Sulfonated poly(arylene thioether phosphine oxide)s (sPTPO) and sPTPO/sulfonated polybenzothiazole blends as proton exchange membranes

Abstract

Sulfonated poly(arylene thioether phosphine oxide)s (sPTPO) were prepared, which showed high swelling at elevated temperatures. To reduce the swelling, sPTPO were blended with sulfonated polybenzothiazoles (sPBT) to prepare composite proton exchange membranes (PEMs) because the sulfonic acid groups of sPTPO and the benzothiazole moieties of sPBT can form strong acid–base interactions. Herein, sPBT were for the first time employed as a blend component to fabricate blend PEMs. The resulting blend membranes exhibited low swelling and remarkably enhanced overall properties compared to the raw sPTPO membranes. For example, one of them exhibited a swelling of only 30% even at 100 °C, a proton conductivity of 0.085 S cm⁻¹ at 80 °C (higher than that of Nafion 117), and a methanol permeability of 1.02 × 10⁻⁷ cm² s⁻¹. Moreover, it also displayed excellent thermal stability, oxidative stability and mechanical properties. Therefore, sPBT are an ideal blend component to improve the resistance to swelling and the other properties of PEMs.

---

Introduction

Polymer electrolyte membrane fuel cells (PEMFC) have attracted considerable attention as a green source owing to the advantages such as high efficiency, fast start-up, and low operating temperature.¹ Proton exchange membranes (PEMs) are a core component of PEMFCs, which possess the functions of conducting proton, supporting catalyst, and separating the fuel and oxygen gas. During the operation of a fuel cell, PEMs are subjected to harsh environments, so that they should meet special material requirements.² Perfluorinated Nafion is still the most widely used PEM, but it has some drawbacks such as a humidity/temperature-sensitive proton conductivity and high methanol permeability.² Therefore, many aromatic membranes were developed as alternative PEMs.^2–11

Our group developed a series of sulfonated poly(arylene ether/thioether phosphine oxide)s for proton exchange membranes.^12–21 The phosphine oxide moieties facilitated them to exhibit excellent water-binding capacity, adhesive ability, and high resistance to oxidation, thus they were suitable for preparing composite PEMs.^12,17 Similar to many aromatic PEMs, some of them exhibited high swelling at elevated temperatures,^15,21 decreasing their mechanical properties. Many methods were adopted to depress their swelling while retaining high proton conductivity, including physical blending, cross-linking, and an adjustment of the microscopic structure.^14–16 Among them, the physical blend was used extensively to restrain the swelling of PEMs. In most cases of physical blending, a basic polymer was used as a component to blend with sulfonated polymers to decrease the swelling.^22,23 These basic groups could form acid–base interactions with sulfonated polymers; thus, a part of their sulfonic acid groups were “consumed” and unavailable for the transport of protons.²⁴ Furthermore, these basic polymers themselves generally exhibit no or very low proton conductivity.^23,24 Therefore, the proton conductivity of the corresponding blend membranes reduced obviously. On the other hand, the abovementioned acid–base interactions often greatly weakened at high temperature, so that the swelling of membranes could not be reduced effectively.^22,23

Our group developed sulfonated polybenzothiazoles (sPBT) PEMs several years ago.^25–28 Their swelling was very low and displayed almost no increase with temperature in the range of 80–100 °C, but still possessed high proton conductivity.^25,26 Their outstanding dimensional stability (i.e. very low swelling) was because the benzothiazole and sulfonic acid groups formed strong acid–base interactions.²⁵ Therefore, sPBT could be used as a blend component to depress the swelling of sulfonated membranes. At the same time, sPBT themselves presented high proton conductivity and thus also contributed to the high proton conductivity of blend membranes, which was quite different from the previous basic blend components. To date, there is no report on the blend membranes of sPBT and sulfonated polymer for PEMs.

In this study, sulfonated poly(arylene thioether phosphine oxide)s (sPTPO) were initially prepared by the polymerization of 1,4-benzenedithiol with bis(4-fluorodiphenyl)-3′-sulfonate phenyl phosphine oxide and bis(4-fluorophenyl)phenyl phosphine oxide. These sPTPO membranes exhibited high swelling at elevated temperatures. Therefore, they were blended with sPBT to fabricate the blend membranes. The strong acid–base interactions between the sulfonic acid groups of sPTPO and the benzothiazole moieties of sPBT were expected to reduce the swelling of the blend membranes and improve their overall properties. The properties of the sPTPO and blend membranes as PEMs were explored in detail.

Experimental

Materials

1,4-Benzenedithiol (BZDT) and p-toluenethiol were purchased from Aldrich and used as received. 2,5-Diamino-1,4-benzenedithiol dihydrochloride (DABDT) was bought from TCI and used without further purification. Toluene was distilled prior to use. Tetrahydrofuran was dried with wirelike sodium followed by distillation before use. N-Methyl-2-pyrrolidone (NMP) was purified by vacuum distillation. Polyphosphoric acid (PPA, 81%) and other reagents were obtained from commercial sources and used as received. Bis(4-fluorophenyl)phenyl phosphine oxide (BFPPO), bis(4-fluorophenyl)-3′-sulfonate phenyl phosphine oxide (SBFPPO), and sulfonated polybenzothiazoles (sPBT) with a sulfonation degree of 110% were synthesized according to our previous studies.^22,27 The structure of sPBT is shown in Scheme 1(a).

