- Email: [email protected]
Anion exchange membranes based on semi-interpenetrating polymer network of quaternized chitosan and polystyrene
Journal of Colloid and Interface Science 361 (2011) 219–225
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Anion exchange membranes based on semi-interpenetrating polymer network of quaternized chitosan and polystyrene Jilin Wang a,b, Ronghuan He a,⇑, Quantong Che a a b
Department of Chemistry, College of Sciences, Northeastern University, Shenyang 110819, China School of Petrochemical Engineering, Liaoning Shi Hua University, Fushun 113001, China
a r t i c l e
i n f o
Article history: Received 8 February 2011 Accepted 16 May 2011 Available online 23 May 2011 Keywords: Anion exchange membrane Semi-interpenetrating polymer network Quaternized chitosan Polystyrene Conductivity Strength Emulsion
a b s t r a c t Anion exchange membranes with semi-interpenetrating polymer network (semi-IPN) were prepared based on quaternized chitosan (QCS) and polystyrene (PS). The PS was synthesized by polymerization of styrene monomers in the emulsion of the QCS in an acetic acid aqueous solution under nitrogen atmosphere at elevated temperatures. The semi-IPN system was formed by post-cross-linking of the QCS. A hydroxyl ionic conductivity of 2.80 102 S cm1 at 80 °C and a tensile stress at break of 20.0 MPa at room temperature were reached, respectively, by the semi-IPN membrane containing 21 wt.% of the PS. The durability of the semi-IPN membrane in alkaline solutions was tested by monitoring the variation of the conductivity and the mechanical strength. The degradation of the conductivity at 80 °C was about 5% by immersing the membrane in a 1 mol L1 KOH solution at room temperature for 72 h and at 60 °C for 50 h, respectively. The tensile stress at break at room temperature could maintain about 20.0 MPa for the membrane soaking in a 10 mol L1 KOH solution at ambient temperature for more than 70 h. The water swelling of the semi-IPN membranes was discussed based on the stress relaxation model of polymer chains, and it obeyed the Schott’s second-order swelling kinetics. Ó 2011 Elsevier Inc. All rights reserved.
1. Introduction In recent years, many efforts have been devoted to develop anion exchange membranes (AEMs) for either alkaline fuel cells [1–3] or direct methanol fuel cells (DMFCs) [4,5]. Substitution of liquid electrolytes with solid membranes in the fuel cells offers benefits such as avoiding formation of precipitated carbonate, potential electrolyte weeping, and corrosion. For AEMs-based DMFCs, the alkaline electrolyte results in facile kinetics at cathode and anode [6,7], reduced methanol crossover [6], and more possibilities to use abundant non-Pt electrode catalysts [5,8] comparing to that of the DMFCs with the acidic electrolyte. Polymers are normally functionalized by grafting or copolymerization [9] to prepare the quaternized polymers. The polymers in quaternary ammonium form are afterward doping with bases to obtain the AEMs [10]. There are different polymers that have been quaternized and investigated for aiming as electrolytes in fuel cells such as chitosan-based composites [11,12], polysulfone [3,13], polyethersulfone matrix [14,15], and chloride polymer networks based on polyvinylbenzyl chloride [6] or polyepichlorhydrin polymers [1,16–18]. As one of the most abundant natural polymers, chitosan has been chosen as the polymer matrix to prepare membrane
⇑ Corresponding author. Fax: +86 024 83676698. E-mail address: [email protected] (R. He). 0021-9797/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2011.05.039
electrolyte due to its excellent membrane properties, low cost, and the feasibility to be modified easily with functional groups in the structure. The quaternization of chitosan can be performed with such as (2,3-epoxypropyl) trimethylammonium chloride (EPTMAC) by the conversion of amino groups into 2-hydroxypropyltrimethyl ammonium chloride [19]. The anion exchange conductor of quaternized chitosan (QCS) is thus obtained by replacement of the chloride ions with hydroxide ions. The high quaternization degree of chitosan results in not only high ionic conductivity, but also significant swelling of the QCS membranes and therefore worse membrane strength. To reinforce the strength of the chitosan-based membranes, dialdehydes such as glyoxal [20–22] and glutaraldehyde (GA) [11,12,22,23] are normally used to perform the cross-linking. The stable imine bonds between amine groups of the chitosan polymer and the aldehydic group of the glutaraldehyde are then formed [24,25]. The involved chitosan unities for cross-linking formation may belong, or not, to the same polymeric chain [24]. In addition, it was also found that GA is not an effective cross-linker when the quaternization degree of QCS is above 35% due to the high swelling and poor mechanical properties of the resultant membranes [23]. Besides dialdehydes, diethylene glycol diglycidyl ether (DEGDE) has also been used and it is able to cross-link chitosan molecules by anchoring two epoxy groups of DEGDE, respectively, on two amino groups in different chitosan chains to form a real covalent linkage [26,27]. However, both the quaternization and cross-linking of QCS take place on the sites of –NH2 groups [24–27]; a compromise is thus
220
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
needed in order to reach a reasonable conductivity as well as enough strength of the QCS membranes. Attempts have also been made to develop AEMs by the preparation of chitosan composite membranes with other polymers [19]. With the aim of enhancement of the mechanical strength of the polymer membranes, a technology has been developed to fabricate composite membranes with a structure of so-called semi-interpenetrating polymer network (semi-IPN) [28–31]. The semi-IPN comprises at least one of the networks, in which one or more linear and branched polymers penetrated on a molecular scale of the macromolecules. This technique has been used to improve the mechanical properties of the proton-conducting polymer membranes such as Nafion [32], sulfonated poly (ether ether ketone) [30], and poly (styrenesulfonic) acid membranes [33]. However, the AEMs based on the concept of semi-IPN are seldom to be reported [28,29]. We propose a route to prepare an AEM with a semi-IPN based on QCS and polystyrene (PS). The polymer PS is more hydrophobic than QCS and has excellent mechanical strength. The introduced PS and the formed semi-IPN system are expected to improve the mechanical properties of the composite AEM.
CH2OH O OH
CH 2OH O OH
O
NH2
CH3 CH3 N + Cl-
O
NH2
+
m
Chitosan
EPTMAC 85oC isopropanol
CH 2OH O OH
O
CH2OH O OH
O
Quaternization
+
O
O
m
NH 2
NH (CH 3)3N+CH2CHCH2 ClOH
CH3
O
QCS
GA
Cross-linking CH2OH O OH
2. Materials and methods 2.1. Materials
O
CH2OH O OH
m
N
NH
(2,3-Epoxypropyl) trimethylammonium chloride (EPTMAC, purity P95%) was purchased from Shandong GuoFeng Fine Chemistry Factory. Chitosan and aqueous GA (50 wt.%) were obtained from China National Medicines Corporation Ltd. The deacetylation degree of the chitosan was 95%, which was determined according to the Ref. [34]. All the reagents used were analytical grade. The QCS in chloride form was synthesized by quaternization of chitosan in isopropanol with EPTMAC at 85 °C for 10 h (Scheme 1). The product was afterward isolated and purified according to the reference described elsewhere [35]. The quaternization degree of the QCS was about 75%, which determined by titration with a standard AgNO3 solution [36].
O
(CH3)3N +CH2CHCH2 ClOH CH2OH O OH NH
OH
N O
O O CH2OH
m
(CH3)3N+CH2CHCH 2 ClOH
Cross-linked QCS 2.2. Preparation of QCS/PS-based semi-IPN membranes In a 100-mL round-bottom flask equipped with a magnetic stirrer and a condenser, 2.0 g of chloride form QCS was dissolved in 40 mL of 2% (v/v) acetic acid aqueous solution under nitrogen atmosphere at room temperature. A certain amount of styrene was added into the flask to form a white emulsion under stirring. A potassium persulfate (0.8%, g/v) aqueous solution was then added dropwise to the mixture at a feeding rate of 0.5 mL min1 to initiate the polymerization of the styrene. The mixture was afterward refluxed at 100 °C for 4 h. After cooling to room temperature, the resulted mixture containing QCS and PS was poured on a glass plate to cast membranes. After evaporation of the solvent at room temperature, the obtained membrane as well as the glass plate were soaked in 100 mL 10% (v/v) GA aqueous solution at room temperature for 1 h to cross-link the QCS (Scheme 1) and to form the semi-IPN film. The film was then immersed in a 1.0 mol L1 KOH solution at room temperature for 24 h to obtain the hydroxide form QCS. The hydroxide form semi-IPN membranes of QCS/PS were washed thoroughly with deionized water and then dried at 80 °C until a constant weight was reached. The mass percent of PS (X) in the membrane was estimated by comparing the difference between the total amount of the styrene monomers added and the quantity of the styrene monomers without reaction. The unpolymerized styrene was measured with a gas chromatography (GC-6820, Aglient). A certain amount of the mixture of QCS and PS for PS polymerization was taken into a flask, and the
Scheme 1. Quaternization and cross-linking of chitosan.
