Proton exchange properties of flexible diamine-based new fluorinated sulfonated polyimides

Abstract

A series of new semifluorinated sulfonated polyimide copolymers (DHNHXX) were prepared from 1,4,5,8-naphthalenetetracarboxylic dianhydride (NTDA), 4,4′-diaminostilbene-2,2′-disulfonic acid (DSDSA) and 1,4-bis-[{2′-trifluoromethyl-4′-(4′′-aminophenyl)phenoxy}] benzene (HQA). The degree of sulfonation of DHNHXX was varied by changing the molar ratio of the sulfonated diamine, DSDSA to the non-sulfonated diamine, HQA. The flexible DHNHXX membranes showed high thermal stability (desulfonation temperature above 270 °C) and good mechanical properties. The oxidative stability of DHNHXX improved with increasing the fluorine content. DHNHXX showed anisotropic dimensional changes, and good water stability (did not dissolve for up to 415 h), higher than many other sulfonated polyimides. Transmission electron microscopy (TEM) analysis revealed cluster-like microstructures for the DHNHXX membranes, suggesting good phase-separated morphology. DHNH70 with IEC = 2.10 meq. g⁻¹ exhibited high proton conductivity, with a maximum up to 129 mS cm⁻¹ at 90 °C in water.

---

1. Introduction

Proton exchange membrane fuel cells (PEMFCs) have drawn much attention as promising power sources for automotive vehicles and in many portable devices due to their high efficiency and low emissions.^1–3 Proton conductivity, thermal, mechanical and chemical stability, and fuel or oxidant permeability, as well as the cost of proton exchange membranes (PEMs), are the parameters that influence the performance and commercialisation of fuel cells. Low conductivity at high temperature (temperature > 80 °C) and high cost are the limitation of the current state of the art Nafion^® membranes.⁴ This has encouraged the synthesis of new PEM materials. In this account, sulfonated polyimides (SPIs) have been developed as promising PEM materials mainly because of their excellent thermal stabilities, high mechanical strength and modulus, substantial water and chemical stability as well as good film forming ability.^5,6 Almost all SPIs are generally synthesized through the direct copolymerization of the sulfonated monomers to avoid the possible repercussions of imide hydrolysis in the post-polymerization acidification process. Moreover, the chemical and physical properties could be easily tuned by simply changing the composition of the non-sulfonated diamine comonomers in the direct copolymerization route.

It has been observed that the sulfonated six-membered ring (naphthalenic) polyimides exhibit superior chemical and thermal stabilities and mechanical strength over analogous five-membered ring polyimides because of their strain-free ring structures.^7,8 Mercier and coworkers⁵ first synthesized various sulfonated copolyimides from naphthalene-1,4,5,8-tetracarboxylic dianhydride (NTDA), 2,2′-bendizine sulfonic acid (BDSA) and common nonsulfonated diamine monomers. However, these SPIs from BDSA exhibited poor water stability. Consequently, Okamoto’s group synthesized SPIs using relatively flexible sulfonated diamines such as 4,4′-diaminodiphenyl ether-2,2′-disulfonic acid (ODADS), 4,4′-bis(4-aminophenoxy)biphenyl-3,3′-disulfonic acid (BAPBDS) and various non-sulfonated diamines which showed much better water stability at comparable ion exchange capacity (IEC) values.^9,10 Similarly, McGrath et al.¹¹ also reported the synthesis of SPIs with improved water stability based on a novel sulfonated diamine, 3,3′-disulfonic acid-bis[4-(3-aminophenoxy)phenyl]sulfone (SA-DADPS), containing flexible sulfone and ether linkages. Both groups independently observed that increasing the flexibility of the backbone structure of SPIs resulted in enhanced water stability. Okamoto and his coworkers also reported 2,2′-bis(3-sulfopropoxy)benzidine (2,2′-BSPB), 3,3′-bis(3-sulfopropoxy)benzidine (3,3′-BSPB) and 2,2′-bis (4-sulfobutoxy)benzidine (2,2′-BSBB)-based side chain-type SPIs which showed higher proton conductivities and better water stability than main-chain type SPIs.^12,13 They attributed the improvement in stability and conductivity to the relatively higher basicity of sulfonated diamines and the well-separated micro-phase structures, respectively. However, the proton conductivity of the side chain-type SPIs bearing pendant sulfoalkoxy groups decreased largely at high temperature and low humidity due to the relatively easy cleavage of the sulfoalkoxy groups.¹⁴ It has also been reported that the structure of non-sulfonated diamines plays a vital role in tuning the solubility, water uptake, proton conductivity and the water stability of SPIs along with the structure of sulfonated diamines. Litt’s group^15,16 has reported the influence of inherently developed chain spacing due to some angled and/or bulky non-sulfonated diamine moieties in SPIs in improving the dimensional stability and proton conductivity, particularly at high temperature and low relative humidity (RH). Miyatake et al.¹⁷ and Sun et al.¹⁸ also reported the improvement of the oxidative stability due to the hydrophobic trifluoromethyl group in the SPIs along with interchain spaces for water molecules in aiding the proton transport mechanism.

In a previous article,⁸ we reported fluorinated SPIs, DQNXX {for NTDA–DSDSA/QA(m/(1¬m))} from the commercially available sulfonated diamine, 4,4′-diaminostilbene-2,2′-disulfonic acid (DSDSA) and non-sulfonated diamine, 4,4′-bis [3′-trifluoromethyl-4′(4-aminobenzoxy)benzyl]biphenyl (QA), bearing trifluoromethyl (–CF₃) groups. The DQNXX membranes exhibited high proton conductivity, good phase-separated morphology, low water uptake and improved water stability. However, it is felt necessary for further improvement of the PEM properties of the DQNXX membranes with a high degree of sulfonation (≤80 mol% of disulfonation). Hence, more stable polymer systems need to be explored. Consequently, in continuation of our research into PEMs, an attempt has been made to increase the proton conductivity at a low degree of sulfonation through the use of a low molecular weight and comparatively flexible fluorinated diamine, 1,4-bis-[{2′-trifluoromethyl 4′-(4′′-aminophenyl)phenoxy}] benzene (HQA) in place of QA in DQNXX. Accordingly, this work describes the synthesis and characterization of new SPIs, DHNHXX based on HQA. The properties of the synthesized materials, such as their thermal and oxidative stability, mechanical strength, water uptake behavior and proton conductivity have been investigated in detail and compared with other polymers.

