Size-selected synthesis of PtRu nano-catalysts: reaction and size control mechanism.

Abstract

A rapid synthesis method for the preparation of PtRu colloids and their subsequent deposition on high surface area carbons is presented. The reaction mechanism is shown to involve the oxidation of the solvent, ethylene glycol, to mainly glycolic acid or, depending on the pH, its anion, glycolate, while the Pt(+IV) and Ru(+III) precursor salts are reduced. Glycolate acts as a stabilizer for the PtRu colloids and the glycolate concentration, and hence the size of the resulting noble metal colloids is controlled via the pH of the synthesis solution. Carbon-supported PtRu catalysts of controlled size can be prepared within the range of 0.7-4 nm. Slow scan X-ray diffraction and high-resolution transmission electron microscopy show the PtRu catalysts to be crystalline. The Ru is partly dissolved in the face-centered cubic Pt lattice, but the catalysts also consist of a separate, hexagonal Ru phase. The PtRu catalysts appear to be of the same composition independent of the catalyst size in the range of 1.2-4 nm. Particular PtRu catalysts prepared in this work display enhanced activities for the CH(3)OH electro-oxidation reaction when compared to two commercial catalysts.

Full text

Page 1

NRC Publications Archive (NPArC)
Archives des publications du CNRC (NPArC)
Web page / page Web
http://dx.doi.org/10.1021/ja0495819
http://nparc.cisti-icist.nrc-cnrc.gc.ca/npsi/ctrl?action=rtdoc&an=5113752&lang=en
http://nparc.cisti-icist.nrc-cnrc.gc.ca/npsi/ctrl?action=rtdoc&an=5113752&lang=fr
Size-selected synthesis of PtRu nano-catalysts: reaction and size
control mechanism
Bock, Christina; Paquet, Chantal; Coullard, Martin; Botton, Gianluigi A.;
MacDougall, Barry R.
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
http://nparc.cisti-icist.nrc-cnrc.gc.ca/npsi/jsp/nparc_cp.jsp?lang=en
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE.
Journal of the American Chemical Society, 126, 25, pp. 8028-8037, 2004
Publisher’s version / la version de l'éditeur:
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site
http://nparc.cisti-icist.nrc-cnrc.gc.ca/npsi/jsp/nparc_cp.jsp?lang=fr
LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
Contact us / Contactez nous: nparc.cisti@nrc-cnrc.gc.ca.

Page 2

Size-Selected Synthesis of PtRu Nano-Catalysts: Reaction
and Size Control Mechanism
Christina Bock,*,†Chantal Paquet,†Martin Couillard,‡Gianluigi A. Botton,‡and
Barry R. MacDougall†
Contribution from the National Research Council of Canada, 1200 Montreal Road,
Ottawa, Ontario, Canada, K1A 0R6, and Brockhouse Institute for Materials Research,
McMaster UniVersity, 1280 Main Street West, Hamilton, Ontario, Canada, L8S 4L8
Received January 23, 2004; E-mail: christina.bock@nrc-cnrc.gc.ca
Abstract: A rapid synthesis method for the preparation of PtRu colloids and their subsequent deposition
on high surface area carbons is presented. The reaction mechanism is shown to involve the oxidation of
the solvent, ethylene glycol, to mainly glycolic acid or, depending on the pH, its anion, glycolate, while the
Pt(+IV) and Ru(+III) precursor salts are reduced. Glycolate acts as a stabilizer for the PtRu colloids and
the glycolate concentration, and hence the size of the resulting noble metal colloids is controlled via the
pH of the synthesis solution. Carbon-supported PtRu catalysts of controlled size can be prepared within
the range of 0.7-4 nm. Slow scan X-ray diffraction and high-resolution transmission electron microscopy
show the PtRu catalysts to be crystalline. The Ru is partly dissolved in the face-centered cubic Pt lattice,
but the catalysts also consist of a separate, hexagonal Ru phase. The PtRu catalysts appear to be of the
same composition independent of the catalyst size in the range of 1.2-4 nm. Particular PtRu catalysts
prepared in this work display enhanced activities for the CH3OH electro-oxidation reaction when compared
to two commercial catalysts.
Introduction
Pt- and PtRu-based nanoparticles are of major interest as
anode catalysts for direct methanol and reformate fuel cells.1,2
The latter utilizes H2as anode fuel that contains CO resulting
typically from re-formation of CH3OH. Up to now, bimetallic
PtRu anode catalysts have shown superior activities for these
fuel cells. The oxidation of CH3OH to CO2at “low” potentials
takes place via a bifunctional mechanism that involves the
abstraction reaction of hydrogen through the adsorption of CH3-
OH on Pt, forming a CO-type intermediate species at the catalyst
surface. The complete oxidation of the adsorbed reaction
intermediate to CO2is promoted by -OH-type groups that are
formed via the partial oxidation of H2O on neighboring Ru sites.2
A similar bifunctional mechanism applies to the H2oxidation
reaction in the presence of CO. On the basis of this bifunctional
mechanism, it is clear that the catalytic performance is strongly
dependent on the distribution of Pt and Ru sites on the atomic
level, and it is believed that alloys are most beneficial in this
regard.1-5However, for the successful implementation of direct
methanol fuel cells, in particular, the catalytic performance and
the amount of noble metal catalysts used need to be reduced
significantly. It is well known that the catalyst utilization can
be improved by supporting Pt and PtRu particles in the nanosize
range on high surface area carbons.6Various methods are used
for the preparation of supported catalysts, such as the impregna-
tion of the support with the noble metal precursor salts that are
subsequently reduced, typically in a hydrogen atmosphere at
elevated temperatures.7
An attractive alternative route to preparing supported Pt and
PtRu catalysts is the synthesis of colloidal catalyst solutions
that is followed by the deposition of the nanosized catalysts on
a suitable support material such as high surface area carbon
blacks.8To stabilize the colloids in solution, organic stabilizers
such as polyvinyl pyrolidine that interact with the Pt and PtRu
catalyst surface sites have been used.9,10Once the catalysts are
deposited on the carbon black, the stabilizer molecules are
preferably removed, as they otherwise hinder the access of the
fuel to the catalyst sites. It has been shown that organic
stabilizers can be removed by oxidative heat treatment.11
However, heat treatment can alter the properties of these
catalysts. In the case of PtRu alloys, it is known that heat
treatments as low as 220 °C result in Pt and Ru phase
†National Research Council of Canada.
‡McMaster University.
(1) Bockris, J. O’M.; Wroblowa, H. J. Electroanal. Chem. 1964, 7, 428-451.
(2) Watanabe, M.; Motoo, S. J. Electroanal. Chem. 1975, 60, 267-273.
(3) Gasteiger, H. A.; Markovic, N.; Ross, P. N.; Cairns, E. J. J. Electrochem.
Soc. 1994, 147, 1795-1806.
(4) Iwasita, T.; Hoster, H.; John-Anacker, A.; Lin, W. F.; Vielstich, W.
Langmuir 2000, 16, 522-529.
(5) Bock, C.; MacDougall, B.; LePage, Y. J. Electrochem. Soc., in press.
(6) Wilson, M. S.; Gottesfeld, S. J. Appl. Electrochem. 1992, 1, 1-7.
(7) Richard, D.; Gallezot, P. In Preparation of Catalysis IV; Delmon, B.,
Grange, P., Jacobs, P. A., Poncelet, G., Eds.; Elsevier: Amsterdam, 1988;
pp 71-81.
(8) Schmidt, T. J.; Noeske, M.; Gasteiger, H. A.; Behm, R. J.; Britz, P.; Brijoux,
W.; Boennemann, H. Langmuir 1997, 13, 2591-2595.
(9) Dalmia, A.; Lineken, C. L.; Savinell, R. F. Colloid Interface Sci. 1998,
205, 535-537.
(10) Bonet, F.; Delmas, V.; Grugeon, S.; Urbina, R. H.; Silvert, P. Y.; Tekaia-
Elhsissen, K. Nanostruct. Mater. 1999, 11, 1277-1284.
(11) Dubeau, L.; Coutanceau, C.; Garnier, E.; Leger, J. M.; Lamy, C. J. Appl.
Electrochem. 2003, 33, 419-429.
Published on Web 06/03/2004
8028 9 J. AM. CHEM. SOC. 2004, 126, 8028-8037
10.1021/ja0495819 CCC: $27.50 © 2004 American Chemical Society

