effective membranes with low permeability of vanadium ions for vanadium redox flow battery (VRFB) application, poly(phenyl sulfone) anion exchange membranes with pyridinium groups (PyPPSU) are prepared and first investigated for VRFB application. PyPPSU membranes show much lower vanadium ions permeability (0.07  107e0.15  107 cm2 min1) than that of Nafion 117 membrane (31.3  107 cm2 min1). As a result, the self-discharge duration of the VRFB cell with PyPPSU membrane (418 h) is about four times longer than that of VRFB cell with Nafion 117 membrane (110 h). Furthermore, the VRFB cell with PyPPSU membrane exhibits higher battery efficiency (coulombic efficiency of 97.8% and energy efficiency of 80.2%) compare with that of VRFB cell with Nafion 117 membrane (coulombic efficiency of 96.1% and energy efficiency of 77.2%) at a high current density of 100 mA cm2. In addition, PyPPSU membrane exhibits stable performance in 100-cycle test. The results indicate that PyPPSU membrane is high performance and low-cost alternative membrane for VRFB application. © 2015 Elsevier B.V. All rights reserved.

Keywords: Vanadium redox flow battery Anion exchange membrane Poly(phenyl sulfone) Pyridinium

1. Introduction Vanadium redox flow battery (VRFB) reported by M. SkyllasKazacos and coworkers [1] is a promising technology for energy storage system which provide solutions to overcome the intermittent nature of renewable energy and provide stable, reliable

* Corresponding author. E-mail address: [email protected] (B. Zhang). http://dx.doi.org/10.1016/j.jpowsour.2015.02.070 0378-7753/© 2015 Elsevier B.V. All rights reserved.

electricity for power grids [2]. In VRFB, membrane serves as physical separator to prevent cross-mixing of the positive and negative electrolytes while allowing the transport of ions (such as Hþ, SO2 4 , and HSO-4) to complete the electrical circuit. Nafion membrane served as the benchmark membrane for VRFB because of its high conductivity and good chemical stability. However, Nafion membrane has suffered from high cost and severe permeability of vanadium ions. Various modification methods (such as nanopartical incorporation [3,4], composite construction [5], multilayer

B. Zhang et al. / Journal of Power Sources 282 (2015) 328e334

structuring [6,7] and surface modification [8,9]) have been carried out to decrease the vanadium ions permeability of Nafion membrane. As a result, these membranes showed lower vanadium ions permeability and improved cell performance in VRFB cells. Clearly, these insights obtained from Nafion-based works would help the design of low-cost alternative membranes with superior properties. Recently, alternative membranes such as cation exchange membranes [10e15], nanofiltration membranes [16e20] and anion exchange membranes [21e29] have been reported for VRFB application. Among these membranes, anion exchange membrane have received considerable attention for VRFB application, because the membrane enjoys lower vanadium permeability related to the Donnan exclusion effect between positively charged groups and vanadium ions. Both traditional quaternary ammonium group and pyridinium group can be served as ion exchange groups for anion exchange membrane. Pyridinium group showed more stability than quaternary ammonium group in oxidative vanadium electrolyte [28]. Moreover, pyridinium group, as a weak alkaline, can promote proton transport of membrane due to acidebase interactions [27]. In this paper, the low-cost poly(phenyl sulfone) with good commercial availability and chemical resistance was used to fabricate novel cost-effective poly(phenyl sulfone) anion exchange membranes with pyridinium groups (PyPPSU), and these membranes were first investigated in VRFB application. The membrane properties and VRFB cell performance were studied. 2. Experimental 2.1. Materials Poly(phenyl sulfone) (PPSU, Radel R-5500) (Solvay Advanced Polymers). Chloromethyl ethyl ether (CMEE) (Henan wanxiang technology & trade co., Ltd), Tin tetrachloride (SnCl4) (Sinopharm Chemical Reagent Co., Ltd), Vanadyl sulfate (VOSO4) (Shanghai Luyuan Fine Chemical Plant), Pyridine (Sinopharm Chemical Reagent Co., Ltd) and other chemicals were commercially obtained and used as received. 2.2. Preparation of anion exchange membrane Chloromethylated poly(phenyl sulfone) was prepared by chloromethylation reaction: after 5.0 g of poly(phenyl sulfone) had been dissolved in 75 mL tetrachloroethane, 0.7 mL SnCl4 was added and then 5 mL of chloromethyl ethyl ether (CMEE) was added to the solution. The mixture was maintained at 50  C for 10 h and poured into excess of ethanol. The precipitated polymer was dried at 60  C under vacuum for 48 h. The degree of substitution of resulting polymer was calculated from 1H NMR spectroscopy (Bruker Avance 500 M). The degree of substitution (DS) of chloromethylated poly(phenyl sulfone) (CMPPSU) was defined as the number of chloromethyl groups per polymer repeat unit. After the CMPPSU had been dissolved in 1-methyl-2pyrrolidone as a 10 wt. % solution, the solution was casted onto a smooth glass plate and evaporating the solvent at 60  C for 9 h to obtain base membrane, and then the base membrane was treated with the method according to Zhang et al. [28]. The base membrane was immersed in 1 mol L1 pyridine water solution at 40  C for 72 h to obtain poly(phenyl sulfone) anion exchange membrane with pyridinium groups (PyPPSU) (Scheme 1). After that, the membrane was immersed into 5 wt. % HCl to neutralize residual pyridine, and then washed with excess deionized water to remove residual HCl. The thickness of PyPPSU membranes were in the range of 40e45 mm.

