Tuning of the spacing and thickness of metal latticeworks by modulation of self-organized potential oscillations in tin (Sn) electrodeposition

Article abstract of DOI:10.1016/j.electacta.2005.01.055

The spacing and thickness of tin (Sn) latticeworks, formed in oscillatory electrodeposition, were tuned through modulation of the oscillation by changing the applied current density (j_ap) and the concentration of Sn(II) ions (C_Sn) in the electrolyte. When the j_ap was made higher or the C_Sn was made lower, the time of transition from the negative-side to positive-side potential (t₁) of the oscillation became longer and the lattice spacing (d) of the latticework was enlarged. The modulation by the j_ap change was especially interesting because it was simple and produced, e.g., a latticework with different spacings. The tuning of the latticework by the oscillation modulation was understood within the framework of our previously proposed model for the formation of latticeworks by oscillatory electrodeposition.

Full text of DOI:10.1016/j.electacta.2005.01.055

Electrochimica Acta 50 (2005) 5050–5055
Tuning of the spacing and thickness of metal latticeworks
by modulation of self-organized potential oscillations
in tin (Sn) electrodeposition
Toshio Tada, Kazuhiro Fukami, Shuji Nakanishi, Haruka Yamasaki, Satoshi Fukushima,
Tomoyuki Nagai, Sho-ichiro Sakai, Yoshihiro Nakato∗
Division of Chemistry, Graduate School of Engineering Science, Osaka University, Toyonaka, Osaka 560-8531, Japan
Received 11 November 2004; received in revised form 20 January 2005; accepted 20 January 2005
Available online 11 July 2005
Abstract
The spacing and thickness of tin (Sn) latticeworks, formed in oscillatory electrodeposition, were tuned through modulation of the oscillation
by changing the applied current density (jap) and the concentration of Sn(II) ions (CSn) in the electrolyte. When the jap was made higher or
the CSn was made lower, the time of transition from the negative-side to positive-side potential (t1) of the oscillation became longer and the
lattice spacing (d) of the latticework was enlarged. The modulation by the jap change was especially interesting because it was simple and
produced, e.g., a latticework with different spacings. The tuning of the latticework by the oscillation modulation was understood within the
framework of our previously proposed model for the formation of latticeworks by oscillatory electrodeposition.
© 2005 Elsevier Ltd. All rights reserved.
Keywords: Oscillation; Electrodeposition; Self-organization
1. Introduction
of deposits are very attractive from the point of view of
two- or three-dimentional structurization because the elec-
trodeposits grow vertically from substrates in this case. Very
recently, we found [11] that metal latticeworks lying ver-
tical to the electrode surface were spontaneously produced
in synchronization with a potential oscillation in tin (Sn)
electrodeposition. It was shown that periodic occurrence of
three successive processes (autocatalytic growth of needle-
like tin under the diffusion-limited condition, steady growth
of cuboid-shaped tin under the reaction-limited condition and
autocatalytic surface oxidation of the needles and cuboids)
led to the oscillation and the formation of latticeworks. For
normal dendrite formation, only the autocatalytic growth of
needle-like metals continues to occur with occasional branch-
ing. The mechanism of the above oscillatory electrodeposi-
tion is quite unique, operating under particular conditions,
and the formation of metal latticeworks has never been real-
ized by other methods. It should be noted that the proposed
mechanism has a certain generality because we found [12]
Oscillatory electrodeposition is an interesting target from
the point of view of production of micro- and nano-structured
materials because it has a possibility to produce ordered
electrodeposits by recording ever-changing self-organized
spatiotemporal patterns during the oscillation. To date, some
examples have been reported on formation of regulated struc-
turesbyoscillatoryelectrodeposition. Oneexampleislayered
structures, which were formed in a number of systems, such
as Cu/Sn [1,2], Cu/Cu2O [3], Ni/NiP [4,5] and Ag/Sb [6]
deposition. It is expected that this type of deposits is formed
by layer-by-layer electrodeposition synchronizing with spon-
taneous electrochemical oscillations.
Another example is dendritic crystal growth accompa-
nied by potential or current oscillations [7–12]. These type
∗
Corresponding author. Tel.: +81 6 6850 6235; fax: +81 6 6850 6236.
E-mail address: nakato@chem.es.osaka-u.ac.jp (Y. Nakato).
0013-4686/$ – see front matter © 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2005.01.055

