Electrochemical reduction of oxygen on palladium nanocubes in acid and alkaline solutions

Article abstract of DOI:10.1016/j.electacta.2011.10.074

The electrocatalytic properties of cubic palladium nanoparticles towards the electrochemical reduction of oxygen were studied in acid and alkaline solutions and compared with those of spherical nanoparticles and bulk Pd. The synthesised Pd nanoparticles were characterised by transmission electron microscopy (TEM) and X-ray diffraction (XRD). Electrooxidation of pre-adsorbed CO was employed for cleaning the palladium catalyst surface. Oxygen reduction was studied using the rotating disk electrode (RDE) method and enhanced electrocatalytic activity of Pd nanocubes was revealed both in acid and alkaline solutions, which was attributed to the prevalence of Pd(1 0 0) facets. The mechanism of oxygen reduction on Pd nanoparticles was similar to that on bulk Pd, the first electron transfer being the rate-limiting step, and the reaction predominantly followed a four-electron pathway in both solutions.

Full text of DOI:10.1016/j.electacta.2011.10.074

Electrochimica Acta 59 (2012) 329–335
Contents lists available at SciVerse ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Electrochemical reduction of oxygen on palladium nanocubes in acid and
alkaline solutions
H. Eriksona, A. Sarapuua, N. Alexeyevaa,1, K. Tammeveskia,∗,1, J. Solla-Gullónb,1, J.M. Feliub,1
a Institute of Chemistry, University of Tartu, Ravila 14a, 50411 Tartu, Estonia
b Instituto de Electroquímica, Universidad de Alicante, Apartado 99, 03080 Alicante, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 24 August 2011
Received in revised form 20 October 2011
Accepted 24 October 2011
Available online 3 November 2011
The electrocatalytic properties of cubic palladium nanoparticles towards the electrochemical reduction
of oxygen were studied in acid and alkaline solutions and compared with those of spherical nanoparticles
and bulk Pd. The synthesised Pd nanoparticles were characterised by transmission electron microscopy
(TEM) and X-ray diffraction (XRD). Electrooxidation of pre-adsorbed CO was employed for cleaning the
palladium catalyst surface. Oxygen reduction was studied using the rotating disk electrode (RDE) method
and enhanced electrocatalytic activity of Pd nanocubes was revealed both in acid and alkaline solutions,
which was attributed to the prevalence of Pd(1 0 0) facets. The mechanism of oxygen reduction on Pd
nanoparticles was similar to that on bulk Pd, the first electron transfer being the rate-limiting step, and
the reaction predominantly followed a four-electron pathway in both solutions.
Keywords:
Oxygen reduction
Palladium nanocubes
Palladium nanoparticles
Pd-based catalysts
Electrocatalysis
© 2011 Elsevier Ltd. All rights reserved.
1. Introduction
ness in H2SO4 solution [11,12], but it was rather constant in HClO4
[11]. Improved electrocatalytic properties for O2 reduction have
Nanostructured palladium catalysts have received much atten-
tion in recent years as rather active catalysts for the electrochemical
reduction of oxygen [1,2]. Pd has properties similar to those of plat-
inum and the mechanism of O2 reduction in both acid and alkaline
solutions is the same on these metals [3,4]. As compared to Pt, Pd
is more abundant on the Earth and considerably cheaper [2], but
the activity of pure Pd towards O2 reduction is lower and Pd is
less stable in acid media [1]. Bimetallic Pd-based catalysts, such as
alloys and core–shell particles with non-noble metals (Co, Fe, etc.),
have been shown to possess higher electrocatalytic activity for oxy-
gen reduction than pure Pd [1,2]. A further advantage of Pd and its
alloys is their good selectivity for oxygen reduction reaction (ORR)
in the presence of alcohols, which makes them attractive catalysts
for direct alcohol fuel cells [2].
There are several studies of oxygen reduction on nanostructured
Pd catalysts in acid media [5–19]. The activity of Pd/C catalysts
is generally considerably lower than that of Pt/C [8,9]. However,
very high catalytic activity has been observed for Pd nanorods
prepared by electrodeposition [10]. For vacuum-evaporated thin
Pd films, the specific activity of O2 reduction was similar to that
of bulk Pd and slightly decreased with decreasing the film thick-
been observed on Pd nanoparticles (PdNPs) supported on carbon
nanotubes [13–17]. For Pd/C catalysts, the activity can also be influ-
enced by pre-treatment of the carbon support [18].