air in the flask was excluded with nitrogen. The flask was then sealed, and it was afterward heated at 80 °C in a water bath for about 10 min to gasify the styrene monomers for determination. The membrane was termed as semi-IPN-X accordingly. 2.3. Instruments and techniques The FT-IR spectra of the membrane samples were recorded on a Perkin–Elmer spectrum one (B) spectrometer (Perkin–Elmer, America), and all the samples were prepared as the KBr pellets. The cross-section SEM images of the samples were taken with a SSX-550 scanning electron microscope (Shimadzu, Tokyo, Japan) by gold coating. Thermogravimetric analysis (TGA) was performed on a TGA 290C analyzer (Netzsch Company, Germany) at a heating rate of 10 °C min1 under nitrogen atmosphere. The mechanical strength of the membranes was determined with an instrument CMT6502 (SANS Company, China). Dumbbell-shaped membrane samples of 25 mm 4 mm were prepared, and the measurements were taken by setting a constant separating speed of 5.00 mm min1 in air at the ambient temperature. The tensile stress at break E was calculated by Eq. (1) [37].
E¼
F A0
ð1Þ
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
where F is the applied force at break, A0 is the initial cross-section area of the sample, which is equal to 4 L mm2, and L is the thickness of the membrane.
resistance of the membrane (RM) by comparing the difference. The anionic conductivity (r) of the membrane was calculated by
r ðS cm1 Þ ¼ 2.4. Water uptake and swelling of the membranes The membrane samples were soaked in distilled water at room temperature. The variations on weight and dimensions of the membranes were monitored at different times to study the swelling kinetics of the semi-IPN membranes. For measuring of the saturated water uptake, the membranes were immersed in distilled water for 24 h. The weight and the dimensions of the wet membrane (Pwet) were measured rapidly after wiping the excessive surface water with a tissue paper, and those of the dry membranes (Pdry) was obtained by drying the samples at 50 °C in a vacuum oven until a constant weight was reached. The water uptake and swelling of the membranes were determined by
water uptake or swelling ð%Þ ¼
Pwet Pdry 100% Pdry
ð2Þ
2.5. Ion exchange capacity The semi-IPN membrane in hydroxide form was washed thoroughly with distilled water and then dried in a vacuum oven at 60 °C to reach a constant weight (wOH in gram). It was afterward immersed in a 0.1 mol L1 HCl standard solution at ambient temperature for 48 h under stirring to neutralize the hydroxide ions containing in the membrane. The mole number (equivalent) of the neutralized hydroxide ions (n) was determined by titration of the acid solution with a 0.1 mol L1 KOH standard solution. The ion exchange capacity (IEC) of the semi-IPN membrane, termed as mili-equivalents (meq) of hydroxide ions per gram of the dry hydroxide form membrane, was obtained by calculation with Eq. (3) [38].
IEC ðmeq g1 Þ ¼
1000n wOH
ð3Þ
2.6. Ionic conductivity The conductivity measurement cell designed based on the Ref. [39] was illustrated in Fig. 1. The alternative current (AC) with a frequency of 2 kHz was supplied via a pair of platinum electrodes. A 0.1 mol L1 KOH aqueous solution was used as the electrolyte of the cell, and meanwhile, it was used to maintain the full hydration of the membrane sample. The resistance between the two electrodes was measured with and without the membrane, respectively, to obtain the
221
l RM S
ð4Þ
where l is the thickness of the membrane (cm) and S is the membrane surface (cm2) for ion transport, respectively. The temperature was controlled by putting the conductivity measurement cell in an oven. The membrane samples were previously swelled in the electrolyte solution before measurement in order to avoid deviation on the conductivity value due to the distortion of the membrane by swelling. 2.7. Membrane stability The stability of the membrane in alkaline medium was investigated by monitoring both the conductivity and the strength of the membranes as a function of time of soaking in KOH solutions [40]. The conductivity of the membranes immersed in 1.0 mol L1 KOH at both room temperature and 60 °C for different times, respectively, was determined using the same cell schematically described in Fig. 1. Before making the measurements, the membrane was washed with deionized water thoroughly to remove the absorbed potassium hydroxide. The stability in mechanical strength of the membrane in alkaline medium was evaluated by soaking the membrane in a 10 mol L1 KOH solution at ambient temperature for different times and testing the tensile stress of the membrane correlatively. 3. Results and discussion 3.1. Production of QCS/PS-based semi-IPN membranes The FT-IR spectra of the semi-IPN-X (X = 0, 8, 21) membranes as well as the pristine chitosan are shown in Fig. 2. The two absorption peaks at 1660 and 1590 cm1 for commercial chitosan are assigned to the C@O stretch of the secondary amide and the N–H bending of the primary amine, respectively [35]. The C–H bending of trimethylammonium group is registered at 1480 cm1 for semiIPN-0 confirming the existence of the quaternary ammonium salt in the QCS [27], and this absorption is shielded by the stretching vibration of the skeletal aromatic rings of the PS in the spectra of semi-IPN-8 and semi-IPN-21 [28]. It is also observed that the peak corresponding to the primary amine (1590 cm1) of chitosan dwindles and a new peak at around 1640 cm1 for semi-IPN-X (X – 0) membranes was recorded, indicating the change of the primary amine to the secondary amine structure due to the reactions at NH2 sites on the chitosan chains [23]. The peaks at 3025, 1600, 1500, and 1493 cm1 in the spectra of the semi-IPN-8 and semi-IPN-21 are attributed to the stretching vibration of the
Fig. 1. Schematic representation of the conductivity measurement cell.
222
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
3.2. Thermal stability
Fig. 2. FT-IR spectra of chitosan and semi-IPN membranes: (a) chitosan, (b) semiIPN-0, (c) semi-IPN-8, and (d) semi-IPN-21.
skeletal aromatic ring of the PS [28]. These results demonstrate that quaternary amino groups have been effectively grafted onto the chitosan backbones to form the QCS, and the linear PS has been synthesized to form the semi-IPN system with the cross-linked QCS together. All the membranes prepared are translucent with a pale yellow color. Fig. 3 gives the cross-section scanning electron micrograph (SEM) of the semi-IPN-0 and semi-IPN-21 membranes in both chloride and hydroxide forms. It is seen the good adhesion at interfaces of the QCS and PS, and no cracks and no obvious phase separation are observed. This indicates that the polymerization of the styrene performed uniformly in the QCS network, and the semi-IPN membranes are compacted.
Thermal stability of the membranes was determined, and the results are represented in Fig. 4. The mass loss of the QCS (semiIPN-0) in a range from 55 °C to about 125 °C is due to the absorbed water by the polymer. Thereafter, the quaternary ammonium groups in the QCS decompose when the temperature is above 125 °C, i.e., 127.6 °C [27]. The decomposition of the PS occurs at around 250 °C, which is more stable than that of the QCS. We observe that the mass loss decreases with the increase in the PS percent in the semi-IPN-X membranes at the same temperature. Taking 200 °C as an example, the mass loss of the membranes is about 14% for semi-IPN-0, 9% for semi-IPN-8, and 5% for semi-IPN-21, respectively. Since the quaternization degree of the semi-IPN-X is the same (75%) and there are no interactions between the PS and QCS, the difference in mass loss of the membranes is presumably due to the increased more stable content of the PS in the membrane samples [23].
3.3. Mechanical properties Mechanical properties of the membranes including tensile stress and elongation at break are collected in Table. 1. The tensile stress at break of semi-IPN-X (X – 0) is in a range of 13.8–20.0 MPa, and the elongation at break is around 3–16%, accordingly. The semi-IPN-X (X – 0) membranes exhibit higher tensile stress but lower elongation than some organic membranes, but the mechanical properties are slightly lower than the hybrid membranes of poly (2,6-dimethyl-1,4-phenylene oxide) with SiO2, i.e., 20– 27 MPa and 7–47%, respectively [41]. The highest stress of 20.0 MPa is achieved by the semi-IPN-21 membrane. The results
Fig. 3. Cross-section SEM graphs of semi-IPN membranes (500): (a) chloride form semi-IPN-0, (b) chloride form semi-IPN-21, (c) hydroxide form semi-IPN-0, and (d) hydroxide form semi-IPN-21.
