Growth of carbon nanotubes on cobalt disilicide precipitates by chemical vapor deposition

Article abstract of DOI:10.1063/1.121629

We have successfully grown carbon nanotubes on cobalt-implanted silicon with various doses. The morphology of such tubes has been examined by scanning electron microscopy, transmission electron microscopy, and Raman scattering. On contrary to the commonly

Full text of DOI:10.1063/1.121629

Growth of carbon nanotubes on cobalt disilicide precipitates by chemical vapor
deposition
J. M. Mao, L. F. Sun, L. X. Qian, Z. W. Pan, B. H. Chang, W. Y. Zhou, G. Wang, and S. S. Xie
Citation: Applied Physics Letters 72, 3297 (1998); doi: 10.1063/1.121629
View online: http://dx.doi.org/10.1063/1.121629
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/72/25?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Carbon impurities on graphene synthesized by chemical vapor deposition on platinum
J. Appl. Phys. 116, 044303 (2014); 10.1063/1.4891200
Initial growth process of carbon nanowalls synthesized by radical injection plasma-enhanced chemical vapor
deposition
J. Appl. Phys. 106, 094302 (2009); 10.1063/1.3253734
Effect of ZnO catalyst on carbon nanotube growth by thermal chemical vapor deposition
J. Vac. Sci. Technol. B 26, 1765 (2008); 10.1116/1.2981084
Synthesis of high yield single helical carbon microsprings by catalytic chemical vapor deposition and an
experimental investigation of their growth mechanism
J. Appl. Phys. 101, 114903 (2007); 10.1063/1.2737381
Nucleation and growth of carbon nanotubes by microwave plasma chemical vapor deposition
Appl. Phys. Lett. 77, 2767 (2000); 10.1063/1.1319529
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
132.174.254.155 On: Tue, 23 Dec 2014 01:03:08

APPLIED PHYSICS LETTERS
VOLUME 72, NUMBER 25
22 JUNE 1998
Growth of carbon nanotubes on cobalt disilicide precipitates by chemical
vapor deposition
J. M. Mao,a) L. F. Sun, L. X. Qian,b) Z. W. Pan, B. H. Chang, W. Y. Zhou, G. Wang,
and S. S. Xie
Institute of Physics, Center for Condensed Matter Physics, Chinese Academy of Sciences,
Beijing 100080, People’s Republic of China
͑
Received 2 February 1998; accepted for publication 22 April 1998͒
We have successfully grown carbon nanotubes on cobalt-implanted silicon with various doses. The
morphology of such tubes has been examined by scanning electron microscopy, transmission
electron microscopy, and Raman scattering. On contrary to the commonly used transition-metal
nanoparticle catalysts, nanometer-sized CoSi precipitates produced in the as-implanted substrates
2
are believed to act as nucleation centers for the formation of carbon nanotubes. © 1998 American
Institute of Physics. ͓S0003-6951͑98͒03625-0͔
Since its observation in 1991,1 the carbon nanotube
the literature.11 Previous transmission electron microscopy
͑
CNT͒ has been widely investigated both for its novel elec-
͑TEM͒ and high-resolution transmission electron microscopy
2
–5
tronic properties and application potentials.
Ideally, a
͑HRTEM͒ studies have revealed that two types of CoSi pre-
2
nanotube consists of concentric cylindrical shells of graphitic
sheets, which can be viewed as rolling up the graphitic layer
into a tubule. The nanotube can be a metallic conductor or
semiconductor, which is uniquely determined by a pair of
cipitates are formed in Co-implanted silicon, which are de-
noted as A type and B type. Generally, no purely Co par-
ticles were found in the implanted samples.11 These
precipitates distribute from the sample surface down to about
6
7
2
00 nm in the bulk. The density of the precipitates shows a
integer indices (n,m). Both theoretical predictions and ex-
8
Gaussian distribution, which is correlated with the Co depth
profile. It is expected that only the surface precipitates ͑part
of the precipitate is above the silicon surface͒ can act as the
nucleation centers for the growth of carbon nanotubes. The
size of the precipitates above the surface is nearly the same
as that of the precipitate itself,11 and the size of the precipi-
tates increases with the implantation dose, e.g., 15 nm for a
sample implanted with a dose of 1ϫ1017 cm . The as-
implanted samples have been investigated by conducting
atomic force microscopy, i.e., simultaneously monitoring the
surface morphology and current image. The position of the
surface precipitates and their distribution can be obtained
from the current image.12
perimental results show that it is possible for a tube to show
different conducting properties at different parts. Thus, nan-
odevices based on a tube can be realized. Nowadays, large
quantity carbon nanotubes can be systhesized by arc dis-
charge, laser ablation, and chemical vapor deposition
6
͑
CVD͒. Among them, the latter method is a cheaper way,
and the reaction process can be easily controlled. Moreover,
the required deposition temperature by CVD is relatively
low, the commonly used reaction temperature is around
Ϫ2
700 °C. Using the CVD method, carbon nanotubes ͑CNTs͒
have been prepared by decomposition of hydrocarbon on
various substrates containing transition-metal nanoparticles9
or through decomposing the metal complex which contains
both metal and carbon.10 We report here a catalyst or nucle-
ation center other than a transition-metal catalyst for the sys-
thesis of carbon nanotubes by CVD, i.e., cobalt disilicide
formed in Co-implanted silicon. To our knowledge, there is
yet no report on carbon nanotube production based on such
substrates. On the other hand, growth of CNTs on silicon is
particularly useful for future integration of CNTs-based nan-
odevices since silicon is the headstone of the modern semi-
conductor industry.
15
15
The implantation doses used are 1ϫ10 , 5ϫ10 , 1
16
16
17
17
Ϫ2
ϫ10 , 5ϫ10 , 1ϫ10 , and 2ϫ10 cm , the corre-
sponding CoSi2 size ranges from several nanometers to 15
nm for the as-implanted samples. These samples were put
into a quartz boat, then transferred to a stainless-steel cham-
ber for reduction and growth. Reduction was performed at
500 °C with pressure about 150 Torr, a mixture of hydrogen
and nitrogen was used with flux rates 10 and 50 cc/min,
respectively. 10% acetylene in N2 with a total flowing rate of
110 cc/min was introduced for the growth at 650 °C in 150
Torr. The growth process continues from 2 to 5 h, and the
reduction time is about 1 h. To compare, some samples were
directly sent to the growth chamber without reduction.
Nanotubes from samples implanted with doses ranging
The Co implantation was performed with a metal vapor
vacuum arc ion source into ͑100͒ or ͑111͒ n-type silicon
wafers with resistivity of 10–20 ⍀ cm at an extraction volt-
age of 30 kV to a dose of 2ϫ1017 cm . The substrate
temperature during implantation was kept lower than 200 °C.
Details of the implantation process and characterization of
the as-implanted and postannealing samples can be found in
Ϫ2
from 1ϫ1015 to 2ϫ10 cm
17
Ϫ2
have been examined by
scanning electron microscopy ͑SEM͒ ͑Hitachi, S-4200͒. As
can be seen with the naked eye, the mirror-like silicon sur-
face was covered with a black layer after several hours
growth. Under SEM, these black layers were the nanotubes
with a diameter around 15 nm, no significant difference of
a͒
Electronic mail: jmmao@aphy01.iphy.ac.cn
Also with the Department of Physics, Central University of Nationalities,
b͒
Beijing 100081, People’s Republic of China.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
003-6951/98/72(25)/3297/3/$15.00 3297 © 1998 American Institute of Physics
32.174.254.155 On: Tue, 23 Dec 2014 01:03:08
0
1

3298
Appl. Phys. Lett., Vol. 72, No. 25, 22 June 1998
Mao et al.
FIG. 3. SEM image of carbon nanotubes from a sample grown for 3 h
without the reduction process. The Co implantation dose is
5
ϫ1016 cm
Ϫ2
.
FIG. 1. Low-magnification SEM image of the carbon nanotubes grown for
h, reduced for 1 h on cobalt-implanted silicon with a dose of 5
3
1
5
Ϫ2
ϫ10 cm
.
2
͑b͒. It can be seen that the purity of the CNTs is quite high,
no amorphous carbon and graphitic debris were found. The
tubes tend to aggregate which is similar to that from the
the nanotube diameter was found for the samples implanted
with different doses. Figure 1 is the low magnification SEM
sample with a dose of 5ϫ101 cm . Since CoSi precipi-
5
Ϫ2
picture for the samples with a Co dose of 5ϫ1015 cmϪ2
2
tates are resistant to oxidation,12 we have grown CNTs with-
out prereduction. In Fig. 3, we present the SEM results of a
sample grown for 3 h without the reduction process, the Co
implantation dose is 5ϫ1016 cm . The nanotube yield is as
high as that with reduction. However, the quality is not as
good as that with prereduction. The average nanotube diam-
eter is also about 15 nm, but it has a broader distribution.
This can be considered as the following. Besides deoxida-
tion, the precipitates coarsen during reduction, the coarsen-
ing and coalescence effect is expected to be reduced at depo-
sition time. Therefore, the grown tubes’ diameter will show a
more uniform distribution. Generally, the coarsening and
coalescence effect accompanied the growth. The movement
and aggregation of the precipitates led to the coarsening and
clustering of the tubes. This process continued to the end of
the deposition, but it is stronger at the beginning of the
growth, especially for the sample without the reduction pro-
cess. Hence, CNTs grown with no prereduction are easy to
cluster and will show a nonuniform distribution of the tube
size.
grown for 3 h with 1 h reduction. The nanotubes are uni-
formly sprayed over the substrates and tend to form a tangled
structure. The yield of the nanotubes increases slowly with
Ϫ2
the implantation dose. CNTs grown on the sample for 2 h
with a dose of 5ϫ1016 cmϪ2 are depicted in Figs. 2͑a͒ and
The specimens for TEM ͑H-9000, NAR͒ analysis were
prepared by dissolving the deposit in ethanol. After ultra-
sonic treatment, a drop of the liquid was then sprayed to a
holey carbon copper grid. Well-defined graphitic fringes
were observed from the HRTEM images of the tubes grown
with the reduction process ͑not shown͒. The outer periphery
of the tube was coated by a layer of amorphous carbon,
which is the ordinary phenomenon observed in the tubes
produced by the catalyst thermal decomposition method,
where a relatively low temperature ͑about 700 °C͒ was used
during reaction. A typical HRTEM image of the tube grown
without prereduction is shown in Fig. 4, and graphitic fringes
can be seen too. However, the direction of the veins was
tilted with respect to the axis of the tube. This may be due to
the coarsening and coalescence effect during CNT growth as
mentioned above. Microarea Raman scattering spectrum of a
typical sample deposited with the reduction process is dis-
played in Fig. 5. The experiment was conducted in a Ren-
FIG. 2. Low-magnification ͑a͒ and high-magnification ͑b͒ SEM image of
the carbon nanotubes grown for 2 h, reduced for 1 h on cobalt-implanted
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
1
6
Ϫ2
silicon with a dose of 5ϫ10 cm
.
ishaw equipment. The spectrum of the deposit shows two
1
32.174.254.155 On: Tue, 23 Dec 2014 01:03:08

Appl. Phys. Lett., Vol. 72, No. 25, 22 June 1998
Mao et al.
3299
application, the growth of nanotubes on CoSi embedded in
2
silicon is particularly useful since CoSi is a suitable candi-
2
date for wide applications in microelectronic devices such as
contact materials, gate electrodes, and interconnects. It is in-
teresting to resolve the nanotubes grown from each type of
the precipitate since the A type is coherent while the B type
is semicoherent. This can be studied by TEM cross-section
investigation, and such work is under progress in our labo-
ratory.
This work was supported in part by the Presidential
Foundation of Chinese Academy of Sciences and the Na-
tional Natural Science Foundation of China.
1
S. Iijima, Nature ͑London͒ 354, 56 ͑1991͒.
T. W. Ebbesen and P. M. Ajayan, Nature ͑London͒ 358, 220 ͑1992͒.
P. M. Ajayan, O. Stephan, C. Colliex, and D. Trauth, Science 265, 1212
2
FIG. 4. High-resolution TEM image of a carbon nanotube from the sample
deposited without the reduction process.
3
͑
1994͒.
4
W. A. de Heer, W. S. Bacsa, A. Chatelain, T. Gerfin, R. H. Baker, L.
Forro, and D. Ugarte, Science 268, 845 ͑1995͒.
H. Dai, E. W. Wong, and C. M. Lieber, Science 272, 523 ͑1996͒.
See, for example, Carbon Nanotubes: Preparation and Properties, edited
by Thomas W. Ebbesen ͑CRC, Boca Raton, FL, 1996͒.
Ph. Lambin, A. Fonseca, J. Vigneron, J. B. Nagy, and A. A. Lucas, Chem.
Phys. Lett. 245, 85 ͑1995͒; R. Saito, G. Dresselhaus, and M. S. Dressel-
haus, Phys. Rev. B 53, 2044 ͑1996͒; L. Chico, V. H. Crespi, L. X. Bene-
dict, S. G. Louie, and M. L. Cohen, Phys. Rev. Lett. 76, 971 ͑1996͒.
P. G. Collins, A. Zettl, H. Bando, A. Thess, and R. E. Smalley, Science
278, 100 ͑1997͒.
clear peaks at 1324 and 1587 cmϪ1. The relative intensity of
the two peaks and the narrower half width of them indicate a
good degree of sp to sp bonding.
In summary, carbon nanotubes with high quality have
been obtained on cobalt-implanted silicon by thermal decom-
5
6
2
3
6
7
position of hydrocarbon. This indicates that CoSi precipi-
2
tates embedded in a silicon matrix are a good catalyst for the
production of carbon nanotubes. For the point of view of
8
9
S. S. Xie, W. Z. Li, C. Y. Wang, L. W. Xu, H. Zhang, Y. Zhang, L. X.
Qian, Sci. China A 40, 971 ͑1997͒; M. Yudasaka, R. Kikuchi, T. Matsui,
Y. Ohki, S. Yoshimura, and E. Ota, Appl. Phys. Lett. 67, 2477 ͑1995͒; M.
Endo, K. Takeuchi, S. Igarashi, K. Kobori, M. Shirashi, and H. W. Kroto,
J. Phys. Chem. Solids 54, 1841 ͑1993͒; V. Ivanov, J. B. Nagy, Ph.
Lambin, A. Lucas, X. B. Zhang, X. F. Zhang, D. Bernaerts, G. van Ten-
deloo, S. Amelinckx, and J. van Landuyt, Chem. Phys. Lett. 223, 329
͑
1994͒; 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; K. Hernadi,
A. Fonesca, J. B. Nagy, D. Bernaerts, A. Fudala, and A. A. Lucas, Zeo-
lites 17, 416 ͑1996͒.
1
1
0
1
M. Yudasaka, R. Kikuchi, Y. Ohki, and S. Yoshimura, Carbon 35, 195
͑
1997͒.
C. D. Meekison, G. R. Booker, K. J. Reason, R. S. Spaggs, R. M. Gwil-
lian, and B. J. Sealy, J. Appl. Phys. 74, 7129 ͑1993͒; D. P. Adams, S. M.
Yalisove, and D. J. Eagleham, J. Appl. Phys. 76, 5190 ͑1994͒.
J. M. Mao, J. B. Xu, Q. C. Peng, S. P. Wong, and I. H. Wilson J. Mater.
Sci. Lett., 17, 219 ͑1998͒.
1
2
FIG. 5. Raman scattering of the carbon nanotubes grown on a sample with
1
5
Ϫ2
the reduction process, the implantation dose is 5ϫ10 cm
.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
32.174.254.155 On: Tue, 23 Dec 2014 01:03:08
1