Home
Search
Collections
Journals
About
Contact us
My IOPscience
Air-assisted growth of ultra-long carbon nanotube bundles
This content has been downloaded from IOPscience. Please scroll down to see the full text.
2008 Nanotechnology 19 455609
(http://iopscience.iop.org/0957-4484/19/45/455609)
View the table of contents for this issue, or go to the journal homepage for more
Download details:
IP Address: 128.226.37.5
This content was downloaded on 14/05/2015 at 08:01
Please note that terms and conditions apply.
IOP PUBLISHING
NANOTECHNOLOGY
Nanotechnology 19 (2008) 455609 (7pp)
doi:10.1088/0957-4484/19/45/455609
Air-assisted growth of ultra-long carbon
nanotube bundles
Xuesong Li1,3 , Xianfeng Zhang2 , Lijie Ci1 , Rakesh Shah2 ,
Christopher Wolfe2 , Swastik Kar1 , Saikat Talapatra2,3 and
Pulickel M Ajayan1
1
Department of Materials Science and Engineering, Rensselaer Polytechnic Institute, Troy,
NY 12180, USA
2
Department of Physics, Southern Illinois University Carbondale, IL 62901, USA
E-mail: lmfe2005@gmail.com (X Li) and stalapatra@physics.siu.edu
Received 11 July 2008, in final form 25 August 2008
Published 9 October 2008
Online at stacks.iop.org/Nano/19/455609
Abstract
We report an air-assisted chemical vapor deposition (CVD) method for the synthesis of
super-long carbon nanotube (CNT) bundles. By mixing a small amount of air in the vapor phase
catalyst CVD process, the catalyst lifetime can be dramatically increased, and extremely long
dense and aligned CNT bundles up to 1.5 cm can be achieved. Electron microscopy
characterization shows that the injection of air does not damage the CNT structures. Further, we
have estimated that individual ultra-long CNTs can carry moderate current densities
∼105 A cm−2 , indicating their possible use in nanoelectronic devices.
(Some figures in this article are in colour only in the electronic version)
years have seen tremendous progress in synthesizing bulk CNT
architectures [6–18].
Currently, the most popular method used for CNT
growth is chemical vapor deposition (CVD) [19–21] since
this approach provides enough flexibility for controlling two
main aspects of CNT growth: growth rate and growth
duration. Using this method, individual single-walled carbon
nanotubes (SWNTs) can grow up to 4 cm at a high growth
rate of 11 μm s−1 [13] and multi-walled carbon nanotubes
(MWNTs) up to ∼10 cm within 10 h [14]. However, for
growing dense aligned CNTs (ACNTs), the growth rates are
much lower, typically between several microns to tens of
microns per minute [7, 10–12, 16–18, 22]. Typically, the
growth rates of CNTs decrease with reaction time increase
and consequently ACNT growth terminates after a certain
time. Some researchers attributed the growth termination
of ACNTs to the lack of diffusion of carbon—the carbon
source is increasingly blocked by the ACNT film height
increase [7, 23, 24]. However, in some other cases it has
been shown that vapor can freely diffuse through the as-grown
CNT forest to the substrate surface to initiate the growth of
a new ACNT layer [25–27]. In these cases ACNT growth
termination is attributed to catalyst deactivation, due to the
graphitic encapsulation of the catalyst particle with amorphous
carbon [28–31]. This graphitic encapsulation detrimentally
1. Introduction
Carbon nanotubes (CNTs), because of their extremely appealing physical properties, show promise for various applications
ranging from vacuum microelectronics to biological sensors. It
is also envisioned that there are technologies [1–4] which will
need macro-architectures fabricated using CNTs as building
blocks. Since these architectures (for example thick CNT films
and fibers) possess superb properties, such as high strength,
low density, high specific surface area, and excellent thermal
and electrical conductivity, they are considered as potential
engineering materials for developing applications in a variety
of fields, such as composites, field emission display devices,
electrochemical sensing and electrochemical energy storage
devices. Recently it was also reported that metallic CNTs or
long dense CNT bundles (∼1 mm) are expected to outperform
copper in terms of failure current density, power dissipation,
and on-chip signal transfer delays as far as their use as
nanoscale interconnects are concerned [5]. Further advantages
of growing long vertically aligned CNTs are that small bundles
of requisite diameters can then be easily separated from the
parent CNT forest, which then forms microscopic multifilament wires for electronic applications. Thus, the past few
