Patterned growth of individual and multiple vertically aligned carbon nanofibers

Article abstract of DOI:10.1063/1.126705

The results of studies of patterned growth of vertically aligned carbon nanofibers (VACNFs) prepared by plasma-enhanced chemical vapor deposition are reported. Nickel (Ni) dots of various diameters and Ni lines with variable widths and shapes were fabricated using electron beam lithography and evaporation, and served for catalytic growth of VACNFs whose structure was determined by high resolution transmission election microscopy. It is found that upon plasma pre-etching and heating up to 600-700°C, thin films of Ni break into droplets which initiate the growth of VACNFs. Above a critical dot size multiple droplets are formed, and consequently multiple VACNFs grow from a single evaporated dot. For dot sizes smaller than the critical size only one droplet is formed, resulting in a single VACNF. In the case of a patterned line, the growth mechanism is similar to that from a dot. VACNFs grow along the line, and above a critical linewidth multiple VACNFs are produced across the line. The mechanism of the formation of single and multiple catalyst droplets and subsequently of VACNFs is discussed.

Full text of DOI:10.1063/1.126705

Patterned growth of individual and multiple vertically aligned carbon
nanofibers
V. I. Merkulov, D. H. Lowndes, Y. Y. Wei, G. Eres, and E. Voelkl
Citation: Appl. Phys. Lett. 76, 3555 (2000); doi: 10.1063/1.126705
View online: http://dx.doi.org/10.1063/1.126705
View Table of Contents: http://apl.aip.org/resource/1/APPLAB/v76/i24
Published by the AIP Publishing LLC.
Additional information on Appl. Phys. Lett.
Journal Homepage: http://apl.aip.org/
Journal Information: http://apl.aip.org/about/about_the_journal
Top downloads: http://apl.aip.org/features/most_downloaded
Information for Authors: http://apl.aip.org/authors
Downloaded 22 Aug 2013 to 128.248.155.225. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions

APPLIED PHYSICS LETTERS
VOLUME 76, NUMBER 24
12 JUNE 2000
Patterned growth of individual and multiple vertically aligned
carbon nanofibers
V. I. Merkulov,a) D. H. Lowndes, Y. Y. Wei, and G. Eres
Solid State Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831
E. Voelkl
Metals and Ceramics Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831
͑Received 29 November 1999; accepted for publication 17 April 2000͒
The results of studies of patterned growth of vertically aligned carbon nanofibers ͑VACNFs͒
prepared by plasma-enhanced chemical vapor deposition are reported. Nickel ͑Ni͒ dots of various
diameters and Ni lines with variable widths and shapes were fabricated using electron beam
lithography and evaporation, and served for catalytic growth of VACNFs whose structure was
determined by high resolution transmission electron microscopy. It is found that upon plasma
pre-etching and heating up to 600–700 °C, thin films of Ni break into droplets which initiate the
growth of VACNFs. Above a critical dot size multiple droplets are formed, and consequently
multiple VACNFs grow from a single evaporated dot. For dot sizes smaller than the critical size
only one droplet is formed, resulting in a single VACNF. In the case of a patterned line, the growth
mechanism is similar to that from a dot. VACNFs grow along the line, and above a critical linewidth
multiple VACNFs are produced across the line. The mechanism of the formation of single and
multiple catalyst droplets and subsequently of VACNFs is discussed. © 2000 American Institute
of Physics. ͓S0003-6951͑00͒00824-X͔
Carbon nanotubes1 ͑CNTs͒ are of great interest to the
scientific and industrial communities due to their astounding
properties2 which include ballistic electron transport and
high mechanical strength and flexibility. CNTs are much
smaller than most of the vapor-grown carbon fibers ͑VGCFs͒
that have been studied for many years but the smallest
VGCFs often closely resemble CNTs in their external mor-
phology. CNTs and VGCFs also have been observed to grow
concurrently.3 However, CNTs are clearly distinguished
