Carbon Nanomaterials

Structure of Single-Wall Carbon Nanotubes:

A Graphene Helix
Jae-Kap Lee,* Sohyung Lee, Jin-Gyu Kim, Bong-Ki Min, Yong-Il Kim, Kyung-Il Lee,
Kay Hyeok An, and Phillip John

Evidence is presented in this paper that certain single-wall carbon nanotubes are not
seamless tubes, but rather adopt a graphene helix resulting from the spiral growth of
a nano-graphene ribbon. The residual traces of the helices are confirmed by highresolution transmission electron microscopy and atomic force microscopy. The
analysis also shows that the tubular graphene material may exhibit a unique armchair
structure and the chirality is not a necessary condition for the growth of carbon
nanotubes. The description of the structure of the helical carbon nanomaterials is
generalized using the plane indices of hexagonal space groups instead of using chiral
vectors. It is also proposed that the growth model, via a graphene helix, results in a
ubiquitous structure of single-wall carbon nanotubes.

Dr. J.-K. Lee, S. Lee

Interface Control Research Center
Korea Institute of Science and Technology (KIST)
Seoul 130650, Korea
E-mail: jklee@kist.re.kr
S. Lee
Department of Semiconductor Science
Dongguk University
Seoul 100715, Korea
Dr. J.-G. Kim
Division of Electron Microscopic Research
Korea Basic Science Institute
Daejeon 305333, Korea
Dr. B.-K. Min
Instrumental Analysis Center
Yeungnam University
Daegu 712749, Korea
Dr. Y.-I. Kim
Korea Research Institute of Standards and Science
Daejeon 305600, Korea
Dr. K.-I. Lee
Center for Spintronics Research
KIST
Dr. K. H. An
Nano Material Research Department
Jeonju Machinery Research Center
Jeonju 561844, Korea
Prof. P. John
School of Engineering and Physical Sciences
Heriot-Watt University
Riccarton, Edinburgh, EH14 4AS, UK
DOI: 10.1002/smll.201400884
small 2014,
DOI: 10.1002/smll.201400884

1. Introduction
Single-wall carbon nanotubes (SWNTs) have raised considerable interest worldwide[1,2] because of their unique
shape and the resulting versatile and unique properties.
Numerous studies have been performed since they were
first reported.[3,4] Previous work has been mainly focused
on understanding the structure of SWNTs and the properties of this scientifically and technically important material.
The deceptively simple structural model of SWNTs, that is,
a seamless concentric tube has been virtually universally
adopted.[24] The model was based on the currently accepted
structure of turbostratic stacking of seamless concentric graphene sheets[5] of multi-wall carbon nanotubes (MWNTs).
Recently, however, we have shown that MWNTs are composed of graphene helices with widths of approximately five
nanometres.[6] The structure of MWNTs was attributed to the
helical growth of the AA stacked graphene ribbons as the
minimum energy configuration in contrast to circular growth
at the tube edges. This suggests that the growth of a curved
sheet of graphene, that is, SWNTs, may occur along the minimum energy pathway. It is questionable whether the growth
of seamless narrow diameter cylindrical SWNTs is energetically possible given the high level of stress[7] and propensity to form lower energy structures. In this study, we have
analyzed the structure of commercial SWNTs, in the context of the previous experimental evidence in the literature.
In contrast to the conventional view we show that SWNTs
are tubular graphene in the unique armchair configuration,

2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

full papers

J.-K. Lee et al.

