Etched single-ion-track templates for single nanowire synthesis

Article abstract of DOI:10.1021/jp031368w

Polycarbonate membranes with one pore only are created by the track-etching technique. The shape and size of the pore are determined by the type of etchant as well as by the temperature and etching time. The dynamics of single-pore formation during etching is investigated to determine the breakthrough time and the tracketching rate. The pore is characterized by electrical conductivity measurements and scanning electron microscopy. This kind of template is employed for the electrochemical deposition of a single bismuth wire that is left in the polymer for further measurements of its electrical properties.

Full text of DOI:10.1021/jp031368w

9950
J. Phys. Chem. B 2004, 108, 9950-9954
Etched Single-Ion-Track Templates for Single Nanowire Synthesis
N. Chtanko,* M. E. Toimil Molares, T. Cornelius, D. Dobrev, and R. Neumann
Gesellschaft f u¨ r Schwerionenforschung (GSI), Planckstrasse 1, D-64291 Darmstadt, Germany
ReceiVed: December 31, 2003
Polycarbonate membranes with one pore only are created by the track-etching technique. The shape and size
of the pore are determined by the type of etchant as well as by the temperature and etching time. The dynamics
of single-pore formation during etching is investigated to determine the breakthrough time and the track-
etching rate. The pore is characterized by electrical conductivity measurements and scanning electron
microscopy. This kind of template is employed for the electrochemical deposition of a single bismuth wire
that is left in the polymer for further measurements of its electrical properties.
Introduction
In recent years, nanowires have attracted considerable atten-
tion because of their extraordinary physical properties and
enormous potential for applications in numerous fields of
research and technology.1-3 Among the methods widely used
to create nanowires, electrochemical deposition in the pores of
4
,5
ion-track polymer membranes turned out to be very suitable.
These membranes are produced by the irradiation of foils with
energetic heavy ions and subsequent chemical etching of the
so-called latent tracks (i.e., the extremely narrow damage zones
created along the ion trajectories). By varying the ion fluence
and the etching conditions, pores of different sizes and
9
2
geometries and with area densities up to 10 pores/cm are
feasible. Because the material deposited in a pore acquires the
exact pore shape, this template technique enables the fabrication
of wires with various shapes, in particular, conical or cylindrical.
The highly selective etching of tracks in polycarbonate makes
it possible to produce membranes with pore diameters down to
Figure 1. Schematic of the cell used for cylindrical pore etching and
wire electrodeposition. (a) Etching in 2 M NaOH at 40 °C, applying
400 mV. (b) Electrochemical deposition of a Cu layer after the
sputtering of a gold layer. (c) Bi single-wire deposition.
6
about 10 nm.
Until now, most publications presented results concerning the
synthesis and study of metallic nanowires deposited in multitrack
µm (both materials from Bayer AG, Leverkusen). Predetermined
by the technology used for production (cast process), the foils
had a glossy (smooth) and a matte (rough) side.
To produce membranes containing one single pore, a metallic
mask with a circular aperture of diameter 0.1 mm at its center
was placed in front of each sample. The signal of a semiconduc-
tor detector behind the sample, created by the first ion passing
through the foil stack, was used to prohibit the impact of further
ions by blocking the accelerator beam within 15 ms.
7
-10
membranes.
The irradiation of a target with one heavy ion
only, a technique available at the UNILAC linear accelerator
of GSI, provides, after track etching, membranes containing a
single channel. This article reports procedures suitable for the
fabrication of single-channel templates with diameters ranging
from 400 down to 50 nm and with well-defined geometry.
Electrochemical deposition in these membranes allowed us to
generate and contact single nanowires with low-ohmic contact
resistances. We chose Bi because exceptional electronic features
To achieve cylindrical pores, we irradiated each foil prior to
etching with UV light, thus increasing the track etching
11-12
are expected to occur with decreasing wire dimensions.
The
6
first studies of how electrical properties of Bi nanowires differ
selectivity (i.e., the track-to-bulk etch-rate ratio, V /V .) The
T
B
2
1
from those of the macroscopic bulk were presented recently.
etching and electrodeposition processes were both performed
in an electrochemical two-compartment cell (Figure 1). After
inserting the foil, we filled both compartments with aqueous 2
M NaOH solution and kept them at 40 °C; we then applied a
voltage of 400 mV between two Au electrodes immersed in
the etchant on both sides of the foil as shown in Figure 1a. The
process of pore opening and growth was monitored by recording
the electrical current. After etching, the cell was flushed with
distilled water, subsequently filled with 1 M KCl (pH 7), and
the electrical conductance was measured using Ag/AgCl elec-
trodes. Assuming a cylindrical shape, the result was used to
Experimental Methods
Polymer foils were irradiated normal to the surface by Xe,
Au, and U ions of energy 11.4 MeV/u at the UNILAC. The
samples consisted of stacks of either four Makrofol KG foils
of thickness 20 µm or three Makrofol N foils of thickness 30
*
Corresponding author. E-mail: n.chtanko@gsi.de. Phone: (49) 6159-
7
1-21-71. Fax: (49) 6159-71-21-79.