Page 3

separation.5This in turn alters the surface concentration and
distribution of Pt and Ru, typically resulting in lower catalyst
activities.
In this work, a simple chemical reduction route is discussed
that results in the formation of PtRu nanoparticles that can be
supported on substrates such as high surface area carbons. The
synthesis method can be used to prepare PtRu catalysts
selectively in the nanoscale particle size range of less than 4
nm without changing the composition of the PtRu catalysts. The
mechanism of the PtRu colloid formation reaction is discussed,
and PtRu nano-catalysts deposited on high surface area carbons
are characterized using X-ray diffraction, X-ray photon spec-
troscopy, and transmission electron microscopy (TEM). The
activities of carbon-supported PtRu particles for the electro-
chemical CH3OH oxidation reaction are also discussed and
compared to those of commercial PtRu catalysts.
Experimental Section
Preparation of Carbon-Supported PtRu Catalysts. The synthesis
of the bimetallic PtRu colloids was carried out in ethylene glycol
(Anachemia) solutions containing different concentrations of sodium
hydroxide (EM Science). First, 0.2326 g of PtCl4(Alfa Aesar, 99.9%
metals basis) and 0.1383 g of RuCl3(Alfa Aesar, 99.9% metals basis)
were dissolved in 50 mL of ethylene glycol containing between 0.2
and 0.04 M NaOH. The solutions were stirred for 30 min in air at
room temperature, subsequently heated under reflux to 160 °C for 3 h,
and then cooled in air. Temperatures of 160 °C were achieved within
less than 20 min. Dark brown solutions containing the PtRu colloids
were formed in this manner and are referred to as colloidal solutions
in this work. Appropriate aliquots of the colloidal solutions were mixed
with carbon blacks in a large and open beaker for up to 24 h, resulting
in the deposition of the PtRu colloids on the carbon substrates, Vulcan
XC-72R (Cabot). The carbon-supported PtRu catalysts were then filtered
and extensively washed with water. Initial, that is, undiluted, filtrate
solutions were analyzed for organic acids using high performance liquid
chromatography (HPLC). The carbon-supported PtRu catalysts were
generally dried at 160 °C in air for 1 h. A mortar was used to
homogeneously ground the carbon-supported catalyst powders. The
supported catalysts were stored in air at room temperature and were
found to have a shelf life of at least 1 year. Prior to the synthesis of
the colloidal solutions, the PtCl4and RuCl3were dried in an air oven
at 135 °C. The ethylene glycol was dried using molecular sieves type
3a (BDH), while the carbon black powder was used as received.
Instrumentations/Techniques. XPS spectra were obtained using a
PHI 5500 spectrometer equipped with a monochromatized Al KR
source. The data were collected using an aperture of 4, a 70° takeoff
angle, and a pass energy of 23.50 eV. The position of the C 1s peak,
that is, 284.6 eV, was used to correct the binding energies for all
catalysts for possible charging effects. The catalyst powders were
attached to conductive carbon tape for the XPS analyzes. For each
catalyst, a survey spectrum was collected before high-resolution spectra
were recorded. Deconvolutions of the XPS spectra were performed
using a Grams_32 Spectral Note base software program. A Philips CM
20 TEM was employed to measure the size of the PtRu catalysts. A
JEOL 2010F operated at 200 keV and equipped with an energy-
dispersive X-ray spectrometer (EDS) and a Gatan annular dark-field
(ADF) detector was also employed. For the TEM analyses, the carbon-
supported catalyst powders were ultrasonically suspended in ethanol,
and a drop of the catalyst powder suspension was applied to a holey
amorphous carbon film on a 300 mesh Cu grid (Marivac, Limited). A
Scintag XDS2000 system was employed using a Cu KR source to obtain
XRD spectra of the carbon-supported catalysts. The angle extended
from 20° to 80° and varied using a step size of 0.06°, accumulating
data for 60 s per step. Silicon powder (typically 1-20 µm, 99.9985%
purity, Alfa Aesar) that was homogeneously grounded with the carbon-
supported catalysts was used as an internal standard. XRD spectra of
the Si-free and Si-containing samples were obtained for all catalysts.
The software program Topas 2 (DIFFRACPLUSTopas, Bruker axs, Inc.)
was employed to extract lattice parameter constants from the experi-
mental XRD spectra. The entire XRD spectra were employed to analyze
the data. The XRD, XPS, and TEM characterizations were carried out
on the unused, that is, as-prepared, catalyst powders. All electrochemical
experiments were carried out using a Solarotron SI 1287 electrochemical
interface (Solarotron Group, Ltd.) driven by the Corrware software
program (Scribner, Assoc.). Undiluted filtrates of the colloidal synthesis
solutions were analyzed for carboxylic acids using a HP series 1100
HPLC. An ion-interaction column (Mandel), thermostated at 46 °C,
was used. The UV detector was set at 195 nm. 0.01 M H2SO4 was
used as the mobile phase at a flow rate of 0.3 mL min-1. A Fisher
Accumet pH meter, model 805 MP, and a combined glass pH electrode
(Corning) were used for pH measurements. A digital Mirak hotplate
was used for the synthesis of the colloids that allowed controlled stirring
and heating conditions.
Electrochemical Measurements. 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. The cells were equipped with a water jacket
and condensers for the electrochemical oxidation studies that were
carried out at 60 °C. A large surface area Pt gauze served as counter
electrode, and a saturated calomel electrode (SCE) was used as the
reference electrode. All potentials in this study are reported with respect
to the SCE. The catalyst powders were formed into electrodes by
sonicating 13 mg of the carbon-supported catalyst powders in 1 mL of
H2O and 300 µL of Nafion solution (5 wt %, Aldrich) for 15 min,
forming a catalyst ink. Subsequently, 10 µL of catalyst ink was applied
to form a thin layer of ca. 0.5 cm2geometrical area on a gold foil (0.5
× 2 × 0.1 cm3, Goodfellow, 99.95% metal basis). The catalyst layer
was then dried at 80 °C in an air oven for 30 min. Electrical contact
was made by firmly attaching an Au wire via a small hole to the Au
foil. Commercially available carbon-supported PtRu catalysts were also
tested for their CH3OH electro-oxidation activity. 20 wt % Pt and 10
wt % Ru catalysts supported on carbon black (Vulcan XC72-R) were
obtained from Alfa Aesar and E-Tek, Inc.
Solutions and Chemicals. All chemicals used in this work were
A.C.S. grade, and high resistivity 18 MΩ H2O was used. Methanol
oxidation studies were carried out in deoxygenated 0.5 M H2SO4 +
0.5 M CH3OH solutions.
Results and Discussion
Synthesis of Bimetallic PtRu Colloids in Ethylene Glycol
Solutions. The synthesis of monometallic noble metal colloids
in ethylene glycol solution has been suggested in previous
work.9,10Various monometallic noble metal colloids were
synthesized in ethylene glycol using organic stabilizers such as
poly(N-sulfonatopropyl p-benzamide)9and polyvinyl pyrolidine
(PVP).10The synthesis of monometallic colloids in ethylene
glycol solution of alkaline pH (>12) without the addition of a
stabilizer molecule has also been discussed.12It was proposed
that the noble metal precursor salts such as H2PtCl6are reduced
by nitrogen passed through the ethylene glycol solution and that
stabilizer-free, that is, “unprotected”, colloids were prepared.
The colloids were suggested to be solely stabilized by the
“dielectric” properties of the ethylene glycol solvent. In the
following sections, the size-selected synthesis of bimetallic PtRu
colloids in ethylene glycol solutions is discussed. The reaction
mechanism and the stabilization of the colloidal particles are
(12) Wang, Y.; Ren, J.; Deng, K.; Gui, L.; Tang, Y. Chem. Mater. 2000, 12,
1622-1627.
Size-Selected Synthesis of PtRu Nano-CatalystsA R T I C L E S
J. AM. CHEM. SOC. 9 VOL. 126, NO. 25, 2004 8029