3 Authors to whom any correspondence should be addressed.
0957-4484/08/455609+07$30.00
1
© 2008 IOP Publishing Ltd Printed in the UK
Nanotechnology 19 (2008) 455609
X Li et al
Figure 1. (a) Photograph of the 1.5 cm long densely aligned CNTs (growth parameters: g0 = 200 sccm, gair = 5 sccm,
CFerrocene = 0.02 g ml−1 , v = 0.12 ml min−1 , T1 = 250 ◦ C, T2 = 800 ◦ C, t = 15 h) and SEM images of different parts of the film: (b) bottom,
(c) middle, and (d) top, respectively. Scale bars: 1 μm.
affects the availability of its surface to hydrocarbon precursors
and subsequently deters CNT formation. When the catalyst
particle is completely encapsulated, CNT growth terminates.
Adding an optimal amount of oxidizer such as water [12] or
oxygen [15] has been proven to effectively increase catalyst
lifetime to keep CNT growing. In this way densely aligned
SWNTs can grow up to 2.5 mm [12] and MWNTs up to
7 mm [16] by catalytic deposition of ethylene on pre-deposited
catalyst. In the present work we show that the growth of
aligned MWNTs by vapor phase catalyst CVD can be greatly
enhanced by mixing a small amount of air during the growth
process. By optimizing the growth parameters, 1.5 cm long
densely aligned MWNTs can be produced. The oxygen in air
serves as the oxidizer that can maintain catalyst activity for a
very long time. We also tested the failure current densities of
small bundles of these long ACNTs in ambient conditions.
(catalyst precursor) and xylene (carbon source) at a determined
concentration (CFerrocene ) was injected into the evaporator by
a syringe pump at a constant rate (v ). The ferrocene/xylene
vapor was carried by Ar/H2 at an optimal rate (g0 ) into the
furnace tube. In this particular work a small amount of air (gair )
was mixed in the reaction environment to help maintain the
catalyst activity. The reaction time was controlled by adjusting
the feeding time of the ferrocene/xylene solution. Densely
aligned MWNTs were grown on SiO2 substrates.
The as-grown samples were characterized by scanning
electron microscopy (SEM), transmission electron microscopy
(TEM), and Raman spectroscopy. Electrical measurement was
performed by the two-terminal current–voltage method.
3. Results and discussion
3.1. Characterization of the as-grown long aligned MWNTs
2. Experimental details
Figure 1(a) shows a photograph of the extremely long aligned
MWNT block up to 1.5 cm grown at 800 ◦ C for 15 h. Different
parts (bottom, middle, and top) of the nanotube block were
examined by SEM (figures 1(b)–(d)) and TEM (figures 2(a)
and (b)). Compared to the uniformly pyrolyzed carbon coated
CNTs grown by normal CVD (without air) we previously
reported [22], here the coatings vary along CNTs: the CNTs
close to the substrate (bottom part) are very clean and the
upper portions are coated with more and more pyrolyzed
carbon. CNT diameter distribution was analyzed based on
TEM images by counting more than 100 tubes (figure 2(c)).
The average outer diameter of the bottom CNTs is 35 nm,
A typical xylene–ferrocene injection CVD system was used
here for CNT growth [21] (horizontal tube furnace with glass
tube 4 feet in length and 1.8 inch in inner diameter). The
growth process was similar to that reported previously [22].
A mixture of Ar/H2 (85%/15%) was fed into the system
at a flow rate of 300 sccm for ∼20 min to purge out the
system. In the mean time, the evaporator (for catalyst and
carbon source solution evaporation) and the tube furnace
(for CNT growth) were heated to the desired temperatures
(T1 and T2 , respectively). Then the solution of ferrocene
2
Nanotechnology 19 (2008) 455609
X Li et al
0
Figure 2. TEM images of (a) bottom and (b) top parts of the 1.5 cm long densely aligned CNTs, and (c) diameter distributions of bottom and
top CNTs. The inset in (a) is the high resolution TEM image showing good graphitization of the air-assisted CVD grown CNTs.
similar to our previously reported results [22]. However,
the pyrolyzed carbon shell coated on the top CNTs is much
thinner than that grown in normal CVD within equal growth
time (i.e. 100 nm versus >400 nm for a 15 h growth).
