A series of organic–inorganic alkaline hybridmembraneswere prepared based on crosslinked quaternized chito-san (QCS, the quaternization degree (DQ) of the QCS was 80.8 ± 3.5%) with different contents oftetraethoxysilanes (TEOS) for alkaline direct methanol fuel cells (ADMFC). These hybrid membranes were char-acterized by Fourier-transform infrared (FT-IR), thermo gravimetric analysis (TGA). The ion exchange capacity(IEC), water uptake, alkaline stability, methanol permeability and mechanical properties of the membraneswere also investigated to evaluate their applicability in ADMFC. The results showed that the anionic conductivi-ties of the hybridmembranes at a level of 10−2 S cm−1were obtained at 80 °C. Themethanol permeability of themembranes was in a range from 7.5 × 10−6 to 1.5 × 10−6 cm2 s−1. The obtained anion exchange membranes(AEM)were stable up to 200 °C in atmosphere according to the TGA analysis. The tensile strength and elongationat break of the membranes were in a range from 15.78 to 27.56 MPa and 10.32% to 3.21% at room temperature,respectively. The results of alkaline stability show that a tensile strength about 20.0 MPa was maintained afterimmersing the hybrid membrane in a 10 mol L−1 KOH solution at room temperature for more than 96 h.
Direct methanol fuel cells (DMFC) represent one of the most attrac-tive power sources because of their stability at relatively low tempera-tures, high energy density, and operation simplicity [1,2]. One of thevital components of the DMFC is the conductive membrane, whichserves as a physical separator between the anode and the cathodewith the function of transporting ions and blocking electrons. For a poly-mer electrolyte membrane to be a candidate for DMFC the followingcriteria should be met: (i) adequate chemical and electrochemical sta-bility, (ii) high conductivity to enable high current densities and low in-ternal resistance, (iii) low permeability to reactants and products, and(iv) good mechanical stability. Nowadays, the most popular proton ex-change membrane (PEM) for DMFC is Nafion, which is a perfluorinatedpolymeric membrane developed by DuPont in the late 1960s [3]. Theworking circumstance of the DMFC with Nafion membrane is acidic.However, the acidic DMFC has faced several serious problems: (1)slow methanol oxidation kinetics [4–6], (2) the poisoning of CO inter-mediate on the Pt surface [7], (3) the highmethanol cross-over throughthe polymer membrane [8–11], and (4) the high costs of the Nafionmembrane and Pt catalyst, which have obstructed the commercializa-tion of DMFC. Many efforts, therefore, have been directed towards de-veloping new types of polymer electrolyte membranes for DMFCs.Besides unceasingly exploring new types of proton exchange mem-branes, anion-exchange membrane (AEM) [12–25] has also aroused
ghts reserved.
new interest in recent years. This type of membrane works underbasic circumstance, where the electrochemical reactions are more facilethan in acidic medium, and non-noble metals can be used as catalyst,making the fuel cell more cost effective [26]. Even though there aremore and more research focused on the AEM. Unfortunately, untilnow no commercially available AEM as Nafion does in the field ofPEM. For anion exchange membrane fuel cells (AEMFC) applications,AEM needs necessary conductivity, mechanical strength and chemicalstability.
As one of the most abundant natural polymers, chitosan has beenchosen as the polymer matrix to prepare membrane electrolyte due toits excellent membrane properties, low cost and the feasibility to bemodified easily with functional groups in the structure. Thequaternization of chitosan can be performed with such as (2, 3-epoxypropyl) trimethylammonium chloride (EPTMAC) by conversionthe amino groups into 2-hydroxypropyltrimethyl ammonium chloride[27]. The anion exchange conductor of quaternized chitosan (QCS) isthus obtained by replacement of the chloride ions with hydroxideions. The high DQ of chitosan results in not only high ionic conductivity,but also significant swelling of the QCSmembranes and thereforeworsemembrane strength. To reinforce the strength of the chitosan basedmembranes, dialdehydes such as glyoxal [28–30] and glutaraldehyde(GA) [30,31] are normally used to perform the crosslinking. The stableimine bonds between amine groups of the chitosan polymer and the al-dehydic group of the glutaraldehyde is then formed [32,33]. However, itwas found that crosslinkage is not an effective method to improve theproperties of membrane when the quaternization degree of QCS isabove 35% due to too high swelling and poor mechanical properties of
97J. Wang, L. Wang / Solid State Ionics 255 (2014) 96–103
the resultantmembranes [31]. Therefore, attempts have also beenmadeto develop AEM by preparation of chitosan composite membranes withother polymers [34] or additives [35,36]. With the aim of enhancementof the mechanical strength of the polymer membranes, a technologyhas been developed to fabricate composite membranes with inorganicadditives of so-called organic–inorganic hybrid membrane [37–43].Organic–inorganic hybrid materials have been regarded as promisingmaterials for many applications due to their unique performance tocombine the remarkable functionality of organic materials with the sta-bility of inorganicmaterials [44,45]. In PEM fuel cellfield,manyorganic–inorganic hybrid membranes have also been prepared [46–51]. The re-sults showed that the tendency for addition of inorganic materials intothe organic polymers are: (1) to reduce the transport channel for meth-anol resulting in a decrease in the methanol permeability; (2) to en-hance the thermal stability and intensity of the membranes; (3) toincrease the ion conductivity [52]. However, the similar research inAEM field for DMFC was not “booming”, moreover the organic polymerwith high degree of quaternization (DQ) (DQ ≥ 50%) was taken as thematerials were seldom. Therefore, in this paper we propose a route toprepare a series of organic–inorganic hybrid membranes with organicpolymer (QCS, DQ = 80.8 ± 3.5%) and inorganic filler (silica) via sol–gel reaction using TEOS as a precursor. By adding a desired amount ofTEOS, a series of hybrid anion exchangemembraneswith different silicacontents have been synthesized. The characteristics of the membraneswere evaluated for potential application in alkaline DMFC.