223
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
the PS. In addition, the dimensional changes of the membranes are decreased obviously in the presence of the PS (Table. 2). This indicates that the introduced hydrophobic PS not only improved the strength of the semi-IPN-X membranes, it also reduced the water uptake and thus decreased the membrane swelling as well. In order to investigate the stress relaxation model of polymer chains as a function of the absorbed water of the semi-IPN membranes, Schott’s second-order swelling kinetics were used and it is expressed as Eq. (5) [44].
t 1 t ¼ þ W t K s W 21 W 1
Fig. 4. TGA curves of the semi-IPN membranes at a heating rate of 10 °C min1 under nitrogen atmosphere.
Table 1 Tensile stress and elongation at break of the semi-IPN membranes. Semi-IPN-X
X = 0%
X = 8%
X = 14%
X = 21%
X = 29%
X = 38%
Tensile stress (MPa) Elongation (%)
12.5
13.8
18.6
20.0
18.0
17.5
47.1
15.6
11.8
8.1
6.1
3.3
indicate that the formation of the semi-IPN structure benefits the mechanical strength of the membranes. However, the increased percent of the PS results in the decreased elongation at break of the membranes as PS is less flexible and similar result has been reported [42,43].
ð5Þ
where Wt is the water uptake of the membrane at time t, Ks is the constant of the swelling rate, W1 is the saturated water uptake, and K s W 21 is the initial water uptaking rate of the membrane, which determines the entire swelling process. As shown in Fig. 6, the plots of t/Wt versus t exhibit linear curves with linear correlation coefficients above 0.98. The theoretical saturated water uptake (W1) can be calculated according to the slope of the curves. The initial swelling rate (K s W 21 ) as well as the swelling rate constant (Ks) can be obtained from the intercept of the curves. Those data are given in Table. 2. It is seen that the experimental saturated water uptakes of the semi-IPN-X (X – 0) membranes are in good agreements with those theoretical results indicating that the swelling process of semi-IPN membranes in water obeys the Schott’s swelling model. In addition, the swelling rate constant Ks increased with the increase in the hydrophobic PS component. A highest initial swelling rate is obtained by semi-IPN21 membrane. The initial swelling rate by water is related to the relaxation rate of the chain segments of the dry membrane involving such as the balances of the rigidity/flexibility and hydrophobic/ hydrophilic properties, the degree of polymer cross-linking, the distribution of amorphous/crystalline domains, and the thickness of the dry membrane [45]. The cross-linking degree of the
3.4. Water uptake It is known that quaternized chitosan can readily absorb water or even dissolve in it because of the introduced hydrophilic quaternary ammonium sites [19]. Therefore, the water uptake and the swelling of the semi-IPN membranes are of critical concerns when the membranes are used as the electrolyte in fuel cells. The water uptake of the semi-IPN membranes was given in Fig. 5. It is seen that all the semi-IPN membranes absorb water very fast and plateaus are reached within 5–10 min. A saturated water uptake as high as 555% is reached by the semi-IPN-0 membrane. With the increase in the PS percent, the water uptake of the membranes decreases gradually due to the enhanced hydrophobic properties by
Fig. 5. Water swelling of the semi-IPN membranes versus time.
Table 2 Swelling kinetic parameters and volume variation of the semi-IPN membranes.
a b c
Sample
Wsa
W1b
Ks c
K s W 21 c
Volume variation (%)
Semi-IPN-0 Semi-IPN-8 Semi-IPN-14 Semi-IPN-21 Semi-IPN-29 Semi-IPN-38
555 254 207 170 131 63
714 277 222 172 133 64
5.60 104 4.45 103 8.44 103 3.74 102 4.33 102 8.11 102
285.71 344.83 416.67 1111.11 769.23 333.33
187 92 65 46 42 37
Experimental equilibrium swelling in gwater/gmembrane. Theoretical equilibrium swelling in gwater/gmembrane. Units in (gmembrane/gwater) min1.
Fig. 6. Plots of Schott’s second-order swelling kinetics for semi-IPN membranes.
224
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
membranes is the same, and the thickness of the membranes varies stochastically in a small range. Thus, the hydrophobic/hydrophilic balance by PS/QCS should be the dominate parameter. The swelling and water uptake of the semi-IPN membranes are decreased by comparing the data with those of the QCS (semi-IPN0). According to the results, the semi-IPN-21 membrane exhibits the fastest initial swelling rate, enough water uptake, and acceptable dimensional variation.
higher base doping as well as a bigger distortion of the membranes due to the swelling, which might affect the ionic conductivity especially at higher temperatures. This phenomenon has also been observed by other researchers [46,47]. Therefore, a relative lower concentration of the 0.1 mol L1 KOH was employed as the electrolyte to determine the conductivity of the membranes.
3.5. IEC and conductivity
The chemical stability of an anion exchange membrane based on quaternary ammonia is always a concern due to its limited durability in basic medium. The conductivities of the semi-IPN21 membranes at 80 °C were monitored after immersing the membrane in 1 mol L1 KOH solutions at room temperature and at 60 °C for different times (Fig. 8). As a reference, the variation in conductivities of semi-IPN-0 at room temperature was also showed in Fig. 8. However, the semi-IPN-0 membrane is too weak to perform the conductivity measurement after pretreated in the KOH solution at 60 °C. A slight degradation on the conductivity of the semi-IPN-21 membrane pretreated at both room temperature and 60 °C is observed from the figure. Taking the membranes pretreated at room temperature as an example, within 72 h, the decrease in the conductivity is about 5% for the semi-IPN-21 membrane, but it is about 10% for the semi-IPN-0 membrane, and a significant drop thereafter is observed although its original conductivity value is higher than that of the semi-IPN-21 membrane. The membrane degradation due to the nucleophilicity of OH is of a great concern for the quaternary ammoniums polymers. The well-known Hoffman degradation may cause the elimination of quaternary ammonium and thus low conductivity [48]. The substantive drop on the conductivity around 96 h for semi-IPN-0 is mainly attributed to the Hoffman degradation. The changes on the conductivity within prior 48 h are apparently due to the distortion of the membrane since the semi-IPN-0 swells significantly in the alkaline solution comparing with the semi-IPN-21 membrane. The presence of the hydrophobic PS reduces the membrane degradation, which probably due to the steric hindrance, however, further investigations are needed. The mechanical strength of the semi-IPN-21 membrane immersed in a 10 mol L1 KOH solution at room temperature for different times was tested, and the results are shown in Fig. 9. The tensile stress at break of the membrane at room temperature could maintain about 20.0 MPa by soaking in the KOH solution at ambient temperature for more than 70 h. However, the tensile stress at break of the QCS membrane (semi-IPN-0) was only stable within less than 48 h and it decreased to 2.5 MPa after immersing in the basic solution at room temperature for 96 h.
The IEC and the conductivity at 75 °C of the semi-IPN membranes are given in Table 3. As expected, the membranes with a higher PS exhibit lower value of IEC and thus lower ionic conductivity because the PS is a nonionic conductor. The ionic conductivity of the membranes at different temperatures is given in Fig. 7. The semi-IPN-X membranes exhibit increased conductivity under full hydration with the increase in the temperature. For an instance, the ionic conductivity of the semi-IPN-21 membrane increases from 6.10 103 to 2.80 102 S cm1 when the temperature changes from 30 to 80 °C. However, the presence of the PS results in the decreased conductivity. The activation energy (Ea) of the membrane conductance was calculated according to the slope of the Arrhenius plots, and the data are given in Table 3. It is seen from Fig. 7 that departures from the liner curves for the semi-IPN-X (X = 0, 8, 14, and 21) membranes are observed, which indicate that the activation energy is affected by the temperature. With the increase in the PS content, the activation energy is decreased and the influence of the temperature on the activation energy is becoming negligible. It is worthwhile to mention that the ionic conductivity of the semi-IPN membranes is affected by the concentration of the KOH electrolyte solution used for the measurements although the contribution of the KOH electrolyte has been subtracted. Our experimental results (without given here) indicated that a higher concentration of the KOH, i.e., 0.5 mol L1 KOH, would result in a
Table 3 IEC, conductivity, and activation energy of the semi-IPN membranes. Membrane
IEC (meq g1)
r at 75 °C (S cm1)
Ea (kJ mol1)
Semi-IPN-0 Semi-IPN-8 Semi-IPN-14 Semi-IPN-21 Semi-IPN-29 Semi-IPN-38
1.39 1.19 0.99 0.85 0.73 0.68
0.055 0.052 0.037 0.027 0.015 0.007
31.93 35.00 32.01 30.26 21.03 8.48
3.6. Stability of the semi-IPN-21 membrane in alkaline solutions
Conductivity (S cm-1)
0.07 0.06
a
0.05 0.04 0.03
b 0.02
c
0.01 0.00
0
50
100
150
200
250
300
Time (h)
Fig. 7. Arrhenius plots for the semi-IPN membranes.
Fig. 8. Stability of conductivities at 80 °C by pre-treating the membranes in 1 mol L1 KOH solution at room temperature, (a) semi-IPN-0, (b) semi-IPN-21, and at 60 °C, (c) semi-IPN-21, respectively.
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
Fig. 9. Degradation of the semi-IPN-0 and semi-IPN-21 membranes on tensile stress in 10 mol L1 KOH solution at room temperature.
According to the experimental results, we conclude that the semi-IPN structure allows the quaternary ammonia-based membranes to have a high tolerance to the bases. The hydrophobic PS in the semi-IPN should be of the benefit. 4. Conclusions Novel semi-IPN anion exchange membranes based on QCS and PS were prepared by polymerization of styrene monomers in the emulsion of the QCS in the acetic acid aqueous solution first and afterward by cross-linking of the QCS to form the PS interpenetrated QCS network. The physical–chemical properties of the semi-IPN membranes were investigated. The hydroxyl ionic conductivity at a level of 102 S cm1 was reached at 80 °C. A stress at break of 20 MPa at room temperature was achieved, and it could maintain for more than 70 h by soaking in the 10 mol L1 KOH solution at ambient temperature. The swelling kinetics of the membrane in water was found to obey the Schott’s second-order swelling model. The semi-IPN structure based on the hydrophobic PS provided the membranes both mechanical and chemical stabilities. Acknowledgments We are grateful for the financial supports by the PhD Program Foundation of Ministry of Education of China (20090042110002) and the Fundamental Research Funds for the Central University of China (Nos. 90105001 and 90605004). References [1] E. Agel, J. Bouet, F. Fauvarque, J. Power Sources 101 (2001) 267. [2] J.R. Varcoe, R.C.T. Slade, Fuel Cells 5 (2005) 187.
225
[3] S. Gu, R. Cai, T. Luo, Z.W. Chen, M.W. Sun, Y. Liu, G.H. He, Y.S. Yan, Angew. Chem., Int. Ed. 48 (2009) 6499. [4] E.H. Yu, K. Scott, J. Power Sources 137 (2004) 248. [5] Y. Wang, L. Li, L. Hu, L. Zhuang, J. Lu, B. Xu, Electrochem. Commun. 5 (2003) 662. [6] R.C.T. Slade, J.R. Varcoe, Solid State Ionics 176 (2005) 585. [7] Z. Ogumi, K. Matsuoka, S. Chiba, M. Matsuoka, Y. Iriyama, T. Abe, M. Inaba, Electrochemistry 70 (2002) 980. [8] A.E.S. Sleightholme, J.R. Varcoe, A.R. Kucernak, Electrochem. Commun. 10 (2008) 151. [9] M. Guo, J. Fang, H. Xu, W. Li, X. Lu, C. Lan, K. Li, J. Membr. Sci. 362 (2010) 97. [10] V.V. Shevchenko, M.A. Gumennaya, Theor. Exp. Chem. 3 (2010) 139. [11] Y. Wan, K.A.M. Creber, B. Peppley, V.T. Bui, J. Membr. Sci. 280 (2006) 666. [12] Y. Wan, K.A.M. Creber, B. Peppley, V.T. Bui, J. Membr. Sci. 284 (2006) 331. [13] J. Pan, S. F Lu, Y. Li, A.B. Huang, L. Zhuang, J.T. Lu, Adv. Funct. Mater. 20 (2010) 312. [14] J. Fang, P.K. Shen, J. Membr. Sci. 285 (2006) 317. [15] L. Li, Y. Wang, J. Membr. Sci. 262 (2005) 1. [16] N. Vassal, E. Salmon, J.F. Fauvarque, Electrochim. Acta 45 (2000) 1527. [17] D. Stoica, L. Ogier, L. Akrour, F. Alloin, J.F. Fauvarque, Electrochim. Acta 53 (2007) 1596. [18] N.P. Berezina, N.A. Kononenko, O.A. Dyomina, N.P. Gnusin, Adv. Colloid Interface Sci. 139 (2008) 3. [19] Y. Xiong, Q.L. Liu, Q.G. Zhang, A.M. Zhu, J. Power Sources 183 (2008) 447. [20] V.R. Patel, M.M. Amiji, Pharm. Res. 13 (1996) 588. [21] A. Testouri, C. Honorez, A. Barillec, D. Langevin, W. Drenckhan, Macromolecules 43 (2010) 6166. [22] R. Pauliukaite, M.E. Ghica, O.F. Filho, C.M.A. Brett, Electrochim. Acta 55 (2010) 6239. [23] Y. Wan, B. Peppley, K.A.M. Creber, V.T. Buia, E. Halliop, J. Power Sources 185 (2008) 183. [24] O.A.C. Monteiro, C. Airoldi, Int. J. Biol. Macromol. 26 (1999) 119. [25] J.Z. Knaul, S.M. Hudson, K.A.M. Creber, J. Polym. Sci., Part B: Polym. Phys. 37 (1999) 1079. [26] T. Uragami, M. Takuno, T. Miyata, Macromol. Chem. Phys. 203 (2002) 1162. [27] Y. Wan, B. Peppley, K.A.M. Crebera, V.T. Buia, J. Power Sources 195 (2010) 3785. [28] Y.J. Choi, M.S. Kang, S.H. Moon, J. Appl. Polym. Sci. 88 (2003) 1488. [29] S.J. Lee, S.S. Kim, Y.M. Lee, Carbohydr. Polym. 41 (2000) 197. [30] P.P. Chu, C.S. Wu, P.C. Liu, T.H. Wang, J.P. Pan, Polymer 51 (2010) 1386. [31] Q.W. Tang, X.M. Sun, Q.H. Li, J.H. Wu, J.M. Lin, J. Colloid Interface Sci. 339 (2009) 45. [32] K.Y. Cho, H.Y. Jung, S.S. Shin, N.S. Choi, S.J. Sung, J.K. Park, Electrochim. Acta 50 (2004) 589. [33] G.K.P. Surya, M.C. Smart, Q.J. Wang, A. Atti, V. Pleynet, B. Yang, J. Fluorine Chem. 125 (2004) 1217. [34] W. Xu, X. Liao, Chin. J. Anal. Lab. 27 (2008) 218. [35] S.H. Lim, S.M. Hudson, Carbohydr. Res. 339 (2004) 313. [36] J. Wu, Z.G. Su, G.H. Ma, Int. J. Pharm. 315 (2006) 1. [37] R.H. He, Q.F. Li, A. Bach, J.O. Jensen, N.J. Bjerrum, J. Membr. Sci. 277 (2006) 38. [38] R.H. Michael, A.H.M.A. Todd, K.M. Sarah, J.C. Chris, Chem. Mater. 20 (2008) 2566. [39] R. He, Q. Li, G. Xiao, N.J. Bjerrum, J. Membr. Sci. 226 (2003) 169. [40] J.H. Wang, S.H. Li, S.B. Zhang, Macromolecules 43 (2010) 3890. [41] J.Y. Luo, C.M. Wu, Y.H. Wu, T.W. Xu, J. Membr. Sci. 347 (2010) 240. [42] Z.B. Wang, H.L. Zhao, J. Zhao, X. Wang, J. Appl. Polym. Sci. 5 (2010) 2523. [43] Q.K. Zhang, S.J. Zhang, S.X. Chen, P.Y. Li, T.Y. Qin, S.G. Yuan, J. Colloid Interface Sci. 322 (2008) 421. [44] E.H. Hany, Eur. Polym. J. 43 (2007) 4830. [45] Y.H. Yin, Y.J. Yang, H.B. Xu, J. Appl. Polym. Sci. 83 (2002) 2835. [46] V.I. Zabolotsky, V.V. Nikonenko, N.D. Pimenskaya, J. Membr. Sci. 119 (1996) 171. [47] D. Elattar, A. Elmidoui, N. Pimenskaia, C. Gavach, G. Pourcelly, J. Membr. Sci. 143 (1998) 249. [48] C. Iojoiu, F. Chabert, M. Maréchal, N.El. Kissi, J. Guindet, J.Y. Sanchez, J. Power Sources 153 (2006) 198.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Anion exchange membranes based on semi-interpenetrating polymer network of quaternized chitosan and polystyrene Jilin Wang a,b, Ronghuan He a,⇑, Quantong Che a a b
Department of Chemistry, College of Sciences, Northeastern University, Shenyang 110819, China School of Petrochemical Engineering, Liaoning Shi Hua University, Fushun 113001, China
a r t i c l e
i n f o
Article history: Received 8 February 2011 Accepted 16 May 2011 Available online 23 May 2011 Keywords: Anion exchange membrane Semi-interpenetrating polymer network Quaternized chitosan Polystyrene Conductivity Strength Emulsion
a b s t r a c t Anion exchange membranes with semi-interpenetrating polymer network (semi-IPN) were prepared based on quaternized chitosan (QCS) and polystyrene (PS). The PS was synthesized by polymerization of styrene monomers in the emulsion of the QCS in an acetic acid aqueous solution under nitrogen atmosphere at elevated temperatures. The semi-IPN system was formed by post-cross-linking of the QCS. A hydroxyl ionic conductivity of 2.80 102 S cm1 at 80 °C and a tensile stress at break of 20.0 MPa at room temperature were reached, respectively, by the semi-IPN membrane containing 21 wt.% of the PS. The durability of the semi-IPN membrane in alkaline solutions was tested by monitoring the variation of the conductivity and the mechanical strength. The degradation of the conductivity at 80 °C was about 5% by immersing the membrane in a 1 mol L1 KOH solution at room temperature for 72 h and at 60 °C for 50 h, respectively. The tensile stress at break at room temperature could maintain about 20.0 MPa for the membrane soaking in a 10 mol L1 KOH solution at ambient temperature for more than 70 h. The water swelling of the semi-IPN membranes was discussed based on the stress relaxation model of polymer chains, and it obeyed the Schott’s second-order swelling kinetics. Ó 2011 Elsevier Inc. All rights reserved.
1. Introduction In recent years, many efforts have been devoted to develop anion exchange membranes (AEMs) for either alkaline fuel cells [1–3] or direct methanol fuel cells (DMFCs) [4,5]. Substitution of liquid electrolytes with solid membranes in the fuel cells offers benefits such as avoiding formation of precipitated carbonate, potential electrolyte weeping, and corrosion. For AEMs-based DMFCs, the alkaline electrolyte results in facile kinetics at cathode and anode [6,7], reduced methanol crossover [6], and more possibilities to use abundant non-Pt electrode catalysts [5,8] comparing to that of the DMFCs with the acidic electrolyte. Polymers are normally functionalized by grafting or copolymerization [9] to prepare the quaternized polymers. The polymers in quaternary ammonium form are afterward doping with bases to obtain the AEMs [10]. There are different polymers that have been quaternized and investigated for aiming as electrolytes in fuel cells such as chitosan-based composites [11,12], polysulfone [3,13], polyethersulfone matrix [14,15], and chloride polymer networks based on polyvinylbenzyl chloride [6] or polyepichlorhydrin polymers [1,16–18]. As one of the most abundant natural polymers, chitosan has been chosen as the polymer matrix to prepare membrane
⇑ Corresponding author. Fax: +86 024 83676698. E-mail address: [email protected] (R. He). 0021-9797/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2011.05.039
electrolyte due to its excellent membrane properties, low cost, and the feasibility to be modified easily with functional groups in the structure. The quaternization of chitosan can be performed with such as (2,3-epoxypropyl) trimethylammonium chloride (EPTMAC) by the conversion of amino groups into 2-hydroxypropyltrimethyl ammonium chloride [19]. The anion exchange conductor of quaternized chitosan (QCS) is thus obtained by replacement of the chloride ions with hydroxide ions. The high quaternization degree of chitosan results in not only high ionic conductivity, but also significant swelling of the QCS membranes and therefore worse membrane strength. To reinforce the strength of the chitosan-based membranes, dialdehydes such as glyoxal [20–22] and glutaraldehyde (GA) [11,12,22,23] are normally used to perform the cross-linking. The stable imine bonds between amine groups of the chitosan polymer and the aldehydic group of the glutaraldehyde are then formed [24,25]. The involved chitosan unities for cross-linking formation may belong, or not, to the same polymeric chain [24]. In addition, it was also found that GA is not an effective cross-linker when the quaternization degree of QCS is above 35% due to the high swelling and poor mechanical properties of the resultant membranes [23]. Besides dialdehydes, diethylene glycol diglycidyl ether (DEGDE) has also been used and it is able to cross-link chitosan molecules by anchoring two epoxy groups of DEGDE, respectively, on two amino groups in different chitosan chains to form a real covalent linkage [26,27]. However, both the quaternization and cross-linking of QCS take place on the sites of –NH2 groups [24–27]; a compromise is thus
220
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
needed in order to reach a reasonable conductivity as well as enough strength of the QCS membranes. Attempts have also been made to develop AEMs by the preparation of chitosan composite membranes with other polymers [19]. With the aim of enhancement of the mechanical strength of the polymer membranes, a technology has been developed to fabricate composite membranes with a structure of so-called semi-interpenetrating polymer network (semi-IPN) [28–31]. The semi-IPN comprises at least one of the networks, in which one or more linear and branched polymers penetrated on a molecular scale of the macromolecules. This technique has been used to improve the mechanical properties of the proton-conducting polymer membranes such as Nafion [32], sulfonated poly (ether ether ketone) [30], and poly (styrenesulfonic) acid membranes [33]. However, the AEMs based on the concept of semi-IPN are seldom to be reported [28,29]. We propose a route to prepare an AEM with a semi-IPN based on QCS and polystyrene (PS). The polymer PS is more hydrophobic than QCS and has excellent mechanical strength. The introduced PS and the formed semi-IPN system are expected to improve the mechanical properties of the composite AEM.
CH2OH O OH
CH 2OH O OH
O
NH2
CH3 CH3 N + Cl-
O
NH2
+
m
Chitosan
EPTMAC 85oC isopropanol
CH 2OH O OH
O
CH2OH O OH
O
Quaternization
+
O
O
m
NH 2
NH (CH 3)3N+CH2CHCH2 ClOH
CH3
O
QCS
GA
Cross-linking CH2OH O OH
2. Materials and methods 2.1. Materials
O
CH2OH O OH
m
N
NH
(2,3-Epoxypropyl) trimethylammonium chloride (EPTMAC, purity P95%) was purchased from Shandong GuoFeng Fine Chemistry Factory. Chitosan and aqueous GA (50 wt.%) were obtained from China National Medicines Corporation Ltd. The deacetylation degree of the chitosan was 95%, which was determined according to the Ref. [34]. All the reagents used were analytical grade. The QCS in chloride form was synthesized by quaternization of chitosan in isopropanol with EPTMAC at 85 °C for 10 h (Scheme 1). The product was afterward isolated and purified according to the reference described elsewhere [35]. The quaternization degree of the QCS was about 75%, which determined by titration with a standard AgNO3 solution [36].
O
(CH3)3N +CH2CHCH2 ClOH CH2OH O OH NH
OH
N O
O O CH2OH
m
(CH3)3N+CH2CHCH 2 ClOH
Cross-linked QCS 2.2. Preparation of QCS/PS-based semi-IPN membranes In a 100-mL round-bottom flask equipped with a magnetic stirrer and a condenser, 2.0 g of chloride form QCS was dissolved in 40 mL of 2% (v/v) acetic acid aqueous solution under nitrogen atmosphere at room temperature. A certain amount of styrene was added into the flask to form a white emulsion under stirring. A potassium persulfate (0.8%, g/v) aqueous solution was then added dropwise to the mixture at a feeding rate of 0.5 mL min1 to initiate the polymerization of the styrene. The mixture was afterward refluxed at 100 °C for 4 h. After cooling to room temperature, the resulted mixture containing QCS and PS was poured on a glass plate to cast membranes. After evaporation of the solvent at room temperature, the obtained membrane as well as the glass plate were soaked in 100 mL 10% (v/v) GA aqueous solution at room temperature for 1 h to cross-link the QCS (Scheme 1) and to form the semi-IPN film. The film was then immersed in a 1.0 mol L1 KOH solution at room temperature for 24 h to obtain the hydroxide form QCS. The hydroxide form semi-IPN membranes of QCS/PS were washed thoroughly with deionized water and then dried at 80 °C until a constant weight was reached. The mass percent of PS (X) in the membrane was estimated by comparing the difference between the total amount of the styrene monomers added and the quantity of the styrene monomers without reaction. The unpolymerized styrene was measured with a gas chromatography (GC-6820, Aglient). A certain amount of the mixture of QCS and PS for PS polymerization was taken into a flask, and the
Scheme 1. Quaternization and cross-linking of chitosan.
air in the flask was excluded with nitrogen. The flask was then sealed, and it was afterward heated at 80 °C in a water bath for about 10 min to gasify the styrene monomers for determination. The membrane was termed as semi-IPN-X accordingly. 2.3. Instruments and techniques The FT-IR spectra of the membrane samples were recorded on a Perkin–Elmer spectrum one (B) spectrometer (Perkin–Elmer, America), and all the samples were prepared as the KBr pellets. The cross-section SEM images of the samples were taken with a SSX-550 scanning electron microscope (Shimadzu, Tokyo, Japan) by gold coating. Thermogravimetric analysis (TGA) was performed on a TGA 290C analyzer (Netzsch Company, Germany) at a heating rate of 10 °C min1 under nitrogen atmosphere. The mechanical strength of the membranes was determined with an instrument CMT6502 (SANS Company, China). Dumbbell-shaped membrane samples of 25 mm 4 mm were prepared, and the measurements were taken by setting a constant separating speed of 5.00 mm min1 in air at the ambient temperature. The tensile stress at break E was calculated by Eq. (1) [37].
E¼
F A0
ð1Þ
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
where F is the applied force at break, A0 is the initial cross-section area of the sample, which is equal to 4 L mm2, and L is the thickness of the membrane.
resistance of the membrane (RM) by comparing the difference. The anionic conductivity (r) of the membrane was calculated by
r ðS cm1 Þ ¼ 2.4. Water uptake and swelling of the membranes The membrane samples were soaked in distilled water at room temperature. The variations on weight and dimensions of the membranes were monitored at different times to study the swelling kinetics of the semi-IPN membranes. For measuring of the saturated water uptake, the membranes were immersed in distilled water for 24 h. The weight and the dimensions of the wet membrane (Pwet) were measured rapidly after wiping the excessive surface water with a tissue paper, and those of the dry membranes (Pdry) was obtained by drying the samples at 50 °C in a vacuum oven until a constant weight was reached. The water uptake and swelling of the membranes were determined by
water uptake or swelling ð%Þ ¼
Pwet Pdry 100% Pdry
ð2Þ
2.5. Ion exchange capacity The semi-IPN membrane in hydroxide form was washed thoroughly with distilled water and then dried in a vacuum oven at 60 °C to reach a constant weight (wOH in gram). It was afterward immersed in a 0.1 mol L1 HCl standard solution at ambient temperature for 48 h under stirring to neutralize the hydroxide ions containing in the membrane. The mole number (equivalent) of the neutralized hydroxide ions (n) was determined by titration of the acid solution with a 0.1 mol L1 KOH standard solution. The ion exchange capacity (IEC) of the semi-IPN membrane, termed as mili-equivalents (meq) of hydroxide ions per gram of the dry hydroxide form membrane, was obtained by calculation with Eq. (3) [38].
IEC ðmeq g1 Þ ¼
1000n wOH
ð3Þ
2.6. Ionic conductivity The conductivity measurement cell designed based on the Ref. [39] was illustrated in Fig. 1. The alternative current (AC) with a frequency of 2 kHz was supplied via a pair of platinum electrodes. A 0.1 mol L1 KOH aqueous solution was used as the electrolyte of the cell, and meanwhile, it was used to maintain the full hydration of the membrane sample. The resistance between the two electrodes was measured with and without the membrane, respectively, to obtain the
221
l RM S
ð4Þ
where l is the thickness of the membrane (cm) and S is the membrane surface (cm2) for ion transport, respectively. The temperature was controlled by putting the conductivity measurement cell in an oven. The membrane samples were previously swelled in the electrolyte solution before measurement in order to avoid deviation on the conductivity value due to the distortion of the membrane by swelling. 2.7. Membrane stability The stability of the membrane in alkaline medium was investigated by monitoring both the conductivity and the strength of the membranes as a function of time of soaking in KOH solutions [40]. The conductivity of the membranes immersed in 1.0 mol L1 KOH at both room temperature and 60 °C for different times, respectively, was determined using the same cell schematically described in Fig. 1. Before making the measurements, the membrane was washed with deionized water thoroughly to remove the absorbed potassium hydroxide. The stability in mechanical strength of the membrane in alkaline medium was evaluated by soaking the membrane in a 10 mol L1 KOH solution at ambient temperature for different times and testing the tensile stress of the membrane correlatively. 3. Results and discussion 3.1. Production of QCS/PS-based semi-IPN membranes The FT-IR spectra of the semi-IPN-X (X = 0, 8, 21) membranes as well as the pristine chitosan are shown in Fig. 2. The two absorption peaks at 1660 and 1590 cm1 for commercial chitosan are assigned to the C@O stretch of the secondary amide and the N–H bending of the primary amine, respectively [35]. The C–H bending of trimethylammonium group is registered at 1480 cm1 for semiIPN-0 confirming the existence of the quaternary ammonium salt in the QCS [27], and this absorption is shielded by the stretching vibration of the skeletal aromatic rings of the PS in the spectra of semi-IPN-8 and semi-IPN-21 [28]. It is also observed that the peak corresponding to the primary amine (1590 cm1) of chitosan dwindles and a new peak at around 1640 cm1 for semi-IPN-X (X – 0) membranes was recorded, indicating the change of the primary amine to the secondary amine structure due to the reactions at NH2 sites on the chitosan chains [23]. The peaks at 3025, 1600, 1500, and 1493 cm1 in the spectra of the semi-IPN-8 and semi-IPN-21 are attributed to the stretching vibration of the
Fig. 1. Schematic representation of the conductivity measurement cell.
222
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
3.2. Thermal stability
Fig. 2. FT-IR spectra of chitosan and semi-IPN membranes: (a) chitosan, (b) semiIPN-0, (c) semi-IPN-8, and (d) semi-IPN-21.
skeletal aromatic ring of the PS [28]. These results demonstrate that quaternary amino groups have been effectively grafted onto the chitosan backbones to form the QCS, and the linear PS has been synthesized to form the semi-IPN system with the cross-linked QCS together. All the membranes prepared are translucent with a pale yellow color. Fig. 3 gives the cross-section scanning electron micrograph (SEM) of the semi-IPN-0 and semi-IPN-21 membranes in both chloride and hydroxide forms. It is seen the good adhesion at interfaces of the QCS and PS, and no cracks and no obvious phase separation are observed. This indicates that the polymerization of the styrene performed uniformly in the QCS network, and the semi-IPN membranes are compacted.
Thermal stability of the membranes was determined, and the results are represented in Fig. 4. The mass loss of the QCS (semiIPN-0) in a range from 55 °C to about 125 °C is due to the absorbed water by the polymer. Thereafter, the quaternary ammonium groups in the QCS decompose when the temperature is above 125 °C, i.e., 127.6 °C [27]. The decomposition of the PS occurs at around 250 °C, which is more stable than that of the QCS. We observe that the mass loss decreases with the increase in the PS percent in the semi-IPN-X membranes at the same temperature. Taking 200 °C as an example, the mass loss of the membranes is about 14% for semi-IPN-0, 9% for semi-IPN-8, and 5% for semi-IPN-21, respectively. Since the quaternization degree of the semi-IPN-X is the same (75%) and there are no interactions between the PS and QCS, the difference in mass loss of the membranes is presumably due to the increased more stable content of the PS in the membrane samples [23].
3.3. Mechanical properties Mechanical properties of the membranes including tensile stress and elongation at break are collected in Table. 1. The tensile stress at break of semi-IPN-X (X – 0) is in a range of 13.8–20.0 MPa, and the elongation at break is around 3–16%, accordingly. The semi-IPN-X (X – 0) membranes exhibit higher tensile stress but lower elongation than some organic membranes, but the mechanical properties are slightly lower than the hybrid membranes of poly (2,6-dimethyl-1,4-phenylene oxide) with SiO2, i.e., 20– 27 MPa and 7–47%, respectively [41]. The highest stress of 20.0 MPa is achieved by the semi-IPN-21 membrane. The results
Fig. 3. Cross-section SEM graphs of semi-IPN membranes (500): (a) chloride form semi-IPN-0, (b) chloride form semi-IPN-21, (c) hydroxide form semi-IPN-0, and (d) hydroxide form semi-IPN-21.
223
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
the PS. In addition, the dimensional changes of the membranes are decreased obviously in the presence of the PS (Table. 2). This indicates that the introduced hydrophobic PS not only improved the strength of the semi-IPN-X membranes, it also reduced the water uptake and thus decreased the membrane swelling as well. In order to investigate the stress relaxation model of polymer chains as a function of the absorbed water of the semi-IPN membranes, Schott’s second-order swelling kinetics were used and it is expressed as Eq. (5) [44].
t 1 t ¼ þ W t K s W 21 W 1
Fig. 4. TGA curves of the semi-IPN membranes at a heating rate of 10 °C min1 under nitrogen atmosphere.
Table 1 Tensile stress and elongation at break of the semi-IPN membranes. Semi-IPN-X
X = 0%
X = 8%
X = 14%
X = 21%
X = 29%
X = 38%
Tensile stress (MPa) Elongation (%)
12.5
13.8
18.6
20.0
18.0
17.5
47.1
15.6
11.8
8.1
6.1
3.3
indicate that the formation of the semi-IPN structure benefits the mechanical strength of the membranes. However, the increased percent of the PS results in the decreased elongation at break of the membranes as PS is less flexible and similar result has been reported [42,43].
ð5Þ
where Wt is the water uptake of the membrane at time t, Ks is the constant of the swelling rate, W1 is the saturated water uptake, and K s W 21 is the initial water uptaking rate of the membrane, which determines the entire swelling process. As shown in Fig. 6, the plots of t/Wt versus t exhibit linear curves with linear correlation coefficients above 0.98. The theoretical saturated water uptake (W1) can be calculated according to the slope of the curves. The initial swelling rate (K s W 21 ) as well as the swelling rate constant (Ks) can be obtained from the intercept of the curves. Those data are given in Table. 2. It is seen that the experimental saturated water uptakes of the semi-IPN-X (X – 0) membranes are in good agreements with those theoretical results indicating that the swelling process of semi-IPN membranes in water obeys the Schott’s swelling model. In addition, the swelling rate constant Ks increased with the increase in the hydrophobic PS component. A highest initial swelling rate is obtained by semi-IPN21 membrane. The initial swelling rate by water is related to the relaxation rate of the chain segments of the dry membrane involving such as the balances of the rigidity/flexibility and hydrophobic/ hydrophilic properties, the degree of polymer cross-linking, the distribution of amorphous/crystalline domains, and the thickness of the dry membrane [45]. The cross-linking degree of the
3.4. Water uptake It is known that quaternized chitosan can readily absorb water or even dissolve in it because of the introduced hydrophilic quaternary ammonium sites [19]. Therefore, the water uptake and the swelling of the semi-IPN membranes are of critical concerns when the membranes are used as the electrolyte in fuel cells. The water uptake of the semi-IPN membranes was given in Fig. 5. It is seen that all the semi-IPN membranes absorb water very fast and plateaus are reached within 5–10 min. A saturated water uptake as high as 555% is reached by the semi-IPN-0 membrane. With the increase in the PS percent, the water uptake of the membranes decreases gradually due to the enhanced hydrophobic properties by
Fig. 5. Water swelling of the semi-IPN membranes versus time.
Table 2 Swelling kinetic parameters and volume variation of the semi-IPN membranes.
a b c
Sample
Wsa
W1b
Ks c
K s W 21 c
Volume variation (%)
Semi-IPN-0 Semi-IPN-8 Semi-IPN-14 Semi-IPN-21 Semi-IPN-29 Semi-IPN-38
555 254 207 170 131 63
714 277 222 172 133 64
5.60 104 4.45 103 8.44 103 3.74 102 4.33 102 8.11 102
285.71 344.83 416.67 1111.11 769.23 333.33
187 92 65 46 42 37
Experimental equilibrium swelling in gwater/gmembrane. Theoretical equilibrium swelling in gwater/gmembrane. Units in (gmembrane/gwater) min1.
Fig. 6. Plots of Schott’s second-order swelling kinetics for semi-IPN membranes.
224
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
membranes is the same, and the thickness of the membranes varies stochastically in a small range. Thus, the hydrophobic/hydrophilic balance by PS/QCS should be the dominate parameter. The swelling and water uptake of the semi-IPN membranes are decreased by comparing the data with those of the QCS (semi-IPN0). According to the results, the semi-IPN-21 membrane exhibits the fastest initial swelling rate, enough water uptake, and acceptable dimensional variation.
higher base doping as well as a bigger distortion of the membranes due to the swelling, which might affect the ionic conductivity especially at higher temperatures. This phenomenon has also been observed by other researchers [46,47]. Therefore, a relative lower concentration of the 0.1 mol L1 KOH was employed as the electrolyte to determine the conductivity of the membranes.
3.5. IEC and conductivity
The chemical stability of an anion exchange membrane based on quaternary ammonia is always a concern due to its limited durability in basic medium. The conductivities of the semi-IPN21 membranes at 80 °C were monitored after immersing the membrane in 1 mol L1 KOH solutions at room temperature and at 60 °C for different times (Fig. 8). As a reference, the variation in conductivities of semi-IPN-0 at room temperature was also showed in Fig. 8. However, the semi-IPN-0 membrane is too weak to perform the conductivity measurement after pretreated in the KOH solution at 60 °C. A slight degradation on the conductivity of the semi-IPN-21 membrane pretreated at both room temperature and 60 °C is observed from the figure. Taking the membranes pretreated at room temperature as an example, within 72 h, the decrease in the conductivity is about 5% for the semi-IPN-21 membrane, but it is about 10% for the semi-IPN-0 membrane, and a significant drop thereafter is observed although its original conductivity value is higher than that of the semi-IPN-21 membrane. The membrane degradation due to the nucleophilicity of OH is of a great concern for the quaternary ammoniums polymers. The well-known Hoffman degradation may cause the elimination of quaternary ammonium and thus low conductivity [48]. The substantive drop on the conductivity around 96 h for semi-IPN-0 is mainly attributed to the Hoffman degradation. The changes on the conductivity within prior 48 h are apparently due to the distortion of the membrane since the semi-IPN-0 swells significantly in the alkaline solution comparing with the semi-IPN-21 membrane. The presence of the hydrophobic PS reduces the membrane degradation, which probably due to the steric hindrance, however, further investigations are needed. The mechanical strength of the semi-IPN-21 membrane immersed in a 10 mol L1 KOH solution at room temperature for different times was tested, and the results are shown in Fig. 9. The tensile stress at break of the membrane at room temperature could maintain about 20.0 MPa by soaking in the KOH solution at ambient temperature for more than 70 h. However, the tensile stress at break of the QCS membrane (semi-IPN-0) was only stable within less than 48 h and it decreased to 2.5 MPa after immersing in the basic solution at room temperature for 96 h.
The IEC and the conductivity at 75 °C of the semi-IPN membranes are given in Table 3. As expected, the membranes with a higher PS exhibit lower value of IEC and thus lower ionic conductivity because the PS is a nonionic conductor. The ionic conductivity of the membranes at different temperatures is given in Fig. 7. The semi-IPN-X membranes exhibit increased conductivity under full hydration with the increase in the temperature. For an instance, the ionic conductivity of the semi-IPN-21 membrane increases from 6.10 103 to 2.80 102 S cm1 when the temperature changes from 30 to 80 °C. However, the presence of the PS results in the decreased conductivity. The activation energy (Ea) of the membrane conductance was calculated according to the slope of the Arrhenius plots, and the data are given in Table 3. It is seen from Fig. 7 that departures from the liner curves for the semi-IPN-X (X = 0, 8, 14, and 21) membranes are observed, which indicate that the activation energy is affected by the temperature. With the increase in the PS content, the activation energy is decreased and the influence of the temperature on the activation energy is becoming negligible. It is worthwhile to mention that the ionic conductivity of the semi-IPN membranes is affected by the concentration of the KOH electrolyte solution used for the measurements although the contribution of the KOH electrolyte has been subtracted. Our experimental results (without given here) indicated that a higher concentration of the KOH, i.e., 0.5 mol L1 KOH, would result in a
Table 3 IEC, conductivity, and activation energy of the semi-IPN membranes. Membrane
IEC (meq g1)
r at 75 °C (S cm1)
Ea (kJ mol1)
Semi-IPN-0 Semi-IPN-8 Semi-IPN-14 Semi-IPN-21 Semi-IPN-29 Semi-IPN-38
1.39 1.19 0.99 0.85 0.73 0.68
0.055 0.052 0.037 0.027 0.015 0.007
31.93 35.00 32.01 30.26 21.03 8.48
3.6. Stability of the semi-IPN-21 membrane in alkaline solutions
Conductivity (S cm-1)
0.07 0.06
a
0.05 0.04 0.03
b 0.02
c
0.01 0.00
0
50
100
150
200
250
300
Time (h)
Fig. 7. Arrhenius plots for the semi-IPN membranes.
