Improved performance of poly(vinyl pyrrolidone)/phosphonated poly(2,6-dimethyl-1,4-phenylene oxide)/graphitic carbon nitride nanocomposite membranes for high temperature proton exchange membrane fuel cells

Abstract

To achieve desirable performance of a polymer electrolyte membrane with higher proton conduction and better mechanical strength is a challenging work in the development of the phosphoric acid (PA) doped solid-state membrane for high temperature proton exchange membrane fuel cells. Firstly, phosphonated poly(2,6-dimethyl-1,4-phenylene oxide) (pPPO) was prepared by the bromination and phosphonation of poly(2,6-dimethyl-1,4-phenylene oxide). Afterward, a series of poly(vinyl pyrrolidone)–phosphonated poly(2,6-dimethyl-1,4-phenylene oxide) (PVP/pPPO) blend membranes and poly(vinyl pyrrolidone)–phosphonated poly(2,6-dimethyl-1,4-phenylene oxide)–graphitic carbon nitride (PVP/pPPO/g-C₃N₄) nanocomposite membranes were prepared by a solution casting method. The PA uptake, volume swelling ratio, and proton conductivity of the PVP/pPPO blend membrane increased with increasing PVP content. But PA molecules drastically reduced the mechanical strength of the PVP/pPPO blend membrane. The incorporation of g-C₃N₄ improved the proton conduction and mechanical properties of the nanocomposite membrane due to the proton hopped sites provided by NH₂ and the interaction of g-C₃N₄ and polymer chains. A higher proton conductivity of 74.4 mS cm⁻¹ and a higher power density of 294 mW cm⁻² at 180 °C without additional humidifying were observed for the PA doped PVP/pPPO nanocomposite membrane containing 5 wt% g-C₃N₄. The results show the PVP/pPPO nanocomposite membrane as a potential polymer electrolyte membrane for high temperature fuel cells.

---

1. Introduction

In recent decades, proton exchange membrane fuel cells (PEMFCs) have been paid wide attention in the area of clean and efficient energy conversion devices.^1–3 The traditional and commercial polymer electrolyte materials are perfluorosulfonic acid ionomer (PFSI) membranes, which have high proton conductivity and excellent long-term performance for fuel cells at moderate temperatures (60–90 °C), but the PFSI membranes suffer from low proton conductivity, and poor mechanical and chemical stability at high temperature (above 100 °C).⁴ PEMFCs operating at 100–200 °C will benefit from the high temperature in simplified water and heat management systems, high activity for the electrode reaction, high tolerance to CO poisoning and low Pt catalyst loading.^5,6 Thus the development of polymer electrolyte materials for high temperature PEMFCs is another interesting research topic.

Among the alternative and promising polymers, phosphoric acid doped polybenzimidazole (PBI) membranes have been widely studied for the high temperature proton exchange membranes (HTPEMs) because of good thermal stability and high mechanical strength.^7–9 The imidazole groups in the main chains of PBI form acid–base pairs with PA molecules and build a hydrogen bond network.^10,11 Proton transportation depends on the dynamic hydrogen bond networks.^12,13 To achieve high proton conductivity, the membranes should have a high PA doping level, but abundant PA molecules will rapidly reduce the mechanical properties of the PBI membranes at the same time.¹⁴ High molecule weight can enhance the mechanical stability of the PBI membranes with high PA doping level, while the processibility of PBI will be a problem because of the poor solubility.¹⁵ Many other alternative materials bearing basic groups have also been developed for the HTPEMs including poly(vinyl pyrrolidone) (PVP).¹⁶ PVP is a tough hydrophilic polymer containing N-heterocycle which can provide the PA absorption sites. Many investigations about PVP blending with other polymers like PBI,¹⁷ poly(vinylidene fluoride-co-hexafluoropropylene) (PVdF-HFP),¹⁸ chitosan (CS),¹⁹ poly(vinyl alcohol) (PVA)²⁰ and polyethersulfone (PES)^21,22 have been well reviewed. The second component should be miscible with PVP and have high mechanical strength and good thermal stability.

It is well known that aromatic polymers have been considered promising candidates for the development of proton exchange membranes for fuel cells.^23,24 Poly(2,6-dimethyl-1,4-phenylene oxide) (PPO) is an engineering polymer which is commercially available and has important properties similar to other aromatic polymers. PPO has also been widely studied for its excellent mechanical properties, outstanding chemical and thermal stability.^25–27 To improve the processibility, PPO can be readily functionalized to the brominated poly(2,6-dimethyl-1,4-phenylene oxide) (BPPO) and the phosphonated poly(2,6-dimethyl-1,4-phenylene oxide) (pPPO) via the substitution reaction.^27–31 Moreover, phosphonated poly(2,6-dimethyl-1,4-phenylene oxide) (pPPO) contains the phosphonic acid groups which could serve as an amphoteric proton conductor.³²

In this work, PPO was firstly modified by bromination and then the benzylic methyl groups was functionalized to phosphonic acid groups. PVP/pPPO blend membranes were prepared by solution casting with different weight ratios. The conductivity of the PVP/pPPO blend membranes increased with the increasing of PVP content, but the mechanical strength drastically decreased. In order to optimize the blend membranes properties, inorganic fillers were added to enhance the mechanical stability at high PA doping level. Graphitic carbon nitride (g-C₃N₄) is a kind of graphene-like inorganic material.^33–35 The chemical structure of g-C₃N₄ is composed of tri-s-triazine (C₆N₇) units with pendant amino groups (Scheme 1). The chemical bond energy of C–N and C=N are 305 kJ mol⁻¹ and 615 kJ mol⁻¹ respectively, so g-C₃N₄ has good thermal stability. Moreover, the –NH₂ groups can form acid–base pairs with the phosphoric acid and enhancing the Grotthuss-type transfer of protons.³⁶ The interfacial interaction between nanofiller and pPPO might also have an effect on the mechanical properties of PA doped composite membrane. Thus, g-C₃N₄ is a good filler candidate to improve the thermal and mechanical stability of PA doped membrane. The preparation of g-C₃N₄ was investigated by FTIR and SEM. The PA absorption ability, proton conductivity, thermal and mechanical properties of the nanocomposite membranes were also studied. And, the results indicated that the PVP/pPPO/g-C₃N₄ nanocomposite membranes had a better overall performance than that of the PVP/pPPO blend membranes when appropriate amount of fillers were added.

