lable at ScienceDirect

Journal of Power Sources 294 (2015) 299e304

Contents lists avai

Journal of Power Sources

journal homepage: www.elsevier .com/locate/ jpowsour

Comparative assessment of synthetic strategies toward activeplatinumerhodiumetin electrocatalysts for efficient ethanolelectro-oxidation

Nina Erini a, Paul Krause a, Manuel Gliech a, Ruizhi Yang b, Yunhui Huang c,Peter Strasser a, d, *

a Department of Chemistry, Technical University Berlin, Straße des 17. Juni 124, 10623 Berlin, Germanyb School of Energy, Soochow University, No. 1 Shizi Street, Suzhou, Jiangsu 215006, Chinac School of Materials Science and Engineering, Huazhong University of Science and Technology, 1037 Luoyu Road, Wuhan 430074, Chinad Ertl Center for Electrochemistry and Catalysis, Gwangju Institute of Science and Technology, Gwangju 500-712, South Korea

h i g h l i g h t s

* Corresponding author. Department of ChemistryStraße des 17. Juni 124, 10623 Berlin, Germany.

E-mail address: pstrasser@tu-berlin.de (P. Strasser

http://dx.doi.org/10.1016/j.jpowsour.2015.06.0420378-7753/© 2015 Elsevier B.V. All rights reserved.

g r a p h i c a l a b s t r a c t

� The electrooxidation of EtOH onPtRhSn/C catalysts was investigatedin acidic media.

� Small metal nanoparticles with sizesin the range of 6e12 nm weresynthesized.

� Variation of pressure and carbonsupporting conditions during polyolsynthesis.

� Ambient pressure conditions result incatalysts with higher EOR massactivities.

a r t i c l e i n f o

Article history:Received 16 March 2015Received in revised form21 May 2015Accepted 9 June 2015Available online xxx

Keywords:Ethanol oxidation reactionFuel cellsElectrochemistryMicrowave assisted synthesisElectrocatalysts

a b s t r a c t

The present work explores the effect of autoclave-based autogenous-pressure vs. ambient pressureconditions on the synthesis and properties of carbon-supported PteRheSn nanoparticle electrocatalysts.The PteRheSn nanoparticles were characterized by X-ray spectroscopy, electron microscopy and massspectroscopy and deployed as catalysts for the electrocatalytic ethanol oxidation reaction. PteRheSncatalysts precipitated with carbon already present showed narrow particle size distribution around 7 nm,while catalysts supported on carbon after particle formation showed broader size distribution rangingfrom 8 to 16 nm, similar metal loadings between 40 and 48 wt.% and similar atomic ratios of Pt:Rh:Sn of30:10:60. The highest ethanol oxidation activity at low overpotentials associated with exceptionally earlyethanol oxidation onset potential was observed for ambient-pressure catalysts with the active ternaryalloy phase formed in presence of the carbon supports. In contrast, catalysts prepared under ambientpressure in a two-step approach, involving alloy particle formation followed by particle separation andsubsequent deposition on the carbon support, yielded the highest overall mass activities. Based on theobserved synthesiseactivity correlations, a comparative assessment is provided of the synthetic tech-niques at high vs. low pressures, and in presence and absence of carbon support. Plausible hypotheses in

, Technical University Berlin,

).

N. Erini et al. / Journal of Power Sources 294 (2015) 299e304300

terms of particle dispersion and interparticle distance accounting for these observed differences arediscussed.

© 2015 Elsevier B.V. All rights reserved.

1. Introduction

Platinum is a commonly used anodic material in acidic lowtemperature fuel cells. Since alcohol oxidation on pure platinumdoesn't reach the desired activities, research in the field of DirectEthanol Fuel Cells (DEFC) has focusedwidely on the development ofbinary and ternary Pt-based alloys [1e3]. The introduction ofternary electrocatalytical systems for the ethanol oxidation reac-tion (EOR) in recent research efforts has brought the developmentof DEFC as alternate power sources a big step forward [4,5]. Ethanolis of a particular interest for mobile applications such as electricvehicles, due to high energy density 8 kWh kg�1, low toxicity,biocompatibility and abundant availability. It is, however, not easilyoxidized completely to CO2 and water. This is due to difficulties inthe CeC bond cleavage in ethanol and the reaction may involveseveral different mechanism pathways with the formation of a highnumber of reaction intermediates such as CHx species or acetal-dehyde and, to some extent, to the formation of CO-intermediatesleading to poisoning of the active sites on Pt catalysts [3,6e11].Efforts to develop highly active and selective EOR electrocatalystshave therefore concentrated on the addition of co-catalysts toplatinum [12e16,34�38].