As on platinum, oxygen reduction on Pd is a structure-sensitive
reaction [1]. It has been demonstrated that in HClO4 solution,
the ORR activity on Pd single crystals increases in the order of
Pd(1 1 0) < Pd(1 1 1) < Pd(1 0 0) [20]. In contradiction to that, high
catalytic activity of Pd nanorods as compared to spherical Pd
nanoparticles has been attributed to the prevalence of Pd(1 1 0)
facets [10]. Very recently, the oxygen reduction studies on Pd
nanocubes with a preferential (1 0 0) surface orientation have been
published [6,7]. The specific activity of cubic Pd particles was com-
parable to that of Pt, but the activity of Pd octahedra, however,
was 10 times lower [6]. The higher ORR activity of Pd nanocubes
was attributed to the lower coverage of chemisorbed OH, which in
turn offers more available reaction sites [6], or, in H2SO4 solution, to
the structure-dependent adsorption of (bi)sulphate ions [7]. In con-
trast, studies of the ORR on stepped surfaces of n(1 1 1)–(1 0 0) series
of Pd in HClO4 revealed that the activity increases with increas-
ing the terrace atom density, although the oxide coverage also
increases. It has been proposed that on Pd, the active sites for the
ORR are the terraces [21].
In alkaline solutions, Pd catalysts have shown remarkably high
activity and good stability and are therefore considered as promis-
ing electrocatalysts for alkaline fuel cells [19,22–29]. In some cases,
activities comparable to that of Pt [23] or even higher [25,28] have
∗
Corresponding author. Tel.: +372 7375168; fax: +372 7375181.
E-mail addresses: kaido@chem.ut.ee, kaido.tammeveski@ut.ee (K. Tammeveski).
ISE member.
1
0013-4686/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2011.10.074

330
H. Erikson et al. / Electrochimica Acta 59 (2012) 329–335
been observed. The specific activity of the Pd/C catalysts depends
on the Pd particle size, decreasing by a factor of about three with
decreasing particle size from 16.7 to 3 nm [24]. However, graphene-
supported Pd nanoparticles with a mean diameter of only 1.8 nm
showed significantly high catalytic activity for the ORR [28].
The aim of this study was to synthesise cubic Pd nanoparti-
cles and to compare their electrocatalytic activity towards oxygen
reduction with that of spherical Pd nanoparticles and bulk Pd in
acidic as well as in alkaline solutions.
PGSTAT30 (Eco Chemie B.V., The Netherlands). An EDI101 rotator
and a CTV101 speed control unit (Radiometer, Copenhagen) were
used for the RDE experiments. All experiments were carried out at
room temperature (23 1 ◦C).
Prior to the O2 reduction measurements, the electrodes were
electrochemically pre-treated in Ar-saturated 0.05 M H2SO4 or
0.1 M KOH solution by scanning the potential between 0.1 and 0.7 V.
To clean the Pd surface, carbon monoxide (AGA) was adsorbed onto
the Pd catalyst by bubbling CO through the electrolyte at 0.1 V
until complete blockage of the surface, which was monitored by
cycling the electrode between 0.1 and 0.35 V [32]. After that, CO was
removed from the solution by bubbling argon for 45 min. Finally,
CO was oxidatively stripped off from the surface by scanning the
potential up to 1.0 V and the voltammogram corresponding to the
CO-free surface was again recorded. After that the electrode was
immediately transferred into an O2-saturated solution in another
cell in order to avoid surface contamination in air. For O2 reduction
measurements, the potential was scanned between 0.1 and 1 V at
2. Experimental
2.1. Synthesis and characterisation of Pd nanoparticles
Pd nanocubes were synthesised using a previously described
methodology [30] in which H2PdCl4 solution was reduced with
ascorbic acid in the presence of cetyltrimethylammonium bromide
(CTAB) at 95 ◦C. The sample was twice centrifugated and redis-
persed in water and the PdNPs were then cleaned with strong
basic aqueous solution followed by washing 3–4 times in ultrapure
water to finally achieve a water suspension. The synthesis of spher-
ical Pd nanoparticles has been adapted from the citrate method
usually employed for the synthesis of gold nanoparticles [31]. In
brief, the metallic precursor (H2PdCl4) was reduced with ice-cold
sodium borohydride in the presence of sodium citrate (which acts
as stabiliser). Afterwards, solid NaOH was added to produce the
precipitation of the nanoparticles. After complete precipitation, the
nanoparticles were washed 3–4 times with ultrapure water.
The nanoparticles were characterised using a Transmission Elec-
tron Microscope (TEM) (JEM-2010, JEOL) working at 200 kV. The
samples for TEM analysis were obtained by placing a droplet of
the PdNPs suspension onto a formvar/carbon coated copper grid
and allowing the solvent to evaporate in air at room temperature.
X-ray diffractograms were recorded in a Bruker D8 advance diffrac-
tometer, using Cu K␣ radiation (ꢀ = 0.15418 nm) operated at 40 kV
and 40 mA. Spectra were recorded between 30◦ and 90◦ (2ꢁ) with a
step of 0.1◦ and a time per step of 30 seconds. The PdNP samples for
XRD analysis were prepared by evaporating a drop of nanoparticle
suspension on a Si substrate.
10 mV s−1
.
3. Results and discussion
3.1. Surface characterisation of Pd nanoparticles
The TEM micrographs for cubic and spherical PdNPs synthe-
sised are presented in Fig. 1. Fig. 1a shows the presence of a high
percentage of Pd nanocubes, for which the (1 0 0) preferential sur-
face structure is expected. In contrast, Fig. 1b shows quasi-spherical
Pd particles which can be considered as representative of polyori-
ented, nonspecifically structured nanoparticles. The particle size
determined from the TEM images was 26.9 3.9 and 2.8 0.4 nm
for cubic and spherical PdNPs, respectively.
Fig. 2 shows the XRD patterns of the ∼27 nm Pd nanocubes and
∼3 nm Pd spheres. All the diffraction peaks can be well-indexed
to face-centered cubic (fcc) palladium metal. Remarkably, the XRD
pattern of Pd nanocubes shows an abnormally intense (2 0 0) peak,
which is not observed for the Pd spheres, suggesting that most
of the Pd nanocubes are preferentially oriented with their (1 0 0)
facets parallel to the substrate. This feature has been previously
observed with ∼50–70 nm Pd nanocubes [30,33,34]. However, this
(2 0 0) abnormally intense peak was not observed for 6–7 nm Pd
nanocubes [6] which could suggest a “size effect”. Nevertheless,
∼60 [35] and ∼24 nm Pd nanocubes [36] also showed the absence
of this (2 0 0) abnormally intense peak. Additional experiments are
required to clarify these discrepancies.
2.2. Electrochemical measurements
Glassy carbon (GC) and bulk polycrystalline palladium elec-
trodes were prepared by mounting GC (GC-20SS, Tokai Carbon)
and Pd (99.95%, Alfa Aesar) disks into Teflon holders. The geometric
area of the electrodes (A) was 0.2 cm2. The surface of the GC elec-
trodes was polished to a mirror finish with 1.0 and 0.3 ␮m alumina
slurries (Buehler); bulk Pd electrodes were finished by polishing
with a 0.05 ␮m alumina slurry. After polishing, the electrodes were
ultrasonically cleaned in Milli-Q (Millipore, Inc.) water for 5 min.
An aliquot of the suspension of PdNPs in water was pipetted onto
a polished GC substrate and the solvent was allowed to evaporate
in air. In this work 4–5 electrodes were tested in each electrolyte.
The GC electrode surface was freshly modified with PdNPs before
electrochemical testing. Oxygen reduction was studied using the
rotating disk electrode (RDE) technique. The solutions were pre-
pared from 96% H2SO4 (Suprapur, Merck) or from KOH pellets (pro
analysi, Merck) and Milli-Q water and were saturated with pure
O2 (99.999%, AGA) or deaerated with Ar gas (99.999%, AGA). A
reversible hydrogen electrode (RHE) connected to the cell through a
Luggin capillary was employed as a reference and all the potentials
are referred to this electrode. A Pt wire served as a counter electrode
and the counter electrode compartment of the three-electrode cell
was separated from the main cell compartment by a glass frit.
Potential was applied with an Autolab potentiostat/galvanostat
3.2. Cyclic voltammetry (CV) and CO stripping
In order to clean the surface of PdNPs without altering their
initial morphology, residual impurities on the surface were dis-
placed by adsorption of CO, followed by its oxidation [32]. Typical
CO stripping voltammograms are presented in Fig. 3. In acid solu-
tion, a well-defined stripping peak of adsorbed CO was observed at
0.88 V for cubic PdNPs and at 0.91 V for spherical PdNPs. The second
anodic peak at 0.94 V can be attributed to the surface oxidation on
Pd(1 0 0) terraces. This characteristic signal has been well-studied
in Pd(S) − [n(1 0 0) × (1 1 1)] [37] and Pd(S) − [n(1 0 0) × (1 1 0)] [38]
electrodes and its charge has been reported to exclusively depend
on the terrace atom density without contributions from step sites.
The cathodic peak at ca 0.77 V corresponds to the reduction of Pd
surface oxides. The CV response of PdNPs recorded after CO strip-
ping (inset of Fig. 3a) showed improved definition and symmetry
of the CV peaks as a result of increased cleanliness of the surface.
The voltammetric profile of the cubic PdNPs resembles that char-
acteristic of a Pd(1 0 0) single crystal and shows well-defined peaks

H. Erikson et al. / Electrochimica Acta 59 (2012) 329–335
331
Fig. 2. XRD patterns of Pd catalysts: (a) cubic PdNPs and (b) spherical PdNPs.
212 ␮C cm−2) [41] and the Ar values obtained by these two methods
were in a good agreement with each other.
Fig. 1. TEM images of Pd catalysts: (a) cubic PdNPs and (b) spherical PdNPs.
3.3. Oxygen reduction in acidic media
The electroreduction of O2 on Pd nanoparticles and bulk Pd
was studied in O2-saturated 0.05 M H2SO4 solution using the RDE
method. Fig. 5 shows the representative polarisation curves of O2
reduction on cubic PdNPs at various rotation rates; the background
current registered in O2-free solution has been subtracted from
these data and only the anodic sweeps are presented and subjected
to further analysis. Similar single-wave polarisation curves were
also observed for spherical PdNPs and bulk Pd. A comparison of the
O2 reduction data at a single electrode rotation rate is presented in
Fig. 6.
at 0.35 and 0.22 V [37,38]. In alkaline solution (Fig. 3b) sharp CO
stripping peaks appear at the potentials of 0.72 and 0.75 V for Pd
nanocubes and spherical Pd nanoparticles, respectively. The peaks
of hydrogen adsorption and desorption on CVs recorded after CO
stripping are not as well-defined as in acid solution, but the peak
characteristic to Pd(1 0 0) terraces can be observed at ca 0.55 V [39].
After the O2 reduction measurements, the CV curves were again
recorded in Ar-saturated 0.05 M H2SO4 or 0.1 M KOH by extending
the anodic potential to 1.4 V (Fig. 4). The anodic peaks at E > 0.7 V
and the cathodic peaks at ca 0.65–0.7 V correspond to the forma-
tion of Pd surface oxides and to their reduction, respectively. For
cubic PdNPs in alkaline solution (Fig. 4b), characteristic peaks of
hydrogen desorption and oxide formation at Pd(1 0 0) terraces are
in evidence at the potentials of 0.58 and 0.87 V, respectively [39]. By
integrating the charge under the cathodic peak, the real electroac-
tive area (Ar) of Pd catalysts was estimated, assuming the value of
424 ␮C cm−2 as the charge density for the reduction of a monolayer
of PdO [40]. For comparison, the real surface area of PdNPs was also
determined from the Hupd desorption peaks on the CV curves reg-
istered in Ar-saturated 0.05 M H2SO4 (using the charge density of
The RDE data were analysed using the Koutecky–Levich (K–L)
equation [42]:
1
j
1
jk
1
jd
1
1
=
+
= −
−
(1)
nFkCOb
0.62nFD2O/3
ꢂ
−1/6COb ω1/2
2
2
2
where j is the measured current density, jk and jd are the kinetic and
diffusion-limited current densities, respectively, n is the number of
electrons transferred per O2 molecule, k is the rate constant for O2
reduction, F is the Faraday constant (96 485 C mol−1), ω is the elec-
trode rotation rate, COb is the concentration of oxygen in the bulk
2

332
H. Erikson et al. / Electrochimica Acta 59 (2012) 329–335
Fig. 3. Oxidation of pre-adsorbed CO on PdNP modified GC electrodes in Ar-
saturated 0.05 M H2SO4 (a) and 0.1 M KOH (b). ꢂ = 20 mV s−1. Insets: CV curves for
PdNPs after CO stripping, ꢂ = 50 mV s−1. Current densities are normalised to the real
surface area of Pd.
Fig. 4. Cyclic voltammograms for PdNP modified GC electrodes after O2 reduction
measurements in Ar-saturated 0.05 M H2SO4 (a) and 0.1 M KOH (b). ꢂ = 50 mV s−1
.
Current densities are normalised to the real surface area of Pd.
polyoriented PdNPs and bulk Pd as compared to cubic PdNPs that
have mostly (1 0 0) facets on the surface. This is also supported by an
IRAS study, which shows that the coverage of (bi)sulphate anions
on Pd(1 0 0) is lower than that of Pd(111) [47].
It is evident from Fig. 6 that for PdNPs (both cubic and spheri-
cal) the oxygen reduction current does not increase as sharply as for
bulk Pd and the kinetic region is wider. We suggest that this is due to
(1.22 × 10−6 mol cm−3) [43], DO is the diffusion coefficient of oxy-
2
gen (1.93 × 10−5 cm2 s−1) [43] and ꢂ is the kinematic viscosity of the
solution (0.01 cm2 s−1) [44]. The K–L plots were constructed (inset
of Fig. 5) and the number of electrons transferred per O2 molecule
(n) was calculated from Eq. (1). On PdNPs, the value of n was close
to four at negative potentials, which is in accordance with the ear-
lier studies [5,7,11,12,19]. At more positive potentials, however, the
n value decreased to about 3.5 at E = 0.7 V, indicating that H2O2 is
produced to some extent. This may be due to the decreased activ-
ity of hydrogen peroxide reduction on Pd nanoparticles in H2SO4
solution [45]. On bulk Pd the 4e− reduction proceeds in the whole
range of potentials studied.
Comparison of the j–E curves of oxygen reduction on PdNPs and
bulk Pd (Fig. 6) shows that the onset potential of O2 reduction
is about 60 mV more positive for Pd nanocubes than for spher-
ical PdNPs and bulk Pd. This increased activity of O2 reduction
is most probably related to the morphological features of cubic
PdNPs, specifically, to the predominance of Pd(1 0 0) surface sites.
The ORR activity of Pd single crystals in HClO4 has been shown to
increase in the following order: Pd(1 1 0) < Pd(1 1 1) < Pd(1 0 0) [20].
In addition, in H2SO4 solution, the adsorption of (bi)sulphate ions
may also have an influence to the O2 reduction activity. It is well
known that on the surface of platinum, (bi)sulphate ions show the
strongest specific adsorption on Pt(1 1 1) sites, thereby blocking the
O2 adsorption centers and also having a negative electronic effect
on the ORR kinetics [46]. On Pd the adsorption of anions is even
stronger than on Pt [1] and therefore it is also expected that the
presence of Pd(1 1 1) sites decreases the O2 reduction activity on
Fig. 5. RDE voltammetry curves for O2 reduction on cubic PdNP modified GC elec-
trodes in O2-saturated 0.05 M H2SO4. ꢂ = 10 mV s−1. Inset: Koutecky-Levich plot for
O2 reduction at 0.3 V.

H. Erikson et al. / Electrochimica Acta 59 (2012) 329–335
333
Fig. 6. A comparison of RDE voltammetry curves on PdNP modified GC electrodes
(Ar = 0.56 cm2 and Ar = 0.53 cm2 for cubic and spherical PdNPs, respectively) and bulk
Fig. 7. Mass-transfer corrected Tafel plots for oxygen reduction on PdNP modified
GC electrodes and bulk Pd in 0.05 M H2SO4. ω = 1900 rpm.
Pd (Ar = 0.57 cm2) in O2-saturated 0.05 M H2SO4. ω = 1900 rpm, ꢂ = 10 mV s−1
.
the values of the Tafel slope for Pd nanocatalysts were slightly
higher than −60 mV, which is the typical value for bulk polycrys-
talline Pd [3] and nanostructured Pd catalysts [11,12,18,51]. At
higher current densities, the Tafel slope values near to −120 mV
have been observed for bulk Pd [3], Pd/C catalysts [18] and thin Pd
films [11,12], whereas a constantly changing Tafel slope has been
reported in Ref. [51]. The Tafel slope values of −60 and −120 mV are
also characteristic for O2 reduction on Pt catalysts [52]. A change in
the slope has been attributed to the change from Temkin to Lang-
muir conditions for the adsorption of reaction intermediates, but
the transfer of the first electron to O2 is the rate determining step in
both regions [52]. The Tafel slope values obtained in this work sug-
gest that the reaction mechanism on Pd nanoparticles is the same
as on bulk Pd and Pt.
the uneven coating of the electrode substrate with PdNPs [48]. Bulk
Pd shows a higher reduction current at E < 0.8 V, but this potential
is already close to the half-wave potential of oxygen reduction and
therefore not relevant from the electrode kinetics point of view.
Based on these considerations, it appears that 0.85 V is the most
appropriate potential for the determination of the intrinsic elec-
trocatalytic activity of Pd catalysts for oxygen reduction in H2SO4
solution.
The specific activities (SA) of Pd catalysts at 0.85 V were calcu-
lated:
Ik
SA =
(2)
Ar
where Ik is the kinetic current at a given potential and Ar is the
real electroactive area of palladium. The SA value for Pd nanocubes
appeared to be considerably higher than these of spherical Pd
nanoparticles and bulk Pd (Table 1). This is in accordance with the
recent studies on Pd nanocubes in acid solutions, where about three
times higher SA value was obtained for Pd nanocubes deposited
onto Au electrode in 0.5 M H2SO4 solution than for spherical Pd
nanoparticles [7]. In addition, an activity increase of Pd nanocubes
of about 10 and 6 times over Pd/C octahedra and a commercial
Pd/C catalyst, respectively, has been recently found in HClO4 solu-
tion [6]. Similarly, for Pt nanocubes with preferential (1 0 0) surface
orientation, an enhancement in oxygen reduction as compared
to polycrystalline Pt nanoparticles has been observed [49,50]. On
spherical PdNPs, the ORR activity is slightly lower than that on
bulk Pd, which is in accordance with the results obtained on poly-
oriented nanostructured Pd electrodes, where the SA values were
somewhat lower than that of bulk Pd [12]. This can be attributed
to the influence of Pd particle size and/or morphology.
3.4. Oxygen reduction in alkaline media
For comparison purposes, the reduction of oxygen on PdNPs was
studied also in alkaline solution. In Fig. 8, a series of RDE voltamme-
try curves of O2 reduction on Pd nanocubes is shown. As compared
to the results in acid, the electrocatalytic activity of Pd is higher
in alkaline solution and a well-defined diffusion-limited current
plateau is formed. Based on the RDE data, the K–L plots were con-
structed (inset of Fig. 8) and from Eq. (1), the value of n ≈ 4 was
On the basis of the RDE data at 1900 rpm (Fig. 6) the mass-
transfer corrected Tafel plots were constructed (Fig. 7) and the Tafel
slopes were determined (Table 1). In the low current density region,
Table 1
Kinetic parameters for oxygen reduction on PdNPs and bulk Pd in 0.05 M H2SO4.
ω = 1900 rpm.
Catalyst
Tafel slope
Tafel slope
SA at 0.85 V
(mA cm−2
(mV) I regiona
(mV) II regiona
)
Cubic PdNPs
Spherical PdNPs
Bulk Pd
−67 11
−81 12
−151 27
−156 20
0.35 0.09
0.15 0.02
0.20 0.02
−61
1
−126
2
Fig. 8. RDE voltammetry curves for O2 reduction on cubic PdNP modified GC elec-
trodes in O2-saturated 0.1 M KOH. ꢂ = 10 mV s−1. Inset: Koutecky–Levich plot for O2
reduction at 0.3 V.
a
Region I corresponds to low current densities and Region II to high current
densities.

334
H. Erikson et al. / Electrochimica Acta 59 (2012) 329–335
Table 2
Kinetic parameters for oxygen reduction on PdNPs and bulk Pd in 0.1 M KOH.
ω = 1900 rpm.
Catalyst
Tafel slope
Tafel slope
SA at 0.95 V
(mA cm−2
(mV) I regiona
(mV) II regiona
)
Cubic PdNPs
Spherical PdNPs
Bulk Pd
−79
3
−180 26
−181 11
3
0.28 0.06
0.073 0.005
0.13 0.01
−77 13
−75
a
Region I corresponds to low current densities and Region II to high current
densities.
The values of the specific activities at 0.95 V calculated from Eq.
(2) are given in Table 2. A rather positive potential was chosen to
allow a better comparison with state-of-the-art cathode catalysts.
Similarly to the results obtained in acid solution, the SA increases in
the sequence: spherical PdNPs < bulk Pd < cubic PdNPs and it is most
likely due to the morphological differences of PdNPs. To our knowl-
edge, the ORR on Pd single crystals has not yet been systematically
studied in alkaline solution, but the results in perchloric acid sug-
gest that Pd(1 0 0) is the most active facet [20]. The lowest activity
of spherical PdNPs may also be related to the particle size effect,
as it has been shown that in alkaline solution the specific activity
of Pd particles towards the ORR decreases by a factor of about 3
with decreasing the particle size from 16.7 to 3 nm [24]. This has
been attributed to the stronger adsorption of OH on smaller par-
ticles that block the active reaction sites [24]; similar explanation
has been given in case of Pt catalysts in alkaline solution [54–56].
The lowest OH coverage on Pd nanocubes has been also proposed
as the reason of their high activity in HClO4 solution [6]. However,
it has been shown recently that in case of Pd single crystal elec-
trodes in HClO4, the oxide coverage is not relevant to the ORR and
the activity depends only on the width of terraces, Pd(1 0 0) being
the most active site [20,21]. In addition, for nanostructured Pd elec-
trodes in 0.1 M KOH, no particle size dependence of the ORR was
observed [12]. Therefore, it is most likely that the high activity of
Pd nanocubes is due to the prevalence of Pd(1 0 0) surface sites, but
it is not entirely clear at this stage of work, what is the reason of
the high activity of Pd(1 0 0) in alkaline solutions.
Fig. 9. A comparison of RDE voltammetry curves on PdNP modified GC electrodes
(Ar = 0.64 cm2 and Ar = 0.91 cm2 for cubic and spherical PdNPs, respectively) and bulk
Pd (Ar = 0.67 cm2) in O2-saturated 0.1 M KOH. ω = 1900 rpm, ꢂ = 10 mV s−1
.
determined for cubic as well as spherical PdNPs and bulk Pd over
the whole range of potentials, using the following values of diffu-
sion coefficient and solubility of oxygen: DO = 1.9 × 10−5 cm2 s−1
2
and COb = 1.2 × 10−6 mol cm−3 [53]. The 4e− reduction of O2 in
2
alkaline solutions has been also observed on Pd nanoparticles in
earlier studies [24,25,28], however, it has been proposed that on
Pt-group metals this reaction proceeds at least partly via peroxide
intermediate [4].
In Fig. 9, the data at a single electrode rotation rate are compared
for PdNPs and bulk Pd. The onset potential of O2 reduction is the
highest for Pd nanocubes and the sharpest current increase was
observed for bulk Pd. The mass-transfer corrected Tafel plots were
constructed from these data (Fig. 10) and the values of Tafel slope
were determined (Table 2). At low current densities, Tafel slopes are
similar for PdNPs and bulk Pd and slightly higher than −60 mV that
is usually observed on Pd nanocatalysts [24,25,29]. For PdNPs, the
slope value gradually increases at more negative potentials, up to
about −180 mV. For bulk Pd, however, the slope is rather constant
in the range of Tafel analysis, similarly to the results obtained by
Vracar et al. in strongly alkaline solutions [3]. On Pt electrodes, two
Tafel slope values (−60 and −120 mV) have been confirmed in many
researches and it has been suggested that the ORR mechanism is
the same on Pd [4].
4. Conclusions
The cubic palladium nanoparticles synthesised in this work
showed enhanced electrocatalytic activity towards electroreduc-
tion of oxygen, as compared to spherical Pd nanoparticles or bulk
Pd. This effect was observed in acidic as well as in alkaline solution
and is most probably related to the predominance of the Pd(1 0 0)
surface sites on Pd nanocubes. The four-electron reduction of oxy-
gen to water was observed on Pd catalysts studied and the Tafel
analysis revealed that the mechanism of O2 reduction is the same
on Pd nanoparticles as on bulk palladium or platinum in acidic solu-
tion. The shape-controlled synthesis of metal nanoparticles enables
to design more active catalysts for O2 reduction that have potential
applications in low temperature fuel cells.
Acknowledgements
This research was supported by the Estonian Science Founda-
tion (Grant No. 8380). Partial financial support by the CRDF-ETF
grant (Project No. ESC2-2975-TR-09) is gratefully acknowledged.
This work has been also financially supported by the MICINN of
Spain through the project CTQ2010-16271 (FEDER).
References
Fig. 10. Mass-transfer corrected Tafel plots for oxygen reduction on PdNP modified
GC electrodes and bulk Pd in 0.1 M KOH. ω = 1900 rpm.
[1] M. Shao, J. Power Sources 196 (2011) 2433.
[2] E. Antolini, Energy Environ. Sci. 2 (2009) 915.

H. Erikson et al. / Electrochimica Acta 59 (2012) 329–335
335
[3] L.M. Vracar, D.B. Sepa, A. Damjanovic, J. Electrochem. Soc. 133 (1986) 1835.
[4] J.S. Spendelow, A. Wieckowski, Phys. Chem. Chem. Phys. 9 (2007) 2654.
[5] J.J. Salvador-Pascual, S. Citalan-Cigarroa, O. Solorza-Feria, J. Power Sources 172
(2007) 229.
[6] M.H. Shao, T.Y. Yu, J.H. Odell, M.S. Jin, Y.N. Xia, Chem. Commun. 47 (2011) 6566.
[7] H. Erikson, A. Sarapuu, K. Tammeveski, J. Solla-Gullon, J.M. Feliu, Electrochem.
Commun. 13 (2011) 734.
[8] M.H. Shao, K. Sasaki, R.R. Adzic, J. Am. Chem. Soc. 128 (2006) 3526.
[9] J.H. Yang, J.Y. Lee, Q.B. Zhang, W.J. Zhou, Z.L. Liu, J. Electrochem. Soc. 155 (2008)
B776.
[10] L. Xiao, L. Zhuang, Y. Liu, J.T. Lu, H.D. Abruna, J. Am. Chem. Soc. 131 (2009) 602.
[11] A. Sarapuu, A. Kasikov, N. Wong, C.A. Lucas, G. Sedghi, R.J. Nichols, K. Tam-
meveski, Electrochim. Acta 55 (2010) 6768.
[30] W.X. Niu, Z.Y. Li, L.H. Shi, X.Q. Liu, H.J. Li, S. Han, J. Chen, G.B. Xu, Cryst. Growth
Des. 8 (2008) 4440.
[31] M.C. Daniel, D. Astruc, Chem. Rev. 104 (2004) 293.
[32] J. Solla-Gullon, V. Montiel, A. Aldaz, J. Clavilier, J. Electroanal. Chem. 491 (2000)
69.
[33] W.X. Niu, L. Zhang, G.B. Xu, ACS Nano 4 (2010) 1987.
[34] L. Zhang, W.X. Niu, G.B. Xu, Nanoscale 3 (2011) 678.
[35] F.R. Fan, A. Attia, U.K. Sur, J.B. Chen, Z.X. Xie, J.F. Li, B. Ren, Z.Q. Tian, Cryst.
Growth Des. 9 (2009) 2335.
[36] Y.C. Yu, Y.X. Zhao, T. Huang, H.F. Liu, Mater. Res. Bull. 45 (2010) 159.
[37] N. Hoshi, K. Kagaya, Y. Hori, J. Electroanal. Chem. 485 (2000) 55.
[38] N. Hoshi, M. Kuroda, Y. Hori, J. Electroanal. Chem. 521 (2002) 155.
[39] N. Hoshi, M. Nakamura, N. Maki, S. Yamaguchi, A. Kitajima, J. Electroanal. Chem.
624 (2008) 134.
[12] H. Erikson, A. Kasikov, C. Johans, K. Kontturi, K. Tammeveski, A. Sarapuu, J.
Electroanal. Chem. 652 (2011) 1.
[13] S. Chakraborty, C.R. Raj, Carbon 48 (2010) 3242.