Scheme 1 Chemical structure of graphitic carbon nitride.

2. Experimental

2.1 Materials

Poly(vinyl pyrrolidone) (PVP, M_w = 1 300 000) were purchased from Aladdin Industrial Co. Poly(2,6-dimethyl-1,4-phenylene oxide) (PPO, Asahi Kasei) was purchased from Gaodong Plastic Company. Chlorobenzene (99%), N-methyl-2-pyrrolidinone (NMP, 99%), phosphoric acid solution (85 wt%) were supplied by Tianjin Guangfu Fine Chemical Research Institute. Liquid bromine (99%), benzoyl peroxide (BPO, 99%), triethyl phosphite (TEP, 98%), diethyl carbitol (DEC, 98%), cyanamide (99%) were purchased from Sinopharm Chemical Reagent Co., Ltd. Methanol (99%), dichloromethane (99%), hydrochloric acid (36–38%) were obtained from Beijing Chemical Works. All the materials and chemicals were used as received.

2.2 Bromination and phosphonation of poly(2,6-dimethyl-1,4-phenylene oxide)

The preparation of phosphonated poly(2,6-dimethyl-1,4-phenylene oxide) (PPO) was carried on by three steps including benzyl bromination, phosphonation and hydrolysis,²⁸ as shown in Scheme 2. In detail, 12 g PPO (0.1 mol repeat unit), 0.24 g BPO (0.001 mol) and 100 mL chlorobenzene were added into a three-necked flask at 80 °C. The mixture was heated to 130 °C.³⁰ A solution of bromine (8 g, 0.05 mol) in 80 mL chlorobenzene was dropwise added for 3 h. The mixture maintained at 130 °C for 12 h under the nitrogen atmosphere. The brominated product was precipitated out in excess methanol and purified by dissolution in dichloromethane and precipitation with methanol, and then dried at 60 °C in a vacuum oven (13.5 g, yield = 112.5%).

Scheme 2 Preparation of phosphonated poly(2,6-dimethyl-1,4-phenylene oxide).

5 g brominated poly(2,6-dimethyl-1,4-phenylene oxide) (BPPO), 70 mL NMP, 30 mL DEC and 5 mL TEP were added into a flask.³¹ The mixture was refluxed at 140 °C for 8 h. After the reaction, the mixture solution was cooled to the room temperature and poured into excess deionized water with stirring. The product was dried at 60 °C in a vacuum oven (3.6 g, yield = 72%). At last, the phosphonic acid ester polymers (1 g, PPO-PE) were placed in hydrochloric acid (30 mL) and refluxed at 110 °C for 10 h. Phosphonated PPO (pPPO) was obtained after wash in water and dried at 80 °C (0.92 g, yield = 92%).

2.3 Synthesis of graphitic carbon nitride (g-C₃N₄)

The g-C₃N₄ powers were prepared from cyanamide by heating at high temperature.³⁷ Cyanamide (5 g) were placed in an alumina crucible and heated to 550 °C in a muffle furnace at a heating rate of 10 °C min⁻¹. After 4 hours, the furnace was cooled to room temperature. The products were collected and ground into a light yellow powders (2.6 g, yield = 52%).

2.4 Preparation of membranes

PVP and pPPO were completely dissolved into N-methyl-2-pyrrolidinone, respectively. The PVP and pPPO solution were mixed with a certain ratio in N-methyl-2-pyrrolidinone, and a transparent and homogeneous casting solution was obtained. The total polymer concentration was about 5 wt%. The mixture solution was coated in a glass dish and dried at 80 °C for 24 h to remove the solvent and obtained the blend membrane. The blend membrane was denoted as x PVP/pPPO, where x represents the weight ratio of PVP in the blend membrane (i.e., 20–80 wt%). The nanocomposite membranes were prepared by mixing a certain amount of g-C₃N₄ in 70%PVP/pPPO solution. At room temperature, g-C₃N₄ was dispersed in the PVP/pPPO mixture solution using an ultrasonic bath. The solution was also cast onto a glass plate to obtain the nanocomposite membrane at 80 °C for 24 h. The thickness of blend membrane and nanocomposite membrane was controlled in a range between 200 and 250 μm.

2.5 Characterization of membranes

2.5.1 FTIR, ¹H NMR and XRD. Fourier transform infrared spectroscopy (FTIR) was performed on a Nicolet Nexus 670 FTIR-ATR spectrometer system between 4000 and 600 cm⁻¹. ¹H NMR spectra of the polymers were measured on a JNM-ECA300 spectrometer (¹H, 300 MHz) using CDCl₃ or DMSO_d6 as solvent according to the solubility of polymers. The level of phosphonation of PPO was 42.36% calculated from the data of ¹H-NMR (seen in ESI†). The crystal structure of g-C₃N₄ was investigated using X-ray diffraction (Bruker D8 Advance, X-ray diffractometer) with a 2θ range of 5–60 degree at a scan rate of 4° min⁻¹.

2.5.2 Morphology. The morphology of g-C₃N₄ and membranes were studied by a Quanta 200F field emission-environment scanning electron microscope (SEM) and a FEI Tecnai G2 F20 transmission electron microscopy (TEM).

2.5.3 Water uptake, PA uptake and swelling ratio. For the determination of water uptake, membrane samples of 1 × 4 cm² size were dried in vacuum at 80 °C for 24 h. The weight of the dry membrane was obtained. Subsequently, the membranes were immersed into deionized water at room temperature for 24 h. Then the membranes were taken out and wiped the water on the surface. The wet membrane was obtained. For the determination of PA uptake, the PA doping treatment of membrane was conducted by immersing into 85 wt% phosphoric acid at 80 °C for 24 h. The membranes were taken out, wiped the PA on surface and keep the membranes in vacuum oven at 80 °C for 24 h to reduce the effect of the absorbed water.³⁸ The weight of PA doped membrane was obtained immediately. The water uptake and PA uptake were calculated as below:

Water uptake (%) = (W_wet − W_dry)/W_dry × 100% (1)

PA uptake (%) = (W_PA − W_dry)/W_dry × 100% (2)

where W_wet and W_dry are the membranes weights after and before immersing into deionized water. W_PA and W_dry are the membranes weights after and before PA doping treatment. The volume swelling ratio after PA doping treatment were calculated by the following equation:

Volume swelling ratio (%) = (V_PA − V_dry)/V_dry × 100% (3)

where V_PA and V_dry were the volume of membranes after and before PA doping treatment.

2.5.4 Thermal and mechanical properties. Thermal analysis of the membranes was conducted by using thermo gravimetric analysis, with TGA2050, at 10 °C min⁻¹ up to 800 °C under nitrogen atmosphere. The mechanical properties of membranes was measured on a tensile test instrument at room temperature. All membranes were cut into 40 mm × 10 mm and the thicknesses were recorded by a micrometer. Each sample was performed with a constant speed of 5 mm min⁻¹ for three times.

2.5.5 Proton conductivity. Through-plane conductivity measurements were carried out in a test instrument. Each membrane was sandwiched between two electrodes and measured on electrochemical workstation (CHI 660D) with an AC impedance from 10⁶ Hz to 100 Hz. The thickness of membrane was recorded. The proton conductivity of membranes were calculated by the following equation:

σ = L/RA (4)

where the L was the thickness of membranes, R was the resistance of membranes and A was the area of roundness electrodes.

2.5.6 Single cell performance test. The PVP–pPPO composite membranes were sandwiched between two electrodes. The electrode was composed of a Pt/C catalyst layer with Pt loading of 0.5 mg cm⁻² and a porous carbon substrate. Membrane electrolyte assemblies (MEAs) were tested on a fuel cell system (Greenlight G20, Canada) with dry H₂ and O₂ of 150 mL min⁻¹.

3. Results and discussion

3.1 The PA doped PVP/pPPO blend membranes

3.1.1 Bromination and phosphonation of PPO. PPO was firstly brominated to introduce functional groups onto the aromatic rings. Then phosphonation of BPPO was performed by classical Michaelis–Arbuzov reaction.³¹ It could be found clearly the characteristic absorption at 782, 1053 and 1255 cm⁻¹ corresponding to the phosphonic ester (PPO-PE), which disappeared after the hydrolysis of ester groups (Fig. 1(a)).^29,39 Phosphonation of BPPO could also be confirmed by ¹H-NMR (Fig. 1(b)). The proton peaks of PPO were observed at 6.5 ppm and 2.1 ppm for aromatic proton ring and –CH₃ respectively. The success of benzyl bromination could be demonstrated by the appearance of proton peak at 4.34 ppm (–CH₂Br).²⁸ Benzyl bromide could further react with triethyl phosphite and the introduction of phosphonic ester would show the peaks at 4.02 ppm and 1.2 pm for methylene and methyl in –PO(OCH₂CH₃)₂.³¹ The peaks at 2.5 ppm and 2.54 ppm were the signals of dimethyl sulfoxide (DMSO) and deuterated dimethyl sulfoxide (DMSO_d6). A hump between 3.2 and 3.7 ppm for pPPO indicated the presence of phosphonic acid linkage on the aromatic ring.^40,41 The results of FTIR and ¹H-NMR confirmed the successful preparation of pPPO.

Fig. 1 FT-IR (a) and ¹H-NMR (b) of phosphonated poly(2,6-dimethyl-1,4-phenylene oxide).

3.1.2 Properties of PA doped PVP/pPPO blend membranes. The blend membranes containing different PVP content were prepared by solution blending method. The miscibility of two polymers was discussed in the ESI† due to limited space. The water uptake, PA uptake of PVP/pPPO blend membranes are shown in Fig. 2(a). The water uptake of blend membranes increased with the content of PVP because of the hydrophilicity of PVP. The PA uptake of blend membranes also increased with the PVP weight ratio because the N-heterocycle of PVP provided the PA absorption sites. And the acid doping level (ADL) of blend membranes increased from 5.74 for 20%PVP/pPPO to 8.59 for 80%PVP/pPPO (Fig. S6†). Fig. 2(b) was the volume swelling ratios of blend membranes after PA doping treatment. With more PVP content, the blend membranes could be faster to reach absorption equilibrium (Fig. S7†), and had more PA uptake and larger volume swelling ratio. The blend membrane containing 80%PVP had the highest PA uptake of 606.72% but it could not maintain the dimensional stability because of the excessive swelling. The thickness and area of PVP/pPPO blend membranes would increase after the PA doping treatment. The dimensional change of PVP/pPPO blend membranes after PA doping was shown in the Table S2.† Certainly, too much PA amount will be a problem for both PBI and PVP/pPPO blend. The PVP was a highly hydrophilic polymer, so the water absorption and PA doping amount increased with the content of PVP. The volume swelling ratio of 70%PVP/pPPO blend membrane was comparable to that of PBI.⁴² To consider both the dimensional stability and proton conductivity of PVP/pPPO blend, the 70% content PVP was chosen for further study.

Fig. 2 The properties of PA doped PVP/pPPO blend membranes: (a) water uptake and PA uptake, (b) volume swelling ratio, (c) proton conductivity, (d) tensile strength.

