CARBON 4 9 ( 2 0 1 1 ) 1 9 7 2 –1 9 7 9

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/carbon

Sub-millimeter-long carbon nanotubes repeatedly grown

on and separated from ceramic beads in a single fluidized
bed reactor

Dong Young Kim, Hisashi Sugime, Kei Hasegawa, Toshio Osawa, Suguru Noda *

Department of Chemical System Engineering, School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan

A R T I C L E I N F O A B S T R A C T

Article history: A semi-continuous fluidized-bed process is reported which rapidly converts acetylene into
Received 2 September 2010 carbon nanotubes (CNTs). Catalysts are first immobilized on ceramic beads and CNTs are
Accepted 12 January 2011 then grown on the beads and then separated from them in a repetitive process accom-
Available online 18 January 2011 plished within a single reactor simply by switching gases at a fixed temperature. CNTs of
6–10 nm diameter, three walls on average, 0.4 mm length and 99 wt.% purity were synthe-
sized at an yield of over 70% in a reactor residence time shorter than 0.3 s. The easy and
efficient production of such CNTs with in situ separation from the catalysts may accelerate
the development of CNT-based nanotechnology industries.
 2011 Elsevier Ltd. All rights reserved.

1. Introduction are typically as small as a few tens of g m 2, which is the ma-

jor obstacle for industrial nanotechnology applications based
Easy and efficient production of inexpensive, high quality on CNTs.
carbon nanonotubes (CNTs) accelerates the development of One promising approach for the mass production of verti-
CNT-based nanotechnology industries such as nanoscale cally aligned CNT (VA-CNT) arrays is to substitute flat sub-
electronic devices [1,2], flexible displays [3], solar cells/batter- strates with beads. Xiang et al. synthesized VA-CNT arrays
ies [4,5], and composites [6,7]. However, production chal- on ceramic beads by simultaneously feeding catalyst and car-
lenges remain with regard to the control of CNT structures bon sources, and proposed a (semi)continuous operation
and reduction of capital costs. Arc discharge [8], laser ablation repeating CNT growth, collection of CNT-coated beads and re-
[9], molten salt intercalation [10], and chemical vapor deposi- fill of fresh beads [22]. Although their VA-CNT arrays were as
tion (CVD) [11–13] methods have been developed to synthe- thick as 0.3 mm, their CNTs were multi-walled, with diame-
size high-quality CNTs at large-scale. In particular, fluidized ters as large as 35 nm. Later, they optimized their production
beds, using ceramic powders as catalyst supports, are method [23,24] and achieved a carbon yield of 50% [24].
effective for large-scale production of CNTs by CVD [14–17]. Simultaneous feeding of catalyst and carbon sources is effec-
However, CNTs entangle with catalysts and supports so puri- tive to prolong CNT growth, however it can cause continuous
fication is usually needed, which causes additional costs and nucleation and growth of short CNTs. Production of VA-CNT
possibly damages the CNTs. On the other hand, since the arrays by catalysts predoposited on supports with large sur-
development of the water-assisted rapid growth method in face areas has also been reported. Pint et al. fabricated cata-
2004 [18], impurity-free, millimeter-long, single walled CNTs lyst flakes by role-to-role e-beam deposition in vacuum
(SWCNTs) can now be grown by CVD as vertically aligned ar- followed by release and collection in solution, grew CNTs on
rays on catalyst-supported flat substrates [19–21]. Unfortu- the flakes by hot filament CVD, and obtained ‘‘flying carpets’’
nately, the amounts of CNTs synthesized on flat substrates of sub-millimeter-long SWCNTs at a purity 75 wt.% (CNT/

* Corresponding author: Fax: +81 3 5841 7332.

