Nanofiber mats electrospun from composite proton exchange membranes prepared from poly(aryl ether sulfone)s with pendant sulfonated aliphatic side chains

Abstract

A series of cardo poly(aryl ether sulfone) copolymers bearing pendant sulfonated aliphatic side chains were synthesized and electrospun into nanofibers. The sulfonated poly(aryl ether sulfone) nanofibrous mats were filled with appropriate amounts of Nafion solution. The composite membranes showed significantly reduced swelling and excellent mechanical properties as well as appropriate proton conductivity. These membranes in particular exhibited much lower methanol permeability, in the range of 10⁻⁷ cm² s⁻¹, and higher selectivity, at about 10⁵ S s cm⁻³, than did Nafion 117 in our experiments. The results show that the fabricated nanofiber-based composite membranes can be used as a promising proton exchange membrane for direct methanol fuel cell applications.

---

1. Introduction

Of all types of fuels cells, direct methanol fuel cells (DMFC) are the focus of the most attention because of their versatile and multiple applications for automobiles and other forms of transportation.^1,2 As a key element of DMFC, proton exchange membranes (PEM) have become highly developed because of the contribution from perfluorinated acid polymers (e.g., Nafion) with high proton conductivity, good mechanical properties, and excellent stability.³ However, these membranes have some drawbacks such as high cost, low conductivity at elevated temperatures, and high methanol crossover, which hinder the widespread commercialization of DMFC.⁴

To overcome these defects, inorganic nanoparticles such as silica⁵ and titania⁶ have been incorporated into Nafion to form hybrid membranes. Unfortunately, most filler particles used for these membranes lack protogenic groups, resulting in relatively slow proton transport. While filler materials such as the sulfonic acid functionalized single-walled carbon nanotubes could offer additional protogenic groups for acid–water clusters,⁷ the low content of dispersed filler particles makes it difficult for them to form effective proton-transport pathways in membranes. Recently, Nafion has been entrapped within a poly(vinyl pyrrolidone) (PVP) network as a semi-interpenetrating polymer network membrane.⁸ In this association, interactions between the PVP amino groups and the sulfonic acid groups of Nafion form a crosslinked ionomer network, which decreases methanol permeation and promotes proton transfer. The overall performance of these Nafion-based modified membranes did not, however, meet the requirements of practical fuel cell applications. Therefore, new strategies are necessary and need to be explored.

Electrospinning is a straightforward method for producing fine polymeric fibers with a small diameter.⁹ Electrospun nanofibrous mats have gained considerable interest since they provide high porosity and a three-dimensional network structure.¹⁰ These distinct features make electrospun mats very useful for numerous applications such as ideal porous substrates for filters¹¹ and conductive materials.¹² Nafion-filled electrospun nanofibrous poly(vinylidene fluoride) composite membranes reported by Choi et al.¹³ showed improved mechanical properties, but decreased proton conductivity with the addition of the non-ionic nanofiber mats. Ballengee and Pintauro reported Nafion film reinforced by poly(phenyl sulfone) nanofibers, with the nanofiber composite membranes exhibiting very low in-plane water swelling and relatively good mechanical properties.¹⁴ Shabani and co-workers,^15,16 prepared composite PEM that are sulfonated poly(ether sulfone) nanofiber webs filled with Nafion. The composite PEM from these polyelectrolyte nanofibers display a higher electrochemical performance than do commercial Nafion membranes. However, the main chain-type sulfonated poly(ether sulfone) was used as an electrospinning material, which leads to high water uptake of the composite film (i.e., the SPES-N-N and SPES-N-N 112 were 13.5% and 16.3%, respectively). According to these results, methanol permeability of the bilayer or triple-layer membrane is still in the same order of magnitude as for Nafion 117.

In order to optimize DMFC performance, it is necessary to find and use polyelectrolyte nanofibers with extraordinary properties such as small diameter, high resistance to swelling, low methanol permeability, superior proton conductivity and good mechanical properties. The water uptake and swelling ratio of the side chain-containing sulfonated poly(aryl ether sulfone) (SPES) were shown to be lower than that of the sulfonated polymers lacking side chains and with similar ion-exchange capacities (IEC).¹⁷ Moreover, the sulfonic acid groups on the flexible aliphatic side chains of the SPES could decrease the limitation of main-chain rigidity on the mobility of sulfonic acid groups, thus enhancing proton conductivity.¹⁸ Therefore, in this work, the side chain-containing SPES samples with different degrees of sulfonation were prepared first, and then they were electrospun into nanofibers. The fabricated composite membrane based on Nafion-filled material was evaluated to identify the effects of these nanofibers, compared to those of a pristine membrane, on proton conductivity, dimensional stability, morphology and so forth.

