Carbon nanotube in

different shapes
Carbon nanotubes (CNTs) have been well studied theoretically and
experimentally. Perfect CNTs have a crystalline structure formed by
hexagonal network; defects cause the tubule to curve. CNTs with
different tubule morphologies have their own special properties and
potential applications. So far, many different shapes, such as straight,
waved, coiled, and branched, are predicted, observed, and target
synthesized. This article reviews CNTs in different shapes formed during
growth, their morphologies and their possible applications.

Mei Zhang* and Jian Li

High-Performance Materials Institute, Florida State University; Department of Industrial and Manufacturing Engineering, FAMU-FSU College of
Engineering, Tallahassee, FL 32310, USA
*E-mail: mzhang@eng.fsu.edu

Carbon Nanotubes (CNTs) are cylindrical shells made, in concept, Iijima’s report4 in 1991 brought carbon nanotubes into the awareness
by rolling graphene sheets into a seamless cylinder. CNTs exist as of the scientific community and triggered a deluge of interest in carbon
either single-wall nanotubes (SWNTs) or multi-walled nanotubes nanotubes. Shortly after this “rediscovery”, CNTs in different shapes
(MWNTs). The SWNT consists of a single graphene sheet, which (toroidal5, coiled6, 7, and branched8-10) other than the straight were
is a planar array of benzene molecules, involving only hexagonal predicted theoretically. Figs. 1f-g show the illustrations of bent (kink)11,
rings with double and single carbon-carbon bonding. The choice branched12, and coiled5 SWNT structures. All these structures are
of rolling axis relative to the hexagonal network of the graphene based on the insertion of non-hexagonal defects in seamless hexagonal
sheet and the radius of the closing cylinder allows for different networks. In particular, the models of the regular helical coils of CNT’s
types of SWNTs, which vary from insulating to conducting1. are based on a very specific arrangement of pentagons and heptagons
Figures 1a, 1b, and 1c show SWNTs of three different types2: in a perfect hexagonal lattice7, 13. If the regular arrangement is perturbed
armchair, zigzag, and chiral. The twist of the chiral nanotube is by misplacing one single non-hexagonal ring, the structure will not be a
clearly evident in the lower Fig. 1c, a perspective view along the regular coil any more13. The model also shows that the regular helically
tube axis, and in Fig. 1d, a scanning probe microscope (SPM) picture coiled nanotubes can be built with non-hexagonal/hexagonal ratio
of a chiral SWNT3. MWNTs comprise an array of such nanotubes higher than unity14, 15.
that are concentrically nested. A transmission electron microscope Experimentally, CNTs are formed during synthesis. In most cases,
(TEM) image of a nine-walled CNT is shown in Fig. 1e. CNTs with CNT grows away from the catalyst particle by the deposition of carbon
perfect crystalline structure are straight cylinders. in the contact region between the particle and the already formed

12 JUNE 2009 | VOLUME 12 | NUMBER 6 ISSN:1369 7021 © Elsevier Ltd 2009 Open access under CC BY-NC-ND license.
 Carbon nanotube in different shapes REVIEW

(a) (b) (c) (d) (e)

(a) (b)

(c) (d)

(f) (g) (h)

(e) (f)

