Adsorption and oxidation of ethanol on colloid-based Pt/C, PtRu/C and Pt3Sn/C catalysts: In situ FTIR spectroscopy and on-line DEMS studies

Article abstract of DOI:10.1039/b700676b

The interaction of colloid-based, carbon supported Pt/C (40 wt%), PtRu/C (45 wt%) and Pt₃Sn/C (24 wt%) catalysts with ethanol and their performance for ethanol electrooxidation were investigated in model studies by electrochemical, in situ infrared spectroscopy and on-line differential electrochemical mass spectrometry measurements. The combined application of in situ spectroscopic techniques on realistic catalysts and under realistic reaction (DEMS, IR) and transport conditions (DEMS) yields new insight on mechanistic details of the reaction on these catalysts under the above reaction and transport conditions. Based on these results, the addition of Sn or Ru, though beneficial for the overall activity for ethanol oxidation, does not enhance the activity for C-C bond breaking. Dissociative adsorption of ethanol to form CO₂ is more facile on the Pt/C catalyst than on PtRu/C and Pt₃Sn/C catalysts within the potential range of technical interests (<0.6 V), but Pt/C is rapidly blocked by an inhibiting CO adlayer. In all cases acetaldehyde and acetic acid are dominant products, CO₂ formation contributes less than 2% to the total current. The higher ethanol oxidation current density on the Pt₃Sn/C catalyst at these potentials results from higher yields of C2 products, not from an improved complete ethanol oxidation to CO₂. the Owner Societies.

Full text of DOI:10.1039/b700676b

View Article Online / Journal Homepage / Table of Contents for this issue

PAPER
www.rsc.org/pccp | Physical Chemistry Chemical Physics
View Article Online
Adsorption and oxidation of ethanol on colloid-based Pt/C, PtRu/C and
Pt3Sn/C catalysts: In situ FTIR spectroscopy and on-line DEMS
studiesw
Q. Wang,a G. Q. Sun,*a L. H. Jiang,ab Q. Xin,a S. G. Sun,*c Y. X. Jiang,c
S. P. Chen,c Z. Jusysb and R. J. Behm*b
Received 15th January 2007, Accepted 18th April 2007
First published as an Advance Article on the web 8th May 2007
DOI: 10.1039/b700676b
The interaction of colloid-based, carbon supported Pt/C (40 wt%), PtRu/C (45 wt%) and Pt3Sn/
C (24 wt%) catalysts with ethanol and their performance for ethanol electrooxidation were
investigated in model studies by electrochemical, in situ infrared spectroscopy and on-line
differential electrochemical mass spectrometry measurements. The combined application of in situ
spectroscopic techniques on realistic catalysts and under realistic reaction (DEMS, IR) and
transport conditions (DEMS) yields new insight on mechanistic details of the reaction on these
catalysts under the above reaction and transport conditions. Based on these results, the addition
of Sn or Ru, though beneficial for the overall activity for ethanol oxidation, does not enhance the
activity for C–C bond breaking. Dissociative adsorption of ethanol to form CO2 is more facile on
the Pt/C catalyst than on PtRu/C and Pt3Sn/C catalysts within the potential range of technical
interests (o0.6 V), but Pt/C is rapidly blocked by an inhibiting CO adlayer. In all cases
acetaldehyde and acetic acid are dominant products, CO2 formation contributes less than 2% to
the total current. The higher ethanol oxidation current density on the Pt3Sn/C catalyst at these
potentials results from higher yields of C2 products, not from an improved complete ethanol
oxidation to CO2.
modern spectroscopy techniques, e.g., in situ infrared spectro-
1. Introduction
scopy (FTIR) and on-line differential electrochemical mass
spectrometry (DEMS) techniques.7,8,15 If applied, however,
In the last decade, ethanol has attracted considerable interest
as a fuel for direct alcohol fuel cells due to its lower toxicity
and higher energy density (8.0 kW h kgÀ1) compared with
massive electrodes and under idealized conditions,4,16,17 hardly
these spectroscopic techniques were used for model studies on
methanol.1,2 In addition, ethanol can be easily produced in
ever on supported catalysts and under more realistic reaction
large amounts by fermentation of sugar-containing raw ma-
and transport conditions.
terials. The slow kinetics of ethanol electrooxidation at low
This is subject of the present spectro-electrochemical study,
temperatures, however, is still an obstacle for the further
where we investigated the interaction of ethanol with carbon
development of direct ethanol fuel cells (DEFCs).3 It is well
supported Pt/C, PtRu/C and Pt3Sn/C catalysts by a combina-
known that platinum is an active catalyst for the electrooxida-
tion of different in situ spectroscopic techniques, in situ FTIR
tion of many small organic molecules. However, at low
and on-line DEMS, and electrochemical measurements under
temperatures, pure platinum is not an effective anode catalyst
continuous reaction (DEMS, FTIR) and continuous flow
for ethanol electrooxidation, because it is readily poisoned by
(DEMS) conditions. Up to now, PtRu and Pt3Sn have been
strongly adsorbed intermediates, especially by adsorbed
identified as the most effective catalysts for the ethanol oxida-
CO.4–13 To improve the mechanistic understanding of this
tion reaction (EOR).13,18–21 However, only a few reports on
complex reaction,14 increasing attention has been paid to
the EOR on realistic carbon supported catalysts can be found
in the literature which also include the application of spectro-
scopic techniques,2,14,22–28 and especially, the role of Ru and
Sn in the EOR is still not well understood. The present study
a Dalian Institute of Chemical Physics, Chinese Academy of Sciences,
Dalian, 116023, Graduate School of the Chinese Academy of
provides new mechanistic insight by the combination of the
above in situ techniques, allowing the identification of ad-
sorbed reaction intermediates and side products and the
quantitative evaluation of their coverage by in situ FTIR
measurements as well as the determination of the (volatile)
reaction products and their distribution at technically relevant
potentials by on-line DEMS.
Sciences, Beijing, 100039, China. E-mail: gqsun@dicp.ac.cn
b Institute of Surface Chemistry and Catalysis, Ulm University,
D-89069 Ulm, Germany. E-mail: juergen.behm@uni-ulm.de
c State Key Laboratory for Physical Chemistry of Solid Surfaces,
Department of Chemistry, Xiamen University,
Xiamen, 361005, China. E-mail: sgsun@xmu.edu.cn
w Electronic supplementary information (ESI) available: Cyclic vol-
tammograms of the potentiodynamic ethanol electrooxidation (Fig.
S1) and polarization curves for ethanol electrooxidation on the
catalysts (Fig. S2). See DOI: 10.1039/b700676b
In previous in situ FTIR spectroscopy studies of ethanol
electrooxidation on Pt or PtRu films deposited on a gold
ꢀc
This journal is the Owner Societies 2007
2686 | Phys. Chem. Chem. Phys., 2007, 9, 2686–2696