2. Experimental
2.1. Materials
(2, 3-Epoxypropyl) trimethylammonium chloride (EPTMAC,purity ≥95%) was purchased from Shandong GuoFeng Fine ChemistryFactory. Chitosan, aqueous glutaraldehyde (GA, 50 wt.%) andtetraethoxysilane(TEOS, SiO2 ≥28%) were obtained from China Nation-al Medicines Corporation Ltd. The deacetylation degree of the Chitosanwas 95%, which was determined according to the reference [53]. Allthe reagents used were analytical grade. The QCS in chloride form wassynthesized by quaterinization of chitosan in isopropanol withEPTMAC at 85 °C for 10 h and itwas afterwards isolated and purified ac-cording to the reference described elsewhere [54]. The quaternizationdegree of theQCSwas80.8 (±3.5) %,whichwasdetermined by titrationwith a standard AgNO3 solution [55]. The same quaternization degree ofQCS was hereafter used for all the membrane preparation.
2.2. Methods
2.2.1. Membrane preparationIn a 100 mL round-bottom flask equipped with a magnetic stirrer
and a condenser, 2.0 g chloride form QCS (DQ = 80.8 (±3.5) %) wasdissolved in 40 mL 2% (v/v) acetic acid aqueous solution in nitrogen at-mosphere at room temperature. Then a required amount of TEOS and10 ml of ethanol was added into the solution and the mixture wasstirred for 1 h at 30 °C to obtain a clear homogeneous solution. After-ward amixture of ethanol (5 ml) and distilledwater (10 ml) containing0.1 wt.% HCl was slowly added to the solution under continuous stir-ring. A given amount of crosslinker glutaraldehyde (GA, 2%, v/v) aque-ous solution was then added dropwise to the mixture at a feeding rateof 0.5 mL min−1 to crosslink the QCS (in this research the amount ofGA was 0.2 wt.% in all membranes). The mixture was stirred at roomtemperature for 1 h and it was then sonicated for another 30 min. Theresultant mixture was poured onto a glass plate and the solvent wasthen evaporated in an oven at 40 °C until a constant weight wasreached. The alkaline doping of the membranes was performed byimmersing the membranes in 1.0 mol L−1 KOH solution at room tem-perature for 24 h. The membranes were then washed thoroughly withde-ionizedwater and dried at 40 °C until a constantweightwas reached
to obtain the hydroxide formmembranes. By varying the ratio of SiO2 toQCS, different organic–inorganic hybrid membranes with 3, 6, 9, 12 and15 wt.% SiO2were prepared. The hybridmembraneswere designated asM-X, where X is the SiO2 content (wt.%) in the membrane phase.
2.2.2. Instruments and techniquesThe 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 KBr pellets. Thermo gravimetricanalysis (TGA) was performed on a TGA 290C analyzer (Netzsch Com-pany, Germany) at a heating rate of 10 °C min−1 under air atmosphere.The mechanical strength of the dry membranes was determined withan instrument CMT6502 (SANS Company, China). Dumbbell-shapedmembrane samples of 25 mm × 4 mm were prepared and the mea-surements were carried out by setting a constant separating speed of5.00 mm min−1 under the ambient atmosphere. The tensile stress atbreak E was calculated by Eq. (1) [56].
E ¼ FA0
ð1Þ
Where F is the applied force at break, A0 is the initial cross-sectionarea of the sample which is equal to 4 × L mm2, and L is the thicknessof the membrane.
2.2.3. Water uptake and swelling of the membranesThemembrane samples were soaked in de-ionizedwater for 24 h at
room temperature tomonitor the variations onweight of thewetmem-branes. The weight of the wet membrane (Pwet) was measured rapidlyafter wiping the excessive surface water with a tissue paper, and thatof the dry membrane (Pdry) was obtained by drying the samples at50 °C in a vacuum oven until a constant weight was reached. Thewater uptake and the swelling of membranes were determined byEq. (2):
Water uptake or swelling %ð Þ ¼ Pwet−Pdry
Pdry� 100% ð2Þ
2.2.4. Ion exchange capacityThe anion exchangemembrane in hydroxide formwaswashed thor-
oughly with distilled water and then dried in a vacuum oven at 60 °C toreach a constant weight (wOH in gram). It was afterwards immersed in a0.1 mol L−1 HCl standard solution at ambient temperature for 48 hunder stirring to neutralize the hydroxide ions contained in the mem-brane. The mole number (equivalent) of the neutralized hydroxideions (n) was determined by titration of the remanent acid with a0.1 mol L−1 KOH standard solution. The ion exchange capacity (IEC)of the anion exchange membrane, termed as mili-equivalents (meq)of hydroxide ions per gram of the dry hydroxide form membrane, wasobtained by calculation with Eq. (3) [57].
IEC meq g−1� �
¼ 1000nwOH
ð3Þ
2.2.5. Anionic conductivityThe anionic conductivity measurement cell has been described else-
where [58]. The conductivity of the membrane was measured by usingalternating current (AC)with a frequency of 2 KHz supplied via a pair ofplatinum electrodes. The resistance between the two electrodes wasmeasured with and without the membrane, respectively, to obtain theresistance of the membrane (RM) by comparing the difference. Inorder to keep that the electrodes were held at a fixed distance apartthe mini-gasket was used under the without membrane to cover thedistance, which presents the thickness of the membrane, between the
Fig. 1. FT-IR spectra of the chitosan, QCS and the hybrid membranes.
