Support effects on the oxidation of ethanol at Pt nanoparticles

Article abstract of DOI:10.1016/j.electacta.2012.01.042

The effects of metal oxide supports on ethanol oxidation have been investigated by drop coating Pt nanoparticles onto glassy carbon electrodes coated with thin layers of ruthenium oxide, tin oxide, a mixed Ru + Sn oxide, and onto an indium-tin oxide (ITO) electrode. All four oxide supports exhibited significant co-catalytic effects, with their effectiveness at low potentials increasing in the order Ru oxide < ITO < Ru + Sn oxide < Sn oxide. However, at higher potentials (e.g. 0.4 V vs. SCE) currents were higher for Pt supported on Ru oxide or Ru + Sn oxide than on Sn oxide, revealing mechanistic differences between the roles of Ru and Sn oxide. Although Sn oxide produced very high initial activities, ITO and Ru + Sn oxide provided more stable performances.

Full text of DOI:10.1016/j.electacta.2012.01.042

Electrochimica Acta 65 (2012) 210–215
Contents lists available at SciVerse ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Support effects on the oxidation of ethanol at Pt nanoparticles
Reza B. Moghaddam, Peter G. Pickup∗
Department of Chemistry, Memorial University of Newfoundland, St. John’s, Newfoundland, Canada A1B 3X7
a r t i c l e i n f o
a b s t r a c t
Article history:
The effects of metal oxide supports on ethanol oxidation have been investigated by drop coating Pt
nanoparticles onto glassy carbon electrodes coated with thin layers of ruthenium oxide, tin oxide, a
mixed Ru + Sn oxide, and onto an indium–tin oxide (ITO) electrode. All four oxide supports exhibited
significant co-catalytic effects, with their effectiveness at low potentials increasing in the order Ru
oxide < ITO < Ru + Sn oxide < Sn oxide. However, at higher potentials (e.g. 0.4 V vs. SCE) currents were
higher for Pt supported on Ru oxide or Ru + Sn oxide than on Sn oxide, revealing mechanistic differences
between the roles of Ru and Sn oxide. Although Sn oxide produced very high initial activities, ITO and
Ru + Sn oxide provided more stable performances.
Received 10 November 2011
Received in revised form 11 January 2012
Accepted 12 January 2012
Available online 20 January 2012
Keywords:
Ethanol
Oxidation
Oxide
© 2012 Elsevier Ltd. All rights reserved.
Support
Platinum
1. Introduction
[14], TiO2 containing polyoxometallate modified Au nanoparticle
[15], Ru containing pervosites [16] and Sb2O5·SnO2 [17] supports
The electrochemical oxidation of ethanol is currently attracting
considerable interest due to its central importance in the devel-
opment of direct ethanol fuel cells (DEFCs) [1–5]. A 2007 review of
electrocatalysts for ethanol oxidation concluded that PtSn catalysts
were the most active binary systems in acidic media, while addition
of Ru resulted in the most promising ternary systems [3]. How-
ever, these systems promote the incomplete oxidation of ethanol
to acetaldehyde and acetic acid in favour of its complete oxidation
to carbon dioxide [3,6,7]. This leads to very low fuel efficiencies
in DEFCs [2] and high levels of byproducts that would need to be
destroyed or recycled.
The key challenge in developing catalysts for ethanol oxida-
tion is to find systems that can efficiently break the C C bond so
that the complete oxidation to CO2 can occur. To date, the high-
est yields of CO2 have been obtained with pure Pt catalysts [7],
which have low activities. Deposition of Pt on Sn oxide has been
reported to increase its activity relative to codeposited PtSn [8],
while a ternary PtRhSnO2/C catalyst has been reported to split
the C C bond in ethanol at ambient temperature with close to
50% efficiency [9,10]. Density functional calculations indicated that
interactions between the PtRh alloy and SnO2 play a crucial role
in C C bond breaking [9]. A study of the effects of SnOx islands on
ethanol oxidation at a Pt(1 1 1) surface has suggested that the Sn
oxide promotes the removal of CO at the Pt/SnOx interfaces [11].
SnInO [12], SnO2 treated with sulphuric acid [13], TiO2 nano-rods
have also been shown to be effective for ethanol oxidation at Pt
nanoparticles. These results suggest that the best approach to the
development of high efficiency catalysts for ethanol oxidation may
be through the use of oxide supports, since it is clear that the inter-
actions between various oxides and Pt can significantly increase
the current density for ethanol oxidation, and there is evidence that
this can occur without greatly inhibiting its complete oxidation to
carbon dioxide [9,10].
