PVS–Nafion composite membrane for reduced methanol crossover and enhanced DMFC performance

PVS–Nafion composite membrane for reduced methanol crossover and enhanced DMFC performance

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 8 7 7 e1 8 8 5 Available online at www.sciencedirect.com S...

2MB Sizes 0 Downloads 8 Views

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 8 7 7 e1 8 8 5

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Layer-by-layer self-assembly of CHI/PVSeNafion composite membrane for reduced methanol crossover and enhanced DMFC performance Yanhong Xue a,*, Siewhwa Chan a,b,** a

Energy Research Institute, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore


article info


Article history:

A novel CHI/PVSeNafion composite membrane was fabricated to address the operational

Received 1 September 2014

issue of severe methanol crossover in DMFC by layer-by-layer (LbL) self-assembly of

Received in revised form

cationic polyelectrolyte Chitosan (CHI) and anionic polyelectrolyte polyvinyl sulfuric acid

5 November 2014

(PVS) on Nafion membrane. The bilayers were successfully deposited onto Nafion mem-

Accepted 22 November 2014

brane surface, which was confirmed by FTIR spectra, FESEM and AFM images. The thermal

Available online 18 December 2014

property and the water uptake of the LbL composite membrane were investigated. With bilayers formed onto the Nafion membrane, methanol permeability has been reduced


expectedly but at an expense of proton conductivity to some extent. It was found that the

Polyelectrolyte membrane

concentration of polyelectrolyte solutions (used to modify the Nafion membrane) and the

Layer-by-layer self assembly

number of deposited bilayers are key factors affecting the proton conductivity and meth-

Direct methanol fuel cell

anol permeability of the composite membrane. However, the characteristic factors (ratio of


proton conductivity to methanol permeability) of the composite membranes were all higher than that of Nafion 117 and the single cell performance has been dramatically enhanced








membrane. Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

Introduction The advanced development of wireless and portable communication technologies such as laptop computers, mobile phones have encountered the pressing need for high energy-density power sources for consumer use. However, the

lithium-ion based battery or other commercially available rechargeable battery are unable to meet the requirement of increasingly high energy density demand and long lifespan. Direct methanol fuel cell (DMFC), which is one of the most promising alternatives, has attracted tremendous attention for such application because of much higher volumetric and gravimetric energy density due to the use of liquid methanol,

* Corresponding author. Tel./fax: +65 67905591. ** Corresponding author. Tel.: þ65 67906957; fax: þ65 67954634. E-mail addresses: [email protected] (Y. Xue), [email protected] (S. Chan). http://dx.doi.org/10.1016/j.ijhydene.2014.11.129 0360-3199/Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.


i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 8 7 7 e1 8 8 5

more durable, instant recharging capability, lower weight and more compact structure [1e4]. Proton-conducting polyelectrolyte membrane (PEM) is a key component of DMFC. An ideal PEM must fulfill a number of operational requirements: an electronic insulator, a fuel barrier and a good proton conductor. So far, perfluorinated sulfonic acid membrane, such as Nafion® from Dupont, is widely used as the proton exchange membrane in direct methanol fuel cells because of its excellent chemical, mechanical and thermal stability, as well as its relatively high proton conductivity when fully hydrated. However, its high rate of fuel crossover has led to low fuel cell efficiency and hindered the practical realization of the DMFC. In DMFC, the methanol can permeate through the proton conductive membrane by diffusion as a result of the concentration gradient and electro-osmotic drag when under polarization. The former dominates when the fuel cell operates at low current density, whereas the latter dominates when the fuel cell operates at high current density. As a result of methanol crossover, fuel efficiency is reduced and a mixed potential occurs due to the methanol oxidation and oxygen reduction occur concurrently at the cathode, leading to competition in reactions and hence performance degradation of the cell. Therefore, low fuel permeability is desirable [5]. There have been continuing research efforts to reduce the methanol permeability of Nafion membranes or to look out for alternative materials to replace it, such as in situ impregnation of inorganic components, blending with another low methanol permeable polyelectrolytes, introduction of other sulphonated polyelectrolyte, such as Sulfonated poly (ether ether ketone) (SPEEK), Sulfonated poly(phenylene oxide) SPPO, Sulfonated Polysulfone(SPSU), etc. [6e14]. However, the leaching of the inorganic component, the compatibility between two blending polymers and the compatibility between membrane and electrode in membrane electrode assembly (MEA) are still issues associated with these approaches. The layer-by-layer (LbL) self-assembly technique is essentially a method based upon alternate depositions of oppositely charged polyelectrolytes onto a membrane substrate. LbL thin films with controlled structure and composition in the nanometer scale are formed due to the electrostatic attraction, without affecting the mechanical and chemical stability of the substrate. A wide variety of materials could be deposited by LbL method, including polyions, metals, ceramics, nanoparticles, and biological molecules [15e17]. Tieke and coworkers reported substantial works on self-assembled polyelectrolyte multilayer membranes based on LbL technique, showing high selective ion permeation and separation factor [18,19]. In the other study, Jiang and co-workers reported that Nafion 1135 membrane was deposited by poly(diallyldimethylammoniumchloride) (PDDA) and polystyrene sulfonic acid (PSS) and observed that the methanol crossover current density significantly decreased with an increase in the number of deposited layer [20], while Hu¨seyin Deligoz investigated the effect of PDDA/PSS LbL deposition conditions such as pH level, concentration of polyelectrolyte solutions and dipping time on the proton conductivity and degree of methanol crossover [21]. Chitosan (CHI), which is a cationic polyelectrolyte similar to PDDA, was reported to have excellent alcohol/water separation factor, thus it was expected to reduce methanol

crossover dramatically when used to modify the Nafion membrane [22]. PVS, on the other hand, with high charge density was used in LbL with chitosan for obtaining a small hydrophobic/hydrophilic separation between polymeric Nbases and polymeric sulphonic acids. The less connected hydrophilic channels on the Nafion surface were designed to reduce the methanol permeation while maintaining high proton conductivity. In this study, the effect of the concentration of polyelectrolyte solution and number of deposited bilayers on proton conductivity, methanol permeability and single cell performance were investigated thoroughly.

Experimental Membrane modification The pre-treatment of Nafion membrane to remove the impurities was conducted by successively immersing the membrane in 3 wt.% H2O2, distilled water, 8 wt.% H2SO4 and then in distilled water again, all at 80  C for 30 min in each step. A moderate amount of Chitosan powder (medium molecular weight, deacetylation degree 75%e85%, Sigma Aldrich) was dissolved into 1% (v/v) acetic acid to achieve different concentration solution. The solution was mechanically stirred for about 4 h until the powder was dissolved completely. A moderate volume of Polyvinyl sulfuric acid salt was diluted with DI-water to make the solution. Both solutions with three different concentrations of 0.01 M, 0.025 M and 0.05 M were prepared. The LbL self-assembled CHI/PVS multilayers was achieved by alternate dipping of the pretreated Nafion membrane in 100 ml CHI and PVS solutions for 15 min at room temperature. Before dipping, the pH level of both solutions was adjusted to 1.8 by using hydrochloric acid and sodium hydroxide. After each dipping step, the membrane was rinsed with Milli-Q water to remove residual CHI or PVS. The alternate dipping of CHI and PVS were repeated to achieve the desired number of deposited bilayers. The LbL self-assembled membrane was finally cleaned with Milli-Q water. In order to confirm the successful deposition of CHI and PVS on the Nafion surface

Fig. 1 e Schematic procedure of layer-by-layer self assembly.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 8 7 7 e1 8 8 5

after LbL treatment, the sample only modified with CHI was prepared by dipping Nafion in 0.05 M chitosan for 15 min and then rinsed with DI-water to remove the residual chitosan, and then dried at room temperature. The sample was denoted as Nafionechitosan. The schematic of LbL self-assembly process is shown in Fig. 1. In this study, all the membranes were denoted as Nafion-xbilayer-y, where x is the number of deposited bilayers, while y is the concentration of the polyelectrolyte solution.

FTIR spectra of the membranes in the wavenumber range of 4000e500 cm1 were obtained on a Thermo Scientific Nicolet 6700 FT-IR spectrometer. The cross sections of the membrane were examined by a JEOL JSM-7600F field emission scanning electron microscope (FESEM). Atomic force microscopy (AFM) is a very high-resolution scanning probe microscopy, which is believed to be one of the most powerful tools for imaging, measuring, and manipulating matter at the nanoscale. Surface morphological information of dry samples was collected by 154 atomic force microscopy (XE-100, PSIA) in non-contact mode (NC-155 AFM). The thermal gravimetric analysis (TG) was carried out on a TGA Q500 analyzer. The thermal measurements were carried out in N2 with temperature ramping from room temperature to 700  C at the rate of 20  C per minute. DSC measurements were conducted on a DSC Q200 differential scanning calorimeter from TA Instruments. A sample mass of about 10 mg was placed in aluminum pan, purged with nitrogen, heating from 0  C to 400  C at a heating rate of 20  C min1.