98 J. Wang, L. Wang / Solid State Ionics 255 (2014) 96–103
two electrodes. As discussed in our previouswork [58], a lower concen-tration of the 0.1 mol L−1 KOHwas employed as the electrolyte becausea higher concentration of the KOHmight affect the anionic conductivityespecially at higher temperatures, although the contribution of the KOHelectrolyte has been subtracted. The anionic conductivity (σ) of themembrane was calculated by Eq. (4).
σ S cm‐1� �
¼ lRM � S
ð4Þ
Where l is the thickness of the wet membrane (cm); and S is themembrane surface (cm2) for ion transport, respectively. The tempera-ture was controlled by putting the conductivity measurement cell inan oven.
2.2.6. Methanol permeability through the membraneThe methanol permeability of the membranes was determined at
room temperature using a home-made diffusion cell. The cell consistsof two conical flasks with a volume of about 100 cm3 and each flaskhas a columniform side-arm (area 0.4126 cm2) to clamp themembraneand provide a channel for methanol permeation. One flask was filledwith 1 mol L−1 methanol aqueous solution (flask A) and the otherwith de-ionized water (flask B). The methanol concentration in theflask (B) by permeation at any time t (in second), i.e., CB in mmol L−1,was monitored by using a gas chromatograph (GC-6820, Agilent,USA). The methanol permeability P (cm2 s−1) through the membranewas calculated by Eq. (5) [59].
P ¼ LA
CBVB
CA−CBð Þt ð5Þ
Where A (cm2) and L (cm) are themembrane area and thickness, re-spectively, CA (mmol L−1) is the initial concentration of themethanol inthe flask (A), VB (mL) is the water volume in the flask (B).
2.2.7. Membrane stabilityThe stability of the membrane in alkaline medium was investigated
by monitoring both the conductivity and the tensile strength of themembranes as a function of soaking time in KOH solutions [60]. Theconductivity of the membranes immersed in 1.0 mol L−1 KOH atroom temperature for different times was determined. Before makingthe measurements, the membrane was washed with de-ionized waterthoroughly to remove the absorbed potassium hydroxide. The stabilityin tensile strength of the membrane in alkaline mediumwas evaluatedby soaking the membrane in a 10 mol L−1 KOH solution at ambienttemperature for different times and testing the tensile strength of themembrane correlatively.
All experimental data in this publication are based on at least threereplications of the respective experiments. The standard deviationswere as follows: for tensile strength the standard deviation wasbelow ±0.42 MPa; for elongation at break: ±1.00%; for IEC values:±0.05 meq g−1; for water uptake: ±4.00% and for anionic conduc-tivity: ±0.0006 cm S−1.
3. Results and discussion
3.1. Preparation of the composite membranes
The FT-IR spectra of the QCS and its compositemembranes aswell asthe pristine chitosan (CS) are shown in Fig. 1.
A broad peak at 3458 cm−1 is observed for all the membrane sam-ples, which may be caused by both N–H and O–H as a result of overlap.A band at 1586 cm−1 can be attributed to amino groups of CS, and ashoulder at around 1660 cm−1 is assigned to the C=O stretch resultedfrom the residual acetyl in commercial CS, which is usually recorded forchitosan with a high degree of deacetylation [61]. There are no distinct
absorptions of the N–H bending to the primary amine at 1586 cm−1
for the QCS [54] and its hybrid membranes. It is also noted that thepeak originally corresponding to the primary amine (1586 cm−1) ofCS disappears, and the peak at around 1660 cm−1 for CS was shiftedto 1640 cm−1 for QCS and its hybridmembraneswere observed, clearlyrevealing that the primary amine of chitosan backbone was alreadychanged into a secondary amine structure due to the reactions associat-ed with EPTMAC.
In addition, the increased sharp absorptions at around 2927, 2850,and 1480 cm−1 resulted from C–H vibrations and bendings of thetrimethylammonium group for QCS membranes also indicate the exis-tence of the quaternary ammonium salt in the QCS [61]. The absorptionpeak around 3450 cm−1 in the CS and QCS membrane belongs to –OHgroups. Also in the hybrid membranes, the peak at 3450 cm−1 wasdue to the existence of –OH, which come from the backbone of CS orunreacted -Si–OH. In the QCS and hybrid membranes, the peak at1177 and 1083 cm−1 is attributed to Si–O–C rocking bands [62] andSi–O–Si bonds stretching [63,64], respectively. The absorption peakaround 750 cm−1 in the hybridmembranes belongs to stretching vibra-tion for O–Si–O groups [65] that was due to the sol–gel process. Theseresults confirm the formation of SiO2 inorganic network inside the hy-brid membrane structure due to the sol–gel reaction, in which thesilanol groups in the tetrasilanol silane, produced by the hydrization ofTEOS, reacted with the hydroxyl groups in the QCS molecule.
