Controllable growth of single wall carbon nanotubes by pyrolizing acetylene on the floating iron catalysts

Article abstract of DOI:10.1016/S0009-2614(01)01098-3

Single wall carbon nanotubes (SWNTs) without amorphous carbon coating were prepared by thermally decomposing acetylene (C₂H₂) at the temperature range 750-1200 °C in a floating iron catalyst system. The C₂H₂ partial pressure was controlled to make a carbon supply limiting growth of SWNTs. The higher reaction temperature above 1100 °C seemed not to favor the SWNT production due to the quick thermal decomposition of C₂H₂.

Full text of DOI:10.1016/S0009-2614(01)01098-3

30 November 2001
Chemical Physics Letters 349 .2001) 191±195
www.elsevier.com/locate/cplett
Controllable growth of single wall carbon nanotubes
by pyrolizing acetylene on the ¯oating iron catalysts
*
Lijie Ci, Sishen Xie , Dongsheng Tang, Xiaoqin Yan, Yubao Li, Zuqin Liu,
Xiaoping Zou, Weiya Zhou, Gang Wang
Institute of Physics, Center for Condensed Matter Physics, Chinese Academy of Science, Group 412, P.O. Box 603-92#,
Beijing 100080, PR China
Received 7 August 2001; in ®nal form 12 September 2001
Abstract
Single wall carbon nanotubes .SWNTs) without amorphous carbon coating were prepared by thermally decom-
posing acetylene .C2H2) at the temperature range 750±1200 °C in a ¯oating iron catalyst system. The C2H2 partial
pressure was controlled to make a carbon supply limiting growth of SWNTs. The higher reaction temperature above
1100 °C seemed not to favor the SWNT production due to the quick thermal decomposition of C2H2. Ó 2001 Elsevier
Science B.V. All rights reserved.
1. Introduction
which SWNTs are grown in a ¯owing gaseous
feedstock mixture, was developed to provide a
mass-production method of SWNTs [8±11]. Cheng
et al. [8] ®rst reported that they obtained ropes of
SWNTs in high yield by pyrolizing a benzene±
ferrocene±thiophene mixture at a temperature of
1200 °C. After that, Rao and coworkers [9],
Nikolaev et al. [10] and Bladh et al. [11] also
demonstrated that SWNTs could be produced in
similar experiments by pyrolyzing other carbon-
containing gases .C2H2, CO and CH4) at a tem-
perature range 800±1200 °C.
In those studies, CO was shown to be a better
carbon feedstock to produce clean SWNTs with-
out amorphous carbon coating [10]. In contrast,
SWNTs produced from hydrocarbon sources such
as C2H2 [9] and C6H6 [8] often su€er from amor-
phous carbon coating due to their self-pyrolysis at
the reaction temperature. However, the reaction
Single wall carbon nanotubes .SWNTs) exhibit
many exceptional and useful electronic and me-
chanical properties and will be a promising can-
didate for application in nano-scale technologies
[1,2]. The most established methods to produce
SWNTs currently include laser-ablation method
[3] and carbon arc method [4]. By pyrolizing car-
bon-containing molecules such as CO [5], CH4 [6]
and C2H4 [7] on supported nanometer-sized metal
particles, chemical vapor deposition .CVD)
method can be a promising way to up-scale the
production of SWNTs. During the past three
years, a continuous CVD synthetic method, in
* Corresponding author. Fax: +86-10-82-64-95-31.
E-mail address: ssxie@aphy.iphy.ac.cn .S. Xie).
0009-2614/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved.
PII: S0009-2614.01)01098-3

