Nucleation of tin and tin-silver alloy on copper and nickel in acid plating baths

Article abstract of DOI:10.1149/1.2768905

The rate of nucleation and number density of nuclei of tin and tin-silver deposits on copper and nickel substrates in acid sulfate and acid methanesulfonate baths were measured by application of a nucleation model to chronoamperometric data. Based on the dependence of nucleation rate on overpotential, the critical Gibbs free energy of nucleation and the critical nucleation size were calculated. A mechanism of alloy nucleation in the presence of thiourea is proposed.

Full text of DOI:10.1149/1.2768905

D550
Journal of The Electrochemical Society, 154 ͑10͒ D550-D556 ͑2007͒
0013-4651/2007/154͑10͒/D550/7/$20.00 © The Electrochemical Society
Nucleation of Tin and Tin–Silver Alloy on Copper and Nickel
in Acid Plating Baths
,z
*
*
Z. Zhang and D. P. Barkey
Department of Chemical Engineering, University of New Hampshire, Durham, New Hampshire 03824,
USA
The rate of nucleation and number density of nuclei of tin and tin–silver deposits on copper and nickel substrates in acid sulfate
and acid methanesulfonate baths were measured by application of a nucleation model to chronoamperometric data. Based on the
dependence of nucleation rate on overpotential, the critical Gibbs free energy of nucleation and the critical nucleation size were
calculated. A mechanism of alloy nucleation in the presence of thiourea is proposed.
© 2007 The Electrochemical Society. ͓DOI: 10.1149/1.2768905͔ All rights reserved.
Manuscript submitted December 8, 2006; revised manuscript received June 8, 2007. Available electronically August 20, 2007.
Electrodeposited solders used in semiconductor packaging and
solderable finishes on printed wiring boards are largely based on
tin–lead alloys. Among alloys that can be electrodeposited, the
96.5% Sn–3.5% Ag eutectic is a promising alternative to lead-
bearing solders. Tin–silver alloys are suitable for use as lead-free
solder from the viewpoints of mechanical properties, solderability,
wettability, creep, and thermal cyclic fatigue characteristics.1-5
Because of the large difference in reduction potential between
the metals, it is necessary to reduce the silver activity in the alloy-
plating bath by addition of a complexing agent. Tin–silver alloy has
been electrodeposited from complexing alkaline baths, such as silver
cyanide/tin pyrophosphate6-8 and silver-iodide/tin pyrophosphate
baths.1 A cyanide-free bath for tin–silver alloy plating composed of
acid sulfate/sulfuric acid solution with thiourea has been used to
electrodeposit a wide range of alloy compositions.9
The purpose of this study was to investigate nucleation of tin and
tin–silver alloy electrodeposits on single-crystal copper, electrode-
posited copper, and electrodeposited nickel substrates in acid tin-
sulfate and acid tin-methanesulfonate baths with or without thiourea
or with silver sulfate in combination with thiourea. Electroanalytical
methods were used to measure the nucleation rate and nucleation
density. Nucleation densities were also measured by ex situ atomic
force microscopy ͑AFM͒.
Nucleation of tin electrodeposits with and without additives has
been examined by a few authors by means of chronoamperometry.
Gomez et al. studied tin electrodeposition on vitreous carbon from
sulfuric acid solution without additives.10 Torrents-Burgues et al.
