- Email: [email protected]
Thermal and hydrolytic stability of sulfonated polyimide membranes with varying chemical structure
Polymer Degradation and Stability 90 (2005) 431e440 www.elsevier.com/locate/polydegstab
Thermal and hydrolytic stability of sulfonated polyimide membranes with varying chemical structure Wonbong Jang a, Choonkeun Lee b, Saimani Sundar c, Yong Gun Shul a, Haksoo Han a,* b
a Department of Chemical Engineering, Yonsei University, 134 Shinchon-dong, Sedaemun-gu, Seoul 120-749, Republic of Korea eMD Lab, SAMSUNG Electro-Mechanics Co., 314 Maetan3-dong, Yeongtong-gu, Suwon, Gyunggi-do 443-743, Republic of Korea c Yonsei Center for Clean Technology, Yonsei University, 134 Shinchon-dong, Sedaemun-gu, Seoul 120-749, Republic of Korea
Received 7 February 2005; received in revised form 5 April 2005; accepted 12 April 2005 Available online 23 May 2005
Abstract Many important properties required for fuel cell applications including hydrolytic stability, depend on various factors like flexibility of the polymer backbone, ring structure and phase separation. This paper is primarily focused on studying the effect of the chemical backbone structure on the hydrolytic stability and other properties. To study the difference in the hydrolytic stability with change in the chemical backbone structure of sulfonated polyimides we synthesized phthalic sulfonated polyimides and naphthalenic sulfonated polyimides. Two series of phthalic sulfonated polyimides were prepared using 4,4#-oxydiphthalic anhydride (ODPA) and 4,4#-methylene dianiline (MDA), and 4,4#-(hexafluoroisopropylidine) diphthalic anhydride (6FDA) and oxydianiline (ODA). 4,4#-Diaminobiphenyl-2,2#-disulfonic acid (BDSA) was used to introduce sulfonic acid group into both series. Naphthalenic polyimides were synthesized from 1,4,5,8-naphthalenetetra-carboxylic dianhydride, BDSA, MDA and ODA. Also to observe other properties according to variation of sulfonic acid content, the degree of functionalisation was effectively controlled by altering the mole ratio between the sulfonated and non-sulfonated diamine monomers in phthalic sulfonated polyimides. The hydrolytic stability of the polyimides was followed by FT-IR spectroscopy at regular intervals. Polyimides prepared using naphthalenic dianhydride, NTDA, exhibited higher hydrolytic stability than the phthalic dianhydrides. The proton conductivity, ion exchange capacity (IEC) and water uptake measurements revealed the dependence on the molecular weight of the repeating unit. The proton conductivity of the sulfonated polyimides was found to vary with chemical backbone structure. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Sulfonated polyimides; Hydrolytic stability; Morphology; Proton conductivity; Fuel cells
1. Introduction Fuel cells have been identified as a more feasible alternative energy source that is free from undesirable emissions [1,2]. At present the polymer electrolyte fuel cell (PEFC) is the most promising of all the fuel cell systems, which are viewed as portable power systems
* Corresponding author. Tel.: C82 2 2123 2764; fax: C82 2 312 6401. E-mail address: [email protected] (H. Han). 0141-3910/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.polymdegradstab.2005.04.012
for use in vehicles. The most important component of a PEFC is polymer electrolyte membrane itself [3]. Perfluoro sulfonic acid polymers (NafionÒ) are currently used as proton exchange membrane (PEM) in commercial systems [4e6]. These membranes are very expensive; moreover, their conductivity decreases at low humidity and at higher temperatures. This lead to extensive research for alternate PEM materials, and various sulfonated polymers have been synthesized and reviewed for their possible application as PEM material [7e13]. Aromatic polyimides, due to their salient features like high thermal stability, thermo-oxidative stability, high
432
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
mechanical strength and superior chemical resistance [14], are viewed as suitable PEM material to work in the more difficult fuel cell environment. Recently several aromatic polyimides are successfully synthesized with sulfonyl substitution and were tested in fuel cells [15e 17]. Generally, the introduction of sulfonic groups into polyimides is achieved by using the sulfonated aromatic diamines [15e19]. Control of sulfonation degree is of practical importance since high sulfonation lead to high swelling and even dissolution of the membrane in water whereas low sulfonation results in low conductivity. Hence it is often required to optimise the degree of sulfonation for the better performance of the polymer membrane. We can tailor the degree of sulfonation according to our requirement by using another nonsulfonated diamine comonomer and just adjusting the mole ratio of them. The rigidity and brittleness, which are the problems often encountered when polyimide based PEM materials are used in the fuel cell [16,20] can be overcome by using monomers that impart flexibility to the system. Moreover, the flexible systems are reported to have better stability towards water than the rigid polyimides [18,19]. Hence all the monomers except BDSA were selected as if it contains a flexible bridge, to increase the flexibility of the system. Hydrolytic stability of the sulfonated polyimide membrane is a limiting factor, which affects the lifetime of these membranes in fuel cell applications. Hence several studies focusing on the various aspects to increase the water stability were carried out. The hydrolytic stability and proton conductivity exhibited dependence on chemical structure of the polymer backbone and polymer morphology. The effect of various morphological aspects including phase separation [21], graft and random copolymerisation [18,21], flexibility of the main chain [18,19], the ring structure of the polyimide [22,23] etc. on the proton conductivity and hydrolytic stability were studied. The flexible systems exhibit better hydrolytic stability than the rigid polymer backbone [18,19]. Likewise the hydrophilicehydrophobic phase separation facilitates the formation of ionic clusters and results in better ionic conductivity. To study the hydrolytic degradation of polyimides, two series of polyimides with different chemical backbone structure were synthesized from dianhydrides and diamine comonomers with flexible linkages along with 4,4#-diaminobiphenyl-2,2#-disulfonic acid (BDSA) which was used to provide the required sulfonic acid groups. To study hydrolytic stability, less stable phthalic derivatives were used in the polyimide synthesis and they were compared with the naphthalenic derivatives. The polyimides were characterized by FT-IR and thermal stability. The proton conductivity, water uptake and ion exchange capacity of the polyimides with different chemical backbone structure were investigated.
2. Experimental 2.1. Materials 4,4#-Oxydiphthalic anhydride (ODPA), oxydianiline (ODA) and 4,4#-methylene dianiline (MDA) were purchased from Aldrich chemical company. 4,4#-(Hexafluoroisopropylidine) diphthalic anhydride (6FDA) was purchased from Chriskev Co. 4,4#-Diaminobiphenyl2,2#-disulfonic acid (BDSA), 1,4,5,8-naphthalenetetracarboxylic dianhydride (NTDA) and m-cresol were purchased from TCI. Triethylamine (TEA) was purchased from TEDIA. ODPA was purified by recrystallisation from acetic acid. ODA was purified by recrystallisation from ethanol. NTDA, MDA, 6FDA and BDSA were dried under vacuum before use. m-Cresol and TEA were used without further purification. 2.2. Synthesis of random copoly(amic acid)s of phthalic systems In a three-necked round bottom flask fitted with a magnetic stirrer, thermometer and nitrogen inlet, 0.465 g (1.35 mmol) of BDSA, 0.375 mL (2.7 mmol) of triethylamine and m-cresol were added and kept at 30 C till BDSA completely dissolved in m-cresol. To this solution ODA or MDA with different mole ratio was added under nitrogen atmosphere. To this solution, equimolecular of 6FDA or ODPA to total diamines was added in N2 atmosphere and stirred for 24 h at 0 C and the viscous copoly(amic acid) solutions were obtained. All other polyimides were prepared by adopting the same procedure, by varying the mole ratio of the monomers and the solid content was fixed at 15% by adjusting the amount of solvent (Scheme 1). 2.3. Synthesis of random copolyimides and proton exchange in phthalic systems The poly(amic acid) solutions were spin coated onto substrates like silicon wafers. The thickness of the films can be adjusted by varying the spin speed. Then, the poly(amic acid) films were kept in the curing oven. The following curing cycle was followed e the films were cured at 80 C for 1 h. Then, the temperature was increased to 200 C at a rate of 2 C/min and the films were cured for another 24 h at 200 C. Then they were cooled to 50 C at a cooling rate of 1 C/min. The copolyimide films were obtained and they were removed from the substrate, soaked in boiling methanol and then in 1.0 M HCl for 10 h to regenerate the sulfonic acid. The films were thoroughly washed with deionised water and then dried. The copolymides were named as the first two letters representing the anhydride and the next letter the comonomer. The following numeral gives the percent of BDSA content. For e.g. 6FDO 20 gives the
433
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
O
O
X
(m+n) O
SO3H +
O
O
m H2N
NH2 + n H2N
Y
NH2 HO3S
O
m-Cresol / NMP Et3N, N2
X: -O-, -(CF3)2Y: -CH2-, -OH O N C
O C OH
X
HO C O
O HO C
C N O H
Y
O H C N
X
SO3– N+HEt3
+ – C OH Et3H N O3S O
N C H O
n
m
O
O
X
O
N
N
Y
N
N m
O
O
SO3– N+HEt3
O
X
O
O
Et3H+N–O3S
n
0.1M HCl 6hrs O N O
X
O N
O Y
O
X
N m
O
O
SO3H
N O
HO3S
n
Scheme 1. Synthesis of five-membered sulfonated copolyimides.