3.2. Thermal stability
Durability of polymer electrolyte membranes is a very importantfactor for limiting the commercial application of polymer electrolytemembrane fuel cells. Thermal stability of the membranes apparentlyplays a key role since these membranes have to function steadily anddurably in different environments at various temperatures for a longterm. The mass-loss behaviors of the hybrid membranes were studiedusing TGA at a heating rate of 10 °C min−1 in flowing air atmosphereand the results were shown in Fig. 2. As a comparison the thermogramsof the pure chitosan membrane and pure QCS membrane were alsoshown in Fig. 2. The first mass-loss stepping from room temperatureto 200 °C for all the samples resulted from the evaporation of theabsorbed water or ethanol generated from the sol–gel reaction ofTEOS in the polymers. On the other hand, the quaternary ammoniumgroups in the QCS decomposed when the temperature was above125 °C, i.e., 127.6 °C [61], which will also result in mass loss of the sam-ples in the same temperature scope. In this scope suchmass loss is about0.5% for CS, whereas QCS and its hybrid membranes exhibited slightlyhigher loss ranging from 2 to 7%. Moreover, this was decreased corre-spondingly with increase of the amount of inorganic filler in the mem-brane. This fact indicates that the QCS and its hybrid membranes have
Fig. 2. TGA curves of the hybrid membranes and crosslinked QCS membrane at a heatingrate 10 °C min−1 in air atmosphere.
Fig. 3. Water uptake and swelling of the hybrid membranes.
99J. Wang, L. Wang / Solid State Ionics 255 (2014) 96–103
higher hydrophilicity than CS. That was due to the quaternary ammoni-um groups, which was grafted onto the QCS.
The second stage of decomposition occurred between 200 and300 °C, and corresponds to weight loss of 30–50%. This is due to degra-dation of side chains, main chains and deacetylation of QCS [66,67]. Al-though it was expected that the inorganic filler will affect the thermalstability, whereas, the second step showed almost similar degradationtemperature for QCS and its hybrid membranes at the beginning ofthis steps. This result can be attributed to the following reasons: thereare a lot of less thermally stable groups such as quaternary ammoniumgroups (DQ = 80.8 (±3.5) % for QCS) and hydroxyl groups into thestructure of the hybrid membranes in this research. And hence organ-ic–inorganic hybrid did not have a pronounced effect on the decompo-sition at the beginning of the second step. The effect of inorganic fillerwill be enhanced when the amount of the quaternary ammoniumgroups was decreased, which results in the increasing of decompositiontemperature. If we consider the decomposition temperature at 10%weight loss, the hybrid membrane M-15, M-9 and M-3 exhibits 55 °C,32 °C and 17 °C higher than that of QCS membrane, respectively. Thisis because of the synergetic effect of SiO2 filler and QCS on the thermalproperties. The improvement of thermal stability of the hybrid mem-branes may be explained as follows [68,69]: good heat barrier proper-ties of inorganic materials for polymer matrix during formation ofcrosslinked silica network structure, which could hinder the flux of deg-radation product and heat flowing into underlying materials, or thethermal stability was enhanced by decreasing content of quaternaryammoniumgroups result in the content of SiO2was increased, or stronginteraction between QCS, which was introduced by GA could also re-strict the polymer motion during heating. As for the thermal stabilityof the membranes, which was mentioned above was considerablylower than the CS membrane. This is because the quaternary graftingside chains onto CS backbone and the introduce of inorganic fillerwould definitely suppress the recrystallized ability of main chains [61]and result in remarkably decreased crystallinity in QCS membrane andits hybrid membranes. The third stage is more than 300 °C, indicatingthe cleavage of the C–C backbone of the polymer matrix.
Although the thermal stability of all hybrid membranes shown inFig. 2 is still much lower than that of the pure chitosan membrane,
Table 1The mechanical strength of QCS/SiO2 hybrid membranes.
these membranes are considered to be thermally stable enough forpractical applications because they will be possibly used only in low-temperature alkaline fuel cells where the operation temperature isusually lower than 100 °C.
3.3. Tensile strength
The hybrid membranes prepared in this research are all of good ten-sile strength to the eyes. For illustration,M-3,M-6, M-9,M-12 andM-15were selected for themeasurement of tensile properties. Specific valuesof the tensile properties of hybrid membranes, including tensilestrength (TS) and elongation at break (Eb) are collected in Table 1. Inorder to evaluate the value of tensile strength of the hybridmembranes,the Nafion 117 was also measured and the results were listed into theTable 1. Obviously, membranes have relatively high TS and Eb valuesas compared to the Nafion 117, and the obtained value in this researchis similar to that of the reference [40]. The elongation at break for allthe hybridmembranes is lower in relation to theNafion 117membrane.As mentioned in Section 2.2.1, all of the hybrid membranes wereprepared by the QCS with the same degree of quaternization(DQ = 80.8 ± 3.5%) and the same content of GA. Therefore the valueof TS and Eb was only depending upon the content of inorganic SiO2
filler.Through observation of Table 1, it was concluded that the tensile
strength of all the hybrid membranes increases gradually with increasein the inorganic filler content. This is obviously due to the presence ofinorganic filler that results in the formation of the silica network [70].On the contrary, the trends of Eb are opposite. The value of Eb was de-creased with an increase in the inorganic filler content. The reason forthis phenomenon was due to the increasing silica content in the mem-branes i.e. the higher cross-linking at higher loading, therefore it willapparently increase the stiffness of the membrane and reduce themembrane's flexibility, so that the Eb of the hybrid membranes, tendsto be opposite to the TS. It can also be seen from Table 1 that at thegiven different feeding composition, the values, especially TS, differ sig-nificantly among the membranes. M-15 has a TS value as high as27.56 MPa while the TS values of the other hybrid membranes aremuch lower. The TS values of the preparedmembranes aremuch higherthan those of some reported membranes [37,71].
Fig. 5. Arrhenius plots for the QCS/SiO2 hybrid membranes.
100 J. Wang, L. Wang / Solid State Ionics 255 (2014) 96–103
3.4. Water uptake and swelling of the membranes
The water uptake of the membranes was shown in Fig. 3. It can beseen that for all the hybridmembranes, thewater uptake values are rel-atively high. This is due to the high hydrophilic quaternary ammoniumgroups of the QCS component [72,73], with high DQ. And the hydrophil-ic Si\OH groups, which come from the hydrolysis of TEOS which wasdoped into the hybrid membranes, will also avail for increasing ofwater uptake.