This can be explained as follows. In the absence of air,
the catalysts easily lose their activities. The hydrocarbons
accordingly thermally decompose and CNTs are evenly coated
with pyrolyzed carbon. On the other hand in the presence of
air, catalysts can retain their activities for a longer time. As
a result, the hydrocarbons surrounding the lower portion of
CNTs are decomposed catalytically for the use of CNT growth
instead of forming pyrolyzed carbon coatings (initial growth
process). Although thermal decomposition of hydrocarbons
can still occur far from the catalysts for a long period of growth
(in the middle and top portion of the CNTs), because of oxygen
etching, the coating rate of pyrolyzed carbon is slow compared
to that without air. The gradient of pyrolyzed carbon coatings
from the top through to the bottom also provides evidence of
the base-growth mode since the fresh CNTs are at the bottom.
The high resolution TEM image (the inset in figure 2(a))
shows that the graphitization of the air-assisted grown CNTs
is comparable to that grown by normal CVD (although the
crystal structure is somewhat defective, it is the nature of the
CVD method, no matter whether the normal CVD or the airassisted CVD) [22]. Raman (532 nm laser wavelength) spectra
also showed that CNTs grown by the two methods have similar
Ig /Id (the ratio of the intensity of the graphitic-like G peak to
that of the disorder-induced D peak) values: 2 for nanotubes
and 1.4 for pyrolyzed carbon coatings (figure 3). This is
because amorphous carbon is more active than CNTs and
Figure 3. Raman spectra of ACNTs grown at 800 ◦ C: (1) no air,
30 min; (2) 5 sccm air, 5 h, bottom; (3) no air, 5 h; (4) 5 sccm air, 5 h,
top.
hence oxygen prefers to react with amorphous carbon rather
than CNTs. This graphitization similarity indicates that mixing
an optimized amount of air does not damage CNT structures.
3
Nanotechnology 19 (2008) 455609
X Li et al
Figure 5. (a) Temperature effects on ACNT growth kinetics.
(b) Arrhenius plot of CNT growth rate.
Figure 4. (a) Air flow effects on ACNT growth kinetics. (b) Air flow
effects on average outer diameters of CNTs. Growth parameters:
g0 = 200 sccm, CFerrocene = 0.01 g ml−1 , v = 0.11 ml min−1 ,
T1 = 250 ◦ C, T2 = 770 ◦ C.
0.01–0.05 g ml−1 , v = 0.11 ± 0.005 ml min−1 , T1 =
170–300 ◦ C. For other temperatures the values shift slightly
but keep wide ranges. This wide growth window makes it
easier to achieve super-long aligned MWNTs.
3.2. Air flow effects
Figure 4(a) shows the effects of air flow rate on the growth
kinetics of the aligned MWNTs. When no air is injected, CNT
growth shows an obvious lengthening–thickening process. In
the lengthening process, catalysts are active and CNTs grow at
a constant rate. In the thickening process, CNTs stop growing
in length and their diameters are increased by being coated with
pyrolyzed carbon. When an optimal amount of air is injected,
catalyst lifetime is dramatically increased, and hence the CNT
lengthening process can extend to a very long time. The
longest experiment we carried out is for 15 h, and there was
still no obvious decrease in CNT growth rate. The amount of
air also affects the thickness of the pyrolyzed carbon coatings
on CNTs, as shown in figure 3(b). When there is no air, the
CNTs are uniformly coated. In the presence of air, the bottom
CNTs are always fresh without coatings no matter how much
air (since the plots are similar for the case of air = 1 sccm
and 2 sccm, only one is shown in figure 4(b)). However, the
thickness of pyrolyzed carbon coated on upper CNTs decreases
with the increase of the air amount.
It should be noted that with the introduction of air the
optimum growth parameters become wider than that without
air. For example, the optimized values for T2 = 770 ◦ C
are: g0 = 100–400 sccm, gair = 2–10 sccm, CFerrocene =
3.3. Growth temperature effects
Figure 5(a) shows the data for the best CNT growth results
obtained at different temperatures. With growth temperature
increasing, more air was required to maintain the catalyst
activity. The lengthening–thickening process appears again at
830 ◦ C, though 10 sccm air has been injected. This could be
due to the fact that hydrocarbons tend to be easily pyrolyzed
at higher temperatures, and the pyrolyzed carbon encapsulates
the catalyst particles and terminates CNT growth.
The CNT growth rate can be determined from the slopes
of the plots in figure 5(a). The activation energy can be
estimated from the Arrhenius plot similarly to the reported
method [32]. The growth rate of CNTs is expressed by v =
A[C]sat exp(−E act /RT ), where A is a constant independent of
temperature, [C]sat is the saturated concentration of carbon in
γ -Fe, and E act is the activation energy. The Arrhenius plot of
ln(v/[C]sat ) versus 1/T is shown in figure 5(b). The linear
fit gives a value of E act , 145 ± 2 kJ mol−1 , which is close to
the data reported by other groups [31–33], suggesting that the
diffusion of C atoms in metal catalyst particles would be a ratedetermining step for CNT growth in the present system.
4
Nanotechnology 19 (2008) 455609
X Li et al
Figure 6. (a)–(c) SEM images of typical bundles used for two-terminal room temperature electrical transport. (d) Cyclic current–voltage
measurement of a long CNT bundle showing the I –V characteristics (at 8.75 V) and the breakdown (at 9 V). The inset in (d) shows a 40 μm
CNT bundle after breakdown. Note that the CNT bundle breaks in the middle and not at the contacts.
3.4. Failure current densities of long ACNT bundles
at high temperatures caused by Joule heating (see later), were
able to carry a nominal current density of ∼103 A cm−2 .
However, in a carbon nanotube bundle, the nanotubes occupy
only a fraction of the total volume, and this has to be taken
into account while calculating the actual current density in the
individual carbon nanotubes. Assuming that the nanotubes
occupy about 7% of the cross-sectional area (estimated using
a CNT lattice with 40 nm diameter CNTs having a separation
of 120 nm), we calculated that the individual tubes in these
long CNT bundles are capable of carrying current densities
∼105 A cm−2 before they break down. Typical resistivities (ρ )
of these bundles were found to be ∼0.8 μ cm (in comparison
the MWNT yarns have reported ρ of ∼0.185 μ cm). This
increase in resistivity could be due to residual oxidation of
the nanotubes due to the usage of air during the growth
process.
At low biases (V < 3V ), the I –V curves are approximately linear, indicating low-resistance Ohmic contacts, with
a resistance of RRT = 456.6 . With increasing bias, we
see that at around V ∼ 3 V the I –V curves begin to show
nonlinear behavior. We attribute this to the rise in temperature
of the bundles due to currents in excess of a few mA (ohmic
self-heating). The nonlinearity increases with the increase in
current, and the value of resistance just prior to breakdown
was noted as Rbd = 345.8 . By varying the sample
temperature (280 K < T < 300 K), we found that the
sample had a negative temperature coefficient of resistance
(TCR) of d R/dT ≈ −0.28 K−1 . Assuming that this
value remains more or less constant, the approximate sample
temperature could be estimated to be T = 296 K+(d R/dT )−1 .
(Rbd − RRT ) = 691.7 K = 418.6 ◦ C. Past works [38] on
thermogravimetric analysis (TGA) have shown that in multiwall carbon nanotubes, oxidation degradation indeed onsets at
T = 410 ◦ C. From this, we conclude that the breakdown seen
in our carbon nanotube bundles is triggered by the onset of oxidation of the tubes at T = 410 ◦ C, and perhaps is also further
assisted by electromigration damage of the oxidized carbon.
This is also consistent with the fact that the nanotube bundles
In many applications, multi-filament wires are preferable over
monofilament ones of the same diameter, due to improved
flexibility, mechanical strength, and fatigue characteristics.