Fig. 1 Schematic illustrations of the structures of (a) armchair, (b) zigzag, and
(c) chiral SWNTs. Projectors normal to the tube axis and perspective views
along the tube axis are on the top and bottom, respectively. (d) SPM picture
of a 1.3 nm diameter chiral SWNT.3 (e) TEM image of a MWNT containing a
concentrically nested array of nine SWNTs. (f) illustration of the carbon-bone Fig. 2 (a) Scanning electron microscope (SEM) image of SWNT cylindrical pillars
network of a kink junction constructed between an ‘armchair’ tube and a ‘zigzag’ with 150 μm radius, 250 μm pitch, and ~1 mm height. Insert, SEN image of a
tube, where 5 denotes a pentagon, 7 denotes a heptagon, and the atoms in the root of a pillar. (b) SEM images of SWNT sheets 10 μm thick. (c) High-resolution
pentagon and heptagon are highlighted by dark balls. (g) Structural model of TEM image of the SWNTs. (d) TEM image of a bundle consists of collapsed
SWNT asymmetric, zig-zag Y junction. n-H rings highlighted in dark. (h) Scheme single-wall, double-wall, and triple-wall nanotubes with big diameters. Note
of the helically coiled SWNT (Helix C360) that have the lowest cohesive energy both the “dog-bone” cross section of these tubes and the non-collapsed tubes
per atom. (Parts (a), (b), (c), and (e) are adapted and reprinted with permission at the edge of the bundle. (e) and (f) SEM images of arrays of almost exclusively
from2. © 2002 AAAS. Part (d) is adapted and reprinted with permission from3. semiconducting SWNTs at different magnifications. The bright and parallel
© 1998 Nature Publishing Group. Part (f) is adapted and reprinted with horizontal lines visible in the images are catalyst lines. (Parts (a), (b), and (c) are
permission from11. © 1999 Nature Publishing Group. Part (g) is adapted and adapted and reprinted with permission from28. © 2004 AAAS. Part (d) is adapted
reprinted with permission from12. © 2001 AIP. Part (h) is adapted and reprinted and reprinted with permission from29. © 2007 WILEY. Parts (e) and (f) are
with permission from5. © 1993 APS.) adapted and reprinted with permission from48. © 2009 ACS)
tubule segment16. To date, a large variety of tubule morphologies have
been observed and synthesized. They can be classified into the following of aligned arrays in catalytic CVD has been demonstrated and is
categories: straight, waved, coiled, regularly bent, branched, and beaded. reviewed26. There are two major breakthroughs in the synthesis of CNT
arrays. In 1999, Fan et al.27 used porous silicon substrates with a catalyst
Straight carbon nanotubes patterned by electron-beam evaporation through shadow masks to
A variety of techniques have been developed to produce CNTs. MWNTs produce MWNT blocks that grew perpendicular to the substrate. Five
were synthesized and observed a few decades ago17, 18. In 1993, years later, Hata et al.28 first successfully produced millimeter high
Iijima et al.19 and Bethune et al.20 independently reported the synthesis SWNT arrays using water assisted CVD (Fig. 2a-c). It is observed that
of SWNTs. Primary synthesis methods for single-wall and multi-walled the SWNTs or double-wall CNTs easily collapse, generating stacks of
CNTs include arc discharge4, 19-21, laser ablation22, 23, gas-phase catalytic parallel graphene layers, when their diameters are larger than ~5 nm
growth from carbon monoxide24, and chemical vapor deposition (CVD) (Fig. 2d)29.
from hydrocarbons25. Synthesis methods such as arc discharge, laser Horizontally well-aligned CNT arrays on suitable substrates are
ablation, and certain types of CVD with floating catalysts produce highly desired for SWNT device applications, such as field effect
nonaligned, entangled ropes of nanotubes. The diameter of a CNT transistors30, sensors31, light emitters32, logic circuits33, and other
is in the order of nanometers, while its length can be up to several systems. The most attractive approach is direct growth34 by CVD with
millimeters. Long CNTs usually can remain straight when they are in an external force. The external forces can originate from an electric
oriented structures. field35, the gas flow36, 37, or interactions with the substrate surface38-44.
Vertical alignment conventionally means that CNTs are oriented Among them, the surface-guided growth on single crystal substrates
perpendicular to the substrate. A variety of methods for the production such as sapphire38 or quartz39-41 provides high density and perfect