polyimide from 6FDA, ODA, and the BDSA content is 20%. 2.4. Synthesis of six-membered sulfonated polyimides; film formation and proton exchange in naphthalenic systems
acid at room temperature for 10 h successively. The proton-exchanged films were thoroughly washed with deionised water and then dried in vacuo at 100 C for 20 h. The thickness of the films was in the range 20e40 mm. 2.5. Measurements
To a 100 mL of completely dried three-neck flask were added 2.0 mmol of BDSA, 12 mL of m-cresol, and 4.8 mmol of triethylamine successively under nitrogen flow. After the BDSA was completely dissolved, 2.0 mmol of non-sulfonated diamine was added. To this solution 4.0 mmol of dianhydride and 5.6 mmol of benzoic acid were added. The mixture was stirred at room temperature for a few minutes and then heated at 80 C for 4 h and 180 C for 20 h. After the reaction mixture was cooled to 80 C, an additional 20 mL of m-cresol was added to dilute the highly viscous solution, and then the solution was poured into acetone. The fibre-like precipitate was filtered off, washed thoroughly with acetone, and dried in vacuo at 80 C for 15 h. The triethylammonium salt of sulfonated polyimides were dissolved in m-cresol, and the solution was cast onto glass plates using spin coater and dried at 120 C for 6 h and 180 C for 4 h. The as-cast films were soaked in methanol at 60 C for 1 h and in 1.0 N hydrochloric
The infrared (FT-IR) spectra were collected using a Genesis series FT-IR (ATI Mattson Co.) instrument. Thermogravimetric results were obtained from thermogravimetric analyser (TGA1000, TA instrument), at a heating rate of 10 C/min in N2 atmosphere. The polyimide thin films after exchange of protons were dried in vacuum oven at 100 C for 24 h. For these polyimide thin films water uptake values were measured at four different temperatures 25 C, 40 C, 60 C and 80 C at 100% relative humidity and also as a function of time by using a thin film diffusion analyser (Cahn Instruments Co., model D-200) with a resolution of 0.1 mg. The detailed procedures were given in our previous studies [24e26]. The ion exchange capacity (IEC) was measured by titration. The sulfonated polyimide films were soaked in 1 M HCl for 10 h to regenerate the protons from the salt form, if any. They were thoroughly washed with
434
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
deionised water for several times and were soaked in 0.1 M NaCl for 15 h. The protons released due to the exchange reaction with NaC ions were titrated against 0.1 M NaOH, using phenolphthalein indicator. The IEC was determined from the equation IEC ðmeq=gÞZX!NNaOH =weight ðpolymerÞ where X is the volume of NaOH consumed; NNaOH is the normality of NaOH. The films were washed in deionised water and then they were soaked in 1.0 M HCl to convert the salt into acid, if any. The ionic conductivity of the sulfonated polyimides was measured using electrochemical impedance spectroscopy over the frequency range from 0.01 to 10,000 kHz. The resistance was measured using an Autolab Impedance Analyser in which the surface area of the Pt electrodes is 1.77 cm2 and the distance between the two electrodes is 4 cm. The ionic conductivity at temperatures 25 C, 40 C, 60 C, and 80 C were measured. The electrochemical cell used in this experiment is described in our previous study [16,17].
3. Results and discussion There are several major drawbacks of post sulfonation methods including the lack of control over the degree and location of functionalisation, which is usually a problem when dealing with macromolecules. Since the sulfonated polyimide could be prepared by the direct copolymerisation of sulfonated and non-sulfonated monomers, control of the polymer structures, molecular weight and degree of sulfonation could be achieved.
Transmittance
3.1. FT-IR spectrum The FT-IR spectra of the five-membered sulfonated polyimide membranes prepared using two different dianhydrides with different mole ratio of sulfonated diamine are given in Fig. 1. The peaks at 1784 and 1726 cm1 are due to the OCaO stretching of the polyimides and peaks at 1380 and 720 cm1 represent the CeN vibration. These confirm the formation of imides from poly(amic acid)s. The peaks at 1178 and 1095 cm1 are due to the stretching vibration of the SO2 [27]. The intensity of this peak was found to increase with the increase in the amount of BDSA, suggesting an increase in the incorporation of SO3H groups. This reveals that the amount of sulfonyl group, which is critical for the good behaviour of the film as a fuel cell membrane, can be effectively controlled by adjusting the mole ratio of the diamines with and without sulfonic acid groups.
S=O
S=O
a:6FDO 20
a:OMD 20
b:OMD 30
b:6FDO 30
c:OMD 40
c:6FDO 40
d:OMD 50
d:6FDO 50
1200
Two different series of sulfonated polyimides were synthesized with different chemical backbone structure; one from ODPA, BDSA and MDA and the other from 6FDA, BDSA and ODA. These two phthalic dianhydride based systems were compared with naphthalenic anhydride based systems. The diamine BDSA was used to incorporate the sulfonic acid group, and MDA and ODA were used as comonomer to control the content of sulfonyl incorporation. The polyimides were synthesized by conventional thermal imidization process and the curing process as given in Scheme 1. They are compared with napthalenic sulfonated polyimides for their degradation behaviour.
1000
1200
800
1000
800
Wavenumbers (cm-1) Fig. 1. Experimental apparatus for measuring conductivity of sulfonated copolyimide membranes.
435
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
The FT-IR spectra of NTDA based polyimides showed that the completion of imidization step was confirmed by the absence of anhydride carbonyl peak (around 1778 and 1737 cm1) and appearance of imide peak. Typical naphthalimide carbonyl absorptions at 1710 cm1 (asymmetric) and 1666 cm1 (symmetric) were observed which were lower than the phthalimide (for e.g. 1780 and 1720 cm1). The broad bands around 1178 and 1083 cm1 correspond to the SO2 stretching vibrations. The peak around 720 cm1 corresponds to the out-of-phase bending of the imide ring. The CeNeC stretching vibration was observed around 1342 cm1. 3.2. Thermal properties
3.3. Ion exchange capacity Table 1 shows the ion exchange capacity of the polyimide films. Generally, the IEC, proton conductivity and water uptake of the polyimide membrane increases with increase in sulfonic acid groups [3,29,30]. However, with the increase in the amount SO3H groups, the polymer become more hydrophilic and the stability towards water and other mechanical properties decreases. Hence it is essential to optimise the amount of sulfonic acid groups such that it contains only required amount. The sulfonated polyimides ODM 20e50 showed IEC of 0.73, 1.05, 1.42 and 1.74 meq/g, respectively, and the sulfonated polyimides 6FDO 20e50 showed IEC of 0.68, 0.92, 1.365 and 1.69 meq/g, respectively. The IEC values increased with increase in the amount of BDSA. Table 1 shows that ODM 50, ODM 40, 6FDO 50 and 6FDO 40 exhibited higher IEC than the reported IEC value of NafionÒ 115 (0.91 meq/g, [31]). These polymers might
110
110
100
100
90
90
80
80
70
70
Weight (%)
Weight (%)
The thermal stability of the five-membered sulfonated polyimides and the NafionÒ 115 was analysed by thermogravimetric analysis and the results are given in Fig. 2. All the polyimide thin films exhibited a three-step degradation pattern. The first weight loss observed around 100 C was due to the loss of water molecules, absorbed by the highly hygroscopic SO3 groups [18,19]. The second step of degradation was observed around 300 C due to the decomposition of sulfonyl groups, leading to expulsion of sulfur monoxide and sulfur dioxide gases [18,19,28]. The onset of the second step is almost similar for both the NafionÒ 115 and the sulfonated polyimides. The similar onset points for the two series of polyimides and the NafionÒ 115 suggest that irrespective of the polymer backbone, the degradation of the sulfonyl groups occurs around 300 C. Also, the percentage of weight loss in the second step corresponding to the elimination of SO3 groups increase with increase in BDSA content. This also proves that by altering the mole ratios of the diamines with and without sulfonyl groups we can effectively alter the amount of sulfonyl
incorporation. The third step indicates the decomposition of the polymer backbones. All the polyimides showed better thermal stability than the NafionÒ 115. For the NafionÒ 115, the onset for third step of degradation started around 380 C and there was a rapid weight loss observed beyond 380 C. But the polyimides showed an onset around 500 C indicating their better thermal stability. Thermal stability of the naphthalenic polyimides was higher than that of phthalic polyimides by showing third onset around over 500 C. This can be attributed to the more strain-free six-membered configuration of the polyimide than the strained five-membered ring structure. These results show that the synthesized sulfonated polyimide films can be used as polymer electrolyte membranes for operation at high temperature due to their thermal stability being higher than the intended operational temperature around 130e150 C.
60 50 ODM 50 ODM 40 ODM 30 ODM 20 Nafion115
40 30 20
50 6FDO 50 6FDO 40 6FDO 30 6FDO 20 Nafion115
40 30 20 10
10 0
60
0
100
200
300
400
Temperature
500
(oC)
600
700
0
0
100
200
300
400
500
600
Temperature (oC)
Fig. 2. TGA thermograms of sulfonated copolyimide (ODM and 6FDO) and NafionÒ 115 membranes.
