Chemical Physics Letters 380 (2003) 496–502

www.elsevier.com/locate/cplett

A catalytic chemical vapor deposition synthesis

of double-walled carbon nanotubes over metal
catalysts supported on a mesoporous material
a,*
Jin Zhu , Masako Yudasaka a, Sumio Iijima a,b,c

a
Carbon Nanotube Project, Japan Science and Technology Corporation, c/o NEC, 34 Miyukigaoka, Tsukuba 305-8501, Japan
b
Department of Physics, Meijo University, Tenpaku-ku, Nagoya 468-8502, Japan
c
NEC Corporation, 34 Miyukigaoka, Tsukuba 305-8501, Japan
Received 4 July 2003; in final form 20 August 2003
Published online: 7 October 2003

Abstract

Double-walled carbon nanotubes (DWNTs) have been synthesized by catalytic chemical vapor deposition (CCVD)
over supported metal catalysts decomposed from Fe(CH3 COO)2 and Co(CH3 COO)2 on mesoporous silica. Bundles of
tubes with relatively high percentage of DWNTs, in areas where tubular layered structures could be clearly resolved,
have been observed by transmission electron microscopy (TEM). In other areas, crystal-like alignment of very uniform
DWNTs was observed for the first time, suggesting that mesoporous silica might play a templating role in guiding the
initial nanotube growth. In addition, compatible with nano-electronics research, bridging of catalytic islands by
DWNTs has also been demonstrated.
Ó 2003 Elsevier B.V. All rights reserved.

1. Introduction
Investigation into the fundamental properties
and applications of carbon nanotubes continues to
attract peopleÕs attention due to their huge struc-
tural variations and associated interesting chemical,
physical properties [1–6]. Major focus has been put
on single-walled carbon nanotubes (SWNTs) for
their structural simplicity [7,8] and synthetic avail-
ability [9–18]. Their synthetic methods reported so
*
Corresponding author. Fax: +81-29-850-1366.
E-mail address: jinzhu@nlp.jst.go.jp (J. Zhu).
far include pulsed laser oven [9], electric arc dis-
charge [10], and catalytic chemical vapor deposition
(CCVD) [11–18]. Among those, CCVD technique is
the most promising since it allows easy scale-up and
directed placement, alignment of nanotubes and
construction of electronic components [15–17].
With concentric arrangement of tubular layers,
multi-walled carbon nanotubes (MWNTs) offer
new structural and property possibilities through
interactions of graphitic carbon sheets. However, it
is difficult to pinpoint exact structures of every
graphitic layer with the current analytical know-
how. Double-walled carbon nanotubes (DWNTs)
constitute a unique class of materials, since they are