investigated tin electrodeposition on vitreous carbon in a sulfuric
acid solution containing sodium gluconate as a chelating agent.11
Barry et al. studied the synergistic effects of organic additives on the
nucleation and growth mechanisms of tin electrodeposits on a
electrodeposited-copper rotating disk electrode in acid-sulfate
solution.12 Petersson et al. studied the kinetics of tin and tin–lead
alloy deposition on glassy carbon electrodes.13
1.2564
tm
=
͓1͔
͓2͔
N␲kD
1/2
im = 0.6382zFDC͑kN͒
In the case of progressive nucleation, the product of nucleation
rate per nucleation site A and the number of nucleation sites per unit
area N can be determined. The product AN is the nucleation rate per
unit area
1/2
4.6733
tm
=
͓3͔
ͩ ͪ
AN␲kD
im = 0.4615zFD3/4C͑kAN͒
͓4͔
1/4
The model of Scharifker and Mostany allows separate determination
of A and N24
imt1m/2
ln 1 −
+ x − ␣͑1 − e−x/␣͒ = 0
͓5͔
͓6͔
͓7͔
ͫ ͬ
a
ln͓1 + 2x͑1 − e−x/␣͔͒ − x − ␣͑1 − e−x/␣͒ = 0
N␲kD
␣ =
A
zFD1/2C
a =
1/2
␲
͓8͔
͓9͔
x = N␲kDtm
1/2
8␲CM
k =
͓10͔
ͩ ͪ
␳
Thermodynamic information can be obtained from the rate data
under the assumption that nuclei are three-dimensional clusters, an
approximation that neglects interactions with the substrate.47 The
critical Gibbs free energy of critical nucleus formation ⌬Gcrit and
the critical nucleus size ncrit may then be defined by either classical
theory or by atomistic theory. In the classical theory, the critical
Gibbs free energy and critical nucleus are given by
Nucleation Models
Several models have been developed to relate potentiostatic cur-
rent transients to nucleation processes.14-29 Investigators have used
both scanning electron microscopy ͑SEM͒ and AFM to count and
examine nuclei.30-44 In a few cases, nucleation rates extracted from
chronoamperometric data with the help of the models have been
compared with counts obtained by microscopic inspection.45,46
In this study, we used models developed by Scharifker and Hills
and by Scharifker and Mostany.22-24 Potentiostatic current transients
were analyzed by determination of the maximum current im and the
time tm at which the maximum current is attained. For instantaneous
nucleation, the number density of nucleation sites per unit area N
and the bulk diffusivity D can be determined from im and tm
ncritze␩
⌬Gcrit
=
͓11͔
2
2
3
8BVm␴
ncrit
=
͓12͔
3
27͑ze␩͒
where Vm is the atomic volume, ␴ the surface excess free energy, ze
the ionic charge, and B a constant that depends on the shape of the
nucleus. For a sphere, B = 36␲, and for a cube B = 216.
The product AN depends on the critical Gibbs free energy ac-
cording to
*
Electrochemical Society Active Member.
z E-mail: dpb@unh.edu
Downloaded on 2014-10-11 to IP 129.97.58.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 154 ͑10͒ D550-D556 ͑2007͒
⌬Gcrit
D551
AN = K exp −
͓13͔
ͩ ͪ
kT
where K is a constant. Equations 11-13 together provide
2
3
4BVm␴
AN = K exp −
͓14͔
ͩ
ͪ
27͑ze␩͒2kT
A plot of ln͑AN͒ vs 1/␩2 provides K and ␴. ⌬Gcrit and ncrit can then
be calculated with Eq. 11 and 12
According to atomistic theory, the nucleation rate follows
͑␤* + ncrit͒ze␩
AN = K exp
͓15͔
Ј
ͫ
ͬ
kT
The factor ␤* depends on the mechanism of attachment of one atom
to the critical cluster. ␤* is equal to one half in the case of direct
transfer and one in the case of adatom attachment. A plot of ln͑AN͒
vs ␩ provides ncrit
.
Experimental
The electroanalytical measurements were carried out in a
100 mL three-electrode cell open to air. The working electrode was
a single-crystal copper disk ͑Goodfellow͒ with a surface area of
0.785 cm2 or an electrodeposited copper or nickel layer on a plati-
num disk with a surface area of 0.125 cm2. The electrodes were
cleaned sonically for 2 min in ethanol ͑99.5 vol %͒ and then rinsed
with distilled, ultrafiltered water and mounted vertically in the cell.
The counter electrode was a carbon rod. A silver/silver chloride
saturated potassium chloride reference electrode was connected to
the plating solution through a fritted bridge tube filled with 1 M
H2SO4. The power supply was a Pine Instruments model AFCBP1
potentiostat. The measurements were carried out at room tempera-
ture.
Copper single-crystal disk electrodes were first mechanically
polished and then chemical-mechanically polished with a mix-
ture of 0.04 ␮m alumina suspension, concentrated ammonium hy-
droxide, and hydrogen peroxide. Bright electrodeposited copper
substrates were prepared by galvanostatic deposition in a cell con-
taining copper counter and reference electrodes. At a rotation speed
of 800 rpm, a constant current of 2 mA was applied for 29 min to
produce a 10 ␮m thick bright electrodeposited copper layer on the
Pt rotating disk electrode ͑RDE͒. The bright copper solution con-
sisted of 0.25 M CuSO4·5H2O + 1.8 M H2SO4 + 1 mM HCl
+ 300 mg/L poly͑ethylene glycol͒ ͑PEG͒ + 10 ␮M 3-mercapto-1-
propanesulfonic acid ͑MPSA͒. The electrodeposited nickel sub-
strates were prepared in a similar fashion. A constant cathodic cur-
rent of 1.5 mA was applied to the Pt RDE electrode for 25 min to
produce an electrodeposited nickel layer. The temperature was con-
trolled at 50 + 0.5°C during electroplating. The Watts nickel bath
composition was 1.2 M NiSO4 + 0.25 M NiCl + 1.7 M H3BO3.
The tin and tin-alloy plating solutions consisted of 0.02 M
SnSO4 + 1.0 M H2SO4 or 0.02 M tin methanesulfonate ͑SnMS͒
+ 1.0 M methanesulfonic acid ͑MSA͒. In addition to each of these
two base solutions, there were two formulations with additives: base
solution + 20 mM thiourea ͑TU͒ and base solution + 20 mM Tu
+ 2 mM AgNO3. In the voltammetry experiments, the working elec-
trode was first held at a potential of −400 mV, roughly 100 mV
positive of the tin deposition potential. The potential was then swept
in the negative direction, reversed at −560 mV and returned to
400 mV, at rates of 20, 40, 60, 80, and 100 mV/s. In the chrono-
amperometry experiments, the applied potential was stepped from
open circuit ͑−465 mV for tin solution with and without TU,
−430 mV for tin solution with TU and silver͒ to a constant negative
value Vf and held for 200 ms.
Figure 1. Cyclic voltammograms of tin electrodeposition on single-crystal
copper in 0.02 M SnSO4 + 1.0 M H2SO4, 0.02 M SnSO4 + 1.0 M H2SO4
+ 0.02 M TU, and 0.02 M SnSO4 + 1.0 M H2SO4 + 0.02 M TU +2 mM
AgNO3 at a sweep rate of 40 mV/s.
step simulated the chronoamperometric experiments, with a poten-
tial sufficiently negative to produce nucleation. In the second step, a
less-negative potential was applied for a period of tens of seconds to
allow existing nuclei to grow without formation of new nuclei. The
deposition time in the second step was chosen to obtain visible and
stable nuclei and to minimize coalescence of nuclei. The sample was
then removed from the electroplating bath immediately, rinsed in a
stream of pure water, dried in air, and transferred to the AFM.
Results
Cyclic voltamograms obtained with a copper single-crystal elec-
trode in sulfuric acid/stannous sulfate solution are shown in Fig. 1.
In the absence of TU, tin reduction began at −495 mV, and the
current peaked at −525 mV. With TU, tin reduction started at
−509 mV, and the peak occurred at −535 mV. In solution contain-
ing both TU and silver, two distinct cathodic peaks were observed, a
silver-reduction peak at −480 mV and a tin-reduction peak at
−529 mV. The curve obtained in the presence of both TU and silver
is more similar to that obtained with neither silver nor TU than to
that obtained with TU alone. Evidently, TU suppresses tin nucle-
ation on copper but not silver nucleation on copper. After the for-
mation of silver nuclei, tin is deposited on the silver even in the
presence of TU.
The effect of sweep rate on the cyclic voltamograms is shown for
additive-free acid sulfate solution in Fig. 2. Consecutive voltamo-
grams are completely overlapped for at least ten cycles. Similar
reproducibility was obtained in the other two solutions. The scan-
rate dependences of the potential and current density for the ca-
thodic peak in the cyclic voltamograms are shown in Fig. 3. As the
scan rate was increased, the peak potential shifted in the negative
direction, and the current function, defined as the peak current den-
sity divided by the square root of scan rate, decreased.
