Influence of Structural Defects on the Electrocatalytic

ORIGINAL PAPER

Influence of structural defects on the electrocatalyticactivity of platinum

O. V. Cherstiouk & A. N. Gavrilov & L. M. Plyasova &

I. Yu. Molina & G. A. Tsirlina & E. R. Savinova

Received: 19 September 2006 /Revised: 20 August 2007 /Accepted: 20 August 2007 / Published online: 11 October 2007# Springer-Verlag 2007

Abstract Structural defects play major role in catalysis andelectrocatalysis. Nanocrystalline (or nanostructured) mate-rials composed of nanometer-sized crystallites joined viagrain boundaries have been recognized for their specificstructure and properties, differentiating them from singlecrystals, coarsely grained materials or nanometer-sizedsupported single-grained particles (Gleiter, NanostructMater 1:1–19, 1992). In this paper, we use Pt electrodes,prepared by electrodeposition on glassy carbon and goldsupports, as model nanocrystalline materials to explore theinfluence of grain boundaries and other structural defectson electrocatalysis of CO and methanol oxidation. We buildon the recently established correlations between the nano-structure (lattice parameter, grain size, and microstrains) ofelectrodeposited Pt and the deposition potential (Plyasova

et al., Electrochim. Acta 51:4447–4488, 2006) and use thelatter to obtain materials with variable density of grainboundary regions. The activity of electrodeposited Pt in theoxidation of methanol and adsorbed CO exceeds greatlythat for Pt(111), polycrystalline Pt, or single-grained Ptparticles. It is proposed that active sites in nanostructured Ptare located at the emergence of grain boundaries at thesurface. For methanol electrooxidation, the electrodes withoptimal nanostructure exhibit relatively high rates of the“direct” oxidation pathway and of the oxidation of stronglyadsorbed poisoning intermediate (COads), but not-too-highmethanol dehydrogenation rate constant. These electrodesexhibit an initial current increase during potentiostaticmethanol oxidation explained by the COads oxidationrate constant exceeding the methanol decomposition rateconstant.

Keywords Nanostructured materials . Structural defects .

Grain boundary . Electrocatalysis . Platinum . CO oxidation .

Methanol oxidation

Introduction

Understanding relations between the structure of materialsand their adsorptive and catalytic properties is the focalpoint of electrocatalysis and heterogeneous catalysis and isof both fundamental and practical importance. Structuraleffects may be concerned with the influence of (1) surfacecrystallography, (2) size confinement, and (3) structuraldefects [1]. These factors affect both geometric andelectronic properties of materials.

Low-index single crystals appear to be the best modelsurfaces for the investigation of the influence of surfacecrystallography on catalysis and adsorption. Numerous

J Solid State Electrochem (2008) 12:497–509DOI 10.1007/s10008-007-0436-8

Dedicated to the 65th birthday of Teresa Iwasita, who has made amajor contribution to the fields of electrocatalysis andelectrochemistry.

O. V. Cherstiouk : L. M. Plyasova : I. Y. Molina : E. R. SavinovaBoreskov Institute of Catalysis,Siberian Division of the Russian Academy of Sciences,Prospekt Ak. Lavrentieva 5,Novosibirsk 630090, Russia

A. N. Gavrilov :G. A. TsirlinaDepartment of Electrochemistry, Moscow State University,Leninskie Gory, 1 - str.3,Moscow GSP-2 119992, Russia

E. R. Savinova (*)l’Ecole Européenne Chimie Polymères Matériaux,Université Louis Pasteur,UMR 7515, 25, rue Becquerel,F 67087 Strasbourg Cedex 2, Francee-mail: [email protected]

studies confirmed structure sensitivity of various adsorptionprocesses, heterogeneous catalytic and electrocatalyticreactions (for recent books and reviews on structural effectsin electrocatalysis see, e.g., [2–5]). It has been demonstratedthat catalytic activity of metal surfaces in structure-sensitiveprocesses may be exclusively determined by the presenceof defects such as low coordinated sites [6]. Stepped singlecrystals have served as models for unveiling the depen-dence of the electrocatalytic activities on the step densityand have been used to study, e.g., oxygen reduction [7–9],CO [10–12], and methanol oxidation [13]. Recently, nano-faceted surfaces have been introduced into electrocatalysiswith the aim to better understand an interplay of thereaction kinetics at vicinal facets [14, 15].

Catalysts utilized for practical applications are oftencomposed of nanometer-sized (metal) particles on inert(e.g., carbon) supports. The influence of the particle sizeand surface crystallography on the electrocatalytic activityof supported metal nanoparticles has been widely discussedin the literature (see Wieckowski et al. [5] and referencestherein). The so called “negative-particle-size effects”, withspecific catalytic activity decaying with the decrease of theparticle size, have been documented for a number ofpractically relevant electrochemical processes, e.g., oxygenreduction reaction [16, 17], methanol [18–20], and carbonmonoxide oxidation [21–25]. Some authors, however,dispute particle size effects in the above reactions [26, 27].

Much less is known about the influence of defects in thebulk crystalline structure of materials on their sorptive andcatalytic properties. This influence may be manifold. Inparticular, bulk defects may affect sorption of foreign atomsin the material bulk [28]. Extended bulk defects, emergingat the surface, may create surface defects, and thus,influence surface processes. In the 1960s–1970s of the20th century, a number of studies was published by theRussian school of electrochemistry devoted to the influenceof various thermal, mechanical, chemical, and electrochem-ical treatments on the adsorptive and electrocatalyticproperties of polycrystalline metals: Pt [29–31], Ni [32],Fe [33]. These led the authors to the conclusion thatdislocations, emerging at the surface, create surface defectsincreasing hydrogen adsorption energy.

Of particular interest are nanocrystalline (addressed alsoas “nanostructured”) materials composed of nanometer-sized crystallites joined via grain boundaries. High volumedensity of disordered grain boundary regions entail uniquephysical properties differentiating these materials fromsingle crystals, coarsely grained materials or nanometer-sized supported single-grained particles (see, e.g., [34–36]).Nanocrystalline materials of this sort are very well-knownto material scientists, but have attracted undeservedly littleattention of the catalysis community. Meanwhile, positiverole of grain boundaries in heterogeneous catalysis was

outlined by Tsybulya et al. [37]. A correlation between thereaction rate of ethylene epoxidation and parametersresponsible for the relationship among the regular anddefect regions in the bulk structure of silver particles hasbeen established.

In fact, nanocrystalline materials have been widely usedin electrochemistry in the form of multilayer metal deposits,which are composed of nanometer-sized grains and containhigh volume concentrations of grain boundaries anddislocations [38]. Investigations devoted to the preparationof electrodeposited metals and alloys, analyses of theirstructure, and their application in electrocatalysis areabundant and cannot be comprehensively reviewed withinthis publication. We would like to point out that TeresaIwasita has contributed significantly to the investigation ofelectrocatalytic properties of electrodeposited metals[39–44].

It should be noted, however, that only a few publicationsaddressed specific role of grain boundary regions inelectrochemical and electrocatalytic processes. In a seriesof publications, Tsirlina et al. [45–48] investigated hydro-gen sorption in nanostructured Pd, obtained throughelectrodeposition. They discovered a strong dependence ofthe hydrogen sorption [46–48] and adsorption [45] on thePd deposition potential and associated it with the density ofdefect regions in the metal lattice. In electrocatalysis, ofimportance are greatly enhanced specific catalytic activitiesof electrodeposited Pt in CO, methanol, and ethylene glycoloxidation compared to polycrystalline Pt foil or single-grained supported Pt nanoparticles documented, .e.g., in[23, 24, 49, 50]. The activity enhancements were attributedto the high grain boundary density. It was proposed thatcatalytically active sites in nanostructured materials arelocated at the emergence of grain boundaries at the surface.It is likely that unique properties of electrodeposited metalsare largely due to the interplay of bulk and surface defectsin their structure.

It is worth mentioning that nanocrystalline materials arehighly relevant to the practical applications. For example,catalytic layers of low-temperature fuel cells containsupported catalysts, usually with high metal percentage.High density of metal particles on the surfaces of supportmaterials occasionally results in formation of nanostruc-tured metals, and may have far-reaching consequences forelectrocatalysis as confirmed for carbon-supported mono-metallic Pt and bimetallic PtRu electrodes in [50, 51].

In our recent publication [52], it was demonstrated thatthe deposition potential can be used as a means for tuningnanostructure of electrodeposited Pt and preparing materialswith defined defect density. The present work is aimed atfurthering understanding of the relationships between thedefect density in Pt materials, on the one hand, and theiradsorptive and electrocatalytic properties, on the other

498 J Solid State Electrochem (2008) 12:497–509

hand. Platinum electrodeposits on glassy carbon (GC) andAu are utilized as model materials reflecting complexstructural features of real catalysts, which cannot beaccounted for by single crystals or supported single-grainednanoparticles.

Experimental

Materials

Solutions were prepared from Milli-Q water (18 MOhm·cm),H2SO4 (puriss. p.a., Fluka) and methanol (99.9%, redis-tilled). GC plates (1-mm thick; NIIGRAFIT, Russia; heattreated at 1,300°C) were mechanically polished to a mirrorfinish and cleaned repeatedly with ethanol, acetone, andwater in an ultrasonic bath before the electrode preparation.H2PtCl4 was prepared from Pt foil (99.999) using theprocedure described in [53].

Electrode preparation

Working electrodes were prepared via Pt electrodepositionon either gold foil or GC plates. The samples are furtherdesignated as Pt(ed#Ed)/Au and Pt(ed#Ed)/GC, respective-ly, with the deposition potential Ed indicated in thebrackets. Thus, Pt(ed#0.1)/GC denotes Pt electrodepositedon GC at 0.1 V vs RHE. Pt deposition was performed fromH2PtCl6 + HCl solutions at a constant potential, which wasvaried in the range from 0.025 to 0.55 V RHE. Typical Ptloadings amounted to 0.5–0.7 mg cm-2 for Pt/Au and to0.1–0.2 mg cm-2 for Pt/GC, as for the latter, the adhesion athigher loadings was not satisfactory. For further details ofthe sample preparation, the reader is referred to [52]. Beforeelectrocatalytic studies, the samples were aged underpotentiodynamic conditions as described in [52].

For comparison, some measurements were performedwith polycrystalline foil Pt(pc) and with GC-supportedsingle-grained Pt nanoparticles Pt(nano)/GC preparedby chemical deposition. Their preparation procedure isdescribed in [24].

Electrochemical measurements

Electrochemical measurements were carried out in three-electrode glass cells at 20±1°С after deaeration with Ar.The counter electrode was Pt foil and the referenceelectrode—either a reversible hydrogen electrode (RHE)or a mercury sulfate electrode (MSE) Hg|Hg2SO4|0.1 MH2SO4 (aq) connected to the cell via a Luggin capillary.Electrode potentials were controlled using AutolabPGSTAT30 potentiostat, and hereinafter, are given vsRHE (EMSE=0.73 V vs RHE).

For CO stripping, CO was bubbled through the electro-lyte for 15 min while the electrode was kept at a constantpotential of 0.1 V, and then, dissolved CO was removed bypurging the solution with Ar for 35 min. Methanol (MeOH)oxidation experiments were performed for Pt/GC usingcyclic voltammetry and chronoamperometry in 0.1 MH2SO4+0.1 M MeOH and for Pt/Au with steady-statevoltammetry in 0.5 М H2SO4+0.1 M MeOH. Steady-statepolarization curves were acquired under potentiostaticmode by varying the electrode potential in 20 mV incre-ments starting from 0.8 down to 0.5 V, with an automaticswitching to the next potential after satisfying thequasi-steady-state criterion. The latter was set at 1% currentdeviation per minute. Measuring one polarization curvetypically required 5–15 h. Such long measurements werenot feasible for Pt/GC samples, which contained smalleramounts of Pt and were thus more readily contaminated.Current densities were normalized to the true surface area ofPt-determined coulometrically from the hydrogen desorp-tion in supporting sulfuric acid solutions.

Electrode characterization

X-ray diffraction (XRD) patterns were acquired for Ptelectrodeposits without their separation from Au or GCsubstrate and were recorded using URD-63 diffractometer(Germany) with monochromatic Cu Ka� radiation in thereflection mode. The X-ray diffraction patterns werecollected in 2θ interval from 30 to 100° in 0.02° steps,the accumulation time was set in the range from 10 to 100 sdepending on the reflection intensities. The averageeffective size of coherently scattering domains D and thecrystal lattice microstrains Δd/d were determined usingVoight analysis [54]. For more details see [52].

Pt(nano)/GC samples were characterized with trans-mission electron microscopy (TEM) with the help ofJEM-2010 instrument with a lattice resolution of 1.4 Ǻand an accelerating voltage of 200 kV. Samples forelectron microscopy were prepared by scraping the activelayer off GC support and fixing the resulting powder on«holey» carbon films supported on copper grids.The average dN, and the surface average dS particlesize were calculated from size distributions as follows:

dN ¼Pidi

�PiNi; dS ¼P

id3i

�Pid2i.

Results and discussion

Structural characterization

In depth structural analyses of Pt electrodeposits on Au andGC supports were performed using XRD, TEM, scanningelectron microscopy (SEM), and scanning tunneling

J Solid State Electrochem (2008) 12:497–509 499

microscopy and described in [52, 55, 56]. In Plyasovaet al. [52], we explored the influence of the electrodeposi-tion potential on the structure. For convenience, we furnishthe reader with a synopsis of this study, while the detailscan be found in the original publications [52, 55, 56].

Structural characterization complemented with the anal-ysis of deposition transients brought us to the conclusionthat specific structural features of multilayered Pt depositsresult from the interplay between the primary (on thesupport surface) and the secondary (on the surface ofelectrodeposited Pt nanoparticles) nucleation. The resultantnanocrystalline materials, composed of randomly orientednanometer-size domains (grains), are formed in the se-quence of primary and secondary nucleation events, growthof metal nuclei in close proximity to each other, and theircoalescence. Characteristic features of these materials,compared to polycrystalline Pt, are: (1) decreased latticeparameter a, (2) microstrains Δd/d, and (3) high concen-tration of dislocations and grain boundaries, all of thembeing determined by the deposition potential. As thedeposition potential is shifted towards the positive, thelattice parameter is reduced, while microstrains andthe concentration of dislocations and grain boundaries rise(Table 1).

Dislocation densities N in the boundary regions ofnanocrystalline materials may be estimated from X-raydiffraction data as follows (see [57, 58] for details):

N ¼ 3n�D2 ð1Þ

Equation 1 is based on the assumption of fully coalescedcubic crystallites. Here, n is the number of dislocations perside of a cube. As the value of n is unknown, the lowerlimit of N was estimated with n set equal to 1. The resultant

N varies markedly with the deposition potential (Table 1).Due to the nanometer size of Pt grains, N exceeds the valuecharacteristic of Pt(pc) by more than an order of magnitudefor the most defective samples.

Other specific structural features of Pt electrodeposits aremicrostrains Δd/d (Table 1), which, depending on thedeposition potential, change from ~0.3 to ~0.7%. These areabout an order of magnitude smaller than the values typicalfor strained materials obtained under severe plastic defor-mations [57, 58]. Plastic deformations of materials areknown to result in the emergence of random dislocations inthe bodies of crystalline grains, which density ρhkl may beestimated as follows [57, 58]:

ρhkl ¼2ffiffiffi3

p$d=dð ÞhklbDhkl

ð2Þ

Here b is Burgers vector.The values of ρ111 calculated using Eq. 2 are typical for

electrodeposited metals and are listed in Table 1. Note that forsingle crystals, ρhkl~10

2–103 cm-2. One may also see that thevalue of ρ is significantly smaller than that of N, suggestingthat dislocations, localized in the grain boundary regions, arethe most abundant defects in multilayered Pt electrodeposits,dominating their specific structural properties.

Note that variation of nanostructure with the depositionpotential affects also the specific surface areas of Ptdeposits which range from 7 to 15 m2 g-1 for Pt/Au andfrom 14 to 50 m2 g-1 for Pt/GC. The highest specificsurface areas were achieved at Ed=100 mV both for Pt(ed)/GC and for Pt(ed)/Au, suggesting that the degree of graincoalescence under these conditions is the lowest. This is inagreement with the lower values of microstrains and closeto Pt(pc) lattice parameter.

Table 1 Grain size D, microstrains Δd/d, lattice parameter a, volume density of dislocations in the grain bodies ρ, and in the grain boundaryregions N calculated for Pt(pc), for Pt(ed)/GC and for Pt(ed)/Au from X-ray diffractograms in <111> direction (See text for details)

Sample Ed (V vs RHE) D111 (nm) (Δd/d)111 100% a (Å) ρ111 (1011 cm-2) N111 (10

11 cm-2)

Pt(ed)/GC 0.025 19 0.33 3.918 (5) 2.1 8.00.100 18 0.27 3.91 (9) 1.9 9.20.200 14 0.44 3.91 (7) 4.0 15.50.300 14 0.49 3.91 (2) 4.2 14.30.400 16 0.56 3.91 (4) 4.4 11.70.550 14 0.69 3.91 (4) 6.2 14.9

Pt(ed)/Au 0.025 23 0.32 3.917 (8) 1.8 5.70.050 28 0.29 3.918 (2) 1.3 3.80.100 20 0.40 3.916 (8) 2.5 7.50.200 24 0.59 3.91 (9) 3.1 5.20.300 14 0.49 3.913 (9) 4.4 15.30.400 18 0.60 3.915 (4) 4.6 9.30.500 19 0.66 3.91 (6) 4.4 8.3

Pt(pc)a 50 3.9231 1.2

a Due to the low intensity of <111> reflection for textured Pt(pc), D<111> was recalculated from D<200>.

500 J Solid State Electrochem (2008) 12:497–509

Interfacial properties

Typical cyclic voltammograms (CVs) of Pt electrodepositedon GC in supporting H2SO4 electrolyte are shown in Fig. 1.For clarity, the figure is furnished with labels, which markhydrogen and oxygen adsorption/desorption peaks. A shiftof the deposition potential towards the positive results in adecrease in the amplitude of hydrogen adsorption/desorp-tion peaks at ~0.28 V (“strongly adsorbed” hydrogenHUPD2) and an increase in the amplitude of peaks at~0.12 V (“weakly adsorbed” hydrogen HUPD1). Compari-son of the hydrogen underpotential deposition (UPD)region for electrodeposited Pt with those for singlecrystalline electrodes [59, 60] allows to tentatively attributethese changes to a growth of the contribution of (110)-typeand a decrease of the contribution of (100)-type sites to thesurface. This is in agreement with the increase of thedensity of the grain boundary regions. The ratio of HUPD1

and HUPD2 amplitudes calculated from the positive sweepsof CVs is shown in Fig. 2a for Pt(ed)/GC and Pt(ed)/Auand demonstrates similar trends regardless of the nature ofsupport, although the magnitude of the changes is larger forPt(ed)/GC. For more information and discussion of thesupport effect, the reader is referred to Plyasova et al.[52].

The deposition potential affects also the position of thepeak, corresponding to the reduction of Pt surface oxides(“oxygen reduction” peak EOd) on the negative sweeps ofCVs. The higher the deposition potential, the lower EOd

(Fig. 2b), suggesting that oxygen is stronger and moreirreversibly bound to the surface of Pt deposited at morepositive deposition potentials. This effect is observed bothfor Pt(ed)/GC and for Pt(ed)/Au electrodes, but is muchstronger in the former case.

The data obtained indicate that a positive shift of thedeposition potential, which is associated with the changes

in the nanostructure of Pt deposits, leads to a systematicalteration in the state of hydrogen and oxygen adsorbateson the metal surface. These may be tentatively ascribedto an increase of the contribution of defect sites formed atthe emergence of domain boundaries at the surface ofelectrodeposited Pt.

Electrocatalytic activity of Pt/GC

CO oxidation

Figure 3 shows CO stripping voltammograms for Pt(ed)/GC prepared at different deposition potentials (curves 1,2).The position of CO oxidation peak shows little dependenceon the deposition potential, but for all electrodepositedsamples, the peak is negatively displaced compared topolycrystalline Pt (curve 3) and Pt(nano)/GC (curve 4).Note that the saturation CO coverages for the samplesexplored are not the same, hence, the differences in thestripping peak areas.

More facile CO oxidation kinetics on Pt(ed)/GC com-pared to Pt(nano)/GC and Pt(pc) is confirmed also bychronoamperometry. Figure 4 shows CO monolayer oxida-tion transients registered at constant electrode potential of0.73 V for Pt electrodes with different nanostructures.

0.0 0.2 0.4 0.6 0.8 1.0 1.2

-20

0

20

Od

HUPD2

j /

µAcm

-2

E / V vs. RHE

1

2

HUPD1

Fig. 1 Cyclic voltamogramms of Pt deposits on GC in 0.1 M H2SO4

prepared at 1 0.025, 2 0.4 V vs RHE. Sweep rate 0.02 V s-1

0.78

0.79

0.80

0.81

E / V vs. RHE

a

EO

d /

V v

s. R

HE

0.0 0.1 0.2 0.3 0.4 0.5 0.60.8

1.0

1.2

1.4

1.6

Pt (pc)

b

HU

PD

1/H

UP

D2

Pt (pc)

Fig. 2 Influence of the deposition potential Ed on the a ratio ofHUPD1 and HUPD2 peak amplitudes calculated from the positivesweeps of CVs taken in 0.1 M H2SO4 for Pt/GC (open symbols) andin 0.5 M H2SO4 for Pt/Au (filled symbols) and b surface oxidereduction peak EOd for Pt/GC. Sweep rate 0.02 V s-1. The notationsare explained in Fig. 1. The corresponding values for Pt(pc) areshown with the dashed lines

J Solid State Electrochem (2008) 12:497–509 501

These exhibit peaked shape which is in agreement with theLangmuir–Hinshelwood mechanism of CO oxidation [61]:

H2O $ OHads þ Hþ þ e� ð3Þ

COads þ OHads�!kox CO2 þ Hþ þ e� ð4Þ

Investigation of CO oxidation at stepped Pt single crystalsurfaces both in UHV [62–65] and in electrochemicalenvironment [11, 12] has decisively proven that formationof chemisorbed oxygen-containing species (reaction 3)

occurs preferentially at low coordinated sites—steps anddefects, COads from terraces reacting with Oads/OHads at thesteps. Different kinetic models have been set up and appliedfor fitting experimental transients on single crystal andnanostructured surfaces. Their detailed discussion is beyondthe scope of this paper and is presented, e.g., in Andreauset al. [66]. To account for heterogeneous surfaces ofnanoparticulate and nanostructured electrodes, a mathemat-ical model employing the active site concept was proposedin Andreaus et al. [66]. This model, assuming that OHad

formation is restricted to the active sites, reproduces verywell experimental transients for single- and multiple-grained GC-supported Pt nanoparticles. In depth analysesof various kinetic parameters affecting the peak shape inchronoamperograms were provided [66]. The time of themaximum (tmax) is determined by the fraction of the activesites on the surface, the rate constants for OHads formationat the active sites, and COads + OHads recombination.

Figure 5 shows log(tmax) vs Eox for Pt electrodepositedon GC at various deposition potentials, Pt(nano)/GC, andfor comparison, for Pt(111) and Pt(110) from Lebedevaet al. [11]. The decrease of tmax(111) > tmax(110) ≥ tmax(ed)reflects acceleration of the CO oxidation kinetics. tmax

exhibits minor dependence on the deposition potential; atlow oxidation potentials electrodeposited Pt is even moreactive than Pt(110).

The grain size of electrodeposited Pt varied from ca.14 nm at Ed=0.550 V to ca. 19 nm at Ed=0.025 V vs ca.

10 20 30 40 500

50

100

150

1 2

3

4

5

j /

µAcm

-2

time / s

Fig. 4 Current transients for CO monolayer oxidation in 0.1 MH2SO4 after a potential step from 0.1 to 0.73 V vs RHE for Pt(ed)/GCprepared at different deposition potentials: 0.025 (1), 0.100 (2), 0.300(3) V vs RHE. For comparison data for polycrystalline Pt (4) and forPt nanoparticles (dN=3.3 nm; dS=5.3 nm) chemically deposited onGC (5) are shown. The latter are replotted from Maillard et al. [96]with permission from Elsevier. CO adsorption potential 0.1 V vs. RHE

0.6 0.7 0.8 0.9 1.0-1

0

1

2

3 1

2

3

4

5

6

7

log

(t m

ax)

Eox / V vs. RHE

Fig. 5 Plot of the logarithm of time (in seconds) corresponding to thecurrent maximum (tmax) in CO oxidation transients against the finalpotential for Pt(ed)/GC prepared at different deposition potentials: 10.100, 2 0.300, 3 0.550 V vs RHE. 4 shows data for Pt nanoparticles(dN=3.3 nm; dS=5.3 nm) chemically deposited on GC and replottedfrom Maillard et al. [96] with permission from Elsevier. 5 and 6 showdata for Pt(110) and Pt(111), respectively, measured in 0.5 M H2SO4

and replotted from Lebedeva et al. [11] with permission from ACS.Data for polycrystalline Pt are shown in 7. CO adsorption potential0.1 V vs RHE. The lines are linear fits of the experimental data

0.6 0.7 0.8 0.9 1.00

50

100

150 4

31

j /

µAcm

-2

E / V vs. RHE

2

Fig. 3 CO stripping voltammograms in 0.1 M H2SO4 for Pt(ed)/GCelectrodeposited at 0.100 (1) and at 0.300 (2) V vs RHE. Forcomparison, data for polycrystalline Pt (3) and for Pt(nano)/GC (dN=2.8 nm; dS=3.4 nm) chemically deposited on GC (4) are shown. Thelatter are replotted from Maillard et al. [96] with permission fromElsevier. Sweep rate 0.02 V s-1, CO adsorption potential 0.1 V RHE

502 J Solid State Electrochem (2008) 12:497–509

50 nm for Pt(pc). These grain dimensions are well beyondthe particle size range where size effects may be expected.Indeed, it has recently been demonstrated [23–25, 67] thatparticle size effects in CO monolayer oxidation occur in therange from ~1 to ~4 nm, the reaction overpotentialdecreasing with the size. Given the grain dimensions ofelectrodeposited Pt and its higher [compared to Pt(pc) or Pt(nano)/GC] electrocatalytic activity in COads oxidation, theobserved structural effects can hardly be attributed to theinfluence of the grain size. Accelerated CO oxidation onnanostructured Pt has been mentioned in a number ofpublications [68–70]. Cherstiouk et al. [23] and later onMaillard et al. [50] associated this activity enhancement tothe catalytic role of grain boundaries interconnecting Ptcrystallites. In what follows, this hypothesis will be furtherverified.

Methanol oxidation

Cyclic voltammetry CVs in 0.1 M MeOH+0.1 M H2SO4

are shown in Fig. 6. Typical of nonsteady-state CVs forMeOH oxidation on Pt electrodes, these exhibit stronghysteresis between the positive and the negative going

scans. All electrodeposited electrodes show significantlyenhanced peak current densities compared to Pt(pc) and Pt(nano)/GC.

Comparison of CVs for MeOH oxidation with thecorresponding CO stripping voltammograms shown inFig. 6 with thick solid lines indicates that (1) MeOHoxidation starts in the potential interval below the onset ofthe main CO stripping peak; (2) MeOH oxidation currentdensities (jMeOH) for Pt(ed)/GC exceed greatly thosecorresponding to CO ML oxidation (jCO), while for Pt(nano)/GC, jMeOH and jCO are comparable. To analyze theexperimental data, we briefly review the current state of theunderstanding of the methanol electrooxidation mechanism.

MeOH oxidation on Pt electrodes has been under carefulattention of electrochemists for several decades (see, e.g.,review articles [71–77]). The main features of this complexmultistep reaction have already been described in theseminal paper by Petrii et al. published in 1965 [77]. Thereaction occurs through a sequence of MeOH dehydroge-nation/decomposition steps leading to formation of astrongly adsorbed intermediate, inhibiting the reaction. Alarge gap has been noticed between the initial and thesteady-state oxidation rates and attributed to the reactionself-inhibition. During the last decades, most of theconclusions of this paper found additional experimentalsupport from spectroscopic techniques, including infraredspectroscopy (IR) and differential electrochemical massspectroscopy (DEMS).

According to the present understanding, methanoloxidation on Pt electrodes occurs via the “dual-path”mechanism [78]. The so-called “indirect” pathway occursas described above and involves (1) MeOH dissociativeadsorption and formation of strongly adsorbed CHxOads

intermediates and (2) oxidation of these adsorbates in aLangmuir–Hinshelwood type reaction via interaction withchemisorbed oxygen-containing species (usually denoted asOHads). Equation 5 describes complete dehydrogenationleading to COads, while Eq. 4 describes the oxidation of thelatter. Based on the investigations of MeOH oxidation withthe electrochemical methods [79, 80], in situ IR spectros-copy [81–85] and online DEMS [81, 86–91], COHads andCOads have been identified as the main adsorbates, the latterbeing abundant in the potential interval of continuousMeOH oxidation. The second, the so-called “direct”pathway is postulated to occur through a “reactiveintermediate” (RI), which is different from COads and isless strongly bonded to the electrode surface. Formation ofsome weakly bonded intermediates which can be easilywashed away from the electrode surface has been men-tioned already by Petrii et al. [77]. The term “directpathway” may sound somewhat misleading, but corre-sponds to the pathway which avoids formation of strongly

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

0

250

500

Cu

rren

t den

sity

/µA

cm-2

E / V vs. RHE

a

0

250

500 b

0

250

500 c

0

250

500 d

-40

0

40

e

Fig. 6 CVs in 0.1 M MeOH+0.1 M H2SO4 (thin solid lines)measured at 0.02 V s-1 for Pt(ed)/GC prepared at different depositionpotentials: a 0.100, b 0.200, c 0.550 V vs RHE. d shows data forpolycrystalline Pt foil and e for Pt(nano)/GC (dN=1.3 nm; dS=1.7 nm). Thick solid lines show CO stripping voltammograms forrespective samples measured at the same sweep rate. In e, COstripping was performed at 0.1 V s-1, but the resulting current wasnormalized by multiplying by 0.2 and background-subtracted

J Solid State Electrochem (2008) 12:497–509 503

adsorbed poisoning intermediates. It results in a stepwiseoxidation of MeOH (Eq. 6) to form formaldehyde, formicacid, and ultimately CO2 [74, 84, 85, 92].

CH3OH�!kdec COads þ 4Hþ þ 4e� ð5Þ

CH3OH�!kd RI ! partial oxidation products ! CO2 ð6ÞRecently, Batista et al. [85, 93] proposed that partition-

ing between “direct” and “indirect” pathway occurs alreadyat the first dehydrogenation step. C–H bond splitting andformation of hydroxymethyl CH2OHads adsorbate furtherleads to COads (“indirect” pathway) as earlier proposed byPetrii et al. [77]. Meanwhile, O–H bond splitting andformation of reactive methoxy-adsorbate CH3Oads is be-lieved to lead to formaldehyde (“direct” pathway). Neu-rock, using ab initio quantum chemical methods, providedsolid basis for the above hypothesis [94], and demonstratedthe influence of the electrode potential and the solvent onthe elementary adsorption and oxidation steps. Furthersupport has been provided by Housmans et al. [95] in theironline electrochemical mass spectrometry investigation ofthe methanol oxidation pathways on the basal and steppedPt single crystalline electrodes in sulfuric and perchloricacid electrolytes. The results indicate that branchingbetween “direct” and “indirect” pathways is criticallydependent on the surface crystallography and presence ofsurface defects (steps). It was suggested that (110) stepscatalyze the direct pathway.

Comparison of CVs for methanol and for CO mono-layer oxidation suggests that for Pt(ed)/GC, the “direct”pathway is much more important than for Pt(pc) andparticularly for Pt(nano)/GC, and under some experimen-tal conditions, may become predominant. Indeed, whenthe electrode potential is cycled in the presence of MeOH,HUPD region is strongly suppressed, suggesting that thesurface of Pt is poisoned by strongly adsorbed COads.Under these conditions, observation of MeOH oxidationcurrents greatly exceeding the current associated withCOads oxidation, suggests occurrence of an alternativereaction pathway on the free surface sites. It is thus likelythat the “direct” pathway is accelerated by the presence ofthe grain boundary regions. For gaining more informationon the influence of nanostructure on the MeOH oxidationkinetics, chronoamperometry was used.

Chronoamperometry MeOH oxidation transients registeredfor selected Pt(ed)/GC at three different electrode potentialsare shown in Fig. 7. Apparently, the electrode structureexerts strong influence on the current densities and on theshapes of the transients. For Pt(ed)/GC obtained at Ed≥0.2 V decaying transients typical for processes with selfinhibition are observed (see Fig. 7a,b). For Pt(ed)/GCprepared at low deposition potentials (0.025 and 0.1 V),transients with the initial current increase are observed atselected oxidation potentials (see Fig. 7c). Rising transientshave also been obtained for Pt(ed#0.1)/Au. Similar tran-sients were observed for Pt(111) by Housmans and Koper

0 50 100 150 200

0

200

400

600

0.83 V

0.73 V

Curr

ent

den

sity

/ µ

Acm

-2

a

0.63 V

0 50 100 150 200

0

200

400

600

0.63 V

0.83 V

0.73 V

0.63 V

c

b

0 50 100 150 2000

200

400

600

0.83 V

0.73 V

time / s

0 50 100 150 2000

20

40

60

0.73 V

0.83 V

0.63 V

d

time / s

Fig. 7 Current transients in0.1 M MeOH+0.1 M H2SO4

after potential steps from 0.1 Vto 0.63 V, 0.73 V and 0.83 V forPt(ed)/GC prepared at differentdeposition potentials: (a) 0.550,(b) 0.200, (c) 0.100 V vs. RHEand (d) for Pt(nano)/GC(dN=1.3 nm; dS=1.7 nm). Steppotentials are indicated in theplots. Note different scale inpanel (d)

504 J Solid State Electrochem (2008) 12:497–509

[13] and attributed to CO oxidation being faster than themethanol dissociative adsorption. However, in contrast tolow catalytic activity of Pt(111), electrodeposited Ptsamples are characterized by very high specific catalyticactivities.

To better understand the influence of the depositionpotential (and ultimately the electrode structure) on theMeOH oxidation kinetics, we tried to fit the experimentaltransients with a simplified kinetic model, including (1)methanol dissociative adsorption (or decomposition) withthe rate constant kdec (Eq. 5), (2) oxidation of the adsorbedintermediates with the rate constant kox (Eq. 4), and (3)“direct” MeOH oxidation with the rate constant kd (Eq. 6).For simplicity, COads was considered as an abundantadsorption species and the “direct” oxidation pathwayassumed to have no common steps with the “indirect”pathway. Chemisorbed oxygen-containing adsorbates weredesignated as OHads. Under the above assumptions, theoverall MeOH oxidation current density can be expressedas a sum of currents (Eq. 7), corresponding to reactions 4,5, and 6:

jtotal ¼ jdec þ jox þ jd ð7ÞCurrent densities corresponding to individual reaction stepscan be expressed as [13]:

jdec ¼ 4eNPtkdec 1�Xi

qi

!n

ð8Þ

jox ¼ 2eNPtk0oxqCOqOH ð9Þ

jd ¼ j0d � 1�Xi

qi

!ð10Þ

Here, NPt is the number of Pt atoms per cm2 (1.5×1015), eis the elementary charge (1.6022×10-19 C), j0d correspondsto the current density of “direct” pathway on the Pt surfacefree from the adsorbates. The rate constant kdec implicitlyincludes MeOH concentration.

To diminish the number of adjustable parameters andsimplify the resulting kinetic equation the followingassumptions were made: (1) n≈1 (as estimated from theinitial part of the current transients); (2)

Piθi � θCO; (3)

CMeOH = const; (4) θmaxCO ¼ 1 and (5) θOH = const and equal

to the fraction of active sites on the surface. The modelconsidered is very similar to the model proposed byHousmans and Koper [13] for fitting MeOH electrooox-idation transients at stepped Pt single crystals exceptfor (1) and (5). In the model considered in Housmans andKoper [13], n=2 and θOH ¼ 1� θCO. We consider the

assumption θOH ¼ 1� θCO to be hardly applicable to veryheterogeneous surfaces of nanostructured electrodes uti-lized in this work.

Under the above simplifying assumptions, one obtains:

jtotal ¼ j0d � 1� qCOð Þ þ 4eNPtkdec 1� qCOð Þþ 2eNPtkoxqCO ð11Þ

d qd t

¼ kdec 1� qCOð Þ � koxqCO ð12Þ

Figure 8 shows some selected experimental transientsand their fits with the model described. Although thequality of the fits is not always satisfactory, the simplifiedmodel is able to predict a change in the shape of thetransients from decaying to rising when kox exceeds kdec,similar to the model considered by Housmans and Koper[13]. The limitations of the model are (1) the inability todescribe the decrease of the current at t>100 s and (2) zerovalues of jd which is admittedly incorrect. We have triedfitting with different models, including the one proposed byHousmans and Koper [13]. However, this did not improvesubstantially the quality of the fits. We suppose that kinetics

0 10 20 30 40 50

40

50

60

70

Ed=0.550 V

a

Curr

ent

den

sity

/ µ

Acm

-2

time / s

30

40

50

60

Ed=0.200 V

c

b

Ed=0.025 V

0 10 20 30 40 50

50

60

Fig. 8 Current transients in 0.1 M MeOH+0.1 M H2SO4 afterpotential steps from 0.1 to 0.63 V for Pt(ed)/GC prepared at differentdeposition potentials: a 0.550, b 0.200, c 0.025 V vs RHE. Thesymbols show the experimental data. The lines correspond to the fits.The following parameters are obtained from fitting: a kdec=0.049,kox/kdec=1.17; b kdec=0.039, kox/kdec=1.09; c kdec=0.056, kox/kdec=2.80. See text for further details

J Solid State Electrochem (2008) 12:497–509 505

of MeOH oxidation on highly heterogeneous surfaces ofelectrodeposited Pt, which comprise both regular and defectcrystal lattice domains, is more complicated than on singlecrystalline surfaces.Steady-State polarization Figure 9 shows E vs logj-curves(Tafel plots) obtained for Pt(ed)/Au using steady-statevoltammetry in the electrode potential interval from 0.55 to0.65 V vs RHE. The slopes @E

@ log j are close to linear and varyfrom 65 to 85 mV decj1: the higher the electrocatalyticactivity, the lower the slope. The Tafel slopes and the currentdensities for MeOH oxidation on Pt depend on the electrodepotential and on the degree of the electrode poisoning. Petriiet al. [77] documented @E

@ log j � 60 mVdec�1 for platinizedplatinum under steady-state MeOH oxidation and explainedthis by a reversible electrochemical followed by chemicalrds, the latter attributed to the interaction of chemisorbedmethanol derivatives with the oxygen-containing species.

Figure 10 compares quasistationary activities in MeOHoxidation. Similar trends have been documented for Pt(ed)/GC and Pt(ed)/Au: the activities are the highest for thesamples prepared at 0.1 V, decreasing for both higher andlower deposition potentials. First of all, this suggests thatthe electrocatalytic activities are determined by the deposi-tion potential rather than by the type of support. Second,this attests for reliability of the trends described, asPt(ed)/Au and Pt(ed)/GC samples were prepared andstudied in two different laboratories in Moscow andNovosibirsk, respectively. The same trend has beenobserved for specific surface areas which showed maximafor the electrodes prepared at 0.1 V [52]. The observed factformally leads us to the conclusion that the highest quasi-

steady-state activities of MeOH oxidation are achieved forelectrodeposited samples with the lowest degree of coales-cence of the crystalline domains, and hence, the lowestdefect densities. This seems to be in contradiction to thecatalytic role of grain boundaries declared in “Introduc-tion”. We will attempt to resolve this seeming discrepancyin “Catalytic properties as related to the defect density”.

Catalytic properties as related to the defect density

Although some indications concerning the influence of thedensity of defects in the bulk crystalline structure ofmaterials on the adsorptive and (electro)catalytic propertiesof materials can be found in the literature, direct correla-tions between these are scarce. A rare and impressiveexample is the already mentioned correlation between therate of ethylene epoxidation and the ratio between theparticle size measured by TEM and XRD (DTEM/DXRD),which is proportional to the number of grain boundaries inmulti-grain particles [37].

In this article, we seek for a correlation between thedensity of structural defects in Pt electrode materials andtheir electrocatalytic properties. As discussed in “Structuralcharacterization”, dislocations in the grain boundary

0.0 0.1 0.2 0.3 0.4 0.5 0.60

20

40

a

Ed / V vs. RHE

0.0 0.1 0.2 0.3 0.4 0.5 0.6

0

10

20

j /

µA

cm-2

j /

µAcm

-2

b

Fig. 10 Influence of the deposition potential on the quasi-steady-statecurrent densities for MeOH oxidation for Pt(ed)/GC measured fromchronoamperograms at 0.63 V and t=650 s (a) and for Pt(ed)/Aumeasured from stationary polarization curves at 0.64 V (b). Lines areguides for the eye

0.56 0.58 0.60 0.62 0.64

-3.0

-2.5

-2.0

E / V vs. RHE

lg j

(m

Acm

-2)

1

2

3

4

Fig. 9 Stationary polarization curves for methanol oxidation on Pt/Aumeasured in 0.1 M CH3OH+0.5 M H2SO4 solution. Ed=0.500 (1),0.400 (2), 0.100 (3), 0.025 (4). For each electrode, two curves areshown and indicated by arrows: one measured in cathodic, anotherone in anodic direction

506 J Solid State Electrochem (2008) 12:497–509

regions have been identified as abundant defects in electro-deposited Pt. In Fig. 11, the current densities for MeOHoxidation at 0.73 V RHE are plotted vs N1. Except for Pt(ed)/GC, the data for Pt(pc) and Pt(nano)/GC are included toexplore the versatility of the correlation. For Pt(nano)/GC,the grain boundary concentration was assumed zero, asaccording to HRTEM data, each Pt particle in the samplecomprised a single grain. Current densities were calculatedfrom potentiostatic current transients. Current density att=0.2 s, j(0.2), was taken as an apparent characteristic of thecatalytic activity of an essentially unpoisoned surface, whilecurrent density after 650 s, j(650), as a characteristic ofquasi-steady-state activity of the poisoned surface. It shouldbe noted that at 0.2 s, the charging current is negligiblysmall.

One may see that j(0.2) increases with the growth of thedefect density, being the lowest for Pt(nano)/GC, increasingsomewhat for polycrystalline Pt and increasing again forelectrodeposited Pt. The spread between different samplesmakes up to the factor of 5 (note the log scale). Meanwhile,

j(650) shows a volcano type behavior. Pt(nano)/GC (N~0)exhibits the lowest j(650), as the current drops by a factorof ca. 30 during 650 s of the electrolysis. A smaller but stillsizable current density drop is observed for Pt(pc) with(N=1.2 × 1011 cm-2). Meanwhile, for Pt(ed#0.1)/GC (N=9.2 × 1011 cm-2), no poisoning is observed, the activityincreasing with the time. Further growth of N leads,however, to the increase of the extent of the electrodepoisoning, and thus, results in lower quasi-steady-stateactivities.

Pt(ed#0.1)/GC seems to comprise an optimal density ofdefects. This can be rationalized as follows. According tothe present understanding, the activity of unpoisoned Ptelectrodes at short reaction times is determined by the rateof MeOH dehydrogenation (Eq. 5) and the rate of “direct”MeOH oxidation (Eq. 6). The positive effect of defects on j(0.2) indicates that either of these reactions or both becatalyzed by defects. This is in agreement with the resultsobtained by Housmans and Koper [13] on the positiveeffect of steps on kdec and kd at high index Pt singlecrystals. An increase of kdec, unless it is accompanied by aconcomitant increase of kox, must result in an increase inthe extent of poisoning. As demonstrated in “CO oxida-tion”, the CO oxidation activity increases greatly from Pt(nano) to Pt(pc) and then Pt(ed). The shape of the currenttransients changes from decaying for Pt(nano)/GC (Fig. 7d)and Pt(pc) (not shown) to rising for Pt(ed#0.100) and Pt(ed#0.025). Obviously, the increase of kox overweighs thatof kdec, thus, leading to the decrease of the extent of self-poisoning, the change in the shape of the transients, andthus, to the strong increase in the quasi-steady-state activityj(650). However, further increase of the defect density doesnot result in the further enhancement of the CO oxidationrate, as discussed in CO oxidation. Meanwhile, kdec is likelyto further increase, as suggested by the growth of j(0.2) andby the change of the shape of the transients, which forsamples obtained at Ed≥0.2 V again becomes decaying(Fig. 7a,b). This results in an increase of the degree of self-poisoning and the concomitant decrease of j(650).

We further speculate that the mechanism through whichdislocations in the grain boundary regions influence theelectrocatalytic activity of Pt operates via formation ofdefects on the surface of nanostructured electrodes at theemergence of grain boundaries at the surface. Metal atomsin the vicinity of grain boundaries usually have decreasednumber of neighbors in the first coordination shell, andthus, are expected to strongly bind adsorbates and catalyzebond-breaking reactions like water dissociation and MeOHdissociative chemisorption. For complex multistep electro-chemical reactions, like MeOH oxidation accompanied withself inhibition, the concentration of defects must besufficiently high to provide high electrocatalytic activitybut not-too-high to avoid high degree of self-inhibition.

Fig. 11 MeOH oxidation current density at 0.73 V vs RHE measuredin 0.1 M CH3OH+0.1 M H2SO4 vs the grain boundary density N forvarious Pt electrodes. Filled symbols show current density measuredafter 0.2 s, empty symbols show current density after 650 s of theelectrolysis. Lines are guides for the eye. See text for details

1It is worth mentioning that qualitatively the same trends are observedwhether the current is plotted vs N or >.

J Solid State Electrochem (2008) 12:497–509 507

Hence, so that the highest steady-state electrocatalyticactivity is attained, the electrode surfaces must comprisean optimal ratio between defect regions and orderedcrystalline domains.

It has been noticed that quasi-steady-state activity in themethanol oxidation correlates with the strength of oxygenbonding as accessed from the peak potential of surfaceoxide reduction: the stronger oxygen is bonded to thesurface, the lower is CO and methanol oxidation activity.This applies reasonably well within the series of electro-deposited samples. However, comparison of Pt electrodesprepared by different means reveals that the position of thesurface oxide reduction peak alone cannot serve as ameasure of the electrode electrocatalytic activity. Thisfinding suggests that it is not the strength of oxygenbonding which directly influences the catalytic activity, butit is rather structural parameters which affect both the oxidereduction peak and the methanol oxidation activity.

Conclusions

CO and methanol oxidation have been studied on electro-deposited Pt. The deposition potential has been found toconsiderably influence the catalytic activity of Pt in MeOHoxidation, and to a lesser extent, in CO monolayeroxidation. This can be explained by the dependence of Ptnanostructure on the deposition potential.

The correlation between the volume density of grainboundaries between Pt crystallites and electrocatalyticactivity in methanol and CO oxidation has been suggested.The initial activity of clean Pt surface in MeOH oxidationincreases with the increase in the grain boundary concen-tration, while the quasi-steady-state activity shows avolcano-type behavior with a maximum at N~9 × 1011 cm-2.The results obtained suggest that defect regions located atthe emergence of grain boundaries at the surface play animportant role in electrocatalytic processes.

Materials comprising high density of grain boundaries maybe of interest for practical applications, as contrary to thesurface defects, like steps and kinks on flat surfaces, defects inthe bulk crystalline structure usually have much longerlifetimes, unless high temperature annealing procedures areemployed.

Acknowledgements Financial support by the Russian Foundationfor Basic Research (RFBR) under grants No. 01-03-33132 and No.06-03-32737 is gratefully acknowledged. The authors would like tothank Prof. S.V. Tsybulya (BIC, Novosibirsk) for stimulatingdiscussions concerning the influence of structural defects on thecatalytic properties, Prof. O.A. Petrii (MSU, Moscow) for his remarkson methanol oxidation mechanism, and Prof. M.T.M. Koper for kindlygranting permission to reproduce the data from Ref. [11].

References

1. Skundin AM (1982) In: Itogi nauki i tekhniki. Elektrokhimija(Science and Engeneering Accounts. In Russ.), vol 18. VINITI,Moscow, p 228

2. Lipkowski J, Ross PN (eds) (1998) Electrocatalysis (frontiers inelectrochemistry). John Wiley & Sons, New York

3. Wieckowski A (ed) (1999) Interfacial electrochemistry: theory,experiment and applications. Marcel Dekker, New York, Basel

4. Vielstich W, Lamm A, Gasteiger HA (eds) (2003) Handbook offuel cells. Fundamentals, Technology and Applications, vol 2:Electrocatalysis. Wiley, Chichester

5. Wieckowski A, Savinova ER, Vayenas CG (eds) (2003) Catalysis &electrocatalysis at nanoparticle surfaces. Marcel Dekker, New York

6. Zambelli T, Wintterlin J, Trost J, Ertl G (1996) Science 273:16887. Strbac S, Anastasijevic NA, Adzic RR (1994) Electrochimica

Acta 39:9838. Maciá MD, Campiña JM, Herrero E, Feliu JM (2004) J

Electroanal Chem 564:1419. Kuzume A, Herrero E, Feliu JM (2007) J Electroanal Chem

599:33310. Lebedeva NP, Koper MTM, Herrero E, Feliu JM, van Santen RA

(2000) J Electroanal Chem 487:3711. Lebedeva NP, Koper MTM, Feliu JM, van Santen RA (2002) J

Phys Chem B 106:1293812. Lebedeva NP, Rodes A, Feliu JM, Koper MTM, van Santen RA

(2002) J Phys Chem B 106:986313. Housmans THM, Koper MTM (2003) J Phys Chem B 107:855714. Komanicky V, Menzel A, Chang KC, You H (2005) J Phys Chem

B 109:2354315. Komanicky V, Menzel A, You H (2005) J Phys Chem B

109:2355016. Kinoshita K (1992) Electrochemical oxygen technology. Wiley,

New York17. Gasteiger HA, Kocha SS, Sompalli B, Wagner FT (2005) Appl

Catal B Environ 56:918. Takasu Y, Iwazaki T, SugimotoW, Murakami Y (2000) Electrochem

Commun 2:67119. Frelink T, Visscher W, van Veen JAR (1995) J Electroanal Chem

382:6520. Park S, Xie Y, Weaver MJ (2002) Langmuir 18:579221. Yahikosawa K, Tateishi N, Nishimura K, Takasu Y (1992) Chem

Express 7:422. Friedrich KA, Marmann A, Stimming U, Unkauf W, Vogel R

(1997) Fresenius J Anal Chem 358:16323. Cherstiouk OV, Simonov PA, Chuvilin AL, Savinova ER (2001)

In: Wieckowski A, Brooman EW, Rudd EJ, Fuller TF, Leddy J(eds) The global climate change: a coordinated response byelectrochemistry and solid-state science and technology ECS. ECSInc., Phoenix

24. Cherstiouk OV, Simonov PA, Savinova ER (2003) ElectrochimicaActa 48:3851

25. Maillard F, Eikerling M, Cherstiouk OV, Schreier S, Savinova E,Stimming U (2004) Faraday Discuss 125:357

26. Watanabe M, Saegusa S, Stonehart P (1988) Chem Lett148727. Yano H, Inukai J, Uchida H, Watanabe M, Babu PK, Kobayashi T,

Chung JH, Oldfield E, Wieckowski A (2006) Phys Chem ChemPhys 8:4932

28. Bucur RV (1987) Acta Met 35:132529. Balasheva NA, Zhmakin GG (1962) Dokladi Akademii Nauk

USSR 143:35830. Pyshnograeva II, Skundin AM, Vassilyev YB, Bagotskii VS

(1969) Electrokhimija 5:146931. Pyshnograeva II, Vassilyev YB, Bagotskii VS (1970) Electrokhimija

6:1545

508 J Solid State Electrochem (2008) 12:497–509

32. Kudryashev IV, Izmajlov AV, Lelikov YA (1975) Zh. Fiz. Khim.(Russian Journal of Physical Chemistry) 49:929

33. Iofa ZA, Batrakov VV, Nikiforova YA (1967) Vestnik MGU,Serija Khim 6:11

34. Gleiter H (1992) Nanostruct Mater 1:135. Birringer R, Gleiter H (1994) In: Haberland H (ed) Clusters of

atoms and molecules. Springer Series in Chemical Physics. vol.56. Springer, Berlin, p. 384

36. Gleiter H (2000) Acta Mater 48:137. Tsybulya SV, Kryukova GN, Goncharova SN, Shmakov AN,

Balzhinimaev BS (1995) J Catal 154:19438. Gamburg YD (1997) Electrodeposited metals. Yanus-K, Moscow39. de Souza JPI, Iwasita T, Nart FC, Vielstich W (2000) J Appl

Electrochem 30:4340. de Lima RB, Paganin V, Iwasita T, Vielstich W (2003) Electro-

chimica Acta 49:8541. Brisard GM, Camargo APM, Nart FC, Iwasita T (2001) Electrochem

Commun 3:60342. Camara GA, de Lima RB, Iwasita T (2004) Electrochem Commun

6:81243. Camara GA, de Lima RB, Iwasita T (2005) J Electroanal Chem

585:12844. de Lima RB, Massafera MP, Batista EA, Iwasita T (2007) J

Electroanal Chem 603:14245. Tsirlina GA, Baronov SB, Spiridonov FM, Rusanova MY,

Safonova TY, Petrii OA (2000) Russ J Electrochem 36:117946. Rusanova MY, Tsirlina GA, Petrii OA, Safonova TY, Vassiliev SY

(2000) Russ J Electrochem 36:45747. Petrii OA, Safonova TY, Tsirlina GA, Rusanova MY (2000)

Electrochim Acta 45:411748. Rusanova MY, Grden M, Czerwinski A, Tsirlina GA, Petrii OA,

Safonova TY (2001) J Solid State Electrochem 5:21249. Lebedeva NP, Kryukova GN, Tsybulya SV, Salanov AN,

Savinova ER (1998) Electrochimica Acta 44:143150. Maillard F, Schreier S, Hanzlik M, Savinova ER, Weinkauf S,

Stimming U (2005) Phys Chem Chem Phys 7:38551. Gavrilov AN, Savinova ER, Simonov PA, Zaikovskii VI,

Cherepanova SV, Tsirlina GA, Parmon VN (2007) Phys ChemChem Phys (Advance Article). doi:10.1039/b707598g

52. Plyasova LM, Molina IY, Gavrilov AN, Cherepanova SV,Cherstiouk OV, Rudina NA, Savinova, ER, Tsirlina GA (2006)Electrochim Acta 51:4477–4488

53. Brauer G (1985) Handbook of preparative inorganic chemistry,vol. 5. Mir, Moscow (Russian translation)

54. Langford JI, Delhez R, Keijser TH, Mittemeijer EJ (1988) Aust JPhys 41:173

55. Sherstyuk OV, Pron'kin SN, Chuvilin AL, Salanov AN, SavinovaER, Tsirlina GA, Petrii OA (2000) Russ J Electrochem 36:741

56. Plyasova LM, Molina IJ, Cherepanova SV, Rudina NA, Cher-stiouk OV, Savinova ER, Pron'kin SN, Tsirlina GA (2002) Russ JElectrochem 38:1116

57. Williamson GK, Smallman RE (1956) Phil Mag 1:3458. Valiev RZ, Alexandrov IV (2000) Nanostructural materials

obtained by severe plastic deformation. Logos, Moscow59. Clavilier J (1980) J Electroanal Chem 107:21160. Lamy C, Leger JM, Clavilier J (1982) J Electroanal Chem 135:32161. Gilman S (1964) J Phys Chem 68:7062. Yates JT (1995) J Vac Sci Technol A-Vacuum Surfaces Films

13:135963. Xu JZ, Yates JT (1993) J Chem Phys 99:725

64. Xu JZ, Henriksen P, Yates JT (1992) J Chem Phys 97:525065. Szabo A, Henderson MA, Yates JT (1992) J Chem Phys 96:619166. Andreaus B, Maillard F, Kocylo J, Savinova ER, Eikerling M

(2006) J Phys Chem B 110:2102867. Cherstiouk OV, Simonov PA, Zaikovskii VI, Savinova ER (2003)

J Electroanal Chem 554:24168. Jiang JH, Kucernak A (2002) J Electroanal Chem 533:15369. Chen W, Sun SG, Zhou ZY, Chen SP (2003) J Phys Chem B

107:980870. Arenz M, Mayrhofer KJJ, Stamenkovic V, Blizanac BB,

Tomoyuki T, Ross PN, Markovic NM (2005) J Am Chem Soc127:6819

71. Markovic NM, Ross PN (2002) Surf Sci Rep 45:12172. Wasmus S, Kuever A (1999) J Electroanal Chem 461:1473. Lamy C, Leger JM (1991) Journal De Chimie Physique Et De

Physico-Chimie Biologique 88:164974. Sriramulu S, Jarvi TD, Stuve EM (1999) In: Wieckowski A (ed)

Interfacial electrochemistry: theory, experiment, and applications.Marcel Dekker, New York, p 793

75. Waszczuk P, Lu GQ, Wieckowski A, Lu C, Rice C, Masel RI(2002) Electrochim Acta 47:3637

76. Hamnett A (1999) In: Wieckowski A (ed) Interfacial electrochem-istry: theory, experiment, and applications. Marcel Dekker, NewYork, p 843

77. Petrii OA, Podlovchenko BI, Frumkin AN, Lal H (1965) JElectroanal Chem 10:253

78. Herrero E, Chrzanowski W, Wieckowski A (1995) J Phys Chem99:10423

79. Bagotzky VS, Vassiliev YB (1967) Electrochim Acta 12:132380. Bagotzky VS, Vassiliev YB, Khazova OA (1977) J Electroanal

Chem 81:22981. Iwasita T, Vielstich W, Santos E (1987) J Electroanal Chem

229:36782. Castro Luna AM, Giordano MC, Arvia AJ (1989) J Electroanal

Chem 259:17383. Beden B, Lamy C, Bewick A, Kunimatsu K (1981) J Electroanal

Chem 121:34384. Iwasita T (2003) In: Vielstich W, Gasteiger HA, Lamm A (eds)

Handbook of fuel cells—fundamentals, technology and applications,vol 2: electrocatalysis. Wiley, Chichester, p p 603

85. Batista EA, Malpass GRP, Motheo AJ, Iwasita T (2004) JElectroanal Chem 571:273

86. Iwasita T, Vielstich W (1986) J Electroanal Chem 201:40387. Wilhelm S, Iwasita T, Vielstich W (1987) J Electroanal Chem

238:38388. Jusys Z, Behm RJ (2001) J Phys Chem B 105:1087489. Jusys Z, Kaiser J, Behm RJ (2003) Langmuir 19:675990. Wang H, Loffler T, Baltruschat H (2001) J Appl Electrochem

31:75991. Wang HS, Wingender C, Baltruschat H, Lopez M, Reetz MT

(2001) J Electroanal Chem 509:16392. Breiter M (1967) Electrochim Acta 12:121393. Batista EA,Malpass GRP,Motheo AJ, Iwasita T (2003) Electrochem

Commun 5:84394. Cao D, Lu G-Q, Wieckowski A, Wasileski SA, Neurock M (2005)

J Phys Chem B 109:1162295. Housmans THM, Wonders AH, Koper MTM (2006) J Phys Chem

B 110:1002196. Maillard F, Savinova ER, Stimming U (2007) J Electroanal Chem

599:221

J Solid State Electrochem (2008) 12:497–509 509