in Figure 2. The line scans along the tubule axis (Figure 2d)
reveal the trace of the nodal morphology.
The nodal morphology of SWNTs has been evident in
previous HRTEM[4,912] and scanning tunnelling microscopy (STM)[13] images. Particularly, the digitalized HRTEM
images of Meyer et al.[9] and Suenaga et al.[12] show atomically resolved nodal morphologies of SWNTs where the
traces of graphene helices are observable (Supporting Information Figure S1). Other morphological evidence has been
provided by the work of Kiang et al.[14] who showed, via
HRTEM images, that SWNTs were twisted and folded but
not fractured. Such twisting and folding of a perfect tube
are impossible without fracture since the three-dimensional
structures of SWNTs have acceptable strains of only 25%
in both the zigzag and armchair structures.[15] On the other
hand, a tubule comprising graphene ribbons can be twisted or
folded without fracture. The thermal conversion of the bun2. Results and Discussion
dles of SWNTs into graphitic ribbons (2200 C under a high
Figure 1 shows typical high-resolution transmission elec- vacuum of 105 Torr where graphite is stable at 3000 C)
tron microscopy (HRTEM) images of bundles of purified reported by Gutirrez et al.[16] provides indirect evidence
SWNTs synthesized by an arc-discharge technique. At rela- that SWNTs are composed of graphene ribbons. Morphotively low magnification (observed by JEM-2100F operating logical evidence from the literature[3,4,1012,14] indicates that
at 200 kV), the contours of SWNTs are observable and a graphene ribbons are common structural building blocks
node-like morphology appears as shown in Figure 1a. Under of SWNTs. The direct observation of the rotational aspects
high magnification (observed by FEI Titan Cubed with aber- of the growth of SWNTs, via a field emission microscopy
ration corrector and a monochromater operating at 80 kV), reported by Marchand et al.,[17] supports the spiral model of
Figure 1b, wavy and disconnected atomic lattice fringes, which a graphene ribbon growth although the authors explained the
correspond to the nodal tubules, are evident (Figure 1b and phenomenon in terms of a screw-dislocation where an atom
b). The profile of the nodal morphology is evident in the is the unit of the screw growth forming a concentric tube. But,
atomic force microscopy (AFM) image of a SWNT as shown this model cannot explain the prevailing nodal morphology
of SWNTs. Furthermore, helical growth of
graphene ribbons reduces the conformational energy from 0.85 to 0.22 eV/atom
for a 0.4 nm diameter armchair SWNT as
shown in Figure 3.
Graphene is a two-dimensional net
that appears morphologically as a layer of
honeycombed sp2 carbon. We can describe
the structure of graphene using the space
group for the simple hexagonal #191 for
AA graphite wherein c is 0 and the other
structural factors are the same as those of
AA crystals (Figure 4a). Here, the structural lines (we use the term line for twodimensional graphene) can be described as
plane indexes (i.e., (hk0)), which are more
general and directly comparable with
other graphene based structures, namely,
AA, AB, and AA stacking.[6,18] In the
crystal structure of graphene, there are two
kinds of distinctive lines, that is, (100) and
(110), wherein the corresponding d-values
are 2.13 and 1.23 , respectively. The
zigzag (100) line is the closest-packed
line, indicating that the <100> direction
of a graphene sheet is the preferential
Figure 1. HRTEM images of commercial arc-SWNTs. a) A low resolution TEM image of SWNTs. growth direction. Moreover the higher
a) An enlarged image of the region indicated in (a). b) HRTEM image (FEI Titan Cubed) of the surface energy of the (100) line, 1.3 eV/,
SWNTs. b,b) Inverse FFT images taken from the regions indicated in (b).
compared with that of the armchair

resulted from a diagonal (30 to the tubule axis) growth of a

zigzag graphene ribbon.
The strain energy for a concentric SWNT depends only
on the radius of the tubules.[7] For the smallest diameter
SWNT, of about 0.4 nm, which has been experimentally
observed[8] the energy is extremely large, that is, up to about
0.48 eV/atom and 0.85 eV/atom for the zigzag and armchair
configurations, respectively.[7] The strain energy of the helical
growth of a zigzag or armchair graphene ribbon is just about
a quarter that of seamless cylindrical SWNTs. This calculation suggests that the growth of seamless SWNTs may be
energetically prohibitive and uncompetitive with the structure proposed here under the conditions of conventional
CVD processes.