700
436
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
Table 1 Ion exchange capacity and water sorption of sulfonated copolyimides (ODM and 6FDO) and NafionÒ 115 membranes Polymer
Nafion 115 ODM 20 ODM 30 ODM 40 ODM 50 6FDO 20 6FDO 30 6FDO 40 6FDO 50
Experimental IEC (meq/g)
Water uptake (%)
0.730 1.050 1.420 1.740 0.680 0.920 1.365 1.690
25 C
40 C
60 C
80 C
22.90 6.69 8.98 15.84 17.37 5.16 7.21 12.80 14.00
24.32 6.92 10.57 16.02 18.98 5.72 8.43 12.16 15.14
25.80 7.38 11.02 16.84 21.90 6.25 8.62 12.64 15.80
29.11 7.81 11.37 18.24 22.78 8.08 9.57 14.14 17.52
contain more sulfonyl groups than the NafionÒ 115 and hence resulted in higher IEC value. Between the two types of polyimides, for the same degree of sulfonyl substitution, ODM showed higher IEC compared to the 6FDO. The molecular weight of the repeating units in ODM is 472.438 g/mol for the MDA segment and 618.538 g/mol for the BDSA segment. For the 6FDO it is 608.458 g/mol for the ODA segment and 752.588 g/mol for the BDSA segment. It can be seen that, the molecular weight of the repeating segments in ODM is lesser than the 6FDO and hence, in a unit volume, more number of SO3H groups will be present in the ODM than 6FDO. Hence, at a given degree of sulfonization, ODM showed better IEC value than 6FDO. The same trend was obtained in the proton conductivity and water uptake experiments. ODM showed better conductivity and higher water uptake compared to 6FDO at a given sulfonyl concentration.
(Table 1 and Fig. 3). The polymer films were kept in a humidified chamber at the mentioned temperature for 6 h before evaluating the water uptake properties and the weight of the polyimide membrane before and after humidification was measured by a Cahn microbalance. At a constant temperature the amount of water uptake with time is measured at 10 min interval up to 4 h. All the samples exhibited rapid water uptake initially and reached an equilibrium beyond which the variation in the water uptake is small with time. At 25 C, NafionÒ 115 showed a water uptake of 22.9%. The ODM 20e50 exhibited an increase in water uptake from 6.7 to 17.4% where as the 6FDO 20e50 showed increase in the water uptake from 5.2 to 13.2%. This is because the amount of water uptake increases with increase in the amount of sulfonyl group [3,21]. Generally, the proton conductivity of the polymer increase with increase in water uptake. This is because with the more water uptake it improves the formation of the hydrophilic domain carrying the proton conductivity [36]. The variation in the water uptake with the increase in temperature is given in Fig. 3. For NafionÒ 115 and all the polyimides, a linear increase in water uptake with increase in temperature was observed. This may be because, with the increase in temperature the molecular mobility increases and hence the water molecules can easily penetrate through the polymer backbone and hydrate a maximum number of SO3H groups. Between the two series of polyimides the ODM showed a higher uptake of water than the 6FDO. This is because of the high hydrophobic interactions of the fluorine groups present in the backbone of the 6FDO polymers compared to ODM.
3.4. Water uptake 3.5. Proton conductivity Water uptakes of the polyimide films at four different temperatures namely, 25 C, 40 C, 60 C and 80 C were evaluated and compared with that of NafionÒ 115
The conductivities for all the sulfonated polyimides and the NafionÒ 115 were determined at four different 30
30 Nafion115
25
Nafion115
ODM 50
20
ODM 40
15
25
20 6FDO 50
15
6FDO 40
10
6FDO 30
ODM 30
10
ODM 20
6FDO 20
5
5
0 20
30
40
50
60
Temperature
70
(oC)
80
90
0 20
30
40
50
60
70
80
90
Temperature (oC)
Fig. 3. Water sorption of sulfonated copolyimide (ODM and 6FDO) and NafionÒ 115 membranes as a function of the temperature.
437
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
temperatures namely 25 C, 40 C, 60 C and 80 C (Fig. 4). Fig. 4 shows that there is increase in the conductivity with increase in the temperature for all the polyimides and NafionÒ 115. The increase in temperature favours both the dynamics involved in proton transfer and structural reorganization, which are required for fast proton conductivity. With increase in temperature both the proton diffusion and molecular diffusion merge together and result in the increased proton conductivity. However, it is reported that molecular diffusion predominates at elevated temperatures like 300 C [32]. In both series conductivity increased with increase in sulfonic acid content as expected. The sulfonated polyimides ODM 40 and 50 and 6FDO 40 and 50 showed better conductivity to be used as PEM in fuel cell than the reported sulfonated poly sulfone (SPSF), phosphotungstic acid (PWA) and polybenzimidazole (PBI) membranes [33,34]. It can be seen from Table 1 and Fig. 4 that the polyimides ODM 50 and 40 and 6FDO 50 and 40 showed a higher IEC than the NafionÒ 115 but the conductivity was less than NafionÒ 115. Generally, materials with higher IEC are expected to show better conductivity. The observed lower conductivity than the NafionÒ 115 despite the higher IEC value may be due to the difference in chemical structure and morphology. NafionÒ 115 having an ethylenic backbone has a better rotational freedom compared to the rigid, aromatic polyimide backbone. Hence the NafionÒ 115 will have loose molecular packing, when compared to the densely packed polyimides, which makes the passage of proton through the polymer membrane much easier. The rigidity of the NafionÒ 115 is less compared to that of polyimides and hence it is more flexible facilitating the easy formation of ionic clusters [35] and also the mobility of the main chain of the NafionÒ 115 is more compared to that of polyimides due to less molecular
interactions and much flexibility. The presence of groups like OCaO, N, eOe, CeH involve in more molecular interactions than the hydrophobic CeF backbone and hence restrict the molecular mobility. Moreover, NafionÒ 115, due to its high phase separation between the hydrophilic SO3H groups and hydrophobic backbone, augments the formation of ionic cluster and hence the ionic cluster density is higher compared to nonphase separated imides. Due to these factors the ionic conductivity of the NafionÒ 115 is higher compared to sulfonated polyimides at all temperatures but the conductivity of NafionÒ 115 decrease in elevated temperatures. These results prove that the ionic conductivity is not solely dependent on the IEC of the polymers. Between the two types of polyimides 6FDO and ODM, ODM showed better conductivity due to the less molecular weight of the repeating unit as explained in IEC. This shows that apart from the degree of sulfonation, the morphology also affects the proton conductivity. 3.6. Hydrolytic stability The polyimides were evaluated for the hydrolytic stability by soaking in water at 80 C. Their corresponding stability in water was followed by monitoring the intensity of the asymmetric stretching vibration of imide carbonyl peak at 1784 cm1 in FT-IR. Genies et al. reported the variation in hydrolytic stability of imides with naphthalenic and phthalic ring structure using NMR studies. In that study the polyimide chain scissions were followed by 1H NMR and 13C NMR in which the appearance of new peaks in 13C NMR have been found and correlated. The effect of chemical backbone structure on the hydrolytic stability of the system was systematically studied. The five-membered sulfonated polyimides were soaked in water at 80 C and their structural changes
Nafion115 ODM 50
1e-2
1e-2
ODM 40
1e-3
Conductivity (S/cm)
Conductivity (S/cm)
Nafion115
SPSF [34]
PBI (PWA60) [33]
ODM 30
1e-4
6FDO 50 6FDO 40
1e-3
SPSF [34]
PBI (PWA60) [33]
6FDO 30
1e-4 ODM 20
6FDO 20
1e-5 20
30
40
50
60
Temperature
70
(oC)
80
90
100
1e-5 20
30
40
50
60
70
80
90
100
Temperature (oC)
Fig. 4. Proton conductivity of sulfonated copolyimide (ODM and 6FDO) and NafionÒ 115 membranes as a function of the temperature.