329

Scheme 1. Synthesis of poly(phenyl sulfone) anion exchange membrane with pyridinium groups (PyPPSU).

2.3. Characterization of membrane The chemical structure of chloromethylated poly(phenyl sulfone) and PyPPSU membrane was confirmed using FT-IR analysis with a Varian 6400 FT-IR spectrometer. The ion exchange capability (IEC) of membrane was determined by titration [21]. A dried PyPPSU membrane (in chloride form) was obtained under vacuum at 60  C for 48 h, and then it was weighed and immersed in 25 mL 0.1 mol L1 NaNO3 solution for 48 h at room temperature. After that the solution was back titrated with 0.1 mol L1 AgNO3, and K2CrO4 was employed as indicator. IEC was calculated according to the following equation:

IEC ¼

VAgNO3  CAgNO3 M

where VAgNO3 was the volume of AgNO3 solution, CAgNO3 was the concentration of AgNO3 solution, and M was the weight of dried membrane, respectively. Water uptake of PyPPSU membrane was determined by equilibrating the sample of membrane with deionized water at room temperature for 24 h. The membrane was wiped using absorbent paper and weighed immediately. Then the membrane was dried at 60  C under vacuum for 48 h. Water uptake was calculated according to the following equation:

Water

uptake ð%Þ ¼

Wwet  Wdry  100% Wdry

where Wwet and Wdry were the weights of the membrane in wet and dry state, respectively. To evaluate the swelling ratio of the PyPPSU membranes in deionized water and vanadium electrolyte, membrane samples were dried at 60  C under vacuum for 48 h, and then the length of membrane were measured. After that the membranes were soaked in deionized water and 1.5 M VO2þ in 3 M H2SO4 solution for 24 h, respectively. The swelling ratio was calculated according to the following equation:

Swelling

ratio ¼

L1  L0  100% L0

330

B. Zhang et al. / Journal of Power Sources 282 (2015) 328e334

where L1 and L0 were the length of the soaked membrane and dry membrane, respectively. The area resistances of the membranes were measured by the method [30]. A conductivity cell was separated into two compartments filled with 1.5 M VOSO4 in 3 M H2SO4. The electric resistances of the conductivity cell with membrane (r1) and without membrane (r2) were measured by electrochemical impedance spectroscopy (EIS) using CH Instruments Electrochemical Workstation (CHI 660E) over a frequency range from 100 kHz to 1 Hz. The effective membrane area S of the conductivity cell was 1 cm2. The area resistance of the membrane r (U cm2) was calculated by the following equation:

25 mL min1. The volume of electrolyte solution was 30 mL for each half-cell. Chargeedischarge tests were conducted with constant current mode at 25  C on a battery test system CT-3008-5V/3A (Neware Co., Ltd, China), and the upper limit of charge voltage and lower limit of discharge were 1.65 V and 0.8 V, respectively. The coulombic efficiency (CE) was calculated as the discharge capacity divided by the charge capacity; Energy efficiency (EE) was calculated as the discharge energy divided by the charge energy; Voltage efficiency (VE) was calculated from VE ¼EE/CE. Open-circuit voltage (OCV) of the cell was monitored after the cell was charged to 1.65 V in the first cycle, and stopped when the OCV reached 0.8 V.

r ¼ ðr1  r2 Þ  S The permeability of VO2þ through membrane was evaluated by the methods described in literature [31,32]. The membrane was sandwiched between two half-cells (Scheme 2). The left half-cell was filled with 1.5 M VOSO4 in 3 M H2SO4, and the right half-cell was filled with 1.5 M MgSO4 in 3 M H2SO4. MgSO4 was used to balance the ionic strength and minimize the osmotic pressure between the two solutions. The active area of the membrane was 1.76 cm2 and volume of solutions in both half-cells was 30 mL. The concentration of VO2þ ion in MgSO4 solution was measured by using an UV/Vis spectrophotometer. The ion permeability was calculated by using the Fick Law:

VB

dCB ðtÞ P ¼ S ðCA  CB ðtÞÞ dt L

where VB is the volume of solution in the right half-cell; S and L are the effective area and thickness of the membrane, respectively; P is the permeability of VO2þ; CA is the concentration of VO2þ in the left half-cell; CB is the concentration of VO2þ in the right half-cell; and t is test time. Ex-situ chemical stability of the membranes was evaluated by the methods described in literature [23,33]. Dried membrane sample (0.1 g) was immersed in 50 mL 0.15 M VOþ 2 solution at a temperature of 40  C. The 0.15 M VOþ 2 solutions was prepared by 2þ diluting the 1.5 M VOþ was used as an 2 in 3 M H2SO4. The VO indicator to measure the stability of the membrane, because the 2þ oxidation of membrane leads the VOþ 2 reduce to VO . The con2þ centration of VO in the solution was measured by UV/Vis spectroscopy. The VRFB cell was assembled by sandwiching a membrane between two pieces of carbon felts (5 mm of thickness, 5 cm2). Two pieces of graphite plates were served as current collectors. The thickness of electrode frame was 4 mm. The negative electrolyte and positive electrolyte were 1.5 M V2þ/V3þ in 3.0 M H2SO4 and 1.5 M VO2þ/VOþ 2 in 3.0 M H2SO4, respectively, and they were circulated into the corresponding half-cell at a flow rate of about Membrane

1.5 M VOSO4 in 3 M H2SO4

3. Results and discussion 3.1. Membrane synthesis and properties The chloromethylated poly(phenyl sulfone) with different degrees of substitution were synthesized by varying the amount of SnCl4 (Table 1). In the 1H NMR spectrum of CMPPSU (Fig. 1), a new peak at 4.64 ppm was assigned to the methylene protons in the chloromethyl groups [34]. As mentioned in the experimental part, degree of substitution (DS) of chloromethylated poly(phenyl sulfone) was calculated from 1H NMR spectroscopy by taking the integral ratio of eCH2Cl (H(2)) at 4.64 ppm to the poly(phenyl sulfone) backbone protons at 7.90 ppm (H(11,12,13,14)) which were intact during the chloromethylation reaction, and the calculation equation was shown as following:

DS ¼

2A2 A11;12;13;14

where A2 was the integration value of H(2), A11,12,13,14 was the integration value of H(11,12,13,14), respectively. The DS values for CMPPSU-1 and CMPPSU-2 were 0.86, and 1.08, respectively. Anion exchange membranes were obtained by utilizing the high reactivity of the chloromethyl groups with pyridine to form pyridinium groups [28]. The membranes derived from the chloromethylated poly(phenyl sulfone) base membranes were named PyPPSU-1 and PyPPSU-2, respectively (Table 2). The chemical structures of PyPPSU membranes were confirmed using FT-IR analysis. As shown in Fig. 2 a new absorption peak at 1631 cm-1 for the PyPPSU membrane was observed, which attributed to the stretching vibration of C]C and C]N bonds in the pyridine ring [28,35]. The results indicated that pyridinium groups were successfully introduced into the PyPPSU membrane. Additionally, the PyPPSU membrane displayed absorbance at around 3391 cm-1, which was attributed to water bound to the ionic groups [26]. The properties of PyPPSU membrane including ion exchange capacity (IEC), water uptake and swelling ratio were summarized in Table 2. Water uptake increased with increasing IEC because of the water uptake of anion exchange membrane is strongly dependent upon the amount of hydrophilic ion exchange groups. As the IEC values increase from 1.60 to 1.95 mmol g1, the water uptake values increase from 21.3% to 35.4%, respectively. Swelling of membrane is an important issue for VRFB stacks because membrane stability

1.5 M MgSO4 in 3 M H2SO4 Table 1 Degree of substitution of CMPPSU.

Magnetic Stirring Bar

Magnetic Stirring Bar Bar Magnetic Stirring

Scheme 2. Diffusion cell for membrane vanadium permeation measurement.

Polymer

CMEE (mL)

SnCl4 (mL)

DS

CMPPSU-1 CMPPSU-2

5 5

0.7 0.9

0.86 1.08

B. Zhang et al. / Journal of Power Sources 282 (2015) 328e334

Fig. 1. 1H NMR spectra of PPSU and CMPPSU in CDCl3.

Table 2 Properties of PyPPSU prepared from CMPPSU. Polymer

Membrane

CMPPSU-1 CMPPSU-2

PyPPSU-1 PyPPSU-2

IEC (mmol g1)

Water uptake (%)

Swelling ratio (%) H2O

1.5 M VOSO4 in 3 M H2SO4

1.60 1.95

21.3 35.3

6.4 12.1

3.6 5.4

affected by scaling in membrane size [36]. However, very limited data is reported that the swelling ratio of membranes in vanadium electrolyte [21,28]. In this work, the swelling ratios of membrane in deionized water as well as in 1.5 M VOSO4 in 3 M H2SO4 were investigated. The swelling ratio increased with increasing IEC

Absorbance

3391 cm

because of the hydrophilicity of ion exchange groups. Furthermore, the swelling ratios of membranes in water were slightly higher than that in vanadium electrolyte, because swelling ratio depended on the osmotic pressure difference between the external solutions and internal solutions of the membrane, which is larger in deionized water than in vanadium electrolyte. The swelling ratio of PyPPSU-2 membranes was 5.4% in vanadium electrolyte indicating that PyPPSU-2 membranes have good dimension stability and would benefits the stability of membrane in VRFB stacks. Crossover of vanadium ions initiates undesired side reactions in the electrolytes, which would lower the voltage and reduces the system capacity, resulting in frequent maintenance of the electrolyte and increased operating cost of the VRFB system [37]. Therefore, the permeability of VO2þ through membranes is an important parameter. In this study, the permeability of VO2þ through PyPPSU membrane and Nafion117 membrane were measured and the results were summarized in Fig. 3. Clearly,VO2þ ions were depressed greatly in PyPPSU membrane compared with Nafion117 membrane. The vanadium permeability P were calculated and illustrated in Table 3. The P value of PyPPSU membranes (0.07  107e0.15  107 cm2 min1) is much lower than that of Nafion 117 (31.30  107 cm2 min1). Compared with perfluorinated Nafion117 membrane, the reduced permeability of PyPPSU membrane could be result by the Donnan exclusion effect of positively charged ion exchange groups with vanadium ions and the narrower, less connected hydrophilic channels in aromatic polymer based ion conducting membranes [24]. The low vanadium permeability of PyPPSU membrane would have a positive effect on the open-circuit voltage of VRFB assembled with PyPPSU membrane. Moreover, PyPPSU membranes have a similar P values to that of anion exchange membranes (0.02  107e0.09  107 cm2 min1) reported by Mai et al. [23]. As shown in Fig. 4, open circuit voltage (OCV) of VRFB cell with PyPPSU-2 and Nafion 117 membrane were monitored to investigate the self-discharge of the cell resulted by transferring of the vanadium ions across the membranes. As listed in Table 2, the selfdischarge duration of the VRFB cell with Nafion 117 lasted for about 110 h, which is about one fourth of that for VRFB cell with PyPPSU-2 (418 h). This result confirms that the PyPPSU membrane has significantly better suppressing vanadium crossover than Nafion117 membrane. In addition, this result is also in good accordance with the results of vanadium permeability test.