Through observation of Fig. 3, it can be seen that thewater uptake ofmembranes generally decrease with SiO2 content increases, as reflectedfrom themuch lowerwater uptake value of the hybridmembranes thanQCSmembrane. Such trend is may be due to the presence of hydropho-bic silica, which forms a very rigid [74] three-dimensional networks ofSiO2; it is reported that the crystallinity of the membrane was remark-ably inhibited by silica network [75]. The chemical structure of the hy-brid membrane was more compact with the increasing of the SiO2
content. Thatwas due to the SiO2 feed ratio increases, resulting in higherdegree of cross linkage of inorganic silica network in the hybridmembranes.
3.5. Ion exchange capacity (IEC)
Table 2 illustrates the IEC ofQCS/SiO2 hybridmembraneswith differ-ent usage of SiO2. The IEC values of the membranes were gradually de-creased from 1.82 meq g−1 to 0.93 meq g−1 that corresponded to themembrane M-3 to M-15. Due to the QCS, which was used to synthesizethe hybrid membrane which is positively charged, IEC values of hybridmembranes will increase with the increase of its content, as can bereflected from the gradually lower IEC from membrane M-3 to M-15.GA and TEOS, on the other hand, are not charged, so an increase intheir content will decrease the membrane IEC, as illustrated by thelower IEC values of membrane, with higher content of the SiO2.
3.6. Anionic conductivity
Anionic conductivity of the alkaline anion exchange membranes is acrucial criterion for evaluating its performance in a fuel cell. Fig. 4 showsthe conductivity of the hybrid membranes. The highest conductivityfor themembraneM-3 is 1.89 × 10−2 S cm−1 at 80 °C, which is higher
Fig. 4. Anionic conductivity of QCS/SiO2 hybrid membranes.
than the other QCS/SiO2 hybrid membranes at the same temperature.The anionic conductivity of the QCS/SiO2 hybrid membrane decreaseswith increase of SiO2 content.
It hypothesizes that the transport of OH− anion follows the Grottusmechanism. The water in the QCS/SiO2 hybrid membranes can be clas-sified as free water and bound water. Only free water can affect theOH− transport. As the doping of TEOS content increases, the un-reacted –Si–OH increases and more water will be restricted that cannottake part in OH− transport [76]. On the other hand, the linkage of SiO2
and QCS makes the membrane structure much more compact and alsoweakens OH− transport in the QCS/SiO2 hybrid membranes. As for IECvalues for the hybrid membranes were also decreased with increasesin the inorganic filler content. These three actions make the membraneM-15 have a lower conductivity than the other hybridmembranes withmore silica content. However, other properties of the hybrid mem-branes (for example the tensile strength, swelling property and so on)improved because of the aforementioned reasons. Therefore, the anionicconductivity of such hybrid membranes improved not only depends onincreasing the hydrophilicity or quaternization degree (80.8 (±3.5) %,in this research). Maybe building the ion channel for OH− transfer is agood choice.
In addition, it was worthwhile to mention that the conductivity ofQCS/SiO2 hybrid membranes was increased with the increasing of themeasured temperature. That was due to the fact that higher tempera-ture enhanced the mobility of OH− within the membrane.
Fig. 5 shows the relation between logσ and 1000/T, in which byusing the Arrhenius equation, allows the determination of the ion trans-port activation energy (Ea) of the hybrid membranes. The membraneM-3 has the lowest Ea (21.41 kJ mol−1) of the QCS/SiO2 hybrid mem-branes, which is consistent with the OH− conductivity results. With
Fig. 6. Methanol permeability of QCS/SiO2 hybrid membranes and Nafion 115.
101J. Wang, L. Wang / Solid State Ionics 255 (2014) 96–103
the increasing of SiO2 content, the activation energy increases due to themore compact structure formed by SiO2 and QCS polymer matrix. Sim-ilarly experimental results and phenomenon were also observed byXiong et al. [38]. Compared with the other membranes [34,77], theQCS/SiO2 hybrid membranes synthesized in this search with higher ac-tivation energies, are indicating that the structure of the hybrid mem-brane is more compact than those of the other composite membranesreported in the reference.
3.7. Methanol permeability
Themethanolmolecule, which exhibits a high affinitywithwater forinterdiffusion, transport in the membrane is similarly as the ions do.Therefore, high IEC value and water uptake of the composite mem-branes are normally correlative to high swelling of the polymer andthus high methanol permeability.
Fig. 6 gives methanol permeability of the QCS/SiO2 hybrid mem-branes. As a comparison, themethanol permeability of Nafion 115mea-sured using the samemethod was also given in Fig. 6. It is seen that themethanol permeability of the hybrid membranes decreased with theincrease of the amount of SiO2. Among them, the lowest methanolpermeability was 1.55 × 10−6 cm S−1, which was corresponding tothe membrane M-15, the value was lower than Nafion 115, i.e.,2.42 × 10−6 cm2 s−1. It was indicated that methanol permeability ofthe QCS/SiO2 hybrid membranes was reduced by doping the inorganicfiller. Though, methanol permeability of QCS/SiO2 hybrid membraneswas still too high for application in DMFC. However, one point thatmust be paid more attention was that the methanol permeability ofthe QCS/SiO2 hybrid membranes was gradually decreased with the
Fig. 7. The morphology of the QCS/SiO2 hybrid membranes: (a) M-3, surfa
increase of inorganic SiO2filler. That probably attributes to the introduc-tion of silica network [78] and crosslinker GA makes the membranemore compact.
3.8. Morphology of the hybrid membranes
It has beenwell documented that the compatibility between organicpolymers and inorganic (such as silica and titania) components has asignificant effect on the thermal and tensile properties of hybrid mate-rials or membranes [79–81]. So morphologies of hybrid membraneswere observed through SEM. The surface and cross-sectionmorphologyof the QCS/SiO2 hybrid membranes M-3 and M-15 was shown in Fig. 7.The surface and cross section of themembraneM-3 are smooth and ho-mogenous, indicating that the compatibility between the organic andinorganic parts is excellent. However, with increase of silica content,particles can be observed on the surfaces and cross section of themem-braneM-15 and the number of particles tends to bemore. The particlesare formed by self-polymerization of silica and distribute on the surfaceof the hybrid membranes uniformly. But it was worthwhile to note thatthere was no discernible crack and pore formation into the structure ofthe QCS/SiO2 hybrid membranes.
Although the M-15 appeared to have rough surfaces, in general itwas uniform and homogeneous, and indeed all of the hybrid mem-branes were translucent. On the other hand, the incorporation of thesesilica particles significantly changed the properties of the QCS/SiO2 hy-bridmembranes as shown bymethanol permeability. Methanol perme-ability of the hybrid membranes with the increasing of SiO2 wasdecreased due to the presence of these particles within the membrane.
Fig. 8. Conductivities of the QCS/SiO2 hybrid membranes at 80 °C versus the immersedtime in 1 M KOH solution.
102 J. Wang, L. Wang / Solid State Ionics 255 (2014) 96–103
3.9. Stability of the hybrid membranes in alkaline solutions
The chemical stability of an anion exchange membrane based onquaternary ammonia is always a concern due to its limited durabilityin basic medium. Here, alkaline stability of the hybrid membranes wasinvestigated by immersing membrane samples into 1 mol L−1 KOHand 10 mol L−1 KOH solutions at room temperature. The changes of an-ionic conductivity and tensile strength values of the tested membraneswere noted.
Fig. 8 shows the hydroxide conductivity at 80 °C of the hybridmem-branes after treatment with 1 mol L−1 KOH at room temperature.
Within 120 h, the decreases in the conductivity are 45.50%, 28.82%,19.21%, 16.28% and 12.75% for the membrane M-3, M-6, M-9, M-12andM-15, respectively. The membrane degradation was due to the dis-placement of the ammonium group by the OH− anions via directnucleophillic displacement and/or Hofmann elimination reactionwhenα, β hydrogen atoms andα carbon atoms are present. In the pres-ent study, there are α, β hydrogen atoms and α carbon atoms aroundthe quaternary ammonium groups, and the direct nucleophillic dis-placement andHofmannelimination reaction on thequaternary ammo-nium salt functionalized AEMs could not be eliminated. Thus, thehydroxide conductivity of the hybrid membranes was decreased. How-ever, it wasworthwhile tomention that a slight degradation on the con-ductivity of the hybrid membranes with high SiO2 content in themembrane structure pretreated at room temperature is observed fromthe figure but a significant drop in the conductivity for the membranewith high content QCS in membrane. That was due to the presence ofthe inorganic filler SiO2 and the formation of inorganic network retardthe membrane degradation, which was probably due to the sterichindrance.
Fig. 9.Degradation of the QCS/SiO2 hybridmembranes on tensile strength in a 10 mol L−1
KOH solution at room temperature.
The tensile strength of the hybrid membranes immersed in a10 mol L−1 KOH at room temperature for different times was testedand the results are shown in Fig. 9. The tensile stress at break of themembrane M-15 at room temperature could maintain about 20 MPaby soaking in a 10 mol L−1 KOH at ambient temperature for morethan 96 h. However, the tensile stress at break of the M-3 membranewas only 3.7 MPa after immersing in the basic solution at room temper-ature for 96 h.
According to the experimental results we conclude that the forma-tion of SiO2 network structure allows the quaternary ammonia basedmembranes to have a high tolerance to the bases. The inorganic filler(SiO2) in the hybrid membrane should be of the benefit.
4. Conclusions
The QCS/SiO2 hybrid anion exchange membranes based on QCS andTEOS were prepared by sol–gel method. The physical chemical proper-ties of the hybrid membranes were investigated. The hydroxyl ionicconductivity at a level of 10−2 S cm−1 was reached at 80 °C. A stressat break of 20 MPa at room temperature was achieved. A tensilestrength of about 20.0 MPa wasmaintained after immersing the hybridmembrane in a 10 mol L−1 KOH solution at room temperature formorethan 96 h. The inorganic additive SiO2 together with the network struc-ture increased the thermal stability and chemical stability, decreasedthe water uptake, methanol permeability and swelling of the mem-branes therefore more competitive properties than pure QCS mem-brane were achieved.
Acknowledgments
We are grateful for the financial supports by the Natural ScienceFoundation of China (51172039), the Scientific Research Fund of Liao-ning Provincial Education Department (L2013153) and the Fundamen-tal Research Funds for the Doctoral of Liaoning Shi Hua University(2013XJJ-006).
