- Email: [email protected]
Enhancement of fuel cell performance with less-water dependent composite membranes having polyoxometalate anchored nanofibrous interlayer
Journal of Power Sources 326 (2016) 482e489
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Enhancement of fuel cell performance with less-water dependent composite membranes having polyoxometalate anchored nanofibrous interlayer Ebrahim Abouzari-lotf a, *, Mohan V. Jacob b, Hossein Ghassemi c, Arshad Ahmad a, Mohamed Mahmoud Nasef a, d, Masoumeh Zakeri a, Shahram Mehdipour-Ataei e a
Advanced Materials Research Group, Center of Hydrogen Energy, Institute of Future Energy, Universiti Teknologi Malaysia, 54100, Kuala Lumpur, Malaysia Electronics Materials Lab., College of Science, Technology and Engineering, James Cook University, Townsville, Queensland 4811, Australia Department of Macromolecular Science & Engineering, Case Western Reserve University, 44106-7202 Cleveland, OH, USA d MalaysiaeJapan International Institute of Technology, Universiti Teknologi Malaysia, 54100 Kuala Lumpur, Malaysia e Iran Polymer and Petrochemical Institute, P.O. Box 14965/115, Tehran, Iran b c
h i g h l i g h t s Composite membranes consisting ion conductive nanofibers are proposed for PEMFCs. Dimensional stability improved and water uptake reduced compared to Nafion 112. The proton conductivity at low humidity is improved in composite membranes. OCV enhanced by about 100 mV than Nafion 112 when tested at 60 C and 40% RH. Maximum power density increased by 35% in the composite.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 March 2016 Received in revised form 4 July 2016 Accepted 7 July 2016 Available online 15 July 2016
Polyoxometalate immobilized nanofiber was used to fabricate low gas permeable layer for composite membranes designed for proton exchange membrane fuel cell (PEMFC) operating at low relative humidity (RH). The composite membranes revealed enhanced proton conductivity in dry conditions compared with state-of-the-art pristine membrane (Nafion 112, N112). This was coupled with a low fuel crossover inheriting the composite membranes about 100 mV higher OCV than N112 when tested in PEMFC at 60 C and 40% RH. A maximum power density of up to 930 mW cm2 was also achieved which is substantially higher than the N112 under similar conditions (577 mW cm2). Such remarkable performance enhancement along with undetectable leaching of immobilized polyoxometalate, high dimensional stability and low water uptake of the composite membranes suggest a strong potential for PEMFC under low RH operation. © 2016 Elsevier B.V. All rights reserved.
Keywords: Polymer electrolyte membrane fuel cells Conductive nanofibrous membrane Polyoxometalate Membrane electrode assembly High energy density
1. Introduction Proton Exchange Membrane fuel cell (PEMFC) is one of the key emerging technologies that is attracting remarkable effort with the aim to provide alternative environmentally friendly and efficient power sources. The advances in proton exchange membranes (PEMs) is critical for improving the performance of PEMFC [1e4].
* Corresponding author. E-mail address: [email protected] (E. Abouzari-lotf). http://dx.doi.org/10.1016/j.jpowsour.2016.07.027 0378-7753/© 2016 Elsevier B.V. All rights reserved.
Membranes comprising perfluorosulfonic acid polymers such as Nafion have been used extensively due to their high conductivity and stability. However, these materials need to be saturated with water to obtain practical levels of proton conductivity. Such hydration undermines the permeability of H2 and O2 due to the involvement of the formed hydrated ionic clusters in the gas permeation mechanism [5,6]. On the other hand, the retention and management of water within these types of membranes is challenging, costly and in most cases reduce the PEMFC performance. Therefore there is a strong demand for the PEMs to work at lower relative humidity and/or under anhydrous conditions.
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
Accelerated electrochemical reactions and simpler water management of fuel cells under lower RH and/or anhydrous conditions is challenged by dramatic losses of proton conductivity of perfluorinated sulfonic acid based membranes [2,7,8]. Typically, N112 losses one order of magnitude of its proton conductivity at around 60% RH which, increases two orders of magnitude with the reduction of RH to 25%. So, it was concluded that the current Nafion based membranes cannot meet the requirements for the practical applications in these conditions. Various designing strategies and advanced materials have been developed to improve the performance of PEMs operating under low RH [9]. One of the main strategies is based on reduction of the membrane thickness in a way leading to thinner membranes with lower internal resistance and improved water management during operation at lower RH and or elevated temperatures [7]. This is the main reason of progressive reduction in the thickness of commercial Nafion membranes for PEMFC over the past 2 decades (e.g. from ~183 mm in N117 to ~125 mm in N115 and then to 25 mm in N211) [10]. However, the thinner membranes are challenged by high fuel crossover and weak mechanical properties. On the other hand, low open-circuit voltage (OCV) and short PEM lifetime are the major concerns that have arisen on the practical application of 25 mmthick Nafion membranes in hydrogen fuel cells. Other strategies of designing less water dependent polymer electrolytes [11e13], modification of available electrolytes with water retention fillers [8,14e18], and introducing anhydrous proton conductive additives are also widely discussed to improve membrane limitations under lower RHs [19e23]._ENREF_15 Particularly, the incorporation of inorganic proton conductors such as heteropolyacids was found to increase the conductivity under dry and humidified conditions [20,24e27]. Heteropolyacids (such as phosphotungstic acid H3PW12O40, PTA) have a very strong Brønsted acidity, higher than both pure sulfuric acid and Nafion and approaching the superacid region. This high acidic strength is accomplished with fast reversible redox transformations under quite mild conditions [28]. In addition, this hydrophilic inorganic metal oxides have a high tendency to strongly accommodate water in their interlayer regions [29,30]. On the other hand, heteropolyacids could improve the stability of membranes by inhibiting the formation of peroxide/or catalyzing its decomposition [31e33]. The latter merit was confirmed by intense reduction in the release rate of F originated from fluorinated backbone degradation during PEMFC operating under relatively hot and dry conditions. However, although introducing of PTA to the membranes led to improved properties, yet PTA loading level and its direct contact with water are the serious concerns which need to be addressed. Low amount of PTA could be introduced into the membrane either through impregnation of porous substrate with an aqueous solution of PTA [22,34], or blending of PTA or its derivatives with a polymer followed by casting [35,36]. The composite membranes with PTA showed improved methanol barrier properties and/or water retention in DMFC [37e39], and enhanced PEMFC performance at low humidity and/or elevated temperatures [22,36,40]. However, due to the low level of PTA, loose interactions with polymer substrate and agglomeration, resulting composite membranes show low proton conductivity and PTA leaching during the fuel cell operation [41]. Copolymerization of PTA containing monomers was proposed to increase the loading level and minimize the leaching of PTA [42,43]. Such monomers are commonly prepared in the multistep synthetic procedure and therefore this method was not used widely. Recently, we proposed a simple and flexible method to introduce high level of PTA onto membranes [44]. Various amounts up to 50 wt% of PTA could be simply immobilized onto nylon nanofibers by self-immobilization. Such ion conductive nanofibers are capable
483
of fast proton transport [45]. On the other hand, leaching of PTA was addressed through assembling the membrane in a layered structure with two recast Nafion layers. The membranes showed an improved performance in DMFC and the conductivity remained quite stable even after more than 100 h of continuous testing. In this work, a modified method was used to immobilize high level of PTA on nanofibrous structure and sandwiched membranes were fabricated for low humidity PEMFC application. The effects of PTA loading level and nanofiber diameter on the water uptake, dimensional stability and proton conductivity were investigated in comparison with commercially available N112. Moreover, the effect of introducing PTA on open-circuit voltage (OCV) as well as singlecell performance under lower RH were also evaluated. 2. Experimental 2.1. Materials Technical grade nylon-6,6 (medium viscous, DSM Co., The Netherlands) was used to prepare the nanofibrous sheets using Electrospinner (Nanolab Instruments) electrospinning system. Thin Nafion layers with a thickness of 20 mm were prepared by casting a 15% concentrated aqueous solution of Nafion (5%, supplied from DUPONT). FeSO4.7H2O (98.1%, from Fischer Scientific Company, Fairlawn, NJ) was used for the preparation of Fenton agent for chemical stability test. Commercial catalyst of 40 wt % Pt/C (Johnson Matthey) was used in both the cathode and the anode. 2.2. Characterizations Transmission electron microscope (TEM) images were obtained with a TECNAI G2 F20 with an accelerating voltage of 200 kV. A piece of the sample was milled at liquid nitrogen using Ika Ultra Turrax tube disperser. The dark areas in the HRTEM images clearly show hydrophilic domains and the light areas represent hydrophobic domains. The atomic force microscope (AFM) images were obtained with a Park NX10 in a non-contact mode. Scanning Electron Microscope (SEM) images of nanofibrous sheets and composite membranes were taken on the Philips XL30 Field Emission Scanning Electron Microscope (FESEM) after coating with 5 nm Au. The cross-sectional morphology was examined using cryofracturing of samples in liquid nitrogen. Fourier transform-infrared attenuated total reflection (FT-IR-ATR) spectra were performed on Agilent Cary 660 spectrometer. X-ray diffraction pattern was obtained on the Philips X’Pert 1 X-ray diffractometer with graphite monochromatized Cu Ka radiation (l ¼ 1.5401 Å) at scanning rate of 2 min1 over a range of 2q ¼ 4e80 . Thermogravimetric analysis (TGA) was performed on the Perkin Elmer TGA7 under nitrogen atmosphere at a heating rate of 10 C min1. Leaching of the PTA from the membrane was measured by immersing the composite membranes in deionized water at room temperature and monitoring the concentration of PTA in water every 12 h using a UV-1800 UVevis spectrometer. The in-plane (sk ) and through-plane (s⊥ ) proton conductivities of the membranes were measured by using a four-point probe of Bekk Tech conductivity cells (BT-112) and two-point probe of homemade stainless steel cylindrical electrodes with diameter of 20 mm. The resistance of the membranes was measured by using a DC conductivity testing (Keithley 2400 sourcemeter) controlled by a Labview software. The potentiostat was set to apply a specific voltage between the two inner probes and measure resulting current. The slope of the data from current versus voltage measurement was used to calculate the resistance (R) and proton conductivity (s) was calculated according to the equation (1):
484
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
LðcmÞ
s S=cm ¼ RðUÞ AðcmÞ2 sk= ¼ ⊥
sk s⊥
(1)
(2)
where L is the distance in the direction of ion flow between voltage measurement probes and A is the area of the membrane. Since the membrane swell during fuel cell operation, it is important to employ the hydrated thickness to calculate the conductivity and the resistivity of the membranes. Anisotropic proton conductivity ratio was calculated from in-plane and through-plane conductivity values according to equation (2). In order to control the RH in a desired range, various saturated salt solutions were used and measurements were performed in a homemade chamber using BT112 cell. A chamber brought to equilibrium at a controlled temperature and humidity controlled by SHT75 sensor (with ±1.8% RH accuracy). Generally, the measured proton conductivity data have errors of ±8%. Membrane water uptake, defined as the difference in mass between fully hydrated and completely dry membrane, 4w, was calculated using the following equation:
4w ¼
Wwet Wdry 100 Wdry
(3)
where Wwet is the weight of the membrane after equilibration in water for 24 h and Wdry is the weight after complete drying in a vacuum oven at 80 C for 24 h. In a similar manner, the in-plane (Sk ), through-plane (S⊥ ), and volume swelling (SV ) of the membranes in water were calculated using the following equation:
Sx ¼
Xwet Xdry 100 Xdry
(4)
where x represents in-plane, through plane, or volume swelling measurements and Xwet and Xdry are the length, thickness, or volume of swollen and dry membranes, respectively. In order to obtain an accurate hydrated mass and swelling data, after removing the hydrated membranes from water, the sample was blotted quickly to remove surface water and measured immediately (the measurement was repeated 3 times). Oxidative stability of the membranes was evaluated by the Fenton agent consisting of 3% H2O2 as a source of peroxide and hydroperoxide radicals and 2 ppm FeSO4. Pre-weighed dry membranes were soaked in a Fenton aqueous solution at 80 C for up to 96 h and the stability was evaluated by recording the retained weight of the membranes after complete drying. 2.3. Preparation of nanofibers and composite membranes A polymer solution of 15e18 wt% concentrated nylon-6,6 was prepared in a mixture of 9/1 (v/v) of formic acid and water in a glass reactor. The polymer solution was electrospun under an applied voltage of 20 kV, a distance of 12 cm between the tip of the metal needle and the aluminum foil collector, a solution flow rate was 0.3e0.4 mL h1 and the collector drum rotation speed of 300 rpm. After 14 h of electrospinning, the fibrous sheet was collected and dried in a vacuum oven to completely remove the solvent. With adjusting the electrospinning parameters, two types of beads free nanofibrous sheets with mean diameters of 70 and 110 nm (based on the SEM micrographs) were obtained under two combinations of electrospinning parameters. PTA self-immobilization was performed by immersing the
nanofibrous sheet in an aqueous solution of PTA (5e12 wt%) at room temperature for periods up to 4 days under slow shaking. Successively the nanofibrous sheet was removed, washed with plenty of deionized water to remove unbounded PTA and sandwiched between two glass plates prior to vacuum drying for 24 h. Two immobilization levels of 39.0 and 53.9 wt% were obtained. A layered membrane composed of the PTA-immobilized nanofibers sandwiched between thin Nafion layers was prepared by mechanical compression between two ETFE films at 130 C. To enhance the adhesion between layers, Nafion solution was sprayed on the both sides of the PTA-immobilized nanofiber sheet before assembling. Hot-press with a pressure of 1700 psi was applied for 10 s for four times (samples were rotated 90 three times to ensure a uniform compression) followed by heat treatment at 135 C for 5 h in a vacuum oven. Finally, the membranes were soaked in 1 N sulfuric acid for 24 h, rinsed with deionized water and dried in a vacuum oven. 2.4. Preparation of membrane electrode assembly (MEA) and single cell operation The catalyst ink was prepared by adding electrocatalyst and 5 wt % of Nafion ionomer solution into a mixture of water and isopropanol (1:1) followed by ultrasonic treatment for 1 h. For MEA fabrication, composite membranes were pretreated by 1 h boiling consecutively in, 3 wt% H2O2, deionized water, 1 M H2SO4 and finally Milli-Q water. The catalyst-coated membrane was prepared by spraying the catalyst onto a membrane with a controlled loading of around 4 mg cm2 for both anode and cathode sides. The obtained MEA was dried for 4 h under atmospheric conditions and heat-treated in a vacuum oven at 60 C to remove residual solvents. Finally, it was sandwiched between two pieces of gas-diffusion layers (Toray TGPH-030 Carbon paper) and assembled in a single cell hardware with an active area of 5 cm2. For a comparison purpose, MEA with the same catalyst loading was prepared with commercial N112 membrane. The cell was operated at 60 C and the fluxes of H2 and O2 were controlled at 10 and 18 mL min1, respectively. The tests were started with the humidified gases with the humidifier temperatures of 70 C for both gases. After reaching a stable performance, to evaluate the effect of introducing PTAimmobilized nanofibrous sheet on the membrane performance at low humidity, the single cell was kept running with a humidity of around 40% for both gases. The cell voltages at different current densities were recorded after reaching stable cell performance. 3. Results and discussions 3.1. Membrane fabrication and characterization The concentration and resulting viscosity of the polymer solution played an important role in the morphology of electrospun nanofibers. For instance, at concentrations lower than 15 wt% nylon-6,6, many beads were formed due to the low viscosity of polymer solution. However, further increase in the concentration to 18 wt%, the spinnability was improved and beads free nanofibrous sheets were obtained. Moreover, the increase in the concentration of polymer solution resulted in an increase in the diameter of the electrospun nanofibers from 60e90 nm to 85e145 nm. Typical SEM micrographs of electrospun nanofibers is shown in the Fig. S1. In order to immobilize phosphotungstic acid on the nanofibrous sheet, the nylon mat was immersed in aqueous solution of PTA [44]. FTIR-ATR spectra of pristine nylon-6,6 nanofibers and corresponding PTA immobilized nanofibers are shown in Fig. 1. The spectrum clearly confirms the immobilization. There are four kinds of oxygen atoms in the Keggin-type PTA including the central
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
485
Fig. 1. Comparison of FTIR-ATR spectra of nanofibers (blue) and PTA immobilized nanofibers (red) in the expanded range of 500e1800 cm1. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
oxygen atoms (PeOa), two types of bridging oxygen atoms (WeObeW and WeOceW), and terminal oxygen atoms (W ¼ Od). The IR spectrum shows characteristic bands for each kinds at 1080 for PeOa, 983 for W ¼ Od, 893 for WeObeW, and 797 cm1 for WeOceW (Fig. S2). Upon anchoring onto nanofibers, as shown in Fig. 1, the frequency of the peak correspond to the WeOceW changes significantly (from 797 to 819 cm1). Smaller shift to the lower frequency was also observed for the WeObeW peak (from 893 to 896 cm1). The representing peaks for terminal and central oxygen atoms almost remain constant. So, it could be concluded that PTA is immobilized due to the preferential interactions reside at bridge-sharing (Oc) oxygen sites rather than the terminal (Od) site. Similarly, DFT calculations with 31P solid-state magic-anglespinning NMR confirm that these oxygen sites are preferred location of the Brønsted protons [46]. As shown, a minor shift in the carbonyl frequency of the amide group (from 1636 to 1633 cm1) was also observed which could be attributed to the formation of hydrogen bonding with PTA.
The PTA immobilization on nylon-6,6 nanofibers was further confirmed by high-resolution TEM imaging as shown in Fig. 2. As can be seen in high-resolution TEM images, there is no evidence for crystalline or agglomerated PTA molecules which reflecting a uniform immobilization. Accordingly, it can be suggested that the immobilization took place by chemical interaction which was further confirmed by disappearance of the PTA peaks and appearing of new peaks in X-ray diffraction (XRD) pattern (shown in Fig. 4b). Achieving a uniform immobilization of PTA in nylone-6,6 nanofibrous sheets depends on a number of parameters such as initial concentration of PTA in immobilization bath, immersion time and shaking speed. Particularly, high PTA concentration and prolonged immersion time provoke inhomogeneous distribution during immobilization. A typical example for non-uniform PTA distribution (57 wt%) onto the nanofibrous support prepared by immersion in highly concentrated PTA solution (12 wt%) for 10 days is shown in Fig. 3. Obviously, the PTA immobilized nanofibrous
Fig. 2. TEM image of PTA immobilized nanofiber, showing fine PTA immobilization onto nanofibrous matrix. The regions with different hydrophobicity could be seen clearly in c.
486
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
3.2. Membrane stability and proton conductivity
Fig. 3. SEM images of non-uniformly impregnated PTA on nylon-6,6 nanofibers.
display not only fragility and non-homogeneity but also leaching of the loaded PTA as indicated by the loss of 20 wt% of PTA upon immersion in water for 24 h. Immobilization level could be simply tuned by controlling the concentration of immersion bath and immersion time. Herein, by manipulating the immobilization parameters two different immobilization levels of 53.9 and 39.0% were obtained. The immobilization level was measured with gravimetric and confirmed by thermogravimetric analysis. Typical TGA thermograms of PTA immobilized nylon-6,6 nanofibers of different loading is shown in Fig. 4a. Uniform immobilization of PTA on the nanofiber surface was further confirmed in AFM images. Fig. 5 reveals the surface features of pristine and PTA immobilized nanofibers. In the vertical direction, darker areas represent the deeper side, contrary to lighter ones. Interestingly, the calculated average roughness of the surfaces for nanofibers before and after PTA immobilization was at the same range. Moreover, the PTA molecules could not be clearly recognized on the fiber surface. Four composite membranes (CM) were assembled through hot pressing of PTA immobilized nanofiber with two outer layers of recast Nafion. Fig. 6 shows FESEM image depicting cross-sectional view of 3-layered membrane having 65 mm thickness in comparison with corresponding PTA immobilized nylon-6,6 sheet. It should be noticed that the spraying of Nafion solution on both sides of nanofiber is required before hotpressing to enhance the interlayer adhesion. Strong interactions between HPAs and sulfonic acid groups of ionomers [47,48] will lower the risk of delamination.
Table 1 summarizes the details of various assembled membranes as well as their conductivity and degree of anisotropy. As can be seen, proton conductivity of around 75e108 mS cm1 at 100% RH and 30 C was achieved which is comparable with that of N112 under the same condition. A small degree of anisotropy was observed for all membrane samples. Such conductance anisotropy behavior is due to the difference in the orientation of the ionic clusters in the recast Nafion layers and in PTA immobilized nanofibrous layer. Such anisotropic behavior was previously reported for the extruded Nafion membranes [49]. The conductivity of pristine N112, CM-II and CM-IV samples as a function of the RH at a constant temperature (80 C) is shown in Fig. 7. Each point was obtained after equilibrating the sample at given conditions until the conductivity became stable. Following the general behavior of proton conducting membranes, the proton conductivity increased with the increase in RH for all membranes. This increase is due to the enhancement of the proton mobility resulting from the rise in water content. As the humidity level decreased, the difference in the performance between the composite membranes and N112 became obvious. For instance, CM-II membranes displayed approximately one order of magnitude higher conductivity than N112 at 26% RH. Such remarkable improvement in the conductivity along with the high durability (no effective decline even after 100 h of continuous monitoring) were encouraging for evaluating the dimensional stability, water swelling and finally the single test performance of the membrane. The water uptake values of the membranes were monitored after 12, 24, and 48 h of immersion in DI to find the maximum values. The steady-state was reached almost around 24 h and it was therefore chosen as the immersion time for water swelling and uptake measurements. The water swelling values of the membranes are presented in Table 2. The data are average values of 3 measurements with the relative standard deviation (RSD; n ¼ 3) of below 5%. Compared to high water uptake of Nafion membrane (13%), sandwiched membrane showed considerably lower uptake of around 5%. Interestingly, the composite membranes also displayed a tendency for an isotropic swelling with slightly larger size change in the thickness direction. 3.3. Fuel cell performance Fig. 8 (a and b) shows the polarization and power curves for PEMFC single cells with two composite membranes in comparison with N112 under low humidity condition. As shown in Fig. 8a, both composite membranes reveled ~100 mV higher OCV than N112.
Fig. 4. Comparison of thermogravimetric analysis of pristine nanofiber with nanofibrous sheets immobilized with various level of PTA (a) and XRD patterns of the pristine nanofiber, PTA, and PTA immobilized fiber.
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
487
Fig. 5. Non-contact mode AFM topography (5 mm 5 mm) of pristine (a) and PTA immobilized (d) nanofibers. 3D rendering images and edge enhanced color for pristine nanofiber (b, c) and nanofiber with PTA (e, f) are also shown.
Fig. 6. FESEM images of the PTA anchored nanofiber (a and b), cross-section of sandwiched membrane in which the recast Nafion are the amorphous regions on both sides (z20 mm) and the PTA anchored nanofibrous is the central layer (z25 mm) (c), and zoom in the central layer (d).
Since the OCV is directly related to the fuel cross over, it could be concluded that the crossover of the reactant gases was remarkably suppressed by introducing PTA immobilized nanofiber. Similar observation was reported in literature for HPA containing membranes [33,50]. It is noteworthy that the cell OCV for the composite membrane with smaller nanofiber diameter (CM II) is slightly
higher. This could be due to the more uniform distribution of PTA and more pack structure of central layer. Interestingly, the OCV of composite membranes was considerably higher than thicker Nafions 117 (175 mm) operated under similar conditions. These observations provide evidence for remarkable improvement in the barrier properties of composite membranes.
488
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
Table 1 Physical characteristics, conductivity and area resistance of the membranes at 30 C. Sample
Nanofiber diameter (nm)a
Thickness (mm)c
PTA immobilization level (wt%)b
CM.I CM.II CM.III CM.IV N112 a b c d e
70 70 110 110 e
39.0 53.9 38.7 53.8 e
60 67 67 73 58
Proton conductivity (mS cm1)d
sk
s⊥
75.5 108.3 72.2 95.4 119.6
66.6 100.9 63.5 86.6 104.8
sk=
⊥
Area resistance (Ucm2)e
1.13 1.07 1.14 1.10 1.14
0.093 0.066 0.105 0.084 0.055
Based on the through-plane measurements. Indicated by thermal analysis and confirmed by weight change after immobilization. Hydrated membrane thickness were measured at 10 random points using a digital micrometer. In-plane (sk ) and through-plane (s⊥ ) proton conductivities at 100% RH. Calculated as R ðUÞ A ðcmÞ2 ; based on the through-plane measurements.
As indicated from Fig. 8 b, the CM-II shows a high current density value of 1500 mA cm2 at 0.6 V which is more than twice the values found with N112 membranes. When Nafion used as a membrane, the cell potential and resulting power density decrease rapidly at currents above 1200 mA cm2. Moreover, the CM-IV membrane showed a maximum power density >810 mW cm2 compared to 577 mW cm2 for N 112 membrane under the same conditions using same Pt loading. Among all membranes, one can observe that the best performance was obtained for MEAs with CMII, achieving a maximum power density of above 932 mW cm2 which is at least 40% higher than N112. This performance enhancement in the composite membrane can be attributed to the significantly increased proton conductivity and the reduction in the fuel gas crossover of composite membranes. Fig. 7. In-plane proton conductivity of membrane compared with Nafion 112 in various relative humidity at 80 C.
Table 2 Dimensional, chemical and mechanical stabilities of phosphonated membranes in comparison with N112. Sample
CM.II CM.IV N112 a b
Water uptake (%)
4.8 5.0 13.4
Dimensional stabilitya
Chemical stabilityb
Mechanical stability
Sk
S⊥
SV
RW1h (wt%)
RW5h (wt%)
RW24h (wt%)
Tensile strength (MPa)
Elongation at break (%)
2.8 3.1 9.5
3.1 3.3 12.4
8.7 9.8 34.7
97.4 97.8 97.4
90.8 90.4 91.5
89.2 88.4 90.2
16.5 ± 2 17.6 ± 1 20.2 ± 1.1
23.4 ± 4 24.3 ± 2 173.2 ± 3
In-plane (Sk ), through-plane (S⊥ ) and volume swelling (Sv) of membranes after soaking in water at 30 C for 24 h. Retained weight (RW) after 1, 5 and 24 h in Fenton’s reagent at 80 C.
Fig. 8. Polarization and power curves of the PEMFC single cells employing MEAs with Nafion 112 (50 mm) and composite membranes of CM-II (65 mm) and CM-IV (70 mm). The operation conditions are 40% RH and cell temperature of 60 C.