-1

1631 cm

-1

PyPPSU-2

PyPPSU-1

CMPPSU-1

4000

3500

3000

1500

1000

500

-1

Wave numbers (cm ) Fig. 2. FT-IR spectra of CMPPSU and PyPPSU membrane.

331

Fig. 3. Concentration of VO2þ in MgSO4 solution versus diffusion time.

332

B. Zhang et al. / Journal of Power Sources 282 (2015) 328e334

Table 3 Vanadium permeability, self-discharge time and area resistance of PyPPSU and Nafion 117 membrane. Membrane

Vanadium permeability P (10-7 cm2 min1)

Self-discharge time (h)

Area resistance (U cm2)

Nafion 117 PyPPSU-1 PyPPSU-2

31.30 0.07 0.15

110 e 418

0.60 1.81 0.58

3.2. Ex-situ chemical stability of the membrane Stability of the alternative membrane in VOþ 2 solutions is a critical issue for VRFB application. Ex-situ chemical stability of the membranes was test by immersing the membrane in 0.15 M VOþ 2 solutions. PyPPSU membranes changed their color (transparent to red), however, the solutions remained its color during the test. PyPPSU membranes are significantly stable than sulfonated poly(phenyl sulfone) membranes (the color of solution changed to green-colored only after 170 h) [33] indicating that membranes with positive groups have better oxidative stability than sulfonated polymer membranes. Moreover, this phenomenon has also been observed by Mai et al. [23]. This result may causes by Donnan exclusion effect between positive charged ion exchange groups and þ VOþ 2 , which limited the oxidative VO2 to permeate in to the membrane, and lead a slower oxidation. As shown in Fig. 5, the VO2þ concentration in the solution immersing PyPPSU membranes slightly increased with time. Nafion117 still remained its original color during the test and showed high stability because of its perfluorinated polymer backbone. In addition, the concentration of VO2þ of PyPPSU-2 membrane was higher than that of PyPPSU-1, indicated that PyPPSU-1 was more stable than PyPPSU-2, this could be attributed to the lower IEC and swelling ratio of PyPPSU-1, which provides less opportunity for the VOþ 2 to get into the hydrophilic domains to attack the membrane. This phenomenon has also been observed in sulfonated poly(phenylene) membranes [38].

3.3. VRFB cell performance The typical chargeedischarge curves of the VRFB cell with PyPPSU and Nafion 117 membrane are plotted in Fig. 6. Compared with the cell with Nafion 117 membrane, the cell with PyPPSU-2

Fig. 4. The open-circuit voltage (OCV) curves of VRFB cell with PyPPSU-2 and Nafion 117 membrane in self-discharge tests.

