The initial stages of palladium deposition onto Pt(1 1 1)

Article abstract of DOI:10.1016/j.electacta.2003.07.008

The electrodeposition of Pd onto Pt(1 1 1) from PdCl₂ and PdSO₄ containing solutions was studied by cyclic voltammetry and in situ scanning tunnelling microscopy. Pd deposition starts by forming a pseudomorphic monolayer in both cases. While in the presence of chloride this monolayer is deposited at underpotentials, its formation in chloride-free solution is kinetically hindered to such an extent that the deposition peak is shifted negative of the equilibrium potential. Detailed structure information has been obtained from STM data about the Pd layer as well as the co-adsorbed anions. Bulk deposition from the chloride-containing solution proceeds via a quasi layer-by-layer fashion. However, the particular electrochemical properties of the first Pd monolayer disappear only after deposition of the equivalent of four more Pd layers. The electrochemical behaviour of such films is similar to that of a rough Pd(1 1 1) surface. Pd bulk deposition from chloride-free solution leads to the formation of three-dimensional clusters from the very beginning. About 10 ML equivalents are needed in that case to completely cover the first Pd monolayer.

Full text of DOI:10.1016/j.electacta.2003.07.008

Electrochimica Acta 49 (2003) 63–72
The initial stages of palladium deposition onto Pt(1 1 1)
R. Hoyer, L.A. Kibler∗,1, D.M. Kolb1
Department of Electrochemistry, University of Ulm, 89069 Ulm, Germany
Received 26 May 2003; accepted 16 July 2003
Abstract
The electrodeposition of Pd onto Pt(1 1 1) from PdCl2 and PdSO4 containing solutions was studied by cyclic voltammetry and in situ
scanning tunnelling microscopy. Pd deposition starts by forming a pseudomorphic monolayer in both cases. While in the presence of chloride
this monolayer is deposited at underpotentials, its formation in chloride-free solution is kinetically hindered to such an extent that the deposition
peak is shifted negative of the equilibrium potential. Detailed structure information has been obtained from STM data about the Pd layer as
well as the co-adsorbed anions. Bulk deposition from the chloride-containing solution proceeds via a quasi layer-by-layer fashion. However,
the particular electrochemical properties of the first Pd monolayer disappear only after deposition of the equivalent of four more Pd layers.
The electrochemical behaviour of such films is similar to that of a rough Pd(1 1 1) surface. Pd bulk deposition from chloride-free solution
leads to the formation of three-dimensional clusters from the very beginning. About 10 ML equivalents are needed in that case to completely
cover the first Pd monolayer.
© 2003 Elsevier Ltd. All rights reserved.
Keywords: Pt(1 1 1); Palladium; Scanning tunnelling microscopy; Cyclic voltammetry; Growth; Mechanism; Surface structure; Metal deposition
1. Introduction
by measuring the oxidation of small organic molecules
like carbon monoxide [12,13] and formic acid [4,14], the
oxygen reduction [15] and the hydrogen evolution and oxi-
dation reaction [16]. Since the activity of the Pd overlayers
depends strongly on the surface structure, studies on the
growth mechanism and a structural characterisation on an
atomic level are essential to develop a better understanding.
In addition, fundamental electrochemical properties like
anion adsorption or the potential of zero total charge as
a function of the Pd coverage have been studied [12,13].
Similar experiments have been also performed for Pd on
Au single crystal electrodes [17,18].
STM measurements for Pd deposition onto Au(1 1 1) from
H2PdCl4 solution revealed for low overpotentials that nucle-
ation starts exclusively at the lower part of the step edges.
The first layer covers the whole surface exhibiting the same
morphology like the substrate. The second layer is growing
in the same way, but the third layer already shows dendritic
growth [17]. For deposition from chloride-free solution, al-
ready the second layer has a high number of defects while
the following layers clearly grow in a 3D fashion [19].
The electrochemical behaviour of submonolayer amounts
of Pd on Pt(1 1 1) has given indirect evidence for an island
growth [12,13]. These islands are monoatomic high, leaving
the long-range order for the sulphate adlayer on uncovered
Thin noble-metal overlayers on well-ordered single crys-
tal surfaces have been investigated intensively in recent years
because they represent interesting systems for surface sci-
ence and electrocatalysis (see, for example [1–4]). Further-
more, they were considered to be ideal substitutes for the
respective bulk materials, which cannot be easily prepared
by conventional flame-annealing. Particularly, palladium has
attracted great interest because the use of thin films provides
the opportunity to study the electrochemical behaviour of Pd
without the problem of massive hydrogen absorption [5].
