Nucleation and growth in electrodeposition of metals on n-Si(111)

Article abstract of DOI:10.1016/S0013-4686(00)00418-7

Initial stages of electrodeposition of Tl, Cd and Cu on n-Si(111) from sulfate electrolyte solutions are studied using cyclic voltammetry and chronoamperometry. Capacitance measurements in metal ion free solutions were used to determine the flat band potential of n-Si. Results show that the relative position of the metal equilibrium potential with respect to the substrate flat band potential influences significantly nucleation and growth kinetics. Within appropriate potential ranges the initial deposition kinetics corresponds to a model including progressive nucleation and diffusion controlled cluster growth. Nucleation rate and the number of atoms in the critical nucleus are determined from the analysis of current transients at different overpotentials. Results are compared with data obtained previously for electrochemical nucleation of these metals on other semiconductors and foreign metal substrates.

Full text of DOI:10.1016/S0013-4686(00)00418-7

Electrochimica Acta 45 (2000) 3255–3262
www.elsevier.nl/locate/electacta
Nucleation and growth in electrodeposition of metals on
n-Si(111)
a
b
a
a,
R. Krumm , B. Guel , C. Schmitz , G. Staikov *
a
Institut f u¨ r Physikalische Chemie und Elektrochemie, Heinrich-Heine-Uni6ersit a¨ t D u¨ sseldorf, Uni6ersit a¨ tsstraße 1, Geb. 26.32,
4
0225 D u¨ sseldorf, Germany
b
Facult e´ des Sciences et Techniques, Uni6ersi e´ de Ouagadougou, Ouagadougou, Burkina Faso
Received 30 November 1999; received in revised form 15 January 2000
Abstract
Initial stages of electrodeposition of Tl, Cd and Cu on n-Si(111) from sulfate electrolyte solutions are studied using
cyclic voltammetry and chronoamperometry. Capacitance measurements in metal ion free solutions were used to
determine the flat band potential of n-Si. Results show that the relative position of the metal equilibrium potential
with respect to the substrate flat band potential influences significantly nucleation and growth kinetics. Within
appropriate potential ranges the initial deposition kinetics corresponds to a model including progressive nucleation
and diffusion controlled cluster growth. Nucleation rate and the number of atoms in the critical nucleus are
determined from the analysis of current transients at different overpotentials. Results are compared with data
obtained previously for electrochemical nucleation of these metals on other semiconductors and foreign metal
substrates. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Electrodeposition; Nucleation; Crystal growth; Semiconductors; Flat band potential
(
El)
1
. Introduction
corresponding electrochemical potentials (v˜
=
−
Mez+
(El)
z+
(Me)
Me
v˜ z+). The deviation from equilibrium D v˜ = v˜ Me
Electrochemical deposition and dissolution of metals
(Me)
Mez+
v˜
is given by:
can be generally expressed by
D v˜ = −zF(E−EMe/Mez+)= −zFp
(2)
z+
solv
z+
Me (El)UMe (Me)
(1)
where E and p are the actual electrode potential and the
overpotential, respectively. If the kinetics of metal de-
position and dissolution are not controlled by charge
transfer, ion transport or additional chemical reaction
steps, p represents the so-called crystallization overpo-
tential [1]. In this case Eq. (2) defines the thermody-
namic driving forces of the phase formation and
dissolution processes, i.e. supersaturation (D v˜ \0 for
pB0) and undersaturation (D v˜ \0 for pB0).
z+
solv
where Me (El) represents the metal ions in the elec-
z+
trolyte phase (El) and Me (Me) denotes the metal
ions in the metal bulk phase (Me), which are coupled to
−
z+
the electrons e (Me) in the Me-crystal lattice (Me
−
(
Me)+ze (Me)=Me). At the so-called Nernst poten-
z+
solv
z+
tial EMe/Mez+, Me (El) and Me (Me) are in
equilibrium, which is defined by the equality of the
Mechanism of metal electrodeposition on foreign
substrates (S) depends strongly on the Me–S interac-
tion [1]. In systems with weak Me–S interaction (weak
adhesion) metal deposition starts at supersaturation in
*
Corresponding author. Tel.: +49-211-8113686; fax: +49-
2
11-8112803.
E-mail address: georgi.staikov@uni-duesseldorf.de (G.