Our previous research focused on the promising family of EORnanocatalysts based on mixtures of Pt, Rh and Sn [3,5,17e23]. In arecent comprehensive study on a set of PteRheSnO2 nanoparticlecatalysts an optimal PteRheSn atomic ratio of 3:1:4 has beenproposed [5]. In our previous work we addressed the optimalstructural arrangement of the atoms of the three components in thesurface and bulk of the final active catalyst. On its surface, metallicPt and Rh are atomically mixed with Sn, giving rise to active-surface-site ensembles. Our aim was to maximize activity andselectivity and find a single-phase Rh-doped PteSn Niggliitestructure as the preferred and catalytically most active nano-crystalline phase [22]. Synthesis routes to nanoparticle EOR cata-lysts containing Platinum, Rhodium and Sn range fromimpregnation-reductions methods [24] to deposition of metalatoms on oxide surfaces followed by galvanic displacement [3].Weimplemented in our work amodified polyol method in dioctylethersolvent [25], controlling the temperature during the reactionwith aheating mantel. This approach yields ternary single phased catalystwith SnOx next to metallic Pt an Rh on the surface in close prox-imity. Recent reports also claim improved electrocatalytic stabilityand elevated activities for ternary electrocatalysts by microwave-assisted selective deposition of nanoparticles onto carbon [26]and ternary PtSn@Rh/C systems by a two-step microwave-assis-ted polyol method as a promising catalyst preparation method foroptimizing the PteSneRh ternary system for EOR application [27].

In order to compare the two synthesis approaches and clearlyestablish a preferable synthetic approach towards PtRhSncatalysts, we first compared a one-pot reduction of metal pre-cursors at ambient pressures both in the presence and absence ofcarbon support (referred to as “ap-PtRhSnþC” and “ap-PtRhSn/C”;ap materials). Synthesis conditions followed our previously usedpolyol method under a temperature control using a standardlaboratory heating mantel device. Thereafter, we compared twovariations of the two synthesis routes involving microwave-assisted temperature control in an autoclave associated with

autogenous overpressure conditions (referred to as “op-PtRhSnþC” and “op-PtRhSn/C; op materials).

2. Experimental section

2.1. Catalyst preparation

All electrocatalysts (40 wt.% of metal loadings, Pt:Rh:Sn atomicratios of 3:1:4) were prepared using Pt(acac)2, Rh2(OAc)4, andSn(acac)2 as metal precursors, 1,2-tetradecandiol, oleic acid andoleylamine in dioctylether as reducing and capping agents, andKetjen Black as support. All precursors were mixed together,including carbon for the direct supported electrocatalysts, heatedup to 260 �C and stirred under reflux at that temperature for30 min. For the heating mantel temperature controlled process thereaction mixture was under ambient pressure conditions, resultingin the “ap-PtRhSnþC” and the “ap-PtRhSn/C” materials. In themicrowave-assisted temperature-controlled process autogenouspressures up to 30 bar and could built up inside the reactionmixture, resulting in the “op-PtRhSnþC” and the “op-PtRhSn/C”materials. Ketjen black carbon was sonicated for 1 h one ice.Then, the unsupported alloy particles (i.e. “ap-PtRhSn/C” and“op-PtRhSn/C”) were precipitated with isopropanol, redispersed inn-hexane and added to the Ketjen black carbon dispersion, soni-cated for another hour on ice, and stirred overnight at RT. All par-ticles were finally separated by centrifugation, freeze-dried, and theresidue dried in a furnace under N2, oxidized under O2/N2, andcalcinated by heating under H2/Ar atmosphere (see detailed in-formation in previous publication [22]).

2.2. Catalyst characterization

Inductively coupled plasma-optical emission spectroscopy (ICP-OES) was used for compositional analysis; performed using a 715-ES-inductively coupled plasma analysis system (Varian). Trans-mission Electron Microscopy (TEM) and Energy dispersive x-rayspectroscopy (EDX) were carried out using a FEI TECNAI G2 20 S-TWIN microscope operated at 200 kV, equipped with a GATANMS794 P CCD-detector to study morphology and composition. Themean particle sizewas determined from TEM images by counting ofat least 50 particles. Cu Ka X-ray diffraction patterns were collectedusing a D8 Advance diffractometer (Bruker) equipped with a LynxEye Detector and KFL Cu 2K X-ray tube. The diffraction patternswere collected in a 20e80� 2q range with a step size of 0.00142�

dwelling for 30 s at every step. The XRD patterns were analyzedusing the MDI Jade 8 software package. Bragg peak positions werecompared with the reference XRD patterns (PDF data files, NationalInstitute of Science and Technology).

2.3. Electrochemical measurements

All electrochemical measurements were carried out in a three-compartment electrochemical glass cell at room temperature us-ing a Biologic SP 150 potentiostat. All potentials reported here aregiven in respect to a reversible hydrogen electrode (RHE). The 0.5 MC2H5OH þ 0.1 M HClO4 electrolyte was deaerated with high-purityN2 before every measurement. During the experiments N2 was

Table 1Molar composition, metal weight loading and crystallite sizes of the supported PtRhSn electrocatalysts.

Catalysts Pt:Rh:Sn atomic ratio (ICP) Total metal loading/wt% Pt:Rh:Sn atomic ratio (EDX) Particle size (TEM)/nm

1) apPtRhSn/C 32:12:56 42 42:7:51 11.3 ± 1.72) apPtRhSnþC 30:12:58 40 44:16:50 6.5 ± 0.83) opPtRhSn/C 33:11:57 48 38:12:50 11.2 ± 2.34) opPtRhSnþC 30:11:59 42 32:6:62 7.1 ± 0.8

N. Erini et al. / Journal of Power Sources 294 (2015) 299e304 301

purged over the electrolyte. A large surface area Pt counter elec-trode was contained in a separate compartment. A saturated mer-curyemercury sulfate electrode (MMS) was inserted in a separatecompartment of the cell via a Luggin capillary for setting desiredoverpotentials. As the working electrode, a polished glassy carbondisc electrode (0.196 cm2 geometrical surface area) was coated with10 ml of a catalyst ink solution. Catalysts inks were prepared byultrasonication of 6 mg of electrocatalyst powder in a mixture of2.0 ml of ultrapure water, 0.5 ml isopropanol, and 20 ml of 5 wt. %Nafion solution for 15 min.

3. Results and discussion

3.1. Structural characterization of the PteRheSn catalysts

The Pt:Rh:Sn atomic ratios obtained by ICP and EDX weresimilar to the intended ratio (Table 1). The local atomic ratios fromEDX measurement differ slightly from the overall ratios in the ICPresults, which can be expected since they represent a more local

Fig. 1. TEM micrographs and particle size distribution diagrams of the as-prepared elecExemplary insets with higher magnification for each sample show different lattice spacing

estimation of the atomic concentration. TEMmicrographs and theircorresponding size distribution histograms of the electrocatalysts(Fig. 1) evidenced that the nanoparticles were well distributedacross the carbon support and largely spherical in shape. The meandiameters for the nanoparticles reduced in the present of carbon(PtRhSnþC) were around 7 nm and showed little agglomeration.The nanoparticles that were supported after reduction (PtRhSn/C)showed stronger agglomeration, while the mean diameter of thenanoparticles is around 11 nm. Furthermore, the size distribution ofmetal particles is broader, for particles supported on carbon afterreduction and precipitation, as pictured in the diagrams in Fig. 1.The higher magnification images, also depicted in Fig. 1, revealdifferent lattice spacing found in all samples and are shown hereexemplary with the four different catalysts: 2.3 Å can be attributedto Pt fcc (111) or Niggliite (101) (Fig. 1 op-PtRhSnþC); 2.4 Å can beattributed to SnO2 rutile (202) (Fig. 1 ap-PtRhSn/C) and 2.6 Å to theSnO2 rutile (101) (Fig. 1 ap-PtRhSnþC) plane, respectively. Thelattice spacing of 3.0 Å can be attributed to the Niggliite (101) plane(Fig. 1 op-PtRhSn/C). The metal loadings ranged in the narrow

trocatalyst: 1) ap-PtRhSn/C, 2) ap-PtRhSnþC, 3) op-PtRhSn/C and 4) op-PtRhSnþC.found in all samples.

Fig. 2. Cu Ka XRD patterns of electrocatalyst: 1) ap-PtRhSn/C, 2) ap-PtRhSnþC,3) op-PtRhSn/C and 4) op-PtRhSnþC. Pure Pt fcc and pure PtSn Niggliite diffractionpatterns are indicated in black and grey, respectively.

Fig. 3. Electrocatalytic activity for the ethanol oxidation reaction (EOR) on1) ap-PtRhSn/C, 2) ap-PtRhSnþC, 3) op-PtRhSn/C and 4) op-PtRhSnþC recorded in0.5 M C2H5OH þ 0.1 M HClO4. A) First forward scan with a scan rate of 20 mV/s. Thevertical dashed line marks the potential of the chronoamperometric experiment inB) Potentiostatic chronoamperometry at E ¼ þ0.45 V vs. RHE. Current densities arenormalized by the amount of Pt present on the working electrode surface (~3 mg Pt).

N. Erini et al. / Journal of Power Sources 294 (2015) 299e304302

region from 40 to 48% by weight.The XRD patterns of the PteRheSn electrocatalysts are depicted

in Fig. 2. All diffractogramms show relatively broad Bragg peaks,typical for nanosized particles with limited structural coherencelengths. All samples showed the typical cubic phased (fcc) re-flections between the respective Pt and Rh literature values [(111):39.764 and 41.069�], evidencing alloy formation and bulk latticecompression compared to pure Pt; as well as more prominenthexagonal phased PtSn (Niggliite) diffraction patterns [(102):41.846�] where Rh statistically substituted Pt in the lattice [22].Both Pt and Rh show a face-centered cubic (fcc) space group Fm3mwith very similar lattice constants of 3.9231 Å and 3.8031 Å,respectively. This favors the formation of homogeneous PtRh binarysolid solutions over wide compositional ranges, as indicated by thePteRh phase diagram [17]. This is why Pt and Rh diffraction peakscannot be separately resolved in the XRD patterns. In general, theap materials showed higher intensities and higher ratios of hex-agonal to cubic diffraction peak intensities than the op materials.The as-reduced supported electrocatalysts materials show incomparison broader diffraction peaks and smaller intensities as thecorresponding first reduced, precipitated and then supported ma-terials; indicating that the crystallite sizes are larger when theprecursors are reduced without carbon present. Although absent inthe XRD pattern of Fig. 2, the presence of minor nanocrystalline oramorphous SnOx phases, especially near the surface, cannot beruled out. This is also indicated by the lattice spacing on the surfaceof some nanoparticles found by TEM throughout all samples.

3.2. Electrochemical characterization of the PteRheSn catalysts

To correlate catalyst structure and composition with the elec-trocatalytic EOR polarization behavior, cyclic voltammograms (CV)in 0.1 M HClO4 as well as in 0.5 M Ethanol (EtOH) and 0.1 M HClO4(Fig. 3A) were collected. Here, only the first forward scan of theethanol oxidation is presented while the first two full scans areshown in the Supplementary Information (Fig. S1 and S2). Inaddition, potentiostatic chronoamperometric (CA) experimentswere performed to examine the catalytic performance and catalyststability at an electrode potential of technical interest (0.45 V vs.RHE) for extended period of times (Fig. 3B). While current densitieswere normalized only by the amount of Platinum present on theelectrode surface, the reduction charge of Hydrogen under poten-tial deposition (HUPD) was additionally employed to estimate

electrochemically active surface area (ECSA) values for the twocatalysts based on the charge density for the formation of a fullycovered PteH monolayer (210 mC/cm2) [28]. CO-ECSA values [29]from the CO-oxidation charge could not be used to compare thetwo surface systems due to the poor CO oxidation activity of thePteRheSn surface. Mass based current density forward scans of thefirst CV in Ethanol containing electrolytes are shown in Fig. 3A. TheECSA values are given in Table 2 alongside mass activity values fordifferent potentials and after 1800 s.

The electrocatalysts synthesized using the ambient pressuremethods with conventional heating (ap materials) showed clearlyhigher activity for the EOR compared to the catalysts that wereprepared under elevated pressures (op materials). In particular,ap-PtRhSnþC showed an exceptionally early onset potential of lessthan 0.20 V vs. RHE and the highest mass activity for the low po-tential region until 0.60 V vs. RHE. At this potential ap-PtRhSn/Cstarted to outperform ap-PtRhSnþC, showing the highest Pt massactivity at around 0.80 V vs. RHE. While the mass activity ofop-PtRhSnþC showed a similar onset potential and lower peakcurrent densities compared to the ap materials, op-PtRhSn/Cshowed low EOR activity over the entire potential range. The ECSAfor ap-PtRhSn/C is the highest with 8.4 mPt

2 gPt�1 followed by

Table 2Electrochemical EOR activity of PtRhSn catalysts: electrochemical surface area (ECSA), current densities at 0.45 V vs. RHE and 0.80 V vs. RHE during voltammetric scans, andcurrent density at constant 0.45 V vs. RHE after 1800 s.

Catalysts ECSAmPt

2 gPt�1Pt loadingmg cmgeo

�2j@0.45 VmA mgPt�1

j@0.80 VmA mgPt�1

j@0.45 V@1800 smA mgPt�1

1) apPtRhSn/C 8.4 13 14.3 142.3 2.92) apPtRhSnþC 5.4 20 31.3 85.2 6.03) opPtRhSn/C 3.3 27 5.6 15.0 0.14) opPtRhSnþC 3.6 22 14.4 43.8 1.4

N. Erini et al. / Journal of Power Sources 294 (2015) 299e304 303

5.4 mPt2 gPt�1 for ap-PtRhSnþC. The op materials showed in general

only about half of this activity around 3 mPt2 gPt�1. This indicates a Pt

richer surface of the ap materials.The observed synthesis-activity correlations can be explained by

the formation of well-dispersed catalyst nanoparticles with limitedagglomeration, in case that the carbon support is present duringprecursor reduction. This would lead to low EOR onset potentialscompared to the more agglomerated step wise prepared catalysts.The activity cross-over observed for the ap materials would beplausibly explained by a prevalent selectivity for the 2 and 4 elec-tron products (acetaldehyde or acetic acid) for the more dispersedparticles, whereas the more agglomerated particles with theirsmaller interparticle distance, thanks to enhanced readsorption,would show a higher selectivity for the higher electron productsassociated with a higher mass activity at higher overpotentials[30e32]. The lowest overall activity of op-PtRhSn/C correlates withthe lowest ratio of hexagonal to cubic phased alloy formation asseen from XRD, supporting the notion that a high concentration ofactive surface site ensembles formed by a Niggliite structure of theparticle is of importance for the achievement of high EOR activities[22]. The relative chronoamperometric performances at 0.45 V vs.RHE of the four distinct electrocatalysts (Fig. 3B) are in goodaccordance with the corresponding mass activities in Fig. 3A at thesame potentials. All materials showed stable catalytic behavior overextended periods of time.

In order to estimate the catalytic performance related to the

Fig. 4. Current density (mA cmgeo�2 at 0.70 V vs. RHE) dependency on the platinum

loading (mg cm�2) for ethanol electro-oxidation on 1) ap-PtRhSn/C, 2) ap-PtRhSnþC, 3)op-PtRhSn/C and 4) op-PtRhSnþC recorded in 0.5 M C2H5OH þ 0.1 M HClO4 with ascan rate of 20 mV/s according to Brouzgou et al. [33] Current densities are normalizedby the geometrical area of the working electrode surface (0.196 cm2). For comparisonPtRhSnO2/C by Kowal et al. is included [3].

state-of-the-art in ternary EOR electrocatalysis, we comparedthe performance to the most promising catalysts in literature. Theap-PtRhSnþC catalyst displayed four times higher current densitiesat þ0.45 V in acidic media than a similar PtRhSnO2/C catalyst [20]and as an earlier PtRhSn catalyst with the same 3:1:4 ratio [5].Brouzgou et al. compared a variety of DEFC anode catalysts withincreasing Pt loading on the working electrode surface and estab-lished a certain region between 0.1 and 1.0 mA cmPt

2 @0.70 V vs. RHEin where most of the reported catalyst to date belong [33]. Fig. 4shows that the ap-PtRhSnþC lays already within this region whiledisplaying the lowest metal loading of all electrocatalysts.

4. Conclusions

Both ambient and overpressure synthesis approaches resultedin carbon supported PteRheSn electrocatalysts that exhibitedhigh performance for ethanol oxidation in acidic medium. The X-ray diffractogramms show fcc PtRh and hexagonal PtRhSn pha-ses. The highest overall activity is shown by materials synthe-sized at ambient pressures, with very early onset potentials foran on-carbon-reduced-precursor synthesis approached catalystap-PtRhSnþC and the highest mass activity for ap-PtRhSn/C. Thiscould indicate the formation of a high concentration of surfacesite ensembles on the carbon during the reduction of the pre-cursors that lead to higher activities at lower potentials, possiblytowards the complete ethanol oxidation. Further work is nownecessary to investigate the product and intermediate distribu-tion on the different surfaces with qualitative in-situ techniquesin order to help elucidate the mechanisms of ethanol-electrooxidation.

Acknowledgments

The project on which this Report is based was promoted withfunds from the Federal Ministry of Education and Research underthe promotional reference number 16N11929. Responsibility forthe contents of this publication lies with the author. Partial finan-cial support by the German Research Foundation (DFG) throughgrant STR 596/4-1 (“Pt stability”) is gratefully acknowledged.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jpowsour.2015.06.042.

References

[1] E. Antolini, J. Power Sources 170 (2007) 1e12.[2] S. Beyhan, C. Coutanceau, J.-M. L�eger, T.W. Napporn, F. Kadırgan, Int. J.

Hydrogen Energy 38 (2013) 6830e6841.[3] A. Kowal, M. Li, M. Shao, K. Sasaki, M.B. Vukmirovic, J. Zhang, N.S. Marinkovic,

P. Liu, A.I. Frenkel, R.R. Adzic, Nat. Mater. 8 (2009) 325e330.[4] J.J. Linares, T.A. Rocha, S. Zignani, V.A. Paganin, E.R. Gonzalez, Int. J. Hydrogen

Energy 38 (2013) 620e630.[5] M. Li, A. Kowal, K. Sasaki, N. Marinkovic, D. Su, E. Korach, P. Liu, R.R. Adzic,

Electrochim. Acta 55 (2010) 4331e4338.

N. Erini et al. / Journal of Power Sources 294 (2015) 299e304304

[6] H. Hitmi, E.M. Belgsir, J.M. Leger, C. Lamy, R.O. Lezna, Electrochim. Acta 39(1994) 407e415.

[7] T. Iwasita, E. Pastor, Electrochim. Acta 39 (1994) 531e537.[8] F. Colmati, E. Antolini, E.R. Gonzalez, Appl. Catal. B-Environ. 73 (2007)

106e115.[9] A. Rabis, P. Rodriguez, T.J. Schmidt, ACS Catal. 2 (2012) 864e890.

[10] J.C.M. Silva, R.F.B. De Souza, L.S. Parreira, E.T. Neto, M.L. Calegaro, M.C. Santos,Appl. Catal. B 99 (2010) 265e271.

[11] J.M. L�eger, S. Rousseau, C. Coutanceau, F. Hahn, C. Lamy, Electrochim. Acta 50(2005) 5118e5125.