Fig. 8. Stability of conductivities at 80 °C by pre-treating the membranes in 1 mol L1 KOH solution at room temperature, (a) semi-IPN-0, (b) semi-IPN-21, and at 60 °C, (c) semi-IPN-21, respectively.
J. Wang et al. / Journal of Colloid and Interface Science 361 (2011) 219–225
Fig. 9. Degradation of the semi-IPN-0 and semi-IPN-21 membranes on tensile stress in 10 mol L1 KOH solution at room temperature.
According to the experimental results, we conclude that the semi-IPN structure allows the quaternary ammonia-based membranes to have a high tolerance to the bases. The hydrophobic PS in the semi-IPN should be of the benefit. 4. Conclusions Novel semi-IPN anion exchange membranes based on QCS and PS were prepared by polymerization of styrene monomers in the emulsion of the QCS in the acetic acid aqueous solution first and afterward by cross-linking of the QCS to form the PS interpenetrated QCS network. The physical–chemical properties of the semi-IPN membranes were investigated. The hydroxyl ionic conductivity at a level of 102 S cm1 was reached at 80 °C. A stress at break of 20 MPa at room temperature was achieved, and it could maintain for more than 70 h by soaking in the 10 mol L1 KOH solution at ambient temperature. The swelling kinetics of the membrane in water was found to obey the Schott’s second-order swelling model. The semi-IPN structure based on the hydrophobic PS provided the membranes both mechanical and chemical stabilities. Acknowledgments We are grateful for the financial supports by the PhD Program Foundation of Ministry of Education of China (20090042110002) and the Fundamental Research Funds for the Central University of China (Nos. 90105001 and 90605004). References [1] E. Agel, J. Bouet, F. Fauvarque, J. Power Sources 101 (2001) 267. [2] J.R. Varcoe, R.C.T. Slade, Fuel Cells 5 (2005) 187.
225
[3] S. Gu, R. Cai, T. Luo, Z.W. Chen, M.W. Sun, Y. Liu, G.H. He, Y.S. Yan, Angew. Chem., Int. Ed. 48 (2009) 6499. [4] E.H. Yu, K. Scott, J. Power Sources 137 (2004) 248. [5] Y. Wang, L. Li, L. Hu, L. Zhuang, J. Lu, B. Xu, Electrochem. Commun. 5 (2003) 662. [6] R.C.T. Slade, J.R. Varcoe, Solid State Ionics 176 (2005) 585. [7] Z. Ogumi, K. Matsuoka, S. Chiba, M. Matsuoka, Y. Iriyama, T. Abe, M. Inaba, Electrochemistry 70 (2002) 980. [8] A.E.S. Sleightholme, J.R. Varcoe, A.R. Kucernak, Electrochem. Commun. 10 (2008) 151. [9] M. Guo, J. Fang, H. Xu, W. Li, X. Lu, C. Lan, K. Li, J. Membr. Sci. 362 (2010) 97. [10] V.V. Shevchenko, M.A. Gumennaya, Theor. Exp. Chem. 3 (2010) 139. [11] Y. Wan, K.A.M. Creber, B. Peppley, V.T. Bui, J. Membr. Sci. 280 (2006) 666. [12] Y. Wan, K.A.M. Creber, B. Peppley, V.T. Bui, J. Membr. Sci. 284 (2006) 331. [13] J. Pan, S. F Lu, Y. Li, A.B. Huang, L. Zhuang, J.T. Lu, Adv. Funct. Mater. 20 (2010) 312. [14] J. Fang, P.K. Shen, J. Membr. Sci. 285 (2006) 317. [15] L. Li, Y. Wang, J. Membr. Sci. 262 (2005) 1. [16] N. Vassal, E. Salmon, J.F. Fauvarque, Electrochim. Acta 45 (2000) 1527. [17] D. Stoica, L. Ogier, L. Akrour, F. Alloin, J.F. Fauvarque, Electrochim. Acta 53 (2007) 1596. [18] N.P. Berezina, N.A. Kononenko, O.A. Dyomina, N.P. Gnusin, Adv. Colloid Interface Sci. 139 (2008) 3. [19] Y. Xiong, Q.L. Liu, Q.G. Zhang, A.M. Zhu, J. Power Sources 183 (2008) 447. [20] V.R. Patel, M.M. Amiji, Pharm. Res. 13 (1996) 588. [21] A. Testouri, C. Honorez, A. Barillec, D. Langevin, W. Drenckhan, Macromolecules 43 (2010) 6166. [22] R. Pauliukaite, M.E. Ghica, O.F. Filho, C.M.A. Brett, Electrochim. Acta 55 (2010) 6239. [23] Y. Wan, B. Peppley, K.A.M. Creber, V.T. Buia, E. Halliop, J. Power Sources 185 (2008) 183. [24] O.A.C. Monteiro, C. Airoldi, Int. J. Biol. Macromol. 26 (1999) 119. [25] J.Z. Knaul, S.M. Hudson, K.A.M. Creber, J. Polym. Sci., Part B: Polym. Phys. 37 (1999) 1079. [26] T. Uragami, M. Takuno, T. Miyata, Macromol. Chem. Phys. 203 (2002) 1162. [27] Y. Wan, B. Peppley, K.A.M. Crebera, V.T. Buia, J. Power Sources 195 (2010) 3785. [28] Y.J. Choi, M.S. Kang, S.H. Moon, J. Appl. Polym. Sci. 88 (2003) 1488. [29] S.J. Lee, S.S. Kim, Y.M. Lee, Carbohydr. Polym. 41 (2000) 197. [30] P.P. Chu, C.S. Wu, P.C. Liu, T.H. Wang, J.P. Pan, Polymer 51 (2010) 1386. [31] Q.W. Tang, X.M. Sun, Q.H. Li, J.H. Wu, J.M. Lin, J. Colloid Interface Sci. 339 (2009) 45. [32] K.Y. Cho, H.Y. Jung, S.S. Shin, N.S. Choi, S.J. Sung, J.K. Park, Electrochim. Acta 50 (2004) 589. [33] G.K.P. Surya, M.C. Smart, Q.J. Wang, A. Atti, V. Pleynet, B. Yang, J. Fluorine Chem. 125 (2004) 1217. [34] W. Xu, X. Liao, Chin. J. Anal. Lab. 27 (2008) 218. [35] S.H. Lim, S.M. Hudson, Carbohydr. Res. 339 (2004) 313. [36] J. Wu, Z.G. Su, G.H. Ma, Int. J. Pharm. 315 (2006) 1. [37] R.H. He, Q.F. Li, A. Bach, J.O. Jensen, N.J. Bjerrum, J. Membr. Sci. 277 (2006) 38. [38] R.H. Michael, A.H.M.A. Todd, K.M. Sarah, J.C. Chris, Chem. Mater. 20 (2008) 2566. [39] R. He, Q. Li, G. Xiao, N.J. Bjerrum, J. Membr. Sci. 226 (2003) 169. [40] J.H. Wang, S.H. Li, S.B. Zhang, Macromolecules 43 (2010) 3890. [41] J.Y. Luo, C.M. Wu, Y.H. Wu, T.W. Xu, J. Membr. Sci. 347 (2010) 240. [42] Z.B. Wang, H.L. Zhao, J. Zhao, X. Wang, J. Appl. Polym. Sci. 5 (2010) 2523. [43] Q.K. Zhang, S.J. Zhang, S.X. Chen, P.Y. Li, T.Y. Qin, S.G. Yuan, J. Colloid Interface Sci. 322 (2008) 421. [44] E.H. Hany, Eur. Polym. J. 43 (2007) 4830. [45] Y.H. Yin, Y.J. Yang, H.B. Xu, J. Appl. Polym. Sci. 83 (2002) 2835. [46] V.I. Zabolotsky, V.V. Nikonenko, N.D. Pimenskaya, J. Membr. Sci. 119 (1996) 171. [47] D. Elattar, A. Elmidoui, N. Pimenskaia, C. Gavach, G. Pourcelly, J. Membr. Sci. 143 (1998) 249. [48] C. Iojoiu, F. Chabert, M. Maréchal, N.El. Kissi, J. Guindet, J.Y. Sanchez, J. Power Sources 153 (2006) 198.