438
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
C=O (asymmetric) 0h
C=O (symmetric)
C-N-C
S=O (symmetric) S=O (asymmetric)
30min
60min
Transmittance
with time were studied by FT-IR at specific time intervals and are given in Figs. 5 and 6. Both series showed a decrease in the intensity of the carbonyl absorption of the imide around 1784 cm1 with increase in time indicating the rupture of the imide rings of the polymer. Also the peak around 1038 cm1, which is due to the SO2 vibration, decreased with increase in time. This clearly shows that there is a decrease in the amount of SO3H groups in the polymer backbone. It is well known that there is a decrease in ionic conductivity of the sulfonated polyimide membranes with time, which was attributed to the polymer degradation. The degradation of the polyimides must be caused by the hydrolysis of imide structure. In our study with FT-IR we found the decrease in the intensity of SO2 vibrations. Due to chain scission, the large polymer molecules are broken down to smaller molecules, which were leached out by water, accounting for the loss of SO3H groups. Hence these polyimides suffer a loss in molecular weight which account for the brittleness and loss of mechanical strength of the polyimide membranes after hydrolysis which was observed by several investigators. The possible hydrolytic degradation mechanism which account for the loss of sulfonic acid groups is given in Fig. 7. The careful FT-IR analysis reveals that absorption due to both imide carbonyl and sulfonic acid decreased. But the decrease in imide carbonyl absorption was rapid and started at 30 min whereas the loss in sulfonyl group absorption was observed after 4 h. This indicate that first one carbonyl group of polyimides are attacked leading to the formation of amic acid structure in which although there is loss in imide carbonyl absorption the polymer chain is intact and hence there is no loss
90min 2h
4h
24h
2000
1800
1600
1400
1200
1000
800
600
400
Wave number (cm-1) Fig. 6. FT-IR spectra of the sulfonated polyimides derived from 6FDO.
in sulfonyl group. With increase in time the second carbonyl group also undergoes nucleophilic attack by OH group leading to the formation of acid and amine structure. This led to chain scission and loss of sulfonyl absorption. The chemical backbone structure of the two different series of five-membered ring polyimide affected the hydrolytic stability. When compared at a specific time, the 6FDO series showed better water stability than the
O
O C=O (symmetric) C-N-C
0h
N
N
S=O (asymmetric)
C
C
O
2h
Transmittance
C
C
S=O (symmetric)
C=O (asymmetric)
HO
OH HO
4h
C
OH
O
36h
O
C
N H C OH
24h
N H C OH O
O 52h
O C
OH
C
OH
+ 2000
O
1800
1600
1400
1200
1000
800
600
400
Wave number (cm-1) Fig. 5. FT-IR spectra of the sulfonated polyimides derived from ODM.
H2N
O Fig. 7. Mechanism of hydrolytic degradation in polyimides.
439
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
4. Conclusion A series of polyimides was synthesized and characterized to evaluate the suitability for the fuel cell application. Sulfonyl groups are used as hydrophilic nanophores to affect transportation of protons dissociated by the dynamics of water. The level of sulfonyl substitution was effectively controlled and revealed by the FT-IR and
6h
21h
Transmittance
43h
80h
S=O (symmetric) C=O (symmetric) S=O (asymmetric) C=O (asymmetric)
2000
1800
1600
C-N-C
1400
1200
1000
800
Wavenumber (cm-1) Fig. 8. FT-IR spectra of the sulfonated polyimides derived from NTDA-BDSA/ODA.
6h
21h
Transmittance
ODM series. This may be due to the highly hydrophobic nature of the fluorine backbone, which resists the approach of water to a greater extent and hence prevents the rapid attack of water, increasing the overall stability. When the polyimides from six-membered and fivemembered anhydrides are compared, the six-membered rings exhibited pronounced stability. When the intensity of imide carbonyl peaks at 1784 cm1 was compared for a specific time, for e.g. 4 h, the six-membered anhydride based system exhibited higher intensity than the fivemembered anhydride system (Figs. 8 and 9). The sixmembered polyimides in contrast showed no decrease in carbonyl absorption or loss in SO2 absorption. Moreover, the polyimides retained their mechanical integrity until three days against 52 h of five-membered analogues. This indicates that six-membered anhydrides based polyimides were more intact and hence have more hydrolytic stability than the five-membered ones. This must be due to the fact that the six-membered rings have a strain-free conformation compared to five-membered ring structures. This is the reason for the high temperature reaction conditions for the naphthalenic polyimides.
43h
80h S=O (symmetric)
C=O (symmetric) S=O (asymmetric) C=O (asymmetric) C-N-C
2000
1800
1600
1400
1200
1000
800
Wavenumber (cm-1) Fig. 9. FT-IR spectra of the sulfonated polyimides derived from NTDA-BDSA/MDA.
TGA studies. The intensity of the sulfonyl absorption band increased with the increase in the amount of BDSA. Also the percentage weight loss in the second step, which is due to the decomposition of SO3 group, increased with increase in BDSA content. Both confirmed that the percent sulfonization could be altered by using comonomers without SO3H group. The IEC values were found to depend on percent sulfonization and as well as on the molecular weight of the repeating unit. The proton conductivity measurements showed that the flexibility of the polymer backbone and phase separation facilitates more ionic conduction. Water uptake values showed temperature dependence. Rapid water uptake was observed initially and reaches an equilibrium beyond which the variation in the water uptake is small with time. Hydrophobicity of the polymer backbone affects the amount of water uptake. 6FDO series showed better hydrolytic stability than the ODM due to the hydrophobic backbone. When the intensity of imide carbonyl peaks at 1784 cm1 was compared for a specific time, the six-membered anhydrides based system exhibited higher intensity than the five-membered anhydrides system. The six-membered polyimides showed no decrease in carbonyl absorption or loss in SO2 absorption for a longer time. Moreover, the polyimides retained their mechanical integrity until three days against 52 h for five-membered analogues. This indicates that six-membered anhydrides based polyimides were more intact and hence have more hydrolytic stability than the five-membered ones. This must be due to the fact that the six-membered rings have a strain-free conformation compared to five-membered ring structures. This is the reason for the high temperature reaction conditions for the naphthalenic
440
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
polyimides. Naphthalenic polyimides showed the possibility to be suitable for fuel cell with higher hydrolytic stability and thermal stability than the phthalic polyimides.
Acknowledgement This work was supported by the Ministry of Science and Technology of Korea through the National Research Laboratory Program.
References [1] Costamanga P, Srinivasan S. J power sources 2001;102:242. [2] Larmine J, Ducks A, editors. Fuel cell systems explained. Wiley publications; 2000. p. 1e15. [3] Poppe D, Frey H, Kruer KD, Heinzel A, Mulhaupt R. Macromolecules 2002;35:7936. [4] Yeo RS, Chin DT. J Electrochem Soc 1980;127:546. [5] Wilson MS, Gottesfeld S. J Appl Electrochem 1992;22:1. [6] Ren XM, Wilson MS, Gottesfeld S. J Electrochem Soc 1996; 143:12. [7] Hickner MA, Ghassemi H, Kim YS, Einsla BR, Mcgrath JE. Chem Rev 2004;104:4587 [and references therein]. [8] Savadogo O. J Power Sources 2004;127:135 [and references therein]. [9] Miyatake K, Fukushima K, Takeoka S, Tsuchida E. Chem Mater 1999;11:1172. [10] Hasiotis C, Deimede V, Kontoyannis C. Electrochim Acta 2001;46:401. [11] Caretta N, Tricoli V, Piccioni F. J Membr Sci 2000;166:189. [12] Notle R, Ledjeff K, Bauer M, Mulhaupt R. J Membr Sci 1993;83:211.
[13] Bailey C, William DJ, Karasz FE, Macknight WJ. Polymer 1987;28:1009. [14] Ghosh MK, Mittal KL, editors. Polyimides fundamental and application. NY: Marcel Dekker Inc.; 1996. [15] Vallejo E, Pourcelly G, Gavach C, Mercier R, Pineri M. J Membr Sci 1999;60:127. [16] Lee C, Sundar S, Kwon J, Han H. J Polym Sci Part A Polym Chem 2004;42:3612. [17] Lee C, Sundar S, Kwon J, Han H. J Polym Sci Part A Polym Chem 2004;42:3621. [18] Fang J, Guo X, Harada S, Watari T, Tanaka K, Kita H, et al. Macromolecules 2002;35:9022. [19] Guo X, Fang J, Watari T, Tanaka K, Kita H, Okamoto K. Macromolecules 2002;35:6707. [20] Miyatake K, Oyaizu K, Tsuchida E, Hay AS. Macromolecules 2001;34:2065. [21] Ding J, Chuy C, Holdcroft S. Chem Mater 2001;13:2231. [22] Genius C, Mercier R, Sillion B, Petiaud R, Cornet N, Gebel G, et al. Polymer 2001;42:5097. [23] Guo X, Fang J, Tanaka K, Kita H, Okamoto K. J Polym Sci Part A Polym Chem 2004;42:1432. [24] Han H, Seo J, Ree M, Pyo SM, Gryte CC. Polymer 1998;39:2963. [25] Seo J, Lee A, Lee C, Han H. J Appl Polym Sci 2000;76:1315. [26] Seo J, Lee A, Oh J, Han H. Polymer J 2000;32:583. [27] Pavia DL, Lampman GR, Kriz GS, editors. Introduction to spectroscopy. NY: Harcourt college publishers; 2001. [28] Rikukawa M, Sauni K. Prog Polym Sci 2000;25:1463. [29] Miyatake K, Shouji E, Yamamoto K, Tsuchida E. Macromolecules 1997;30:2941. [30] Kobayashi T, Rikukawa M, Sanui K, Ogata N. Solid State Ionics 1998;106:219. [31] Yang C, Srinivasan S, Bocarsly AB, Tulyani S, Benziger JB. J Membr Sci 2004;237:145. [32] Kreuer KD. Chem Mater 1996;8:610. [33] Staiti P, Minutoli M, Hocevar S. J Power Sources 2000;90:231. [34] Dimitrova PG, Baradie B, Foscallo D, Poinsignon C, Sanchez JY. J Membr Sci 2001;185:59. [35] Eisenberg A, Hird B, Moore RB. Macromolecules 1990;23:4098. [36] Kreuer KD. J Membr Sci 2001;185:29.
