Electrochemical layer-by-layer growth of palladium on an Au(111) electrode surface: Evidence for important role of adsorbed Pd complex

Article abstract of DOI:10.1021/jp980624f

The adsorption and electrochemical reduction of the tetrachloropalladate complex, PdCl₄^2-, on an Au(111) electrode in H₂SO₄ solution containing PdCl₄^2- was investigated using an electrochemical quartz crystal microbalance (EQCM) and in situ electrochemical scanning tunneling microscopy (STM). The adlayer of PdCl₄^2- with a (√7 × √7)R19.1° structure on the Au(111) substrate was observed by in situ STM measurement, and the amount of adsorbed PdCl₄^2- determined by the EQCM measurement was in good agreement with that estimated from this adlayer structure. In situ STM observation showed also that the electrochemical deposition of palladium proceeded with an epitaxial layer-by-layer growth mode and the PdCl₄^2- complex adsorbed on the deposited palladium layer with the same adlayer structure. The adsorption of the PdCl₄^2- complex seemed to play an important role for the layer-by-layer growth of palladium in a large area, since it inhibits the vertical but favors the lateral growth of the palladium layer. The formation of the Pd(111) bulk phase and the surface structure of Pd(111) was confirmed by X-ray diffraction (XRD) and the underpotential deposition of copper, respectively.

Full text of DOI:10.1021/jp980624f

4366
J. Phys. Chem. B 1998, 102, 4366-4373
Electrochemical Layer-by-Layer Growth of Palladium on an Au(111) Electrode Surface:
Evidence for Important Role of Adsorbed Pd Complex
Hideo Naohara, Shen Ye, and Kohei Uosaki*
Physical Chemistry Laboratory, DiVision of Chemistry, Graduate School of Science, Hokkaido UniVersity,
Sapporo 060-0810, Japan
ReceiVed: January 5, 1998; In Final Form: March 24, 1998
The adsorption and electrochemical reduction of the tetrachloropalladate complex, PdCl42-, on an Au(111)
2-
electrode in H2SO4 solution containing PdCl4 was investigated using an electrochemical quartz crystal
microbalance (EQCM) and in situ electrochemical scanning tunneling microscopy (STM). The adlayer of
PdCl42- with a (x7 × x7)R19.1° structure on the Au(111) substrate was observed by in situ STM measurement,
2-
and the amount of adsorbed PdCl4 determined by the EQCM measurement was in good agreement with
that estimated from this adlayer structure. In situ STM observation showed also that the electrochemical
2-
deposition of palladium proceeded with an epitaxial layer-by-layer growth mode and the PdCl4 complex
2-
adsorbed on the deposited palladium layer with the same adlayer structure. The adsorption of the PdCl4
complex seemed to play an important role for the layer-by-layer growth of palladium in a large area, since
it inhibits the vertical but favors the lateral growth of the palladium layer. The formation of the Pd(111) bulk
phase and the surface structure of Pd(111) was confirmed by X-ray diffraction (XRD) and the underpotential
deposition of copper, respectively.
Introduction
available for in situ observation of so-called underpotentially
deposited (UPD) metal layers such as copper,7 silver,8 lead,9
bismuth,10 and tellurium,11 the investigation of the electrochemi-
cal deposition process with atomic resolution is rather limited
and only limited evidence has been provided for the adsorption
of the reactant,12-15 i.e., metal ions, on the electrode surface
during the electrochemical deposition.
Modifications of the properties of a solid surface by foreign
atoms and molecules that are usually realized by physical vapor
deposition in ultrahigh vacuum (UHV) or electrochemical
plating in a electrolyte solution are very important not only in
fundamental studies but also in industrial applications such as
catalysis, electronics, and corrosion protection. In UHV, the
epitaxial growth of a foreign metal layer on a substrate has been
achieved by vapor deposition, molecular beam epitaxy (MBE).
The structure of the epitaxial layer was analyzed by various
methods in UHV, such as reflection high-energy electron
diffraction (RHEED), low-energy electron diffraction (LEED),
auger electron spectroscopy (AES), and X-ray photoelectron
spectroscopy (XPS). The metal deposition by an electrochemi-
cal method in a solution containing a metal ion or metal complex
should be more economical and convenient than the growth in
UHV because the expensive UHV environment is not necessary.
The structure of the electrodeposited metal layer is known to
be controlled by many factors including deposition potential/
current, concentration of metal ion/complex, nature of the
additives, and surface structure of the substrate. To establish a
method to electrochemically deposit a smooth epitaxial layer
of foreign metals, it is essential to monitor the surface structure
of the electrode during the electrochemical deposition process.
The observation of the structure of the substrate and the metal
adlayer with atomic resolution in an electrochemical environ-
ment is more difficult than that in UHV because of the existence
of the electrolyte solution. Recently, in situ techniques for
surface characterization in an electrolyte solution such as STM,1
atomic force microscopy (AFM),1-3 EQCM,4 and surface X-ray
scattering (SXS)5,6 were applied for the investigation of the
electrochemical deposition process. Although many reports are
One of the most important metals for industrial applications
and fundamental studies is palladium because of its high
catalytic activities for many chemical reactions and ability to
absorb hydrogen.16 The growth of the ultrathin palladium layer
on Au(111) prepared by vapor deposition in UHV was
investigated by AES, LEED, and STM.17-19 The results of the
AES and LEED measurements showed that the palladium has
grown on the Au(111) substrate in an epitaxial layer-by-layer
mode for the first few layers at 150 K and that the surface alloy
between the gold and palladium was formed at more than 500
K.17,18 STM images showed, however, a rough morphology
for the deposited palladium layer in which a simultaneous
evolution of a number of layers was observed when 3.0
monolayers (ML) of palladium was dosed.19 Recently, the
epitaxial growth of a palladium layer on gold single-crystal
electrodes by electrochemical deposition was confirmed by
Baldauf and Kolb with in situ STM measurements.20 The
reaction rates of hydrogen adsorption and formic acid oxidation
were investigated as a function of the thickness of the palladium
layer on the gold substrate. A large difference between the
palladium epitaxial thin layer and a bulk single crystal was
observed,20,21 suggesting an effect of the interaction with the
substrate. However, the observation of electrochemical deposi-
tion of palladium with atomic resolution has not been carried
out and the mechanism for the two-dimensional (2D) growth
of palladium on gold is not clear yet.
2-
In this paper, the adsorption of reactant, i.e., the PdCl4
* Corresponding author. Telephone: +81-11-706-3812. Fax: +81-11-
706-3440. E-mail: uosaki@PCL.sci.hokudai.ac.jp.
complex, and the electrochemical deposition process of the
S1089-5647(98)00624-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 05/07/1998

Growth of Pd on Au(111)
J. Phys. Chem. B, Vol. 102, No. 22, 1998 4367
PdCl42- to palladium on an Au(111) electrode in H2SO4 solution
2-
containing PdCl4 are investigated by in situ electrochemical
STM and EQCM measurements and the important role of the
adsorbed PdCl42- species on the electrode surface for promoting
the electrochemical 2D growth of the palladium layer is
discussed.
Experimental Section
Materials. Electrochemical STM, electrochemical measure-
ments, and XRD measurements were carried out at the surface
of an Au(111) single crystal, which was prepared by Clavilier’s
method.22 It was mechanically polished, annealed by a gas-
oxygen flame for at least 4 h at ca. 800 °C and quenched in
argon-saturated ultrapure water (Milli-Q, Millipore). For the
EQCM measurements, the gold electrode was prepared by the
vacuum evaporation of 10 nm of titanium followed by 150 nm
of gold onto a 5 MHz AT-cut quartz crystal plate (double-side
polished) at 300 °C with an evaporation rate of less than 0.01
nm/s. This procedure is known to provide a highly ordered
Au(111) phase with wide terraces.23 Before each measurement,
the electrode was annealed by a gas-oxygen flame for a few
seconds and then quenched in argon-saturated Milli-Q water.
Electrolyte solutions were prepared using H2SO4 (Suprapure
reagent grade, Wako Pure Chemicals), HClO4 (Suprapure
reagent grade, Wako Pure Chemicals), K2PdCl4 (Reagent grade,
Wako Pure Chemicals), CuSO4 (Reagent grade, Wako Pure
Chemicals), and Milli-Q water.
Electrochemical STM Measurements. In situ electrochemi-
cal STM measurements were carried out using a NanoScope E
(Digital Instruments) control unit and a PicoSPM (Molecular
Imaging) scanning unit with an airtight chamber and solution
inlet tubes. STM images with atomic resolution were recorded
as current images in a constant-height mode, and wide-range
STM images were recorded as height images in a constant-
current mode. A homemade electrochemical STM cell, which
can accommodate a single-crystal electrode, was used. A small
quasi-reversible hydrogen electrode and a platinum wire were
used as the reference and the counter electrode, respectively.
The electrolyte solution was deaerated by passing purified argon
gas for at least 20 min before it was introduced into the STM
cell. Electrochemical potentials of the Au(111) substrate (ES)
and STM tip (ET) were independently controlled by a bipoten-
tiostat (PicoSTAT, Molecular Imaging). STM tips were me-
chanically cut Pt/Ir wire (80/20, φ ) 0.25 mm) insulated with
nail polish. Before each measurement, pure argon gas was
passed to the STM chamber to purge the oxygen.
Figure 1. Time dependence of frequency change at +0.95 V when
the PdCl42- solution was added to the EQCM cell that contained a 50
mM H2SO4 solution. The final concentration of PdCl42- in the cell was
0.1 mM.