View Article Online
substrate, Iwasita et al. determined acetaldehyde, acetic acid
and CO2 as the main products.23,24 Variation of neither the Ru
content nor the ethanol concentration led to other detectable
products, but acetaldehyde was found to become increasingly
dominant with increasing ethanol concentration.24,29 Re-
cently, it was concluded from in situ IR measurements that
the addition of tin, e.g., in Pt3Sn/C catalysts, is beneficial for
the C–C bond breaking activity in ethanol oxidation and in
addition offers OH species at lower potentials.22,30,31 On the
other hand, results of a recent DEMS study did not support an
increased selectivity for CO2 formation of the PtRu/C or
Pt3Sn/C catalysts compared to Pt/C.15,25,27
2.2 Physical characterization
XRD patterns of all samples were obtained with an X’Pert PR
X-ray diffractometer using Cu Ka radiation. The 2y angular
scan was measured from 15–851 with a scan rate of 0.021
stepÀ1 and 10 s stepÀ1 (tube current 30 mA, tube voltage 40
kV). Transmission electron microscopy (TEM) was employed
for the structural characterization of the catalysts, using a
TECNAI F30 transmission electron microscopy (FEI Corp.).
2.3 Electrochemical and DEMS measurements
The electrochemical measurements (cyclic voltammetry (CV)
and linear sweep voltammetry (LSV) measurements) were
performed in a standard three-electrode RDE set-up (Pine
Instruments) without rotation. A Pt foil and a saturated
calomel electrode (SCE) were used as counter and reference
electrodes, respectively. The preparation of the working elec-
trodes was described previously.32 In short, 5 mg of the
respective catalyst was dispersed in a mixture of 1 ml water,
1 ml ethanol and 50 ml Nafions solution (5 wt%) under
ultrasonic stirring to form a homogeneous catalyst ink. 10 ml
of this mixture was pipetted onto a polished glassy carbon
(GC) disk with a diameter of 6 mm to form a homogeneous
catalyst layer.
In the following we will, after a brief description of the
experimental set-up and procedures used in this study, first
present results on the physical characterization of the carbon
supported Pt (40 wt% Pt), PtRu (30 wt% Pt, 15 wt% Ru) and
Pt3Sn (20 wt% Pt, 4 wt% Sn) catalysts by transmission
electron microscopy (TEM) and X-ray diffraction (XRD)
(section 3.1). This is followed by electrochemical and on-line
DEMS measurements of the activity and selectivity of these
catalysts towards ethanol oxidation (sections 3.2 and 3.3). The
last section concentrates on the formation and identification of
stable adsorbates and (volatile) reaction products upon inter-
action with ethanol at specific potentials (section 3.4) by
electrochemical and DEMS measurements (‘oxidative strip-
ping’) (section 3.4.1) and by in situ IR spectroscopy (section
3.4.2). The catalysts were prepared by a modified polyol
method,32 preliminary DEMS data on the EOR on similar
type catalysts were reported in ref. 25.
The DEMS set-up used for the electrochemical measure-
ments and the preparation of the carbon supported Pt and Pt
alloy catalyst thin film electrodes have been described in
previous publications.33–36 The DEMS set-up consists of two
differentially pumped chambers, a Balzers QMS 112 quadru-
pole mass spectrometer, and a Pine Instruments potentiostat.
In order to avoid interference with ClÀ ions, the reference
electrode was separated from the main electrochemical cell by
using a salt bridge containing the same solution as in the cell.
Prior to the measurements the electrode potential was
scanned in the supporting electrolyte (0.5 M H2SO4) at a scan
rate of 100 mV sÀ1 to clean the surface (from 0.06 to 1.16 V,
0.6 V, and 0.7 V for Pt/C, Pt3Sn/C, PtRu/C, respectively). To
quantify the adsorbates produced on the surface of the carbon
supported Pt, Pt3Sn and PtRu catalysts upon interaction with
ethanol at different potentials, adsorbate stripping experi-
ments were carried out after accumulation of ethanol adsor-
bate for a fixed time in ethanol containing electrolyte in
potentiostatic measurements. The procedure of these measure-
ments was as follows: After electrode cleaning, the electrolyte
was switched from the supporting electrolyte bottle to the
ethanol containing electrolyte bottle, keeping the electrode
potential of 0.06 V until the baseline was stable. After that, the
potential was stepped to the respective adsorption/reaction
potential for 5 or 15 min to obtain (close to) steady-state
currents. Then the electrolyte was switched from the ethanol-
containing electrolyte back to the supporting electrolyte and,
at the same time, the potential was stepped back to 0.06 V.
After rinsing the cell with supporting electrolyte, the adsorbed
species deposited on the surface under reaction conditions,
were oxidized in a potential scan from 0.06 to 1.16 V, record-
ing the Faradaic current and the mass spectrometric currents
2. Experimental
2.1 Catalyst preparation
H2PtCl6 Á 6H2O, RuCl3 Á xH2O and SnCl2 Á 2H2O (all AR
grade) were used for the synthesis of the Pt/C, PtRu/C and
Pt3Sn/C catalysts; Vulcan XC-72R carbon black (Cabot
g
Corp., SBET = 237 m2 À1) served as support material for
all samples. The required amounts of Pt, Ru and Sn com-
pounds were added to ethylene glycol (EG) to obtain a
mixture containing 2 mg metal per ml solvent. The mixture
was stirred for 1/2 h at room temperature to obtain a homo-
geneous solution. The required amount of sodium hydroxide
(1 M in EG) was added instantaneously to increase the pH
value of the solution to above 13. Subsequently, the mixture
was heated to 160 1C at a rate of 10 1C minÀ1 in an oil bath
and kept at this temperature for 3 h. The required amount
of Vulcan XC-72R was then added and the resulting mixture
was stirred for another 2 h to deposit the metal particles
on the support. After cooling down, the mixture was filtered
to get a black filtered cake, which was washed in hot deionized
water (80 1C) to remove organic residues and other ions.
Then the filtered cake was dried at 80 1C for 10 h with a
heating rate of about 5 1C minÀ1 in a vacuum oven. The noble
metal loadings obtained for the Pt/C, PtRu/C and Pt3Sn/C
catalysts were 40 wt%, 45 wt% (nominal atomic ratio Pt : Ru
= 1 : 1) and 24 wt% (nominal atomic ratio Pt : Sn = 3 : 1),
respectively.
+
simultaneously. To avoid any interference between the CO2
and CH3CHO+ ion currents (both at m/z = 44), the forma-
tion of CO2 was monitored at m/z = 22 (CO22+).15,37 For
ꢀc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 2686–2696 | 2687