Thermal and hydrolytic stability of sulfonated polyimide membranes with varying chemical structure Wonbong Jang a, Choonkeun Lee b, Saimani Sundar c, Yong Gun Shul a, Haksoo Han a,* b
a Department of Chemical Engineering, Yonsei University, 134 Shinchon-dong, Sedaemun-gu, Seoul 120-749, Republic of Korea eMD Lab, SAMSUNG Electro-Mechanics Co., 314 Maetan3-dong, Yeongtong-gu, Suwon, Gyunggi-do 443-743, Republic of Korea c Yonsei Center for Clean Technology, Yonsei University, 134 Shinchon-dong, Sedaemun-gu, Seoul 120-749, Republic of Korea
Received 7 February 2005; received in revised form 5 April 2005; accepted 12 April 2005 Available online 23 May 2005
Abstract Many important properties required for fuel cell applications including hydrolytic stability, depend on various factors like flexibility of the polymer backbone, ring structure and phase separation. This paper is primarily focused on studying the effect of the chemical backbone structure on the hydrolytic stability and other properties. To study the difference in the hydrolytic stability with change in the chemical backbone structure of sulfonated polyimides we synthesized phthalic sulfonated polyimides and naphthalenic sulfonated polyimides. Two series of phthalic sulfonated polyimides were prepared using 4,4#-oxydiphthalic anhydride (ODPA) and 4,4#-methylene dianiline (MDA), and 4,4#-(hexafluoroisopropylidine) diphthalic anhydride (6FDA) and oxydianiline (ODA). 4,4#-Diaminobiphenyl-2,2#-disulfonic acid (BDSA) was used to introduce sulfonic acid group into both series. Naphthalenic polyimides were synthesized from 1,4,5,8-naphthalenetetra-carboxylic dianhydride, BDSA, MDA and ODA. Also to observe other properties according to variation of sulfonic acid content, the degree of functionalisation was effectively controlled by altering the mole ratio between the sulfonated and non-sulfonated diamine monomers in phthalic sulfonated polyimides. The hydrolytic stability of the polyimides was followed by FT-IR spectroscopy at regular intervals. Polyimides prepared using naphthalenic dianhydride, NTDA, exhibited higher hydrolytic stability than the phthalic dianhydrides. The proton conductivity, ion exchange capacity (IEC) and water uptake measurements revealed the dependence on the molecular weight of the repeating unit. The proton conductivity of the sulfonated polyimides was found to vary with chemical backbone structure. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Sulfonated polyimides; Hydrolytic stability; Morphology; Proton conductivity; Fuel cells
1. Introduction Fuel cells have been identified as a more feasible alternative energy source that is free from undesirable emissions [1,2]. At present the polymer electrolyte fuel cell (PEFC) is the most promising of all the fuel cell systems, which are viewed as portable power systems
* Corresponding author. Tel.: C82 2 2123 2764; fax: C82 2 312 6401. E-mail address: [email protected] (H. Han). 0141-3910/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.polymdegradstab.2005.04.012
for use in vehicles. The most important component of a PEFC is polymer electrolyte membrane itself [3]. Perfluoro sulfonic acid polymers (NafionÒ) are currently used as proton exchange membrane (PEM) in commercial systems [4e6]. These membranes are very expensive; moreover, their conductivity decreases at low humidity and at higher temperatures. This lead to extensive research for alternate PEM materials, and various sulfonated polymers have been synthesized and reviewed for their possible application as PEM material [7e13]. Aromatic polyimides, due to their salient features like high thermal stability, thermo-oxidative stability, high
432
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
mechanical strength and superior chemical resistance [14], are viewed as suitable PEM material to work in the more difficult fuel cell environment. Recently several aromatic polyimides are successfully synthesized with sulfonyl substitution and were tested in fuel cells [15e 17]. Generally, the introduction of sulfonic groups into polyimides is achieved by using the sulfonated aromatic diamines [15e19]. Control of sulfonation degree is of practical importance since high sulfonation lead to high swelling and even dissolution of the membrane in water whereas low sulfonation results in low conductivity. Hence it is often required to optimise the degree of sulfonation for the better performance of the polymer membrane. We can tailor the degree of sulfonation according to our requirement by using another nonsulfonated diamine comonomer and just adjusting the mole ratio of them. The rigidity and brittleness, which are the problems often encountered when polyimide based PEM materials are used in the fuel cell [16,20] can be overcome by using monomers that impart flexibility to the system. Moreover, the flexible systems are reported to have better stability towards water than the rigid polyimides [18,19]. Hence all the monomers except BDSA were selected as if it contains a flexible bridge, to increase the flexibility of the system. Hydrolytic stability of the sulfonated polyimide membrane is a limiting factor, which affects the lifetime of these membranes in fuel cell applications. Hence several studies focusing on the various aspects to increase the water stability were carried out. The hydrolytic stability and proton conductivity exhibited dependence on chemical structure of the polymer backbone and polymer morphology. The effect of various morphological aspects including phase separation [21], graft and random copolymerisation [18,21], flexibility of the main chain [18,19], the ring structure of the polyimide [22,23] etc. on the proton conductivity and hydrolytic stability were studied. The flexible systems exhibit better hydrolytic stability than the rigid polymer backbone [18,19]. Likewise the hydrophilicehydrophobic phase separation facilitates the formation of ionic clusters and results in better ionic conductivity. To study the hydrolytic degradation of polyimides, two series of polyimides with different chemical backbone structure were synthesized from dianhydrides and diamine comonomers with flexible linkages along with 4,4#-diaminobiphenyl-2,2#-disulfonic acid (BDSA) which was used to provide the required sulfonic acid groups. To study hydrolytic stability, less stable phthalic derivatives were used in the polyimide synthesis and they were compared with the naphthalenic derivatives. The polyimides were characterized by FT-IR and thermal stability. The proton conductivity, water uptake and ion exchange capacity of the polyimides with different chemical backbone structure were investigated.
2. Experimental 2.1. Materials 4,4#-Oxydiphthalic anhydride (ODPA), oxydianiline (ODA) and 4,4#-methylene dianiline (MDA) were purchased from Aldrich chemical company. 4,4#-(Hexafluoroisopropylidine) diphthalic anhydride (6FDA) was purchased from Chriskev Co. 4,4#-Diaminobiphenyl2,2#-disulfonic acid (BDSA), 1,4,5,8-naphthalenetetracarboxylic dianhydride (NTDA) and m-cresol were purchased from TCI. Triethylamine (TEA) was purchased from TEDIA. ODPA was purified by recrystallisation from acetic acid. ODA was purified by recrystallisation from ethanol. NTDA, MDA, 6FDA and BDSA were dried under vacuum before use. m-Cresol and TEA were used without further purification. 2.2. Synthesis of random copoly(amic acid)s of phthalic systems In a three-necked round bottom flask fitted with a magnetic stirrer, thermometer and nitrogen inlet, 0.465 g (1.35 mmol) of BDSA, 0.375 mL (2.7 mmol) of triethylamine and m-cresol were added and kept at 30 C till BDSA completely dissolved in m-cresol. To this solution ODA or MDA with different mole ratio was added under nitrogen atmosphere. To this solution, equimolecular of 6FDA or ODPA to total diamines was added in N2 atmosphere and stirred for 24 h at 0 C and the viscous copoly(amic acid) solutions were obtained. All other polyimides were prepared by adopting the same procedure, by varying the mole ratio of the monomers and the solid content was fixed at 15% by adjusting the amount of solvent (Scheme 1). 2.3. Synthesis of random copolyimides and proton exchange in phthalic systems The poly(amic acid) solutions were spin coated onto substrates like silicon wafers. The thickness of the films can be adjusted by varying the spin speed. Then, the poly(amic acid) films were kept in the curing oven. The following curing cycle was followed e the films were cured at 80 C for 1 h. Then, the temperature was increased to 200 C at a rate of 2 C/min and the films were cured for another 24 h at 200 C. Then they were cooled to 50 C at a cooling rate of 1 C/min. The copolyimide films were obtained and they were removed from the substrate, soaked in boiling methanol and then in 1.0 M HCl for 10 h to regenerate the sulfonic acid. The films were thoroughly washed with deionised water and then dried. The copolymides were named as the first two letters representing the anhydride and the next letter the comonomer. The following numeral gives the percent of BDSA content. For e.g. 6FDO 20 gives the
433
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
O
O
X
(m+n) O
SO3H +
O
O
m H2N
NH2 + n H2N
Y
NH2 HO3S
O
m-Cresol / NMP Et3N, N2
X: -O-, -(CF3)2Y: -CH2-, -OH O N C
O C OH
X
HO C O
O HO C
C N O H
Y
O H C N
X
SO3– N+HEt3
+ – C OH Et3H N O3S O
N C H O
n
m
O
O
X
O
N
N
Y
N
N m
O
O
SO3– N+HEt3
O
X
O
O
Et3H+N–O3S
n
0.1M HCl 6hrs O N O
X
O N
O Y
O
X
N m
O
O
SO3H
N O
HO3S
n
Scheme 1. Synthesis of five-membered sulfonated copolyimides.
