Enhanced hydroxide conductivity of imidazolium functionalized polysulfone anion exchange membrane by doping imidazolium surface-functionalized nanocomposites

Abstract

A new method was proposed to prepare a membrane with regional aggregation of functional groups by incorporating nanocomposites surface-functionalized with a large number of functional groups. Imidazolium surface-functionalized SiO₂ (SiO₂-Im) nanocomposites were synthesized by the reaction of 1,2-dimethylimidazole, γ-chloropropyl triethoxysilane and SiO₂ nanocomposites. The obtained SiO₂-Im nanocomposites were incorporated into imidazolium functionalized polysulfone (PSf-Im) to fabricate composite alkaline anion exchange membranes. The uniform dispersion of nanocomposites in the membrane was demonstrated by SEM. With increasing mass ratio of SiO₂-Im from 0% to 20%, hydroxide conductivity of composite membrane dramatically increases at first and then decreases. The composite membrane with 12 wt% of SiO₂-Im shows the highest conductivity, e.g., the hydroxide conductivity of the composite membrane based on PSf-Im with functionalization degree of 76% reaches 32 mS cm⁻¹ (at 20 °C) that is 68% higher than the membrane's without doping SiO₂-Im (19 mS cm⁻¹). In addition, adding SiO₂-Im has a slight effect on water uptake and swelling ratio of composite membrane. It indicates that doping surface-functionalized nanocomposites is a simple and effective method to enhance the hydroxide conductivity without increasing swelling.

---

Introduction

Alkaline anion exchange membrane fuel cells (AAEMFCs) have been recognized as one of the most promising clean power technologies applied in vehicles and mobile electronic devices.^1,2 By using the hydroxide (OH⁻) as conductive ions, AAEMFCs have the potential to solve the problems of high cost and low durability of electrocatalyst (platinum) in the acid environment existed in proton exchange membrane fuel cells (PEMFCs), it shows advantage of high electrode kinetics which allows the use of non-precious metal catalysts, like metal–nitrogen–carbon (M–N–C).^3–8

As one of the key components of AAEMFCs, alkaline anion exchange membranes (AAEMs) serve as fuel/oxidant separator and hydroxide conductor simultaneously.^9–13 A common method for producing the AAEMs is the chloromethylation of polymers and following introduction of functional groups. Unfortunately, low hydroxide conductivity is still a big challenge for commercialization.^14–17 In general, the main chains of most polymer matrixes are stiffness and difficult to move freely, making the functional groups uniformly distributed in the membrane. As a result, it is difficult to interconnect functional groups to form continuous hydroxide transport channels. In order to enhance the hydroxide conductivity of AAEMs, an effective way is increasing ion exchange capacity (IEC).¹⁸ By increasing IEC, the concentrations of functional groups are effectively raised, and large amounts of functional groups dispersed in the membrane make the channels easy to be interconnected, leading to high conductivity. However, the large IEC also results in excessive water uptake, thus causing severe swelling and poor mechanical strength of the membrane. Introducing crosslinked structures has been frequently used to limit the swelling of the membrane with high IEC, but the degree of crosslinking is hard to control and some of the crosslinked membranes are a little fragile in the dry state.^19–27 Recently, it was found that raising the regional aggregation of functional groups could facilitate the formation of the continuous channels at a low IEC.^28–32 Several methods have been developed to achieve the regional aggregation of functional groups. Grafting multiple functional groups in one repeat unit of polymers could make the naturally regional aggregation of function groups.^33,34 The introduction of a long hydrophobic side chain between the polymer main chain and the functional group could enhance the mobility of ionic functional groups, driving their aggregation during the membrane formation.^35–37

In this work, a new method was proposed to prepare the membrane with the regional aggregation of functional groups by incorporating nanocomposites surface-functionalized with a large number of functional groups. The SiO₂-Im nanocomposites were prepared by the reaction of 1,2-dimethylimidazole, γ-chloropropyl triethoxysilane and SiO₂ nanocomposites.³⁸ Then the composite membrane was prepared by doping the SiO₂-Im nanocomposites into the imidazolium functionalized polysulfone. Large amounts of imidazolium functional groups clustering on the SiO₂-Im nanocomposites surfaces lead to the regional aggregation of functional groups in the composite membrane, which could promote the formation of continuous hydroxide transport channels. The properties of water uptake, swelling ratio, hydroxide conductivity and morphology were investigated. The effect of the SiO₂-Im nanocomposites dosage on the properties of the membrane was also studied.