View Article Online
each potential studied a freshly prepared electrode was used to
avoid effects resulting from EOR residues or possible losses of
Sn or Ru.
hydrogen electrode (RHE), the measurements were performed
at room temperature of around 25 Æ 2 1C.
3. Results and discussion
2.4 In situ FTIR spectroscopy measurements
3.1 Physical characterization of the catalysts
Electrochemical in situ FTIR spectroscopic studies were per-
formed on a Nexus 870 spectrometer equipped with a liquid
nitrogen cooled MCT-A detector and an EverGlo IR source,
as described in ref. 38. The electrode used for the IR study was
prepared in the same way as that used for the CV and LCV
measurements. A CaF2 disk was used as the IR window, and
the electrode was pressed against the IR window to create a
thin solution layer with a thickness of a few micrometers for in
situ FTIR measurements. The IR radiation from the IR source
passed through the CaF2 window with an incident angle of 601
and the thin-layer solution, then reflected from the electrode
surface. The spectra were collected using single potential-
alteration infrared spectroscopy (SPAIRS)39 procedures at
stepwise increased potentials. At each potential, 1000 inter-
ferograms (each interferogram takes 0.46 s) at a spectral
resolution of 8 cmÀ1 were co-added into a spectrum. The
resulting spectrum is defined as DR/R = (R(ES) À R(ER))/
(R(ER) À R(EW)), where R(ES) and R(ER) are single-beam
spectra recorded at sample potential (ES) and reference po-
tential (ER), respectively. R(EW) is the single-beam spectrum
of the IR window, which depends on the material and the
shape of the window. Following this definition, a negative
peak in the resulting spectrum represents IR absorption at ES,
equivalent to a production of species, while a positive peak
signifies IR absorption at ER and therefore consumption of the
reagent. Pure N2 and dry air (CO2 excluded) were used to
purge the electrolyte and the IR spectroscope, respectively, to
reduce contributions from CO2 and water vapor. All poten-
tials given in this paper are referenced to that of the reversible
Fig. 1 shows TEM images of Pt/C, PtRu/C, and Pt3Sn/C
catalysts and the corresponding histograms of the particle size
distribution. The metal particles, with a mean diameter of 2–3
nm, are uniformly distributed on carbon supports. From the
histograms of the particle size distribution, it is clear that 90%
of the particles in the Pt/C catalyst have diameters between
1.7–3.2 nm, while for the PtRu/C and Pt3Sn/C catalysts the
corresponding values are 1.2–2.7 nm. Therefore, for bimetallic
catalysts there is a trend of decrease in the particle size,
compared to pure Pt nanoparticles. The results suggest that
the addition of Ru or Sn to Pt can reduce the agglomeration of
Pt particles during the synthesis process, which was also
reported in the literature.32,40–43 The physical origin for this
phenomenon, however, is still under discussion.
In the XRD patterns (Fig. 2), the diffraction peaks of the
Pt/C catalyst are much sharper than those of the PtRu/C and
Pt3Sn/C catalysts, supporting the TEM based conclusion that
the metal particles in the Pt/C catalyst are larger than those in
the two bimetallic catalysts. A quantitative evaluation of the
Pt(220) line based on the Scherrer equation yielded a mean
particle size of 3.3 nm for the Pt/C catalyst; for the other two
catalysts the particle size could not be evaluated because of a
too low intensity of that line (PtRu/C) or because of the
presence of a SnO2 related line overlapping the Pt(220) line
(Pt3Sn/C). Only Pt diffraction peaks are observed for PtRu/C,
with no indications of Ru or Ru oxides. The diffraction peaks
of Pt for the PtRu/C catalyst are shifted to higher angles,
indicative of PtRu alloy formation.32 For the Pt3Sn/C catalyst,
Fig. 1 TEM images and the corresponding particle size distributions of (a) Pt/C, (b) PtRu/C, (c) Pt3Sn/C catalysts.
ꢀc
This journal is the Owner Societies 2007
2688 | Phys. Chem. Chem. Phys., 2007, 9, 2686–2696

View Article Online
carried out at a slow scan rate of 1 mV sÀ1 on the Pt/C, PtRu/
C and Pt3Sn/C catalysts (for polarization curves see Fig. S2 in
the ESIw). The profiles show that the onset of ethanol oxida-
tion is just above 0.1 V on the Pt3Sn/C catalyst, while on the
PtRu/C and Pt/C catalysts the reaction commences at 0.3 and
0.4 V, respectively. In addition, at higher potentials (0.6–0.7
V), the EOR Faradaic current density was much higher on
Pt3Sn/C than on the other two catalysts. At potentials of
technical interest, around 0.4 V, the current density on the
Pt/C catalyst is negligible, while on the PtRu/C and Pt3Sn/C
catalysts it is 0.008 and 0.026 mA cmÀ2, respectively. Thus, the
Pt3Sn/C catalyst is much more active under these conditions
than the Pt/C and even the PtRu/C catalyst, with similar Pt
loadings for the alloy catalysts (half of that of the Pt/C
catalyst) and all catalysts prepared by the same modified
polyol method. This result resembles findings by Lamy et al.,
who reported a higher activity for PtxSn/C catalysts prepared
Fig. 2 X-Ray diffraction patterns of (a) Pt/C, (b) PtRu/C, (c)
Pt3Sn/C catalysts.
via a co-impregnation/reduction method14 or via the ‘‘Bonne-
¨
mann’’ method2 for ethanol oxidation than for PtRu/C and
the Pt related diffraction peaks are shifted to lower angles,
indicating Pt3Sn alloy formation.32,40 In addition, we observed
two weak peaks for tin oxide at around 341 and 521 (PCPDF
411445),44 which indicates that after preparation tin oxide co-
exists with the Pt3Sn alloy in the Pt3Sn/C catalyst.45
Pt/C catalysts.
3.3 Selectivity of Pt/C, Pt3Sn/C, and PtRu/C catalysts
towards the EOR
3.2 Activity of Pt, PtRu, and Pt3Sn towards ethanol
electrooxidation
The selectivity of the Pt/C, Pt3Sn/C, and PtRu/C catalysts in
the EOR was investigated by quantitative potentiostatic elec-
trochemical and DEMS measurements of the EOR product
distribution, performed at room temperature under fuel cell
relevant mass-transport and continuous reaction conditions.
Faradaic and mass spectrometric current transients, follow-
ing the formation of CO2 (m/z = 22) and acetaldehyde (m/z =
29) after a potential step from 0.06 V to the respective reaction
potential, are presented in Fig. 3 (0.5 M H2SO4 + 0.1 M
ethanol solution). For the Pt/C catalyst, we investigated
reaction potentials of 0.5, 0.6, and 0.7 V. At lower potentials,
the Faradaic current is negligible due to the surface poisoning
by COad resulting from the dissociative adsorption of ethanol.
For the PtRu/C and Pt3Sn/C catalysts, the lowest reaction
potential could be reduced to 0.4 V due to their higher
reactivity. On all three catalysts, the steady-state Faradaic
current density (data obtained 15 min after the potential step)
increased with potential. At technically interesting potentials
lower than 0.5 V, the Faradaic current densities on Pt3Sn/C
and PtRu/C are similar, and both are distinctly higher than
that on Pt/C. On the other hand, the Faradaic current density
on the Pt3Sn/C catalyst is much higher than on the PtRu/C
and Pt/C catalysts at potentials of 0.5 V or above, where
adsorbed residues start to be oxidized on the Pt3Sn/C and Pt/C
catalysts, and are largely oxidized on the PtRu/C catalyst (see
section 3.4.1). The lower onset potential for ethanol adsorbate
oxidation on the PtRu/C catalyst compared to the other
catalysts indicates that PtRu/C is most efficient for adsorbate
oxidation among the three catalysts. However, the Faradaic
currents for the EOR on these catalysts mainly originate from
the formation of C2 products rather than from complete
oxidation of the adsorbates to CO2, as evidenced by the
DEMS results in section 3.4.1. The Faradaic current of the
EOR on the Pt3Sn/C catalyst is the highest for the same
reason, dominant C2 formation, even though the adsorbate
The activity of the three catalysts was first evaluated in CVs
(see Fig. S1 in the electronic supplementary informationw),
together with the base CVs. The upper potential limit of the
base CV and of CVs in ethanol containing electrolyte (poten-
tiodynamic ethanol oxidation) on the PtRu/C and Pt3Sn/C
catalysts was set to 0.7 and 0.6 V, respectively, to avoid
dissolution of the non-noble metal in the catalysts.46,47 The
Faradaic currents are normalized to the active surface areas,
which were determined by COad stripping36 to be 81, 111 and
À1
88 m2 gmetal for the Pt/C, PtRu/C and Pt3Sn/C catalyst,
respectively. The COad stripping curves on the three catalysts
are presented later in Fig. 4–6, respectively.
In these measurements the current densities at 0.6 V de-
crease in the order of Pt3Sn/C 4 Pt/C 4 PtRu/C. At
potentials lower than 0.57 V, this changes to Pt3Sn/C 4
PtRu/C 4 Pt/C. The better performance of Pt/C at potentials
positive to 0.6 V is ascribed to a combination of two effects,
the higher efficiency of Pt for C–C bond breaking and the
increased formation of OH species at these potentials which
can remove the strongly adsorbed COad species from the Pt
surface. Due to the rather low upper potential limit in the
scans, the curves show no distinct maximum, which based on
literature reports would be expected to be around 0.8 V on
PtRu/C15,29 and 0.95 V on the Pt3Sn/C catalyst,2,15,32 respec-
tively. Hydrogen adsorption is clearly inhibited on the three
catalysts, indicating that the active sites of the catalysts are
occupied by the reaction intermediates/side products gener-
ated from ethanol dissociation at higher potentials in the
preceding negative-going scan. The identification of the ad-
sorbed species and products will be discussed later, together
with in situ FTIR and on-line DEMS results. In order to
determine the onset potentials for ethanol electrooxidation on
the different catalysts more precisely, LSV experiments were
ꢀc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 2686–2696 | 2689

View Article Online
Fig. 3 Simultaneously recorded Faradaic and ion current (m/z = 22, m/z = 29) transients for ethanol oxidation on (a) Pt/C, (b) PtRu/C, and (c)
Pt3Sn/C in 0.1 M CH3CH2OH + 0.5 M H2SO4 solution at constant electrode potentials, recorded upon stepping the potential from 0.06 V to the
respective reaction potential (catalyst loading: 140 mg cmÀ2).
oxidation is not as facile as on the PtRu/C catalyst. The
adsorption/dissociation of ethanol most likely occurs on Pt
sites, while the second component provides OH species for
COad oxidation. Therefore, although Ru and Sn act as OH
donors, the number of Pt sites at the surface, and thus
dissociative adsorption, is decreased, without any improve-
ment of the EOR selectivity for complete oxidation.
ing on the reaction potentials, the CO2 current efficiency on
these catalysts reaches at most 2% (the corresponding CO2
yield is less than 1%). At the same potential, the current
efficiency for CO2 formation is slightly higher on the Pt3Sn/C
than on the PtRu/C catalyst, but lower than that on the Pt/C
catalyst. At a potential of 0.6 V, the CO2 current efficiencies
are B1% for Pt3Sn/C and PtRu/C, and B2% for Pt/C,
respectively. Correspondingly, acetaldehyde and acetic acid
are the majority products in the EOR within the investigated
potentials. On the Pt3Sn/C catalyst, the current efficiency for
acetaldehyde formation increases with potential, while that for
acetic acid decreases correspondingly, in good agreement with
previous results obtained on the same catalyst in 1 M ethanol
+ 0.5 M H2SO4 solution.25 Comparing two different ethanol
concentrations (0.1 M and 1 M ethanol solution), the current
efficiency and product yield for acetic acid increases with
decreasing ethanol concentration (0.1 M ethanol), similar
to the behavior of Pt/C.37 This trend can be explained by the
To analyze the EOR product distribution on the three
catalysts in detail, the current efficiencies (Ai) and product
yields (Wi) for CO2, acetaldehyde and acetic acid formation at
different potentials in 0.1 M ethanol + 0.5 M H2SO4 solution
were calculated and listed in Table 1 (for details of the
evaluation see ref. 37). At lower potentials, at 0.4 V on
Pt3Sn/C and at 0.5 V on Pt/C, the CO2 formation rate is too
low to calculate the efficiency for CO2 formation. In general,
on all three catalysts the current efficiency and product yield
for CO2 decrease with increasing potential, although the
absolute amount of CO2 formation increases slightly. Depend-
Table 1 Faradaic currents, current efficiencies (Ai) and product yields (Wi) for CO2, acetaldehyde and acetic acid formation during ethanol
oxidation on Pt/C, PtRu/C, and Pt3Sn/C catalysts at different potentials in 0.1 M ethanol + 0.5 M H2SO4 solution (data recorded 15 min after a
potential step from 0.06 V to the respective reaction potential)
b
Inormalized
mA cmÀ2
/
Ai(CO2)
(%)
Ai(CH3CHO)
(%)
Ai(CH3COOH)
(%)
Wi(CO2)
(%)
Wi(CH3CHO)
(%)
Wi(CH3COOH)
(%)
Catalyst
Pt3Sn/C
E/V vs. RHE
0.4a
0.5
0.6
0.7
0.4a
0.5
0.6
0.7
0.5
0.6
0.7
0.003
0.014
0.055
0.080
0.004
0.013
0.030
0.043
0.007c
0.035
0.060
—
53.1
46.0
53.9
49.7
41.5
32.2
31.7
30.9
50.0
43.1
46.4
46.9
51.7
44.9
49.5
57.4
66.7
67.5
68.5
50.0
54.8
52.1
—
1.1
0.5
0.4
0.5
0.6
0.4
0.3
—
69.4
63.3
70.2
66.5
58.8
48.8
48.2
47.3
66.7
60.5
63.6
30.6
35.6
29.3
33.1
40.7
50.6
51.4
52.4
33.3
38.5
35.7
2.3
1.2
0.8
1.1
1.1
0.8
0.6
—
PtRu/C
Pt/C
2.1
1.5
1.0
0.7
b
a
Larger error in the determination of the product distribution due to a low signal-to-noise ratio. The Faradaic current is normalized by division
c
by the active surface area Sact determined by COad stripping. Values obtained 5 min after the potential step to 0.5 V.
ꢀc
This journal is the Owner Societies 2007
2690 | Phys. Chem. Chem. Phys., 2007, 9, 2686–2696

View Article Online
re-adsorption and further oxidation of incomplete oxidation
products, which is more efficient at lower concentrations.48 On
the PtRu/C catalyst, the current efficiency for acetaldehyde
decreases from 42% to 32% when the reaction potential is
changed from 0.4 to 0.5 V, and then remains constant for
increasing potential up to 0.7 V, which agrees closely with the
trend on a similar type of 40 wt% PtRu catalyst in 1 M
ethanol.25 Also in this case, more acetic acid was formed in the
0.1 M ethanol electrolyte, which can be explained by the same
‘re-adsorption and oxidation’ mechanism. Furthermore, more
acetic acid was produced on the PtRu/C catalyst at each
investigated potential than on the other catalysts (0.6 and
0.7 V: 67% current efficiency/52% product yield) (see Table 1).
The higher production of acetic acid on the PtRu/C catalyst
can be explained by the larger amount of available OH species
on Ru sites than on Sn or Pt sites, making re-adsorption and
further oxidation of acetaldehyde more effective on this cata-
lyst than on the other ones.
3.4 Ethanol adsorbates on carbon supported Pt, PtRu, and
Pt3Sn catalysts
Topics of this section are the identification and quantification
of the adsorbed decomposition products, which are produced
during the interaction with ethanol under reaction conditions
on the Pt/C, PtRu/C, and Pt3Sn/C catalysts, by on-line DEMS
and in situ FTIR techniques. It should be noted that the
adsorbate coverage determined from DEMS measurements
is calculated on the premise that only C1 species are adsorbed.
In the FTIR measurements, approximate values of the COad
coverage are determined assuming that the COL intensity
detected on the respective catalyst is proportional to the COad
coverage. It should be noted that COad adsorbed on bridged
and multiple sites (COM) have been observed upon ethanol
oxidation on a Pt film electrode using in situ attenuated
reflection Fourier transform IR spectroscopy (ATR-FTIRS)
techniques,49,50 which will contribute to the COad coverage.
These species were not observed in the present measurements,
most likely because of the lower sensitivity, and therefore not
included in the coverage determination.
Fig. 4 CVs and MSCVs (m/z = 22) for the oxidation of ethanol
adsorbates accumulated upon adsorption/reaction at different poten-
tials (5 min) and of pre-adsorbed CO on a Pt/C catalyst in 0.5 M
H2SO4 (catalyst loading: 140 mg cmÀ2).
obtained after 5 min adsorption time are shown in the
presentation. The oxidation of ethanol adsorbates on the
Pt/C and Pt3Sn/C catalysts occurs in two potential regions,
as described in ref. 51, in the low potential region below 0.85 V
and in the high potential region above 0.85 V. For the PtRu/C
catalyst, these two regions are not that clearly separated. On
the Pt/C and Pt3Sn/C catalysts, the oxidation of ethanol
adsorbates commences at 0.5 V and passes through a broad
maximum centered at about 0.7 V, while on the PtRu/C
catalyst, the oxidation of ethanol adsorbates starts at 0.3 V
and peaks at 0.6 V. These results compare qualitatively with
the onset potential for CO2 formation detected in the FTIR
spectra (see section 3.4.2). On all three catalysts, most of the
ethanol adsorbates could be oxidized to CO2 in the lower
potential region; only a small fraction of them persists and is
oxidized to CO2 in the higher potential region.51 In contrast,
oxidation of pre-adsorbed CO (COad) starts at 0.65, 0.4, and
0.3 V on the Pt/C, PtRu/C, and Pt3Sn/C catalysts, respec-
tively. Compared with the respective ignition potential for
COad oxidation, the oxidation of ethanol adsorbates starts at
significantly lower potentials on the Pt/C or PtRu/C catalysts.
These results agree well with observations from IR measure-
ments on Pt film electrodes.49,50 In contrast, on the Pt3Sn/C
catalyst, ethanol adsorbate oxidation starts at much higher
potentials than the ignition potential for COad oxidation,
which is also consistent with the findings from FTIR spectro-
scopy (section 3.4.2).
3.4.1 DEMS measurements upon ethanol adsorption/oxida-
tion. The coverages of the carbon containing adsorbates
resulting from the dissociative adsorption of ethanol (‘ethanol
adsorbates’) on the Pt/C, PtRu/C, and Pt3Sn/C catalysts at
specific adsorption/reaction potentials were quantified by
oxidative stripping of the adsorbed species using on-line mass
spectrometric detection of CO2. The results are compared with
the data from COad stripping experiments (oxidative stripping
of a pre-adsorbed CO adlayer).
The simultaneously recorded CVs and MSCVs (m/z = 22)
recorded after interaction of the catalyst with ethanol at
different adsorption/reaction potentials between 0.06 and
0.70 V (adsorption/reaction time: 5 min) and after pre-adsorp-
tion of CO on Pt/C, PtRu/C, and Pt3Sn/C catalysts in 0.5 M
H2SO4 are shown in Fig. 4–6, respectively. To evaluate the
effect of the adsorption time on the adsorbate coverage for the
three catalysts, similar adsorption experiments were also per-
formed with 15 min adsorption times. Since the shapes of the
CVs and MSCVs are similar in both cases, only the curves
The relative coverage of ethanol adsorbates was calculated
as the ratio between the mass spectrometric charge of the
ꢀc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 2686–2696 | 2691

View Article Online
Fig. 5 CVs and MSCVs (m/z = 22) for the oxidation of ethanol
adsorbates accumulated upon adsorption/reaction at different poten-
tials (5 min) and of pre-adsorbed CO on a PtRu/C catalyst in 0.5 M
H2SO4 (catalyst loading: 140 mg cmÀ2).
Fig. 6 CVs and MSCVs (m/z = 22) for the oxidation of ethanol
adsorbates accumulated upon adsorption/reaction at different poten-
tials (5 min) and of pre-adsorbed CO on a Pt3Sn/C catalyst in 0.5 M
H2SO4 (catalyst loading: 140 mg cmÀ2).
m/z = 22 peak (CO22+) for the oxidation of ethanol adsor-
bates (from Fig. 4–6) and that obtained from oxidation of a
saturated CO adlayer on the respective catalyst, yCO,sat (CO
adsorption at 0.06 V, m/z = 22 signal). The resulting cov-
erages yCO,rel are shown for different adsorption potentials in
Fig. 7. Note that this calculation does not correct for small
amounts of adsorbed C2 species possibly present on the sur-
face. For comparison, the relative coverage of ethanol adsor-
bates on a commercial 20 wt% Pt/C catalyst51 is also plotted in
Fig. 7. On all catalysts, the adsorbate coverage increases
significantly when increasing the adsorption time from 5 to
15 min, indicating that ethanol adsorption is rather slow. For
15 min adsorption time, the (relative) adsorbate coverage on
the 20 wt% Pt/C is comparable to that on the 40 wt% Pt/C
catalyst. In addition, the potential dependence of the adsor-
bate uptake is similar on both Pt/C catalysts in the potential
regime investigated. At the lowest adsorption potential of 0.06
V, the adsorbate coverage on the carbon supported Pt cata-
lysts is negligible. Apparently, the relatively strong adsorption
of Hupd almost completely inhibits the dissociative adsorption
of ethanol at low potentials (o0.1 V).51 The relative adsorbate
coverage on the PtRu/C and Pt3Sn/C catalysts at such low
potentials is slightly higher than on the Pt/C catalyst, indicat-
ing that dissociative adsorption of ethanol on these two
catalysts is more facile. This can be understood by the weaker
and less efficient Hupd adsorption on the former catalysts (see
Fig. S1 of the ESIw), which makes it more facile for these
catalysts to (dissociatively) adsorb ethanol within the Hupd
potential range.
Leger et al. also reported that the addition of Sn to Pt
´
activates the dissociative adsorption of ethanol and the clea-
vage of the C–C bond at lower potentials.30,31 However, for
adsorption potentials above 0.1 V, the adsorbate coverage on
Pt/C increases sharply and exceeds that on the PtRu/C and
Fig. 7 Relative coverage of ethanol adsorbates compared to that of a
saturated CO adlayer for different adsorption potentials on Pt/C (&
adsorption time: 5 min, ’ adsorption time: 15 min), PtRu/C (D: 5 min
adsorption time, m: 15 min adsorption time), and Pt3Sn/C ($: 5 min
adsorption time, %: 15 min adsorption time). A commercial 20 wt%
Pt/C catalyst (E-Tek) (J)51 is included for comparison (catalyst
loading: 140 mg cmÀ2, 10 min adsorption time).
ꢀc
This journal is the Owner Societies 2007
2692 | Phys. Chem. Chem. Phys., 2007, 9, 2686–2696