Staikov)
0
013-4686/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0013-4686(00)00418-7

3
256
R. Krumm et al. / Electrochimica Acta 45 (2000) 3255–3262
the so-called overpotential deposition (OPD) range EB
EMe/Mez+ with nucleation and growth of the 3D Me
bulk phase. In the case of a strong Me–S interaction
band gap, and is more positive than the flat band
potential Efb (EMe/Mez+ ꢁE ) a depletion layer is
fb
formed in the n-type semiconductor at E=EMe/Mez+.
Thus, electrode potentials more negative than the flat
band potential (EBEfb) are needed in this case to reach
a sufficient surface electron density (accumulation con-
ditions) and electrodeposition of metals in such systems
occurs at high cathodic overpotentials (Fig. 1b).
Mechanism of metal deposition and structures of
electrodeposits on foreign metal substrates have been
extensively studied using various electrochemical meth-
ods and in situ SPM (scanning probe microscopy)
techniques [1]. Despite its technological importance,
electrodeposition of metals on semiconductor substrates
has been the object of relatively few studies to date
(
strong adhesion), however, the deposition process can
start even at undersaturation in the so-called underpo-
tential deposition (UPD) range E\EMe/Mez+ with for-
mation of low-dimensional metal phases, which act as
precursors for the nucleation and growth of the 3D Me
bulk phase in the OPD range.
Kinetics and mechanism of metal electrodeposition
and the involved phase formation phenomena can also
be influenced significantly by electronic properties of
the foreign substrate. In the case of deposition on
foreign metal substrates the substrate surface is a con-
tinuum of electronic states and thus, sufficient supersat-
urations for nucleation of the Me bulk phase are
usually reached by application of relatively low ca-
thodic overpotentials. The situation, however, becomes
more complicated in the case of electrodeposition of
metals on semiconductor substrates. The influence of
the space charge layer in the semiconductor has to be
taken into account in this case [2–6]. For a not very
high doped n-type semiconductor, the surface concen-
tration of electrons and the band bending at the actual
electrode potential play an important role in the elec-
trodeposition process. Neglecting the influence of sur-
face states, the mechanism of metal deposition in this
case depends strongly on the relative position of the
equilibrium potential EMe/Mez+with respect to the flat
band potential Efb of the semiconductor substrate (Fig.
[
2–12]. Recent investigations of metal electrodeposition
on n-GaAs(100) [7,8], n-Si(111) [9–11] and n-Si(100)
12] provided new important information about the
[
nucleation and growth kinetics, and the role of sub-
strate surface inhomogeneities in the deposition pro-
cess. More studies are needed, however, to understand
the complex nature and mechanism of electrochemical
metal phase formation on semiconductor substrates.
In this paper we present new results on the mecha-
nism and kinetics of the initial stages of metal elec-
trodeposition, on H-terminated n-Si(111) substrates
from acidic sulfate electrolyte solutions. The systems
+
2+
n-Si(111)/Tl and n-Si(111)/Cd
are selected as typi-
cal examples for systems characterized by E
ꢀ
Me/Mez+
2+
E , whereas n-Si(111)/Cu
is chosen as an example
fb
for a system with EMe/Mez+ ꢁE . The experimental
fb
1). If EMe/Mez+ ꢀEfb accumulation layer is formed at
results are compared with data obtained by electrode-
position of the same metals on other substrates.
E=EMe/Mez+and the deposition of the Me bulk phase
usually occurs at relatively small cathodic overpoten-
tials, such as on metal substrates (Fig. 1a). However, if
the equilibrium potential EMe/Mez+, is located in the
2. Experimental
The experiments were carried out in the following
systems:
1
2
3
. n-Si(111)/5 mM Tl SO , 0.5 M Na SO , 5 mM
2 4 2 4
H SO (pH 2),
2
4
. n-Si(111)/5 mM CdSO , 0.5 M Na SO , 5 mM
4
2
4
H SO (pH 2),
2
4
. n-Si(111)/10 mM CuSO , 0.5 M H SO (pH 0.3).
4
2
4
Electrolyte solutions were prepared from Suprapure
®
chemicals and Millopore water. Prior to each mea-
surement, the solutions were deaerated in the electro-
chemical cell with pure nitrogen (99.999%). The
electrochemical cell consisted of a Pt counter electrode,
a reference Hg/HgSO /Na SO (sat.) electrode (MSE)
4
2
4
and a n-Si(111) working electrode. All electrode poten-
tials are referred to the SHE potential and/or to the
corresponding Nernst equilibrium potential EMe/Mez+.
Silicon working electrodes were made from n-doped
Si(111) wafers (Siltronix, France) of 1–5 V cm resistiv-
ity and 0° misorientation. Samples were subsequently
Fig. 1. Schematic band diagrams for electrodeposition of
different metals (Me1 and Me ) on a n-semiconductor with a
2
flatband potential E . Relative positions of corresponding
fb
metal equilibrium potentials with respect to Efb are also indi-
cated. (a) EMe1/Me ꢀEfb; deposition potential E , (b)
z+
1
1
EMe2/Me ꢁE ; deposition potential E .
z+
fb
2
2

R. Krumm et al. / Electrochimica Acta 45 (2000) 3255–3262
3257
3. Results and discussion
3.1. Electrochemical beha6ior of n-Si(111) in metal ion
free electrolytes
Fig. 2a shows cyclic voltammograms of n-Si(111) in
metal ion free electrolytes at pH 0.3 and 2. The sharp
current increase at −0.6 and −0.8 V versus SHE is
related to the H evolution reaction. Potential depen-
2
dencies of space charge capacitance measured under
depletion conditions are shown in Fig. 2b. From the
−
2
corresponding Mott–Schottky (C –E) plots flatband
potentials of E = −0.24 and −0.14 V versus SHE
fb
could be estimated for the solutions at pH 2 and 0.3,
respectively. In Fig. 3 these values are compared with
results obtained previously by Allongue et al. [13], in
NH F solutions at various pH values and by Itaya et
4
al. [14] in H SO solution at pH 1. All data seem to
2
4
correlate well to a linear E –pH dependence with a
fb
slope of −45 mV/pH deviating from the Nernstian
slope of −60 mV/pH observed in many other systems.
The results indicate that anions do not affect signifi-
cantly the flat band potential of H-terminated n-Si(111)
substrates.
Fig. 2. Cyclic voltammograms (a) (cathodic sweep) and Mott–
2
−
+
Schottky plots (b) in the system: n-Si(111)/ SO , H (pH 0.3
4
−
1
and 2). Scan rate ꢀdE/dtꢀ=5 mV s ; T=298 K; capacitance
measurements at f=1014 Hz.
3.2. Deposition of metals with EMe/Mez+ ꢀEfb
Equilibrium potentials of Tl and Cd in systems (i)
and (ii) are about 270 mV more negative than the
flatband potential of n-Si(111) in the corresponding
cleaned in trichloroethylene, acetone and ethanol and
etched for 1 min in 2% HF solution and 6 min in
electrolyte solutions (E = −0.24 V). Fig. 4 shows
fb
deaerated 40% NH F solution. Si(111) surfaces pre-
4
typical cyclic voltammograms of deposition and disso-
lution of Tl and Cd on n-Si(111) in systems (i) and (ii),
respectively. Reversible deposition/dissolution reactions
observed in both systems indicate a formation of n-
Si(111)/Me contacts with ohmic behavior. Equal charge
densities are estimated for metal deposition and dissolu-
pared in such a manner are H-terminated and exhibit
atomically flat terraces. The substrates were mounted
into a Teflon holder and have an active surface area of
2
0
.3 cm . An ohmic contact at the back of the substrate
face was established by GaꢀIn alloy. All experiments
were performed at T=298 K in dark using standard
electrochemical measurement equipment.
Fig. 4. Cyclic voltammograms for Tl and Cd deposition and
z+
2−
+
z+
Fig. 3. Flat band potentials Efb of n-Si in NH F and H SO
electrolyte solutions at different pH values.
dissolution in the systems: n-Si(111)/Me , SO , H ( Me
4
2
4
4
+
2+
−1
=Tl and Cd , pH 2). ꢀdE/dtꢀ=1 mV s ; T=298 K.

3
258
R. Krumm et al. / Electrochimica Acta 45 (2000) 3255–3262
tion processes in both cases, which excludes the occur-
rence of a parallel H evolution reaction. No UPD of
2
Tl and Cd could be observed in any of the two systems
indicating a weak interaction (weak adhesion) of these
metals with n-Si(111). A comparison of cyclic voltam-
mograms in Fig. 4 shows that initial deposition kinetics
of Tl is faster. This is reflected in the higher onset
overpotential for deposition.