Page 4

investigated in detail. It should be noted that ethylene glycol is
an alcohol and is readily oxidized. Therefore, ethylene glycol
is more likely to act as a reducing agent for the Pt and Ru
precursor salts rather than nitrogen that is generally utilized to
remove oxygen from solutions. Furthermore, oxidation products
resulting from the ethylene glycol oxidation reaction could
interact with the noble metal colloids and hence act as their
stabilizers.
Reaction Mechanism: Reduction of the Precursor Salts
Accompanied by the Oxidation of Ethylene Glycol. Ethylene
glycol is an alcohol and a very weak acid (pKa≈ 1513) that can
be oxidized to aldehydes, carboxylic acids, and CO2, as shown
in the following reaction scheme:
For this reaction pathway to take place, the -OH groups of
ethylene glycol (A) interact with Pt- and Ru-ion sites, resulting
in the oxidation of the alcohol groups to aldehydes (B and C).
These aldehydes are not very stable and are easily oxidized to
glycolic (D) and oxalic acid (E), respectively. The two car-
boxylic acids may be further oxidized to CO2or carbonate in
alkaline media. The electrons donated from these oxidation
reactions result in the reduction of the Pt and Ru metal ions. It
is also well known that platinum is an excellent catalyst to
abstract hydrogen from carbon atoms.14Therefore, the oxidation
of ethylene glycol could take place via an alternative or parallel
reaction pathway that involves the hydrogen abstraction from
the carbon atoms of ethylene glycol by platinum, as shown in
eq 2:
This reaction pathway results in the adsorption of CO (COads)
on platinum that may be further oxidized to CO2. In parallel
work, we prepared Pt colloids via this ethylene glycol synthesis
route and deposited the colloids on Au and Teflon substrates.15
Infrared studies confirmed the presence of linearly adsorbed CO
as well as carboxylic acids on the electrode surface, thus
suggesting that ethylene glycol oxidation takes place via both
pathways, eqs 1 and 2. The infrared data also support the view
that oxidation products containing carboxyl groups act as
stabilizers for these colloids. Furthermore, it was shown that
these organics, COads, and carboxyl groups can be removed
either by electrochemical oxidation or by oxidative heat
treatment at “low” temperatures (<160 °C), thus freeing up
catalyst surface sites without changing the properties of the PtRu
catalysts.5,15To establish the reaction mechanism and analyze
the dominating reaction pathway, HPLC analysis of the synthesis
solutions was carried out after the PtRu colloids were adsorbed
on carbon, that is, in the final synthesis solution, as described
in the Experimental Section. Both oxalic and glycolic acids were
detected in the chromatograms with retention times of 4.2 and
7.2 min, respectively. The presence of these two species in the
final synthesis solution confirms that ethylene glycol is at least
partly oxidized according to the reaction scheme shown in eq
1. The oxalic and glycolic acid concentrations in the final
synthesis solution were determined as 6 × 10-5and 1.5 × 10-2
M, respectively, independent of the NaOH concentration. The
quantitative analysis indicates that glycolic acid is the dominat-
ing product of the ethylene glycol oxidation reaction. The
oxidation of one ethylene glycol molecule to one oxalic and
one glycolic acid molecule yields 8 and 4 electrons, respectively.
This suggests that “electron concentrations” of at least 5 × 10-4
and 6 × 10-2M, respectively, are generated by the oxidation
of ethylene glycol to these two acids. The reduction reactions
of the 1.4 × 10-2M Ru3+and 1.4 × 10-2M Pt4+, if reduced
to metals, require electron concentrations of 4.2 × 10-2and
5.6 × 10-2M, respectively. Therefore, the complete reduction
of the two noble metal precursor salts to metals requires a total
electron concentration of 9.8 × 10-2M, of which at least 6 ×
10-2M, that is, more than 60%, are produced by the oxidation
of ethylene glycol to glycolic acid. This suggests that the
majority of the electrons needed for the noble metal salt
reduction are provided by the oxidation of ethylene glycol to
glycolic acid.
Adjustment of the PtRu Particle Size. The glycolic acid
concentration of 1.5 × 10-2M found in the final synthesis
solution is comparable to the precursor salt concentration, and
hence this molecule could be the stabilizer of the PtRu colloids,
as discussed in this section. The dissociation constant of glycolic
acid is 1.48 × 10-4mol L-1at 25 °C;13that is, glycolic acid is
present in its deprotonated form as the glycolate anion, A-, in
alkaline solutions and in its protonated form, HA, in acidic
solutions. Glycolate is believed to act as a good stabilizer for
the colloids, possibly forming chelate-type complexes via its
carboxyl groups. Such interactions between the PtRu catalysts
and the neutral, acidic form, that is, glycolic acid, are smaller,
and glycolic acid is believed to be a poor stabilizer. Therefore,
the pH of the synthesis solution is expected to greatly influence
the stability and size of the resulting colloids, in the range where
the glycolate concentration changes. The theoretical concentra-
tions of the HA and A-species as a function of pH are shown
in Figure 1. An initial glycolic acid concentration of 1.5 × 10-2
M and the dissociation constant of 1.48 × 10-4mol L-1were
used for the calculation. Changes in the glycolate and glycolic
acid concentrations are seen to take place within the pH range
between 6 and 2. At pH values higher than 6, a constant
glycolate concentration is reached, while below pH 2, glycolate
is essentially nonexisting. On the basis of these data, and
assuming that glycolate acts as stabilizers, the PtRu catalyst size
is expected to be the smallest and essentially independent of
the solution pH when larger than 6. The catalyst size is further
expected to continuously increase with decreasing pH in the
pH range from 6 to 2, and again to be essentially independent
of the solution pH, with, however, larger particle sizes at pH
values less than 2.
This proposed pH influence on the resulting PtRu catalyst
size was tested by adjusting the pH of the synthesis solution
(13) CRC Handbook of Chemistry and Physics, 71st ed.; Lide, D. R., Ed.; CRC
Press: Boston, 1990.
(14) Bagotsky, V. S.; Vassiliev, Y. B. Electrochim. Acta 1967, 12, 1323-1343.
(15) Bock, C.; Krasinski, A.; MacDougall, B., in preparation.
A R T I C L E S Bock et al.
8030 J. AM. CHEM. SOC.9VOL. 126, NO. 25, 2004