1
0.1021/jp031368w CCC: $27.50 © 2004 American Chemical Society
Published on Web 06/17/2004

Etched Templates for Single Nanowire Synthesis
J. Phys. Chem. B, Vol. 108, No. 28, 2004 9951
Figure 3. Typical etching curve of single ion track. Etching occurred
from both sides of the foil (20-µm Makrofol KG) in 2 M NaOH at 40
°
C and a voltage of 400 mV. The arrow indicates the breakthrough.
Figure 2. Schematic of the cell used for conical pore etching and wire
electrodeposition. First procedure: (a) Electrochemical deposition of
a Cu layer after the sputtering of a gold layer on the rough side of the
foil. (b) Etching in 9 M NaOH with 30% methanol at 30 °C, applying
1
00 mV. Second procedure: (c) Etching in 2 M NaOH with 40%
methanol at 50 °C with stopping solution from the opposite foil side
rough side with the gold layer). (d) Electrochemical deposition of a
(
Cu layer. (e) Bi single-wire deposition in both procedures.
calculate the effective diameter deff of a single pore according
1
8
to
1
/2
4
πκR
l
deff
)
(
)
where l is the thickness and R is the resistance of the cell and
κ denotes the specific conductivity of the 1 M KCl solution,
-
1
amounting to 10.2 S m at 20 °C. The uncertainty of the pore
diameter (∆deff) was calculated, taking into account the errors
of contributing experimental parameters (current, voltage, foil
thickness, and specific conductivity). The sample was taken out
of the cell, rinsed, and dried, and a gold layer was sputtered on
the matte side using an Edwards Sputter Coater S150 B.
Afterward, the foil was returned to its former place in the cell.
A Cu layer was then deposited electrochemically (Figure 1b)
onto the gold and served as a cathode during the subsequent
electrodeposition of Bi. The deposition into a single pore was
performed at 50 °C using a Bi electrolyte with a Bi anode
immersed into it at voltages from 20 to 50 mV between this
anode and the copper layer (Figure 1c). The resulting wires were
almost cylindrical, confirming a large VT/VB.
Two procedures were employed for the generation of conical
Bi nanowires (Figure 2). In both cases, we decreased VT/VB by
a controlled increase of VB. The first technique allowed us to
obtain single conical wires with a large opening angle. After
sputtering a 100-nm gold layer on the rough foil side, the sample
was inserted into the electrochemical cell, and the gold layer
was reinforced with several micrometers of Cu (Figure 2a).
Then, the single track was etched from the uncoated side with
Figure 4. Current (a) and charge (b) as a function of time for Bi wire
deposition in a cylindrical single pore (diameter 130 nm) in polycar-
bonate Makrofol KG, thickness 20 µm. The deposition occurred at 30
mV at 50 °C.
a certain current through the etched pore. After etching, a Cu
layer was deposited on the gold (Figure 2d). In both procedures,
bismuth nanowires were electrodeposited potentiostatically
(Figure 2e) in the single pores at 50 °C using the following
electrolytes: the Bi solution contained 0.2 M BiCl3, 0.3 M
tartaric acid, 0.2 M NaCl, 1.3 M HCl, and 100 g/L glycerol;14
the Cu solution consisted of 238 g/L CuSO4‚5 H2O and 21 g/L
H2SO4; the stopping solution was prepared by mixing 2 M KCl
and 2 M HCOOH (1:1 v/v). The water contact angles were
measured for original polycarbonate foils using an optical
microscope (model Biolam, LOMO, Russia). The current
flowing between the two electrodes was monitored by a Keithley
9
1
M NaOH in H2O + CH3OH (7:3 or 8:2 v/v) at 30 °C, applying
00 mV with the positive potential at the Au electrode in the
etchant (Figure 2b). To obtain a single pore with small opening
angle, we used a second procedure as follows. After sputtering
gold on one side of the PC foil, we filled one-half of the cell
with 2 M NaOH in H2O + CH3OH (6:4 or 7:3 v/v) and the
other half with KCl + H2O + HCOOH (Figure 2c) that was in
contact with the gold-coated side and served as stopping
solution.13 In both cases, etching was stopped when registering
6
485 picoammeter.
Results and Discussion
1. Etching of Single Cylindrical Pores and Bi Wire
Deposition in Makrofol KG. The main disadvantage of
polycarbonate templates is the hydrophobicity of these materials

9952 J. Phys. Chem. B, Vol. 108, No. 28, 2004
Chtanko et al.
Figure 6. Etching of a single PC track in 9 M NaOH + 30% CH
at 30 °C (a) and 2 M NaOH + 40% CH OH (b) with stopping solution
at 50 °C, in both cases applying 100 mV.
3
OH
3
foil originating from the production procedure.15 As a result,
each single track showed individual behavior with regard to
the time needed for perforation, in contrast to many-track
6
9
-2
Figure 5. Morphologies of caps on top of Bi single wires (d
nm) deposited at 50 °C and different voltages: (a) 50 mV, (b) 20 mV.
w
) 200
samples (track densities of 10 -10 cm ), which exhibited well
reproducible current versus etching-time curves due to averaging
over the large track number. Additional UV treatment before
etching increased the etching selectivity and had a significant
and, as a result, the difficulty of filling the pores for conductivity
1
6,17
measurements and the electrodeposition procedure.
Our
6
influence on the final pore shape and size. Some authors
measurement of the water contact angles for both types of
pristine Makrofol resulted in 74 and 84° for KG and N foils,
respectively. The Young-Laplace capillary pressure P depends
significantly on the wetting contact angle according to
mention that “hard”, thin surface layers on both sides of a
16,17
polycarbonate foil influence the etching process.
A problem
that may arise is the occurrence of a “cigarlike” or tapered shape
of the resulting pore in the polycarbonate. Such shapes would
affect the electrochemical synthesis and properties of the wires.
For the foils used in our work, we found this effect to be
insignificant for pores with diameters around 100 nm. Our
investigations, together with the use of selective etchant
(aqueous 2 M NaOH), enabled us to produce cylindrical pores
with high aspect ratios.
θ
P ) 2γ cos
r
where γ is the surface tension, θ is the wetting contact angle,
and r is the radius of the capillary. An increase of θ from 74 to
8
4° changes the capillary pressure by a factor of 5 at the same
capillary radius. This means that filling the pores with aqueous
solutions would occur much more easily in the case of Makrofol
KG. However, one should take into account that the contact
angles have been determined macroscopically, whereas on a
microscopic scale, θ may fluctuate and thus exceed the critical
limit of 90°. Such fluctuations can be due to heterogeneities
The curve in Figure 3 displays the current recorded during
two-sided etching of a single cylindrical pore. The breakthrough
time indicated by a sharp current increase was 3 min, and the
track-etching rate amounted to 3.3 µm/min. For individual single
pores etched from both sides, the opening time varied from 3
to 9 min, which corresponded to a track-etching rate VT ranging
from 1.1 to 3.3 µm/min. The bulk etch rate VB was about 1
(
in topography and chemical composition) always present at a
1
9
3
solid surface. Heterogeneities along the narrow pore can
prohibit the imbibition of a liquid into the channel. For these
reasons, we selected the more hydrophilic Macrofol KG for the
fabrication of single thin cylindrical wires. Because of the
wetting problems in pores, the wire diameters were limited to
values larger than 80 nm.
nm/min, resulting in VT/VB > 10 . A pore opening angle e0.05°
was calculated from arcsin(VB/VT).
The electrodeposition process consisted of several steps. Parts
a and b of Figure 4 show the current and charge versus time
curves for the growth of single Bi wires of 130 ( 20 nm
diameter in 20-µm-thick polycarbonate. At the beginning of the
deposition (phase I), the current increased rapidly because of
the charging of the electrical double layer. The immediately
following decrease in the current indicated the formation of the
diffusion layer. During the wire growth inside the pore (phase
The etching behavior of a single ion track is influenced by
the fact that the energy loss of the ion along its path is not
smooth but fluctuates stochastically. Furthermore, it also
depends on the inhomogeneous structure of the polycarbonate

Etched Templates for Single Nanowire Synthesis
J. Phys. Chem. B, Vol. 108, No. 28, 2004 9953
the area under the experimental curve, shown in Figure 4b until
the moment of the fast current increase, provided the total charge
Qexp ) 3.6 nC. From Qexp, we calculated a wire diameter of
1
1
28 ( 16 nm, in good agreement with the pore diameter of
30 ( 20 nm estimated from the conductance of the pore in 1
M KCl. The 16-nm error was mainly due to the uncertainty of
the point in time at which the metal deposition into the pore
was completed. After the growth of the single wire, the
deposition was continued until a micrometer-sized cap was
grown on top. As was shown earlier for Cu nanowires, the cap
20
morphology indicates the wire crystallinity. We controlled the
crystallinity by varying the deposition voltage. Parts a and b of
Figure 5 depict the caps of two Bi wires (200 nm in diameter)
generated by applying 50 and 20 mV, respectively. A voltage
decrease causes a smaller current density and leads to larger
grains.
2. Etching of Single Conical Pores and Bi Wire Deposition
in Makrofol N. For the fabrication of conical wires, we
employed highly hydrophobic Makrofol N. In principle, the
conical pore shape may help to eliminate the wetting problems
in the process of electrodeposition. For a wetting angle close
to 90°, the imbibition of a liquid into the pore is facilitated when
the meniscus moves from a wider to a narrower part of the
channel. To produce narrow conical channels, a NaOH solution
with a certain amount of methanol was employed for the etching
of single tracks from one side in Makrofol N. Figure 6a and b
shows two etching curves of conical wires created using the
procedures depicted in Figure 2. In the case of Figure 6a, a
conical pore was etched at 30 °C from one side of the foil,
employing 9 M NaOH with 30% methanol. The etching curve
in Figure 6b was recorded during etching from one side of the
foil at 50 °C in 2 M NaOH and by applying a stopping solution
from the other side. The irregular current behavior after the
breakthrough displayed in Figure 6b may be caused by the
meeting of the etchant and the stopping medium, the fluctuation
probably depending on the transient prevalence of electromi-
gration or diffusion fluxes on consumption of highly mobile
w w
Figure 7. Geometrical parameters, D (a) and d (b), of the Cu conical
wires in Makrofol N as a function of etching time in 9 M NaOH at
different concentrations of methanol at 30 °C, determined by SEM.
II), the current increased slightly. By registering the current
during deposition (Figure 4a), we determined the end of wire
formation inside the pore and the starting of cap growth (phase
III) characterized by a very fast current increase.
Faraday’s law determines the charge Q required to deposit a
certain mass m of Bi
F
M
2
w
Q ) mz
with
m ) Fπr l
w
-
OH ions in the neutralization reaction and on convection. After
the breakthrough, the stopping solution neutralized the etchant,
thus preventing the narrow part of the pore from fast growth.
where M is the molar mass of the deposited Bi, z is the valence
of the Bi ions, F is Faraday’s constant, F is the metal density,
rw is the wire radius, and lw is the wire length. Integration of
A series of etching experiments were performed with mul-
Figure 8. SEM image of a conical Bi wire with D
w
) 610 nm and d ) 50 nm.
w

9954 J. Phys. Chem. B, Vol. 108, No. 28, 2004
Chtanko et al.
Figure 9. SEM image of a slightly conical Bi wire with D
w
) 289 nm and d ) 90 nm.
w
tipore membranes under the same conditions as with the single-
pore membranes to study the influence of methanol concentra-
tion on pore geometry. After the electrochemical Cu deposition
and dissolution of the membrane in dichloromethane, the conical
wires were imaged with SEM, and the large (Dw) and small
ments with multitrack membranes. Bismuth single wires,
embedded in polymer, can be employed to find a correlation
between diameter and electrical resistance. The creation of single
wires with diameters <100 nm should be achievable with more
hydrophilic polymers.
(
dw) diameters were determined. Figure 7 illustrates that the
Acknowledgment. We are grateful to Dr. P. Yu. Apel for
fruitful discussions and for critically reading this manuscript.
cone angle became larger when increasing the methanol
concentration from 20 to 30%. Furthermore, parts a and b of
Figure 7 show that for the same etching time a higher methanol
concentration resulted in a larger Dw, whereas there was almost
no influence on dw. The latter result was probably due to the
electrostopping effect of a voltage applied between the two
References and Notes
(
(
(
1) Buchachenko, A. Usp. Khim. 2003, 72, 419.
2) Xia, Y.; Yang, P. AdV. Mater. 2003, 15, 351.
3) Kiang, C.-H.; Choi, J.-S.; Tran, T. T.; Bacher, A. D. J. Phys. Chem.
1
3
electrodes with the positive potential on the etchant side. We
used the graphs in Figure 7 as calibration curves and indirectly
deduced the geometrical parameters of single wires. Under these
etching conditions, we obtained single conical wires, the
narrowest having dw ) 50 nm and Dw ) 610 nm, as illustrated
in Figure 8. Etching at 50 °C provided pores with a smaller
cone angle (probably due to the partial evaporation of methanol)
than etching at room temperature and, accordingly, slightly
conical wires as shown in Figure 9, with dw ) 90 nm and Dw
B 1999, 103, 7449.
(4) Hulteen, J. C.; Martin, C. R. J. Mater. Chem. 1997, 7, 1075.
(
(
(
5) Martin, C. R. Science 1994, 266, 1961.
6) DeSorbo, W. Nucl. Tracks 1979, 3, 13.
7) Toimil Molares, M. E.; Buschmann, V.; Dobrev, D.; Neumann,
R.; Scholz, R.; Schuchert, I. U.; Vetter, J. AdV. Mater. 2001, 13, 62.
8) Dobrev, D.; Vetter, J.; Angert, N.; Neumann, R. Appl. Phys. A 1999,
9, 233.
9) Reutov, V. F.; Dmitriev, S. N.; Sokhatsky, A. S. Nucl. Instrum.
Methods Phys. Res., Sect. B 2003, 201, 460.
(
6
(
(
10) Schwanbeck, H.; Schmidt, U. Electrochim. Acta 2000, 45, 4389.
(11) Heremans, J.; Thrush, C. M.; Lin, Y.-M.; Cronin, S.; Zhang, Y.;
)
289 nm. The wires depicted in Figures 8 and 9 resulted from
Dresselhaus, M. S.; Mansfield, J. F. Phys. ReV. B 2000, 61, 4850.
(12) Liu, K.; Chien, C. L.; Searson, P. C.; Zhang, K. Y. Appl. Phys.
Lett. 1998, 73, 1436.
multipore membranes.
(
13) Apel, P. Yu.; Korchev, Yu. E.; Siwy, Z.; Spohr, R.; Yoshida, M.
Summary and Conclusion
Nucl. Instrum. Methods Phys. Res., Sect. B 2001, 184, 337.
(14) Dobrev, D.; Neumann, R. GSI Sci. Rep. 2002 2003, 147.
In this work, we aimed at the fabrication of bismuth
nanowires suitable for measurements of electrical properties,
eventually visualizing quantum effects. Different procedures for
the preparation of single metallic nanowires were developed.
Two types of polycarbonate foils were tested as matrixes for
the production of single-ion track templates. The dynamics of
single-pore formation in both materials was studied by electrical
conductivity measurements. The pore geometry was controlled
by the etching mode (one-side or two-side), alkali concentration,
and methanol content in the etching solution. Cylindrical and
conical single pores with diameters between 50 and 400 nm
were filled with Bi by electrochemical deposition. The size and
shape of the conical single pores and wires were determined
by means of SEM from etching and electrodeposition experi-
(
15) Oganessian, V. R.; Trofimov, V. V.; Vetter, J.; Danziger, M.;
D o¨ rschel, B.; Hermsdorf, D. Nucl. Instrum. Methods Phys. Res., Sect. B
003, 208, 166.
(16) Ferain, E.; Legras, R. Nucl. Instrum. Methods Phys. Res., Sect. B
003, 208, 115.
2
2
2
(17) Ferain, E.; Legras, R. Nucl. Instrum. Methods Phys. Res., Sect. B
001, 174, 116.
(18) Apel, P. Yu.; Blonskaya, I. V.; Didyk, A. Yu.; Dmitriev, S. N.,
Orelovitch, O. L.; Root, D, Samoilova L. I., Vutsadakis, V. Nucl. Instrum.
Methods Phys. Res., Sect. B 2001, 179, 55.
(19) De Gennes, P. G. ReV. Mod. Phys. 1985, 57, 827.
(20) Toimil Molares, M. E.; Br o¨ tz, J.; Buschmann, V.; Dobrev, D.;
Neumann, R.; Scholz, R.; Schuchert, I. U.; Trautmann, C.; Vetter, J. Nucl.
Instrum. Methods Phys. Res., Sect. B 2001, 185, 192.
(21) Toimil Molares, M. E.; Chtanko, N.; Cornelius, Th.; Dobrev, D.;
H o¨ hberger, E. M.; Enculescu, I.; Blick, R. H.; Neumann, R. Nanotechnology
2004, 15, 201.