The sergeants-and-soldiers effect: Chiral amplification in naphthalenediimide nanotubes

Article abstract of DOI:10.1039/c0ob00027b

Self-assembling naphthalenediimide (NDI)-based helical organic nanotubes display sergeants-and-soldiers behaviour, chiral monomers imposing a supramolecular structure upon achiral monomers. Several achiral NDI monomers were synthesised and their ability to incorporate into supramolecular nanotubes was studied. Nanotubes containing predominantly the most effective soldier, derived from 1-aminocyclopropane-carboxylic acid, are effective hosts for C ₆₀.

Full text of DOI:10.1039/c0ob00027b

Organic &
Biomolecular
Chemistry
www.rsc.org/obc
Volume 8 | Number 19 | 7 October 2010 | Pages 4185–4484
ISSN 1477-0520
FULL PAPER
Tom W. Anderson et al.
EMERGING AREA
Lalla Aicha Ba et al.
Tellurium: an element with great
biological potency and potential
The sergeants-and-soldiers
effect: chiral amplification in
naphthalenediimide nanotubes
1477-0520(2010)8:19;1-V

PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
The sergeants-and-soldiers effect: chiral amplification in naphthalenediimide
nanotubes†
Tom W. Anderson, Jeremy K. M. Sanders and G. Dan Pantos¸*
Received 21st April 2010, Accepted 9th June 2010
DOI: 10.1039/c0ob00027b
Self-assembling naphthalenediimide (NDI)-based helical organic nanotubes display
sergeants-and-soldiers behaviour, chiral monomers imposing a supramolecular structure upon achiral
monomers. Several achiral NDI monomers were synthesised and their ability to incorporate into
supramolecular nanotubes was studied. Nanotubes containing predominantly the most effective
soldier, derived from 1-aminocyclopropane-carboxylic acid, are effective hosts for C60.
Introduction
We report here that achiral building blocks can be incorporated
into chiral helical nanotubes, exhibiting ‘sergeants-and-soldiers’
organisation1 as judged by both spectroscopic and functional
binding behaviour. Many chiral amino acid derivatives of naph-
thalenediimides (NDIs) may be readily obtained in good yield by
the microwave reaction of the relevant amino acid with naphtha-
lene dianhydride (NDA), as shown in Scheme 1.2 These molecules
were found to spontaneously self-assemble into helical nanotubes
when dissolved in an aprotic solvent such as chloroform.3
Scheme 1 Generalised microwave synthesis of NDIs from NDA and
amino acids (or esters). For more details, such as the range of R-groups
employed, see ref. 2
This assembly is due to two sets of hydrogen bonds.
Firstly, a classical O–H ◊ ◊ ◊ O interaction between the carboxylic
acid-functionalised ends of each NDI monomer produces a
supramolecular polymer. Secondly, a non-classical C–H ◊ ◊ ◊ O
interaction between an aromatic hydrogen of the naphthyl core
Fig. 1 Brief overview of the two forms of hydrogen bonding in the NDI
and a diimide carbonyl oxygen, specifically between NDI i and
nanotubes: (a) the classical O–H ◊ ◊ ◊ O carboxylic acid dimerisation and
i + 3 (see Fig. 1), is responsible for aligning the NDI cores to be
coplanar.
(b) the non-classical C–H ◊ ◊ ◊ O cross-bonding between NDI i and i + 3
(immediately below it in the helix).3
The presence of these two types of hydrogen bonding along
with the amino acid chiral centre impose a helical twist to the
structure, meaning that this is not simply a linear supramolecular
polymer, but a supramolecular nanotube. This structure resulting
from hydrogen-bond-directed assembly is preferred in this case
over the more usual p–p stacking geometry.4,5
University Chemical Laboratory, University of Cambridge, Lensfield Road,
Cambridge, UK CB2 1EW. E-mail: gdp26@cam.ac.uk; Fax: (+44) 1223-
336-017
† Electronic supplementary information (ESI) available: X-Ray data,
majority rule study, CD data, 13C NMR data. CCDC reference numbers
743751. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/c0ob00027b
4274 | Org. Biomol. Chem., 2010, 8, 4274–4280
This journal is
The Royal Society of Chemistry 2010
©

The direction of the helical twist (M or P) is determined by
the chirality at the a-carbon of the constituent amino acid but
is independent of the chemical nature of its side chains: an L-
amino acid-derived NDI produces a P-helix, while a D-amino
acid-derived NDI produces an M-helix. Due to their chirality
the nanotubes may be studied by circular dichroism spectroscopy
(Fig. 2).
Fig. 3 (a) C60 13C NMR upfield shift of 1.4 ppm observed upon the
addition of NDI 1; (b) computer modelling of C60 inside the nanotube
cavity, adapted from ref. 3
the destruction of the nanotube and templates the formation
of a hexameric NDI ‘capsule’-like receptor for individual C70
molecules.10
Some chiral self-assembling systems are able to incorporate
enantiomers of opposite chirality because the chiral centre is rela-
tively remote from the moiety that is involved in self-assembly.11,12
By contrast, NDI monomers are highly discriminating when it
comes to the formation of chiral nanotubes: one enantiomer
cannot be incorporated into a helix made of the other enantiomer
because the directionality created at the chiral centre is critical to
self-assembly. Therefore, when mixed, the two enantiomers self-
sort. This is shown by the observation that a 1 : 1 mixture of L-1 and
D-1 continues to show good receptor behaviour for C60, indicating
that nanotubes are still present (even though they are not visible
by CD due to the equal proportions of L- and D-enantiomers).
The high degree of self-sorting is also shown by a ‘majority rules’
study12–14 (see ESI†). Sergeants-and-soldiers behaviour therefore
requires that L- and D-enantiomers do not reject an achiral NDI
as they would one another.
Fig. 2 Example CD traces showing the inverted signal of D-NDI
nanotubes with respect to L and the collapse seen when nanotubes are
destroyed by the addition of methanol. The NDI used was 1 (see Fig. 4) at
a concentration of 1 ¥ 10-3 mol dm-3.
Computational studies have shown that the characteristic CD
trace produced by the nanotube helix is due to a dipole–dipole
interaction between adjacent NDIs (i and i + 1) which is
strongly dependent on the angle between these naphthyl cores.6
As expected, the CD trace of a P-helix is inverted with respect to
that of an M-helix. The nanotubes may be destroyed by heating
or by adding a polar solvent such as methanol. These actions
result in the collapse of the CD signal (see Fig. 2), showing how
CD spectroscopy may be used as an indicator of the presence
or absence of a nanotube helix. Esterifying the NDI carboxylic
acid groups prevents hydrogen bonding and inhibits nanotube
formation. Solutions of the esters give similar CD traces to those
of the acids plus methanol.
The sergeants-and-soldiers effect
The ‘sergeants-and-soldiers’ effect was first proposed in the field
of polymer chemistry in the 1960s and was named by M. M. Green
and co-workers in 1989.1,15 Since that time, E. W. Meijer and
others have broadened the scope of the concept to take in many
facets of supramolecular as well as polymer chemistry.16–19 In a
system with ‘sergeants-and-soldiers’ behaviour a chiral derivative
(the ‘sergeant’) imposes its chirality on a structure formed mainly
out of achiral derivatives, the ‘soldiers’. Many systems of this type
have been reported.17,20a–e
In the case of NDI nanotubes, investigating the possibility of a
sergeants-and-soldiers effect requires the synthesis of carboxylate-
functionalised NDIs from achiral amino acids. Glycine is perhaps
the most obvious choice, but the resulting NDI had previously
been found to be highly insoluble in organic solvents. Several
different achiral amino acids outside the ‘natural 20’ were therefore
employed to synthesise a range of achiral NDIs.
The nanotubes have been found to be an effective receptor for a
˚
number of guests. C60, with a diameter of 10.3 A, is readily taken up
˚
into the 12.4 A interior cavity of the nanotube. The nanotubes are
capable of solvating C60 in solvents where the fullerene alone has
poor solubility, such as chloroform. An observed induced Cotton
effect on the C60 absorbances indicates that C60 senses the chirality
of the nanotubes. Most relevantly for the work presented here, the
ring current of the naphthyl cores of the NDIs acts to shield the
C60 carbons, producing a significant upfield shift in their 13C NMR
signal (Fig. 3).7
Recently, the nanotubes have also been shown to be effective
receptors for many aromatic molecules, with a strong dependence
on size due to the limits of the interior cavity. More surprisingly,
they can also serve as hosts for a wide variety of ions, including
those which form donor–acceptor complexes either with or within
the nanotube itself.8,9
In this work, we have used chiral NDIs derived from S-trityl
cysteine (1 and control ester 2), or N-Boc lysine (3 and control
ester 4) as the sergeant, while achiral derivatives 5–8 were used as
soldiers (Fig. 4). For reasons that will become clear below, most
As C70 has the same diameter as C60, it was expected that
C70 would be complexed in a similar fashion by the nanotube.