E-mail address: noda@chemsys.t.u-tokyo.ac.jp (S. Noda).
0008-6223/$ - see front matter  2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carbon.2011.01.022
 CARBON 4 9 ( 20 1 1 ) 1 9 7 2–19 7 9 1973

Fig. 1 – Schematics of the novel semi-continuous fluidized-bed process (a) and the novel catalyst re-deposition method,
which overcomes the problem of catalyst deactivation and coarsening (b).

catalyst weight ratio of 400%) [25]. Zhang et al. developed a were performed in a single fluidized bed reactor, consisting of
lamellar catalyst by impregnation in solution followed by cal- a quartz tube (22 mm inner diameter, 900 mm length, and
cination, and grew several-tens-micrometer-long MWCNTs 100–150 mm heating zone length). About 40 g of commercially
among the layers of vermiculite at a purity of 84 wt.% by available Al2O3 beads (purity > 99.95 wt.%, 0.5 mm average
fluidized-bed CVD [26]. They recently reported the mass pro- diameter, and 3.89 g/cm3 density, Taimei Chemicals, Japan)
duction of MWCNTs at a rate of 3.0 kg/h using a pilot plant were loaded into the fluidized bed reactor, heated to the tar-
fluidized bed reactor with an inner diameter of 0.5 m [27]. get temperature (1093 K), and treated under Ar at a flow rate
These approaches realized better control of CNT diameters of 3–4 standard litter per minute (slm). The subsequent steps
than the simultaneous feeding of catalyst and carbon sources of catalyst (re-)deposition, CNT growth, separation of CNTs,
[22–24], however they needed additional apparatuses and and removal of residual carbons were carried out under the
times for catalyst predeposition and allowed catalyst contam- fluidized bed condition (see Supplementary data, p. 2) by
ination (16–25 wt.%). switching gases at a fixed temperature of 1093 K. Fe/Al2O3 cat-
In this work, targeted towards mass production of sub- alyst was prepared on the beads by feeding metalorganic va-
millimeter-long CNTs, we coupled rapid growth on flat sub- pors of aluminum-isopropoxide (95%, Wako, Japan) and
strates with fluidized-bed CVD using spherical ceramic beads ferrocene (98%, Wako, Japan) sequentially using a carrier gas
instead of ceramic powders. To obtain better control of cata- of 4 vol.% O2/Ar balance at a flow rate of 10 slm. Aluminum-
lyst diameter and thus CNTs, and to ensure the continuous isopropoxide vapor was supplied using an evaporator by flow-
growth of CNTs from the catalysts on the beads surface, we ing 0.15 slm Ar at 403 K for 1 min, and then ferrocene vapor
separated the processes for catalyst deposition and CNT was supplied using another evaporator by flowing 0.05 slm
growth. To make the most effective use of the high CNT Ar at 393 K for 1 min. The distributor was set the bottom of
growth rate during the initial 10 min of CVD [28], and to sup- the furnace so as to prevent deposition of Al and Fe on the
press the catalyst contamination to the CNTs, we separated distributor. The catalyst was reduced by flowing 26 vol.% H2/
the VA-CNTs from the beads in situ by vigorous fluidization, 600 ppmv H2O/Ar balance at 3.16 slm for 10 min. Then, sub-
deposited the catalysts in situ by CVD, grew CNTs by CVD, millimeter-long CNTs were synthesized by fluidized-bed
and repeated these processes rapidly (Fig. 1a). To overcome CVD at standard conditions of 1.1 vol.% C2H2/26 vol.% H2/
catalyst deactivation and coarsening, we developed a novel 600 ppmv H2O/Ar at 3.16 slm, 1 atm for 10 min. The synthe-
catalyst deposition method that involves combustion of the sized CNTs were separated from the beads by vigorously
residual carbon on the beads, coating the beads with a fresh fluidizing the bed by flowing Ar carrier gas at a flow rate of
supporting layer, and depositing a thin catalyst layer on the 10–15 slm for 1 min. The residual carbon on the beads was
supporting layer (Fig. 1b). Furthermore, to minimize the time removed by combustion by flowing 20 vol.% O2/Ar at 5 slm
and energy required for temperature control, all of these pro- for 5 min.
cesses were performed simply by switching the gas flow at a The synthesized CNT samples were characterized by field
fixed temperature in a single reactor. emission scanning electron microscopy (SEM; Hitachi, S-4700
and S-4800), transmission electron microscopy (TEM, JEOL,
2. Experimental 2000EX), Raman spectroscopy (HORIBA, HR 800) with an
excitation laser of wavelength 488 nm, and thermogravimet-
All experiments for catalyst (re-)deposition, CNT growth, sep- ric-differential thermal analysis (TG-DTA; Rigaku, TG8120).
aration of synthesized CNTs, and removal of residual carbons Brunauer–Emmett–Teller (BET) surface area of the CNT
1974 CARBON 4 9 ( 2 0 1 1 ) 1 9 7 2 –1 9 7 9

sample was determined by N2 (77 K) volumetric adsorption perse, the bead surface was covered with an fresh Al2O3 layer
(Quantachrome, Autosorb1-MP) after pre-evacuation at 423 K and Fe catalyst particles were re-deposited (i). This cycle was
and 10 4 Pa for 2 h. repeated every 30 min simply by gas flow modulation at a
fixed temperature of 1093 K in a single fluidized bed reactor.
3. Results and discussion During the semi-continuous operation, the synthesized CNTs
were collected in a 1 L polycarbonate bottle for 20 cycles in
3.1. Semi-continuous production of CNTs 10 h (Fig. 2b, also see Supplementary data, Fig. S2). The CNT
yield increased and stabilized after the first few cycles
A snapshot of the fluidized bed reactor was taken at ambient (Fig. 1b) due to the well-developed multilayer structure of
temperature without gas flow at each step by interrupting the the catalyst on the beads’ surface, as discussed in Section
semi-continuous operation (Fig. 2a). The operation was 3.3. Because the beads’ diameter (0.5 mm) was over ten thou-
started by loading pristine Al2O3 beads into the fluidized sand times larger than the catalyst and support layer thick-
bed reactor and heating to 1093 K. The Al2O3 support layer nesses (10 nm), we estimate the reusability of the beads to
(underlayer) and Fe catalysts were deposited on the beads be P10,000 cycles, corresponding to a continuous operation
by flowing aluminum-isopropoxide and ferrocene vapors, of P200 days.
sequentially, for 1–2 min (i). The catalyst was then reduced Fig. 3a and b show an illustration and a photograph,
by flowing H2/H2O/Ar for 10 min and CNTs were subsequently respectively, of the fluidized bed apparatus, which clearly
grown by flowing C2H2/H2/H2O/Ar for 10 min [21] under slug- shows that the CNT container was much larger than the reac-
ging fluidized bed (see Supplementary data, Movie S4). The tor. The easy separation and collection, together with the ra-
bead bed increased in height remarkably, by over five times, pid catalyst (re-)deposition, are advantageous for continuous
from 30 to 155 mm due to the sub-millimeter-long CNT arrays production of sub-millimeter-long CNTs. Since the first
grown on the bead surfaces (ii) (see Supplementary data, large-scale synthesis of aligned CNTs in 1996 [30], safe collec-
Fig. S1). These CNT arrays were readily separated from the tion of synthesized CNTs has been an important issue be-
beads by vigorously fluidizing the beads with an increased cause of the potential risks caused by peculiar
Ar carrier gas flow of 10 slm within 1 min and collected in nanostructures in humans and the environment [31–33]. In
a 1 L bottle (see Supplementary data, Movie S5). The bead our system, the separated CNTs from the beads were carried
bed decreased its height from 155 to 30 mm during vigorous by the gas flow to a large diameter tube, where most of the
fluidization (iii), indicating an almost perfect separation of CNTs fell down by gravity to the container bottle due to the
CNTs from the beads. Then, the residual carbon on the beads reduced gas velocity. The other CNTs were once captured by
was removed by oxidization by flowing O2/Ar for 5 min. Be- a filter (membrane filter, pore size 5 lm) set at the exit of
cause the Fe catalyst particles on the Al2O3 underlayer are the large diameter tube and then fell down to the container
coarsened after CVD [29] and therefore become hard to re-dis- bottle by stopping the gas flow for a moment (see Supplemen-

Fig. 2 – Semi-continuous production of sub-millimeter-long CNTs. (a) Snapshots of the fluidized-bed reactor at each step of
the semi-continuous operation. All steps were performed simply by gas flow modulation at a fixed temperature in a single
fluidized bed reactor. (b) The container bottle filled with sub-millimeter-long CNTs after 20 cycles over 10 h.
 CARBON 4 9 ( 20 1 1 ) 1 9 7 2–19 7 9 1975

Fig. 3 – (a) Schematic of the semi-continuous fluidized-bed

CVD; (b) a photo of the 50 mL fluidized bed reactor showing
the 1 L container bottle filled with sub-millimeter-long
CNTs.

tary data, Movie S5). This system enables a complete and con-
tinuous collection of CNTs by a filter with minimal pressure
drop. We expect that our easy collection method can enable
safe and efficient mass production of CNTs or other advanced
nanomaterials. The amount of sub-millimeter-long CNTs col-
lected in the bottle was about 0.26 g per cycle using a 50 mL
reactor volume, which corresponds to a productivity over
0.2 kg-CNT per L-reactor per day. In addition, our semi-
continuous fluidized-bed CVD method provides an efficient
synthesis of CNTs at a high yield of over 70% (0.26 g-CNT/ Fig. 4 – Characterization of CNTs on and off the beads. SEM
(0.386 g-C2H2 · 24/26) = 73 at.%), which is higher than that of images were taken for the pristine beads (a) and beads after
conventional production methods reported previously [34]. It CVD (b, c). An enlarged image of the CNTs in (c) is shown in
is noteworthy that this high yield was obtained for a gas res- (d). SEM images were taken also for the CNTs separated
idence time shorter than 0.3 s. The bead diameter of 0.5 mm in situ from the beads and collected in the container bottle
allowed a large gas flow rate (i.e. a large carbon feed rate), (e, f).
while offering a sufficiently large surface area for supporting
catalyst particles.
larged image (Fig. 4f), some disorder can be seen in the verti-
3.2. Characterization of CNTs cal alignment of the CNTs, which is typical for the vertically
aligned CNTs [28].
Figs. 4a–d show the SEM images of the ceramic beads before Then the CNTs shown in Fig. 4 were transferred from the
and after CNT growth. The beads with CNTs are taken out bead to a TEM grid using tweezers and characterized by
from the fluidized bed reactor after 20 cycles without sepa- TEM. Fig. 5 shows the TEM images and histograms of diame-
rating the CNTs from the beads. A clean spherical surface was ter distributions of the CNTs taken near the top (a), middle (b),
observed for the pristine beads (Fig. 4a) whereas a ‘‘chestnut’’ and bottom (c) of a single piece of the CNTs taken out from
morphology was observed for the beads after CVD, with the the forest. The individual CNTs were thin-walled, involving
vertically aligned CNT forests 0.4 mm in height (Figs. 4b–d). three walls on average. Their diameter gradually increased
The outer diameter of the beads increased from 0.5 to from the top (a) to the bottom (c), which is in good agreement
1.3 mm, which was related to the 5 times increased height with the diameter increase in growing SWCNTs during their
of the bead bed in the reactor (Fig. 2a). The average CNT rapid growth on flat substrates by CVD [29]. The gradual
growth rate was 0.7 lm s 1 during CVD of 10 min, which is coarsening of the catalyst particles during CVD possibly
higher than the rate generally reported for fluidized bed caused the diameter increase in CNTs during growth as dis-
methods [34]. Figs. 4e,f show the SEM images of the CNTs sep- cussed in detail in the previous report [29]. The overall aver-
arated in situ from the beads and collected in the container age diameter of CNTs was 8.1 ± 2.1 [nm].
bottle for 20 cycles. The CNTs preserved their sub-millime- Then the CNTs collected in the container bottle (Fig. 3b) for
ter length and vertical alignment after the in situ separation 20 cycles were characterized. The N2 adsorption measure-
(Fig. 4e), owing to the reduced gas flow rate for CNT growth ment (Fig. 6a) showed that the CNTs had a BET surface area
(3.16 slm) than for CNT separation (10–15 slm). At the en- of 4.0 · 102 m2 g 1, which is about one sixth of the theoretical
1976 CARBON 4 9 ( 2 0 1 1 ) 1 9 7 2 –1 9 7 9