Water uptake and contact angle Water uptake was measured in the temperature range 25e60  C. For the water uptake measurement, a membrane was first equilibrated in deionized water at the desired temperature for 24 h and weighted immediately after the removal of surface water (Wwet). The membrane was then dried at 80  C until a constant weight was reached (Wdry). The water uptake (WU) of the membrane was calculated as: WU ¼ Wwet  Wdry

Wdry  100%

measuring cell was mounted on a Teflon holder with two stainless steel flat outer current-carrying electrodes (2 cm apart) and two platinum wire inner potential-sensing electrodes (1 cm apart). A membrane strip (5 cm  1 cm) was mounted on the holder. The impedance was measured with a Solartron Electrochemical Interface (1255B) operating in the frequency range of 1 MHz down to 50 Hz. Proton conductivity (k) was calculated based on the following expression: k ¼ L=RWd

Characterization of membrane


To measure the hydrophobic/hydrophilic behavior of membrane, the contact angle of membrane was detected by using a contact angle measuring instrument (KYOWA Interface Measurement and Analysis System FAMAS). The wet membrane was firstly dried under vacuum at 80  C for 12 h. Dry Membrane samples were then mounted on glass slides to provide a flat surface for analysis. The drop method was used to measure the contact angle of deionized water on the surface of the membranes at ambient temperature.

Proton conductivity Proton conductivity is one of the crucial properties of proton conductive membranes, which can directly affect the fuel cell performance. Proton conductivities of the membranes were measured by the standard four-point probe method. The



where R is the measured membrane resistance, L is the distance between the potential-sensing electrodes, and W and d are the width and thickness of the membrane strip, respectively.

Methanol permeability Methanol permeability measurements were carried out in a home-made diffusion cell with the membrane fastened firmly in between two compartments. One compartment was filled with 100 ml 2 M methanol solution, and the other with 100 ml de-ionized water. The contents of the two compartments were homogenized with magnetic stir bars. The effective diffusion area was 3.4 cm2. The concentration gradient set up across the membrane generated a net methanol flux from the methanol compartment to the deionized water compartment. The methanol concentration in the deionized water compartment was monitored as a function of diffusion time by gas chromatography (Shimadzu GC2010 with flame ionization detector, using 1-butanol as the internal standard). It is noted that the methanol permeability reported in this study is actually diffusion control with the methanol concentration gradient as the driving force. Methanol permeability (P) was calculated based on the following relationship:  A$P$CA $t CB t ¼ VB $L


where CA is the initial methanol concentration of the methanol solution, CB(t) is the methanol concentration in the water compartment at time t, VB the volume of deionized water in the water compartment, L the thickness of the membrane and A the effective permeation area.

Single cell performance To evaluate the performance of the DMFCs, a membrane electrode assembly (MEA) was fabricated by placing a LbL selfassembly composite membrane in between the anode and cathode followed by hot pressing at 135  C, 50 kg/cm2 for 150 s. The anode and cathode were obtained from brushing asprepared catalyst ink onto the gas diffusion layer. Electrocatalysts used in the anode and cathode were 60% PtRu/C and 60% Pt/C (both from Alfa-Aesar). The catalyst, Nafion solution (Dupont, 5 wt%), distilled water and isopropyl alcohol were mixed with a certain ratio and then ultrasonic for 30 min to prepare the catalyst ink. The gas diffusion layers used were GDS2240 (Ballard) for anode and 30% PTFE treated TGP-H-090


i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 8 7 7 e1 8 8 5

(Toray) for cathode. The metal loading in the catalyst layers of both anode and cathode were 3.0 mg/cm2. DMFC performance were evaluated with single-cell fixture (with an active area of 4 cm2) by feeding 2 M methanol solution and dry oxygen at flow rates of 5 and 60 mL min1 (without backpressure) at ambient temperature, respectively. The steady-state fuel cell polarization data were collected by Solartron Electrochemical Interface (1255B).

Result and discussion FTIR spectrum FTIR of the solid membrane was carried out to confirm the presence of the bilayers on the Nafion surface after LbL treatment. The FTIR spectra of Nafion, Nafionechitosan and Nafion-1bilayer-0.05 are shown in Fig. 2. The peaks at 3460, 1200, 1147, 1056, 981, 969 and 625 cm1 were ascribed to the characteristic functional groups of Nafion [23,24]. Characteristic peaks at 1635 cm1 and 1530 cm1 were observed in the spectrum of Nafionechitosan, which were assigned to the anti-symmetrical and symmetric deformation of the protonation of chitosan amine group NHþ 3 [25,26]. The presence of protonation of amine group NHþ 3 has confirmed the successful deposition of chitosan on the Nafion. The other peaks at 1380 cm1 (CH3 symmetric deformation vibration and CH bending vibration), 1095 cm1 (eCeOeC), 2935 cm1, (CeH) and 2890 cm1, (CeH) were also observed in the spectrum of NafioneChitosan [27]. The broad peak centered at 3437 cm1 was assigned to the stretching vibration of OeH and NeH [25]. The peaks in the sample of Nafion-1bilayer-0.05 were similar to that of Nafionechitosan except the intensity of the peaks ascribed to chitosan were a little bit weaker, which might be due to the introduction of the PVS layer on the CHI layer. The absorptions ascribed to the sulfonic acid group of PVS might be overlapped with the peaks of Nafion.

Fig. 2 e FTIR spectra of Nafion, Nafionechitosan and Nafion-1bilayer-0.05.

Morphology The SEM image of Nafion-1bilayer-0.05 cross section is shown in Fig. 3, it was clear that there was a sharp boundary between bulk Nafion and the bilayers of CHI and PVS (arrows pointing part), indicating that only the polyelectrolyte was deposited onto the surface, but not penetrating into the Nafion membrane. The cross section of pristine Nafion membrane was very homogeneous. The thickness of the bilayer was less than 1 mm, similar results was reported for (poly(allylamine hydrochloride)) PAH/PSSeNafion membrane and PAH/PVSeNafion composite membrane [28,29]. Fig. 4 shows the AFM image of pure Nafion 117 membrane and Nafion 117-3bilayer-0.025, Nafion 117-5bilayer-0.025 in dry state (scan rate, 0.8 Hz; scan area, 1 mm  1 mm). It can be seen that the fluctuation along the altitude on the surface of each membrane was less than 20 nm. The surface of pristine Nafion membranes was generally smooth (Fig. 4a). The fluctuation along the altitude on the surface was 10 nm. After 3 CHI-PVS bilayers were deposited, the surface roughness of the membrane has increased dramatically with the formation of tiny pinnacles, but the fluctuation along the altitude on the surface remains as 10 nm (Fig. 4b). The surface morphology of Nafion 1175bilayer-0.025 was similar to that of Nafion 117-3bilayers0.025, but the fluctuation along the altitude on the surface was double of that of pristine Nafion membrane and Nafion 1173bilayers-0.025 (Fig. 4c). AFM results show the direct evidence of the formation of CHI-PVS multilayers on Nafion 117 by LbL self-assembly processes. Similar morphology was reported for other LbL self-assembly membranes in the literature [30].

Thermal property Fig. 5a shows the TG/DTG curves of Nafion membrane and LbL modified membranes. The Nafion membrane seemed to decompose in 4 stages, which could be clearly seen in the derivative trace. The mass loss between 25 and 290  C was mainly attributed to the loss of residual moisture, while the second loss stage between 290 and 400  C was associated to

Fig. 3 e FE-SEM images of the cross section of Nafion1bilayer-0.05.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 8 7 7 e1 8 8 5

Fig. 4 e AFM image of a: Nafion, b: Nafion-3 bilayer-0.025, c: Nafion-5bilayer-0.025.

Fig. 5 e Thermal properties of Nafion and LbL composite membrane: (a) TG-DTG curves; (b) DSC curves.



i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 8 7 7 e1 8 8 5

desulfonation process combined with the decomposition of the ether groups (CeOeC) on the side-chains, and the third stage at 400e470  C was due to side-chain decomposition, the final stage at 470e560  C was related to decomposition of the poly(tetrafluoroethylene) backbone [31,32]. The LbL composite membranes show similar TG/DTG character with Nafion membrane. The DTG curves indicated that the second mass loss stage started at a lower temperature than that for the bare Nafion, which was probably due to the decomposition of chitosan and PVS [33]. The difference in the TG/DTG curves also suggested the existence of CHI and PVS. Fig. 5b shows the DSC curves of Nafion membrane and LbL modified membranes. A typical DSC curve of Nafion had three endothermic peaks. The first endothermic peak appeared at about 50e120  C, which may be taken as the cluster transition temperature. That can be accompanied by polymer contraction due to the vaporization and desorption of absorbed moisture with an increase in entropy. The second peak was a weak and broad endothermic peak and was present at around 200  C, which was assigned to the melting peak of the nonpolar crystallite backbone, while the weak endothermic peak at ~325  C could be related to the desulfonation process that occurred during the heating, as observed in the TG curve [32,34e36]. A sharp peak at 230e250  C was observed, which could be related to SO2 liberation during the heating process [32]. The composite membranes after Layer by Layer modification show similar DSC trace with Nafion membrane because the surface modification can't affect the polymer matrix and the inside sulfonic acid groups on the segment motions of perfluoro chains. It is notable that an exothermic peak near 312  C appeared in sample Nafion-5bilayer-0.05, which was possibly related to decomposition procedure of chitosan [37]. However, the exothermic peak was not found in the other two LbL composite membrane, which was may be due to the absorbed amount of chitosan was too small to be detected.