Fig. 5. VO2þ concentration of 0.15 M VOþ 2 solutions versus membranes immersion time at 40  C.

membrane exhibited slightly higher discharge voltage and comparable charge voltage, which suggested that the PyPPSU-2 membrane could have high voltage efficiency in VRFB. Furthermore, the cell with PyPPSU-2 membrane and cell with Nafion 117 membrane showed similar discharge capacity and charge capacity, respectively. Rate capability is a key parameter to evaluate the quality of VRFB system [20]. For a given output power, VRFB running at higher current density would require smaller stack size and area of electrode and membrane, which could lower the overall cost of the VRFB system and better for the practical application of VRFB. Therefore, the VRFB cell performances under different current densities were evaluated. Fig. 7 showed that the current efficiency (CE) of VRFB increased with the increase of chargeedischarge current density. From 40 mA cm2 to 100 mA cm2, the CE of VRFB with PyPPSU membrane was higher than that of VRFB with Nafion117 membrane. Because anion exchange membrane enjoys lower permeability of vanadium ions than Nafion117 membrane in VRFB. Moreover, similar trend has also been observed in VRFB with other anion exchange membranes [21e23].

Fig. 6. Chargeedischarge curves for VRFB cell with PyPPSU-2 membrane and Nafion 117 membrane at 80 mA cm2.

B. Zhang et al. / Journal of Power Sources 282 (2015) 328e334

Fig. 7. Coulombic efficiencies of VRFB cell with membranes at different current densities.

As shown in Fig. 8, because of the increase of ohmic resistance and the over potentials led by the increase of current densities, both the voltage efficiency (VE) of the cell with PyPPSU membrane and cell with Nafion 117 membrane decreased with the increasing chargeedischarge current densities. From 40 mA cm2 to 100 mA cm2, the VE of cell with PyPPSU-2 membrane was slightly higher than that of cell with Nafion 117 membrane. This is mainly because the area resistance of PyPPSU-2 (0.58 U cm2, Table 3) was slightly lower than that of Nafion117 membrane (0.60 U cm2). In addition, the VE of PyPPSU-1 membrane was lower than that of PyPPSU-2 and Nafion 117 membrane due to its higher area resistance (1.81 U cm2). Energy efficiency (EE) of cell with PyPPSU membrane and cell with Nafion117 membrane were outlined in Fig. 9. As shown, the EE of all cells decreased with the increase of current densities. From 40 mA cm2 to 100 mA cm2, the EE values (87.5%, 85.7%, 83.5%, 80.2%, at 40, 60, 80, 100 mA cm2, respectively) of cell with PyPPSU-2 membrane was higher than that (85.7%, 83.5%, 80.0%, 77.2% at 40, 60, 80, 100 mA cm2, respectively) of cell with Nafion 117 membrane. Furthermore, at a current density of 80 mA cm2 the cell with PyPPSU-2 membrane showed an EE of

Fig. 8. Voltage efficiencies of VRFB cell with membranes at different current densities.

333

Fig. 9. Energy efficiencies of VRFB cell with membranes at different current densities.

83.5%, which was comparable with the EE (84%) reported by Zhang et al. [26] and EE (83.6%) by Zhang et al. [28] under the similar test condition (80 mA cm2, 1.5 M V2þ/V3þ and 1.5 M VO2þ/VOþ 2 in 3.0 M H2SO4). Additionally, the cell with PyPPSU-2 showed high EE of 80.2% even at high current density of 100 mA cm2, which was higher than that of cell with Nafion 117 membrane (77.2%) under the same test condition. The results indicated that the cell with PyPPSU-2 membrane have high power density at high current density (>80 mA cm2). When the PyPPSU-2 membrane applied in VRFB system, smaller stack size, area of membrane and electrode would be required, and it could lower the overall cost of the VRFB system and better for the practical application of VRFB. Membrane durability is also a critical property for alternative membrane used in VRFB. To investigate the durability of PyPPSU membrane in VRFB, the cycling test was conducted. Due to time constrains, the cycling test was conducted for 100 cycles. As shown in Fig. 10, after 100 cycles, the VRFB cell with PyPPSU-2 membrane maintained high efficiencies at high chargeedischarge current density of 80 mA cm2. The CE, VE and EE remained about 98%, 80% and 79%, respectively, which indicated that PyPPSU membrane was

Fig. 10. Chargeedischarge cycling performance (80 mA cm2) of VRFB with PyPPSU membrane.