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
4. Conclusion In conclusion, new composite membranes were developed to address the challenges of overwhelming fuel crossover associated with the thin Nafion membranes. The new membrane was fabricated from the electrospun nylon-6,6 nanofibrous layer loaded with high level of PTA as a center layer that was covered by thin recast Nafion layer from both sides. The AFM, SEM, and TEM microscopy studies clearly revealed that the PTA immobilized uniformly. This highly proton conductive sheet was then assembled with 2 recast Nafion layers into membranes for PEMFC applications. No delamination was observed due to the strong interactions between PTA and sulfonic acid groups. Among the four new composite membranes evaluated, CM-II and CM-IV exhibited better proton conductivity under low RH. The membranes reveled better dimensional stability than N112, and good thermal, chemical and acceptable mechanical stabilities. The composite membranes further exhibited promising fuel cell performance at 60 C and 40% RH. In particular, high power densities of >930 mW cm2 and OCV greater than 1000 mV were obtained which are higher than those of N112 under the same conditions (577 mW cm2 and 904 mV). It can be concluded that the strategy of using PTA immobilized nanofibrous interlayer is highly promising in designing new effective membranes for applications of PEM fuel cells under low RH. More work is yet to be carried out to further explore the full potential of this class of composite membranes. Acknowledgements This research was funded by the Malaysian Ministry of Higher Education under the LRGS program (vote no. #4L817), Universiti Teknologi Malaysia (vote no. #00D20), and joint partnership research program of Innovative Research Universities of Australiaand Malaysia Research University Network (IRU-MRUN) (vote no. #00M84). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jpowsour.2016.07.027. References [1] W. Lehnert, C. Wannek, R. Zeis, Chapter 3 trends in high-temperature polymer electrolyte fuel cells, in: Innovations in Fuel Cell Technologies, The Royal Society of Chemistry, 2010, pp. 41e75. [2] M.M. Nasef, Chem. Rev. 114 (2014) 12278e12329. [3] L. Akbarian-Feizi, S. Mehdipour-Ataei, H. Yeganeh, Int. J. Hydrogen Energy 35 (2010) 9385e9397. [4] A. Chandan, M. Hattenberger, A. El-Kharouf, S. Du, A. Dhir, V. Self, B.G. Pollet, A. Ingram, W. Bujalski, J. Power Sources 231 (2013) 264e278. €z, Appl. Surf. Sci. 258 (2012) 3139e3146. [5] S. Yılmaztürk, N. Ercan, H. Deligo [6] K. Broka, P. Ekdunge, J. Appl. Electrochem 27 (1997) 117e124. [7] Q. Li, R. He, J.O. Jensen, N.J. Bjerrum, Chem. Mater. 15 (2003) 4896e4915. [8] K. Ketpang, S. Shanmugam, C. Suwanboon, N. Chanunpanich, D. Lee, J. Membr. Sci. 493 (2015) 285e298. [9] S. Rao, R. Xiu, J. Si, S. Lu, M. Yang, Y. Xiang, ChemSusChem 7 (2014) 822e828. [10] R. Wang, W. Zhang, G. He, P. Gao, J. Mater. Chem. A 2 (2014) 16416e16423. [11] E. Abouzari-Lotf, H. Ghassemi, A. Shockravi, T. Zawodzinski, D. Schiraldi, Polymer 52 (2011) 4709e4717.
489
[12] D. Henkensmeier, N.M.H. Duong, M. Brela, K. Dyduch, A. Michalak, K. Jankova, H. Cho, J.H. Jang, H.-J. Kim, L.N. Cleemann, Q. Li, J.O. Jensen, J. Mater. Chem. A 3 (2015) 14389e14400. [13] E. Abouzari-Lotf, H. Ghassemi, S. Mehdipour-Ataei, A. Shockravi, J. Membr. Sci. 516 (2016) 74e82. [14] K.T. Adjemian, R. Dominey, L. Krishnan, H. Ota, P. Majsztrik, T. Zhang, J. Mann, B. Kirby, L. Gatto, M. Velo-Simpson, J. Leahy, S. Srimvasan, J.B. Benziger, A.B. Bocarsly, Chem. Mater. 18 (2006) 2238e2248. [15] K. Ketpang, K. Lee, S. Shanmugam, ACS Appl. Mater. Interfaces 6 (2014) 16734e16744. [16] Y. Patil, S. Sambandam, V. Ramani, K. Mauritz, ACS Appl. Mater. Interfaces 5 (2013) 42e48. [17] R.P. Pandey, V.K. Shahi, J. Power Sources 299 (2015) 104e113. [18] M.M. Nasef, T. Fujigaya, E. Abouzari-Lotf, N. Nakashima, Z. Yang, Int. J. Hydrogen Energy 41 (2016) 6842e6854. [19] K. Li, B. Zhou, G. Ye, M. Pan, H. Zhang, J. Power Sources 298 (2015) 68e73. [20] Y. Kim, K. Ketpang, S. Jaritphun, J.S. Park, S. Shanmugam, J. Mater. Chem. A 3 (2015) 8148e8155. [21] A. Donnadio, M. Pica, D. Capitani, V. Bianchi, M. Casciola, J. Membr. Sci. 462 (2014) 42e49. [22] J.L. Lu, Q.H. Fang, S.L. Li, S.P. Jiang, J. Membr. Sci. 427 (2013) 101e107. rez Robles, J. Power [23] N. Guerrero Moreno, D. Gervasio, A. Godínez García, J.F. Pe Sources 300 (2015) 229e237. [24] P. Knauth, M.L. Di Vona, Solid State Proton Conductors: Properties and Applications in Fuel Cells, John Wiley and Sons, 2012. [25] M.M. Hasani-Sadrabadi, E. Dashtimoghadam, F.S. Majedi, J.J. VanDersarl, A. Bertsch, P. Renaud, K.I. Jacob, Nano Energy. [26] M. Yamada, I. Honma, J. Phys. Chem. B 110 (2006) 20486e20490. [27] J. Zeng, S.P. Jiang, J. Phys. Chem. C 115 (2011) 11854e11863. [28] F. Cavani, Catal. Today 41 (1998) 73e86. [29] Y. Zhou, J. Yang, H. Su, J. Zeng, S.P. Jiang, W.A. Goddard, J. Am. Chem. Soc. 136 (2014) 4954e4964. [30] T. Haolin, L. Junrui, W. Zhao, Z. Huijie, P. Mu, J. San Ping, Self-assembly of nanostructured proton exchange membranes for fuel cells, in: Nanotechnology for Sustainable Energy, American Chemical Society, 2013, pp. 243e263. [31] A.M. Herring, H.M. Gregory, M. Fanqin, N.V. Aieta, J.L. Horan, M.H. Frey, S.J. Hamrock, M.C. Kuo, ECS Trans. 3 (2006) 551e559. [32] G.M. Haugen, F. Meng, N.V. Aieta, J.L. Horan, M.-C. Kuo, M.H. Frey, S.J. Hamrock, A.M. Herring, Electrochem. Solid-State Lett. 10 (2007) B51eB55. [33] M.P. Rodgers, L.J. Bonville, H.R. Kunz, D.K. Slattery, J.M. Fenton, Chem. Rev. 112 (2012) 6075e6103. [34] S. Malhotra, R. Datta, J. Electrochem. Soc. 144 (1997) L23eL26. [35] Z.-G. Shao, H. Xu, M. Li, I.M. Hsing, Solid State Ionics 177 (2006) 779e785. [36] K.H. Lee, J.Y. Chu, A.R. Kim, K.S. Nahm, C.-J. Kim, D.J. Yoo, J. Membr. Sci. 434 (2013) 35e43. [37] H. Wu, X. Shen, Y. Cao, Z. Li, Z. Jiang, J. Membr. Sci. 451 (2014) 74e84. [38] M. Yang, S. Lu, J. Lu, S.P. Jiang, Y. Xiang, Chem. Commun. 46 (2010) 1434e1436. [39] Q. Wu, H. Wang, S. Lu, X. Xu, D. Liang, Y. Xiang, J. Membr. Sci. 500 (2016) 203e210. [40] A.B. Bose, S. Gopu, W. Li, J. Power Sources 263 (2014) 217e222. [41] A.M. Herring, J. Macromol. Sci. C 46 (2006) 245e296. [42] J.L. Horan, A. Genupur, H. Ren, B.J. Sikora, M.-C. Kuo, F. Meng, S.F. Dec, G.M. Haugen, M.A. Yandrasits, S.J. Hamrock, M.H. Frey, A.M. Herring, ChemSusChem 2 (2009) 226e229. [43] J.L. Horan, A. Lingutla, H. Ren, M.-C. Kuo, S. Sachdeva, Y. Yang, S. Seifert, L.F. Greenlee, M.A. Yandrasits, S.J. Hamrock, M.H. Frey, A.M. Herring, J. Phys. Chem. C 118 (2013) 135e144. [44] E. Abouzari-Lotf, M.M. Nasef, H. Ghassemi, M. Zakeri, A. Ahmad, Y. Abdollahi, ACS Appl. Mater. Interfaces 7 (2015) 17008e17015. [45] M. Tanaka, Polym. J. 48 (2016) 51e58. [46] N. Feng, A. Zheng, S.-J. Huang, H. Zhang, N. Yu, C.-Y. Yang, S.-B. Liu, F. Deng, J. Phys. Chem. C 114 (2010) 15464e15472. [47] Y.S. Kim, F. Wang, M. Hickner, T.A. Zawodzinski, J.E. McGrath, J. Membr. Sci. 212 (2003) 263e282. [48] F. Meng, N.V. Aieta, S.F. Dec, J.L. Horan, D. Williamson, M.H. Frey, P. Pham, J.A. Turner, M.A. Yandrasits, S.J. Hamrock, A.M. Herring, Electrochimica Acta 53 (2007) 1372e1378. [49] T. Soboleva, Z. Xie, Z. Shi, E. Tsang, T. Navessin, S. Holdcroft, J. Electroanal. Chem. 622 (2008) 145e152. [50] R.J. Stanis, M.-C. Kuo, A.J. Rickett, J.A. Turner, A.M. Herring, Electrochimica Acta 53 (2008) 8277e8286.
