Fabrication of nanopatterned metal layers on silicon by nanoindentation/nanoscratching and electrodeposition

Article abstract of DOI:10.1016/j.electacta.2010.01.024

The present work illustrates a novel approach for the maskless and resistless fabrication of nanopatterned metal layers on Si substrates, based on the combination of nanomechanical surface modification techniques (such as nanoindentation and nanoscratching) and electrodeposition. Single crystal (1 0 0) n-doped Si substrates were first cleaned from native oxide. Nanoindentation and nanoscratching were then used to locally change the substrate microstructure and create regions with reduced electrical conductivity. The substrates were finally mounted as cathode electrodes in a three-electrode electrochemical cell to potentiostatically deposit a Ni layer. Electrodeposition was prevented in regions with modified microstructure, enabling the formation of a patterned Ni layer. The fabrication of several patterns including continuous Ni lines of 200 nm width and several microns length was obtained.

Full text of DOI:10.1016/j.electacta.2010.01.024

Electrochimica Acta 55 (2010) 3355–3360
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Fabrication of nanopatterned metal layers on silicon by
nanoindentation/nanoscratching and electrodeposition
a,∗
a
a
b
b
R. Cecchini , A. Fabrizi , C. Paternoster , Wei Zhang , G. Roventi
a
VINF/Marche Polytechnic University, Department of Mechanics, v. Brecce Bianche, 60131 Ancona, Italy
Marche Polytechnic University, Department of Physics and Material and Territory Engineering, v. Brecce Bianche, 60131 Ancona, Italy
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 July 2009
Received in revised form 7 January 2010
Accepted 7 January 2010
Available online 15 January 2010
The present work illustrates a novel approach for the maskless and resistless fabrication of nanopatterned
metal layers on Si substrates, based on the combination of nanomechanical surface modification tech-
niques (such as nanoindentation and nanoscratching) and electrodeposition. Single crystal (1 0 0) n-doped
Si substrates were first cleaned from native oxide. Nanoindentation and nanoscratching were then used to
locally change the substrate microstructure and create regions with reduced electrical conductivity. The
substrates were finally mounted as cathode electrodes in a three-electrode electrochemical cell to poten-
tiostatically deposit a Ni layer. Electrodeposition was prevented in regions with modified microstructure,
enabling the formation of a patterned Ni layer. The fabrication of several patterns including continuous
Ni lines of 200 nm width and several microns length was obtained.
Keywords:
Maskless nanopatterning
Nanoindentation
Electrodeposition
Silicon
© 2010 Elsevier Ltd. All rights reserved.
Phase transformation
1
. Introduction
Microelectronic and micromechanical devices industry is in
patterned magnetic metal layers [6–7]. The use of a resist for
substrate patterning has the disadvantage of requiring several fab-
rication steps, including resist deposition, curing and removal.
Several approaches for maskless and resistless nanopatterning
have been attempted, including the use of nanomechanical tech-
niques. Nanoindentation and nanoscratching are well known tools
for changing as well as for measuring the properties of sur-
faces at nanometric scale. The combination of nanoindentation
and nanoscratching with electrochemical deposition as a way
for obtaining metal nanopatterning on semiconductor surfaces
has been investigated by other authors [8–14]. Selective metal
electrodeposition and electroless deposition within the scratches
produced by an atomic force microscope (AFM) or a nanoinden-
ter in an oxide or organic insulating layer has been demonstrated
[8–12]. In other works defects were created by nanoindentation
on hydrogen-terminated Si surface, where metal deposits were
grown by electroless deposition, as a consequence of modified local
surface activity [13,14]. In these types of positive lithographic pro-
cesses deposition occurs on nanomechanically modified areas of
the substrate.
constant need of high resolution and low cost fabrication pro-
cesses. One critical aspect is represented by the processes involved
in the transfer of a given pattern onto a substrate. Often the sub-
strate is constituted by a semiconductor material, like Si, onto
which a patterned metal layer needs to be grown for the fabrica-
tion of micro-electro-mechanical systems (MEMS) and electronic,
optoelectronic and magnetic devices. Metal nanostructures (e.g.,
nanometric clusters) on semiconductors can also be used as cat-
alysts for chemical reactions and for the fabrication of sensors.
Moreover, metal/semiconductor Schottky junctions have a huge
number of applications (e.g., solar cells). Several processes, includ-
ing electrodeposition, are studied for the fabrication of Schottky
diodes [1–4].