References
[1] N.A. Hampson, M.J. Wilars, B.D. McNicol, J. Power Sources 4 (3) (1979) 191–201.[2] X. Ren, M.S. Wilson, S. Gottefeld, J. Electrochem. Soc. 143 (12) (1996) L12–L15.[3] K.A. Mauritz, R.B. Moore, Chem. Rev. 104 (2004) 4535–4585.[4] W.H. Lizcano, V.A. Paganin, E.R. Gonzalez, Electrochim. Acta 47 (22) (2002)
3715–3722.[5] N. Nakagawa, Y. Xiu, J. Power Sources 118 (1–2) (2003) 248–255.[6] G.G. Park, T.H. Yang, Y.G. Yoon, W.Y. Lee, C.S. Kim, Int. J. Hydrogen Energy 28 (6)
(2003) 645–650.[7] T.C. Deivaraj, J.Y. Lee, J. Power Sources 142 (1–2) (2005) 43–49.[8] V. Baglio, A.S. Arico, A.D. Blasi, V. Antonucci, P.L. Antonucci, S. Licoccia, E. Traversa,
F.S. Fiory, Electrochim. Acta 50 (5) (2005) 1241–1246.[9] S. Panero, P. Fiorenza, M.A. Navarra, J. Romanowska, B.J. Scrosati, Electrochem. Soc.
152 (12) (2005) A2400–A2405.[10] J.H. Choi, Y.M. Kim, J.S. Lee, K.Y. Cho, H.Y. Jung, J.K. Park, I.S. Park, Y.E. Sung, Solid
State Ionics 176 (39) (2005) 3031–3034.[11] V.S. Silva, S. Weisshaar, R. Reissner, B. Ruffmann, S. Vetter, A. Mendes, L.M. Madeirs,
S. Nunes, J. Power Sources 145 (2) (2005) 485–494.[12] E.H. Yu, K. Scott, J. Power Sources 137 (2) (2004) 248–256.[13] Y. Wang, L. Li, L. Hu, L. Zhuang, J.T. Lu, B. Xu, Electrochem. Commun. 5 (8) (2003)
662–666.[14] M.L. Guo, J. Fang, H.K. Xu, W. Li, X.H. Lu, C.H. Lan, K.Y. Li, J. Membr. Sci. 362 (1-2)
(2010) 97–104.[15] Y. Wan, K.A.M. Creber, B. Peppley, V.T. Bui, J. Membr. Sci. 280 (1–2) (2006) 666–674.[16] Y. Wan, K.A.M. Creber, B. Peppley, V.T. Bui, J. Membr. Sci. 284 (1–2) (2006) 331–338.[17] 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 (35) (2009) 6499–6502.[18] J. Pan, S.F. Lu, Y. Li, A.B. Huang, L. Zhuang, J.T. Lu, Adv. Funct. Mater. 20 (2) (2010)
312–319.[19] J. Fang, P.K. Shen, J. Membr. Sci. 285 (1–2) (2006) 317–322.[20] L. Li, Y. Wang, J. Membr. Sci. 262 (1–2) (2005) 1–4.[21] R.C.T. Slade, J.R. Varcoe, Solid State Ionics 176 (5–6) (2005) 585–597.[22] E. Agel, J. Bouet, F. Fauvarque, J. Power Sources 101 (2) (2001) 267–274.[23] N. Vassal, E. Salmon, J.F. Fauvarque, Electrochim. Acta 45 (8) (2000) 1527–1532.[24] D. Stoica, L. Ogier, L. Akrour, F. Alloin, J.F. Fauvarque, Electrochim. Acta 53 (4) (2007)
103J. Wang, L. Wang / Solid State Ionics 255 (2014) 96–103
[26] C. Qu, H.M. Zhang, F.X. Zhang, B. Liu, J. Mater. Chem. 22 (2012) 8203–8207.[27] Y. Xiong, Q.L. Liu, Q.G. Zhang, A.M. Zhu, J. Power Sources 183 (2) (2008) 447–453.[28] V.R. Patel, M.M. Amiji, Pharm. Res. 13 (4) (1996) 588–593.[29] A. Testouri, C. Honorez, A. Barillec, D. Langevin, W. Drenckhan, Macromolecules 43
6239–6247.[31] Y. Wan, B. Peppley, K.A.M. Creber, V.T. Buia, E. Halliop, J. Power Sources 185 (1)
(2008) 183–187.[32] O.A.C. Monteiro, C. Airoldi, Int. J. Biol. Macromol. 26 (2–3) (1999) 119–128.[33] J.Z. Knaul, S.M. Hudson, K.A.M. Creber, J. Polym. Sci., Part B: Polym. Phys. 37 (11)
(1999) 1079–1094.[34] T.Y. Guo, Q.H. Zeng, C.H. Zhao, Q.L. Liu, A.M. Zhu, I. Broadwell, J. Membr. Sci. 371
(1–2) (2011) 268–275.[35] L.H. Li, J.C. Deng, H.R. Deng, Z.L. Liu, L. Xin, Carbohydr. Res. 345 (8) (2010) 994–998.[36] S.K. Choudhari, M.Y. Kariduraganavar, J. Colloid Interface Sci. 338 (1) (2009) 111–120.[37] Y.H. Wu, C.M. Wu, F. Yu, T.W. Xu, Y.X. Fu, J. Membr. Sci. 307 (1) (2008) 28–36.[38] Y. Xiong, Q.L. Liu, A.M. Zhu, S.M. Huang, Q.H. Zeng, J. Power Sources 186 (2) (2009)
328–333.[39] C.M. Wu, Y.H. Wu, T.W. Xu, Y.X. Fu, J. Appl. Polym. Sci. 107 (3) (2008) 1865–1871.[40] A.K. Sahu, S.D. Bhat, S. Pitchumani, P. Sridhar, V. Vimalan, C. George, N.