The reality proved to be quite different: addition of C70 causes
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4274–4280 | 4275
©

structures are different from the nanotubes observed for the chiral
derivatives.
Results and discussion
In order to test the ‘sergeants-and-soldiers’ effect in the NDI
nanotube system, solutions of L-1 and 5 (2.1 ¥ 10-4 mol dm-3
each) were prepared in TCE. Chiral amplification and propagation
experiments were performed by addition of L-1 to a solution of 5
and addition of 5 to solutions of L-1, respectively (see Experimental
Section for details). Both experiments were then repeated with a
solution of the non-hydrogen bonding ester L-3 substituted for 5
to act as a control. The overall concentration of NDI moieties
therefore remains constant throughout the additions. An example
of the CD traces across the full scanning range taken for one
experiment is shown in Fig. 6(a). The CDmax points were then
collated and plotted against percentage of L-1 present to produce
Fig. 6(b).
It is immediately clear that the presence of 5 amplifies the
CD signal relative to that which the dilute solution of L-1 can
Fig. 4 Chiral NDIs 1 and 2 derived from protected cysteine and lysine,
respectively; esterified inactive variants 3 and 4 for control experiments;
achiral NDIs 5, 6, 7 and 8 derived from the corresponding achiral amino
acids (Trt = trityl, Boc = tert-butyloxycarbonyl-).
efforts have been focused on NDI 5. The synthesis of this derivative
followed the published procedures,1 producing 5 in good yield.
Crystals of derivative 5 suitable for X-ray diffraction study were
obtained from DMSO. In this hydrogen bond acceptor solvent,
the NDIs are arranged in a p-stacking structure, with the NDI
cores n and n + 1 tilted at ca. 30◦ with respect to each other, while
n and n + 2 are parallel to each other (dihedral angle 0.6◦, Fig. 5).
Fig. 5 Crystal structure of 5 showing tilt of NDI n + 1 with respect to
parallel NDIs n and n + 2. Note that this is only part of the unit cell; see
Experimental Section for details.
Compound 5 is much less soluble than chiral NDIs 1 and 2 in
chloroform, but it can be solubilised by heating or by adding protic
solvents to break up the p-stacked/hydrogen-bonded structures
that may be formed. 1,1,2,2-Tetrachloroethane (TCE) was found
to be a superior solvent to chloroform for the achiral NDI. Binding
experiments with C60 (vide infra) in both solvents indicate that these
Fig. 6 (a) Overlaid CD traces across the full scanning range for a
typical propagation experiment, 5 being added to L-1. (b) Plotted CDmax
(383.5 nm) for all experiments. Concentration of all NDIs = 2.1 ¥ 10-4 mol
dm-3.
4276 | Org. Biomol. Chem., 2010, 8, 4274–4280
This journal is
The Royal Society of Chemistry 2010
©

produce alone. In the centre of the plot, the ratio of L-1 to 5
is 1 : 1. Maximum amplification (2.5-fold) is observed around
30% L-1. Since nanotubes are clearly formed in the presence
of as little as 10% chiral NDI,21 the effective heterogeneous
nanotubes’ association constant must be at least 105 mol-1 dm3;
given that the equilibrium constant for carboxylic acid hydrogen
bonding in TCE is only of the order of 103, this confirms that
there is cooperative binding between the NDI monomers, as in
homogeneous nanotubes.2 The curvature of both the ‘sergeants-
and-soldiers’ and control plots in Fig. 6(b), which is also observed
even in simple dilution experiments of solutions of nanotubes is
likely to be due to the cumulative and cooperative nature of the
hydrogen bonding.
A further experiment with the same protocol was used to test
that the ‘sergeants-and-soldiers’ behaviour was not unique to L-1.
Separate solutions of NDIs D-2 and L-2 were each diluted with 5
and with the appropriate ester (D-4 or L-4) as control. When plotted
on one graph (Fig. 7) this also illustrates that—as expected—
achiral 5 amplifies both M- and P-helices equally, with no intrinsic
preference.
Fig. 8 13C NMR (125 MHz, 298 K, TCE-d2) traces of the C60 peak; ratio
of L-1 to 5. Molar ratio NDI : C60 is 4 : 1, [NDI] = 7.21 ¥ 10-2 mol dm-3.
of 5 rises. However, even nanotubes composed of more 5 than L-
1 (fourth trace, Fig. 8) still lead to an upfield shift far greater
than could be achieved by the quantity of L-1 present alone.
The peak represents a weighted average of C60 inside and outside
the nanotube, an exchange which is fast on the NMR timescale.
The loss of intensity results from a fast exchange between rapidly
relaxing bound C60 and the slow relaxing free C60. These results
therefore indicate that heterogeneous nanotubes become slightly
poorer receptors for C60 as the proportion of achiral NDI 5 rises.
This may suggest that the superstructure of the heterogeneous
nanotubes is subtly different to that of the homogenous L-1
nanotube. This difference could be due to the achiral 5 lacking
the usual amino acid a-proton. The steric repulsion of two alkyl
groups on 5 is balanced by the cyclopropyl ring holding them
tightly together, meaning the N–C–C(O) angle in 5 is within the
range of those of 1 and 2, both of which readily form nanotubes
with similar binding properties (Fig. 9). However the added
rigidity imposed by the cyclopropyl moiety forces 5 to take up
conformations that are slightly different to the minimum energetic
conformation for the chiral amino acid-derived NDIs (which
present the a-proton in the most sterically crowded position).
Fig. 7 Plotted CDmax points for mixtures of L-2/5 (ꢀ) or L-2/esterified
L-4 (ꢁ); and D-2/5 mixtures (᭹) or D-2/esterified D-4 (᭺). Concentration
of all NDIs = 2.1 ¥ 10-4 mol dm-3.
C60 Studies
As described above, homogenous nanotubes, such as those
composed entirely of L-1, are known to take up C60. Experiments
were performed using 13C NMR in deuterated TCE to study if this
functional behaviour is the same for heterogeneous nanotubes
made up of a mixture of chiral L-1 and achiral 5. NDI solutions
of several different ratios of L-1 to 5 were allowed to stand over
Fig. 9 Ring strain in NDI 5 counteracts the repulsion of the alkyl carbons
to bring the N–C–C(O) angle back within the range of an NDI derived
from a chiral amino acid with only one R group (1,2). The lesser ring
strain in 7 produces a reduced version of this effect, while it is not present
at all in unstrained 6 or 8. The degree of difference in the N–C–C(O) angle
is exaggerated for clarity in the diagrams. Where crystal structures of the
NDIs were not available, estimates have been made by averaging the values
found in crystal structures of related compounds from CSD.22
C60 for 3 h, after which the resulting solution was analysed by 13
C
NMR. Control experiments were also performed with solutions
containing 100% L-1, 100% 5, and C60 alone in deuterated TCE
to provide ‘goalposts’ delineating the extreme possibilities. The
chemical shifts of the C60 peak were then compared as shown in
Fig. 8.
The magnitude of the upfield shift of the C60 chemical shift,
caused by shielding by the ring currents of the naphthalene system
and/or adjacent fullerenes, decreases slightly as the proportion
It is notable that 5 (and 7) alone does not induce a shift of the
peak due to dissolved C60 in pure solvent, strikingly suggesting
that 5 cannot form a racemic mixture of helical nanotubes.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4274–4280 | 4277
©

Other achiral NDIs
Conclusions
The importance of 5’s three-membered ring strain is demonstrated
by the fact that achiral NDIs 6 and 8 derived from achiral
amino acids without this strain (Fig. 4) were observed to have no
significant sergeants-and-soldiers activity, apparently interacting
no more with chiral NDI L-1 than the control ester does (see ESI†).
Based on this interpretation, we might naively expect an achiral
NDI possessing an N–C–C(O) angle somewhere between that
of the strained 5 and unstrained 6 and 8 to be a soldier but a
poorer one than 5. Such an achiral NDI is 7, which is derived
from 1-aminocyclobutane-1-carboxylic acid. A four-membered
ring naturally has internal angles of 90◦ rather than the 60◦
of a three-membered ring, which commensurately means less
counteraction of the repulsion compression of the N–C–C(O)
angle but some degree remains, unlike the unstrained internal
angles of 6 and 8 (Fig. 9).
NDI 7 was found to indeed act as a ‘soldier’, but an inferior one
to 5, as predicted (Fig. 10). A similar behaviour was observed in
the C60 complexation studies in which 7 was used as counterpart
for L-1 (see ESI†). This supports the idea that bond angle must be
key to the molecule’s suitability as a soldier: it is the only significant
molecular distinction between NDIs 5, 6 and 7, yet accounts for
their very variant properties. It may be that future molecular
modelling studies will help elucidate the exact consequences
of changing the bond angle upon the structure of the hybrid
nanotube, and whether a hybrid nanotube can form at all.
A new ‘sergeants-and-soldiers’ supramolecular system is pre-
sented. Chiral amplification behaviour is observed with different
chiral NDI ‘sergeants’ and equally with both enantiomers. An
achiral NDI does not form functional nanotubes alone, yet builds
upon existing ones. A negative result with other achiral NDIs
highlights the high specificity with which NDI monomers may
be incorporated into the nanotube. Uptake of C60 suggests that
the structure of these resulting heterogeneous nanotubes is subtly
different and depends on the “soldier“ NDI used.
Experimental
General methods
All solvents were of reagent grade quality (DMF, CH3CN) or
HPLC grade (CHCl3) and purchased commercially. All starting
materials were purchased from Aldrich and ChemImpex and used
without further purification. Melting points were determined with
a Gallenkamp apparatus and are uncorrected. NMR spectra were
recorded on Bruker DRX 400 MHz or 500 MHz instruments. The
NMR spectra were referenced to solvent and the spectroscopic
solvents were purchased from Euriso-Top (C. E. Saclay). All the
spectra were recorded at 298 K. 1H NMR data are reported
as follows: chemical shift in ppm on the d scale, integration,
multiplicity (s: singlet, d: doublet, t: triplet, q: quartet, dd: doublet
of doublets, bs: broad singlet, bt: broad triplet), coupling constants
(Hz) and assignment. All high-resolution (HR) electrospray
ionization mass spectra were recorded on a Waters LCT Premier
XE instrument. The sonication was performed using a Bransonic
1210E tabletop ultrasonic cleaner. The microwave experiments
were conducted using a CEM DiscoverTM Microwave Synthesizer.
The CD analyses were performed on an Applied Photophysics
Chirascan circular dichroism spectrometer. Compounds L-1, L-2,
D-2, L-3, L-4 and 8 were synthesized using previously reported
methods.1
Synthesis of achiral NDI (5)
1,4,5,8-Naphthalenetetracarboxylic dianhydride (200 mg, 0.746
mmol) and 1-aminocyclopropane-1-carboxylic acid (151 mg,
1.494 mmol) were suspended in 6 mL of DMF in a pressure-tight
8 mL microwave vial. To this suspension was added 0.2 mL of
dry Et3N. The suspension was sonicated until the mixture became
homogenous. The reaction mixture was heated for 5 min at 140
Fig. 10 Example datasets for NDIs 5 (black points) and 7 (grey points)
acting as soldiers to L-1 as sergeant, collected using the same solution
of L-1 to reduce experimental errors. Control data using L-3 as ester
as inactive soldier is also plotted (hollow points). Concentration of all
NDIs = 2.1 ¥ 10-4 mol dm-3. Note the black L-1 + 7 trace has a larger CD
signal at 90% chiral than at 100%.
5
◦C (direct flask temperature measurement) under microwave
irradiation using a dedicated microwave system. The solvent was
removed under reduced pressure, and the reaction mixture was
worked-up by being re-dissolved in EtOH (~5–10 mL) and poured
into constantly stirred acidified water (5 mL 37% aqueous HCl in
500 mL water). This was allowed to stir at room temperature for a
further 1 h, then filtered under suction, the product was collected
as a tan solid. The product was thoroughly dried under reduced
pressure.
In most cases the product was obtained without need for
purification, with a yield averaging 91%. In rare cases requiring
purification due to the presence of starting materials or mono-
substituted side products, the product was recrystallised from
hot chloroform–methanol (7 : 3 ratio). Unreacted NDA did not
It is notable in Fig. 6, 7, and 10 that solutions of chiral NDIs
1 and 2 containing small percentages of 5 tend to show stronger
CD signals than 100% chiral solutions of the same total NDI
concentration. The effect is small but reproducible, and must result
from a subtly different geometry modifying the size of the exciton
coupling between adjacent chromophores.6 A modified geometry
present in the hybrid nanotubes might also explain the inferior C60
receptor ability.