1/2
Plots of cathodic peak current ip vs ␷ ͑Randles–Sevcik plot͒
The AFM measurements were performed with a Digital Instru-
ments NanoScope E atomic force microscope equipped with a
12 ␮m scanner. Images were captured in height and deflection mode
simultaneously. Samples were prepared by two-step potentiostatic
electrodeposition of tin on a single-crystal copper substrate. The first
are shown in Fig. 4. Potentiostatic current transients obtained on a
stationary electrode were used to construct Cottrell plots ͑Fig. 5͒.
The resulting value of the diffusion coefficient for Sn2+ is in good
agreement with values reported by Tzeng48 ͑see Table I͒. Linear fits
to the data did not pass through the origin on the Randles–Sevcik
Downloaded on 2014-10-11 to IP 129.97.58.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D552
Journal of The Electrochemical Society, 154 ͑10͒ D550-D556 ͑2007͒
Figure 2. Overlayed fifth and tenth cyclic voltammograms of tin elec-
trodeposition on single-crystal copper in 0.02 M SnSO4 + 1.0 M H2SO4 so-
lution at scan rates of 40 and 80 mV/s.
Figure 4. Randle–Sevcik plots for tin electrodeposition on single-crystal
copper in 0.02 M SnSO4 + 1.0 M H2SO4, 0.02 M SnSO4 + 1.0 M H2SO4
+ 0.02 M TU, and 0.02 M SnSO4 + 1.0 M H2SO4 + 0.02 M TU + 2 mM
AgNO3 at scan rates from 20 to 100 mV/s.
plots but came close to the origin on the Cotrell plots. Hence, the
diffusion coefficients obtained from the Cottrell plots are probably
more reliable.
Figure 7 shows a comparison of experimental and theoretical
current transients obtained from the Scharifker and Mostany model,
using the values of A and N obtained from the current maxima. The
calculated transients are in good agreement with the experimental
rising part. After the peak current, a deviation from the model in the
region controlled by linear diffusion was observed. This positive
deviation may be the result of a parallel faradaic reaction.49 The
Current transients obtained at different potentials during tin elec-
trodeposition on single-crystal copper in 0.02 M SnSO4 + 1.0 M
H2SO4 are shown in Fig. 6. The electrodeposited copper substrate
produced curves of similar shape. In the case of electrodeposited
nickel substrate no maximum currents were observed. Therefore, the
process occurring at electrodeposited nickel electrode from additive-
free tin solution is uniform electrodeposition of tin without nucle-
ation, probably initiated by underpotential deposition.
diffusion coefficient obtained from the model was 5.72 ϫ 10−6
.
Figure 5. Cottrell plots for tin electrodeposition on single-crystal copper in
0.02 M SnSO4 + 1.0 M H2SO4, 0.02 M SnSO4 + 1.0 M H2SO4 + 0.02 M
TU, and 0.02 M SnSO4 + 1.0 M H2SO4 + 0.02 M TU + 2 mM AgNO3.
Figure 3. Cyclic voltammograms of tin electrodeposition on single-crystal
copper in 0.02 M SnSO4 + 1.0 M H2SO4 solution.
Downloaded on 2014-10-11 to IP 129.97.58.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 154 ͑10͒ D550-D556 ͑2007͒
Table I. The apparent diffusion coefficient of Sn2+ in acid-sulfate
D553
solutions
Diffusion coefficient ͑cm2 s−1
͒
Cottrell
Randle–Sevcik
Tin
6.19 ϫ 10−6
6.68 ϫ 10−6
6.42 ϫ 10−6
5.78 ϫ 10−6
7.69 ϫ 10−6
5.68 ϫ 10−6
Tin with TU
Tin with TU Ag
Figure 8 shows the dependence of N on ␩ for tin on single-
crystal and electrodeposited copper. Figures 9 and 10 show the de-
2
pendence of AN on 1/␩ and ␩ for single-crystal and electrodepos-
ited copper substrates, respectively. Values of A were of the order
105 s−1, and the process can be regarded as instantaneous nucle-
ation. Critical Gibbs energies of nucleation ⌬Gcrit based on Fig. 9
and Eq. 11, 12, and 14 are roughly 10−20 J. For tin deposited at an
overpotential of 80 mV, for example, the Gibbs free energy of
nucleation on Cu–S and Cu–P was found to be 0.649 and 0.755
ϫ 10−20 J, respectively. For tin–silver on Cu–S, Cu–P, and Ni, the
values were 1.03, 0.75, and 0.82 ϫ 10−20 J, respectively. The num-
ber of atoms forming the critical nucleus ͑ncrit͒ ranged from 0.3 to
1.2. Numbers of this magnitude are not unusual.47 The critical
nucleus size obtained from Eq. 15 ranged from 0 to 2.
Figure 7. Comparison of theoretical and experimental current transients for
deposition on single-crystal copper.
Figure 11 shows current transients for tin–TU solution on the
nickel substrate. In the presence of TU, a nucleation-limited re-
sponse is obtained. Figure 12 shows the dependence of N on ␩ for
tin–silver nucleation on a single crystal, electrodeposited copper,
and electrodeposited nickel. In comparison with tin solution without
additives, the value of A is significantly smaller in the presence of
TU, but N is much higher.
Figure 16 shows the linear dependence of ln͑AN͒ on ␩. Accord-
ing to the small cluster model of the atomistic approach, assuming
␣ = 0.5, the number of atoms in the critical nucleus is 0–1 over the
entire range of investigated overpotentials. The critical nucleus size
obtained from the classical theory is in good agreement with atom-
istic theory.
The effects of TU–silver complexes on tin nucleation and growth
kinetics can be explained by an induced adsorption and ion-bridging
mechanism.50,51 TU is adsorbed on copper52-57 and silver
electrodes.53,58-61 Therefore, the TU–silver complexes are more
readily adsorbed on copper and silver substrates than uncomplexed
2
The relationship between AN and 1/␩ or ␩ is shown in Fig. 13
and 14 for tin–silver on single-crystal copper, electrodeposited cop-
per, and electrodeposited nickel substrates. Figure 15 shows the lin-
ear relationship between ln͑AN͒ and ␩. The critical nucleus size
based on both models is 0–2 atoms on all three substrates. This
system has a high nucleation rate and number density of active sites,
and nucleation is instantaneous.
Figure 6. Current transients for tin electrodeposition on single-crystal cop-
per substrate at the indicated final potentials in 0.02 M SnSO4 + 1.0 M
H2SO4 solution. The initial potential was -465 mV.
Figure 8. Dependence of N on ␩ for tin electrodeposition on single-crystal
copper and electrodeposited copper, in 0.02 M SnSO4 + 1.0 M H2SO4 solu-
tion.
Downloaded on 2014-10-11 to IP 129.97.58.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D554
Journal of The Electrochemical Society, 154 ͑10͒ D550-D556 ͑2007͒
Figure 11. Current transients for tin electrodeposition on electrodeposited
nickel in 0.02 M SnSO4 + 1.0 M H2SO4 + 0.02 M TU solution at the indi-
cated potentials. The initial potential was -465 mV.
2
Figure 9. Dependence of AN on 1/␩ for tin electrodeposition on single-
crystal copper and electrodeposited copper in 0.02 M SnSO4 + 1.0 M
H2SO4 solution.
absence of TU. Apparently, TU inhibits nucleation. Addition of sil-
ver lifts the inhibition by permitting rapid nucleation of silver, fol-
lowed by deposition of tin on the silver nuclei.
tin ions. It is postulated that the adsorbed TU–silver complex ions
are discharged and deposited, forming nucleation sites for tin depo-
sition. In the ion-bridging mechanism, it is assumed that the TU–
silver complex acts as an electron bridge and accelerates reduction.
Figure 17 shows current transients on single-crystal copper for
the three solutions. Addition of TU changes the current transient in a
manner consistent with a shift from instantaneous to progressive
nucleation with substantial inhibition. Addition of silver to the solu-
tion restores the less-inhibited behavior seen in tin deposition in the
Figure 18 shows a typical AFM image. The uniform nucleus size
is an indication that the nuclei were formed instantaneously. At
−530 mV, the measured number of nuclei per unit area was 5.72
ϫ 107 cm−2. At −550 mV, it was 1.01 ϫ 108 cm−2. At −570 mV,
the number density of nuclei per unit area was 2.55 ϫ 108 cm−2
.
The calculated and measured nucleus densities are shown in Table
II. At low deposition potential, the difference in nucleus density
Figure 10. Dependence of AN on ␩ for tin electrodeposition on single-
crystal copper and electrodeposited copper in 0.02 M SnSO4 + 1.0 M
H2SO4 solution.
Figure 12. Dependence of N on ␩ for tin electrodeposition on single-crystal
copper, electrodeposited copper, and electrodeposited nickel in 0.02 M
SnSO4 + 1.0 M H2SO4 + 0.02 M TU + 2 mM AgNO3 solution.
Downloaded on 2014-10-11 to IP 129.97.58.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 154 ͑10͒ D550-D556 ͑2007͒
D555
Figure 16. Dependence of AN on ␩ for tin electrodeposition on single-
crystal copper in 0.02 M SnMS + 1.0 M MSA solution.
2
Figure 13. Dependence of AN on 1/␩ for tin electrodeposition on single-
crystal copper, electrodeposited copper, and electrodeposited nickel, in
0.02 M SnSO4 + 1.0 M H2SO4 + 0.02 M TU + 2 mM AgNO3 solution.
Figure 17. Current transients for tin electrodeposition on single-crystal cop-
per in 0.02 M SnSO4 + 1.0 M H2SO4, 0.02 M SnSO4 + 1.0 M H2SO4
+ 0.02 M TU, and 0.02 M SnSO4 + 1.0 M H2SO4 + 0.02 M TU + 2 mM
AgNO3 solution.
Figure 14. Dependence of AN on ␩ for tin electrodeposition on single-
crystal copper, electrodeposited copper, and electrodeposited nickel in
0.02 M SnSO4 + 1.0 M H2SO4 + 0.02 M TU + 2 mM AgNO3 solution.
between the calculated and measured value is small. But at more
negative deposition potential, the measured value is significantly
higher. At −570 mV, the measured nucleus density is 2.5 times as
great as the calculated value. A possible explanation for the differ-
ence could be the inability of the mathematical model to distinguish
nucleation phenomena that could occur within the diffusion zone,
e.g., more than one nucleus within a single diffusion zone.42
Table II. Values of N based on chronoamperometry and AFM.
N ͑Nuclei/cm−2
͒
Potential
Solution
Sulfate
Additions
None
͑mV͒
Model
—
Measured
−530
−550
−570
−550
−550
−550
−580
0.57 ϫ 108
0.62 ϫ 108 1.01 ϫ 108
1.02 ϫ 108 2.55 ϫ 108
TU
TU + Silver
—
1.87 ϫ 108
0.91 ϫ 108 6.16 ϫ 108
0.87 ϫ 108
2
Methanesulfonate None
—
Figure 15. Dependence of AN on 1/␩ for tin electrodeposition on single-
crystal copper in 0.02 M SnMS + 1.0 M MSA solution.
0.78 ϫ 108
—
Downloaded on 2014-10-11 to IP 129.97.58.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D556
Journal of The Electrochemical Society, 154 ͑10͒ D550-D556 ͑2007͒
Figure 18. ͑Color online͒ AFM images of
tin deposits obtained at -530 mV for 4 s in
0.02 M SnSO4/1.0 M H2SO4 solution.
25. M. V. Mirkin and A. P. Nilov, J. Electroanal. Chem. Interfacial Electrochem., 283,
35 ͑1990͒.
Conclusion
In acid tin solution without TU or silver ion, nucleation is rate
limiting on copper but not on nickel. Addition of TU inhibits nucle-
ation on copper and introduces a rate-limiting nucleation step on
nickel. Addition of silver ion to the tin bath containing TU lifts the
inhibition, resulting in fast nucleation on both copper and nickel.
Nucleation in the alloy-deposition bath proceeds by nucleation of
silver across a film of adsorbed TU followed by deposition of tin
and silver onto the silver nuclei.
26. P. A. Bobbert, M. M. Wind, and J. Vlieger, Physica A, 146, 69 ͑1987͒.
27. M. Y. Abyaneh, Electrochim. Acta, 36, 727 ͑1991͒.
28. A. Milchev, J. Electroanal. Chem., 457, 35 ͑1998͒; , 457, 47 ͑1998͒.
29. A. Milchev, Electrochim. Acta, 30, 125 ͑1985͒.
30. K. Kowal, L. Xie, R. Hug, and G. Farrington, J. Electrochem. Soc., 141, 116
͑1994͒.
31. R. Rynders, and R. Alkire, J. Electrochem. Soc., 141, 1166 ͑1994͒.
32. E. Gomez, E. Valles, P. Gorostiza, J. Servat, and F. Sanz, J. Electrochem. Soc.,
142, 4091 ͑1995͒.
33. J. V. Zoval, R. M. Stiger, P. R. Biernacki, and R. M. Penner, J. Phys. Chem., 100,
837 ͑1996͒.
University of New Hampshire assisted in meeting the publication costs of
this article.
34. B. Rashkova, B. Guel, R. T. Potzschke, G. Staikov, and W. J. Lorenz, Electrochim.
Acta, 43, 3021 ͑1998͒.
35. F. Li, R. Newman, and G. Thompson, Electrochim. Acta, 42, 2455 ͑1997͒.
36. Y. G. Li and A. Lasia, J. Appl. Electrochem., 27, 643 ͑1997͒.
37. K. Ondo, T. Murakami, F. Cerwinski, and K. Shinohara, ISIJ Int., 37, 140 ͑1997͒.
38. E. Barrera, M. Pardave, N. Batina, and I. Gonzalez, J. Electrochem. Soc., 147,
1787 ͑2000͒.
References
1. S. Arai and T. Watanabe, Mater. Trans., JIM, 39, 439 ͑1998͒.
2. E. Bradley, K. Snowdown, and R. Gedney, Circuits Assem., 10, 91 ͑1999͒.
3. M. Jordan, Galvanotechnik, 92, 1225 ͑2001͒.
39. D. Grujicic and B. Pesic, Electrochim. Acta, 47, 2901 ͑2002͒.
40. S. Floate, M. Hyde, and R. Compton, J. Electroanal. Chem., 523, 49 ͑2002͒.
41. M. Hyde, R. Jacobs, and R. Compton, J. Phys. Chem. B, 106, 11075 ͑2002͒.
42. A. Radisic, A. West, and P. Searson, J. Electrochem. Soc., 148, C41 ͑2001͒.
43. A. Radisic, A. West, and P. Searson, J. Electrochem. Soc., 149, C94 ͑2002͒.
44. K. Marquez, G. Staikov, and J. W. Schultze, Electrochim. Acta, 48, 875 ͑2003͒.
45. S. Toschev, A. Milchev, and E. Vassileva, Electrochim. Acta, 21, 1055 ͑1976͒.
46. A. Milchev, B. R. Scharifker, and G. J. Hills, J. Electroanal. Chem. Interfacial
Electrochem., 132, 277 ͑1982͒.
47. E. Budevski, G. Staikov, and W. J. Lorenz, Electrochemical Phase Formation and
Growth: An Introduction to the Initial Stages of Metal Deposition, VCH, Weinheim
͑1996͒.
48. G. S. Tzeng, Plat. Surf. Finish., 66, 67 ͑1995͒.
49. M. Palomar-Pardav, B. R. Scharifker, E. M. Arce, and M. Romero-Romo, Electro-
4. I. Artaki and A. M. Jackson, J. Electron. Mater., 23, 757 ͑1994͒.
5. W. Yang, L. E. Felton, and R. W. Messler, J. Electron. Mater., 24, 1465 ͑1995͒.
6. N. Kubota and E. Sato, Electrochim. Acta, 29, 361 ͑1984͒.
7. N. Kubota and E. Sato, Electrochim. Acta, 30, 305 ͑1985͒.
8. J. C. Puippe and W. Fluehmann, Plat. Surf. Finish., 70, 46 ͑1983͒.
9. M. Bestetti, A. Vicenzo, and P. L. Cavallotti, AESF SUR/FIN 2002, AESF Annual
Technical Conference, p. 745 ͑2002͒.
