Underpotential deposition of Cd on Ag(1 1 1): An in situ STM study

Article abstract of DOI:10.1016/j.susc.2004.11.037

The kinetics and mechanism of Cd underpotential deposition (UPD) and involved surface alloy formation processes in the system Ag(1 1 1)/Cd ²⁺, SO42-, are studied by means of combined electrochemical measurements and in situ scanning tunneling microscopy (STM). The results show that the UPD process starts with a formation of an expanded (diluted) adlayer with a superlattice structure Ag(1 1 1)-(3×19)R23.4°. In the underpotential range 50 mV < ΔE < 80 mV this adlayer transforms to a condensed close packed Cd monolayer via a first order phase transition. At long polarization times the condensed monolayer undergoes structural changes involving place exchange processes between Cd atoms and surface Ag atoms. A formation of a second Cd monolayer and a significant Ag-Cd surface alloying take place at lower underpotentials (ΔE < 50 mV). The kinetics of surface alloying are analyzed on the basis of a recently proposed diffusion model including a relatively fast initial formation of a very thin surface alloy film and a subsequent slow alloy growth controlled by solid state diffusion. The anodic dealloying results in an appearance of monatomically deep pits, which disappear quickly at relatively high underpotentials (ΔE > 550 mV) indicating a high mobility of surface Ag atoms.

Full text of DOI:10.1016/j.susc.2004.11.037

Surface Science 576 (2005) 9–18
www.elsevier.com/locate/susc
Underpotential deposition of Cd on Ag(111):
an in situ STM study
a
S.G. Garc ´ı a , D.R. Salinas , G. Staikov
a
b,
*
a
Instituto de Ing. Electroqu ı´ mica y Corrosi o´ n, Universidad Nacional del Sur, Avda. Alem 1253, (8000) Bah ı´ a Blanca, Argentina
b
Institute of Thin Films and Interfaces (ISG 3), and cni-Centre of Nanoelectronic Systems for Information Technology,
Research Centre J u¨ lich, D-52425 J u¨ lich, Germany
Received 12 August 2004; accepted for publication 24 November 2004
Available online 10 December 2004
Abstract
The kinetics and mechanism of Cd underpotential deposition (UPD) and involved surface alloy formation processes
2
in the system Ag(111)/Cd , SO , are studied by means of combined electrochemical measurements and in situ scan-
+
2ꢀ
4
ning tunneling microscopy (STM). The results show that the UPDprocess starts with a formation of an expanded
ꢁ
pffi ffiffi pffiffiffiffiffi
(
diluted) adlayer with a superlattice structure Ag(111)-ð 3 Â 19ÞR23:4 . In the underpotential range 50 mV < DE <
8
0 mV this adlayer transforms to a condensed close packed Cd monolayer via a first order phase transition. At long
polarization times the condensed monolayer undergoes structural changes involving place exchange processes between
Cd atoms and surface Ag atoms. A formation of a second Cd monolayer and a significant Ag–Cd surface alloying take
place at lower underpotentials (DE < 50 mV). The kinetics of surface alloying are analyzed on the basis of a recently
proposed diffusion model including a relatively fast initial formation of a very thin surface alloy film and a subsequent
slow alloy growth controlled by solid state diffusion. The anodic dealloying results in an appearance of monatomically
deep pits, which disappear quickly at relatively high underpotentials (DE > 550 mV) indicating a high mobility of sur-
face Ag atoms.
ꢀ
2004 Elsevier B.V. All rights reserved.
Keywords: STM; Underpotential deposition; Alloy formation; Cadmium; Silver
1. Introduction
Underpotential deposition (UPD) of metals
taking place at electrode potentials (E) more posi-
*
Corresponding author. Tel.: +49 2461 612064; fax: +49
tive than the equilibrium Nernst-potential (E3DMe
)
of the corresponding three-dimensional metal bulk
2461 613907.
E-mail address: g.staikov@fz-juelich.de (G. Staikov).