192
L. Ci et al. / Chemical Physics Letters 349 /2001) 191±195
rate of CO disproportionation at a given temper-
ature and pressure is much slower than hydro-
carbon thermal decomposition, and a larger CO
¯ow amount .1000±2000 sccm) is always needed to
grow SWNTs [10].
Other reports showed that it was dicult to
produce pure SWNTs from hydrocarbon precur-
sors such as C6H6 [12] and C2H2 [11] under their
experimental conditions. It means that this kind of
SWNT synthesis process needs to be improved.
Hydrocarbons are more common carbon-con-
taining gases, and synthesizing SWNTs from hy-
drocarbon gas by this kind of process may provide
a good way to scale up SWNT production for
commercial application.
Scanning electron microscopy .SEM), trans-
mission electron microscopy .TEM) and micro-
Raman spectroscopy were performed to charac-
terize the products. Because the ®lm of the prod-
ucts were not easy to disperse even after ultrasonic
treatment for several hours, TEM samples were
prepared by simply tearing apart a small piece of
product and attaching it on the carbon copper
grid, then dropping a drop of ethanol and drying
it.
3. Results and discussion
The products in our typical experiments can be
peeled easily o€ the water-cooled collector in large
sheets. Fig. 1a shows a SEM macro-appearance of
a small piece of the products, and Fig. 1b is a local
enlarged image of this piece. It shows that the
products mainly consist of thousands of entangled
®laments with very small diameter and small par-
ticles, and most of the small particles seem deco-
rate on those ®laments and are shown as bright
color in the image. High resolution TEM image
.Fig. 1c) shows that all of those ®laments are
SWNT bundles or individual SWNTs, and the
particles are metal catalysts encapsulated by two
or three graphite layers. Analysis of XRD indi-
cates that the particles are iron. In the as-grown
®lm .as shown in Fig. 1c), SWNTs are not easy to
form large bundles due to the small particles sup-
porting. SWNTs tend to congregate to form a
large bundles where there are no particles sup-
porting. No amorphous carbon was found in the
sample, which indicates that there is no hydro-
carbon thermal decomposition on the SWNTs and
the encapsulated particles.
The partial pressure of the C2H2 was shown as a
crucial factor to produce SWNTs in our experi-
ments. Fig. 2 shows the partial pressure of C2H2
vs. the yield of SWNTs. We found that when the
C2H2 partial pressure was higher than 12 Torr,
almost no SWNTs could be obtained from the
collector. Hafner et al. [7] showed evidences that
the growth rate of SWNTs, which are prepared by
catalytic decomposition of C2H4 and CO on the
supported catalyst, is limited by the supply of
carbon to the catalyst particles. The above exper-
In this communication, we report that C2H2 can
also be a very ideal carbon feedstock to produce
clean and pure SWNTs by controlling the experi-
mental parameters carefully. In our studies, we
found that the reaction temperature of SWNT
from C2H2 can be in a wide range 750±1200 °C,
which is di€erent from other report [9]. We also
found that the partial pressure of hydrocarbon is
much crucial to synthesize SWNTs. Experiments
here may be helpful to grow SWNTs from other
hydrocarbon by the ¯oating catalyst process [13].
2. Experimental
Our experimental set-up is a two-stage furnace
system ®tted with a quartz tube .30 mm i.d.) [12].
Ferrocene .dicyclopentadienyl iron), acting as the
source of catalyst, was sublimed in the ®rst furnace
at the temperature 60±90 °C. The sublimed ferro-
cene was carried by the ¯owing Argon .1200 sccm)
and acetylene .3±10 sccm) mixture into the second
furnace. The SWNT growth temperature .the
temperature of the second furnace) can vary from
750 to 1200 °C. The pressure inside the quartz tube
was held constant at 1 atm in all the following
experiments. To collect the SWNT product, we
®tted a water-cooled collector in the end of second
furnace. We found that most of the products
condense on the water-cooled collector to form a
SWNT ®lm, and only a small faction of the
products were found on the cold part of the quartz
tube.