Fig. 5 – Characterization of CNTs transferred from the bead

to a TEM grid using tweezers. TEM images and histograms
of diameter distributions of CNTs were obtained at the top
(a), middle (b) and bottom (c) of a single piece of the CNTs
taken out from the forest.
Fig. 6 – Characterization of the CNTs collected in the
container bottle. (a) N2 adsorption isotherms of the sub-
value for monolayer graphene (2.7 · 103 m2 g 1) and thus millimeter-long CNTs at 77 K. CNTs were separated from the
about one third that of SWCNTs with closed ends [35]. This re- beads by vigorous fluidization. The open and filled symbols
sult indicates that the number of walls of the synthesized indicate adsorption and desorption processes, respectively.
CNTs was three on average, which is in good agreement with The inset shows N2 adsorption isotherms in logarithmic
the TEM observation (Fig. 5). Note that SWCNTs grown by CVD scale. (b) TG-DTA curves of the CNTs separated from beads
typically have smaller surface areas (300–1200 m2 g 1) than by vigorous fluidization.
theoretical value due to the presence of impurities [36–38].
The TG-DTA curves (Fig. 6b) show that the CNTs had a high served at low number density. Whereas after 5–10 cycles,
carbon purity of 99 wt.%, a small content of amorphous car- these small catalyst particles were clearly observed at a high
bon of 2 wt.%, and a high oxidation stability of about 920 K. density with a few large particles (about 30 nm in diameter).
These results show that the catalyst structure (i.e. Fe particles
3.3. Characterization of catalysts and CNTs for different on Al2O3 underlayers) got well developed after a few cycles,
cycles which is in good agreement with the change in the CNT
growth yields (Fig. 2b).
Fig. 7 shows the SEM images of the surfaces of the pristine Fig. 8 shows the TEM images and histograms of diameter
beads (a) and beads used for 1–10 cycles after reduction by distributions of as-collected CNTs for different cycles. The
hydrogen (b–e). The images at the upper column are low mag- average diameter was about 8 nm for any cycles, which is al-
nification images showing the distribution of the surface most the same as that for the CNTs grown on the beads after
roughness. The pristine beads (a) had two regions; i.e. a 20 cycles (8.1 ± 2.1 nm, Fig. 5). For the first few cycles, the cat-
smooth surface and a rough, faceted surface. As the number alyst structure was not well developed (Fig. 7) and the CNT
of the cycles increased from (b) to (e), the rough region be- yield was small (Fig. 2). But after the few cycles, the catalyst
came smoother, showing that the surfaces were getting cov- structure got well developed and both the yield (Fig. 2) and
ered with the Al2O3 layer. The images at the lower column diameter (Fig. 8) of the CNTs were stabilized. These results
are high magnification images at the smooth region showing show the validity of our concept of the repetitive production
the catalyst particle structure. After the first 1–3 cycles (b–c), of CNTs without changing the beads (Fig. 1a) and the effec-
small catalyst particles (about 5–15 nm in diameter) were ob- tiveness of our catalyst re-deposition method (Fig. 1b).
 CARBON 4 9 ( 20 1 1 ) 1 9 7 2–19 7 9 1977

Fig. 7 – Characterization of the catalysts on the beads. SEM images were taken for the surfaces of the pristine beads (a) and
beads used for 1–10 cycles after reduction by hydrogen (b–e) at a tilt angle around 45. The low magnification images at the
upper column show the distribution of the surface roughness. The high magnification images at the lower column show the
catalyst particle structure taken at a smooth region in the low magnification images.

Fig. 8 – Characterization of the CNTs collected for different cycles of (a) 1–5, (b) 6–10, (c) 11–15 and (d) 16–20. TEM images were
taken for the as-collected CNTs by transferring them to each TEM grid using tweezers.

The catalyst particles formed under the condition of this ticles by CVD, targeting at the efficient production of sub-
work had a diameter around 5–15 nm. This is why the CNTs millimeter-long SWCNTs.
synthesized in this work were mostly 6–10 nm in diameter
and triple-walled on average. Actually we can grow SWCNTs 4. Conclusions
by fluidized bed when we use beads with sputter-deposited
catalysts [39] instead of CVD-catalysts. And the negative ef- Aiming at the efficient production of CNTs with well-
fect of H2 forming by the decomposition of carbon feedstock controlled structures, a novel, semi-continuous fluidized
[17] does not appear in our growth condition (note that we CVD method was proposed and developed, which rapidly
add excess H2 (26 vol.%) to C2H2 (1.1 vol.%) but SWCNTs grow converted C2H2 into sub-millimeter-long CNTs at a yield of
from the sputter-deposited catalysts). Reducing the nominal over 70 at.%, with a residence time of <0.3 s. Processes for cat-
thickness of the catalysts is usually effective to reduce the alyst (re-)deposition, CNT growth, CNT separation, and oxida-
catalyst particle size in case of the sputter-deposited catalysts tion of residual carbon were repeated simply by switching the
[19,20], however reducing the feed of ferrocene resulted in the gas flow at a fixed temperature, which minimized the time
significant decrease in the CNT yield in case of the CVD-cata- and energy required for temperature control. By separating
lysts in this work. This is possibly due to the different mech- the steps in the process, CNT structures were well controlled
anism of catalyst deposition between sputter-deposition and at diameters of 6–10 nm, the number of walls at 3, with a
CVD: incubation time for deposition usually appears not for high carbon purity of 99 wt.%, a burning temperature of
the former [40] but for the latter [41]. Further study is ongoing 920 K, and a length of 0.4 mm. CNTs were produced in a
to reduce the size and increase the density of the catalyst par- 50 mL reactor and 1 L of CNTs was collected over 20 cycles
1978 CARBON 4 9 ( 2 0 1 1 ) 1 9 7 2 –1 9 7 9