View Article Online
Pt3Sn/C catalysts, achieving a maximum values of 0.28 yCO,sat
and 0.63 yCO,sat at an adsorption potential of around 0.4 V
after 5 min and 15 min adsorption time, respectively. On the
PtRu/C and Pt3Sn/C catalysts, maximum coverages of 0.2
yCO,sat and 0.27 yCO,sat, respectively, are reached at 0.2 and
0.4 V, respectively, after 5 min adsorption time. Increasing the
adsorption time to 15 min, the adsorbate coverage increases to
0.26 and 0.53 yCO,sat for the PtRu/C and Pt3Sn/C catalysts at
0.4 V, respectively.
In total, the more facile ethanol adsorption on the bimetallic
catalysts at low potentials (o0.1 V) results from the more
efficient blocking of the Pt/C catalyst by Hupd, while at more
positive potentials the higher activity of Pt for C–C bond
breaking results in faster ethanol adsorption and a higher
saturation coverage on the Pt/C catalyst. The ignition poten-
tials for CO2 formation, which are similar for Pt/C and Pt3Sn/
C, and lower for PtRu/C, reflect the lower potential for OH
formation on the latter catalysts.
3.4.2 Identification of the intermediates and products by in
situ FTIR spectroscopy. Fig. 8 shows sequences of in situ FTIR
spectra recorded during ethanol electrooxidation on the Pt/C,
PtRu/C and Pt3Sn/C catalysts, while varying the electrode
potential ES stepwise from 0.14 to 0.84 V (reference potential
0 V). One dominant feature is a bipolar band with its negative
peak at 2030 cmÀ1 and its positive peak gradually shifting to
higher frequencies as ES increases.38,52 The bipolar band is
assigned to IR absorption by linearly adsorbed CO (COL),
produced by the dissociative adsorption of ethanol on the
electrocatalyst surface. The bipolar shape results from the
subtraction of the two Stark-shifted, positive COL bands at
sample and reference potential, respectively (see section 2),
which leads to a positive peak at ER. Anomalous IR properties
of CO adsorbed on nanoscale materials of noble metals, as
evidenced by the positive ‘absorption’ peak, have been re-
ported previously and discussed in detail by Sun’s group38,53
(see also the related discussion in ref. 54 and 55). The IR
features of the COL bipolar band are similar on the Pt/C,
PtRu/C and Pt3Sn/C catalysts (see Fig. 8b and 8c), while the
peak position and intensity vary between the catalysts due to
their different adsorption properties. A negative peak (corre-
sponding to IR absorption at ES) near 2345 cmÀ1 is attributed
to the asymmetric stretch vibration of CO2 which appears at
about 0.49 V on the Pt/C and Pt3Sn/C catalysts, but already at
0.34 V on the PtRu/C catalyst.
Another band located at around 1715 cmÀ1, assigned to the
stretch vibration of the CQO bond in acetaldehyde and/or
acetic acid as partially oxidized reaction products, appeared at
a potential of about 0.49, 0.34, and 0.24 V on the Pt/C, PtRu/
C, and Pt3Sn/C catalysts, respectively, indicating that the C2
products are more easily formed on the Pt3Sn/C catalyst than
on the other ones. Actually, it is difficult to distinguish
between the carbonyl groups in acetaldehyde and acetic acid,
since the CQO bands in these compounds are very close, at
1713 and 1715 cmÀ1, respectively.7 In addition, we find four
other negative bands located between 1400 and 100 cmÀ1. For
the Pt3Sn/C catalyst, where these IR bands are strongest, they
appear at 1391, 1368, 1277, and 1108 cmÀ1 at potentials higher
than 0.24 V. The bands at 1391 and 1277 cmÀ1 are attributed
Fig. 8 In situ FTIR spectra recorded during ethanol electrooxidation
on the (a) Pt/C, (b) PtRu/C, and (c) Pt3Sn/C catalyst in 0.1 M H2SO4
+ 0.1 M C2H5OH solution. 1000 interferograms (resolution 8 cmÀ1
)
were collected and co-added into each spectrum (catalyst loading:
86 mg cmÀ2).
to the C–O stretch and O–H deformation vibrations in acetic
acid, the peaks at 1368 and 1108 cmÀ1 are due to the CH3
symmetric deformation and C–H wagging vibration in acet-
aldehyde (see Table 2). At potentials above 0.49 V, a wide
peak around 2622 cmÀ1 appears, which may be ascribed to an
overlap of the C–H stretch (2700 to 3000 cmÀ1) and a broad
band of the O–H stretch of the carboxylic group between 2500
and 3000 cmÀ1 in acetic acid. The positive bands (correspond-
ing to more pronounced IR absorption at ER) at 2977 and
1455 cmÀ1 are assigned to a C–H stretch and a coupled
vibration of the O–H in-plane deformation and the C–O
stretch vibration in ethanol in the solution, whose intensity
decays due to ethanol consumption in the thin electrolyte
layer. In addition, the positive absorption band located close
to 1650 cmÀ1 indicates the consumption of interfacial water.
The negative peaks at 1190 and 1057 cmÀ1 can be attributed to
IR absorption by adsorbed (bi-)sulfate upon COad removal.
For the Pt/C and PtRu/C catalysts, which exhibit similar
ꢀc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 2686–2696 | 2693

View Article Online
Table 2 Assignments of the fundamental bands in the spectra in Fig. 8
Wavenumber/cmÀ1
Assignment
Ref.
2977, 2901
2062, 2030
2342
1715
1650
1455
1391, 1277
1368, 1108
1190, 1057
C–H stretching
COL, (CQO stretching)
CO2, asymmetric stretching
CQO stretching of acetaldehyde/acetic acid
O–H bending in H2O
Coupled vibration of in plane O–H deformation and C–O stretching in C2H5OH
C–O stretching and O–H deformation in CH3COOH
CH3 symmetric deformation and C–H wagging in CH3CHO
HSO4,ad
12, 56
12, 14, 22
12, 14, 22
12, 14, 22
12, 57
58
2, 7, 23
2, 7, 58
58
spectral features, the same assignment can be used (see com-
pilation in Table 2).
on the other catalysts. On the Pt/C catalyst, the resulting COL
could only be oxidized when the potential is above 0.45 V, in
contrast to the bimetallic catalysts, where, after passing the
maximum at 0.36 V, the COL band intensity descends slowly
for the PtRu/C catalyst and even slower for the Pt3Sn/C
catalyst (see also section 3.4.1). This implies that COL oxida-
tion is faster on the PtRu/C catalysts than on Pt3Sn/C at
potentials higher than 0.36 V, which is also confirmed by the
more pronounced increase of the CO2 signal on the PtRu/C
catalyst (see the next paragraph).
The COL band intensities on the Pt/C, PtRu/C, and Pt3Sn/C
catalysts at different potentials are plotted in Fig. 9. Because of
the problems in integrating the intensity of the COL band
caused by its bipolar shape, the peak-to-peak values of the
bipolar band were used instead. For comparison, the inte-
grated intensity of the CO2 band on the three catalysts is also
included in Fig. 9. On all three catalysts, the COL band
intensity follows a volcano-type behavior, with a low COL
intensity at low potentials, an increase with higher potential up
to a maximum value which is reached between 0.4 and 0.5 V
on the Pt/C catalyst and at 0.36 V on the PtRu/C and Pt3Sn/C
catalysts, and a subsequent decay due to increasing COad
oxidation and/or decreasing COad formation (C–C bond
splitting). At each potential, the COL band intensity is highest
on the Pt/C catalyst, indicating that C–C bond splitting is
more facile and COad is less efficiently removed on Pt/C than
Further information on the COad oxidation behavior is
obtained from the CO2 band intensity, which during COad
oxidation increases at the expense of the COL intensity on the
three catalysts (see Fig. 9). The formation of CO2 commences
at about 0.25 V on the PtRu/C catalyst and at 0.45 V on the
Pt/C and Pt3Sn/C catalysts, indicating that under present
conditions COL species are more easily removed on the
PtRu/C catalyst than on the other ones. This contrasts the
behavior observed in COad oxidation (COad from CO adsorp-
tion), where CO2 formation starts at substantially lower
potential on Pt3Sn/C than on PtRu/C or even Pt/C). One
might speculate whether this is due to a different bonding
situation of the COad or whether this difference simply results
from a different COad coverage and adlayer composition
obtained upon ethanol adsorption or CO adsorption. The
CO2 intensity increases with potential until 0.69 V, in good
agreement with the results reported in ref. 30. For the Pt/C
and PtRu/C catalysts, the CO2 band intensity decreases slowly
from 0.69 to 0.84 V, with the decay on the Pt3Sn/C catalyst
being even slower. When the potential is above 0.5 V, more
CO2 is detected over the Pt/C catalyst than over the Pt3Sn/C
catalyst at the same reaction potential. This latter result is
similar to our findings from the DEMS measurements (section
3.4.1). The slight differences in the potential dependence of the
adsorbate coverages derived from the DEMS and FTIR
measurements are most simply explained by the differences
in the experimental protocol and the transport conditions in
the two types of experiments. The DEMS experiments were
carried out under continuous flow conditions and at a fixed
potential, starting with a freshly prepared electrode at each
potential, while the IR measurements were performed in
stagnant electrolyte, changing the potential stepwise to higher
potentials. The different transport characteristics lead to a
continuous removal of acetaldehyde and acetic acid from the
electrode, while in the thin-layer IR configuration, intermedi-
ates are trapped in the catalyst layer and can be further
oxidized. In addition, the metal loading in the three catalysts
Fig. 9 IR band intensities of COL (peak-to-peak value in the bipolar
peak) (’) and CO2 (integrated intensity) (m) during adsorption/
oxidation of ethanol at different electrode potentials (data from
Fig. 8).
ꢀc
This journal is the Owner Societies 2007
2694 | Phys. Chem. Chem. Phys., 2007, 9, 2686–2696