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Enhancement of fuel cell performance with less-water dependent composite membranes having polyoxometalate anchored nanofibrous interlayer Ebrahim Abouzari-lotf a, *, Mohan V. Jacob b, Hossein Ghassemi c, Arshad Ahmad a, Mohamed Mahmoud Nasef a, d, Masoumeh Zakeri a, Shahram Mehdipour-Ataei e a
Advanced Materials Research Group, Center of Hydrogen Energy, Institute of Future Energy, Universiti Teknologi Malaysia, 54100, Kuala Lumpur, Malaysia Electronics Materials Lab., College of Science, Technology and Engineering, James Cook University, Townsville, Queensland 4811, Australia Department of Macromolecular Science & Engineering, Case Western Reserve University, 44106-7202 Cleveland, OH, USA d MalaysiaeJapan International Institute of Technology, Universiti Teknologi Malaysia, 54100 Kuala Lumpur, Malaysia e Iran Polymer and Petrochemical Institute, P.O. Box 14965/115, Tehran, Iran b c
h i g h l i g h t s Composite membranes consisting ion conductive nanofibers are proposed for PEMFCs. Dimensional stability improved and water uptake reduced compared to Nafion 112. The proton conductivity at low humidity is improved in composite membranes. OCV enhanced by about 100 mV than Nafion 112 when tested at 60 C and 40% RH. Maximum power density increased by 35% in the composite.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 March 2016 Received in revised form 4 July 2016 Accepted 7 July 2016 Available online 15 July 2016
Polyoxometalate immobilized nanofiber was used to fabricate low gas permeable layer for composite membranes designed for proton exchange membrane fuel cell (PEMFC) operating at low relative humidity (RH). The composite membranes revealed enhanced proton conductivity in dry conditions compared with state-of-the-art pristine membrane (Nafion 112, N112). This was coupled with a low fuel crossover inheriting the composite membranes about 100 mV higher OCV than N112 when tested in PEMFC at 60 C and 40% RH. A maximum power density of up to 930 mW cm2 was also achieved which is substantially higher than the N112 under similar conditions (577 mW cm2). Such remarkable performance enhancement along with undetectable leaching of immobilized polyoxometalate, high dimensional stability and low water uptake of the composite membranes suggest a strong potential for PEMFC under low RH operation. © 2016 Elsevier B.V. All rights reserved.
Keywords: Polymer electrolyte membrane fuel cells Conductive nanofibrous membrane Polyoxometalate Membrane electrode assembly High energy density
1. Introduction Proton Exchange Membrane fuel cell (PEMFC) is one of the key emerging technologies that is attracting remarkable effort with the aim to provide alternative environmentally friendly and efficient power sources. The advances in proton exchange membranes (PEMs) is critical for improving the performance of PEMFC [1e4].
* Corresponding author. E-mail address: [email protected] (E. Abouzari-lotf). http://dx.doi.org/10.1016/j.jpowsour.2016.07.027 0378-7753/© 2016 Elsevier B.V. All rights reserved.
Membranes comprising perfluorosulfonic acid polymers such as Nafion have been used extensively due to their high conductivity and stability. However, these materials need to be saturated with water to obtain practical levels of proton conductivity. Such hydration undermines the permeability of H2 and O2 due to the involvement of the formed hydrated ionic clusters in the gas permeation mechanism [5,6]. On the other hand, the retention and management of water within these types of membranes is challenging, costly and in most cases reduce the PEMFC performance. Therefore there is a strong demand for the PEMs to work at lower relative humidity and/or under anhydrous conditions.
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
Accelerated electrochemical reactions and simpler water management of fuel cells under lower RH and/or anhydrous conditions is challenged by dramatic losses of proton conductivity of perfluorinated sulfonic acid based membranes [2,7,8]. Typically, N112 losses one order of magnitude of its proton conductivity at around 60% RH which, increases two orders of magnitude with the reduction of RH to 25%. So, it was concluded that the current Nafion based membranes cannot meet the requirements for the practical applications in these conditions. Various designing strategies and advanced materials have been developed to improve the performance of PEMs operating under low RH [9]. One of the main strategies is based on reduction of the membrane thickness in a way leading to thinner membranes with lower internal resistance and improved water management during operation at lower RH and or elevated temperatures [7]. This is the main reason of progressive reduction in the thickness of commercial Nafion membranes for PEMFC over the past 2 decades (e.g. from ~183 mm in N117 to ~125 mm in N115 and then to 25 mm in N211) [10]. However, the thinner membranes are challenged by high fuel crossover and weak mechanical properties. On the other hand, low open-circuit voltage (OCV) and short PEM lifetime are the major concerns that have arisen on the practical application of 25 mmthick Nafion membranes in hydrogen fuel cells. Other strategies of designing less water dependent polymer electrolytes [11e13], modification of available electrolytes with water retention fillers [8,14e18], and introducing anhydrous proton conductive additives are also widely discussed to improve membrane limitations under lower RHs [19e23]._ENREF_15 Particularly, the incorporation of inorganic proton conductors such as heteropolyacids was found to increase the conductivity under dry and humidified conditions [20,24e27]. Heteropolyacids (such as phosphotungstic acid H3PW12O40, PTA) have a very strong Brønsted acidity, higher than both pure sulfuric acid and Nafion and approaching the superacid region. This high acidic strength is accomplished with fast reversible redox transformations under quite mild conditions [28]. In addition, this hydrophilic inorganic metal oxides have a high tendency to strongly accommodate water in their interlayer regions [29,30]. On the other hand, heteropolyacids could improve the stability of membranes by inhibiting the formation of peroxide/or catalyzing its decomposition [31e33]. The latter merit was confirmed by intense reduction in the release rate of F originated from fluorinated backbone degradation during PEMFC operating under relatively hot and dry conditions. However, although introducing of PTA to the membranes led to improved properties, yet PTA loading level and its direct contact with water are the serious concerns which need to be addressed. Low amount of PTA could be introduced into the membrane either through impregnation of porous substrate with an aqueous solution of PTA [22,34], or blending of PTA or its derivatives with a polymer followed by casting [35,36]. The composite membranes with PTA showed improved methanol barrier properties and/or water retention in DMFC [37e39], and enhanced PEMFC performance at low humidity and/or elevated temperatures [22,36,40]. However, due to the low level of PTA, loose interactions with polymer substrate and agglomeration, resulting composite membranes show low proton conductivity and PTA leaching during the fuel cell operation [41]. Copolymerization of PTA containing monomers was proposed to increase the loading level and minimize the leaching of PTA [42,43]. Such monomers are commonly prepared in the multistep synthetic procedure and therefore this method was not used widely. Recently, we proposed a simple and flexible method to introduce high level of PTA onto membranes [44]. Various amounts up to 50 wt% of PTA could be simply immobilized onto nylon nanofibers by self-immobilization. Such ion conductive nanofibers are capable
483
of fast proton transport [45]. On the other hand, leaching of PTA was addressed through assembling the membrane in a layered structure with two recast Nafion layers. The membranes showed an improved performance in DMFC and the conductivity remained quite stable even after more than 100 h of continuous testing. In this work, a modified method was used to immobilize high level of PTA on nanofibrous structure and sandwiched membranes were fabricated for low humidity PEMFC application. The effects of PTA loading level and nanofiber diameter on the water uptake, dimensional stability and proton conductivity were investigated in comparison with commercially available N112. Moreover, the effect of introducing PTA on open-circuit voltage (OCV) as well as singlecell performance under lower RH were also evaluated. 2. Experimental 2.1. Materials Technical grade nylon-6,6 (medium viscous, DSM Co., The Netherlands) was used to prepare the nanofibrous sheets using Electrospinner (Nanolab Instruments) electrospinning system. Thin Nafion layers with a thickness of 20 mm were prepared by casting a 15% concentrated aqueous solution of Nafion (5%, supplied from DUPONT). FeSO4.7H2O (98.1%, from Fischer Scientific Company, Fairlawn, NJ) was used for the preparation of Fenton agent for chemical stability test. Commercial catalyst of 40 wt % Pt/C (Johnson Matthey) was used in both the cathode and the anode. 2.2. Characterizations Transmission electron microscope (TEM) images were obtained with a TECNAI G2 F20 with an accelerating voltage of 200 kV. A piece of the sample was milled at liquid nitrogen using Ika Ultra Turrax tube disperser. The dark areas in the HRTEM images clearly show hydrophilic domains and the light areas represent hydrophobic domains. The atomic force microscope (AFM) images were obtained with a Park NX10 in a non-contact mode. Scanning Electron Microscope (SEM) images of nanofibrous sheets and composite membranes were taken on the Philips XL30 Field Emission Scanning Electron Microscope (FESEM) after coating with 5 nm Au. The cross-sectional morphology was examined using cryofracturing of samples in liquid nitrogen. Fourier transform-infrared attenuated total reflection (FT-IR-ATR) spectra were performed on Agilent Cary 660 spectrometer. X-ray diffraction pattern was obtained on the Philips X’Pert 1 X-ray diffractometer with graphite monochromatized Cu Ka radiation (l ¼ 1.5401 Å) at scanning rate of 2 min1 over a range of 2q ¼ 4e80 . Thermogravimetric analysis (TGA) was performed on the Perkin Elmer TGA7 under nitrogen atmosphere at a heating rate of 10 C min1. Leaching of the PTA from the membrane was measured by immersing the composite membranes in deionized water at room temperature and monitoring the concentration of PTA in water every 12 h using a UV-1800 UVevis spectrometer. The in-plane (sk ) and through-plane (s⊥ ) proton conductivities of the membranes were measured by using a four-point probe of Bekk Tech conductivity cells (BT-112) and two-point probe of homemade stainless steel cylindrical electrodes with diameter of 20 mm. The resistance of the membranes was measured by using a DC conductivity testing (Keithley 2400 sourcemeter) controlled by a Labview software. The potentiostat was set to apply a specific voltage between the two inner probes and measure resulting current. The slope of the data from current versus voltage measurement was used to calculate the resistance (R) and proton conductivity (s) was calculated according to the equation (1):