from VGCFs by the fact that a perfect CNT is a single crys-
talline grain, and even a defective CNT is a single crystal in
cross section, except where the defect occurs. In contrast, a
VGCF has a grain size ͑or structural coherence length4͒ that
is small compared with its circumference. This distinction is
crucial because it is the single crystal nature of CNTs that is
responsible for their one-dimensional quantum effects, and
which permits the use of a Brillouin zone-folding approach
to obtain their electron and phonon dispersion relations.2
In addition to their fundamental scientific importance,
CNTs have numerous potential applications in nanoelectron-
ics, nanoscale structural materials, hydrogen storage, and
field-emission devices, among others. Unfortunately for elec-
trical applications and integration with microelectronics, the
structurally nearly perfect CNTs are prepared by a variety of
methods including laser ablation,5 arc discharge,6 and chemi-
cal vapor deposition7 ͑CVD͒, all of which involve high syn-
thesis temperatures and produce randomly oriented CNTs
that must be laboriously separated, purified, and cut to length
before use. However, the CVD method has been used to
produce ‘‘carpets’’ of vertically aligned ͑perpendicular to the
substrate surface͒ multiwalled nanotubes and nonfibers,8–11
although control of their placement was limited. Recently,
patterned growth of individual, vertically aligned carbon fi-
bers also was demonstrated by Ren et al. using a plasma-
enhanced CVD ͑PECVD͒ method at a low substrate tem-
perature estimated to be less than 660 °C.12 In this case,
arrays of nanoscale carbon fibers were grown from Ni dots,
one fiber per dot. However, the resulting vertically aligned
fibers are not single crystalline grains, so are more properly
called vertically aligned carbon nanofibers ͑VACNFs͒, not
CNTs. Their defective structure is due at least in part to the
low growth temperature. Nevertheless, the PECVD growth
of VACNFs is of particular interest for possible applications
in field-emission arrays and scanning-probe microscopy.
There are several important issues about PECVD growth
of VACNFs that remain to be addressed. First, the maximum
size of a catalyst dot that produces only one VACNF has not
been determined. ͑It is known that for a continuous Ni film a
forest of VACNFs is formed.8͒ Second, growth of VACNFs
on a continuous line of catalyst rather than a dot has not been
explored. And, finally, the mechanism and conditions for the
formation of single and multiple VACNFs on an evaporated
catalyst dot or line have not been explained. In this work we
studied various stages of VACNF growth on patterned Ni
dots and lines in order to address all of these issues.
For VACNF preparation, n-type ͑001͒-oriented Si wa-
fers were used as substrates and Ni was employed as the
catalyst. The catalyst patterns ͑dots and lines of various
sizes͒ were fabricated by conventional electron beam lithog-
raphy using a bilayer resist in conjunction with electron-gun
metal evaporation. Following e-beam exposure and pattern
development, a buffer layer consisting of 10 nm of Ti was
deposited first to prevent the formation of Ni silicide at el-
evated temperatures and then a 15 nm thick Ni catalyst layer
was deposited.
VACNFs were prepared by PECVD in a vacuum cham-
ber evacuated by a mechanical pump to a base pressure
Ͻ5ϫ10Ϫ3 Torr. Acetylene (C2H2) and a mixture of 10%
a͒
Electronic mail: merkulov@solid.ssd.ornl.gov
0003-6951/2000/76(24)/3555/3/$17.00
3555
© 2000 American Institute of Physics
Downloaded 22 Aug 2013 to 128.248.155.225. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions

3556
Appl. Phys. Lett., Vol. 76, No. 24, 12 June 2000
Merkulov et al.
FIG. 1. High-resolution TEM image of VACNFs grown by PECVD at
700 °C. The VACNFs have a highly disordered ‘‘bamboo’’ structure ͑see
Ref. 3͒. The Ni catalyst particles ͑dark͒ are seen at the nanotube tips.
ammonia (NH3) and 90% helium ͑He͒, with gas flows of 15
and 200 sccm, respectively, were used as the gas source.
NH3 is needed to etch away graphitic C film that continu-
ously forms during the growth from the plasma discharge
and passives the Ni catalyst, thereby preventing the forma-
tion of nanofibers. Then Ni was pre-etched by the NH3 /He
plasma for ϳ2 min to remove oxide, after which the acety-
lene flow was introduced. The total pressure during VACNF
growth was ϳ6 Torr, the dc plasma discharge was operated
at 50 mA and ϳ500 V, and the growth rate was ϳ150 nm/
min. Changing the pressure by a factor of 2–3 did not sig-
nificantly affect the growth. In contrast to hot filament
PECVD,8 the substrates were heated directly by placing them
on a heater plate. This method has the advantage of enabling
large-area deposition and excellent uniformity of VACNF
films. The substrate temperature during growth was
ϳ700 °C, as measured by a thermocouple. Growth at lower
temperatures down to 600 °C was possible, although the
growth rate in this case was substantially lower.
The formation of VACNFs was studied using a Philips
XL30/FEG high resolution scanning electron microscope
͑HRSEM͒ with x-ray energy dispersive spectroscopy ͑EDS͒
capability. The structure of VACNFs was investigated using
a Hitachi HF-2000 transmission electron microscope ͑TEM͒
equipped with a field emission gun. The VACNFs are quite
disordered and have a bamboo-like structure, as shown in
Fig. 1. This high degree of structural disorder is typical for
PECVD growth at relatively low temperatures.8
Scanning electron microscopy ͑SEM͒ images presented
in Fig. 2 demonstrate the combined effects of pre-etching
and heating up to ϳ700 °C on the Ni/Ti dots deposited on Si.
While the Ti layer continues to adhere to Si, the initially
continuous Ni layer breaks into little droplets. This is con-
sistent with the observations by Yudasaka et al.13 For ini-
tially large dots, multiple droplets are formed ͓Fig. 2͑a͔͒. As
the size of the dots is reduced, the number of Ni droplets also
decreases ͓Fig. 2͑b͔͒. Finally, for a dot size less than ϳ350
nm only a single Ni droplet is formed ͓Fig. 2͑c͔͒. These
droplets are the necessary precursors for the catalytic growth
of nanofibers. After the acetylene is introduced, the VACNF
growth begins. The mechanism of the vertically oriented
growth is not completely understood at this time. However,
‘‘crowding’’ of the nanofibers ͑suggested by others9,11 as a
mechanism for vertical alignment of nanotube ‘‘forests’’͒
clearly can be eliminated as an alignment mechanism in this
case, since regular arrays of single VACNFs were grown at a
FIG. 2. Formation of multiple ͑a͒, ͑b͒ and single ͑c͒ Ni droplets on the
patterned evaporated dots ͑15 nm Ni/10 nm Ti on Si͒ and subsequent growth
of multiple ͑d͒, ͑e͒ and single ͑f͒ VACNFs. The Ni droplets were formed
during NH3 /He plasma pre-etching and heating up to 600–700 °C. The
SEM images were taken at 15 kV and 0° tilt ͑a͒–͑c͒ and at 50° tilt ͑d͒–͑f͒.
5 ␮m pitch. We believe instead that preferential plasma etch-
ing in the direction parallel to the substrate surface and per-
haps the alignment of nanofiber growth along the plasma
electric-field direction are responsible for the vertical orien-
tation of CNFs.
Figures 2͑d͒–2͑f͒ show VACNFs grown from the pat-
terned Ni dots of various sizes. Once can see that for a large
dot that produced multiple Ni droplets multiple VACNFs are
grown ͓Fig. 2͑d͔͒, and the number of VACNFs decreases
with decreasing dot size, corresponding to fewer Ni droplets
͓Fig. 2͑e͔͒. Each Ni droplet initiates the growth of a single
VACNF, and then resides on top of the nanofiber and pro-
vides for its continued catalytic growth.14 For evaporated
dots smaller than ϳ350 nm in diameter only a single Ni
droplet is formed and only a single VACNF is grown ͓Fig.
2͑f͔͒. Thus we estimate the critical dot size to be ϳ350 nm.
Of course, it has to be realized that the critical size will
depend upon several parameters, including the choice of the
buffer layer, the substrate, and the type and thickness of the
catalyst used ͑15 nm of Ni here͒. Nevertheless, the mecha-
nism making possible the deterministic growth of a single
VACNF from a single patterned Ni dot will remain the same.
When patterned lines are used instead of dots, the
VACNFs grow along the lines, as illustrated in Fig. 3͑a͒.
There may be a single or multiple VACNFs across the line,
depending upon the linewidth. The critical width is found to
be ϳ200 nm under our preparation conditions. The VACNF
formation mechanism is very similar to that in the case of
evaporated dots. The patterned Ni line breaks into individual
droplets, each of which catalyzes the growth of VACNFs.
After the growth starts, the Ni particles reside on top of the
Downloaded 22 Aug 2013 to 128.248.155.225. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions

Appl. Phys. Lett., Vol. 76, No. 24, 12 June 2000
Merkulov et al.
3557
lets and subsequent catalytic growth of one VACNF per
droplet. For an initial 15 nm Ni/10 nm Ti dot on Si, the
critical diameter below which only a single VACNF will
grow is ϳ350 nm. Above this critical diameter multiple
VACNFs are formed. VACNF growth along patterned cata-
lyst lines also was demonstrated, and the growth mechanism
is found to be similar, with a critical linewidth of ϳ200 nm
for the same catalyst deposition. We also found that direct
substrate heating during PECVD growth has a significant
advantage in producing large-area growth of VACNFs with
excellent uniformity.
The authors would like to acknowledge and thank I. A.
Merkulov for a stimulating discussion regarding the physical
mechanisms of the cluster formation. They also would like to
thank the SHaRE facility and E. A. Kenik for providing the
XL30/FEG microscope used for the SEM measurements, X.
Fan for preliminary TEM measurements, and P. H. Fleming
for assistance with sample preparation. This research was
partially sponsored by the Defense Advanced Research
Projects Agency under Contract No. DARPA-MIPR-97-
1357 with Oak Ridge National Laboratory ͑ORNL͒ and by
the Office of Basic Energy Sciences, Division of Materials
Sciences, U.S. Department of Energy. The research also was
supported in part by an appointment to the Oak Ridge Na-
tional Laboratory Postdoctoral Research Associates Program
administered jointly by ORNL and the Oak Ridge Institute
for Science and Education and was carried out at ORNL,
managed by Lockheed Martin Energy Research Corp. for the
U.S. Department of Energy, under Contract No. DE-AC05-
96OR22464.
FIG. 3. SEM images taken at 50° tilt ͑a͒ and 0° tilt ͑b͒ showing VACNFs
grown on Ni catalyst lines. ͑a͒ VACNF growth follows the catalyst line
pattern. ͑b͒ The Ni film breaks into little droplets that initiate CNF formation
and then reside on top of VACNFs that grow perpendicular to the substrate.
nanofibers, providing for their continued growth ͓Fig. 3͑b͔͒.
The formation of single or multiple catalyst droplets on a
single patterned catalyst dot can be explained as follows.
As-deposited Ni films have intrinsic tensile stress of ϳ8
ϫ109 dynes/cm2.15 However, upon heating a compressive
stress develops in the films16 due to the different expansion
coefficients for Ni (13ϫ10Ϫ6/°C) and Si (3ϫ10Ϫ6/°C). At
600–700 °C this thermal compressive stress is equal to ϳ2
ϫ1010 dynes/cm2 and thus the net stress will be compressive.
Furthermore, the compressively strained Ni film will develop
propagating sinusoidal wrinkles to minimize the strain
energy.17 This will tend to break large-area films into smaller
patches. However, the wrinkle formation and the subsequent
film breakup increase the surface energy. Consequently, be-
low a critical dot size the breakup does not occur because the
increase in the surface energy would be larger than the re-
duction of the strain energy. Next, at elevated temperatures
the metastable Ni film patches will tend to form clusters to
further reduce the strain energy.18 This results in the
Stranski–Krastanov growth mode: Ni clusters are formed on
top of one or a few monolayers of Ni attached to the sub-
strate surface. The characteristic cluster size will be dictated
by the minimum of the sum of the strain and surface ener-
gies. In the end, large-dot Ni films that break into multiple
patches produce multiple Ni clusters, whereas Ni films on
smaller dots below a critical size form only a single droplet.
The Ti layer, on the other hand, does not form clusters due to
its lower (8.5ϫ10Ϫ6/°C) expansion coefficient ͑lower strain
energy͒ and higher melting point ͑lower mobility of the at-
oms at a given temperature͒.
1 S. Iijima Nature ͑London͒ 354, 56 ͑1991͒.
2 See, for example, R. Saito, G. Dresselhaus, and M. S. Dresselhaus, Physi-
cal Properties of Carbon Nanotubes ͑Imperial College Press, London,
1998͒.
3 M. Endo, R. Saito, M. S. Dresselhaus, and G. Dresselhaus, in Carbon
Nanotubes Preparation and Properties, edited by T. W. Ebbesen ͑Chemi-
cal Rubber, Boca Raton, FL, 1997͒, p. 37.
4 In Ref. 3, p. 45.
5 T. Guo, P. Nikolaev, A. Thess, D. T. Colbert, and R. E. Smalley, Chem.
Phys. Lett. 243, 49 ͑1995͒.
6 C. Journet, W. K. Maser, P. Bernier, A. Loiseau, M. L. Chapelle, S.
Lefrant, P. Deniard, R. Lee, and J. E. Fisher, Nature ͑London͒ 388, 756
͑1997͒.
7 V. Ivanov, J. B. Nagy, Ph. Lambin, A. Lucas, X. B. Zhang, X. F. Zhang,
D. Bernaerts, G. Van Tendeloo, S. Amelinckx, and J. Van Landuyt, Chem.
Phys. Lett. 223, 329 ͑1994͒.
8 Z. F. Ren, Z. P. Huang, J. W. Xu, J. H. Wang, P. Bush, M. P. Siegal, and
P. N. Provencio, Science 282, 1105 ͑1998͒.
9 S. Fan, M. G. Chapline, N. R. Franklin, T. W. Tombler, A. M. Cassell,
and H. Dai, Science 283, 512 ͑1999͒.
10 W. Z. Li, S. S. Xie, L. X. Qian, B. H. Chang, B. S. Zou, W. Y. Zhou, R.
A. Zhao, and G. Wang, Science 274, 1701 ͑1996͒.
11 R. Andrews, D. Jacques, A. M. Rao, F. Derbyshire, D. Qian, X. Fan, E. C.
Dickey, and J. Chen, Chem. Phys. Lett. 303, 467 ͑1999͒.
12 Z. F. Ren, Z. P. Huang, D. Z. Wang, J. G. Wen, J. W. Xu, J. H. Wang, L.
E. Calvet, J. Chen, J. F. Klemic, and M. A. Reed, Appl. Phys. Lett. 75,
1086 ͑1999͒.
13 M. Yudasaka, R. Kikuchi, Y. Ohki, E. Ota, and S. Yoshiumura, Appl.
Phys. Lett. 70, 1817 ͑1997͒.
14 R. T. K. Baker, Carbon 27, 315 ͑1989͒.
15 M. Ohring, The Materials Science of Thin Films ͑Academic, San Diego,
1992͒, p. 422.
In conclusion, the mechanism for PECVD growth of
vertically aligned carbon nanofibers from catalyst dot and
line patterns has been studied. Experiments reveal that the
VACNFs are produced due to the formation of catalyst drop-
16 In Ref. 15, p. 419.
17 K. M. Crosby and R. M. Bradley, Phys. Rev. E 59, R2542 ͑1999͒.
18 M. H. Grabow and G. H. Gilmer, Surf. Sci. 194, 333 ͑1988͒.
Downloaded 22 Aug 2013 to 128.248.155.225. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions