Carbon nanotube-nucleobase hybrids: Nanorings from uracil-modified single-walled carbon nanotubes

Article abstract of DOI:10.1002/chem.201100312

Single-walled carbon nanotubes (SWCNTs) have been covalently functionalized with uracil nucleobase. The hybrids have been characterized by using complementary spectroscopic and microscopic techniques including solid-state NMR spectroscopy. The uracil-functionalized SWCNTs are able to self-assemble into regular nanorings with a diameter of 50-70 nm, as observed by AFM and TEM. AFM shows that the rings do not have a consistent height and thickness, which indicates that they may be formed by separate bundles of CNTs. The simplest model for the nanoring formation likely involves two bundles of CNTs interacting with each other via uracil-uracil base-pairing at both CNT ends. These nanorings can be envisaged for the development of advanced electronic circuits. Copyright

Full text of DOI:10.1002/chem.201100312

DOI: 10.1002/chem.201100312
Carbon Nanotube–Nucleobase Hybrids: Nanorings from Uracil-Modified
Single-Walled Carbon Nanotubes
Prabhpreet Singh,[a] Francesca Maria Toma,[b] Jitendra Kumar,[c] V. Venkatesh,[c]
Jesus Raya,[d] Maurizio Prato,[b] Sandeep Verma,*[c] and Alberto Bianco*[a]
Abstract: Single-walled carbon nano-
tubes (SWCNTs) have been covalently
functionalized with uracil nucleobase.
The hybrids have been characterized
by using complementary spectroscopic
and microscopic techniques including
solid-state NMR spectroscopy. The
uracil-functionalized SWCNTs are able
to self-assemble into regular nanorings
with a diameter of 50–70 nm, as ob-
served by AFM and TEM. AFM shows
that the rings do not have a consistent
height and thickness, which indicates
that they may be formed by separate
bundles of CNTs. The simplest model
for the nanoring formation likely in-
volves two bundles of CNTs interacting
with each other via uracil–uracil base-
pairing at both CNT ends. These nano-
Keywords: carbon nanotubes
functionalization nanorings
nucleobases · self-assembly
·
·
AHCTUNGTRENNUNG
·
Introduction
an interesting structural variation of carbon nanotubes. They
exhibit remarkable transport and field emission proper-
ties[12–13] and can be also exploited for building nanoscale
electronic circuits. Nevertheless, the observation of ring- or
coil-like structures based on CNTs remains challenging.
Until now, various CNT-based rings have been produced by
the catalytic synthesis or different post-treatment meth-
ods,[14–19] including chemical modifications and physical ma-
nipulations.[15–19] Single-walled CNT (SWCNT) rings were
generated from tip-oxidized tubes by covalent closure reac-
tions,[17] by controlled bending on patterned templates[18] or
by ultrasonic irradiation of straight nanotube segments by
using strong oxidizing conditions.[19]
Nanostructured materials can adopt different shapes includ-
ing wires, rods, ribbons, or rings. These well-defined forms
can be extremely useful for applications in the field of nano-
technology and biomedicine.[1–2] Within the different classes
of nanomaterials currently developed, carbon nanotubes
(CNTs)[3,4] are attracting a lot of interest due to their unique
mechanical, thermal, optical, and electronic properties.[5] For
the implementation of CNTs into advanced devices, the co-
valent functionalization of CNTs is one of the most power-
ful ways to improve both dispersion and interactions.[6–8]
CNTs can build up complex superstructures, such as heli-
ces,[9] loops,[10] and coiled structures.[11] Carbon nanorings are
The nucleobases are the fundamental constituents of nu-
cleic acids and can provide hydrogen-bonding aided molecu-
lar recognition for the formation of duplexes, triplexes, tet-
raplexes, and higher-order architectures.[20] In the recent
years, a number of groups have carried out studies on how
different nucleobases and nucleobase pairs self-assemble on
a surface.[21,31–33] The molecular self-assembly involving ad-
sorption of nucleic acid bases on a graphite surface has been
characterized by different spectroscopic techniques. The nu-
cleobases self-assemble to form monolayers that are stabi-
lized by hydrogen bonds. Based on our recent interests on
the covalent functionalization of SWCNTs with the nucleo-
bases, we have explored the possibility to create original ar-
chitectures.[22a] The noncovalent interaction and the ability
of the nucleobases to disperse SWCNTs have been evaluat-
ed theoretically and experimentally.[23,24] However, the nu-
cleobases have been scarcely used for the covalent attach-
ment on carbon nanotubes.[22]
[a] Dr. P. Singh, Dr. A. Bianco
CNRS, Institut de Biologie Molꢀculaire et Cellulaire
Laboratoire d’Immunologie et Chimie Thꢀrapeutiques
67000 Strasbourg (France)
Fax : (+33)388 610680
E-mail: a.bianco@ibmc-cnrs.unistra.fr
[b] Dr. F. M. Toma, Prof. Dr. M. Prato
Dipartimento di Scienze Chimiche e Farmaceutiche
Universitꢁ di Trieste, 34127 Trieste (Italy)
[c] Dipl.-Chem. J. Kumar, Dipl.-Chem. V. Venkatesh, Prof. Dr. S. Verma
Department of Chemistry
Indian Institute of Technology-Kanpur
Kanpur-208016 UP (India)
E-mail: sverma@iitk.ac.in
[d] Dr. J. Raya
Institut de Chimie
Laboratoire de RMN et de biophysique des membranes
UMR 7177 CNRS Universitꢀ de Strasbourg
67000 Strasbourg (France)
In this contribution, we describe the self-assembly of
nanorings formed by the uracil nucleobase, covalently at-
tached to SWCNTs, on highly oriented pyrolytic graphite
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201100312.
6772
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6772 – 6780

FULL PAPER
Scheme 1. Synthesis of SWCNT derivatives 1 and uracil–CNT hybrids 2 and 3.