Typically, lithographic processes such as photolithography are
used for transferring the pattern from a mask to a photoresist mate-
rial layer, previously deposited on the substrate. Subsequently,
deposition, etching or other steps are performed in order to
add and subtract material as required for transferring the pat-
tern on the substrate [5]. Among other deposition processes,
electrodeposition in combination with photolithography repre-
sents an effective and cost efficient way for fabricating, e.g.,
On the other hand, nanoindentation and nanoscratching can
also be used to change the microstructure, e.g., the phase, of
the indented or scratched material. It is for instance known that
nanoindentation and nanoscratching of single crystal Si can induce
amorphisation in a limited volume of material near the surface
[
15–18]. The use of scratching was for instance considered for the
∗ Corresponding author. Tel.: +39 071 2204794; fax: +39 071 2204801.
E-mail address: r.cecchini@univpm.it (R. Cecchini).
fabrication of amorphous patterns on single crystal Si, subsequently
used as an etching masking [19].
0
013-4686/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2010.01.024

3
356
R. Cecchini et al. / Electrochimica Acta 55 (2010) 3355–3360
Aim of this work is to study a novel approach to the mask-
less and resistless fabrication of nanopatterned metal layers on Si,
where nanomechanically induced local amorphisation of the sub-
strate is used as a mask for electrodeposition. By this approach,
the combination of nanomechanical surface modification and elec-
trodeposition results in a negative type of lithography for metal
nanopatterning.
Patterned Ni layers on n-doped single crystal (1 0 0) Si substrates
were fabricated using nanoindentation and nanoscratching fol-
lowed by electrodeposition and characterized by scanning electron
microscopy (SEM).
2
. Experimental
Commercially available 380 ␮m thick n-type single crystal
(
1 0 0) Si wafers (phosphorous doped with resistivity of 1–10 ꢀ cm)
with one mechanically polished side were cut into 1 cm × 1 cm sam-
ples and immersed in 3.5% HF solution for 10 min to remove native
SiO2 layer and leave a hydrogen-terminated surface. A Hysitron®
UBI nanoindenter with a Berkovich diamond tip (curvature radius
∼
1 ␮m) was used to perform nanoindentation and nanoscratching
on the Si samples top surface. Nanoindentations were performed by
applying a force perpendicularly to the substrate surface (z direc-
tion) with loading and unloading sequences of 10 s and peak forces
of 1000–10000 ␮N. Using its scanning probe mode, the nanoin-
denter was also operated for nanoscratching, by simultaneously
applying a force along the z direction and scanning the tip in the
lateral, x–y, directions. Nanoscratching was done at constant forces
Fig. 1. Schematic description of the process steps for maskless and resistless pat-
terning: (a) n-doped single crystal (1 0 0) Si substrate covered with native oxide
layer; (b) native oxide removal by immersion in HF solution; (c) Si surface nanoin-
dentation or nanoscratching by diamond tip, which generates amorphous regions
on Si; (d) Ni electrodeposition on unmodified Si regions.
−
1
of 200, 300, 400 and 500 ␮N and scanning rate of 130 ␮m s
.
After nanomechanical surface modification step, an ohmic con-
tact was fabricated on the Si samples back surface using InGa
eutectic (99.99%). The samples were then mounted on a copper
5
49 potentiostat connected to an AMEL® model 568 programmable
support. An insulating adhesive tape was used to mask off the sup-
port and the sample leaving exposed an area of 0.196 cm2 on the Si
generator function for deposition times in the range 0.2–2 s. The
current density (j)–time (t) curves measured during electrodepo-
sition were recorded by a computer connected to the potentiostat.
top surface.
Si amorphisation in the scratched regions was investigated by
®
®
After electrodeposition, samples were investigated by a Philips
micro-Raman spectroscopy, using a Jobin-Yvon micro-Raman sys-
XL20 SEM equipped with an Edax® energy dispersive X-ray (EDX)
spectroscopy system.
tem with 632.8 nm excitation laser, and by conductive atomic force
®
microscopy (AFM), using a NT-MDT Solver P47 Pro AFM operated
A schematic description of the present approach for maskless
and resistless metal nanopatterning on Si is illustrated in Fig. 1.
After native oxide layer removal by immersion in HF solution, Si
surface is modified by nanoindentation or nanoscratching, which
generates amorphous regions on Si. Finally Ni electrodeposition is
carried out and a Ni layer is grown on unmodified Si regions.
with Pt coated probes.