Scheme 1 (a) Chemical structure of sPBT; (b) synthesis of sPTPO.

Measurements

Fourier transform infrared (FT-IR) spectra were obtained on a Bruker Equinox-55 spectrometer. The nuclear magnetic resonance (NMR) spectra were obtained on a MERCURYplus 400 spectrometer using tetramethylsilane as an internal standard. The molecular weight and polydispersity index (PDI) were measured by gel permeation chromatography (GPC, Series 200), whereas polystyrene standards were used for calibration. The thermal stability was investigated on a Q5000IR thermogravimetric analyzer in a nitrogen atmosphere at a heating rate of 20 °C min⁻¹. Differential scanning calorimetry (DSC) was performed on a Pyris 1 apparatus. The acid form samples for thermal analysis were preheated at 150 °C for 0.5 h to eliminate moisture. The DSC curves were recorded over the range of 100–320 °C at 20 °C min⁻¹.

The theoretical ion exchange capacity (IEC) of sPTPO was calculated by molecular formulae. Similarly, the theoretical IEC of the blend membranes could be obtained by this formula as follows:

where IEC_sPTPO and IEC_sPBT are the IEC of sPTPO and sPBT and W_sPTPO and W_sPBT are the weight percentage of sPTPO and sPBT, respectively. In addition, the experimental IEC value was determined by titration. The operation processes were as follows. First, the membrane sample was immersed in the saturated NaCl solution for two days to release H⁺ entirely. Furthermore, the released H⁺ was titrated with 0.01 mol L⁻¹ NaOH solution. The experimental IEC value could be deduced by the amount of NaOH consumed.

The oxidative stability of membranes was investigated following a typical procedure.¹² The membranes (0.5 cm × 1 cm × ∼40 μm) were placed in Fenton's reagent (3% H₂O₂, 2 ppm FeSO₄) at 80 °C; their oxidative stability was then characterized by the consumed time when they began to dissolve (τ₁) and fully dissolved (τ₂).

The mechanical properties were tested on an electronic tensimeter (Instron 4456) at a crosshead speed of 1 mm min⁻¹. The samples were maintained at ∼25 °C and 50% relative humidity for one day before measurement following a typical procedure.²⁷

Small angle X-ray scattering (SAXS) was measured on a SAXSess mc2 apparatus, using Cu Kα radiation. The Pb²⁺ form membranes were employed instead of the acid from membranes to improve the scattering intensity.¹² They were prepared by soaking the acid form membranes in 0.5 mol L⁻¹ lead acetate for one day, followed by drying in vacuo. The scattering vector (q) obeys the formula of q = 4π/λ × sin θ, in which λ represents the wavelength and 2θ means the scattering angle. The characteristic periodicity (d) between the hydrophilic and hydrophobic domains was calculated by the equation: d = 2π/q.¹² Scanning transmission electron microscopy (STEM) was carried out with a JEM-2100F microscope at an accelerating voltage of 200 kV. The STEM samples were ultrathin sections. Their preparation procedure was as follows. A strip (1 × 3 mm²) was cut from the lead-stained membrane and then embedded to an epoxy resin. After curing, the strip sample coated with epoxy resin was sectioned to 90–120 nm thickness films using an ultramicrotome.¹³

The water uptake, swelling, proton conductivity, and methanol permeability were measured according to the previous procedures.^12,27 The activation energy (E_a) of proton conduction was calculated using the equation as follows:¹⁴

ln σ = −E_a/RT

where σ is the proton conductivity, R is the universal gas-constant (8.314 J mol⁻¹ K⁻¹), and T represents the absolute temperature.

Synthesis of sulfonated poly(arylene thioether phosphine oxide)s

Typically, the synthesis of sPTPO-60 was depicted as follows. 0.5690 g (4 mmol) of BZDT, 0.9991 g (2.4 mmol) of SBFPPO, 0.5028 g (1.6 mmol) of BFPPO, 0.6081 g (4.4 mmol) of K₂CO₃, 8 mL of toluene and 8 mL of NMP were put to a 150 mL three-necked flask, which was assembled with a condenser, Dean–Stark trap, argon inlet/outlet, and stirrer. The reaction mixture was stirred under argon for 0.5 h to dissolve the reactants, and then heated to 150 °C for approximately 4 h to remove water in the mixture. Subsequently, the toluene was eliminated and the reaction system was then heated to 195–200 °C. The reaction system was maintained at this temperature for several days until it became quite viscous. After cooling to 110 °C, 2 mL of NMP was poured to the reaction flask to dilute the mixture, and then it was placed into de-ionized water with stirring. The fibrous polymer was immersed in hot de-ionized water to eliminate the inorganic salts. The product was dried in vacuo at 110 °C for about 36 h.

Yield: 94%. ¹³C NMR (DMSO-d₆, ppm): 148.895 (d, J_CP = 11.6 Hz), 141.469, 134.157, 133.005, 132.711 (d, J_CP = 109 Hz), 133.182 (d, J_CP = 9.6 Hz), 132.703, 132.427 (d, J_CP = 116 Hz), 131.941 (d, J_CP = 10 Hz), 130.851 (d, J_CP = 104 Hz), 130.605 (d, J_CP = 104 Hz), 129.782, 129.189 (d, J_CP = 10.3 Hz), 129.085 (d, J_CP = 10.4 Hz), 129.209 (d, J_CP = 13.6 Hz), 128.768 (d, J_CP = 10.3 Hz). FT-IR (film, cm⁻¹): 1190 (P=O), 1228, 1035, 619 (–SO₃Na).

sPTPO-100 and poly(arylene thioether phosphine oxide)s (PTPO) were also prepared for assisting to confirm the chemical structure of the copolymer sPTPO. Herein, sPTPO-100 was synthesized by polycondensation of SBFPPO and BZDT, whereas PTPO was prepared by the polymerization of BFPPO and BZDT according to the abovementioned preparation process.

Preparation of membranes

The salt-form polymer of sulfonated polybenzothiazoles was initially dissolved in 4 mL NMP, and sPTPO was added to this solution. After the mixture formed a homogenous solution, it was cast to the glass plate (6 cm × 6 cm), followed by drying at 65 °C for 24 h. The blend membrane was separated from the plate by soaking in de-ionized water. The membrane was acidified in 0.5 mol L⁻¹ sulfuric acid at ∼25 °C for 24 h and at 90 °C for 2 h, and then immersed in de-ionized water at ∼25 °C for one day to get rid of impurity.¹² The acid-form blend membrane was dried in vacuo at 110 °C for 24 h.

Results and discussion

Synthesis and characterization of sulfonated polymers

Sulfonated poly(arylene thioether phosphine oxide)s (sPTPO) were prepared by the polymerization of 1,4-benzenedithiol (BZDT), bis(4-fluorophenyl)-3′-sulfonate phenyl phosphine oxide (SBFPPO) and bis(4-fluorophenyl)phenyl phosphine oxide (BFPPO), as indicated in Scheme 1(b). The resulting polymers were denoted as sPTPO-x, where “x” represents the molar percentage of SBFPPO in the feed mixture of SBFPPO and BFPPO. As shown in Table 1, the sulfonated polymers presented a number average molecular weight (M_n) higher than 3 × 10⁴ g mol⁻¹, and their polydispersity index (PDI) is between 1.5 and 1.9. Furthermore, they could all form flexible, transparent and tough membrane by casting. Therefore, they possessed a high molecular weight.

Table 1 Reactant ratios and results of polycondensation

Polymer	SBFPPO/BFPPO/BZDT (molar ratio)	GPC^a			Yield (%)
		Mn × 10^−4	Mw × 10^−4	PDI
a Measured in DMF solution.
sPTPO-60	60/40/100	3.48	6.79	1.89	96
sPTPO-65	65/35/100	4.25	6.48	1.52	95
sPTPO-100	100/0/100	4.77	8.86	1.85	96

---

sPTPO-100 was utilized as a “model” polymer to assign the NMR spectra of sPTPO-x. Moreover, poly(arylene thioether phosphine oxide)s (PTPO) were also synthesized to assist the analysis of the FT-IR spectra of sPTPO polymers.

sPTPO-60 was chosen as an example to analyze the chemical structure of sPTPO-x. To analyze the ¹³C NMR spectra of sPTPO-60, the model compound (TTPPO) was also designed and synthesized. Its preparation is described in the ESI† and its chemical structure is displayed in Scheme S1 (ESI†). The dept135 and ¹³C NMR spectra of the model compound, sPTPO-100 and sPTPO-60 are exhibited in Fig. S1 (ESI†). The signal peaks in the dept135 and ¹³C NMR spectra of sPTPO-60 were well assigned with the assistance of the dept135 and ¹³C NMR spectra of the model compound and sPTPO-100. The assignments of the signal peaks in their NMR spectra are indicated in the Fig. S1 (ESI†). Based on the NMR analysis in the ESI,† the chemical structure of sPTPO-60 was confirmed.

The FT-IR spectra of the resulting polymers are exhibited in Fig. 1. They all showed a strong band at ∼1190 cm⁻¹, ascribed to the stretching vibration of the P=O groups.^15,21 Compared to PTPO, the sulfonated polymers (sPTPO-x) still displayed the absorption peaks at 1035 and 619 cm⁻¹ and a shoulder peak at 1228 cm⁻¹, which were the characteristic bands of the –SO₃Na groups.²⁹ These characteristic bands verified that the phosphine oxide and sodium sulfonate groups were incorporated into the molecular chain of sPTPO by a polycondensation reaction.

Fig. 1 FT-IR spectra of sPTPO-x and PTPO.

Preparation of the blend membranes

As mentioned in the Introduction, sulfonated polybenzothiazoles membranes exhibited outstanding resistance to swelling at high temperature because of the strong acid–base interactions between the sulfonic acid and benzothiazole groups.^25,26 When sPTPO were blended with sulfonated polybenzothiazoles (sPBT), the sulfonic acid groups of sPTPO could form strong acid–base interactions with the benzothiazole groups of sulfonated polybenzothiazoles. As a result, these strong interactions could facilitate the sPTPO/sPBT blend membranes to show low swelling. The effect of the amount of sPBT and the sulfonation degree of sPTPO on the properties of blend membranes was investigated in the following text. The abbreviations of the blend membranes were sPTPOx/sPBTy, where “x” is the percent sulfonation degree of sPTPO and “y” is the weight percentage of the sPBT component. There are two series of blend membranes, which are compiled in Table 2. They all were flexible and tough membranes.

Table 2 sPBT content, IEC, T_d5 and T_g of the membranes

Polymer	sPBT (wt%)	IEC (meq g^−1)		Td5^a (°C)	Tg (°C)
		Calculated	Measured
a The 5% weight-loss temperature.
sPTPO-60	0	1.29	1.24	379	243
sPTPO60/sPBT6.25	6.25	1.34	1.29	384	255
sPTPO60/sPBT6.5	6.5	1.34	1.28	393	260
sPTPO-65	0	1.39	1.33	377	258
sPTPO65/sPBT10.75	10.75	1.46	1.41	384	275
sPTPO65/sPBT11	11	1.46	1.39	386	278

---

Thermal properties of membranes

The TGA traces of the raw and blend membranes are presented in Fig. 2a and b. As indicated, both series of membranes displayed a two-step degradation process. The first step weight-loss at ∼400 °C was because of the degradation of –SO₃H groups, and the second step weight-loss at about 500 °C was associated with decomposition of the polymer backbone. This decomposition process was the typical degradation mode of sulfonated polymers.^15,21 It could also be observed that for each series of membranes, the TGA curves of blend membranes were on the “top” of the one of the corresponding raw membranes and the TGA curve of the blend membrane with a higher content of sPBT was on the “top” of the one of the blend membrane with lower content of sPBT. These results illustrated that the blend membranes exhibited higher thermal stability than the corresponding raw membrane and that the blend membrane with higher content of sPBT showed better thermal stability. The results were attributed to the acid–base interactions between the benzothiazole and sulfonic acid groups.²⁵ In Table 2, the 5% weight-loss temperature (T_d5) of the blend membranes was higher than that of the corresponding raw membrane. Moreover, the T_d5 of two series of membranes was higher than 370 °C, higher than that of many sulfonated PEMs with equivalent ion exchange capacity.³⁰

Fig. 2 TGA traces of the sPTPO-60 (a) and sPTPO-65 (b) series of membranes.

The DSC curves of sPTPO and the blend membranes are exhibited in Fig. 3. The glass transition of all these membranes was obvious. Their glass transition temperature is listed in Table 2. It could be observed that sPTPO-60 and -65 exhibited a glass transition temperature (T_g) of 243 and 258 °C, respectively. For each series of membranes, the glass transition temperature of the blend membranes was higher than that of the raw membrane. Moreover, the glass transition temperature of each series of blend membranes increased with the increase of sPBT content. These results showed that the sulfonic acid groups of sPTPO and the benzothiazole moieties of sPBT formed the acid–base interactions, thus restraining the segment movement of sPTPO and enhancing the glass transition temperature. These results also implied that sPTPO and sPBT were compatible. In addition, it could be observed from Table 2 that the T_g of all the blend membranes was higher than 250 °C. The TGA and DSC results suggested that the blend membranes possessed excellent thermal stability.

Fig. 3 DSC curves of the raw sPTPO and blend membranes.

Ion exchange capacity, water uptake and swelling of membranes

The ion exchange capacity (IEC) is an important parameter for proton exchange membranes. The theoretical IEC of the raw and blend membranes could be calculated by their molecular formulae and blend ratio, and the experimental IEC was acquired by titration.^12,27 The IEC of membranes are collected in Table 2. As indicated, the IEC values of the membranes ranged from 1.29 to 1.46 meq g⁻¹, higher than that (0.909 meq g⁻¹) of Nafion 117. The IEC of sPTPO was controlled by the feed ratio to restrict the excessive swelling. Otherwise, the swelling of blend membranes could not be maintained in an acceptable range even if large amount of sPBT was introduced to the blend membranes.

The water uptake and swelling of PEMs markedly affected their proton conductivity and mechanical properties.⁷ A certain amount of water uptake was needed to promote the transport of protons, thus attaining high proton conductivity. However, excessive water uptake led to high swelling as well as large loss of mechanical strength.⁷ Therefore, water uptake should be maintained in a reasonable range. The water uptake of the membranes as a function of temperature is presented in Fig. 4. As expected, the water uptake of all the membranes augmented with increasing temperature and IEC. sPTPO-60 and -65 denoted a rapidly enhanced water uptake when the temperature exceeded 80 °C. On the contrary, the blend membranes displayed much lower water uptake than the corresponding raw membrane. Furthermore, the water uptake of each series of blend membranes depressed obviously with increasing sPBT content within the blend membranes. These results implied that the component of sPBT decreased the water uptake of the blend membranes. The blend membranes presented a water uptake of 34–48% at 80 °C, which was close to that of many other aromatic PEMs.^17,26

Fig. 4 Water uptake of the membranes at various temperatures.

The in-plane swelling of membranes is shown in Fig. 5a. The outline of the in-plane swelling curves was analogous to that of the water-uptake curves. sPTPO-60 and -65 showed high swelling when the temperature was higher than 80 °C. This high in-plane swelling caused them to show low dimensional stability as well as low mechanical properties at elevated temperatures. However, the in-plane swelling of two series of blend membranes increased slowly. For example, sPTPO60/sPBT6.25 and sPTPO60/sPBT6.5 displayed an in-plane swelling of 30% and 27%, whereas sPTPO65/sPBT10.75 and sPTPO65/sPBT11 exhibited an in-plane swelling of 32% and 28%, respectively, even at 100 °C. Their in-plane swelling approached that (23%) of Nafion 117, and was much lower than that of aromatic PEMs with equivalent proton conductivity.³¹

Fig. 5 In-plane (a) and through-plane (b) swelling of the membranes at various temperatures.

The through-plane swelling of membranes is displayed in Fig. 5b. The outline of the through-plane swelling was roughly similar to that of the in-plane swelling. The through-plane swelling of the blend membranes was slightly lower than their in-plane swelling, but the through-plane swelling of sPTPO-60 and -65 was much lower than their in-plane swelling at elevated temperatures (>80 °C).

In short, these blend membranes displayed excellent dimensional stability. This was because the strong acid–base interactions between the benzothiazole and sulfonic acid groups reduced the swelling, as revealed by DSC.

Proton conductivity, methanol permeability and selectivity

The proton conductivity of membranes is very important for determining their performance in fuel cells. The proton conductivity of the sPTPO and blend membranes as a function of temperature is exhibited in Fig. 6. sPTPO-60 and -65 showed large swelling at temperatures higher than 80 °C; therefore, their proton conductivity at 90 °C was not obtained. As expected, the proton conductivity of all the membranes increased with temperature. For each series of membranes, the proton conductivity decreased with the content of sPBT. A small part of the sulfonic acid groups formed the acid–base interactions with the benzothiazole groups, thus they had no contribution to proton transport.²⁴ These acid–base interactions made the blend membranes with more sPBT content show lower proton conductivity than those with lower sPBT content. However, sPBT themselves showed high proton conductivity.²⁷ Consequently, the blend membranes still showed high proton conductivity. For example, sPTPO60/sPBT6.25 and sPTPO65/sPBT10.75 displayed a proton conductivity 0.085 and 0.087 S cm⁻¹ at 80 °C, respectively, which is higher than that (0.082 S cm⁻¹) of Nafion 117. As mentioned above, both sPTPO60/sPBT6.25 and sPTPO65/sPBT10.75 displayed low swelling. Therefore, these blend membranes not only displayed excellent dimensional stability, but also retained high proton conductivity.

Fig. 6 Proton conductivity of the membranes at various temperatures.

The activation energy of proton conduction for sPTPO-60 and -65 membranes was 15.7 and 14.0 kJ mol⁻¹, respectively, which was close to that of other PEMs.¹⁴ In contrast, the E_a of proton conductivity for the blend membranes was in the range of 16.3–24.4 kJ mol⁻¹. These blend membranes showed higher activation energy than the corresponding raw membranes probably because the former possessed lower water uptake and thus their proton conduction became difficult.

The methanol permeability of PEMs is a crucial property for their applications in direct methanol fuel cells. The methanol permeability of two series of membranes is compiled in Table 3. sPTPO-60 and -65 showed a methanol permeability of 7.46 × 10⁻⁷ and 10.6 × 10⁻⁷ cm² s⁻¹, respectively. However, the methanol permeability of the blend membranes was much lower than the corresponding raw membranes. For instance, sPTPO60/sPBT6.25 and sPTPO65/sPBT10.75 displayed a methanol permeability of 1.02 × 10⁻⁷ and 3.01 × 10⁻⁷ cm² s⁻¹, which is only about 1/15 and 1/5 of that (15 × 10⁻⁷ cm² s⁻¹) of Nafion 117, respectively. Moreover, for each series of membranes, the methanol permeability decreased with increasing sPBT content. As a result, the sPBT component of the blend membranes reduced the methanol permeability sharply, which was attributed to the acid–base interactions between the benzothiazole and sulfonic acid groups.

Table 3 Oxidative stability, mechanical properties, methanol permeability, and selectivity of membranes

Membrane	Oxidative stability		Mechanical properties^b			Methanol permeability (×10^−7 cm^2 s^−1)	Selectivity (10^4 S s cm^−3)
	τ1^a (h)	τ2^a (h)	Tensile strength (MPa)	Young's modulus (GPa)	Elongation at break (%)
a τ1 and τ2 are the time when the samples began to dissolve and fully dissolved, respectively.b Equilibrated at 50% RH and 25 °C for 24 h prior to measurement.
sPTPO-60	1.5	6.5	24.8	0.889	23	7.46	4.80
sPTPO60/sPBT6.25	2.5	18	37.2	1.32	10	1.02	31.5
sPTPO60/sPBT6.5	3.0	19	39.1	1.34	9.9	0.943	30.6
sPTPO-65	0.5	3.5	24.9	0.613	29	10.6	3.88
sPTPO65/sPBT10.75	1.5	10	35.5	0.861	9.6	3.01	11.5
sPTPO65/sPBT11	2.0	12	35.9	0.872	8.9	1.78	17.6
Nafion 117	—	>24	25.6	0.120	193	15.0	2.59

---

The selectivity (ratio of proton conductivity to methanol permeability) of the membranes was also an important property of PEMs. The selectivity of these membranes was compiled in Table 3. As shown, the selectivity of blend membranes was much higher than that of the corresponding raw membranes because of the acid–base interactions. sPTPO60/sPBT6.25 exhibited a selectivity of 31.5 × 10⁴ S s cm⁻³. It was more than 12 times that of Nafion 117 (2.59 × 10⁴ S s cm⁻³). The selectivity of Nafion 117 was close to the reported value.³²

Oxidative stability of membranes

The oxidative stability of aromatic PEMs is an important issue for their practical applications.¹ In general, the oxidative stability was explored using Fenton's reagent (3% H₂O₂, 2 ppm FeSO₄) following the widely used procedure,¹² which was evaluated by the consumed time when the membrane began to dissolve (τ₁) and entirely dissolve (τ₂) at 80 °C. The consumed time (τ₁ and τ₂) of the raw and blend membranes are listed in Table 3. sPTPO-60 and sPTPO-65 exhibited a τ₂ of 6.5 and 3.5 h, respectively. Their consumed time (τ₂) was longer than that (2 h) of most aromatic PEMs, showing excellent oxidative stability.^33,34 The phosphine oxide-containing aromatic PEMs often denote excellent oxidative stability.^15,17 It can be noted that the blend membranes showed much longer consumed time (τ₂) than the corresponding raw membranes. For example, sPTPO60/sPBT6.25 and sPTPO65/sPBT10.75 displayed a τ₂ of 18 and 10 h, respectively, indicating excellent oxidative stability. In contrast to proton transport and methanol permeability, the consumed time of blend membranes increased with increasing sPBT content. The oxidative degradation of PEMs resulted from the hydrated HO˙ and HOO˙ radicals during the running of PEMFC. Lower water uptake in the PEMs depressed the diffusion of hydrated HO˙ and HOO˙ radicals and weakened the degradation of PEMs.^15,27 On the other hand, the sPBT component reduced the water uptake of the blend membranes and their water uptake decreased with increasing sPBT content. Therefore, the blend membranes containing higher sPBT content exhibited lower water uptake but longer consumed time (i.e. better oxidative stability).

Mechanical properties of membranes

Proton exchange membranes have three functions in fuel cells, including the transport of protons, the support of the catalyst, and the isolation of oxygen and fuel gas; therefore, appropriate mechanical properties are required.^2,14 The mechanical properties of two series of membranes were measured under the reported conditions.^12,27 The testing results are compiled in Table 3. The tensile strength and Young's modulus of blend membranes greatly improved but the elongation at break obviously decreased with increasing sPBT content. As aforementioned, there were acid–base interactions between benzothiazole and sulfonic acid groups in these blend membranes, therefore these interactions made the blend membranes display higher tensile strength and Young's modulus but lower elongation at break as the sPBT content increased. The blend membranes exhibited a tensile strength of 35.5–39.1 MPa and a Young's modulus of 0.861–1.34 GPa, which is much higher than those of Nafion 117. Although the elongation at break of the blend membranes decreased in contrast to the corresponding raw membranes, it was still up to 8.9–10%. Compared to the previous aromatic PEMs, these membranes presented excellent mechanical properties.^35,36

Microstructure of membranes

The hydrophobic domains of PEMs consist of hydrophobic polymer backbone and make the membranes retain their mechanical strength and dimensional stability. In contrast, the hydrophilic domains are composed of hydrated sulfonic acid groups, which form the channels for the diffusion of water and the transport of proton.^7,12 The microstructure of these membranes was explored by SAXS. The characteristic periodicity (d) between the hydrophobic and hydrophilic domains was used to characterize the microscopic structure of these PEMs.¹² To improve the electron-density contrast, the membranes were exchanged with the Pb²⁺ form before measurement following the previous procedures.^12,37 The SAXS curves are presented in Fig. 7. sPTPO-60 showed an ionomer peak with a slightly higher q value than sPTPO-65. It exhibited a characteristic periodicity (d) of 2.5 nm, which is slightly lower than that (2.6 nm) of sPTPO-65. This was because sPTPO-60 showed a lower degree of sulfonation than sPTPO-65. On the other hand, the two series of blend membranes both presented a slightly lower q value than the corresponding raw membranes, thus the blend membranes had slightly larger characteristic periodicity (d) than the raw membranes. This result was probably because the basic benzothiazole groups of sPBT facilitated the acid ionic-domains to gather and thus the size of the ionic-domains increased. Similar phenomena were observed in previous reports.^38,39 For each series of blend membranes, the characteristic periodicity (d) of two blend membranes was quite close because the difference of their sPBT content was very little (only 0.25%).

Fig. 7 SAXS profiles of membranes.

As typical examples, the microstructures of sPTPO-60 and sPTPO60/sPBT6.5 were investigated further by STEM for comparison. The STEM images of sPTPO-60 and sPTPO60/sPBT6.5 are presented in Fig. 8. Both of them exhibited dark and light regions, indicating distinct nanophase separation.³⁷ The light regions represented the hydrophobic domains containing a polymer backbone, which have the function to endow the membrane with dimensional stability. The dark regions are assigned to the hydrophilic ionic domains, forming the channels for the transport of water and protons.^7,12 In Fig. 8, the average characteristic periodicity (d) of sPTPO60/sPBT6.5 was slightly larger than that of sPTPO-60. Moreover, the average characteristic periodicity (d) of sPTPO-60 and sPTPO60/sPBT6.5 was similar to the value derived from their SAXS curves. It could also be observed from Fig. 8 that the connectivity of the hydrophilic ionic networks of sPTPO60/sPBT6.5 was worse than that of sPTPO-60 to some extent. This factor depressed the transport of water and hydrated protons, facilitating sPTPO60/sPBT6.5 to display lower swelling, less methanol permeability, and higher oxidative stability than sPTPO-60.

Fig. 8 STEM images of (a) sPTPO-60 and (b) sPTPO60/sPBT6.5.

Conclusions

Sulfonated poly(arylene thioether phosphine oxide)s (sPTPO) with a sulfonation degree of 60% and 65% were synthesized via the polymerization of SBFPPO and BFPPO with 1,4-benzenedithiol, which exhibited large swelling at high temperatures. Sulfonated polybenzothiazoles were for the first time used as a blend component to prepare the composite membranes and to reduce the swelling and improve the overall properties. The resulting blend membranes displayed markedly enhanced comprehensive properties compared to the raw sPTPO membranes. As a consequence, sPTPO60/sPBT6.25 showed a proton conductivity higher than that of Nafion 117, a swelling of only 30% even at 100 °C, and a methanol permeability of 1.02 × 10⁻⁷ cm² s⁻¹. Furthermore, it also showed excellent thermal/oxidative stability and mechanical properties. The property improvement of blend membranes was attributed to the acid–base interactions between the benzothiazole and sulfonic acid groups. Therefore, sulfonated polybenzothiazoles can be employed as an excellent blend component to enhance the dimensional stability and other properties of aromatic PEMs.

Acknowledgements

The authors wish to thank the financial support from the National Natural Science Foundation of China (No. 21274089, 21320102006, 21404071, and 91127047) and the National Basic Research Program (No. 2013CB834506).