JUNE 2009 | VOLUME 12 | NUMBER 6 13

REVIEW Carbon nanotube in different shapes

alignment45. It shows that uniform nanotube arrays fabricated on single illustrated in Fig. 1f or a local mechanical deformation in a uniform
crystal substrates can be used directly as a thin film for making a large nanotube. A nanotube elastically deforms under a small bending stress,
amount of devices45-47. It has been demonstrated that a large variety and buckles if the local curvature exceeds a critical value49, 50. During
of metals (including Fe, Co, Ni, Cu, Pt, Pd, Mn, Mo, Cr, Sn, Au, Mg, and growth, the bending stress can come from the nanotube’s own weight,
Al) can catalyze SWNTs growth. They all show horizontally aligned interaction with neighbor nanotubes, or limited growing space.
lattice growth of CNTs on quartz substrates under the same growth A group of CNTs can form a randomly oriented CNT mat or super-
conditions43. A hypothesis is proposed in which the precipitated carbon aligned CNT arrays, depending on the density of catalyst and their
shell on the outer surface of the metal catalysts guides the alignment activities under the same synthesis conditions. Figs. 3a-b show a
along the crystal lattice but not the catalysts themselves. Recently, MWNT thin sheets array. These ~100 μm height sheets were grown
high-density arrays of horizontally aligned semiconducting SWNTs from 0.2 μm wide and 40 μm long catalyst thin films patterned by
were successfully grown on ST-cut single crystal quartz substrates48. As e-beam lithography. There were no external forces during sheets
shown in Figs. 2e-f, the degree of alignment and linearity in the SWNT growth. The sheets bend when their height is over a certain level. The
arrays is extremely high. bending directions and angles depend on each sheet’s morphologies.
The nanotubes within each sheet confine the nearest neighbors and
Waved carbon nanotubes attract the outermost nanotubes to their neighbors via van der Waals
A single nanotube naturally curves (in bending status) during growth force, thereby producing oriented growth. However, the CNTs in such
if no external forces exist. In principle, the CNT bending (kinks) thin sheets present random curvatures and are tangled (Fig. 3c) because
can originate from a pentagon-heptagon topological defect pair as of the weak confinement in thickness direction. As the thickness of the

(a) (b) (c)

50 μm 30 μm 1 μm
m

(d) (e) (f)

200 μm 0 2 μm
0.2 0 2 μm
0.2

(g) (h) (i)

20 μm 2μm 1 μm

Fig. 3 (a) and (b) SEM images of CNT thin sheets array. The ~100 μm height sheets were grown from 0.2 μm wide and 40 μm long catalyst thin films. (c) CNTs in the
thin sheet. (d) and (e) super-aligned MWNT array. (f) Straight and waved MWNTs in an array. (g) to (i) SEM images of a MWNT array with wavy structure at different
magnifications.

14 JUNE 2009 | VOLUME 12 | NUMBER 6

 Carbon nanotube in different shapes REVIEW

sheet increases, the alignment of the CNTs could be improved due to (a)
the crowding effect. Fig. 3d shows a block of MWNT array, in which
nanotubes are super-aligned (Fig. 3e).
Fig. 3g-i show a special CNT array, in which more than 80% of the
CNTs are not straight; they periodically bend within fixed intervals
throughout their entire length. As a result of this regular bending,
a wavy structure is formed. It is believed that the wavy structure is
formed because there are roughly two groups of catalysts uniformly
distributed on the substrate: one is more active and results in higher
CNT growth rate than the other. Due to van der Walls force, which
sticks nanotubes together whenever they touch, the growth rate of
the array is limited by the nanotubes with the relatively slow growth 20 μm
rate when catalysts stay on the surface of the substrate. The nanotubes
with higher growth rate are forced to bend periodically. The period of (b) (c)
the wavy is related with the ratio of growth rates of these two groups.
When the distribution of the catalyst activity is broad but the density
is high, the array will have the morphology as shown in Fig. 3f. The
straight CNTs bundled while the waved CNTs switch between different
bundles. Such structure is believed important for assembling CNT sheets
or yarns by drawing CNTs directly from the array51-53.
(d)

