A New, Simple and Versatile Strategy for the Synthesis of Short Segments of Zigzag-Type Carbon Nanotubes

Article abstract of DOI:10.1002/chem.201503693

Short segments of zigzag single-walled carbon nanotubes (SWCNTs) were obtained from a calixarene scaffold by using a completely new, simple and expedited strategy that allowed fine-tuning of their diameters. This new approach also allows for functionalise

Full text of DOI:10.1002/chem.201503693

DOI: 10.1002/chem.201503693
Full Paper
&
Synthesis Design
A New, Simple and Versatile Strategy for the Synthesis of Short
Segments of Zigzag-Type Carbon Nanotubes
Etienne AndrØ,[a] Baptiste Boutonnet,[a] Pauline Charles,[a] Cyril Martini,[a] Juan-
Manuel Aguiar-Hualde,[b] Sylvain Latil,[b] Vincent GuØrineau,[c] Karim Hammad,[c]
Priyanka Ray,[a] RØgis Guillot,[a] and Vincent Huc*[a]
Abstract: Short segments of zigzag single-walled carbon
nanotubes (SWCNTs) were obtained from a calixarene scaf-
fold by using a completely new, simple and expedited strat-
egy that allowed fine-tuning of their diameters. This new
approach also allows for functionalised short segments of
zigzag SWCNTs to be obtained; a prerequisite towards their
lengthening. These new SWCNT short segments/calixarene
composites show interesting behaviour in solution. DFT
analysis of these new compounds also suggests interesting
photophysical behaviour. Along with the synthesis of various
SWCNTs segments, this approach also constitutes a powerful
tool for the construction of new, radially oriented p systems.
Introduction
wards the synthesis of zigzag CNTs are based on cyclacene-
type repeating units as a starting point (Figure 1A). Because
the synthesis of these long-pursued compounds is still elusive,
we decided to consider a different approach.
The rediscovery[1] of carbon nanotubes (CNTs) by Ijima in
1991,[2] and most importantly the discovery of single-walled
carbon nanotubes (SWCNTs) by Ijima and Ichihashi[3] and
Kiang[4] in 1993, opened up a new world in chemistry and
physics, especially regarding electron transport characteristics
at the nanoscale.[5–7] It was shown that the electronic behav-
iour of SWCNTs was directly related to the way in which the
constituting graphene sheet was rolled.[8] So far, most known
syntheses of CNTs offer a limited degree of control over this
rolling process.[9–12] Post-synthesis sorting processes are thus
required.[13–16] A radical solution would be the total synthesis of
CNTs through synthetic organic chemistry.
With few exceptions,[17–19] most of the experimental ap-
proaches developed so far for the synthesis of SWCNTs are di-
rected towards the synthesis of armchair-type nanotubes.[20–38]
All of these tubes will be metallic, regardless of their diame-
ter.[8] Conversely, a diameter-controlled strategy, for the synthe-
sis of zigzag CNTs would open up the way to both metallic
and semiconducting nanotubes. Indeed, it was shown that the
electronic properties of zigzag nanotubes were solely depen-
dent on their diameters.[8] Most of the proposed strategies to-
Figure 1. Topological representation of cyclacene (A) and cyclo[n](m-
phenylene) (B) repeating units for the total synthesis of zigzag SWCNTs.
Analysis of the structure of zigzag CNTs led us to conclude
that a different repeating unit, namely, cyclo[n](m-phenylene)
(CMP),[39–42] could also be considered as an interesting starting
point (Figure 1B).
[a] Dr. E. AndrØ, Dr. B. Boutonnet, P. Charles, Dr. C. Martini, Dr. P. Ray,
Dr. R. Guillot, Dr. V. Huc
Institut de Chimie MolØculaire et MatØriaux d’Orsay (ICMMO)
15 rue Georges ClØmenceau, 91405 Orsay (France)
E-mail: Vincent.huc@u-psud.fr
Reports in the literature (along with molecular modelling)
shows that CMPs exhibit planar or saddle-like geometries.[39–43]
This definitively precludes their direct use as a starting point
for the synthesis of radially conjugated compounds, such as
CNTs. It is thus necessary to constrain the geometry of these
macrocycles to force them to adopt a “tubular” geometry,
closer to the final radially oriented p system of the target
CNTs. Herein, we report the first experimental steps towards
[b] Dr. J.-M. Aguiar-Hualde, Dr. S. Latil
CEA Saclay - Orme des Merisiers
91191 Gif-sur-Yvette Cedex (France)
[c] V. GuØrineau, K. Hammad
Institut de Chimie des Substances Naturelles (ICSN)
CNRS, 91300 Gif-Sur-Yvette Cedex (France)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201503693.
Chem. Eur. J. 2016, 22, 3105 – 3114
3105
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
the synthesis of zigzag CNTs with a priori well-defined
diameters by using this completely new CMP-based approach.
An extensive NMR spectroscopy analysis of these new calix-
arene-scaffolded zigzag SWCNT short segments demonstrated
their cyclic structure. It also showed that these compounds
exhibited complex dynamic behaviour in solution. These
structural characterisations were supported by DFT analysis.
Results and Discussion
Our strategy relies on the use of conformationally locked
calix[4]arene, calix[5]arene and calix[6]arene as scaffolds to
build the corresponding zigzag SWCNTs (Figure 2).