L. Ci et al. / Chemical Physics Letters 349 /2001) 191±195
193
(a)
Fig. 2. The partial pressure of C2H2 vs. the yield of SWNT.
The reaction temperature is 900 °C and the amount of sublimed
ferrocene 3:74 Â 10À5 mol/h.
imental data indicate that the rate-limiting step in
our ¯oating catalyst system is also the carbon
supply to the catalysts, which was achieved by
controlling the partial pressure of C2H2. Fig. 2
shows that at the controllable C2H2 partial pres-
sures blow 5.0 Torr, the product mass were grown
with the C2H2 partial pressure increased. How-
ever, when the C2H2 partial pressure excess 5.0
Torr, more thermal decomposed carbon was pro-
duced, which may make the catalyst particles in-
active, and the SWNT production began to
decrease.
(b)
Reaction temperature is also an important pa-
rameter to grow SWNT. The growth temperature
of SWNTs from hydrocarbon were reported at
1100±1200 °C [8,9]. In our experiments, we found
that SWNTs could be synthesized at a wide tem-
perature range 750±1200 °C. Table 1 shows the
yield of SWNTs vs. the reaction temperature. At a
low temperature below 800 °C, we found that
there is only small piece of the product covered on
the collector. On the other hand, most of the
products obtained at a temperature over 1100 °C
are amorphous carbon, which looks gray and
cannot peel them o€ the collector as sheets. The
optimized temperatures for the growth of SWNTs
are in the range 900±1000 °C, which is di€erent
from other reports that the highest SWNTs' yield
(c)
Fig. 1. .a) SEM macro-appearance of a small piece of the as-
grown products; .b) a local enlarged image of a; .c) HETEM
image of the as-grown ®lm, and the inserted bar is 10 nm.

194
L. Ci et al. / Chemical Physics Letters 349 /2001) 191±195
Table 1
The relation between production rate and reaction temperature
Temperature
.°C)
Sublimed ferrocene Â10À5
C2H2 Partial pressure
.Torr)
Production rate
.mg/h)
.mol/h)
800
850
4.68
4.68
3.74
4.68
4.33
5.0
5.0
5.0
5.0
5.0
0.16
0.36
1.2
900
1000
1100
1.1
0.72
Table 2
The relation between the production rate and the amount of the sublimed ferrocene
Sublimed ferrocene Â10À5
Temperature
.°C)
C2H2 Partial tressure
.Torr)
Production rate
.mg/h)
.mol/h)
11.29
8.55
5.91
3.74
900
900
900
900
6.3
6.3
6.3
6.3
4.4
3.0
1.52
1.2
were often achieved at their highest available
temperature up to 1200 °C [10,11]. The reason that
we did not obtain highest SWNTs' yield at the
higher temperature may be ascribed to the rate-
limiting step for growth of SWNT is changed due
to the much quicker thermal decomposition of
C2H2 at the higher temperature.
spectra for an as-grown ®lm, measured with a laser
excitation wavelength of 514.5 nm. The Raman
spectra consist of two main groups of peaks, which
are typical for SWNTs and indicate SWNT exist-
ing in the ®lm. The ®rst group includes the peaks
As shown above, the main products in our ex-
periments are SWNTs and encapsulated iron
particles. We found that the fraction of the en-
capsulated particles was controllable by adjusting
the sublimed temperature of ferrocene. More fer-
rocene was introduced into the second furnace
with the temperature of the ®rst furnace increased,
and the yield of the products was increased due to
the mass fraction of the encapsulated particles in-
creased .as listed in Table 2). The purer SWNT
®lm can be obtained at the lower sublimed tem-
perature.
The diameter distribution of the SWNTs was
shown in Fig. 3, which were measured from
HRTEM images. The average diameter we mea-
sured is 1.1 nm, and we did not observed SWNTs
with diameter larger than 2 nm. However, many
SWNTs as small a 0.7 nm in diameter, which is
about the diameter of a C60 molecule, were found
in our samples.
Raman spectroscopy was always employed to
identify SWNTs. Fig. 4 shows micro-Raman
Fig. 3. The diameter distribution of the typical SWNT prod-
ucts.