Page 5

using different NaOH concentrations, as listed in Table 1. The
pH of the freshly prepared, that is, not yet heated, solutions
was found to decrease rapidly and reach a steady-state value
within ca. 30 min. The steady-state values of a particular solution
containing the noble metal salts were always lower than those
measured for ethylene glycol solutions of corresponding NaOH
concentration, that is, in the absence of the noble metal salts,
indicating the influence of these two salts that are present in
high concentrations on the solution pH. Using this NaOH
concentration range, the pH of the unheated synthesis solution
was varied between ca. 11 and 7. In all cases, the initial pH
value dropped by 4-5 orders of magnitude as a result of the
reduction of the noble metal salts at higher temperatures. The
drop in solution pH is consistent with the reaction schemes
shown in eqs 1 and 2.
The resulting particle size and particle size distributions
obtained from TEM images for the PtRu colloids synthesized
using the different NaOH concentrations are shown in Table 1
and Figure 2. All TEM images were obtained for carbon-
supported PtRu catalysts. The data suggest that the size of the
PtRu catalysts can be controlled by varying the NaOH concen-
trations between ca. 0.1 and 0.06 M. It is seen that the catalyst
sizes change within the (final) solution pH range between 6
and 3. This is the same pH range where the glycolate con-
centration changes, consistent with the view that glycolate acts
as a stabilizer for the PtRu colloids. The size of the PtRu
catalysts prepared using solutions of higher pH values, that is,
a final solution pH more positive than 6, was found to be
essentially independent of the solution pH, while the use of
lower NaOH concentrations, which resulted in final solution
pH values of less than 3, resulted in significantly larger PtRu
particles that were in fact solid powders rather than colloidal
solutions. These results are also consistent with the view that
glycolate acts as a stabilizer for the PtRu colloids. Furthermore,
the results show that the synthesis solution pH is a key factor
that influences the catalyst particle size. The possibility of the
reaction of the carboxylic acids and the ethylene glycol to form
an ester has not been investigated. The quantity of esters formed
would be smaller than the estimated amount of glycolic acid.
An ester could potentially act as a stabilizer for the PtRu
particles. However, the control of its ability to stabilize the
particles and hence control the particle size would not be pH
dependent (unlike the situation with glycolic acid).
Acidifying “prepared” colloidal solutions to pH values of less
than 2, by adding small amounts of concentrated H2SO4, had
no effect on the properties of the PtRu catalysts deposited on
carbon black from these solutions. This indicates that once the
colloids have been synthesized, they are no longer influenced
by changes in solution pH. The important factor is the solution
pH during the “synthesis” (i.e., preparation) process. It is
important to note that the particle size/pH correlation reported
here is characteristic of the conditions used in this work. The
use of different noble metal concentrations influences the
solution pH and hence is expected to yield PtRu catalysts of
different particle sizes; thus, the NaOH concentration will need
to be adjusted accordingly. Furthermore, the possibility of
particle size control due to electrostatic effects introduced by
the different amounts of NaOH added can be ruled out. This
has been confirmed in separate PtRu catalyst synthesis experi-
ments, where H2PtCl6instead of PtCl4was used. This use of
the platinic acid instead of PtCl4lowers the pH and increases
the ionic strength of the synthesis solution. The resulting PtRu
particle sizes were found to be conclusively larger than those
for PtRu particles synthesized in a correspondingly more alkaline
solution made using PtCl4under otherwise the same conditions.
This confirms that the pH and not the ionic strength of the
Figure 1. Dependence of glycolic acid (HA; gray line) and glycolate (A-,
black line) concentration on the pH for a total HA + A-concentration of
1.5 × 10-2M. The acid dissociation constant (KD) value of 1.48 × 10-4
mol L-1 13was used for the calculation.
Table 1. Influence of the NaOH Concentration on the Resulting
PtRu Particle Sizea
cNaOH/mol L-1
initial pHb
final pHc
PtRu particle sizea/nm
0.1
0.085
0.075
0.071
0.069
0.068
0.063
∼11.1
∼10.5
∼10
∼9.5
∼7.8
∼7.5
∼7.2
∼6 0.7 ( 0.5
1.2 ( 0.5
1.5 ( 0.8
2 ( 0.8
2.2 ( 0.8
3 ( 1.5
4 ( 1.5
5.5
5
4.5
4
3.5
3.3
aPtRu particle sizes were estimated from TEM images obtained for
carbon-supported catalysts.bpH of the not yet heated synthesis solutions
measured 30 min after dissolving the noble metal precursor salts in the
ethylene glycol + NaOH solutions.cpH of the synthesis solutions that were
heated at 160 °C for 3 h. The solutions were cooled to room temperature
prior to the pH measurements.
Figure 2. Average particle size dependence on the NaOH concentration
in the ethylene glycol synthesis solution for initial 1.4 × 10-2M PtCl4+
1.4 × 10-2M RuCl3solutions. The particle size distributions characteristic
for the particles synthesized using a particular NaOH concentration are
indicated by the vertical lines and diamonds. The particle sizes were
extracted from TEM images obtained for PtRu catalysts supported on Vulcan
XC-72R.
Size-Selected Synthesis of PtRu Nano-CatalystsA R T I C L E S
J. AM. CHEM. SOC. 9 VOL. 126, NO. 25, 2004 8031

Page 6

synthesis solution is the controlling factor for the particle size.
Figure 3a-c shows examples of TEM images obtained for PtRu
particles supported on carbon black. The PtRu particles were
synthesized using the following NaOH concentrations: (a) 0.1,
(b) 0.075, and c) 0.068 M. The corresponding histograms are
shown in Figure 4a-c. It is seen that the NaOH concentration
clearly influences the PtRu particle size, indicating that the
NaOH concentration can be used to prepare PtRu nanoparticles
of selected size. The size can be controlled in the range of 0.7-4
nm. It should, however, be noted that the particle size distribu-
tion is broader for the larger PtRu particles that are synthesized
using lower NaOH concentrations, as indicated in the vertical
size range bars shown in Figure 2.
Changes in the synthesis solution pH during the reduction of
the noble metal salts give an indication of the completion of
the reaction. For the heating conditions used in this work, the
synthesis solution pH was observed to drop rapidly upon heating
and to reach a steady state after ca. 5 min. Within this time
period, the synthesis solution temperature approached 130 °C.
The fact that a steady-state pH was reached within 5 min
suggests that the reduction reaction is complete within this very
short time period. It is obvious that changes in the synthesis
Figure 3. TEM images of 6 wt % Pt and 3 wt % Ru catalysts deposited on Vulcan XC-72R. The PtRu catalysts were synthesized in ethylene glycol
solutions using initial 1.4 × 10-2M PtCl4+ 1.4 × 10-2M RuCl3and the following NaOH concentrations: 0.1 M (a), 0.075 M (b), and 0.068 M (c). The
bars in (a) and (b) indicate a 20 nm scale, and the bar in (c) indicates a 50 nm scale; that is, the following magnifications were used to obtain the TEM
images: 1075 kx (a and b) and 470 kx (c).
A R T I C L E S Bock et al.
8032 J. AM. CHEM. SOC.9VOL. 126, NO. 25, 2004