Experimental

Materials

Udel P3500 polysulfone (PSf) were obtained from Amoco Performance Products Inc. Chloromethyl octylether (CMOE) was synthesized according to the ref. 39. Acetone, dichloromethane, ethanol, potassium hydroxide, 1-methyl-2-pyrrolidone (NMP) and toluene were purchased from Tianjin Guangfu Fine Chemical Research Institute (China). γ-Chloropropyl triethoxysilane, stannic chloride, silica (SiO₂) and 1,2-dimethylimidazole were obtained from Aladdin company and used as received without further purification. All the commercial chemicals used in the experiments are analytical grade. In order to reduce the effect of CO₂ dissolved in deionized water on the prepared membranes, fresh deionized water before being used was boiled to remove CO₂.

Synthesis of chloromethylated PSf (CMPSf)

Polysulfone was chloromethylated using the CMOE as chloromethylating agent, stannic chloride as catalyst, and dichloromethane as solvent. Specifically, 1 g PSf was dissolved in 30 mL dichloromethane at 30 °C with stirring, and then 6 mL freshly synthesized CMOE and 0.3 mL stannic chloride was added. The reaction was kept at 30 °C for a certain time. After that, the product polymer, CMPSf, was obtained by precipitation in ethanol, filtration and washing with ethanol several times, and then drying in vacuum at 50 °C for 24 h. The chemical structure of the CMPSf is shown in Fig. 1.

Fig. 1 ¹H NMR spectra of CMPSf and PSf-Im.

Preparation of imidazolium surface-functionalized SiO₂ (SiO₂-Im) nanocomposites

SiO₂-Im nanocomposites were synthesized referring to the literature.³⁸ 1,2-Dimethylimidazole (2 g, 0.02 mol), γ-chloropropyl triethoxysilane (2 g, 0.008 mol) and 1 g SiO₂ were mixed with 40 mL of dry toluene, and then the reaction solution was refluxed with stirring under nitrogen atmosphere for 24 h. After cooling, the reaction mixture was filtered and washed with toluene repeatedly, a white powder was obtained, and then vacuum drying at 50 °C for 24 h. The zeta potential of the particle surface was used to confirm the successful synthesis of the product. Comparing to the SiO₂ surface potential of −42.8 mV, the product-nanocomposites presented surface potential of 46.2 mV.

Preparation of imidazolium functionalized polysulfone/SiO₂-Im composite membranes

Typically, a solution was made by dissolving 0.15 g CMPSf into 3 mL NMP, and stirred at 80 °C for 1 h to keep homogeneous mixing. After that, 0.05 g 1,2-dimethylimidazole was added into the solution quickly. The mixture reacted at 80 °C for 5 h with stirring. By pouring the reaction mixture into acetone, pale yellow precipitation sank to the bottom of the solution, imidazolium functionalized polysulfone (PSf-Im) was obtain. The chemical structure of the PSf-Im was shown in Fig. 1.

To get composite membranes, the reaction solution jumping over the precipitation process was mixed with a certain quality of SiO₂-Im nanocomposites, and sonicated for 1 h to keep homogeneous mixing. After that, the solution was poured onto a glass plate to cast the membrane. After curing and drying at 60 °C for 2 days, the membrane of SiO₂-Im composite imidazolium functionalized polysulfone was formed and peeled off from the glass plate. To obtain the hydroxide-functionalized composite membrane (PSf-Im/SiO₂-Im), the membrane was soaked in 1 M KOH solution at room temperature for 48 h for ion-exchanging, followed by washing and immersion with deionized water for additional 48 h to completely remove the residual KOH.

As a comparison, we prepared the pristine imidazolium functionalized polysulfone membrane (PSf-Im) as the same process.

¹H NMR

¹H NMR was performed using INOVA 400 MHz spectrometer. ¹H NMR spectroscopy was used to confirm the chemical structures of CMPSf and PSf-Im, and predict the degree of chloromethylation (DCs) of CMPSf. CMPSf was directly dissolved in deuterated chloroform, and PSf-Im solution was prepared by deuterated dimethyl sulfoxide (DMSO), where tetramethylsilane (TMS) was used as the internal standard in all cases.