In high-frequency electronics, this architecture distributes the
current over individual filaments, rather than conducting near
the surface of a single filament. The multi-filament architecture
obtained from the long ACNTs is superior to ‘ropes’ [34] or
‘fibers’ [35] reported before, since every individual carbon
nanotube runs end-to-end in the bundles. In the past, short
carbon nanotubes (up to a few microns in length) have been
demonstrated to be resilient against electromigration failure,
conducting high current densities [36] (up to 109 A cm−2 ) due
to their ballistic charge conduction process. However, these
experiments were mostly performed in vacuum and on short
(∼μm) nanotubes. At larger length-scales ( L mean free
path), scattering events give rise to finite resistances. At higher
current densities, this would give rise to significant heating of
the nanotubes which may lead to decomposition/breakdown
of the carbon nanotubes, especially if they are used in
applications in ambient conditions such as sensors and antenna
technologies. Hence, we have tested the failure current
densities of small bundles of these long MWNTs in ambient
conditions.
Figure 6(a) shows a typical bundle isolated from the parent
MWNT forest of diameter ∼40 μm. Figures 6(b) and (c)
show the structural integrity and alignment at the individual
nanotube level even after mechanical separation from the
parent forest. Figure 6(d) shows the breakdown cycle (in
air) of a typical device of length ∼1.5 mm, which failed at
about 9 V carrying about 25 mA of current. The inset is
an SEM image of the breakdown region in the middle of the
bundle. Past work has shown that such current-driven Jouleheating causes centimeter-long MWNT yarns to degrade and
finally break down in vacuum at nominal current densities
of ∼104 A cm−2 . [37]. In comparison, our MWNT bundles,
which break down in air due to a possible onset of oxidation
5
Nanotechnology 19 (2008) 455609
X Li et al
break down at the middle (inset, figure 6(d)) and not at the
contacts.
Using an approximate value of 50 000 nanotubes in our
bundle, the maximum current carried by each nanotube was
found to be ∼0.5 μA. We expect that in vacuum, this
value would be much higher. The average nanotube has a
resistance of ∼22.8 M. As mentioned before, this is due
to the diffusive nature of transport in the long MWNTs as
compared to the ballistic transport reported for short (few
100 nm) nanotubes, which have been shown to have quantum
resistances as low as h/4e2 ∼ 6.5 k. The Drude mean
free path (le ) for our individual nanotubes is given by le =
h L NT /(4e2 RNT ) ≈ 0.5 μm (here L NT is the effective length
of the nanotube device and RNT is its resistance), comparable
to typical values used in the literature [39]. Hence, in addition
to applications in intermediate and global-scale interconnects,
our ACNT bundles can have potential uses in lightweight,
low-cost, environment-friendly, flexible electronic applications
such as sensors [40] and antennae [41].
[5] Li H, Yin W Y, Banerjee K and Mao J F 2008 Circuit modeling
and performance analysis of multi-walled carbon nanotube
interconnects IEEE Trans. Electron. Devices 55 1328–37
[6] Pan Z W, Xie S S, Chang B H, Wang C Y, Lu L, Liu W,
Zhou W Y, Li W Z and Qian L X 1998 Very long carbon
nanotubes Nature 394 631–2
[7] Zhang X F, Cao a Y, Wei B Q, Li Y H, Wei J Q, Xu C L and
Wu D H 2002 Rapid growth of well-aligned carbon
nanotube arrays Chem. Phys. Lett. 362 285–90
[8] Zhu H W, Xu C L, Wu D H, Wei B Q, Vajtai R and
Ajayan P M 2002 Direct synthesis of long single-walled
carbon nanotube strands Science 296 884–6
[9] Huang S, Maynor B, Cai X and Liu J 2003 Ultralong,
well-aligned single-walled carbon nanotube architectures on
surfaces Adv. Mater. 15 1651–5
[10] Christen H M, Puretzky a A, Cui H, Belay K, Fleming P H,
Geohegan D B and Lowndes D H 2004 Rapid growth of
long, vertically aligned carbon nanotubes through efficient
catalyst optimization using metal film gradients Nano Lett.
4 1939–42
[11] Eres G, Puretzky a A, Geohegan D B and Cui H 2004 In situ
control of the catalyst efficiency in chemical vapor
deposition of vertically aligned carbon nanotubes on
predeposited metal catalyst films Appl. Phys. Lett.
84 1759–61
[12] Hata K, Futaba D N, Mizuno K, Namai T, Yumura M and
Iijima S 2004 Water-assisted highly efficient synthesis of
impurity-free single-walled carbon nanotubes Science
306 1362–4
[13] Zheng L X et al 2004 Ultralong single-wall carbon nanotubes
Nat. Mater. 3 673–6
[14] Hong B H, Lee J Y, Beetz T, Zhu Y, Kim P and Kim K S 2005
Quasi-continuous growth of ultralong carbon nanotube
arrays J. Am. Chem. Soc. 127 15336–7
[15] Zhang G et al 2005 Ultra-high-yield growth of vertical
single-walled carbon nanotubes: hidden roles of hydrogen
and oxygen Proc. Natl Acad. Sci. 102 16141–5
[16] Chakrabarti S, Nagasaka T, Yoshikawa Y, Pan L and
Nakayama Y 2006 Growth of super long aligned brush-like
carbon nanotubes Japan. J. Appl. Phys. 45 L720–2
[17] Hart A J and Slocum A H 2006 Rapid growth and
flow-mediated nucleation of millimeter-scale aligned carbon
nanotube structures from a thin-film catalyst J. Phys. Chem.
B 110 8250–7
[18] Xiong G Y, Wang D Z and Ren Z F 2006 Aligned
millimeter-long carbon nanotube arrays grown on single
crystal magnesia Carbon 44 969–73
[19] Li W Z, Xie S S, Qian L X, Chang B H, Zou B S, Zhou W Y,
Zhao R A and Wang G 1996 Large-scale synthesis of
aligned carbon nanotubes Science 274 1701–3
[20] Rao C N R, Sen R, Satishkumar B C and Govindaraj A 1998
Large aligned-nanotube bundles from ferrocene pyrolysis
Chem. Commun. 15 1525–6
[21] Andrews R, Jacques D, Rao a M, Derbyshire F, Qian D, Fan X,
Dickey E C and Chen J 1999 Continuous production of
aligned carbon nanotubes: a step closer to commercial
realization Chem. Phys. Lett. 303 467–74
[22] Li X, Ci L, Kar S, Soldano C, Kilpatrick S J and Ajayan P M
2007 Densified aligned carbon nanotube films via vapor
phase infiltration of carbon Carbon 45 847–51
[23] Singh C, Shaffer M S P and Windle a H 2003 Production of
controlled architectures of aligned carbon nanotubes by an
injection chemical vapour deposition method Carbon
41 359–68
[24] Zhu L, Hess D W and Wong C-P 2006 Monitoring carbon
nanotube growth by formation of nanotube stacks and
investigation of the diffusion-controlled kinetics J. Phys.
Chem. B 110 5445–9
[25] Li X, Cao A, Jung Y J, Vajtai R and Ajayan P M 2005
Bottom-up growth of carbon nanotube multilayers:
unprecedented growth Nano Lett. 5 1997–2000
4. Conclusion
In summary, in this work we have shown an air-assisted CVD
method for obtaining super-long CNTs and that the presence
of oxygen (from air) is crucial for maintaining catalyst activity.
This method can extend catalyst lifetime, and long CNT
bundles can be synthesized rapidly. The wide optimum
CVD growth window provides flexibility for the synthesis
of long CNTs. Their current carrying capabilities indicate
their potential use as components in advanced nanotechnology
based systems.
Acknowledgments
The authors (X Li, L Ci, S Kar, and P M Ajayan) acknowledge
financial support from the NSF Nanoscale Science and
Engineering Center for directed assembly of nanostructures
and the Rensselaer Interconnect Focus Center, New York.
S Talapatra acknowledges the financial support provided by the
Office of Research Development and Administration at SIUC
through faculty start-up funds and a seed grant.
References
[1] Jiang K L, Li Q Q and Fan S S 2002 Nanotechnology: spinning
continuous carbon nanotube yarns Nature 419 801
[2] Li Y L, Kinloch L A and Windle a H 2004 Direct spinning of
carbon nanotube fibers from chemical vapor deposition
synthesis Science 304 276–8
[3] Zhang M, Atkinson K R and Baughman R H 2004
Multifunctional carbon nanotube yarns by downsizing an
ancient technology Science 306 1358–61
[4] Zhang M, Fang S L, Zakhidov a A, Lee S B, Aliev a E,
Williams C D, Atkinson K R and Baughman R H 2005
Strong, transparent, multifunctional, carbon nanotube sheets
Science 309 1215–9
6
Nanotechnology 19 (2008) 455609
X Li et al
[26] Pinault M, Pichot V, Khodja H, Launois P, Reynaud C and
Mayne-L’hermite M 2005 Evidence of sequential lift in
growth of aligned multiwalled carbon nanotube multilayers
Nano Lett. 5 2394–8
[27] Zhu L, Xiu Y, Hess D W and Wong C-P 2005 Aligned carbon
nanotube stacks by water-assisted selective etching Nano
Lett.
5 2641–5
[28] Baker R T K 1989 Catalytic growth of carbon filaments Carbon
27 315–23
[29] Bower C, Zhou O, Zhu W, Zhu W, Werder D J, Jin S and
Jin S 2000 Nucleation and growth of carbon nanotubes by
microwave plasma chemical vapor deposition Appl. Phys.
Lett. 77 2767–9
[30] Li Y M, Kim W, Zhang Y G, Rolandi M, Wang D W and
Dai H J 2001 Growth of single-walled carbon nanotubes
from discrete catalytic nanoparticles of various sizes J. Phys.
Chem. B 105 11424–31
[31] Helveg S, Lopez-Cartes C, Sehested J, Hansen P L,
Clausen B S, Rostrup-Nielsen J R, Abild-Pedersen F and
Norskov J K 2004 Atomic-scale imaging of carbon nanofibre
growth Nature 427 426–9
[32] Kim K E, Kim K J, Jung W S, Bae S Y, Park J, Choi J and
Choo J 2005 Investigation on the temperature-dependent
growth rate of carbon nanotubes using chemical vapor
deposition of ferrocene and acetylene Chem. Phys. Lett.
401 459–64
[33] Baker R T K, Barber M A, Harris P S, Feates F S and
Waite R J 1972 Nucleation and growth of carbon deposits
from the nickel catalyzed decomposition of acetylene
J. Catal. 26 51–62
[34] Baughman R H, Zakhidov A A and Deheer W A 2002 Carbon
nanotubes–the route toward applications Science
297 787–92
[35] Harris P J F 2004 Carbon nanotube composites Int. Mater.
49 31–43
[36] Wei B Q, Vajtai R and Ajayan P M 2001 Reliability and current
carrying capacity of carbon nanotubes Appl. Phys. Lett.
79 1172–4
[37] Wei Y, Jiang K L, Liu L, Chen Z and Fan S S 2007
Vacuum-breakdown-induced needle-shaped ends of
multiwalled carbon nanotube yarns and their field emission
applications Nano Lett. 7 3792–7
[38] Chena C-M, Chenb M, Leub F-C, Hsub S-Y, Wangb S-C,
Shia S-C and Chena C-F 2004 Purification of multi-walled
carbon nanotubes by microwave digestion method Diamond
Relat. Mater. 13 1182–6
[39] Cho H, Koo K-H, Kapur P and Saraswat K C 2008
Performance comparisons between Cu/low-κ ,
carbon-nanotube, and optics for future on-chip interconnects
IEEE Electron Device Lett. 29 122
[40] Zhao L, Choi M, Kim H S and Hong S H 2007 The effect of
multiwalled carbon nanotube doping on the CO gas
sensitivity of SnO2-based nanomaterials Nanotechnology
18 445501
[41] Wang Y, Kempa K, Kimball B, Carlson J B, Benham G, Li W
Z, Kempa T, Rybczynski J, Herczynski A and Ren Z F 2004
Receiving and transmitting light-like radio waves: Antenna
effect in arrays of aligned carbon nanotubes Appl. Phys. Lett.
85 2607–9
7