T. Tada et al. / Electrochimica Acta 50 (2005) 5050–5055
5051
that zinc (Zn) electrodeposition leading to periodic growth of
stacked hexagonal wafers also proceeded by a similar mech-
anism.
between the working electrode and the reference electrode
was not corrected in the present work.
In the present paper, we report that the spacing and thick-
ness of the latticeworks for Sn are tunable by the modulation
of the oscillation, induced by changing the applied current
density (jap) and the concentration of Sn(II) ions (CSn).
3. Results
Fig. 2 shows a jap versus U curve (for Sn deposition) in
4.0 M NaOH containing 0.4 M Sn(II), obtained in the first
negative scan of jap under a jap-controlled condition. With
the increasing jap in the absolute value, the U shifts gradually
from −1.15 V toward the negative. When the jap reaches and
exceeds the diffusion-limited j (jdl) for the Sn deposition, the
U jumps to a large negative value and a potential oscillation
starts to occur, as reported previously [11].
2. Experimental
Theexperimentalsetupforelectrochemicalmeasurements
and in situ observation of deposits is schematically shown in
Fig. 1. ApolycrystallineSndiscwasusedastheworkingelec-
trode. It was etched with 1 mM HCl just before experiment.
A Pt plate (1 cm2) and an Ag|AgCl|saturated KCl electrode
were used as the counter and reference electrodes, respec-
tively. The applied current density (jap) versus potential (U),
and U versus time (t) curves were measured with a potentio-
galvanostat (Nikko-Keisoku, NPGS-301) and recorded with
a data-storing system (Keyence, NR-2000) at a sampling fre-
quency of 100 Hz. The deposits formed on the surface of the
working electrode was observed with an optical microscope
with a digital CCD camera (OM, VH-5000, Keyence), placed
on a side of the electrode (see Fig. 1). Detailed inspection of
electrodeposits was carried out with a high-resolution scan-
ning electron microscope (SEM, Hitachi S-5000).
Fig. 3showswaveformsofthepotentialoscillationatfixed
j
ap values in 4.0 M NaOH with (a) 0.2 M and (b) 0.4 M Sn(II).
The Sn(II) solution for Sn deposition was prepared by
addition of SnO powder into 4.0 M NaOH. Special grade
chemicals were used together with pure water, which was
obtained by purification of deionised water with a Milli-Q
water purification system. The solution was kept stagnant in
all experiments in order to avoid disruption of fragile metal
latticeworks as the deposits. The ohmic drop in the solution
Fig. 2. Current-controlled jap vs. U curve in 4.0 M NaOH containing
0.4 M Sn(II), obtained in the first negative jap scan. The scan rate is
0.71 mA cm−2s−1
.
Fig. 3. Potential oscillations observed at a constant jap of: (a) −36 mA cm−2
in 0.2 M Sn(II) + 4.0 M NaOH and (b) −71 mA cm−2 in 0.4 M Sn(II) + 4.0 M
NaOH.
Fig. 1. Schematic illustration of an experimental setup used in this work.

5052
T. Tada et al. / Electrochimica Acta 50 (2005) 5050–5055
The fixed jap was chosen to be around two times of the jdl
for better comparison with each other, by taking into account
that the jdl depends on the concentration of Sn(II) ions (CSn).
Hereafter, we focus on two features of the wave form of the
oscillation: (1) the transition time, t1, from the negative-side
to the positive-side potential of the potential oscillation and
(2)theamplitude, A, oftheoscillation. ComparisonofFig. 3(a
and b) shows that the t1 is shorter in the higher CSn solution,
whereas the A is almost independent of CSn. It should be
noted that stable oscillation continued only for about first 120
and 60 s in the solutions of 0.2 and 0.4 M Sn(II), respectively.
The wave form then gradually changed with time and, finally,
the oscillation stopped with U at the positive-side potential.
The wave form and the latticeworks described below were
investigated during the stable oscillation.
Fig. 4 indicates how the latticework as the electrodeposit
is formed in synchronization with the potential oscillation.
A similar result was reported in our previous letter [11]. Pic-
tures in Fig. 4(b) were taken under in situ conditions with an
optical microscope (OM) during the latticework formation.
The numbers added to the pictures mean that they were taken
at stages of the potential oscillation marked by the same num-
bers in Fig. 4(a). Fig. 4(c) is a schematic illustration of the
latticework for easy understanding. These results show that
sharp needles grow during the period of the transition from
the negative-side to the positive-side potential of the oscilla-
tion (stages (1)–(3)), whereas the needles stops to grow when
the potential reaches the positive-side potential. It should be
mentioned that long-range synchronization of the periodic
needle growth processes all over the needles leads to a highly
ordered latticework. Fig. 4(d) shows an SEM image of the lat-
ticework, taken at the positive-side potential, in which cuboid
crystals are formed at the tips of the needles, indicating the
cuboid crystals are formed during the positive-side poten-
tial. The cuboid crystal is made of tetragonal Sn (␤-Sn) and
surrounded with the closest packed, thermodynamically sta-
ble (1 1 0) and (0 1 1) faces. Since the following growth of
needles starts at the tips of the cuboid crystals in three crys-
tallographically equivalent directions as seen in Fig. 4(b and
c), the cuboid crystals are located at lattice-points of the lat-
ticework.
Fig. 5 compares the SEM and OM images of the lattice-
works obtained in the solutions of different CSn, which are
the same solutions as in Fig. 3. The latticeworks obtained in
the low CSn solution (Fig. 5(a and b)) have the larger lattice
spacing (d, see Fig. 4(c)) and smaller sized cuboid crystals
than those obtained in the high CSn solution (Fig. 5(c and d)).
An experimental error for the d was of the order of several
micrometers.
The potential oscillation can also be modulated by con-
trolling the jap value. The t1 at a high jap (Fig. 6(a)) is longer
than that at a low jap (Fig. 6(c)). Fig. 6(b and d) are the OM
images of the latticeworks obtained by oscillations of Fig. 6(a
and c), respectively. The results show that the higher the jap,
the larger the d is. Thus, the d and t1 change simultaneously
through the modulation of the jap.
Fig. 4. (a)
A
potential oscillation at jap = −36 mA cm−2 in 0.2 M
Sn(II) + 4.0 M NaOH. (b) Optical microscopic (OM) images of Sn deposits
taken at various stages of the potential oscillation. The number, (1)–(4),
added on the OM images mean that they were obtained at stages of the
potential oscillation marked by the same numbers in (a). (c) A schematic
illustration of the latticework in (b). (d) A scanning electron microscopic
(SEM) image of cuboid crystals formed at the stage of the positive-side
potential of the oscillation.

T. Tada et al. / Electrochimica Acta 50 (2005) 5050–5055
5053
Fig. 5. (a and c): OM images of latticeworks. (b and d): SEM images
of cuboid crystals formed at the positive-side potential. The electrolyte is
0.2 M Sn(II) + 4.0 M NaOH for (a and b) and 0.4 M Sn(II) + 4.0 M NaOH for
(c and d).
Fig. 7. (a) Potential oscillation in 0.2 M Sn(II) + 4.0 M NaOH and (b) an
OM image of a latticework formed by this oscillation, in which the jap was
changed from −43 to −76 mA cm−2 at the position of the arrow.
Fig. 7 shows results obtained when the jap is jumped
instantaneously from a low to a high value in the course
of the oscillation. At the position of the arrow in Fig. 7(a),
the jap was jumped from −43 to −71 mA cm−2. The wave
form of the oscillation changed simultaneously with the jap
jump. Fig. 7(b) shows the OM image of the resultant deposit,
in which the arrow indicates the position where the jap was
jumped. We can see clearly that the d becomes larger after
the jap is increased in the absolute value, in good harmony
with the increase in the opening by the |jap| increase in
Fig. 6.
4. Discussion
First, let us explain the mechanism of the latticework for-
mation in synchronization with the self-organized potential
oscillation, reported in a previous letter, for later discussion.
We have to note that the experiment is done under a galvanos-
tatic condition with |jap| > |jdl|. In the foregoing stage of stage
(1) of Fig. 4, the cuboids of ␤-Sn, surrounded with the ther-
modynamically stable (1 1 0) and (0 1 1) faces, were prepared
and lastly covered with thin layers of Sn oxide or hydroxide
and passivated, as will be explained later. This implies that
the effective area (Seff) of the active electrode surface for Sn
deposition became small, which led to the negative shift in
U until hydrogen evolution occurred in order to maintain the
constant jap. Thus, the U reaches the negative-side potential
of the oscillation, namely, stage (1) is attained. At such a
large negative potential, the potential-regulated rate of elec-
trodeposition of Sn is very high and the Sn deposition soon
becomes diffusion-limited.
Under the diffusion-limited condition, autocatalytic crys-
tal growth starts (Fig. 8). Namely, under the diffusion-
controlled condition, a spherical diffusion layer is formed at a
peaked part of the deposit surface. This leads to a high current
densityatthepeakedpartcomparedwithotherflatparts, caus-
ing a high growth rate for Sn at the peaked part. Thus, peaked
parts grow rapidly, resulting in formation of sharp needles,
as really observed at stage (2) of Fig. 4. The formation of a
large number of needles at stage (2), on the other hand, leads
to a large increase in the Seff and hence a large decrease in
Fig. 6. (a and c): Potential oscillations in 0.2 M Sn(II) + 4.0 M NaOH at (a)
j
latticeworks formed by the oscillations of (a and c), respectively.
ap = −107 mA cm−2 and (c) jap = −43 mA cm−2. (b and d): OM images of