in 10 h, which corresponds to a productivity over 0.2 kg-CNTs [15] Qian W, Wei F, Wang Z, Liu T, Yu H, Luo G, et al. Production of
per L-reactor per day. The fluidized bed is compatible with carbon nanotubes in a packed bed and a fluidized bed. AIChE
scale-up, which is now underway. J 2003;49:619–25.
[16] Pauluhn J. Multi-walled carbon nanotubes (Baytubes):
approach for derivation of occupational exposure limit. Regul
Acknowledgements Toxicol Pharmacol 2010;57:78–89.
[17] Liu Y, Qian W, Zhang Q, Ning G, Luo G, Wang Y, et al.
Synthesis of high-quality, double-walled carbon nanotubes
This work is financially supported by Nanotech Challenge
in a fluidized bed reactor. Chem Eng Technol 2009;32:73–9.
(No. 07005623-0) by NEDO, Japan, by cooperative research pro-
[18] Hata K, Futaba DN, Mizuno K, Namai T, Yumura M, Iijima S.
ject with Hitachi Chemical, Co., Ltd., Japan, and by KAKENHI Water-assisted highly efficient synthesis of impurity-free
(Nos. 18686062 and 21686074), by MEXT, Japan. single-walled carbon nanotubes. Science 2004;306:1362–4.
[19] Noda S, Hasegawa K, Sugime H, Kakehi K, Zhang Z,
Maruyama S, et al. Millimeter-thick single-walled carbon
Appendix A. Supplementary data
nanotube forests: hidden role of catalyst support. Jpn J Appl
Phys 2007;46:L399–401.
Supplementary data associated with this article can be found,
[20] Sugime H, Noda S. Millimeter-tall single-walled carbon
in the online version, at doi:10.1016/j.carbon.2011.01.022. nanotube forests grown from ethanol. Carbon
2010;48:2203–11.
[21] Hasegawa K, Noda S. Millimeter-tall single-walled carbon
R E F E R E N C E S nanotubes rapidly grown with and without water. ACS Nano;
in press. doi:10.1021/nn102380j.
[22] Xiang R, Luo G, Qian W, Wang Y, Wei F, Li Q. Large area growth
of aligned CNT arrays on spheres: towards large scale and
[1] Javey A, Guo J, Wang Q, Lundstrom M, Dai H. Ballistic carbon continuous production. Chem Vapor Depos 2007;13:533–6.
nanotube field-effect transistors. Nature 2003;424:654–7. [23] Zhang Q, Huang J, Zhao M, Qian W, Wang Y, Wei F. Radial
[2] Kang SJ, Kocabas C, Kim H, Cao Q, Meitl MA, Khang D, et al. growth of vertically aligned carbon nanotube arrays from
Printed multilayer superstructures of aligned single-walled ethylene on ceramic spheres. Carbon 2008;46:1152–8.
carbon nanotubes for electronic applications. Nano Lett [24] Xiang R, Luo G, Yang Z, Zhang Q, Qian W, Wei F. Large area
2007;7:3343–8. growth of aligned CNT arrays on spheres: cost performance
[3] Wu Z, Chen Z, Du X, Logan JM, Sippel J, Nikolou M, et al. and product control. Mater Lett 2009;63:84–7.
Transparent, conductive carbon films. Science [25] Pint CL, Pheasant ST, Pasquali M, Coulter KE, Schmidt HK,
2004;305:1273–6. Hauge RH. Synthesis of high aspect-ratio carbon nanotube
[4] Li G, Wang F, Jiang Q, Gao X, Shen P. Carbon nanotubes with ‘‘flying carpets’’ from nanostructured flake substrates. Nano
titanium nitride as a low-cost counter-electrode material for Lett 2008;8:1879–83.
dye-sensitized solar cells. Angew Chem Int Ed [26] Zhang Q, Zhao M, Huang J, Liu Y, Wang Y, Qian W, et al.
2010;49:3653–6. Vertically aligned carbon nanotube arrays grown on a
[5] Yang L, Fishbine BH, Migliori A, Pratt LR. Molecular lamellar catalyst by fluidized bed catalytic chemical vapor
simulation of electric eouble-layer capacitors based on deposition. Carbon 2009;47:2600–10.
carbon nanotube forests. J Am Chem Soc 2009;131:12373–6. [27] Zhang Q, Zhao M-Q, Huang J-Q, Nie J-Q, Wei F. Mass
[6] Hou Y, Tang J, Zhang H, Qian C, Feng Y, Liu J. Functionalized production of aligned carbon nanotube arrays by fluidized
few-walled carbon nanotubes for mechanical reinforcement bed catalytic chemical vapor deposition. Carbon
of polymeric composites. ACS Nano 2009;3:1057–62. 2010;48:1196–209.
[7] Eder D. Carbon nanotube-inorganic hybrids. Chem Rev [28] Hasegawa K, Noda S. Real-time monitoring of millimeter-tall
2010;110:1348–85. vertically aligned single-walled carbon nanotube growth on
[8] Shi Z, Lian Y, Zhou X, Gu Z, Zhang Y, Iijima S, et al. Mass- combinatorial catalyst library. Jpn J Appl Phys
production of single-wall carbon nanotubes by arc discharge 2010;49:085104-1–6.
method. Carbon 1999;37:1449–53. [29] Hasegawa K, Noda S. Diameter increase in millimeter-tall
[9] Maser WK, Muñoz E, Benito AM, Martı́nez MT, de la Fuente vertically aligned single-walled carbon nanotubes during
GF, Maniette Y, et al. Production of high-density single- growth. Appl Phys Express 2010;3:045103-1–3.
walled nanotube material by a simple laser-ablation method. [30] Li WZ, Xie SS, Qian LX, Chang BH, Zou BS, Zhou WY, et al.
Chem Phys Lett 1998;292:587–93. Large-scale synthesis of aligned carbon nanotubes. Science
[10] Hsu WK, Hare JP, Terrones M, Kroto HW, Walton DRM, Harris 1996;274:1701–3.
PJF. Condensed-phase nanotubes. Nature 1995;377:687. [31] Chou CC, Hsiao H, Hong Q, Chen C, Peng Y, Chen H, et al.
[11] Nikolaev P, Bronikowski MJ, Bradley RK, Rohmund F, Colbert Single-walled carbon nanotubes can induce pulmonary
DT, Smith KA, et al. Gas-phase catalytic growth of single- injury in mouse model. Nano Lett 2008;8:437–45.
walled carbon nanotubes from carbon monoxide. Chem Phys [32] Zhao Y, Xing G, Chai Z. Are carbon nanotube safe? Nat
Lett 1999;313:91–7. Nanotech 2008;3:191–2.
[12] Jeong HJ, Kim KK, Jeong SY, Park MH, Yang CW, Lee YH. High- [33] Plata PL, Hart AJ, Reddy CM, Gschwend PM. Early evaluation
yield catalytic synthesis of thin multiwalled carbon of potential environmental impacts of carbon nanotube
nanotubes. J Phys Chem B 2004;108:17695–8. synthesis by chemical vapor deposition. Environ Sci Technol
[13] Endo M, Muramatsu H, Hayashi T, Kim YA, Terrones M, 2009;43:8367–73.
Dresselhaus MS. ‘Buckypaper’ from coaxial nanotubes. [34] See CH, Harris AT. A review of carbon nanotube synthesis via
Nature 2005;433:476. fluidized-bed chemical vapor deposition. Ind Eng Chem Res
[14] Bachilo SM, Balzano L, Herrera JE, Pompeo F, Resasco DE, 2007;46:997–1012.
Weisman RB. Narrow (n, m)-distribution of single-walled [35] Hiraoka T, Izadi-Najafabadi A, Yamada T, Futaba DN, Yasuda
carbon nanotubes grown using a solid supported catalyst. J S, Tanaike O, et al. Compact and light supercapacitor
Am Chem Soc 2003;125:11186–7. electrodes from a surface-only solid by opened carbon
 CARBON 4 9 ( 20 1 1 ) 1 9 7 2–19 7 9 1979

nanotubes with 2200 m2g–1 surface area. Adv Fun Mater [39] Kim DY, Fukai H, Sugime H, Hasegawa K, Osawa T, Noda S.
2010;20:422–8. Fluidized-bed CVD for sub-millimeter-long single-walled
[36] Yang CM, Kim DY, Lee YH. Formation of densely packed carbon nanotubes. In: 10th International conference on the
single-walled carbon nanotube assembly. Chem Mater science and application of nanotubes (Beijing); 2009. p. 69.
2005;17:6422–9. [40] Hu M, Noda S, Komiyama H. Amorphous-to-crystalline
[37] Kim DY, Yang CM, Park YS, Kim KK, Jeong SY, Han JH, et al. transition during the early stages of thin film growth of Cr on
Characterization of thin multi-walled carbon nanotubes SiO2. J Appl Phys 2003;93:9336–44.
synthesized by catalytic chemical vapor deposition. Chem [41] Kajikawa Y, Tsuchiya T, Noda S, Komiyama H. Incubation
Phys Lett 2005;413:135–41. time during the CVD of Si onto SiO2 from silane. Chem Vap
[38] Miyamoto J, Hattori Y, Noguchi D, Tanaka H, Ohba T, Utsumi Deposition 2004;10:128–33.
S, et al. Efficient H2 adsorption by nanopores of high-purity
double-walled carbon nanotubes. J Am Chem Soc
2006;128:12636–7.