Ion exchange capacity (IEC)

The IEC of the PSf-Im/SiO₂-Im membrane was measured by the conventional back titration method. Typically, a membrane sample was soaked in a known volume (15 mL) of 0.01 M HCl standard solution for 24 h, the solution was then titrated against 0.01 M NaOH standard solution with phenolphthalein as the indicator. The IEC of the PSf-Im/SiO₂-Im membrane was calculated by the following equation:where V_HCl and C_HCl are the volume and concentration of HCl solution before titration, respectively; V_NaOH and C_NaOH are the volume and concentration of NaOH solution consumed in titration, respectively; M_dry is the mass of the dry membrane sample.

Water uptake and swelling ratio test

PSf-Im/SiO₂-Im membranes were immersed in deionized water for 12 h to ensure the full hydration at a certain temperature. Subsequently, the wet membrane was taken from the deionized water, and then the liquid water on the surface was removed quickly with filter paper. After that, the weight, lengths and widths of the wet membrane were measured rapidly. The test would be continued from 20 °C to 60 °C. After which the membrane samples were vacuum dried at 50 °C for 12 h, water uptake and swelling ratio were determined by measuring the difference in weight and dimension between dry and wet membrane. They can be calculated by the two following equations, respectively:where W_wet and W_dry are the weights of the wet and the dry membrane sample, respectively; l_wet and l_dry are the average lengths [l_wet = (l_wet1 × l_wet2)^1/2, l_dry = (l_dry1 × l_dry2)^1/2] of the wet and the dry membrane sample, respectively; l_wet1, l_wet2 and l_dry1, l_dry2 are the lengths and widths of the wet and the dry membrane sample, respectively.

Hydroxide conductivity test

Hydroxide conductivity was measured under the condition of complete immersion in deionized water by a typical four-probe AC impedance spectroscopy with the IVIUM-N-STAT N27133, and 1 to 10⁵ Hz was chosen as the scanning frequency range. The measurement apparatus used two platinum foils as the current carriers and two platinum wires as the potential sensors. Prior to the measurement, all the membranes were fully hydrated in deionized water for 24 h. The membrane conductivity (σ, S cm⁻¹) is calculated from the following equation:where L is the distance (cm) between the two potential electrodes, d and W are the thickness (cm) and width (cm) of the membrane sample, respectively, and R is the membrane resistance obtained from the AC impedance data.

Thermogravimetric analysis (TGA)

Thermogravimetric (TG) analysis was performed on a METTLER TOLEDO thermogravimetric analyzer. In the atmosphere, samples (ca. 10 mg) were heated from 30 °C to 800 °C at a heating rate of 10 °C min⁻¹.

Alkaline stability

The alkaline stability of the PSf-Im/SiO₂-Im membrane was examined by immersing a membrane sample in a 1 M KOH solution at 60 °C. The degradation of the membrane sample was evaluated by testing the change of hydroxide conductivity.

Results and discussion

Structure confirmation

As an electrophilic substitution, the chloromethylation reaction takes place mainly at the H_a position due to its relatively high electron density. As it is shown in Fig. 1, a new characteristic single-peak of protons (H_f) in chloromethyl groups (–CH₂Cl) appears at 4.64 ppm, confirming the successful chloromethylation of PSf. The DCs of CMPSf is calculated by the following equation: DC = 2A(H_f)/A(H_d), where A(H_f) and A(H_d) are the integral area of the H_f and H_d peak, respectively. After the Menshutkin reaction between CMPSf and 1,2-dimethylimidazole, the single-peak at 4.64 ppm almost disappears and a new peek arises at 5.4 ppm (H′_f), which indicate chloromethyl groups completely convert to imidazolium groups. The proton signal of methyl groups of imidazole ring appear (7.6 ppm for H_i and H_h, 3.6 ppm for H_j and 2.6 ppm for H_g). All these indicate the successful synthesis of PSf-Im and the complete conversion from chloromethyl groups to imidazolium groups.

Morphology of the nanocomposites and composite membrane

Fig. 2 shows the TEM micrographs for SiO₂ and SiO₂-Im nanocomposites. The size of the nanoparticles is about 30 nm. The structural features for the nanocomposites remain almost unchanged before and after functionalization.

Fig. 2 TEM images of SiO₂ and SiO₂-Im nanocomposites. (a) SiO₂ nanocomposites and (b) SiO₂-Im nanocomposites.