Coiled carbon nanotubes

Coiled CNTs were predicted in the early 1990s6, 54. These CNTs are
created when paired pentagon-heptagon atomic rings arrange themselves
periodically within the hexagonal carbon network55. Theoretical
Fig. 4 Coiled CNTs. (a) SEM image of large amount of helically coiled CNTs61.
calculation also predicted that various forms of helically coiled structures Each coil grows with its own diameter and pitch. (b) The tip of a coil. (c)
are possible and those structures are energetically and thermodynamically TEM image of a coil formed by two tubules with the same pitch but different
diameters and a slight shift in phase. (d) TEM image of typical coiled CNTs.
stable7. Helical structures of carbon have been observed as early as
(Part (d) is adapted and reprinted with permission from62. © 2003 AIP.)
1950s56. Coiled CNTs were observed experimentally in 199416.
Fig. 4a shows coiled CNTs grown on iron-coated indium tin oxide Coiled CNTs are attracting because their peculiar morphology
substrate by catalytic CVD57. More than 95% of CNTs are in helical enables them to be used as high-performance electromagnetic wave
structures. The coils have various diameters and pitches. They grow absorbers, sensors, resonators, nanoscale mechanical springs, electrical
out of the substrate and maintain their self-organization well during inductors, and generators of magnetic beams. For applications, it would
growth. It is interesting to note that each coil grows with its own pitch be desirable to have control over the coil morphology and geometry
and diameter. Later investigations evidenced that indium and tin play (diameter, pitch, length, and turning direction).
an important role in the formation of coiled CNTs58, 59. The double and
triple helices are often observed too (Fig. 4c)57, 60. So far, synthesized Regularly bent carbon nanotubes
coiled CNTs are MWNTs with more or less incomplete crystalline Regularly bent CNTs are proposed for applications including mechanical
structures. Coiled SWNTs have not been recorded. nanospring devices, high-resolution AFM tips, and nanocircuit
Several mechanisms were hypothesized for the formation of coiled interconnections in device. It is known that CNTs can be aligned during
CNTs. Regular insertion of pentagon-heptagon pairs at the junction, growth with an external force originated from an electric field35, the
as predicted theoretically, is for CNTs without other crystallographic gas flow36, 37, or interactions with the substrate surface38, 67. Properly
defects. The model of localized stresses and anisotropic rates of combining those external forces is a way to make regularly bent CNTs.
carbon deposition57, 61, 62 on catalyst particles is most widely accepted. Fig. 5a shows a zigzag shaped SWNT, which is formed due to the SWNT-
Recently, a thermodynamic model63 was proposed for providing a substrate lattice interaction and the gas flow67.
comprehensive explanation, where helix/coil formation is explained Arrays of CNTs with zigzag morphology are shown in Fig. 5b. The
on the basis of the interactions between specific catalyst particles and zigzag morphologies consist of 2-4 very sharp and alternating ~90°
growing nanostructures. Coiled CNTs can be produced at high yield bends. They were grown using a dc plasma enhanced CVD process68.
by catalytic CVD16, 57, 61, 64-66. The bending of the CNTs during growth was caused by changing the

JUNE 2009 | VOLUME 12 | NUMBER 6 15

REVIEW Carbon nanotube in different shapes

(a) (b) (a)

(c)

(b)

10 μm
μm

Fig. 5 (a) SEM image of a self-organized SWNT serpentines formed due to the
combined alignment effects from the quartz substrate and gas flow. (b) Array
of CNTs grown with zigzag morphology using a three-stage growth process.
Sample tilted 45° for SEM analysis. (c) A small group of MWNTs forms a coil-like
structure by self assembly. (Part (a) is adapted and reprinted with permission
from67. © 2007 ACS. Part (b) is adapted and reprinted with permission from68.
© 2004 ACS.)

direction of the electric field lines in the growth region of the sample. It
is anticipated that such a sharp bend in a nanotube is likely to contain
many defects, not only pentagon and heptagon defects. To utilize Fig. 6 (a) SEM image of branched CNTs. (b) TEM images of a typical
Y-junction SWNT. (Part (a) is adapted and reprinted with permission
CNTs as interconnectors and other device components, the ability to from88. © 2005 Elsevier. Part (b) is adapted and reprinted with permission
control their growth morphology is desired, especially if SWNTs or from90. © 2005 Elsevier.)
small diameter MWNTs can also be made to respond to electrical field
manipulations and bend in a similar fashion. CVD80-88. Several processes and mechanisms were presented. A
Self assembly can form some novel structures. Fig. 5c records a small single-particle process83, 84, 88 attributed the formation of branched
group of CNTs that bundle together, regularly bend during growth, and CNTs to fluctuation in temperature, gas flow, or carbon source
result in a coil like structures. feeding, which could change the distribution of carbon atoms on
the catalyst particles and alter the growth direction to form a
Branched carbon nanotubes multi-stem CNT88 (Fig. 6a). A splitting process 85, 86 posited that the
The first structural models69, 70 for CNT Y-junctions are based on the liquid-phase metal-carbon solution encapsulated in a CNT could split
insertion of non-hexagonal rings in the hexagonal network in the into smaller particles and lead to the growth of branched CNTs. A
region where the three branches of the Y are joined together. All the merging process81, 87 proposed that catalyst particles encapsulated in
subsequent structural models71-73 follow the same construction principle two neighboring CNTs could weld together to form a larger catalyst
of conserving the sp2 hybridization of the carbon network, differing particle which then catalyzes the growth of a third CNT branch. Catalyst
only in the kind, number, and placement of the non-hexagonal rings. particles attached to the CNT sidewall could also result in branched
These variations make possible the constructions of various symmetric growth, and the doping metals (Ti89 or Mo90) play important roles in
and asymmetric model junctions74, 75 and various angles from Y to T the attachment process. A multibranching CNT structure synthesized by
shapes12, 71. Since the electrical properties of CNTs are dependent on dc plasma enhanced CVD has been reported 91. The structure consists of
the tube structure (chirality and diameter), various combinations of aligned CNTs which have branches of smaller diameters growing aligned
metallic and semiconducting tubes can be built to form CNT junctions along a direction perpendicular to the original CNT. Most of these
as building parts for nanoscale integrated circuits76, 77. processes generate branched MWNTs, some of them are with poor
Branched nanotubes with T, Y, L, and more complex junctions crystallographic structures.
were initially observed in arc-discharge produced nanotubes78. The Branched SWNTs are believed to have a profound impact on next-
first synthesis was reported in 1999 by pyrolysis of acetylene in generation electronic devices since they have a potential to be used
Y-shaped templates79. Since then, most work has focused on catalytic in nano-electronic devices as nano-diode, nano-transistor, and nano-

16 JUNE 2009 | VOLUME 12 | NUMBER 6

 Carbon nanotube in different shapes REVIEW

(e)

(a)
(f) (g)
(b)

(c)

(h)

(d)

Fig. 7 Various beaded CNTs. (a) TEM image of a MWNT with many beads. (c) TEM image of a small, elongated bead on a 15-layer MWNT. (b) and (d) High-resolution
TEM images of the furthest extent of the meniscus on the left side of the bead and the bead meniscus to the right. (e-h) SEM images of beads in different shapes95:
sphere (e), cubic (f), polyhedral (g), and bloom-shaped beads (h). (Parts (a) to (d) are adapted and reprinted with permission from96. © 2005 AAAS.)

interconnect. The electronic properties of SWNT Y-junctions have evidence for graphitic layering in the bead, nor does the bead distort
been modeled theoretically92 and measured experimentally93, 94. They the nanotube.
show great potential as elements in simple nanoelectronic devices. The beaded CNTs also formed in relatively low temperature
Y-junction SWNTs have been successfully synthesized using controlled (~700 °C) CVD process95. Various shapes, such as spherical, cubic,
catalysts by thermal CVD90, where Mo-doped Fe nanoparticles other polyhedron, and bloom shaped structures, were observed
supported by aluminum oxide particles are used as catalysts. It is (Fig. 7e-h). The dense packaged beads form thick-stick like structures
found that distribution of Mo-doped Fe particles plays an important around the CNT. It is believed that those chaplets are formed during
role in Y-junction formation. Fig. 6b shows a typical TEM image of nanotube growth and the shape of the beads is related to catalyst-
SWNT Y-junction with diameters of 2-5 nm. included seeds95.
Beaded CNTs are expected as fillers in composites to enhance
Carbon nanotubes with beads the electric conductivity and/or mechanical properties of the matrix
CNTs with beads were observed from different processes18, 95-98. Beads materials because the beads on CNTs prevent slipping of the nanotubes
appear in various patterns and their structures are either amorphous or in the composite materials.
polycrystalline graphite. The beads form either with18, 95 or after CNT
formation96, 97 during synthesis. CNT’s crystallographic structures may Summary
not be disturbed in some cases. CNTs have various tubule morphologies and appear in different shapes.
Fig. 7a shows a MWNT with 100 to 200 nm diameter beads. These Each shape has its own special properties and potential applications. CNTs
beaded nanotubes are occasionally found on the surface of the columns in different shapes will have significant impacts only when they can be
harvested from the interior of the arc deposition96. The beads are carbon produced with uniform properties in large quantities. Precise control of a
glass (amorphous phase). They formed on nanotube because carbon CNT’s morphology at specific positions by synthesis is still challenge.
coated on the nanotube was a viscous liquid and cooling caused the
viscosity to increase to a degree that the beading process stagnated. Acknowledgments
The nanotube is clearly visible inside the bead (Fig. 7b-d). There is no This work was supported by the Air Force Research Laboratory.

