Electrochemical deposition of metallic nanowires as a scanning probe tip

Article abstract of DOI:10.1149/1.3187215

This paper describes Ni electrodeposition as a method of producing metallic nanowires on a scanning probe tip. A polycarbonate porous membrane was used as a template for the growth of nanowires on a Si atomic force microscope (AFM) probe tip. A thin Pt ca

Full text of DOI:10.1149/1.3187215

Journal of The Electrochemical Society, 156 ͑10͒ D431-D438 ͑2009͒
D431
0
013-4651/2009/156͑10͒/D431/8/$25.00 © The Electrochemical Society
Electrochemical Deposition of Metallic Nanowires as a
Scanning Probe Tip
a,
a
a,b, ,z
Munekazu Motoyama, * Neil P. Dasgupta, and Fritz B. Prinz *
a
b
Department of Mechanical Engineering, and Department of Materials Science and Engineering,
Stanford University, Stanford, California 94305, USA
This paper describes Ni electrodeposition as a method of producing metallic nanowires on a scanning probe tip. A polycarbonate
porous membrane was used as a template for the growth of nanowires on a Si atomic force microscope ͑AFM͒ probe tip. A thin
Pt cathode was deposited on the tip apex and onto the underlying porous membrane with the help of a focused ion beam. The
number of nanowires deposited on the AFM probe tip was controlled by the cathode area and the duration of Ni deposition.
Surface imaging was performed with the nanowire AFM probes. Elastic and inelastic deformations of the nanowires were
observed, and the related beam mechanics were calculated.
©
2009 The Electrochemical Society. ͓DOI: 10.1149/1.3187215͔ All rights reserved.
Manuscript submitted April 24, 2009; revised manuscript received June 29, 2009. Published August 20, 2009. This was Paper
2
610 presented at the Honolulu, Hawaii, Meeting of the Society, October 12–17, 2008.
Scanning probe technology has played an essential role in nano-
technology since the scanning tunneling microscope was invented in
the inherent shape anisotropy, and diameters that are comparable to
25
magnetic domains significantly increase coercivity.
1,2
the early 1980s. Subsequently, the introduction of the atomic force
microscope ͑AFM͒ provided a mechanism to detect surface proper-
ties via mechanical interactions between a piezomanipulated tip and
Electron- or ion-beam-induced metal deposition can individually
grow metal nanowires with diameters below 100 nm at desired po-
sitions. Yasutake et al. deposited single tungsten pillars with diam-
eters of 90 nm on flattened AFM probe ends by focused-ion-beam-
3-5
a sample surface at the atomic scale.
2
6
A range of surface observations may be facilitated through the
͑FIB͒-induced deposition. However, the beam-induced deposition
introduces deposition of other elements such as C from the organo-
metallic precursors and Ga from the ion beam. Eventually, the com-
position of these structures is typically 20–40 atom % of the desired
6
use of conductive tips, including scanning capacitance microscopy,
7
8
Kelvin probe force microcopy, AFM impedance spectroscopy, and
scanning spreading resistance microscopy. Moreover, conductive
9
2
7,28
tips can also be used to locally induce nanoscale electrochemical
metal, with the balance being C, Ga, and other contaminants.
1
0-12
surface reactions such as scanning probe lithography.
Arrays of metallic nanowires can also be produced by electro-
chemical deposition into porous thin-film templates such as polycar-
Metal-coated Si probes are commonly used as conductive AFM
probes. However, in most instances, the tip geometry prohibits the
generation of highly localized electric fields. Due to the pyramid-
shaped tip geometry, the electric field between the tip and the
sample extends in the lateral direction from the tip apex. Moreover,
the metal layer on the tip degrades with the number of scans, result-
ing in the removal of the conductive surface from the tip. Carbon-
nanotube ͑CNT͒ AFM tips have a very sharp tip and high aspect
2
9,30
31,32
bonate track-etched membranes
and anodic alumina films.
The diameter of the nanowires can be as small as 10 nm, and aspect
2
ratios of greater than 10 are achievable. A wider range of metals
can be deposited compared to that by beam-induced deposition, and
the atomic percent of metals in such structures is almost 100. How-
ever, this method produces several wires per unit area
8
11
−2
͑10 –10 cm ͒. This method does not have the ability to control
the growth of individual nanowires. The number density of the wires
is essentially determined by the pore density of the template. The
present track-etching technology can produce single-pore-containing
1
3,14
ratio,
to control.
is difficult to maintain. Metal-coated CNT tips provide a possible
but the electrical conductivity of the nanotubes is difficult
15,16
The structural stability under large current conditions
17
1
8
30
solution. However, wear of the metal layer poses significant chal-
lenges. Alternatively, excellent electronic transport properties can be
expected in metallic nanowires with high aspect ratios. Metal nano-
wire probe tips are usable until they are fully worn.
membranes. Still, positional selectivity is not available. The fabri-
cation of metal nanowire AFM probes requires an effective strategy
for growing the nanowires at a desired position and for controlling
their number.
The ability to control the electrical potential of a nanowire AFM
probe is attractive for measurements in liquids. Several methods for
producing metal tip electrodes have been reported for scanning elec-
trochemical microscope ͑SECM͒ applications. However, the ex-
We report here an approach to the production of metallic nano-
wire AFM probes by electrochemical deposition. We used the FIB-
induced deposition technique to prepare the cathode on a limited
area of the template film, exactly where the AFM probe tip is posi-
tioned. We also show that the fabricated metallic nanowire AFM
probes are force sensitive enough to obtain a surface topographic
image at the nanometer scale.
1
9-21
posed electrode diameters are mostly in the micrometer range.