The microstructures of the pristine PSf-Im membrane and PSf-Im/SiO₂-Im composite membrane are shown in Fig. 3. The uniform dispersion of the SiO₂-Im nanocomposites is observed in the composite membrane, suggesting the good compatibility. There are no remarkable asymmetric dispersion of SiO₂-Im nanocomposites domains and formation of cracks on the SiO₂-Im nanocomposites rich surface.

Fig. 3 SEM images of PSf-Im and PSf-Im/SiO₂-Im membranes. (a) PSf-Im membrane, (b) PSf-Im/SiO₂-Im composite membrane.

TEM images for PSf-Im and PSf-Im/SiO₂-Im membranes are shown in Fig. 4. The pristine PSf-Im membrane shows a homogeneous microstructure, indicating a uniform dispersion of functional groups. A micro-phase separation structure was observed for PSf-Im/SiO₂-Im membrane. It suggests that doping imidazolium-functionalized nanocomposites improved the micro-phase separation structure, thus constructing effective hydroxide conducting channels.

Fig. 4 TEM images for PSf-Im and PSf-Im/SiO₂-Im membranes. (a) PSf-Im membrane, (b) PSf-Im/SiO₂-Im composite membrane.

IEC, water uptake and swelling ratio

The IEC significantly affects water uptake, swelling ratio and hydroxide conductivity of the membrane. The IEC of composite membranes with different mass ratio of SiO₂-Im are listed in Table 1. To be convenience, the membrane is named as PSf-Im-xx%/SiO₂-Im yy%, where xx% is the DCs of the membrane, and yy% represents the mass ratio of SiO₂-Im in the membrane. The pristine PSf-Im membranes are denoted as PSf-Im-xx%, where xx% is the DC of PSf-Im. The IEC increases with raising the degree of functionalization of PSf. When the mass ratio of SiO₂-Im increases from 0% to 20%, the IEC slightly deceases suggesting the nanocomposites have a little lower IEC than PSf-Im.

Table 1 The IEC of PSf-Im/SiO₂-Im composite membranes with different contents of SiO₂-Im

Membrane	IEC (mmol g^−1)
	0^a%	4^a%	8^a%	12^a%	16^a%	20^a%
a Mass ratio of SiO2-Im contained in composite membranes.b To represent a series of composite membranes with same value of DCs and different mass ratio of SiO2-Im.
PSf-Im-76%/SiO2-Im^b	1.35	1.36	1.34	1.32	1.30	1.28
PSf-Im-92%/SiO2-Im^b	1.42	1.41	1.39	1.37	1.36	1.35
PSf-Im-102%/SiO2-Im^b	1.64	1.60	1.59	1.55	1.54	1.52

---

As shown in Fig. 5, with increasing the mass ratio of SiO₂-Im from 0% to 20%, water uptake of PSf-Im-102%/SiO₂-Im composite membranes slightly increases from 100% to 105% at 20 °C. It indicates that adding SiO₂-Im nanocomposites has no apparent effect on the water uptake of composite membranes. For Fig. 6, it shows the water uptake vary with the temperature. For example, with the temperature increasing from 20 °C to 60 °C, water uptake of PSf-Im-102%/SiO₂-Im 12% membrane slightly rises from 105% to 114%, because higher temperature can make the polymer chains have more energy to move freely, and further causes composite membrane to absorb more water. By comparing water uptake of composite membranes which are different in DCs, it is clear that DC is the main factor of making the water uptake different. When DCs increases from 76% to 102%, water uptake dramatically increases from 50% to 100% at 20 °C, because the higher IEC would make the ion clusters larger and hydrophilic channels easier and better to be continuous.

Fig. 5 Water uptake of PSf-Im/SiO₂-Im composite membranes with different content of SiO₂-Im at 20 °C.

Fig. 6 Water uptake of PSf-Im/SiO₂-Im composite membranes with 12 wt% of SiO₂-Im at different temperature.

Fig. 7 illustrates the swelling ratio of composite membranes with different content of SiO₂-Im. With increasing SiO₂-Im content, there is almost no change in swelling ratio at 20 °C. The possible reason is that the water uptake and swelling ratio mainly depend on the properties of the pristine PSf-Im membrane. As shown in Fig. 8, the temperature also has a little effect on the swelling ratio of composite membranes.

Fig. 7 Swelling ratio of PSf-Im/SiO₂-Im composite membranes with different content of SiO₂-Im at 20 °C.

Fig. 8 Swelling ratio of PSf-Im/SiO₂-Im composite membranes with 12 wt% of SiO₂-Im at different temperature.