The mass change (∆m) was calculated from the resonant
frequency shift (∆f) using the Sauerbrey equation.24 The mass
sensitivity of the 5 MHz AT-cut quartz crystal was calibrated
by the deposition of silver and lead.13,25 Both measurements
gave similar values, and -19.3 ng/Hz cm2 was used for the
calculation of the mass change.
XRD Measurements. The XRD measurement was carried
out using a RINT-2200 (RIKAKU) with a four-circle diffrac-
tometer. After the electrochemical deposition of palladium on
the Au(111) electrode, the electrode was moved from the
electrolyte solution under potential control, then rinsed with
Milli-Q water, and finally dried with purified argon gas. A
palladium foil and the Au(111) single crystal were used for
comparison.
Results and Discussion
2-
Adsorption of PdCl4 Species on an Au(111) Electrode.
Figure 1 shows the time dependence of frequency change at
+0.95 V where neither a cathodic nor an anodic current was
observed when 0.17 mL of 30.6 mM K2PdCl4 solution was
added into the EQCM cell that contained 50 mL of 50 mM
H2SO4 solution while the solution was stirred with a stirring
bar. The final concentration of PdCl42- in the cell was 0.1 mM.
The frequency was increased by 0.5 Hz in the first ca. 20 s of
2-
the induction period after the addition of PdCl4 and then
abruptly decreased by 4.5 Hz, which corresponded to the surface
mass increase of 86.9 ng/cm2. The frequency did not change
except for a little perturbation when the same volume of 50
mM H2SO4 solution was added to the EQCM cell. Thus, the
2-
mass increase should be related to the injection of the PdCl4
Electrochemical and EQCM Measurements. Electrochemi-
cal and EQCM measurements were carried out in a three-
compartment cell with a Luggin capillary. The electrochemical
potential was controlled by a potentiostat (HA-151, Hokuto
Denko), and an external potential was provided by a function
generator (HB-111, Hokuto Denko). A quasi-reversible hy-
drogen electrode and a platinum foil were used as the reference
and the counter electrode, respectively. The potential is
referenced to the reversible hydrogen electrode (RHE) in this
paper. The resonant frequency of the quartz crystal electrode,
which was oscillated by a homemade oscillation circuit, was
simultaneously monitored with the electrode potential and
current by a frequency counter (HP53131A, Hewlett-Packard)
controlled by a personal computer (PC9821cb2, NEC) through
a GPIB interface. The frequency stability of the EQCM system
was better than 0.1 Hz for a sampling gate time of 0.1 s.
All the measurements were carried out after the solution was
purged by passing pure argon gas for at least 20 min through
the solution.
species. It is known, however, that the bisulfate anion is
adsorbed on the Au(111) electrode surface at this potential.26,27
A similar experiment was carried out in a 50 mM HClO4
-
solution in which the ClO4 is considered as an unadsorbed
species. The surface mass was changed in the same manner as
the case in H2SO4 solution when the same amount of K2PdCl4
solution was introduced into HClO4 solution. We therefore
2-
attributed the mass increase to the adsorption of PdCl4 on
2-
the Au(111) electrode surface. The number of adsorbed PdCl4
species calculated from the surface mass change was 2.1 × 1014/
cm2, i.e., ca. 15% of that of the Au(111) surface atoms. This
value was the same as that found for the adsorbed hexachlo-
roplatinate complex, PtCl62-, on Au(111) with the (x7 × x7)-
R19.1° structure that we recently reported.13
2-
The structure of the PdCl4 adlayer was investigated by in
situ electrochemical STM at +0.95 V in a 50 mM H2SO4
solution containing 0.05 mM PdCl42-. The atomically ordered
Au(111)-(1 × 1) structure (nearest-neighbor distance of ca.

4368 J. Phys. Chem. B, Vol. 102, No. 22, 1998
Naohara et al.
Figure 3. Potential dependence of (a) current and (b) mass change
(∆m) at an Au(111) electrode in a 50 mM H2SO4 solution containing
0.1 mM PdCl42-. Potential sweep rate was 5 mV s-1
.
Figure 2. STM images (10 × 10 nm2) of (a) an Au(111)-(1 × 1)
electrode obtained at +0.95 V in a 50 mM H2SO4 solution and (b) the
2-
same electrode in a 50 mM H2SO4 solution containing 0.1 mM PdCl4
while the electrode potential was kept constant at +0.95 V.
0.29 nm) was observed at +0.95 V before the addition of
PdCl42- (Figure 2a). When a 50 mM H2SO4 solution containing
0.1 mM PdCl42- was added into the electrochemical STM cell
through the solution inlet tube which was attached to the airtight
STM chamber (final concentration of PdCl42- of ca. 0.05 mM)
while keeping the electrode potential at 0.95 V, a new structure
of hexagonal symmetry with a nearest-neighbor distance of ca.
0.78 nm was observed (Figure 2b). The distance of 0.78 nm is
x7 times that of the Au(111) lattice. The symmetry of the
adlattice was found to be rotated by ca. 20° with that of the
Au(111) substrate. This structure is different from that of a
bisulfate adlayer, i.e., (x3 × x7) 26,28 but similar to that of the
Figure 4. Relation between the mass change (∆m) and the charge
(Q) obtained from Figure 3. The solid line and the dotted line represent
the results obtained during the cathodic scan and the anodic scan,
respectively.
2-
PtCl6 adlayer on Au(111).13 Thus, we concluded that the
PdCl42- complex was adsorbed on the Au(111) with the adlattice
structure of (x7 × x7)R19.1°. The number of adsorbed
2-
To analyze the result more quantitatively, the mass change
(∆m) was plotted against the charge, Q (solid line in Figure 4).
A good linear relation was obtained in this potential region
between +0.95 and +0.35 V during the cathodic scan. The
slope of this plot gives the mass change per charge equivalent
to one mole of electrons, mpe, as 51.3, which is in good
agreement with the value expected for the following reduction
PdCl4 estimated from this structure is equal to 14% of that
of the Au(111) surface atoms and is in very good agreement
with that observed by EQCM measurements.
As the potential became more positive than +1.2 V, no clear
STM image of the adlayer was obtained.
Electrochemical Deposition of Palladium. EQCM Mea-
2-
surements. After the adsorption of PdCl4 was completed in
2-
process of the PdCl4 complex to palladium:29
the H2SO4 solution at +0.95 V, the potential was scanned
negatively to +0.35 V and then positively to +1.2 V. Simul-
taneously recorded current and mass changes during the potential
sweep (5 mV/s) are shown in Figure 3 as a function of potential.
A cathodic current started to flow as soon as the potential was
made more negative than +0.95 V, reaching a maximum at
+0.88 V and decreasing to a limiting value. The mass increase
was observed at all the potentials during the cathodic scan from
+0.95 to +0.35 V. The cathodic current observed in this region
may be due to the reduction of gold oxide, the reduction of
PdCl42- + 2e- f Pd + 4Cl-
(I)
If the preadsorbed palladium complex is reduced to palladium,
the mass loss due to Cl- desorption should be observed. Thus,
the mass gain with an mpe of 51.3 means the reduction of the
complex in the solution. Actually, the STM investigation
described later showed that the palladium complex covered not
only the gold substrate but also the deposited palladium surface
during the deposition. Thus, the complex in solution seems to
2-
PdCl4 to palladium, or both.

Growth of Pd on Au(111)
J. Phys. Chem. B, Vol. 102, No. 22, 1998 4369
be adsorbed on the surface as soon as the adsorbed complex is
reduced and only the net mass change due to the Pd deposition
was observed. The amount of the deposited palladium deter-
mined from the mass change from +0.95 to +0.35 V in this
particular case was 1.2 monolayers (ML).
As the potential was swept in the anodic direction from +0.35
V, a relatively constant cathodic current flowed and the mass
was uniformly increased to +0.80 V (Figure 3). The slope of
∆m vs Q plot (broken line in Figure 4), mpe, in this region is
close to the one observed during the cathodic scan, showing
that the two-electron reduction of PdCl42- to palladium shown
by eq 1 proceeded in this potential region. The mass increase
reached a maximum value at +0.80 V, which was equivalent
to a palladium deposition of 1.7 ML.
The anodic current started to flow at +0.80 V and increased
as the potential became more positive, and two anodic peaks
were observed at +0.89 and +0.96 V. A mass decrease was
observed as soon as the anodic current started to flow. The
anodic current observed at potentials more positive than +0.8
V could be due to the oxidation of palladium to palladium
hydroxide,30 the dissolution of palladium, or both. A very good
linear relation in the ∆m vs Q plot was obtained within this
potential region (broken line in Figure 4) and the mpe was 63.4,
which is in agreement with the value expected for the reverse
reaction of reaction I, i.e., dissolution of palladium to the
PdCl42- complex. If the (hydro-)oxide was formed, the surface
mass should increase and mpe should be less than the above
value. It is clear, therefore, that the observed anodic current
was due to the dissolution of palladium. The amount of mass
decrease in this potential region shows that all of the deposited
palladium layer was dissolved.
Figure 5. (a) Time dependence of current (dotted line) and the mass
change (∆m, solid line) at an Au(111) electrode in a 50 mM H2SO4
2-
solution containing 0.1 mM PdCl4 when the potential was stepped
from +0.95 to +0.70 V. (b) Relationship between the mass change
(∆m) and the charge (Q) after the potential was stepped from +0.95
to +0.75 V.
section along the white dotted line indicated in the STM image
is shown as Figure 6a′. Atomically flat and large terraces
(Terrace-1 and Terrace-2) and monatomic steps of gold of the
Au(111) substrate were observed. Parts b-h of Figure 6
sequentially show obtained STM images of the same area (b) 0
min, (c) 1 min, (d) 2 min, (e) 3 min, (f) 28 min, (g) 48 min,
and (h) 90 min after the electrochemical potential of the substrate
(ES) was stepped from +0.95 to +0.80 V as indicated by the
thick arrow in Figure 6b where palladium deposition is expected
to take place. The scan direction is shown by the arrow at the
upper-left side of each figure. The electrochemical potential
of the STM tip (ET) was held at +1.0 V during the measurement.
As soon as ES was stepped from +0.95 to +0.80 V, the
palladium nuclei were being generated on the Au(111) substrate.
The cross section (Figure 6b′) along the dotted line shows the
formation of the palladium islands with a monatomic height on
the large terraces of the gold substrate. In the upper part of
Figure 6b two-dimensional growth of the first palladium
monolayer (Pd-1) progressed not only on the large terrace but
also on a narrow terrace between the step lines. This result is
in contrast to that observed by Baldauf and Kolb. They reported
that the palladium layer started to preferentially grow from the
step site.20 This discrepancy should be due to the difference in
the potential of the Au(111) substrate. As shown in parts c
and c′ of Figure 6, after the first palladium layer (Pd-1) was
completely formed on the gold substrate, islands of the second
palladium layer (Pd-2) were generated on both Terrace-1 and
Terrace-2 and the first palladium layer (Pd-1) on the narrow
terrace started to grow laterally further onto the Pd-1 layer on
Terrace-2 from the step line. Both the Pd-2 layer on Terrace-2
and the palladium layer extending from the Pd-1 layer on the
narrow terrace grew two-dimensionally and merged, resulting
in a large flat terrace of a second palladium layer on Terrace-2
(parts d and e of Figure 6). The cross section (parts c′, d′, and
During the negative potential sweep from +1.2 to +0.95 V,
a small cathodic current flowed and the surface mass was
slightly increased, showing the redeposition of palladium.
Figure 5a shows the time dependence of the current and the
surface mass change simultaneously recorded in 50 mM H2-
2-
SO4 solution containing 0.1 mM PdCl4 when the potential
was stepped from +0.95 to +0.70 V. A sharp cathodic current
spike was observed after the potential step and the surface mass
gradually increased. A very good linear relation was obtained
between the surface mass change and the charge calculated from
the current in Figure 5a. The mpe calculated from the slope is
54.7, which is in very good agreement with the value expected
for the two-electron reduction process given by eq I. It is
interesting to note that an abrupt apparent surface mass increase
of ca. 30 ng/cm2, i.e., a frequency decrease of ca. 1.7 Hz, was
observed around 190 s after the potential step. By this time,
290 ng/cm2, i.e., one monolayer, of palladium was deposited.
This abrupt change was observed in several independent
experiments when the substrate was covered with ca. 1 ML of
palladium. It is rather difficult to imagine that this change was
due to the actual mass change. One possible explanation is the
surface stress. It was reported that the resonance frequency of
a quartz resonator was decreased when the surface stress was
increased.31 When the first monolayer of palladium was
completed, the surface stress may increase, since there is a
mismatch of 4% in the lattice spacing between gold and
palladium.
In Situ STM Measurements. The deposition of PdCl42- was
also investigated by in situ electrochemical STM. Figure 6
shows STM images (300 × 300 nm2) of the Au(111) surface
in a 50 mM H2SO4 solution containing 0.5 mM PdCl42-. It
took ca. 50 s to capture one image. Figure 6a is an STM image
at +0.95 V where no palladium deposition takes place. A cross

4370 J. Phys. Chem. B, Vol. 102, No. 22, 1998
Naohara et al.
Figure 6. (a) STM image (300 × 300 nm2) of an Au(111) substrate at +0.95 V in a 50 mM H2SO4 solution containing 0.5 mM PdCl42-. Sequentially
obtained STM images (300 × 300 nm2) of the palladium deposition process on the Au(111) substrate (b) 0 min, (c) 1 min, (d) 2 min, (e) 3 min,
(f) 28 min, (g) 48 min, and (g) 90 min after the potential were stepped from +0.95 to +0.80 V in the same solution. A cross section along with
the white dotted line in each figure is shown below each image.
2-
e′ of Figure 6) clearly shows the two-dimensional growth of
the palladium monatomic layer. After the complete second
palladium layer was formed on both Terrace-1 and Terrace-2,
the third palladium layer (Pd-3) started to grow two-dimension-
ally on the terraces and from the step line (parts f and g of
Figure 6). The growth rate of the third palladium layer was,
however, slower than that of the first and the second layers.
This should be due to the decrease in the local concentration of
the reactant, i.e., PdCl42- near the surface, leading to a negative
shift in the equilibrium potential for palladium deposition. The
palladium layer, which was originally first deposited on Terrace
1, was also grown laterally and finally merged with Pd-3 on
Terrace 2 (parts f-h of Figure 6).
atomically flat palladium surface on which PdCl4 adsorbed
with a (x7 × x7)R19.1° structure was obtained (Figure 7c).
The second layer of palladium started to grow on top of the
first palladium layer as observed at the right-bottom corners of
parts b and c of Figure 7. .
The palladium layer grown electrochemically under the
present conditions is very flat, and its domain is much larger
than that of the vacuum-deposited palladium layer.19 The
2-
adsorbed PdCl4 with a (x7 × x7)R19.1° structure on both
the gold substrate and the deposited palladium layer should play
a very important role in the present preferential 2D growth of
the palladium layer.
Figure 8 shows a schematic model for the palladium
2-
Thus, palladium deposition proceeds in a layer-by-layer
growth mode. Under these experimental conditions, an STM
tip-induced local deposition, which we previously reported,32
was not observed because ET was more positive than ES.
To investigate the deposition process more in detail, higher
resolution images were obtained during the palladium deposition
in 50 mM H2SO4 solution containing 0.5 mM PdCl42-. Figure
7 shows the STM images (40 × 40 nm2) (a) 20 s, (b) 40 s, and
(c) 2 min after the potential was stepped from +0.95 to +0.87
V. In Figure 7a, monatomic-height palladium deposited layers
are observed at the upper left and in the right region in the
image. White spots are observed everywhere on the surface.
Since the arrangement of these spots is the same as that of Figure
deposition in the presence of the PdCl4
adlayer. The
2-
nucleation of palladium occurred at the surface defects. PdCl4
was also adsorbed on a deposited palladium island. The PdCl4
2-
adlayer on the terrace of the palladium island should be more
stable than that at the edge of the palladium island because of
the interaction between the adsorbed species; therefore, the
deposition rate of palladium on the terrace was greatly reduced.
The deposition of palladium is easier at the edge site of an island
where the blocking by PdCl42- is weaker than that on the terrace.
Thus, the ordered PdCl42- adlayer inhibits the vertical but favors
lateral, i.e., 2D, growth of the palladium layer as observed in
the present experiment. A similar enhanced 2D growth of
copper on an Au(111) electrode has been observed when organic
additives are added to the solution.7,33,34 The present result is,
however, the first evidence for the effect of the adsorption of
reactant itself, i.e., PdCl42-, during the electrochemical dimen-
sion growth.
2-
2b, these white spots should correspond to the adsorbed PdCl4
2-
complex. Thus, the PdCl4 adsorbed not only on the gold
substrate but also on the palladium layer with an ordered
structure of (x7 × x7)R19.1°. As time passed, palladium
deposition proceeded two-dimensionally (Figure 7b) and the
gold substrate was covered with the palladium, and finally an
Characterization of Deposited Pd Layer. The structure of
the deposited palladium layer was examined by means of X-ray

Growth of Pd on Au(111)
J. Phys. Chem. B, Vol. 102, No. 22, 1998 4371
Figure 7. Sequentially obtained higher resolution STM images (40 ×
40 nm2) of the palladium deposition process on an Au(111) substrate
(a) 20 s, (b) 40 s, and 2 min after the potential was stepped from +0.95
2-
to +0.87 V in 50 mM H2SO4 solution containing 0.5 mM PdCl4
.
Figure 9. XRD patterns of (a) an Au(111) single-crystal electrode,
(b) an Au(111) electrode followed by electrochemical deposition of
palladium at 0.7 V for 1 h in a 50 mM H2SO4 solution containing 1
mM PdCl42- (amount of palladium deposited layer is equal to 71 ML)
and (c) a polycrystalline Pd foil.
ML. The assignments based on the JCPDS data are given for
the respective reflection peaks in the figure. For the Au(111)
single crystal, only Au(111) and Au(222) peaks were observed.
In addition to the Au(111) and Au(222) peaks, the (111) peak
of palladium was observed in the XRD pattern of the Pd(71
ML)/Au(111) (Figure 9b). No other peaks due to palladium
such as (200), (220), and (311), which were observed in the
case of the polycrystalline palladium (Figure 9c), were observed.
Thus, it is confirmed that the Pd(111) phase was deposited
electrochemically on the Au(111) substrate.
The surface structure of the palladium deposited layer was
also examined by recording the cyclic voltammograms (CVs)
in H2SO4 solution and for the underpotential deposition (UPD)
of copper (Figure 10). The shape of CV in general and the
positions of the UPD peaks in particular of metal electrodes
are known to be very sensitive to the surface structure and
order.7-11 After the deposition of palladium in 50 mM H2SO4
solution containing 0.01 mM PdCl42- at 0.7 V for 30 min (ca.
2 ML), the electrode was rinsed with milli-Q water. The CVs
were recorded in another cell that contained a 50 mM H2SO4
Figure 8. Schematic model for the palladium deposition process on
an Au(111) substrate in the presence of the PdCl42- adlayer. The ordered
PdCl42- adlayer inhibits the vertical but favors lateral, i.e., 2D, growth
of the palladium layer as observed in the present experiment.
diffraction (XRD). Figure 9 shows the XRD patterns of (a) an
Au(111) single crystal, (b) an Au(111) single crystal on which
palladium was deposited electrochemically at 0.7 V for 1 h in
2-
a 50 mM H2SO4 solution containing 1 mM PdCl4 (Pd/Au-
(111)), and (c) a bare polycrystalline Pd foil. The amount of
the deposited palladium estimated from the charge was ca. 71

4372 J. Phys. Chem. B, Vol. 102, No. 22, 1998
Naohara et al.
alloy formation between gold and palladium, which was
observed in UHV (>500 K),17,18 did not take place under this
condition.
Conclusion
The adsorption and electrochemical reduction of PdCl42- on
an Au(111) electrode in a H2SO4 solution was investigated by
EQCM and in situ electrochemical STM. The adlayer of
2-
PdCl4 with a (x7 × x7)R19.1° structure on an Au(111)
substrate was observed by STM measurement, and the amount
of the adsorbate determined by EQCM measurement was in
good agreement with the values expected for this structure.
Palladium deposition proceeded with an epitaxial layer-by-layer
2-
growth mode, and the adlayer of PdCl4 with a (x7 × x7)-
R19.1° structure was observed on the deposited palladium layer.
2-
The important role of the PdCl4 adlayer on the growth
mechanism was suggested.
Acknowledgment. This work was partially supported by a
Grant-in-Aid for Scientific Research on Priority Area of
“Electrochemistry of Ordered Interfaces” (No. 09237101) from
the Ministry of Education, Science, Sports and Culture, Japan.
References and Notes
(1) Gewirth, A. A.; Niece, B. K. Chem. ReV. 1997, 97, 1192.
(2) LaGraff, J. R.; Gewirth, A. A. Nanoscale Probes of the Solid/Liquid
Interface; Gewirth, A. A., Siegenthaler, H., Eds.; Kluwer Academic
Publishers: Dordrecht, 1995; pp 83-101.
(3) Uosaki, K.; Koinuma, M.; Sekine, N.; Ye, S. Solid-Liquid
Electrochemical Interfaces; Jerkoiewicz, G., Soriaga, M. P., Uosaki, K.,
Wieckowski, A., Eds.; American Chemical Society: Washington, DC, 1997;
pp 189-201.
(4) Buttry, D. A.; Ward, M. D. Chem. ReV. 1992, 92, 1355.
(5) Toney, M. F. Synchrotron Techniques in Interfacial Electrochem-
istry; Melendres, C. A., Tadjeddine, A., Eds.; Kluwer Academic Publish-
ers: Dordrecht, 1994; pp 109-125.
(6) Ocko, B. M.; Wang, J. Synchrotron Techniques in Interfacial
Electrochemistry; Melendres, C. A., Tadjeddine, A., Eds.; Kluwer Academic
Publishers: Dordrecht, 1994; pp 127-155.
(7) Will, T.; Dietterle, M.; Kolb, D. M. Nanoscale Probes of the Solid/
Liquid Interface; Gewirth, A. A., Siegenthaler, H., Eds.; Kluwer Academic
Publishers: Dordrecht, 1995; pp 137-162.
(8) Ogaki, K.; Itaya, K. Electrochim. Acta 1995, 40, 1249.
(9) Tao, N. J.; Pan, J.; Li, Y.; Oden, P. I.; DeRose, J. A.; Lindsay, S.
M. Surf. Sci. Lett. 1992, 271, L338.
Figure 10. Cyclic voltammograms of the deposited palladium layer
(ca. 2 ML) on an Au(111) electrode in (a) 50 mM H2SO4 solution
(sweep rate of 50 mV s-1) and (b) 5 mM CuSO4 + 50 mM H2SO4
solution (sweep rate of 5 mV s-1).
solution (Figure 10a) and a 50 mM H2SO4 + 5 mM Cu2+
solution (Figure 10b). The shape of the current-potential curve
and the charge of hydrogen adsorption (214 µC cm-2) on a Pd
thin layer in a 50 mM H2SO4 solution were in good agreement
with those reported by Baldauf et al.20 Very sharp reduction
and oxidation peaks with a full width at half-maximum (fwhm)
of ca. 18 mV were observed in a 50 mM H2SO4 solution
containing 5 mM Cu2+ at +0.52 and +0.54 V, respectively
(Figure 10b). The copper stripping charge was 436 µC/cm2.
These values are close to those reported for copper UPD on a
Pd(111) electrode in a 0.5 M NaClO4 + 10-2 M HClO4 + 10-3
M Cu2+ solution (sweep rate of 10 mV/s). The CV of copper
UPD on a Pd(111) electrode shows a pair of very sharp redox
peaks (+0.45 and +0.47 V, fwhm of 20 mV). The stripping
charge was 440 µC/cm2. The small difference in peak potentials
reflects the difference in Cu2+ concentration. The CV of copper
UPD in a polycrystalline palladium electrode is totally different
from these CVs and shows three pairs of redox peaks (+0.26/
+0.27, +0.46/+0.47, and +0.48/+0.49 V) in the same solu-
tion.35 Furthermore, the potentials of the copper UPD peaks
are completely different from those of copper UPD on an Au-
(111) electrode (+0.56/+0.57 and +0.38/+0.41 V). Thus, it
was confirmed that the deposited palladium layer on the Au-
(111) electrode was the epitaxially grown Pd(111) phase. The
(10) Chen, C.-h. C.; Gewirth, A. A. J. Am. Chem. Soc. 1992, 97, 9754.
(11) Ikemiya, N.; Iwai, D.; Yamada, K.; Vidu, R.; Hara, S. Surf. Sci.
1996, 369, 199.
(12) Sawaguchi, T.; Yamada, T.; Okinaka, Y.; Itaya, K. J. Phys. Chem.
1995, 99, 14149.
(13) Uosaki, K.; Ye, S.; Oda, Y.; Haba, T.; Hamada, K. Langmuir 1997,
13, 594.
(14) Uosaki, K.; Ye, S.; Naohara, H.; Oda, Y.; Haba, T.; Kondo, T. J.
Phys. Chem. B 1997, 101, 7566.
(15) Haiss, W.; Sass, J.-K. J. Electroanal. Chem. 1997, 431, 15.
(16) Somorjai, G. A. Introduction to Surface Chemistry and Catalysis;
John Wiley & Sons: New York, 1994.
(17) Koel, B. E.; Sellidj, A.; Paffett, M. T. Phys. ReV. B 1992, 46, 7846.
(18) Baddely, C. J.; Barnes, C. J.; Wander, A.; Ormerod, R. M.; King,
D. A.; Lambert, R. M. Surf. Sci. 1994, 314, 1.
(19) Baddely, C. J.; Ormerod, R. M.; Stephenson, A. W.; Lambert, R.
M. J. Phys. Chem. 1995, 99, 5146.
(20) Baldauf, M.; Kolb, D. M. Electrochim. Acta 1993, 38, 2145.
(21) Baldauf, M.; Kolb, D. M. J. Phys. Chem. 1996, 100, 11375.
(22) Clavilier, J. J. Electroanal. Chem. 1980, 107, 211; 1980, 107, 205.
(23) Uosaki, K.; Ye, S.; Kondo, T. J. Phys. Chem. 1996, 99, 1411.
(24) Sauerbrey, G. Z. Z. Phys. 1959, 155, 206.
(25) Bruckenstein, S.; Swathirajan, S. Electrochim. Acta 1985, 30, 851.
(26) Edens, G. J.; Gao, X.; Weaver, M. J. J. Electroanal. Chem. 1994,
375, 357.
(27) Shi, Z.; Lipkowski, J.; Gamboa, M.; Zelenay, P.; Wieckowski, A.
J. Electroanal. Chem. 1994, 366, 317.
(28) Magnussen, O. M.; Hagebo¨ck, J.; Hotlos, J.; Behm, R. J. Faraday
Discuss. 1992, 94, 329.

Growth of Pd on Au(111)
J. Phys. Chem. B, Vol. 102, No. 22, 1998 4373
(29) Llopis, J. F.; Colom, F. Encyclopedia of Electrochemistry of the
Elements; Bard, A. J., Ed.; Marcel Dekker: New York, 1976; Vol. VI.
(30) Kessler, T.; Visntin, A.; Bolzan, A. E.; Andreasen, G.; Salvarezza,
R. C.; Triaca, W. E.; Arvia, A. J. Langmuir 1996, 12, 6587.
(31) EerNisse, E. P. J. Appl. Phys. 1972, 43, 1330.
(33) Nichols, R. J.; Bach, C. E.; Meyer, H. Ber. Bunsen-Ges. Phys.
Chem. 1993, 97, 1012.
(34) Nichols, R. J. Nanoscale Probes of the Solid/Liquid Interface;
Gewirth, A. A., Siegenthaler, H., Eds.; Kluwer Academic Publishers:
Dordrecht, 1995; pp 163-182.
(32) Naohara, H.; Ye, S.; Uosaki, K. Appl. Phys. A, in press.
(35) Chierchie, T.; Mayer, C. Electrochim. Acta 1988, 33, 341.