Fig. 2(c) shows the proton conductivity of PVP/pPPO blend membranes from 20 °C to 180 °C. As expected, the conductivity of PVP/pPPO blend membranes increased with the increasing of temperature and the content of PVP. The PA doped 70%PVP/pPPO blend membrane had the highest proton conductivity of 61.3 mS cm⁻¹ at 180 °C without additional humidifying, which was close to the value of PA doped PBI membrane (ADL = 5.6).⁴³ The calculated activation energies for PVP/pPPO blend membranes were within the range from 12 to 14 kJ mol⁻¹, which suggested that the conductivity mechanism in PA doped PVP/pPPO membranes was similar to that of a concentrated H₃PO₄ solution.^22,43

Similar to PBI, the interaction of PA molecules with the N-containing groups allows PVP to retain PA. The interaction of PA molecules with the N-containing groups contributes to the proton conductivity of the PVP/pPPO blend membrane. Ma et al. proposed that protons migrates by different pathway for different doping level.⁴⁴ At high acid doping level, there is more excess PA molecules, which build a network of hydrogen bond and form the anionic chain (H₂PO₄⁻⋯H₃PO₄).⁴² Protons transfer as well between other species, like N–H⁺⋯H₂PO₄⁻, N–H⁺⋯H₂O, N–H⁺⋯N–H.^44,45 The phosphonic acid groups anchoring on the chain of pPPO could also serve as an amphoteric proton conductor.³² In addition, the presence of the phosphonic acid groups of pPPO had a contradictory effect on the PA absorption of PVP.³ Even so, the blending of pPPO with PVP could improve the mechanical properties of the membrane.

The tensile strength of the PVP/pPPO blend membranes and the PA doped membranes are shown in Fig. 2(d). The tensile strength of blend membranes reduced from 43.12 MPa to 22.91 MPa with the increasing of PVP content. As same as the PA doped PBI membrane, PA molecules reduced dramatically the mechanical properties of PVP/pPPO blend membranes due to the larger free volume and the decrease of intermolecular force.^21,22 The PA doped 70%PVP/pPPO blend membrane only had a tensile strength of 1.49 MPa. In order to achieve fulfilling performance of PEM, the g-C₃N₄ as fillers were applied to enhanced the properties of PVP/pPPO blend membrane.

3.2 The PA doped PVP/pPPO/g-C₃N₄ nanocomposite membranes

3.2.1 FTIR of g-C₃N₄, PVP/pPPO/g-C₃N₄ nanocomposites membranes. Fig. 3 shows the FTIR spectra of g-C₃N₄, 70%PVP blend membrane and 70%PVP nanocomposites membranes with different content of g-C₃N₄. In the spectrum of g-C₃N₄ shown in Fig. 3(a), the broad absorption at around 3000–3300 cm⁻¹ was assigned to the antisymmetric stretching vibration of NH₂ and NH. The absorption at around 1640 cm⁻¹ was attributed to the bending vibration of NH₂. The absorptions at around 810 cm⁻¹ and 1571 cm⁻¹ was due to the deformation vibration of the triazine rings and the stretching vibration of C=N in the triazine rings. The strong absorption peaks at around 1237 cm⁻¹, 1317 cm⁻¹, 1406 cm⁻¹ and 1460 cm⁻¹ were assigned to the stretching vibration of C–N. The spectrum result was in good agreement with the reported literatures.^34,46

Fig. 3 The FT-IR spectra of (a) g-C₃N₄, (b) 70%PVP/pPPO blend membrane and PVP/pPPO nanocomposites membranes with different amount of g-C₃N₄: (c) 1%, (d) 5%, (e) 10%.

Fig. 3(b) shows the spectrum of 70%PVP/pPPO blend membrane. The absorption around 3000–3600 cm⁻¹ was due to the absorption water. The strong absorption at 1639 cm⁻¹ was attributed to the stretching vibration of C=O for PVP. The absorptions at around 1294 cm⁻¹ and 1424 cm⁻¹ were assigned to the stretching vibration of C–N and the bending vibration of C–H for PVP.⁴⁷ The absorptions at 1466 cm⁻¹ and 1446 cm⁻¹ belonged to the deformation vibration of the aromatic rings for pPPO. The characteristic absorption peak of P=O appeared at around 1187 cm⁻¹ for phosphonic acid groups. Compared to the 70%PVP blend membrane, the nanocomposite membranes containing 1 wt%, 5 wt% and 10 wt% g-C₃N₄ had absorptions at 1237 cm⁻¹ and 1406 cm⁻¹ due to the stretching vibration of C–N of g-C₃N₄, which confirmed the preparation of the nanocomposite membranes.

3.2.2 Morphology of PVP/pPPO/g-C₃N₄ nanocomposites membranes. Fig. 4 shows the cross-section morphologies of the 70%PVP/pPPO blend membrane and 70%PVP/pPPO nanocomposite membranes. In the Fig. 4(a), the image of 70%PVP/pPPO blend membrane appeared a uniform microstructure. For nanocomposite membranes with the weight ratio of g-C₃N₄ from 1% to 5%, the morphologies were similar with the 70%PVP/pPPO blend membrane which indicated that the dispersion of fillers in the blend membrane was homogeneous. When the amount of g-C₃N₄ increased, some g-C₃N₄ fillers agglomerated together for the nanocomposite membranes with 7% and 10% g-C₃N₄ like other inorganic fillers.^9,48 The amine groups might help g-C₃N₄ particles to disperse in the polymer matrix and provide more proton pathways.

Fig. 4 The cross-section morphologies of the membranes: (a) 70%PVP/pPPO and 70%PVP/pPPO containing (b) 1%, (c) 3%, (d) 5%, (e) 7%, (f) 10% g-C₃N₄.

3.2.3 Water uptake, PA uptake and volume swelling ratio. Water absorption and PA absorption are important parameters for the PEM because water molecules and PA molecules not only work as plasticizer to improve the motion of polymer chains but also facilitate the transportation of protons as medium. The water uptake, PA uptake and volume swelling ratio of PVP/pPPO/g-C₃N₄ nanocomposite membranes are shown in Fig. 5. The water uptake of nanocomposite membranes have shown a decreasing tendency with the increasing of g-C₃N₄ amount because of the low water absorption of fillers. The PA uptake behaviour was similar with the water absorption. It could be seen that the value of PA uptake of nanocomposite membranes decreased to 445.2% with 10% g-C₃N₄ from 480.8% with 1% g-C₃N₄, which might be attributed to the low PA absorption of g-C₃N₄. And the aggregation of fillers also leaded to the decrease of the PA absorption. The volume swelling ratio of nanocomposite membranes was closely related to the PA uptake and gradually decreased with the increase of inorganic filler content. The nanocomposite membrane with the least PA uptake also had the smallest volume swelling ratio of 178.8% at the g-C₃N₄ content of 10%. Incorporation of g-C₃N₄ would improve the structural stability of the nanocomposite membranes.