A
few groups reported the fabrication of SECM probes, which can also
be used for AFM, but the tip geometries have pyramidal
1
8,22,23
shapes.
Experimental
Unique mechanical and magnetic properties are expected from
one-dimensional ͑1D͒ metal nanostructures. For instance, CNT tips
are considered not to damage delicate organic and biological
samples because they buckle under a low limit of vertical forces and
A polycarbonate track-etched membrane ͑Toyo Roshi Kaisha
Ltd.͒ was used as a template for the growth of nanowires. The nomi-
nal values of the pore diameter, membrane thickness, and pore den-
2
4
8
−2
elastically bend in a wide range of lateral deflections. Metal nano-
wire probes could provide a conductive probe for electrical mea-
surements in parallel with topography measurements while retaining
significant tip compliance. Furthermore, ferromagnetic nanowires
may improve the sensitivity of magnetic force microscopy because
the magnetization is strongly preferred in the axial direction due to
sity were 200 nm, 10 ␮m, and 3 ϫ 10 cm , respectively. A com-
+
mercial n -Si AFM probe with an angled tip for the contact mode
͑NANOSENSORS AdvancedTEC͒ was used as the substrate for
nanowires. The guaranteed ranges of the resistivity, force constant,
−
1
and resonance frequency were 0.01–0.02 ⍀ cm, 0.02–0.75 N m ,
and 7–25 kHz, respectively. The probe was designed with a tip that
protrudes at an angle from the end of the cantilever. The tip was thus
visible from above.
Figure 1 shows scanning electron microscope ͑SEM͒ images of
an identical Si probe tip at different viewing angles. The overall
*
Electrochemical Society Active Member.
E-mail: fbp@cdr.stanford.edu
z
Downloaded on 2015-03-11 to IP 132.174.255.116 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D432
Journal of The Electrochemical Society, 156 ͑10͒ D431-D438 ͑2009͒
Figure 3. Schematic illustration of the configuration of the electron- and
ion-beam columns and the sample.
the ion beam, 5 mm below the lower edge of the electron beam
column. The distance from the ion-beam column to the sample
surface was 18 mm. The capillary needle used to inject Pt precursor
gas molecules ͓methylcyclopentadienyl platinum trimethyl,
Figure 1. SEM images of a commercial Si AFM probe having a tip visible
from the top. ͑A͒ Plain view ͑tip side͒ and ͑B͒ side view. Scale bar: 10 ␮m.
͑
CH ͒ ͑CH C H ͒Pt͔ was inserted above the sample so that the lin-
3 3 3 5 4
vertical length of the tip was ϳ30 µm, although these lengths varied
among individual probes. The tip was tilted at an angle of approxi-
mately 40° ahead from the vertical direction to have the end at the
front of the cantilever. The top of the Si chip and the cantilever were
coated with a thin Au layer ͑ϳ30 nm thick͒ by sputtering. The
bottom of the Si chip was bonded to the top of the template mem-
brane with epoxy resin ͑Fig. 2͒. Because the end of the probe pro-
truded below the bottom of the Si chip, it was also in contact with
the top of the membrane. Indeed, placing an AFM tip onto the tem-
plate membrane only caused bending of the cantilever but did not
cause damage to the tip. After allowing for sufficient time to solidify
the resin, the AFM probe with a template membrane on the bottom
was transferred into the chamber of an FIB machine ͑FEI Strata DB
ear distance from the injection nozzle to the ion-beam incident point
on the sample was approximately 160 ␮m.
Platinum precursor molecules were directed from the nozzle to-
ward the sample. The precursor molecules that adsorbed on the sur-
face decomposed by incident Ga beams whereby nonvolatile prod-
ucts from the precursors, e.g., Pt, were deposited. Volatile products
were readily evacuated from the chamber.
After the FIB-induced Pt deposition, the sample was transferred
to a Cu sheet with the Au-coated side of the Si chip in electrical
contact with the Cu. The approximate size of the Cu sheet was
1 cm ϫ 1 cm, and a lead wire was soldered on an edge. The probe
was then immobilized on the Cu sheet using epoxy resin. Subse-
quently, the Cu sheet was horizontally set on the bottom of a glass
beaker cell while keeping the AFM probe side facing up.
+
235 dual-beam FIB/SEM͒ to locally deposit a Pt layer on the Si
probe tip apex in the center.
Figure 4 shows an optical microscope image of a Si AFM probe
Figure 3 shows the column arrangement in the FIB chamber
during the Pt deposition. The sample was placed with the membrane
template adjacent to the stage. The stage was tilted 52° around an
axis coinciding with the intersection of the electron beam and
Figure 2. Schematic illustration of a Si AFM probe and the template mem-
brane used to grow nanowires from the tip. A small amount of epoxy resin
was pasted on the bottom of the Si chip to couple with the template.
Figure 4. ͑Color online͒ Optical microscope image of the side view of the Si
AFM probe placed on the polycarbonate membrane. The cantilever length is
approximately 450 ␮m. The current flows through a thin Au layer on top.
Downloaded on 2015-03-11 to IP 132.174.255.116 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, 156 ͑10͒ D431-D438 ͑2009͒
D433
Table I. Summary of FIB-induced Pt deposition. Processed area,
ion-beam current, and thickness are preset parameters.
Processed area Ion-beam current Deposited thickness Halo radius
2
͑
␮m ͒
͑pA͒
͑␮m͒
͑␮m͒
3
0
0
ϫ 3
.8 ϫ 0.8
.4 ϫ 0.4
29
1
1
0.7
0.4
0.1
13–14
3–4
ϳ1
ered that a circular halo creation on the substrate is also associated
with complicated interactions of secondary particles such as emitted
3
4,35
electrons and sputtered atoms with the substrate.