A conventional three electrode cell was used to perform cyclic
voltammetry (CV) and electrodeposition; a platinum spiral was
the counter-electrode and an Ag/AgCl (E = +0.208 vs. SHE) was the
reference electrode, to which all potentials are referred to. The ref-
erence electrode was mounted inside a Luggin capillary, whose
tip was placed next to the working electrode surface. A Watts
type bath was used for electrochemical tests: NiSO ·6H O 1.14 M,
3. Results and discussion
4
2
−
3
NiCl ·6H O 0.15 M, H BO 0.65 M and saccharine 1 g dm . All the
2
2
3
3
solutions were prepared with double-distilled water and analytical
grade reagents. All electrochemical experiments were performed
First, force (P)–penetration depth (h) curves were used to char-
acterize Si substrates behaviour during nanoindentation. A typical
example of recorded P–h curves is reported in Fig. 2. These curves
are similar to those reported in the literature [15–17] and are
related to the indentation parameters, to the substrate mechani-
cal properties and to mechanically induced phase transformations
in the substrate. It is known that in nanoindentation atmospheric
phase Si-I (diamond structure) transforms into a metallic Si-II (b-Sn
structure) phase on loading when a critical pressure (11–13 GPa) is
attained under the indenter tip [20]. Other phase transformations
are possible on unloading. The metallic phase can transform into
a-Si or into high pressure polycrystalline phases Si-III (bc8, body-
centered cubic structure) and Si-XII (r8, rhombohedral structure).
These transformations are accompanied either by a discontinuity,
called “pop-out”, or by a so-called “elbow” in the unloading part
of the P–h curve. Strong correlation has been found between the
“pop-out” and the appearance of Si-III and Si-XII, while the pres-
ence of the “elbow” is linked to the appearance of a-Si [15]. It has
◦
at room temperature (T = 25 C). Bath composition and deposition
temperature were chosen in order to obtain nanocrystalline Ni
deposits. In fact, preliminary tests indicated that addition of sac-
charine and deposition at room temperature favours the reduction
of Ni grain size, as measured by Scherrer’s equation, using the full-
width at half maximum (FWHM) of diffraction peaks recorded from
®
electrodeposited Ni layers with a Philips PV 1730 X-ray diffrac-
tometer (XRD) with ꢁCu K␣ = 0.154 nm. Before each electrochemical
experiment, solutions were deaerated by N2 bubbling inside the
cell. CV was performed to characterize the electrochemical sys-
tem and establish Ni deposition potential on Si. CV tests were
carried out using an EG&G Princeton Applied Research poten-
tiostat/galvanostat model 273 controlled by a personal computer;
the scanning rate was 20 mV s . Electrodeposition experiments
were performed potentiostatically with deposition potential in the
range of −1.000 to −3.000 V and carried out by an AMEL® model
®
−1

R. Cecchini et al. / Electrochimica Acta 55 (2010) 3355–3360
3357
Fig. 2. Force (P) as function of penetration depth (h) plot for loading and
unloading during nanoindentation of n-doped single crystal (1 0 0) Si sub-
strate with a Berkovich diamond tip: peak load = 10000 ␮N; loading–unloading
−1
rate = 1000 ␮N s . The arrow marks the “elbow” in the unloading part of the curve,
indicating formation of a-Si.
been also proved that the above mentioned transformations on the
Si indented volume can be controlled by controlling the indenta-
tion parameters: tip geometry, peak load, loading and unloading
rates [21]. It was argued that Si-II → Si-XII/Si-III transformation can
only occur when sufficient time for lattice reconstruction is given
to the indented material. a-Si formation is therefore favoured by
fast unloading. a-Si formation is also favoured when unloading is
performed from low peak loads, as the volume of modified material
is in such cases too low to allow a sufficient number of nucleation
sites to be created for the reconstructive Si-II → Si-XII/Si-III transi-
tion. The recorded curves, such as the one reported in Fig. 2, with the
presence of an “elbow” on the unloading part of the P–h curve, sug-
gest that under present experimental conditions, and in particular
the relatively low peak loads used, a-Si is formed during nanoinden-
tation. Using the nanoindenter for nanoscratching, a-Si lines were
also created on the Si surface. Formation of a-Si during lubricated
and dry scratching has previously been observed [17–19].