Page 7

solution temperature and/or rate of heating will affect the
reaction rate as well as the characteristics of the resulting
particles such as their size. The examination of the exact
influence of the temperature and heating rate on the resulting
particle characteristics was not within the scope of this work
and is not further discussed here. Furthermore, for all Pt-Ru
particles synthesized in this work, the Pt:Ru ratio was found to
be the same. The Pt:Ru ratio of the final catalysts was also the
same as the Pt:Ru ratio of the precursor salts, that is, 50:50
atom %, as confirmed by EDX measurements.
Characterization of the Carbon-Supported PtRu Cata-
lysts: (a) XRD Characterization. Figure 5 shows slow scan
XRD spectra obtained for a range of carbon-supported PtRu
catalysts of different sizes. Figure 5a shows the XRD spectra
for the larger PtRu catalysts (>1.2 nm), while Figure 5b shows
the same for smaller (<1 nm) PtRu catalysts. In the XRD spectra
of the larger PtRu particles, the diffraction peaks of the face-
centered cubic (fcc) Pt lattice, particularly the highest intensity
Pt(111) plane, are clearly recognizable in all spectra. Due to
the small particle size of these catalysts, the diffraction peaks
are very broad. In fact, the peaks broaden as the particle size
decreases as predicted by Scherrer’s equation.16Particle size
values extracted from XRD are not listed here, as accurate
particle sizes measurements from XRD data are questionable
and, in fact, proper particle size measurements require TEM
analysis, as carried out in this work. Overall, the XRD spectra
of the larger sized catalysts (Figure 5a) are seen to be identical
with the exception of the Pt diffraction peak widths. Diffraction
peaks for Ru for the larger (>1.2 nm) catalysts were not
observed in the raw XRD data that were collected with and
without the Si standard. The “absence” of diffraction peaks
typical for Ru can be due to a number of reasons such as Ru
not being dissolved in the Pt lattice, that is, forming a PtRu
alloy, and/or the Ru being present in the amorphous form, as
further discussed below. The position of the Pt(111) peak for
the larger (>1.2 nm) sized catalysts is shifted to higher 2θ
positions than that for pure Pt that has a maxima at 39.7645°.17
This indicates that Ru is at least partially dissolved in the Pt
fcc lattice. Lattice parameter values for the Pt fcc phase (aPt)
were obtained by fitting the entire XRD spectra using the Topas
(16) Cullity, B. D. Elements of X-ray Diffraction; Addison-Wesley Publishing
Co., Inc.: Reading, MA, 1956; p 150.
(17) X-ray Diffraction Database; Diffraction Management System for Windows
NT; Scintag, Inc.: Cupertina, CA, 1997.
Figure 4. Histograms to TEM images shown in Figure 3a-c.
Size-Selected Synthesis of PtRu Nano-Catalysts A R T I C L E S
J. AM. CHEM. SOC. 9 VOL. 126, NO. 25, 2004 8033

Page 8

2 computer program. For these PtRu particles, an aPt lattice
parameter value of 0.3906 ( 0.001 nm was estimated. This
lattice parameter value could indicate that a PtRu alloy phase
of Pt:Ru atom % ratio of 85:15 ((8%) is formed. It should be
noted that this value is estimated using a Vegard’s law
relationship established for unsupported, that is, bulk, PtRu
alloys.5Possible influences of the support on the lattice structure
of these nanosized Pt-based catalysts are not accounted for in
this relationship. However, such influences are expected to be
small for the systems studied here, as the majority of the
particles are of sizes for which more atoms are in the bulk than
on the surface. Furthermore, XRD data for PtRu catalysts of
1.2 nm, size deposited on different substrates such as Si and
WOxpowders, yielded the same aPtvalue as estimated for the
carbon black substrates. This supports the use of Vegard’s law
relationship established for bulk alloys to estimate the lattice
parameter values for the supported catalysts discussed here. The
aPtvalue estimated from the raw XRD data was found to be
independent of the particle size for the larger (>1.2 nm)
particles, although it should be noted that the error in the
estimated lattice value increases to (8% as the particle size
decreases. It is important to note that best fits of the experimental
XRD spectra were only obtained considering a Pt fcc as well