Hydroxide conductivity

Hydroxide conductivity of composite membranes with different mass ratios of SiO₂-Im are shown in Fig. 9. For all the composite membranes, hydroxide conductivity firstly increases and then decreases with increasing SiO₂-Im doping content. The composite membrane with 12% of SiO₂-Im has the highest conductivity, e.g. the conductivity of PSf-Im-76%/SiO₂-Im 12% membrane at 20 °C reaches 32 mS cm⁻¹, 68% higher than that of pristine PSf-Im-76% membrane (19 mS cm⁻¹). Since all the composite membranes have similar IEC with pristine PSf-Im membrane, it is speculated that doping SiO₂-Im nanocomposites may improve hydroxide conducting channels, further enhancing the conductivity. Highly clustered imidazolium groups on the surface of SiO₂-Im nanocomposites could facilitate the formation of the continuous hydroxide conducting channels. According to the convection and Grotthuss transport mechanisms, those channels could greatly promote the hydroxide conductivity, as shown in Fig. 10. For the composite membranes at lower mass ratio (no more than 12%), adding SiO₂-Im could effectively help interconnecting the ion clusters in the membrane to form channels and micro phase separation, so hydroxide conductivity dramatically increases, but when mass ratio of the SiO₂-Im overs a certain value (higher than 12%), too many additions would block ion clusters channels, and then the hydroxide conductivity slopes down. With increasing IEC of pristine PSf-Im membranes, the increment of hydroxide conductivity induced by doping SiO₂-Im nanocomposites decreases. Hydroxide conductivity of PSf-Im-102%/SiO₂-Im 12% at 20 °C is 45 mS cm⁻¹ that is an increase of 25% compared to pristine PSf-Im-102% membrane (36 mS cm⁻¹). It is because that higher IEC would make the ion clusters larger and hydrophilic channels easier and better to be continuous, so that doping SiO₂-Im nanocomposites has a decreased effect on the conductivity of the composite membrane.

Fig. 9 Hydroxide conductivity of PSf-Im/SiO₂-Im composite membranes with different content of SiO₂-Im at 20 °C.

Fig. 10 Schematic for the convection and Grotthuss transport mechanisms.

As shown in Fig. 11, hydroxide conductivity for both membranes with and without doping SiO₂-Im increases with temperature. Hydroxide conductivity of PSf-Im-76%/SiO₂-Im 12% membrane increases from 32 mS cm⁻¹ to 60 mS cm⁻¹ with increasing temperature from 20 °C to 60 °C. By contrast, the conductivity of pristine PSf-Im-76% membrane increases from 19 mS cm⁻¹ to 36 mS cm⁻¹. It is a considerable improvement. The conductivity of PSf-Im-102%/SiO₂-Im 12% membrane reaches 83 mS cm⁻¹ that is also much higher than that of PSf-Im-102% membrane (74 mS cm⁻¹). Based on the Arrhenius relationship between hydroxide conductivity and temperature, the apparent activation energies (ΔE_a) of hydroxide conductivity for PSf-Im-76%/SiO₂-Im 12% and PSf-Im-76% membranes are 13.1 and 12.5 kJ mol⁻¹, respectively. It is indicated that the temperature has similar effects on the conductivity for composite membrane than for pristine one.

Fig. 11 Hydroxide conductivity of PSf-Im/SiO₂-Im and PSf-Im membranes at different temperature.

Thermal and alkaline stabilities

Fig. 12 shows TGA and DTG curves of the PSf-Im-76% and PSf-Im-76%/SiO₂-Im 12% membrane. T_OD represents the onset decomposition temperature, and T_FD stands for the fastest decomposition temperature, respectively. The decomposition of PSf-Im-76% could be divided into three steps. The first step (T_OD: 171 °C, T_FD1: 233 °C) is due to the decomposition of side chain imidazolium groups. The second step (T_FD2: 401 °C) is attributed to the decomposition of the residual functional groups, e.g., –CH₂Cl, hydroxylmethylene, methyl groups. The third step (T_FD3: 674 °C) corresponds to the degradation of polymer main chains. There are also three decomposition steps in the TGA for PSf-Im-76%/SiO₂-Im 12%: first step (T_OD: 165 °C, T_FD1: 259 °C), second step (T_FD2: 446 °C) and third step (T_FD3: 668 °C). The TGA results indicate that doping SiO₂-Im has a little effect on thermal property of the membranes.

Fig. 12 TGA curves of PSf-Im-76% and PSf-Im-76%/SiO₂-Im 12% membrane.

As shown in Fig. 13, the hydroxide conductivity of the membranes decreased with immersion time in the KOH solution. The hydroxide conductivity dropped more rapidly for composite membranes than for pristine membranes in the first 12 h, after that, the composite membrane showed the same conductivity with pristine membrane. It indicates that the SiO₂-Im completely degrades, causing the disappearance of hydroxide conducting channels constructed by SiO₂-Im. Although the SiO₂-Im nanocomposites are non-stable in hot alkaline solution, doping functionalized nanocomposites is still a simple and effective method to improve the hydroxide conductivity without increasing swelling ratio. The alkaline stability could be improved by using other stable nanocomposites instead of SiO₂.

Fig. 13 Hydroxide conductivity of PSf-Im/SiO₂-Im and PSf-Im in 1 M KOH solution at 60 °C.

Conclusions

Imidazolium surface-functionalized SiO₂ nanocomposites were synthesized with 1,2-dimethylimidazole, γ-chloropropyl triethoxysilane and SiO₂ nanocomposites. By mixing different mass ratios of SiO₂-Im nanocomposites with imidazolium functionalized polysulfone, composite anion exchange membranes were prepared. SEM images confirmed the uniform dispersion of nanocomposites in membrane. With increasing mass ratio of the nanocomposites from 0% to 20%, hydroxide conductivity of composite membranes dramatically increases at first and then decreases. The composite membrane with 12 wt% of SiO₂-Im shows the highest conductivity, e.g., the hydroxide conductivity of PSf-Im-76%/SiO₂-Im 12% reaches 32 mS cm⁻¹ (at 20 °C) that is 68% higher than PSf-Im-76% (19 mS cm⁻¹). In addition, adding nanocomposites has a slight effect on water uptake and swelling ratio of the membranes. It indicates that doping surface-functionalized nanocomposites is an effective method to enhance the hydroxide conductivity without the challenge of excessive swelling.

Acknowledgements

The authors thank the support of Major National Scientific Instrument Development Project (Grant no. 21527812), National Natural Science Foundation of China (Grant no. 21406031 and 21476044), the Changjiang Scholars Program (Grant no. T2012049), the State Key Laboratory of Fine Chemicals (KF1507), and State Key Laboratory of fine chemicals (Panjin) project (Grant No. JH2014009).

Notes and references

1. G. Merle, M. Wessling and K. Nijmeijer, J. Membr. Sci., 2011, 377, 1–35.
2. J. R. Varcoe, P. Atanassov, D. R. Dekel, A. M. Herring, M. A. Hickner, P. A. Kohl, A. R. Kucernak, W. E. Mustain, K. Nijmeijer, K. Scott, T. Xu and L. Zhuang, Energy Environ. Sci., 2014, 7, 3135–3191.
3. W. Ding, L. Li, K. Xiong, Y. Wang, W. Li, Y. Nie, S. Chen, X. Qi and Z. Wei, J. Am. Chem. Soc., 2015, 137, 5414–5420.
4. Z. Li, G. Li, L. Jiang, J. Li, G. Sun, C. Xia and F. Li, Angew. Chem., 2015, 54, 1494–1498.
5. H. Peng, F. Liu, X. Liu, S. Liao, C. You, X. Tian, H. Nan, F. Luo, H. Song, Z. Fu and P. Huang, ACS Catal., 2014, 4, 3797–3805.
6. K. P. Singh, E. J. Bae and J. S. Yu, J. Am. Chem. Soc., 2015, 137, 3165–3168.
7. Y. Wu, C. Wu, F. Yu, T. Xu and Y. Fu, J. Membr. Sci., 2008, 307, 28–36.
8. W. Xia, R. Zou, L. An, D. Xia and S. Guo, Energy Environ. Sci., 2015, 8, 568–576.
9. S. Gu, R. Cai, T. Luo, Z. Chen, M. Sun, Y. Liu, G. He and Y. Yan, Angew. Chem., 2009, 48, 6499–6502.
10. X. Lin, M. Gong, Y. Liu, L. Wu, Y. Li, X. Liang, Q. Li and T. Xu, J. Membr. Sci., 2013, 425, 190–199.
11. X. Lin, L. Wu, Y. Liu, A. L. Ong, S. D. Poynton, J. R. Varcoe and T. Xu, J. Power Sources, 2012, 217, 373–380.
12. K. J. Noonan, K. M. Hugar, H. A. t. Kostalik, E. B. Lobkovsky, H. D. Abruna and G. W. Coates, J. Am. Chem. Soc., 2012, 134, 18161–18164.
13. C. Qu, H. Zhang, F. Zhang and B. Liu, J. Mater. Chem., 2012, 22, 8203–8207.
14. J. Han, H. Peng, J. Pan, L. Wei, G. Li, C. Chen, L. Xiao, J. Lu and L. Zhuang, ACS Appl. Mater. Interfaces, 2013, 5, 13405–13411.
15. X. Yan, S. Gu, G. He, X. Wu and J. Benziger, J. Power Sources, 2014, 250, 90–97.
16. Y. Zha, M. L. Disabb-Miller, Z. D. Johnson, M. A. Hickner and G. N. Tew, J. Am. Chem. Soc., 2012, 134, 4493–4496.
17. F. Zhang, H. Zhang and C. Qu, J. Mater. Chem., 2011, 21, 12744–12752.
18. X. Yan, S. Gu, G. He, X. Wu, W. Zheng and X. Ruan, J. Membr. Sci., 2014, 466, 220–228.
19. X. Lin, Y. Liu, S. D. Poynton, A. L. Ong, J. R. Varcoe, L. Wu, Y. Li, X. Liang, Q. Li and T. Xu, J. Power Sources, 2013, 233, 259–268.
20. G. Nie, X. Li, J. Tao, W. Wu and S. Liao, J. Membr. Sci., 2015, 474, 187–195.
21. J. Ran, L. Wu, Q. Ge, Y. Chen and T. Xu, J. Membr. Sci., 2014, 470, 229–236.
22. J. Wang, G. He, X. Wu, X. Yan, Y. Zhang, Y. Wang and L. Du, Journal of Membrane Science, 2014, 459, 86–95.
23. J. Wang, R. He and Q. Che, J. Colloid Interface Sci., 2011, 361, 219–225.
24. J. L. Wang, Q. T. Che and R. H. He, J. Electrochem. Soc., 2012, 160, 168–174.
25. L. Wu, Q. Pan, J. R. Varcoe, D. Zhou, J. Ran, Z. Yang and T. Xu, J. Membr. Sci., 2015, 490, 1–8.
26. Y. Wu, J. Luo, L. Zhao, G. Zhang, C. Wu and T. Xu, J. Membr. Sci., 2013, 428, 95–103.
27. Y. Xu, J. Yang, N. Ye, M. Teng and R. He, Eur. Polym. J., 2015, 73, 116–126.
28. L. Liu, C. Tong, Y. He, Y. Zhao and C. Lü, J. Membr. Sci., 2015, 487, 99–108.
29. R. P. Pandey, A. K. Thakur and V. K. Shahi, J. Membr. Sci., 2014, 469, 478–487.
30. B. Shi, Y. Li, H. Zhang, W. Wu, R. Ding, J. Dang and J. Wang, J. Membr. Sci., 2016, 498, 242–253.
31. Y. Wan, B. Peppley, K. A. M. Creber and V. T. Bui, J. Power Sources, 2010, 195, 3785–3793.
32. Z. Yi, C. Chaorong, Z. Danying and Z. Hongwei, Electrochim. Acta, 2015, 177, 128–136.
33. Z. Zhang, K. Shen, L. Lin and J. Pang, J. Membr. Sci., 2016, 497, 318–327.
34. Y. Z. Zhuo, A. Nan Lai, Q. G. Zhang, A. M. Zhu, M. L. Ye and Q. L. Liu, J. Membr. Sci., 2015, 491, 138–148.
35. N. Li, Y. Leng, M. A. Hickner and C. Y. Wang, J. Am. Chem. Soc., 2013, 135, 10124–10133.
36. C. Wang, B. Shen, C. Xu, X. Zhao and J. Li, J. Membr. Sci., 2015, 492, 281–288.
37. L. Zhu, J. Pan, Y. Wang, J. Han, L. Zhuang and M. A. Hickner, Macromolecules, 2016, 49, 815–824.
38. M. Sankar, T. G. Ajithkumar, G. Sankar and P. Manikandan, Catal. Commun., 2015, 59, 201–205.
39. A. Warshawsky and A. Deshe, J. Polym. Sci., Polym. Chem. Ed., 1985, 23, 1839–1841.

---