4278 | Org. Biomol. Chem., 2010, 8, 4274–4280
This journal is
The Royal Society of Chemistry 2010
©

dissolve and was filtered from the hot mixture; the mono-
substituted side product crystallised out first, allowing separation
from the desired di-substituted product, 5.
Synthesis of other achiral NDIs
The procedure followed was identical to that for 5, with the
exception that the respective achiral amino acid was used in
the appropriate quantity (for 6, 2-aminoisobutyric acid; for 7, 1-
aminocyclobutane-1-carboxylic acid). The synthesis of 7 produces
a similar yield as 5 (average 89%) and also rarely requires
purification; 6 and 8 have smaller yields (50–60%) and almost
always require purification by the method quoted above.
Characterisation data
5. (1,1¢(1,3,6,8-tetraoxobenzo[lmn][3,8]phenanthroline-2,7-
(1H,3H, 6H,8H)-diyl)dicyclopropanecarboxylic acid): mp >
300 ◦C; 1H NMR (400 MHz, DMSO-d6) d (ppm): 12.80 (bs,
2H), 8.70 (s, 4H), 1.81 (dd, J = 2.9, 4H), 1.47 (t, J = 2.9, 4H);
13C NMR (1H) (100.62 MHz, DMSO-d6) d (ppm): 172.5, 163.2,
131.1, 126.7, 35.2, 18.7; HRMS (ESI+) calcd for: C22H15N2O8 [M
+ H]+ (m/z): 435.0828, found: 435.0847; Elemental analysis for
5C22H14N2O8·1H2O, calcd: C 60.28%, H 3.40%, N 6.39%, found
C 60.13%, H 3.29%, N 6.30%.
Fig. 11 (a) Molecular structure of 5 showing the atom labelling scheme.
Displacement ellipsoids are scaled to the 50% probability level; b) Unit
cell packing diagram for single crystals of 5.
shown that such a solution is experimentally viable for over 5 h
at room temperature without reheating, significant precipitation
only occurring after more than 16 h). Solutions of the same
concentration of the required chiral NDIs were prepared in TCE.
Quartz cuvettes of path length 5 mm and 10 mm were used
(volume of 2 mL and 4 mL, respectively) for all experiments. An
example of a full data run from 100% chiral to 100% achiral was
recorded as follows, using a 5 mm, 2 mL cuvette with chiral NDI L-
1 and achiral NDI 5 (as seen in Fig. 6(b)). 1 mL of the L-1 solution
was added to the cuvette and the CD signal recorded from 400 to
270 nm three times, smoothed and averaged. 0.1 mL of a solution
of 5 of equal concentration was added, the cuvette shaken and
allowed to stand for 1 min, and the signal again recorded by the
same procedure. This was repeated until the cuvette was filled
and had therefore reached 50% achiral/50% chiral. The whole
experiment was then repeated three times and the results averaged
to minimise experimental error. The averaged CDmax points of each
recording were plotted against percentage chiral NDI present. This
provides the 100–50% chiral part of the data run covering the right
side half of Fig. 6(b). A second experiment was then performed
using the same procedure, but this time starting with achiral 5 and
then adding 0.1 mL volumes of chiral L-1 until 50% achiral/50%
chiral was reached, and this too was repeated three times, averaged
and the CDmax points plotted. This provides the 50–100% achiral
part of the data run covering the left side half of Fig. 6(b). Finally
the experiments were performed again (only a single time) using
H-bond-inactive ester L-3 instead of 5 to provide the control data.
The results shown in Fig. 7 were obtained by the same procedure,
except now using L-2 and D-2 as the chiral NDIs in question, and
L-4 and D-4 for the control esters. Raw CD data plots are shown
in the ESI.†
6. (2,2¢(1,3,6,8-tetraoxobenzo[lmn][3,8]phenanthroline-2,7-
(1H,3H, 6H,8H)-diyl)bis(2-methylpropanoic acid): mp > 300 ◦C;
1H NMR (400 MHz, DMSO-d6) d (ppm): 12.57 (bs, 2H), 8.63
(s, 4H), 1.83 (s, 6H); 13C NMR (1H) (100.62 MHz, DMSO-d6) d
(ppm): 174.3, 164.13, 127.63, 126.40, 63.50, 25.19; HRMS (ESI+)
calcd for: C22H19N2O8 [M + H]+ (m/z): 439.1141 found: 439.1136;
Elemental analysis for 3C22H18N2O8·2H2O, calcd: C 58.67%, H
4.33%, N 6.22%, found: C 58.76%, H 4.21%, N 6.26%.