Fig. 3. (a) Raman shift obtained from an unmodified crystal Si area (solid line) and
on an area after nanoscratching with a constant force of 300 ␮N (dotted line); (b)
Raman shift obtained by subtraction of Raman spectra obtained from unmodified
crystal Si area to the Raman spectra obtained from the scratched area. The arrows
mark the bands corresponding to a-Si.
Fig. 3(a) reports the results of micro-Raman investigation on a
scratched and on an unmodified Si region. Subtraction of the two
spectra (scratched Si and reference unmodified Si) is reported in
Fig. 3(b). The comparison with Raman measurements reported in
the literature [15–17] confirms the attribution of the broad bands
Si substrates. No anodization peak is observed during the reverse
scan on native oxide covered Si substrates, and SEM investigations
and EDX analysis confirmed that no Ni dissolution occurs during
the reverse scan on such samples. Hence, the presence of native
−1
at 160, 310, 400 and 470 cm to a-Si phase.
Electrical characterisation of the a-Si layer formed dur-
ing scratching was done by conductive AFM. Fig. 4 reports
the current–voltage curves measured on both scratched and
unmodified Si areas. The current–voltage curve measured on
unmodified crystal Si is determined by the properties of the
metal/semiconductor junction formed between the Pt of the AFM
tip and the crystal n-Si: the initial exponential increase of the cur-
rent with forward voltage bias (n-Si substrate negative with respect
to AFM tip) is compatible with the presence of a Schottky barrier
at the junction. On the other hand, no forward current is observed
on scratched areas up to +3.0 V forward voltage bias; while only a
low value of current is present above this value of bias.
Characteristics of the electrolyte/Si systems were investigated
by CV tests, performed both for as received native oxide covered
and for HF cleaned Si samples (Fig. 5). Ni electrodeposition starts at
about −0.920 V on HF cleaned Si, while it starts at about −1.300 V
on native oxide covered Si. During the reverse scan Ni deposition
continues also for more positive potential values; this is due to the
fact that in the reverse scan Ni deposition occurs on the metallic
layer deposited during the direct scan. An anodization peak, due to
Ni dissolution, is observed during the reverse scan on HF cleaned
Fig. 4. Current–voltage measurement by conductive AFM recorded on an unmodi-
fied crystal Si area (-ꢀ-) and on an area scratched with a constant force of 300 ␮N
(
-ꢁ-). Positive voltage bias corresponds toAFM tip at apositivepotential with respect
to Si substrate.

3
358
R. Cecchini et al. / Electrochimica Acta 55 (2010) 3355–3360
Fig. 5. Cyclic voltammogram obtained for as received (dashed line) and HF
cleaned (solid line) n-doped (1 0 0) Si samples in the Watts type bath. Scanning
−1
rate = 20 mV s
.
SiO2 layer on the Si substrate surface shifts Ni deposition potential
towards more negative values compared to oxide free Si surface
during the direct scan, as well as hindering Ni dissolution during
the reverse scan. The absence of anodization during the reverse
scan of the CV in presence of SiO2 can be in fact explained as a
result of the electric properties of the Ni/substrate junction formed
during the forward scan of the CV. Current–voltage measurements
on electrodeposited Ni/n-Si and Ni/SiO /n-Si samples showed a rec-
2
tifying behaviour for both types of junctions, with an exponential
increase of current as a function of voltage for forward bias (n-Si
negative with respect to Ni layer, as during forward scan of the
CV) and a constant value of current recorded for reverse bias (n-
Si positive with respect to Ni layer, as during reverse scan of the
CV). For Ni/n-Si samples, the value of measured reverse current
−
2
(
∼1 mA cm ) is high enough to allow Ni dissolution during the
reverse scan of the CV. For Ni/SiO /n-Si reverse current is extremely
2
−
3
−2
low (∼1 × 10 mA cm ) and is not sufficient to allow significant
Ni dissolution during the reverse scan of the CV. Similar properties
of electrodeposited metal/semiconductor junctions were reported
in other works [1–4]. The effect of SiO2 interlayer on the reduction
of the reverse current of the junctions is also expected [5].
