Design, synthesis and computational modelling of aromatic tweezer-molecules as models for chain-folding polymer blends

Article abstract of DOI:10.1016/j.tet.2008.05.077

Novel 'tweezer-type' complexes that exploit the interactions between π-electron-rich pyrenyl groups and π-electron deficient diimide units have been designed and synthesised. The component molecules leading to complex formation were accessed readily from

Full text of DOI:10.1016/j.tet.2008.05.077

Tetrahedron 64 (2008) 8346–8354
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Design, synthesis and computational modelling of aromatic tweezer-molecules
as models for chain-folding polymer blends
*
*
Barnaby W. Greenland, Stefano Burattini, Wayne Hayes , Howard M. Colquhoun
Department of Chemistry, University of Reading, Whiteknights, Reading RG6 6AD, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Novel ‘tweezer-type’ complexes that exploit the interactions between
p-electron-rich pyrenyl groups
Received 29 February 2008
Received in revised form 9 May 2008
Accepted 16 May 2008
and -electron deficient diimide units have been designed and synthesised. The component molecules
p
leading to complex formation were accessed readily from commercially available starting materials
through short and efficient syntheses. Analysis of the resulting complexes, using the visible charge-
transfer band, revealed association constants that increased sequentially from 130 to 11,000 Mꢀ1 as
Available online 21 May 2008
increasing numbers of
modelling was used to analyse the structures of these complexes, revealing low-energy chain-folded
conformations for both components, which readily allow close, multiple -stacking and hydrogen
bonding to be achieved. In this paper, we give details of our initial studies of these complexes and outline
how their behaviour could provide a basis for designing self-healing polymer blends for use in adaptive
coating systems.
p–p-stacking interactions were introduced into the systems. Computational
This paper is dedicated to Professor
Sir J. Fraser Stoddart on the occasion of his
receiving the 2007 Tetrahedron Award for
Creativity in Organic Chemistry
p–p
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
contains a dispersion of microcapsules filled with a monomeric
‘healing agent’. Should a crack propagate through the polymer, it
Materials chemistry has seen an explosion of interest in poly-
mers with tuneable properties in recent years. One of the many
driving forces for this research has been the desire to produce
polymers with healable or even self-healing characteristics.1 Poly-
mers of this type are typified by their ability to re-organise their
molecular architecture in response to a controllable external
stimulus such as heat, light or an electric field.2 Thus, if subjected to
stress or even fracture, the material is able to recover the physical
properties of the original system. Significant progress towards the
goal of self-healing polymers has been made by several groups.
Notably, Wudl et al. have synthesised epoxy resins containing diene
(furan) and dieneophile (maleimide) residues.3 These components
are able to react via a thermally reversible Diels–Alder reaction to
heal any structural damage within the resin. Thus, should a crack
form in the material, the system can be subjected to a thermal
annealing process during which time new covalent bonds form
between monomers between either side of the crack and the ma-
terial can regain its mechanical properties.
ruptures the microcapsules, releasing the monomer into the crack.
The healing agent is then polymerised by contact with a catalyst,
which is also impregnated within the bulk material, thus sealing
the crack with new polymer.
In contrast to these covalent-bond-forming approaches to heal
a fracture in a polymeric material, it has recently been shown that
non-covalent bonds may be used to effect healing processes within
supramolecular polymer materials.5 Supramolecular materials
generally comprise low to medium molecular weight species that
form extended polymeric architectures via non-covalent inter-
actions.6 Such self-assembled materials have been found to exhibit
physical properties similar to those of high molecular weight
polymers, including high viscosity7 and a high storage modulus.8 It
has been postulated5 that, in polymer systems of this type, a frac-
ture will propagate via dissociation of the weaker supramolecular
bonds between the components, rather than by scission of the
stronger covalent bonds within these units. In this model, as the
components of the material remain chemically unaltered during
the fracturing process, application of a suitable stimulus may allow
the supramolecular array to re-assemble and so ultimately restore
the original properties of the material.
An alternative approach to self-healing polymers has been
investigated by White et al.4 In these systems, the bulk polymer
Studies by Sijbesma and Meijer have shown that high binding
constants (>107 Mꢀ1) are required to generate stable and pro-
cessable supramolecular polymers.9 However, in addition to the
requirement for high binding efficiencies between the recognition
* Corresponding authors. Tel.: þ44 118 378 6491; fax: þ44 118 378 6331.
E-mail addresses: w.c.hayes@rdg.ac.uk (W. Hayes), h.m.colquhoun@rdg.ac.uk
(H.M. Colquhoun).
0040-4020/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.05.077

B.W. Greenland et al. / Tetrahedron 64 (2008) 8346–8354
8347
units of the monomeric components, the purity of the monomers
2. Results and discussion
used also constrains the ability to produce materials with proper-
ties consistent with high molecular weight polymers.7
2.1. Computational modelling
As part of a programme working towards the development of
healable and self-healing polymers, we are investigating whether
a blend of two interacting oligomeric components, in which one
of the species possesses multiple but equivalent binding sites, will
enable the generation of ‘infinite’ supramolecular arrays. A su-
pramolecular polymer of this type would possess a cross-linked
architecture and should therefore be tolerant of minor quantities
of impurities such as mono-functional components. In order to
realise self-assembling, supramolecular polymer systems of this
Key to success of the proposed two-component, complementary
polymer blend is the design of the chain-folding, diimide-oligomer
host system. It is vital that the separation of the p-electron deficient
imide groups allows them to attain the correct spatial configuration
in order to form a strong tweezer-like complex with a pyrenyl unit.