10. E. Gomez, E. Guaus, F. Sanz, and E. Valles, J. Electroanal. Chem., 465, 63 ͑1999͒.
11. J. Torrent-Burgues, E. Guaus, and F. Sanz, J. Appl. Electrochem., 32, 225 ͑2002͒;
D. Pletcher, A First Course in Electrode Processes, Alresford Press Ltd., Hants
͑1991͒.
12. F. J. Barry and V. J. Cunnane, J. Electroanal. Chem., 537, 151 ͑2002͒.
13. I. Petersson and E. Ahlberg, J. Electroanal. Chem., 485, 166 ͑2000͒; I. Petersson
and E. Ahlberg, J. Electroanal. Chem., 485, 179 ͑2000͒.
14. R. D. Armstrong, M. Fleischmann, and H. R. Thirsk, Electroanal. Chem., 11, 208
͑1966͒.
15. D. J. Astley, J. A. Harrison, and H. R. Thirsk, Trans. Faraday Soc., 64, 192
͑1968͒.
chim. Acta, 50, 4736 ͑2005͒.
50. T. C. Franklin, Surf. Coat. Technol., 30, 415 ͑1987͒.
51. T. C. Franklin, Plat. Surf. Finish., 81, 62 ͑1994͒.
52. B. H. Loo, Chem. Phys. Lett., 89, 346 ͑1982͒.
16. G. J. Hills, D. J. Schiffrin, and J. Thompson, Electrochim. Acta, 19, 657 ͑1974͒.
17. G. J. Hills, D. J. Schiffrin, and J. Thompson, Electrochim. Acta, 19, 671 ͑1974͒.
18. M. Y. Abyaneh, Electrochim. Acta, 27, 1329 ͑1982͒.
19. M. Y. Abyaneh and M. Fleischmann, Electrochim. Acta, 27, 1513 ͑1982͒.
20. E. Bosco and S. K. Rangarajan, J. Electroanal. Chem. Interfacial Electrochem.,
134, 213 ͑1982͒.
53. M. Fleischmann, I. R. Hill, and G. Sundholm, J. Electroanal. Chem. Interfacial
Electrochem., 157, 359 ͑1983͒.
54. Z. Q. Tian, Y. Z. Lian, and M. Fleischmann, Electrochim. Acta, 35, 879 ͑1990͒.
55. A. Lukomska, S. Smolinski, and J. Sobkowski, Electrochim. Acta, 46, 3111
͑2001͒.
56. D. Papapanayiotou, R. N. Nuzzo, and R. C. Alkire, J. Electrochem. Soc., 145,
3366 ͑1998͒.
21. G. Gunawardena, G. J. Hills, I. Montenegro, and B. Scharifker, J. Electroanal.
Chem. Interfacial Electrochem., 138, 225 ͑1982͒.
57. M. Alodan and W. Smyrl, Electrochim. Acta, 44, 299 ͑1998͒.
58. B. Reents, W. Plieth, V. A. Macagno, and G. I. Lacconi, J. Electroanal. Chem.,
453, 121 ͑1998͒.
59. M. Milkowska, Electrochim. Acta, 36, 965 ͑1991͒.
60. M. Milkowska, Electrochim. Acta, 32, 159 ͑1987͒.
61. M. Milkowska, Electrochim. Acta, 33, 161 ͑1988͒.
22. B. Scharifker and G. Hills, Electrochim. Acta, 28, 879 ͑1983͒.
23. B. R. Scharifker and J. Mostany, J. Electroanal. Chem. Interfacial Electrochem.,
177, 13 ͑1984͒.
24. M. Sluyters-Rehbach, J. H. O. J. Wijenberg, E. Bosco, and J. H. Sluyters, J. Elec-
troanal. Chem. Interfacial Electrochem., 236, 1 ͑1987͒.
Downloaded on 2014-10-11 to IP 129.97.58.73 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).