Ni electrodeposition was then carried out on nanomechanically
modified Si substrates. Fig. 6(a) shows an SEM image of a Si sample
where Ni electrodeposition was performed after nanoindentation
with peak load of 10000 ␮N. It is seen that a continuous Ni film
is formed on the unmodified Si surface, whereas no film growth
takes place in the indented region, as a result of amorphisation of
Si and consequent reduced electrical conductivity, as described by
Fig. 4. Fig. 6(b) reports an SEM image of a sample where indenta-
tion was performed with 1000 ␮N. As it can be seen, also for this
value of peak load amorphisation of Si was obtained. Some Ni depo-
sition defects appear near the edges of the imprint, probably as a
consequence of not complete amorphisation of Si in the indented
area.
This indicates the possible use of nanomechanical modifica-
tion of Si surface and electrodeposition for fabricating patterned
metal layers. Incidentally, this procedure can also be useful as a
complementary technique for studying nanomechanically induced
phase transformation on Si. This has been the subject of several
investigations, by transmission electron microscopy (TEM), Raman
spectroscopy and electrical conductivity measurements by con-
ductive indentation tips and scanning probe microscopy (SPM)
techniques [15,22–25]. The extension and shape near surface of the
Si region with significant reduction of electrical conductivity, due
Fig. 6. SEM image of
load = 10000 ␮N and loading–unloading rate = 1000 ␮N s
a
Si sample after: (a) nanoindentation with peak
−1
followed by Ni elec-
trodeposition at −1.200 V for 4.0 s; (b) nanoindentation with peak load of 1000 ␮N
−
1
and loading–unloading rate = 1000 ␮N s
−2.000 V for 0.6 s (b).
followed by Ni electrodeposition at
to microstructural changes, can be quickly evaluated as it corre-
sponds to the region where no electrodeposition takes place. From
Fig. 6(a) and (b) it can be seen that the amorphised region follows
the shape of the indenter Berkovich tip.
A typical j–t curve recorded during Ni electrodeposition is
shown in Fig. 7. Cathodic current density rises up with t, reaches
a maximum value and then starts falling. This is consistent with
a deposition process that proceeds by nucleation followed by
nuclei growth. Three-dimensional nucleation and growth is often
observed for metal electrodeposition on semiconductors, as a
consequence of weak interaction between metal adatoms and sub-
strate [26,27]. Film growth by coalescence of Ni nuclei is confirmed
by SEM images (Fig. 6).
Using the nanoindenter for nanoscratching, a-Si lines were also
created on the Si surface. Formation of a-Si during lubricated and
dry scratching has previously been observed [17–19]. Using the
nanoindenter for nanoscratching and by controlling the nanoin-
denter tip movement in the x–y directions different a-Si patterns,
such as lines, squares and arrays were created on Si. Electrodepo-
sition followed, generating a Ni patterned layer.

R. Cecchini et al. / Electrochimica Acta 55 (2010) 3355–3360
3359
Fig. 7. Current density transient (j–t) recorded during Ni electrodeposition from the
Watts type bath at deposition potential of −1.600 V.
In Fig. 8 is reported the image of two 15 ␮m × 30 ␮m rectangu-
lar patterns separated by a 1 ␮m wide line, fabricated by scanning
the nanoindenter tip with a constant force of 300 ␮N along the x
direction followed by electrodeposition of a 200 nm thick Ni film.
The distance between the scanned lines, 100 ␮m in this case, was
chosen so that the amorphised lines were overlapping. This pre-
vents Ni film to grow inside the scanned areas. Absence of Ni in
the scanned area was verified by EDX measurements. It is impor-
tant to notice here that the masking properties of the amorphised
Si are extremely good. No Ni electrodeposition occurred inside the
patterns fabricated by scratching at constant force of 300 ␮N for
deposition potentials as negative as −3.00 V. This is a much more
negative potential than the value at which electrodeposition starts
on Si covered by native SiO2 (Fig. 4), increasing the deposition
selectivity of the present approach compared to those based on
masking by native oxide. Considering that electrodeposition starts
on hydrogen-terminated samples at −0.920 V, this gives a wide
range of deposition potentials to be used for Ni electrodeposition.
It has also to be noticed that, if needed, the amorphous Si layer
produced by nanoindentation and nanoscratching can be easily
removed after electrodeposition by HF solution clean [19].
Fig. 9. SEM image at low (a) and high (b) magnification of parallel 200 nm wide and
100 nm thick Ni lines fabricated by scanning the nanoindenter with a constant force
of 300 ␮N along lines separated by 250 nm and Ni electrodeposition performed with
potential of −2.000 V for 0.6 s.
phisation of Si is still obtained. In [19] diamond tips with curvature
radius of 300–1000 nm were used for scratching a Si substrate,
forming a-Si patterns. It was found that vertical forces as low as
1
80 ␮N are sufficient to create a layer of a-Si of few nanometres of
thickness.