Fig. 5 Water uptake, PA uptake (a) and volume swelling ratio (b) of PVP/pPPO nanocomposite membranes with different g-C₃N₄ content.

3.2.4 Thermal stability. Thermo gravimetric analysis (TGA) was performed to evaluate the thermal stability of the PVP/pPPO blend membranes and the PVP/pPPO nanocomposite membranes in the application of HTPEMFCs. The thermograms of g-C₃N₄ and 70%PVP/pPPO blend and nanocomposite membranes are shown Fig. 6(a). For the PVP/pPPO blend membranes, the TGA curves had three weight loss stages: 30–150 °C, 200–350 °C, 350–500 °C. The initial weight loss about 10% was reasonably due to loss of water molecules, which was the absorbed moisture by the hydrophilic PVP. The second weight loss at 200–350 °C was attributable to the solvent residues, or possibly anhydride formation. The condensation of phosphonic acid of pPPO would also take place with elevated temperature.^31,49 The third weight loss occurred around 350–500 °C which mainly caused by the degradation of polymer chains.

Fig. 6 TGA curves of (a) g-C₃N₄, 70%PVP/pPPO blend and nanocomposite membranes, (b) PA doped 70%PVP/pPPO blend and nanocomposite membranes.

The thermal degradation temperature of pure g-C₃N₄ was as high as 550 °C. The temperature at the max decomposition rate was about 700 °C. For PVP/pPPO nanocomposite membranes, a slight weight loss below 150 °C was caused by the evaporation of water or residual solvent. The second weight loss happened around 350–450 °C, which was apparently ascribed to the decomposition of polymer chains. Above 450 °C, the decomposition rate of polymer chains reduced which might be affected by the interaction between the polymer chains and g-C₃N₄ particles. The decomposition of g-C₃N₄ particles would also proceed in advance.

Fig. 6(b) presents the TGA curves of PA doped 70%PVP/pPPO blend and nanocomposite membranes with different g-C₃N₄ content. The PA doped membranes had a large weight loss below 150 °C due to the strong hygroscopicity of the phosphoric acid. And the auto-polymerisation of the phosphoric acid to pyrophosphoric and triphosphoric acids happened around 200 °C. After all, the PVP/pPPO nanocomposite membranes shown adequate thermal stability for HTPEMs.

3.2.5 Mechanical properties. Mechanical stability is another important performance for the application of HTPEMs in fuel cells. The mechanical properties of 70%PVP/pPPO blend membrane and PVP/pPPO/g-C₃N₄ nanocomposite membranes after the PA doping treatment were summarized in Table S3,† and Fig. 7 shows the stress–strain curves of those membranes. The PA doped 70%PVP/pPPO blend membrane had the low mechanical strength and Young's modulus of 1.49 MPa and 8.76 MPa. Because the PA molecules facilitated the motion of polymer chains and reduced the intermolecular force between the polymer chains. It could be clearly found that the fillers enhanced the mechanical strength of PA doped membranes. The maximum elongation and Young's modulus also increased after the addition of g-C₃N₄. At 5% content of g-C₃N₄, the nanocomposite membrane had the highest tensile strength of 6.48 MPa suggesting the reinforcing effect of g-C₃N₄ in the polymer matrix. The mechanical properties of nanocomposite membranes decreased for 7% and 10% g-C₃N₄ which could be explained by the aggregation of the fillers (Fig. 4(e) and (f)). Compared to the PA doped PVP/pPPO blend membranes, the mechanical properties of PVP/pPPO nanocomposite membranes were enhanced by the inorganic fillers of g-C₃N₄.

Fig. 7 Mechanical properties of PA doped PVP/pPPO/g-C₃N₄ nanocomposite membranes.

3.2.6 XRD. Fig. S9† shown the XRD patterns of g-C₃N₄, 70%PVP/pPPO blend and nanocomposite membranes. For 70%PVP/pPPO blend membrane, two brand peaks at about 2θ = 12° and 23° were determined due to the amorphous phase. When high content g-C₃N₄ nanofiller (5% and 10%) was added, the main peak of g-C₃N₄ (2θ = 27.4°) would appear and the intensity of g-C₃N₄ peaks in the composite increased. It illustrated that the structure of polymer blend membrane did not change when g-C₃N₄ was introduced into the blend membrane. The characteristic peaks of g-C₃N₄ could not be found for the PA doped composite membranes due to the effect of PA molecules.⁵⁰ So it can be concluded that the PA doped composite membranes were all amorphous.

3.2.7 Proton conductivity. For PA doped polymer electrolyte membrane, the proton conductivity always depend on PA absorption amount of membranes because the proton pathway was provided by the hydrogen bond network of PA molecules.⁴⁸ The proton conductivities of PVP/pPPO nanocomposite membranes at different temperature without additional humidification are shown in the Fig. 8. As expected, for all nanocomposite membranes, the proton conductivity increased with increasing the test temperature.

Fig. 8 Proton conductivity of PA doped PVP/pPPO/g-C₃N₄ membranes at different temperature without additional humidification.

The temperature was not the only factor in elevating proton conduction ability of PA doped PVP/pPPO nanocomposite membranes. The fillers of g-C₃N₄ had an effect on the PA absorption and the proton transportation. With increasing g-C₃N₄ content, the proton conductivity was supposed to reduce because of the lower PA uptake. But the nanocomposite membranes with 3% and 5% g-C₃N₄ had higher proton conductivity than that of the 70%PVP/pPPO blend membrane under same condition. The membrane with 5% g-C₃N₄ had the highest proton conductivity of 74.4 mS cm⁻¹ at 180 °C.

At high temperature and under anhydrous condition, the proton transport mainly depends on the hydrogen bonds network of the PA doped membrane by a Grotthuss mechanism.⁴² The interaction between the graphitic carbon nitride and the acid groups provided another proton pathway due to the NH₂ of g-C₃N₄ formed the Lewis acid–base pairs with acid groups.³⁶ In order to have a deeper understand of the role of g-C₃N₄ on nanocomposite membranes, the activation energy (E_a) for proton conductivity had also been investigated in Fig. S10.† The E_a value of composite membrane containing 5 wt% g-C₃N₄ decreased that indicated the lower energy barrier for proton transfer.⁵¹ Compared to other composite membranes, the 5% g-C₃N₄ composite membrane probably had a better proton transport pathway and more contribution to the Grotthuss mechanism for proton hopping transport. Incorporation of high g-C₃N₄ content decreased the PA absorption which resulted in the lower proton conductive for nanocomposite membranes compared to that of blend membrane.

We made a comparison about the proton-conduction performance of PA doped 70%PVP/pPPO blend and nanocomposite membrane with several PA doped polymer systems (Table S4†). The proton conductivity of PA doped 70%PVP/pPPO blend membrane and PA doped 70%PVP/pPPO/g-C₃N₄ (5%) were higher than that of PA doped PBI and were comparable to several PA doped membranes. All performances considered, it indicated that the system of PVP/pPPO/g-C₃N₄ could be a promising alternative membrane for high temperature proton exchange membrane fuel cells.

3.2.8 Single cell performance test. The PEMFC single cell performances of 70%PVP/pPPO blend membrane and 70%PVP/pPPO/g-C₃N₄ nanocomposite membrane were evaluated by the polarization and power density curves at different temperature and without additional humidification in H₂/O₂ (Fig. 9). The open circuit voltages (OCV) of fuel cells were above 0.9 V, which indicates the membranes were pore-free and had a low gas permeability.⁹ The PA doped 70%PVP/pPPO blend membrane shown 138 mW cm⁻² power density and 420 mA cm⁻² current density at 0.33 V and 180 °C. The maximum power density improved to 294 mW cm⁻² when 5 wt% g-C₃N₄ nanofillers were incorporated into the 70%PVP/pPPO blend membrane. The 70%PVP/pPPO blend membranes with 5 wt% g-C₃N₄ had a higher power density than the blend membrane due to the higher proton conductivity. The proton conductivity of 70%PVP/pPPO blend membrane and 70%PVP/pPPO/g-C₃N₄ nanocomposite membrane measured by the single cell performance test were 0.14 S cm⁻¹ and 0.27 S cm⁻¹ at 180 °C. It was well understood that the maximum power density increased from 235 mW cm⁻² to 294 mW cm⁻² with elevated temperature due to the improved proton conductivity and the enhanced reaction kinetics. Overall, the PVP/pPPO blend membrane and PVP/pPPO/g-C₃N₄ nanocomposite membranes are potential polymer electrolyte membrane for high temperature fuel cells.

Fig. 9 Single cell performances of (a) PA doped 70%PVP/pPPO membrane and (b) PA doped PVP/pPPO/5 wt% g-C₃N₄ membrane at different temperature.

4. Conclusions

In summary, pPPO was successfully prepared by bromination and phosphonation of PPO. Then, a series of PVP/pPPO blend membranes was prepared by solution blending method as PA doped polymer electrolyte membrane for HTPEMs. The water uptake, PA uptake and volume swelling ratio of PVP/pPPO blend membranes increased with the increasing of PVP content. More PA uptake will have higher proton conduction but also lead to larger swelling and worse mechanical strength. The 70%PVP/pPPO blend membrane had the proton conductivity of 61.3 mS cm⁻¹ and the power density of 138 mW cm⁻² at 180 °C without additional humidifying. In a further study, a graphite-like inorganic filler of g-C₃N₄ was chosen to reinforce the properties of 70%PVP/pPPO blend membrane. The PVP/pPPO nanocomposite membranes display better structural stability and enhanced mechanical properties. Even the PA absorption of PVP/pPPO nanocomposite membranes reduced, the amine groups of g-C₃N₄ would provide extra proton pathway to improve the proton conduction. With 5 wt% g-C₃N₄ content, a higher proton conductivity of 74.4 mS m⁻¹ and a higher power density of 294 mW cm⁻² were achieved at 180 °C for PVP/pPPO nanocomposite membrane.

Acknowledgements

We thank Prof. Juntao Wu for his help and for the discussions. Acknowledge the financial support of the National Nature Science Foundation of China (51303211).