0
039-6028/$ - see front matter ꢀ 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.susc.2004.11.037

10
S.G. Garc ı´ a et al. / Surface Science 576 (2005) 9–18
phase (3DMe) is of great importance for the
chemistry and physics of solid/liquid interfaces,
electrocrystallisation, electrocatalysis, and electro-
deposition of alloys and heterostructured ultrathin
films [1]. In the last decade the UPDprocesses
were studied extensively combining electrochemi-
cal measurements with in situ scanning probe tech-
niques such as scanning tunnelling microscopy
panded Au(100)-c(2 · 2)Cd structure. A similar
behaviour has also been observed previously in
2
+
the system Ag(100)/Cd , SO , where the forma-
2ꢀ
4
tion of an initial UPDCd-overlayer with a
Ag(100)-c(2 · 2)Cd superlattice structure was sug-
gested on the basis of charge measurements [12].
The identical superlattice structures of the initial
Cd UPDadlayers in both systems are obviously
related to very similar crystal structures of
Ag(100) and Au(100) substrates. However, exper-
imental results clearly show that on Au(100) the
Cd UPDstarts at much higher underpotentials,
which indicates that the chemical nature and elec-
tronic properties of the substrate play also an
important role in this process. Comparative kinetic
studies of surface alloy formation processes occur-
ring at lower underpotentials in both systems show
also a much higher rate of surface alloying on the
Au(100) substrate [24]. Very interesting in situ
STM studies of the initial stages of Cd electrode-
position on a herringbone reconstructed Au(111)
surface have been reported recently by Maupai
et al. [25,26]. They observed during Cd UPDa
preferential formation of nanoclusters of mon-
atomic height and very uniform lateral size at the
elbow sites of the reconstructed Au(111) surface.
According to the so-called low-dimensional system
(LDS) concept [8,9], these clusters can be consid-
ered as 0Dsystems stabilized by the elbow sites.
An understanding of the role of numerous factors
affecting the kinetics and mechanism of UPDpro-
cesses, however, requires further in situ STM and
AFM investigations in well-defined model systems.
The UPDof Cd on Ag(111) has been investigated
up to now primarily with classical electrochemical
methods [12,13,27]. In situ STM studies are re-
stricted only to the participation of this process
in the formation of compound films such as CdS
by ECALE [28].
(
7
STM) and atomic force microscopy (AFM) [1–
]. It has been shown that the substrate surface
inhomogeneities play an important role in the
UPDprocesses, which cause the formation of var-
ious low-dimensional metal phases (systems) [8,9].
The formation mechanism and the atomic struc-
ture of low-dimensional metal phases were found
to be strongly influenced by specifically adsorbed
electrolyte anions [1–9].
In many UPDsystems the strong metal–sub-
strate interaction induces place exchange pro-
cesses, which lead to the formation of surface
alloys [1]. A pronounced surface alloy formation,
however, is usually observed in systems exhibiting
sufficient miscibility between the substrate and the
UPDmetal. Typical examples represent the UPD
2
+
2+
systems Au/Cd and Ag/Cd [1,10–14]. These
systems attracted in the last years a renewed inter-
est not only as model systems for the study of sur-
face alloy formation processes but also due to their
application in the electrochemical atomic layer epi-
taxy (ECALE) [15,16]. Recent in situ STM and
AFM studies of Cd UPDon gold single crystal
substrates provided new important information
on the atomic structure of initial low-dimensional
Cd phases, which act as precursors for the subse-
quent Cd–Au surface alloy formation [17–23].
Expanded UPDadlayers with novel ‘‘linear’’
structures have been observed by in situ STM at
relatively high underpotentials in the system
2
+
2ꢀ
Au(111)/Cd , SO [17,18]. The appearance of
these structures has been related to a partial charge
transfer between Au substrate atoms and Cd ada-
4
In this paper, we present first in situ STM stud-
ies of Cd UPDon Ag(111). A special attention is
paid to the atomic structure and the local mecha-
nism of formation of different low-dimensional
UPDphases including Cd–Ag surface alloys. Ki-
netic information is obtained combining in situ
STM observations with voltammetric and charge
measurements.
2
4
ꢀ
toms as well as to the specific adsorption of SO
anions. Detailed studies on Cd deposition and
2
stripping in the system Au(100)/Cd , SO apply-
+
2ꢀ
4
ing in situ AFM have been performed by Vidu and
Hara [19–22]. They showed that the UPDprocess
in this system starts with the formation of an ex-

S.G. Garc ı´ a et al. / Surface Science 576 (2005) 9–18
11
2
. Experimental
The experiments were performed in the system
Ag(111)/5 mM CdSO + 5 mM H SO + 0.5 M
Na SO at a temperature T = 298 K. The electro-
4
2
4
2
4
lyte solution was prepared from suprapure chemi-
cals (Merck, Darmstadt) and fourfold quartz-
distilled water. Prior to each experiment the solu-
tion was deaerated by nitrogen bubbling. In situ
STM studies were carried out with a standard
NanoScope III (Digital Instruments, Santa Bar-
bara, CA, USA) equipment using Pt–Ir tips coated
with Apiezon wax. Platinum wires served as coun-
ter and quasi-reference electrodes. In each STM
experiment the tip potential, E was held constant
tip
Fig. 1. Cyclic voltammogram for Cd UPDin the system
2ꢀ
2
+
at a value of minimum faradaic current. The corre-
Ag(111)/Cd , SO
4
. (jdE/dtj = 10 mV/s; DE = E ꢀ E3DCd).
sponding tunneling voltage, Ebias = Esample ꢀ E
tip
and the tunneling current, Itun are indicated in the
figure captions. Additional conventional electro-
chemical measurements were performed in a stan-
dard three-electrode electrochemical cell with
saturated mercury sulfate (SSE) reference electrode
the underpotential range DE = E ꢀ E
P
3
DCd
0 mV. The voltammogram corresponds well to
that observed in a previous electrochemical study
of this system [12] and is characterized by four
adsorption(A)/desorption(D) peak pairs. In situ
STM imaging of the substrate surface in the poten-
tial range positive of the voltammetric peaks
(DE > 400 mV) shows a two-dimensional hexago-
nal lattice with an interatomic spacing of
0.29 ± 0.02 nm, which corresponds to the atomic
structure of the Ag(111) surface (Fig. 2). The
nearly symmetrical voltammetric peak pairs A1/
D1 and A /D observed in the UPDrange
2
and a platinum sheet (1 cm ) counter electrode. All
electrode potentials are referred to the Nernst equi-
librium potential of the pure 3DCd bulk phase
E3DCd = ꢀ1150 mV vs. SSE (for c 2þ ¼ 5 mM).
Cd
Ag(111) single crystals with a diameter of
.65 cm were used as substrates. Before each
0
experiment the substrate surface was mechanically
polished with diamond paste of decreasing grain
size down to 0.25 lm and subsequently chemically
etched according to a standard procedure [29],
including etching for 10 s in a solution containing
2
2
100 mV 6 DE 6 400 mV indicate the occurrence
of reversible adsorption/desorption processes,
which have been attributed previously to the for-
4
4
.8% CrO , 4% HCl and 20% HClO , for 20 s in
3 4
M H SO and finally for 300 s in 25% aqueous
2
mation of an expanded adlayer with a superlattice
ꢁ
4
pffi ffiffi pffi ffiffi
solution of NH . After etching, the electrodes were
3
structure Ag(111)-ð 3 Â 3ÞR30 –Cd [12]. How-
thoroughly rinsed with fourfold distilled water.
This procedure was repeated until a highly reflect-
ing surface exhibiting relatively large atomically
flat terraces separated by monatomic steps was
obtained.
ever, in situ STM imaging with lateral atomic
resolution at underpotentials 100 mV 6 DE 6
150 mV reveals overlayer structures, which are
very similar to those reported for the UPDof Cd
on Au(111) [17,18]. Fig. 3a shows the appearance
of such structure by stepping the underpotential
from DE = 450 mV to DE = 125 mV. The 2D
hexagonal lattice of the Ag(111) surface can be
identified in the lower part of the STM image
in Fig. 3a (corresponding to DE = 450 mV). The
upper part of the STM image (corresponding to
3
. Results and discussion
Fig. 1 shows a typical cyclic voltammogram for
deposition and dissolution of Cd on Ag(111) in

12
S.G. Garc ı´ a et al. / Surface Science 576 (2005) 9–18
Fig. 2. In situ STM image with lateral atomic resolution of the
2ꢀ
2+
Ag(111) surface obtained in the system Ag(111)/Cd , SO at
DE = 450 mV. Grayscale range: 15 nA; Ebias = ꢀ30 mV; Itun
0 nA.
4
=
2
DE = 125 mV) shows clearly the appearance of the
specific structure characterized by two types of
parallel rows, which are composed of brighter
spots (A) and darker spots (B), respectively. These
rows are rotated by 30ꢁ with respect to the
Ag[110] atomic rows of the Ag(111) surface and
alternate with each other at a distance of 0.55 ±
0
.02 nm. The spacing between the spots along both
types of rows is 0.5 ± 0.02 nm and corresponds to
pffiffiffi
3
times the Ag–Ag distance in the Ag[110] atom-
ic rows of the Ag(111) substrate. The different
intensities of the spots labeled as (A) and (B) indi-
cate that the adatoms in the overlayer are located
on surface sites with different symmetry. The ob-
Fig. 3. (a) In situ STM image showing the formation of an
,
2+
expanded overlayer structure in the system Ag(111)/Cd
2ꢀ
SO after a potential step from DE = 450 mV to DE = 125 mV.
4
Grayscale range: 10 nA; Ebias = ꢀ30 mV; Itun = 10 nA. (b)
served superlattice structure can be denoted as
ꢁ
pffi ffiffi pffiffiffiffiffi
Model for the overlayer structure shown in (a).
Ag(111)-ð 3 Â 19ÞR23:4 and is shown sche-
matically in Fig. 3b. The charge density for the for-
mation of a Cd adlayer with such structure,
calculated assuming ideal charge/coverage stoichio-
metry (i.e. ðoq=oC 2þ Þ ¼ 2F ) is q = 111 lC/
potentials of both Ag and Cd [30,31]. Thus, we
suggest that the overlayer contents also adsorbed
sulfate anions with a coverage remaining nearly
constant during the UPDof Cd. A similar model
has been proposed recently also for the UPDof
Cd
E
calc
2
cm . This value is quite close to the experimental
2
cathodic charge density jDqexpj ꢂ 120 lC/cm ,
+
2ꢀ
obtained from the integration of the cyclic voltam-
mogram in the UPDrange 100 mV 6 DE 6
Ag in the system Au(111)/Ag , SO [32] and
2
4
+
for the UPDof Cd in the systems Cu(111)/Cd
ꢀ
,
2+
ꢀ
400 mV. However, this UPDrange is located at
potentials more positive than the zero charge
Cl [33] and Ag(111)/Cd , Cl [27]. We suppose
that the specific overlayer structure observed in the

S.G. Garc ı´ a et al. / Surface Science 576 (2005) 9–18
13
present study (cf. Fig. 3) is stabilized by the coad-
sorbed sulfate anions, as already suggested for
other UPDsystems [1,18,32,33]. The identity of
and jDq j) indicates that the spots in the STM
exp
image could be attributed to the Cd adatoms.
The sequence of in situ STM images shown in
Fig. 4 presents the changes of surface topography
during the UPDof Cd. Fig. 4a shows the initial Cd
free substrate surface at an underpotential DE =
450 mV. No changes of this surface topography
2ꢀ
the species (Cd or SO ) corresponding to the
4
spots seen in the STM image in Fig. 3a is un-
known. However, the consistency between the the-
oretical and experimental charge densities (qcalc
2+
2ꢀ
Fig. 4. Sequence of in situ STM images obtained in the system Ag(111)/Cd , SO at different polarization conditions. (a) Surface
topography of the initial Ag(111) surface at DE = 450 mV. Grayscale range: 8 nm; Ebias = ꢀ550 mV; Itun = 15 nA. (b–d) Topography
4
changes in function of t
range: 8 nm; Itun = 15 nA. (e) Atomic structure of the condensed close packed monolayer at DE = 70 mV. Grayscale range: 15 nA;
p
after a potential step from DE = 125 mV (Ebias = ꢀ875 mV) to DE = 70 mV (Ebias = ꢀ930 mV); Grayscale
E
bias = ꢀ930 mV; Itun = 15 nA. (f) Surface topography after changing the underpotential to DE = 10 mV. Grayscale range: 8 nm;
Ebias = ꢀ990 mV; Itun = 15 nA. The arrows in (b), (c) and (f) indicate the formation of the 1st and 2nd monolayers (ML).

14
S.G. Garc ı´ a et al. / Surface Science 576 (2005) 9–18
could be observed in the UPDrange 100 mV 6
DE 6 400 mV, where the formation of the ex-
panded Cd adlayer (cf. Fig. 3) takes place. As
shown in Fig. 4b, however, a growth front appears
in the in situ STM image after stepping the under-
potential from DE = 125 mV to a value of DE =
7
0 mV, corresponding to the peak A of the cyclic
3
voltammogram. This observation gives clear evi-
dence for the occurrence of a first order 2Dtrans-
formation of the expanded gas-like 2DCd adlayer
to a condensed Cd monolayer (ML). The ML for-
mation starts preferentially at monatomic step
edges (cf. Fig. 4b), whereas the formation of 2D
Cd islands of the condensed ML on top of sub-
strate islands (i.e. on atomically smooth substrate
terraces) is observed at longer times (cf. Fig. 4c).
This behavior can be attributed to the lower bar-
rier for 2Dnucleation and/or to the higher resi-
dence time of Cd adatoms at monatomic step
edges, in good agreement with theoretical predic-
tions [1,8]. The completion of the condensed
monolayer coverage of the substrate surface is
shown in Fig. 4c and Fig. 4d. The in situ STM
imaging of the condensed monolayer with atomic
resolution shows a close-packed 2Datomic
structure (Fig. 4e) characterized by an interatomic
spacing of 0.3 ± 0.02 nm, which is close to the
nearest-neighbor distance in the bulk Cd crystal
lattice. The observed formation of a close-packed
Cd monolayer under these conditions correlates
Fig. 5. Anodic stripping curves obtained in the system
2ꢀ
2
+
Ag(111)/Cd , SO4 after different polarization times, t
DE = 70 mV (dE/dt = 10 mV/s).
p
at
the stripping peak D to more positive potentials,
3
indicate the occurrence of surface transformations.
The corresponding stripping charge density, Dq,
however, does not change with t and is limited
p
to the charge density required for the formation
of a closed packed Cd monolayer. These results
suggest that under these conditions the close
packed Cd monolayer formed at DE = 70 mV
transforms to a very thin surface alloy layer by a
place exchange between Cd atoms and surface
Ag atoms. The occurrence of these processes is
supported by the behavior observed during a
repetitive potential cycling within the potential
range 65 mV 6 DE 6 155 mV. As shown in Fig.
6, a cycling for more than one hour leads to almost
complete disappearance of voltammetric peaks A3
well with the experimental charge density of
2
jDq j ꢂ 410 lC/cm obtained in the UPDrange
exp
5
STM image obtained after application of a lower
0 mV 6 DE 6 400 mV. Fig. 4f shows an in situ
underpotential, DE = 10 mV, corresponding to
the peak A of the cyclic voltammogram. A forma-
and D . A further Cd deposition does not occur in
3
4
tion of a second Cd monolayer is recognized in the
STM image (cf. the arrow in Fig. 4f). It should be
noted, that extended polarization under these con-
ditions (electrode potentials close to E3DCd) leads
to instability of the STM imaging (cf. Fig. 4f).
Therefore, it was not possible to perform a more
detailed in situ STM study of long time surface
topography changes in this underpotential range.
Fig. 5 shows anodic stripping curves obtained
after extended polarization at DE = 70 mV. The
observed change of stripping peaks with increasing
this potential range because the UPDof Cd on the
alloyed surface requires more negative potentials
(lower underpotentials DE). The stripping curve
obtained after the long time cyclic polarization
(curve (3) in Fig. 6) shows two new stripping peaks
D01 and D02 located at more anodic potentials.
The corresponding stripping charge density
2
(jDqj = 390 ± 20 lC/cm ) is very close to that
required for the formation of a close packed
Cd monolayer. The electrochemical behavior
shown in Fig. 6 is very similar to that observed
during the surface alloy formation in the system
polarization time, t and in particular the shift of
p

S.G. Garc ı´ a et al. / Surface Science 576 (2005) 9–18
15
Fig. 6. Voltammetric adsorption/desorption behavior of the
2+
2ꢀ
system Ag(111)/Cd , SO4 after repetitive cycling within the
potential range 65 mV 6 E 6 155 mV (dE/dt = 10 mV/s). (1)
CV after 1 min potential cycling. (2) CV after 60 min potential
cycling. (3) Anodic stripping curve after 60 min potential
cycling.
2
+
Ag(111)/Pb [1,34–38]. The present results, how-
ever, show that in contrast to the system Ag(111)/
2
Pb , the surface alloying in the system Ag(111)/
+
2
+
Cd seems to be not accompanied by a measur-
able partial desorption of the original Cd mono-
layer coverage. More detailed voltammetric and
in situ STM investigations are in progress in order
to clarify the kinetics and mechanism of surface al-
loy formation processes occurring under these
conditions.
Fig. 7. (a) Anodic stripping curves obtained in the system
2+
2ꢀ
Ag(111)/Cd , SO4 after different polarization times, t
DE = 10 mV (dE/dt = 10 mV/s). (b) Stripping charge density,
Dq, as a function of polarization time, t at DE = 10 mV.
p
at
p
The formation of the second Cd monolayer
indicated above (cf. Fig. 4f) is accompanied by a
significant surface alloy formation, which is re-
flected in the anodic stripping curves obtained
after extended polarization at DE = 10 mV (Fig.
parabolic dependence of Dq on t at relatively long
p
polarization times (t > 100 s) has been discussed
previously in terms of a model including non-sta-
tionary mutual diffusion of Ag and Cd in the Ag/
p
7
similar to those observed in Fig. 5 and show the
a). The changes of the stripping curves are very
2
+
Ag–Cd/Cd system [12]. It has been suggested
that the alloy formation process proceeds by a
movement of Ag atoms through a highly distorted
(vacancy-rich) Ag–Cd alloy layer and simulta-
0
appearance of an additional stripping peak D at
3
long polarization times (t > 100 s). In contrast
to the data in Fig. 5, however, with increasing
polarization time, t the stripping charge density
p
2
+
neous Cd deposition at the Ag–Cd/Cd interface.
1
p
=2
p
Dq increases significantly, exceeding that required
for the deposition of two close-packed Cd mono-
layers (Fig. 7b). Experimental data presented in
The linear Dq ꢀ t dependence observed in Fig.
7b at short polarization times can be attributed
to the occurrence of place exchange processes lead-
ing to a formation of a thin initial Ag–Cd surface
alloy film on the substrate surface.
1=2
Fig. 7b show two different linear Dq ꢀ t depen-
p
dences for short and long polarization times. The

16
S.G. Garc ı´ a et al. / Surface Science 576 (2005) 9–18
The results presented in Fig. 7b are very similar
to those reported recently by Vidu and Hara for the
topography changes observed after anodic strip-
ping. A typical in situ STM image obtained after
extended polarization at E = 10 mV and subse-
quent anodic stripping at DE = 450 mV is shown
in Fig. 9. As seen the removal of Cd leads to an
appearance of new terraces and a large number
of pits and 2Dislands of monatomic thickness
(cf. Fig. 4a and Fig. 9). This behavior is character-
istic for alloy dissolution and has been observed
2
+
2+
systems Au(100)/Cd [22] and Ag(100)/Cd [24].
These authors proposed a diffusion model, which
can be applied also for the analysis of kinetic data
in Fig. 7b. In this model the Ag–Cd alloy layer is
considered as consisting of two regions character-
ized by different diffusion properties as illustrated
schematically in Fig. 8 [22]. The region ꢀl < x < 0
corresponds to a thin surface alloy layer (‘‘skin’’)
and is characterized by an effective diffusion coef-
ficient D which is larger than the diffusion coeffi-
1
cient D for the rest region x > 0 (D ꢃ D ). The
2
1
2
Cd concentration c at the electrode/electrolyte
0
interface (x = ꢀl) is considered to be constant for
all times. Present in situ STM observations showed
2
+
that in the system Ag(111)/Cd the alloy forma-
tion process is preceded by the formation of a close
packed Cd monolayer (cf. Fig. 4). Thus, in the
present case we considered a surface concentration
3
of c = 0.0735 mol/cm corresponding to a close
0
packed Cd monolayer. The diffusion coefficients
D1 and D were estimated from the linear
2
1=2
Dq ꢀ t dependences in Fig. 7b following the pro-
p
cedure described in Ref. [22]. Values of
D = 1.5 · 10
ꢀ16
2
ꢀ19
2
cm /s
cm /s and D = 3.5 · 10
1
2
were found in good agreement with those reported
for the surface alloy formation in the system
Fig. 9. In situ STM image obtained in the system Ag(111)/
2
Ag(100)/Cd [24].
+
2+
2ꢀ
Cd , SO4 after 10 min polarization at DE = 10 mV and
subsequent anodic stripping at DE = 450 mV. Grayscale range:
5 nm; Ebias = ꢀ200 mV; Itun = 15 nA.
The surface alloy formation occurring in the
system is recognized also by substantial surface
Fig. 8. Schematic representation of the diffusion model [22] applied for the analysis of experimental results shown in Fig. 7b.

S.G. Garc ı´ a et al. / Surface Science 576 (2005) 9–18
17
2
+
2ꢀ
Fig. 10. (a) In situ STM image obtained in the system Ag(111)/Cd , SO after 2 min polarization at DE = 10 mV and subsequent
4
anodic stripping at DE = 450 mV. Grayscale range: 5 nm; Ebias = ꢀ300 mV; Itun = 10 nA. (b–d) Topography changes after potential
step from DE = 450 mV to DE = 650 mV. Grayscale range: 5 nm; Ebias = ꢀ100 mV; Itun = 10 nA.
previously in other UPDsystems exhibiting sur-
face alloying [1]. Anodic formation of pits and is-
lands on terraces is observed even after relatively
short polarization time at DE = 10 mV (Fig. 10).
The clustering of vacancies created during removal
of Cd from the surface alloy at DE = 450 mV leads
to formation of larger pits, which disappear
quickly by changing the underpotential to DE =
Voltammetric measurements indicate that the
close packed monolayer is not stable and under-
goes slow transformations involving a place ex-
change between Cd atoms and Ag surface atoms.
At lower underpotentials DE < 50 mV the UPD
process continues with the formation of a second
Cd monolayer, which is accompanied by a signifi-
cant Ag–Cd surface alloying. Kinetic results are in
agreement with those reported recently by Vidu
and Hara [24] for the surface alloy formation in
the system Ag(100)/Cd . The anodic dealloying
leads to an appearance of a large number of 2Dis-
lands and monatomically deep pits. The pits disap-
pear quickly at high underpotentials DE > 550 mV
suggesting a relatively high mobility of surface Ag
atoms under these conditions.
6
50 mV (cf. Fig. 10a–d). This behavior suggests a
relatively high mobility of Ag surface atoms at this
substrate potential in good agreement with obser-
vations reported previously for Ag(100) substrates
2
+
[
39].
4
. Conclusions
Present in situ STM studies show that the UPD
of Cd on Ag(111) starts with formation of an
Acknowledgments
expanded adlayer with a superlattice structure
ꢁ
pffi ffiffi pffiffiffiffiffi
Ag(111)-ð 3 Â 19ÞR23:4 . This adlayer trans-
The authors gratefully acknowledge the finan-
cial support of this work by the Universidad Nac-
ional del Sur, the Forschungszentrum J u¨ lich, and
forms in the underpotential range 50 mV 6 DE 6
8
0 mV to a condensed close packed Cd monolayer.

18
S.G. Garc ı´ a et al. / Surface Science 576 (2005) 9–18
the Deutscher Akademischer Austauschdienst
DAAD).
[17] J.C. Bondos, A.A. Gewirth, R.G. Nuzzo, J. Phys. Chem.
00 (1996) 8617.
1
(
[
[
[
18] S-J. Hsieh, A.A. Gewirth, Langmuir 16 (2000) 9501.
19] R. Vidu, S. Hara, Scr. Mater. 41 (1999) 617.
20] R. Vidu, S. Hara, J. Electroanal. Chem. 475 (1999) 171.
References
[21] R. Vidu, S. Hara, J. Vac. Sci. Technol. B 17 (1999)
423.
[22] R. Vidu, S. Hara, Surf. Sci. 452 (2000) 229.
2
[
1] E. Budevski, G. Staikov, W.J. Lorenz, Electrochemical
Phase Formation and Growth, VCH, Weinheim, 1996.
2] K. Itaya, Prog. Surf. Sci. 58 (1998) 121.
´
[23] M.C. del Barrio, S.G. Garcıa, D.R. Salinas, Electrochem.
Comm. 6 (2004) 762.
[
[
3] T.P. Moffat, in: A.J. Bard, I. Rubinstein (Eds.), Electro-
analytical Chemistry, vol. 21, Marcel Dekker, New York,
[24] R. Vidu, N. Hirai, S. Hara, Phys. Chem. Chem. Phys. 3
(2001) 3320.
1
999, p. 211.
[25] S. Maupai, Y. Zhang, P. Schmuki, Electrochem. Solid-
State Lett. 6 (2003) C63.
[
[
4] R. Nichols, in: J. Lipkowski, P.N. Ross (Eds.), Imaging of
Surfaces and Interfaces, Wiley-VCH, New York, 1999, p.
[26] S. Maupai, Y. Zhang, P. Schmuki, Surf. Sci. 527 (2003)
L165.
99.
5] D.M. Kolb, in: R.C. Alkire, D.M. Kolb (Eds.), Advances
in Electrochemical Science and Engineering, vol. 7, Wiley-
VCH, Weinheim, 2001, p. 107.
[27] V.D. Jovic, B.M. Jovic, Electrochim. Acta 47 (2002)
1777.
[28] M.L. Foresti, G. Pezzatini, M. Cavallini, G. Aloisi, M.
Innocenti, R. Guidelli, J. Phys. Chem. B 102 (1998)
7413.
[
[
6] O.M. Magnussen, Chem. Rev. 102 (2002) 679.
7] Th. Wandlowski, in: M. Urbakh, E. Gileadi (Eds.),
Encyclopedia of Electrochemistry, vol. 1, Wiley-VCH,
New York, 2002, p. 383.
[29] S. Garcia, D.R. Salinas, C.E. Mayer, W.J. Lorenz, G.
Staikov, Electrochim. Acta 48 (2003) 1279.
[30] A. Hamelin, T. Vitanov, E. Sevastyanov, A. Popov, J.
Electroanal. Chem. 145 (1983) 225.
[
[
8] G. Staikov, W.J. Lorenz, E. Budevski, in: J. Lipkowski,
P.N. Ross (Eds.), Imaging of Surfaces and Interfaces,
Wiley-VCH, New York, 1999, p. 1.
[31] D. Leikis, V. Panin, K. Rybalka, J. Electroanal. Chem. 40
(1972) 9.
9] W.J. Lorenz, G. Staikov, W. Schindler, W. Wiesbeck, J.
Electrochem. Soc. 149 (2002) K47.
[32] S.G. Garcia, D.R. Salinas, C.E. Mayer, E. Schmidt, G.
Staikov, W.J. Lorenz, Electrochim. Acta 43 (1998) 3007.
[33] C. Stuhlmann, Z. Park, C. Bach, K. Wandelt, Electrochim.
Acta 44 (1998) 993.
[
[
[
[
[
[
10] E. Schmidt, M. Christen, P. Beyeler, J. Electroanal. Chem.
2 (1973) 275.
4
11] J.W. Schultze, F.D. Kopitz, M.M. Lohrengel, Ber. Buns-
enges. Phys. Chem. 78 (1974) 693.
[34] E. Schmidt, H. Siegenthaler, J. Electroanal. Chem. 150
(1983) 59.
12] H. Bort, K. J u¨ ttner, W.J. Lorenz, G. Staikov, Electrochim.
Acta 28 (1983) 993.
[35] T. Vitanov, A. Popov, G. Staikov, E. Budevski, W.J.
Lorenz, E. Schmidt, Electrochim. Acta 31 (1986) 981.
[36] D. Carnal, P.I. Oden, U. M u¨ ller, E. Schmidt, H. Siegent-
haler, Electrochim. Acta 40 (1995) 1223.
[37] E. Ammann, P.I. Oden, H. Siegenthaler, in: W.J. Lorenz,
W. Plieth (Eds.), Electrochemical Nanotechnology,
Wiley-VCH, Weinheim, New York, 1998, p. 1.
[38] R. Widmer, H. Siegenthaler, J. Electrochem. Soc. 151
(2004) E238.
13] V.D. Jovic, B.M. Jovic, A.R. Despic, J. Electroanal.
Chem. 288 (1990) 229.
14] M.E. Martins, A. Hernandes-Creus, R.C. Salvarezza, A.J.
Arvia, J. Electroanal. Chem. 375 (1994) 141.
15] J.L. Stickney, in: A.J. Bard, I. Rubinstein (Eds.), Electro-
analytical Chemistry, vol. 21, Marcel Dekker, New York,
1
999, p. 75.
[
16] J.L. Stickney, in: R.C. Alkire, D.M. Kolb (Eds.), Advances
in Electrochemical Science and Engineering, vol. 7, Wiley-
VCH, Weinheim, 2001, p. 1.
[39] N. Hirai, H. Tanaka, S. Hara, Appl. Surf. Sci. 130–132
(1998) 506.