Understanding a Degradation Mechanism of Direct Methanol Fuel Cell

12 Dec 2007 - From the mass-resolved imaging technique of TOF-SIMS combined with a peel-off method, it was indicated that the traveled ruthenium from ...
0 downloads 0 Views 551KB Size
J. Phys. Chem. C 2008, 112, 313-318

313

Understanding a Degradation Mechanism of Direct Methanol Fuel Cell Using TOF-SIMS and XPS Youngsu Chung,*,† Chanho Pak,*,‡ Gyeong-Su Park,† Woo Sung Jeon,† Ji-Rae Kim,‡ Yoonhoi Lee,‡ Hyuk Chang,‡ and Doyoung Seung‡ Analytical Engineering Center and Energy and EnVironment Laboratory, Samsung AdVanced Institute of Technology, P. O. Box 111, Suwon, 440-600, Korea ReceiVed: July 27, 2007

The catalyst layers, which were obtained from the aged membrane electrode assemblies (MEAs) showing performance loss, 17%, 37%, 45% as compared with that from the pristine MEA were investigated using X-ray photoelectron spectroscopy (XPS) and time-of-flight secondary ion mass spectroscopy (TOF-SIMS) to understand the degradation mechanism of single-cell performance in direct methanol fuel cell. The metallic components in the PtRu anode catalyst layer was decreased significantly with the performance drop revealed from the change of the valance bands. It has been found that ruthenium and platinum were driven to the cathode by the gradient of electric field during the durability test. From the mass-resolved imaging technique of TOF-SIMS combined with a peel-off method, it was indicated that the traveled ruthenium from anode to cathode was not uniformly distributed within the cathode catalyst layer, but more accumulated at the surface of the catalyst layer, that is, the interface with the gas diffusion layer. XPS and TOF-SIMS results suggested that the ruthenium ions are electro-deposited to RuOx rather than metallic Ru, which seems to more closely correlate with the degradation of performance.

Introduction Direct methanol fuel cell (DMFC) is getting attractive as a promising power source for portable electronics because of its advantages such as a high-energy density, green emission, convenient refueling of liquid fuel, and ambient operating condition.1 However, there still remain several critical issues to be resolved for commercialization of the DMFC. Among them, a working lifetime is one of the critical issues to be addressed. Nowadays, portable power sources for a note PC, portable multimedia player, 4G cellular phone, and so forth are requested to ensure working lifetime for thousands of hours, which is not satisfied by present DMFC systems. Thus, identifying the factors that affect the durability of DMFC and understanding the degradation mechanism are essential to improve the lifetime. A number of factors, which cause performance loss, have been reported to date and can be classified with regard to the components of membrane electrode assembly (MEA), which is a core compartment of DMFC as well as polymer electrolyte membrane fuel cell (PEMFC). MEA is composed of a membrane, gas diffusion layer (GDL), and catalyst layer, and the durability of each component is an important concern to achieve a long working lifetime. The membrane degradation came mainly from decomposition of the polymer, resulted in the decrease of ionic conductivity and thickness of the membrane, and, eventually, formation of a pinhole in the membrane.2 The hydrogen peroxide radical has been suspected as a potential cause of membrane decomposition.3,4 The physical changes such * Corresponding authors. Youngsu Chung: phone: +82-31-280-8371; fax: +82-31-280-9157; e-mail: [email protected]. Chanho Pak: phone: +82-31-280-6884; fax: +82-31-280-9359; e-mail: chanho.pak@ samsung.com. † Analytical Engineering Center. ‡ Energy and Environment Laboratory.