Notes and references

1. N. Gourdoupi, A. K. Andreopoulou, V. Deimede and J. K. Kallitsis, Novel proton-conducting polyelectrolyte composed of an aromatic polyether containing main-chain pyridine units for fuel cell applications, Chem. Mater., 2003, 15, 5044–5050.
2. K. D. Papadimitriou, F. Paloukis, S. G. Neophytides and J. K. Kallitsis, Cross-linking of side chain unsaturated aromatic polyethers for high temperature polymer electrolyte membrane fuel cell applications, Macromolecules, 2011, 44, 4942–4951.
3. T.-H. Tang, P.-H. Su, Y.-C. Liu and T. L. Yu, Polybenzimidazole and benzyl-methyl-phosphoric acid grafted polybenzimidazole blend crosslinked membrane for proton exchange membrane fuel cells, Int. J. Hydrogen Energy, 2014, 39, 11145–11156.
4. H. Zhang and P. K. Shen, Recent development of polymer electrolyte membranes for fuel cells, Chem. Rev., 2012, 112, 2780–2832.
5. W. Ma, C. Zhao, H. Lin, G. Zhang, J. Ni, J. Wang, S. Wang and H. Na, High-temperature water-free proton conducting membranes based on poly(arylene ether ketone) containing pendant quaternary ammonium groups with enhanced proton transport, J. Power Sources, 2011, 196, 9331–9338.
6. S. Authayanun, K. Im-orb and A. Arpornwichanop, A review of the development of high temperature proton exchange membrane fuel cells, Chin. J. Catal., 2015, 36, 473–483.
7. J. Lobato, P. Canizares, M. A. Rodrigo, J. J. Linares and J. A. Aguilar, Improved polybenzimidazole films for H₃PO₄-doped PBI-based high temperature PEMFC, J. Membr. Sci., 2007, 306, 47–55.
8. S. Wang, C. Zhao, W. Ma, G. Zhang, Z. Liu, J. Ni, M. Li, N. Zhang and H. Na, Preparation and properties of epoxy-cross-linked porous polybenzimidazole for high temperature proton exchange membrane fuel cells, J. Membr. Sci., 2012, 411–412, 54–63.
9. C. Xu, X. Liu, J. Cheng and K. Scott, A polybenzimidazole/ionic-liquid–graphite-oxide composite membrane for high temperature polymer electrolyte membrane fuel cells, J. Power Sources, 2015, 274, 922–927.
10. R. Kannan, H. N. Kagalwala, H. D. Chaudhari, U. K. Kharul, S. Kurungot and V. K. Pillai, Improved performance of phosphonated carbon nanotube–polybenzimidazole composite membranes in proton exchange membrane fuel cells, J. Mater. Chem., 2011, 21, 7223–7231.
11. M. Hazarika and T. Jana, Proton exchange membrane developed from novel blends of polybenzimidazole and poly(vinyl-1,2,4-triazole), ACS Appl. Mater. Interfaces, 2012, 4, 5256–5265.
12. I. Nicotera, V. Kosma, C. Simari, S. Angioni, P. Mustarelli and E. Quartarone, Ion dynamics and mechanical properties of sulfonated polybenzimidazole membranes for high-temperature proton exchange membrane fuel cells, J. Phys. Chem. C, 2015, 119, 9745–9753.
13. X. Li, H. Ma, H. Wang, S. Zhang, Z. Jiang, B. Liu and M. D. Guiver, Novel PA-doped polybenzimidazole membranes with high doping level, high proton conductivity and high stability for HT-PEMFCs, RSC Adv., 2015, 5, 53870–53873.
14. S. Wang, G. Zhang, M. Han, H. Li, Y. Zhang, J. Ni, W. Ma, M. Li, J. Wang, Z. Liu, L. Zhang and H. Na, Novel epoxy-based cross-linked polybenzimidazole for high temperature proton exchange membrane fuel cells, Int. J. Hydrogen Energy, 2011, 36, 8412–8421.
15. J. Wang, J. Zheng, Z. Zhao and S. Zhang, Synthesis and characterization of a novel poly(arylene ether sulfone) containing pendent imidazole groups for high temperature proton exchange membranes, J. Mater. Chem., 2012, 22, 22706–22712.
16. A. Bozkurt and W. H. Meyer, Proton-conducting poly(vinylpyrrolidone)–polyphosphoric acid blends, J. Polym. Sci., Part B: Polym. Phys., 2001, 39, 1987–1994.
17. H. Pu, Q. Liu, L. Qiao and Z. Yang, Studies on proton conductivity of acid doped polybenzimidazole/polyimide and polybenzimidazole/polyvinylpyrrolidone blends, Polym. Eng. Sci., 2005, 1395–1400.
18. Z. Wang and Z. Tang, Characterization of the polymer electrolyte based on the blend of poly(vinylidene fluoride-co-hexafluoropropylene) and poly(vinyl pyrrolidone) for lithium ion battery, Mater. Chem. Phys., 2003, 82, 16–20.
19. B. Smitha, S. Sridhar and A. A. Khan, Chitosan–poly(vinyl pyrrolidone) blends as membranes for direct methanol fuel cell applications, J. Power Sources, 2006, 159, 846–854.
20. Y. F. Huang, L. C. Chuang, A. M. Kannan and C. W. Lin, Proton-conducting membranes with high selectivity from cross-linked poly(vinyl alcohol) and poly(vinyl pyrrolidone) for direct methanol fuel cell applications, J. Power Sources, 2009, 186, 22–28.
21. X. Xu, H. Wang, S. Lu, Z. Guo, S. Rao, R. Xiu and Y. Xiang, A novel phosphoric acid doped poly(ethersulphone)–poly(vinylpyrrolidone) blend membrane for high-temperature proton exchange membrane fuel cells, J. Power Sources, 2015, 286, 458–463.
22. S. Lu, R. Xiu, X. Xu, D. Liang, H. Wang and Y. Xiang, Polytetrafluoroethylene (PTFE) reinforced poly(ethersulphone)–poly(vinyl pyrrolidone) composite membrane for high temperature proton exchange membrane fuel cells, J. Membr. Sci., 2014, 464, 1–7.
23. G. Maier and J. Meier-Haack, Sulfonated aromatic polymers for fuel cell membranes, Adv. Polym. Sci., 2008, 216, 1–62.
24. A. L. Rusanov, P. V. Kostoglodov, M. J. M. Abadie, V. Y. Voytekunas and D. Y. Likhachev, Proton-conducting polymers and membranes carrying phosphonic acid groups, Adv. Polym. Sci., 2008, 216, 125–155.
25. A. R. Khodabakhshi, S. S. Madaeni, T. W. Xu, L. Wub, C. Wub, C. Li, W. Nab, S. A. Zolanvari, A. Babayi, J. Ghasemi, S. M. Hosseini and A. Khaledi, Preparation, optimization and characterization of novel ion exchange membranes by blending of chemically modified PVDF and SPPO, Sep. Purif. Technol., 2012, 90, 10–21.
26. Y. Yang and D. M. Knauss, Poly(2,6-dimethyl-1,4-phenylene oxide)-b-poly(vinylbenzyltrimethylammonium) diblock copolymers for highly conductive anion exchange membranes, Macromolecules, 2015, 48, 4471–4480.
27. T. Xu, D. Wu and L. Wu, Poly(2,6-dimethyl-1,4-phenylene oxide) (PPO)—A versatile starting polymer for proton conductive membranes (PCMs), Prog. Polym. Sci., 2008, 33, 894–915.
28. I. Cabasso, J. J. Grodzinski and D. Vofs, Synthesis and characterization of polymers with pendent phosphonate groups, J. Appl. Polym. Sci., 1974, 18, 1969–1986.
29. S. Subianto, N. R. Choudhury and N. K. Dutta, Palladium-catalyzed phosphonation of SEBS block copolymer, J. Polym. Sci., Part A: Polym. Chem., 2008, 46, 5431–5441.
30. Y. Wu, C. Wu, T. Xu, X. Lin and Y. Fu, Novel silica/poly(2,6-dimethyl-1,4-phenylene oxide) hybrid anion-exchange membranes for alkaline fuel cells: Effect of heat treatment, J. Membr. Sci., 2009, 338, 51–60.
31. N. Y. A. Thabit, S. A. Ali and S. M. J. Zaidi, New highly phosphonated polysulfone membranes for PEM fuel cells, J. Membr. Sci., 2010, 360, 26–33.
32. V. Atanasov, D. Gudat, B. Ruffmann and J. Kerres, Highly phosphonated polypentafluorostyrene: Characterization and blends with polybenzimidazole, Eur. Polym. J., 2013, 49, 3977–3985.
33. X. Wang, X. Chen, A. Thomas, X. Fu and M. Antonietti, Metal-containing carbon nitride compounds: a new functional organic–metal hybrid material, Adv. Mater., 2009, 21, 1609–1612.
34. J. Liu, T. Zhang, Z. Wang, G. Dawson and W. Chen, Simple pyrolysis of urea into graphitic carbon nitride with recyclable adsorption and photocatalytic activity, J. Mater. Chem., 2011, 21, 14398–14401.
35. C. Han, L. Wu, L. Ge, Y. Li and Z. Zhao, AuPd bimetallic nanoparticles decorated graphitic carbon nitride for highly efficient reduction of water to H₂ under visible light irradiation, Carbon, 2015, 92, 31–40.
36. M. Gang, G. He, Z. Li, K. Cao, Z. Li, Y. Yin, H. Wu and Z. Jiang, Graphitic carbon nitride nanosheets/sulfonated poly(ether ether ketone) nanocomposite membrane for direct methanol fuel cell application, J. Membr. Sci., 2016, 507, 1–11.
37. L. Ge, C. Han, X. Xiao and L. Guo, Synthesis and characterization of composite visible light active photocatalysts MoS₂–g-C₃N₄ with enhanced hydrogen evolution activity, Int. J. Hydrogen Energy, 2013, 38, 6960–6969.
38. N. Zhang, B. Wang, C. Zhao, Y. Zhang, F. Bu, Y. Cui, X. Li and H. Na, Dual cross-linked organic–inorganic hybrid polymer electrolyte membranes based on quaternized poly(ether ether ketone) and (3-aminopropyl)triethoxysilane, J. Power Sources, 2015, 275, 815–822.
39. N. Fukuzaki, K. Nakabayashi, S. Nakazawa, S. Murata, T. Higashihara and M. Ueda, Highly phosphonated poly(N-phenylacrylamide) for proton exchange membranes, J. Polym. Sci., Part A: Polym. Chem., 2011, 49, 93–100.
40. K. D. Papadimitriou, A. K. Andreopoulou and J. K. Kallitsis, Phosphonated fully aromatic polyethers for PEMFCs applications, J. Polym. Sci., Part A: Polym. Chem., 2010, 48, 2817–2827.
41. A. Kumar, W. Pisula, D. Markova, M. Klapper and K. Müllen, Proton-conducting poly(phenylene oxide)–poly(vinyl benzyl phosphonic acid) block copolymers via atom transfer radical polymerization, Macromol. Chem. Phys., 2012, 213, 489–499.
42. D. Plackett, A. Siua, Q. Li, C. Pan, J. O. Jensen, S. F. Nielsenc, A. A. Permyakovab and N. J. Bjerrum, High-temperature proton exchange membranes based on polybenzimidazole and clay composites for fuel cells, J. Membr. Sci., 2011, 383, 78–87.
43. R. He, Q. Li, G. Xiao and N. J. Bjerrum, Proton conductivity of phosphoric acid doped polybenzimidazole and its composites with inorganic proton conductors, J. Membr. Sci., 2003, 226, 169–184.
44. Y.-L. Ma, J. S. Wainright, M. H. Litt and R. F. Savinell, Conductivity of PBI membranes for high-temperature polymer electrolyte fuel cells, J. Electrochem. Soc., 2004, 151, 8–16.
45. Q. Li, J. O. Jensena, R. F. Savinell and N. J. Bjerrum, High temperature proton exchange membranes based on polybenzimidazoles for fuel cells, Prog. Polym. Sci., 2009, 34, 449–477.
46. Y. Xu, H. Xu, L. Wang, J. Yan, H. Li, Y. Song, L. Huang and G. Cai, The CNT modified white C₃N₄ composite photocatalyst with enhanced visible-light response photoactivity, Dalton Trans., 2013, 42, 7604–7613.
47. A. Bozkurt and W. H. Meyer, Proton-conducting poly(vinylpyrrolidon)–polyphosphoric acid blends, J. Polym. Sci., Part B: Polym. Phys., 2001, 39, 1987–1994.
48. N. Zhang, B. Wang, Y. Zhang, F. Bu, Y. Cui, X. Li, C. Zhao and H. Na, Mechanically reinforced phosphoric acid doped quaternized poly(ether ether ketone) membranes via cross-linking with functionalized graphene oxide, Chem. Commun., 2014, 50, 15381–15384.
49. Z. Shao, A. Sannigrahi and P. Jannasch, Poly(tetrafluorostyrenephosphonic acid)–polysulfone block copolymers and membranes, J. Polym. Sci., Part A: Polym. Chem., 2013, 51, 4657–4666.
50. A. Carollo, E. Quartarone, C. Tomasi, P. Mustarelli, F. Belotti, A. Magistris, F. Maestroni, M. Parachini, L. Garlaschelli and P. P. Righetti, Developments of new proton conducting membranes based on different polybenzimidazole structures for fuel cells applications, J. Power Sources, 2006, 160, 175–180.
51. J. T. Wang, X. J. Yue, Z. Z. Zhang, Z. Yang, Y. F. Li, H. Zhang, X. L. Yang, H. Wu and Z. Y. Jiang, Enhancement of proton conduction at low humidity by incorporating imidazole microcapsules into polymer electrolyte membranes, Adv. Funct. Mater., 2012, 22, 4539–4546.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra17243a

---