2 www.small-journal.com

2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

small 2014,
DOI: 10.1002/smll.201400884

Structure of Single-Wall Carbon Nanotubes: A Graphene Helix

Figure 3. Strain energy per atom versus the radius for helical growth of
a graphene ribbon and tubular growth of a seamless graphene tube,
calculated by Xin et al. method,[7] which was based on the electronic
band structure without introducing any empirical potential. We assumed
that the graphene ribbon was sufficiently narrow not to induce relevant
additional strain energy.

Figure 2. AFM analysis of a SWNT. a) A non-contact mode AFM image

of a SWNT. It appears to be broadened (15 nm in width) due to the
AFM tip effects. b) A corrected AFM image where the nodal morphology
is evident. c) From the heights of the tubule, 0.8 nm, which is much
smaller than the diameters of the tubules, 23 nm (Figure 1), we expect
that the tubule (i.e., a graphene helix) was flattened (4 nm in width)
during the severe sonication. The green and red arrows indicate the
scan directions of the line profiles shown in (c) and (d), respectively.
The blue arrows indicate the periods of the helix, that is, the width of
the helix.

(110) line, 1.0 eV/,[19] can strongly drive the growth of

graphene along the <100> direction.
The conceptual structure of graphene can be extended
to describe SWNTs. This enables us to use plane indexes,
which are convenient for describing the structures of SWNTs
whether they are armchair or zigzag in structure as shown
in Figure 4b,c. This structural unit cannot describe the entire
tubule. The preferential growth direction, <100>, appears
diagonally at 30 for the armchair and 60 for the zigzag configurations to the tubule axis (Figure 4b,c). This indicates that
the armchair configuration is the energetically lower form
of SWNTs because its strain energy is about one-third that
of a zigzag (ribbon) SWNT as shown in Figure 5. The strain
energy for a commonly observable armchair (ribbon) SWNT
with a diameter of 1 to 2 nm is as low as 0.02 eV per atom.
Longitudinal, that is, overlapping views of a SWNT differ
significantly from the azimuthal orientations. This difference
implies that their electron diffraction (ED) patterns depend
on the incident direction of the electron beam.[20] Two views
small 2014,
DOI: 10.1002/smll.201400884

Figure 4. Schematic showing the structural unit of graphene and SWNTs.

a) The structural unit of graphene described by the space group, simple
hexagonal (# 191) (a = b = 2.46 , c = 0, and the angle between a and
b = 120). The (100) line of the two crystalline lines is the close-packed
line, and the <100> directions become the preferred growth orientation. b),
c) The structural units of an armchair and a zigzag SWNTs. Both structures
can be described by the plane indexes of hexagonal groups, such as in
graphene. The angles of the preferred growth direction <100> to the tubule
axis are 30 and 60, respectively. The red dotted lines represent imaginary
boundaries of a zigzag graphene helix that drives the helical growth.

2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.small-journal.com

full papers

J.-K. Lee et al.

Figure 5. Strain energy per atom versus the radius for armchair and
zigzag SWNTs formed by helical growth of a graphene ribbon, calculated
by Xin et al. method,[7] which was based on the electronic band structure
without introducing any empirical potentials. The red dotted lines in b
indicate the imaginary boundaries of the graphene ribbons in the form
of a tubule. We assumed that the graphene ribbon was sufficiently
narrow not to induce any relevant additional strain energy.

of AA and AA armchair SWNTs and their simulated ED

patterns are presented in Figures 6ad. The ED patterns
also vary with inclinational tilts of SWNTs on the TEM grid
as shown in Figure 7. The ED pattern for an untilted AA
armchair SWNT has the 2 mm symmetry, in which each 6
spots for the (100) and (110) planes appear on each Ewalds
sphere, and is directly comparable to that of the AA stacked
MWNTs.[6]
The digital diffractogram for a tubule reveals evidence
of the armchair configuration as shown in Figure 6e. Many
ED patterns identical or similar to the expected patterns
of the armchair structure have been reported in the literature.[3,2129] Arepalli et al.[30] showed an ED pattern identical
to the expected pattern for the untilted AA armchair SWNT
(Figure 6c) from a bundle of SWNTs using a nano-diffraction
technique.
The majority of other reported ED patterns have
revealed the features of distorted armchair SWNTs, in
which graphene was rotated by 312 and each ED spot was
split into two and appeared as a twin. Previous authors have
interpreted the split of ED spots as due to chirality.[3,2128]
All of the ED patterns of SWNTs reported in the literature, with the exception of two, reported by Liu et al.,[22]
showed the features of armchair SWNTs.[3,2330] It is true that
the data of Liu et al. are exceptions since they show zigzag
structure-like patterns (Supporting Information Figure S2).
The observations of these patterns cannot be unequivocal
evidence of zigzag SWNTs because the (110) spots of the
two ED patterns were not presented. Such ED patterns,
that is, six (100) spots for the zigzag configuration appeared
whereas (002) spots were absent were obtained from
MWNTs with only a few layers, that is, double-wall carbon
nanotubes (DWNTs). DWNTs exhibit a zigzag configuration,[6] and their (002) signals are absent. Thus, the previous
ED analyses in the literature support our interpretation

4 www.small-journal.com

Figure 6. Schematic showing two typical overlapping views of armchair

SWNTs and their simulated and expected ED patterns. a,b) AA and AA
overlapping views. The structural units and planes are drawn in the
schematic views. c,d) Simulated ED patterns for AA and AA overlapping
views of an armchair tubule (diameter, 2 nm and length, 4 nm).
e) HRTEM image (FEI Titan Cubed) of a tubule. e) Inverse FFT image
taken from the region indicated in (e). e) Digital diffractogram taken
from the region indicated in (e). The arrows in (e) and (e) indicate the
trace of graphene helix.

namely the armchair configuration is the preferred structure

of SWNTs.
Several STM images interpreted as zigzag SWNTs have
been reported previously.[3136] However, the present analysis
of the STM images shows no evidence for zigzag SWNTs
(i.e., STM images directly revealing the graphene lattices in
the zigzag direction of the tubules). Venema et al.[32] interpreted the triangular (or rhombic) lattices in the STM images
of a tubule in terms of a zigzag structure, in which the lattices were indicative of zigzag SWNTs (Supporting Information Figure S3).[34,35] However it is unreasonable for zigzag
SWNTs to reveal the rhombic lattices instead of the hexagonal configuration of graphene, which appears for armchair
SWNTs.[31,34] It is conceivable that the samples exhibiting the
rhombic lattice in the STM images were MWNTs and could

2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

small 2014,
DOI: 10.1002/smll.201400884

Structure of Single-Wall Carbon Nanotubes: A Graphene Helix

Figure 7. Effect of inclination tilt of a tubule on ED spots. With the inclination angle (), some of ED spots for a) AA and b) AA overlapping of an
armchair SWNT are disappeared ( = 30 of AA overlapping) or evolved ( = 5, 10, 20, 30 of AA overlapping). The inset defines the inclination
angle.

be DWNTs consistent with the diameter of their sample,

1.4 nm.[32] When the graphene sheets of DWNTs, which
have the AA stacked zigzag configuration,[6] are circumferentially translated by several percent, the rhombic symmetry
can be projected in their STM images due to an additional
effect of the electrons of the alternate atoms from beneath
the AA stacked graphene sheet. The rhombic images are the
signals amplified from incoherently overlapped two atoms
(Supporting Information Figure S3). An STM image showing
rhombic lattices was presented[37] from a clear MWNT with
a diameter of about 25 nm. Furthermore, the zigzag hexagonal STM image was directly measured from a MWNT.[35]
The two STM images obtained from MWNTs indicate that
AA stacked zigzag MWNTs can project the zigzag hexagonal directly or the triangular lattice-like images indirectly
depending on the degree of the translation of the graphene.
The hypothesis that the samples that produce triangular STM
images are derived from MWNTs needs further confirmation.
Venema et al.[32] claimed that their sample comprised
SWNTs from measuring the nanotube diameters to be 1.4 nm,
although MWNTs (i.e., DWNTs) can also exhibit similar
diameters[38] although the means by which the structures of
the samples[31,3335] were established were absent. Furthermore, all types of CNTs generally coexist.[38] Unlike the result
small 2014,
DOI: 10.1002/smll.201400884