5054
T. Tada et al. / Electrochimica Acta 50 (2005) 5050–5055
face is gradually oxidized with time. The partial formation
of Sn oxide at this stage produces electrically polar bonds,
Snδ+ Oδ−, here and there at the surface and the resultant
Snδ+ atoms induce positive polarization of adjacent surface
Sn atoms, which in turn causes an easier attack of negatively
charged species OH− on them. This autocatalytic mechanism
works most effectively at thermodynamically stable close-
packed faces [13]. Thus, the (1 1 0) and (0 1 1) faces of the
cuboids are rapidly covered with Sn oxide as soon as they
are formed. The rapid oxidation leads to a rapid decrease in
Seff and a rapid increase in jeff under the constant jap, which
causes a rapid negative shift in U. Thus, the negative potential
state [stage (4) which is equivalent to stage (1)] is reached
again. Repeated occurrence of such processes gives rise to a
potential oscillation together with the latticework formation.
The experimental results shown in the preceding section
can be explained within the framework of the above mecha-
nism. First, let us consider the CSn dependence of the oscilla-
tion (Fig. 3) and the latticework (Fig. 5). The crystal growth
in the higher CSn solution is expected to be faster than that in
the lower CSn solution. This means that the period of time of
the transition from the negative-side to positive-side poten-
tial (t1) becomes shorter with the increasing CSn, explaining
the CSn dependence of the wave form in Fig. 3. On the other
hand, the reason for the CSn dependence of the d in Fig. 5
is not clear yet. A possible explanation is that thick needles
are formed in a high CSn solution (Figs. 5 and 9) and for
such thick needles, the effective surface area (Seff) rapidly
increases compared with thin needles because the Seff is in
proportion to the square of the radius of the needles, in con-
trast to the length of the needle. Namely, the Seff in higher
CSn solution can reach the value giving the reaction-limited j
at a stage of short needles, as schematically shown in Fig. 9.
Second, let us consider the jap dependence of the d of the
latticework. All the experiments in this work were performed
Fig. 8. Schematic drawings of Sn deposits in the course of the transition
from the negative-side to positive-side potential of the oscillation.
the effective current density (jeff) under the constant jap. The
decrease in jeff in turn leads to a positive shift in U, which con-
tinues until the autocatalytic crystal growth stops, namely, the
diffusion-controlled condition disappears. Accordingly, the
U reaches the positive-side potential of the oscillation (stage
(3)). At stage (3), a normal potential-regulated electrodeposi-
tion reaction proceeds under the reaction-limited condition,
leading to production of cuboid crystals surrounded with the
thermodynamically stable (1 1 0) and (0 1 1) faces.
At stage (3), another autocatalytic process, autocatalytic
oxidative passivation, may proceed on the thermodynami-
cally stable faces. Namely, the U at this stage is more positive
than the redox potential for Sn oxidation and the Sn sur-
Fig. 9. Schematic illustrations of growth of needles in solutions of high and
low CSn for explanation of the CSn dependence of d.

