TMCnet News

Fabrication of Surface-Patterned Membranes by Means of a ZnO Nanorod Templating Method for Polymer Electrolyte Membrane Fuel-Cell Applications [ChemPlusChem]
[August 27, 2014]

Fabrication of Surface-Patterned Membranes by Means of a ZnO Nanorod Templating Method for Polymer Electrolyte Membrane Fuel-Cell Applications [ChemPlusChem]


(ChemPlusChem Via Acquire Media NewsEdge) Surface-patterned Nafion films were prepared by using different types of ZnO nanorods as templates for polymer electrolyte membrane fuel-cell (PEMFC) applications. Varying the concentrations of the Zn precursor produced ZnO nanorods with different diameters and pore sizes in the Nafion films. The surface-tuned structure of the films improved fuel-cell performance owing to effective Pt loading on the membranes through an enhancement in surface area. Also, the interconnected morphology resulted in a reduction in charge-transfer resistance at the interface between the electrode and membrane. The optimum surface-patterned Nafion film recorded a current density of 1.19 A cm-2 at 0.6 V and a maximum power density of 0.95 W cm-2. These values are much higher than those of non-patterned Nafion (0.85 A cm-2 and 0.59 W cm-2, respectively). We expect that the patterning process using ZnO nanorod templates will improve the performance of any electrochemical device by allowing for the tuning of the interfacial resistance.



Keywords : fuel cells · membranes · polymers · template synthesis · zinc (ProQuest: ... denotes formulae omitted.) Introduction Polymer electrolyte membrane fuel cells (PEMFCs) have been the focus of much research and have great potential as clean- energy power sources owing to their high power densities, low operation temperatures, and zero greenhouse-gas emis- sions.[1] One of the important considerations in PEMFCs is find- ing a way to control the interface of membrane electrode as- semblies (MEAs). Unoptimized interface properties are the main hindrance impeding and retarding the commercial appli- cation of PEMFCs.[2] Effectively using Pt in the catalyst layer represents one of the main barriers to PEMFC commercializa- tion.[3] Therefore, a technique must be found to evenly layer the Pt catalyst onto the membrane to increase the reaction rate and improve fuel-cell performance.

The most commonly used method to influence the per- formance of energy-conversion devices is to control the sur- face area of the material. For example, high-contact-area struc- tures are important for solar cells because they allow for the creation of high-surface-area semiconductor materials, which result in higher energy-conversion efficiency.[4] Controlling the surface-tuned morphology with the high contact area of the interfacial layer between the semiconductor and the substrate can facilitate electron movement and decrease interfacial re- sistance. When this concept is applied to Nafion, the typical polymer electrolyte membrane used in PEMFCs, the surface- tuned structure is modified at the membrane surface to reduce the interfacial resistance between the membrane and the electrode.


Recently, there have been some results that indicate that the patterning of the polymer electrolyte membrane (PEM) en- hances the electrochemical property and fuel-cell per- formance.[5] The patterning results in an increased membrane surface area, which allows for increased catalytic reaction at the cathode side.[6] Additionally, the layered Pt catalyst nano- particles on the increased contact area of polymer electrolyte membrane have the potential to quickly transport gases through the high-surface-area catalyst[7] and utilize a large amount of Pt during the reaction.[8] Most importantly, the sur- face-tuned structure produces better interfacial contact be- tween the electrode and the membrane, forming high me- chanical strength and electrochemical stability owing to the strong interaction with the deeply penetrated Pt catalyst nano- particles.

Imprint/mold lithography is the most commonly used method to generate a surface-patterned structure.[5a-e] For ex- ample, Aizawa et al. fabricated an ordered micropillar-array- patterned Nafion membrane with controllable pillar widths and heights.[5a,b] Deshmukh et al. also reported a mircoline pat- terned structure by soft lithography, which improved water management and reactant distribution.[5c] Omosebi et al. dem- onstrated the use of electron-beam lithography coupled with dry etching for the nano/micropatterning of Nafion membra- nes.[5d,e] Also, Wang et al. developed a surface-patterned struc- ture by incorporating a porous polytetrafluoroethylene (PTFE) matrix within the MEA.[5f] However, these methods contain some limitations such as high cost and a complicated fabrica- tion method.

Therefore, in this study, we prepared 1 mm-thick surface-pat- terned Nafion membranes using different types of zinc oxide (ZnO) nanorods as templates, which is a relatively simple and low-cost process, for enhancement of fuel-cell performance. The designed MEA, consisting of a surface-modified membrane and electrode, provides an enhanced electrochemical surface area (ECSA) with higher Pt utilization and electrochemical ac- tivity on the cathode side of the PEMFC. These enhancements are due to the even distribution of Pt nanoparticles spread out on the surface-patterned Nafion surface to increase the contact area. The controllable porosity and connectivity of surface-pat- terned Nafion membranes were obtained by using ZnO nano- rods with different diameters as templates to control the sur- face area. The current density-voltage (I-V) curves, interfacial ohmic resistance, charge-transfer resistance, and catalyst load- ing amounts of MEAs were characterized in detail.

Results and Discussion Formation of ZnO nanorod templates The fabrication of surface-patterned Nafion membranes using ZnO nanorod templates is illustrated in Scheme 1. The zinc oxide was seeded on a glass dish and grown by means of a hy- drothermal method with different precursor concentrations.[9] Next, the Nafion solution in an alcohol/H2O mixture was poured onto the ZnO nanorod template. Upon solvent evapo- ration and ZnO etching, the surface-patterned Nafion mem- branes were produced. To change the nanorod diameters while maintaining similar thicknesses, the amounts of zinc ni- trate hexahydrate and hexamethylenetetramine were varied (using the same amount of methanol solution).[10] Figure 1 shows scanning electron microscopy (SEM) images of grown ZnO nanorods with diverse diameters. Overall, higher solution concentrations produced ZnO nanorods with larger diameters. The average thickness of all ZnO nanorods was controlled to be approximately 1 mm. This was done by changing the amount of polyethyleneimine and ammonium hydroxide, which largely affects the growth rate. Applying higher temper- atures to the hydrothermal reaction can also generate longer ZnO nanorods. The thickness of the ZnO nanorods was ap- proximately 1 mm. This is thick enough to minimize the interfa- cial resistance of the membrane/electrode. The ZnO nanorod template synthesized with a 50 mm zinc precursor concentra- tion showed nanorods with a diameter of 50 nm. This sample also contained a low density of nanorods without any large ag- gregates. Upon increasing the concentration of the zinc pre- cursor solution, the diameter of the ZnO nanorods becomes larger. ZnO nanorods with 100 and 300 nm average diameters were obtained using 75 and 100 mm zinc precursor solutions, respectively. When using a 125 mm zinc precursor solution, a greater number of ZnO interconnected nanorods were formed, which resulted in fewer void spaces.

Fabrication of surface-patterned Nafion films By using the ZnO nanorod templating method, we produced fine surface-patterned Nafion membranes with an average film thickness of 47 mm and porous surface-patterned thickness of 1 m m. Figure 2 displays surface SEM images of surface-pat- terned Nafion films prepared with various ZnO templates. When the Nafion solution was cast onto the ZnO nanorod template, it infiltrated into the void spaces near the ZnO nano- rods and developed a closely connected, surface-patterned Nafion film (after the ZnO template was etched away). The contact area of the Nafion films increases with increasing ZnO nanorod diameters, whereas the interconnectivity remains con- stant. However, the diameter of the ZnO nanorod template from the ZnO precursor-solution concentration of 50 mm was small, which provides a limited porosity compared to the Nafion film prepared with the 100 mm ZnO precursor solution as shown in Figure 2d. However, the Nafion film prepared by using the ZnO nanorod template utilizing the highest ZnO pre- cursor-solution concentration decreased in porosity owing to the fact that the 125 mm precursor ZnO nanorod template is too densely packed to produce a high contact area and con- nected patterned Nafion film at high precursor-solution con- centrations. The template-inhibiting infiltration of Nafion solu- tion evenly generated the Nafion film with lower surface area. The surface-patterned Nafion film obtained with the template derived from the 100 mm ZnO precursor solution is the best of our samples, and yielded the highest porosity structure. This sample decreases the interfacial resistance and increases the surface area for high catalyst loading to obtain improved MEA performance. Specifically, the patterned structure was generat- ed only on the membrane surface, as confirmed by the cross- sectional images of nonpatterned and surface-patterned Nafion films shown in Figure 3. For a detailed comparison, photographs of pristine glass, a ZnO nanorod template on glass, a nonpatterned Nafion membrane, and a surface-pat- terned Nafion membrane were taken and compared to one an- other, as shown in Figure S1 in the Supporting Information. Whereas pristine glass is transparent, the ZnO nanorods grown on the glass are opaque. This opacity is due to the formation of vertically grown nanorods, which lead to a difference in the refractive index. Similarly, the nonpatterned Nafion was trans- parent, whereas the surface-patterned Nafion membrane was opaque due to the patterned structure at the surface.

Electrochemical properties The electrochemical properties and performance of the MEAs prepared with surface-patterned membranes were evaluated. The synthesized membranes were coated by the catalyst of 0.4 mg cm^2 at both sides with an active surface of 1 cm2. MEAs were prepared by the assembly of catalyst-coated mem- brane, gas diffusion media, and Teflon gaskets including the pressing process. The electrochemical properties of the catalyst layers in the MEAs were characterized by cyclic voltammetry (CV) measurements. These can be used to determine the ECSA. The goal of this structure design is to maximize the triple- boundary phase at which O2 molecules, electrons, and protons combine at active catalyst sites.[7a] Figure 4 shows typical CV curves with strong hydrogen adsorption/desorption peaks without any vertical shifts, regardless of the surface-patterned structures of Nafion membranes. From the magnified plots, we can see that the hydrogen-adsorption peaks of the surface-pat- terned membranes moved to the higher current density region, which indicates an enhancement of the triple-boundary phase in the catalyst layer.

ECSA provides a numerical way to express the quality of the catalyst layer because it reflects particle location, particle size, and size distribution in the cathode. ECSA values [m2 g^1]of the Pt/C electrocatalyst in the cathode side scan were calculat- ed by the electrooxidation charge (QH) of the hydrogen ad- sorption on the smooth Pt surface by integrating the hydro- gen-adsorption peak obtained from the CV profile in Figure 4 (after correcting for double-layer capacitance). The electrical charge (c) associated with monolayer adsorption of the hydro- gen on the Pt catalyst surface (generally 21 mC cm^2) and the Pt mass (m) in the cathode [mg] were applied to define the fol- lowing equation [Eq. (1)] .

... (1) In a conventional electrode, it is common that some Pt/C catalyst particles may not be in contact with the Nafion binder, which results in poor electronic conductivity and a lower amount of accessible ECSA. As shown in Table 1, however, the ECSAs of the surface-patterned Nafion membranes were rather high with values of 34.03, 41.02, 55.77, and 32.78 m2 g^1 for 50, 75, 100, and 125 mm, respectively, compared with the pristine membrane (32.59 m2 g^1). The membrane fabricated by the 100 mm precursor achieved the highest value, whereas the membranes from 50 and 125 mm have values similar to that of the pristine membrane. This is likely caused by the fact that the 50 mm ZnO nanorods are too thin and the 125 mm ZnO nanorods are excessively large ; this inhibits their ability to ef- fectively influence the roughness of the membrane surface. However, all of the surface-patterned membranes showed an increase in the ECSA, which is attributed to a more uniform distribution of Pt/C particles. This allows for better proton transport and oxygen diffusion to catalyst sites during fuel-cell operation.

It is also possible to determine the amount the Pt/C catalyst particles in contact with the Nafion binder by calculating the Pt utilization value (UPt). For good electronic conductivity through the MEA, it is important to have high UPt since the less UPt starved with protons during fuel-cell operation. The UPt can be obtained from the CV measurements with respect to the Pt particle size. The particle size is also crucial because electrochemical reactions can only occur at the surface of Pt particles. The surface area of Pt (SPt) can be obtained by using the following equation (assuming uniform, spherical particles) [Eq. (2)].

.... (2) in which 1 is the mass density of Pt (21.4 g cm^3) and D is the mean particle size of Pt calculated from the Pt(111) diffraction peak in the X-ray diffraction (XRD) patterns (Figure S2 in the Supporting Information) of each catalyst layer in the cathode. In Table 1, the mean particle sizes were in the typical range of 3-4 nm for all MEAs as calculated by Scherrer's equation using a shape constant ( K ) of 0.9. A slight difference in the XRD peak intensity between the patterned and flat membranes was de- tected, maybe due to the change in the diffraction angle through a rough surface of the catalyst layers by a direct spraying of the catalyst ink to the patterned membrane sur- face. Because the particles were all similar in size, the values were similar and ranged from 78.37 to 82.11 m2 g-1. This is due to the usage of the same Pt/C, which leads to similarly effec- tive surface areas.

The UPt, which expresses the amount of active surface Pt atoms for electrochemical reactions, can be estimated from the ratio of the ECSA to the calculated SPt, as shown in the following equation [Eq. (3)].[8b] ... (3) As shown in Table 1, we obtained significantly higher UPt values of 41.87, 50.18, 69.82, and 43.49% for 50, 75, 100, and 125 mm, respectively, compared with that of the pristine membrane (39.41 %).

It is noticeable that the utilization of Pt increased as the ECSA increased, whereas the effective surface areas remained fairly constant. Therefore, we can conclude that the roughness of the patterned membrane is closely related to the distribution of active Pt nanoparticles and can influence the formation of effective electron and proton pathways. Therefore, the addition of a surface-patterned structure to the Nafion membranes is likely to enhance the electrochemical performance of MEAs owing to high Pt utilization and a high active surface area; these properties are closely associated with high performance in PEMFCs.

Impedance study of the prepared MEAs ITo investigate the resistances in the membrane and at the interface of the electrode/membrane, electrochemical impedance spectroscopy (EIS) was performed. The resulting Nyquist plots are shown in Figure 5. Nyquist plots show semicircles where the ohmic resistance value is equivalent to the x-axis intercept at a high frequency. This value is generally taken as the membrane resistance (ohmic). Generally, the ohmic loss is mainly related to the thickness of the membrane and defects in the membrane. A slight decrease in the membrane resistance for the surface-patterned membranes was observed (decreasing from 0.116 to 0.108 Wcm2). However, because all of the membranes used in this study were 46-47 mm thick, mem- brane thickness may not contribute to the resistance decrease. This suggests that the patterned features may influence the transport of ions through the MEA.

Charge-transfer resistance at the interface of the catalyst layer and the membrane was obtained from the diameter of the semicircle by EIS analysis. As shown in Figure 5 and Table 2, the flat and surface-patterned MEAs demonstrated a substantial decrease in charge-transfer resistance from 0.139 to 0.129, 0.111, 0.104, and 0.119 W cm2 for pristine, 50, 75, 100, and 125 mm, respectively. The MEA with the surface-patterned membrane from the 100 mm precursor had the lowest charge-transfer resistance due to its modified electrode structure (which had the highest specific area among the MEAs fabricat- ed). This demonstrates that en- hancing the interface of the tuned MEAs can lead to better mass transfer and high-per- formance MEAs.

PEMFC performance Polarization curves (I-V) and power density plots for H2/O2 single PEMFC, employing different surface-patterned structures at the cathode (pristine, 50, 75, 100, and 125 mm), are shown in Figure 6. The Pt loading is controlled at 0.4 mg cm^2 on both sides. The open-circuit voltage for all MEAs was approxi- mately 0.98 V at the similar membrane thickness (^ 47 mm) in the fabricated MEAs. MEA performance from the I-V curves can be divided into three parts : the activation-loss region, the ohmic polarization region, and the mass-transfer-controlled re- gion.[6c] The performance of the MEAs with surface-patterned membranes showed the low-current activation-loss region at 0.8 V. Using a patterned membrane yielded a slight improve- ment by increasing the utilization of Pt. Pt utilization is respon- sible for generating the activation reaction. As described above, the increased performance of the surface-patterned MEAs is due to improved interfacial properties at the elec- trode/membrane, which leads to an increase in the electro- chemically active surface area and better Pt utilization at the cathode.

The ohmic polarization region (mid-current range) at 0.6 V shows relatively reduced resistance and good contact between the catalyst, Nafion binders, and the surface-patterned Nafion membrane. This improvement is attributed to the enhanced triple-boundary phase in the cathode. It can be seen in Table 3 that the current densities of the 50, 75, and 125 mm mem- brane-based MEAs were 0.93, 1.02, and 0.96 A cm^2 at 0.6 V, re- spectively, showing a trend similar to what was observed for the ECSAs. The highest current density at 0.6 V (1.19 A cm^2) occurred when employing the surface-patterned membrane from the 100 mm precursor at the cathode side. Its maximum peak power density is high enough (0.95 W cm^2) for practical application of this MEA in PEMFCs. The polarization curves demonstrate better hydrogen electrooxidation activity and mass transfer through the cathode for all of the surface-pat- terned membranes synthesized in this study when compared to the pristine Nafion membrane (0.85 A cm^2 at 0.6 V).