polyimide from 6FDA, ODA, and the BDSA content is 20%. 2.4. Synthesis of six-membered sulfonated polyimides; film formation and proton exchange in naphthalenic systems
acid at room temperature for 10 h successively. The proton-exchanged films were thoroughly washed with deionised water and then dried in vacuo at 100 C for 20 h. The thickness of the films was in the range 20e40 mm. 2.5. Measurements
To a 100 mL of completely dried three-neck flask were added 2.0 mmol of BDSA, 12 mL of m-cresol, and 4.8 mmol of triethylamine successively under nitrogen flow. After the BDSA was completely dissolved, 2.0 mmol of non-sulfonated diamine was added. To this solution 4.0 mmol of dianhydride and 5.6 mmol of benzoic acid were added. The mixture was stirred at room temperature for a few minutes and then heated at 80 C for 4 h and 180 C for 20 h. After the reaction mixture was cooled to 80 C, an additional 20 mL of m-cresol was added to dilute the highly viscous solution, and then the solution was poured into acetone. The fibre-like precipitate was filtered off, washed thoroughly with acetone, and dried in vacuo at 80 C for 15 h. The triethylammonium salt of sulfonated polyimides were dissolved in m-cresol, and the solution was cast onto glass plates using spin coater and dried at 120 C for 6 h and 180 C for 4 h. The as-cast films were soaked in methanol at 60 C for 1 h and in 1.0 N hydrochloric
The infrared (FT-IR) spectra were collected using a Genesis series FT-IR (ATI Mattson Co.) instrument. Thermogravimetric results were obtained from thermogravimetric analyser (TGA1000, TA instrument), at a heating rate of 10 C/min in N2 atmosphere. The polyimide thin films after exchange of protons were dried in vacuum oven at 100 C for 24 h. For these polyimide thin films water uptake values were measured at four different temperatures 25 C, 40 C, 60 C and 80 C at 100% relative humidity and also as a function of time by using a thin film diffusion analyser (Cahn Instruments Co., model D-200) with a resolution of 0.1 mg. The detailed procedures were given in our previous studies [24e26]. The ion exchange capacity (IEC) was measured by titration. The sulfonated polyimide films were soaked in 1 M HCl for 10 h to regenerate the protons from the salt form, if any. They were thoroughly washed with
434
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
deionised water for several times and were soaked in 0.1 M NaCl for 15 h. The protons released due to the exchange reaction with NaC ions were titrated against 0.1 M NaOH, using phenolphthalein indicator. The IEC was determined from the equation IEC ðmeq=gÞZX!NNaOH =weight ðpolymerÞ where X is the volume of NaOH consumed; NNaOH is the normality of NaOH. The films were washed in deionised water and then they were soaked in 1.0 M HCl to convert the salt into acid, if any. The ionic conductivity of the sulfonated polyimides was measured using electrochemical impedance spectroscopy over the frequency range from 0.01 to 10,000 kHz. The resistance was measured using an Autolab Impedance Analyser in which the surface area of the Pt electrodes is 1.77 cm2 and the distance between the two electrodes is 4 cm. The ionic conductivity at temperatures 25 C, 40 C, 60 C, and 80 C were measured. The electrochemical cell used in this experiment is described in our previous study [16,17].
3. Results and discussion There are several major drawbacks of post sulfonation methods including the lack of control over the degree and location of functionalisation, which is usually a problem when dealing with macromolecules. Since the sulfonated polyimide could be prepared by the direct copolymerisation of sulfonated and non-sulfonated monomers, control of the polymer structures, molecular weight and degree of sulfonation could be achieved.
Transmittance
3.1. FT-IR spectrum The FT-IR spectra of the five-membered sulfonated polyimide membranes prepared using two different dianhydrides with different mole ratio of sulfonated diamine are given in Fig. 1. The peaks at 1784 and 1726 cm1 are due to the OCaO stretching of the polyimides and peaks at 1380 and 720 cm1 represent the CeN vibration. These confirm the formation of imides from poly(amic acid)s. The peaks at 1178 and 1095 cm1 are due to the stretching vibration of the SO2 [27]. The intensity of this peak was found to increase with the increase in the amount of BDSA, suggesting an increase in the incorporation of SO3H groups. This reveals that the amount of sulfonyl group, which is critical for the good behaviour of the film as a fuel cell membrane, can be effectively controlled by adjusting the mole ratio of the diamines with and without sulfonic acid groups.
S=O
S=O
a:6FDO 20
a:OMD 20
b:OMD 30
b:6FDO 30
c:OMD 40
c:6FDO 40
d:OMD 50
d:6FDO 50
1200
Two different series of sulfonated polyimides were synthesized with different chemical backbone structure; one from ODPA, BDSA and MDA and the other from 6FDA, BDSA and ODA. These two phthalic dianhydride based systems were compared with naphthalenic anhydride based systems. The diamine BDSA was used to incorporate the sulfonic acid group, and MDA and ODA were used as comonomer to control the content of sulfonyl incorporation. The polyimides were synthesized by conventional thermal imidization process and the curing process as given in Scheme 1. They are compared with napthalenic sulfonated polyimides for their degradation behaviour.
1000
1200
800
1000
800
Wavenumbers (cm-1) Fig. 1. Experimental apparatus for measuring conductivity of sulfonated copolyimide membranes.
435
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
The FT-IR spectra of NTDA based polyimides showed that the completion of imidization step was confirmed by the absence of anhydride carbonyl peak (around 1778 and 1737 cm1) and appearance of imide peak. Typical naphthalimide carbonyl absorptions at 1710 cm1 (asymmetric) and 1666 cm1 (symmetric) were observed which were lower than the phthalimide (for e.g. 1780 and 1720 cm1). The broad bands around 1178 and 1083 cm1 correspond to the SO2 stretching vibrations. The peak around 720 cm1 corresponds to the out-of-phase bending of the imide ring. The CeNeC stretching vibration was observed around 1342 cm1. 3.2. Thermal properties
3.3. Ion exchange capacity Table 1 shows the ion exchange capacity of the polyimide films. Generally, the IEC, proton conductivity and water uptake of the polyimide membrane increases with increase in sulfonic acid groups [3,29,30]. However, with the increase in the amount SO3H groups, the polymer become more hydrophilic and the stability towards water and other mechanical properties decreases. Hence it is essential to optimise the amount of sulfonic acid groups such that it contains only required amount. The sulfonated polyimides ODM 20e50 showed IEC of 0.73, 1.05, 1.42 and 1.74 meq/g, respectively, and the sulfonated polyimides 6FDO 20e50 showed IEC of 0.68, 0.92, 1.365 and 1.69 meq/g, respectively. The IEC values increased with increase in the amount of BDSA. Table 1 shows that ODM 50, ODM 40, 6FDO 50 and 6FDO 40 exhibited higher IEC than the reported IEC value of NafionÒ 115 (0.91 meq/g, [31]). These polymers might
110
110
100
100
90
90
80
80
70
70
Weight (%)
Weight (%)
The thermal stability of the five-membered sulfonated polyimides and the NafionÒ 115 was analysed by thermogravimetric analysis and the results are given in Fig. 2. All the polyimide thin films exhibited a three-step degradation pattern. The first weight loss observed around 100 C was due to the loss of water molecules, absorbed by the highly hygroscopic SO3 groups [18,19]. The second step of degradation was observed around 300 C due to the decomposition of sulfonyl groups, leading to expulsion of sulfur monoxide and sulfur dioxide gases [18,19,28]. The onset of the second step is almost similar for both the NafionÒ 115 and the sulfonated polyimides. The similar onset points for the two series of polyimides and the NafionÒ 115 suggest that irrespective of the polymer backbone, the degradation of the sulfonyl groups occurs around 300 C. Also, the percentage of weight loss in the second step corresponding to the elimination of SO3 groups increase with increase in BDSA content. This also proves that by altering the mole ratios of the diamines with and without sulfonyl groups we can effectively alter the amount of sulfonyl
incorporation. The third step indicates the decomposition of the polymer backbones. All the polyimides showed better thermal stability than the NafionÒ 115. For the NafionÒ 115, the onset for third step of degradation started around 380 C and there was a rapid weight loss observed beyond 380 C. But the polyimides showed an onset around 500 C indicating their better thermal stability. Thermal stability of the naphthalenic polyimides was higher than that of phthalic polyimides by showing third onset around over 500 C. This can be attributed to the more strain-free six-membered configuration of the polyimide than the strained five-membered ring structure. These results show that the synthesized sulfonated polyimide films can be used as polymer electrolyte membranes for operation at high temperature due to their thermal stability being higher than the intended operational temperature around 130e150 C.
60 50 ODM 50 ODM 40 ODM 30 ODM 20 Nafion115
40 30 20
50 6FDO 50 6FDO 40 6FDO 30 6FDO 20 Nafion115
40 30 20 10
10 0
60
0
100
200
300
400
Temperature
500
(oC)
600
700
0
0
100
200
300
400
500
600
Temperature (oC)
Fig. 2. TGA thermograms of sulfonated copolyimide (ODM and 6FDO) and NafionÒ 115 membranes.