Figure 4. Two-step synthesis of a cyclo[10](m-phenylene)-type SWCNT
segment on a calix[5]arene scaffold (9). Reagent and conditions: a) 2,
CH3CN, K2CO3, overnight; b) 5, [Pd(PPh3)4], CsF, dioxane, reflux, 6 h.
Figure 2. The calixarene/SWCNT composites described herein.
Figure 3. Two-step synthesis of the cyclo[8](m-phenylene)-type SWCNT
segment on a calix[4]arene scaffold (6). Reagent and conditions: a) THF/
DMF/NaH, À788C!RT, overnight; b) 5, [Pd(PPh3)4], CsF, dioxane, reflux, 10 h.
Figure 5. Five-step synthesis of the cyclo[12](m-phenylene)-type SWCNT
segment from a calix[6]arene scaffold (15). Reagent and conditions: a) tri-
ethyleneglycol bis(tosylate), tBuOK, toluene, 1008C overnight; b) MeI, DMF,
RT, overnight;[45] c) Pd(OH)2/C, H2, 48 h; d) 2, THF/DMF 508C, overnight; e) 5,
[Pd(PPh3)4], CsF, dioxane, reflux, 6 h.
As shown in Figures 3, 4, and 5, the proposed strategy is
general and has been successfully applied to different calixar-
ene platforms. It relies on the synthesis of key poly-halogenat-
ed scaffolds of different diameters (compounds 3, 8 14 and
19) starting from the corresponding calixarenes 1, 7 and 10.
The poly-halogenated calixarene scaffolds 3, 8, 14 and 19
are then subjected to cascade Suzuki couplings to introduce
the missing aromatic units required for the target SWCNT short
segments.[2]
Synthesis of the poly-halogenated calixarene scaffolds
Regarding the synthesis of 6 (Figure 3), calixarene scaffold 3 is
obtained from commercially available p-tert-butylcalix[4]arene
1 by NaH-promoted alkylation, by using the commercially
available 3,5-(dibromo)(bromomethyl)benzene (2).[45]
Chem. Eur. J. 2016, 22, 3105 – 3114
www.chemeurj.org
3106
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
We observed that the proportion of expected cone deriva-
tive 3 (Figure S1a–d in the Supporting Information) versus par-
tial cone derivative 4 (Figure S2a–d in the Supporting Informa-
tion) dramatically increased at low temperature. Starting the
reaction at À788C resulted in 90% yield of scaffold 3. This
ment (6), only a combination of CsF in dioxane, along with
[Pd(PPh3)4][48] finally allowed us to obtain 6 in 4% yield (Fig-
ure S4 in the Supporting Information). Surprisingly, by using
the same conditions with the calix[5] and -[6]-derived scaffolds
8 and 14, no detectable amounts of the expected segments 9
and 15 were obtained. A second extensive optimisation proce-
dure finally allowed us to obtain these compounds in 4 and
1% yield, respectively. Albeit low, these yields compare with
those described for the first synthesis of cyclo-p-poly(pheny-
lene)s by Jasti et al.[20] (ranging from 0.2 to 1.4%). Interestingly,
the catalytic conditions optimised for 9 and 15 are quite differ-
ent from those used to obtain 6. A 10-fold increase in the
amount of catalyst proved to be necessary (to accelerate the
reaction; thus reducing debrominations/deborylations), along
with a 10-fold dilution of the reaction medium (to reduce the
number of Suzuki over-couplings; see Figures S4j, S11s and S6r
in the Supporting Information). However, once optimised, the
purification of 9 and 15 is surprisingly easy, considering that
more than 50 calixarene-derived by-products are observed in
both cases in the crude samples (Figures S6q and S11p in the
Supporting Information). Purification only involves three suc-
cessive washing steps with suitable solvents and a final prepa-
rative TLC step (Figures S6a and S11a in the Supporting Infor-
mation). Interestingly, this purification protocol is very similar
in both cases, despite the very different functionalisation
patterns of 9 and 15. This shows that the behaviour of these
compounds during purification is mostly determined by the
upper SWCNT short segment annulus. This may be due to the
rigidification effect associated with the presence of the closed
SWCNT upper segment, in sharp contrast with the
more flexible acyclic by-products.
1
compound was fully characterised by H/13C NMR spectroscopy
and mass spectrometry (Figure S1a–d in the Supporting Infor-
mation). The 1H NMR spectrum is especially informative: the
cone conformation of 3 is demonstrated by the presence of
a characteristic pair of doublets (at d=2.8 and 3.95 ppm),
which corresponds to the bridging methylene groups.
Regarding the synthesis of 9 (Figure 4), the key cone-locked
decabromo derivative scaffold 8 was obtained by direct alkyl-
ation of the starting p-(benzyloxy)calix[5]arene[46,47] in aceto-
nitrile. The presence of the cone conformer was demonstrated
by both 1H NMR spectroscopy experiments (Figure S5 in the
Supporting Information) and single-crystal XRD analysis
(Figure 6 and Figures S5a and S19 in the Supporting
Information).
Figure 6. Structural determination of the (decabromo)calix[5]arene scaffold 8
by X-ray diffraction (R=benzyloxy). Colour code: H, white; C, black; O, red;
Br, brown.
Characterisation of the zigzag SWCNT short segments
Compounds 6, 9, 15 and 20 are the first examples of diameter-
controlled short segments of zigzag SWCNTs. They were fully
characterised by mass spectrometry and a combination of 1D
and 2D NMR spectroscopy experiments (see Figures S4a–i,
S6a–r, S11a–r and S13–S17 in the Supporting Information). The
1H NMR spectra of 9, 6 and 12 are especially exotic, and clearly
demonstrate the tubular structure of these products (Figures 7
to 14, below).
Regarding the synthesis of the calix[6]arene-derived SWCNT
short segment 15, (Figure 5), the synthesis of the dodecabro-
minated scaffold 14 was more complex than the two previous
cases. Indeed, locking the starting p-(benzyloxy)calix[6]arene
10[46,47] in the cone conformation requires the two-step pre-
organisation sequence depicted in Figure 5.[44] Compound 10
was first alkylated by triethyleneglycol bis(tosylate) (step a in
Figure 5), and then by methyl iodide (step b in Figure 5).
During this second alkylation process, a conformational switch
is observed from alternate (11) to cone (12).[44] Debenzylation
of the resulting cone-shaped platform 12 quantitatively result-
ed in 13, which was then alkylated in nearly quantitative yields
with commercially available 2. The resulting (dodecabromo)-
calix[6]arene precursor 14 is easily obtained on a 10 g scale
(Figure S10 in the Supporting Information).
1
Regarding segment 6, the H NMR spectrum is symmetric, as
expected. The preservation of the cone conformation of the
lower calixarene stage is evidenced by the presence of charac-
teristic bridging methylene doublets. This implies that the
upper cyclophenylene unit adopts a tubular shape. Very high
chemical shifts are observed for the bridging methylene
protons (Figure 7, green hydrogen atoms). Those protons are
observed at d=5 and 3.7 ppm, which is shifted by +0.5 ppm
relative to most previously described calix[4]arene derivatives
in the same cone conformation.[45] This effect is attributed to
the convergent influences of the deshielding cone of the three
surrounding phenyl units (green arrows).
Assembly of the SWCNT short segments by cascade Suzuki
couplings
The final Suzuki-type coupling step required extensive
optimisation work with all of the previously described poly-
brominated scaffolds. Regarding the first SWCNT short seg-
Conversely, the methylene protons from the benzyl groups
(Figure 7, blue hydrogen atoms) are observed at d=4.2 ppm,
which is shielded by nearly 1 ppm compared with its usual
Chem. Eur. J. 2016, 22, 3105 – 3114
www.chemeurj.org
3107
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 7. A) 1H NMR spectra of 6 (solvents marked with asterisks) and B) in-
fluence of the presence of the cyclo[8](m-phenylene) stage on the chemical
shifts of the calixarene protons. The magnetic influence of the aromatic
groups on the chemical shifts of some protons is indicated by arrows.
Figure 8. A) 1H NMR spectrum of the SWCNT short segment/calixarene com-
posite 9 (signals from the self-included phenyl group are marked with red
asterisks; inset: magnification of these signals). B) COSY analysis provides
evidence of coupling between the protons of the included phenyl group.
C) The self-inclusion phenomenon affecting one benzyloxy group is
highlighted by the blue arrow. D) In-plane projection of 9 (top view), high-
lighting its symmetry plane (R=benzyloxy). The self-included benzyloxy
groups is highlighted in blue.
value (dꢀ5 ppm). This effect is attributed to the proximity of
the shielding cone of the aromatic groups from the calix-
[4]arene lower stage (Figure 7, blue arrow).
1
A third piece of structural information provided by H NMR
spectroscopy is the deshielding effect observed for the aromat-
ic hydrogen atoms from the calixarene scaffold (Figure 7, red
hydrogen atoms). These are shifted by +0.5 ppm compared
with their usual values (dꢀ7 ppm), due to the reduction of
the cone angle of the calixarene lower stage upon the forma-
tion of the upper m-(phenylene) macrocycle. The two aromatic
protons are then brought in close proximity, which increases
the influence of the deshielding cone from the neighbouring
aromatic moiety (Figure 7, red arrows). Lastly, NOESY analysis
(Figures S4g–i in the Supporting Information) also confirmed
the cylindrical shape of 6. Indeed, the NOESY spectrum shows
the expected through-space couplings between the protons
from the SWCNT stage.
This interpretation was first confirmed by COSY analysis (Fig-
ure 8B and Figure S6j in the Supporting Information), which
showed that these three signals were indeed coupled. Lastly,
HSQC analysis (Figure S6i in the Supporting Information)
shows that these three signals are associated with phenyl-type
carbon atoms (characteristic chemical shifts at dꢀ130 ppm).
This upfield shift phenomenon is characteristic of an
inclusion process inside the calixarene cavity. Indeed, it was
shown that such an inclusion process places the guest in the
convergent shielding cone of all the aromatic groups of the
calixarene host.[49] As the intensity of the signals of the includ-
ed aromatic group corresponds to exactly one aromatic group
(Figure 8A), this led us to conclude that one of the benzyloxy
legs of the molecule was self-included inside the lower-stage
calixarene cavity (Figure 8C, blue).
1
Regarding segment 9, the H NMR spectrum is surprisingly
complex (Figure 8A and Figure S6c in the Supporting Informa-
tion). This shows that a symmetry-lowering effect occurs,
which is not observed with 6. Careful examination of the spec-
trum shows the presence of a characteristic pattern of three
signals tentatively associated with an aromatic group, but
shifted upfield by 2 ppm compared with the others aromatic
groups (Figure 8A, red asterisks (magnification in the inset)).
This self-inclusion phenomenon results in a symmetry plane
1
crossing the molecule (Figure 8D, red dotted line). H/13C NMR
Chem. Eur. J. 2016, 22, 3105 – 3114
www.chemeurj.org
3108
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
spectroscopy analyses are also fully supported by HMBC (Fig-
ure S6n in the Supporting Information) and ROESY (Figure S6o
and p in the Supporting Information) experiments.
The identity and purity of the product were further con-
firmed by mass spectrometry. Figure S6q in the Supporting
Information shows the MALDI mass spectra (Cs+ cationised) of
the same sample as that used for the previously described
NMR spectroscopy experiments. Only one single signal is ob-
served. Accordingly, TLC analysis of this sample shows only
a single spot.
The self-inclusion phenomenon of a benzyloxy group inside
the calixarene cavity of 9 is not observed for 8 because, even
if the latter exists in the cone conformation, it is still quite
a flexible compound. This results in large fluctuations of the di-
ameter/shape of the calix[5]arene cavity. This reduces fitting
between the aromatic cavity of the calixarene and benzyloxy
group; thus lowering the association constant.
In the case of 9, the calixarene structure is made extremely
rigid by the presence of the SWCNT short segment. This pre-or-
ganises and reduces the size of the cavity, and thus, reinforces
the inclusion of the benzyloxy leg.
This phenomenon can be considered as characteristic of the
rigidification of the calixarene cavity (i.e., of the successful
assembly of the SWCNT short segment), and thus, confirms the
structure of 9.
Regarding the SWCNT short segment/calixarene composite
1
15, analysis of the H NMR spectrum proved to be more diffi-
Figure 9. Expected (A) and observed (B) patterns for the bridging methylene
area of 9 (calixarene scaffold) in the H NMR spectrum. The three doublets
1
cult and revealed complex dynamic behaviour in solution. The
ambient-temperature 1H NMR spectrum of 15 in CDCl3 is
shown in Figure 10A (see also Figure S11b in the Supporting
Information). Only broad, ill-defined features are observed.
We suggest that this phenomenon results from slow rota-
tions (at ambient temperature) of the upper nanotube seg-
ment versus the lower calixarene stage (Figure 10C, blue
arrows), due to rotation around the benzyloxy bonds. Indeed,
analysis of molecular models shows that the ArÀOÀCH2 benzyl-
ic bonds linking the two macrocyclic stages are almost free to
rotate (Figure 10C, green arrows). This quasi-absence of any
defined signals at ambient temperature made it extremely
difficult to track the product during the early design of its
purification procedure.
are highlighted with blue, red and green arrows. C) Expected pattern for the
benzyloxy groups. The blue and green circles highlight the two sets of
benzyloxy groups in 9. D) Magnification of the benzyloxy area of the HSQC
experiment. The blue/green arrows highlight the two benzyloxy groups that
appear as singlets. E) Expected pattern for the hydroquinone groups.
F) Magnification of the hydroquinone groups area of the HSQC experiment.
spectroscopy analysis of 9 is fully consistent with the afore-
mentioned results (Figure 9). Three doublets in an expected in-
tensity ratio of 2/2/1 (Figure 9A) are observed (Figure 9B) for
the bridging methylenes of the calixarene scaffold (highlighted
as blue, red and green arrows in Figure 9).
In the same way, six different benzyloxy-type groups are ex-
pected (Figure 9C) and observed (Figure 9D and Figure S6g in
the Supporting Information; 13C chemical shifts at dꢀ80 ppm).
Indeed, compound 9 exhibits two sets of such groups, each in-
cluding three inequivalent benzyloxy bonds. The first set links
the SWCNT stage to the calixarene one (Figure 9C, G1), and
the second one belongs to the calixarene stage (Figure 9C,
G2). Two of these benzyloxy groups appear as singlets (Fig-
ure 9C and D, green/blue arrows) because they are crossed by
the symmetry plane.
To confirm this interpretation, both low- and high-
temperature NMR experiments were undertaken. Due to the
limited stability of 15 at high temperatures (see below), we
first performed the NMR spectroscopy analysis at low tempera-
ture to slow down and stop the rotational motion depicted in
Figure 10C. As expected, we observed the progressive appear-
ance of sharp signals as the temperature decreased (Fig-
ure S11c in the Supporting Information), showing the existence
of a well-defined, low-symmetry conformation at 233 K (Fig-
ure 10B and Figure S11c in the Supporting Information). This
temperature was chosen for the subsequent low-temperature
analysis as the best compromise between different trends:
1) easily observed, sharp signals; 2) sufficient signal-to-noise
ratio; 3) no precipitation; and 4) no freezing of the solvent.
Indeed, at 233 K, the solution of 15 in chloroform is nearly sa-
turated, and we are close to the freezing point of this solvent.
Lastly, five hydroquinone-type protons are expected
(Figure 9E) and observed (Figure 9F and Figures S6h in the
Supporting Information), with characteristic 13C chemical shifts
falling in the d=112–120 ppm range. All of these analysis are
perfectly confirmed by COSY experiments (Figure S6k–m in the
Supporting Information), which in each case show the expect-
ed coupling patterns. All of these previously described NMR
Chem. Eur. J. 2016, 22, 3105 – 3114
www.chemeurj.org
3109
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 11. DFT-optimised geometry of 15 (side view (A) and top view (B)).
C) Simplified in-plane projection of 15 (top view), highlighting the twist
angle between the SWCNT short segment stage and the calixarene. Methoxy
groups are omitted for clarity and the symmetry axis lying in the centre of
the molecule is highlighted as a yellow cross (see the Supporting Informa-
tion, Figure S11a).
1
Figure 10. A) Ambient- and B) low-temperature H NMR spectroscopy
analyses of 15. C) Rotational motion of the two macrocyclic stages (blue
arrows) around the benzyloxy bonds (green arrows) at room temperature.
According to DFT calculations, the most stable conformer of
15 shows the calixarene stage and the SWCNT short segment
“twisted” together (Figure 11 A, side view, and B, top view, see
also Figure S18 in the Supporting Information). Fig-
ure 11 Cshows a simplified top view (i.e. in plane projection) of
this conformation, depicting this twist. This conformation
shows a pseudo-symmetry axis, crossing the centre of the mol-
ecule, and running perpendicular to the drawing (Figure 11 C,
yellow cross). At 233 K, NMR spectroscopy analysis is fully con-
sistent with this structure (Figure 12).
As expected, six doublets are observed for the bridging
methylenes belonging to the calixarene stage (Figure 12A and
B, red arrows; and Figure S11d and e in the Supporting Infor-
mation). This was confirmed by 2D COSY analysis (Figure S11d
and e in the Supporting Information), which showed the three
expected correlation rectangles.
In the same way, six doublets are expected for the three
inequivalent methylene groups belonging to the benzyloxy
moieties (Figure 12C). These signals are indeed observed
(Figure 11 D; see also the 2D COSY analysis, Figure S11d and f
in the Supporting Information).
Figure 12. 1H NMR spectroscopy analysis of 15 at 233 K. Symmetry-induced
equivalences (A, red arrows) to explain the three observed doublet of dou-
blets for the bridging methylenes from the calixarene (B). Symmetry-induced
equivalences (C, green arrows) to explain the sets of doublet of doublets ob-
served for the benzyloxy groups protons (D). E) Symmetry-induced equiva-
lences (E, blue arrows) to account for the six different hydroquinone signals
Lastly, six correlated signals are expected for the six hydro-
quinone protons belonging to the calixarene (Figure 12E (blue
arrows) and Figure S11g in the Supporting Information). These
signals are indeed observed (Figure 12F), and pairwise
correlated, as expected (see 2D COSY analysis, Figure S11d and
g in the Supporting Information).
1
observed in the H NMR spectrum (F). Red asterisks: signals belonging to
the protons from the SWCNT short segment upper stage.
Chem. Eur. J. 2016, 22, 3105 – 3114
www.chemeurj.org
3110
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Extra signals are also observed in the same area (Figure 12F,
red asterisks, and Figure S11h in the Supporting Information,
red asterisks/red arrows). Analysis of the edited HSQC (Fig-
ure S11j in the Supporting Information) and COSY 2D experi-
ments (Figure S11d in the Supporting Information) shows that
these signals belong to the m-[cyclopoly(phenylene)] ring
protons. The extensive overlapping of these signals and the
low symmetry of the molecule preclude a more precise
assignment.
The four methoxy groups of 15 are observed as two distinct
signals (as expected for symmetry reasons) at d=3.76 and
1
2.38 ppm on the H NMR spectrum (HSQC, Figure S11j in the
Supporting Information). The signals belonging to the PEG
chain are observed as a complex set of overlapping doublets
(Figure S11e and j in the Supporting Information). For all
relevant signals, the observed intensity ratios are in good
agreement with those expected.
The high-temperature analysis of 15 proved to be difficult
because this compound was quite unstable under these condi-
tions. Indeed, fast evolution of the signals was observed upon
heating in high-boiling solvents, such as [D6]DMSO, which was
indicative of rapid decomposition. However, compound 15
proved to be stable for a limited period of time (a few hours)
at 908C in [D7]DMF under argon (sealed NMR tube).
The evolution of the 1H NMR spectrum with temperature
under these conditions is shown in Figure 13. Interestingly, the
ambient-temperature spectrum in [D7]DMF (Figure 13, 299 K)
appears to be more resolved than the corresponding one in
CDCl3 (Figure 10A). This shows that the intramolecular motions
(sketched in Figure 10C, blue arrows) are faster in [D7]DMF. As
the temperature increases, the signals become sharper.
Figure 14. A) 1H NMR spectrum of 15 at 363 K. B) Magnification of the
bridging methylene area of the COSY NMR spectrum.
periment (Figure S11o in the Supporting Information) provides
1
evidence of all expected H/13C correlations. All other signals
are fully coherent with the proposed structure of 15
(Figure S11l in the Supporting Information). Unfortunately, the
limited stability of 15 at high temperature prevented us from
performing further analysis.
The identity and purity of 15 were further confirmed by
mass spectrometry. Only one signal is observed after purifica-
tion (Figure S11q and r in the Supporting Information), which
is in accordance with the observation of a single spot by TLC.
Importantly, the tracking of 15 was made easier by its
characteristic V-shaped spot on the silica gel TLC plate.
Along with the short nanotube segments 9, 6 and 15, in-
completely cyclised by-products are observed. These by-prod-
ucts are the result of deborylation/debromination (Figure S4j
in the Supporting Information), Suzuki over-couplings and
ligand scrambling (Figures S6q and S11s in the Supporting In-
formation). Although these side reactions are known,[50–53] their
extent here may be due to the very high number of consecu-
tive Suzuki coupling steps (8, 10 and 12 for 6, 9, and 15, re-
spectively). This makes the complete assembly of the upper
SWCNT short segments a slow process, which leaves enough
time for such “faults” to occur. This trend is reinforced by the
strained nature of 6, 9, and 15. As a consequence, the reaction
time was optimised: 10 h for compound 6, and 6 h for com-
pounds 9 and 15. These comparatively short reaction times
decrease the overall yield, but, most importantly, reduce the
amount of faulty products, which results in simpler
purification. We indeed observed that these faulty products
were difficult to remove.
Figure 13. Variable-temperature analysis of 15 in [D7]DMF (solvents marked
with asterisks).
1
At 363 K, the H NMR spectrum is characteristic of C2v sym-
metry (Figure 14A). This is in accordance with faster rotational
motion of the two stages of 15 (Figure 10C), resulting in this
averaged highly symmetric spectrum (Figure 14A, see also Fig-
ure S11k in the Supporting Information). Detailed NMR spec-
troscopy analysis is fully consistent with this interpretation. For
example, at 363 K, two sets of doublets corresponding to the
bridging methylenes of the calixarene stage of 15 are ob-
served at d=2.84/3.69 and 3.47/4.5 ppm (Figure 14B, red). All
other protons show the expected coupling patterns (Fig-
ure S11m in the Supporting Information, COSY). The HSQC ex-
Extensive DFT calculations were performed on compounds
6, 9 and 15 (Figure S18 in the Supporting Information).
Figures 15, 16 and 17 show the HOMO/LUMO systems of the
different SWCNT short segments.
Chem. Eur. J. 2016, 22, 3105 – 3114
www.chemeurj.org
3111
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
structures for 6, 9 and 12. An electronic transition in one of
these compounds will result in an excited state that will relax
very rapidly through charge transfer from/to the other chro-
mophore (Figure S18e in the Supporting Information). Consid-
ering the small electronic coupling between the calixarene and
m-(phenylene) stages, electron/hole recombination should be
disfavoured, resulting in long-lived charge-separated states.
Synthesis of a functionalised zigzag SWCNT short segment
To really envision the growth of SWCNTs by using our short
segments as starting points, the introduction of suitable
functional groups is necessary. We thus demonstrated this pos-
sibility by a proof-of-concept experiment with the methoxy-
functionalised calixarene scaffold 19 (Figure 18).
Figure 15. DFT-computed LUMOs and HOMOs of the calix[4]arene-derived
SWCNT short segment 6 (A) and calix[5]arene-derived SWCNT short segment
9 (B).
Figure 16. DFT-computed LUMOs of the calix[6]arene-derived SWCNT short
segment 15.
Figure 18. Synthesis of a p-methoxy-functionalised zigzag SWCNT short
segment on a calix[4]arene scaffold.
Key diiodo derivative 18 is easily available on a 20 g scale,
starting from inexpensive and commercially available 4-
hydroxy-3,5-diiodobenzaldehyde (Figures S12–S15 in the Sup-
porting Information). From octaiodo scaffold 19 (Figure S16 in
the Supporting Information), the corresponding SWCNT short
segment 20 was obtained under similar conditions to those
used for 9 and 15 (Figure S17 in the Supporting Information).
This proof-of-concept experiment shows that, in our case, the
steric effects associated with the presence of substituents next
to the reactive sites do not prevent the final cyclisation steps.
It could thus open up the way to synthesise lengthened
SWCNT segments, for example, by means of a demethylation/
triflation/Suzuki coupling sequence. Other, more easily re-
moved, protecting groups than methyl may also be envi-
sioned. This could allow the introduction of extra aromatic
units, oriented parallel (e.g., linear p-poly(phenylene)-type olig-
omers), for the construction of longer tubes. This should open
up, for the first time, simple strategies for the synthesis of arbi-
trarily long SWCNTs; a fundamental improvement compared
with the previously envisioned lengthening strategies for
SWCNT short segments.
Figure 17. DFT-computed HOMOs of the calix[6]arene-derived SWCNT short
segment 15.
Regarding calix[5]-derived segment 9, the calculations were
performed with tBu groups at the para position to reduce the
calculation time (Figure 15B).
In all three cases, the SWCNT short segments are fully conju-
gated. The lower symmetry of compound 15 compared with
that of 6 and 9 splits the HOMO and LUMO of the correspond-
ing SWCNT short segment into different quasi-degenerate
ones. However, the observed overlapping patterns are similar
to those observed for 6 and 9.
Moreover, there are no low-lying orbitals delocalised over
both the short SWCNT and calixarene scaffold. This means that
the m-phenylene and calixarene stages are electronically
decoupled. However, their respective energy levels allow for
electron/hole transfer upon excitation. Figure S18d in the
Supporting Information shows the calculated electronic
Chem. Eur. J. 2016, 22, 3105 – 3114
www.chemeurj.org
3112
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Agency for Research (ANR) for funding, grant number ANR-08-
JCJC-0085.
Conclusion
We described the first synthesis of short segments of zigzag-
type CNTs with tuned diameters. In the case of the calix[4]ar-
ene scaffold, this approach involved a very simple two-step
protocol, starting from cheap, commercially available reagents.
This strategy appeared to be far simpler than those previously
described.[17–24,29] Our synthetic strategy is extremely versatile
because it opens up the possibility of using a huge variety of
bis(boronic)derivatives in combination with different calixarene
scaffolds. This opens up the way to synthesise new families of
macrocyclic compounds, with well-defined diameters and con-
strained into unconventional tubular geometries. DFT calcula-
tions show that these compounds should exhibit interesting
photophysical properties. It should also be underlined that
compound 15 exhibits (to the best of our knowledge) the
largest rigid cavity known so far for a calixarene derivative,
which is interesting for supramolecular chemistry.
Keywords: calixarenes · carbon · nanotubes · macrocycles ·
total synthesis
[1] M. Monthioux, V. Kuznetsov, Carbon 2006, 44, 1621.
[2] S. Iijima, Nature 1991, 354, 56.
[3] S. Iijima, T. Ichihashi, Nature 1993, 363, 603.
[4] D. S. Bethune, C. H. Kiang, M. S. de Vries, G. Gorman, R. Savoy, J. Vaz-
quez, R. Beyers, Nature 1993, 363, 605.
[5] R. Martel, T. Schmidt, H. R. Shea, T. Hertel, P. Avouris, Appl. Phys. Lett.
1998, 73, 2447.
[6] S. J. Tans, A. R. M. Verschueren, C. Dekker, Nature 1998, 393, 49.
[7] P. Bondavalli, P. Legagneux, D. Pribat, Sensors and Actuators B 2009,
140, 304.
[8] R. Saito, M. Fujita, G. Dresselhaus, M. S Dresselhaus, Appl. Phys. Lett.
1992, 60, 2204.
[9] S. M. Bachilo, L. Balzano, J. E. Herrera, F. Pompeo, D. E. Resasco, R. B.
Weisman, J. Am. Chem. Soc. 2003, 125, 11186.
[10] X. Li, X. Tu, S. Zaric, K. Welsher, W. S. Seo, W. Zhao, H. Dai, J. Am. Chem.
Soc. 2007, 129, 15770.
[11] M. He, A. I. Chernov, P. V. Fedotov, E. D. Obraztsova, J. Sainio, E. Rikkinen,
H. Jiang, Z. Zhu, Y. Tian, E. I. Kauppinen, M. Niemelä, A. O. I. Krause, J.
Am. Chem. Soc. 2010, 132, 13994.
[12] L. Ding, A. Tselev, J. Wang, D. Yuan, H. Chu, T. P. McNicholas, Y. Li, J. Liu,
Nano Lett. 2009, 9, 800.
[13] S. Ghosh, S. M. Bachilo, R. B. Weisman, Nat. Nanotechnol. 2010, 5, 443.
[14] X. Tu, S. Manohar, A. Jagota, M. Zheng, Nature 2009, 460, 250.
[15] T. Tanaka, H. Jin, Y. Miyata, S. Fujii, H. Suga, Y. Naitoh, T. Minari, R. Miya-
dera, K. Tsukagoshi, H. Katura Nano Lett. 2009, 9, 1497.
[16] M. C. Hersam, Nat. Nanotechnol. 2008, 3, 387.
[17] S. Hitosugi, W. Nakanishi, T. Yamasaki, H. Isobe, Nat. Commun. 2011, 2,
492.
[18] H. Omachi, Y. Segawa, K. Itami, Org. Lett. 2011, 13, 2480.
[19] S. Hitosugi, T. Yamasaki, H. Isobe, J. Am. Chem. Soc. 2012, 134, 12442.
[20] R. Jasti, J. Bhattacharjee, J. B. Neaton, C. R. Bertozzi, J. Am. Chem. Soc.
2008, 130, 17646.
Apart from its “scaffolding” effect, the calixarene stage itself
provides interesting possibilities for the future synthesis of
original SWCNT/calixarene composites, combining the proper-
ties of both families of compounds. For example, it seems
reasonable to consider that the presence of the calixarene
fragment will confer ion complexation capacities on our
SWCNT short segment/calixarene composites (especially for 15,
with its PEG chain). At the same time, the very large diameter
of the SWCNT short segment will make these compounds in-
teresting hosts for large-molecule complexation. These two ef-
fects pave the way to interesting ion-pair complexation effects
for large molecules (biology) that cannot be addressed with
current calixarene receptors, with comparatively smaller
cavities.
As a longer term prospect, our strategy for the lengthening
of SWCNT short segments should result in very interesting
SWCNT/calixarene composites, with only one calixarene unit at
one end. Providing that ions are then complexed into these
units, this will result in a completely new object, that is,
a SWCNT with only a single charge (if alkaline ions are used) at
one end. This will lead to new possibilities for the organisation
of these nanotubes on surfaces, for example, on electrically
polarised substrates or by using polyelectrolytes.
[21] R. Gleiter, B. Esser, S. C. Kornmayer, Acc. Chem. Res. 2008, 41, 1108.
[22] T. J. Sisto, M. R. Golder, E. S. Hirst, R. Jasti, J. Am. Chem. Soc. 2011, 133,
15800.
ˇ
[23] Y. Segawa, S. Miyamoto, H. Omachi, S. Matsuura, P. Senel, T. Sasamori, N.
Tokitoh, K. Itami, Angew. Chem. Int. Ed. 2011, 50, 3244; Angew. Chem.
2011, 123, 3302.
[24] S. Yamago, Y. Watanabe, T. Iwamoto, Angew. Chem. Int. Ed. 2010, 49,
757; Angew. Chem. 2010, 122, 769.