The beam in-
Figure 5. Scanning ion-beam images of Si probe tips in contact with the
polycarbonate template surfaces. The patterned Pt sizes are ͑A͒
tensity exterior to the defined beams exponentially decreases with
increasing radial distance. Thus, the thickness of the Pt layer expo-
nentially decreases with increasing radial distance. The intended
rectangular Pt structure was better defined at the tip apex ͑indicated
by the arrow͒. As a result, the actual Pt deposition area was larger by
a factor of approximately 20 than the intended area due to the halo
deposition.
3
2
2
ϫ 3 ␮m ͑scale bar: 10 ␮m͒ and ͑B͒ 0.4 ϫ 0.4 ␮m ͑scale bar: 5 ␮m͒.
The bright Pt region appears to extend in a circle ͑halo͒ much larger than the
intended square pattern indicated by the arrow in ͑A͒. Although the region is
more blurred compared to the circle in ͑A͒, the bright Pt area ͑indicated as
FIB-induced Pt͒ is much larger than the intended square area of 0.4
3
4
2
ϫ 0.4 ␮m in ͑B͒.
The Pt deposition results are summarized in Table I. 29 pA was
too large of a current to process a smaller geometric area because
the sputter-etching effect of ion beams starts to overwhelm Pt depo-
sition. Hence, a beam current of 1 pA was applied for a smaller size
of Pt deposition.
on the template membrane after FIB-induced Pt deposition. The
sputter-deposited Au layer was used to apply a potential to the Pt
region at the tip apex.
−
1
The electrolyte composition was 262 g L
NiSO ·6H O,
4 2
The probe tip after Pt deposition with the smallest processed area
is shown in Fig. 5B. The actual Pt deposition occurred more exten-
−
1
−1
4
5 g L NiCl ·6H O, and 38 g L H BO ͑pH 3.4͒. The elec-
2 2 3 3
trolyte was carefully poured into the beaker up to a level approxi-
mately 5–10 mm above the AFM tip. A Ni wire ͑1.0 mm diameter͒
with an insulated lateral surface was designed as a reference elec-
trode and was fixed with the tip located 1–2 mm above and 2–4 mm
laterally distant from the AFM tip. A Ni sheet counter electrode was
vertically installed into the cell.
2
sively over an area of ϳ3 ␮m compared to the intended
0
.4 ␮m ϫ 0.4 ␮m square area because of halo deposition, as ob-
served in Fig. 5A. A roughly fourfold increase from the processed
length to the halo radius was always observed. Resistivities in the
films were supposed as more than 10 ␮⍀ cm relative to bulk pure
5
27
Pt with a resistivity of 10 ␮⍀ cm.
We deposited Ni potentiostatically at E = −0.9 V with respect to
a reference electrode. Two deposition durations of 330 and 180 s
were tested. After electrodeposition, the template membrane was
Electrochemical deposition of metallic nanowires.— An exami-
nation of whether an FIB-induced Pt layer on the polycarbonate
template can serve as the cathode for the deposition of nanowires
was conducted as a function of the change in the processed area.
Figure 6A shows an SEM image of the AFM probe tip stripped from
the template after the electrodeposition of Ni nanowires. FIB-
induced Pt was deposited in the vicinity of the Si tip apex, with a
halo radius of 13–14 ␮m ͑top row in Table I͒. The Ni deposition
duration was 330 s. The Si surface was substantially covered by a Ni
film. A circular tablelike structure formed on the Si tip where several
nanowires stood entirely on top ͑Fig. 6B͒. The height of the longest
nanowire located directly above the Si tip was approximately 7.0 µm
dissolved by slowly injecting a dichloromethane ͑CH Cl ͒ solution
2
2
into the cell. The volume of the separate phase of dichloromethane
from the aqueous electrolyte was slowly increased until the mem-
brane was submerged in the dichloromethane phase. Polycarbonate
readily dissolves in a dichloromethane solution.
After the template membrane disappeared from the AFM probe,
the probe was separated from the Cu sheet and kept in sequentially
fresh dichloromethane, chloroform, and ethanol solutions at 70°C
for more than 1 h each.
Results and Discussion
͑
ϭ4.7/cos 48°͒. The location of the bottom end of the nanowire was
FIB-induced Pt deposition.— It is difficult to observe a noncon-
ducting polycarbonate surface with an electron beam because of
surface charging, resulting in thermal distortions. Therefore, we
used ion beams with small currents when viewing polycarbonate
regions.
Figure 5A shows a scanning ion-beam image of the AFM probe
tip placed on the template after deposition of ion-beam-induced Pt.
An ion-beam current of 29 pA was used for a square area of
determined in Fig. 6A. The geometry of the nanowire region was
influenced by the local Pt deposition condition because the growth
of nanowires was suppressed on the left side in Fig. 6A ͑the side
nearer to a reader in B͒, where the Si probe blocked the Pt deposi-
tion. Nanowires standing near the edge were shorter than those in
the center. The areal diameter of the nanowire region was approxi-
mately 36 ␮m, which extended from the halo diameter.
To decrease the geometric area of the nanowire region, we em-
2
3
␮m ϫ 3 ␮m with the deposited Pt centered at the tip apex.
ployed a smaller Pt cathode with a processed area of 0.64 ␮m
A scanning ion-beam image was produced by collecting second-
͑middle row in Table I͒. The obtained probe tip structure is shown in
Fig. 7A. A drastic change in the size of the tip was observed by
reducing the Pt deposition area. The circular geometry of the nano-
wire region, which was significantly large in Fig. 6, changed into a
small circle with a diameter of approximately 12 ␮m in Fig. 7A. It
was slightly larger than the produced halo diameter of ϳ7 ␮m ͑see
Table I͒. Although the remaining residue covers a major portion of
the tip in Fig. 7A, the number of nanowires is smaller than that in
Fig. 6.
ary electrons generated from the surface due to the incidence of the
3
3
ions. The nonconducting region of a sample accumulates a net
+
positive charge due to the impinging Ga ions, thereby inhibiting the
escape of the secondary electrons emitted from the surface. The
insulating polycarbonate region therefore appears darker. The ion-
beam-induced Pt region in the vicinity of the probe tip appears
brighter. It was observed that a circular thin Pt region with a diam-
eter of approximately 13–14 ␮m resulted extending from the in-
tended square area of 3 ␮m ϫ 3 ␮m. This is referred to as a
Finally, the Pt deposition was performed with a processed area of
2
“
halo”, which is caused by a broad tail of an FIB radially spreading
0.16 ␮m ͑bottom row in Table I͒, which was the minimum achiev-
out of the defined beam region with a Gaussian profile. It is consid-
able size. The obtained probe tip structure is shown in Fig. 7B. The
Downloaded on 2015-03-11 to IP 132.174.255.116 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D434
Journal of The Electrochemical Society, 156 ͑10͒ D431-D438 ͑2009͒
Pores
Polycarbonate
Pt
Ni
Si tip
Figure 8. Schematic illustration of the deposition process of the nanowires
on an AFM probe tip.
Figure 6. SEM images of the produced Ni nanowire AFM tip with a halo
deposition of 26 ␮m diameter. ͑A͒ 0°- and ͑B͒ 52°-tilted views. The Ni
deposition duration was 330 s. Scale bar: 10 ␮m.
probe axis, which was caused by the much larger dimensions of the
Pt halo compared to the Si tip apex, was observed in Fig. 6 and 7A
but was no longer obvious in Fig. 7B.
The areal diameter of the nanowire region increased with the Pt
deposition area. It was also observed that the height of the nano-
wires tends to be shorter toward the edge of the region. The vertical
length of the longest nanowire on each probe was comparable to the
radial expansion length of the nanowire region from the Pt halo
diameter.
tip formed one hundred and several tens of nanowires on top of the
flat surface with a diameter of approximately 8 ␮m. Nanowires
located in the center are longer than near the edge. The vertical
height of the longest wire was 4.2 ␮m. The probe tip formed a
frustum shape. The radial expansion of the nanowire region from the
FIB-induced Pt including a halo is supposed to have a nonuni-
form resistivity distribution such that the resistivity increases with
radial distance in a halo. Pores located in the low resistivity Pt
region readily start to grow wires after the deposition begins. How-
ever, a large potential drop is produced in the high resistivity region,
and therefore the applied cathodic overpotential is decreased with
increasing radial distance. Ni grows not only into the pores, but also
along the membrane surface expanding the tip cross section. As the
tip radius expands, the edge of the tip enters into the pores more
distant from the cathode center, increasing the number of wires pro-
truding from the tip ͑Fig. 8͒. It is therefore concluded that the
growth duration of the nanowires becomes shorter with increasing
radial distance from the Pt cathode. The trend that the height of the
nanowires becomes shorter toward the edge of the circular flat sec-
tion is attributed to the potential drop increase with the radial dis-
tance of the cathode layer and the lateral expansion of the tip section
during the deposition. Hence, the longest nanowires were always
observed in the center, which is located directly above the Si tip. It
can also be deduced that Ni growth rates into the pores and in the
lateral direction are close to each other.
Approximately 149 nanowires are found within a flat-sectional
2
area of approximately 54 ␮m in Fig. 7B. The number density of
8
−2
the nanowires can be calculated as 3 ϫ 10 cm , which is exactly
the same as the pore density of the template.
The Ni deposition duration was subsequently reduced from 330
to 180 s while maintaining the minimum Pt cathode area, which was
2
the 0.4 ϫ 0.4 ␮m processed area. The obtained tip structure is
shown in Fig. 9A. The flat section at the tip end became even
smaller than that in Fig. 7B, and the number of nanowires on the tip
was reduced to ϳ66. The nanowire ends show diameters of 320 nm,
which is larger than the nominal diameter of 200 nm. This is be-
cause the pore interiors are slightly swollen inside the membrane
Figure 7. SEM images of the produced Ni nanowire AFM tips with halo
depositions of ͑A͒ 3–4 ␮m and ͑B͒ ϳ1 ␮m diameters. The Ni deposition
duration was 330 s. Scale bar: 5 ␮m.
Downloaded on 2015-03-11 to IP 132.174.255.116 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, 156 ͑10͒ D431-D438 ͑2009͒
D435
Figure 10. ͑A͒ TEM bright field image ͑scale bar: 1 ␮m͒ and ͑B͒ electron
diffraction pattern of Ni nanowires deposited without an AFM tip at
2
9
E = −0.9 V by the method presented in a previous work.
pitch of the steps is thus 3 ␮m, as expected. Figure 11B shows the
surface profile in the transverse direction ͑X͒ in the middle of the
image. The vertical distance between two parallel lines through the
top and the bottom of the structure is 17 nm, which agrees with the
nominal step height.
The accompanying vertical force variation is also shown below.
The scanning direction was parallel to the cantilever axis ͑transverse
direction in Fig. 11A͒. The vertical deflection of the cantilever head
Figure 9. ͑A͒ SEM image of the produced Ni nanowire AFM tip with a Ni
deposition duration of 180 s. The halo area was approximately 1 ␮m radius.
Scale bar: 2 ␮m. ͑B͒ Amplitude vs excited frequency variation.
29,36
compared to the mouths.
The vertical length of the highest nano-
wire was 2.3 µm. The ratio of 330 to 180 equals that of 4.2–2.3. This
result indicates a steady growth of the nanowires along their length.
−
1
The deposition rate was calculated roughly as 13 nm s , which
−
2
corresponds to a reasonable current density of 38 mA cm for this
2
9
potential.
The resonance frequency of the produced nanowire AFM probe
was measured by a forced oscillation. The maximum amplitude ap-
peared at 19.6 kHz ͑Fig. 9B͒, and the quality factor is 156.7. Sader’s
1
method estimates
451 ␮m; width = 50 ␮m͒.
For comparison with normal experiments, Fig. 10A shows a
a
force 37constant of 0.91 N m− ͑length
=
transmission electron microscope ͑TEM͒ image of Ni nanowires de-
2
posited in a wider area ͑0.0314 cm ͒ of the template. Experimental
29
details are described elsewhere. Consequently, the obtained nano-
wires also showed lengths of 4–5 ␮m with a deposition duration of
3
1
30 s. The electron diffraction pattern indicates crystallite Ni ͑Fig.
0B͒.
AFM measurements.— Before testing the nanowire AFM probe,
we obtained reference data using a bare Si probe tip without nano-
wires. The test sample was a 1D grating structure of SiO rectangu-
2
lar steps ͑MikroMasch͒. Figure 11A shows the plain view of the
surface topography of the test sample measured with a Si probe tip
in the contact mode AFM ͑JEOL JSPM-5200͒. The nominal step
height and pitch of steps were 18 Ϯ 1 nm and 3 ␮m, respectively.
The scanning area was 5 ␮m ϫ 5 ␮m. The scanning rate was
−
1
5
␮m/51.2 ms = 9.8 ϫ 10 ␮m s . Two parallel steps are visual-
ized clearly, and one trench is seen in the middle. The respective
Figure 11. ͑A͒ AFM image of parallel rectangular SiO steps obtained with
a Si probe tip. ͑B͒ Z-profile along the broken arrow.
2
widths of each step and trench are approximately 1.5 ␮m, and the
Downloaded on 2015-03-11 to IP 132.174.255.116 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D436
Journal of The Electrochemical Society, 156 ͑10͒ D431-D438 ͑2009͒
Figure 12. Surface topographic profile of the same sample as used in Fig. 11
and the accompanying vertical force variation measured using the nanowire
AFM probe presented in Fig. 9.
Figure 13. Schematic illustration of an inverted-conical-shaped beam ͑bend-
ing nanowire͒. The diameter at the free end ͑d ͒ is larger than that at the
A
fixed end ͑d ͒. The lateral force, F, is exerted on the free end, and the beam
B
is bent with an end deflection of ␦x=L. The x coordinate axis is arranged from
the fixed end along the length of L.
was proportional to a vertical load exerted on the tip following
Hooke’s law. Sharp peaks and valleys appeared when the tip
scanned up and down the steps, respectively. They coincide with
inflection points of the Z-profile variation.
When the tip is released from the step edge ͑a, Fig. 11B͒, the
vertical deflection of the cantilever recovers toward its noncontact
state. At the moment before the tip touches the bottom, the cantile-
ver senses the smallest force ͑b͒. The vertical deflection increases
again as the tip drags on the sample surface. When the tip contacts
the step edge ͑c͒, the cantilever is deflected vertically, exerting a
larger force on the sample. The deflection reaches a maximum when
the tip end was lifted along the step wall up to the edge ͑d͒. The
deflection then begins to decrease slowly.
more than 200 nm ͑see Fig. 10͒, which is larger than a Si tip radius.
The increase in the error distances of Fig. 12B from Fig. 11B is
38
attributed to the effect of the finite length of the tip radius.
Next, we consider the bending stiffness of a nanowire. A nano-
wire can be approximated as having an inverted-conical shape ͑Fig.
1
3͒ because the free ends of nanowires are slightly wider than the
fixed ends, as we have mentioned previously. The lateral force, F,
exerted on the free end of a nanowire was considered as a point
load. Based on the classical mechanics of bending beams, the rela-
Figure 12A shows the plain view of the topography of the same
tionship between F and the free-end deflection of a nanowire, ␦x=L
is given as follows ͑Appendix A͒
,
SiO rectangular step arrays, measured with the nanowire AFM tip
2
presented in Fig. 9. The measurement conditions, such as scanning
rate and scanning area, were the same as in Fig. 11. Two parallel
steps and two parallel trenches are visualized with widths of
3
3␲EdAdB
F =
␦x=L
͓1͔
4L3
6
1.5 ␮m. The pitch of steps thus successfully indicates 3 ␮m. Fig-
where dA and dB are the diameters at the free and fixed ends, re-
spectively. Thus, the bending stiffness of a nanowire can be defined
as
ure 12B shows the transverse surface profile and the accompanying
vertical force variation in the middle of the image. The vertical step
height is 21 nm, which is actually slightly higher than the nominal
value. The accompanying vertical force variation is similar to Fig.
3
␲EdAd3
B
1
1B. However, error distances of ͑a͒ to ͑b͒ and ͑c͒ to ͑d͒ appeared
ktip =
͓2͔
6
4L3
longer with the nanowire probe tip than with a Si probe tip. The
error distances are 40–100 nm in Fig. 11B and 160–200 nm in Fig.
Lantz et al. reported the lateral spring constants of the tip apexes
along the last 300–400 nm of the axes of Si and Si N tips as 84 and
12B. As far as we have observed, most tip radii of Ni nanowires are
3
4
Downloaded on 2015-03-11 to IP 132.174.255.116 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, 156 ͑10͒ D431-D438 ͑2009͒
D437
were 0.4–1.3 ␮m. The force constant of this cantilever was
−
1
ϳ1 N m , as mentioned previously ͑Fig. 9B͒. Because the tip
protruded below the bottom of the Si chip, the vertical displacement
of the cantilever head at the time of impact on a flat substrate cor-
responded to the overall protruding length of the tip ͑20–30 ␮m͒.
Hence, the normal load exerted on the tip was approximately on the
order of ϳ2 ϫ 10 to ϳ3 ϫ 10 ␮N. A cluster of nanowires, which
was seen at location ͑a͒, completely disappeared from the fixed end
in Fig. 14. The nanowire of ͑b͒ was inelastically deformed in the
middle where the lower wire provided the fixed support.
An isolated nanowire of ͑c͒ is suitable to examine the rightmost
term of Eq. 4. The L, d , and dB were 1.0 ␮m, 320 nm, and 200
A
nm, respectively. The free end of the nanowire ͑c͒ horizontally trav-
2
eled 4.4 ϫ 10 nm due to elastic breakdown at the fixed end while
maintaining the straight axis. The reported critical shear stress of Ni
is usually 1/10th to 1/30th of the Ni bulk shear modulus, G, of 80
4
1,42
GPa.
The maximum free-end deflection, ␦x=L, in the elastic
2
range is thus ͑1–3͒ ϫ 10 nm by Eq. 4, which is below 4.4
2
ϫ 10 nm. The inelastic deformation of the nanowire ͑c͒ agrees
with classical mechanics.
Mechanical strength and stress measurements using the AFM
4
3-45
technique were reported recently,
but the tested nanowires were
basically distributed on the substrate. The nanowire AFM probe can
also be a tool to investigate the mechanical properties of metal nano-
wires by different approaches.
Figure 14. SEM image of inelastically deformed nanowires on the AFM
probe tip previously shown in Fig. 9. The resulting displacements are indi-
cated by the vectors. Scale bar: 1 ␮m.
In summary, we have demonstrated that electrodeposition into a
porous template is an effective method of producing metal nanowire
AFM probes. The produced nanowire probe is force sensitive
enough to obtain a surface topographic image. Still, the number of
nanowires protruding from the tip was more than 1. The average
distance of the first nearest neighbors of the nanowires is ϳ300 nm
͑center to center͒ with a number density of 3 ϫ 10 cm . This
distance is so short that the nearby wires have similar heights. The
second-longest wire may interfere in the surface imaging. It is there-
fore necessary to singularize the nanowires on the tip. We have
recently accomplished the production of a single metal nanowire
AFM tip, and the wire diameter is also smaller. The details will be
published elsewhere. This is the preliminary report to explore the
applicability of the nanowire tips as probe tools.
−
1
39
3
9 N m , respectively, simulated by the finite element method.
A Ni nanowire has a softer spring constant ͓ktip = 9.9 N m− with
dA = 320 nm, d = 200 nm, L = 2 ␮m, and E = 2.1 ϫ 10 Pa
bulk Ni͒ ͑Appendix B͔͒, which is likely more stable than the tip
apex stiffness of Si and Si N tips during a scan because those tips
1
1
1
B
8
−2 46
͑
3
4
degrade easily.
Typically, the lateral forces in the contact mode are less than 100
nN. Equation 1 therefore calculates the corresponding free-end de-
flections of a nanowire, ␦x=L, as less than 10 nm. Hence, bending of
a nanowire did not account for the error distances in Fig. 12B.
Tip-sample contact stiffness for metal or ceramic samples is typi-
cally in the range of 10 –10 N m , which is even greater than
the bending stiffness of a nanowire. The used AFM probe tip was
observed with an SEM. The ends of all nanowires were observed to
be at the same positions as before. It is supposed that the deforma-
tion of the nanowire, which contacted the surface, was in the elastic
range during the scan.
2
3
−1 40
Conclusions
We described an electrochemical approach for producing metal-
lic nanowire AFM probes. The FIB-induced metal deposition tech-
nique was used to deposit a Pt cathode on the tip apex of a Si AFM
probe in contact with a porous template for the growth of nanowires.
Control of the Pt deposition area and the electrodeposition duration
is critical for the resulting probe tip structure. Although Ni elec-
trodeposition was performed in this report, our procedure is general
and can be extended to the electrodeposition of several metals.
The produced nanowire AFM probe was force sensitive enough
to show compatibility with a conventional AFM instrument. The
probe nanowire produced a surface topographic image. Inelastic de-
formations of the nanowires were also observed when an excess
load was applied on the tip.
When a nanowire has a moment distribution along its length due
to a lateral force on the free end, the maximum normal stress at any
cross section is given by the equation below
F͑L − x͒
32F͑L − x͒
␴x
=
=
͓3͔
3
Sx
x
␲
ͫ
dB + ͑dA − dB͒
ͬ
L
where S is the section modulus ͑Eq. A-1͒. ␴ is the tensile at one
x
x
side of the probe nanowire and compressive at the other side. The ␴
x
monotonically decreases with x, whereby the maximum stress at the
section of the largest bending moment can be found by substituting
x = 0
Acknowledgments
The authors thank Wonyoung Lee for the AFM measurements
and Professor Turgut Gür and Dr. Rainer Fasching for their valuable
discussions.
3
2FL 3EdA
␴max
=
=
␦x=L
͓4͔
3
2L2
␲
d
B
Hence, the fixed end breaks the elasticity if a lateral force on the end
increases gradually.
Stanford University assisted in meeting the publication costs of this ar-
ticle.
Figure 14 shows the nanowire AFM probe tip, which is identical
to the tip shown in Fig. 9. This tip was viewed after being acciden-
tally flipped, with the nanowires facing down on a flat substrate. It is
apparent that several nanowires located around the center were in-
elastically deformed in Fig. 14. A comparison between Fig. 9 and 14
reveals that the horizontal travel distances of the nanowire ends
Appendix A
The moment of inertia, Ix, and section modulus, Sx, of an inverted-conical-shaped
beam having a larger diameter dA at the free end and a smaller diameter dB at the fixed
4
7
end are given by
Downloaded on 2015-03-11 to IP 132.174.255.116 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

D438
Journal of The Electrochemical Society, 156 ͑10͒ D431-D438 ͑2009͒
4
3
8. R. O’Hayre, G. Feng, W. D. Nix, and F. B. Prinz, J. Appl. Phys., 96, 3540 ͑2004͒.
9. P. De Wolf, R. Stephenson, T. Trenkler, T. Clarysse, T. Hantschel, and W. Vande-
vorst, J. Vac. Sci. Technol. B, 18, 361 ͑2000͒.
0. R. D. Piner, J. Zhu, F. Xu, S. Hong, and C. A. Mirkin, Science, 283, 661 ͑1999͒.
x
x
␲
ͫ
dB + ͑dA − dB͒
ͬ
␲
ͫ
dB + ͑dA − dB͒
ͬ
L
L
Ix =
Sx =
͑dA Ͼ dB͒ ͓A-1͔
6
4
32
1
1
1
1. R. Garcia, R. V. Martinez, and J. Martinez, Chem. Soc. Rev., 35, 29 ͑2006͒.
where L is the length of the beam, and x is a given distance from the fixed end along the
2. W. Lee and F. B. Prinz, J. Electrochem. Soc., 156, G125 ͑2009͒.
3. H. Dai, J. H. Hafner, A. G. Rinzler, D. T. Colbert, and R. E. Smalley, Nature
(London), 384, 147 ͑1996͒.
length
1
d␦ 2x F͑L − x͒
=
͓A-2͔
͓A-3͔
dx2
14. J. H. Hafner, C.-L. Cheung, and C. M. Lieber, J. Am. Chem. Soc.,
121, 9750
EIx
͑1999͒.
1
5. T. W. Ebbesen, H. J. Lezec, H. Hiura, J. W. Bennett, H. F. Ghaemi, and T. Thio,
d␦x 64F
L − x
Nature (London), 382, 54 ͑1996͒.
16. H. Dai, E. W. Wong, and C. M. Lieber, Science, 272, 523 ͑1996͒.
17. P. G. Collins, M. Hersam, M. Arnold, R. Martel, and Ph. Avouris, Phys. Rev. Lett.,
=
4dx
͵
dx
␲E
x
ͫ
dB + ͑dA − dB͒
ͬ
L
8
6, 3128 ͑2001͒.
Here X is defined as
18. D. P. Burt, N. R. Wilson, J. M. R. Weaver, P. S. Dobson, and J. V. Macpherson,
Nano Lett., 5, 639 ͑2005͒.
x
1
9. J. V. Macpherson and P. R. Unwin, Anal. Chem., 73, 550 ͑2001͒.
20. C. Kranz, G. Friedbacher, and B. Mizaikoff, Anal. Chem., 73, 2491 ͑2001͒.
1. J. L. Fernández, D. A. Walsh, and A. J. Bard, J. Am. Chem. Soc., 127, 357 ͑2005͒.
X = dB + ͑dA − dB͒
͓A-4͔
͓A-5͔
͓A-6͔
L
2
d␦x 64F L2
1
3X3
1
1
dB
22. H. Shin, P. J. Hesketh, B. Mizaikoff, and C. Kranz, Sens. Actuators B, 134, 488
͑2008͒.
3. S.-J. Bai, T. Fabian, F. B. Prinz, and R. J. Fasching, Sens. Actuators B, 130, 249
͑2008͒.
=
ͫ
ͫ
−
−
ͩ
−
+
ͪ
ͬ
+ C1
dx
␲E dA − dB
dA − dB
2X2 3X3
2
L3
6
4F
1
1
1
dB
␦
x =
−
ͩ
−
ͪ
ͬ
+ C1x + C2
24. C. M. Lieber, Solid State Commun., 107, 607 ͑1998͒.
␲
E ͑dA − dB͒2 6X2 dA − dB 2X 6X2
2
5. T. M. Whitney, J. S. Jiang, P. C. Searson, and C. L. Chien, Science, 261, 1316
͑1993͒.
d␦x/dx ͉x=0 = 0 at x = 0 ͑X = dB͒
2
2
6. M. Yasutake, T. Kaito, and S. Wakiyama, AIP Conf. Proc., 696, 256 ͑2003͒.
7. T. Tao, J. Ro, J. Melngallis, Z. Xue, and H. D. Kaesz, J. Vac. Sci. Technol. B, 8,
1826 ͑1990͒.
6
4F L2
1
1
C1 =
−
͓A-7͔
͓A-8͔
͓A-9͔
ꢀ
ͫ
3
B
2
ꢁ
␲
E dA − dB 3d
B
A
6d ͑d − dB͒
2
8. D. Spoddig, K. Schindler, P. Rödiger, J. Barzola-QuiQuia, K. Fritsch, H. Mulders,
and P. Esquinazi, Nanotechnology, 18, 495202 ͑2007͒.