2. Experimental

2.1. Materials

NTDA was purchased from Aldrich (USA) and DSDSA was purchased from Alfa Aesar (India) and were both dried overnight under vacuum at 120 °C prior to use. The detailed synthesis of 1,4-bis-[{2′-trifluromethyl 4′-(4′′-aminophenyl)phenoxy}] benzene (HQA) has been reported previously.¹⁹ Triethylamine (TEA, 99%), m-cresol (99%) and benzoic acid (BA, >99.5%) were purchased from E. Merck (India) and were used as received. N,N-Dimethylacetamide (DMAc) (Spectrochem, India) was used as a solvent for membrane casting. Nafion^® 117 membrane was purchased from Alfa Aesar and was treated as follows: the as-received membrane was first placed in hot 3 vol% H₂O₂ for 1 h and was rinsed in deionized water several times to remove the H₂O₂. Subsequently, the membrane was immersed in boiling 0.5 M H₂SO₄ solution for 1 h, followed by rinsing in deionized water several times to remove excess H₂SO₄. The membrane was finally stored in deionized water at room temperature.

2.2. Synthesis of DHNHXX polymers

DHNHXX polymers with different molar proportions of disulfonation were synthesised by a one step high temperature polycondensation reaction in m-cresol, as presented in Scheme 1. The mole percentage of DSDSA used in the polymerization of a particular polymer is indicated by XX. The salt (TEA) form of the polymer is represented by DHNXX, whereas the acid (H) form of the polymer is represented by DHNHXX. A representative procedure for the preparation of DHN50 {NTDA–DSDSA/HQA (50 : 50)} is described as an example. First, 0.340244 g (0.0918 mmol) of DSDSA, 10 mL of m-cresol and 0.2044 g (0.28 mL) of TEA were added into a dried 250 mL, three-necked round-bottomed flask equipped with a condenser and N₂ inlet. The mixture was then heated to 80 °C and stirred under N₂ until DSDSA was dissolved. Thereafter, 0.533257 g (0.0918 mmol) of HQA, 0.492692 g (0.1837 mmol) of NTDA and 0.44871 g of BA were added successively, followed by the addition of ∼10 mL of m-cresol. The mixture was stirred at 80 °C for 4 h, at 180 °C for 16 h and at 200 °C for 3 h, respectively. After cooling to room temperature, 3 mL of m-cresol was added to dilute the highly viscous solution. After completion of the reaction, the viscous polymer solution was slowly poured into excess isopropyl alcohol under constant stirring. The fibrous precipitate was collected by filtration. The fibrous polymer was further washed thoroughly by methanol to remove any residual solvent and impurities. Finally, fibre-like polymer was collected after drying at 120 °C under vacuum overnight. Films were prepared from the polymer solutions in DMAC (15% w/v) by pouring them in to Petri dishes and heating sequentially at different temperature (at 80 °C (12 h), 100 °C, 120 °C, 140 °C and 160 °C each for 2 h) for slow removal of the solvent. The acidification of the membranes was performed by immersing them in 1.5 M H₂SO₄ at room temperature for three days. Except for DHNH40, transparent, light brown and flexible membranes were obtained from the polymers. The thicknesses of the membranes (in the range ∼40–60 μm) were measured as the average of the ten different measurements of the same membrane sample using a Mitutoyo Digimatic micrometer with an accuracy of 0.001 mm, and in each case the standard deviation was within 3 μm of the average thickness of the membrane.

Scheme 1 Synthesis and structure of the sulfonated polyimides, DHNHXX.

NMR data of the acid form: ¹H-NMR (DMSO-d₆): δ (ppm) 8.83–8.61 (H⁵, H⁶); 8.30 (H¹); 8.13–8.07 (H⁹, H¹⁰); 7.92–7.83 (H², H³, H⁸); 7.59–7.50 (H⁴, H⁷); 7.38–7.28 (H¹¹, H¹²).

2.3. Polymer analysis and measurements

The ¹H NMR spectrum of DHNHXX was recorded on a Bruker 400 MHz instrument (Switzerland), using DMSO-d₆ as the solvent and internal reference (δ(¹H) = 2.50 ppm). FTIR spectra of DHNHXX were recorded on a Bruker (TENSOR 27) spectrophotometer using KBr pellets at room temperature and a humidity-free atmosphere. Differential scanning calorimetry (DSC) measurements were measured using a NETZSCH DSC 200PC instrument at a heating rate of 10 °C min⁻¹. Thermogravimetric (TGA) measurements were performed on a TGA Q5000 from TA Instruments at a heating rate of 10 °C min⁻¹ to determine the decomposition temperature under synthetic air. Stress–strain behavior of the thin polymer films (10 mm × 25 mm) was measured (in both dry and wet conditions) using UTM-Instron, Plus-8800 at a cross-head speed of 5% of the specimen length per minute at room temperature (∼30 °C) and 60% RH. Transmission electron microscopy (TEM) was undertaken of the ultramicrotome membranes using a TEM instrument (FEI-TECNAI G2 20-TWIN, USA). In order to stain the ionic domains, DHNHXX membranes were converted into the Ag⁺ form by immersing them in 0.5 M AgNO₃ aqueous solution overnight. The samples were thoroughly rinsed with deionised water and dried at room temperature for 24 h. Thin slices (200 nm) of the embedded polymer samples were cut with a cryo-ultramicrotome using a Leica Ultracut UCT Leica EM FCS, Austria, fitted with freshly sharpened glass knives held at a cutting edge of 45 degrees, and were transferred onto carbon-coated copper grids for TEM analysis.