of ED analysis, the frequent observation of SWNTs with the

zigzag features in STM measurement indirectly suggests that
a portion of their samples were indeed MWNTs.[3134] On the
other hand, STM images of distorted SWNTs, showing armchair hexagonal lattices, have been reported[31,32,34] in which,
graphene revealed mismatched and contorted lattices with
distortion angles of 10, which are in the range of the angles
measured by the ED signals. In summary, the inspection
of STM images cannot unequivocally distinguish between
the types of CNTs and, moreover, the existence of a zigzag
SWNT has yet to be confirmed.
The growth mechanism of helical SWNTs can be inferred
from the different lengths of the nodes which range from 1 nm
to 5 nm as shown in Figure 1a and b. The growth of a
narrow and thin graphene ribbon[39] (i.e., 1 nm, 4 zigzag
hexagon width) in the zigzag configuration drives the helical
growth (Figures 8a,b) process. As the top of the graphene
ribbon rises from the nucleation site, the edges of the helix
become wider to minimize surface energy by lateral growth
along to the <100> directions of the edges (Figures 8b,c).
The secondary lateral growth of the zigzag graphene helix
fills the gaps between the nodes preformed by the primary
helical growth, and completes the tubular shape in the armchair configuration. Such two-step growth produces a unique

2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.small-journal.com

full papers

J.-K. Lee et al.

Figure 8. The growth model of a SWNT. a) Schematic showing a

graphene ribbon diagonally nucleated in the zigzag configuration. The
<100> direction, which is normal to the cross-packed (100) line, is the
preferred growth direction. b) Schematic showing the initial stage of
the growth of the graphene ribbon. c) Schematic showing the two-step
growth of the graphene helix; lateral growth, followed by helical growth
of the zigzag graphene ribbon. The length of a node (i.e., a cycle of
the helix), l, is smaller than 5.3d where d is the diameter of a tubule.
d) Schematic showing a completed armchair SWNT with a diagonal
dislocation in which is the distortion angle.

back and belly nodal morphology (Figure 8d), observable

in the digitalized HRTEM images of Meyer et al. (Supporting
Information Figure S1a).[9] The unique nodal morphology
for armchair SWNTs is also observable in the STM image
for the 1-nm tubule reported by Ge et al.[36] although they
interpreted the tubule as a zigzag structure. The angle of their
prominent rows, 50 to tubule axis (Supporting Information Figure S4), demonstrates that the tubule is an armchair
SWNT distorted by about 10 which is frequently observed in
the literature.[2129,31,32,34] A zigzag SWNT, if it exists, should
reveal prominent rows oriented by about 30 to the tubule
axis (Supporting Information Figure S3). We anticipate
that the slightly deformed rhombic lattices (which can be
regarded as a part of the hexagon-like lattices) is attributed
to the steep curvature of the 1 nm-tubule which makes the
atoms be seen to be closer in STM analysis and projects the
derived images, as in the case of the AA stacking. The current analysis for the STM image can explain the inconsistent
and diverse STM images for SWNTs in the literature where
distorted graphene is the unique feature.[31,32,34,36] The direct
armchair hexagons of graphene with lattice distortions[31,32,34]

6 www.small-journal.com

and indirect derived symmetries can appear according to

diameters and local positions of nodal samples comprising
the back and the belly zones (Supporting Information
Figures S1 and S4). The line defects in the STM images for
SWNTs[31,32,34] are characteristic of the unzipped zones.
Due to the STM imaging mechanism, the three-dimensionally mismatched graphene helices appear as line defects in
their STM images.
We predict that two-step growth, initially via a helical
structure followed by zipping of the single strand, distorts
the graphene helix (Figure 8d) and the distorted graphene
helix may have been previously interpreted as a chiral SWNT.
The atoms on edges of the unzipped zones of a helix may
conceivably be H-terminated or modified (i.e., forming a
five- or seven-membered ring) to minimize the conformation
energy.
The new model explains the diverse HRTEM morphologies of SWNTs in the literature as well as the different diameters of the tubules that are determined by the width of the
initial graphene ribbons. Ideal matching of the two spatially
approaching graphene edges should form a smooth nodal
morphology, such as in the HRTEM image of Hashimoto
et al.[23] This explanation addresses the prominent appearance of the D band (1350 cm1) as well as its large variation
in Raman spectra for SWNTs reported in the literature.[40,41]
The D band can be strong for some samples (incoherently
scrolled) or weak for other samples (coherently scrolled).
We expect that the D band should be negligible if SWNTs
are seamless graphene tubes because the fraction of the disordered portion at the ends of a seamless tube will be very
small due to the large aspect ratio of SWNTs. Our graphene
helix model also explains the inconsistent physical properties of SWNTs, that is, an approximately four-fold deviation
in breaking strength and Youngs modulus,[42] which should
be critically dependent on the perfection of the graphene
helix.
Our analysis of the ED patterns and STM images in the
literature shows that the preferred armchair configuration
is possibly the unique structure of SWNTs. Theoretically,
SWNTs in the armchair configuration are metallic.[41,43] The
theory is based on the assumption that SWNTs are perfect
tubules. The electrical properties of SWNTs, as metallic or
semiconducting, may be attributed to imperfections in the
graphene helix and experimental errors in sampling. It is
well understood that strain can change the band structure
of SWNTs and defects act as conductance barriers.[44] We
propose, on these grounds, that the semiconducting properties may be an intrinsic feature of helical nanotubes. There
is evidence that some of the previous STM data[3135] have
been obtained from MWNTs (mostly DWNTs) which show
metallic conductivity.[40] Straight and undistorted tubules
with a diameter of 2 nm are DWNTs which reveal the
rhombic STM symmetry where the angle of the prominent
rows to the tubule axis is about 30 (Supporting Information Figure S3).[3135] DWNTs can be relatively free from
stress due to the growth mechanism where AA stacked
graphene nanoribbons collectively grow,[6] unlike SWNTs
with the typical nodal morphology which is evident in AFM
images.[45]

2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

small 2014,
DOI: 10.1002/smll.201400884

Structure of Single-Wall Carbon Nanotubes: A Graphene Helix

3. Conclusion
The analysis and experimental data presented in this paper
show that SWNTs may be derived from the initial helical
growth of a zigzag oriented graphene ribbon. These conclusions address previous experimental evidence in the literature, diverse ED patterns, HRTEM and STM morphologies
as well as inconsistencies in the measured mechanical and
electrical properties of SWNTs. The prevailing nodal morphology of SWNTs is direct evidence for our graphene
helix model. The results clarify the interpretation of the scientifically and technically important electrical properties of
SWNTs.

4. Experimental Section
Commercial purified SWNTs were synthesized by an arc-discharge
technique (Hanwha nanotech) and analyzed by HRTEM and AFM.
Samples were sonicated in ethyl alcohol with an ultrasonic power
density of 20 W/mL for 2 h. HRTEM images were acquired by a JEM2100F operating at 200 kV and a FEI Titan Cubed with aberration
corrector and a monochromater operating at 80 kV. We confirmed
that the images of the samples did not change when subject to
further FEI Titan TEM observation (i.e., there was no any electron
damage to the samples). A droplet of the SWNTs suspension was
dispersed on a single crystal silicon (100) substrate for AFM sampling The surface morphology was probed by a non-contact-mode
AFM (PSIA, XE-100) using a commercial cantilever with a tip radius
of curvature <10 nm. The tubules appeared to be broadened
(15 nm in width) due to the AFM tip effects, and we corrected the
broadened tubule with the average diameter of the samples measured by HRTEM observation.

Acknowledgements
The authors thank Young-Su Lee for providing the coordinates
of SWNTs for simulations of ED patterns and valuable discussions. This work was supported by KIST Future Resource Program
(2V01900, 2V02120, and 2E24011) and the Korea Research
Council of Fundamental Science & Technology under the STRM
project (2013).

[1] I. Ibrahim, A. Bachmatiuk, J. H. Warner, B. Bchner, G. Cuniberti,

M. H. Rmmeli, Small 2012, 8, 1973.
[2] R. H. Baughman, A. A. Zakhidov, W. A. de Heer, Science 2002,
297, 787.
[3] S. Iijima, T. Ichihashi, Nature 1993, 363, 603.
[4] D. S. Bethune, C. H. Klang, M. S. de Vries, G. Gorman, R. Savoy,
J. Vazquez, R. Beyers, Nature 1993, 363, 605.
[5] S. Iijima, Nature 1991, 354, 56.
[6] J.-K. Lee, S. Lee, Y.-I. Kim, J.-G. Kim, K.-I. Lee, J.-P. Ahn, B.-K. Min,
C.-J. Yu, K. H. Ghae, P. John, Appl. Phys. Lett. 2013, 102,
161911.

small 2014,
DOI: 10.1002/smll.201400884

[7] Z. Xin, Z. Jianjun, O.-Y. Zhong-can, Phys. Rev. B 2000, 62,

13692.
[8] N. Wang, G. D. Li, Z. K. Tang, Chem. Phys. Lett. 2001, 339, 47.
[9] R. R. Meyer, S. Friedrichs, A. I. Kirkland, J. Sloan, J. L. Hutchison,
M. L. H. Green, J. Microsc. 2003, 212, 152.
[10] A. Hashimoto, K. Suenaga, A. Gloter, K. Urita, S. Iijima, Nature
2004, 430, 870.
[11] A. Thess, R. Lee, P. Nikolaev, H. Dai, P. Petit, J. Robert, C. Xu,
Y. H. Lee, S. G. Kim, A. G. Rinzier, D. T. Colbert, G. E. Scuseria,
D. Tomnek, J. E. Ficher, R. E. Smalley, Science 1996, 273,
483.
[12] K. Suenaga, Y. Sato, Z. Liu, M. Koshino, C. Jin, JEOL News 2009,
44, 32.
[13] A. Fonseca, K. Hernadi, P. Piedigrosso, J.-F. Colomer,
K. Mukhopadhyay, R. Doome, S. Lazarescu, L. P. Biro, Ph. Lambin,
P. A. Thiry, D. Bernaerts, J. Nagy, Appl. Phys. A 1998, 67,
11.
[14] C.-H. Kiang, W. A. Goddard III, R. Beyers, D. S. Bethune, J. Phys.
Chem. 1996, 100, 3749.
[15] M. Meo, M. Rossi, Eng. Fracture Mech. 2006, 73, 2589.
[16] H. R. Gutirrez, U. J. Kim, J. P. Kim, P. C. Eklund, Nano Lett. 2005,
5, 2195.
[17] M. Marchand, C. Journet, D. Guillot, J.-M. Benoit, B. I. Yakobson,
S. T. Purcell, Nano Lett. 2009, 9, 2961.
[18] J.-K. Lee, S.-C. Lee, J.-P. Ahn, S.-C. Kim, J. I. B. Wilson, P. John,
J. Chem. Phys. 2008, 129, 234709.
[19] P. Koskinen, S. Malola, H. Hkkinen, Phys. Rev. Lett. 2008, 101,
115502.
[20] Z. Liu, L.-C. Qin, Chem. Phys. Lett. 2005, 412, 399.
[21] J.-F. Colomer, L. Henrard, Ph. Lambin, G. Van Tendeloo, Phys. Rev.
B 2001, 64, 125425.
[22] Z. Liu, L.-C. Qin, Chem. Phys. Lett. 2004, 400, 430.
[23] J.-F. Colomer, L. Henrard, Ph. Lambin, G. Van Tendeloo, Eur. Phys.
J. B 2002, 27, 111.
[24] J. C. Meyer, M. Paillet, G. S. Duesberg, S. Roth, Ultramicroscopy
2006, 106, 176.
[25] M. Gao, J. M. Zuo, R. D. Twesten, I. Petrov, L. A. Nagahara,
R. Zhang, Appl. Phys. Lett. 2003, 82, 2703.
[26] H. Jiang, A. G. Nasibulin, D. P. Brown, E. I. Kauppinen, Carbon
2007, 45, 662.
[27] T. Beetz, L. Wu, Y. Zhu, Microsc. Microanal. 2005, 11, 560.
[28] J.-M. Zuo, T. K. Kim, A. Celik-Aktas, J. Tao, Z. Kristallogr. 2007,
222, 625.
[29] R. Arenal, P. Lthman, M. Pichen, T. Than, M. Pamer, V. Jouruam,
J. Phys. Chem. C 2012, 116, 14103.
[30] S. Arepalli, O. Gorelik, C. Huffman, P. Nikolaev, preprint at http://
research.jsc.nasa.gov/PDF/Eng-10.pdf (accessed: 2006).
[31] J. W. G. Wilder, L. C. Venema, A. G. Rinzler, R. E. Smalley,
C. Dekker, Nature 1998, 391, 59.
[32] L. C. Venema, J. W. G. Wilder, C. Dekker, G. A. Rinzler,
R. E. Smalley, Appl. Phys. A 1998, 66, S153.
[33] T. W. Odom, J.-L. Huang, P. Kim, C. M. Lieber, Nature 1998, 391,
62.
[34] P. Kim, T. W. Odom, J. Huang, C. M. Lieber, Carbon 2000, 38,
1741.
[35] A. Hassanien, A. Mrzel, M. Tokumoto, D. Tomnek, Appl. Phys.
Lett. 2001, 79, 4210.
[36] M. Ge, K. Sattler, Appl. Phys. Lett. 1994, 65, 2284.
[37] Z. Osvth, G. Vrtesy, L. Tapaszt, F. Wber, Z. E. Horvth,
J. Gyulai, L. P. Bir, Phys. Rev. B 2005, 72, 045429.
[38] T. Yamada, T. Nanai, K. Hata, D. N. Futaba, K. Mizuno, J. Fan,
M. Yudasaka, M. Yumura, S. Iijima, Nat. Nanotechnol. 2006, 1,
131.
[39] H. E. Lim, Y. Miyata, R. Kitaura, Y. Nishimura, Y. Nishimoto, S. Irle,
J. H. Warner, H. Kataura, H. Shinohara, Nat. Commu. 2013, 4,
2548.

2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.small-journal.com

full papers

J.-K. Lee et al.

[40] J. Hodkiewicz, Characterizing Carbon Materials with Raman Spectroscopy Application note:51901, Thermo Fisher Scientific Inc.
2010.
[41] (Eds: A. Jorio, M. S. Dresselhaus, G. Dresselhaus), in Carbon Nanotubes: Advanced Topics in the Synthesis, Structure, Properties and
Applications Springer-Verlag, Berlin, Germany 2008, pp 84, 256,
387, 502, 548-552, 590.
[42] M.-F. Yu, B. S. Files, S. Arepalli, R. S. Ruoff, Phys. Rev. Lett. 2000,
84, 5552.

8 www.small-journal.com

[43] M. S. Dresselhaus, G. Dresselhaus, R. Saito, Carbon 1995, 33, 883.

[44] (Eds: A. Jorio, M. S. Dresselhaus, G. Dresselhaus), in Carbon Nanotubes: Advanced Topics in the Synthesis, Structure, Properties and
Applications Springer-Verlag, Berlin, Germany 2008, pp 463464.
[45] L. Ding, D. Yuan, J. Liu, J. Am. Chem. Soc. 2008, 130, 5428.

2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Received: April 1, 2014

Revised: April 19, 2014
Published online:

small 2014,
DOI: 10.1002/smll.201400884