␦
x=0 = 0 at x = 0 ͑X = dB͒
L3
29. M. Motoyama, Y. Fukunaka, T. Sakka, Y. H. Ogata, and S. Kikuchi, J. Electroanal.
6
4F
1
1
C2 =
−
+
ͬ
Chem., 584, 84 ͑2005͒.
30. T. W. Cornelius, M. E. Toimil-Molares, S. Karim, and R. Neumann, Phys. Rev. B,
77, 125425 ͑2008͒.
1. D. Al-Mawlawi, C. Z. Liu, and M. Moskovits, J. Mater. Res., 9, 1014 ͑1994͒.
32. G. Sauer, G. Brehm, S. Schneider, K. Nielsch, R. B. Wehrspohn, J. Choi, H.
E ͑dA − dB͒2
By substituting x = L ͑X = dA͒ into Eq. A-6
4F L3
E 3dAd3
2
B
3dB͑dA − dB͒
␲
6d
3
6
␦
x=L =
␲
Hofmeister, and U. Gösele, J. Appl. Phys., 91, 3243 ͑2005͒.
B
3
3
3. Introduction to Focused Ion Beams—Instrumentation, Theory, Technique and Prac-
tice, L. A. Giannuzzi and F. A. Stevie, Editors, Springer Science, New York ͑2005͒.
4. A. Yasaka, T. Yamaoka, T. Kaito, and T. Adachi, J. Vac. Sci. Technol. B, 9, 2642
͑1991͒.
Appendix B
The following relation is given by the classical explanation ͑Frank–Read source͒ of
the propagation of multiple dislocations during plastic deformation of a metal
4
8
35. P. Chen, P. F. A. Alkemade, and H. W. M. Salemink, Jpn. J. Appl. Phys., 47, 5123
2008͒.
͑
2
Gb
s
3
6. C. Schönenberger, B. M. I. van der Zande, L. G. J. Fokkink, M. Henny, C. Schmid,
␶
=
͓B-1͔
M. Krüger, A. Bachtold, R. Huber, H. Birk, and U. Staufer, J. Phys. Chem. B, 101,
5
497 ͑1997͒.
where ␶ is the shear stress exerted on a slip plane, b is the Burgers vector of the
dislocation, and s is the distance between the pinning sites of the dislocation line. By
substituting ␶ Ϸ 10 GPa, G = 80 GPa, and b Ϸ 10 nm into Eq. B-1, s is on the
order of 1 nm. This value is much smaller than the nanowire diameter and grain sizes
that we observed by TEM ͑not shown͒. The dimensions of the produced nanowires are
still sufficiently large that plastic deformations are not influenced by the finite size. The
bulk Ni properties are therefore applicable to the calculations in this paper.
3
3
7. J. E. Sader, J. W. M. Chon, and P. Mulvaney, Rev. Sci. Insrum., 70, 3967 ͑1999͒.
8. J. S. Villarrubia, in Applied Scanning Probe Methods, B. Bhushan, H. Funds, and
S. Hosaka, Editors, Springer-Verlag, Berlin ͑2004͒.
0
−1
3
9. M. A. Lantz, S. J. O’Shea, A. C. F. Hoole, and M. E. Welland, Appl. Phys. Lett.,
8
0, 970 ͑1997͒.
4
4
4
4
0. U. Rabe, K. Janser, and W. Arnold, Rev. Sci. Instrum., 67, 3281 ͑1996͒.
1. A. C. Lund and C. A. Schuh, Acta Mater., 53, 3193 ͑2005͒.
2. S. G. Srinivasan and M. I. Baskes, Metall. Mater. Trans. A, 38, 2716 ͑2007͒.
3. E. W. Wong, P. E. Sheehan, and C. M. Lieber, Science, 277, 1971 ͑1997͒.
References
1
2
3
4
5
6
7
. G. Binnig, H. Rohrer, Ch. Gerber, and E. Weibel, Phys. Rev. Lett., 50, 120 ͑1983͒.
. G. Binnig and H. Rohrer, IBM J. Res. Dev., 30, 355 ͑1986͒.
. G. Binnig, C. F. Quate, and Ch. Gerber, Phys. Rev. Lett., 56, 930 ͑1986͒.
. Y. Martin and H. K. Wickramasinghe, Appl. Phys. Lett., 50, 1455 ͑1987͒.
. W. A. Ducker, T. J. Senden, and R. M. Pashley, Nature (London), 353, 239 ͑1991͒.
. R. C. Barrett and C. F. Quate, J. Appl. Phys., 70, 2725 ͑1991͒.
44. B. Wu, A. Heidelberg, and J. J. Boland, Nature Mater., 4, 525 ͑2005͒.
45. J. Song, X. Wang, E. Riedo, and Z. L. Wang, Nano Lett., 5, 1954 ͑2005͒.
46. A. Milchev, W. S. Kruijt, M. Sluyters-Rehbach, and J. H. Sluyters, J. Electroanal.
Chem., 350, 89 ͑1993͒.
47. J. M. Gere, Mechanics of Materials, 6th ed., p. 331, Brooks/Cole-Thomas Learn-
ing, Belmont ͑2004͒.
. M. Nonnenmacher, M. P. O’Boyle, and H. K. Wickramasinghe, Appl. Phys. Lett.,
48. M. A. Meyers and K. K. Chawla, Mechanical Behavior of Materials, p. 231,
Prentice-Hall, Englewood Cliffs, NJ ͑1999͒.
5
8, 2921 ͑1991͒.
Downloaded on 2015-03-11 to IP 132.174.255.116 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).