We have recently demonstrated that support effects in elec-
trocatalysis can be unambiguously identified and quantitatively
compared by applying monolayer quantities of preformed cata-
lyst nanoparticles onto modified glassy carbon electrodes [18,19].
For example, it was shown that a thin layer of Ru oxide on the
glassy carbon strongly promoted the oxidation of methanol at Pt,
while a layer of polyaniline had mainly inhibitory effects [18]. This
methodology has been applied here to investigate and compare the
effects of various oxides on the oxidation of ethanol at supported Pt
nanoparticles. Glassy carbon was used as a control support, while
indium–tin oxide (ITO) and thin layers or Ru oxide, Sn oxide, and a
mixed Ru Sn oxide on glassy carbon are shown to provide activat-
ing support effects.
2. Experimental
2.1. Chemicals
Sulphuric acid (Fisher Scientific), anhydrous ethanol (Com-
mercial Alcohols Inc.), KRuO4 (Alfa Aesar), SnCl4·5H2O (Fisher
Scientific), H2PtCl6·6H2O (Alfa Aesar), potassium hydroxide (ACP
∗
Corresponding author. Tel.: +1 709 864 8657; fax: +1 709 864 3702.
E-mail address: ppickup@mun.ca (P.G. Pickup).
0013-4686/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2012.01.042

R.B. Moghaddam, P.G. Pickup / Electrochimica Acta 65 (2012) 210–215
211
Chemical Inc.), sodium citrate (Anachemia), sodium borohydride
(Sigma–Aldrich), and NafionTM solution (5%; Dupont) were used as
received. All measurements were recorded at ambient temperature
under a nitrogen atmosphere following purging for 15 min.
2.2. Preparation of Pt nanoparticles
NaBH4(aq) (1.5 mL; 120 mM) was added dropwise to a stirred
solution of 10 mL of 3 mM H2PtCl6(aq) mixed with 0.6 mL of 50 mM
aqueous sodium citrate [20]. Following stirring for a further 2 h,
the resulting grey colloidal Pt nanoparticle solution was stored in
a fridge. This stock solution was diluted by a factor of 20 prior to
use with water and Nafion solution to give 1.2 ␮g mL−1 Nafion. X-
ray diffraction measurements indicated that the average particle
diameter was 5.0 0.4 nm.
2.3. Working electrode preparation
Glassy carbon electrodes (GC; CH Instruments; 0.071 cm2)
were polished with 0.05 ␮m alumina and rinsed well with water
before use. Electrodes were coated with a thin film (ca. 25 nm)
of hydrous Ru oxide (GC/Ru oxide) as previously described [18].
Spontaneous deposition was also used to prepare GC/Sn oxide
electrodes, by immersion of preconditioned (E = −0.5 V, t = 300 s
in 0.1 M H2SO4) GC electrodes in a neutral aqueous SnCl4 solu-
tion (0.05 M) for 15 min. Spontaneous codeposition was used to
prepare GC/RuxSnx−1 oxide electrodes, whereby a preconditioned
(E = −0.5 V, t = 300 s in 0.1 M H2SO4) GC electrode was placed in
0.05 M KRuO4 + 0.05 M SnCl4 (in 0.1 M KOH) for 15 min. The GC, ITO
(on glass; Donnelly Corp.), GC/Sn oxide and GC/RuxSn1−x oxide/Pt
electrodes were then drop coated with 12.5 ␮L of the diluted Nafion
containing Pt colloid to obtain GC/Pt, ITO/Pt, GC/Ru oxide/Pt, GC/Sn
oxide/Pt and GC/RuxSn1−x oxide/Pt electrodes with Pt loadings of
3.0 × 10−7 g (4.3 ␮g cm−2) with 5% Nafion by mass relative to Pt.
Analysis of two of these electrodes by ICP-MS gave the following
results:
Fig. 1. TEM of the Pt nanoparticles.
3. Results and discussion
3.1. Characterization of the nanoparticles and electrodes
Fig. 1 shows a TEM image of the Pt nanoparticles. There is a
relatively narrow size distribution with an average diameter of
5.0 1.3 nm that is consistent with the XRD measurement. Based
on the surface area to mass ratio calculated for particles of this
size, the loading of 3.0 × 10−7 g Pt applied to each electrode would
correspond to a real geometric Pt surface area of 0.17 cm2.