Systematic studies have shown that metal deposition onto
a foreign metal substrate allows to modify and to fine-tune
the physical and chemical properties of the surface. Pd
deposition onto Pt single crystal surfaces has been carried
out by chemical vapour deposition in ultra-high vacuum
(uhv) [6,7], irreversible adsorption [8,9], forced deposi-
tion [10,11] or by electrochemical deposition [4,11]. These
Pd films have been checked for their catalytic properties
∗
Corresponding author. Tel.: +49-731-5025419;
fax: +49-731-5025409.
E-mail address: ludwig.kibler@chemie.uni-ulm.de (L.A. Kibler).
1
ISE members.
0013-4686/$ – see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2003.07.008

64
R. Hoyer et al. / Electrochimica Acta 49 (2003) 63–72
Pt sites, and they grow pseudomorphic [6,7,9,11,16,20,21].
In sulphuric acid solutions, hydrogen adsorption–desorption
and (bi)sulphate desorption–adsorption on the Pd islands
give rise to a sharp voltammetric peak at 0.23 V versus RHE.
Until the monolayer is completed, the characteristic butterfly
feature for Pt(1 1 1) surface is still visible [12,13]. As soon as
a second Pd layer is formed, another peak emerges at 0.28 V,
corresponding to adsorption on bulk Pd. However, the cor-
rect assignment of this peak is still under debate. Attard et al.
[7] have vapour-deposited Pd films on Pt(1 1 1) in UHV and
they correlated LEED patterns with the corresponding cyclic
voltammogram (CV). Since they found a (1×1) LEED pat-
tern for the whole coverage region, they concluded a pseu-
domorphic growth and therefore assigned the peak in the CV
to hydrogen and anion adsorption–desorption on the terraces
of the Pd multilayers. In a recent paper by Ball et al. [22]
on electrochemically deposited Pd films on Pt(1 1 1), which
were studied by X-ray scattering (SXS), a so-called pseudo-
morphic Stranski–Krastanov (SK) growth mode (3D growth
on a pseudomorphic monolayer) was reported. This led to the
conclusion that the voltammetric peak at 0.28 V is related to
hydrogen adsorption–desorption combined with (bi)sulphate
desorption–adsorption at step sites rather than terraces sites.
Since scanning tunnelling microscopy (STM) can provide
structural information in real space and on an atomic level,
we applied this meted to study the electrochemical Pd depo-
sition onto Pt(1 1 1) in the presence of two different anions.
A correlation between electrochemical behaviour and sur-
face structure is attempted by comparing CV and STM data.
In addition, the reactivity of the Pd overlayers for CO ad-
layer oxidation has been investigated and is correlated with
that of bulk palladium.
The STM images were recorded with a Topometrix TMX
Discoverer 2010. W and Pt/Ir (80/20) tips were prepared
by etching a 0.25 mm wire in aqueous NaOH or NaCN, re-
spectively, and by coating the tips with an electrophoretic
paint to reduce the Faradaic current at the tip/electrolyte in-
terface. Two platinum wires were used as counter and refer-
ence electrodes in the STM cell, while a saturated calomel
electrode (SCE) was employed for cyclic voltammetry. In
the following all potentials are quoted against SCE.
All electrolytes were prepared from H2SO4 (Merck,
suprapur), HCl (Merck, suprapur), PdSO4 (Alfa Aesar),
PdCl2 (Aldrich) and ultra-pure water (USF and Satorius,
18.2 Mꢀ cm, <2 ppb TOC). The solutions were de-aerated
with nitrogen 6.0 (MTI IndustrieGase AG). For the experi-
ments with carbon monoxide, CO 4.7 (MTI IndustrieGase
AG) was used. All experiments were carried out at room
temperature.
3. Results and discussion
3.1. Electrochemical Pd deposition
Cyclic voltammograms for Pd deposition onto Pt(1 1 1)
from sulphate and chloride containing solutions are shown in
Fig. 1. The freshly prepared Pt electrode was immersed into
the Pd2+ containing solution at 0.7 and 0.75 V, respectively,
where no Pd deposition occurred. For Pd deposition from
PdSO4, a peak at 0.45 V is observed (Fig. 1a) with a charge
of about 480 ␮C/cm2, which corresponds to roughly 1 ML.
Since this potential value is negative of the equilibrium po-
tential for bulk Pd, which was measured to be at 0.52 V in
that solution, there is no direct evidence for under-potential
deposition (upd) in this system. However, Pd deposition in
the absence of chloride is kinetically strongly hindered and
for this reason the deposition peak may be shifted towards
more negative potentials. The onset of bulk deposition can be
seen in Fig. 1a around 0.4 V, but the deposit, instead of being
dissolved in the anodic scan, simply forms a passivating ox-
ide layer at positive potentials. In a chloride-containing solu-
tion, the kinetic hindrance for Pd deposition is reduced and
the reaction becomes less irreversible. The cyclic voltammo-
gram in Fig. 1b shows a cathodic peak at 0.53 V, i.e. in the
upd region, the charge of which again corresponds roughly
to 1 Pd ML. On the reverse scan, dissolution of bulk pal-
ladium around 0.53 V is followed by a peak at about 0.7 V,
which can be assigned to the dissolution of the Pd mono-
layer, accompanied by the beginning oxidation of the Pt sur-
face. The initial state of the substrate is not easily recovered
due to incomplete Pd dissolution and structural changes due
to surface oxidation and reduction.
2. Experimental
Single crystal beads of platinum and palladium were pre-
pared by melting one end of a 0.5 mm wire [23,24]. On
the metal beads facets of (1 1 1) orientation could be easily
found, which were directly used for STM measurements.
Prior to each measurement the small crystals were annealed
in a hydrogen flame, cooled in argon and quenched in wa-
ter in the case of palladium or cooled in a H2 + Ar mixture
and quenched in H2 saturated water in the case of platinum.
The crystals were transferred to the STM cell protected by a
droplet of water. The platinum single crystal used for cyclic
voltammetry was a cylindrical disc with a diameter of 4 mm
(MaTecK, Jülich, Germany), which was oriented better than
1◦ and polished down to 0.03 ␮m.
This electrode was flame-annealed before each measure-
ment and slowly cooled down in CO for 10 min. The ad-
sorbed CO prevented contamination of the surface during
the transfer through air to the electrochemical cell. The CO
adlayer was stripped in Pd2+-free 0.1 M H2SO4, where both
the cleanliness of the system and the quality of the Pt(1 1 1)
surface was checked by recording a cyclic voltammogram.
3.2. In situ STM measurements
The initial stages of the deposition process were recorded
by stepwise decreasing the potential from 0.6 V until the

R. Hoyer et al. / Electrochimica Acta 49 (2003) 63–72
65
Fig. 1. Cyclic voltammograms (first cycle each) for Pd deposition onto Pt(1 1 1) from (a) 0.1 M H2SO4 + 0.1 mM PdSO4 (scan rate: 0.1 mV/s) and (b)
0.1 M H2SO4 + 0.1 mM H2PdCl4 (scan rate: 1 mV/s).
deposition started. Fig. 2 shows a sequence of STM images
for the deposition of the first Pd monolayer onto Pt(1 1 1)
from 0.1 M H2SO4 + 0.1 mM H2PdCl4, beginning with
the bare surface at +0.6 V. Unfortunately, it was not pos-
sible to image [PdCl4]2− on the bare Pt(1 1 1) as in the
case of Au(1 1 1), despite similar experimental conditions
[17]. Possibly, the [PdCl4]2− anion is not adsorbed in this
potential region, because some oxygen species are still
interacting with the Pt(1 1 1) surface. This assumption is
supported by the fact that on Pt(1 1 1) the ordered sulphate
adlayer has only been imaged in the potential region from
0.2 to 0.4 V versus SCE, although surface oxidation occurs
at much more positive potentials [25,26]. After stepping the
potential to 0.55 V, Pd deposition starts by forming small
Fig. 2. (a–d) Sequence of STM images showing the formation of a Pd monolayer on Pt(1 1 1) in 0.1 M H2SO4 + 0.1 mM H2PdCl4 after stepping the
potential from +0.6 to +0.55 V. The hexagonal structure of the [PdCl4]2− adlayer on the Pd monolayer is shown in image (d).

66
R. Hoyer et al. / Electrochimica Acta 49 (2003) 63–72
monoatomic high islands, both at the steps as well as on
the terraces (Fig. 2b). This confirms previous results, which
showed that Pd forms dense islands and is not distributed
homogenously over the surface as in the case of Cu [12,27].
The STM image in Fig. 2c still shows some vacancies in the
Pd monolayer. More than 30 min are needed until the islands
grow together and form a smooth Pd monolayer, completely
covering the Pt(1 1 1) substrate. A hexagonal superstructure
on top of this first Pd monolayer is visible in Fig. 2d, with
different rotational domains. The distance between neigh-
bouring spots is about 0.8 nm. This is similar to the distances
found in the ordered structures of [PdCl4]2− adsorbed on
Au(1 1 1) and on the first Pd layer on Au(1 1 1) [28]. Hence,
it is concluded that the structure in Fig. 2d can also be at-
tributed to adsorbed [PdCl4]2− on the Pd monolayer. Unfor-
tunately, an exact determination of the adsorbate structure
was not possible, due to the lack of an internal calibration
(e.g. via atomically resolved images of the bare substrate).
After decreasing the potential to 0.5 V, Pd bulk deposi-
tion is initiated as shown in the sequence of STM images
in Fig. 3. On the smooth upd layer (Fig. 3a), Pd forms
monoatomic high islands with a dendritic shape nucleating
at steps as well as at defects on the terraces. In analogy to
Au(1 1 1), a complete second Pd layer can be formed [17]
except for a few holes (Fig. 3c). The third layer then grows
in the same fashion, but the fourth layer starts to grow be-
fore completion of the third layer. Thus, with the increasing
number of layers the number of defects is slightly increas-
ing, leading to a relatively flat surface with a fractal-like
shape. This is shown is Fig. 3d, where the fourth Pd layer
can be seen on top as bright islands.
When using PdSO4, deposition of the first Pd layer starts
at 0.47 V in the STM experiment (Fig. 4). As mentioned
above, no ordered adion structure can be observed in sul-
phuric acid solution at potentials positive of 0.4 V, which
means that the initial state of the Pt(1 1 1) surface is likely
to be the same in the presence and in the absence of chlo-
ride. This may explain why deposition of the first Pd layer
looks similar in both cases. Again, small monoatomic high
Pd islands nucleate on the surface with no noticeable pref-
erence of nucleation sites. In Fig. 4c, the first monolayer is
almost completed, but still having some holes in the layer.
A full Pd monolayer without detectable holes can be formed
only at long deposition times as in the case of deposition
from chloride-containing solution. The high-resolution im-
age in Fig. 4d reveals the presence of an ordered sulphate
adlayer on the Pd monolayer. Sulphate commonly forms a
√
√
( 3 × 7)R19.1◦ structure on the (1 1 1) faces of noble
Fig. 3. Sequence of STM images showing Pd deposition on Pt(1 1 1) from 0.1 M H2SO4 + 0.1 mM H2PdCl4. (a) Completion of the first layer; (b) growth
of the second layer; (c) completed second layer; (d) Pt(1 1 1) covered by an equivalent of almost 4 Pd ML.

R. Hoyer et al. / Electrochimica Acta 49 (2003) 63–72
67
Fig. 4. STM images of Pt(1 1 1) in 0.1 mM PdSO4 + 0.1 M H2SO4 showing the formation of the first Pd monolayer (a–c). Image (d) shows the
√
√
(
3 × 7)R19.1◦ structure for adsorbed sulphate on the first Pd monolayer on Pt(1 1 1) (ESCE = +0.47 V).
metals (Au [29,30], Pd [31], Pt [25,26], Rh [32], Ir [33],
and Ru [34]), which is also the case for the Pd monolayer
on Pt(1 1 1). It is well accepted that the first Pd layer grows
pseudomorphic on Pt(1 1 1) [6,7,9,11,16,20,21]. Since the
difference in the next-neighbour distances for Pd (0.275 nm)
and for Pt (0.277 nm) lies within the experimental error of
the STM measurement, the pseudomorphy cannot be proven
here. Therefore, we also can not determine how many lay-
ers grow pseudomorphic, and when the Pd distance relaxes
into the Pd bulk value.
sulphate structure. It turned out, that the morphology of the
Pd clusters is very sensitive to the deposition condition, i.e.
in other experiments the clusters were too small to observe
an ordered sulphate adlayer. Bulk Pd deposition continues
to proceed in a 3D way, leading to a rough surface with
the first Pd monolayer still visible (black parts in Fig. 5f).
The Pd deposition onto Pt(1 1 1) in a PdSO4 electrolyte can
therefore be described by a SK growth mode.
3.3. Electrochemical behaviour of thin palladium films
In order to monitor the Pd bulk deposition, the potential
was set to 0.41 V in the chloride-free 0.1 mM PdSO4+0.1 M
H2SO4 solution (Fig. 5). In contrast to Pd deposition from
a chloride-containing solution, Pd forms small 3D clusters,
which are one to four layers high, instead of monoatomic
high islands. The clusters nucleate preferentially at steps,
decorating them with a row of Pd clusters (Fig. 5b). Thereby,
the upper and the lower part of the step appear to be cov-
ered by palladium. Simultaneously, the defects in the first
Pd monolayer seem to be covered by Pd clusters as well.
For longer deposition times, the Pd clusters cover more and
more of the terraces (Fig. 5e). In this case, the plateaus on
top of the clusters were big enough to image the ordered
3.3.1. A comparison between deposits from chloride and
sulphate containing solution
The electrochemical behaviour of thin Pd films on
Pt(1 1 1) has been intensively studied, especially in the
sub-monolayer regime [12,13,35,36], but also higher Pd
coverages have been investigated [4,7,9,37]. For the lat-
ter, two characteristic sharp voltammetric peaks have been
observed in sulphuric acid solution. This can also be seen
in Fig. 6, where the cyclic voltammograms for 3 and 5
ML equivalents Pd on Pt(1 1 1), deposited from chloride
(a, c) and sulphate (b, d) containing solutions, are com-
pared. The Pd deposition was carried out in the metal-ion

68
R. Hoyer et al. / Electrochimica Acta 49 (2003) 63–72
Fig. 5. Sequence of six STM images (a–f) showing Pd bulk deposition from 0.1 M H2SO4 +0.1 mM PdSO4 onto the first (upd) Pd monolayer on Pt(1 1 1).
containing solution by scanning the potential with 1 mV/s
to a certain value, at which Pd was deposited, and by mon-
itoring the charge flow. After a certain charge was reached
(480 ␮C/cm2 for 1 ML Pd), the electrode was removed from
solution, rinsed with water and immersed into a second cell,
containing 0.1 M H2SO4, for characterisation.
The cyclic voltammograms clearly show two sharp peaks
at −0.08 and −0.03 V independent of the deposition solu-
tion. The peak at −0.08 V is due to hydrogen and sulphate
adsorption–desorption on the first Pd monolayer, which is
structurally and electronically influenced by the Pt sub-
strate, leading to a negative potential shift with respect to
bulk Pd [18]. The potential of the more positive peak agrees
well with that for hydrogen adsorption on massive Pd(1 1 1)
[18]. The peak at −0.03 V was therefore related to hydro-
gen and sulphate adsorption–desorption on the second and
following Pd layers, which should exhibit Pd bulk proper-
ties. From the intensity of the Pd monolayer and multilayer
current peaks (at −0.08 and −0.03 V) in Fig. 6, a much
more uniform film growth can be inferred for deposits ob-
tained in chloride-containing solutions. For example, for 5
ML Pd on Pt(1 1 1) the current peak at −0.08 V for the first
layer has disappeared almost completely (Fig. 6c), whereas
an equivalent of 10 ML is required for deposition from
sulphate solutions to fully burry the Pd upd layer (dashed
curve in Fig. 6d). Likewise, a 3 ML deposit will cover
either 80 or 60% of the substrate surface, depending on
the presence or absence of Cl− (Fig. 6a and b). The much

R. Hoyer et al. / Electrochimica Acta 49 (2003) 63–72
69
Fig. 6. Cyclic voltammograms for thin Pd films on Pt(1 1 1) in 0.1 M H2SO4, deposited from chloride and sulphate containing solutions in various
amounts: (b, d) 0.1 M H2SO4 + 0.1 mM PdSO4; (a, c) 0.1 M H2SO4 + 0.1 mM H2PdCl4 (scan rate: 10 mV/s).
more uniform film growth in chloride solutions is directly
evidenced by STM measurements, which show rather large
terraces in the corresponding images. Furthermore, we re-
voltammograms for the oxidation of CO adlayers on bare
Pt(1 1 1), on Pd films on Pt(1 1 1) for various coverages and
on a massive Pd(1 1 1) electrode are shown in Fig. 7. The
oxidation peak for adsorbed CO on bare Pt(1 1 1) is located
at about 0.45 V (Fig. 7a). The subsequent cycle shows the
well-known butterfly feature of Pt(1 1 1) in sulphuric acid
[39]. In the presence of Pd on the Pt(1 1 1) surface, the CO
oxidation peak is shifted to more positive potentials, between
0.52 and 0.6 V (Fig. 7b–d). The observation of features in
the cyclic voltammogram that are characteristic of Pt(1 1 1)
up to a monolayer of Pd was interpreted as an indication
for the formation of monoatomic high Pd islands [12,13]. It
is also evident from STM images shown above that Pd nu-
cleates at step edges and thus blocks the Pt steps, which are
crucial for the CO oxidation [38]. For a Pd coverage in the
submonolayer regime (see Fig. 7b), the peak for the CO oxi-
dation on uncovered Pt(1 1 1) is already absent, which means
that the Pt step sites are no longer available. This raises the
question about the sites that are responsible for the elec-
trooxidation of CO at 0.52 V. Those can be either Pt(1 1 1)
terraces sites or the newly created Pd steps, including the
rims of the pseudomorphic Pd islands. We should mention
here that the anodic current peak around 0.8 V is due to sur-
face oxidation of Pd. The height of the CO oxidation peak
has a minimum for 1 ML Pd on Pt(1 1 1) (Fig. 7c), suggest-
ing slow kinetics under these conditions. Since the complete
Pd monolayer possesses exactly the same morphology as
the substrate, it is obvious that the defect density is lowest
√
√
call that the ( 3 × 7)R19.1◦ structure of the sulphate
adlayer is sensitive to long-range order of the substrate. We
therefore assume that the peak at −0.03 V is due to hy-
drogen and sulphate adsorption–desorption on the terraces
of the bulk deposit. This is in contrast to the work of Ball
et al. [22] who suggested that adsorption and desorption at
step sites of the Pd clusters lead to the second peak in the
CV. With the assignment of the sharp peaks at −0.08 and
−0.03 V to adsorption on (1 1 1) terraces sites of mono-
and multilayer Pd deposits, cyclic voltammetry can be used
for structural characterisation. We mention in passing that
the charge under both peaks in the anodic scan amounts to
about 130 ␮C/cm2, not taking the background into account
(see Fig. 6a). This is equal within experimental error to the
charge, which is measured under the peak of the Pd mono-
layer (see, e.g. Fig. 7c). This value differs significantly from
the 320 ␮C/cm2, reported by Álvarez et al. [12], the reason
for the discrepancy being unclear to us.
3.3.2. CO oxidation on thin Pd films
The oxidation of an irreversibly adsorbed CO adlayer
has not only been used as a model reaction to measure the
activity of a certain surface, but it also served as a probe for
the surface structure since oxidative CO stripping is highly
influenced by the surface defect density [18,38]. Cyclic

70
R. Hoyer et al. / Electrochimica Acta 49 (2003) 63–72
(a)
(b)
(c)
200
100
0
Pt(111)
-100
-200
1 ML Pd
0.5 ML Pd
-0.4
0.0
0.4
0.8
ESCE / V
(d)
(e)
200
100
0
Pd(111)
5 ML Pd
well-ordered
perturbated
-100
-200
-0.4
0.0
0.4
ESCE / V
0.8 -0.4
0.0
0.4
ESCE / V
0.8
Fig. 7. Cyclic voltammograms (1st and 2nd cycle) in 0.1 M H2SO4 for the oxidation of irreversibly adsorbed CO on (a) Pt(1 1 1), (b–(d) on Pd of
different coverages on Pt(1 1 1) and (e) on well-ordered and electrochemically roughened Pd(1 1 1). Scan rate: 10 mV/s.
in that case. Parallel to the minimum for CO oxidation, the
spike for the hydrogen and sulphate adsorption–desorption
on the terraces of the Pd monolayer at −0.08 V reaches a
maximum in current density. Anion adsorption/hydrogen
desorption on the first Pd monolayer layer is shifted by
50 mV to more negative potentials compared to bulk Pd,
probably because of electronic modification of the Pd on
Pt(1 1 1) [40]. When the coverage exceeds 1 ML the peak at
−0.03 V emerges in the CV for 0.1 M H2SO4, which corre-
sponds to hydrogen and sulphate adsorption–desorption on
the terraces of the Pd multilayer deposits (see above). The
peak at −0.08 V for the first Pd layer disappears, when the
latter is fully covered by bulk palladium (Fig. 7d).
cycle (Fig. 8b) and of a 5 ML Pd film on Pt(1 1 1) (Fig. 8c).
The well-ordered Pd(1 1 1) surface exhibits large terraces,
separated by monoatomic high steps, while surface oxida-
tion introduces many holes and islands. The defect densities
of the surfaces in Fig. 8b and c are comparable, hence, thin
Pd overlayers on Pt(1 1 1) can be considered as defect-rich
Pd(1 1 1). For reactions which are not strongly influenced
by surface defects, e.g. formic acid oxidation, the thick
Pd films show properties very similar to Pd(1 1 1) and can
therefore be used as ideal substitutes of massive single
crystals.
When increasing the Pd coverage from 1–5 ML, the poten-
tial for the CO oxidation remains in essence constant, only
the current density increases. From the CV in Fig. 7e, we
see that CO is oxidised at 0.8 V on a well-ordered Pd(1 1 1)
surface, but this potential is shifted by about 0.2 V to more
negative values when the number of surface defects had
been increased by electrochemical oxidation–reduction cy-
cles. This confirms the previously expressed view that CO
oxidation is much faster at defects than on terraces [18]. The
CO oxidation potential for the perturbed Pd(1 1 1) surface
agrees quite well with that for the 5 ML deposit. Therefore,
it can be stated that the peak at 0.59 V in Fig. 7d is indeed
related to CO oxidation at defect sites of the Pd overlayer
having Pd bulk properties.
4. Summary
We have investigated the electrochemical deposition of
Pd onto Pt(1 1 1) by in situ STM. Both in PdSO4 and
H2PdCl4 containing sulphuric acid solutions, an (almost)
complete pseudomorphic monolayer of Pd is formed prior
to bulk deposition. The structure of the deposit depends
strongly on the anions present in the solution. While in
chloride-free solution Pd grows in a 3D fashion leading to a
SK growth mode, addition of chloride yields a more layer-
by-layer type growth, with the Pd layers having a dendritic
shape.
The voltammogram in sulphuric acid for a multilayer de-
posit of Pd on Pt(1 1 1) shows two characteristic peaks in the
hydrogen adsorption region as long as the first Pd monolayer
is not completely covered. The peak at −0.08 V corresponds
to hydrogen adsorption and anion desorption on the terraces
Fig. 8 shows STM images of a freshly prepared,
well-ordered Pd(1 1 1) surface (Fig. 8a), of a Pd(1 1 1) sur-
face which had been subjected to an oxidation–reduction

R. Hoyer et al. / Electrochimica Acta 49 (2003) 63–72
71
required to cover the first Pd monolayer, are in agreement
with a 3D growth of bulk Pd, when deposited from sulphate
solution.
The electrochemical characterisation of thin Pd films de-
posited onto Pt(1 1 1) and of a Pd(1 1 1) surface reveals
great similarities. Above 1 ML, the Pd deposit exhibits bulk
properties. The defect density of thicker Pd films is simi-
lar to that of a Pd(1 1 1) surface, which was subjected to an
oxidation–reduction cycle. This was demonstrated by CO
oxidation experiments.
References
[1] J.A. Rodriguez, Surf. Sci. Rep. 24 (1996) 223.
[2] E. Herrero, L.J. Buller, H.D. Abruna, Chem. Rev. 101 (2001)
1897.
[3] W. Chrzanowski, A. Wieckowski, Langmuir 13 (1997) 5974.
[4] M. Baldauf, D.M. Kolb, J. Phys. Chem. 100 (1996) 11375.
[5] M. Baldauf, D.M. Kolb, Electrochim. Acta 38 (1993) 2145.
[6] M. Han, P. Mrozek, A. Wieckowski, Phys. Rev. B 48 (1993) 8329.
[7] G.A. Attard, R. Price, A. Al-Akl, Electrochim. Acta 39 (1994) 1525.
[8] G.A. Attard, A. Bannister, J. Electroanal. Chem. 300 (1991) 467.
[9] M.J. Llorca, J.M. Feliu, A. Aldaz, J. Electroanal. Chem. 351 (1993)
299.
[10] J. Clavilier, M.J. Llorca, J.M. Feliu, A. Aldaz, J. Electroanal. Chem.
310 (1991) 429.
[11] J. Inukai, M. Ito, J. Electroanal. Chem. 358 (1993) 307.
[12] B. Álvarez, V. Climent, A. Rodes, J.M. Feliu, J. Electroanal. Chem.
497 (2001) 125.
[13] B. Álvarez, V. Climent, A. Rodes, J.M. Feliu, Phys. Chem. Chem.
Phys. 3 (2001) 3269.
[14] M.J. Llorca, J.M. Feliu, A. Aldaz, J. Clavilier, J. Electroanal. Chem.
376 (1994) 151.
[15] V. Climent, N.M. Markovic, P.N. Ross, J. Phys. Chem. B 104 (2000)
3116.
