Novel method for the estimation of the electroactive Pt area

Article abstract of DOI:10.1149/1.1586924

The (COOH)₂ oxidation reaction was studied at potentials below which the oxygen evolution reaction (OER) takes place. Pt was found to catalyze the (COOH)₂ oxidation reaction more strongly than Au, while Ru did not display any activity toward the (COOH)₂ oxidation reaction. Furthermore, under rapid stirring conditions, the (COOH)₂ oxidation reaction using Pt electrodes was shown to be activation controlled. Therefore, the (COOH)₂ oxidation currents can be related to the electroactive Pt area, as shown for a range of polycrystalline, bulk metal Pt, and Pt powder electrodes. The Pt surface area for multicomponent catalyst systems can also be estimated by combining (COOH)₂ oxidation data with the charge needed to oxidize adsorbed CO to CO₂ (CO_ads charge), as shown for a range of Pt- and Ru-containing powder electrodes. In fact, the combination of the two methods [(COOH)₂ oxidation current and CO_ads charge] can be used as an in situ probe to estimate the fraction of Ru in the metallic state in the potential region where CO is adsorbed provided the surface ratio of Pt vs. Ru is known.

Full text of DOI:10.1149/1.1586924

Journal of The Electrochemical Society, 150 ͑8͒ E377-E383 ͑2003͒
E377
0
013-4651/2003/150͑8͒/E377/7/$7.00 © The Electrochemical Society, Inc.
Novel Method for the Estimation of the Electroactive Pt Area
,
z
C. Bock* and B. MacDougall*
National Research Council Canada, Ottawa, Ontario, Canada K1A 0R6
The ͑COOH͒ oxidation reaction was studied at potentials below which the oxygen evolution reaction ͑OER͒ takes place. Pt was
2
found to catalyze the ͑COOH͒2 oxidation reaction more strongly than Au, while Ru did not display any activity toward the
͑
COOH͒ oxidation reaction. Furthermore, under rapid stirring conditions, the ͑COOH͒ oxidation reaction using Pt electrodes was
2
2
shown to be activation controlled. Therefore, the ͑COOH͒ oxidation currents can be related to the electroactive Pt area, as shown
2
for a range of polycrystalline, bulk metal Pt, and Pt powder electrodes. The Pt surface area for multicomponent catalyst systems
can also be estimated by combining ͑COOH͒2 oxidation data with the charge needed to oxidize adsorbed CO to CO2 (COads
charge͒, as shown for a range of Pt- and Ru-containing powder electrodes. In fact, the combination of the two methods ͓͑COOH͒2
oxidation current and COads charge͔ can be used as an in situ probe to estimate the fraction of Ru in the metallic state in the
potential region where CO is adsorbed provided the surface ratio of Pt vs. Ru is known.
©
2003 The Electrochemical Society. ͓DOI: 10.1149/1.1586924͔ All rights reserved.
Manuscript submitted December 26, 2002; revised manuscript received February 11, 2003. Available electronically June 23, 2003.
Pt and binary Pt-Ru catalysts have attracted much attention due
In the present work, the currents observed as a result of the
1
,2
to their promising use in low-temperature fuel cells. Despite many
studies and thorough characterization of these catalysts, the estima-
tion of the true electroactive surface area of these catalysts remains
electrochemical ͑COOH͒ oxidation reaction to CO are related to
2
2
the electroactive Pt area. The use of this ͑COOH͒ oxidation method
2
is probed for a range of unsupported powder catalysts such as Pt,
Ru, and two component Pt-Ru systems. First, the faradaic ͑COOH͒2
oxidation characteristics are discussed using Pt, Ru, and Au elec-
trodes. The ͑COOH͒2 oxidation reaction is then studied in detail
using polycrystalline Pt and Pt powder electrodes distinguishing ac-
tivation vs. mass-transport reaction rate conditions and testing the
a challenge. Anodic stripping voltammetry of adsorbed CO (COads
)
has been suggested as a possible in situ probe for these catalyst
systems yielding some information about catalyst composition and
surface areas.3 This method involves the adsorption of CO on the
catalyst surface at negative potentials and the oxidation of the COads
,4
to CO in a subsequent positive potential scan. Onto Pt surfaces, ca.
one monolayer of CO is adsorbed at sufficiently negative potentials,
accuracy of the ͑COOH͒ oxidation method. Pt areas estimated from
2
2
the ͑COOH͒ oxidation method are also compared to the experimen-
2
thus allowing the conversion of the experimentally obtained CO-to-
tally observed COads charges for a range of Pt and Pt-Ru powders.
Furthermore, the electrochemistry of the differently prepared Pt-Ru
powders is briefly discussed using cyclicvoltammetry ͑CV͒ charac-
teristics that are normalized for the Pt area estimated using the
3
CO oxidation charge to the electroactive Pt area. The CO strip-
2
ads
ping method can be employed to estimate the electroactive Pt area
for multicomponent catalyst systems provided that CO does not ad-
sorb on the non-Pt catalyst components. This is typically the case for
carbon as well as metal oxides. However, it is known that CO can
also adsorb on Au, Hg, and Ru.5 The case of CO adsorption on
surfaces that contain Ru is particularly complicated. Ru oxides, par-
͑
COOH͒ oxidation method.
2
,6
5
Experimental
Pt, Ru, and Pt-Ru powder preparation.—A range of Pt, Ru, and
ticularly in the presence of Pt, can be reduced to metallic Ru within
and below the potential region where molecular hydrogen is
adsorbed/desorbed on Pt, i.e., within the typically used potential
range for CO adsorption region. The electrochemistry of Ru is mani-
fold. Ru can be in the metallic state as well as form oxides of
various oxidation and hydration states over the potential ranges typi-
Pt-Ru powders were synthesized. The atomic Pt-to-Ru ratio of the
starting, i.e., the total bulk powder composition for the Pt-Ru pow-
ders was generally maintained constant at 70 to 30 atom %. The
following powders were prepared: powder 1, Pt powder formed by
the reduction of 3.2 g of H PtCl •6H O ͑Alfa Aesar͒ dissolved in
7
cally investigated for fuel cell applications. Clear knowledge of the
2
6
2
dependence of the Ru oxidation state and oxide form on potential
for Pt-Ru catalyst systems is lacking. Furthermore, it has been pro-
posed that different numbers of CO molecules ͑in some cases up to
1
50 mL H O with 75 mL of 0.2 M NaBH ͑Anachemia͒; powder 2,
2
4
Ru powder, formed as Pt powder 1, but using 1.6 g RuCl ͑Alfa
3
Aesar͒ instead of H PtCl •6H O; powder 3, PtRu alloy formed by
5
2
6
2
two͒ can adsorb onto a single Ru atom. It is therefore clear that the
the simultaneous reduction of 3.256 g H PtCl •6H O and 0.5603 g
2
6
2
use of only COads stripping voltammetry for the electroactive Pt area
estimation for binary, Ru-containing catalyst systems is very com-
plicated and questionable.
RuCl dissolved in 150 mL H O with 75 mL 0.2 M NaBH . Powder
3
2
4
4, Pt/RuO2 ͑thermal͒ formed by the reduction of 3.256 g
H PtCl •6H O dissolved in 150 mL H O with 75 mL 0.2 M
2
6
2
2
More recently, the use of Cu underpotential deposition ͑UPD͒
followed by subsequent oxidative stripping of the UPD Cu has been
NaBH , after successive washing with H O, filtering, and drying in
4
2
8
,9
an air oven at 100°C, 0.5603 g of RuCl was distributed ͑using a
discussed for the estimation of the electroactive Pt and Ru area.
3
mortar͒ within the Pt powder. This powder mixture was subse-
Green and Kucernak have shown that the Cu UPD method can be
8
quently heated at 500°C under O flow in a tube furnace for 2 h:
applied to supported and high-surface area catalyst systems. The Cu
2
Powder 5, Pt/RuO2 ͑ballmilled͒, 2.83 g Pt powder ͑200 mesh,
99.98%, Alfa Aesar͒ and 0.85 g RuO2 ͑electronic grade, Premion,
99.95%, Alfa Aesar͒ were ballmilled using a Spex 2000 ballmill
UPD method takes advantage of the fact that Cu is of similar size to
Pt and Ru and, hence, one Cu atom per Pt or Ru atom is deposited
in the UPD region. It has been shown that one Cu atom is deposited
by UPD per metallic Pt and Ru atom using unsupported PtRu alloy
powders.8 However, the application of the Cu UPD stripping
method is limited, as this method requires the exact knowledge of
the fraction of Ru present in the metallic state in the Cu UPD region.
1
0
mixer, as described in detail elsewhere. Powder 6, Pt/Ru, the
Pt/RuO2 ͑thermal͒ powder 4 was reduced in a H2 atmosphere at
100°C for 2 h. After preparation, powders 1-3 were washed thor-
oughly with H O, filtered, and dried in an air oven at 100°C. A PtRu
2
alloy powder of Pt to Ru atom % composition of 54:46 was also
prepared in the same manner as powder 3 by the simultaneous re-
duction of 3.256 g H PtCl •6H O and 0.8591 g RuCl . This pow-
*
Electrochemical Society Active Member.
E-mail: christina.bock@nrc.ca
2
6
2
3
z
der catalyst is referred to as powder 3a.

E378
Journal of The Electrochemical Society, 150 ͑8͒ E377-E383 ͑2003͒
Table I. Summary of catalyst powder characteristics for unused powders.
Powder
Characteristics identified using XRD analyzes10
Surface characteristics identified using XPS analyzes10
1
2
3
͑Pt͒
͑Ru͒
and 3a
Pt powder
Ru powder
PtRu alloy powders
Metallic Pt, PtOH
Mainly metallic Ru and RuO
Metallic Pt, PtOH, Ru and RuO
͑PtRu alloys͒
4
5
6
(Pt/RuO thermal͒
Pt and RuO powder
PtOH, PtO, PtO , RuO and RuO
2
Metallic Pt, PtOH, PtO, RuO and RuO2
Metallic Pt, PtOH, metallic Ru, RuO
2
2
2
(Pt/RuO ball milled͒
Pt and RuO powder
2
2
Pt/Ru (H reduced͒
Pt and Ru powder
2
Working electrode preparation.—The catalyst powders were
formed into anodes by sonicating 13 mg of a particular powder in 5
were deoxygenated using high purity argon gas prior to the electro-
chemical studies. All electrochemical experiments were carried out
at room temperature. ACS grade chemicals and high resistivity 18
M⍀ water were used. The RuCl3 powder was dried in air at 135°C
prior to its use for the catalyst powder synthesis.
to 20 mL of H O for 30 min. Appropriate amounts ͑1-2 ␮L͒ of the
2
2
suspension were then pipetted onto and well spread out on 0.2 cm
͑
geometrical area͒ glassy carbon disk electrodes ͑Pine͒. In some
cases, as stated in the appropriate section of the text, catalyst powder
suspensions were also pipetted onto ca. 0.35 cm Au foil electrodes
2
Results and Discussion
͓
99.9% Au, 0.1 mm thick ͑Goodfellow͔͒. The electrodes were dried
Characteristics of the catalyst powders.—In parallel work, XRD
and XPS spectroscopy were used for the characterization of the
catalysts powders used here and the results are summarized in Table
I. The raw XRD and XPS data are shown and discussed in detail
in air at room temperature and subsequently washed thoroughly with
H O. The Au foils were firmly attached to Au wire electrodes. The
Au electrodes were carefully wrapped with Teflon tape to mask Au
not covered with the catalyst powder and to expose a geometrical
area of ca. 0.12 cm . In some cases, polycrystalline Pt rotating disk
electrodes ͑RDEs͒ (0.2 cm geometrical area, Pine͒ were also used
as working electrodes. All ͑COOH͒ oxidation studies were carried
2
10
elsewhere. The XPS studies showed that both the Pt and Ru salt
2
precursors are reduced to mainly metallic Pt and Ru for the case of
powders 1 and 2, respectively. Furthermore, the XRD data showed
the Pt-Ru based powders 3 and 3a to consist of PtRu alloys, while
powders 4 and 6 exist of mixed phases of nonalloyed Pt and Ru
components. The Ru was found to consist in different oxide forms,
2
2
out using the powder electrodes prepared without the typically used
Nafion solution,11 as preliminary studies showed that Nafion slows
down the transport of ͑COOH͒ to the catalyst sites.
namely RuO and RuO ͑no metallic Ru͒, for powders 4 and 5 and
2
2
In some cases, as stated in the appropriate section of the text,
electrodes were also prepared as described above, however, 13 mg
of the powder was dispersed and sonicated in solutions consisting of
Ru in the mainly metallic state and lower oxide forms namely RuO
͑
no RuO ) for powders 3, 4, and 6.
2
300 ␮L Nafion 117 solution ͑5 wt % Nafion dissolved in lower
(COOH)2 oxidation characteristics. CV studies.—Figure 1
Ϫ1
Ϫ2
alcohols, Fluka͒ and 1 mL H O. Appropriate amounts ͑10-50 ␮L͒
shows typical CV characteristics recorded at 20 mV s in 10
M
2
were pipetted on Au foil electrodes. The general CV and CO strip-
ping characteristics of a particular powder electrode prepared with
and without Nafion were found to be the same.
͑COOH͒2 ϩ 0.5 M H2SO4 solutions using Pt and Ru powder elec-
trodes and a Au foil electrode. The Pt and Ru powders were sup-
ported using Au foil electrodes, as described in the Experimental
section. Steady-state CV characteristics were generally obtained
within less than five to seven complete oxidation/reduction cycles
Cells and electrodes.—Three compartment cells, in which the
reference electrode was separated from the working and counter-
electrode compartment by a Luggin capillary, were employed for the
electrochemical studies. A saturated calomel electrode ͑SCE͒ was
used as the reference electrode. All potentials reported in this paper
are vs. the SCE unless otherwise stated. Large surface area Pt gauzes
served as counterelectrodes.
for all ͑COOH͒ oxidation studies reported in this work. For the Pt
2
powder and Au foil electrodes, a ͑COOH͒ oxidation current is ob-
2
served, while the Ru powder electrode appears not to have any
activity for the ͑COOH͒2 oxidation reaction within the potential
Techniques and instrumentation.—Electrochemical experiments
were performed using either an EG&G PAR 273 potentiostat or a
Solartron SI 1287 electrochemical interface ͑Solartron Group, Ltd.͒
both driven by Corrware software program ͑Scribner, Assoc.͒. A
Pine rotating disc electrode ͑RDE͒ Rotator controlled by a MSR
speed control system ͑Pine Instrument Company͒ was employed for
the electrochemical oxidation studies carried out under controlled
stirring conditions. A JEOL JSM-5300 scanning microscope, a Kra-
tos Axis X-ray photoelectron spectrometer ͑XPS͒ and X-ray diffrac-
tion spectrometer ͑XRD, Scintag 2000͒ equipped with a Cu K␣
source were also employed to characterize the catalyst powders.
COads stripping voltammetry.—CO was adsorbed onto the Pt-
based powder electrodes at Ϫ0.1 V by bubbling CO gas ͑Matheson
purity, Matheson gas͒ through the 0.5 M H SO solution for 20 min.
2
4
Solution CO was subsequently removed by bubbling argon gas ͑Air
Products͒ for 40 min. holding the potential at Ϫ0.1 V. The potential
Ϫ1
was then cycled at 20 mV s between Ϫ0.1 V and specific upper
limit (Eϩ) for two complete oxidation/reduction cycles.
Solutions.—All ͑COOH͒ oxidation studies were carried out us-
Figure 1. Typical CVs for Au foil, Pt and Ru powder electrodes recorded at
20 mV s in quiescent 10 M (COOH)2 ϩ 0.5 M H SO solutions.
2 4
2
Ϫ2
Ϫ1
Ϫ2
ing 10 M ͑COOH͒ ϩ 0.5 M H SO solutions. The solutions
2
2
4

Journal of The Electrochemical Society, 150 ͑8͒ E377-E383 ͑2003͒
range studied here. The exact amount of the mainly metallic ͑see
E379
Table I͒ Ru powder deposited onto the Au support electrode was not
estimated, but a Ru powder layer was clearly visible, indicating the
presence of a sufficient amount of Ru. Furthermore, as observed for
the Ru powder electrode, the glassy carbon ͑GC͒ RDE was also
found not to display activity for the ͑COOH͒ oxidation reaction,
2
thus justifying the use of this electrode as support for the Pt and
Pt-Ru powders. The ͑COOH͒2 oxidation reaction is seen to take
place at more negative potentials ͑ca. 80 mV estimated roughly by
extrapolation as shown in Fig. 1͒ for the Pt powder than for the Au
electrode. This result suggests that Pt catalyzes the ͑COOH͒ oxida-
2
tion reaction more strongly than Au and much more effectively than
Ru. The fact that the ͑COOH͒2 oxidation reaction takes place at
more negative potentials on Pt than on Au and not at all on Ru
suggests that the ͑COOH͒2 oxidation currents observed for Pt-
containing electrodes are directly related to the Pt area, particularly
for electrodes exposing no or little Au to the electrolyte solution.
The reduction peak seen in the CV in Fig. 2 is due to the reduction
of Pt oxide formed during the positive cycle.
The electrochemical ͑COOH͒ oxidation reaction that takes place
2
at potentials below the OER using polycrystalline Pt electrodes has
4Ϫ
Figure 2. ip,a ͑छ͒ and ilim ͑᭝͒ dependence for the (COOH) and Fe(CN)6
2
been studied previously.12 It has been shown that in acidic solutions
1/2
2
oxidation reaction, respectively, on ␻ using a 0.2 cm ͑geometrical area͒ Pt
RDE. 10 M (COOH) in 0.5 M H SO ͑छ͒ and 5 mM K Fe(CN) in 1 M
KCl ͑᭝͒ solutions were used. The solid lines represent the best fit to the
Ϫ2
͑
COOH͒2 is oxidized to CO2 according to the following net
2
2
4
4
6
1
2
reaction
experimental data.
ϩ
Ϫ
͑
COOH͒ → 2CO ϩ 2H ϩ 2e
͓1͔
2
2
Johnson et al.12 suggested three possible, detailed reaction pathways
rotations per minute ͑rpm͔͒, the ip,a values are independent of the
rotation rate. The ip,a values measured at higher rpm (у2000 rpm)
for the oxidation reaction of ͑COOH͒ using Pt electrodes. All three
2
reaction pathways involve adsorptive interaction of the oxalic acid
with the Pt surface ͑according to a Temkin isotherm͒ thus empha-
sizing the importance of the electrode state/nature on the ͑COOH͒2
oxidation reaction. It should also be noted that the electrochemical
Ϫ4
Ϫ2
were found to vary by Ϯ2 • 10 A cm , thus accounting for the
scatter in the experimental i data. At lower (р1000) rpm, a de-
p,a
1/2
crease of the ip,a value with decreasing ␻ is observed, indicating
that elements of mass-transport control are present at these lower
rotation speeds. The result indicates that by operating at rotation
rates larger than 2000 rpm, it is possible to study the ͑COOH͒2
oxidation reaction under activation-controlled conditions.
͑
COOH͒2 oxidation reaction is fundamentally different from the
CH OH and CO oxidation reaction. In the case of ͑COOH͒ the
3
2
rate-determining step is the breaking of the carbon-carbon bond and
the ͑COOH͒2 molecule itself has a sufficient number of oxygen
It is well known that the mass-transport limited current (i ) is
L
atoms to be electrochemically converted to CO . Therefore, addi-
1/2
14
2
proportional to ␻ , according to the Levich equation, as follows
iL ϭ B␻1/2 ϭ 0.20nFAD2/3␯Ϫ1/6␻1/2co
͓2͔
tional oxygen, supplied from, e.g., adsorbed OH species, is not
needed for the ͑COOH͒ oxidation reaction, as is the case for the
2
CH OH and CO oxidation reaction to CO .
3
2
In Eq. 2, n is the number of electrons ͓2 in the case of the
RDE (COOH)2 oxidation studies using thin-layer Pt powder
electrodes.—Previously conducted outsweep rate studies and limit-
ing current calculations using unstirred solutions showed that the
͑
A
(
␻
COOH͒ oxidation reaction studied here͔, F is Faraday’s constant,
2
the
surface
area,
D
the
diffusion
coefficient
Ϫ6
2
Ϫ1 13
2
Ϫ1 15
5 • 10 cm s
͒, ␯ the kinematic viscosity (0.01 cm s
͒,
electrochemical ͑COOH͒ oxidation reaction observed using Pt cata-
2
o
the rotation rate in rpm, and c the concentration ͑in this case:
lyst electrodes takes place under mass-transport controlled
Ϫ5
Ϫ3
1/2
1
3
10 mol cm ). B is the slope of a plot of iL vs. ␻ and is defined
as shown in Eq. 2. Therefore, for a 0.2 cm electrode, a theoretical
conditions. However, the ͑COOH͒2 oxidation reaction needs to
take place under activation-controlled conditions in order to relate
2
Ϫ5
Ϫ1/2
slope of 4.8 • 10 A rpm
is calculated using this relationship,
the ͑COOH͒ oxidation current to the real Pt area. Hence, in this
2
Ϫ6
2
Ϫ1
and a D value of 5 • 10 cm s . This value is of the same order
work, the ͑COOH͒ electrolyte solutions were rapidly stirred em-
2
of magnitude as the experimentally obtained slope value of
ploying a rotating disk electrode arrangement to ensure that the
Ϫ5
Ϫ1/2
3
• 10 A rpm
for the ͑COOH͒2 oxidation data observed at
͑
COOH͒ oxidation reaction takes place under activation-controlled
2
lower (р1000) rpm, although only two data points are used in this
conditions. Furthermore, the powder electrodes were prepared as
thin layers of small total surface areas to allow all Pt catalyst sites to
be involved in the activation-controlled ͑COOH͒2 oxidation reac-
tion, as discussed in the following sections. The total Pt surface area
4
6
Ϫ
experiment. The i dependence for the Fe͑CN͒ oxidation reaction
L
1
/2
͑
⌬͒ on ␻ is also shown in Fig. 2. An experimental slope of
Ϫ5
Ϫ1/2
1.2 • 10 A rpm
is found for this one-electron reaction using 5
2
was generally less than 0.2 cm for studies involving positive poten-
mM K4Fe͑CN͒6 in 1 M KCl solution. The experimental slope for the
tials, and hence high anodic currents, while a higher total surface
area could be employed for studies carried out at lower potentials. In
all cases, the experimentally observed currents were compared to the
Fe oxidation reaction is very close to the theoretical slope of
Ϫ5
Ϫ1/2
1.4 • 10 A rpm
,
calculated using
a
D
value of
Ϫ6
2
Ϫ1 16
6.5 • 10 cm s
.
Ϫ1
theoretical limiting current to ensure that the studied ͑COOH͒ oxi-
Figure 3 shows a typical i-E curve recorded at 20 mV s in a
2
10Ϫ2 M ͑COOH͒ ϩ 0.5 M H SO solution using a polycrystalline
dation reactions took place under activation controlled conditions.
2
2
4
2
Figure 2 shows a plot for ͑COOH͒ oxidation current, i , ex-
(0.2 cm , geometrical area͒ Pt RDE rotated at 2000 and 5000 rpm.
It is seen that the i-V curves are qualitatively and quantatively es-
sentially the same ͑within an experimental current density variation
2
p,a
tracted from raw i-E curves at 1.1 V as a function of the square root
of the rotation rate (␻1 ) using the polycrystalline (0.2 cm , geo-
/2
2
Ϫ4
Ϫ2
metrical area͒ Pt RDE. ͑The i-E curves were recorded at
of Ϯ2 • 10 A cm estimated at the peak maximum͒, i.e., inde-
Ϫ1
2
0 mV s .) It is seen that at sufficiently high rotation rates ͓у2000
pendent of the rotation rate, as typically observed for a faradaic

E380
Journal of The Electrochemical Society, 150 ͑8͒ E377-E383 ͑2003͒
Table II. „COOH…2 oxidation currents and RAPt factors for a
range of Pt electrodes.
ip,a (COOH) a
b
b
APt
RAPt(ϭip,a /APt ͒
2
2
Ϫ2
͑
A͒
(cm )
(A cm )
Ϫ3
Ϫ3
Ϫ2
Ϫ2
Ϫ2
Ϫ2
Ϫ2
Pt RDEs
1.65 Ϯ 0.2 • 10
0.16 Ϯ 0.01
1 Ϯ 0.05 • 10
1
2
3
4
st Pt powder 1.02 Ϯ 0.1 • 10 0.104 Ϯ 0.005 0.98 Ϯ 0.05 • 10
Ϫ4
nd Pt powder 1.8 Ϯ 0.2 • 10 0.018 Ϯ 0.0001
1 Ϯ 0.05 • 10
Ϫ3
rd Pt powder 1.65 Ϯ 0.1 • 10 0.173 Ϯ 0.009 0.95 Ϯ 0.05 • 10
Ϫ4
th Pt powder 7.9 Ϯ 0.8 • 10 0.073 Ϯ 0.004 1.09 Ϯ 0.05 • 10
a
Steady-state ip,a values extracted at 1.1 V vs. SCE from raw iϪV
Ϫ1
Ϫ2
curves recorded at larger than 3000 rpm and 20 mV s in 10
M
(
COOH)2 ϩ 0.5 M H SO4 solutions. Experiments were carried out
2
at different rpm confirming that the ip,a value is independent of the
rotation rate.
b
Pt surface area (A ) estimated from CO stripping voltammetry
Pt
ads
Ϫ2
3
using a charge-to-area conversion factor of 420 ␮C cm
.
Figure 3. Typical CVs recorded at 20 mV sϪ1 in 10Ϫ2 M (COOH)2
In Eq. 3, n is the number of electrons ͑i.e., two for the ͑COOH͒ to
2
2
ϩ 0.5 M H SO solution using a polycrystalline (0.2 cm , geometrical area͒
2
4
CO oxidation reaction͒, F is Faraday’s constant, k is the standard
2
o
Pt RDE at 2000 and 5000 rpm. The 7th steady-state cycles are shown.
heterogeneous rate constant for the ͑COOH͒ oxidation reaction, ␣a
2
is the transfer coefficient, R is the gas constant, T is the temperature,
o
and E is the standard potential. Equation 3 shows that the
reaction that is activation-controlled. Consistent with this view of an
activation controlled mechanism, the experimentally observed peak
activation-controlled oxidation current, i , is proportional to the
electrode area (A), and hence, this equation can be used to relate the
a
currents are also smaller than the iL currents predicted for this oxi-
ratio of the experimentally observed i values to the real Pt area,
a
Ϫ3
dation reaction, as a i value of 2.2 • 10 A is predicted by Eq. 2
A . This ratio, R , is defined as follows
L
Pt
APt
2
for 2000 rpm and a 0.2 cm electrode. Additional support for an
ia
o
activation as opposed to a mass-transport controlled reaction was
given from sweep rate studies. i-E curves recorded for the same
␣ nF/RT(EϪE )
RAPt
ϭ
ϭ nFkoc͑COOH͒2 exp a
͓4͔
APt
electrode and conditions, but employing different sweep
Ϫ1
(
10-30 mV s ) yielded the same i-V curves. However, the shape
Equation 4 predicts the RAPt ratio to be constant for a particular
of the experimentally observed i-V curves ͑Fig. 3͒ is different than
typically observed for a classical activation-controlled reaction. For
a classical electron transfer reaction taking place under activation-
controlled conditions, an increase in current with increasing poten-
tial in an exponential manner is predicted15 unlike the decrease in
current, and hence current maximum, observed here for the
͑COOH͒ concentration and potential, even if the Pt electrodes have
2
different areas. This is shown in Table II, which summarizes the
experimentally observed ip,a and APt values for Pt RDE and Pt pow-
der electrodes ͑i.e., all electrodes contain only Pt and no Ru͒. The
ratio of the ip,a to APt values, i.e., the value of the RAPt factor at 1.1
V, is also given in Table II. The APt values were estimated from the
͑
COOH͒ oxidation reaction. Nevertheless, for the ͑COOH͒ oxida-
2
2
COads charge using a charge-to-area conversion factor of
tion reaction studied here a current maximum ͑and not a limiting
current͒ at ca. 1.1 V ͑vs. SCE͒ is observed. This experimentally
Ϫ2 3
Ϫ2
A cmϪ2
4
20 ␮C cm . Average RAPt values of 1 (Ϯ0.05) • 10
are found for the Pt powder and polycrystalline Pt electrodes tested.
observed current maximum and decrease in the ͑COOH͒ oxidation
2
The RAPt ratios are essentially the same for the Pt powders and the
current is believed to be related to a continuous passivation of the Pt
surface taking place with increasing potentials.17 In previous work,
Pt RDEs indicating that the ͑COOH͒ oxidation reaction is not in-
2
fluenced by the particle size of the Pt powders studied here.
it has been reported that a steady decrease in ͑COOH͒ oxidation
2
The relationship of Pt area and COads charge. The conversion
factor (fAPt).—We introduce a conversion factor, fAPt , that relates
the electroactive Pt area to the experimentally obtained COads strip-
ping charge, as follows
current takes place at potentials more positive than 1.3 V vs. revers-
ible hydrogen electrode, ͑i.e., ca. 1.05 V vs. SCE͒ due to passivation
of the Pt surface.17 This passivation is believed to be linked to the
formation of higher Pt oxide films. Consistent with this view, it has
also been observed, that a preformed, thick Pt oxide film displays
QCOads
little or no activity to oxidize ͑COOH͒ in the more positive poten-
2
APt
ϭ
fAPt
͓5͔
18
420 ␮C cmϪ2
tial region where O is evolved.
2
It is noteworthy that rotation rates of at least 3000 rpm were used
for the ͑COOH͒2 oxidation experiments carried out in the subse-
quent parts of this work. The influence of the rotation rate was also
tested in each ͑COOH͒2 oxidation experiments to ensure that the
reaction takes place under activation-controlled conditions.
In this work, the fAPt factor is taken as unity for Pt ͑Ru-free͒
electrodes. However, it should be noted that only close to one mono-
3
layer of CO ͑ca. 80-90% ͒ is adsorbed on Pt surfaces, and hence, Pt
areas estimated using the COads stripping charge, i.e., Eq. 5 and an
fAPt factor of unity, are somewhat lower than the true Pt areas.
Despite this fact, a fAPt factor of unity is used here, as this error is
eliminated in the Pt area estimation using the ͑COOH͒2 method
introduced in this work.
As discussed above, the fAPt factor for the Pt electrodes is ex-
pected to be unity, while the fAPt value for Pt-Ru electrodes is likely
to be less than unity and expected to be influenced by the electrode
composition. Metallic Ru contributes to the COads charge, however,
Conversion of (COOH)2 oxidation currents to Pt surface
area.—The following relationship between the oxidation current
(
i ) and potential, E, applies for an activation-controlled oxidation
a
15
reaction at sufficiently high overpotentials
o
␣
nF/RT(EϪE )
a
i ϭ nFAk c exp
o ͑COOH͒
2
͓3͔
a

Journal of The Electrochemical Society, 150 ͑8͒ E377-E383 ͑2003͒
E381
it is inactive toward the ͑COOH͒2 oxidation reaction ͑as shown
above͒, thus reducing the value of the fAPt conversion factor. With
increasing surface fraction of metallic Ru vs. Pt, the contribution of
the Ru sites to the COads charge increases, and hence the fAPt value
is predicted to decrease. In fact, the value of the fAPt factor for a
particular catalyst yields the fraction of the COads molecules that
adsorb onto Pt sites, and vice versa, the value of 1-fAPt yielding the
fraction of COads molecules on the other catalyst components ͑i.e.,
Ru in the present study͒. It is clear that the fAPt value needs to be
estimated for each individual powder. The fAPt value for Pt-Ru pow-
ders can be obtained from the experimentally determined Pt area
values estimated from ͑COOH͒ oxidation experiments, e.g., using
2
Ϫ2
Ϫ2
the RAPt value of 1.00 Ϯ 0.05 • 10 A cm estimated at ip,a , i.e.,
Ϫ2
1
.1 V for 10 M ͑COOH͒ in 0.5 M H SO solutions, and com-
2 2 4
bining Eq. 4 and 5 as follows
ia
RAPt
APt
fAPt
ϭ
ϭ
͓6͔
QCOads
QCOads
Figure 4. Typical J-V curves for Pt powder 1 ͑thick line͒ and Pt/RuO2
͑thermal͒ powder 4 ͑thin line͒ electrodes recorded at 8000 rpm and
20 ␮C cmϪ2
420 ␮C cmϪ2
4
Ϫ1
Ϫ2
20 mV s
in 10 M (COOH)2 ϩ 0.5 M H SO4 solutions. The currents
2
were normalized using the experimentally observed COads charge and a fAPt
factor of unity and 0.55 for the Pt powder and PtRuO2 ͑thermal͒ powder 4
electrodes, respectively.
It should be noted that any ͑COOH͒ oxidation current can be
used for the Pt area estimation carried out in such a manner provided
2
that the RAPt factor is estimated for the corresponding potential.
Using additional ͑COOH͒ oxidation current data and R
found and estimated for a range of potentials, clearly increases the
accuracy of the method. Furthermore, it is possible to estimate the
values
2
APt
These fAPt factors listed in Table III can be used to calculate the
Pt area from the experimentally observed COads charge for any elec-
trode made up of the corresponding Pt-based catalyst and using Eq.
͑
COOH͒ oxidation currents at potentials less positive than 1.1 V.
2
5. Furthermore, the value of the fAPt factor indicates the fraction of
This is particularly important for catalysts that may corrode at such
positive potentials, as is often observed for catalysts containing Ru.
In this work, COads stripping voltamograms were also recorded be-
fore and after ͑COOH͒2 oxidation studies were carried out for a
particular catalyst. A decrease in COads charge would indicate a loss
of catalyst due to either catalyst detachment from the support and/or
partial dissolution of the catalyst, while changes in the shape of the
COads stripping peak ͑typically a shift of the oxidation wave to more
positive potentials͒ likely indicates the preferential dissolution of Ru
from the catalyst system. In this work, it was found that all Pt
powders 1, the Pt/RuO2 ͑thermal͒ powder 4, Pt/Ru (H2 reduced͒
CO that is adsorbed on Pt sites, and hence this factor also yields
information about the fraction of CO adsorbed on the Ru sites of the
catalysts studied here. This is clearly seen for the PtRu alloy catalyst
powders 3 and 3a. The fAPt factor of 0.7 found for the PtRu powder
3
suggests that 70% of the adsorbed CO molecules adsorb on Pt
powder 5, and the PtRuO ͑ballmilled͒ powder 6 electrodes did not
2
dissolve or detach from the glassy carbon support, as a result of the
͑
COOH͒ oxidation studies carried out involving potentials as high
2
as 1.2 V. However, it was necessary to restrict the potential range of
the other PtRu alloy electrodes powders 3 and 3a, to potentials less
negative than 0.8 V to avoid changes in the catalyst area as well as
dissolution of the Ru component.
It is noteworthy that the shapes of the experimentally observed
current density J-V curves for the individual Pt and Pt-Ru powders
were found to be essentially the same for all powders investigated in
this work. This is shown in Fig. 4 for Pt powder 1 and Pt/RuO2
͑
thermal͒ powder 4 electrodes. The experimentally observed cur-
rents were normalized using the COads charges and fAPt factors of
unity and 0.55 for the Pt powder 1 and the Pt/RuO ͑thermal͒ pow-
2
der 4, respectively. Figure 5 shows J-V curves for Pt powder 1 and
PtRu alloy powder 3 electrodes for the limited potential range of up
to 0.8 V. The current scales were normalized using a fAPt factor of
unity and 0.7 for the Pt powder 1 and the PtRu alloy powder 3,
respectively. The limited E range was selected to avoid preferential
dissolution of Ru from the catalyst surface, as discussed above. It is
seen that the shape of the J-V curves for the Pt and the PtRu alloy
powders are essentially the same within the E range of ca. 0.75 to
Figure 5. Typical J-V curves for Pt powder 1 ͑thick line͒ and PtRu alloy
͑
8
Pt:Ru atomic ratio of 0.7:0.3͒ powder 3 ͑thin lines͒ electrodes recorded at
Ϫ1
Ϫ2
000 rpm and 20 mV s in 10 M (COOH)2 ϩ 0.5 M H SO solutions.
2 4
The currents were normalized using the experimentally observed COads
charge and a fAPt factor of unity and 0.7 for the Pt powder and PtRu alloy
powder 3 electrodes, respectively. The J-V curve for the Pt powder elec-
trodes is the average of experimental data collected for three individual Pt
powder electrodes, while the actual raw data for three individual PtRu alloy
electrodes are shown in Fig. 6.
0.8 V, i.e., within the E range used in this work to estimate the fAPt
factors for the PtRu alloy powder electrodes. The meaning of the
value of these fAPt factors is discussed in more detail in the follow-
ing paragraph.

E382
Journal of The Electrochemical Society, 150 ͑8͒ E377-E383 ͑2003͒
Table III. fAPt values and XPS data for a range of Pt and Pt-Ru
powder electrodes.
fAPta
Pt ͑atom %b͒
Pt-based powder type
Pt ͑1͒
PtRu alloy ͑3͒
PtRu alloy ͑3a͒
PtRuO ͑thermal͒ ͑4͒
PtRu (H reduced͒ ͑5͒
1 Ϯ 0.05
0.7 Ϯ 0.05
0.54 Ϯ 0.03
0.55 Ϯ 0.03
0.45 Ϯ 0.02
0.5 Ϯ 0.03
1
0.7
0.55
0.36
0.44
0.54
2
2
PtRuO ͑ballmilled͒ ͑6͒
2
a
The fAPt factors are estimated from (COOH)2 oxidation studies and
COads stripping voltammetry, as described in the text.
b
Atomic fraction of Pt ͑atomic Pt concentration divided by the total
atomic concentration of Pt and Ru͒ obtained from XPS data, as dis-
1
0
cussed in detail elsewhere. The atomic fraction of Pt data indicated
in the third column for the powders 4, 5, and 6 are thought to be
lower than the actual values due to possible inhomogeneities of the Pt
and Ru phase within the catalyst particles that could complicate the
XPS analyses. Relative errors in the atom % Pt values are ca. 2%.10
sites, while 30% are adsorbed on other sites, i.e., Ru. Similarly, the
fAPt factor of 0.54 estimated for the PtRu alloy powder 3a suggests
that 54% of the adsorbed CO molecules are adsorbed on Pt sites,
while 46% are adsorbed on Ru sites. This, and the fact that the
surface atomic fraction of Pt to Ru for these two PtRu alloy powders
3
and 3a is 0.3:0.7 and 0.54:0.46, respectively, suggests that one CO
molecule adsorbs per Pt as well as Ru site. These results further
suggest that the Ru sites are likely present in their metallic state at
the potential of Ϫ0.1 V where CO is adsorbed, as CO adsorption
onto the oxide covered Ru surface is generally not observed.
The fAPt factors for the PtRuO ͑thermal͒ powder 4, PtRu (H
2
2
reduced͒ powder 5, and PtRuO ballmilled powder 6 are less than
2
unity indicating that CO also adsorbs onto Ru sites. In fact, 45%
͑
powder 4͒, 55% ͑powder 5͒, and 50% ͑powder 6͒ of the COads
molecules adsorb onto Ru sites, thus further suggesting that a sig-
nificant fraction of the Ru sites of these powders are in the reduced
Ru metal state.
Current-normalized background CVs (in 0.5 M H SO ) for
2
4
different powders.—In this section, current-normalized background
CV characteristics for the Pt and Pt/Ru powders are discussed. The
CVs were recorded in 0.5 M H SO and the currents were normal-
2
4
ized for the Pt area using the fAPt factors shown in Table III. The
powder electrodes were prepared by sonicating 13 mg of the powder
Figure 6. Typical background CVs for Pt and Pt-Ru powder electrodes in
Ϫ1
0
.5 M H SO . The potential was cycled at 10 mV s between Ϫ0.22 and
2 4
in 1 mL H O and 0.3 mL Nafion solution, as discussed in the Ex-
2
1.1 V for the Pt powder electrode, while an upper cycling limit of 0.55 V was
perimental section. Figures 6a and b show the normalized CV char-
employed for the Pt-Ru powder electrodes. The current scale is normalized
for the Pt area calculated using the fAPt factors listed in Table III. ͑a͒ shows
the CVs for Pt powder ͑no. 1͒, PtRu alloy ͑no. 3͒, and Pt/RuO2 ͑thermal͒
͑powder no. 4͒ electrodes, while ͑b͒ shows the same for Pt powder ͑no. 1͒,
Pt/Ru (H2 reduced͒ ͑no. 5͒ and Pt/RuO2 ͑ballmilled͒ ͑powder no. 6͒ elec-
trodes.
acteristics for the Pt and Pt-Ru powder electrodes recorded at
Ϫ1
1
0 mV s , i.e., at a sweep rate sufficiently low to allow time for all
catalyst sites to be electrochemically converted. Figures 6a and b
show that the characteristics CV shape for polycrystalline Pt is ob-
served for the Pt powder electrode, while for all Pt/Ru powders a
redox reaction ascribed to the oxidation/reduction of the Ru sites is
observed within the potential of ca. 0.05 V and an arbitrarily se-
lected upper cycling limit of 0.55 V. This indicates that all Pt-Ru
powders studied in this work contain Ru sites that display oxidation/
reduction behavior within a similar potential range, in fact, within a
potential range of interest for practical fuel cell applications. The
shape and magnitude of the normalized currents are different for the
individual powders. The latter is not unexpected, as in fact, the Ru
surface concentration ͑i.e., the amount of Ru on the surface per Pt
surface area͒ is different for all powders studied here, as shown in
Table III. It is also noteworthy that, between 0.05 and 0.5 V, the
The charge observed between 0.05 and 0.55 V was calculated for
the Pt-Ru powders. For this calculation, the current observed for the
Pt powder electrode was used as the baseline and subtracted from
the monitored currents. Table IV shows the oxidation charges nor-
malized for the Pt area ͓QO͑APt͔͒ obtained in this manner for differ-
ent Pt-Ru powder electrodes. It should be noted that the normalized
charge value for the Pt/Ru (H2 reduced͒ powder electrode is very
7
close to the value observed in previous work, thus further support-
shape of the CV curve for the Pt/Ru (H -reduced͒ powder 5 is
ing the Pt surface-area estimation introduced in this work. The num-
2
Ϫ
Ϫ
essentially the same as that found for bare Ru metal cycled under the
ber of electrons per Pt and Ru atom ͓e
and e
, respec-
Pt͑atom͒
Ru͑atom͒
7
same experimental conditions, indicating that the Ru component of
tively͔ are also shown in Table IV. The numbers of electrons per Pt
this particular powder behaves like bare Ru metal.
and Ru were calculated as shown in Eq. 7 and 8, respectively

Journal of The Electrochemical Society, 150 ͑8͒ E377-E383 ͑2003͒
E383
adsorption in this work. Furthermore, valuable information about the
electrochemistry of the Ru component of Pt-Ru catalysts can be
gained from background CV characteristics that are normalized us-
Table IV. Pt and Ru area normalized oxidation charge values for
different PtÕRu powder electrodes.
Qox/APta (C cmϪ2
)
Ϫ
b
Ϫ
c
ing the Pt area estimated using the ͑COOH͒ oxidation method.
e
e
2
Pt(surface-atom)
Ru(surface-atom)
However, care must be taken to apply this method for the esti-
mation of the electroactive Pt area and possible extraction of the
fraction of CO molecules that adsorb on the individual catalyst sites.
The catalysts must be stable within the potential range used to mea-
Ϫ4
PtRu alloy ͑3͒
1 " 10
0.4
1
1.9
1.1
0.9
1.2
1.5
1.1
Ϫ4
PtRuO ͑thermal͒ ͑4͒
2.6 " 10
5.3 " 10
2
Ϫ4
PtRu (H reduced͒ ͑5͒
2
Ϫ4
PtRuO ͑ballmilled͒ ͑6͒
2.8 " 10
2
sure the ͑COOH͒ oxidation current. Furthermore, a thin layer of
2
a
b
c
catalyst particles must be applied to the electronically conductive
Experimentally observed oxidation charge per Pt area (Qox/APt), cal-
support, and the ͑COOH͒ oxidation reaction must be carried out
culated as described in the text.
Number of electrons per Pt atom (e
2
Ϫ
under controlled and rapid stirring conditions. It is also noteworthy
that activation-controlled currents are more susceptible to impurities
and electrode poisoning effects than mass-transport limited currents.
This is due to the fact that for mass-transport controlled conditions,
the electrode area collapses to the geometrical area, while the real
electroactive area determines the activation-controlled currents.
), calculated using Eq.
), calculated using
Pt(surface-atom)
7
.
Ϫ
Number of electrons per Ru atom (e
Eq. 8.
Ru(surface-atom)
QO͑APt͒APt͑atom͒
Ϫ
Acknowledgments
e
ϭ
͓7͔
͓8͔
Pt͑atom͒
Qe
Ϫ
The authors thank P. L’Abbe ͑National Research Council
Canada͒ for making the electrochemical cells used in this work and
G. Pleizier ͑National Research Council Canada͒ for the XPS ana-
lyzes. Helpful discussions with Dr. E. Gileadi ͑Tel-Aviv University,
Isreal͒ are also greatly appreciated.
Ϫ
Qe
fAPt
Pt͑atom͒Ϫ
Ϫ
e
ϭ
Ru͑atom͒
1
Ϫ fAPt
In Eq. 7, the area of a Pt atom (APt͑atom͒) is taken as
The National Research Council Canada assisted in meeting the publica-
tion costs of this article.
Ϫ16
Ϫ19
2 19
6
1
.1 • 10
cm
and the charge of an electron (QeϪ
) as
20
.6 • 10
C. Of particular interest is the comparison of the val-
References
ues of the number of electrons per Ru atom for the different Pt-Ru
powders. The data shown in Table IV suggest that almost one elec-
tron per Ru atom is passed within the potential range investigated
here ͑i.e., up to 0.55 V͒ for the PtRu alloy powder 3, while a larger
number of electrons is suggested to be involved for the other three
powders. For powders 4 and 6 ͑i.e., the two powders containing the
1. J. O’M. Bockris and H. Wroblowa, J. Electroanal. Chem., 7, 428 ͑1964͒.
2. M. Watanabe and S. Motoo, J. Electroanal. Chem., 60, 267 ͑1975͒.
3
. H. Gasteiger, N. Markovic, P. N. Ross, and E. Cairns, J. Electrochem. Soc., 141,
795 ͑1994͒.
1
4
. H. N. Dinh, X. Ren, F. H. Garzon, P. Zelenay, and S. Gottesfeld, J. Electroanal.
Chem., 491, 222 ͑2000͒.
5
6
. K. Kinoshita and P. N. Ross, J. Electroanal. Chem., 78, 313 ͑1977͒.
. K. A. Friedrich, F. Henglein, U. Stimming, and W. Unkauf, Electrochim. Acta, 45,
RuO and RuO phase͒ the number of electrons per Ru atom are very
2
similar, namely 1.2 and 1.1, respectively. This is possibly consistent
with the fact that the Ru phases of these two powders are essentially
3
283 ͑2000͒.
7
. S. Hadzi-Jordanov, H. Angerstein-Kozlowska, and B. E. Conway, J. Electroanal.
Chem., 60, 359 ͑1975͒.
1
0
the same. These results suggest that the normalized-CV approach
obtained using the fAPt factors can yield useful information about
the Ru electrochemistry of Pt-Ru electrocatalysts, which is likely
relevant to the development and understanding of fuel cell catalyst
systems. It should be noted that influences of double-layer charging
have not been accounted for in the above calculation of the number
of electrons per Pt and Ru atom.
8
9
. C. L. Green and A. Kucernak, J. Phys. Chem. B, 106, 1036 ͑2002͒.
. S. A. S. Machado, A. A. Tanaka, and E. R. Gonzalez, Electrochim. Acta, 36, 1325
͑
1991͒.
10. C. Bock, M.-A. Blackely, and B. MacDougall, In preparation.
11. T. J. Schmidt, H. A. Gasteiger, and R. J. Behm, J. Electrochem. Soc., 146, 1296
͑
1999͒.
2. J. W. Johnson, H. Wroblowa, and J. O’M. Bockris, Electrochim. Acta, 9, 639
1964͒.
1
͑
1
1
3. C. Bock, A. Smith, and B. MacDougall, Electrochim. Acta, 48, 57 ͑2002͒.
4. E. Gileadi, Electrode Kinetics for Chemists, Chemical Engineers, and Material
Scientists, VCH Publishers, Inc., New York ͑1993͒.
5. A. J. Bard and L. R. Faulkner, Electrochemical Methods, Fundamentals and Ap-
plications, John Wiley & Sons, Inc., New York ͑1980͒.
Conclusions
It has been shown that the activation-controlled ͑COOH͒ oxida-
tion current can be related to the electroactive Pt area for Pt-based
2
1
catalysts. The combination of the COads stripping voltammetry and
16. L. Lipp and D. Pletcher, Electrochim. Acta, 42, 1091 ͑1997͒.
1
1
7. J. Giner, Electrochim. Acta, 4, 42 ͑1961͒.
the activation-controlled ͑COOH͒ current was applied to estimate
2
8. C. Bock, A. Smith, and B. MacDougall, in Energy and Electrochemical Processes
for a Cleaner Environment, C. Comninellis, M. Dogle, and J. Winnick, Editors, PV
2001-23, p. 23, The Electrochemical Society Proceeding Series, Pennington, NJ
͑2001͒.
the Pt surface area for a range of Pt-Ru powder electrodes. It was
also found that the combination of these two methods can yield in
situ information about the fraction of CO molecules that adsorb on
Pt vs. the ad-metal ͑i.e., Ru in this work͒ catalyst sites. For the case
of PtRu alloy electrodes, it was found that one mole of CO adsorbs
per Pt as well as Ru sites at Ϫ0.1 V, i.e., the potential used for CO
1
2
9. T. D. Jarvi and E. M. Stuve, in Electrocatalysis, J. Lipkoswki and P. R. Ross,
Editors, Wiley-VCH, Inc. New York ͑1998͒.
0. Handbook of Chemistry and Physics, 71st ed., D. R. Lide, Editor, CRC Press, Inc.,
Boston, MA ͑1990-1991͒.