334

B. Zhang et al. / Journal of Power Sources 282 (2015) 328e334

stable during the long-term test. In addition, the VE of VRFB decreased gradually with cycle numbers. This is because concentration polarization which caused by the crossover of vanadium ions in negative electrolyte and positive electrolyte led to voltage loss during the cycle test [39]. And, similar phenomenon has also been observed in VRFB with other membranes [40]. The results of the VRFB cell tests suggested that the cost-effective PyPPSU membranes could be promising alternative membrane for VRFB application. Moreover, PyPPSU-2 membrane is significantly stable than sulfonated poly(phenyl sulfone) membrane which was only maintained for about 42 cycles at a similar test condition [33]. 4. Conclusions

[6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]

Poly(phenyl sulfone) anion exchange membranes with pyridinium groups (PyPPSU) were prepared and their properties were investigated for vanadium redox flow battery (VRFB) applications. The PyPPSU membranes had significantly lower permeability of vanadium ions than Nafion117 membrane, resulting in the selfdischarge duration of the VRFB cell with PyPPSU membrane (418 h) was about four times longer than that of VRFB cell with Nafion 117 membrane (110 h). The VRFB cell with PyPPSU-2 membrane showed higher energy efficiency and coulombic efficiency than that of cell with Nafion 117 membrane at current densities ranging from 40 to 100 mA cm2. At 100 mA cm2, EE of 80.2% was achieved with the PyPPSU-2 membrane, compared to 77.2% with Nafion 117 membrane. In addition, PyPPSU membranes showed good stability in the ex-situ chemical stability test and the 100-cycle test. Acknowledgment

[18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]

We greatly acknowledge financial support from the Natural Science Foundation of China (no. 21444006) and China Liaoning Provincial Education Department (no. L2013163). References [1] [2] [3] [4]

M. Rychcik, M. Skyllas-Kazacos, J. Power Sources 22 (1988) 59e67. C. Ding, H. Zhang, X. Li, T. Liu, F. Xing, J. Phys, Chem. Lett. 4 (2013) 1281e1294. J. Xi, Z. Wu, X. Qu, L. Chen, J. Power Sources 166 (2007) 531e536. J. Zeng, C. Jiang, Y. Wang, J. Chen, S. Zhu, B. Zhao, R. Wang, Electrochem. Commun. 10 (2008) 372e375. [5] B. Schwenzer, S. Kim, M. Vijayakumar, Z. Yang, J. Liu, J. Membr. Sci. 372 (2011) 11e19.

[34] [35] [36] [37] [38] [39] [40]

C. Jia, J. Liu, C. Yan, J. Power Sources 203 (2012) 190e194. J. Kim, J. Jeon, S. Kwak, Electrochem. Commun. 38 (2014) 68e70. Q. Luo, H. Zhang, J. Chen, P. Qian, Y. Zhai, J. Membr. Sci. 311 (2008) 98e103. J. Ma, S. Wang, J. Peng, J. Yuan, C. Yu, J. Li, X. Ju, M. Zhai, Eur. Polym. J. 49 (2013) 1832e1840. S. Winardi, S.C. Raghu, M. Oo, Q. Yan, N. Wai, T. Lim, M. Skyllas-Kazacos, J. Membr. Sci. 450 (2014) 313e322. Z. Li, J. Xi, H. Zhou, L. Liu, Z. Wu, X. Qiu, L. Chen, J. Power Sources 237 (2013) 132e140. H. Zhang, H. Zhang, X. Li, Z. Mai, W. Wei, Y. Li, J. Power Sources 217 (2012) 309e315. X. Ling, C. Jia, J. Liu, C. Yan, J. Membr. Sci. 415e416 (2012) 306e312. N. Wang, J. Yu, Z. Zhou, D. Fang, S. Liu, Y. Liu, J. Membr. Sci. 437 (2013) 114e121. J. Pan, S. Wang, M. Xiao, M. Hickner, Y. Meng, J. Membr. Sci. 443 (2013) 19e27. H. Zhang, H. Zhang, X. Li, Z. Mai, J. Zhang, Energy Environ. Sci. 4 (2011) 1676e1679. H. Zhang, H. Zhang, X. Li, Z. Mai, W. Wei, Energy Environ. Sci. 5 (2012) 6299e6303. J. Cao, H. Zhang, W. Xu, X. Li, J. Power Sources 249 (2014) 84e91. Y. Li, X. Li, J. Cao, W. Xu, H. Zhang, Chem. Commun. 50 (2014) 4596e4599. X. Wei, Z. Nie, Q. Luo, B. Li, B. Chen, K. Simmons, V. Sprenkle, W. Wang, Adv. Energy Mater. 3 (2013) 1215e1220. B. Zhang, S. Zhang, D. Xing, R. Han, C. Yin, X. Jian, J. Power Sources 217 (2012) 296e302. F. Zhang, H. Zhang, C. Qu, J. Phys, Chem. B 116 (2012) 9016e9022. Z. Mai, H. Zhang, H. Zhang, W. Xu, W. Wei, H. Na, X. Li, ChemSusChem 6 (2013) 328e335. S. Yun, J. Parrondo, V. Ramani, J. Mater. Chem. A 2 (2014) 6605e6615. M. Jung, J. Parrondo, C. Arges, V. Ramani, J. Mater. Chem. A 1 (2013) 10458e10464. F. Zhang, H. Zhang, C. Qu, ChemSusChem 6 (2013) 2290e2298. H. Zhang, H. Zhang, F. Zhang, X. Li, Y. Li, I. Vankelecom, Energy Environ. Sci. 6 (2013) 776e781. S. Zhang, B. Zhang, D. Xing, X. Jian, J. Mater. Chem. A 1 (2013) 12246e12254. D. Chen, M. Hickner, E. Agar, E. Kunbur, ACS Appl. Mater. Interfaces 5 (2013) 7559e7566. Z. Mai, H. Zhang, X. Li, C. Bi, H. Dai, J. Power Sources 196 (2011) 482e487. T. Sukkar, M. Skyllas-Kazacos, J. Appl. Electrochem. 34 (2004) 137e145. J. Qiu, M. Li, J. Ni, M. Zhai, J. Peng, L. Xu, H. Zhou, J. Li, G. Wei, J. Membr. Sci. 297 (2007) 174e180. S. Kim, T. Tighe, B. Schwenzer, J. Yan, J. Zhang, J. Liu, Z. Yang, M. Hickner, J. Appl. Electrochem. 41 (2011) 1201e1213. M. Hibbs, M. Hickner, T. Alam, S. Mcintyre, C. Fujimoto, C. Cornelius, Chem. Mater. 20 (2008) 2566e2573. Y. Li, T. Xu, M. Gong, J. Membr. Sci. 279 (2006) 200e208. B. Schwenzer, J. Zhang, S. Kim, L. Li, J. Liu, Z. Yang, ChemSusChem 4 (2011) 1388e1406. S. Kim, J. Yan, B. Schwenzer, J. Zhang, L. Li, J. Liu, Z. Yang, M. Hickner, Electrochem. Commun. 12 (2010) 1650e1653. C. Fujimoto, S. Kim, R. Stains, X. Wei, L. Li, Z. Yang, Electrochem. Commun. 20 (2012) 48e51. X. Xi, C. Ding, H. Zhang, X. Li, Y. Cheng, H. Zhang, J. Power Sources 274 (2015) 1126e1134. A. Chromik, A. Santos, T. Turek, U. Kunz, T. Haring, J. Kerres, J. Membr. Sci. 476 (2015) 148e155.