[40] M. Grden, M. Lukaszewski, G. Jerkiewicz, A. Czerwinski, Electrochim. Acta 53
(2008) 7583.
[14] J.S. Ye, Y.C. Bai, W.D. Zhang, Microchim. Acta 165 (2009) 361.
[15] Y.H. Lin, X.L. Cui, X.R. Ye, Electrochem. Commun. 7 (2005) 267.
[16] S. Takenaka, N. Susuki, H. Miyamoto, E. Tanabe, H. Matsune, M. Kishida, Chem.
Commun. 46 (2010) 8950.
[41] R. Woods, in: A.J. Bard (Ed.), Electroanalytical Chemistry, vol. 9, Marcel Dekker,
New York, 1976, p. 1.
[42] A.J. Bard, L.R. Faulkner, Electrochemical Methods, 2nd ed., Wiley, New York,
2001.
[17] S. Takenaka, N. Susuki, H. Miyamoto, E. Tanabe, H. Matsune, M. Kishida, J. Catal.
279 (2011) 381.
[18] S.M. Senthil Kumar, J. Soler Herrero, S. Irusta, K. Scott, J. Electroanal. Chem. 647
(2010) 211.
[19] N. Alexeyeva, A. Sarapuu, K. Tammeveski, F.J. Vidal-Iglesias, J. Solla-Gullon, J.M.
Feliu, Electrochim. Acta 56 (2011) 6702.
[20] S. Kondo, M. Nakamura, N. Maki, N. Hoshi, J. Phys. Chem. C 113 (2009) 12625.
[21] A. Hitotsuyanagi, S. Kondo, M. Nakamura, N. Hoshi, J. Electroanal. Chem. 657
(2011) 123.
[43] R.R. Adzic, J. Wang, B.M. Ocko, Electrochim. Acta 40 (1995) 83.
[44] D.R. Lide (Ed.), CRC Handbook of Chemistry and Physics, CRC Press, Boca Raton,
2001.
[45] D.X. Cao, L.M. Sun, G.L. Wang, Y.Z. Lv, M.L. Zhang, J. Electroanal. Chem. 621
(2008) 31.
[46] J.X. Wang, N.M. Markovic, R.R. Adzic, J. Phys. Chem. B 108 (2004) 4127.
[47] N. Hoshi, M. Kuroda, O. Koga, Y. Hori, J. Phys. Chem. B 106 (2002) 9107.
[48] Y. Garsany, O.A. Baturina, K.E. Swider-Lyons, S.S. Kocha, Anal. Chem. 82 (2010)
6321.
[22] Y.F. Yang, Y.H. Zhou, C.S. Cha, Electrochim. Acta 40 (1995) 2579.
[23] F.H.B. Lima, J. Zhang, M.H. Shao, K. Sasaki, M.B. Vukmirovic, E.A. Ticianelli, R.R.
Adzic, J. Phys. Chem. C 111 (2007) 404.
[24] L. Jiang, A. Hsu, D. Chu, R. Chen, J. Electrochem. Soc. 156 (2009) B643.
[25] L. Jiang, A. Hsu, D. Chu, R. Chen, J. Electrochem. Soc. 156 (2009) B370.
[26] T.G. Nikiforova, Y.V. Kabeneva, O.A. Runova, Russ. J. Appl. Chem. 83 (2010) 1001.
[27] F.H.B. Lima, J. Zhang, M.H. Shao, K. Sasaki, M.B. Vukmirovic, E.A. Ticianelli, R.R.
Adzic, J. Solid State Electrochem. 12 (2008) 399.
[49] C. Wang, H. Daimon, Y. Lee, J. Kim, S. Sun, J. Am. Chem. Soc. 129 (2007) 6974.
[50] M. Inaba, M. Ando, A. Hatanaka, A. Nomoto, K. Matsuzawa, A. Tasaka, T. Kinu-
moto, Y. Iriyama, Z. Ogumi, Electrochim. Acta 52 (2006) 1632.
[51] G.F. Alvarez, M. Mamlouk, S.M. Senthil Kumar, K. Scott, J. Appl. Electrochem. 41
(2011) 925.
[52] M.R. Tarasevich, Elektrokhimiya 9 (1973) 599.
[53] R.E. Davis, G.L. Horvath, C.W. Tobias, Electrochim. Acta 12 (1967) 287.
[54] J. Perez, E.R. Gonzalez, E.A. Ticianelli, Electrochim. Acta 44 (1998) 1329.
[55] L. Genies, R. Faure, R. Durand, Electrochim. Acta 44 (1998) 1317.
[56] K. Tammeveski, T. Tenno, J. Claret, C. Ferrater, Electrochim. Acta 42 (1997) 893.
[28] M.H. Seo, S.M. Choi, H.J. Kim, W.B. Kim, Electrochem. Commun. 13 (2011) 182.
[29] M.H. Shao, K. Sasaki, P. Liu, R.R. Adzic, Z. Phys. Chem. 221 (2007) 1175.