[12] S. Song, P. Tsiakaras, Appl. Catal. B: Environ. 63 (2006) 187e193.[13] M. Oezaslan, F. Hasche, P. Strasser, J. Electrochem. Soc. 159 (2012)

B444eB454.[14] F. Hasche, M. Oezaslan, P. Strasser, ChemCatChem 3 (2011) 1805e1813.[15] M. Oezaslan, F. Hasche, P. Strasser, J. Electrochem. Soc. 159 (2012)

B394eB405.[16] L. Gan, M. Heggen, R. O'Malley, B. Theobald, P. Strasser, Nano Lett. 13 (2013)

1131e1138.[17] F. Colmati, E. Antolini, E.R. Gonzalez, J. Alloys Compd. 456 (2008) 264e270.[18] J.C.M. Silva, L.S. Parreira, R.F.B. De Souza, M.L. Calegaro, E.V. Spinac�e, A.O. Neto,

M.C. Santos, Appl. Catal. B 110 (2011) 141e147.[19] E.V. Spinace, R.R. Dias, M. Brandalise, M. Linardi, A.O. Neto, Ionics 16 (2010)

91e95.[20] A. Kowal, S.L. Gojkovi�c, K.S. Lee, P. Olszewski, Y.E. Sung, Electrochem. Com-

mun. 11 (2009) 724e727.[21] S. García-Rodríguez, S. Rojas, M.A. Pe~na, J.L.G. Fierro, S. Baranton, J.M. L�eger,

Appl. Catal. B: Environ. 106 (2011) 520e528.[22] N. Erini, R. Loukrakpam, V. Petkov, E.A. Baranova, R. Yang, D. Teschner,

Y. Huang, S.R. Brankovic, P. Strasser, ACS Catal. 4 (2014) 1859e1867.[23] N. Erini, S. Rudi, V. Beermann, P. Krause, R. Yang, Y. Huang, P. Strasser,

ChemElectroChem 2 (2015) 903e908.[24] S. García-Rodríguez, S. Rojas, M.A. Pe~na, J.L.G. Fierro, S. Baranton, J.M. L�eger,

Appl. Catal. B 106 (2011) 520e528.[25] R. Loukrakpam, B.N. Wanjala, J. Yin, B. Fang, J. Luo, M. Shao, L. Protsailo,

T. Kawamura, Y. Chen, V. Petkov, C.-J. Zhong, ACS Catal. 2 (2012) 1461.[26] P.A. Russo, M. Ahn, Y.-E. Sung, N. Pinna, RSC Adv. 3 (2013) 7001e7008.[27] S. Song, C. He, J. Liu, Y. Wang, A. Brouzgou, P. Tsiakaras, Appl. Catal. B 119e120

(2012) 227e233.[28] M. Shao, J.H. Odell, S.-I. Choi, Y. Xia, Electrochem. Commun. 31 (2013) 46e48.[29] S. Rudi, C. Cui, L. Gan, P. Strasser, Electrocatalysis 5 (2014) 408e418.[30] S.C.S. Lai, M.T.M. Koper, Faraday Discuss. 140 (2008) 399e416.[31] L. Ono, B. Rold�an-Cuenya, Catal. Lett. 113 (2007) 86e94.[32] A. Schneider, L. Colmenares, Y.E. Seidel, Z. Jusys, B. Wickman, B. Kasemo,

R.J. Behm, PCCP 10 (2008) 1931e1943.[33] A. Brouzgou, S.Q. Song, P. Tsiakaras, Appl. Catal. B 127 (2012) 371e388.[34] M. Oezaslan, F. Hasche, P. Strasser, J Phys Chem Lett 4 (2013) 3273e3291.[35] P. Strasser, J. Comb. Chem. 10 (2008) 216e224.[36] Z. Liu, S. Koh, C. Yu, P. Strasser, J. Electrochem. Soc. 154 (2007) B1192eB1199.[37] F. Hasche, M. Oezaslan, P. Strasser, ChemCatChem 3 (2011) 1805e1813.[38] M. Oezaslan, F. Hasche, P. Strasser, J. Electrochem. Soc. 159 (2012)

B394eB405.