Supramolecular chemistry of monochiral naphthalenediimides

Article abstract of DOI:10.1039/c1ob06147j

Three new N-desymmetrised naphthalenediimides (NDIs) are described, each containing one chiral and one achiral centre. The ability of such 'monochiral' NDIs to self-assemble into hydrogen-bonded helical nanotubes, to act as a sergeant in a 'sergeants-and-soldiers' system and to form a hexameric receptor for C₇₀ was examined. Small differences at the achiral centre were found to have significant effects on the supramolecular properties of the NDI. All three new NDIs form nanotubes that bind C₆₀, but with different efficiencies, and one is a better sergeant than any of the dichiral NDIs investigated to date.

Full text of DOI:10.1039/c1ob06147j

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 7547
www.rsc.org/obc
PAPER
Supramolecular chemistry of monochiral naphthalenediimides†
Tom W. Anderson,a G. Dan Pantos¸*a,b and Jeremy K. M. Sanders*a
Received 12th July 2011, Accepted 15th August 2011
DOI: 10.1039/c1ob06147j
Three new N-desymmetrised naphthalenediimides (NDIs) are described, each containing one chiral and
one achiral centre. The ability of such ‘monochiral’ NDIs to self-assemble into hydrogen-bonded helical
nanotubes, to act as a sergeant in a ‘sergeants-and-soldiers’ system and to form a hexameric receptor
for C70 was examined. Small differences at the achiral centre were found to have significant effects on
the supramolecular properties of the NDI. All three new NDIs form nanotubes that bind C60, but with
different efficiencies, and one is a better sergeant than any of the dichiral NDIs investigated to date.
soldiers by dichiral NDI sergeants in the so-called ‘sergeants-and-
Introduction
soldiers effect’,10–19 leading to chiral amplification that is detectable
using circular dichroism (CD) spectroscopy.9
Naphthalenediimides (NDIs), which are readily obtained from
naphthalene dianhydride (NDA) and two amino acids via a
microwave synthesis, self-assemble in chloroform or 1,1,2,2-
tetrachloroethane (TCE) to form hydrogen-bonded helical
nanotubes.1,2 When two equivalents of a chiral amino acid are
used, the resulting NDI possesses two identical chiral a-carbon
centres, such as NDI 1 and its ester analogue 2. The chirality
of the amino acid determines the helical twist of the resulting
nanotube: L-1 gives P-helices, while D-1 gives M-helices. A mixture
of L-1 and D-1 self-sorts into two opposing helices. Majority-rules
behaviour3–9 is not observed in these systems because, unlike in
other systems, the chiral centres are intimately involved in creating
the helical structure. A pair of achiral amino acids may also be
incorporated, producing an achiral NDI, such as 3 or 4 (Fig. 1). We
previously reported that such achiral NDIs may be organised as
The NDI nanotube is a receptor for numerous guests, most
relevantly C60.20,21 This uptake can be observed using 13C NMR,
with the C60 carbon peak exhibiting a characteristic upfield shift
due to shielding by the ring currents of the NDIs’ aromatic
cores and of neighbouring C60 molecules. The ‘sergeants-and-
soldiers’ experiments revealed that hybrid nanotubes, made up
of a mixture of chiral NDI L-1 and achiral 3 or 4, take up C60,
producing a slightly smaller upfield shift in the 13C NMR. This
was interpreted as the new achiral component introducing a subtle
change in geometry or kinetic stability to the nanotube, which
slightly reduces its ability to act as a C60 receptor.9 In contrast,
when C70 is added to a nanotube made up entirely of L-1, the
nanotube is destroyed, the NDIs instead assembling into discrete
hexameric C70-receptor complexes with characteristic changes to
the NDI 1H NMR.22 It does not appear that this receptor is formed
by any of the achiral NDIs.23
The ‘sergeants-and-soldiers’ experiments demonstrated that
small changes in the groups attached to the a-carbon of the achiral
‘soldier’ led to disproportionate changes in the behaviour of NDI
analogues. NDI 3 was found to be the best ‘soldier’ achiral NDI
(i.e. was most incorporated into L-1 nanotubes), NDI 4 was an
inferior but functional soldier and an NDI with two unconnected
methyl groups at the a-carbon (derived from dimethylglycine, not
pictured) showed no soldier activity at all. The most obvious
differences between the NDIs were the bond angles and rigidity
at the a-carbon, leading to a tentative explanation based on these
factors.9
Fig. 1 The dichiral and achiral NDIs used in this work. Trt = trityl,
–CPh3.
Having explored the very different behaviours of dichiral and
achiral NDIs, we now report the synthesis and behaviour of
‘monochiral’ NDIs containing two different amino acids, one
chiral and one achiral. These new building blocks help to answer
questions regarding the formation of nanotubes and the ability
of a single chiral centre to impose its preferred structure in a
supramolecular assembly. The behaviour of monochiral NDIs
towards C60 and C70 is also reported.
aUniversity Chemical Laboratory, University of Cambridge, Lensfield Rd.,
Cambridge, UK, CB2 1EW. E-mail: jkms@cam.ac.uk
bDepartment of Chemistry, University of Bath, Bath, UK, BA27 AY.
E-mail: g.d.pantos@bath.ac.uk
† Electronic supplementary information (ESI) available: CD and NMR
data. See DOI: 10.1039/c1ob06147j
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7547–7553 | 7547
©

View Article Online
The dramatic changes in the sergeants-and-soldiers behaviour
observed due to small changes in the nature of the achiral a-carbon
on the achiral soldier suggested another aspect to this enquiry:
if several different achiral amino acids were used to make several
different monochiral NDIs (each sharing a common amino acid as
the chiral component), would this produce analogously significant
differences in their properties?
Results and discussion
The synthesis employed here follows a recently published
procedure24 using a lower microwave temperature for the first
synthetic step. This method allows a monosubstitution of NDA by
a chiral amino acid to form a naphthalene monoimide (NMI). The
resulting NMI can then be worked up and used in lieu of NDA
with one equivalent of the desired second amino acid, producing
an N-desymmetrised NDI (Scheme 1). S-trityl-cysteine was the
amino acid of choice in the first step, making NMI 5, which in
turn will lead to monochiral NDIs best suited for comparison
with the dichiral NDI 1.
Scheme 2 Achiral amino acids and the resulting monochiral NDIs
produced by their reaction with L-5. Yields given are for this final step
only.
Nanotube formation
Equally concentrated solutions of monochiral NDIs L-6, L-7 and
L-8 were prepared in TCE and their CD traces were compared
to that of a solution of dichiral NDI L-1 (Fig. 2). The three
monochiral NDIs were found to produce CD signals of different
intensities, which are indicative of different levels of nanotube
formation.25 The signal of NDI L-6 is almost identical in strength
to that of L-1, showing it forms nanotubes just as capably as
its dichiral counterpart. However, the characteristic peak of the
nanotube CD is slightly shifted to shorter wavelengths.
Scheme 1 The two-step synthesis of a monochiral NDI via NMI. R =
two identical groups on an achiral amino acid.
Three monochiral NDIs were synthesised, as shown in Scheme
2. Two were derived from the same cyclopropyl- and cyclobutyl-
derived amino acids as were used to synthesise the achiral NDIs 3
and 4, while the third was glycine. A glycine-based achiral NDI had
previously been synthesised and considered as a potential soldier
for ‘sergeants-and-soldiers’ experiments, but proved insoluble in
most organic solvents. These monochiral NDIs therefore represent
both the investigation of two already studied achiral centres in a
new molecular context and a way of looking at a centre that had
been impossible to work with in an achiral NDI.
Fig. 2 CD traces of TCE solutions of monochiral NDIs L-6, L-7 and L-8
compared to that of dichiral L-1. Concentration of all solutions = 2.1 ¥
10-4 mol dm-3.
This effect is also observed with ‘sergeants-and-soldiers’ mix-
tures of L-1 and achiral 3, suggesting that L-6 can be thought of as
an ‘idealised’ form of such a mixture, with the achiral 3 centres an
integral part of each monomer.9 The UV/vis absorbance spectra
of L-6 is also shifted relative to that of L-1, indicating the CD
shift is due to the asymmetry of the NDI chromophore, rather
7548 | Org. Biomol. Chem., 2011, 9, 7547–7553
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
than changes in the interactions between adjacent molecules in
the nanotube.25
solution they distinctively split further into four doublets, which
is indicative of different chemical environments being imposed,
presumably by hydrogen bonding (Fig. 4).26
The CD signals of NDIs L-7 and L-8 are progressively reduced
in strength relative to L-6, indicating that nanotube formation
is impaired. However, hydrogen-bonded assemblies are clearly
present, as these signals collapse on the addition of 5% methanol
(see Supporting Information†). While L-7 unambiguously forms
the nanotube, albeit not as well as L-6, the behaviour of L-8 was
less clear. The possibility existed that the hydrogen bonds formed
through its glycyl end were too flexible to allow the formation of
a nanotube and the small signal we observe is simply due to a
dimer. In order to test this possibility, the CD spectrum of L-8 was
compared to that of an equally concentrated solution of NDI L-9
(Fig. 1), a hybrid between L-1 and L-2 in which only one carboxylic
acid is ‘capped’ with a methyl ester, so that it can only form dimers
(Fig. 3).
Fig. 4 A comparison of the naphthyl proton signals observed in the
1H-NMR spectrum of NDI L-6 under conditions in which the nanotube
can and cannot form.
Fig. 3 A comparison of the CD traces of monochiral NDI L-8 with the
dichiral monoester L-9 as solutions in TCE. The concentration of both
solutions is 5.4 ¥ 10-3 mol dm-3.
This result is significant as the first direct evidence of CH ◊ ◊ ◊ O
hydrogen bonding in the nanotube. However, it is also important
because the fact that we can observe these different proton
environments suggests that the assembly and disassembly of the
L-6 nanotube is significantly kinetically slower than that of the L-1
nanotube.27
The observed difference conclusively demonstrated that while
the L-8 CD signal is small, it definitely forms nanotubes rather
than only dimers, and is therefore capable of hydrogen bonding
through its glycyl ends; this observation is significant for the later
experiments described below.
Uptake of C60
Our previously published work showed that hybrid nanotubes
made up of a 1 : 1 mixture of L-1 sergeant and either 3 or 4
soldier were effective receptors for C60, but slightly inferior to
nanotubes composed only of L-1, and this difference seemed to be
an intrinsic property of the hybrid nanotube.9 If this interpretation
was correct, we might expect nanotubes made up of L-6 and L-7
to be analogously poorer C60 receptors and perhaps comparable
to the 1 : 1 mixture of L-1 and the corresponding achiral NDI (as
they contain the same proportions of chiral and achiral centres).
13C NMR experiments carried out under the same conditions
as previously showed that this prediction was broadly correct,
although the monochiral nanotubes were actually slightly poorer
receptors than the 1 : 1 mixtures (Fig. 5). The C60 peak represents
a weighted average of the fast-exchanging C60 molecules inside the
nanotube (shielded) and outside it (not shielded). Therefore, the
more effective the nanotube is as a receptor, the more this peak is
shifted upfield.
Evidence for hydrogen bonding
One issue with the previously published NDI nanotube work
is that while everything points to these assemblies being held
together by hydrogen bonds, the evidence for this is indirect.
The hydrogen bonding of a nanotube made up of L-1 cannot be
observed using 1H NMR, because although the hydrogen-bonded
naphthyl protons should be in different chemical environments, the
dynamic assembly and exchange of the NDIs to form the nanotube
is sufficiently fast on the NMR chemical shift timescale that the
naphthyl protons average to a single peak. However, the same is
not true of monochiral NDI L-6, as can be seen by comparing the
naphthyl region of the 1H NMR spectra in 5% MeOD in CDCl3
(in which nanotubes cannot form) and in pure CDCl3 (in which
they can). In the methanolic solution, the asymmetry of the L-6
monomer means that the naphthyl protons are observed as two
doublets coupling to one another. However, in the pure CDCl3
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7547–7553 | 7549
©

View Article Online
to the a-proton, and fits with the crystal structure obtained of L-1
where all of the –STrt groups must point away from the nanotube
to avoid steric clash.2
By contrast, the a-proton of L-6 has J1 = 9.8 Hz, J2 = 5.1 Hz.
This is consistent with conformation A, where the a-proton is
gauche to one of the b-protons but anti to the other. Importantly,
this effect is only observed under conditions when the nanotube
can form: in the methanolic solutions in which it cannot assemble,
L-1 and L-6 have almost identical a-proton coupling constants
(see Supporting Information†).
This change in conformation can be rationalised based on the
observation that L-6 only has a bulky –STrt group at one end of
the NDI. In an L-1 nanotube there are possible conformations
that are not adopted because they would place the –STrt group in
a position where it can clash with another –STrt on an adjacent
NDI, but in L-6 the second –STrt is not present as the adjacent NDI
presents its achiral end to the first NDI. This idea can also help us
understand how individual molecules of NDI L-6 are positioned
in the nanotube (Fig. 7).
Fig. 5 13C NMR showing the C60 resonance and its shift when exposed to
nanotubes made up of different NDIs in a solution of deuterated toluene.
Vertical lines: (a) = C60 + a 1 : 1 mixture of L-1 and 4; (b) = C60 + a 1 : 1
mixture of L-1 and 3; (c) = C60 + L-1.9 Each solution has the same molar
quantity of NDI and C60.
We had already hypothesised that the reduced ability of the
hybrid nanotubes to act as C60 receptors was due to geometric
differences in the way that the NDIs are arranged in the nanotube
and this explanation can also be applied to the monochiral
nanotubes. Clear confirmation that the geometry is different in the
L-6 nanotube to the L-1 nanotube can be found by examination of
the 1H proton signal corresponding to the a proton (Fig. 6), which
reveals major differences in the vicinal couplings to the b-protons
in the two species.
Fig. 7 Possible arrangements of NDI L-6 in its nanotube (which, for
clarity, is represented as a linear polymer). In the real helical nanotube,
NDI i + 3 is immediately below NDI i as there are three NDIs per turn of
the helix.
Case Z, a random arrangement, can be ruled out as it would
not explain this change in conformation. Either of case X or Y is
possible: it depends on whether the dominant steric clash between
–STrt groups is between adjacent NDIs along the hydrogen-
bonded chain (e.g. NDIs i and i + 1) or between NDIs immediately
above or below each other in the helix (e.g. NDIs i and i + 3). The
first would mean case X is correct as it places the chiral end of i
next to the achiral end of i + 1, while the second would favour case
Y as it places the chiral end of i immediately above the achiral end
of i + 3 in the helix. Currently, we cannot assess which of these
arrangements is the correct one.
Fig. 6 1H NMR of the a-proton of NDIs L-1 and L-6 in deuterated TCE,
together with Newman projections of the possible conformations of the
a-carbon (back) and b-carbon (front). Note that the NMR spectra are not
on the same vertical scale, as L-1 has two a-protons per NDI whereas L-6
only has one.
Formation of the C70 receptor
A second question regarding the interaction with fullerenes is
whether the new monochiral NDIs could also form the hexameric
receptor for C70 observed with dichiral NDIs.22 Initially these
experiments were performed in deuterated TCE due to the fact that
L-8 has too low a solubility in chloroform. In the process of doing
this work an unexpected solvent dependence for the formation of
the C70 receptor was discovered: even NDI L-1 does not form it in
TCE or in toluene, chloroform being the only solvent in which this
self-assembled receptor has yet been found to form. This was an
intriguing discovery and the reasons behind it remain under study.
In both L-1 and L-6, the a-proton couples to the two b-protons
on the bridging b-carbon to the bulky –STrt group. These protons
are diastereotopic, so the a-proton is split into a doublet of
doublets. In L-1, the two coupling constants are similar: J1
=
4.9 Hz, J2 = 4.1 H, which are consistent with a conformation close
to that labelled B in Fig. 6, in which both b-protons are gauche
7550 | Org. Biomol. Chem., 2011, 9, 7547–7553
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
In chloroform, all three monochiral NDIs were found to form
the C70 receptor in the presence of an excess of C70. However, unlike
L-1, they did not form the receptor exclusively, the equilibrium
position between receptor and free monomer/nanotube being
different for each monochiral NDI. This can be observed in the 1H
NMR of the naphthyl protons for each NDI, although integration
of the peaks associated with the two species is approximate as they
overlap. The most useful guide is the fact that, of the four doublets
produced by the C70 receptor, one is much more downfield than
the other three and separated from the peaks associated with the
free NDI or nanotube, recalling from the earlier discussion that
monochiral NDI nanotubes already produce multiple naphthyl
peaks as opposed to a singlet in nanotubes made up of symmetrical
NDIs like L-1. A second indicator is the a-proton shifting to a more
downfield position. The change on the addition of an excess of C70
to L-6 is shown in Fig. 8.
Fig. 9 The possible arrangements of NDI L-6 in the C70 receptor and a
schematic representation of the hydrogen bonding patterns at the pole28
and equator of the hexameric capsule. The single a-proton signal means
either A or B are possible, but C can be eliminated.
the soldier solution with one of NDI ester L-2, which cannot be
incorporated into the nanotubes and provides a baseline by which
the incorporation of soldiers may be judged. All four datasets are
then plotted on the same graph, CD intensity against % chiral
(i.e. % sergeant NDI) in the mixed solution.
Fig. 8 1H NMR signals for the naphthyl and a-protons of NDIs L-6 in
deuterated chloroform, before (black) and after (red) the addition of C70.
NDI L-8’s lower solubility in chloroform made the study
of its NMR more problematic (see Supporting Information†),
but it appears that less than 15% of the NDI forms the C70
receptor under these conditions. By contrast, integration shows
that approximately 70% of NDI L-6 and 30% of NDI L-7 forms the
receptor (see Supporting Information†). Perhaps significantly, this
is the same trend across the three molecules as in their nanotubes’
ability to act as a receptor for C60.
An insight into the possible arrangements of monochiral NDIs
around C70 was obtained from the observation that only one new
a-proton signal is seen emerging as the C70 receptor is formed.
This means that all of the chiral ends of the NDIs are in the same
environment; with L-1, two a-proton signals are observed as those
at the poles are in a different chemical environment to those at
the equator. We cannot yet assign whether the monochirals’ single
a-proton signal is associated with the equatorial or axial position,
but we can say that the arrangement of NDIs in the receptor is
ordered rather than random (Fig. 9).
For these experiments, the sergeant solution consisted of a
solution of one of the monochiral NDIs. The results were
surprising. While we might perhaps have expected monochiral
NDIs to be inferior sergeants, the manner of their interaction with
achiral 3 and 4 was unprecedented. NDIs L-6 and L-7 (L-6: Fig. 10;
for L-7, see Supporting Information†) do act as sergeants, but only
when soldiers 3 and 4 are present in higher concentrations than the
sergeants. By halving the concentration of the starting solution of
sergeant, a more familiar curve is obtained, but a comparison to L-
1 under the same conditions shows that the latter is still a superior
sergeant (Fig. 10). The same effect is seen with 4 as the soldier (see
Supporting Information†) save that, as previously reported, 4 is
an inferior soldier to 3. L-7 also acts in this manner as the sergeant.
At present, we do not yet have a convincing theoretical
explanation for this phenomenon. Fig. 2 demonstrates that NDIs
L-1 and L-6 appear to form the nanotube equally readily, so any
model to explain the difference in sergeant behaviour must take
this into account.
Monochiral NDIs as ‘sergeants’
NDI L-8 has very different behaviour again. Far from a
weaker and concentration-dependent sergeant activity, it is the
strongest sergeant discovered for this system thus far, surpassing
L-1. Because the CD signal for L-8 is small compared to other
monochiral NDIs, the protocol was slightly modified by averaging
additional data sets and using a dilution method for the control
data (See Supporting Information†) to minimise error, but the
results (Fig. 11) are still comparable to other data sets based on
equally concentrated solutions of sergeant and soldier.
‘Sergeants-and-soldiers’ studies were carried out with monochiral
NDIs L-6, L-7 and L-8 as potential sergeants, using the same
protocol as previously reported.9 This involves beginning with
a cuvette 50% filled with soldier NDI solution and recording the
CD after aliquots of 10% by volume of sergeant NDI solution
(of the same concentration) are added until the cuvette is filled to
a 1 : 1 mixture, giving a series of data points from 0–50% chiral
solution. The protocol is then repeated with the two solutions
swapped, giving data points from 50–100% chiral. A control data
set is recorded by repeating these two experiment but replacing
This very large chiral amplification can be rationalised if we
recall the idea that NDI L-8 forms nanotubes less readily than
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7547–7553 | 7551
©

View Article Online
Fig. 11 The sergeants-and-soldiers data set for NDI L-8 as the sergeant
and 3 as the soldier. For the details of the control data set, see Supporting
Information†.
NDI both helps us rationalise existing differences in the activity
of nanotubes as C60 receptors and provide new ideas for future
studies in this area. They have also provided us with more infor-
mation about the nature of the C70 receptor. Monochiral NDIs
allowed us to demonstrate how small changes in the molecular
structure lead to markedly different behaviour in the macroscopic
world.
Experimental
General methods
All solvents were of reagent grade quality (DMF, CH3CN) or
HPLC grade (CHCl3) and purchased from standard suppliers. 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) and Aldrich. 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 circu-
lar dichroism spectrometer. Compounds L-1, D-1, L-2, D-2,3,4,
L-5, L-9 and L-9 were synthesized using previously reported
methods.1,2,24
Fig. 10 The greater displacement above the control traces with inactive
ester L-2, the more effective the ‘sergeants-and-soldiers’ behaviour. Using
different mixtures of equally concentrated solutions of sergeant and soldier,
L-6 only acts as a sergeant when 3 is present in high concentration (low
% chirality), which is made clearer when half-concentrated ‘sergeant’
solutions are used.
the other monochiral NDIs and this is presumably due to less
favourable hydrogen bonding through its glycyl ends. NDI L-8
also forms the least proportion of the C70 receptor, probably for the
same reason. We can hypothesise that NDI 3 can act to bridge two
molecules of L-8 through their glycyl ends and allow the formation
of longer oligomers than L-8 alone can form. This explains why for
~60–90% L-8, the hybrid actually produces a stronger CD signal
than L-8 alone.
Conclusions
In this paper, we present a new set of NDI derivatives that helped
us to explore the key role of chirality in the supramolecular
nanotubes. All three new NDIs form nanotubes that bind C60,
but with different efficiencies, and one is a better sergeant than
any of the dichiral NDIs investigated to date. Monochiral NDIs
are not only interesting in their own right, but provide a tool to
better understand the sergeants-and-soldiers interaction between
dichiral and achiral NDIs. A geometric shift in the structure of the
Synthesis of monochiral NDIs
NMI L-5 (200 mg, 0.326 mmol) was dissolved in 5 mL DMF
together with the corresponding achiral amino acid (0.326 mmol).
7552 | Org. Biomol. Chem., 2011, 9, 7547–7553
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Et3N (100 mL) was added. The suspension was sonicated until
Acknowledgements
the mixture became homogenous. The reaction mixture was
5
◦C (direct flask temperature
We thank Pembroke College, Cambridge, and the University of
Bath (GDP) and EPSRC for financial support.
heated for 5 min at 140
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
16 h, then filtered under suction. The product was collected as
a yellow solid. The product was thoroughly dried under reduced
pressure.
Notes and references
1 P. Pengo, G. D. Pantos¸, S. Otto and J. K. M. Sanders, J. Org. Chem.,
2006, 71, 7063.
2 G. D. Pantos¸, P. Pengo and J. K. M. Sanders, Angew. Chem., Int. Ed.,
2007, 46, 194.
3 M. M. Green, B. A. Garetz, B. Munoz, H. P. Chang, S. Hoke and R.
G. Cooks, J. Am. Chem. Soc., 1995, 117, 4181.
4 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.
5 J. van Gestel, Macromolecules, 2004, 37, 3894.
6 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.
Characterisation data
L-6 ((R)-1-(7-(1-carboxy-2-(tritylthio)ethyl)-1,3,6,8-tetraoxo-7,8-
dihydrobenzo[lmn][3,8]phenanthrolin - 2 ( 1H , 3H , 6H) - yl)cyclo-
propanecarboxylic acid): Yiel◦d was 81% and further purification
7 A. R. A. Palmans and E. W. Meijer, Angew. Chem., Int. Ed., 2007, 46,
8948.
8 E. Yashima, K. Maeda, H. Iida, Y. Furusho and K. Nagai, Chem. Rev.,
2009, 109, 6102.
1
was not required. mp >300 C; H NMR (400 MHz, DMSO-
d6) d (ppm): 12.16 (bs, 1H) 12.93 (bs, 1H), 8.72 (s, 4H), 7.20 (m,
15H), 5.55 (dd, J1 = 10.52, J2 = 4.6, 1H), 3.12 (dd, J1 = 12.76,
J2 = 4.8, 1H), 2.92 (dd, J1 = 12.76, J2 = 10.52, 1H), 1.81 (dd, J1 =
9.8, J2 = 4.4, 2H), 1.45 (dd, J1 = 9.8, J2 = 4.4, 2H); 13C NMR
(1H) (125 MHz, TCE-d4) d (ppm): 172.15, 169.18, 162.82, 161.95,
143.12, 129.03, 128.88, 128.07, 127.94, 126.97, 126.48, 126.21,
125.16, 66.47, 52.18, 34.82, 30.25, 18.38; HRMS (ESI+) calcd for:
C40H28N2O8S [M+H]+ (m/z): 697.1639 found: 697.1582. Elemental
analysis for 4 C40H28N2O8S·1 H2O: calcd C, 68.51%; H, 4.10%; N,
3.99%; S, 4.57%; found: C, 68.53%; H, 4.11%; N, 4.00%; S, 4.58%.
L-7 (R)-1-(7-(1-carboxy-2-(tritylthio)ethyl)-1,3,6,8-tetraoxo-7,
8-dihydrobenzo[lmn][3,8]phenanthrolin-2(1H,3H,6H)-yl)cyclo-
butanecarboxylic acid): Yield◦was 73% and further purification
9 T. W. Anderson, J. K. M. Sanders and G. D. Pantos¸, Org. Biomol.
Chem., 2010, 8, 4274.
10 M. M. Green, M. P. Reidy, R. J. Johnson, G. Darling, D. J. Oleary and
G. Willson, J. Am. Chem. Soc., 1989, 111, 6452.
11 M. M. Green, N. C. Peterson, T. Sato, A. Teramoto, R. Cook and S.
Lifson, Science, 1995, 268, 1860.
12 C. Carlini, F. Ciardelli and P. Pino, Makromol. Chem., 1968, 119,
244.
13 A. R. A. Palmans, J. A. J. M. Vekemans, E. E. Havinga and E. W.
Meijer, Angew. Chem., Int. Ed. Engl., 1997, 36, 2648.
14 L. Brunsveld, B. G. G. Lohmeijer, J. A. J. M. Vekemans and E. W.
Meijer, Chem. Commun., 2000, 2305.
15 R. B. Prince, L. Brunsveld, E. W. Meijer and J. S. Moore, Angew. Chem.,
Int. Ed., 2000, 39, 228.
16 R. B. Prince, J. S. Moore, L. Brunsveld and E. W. Meijer, Chem.–Eur.
J., 2001, 7, 4150.
17 M. M. J. Smulders, A. P. H. J. Schenning and E. W. Meijer, J. Am.
Chem. Soc., 2008, 130, 606.
1
was not required. mp >300 C; H NMR (400 MHz, CDCl3)
d (ppm): 8.60 (s, 2H), 8.56 (s, 2H), 7.23 (m, 15H), 5.52 (dd,
J1 = 4.6, J2 = 5.2, 1H), 3.18 (dd, J1 = 12.1, J2 = 5.2), 3.10 (dd,
J1 = 12.1, J2 = 4.6, 1H), 2.69 (dd, J1 = 12.4, J2 = 8.9, 2H), 2.60
(dd, J1 = 11.2, J2 = 12.4, 2H), 2.45 (dd, J1 = 9.6, J2 = 8.9), 1.91
(dd, J1 = 11.2, J2 = 9.6); 13C NMR (1H) (125 MHz, TCE-d4)
d (ppm): 161.75, 161.51, 146.59, 130.75, 129.43, 128.28, 127.91,
126.79, 67.52, 62.99, 52.28, 36.46, 30.00, 16.78; HRMS (ESI+)
calcd for: C41H30N2O8S [M+H]+ (m/z): 711.1796 found: 711.1783.
Elemental analysis for 6 C41H30N2O8S·1 H2O: calcd C, 68.99%; H,
4.28%; N, 3.92%; S, 4.49%, found C, 68.97%; H, 4.29%; N, 3.91%;
S, 4.50%.
18 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.
19 L. J. Prins, P. Timmerman and D. N. Reinhoudt, J. Am. Chem. Soc.,
2001, 123, 10153.
20 G. D. Pantos¸, J. L. Wietor and J. K. M. Sanders, Angew. Chem., Int.
Ed., 2007, 46, 2238.
21 E. Tamanini, G. D. Pantos¸ and J. K. M. Sanders, Chem.–Eur. J., 2010,
16, 81.
22 J. L. Wietor, G. D. Pantos¸ and J. K. M. Sanders, Angew. Chem., Int.
Ed., 2008, 47, 2689.
23 T. W. Anderson, PhD thesis, University of Cambridge,
2011.
24 K. Tambara, N. Ponnuswamy, G. Hennrich and G. D. Pantos¸, J. Org.
L-8 ((R)-2-(7-(carboxymethyl)-1,3,6,8-tetraoxo-7,8-dihydro-
benzo[lmn][3,8]phenanthrolin-2(1H ,3H ,6H)-yl)-3-(tritylthio)-
propanoic acid): Yield was 67% and required further purification:
the product was recrystallised in hot chloroform/methanol (7 : 3
ratio) with the less soluble NDA starting material being removed
Chem., 2011, 76, 3338.
25 B. M. Bulheller, G. D. Pantos¸, J. K. M. Sanders and J. D. Hirst, Phys.
Chem. Chem. Phys., 2009, 11, 6060.
26 p–p interactions, which may lead to chiral aggregates that could
have a similar set of naphthyl proton resonances, are unlikely to
play a role in the self-assembly of the NDIs in chloroform, as
the association constant between two molecules in this solvent
is <1 M-1 (M. S. Cubberley and B. L. Iverson, J. Am. Chem. Soc.,
2001, 123, 7560).
27 The nanotubes made up of monochiral L-6 must cope with a constant
‘mismatch penalty’ from the fact that their components have both chiral
and achiral centre-based carboxylic acid ends, which will slow their
hydrogen-bonded assembly. The slower disassembly of L-6 nanotubes
may be due to the less sterically demanding packing of the smaller
sidechains in L-6 compared to L-1.
◦
1
via hot filtration. mp >300 C; H NMR (400 MHz, TCE-d4) d
(ppm): 8.76 (d, J = 7.6, 2H), 8.70 (d, J = 7.6, 2H), 7.25 (m, 15H),
5.00 (dd, J1 = 10.2, J2 = 5.1), 3.28 (dd, J1 = 6.8, J2 = 5.1, 1H),
3.15 (dd, J1 = 10.2, J2 = 6.8, 1H), 2.9 (d, J = 28.1, 2H); 13C NMR
(1H) (125 MHz, TCE-d4) d (ppm): 161.94, 161.48, 143.88, 129.33,
128.28, 127.88, 127.19, 126.77, 126.21, 67.47, 56.71, 52.40, 29.62;
HRMS (ESI+) calcd for: C38H26N2O8S [M+H]+ (m/z): 671.1488
found: 671.1479. Elemental analysis for 4 C38H26N2O8S·1 H2O:
calcd C, 67.60%; H, 3.96%; N, 4.15%; S, 4.75%; found C, 67.59%;
H, 3.97%; N, 4.15%; S, 4.74%.
28 A similar hydrogen pattern has been previously observed in alanine
phthalimide tapes: J. Li and Z.-P. Liang, Acta Crystallogr. Sect. E,
2006, 62, o4915.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7547–7553 | 7553
©