Interestingly, a dramatic improvement in the mass-transfer- controlled region (high-current range) at 0.3 V is shown in Figure 6. In general, this region reflects performance associated with the interfacial design of the MEA (especially the triple- boundary phase). A high current density of 2.98 A cm^2 for the surface-patterned membrane from the 100 mm precursor was achieved ; this is approximately 58 % higher than that of the pristine Nafion membrane (1.73 A cm^2 at 0.3 V). This result re- flects a higher catalytic activity in the wider cathode catalyst region in which the oxygen-reduction reaction occurs. It is also indicative of shorter proton and electron pathways through the three-dimensional pore structures, which results in the lower electrode resistance.[5a-c] These results combine to yield effective mass transfer of both reactants and products.[11] Therefore, we can conclude that a surface-patterned structure on the Nafion membrane improves PEMFC performance.

Conclusion In this study, we demonstrate a simple synthesis method for a surface-patterned polymer electrolyte membrane for en- hanced fuel-cell performance through interface improvements. Interconnected modified structures of Nafion membranes were fabricated by utilizing ZnO nanorods as templates. The porosi- ty and interconnectivity of the Nafion membranes were con- trolled by using different ZnO nanorod templates that were hydrothermally fabricated by changing the Zn precursor con- centration from 50 to 125 mm (with 25 mm step size). Nano- rods with various diameters were fabricated to control the sur- face area. Use of surface-patterned Nafion membranes in- creased the ECSA and the utilization of the catalyst, which re- sulted in improved electrochemical performance of the MEA. Furthermore, we obtained a 53 % improvement in the current density and a 59 % increase in the power density in a PEMFC when using the surface-patterned MEA with the highest sur- face area. Our approach will help facilitate the manufacture of PEMFCs with improved Pt utilization and an increase in the electrochemically active electrode surface area. Furthermore, our patterning process can be used to produce a wide variety of structures and shapes depending on the desired applica- tions.

Experimental Section Preparation of ZnO nanorod templates ZnO nanorod arrays were grown directly on glass dishes. To depos- it the ZnO seed layer, the ZnO seed solution was prepared by dis- solving diethylzinc solution (0.37 mL; 1.0 m in n-hexane) in tetrahy- drofuran (THF ; 5 mL), which was stirred for 1 h at room tempera- ture. The solution was spin-coated on glass dishes at 2000 rpm for 20 s. This was followed by thermal decomposition at 450 8C for 30 min. For hydrothermal synthesis of ZnO nanorods, aqueous sol- utions of zinc nitrate hydrate (Zn(NO3)2·6 H2O) and hexamethylene- tetramine were prepared and stirred to make homogeneous solu- tions. The same concentration of zinc nitrate hydrate and hexame- thylenetetramine was used ; these concentrations varied from 50 to 125 mm (with a 25 mm step size) in water. Then, the seeded sub- strates were put face-down into Teflon-sealed glass bottles con- taining growing solution (100 mL). The growing process was per- formed for 4 h in a 90 8C oven. The resulting arrays were rinsed with deionized water and sintered at 450 8C for 30 min to remove any residual salts and organics.

Preparation of surface-patterned Nafion membranes To prepare surface-patterned Nafion films, Nafion perfluorinated resin solution (3 mL; 5 wt % solution in lower aliphatic alcohol/H2O mix, equivalent weight (EW) = 1100, Aldrich) was cast on the ZnO patterned glass dish. This dish was covered with hole-treated alu- minum foil to evaporate the solvent slowly to fabricate perfect Nafion membranes without any defects. Samples were dried at 508C for a day. The patterned Nafion was detached using a 1 n HCl solution etching process. The patterned Nafion film was washed with distilled water and kept in a vacuum oven for 24 h to elimi- nate water. Five kinds of Nafion films using ZnO nanorods (50, 75, 100, and 125 mm) and a nonpatterned membrane were prepared.

For the characterization of the physical properties, five different measurements were utilized. Morphologies of ZnO nanorod tem- plates and surface-patterned Nafion films were observed by using a field-emission scanning electron microscope (SUPRA 55VP, Ger- many, CarlZeiss) operated at 15 kV after the samples were sput- tered with platinum at 10 mA for 100 s. XRD was recorded by Riga- kuMiniflex AD11605 with CuKa radiation (l = 1.5405 nm) to study the particle size of the electrocatalysts. The tube current was 30 mA and the tube voltage was 30 kV. The 2q angular regions be- tween 25 and 858 were explored at a scan rate of 58 min^1.

Fabrication of the MEAs Carbon-supported Pt catalysts (Johnson Matthey, 40 wt % Pt on carbon black) were used for the MEA preparation. To spray the cat- alyst onto the Nafion membrane with an airbrush gun, catalyst ink was prepared by mixing carbon-supported Pt (0.3 g), deionized water (1.2 g), Nafion ionomer solution (0.8 g), and isopropyl alcohol (3.6 g ; Aldrich). The catalyst slurries were mechanically stirred and ultrasonicated alternately five times to allow for complete mixing of the ionomer and the Pt particles. The catalyst (0.4 mg cm^2) was loaded directly on the prepared membranes at both sides with an active surface area of 1 cm2. Then, the catalyst-coated membranes were dried at room temperature for 1 h to remove residual solvent. MEAs were prepared by the assembly of catalyst-coated Nafion membranes, gas diffusion media (SGL 10BC), and Teflon gaskets. This step included a pressing process (0.9 ton for 5 min at room temperature).

Measurement of single-cell performance The single-cell performances of the prepared MEAs were evaluated by a single PEMFC station (Fuel Cell Technology, USA). Pure hydro- gen and oxygen were fed into the cell as a fuel and oxidation gas. The flow rates of H2 and O2 were 0.2 and 0.3 L min^1, respectively, with a stoichiometry of 1.5 :2 (H2/O2). To control the relative humid- ity, the feed gases of hydrogen and oxygen were preheated to the same temperature by a water humidifier. All the electrochemical experiments were performed at 758C with 100 % relative humidity (RH) without giving a back-pressure. An activation step at a con- stant current of 3 A cm^2 for 1 h and I-V curve plotting were each carried out three times. I-V curves, CV, and EIS were measured for the surface-enhanced MEAs. A KFM2030 instrument (KIKUSUI) was used as the electric load to determine the polarization curves (cur- rent density versus voltage). A PGSTAT-30 instrument (Autolab) was used for in situ EIS and CV. EIS was performed to examine the membrane/electrode resistances with an amplitude of 10 mV be- tween 100 MHz and 10 kHz. Moreover, the potential of the CV was scanned between 0 and 1.35 V with a scan rate of 50 mV s^1 to an- alyze the electrochemical properties of the catalyst layer.

Acknowledgements This study was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (MEST) (2009-C1AAA001-2009-0092926), the Active Polymer Center for Pattern Integration (2007-0056091), and the Core Research Pro- gram (2012R1A2A2A02011268).

[1] a) M. K. Debe, Nature 2012, 486, 43- 51; b) F. Barbir, T. G^mez, Int. J. Hy- droqen Energy 1997, 22, 1027 - 1037; c) T. B. Norsten, M. D. Guiver, J. Murphy, T. Astill, T. Navessin, S. Holdcroft, B. L. Frankamp, V. M. Rotello, J. Ding, Adv. Funct. Mater. 2006, 16, 1814 - 1822.

[2] a) H. Zhang, P. K. Shen, Chem. Soc. Rev. 2012, 41, 2382-2394; b) H. Zhang, P. K. Shen, Chem. Rev. 2012, 112, 2780-2832; c) R. Borup, J. Meyers, B. Pivovar, Y. S. Kim, R. Mukundan, N. Garland, D. Myers, M. Wilson, F. Garzon, D. Wood, P. Zelenay, K. More, K. Stroh, T. Zawodzinski, J. Boncella, J. E. McGrath, M. Inaba, K. Miyatake, M. Hori, K. Ota, Z. Ogumi, S. Miyata, A. Nishikata, Z. Siroma, Y. Uchimoto, K. Yasuda, K.-i. Ki- mijima, N. Iwashita, Chem. Rev. 2007, 107, 3904 - 3951.

[3] a)B. Liu, S. Creager, J. Power Sources 2010, 195, 1812-1820; b)S. Yin, S. Mu, M. Pan, Z. Fu, J. Power Sources 2011, 196, 7931 - 7936 ; c) D. Dixon, J. Melke, M. Botros, J. Rathore, H. Ehrenberg, C. Roth, Int. J. Hydrogen Energy 2013, 38, 13393-13398; d) B. Liu, S. Creager, Electrochim. Acta 2010, 55, 2721 -2726.

[4] a) M. Zukalov^, A. Zukal, L. Kavan, M. K. Nazeeruddin, P. Liska, M. Gr^tzel, Nano Lett. 2005, 5, 1789-1792; b)J. T. Park, W. S. Chi, D. K. Roh, S. H. Ahn, J. H. Kim, Adv. Funct. Mater. 2013, 23, 26-33; c)S. H. Ahn, W. S. Chi, J. T. Park, J. K. Koh, D. K. Roh, J. H. Kim, Adv. Mater. 2012, 24, 519 - 522.

[5] a) M. Aizawa, H. Gyoten, A. Salah, X. Liu, J. Electrochem. Soc. 2010, 157, B1844; b) M. Aizawaz, H. Gyoten, J. Electrochem. Soc. 2013, 160, F417 - F428 ; c) A. B. Deshmukh, V. S. Kale, V. M. Dhavale, K. Sreekumar, K. Vi- jayamohanan, M. V. Shelke, Electrochem. Commun. 2010, 12, 1638 - 1641; d) A. Omosebi, R. S. Besser, J. Electroanal. Chem. 2011, 158, D603 - D610; e) A. Omosebi, R. S. Besser, J. Power Sources 2013, 228, 151 - 158; f) L. Wang, S. G. Advani, A. K. Prasad, J. Phys. Chem. C 2013, 117, 945 - 948 ; g) J. K. Koh, Y. Jeon, Y. Cho, J. H. Kim, Y. G. Shul, J. Mater. Chem. A DOI : 10.1039/C4A00674G.

[6] a) Z. Zhou, R. N. Dominey, J. P. Rolland, B. W. Maynor, A. A. Pandya, J. M. DeSimone, J. Am. Chem. Soc. 2006, 128, 12963 - 12972 ; b) J. W. Lee, C.W. Yi, K. Kim, Bull. Korean Chem. Soc. 2012, 33, 1788-1790; c)J. Li, H. Tang, L. Chen, R. Chen, M. Pana, S. P. Jiang, Chem. Commun. 2013, 49, 6537-6539; d)Y. H. Cho, J. W. Bae, Y. H. Cho, J. W. Lim, M. Ahn, W. S. Yoon, N. H. Kwon, J. Y. Jho, Y. E. Sung, Int. J. Hydrogen Energy 2010, 35, 10452-10456.

[7] a) W. Zhang, P. N. Pintauro, ChemSusChem 2011, 4, 1753 - 1757; b) P. C. Sherrell, W. Zhang, J. Zhao, G. G. Wallace, J. Chen, A. I. Minett, ChemSu- sChem 2012, 5, 1233 - 1240 ; c) M. S. Saha, R. Li, X. Sun, J. Power Sources 2008, 177, 314 - 322.

[8] a)Z.N. Farhat, J. Power Sources 2004, 138, 68-78; b)M. Uchida, Y. C. Park, K. Kakinuma, H. Yano, D. A. Tryk, T. Kamino, H. Uchidab, M. Wata- nabe, Phys. Chem. Chem. Phys. 2013, 15, 11236 - 11247; c) D. Zhao, B. Q. Xu, Angew. Chem. 2006, 118, 5077 - 5081.

[9] T. Ma, M. Guo, M. Zhang, Y. Zhang, X. Wang, Nanotechnology 2007 , 18, 035605.

[10] a) L. Vayssieres, Adv. Mater. 2003, 15, 464 - 466 ; b) M. Guo, P. Diao, S. Cai, J. Solid State Chem. 2005, 178, 1864 -1873.

[11] S. Sambandam, J. Parrondo, V. Ramani, Phys. Chem. Chem. Phys. 2013, 15, 14994.

Received : March 25, 2014 Published online on May 21, 2014 Won Seok Chi, Yukwon Jeon, Se Jun Park, Jong Hak Kim,* and Yong-Gun Shul*[a] [a] W. S. Chi,+ Y. Jeon,+ S. J. Park, Prof. J. H. Kim, Prof. Y.-G. Shul Chemical and Biomolecular Engineering Yonsei University, 50 Yonsei-ro Seodaemun-gu, Seoul 120-749 (Korea) E-mail : [email protected] [email protected] * These authors contributed equally to this study.

* Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/cplu.201402083.

(c) 2014 Blackwell Publishing Ltd.

[ Back To TMCnet.com's Homepage ]