concentration polyelectrolyte was used to modify the membrane. The water uptake of the bulk Nafion membrane before and after LbL modification should be similar, the gradually increase in water uptake after LbL modification is attributed to the incorporation of the strong hydrophilic of chitosan and PVS [26]. The hydrophilic properties of the Nafion surface before and after LbL modification were investigated by contact angle measurements. Fig. 6 shows the differences in the form of the water drops on the surface of the Nafion-117 membrane and Nafion-5bilayer-0.05. It can be found that the surface of pristine Nafion membrane was a bit of hydrophobic due to it is based on perfluorinated polymer(Fig. 6a) [38]. However, after modified with 5 bilayers of 0.05 CHI and PVS (Fig. 6b), the surface became very hydrophilic, confirming the strong hydrophilic of CHI and PVS bilayers. All the contact angle results of the modified membranes were listed in Table 1. It can be seen that the hydrophilicity of the surface of the LbL modified Nafion membrane was gradually increased with the increase of the deposited bilayer numbers or the modified polyelectrolyte concentration. The trend of the effect of bilayer numbers and modified polyelectrolyte concentration on the contact angle was consistent with that on water uptake, which was a clear indication of the changes due to the incorporation of more hydrophilic CHI and PVS. Similar trend was reported with the other ion pairing polyelectrolyte mutilayers [26].

Water uptake and contact angle Table 1 shows the extents of water uptake in the temperature range 25  C to 60  C. It can be found that the water uptake gradually increased when more bilayers deposited on the Nafion surface at each polyelectrolyte concentration, and at the same deposited bilayers, it gradually increased when high

Table 1 e Water uptake of Nafion and LbL composite membranes at different temperatures and their contact angles at room temperature. Sample

25  C

40  C

60  C

Contact angle (Deg)

Nafion 117 Nafion-1bilayer-0.01 Nafion-3bilayer-0.01 Nafion-5bilayer-0.01 Nafion-1bilayer-0.025 Nafion-3bilayer-0.025 Nafion-5bilayer-0.025 Nafion-1bilayer-0.05 Nafion-3bilayer-0.05 Nafion-5bilayer-0.05

18.89 19.57 19.93 20.82 20.86 22.98 23.26 22.75 24.22 25.82

21.37 22.95 23.48 23.51 22.98 25.43 26.63 26.28 26.39 26.74

25.64 25.82 27.21 27.27 27.94 29.06 29.25 30.35 30.75 31.87

99.4 93.0 92.7 92.7 91.6 90.5 90.5 87.9 61.1 50.4

Fig. 6 e Contact angle image of (a) Nafion 117 and (b) Nafion-5bilayer-0.05.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 8 7 7 e1 8 8 5

Proton conductivity Fig. 7 shows the effect of the concentration and number of bilayers on proton conductivity. It can be observed that the concentration of the polyelectrolyte solution and the number of bilayers can affect the proton conductivity significantly. At each concentration, the higher the number of bilayers deposited on the Nafion surface, the lower the proton conductivity will be. For example, the proton conductivity of the pristine Nafion was 9.52  102 S cm1. When modified with 0.01 M chitosan and PVS, the proton conductivity decreased from 7.89  102 S cm1 for 1 bilayer to 7.73  102 S cm1 for 3 bilayers to 7.38  102 S cm1 for 5 bilayers. Similar phenomena were observed when Nafion was modified with 0.025 M and 0.05 M of CHI and PVS. The reduction in proton conductivity of self-assembled composite membrane with increasing number of bilayers on the Nafion surface was attributed to the blocking of charge carrier species that were responsible for proton conduction [21]. The higher the number of bilayers deposited on the surface, the more the seriousness of blocking of charge carriers, and hence the lower the proton conductivity will be. In order to evaluate the effect of the concentration of polyelectrolyte solution on proton conductivity, we fixed the number of bilayers but varying the concentration of the solution. Results in Fig. 7 show that the proton conductivity decreases with the increase of the concentration of polyelectrolyte solution concentration for a fixed number of deposited bilayers. For example, when 1 bilayer was formed on Nafion surface, the proton conductivity of the composite membrane with 0.01 M CHI and PVS solution was 7.89  102 S cm1, but 7.68  102 S cm1 for 0.025 M CHI and PVS solution and 7.01  102 S cm1 for 0.05 M CHI and PVS solutions. The reduction in proton conductivity was attributed to more polyelectrolyte absorbed when dipping in a higher concentration of polyelectrolyte. As a result, the thickness of deposited bilayers would be thicker, causing a reduction in proton conductivity of the composite membrane.

Methanol permeability Fig. 8 shows the effect of the concentration and number of bilayers on the methanol permeability. It can be seen that

Fig. 7 e The effect of concentration & absorbed bilayer numbers on proton conductivity.


both the concentration of the polyelectrolyte solutions and the number of deposited bilayers are strong factors affecting the permeability of the membrane. The methanol permeability trend in Fig. 8 mirrors that of proton conductivity in Fig. 7. Under the same number of bilayers, the higher the concentration of deposited polyelectrolyte solution, the lower the methanol permeability will be. For example, when 1 bilayer of CHI and PVS was deposited on Nafion surface, the methanol permeability of Nafion 117-1bilayer-0.01 was approximately three fourth of that of pristine Nafion, while the methanol permeability of Nafion 117-1bilayer-0.05 was half of that of pristine Nafion membrane. The same phenomena were reported for other polyelectrolyte [21]. The reason was that more deposition will occur with higher concentration over the same fixed area and over the same period of absorption. More deposition of methanol-blocking polyelectrolyte, like chitosan in this study, would mean that thicker methanol-resistant layers formed on the surface, and hence, leading to a reduction in methanol permeability. For each fixed concentration of polyelectrolyte solution used, the higher the number of bilayers deposited, the lower the methanol permeability will be. For example, the methanol permeability of pristine Nafion membrane was around 1.21  106 cm2/s, but when modified with 0.01 M polyelectrolyte solution, the methanol permeability was reduced to 9.02  107 cm2/s for 1 bilayer and further decreased to 8.59  106 cm2/s for 3 bilayers and 7.86  106 cm2/s for 5 bilayers. The decrease of methanol permeability with an increase in the number of bilayers was believed to be due to more methanol resistant layers formed on the surface. Jiang et al. employed polycations PDDA and PAH to form polyelectrolyte multilayers on Nafion membrane, and found that the reduction in methanol permeability is in the range of 24e43% for PDDA/polyanion bilayers and 8e14% for PAH/ polyanion bilayers [30]. In this study, the lowest permeability value obtained was Nafion-5bilayer-0.05, which presented a 50% decrease, suggesting that chitosan might have a better methanol resistance compared to those of PDDA and PAH. While both proton conductivity and methanol permeability decreased with the layer by layer modification, they do not decrease at the same rate. In this case, the characteristic

Fig. 8 e The effect of concentration & absorbed bilayer numbers on methanol permeability.


i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 8 7 7 e1 8 8 5

factor (also known as selective factor), which is defined as the ratio of proton conductivity to methanol permeability, is generally used as a measure of relative merits. The higher value of characteristic factor indicates higher conductivity and lower methanol permeability, and thus better cell performance would be expected. The characteristic factors of pristine Nafion membrane and the modified membranes at room temperature were shown in Fig. 9. It can be seen that the characteristic factor of all the modified membrane were higher than that of pristine Nafion membrane. The optimum number of deposited bilayers varies under different concentration of polyelectrolyte solutions. At lower concentration, for example 0.01 M and 0.025 M, the characteristic factor increases with an increase in the number of bilayers, and the optimum number of bilayers has yet to reach. However, for higher concentration of 0.05 M, the trend was opposite. Further increase in the number of bilayers would lead to a decrease in the selective factor, which is attributed to decreased conductivity under the more or less constant methanol permeability.

Fig. 10 e The single cell performance of Nafion 117 & membrane modified with 0.05 M CHI-PVS.

Conclusion Single cell performance The cell performance of DMFCs prepared with pristine Nafion membrane and after LbL self assembly treatment in 0.05 M CHI and PVS are shown in Fig. 10. The open circuit voltage (OCV) was 0.665 V for the cell with an unmodified membrane. OCV increased after self-assembled modification, which was due to less methanol transport through the membrane to cathode side. From Fig. 10, it was clear that the best performance was obtained for the cell with 1 self-assembled bilayer, and the performance is in the following order: Nafion 1171bilayer-0.05 > Nafion 117-3bilayer-0.05 > Nafion 117-5 bilayer-0.05 > Nafion 117, which is in the same order as that of characteristic factors. The power density of Nafion 1171bilayer-0.05 was 24 mW/cm2, which is 33% increase in power density compared with that of a cell with unmodified Nafion 117 membrane (18 mW/cm2), indicating a significant improvement in cell performance by LbL self assembly modification.