Nucleation and growth kinetics in the initial stages of
Tl and Cd deposition on n-Si(111) were studied using
the current transient technique. Figs. 5a and 6a show
typical current transients for deposition of Tl and Cd at
relatively low cathodic overpotentials. The initial parts
of the current transients are represented in Figs. 5b and
2
/3
6b as i
versus t plots. The obtained linear relation-
free
ships show that the kinetics of the initial stages of metal
deposition in these systems can be explained by a model
including progressive nucleation and cluster growth
controlled by hemispherical diffusion [1,15]. According
to this model the current density ifree(t) at relatively
short times, i.e. for deposition without overlapping of
diffusion zones, can be generally described by the
equation
3/
2
2
1/
3/
ꢁ ꢂzFpꢃn
2
2
ifree=−zF yJ(Vm) (2DcMez+) 1−exp
3
RT
Fig. 5. System n-Si(111)/Tl , SO 24 −, H (pH 2). T=298 K.
+
+
3/
2
(
a) Current transients for Tl deposition at different overpoten-
(t−to)
(3)
2/3
tials p; (b) i
vs. t plot of the initial parts of the transients
free
where J represents the nucleation rate, t the induction
o
shown in (a).
period of nucleation, Vm the molar volume of Me, D
the diffusion coefficient and c
the concentration of
Mez+
z+
Me
in the electrolyte. At relatively high cathodic
overpotentials the nucleation sites are virtually con-
verted to nuclei instantaneously and the initial current
1/2
density ifree is proportional to t . With increasing t
hemispherical diffusion zones of the growing clusters
start to overlap leading to the appearance of a maxi-
mum in current transients. A theoretical model taking
into account the overlapping of diffusion zones has
been developed by Gunawardena et al. [16]. According
to this model the overall current transients for progres-
sive and instantaneous nucleation can be described by
2
ꢂ
ꢂ
ꢂ
ꢂ
i
ꢃ
(t−t
o
)
max
ꢁ
=
1.9542
1−exp
imax
(t−to)
inst
2
(t−t
)
ꢃn
o
−
1.2564
(4)
(
t−t )
o max
2
i
ꢃ
(t−t
)
ꢁ
o
max
=
1.2254
1−exp
imax
(t−t )
prog
o
2
2
(t−t
)
ꢃn
o
2
−
2.3367
(5)
(
t−t )
o max
2
max
The products i (t−t )
are given, respectively, by
o max
Fig. 6. System n-Si(111)/Cd2 , SO 42 −, H (pH 2). T=298 K.
+
+
2
max
[i
(t−t )
]
o max inst
(
a) Current transients for Cd deposition at different overpo-
2/3
tentials p; (b) i
shown in (a).
vs. t plot of the initial parts of the transients
2
free
ꢁ ꢂzFpꢃn
2
=0.1629(zFcMez+) 1−exp
D
(6)
RT

R. Krumm et al. / Electrochimica Acta 45 (2000) 3255–3262
3259
2
max
[
i
(t−t ) ]
o max prog
2
ꢁ
ꢂ
zFpꢃn
2
=
0.2598(zFcMez+) 1−exp
D
(7)
RT
Fig. 7a represents as a typical example, a current
transient for Tl deposition at an overpotential p= −15
mV. The corresponding (i/imax) versus (t/tmax) plot in
2
Fig. 7b shows, that at this relatively high cathodic
overpotential the nucleation of Tl can be considered as
instantaneous. Similar behavior was observed also in
the case of Cd deposition at relatively high cathodic
overpotentials. From experimental data and Eq. (6)
−
5
diffusion coefficients of D=1.4×10
and D=9.5×
−
6
2
−1
+
2+
Fig. 8. Nucleation rates J as a function of supersaturation
10
cm
s
were estimated for Tl and Cd
,
(
zFꢀpꢀ) for electrodeposition of Tl and Cd on n-Si(111) derived
respectively. These values and Eq. (3) were used to
from experimental data in Figs. 5 and 6. Previous results
obtained in the system n-Si(111)/Pb
son.
derive nucleation rates J for Tl and Cd deposition at
2+
are shown for compari-
2
/3
low cathodic overpotentials from the slopes of the i
free
versus t plots presented in Figs. 5b and 6b. In the initial
stages of Tl and Cd electrodeposition on n-Si(111) the
driving force of nucleation, the supersaturation D v˜ , is
related to the overpotential by Eq. (2). Thus, in Fig. 8
the corresponding nucleation rates J are presented as a
function of supersaturation D v˜ =zFꢀpꢀ. Experimental
2
+
data obtained previously in the system n-Si(111)/Pb
9,10] are also shown for comparison.
The number of atoms Ncrit in the critical nuclei can
[
be determined using the equation
d ln J
Ncrit=RT
(8)
d(zFꢀpꢀ)
This equation is general and can be applied without any
restrictions for the size and form of the nucleus [1,17].
It allows an estimation of Ncrit with an accuracy of
about 90.5. Relatively small Ncrit-values of 18, 11 and
6
were derived from the slopes of the linear dependen-
cies in Fig. 8 for Tl, Pb and Cd, respectively. The small
values show that the so-called atomistic approach [18–
20] has to be applied to describe the nucleation pro-
cesses in these systems.
3.3. Deposition of metals with EMe/Mez+ ꢁEfb
Fig. 9 shows a cyclic voltammogram of the system
2
+
n-Si(111)/Cu . The voltammogram is characterized by
a relatively high critical cathodic overpotential for de-
position and an absence of an anodic stripping peak. A
similar electrochemical behavior was observed also by
Cu electrodeposition on n-GaAs(100) [7,8] and n-
Si(100) [12] and seems to be typical for many systems in
which the equilibrium potential EMe/Mez+ is located in
the band gap of the n-type semiconductor substrate and
is more positive than the flat band potential E . It is
fb
interesting to note, that in the absence of surface states,
a formation of low dimensional metal phases by UPD
is impossible in systems with EMe/Mez+ ꢁE , even in
Fig. 7. System n-Si(111)/Tl , SO 24 −, H ; T=298 K, pH 2. (a)
+
+
fb
the cases of a strong metal–semiconductor interaction.
Typical current transients for nucleation and growth
of Cu on n-Si(111) at different overpotentials are shown
Current transient for Tl deposition at p= −15 mV (t =0).
0
2
(
b) (i/imax
)
vs. (t−t )/(t−t )
plot of the transient in (a)
0
0 max
compared with theoretical transients for progressive (dashes)
and instantaneous (solid line) nucleation according to Eqs. (4)
and (5).
2
/3
in Fig. 10a. The obtained linear i
versus t relation-
free
ships for the initial parts of the transients (Fig. 10b)

3
260
R. Krumm et al. / Electrochimica Acta 45 (2000) 3255–3262
2
ing a maximum at large times. The normalized (i/imax
)
versus (t−t )/(t−t ) plot of the transient presented
o
o max
in Fig. 11b is in good agreement with Eq. (5), corre-
sponding to the model including progressive nucleation
and cluster growth controlled by hemispherical diffu-
2
+
−5
sion. A Cu
diffusion coefficient of D=1.1×10
2
−1
cm
s
was estimated from experimental data using
Eq. (7). This value is in good agreement with literature
data and was used to derive the nucleation rate J at
2
/3
different overpotentials from the slopes of the i
free
versus t plots in Fig. 10b and Eq. (3). Nucleation rate
as a function of supersaturation zFꢀpꢀ is shown in Fig.
1
2. Applying Eq. (8), a value of N =190.5 was
crit
estimated from experimental data in Fig. 12. In terms
of the so-called small cluster nucleation model [1,18–
2
0] this result indicates that under given conditions a
2
+
Fig. 9. Cyclic voltammogram in the system n-Si(111)/Cu
SO4 , H (pH 0.3), ꢀdE/dtꢀ=5 mV s ; T=298 K.
,
single adatom or an adatom pair represent stable clus-
ters which can grow spontaneously. Ncrit-values ranging
2
−
+
−1
Fig. 10. System n-Si(111)/Cu2 , SO 42 −, H (pH 0.3). T=298
+
+
K. (a) Current transients for Cu deposition at different over-
potentials p; (b) i2 vs. t plot of the initial parts of the
/3
free
Fig. 11. System n-Si(111)/Cu2+, SO 24 −, H+; T=298 K, pH
transients shown in (a).
0
.3. (a) Current transient for Cu deposition at p= −509 mV
2
(
t0=280 ms). (b) (i/imax
)
vs. (t−t )/(t−t )
plot of the
0
0 max
correspond to a progressive nucleation and diffusion
controlled cluster growth described by Eq. (3). Fig. 11a
shows the current transient for p= −509 mV exhibit-
transient in (a) compared with theoretical transients for pro-
gressive (dashes) and instantaneous (solid line) nucleation
according to Eqs. (4) and (5).

R. Krumm et al. / Electrochimica Acta 45 (2000) 3255–3262
. Conclusions
The studies of the early stages of electrodeposition of
3261
4
Tl, Cd and Cu on H-terminated n-Si(111) substrates
presented in this paper demonstrate the important role
of the relative position of EMe/Mez+ with respect to the
substrate flat band potential E . In systems with
fb
+
E
Cd
ꢀEfb such as n-Si(111)/Tl and n-Si(111)/
the deposition process occurs as on metal sub-
Me/Mez+
2
+
strates and the initial nucleation and growth kinetics can
be analyzed at relatively low cathodic overpotentials
(
supersaturations). On the contrary, in systems with
EMe/Mez+ ꢁE , a study of deposition processes is usu-
fb
ally possible only at relatively high cathodic overpoten-
tials corresponding to high supersaturations. The
relative position of the metal equilibrium potential with
respect to the substrate flat band potential can be altered
z+
by a change of the concentration of Me and/or pH as
z+
well as by complexation of Me
introducing complex-
ing agents in the electrolyte solution [12]. Experimental
results for electrochemical nucleation of Cd, Pb, Tl and
Cu on various substrates are compared in Table 1.
Results indicate that the main factors determining the
size of the critical nuclei and the Gibbs energy of nucleus
formation, are supersaturation and Me–Me interaction.
The substrate influence seems to be related primarily to
the density and activity of nucleation sites. The
relatively small number of atoms in the critical
nuclei, obtained even at low supersaturations (low ca-
thodic overpotentials), show that thermodynamics and
Fig. 12. Nucleation rate J as a function of supersaturation
zFꢀpꢀ for Cu electrodeposition on n-Si(111) derived from ex-
perimental data in Fig. 10.
between 0 and 1 were recently found also in other
systems characterized by EMe/Mez+ ꢁEfb [7,8,12]. Such
low Ncrit-values, however, are not specific for semicon-
ductor substrates and are often obtained at high ca-
thodic overpotentials (supersaturations) also by
electrodeposition of metals on foreign metal substrates,
glassy carbon and HOPG [1,20–28].
Table 1
Ncrit-values for electrochemical nucleation of Cd, Cu, Pb and Tl on various substrates in different overpotential (supersaturation)
ranges
Metal
Cd
Substrate
Overpotential range (mV)
Supersaturation range (kJ mol−1)
Ncrit (atoms)
Ref.
n-Si(111)
Pt
−15 to −21
−22 to −30
2.89 to 4.05
4.25 to 5.79
5.79 to 7.33
6
2
1
This paper
[22]
[22]
−
30 to −38
n-Si(111)
n-GaAs(100)
−474 to −520
−510 to −570
91.47 to 100.34
98.41 to 109.99
121.76 to 170.01
13.70 to 15.82
7.72 to 10.42
10.42 to 15.82
15.82 to 27.02
4.25 to 6.17
1 to 0
1 to 0
0
2
4
1
0
11
0
This paper
[7]
[8]
[30]
[25]
[25]
[25]
[31]
[26]
−
631 to −881
Au(111)
Pd
−71 to −82
−40 to −54
Cu
−
−
54 to −82
82 to −140
Pt
W
−22 to −32
50 to −100
−
9.65 to 19.30
7.72 to 10.61
−40 to −55
4
[31]
n-Si(111)
Ag(111)
Ag(100)
C
−6 to −10
−15 to −19
−13 to −18
−50 to −300
−4 to −7
1.16 to 1.93
2.89 to 3.67
2.51 to 3.47
9.65 to 57.89
0.77 to 1.35
11
11
13
0
[9]
[29]
[29]
[24]
[9]
Pb
Tl
HOPG
11
n-Si(111)
−3 to −7
0.29 to 0.68
18
This paper

3
262
R. Krumm et al. / Electrochimica Acta 45 (2000) 3255–3262
kinetics of electrochemical nucleation must be generally
described by the so-called small cluster or atomistic
model [1,18–20].
[10] R.T. P o¨ tzschke, G. Staikov, W.J. Lorenz, W. Wiesbeck,
J. Electrochem. Soc. 146 (1999) 141.
[
[
[
11] J.C. Ziegler, R.I. Wielgosz, D.M. Kolb, Electrochim.
Acta 45 (1999) 827.
12] G. Oskam, J.G. Long, A. Natarajan, P.C. Searson, J.
Phys. D: Appl. Phys. 31 (1927) 1998.
Acknowledgements
13] P. Allongue, V. Kieling, H. Gerischer, Electrochim. Acta
The authors gratefully acknowledge the financial sup-
port of this work by the Deutsche Forschungsgemein-
schaft (DFG), the Arbeitsgemeinschaft industrieller
Forschungs-vereinigungen (AiF), the Bundesminis-
terium f u¨ r Wirtschaft (BMWi), the Ministerium f u¨ r
Schule und Weiterbildung, Wissenschaft und
Forschung des Landes Nordrhein-Westfalen (MSWWF
NRW), and the Deutscher Akademischer Austauschdi-
enst (DAAD).
4
0 (1995) 1353.
[14] E. Tomita, N. Matsuda, K. Itaya, J. Vac. Sci. Technol. A
8 (1990) 534.
[15] J.A. Harrison, H.R. Thirsk, in: A.J. Bard (Ed.), Electro-
analytical Chemistry, vol. 5, Marcel Dekker, New York,
1
971, p. 67.
[
16] G. Gunawardena, G. Hills, I. Montenegro, B. Scharifker,
J. Electroanal. Chem. 138 (1982) 225.
17] D. Kashchiev, J. Chem. Phys. 76 (1982) 5098.
18] S. Stoyanov, Thin Solid Films 18 (1973) 91.
19] A. Milchev, S. Stoyanov, R. Kaischev, Thin Solid Films
22 (1974) 255.
[
[
[
References
[
20] A. Milchev, Contemp. Phys. 32 (1991) 321.
[
1] E. Budevski, G. Staikov, W.J. Lorenz, Electrochemical
Phase Formation and Growth — An Introduction to the
Initial Stages of Metal Deposition, VCH, Weinheim,
[21] A. Milchev, S. Stoyanov, J. Electroanal. Chem. 72 (1976)
33.
[22] S. Toschev, I. Markov, Electrochim. Acta 12 (1967) 281.
[23] A. Milchev, E. Vassileva, J. Electroanal. Chem. 107
(1980) 337.
[24] J. Mostany, J. Mozota, B.R. Scharifker, J. Electroanal.
Chem. 177 (1984) 25.
[25] P.M. Rigano, C. Mayer, T. Chierchie, J. Electroanal.
Chem. 248 (1988) 219.
1
996.
2] P. Bindra, H. Gerischer, D.M. Kolb, J. Electrochem. Soc.
24 (1977) 1012.
[
[
1
3] P. Allongue, in: H. Gerischer, Ch.W. Tobias (Eds.), Ad-
vances in Electrochemical Science and Engineering, vol. 4,
VCH, Weinheim, 1995, p. 1.
[
[
[
[
[
[
4] P. Allongue, E. Souteyrand, J. Electroanal. Chem. 286
[26] A.I. Danilov, E.B. Molodkina, Ju.M. Polukarov, Russ. J.
Electrochem. 30 (1994) 674.
[27] A. Kelaidopoulou, G. Kokkinidis, A. Milchev, J. Elec-
troanal. Chem. 444 (1998) 195.
[28] D.R. Salinas, E.O. Cobo, S.G. Garc ´ı a, J.B. Bessone, J.
Electroanal. Chem. 470 (1999) 120.
[29] H. Bort, K. J u¨ ttner, W.J. Lorenz, G. Staikov, E. Bude-
vski, Electrochim. Acta 28 (1983) 985.
[30] M.H. H o¨ lzle, V. Zwing, D.M. Kolb, Electrochim. Acta 40
(1995) 1237.
[31] E. Michailova, I. Vitanova, D. Stoychev, A. Milchev,
Electrochim. Acta 38 (1993) 2455.
(
1990) 217.
5] P. Allongue, E. Souteyrand, J. Electroanal. Chem. 362
1993) 79.
(
6] P. Allongue, E. Souteyrand, L. Allemand, J. Electroanal.
Chem. 362 (1993) 89.
7] G. Scherb, D.M. Kolb, J. Electroanal. Chem. 396 (1995)
1
51.
8] P.M. Vereecken, K. Strubbe, W.P. Gomes, J. Electroanal.
Chem. 433 (1997) 19.
9] B. Rashkova, B. Guel, R.T. P o¨ tzschke, G. Staikov, W.J.
Lorenz, Electrochim. Acta 43 (1998) 3021.
.