(HOPG) at the liquid–solid interface. We selected uracil be-
cause it has two adjacent cis amide groups giving rise to
four double hydrogen-bonding sites. By using these binding
ty of impurities can be effectively eliminated, this process
also leads to a certain loss of material and increases the de-
fects in the graphene layers. Pristine SWCNTs were first oxi-
dized and functionalized as shown in Scheme 1. The loading
(amount of functional groups) was calculated after derivati-
zation of the carboxylic groups (activated as acid chlorides)
with tert-butoxycarbonyl (Boc)-monoprotected diaminotri-
ethylene glycol. The Boc group was removed with HCl solu-
tion in dioxane to afford functionalized f-SWCNT 1. The
amount of amino groups corresponded to 0.58 mmol per
gram of CNTs as assessed by the Kaiser test. f-SWCNT 2
was synthesized by mixing Boc-deprotected f-SWCNT 1
with 3-(1-uracilyl)propionic acid[26a] activated, in turn, with
HOBt (N-hydroxybenzotriazole) and EDC·HCl (1-ethyl-3-
(3-dimethyl-aminopropyl)carbodiimide hydrochloride). f-
SWCNT 3 was instead synthesized by following the acyl
chloride route, allowing the oxidized SWCNTs (ox-
SWCNTs) to react with (COCl)2 under N2 at refluxing con-
ditions. The resulting acid chloride derivative was then
heated at reflux with 1-(2-aminoethyl) uracil[26b] to afford f-
SWCNT 3.
À
À
sites, namely, N1 H, N3 H donor and C2=O, C4=O accept-
or sites, several distinguishable doubly hydrogen-bonded
U···U dimers are formed.[25] The uracil-functionalized
SWCNT (f-SWCNTs) hybrids have been characterized by
using complementary microscopic and spectroscopic tech-
niques including solid-state NMR spectroscopy. The nano-
ACHTUNGTRENNUNGrings have been evidenced by AFM and HRTEM.
Results and Discussion
Synthesis of the functionalized nanotubes: Pristine SWCNTs
are always present in the form of bundles with a diameter of
around 10–200 nm, which makes them difficult to disperse
in any solvent. We followed a direct solubilization strategy,
achieved by reducing the length of SWCNTs by a chemical
oxidation method.[22] Although through the oxidative disso-
lution of the carbon-encapsulated metal particles the majori-
Chem. Eur. J. 2011, 17, 6772 – 6780
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6773

A. Bianco, S. Verma et al.
Characterization of the different CNT hybrids: The nano-
tube samples were characterized by standard analytical tech-
niques, such as thermogravimetric analysis (TGA), Raman
spectroscopy, solid-state magic-angle spinning (MAS) NMR
spectroscopy, and microscopic techniques including TEM,
HRTEM, and AFM. The degree of functionalization for the
ox-SWCNT and the functionalized conjugates (f-SWCNT 1,
2, and 3) was assessed by TGA under a N2 atmosphere
(Figure 1). The TGA of f-SWCNT 2 and 3 shows a gradual
weight loss of about 27.2 and 23.6% at 5008C, respectively,
relative to 24.5% of the f-SWCNT 1 and 16.3% of the ox-
also calculate the average diameter distribution by studying
the radial breathing mode (RBM) of the Raman spectrum.
Usually, the CNT samples are made of heterogenous materi-
als with different diameters and different helicities so we
need different excitation wavelengths that can selectively
excite nanotubes with different diameters. We performed
the Raman spectra by using all the excitation lasers avail-
able in our setup (Figure 2). To achieve a clear representa-
tion, the RBM peaks were normalized at 150 cmÀ1, whereas
the entire spectra were normalized by the maximum value
of the G band. The analysis of the RBM peaks gave a diam-
eter mean value of 1.1Æ0.1 nm (see the Supporting Infor-
mation, Table S1). D/G ratios were calculated by comparing
the values in the extended 633 nm excitation wavelength
spectra. Unfortunately, the fluorescence increased the back-
ground and it was not possible to precisely calculate this
ratio, especially for compound f-SWCNT 3. In general, we
can affirm that the D/G ratio, taking into account the ampli-
tude of the peaks, is equal to 0.14–0.15. The D/G ratio pro-
vided information on the number of defects, induced by the
functionalization process that could also affect the softness
of materials (see below).
We have recently characterized the functionalized carbon
nanotubes by using solid-state MAS NMR spectrosco-
py.[22a,28] Although this technique does not allow us to unam-
biguously assign the functional groups bound to the nano-
tubes, the changes on the resonance position and intensity
support the success of the covalent functionalization. There-
fore, we have decided to carry out 13C NMR spectroscopy
Figure 1. TGA of ox-SWCNT (a), f-SWCNT 1 (a), f-SWCNT 2
(c), and f-SWCNT 3 (c) under a N2 atmosphere by using a ramp of
108CminÀ1
.
SWCNT. On the basis of TGA
weight loss, we have estimated
that the amount of functional
groups per gram (Fw, mmolgÀ1)
of f-SWCNTs 2 and 3 is 0.35
and 0.48, respectively, which
correlates with the value ob-
tained from the Kaiser test
(0.58 mmolgÀ1). On the basis of
weight loss, we estimated that
the degree of functionalization
for f-SWCNT 2 and 3 is one
functional group every 77 and
49 carbon atoms, respectively.
Raman spectroscopy is
a
powerful tool to investigate to
what extent the sp2-bonded net-
work of the SWCNTs is affect-
ed after the functionalization.[27]
This could be done by evaluat-
ing the degree of functionaliza-
tion by measuring the D/G
band ratio. The D band is con-
sidered as an indication of the
disorder in the graphite lattice
or the presence of defects in
Figure 2. Raman spectra of ox-SWCNT (c), f-SWCNT 1 (c), f-SWCNT 2 (c), and f-SWCNT 3 (c).
In a–c) are presented the RBM spectra at 532, 633, and 785 nm, respectively. Panel d) shows extended spectra
the nanotube. Besides, we could with D and G peaks at 633 nm.
6774
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6772 – 6780

Carbon Nanotube–Nucleobase Hybrids
FULL PAPER
by using magic-angle spinning. Synchronized Hahn–Echo
and 1H–13C cross polarization (CP) pulse sequences were ap-
plied to the samples for the characterization of ox-SWCNTs
and their derivatives. Figure 3A shows a strong 13C NMR
to the carbon nanotube surface probably due to covalent at-
tachment. This excludes the possibility of a nonspecific in-
teraction of the linker protons with the surface of the nano-
tubes by simple adsorption. To further confirm the covalent
attachment we simply mixed f-SWCNT 3 and 1-(2-amino-
AHCTUNGTREGeNNNU thyl)uracil by grinding in a pestle mortar. We observed
sharp signals only for uracil due to the much higher trans-
verse 13C relaxation of the small molecule in comparison to
the SWCNT hybrids. In the case of a strong interaction be-
tween uracil and the nanotubes, the signals of the former
should also be as broad as the carbon atom resonances be-
cause they should automatically have the same dispersion of
the isotropic chemical shifts similar to amorphous substan-
ces (see the Supporting Information, Figure S1).
Formation of the nanorings: By following the characteriza-
tion of the different uracil-modified CNT conjugates, we
have studied their capacity to self-assemble on different
types of surfaces typically used for microscopic analysis. We
performed HRTEM of these uracil-functionalized SWCNTs
conjugates. Samples of f-SWCNT 2 were prepared in H2O/
MeOH and deposited onto a carbon-coated grid. As shown
in Figure 4, we could observe the nanoring like feature with
Figure 3. Solid-state MAS 13C NMR spectroscopic data for a) precursor
ox-SWCNT, b) Boc-protected f-SWCNT 1, c) f-SWCNT 2, and d) free 1-
(2-aminoethyl)uracil base. SSB indicate the spinning sidebands. The
arrows show the apparent changes in the spectrum.
spectroscopic resonance at d=110 ppm for the aromatic
core of ox-SWCNT, and broad chemical shifts in the range
of d=135–180 ppm corresponding to peripheral aromatic
and carbonyl carbon atoms. The f-SWCNT 1 (Boc-protect-
ed; Figure 3B) exhibits a peak around d=20–30 ppm for the
methyl of the Boc group and peaks around d=30–40 and
50–80 ppm corresponding to the methylene carbon atoms of
the triethylene glycol (TEG) linker. Apparent changes were
observed at d=135–180 ppm due to conversion of the acid
groups into amide functionalities. The MAS spectrum of f-
SWCNT 2 (Figure 3C) clearly shows the disappearance of
the Boc group 13C resonance and changes in the d=30–40
and 135–180 ppm regions due to the presence of uracil sig-
nals that fall in the same region as observed for free 1-(2-
aminoethyl)uracil (Figure 3D). This observation suggests the
presence of the TEG linker and the uracil moiety on f-
SWCNT 2. Similarly, we observed the absence of the TEG
and the presence of the uracil moiety in the case of f-
SWCNT 3 (see the Supporting Information, Figure S1).
Figure 4. HRTEM images of f-SWCNT 2 obtained by dispersing the
nanotubes in a 1:1 methanol/water solution and deposited on a TEM
grid. The circle highlights the area in which a few nanotubes seem to in-
teract with each other to form a closed end in a nanoring arrangement.
the thickness of the bundles of nanotubes about 18–20 nm.
By a careful inspection, we could see that actually two or
three nanotube bundles are closing to form the nanoring-
like structures almost joined on one end while the other end
remains still opened. Although it will be difficult to measure
the diameter of this ring, we roughly estimated a diameter
ranging from 90–110 nm. We propose that uracil–uracil nu-
cleobases are interacting through hydrogen bonds that lead
to the formation of the nanorings, whereas the presence of
the TEG chains helps in the interdigitation to produce
smaller bundles of nanotubes and probably participates in
the hydrogen-bonding interaction. Low-resolution TEM
images of f-SWCNT 2 and 3 also display bundles of few
nanometers in diameter that are more dissociated and less
tightly bound within each other than ox-SWCNTs (see the
1
We next utilized H–13C CP-MAS to gain more informa-
tion about the proximity of the protons on f-SWCNT 2 and
3. We observed that the transfer of magnetization occurred
due to the dipolar interaction between the protons and the
carbon atoms for the TEG linker (signal at d=50–90 ppm)
and for the uracil (signals at d=10–50 and 140–190 ppm) for
both f-SWCNT 2 and 3 (see the Supporting Information,
Figure S1). The proton network of the TEG linker and
uracil is evidently not efficient in transferring magnetization
Chem. Eur. J. 2011, 17, 6772 – 6780
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6775

A. Bianco, S. Verma et al.
Supporting Information, Figure S2–S4). This is probably due
to the impact of the functionalization with uracil on the
morphology of the starting nanotubes and the effect of the
presence of the TEG linker, which favors the debundling, in
comparison to ox-SWCNTs that show more complex aggre-
gates of nanotubes.
The AFM images of f-SWCNT 2 taken on the freshly
cleaved HOPG surface revealed the formation of fine
nanoring structures when dispersed in 1:1 H2O/MeOH and
randomly placed all over the HOPG surface (Figure 5).
a chain through weak interactions thus creating a long heli-
cal morphology in most of the surface. In another experi-
ment, we evaporated the solution of f-SWCNT 2 in a lyo-
philizer, dried the sample, and dissolved it again in the same
mixture. We collected a new image with AFM and observed
the formation of the nanorings. We also deposited the solu-
tion of f-SWCNTs 2 on HOPG after different intervals of
time, up to few months, and we observed that longer incuba-
tions in solution neither seem to affect the formation of the
nanorings nor change the nanoring morphology. In compari-
son to AFM, HRTEM images
indicate a slightly larger diame-
ter although they appear less
regular, with approximately the
same thickness of the nanotube
bundles. The surfaces of both
techniques on which samples
were adsorbed are different,
and HOPG is certainly playing
a crucial role in the formation
of the regular nanorings. The
AFM images of the precursor
ox-SWCNT and f-SWCNT 3 on
the HOPG surface showed in-
stead globular and unresolved
structures (see the Supporting
Information, Figure S6 and S7).
The AFM images of ox-
SWCNT show bundles with
average diameters of around
60–120 nm with a height vary-
ing from 20 to 40 nm. We also
observed that the height of
some nanotube bundles are
only 10 nm and the length of
the bundle of nanotubes is
around 140 nm with regular
gaps.
Figure 5. a–c) AFM images of f-SWCNT 2 taken on HOPG surface showing a wide distribution of the isolated
and long chain interconnected nanorings. d) 3D AFM image of the nanorings showing the heights from the
HOPG surface.
In another experiment we
The diameter of these regular nanorings is approximately
around 50–70 nm and the height between 7 and 11 nm. The
thickness of one nanotube bundle ranges from 20 to 25 nm.
This morphology has been reproducibly observed all over
the surface in a series of repeated experiments by using also
different newly prepared batches of f-SWCNT 2.
From the cross-sectional height and diameter analysis
(Figure 6) of the AFM images, the internal empty cavity of
a single nanoring shows a diameter width around 15–20 nm.
Multiple nanorings were observed and the analysis of their
line scans revealed inconsistent profile features. From the
height–diameter profile of the AFM image in Figure 6, it is
clear that the nanorings consist of bundles of nanotubes. We
have seen steps in the height profile of the nanorings indi-
cating that each nanoring is made up of different nanotube
bundles and these bundles determine the height of the
nanoring (see also the Supporting Information, Figure S5).
Interestingly, these multiple nanorings aligned like a helix in
added a hydrogen-bonding competitor to possibly interfere
or to prevent the formation of the nanorings gaining addi-
tional information on the nature of these rings. For this pur-
pose, we used water-soluble 2-(6-amino-9H-purin-9-yl)etha-
nesulfonic acid. Samples of f-SWCNT 2 were prepared in
H2O/MeOH in the presence of the adenine sulfonic acid de-
rivative (1:1 ratio based on number of uracil molecules).
The sample was incubated for 36 and 72 h. The AFM
images of f-SWCNT 2 taken on HOPG revealed the pres-
ence of nanorings after 36 h incubation. Longer incubation
has no effect and the morphology remained the same (see
the Supporting Information, Figure S8). Though the ring
morphology is slightly changed, the interaction between the
uracil moieties is still so strong that the complementary ade-
nine derivative is not able to compete and disrupt the nano-
AHCTUNGTRENNUNG
6776
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6772 – 6780

Carbon Nanotube–Nucleobase Hybrids
FULL PAPER
It may be possible that, after the adsorption of the
SWCNTs on the HOPG surface, the TEG chains at the tips
of f-SWCNT 2 help in the interdigitation and avoid the for-
mation of globular structures by keeping the nanotubes
apart, whereas uracil moieties present at the end of the
TEG chains undergo self-hydrogen-bonding interactions.
However, when SWCNT 3, devoid of TEG chains, is depos-
ited, fibrils of tubes were observed. If we consider the solu-
tion behavior of such hybrids, the chances of intermolecular
hydrogen bonding, through uracil–uracil base-pairing, in
protic solvents (water/methanol) will be low. As a conse-
quence, hydrogen-bonding and hydrophobic interactions
likely play crucial roles in the assembly process at the
HOPG surface. The fact that the HOPG surface is playing a
crucial role can be evicted when deposition of f-SWCNT 2
on mica hampered the formation of the nanorings (see the
Supporting Information, Figure S9).
We suppose that small strain arises due to a small bending
of the ends of two carbon nanotube bundles to form nano-
AHCTUNGTREGrNNNU ings accommodated through the presence of a significant
concentration of defects. Possibly these defects are arranged
in an incoherent fashion in the nanotube and under stress
they slide over each other reversibly.[37] Martel et al.[19] ex-
plained that the activation to coil SWCNTs into rings is
being provided by ultrasonication. As we performed the
sonication, our hybrids likely bend in solution before ad-
sorption. Another piece of evidence for nanoring formation
came from the observations of Sano et al.[17] The persistence
length is a basic mechanical property quantifying the stiff-
ness of a material. Sano et al. have shown that the oxidized
CNTs need to be 1600 nm long (twice the persistence
length) to have a large enough fluctuation for both ends to
meet and to form nanorings. Thermodynamically, a material
with a length shorter than the persistence length behaves as
a stiff rod. In our case, a nanoring with a radius of about
20–25 nm with a thickness of about 7–11 nm suggests that
the CNTs may be as short as few tenths of a nm (<150 nm).
As a consequence, these f-SWCNTs must have a persistence
length much shorter than the tubes studied by Sano et al.,
which means that our nanotubes may be a little soft but not
so much as to allow both ends of a nanotube bundle to meet
and form the nanoring. However, the softness of the materi-
al could play a role in CNT ring formation and our f-
SWCNTs 2 are highly functionalized, as assessed by TGA
and Raman spectra. Raman spectra also show that these
nanotube posses small diameter and on the basis of the as-
sumption that a thin wire is softer than a thick one, we be-
lieve that the nanotubes are soft enough to allow small
bending. Nevertheless, our f-SWCNTs are not so thin nor
the structures so highly disrupted to completely justify the
ring formation that would have occurred also in the case of
f-SWCNT 3. Therefore, despite the fact that other mecha-
nisms might produce suitable conditions, it seems that the
ring-formation mechanism likely deals the most with the
type of functionalization.
Figure 6. AFM images of f-SWCNT 2 taken on a HOPG surface showing
a) nanorings forming
a helix-type structure and b) multiple isolated
nanorings with their corresponding height–diameter profile of nanorings
in each image (below).
cal in creating ordered surface assemblies.[29,30] The nucleo-
bases can exhibit the formation of organized assemblies
through hydrogen-bonding interactions.[31–33] The formation
mechanism of such assemblies, particularly in aqueous envi-
ronments, involves adsorption followed by the rearrange-
ment of the molecules on the surface into hierarchical as-
semblies.[34,35] Thus, the tendency to form the nanoring struc-
tures by these carbon nanotubes after end-to-end functional-
ization with the uracil nucleobase, probably involves the ad-
sorption of SWCNTs on the HOPG surface. Since the
pyrimidine bases are known to feature poor adsorption en-
ergies on graphite,[36] this organization is further stabilized
À
through uracil–uracil self base-pairing involving N3 H and
C4=O. Contrary to the uracil nucleobase, the adenine nucle-
obase covalently attached to SWCNTs, tends to adsorb in a
flat fashion on the HOPG surface[36] due to a stronger at-
traction of adenine with the p-system of graphite resulting
in an aligned nanotube self-assembly.[22]
Based on our assumptions, we propose a mechanism for
the nanoring formation. The black-colored bars in Figure 7
Chem. Eur. J. 2011, 17, 6772 – 6780
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6777

A. Bianco, S. Verma et al.
sonication, wet chemistry process involving solution evapo-
ration on solid interface and HOPG surface. Encouraged by
these results we are currently exploring the possibility to use
these circular hybrids for electronic and biomedical applica-
tions.
Experimental Section
Materials and methods: HiPco nanotubes were purchased from Carbon
Nanotechnologies (lot R0496). 1-(2-Aminoethyl)uracil was synthesized
according to the literature procedures (see the Supporting Information
for details). Reagents and solvents were purchased from Fluka, Aldrich,
Loba Chemie Pvt. (Mumbai, India), Spectrochem (Mumbai, India), and
S.D. Fine-chem (Mumbai, India), respectively, and used without further
purification unless otherwise stated. Moisture-sensitive reactions were
performed under an argon or N2 atmosphere. CH2Cl2 was freshly distilled
from CaH2, THF from Na/benzophenone, and DMF dried over 4 A mo-
lecular sieves. Chromatographic purification was done with silica gel
Merck (Kiesegel 60, 40–60 mm, 230–400 mesh ASTM) in a standard
column. TLC analysis was performed on aluminum sheets coated with
silica gel 60 F254 (Merck, Darmstadt). 1H and 13C NMR liquid-state spec-
tra were recorded on JEOL-JNM LAMBDA 400 model (operating at
400 and 100 MHz, respectively), JEOL ECX-500 model (operating at
500 and 125 MHz, respectively), and Bruker DPX 300. The peak values
were obtained as ppm (d) and referenced to the solvent. The thermogra-
vimetric analyses were performed by using a TGA Q500 TA instrument
with a ramp of 108CminÀ1 under N2 from 100 to 9008C. Raman spectra
were acquired with a Renishaw instrument, model Invia reflex equipped
with 532, 632.8, and 785 nm lasers. The spectra were registered with the
laser at 532 nm. After acquisition, the spectra were normalized with re-
spect to the G band and then the amplitude of the D peak was calculat-
ed. TEM was performed on a Hitachi H600 microscope and a Philips 208
working at different accelerating voltage and at different magnification.
As a general procedure to perform TEM analyses, samples (0.1 mg) were
dispersed in 0.1 mL methanol/water (1:1) by ultrasonication for 5 min
and kept for 10–12 h. The solution was again ultrasonicated for 5 min
before depositing and 10 mL was deposited onto a holey-carbon TEM
grid and dried. The images are typical and representative of the samples
under observation. HRTEM was performed under a JEOL 2000FX-II
operating at 200 kV. AFM was carried out by using an Agilent Technolo-
gies Atomic Force Microscope (Model 5500) operating in Non contact/
ACAFM mode. Microfabricated silicon nitride cantilevers with a spring
constant (C) of approximately 50 NmÀ1 and resonant frequency (f) of
178 kHz were used. The average dimension thickness (T), width, and
length (L) of cantilever was approximately 670, 37, and 228 mm, respec-
tively. The AFM tips (PPP-NCL-10) with a radius of less than 7 nm from
NANOSENSORS were used for the different experiments. Data acquisi-
tion and analysis was carried out by using PicoView 1.4 and Pico Image
Basic software, respectively. As a general procedure to perform AFM
analyses, samples (0.1 mg) were dispersed in methanol/water (1:1, 0.1 mL)
by ultrasonication for 5 min and kept for 10–12 h. The solution was again
ultrasonicated for 5 min before depositing and 10 mL was deposited onto
a HOPG surface and dried under a UV lamp (100 W) for 15 min fol-
lowed by drying under vacuum for 2 h. For time-dependent experiments,
samples were kept for incubations for different time intervals. Then,
10 mL of each incubated sample was deposited on freshly cleaved HOPG
followed by AFM imaging. Solid-state NMR spectroscopic experiments
were conducted on an AVANCE 500 MHz wide bore spectrometer
(Bruker, Wissembourg, France) operating at a frequency of 125.7 MHz
for 13C and equipped with a triple resonance MAS probe designed for
3.2 mm zirconia rotors (closed with Kel-F caps). All the samples were
spun at 15 kHz spinning frequency. All cross-polarization 1H/13C experi-
ments were carried out with a proton pulse of 3.12 ms, a spin-lock field of
80 for proton and 60 kHz for carbon, a decoupling field of 100 kHz, a
1.2 ms contact time, and a recycle time of 5 s. Owing to the spectral wide
lines and to get undistorted lineshapes we also made use of Hahnꢃs echo
Figure 7. Proposed model for the formation of the nanorings, isolated or
interconnected in a chain. a) and b) The models are overlapped on
images shown in Figure 6. c) Model for the stepwise elongation of the
chain of the nanorings.
represents one bundle of SWCNTs;
rep-
resents the uracil molecules after the tip functionalization of SWCNTs.
represent one bundle of carbon nanotubes and on both ends
of these bars the small dark-gray-colored rings represent the
uracil molecules. An isolated nanoring could be imagined if
two black-colored bars come close to each other and the
uracil molecules on both ends interact to form a U···U
dimer. Similarly, two black-colored bars interact in a head-
to-head, tail-to-tail, or head-to-tail fashion (supposing that
there is no specific head and tail part differentiation, being
similar on both sides) to give the S-shaped structure (two
bundles of nanotubes in a S shape; Figure 7c). The elonga-
tion of the chain of the nanorings occurred when another S-
shaped structure or dimer interacts with each other. The
upper part of first S-dimer interacts with the middle part of
a second S-dimer and simultaneously, the lower part of the
second S-dimer automatically comes closed to the middle
part of first S-dimer for interaction.
Conclusion
In summary, uracil nucleobases have been covalently at-
tached mainly at the tips of the SWCNTs. The nucleobase–
SWCNT hybrids have been characterized by different spec-
troscopic and microscopic techniques. Solid-state MAS
NMR has been successfully applied to support the covalent
functionalization. The uracil functionalized single-walled
carbon nanotube hybrids have been reported to form nano-
ACHTUNGTRENNUNGring structures that result in the formation of aligned helical
morphology characterized through AFM and HRTEM tech-
niques. The height-diameter profile of these nanorings sug-
gests that they are constituted of different small bundles of
SWCNTs. We believe that the formation mechanism of the
self-assembled uracil-modified SWCNTs may due to various
factors, including U···U dimer interaction at the tips, ultra-
6778
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6772 – 6780

Carbon Nanotube–Nucleobase Hybrids
FULL PAPER
pulse sequence synchronized with the rotation (echo time=n rotation pe-
riods). The total echo time was kept identical in all spectra and equal to
two rotation periods (t=66.7 ms).
atomic absorption analyses, Dr. K.B. Joshi for his precious help during
AFM measurements and Prof. M. Sano for helpful discussions.
Preparation of SWCNT-COOH: Pristine HiPco SWCNTs (100 mg) were
suspended in HNO3 (75 mL, 3m) by sonication. The mixture was refluxed
for about 48 h, sonicated for 1 h, and refluxed again for another 48 h.
Then, HNO3 acid (25 mL, 3m) was added and after sonication for 2 h,
the mixture was again refluxed for 12 h. The resultant suspension was
then diluted by deionized water, filtered through a polycarbonate filter
(Isopore, pore size: 100 nm), and rinsed thoroughly with deionized water
several times until the pH value was ~7. The resulting SWCNTs were re-
suspended in deionized water and sonicated for 5 min. The suspension
was then filtered again. The black product obtained was dried and char-
acterized by TEM, AFM, and TGA.
[1] a) W. Y. Kim, Y. C. Choi, K. S. Kim, J. Mater. Chem. 2008, 18, 4510–
4521; b) M. J. Allen, V. C. Tung, R. B. Kaner, Chem. Rev. 2010, 110,
132–145.
[2] a) M. E. Pearce, J. B. Melanko, A. K. Salem, Pharm. Res. 2007, 24,
2335–2352; b) M. F. Beaux II, D. N. McIlroy, K. E. Gustin, Expert
Opin. Drug Delivery 2008, 5, 725–735; c) A. K. Shalek, J. T. Robin-
son, E. S. Karp, J. S. Lee, D.-R. Ahn, M.-H. Yoon, A. Sutton, M. Jor-
golli, R. S. Gertner, T. S. Gujral, G. MacBeath, E. G. Yang, H. Park,
Proc. Natl. Acad. Sci. USA 2010, 107, 1870–1875.
[3] a) M. Monthioux, V. L. Kuznetsov, Carbon 2006, 44, 1621–1623;
b) S. Iijima, Nature 1991, 354, 56–58.
Preparation of f-SWCNT 1:
A suspension of ox-SWCNT (10 mg,
SWCNT-COOH) in oxalyl chloride (4 mL) was stirred at 628C for 24 h
under an Ar atmosphere. The excess of oxalyl chloride was evaporated
under vacuum to give SWCNT-COCl. Boc-NHACHTNURTGNE(UNG CH2CH2O)2-CH2CH2NH2
[4] a) S. Iijima, T. Ichihashi, Nature 1993, 363, 603–605; b) D. S. Be-
thune, C. H. Klang, M. S. de Vries, G. Gorman, R. Savoy, J. Vazquez,
R. Beyers, Nature 1993, 363, 605–607.
(60 mg) in dry THF (10 mL) was added to the resulting mixture, which
was then heated at reflux for 48 h. After cooling at room temperature,
the solvent was evaporated under vacuum. The black suspension of meth-
anol was suspended in diethyl ether to remove excess of protected
amine. The black precipitate separates out after some time and then sus-
pension was centrifuged. The resulting Boc-protected SWCNTs were
dried at room temperature under vacuum. Boc-protected SWCNTs
(7 mg) were suspended in HCl (3 mL of 4m) in dioxane and stirred at
room temperature for 12 h under an Ar atmosphere to cleave the Boc
group at the chain end. The solvent was evaporated. f-SWCNTs 1 were
washed with dichloromethane and methanol several times and finally
with diethyl ether and dried under vacuum. The amount of functional
groups per gram of f-SWCNT 1 (loading) was measured with the quanti-
tative Kaiser test (0.58 mmolgÀ1). The nanotubes were characterized by
TEM, AFM, and TGA.
[5] a) Special issue: Carbon nanotubes (Ed.: R. C. Haddon), Acc.
Chem. Res. 2002, 35, Issue 12; b) Carbon Nanotubes and Related
Structures: Synthesis, Characterization, Functionalization, and Appli-
cations (Eds.: D. M. Guldi, N. Martin), Wiley-VCH, Weinheim,
2010.
[6] P. Singh, S. Campidelli, S. Giordani, D. Bonifazi, A. Bianco, M.
Prato, Chem. Soc. Rev. 2009, 38, 2214–2230.
[7] a) C. Menard-Moyon, E. Venturelli, C. Fabbro, C. Samori, T. Da
Ros, K. Kostarelos, M. Prato, A. Bianco, Expert Opin. Drug Discov-
ery 2010, 5, 691–707; b) H.-C. Wu, X. Chang, L. Liu, F. Zhaoa, Y.
Zhao, J. Mater. Chem. 2010, 20, 1036–1052.
[8] Carbon Nanotubes: Advanced Topics in the Synthesis, Structure,
Properties and Applications (Eds.: A. Jorio, G. Dresselhaus, M. S.
Dresselhaus), Springer, Berlin, 2008.
[9] a) M. L. Terranova, V. Sessa, S. Orlanducci, M. Rossi, D. Manno, G.
Micocci, Chem. Phys. Lett. 2004, 388, 36–39; b) N. Nakashima, Y.
Tanaka, Y. Tomonari, H. Murakami, H. Kataura, T. Sakaue, K.
Yoshikawa, J. Phys. Chem. B 2005, 109, 13076–13082.
[10] T. S. Jespersen, J. Nygꢄrd, Nano Lett. 2005, 5, 1838–1841.
[11] a) M. Lu, H.-L. Li, K.-T. Lau, J. Phys. Chem. B 2004, 108, 6186–
6192; b) S. Amelinckx, X. B. Zhang, D. Bernaerts, X. F. Zhang, V.
Ivanov, J. B. Nagy, Science 1994, 265, 635–639.
[12] a) G. Cuniberti, J. Yi, M. Porto, Appl. Phys. Lett. 2002, 81, 850–852;
b) H. Watanabe, C. Manabe, T. Shigematsu, M. Shimizu, Appl. Phys.
Lett. 2001, 78, 2928–2930.
Preparation of f-SWCNT 2: Boc-deprotected f-SWCNT
1 (10 mg,
5.8 mmol, based on the loading calculated with the quantitative Kaiser
test) was dissolved in DMF (2 mL) and neutralized with DIEA (diisopro-
pylethylamine; 1.0 mL, 5.8 mmol). A solution of compound 3-(1-uracilyl)-
propionic acid (21.3 mg, 116 mmol) in a mixture of CH2Cl2/DMF (1:1,
2 mL) was activated with EDC·HCl (33.2 mg, 174 mmol) and HOBt
(23.5 mg, 174 mmol) for 1 h under argon and subsequently added to the
carbon nanotube solution. The resulting mixture was stirred for 48 h at
room temperature and the reaction mixture was then filtered on a poly-
carbonate filter (Fluoropore, pore size 100 nm). The black solid collected
on the filter was redissolved in DMSO and filtered again. This procedure
was repeated two times each with DMSO and CH2Cl2/MeOH (10:1) mix-
ture to remove the excess of uracil derivative. After final washings with
the diethyl ether, the black solid was dried under vacuum to afford
7.2 mg of f-SWCNT 2. The nanotubes were characterized by TEM,
AFM, and TGA.
[13] a) R. C. Haddon, Nature 1997, 388, 31–32; b) H. R. Shea, R. Martel,
P. Avouris, Phys. Rev. Lett. 2000, 84, 4441–4444.
[14] a) L. Song, L. Ci, C. Jin, P. Tan, L. Sun, W. Ma, L. Liu, D. Liu, Z.
Zhang, Y. Xiang, S. Luo, X. Zhao, J. Shen, J. Zhou, W. Zhou, S. Xie,
Nanotechnology 2006, 17, 2355–2361; b) D. Grimm, P. Venezuela, F.
Banhart, N. Grobert, H. Terrones, P. M. Ajayan, M. Terrones, A.
Latg, Small 2007, 3, 1900–1905.
[15] a) J. Liu, H. Dai, J. H. Hafner, D. T. Colbert, R. E. Smalley, S. J.
Tans, C. Dekker, Nature 1997, 385, 780–781; b) M. Ahlskog, E. Sey-
naeve, R. J. M. Vullers, C. V. Haesendonck, A. Fonseca, K. Hernadi,
J. B. Nagy, Chem. Phys. Lett. 1999, 300, 202–206.
[16] a) L. Song, L. Ci, L. Sun, C. Jin, L. Liu, W. Ma, D. Liu, X. Zhao, S.
Luo, Z. Zhang, Y. Xiang, J. Zhou, W. Zhou, Y. Ding, Z. Wang, S.
Xie, Adv. Mater. 2006, 18, 1817–1821; b) J. F. Colomer, L. Henrard,
E. Flahaut, G. Van Tendeloo, A. A. Lucas, P. Lambin, Nano Lett.
2003, 3, 685–689.
Preparation of f-SWCNT 3: SWCNT-COCl (10 mg), prepared in turn as
reported above, was suspended in a solution of 1-(2-aminoethyl)uracil
(18 mg, 116 mmol) and DIEA (10.16 mL, 34.8 mmol) in dry THF (15 mL).
The resulting suspension was heated under reflux for 48 h. After cooling
to room temperature and removing excess of the uracil derivative by
washing several times with DMF, methanol, and finally with diethyl
ether, the resulting f-SWCNT 3 was dried at room temperature under
vacuum to afford 8.3 mg of f-SWCNT 3. The nanotubes were character-
ized by TEM, AFM, and TGA.
[17] M. Sano, A. Kamino, A. Okamura, S. Shinkai, Science 2001, 293,
1299–1301.
[18] Y. Wang, D. Maspoch, S. Zou, G. C. Schatz, R. E. Smalley, C. A.
Mirkin, Proc. Natl. Acad. Sci. USA 2006, 103, 2026–2031.
[19] a) R. Martel, H. R. Shea, P. Avouris, Nature 1999, 398, 299–299;
b) R. Martel, H. R. Shea, P. Avouris, J. Phys. Chem. B 1999, 103,
7551–7556.
[20] a) R. R. Sinden, DNA Structure and Function, Academic Press, San
Diego, 1994; b) Oxford Handbook of Nucleic Acid Structure (Ed.: S.
Neidle), Oxford University Press, New York, 1999; c) J. Sꢅhnel, Bio-
Acknowledgements
The work was also financed by the CEFIPRA/IFCPAR (project no.
3705–2), the University of Trieste, and INSTM, MIUR (cofin Prot.
20085M27SS). P.S. wishes to thank CEFIPRA/IFCPAR for a postdoctoral
fellowship. TEM images were recorded at the RIO Plate-form of Espla-
nade (Strasbourg, France). We greatly thank Dr. G. Mitaritonna for
Chem. Eur. J. 2011, 17, 6772 – 6780
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6779

A. Bianco, S. Verma et al.
polymers 2002, 61, 32–51; d) S. Verma, A. K. Mishra, J. Kumar, Acc.
Chem. Res. 2010, 43, 79–91.
2007, 46, 3762–3764; b) J. V. Barth, G. Costantini, K. Kern, Nature
2005, 437, 671–679.
[21] a) R. E. A. Kelly, L. N. Kantorovich, J. Mater. Chem. 2006, 16, 1894–
1905; b) G. Gottarelli, S. Masiero, E. Mezzina, S. Pieraccini, J. P.
Rabe, Chem. Eur. J. 2000, 6, 3242–3248; c) S. Xu, M. Dong, E.
Rauls, R. Otero, T. R. Linderoth, F. Besenbacher, Nano Lett. 2006,
6, 1434–1438.
[22] a) P. Singh, J. Kumar, F. M. Toma, J. Raya, M. Prato, B. Fabre, S.
Verma, A. Bianco, J. Am. Chem. Soc. 2009, 131, 13555–13562;
b) M. Quintana, M. Prato, Chem. Commun. 2009, 6005–6007.
[23] a) M. Zheng, A. Jagota, E. D. Semke, B. A. Diner, R. S. Mclean,
S. R. Lustig, R. E. Richardson, N. G. Tassi, Nat. Mater. 2003, 2, 338–
342; b) Y. Wang, J. Phys. Chem. C 2008, 112, 14297–14305.
[24] a) N. Varghese, U. Mogera, A. Govindaraj, A. Das, P. K. Maiti,
A. K. Sood and C. N. R. Rao, ChemPhysChem 2009, 10, 206–210;
b) C. Zhao, Y. Peng, Y. Song, J. Ren, X. Qu, Small 2008, 4, 656–661.
[25] J. A. Frey, A. Muller, M. Losada, S. Leutwyler, J. Phys. Chem. B
2007, 111, 3534–3542.
[26] a) T. O. Soine, M. R. Buchdahl, Org. Synth. 1952, 32, 18; b) B. K.
Das, N. Shibata, Y. Takeuchi, J. Chem. Soc. Perkin Trans. 1 2002,
197–206.
[27] R. Graupner, J. Raman Spectrosc. 2007, 38, 673–683.
[28] R. Marega, V. Aroulmoji, F. Dinon, L. Vaccari, S. Giordani, A.
Bianco, E. Murano, M. Prato, J. Am. Chem. Soc. 2009, 131, 9086–
9093.
[31] S. Sivakova, S. J. Rowan, Chem. Soc. Rev. 2005, 34, 9–21.
[32] a) A. Jatsch, A. Kopyshev, E. Mena-Osteritz, P. Bauerle, Org. Lett.
2008, 10, 961–964; b) S. Piana, A. Bilic, J. Phys. Chem. B 2006, 110,
23467–23471; c) W. M. Heckl, D. P. Smith, G. Binnig, H. Klagges,
T. W. Hansch, J. Maddocks, Proc. Natl. Acad. Sci. USA 1991, 88,
8003–8005.
[33] a) M. Fathalla, C. M. Lawrence, N. Zhang, J. L. Sessler, J. Jayawick-
ramarajah, Chem. Soc. Rev. 2009, 38, 1608–1620; b) W. Xu, R. E. A.
Kelly, R. Otero, M. Schock, E. Loegsgaard, I. Stensgaard, L. N.
Kantorovich, F. Besenbacher, Small 2007, 3, 2011–2014; c) W. Mam-
douh, R. E. A. Kelly, M. Dong, L. N. Kantorovich, F. Besenbacher,
J. Am. Chem. Soc. 2008, 130, 695–702.
[34] A. M. S. Kumar, J. D. Fox, L. E. Buerkle, R. E. Marchant, S. J.
Rowan, Langmuir 2009, 25, 653–656.
[35] a) A. M. S. Kumar, S. Sivakova, J. D. Fox, J. E. Green, R. E. March-
ant, S. J. Rowan, J. Am. Chem. Soc. 2008, 130, 1466–1476; b) I.
Bestel, N. Campins, A. Marchenkoc, D. Fichouc, M. W. Grinstaffd,
P. Barthꢀlꢀmy, J. Colloid Interface Sci. 2008, 323, 435–440.
[36] S. J. Sowerby, C. A. Cohn, W. M. Heckl, N. G. Holm, Proc. Natl.
Acad. Sci. USA 2001, 98, 820–822.
[37] a) O. Zhou, R. M. Fleming, D. W. Murphy, C. H. Chen, R. C.
Haddon, A. P. Ramirez, S. H. Glarum, Science 1994, 263, 1744–
1747; b) M. R. Falvo, G. J. Clary, R. M. Taylor, V. Chi, F. P.
Brooks Jr, S. Washburn, R. Superfine, Nature 1997, 389, 582–584.
[29] S.-S. Li, H.-J. Yan, L.-J. Wan, H.-B. Yang, B. H. Northrop, P. J.
Stang, J. Am. Chem. Soc. 2007, 129, 9268–9269.
[30] a) M. G. Badin, A. Bashir, S. Krakert, T. Strunskus, A. Terfort, C.
Woll, Angew. Chem. 2007, 119, 3837–3839; Angew. Chem. Int. Ed.
Received: January 28, 2011
Published online: May 3, 2011
6780
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6772 – 6780