Scanning the surface along further apart lines allowed the fabri-
cationofparallel Ni linessuchasthose reported inFig. 9. Continuous
and well separated lines with several microns length and sub-
micron width were electrodeposited. The quality of the deposits
obtained with the present approach seems to be improved with
respect to other methods based on the scratching through an oxide
layer when considering the submicrometric range [9,10]. One of
the reasons of such improvement can be found in the fact that in
the present approach the deposition occurs on the unmodified, and
hence flat, Si surface rather than on scratched, and hence rough,
regions (e.g., grooves). Moreover, in Refs. [9,10], it was reported
that nucleation starts at the AFM-scratch edges and metal lines
result from the coalescence of two separate lines into a single one,
as a consequence Si amorphisation in the centre of the groove. This
effect does not represent an issue with the present approach, as
the formation of a-Si is precisely the principle exploited to obtain
nanopatterning. The metal lines formed by our approach do not
present this double feature characteristic.
The smallest force at which no Ni deposition was observed on
scratches was 200 ␮N. It can be inferred that for such load amor-
The smallest line width obtained was below 100 nm, although
these lines were discontinuous (Fig. 10). Finally, by scanning the
nanoindenter in two perpendicular directions, arrays of Ni islands
were electrodeposited. Fig. 11 reports the SEM image of one of
such arrays with nanometric Ni deposits. The array was obtained by
scanning the nanoindenter tip with a force of 300 ␮N along lines
with 250 nm separation first in the x and then in the y direction,
followed by Ni electrodeposition at −2.300 V for 0.2 s.
Fig. 8. SEM image of two rectangular patterns in a 200 nm thick Ni film separated
by 1 ␮m wide Ni line, fabricated by scanning the nanoindenter tip with a constant
force of 300 ␮N along lines separated by 100 nm and Ni electrodeposition performed
with potential of −1.300 V for 3.0 s.

3
360
R. Cecchini et al. / Electrochimica Acta 55 (2010) 3355–3360
electrodeposition. Nanoindentation and nanoscratching were used
to induce localized amorphisation on the surface of n-doped single
crystal (1 0 0) Si substrates. The regions with modified microstruc-
ture were used to mask the substrates during Ni electrodeposition
step from a Watts type bath, leading to the formation of patterned
Ni layers. Several patterns with characteristic dimensions both in
the micrometric and nanometric range were fabricated. As the size
of the amorphous lines created during scratching depends on the
applied force and on tip size, it is foreseen that by using shaper tips
even smaller features could be created. Diamond nanoindentation
tips with curvature radius smaller than 50 nm are already commer-
cially available. In alternative an AFM with diamond or diamond
coated tips could be used instead of a nanoindenter to attain even
smaller feature sizes.
Acknowledgments
Fig. 10. SEM image of parallel Ni lines with width lower than 100 nm. Nanoscratch-
ing was done by scanning the nanoindenter with constant force of 500 ␮N along
parallel lines separated by 250 nm. Ni electrodeposition was done at a potential of
The authors are thankful to Dr. T. Bellezze, Prof. M. Cabibbo, Prof.
R. Fratesi and Prof. S. Spigarelli for useful discussion, to Dr. A. Di
Cristoforo for XRD measurements, to Dr. L. Gobbi and Dr. G. Barucca
for their assistance with SEM investigations, to the Microwave and
Optics Group of Electronic Engineering Dep. (Marche Polytechnic
University) for conductive AFM measurements and to the Physics
Dep. and C.I.G.S. Laboratory of the University of Modena and Reggio
Emilia for Raman and FIB-SEM investigations.
−
2.300 V for 0.2 s.
References
[
1] J. Carrey, K. Bouzehouane, J.M. George, C. Ceneray, T. Blon, M. Bibes, A. Vaurès,
S. Fusil, S. Kenane, L. Vila, L. Piraux, Appl. Phys. Lett. 81 (2002) 760.
2] P. Allongue, K. Souteyrand, J. Electroanal. Chem. 269 (1989) 361.
3] G. Guler, O. Gullu, O.F. Bakkaloglu, A. Turut, Physica B 403 (2008) 2211.
4] T. Zambelli, M.L. Munford, F. Pillier, M.-C. Bernard, P. Allongue, J. Electrochem.
Soc. 148 (2001) C614.
[
[
[
[
[
5] S.M. Sze, Physics of Semiconductor Devices, John Wiley & Sons, New York, 1981.
6] M. Duch, J. Esteve, E. Gómez, R. Pérez-Castillejos, E. Vallés, J. Electrochem. Soc.
149 (2002) C201.
[
7] W. Schwarzacher, O.I. Kasyutich, P.R. Evans, M.G. Darbyshire, G. Yi, V.M.
Fedosyuk, F. Rousseaux, E. Cambril, D. Decanini, J. Magn. Magn. Mater. 198–199
(
1999) 185.
[
[
8] L. Santinacci, T. Djenizian, P. Schmuki, Appl. Phys. Lett. 79 (2001) 1882.
9] L. Santinacci, Y. Zhang, P. Schmuki, Surf. Sci. 597 (2005) 11.
Fig. 11. SEM image of an array of Ni deposits with lattice period of 250 nm, obtained
by scanning the nanoindenter with a constant force of 300 ␮N along two perpen-
dicular directions and Ni electrodeposition performed with a deposition potential
of −2.300 V for 0.2 s.
[
10] L. Santinacci, T. Djenizian, H. Hildebrand, S. Ecoffey, H. Mokadad, T. Campanella,
P. Schmuki, Electrochim. Acta 48 (2003) 3123.
[11] Y. Zhang, E. Balaur, P. Schmuki, Electrochim. Acta 51 (2006) 3674.
12] Y. Zhang, E. Balaur, S. Maupai, T. Djenizian, R. Boukherroub, P. Schmuki, Elec-
trochem. Commun. 5 (2003) 337.
[
Further improvement on quality and scaling down of the pat-
terning could be attained by additional investigation on the effect
of scanning rate during scratching and by use of tips with different
geometry. For what concerns the electrodeposition step, the effect
of deposition potential and other additives besides of saccharine,
on the nucleation and growth stage could be investigated in order
to improve deposit quality.
[13] T. Homma, N. Kubo, T. Osaka, Electrochim. Acta 48 (2003) 3115.
[
14] N. Kubo, T. Homma, Y. Hondo, T. Osaka, T. Osaka, Electrochim. Acta 51 (2005)
34.
15] V. Domnich, Y. Gogotsi, S. Dub, Appl. Phys. Lett. 76 (2000) 2214.
8
[
[16] D. Ge, V. Domnich, Y. Gogotsi, J. Appl. Phys. 93 (2003) 2418.
[
[
[
17] X. Li, J. Lu, Z. Wan, J. Meng, S. Yang, Tribol. Int. 40 (2007) 360.
18] J.W. Park, N. Kawasegi, N. Morita, D.W. Lee, Appl. Phys. Lett. 85 (2004) 1766.
19] J.W. Park, S.S. Lee, B.S. So, Y.H. Jung, N. Kawasegi, N. Morita, D.W. Lee, J. Mater.
Process. Technol. 187–188 (2007) 321.
[
[
[
[
[
[
[
20] J.Z. Hu, L.D. Merkle, C.S. Menoni, I.L. Spain, Phys. Rev. B 34 (1986) 4679.
21] J. Jang, M.J. Lance, S. Wen, T.Y. Tsui, G.M. Pharr, Acta Mater. 53 (2005) 1759.
22] S. Ruffell, J.E. Bradby, N. Fujisawa, J.S. Williams, J. Appl. Phys. 101 (2007) 083531.
23] S. Ruffell, J.E. Bradby, J.S. Williams, J. Appl. Phys. 102 (2007) 063521.
24] S. Ruffell, J.E. Bradby, J.S. Williams, Appl. Phys. Lett. 89 (2006) 091919.
25] S.T. Ho, Y.H. Chang, H.N. Lin, J. Appl. Phys. 96 (2004) 3562.
26] G. Oskam, J.G. Long, A. Natarajan, P.C. Searson, J. Phys. D: Appl. Phys. 31 (1998)
1927.
4
. Conclusions
A novel maskless and resistless processing method for the fab-
rication of patterned metal films on Si was studied. The method
combines nanomechanical surface modification techniques, for
changing the substrate microstructure, and selective film growth by
[27] E. Budevski, G. Staikov, W.J. Lorenz, Electrochim. Acta 45 (2000) 2559.