Chandrakumar, A.K. Shukla, J. Membr. Sci. 345 (1–2) (2009) 305–314.[41] T. Li, Y. Yang, J. Power Sources 187 (1–2) (2009) 332–340.[42] S.S. Lee, S.S. Kim, Y.M. Lee, Carbohydr. Polym. 41 (2) (2000) 197–205.[43] P.P. Chu, C.S. Wu, P.C. Liu, T.H. Wang, J.P. Pan, Polymer 51 (6) (2010) 1386–1394.[44] C. Sanchez, B. Julián, P. Belleveille, M. Popall, J. Mater. Chem. 15 (2005) 3559–3592.[45] G. Kickelbick, Prog. Polym. Sci. 28 (1) (2003) 83–114.[46] D.S. Kim, H.B. Park, J.W. Rhim, Y.M. Lee, J. Membr. Sci. 240 (1) (2004) 37–48.[47] S.P. Nunes, B. Ruffmann, E. Rikowski, S. Vetter, K. Richau, J. Membr. Sci. 203 (1–2)
(2002) 215–225.[48] J.D. Kim, I. Honma, Electrochim. Acta 49 (19) (2004) 3179–3183.[49] P. Staiti, Mater. Lett. 47 (4–5) (2001) 241–246.[50] B. Ruffmann,H. Silva, B. Schulte, S.P. Nunes, Solid State Ionics 162–163 (2003) 269–275.[51] D.H. Jung, S.Y. Cho, D.H. Peck, D.R. Shin, J.S. Kim, J. Power Sources 118 (1–2) (2003)
205–211.[52] Z.Q. Mao, Fuel Cell, Chemical Industry Press, Beijing, 2005.[53] W.F. Xu, X.L. Liao, Chin. J. Anal. Lab. 27 (2008) 218–221.[54] S.H. Lim, S.M. Hudson, Carbohydr. Res. 339 (2) (2004) 313–319.[55] J. Wu, Z.G. Su, G.H. Ma, Int. J. Pharm. 315 (1–2) (2006) 1–11.
[56] R.H. He, Q.F. Li, A. Bach, J.O. Jensen, N.J. Bjerrum, J. Membr. Sci. 277 (1–2) (2006)38–45.
[58] J.L. Wang, R.H. He, Q.T. Che, J. Colloid Interface Sci. 361 (1) (2011) 219–225.[59] D.S. Kim, K.H. Shin, H.B. Park, Y.S. Chung, S.Y. Nam, Y.M. Lee, J. Membr. Sci. 278 (1–2)
(2006) 428–436.[60] J.H. Wang, S.H. Li, S.B. Zhang, Macromolecules 43 (8) (2010) 3890–3896.[61] Y. Wan, B. Peppley, K.A.M. Crebera, V.T. Buia, J. Power Sources 195 (12) (2010)
3785–3793.[62] C.M. Wu, T.W. Xu, M. Gong, W.H. Yang, Eur. Polym. J. 43 (4) (2007)
1573–1579.[63] N. Viart, J.L. Rehspringer, J. Non-Cryst. Solids 195 (3) (1996) 223–231.[64] F. Rubio, J. Rubio, J.L. Oteo, Spectrosc. Lett. 31 (1) (1998) 199–219.[65] A. Tiwari, S.Q. Gong, Electroanalysis 20 (19) (2008) 2119–2126.[66] F.A.A. Tirkistani, Polym. Degrad. Stab. 60 (1) (1998) 67–70.[67] X. Qu, A. Wirsen, A.C. Albertsson, Polymer 41 (13) (2000) 4841–4847.[68] S.F. Wang, L. Shen, Y.J. Tong, L. Chen, I.Y. Phang, P.Q. Lim, T.X. Liu, Polym. Degrad.
Stab. 90 (1) (2005) 123–131.[69] H.Y. Ma, L.F. Tong, Z.B. Xu, Z.P. Fang, Nanotechnology 18 (37) (2007)
375602–375609.[70] R.K. Nagarale, V.K. Shahi, R. Rangarajan, J. Membr. Sci. 248 (1–2) (2005) 37–44.[71] J. Hu, C.X. Zhang, J. Cong, H. Toyoda, M. Nagatsu, Y.D. Meng, J. Power Sources 196
(10) (2011) 4483–4490.[72] C.M. Wu, T.W. Xu, M. Gong, W.H. Yang, J. Membr. Sci. 247 (1–2) (2005) 111–118.[73] T.W. Xu, Z.M. Liu, W.H. Yang, J. Membr. Sci. 249 (1–2) (2005) 183–191.[74] T. Uragami, K. Okazaki, H. Matsugi, T. Miyata, Macromolecules 35 (24) (2002)
9156–9163.[75] K. Nakane, T. Yamashita, K. Iwakura, F. Suzuki, J. Appl. Polym. Sci. 74 (1) (1999)
133–138.[76] C. Li, G. Sun, S. Ren, J. Liu, Q. Wang, Z. Wu, H. Sun, W. Jin, J. Membr. Sci. 272 (1–2)
(2006) 50–57.[77] C.X. Zhang, J. Hu, J. Cong, Y.P. Zhao, W. Shen, H. Toyoda, M. Nagatsu, Y.D. Meng,
J. Power Sources 196 (13) (2011) 5386–5393.[78] D. Wu, R.Q. Fu, T.W. Xu, L. Wu, W.H. Yang, J. Membr. Sci. 310 (1–2) (2008)
522–530.[79] Y.J. Kim, W.C. Choi, S.I. Woo, W.H. Hong, J. Membr. Sci. 238 (1–2) (2004) 213–222.[80] C.W. Lin, P.H. Chang, J. Membr. Sci. 254 (1–2) (2005) 197–205.[81] Y.G. Hsu, F.J. Lin, J. Appl. Polym. Sci. 75 (2) (2000) 275–283.