View Article Online
also influences the intensity of the IR signals. This should lead
to higher CO2 and COad signals for the higher loading Pt/C
catalyst compared to the bimetallic catalysts, while the elec-
trochemical and mass spectrometric signals are normalized to
the active surface area, removing the loading effects.
ments (detection of volatile CO2). These measurements
showed that ethanol adsorption is totally inhibited on the
Pt/C catalyst at the lowest potentials investigated (0.06 V),
while on PtRu/C and Pt3Sn/C catalysts ethanol adsorption
could proceed under these conditions at rather low, but non-
zero rates. The complete (Pt/C) or almost complete (bimetallic
catalysts) inhibition of ethanol adsorption at the lowest po-
tential was attributed to a blocking of adsorption sites by
Finally we would like to note that the intensity of the COL
band assumes its minimum value at the same potential where
the intensity of the CO2 related band reaches its maximum,
similar to reports by Le
´
ger et al.30,31 This result, together with
Hupd, which is most strongly bound on the Pt/C catalyst, and
the time spent at each potential (460 s in the IR measure-
ments), underlines that CO2 formation is not only possible by
oxidation of adsorbed CO species formed at lower potentials,
but can also occur via dissociation and subsequent oxidation
of C2 species at higher potentials, in agreement with the
DEMS results presented in section 3.4.1 and ref. 51.
therefore leads to the most efficient inhibition on that catalyst.
At potentials above 0.1 V, however, the adsorbate coverage on
the Pt/C catalyst increases sharply to exceed that on the other
two catalysts, indicating that in the absence of Hupd dissocia-
tive ethanol adsorption is more facile on the Pt/C catalyst than
on the other ones. The dissociative adsorption of ethanol
results in strongly adsorbed COad and CHx,ad species, which
act as reaction poison for the EOR at lower potentials. There
is no evidence for significant amounts of other stable adsor-
bates formed upon interaction with ethanol on these catalysts.
The electrocatalytic activity of the Pt3Sn/C catalyst towards
the EOR is much higher than that of the Pt/C and PtRu/C
catalysts at potentials below 0.6 V. For all three catalysts and
under all reaction conditions investigated, however, partial
oxidation products (acetaldehyde and acetic acid) are the
major EOR products, with CO2 contributing only less than
2% of current efficiency and less than 1% of the product yield,
depending on the reaction potentials and different catalysts.
The higher EOR current on the Pt3Sn/C catalyst at lower
potentials is due to the higher yield of C2 products, not of
CO2. At the same reaction potential, the current efficiency (or
product yield) for CO2 formation is lowest on the PtRu/C
catalyst and highest on the Pt/C catalyst.
In total, the FTIR measurements have shown that, in
addition to the strongly adsorbed COad species and CO2,
CQO containing products are also formed in the EOR, in
agreement with the DEMS results. The FTIR measurements
cannot distinguish, however, between acetaldehyde or acetic
acid as the two most likely candidates. The increasing forma-
tion of CO2 correlates with a decrease in COL species intensity
and hence in COL coverage. Removal of the COL species
commences at lower potentials on the PtRu/C and Pt3Sn/C
catalysts than on the Pt/C catalyst. With increasing potential,
the decay of the COL intensity and the increase of CO2
intensity are less pronounced for Pt3Sn/C than for PtRu/C,
reflecting a higher activity of the PtRu/C catalyst for COL
oxidation/CO2 formation under these conditions. While the
trends are compatible with the DEMS results presented be-
fore, a quantitative comparison between results from the two
techniques has to consider the different catalysts’ loadings and
experimental conditions in the two types of experiments.
The combined electrochemical and in situ spectroscopic data
underlines that the current efficiency for CO2 formation is
determined by two factors, the activity for C–C bond splitting
and the ability to remove the resulting COad and CHx,ad
species, which act as reaction poison, by oxidation. Both of
these rates will depend on the potential, particularly interest-
ing from a technical point of view are relatively low potentials
(0.4–0.6 V). The resulting COad, however, is difficult to remove
from the Pt surface due to its lower oxidation activity. On the
PtRu/C and Pt3Sn/C catalysts, the dissociative adsorption of
ethanol on the surface is less facile than on the Pt/C surface at
these potentials, but the reaction products are oxidized faster,
resulting in lower steady-state COad coverages at the same
potential and in a maximum of the steady-state COad coverage
at lower potentials compared to Pt/C. Under continuous
reaction conditions at higher, but constant potentials, CO2
formation by consumption of COad species formed at lower
potentials is not possible; the steady-state CO2 formation rate
instead reflects the oxidation of COad and CHx,ad species
produced at this potential.
4. Conclusions
We have presented and discussed results of a comparative in
situ spectro-electrochemical study on the interaction of etha-
nol with carbon supported Pt/C, PtRu/C and Pt3Sn/C cata-
lysts performed under realistic reaction conditions and under
enforced electrolyte flow, i.e., under controlled mass transport
conditions, which allowed the identification and quantification
of adsorbed reaction intermediates and products (in situ
FTIR) and volatile reaction (side) products (on-line DEMS)
on realistic catalysts and under close to realistic conditions.
The measurements provide information on the nature and
(potential dependent) abundance of adsorbed reaction inter-
mediates and products, and on the nature and yields of volatile
reaction products and side products, which is important for
the mechanistic understanding of the EOR and the different
reactions going on during the overall reaction. The catalysts
were synthesized via a modified polyol method which results in
a homogeneous distribution of the metal particles on the
carbon support and very small metal particles with diameters
of 2.2–2.7 nm, as derived from XRD and TEM measurements.
Information on the activity of the catalysts for the disso-
ciative adsorption of ethanol, in particular for C–C bond
breaking as a first step for ethanol oxidation, was derived
from mass spectrometric ethanol adsorbate stripping experi-
Acknowledgements
This work was supported by the Sino-German Center at
Beijing, the National Natural Science Foundation of China
(Grant No. 50575036, 50676093), the Opening Foundation of
the State Key Laboratory for Physical Chemistry of Solid
ꢀc
This journal is the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 2686–2696 | 2695