T. Tada et al. / Electrochimica Acta 50 (2005) 5050–5055
5055
the jap as shown in Fig. 7, in which the history of a dynamic
potential oscillation was recorded directly in the resulting
latticework.
Acknowledgements
The authors thank the Core Research for Evolutional Sci-
ence and Technology (CREST) program of the Japan Science
and Technology Agency (JST) for financial support. This
work was partly supported also by a Grand in Aid of the Min-
istry of Education, Culture, Sport, Science and Technology
(MEXT) for scientific research and by the Kao Foundation
for Arts and Science. One of the authors, K.F., wishes to thank
for financial support by the 21st Century Center of Excellent
(COE) program, “Creation of Integrated EcoChemistry”, of
Osaka University.
Fig. 10. Schematic drawing of the jap vs. U curve to explain the increase of
the amplitude A of the potential oscillation with the increasing |jap|. ꢀA is a
difference in A between oscillations at low and high jap
.
under the constant jap (|jap| > |jdl|). Therefore, a hydrogen
evolution current should contribute to the total current at
the negative-side potential of the oscillation where the Seff is
small. When the jap was made higher, the U on the negative-
side of the oscillation shifts to the more negative to keep the
constant jap or to get a more contribution of hydrogen evo-
lution current, as schematically shown in Fig. 10. This can
explain the expansion of the amplitude A in the negative-
potential direction, as seen in Fig. 6(a and c). The nega-
tive shift of the negative-side potential requires the growth
of longer needles for the potential to reach the reaction-
limited condition at the positive-side potential, resulting in
the increase in d (Figs. 6 and 7). If the rate of Sn electrode-
position is nearly constant, this implies that the t1 becomes
longer.
In conclusion, the present work has revealed that the spac-
ing and thickness of Sn latticeworks formed in synchroniza-
tion with self-organized potential oscillations are tunable by
the modulation of the oscillation through changes in the jap
and CSn. The tuning by the jap change is very attractive for
constructing modulated latticework structures because the j
can be controlled simply in a reversible manner. Actually, the
latticeworks with two spacings were obtained by controlling
References
[1] S.-I. Sakai, S. Nakanishi, K. Fukami, Y. Nakato, Chem. Lett. 31
(2002) 640.
[2] S. Nakanishi, S.-I. Sakai, T. Nagai, Y. Nakato, J. Phys. Chem. B
152 (2005) 1750–1755.
[3] J.A. Switzer, C.J. Hung, L.Y. Huang, E.R. Switzer, D.R. Kammler,
T.D. Golden, E.W. Bohannan, J. Am. Chem. Soc. 120 (1998) 3530.
[4] P. Peeters, G.v.d. Hoorn, T. Daenen, A. Kurowski, G. Staikov, Elec-
trochim. Acta 47 (2001) 161.
[5] Y.Y. Tevtul’, E.P. Pakhomova, V.I. Markovskii, Russ. J. Electrochem.
30 (1994) 245.
[6] I. Krastev, M.T.M. Koper, Physica A 213 (1995) 199.
[7] D.L. Piron, I. Nagatsugawa, C.L. Fan, J. Electrochem. Soc. 137
(1990) 2491.
[8] J. Stpierre, D.L. Piron, J. Electrochem. Soc. 137 (1990) 2491.
[9] N. Kaneko, H. Nezu, N. Shinohara, J. Electroanal. Chem. 252 (1988)
371.
[10] N. Shinohara, N. Kaneko, H. Nezu, Denki Kagaku 61 (1993) 1211.
[11] S. Nakanishi, K. Fukami, T. Tada, Y. Nakato, J. Am. Chem. Soc.
126 (2004) 9556.
[12] K. Fukami, S. Nakanishi, T. Tada, H. Yamasaki, S.-I. Sakai, S.
Fukushima, J. Electrochem. Soc. 152 (2005) C493–C497.
[13] S. Nakanishi, Y. Mukouyama, K. Karasumi, A. Imanishi, N. Furuya,
Y. Nakato, J. Phys. Chem. B 104 (2000) 4181.