JUNE 2009 | VOLUME 12 | NUMBER 6 17

REVIEW Carbon nanotube in different shapes

REFERENCES 49 Iijima, S., et al., J. Chem. Phys. (1996) 104, 2089)

1 Louie, S. G., Top. Appl. Phys. (2001) 80, 113 50 Yakobson, B. I., et al., Phys. Rev. Lett. (1996) 76, 2511
2 Baughman, R. H., et al., Science (2002) 297, 787 51 Zhang, M., et al., Science (2004) 306, 1358
3 Wildoer, J. W. G., et al., Nature (1998) 391, 59 52 Zhang, M., et al., Science (2005) 309, 1215
4 Iijima, S., Nature (1991) 354, 56 53 Zhang, X., et al., Adv. Mater. (2006) 18, 1505
5 Ihara, S., et al., Phys. Rev. B (1993) 47, 12908 54 Itoh, S., et al., Phys. Rev. B (1993) 48, 8323
6 Dunlap, B. I., et al., Phys. Rev. B (1992) 46, 1933 55 Gao, R. P., et al., J. Phys. Chem. B (2000) 104, 1227
7 Ihara, S., et al., Phys. Rev. B (1993) 48, 5643 56 Davis, W. R., et al., Nature (1953) 171, 756
8 Macky, A. L., et al., Nature (1991) 352, 762 57 Zhang, M., et al., Jpn. J. Appl. Phys. (2000) 39, L1242
9 Scuseria, G. E., Chem. Phys. Lett. (1992) 195, 534 58 Wang, W., Adv. Mater. (2008) 20, 179
10 Chernazatonskii, L. A., Phys. Lett. A (1992) 172, 173 59 Pan, L., et al., J. Appl. Phys. (2002) 91, 10058
11 Yao, Z., et al., Nature (1999) 402, 273 60 Su, C. J., et al., Phys.Chem. Comm. (2002) 5, 34
12 Andriotis, A. N., et al., Appl. Phys. Lett. (2001) 79, 266 61 Ivanov, V., et al., Chem. Phys. Lett. (1994) 223, 329
13 Dunlap, B. I., et al., Phys. Rev. B (1994) 49, 5643 62 Zhong, D. Y., et al., Appl. Phys. Lett. (2003) 83, 4423
14 Biro, L. P., et al., Phys. Rev. B (2002) 66, 165405 63 Bandaru, P. R., et al., J. Appl. Phys. (2007) 101, 094307
15 Lambin, Ph., Phys. Rev. B (2003) 67, 205413 64 Hou, H., et al., Chem. Mater. (2003) 15, 3170
16 Amelinckx, S., et al., Science (1994) 256, 635 65 Lu, M., et al., J. Phys. Chem. B (2004) 108, 6186
17 Radushkevich, L. V., and Lukyanovich, V. M., Zurn. Fisc. Chim. 66 Wang, J. N., et al., Cryst. Growth Des. (2008) 8, 1741
(1952) 26, 88 67 Kocabas, C., et al., J. Phys. Chem. C (2007) 111, 17879
18 Oberlin, A., et al., J. Cryst. Growth (1976) 32, 335 68 AuBuchon, J. F., et al., Nano Lett., (2004) 4, 1781
19 Iijima, S., and Ichihashi, T., Nature (1993) 363, 603 69 Scuseria, G. E., Chem. Phys. Lett. (1992) 195, 534.
20 Bethune, D. S., et al., Nature (1993) 363, 605 70 Chernozatonskii, L. A., Phys. Lett. A (1992) 172, 173
21 Journet, C., et al., Nature (1997) 388, 756 71 Menon, M., and Srivastava, D., Phys. Rev. Lett. (1997) 79, 4453.
22 Rinzler, A. G., et al., Appl. Phys. A (1998) 67, 29 72 Treboux, G., et al., Chem. Phys. Lett. (1999) 306, 402.
23 Guo, T., et al., Chem. Phys. Lett. (1995) 243, 49 73 Meunier, V., et al., Appl. Phys. Lett. (2002) 81, 5234
24 Nikolaev, P., et al., Chem. Phys. Lett. (1999) 313, 91 74 Andriotis, A. N., et al., Phys. Rev. B (2002) 65, 165416
25 Ren, Z. F., et al., Science (1998) 282, 1105 75 Xue, B., et al., Comput. Mater. Sci. (2008) 43, 531
26 Huczko, A., Appl. Phys. A: Mater. Sci. Process (2002) 74, 617 76 Popov, V. N., Mater. Sci. Eng. R (2004) 43, 61
27 Fan, S. S., et al., Science (1999) 283, 512 77 Dunlap, B. I., Phys. Rev. B (1992) 46, 1933
28 Hata, K., et al., Science (2004) 306, 1362 78 Zhou, D., and Seraphin, S., Chem. Phys. Lett. (1995) 238, 286
29 Motta, M., et al. Adv. Mater. (2007) 19, 3721 79 Li, J., et al., Nature (1999) 402, 253
30 Tans, S. J., et al., Nature (1998) 393, 49 80 Satishkumar, B. C., et al., Appl. Phys. Lett. (2000) 77, 2530
31 Kong, J., et al., Science (2000) 287, 622 81 Wei, D. C., et al., Nano Lett., (2006) 6, 186
32 Misewich, J. A., et al., Science (2003) 300, 783 82 Luo, C., et al. Carbon (2008) 46, 440
33 Bachtold, A., et al., Science (2001) 294, 1317 83 Su, L. F., et al., Chem. Vapor Depos. (2005) 11, 351
34 Kong, J., et al., Nature (1998) 395, 878 84 Tao, X. Y., et al., Chem. Phys. Lett. (2005) 409, 89
35 Ural, A., et al., Appl. Phys. Lett. (2002) 81, 3464 85 Huang, S. M., et al., Phyica B (2002) 323, 336
36 Huang, S. M., et al., J. Am. Chem. Soc. (2003) 125, 5636 86 Ding, D. Y., et al., Appl. Phys. A (2005) 81, 805
37 Zhong, J., et al., Nano Lett. (2007) 7, 2073 87 Lee, T. J., et al., Carbon (2005) 43, 1341
38 Han, S., et al., J. Am. Chem. Soc. (2005) 127, 5294 88 Heyning, O. T., Chem. Phys. Lett. (2005) 409, 43
39 Ismach, A., et al., Angew. Chem., Int. Ed. (2004) 43, 6140 89 Gothard, N., et al., Nano Lett. (2004) 4, 213
40 Kocabas, C., et al., Small (2005) 1, 1110 90 Choi, Y. C., and Choi, W., Carbon (2005) 43, 2737
41 Hur, S. H., et al., J. Appl. Phys. (2005) 98, 114302. 91 AuBuchon, J. F., et al., Nano Lett. (2006) 6, 324
42 Yoshihara, N., et al., J. Phys. Chem. C (2009) 113, 8030 92 Treboux, G. J., Phys. Chem. B (1999) 103, 10378
43 Yuan, D. et al., Nano Lett. (2008) 8, 2576 93 Bandaru, P. R., J. Mater. Sci. (2007) 42, 1809
44 Zhou, W., et al., J. Nano Res. (2008) 1, 158 94 Papadoupoulos, C., et al., Phys. Rev. Lett. (2000) 85, 3476
45 Kang, S. J., et al., Nat. Nanotechnol. (2007) 2, 230 95 Nakayama, Y., and Zhang, M., Jpn. J. Appl. Phys. (2001) 40, L492
46 Ding, L., et al., J. Am. Chem. Soc. (2008) 130, 5428 96 de Heer, W. A., et al., Science (2005) 307, 907
47 Orofeo, C. M., et al., Appl. Phys. Lett. (2009) 94, 053113 97 Monthioux, M., et al., Carbon (2006) 44, 3183
48 Ding, L., et al., Nano Lett., (2009) 9, 800 98 Ting, J. M., and Lan, J. B. C., Appl. Phys. Lett. (1999) 75, 3309

18 JUNE 2009 | VOLUME 12 | NUMBER 6