7. (1,1¢(1,3,6,8-tetraoxobenzo[lmn][3,8]phenanthroline-2,7-
(1H,3H,6H,8H)-diyl)dicyclobutanecarboxylic acid): mp
>
300 ◦C; 1H NMR (400 MHz, DMSO-d6) d (ppm): 13.08 (bs,
2H), 8.65 (s, 4H), 2.86 (dd, J1 = 9.1, J2 = 10.0, 2H), 2.71 (dd,
J1 = 9.1, J2 = 10.0, 2H), 2.27 (dd, J1 = 9.6, J2 = 10.0, 1H),
1.82 (dd, J1 = 9.6, J2 = 10.0, 1H); 13C NMR (1H) (125 MHz,
TCE-d2) d (ppm): 161.60, 143.90, 129.41, 127.92, 126.80, 126.14,
67.48, 52.50, 45.72; HRMS (ESI+) calcd for: C24H19N2O8 [M
+ H]+ (m/z): 463.1141 found: 463.1136; Elemental analysis for
4C24H18N2O8·1H2O, calcd: C 61.66%, H 4.00%, N 6.04%, found
C 61.65%, H 4.05%, N 6.19%.
X-Ray crystallography of 5
Achiral NDI 5 was crystallised from dimethylsulfoxide solution at
room temperature (Fig. 11). Under such conditions 5 was found
to form a monoclinic unit cell with space group P21/c. Unit cell
˚
dimensions: a = 9.0612(2); b = 11.8489(2); c = 28.6964(5) A;
3
˚
volume = 3047.05(10) A . Further details may be found in the
ESI.†
CD experimental procedures
A
solution of 5 (or another achiral NDI) in 1,1,2,2-
13C NMR experimental procedures (C60 uptake study)
tetrachloroethane (TCE) at the required concentration was heated
to 100 ◦C for 1 h with stirring, to aid solubilisation. The solution
of 5 was allowed to cool to room temperature. (Tests have
Mixtures of NDIs L-1 and 5 were made up in the following
proportions: 1 : 0, 1 : 3, 1 : 1, 3 : 1 and 0 : 1. The total molar content
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4274–4280 | 4279
©

of each sample was 11.12 mmol. Each sample was dissolved in
2.5 mL of 1,1,2,2-tetrachloroethane-d2 and allowed to stand over
excess C60 for 4 h. A control vial containing only solvent and C60
was also made up. 0.7 mL of each mixture was then transferred to
an NMR tube and 13C NMR was taken using a Bruker Avance 500
TCI spectrometer. Data for each nanotube fullerene complex was
11 A. R. A. Palmans and E. W. Meijer, Angew. Chem., Int. Ed., 2007, 46,
8948.
12 (a) M. M. J. Smulders, I. A. W. Filot, J. M. A. Leenders, P. van der
Schoot, A. R. A. Palmans, A. P. H. J. Schenning and E. W. Meijer,
J. Am. Chem. Soc., 2010, 132, 611; (b) M. M. J. Smulders, P. J. M. Stals,
T. Mes, T. F. Paffen, A. P. H. J. Schenning, A. R. A. Palmans and E. W.
Meijer, J. Am. Chem. Soc., 2010, 132, 620.
13 J. Van Gestel, Macromolecules, 2004, 37, 3894.
14 J. Van Gestel, A. R. A. Palmans, B. Titulaer, J. A. J. M. Vekemans and
E. W. Meijer, J. Am. Chem. Soc., 2005, 127, 5490.
1
acquired over 800 scans with a delay of 6 s. The peaks in the H
NMR spectra were integrated and the ratio of L-1 to 5 recalculated
15 C. Carlini, F. Ciardelli and P. Pino, Makromol. Chem., 1968, 119, 244.
16 (a) A. R. A. Palmans, J. A. J. M. Vekemans, E. E. Havinga and E. W.