Fig. 2 shows SEM and AFM images of a GC plate coated with
3.0 × 10−7 g of Pt nanoparticles with 5% Nafion by mass follow-
ing washing with deionized water to remove residual salts from
the synthesis solution. It can be seen from the SEM image that
the Pt + Nafion mixture formed an uneven coating, as a result of
which some regions consisted of aggregates and some regions were
bare. AFM confirmed the uneven Pt + Nafion distribution, display-
ing aggregates as high as 70 nm. The average layer thickness of
14 nm seen in Fig. 2B is reasonably consistent with the quantity
of Pt applied to the electrode, which was roughly equivalent to one
close-packed monolayer of 5 nm particles, and the volume ratio of
Pt to Nafion of ca. 1:1.
GC/Sn oxide/Pt: Sn = 0.33–0.34 ␮g, Pt = 0.39 ␮g
GC/RuxSnx−1 oxide/Pt: Ru = 1.5 ␮g, Sn = 0.32 ␮g, Pt = 0.41 ␮g
The RuxSnx−1 oxide composition was therefore Ru0.85Sn0.15
oxide, and the Pt loadings were close to the targeted value. Based
on the densities of SnO2 and RuO2 (6.95 g cm−3 and 6.97 g cm−3
,
respectively), the thicknesses of the Sn oxide and Ru0.85Sn0.15 oxide
films can be estimated to have been ca. 9 nm and 40 nm, respec-
tively.
Fig. 3 shows overlaid cyclic voltammograms of GC/Pt, ITO/Pt,
GC/Ru oxide/Pt, GC/Sn oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt elec-
trodes in 0.1 M H2SO4. The forward scan (anodic) for the GC/Pt
electrodes shows peaks centered at −0.13 and −0.08 V assigned to
hydrogen desorption, followed by the onset of Pt oxide (PtO) for-
mation from ∼0.4 V. A well-defined peak for PtO reduction can be
seen at ca. 0.47 V, following which hydrogen adsorption and hydro-
gen evolution occur beyond −0.04 V to terminate the cathodic
scan. These features are generally similar for the other electrodes.
Nevertheless, the GC/Ru oxide/Pt, GC/Sn oxide/Pt, and particularly
the GC/Ru0.85Sn0.15 oxide/Pt show larger background currents, and
the signals assigned to hydrogen desorption/adsorption were sup-
pressed in some cases. In contrast, ITO/Pt had a low background
current and its voltammogram was most similar to that of GC/Pt.
Electrochemically active areas estimated from the H adsorp-
tion waves in Fig. 3 were 0.078, 0.071, 0.055, 0.043, and 0.078 cm2,
respectively, for the GC/Pt, ITO/Pt, GC/Ru oxide/Pt, GC/Sn oxide/Pt
and GC/Ru0.85Sn0.15 oxide/Pt electrodes. These values corre-
spond to electrochemical utilizations (active area/geometric area)
2.4. Instrumentation
An EG&G Model 273A Potentiostat/Galvanostat run by a PC
through M270 commercial software was used for voltammetric
measurements. A saturated calomel electrode (SCE) and a platinum
wire formed the reference and counter electrodes, respectively.
A Model FEI Quanta 400 environmental SEM was used to
perform scanning electron microscopy (SEM) measurements. For
transmission electron microscopy (TEM), the colloidal Pt nanopar-
ticle solution was diluted with ethanol and sonicated for some
minutes. A drop of the mixture was then placed on a 200 mesh
copper grid with a carbon support film. The grid was allowed to dry
overnight and then observed under a Model JEOL 2011 transmission
electron microscope (The Microscopy and Microanalysis Facility,
University of New Brunswick). AFM measurements were conducted
with a Quesant Q-Scope 350 using a smooth glassy carbon plate (SPI
Supplies, USA) as the substrate.

212
R.B. Moghaddam, P.G. Pickup / Electrochimica Acta 65 (2012) 210–215
Fig. 4. Cyclic voltammogram (10 mV s−1; 1st scan solid, 2nd scan dashed) in 0.1 M
H2SO4 containing 0.2 M ethanol of a GC electrode coated with 4.3 ␮g cm−2 of Pt
nanoparticles.
ranging from 26% to 46%. Previously, it has been shown that use of
the hydrogen adsorption waves for the type of GC/Pt electrode used
here underestimates the electrochemically active Pt area relative
to use of the oxide stripping peak or Cu underpotential deposition
[18]. It is therefore not clear that the differences in apparent utiliza-
tions between electrodes observed here (in Fig. 3) are meaningful,
or even significant relative to the experimental uncertainty.
3.2. Ethanol oxidation
For each electrode, cyclic voltammetry in the presence of
ethanol was conducted immediately following the cyclic voltam-
metry in 0.1 M H2SO4 shown in Fig. 3. The electrode, initially at
−0.3 V, was raised out of the electrolyte solution while ethanol
was added and then immersed into the mixed 0.1 M H2SO4 + 0.2 M
ethanol solution. A nitrogen atmosphere was maintained during
this procedure which typically took ca. 30 s.
Fig. 4 shows the first two CV scans for a GC/Pt electrode in the
presence of ethanol. The onset of ethanol oxidation during the for-
ward scan of the first cycle is very early, at ca. 0 V. A small peak
appears at ca. −0.05 V, followed by a rising current to a dominant
oxidation peak centered at 0.56 V. There is a weakly defined shoul-
der at about 0.38 V and a second peak at +0.95 V. The reverse scan
of the first cycle consists of a large anodic peak at 0.37 V that can
be attributed to rapid oxidation of ethanol at the clean Pt produced
by reduction of the Pt oxide layer formed at higher potentials. On
the second cycle, the onset of ethanol oxidation is shifted to a much
higher potential (ca. 0.3 V) and the shoulder at ca. 0.38 V is absent.
This is discussed in Section 3.3.
Fig. 2. SEM (A) and AFM (B; 2 × 2 ␮m × 74 nm) images of ca. 4.3 ␮g cm−2 of Pt
nanoparticles + Nafion on GC.
Fig. 5 compares the first forward scans for ethanol oxidation at
GC/Pt, ITO/Pt, GC/Ru oxide/Pt, GC/Sn oxide/Pt and GC/Ru0.85Sn0.15
oxide/Pt electrodes. The reverse scans were not notably influenced
by the presence of the oxide layers, and so are not presented. At
potentials below ∼0.4 V, the GC/Sn oxide/Pt and GC/Ru0.85Sn0.15
oxide/Pt electrodes gave markedly higher currents than the others,
with both showing a large oxidation peak at ca. 0.3 V. The response
of the GC/Ru oxide/Pt electrode in this region was somewhat better
than that at the GC/Pt electrode, while ITO/Pt was somewhat better
than GC/Ru oxide/Pt. These results clearly indicate that Sn oxide
facilitates the decomposition of ethanol and/or has a synergetic/co-
catalytic effect on Pt in the oxidation of ethanol. The presence of
Ru in the mixed Ru0.85Sn0.15 oxide does not significantly change
the effect of the Sn, while the presence of In in the ITO appears to
diminish its effect.
Fig. 3. Cyclic voltammograms (10 mV s−1) in 0.1 M H2SO4 of GC, ITO, GC/Ru oxide,
GC/Sn oxide and GC/Ru0.85Sn0.15 oxide electrodes coated with 4.3 ␮g cm−2 of Pt
nanoparticles.

R.B. Moghaddam, P.G. Pickup / Electrochimica Acta 65 (2012) 210–215
213
Fig. 5. First anodic scans of cyclic voltammograms (10 mV s−1) for GC/Pt, ITO/Pt,
GC/Ru oxide/Pt, GC/Sn oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt electrodes in 0.1 M
H2SO4 containing 0.2 M ethanol.
Fig. 6. Second anodic scans of cyclic voltammograms (10 mV s−1) for GC/Pt, ITO/Pt,
GC/Ru oxide/Pt, GC/Sn oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt electrodes in 0.1 M
H2SO4 containing 0.2 M ethanol.
0.1 V, 0.2 V, 0.3 V, and 0.4 V, but only the 0.1 V and 0.4 V results are
shown here.
The striking contrast between the effects of Ru oxide and Sn
oxide seen in Fig. 5 has important mechanistic implications. Since
Ru oxide is known to be very effective at shifting the onset of CO
oxidation to lower potentials, its relatively small influence here
implies that poisoning of the Pt by CO was not severe at potentials
below 0.4 V. Thus the role of Sn may not be related to its ability to
inhibit CO formation or promote CO oxidation.
Inspection of the linear sweep voltammograms in Fig. 5 over
the 0.4–0.8 V range reveals additional insight into the roles of the
metal oxides. In all cases, the peak at ca. 0.55 V can be primarily
attributed to the oxidation of ethanol to CO2 through an adsorbed
CO intermediate on the Pt nanoparticles. The height of this peak
shows a rough inverse correlation with the onset of ethanol oxida-
tion, with GC/Pt and GC/Ru oxide/Pt showing the highest currents in
the region, while GC/Sn oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt gave
the lowest. This may suggest that the presence of Sn promotes the
direct oxidation of ethanol to acetaldehyde and acetic acid over the
formation of CO2 via an adsorbed CO intermediate.
Although the results shown in Fig. 5 provide some important
insights into the mechanistic role of the Sn oxide supports, they
are not very relevant to their use in fuel cells where high steady
state ethanol oxidation rates are required. In all cases, currents at
potentials below ca. 0.4 V were diminished considerably on the sec-
ond cycle, as illustrated for CG/Pt in Fig. 4. For GC/Pt, ITO/Pt, GC/Ru
oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt electrodes, the changes after
the 2nd cycle were relatively small (not shown). However, for
GC/Sn oxide/Pt electrodes the currents below ca. 0.4 V continued
to decrease significantly over many cycles. Nonetheless, the supe-
rior performances of the ITO/Pt, GC/Sn oxide/Pt and GC/Ru0.85Sn0.15
oxide/Pt electrodes over GC/Pt and GC/Ru oxide/Pt were main-
tained over multi cycles. This is illustrated in Fig. 6 which shows
the 2nd cycles of the voltammograms for all of the electrodes.
These changes in the cyclic voltammograms with cycling make
it very difficult to accurately assess the effects of the various oxide
supports on ethanol oxidation at the Pt nanoparticles. More defini-
tive comparisons were therefore obtained by chronoamperometry.
A voltammogram of each electrode was obtained in 0.1 M H2SO4
prior to these experiments but, in order to preserve their initial
activity, no voltammetry was performed in the presence of ethanol.
Sequential chronoamperometry experiments were then run at 0 V,
Chronoamperometric oxidation of ethanol at 0.1 V at GC/Pt,
ITO/Pt, GC/Ru oxide/Pt, GC/Sn oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt
electrodes is presented in Fig. 7. Consistent with the CVs, GC/Pt
and GC/Ru oxide/Pt were much less active at 0.1 V than the ITO/Pt,
GC/Sn oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt electrodes. The GC/Pt
electrode initially gave a higher current than the GC/Ru oxide/Pt
electrode but its current decayed more rapidly. The superior longer
term activity of the GC/Ru oxide/Pt electrode is consistent with the
ability of Ru oxide to facilitate the oxidation of CO, which is known
to poison Pt during ethanol oxidation.
The presence of Sn oxide in the support for the Pt nanoparti-
cles greatly enhanced their activity for ethanol oxidation at 0.1 V,
with the greatest effect seen for the GC/Sn oxide/Pt electrode
in Fig. 7. The activity of the GC/Ru0.85Sn0.15 oxide/Pt electrode
was only slightly lower, while the activity of the ITO/Pt electrode
was significantly lower. Current decay rates were similar for the
GC/Ru oxide/Pt, GC/Sn oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt elec-
trodes, but slightly higher for the ITO/Pt electrode. Fig. 8 presents
chronoamperometric profiles at a more positive potential of 0.4 V.
Here, the GC/Pt electrode gave the lowest current at all times.
For times less than ca. 150 s, the electrodes with oxide supports
Fig. 7. Chronoamperometry at +0.1 V for GC/Pt, ITO/Pt, GC/Ru oxide/Pt, GC/Sn
oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt electrodes in 0.1 M H2SO4 containing 0.2 M
ethanol.

214
R.B. Moghaddam, P.G. Pickup / Electrochimica Acta 65 (2012) 210–215
that are oxidized to CO2, 2.2–5.9 electrons are involved per CO2
molecule, depending on the adsorption and stripping potentials
[24]. This indicates that various oxidizable Pt CHx and Pt CHyO
species are formed during the oxidative adsorption of ethanol on
Pt.
In light of the above discussion, the large differences between
the first and subsequent anodic voltammetric scans for ethanol
oxidation seen in this work (e.g. Fig. 4) can be explained by the
changing populations of the various adsorbates that are formed.
Before the first scan, it is clear that a diverse range of one and two
carbon adsorbates would be present. The dominate species would
be Pt CO, which is oxidized to CO2 during the main peak seen at ca.
+0.55 V for scans at all electrodes. The small peak seen at ca.−0.05 V
on the first scan for all electrodes may be due to adsorbed H, while
the broad shoulders are presumably due to the oxidation of species
Fig. 8. Chronoamperometry at +0.4 V for GC/Pt, ITO/Pt, GC/Ru oxide/Pt, GC/Sn
oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt electrodes in 0.1 M H2SO4 containing 0.2 M
ethanol.
such as Pt
O CH2CH3, Pt CH(OH)CH3, Pt CH3 and Pt CH2O. The
absence of these features on the second anodic scan at the GC/Pt
electrode (Fig. 4) suggests that CO is the predominant adsorbate
formed during the cathodic scan. The fact that the broad shoul-
der prior to the +0.55 V peak persists to varying extents for the
ITO/Pt, GC/Ru oxide/Pt, GC/Sn oxide/Pt and GC/Ru0.85Sn0.15 oxide/Pt
electrodes (Fig. 6) indicates that the oxide supports facilitate the
oxidation of adsorbed CO and/or may play other roles in promoting
ethanol oxidation.
followed the same activity trend as they did at 0.1 V. At short
times the activity of the GC/Sn oxide/Pt electrode was much
higher than for the others, but it decayed much more rapidly. The
GC/Ru0.85Sn0.15 oxide/Pt electrode also had a relatively high ini-
tial activity and its decay rate was lower than for GC/Sn oxide/Pt.
Consequently, it gave the best longer term (>150 s) activity. ITO/Pt
gave a lower initial current but exhibited the lowest decay rate,
while GC/Ru oxide/Pt was inferior both in terms of initial current
and decay rate.
The role of the Ru oxide support is reasonably clear because it
is consistent with the well known bifunctional mechanism for CO
oxidation at Pt Ru electrodes, as shown in reaction (7).
Pt CO + Ru OH → Pt + Ru + CO2 + H+ + e−
(7)
3.3. Mechanistic insights
Since Ru OH forms at lower potentials than Pt OH, reaction
(7) occurs at lower potentials than reaction (6). It has been pre-
viously demonstrated that this occurs for methanol oxidation at
GC/Ru oxide/Pt electrodes, where the predominant adsorbate is CO
[18]. The shift in the oxidation wave in Fig. 6 for GC/Ru oxide/Pt
relative to GC/Pt is similar to that seen for methanol oxidation,
suggesting a similar mechanistic role for the Ru oxide support.
In the following discussion of the literature on the activities and
mechanisms of PtRu and PtSn catalysts and the roles of the various
oxide supports employed here, we use the term PtM (where M is
Ru or Sn) to refer to results for either PtM alloys or various types
of mixed Pt + M oxide catalysts. Although this may result in some
inaccuracies in the comparisons [26], it reflects the observation that
these two types of catalyst system generally behave very similarly,
and appear to operate by similar mechanisms. This is due to the fact
that Ru and Sn at the surfaces of alloy catalysts are usually present
in the form of oxides [26].
The electrocatalytic oxidation of ethanol at Pt begins with its
oxidative adsorption (reactions (1) and (2)) followed by its oxida-
tive dissociation to a mixture of CHx and oxygen containing
species (reactions (3) and (4)) [21–23].
CH3CH2OH + Pt → Pt
CH3CH2OH + Pt → Pt CH(OH)CH3 + H+ + e−
O
CH2CH3 + H+ + e−
(1)
(2)
(3)
Pt
O
CH2CH3 or Pt CH(OH)CH3 → Pt CHOCH3 + H+ + e−
Pt
O
CH2CH3 or Pt CH(OH)CH3 → Pt CHx + Pt CHyO
+ (5−x−y)H+ + (5−x−y)e−
(4)
Both the adsorbed Pt CHx and Pt CHyO type species can be
further oxidized to adsorbed CO (e.g. reaction (5)) and then to CO2
(reaction (6)).
Although PtRu systems can promote ethanol oxidation by the
bifunctional mechanism shown in reaction (7), it is well known
that PtSn systems are more effective promoters. In contrast, PtRu
systems are much more effective for methanol oxidation than PtSn
systems [27,28]. These differences can possibly be explained by the
observation that Sn species (oxides) promote CO oxidation by elec-
tronic effects (ligand effects) that weaken the Pt CO bond, rather
than the bifunctional mechanism [26].
