Journal of Electroanalytical Chemistry 716 (2014) 58–62

Contents lists available at ScienceDirect

Journal of Electroanalytical Chemistry

journal homepage: www.elsevier .com/locate / je lechem

Structural effects on the oxygen reduction reaction on the high indexplanes of Pt3Ni: n(111)–(111) and n(111)–(100) surfaces

1572-6657/$ - see front matter � 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.jelechem.2013.12.008

⇑ Corresponding author. Tel./fax: +81 43 290 3384.E-mail address: hoshi@faculty.chiba-u.jp (N. Hoshi).

Takeshi Rurigaki, Aya Hitotsuyanagi, Masashi Nakamura, Nanami Sakai, Nagahiro Hoshi ⇑Department of Applied Chemistry and Biotechnology, Graduate School of Engineering, Chiba University, 1-33, Yayoi-cho, Inage-ku, Chiba 263-8522, Japan

a r t i c l e i n f o a b s t r a c t

Article history:Available online 14 December 2013

Keywords:High index planesTerraceStepAlloyFuel cell

Active sites for the oxygen reduction reaction (ORR) have been studied on n(111)–(111) and n(111)–(100) surfaces of Pt3Ni in 0.1 M HClO4 using rotating disk electrode (RDE). The activity for the ORR isdecreased with the increase of the step atom density on n(111)–(111) surfaces. This fact is completelyopposite to that of the same series of Pt on which the activity is enhanced at higher step atom density.The ORR gives the highest activity on n(111)–(100) surfaces of Pt3Ni with terrace atomic rows n = 3and n = 5. These results show that the activity for the ORR of (100) step is superior to that of (111) stepon Pt3Ni electrodes.

� 2013 Elsevier B.V. All rights reserved.

1. Introduction

Pt is used as electrocatalysts of fuel cells. Reduction of Pt load-ing in the electrocatalysts is one of the most important subjects forwidespread of fuel cells, because the amount of natural resourcesof Pt is limited. The oxygen reduction reaction (ORR) is the reactionat the cathodes of fuel cells. The high overpotential of the ORR re-sults in the high consumption of Pt in the cathodes. Developmentof electrocatalysts with high activity for the oxygen reduction(ORR) is necessary for the reduction of Pt loading in fuel cells.

It is well known that reaction rate and selectivity of an electro-chemical reaction depend on the surface structure of the electroderemarkably [1–4]. One of the strategies for the enhancement of theORR activity is the control of the surface structures of the electrode.Activity for the ORR depends on the crystal orientation on the lowindex planes of Pt: Pt(100) < Pt(111) < Pt(110) in 0.1 M HClO4 [5].On the other hand, the activity series on the low index planes of Pdis opposite to that on Pt: Pd(110) < Pd(111) < Pd(100) in 0.1 MHClO4 [6]. However, studies using the low index planes are notadequate to elucidate the active site for electrochemical reactionsprecisely. Structure of surfaces vicinal to the low index planes(high index planes) can be modified systematically. The high indexplanes can determine the structure of the active site for the ORR inatomic scale.

The study on the ORR was extended to the high index planesof Pt [7–10] and Pd [6,11]. A high index plane is notated asn(hkl)–(h0 k0 l0), in which n, (hkl) and (h0 k0 l0) show the number ofterrace atomic rows, terrace and step structures, respectively. On

n(111)–(111) and n(111)–(100) surfaces of Pt, the activity forthe ORR increases with the increase of the step atom density dS ex-cept Pt(110) = 2(111)–(111) and Pt(311) = 2(111)–(100) in0.1 M HClO4 [9,10]. On n(100)–(111) and n(100)–(110) surfacesof Pt, however, the activity does not depend on the surface struc-tures [10]. The activity on the surfaces with (111) terrace is higherthan that with (100) terrace. The structural analysis supports thatthe active site of the ORR is located between the (111) terrace edgeand the (111) terrace atomic rows neighboring to the edge on Ptelectrodes [10]. On Pd electrodes, however, the active site for theORR is (100) terrace [6,11]. Difference of the step structure doesnot affect the ORR activity on Pd electrodes at all.

An alloy catalyst has higher activity than a pure metal. Pt alloysincluding transition metals such as Fe, Ni, and Co enhance the car-bon monoxide tolerance at the anode [12,13] as well as the ORRactivity at the cathode [14–16]. Stamenkovic et al. evaluated theactivity for the ORR on the low index planes of Pt3Ni, and foundthat the activity on Pt3Ni(111) electrode is 10 times as high as thatof Pt(111) [16]. Monolayer of Pt skin is formed on the low indexplanes of Pt3Ni. Wadayama et al. prepared Pt enriched Ni/Pt(111) [17] and Co/Pt(111) [18] surfaces using molecular beamepitaxy (MBE) method. The topmost layers of these surfaces arealso composed of monolayer of Pt. The ORR activities of Pt enrichedNi/Pt(111) and Co/Pt(111) are 8 and 10 times higher than that ofclean Pt(111), respectively. The activity for the ORR is enhanced as(100) < (110) < (111) on Pt3Ni(hkl) alloy and Pt enriched Co/Pt(hkl).

Studies on the structural effects on the ORR have been limitedto the low index planes of Pt alloy electrodes. In this paper, we ex-tend the study to the high index planes of Pt3Ni to elucidate the ac-tive site for the ORR. Our previous study shows that the surfaces

T. Rurigaki et al. / Journal of Electroanalytical Chemistry 716 (2014) 58–62 59

with (111) terrace have higher activity for the ORR in the high in-dex planes of Pt. Therefore, we first study the orientation depen-dence of the ORR on n(111)–(111) and n(111)–(100) series ofPt3Ni (n = 2, 3, 5, 9 and1). The hard sphere models of the surfacesexamined are shown in Fig. 1.

Fig. 2. Laue back reflection pattern of prepared Pt3Ni(111).

2. Experimental

A single crystal bead of Pt3Ni (about 3 mm in diameter) wasprepared according to the following procedure. A Pt3Ni wire (Tana-ka Kikinzoku, 1 mmu) was set in a quartz tube. The end of the wirewas melted in Ar atmosphere (99.9999%) using induction furnace(HOTSHOT 3.5, Alonics) for prevention of the oxidation of thedoped Ni.

No (111) and (100) facets appeared on the surface of the bead.The single crystal bead was oriented with the use of X-ray Laueback reflection method, and then mechanically polished with dia-mond slurries.

The polished crystal was annealed in Ar(95%)/H2(5%) atmo-sphere about 1300 K using induction furnace for the removal ofthe distortions due to the mechanical polishing, and then cooleddown to room temperature. The annealed surface was protectedwith ultrapure water saturated with Ar(95%)/H2(5%), and trans-ferred to an electrochemical cell.

Electrolytic solutions were prepared from ultrapure water trea-ted with Milli-Q Advantage A10 (Millipore) and suprapure gradechemicals (Merck). Voltammograms of the ORR were measuredin the hanging meniscus RDE configuration [7,8,19,20] using elec-trochemical analyzer (ALS 701C) and rotating ring disk electrodeapparatus (BAS:RRDE-3).

Potential was scanned from 0.05 V (RHE) to the positive direc-tion up to 1.0 V (RHE) at scanning rate 0.010 V s�1 and rotation rate1600 rpm. Current was divided by geometric surface area to obtaincurrent density, because a single crystal surfaces is atomically flat,and Pt skin is expected to be formed on the high index planes ofPt3Ni as is the case of the low index planes [16]. The specific activ-ity for the ORR is estimated using the kinetic current density at0.90 V (RHE) jk,0.90V according to the following equation [21,22]:

1j¼ 1

jkþ 1

jL;

where j and jL are total current density and limiting current density,respectively.

3. Results and discussion

Fig. 2 shows Laue back reflection pattern of Pt3Ni(111) pre-pared by the method mentioned above. The Laue patter has hexag-onal symmetry characteristic of a (111) surface of fcc structure.This result indicates the high quality of the single crystal electrode.

Fig. 3 shows voltammograms of n(111)–(111) and n(111)–(100) surfaces of Pt3Ni in 0.1 M HClO4 saturated with Ar. On

Fig. 1. Hard sphere models of the high index planes of Pt3Ni examined (top view):(a) n(111)–(111) and (b) n(111)–(100) surfaces.

n(111)–(111) surfaces, the redox peaks at 0.17 V (RHE) increasewith the decrease of the terrace width n (namely the increase of dS),reaching maximum on Pt3Ni(110). The notation of n(111)–(111)can be also written as (n � 1)(111)–(110); the redox peaks at0.17 V (RHE) are due to the adsorption and the desorption ofhydrogen at (110) structure. However, structural effects on thevoltammogram are less remarkable than those on the high indexplanes of Pt [10,23,24].

On n(111)–(100) surfaces, redox shoulders are found at 0.27 V.These shoulders can be due to the desorption and the adsorption ofhydrogen at (100) step according to the voltammograms of Pt[10,23,25]. However, the peak intensities show smaller structuraleffects than those on n(111)–(111) surfaces of Pt3Ni. We preparedseveral high index planes of Pt3Ni with the same orientation, andall the voltammograms gave the same tendency. These facts mayindicate that perfect Pt skin may not be formed on the high indexplanes of Pt3Ni.

Table 1 shows anodic charges of the adsorbed hydrogen be-tween 0.05 and 0.40 V (RHE) of Pt3Ni and Pt electrodes after thecorrection of the double layer charges. Theoretical charges of purePt electrodes are also shown in the table on the assumption thatone hydrogen atom is adsorbed to one Pt atom of the topmostlayer. The charges of pure Pt electrodes are lower than those of the-oretical values in 0.1 M HClO4. This tendency is general on Pt elec-trodes in 0.1 M HClO4, for example, the charge of the hydrogendesorption is only 2/3 of the theoretical value on Pt(111). The ob-served charge on Pt3Ni(111) is less than that on Pt(111) as re-ported before [16], although perfect Pt skin is formed onPt3Ni(111). The difference between Pt3Ni and Pt gets smaller withthe decrease of n (namely increase of dS) on n(111)–(111) sur-faces. On n(111)–(100) surfaces, the difference between Pt3Niand Pt is smaller on the surfaces with larger terrace (n = 5 and 9),but the difference gets larger on the surfaces with narrower terrace(n = 2 and 3).

Fig. 4 shows ORR voltammograms on n(111)–(111) andn(111)–(100) surfaces of Pt3Ni. The ORR currents depend on thesurface structure between 0.7 and 1.0 V, at which the ORR is kinet-ically controlled. The ORR currents show no structural effect be-tween 0.3 and 0.7 V, because the reaction is diffusion controlledat this potential range. The ORR currents shows smaller structuraleffects on n(111)–(111) surfaces of Pt3Ni in the kinetically con-trolled region. The currents depend on surface structure remark-ably on n(111)–(100) surfaces in the kinetically controlledregion, although the voltammograms show little structural effectsin Ar saturated solution (Fig. 3(b)).

The limiting currents drop in the adsorbed hydrogen region onthe high index planes of Pt as shown in Fig. 4(c) and (d). The ad-sorbed hydrogen atoms hinder the 4 electron reduction of O2, pro-ducing H2O2 via 2 electron reduction [26]. The hindrance of the 4

Fig. 3. Voltammograms in 0.1 M HClO4 saturated with Ar. Scanning rate: 0.050 V s�1: (a) n(111)–(111) and (b) n(111)–(100) surfaces of Pt3Ni.

Table 1Anodic charge in the adsorbed hydrogen between 0.05 and 0.40 V (RHE) after thecorrection of the double layer charge.

QH/lC cm�2

Pt3Ni (observed) Pt (observed) Pt (theoretical)

Single crystal electrodes: n(111)–(111)(111) 86.5 129 239(997) = 9(111)–(111) 90.0 144 228(553) = 5(111)–(111) 141 147 216(331) = 3(111)–(111) 140 165 190(110) = 2(111)–(111) 176 183 220

Single crystal electrodes: n(111)–(100)(111) 86.5 129 239(544) = 9(111)–(100) 131 145 247(322) = 5(111)–(100) 180 162 251(211) = 3(111)–(100) 126 194 254(311) = 2(111)–(100) 130 237 250

60 T. Rurigaki et al. / Journal of Electroanalytical Chemistry 716 (2014) 58–62

electron reduction on Pt3Ni(111) electrode is lower than that onPt(111) electrode as reported previously [16], whereas the hin-drance on n(111)–(111) surfaces of Pt3Ni is as low as that of Pt.On n(111)–(100) series, however, the hindrance on Pt3Ni elec-trodes are less remarkable than that on Pt. H2O2 damages the poly-mer electrolyte of fuel cells. Pt3Ni produces less H2O2 than Pt; Pt3Niis better catalyst for the ORR than Pt.

Fig. 5 shows the value of jk,0.90V plotted against dS. Calculationmethod of dS is described in detail in previous reports[9,10,24,25], however, we briefly summarize the method belowby taking n(111)–(111) surface as an example. Fig. 6(a) and (b)show hard sphere models of n(111)–(111) surface. The area sur-rounded by a yellow rectangle is regarded as a ‘‘unit cell’’ inFig. 6(a). The number of step atoms is one in the unit cell, and thatof the terrace atoms is (n � 2) on the surfaces with n terrace atomic

rows. The area of the unit cell S is calculated according to Eq. (1)[10,24].

S ¼ n� 23

� ��

ffiffiffi3p

2dPt � dPt ¼ n� 2

3

� ��

ffiffiffi3p

2d2

Pt

" #; ð1Þ

where dPt is the diameter of Pt atom. When the angle between theterrace and the tangent to the high index plane (dotted line inFig. 5(b)) is h, the value of dS is obtained according to Eq. (2).

dS ¼cos h

S: ð2Þ

Previous paper reports that Pt3Ni(110) surface is composed of amixture of unreconstructed (1 � 1) and reconstructed (1 � 2)structures [16]. Thus we roughly assume that Pt3Ni(110) is com-posed of (1 � 2) structure at 0.90 V. Surface X-ray scattering(SXS) study verifies that Pt(311) = 2(111)–(100) is reconstructedto (1 � 2) structure [27]. There has been, however, no report onthe surface structure of Pt3Ni(311); Pt3Ni(311) surface is assumedto be (1 � 1) structure. We assume that the other surfaces of Pt3Nihave unreconstructed (1 � 1) structure, because the high indexplanes of Pt with 3 6 n has unreconstructed (1 � 1) structure inelectrochemical environments according to the studies of SXS[28,29].

The ORR activity of Pt3Ni is decreased with the increase of dS onn(111)–(111) series, as shown in Fig. 5(a). This tendency is com-pletely opposite to the structural effects on n(111)–(111) surfacesof Pt on which the ORR activity is increased with the increase of dS

except Pt(110) = 2(111)–(111) [10]. The activity of Pt3Ni is higherthan that of Pt except (331). The ORR activity on Pt3Ni(331) islower than that on Pt(331), although it is alloyed. Previous studyshows that (111) step enhances the ORR activity on Pt electrodes[10]. On Pt3Ni alloy electrodes, however, (111) step deactivates

Fig. 4. ORR voltammograms on Pt3Ni electrodes in 0.1 M HClO4: (a) n(111)–(111) and (b) n(111)–(100) series. The results on (c) n(111)–(111) and (d) n(111)–(100) seriesof Pt are reproduced for comparison [10]. Scanning rate: 0.010 V s�1. Rotation rate: 1600 rpm.

Fig. 5. Kinetic current density jk of Pt (triangles) and Pt3Ni (squares) plotted against the step atom density dS: (a) n(111)–(111) and (b) n(111)–(100) surfaces.

T. Rurigaki et al. / Journal of Electroanalytical Chemistry 716 (2014) 58–62 61

the ORR in 0.1 M HClO4. Larger (111) terrace is appropriate for theORR on n(111)–(111) surfaces of Pt3Ni.

The n(111)–(100) surfaces of Pt3Ni give maxima of the ORRactivity on Pt3Ni(322) = 5(111)–(100) and Pt3Ni(211) = 3(111)–(100) as shown in Fig. 5(b). The activity is 1.4 time as high as thaton Pt3Ni(111) that has the highest activity in the low index planes.These results show that n(111)–(100) surfaces of Pt3Ni have themost active sites on the surfaces with terrace atomic rows n be-tween 3 and 5.

Pt3Ni(533) = 4(111)–(100) is expected to have the highestactivity in n(111)–(100) surfaces. We prepared several Pt3-

Ni(533) = 4(111)–(100) electrodes. However, no reproducible re-sult is obtained in voltammograms in Ar saturated solution andthe ORR activity. Surface structure of Pt3Ni(533) = 4(111)–(100)may be unstable.

Pt3Ni(311) = 2(111)–(100) has the lowest activity in the highindex planes examined, although it has the highest dS assuming(1 � 1) structure. The notation of (311) = 2(111)–(100) can be

written as (311) = 2(100)–(111); Pt3Ni(311) also belongs to thesurfaces with (100) terrace. Pt3Ni(100) has the lowest activity inthe low index planes of Pt3Ni [16]. We also measured the ORRactivity on the high index planes of Pt3Ni with (100) terrace, andfound that the ORR activity on the surfaces with (100) terrace islower than that with (111) terrace. The low activity of (100) ter-race may govern the activity of Pt3Ni(311) electrode.

All the Pt3Ni electrodes in n(111)–(100) series have higheractivity than those of Pt in the same series (Fig. 5(b)). Structural ef-fects on the ORR on Pt3Ni electrodes are similar to those on Pt elec-trodes in n(111)–(100) series: the activities are high on thesurfaces with n = 3 and n = 5, and the surfaces with n = 2 havethe lowest activity. This tendency differs from that on n(111)–(111) surfaces of Pt3Ni, of which structural effects are completelyopposite to those of Pt.

Shift of d-band center can explain the higher activity for theORR on Pt3Ni compared with Pt electrodes [30–32] and the positiveshift of the oxide film formation [16]. Pt3Ni(553) = 5(111)–(111)

Fig. 6. Hard sphere model of n(111)–(111) series of Pt3Ni surface. Spheres paintedin pink show step atoms. Yellow rectangular is regarded as a ‘‘unit cell’’ for thecalculation of the density of step atoms dS. (For interpretation of the references tocolour in this figure legend, the reader is referred to the web version of this article.)

62 T. Rurigaki et al. / Journal of Electroanalytical Chemistry 716 (2014) 58–62

has higher activity than Pt3Ni(110) = 2(111)–(111), and the activ-ity on Pt3Ni(322) = 5(111)–(100) is 3.6 times as high as that onPt3Ni(311) = 2(111)–(100) in Fig. 4. Voltammograms in Fig. 3show, however, that total anodic charges of the oxide film forma-tion on Pt3Ni(553) and Pt3Ni(322) are larger than those on Pt3-

Ni(110) and Pt3Ni(311) at 0.90 V (RHE). These facts indicate thatthe total coverage of the oxide film does not affect the ORR activityon the high index planes of Pt3Ni. Coverage of the oxide film at thestep and the terrace must be measured for more detailed discus-sion of the correlation between the oxide film and the ORR activity.However, we cannot distinguish the oxide film formation at thestep from that at the terrace in voltammograms of Fig. 3. Spectro-scopic study is necessary for the elucidation of the adsorption sitesof the oxide.

Real surface structures were determined on the low indexplanes of Pt3Ni using SXS, whereas there has been no SXS studyon the high index planes of Pt3Ni in electrochemical environments.There is no guarantee that perfect Pt skins are formed on the highindex planes of Pt3Ni, and we have no information on the ratio ofPt/Ni in the inner layers of the high index planes. Calculation usingthe density functional theory (DFT) predicts that the ORR activity isenhanced with the increase of Ni ratio at second layer of PtNi alloy[33]. Information on the inner layers and real surface structures isnecessary for the revelation of the origin of the structural effects onPt3Ni electrodes. Vibrational study of the oxide and SXS study ofthe surface structures are now on progress in our laboratory.

4. Conclusions

Specific activity for the ORR decreases with the increase of thestep atom density on n(111)–(111) series of Pt3Ni: Pt3Ni(331)

n = 3 < Pt3Ni(110) n = 2 < Pt3Ni(553) n = 5 < Pt3Ni(997) n = 9 < Pt3-

Ni(111) n =1. The activity on Pt3Ni(331) n = 3 is lower than thaton Pt(331) n = 3. On n(111)–(100) series of Pt3Ni, following orderof the activity is obtained: Pt3Ni(311) n = 2 < Pt3Ni(544)n = 9 < Pt3Ni(111) n =1 < Pt3Ni(211) n = 3 � Pt3Ni(322) n = 5.Pt3Ni(211) = 3(111)–(100) and Pt3Ni(322) = 5(111)–(100) havethe highest activity for the ORR in the high index planes ofPt3Ni examined. The specific activities for the ORR onPt3Ni(211) = 3(111)–(100) and Pt3Ni(322) = 5(111)–(100) are1.4 times as high as that on Pt3Ni(111) n =1.

Acknowledgement

This work was supported by New Energy DevelopmentOrganization.

References

[1] R. Adzic, in: R.E. White, J.O’M. Bockris, B.E. Conway (Eds.), Reaction kineticsand mechanisms on metal single-crystal electrode surfaces, Modern Aspects ofElectrochemistry, vol. 21, Springer, Berlin, 1990, pp. 163–236 (Chapter 5).

[2] C. Lamy, J.M. Leger, J. Chim. Phys. 88 (1991) 1649–1671.[3] N.M. Markovic, P.N. Ross Jr., Surf. Sci. Rep. 45 (2002) 117–229.[4] S. Ye, T. Kondo, N. Hoshi, J. Inukai, S. Yoshimoto, M. Osawa, K. Itaya,

Electrochemistry 77 (2009) 2–20.[5] N.M. Markovic, R.R. Adzic, B.D. Cahan, E.B. Yeager, J. Electroanal. Chem. 377

(1994) 249–259.[6] S. Kondo, M. Nakamura, N. Maki, N. Hoshi, Active sites for the oxygen reduction

reaction on the low and high index planes of palladium, J. Phys. Chem. C 113(2009) 12625–12628.

[7] M.D. Macia, J.M. Campina, E. Herrero, J.M. Feliu, J. Electroanal. Chem. 564(2004) 141–150.

[8] A. Kuzume, E. Herrero, J.M. Feliu, J. Electroanal. Chem. 599 (2007) 333–343.[9] A. Hitotsuyanagi, M. Nakamura, N. Hoshi, Electrochim. Acta 82 (2012) 512–

516.[10] N. Hoshi, M. Nakamura, A. Hitotsuyanagi, Electrochim. Acta. <http://

dx.doi.org/10.1016/j.electacta.2013.05.045>.[11] A. Hitotsuyanagi, S. Kondo, M. Nakamura, N. Hoshi, J. Electroanal. Chem. 657

(2011) 123–127.[12] F. Maillard, G.Q. Lu, A. Wieckowski, U. Stimming, J. Phys. Chem. B 109 (2005)

16230–16243.[13] M.J. Ball, C.A. Lucas, N.M. Markovic, V. Stamenkovic, P.N. Ross, Surf. Sci. 540

(2003) 295–302.[14] T. Toda, H. Igarashi, H. Uchida, M. Watanabe, J. Electrochem. Soc. 146 (1999)

3750–3756.[15] V.R. Stamenkovic, B.S. Mun, M. Arenz, K.J.J. Mayrhofer, C.A. Lucas, G. Wang, P.N.

Ross, N.M. Markovic, Nat. Mater. 6 (2007) 241–247.[16] V.R. Stamenkovic, B. Fowler, B.S. Mun, G. Wang, P.N. Ross, C.A. Lucas, N.M.

Markovic, Science 315 (2007) 493–497.[17] T. Wadayama, N. Todoroki, Y. Yamada, T. Sugawara, K. Miyamotop, Y. Iijima,

Electrochem. Commun. 12 (2010) 1112–1115.[18] T. Wadayama, H. Yoshida, K. Ogawa, N. Todoroki, Y. Yamada, K. Miyamoto, Y.

Iijima, T. Sugawara, K. Arihara, S. Sugawara, K. Shinohara, J. Phys. Chem. C 115(2011) 18589–18596.

[19] B.D. Cahan, H.M. Villulas, J. Electroanal. Chem. 307 (1991) 263–268.[20] T. Abe, G.M. Swain, K. Sashikata, K. Itaya, J. Electroanal. Chem. 382 (1995) 73–

83.[21] H.M. Villullas, M. Lopez Teijelo, J. Electroanal. Chem. 384 (1995) 25–30.[22] H.M. Villullas, M. Lopez Teijelo, J. Electroanal. Chem. 385 (1995) 39–44.[23] N. Furuya, S. Koide, Surf. Sci. 220 (1989) 18–28.[24] J. Clavilier, K. El Achi, A. Rodes, J. Electroanal. Chem. 272 (1989) 253–261.[25] A. Rodes, K. El Achi, M.A. Zamakhchari, J. Clavilier, J. Electroanal. Chem. 284

(1990) 245–253.[26] N.M. Markovic, H.A. Gasteiger, P.N. Ross, J. Phys. Chem. 99 (1995) 3411–3415.[27] A. Nakahara, M. Nakamura, K. Sumitani, O. Sakata, N. Hoshi, Langmuir 23

(2007) 10879–10882.[28] N. Hoshi, A. Nakahara, M. Nakamura, K. Sumitani1, O. Sakata, Electrochim. Acta

53 (2008) 6070–6075.[29] N. Hoshi, M. Nakamura, O. Sakata, A. Nakahara, K. Naito, H. Ogata, Langmuir 27

(2011) 4236–4242.[30] B. Hammer, J.K. Nørskov, Surf. Sci. 343 (1995) 211–220.[31] J. Greeley, J.K. Nørskov, M. Mavrikakis, Ann. Rev. Phys. Chem. 53 (2002) 319–

348.[32] V. Stamenkovic, B.S. Mun, K.J.J. Mayrhofer, P.N. Ross, N.M. Markovic, J.

Rossmeisl, J. Greeley, J.K. Nørskov, Angew. Chem. Int. Ed. 45 (2006) 2897–2901.

[33] I. Matanovic, F.H. Garzon, N.J. Henson, J. Phys. Chem. C 115 (2011) 10640–10650.