as a hexagonal Ru phase. The need for a Ru phase to fit the
data indicates that the Ru is indeed not entirely dissolved in
the Pt fcc lattice, consistent with the less than 50 (i.e., the
suggested 15) atom % Ru dissolved in the Pt fcc lattice. In all
cases, the extracted lattice parameters for the hexagonal Ru
phase were a ) 0.27 nm and c ) 0.43 nm, respectively; while
the error of these values was high, that is, (0.01 nm, the values
are close to those reported in the literature for bulk Ru metals.17
The diffraction peaks of the smaller PtRu particles (Figure
5b) are very broad and of low intensity, as is expected for these
very small particles. However, diffraction peaks are recognizable
at 2θ values of ca. 39.8° and 44°, suggesting the presence of
Pt(111) as well as Ru(101) planes, respectively. The presence
of the Ru(101) plane indicates that the Ru is not, at least entirely,
dissolved in the Pt fcc lattice. It is unclear whether a partial
PtRu alloy formation takes place for these very small catalyst
particles. The 2θ position of the Pt(111) diffraction peak appears
to be shifted to lower values typical for a Pt-only phase of
39.7645°. However, exact values cannot be obtained from these
spectra due to the broad nature of the diffraction peaks of these
very small catalyst particles.
(b) XPS Characterization. Figure 6a and b shows the XPS
spectra for the Pt 4f and Ru 3p, respectively, core level regions
for PtRu colloids of 2 nm size (synthesized using a 0.071 M
NaOH solution). The colloids were deposited on Vulcan XC-
72R, resulting in loadings of 18 wt % of Pt and 9 wt % of Ru
per carbon. The less intense Ru 3p region was analyzed instead
of the main Ru 5d spectra, as the latter overlays with the carbon
1s region. The results obtained by deconvoluting the XPS spectra
are summarized in Table 2. The Pt core level region was
deconvoluted, as described by Bancroft et al.,18while other
literature sources19,20were used for the Ru 3p region. The mean
free path values for Pt and Ru are in the range of 1.5 nm, and
hence the XPS data for the small particles analyzed here are
believed to yield data for the entire particle and not only for
the surface species. It is noteworthy that the survey XPS spectra
did not show the presence of Cl-, as no peak at 198 eV typical
for the highest intensity Cl 3p core level was observed.20The
XPS data suggest that 38% of the Pt is present as Pt metal and
35% is present as two valent Pt-oxide, PtO, which can be
reduced electrochemically. A fraction, 27%, of the Pt is also
present in a higher oxidation state, likely as platinum-dioxide,
PtO2. The deconvolution of the Ru 3p core level region was
more difficult, due to its low intensity and hence increased
contribution of noise. However, a large fraction, ca. 70%, of
the Ru is suggested to be present as Ru metal, and a smaller
fraction, ca. 30%, is present in a higher oxidation state, likely
RuO2. It is not known whether the “partial” oxidation of both
Pt and Ru takes place after the synthesis of the catalysts as
colloids, for example, when the carbon-supported PtRu colloids
are dried in and exposed to air and/or if the reduction reaction
of the noble metal precursor salts is incomplete. The amount
of RuO2is seen to be very small, possibly explaining its absence
in XRD spectra as well as the possibility that the RuO2may be
amorphous.
(18) Bancroft, G. M.; Adams, I.; Coatsworth, L. L.; Bennewitz, C. D.; Brown,
J. D.; Westwood, W. D. Anal. Chem. 1975, 47, 586-588.
(19) Biloen, P.; Pott, G. T. J. Catal. 1973, 30, 169-178.
(20) Wagner, C. D. Handbook of X-ray Photoelectron Spectroscopy; Perkin-
Elmer Corp., Physical Electronics Division: Eden Prairie, MN, 1979.
Figure 5. Slow scan XRD spectra for carbon-supported PtRu catalysts. Si
powder was added to the catalysts as internal standard. Part a shows the
XRD spectra for the larger than 1.2 nm particles, as follows: (a) 4, (b) 3.2,
(c) 2.2, and (d) 1.5 nm average particle size, while part b shows an example
of an XRD spectrum for smaller carbon-supported PtRu particles of 0.7
nm average size. The spectra were collected between 2θ’s of 20° and 70°,
using an acquisition time of 60 s and a step size of 0.06°.
A R T I C L E S Bock et al.
8034 J. AM. CHEM. SOC.9VOL. 126, NO. 25, 2004