LbL films containing up to 5 CHI/PVS bilayers were assembled on Nafion membrane by alternately immersing the membrane into the CHI and PVS polyelectrolyte solutions at various concentrations. The thermal stability of the LbL modified membrane was similar with that of pristine Nafion membrane, but the water uptake was gradually increased due to the incorporation of more hydrophilic CHI and PVS. The concentration and number of bilayers of LbL film have proved to affect the proton conductivity and methanol permeability significantly. In general, the methanol permeability decreased expectedly with the number of bilayers but at the expense of the proton conductivity of the composite membrane. The characteristic factors of all the modified membrane were higher than that of pristine Nafion membrane. The DMFC single cell performance was dramatically enhanced with the self-assembled bilayers deposition. The membrane modified with 1 bilayer of chitosan and PVS exhibited 33% increase in cell power density compared to that of pure Nafion membrane.


Fig. 9 e The effect of concentration & bilayer numbers on selective factor.

[1] Zhang L, He HQ, Kamal R, Zhou WJ, Xue YH, Ding OL, et al. Fabrication of novel phosphotungstic acid functionalized mesoporous silica composite membrane by alternative gelcasting technique. J Power Sources 2013;221:318e27. [2] Zhou WJ, Li M, Ding OL, Chan SH, Zhang L, Xue YH. Pd particle size effects on oxygen electrochemical reduction. Int J Hydrog Energy 2014;39:6433e42. [3] Xue YH, Zhang L, Zhou WJ, Chan SH. Pd nanoparticles supported on PDDA-functionalized carbon black with enhanced ORR activity in alkaline medium. Int J Hydrog Energy 2014;39:8449e56. [4] Wu RB, Xue YH, Qian X, Liu H, Zhou K, Chan SH, et al. Pt nanodendrites anchored on bamboo-shaped carbon nanofiber arrays as highly efficient electrocatalyst for oxygen reduction reaction. Int J Hydrog Energy 2013;38:16677e84.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 0 ( 2 0 1 5 ) 1 8 7 7 e1 8 8 5

[5] Barragan VM, Ruiz-Bauza C, Villaluenga JPG, Seoane B. Transport of methanol and water through Nafion membranes. J Power Sources 2004;130:22e9. [6] Kerres JA. Blended and cross-linked ionomer membranes for application in membrane fuel cells. Fuel Cells 2005;5:230e47. [7] Smitha B, Sridhar S, Khan AA. Solid polymer electrolyte membranes for fuel cell applicationsda review. J Membr Sci 2005;259:10e26. [8] Xue YH, Fu RQ, Wu CM, Lee JY, Xu TW. Acidebase hybrid polymer electrolyte membranes based on SPEEK. J Membr Sci 2010;350:148e53. [9] Ahmad H, Kamarudin SK, Hasran UA, Daud WRW. Overview of hybrid membranes for direct-methanol fuel-cell applications. Int J Hydrog Energy 2010;35:2160e75. [10] Wu D, Wu L, Woo J-J, Yun S-H, Seo S-J, Xu TW, et al. A simple heat treatment to prepare covalently crosslinked membranes from sulfonated poly(2,6-dimethyl-1,4phenylene oxide) for application in fuel cells. J Membr Sci 2010;348:167e73. [11] Xu TW, Wu D, Wu L. Poly(2,6-dimethyl-1,4-phenylene oxide) (PPO)da versatile starting polymer for proton conductive membranes (PCMs). Prog Polym Sci 2008:894e915. [12] Wu D, Xu TW, Wu L, Wu YH. Hybrid acid-base polymer membranes prepared for application in fuel cells. J Power Sources 2009;186:286e92. [13] Feng K, Tang BB, Wu PY. Selective growth of MoS2 for proton exchange membranes with extremely high selectivity. ACS Appl Mater Interfaces 2013;5:13042e9. [14] Iannaci A, Mecheri B, D'Epifanio A, Licoccia S. Sulfated zirconium oxide as electrode and electrolyte additive for direct methanol fuel cell applications. Int J Hydrog Energy 2014;39:11241e9. [15] Zhang HW, Huang HP, Shen PK. Methanol-blocking Nafion composite membranes fabricated by layer-by-layer selfassembly for direct methanol fuel cells. Int J Hydrog Energy 2012;37:6875e9. [16] Lin HD, Zhao CJ, Ma WJ, Li HT, Na H. Layer-by-layer selfassembly of in situ polymerized polypyrrole on sulfonated poly(arylene ether ketone) membrane with extremely low methanol crossover. Int J Hydrog Energy 2009;34:9795e801. [17] Zhao C, Lin H, Zhang Q, Na H. Layer-by-layer self-assembly of polyaniline on sulfonated poly(arylene ether ketone) membrane with high proton conductivity and low methanol crossover. Int J Hydrog Energy 2010;35:10482e8. [18] Tieke B, van Ackern F, Krasemann L, Toutianoush A. Ultrathin self-assembled polyelectrolyte multilayer membranes. Eur Phys J E 2001;5:29e39. [19] Lutz Krasemann AT, Tieke Bernd. Self-assembled polyelectrolyte multilayer membranes with highly improved pervaporation separation of ethanol/water mixtures. J Membr Sci 2001;181:221e8. [20] Xiang Y, Lu SF, Jiang SP. Layer-by-layer self-assembly in the development of electrochemical energy conversion and storage devices from fuel cells to supercapacitors. Chem Soc Rev 2012;41:7291e321. € € z H, Yılmaztu¨rk S, Karaca T, Ozdemir [21] Deligo H, Koc¸ SN, € ¨ zo € mer F, et al. Self-assembled polyelectrolyte Oksu multilayered films on Nafion with lowered methanol crossover for DMFC applications. J Membr Sci 2009;326:643e9. [22] Cogo LC, Batisti MV, Pereira-da-Silva MA, Oliveira ON, Nart FC, Huguenin F. Layer-by-layer films of chitosan,