References

1. R. Borup, J. Meyers, B. Pivovar, Y. S. Kim, R. Mukundan, N. Garland, D. Myers, M. Wilson, F. Garzon, D. Wood, P. Zelenay, K. More, K. Stroh, T. Zawodzinski, J. Boncella, J. E. McGrath, M. Inaba, K. Miyatake, M. Hori, K. Ota, Z. Ogumi, S. Miyata, A. Nishikata, Z. Siroma, Y. Uchimoto, K. Yasuda, K.-I. Kimijima and N. Iwashita, Chem. Rev., 2007, 107, 3904–3951.
2. M. A. Hickner, H. Ghassemi, Y. S. Kim, B. R. Einsla and J. E. McGrath, Chem. Rev., 2004, 104, 4587–4612.
3. H. W. Zhang and P. K. Shen, Chem. Rev., 2012, 112, 2780–2832.
4. T. J. Peckham and S. Holdcroft, Adv. Mater., 2010, 22, 4667–4690.
5. C. H. Park, C. H. Lee, M. D. Guiver and Y. M. Lee, Prog. Polym. Sci., 2011, 36, 1443–1498.
6. Q. F. Li, J. O. Jensen, R. F. Savinell and N. J. Bjerrum, Prog. Polym. Sci., 2009, 34, 449–477.
7. K. D. Kreuer, J. Membr. Sci., 2001, 185, 29–39.
8. C. X. Zhao, D. He, Y. T. Li, J. F. Xiang, P. Li and H.-J. Sue, RSC Adv., 2015, 5, 47284–47293.
9. Y. X. Zhao, Y. Q. Fu, Y. He, B. Hu, L. D. Liu, J. H. Lü and C. L. Lü, RSC Adv., 2015, 5, 93480–93490.
10. L. S. Xu, H. L. Han, M. Y. Liu, J. M. Xu, H. Z. Ni, H. L. Zhang and Z. Wang, RSC Adv., 2015, 5, 83320–83330.
11. X. Liu, X. Y. Meng, J. T. Wu, J. B. Huo, L. S. Cui and Q. Zhou, RSC Adv., 2015, 5, 69621–69628.
12. H. Y. Liao, K. Zhang, G. S. Tong, G. Y. Xiao and D. Y. Yan, Polym. Chem., 2014, 5, 412–422.
13. H. Y. Liao, G. Y. Xiao and D. Y. Yan, Chem. Commun., 2013, 49, 3979–3981.
14. L. C. Fu, G. Y. Xiao and D. Y. Yan, J. Mater. Chem., 2012, 22, 13714–13722.
15. L. C. Fu, H. Y. Liao, G. Y. Xiao and D. Y. Yan, J. Membr. Sci., 2012, 389, 407–415.
16. D. Liu, H. Y. Liao, N. Tan, G. Y. Xiao and D. Y. Yan, J. Membr. Sci., 2011, 372, 125–133.
17. L. C. Fu, G. Y. Xiao and D. Y. Yan, ACS Appl. Mater. Interfaces, 2010, 2, 1601–1607.
18. L. Y. Gui, C. J. Zhang, S. Kang, N. Tan, G. Y. Xiao and D. Y. Yan, Int. J. Hydrogen Energy, 2010, 35, 2436–2445.
19. X. H. Ma, C. J. Zhang, G. Y. Xiao and D. Y. Yan, J. Power Sources, 2009, 188, 57–63.
20. C. J. Zhang, S. Kang, X. H. Ma, G. Y. Xiao and D. Y. Yan, J. Membr. Sci., 2009, 329, 99–105.
21. X. H. Ma, L. P. Shen, C. J. Zhang, G. Y. Xiao, D. Y. Yan and G. M. Sun, J. Membr. Sci., 2008, 310, 303–311.
22. X. H. Ma, C. J. Zhang, G. Y. Xiao, D. Y. Yan and G. M. Sun, J. Polym. Sci., Part A: Polym. Chem., 2008, 46, 1758–1769.
23. J. A. Kerres, J. Membr. Sci., 2001, 185, 3–27.
24. J. Jouanneau, R. Mercier, L. Gonon and G. Gebel, Macromolecules, 2007, 40, 983–990.
25. G. Wang, Y. F. Yao, G. Y. Xiao and D. Y. Yan, J. Membr. Sci., 2013, 425–426, 200–207.
26. G. Wang, G. Y. Xiao and D. Y. Yan, Int. J. Hydrogen Energy, 2012, 37, 5170–5179.
27. N. Tan, G. Y. Xiao and D. Y. Yan, Chem. Mater., 2010, 22, 1022–1031.
28. N. Tan, Y. Chen, G. Y. Xiao and D. Y. Yan, J. Membr. Sci., 2010, 356, 70–77.
29. N. W. Li, Z. M. Cui, S. B. Zhang and S. H. Li, J. Polym. Sci., Part A: Polym. Chem., 2008, 46, 2820–2832.
30. S. L. Zhong, T. Z. Fu, Z. Y. Dou, C. J. Zhao and H. Na, J. Power Sources, 2006, 162, 51–57.
31. P. X. Xing, G. P. Robertson, M. D. Guiver, S. D. Mikhailenko and S. Kaliaguine, Macromolecules, 2004, 37, 7960–7967.
32. H.-Y. Li and Y.-L. Liu, J. Mater. Chem. A, 2014, 2, 3783–3793.
33. K. Miyatake, H. Zhou and M. Watanabe, Macromolecules, 2004, 37, 4956–4960.
34. B. J. Liu, G. P. Robertson, D. S. Kim, M. D. Guiver, W. Hu and Z. H. Jiang, Macromolecules, 2007, 40, 1934–1944.
35. N. W. Li, S. B. Zhang, J. Liu and F. Zhang, Macromolecules, 2008, 41, 4165–4172.
36. Y. Gao, G. P. Robertson, M. D. Guiver, G. Q. Wang, X. G. Jian, S. D. Mikhailenko, X. Li and S. Kaliaguine, J. Membr. Sci., 2006, 278, 26–34.
37. T. B. Norsten, M. D. Guiver, J. Murphy, T. Astill, T. Navessin, S. Holdcroft, B. L. Frankamp, V. M. Rotello and J. F. Ding, Adv. Funct. Mater., 2006, 16, 1814–1822.
38. W. Li, A. Manthiram and M. D. Guiver, J. Membr. Sci., 2010, 362, 289–297.
39. W. Li, A. Manthiram and M. D. Guiver, Electrochem. Solid-State Lett., 2009, 12, B180–B184.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra23188d

---