[16] N.M. Markoviæ, C.A. Lucas, V. Climent, V. Stamenkoviæ, P.N. Ross,
Surf. Sci. 465 (2000) 103.
[17] L.A. Kibler, M. Kleinert, R. Randler, D.M. Kolb, Surf. Sci. 443
(1999) 19.
[18] A.M. El-Aziz, L.A. Kibler, J. Electroanal. Chem. 534 (2002) 107.
[19] J. Tang, R. Hoyer, M. Petri, L.A. Kibler, D.M. Kolb, in preparation.
[20] R. Gómez, A. Rodes, J.M. Perez, J.M. Feliu, A. Aldaz, Surf. Sci.
327 (1995) 202.
[21] R. Gómez, A. Rodes, J.M. Perez, J.M. Feliu, A. Aldaz, Surf. Sci.
344 (1995) 85.
[22] M.J. Ball, C.A. Lucas, N. M Markovic, V. Stamenkovic, P.N. Ross,
Surf. Sci. 518 (2002) 201.
[23] J. Clavilier, R. Faure, G. Guinet, R. Durand, J. Electroanal. Chem.
107 (1980) 205.
[24] J. Clavilier, D. Armand, S.G. Sun, M. Petit, J. Electroanal. Chem.
205 (1986) 267.
[25] A.M. Funtikov, U. Linke, U. Stimming, R. Vogel, Surf. Sci. Lett.
324 (1995) L343.
[26] A.M. Funtikov, U. Stimming, R. Vogel, J. Electroanal. Chem. 428
(1997) 147.
Fig. 8. STM images of three differently prepared Pd surfaces in
0.1 M H2SO4. (a) freshly flame-annealed, well-ordered Pd(1 1 1) (in-
set: ( 3 × 7)R19.1◦ sulphate structure), (b) electrochemically rough-
ened Pd(1 1 1) and (c) 5 ml Pd on Pt(1 1 1), deposited from 0.1 mM
H2PdCl4 + 0.1 M H2SO4.
√
√
of the first Pd layer, while the peak at −0.03 V is due to
the adsorption–desorption process on the terraces of the Pd
multilayers. The potentials for both adsorption–desorption
processes do not depend on the deposition solution, whereas
their peak shapes and charges do. The decreasing charge for
hydrogen desorption and anion adsorption on Pd(1 1 1) sites
with increasing Pd coverage and the higher amount of Pd
[27] T. Will, M. Dietterle, D.M. Kolb in: A.A. Gewirth, H. Siegenthaler
(Eds.), Nanoscale Probes of Solid Liquid Interface, Nato ASI Series
288 (1995) 137.
[28] M. Kleinert, Dissertation, Universität Ulm, Ulm, 2001.
[29] O.M. Magnussen, J. Hageböck, J. Hotlos, R.J. Behm, Faraday Discuss
94 (1992) 329.
[30] G.J. Edens, X. Gao, M.J. Weaver, J. Electroanal. Chem. 375 (1994)
357.

72
R. Hoyer et al. / Electrochimica Acta 49 (2003) 63–72
[31] J. Okada, J. Inukai, K. Itaya, Phys. Chem. Chem. Phys. 3 (2001)
[36] M. Arenz, V. Stamenkovic, T.J. Schmidt, K. Wandelt, P.N. Ross,
N.M. Markovic, Surf. Sci. 523 (2003) 199.
3297.
[32] L.-J. Wan, S.-L. Yau, K. Itaya, J. Phys. Chem. 99 (1995) 9507.
[33] L.-J. Wan, M. Hara, J. Inukai, K. Itaya, J. Phys. Chem. B 103 (1999)
6978.
[34] P.-C. Lu, C.-H. Yang, S.-L. Yau, M.-S. Zei, Langmuir 18 (2002)
754.
[37] B. Álvarez, A. Rodes, J.M. Pérez, J.M. Feliu, J. Phys. Chem. 107
(2003) 2018.
[38] N.P. Lebedeva, M.T.M. Koper, J.M. Feliu, R.A. van Santen, J. Phys.
Chem. B 106 (2002) 12938.
[39] J. Clavilier, K. ElAchi, A. Rodes, Chem. Phys. 141 (1990) 1.
[40] A. Ruban, B. Hammer, P. Stoltze, H.L. Skiver, J.K. Norskov, J. Mol.
Cat. A: Chem. 115 (1997) 421.
[35] M. Arenz, V. Stamenkovic, T.J. Schmidt, K. Wandelt, P.N. Ross,
N.M. Markovic, Surf. Sci. 506 (2002) 287.