2. Experimental

2.1. Materials

Phenolphthalein was purchased from the Beijing Chemical Reagent Company and purified by recrystallization from a mixed solvent of ethanol and water. 4,4-Difluorodiphenylsulfone was obtained from the Jiangyan Jiasheng Chemical Co., Ltd., China. 1,3-Propanesultone was obtained from Alfa Chemicals and used as received. Nafion 5 wt% solution and Nafion 117 membrane were obtained from DuPont Co. All other chemicals were reagent grade and used as received.

2.2. Synthesis of cardo poly(aryl ether sulfone)s with pendant sulfonic acid groups (SPES-X)

Preparation and characterization of cardo poly(arylene ether sulfone)s containing amide groups (PES-NH) were reported in a previous paper.¹⁹ Then the cardo poly(arylene ether sulfone)s with pendant sulfonated aliphatic side chains were synthesized as shown in Scheme 1. The polymers obtained are denoted by SPES-X, where X represents the degree of sulfonation.

Scheme 1 The typical synthetic route for cardo poly(arylene ether sulfone)s containing sulfonated side chains.

A typical procedure for the preparation of SPES-60 began with PES-NH (10.632 g, 0.02 mol) and NaH (0.96 g, 0.04 mol) being added to 100 mL of N-methyl-pyrrolidone at room temperature and stirred for 30 min. Then 1,3-propanesultone (1.4657 g, 0.012 mol) was added and the reaction was allowed to proceed at room temperature for 12 h. The viscous solution was precipitated into ethanol, and the isolated white polymer was washed several times with deionized water before being dried under vacuum at 120 °C overnight. The yield of this procedure was 97%. Other sulfonated polymers were prepared by the same method, except with different amounts of 1,3-propanesultone.

SPES-X: ¹H-NMR (deuterated dimethyl sulfoxide; DMSO-d₆): 9.80 ppm 1H; 7.93–7.96 ppm 4H; 7.57–7.84 4H; 7.32–7.41 ppm 4H; 7.17–7.24 ppm 8H; 3.43 ppm 2H; 2.26 ppm 2H; 1.23 ppm 2H.

2.3. Preparation of electrospun SPES-X nanofiber mats

Clear and homogeneous solutions of 25, 30, and 35 wt% SPES-X in dimethylacetamide (DMAc) were prepared and electrospun with electrical voltages of 15 kV. A syringe pump was used to feed the polymer solution into the needle tip at a fixed feed rate of 0.1 mL h⁻¹. The electrospun fibers were collected on an aluminum target at a distance of 15 cm from the spinneret. The mats were dried under vacuum to remove residual DMAc, and had specific weights of about 25 g m⁻². The porosity of the electrospun mats was about 60%.²⁰

2.4. Fabrication of the composite membranes containing nanofibers

Appropriate amounts of 5 wt% Nafion solution were poured on the SPES-X nanofiber porous mats, and the solution was dried at room temperature for 3 days. The resulting membranes were then dried under vacuum at 100 °C for 12 h. Finally, the composite membranes in their salt form were then converted to a corresponding acid form by soaking in 1 M H₂SO₄ solution for 48 h followed by immersion in deionized water for 48 h. Tough, ductile ionomer membranes were obtained with a controlled thickness of 40–60 μm.

2.5. Characterizations

¹H-NMR spectra were measured at 300 MHz on an AV 300 spectrometer (Bruker). The morphologies of the electrospun nanofibers and composite membranes were examined using a scanning electron microscope (SEM; Hitachi XE-100 SEM, Japan).

The proton conductivity (σ, S cm⁻¹) of each membrane coupon (size: 1 cm × 4 cm) was obtained using the equation σ = d/L_S × W_S × R (where, d is the distance between reference electrodes, and L_s and W_s are the thickness and width of the membrane, respectively). Ohmic resistance (R) was measured by four-point probe alternating current impedance spectroscopy using an electrode system connected to a Solatron 1260 impedance/gain-phase analyzer with a Solatron 1287 electrochemical interface (Solatron, Farnborough, Hampshire, UK). All samples were equilibrated in water for at least 24 h before the conductivity measurements. At a given temperature, the samples were equilibrated for at least 30 min before any measurements.

The membrane sample (30–40 mg per sheet) was dried at 120 °C under vacuum for 12 h until a constant weight of dry material was obtained. The sample was immersed into deionized water at 20 °C for 8 h and then quickly removed, wiped with tissue paper, and quickly weighed on a microbalance. The water uptake was calculated using the following equation: water uptake (wt%) = (W_s − W_d)/W_d, where W_d and W_s are the weights of the dry and wet membranes, respectively. The water swelling ratio of the membranes was investigated by immersing the round-shaped samples into water at 20 °C for 8 h and the swelling ratio was calculated from the following equation: swelling ratio (%) = [l_wet − l_dry]/l_dry, where l_dry and l_wet are the lengths of the dry and wet samples, respectively.

The methanol permeability was determined by using an HS test cell with a solution that contained (2 mol L⁻¹) methanol in water in one side and pure water in the other side. Magnetic stirrers were used in each compartment to ensure uniformity. The methanol concentration within the water cell was monitored by a GC-1020A series gas chromatograph (SHIMADZU). The methanol permeability was calculated by eqn (1):

In this equation, C_A and C_B are the methanol concentration in the feed side and the concentration that permeated through the membrane, respectively. A, L and V_A are the effective area, the thickness of membrane and the volume of the permeated compartment, respectively. DK is defined as the methanol permeability, and t₀ is the time lag.

Tensile measurements were performed with an Instron-1211 mechanical tester instrument (Instron Co., USA) at a speed of 2 mm min⁻¹ at ambient humidity [∼30% relative humidity (RH)].

3. Results and discussion

3.1. Synthesis and characterization of the polymers

The procedure used to synthesize cardo poly(arylene ether sulfone)s containing sulfonated aliphatic side chains is shown in Scheme 1. The as-synthesized polymers (PES-NH) contain an amide group in each repeating unit, which provides reaction sites to introduce various functional groups, such as sulfonated aliphatic side chains. The sulfoalkylation reaction of PES-NH and 1,3-propanesultone was carried out at room temperature for 12 h with NaH as the base.

Useful information can be extracted from the ¹H-NMR spectra of the SPES-X copolymers. Fig. 1 shows ¹H-NMR spectra of SPES-60, SPES-70 and SPES-80. The peak at 9.80 ppm (proton in amide), which is in the spectrum of the reactant, remains in that of the product, and new peaks appear at 3.43 ppm, 2.26 ppm and 1.23 ppm which correspond to the protons in the sulfoalkyl group and which confirms that some of the amide groups reacted with 1,3-propanesultone to produce N-alkylated product. The integration ratio of H(c) to H(1–8) was close to 2X : 20, suggesting a quantitative transfer of the proton in the amide (NH) to the sulfopropyl groups. The content of the sulfonic acid groups in the copolymers was readily controlled through the feed ratios.

Fig. 1 ¹H-NMR spectra of SPES-X copolymers in DMSO-d₆.

Table 1 shows the properties of the synthesized SPES-X membranes. IEC is a constant representing the amount of the exchangeable protons in ionomer membranes. The IEC values, which were determined by titration, were close to the theoretical values calculated according to the content of –NH groups, indicating a quantitative reaction between –NH groups and 1,3-propanesultone. The IEC values in the range of 0.90–1.51 meq g⁻¹ were readily controlled by adjusting the feed ratios.

Table 1 Properties of the cast SPES-X membranes

Samples	IEC (meq g^−1)		Water uptake (%)		Swelling ratio (%)		σ (S cm^−1)
	IEC^a	IEC^b	20 °C	80 °C	20 °C	80 °C	20 °C
a Calculated IEC values based on complete reaction of –NH groups with 1,3-propanesultone.b Determined from titration.
SPES-100	1.53	1.51	20.0	30.0	7.5	9.8	0.0560
SPES-80	1.27	1.18	14.4	22.3	6.2	8.0	0.0456
SPES-70	1.13	1.08	12.6	20.1	4.8	5.6	0.0392
SPES-60	0.99	0.90	7.3	12.2	4.4	5.0	0.0126

---

As shown in Table 1, the water uptake increased from 7.3% for SPES-60 to 20.0% for SPES-100. Also, the swelling ratio increased from 4.4% for SPES-60 to 7.5% for SPES-100. With the increasing IEC, the proton conductivity was also enhanced from 0.0126 to 0.0560 S cm⁻¹ for SPES-60 and SPES-100, respectively. The increased proton conductivity with IEC implies that the water uptake values were reasonable in enhancing the proton transfer through the fully hydrated membrane.

The homopolymer SPES-100 can achieve suitable conductivities only at high IEC, resulting in high water uptake and large changes in the dimensions of the membrane, which are unsuitable for practical PEM applications. Therefore, it is more suitable to select copolymer-electrospun mats to reinforce the Nafion matrix.

3.2. Electrospun fiber and membrane morphology

Three types of copolymer, SPES-60, SPES-70 and SPES-80, were electrospun into nanofibers. The SEM micrographs of electrospun fibers from SPES-70 in DMAc solution with different concentrations, ranging from 25% to about 35%, are shown in Fig. 2. At concentrations below 25 wt%, the electrospinning process generated a mixture of fibers and droplets. In contrast, the electrospinning fiber population exhibits a wide distribution of diameter sizes, between 300 and 500 nm, at concentrations above 35 wt%. Continuous fibers without beads are obtained at the polymer concentration of 30 wt% (Fig. 2b) and the widths of these fibers are narrowly distributed, with diameters between 120 and 150 nm. This observation about the effect of concentration is consistent with those found in the literature.²¹ The fiber mat is highly porous and cannot be used as a fuel cell membrane, so further processing of the mat is required to create a dense and defect-free film. Moreover, SPES-60 and SPES-80 were also electrospun in to nanofibrous mats at the optimized conditions (e.g., 15 kV, 30 wt%, 0.1 mL h⁻¹ and 15 cm).

Fig. 2 SEM images of the SPES-70 electrospun nanofibers with different concentrations: (a) 25%, (b) 30%, (c) 35%.

Fig. 3a shows the cross-sectional SEM images of the SPES-70 electrospun nanofiber mats, which revealed high porosity and an interconnected open pore structure. The SPES-70 nanofibers are clearly visible and uniform throughout the membrane, which was full of pores (Fig. 3b and c). Absence of fiber pull-outs or cracks indicate good interfacial surface compatibility between SPES-70 nanofibers and the Nafion matrix (Fig. 3c). Fig. 3d and e show the morphology of the composite membrane containing 5% Nafion, where there is a uniform distribution of SPES-70 nanofibers embedded in Nafion. However, image analysis of these nanofibers (Fig. 3e) revealed that they had slightly dissolved because of a strong interaction between the nanofiber and Nafion. The nanofiber structures are clearly seen in the micrographs of CSPES-60-3 composite samples (Fig. 3f). Nevertheless, cracks appear in the composites, indicating poor compatibility between the Nafion matrix and the SPES-60 nanofiber. Furthermore, the preparation of the CSPES-60-5 composite was not successful in this work. Fig. 3g confirmed designed composite structures, but there are a lot of voids. Because of the above results, the remainder of the discussion is limited to SPES-70 and related materials.

Fig. 3 Cross-sections of SEM images of the prepared samples for (a) SPES-70 nanofiber mat, (b) and (c) the CSPES-70-3 composite membrane ((b) for low magnification and (c) for high magnification), (d) and (e) the CSPES-70-5 composite membrane, (f) the CSPES-60-3 composite membrane, and (g) the CSPES-80-3 composite membrane. The composite membrane reinforced to Nafion by SPES-X nanofibers, denoted as CSPES-60, CSPES-70, and CSPES-80, respectively. CSPES-70-3 and CSPES-70-5 represented 3 wt% and 5 wt% Nafion solution, respectively.

As shown in Fig. 4, the swelling ratios of composite membranes were greatly reduced from 15.2% for Nafion 117 to 5.1% for CSPES-70-5 and 4.5% for CSPES-70-3. Meanwhile, the water uptake of composite membranes also showed a similar trend. The three-dimensional network structures formed by nanofibers within the composite membranes place a limit on the swelling ratio.²² These results imply that the Nafion matrix were efficiently reinforced by nanofibers, and the electrospun nanoporous mat was suitable for use as a supporting material in a pore-filling membrane.

Fig. 4 Water uptake and swelling ratio of the membranes at 20 °C.

The proton conductivity of the membranes plays a particularly important role in fuel cell performance: higher levels of ion conductivity lead to higher power density. Proton conductivities of the fabricated membranes were subsequently examined at 20 °C under 100% RH conditions. Fig. 5 shows the dependence of membrane proton conductivity on temperature over the range 20–80 °C. The conductivity of all membranes steadily increases with temperature because the free volume, which favors ion transport and the mobility of cations, is increased as the temperature rises. The composite membranes contain nanofibers displaying the higher proton conductivity at all measured temperatures compared to that of neat SPES-70. The proton conductivity of CSPES-70-5 is lower than that of CSPES-70-3, which was associated with slightly dissolved nanofibers as shown in SEM micrographs (Fig. 3e). The observed increase in conductivity is mainly attributed to Nafion, which contains several sulfonic groups. On the other hand, this conductivity increase is also due to the nanofibers. The proton channel structure in the network of sulfonic acid groups formed within the nanofiber may lead to the rapid transport of the proton.²³

Fig. 5 Impact of temperature on proton conductivity.

One of the most significant limitations to the use of Nafion membranes in DMFC is methanol crossover, which is detrimental to the performance of the fuel cell.²⁴ Table 2 lists the methanol transport behavior of the CSPES-70-3 membranes in comparison to that of Nafion 117 and of SPES-70 cast film.

Table 2 Methanol permeability and selectivity of membranes

Samples	σ (S cm^−1) 20 °C	PM (×10^−7 cm^2 s^−1) 20 °C	Selectivity (×10^5 S s cm^−3)
SPES-70	0.0392	4.98	0.79
CSPES-70-3	0.0624	1.02	6.12
Nafion 117	0.0900	24.0	0.38

---

The methanol permeability value of 1.02 × 10⁻⁷ cm² s⁻¹ for CSPES-70-3 is much less than that for the Nafion 117 measured under the same experimental conditions (2.4 × 10⁻⁶ cm² s⁻¹). The SPES-70 membrane exhibited methanol permeability of 4.98 × 10⁻⁷ cm² s⁻¹, which is also higher than that displayed by CSPES-70-3. The methanol permeability reduction occurs because of the long and tortuous diffusive pathway in the presence of the CSPES-70-3 nanofibers.¹⁶

The possibility of using a membrane in DMFC is often evaluated using the selectivity parameter, which is the ratio of proton conductivity to methanol permeability. The selectivity value of CSPES-70-3 is about 6.12 × 10⁵ S s cm⁻³ compared to 7.9 × 10⁴ S s cm⁻³ and 3.8 × 10⁴ S s cm⁻³ for SPES-70 and Nafion 117, respectively. An improvement in the composite membrane selectivity has been achieved because of the incorporation of the three-dimensional network of nanofibers into Nafion.

The actual stress–strain curves for the two composite membranes and for Nafion 117 are shown in Fig. 6. Mechanical properties are summarized in Table 3. As can be seen, the neat SPES-70 membrane has excellent mechanical properties, with a Young's modulus of 1490.4 MPa, a tensile strength of 72.9 MPa and an elongation at break of 38.6%. For the nanofiber composite membranes, it is observed that the Young's modulus and tensile strength of CSPES-70-3 and CSPES-70-5 are much higher than for Nafion 117. This is attributed to SPES-70 in the fiber form acting as a reinforcing non-woven network. The interfacial adhesion of the nanofibers to the 5% Nafion matrix is stronger than to the 3% Nafion matrix (as indicated by SEM).

Fig. 6 Stress–strain curves for Nafion 117 and composite membranes.

Table 3 Mechanical properties of membranes

Samples	Tensile strength (MPa)	Young's modulus (MPa)	Elongation at break (%)
SPES-70	72.9	1490.4	38.6
CSPES-70-3	25.5	362.2	38.7
CSPES-70-5	33.5	634.2	36.9
Nafion 117	18.3	164.0	190.0

---

4 Conclusions

We have developed a Nafion composite membrane reinforced by electrospun side-chain cardo-SPES nanofibers. The morphological evaluations revealed that the nanofibrous mat was successfully filled by Nafion. The swelling ratio of the composite membranes was reduced because of the incorporation of the three-dimensional network of SPES nanofibers into the Nafion matrices.