[25] T. Nishiuchi, X. Feng, V. Enkelmann, M. Wagner, K. Müllen, Chem. Eur. J.
2012, 18, 16621.
[26] H. Omachi, T. Nakayama, E. Takahashi, Y. Segawa, K. Itami, Nat. Chem.
2013, 5, 572.
[27] F. E. Golling, M. Quernheim, M. Wagner, T. Nishiuchi, K. Müllen, Angew.
Chem. Int. Ed. 2014, 53, 1525; Angew. Chem. 2014, 126, 1551.
[28] E. Kayahara, V. K. Patel, S. Yamago, J. Am. Chem. Soc. 2014, 136, 2284.
[29] B. Esser, F. Rominger, R. Gleiter, J. Am. Chem. Soc. 2008, 130, 6716.
[30] A. M. Butterfield, B. Gilomen, J. S. Siegel, Org. Process Res. Dev. 2012, 16,
664.
Finally, we obtained the first functionalised zigzag SWCNT
short segments. The functionalisation pattern and symmetry
allow for simple strategies to be envisioned for lengthening.
This may lead to the total synthesis of arbitrarily long seg-
ments by simple iterative approaches. This work is in progress
within our group and will be reported in due course.
[31] L. T. Scott, E. A. Jackson, Q. Zhang, B. D. Steinberg, M. Bancu, B. Li, J.
Am. Chem. Soc. 2012, 134, 107.
[32] U. H. F. Bunz, S. Menning, N. Martin, Angew. Chem. Int. Ed. 2012, 51,
7094; Angew. Chem. 2012, 124, 7202.
[33] J. Xia, M. R. Golder, M. E. Foster, B. M. Wong, R. Jasti, J. Am. Chem. Soc.
2012, 134, 19709.
Experimental Section
Experimental details are provided in the Supporting Information.
[34] T. J. Sisto, X. Tian, R. Jasti, J. Org. Chem. 2012, 77, 5857.
[35] A. Yagi, Y. Segawa, K. Itami, J. Am. Chem. Soc. 2012, 134, 2962.
[36] H. Ito, Y. Mitamura, Y. Segawa, K. Itami, Angew. Chem. Int. Ed. 2015, 54,
156; Angew. Chem. 2015, 127, 161.
Acknowledgements
[37] M. R. Golder, R. Jasti, Acc. Chem. Res. 2015, 48, 557.
[38] P. J. Evans, E. R. Darzi, R. Jasti, Nat. Chem. 2014, 6404.
[39] H. A. Staab, F. Binnig, Chem. Ber. 1967, 100, 293.
E.A. and B.B. gratefully acknowledge the French Ministry of Re-
search for PhD thesis grants. We thank the French National
Chem. Eur. J. 2016, 22, 3105 – 3114
www.chemeurj.org
3113
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[40] T. Yamamoto, Y. Hayashi, A. Yamamoto, Bull. Chem. Soc. Jpn. 1978, 51,
2091.
[48] S. V. Shorunova, M. M. Krayushkin, F. M. Stoyanovich, M. Irie, Russ. J. Org.
Chem. 2006, 42, 1490.
[49] M. Baur, M. Frank, J. Schatz, F. Schildbach, Tetrahedron 2001, 57, 6985.
[50] S. Kappaun, M. Zelzer, K. Bartl, R. Saf, F. Stelzer, S. Slugovc, J.Polym. Sci.:
Part A: Polym. Chem. 2006, 44, 2130.
[51] S. T. Handy, H. Bregman, J. Lewis, X. Zhang, Y. Zhang, Tetrahedron Lett.
2003, 44, 427.
[52] O. Navarro, H. Kaur, P. Mahjoor, S. P. Nolan, J. Org. Chem. 2004, 69, 3173.
[53] B. Hohl, L. Bertschi, X. Zhang, A. D. Schlüter, J. Sakamoto, Macromole-
cules 2012, 45, 5418.
[41] K. N. Trueblood, C. B. Knobler, E. Maverick, R. C. Helgeson, S. B. Brown,
D. J. Cram, J. Am. Chem. Soc. 1981, 103, 5594.
[42] D. J. Cram, T. Kaneda, R. C. Helgeson, G. M. Lein, J. Am. Chem. Soc. 1979,
101, 6752.
[43] J. Y. Xue, K. Ikemoto, N. Takahashi, T. Izumi, H. Taka, H. Kita, S. Sato, H.
Isobe, J. Org. Chem. 2014, 79, 9735.
[44] A. Casnati, P. Jacopozzi, A. Pochini, F. Ugozzoli, R. Cacciapaglia, L. Man-
dolini, R. Ungaro, Tetrahedron 1995, 51, 591.
[45] K. Iwamoto, K. Araki, S. Shinkai, J. Org. Chem. 1991, 56, 4955.
[46] V. Huc, V. GuØrineau, Eur. J. Org. Chem. 2010, 11, 2199.
[47] V. Huc, E. Npetgat, V. GuØrineau, S. Bourcier, A. Dos Santos, R. Guillot,
J. P. Baltaze, C. Martini, Eur. J. Org. Chem. 2010, 11, 6186.
Received: September 14, 2015
Revised: November 6, 2015
Published online on January 27, 2016
Chem. Eur. J. 2016, 22, 3105 – 3114
www.chemeurj.org
3114
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim