Full text of DOI:10.1063/1.114613

Specific conditions for Ni catalyzed carbon nanotube growth by chemical
vapor deposition
Masako Yudasaka, Rie Kikuchi, Takeo Matsui, Yoshimasa Ohki, Susumu Yoshimura et al.
Citation: Appl. Phys. Lett. 67, 2477 (1995); doi: 10.1063/1.114613
View online: http://dx.doi.org/10.1063/1.114613
View Table of Contents: http://apl.aip.org/resource/1/APPLAB/v67/i17
Published by the American Institute of Physics.
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 15 May 2013 to 128.206.9.138. 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

Specific conditions for Ni catalyzed carbon nanotube growth by chemical
vapor deposition
Masako Yudasaka, Rie Kikuchi, Takeo Matsui, Yoshimasa Ohki, and Susumu Yoshimura
Yoshimura Pi-Electron Materials Project, ERATO, JRDC, c/o Matsushita Research Institute Tokyo,
Inc., 3-10-1 Higashimita, Tama-ku, Kawasaki 214, Japan
Etsuro Ota
Department of Chemistry, Faculty of Engineering, Gunma University, Kiryu, Gunma 376, Japan
͑Received 15 June 1995; accepted for publication 18 August 1995͒
Chemical vapor deposition using 2-methyl-1,2Ј-naphthyl ketone as a starting material has been done
between 1000 and 600 °C on Ni particles with diameters ranging from 10 to 500 nm. The Ni
particles were prepared by annealing Ni thin film deposited on quartz glass substrates. The size of
the Ni particle was controlled by the thickness of the Ni film. Carbon nanotubes were obtained at
700 °C when the diameter of the Ni particles was about 20–30 nm. © 1995 American Institute of
Physics.
A carbon nanotube comprises, according to Ebbesen,1
cylindrical shells of graphitic sheets. The length of the nano-
tube is greater than 1 ␮m and the aspect ratio ͑length over
diameter͒ is greater than 100. The interlayer distance of the
sheets is about 0.34 nm. Carbon nanotubes have been formed
in the arc plasma with the subsequent purification for sepa-
rating the graphite tubes from the carbon soot.2 The forma-
tion mechanism in the arc has been studied but is still open
to question.3 While much work has been devoted to arc-
discharge-based nanotubes, metal-catalyzed growth of nano-
tubes has been recognized for some time.5,6 Here we report
the formation of multilayer graphite nanotubes by chemical
vapor deposition ͑CVD͒ under a very specific reaction con-
dition. The specific condition includes the temperature of
700 °C and the size of a Ni catalyst with a round shape of 20
to 30 nm in diameter, which is noted for the first time and
may help to clarify the growth mechanism. Precise knowl-
edge on roles of Ni in low-temperature graphitization found
by the authors7,8 is also useful.
Ni films with various thicknesses ͑1–100 nm͒ were
vacuum-deposited (10Ϫ6 Torr) on quartz–glass substrates.
Ni particles were formed by heat treating the Ni film under
vacuum (10Ϫ7 Torr). Immediately after the heat treatment,
carbon was deposited on the Ni particles by CVD without
breaking the vacuum. The starting material for vacuum CVD
was 2-methyl-1,2Ј naphthyl ketone which has been found to
be appropriate for graphite film formation on Ni at tempera-
tures above 600 °C.5–7 The vacuum CVD and the detailed
deposition conditions were described previously.7
age of Fig. 3͑a͒ shows straight carbon nanotubes with uni-
form diameter. The outside and inside diameters of the car-
bon nanotubes are 17–27 nm and 10–17 nm, respectively
͓Figs. 3͑a͒ and 3͑b͔͒. About 10 sheets of graphite layers can
be clearly seen in Fig. 3͑b͒. A selected area diffraction pat-
tern of the carbon nanotube is shown in Fig. 4. The diffrac-
tion spots ͑Fig. 4͒, corresponding to a layer distance of 0.34
and 0.17 nm, are considered to be of the graphite ͑002͒ and
͑004͒ diffractions, respectively. The graphite ͗001͘ is known
to be perpendicular to the tube axis. In Fig. 4 can also be
seen halolines with spots corresponding to layer distances of
0.21 and 0.12 nm. The former corresponds to ͑100͒ and the
latter to ͑110͒ of graphite. This means that ͗100͘ and ͗110͘ of
the graphite tube is parallel with the tube axis.
There is evidence of materials appearing as black shad-
ows in Fig. 3, not only at the top end but also in the middle
of the carbon nanotubes. They were found to be Ni by energy
dispersive spectrometry. Many of the Ni particles take a cy-
lindrical shape with diameter of about 10–13 nm and length
of 40–100 nm, whose volume is 4500–11 000 nm3, which
corresponds to the volume of a sphere with diameters rang-
ing from 20–28 nm. There are many Ni droplets with diam-
eters around 20–30 nm in the scanning electron microscope
͑SEM͒ image for the heat-treated Ni film with an initial
thickness of 5 nm ͓Fig. 1͑a͔͒. The Ni cylinder was found to
be crystallized because the electron diffraction spots corre-
sponding to ͑111͒ and ͑100͒ of Ni could be observed. Be-
sides the carbon nanotubes, graphite particles with Ni par-
ticles inside can also be seen in the TEM images of Fig. 3͑c͒.
The diameter of the Ni particle is larger than 50 nm and the
thickness of graphite is less than 10 nm. When the diameter
of the Ni particle is above 50 nm, the graphite prefers to take
a round particle shape.
The Ni films with a thickness of 5 nm were transformed
into Ni droplets by heat treatment at 700 °C for 2 h, whose
diameter was 10–250 nm ͓Fig. 1͑a͔͒. Long carbon tubes with
length more than several ␮m and the carbon particles with
round shape were obtained by CVD at 700 °C on the Ni
droplets ͓Fig. 1͑b͔͒. It was confirmed that the carbon nano-
tubes and particles were composed of graphite with small
basal domain dimensions, using a Raman scattering spec-
trum ͓Fig. 2͑b͔͒ that exhibited two peaks at 1585 and
1350 cmϪ1.9 The transmission electron microscopy ͑TEM͒
images of the nanotubes are shown in Fig. 3. The TEM im-
The above results indicate that the carbon nanotubes are
able to grow from the Ni droplets with specific diameter,
which is confirmed in the experiments shown below where
the Ni droplet size is changed. After the Ni film with the
initial thickness of 1 nm ͓written as Ni ͑1 nm͒ film͔ was heat
treated at 700 °C, its SEM image did not show any Ni drop-
lets. The carbon deposit obtained by CVD on the heat treated
Appl. Phys. Lett. 67 (17), 23 October 1995
0003-6951/95/67(17)/2477/3/$6.00
© 1995 American Institute of Physics
2477
Downloaded 15 May 2013 to 128.206.9.138. 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

FIG. 1. Scanning electron micrographs of ͑a͒ heat-treated Ni film with
thickness of 5 nm and ͑b͒ deposit obtained by CVD on Ni droplets appear-
ing in ͑a͒. Temperature of heat treatment and CVD was 700 °C. Gold film
with thickness about 10 nm was deposited for SEM observation.
FIG. 2. Raman spectra of deposits obtained by CVD on heat-treated Ni
films with initial thicknesses of ͑a͒ 1, ͑b͒ 5, ͑c͒ 10, and ͑d͒ 100 nm.
Ni ͑1 nm͒ film ͓written as C/Ni ͑1 nm͔͒ was amorphous
carbon, which is apparent from the Raman spectrum showing
broad peaks at around 1600 and 1350 cmϪ1 ͓Fig.2͑a͔͒.9
When the Ni ͑10 nm͒ film is used, the sizes of the Ni drop-
lets generated by the heat treatment were in the range of
10–500 nm. The number of Ni droplets with diameter
around 20–28 nm are much smaller than that observed in the
case of Ni ͑5 nm͒. The C/Ni ͑10 nm͒ achieves a graphite
structure as is apparent from the Raman spectrum of Fig.
2͑c͒. SEM images of the C/Ni ͑10 nm͒ showed only a few
carbon tubes among many carbon particles. The SEM images
of the heat-treated Ni ͑100 nm͒ film and C/Ni ͑100 nm͒
showed a continuous structure. The Raman spectrum of the
C/Ni ͑100 nm͒ ͓Fig. 2͑d͔͒ is characteristic of graphite.9 It is
concluded that the size of the Ni droplet suitable for carbon
nanotube growth is limited to around 20–30 nm.
Furthermore, the optimum temperature range for the car-
bon nanotube formation was investigated. At 1000 °C, no
carbon nanotube was generated even though various thick-
nesses of Ni films were applied. The SEM images of graphite
FIG. 3. Transmission electron micrographs of ͑a͒ carbon nanotubes and particles, ͑b͒ carbon nanotubes, and ͑c͒ carbon particles grown by CVD at 700 °C
using Ni thin film with thickness of 5 nm. The thick network is a microgrid made of organic material. Large dark islands are broken pieces of quartz glass.
2478
Appl. Phys. Lett., Vol. 67, No. 17, 23 October 1995
Yudasaka et al.
Downloaded 15 May 2013 to 128.206.9.138. 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

FIG. 5. Transmission electron micrographs of ͑a͒ carbon nanotubes and
particles and ͑b͒ carbon tubes grown by CVD at 800 °C using an Ni thin
film with thickness of 5 nm. The thick network is a microgrid made of
organic material. Transparent carbon sheets are considered to be amorphous
carbon.
FIG. 4. Selected-area diffraction pattern of carbon nanotube inserted in Fig.
1͑b͒.
deposited on the Ni films at 800 °C showed a few carbon
tubes among round graphite particles. The carbon nanotubes
were featured by an outside diameter of 17–27 nm, a length
of less than 1 ␮m, and a bending structure. The diameters of
the tube are not constant along the tube ͑Fig. 5͒. At 600 °C,
the Ni film could not be transformed into round droplets by
the heat treatment.
It is concluded that the optimum temperature and the
diameter of the Ni droplet for the carbon nanotube formation
are 700 °C and 20–30 nm, respectively. The temperature of
700 °C is understood to be significant for the graphite thin
film formation on Ni by CVD; Ni can assist the graphite
formation via dehydrogenation of the starting material used
in this study.8 Since Ni and carbon do not make a solid
solution at 700 °C,8 the Ni cylinder and the carbon nanotube
can exist independently. The temperature of 700 °C is suffi-
ciently high for Ni to change its shape from thin films to
droplets and then to cylinders. At temperatures above
800 °C, carbon and Ni dissolve into each other, leading to
recrystallization of carbon as graphite. This phenomenon is
suitable for the formation of graphite thin films by CVD8 but
is not suitable for the nanotube formation. The interdissolu-
tion possibly prohibits the formation of long straight carbon
nanotubes with uniform diameter.
stress of the graphite exceeds the elastic limit. This causes
the round shape of the graphite and the Ni droplets to con-
currently transform into the carbon nanotube and the Ni cyl-
inders, respectively. One end of the cylinder must be rooted
on the substrate surface, which was actually recognized
through the SEM observation ͓Fig. 1͑b͔͒. During the growth
of the carbon nanotube by CVD at 700 °C, the Ni cylinder is
likely to exist at the end of the carbon nanotube so that
additional stability can be gained by contact with newly
forming nanotube fragments at the nanotube edge.5 The car-
bon nanotube may be closed when the temperature is suffi-
ciently lowered, causing five-membered carbon rings to be-
come sufficiently stable to cap the nanotube.10
We thank T. Kohzaki of Matsushita Technoresearch, Inc.
for his measuring the TEM, the energy dispersive spectra,
and the selected-area diffraction.
1 T. W. Ebbesen, Annu. Rev. Mater. Sci. 24, 235 ͑1994͒.
2 M. Endo, K. Takeuchi, S. Igarashi, K. Kobori, M. Shiraishi, and H. Kroto,
J. Phys. Chem. Solids 54, 1841 ͑1993͒.
3
´
´
´
M. Jose-Yacaman, M. Miki-Yoshida, and L. Rendon, Appl. Phys. Lett. 62,
657 ͑1993͒.
4
´
C. Gurret-Plecourt, Y. Le Bouar, A. Loiseau, and H. Pascard, Nature 372,
Finally, we give a picture for the mechanism of the car-
bon nanotube formation and explain the specific sizes of the
Ni droplets and unique sizes of the carbon nanotubes and the
Ni cylinders. First, the Ni film with a thickness of 5 nm
transforms into the Ni droplets with 10–250 nm in diameter
by heat treatment. Then, as the CVD begins, the graphite
starts to cover the Ni droplets. When the Ni droplets are
20–28 nm in diameter and the number of graphite sheets
covering the Ni droplets reaches around 10, the mechanical
761 ͑1994͒.
5 A. Oberlin, M. Endo, and T. Koyama, J. Cryst. Growth 32, 335 ͑1976͒.
6 M. Yudasaka, R. Kikuchi, T. Matsui, H. Kamo, Y. Ohki, E. Ota, and S.
Yoshimura, Appl. Phys. Lett. 64, 842 ͑1994͒.
7 M. Yudasaka, R. Kikuchi, T. Matsui, Y. Ohki, E. Ota, and S. Yoshimura,
Appl. Phys. Lett. 65, 46 ͑1994͒.
8 M. Yudasaka, R. Kikuchi, T. Matsui, Y. Ohki, E. Ota, and S. Yoshimura, J.
Vac. Sci. Technol. A 13, 2142 ͑1995͒.
9 K. Kinoshita, Carbon, Electrical and Physicochemical Properties ͑Wiley,
New York, 1988͒.
10 M. Endo and H. Kroto, J. Phys. Chem. 96, 6941 ͑1992͒.
Appl. Phys. Lett., Vol. 67, No. 17, 23 October 1995
Yudasaka et al.
2479
Downloaded 15 May 2013 to 128.206.9.138. 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