0009-2614/$ - see front matter Ó 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2003.09.049
 J. Zhu et al. / Chemical Physics Letters 380 (2003) 496–502
the simplest MWNTs and it is much easier to map
out their structure-function correlations [19]. More
importantly, a study of these simplest MWNT sys-
tems would provide new insight into the inter-layer
coupling properties of more complex structures.
The serious endeavor on the synthesis of DWNTs
was started only very recently and has been bene-
fited by the production methods accumulated on
SWNTs. Electric steady arc discharge [20,21] and
pulsed arc discharge [22] have been applied to the
generation of these materials. In addition, coales-
cence of one-dimensional crystal of C60 inside
SWNTs [23] was used to produce regular double-
layered tubes. On the other hand, CCVD technique
has just been started as an alternative method to
synthesize those structures and several successful
examples have been reported [24–28]. Herein we
report on our effort of producing DWNTs over
metal catalysts supported on a mesoporous mate-
rial. So far, no report on CCVD synthesis of
DWNTs has yet utilized mesoporous silica as a
support materials [24–28]. In view of the strong
substrate–catalyst interaction dependence of
nanotube growth [29], the present results obtained
in our laboratory will likely contribute to the ex-
pansion of DWNT synthetic scopes.
2. Experimental
The mesoporous material support utilized in
this Letter was prepared according to the literature
[30] at a smaller scale. Typically, with constant
stirring, 2.0 g of surfactant Brij 76 was dissolved in
10 g of H2 O. To this solution, 40 g of 2 M HCl was
added slowly. The solution was stirred for about
2 h until all the surfactant was dispersed well. Then
4.40 g of tetraethoxysilane was slowly dropped
into the solution and the homogeneous solution
was stirred for 20 h. The white powder was re-
covered through filtration, washed with copious
amount of H2 O and dried in air. The surfactant
was removed through calcination by heating to
500 °C in 8 h and kept at that temperature for 6 h.
The formation of mesophase P 63 =mmc (SBA-12)
was confirmed with TEM.
Solution method was used for the preparation
of metal catalysts supported on mesoporous ma-
497
terials. Typically, certain amount (varied from
1.5 to 6 mg; correspondingly weight percentage
changed from 1.5% to 6%) of Fe(CH3 COO)2 and
Co(CH3 COO)2
4H2 O was dissolved in 6 mL
CH3 CH2 OH, sonicated for 30 min, and this would
result in a homogeneous solution. Then 0.1 g of
SBA-12 was ground and added into the solution.
The mixture was heated at 100 °C under constant
stirring to remove the solvent until 1 mL dis-
persion was left (usually it takes 7 h). For every
experiment, 50 lL dispersion was cast onto a
quartz plate and the powder was left dried in the
air. Then the quartz plate with the dried powder
was put into a CCVD furnace, heated to a target
temperature and flowed through with the reaction
gas for a certain amount of time. A variety of
conditions have been tried for the synthesis, in-
cluding the pre-treatment heating atmosphere (Ar
or air), the heating rate (room temperature to
target temperature in 2–16 h), CCVD reaction gas
(CH4 vs. CH4 + H2 ), duration of the CCVD reac-
tion (10 min to 1 h).
The powder material after CCVD synthesis was
characterized with scanning electron microscopy
(SEM), TEM (the powder was sonicated and drop
coated onto a TEM grid from CH3 CH2 OH dis-
persion) and Raman spectroscopy (excitation
wavelength: 488 nm, excitation power: 50 mW).
3. Results and discussion
3.1. Experimental conditions for the synthesis of
DWNTs
As shown in other CCVD production of carbon
nanotubes, reaction conditions are the key to the
successful generation of these materials. One crit-
ical parameter for CCVD synthesis of carbon na-
notubes is the pre-treatment heating rate and
reaction temperatures. The production of DWNTs
was successful at the temperature range of 750–870
°C, independent of the pre-treatment heating rate.
When the temperature was between 800 and
870 °C, and the pre-treatment was done in air
(typically from room temperature to target tem-
perature in 2 h), the product included abundant
DWNTs and some other nano-scale carbonaceous
498 J. Zhu et al. / Chemical Physics Letters 380 (2003) 496–502

materials. For light brown or light grey powder
after the synthesis, nanotubes could be clearly
observed by SEM, Fig. 1a. The morphology of the
nanotubes seems to indicate that they are bundles
of either SWNTs or DWNTs since MWNTs would
usually have a lot of irregular kinks along the tube
direction. This conclusion was further supported
by lower magnification TEM observation, Fig. 1b.
Under TEM, spider-web-like appearance was al-
most always the signature morphology of SWNTs
or DWNTs. A close-up at higher magnification
reveals the presence of abundant DWNTs. When
the temperature was between 750 and 700 °C, short
MWNTs dominate the products. If the reaction

Fig. 1. (a) A SEM image of DWNTs synthesized on the mesoporous support (3% metal salts, 800 °C, 0.6 L/min CH4 and 0.4 L/min H2
for 30 min). (b) A lower magnification TEM image of DWNTs synthesized on the mesoporous support (3% metal salts, 800 °C, 0.6 L/
min CH4 and 0.4 L/min H2 for 30 min).
 J. Zhu et al. / Chemical Physics Letters 380 (2003) 496–502
temperature was further reduced to below 650 °C,
hardly any nanotubes could be observed by TEM.
If the temperature was above 900 °C and other
parameters were kept the same, no nanotubes could
be seen by TEM.
A second parameter for controlled synthesis of
carbon nanotubes is the pre-treatment atmosphere
and reaction gases. Ar was used as the pre-treat-
ment atmosphere for controlled decomposition of
metal salts in the initial set of experiments. Under
this condition, no matter what kind of reaction gas
mixtures or what temperature was used, the
powder surface was always covered with a layer of
dark black material after the synthesis. No na-
notubes were observed by TEM, indicating car-
bonization dominated the CCVD process and
amorphous carbon was formed afterwards. The
mesoporous support might play a significant role
in this particular CCVD reaction, since for this
medium, the synthesis was successful only when
the reaction conditions were distinctly different
from those of the other support. If the same
amount of metal salts was dispersed on a zeolite
support, even though the pre-treatment was done
in Ar and only CH4 was used as the reaction gas,
SWNTs would dominate the products. When the
pre-treatment was carried out in air, the use of
reaction gas mixture of H2 and CH4 was critical. If
no H2 was used, usually carbonization would oc-
cur and amorphous carbon would be formed.
After the pre-treatment of metal salts in air, if H2
was used to flow through and reduce the catalysts
for certain time (30 min was enough), no na-
notubes could be produced, suggesting that oxi-
dized metals are active components in the CCVD
process. The role of H2 in the gas mixtures is to
reduce the amorphous carbon formation and as-
sist in the production of DWNTs during the
CCVD process. On the other hand, Maruyama et
al. have observed that reduction of the catalytic
metal could enhance the yield of SWNTs synthe-
sized on metal catalysts over a different support
[31]. On other support materials, DaiÕs group ex-
tensively utilized oxidized Fe nanoparticles for the
synthesis of SWNTs [6,32,33]. Indeed, the role of
support material in the synthesis of carbon na-
notubes is a complex issue and remains to be
clarified [29].
499
The third control parameters are the composi-
tions of metal catalysts and the catalyst/support
ratios. When both Fe and Co were present in the
catalyst/support ratio range between 1.5% and 3%,
abundant DWNTs could be synthesized, as ob-
served by TEM. When the catalyst/support ratio
was 6%, most of the products were MWNTs. If the
synthesis was carried out with only Fe (3%),
DWNTs could be generated, but Co alone (3%)
would not produce any nanotubes. This indicates
that Fe plays a major role in the CCVD process.
The fact that the amount of DWNTs is larger
when both Fe and Co are present indicates that Co
might play a co-catalyst role in this process.
3.2. Characterization of DWNTs
Although the synthesis method in this Letter
has not yet generated pure DWNTs, there are
some phenomena worthy of notice and this might
point to the unique feature of this system and the
direction of possible further improvement in con-
trolled DWNT synthesis.
Bundles of relatively pure DWNTs could often
be observed under TEM. Fig. 2 shows a higher
magnification image of a particular area. In such a
place where tubular layered structures could be
clearly resolved, high-purity DWNTs with rela-
tively uniform diameters (outer diameter around
4.3 nm) could clearly be seen over a wide area,
suggesting that the current CCVD system might be
superior to those reported previously. To our best
estimate, the purity of the DWNTs in this system
exceeds 50%.
Another notable phenomenon in this system is
that regular, crystal-like structure made out of
DWNTs was observed for the first time to our
knowledge, as displayed in Fig. 3. In this area, the
DWNTs with very uniform outer diameters of
around 2.6 nm formed straight, aligned line pat-
terns. The formation mechanism of those struc-
tures remains to be clarified, but its nearly
crystalline nature suggests a possible templating
role of mesoporous silica in guiding the initial
alignment of the growing nanotubes.
In addition to TEM, micro-Raman spectros-
copy was used to characterize the products. For
some samples, a strong G-band (e.g. 1589.2 and
500 J. Zhu et al. / Chemical Physics Letters 380 (2003) 496–502

Fig. 2. A higher magnification TEM image of DWNTs with
outer diameters around 4.3 nm synthesized on the mesoporous
support (3% metal salts, 850 °C, 0.6 L/min CH4 and 0.4 L/min
H2 for 30 min).
Fig. 3. A higher magnification TEM image of crystal-like
alignment of DWNTs with very uniform outer diameters
around 2.6 nm synthesized on the mesoporous support (3%
metal salts, 800 °C, 1.0 L/min CH4 and 0.2 L/min H2 for
30 min).

1570.5 cm1 ) and a weak D-band (e.g. 1343.2

cm1 ) could always be observed at all the locations
in Fig. 4, indicating low defect and high quality of
our DWNTs. However, the Raman spectra in
the radial breathing mode (RBM) area (below
300 cm1 ) change from place to place. The spec-
trum of one particular area is shown in Fig. 4. The
relationship between the Raman RBM shift and
nanotube diameter is complicated by the interlayer
coupling of DWNTs, and the intensity and posi-
tion of the Raman peaks depend on the excitation
wavelength of the incident laser beam. This
drawback in Raman scattering means that it can-
not give a definite answer to the distribution of Fig. 4. Raman spectroscopy of DWNTs in a particular area
nanotube diameters, especially for larger DWNTs, synthesized on the mesoporous support (3% metal salts, 800 °C,
in which case RBM peaks may not even appear. 0.6 L/min CH4 and 0.4 L/min H2 for 30 min).
However, the appearance of both tangential and
RBM modes in the product is consistent with the
presence of abundant DWNTs observed by TEM. hara et al. in an unambiguous assignment of
When the DWNTs are uniform, Raman RBM DWNT Raman shifts [22]. As for our case of the
peaks could provide fruitful information regarding Raman spectrum of the specific sample spot shown
the tube diameters, as first established by Shino- Fig. 4, the identified diameters, 1.22, 1.36, 1.52 and
 J. Zhu et al. / Chemical Physics Letters 380 (2003) 496–502
1.69 nm (201.8, 181.2, 162.2 and 145.4 cm1 re-
spectively), based on d ¼ 246=x [34], where x is
the RBM Raman shift, might arise from the inner
tubes since the scattering from outer tubes might
not appear. These values fall in the diameter range
of the DWNTs observed by TEM, where the outer
and inner diameters are 1.5–5 and 0.8–4.2 nm re-
spectively, with inter-wall distances of 0.34–0.45
nm, to our best estimate.
3.3. Relevance to nano-electronics
Nanotube nano-electronics research has been
benefited by the CCVD method for its spatially
resolvable growth. Here we have demonstrated
that bridging of the islands by DWNTs is possible
if catalytic powder islands are separated by suit-
able distances. Fig. 5 shows that two isolated is-
Fig. 5. SEM images of two separate catalytic islands bridged by
(a) one and (b) two DWNT bundles on a Si surface (2% metal
salts, 800 °C, 0.6 L/min CH4 and 0.4 L/min H2 for 30 min).
501
lands deposited on a Si surface at micrometer scale
separations are connected by one or two DWNT
bundles. This is the advantage of CCVD method
over other alternatives, allowing easy integration
of nanotubes and direct construction of electronic
components [6].
4. Conclusions
DWNTs have been synthesized by the CCVD
method by flowing a gas mixture of CH4 and H2
over metal catalysts decomposed from Fe and Co
salts on a mesoporous support. Bridging of two
catalytic islands by DWNT bundles on a Si surface
has also been demonstrated. We hope that this
method will contribute to expansion of scopes for
the synthesis and the controlled deposition of
DWNTs on solid surfaces and to fundamental
studies on these simplest MWNTs.
References
[1] S. Iijima, Nature 354 (1991) 56.
[2] S. Iijima, T. Ichihashi, Nature 363 (1993) 603.
[3] S. Iijima, Physica B 323 (2002) 1.
[4] M. Ouyang, J. Huang, C.M. Lieber, Acc. Chem. Res. 35
(2002) 1018.
[5] H. Dai, J. Kong, C. Zhou, N. Franklin, T. Tombler, A.
Cassell, S. Fan, M. Chapline, J. Phys. Chem. B 103 (1999)
11246.
[6] H. Dai, Acc. Chem. Res. 35 (2002) 1035.
[7] J.W.G. Wildoer, L.C. Venema, A.G. Rinzler, R.E. Smal-
ley, C. Dekker, Nature 391 (1998) 59.
[8] T.W. Odom, J.L. Huang, P. Kim, C.M. Lieber, Nature 391
(1998) 62.
[9] A. Thess, R. Lee, P. Nikolaev, H. Dai, P. Petit, J. Robert,
C. Xu, Y.H. Lee, S.G. Kim, A.G. Rinzler, D.T. Colbert,
G.E. Scuseria, D. Tomanek, J.E. Fischer, R.E. Smalley,
Science 273 (1996) 483.
[10] C. Journet, W.K. Maser, P. Bernier, A. Loiseau, M.L. de la
Chapelle, S. Lefrant, P. Deniard, R. Lee, J.E. Fischer,
Nature 388 (1997) 756.
[11] S. Maruyama, R. Kojima, Y. Miyauchi, S. Chiashi, M.
Kohno, Chem. Phys. Lett. 360 (2002) 229.
[12] H. Dai, A.G. Rinzler, P. Nikolaev, A. Thess, D.T. Colbert,
R.E. Smalley, Chem. Phys. Lett. 260 (1996) 471.
[13] J. Kong, A.M. Kassell, H. Dai, Chem. Phys. Lett. 292
(1998) 567.
[14] M. Su, B. Zheng, J. Liu, Chem. Phys. Lett. 322 (2000)
321.
502 J. Zhu et al. / Chemical Physics Letters 380 (2003) 496–502
[15] E. Yenilmez, Q. Wang, R. Chen, D. Wang, H. Dai, Appl.
Phys. Lett. 80 (2002) 2225.
[16] Y. Zhang, A. Chang, J. Cao, Q. Wang, W. Kim, Y. Li, N.
Morris, E. Yenilmez, J. Kong, H. Dai, Appl. Phys. Lett. 79
(2001) 3155.
[17] J. Kong, A.M. Cassell, H. Dai, Nature 395 (1998) 878.
[18] J.H. Hafner, C.-L. Cheung, T.H. Oosterkamp, C.M.
Lieber, J. Phys. Chem. B 105 (2001) 743.
[19] M. Kociak, K. Suenaga, K. Hirahara, Y. Saito, T.
Nakahira, S. Iijima, Phys. Rev. Lett. 89 (2002) 155501.
[20] J.L. Hutchison, N.A. Kiselev, E.P. Krinichnaya, A.V.
Krestinin, R.O. Loutfy, A.P. Morawsky, V.E. Muradyan,
E.D. Obraztsova, J. Sloan, S.V. Terekhov, D.N. Zakharov,
Carbon 39 (2001) 761.
[21] Y. Saito, T. Nakahira, S. Uemura, J. Phys. Chem. B 107
(2003) 931.
[22] T. Sugai, H. Yoshida, T. Shimada, T. Okazaki, H.
Shinohara, S. Bandow, Nano Lett. 3 (2003) 769.
[23] S. Bandow, M. Takizawa, K. Hirahara, M. Yudasaka, S.
Iijima, Chem. Phys. Lett. 337 (2001) 48.
[24] R.R. Bacsa, Ch. Laurent, A. Peigney, W.S. Bacsa, Th.
Vaugien, A. Rousset, Chem. Phys. Lett. 323 (2000) 566.
[25] L. Ci, Z. Rao, Z. Zhou, D. Tang, X. Yan, Y. Liang, D.
Liu, H. Yuan, W. Zhou, G. Wang, W. Liu, S. Xie, Chem.
Phys. Lett. 359 (2002) 63.
[26] W.Z. Li, J.G. Wen, M. Sennett, Z.F. Ren, Chem. Phys.
Lett. 368 (2003) 299.
[27] E. Flahaut, R. Bacsa, A. Peigney, C. Laurent, Chem.
Commun. (2003) 1442.
[28] S.C. Lyu, T.J. Lee, C.W. Yang, C.J. Lee, Chem. Commun.
(2003) 1404.
[29] H. Hongo, M. Yudasaka, T. Ichihashi, F. Nihey, S. Iijima,
Chem. Phys. Lett. 361 (2002) 349.
[30] D. Zhao, Q. Huo, J. Feng, B.F. Chmelka, G.D. Stucky,
J. Am. Chem. Soc. 120 (1998) 6024.
[31] Y. Murakami, Y. Miyauchi, S. Chiashi, S. Maruyama,
Chem. Phys. Lett. 374 (2003) 53.
[32] Y. Li, W. Kim, Y. Zhang, m. Rolandi, D. Wang, H. Dai,
J. Phys. Chem. B 105 (2001) 11424.
[33] W. Kim, H.C. Choi, M. Shim, Y. Li, D. Wang, H. Dai,
Nano Lett. 2 (2002) 703.
[34] S. Bandow, M. Takizawa, H. Kato, T. Okazaki, H.
Shinohara, S. Iijima, Chem. Phys. Lett. 347 (2001)
23.