The ion exchange capacities (IECs) of the DHNHXX membranes were determined by the back-titration method as reported elsewhere.²⁰ The membrane in acid form (0.2 g) was converted to the sodium form by immersing the membrane in 50 mL of 1.0 M NaCl aqueous solution for 48 h. The released proton was then titrated with standard 0.01 M NaOH aqueous solution using phenolphthalein as the indicator. The water absorption and dimensional stability of the membranes were tested at the desired temperature from the water uptake ratio and swelling ratio, which was defined as the change in weight and dimension (diameter or thickness) of the membranes after soaking in water for 5 h divided by the weight and dimensions of the dry samples. The oxidative stability of the DHNHXX membranes was evaluated by recording the elapsed time in which the membranes (10 mm × 10 mm) started to break into pieces (τ₁) and dissolved completely (τ₂) under occasional stirring after immersion in Fenton's reagent (2 ppm FeSO₄ in 3% H₂O₂) at 80 °C. The water stability was investigated by measuring the elapsed time in which a membrane lost its mechanical properties (brittle or bending at the corners) after its immersion into water at 80 °C.

The proton conductivity (σ, S cm⁻¹) in the plane direction of the DHNHXX membranes (1 cm × 2 cm) was calculated using σ = l/wdR (l (= 0.8 cm) is the distance between the electrodes, and w and d are the width and thickness of the membrane, respectively). The resistance value (R) was measured over the frequency range of 100 Hz to 2 MHz by the two-point probe method using AC impedance spectroscopy (HIOKI 3532-50 LCR Hi-TESTER). The prehydrated membranes were clamped across the pair of platinum electrodes in a home-made Teflon conductivity cell, as reported elsewhere.^20,21 Conductivity measurements under fully hydrated conditions were then carried out with the cell immersed in deionized water by increasing the temperature from 30 to 90 °C at a heating rate of 1–2 °C min⁻¹.

3. Results and discussions

3.1. DHNHXX polymer synthesis and characterisation

The DHNHXX polymer was prepared from NTDA, DSDSA and a semifluorinated nonsulfonated diamine (HQA) with different degrees of sulfonation (DS) through a one-step high temperature polycondensation reaction in m-cresol (Scheme 1). The preparation of DHNHXX with XX below 40 was not successful, while DHNH40 also formed with some particles (probable precipitation in the form of low molecular weight chain polymers (oligomers)) during the polymerization reaction. The preparation of DHNH80 (towards higher DS) was also tried. However, DHNH80 gave an almost insoluble mass after precipitation from the highly viscous reaction mixture, and could be partially dissolved even after heating at 60 °C for a prolonged time. Thus the preparation and study of higher DS (XX > 70) polymers was abandoned thereafter. In the reaction, BA was used as a catalyst, while TEA was employed to improve the solubility of DSDSA and liberate free (non-zwitterionic) amine groups for polymerization with NTDA. After the initial complete dissolution of DSDSA in the presence of TEA in m-cresol at 80 °C, measured quantities of HQA, NTDA and BA were charged successively into the reaction flask with a few mL of m-cresol and heated at 180–200 °C for ∼19 h for polymerisation into high molar masses. The inherent viscosity of the DHNHXX polymers was evaluated from their solution in N-methyl-2-pyrrolidone (NMP) (0.5 g dL⁻¹), and the values were in the range of 1.80–3.07 dL g⁻¹. The high viscosity values for DHNHXX indicated the formation of high molar masses. The polymerization compositions and properties of the polymers are summarized in Table 1.

Table 1 Composition and properties of the DHNHXX polymers

Polymer	DSDSA (mol%)	ηinh^a (dL g^−1)	DS^b
			Theoretical	NMR
a Inherent viscosity of the DHNHXX polymers in NMP at 30 °C; polymer concentration 0.5 g dL^−1.b Degree of sulfonation: the theoretical value is calculated from the monomer feed ratio; the NMR value corresponds to the experimentally determined value from the content of the DSDSA moiety in the polymer.
DHNH40	40	1.48	0.4	0.38
DHNH50	50	2.17	0.5	0.46
DHNH60	60	1.88	0.6	0.60
DHNH70	70	1.97	0.7	0.66

---

The solubility behavior of the polymers (both in the TEA salt and acid form) at room temperature was tested (as 10% (w/v)) in various common organic solvents, and the results are shown in Table 2. The DHNHXX (acid form) polymers were soluble in polar aprotic solvents such as DMSO, DMAc, DMF, NMP and pyridine, but insoluble in THF or acetone. The DHNXX (TEA salt form) polymers with XX > 40 showed good solubility in m-cresol, but not in their proton form. This might be due to the intensive polymer chain packing in their proton form. Though the polymers were prepared in m-cresol, after isolation by precipitation and drying, DHNXX with XX > 70 (i.e. DHN80) was mostly in the form of an insoluble mass or was sparingly soluble in m-cresol. Similarly, DHNXX with XX < 50 was not completely soluble and formed a suspension of particles. This behavior can be explained on the basis of π–π interactions of planar NTDA moieties.²² Increasing the content of the ionic DSDSA comonomer above XX = 40 improves the solubility of DHNXX. However, after reaching XX = 70, a further increase of the ionic DSDSA moieties might have resulted in higher ordering in the microstructure which complementarily facilitated the increased π–π interactions of planar NTDA moieties, ultimately leading to aggregates. The overall good solubility of DHNHXX, XX = 40–70, was due to the combined effects of flexible vinyl and ether (–O–) linkages which form a different intrasegmental configuration, and the bulky trifluoromethyl (–CF₃) groups in the backbone, which disrupt the regularity of the molecular chains and hinder the dense chain packing.

Table 2 Solubility behavior of the DHNHXX polymers^a

Polymer	NMP	DMAc	DMF	DMSO	Py	THF	m-Cresol
a Data in parentheses refer to the TEA salt form, +: soluble, ±: partially soluble, − : insoluble.
DHNH40	±(±)	±(±)	±(±)	±(±)	±(±)	−(−)	±(±)
DHNH50	+(+)	+(+)	+(+)	+(+)	+(+)	−(−)	+(+)
DHNH60	+(+)	+(+)	+(+)	+(+)	+(+)	−(−)	−(+)
DHNH70	+(+)	±(+)	±(+)	+(+)	+(+)	−(−)	−(±)

---

The chemical structure of DHNHXX was confirmed by both FTIR and NMR spectroscopy. Fig. 1 shows the FTIR spectra of the DHNHXX polymers as a function of DS. The spectra showed strong absorption bands at around 1712 cm⁻¹ and 1674 cm⁻¹ due to the stretching vibrations of the carbonyl groups of the naphthalimide ring, while the band at around 1349 cm⁻¹ appeared due to the C–N asymmetric stretching. These bands clearly indicate the formation of imide rings. The bands at 1024 cm⁻¹ and 1081 cm⁻¹ were due to the symmetric and asymmetric stretching of the sulfonic acid groups. The intensities of the peaks at 1024 cm⁻¹ and 1081 cm⁻¹ increased with the degree of disulfonation. The representative ¹H NMR spectrum of DHNH60 is presented in Fig. 2. The spectral signals can be assigned to all of the magnetically different protons of the polymer repeat unit. No residual amine, amide or carboxylic protons were observed, indicating the imidisation reaction was complete. After acid treatment, the peak related to triethylammonium protons at δ (ppm) = 1.14 and 3.07 disappeared. Extra peaks due to free aromatic amine or free carboxylic acid in moiety were not found. This confirmed that the salt form of the SPI copolymer was successfully converted into the corresponding acid form without the degradation of the polymer backbone. The ¹H NMR spectra were also used to calculate the copolymer composition, similar to the method used in our previous article²⁰ (Table 1). The area under the broad signal due to the protons of the naphthalimide moiety (δ (ppm) = 8.83–8.61) was taken as the integral value = 1 and was used for the integration of the remaining signals. The copolymer composition calculated from the integrals of the signal regions I–III (Fig. 2) was in good agreement with the feed ratio.

Fig. 1 FTIR spectra of DHNHXX polymers.

Fig. 2 ¹H NMR spectrum of DHNH60.

3.2. Thermal properties, mechanical properties, oxidative stability and water stability

The thermal stability of the DHNHXX membranes was investigated by thermogravimetric analysis (TGA) measurements in the range of room temperature–800 °C at a heating rate of 10 °C min⁻¹ in synthetic air. Fig. 3 shows the thermal decomposition curve of DHNHXX as a function of the degree of sulfonation. The DHNHXX polymers showed a common three step degradation pattern similar to many reported articles. The initial weight loss at around 100 °C was due to the loss of absorbed water. The second weight loss which started at around 270–295 °C was due to the decomposition of sulfonic acid moieties. The third weight loss which started at around 510–535 °C was due to the pyrolysis of the polymer backbone. These decomposition patterns are well in accordance with the literature.⁸ With increasing DS, the weight losses at the same temperature increased for the polymers in the series. The thermograms clearly indicates that the DHNHXX polymers possess fairly good thermal stability as their decomposition temperature (2^nd onset decomposition temperature due to the dissociation of –SO₃H groups) is well above the application temperature of PEMs in fuel cells. It is also noted that the DSC plot of the DHNHXX showed no clear signature of melting or glass transition. This can be asserted to high rigidity due to the strong ionic interaction between the ionic groups, and the presence of the naphthalenic and aromatic backbone structure. Hence, we can presume that the decomposition of the DHNHXX polymers started at temperatures (desulfonation at around 270–295 °C in the TGA) below their glass transition temperature.

Fig. 3 TGA thermogram of the DHNHXX polymers.

The mechanical properties of the DHNHXX membranes were measured in both dry and wet conditions. For the wet conditions, the membranes were immersed in DI water at room temperature for 24 h prior to the measurements. The results are summarised in Table 3. The reason for the low value of the tensile strength for the DHNH40 membrane is viewed with scepticism, and might be attributed to the low quality of the film (opaque due to particle formation in the solution) and the low molecular weight of the formed polymers due to their phase-out (precipitation) during the polymerisation reaction. However, it should be noted that the remaining membranes showed comparable tensile strengths and much larger Young's moduli compared to many other reported six-membered SPIs.^18,23 In the dry state the membranes showed tensile strengths in the range of 66–88 MPa and Young's moduli in the range of 1.74–2.36 GPa. The comparatively high elongation at break (14–49%) of the DHNHXX polymers with respect to the polymers reported by Sun et al.¹⁸ (Table 3), despite having similar non-sulfonated segments of the polymer structure, might be due to the flexible vinyl linkage of the sulfonated diamine, DSDSA, used in the preparation of DHNHXX. As is commonly experienced in aromatic PEMs, the elongation at break for the DHNHXX membranes (Fig. 4a) were much lower than that of the Nafion^® 117 membrane (288%). In wet conditions (Fig. 4b), the membranes also showed good mechanical properties with tensile strengths of 28–48 MPa, elongation of 4–23% and Young's moduli of 1.20–1.56 GPa. Similar to the regular observed trend in PEMs, the tensile strength of the DHNHXX membranes decreased with increasing DS. It was also observed that the Young's moduli of the membranes did not show a regular trend with the increasing degree of sulfonation.

Table 3 Mechanical properties of DHNHXX and Nafion^® 117 membranes

Polymer	TS^a (MPa)		Y^b (GPa)		EB^c (%)		Reference
	Dry	Wet	Dry	Wet	Dry	Wet
a Tensile strength at a strain rate of 5% min^−1, 65 ± 2% RH and 30 °C.b Young's modulus.c Elongation at break.
DHNH40	44	28	1.23	1.20	6.6	3.6	This paper
DHNH50	88	73	1.74	1.55	48.5	23.1	This paper
DHNH60	68	61	2.36	1.56	20.8	12.5	This paper
DHNH70	66	48	2.17	1.38	13.7	7.8	This paper
NTDA–BABSA/6FBAB (3 : 1)	73	—	1.85	—	10.3	—	18
NTDA–BABSA/6FAPB (3 : 1)	75	—	2.06	—	9.5	—	18
NTDA–BABSA/BAPB (3 : 1)	70	—	1.89	—	9.7	—	18
Nafion^® 117	22	ND	0.16	ND	288	ND	This paper

---

Fig. 4 Tensile properties of the DHNHXX polymers in the (a) dry state and (b) wet state.

The oxidative and water stability of the PEMs are critical aspects as the fuel cell operates under humidified and oxidative conditions. As shown in Table 4, the DHNHXX membranes showed fairly good oxidative stability which decreased with the increasing degree of sulfonation. Thus, as expected, the DHNH40 membrane showed the highest oxidative stability (τ₂ = 4.6 h) and DHNH70 showed the lowest oxidative stability (τ₂ = 2.9 h) in the series. Nevertheless, these oxidative stability values are better than those of many other reported sulfonated polyimides.^18,24 The high oxidative stability of the DHNHXX membranes is due to the presence of hydrophobic trifluoromethyl groups which could protect the polymer main chain from being attacked by water molecules containing highly oxidizing radical species.^13,17 In addition, although displaying a relatively high fluorine content, the DHNHXX membranes displayed lower oxidative stability in comparison to our previously reported DQNXX membranes.⁸ This might be due to the comparatively slightly higher water absorption due to the higher IEC of the DHNHXX membranes (relatively lower molecular weight of HQA with respect to QA in the co-SPIs) and the relatively greater flexible backbone structure resulting in higher entanglement and more free space.

Table 4 Oxidative stability and water stability of the DHNHXX and other reported SPI membranes

Polymer	IECW,Theo. (meq. g^−1)^a	Thickness (μm)	Oxidative stability^b (h)		Water stability^c (h)	Reference
			τ1	τ2
a IECW,Theo. = (1000/MWrepeat unit) × DSTheo. × 2, where DSTheo. is calculated from the monomer feed ratio.b The time expended for the membrane to begin to break (τ1), or the complete dissolution (τ2) in Fenton's reagent (2 ppm FeSO4 in 3% H2O2) at 80 °C.c Measured as the time in which the membrane lost its mechanical properties in hot water (80 °C).d Completely dissolved.e Started to dissolve.f Reported for Fenton's reagent (30 ppm FeSO4 in 30% H2O2) at 30 °C.g Measured at 100 °C.
DHNH40	1.10	121	4.6	20.5	>450	This paper
DHNH50	1.41	61	3.9	14.7	>450	This paper
DHNH60	1.75	61	3.4	11.2	>450	This paper
DHNH70	2.10	39	2.9	6.3	∼417	This paper
DQN60	1.67	51	3	20	>400	8
DQN70	2.03	45	1.2	9	>400	8
DQN80	2.42	50	0.8	4.5	∼120^d	8
NTDA–BAPBPDS/ODA (2 : 1)	1.99	27	0.95	1.35^e	390	24
NTDA–BAPBPDS/BAPB (3 : 1)	2.02	40	1.42	1.33^e	360	24
NTDA–BABSA/6FBAB (3 : 1)	1.87	50–60	47^f	76^f	191^g	18
NTDA–BABSA/6FAPB (3 : 1)	1.91	50–60	37^f	47^f	237^g	18
NTDA–BABSA/BAPB (3 : 1)	1.95	50–60	29^f	49^f	290^g	18

---

The water stability of the DHNHXX membranes was studied in de-ionised water at 80 °C. The stability was characterized by the elapsed time by which the hydrated membranes lost their mechanical strength. The results of the water stability are listed in Table 4 and compared with the results of other SPIs. The DHNH70 membrane with the highest IEC in the series exhibited fair water stability, as it did not dissolve until the membrane was soaked in hot water for ∼415 h in relation to that of our previously reported DQN80 membrane, which dissolved in 120 h, and various other SPIs.^8,18,24 This suggested that in addition to the validated profound effect of various sulfonated diamines, the structure of the non-sulfonated diamine can influence the water stability of SPI membranes. According to the general trend, for a particular series of polymers with varying DS, the stability of the membranes should decrease with the increase of the IEC (high water uptake). However, it is noticed that despite having higher IEC and water uptake, the DHNHXX membranes showed comparable water stability over DQNXX. It has been reported that a flexible chain can undergo relaxation more easily than rigid chains.⁶ Thus, the non-sulfonated diamine HQA with three benzene rings linked with ether bonds imparted greater flexibility to DHNHXX in comparison to the flexibility contribution of QA for DQNXX. Thus, the adverse effect of high IEC by high water uptake for DHNHXX (in comparison to DQNXX) was offset by the favorable effect of its flexible structure. It may be further noted that the increased fluorine content in DHNHXX compared to DQNXX did not bring significant change in the water stability (the DHNH70 membrane lost mechanical strength after 415 h). This might be due to the selective increased hydrophobicity of the non-sulfonated segment (due to the higher fluorine content) of the DHNHXX, while the hydrolysis degradation would generally take place in the sulfonated segment of the copolymer structure. In addition to the IEC and flexibility of SPIs, the basicity of the sulfonated and non-sulfonated diamine moieties greatly influences their water stability. The water stability is mainly determined by the electronic density of imido rings which is strongly dependent on the basicity of diamine moieties as previously mentioned. The greater the positive character of the imido carbonyl, the poorer the water stability (i.e. faster loss of mechanical strength). This indicates that in the present comparison (up to 450 h testing) between DHNHXX and DQNXX polymers, the effect of the basicity of the non-sulfonated diamine was found to be redundant, whereby both systems used the same common sulfonated diamine, DSDSA.

3.3. IEC, microstructure, water uptake and dimensional change

The IEC_W values of the acid-form (DHNHXX) membranes were determined by acid–base titration and are listed in Table 5. The experimental IEC_W values were in the range from 1.02 to 1.98 meq. g⁻¹, close to the theoretical data calculated from the monomer feed ratios and the IEC_W value calculated from the NMR composition data. This indicates complete exchange with protons after the acidification process of the TEA salt form (DHNXX) of the membrane. As seen from the IEC_W values listed in Table 5, it is obvious that both the experimental IEC_W values and IEC_W,NMR values were a little lower than the IEC_W,Theo. values. This might be due to incomplete drying of the membranes and/or incomplete exchange of the H⁺ ions with Na⁺ ions, while the error in determination of the signal integrals in the ¹H NMR spectra may also contribute. Nevertheless, the washing away of sulfonated oligomers during the purification process is a further reason for the lower value of IEC_W in comparison to IEC_W,Theo..

Table 5 Proton exchange membrane properties of the polymer membranes

Polymer	IECW (meq. g^−1)			Dimensional change		WUW^b (%)	λ^c [H2O/SO3^−]	σ (mS cm^−1)			Reference
	Theo.	Titr.	NMR^a	Δl	Δt			30 °C	80 °C	90 °C
a IECW,NMR = (1000/MWrepeat unit) × DSNMR × 2, where DSNMR corresponds to the molar content of DSDSA units in the polymer structure determined from the ^1H NMR spectra.b WUW (%) = [(Wwet − Wdry)/Wdry × 100], where Wwet and Wdry are the weights of the wet and dry membranes, respectively (measurements were at 30 °C).c λ = WUW (%)/(100 × IECW,Theo. × MW,H2O), where MW,H2O = 18 g mol^−1.
DHNH40	1.10	1.02	1.04	0.01	0.03	15.23	7.71	5	10	12	This paper
DHNH50	1.41	1.32	1.30	0.02	0.06	20.57	8.09	10	28	31	This paper
DHNH60	1.75	1.63	1.75	0.04	0.17	26.29	8.36	28	60	66	This paper
DHNH70	2.10	1.98	1.98	0.05	0.20	33.61	8.88	46	99	129	This paper
DQN70	2.03	1.91	ND	0.02	0.11	29.73	8.2	23	40	ND	8
DQN80	2.42	2.31	ND	0.03	0.12	35.61	8.2	55	81	ND	8
Nafion^® 117	0.91	0.90	NA	0.11	0.19	19	11.4	60	135	150	This paper

---

It is well known that the phase-separated morphology holds a close relation with the proton conductivity of the PEMs. Hence, the microstructure of the DHNHXX membranes stained with Ag⁺ was investigated by TEM. Thus, the dark regions, as shown in Fig. 5, refer to hydrophilic ionic clusters (due to the aggregation of sulfonated groups) and the brighter regions correspond to hydrophobic segments, giving a clear indication of the phase-separated morphology of the DHNHXX membranes. This morphological feature is in contrast to the homogeneous nature without clear phase-separated morphology reported for many other main chain-type SPIs.⁹ From the TEM micrograph it is clearly observed that the DHNH50 membrane consists of ionic clusters in a uniform distribution with average sizes of 27–30 nm. When the DS was 0.6 (DHNH60), it showed uniform distribution of a large amount of medium-sized ionic clusters (33–40 nm), along with a certain amount of smaller (20–25 nm) and larger (40–45 nm) ionic clusters. On increasing the DS further to 0.7 (DHNH70), a largely aggregated cluster morphology was observed with widely varied cluster sizes, with large amounts in the range 35–40 nm with certain small size clusters (17–24 nm) and larger ionic clusters (50–55 nm). These widely varied sizes of ionic clusters in the TEM micrograph was quite similar to many reported in the literature.²⁵ At low DS (DHNH50), hydrophilic ionic clusters were exhibited as isolated domains in the system. However, with an increase in DS (DHNHXX, XX ≥ 60), the individual isolated clusters come close to each other and form clearer connectivity with clusters in close proximity for better proton transport. Although the phase-separated morphology was observed in both the DHNHXX polymers and our previously reported DQNXX polymers,⁸ the distribution of hydrophilic domains (ionic clusters) was significantly better in the former compared to the latter. Also, the variation in sizes of the clusters was very wide in the DQNXX polymers. If the above mentioned observation is true, since the size of the clusters varies in the presence of absorbed water (under different humid conditions), it might be considered that the increased flexibility and the proportion of hydrophobic trifluoromethyl groups in the nonsulfonated diamine moieties of the DHNHXX polymers would result in relatively increased uniformity in the sizes and distribution of clusters in the hydrophobic domains. It is also important to note that the microstructure obtained was similar to the reported microstructure of side chain-type SPIs,^18,25 where well phase-separated morphology is generally expected. It is anticipated that the presence of pendent hydrophobic trifluoromethyl groups rendered the formation of this kind of phase-separated morphology responsible for improved proton conductivity.

Fig. 5 TEM micrograph of Ag⁺-stained DHNHXX membranes.

In keeping with the common trend of PEMs, water uptake for the DHNHXX membranes gradually increased with the increasing IEC_W values (Table 5) and was in the range of 15–34%. Nevertheless, the maximum water uptake value (34%) for the DHNHXX polymers with the highest DS (DHNH70) was rather low as compared to many other reported fluorinated and non-fluorinated SPIs,^6,18 and the value further decreased with an increase in the fluorine content (hydrophobic trifluoromethyl groups) in the DHNHXX polymer series. Hence, the presence of trifluoromethyl groups in the polymer structure can be envisaged as hindrance to excessive dimensional changes under humidified conditions, while helping to maintain acceptable mechanical properties for the membranes. Again, the water uptake values among the SPI membranes with different IEC_w values is often evaluated in terms of the number of water molecules absorbed per sulfonic acid group or the hydration number (λ).²⁶ In this study, the λ values were calculated using the theoretical IEC_w values which were calculated from the molar ratio in the feed. Although the WU values increased with increasing IEC_w, λ remained more or less constant after reaching a certain value (λ ∼ 8), as shown in Table 5. This may be attributed to the higher rigidity of the aromatic SPI chains and the strong ionic interaction among the sulfonic acid groups, which restricted the free volume for water absorption beyond a certain limit.

Table 5 lists the water swelling values for the DHNHXX membranes, as they showed anisotropic character due to the relatively larger dimensional change in thickness than in the plane direction. The anisotropic degree of membrane swelling, Δt/Δl, was in the range of 2.8–4.6, indicating significant anisotropic membrane swelling. It is expected that the lower dimensional change in the plane direction of the membrane is favorable for the high quality fabrication of membrane electrode assemblies for application under the humid conditions of fuel cells. The anisotropic membrane swelling of DHNHXX membranes is considered to be due to the favourable orientation of polymer chains in the plane direction. The rigid structure of the DHNHXX SPIs seemed to facilitate the alignment in the plane direction. The increased swelling (in both the plane and thickness directions) of the DHNHXX membranes in comparison to the DQNXX membranes might be due to the relatively reduced rigidity of the backbone structure and the higher probability of entanglement resulting from the flexible phenyl ether linkages of the HQA moieties in the co-SPIs.

3.4. Proton conductivity

Proton conductivity is the property of prime importance for fuel cell applications. The proton conductivities of the DHNHXX membranes were measured in deionized water (18 MΩ) at different temperatures (30–90 °C) using AC impedance spectroscopy. Fig. 6 shows the complex impedance spectra of the DHNH50 membrane at different temperatures, and the inset spectra of DHNHXX (XX = 40 to XX = 70) at 80 °C. The membrane resistance was derived from the low intersect of the high-frequency semicircle with the Re(Z) axis. The proton conductivity of the DHNHXX membranes was strongly dependent on their IEC values and the temperature, i.e., membranes with higher IECs would tend to exhibit higher proton conductivity and display increased proton conductivity with increasing temperature. This fact was obvious from the decreasing size of the arc in the complex impedance spectra. The conductivity values of the membranes were in the range of 5–46 mS cm⁻¹ at 30 °C, 10–99 mS cm⁻¹ at 80 °C and 12–129 mS cm⁻¹ at 90 °C (Table 5). Nevertheless, these proton conductivity values of the DHNHXX membranes are comparable to or better than many other SPIs reported in the literature with similar IECs and under similar experimental conditions.^8,27 For example, as shown in Table 5, the DHNH70 membrane showed two times higher proton conductivity than our previously reported DQN70 (IEC, 2.03 meq. g⁻¹) membrane, although the former exhibited similar (insignificantly higher) IEC values than the latter for the same DS. Similarly, the DHNH60 membrane showed almost four times higher proton conductivity than the DQN60 (IEC, 1.67 meq. g⁻¹) membrane for similar IEC values. This suggests that the influence of the type of non-sulfonated diamine on proton conductivity became weaker with increasing DS. The difference in the proton conductivity of the two systems (DHNHXX and DQNXX) may be attributed to the variation in the membrane morphology and water uptake (λ values). The proton conductivity of Nafion^® 117 is reported differently throughout the literature,^28,29 because the proton conductivity of Nafion^® depends upon many factors such as the manufacturing history, pretreatment process (purification/acidification), type of cells and measurement (four probe/two probe, through-plane/in-plane, in water/in RH-maintained environment). Thus, researchers are always encouraged to represent the proton conductivity of Nafion^® along with the synthesized PEMs to make a good comparison under the same set of laboratory conditions. Hence, we also measured the proton conductivity of Nafion^® 117 under similar test conditions to make a comparison with the DHNHXX membranes. The conductivity of Nafion^® 117 (Table 5) determined under our test conditions was closer to the reported conductivity data of Nafion^® 117 (77 mS cm⁻¹ at 30 °C and 165 mS cm⁻¹ at 80 °C) by Guiver et al.²⁹ Although the DHNHXX membranes had higher IEC values (IEC, 1.1–2.1 meq. g⁻¹) than Nafion^® 117 (IEC, 0.91 meq. g⁻¹), all of the DHNHXX membranes displayed relatively low proton conductivities in comparison to the conductivity of Nafion^® 117. This may be attributed to the difference in morphology and the point of attachment of the sulfonic acid groups, as well as the rigidity in the polymer which manifests for various reasons. Fig. 7 demonstrates the correlation of IEC with the proton conductivity and water uptake. The percolation in the proton conductivity and water uptake properties of DHNHXX was observed after DS = 0.5, similar to the results obtained for the DQNXX system. This was also quite obviously reflected in terms of the membrane microstructure with the formation of interconnected clusters, as previously explained in the microstructure section.

Fig. 6 Complex impedance spectra of the DHNHXX membranes.

Fig. 7 Variation of the proton conductivity and water uptake with IEC_W of the DHNHXX polymers.

Fig. 8 shows the temperature dependence of the proton conductivity for the DHNHXX and Nafion^® 117 membranes in deionized water. The proton conductivity of membranes showed Arrhenius-type (σ = A exp(−E_a/RT), where σ = proton conductivity (mS cm⁻¹); A = pre-exponential factor; R = universal gas constant (8.314 J mol⁻¹ K⁻¹); T = absolute temperature (K)) temperature dependence. With the rise in temperature, the proton conductivity of all of the membranes increased; DHNH70 exhibited proton conductivity comparable to the level of conductivity of Nafion^® 117 at high temperature. The activation energies, E_a, of the proton conductivity were in the range 13.8–18.46 kJ mol⁻¹ for DHNHXX (i.e. for XX = 40, E_a = 14.97 kJ mol⁻¹; for XX = 50, E_a = 18.46 kJ mol⁻¹; for XX = 60, E_a = 13.80 kJ mol⁻¹ and for XX = 70, E_a = 15.21 kJ mol⁻¹). Under similar laboratory conditions for Nafion^® 117 the value was 13.63 kJ mol⁻¹, whereas the literature reports activation energies for Nafion^® 117 between 7 and 14 kJ mol⁻¹.³⁰ This implied that the activation energies for DHNHXX were close to the activation energies of Nafion^® 117. This suggests that the DHNHXX membranes may involve similar kinds of proton conduction mechanisms.

Fig. 8 Arrhenius temperature dependence of the proton conductivity (σ) of the DHNHXX and Nafion^® 117 membranes.

4. Conclusion

Several new fluorinated SPIs (DHNHXX) with varying DS were prepared by the one pot polycondensation reaction of NTDA with DSDSA and HQA. FTIR and NMR were used to characterize the chemical structure of DHNHXX, while the composition was calculated with accuracy from the NMR signal integration. The DHNHXX with high inherent viscosity suggested the formation of high molar mass polymers. They also displayed good solubility in common aprotic solvents. The tough, flexible and transparent membranes were obtained by solution casting. The DHNHXX membranes exhibited high thermal stability with desulfonation temperatures in the range of 270–295 °C and good mechanical properties under both dry and wet conditions. The good oxidative stability of the DHNHXX membranes over many other SPIs was due to the presence of hydrophobic trifluoromethyl groups, and it increased with the proportion of trifluoromethyl groups in the polymer structures. The DHNHXX membranes displayed obvious anisotropic dimensional changes in water, with much greater expansion in the thickness direction than in the plane direction. The hydrophobic trifluoromethyl groups in the copolymer structure rendered DHNHXX with good phase-separated morphology as the TEM analysis revealed ionic cluster-like microstructures for the membranes, as generally expected in side chain-type sulfonated PEMs. The polymer chemical composition and IEC resulted in differences in the connectivity among the clusters in the membrane micrographs. These HQA-based DHNHXX membranes displayed much better proton conductivities and comparable or improved water stability (up to 450 h of testing) than the corresponding QA-based DQNXX membranes. For example, DHNH70, with the highest IEC (2.10 meq. g⁻¹) in the series, displayed improved proton conductivity (99 mS cm⁻¹ at 80 °C and 129 mS cm⁻¹ at 90 °C) which is approaching that of Nafion^® 117, while it did not dissolve for up to ∼415 h in water at 80 °C, suggesting its high potential as a PEM electrolyte for fuel cell applications.

References

1. S. M. J. Zaidi and M. A. Rauf, in In polymer membranes for fuel cells, ed. S. M. J. Zaidi and T. Matsuura, Springer, New York, 1st edn., 2009, ch. 1, p. 1.
2. 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.
3. C. H. Park, C. H. Lee, M. D. Guiver and Y. M. Lee, Prog. Polym. Sci., 2011, 36, 1443.
4. K. D. Kreuer, J. Membr. Sci., 2001, 185, 29.
5. C. Genies, R. Mercier, B. Sillion, N. Cornet, G. Gebel and M. Pineri, Polymer, 2001, 42, 359.
6. Y. Yin, O. Yamada, K. Tanaka, K. Okamoto, S. Liu, H. Ye, Y. Zhou, J. Zhao, H. Yanai and T. Sato, Polym. J., 2006, 38.
7. E. A. Mistri, A. K. Mohanty and S. Banerjee, J. Membr. Sci., 2012, 411, 117.
8. E. A. Mistri, A. K. Mohanty, S. Banerjee, H. Komber and B. Voit, J. Membr. Sci., 2013, 441, 168.
9. J. Fang, X. Guo, S. Harada, T. Watari, K. Tanaka, H. Kita and K.-I. Okamoto, Macromolecules, 2002, 35, 9022.
10. T. Watari, J. Fang, K. Tanaka, H. Kita, K.-I. Okamoto and T. Hirano, J. Membr. Sci., 2004, 230, 111.
11. B. R. Einsla, Y. T. Hong, Y. Seung Kim, F. Wang, N. Gunduz and J. E. McGrath, J. Polym. Sci., Part A: Polym. Chem., 2004, 42, 862.
12. Y. Yin, J. Fang, T. Watari, K. Tanaka, H. Kita and K.-I. Okamoto, J. Mater. Chem., 2004, 14, 1062.
13. Y. Yin, Q. Du, Y. Qin, Y. Zhou and K.-I. Okamoto, J. Membr. Sci., 2011, 367, 211.
14. Y. Yin, Y. Suto, T. Sakabe, S. Chen, S. Hayashi, T. Mishima, O. Yamada, K. Tanaka, H. Kita and K.-I. Okamoto, Macromolecules, 2006, 39, 1189.
15. M. H. Litt, Y. Zhang, R. F. Savinell and J. S. Wainright, Polym. Prepr. (Am. Chem. Soc. Polym. Div.), 1999, 40, 480.
16. Y. Zhang, M. H. Litt, R. F. Savinell and J. S. Wainright, Polym. Prepr. (Am. Chem. Soc. Polym. Div.), 2000, 41, 1561.
17. K. Miyatake, H. Zhou, T. Matsuo, H. Uchida and M. Watanabe, Macromolecules, 2004, 37, 4961.
18. F. Sun, T. Wang, S. Yang and L. Fan, Polymer, 2010, 51, 3887.
19. P. Bandyopadhyay, D. Bera and S. Banerjee, J. Membr. Sci., 2011, 382, 20.
20. A. K. Mohanty, E. A. Mistri, A. Ghosh and S. Banerjee, J. Membr. Sci., 2012, 409, 145.
21. F. Wang, M. Hickner, Y. S. Kim, T. A. Zawodzinski and J. E. McGrath, J. Membr. Sci., 2002, 197, 231.
22. T. D. M. Bell, S. V. Bhosale, C. M. Forsyth, D. Hayne, K. P. Ghiggino, J. A. Hutchison, C. H. Jani, S. J. Langford, M. A. P. Lee and C. P. Woodward, Chem. Commun., 2010, 46, 4881.
23. C. Liu, L. Li, Z. Liu, M. Guo, L. Jing, B. Liu, Z. Jiang, T. Matsumoto and M. D. Guiver, J. Membr. Sci., 2011, 366, 73.
24. F. Zhai, X. Guo, J. Fang and H. Xu, J. Membr. Sci., 2007, 296, 102.
25. N. Asano, M. Aoki, S. Suzuki, K. Miyatake, H. Uchida and M. Watanabe, J. Am. Chem. Soc., 2006, 128, 1762.
26. A. K. Mohanty, E. A. Mistri, S. Banerjee, H. Komber and B. Voit, Ind. Eng. Chem. Res., 2013, 52, 2772.
27. Y.-Z. Wang, K.-S. Ho, S.-D. Wu, K.-H. Hsieh and C.-H. Lee, J. Fuel Cell Sci. Technol., 2010, 7, 1.
28. S. Slade, S. A. Campbell, T. R. Ralph and F. C. Walsh, J. Electrochem. Soc., 2002, 149, A1556.
29. C. Wang, N. Li, D. W. Shin, S. Y. Lee, N. R. Kang, Y. M. Lee and M. D. Guiver, Macromolecules, 2011, 44, 7296.
30. F. Lufrano, I. Gatto, P. Staiti, V. Antonucci and E. Passalacqua, Solid State Ionics, 2001, 145, 47.

---

Footnote

† Authors made equal contribution.

---