as hydrophobic/hydrophilic properties have been addressed regarding to the deterioration of GDL. The deterioration of catalysts results mainly from the decrease of active sites or the change of electronic properties due to agglomeration,5-7 poisoning, oxidation,8 or crossover of catalytic components.9,10 Especially, Piela et al.9 reported an important phenomenon of ruthenium crossover from the PtRu catalyst in the anode to the cathode during a durability test using X-ray fluorescence (XRF). The crossover ruthenium at the cathode inhibits oxygen reduction kinetics and possibly the cathode’s ability to handle methanol crossover.9 Besides these, ruthenium crossover means the loss of the catalytic anodic component and certainly affects the activity of the anode catalyst itself for methanol oxidation reaction. The atomic scale behavior of catalyst nanoparticles under DMFC operation conditions showed that the decomposition of PtRu nanoparticles at the anode occurred through morphological alteration or cracking by transmission electron microscopy (TEM).11 However, the compositional changes in the catalyst layer accompanied by performance loss are not well investigated at a microscopic scale. In this study, we have used X-ray photoelectron spectroscopy (XPS) and time-of-flight secondary ion mass spectroscopy (TOF-SIMS) to investigate the compositional change with performance loss. XPS is the most frequently applied surface analysis technique in the field of catalysts, and both qualitative and quantitative analysis have been used for all kinds of catalysts.12 In particular, for DMFC, it has played an important role in elucidating the chemical state of ruthenium and platinum involved in the methanol oxidation reaction.13,14 We also used XPS to probe not only the chemical state and quantification of crossed ruthenium but also the change of electronic properties of catalyst layers to trace the reason of the performance loss. Secondary ion mass spectroscopy (SIMS) has been useful for the surface chemistry dealing with the adsorptions and reactions

10.1021/jp0759372 CCC: $40.75 © 2008 American Chemical Society Published on Web 12/12/2007

314 J. Phys. Chem. C, Vol. 112, No. 1, 2008

Chung et al.

Figure 1. Schematic diagram of a MEA. After detachment of GDL from both sides of MEA, the exposed surfaces of the cathode and anode catalyst layers (dot lines) were analyzed.

of adsorbed species in the model systems,15-18 but the applications in the practical fields such as fuel cell or lithium battery have been limited.19,20 The unique advantages such as great sensitivity, isotope sensitivity, detection of molecular cluster ions, and shallow information depth (1-2 nm) are further extending the applicability of SIMS in these fields. In this study, TOF-SIMS was employed successfully to visualize the surface molecular distributions as well as in-depth distribution of crossed ruthenium within the cathode catalyst layer at a microscopic scale. This work has demonstrated the great potential of TOFSIMS in combination with XPS in application of the fuel cell. Experimental Section The catalyst layers were prepared using a commercial PtRu black (HiSpec6000, Johnson Matthey) catalyst for the anode layer, a commercial Pt black (HiSpec1000, Johnson Matthey) catalyst for the cathode layer, respectively, and Nafion ionomer solution (DuPont, 5% dispersion). The commercial black catalysts were used as received without any treatment. The composition of PtRu black was determined to be 61.0 ( 2.0 atom % for Pt and 39 ( 2.0 atom % for Ru by EDS (energydispersive X-ray spectroscopy).11 The polymer electrolyte membrane is a commercial Nafion 115 (DuPont Corp.). To make a MEA, we hot-pressed the anode and cathode catalyst layers onto both sides of a piece of Nafion 115 membrane at 135 °C for 3 min. The catalyst metal amount on both catalyst layers was 6 mg/cm2. The thicknesses of anode layer, membrane, and cathode layer were determined using SEM to 15, 90, and 1030 µm, respectively. The MEA was then assembled into a single cell fixture with serpentine flow channels and gold-coated copper plates. The durability test of DMFC was carried out in a single cell with an active cross-sectional area of 10 cm2. Methanol solution (1 M) was delivered to the anode with a flow rate of 0.3 mL/A, and dry air was fed to the cathode with a flow rate of 55 mL/ A. The temperature of a single cell was controlled to maintain 50 °C. The performances of three MEA were monitored at different conditions to drive different performance degradations. The first sample (M1) with 17% performance drop was obtained by operating at 0.4 V for 2 h on each day over 114 h. The second MEA (M2) and the third MEA (M3) were aged over 136 and 128 h at 0.25 and 0.35 V, respectively, with daily onoff mode operation for 8 h. The performance drop for M2 and M3 is 37% and 45%, respectively. After the test, the MEA samples were taken out from the cell fixture carefully and frozen in liquid nitrogen for 10 min. The frozen MEAs were immediately cut into several pieces for further investigations. Figure 1 shows a schematic cross section of a MEA sample.

Figure 2. Valence band spectra for (a) the cathode surfaces and (b) the anode surfaces of the pristine and durability-tested MEAs.

The GDLs can be detached from both sides of MEA easily, and the exposed surfaces of respective cathode and anode catalyst layers (dot lines in Figure 1) were characterized immediately using XPS and TOF-SIMS. XPS measurements were carried out using a PHI Q2000 equipped with a monochromatic Al KR source. Atomic concentration was determined using the peak areas of C 1s, O 1s, F 1s, Pt 4f, and Ru 3d5 normalized with their ionization cross sections. TOF-SIMS analyses were performed using an IONTOF IV. The surface spectra were acquired using 25 kV Bi+1 primary ions within static limit (1 × 1012 ions/cm2). The peak intensities were normalized using the corrected total ion intensities. Mass-resolved ion images were also acquired using a focused ion beam with a beam size of about 5 µm over 500 × 500 µm2. To investigate the vertical distribution of components to the membrane, we applied the peel-off method to the catalyst layer of the cathode. Using a clean 3M tape, the top layer was peeled off successively and the newly exposed surfaces were analyzed using TOF-SIMS. After the third peeloff, we cannot peel off any more, which implies that the surface exposed by the last peel-off is close to the membrane side. Results and Discussion Figure 2a and b show the valence band spectra of the cathode and anode catalyst layers, respectively, for pristine and durability-tested samples (M1, M2, and M3) with increasing performance drop. Each spectrum was normalized to the maximum peak intensity. Valence bands from the aged anode (Figure 2b) showed significant changes with performance drop, whereas valence bands from the aged cathode (Figure 2a) displayed only slight changes. As shown in Figure 2a, the Pt 5d bands between 0 and 6 eV near the Fermi level remained regardless of performance drop, implying no change of metallic character in the catalytic metal particles at the cathode surface. For the anode, we can observe the decrease in intensities of hybrid bands of the Pt 5d and Ru 4d states between 0 and 7 eV and the increase in intensities around 9.6, 14, and 19.8 eV (indicated by arrows in Figure 2b) with increasing performance drop, which correspond to the bands from the ionomer, that is, perfluorosulfonate polymer, Nafion. This suggests that the performance degradation of MEA may be related to the decrease the metallic character of the anode catalyst layer. Ruthenium crossover of the PtRu catalyst from the anode to the cathode was reported by Piela et al.9 Because the ruthenium loss could be a cause of the decrease of metallic character in the anode side, we investigated the phenomenon of Ru crossover

Degradation Mechanism of DMFC

Figure 3. (a) Pt 4f and (b) C 1s XPS spectra for the surfaces of pristine and aged cathode catalyst layers.

in more detail and focused mainly on the cathode surface to detect Ru transferred from the anode during the DMFC operation. The evidence of Ru crossover could be obtained using XPS. Figure 3 shows the XPS spectra of Pt 4f and C 1s for the cathode surfaces obtained from pristine and aged MEA samples. The Pt 4f spectra showed little change with the performance degradation, which indicates that the oxidation of Pt at the cathode surface is not related to the performance loss, in agreement with a slight change of valence bands from the cathode shown in Figure 2a. The C 1s spectra displayed two main peaks for pristine and M1 cathode surfaces. A strong peak around 290 eV is assigned to the carbon in CF2 from Nafion ionomers, which was introduced in preparation of the catalyst layer. A peak at about 284 eV is ascribed to the carbon in hydrocarbons. For more deteriorated M2 and M3, another peak appeared at about 281 eV. This low binding energy peak can be attributed to Ru 3d5/2, and the peak at 284 eV becomes broader because of the superposition of the Ru 3d3/2 peak with the C 1s peak. It is certain that the Ru traveled from the anode through the polymer membrane to the cathode surface in agreement with previous results.9,10 The atomic concentration of Ru accumulated on the cathode surface of aged MEA samples was less than 0.4 atom % based on quantitative XPS analysis. In our previous study,11 the Pt crossover from the anode was suggested through the quantification results of Pt to Ru ratios for PtRu catalyst particles using energy-dispersive spectroscopy (EDS) and the observation of an amorphous Pt phase at the aged cathodes, which is newly formed after the durability test. Quantitative analysis using XPS in the present study also provides evidence of the loss of catalytic Pt and Ru components in the anode. The sum of metallic Pt and Ru contents decreased to 0.6-1.4 atom % of the aged anodes from 3.1 atom % with respect to the total atoms of the pristine anode. Therefore, the loss of metallic properties of anodes shown in Figure 2b can be explained by the actual loss of active catalytic metal components from the anode catalyst layer due to migration of Ru as well as Pt.10,11 The Ru to Pt ratio was also determined to be 0.65 for the pristine anode surface and was decreased to 0.44 ( 0.05 for the anodes from M2 and M3 by XPS, which means that PtRu particles in the anode lost more Ru atoms than Pt atoms during the operation with DMFC conditions. Therefore, Ru loss is reflected more than Pt loss in the change of anode valence bands shown in Figure 2b. It is certain that the loss of catalytic components from the anode results in a decrease in catalytic activity for the methanol oxidation and is possibly one reason for the performance degradation of DMFC.

J. Phys. Chem. C, Vol. 112, No. 1, 2008 315 However, the question about how ruthenium or platinum could migrate from anode to cathode still remains. Piela et al.9 insisted that an amorphous hydrous RuO2 component in the PtRu black catalyst in the anode is likely to penetrate a membrane to reach the cathode. They also assumed that the Ru dissolution potential is expected to remain much above the typical potential of the DMFC anode, which indicates that the PtRu alloy is maintained mostly under operating conditions. However, according to the Pourbaix diagram,21 hydrous RuO2 will be transformed to metallic Ru easily in a normal DMFC anode operation region, that is, potential range between 0.4 and 0.5 V and in a highly acidic conditions from the Nafion membrane and ionomers.9,22 In addition, PtRu particles can be decomposed into small particles during the operation, which was supported by direct TEM observation of decomposed small metal particles with sizes less than 2 nm after the durability test as shown in the previous study.11 It has been known that the physical and chemicalpropertiesofnanoparticles,suchasmeltingtemperature23-25 and dissolution,26,27 can differ significantly from those of the corresponding bulk materials because of the limited number of atoms within the nanoparticles and high portion of the surface atoms with lower binding energy.28 Recently, it was reported that the solubility of Pt nanoparticles in the supported catalysts was enhanced in acidic solution.29 The dissolution potential of the small nanoparticles could be lowered below the typical anode potential. Furthermore, the small nanoparticles will be more active toward methanol or water to form the positively charged metal ions or particles in the acidic operation conditions of the DMFC anode. These positively charged species could drift facilely across the membrane by the gradient of electric field and reach the cathode. The vertical distribution to the membrane of traveled ruthenium within the catalyst layer of the cathode was investigated by TOF-SIMS with a peel-off method. We peeled off the cathode layer with a tape and obtained the mass-resolved ion images of the newly exposed surfaces using TOF-SIMS. Although it is difficult to estimate the thickness of the detached layer by a peel-off, the interface between the cathode catalyst layer and the membrane could be observed after the third peeloff by SEM (see the Supporting Information Figure S1). Approximately, the thickness of the detached layer by a peeloff seems to be several micrometers. In addition, the detached layers were uniformly adhered to the 3M tape as shown in Figure S1. Therefore, we believe that this peel-off method combined with a TOF-SIMS imaging technique could provide roughly the composition distribution with vertical depth to membrane by repeating the peel-off. Figure 4 shows the sequential lateral distribution of the Ru ions from the M3 cathode at different numbers of peel-off, that is, at different locations in the cathode approaching to the membrane. The first image without peel-off corresponds to the Ru distribution at the top cathode surface, that is, interface with GDL, the first and second peel-off images to middle parts of the cathode catalyst layer, and the last one to the interface adjacent to the membrane (refer to Figure 1). It was suggested that Ru is not uniformly distributed within the catalyst layer of the cathode but is accumulated more at the cathode surface, that is, the interface between the cathode catalyst layer and GDL. This could be explained by considering an electro-deposition of positively charged ruthenium species as described above. The electro-deposition of cationic Ru species could occur easily at this interface between the cathode catalyst layer and GDL

316 J. Phys. Chem. C, Vol. 112, No. 1, 2008

Chung et al.

Figure 4. TOF-SIMS images of the crossover ruthenium ion for the M3 cathode with increasing peel-off times. Field of view: 500 × 500 µm2.

Figure 5. RGB color overlay of the Pt (red), the Ru (green), and the Nafion (blue) for (a) pristine and (b) M3 cathodes with increasing peel-off times. Field of view: 500 × 500 µm2.

because the electron from the current collector located outside of GDL (refer to Figure 1) moves to the catalyst layer through the GDL. Initially, ruthenium presented only in the anode as PtRu black particles. The Ru concentration of the pristine anode surface was about 1.2 atom % when the sum of atomic Pt, Ru, C, O, and F concentrations was set to 100%. This value might appear to be too small when considering the catalyst loading amount. However, this value was calculated including the Nafion ionomers in the catalyst layer, which tend to remain on the surface, so that C and F are the major components at the surface within the escape depth of excited electrons in XPS. This value is similar to the 1.76 atom % estimated by XPS in the previous literature30 where the same commercial PtRu black catalyst as described in the experimental section was used to fabricate the anode of DMFC. During the durability test, ruthenium migrated from the anode and the Ru amount left on the M1 anode surface decreased to about half of the initial value, 0.5 ( 0.05 atom %. However, the Ru was not detected on the M1 cathode surface within the XPS detection limit. For the M2 and M3 anode surfaces, the Ru amount left on the anode surfaces was reduced significantly to about 0.2-0.3 atom % and the detected amount of Ru on the aged M2 and M3 cathode surfaces was only about 0.3 atom %. The sum of the Ru amount detected in the aged MEAs was much lower than that of the original MEA. Where is the undetected ruthenium? We expect that a portion of Ru species crossed from the anode reaches the cathode surface but most of them move further to be accumulated inside the GDL, where the electron density is high. Figure 5 shows RGB overlays of mass-resolved ion images using TOF-SIMS for pristine (a) and M3 (b) cathodes with peeloff numbers. The Pt, Ru, and Nafion (CF3+) distributions are expressed in red, green, and blue, respectively. Distributions of Ru in the cathode of M3 with peel-off numbers are also shown in Figure 4. Areas where the two or three components

are present simultaneously appear as the relevant mixed color. For pristine cathode, the violet color remained within a whole catalyst layer of the cathode even though blue areas (Nafion ionomer) increased slightly and red areas (Pt catalyst) decreased slightly with increasing peel-off numbers, that is, closely approaching the membrane. For the cathode in the M3 sample, the red color (Pt catalyst) diminished suddenly only after the first peel-off, green areas (Ru) increased, and blue (Nafion ionomer) became dominant getting closer to the membrane. The increase of the green area does not imply the increase of Ru concentration as can be seen in Figure 4 but rather reflects the decrease of Pt concentration when approaching the membrane. The Pt catalysts are rich only in the surface region, and the inside of the catalyst layer seems to be devoid of catalyst components for the deteriorated M3 cathode. The loss of the Pt component inside the cathode sustains the drift of cationic metal components by electric field and electro-deposition at the interface between the catalyst layer and GDL. The cationic metal components, which have been observed often in the cathode of the polymer electrolyte membrane fuel cell,31,32 might be generated by the oxidation of metal during the DMFC operation. The loss of Pt component inside of the aged cathode suggested from the peel-off experiment was not reflected in the valence bands of the cathodes in Figure 2a because the valence bands probed only the surface electronic structure less than 6 nm while the images after a peeloff correspond to the distributions of locations far from the surface, about several micrometers as mentioned earlier. It is sure that the drift of Pt and Ru to the cathode surface, that is, the interface with GDL, has a negative effect on the methanol oxidation reaction in the anode as well as on the oxygen reduction reaction in the cathode and, consequently, on the performance of the fuel cell. Therefore, the performance decrease shown in samples M1 and M2 can be explained using the loss of catalytic metallic components.

Degradation Mechanism of DMFC

Figure 6. Intensity plots of PtO- and RuO3- ions, which are the characteristic peaks of oxidized Pt and Ru, respectively, for pristine and aged cathode surfaces. The peak area intensities were averaged from five spectra over 200 × 200 µm2.

The intensities of the Ru+ ion by TOF-SIMS and the Ru amount by XPS on the cathodes increased from pristine to M1 and M2 with performance decrease. However, the amounts of crossed ruthenium to the M2 and M3 cathodes are similar, about 0.3 atom %, despite the difference in the performance drop, 37% and 45%, respectively. In addition, the Pt concentration of the M3 anode is slightly higher than that of the M2 anode. Thus, it is difficult to explain the performance difference between M2 and M3 using only the amounts of crossed ruthenium as well as platinum. It has been reported that the chemical state of ruthenium at the anode is one of the key factors of methanol electro-oxidation.13,14 The ruthenium state not only in the anode but also in the cathode could be another important factor for performance of DMFC because Ru deposition on the Pt particles in the cathode could hinder the active sites of Pt or change the catalytic activity. We have found that the binding energy of the Ru 3d5/2 peak shifted from 280.3 eV for M2 to 280.7 eV for M3 and the fwhm also decreased as shown in Figure 3. The peak for M2 seems to be composed of three components, a metallic ruthenium at ca. 279.9 eV, RuOx at 280.4 eV, and RuO2 at 281.3 eV, respectively.13,14,33 The high binding energy shift from M2 to M3 results from the decrease of a metallic ruthenium component and an increase of oxidized Ru phases, implying that the oxidation of ruthenium was more developed on the more deteriorated cathode surface. However, the intensities of the Ru 3d5/2 peaks are too small to deconvolute for quantification. For more clear correlation between oxidation of catalytic components and performance degradation, we used TOF-SIMS with an excellent sensitivity to detect and quantify very small amounts of oxidized Ru and Pt species on the cathode surfaces. Figure 6 displays the intensity plots of PtO- and RuO3peaks for pristine and aged surfaces of the cathode catalyst layer. PtO- and RuO3- peaks are the characteristic peaks of oxidized Pt and Ru species, respectively. The intensity of the PtO- peak does not display appreciable change for aged cathode surfaces, in agreement with the XPS results shown in Figure 3a. Alternatively, the intensity of the RuO3- peak increased noticeably for the aged cathode, with increasing performance drop. This suggests that the amount of oxidized Ru species generated by the oxidation of crossed Ru is closely correlated to the performance degradation. Ru ions from the PtRu anode could be electro-deposited by accepting electrons at the interface of the catalysts layer and GDL as metallic ruthenium or ruthenium oxides. The formation of ruthenium oxide is preferred to that of metallic Ru as evidenced from the XPS result in Figure 3b. Ruthenium oxides are to be formed through the electro-

J. Phys. Chem. C, Vol. 112, No. 1, 2008 317 chemical reaction of Ru ions with O2 gas supplied to the cathode. The performance decrease from M2 to M3 can be explained using oxidation of crossover ruthenium. This means O2 gas supplied to the cathode is consumed through an unnecessary electrochemical reaction and the product RuOx may also act as a poison of the active Pt sites. Thus, deposition of Ru oxide species in the cathode caused by Ru crossover is suggested to be more detrimental than Pt loss to the durability of DMFC. In addition, lateral distributions of crossed ruthenium shown in Figure 5 are also somewhat interesting although they are microscale images. It seems that most of ruthenium clusters at the cathode surface existed randomly in contact with Pt region. In the middle parts of the cathode catalyst layer, the ruthenium transferred from the anode side through the ion channel of the Nafion membrane seemed to be accumulated around the Pt agglomerates. This supports the movement of positively charged ruthenium species because the Ru deposition can easily occur by electrons conducted through the Pt particles inside the catalyst layer. Conclusions The electronic and chemical changes in the catalyst layers from the degraded MEAs were investigated using XPS and TOF-SIMS combined with a peel-off method. The metallic property in the aged anode surface was decreased with increased performance drop, which was caused by the traveling of metallic cationic components from the anode to the cathode following the electric field through the decomposition of the catalyst particles. The Ru crossover was revealed by the XPS spectra and mass-resolved images of TOF-SIMS. The Ru species was not uniformly distributed within the catalyst layer of the cathode but accumulated more at the interface between the cathode catalyst layer and GDL, indicating the electro-deposition of cationic Ru species at this interface. From the XPS, a peak corresponding to Ru oxides was observed at the cathode surface. Quantitative measurement for the metallic oxide component in the aged cathodes by TOF-SIMS suggested that the decrease in performance was closely correlated to the amount of Ru oxide species in the cathode surface. Supporting Information Available: Available SEM images of the exposed cathode catalyst layer and detached layer on 3M tape after each peel-off. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Carrette, L.; Friedrich, A.; Stimming, U. Fuel Cells 2001, 1, 5. (2) Cleghorn, S.; Kolde, J.; Liu, W. In Handbook of Fuel Cells; Vielstich, W, Gasteiger, H. A., Lamm, A., Eds.; John Wiley & Sons: New York, 2003; Vol. 3, p 566. (3) Inaba, M.; Yamada, H.; Tokunaga, J.; Tasaka, A. Electrochem. Solid-State Lett. 2004, 7, A474. (4) Kinumoto, T.; Inaba, M.; Nakayama, Y.; Ogata, K.; Umebayashi, R.; Tasaka, A.; Iriyama, Y.; Abe, T.; Ogumi, Z. J. Power Sources 2006, 158, 1222. (5) Blom, D. A.; Dunlap, J. R.; Nolan, T. A.; Allard, L. F. J. Electrochem. Soc. 2003, 150, A414. (6) Jeon, M. K.; Lee, K. R.; Oh, K. S.; Hong, D. S.; Won, J. Y.; Li, S.; Woo. S. I. J. Power Sources 2006, 158, 1344. (7) Kim, H.; Shin, S.-J.; Park, Y.-G.; Song, J.; Kim, H.-T. J. Power Sources 2006, 160, 440. (8) Eickes, C.; Piela, P.; Davey, J.; Zelenay, P. J. Electrochem. Soc. 2006, 153, A171. (9) Piela, P.; Eickes, C.; Brosha, E.; Garzon, F.; Zelenay, P. J. Electrochem. Soc. 2004, 151, A2053. (10) Chen, W.; Sun, G.; Guo, J.; Zhao, X.; Yan, S.; Tian, J.; Tang, S.; Zhou, Z.; Xin, Q. Electrochim. Acta. 2006, 51, 2391.

318 J. Phys. Chem. C, Vol. 112, No. 1, 2008 (11) Park, G. -S.; Pak, C.; Chung, Y.; Kim, J. R.; Jun, W. S.; Lee, Y. H.; Kim, K.; Chang, H.; Seung, D. J. Power Sources doi:10.1016/ j.jpowsour.2007.08.068. (12) Niemantsverdriet, J. W. Spectroscopy in Catalysis, an Introduction; Wiley-VCH: Weinheim, 1995. (13) Kim, H.; de Moraes, I. R.; Tremiliosi, G.; Haasch, R.; Wieckowski, A. Surf. Sci. 2001, 474, L203. (14) Park, K.-W.; Sung, Y.-E. J. Phys. Chem. B 2005, 109, 13585. (15) ToF-SIMS: Surface Analysis by Mass Spectrometry; Vickerman, J. C., Briggs, D., Eds.; IM Publications and SurfaceSpectra Limited, 2001. (16) Borg, H. J.; Reijerse, J. F. C.-J. M.; van Santen, R. A.; Niemantsverdriet, J. W. J. Chem. Phys. 1994, 101, 10052. (17) van Hardeveld, R. M.; van Santen, R. A.; Niemantsverdriet, J. W. J. Phys. Chem. B 1997, 101, 998. (18) van Hardeveld, R. M.; van Santen, R. A.; Niemantsverdriet, J. W. J. Phys. Chem. B 1997, 101, 7901. (19) Norrman, K.; Vels Hansen, K.; Mogensen, M. J. Eur. Ceram. Soc. 2006, 26, 967. (20) Ota, H.; Akai, T.; Namita, H.; Yamaguchi, S.; Nomura, M. J. Power Sources 2003, 119-121, 567. (21) Pourbaix, M. Atlas of Electrochemical Equilibria in Aqueous Solutions; National Association of Corrosion Engineers: Houston, TX, 1974.

Chung et al. (22) Waller, F. J.; Scoyoc, R. W. V. CHEMTECH 1987, 17, 438. (23) Ercolessi, F.; Andreoni, W.; Tosatti, E. Phys. ReV. Lett. 1991, 66, 911. (24) Lai, S. L.; Guo, J. Y.; Petrova, V.; Ramanath, G.; Allen, L. H. Phys. ReV. Lett. 1996, 77, 99. (25) Lee, Y. J.; Lee, E.-K.; Kim, S. Phys. ReV. Lett. 2001, 86, 999. (26) Meulenkamp, E. A. J. Phys. Chem. B 1998, 102, 7764. (27) Whetten, R. L. Mater. Sci. Eng. 1993, B19, 8. (28) Kluth, P.; Johannessen, B.; Foran, G. J.; Cookson, D. J.; Kluth, S. M.; Ridgway, M. C. Phys. ReV. B 2006, 74, 014202. (29) Jung, Y.; Kim, S.; Park, S. -J.; Kim, J. M. Colloids Surf., A; doi: 10.1016/j.colsurfa.2007.04.088. (30) Diaz-Morales, R. R.; Liu, R.; Fachini, E.; Chen, G.; Segre, C. U.; Martinez, A.; Cabrera, C. and Smotkin, E. S. J. Electrochem. Soc. 2004, 151, A1314. (31) Guilminot, E.; Corcella, A.; Charlot, F.; Maillard, F.; Chatenet, M. J. Electrochem. Soc. 2007, 154, B96. (32) Bi, W.; Gray, G. E.; Fuller, T. F. Electrochem. Solid-State Lett. 2007, 10, B101. (33) Liu, R.; Iddir, H.; Fan, Q.; Hou, G.; Bo, A.; Ley, K. L.; Smotkin, E. S.; Sung, Y.-E.; Kim, H.; Thomas, S.; Wieckowski, A. J. Phys. Chem. B 2000, 104, 3518.