For ethanol oxidation, it has been reported that PtSn is
more effective than PtRu at low potentials while PtRu becomes
more effective at higher potentials [27,29]. The results shown in
Figs. 6, 7 and 8 are consistent with this, with the GC/Ru oxide/Pt
electrode showing the largest currents at potentials above ca. 0.4 V
(Fig. 6), but all of the Sn oxide containing electrodes being more
effective at 0.4 V and lower. Although we cannot yet provide a full
explanation of these results, it does appear that the electronic effect
of Sn oxide containing supports contribute to the removal of COads
at lower potentials than the bifunctional effect of the Ru oxide
support. In terms of initial activity at low potentials, the pure Sn
oxide support was better than either of the mixed oxides, which
Pt CH2O → Pt CO + 2H+ + 2e−
Pt CO + Pt OH → 2Pt + CO2 + H+ + e−
(5)
(6)
The oxidative adsorption of ethanol at Pt is generally followed
by in situ vibrational spectroscopy [23], usually by monitoring the
adsorbed CO produced by reactions such as (5). Complementary
information can be obtained by measuring the CO2 produced by
oxidative stripping of adsorbates by differential electrochemical
mass spectrometry (DEMS) [24,25], and from quantum chemistry
calculations [22]. The oxidative adsorption of ethanol on Pt can be
completely suppressed by underpotentially deposited hydrogen at
low potentials (e.g. 0.06 V vs. RHE [24]), but produces substan-
tial coverages of CO and other adsorbates at higher potentials
[21,23,24]. Oxidation of these adsorbates to CO2 begins at ca. 0.5 V
vs. RHE [24] (∼0.25 V vs. SCE), which coincides approximately
with a decrease in the intensity of the vibrational response due to
adsorbed CO [21,23]. However, the amount of CO2 produced corre-
sponds to ≤60% of a CO monolayer, indicating the presence of other
adsorbates that cannot readily be oxidized. For the adsorbed species

R.B. Moghaddam, P.G. Pickup / Electrochimica Acta 65 (2012) 210–215
215
can reasonably be attributed to the higher Sn concentration at the
Pt surface. However, the presence of either In or Ru in the oxide sig-
nificantly stabilizes the performance at potentials above ca. 0.1 V
(see Fig. 8), and makes ITO and RuxSnx−1 oxide supports attractive
for further investigation.
electronic (ligand) effects for Sn oxide and mixed oxides containing
Sn.
Acknowledgements
Finally, it is necessary to consider how the oxide supports
employed here are able to induce similar bifunctional and elec-
tronic effects to those previously documented for PtM alloys and
mixed Pt + M oxide catalysts. The bifunctional effects are addressed
by work on systems in which the distance between catalyst pairs
has been experimentally controlled. For example, Abruna et al.
were able to demonstrate “electrocatalytic synergy” in the oxida-
tion of CO between an Ru scanning tunnelling microscopy tip and a
Pt substrate, via the bifunctional mechanism shown in Eq. (7) [32].
CO adsorbed on the Pt was oxidized at lower overpotentials when
the Ru tip was brought close to its surface, with density functional
theory calculations indicating that a critical distance of less than
ca. 0.4 nm would be required for reaction (7) to occur. Sustained
CO oxidation currents were observed due to surface diffusion of CO
to the Ru tip. It is clear from these results that bifunctional oxidation
of adsorbed CO on Pt nanoparticles will occur in a sustained fashion
whenever there is contact (or even very close proximity) between
any regions of the particles and a suitable metal oxide, regardless of
whether the oxide is deposited onto the particles [11] or whether
the particles are deposited on a metal oxide support. In addition, it
is possible for particles of two different metals (or an oxide and a
metal), such as Pt and Ru, in close proximity to exhibit synergistic
effects due to chemical transfer of metal atoms via dissolution and
redeposition [33].
This work was supported by the Natural Sciences and Engi-
neering Research Council of Canada and Memorial University. We
also thank Dr. Louise Weaver at The Microscopy and Microanal-
ysis Facility, University of New Brunswick, for making the TEM
measurements.
References
[1] C. Lamy, A. Lima, V. LeRhun, F. Delime, C. Coutanceau, J.M. Leger, J. Power
Sources 105 (2002) 283.
[2] F. Vigier, S. Rousseau, C. Coutanceau, J.M. Leger, C. Lamy, Top. Catal. 40 (2006)
111.
[3] E. Antolini, J. Power Sources 170 (2007) 1.
[4] S.Q. Song, V. Maragou, P. Tsiakaras, J. Fuel Cell Sci. Technol. 4 (2007) 203.
[5] U.B. Demirci, J. Power Sources 169 (2007) 239.
[6] S. Rousseau, C. Coutanceau, C. Lamy, J.M. Leger, J. Power Sources 158 (2006) 18.
[7] A. Ghumman, G. Li, D.V. Bennett, P.G. Pickup, J. Power Sources 194 (2009)
286.
[8] L.H. Jiang, G.Q. Sun, S.G. Sun, J.G. Liu, S.H. Tang, H.Q. Li, B. Zhou, Q. Xin, Elec-
trochim. Acta 50 (2005) 5384.
[9] 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) 325.
[10] M. Li, A. Kowal, K. Sasaki, N. Marinkovic, D. Su, E. Korach, P. Liu, R.R. Adzic,
Electrochim. Acta 55 (2010) 4331.
[11] W.P. Zhou, S. Axnanda, M.G. White, R.R. Adzic, J. Hrbek, J. Phys. Chem. C 115
(2011) 16467.
[12] J. Mann, N. Yao, A.B. Bocarsly, Langmuir 22 (2006) 10432.
[13] X.W. Zhang, H. Zhu, Z.J. Guo, Y.S. Wei, F.H. Wang, J. Power Sources 196 (2011)
3048.
Electronic effects between oxide supports and Pt nanoparticles
have been extensively investigated [34 and references therein].
For example, in the case on Pt on TiO2 [34] partial charge transfer
from the substrate to Pt has been observed by X-ray photoelectron
spectroscopy, which was correlated to increased activity to oxygen
reduction.
[14] X.S. He, C.G. Hu, J. Power Sources 196 (2011) 3119.
[15] S. Zoladek, I.A. Rutkowska, P.J. Kulesza, Appl. Surf. Sci. 257 (2011) 8205.
[16] A.D. Lan, A.S. Mukasyan, Ind. Eng. Chem. Res. 47 (2008) 8989.
[17] A.O. Neto, M. Brandalise, R.R. Dias, J.M.S. Ayoub, A.C. Silva, J.C. Penteado, M.
Linardi, E.V. Spinace, Int. J. Hydrogen Energy 35 (2010) 9177.
[18] R.B. Moghaddam, P.G. Pickup, Electrochem. Commun. 13 (2011) 704.
[19] R.B. Moghaddam, P.G. Pickup, Electrocatalysis 2 (2011) 159.
[20] J. Yang, J.Y. Lee, H.-P. Too, Anal. Chim. Acta 571 (2006) 206.
[21] F. Vigier, C. Coutanceau, A. Perrard, E.M. Belgsir, C. Lamy, J. Appl. Electrochem.
34 (2004) 439.
4. Conclusions
[22] H.F. Wang, Z.P. Liu, J. Am. Chem. Soc. 130 (2008) 10996.
[23] R.B. Kutz, B. Braunschweig, P. Mukherjee, R.L. Behrens, D.D. Dlott, A. Wieck-
owski, J. Catal. 278 (2011) 181.
[24] H. Wang, Z. Jusys, R.J. Behm, Fuel Cells 4 (2004) 113.
[25] H.S. Wang, H.D. Abruna, in: A. Bocarsly, D.M.P. Mingos (Eds.), Structure and
Bonding: Fuel Cells and Hydrogen Storage, vol. 141, Springer-Verlag, Berlin,
2011, p. 33.
[26] F.J. Scott, S. Mukerjee, D.E. Ramaker, J. Phys. Chem. C 114 (2010) 442.
[27] W.J. Zhou, B. Zhou, W.Z. Li, Z.H. Zhou, S.Q. Song, G.Q. Sun, Q. Xin, S. Douvartzides,
A. Goula, P. Tsiakaras, J. Power Sources 126 (2004) 16.
[28] G. Li, P.G. Pickup, J. Power Sources 173 (2007) 121.
[29] H.Q. Li, G.Q. Sun, L. Cao, L.H. Jiang, Q. Xin, Electrochim. Acta 52 (2007) 6622.
[30] E. Antolini, E.R. Gonzalez, Catal. Today 160 (2011) 28.
[31] C.W. Liu, Y.W. Chang, Y.C. Wei, K.W. Wang, Electrochim. Acta 56 (2011) 2574.
[32] L. Zhuang, J. Jin, H.D. Abruna, J. Am. Chem. Soc. 129 (2007) 11033.
[33] B. Wickman, Y.E. Seidel, Z. Jusys, B. Kasemo, R.J. Behm, ACS Nano 5 (2011)
2547.
The results presented here show unambiguously that both Ru
and Sn containing oxide supports promote the oxidation of ethanol
at Pt nanoparticles relative to a glassy carbon support. The method-
ology employed allows clear comparisons to be made between
supports and will aid in the understanding of real catalysts which
contain complex mixtures of alloyed and oxidized components
[3,30,31] and for which compositions and structures are less cer-
tain. The use of preformed pure Pt nanoparticles ensures that the
support does not influence their composition, particle size and dis-
tribution, or shape, while the use of thin support layers and Pt
nanoparticle layers negates transport effects. The observed differ-
ences between the effects of Sn oxide and Ru oxide can therefore
be attributed to differences in catalytic mechanisms, which appear
to be dominated by a bifunctional mechanism for Ru oxide and
[34] A. Lewera, L. Timperman, A. Roguska, N. Alonso-Vante, J. Phys. Chem. C 115
(2011) 20153.