An additional design consideration is that the systems studied
should be accessible from readily available starting materials
through short and efficient syntheses routes if materials of genuine
utility are to be obtained. This approach ensures that the com-
plexation data, structural understanding and synthetic insights
gained through the present model studies can be transferred with
confidence to analogous but more complex higher molecular
weight materials.
Computational modelling studies (charge-compensated mole-
cular mechanics, using a custom-modified version of the Dreiding II
force field, Cerius3.5, Accelrys Inc., San Diego) revealed that sepa-
rating two naphthalene tetracarboxylic diimide units with a short
but highly flexible spacer-group derived from commercially avail-
able 2,20-(ethylenedioxy)bis(ethylamine) (1) would allow the
imides to adopt the correct spatial orientation to generate the
‘chain-folded-tweezer’ motif required for strong face-to-face
binding to pyrene (Fig. 1). For ease of computation, the model for
the diimide tweezer 2 contained simple methyl substituents at the
imide termini. Encouragingly, the interplanar separation between
the atoms of the diimide units and pyrene 3 within the complex
was found to be close to van der Waals contact distances, at ca.
3.5 Ådconsistent with values observed in the solid state structures
of related tweezers12 and catenanes11 and also with those found in
the co-crystals of pyrene and diimide derivatives.14a
type, we are exploiting the well-known ‘charge-transfer’ (
stacking) interaction between -electron-deficient naphthalene
tetracarboxylic diimide or pyromellitic diimide moieties and
-electron-rich species such as pyrenyl or alkoxynaphthalenyl
p–p-
p
p
residues.10 Complexes between such species have previously been
employed to synthesise a diverse range of systems ranging from
small-molecule catenanes11 and clips or ‘tweezer’-type mole-
cules,12 to extended oligomeric and polymeric complexes.13
Notable contributions in this field have been reported by Iverson
et al. who produced a series self-complementary, chain-folding
‘aedemers’,14 and the interactions occurring in double stranded
‘zipper’ type supramolecular polymeric complexes have also
been studied.15 More recent investigations by Colquhoun et al.
have explored the interactions between small, pyrene-based
tweezer-molecules and high molecular weight, chain-folding
polyimides.16
Utilising the chain-folding concept to provide multiple binding
sites embedded in an oligo-imide backbone, it should be possible to
‘cross-link’ these low molecular weight entities (non-covalently)
with telechelic pyrenyl-terminated oligomers via p–p-stacking and
hydrogen bonding interactions, to produce potentially healable
supramolecular materials. As part of our programme to generate
oligomeric, multi-receptor systems featuring chain-folding aro-
matic diimide units and pyrene-functionalised chains, we set out to
gain a detailed understanding of the interactions between these
two recognition motifs. Specifically, we sought to answer the fol-
lowing questions: (i) what is the simplest chain-folding motif that
can be synthesised and analysed; and (ii) is it possible to modulate
the strength of the interactions between pyrenyl moieties and the
chain-folding diimide to generate complexes with ‘designed-in’
binding constants? The present report details the rational design
and synthesis of a series of well-defined models for the proposed
chain-folding host polyimides and a study of their interactions with
a variety of pyrene-based guest molecules.
Whilst this simulation was most encouraging, it was apparent
that if 1:1 complexation between tweezer diimide 2 and pyrene 3
occurred as predicted, each of the
in tweezer 2 would only be able to maintain one face-to-face
interaction with the -electron-rich pyrene guest. However, to in-
p-electron-poor diimide units
p
crease the number of associative interactions between the species,
a simple approach would be to employ the amide-based, pyrenyl
tweezer-molecule 4, which we have studied extensively over recent
years.12,16 Modelling the ‘tweezer–tweezer’ complex, which could
potentially be formed between compounds 2 and 4 strongly
supported this approach, and suggested that both the pyrenyl
groups of 4 would be able to form strong, associative, face-to-face
interactions with the new bis-diimide tweezer 2. Moreover,
Figure 1. Structure of diimide tweezer 2 (a) and minimised model of the tweezer-type complex between the bis-diimide 2 and pyrene (b). Interplanar spacings between the
-systems average w3.5 Å.
p

8348
B.W. Greenland et al. / Tetrahedron 64 (2008) 8346–8354
additional binding forces should result from the pair of convergent
hydrogen bonds (NH/O¼2.01, 2.39 Å) predicted to occur between
the two amide –NH– groups of 4 and one of the carbonyl groups of
2 (Fig. 2).
Extending this concept still further, the oligomeric diimides, for
which compound 2 is designed to be a model, would contain sev-
eral repetitions of the bis-diimide chain-folding motif along the
backbone. This structural repeat would offer the potential for
a third diimide unit to complete a second chain-fold around the bis-
pyrenyl tweezer and thus bind both faces of each pyrenyl moiety of
4, so further increasing the overall strength of binding between
2.2. Synthesis and binding studies
Guided by the computational modelling studies of the complex
between the diimide tweezer 2 and pyrene (3) it was evident that,
to realise even the simplest such system in practice, it would be
necessary to synthesise a bis-diimide in which each of the diimide
units is unsymmetrically substituted. Simple addition of the re-
quired mono- and di-amine to naphthalene dianhydride 6 would
result in a statistical distribution of products, which in turn would
necessitate extensive purification to isolate the desired compound,
resulting in low yields of the target compound. To circumvent this
problem, the reaction conditions described by Buncel et al. were
employed.17 Thus, by carrying out the reaction stepwise, first in
water with careful control of pH (Scheme 1), unsymmetrically
substituted diimides may be obtained in good yields (typically ca.
75%). In the present work, application of this procedure enabled
the synthesis of diacid imide 7 in excellent yield (85%, Scheme 1).
The branched amine 8 was chosen as a commercial analogue of the
so-called ‘swallow tail’ hydrocarbons, which have been used to
promote solubility in related systems.18 Diacid 7 was used directly
p-electron-rich and p-electron-poor components. An energy-
minimised model of a complex between the trimeric diimide (5),
wherein each of the diimide units is separated by a 2,20-(ethyl-
enedioxy)bis(ethylamine) residue (see 1 in Fig. 3), and the bis-
pyrenyl tweezer-molecule 4 showed that a second chain-fold could
indeed be achieved quite readily. This system could thus meet
our aim of producing a model chain-folding polyimide from the
minimum number of low-cost, commercially available starting
materials.
Figure 2. Molecular formula of the bis-pyrenyl tweezer-molecule 4 (a) and an energy-minimised model of its complex with the bis-diimide 2 (b). Interplanar spacings shown are
the average distances of pyrenyl carbons from the mean planes of the diimide units. Hydrogen bonds (NH/O) are shown in magenta.
Figure 3. Structure of tris-diimide (a) and energy-minimised model of the complex formed between the trimeric diimide 5 and the bis-pyrenyl tweezer 4 (b). Molecule 5 is shown
in red, to distinguish it more clearly from tweezer 4. Hydrogen bonds (NH/O) are shown in magenta.

B.W. Greenland et al. / Tetrahedron 64 (2008) 8346–8354
8349
Scheme 1. Synthesis of bis-diimide tweezer 9. Reagents and conditions: (i) H2O, pH¼6.2, 110 ꢂC, 24 h, 85%; (ii) DMAc, 125 ꢂC, 18 h, 82%.
to synthesise the model tweezer-type compound 9 (82% yield) by
reaction with 0.5 mol equiv of 2,20-(ethylenedioxy)bis(ethylamine).
As noted above, the model complex between 2 and 3 (see Fig. 1)
averaged environments of the ‘naphthalene’ protons, when 9 is
complexed to 10, are quite different so that their associated reso-
nances no longer coincide.
wasoriginallyenergy-minimised usingpyreneasthe
p
-electron-rich
A Job plot, using results from UV/vis spectroscopy based on the
charge-transfer absorption band at 530 nm (Fig. 5a) confirmed for-
mation of a 1:1 complex between the pyrenyl derivative 10 and the
bis-diimide tweezer 9, supporting the stoichiometry predicted for
themodelled complex. Theassociationconstantforthe complexwas
found to be 130(ꢁ20) Mꢀ1 by the UV/vis dilution method19 (Fig. 5b).
Whilst these preliminary binding assays for complexation
between 9 and 10 were encouraging, the relatively low binding
constant would certainly not be sufficient to translate into
the formation of stable supramolecular polymers. To enhance the
binding constant, we next investigated complexation between the
bis-diimide tweezer-molecule (9) and the bis-pyrenyl tweezer-
compound 4. Mixing the two tweezer-type molecules 4 and 9
produced a deep red solution as a result of a strong absorption band
centred on 538 nm. A Job plot based on this absorption revealed
that the complex was formed in a 1:1 ratio of 4 to 9, in accordance
with the proposed mode of binding (Fig. 6a). Encouragingly, the
binding constant for 4 with 9 with was found to be 3500(ꢁ400) Mꢀ1
(Fig. 6b), more than an order of magnitude greater than that
observed between the mono-pyrene derivative 10 and the bis-
diimide 9 (Fig. 5b).
To assess the binding characteristics of an oligomeric, chain-
folding imide with the pyrenyl tweezer-molecule 4, we next syn-
thesised a trimeric diimide, analogous to the system modelled in
our initial computational studies. The target trimeric diimide 16
was thus synthesized as shown in Scheme 3. Mono-protection of
2,20-(ethylenedioxy)bis(ethylamine) afforded the known20 trityl
(Tr) protected amine 13, using a large excess of the diamine to
reduce the quantity of unwanted bis-protected amine formed.
Addition of amine 13 to diacid 7, followed by deprotection with
trifluoroacetic acid, furnished the desymmetrised naphthalene-
diimide 15 in good yield (76%, two steps). The synthesis was com-
pleted by condensation of 2 equiv of amine 15 with naphthalene
species. However, pyrene itself was judged to be an over-simplified
analogue for the pyrenyl-end-capped oligo-amides, which we pro-
pose to use in the construction of supramolecular network poly-
mers. Therefore, we synthesised the benzamidomethyl-substituted
pyrene 10 as a more realistic p-electron-rich component. This was
obtained straightforwardly via condensation of benzoyl chloride (11)
with 1-pyrenemethylamine hydrochloride (12) (Scheme 2).
