Probing n-propanol electrochemical oxidation on bimetallic PtRh codeposited electrodes

Article abstract of DOI:10.1149/1.1532327

The adsorption and reaction of n-propanol was investigated using in situ Fourier transform IR spectroscopy and on-line differential electrochemical mass spectrometry on electrodeposited Pt, Rh, and PtRh with different compositions. It has been observed th

Full text of DOI:10.1149/1.1532327

Journal of The Electrochemical Society, 150 ͑2͒ E89-E94 ͑2003͒
E89
0013-4651/2003/150͑2͒/E89/6/$7.00 © The Electrochemical Society, Inc.
Probing n-Propanol Electrochemical Oxidation on Bimetallic
PtRh Codeposited Electrodes
a
,z
*
I. A. Rodrigues, K. Bergamaski, and F. C. Nart
´
˜
˜
˜
Instituto de Quımica de Sao Carlos-Universidade de Sao Paulo, 13560-970 Sao Carlos, SP, Brazil
The adsorption and reaction of n-propanol was investigated using in situ Fourier transform IR spectroscopy and on-line differential
electrochemical mass spectrometry on electrodeposited Pt, Rh, and PtRh with different compositions. It has been observed that the
bimetallic electrodes were more active than pure platinum below 0.9 V. The pure rhodium electrode was practically inactive.
Differences in product yield show that platinum is more active than the bimetallic electrodes for propionic acid formation, but the
bimetallic electrodes show higher activity for CO2 and propanal production. The electrochemical reduction of the strongly
adsorbed intermediates on pure platinum and on the two bimetallic electrodes gave products with 1, 2, and 3 carbons, while the
pure rhodium electrode produced only methane. The degree of coverage by the irreversibly adsorbed species is about ten times
higher on platinum than on the bimetallic electrodes or rhodium, showing that on the bimetallic electrodes the intermediates are
not as strongly adsorbed as on pure platinum.
© 2003 The Electrochemical Society. ͓DOI: 10.1149/1.1532327͔ All rights reserved.
Manuscript submitted March 21, 2002; revised manuscript received August 13, 2002. Available electronically January 2, 2003.
The electrochemical oxidation of n-propanol on platinum elec-
trodes has been the subject of a detailed study1 using auxiliary tech-
niques like in situ Fourier transform infrared spectroscopy ͑FTIRS͒
and on-line mass spectrometry to detect products and intermediates.
It was found that the oxidation leads to CO2 , propanal, and propi-
onic acid for the oxidation products. Stable intermediates with C-H
stretching bands were observed and the adsorbed species may con-
tain CH3 and CH2 stretching vibrations. However, a very character-
istic band from adsorbed alkoxy species, the C-O-H deformation
band at ca. 1430 cmϪ1, was not reported and a definitive identifica-
tion of the adsorbed species was not given. However, it is clear that
the reduction of the adsorbed intermediates gives ethane and pro-
pane. Ethane is ca. 3.5 more abundant than propane. Detection of
ethane and propane show that C2 and C3 adsorbates are very likely,
because reduction products from adsorbed CO on platinum elec-
trodes, generally yield only C1 hydrocarbon compounds.2
efficiency depends on the electrode material with Ir being a better
electrocatalyst than Rh for selectivity and rhodium for the total oxi-
dation of ethanol to CO2 . The ethanol total oxidation requires C-C
bond cleavage, which is energetically more difficult than only the
C-H bond dissociation and a parallel route leading to the C2 partial
oxidation products takes place with the production of ethanal and
acetic acid. Thus, rhodium seems to provide more active sites for
C-C bond dissociation than iridium electrodes.
As pointed out above, the same has been observed for C3 alco-
hols, where propanal and propionic acid are observed parallel to the
production of CO2 .1 The C-C bond stability increases with the chain
length and the oxidation of C3 alcohols to CO2 is more difficult than
ethanol.3 In a previous communication,9 we showed that the surface
roughness of platinum electrodes improves the C-C bond dissocia-
tion for C3 alcohols. On the other hand, it would be interesting to
find also a combination of metals that could afford a suitable catalyst
able to decrease the C-C bond energy and promote a higher yield to
the total oxidation of C3 alcohols.
In this study we are interested in the examination of the role of
bimetallic electrodes containing rhodium on the C-C bond dissocia-
tion. This can be accomplished by following the production of CO2
͑a measure of C-C bond dissociation͒. Furthermore, rhodium elec-
trodes produce oxygen adsorption and alcohol oxidation at lower
potentials than platinum and this can also improve the C-O bond
formation for the total oxidation. In order to investigate these effects
we use FTIRS and differential electrochemical mass spectrometry
͑DEMS͒ as analytical techniques.
The key step for total oxidation of a multicarbon atom alcohol is
the C-C bond dissociation and the C-O coupling reactions, while
partial oxidation requires only C-H dissociation for propanal and
C-H bond dissociation followed by C-O bond formation for the acid.
Thus a good catalyst for total oxidation must provide sites for C-H
and C-C bond dissociation and sites for active oxygen at low poten-
tials to promote the C-O bond formation, in order to carry out the
oxidation to CO2 . Therefore, a more complete oxidation to CO2 or
acid entails the supply of oxygen species, which may come from the
water in solution or from surface oxides formed on the electrode
surface at different potentials. Indeed, oxides on platinum electrode
surfaces, depending on the oxidation state ͓Pt-OH, Pt-(OH)2 and
PtO͔, can be classified as active species. Oxides of platinum at
Experimental
higher oxidation states Pt(OH) and PtO2] poison the reaction.3
͓
3
Electrodes.—Potentiostatic deposition of Pt or codeposition of Pt
and Rh onto a smooth Au substrate ͑a disk of 0.32 cm2 geometric
area, previously polished to a mirror finish͒. For DEMS the electro-
chemical deposition were made on a gold layer ͑1.13 cm2 area, 50
nm thickness͒ prepared by gold sputtering onto a Scimat membrane
͑thickness 60 ␮m, mean pore size 0.17 ␮m, 50% porosity͒. All
electrodepositions were performed in a 0.1 M HClO4 solution con-
taining the appropriate amount of Pt and Rh salts for 5 min at 0.2 V
vs. reversible hydrogen electrode ͑RHE͒. The electrode active area
was determined by adsorbing CO and recording the amount of CO2
by measuring the charge required to oxidize the CO monolayer. This
charge was used to calculate the active surface area assuming one
CO adsorbed per active site. The active surface area was used to
normalize the current in the cyclic voltammograms. The normaliza-
tion of the FTIRs and DEMS were made by recording the amount of
CO2 released in the solution upon oxidation of the CO monolayer in
a potentiostatic step ͑FTIR͒ and during a voltammetric cycle
͑DEMS͒. The normalization using the FTIR and DEMS results are
Ruthenium has been added to platinum electrodes to supply active
oxygen at overpotentials lower than platinum. This mechanism has
been observed for methanol oxidation in many studies.4 It has been
a consensus in the literature that ruthenium is one of the most prom-
ising second elements in methanol electro-oxidation.5-7 In a recent
study, we showed that ruthenium presents also effects on ethanol
and n-propanol oxidation.8,9 Other elements have not been explored
extensively. Particularly, the interest in the use of rhodium-
containing electrodes has been scarce. It has been reported that a
pure rhodium electrode has smaller catalytic activity for methanol
electro-oxidation than pure platinum electrodes.10 The catalytic ac-
tivity of the rhodium and iridium electrodes for ethanol oxidation
was compared by Tacconi et al.11 They concluded that the oxidation
* Electrochemical Society Active Member.
a
˜
˜
Present address: Universidade Federal do Maranhao, Sao Luis, MA, Brazil.
z E-mail: nart@iqsc.sc.usp.br
Downloaded on 2015-06-22 to IP 128.192.114.19 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

E90
Journal of The Electrochemical Society, 150 ͑2͒ E89-E94 ͑2003͒
Table I. Bulk composition of the different electrodes used for in
situ FTIR and DEMS experiments obtained by energy dispersive
X-ray analysis and the corresponding solution composition used
for the electrodeposition of the electrodes.
Electrode
compositions compositions
Electrode
Solution composition
H2PtCl6•6H2O RhCl3•3H2O HClO4
͑FTIRS͒
͑DEMS͒
Pt
Rh
Pt0.75Rh0.25
Pt0.63Rh0.37
Pt
Rh
Pt0.75Rh0.25
Pt0.55Rh0.45
20 mM
1 M
1 M
1 M
1 M
20 mM
8.5 mM
20 mM
20 mM
20 mM
necessary, since the amount of products detected by these techniques
depends on factors other than the active surface area, as discussed in
previous publications.12,13
An RHE in the electrolyte solution ͑0.1 M HClO4) was used as
reference and a platinized Pt foil as counter electrode.
Chemicals.—Solutions were prepared with Millipore Milli Q wa-
ter and analytical grade chemicals: HClO4 acid ͑70%͒, propanol
͑99.9%͒, H2PtCl5•6H2O ͑Aldrich͒, and RhCl3•3H2O ͑Aldrich͒. Ni-
trogen ͑99.96%͒ was used to degas the solutions and CO ͑99.9%͒ to
carry out the normalization of the electrodes. The electrode compo-
sitions were determined by energy dispersive X-ray fluorescence
͑EDX͒. The composition of the solutions used for the electrodepo-
sition and the final electrode composition are listed in Table I. The
solutions used were 0.1 M HClO4 as a background electrolyte and a
0.1 M n-propanol in 0.1 M HClO4 for the reaction studies.
Instrumentation and procedures.—The FTIRS was a BOMEM
DA8 provided with a liquid nitrogen cooled mercury-cadmium tel-
luride ͑MCT͒ detector. A cell made of poly͑tetrafluoroethylene͒
͑PTFE͒ equipped with a CaF2 window was used for the in situ FTIR
experiments. Each single-beam spectrum was computed after aver-
aging 256 interferometer scans, taken with 8 cmϪ1 resolution, at
different potentials in the range 0.3-1.2 V. The spectra are presented
in the form of the reflectance ratio R/R0 of a single-beam spectrum,
R, obtained at a given potential and a reference spectrum, R0 , ob-
tained at 0.25 V.
Home-made DEMS equipment was used. The quadrupole mass
spectrometer used was a MKS Instrument. Details of this equipment
are given in Ref. 13. The electrochemical cell was constructed as
described in Ref. 14. The experiment consists in recording simulta-
neously the current-potential ͑I-E͒ and mass intensity-potential vol-
tammograms ͑potential positive scans and mass signal͒ for selected
values of m/z ͑mass/charge͒ ion signals. The potential was cycled in
the range 0.05-1.0 V and the scan rate was 0.01 V sϪ1
.
Propanol adsorption experiments were performed in the flow cell
under constant potential conditions. Thus, after the activation of the
working electrode in the 0.1 M HClO4 , with the electrode polarized
at Ead 0.35 V, propanol was added to the cell in order to have a 0.1
M propanol solution. The electrode was held for 5 min in order to
generate the stable adsorbed intermediates. After the adsorption time
the propanol-containing solution in the cell was replaced by the 0.1
M HClO4 solution under potential control. Finally, DEMS measure-
ments were performed in the range 0.35 to Ϫ0.05 V.
Results
In situ infrared spectroscopic study.—All the in situ spectra ob-
tained during the n-propanol oxidation presented the same feature as
observed before for platinum and PtRu bimetallic electrodes.1,9 So
Figure 1. ͑a͒ Normalized integrated band intensity of CO2 obtained from
n-propanal oxidation in a 0.1 M propanol in 0.1 M HClO4 solution on dif-
ferent electrode compositions. ͑b͒ Normalized integrated band intensity of
propionic acid obtained during oxidation of 0.1 M propanol in 0.1 M HClO4
solution on different electrode compositions. The symbols are ͑᭿͒ Pt, ͑᭡͒
Pt75Rh25 , ͑᭢͒ Pt63Rh37 , and ͑ࡗ͒ Rh.
here we use only the integrated band intensity for CO2 ͑2344 cmϪ1
͒
and propionic acid ͑1226 cmϪ1͒ as a function of the oxidation po-
tential ͑Fig. 1a, b, respectively͒ to evaluate the electrocatalytic ac-
tivity and selectivity of the electrodes. Clearly the Pt0.75Rh0.25 elec-
trode produces the highest band intensity for CO2 ͑Fig. 1a͒.
Downloaded on 2015-06-22 to IP 128.192.114.19 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑2͒ E89-E94 ͑2003͒
E91
starts at 0.6 V is followed through the mass signal m/z ϭ 44. How-
ever, since m/z ϭ 29 is also found for acetaldehyde, the possibility
of acetaldehyde production must be tested. The fragmentation of
propanal, under the conditions of the experiments, indicates that
acetaldehyde must present also a mass signal at m/z ϭ 43, which is
not observed in the mass signal shown in Fig. 2. Likewise, the
correction of the m/z ϭ 29 for propanal fragmentation shows that
the mass signal m/z ϭ 29 corresponds only to the fragmentation of
propanal. In that case, the production of lower aldehydes has been
ruled out. Based on this result, it is important to stress that the mass
signal m/z ϭ 44 is due only to CO2 , since propanal does not
present a mass signal at m/z ϭ 44. The direct consequence of this
observation is that once one C-C bond has dissociated, the remain-
ing C2 fragment also reacts to give CO2 .
The products observed by FTIR do not allow the discrimination
between propanal and propionic acid, but the mass signal m/
z ϭ 58 shows unequivocally the production of propanal, as shown
before for platinum electrodes. The lack of detection of propionic
acid by mass spectrometry can be attributed to the low volatility of
propionic acid.11 Recently in a study of ethanol oxidation on Pt and
PtRu porous electrodes15 acetic acid has been detected by DEMS
measurements; however, the mass intensity was about 10 times
smaller in comparison to acetaldehyde and CO2 .
The results of the propanal and CO2 yield for four different elec-
trodes are show in Fig. 3. In the range 0.4-0.8 V, all the bimetallic
electrodes present higher activity than pure Pt for CO2 production.
Related to the higher current density for the alloys is the onset of the
reaction for propanal and CO2 . The partial oxidation to propanal
starts at 0.35 V for the most active Pt0.75Rh0.25 alloy electrode, while
CO2 production starts at ca. 0.5 V. The reaction pathway leading to
propanal on pure platinum electrodes starts at ca. 0.6 V, at the same
potential as for the CO2 onset. The order of activities, based on the
Figure 2. Positive scan voltammetry and mass signal-potential curves ob-
tained for the cyclic oxidation of 0.1 M propanol in 0.1 M HClO4 on elec-
trodeposited Pt electrode.
Ͼ
Ͼ
Ͼ
Pt Rh.
propanal and CO2 yield, is Pt0.75Rh0.25
Pt0.55Rh0.45
These results are in good agreement with those obtained from the
FTIR experiments.
The activity of the pure platinum electrode increases for more
positive potentials. Specifically the propanal production starts to be
more intense for the pure platinum electrode at potentials above 0.9
V. The pure rhodium electrode displays only a very low activity for
both CO2 and propanal.
Comparing the different electrodes for the CO2 yield in the range
between 0.6 and 1.1 V, the electrocatalytic activity for the total
oxidation can be arranged in the following sequence: Pt0.75Rh0.25
Although the adsorbed species generated from the dissociative
adsorption of the n-propanol were not detected in the in situ FTIR
spectra, the products obtained from the reduction reaction of the
strongly adsorbed species can be a clue to identify the stable reac-
tion intermediates.
The cyclic voltammetric curves ͑Fig. 4͒ for the irreversibly ad-
sorbed intermediates show a current density ten times higher for the
platinum electrode, compared to the bimetallic ones. The current
density is directly proportional to the degree of coverage, it is clear
that the surface coverage by the strongly adsorbed intermediates is
strongly inhibited on the bimetallic electrodes.