484
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
LðcmÞ
s S=cm ¼ RðUÞ AðcmÞ2 sk= ¼ ⊥
sk s⊥
(1)
(2)
where L is the distance in the direction of ion flow between voltage measurement probes and A is the area of the membrane. Since the membrane swell during fuel cell operation, it is important to employ the hydrated thickness to calculate the conductivity and the resistivity of the membranes. Anisotropic proton conductivity ratio was calculated from in-plane and through-plane conductivity values according to equation (2). In order to control the RH in a desired range, various saturated salt solutions were used and measurements were performed in a homemade chamber using BT112 cell. A chamber brought to equilibrium at a controlled temperature and humidity controlled by SHT75 sensor (with ±1.8% RH accuracy). Generally, the measured proton conductivity data have errors of ±8%. Membrane water uptake, defined as the difference in mass between fully hydrated and completely dry membrane, 4w, was calculated using the following equation:
4w ¼
Wwet Wdry 100 Wdry
(3)
where Wwet is the weight of the membrane after equilibration in water for 24 h and Wdry is the weight after complete drying in a vacuum oven at 80 C for 24 h. In a similar manner, the in-plane (Sk ), through-plane (S⊥ ), and volume swelling (SV ) of the membranes in water were calculated using the following equation:
Sx ¼
Xwet Xdry 100 Xdry
(4)
where x represents in-plane, through plane, or volume swelling measurements and Xwet and Xdry are the length, thickness, or volume of swollen and dry membranes, respectively. In order to obtain an accurate hydrated mass and swelling data, after removing the hydrated membranes from water, the sample was blotted quickly to remove surface water and measured immediately (the measurement was repeated 3 times). Oxidative stability of the membranes was evaluated by the Fenton agent consisting of 3% H2O2 as a source of peroxide and hydroperoxide radicals and 2 ppm FeSO4. Pre-weighed dry membranes were soaked in a Fenton aqueous solution at 80 C for up to 96 h and the stability was evaluated by recording the retained weight of the membranes after complete drying. 2.3. Preparation of nanofibers and composite membranes A polymer solution of 15e18 wt% concentrated nylon-6,6 was prepared in a mixture of 9/1 (v/v) of formic acid and water in a glass reactor. The polymer solution was electrospun under an applied voltage of 20 kV, a distance of 12 cm between the tip of the metal needle and the aluminum foil collector, a solution flow rate was 0.3e0.4 mL h1 and the collector drum rotation speed of 300 rpm. After 14 h of electrospinning, the fibrous sheet was collected and dried in a vacuum oven to completely remove the solvent. With adjusting the electrospinning parameters, two types of beads free nanofibrous sheets with mean diameters of 70 and 110 nm (based on the SEM micrographs) were obtained under two combinations of electrospinning parameters. PTA self-immobilization was performed by immersing the
nanofibrous sheet in an aqueous solution of PTA (5e12 wt%) at room temperature for periods up to 4 days under slow shaking. Successively the nanofibrous sheet was removed, washed with plenty of deionized water to remove unbounded PTA and sandwiched between two glass plates prior to vacuum drying for 24 h. Two immobilization levels of 39.0 and 53.9 wt% were obtained. A layered membrane composed of the PTA-immobilized nanofibers sandwiched between thin Nafion layers was prepared by mechanical compression between two ETFE films at 130 C. To enhance the adhesion between layers, Nafion solution was sprayed on the both sides of the PTA-immobilized nanofiber sheet before assembling. Hot-press with a pressure of 1700 psi was applied for 10 s for four times (samples were rotated 90 three times to ensure a uniform compression) followed by heat treatment at 135 C for 5 h in a vacuum oven. Finally, the membranes were soaked in 1 N sulfuric acid for 24 h, rinsed with deionized water and dried in a vacuum oven. 2.4. Preparation of membrane electrode assembly (MEA) and single cell operation The catalyst ink was prepared by adding electrocatalyst and 5 wt % of Nafion ionomer solution into a mixture of water and isopropanol (1:1) followed by ultrasonic treatment for 1 h. For MEA fabrication, composite membranes were pretreated by 1 h boiling consecutively in, 3 wt% H2O2, deionized water, 1 M H2SO4 and finally Milli-Q water. The catalyst-coated membrane was prepared by spraying the catalyst onto a membrane with a controlled loading of around 4 mg cm2 for both anode and cathode sides. The obtained MEA was dried for 4 h under atmospheric conditions and heat-treated in a vacuum oven at 60 C to remove residual solvents. Finally, it was sandwiched between two pieces of gas-diffusion layers (Toray TGPH-030 Carbon paper) and assembled in a single cell hardware with an active area of 5 cm2. For a comparison purpose, MEA with the same catalyst loading was prepared with commercial N112 membrane. The cell was operated at 60 C and the fluxes of H2 and O2 were controlled at 10 and 18 mL min1, respectively. The tests were started with the humidified gases with the humidifier temperatures of 70 C for both gases. After reaching a stable performance, to evaluate the effect of introducing PTAimmobilized nanofibrous sheet on the membrane performance at low humidity, the single cell was kept running with a humidity of around 40% for both gases. The cell voltages at different current densities were recorded after reaching stable cell performance. 3. Results and discussions 3.1. Membrane fabrication and characterization The concentration and resulting viscosity of the polymer solution played an important role in the morphology of electrospun nanofibers. For instance, at concentrations lower than 15 wt% nylon-6,6, many beads were formed due to the low viscosity of polymer solution. However, further increase in the concentration to 18 wt%, the spinnability was improved and beads free nanofibrous sheets were obtained. Moreover, the increase in the concentration of polymer solution resulted in an increase in the diameter of the electrospun nanofibers from 60e90 nm to 85e145 nm. Typical SEM micrographs of electrospun nanofibers is shown in the Fig. S1. In order to immobilize phosphotungstic acid on the nanofibrous sheet, the nylon mat was immersed in aqueous solution of PTA [44]. FTIR-ATR spectra of pristine nylon-6,6 nanofibers and corresponding PTA immobilized nanofibers are shown in Fig. 1. The spectrum clearly confirms the immobilization. There are four kinds of oxygen atoms in the Keggin-type PTA including the central
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
485
Fig. 1. Comparison of FTIR-ATR spectra of nanofibers (blue) and PTA immobilized nanofibers (red) in the expanded range of 500e1800 cm1. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
oxygen atoms (PeOa), two types of bridging oxygen atoms (WeObeW and WeOceW), and terminal oxygen atoms (W ¼ Od). The IR spectrum shows characteristic bands for each kinds at 1080 for PeOa, 983 for W ¼ Od, 893 for WeObeW, and 797 cm1 for WeOceW (Fig. S2). Upon anchoring onto nanofibers, as shown in Fig. 1, the frequency of the peak correspond to the WeOceW changes significantly (from 797 to 819 cm1). Smaller shift to the lower frequency was also observed for the WeObeW peak (from 893 to 896 cm1). The representing peaks for terminal and central oxygen atoms almost remain constant. So, it could be concluded that PTA is immobilized due to the preferential interactions reside at bridge-sharing (Oc) oxygen sites rather than the terminal (Od) site. Similarly, DFT calculations with 31P solid-state magic-anglespinning NMR confirm that these oxygen sites are preferred location of the Brønsted protons [46]. As shown, a minor shift in the carbonyl frequency of the amide group (from 1636 to 1633 cm1) was also observed which could be attributed to the formation of hydrogen bonding with PTA.
The PTA immobilization on nylon-6,6 nanofibers was further confirmed by high-resolution TEM imaging as shown in Fig. 2. As can be seen in high-resolution TEM images, there is no evidence for crystalline or agglomerated PTA molecules which reflecting a uniform immobilization. Accordingly, it can be suggested that the immobilization took place by chemical interaction which was further confirmed by disappearance of the PTA peaks and appearing of new peaks in X-ray diffraction (XRD) pattern (shown in Fig. 4b). Achieving a uniform immobilization of PTA in nylone-6,6 nanofibrous sheets depends on a number of parameters such as initial concentration of PTA in immobilization bath, immersion time and shaking speed. Particularly, high PTA concentration and prolonged immersion time provoke inhomogeneous distribution during immobilization. A typical example for non-uniform PTA distribution (57 wt%) onto the nanofibrous support prepared by immersion in highly concentrated PTA solution (12 wt%) for 10 days is shown in Fig. 3. Obviously, the PTA immobilized nanofibrous
Fig. 2. TEM image of PTA immobilized nanofiber, showing fine PTA immobilization onto nanofibrous matrix. The regions with different hydrophobicity could be seen clearly in c.
486
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
3.2. Membrane stability and proton conductivity
Fig. 3. SEM images of non-uniformly impregnated PTA on nylon-6,6 nanofibers.
display not only fragility and non-homogeneity but also leaching of the loaded PTA as indicated by the loss of 20 wt% of PTA upon immersion in water for 24 h. Immobilization level could be simply tuned by controlling the concentration of immersion bath and immersion time. Herein, by manipulating the immobilization parameters two different immobilization levels of 53.9 and 39.0% were obtained. The immobilization level was measured with gravimetric and confirmed by thermogravimetric analysis. Typical TGA thermograms of PTA immobilized nylon-6,6 nanofibers of different loading is shown in Fig. 4a. Uniform immobilization of PTA on the nanofiber surface was further confirmed in AFM images. Fig. 5 reveals the surface features of pristine and PTA immobilized nanofibers. In the vertical direction, darker areas represent the deeper side, contrary to lighter ones. Interestingly, the calculated average roughness of the surfaces for nanofibers before and after PTA immobilization was at the same range. Moreover, the PTA molecules could not be clearly recognized on the fiber surface. Four composite membranes (CM) were assembled through hot pressing of PTA immobilized nanofiber with two outer layers of recast Nafion. Fig. 6 shows FESEM image depicting cross-sectional view of 3-layered membrane having 65 mm thickness in comparison with corresponding PTA immobilized nylon-6,6 sheet. It should be noticed that the spraying of Nafion solution on both sides of nanofiber is required before hotpressing to enhance the interlayer adhesion. Strong interactions between HPAs and sulfonic acid groups of ionomers [47,48] will lower the risk of delamination.
Table 1 summarizes the details of various assembled membranes as well as their conductivity and degree of anisotropy. As can be seen, proton conductivity of around 75e108 mS cm1 at 100% RH and 30 C was achieved which is comparable with that of N112 under the same condition. A small degree of anisotropy was observed for all membrane samples. Such conductance anisotropy behavior is due to the difference in the orientation of the ionic clusters in the recast Nafion layers and in PTA immobilized nanofibrous layer. Such anisotropic behavior was previously reported for the extruded Nafion membranes [49]. The conductivity of pristine N112, CM-II and CM-IV samples as a function of the RH at a constant temperature (80 C) is shown in Fig. 7. Each point was obtained after equilibrating the sample at given conditions until the conductivity became stable. Following the general behavior of proton conducting membranes, the proton conductivity increased with the increase in RH for all membranes. This increase is due to the enhancement of the proton mobility resulting from the rise in water content. As the humidity level decreased, the difference in the performance between the composite membranes and N112 became obvious. For instance, CM-II membranes displayed approximately one order of magnitude higher conductivity than N112 at 26% RH. Such remarkable improvement in the conductivity along with the high durability (no effective decline even after 100 h of continuous monitoring) were encouraging for evaluating the dimensional stability, water swelling and finally the single test performance of the membrane. The water uptake values of the membranes were monitored after 12, 24, and 48 h of immersion in DI to find the maximum values. The steady-state was reached almost around 24 h and it was therefore chosen as the immersion time for water swelling and uptake measurements. The water swelling values of the membranes are presented in Table 2. The data are average values of 3 measurements with the relative standard deviation (RSD; n ¼ 3) of below 5%. Compared to high water uptake of Nafion membrane (13%), sandwiched membrane showed considerably lower uptake of around 5%. Interestingly, the composite membranes also displayed a tendency for an isotropic swelling with slightly larger size change in the thickness direction. 3.3. Fuel cell performance Fig. 8 (a and b) shows the polarization and power curves for PEMFC single cells with two composite membranes in comparison with N112 under low humidity condition. As shown in Fig. 8a, both composite membranes reveled ~100 mV higher OCV than N112.
Fig. 4. Comparison of thermogravimetric analysis of pristine nanofiber with nanofibrous sheets immobilized with various level of PTA (a) and XRD patterns of the pristine nanofiber, PTA, and PTA immobilized fiber.
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
487
Fig. 5. Non-contact mode AFM topography (5 mm 5 mm) of pristine (a) and PTA immobilized (d) nanofibers. 3D rendering images and edge enhanced color for pristine nanofiber (b, c) and nanofiber with PTA (e, f) are also shown.
Fig. 6. FESEM images of the PTA anchored nanofiber (a and b), cross-section of sandwiched membrane in which the recast Nafion are the amorphous regions on both sides (z20 mm) and the PTA anchored nanofibrous is the central layer (z25 mm) (c), and zoom in the central layer (d).
Since the OCV is directly related to the fuel cross over, it could be concluded that the crossover of the reactant gases was remarkably suppressed by introducing PTA immobilized nanofiber. Similar observation was reported in literature for HPA containing membranes [33,50]. It is noteworthy that the cell OCV for the composite membrane with smaller nanofiber diameter (CM II) is slightly
higher. This could be due to the more uniform distribution of PTA and more pack structure of central layer. Interestingly, the OCV of composite membranes was considerably higher than thicker Nafions 117 (175 mm) operated under similar conditions. These observations provide evidence for remarkable improvement in the barrier properties of composite membranes.
488
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
Table 1 Physical characteristics, conductivity and area resistance of the membranes at 30 C. Sample
Nanofiber diameter (nm)a
Thickness (mm)c
PTA immobilization level (wt%)b
CM.I CM.II CM.III CM.IV N112 a b c d e
70 70 110 110 e
39.0 53.9 38.7 53.8 e
60 67 67 73 58
Proton conductivity (mS cm1)d
sk
s⊥
75.5 108.3 72.2 95.4 119.6
66.6 100.9 63.5 86.6 104.8
sk=
⊥
Area resistance (Ucm2)e
1.13 1.07 1.14 1.10 1.14
0.093 0.066 0.105 0.084 0.055
Based on the through-plane measurements. Indicated by thermal analysis and confirmed by weight change after immobilization. Hydrated membrane thickness were measured at 10 random points using a digital micrometer. In-plane (sk ) and through-plane (s⊥ ) proton conductivities at 100% RH. Calculated as R ðUÞ A ðcmÞ2 ; based on the through-plane measurements.
As indicated from Fig. 8 b, the CM-II shows a high current density value of 1500 mA cm2 at 0.6 V which is more than twice the values found with N112 membranes. When Nafion used as a membrane, the cell potential and resulting power density decrease rapidly at currents above 1200 mA cm2. Moreover, the CM-IV membrane showed a maximum power density >810 mW cm2 compared to 577 mW cm2 for N 112 membrane under the same conditions using same Pt loading. Among all membranes, one can observe that the best performance was obtained for MEAs with CMII, achieving a maximum power density of above 932 mW cm2 which is at least 40% higher than N112. This performance enhancement in the composite membrane can be attributed to the significantly increased proton conductivity and the reduction in the fuel gas crossover of composite membranes. Fig. 7. In-plane proton conductivity of membrane compared with Nafion 112 in various relative humidity at 80 C.
Table 2 Dimensional, chemical and mechanical stabilities of phosphonated membranes in comparison with N112. Sample
CM.II CM.IV N112 a b
Water uptake (%)
4.8 5.0 13.4
Dimensional stabilitya
Chemical stabilityb
Mechanical stability
Sk
S⊥
SV
RW1h (wt%)
RW5h (wt%)
RW24h (wt%)
Tensile strength (MPa)
Elongation at break (%)
2.8 3.1 9.5
3.1 3.3 12.4
8.7 9.8 34.7
97.4 97.8 97.4
90.8 90.4 91.5
89.2 88.4 90.2
16.5 ± 2 17.6 ± 1 20.2 ± 1.1
23.4 ± 4 24.3 ± 2 173.2 ± 3
In-plane (Sk ), through-plane (S⊥ ) and volume swelling (Sv) of membranes after soaking in water at 30 C for 24 h. Retained weight (RW) after 1, 5 and 24 h in Fenton’s reagent at 80 C.
Fig. 8. Polarization and power curves of the PEMFC single cells employing MEAs with Nafion 112 (50 mm) and composite membranes of CM-II (65 mm) and CM-IV (70 mm). The operation conditions are 40% RH and cell temperature of 60 C.
E. Abouzari-lotf et al. / Journal of Power Sources 326 (2016) 482e489
4. Conclusion In conclusion, new composite membranes were developed to address the challenges of overwhelming fuel crossover associated with the thin Nafion membranes. The new membrane was fabricated from the electrospun nylon-6,6 nanofibrous layer loaded with high level of PTA as a center layer that was covered by thin recast Nafion layer from both sides. The AFM, SEM, and TEM microscopy studies clearly revealed that the PTA immobilized uniformly. This highly proton conductive sheet was then assembled with 2 recast Nafion layers into membranes for PEMFC applications. No delamination was observed due to the strong interactions between PTA and sulfonic acid groups. Among the four new composite membranes evaluated, CM-II and CM-IV exhibited better proton conductivity under low RH. The membranes reveled better dimensional stability than N112, and good thermal, chemical and acceptable mechanical stabilities. The composite membranes further exhibited promising fuel cell performance at 60 C and 40% RH. In particular, high power densities of >930 mW cm2 and OCV greater than 1000 mV were obtained which are higher than those of N112 under the same conditions (577 mW cm2 and 904 mV). It can be concluded that the strategy of using PTA immobilized nanofibrous interlayer is highly promising in designing new effective membranes for applications of PEM fuel cells under low RH. More work is yet to be carried out to further explore the full potential of this class of composite membranes. Acknowledgements This research was funded by the Malaysian Ministry of Higher Education under the LRGS program (vote no. #4L817), Universiti Teknologi Malaysia (vote no. #00D20), and joint partnership research program of Innovative Research Universities of Australiaand Malaysia Research University Network (IRU-MRUN) (vote no. #00M84). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jpowsour.2016.07.027. References [1] W. Lehnert, C. Wannek, R. Zeis, Chapter 3 trends in high-temperature polymer electrolyte fuel cells, in: Innovations in Fuel Cell Technologies, The Royal Society of Chemistry, 2010, pp. 41e75. [2] M.M. Nasef, Chem. Rev. 114 (2014) 12278e12329. [3] L. Akbarian-Feizi, S. Mehdipour-Ataei, H. Yeganeh, Int. J. Hydrogen Energy 35 (2010) 9385e9397. [4] A. Chandan, M. Hattenberger, A. El-Kharouf, S. Du, A. Dhir, V. Self, B.G. Pollet, A. Ingram, W. Bujalski, J. Power Sources 231 (2013) 264e278. €z, Appl. Surf. Sci. 258 (2012) 3139e3146. [5] S. Yılmaztürk, N. Ercan, H. Deligo [6] K. Broka, P. Ekdunge, J. Appl. Electrochem 27 (1997) 117e124. [7] Q. Li, R. He, J.O. Jensen, N.J. Bjerrum, Chem. Mater. 15 (2003) 4896e4915. [8] K. Ketpang, S. Shanmugam, C. Suwanboon, N. Chanunpanich, D. Lee, J. Membr. Sci. 493 (2015) 285e298. [9] S. Rao, R. Xiu, J. Si, S. Lu, M. Yang, Y. Xiang, ChemSusChem 7 (2014) 822e828. [10] R. Wang, W. Zhang, G. He, P. Gao, J. Mater. Chem. A 2 (2014) 16416e16423. [11] E. Abouzari-Lotf, H. Ghassemi, A. Shockravi, T. Zawodzinski, D. Schiraldi, Polymer 52 (2011) 4709e4717.
489
[12] D. Henkensmeier, N.M.H. Duong, M. Brela, K. Dyduch, A. Michalak, K. Jankova, H. Cho, J.H. Jang, H.-J. Kim, L.N. Cleemann, Q. Li, J.O. Jensen, J. Mater. Chem. A 3 (2015) 14389e14400. [13] E. Abouzari-Lotf, H. Ghassemi, S. Mehdipour-Ataei, A. Shockravi, J. Membr. Sci. 516 (2016) 74e82. [14] K.T. Adjemian, R. Dominey, L. Krishnan, H. Ota, P. Majsztrik, T. Zhang, J. Mann, B. Kirby, L. Gatto, M. Velo-Simpson, J. Leahy, S. Srimvasan, J.B. Benziger, A.B. Bocarsly, Chem. Mater. 18 (2006) 2238e2248. [15] K. Ketpang, K. Lee, S. Shanmugam, ACS Appl. Mater. Interfaces 6 (2014) 16734e16744. [16] Y. Patil, S. Sambandam, V. Ramani, K. Mauritz, ACS Appl. Mater. Interfaces 5 (2013) 42e48. [17] R.P. Pandey, V.K. Shahi, J. Power Sources 299 (2015) 104e113. [18] M.M. Nasef, T. Fujigaya, E. Abouzari-Lotf, N. Nakashima, Z. Yang, Int. J. Hydrogen Energy 41 (2016) 6842e6854. [19] K. Li, B. Zhou, G. Ye, M. Pan, H. Zhang, J. Power Sources 298 (2015) 68e73. [20] Y. Kim, K. Ketpang, S. Jaritphun, J.S. Park, S. Shanmugam, J. Mater. Chem. A 3 (2015) 8148e8155. [21] A. Donnadio, M. Pica, D. Capitani, V. Bianchi, M. Casciola, J. Membr. Sci. 462 (2014) 42e49. [22] J.L. Lu, Q.H. Fang, S.L. Li, S.P. Jiang, J. Membr. Sci. 427 (2013) 101e107. rez Robles, J. Power [23] N. Guerrero Moreno, D. Gervasio, A. Godínez García, J.F. Pe Sources 300 (2015) 229e237. [24] P. Knauth, M.L. Di Vona, Solid State Proton Conductors: Properties and Applications in Fuel Cells, John Wiley and Sons, 2012. [25] M.M. Hasani-Sadrabadi, E. Dashtimoghadam, F.S. Majedi, J.J. VanDersarl, A. Bertsch, P. Renaud, K.I. Jacob, Nano Energy. [26] M. Yamada, I. Honma, J. Phys. Chem. B 110 (2006) 20486e20490. [27] J. Zeng, S.P. Jiang, J. Phys. Chem. C 115 (2011) 11854e11863. [28] F. Cavani, Catal. Today 41 (1998) 73e86. [29] Y. Zhou, J. Yang, H. Su, J. Zeng, S.P. Jiang, W.A. Goddard, J. Am. Chem. Soc. 136 (2014) 4954e4964. [30] T. Haolin, L. Junrui, W. Zhao, Z. Huijie, P. Mu, J. San Ping, Self-assembly of nanostructured proton exchange membranes for fuel cells, in: Nanotechnology for Sustainable Energy, American Chemical Society, 2013, pp. 243e263. [31] A.M. Herring, H.M. Gregory, M. Fanqin, N.V. Aieta, J.L. Horan, M.H. Frey, S.J. Hamrock, M.C. Kuo, ECS Trans. 3 (2006) 551e559. [32] G.M. Haugen, F. Meng, N.V. Aieta, J.L. Horan, M.-C. Kuo, M.H. Frey, S.J. Hamrock, A.M. Herring, Electrochem. Solid-State Lett. 10 (2007) B51eB55. [33] M.P. Rodgers, L.J. Bonville, H.R. Kunz, D.K. Slattery, J.M. Fenton, Chem. Rev. 112 (2012) 6075e6103. [34] S. Malhotra, R. Datta, J. Electrochem. Soc. 144 (1997) L23eL26. [35] Z.-G. Shao, H. Xu, M. Li, I.M. Hsing, Solid State Ionics 177 (2006) 779e785. [36] K.H. Lee, J.Y. Chu, A.R. Kim, K.S. Nahm, C.-J. Kim, D.J. Yoo, J. Membr. Sci. 434 (2013) 35e43. [37] H. Wu, X. Shen, Y. Cao, Z. Li, Z. Jiang, J. Membr. Sci. 451 (2014) 74e84. [38] M. Yang, S. Lu, J. Lu, S.P. Jiang, Y. Xiang, Chem. Commun. 46 (2010) 1434e1436. [39] Q. Wu, H. Wang, S. Lu, X. Xu, D. Liang, Y. Xiang, J. Membr. Sci. 500 (2016) 203e210. [40] A.B. Bose, S. Gopu, W. Li, J. Power Sources 263 (2014) 217e222. [41] A.M. Herring, J. Macromol. Sci. C 46 (2006) 245e296. [42] J.L. Horan, A. Genupur, H. Ren, B.J. Sikora, M.-C. Kuo, F. Meng, S.F. Dec, G.M. Haugen, M.A. Yandrasits, S.J. Hamrock, M.H. Frey, A.M. Herring, ChemSusChem 2 (2009) 226e229. [43] J.L. Horan, A. Lingutla, H. Ren, M.-C. Kuo, S. Sachdeva, Y. Yang, S. Seifert, L.F. Greenlee, M.A. Yandrasits, S.J. Hamrock, M.H. Frey, A.M. Herring, J. Phys. Chem. C 118 (2013) 135e144. [44] E. Abouzari-Lotf, M.M. Nasef, H. Ghassemi, M. Zakeri, A. Ahmad, Y. Abdollahi, ACS Appl. Mater. Interfaces 7 (2015) 17008e17015. [45] M. Tanaka, Polym. J. 48 (2016) 51e58. [46] N. Feng, A. Zheng, S.-J. Huang, H. Zhang, N. Yu, C.-Y. Yang, S.-B. Liu, F. Deng, J. Phys. Chem. C 114 (2010) 15464e15472. [47] Y.S. Kim, F. Wang, M. Hickner, T.A. Zawodzinski, J.E. McGrath, J. Membr. Sci. 212 (2003) 263e282. [48] F. Meng, N.V. Aieta, S.F. Dec, J.L. Horan, D. Williamson, M.H. Frey, P. Pham, J.A. Turner, M.A. Yandrasits, S.J. Hamrock, A.M. Herring, Electrochimica Acta 53 (2007) 1372e1378. [49] T. Soboleva, Z. Xie, Z. Shi, E. Tsang, T. Navessin, S. Holdcroft, J. Electroanal. Chem. 622 (2008) 145e152. [50] R.J. Stanis, M.-C. Kuo, A.J. Rickett, J.A. Turner, A.M. Herring, Electrochimica Acta 53 (2008) 8277e8286.