View Article Online
Surface of Xiamen University, and the Deutsche Forschungs-
gemeinschaft (Be 1201/12–1 and Be 1201/11–1). L. H. Jiang is
grateful for a fellowship from the Alexander von Humboldt
Foundation.
27 L. Jiang, L. Colmenares, Z. Jusys, G. Q. Sun and R. J. Behm,
Electrochim. Acta, DOI: 10.1016/j.electacta.2007.01.047.
28 F. C. Simoes, D. M. dos Anjos, F. Vigier, J.-M. Leger, F. Hahn, C.
´
Coutanceau, E. R. Gonzalez, G. Tremiliosi-Filho, A. R. de
Andrade, P. Olivi and K. B. Kokoh, J. Power Sources, 2006,
167, 1.
29 G. A. Camara, R. B. de Lima and T. Iwasita, Electrochem.
Commun., 2004, 6, 812.
´
30 J.-M. Leger, S. Rousseau, C. Coutanceau, F. Hahn and C. Lamy,
References
1 C. Lamy, A. Lima, V. Le Rhun, F. Delime, C. Coutanceau and J.-
Electrochim. Acta, 2005, 50, 5118.
M. Le
2 C. Lamy, S. Rousseau, E. M. Belgsir, C. Coutanceau and J.-M.
Leger, Electrochim. Acta, 2004, 49, 3901.
´
ger, J. Power Sources, 2002, 105, 283.
31 J.-M. Le
´
ger, Electrochim. Acta, 2007, 50, 3123.
32 W. Zhou, Z. Zhou, S. Song, W. Li, G. Sun, P. Tsiakaras and Q.
Xin, Appl. Catal., B, 2003, 46, 273.
33 T. J. Schmidt, H. A. Gasteiger, G. D. Stab, P. M. Urban, D. M.
¨
Kolb and R. J. Behm, J. Electrochem. Soc., 1998, 145, 2354.
34 T. J. Schmidt, H. A. Gasteiger and R. J. Behm, J. Electrochem.
Soc., 1999, 146, 1296.
´
3 X. Zhao, L. H. Jiang, G. Q. Sun, S. Yang, B. Yi and Q. Xin, Chem.
J. Chin. Univ., 2004, 26, 1304.
4 B. Beden, M. C. Morin, F. Hahn and C. Lamy, J. Electroanal.
Chem., 1987, 229, 353.
5 P. Gao, S.-C. Chang, Z. Zhou and M. J. Weaver, J. Electroanal.
Chem., 1989, 272, 161.
6 J. M. Perez, B. Beden, F. Hahn, A. Aldaz and C. Lamy, J.
Electroanal. Chem., 1989, 262, 251.
7 T. Iwasita, B. Rasch and E. Cattaneo, Electrochim. Acta, 1989, 34,
1073.
8 T. Iwasita and E. Pastor, Electrochim. Acta, 1994, 39, 531.
35 Z. Jusys and R. J. Behm, J. Phys. Chem. B, 2001, 105, 10874.
36 Z. Jusys, J. Kaiser and R. J. Behm, Phys. Chem. Chem. Phys., 2001,
3, 4650.
37 H. Wang, Z. Jusys and R. J. Behm, J. Phys. Chem. B, 2004, 108,
19413.
38 G.-Q. Lu, S.-G. Sun, L.-R. Cai, S.-P. Chen, Z.-W. Tian and K.-K.
Shiu, Langmuir, 2000, 16, 778.
9 H. Hitmi, E. M. Belgsir, J.-M. Le
Electrochim. Acta, 1994, 39, 407.
´
ger, C. Lamy and R. O. Lezna,
39 D. S. Corrigan, L.-W. H. Leung and M. J. Weaver, Anal. Chem.,
1987, 59, 2252.
10 J. Shin, W. J. Tornquist, C. Korzeniewski and C. S. Hoaglund,
Surf. Sci., 1996, 364, 122.
40 L. H. Jiang, G.-Q. Sun, Z. Zhou, W.-J. Zhou and X. Qin, Catal.
Today, 2004, 93–95, 665.
11 T. Iwasita, R. Dalbeck, E. Pastor and X. Xia, Electrochim. Acta,
1994, 39, 1817.
41 Z. Liu, X. Y. Ling, J. Y. Lee, X. Su and L. M. Gan, J. Mater.
Chem., 2003, 13, 3049.
12 X. H. Xia, H.-D. Liess and T. Iwasita, J. Electroanal. Chem., 1997,
437, 233.
42 Z. Liu, J. Y. Lee, W. Chen, M. Han and L. M. Gan, Langmuir,
2004, 20, 181.
13 F. Delime, J.-M. Le
29, 1249.
´
ger and C. Lamy, J. Appl. Electrochem., 1999,
43 W.-X. Chen, J. Y. Lee and Z. Liu, Mater. Lett., 2004, 58, 3166.
44 JCPDS, International Centre for Diffraction Data, 1987.
45 L. H. Jiang, Z. Zhou, W. Li, W.-J. Zhou, S. Somg, H.-Q. Li, G.-Q.
Sun and X. Qin, Energy Fuels, 2004, 18, 866.
46 C.-H. Lee, C.-W. Lee, D.-I. Kim and S.-E. Bae, Int. J. Hydrogen
Energy, 2002, 27, 445.
14 F. Vigier, C. Coutanceau, A. Perrard, E. M. Belgsir and C. Lamy,
J. Appl. Electrochem., 2004, 34, 439.
15 J. K. Wang, N. M. Markovic and R. R. Adzic, J. Phys. Chem. B,
2004, 108, 4127.
16 V. M. Schmidt, R. Ianniello, E. Pastor and S. Gonzalez, J. Phys.
Chem., 1996, 100, 17901.
47 T. Frelink, W. Visscher and J. A. R. van Veen, Electrochim. Acta,
1994, 39, 1871.
17 R. Ianniello, V. M. Schmidt, J. L. Rodriguez and E. Pastor, J.
Electroanal. Chem., 1999, 471, 167.
´
18 C. Lamy, E. M. Belgsir and J.-M. Leger, J. Appl. Electrochem.,
48 Z. Jusys, J. Kaiser and R. J. Behm, Langmuir, 2003, 19, 6759.
49 M. H. Shao and R. R. Adzic, Electrochim. Acta, 2005, 50,
2415.
2001, 31, 799.
19 L. Jiang, G. Sun, S. Sun, J. Liu, S. Tang, H. Li, B. Zhou and Q.
Xin, Electrochim. Acta, 2005, 50, 5384.
20 F. Colmati, E. Antolini and E. R. Gonzalez, J. Power Sources,
2006, 157, 98.
21 X.-S. Zhao, G.-Q. Sun, L. H. Jiang, Q. Mao, Q. Xin, B.-L. Yi and
S.-H. Yang, Chem. J. Chinese Univ., 2005, 26, 1304.
22 F. Vigier, C. Coutanceau, F. Hahn, E. M. Belgsir and C. Lamy, J.
Electroanal. Chem., 2004, 563, 81.
50 M. Heinen, Z. Jusys, Y.-X. Chen and R. J. Behm, to be published.
51 H. Wang, Z. Jusys and R. J. Behm, Fuel Cells, 2004, 4, 113.
52 Q.-S. Chen, S.-G. Sun, J.-W. Yan, J. T. Li and Z.-Y. Zhou,
Langmuir, 2006, 22, 10575.
53 M.-S. Zheng and S.-G. Sun, J. Electroanal. Chem., 2001, 500,
223.
54 C. Pecharroman, A. Cuesta and C. Gutie
Chem., 2002, 529, 145.
55 S.-G. Sun, J. Electroanal. Chem., 2002, 529, 155.
´
rrez, J. Electroanal.
23 G. A. Camara and T. Iwasita, J. Electroanal. Chem., 2005, 578, 315.
24 G. A. Camara, R. B. de Lima and T. Iwasita, J. Electroanal.
Chem., 2006, 585, 128.
25 L. Colmenares, H. Wang, Z. Jusys, L. Jiang, S. Yan, G. Q. Sun and
R. J. Behm, Electrochim. Acta, 2006, 52, 221.
56 A. Kabbabi, R. Faure, R. Durand, B. Beden, F. Hahn, J.-M. Leger
´
and C. Lamy, J. Electroanal. Chem., 1998, 444, 41.
57 T. Yajima, H. Uchida and M. Watanabe, J. Phys. Chem. B, 2004,
108, 2654.
58 J. G. Wu, Modern Fourier Transform Infrared Spectroscopy Tech-
niques and their Applications, Scientific and Technical Document
Publishing House, Beijing, 1994.
26 H. Wang, Z. Jusys and R. J. Behm, J. Power Sources, 2006, 154,
351.
ꢀc
This journal is the Owner Societies 2007
2696 | Phys. Chem. Chem. Phys., 2007, 9, 2686–2696