Meijer, Angew. Chem., Int. Ed. Engl., 1997, 36, 2648; (b) L. Brunsveld,
B. G. G. Lohmeijer, J. A. J. M. Vekemans and E. W. Meijer, Chem.
Commun., 2000, 2305; (c) R. B. Prince, J. S. Moore, L. Brunsveld and
E. W. Meijer, Chem.–Eur. J., 2001, 7, 4150.
17 S. J. George, Z. Tomovic, M. M. J. Smulders, T. F. A. de Greef, P. E. L. G.
Leclere, E. W. Meijer and A. P. H. J. Schenning, Angew. Chem., Int. Ed.,
2007, 46, 8206.
from this to take any precipitation into account.
Acknowledgements
We thank Pembroke College, Cambridge, and EPSRC for financial
support and Dr J. E. Davies for the X-ray crystal structure of 5.
18 (a) M. M. J. Smulders, A. P. H. J. Schenning and E. W. Meijer, J. Am.
Chem. Soc., 2008, 130, 606; (b) T. F. A. de Greef, M. M. J. Smulders, M.
Wolffs, A. P. H. J. Schenning, R. P. Sijbesma and E. W. Meijer, Chem.
Rev., 2009, 109, 5687; (c) M. M. J. Smulders, M. M. L. Nieuwenhuizen,
T. F. A. de Greef, P. Van Der Schoot, A. P. H. J. Schenning and E. W.
Meijer, Chem.–Eur. J., 2010, 16, 362.
Notes and references
1 M. M. Green, M. P. Reidy, R. J. Johnson, G. Darling, D. J. Oleary and
G. Willson, J. Am. Chem. Soc., 1989, 111, 6452.
2 P. Pengo, G. D. Pantos¸, S. Otto and J. K. M. Sanders, J. Org. Chem.,
2006, 71, 7063.
3 G. D. Pantos¸, P. Pengo and J. K. M. Sanders, Angew. Chem., Int. Ed.,
2007, 46, 194.
19 A. J. Wilson, M. Masuda, R. P. Sijbesma and E. W. Meijer, Angew.
Chem., 2005, 117, 2315.
20 (a) J. C. Nelson, J. G. Saven, J. S. Moore and P. G. Wolynes, Science,
1997, 277, 1793; (b) R. B. Prince, T. Okada and J. S. Moore, Angew.
Chem., Int. Ed., 1999, 38, 233; (c) M. L. Bushey, A. Hwang, P. W.
Stephens and C. Nuckolls, Angew. Chem., Int. Ed., 2002, 41, 2828;
(d) S. R. Nam, H. Y. Lee and J.-I. Hong, Chem.–Eur. J., 2008, 14,
6040; (e) D. Monti, M. De Rossi, A. Sorrenti, G. Laguzzi, E. Gatto,
M. Stefanelli, M. Venanzi, L. Luvidi, G. Mancini and R. Paolesse,
Chem.–Eur. J., 2010, 16, 860.
21 In a mixture containing 10% L-1 and 90% 5 about 20% of the
NDI material is assembled into nanotubes. In the control experiment
containing 10% L-1 and 90% L-3 only about 10% of the NDI material
is assembled in nanotubes.
4 F. Wurthner and S. Yao, Angew. Chem., Int. Ed., 2000, 39, 1978.
5 F. Wurthner, C. Thalacker, S. Diele and C. Tschierske, Chem.–Eur. J.,
2001, 7, 2245.
6 B. M. Bulheller, G. D. Pantos¸, J. K. M. Sanders and J. D. Hirst, Phys.
Chem. Chem. Phys., 2009, 11, 6060.
7 G. D. Pantos¸, J. L. Wietor and J. K. M. Sanders, Angew. Chem., Int.
Ed., 2007, 46, 2238.
8 E. Tamamini, G. D. Pantos¸ and J. K. M. Sanders, Chem.–Eur. J., 2010,
16, 81.
9 E. Tamanini, N. Ponnuswamy, G. D. Pantos¸ and J. K. M. Sanders,
Faraday Discuss., 2010, 145, 205.
10 J. L. Wietor, G. D. Pantos¸ and J. K. M. Sanders, Angew. Chem., Int.
Ed., 2008, 47, 2689.
22 Cambridge Structural Database, v.5.30, 2009.
4280 | Org. Biomol. Chem., 2010, 8, 4274–4280
This journal is
The Royal Society of Chemistry 2010
©