Page 9

(c) Scanning TEM Images and EDS Analyses. Figure 7a
shows an annular dark field (ADF) image taken in the scanning
TEM (STEM) mode of a 20:10 wt % Pt:Ru catalyst supported
on carbon black. The catalyst was synthesized using a 0.071 M
NaOH solution. The bright white features seen in this figure
arise from the PtRu noble metal catalysts. Branched-like
structures resulting from agglomeration of particles are observed.
The average particle size is seen to be around 2 nm, consistent
with the particle size and size distribution analysis carried out
individually (see Table 1 and Figure 2). Elemental profiles taken
across two different regions of these catalysts and obtained from
energy-dispersive X-ray spectroscopy (EDS) are shown in
Figure 7b. The EDS profile corresponds to the scanned region,
indicated by the white line shown in Figure 7a. The EDS counts
are shown directly. The counts are proportional to the atomic
concentration of Pt and Ru and reflect changes in the Pt and
Ru atomic concentrations over the scanned distance. However,
Figure 6. XPS core level spectra for the Pt 4f (a) and Ru 3p (b) regions
of a PtRu catalyst of 2 nm average particle size supported on Vulcan XC-
72R. The thick line shows the raw XPS, while the thin lines show the peaks
obtained by deconvoluting the raw XPS data and the squares show the sum
of the deconvoluted spectra.
Table 2. Binding Energies of Pt and Ru Species Obtained from
Curve-Fitted XPS Spectra for Carbon-Supported PtRuaCatalysts
species orbital/spin
binding
energy/eV
peak half
width/eVassignment
relative
concentrations/%b
Pt4f7/2
4f5/2
4f7/2
4f5/2
4f7/2
4f5/2
3p1/2
3p3/2
3p1/2
3p3/2
71.1
74.4
72.1
75.4
74.5
77.7
462.4
484.8
466.1
488.6
1.1
1.1
2
2
2.5
2.5
3.3
3.3
4
4
Pt metal
Pt metal
PtO
PtO
PtO2
PtO2
Ru metal
Ru metal
RuO2
RuO2
24
14
20
15
16
11
47
24
23
6
Ru
aPtRu catalyst (2 nm average particle size) deposited on Vulcan XC-
72R. 20 wt % Pt and 10 wt % Ru per carbon.bRelative concentrations are
equal to the corresponding, deconvoluted peak areas divided by the total
XPS signal area extracted from the experimental XPS core level regions of
either Pt 4f or Ru 3p.
Figure 7. Part a shows an annular dark field image taken in the scanning
TEM (STEM) mode for a 20 wt % Pt and 10 wt % Ru catalyst supported
on Vulcan XC-72R. The PtRu catalyst was synthesized using a 0.071 M
NaOH solution, resulting in a 2 nm average PtRu particle size. The white
line indicates the region scanned for the EDS data shown in (b). The black
line in (b) shows the counts for Pt, while the gray line shows the counts
for Ru.
Size-Selected Synthesis of PtRu Nano-CatalystsA R T I C L E S
J. AM. CHEM. SOC. 9 VOL. 126, NO. 25, 2004 8035