[23] [24]








[32] [33]







poly(vinyl sulfonic acid), and platinum for methanol electrooxidation and oxygen electroreduction. J Power Sources 2006;158:160e3. Ludvigsson M, Lindgren J, Tegenfeldt J. FTIR study of water in cast Nafion films. Electrochim Acta 2000;45:2267e71. Laporta M, Pegoraro M, Zanderighi L. Perfluorosulfonated membrane (Nafion): FT-IR study of the state of water with increasing humidity. Phys Chem Chem Phys 1999;1:4619e28. Amaral IF, Granja PL, Barbosa MA. Chemical modification of chitosan by phosphorylation: an XPS, FT-IR and SEM study. J Biomater Sci-Polym Ed 2005;16:1575e93. Almodovar J, Place LW, Gogolski J, Erickson K, Kipper MJ. Layer-by-layer assembly of polysaccharide-based polyelectrolyte multilayers: a spectroscopic study of hydrophilicity, composition, and ion pairing. Biomacromolecules 2011;12:2755e65. Valentin R, Bonelli B, Garrone E, Di Renzo F, Quignard F. Accessibility of the functional groups of chitosan aerogel probed by FT-IR-monitored deuteration. Biomacromolecules 2007;8:3646e50. Yilmazturk S, Deligoz H, Yilmazoglu M, Damyan H, Oksuzomer F, Koc SN, et al. A novel approach for highly proton conductive electrolyte membranes with improved methanol barrier properties: layer-by-layer assembly of salt containing polyelectrolytes. J Membr Sci 2009;343:137e46. € z H, Yılmazog  lu M, Damyan H, Yılmaztu¨rk S, Deligo € ¨ zo € mer F, Koc¸ SN, et al. Self-assembly of highly charged Oksu polyelectrolyte complexes with superior proton conductivity and methanol barrier properties for fuel cells. J Power Sources 2010;195:703e9. Jiang SP, Tang H. Methanol crossover reduction by Nafion modification via layer-by-layer self-assembly techniques. Colloids Surf A Physicochem Eng 2012;407:49e57. Tan S, Belanger D. Characterization and transport properties of Nafion/polyaniline composite membranes. J Phys Chem 2005;109:23480e90. de Almeida SH, Kawano Y. Thermal behavior of Nafion membranes. J Therm Anal 1999;58:569e77. Wang J-P, Chen Y-Z, Wang Y, Yuan S-J, Sheng G-P, Yu H-Q. A novel efficient cationic flocculant prepared through grafting two monomers onto chitosan induced by Gamma radiation. RSC Adv 2012;2:494e500. Iwai Y, Yamanishi T. Thermal stability of ion-exchange Nafion N117CS membranes. Polym Degrad Stabil 2009;94:679e87. Lin Y-F, Hsiao Y-H, Yen C-Y, Chiang C-L, Lee C-H, Huang C-C, et al. Sulfonated poly(propylene oxide) oligomers/Nafion (R) acid-base blend membranes for DMFC. J Power Sources 2007;172:570e7. Hong L, Zhou YJ, Chen NP, Li K. Association of Nafion with polypyrrole nanoparticles in a hydrophilic polymer network: effects on proton transport. J Colloid Interface Sci 1999;218:233e42. Wan Y, Lu X, Dalai S, Zhang J. Thermophysical properties of polycaprolactone/chitosan blend membranes. Thermochim Acta 2009;487:33e8. Lee HJ, Nam EJ, Woo JJ, Moon SH, Lee J. Improved dimensional stability of Nafion membrane modified using a layer by layer self-assembly of biophilic polymers. Curr Appl Phys 2012;12:1235e8.