L. Ci et al. / Chemical Physics Letters 349 /2001) 191±195
195
produce SWNTs from other hydrocarbon gas by
the ¯oating catalyst method, which is a promising
process for bulk SWNT production.
Acknowledgements
We thank C.Y. Wang and X.H. Chen for their
assistance in SEM and TEM work. This work
is supported by National Natural Science
Foundation of China and `973' National Key
Basic Research Item
Nanostructures'.
± `Nanomaterials and
Fig. 4. The Raman spectrum of an as-grown ®lm, obtained
with an excitation wavelength of 514.5 nm. The inset shows
details of the peak distribution in the low frequency range.
References
[1] C. Dekker, Phys. Today 22 .1999).
[2] E. Hernandez, C. Goze, P. Bernier, A. Rubio, Appl. Phys.
A 68 .1999) 287.
at 137, 166, 187, 223, 266 cmÀ1, which consists of
the radial breathing modes .RBM) whose fre-
quencies depend on the SWNT's diameters .see
inserted graph). According to the relation between
diameter and frequency [14], the SWNT's diameter
in the measured sample is 1.63, 0.84 nm for 137
and 266 cmÀ1, respectively, which agrees with the
HRTEM observation. The second peak group
consists of the peaks at 1573 and 1594 cmÀ1, which
are related to the Eg2 graphite mode. The D-line
around 1344 cmÀ1 indicates that the amount of
amorphous carbon in the sample is very small.
[3] T. Guo, P. Nikolaev, A. Thess, D.T. Colbert, R.E.
Smalley, Chem. Phys. Lett. 243 .1995) 49.
[4] S. Iijima, T. Ichihashi, Nature .London) 364 .1993) 603.
[5] H. Dai, A.G. Rinzler, P. Nikolaev, A. Thess, D.T. Colbert,
R.E. Smalley, Chem. Phys. Lett. 260 .1996) 471.
[6] J. Kong, A.M. Cassell, H.J. Dai, Chem. Phys. Lett. 292
.1998) 567.
[7] J.H. Hafner, M.J. Bronikowski, B.R. Azamian, P. Niko-
laev, A.G. Rinzler, D.T. Colbert, K.A. Smith, R.E.
Smalley, Chem. Phys. Lett. 296 .1998) 195.
[8] H.M. Cheng, F. Li, X. Sun, S.D.M. Brown, M.A. Pimenta,
A. Marucci, G. Dresselhaus, M.S. Dresselhaus, Chem.
Phys. Lett. 289 .1998) 602.
[9] B.C. Satishkumar, A. Govindaraj, R. Sen, C.N.R. Rao,
Chem. Phys. Lett. 293 .1998) 47.
[10] P. Nikolaev, M.J. Bronikowski, R.K. Bradley, F. Roh-
mund, D.T. Colbert, K.A. Smith, R.E. Smalley, Chem.
Phys. Lett. 313 .1999) 91.
4. Conclusion
We have demonstrated that clean SWNTs
without amorphous carbon coating could be pro-
duced by thermally decomposing C2H2 in the
¯oating catalyst system. The rate-limiting step for
the growth of SWNTs must be controlled to be the
supply of carbon to catalyst particles by adjusting
the C2H2 partial pressure and the reaction tem-
perature. Our experiments here may be helpful to
[11] K. Bladh, L.K.L. Falk, F. Rohmund, Appl. Phys. A 70
.2000) 317.
[12] L. Ci, J. Wei, B. Wei, J. Liang, C. Xu, D. Wu, Carbon 39
.2001) 329.
[13] M. Endo, T. Koyama, Japanese patent 60-32818, 1985.
[14] R. Saito, T. Takeya, T. Kimura, G. Dresselhaus, M.S.
Dresselhaus, Phys. Rev. B 57 .1998) 4145.