Page 10

these counts do not yield the actual Pt:Ru atomic concentrations,
as they are not normalized. The EDS profiles show no clear
signs of segregation; that is, no particles consist of only Ru or
Pt. However, because of the low counts, the experimental error
is significant and the analysis is not sensitive to small variations.
It is noteworthy that HRTEM images were obtained that
confirmed the crystalline nature of the PtRu catalysts. However,
due to the limited resolution of the TEM, no additional
information such as lattice spacing values was extracted from
the data and the HRTEM images are not shown and discussed
here.
Electro-oxidation Characteristics of CH3OH. Carbon-
supported PtRu catalysts electrodes were prepared on Au foils,
as described in the Experimental Section. Electrodes consisting
of small amounts (ca. 0.02 mg of Pt and ca. 0.01 mg of Ru) of
catalysts and forming thin catalyst layers were prepared in this
manner. The use of thin layer electrodes ensures that all catalyst
sites are accessible to the reactant, that is, that true catalytic
activities rather than mass transport limitations of the reactant
within the catalyst layer are studied. Figure 8a and b shows the
CH3OH oxidation activity at 60 °C for electrodes made using
carbon-supported PtRu particles of 2 nm average size, that is,
synthesized using a 0.071 M NaOH solution. Electrochemical
data for this particular catalyst are shown here, as studies carried
out in parallel work showed this catalyst to exhibit high CH3-
OH oxidation activity per catalyst weight. For comparison, the
CH3OH oxidation activities for two commercially available cata-
lysts (Alfa Aesar and E-TEK, Inc.) are also shown in Figure 8
a. All electrodes were prepared in the same manner, and
catalyst loadings of 20 wt % Pt and 10 wt % Ru on carbon
were used in all cases. The CVs suggest that PtRu catalysts
prepared in this work exhibit better CH3OH electro-oxidation
activities than the two commercial catalysts tested, as higher
CH3OH oxidation currents are observed at lower potentials. In
parallel work, current-transient measurements at constant po-
tential were carried out over 8 h periods, showing that the
catalytic advantage of these PtRu catalysts is maintained. Each
CV curve shown is the average of three different electrodes
prepared using a particular catalyst type and loading. In the case
of the Alfa Aesar and E-TEK, Inc. catalysts, the same i-V curve
shapes were obtained for multiple electrodes. However, two
distinctively different i-V curve types were obtained for the
PtRu catalysts prepared in this work, as shown in Figure 8b.
The CV type shown in the black curve in Figure 8b (which is
identical to the CV shown in Figure 8a) was observed for ca.
70% of the prepared electrodes made of PtRu catalysts of
particle size larger than 1.2 nm, while the gray curve shown in
Figure 8b was observed for ca. 30% of these electrodes. In the
case of the electrodes prepared using smaller than 1 nm PtRu
catalysts, the latter CV type, that is, the gray curve in Figure
8b, was dominant at ca. 90% observed. The electrodes prepared
in this work are made of very small amounts of catalysts, and
hence the fact that two distinctively different CV types are
observed for the prepared PtRu catalysts possibly indicates that
the individual electrodes consist of PtRu catalysts of two
characteristically different properties, such as, for example, a
mainly PtRu alloy versus separate Pt and Ru phases.
Concluding Remarks
In this study, a simple and rapid synthesis method for the
preparation of PtRu colloids that can be subsequently deposited
on suitable substrates such as high surface area carbon blacks
has been introduced. PtRu catalysts of a nominal and final
composition of 50:50 atom % Pt:Ru were studied. It has been
shown that the size of the PtRu catalyst particles can be varied
in the range of 0.7-4 nm. The formation of the particles
involves the reduction of the Pt and Ru precursor salts by the
solvent, ethylene glycol, which in turn is oxidized. The oxidation
of ethylene glycol is shown to mainly result in glycolic acid
or, depending on the synthesis solution pH, the glycolate anion.
The glycolate anion acts as a stabilizer for the PtRu colloids,
and its concentration, and hence the resulting PtRu particle size,
is controlled via the synthesis solution pH. The fact that the
size of the PtRu catalyst particles is varied via the pH and
glycolate concentration only, that is, maintaining other experi-
mental conditions constant, appears to result in PtRu catalyst
compositions that are independent of the catalyst particle size.
This was supported by XRD data obtained for carbon-
supported PtRu catalyst particles larger than 1.2 nm. The XRD
data suggest that Ru is partially (ca. 15 atom %) dissolved in
Figure 8. Cyclic voltammograms recorded at 10 mV s-1in 0.5 M CH3-
OH + 0.5 M H2SO4solutions at 60 °C. Part a shows the CVs for a 2 nm
average particle size PtRu catalyst synthesized in this work (labeled as EG
catalyst), a PtRu catalyst from Alfa Aesar (A/A catalyst), and a PtRu catalyst
from E-TEK, Inc. Part b shows the two CV types observed for the 2 nm
average particle size PtRu catalysts prepared in this work. In all cases, 20
wt % Pt and 10 wt % Ru on Vulcan XC-72R catalysts loadings were used.
A R T I C L E S Bock et al.
8036 J. AM. CHEM. SOC.9VOL. 126, NO. 25, 2004

Page 11

the fcc Pt lattice and that a separate hexagonal Ru phase is
also present. The composition of the less than 1 nm PtRu
catalysts appears to be different, and the catalyst particles
possibly consist of separate Pt and Ru phases. It is possible
that alloy formation of such very small particles that consist of
only a few atoms (<40) that are mainly (>80%) surface atoms
is very difficult or may even not be possible for the PtRu
system.21
Initial studies showed that particular PtRu catalysts prepared
in this work display better CH3OH electro-oxidation activities
than two commercial catalysts tested. This shows the potential
of these synthesis methods to generate improved catalyst
formulations for low-temperature methanol fuel cells. However,
it needs to be emphasized that long-term studies in real fuel
cell systems need to be carried out for these catalysts. Also,
the synthesis method will need to be further optimized to gain
a better understanding and control of the PtRu alloy versus
separate Pt and Ru phase formation.
It is noteworthy that this synthesis method is also attractive
because the glycolate stabilizer is a simple organic molecule
that can be oxidatively removed either electrochemically or by
oxidative heat treatment at temperatures as low as 160 °C.15
Heat treatments at these “low” temperatures free up valuable
catalyst sites without resulting in changes of the PtRu catalyst
properties.5,15
Acknowledgment. We thank I. Sproule and S. Moisa (NRC,
Ottawa) for the XPS analysis and D. Wang (NRC, Ottawa) for
the TEM analysis. P. L’Abbe’s (NRC, Ottawa) assistance for
the preparation of the glass cells used in this work is also greatly
acknowledged. Financial support from a joint NRC/Helmholtz
grant and the NRC fuel cell program is also gratefully appre-
ciated.
JA0495819
(21) LePage, Y. Program Clusurf, ICPET; National Research Council of Canada,
Ottawa, 2002.
Size-Selected Synthesis of PtRu Nano-CatalystsA R T I C L E S
J. AM. CHEM. SOC. 9 VOL. 126, NO. 25, 2004 8037