With the model host and guest compounds 9 and 10 in hand, it
was possible to study the binding of the components under con-
trolled conditions. Simply mixing chloroform solutions of the two
components afforded a deep red solution resulting from a strong
charge-transfer band, centred on 530 nm, which is diagnostic of
a complementary, aromatic, p–p
-stacked complex.10 Moreover, ti-
tration of the pyrenyl derivative 10 against a fixed concentration of
tweezer-compound 9 (Fig. 4) in CDCl3 revealed a significant upfield
shift (up to 0.6 ppm at equimolar concentration 9 and 10) of the 1H
NMR resonance corresponding to the four ‘naphthalene’ protons of
tweezer-molecule 9. This shift is accompanied by a progressive
increase in the complexity of the resonances, which changes from
an apparent singlet in the starting material (9) to an apparent
AA0BB0 spin pattern (two distorted doublets) at equimolar con-
centrations of the components (Fig. 4). There are clearly two
chemically distinct protons in 9, one type adjacent to the ‘terminal’
imide units and the other adjacent to the ‘internal’ imides. An
AA0BB0 pattern would therefore be expected for 9. However, in the
uncomplexed molecule, the environments of these protons are
evidently indistinguishable and hence their associated resonances
coincide, with complete loss of coupling information. Complexa-
tion with compound 10 must then result in the two different types
of ‘naphthalene’ proton experiencing different degrees of ring-
current shielding by the pyrenyl residue. Association and dissoci-
ation are obviously fast on the NMR timescale, but even so the
Scheme 2. Synthesis of model compound 10. Reagents and conditions: (i) Et3N, CHCl3, 10 min, 80% yield.

8350
B.W. Greenland et al. / Tetrahedron 64 (2008) 8346–8354
Figure 4. Titration of the guest pyrenyl compound 10 against the bis-diimide tweezer 9, as monitored by 1H NMR spectroscopy (250 MHz, CDCl3; note residual CHCl3 resonance at
7.25 ppm).
tetracarboxylic dianhydride to furnish the target trimer 16 in 75%
yield.
Mixing solutions of the bis-pyrenyl tweezer 4 and trimer 16
produced a deep red solution as a result of a strong charge-transfer
band centred on 539 nm signifying complex formation. A Job
plot based on this absorption maximum revealed that the complex
forms in a 1:1 ratio of 4 to 16, with a binding constant of
11,000(ꢁ2000) Mꢀ1 (Fig. 7), an increase of a further factor of 3
compared to that for the tweezer–tweezer complex between 4 and
9 (Fig. 6).
Figure 5. Job plot (a) and UV/vis binding study (b) using the dilution method, for complex formation between 9 and 10 (error bars show standard error).

B.W. Greenland et al. / Tetrahedron 64 (2008) 8346–8354
8351
Figure 6. Job plot (a) and UV/vis binding study (b) using the dilution method, for complex formation between 4 and 9 (error bars show standard error).
Scheme 3. Synthesis of trimeric diimide 16. (i) Tr–Cl (0.1 equiv), Et3N, CH2Cl2, rt, 73%; (ii) DMAc, 125 ꢂC, 18 h, 80%; (iii) TFA/CH2Cl2 (1:10 v/v), 3 h, rt, 95%; (iv) DMAc, 125 ꢂC,
18 h, 75%.
3. Conclusions
pyrenyl guests in solution. Binding assays revealed that by in-
creasing both the number of -electron deficient naphthalene-
diimide units and -electron-rich, pyrene-based moieties, in
p
A
new chain-folding tweezer-molecule comprising two
p
p
-electron deficient naphthalene tetracarboxylic diimide units
conjunction with the use of appropriate flexible spacers between
the receptor sites, it was possible to generate stable assemblies
with binding constants up to 11,000 Mꢀ1. The results of this initial
study are very encouraging and tend to confirm our hypothesis
that the binding motifs described could form the basis of a new
class of supramolecular polymer blends based on non-covalent
networks.
separated by a short ethylene glycol spacer has been designed
and synthesised, and the binding characteristics of this type of
receptor with
evaluated. Computational modelling predicted that stable com-
plexes should be formed and indeed strong charge-transfer
absorption bands were observed on mixing the diimide host and
p-electron-rich pyrenyl guest molecules has been

8352
B.W. Greenland et al. / Tetrahedron 64 (2008) 8346–8354
Figure 7. Job plot (a) and UV/vis binding study (b) using the dilution method, for complex formation between 4 and 16 (error bars show standard error).
4. Experimental
4.1. General
reduce the number of variable parameters during minimisation.
Starting models were generated with pre-folded conformations
baseddwhere possibledon known X-ray structures12,16 and were
brought together by simple docking procedures prior to energy-
minimisation. While the final structures obtained in this work
clearly represent genuine energy-minima, a more detailed com-
putational study involving extended conformational searching
would be required to prove that these structures unambiguously
represent global minima.
Reagents were purchased from either Aldrich or Acros and were
used as received without further purification. Solvents were used as
supplied, with the exception of chloroform (CHCl3), which was
distilled under nitrogen from calcium hydride. DMAc (99.8% an-
hydrous) was purchased from Aldrich and used as received. NMR
spectra were recorded on Bruker AC250 operating at 250 MHz and
62.5 MHz for 1H and 13C nuclei, respectively, or Bruker AMX400
operating at 400 MHz and 100 MHz for 1H and 13C nuclei, re-
spectively. The residual protic signal from the deuterated solvent or
tetramethylsilane was used as internal reference. Infrared (IR)
spectroscopic analyses were performed on a Perkin–Elmer 17 20-X
Infrared Fourier Transform spectrometer on KBr disks using a thin
film for sample preparation. Matrix-assisted laser desorption–ion-
isation time-of-flight mass spectra (MALDI-TOF MS) were obtained
using a Scientific Analysis Instruments LT3 LaserTof (Manchester,
UK) spectrometer using dithranol as the matrix. A typical sample
4.3. Binding studies
Binding constants were determined according to the UV/vis di-
lution method described in detail by Stoddart et al.19 This method-
ology requires measurement of the charge-transfer absorption band
arising from complex formation between equimolar quantities of the
p-electron-rich and p-electron-deficient species at varying con-
centrations. The concentration range was chosen to ensure that the
maximum absorption measured during the experiment remained
below 1 absorbance unit, and thus within the range where the
Beer–Lambert law is applicable. The following concentration ranges
preparation is described as follows: 3
m
L of a solution of the analyte
in chloroform (1–10 mg mLꢀ1) was combined with 10–20
mL of the
were used: 9/10¼0.004500–0.000887 mol Lꢀ1
;
9/4¼0.001250–
freshly prepared matrix (40 mg mLꢀ1 in CHCl3) in a mini-vial, and
from the mixture was taken an aliquot, which was transferred
onto a sample plate and left to air dry prior to analysis. Mass spectra
(MS) were obtained from a Finnigan MAT95 (CI) or a Waters LCT
Premier (ESI). UV/vis absorption measurements were performed in
a 1 cm2 quartz cuvettes on a double-beam Perkin–Elmer Lambda
25 spectrophotometer interfaced with UV WinLab software, in the
0.000060 mol Lꢀ1
;
16/4¼0.000645–0.000278 mol Lꢀ1
. The error
bars shown in Figures 5b, 6b and 7b represent standard errors
of estimate.
4.4. Diacid imide 7
1,4,5,8 Naphthalene tetracarboxylic dianhydride (2.00 g,
7.46 mmol) was dissolved in a solution of KOH (1.96 g) in water
(350 mL) and the resulting brown solution was acidified to pH 6.3
with H3PO4 (1 M). 2-Ethyl hexylamine (962 mg, 7.46 mmol) was
added and the solution was re-acidified to pH 6.4 with H3PO4 (1 M),
then heated to 110 ꢂC for 16 h. After cooling to room temperature,
the cloudy solution was filtered and the filtrate was acidified with
AcOH (5 mL), which produced a cream precipitate. After stirring for
20 min, the solid was collected by filtration and dried under vac-
uum at 50 ꢂC to give 7 as a cream powder (2.52 g, 85%). Mp 173–
wavelength range 400–700 nm, employing
a scan speed of
120 nm minꢀ1, with a slit width of 1 nm. Spectra were recorded
using samples dissolved in analytical grade chloroform mixed 6:1
with hexafluoroisopropanol and blank-corrected for the possible
absorption of the solvent. Elemental analyses were provided by
Medac UK Ltd.
4.2. Computational modelling
Computational modelling studies using charge-compensated
molecular mechanics were carried out with a custom-modified
version of the Dreiding II force field in Cerius2 (version 3.5, Accelrys
Inc., San Diego) running on an SGI-O2 workstation. Customisation
of the force field involved constrained planarisation of the naph-
thalene tetracarboxylic diimide and pyrenyl residues in order to
174 ꢂC (decomp.); 1H NMR (CDCl3/CF3CO2H, 250 MHz)
d¼8.65 (2H,
AA0XX0, Ar–H), 8.18 (2H, AA0XX0, Ar–H), 4.04–3.86 (2H, m, NCH2CH),
1.86–1.81 (1H, m, CH2CH(CH2)2), 1.28–1.22 (8H, m, 4ꢃCH2), 0.89–
0.81 (6H, m, 2ꢃCH3); 13C NMR (CDCl3/CF3CO2H, 250 MHz)
¼168.5,
d
163.5, 137.0, 130.6, 129.6, 128.9, 125.8, 124.8, 43.8, 37.5, 30.4, 28.5,
23.9, 22.8, 14.3, 10.8. IR nmax(KBr)/cmꢀ1 3418, 2956, 2921, 2858,

B.W. Greenland et al. / Tetrahedron 64 (2008) 8346–8354
8353
1702, 1697, 1592, 1450, 1390, 1347, 1298, 1232. ESI-MS (negative ion
mode) m/z calcd for C22H22NO6: 396.1447, found: 396.1466. Anal.
Calcd for C22H23NO6$1/2H2O: C, 65.01; H, 5.95; N, 3.45%. Found: C,
64.73; H, 5.60; N, 3.40%.
1.24 (8H, m, 4ꢃCH2), 0.97–0.85 (m, 6H, 2ꢃCH3); 13C NMR (CDCl3,
62.5 MHz)
d¼163.0, 162.7, 146.0, 131.0, 128.7, 127.8, 126.69, 126.66,
126.59, 126.46, 126.21, 71.4, 70.6, 70.2, 70.0, 67.8, 44.5, 43.0, 39.5,
37.9, 30.6, 28.5, 24.0, 23.0, 14.0, 10.5; IR nmax(KBr)/cmꢀ1 3439, 2959,
2929, 2870, 1708, 1664, 1583, 1450, 1380, 1332, 1244, 1115. HRMS
(CI): C47H49N3O6 m/z calcd: 751.3621; found: 751.3628. Anal. Calcd
for C47H49N3O6: C, 75.08; H, 6.57; N, 5.59%. Found: C, 75.29; H, 6.69;
N, 5.51%.
4.5. Bis-diimide tweezer 9
A solution of diacid 7 (100 mg, 0.287 mmol) and diamine 1
(21.3 mg, 0.144 mmol) in DMAc (2.5 mL) and toluene (0.25 mL)
were stirred at 135 ꢂC for 20 h. After cooling to room temperature,
the dark solution was added dropwise to a stirred mixture of
chloroform/methanol (10 mL 1:1 v/v). After 10 min, the resulting
suspension was collected by filtration and dried at 70 ꢂC under
vacuum to give bis-diimide tweezer 9 (103 mg, 82%) as a peach
4.8. Amino diimide 15
Trityl protected diimide 14 (250 mg, 0.333 mmol) was dissolved
in a stirred mixture of dichloromethane (15 mL) and trifluoroacetic
acid (0.5 mL). After 3 h at room temperature, a saturated aqueous
solution of NaHCO3 (5 mL) was added dropwise and the organic
layer separated, washed again with aqueous NaHCO3 (2ꢃ5 mL) and
water (5 mL) before the organic layer was dried over MgSO4, fil-
tered and the solvent removed under reduced pressure to give an
orange solid, which was subjected to column chromatography
(gradient chloroform to 7:1:0.1 chloroform/methanol/triethyl-
amine) to give 15 (153 mg, 90%) as a tan solid. Mp 179–180 ꢂC
coloured solid. Mp 223–224 ꢂC; 1H NMR (CDCl3, 250 MHz)
d¼8.69
(8H, s, Ar–H), 4.36 (4H, t, J¼5.9 Hz, 2ꢃ(CO2)NCH2CH2), 4.12 (2H, dd,
J¼5.2 and 9.1 Hz, 2ꢃ(CO)2NCHHCH), 4.05 (2H, dd, J¼7.0, 13.0 Hz,
2ꢃ(CO)2NCHHCH), 3.77 (4H, t, J¼5.8 Hz, 2ꢃOCH2CH2), 3.66 (4H, s,
OCH2CH2), 1.95–1.84 (2H, m, CH2CH(CH2)2), 1.43–1.20 (16H, m,
8ꢃCCH2C), 0.94–85 (12H, m, 4ꢃCH3); 13C NMR (CDCl3, 67.5 MHz)
d
¼163.1, 162.8, 130.99, 130.96, 127.01, 126.98, 70.6, 68.1, 45.0,
39.9, 38.3, 31.1, 29.1, 24.4, 23.5, 14.5, 11.0; IR nmax(KBr)/cmꢀ1 2950,
2914, 2872, 1705, 1659, 1581,1453,1375,1333,1241. ESI-MS m/z calcd
(decomp.) 1H NMR (CDCl3, 250 MHz)
d¼8.67 (m, 4H, Ar–H), 4.38
(2H, t, J¼5.8 Hz, (CO2)NCH2CH2), 4.16 (1H, dd, J¼7.7 and 13.0 Hz,
(CO)2NCHHCH), 4.10 (1H, dd, J¼13.0 and 7.3 Hz, (CO)2NCHHCH),
3.79 (2H, t, J¼5.8 Hz, OCH2CH2), 3.65–3.61 (2H, m, 2H, OCH2CH2),
3.54–3.50 (2H, m, OCH2CH2), 3.41 (2H, t, J¼5.1 Hz, OCH2CH2),
2.79–2.74 (2H, m, CH2NH2), 2.27 (br, 2H, NH2), 1.95–1.84 (1H, m,
CH2CH(CH2)2),1.40–1.17 (8H, m, 4ꢃCH2), 0.89–0.78 (6H, m, 2ꢃCH3);
for
C50H54N4O10: 871.3918; found: 871.3943. Anal. Calcd for
C
50H54N4O10$H2O: C, 67.55; H, 6.35; N, 6.30%. Found: C, 67.70; H,
6.06; N, 6.39%.
4.6. N-Pyren-1-ylmethyl-benzamide 10
13C NMR (CDCl3, 62.5 MHz)
d
¼163.5, 163.3, 131.42, 130.37, 127.0,
To
a solution of 1-pyrenemethylamine hydrochloride 11
126.8, 73.0, 70.7, 70.4, 68.3, 45.0, 41.8, 39.0, 38.3, 31.1, 29.0, 24.4,
23.4, 14.5, 11.0. IR nmax(KBr)/cmꢀ1 3416, 2955, 2870, 1700, 1661,
1579, 1452, 1333, 1241, 1093, 766. ESI-MS m/z calcd for C28H36N3O6:
510.2604, found: 510. 2601.
(0.268 g, 1.00 mmol) and dry triethylamine (2 mL, 14.4 mmol) in
dry chloroform (50 mL) was added a solution of benzoyl chloride 12
(0.280 g, 0.231 mL, 2.00 mmol) with stirring over a period of 5 min.
The reaction was poured into n-hexane (200 mL). A pale yellow
precipitate was formed and filtered off under vacuum. The crude
product was washed well with hydrochloric acid solution 0.5 M
(20 mL), water (50 mL) and acetone (10 mL). A white solid 10
(0.2681 g, 80% yield) was obtained and dried under high vacuum
4.9. Tris-diimide 16
A solution of amino diimide 15 (50.0 mg, 0.0982 mmol) and
1,4,5,8 naphthalene dicarboxylicanhydride (12.5 mg, 0.0468 mmol)
were stirred at 135 ꢂC for 20 h in a solution of DMAc (2.5 mL) and
toluene (0.25 mL). After cooling to room temperature, the dark
solution was added dropwise into chloroform/methanol mix (5 mL,
1:1, v/v), which was stirred rapidly for 10 min. The resulting sus-
pension was collected by filtration and dried under vacuum at
70 ꢂC, then subjected to column chromatography (gradient chlo-
roform to chloroform/methanol (20:1 v/v)) to give tris-diimide 16
(43 mg, 75%) as a tan solid. Mp 257–260 ꢂC; 1H NMR (CDCl3,
pump. Mp 198 ꢂC; 1H NMR (CDCl3, 250 MHz)
d 7.98–8.33 (9H, m,
Py–H), 7.34–7.78 (5H, m, Ar–H), 6.48 (br, NH), 5.34–5.32 (2H, d,
NHCH2–Py, J¼5.2 Hz); 13C NMR (62.5 MHz, CDCl3)
d 167.58, 134.68,
131.98, 131.73, 131.64, 131.29, 131.13, 129.55, 129.00, 128.78, 128.03,
127.75, 127.74, 127.42, 126.55, 125.86, 125.81, 125.46, 125.21, 125.09,
123.19, 43.01; IR nmax(KBr)/cmꢀ1 3350, 1665, 1500; HRMS (CI):
C24H17NO m/z calcd: 335.1310; found: 335.1313.
4.7. Trityl protected diimide 14
250 MHz)
d
¼8.89 (8H, s, Ar–H), 8.51 (4H, s, Ar–H), 4.33 (8H, t,
A stirred solution of 2-{2-[2-(trityl-amino)-ethoxy]-ethoxy}-
ethylamine 13 (195 mg, 0.500 mmol) and diacid imide 7 (199 mg,
0.500 mmol) were dissolved in DMAc (5 ml) and toluene (0.5 mL)
and heated at 135 ꢂC for 16 h. After cooling to room temperature,
chloroform (50 mL) was added to the dark suspension and washed
with water (4ꢃ20 mL) and brine (20 mL) before being dried over
MgSO4, filtered and the solvent removed under reduced pressure.
The resulting black oil was added dropwise into methanol and
stirred rapidly for 10 min, producing a cream precipitate, which
was collected by filtration and subjected to column chromatogra-
phy (gradient 1:9 to 1:1 EtOAc/hexane) to give 14 as a tan solid
J¼5.6 Hz, 4ꢃ(CO2)NCH2CH2), 4.12 (2H, dd, J¼13.0 and 7.8 Hz,
(CO)2NCHHCH), 4.08 (2H, dd, J¼13.0 and 7.5 Hz, (CO)2NCHHCH),
3.76 (8H, t, J¼5.6 Hz, 4ꢃOCH2CH2), 3.67 (8H, s, 4ꢃOCH2CH2), 1.94–
1.87 (2H, m, CH2CH(CH2)2), 1.42–1.23 (16H, m, 8ꢃCH2), 0.89–0.78
(12H, m, 4ꢃCH3); 13C NMR (CDCl3/(CF3)2CHOH, 100 MHz)
¼163.6,
d
163.3, 163.2, 131.2, 126.72, 126.69, 126.59, 126.4, 126.3, 69.9, 67.9,
44.8, 39.5, 37.9, 30.6, 28.6, 24.0, 23.0, 14.0, 10.5; IR nmax(KBr)/cmꢀ1
2954, 2912, 2875, 1707, 1663, 1580,1452, 1241; ESI-MS m/z calcd for
C
C
70H70N6O16$Na: 1273.4746; found: 1273.4778. Anal. Calcd for
70H70N6O16$H2O: C, 66.23; H, 5.72; N, 6.62%. Found: C, 66.48; H,
5.76; N, 6.54%.
(319 mg, 85%). Mp 131–132 ꢂC; 1H NMR (CDCl3, 250 MHz)
d
¼8.68
(4H, m, 4ꢃNaphthalene–H), 7.46–7.09 (15H, m, Ar–H), 4.42 (2H, t,
J¼5.8 Hz, (CO2)NCH2CH2), 4.16 (1H, dd, J¼7.7 and 12.9 Hz,
(CO)2NCHHCH), 4.10 (1H, dd, J¼6.0 and 12.9 Hz, (CO)2NCHHCH),
3.80 (2H, t, J¼5.8 Hz, OCH2CH2), 3.67–3.64 (2H, m, OCH2CH2), 3.57–
3.45 (6H, m, 2H, 2ꢃOCH2CH2 and NHCH2CH2), 2.27 (2H, t, J¼5.4 Hz,
NHCH2CH2), 2.00–1.90 (3H, br m, CH2NH2 and CH2CH(CH2)2), 1.44–
Acknowledgements
The authors wish to thank Mr. T. Gasa (Northwestern University)
for help with the determination of binding constants, and EPRSC
(EP/D07434711) for financial support.

8354
B.W. Greenland et al. / Tetrahedron 64 (2008) 8346–8354
11. Hansen, J. G.; Feeder, N.; Hamilton, D. G.; Gunter, M. J.; Becher, J.; Sanders,
References and notes
J. K. M. Org. Lett. 2000, 2, 449–452.
12. Colquhoun, H. M.; Zhu, Z. X.; Williams, D. J. Org. Lett. 2003, 5, 4353–4356.
13. (a) Ghosh, S.; Ramakrishnan, S. Agnew. Chem., Int. Ed 2004, 43, 3264–3268; (b)
Ghosh, S.; Ramakrishnan, S. Macromolecules 2005, 38, 676–686.
14. (a) Lokey, R. S.; Iverson, B. L. Nature 1995, 375, 303–305; (b) Lokey, R. S.; Kwok,
Y.; Guelev, V.; Pursell, C. J.; Hurley, L. H.; Iverson, B. L. J. Am. Chem. Soc. 1997, 119,
7202–7210; (c) Nguyen, J. Q.; Iverson, B. L. J. Am. Chem. Soc. 1999, 121, 2639–
2640.
15. (a) Zhou, Q. Z.; Jiang, X. K.; Shao, X. B.; Chen, G. J.; Jia, M. X.; Li, Z. T. Org. Lett.
2003, 5, 2763; (b) Zhou, Q. Z.; Jiang, X. K.; Shao, X. B.; Chen, G. J.; Jia, M. X.; Li,
Z. T. Org. Lett. 2003, 5, 1955–1958; (c) Zhao, X.; Jia, M. X.; Jiang, X. K.; Wu, L. Z.;
Li, Z. T.; Chen, G. J. J. Org. Chem. 2004, 69, 270–279; (d) Zhou, Q. Z.; Jia, M. X.;
Shao, X. B.; Wu, L. Z.; Jiang, X. K.; Li, Z. T.; Chen, G. J. Tetrahedron 2005, 61,
7117–7124.
16. (a) Colquhoun, H. M.; Zhu, Z. X. Angew. Chem., Int. Ed. 2004, 43, 5040–5045; (b)
Colquhoun, H. M.; Zhu, Z. X.; Cardin, C. J.; Gan, Y. Chem. Commun. 2004, 2650–
2652; (c) Colquhoun, H. M.; Zhu, Z. X.; Cardin, C. J.; Gan, Y.; Drew, M. G. B. J. Am.
Chem. Soc. 2007, 129, 16163–16174.
17. Buncel, E,; Mailloux, N. L.; Brown, R. S.; Kazmaier, P. M.; Dust, J. Tetrahedron Lett.
2001, 42, 3559–3562.
1. Bergman, S. D.; Wudl, F. J. Mater. Chem. 2007, 18, 41–62.
2. Ciferri, A. Macromol. Rapid Commun. 2002, 23, 511–529.
3. Chen, X. X.; Dam, M. A.; Ono, K.; Mal, A.; Shen, H. B.; Nutt, S. R.; Sheran, K.;
Wudl, F. Science 2002, 295, 1698–1702.
4. White, S. R.; Sottos, N. R.; Geubelle, P. H.; Moore, J. S.; Kessler, M. R.; Sriram,
S. R.; Brown, E. N.; Viswanathan, S. Nature 2001, 409, 794–797.
5. (a) Mynar, J. L.; Aida, T. Nature 2008, 451, 895–896; (b) Cordier, P.; Tournilhac, F.;
Soulie´-Ziakovic, C.; Leibler, L. Nature 2008, 451, 977–980.
6. (a) Swiegers, G. F.; Malefetse, T. J. Chem. Rev. 2000, 100, 3483–3537; (b)
Schmuck, C.; Wienand, W. Angew. Chem., Int. Ed. 2001, 40, 4363–4369; (c)
Brunsveld, L.; Folmer, B. J. B.; Sijbesma, R. P.; Meijer, E. W. Chem. Rev. 2001, 101,
4071–4097; (d) Sijbesma, R. P.; Cate, A. T. Macromol. Rapid. Commun. 2002, 23,
1094–1112; (e) Hofmeier, H.; Schubert, U. S. Chem. Soc. Rev. 2004, 33, 373–399;
(f) Hoeben, F. J. M.; Jonkheijm, P.; Meijer, E. W.; Schenning, A. P. H. J. Chem. Rev.
2005, 105, 1491–1546; (g) Sivakova, S.; Rowan, S. J. Chem. Soc. Rev. 2005, 34,
9–21; (h) Perez-Garcia, L.; Amabilino, D. B. Chem. Soc. Rev. 2007, 36, 941–967.
7. Brunsveld, L.; Folmer, B. J. B.; Sijbesma, R. P.; Meijer, E. W.; Beijer, F.; Lange,
R. F. M.; Lowe, J. K. L. Science 1997, 278, 1601–1604.
8. Sivakova, S.; Bohnsack, D. A.; Mackay, M. E.; Suwanmala, P.; Rowan, S. J. J. Am.
Chem. Soc. 2005, 127, 18202–18211.
9. (a) Sontjens, S. H. M.; Sijbesma, R. P.; van Genderen, M. H. P.; Meijer, E. W. J. Am.
Chem. Soc. 2000, 122, 7487–7493; (b) Sontjens, S. H. M.; Sijbesma, R. P.; van
Genderen, M. H. P.; Meijer, E. W. Macromolecules 2001, 34, 3815–3818.
10. Foster, R. Organic Charge Transfer Complexes; Academic: London,1969; pp 33–93.
18. Langhals, H. Heterocycles 1995, 40, 477–500.
19. Nielsen, M. B.; Jeppesen, J. O.; Lau, J.; Lomholt, C.; Damgaard, D.; Jacobsen, J. P.;
Becher, J.; Stoddart, J. F. J. Org. Chem. 2001, 66, 3559–3563.
20. Hauquier, F.; Pastorin, G.; Hapiot, P.; Prato, M.; Bianco, A.; Fabre, B. Chem.
Commun. 2006, 4536–4538.