700
436
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
Table 1 Ion exchange capacity and water sorption of sulfonated copolyimides (ODM and 6FDO) and NafionÒ 115 membranes Polymer
Nafion 115 ODM 20 ODM 30 ODM 40 ODM 50 6FDO 20 6FDO 30 6FDO 40 6FDO 50
Experimental IEC (meq/g)
Water uptake (%)
0.730 1.050 1.420 1.740 0.680 0.920 1.365 1.690
25 C
40 C
60 C
80 C
22.90 6.69 8.98 15.84 17.37 5.16 7.21 12.80 14.00
24.32 6.92 10.57 16.02 18.98 5.72 8.43 12.16 15.14
25.80 7.38 11.02 16.84 21.90 6.25 8.62 12.64 15.80
29.11 7.81 11.37 18.24 22.78 8.08 9.57 14.14 17.52
contain more sulfonyl groups than the NafionÒ 115 and hence resulted in higher IEC value. Between the two types of polyimides, for the same degree of sulfonyl substitution, ODM showed higher IEC compared to the 6FDO. The molecular weight of the repeating units in ODM is 472.438 g/mol for the MDA segment and 618.538 g/mol for the BDSA segment. For the 6FDO it is 608.458 g/mol for the ODA segment and 752.588 g/mol for the BDSA segment. It can be seen that, the molecular weight of the repeating segments in ODM is lesser than the 6FDO and hence, in a unit volume, more number of SO3H groups will be present in the ODM than 6FDO. Hence, at a given degree of sulfonization, ODM showed better IEC value than 6FDO. The same trend was obtained in the proton conductivity and water uptake experiments. ODM showed better conductivity and higher water uptake compared to 6FDO at a given sulfonyl concentration.
(Table 1 and Fig. 3). The polymer films were kept in a humidified chamber at the mentioned temperature for 6 h before evaluating the water uptake properties and the weight of the polyimide membrane before and after humidification was measured by a Cahn microbalance. At a constant temperature the amount of water uptake with time is measured at 10 min interval up to 4 h. All the samples exhibited rapid water uptake initially and reached an equilibrium beyond which the variation in the water uptake is small with time. At 25 C, NafionÒ 115 showed a water uptake of 22.9%. The ODM 20e50 exhibited an increase in water uptake from 6.7 to 17.4% where as the 6FDO 20e50 showed increase in the water uptake from 5.2 to 13.2%. This is because the amount of water uptake increases with increase in the amount of sulfonyl group [3,21]. Generally, the proton conductivity of the polymer increase with increase in water uptake. This is because with the more water uptake it improves the formation of the hydrophilic domain carrying the proton conductivity [36]. The variation in the water uptake with the increase in temperature is given in Fig. 3. For NafionÒ 115 and all the polyimides, a linear increase in water uptake with increase in temperature was observed. This may be because, with the increase in temperature the molecular mobility increases and hence the water molecules can easily penetrate through the polymer backbone and hydrate a maximum number of SO3H groups. Between the two series of polyimides the ODM showed a higher uptake of water than the 6FDO. This is because of the high hydrophobic interactions of the fluorine groups present in the backbone of the 6FDO polymers compared to ODM.
3.4. Water uptake 3.5. Proton conductivity Water uptakes of the polyimide films at four different temperatures namely, 25 C, 40 C, 60 C and 80 C were evaluated and compared with that of NafionÒ 115
The conductivities for all the sulfonated polyimides and the NafionÒ 115 were determined at four different 30
30 Nafion115
25
Nafion115
ODM 50
20
ODM 40
15
25
20 6FDO 50
15
6FDO 40
10
6FDO 30
ODM 30
10
ODM 20
6FDO 20
5
5
0 20
30
40
50
60
Temperature
70
(oC)
80
90
0 20
30
40
50
60
70
80
90
Temperature (oC)
Fig. 3. Water sorption of sulfonated copolyimide (ODM and 6FDO) and NafionÒ 115 membranes as a function of the temperature.
437
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
temperatures namely 25 C, 40 C, 60 C and 80 C (Fig. 4). Fig. 4 shows that there is increase in the conductivity with increase in the temperature for all the polyimides and NafionÒ 115. The increase in temperature favours both the dynamics involved in proton transfer and structural reorganization, which are required for fast proton conductivity. With increase in temperature both the proton diffusion and molecular diffusion merge together and result in the increased proton conductivity. However, it is reported that molecular diffusion predominates at elevated temperatures like 300 C [32]. In both series conductivity increased with increase in sulfonic acid content as expected. The sulfonated polyimides ODM 40 and 50 and 6FDO 40 and 50 showed better conductivity to be used as PEM in fuel cell than the reported sulfonated poly sulfone (SPSF), phosphotungstic acid (PWA) and polybenzimidazole (PBI) membranes [33,34]. It can be seen from Table 1 and Fig. 4 that the polyimides ODM 50 and 40 and 6FDO 50 and 40 showed a higher IEC than the NafionÒ 115 but the conductivity was less than NafionÒ 115. Generally, materials with higher IEC are expected to show better conductivity. The observed lower conductivity than the NafionÒ 115 despite the higher IEC value may be due to the difference in chemical structure and morphology. NafionÒ 115 having an ethylenic backbone has a better rotational freedom compared to the rigid, aromatic polyimide backbone. Hence the NafionÒ 115 will have loose molecular packing, when compared to the densely packed polyimides, which makes the passage of proton through the polymer membrane much easier. The rigidity of the NafionÒ 115 is less compared to that of polyimides and hence it is more flexible facilitating the easy formation of ionic clusters [35] and also the mobility of the main chain of the NafionÒ 115 is more compared to that of polyimides due to less molecular
interactions and much flexibility. The presence of groups like OCaO, N, eOe, CeH involve in more molecular interactions than the hydrophobic CeF backbone and hence restrict the molecular mobility. Moreover, NafionÒ 115, due to its high phase separation between the hydrophilic SO3H groups and hydrophobic backbone, augments the formation of ionic cluster and hence the ionic cluster density is higher compared to nonphase separated imides. Due to these factors the ionic conductivity of the NafionÒ 115 is higher compared to sulfonated polyimides at all temperatures but the conductivity of NafionÒ 115 decrease in elevated temperatures. These results prove that the ionic conductivity is not solely dependent on the IEC of the polymers. Between the two types of polyimides 6FDO and ODM, ODM showed better conductivity due to the less molecular weight of the repeating unit as explained in IEC. This shows that apart from the degree of sulfonation, the morphology also affects the proton conductivity. 3.6. Hydrolytic stability The polyimides were evaluated for the hydrolytic stability by soaking in water at 80 C. Their corresponding stability in water was followed by monitoring the intensity of the asymmetric stretching vibration of imide carbonyl peak at 1784 cm1 in FT-IR. Genies et al. reported the variation in hydrolytic stability of imides with naphthalenic and phthalic ring structure using NMR studies. In that study the polyimide chain scissions were followed by 1H NMR and 13C NMR in which the appearance of new peaks in 13C NMR have been found and correlated. The effect of chemical backbone structure on the hydrolytic stability of the system was systematically studied. The five-membered sulfonated polyimides were soaked in water at 80 C and their structural changes
Nafion115 ODM 50
1e-2
1e-2
ODM 40
1e-3
Conductivity (S/cm)
Conductivity (S/cm)
Nafion115
SPSF [34]
PBI (PWA60) [33]
ODM 30
1e-4
6FDO 50 6FDO 40
1e-3
SPSF [34]
PBI (PWA60) [33]
6FDO 30
1e-4 ODM 20
6FDO 20
1e-5 20
30
40
50
60
Temperature
70
(oC)
80
90
100
1e-5 20
30
40
50
60
70
80
90
100
Temperature (oC)
Fig. 4. Proton conductivity of sulfonated copolyimide (ODM and 6FDO) and NafionÒ 115 membranes as a function of the temperature.
438
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
C=O (asymmetric) 0h
C=O (symmetric)
C-N-C
S=O (symmetric) S=O (asymmetric)
30min
60min
Transmittance
with time were studied by FT-IR at specific time intervals and are given in Figs. 5 and 6. Both series showed a decrease in the intensity of the carbonyl absorption of the imide around 1784 cm1 with increase in time indicating the rupture of the imide rings of the polymer. Also the peak around 1038 cm1, which is due to the SO2 vibration, decreased with increase in time. This clearly shows that there is a decrease in the amount of SO3H groups in the polymer backbone. It is well known that there is a decrease in ionic conductivity of the sulfonated polyimide membranes with time, which was attributed to the polymer degradation. The degradation of the polyimides must be caused by the hydrolysis of imide structure. In our study with FT-IR we found the decrease in the intensity of SO2 vibrations. Due to chain scission, the large polymer molecules are broken down to smaller molecules, which were leached out by water, accounting for the loss of SO3H groups. Hence these polyimides suffer a loss in molecular weight which account for the brittleness and loss of mechanical strength of the polyimide membranes after hydrolysis which was observed by several investigators. The possible hydrolytic degradation mechanism which account for the loss of sulfonic acid groups is given in Fig. 7. The careful FT-IR analysis reveals that absorption due to both imide carbonyl and sulfonic acid decreased. But the decrease in imide carbonyl absorption was rapid and started at 30 min whereas the loss in sulfonyl group absorption was observed after 4 h. This indicate that first one carbonyl group of polyimides are attacked leading to the formation of amic acid structure in which although there is loss in imide carbonyl absorption the polymer chain is intact and hence there is no loss
90min 2h
4h
24h
2000
1800
1600
1400
1200
1000
800
600
400
Wave number (cm-1) Fig. 6. FT-IR spectra of the sulfonated polyimides derived from 6FDO.
in sulfonyl group. With increase in time the second carbonyl group also undergoes nucleophilic attack by OH group leading to the formation of acid and amine structure. This led to chain scission and loss of sulfonyl absorption. The chemical backbone structure of the two different series of five-membered ring polyimide affected the hydrolytic stability. When compared at a specific time, the 6FDO series showed better water stability than the
O
O C=O (symmetric) C-N-C
0h
N
N
S=O (asymmetric)
C
C
O
2h
Transmittance
C
C
S=O (symmetric)
C=O (asymmetric)
HO
OH HO
4h
C
OH
O
36h
O
C
N H C OH
24h
N H C OH O
O 52h
O C
OH
C
OH
+ 2000
O
1800
1600
1400
1200
1000
800
600
400
Wave number (cm-1) Fig. 5. FT-IR spectra of the sulfonated polyimides derived from ODM.
H2N
O Fig. 7. Mechanism of hydrolytic degradation in polyimides.
439
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
4. Conclusion A series of polyimides was synthesized and characterized to evaluate the suitability for the fuel cell application. Sulfonyl groups are used as hydrophilic nanophores to affect transportation of protons dissociated by the dynamics of water. The level of sulfonyl substitution was effectively controlled and revealed by the FT-IR and
6h
21h
Transmittance
43h
80h
S=O (symmetric) C=O (symmetric) S=O (asymmetric) C=O (asymmetric)
2000
1800
1600
C-N-C
1400
1200
1000
800
Wavenumber (cm-1) Fig. 8. FT-IR spectra of the sulfonated polyimides derived from NTDA-BDSA/ODA.
6h
21h
Transmittance
ODM series. This may be due to the highly hydrophobic nature of the fluorine backbone, which resists the approach of water to a greater extent and hence prevents the rapid attack of water, increasing the overall stability. When the polyimides from six-membered and fivemembered anhydrides are compared, the six-membered rings exhibited pronounced stability. When the intensity of imide carbonyl peaks at 1784 cm1 was compared for a specific time, for e.g. 4 h, the six-membered anhydride based system exhibited higher intensity than the fivemembered anhydride system (Figs. 8 and 9). The sixmembered polyimides in contrast showed no decrease in carbonyl absorption or loss in SO2 absorption. Moreover, the polyimides retained their mechanical integrity until three days against 52 h of five-membered analogues. This indicates that six-membered anhydrides based polyimides were more intact and hence have more hydrolytic stability than the five-membered ones. This must be due to the fact that the six-membered rings have a strain-free conformation compared to five-membered ring structures. This is the reason for the high temperature reaction conditions for the naphthalenic polyimides.
43h
80h S=O (symmetric)
C=O (symmetric) S=O (asymmetric) C=O (asymmetric) C-N-C
2000
1800
1600
1400
1200
1000
800
Wavenumber (cm-1) Fig. 9. FT-IR spectra of the sulfonated polyimides derived from NTDA-BDSA/MDA.
TGA studies. The intensity of the sulfonyl absorption band increased with the increase in the amount of BDSA. Also the percentage weight loss in the second step, which is due to the decomposition of SO3 group, increased with increase in BDSA content. Both confirmed that the percent sulfonization could be altered by using comonomers without SO3H group. The IEC values were found to depend on percent sulfonization and as well as on the molecular weight of the repeating unit. The proton conductivity measurements showed that the flexibility of the polymer backbone and phase separation facilitates more ionic conduction. Water uptake values showed temperature dependence. Rapid water uptake was observed initially and reaches an equilibrium beyond which the variation in the water uptake is small with time. Hydrophobicity of the polymer backbone affects the amount of water uptake. 6FDO series showed better hydrolytic stability than the ODM due to the hydrophobic backbone. When the intensity of imide carbonyl peaks at 1784 cm1 was compared for a specific time, the six-membered anhydrides based system exhibited higher intensity than the five-membered anhydrides system. The six-membered polyimides showed no decrease in carbonyl absorption or loss in SO2 absorption for a longer time. Moreover, the polyimides retained their mechanical integrity until three days against 52 h for five-membered analogues. This indicates that six-membered anhydrides based polyimides were more intact and hence have more hydrolytic stability than the five-membered ones. This must be due to the fact that the six-membered rings have a strain-free conformation compared to five-membered ring structures. This is the reason for the high temperature reaction conditions for the naphthalenic
440
W. Jang et al. / Polymer Degradation and Stability 90 (2005) 431e440
polyimides. Naphthalenic polyimides showed the possibility to be suitable for fuel cell with higher hydrolytic stability and thermal stability than the phthalic polyimides.
Acknowledgement This work was supported by the Ministry of Science and Technology of Korea through the National Research Laboratory Program.
References [1] Costamanga P, Srinivasan S. J power sources 2001;102:242. [2] Larmine J, Ducks A, editors. Fuel cell systems explained. Wiley publications; 2000. p. 1e15. [3] Poppe D, Frey H, Kruer KD, Heinzel A, Mulhaupt R. Macromolecules 2002;35:7936. [4] Yeo RS, Chin DT. J Electrochem Soc 1980;127:546. [5] Wilson MS, Gottesfeld S. J Appl Electrochem 1992;22:1. [6] Ren XM, Wilson MS, Gottesfeld S. J Electrochem Soc 1996; 143:12. [7] Hickner MA, Ghassemi H, Kim YS, Einsla BR, Mcgrath JE. Chem Rev 2004;104:4587 [and references therein]. [8] Savadogo O. J Power Sources 2004;127:135 [and references therein]. [9] Miyatake K, Fukushima K, Takeoka S, Tsuchida E. Chem Mater 1999;11:1172. [10] Hasiotis C, Deimede V, Kontoyannis C. Electrochim Acta 2001;46:401. [11] Caretta N, Tricoli V, Piccioni F. J Membr Sci 2000;166:189. [12] Notle R, Ledjeff K, Bauer M, Mulhaupt R. J Membr Sci 1993;83:211.
[13] Bailey C, William DJ, Karasz FE, Macknight WJ. Polymer 1987;28:1009. [14] Ghosh MK, Mittal KL, editors. Polyimides fundamental and application. NY: Marcel Dekker Inc.; 1996. [15] Vallejo E, Pourcelly G, Gavach C, Mercier R, Pineri M. J Membr Sci 1999;60:127. [16] Lee C, Sundar S, Kwon J, Han H. J Polym Sci Part A Polym Chem 2004;42:3612. [17] Lee C, Sundar S, Kwon J, Han H. J Polym Sci Part A Polym Chem 2004;42:3621. [18] Fang J, Guo X, Harada S, Watari T, Tanaka K, Kita H, et al. Macromolecules 2002;35:9022. [19] Guo X, Fang J, Watari T, Tanaka K, Kita H, Okamoto K. Macromolecules 2002;35:6707. [20] Miyatake K, Oyaizu K, Tsuchida E, Hay AS. Macromolecules 2001;34:2065. [21] Ding J, Chuy C, Holdcroft S. Chem Mater 2001;13:2231. [22] Genius C, Mercier R, Sillion B, Petiaud R, Cornet N, Gebel G, et al. Polymer 2001;42:5097. [23] Guo X, Fang J, Tanaka K, Kita H, Okamoto K. J Polym Sci Part A Polym Chem 2004;42:1432. [24] Han H, Seo J, Ree M, Pyo SM, Gryte CC. Polymer 1998;39:2963. [25] Seo J, Lee A, Lee C, Han H. J Appl Polym Sci 2000;76:1315. [26] Seo J, Lee A, Oh J, Han H. Polymer J 2000;32:583. [27] Pavia DL, Lampman GR, Kriz GS, editors. Introduction to spectroscopy. NY: Harcourt college publishers; 2001. [28] Rikukawa M, Sauni K. Prog Polym Sci 2000;25:1463. [29] Miyatake K, Shouji E, Yamamoto K, Tsuchida E. Macromolecules 1997;30:2941. [30] Kobayashi T, Rikukawa M, Sanui K, Ogata N. Solid State Ionics 1998;106:219. [31] Yang C, Srinivasan S, Bocarsly AB, Tulyani S, Benziger JB. J Membr Sci 2004;237:145. [32] Kreuer KD. Chem Mater 1996;8:610. [33] Staiti P, Minutoli M, Hocevar S. J Power Sources 2000;90:231. [34] Dimitrova PG, Baradie B, Foscallo D, Poinsignon C, Sanchez JY. J Membr Sci 2001;185:59. [35] Eisenberg A, Hird B, Moore RB. Macromolecules 1990;23:4098. [36] Kreuer KD. J Membr Sci 2001;185:29.