Because of the properties of the polyelectrolyte nanofibers, the composite membranes obtained exhibited enhanced proton conductivity compared to that displayed by the neat SPES-70 films. Methanol permeability results showed that the incorporation of SPES nanofibers into the Nafion matrix can act as a methanol barrier. Moreover, the nanofiber composite membranes exhibited better mechanical properties. Such a membrane may prove to be promising as a polymer electrolyte membrane and may be potentially useful for application in fuel cells.

Acknowledgements

We thank the National Basic Research Program of China (no. 2012CB932802), the National Natural Science Foundation of China (no. 51133008 and 21304092), the National High Technology Research and Development Program of China (no. 2012AA03A601), and the Program of Jilin Education Department (no. 93201201) for the financial support.

Notes and references

1. A. S. Arico, S. Srinivasan and V. Antonucci, Fuel Cells, 2001, 1, 133–161.
2. N. W. Deluca and Y. A. Elabd, J. Polym. Sci., Part B: Polym. Phys., 2006, 44, 2201–2225.
3. K. Rikukawa and K. Sanui, Prog. Polym. Sci., 2000, 25, 1463–1502.
4. Y. Yang and S. Holdcroft, Fuel Cells, 2005, 5, 171–186.
5. V. D. Noto, N. Boaretto, E. Negro and G. Pace, J. Power Sources, 2010, 195, 7734–7742.
6. E. I. Santiago, R. A. Isidoro, M. A. Dresch, B. R. Matos, M. Linardi and F. C. Fonseca, Electrochim. Acta, 2009, 54, 4111–4117.
7. K. Schmidt-rohr and Q. Chen, Nat. Mater., 2008, 7, 75–83.
8. T. Li, G. Zhong, R. Fu and Y. Yang, J. Membr. Sci., 2010, 354, 189–197.
9. S. V. Fridrikh, J. H. Yu, M. P. Brenner and G. C. Rutledge, Phys. Rev. Lett., 2003, 90, 144502–144505.
10. Y.-J. Kima, et al., Int. J. Biol. Macromol., 2009, 45, 65–71.
11. S. Sh. Homaeigohar, K. Buhr and K. Ebert, J. Membr. Sci., 2010, 365, 68–77.
12. H.-Y. Li and Y.-L. Liu, J. Mater. Chem. A, 2013, 1, 1171–1178.
13. S. W. Choi, Y. Z. Fu, Y. R. Ahn, S. M. Jo and A. Manthiram, J. Power Sources, 2008, 180, 167–171.
14. J. B. Ballengee and P. N. Pintauro, Macromolecules, 2011, 44, 7307–7314.
15. I. Shabani, M. M. H. Sadrabadia and V. H. Asla, J. Membr. Sci., 2011, 368, 233–240.
16. M. M. H. Sadrabadia, I. Shabani, M. Soleimanid and H. Moaddele, J. Power Sources, 2011, 196, 4599–4603.
17. P. X. Xing, G. P. Robertson, M. D. Guiver, S. D. Mikhailenko and S. Kaliaguine, Macromolecules, 2004, 37, 7960–7967.
18. Q. Zhang, F. Gong, S. Zhang and S. Li, J. Membr. Sci., 2011, 367, 166–173.
19. N. Gao, F. Zhang and S. B. Zhang, J. Membr. Sci., 2011, 372, 49–56.
20. Z. Ma, M. Kotaki and T. Yong, et al., Biomaterials, 2005, 26, 2527–2536.
21. K. Lin, K. N. Chua, G. T. Christopherson, S. Lim and H. Q. Mao, Polymer, 2007, 48, 6384–6394.
22. S. H. Yun, J. J. Woo and S. J. Seo, et al., J. Membr. Sci., 2011, 367, 296–305.
23. T. Tamura and H. Kawakami, Nano Lett., 2010, 10, 1324–1328.
24. V. Tricoli, J. Electrochem. Soc., 2000, 147, 1286–1290.

---