The reduction reaction products obtained for the propanol ad-
sorbed fragments were followed with the mass signals m/z ϭ 43,
m/z ϭ 30, and m/z ϭ 15 ͑Fig. 5͒. These mass signal account for
propane, ethane, and methane ͑note that the mass signal m/z ϭ 15
has been corrected for propane and ethane fragmentation and there-
fore it is only due to methane͒. Although the mass signals are cor-
rected for fragmentation, each product has different mass peaks and
quantitative determination needs the determination of all fragments.
Therefore it is not possible to obtain quantitative data, since the
relative amount of products is not known exactly. However, the
large difference in current density and mass signals between pure
platinum and the bimetallic electrode indicates that a large differ-
ence in degree of coverage exists between these two kinds of
electrodes.
Ͼ
Ͼ
Rh. For the partial oxidation to propionic
Ͼ Pt0.63Rh0.37
Pt
acid ͑Fig. 1b͒, the Pt electrode presents the highest activity over the
whole range of potentials studied. The sequence of activities in this
Ͼ
Ͼ
Ͼ
Pt0.63Rh0.37 Rh.
case is Pt
Pt0.75Rh0.25
The rhodium electrode presents the lowest activity for the
electro-oxidation of propanol. On this electrode the CO2 band is
approximately ten times less intense than observed for the
Pt0.75Rh0.25 electrode. The rhodium electrode is almost inactive for
propionic acid production.
The results of Fig. 1 show a very interesting feature. The CO2
production reaches a maximum at 0.8 V, while the propionic acid
production does not present a maximum. This result shows that the
platinum oxide, which is formed above 0.8 V, may act as a poison
for the C-C bond dissociation ͑inhibition of CO2), but not to C-O
coupling, since oxidation to propionic acid is not inhibited.
Differential electrochemical mass spectrometry (DEMS)
studies.—The volatile reaction products of the n-propanol oxidation
were followed by DEMS. In Fig. 2 the mass signal obtained during
the electro-oxidation of n-propanol on electrodeposited Pt are
shown.
The oxidation of n-propanol leads to the production of propanal
and CO2 , as already reported.1 The reaction leading to propanal
starts at 0.5 V, and it is followed using the mass signals at m/
z ϭ 58 and m/z ϭ 29. The former corresponds to the propanal radi-
cal cation and the latter is related to the (COH)ϩ fragment, which is
the main peak of the mass spectrum of propanal.12 CO2 , which
Except on rhodium electrodes, for which only m/z ϭ 15 is ob-
served, the other electrodes present detectable values for the three
Downloaded on 2015-06-22 to IP 128.192.114.19 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

E92
Journal of The Electrochemical Society, 150 ͑2͒ E89-E94 ͑2003͒
Figure 4. Cyclic voltammograms obtained for the products obtained upon
reducing electrochemically the irreversibly adsorbed intermediates in the ab-
sence of n-propanol in solution: ͑
͒ Pt, ͑----͒ Pt0.75Rh0.25 , ͑-•-•-•-͒
Pt0.55Rh0.45 , and ͑-••-••-••-͒ Rh electrodes. The irreversibly adsorbed inter-
mediates were formed by contacting the electrode polarized at 0.35 V in 0.1
M propanol in 0.1 M HClO4 for 5 min and then replacing the solution by 0.1
M HClO4 . The data are normalized by the area occupied by adsorbed CO on
each electrode ͑active surface area͒.
different bulk composition. The bimetallic electrodes presented the
best electrocatalytic activity for n-propanol oxidation at potentials
lower than 0.9 V. The best composition tested in the present paper is
Pt0.75Rh0.25 . The overall electrocatalytic activity correlates well with
the CO2 and propanal yield for the electrodes studied, meaning that
those products control the overall reaction rate for potentials below
0.9 V.
Figure 3. Positive scan current and mass signals as a function of the applied
potential obtained during the voltammetric oxidation of 0.1 M propanol in
The composition of the bimetallic electrode is the bulk compo-
sition and not the surface composition. For bimetallic PtRu elec-
trodes it has been reported that the surface composition differs from
the bulk. There is platinum enrichment of the surface, depending on
the surface pretreatment.16 Surface enrichment for PtRh can also be
expected, but no data on the differences in surface and bulk compo-
sition exists for these electrodes. Independent of the final surface
composition, note that in an early study, using bimetallic PtRh as
CO-tolerant electrode for H2 oxidation, it was suggested that the
increased electrocatalytic effect for H2 oxidation is related to an
electronic effect instead of a bifunctional mechanism. Therefore,
addition of rhodium is not only to supply active oxygen. In that case
the composition at the surface may not be so important as it is in the
case of a bifunctional mechanism. In our case strong evidence that
the bifunctional mechanism is not so important, like in the case of
PtRu bimetallic electrodes, is that the addition of rhodium to the
electrode composition strongly shifts the onset potential for propa-
nal, but not for CO2 . Production of propanal entails only C-H bond
dissociation and not C-O bond formation, for which the presence of
oxygen at lower potentials does not affect the reaction rate. On the
other hand, on PtRu bimetallic electrodes the effect of ruthenium on
0.1 M HClO4 on ͑
͒ Pt, ͑-----͒ Pt0.75Rh0.25 , ͑-•-•-•-͒ Pt0.55Rh0.45 , and
͑-••-••-••-͒ Rh electrodes. The data are normalized to the integrated CO2
mass signal generated by a monolayer of adsorbed CO.
mass signals. As in the case of the cyclic voltammetric measure-
ments of the reduction of the stable adsorbed intermediates, plati-
num presents the largest mass signals for the three adsorbates. It is
very difficult to quantify the total amount of adsorbed material from
the mass signal, including the relation between the different species.
It is interesting that on rhodium, no C2 and C3 adsorbates were
observed. On the other hand, the adsorbates on rhodium present by
far the lowest degree of coverage. In general, it can be concluded
that rhodium practically does not present any meaningful electro-
catalytic activity.
Discussion
The use of a normalization procedure allows a confident com-
parison of the electrode activity for the electro-oxidation of
n-propanol on Pt, Rh, and PtRh coelectrodeposited electrodes with
9
the onset potential for CO2 is in good harmony with the bifunc-
Downloaded on 2015-06-22 to IP 128.192.114.19 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 150 ͑2͒ E89-E94 ͑2003͒
E93
Figure 6. Possible general reaction scheme for the oxidation of n-propanol
on platinum and PtRh bimetallic electrodes in acid solutions. The ͑s͒ means
strongly bonded and ͑w͒ weakly bonded intermediates. C1, C2, and C3 are
for fragments of the adsorbed intermediates containing one, two and three
carbon atoms, respectively.
bates control the overall reaction rate, even for CO2 production, as
already anticipated before for pure platinum electrodes.1 Evidence
connected to this observation is the result that the oxidation of the
isolated stable adsorbates produces only CO2 , as published before.1
If CO2 would be produced only from the strongly adsorbed inter-
mediates, a higher CO2 yield should be expected for platinum.
Therefore, a lower degree of coverage by the strongly adsorbed
intermediates on the bimetallic electrodes facilitates both the total
and partial oxidation via the weakly adsorbed intermediates.
Summarizing, propanal and CO2 can be produced via weakly
adsorbed intermediates, but CO2 can originate from the strongly
bonded intermediates as well. Additionally, simultaneous observa-
tion of propanal, CO2 , and propionic acid suggests that the reaction
follows parallel pathways on all the electrodes. A tentative general
reaction scheme is proposed in Fig. 6.
In this scheme, the basic difference between platinum and PtRh
bimetallic electrode is that on the bimetallic electrodes the coverage
by the strongly bonded intermediates is much less than for platinum
electrodes and the pathways leading to the weakly bonded interme-
diates prevail, leading to a higher reactivity.
The higher degree of coverage with strongly adsorbed interme-
diates on platinum electrodes must relate to the higher reaction onset
potential and lower electrocatalytic activity for propanol oxidation
on platinum, compared to the bimetallic electrodes. As mentioned
above, the strongly adsorbed intermediates poison the electrode sur-
face, requiring higher oxidation potential to cause the reaction to
proceed. The production of propanal starts at 0.35 and 0.45 V on the
Pt0.75Rh0.25 and Pt0.55Rh0.45 electrodes respectively, while pure plati-
num presents the onset for propanal at ca. 0.6 V. Thus, poisoning by
the strongly bonded intermediates inhibits not only the C-C bond
dissociation but also the C-H bond dissociation.
Figure 5. Mass signals obtained during the negative scan for reducing the
strongly adsorbed intermediates in 0.1 M HClO4 : The symbols are ͑
͒
• • • •
Pt, ͑
͒ Pt0.75Rh0.25 , ͑– – –͒ Pt0.55Rh0.45 , and ͑– • – • –͒ Rh. The
irreversibly adsorbed intermediates were formed by contacting the electrode
polarized at 0.35 V in 0.1 M propanol in 0.1 M HClO4 for 5 min and then
replacing the solution by 0.1 M HClO4 . The data are normalized to the
integrated CO2 mass signal generated by a monolayer of adsorbed CO.
Another important observation in relation to the reaction path-
ways is the difference in propionic acid and propanal yield as shown
by the FTIR and DEMS results. Propionic acid presents the best
yield on pure platinum electrodes, while the best yield for propanal
is observed for the Pt0.75Rh0.25 bimetallic electrode. If propionic acid
would originate exclusively from propanal oxidation, then it should
be expected that the best propanal yield detected by DEMS would
lead also to the best propionic acid yield detected by in situ FTIR
spectroscopy. That pure platinum favors the pathway to propionic
acid, while the Pt0.75Rh0.25 favors propanal production, indicates that
there are parallel pathways for propanal and propionic acid instead
of a purely sequential pathway as suggested before.1 The sequential
pathway may also be another way to produce propionic acid, but the
main pathway is the direct oxidation of propanol to the acid. On the
other hand, the higher propionic acid yield on platinum electrodes
than on the bimetallic electrodes suggests that the production of
propionic acid may be related also to the adsorbed C3 stable adsor-
bate, since the production of propanal seems not to depend on the C3
tional mechanism already proposed for methanol oxidation for this
alloy. Unfortunately the propanal and CO2 production cannot be
compared with PtRu bimetallic electrode, since no DEMS results are
reported for this system. Only in situ FTIR results have been
reported,9 but in this case the differences in CO2 and propanal yield
can be due to the differences in surface area.b
A remarkable result is the difference in the current and mass
signal of the stable adsorbates on the different electrodes. Pure plati-
num electrodes present the highest degree of coverage for strongly
adsorbed intermediates on all the electrodes. In addition, the highest
electrocatalytic activity of the bimetallic electrodes indicates that the
important intermediates are only weakly adsorbed and these adsor-
b The in situ FTIR can be strongly affected by the total surface area of the electrode,
since the reaction occurs in a very thin layer of solution between the electrode and the
infrared window, which is practically disconnected from the bulk solution in terms of
diffusion. In that case a larger area will produce larger reactant depletion with a conse-
quent change in the propanal/CO2 ratio.
Downloaded on 2015-06-22 to IP 128.192.114.19 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

E94
Journal of The Electrochemical Society, 150 ͑2͒ E89-E94 ͑2003͒
stable adsorbate coverage. However, no direct evidence can be sup-
plied from the present experiments.
by the platinum oxide, but not the dehydrogenation reaction and the
C-O bond formation, since propionic acid shows an increasing yield
on the pure platinum electrode.
The rhodium electrode is practically inactive compared to pure
platinum or to the codeposited Pt-Rh electrodes. In this connection it
is interesting to discuss the C-C bond dissociation efficiency on
these electrodes. It is clear that, although pure rhodium electrodes
are practically inactive for n-propanol oxidation, the contact of
n-propanol with the rhodium surface gives rise only to stable C1
fragment adsorbates. This contrasts with pure platinum or the code-
posited PtRh, where C1 , C2 , and C3 fragments were found. The
lack of C3 adsorbates on rhodium electrodes indicates that the
rhodium stabilizes only adsorbates with C1 fragments. Indeed, in a
previous study,17 it was found that the presence of rhodium in the
electrode composition increases the C-C bond dissociation for eth-
anol oxidation. This higher ability to dissociate the C-C bond was
assigned to a modified electronic structure caused by the presence of
rhodium on the electrode. However for the case of n-propanol the
CO2/propanal ratio between the bimetallic electrodes and platinum
are comparable. This means that the C-C bond dissociation is not as
effective as in the case of ethanol.15 On the other hand, the higher
yield both for propanal and CO2 on the bimetallic electrodes clearly
indicates a better catalytic efficiency of these materials. As men-
tioned above, this superior catalytic activity seems to be related to
the lower adsorption energy of the reaction intermediates on the
bimetallic electrodes, allowing a higher turnover frequency for the
bimetallic electrodes.
Acknowledgments
The authors would like to acknowledge the FAPESP, CNPq, and
CAPES agencies for financing the project and fellowships.
˜
The University of Sao Paulo assisted in meeting the publication costs of
this article.
References
1. E. Pastor, S. Wasmus, T. Iwasita, M. C. Arevalo, S. Gonzalez, and A. J. Arvia, J.
Electroanal. Chem., 350, 97 ͑1993͒.
2. G. M. Brisard, A. P. M. Camargo, F. C. Nart, and T. Iwasita, Electrochem. Com-
mun., 3, 603 ͑2001͒.
3. N.-H. Li, S.-G. Sun, and S.-P. Shen, J. Electroanal. Chem., 430, 57 ͑1997͒.
4. M. Watanabe and M. Motoo, J. Electroanal. Chem. Interfacial Electrochem., 60,
267 ͑1975͒.
5. H. A. Gasteiger, N. Markovic, P. N. Ross, and E. J. Cairns, J. Phys. Chem., 97,
12020 ͑1993͒.
6. T. Iwasita, F. C. Nart, and W. Vielstich, Ber. Bunsenges. Phys. Chem., 94, 1030
͑1990͒.
7. V. Radmilovic, H. A. Gasteiger, and P. N. Ross, J. Catal., 154, 98 ͑1995͒.
8. J. P. I. de Souza, F. J. B. Rabelo, I. R. de Moraes, and F. C. Nart, J. Electroanal.
Chem., 420, 17 ͑1997͒.
9. I. A. Rodrigues, J. P. I. de Souza, E. Pastor, and F. C. Nart, Langmuir, 13, 6829
͑1997͒.
10. F. Hahn, B. Beden, and C. Lamy, J. Electroanal. Chem. Interfacial Electrochem.,
204, 315 ͑1986͒.
11. N. R. Tacconi, R. O. Lezna, B. Beden, F. Hahn, and C. Lamy, J. Electroanal.
Chem., 379, 329 ͑1994͒.
Conclusions
¨
12. B. Bittins-Cattaneo, E. Cattaneo, P. Konigshoven, and W. Vielstich, Electroanalyti-
cal Chemistry—A Series of Advances, Vol. 17, J. Bard, Editor, p. 181, Marcel
Dekker, Inc., New York ͑1991͒.
13. J. P. I. de Souza, S. L. Queiroz, and F. C. Nart, Quim. Nova, 23, 384 ͑2000͒.
14. R. Ianello and V. M. Schmidt, Ber. Bunsenges. Phys. Chem., 99, 83 ͑1995͒.
15. P. N. Ross, K. Kinoshita, A. J. Scarpellino, and P. Stonehardt, J. Electroanal. Chem.
Interfacial Electrochem., 59, 177 ͑1975͒.
16. H. A. Gasteiger, P. N. Ross, and E. J. Cairns, Surf. Sci., 293, 67 ͑1993͒.
17. J. P. I. de Souza, S. L. Queiroz, E. R. Gonzalez, and F. C. Nart, J. Phys. Chem. B,
106, 9825 ͑2002͒.
Bimetallic platinum-rhodium electrodes present a superior elec-
trocatalytic activity than the pure platinum electrode for applied po-
tentials below 0.9 V. The best electrocatalytic activity is for the
Pt0.75Rh0.25 . The pure rhodium electrode is practically inactive for
n-propanol oxidation.
The reaction rate below 0.9 V is governed by the weakly bound
intermediates. Above 0.9 V the C-C bond dissociation is inhibited
Downloaded on 2015-06-22 to IP 128.192.114.19 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).