Construction of supramolecular multi-component assemblies by using allosteric interactions

Article abstract of DOI:10.1016/j.tet.2008.05.139

Supramolecular complexes between two cavity-appended porphyrin hosts and three bifunctional guests are described. The host with a single cavity exclusively forms dimers with the bifunctional guests, while the double cavity host yields tetramers and higher order assemblies. The role of allosteric interactions in the binding and assembly process is highlighted.

Full text of DOI:10.1016/j.tet.2008.05.139

Tetrahedron 64 (2008) 8535–8542
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Construction of supramolecular multi-component assemblies
by using allosteric interactions
Nico Veling a, Paul J. Thomassen a, Pall Thordarson b, Johannes A.A.W. Elemans a,
,
c,
*
Roeland J.M. Nolte a , Alan E. Rowan
*
a Department of Organic Chemistry, Institute for Molecules and Materials, Radboud University Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands
b School of Chemistry, The University of New South Wales, NSW 2052, Australia
c Department of Molecular Materials, Institute for Molecules and Materials, Radboud University Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Supramolecular complexes between two cavity-appended porphyrin hosts and three bifunctional guests
are described. The host with a single cavity exclusively forms dimers with the bifunctional guests, while
the double cavity host yields tetramers and higher order assemblies. The role of allosteric interactions in
the binding and assembly process is highlighted.
Received 18 March 2008
Received in revised form 23 May 2008
Accepted 30 May 2008
Available online 5 June 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Self-assembly
Host–guest chemistry
Cooperativity
Porphyrins
1. Introduction
both positive and negative in nature, and can act between a host and
identical guests (homotropic allosterism) or between a host and
Cooperative interactions are common in nature; they are, for
instance, employed to influence the composition and function of
hierarchical, self-assembled systems such as the tobacco mosaic
virus, and also to transfer information, like in the binding of oxygen
to haemoglobin.1–4 Cooperative interactions also play a critical role
in gene transcription; for instance, cyclic AMP has a strong co-
operative effect on the binding of the gene transcription regulating
cAMP receptor protein (CRP) to DNA.5 Chemists have recognised
that utilising cooperative interactions might also help in the con-
struction of functional nanoscale objects. In the past decades this
type of interaction has received increasing attention and a myriad
of self-assembled structures that employ cooperative effects have
been developed.6,7
Cooperative effects have been widely used and studied in the
field of supramolecular chemistry and are interesting tools to con-
trol the properties of host–guest systems8 and polymers.9 Allos-
terism is a special case of cooperativity,10 in the sense that a binding
event at one site in a multivalent system like a biomolecule11,12 or
synthetic host13 causes a discrete, reversible alteration in the
structure at a remote binding site. Allosteric interactions can be
different types of guests (heterotropic allosterism).
In previous research in our group a number of cavity-appended
porphyrin hosts that display allosteric binding behaviour have been
developed.14–17 The most extensively studied host molecule is the
monocavity-appended zinc porphyrin ZnMC, which is depicted in
Figure 1 together with its allosteric binding behaviour. It was found
that ZnMC can bind viologen molecules with high binding con-
stants. The addition of a nitrogen-donor ligand that is too bulky to
fit inside the cavity, like 4-tert-butyl-pyridine (tbpy), results in
a positive heterotropic allosteric effect, which is demonstrated by
an increase in binding constant between dimethylviologen (V) and
ZnMC (Fig. 1b). Vice versa, an increase in binding constant between
tbpy and ZnMC was observed upon the addition of viologen mole-
cules to a mixture of the host and the axial ligand.
In the double cavity analogue of ZnMC and ZnDC, the in-
teractions are even more complicated (Fig. 2). For the binding of
viologens, this double cavity porphyrin system shows highly neg-
ative allosteric binding. The free base host molecule has the ability
to bind two dimethylviologen molecules, ꢁb1ut the association con-
stant for the second binding (Ka¼5ꢀ104 M ) is considerably lower
than for the first binding event (Ka¼7ꢀ107 Mꢁ1). A combination of
NMR studies and computational modelling indicated that this
lower binding was not due to electronic repulsion between the two
guests, but the result of conformational changes in the host
molecule.15
* Corresponding authors. Tel.: þ31 24 3652323; fax: þ31 24 3652929.
E-mail addresses: r.nolte@science.ru.nl (R.J.M. Nolte), a.rowan@science.ru.nl
(A.E. Rowan).
0040-4020/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.05.139

8536
N. Veling et al. / Tetrahedron 64 (2008) 8535–8542
Figure 1. (a) Structure of porphyrin host ZnMC, together with the relevant proton numbering. (b) Schematic representation of the allosteric binding properties of this compound.
Figure 2. (a) ZnDC and its schematic representation. (b) Negative homotropic allosteric binding behaviour of DC.
As a continuation of this research we describe here attempts to
construct more complex self-assembled allosteric assemblies based
on bifunctional viologen and pyridine guests, and both ZnMC and
ZnDC as hosts. The ultimate goal of this research is to generate
materials of which both the architecture and the properties can be
influenced by allosteric interactions.
acid by oxidation with CrO3 and then into an acid chloride using
standard conditions. The latter compound was subsequently
reacted with 3-aminopyridine to yield the corresponding amide. In
order to prevent cyclisation and oligomerisation, this amide was
treated with an excess of mono-methylated bipyridine at room
temperature.
2. Results and discussion
2.1. Synthesis
2.2. Self-assembled complexes based on ZnMC
The association constants of the complexes between the three
guest molecules and the host ZnMC were determined by UV–vis
and fluorescence spectroscopy (Table 1). All association constants
were calculated assuming independent (non-cooperative) binding
of the separate binding moieties of the guest molecules.
To study allosteric assembly three different bifunctional guests
were synthesised, one containing two pyridine moieties (PyPy),
another one containing two viologen moieties (VV) and a third one
containing one pyridine and one viologen moiety (VPy) (Fig. 3).
Bis-pyridine bifunctional guest PyPy was synthesised by
a straightforward reaction of suberoyl chloride with 3-amino-
pyridine (yield 73% after recrystallisation). An amide bond linker
was chosen because it was found to increase the binding of the
pyridine guest in ZnMC due to the formation of a weak hydrogen
bond with one of the carbonyl groups in the cavity of the host.14
For the synthesis of diviologen bifunctionalised guest VV, 1,10-
dibromodecane was reacted with an excess of 4,40-bipyridine to
give an
a,u-bis-bipyridyl-functionalised decane. This compound
was then treated with iodomethane, after which anion exchange
was carried out by dissolving the compound in a hot aqueous so-
lution of ammonium hexafluorophosphate. Upon cooling, the de-
sired product VV precipitated as the tetrakis-hexafluorophosphate
salt in a yield of 28%.
Viologen-pyridine bifunctionalisedguest VPy was prepared from
9-bromo-1-nonanol, which was converted into 9-bromononanoic
Figure 3. The bifunctional guests used in this study and their proton assignment (left).
Schematic representation of the guest molecules (right).

N. Veling et al. / Tetrahedron 64 (2008) 8535–8542
8537
Table 1
binding process. A 1:1 stoichiometry was observed for the VV–
ZnMC complex without an axial ligand present, whereas the VV–
ZnMC complex in the presence of 10 equiv of tbpy showed a 1:2
stoichiometry. From the NMR spectra the geometry of the 1:2 VV/
ZnMC complex could be derived (Fig. 5). A splitting of the proton
signals 5–7 and 11–13 of ZnMC into two sets of signals for each
proton had occurred, which can be attributed to the non-sym-
metrical complexation of the viologen guests inside the cavities.
Only four doublets were observed for the viologen protons, and
since the signals of the host and the VV guest are in slow exchange
in the NMR spectrum, this observation indicates that two ZnMC
hosts bind to one VV guest. Only very broad signals were found for
the free tbpy ligand (indicated by an asterisk in Fig. 5), indicating
that it is in relatively fast exchange with axially coordinated tbpy.
Knowing that both VV and PyPy are excellent guests for ZnMC,
the combination of all three host–guest components was attemp-
ted. NMR studies on a 2:1:1 molar mixture of ZnMC, VV and PyPy
in CDCl3/CD3CN 1:1 (v/v) revealed that the discrete self-assembled
complex depicted in Figure 6 had formed exclusively. Again, dou-
bling of the signals of protons 5–7 and of 11–13 indicated a strong
non-symmetrical complexation of the guest inside the host cavi-
ties, whereas the upfield shifts of the signals for the protons of both
the viologen and pyridine moieties of both bifunctional guests
confirmed the complexation to ZnMC.
In the case of the hetero-bifunctional guests VPy and ZnMC, an
association constant of roughly Ka¼1.0ꢀ107 Mꢁ1 between the
viologen moiety and the host could be calculated from the titration
curve obtained by UV–vis spectroscopy (Fig. 7). Since the binding
affinity of the pyridine is at least two orders of magnitude smaller
than that of the viologen, the formation of a 2:1 complex between
ZnMC and VPy in which both ends of the guest molecule are bound
to two different hosts was not included in the fitting model. The
association constant of the complex with VPy is higher than the
association constant of the complex with VV (Ka¼4.6ꢀ106 Mꢁ1),
suggesting that the pyridine moiety acts as an allosteric effector.
a
Association constants Ka and binding free energies
different guests
D
G of complexes of ZnMC and
Guest
Ka (Mꢁ1
)
D
G (kJ molꢁ1
)
VV
4.6ꢀ106
2.2ꢀ108
5ꢀ105
ꢁ38
ꢁ48
ꢁ33
ꢁ38
VVb
PyPy
Vpy
1.0ꢀ107
a
Determined by UV–vis spectroscopy, CHCl3/CH3CN 1:1 (v/v) at 298 K,
[ZnMC]¼1.5ꢀ10ꢁ6 M. Association constants determined by fluorescence spectros-
copy (excitation wavelength¼426 nm; emission wavelength (max)¼605 nm) were
compared with the association constants obtained from UV–vis data. They were
found to be similar and are therefore not mentioned. In their calculation it was
assumed that the quantum yields of the host and of the host–guest complexes were
the same.
b
Association constant in the presence of 500 equiv tbpy.
The association constant obtained for the complex between VV
and ZnMC was approximately five times higher than that of the
complex between this host and V (Ka¼9ꢀ105 Mꢁ1), which is in line
with the general observation that viologens equipped with long
alkyl chains bind more strongly in ZnMC.16 The accuracy of the fit
(Fig. 4) deviated at higher concentrations of the guest suggesting
additional binding geometries.
A 40-fold increase in association constant of VV was observed
upon the addition of an excess tbpy. More remarkable, however,
was the fact that after the addition of only half an equivalent of VV
to ZnMC the titration curve reached saturation, whereas a full
equivalent was needed to reach this level in the absence of tbpy
(Fig. 4). 1H NMR confirmed the observed allosteric effect in the
Figure 5. Computer modelled structure of the 1:2 assembly of VV and ZnMC in the
presence of tbpy (top, please note that the tbpy is in fast exchange). 1H NMR spectrum
(400.15 MHz, CDCl3/CD3CN 1:1 (v/v), 298 K, [ZnMC]¼2.90ꢀ10ꢁ3 M) in a 1:2:10 molar
VV/ZnMC/tbpy mixture (bottom). The proton signals of tbpy are indicated by an as-
terisk. For proton numbering see Figures 1 and 3. See Supplementary data for the 2D
NMR spectra.
Figure 4. Titration curves for the binding of VV in ZnMC obtained by UV–vis spec-
troscopy in CHCl3/CH3CN 1:1 (v/v) in the absence of tbpy (top) and in the presence of
500 equiv of tbpy (bottom).

8538
N. Veling et al. / Tetrahedron 64 (2008) 8535–8542
guest bind to the same host molecule, i.e., the viologen moiety of
VPy inside the cavity of ZnMC and the pyridine moiety to the
porphyrin zinc ion on the outside of the host, despite the increased
effective molarity.
According to 1H NMR spectroscopy, the addition of 1 equiv of
VPy to ZnMC in CDCl3/CD3CN 1:1 (v/v) resulted in the full com-
plexation of the viologen moiety inside the cavity of the host and
the full axial binding of pyridine to the zinc ion. The complexation
of a non-symmetrical viologen inside the cavity, in combination
with the axial complexation of a non-symmetrical pyridine on top
of the host, evidenced by an upfield shift of the signals of the
pyridine protons, causes an extreme level of dissymmetry in ZnMC,
which was indicated by the splitting of the NMR signals of protons
5–7, 9–11 and 13 (Fig. 7). The dissymmetric conformation of the
complex was also revealed by an unequal upfield shift of the signals
of the protons of the two respective pyridine groups in the viologen
moiety of VPy. The formation of the proposed 2:2 assembly was
confirmed by COSY and NOESY 2D NMR studies.
Apparently, upon complexation with ZnMC, the equally sized
spacers of the bifunctional guests favour the formation of discrete
assemblies over supramolecular oligomers or polymers. To in-
vestigate whether a mismatch with regard to the alkyl spacers
between a VV guest and PyPy guest in combination with ZnMC
would disfavour the self-assembly of the components into a dis-
crete complex and favour the formation of polymeric structures,
a fourth bifunctional guest was synthesised, containing two viol-
Figure 6. Computer modelled structure of the 2:1:1 assembly of ZnMC, VV and PyPy
(top). 1H NMR spectrum (400.15 MHz, CDCl3/CD3CN 1:1 (v/v), 298 K,
[ZnMC]¼2.16ꢀ10ꢁ3 M) in a 1:2:1 molar VV/ZnMC/PyPy mixture (bottom). The proton
signals of PyPy are indicated by an asterisk. For proton numbering see Figures 1 and 3.
See Supplementary data for the 2D NMR spectra.
ogen moieties, separated by an alkyl spacer of 16 CH2-units (V–C16
–
V). The addition of half an equivalent of V–C16–V and half an
equivalent of PyPy to a solution of ZnMC in CDCl3/CD3CN 1:1 (v/v)
([ZnMC]¼2ꢀ10ꢁ3 M) also resulted in the formation of a discrete
Molecular modelling studies on a 1:1 complex between ZnMC
and VPy revealed that there is a too high energy penalty for
adopting a conformation in which both binding moieties of the
Figure 7. Titration curve for the binding of VPy in ZnMC as measured by UV–vis spectroscopy in CHCl3/CH3CN 1:1 (v/v) (top left). Computer modelled structure of the 2:2 assembly
of ZnMC and VPy (top right). 1H NMR spectrum (400.15 MHz, CDCl3/CD3CN 1:1 (v/v), 298 K, [ZnMC]¼1.06ꢀ10ꢁ3 M) in a 1:1 mixture of ZnMC and VPy (bottom). For proton
numbering see Figures 1 and 3. See Supplementary data for the 2D NMR spectra.

N. Veling et al. / Tetrahedron 64 (2008) 8535–8542
8539
Table 2
self-assembled 2:1:1 complex. The formation of supramolecular
polymers apparently is entropically not favoured when compared
to the formation of discrete host–guest complexes, which is prob-
ably the result of the increased effective molarity.6
a
Association constants Ka and binding free energies
ZnDC with different guests
D
G for the complexation of
Guest
Ka (Mꢁ1
)
D )
G (kJ molꢁ1
PyPy
VV
4ꢀ105
1ꢀ108
7ꢀ109
ꢁ32
ꢁ46
VVb
VPy
ꢁ56
2.3. Self-assembled complexes containing ZnDC
1ꢀ106; 1ꢀ108
ꢁ34; ꢁ46
a
Determined by UV–vis spectroscopy, CHCl3/CH3CN 1:1 (v/v) at 298 K,
[ZnDC]¼2.6ꢀ10ꢁ6 M. Association constants determined by fluorescence spectros-
copy (excitation wavelength¼426 nm; emission wavelength (max)¼605 nm) were
compared with the association constants obtained from UV–vis data. They were
found to be similar and are therefore not mentioned. In their calculation it was
assumed that the quantum yields of the host and of the host–guest complexes were
the same.
In a second series of experiments the formation of self-assem-
bled allosteric complexes of ZnDC was studied. Since the two
cavities of ZnDC are oriented orthogonally to each other, the
combination of VPy with ZnDC or VV and PyPy with ZnDC cannot
easily lead to complexes of small size, hence the formation of larger
assemblies was expected.
b
Association constant in the presence of 1 equiv PyPy.
First it was investigated whether ZnDC displayed allosteric
binding behaviour upon the complexation of both a pyridine and
a viologen guest. To this end, the association constant of the com-
plex between ZnDC and pyridine (Py) was determined in the
presence and absence of 1 equiv of V. In the absence of V the as-
sociation constant of the complex in CHCl3/CH3CN 4:1 (v/v) was
measured to be Ka¼2.2ꢀ104 Mꢁ1, which is somewhat lower than
the association constant of the complex between Py and ZnMC
(Ka¼7.5ꢀ104 Mꢁ1).18 The association constant between Py and
ZnDC in the presence of 1 equiv of V increased fivefold to
Ka¼1.1ꢀ105 Mꢁ1, which indicates a significant positive heterotropic
allosteric binding behaviour (Fig. 8). This, in combination with the
negative homotropic binding of V, makes ZnDC an ideal host for the
formation of higher molecular weight self-assembled complexes
using the bifunctional guests shown in Figure 3.
enough to completely quench the fluorescence of the zinc por-
phyrin in the titration experiment followed by fluorescence spec-
troscopy, whereas in absence of PyPy a whole equivalent was
required. The deviation in the latter stage of the titration from the
fitting curve seems to indicate that several different self-assembled
structures are formed. The formation of non-discrete self-assem-
bled oligomers, together with the possible formation of [4]-
pseudo-rotaxanes, are the most obvious possibilities. Further
studies are currently in progress to investigate the structural dis-
tribution in detail.
Similar to what was observed for the ZnMC upon the addition of
only half an equivalent of VPy to ZnDC complete binding and
The binding affinities of the three different bifunctional guests
for ZnDC were measured, again assuming independent (non-co-
operative) binding between the two binding moieties of the guests
(Table 2). The bifunctional guest VV gave an association constant of
Ka¼1ꢀ108 Mꢁ1, almost two orders of magnitude higher than the
association constant measured for ZnMC (Ka¼4.6ꢀ106 Mꢁ1). In
fitting the binding data it was assumed that only [3]-pseudo-
rotaxanes were formed, because it is expected that ZnDC displays
similar negative allosteric binding behaviour upon the binding of
two viologen moieties as its free base analogue DC. This assumption
may not completely reflect reality; however, a reasonable fit was
obtained (Fig. 9).
The titration was repeated in the presence of 1 equiv of PyPy to
evaluate whether allosteric binding occurred. A 70-fold increase in
binding constant of VV proved that this was indeed the case (Fig. 9).
In the presence of PyPy already half an equivalent of VV was
Figure 9. Titration curves for the binding of VV in ZnDC obtained by UV–vis spec-
troscopy in CHCl3/CH3CN 1:1 (v/v) in the absence of PyPy (top) and in the presence of
1 equiv of PyPy (bottom).
Figure 8. Positive heterotropic allosteric binding behaviour of ZnDC.

8540
N. Veling et al. / Tetrahedron 64 (2008) 8535–8542
fluorescence quenching was observed. By assuming that only
the viologen moiety would be engaged in binding, no adequate
fit was obtained, however, when it was assumed that both ends
could bind, two association constants were obtained, viz.
Ka(1)¼1ꢀ106 Mꢁ1 and Ka(2)¼1ꢀ108 Mꢁ1 (Fig. 10). These values
might be the association constants for pyridine and viologen
binding to ZnDC, respectively, which corresponds to the association
constants found for the two other bifunctional guests (Table 2 en-
tries 1 and 2). This conclusion would suggest that in this case the
pyridine moiety does not act as an allosteric effector. A possible
reason for this may be that there are other factors that influence the
allosteric binding process, for example, geometric or steric factors.
To obtain more information about the geometry of the formed
complex, 1H NMR experiments were carried out on a 1:1 mixture of
ZnDC and VPy in CDCl3/CD3CN 1:1 (v/v). In contrast to all previous
experiments, a very broad NMR spectrum was obtained in which no
clear structure picture could be observed. When the host and guest
were dissolved in a CDCl3/CD3CN 19:1 (v/v) mixture, the resolution
was increased, but still only the separate regions of the ZnDC could
be assigned. Numerous different signals were observed indicating
that a variety of complexes are present in solution. According to the
isotope pattern of the peak corresponding to a mass of [ZnDC–VPy-
PF6]nnþ in a high resolution ESI MS spectrum of the 1:1 mixture, at
least four molecules of each component were present.
As an alternative investigative tool a diffusion-ordered NMR
(DOSY NMR) spectrum of the 1:1 mixture of ZnDC and VPy was
recorded in CDCl3/CD3CN 1:1 (v/v) applying a 10ꢁ3 M concentration
for ZnDC. Somewhat surprisingly, the DOSY NMR spectrum sug-
gested that the solution contained only one complex with a well-
defined diffusion coefficient. Using the Einstein–Stokes equation it
was possible to estimate an approximate radius of this complex,
which amounted to 193 Å (see Section 4). The measured value
suggests the formation of a complex that contains many host and
guest molecules. However, given the strong allosteric behaviour of
the ZnDC, an equally defined structure is expected as seen for the
ZnMC complexes. All the above data can be explained by a defined
4:4 tetrameric complex (with four different diastereomeric struc-
tures, Fig. 10), although supramolecular polymers cannot be
excluded.
of the assemblies, the bifunctional guests were combined with the
double cavity host ZnDC. The binding behaviour of this host to-
wards the bifunctional guests proved to be complex and suggests
formation of allosterically driven tetrameric assemblies. The com-
bination of allosterism and self-assembly as demonstrated ele-
gantly by nature, opens genuine possibilities for the controlled
construction of hierarchical architectures. The present studies are
just a first step in this direction.
4. Experimental
4.1. Methods and materials
Acetonitrile-d3 was distilled over CaH2 and chloroform-d was
distilled over CaCl2 prior to use. ZnMC18 and ZnDC15 were syn-
thesised according to the literature procedures. NMR spectra were
obtained on Bruker DPX 300, Varian Unity Inova 400 and Bruker
DRX 500 instruments at 298 K. All NMR studies of host–guest
complexes were performed in freshly prepared CDCl3/CD3CN 1:1
(v/v) mixtures at 1ꢀ10ꢁ3 M concentrations. All the ratios that are
mentioned with regard to mixtures of hosts and guests are molar
ratios. UV–vis spectra were measured on a Varian Cary 50 UV–vis
spectrophotometer and MALDI-TOF MS spectra were recorded on
a Bruker Biflex III spectrometer. ESI mass spectra were measured on
a Jeol Accu-TOF-CS. IR spectra were recorded on a ATI Matson
Genesis Series FTIR spectrometer with an ATR cell. Fluorescence
titrations were performed on a Perkin–Elmer Luminescence Ls50B
spectrometer. UV–vis and fluorescence titration experiments and
the calculation of binding constants were carried out following the
standard methods reported earlier by our group assuming in-
dependent binding of the separate binding moieties of the guest
molecules.15,17
Diffusion rate values in DOSY NMR studies were obtained using
data analysis in Mestre-C 4.7.0.0. The diffusion rate that was used
for the calculation of the radius was an average of the diffusion
rates obtained for at least six different peaks in the DOSY NMR
spectra to minimise errors. The radius was calculated from the
diffusion rate, with the assumption that the viscosity of a 1:1 (v/v)
mixture of CDCl3 and CD3CN is the average (0.4435ꢀ
10ꢁ3 kg mꢁ1 sꢁ1) of the viscosity values of the two individual
solvents, using the rearranged Einstein–Stokes equation:
3. Conclusions
An array of bifunctional guests have been synthesised contain-
ing either two viologen or two pyridine moieties, or one viologen
and one pyridine moiety. In combination with ZnMC these guests
were shown to form discrete self-assembled complexes in solution,
which according to UV–vis and fluorescence titrations displayed
allosteric interactions. In order to increase the size and complexity
r ¼ kT=6
D
p h
in which r¼radius (m), k¼Boltzmann’s constant (J Kꢁ1),
T¼temperature (K), D¼diffusion rate (m2 sꢁ1) and
¼viscosity of
h
the medium (kg mꢁ1 sꢁ1).
Figure 10. Titration curve for the binding of VPy in ZnDC obtained by UV–vis spectroscopy in CHCl3/CH3CN 1:1 (v/v) (left) and schematic representation of possible self-assembled
allosteric complexes in which ZnDC and VPy are present in a 1:1 ratio (right).

N. Veling et al. / Tetrahedron 64 (2008) 8535–8542
8541
4.2. Synthesis of PyPy
to a solution of 9-bromononanol (5.4 g, 24.2 mmol) in acetone
(250 mL). After stirring at room temperature for 2 h a green pre-
cipitate had formed, which was filtered off and discarded. The fil-
trate was concentrated and a solution of the resulting solid in Et2O
was washed with water (3ꢀ). The organic layer was dried (MgSO4),
filtered and evaporated, resulting in a solid that was purified by
column chromatography (silica; hexane/ethyl acetate 20:1 (v/v)),
yielding a white solid. A mixture of this solid (0.41 g, 1.7 mmol) and
thionyl chloride (2 mL) was stirred overnight, after which the ex-
cess liquid was evaporated until an oil remained that was dissolved
in Et2O (5 mL). This solution was added dropwise to an ice-cold
solution of 3-aminopyridine (0.16 g, 1.7 mmol) in Et2O (15 mL).
After warming to room temperature, the mixture was stirred
overnight under nitrogen. The resulting suspension was filtered
and the residue dissolved in water, after which NaHCO3 was added
until pH 8 was reached. The product was extracted with Et2O, the
organic layer was washed with a saturated aqueous NaHCO3 solu-
tion, dried with MgSO4, filtered and evaporated, resulting in a white
To triethylamine (3.3 mL, 24 mmol) and 3-aminopyridine (1.6 g,
17 mmol) in CH2Cl2 (40 mL) was added dropwise at 0 ꢂC a solution
of suberoyl chloride (2.1 g, 9.8 mmol) in CH2Cl2 (3 mL). This mix-
ture was stirred at 0 ꢂC for 1 h and subsequently at room temper-
ature for 2 h. It was then poured into water and a white solid was
isolated via filtration, which was dissolved in a small volume of
DMSO and this solution was again poured into water. After filtra-
tion, the resulting solid was recrystallised from boiling MeOH and
by allowing Et2O to diffuse into a saturated MeOH solution of the
compound yielding a white solid (1.5 g, 6.4 mmol, 73%). 1H NMR
(300 MHz, CD3CN/CDCl3 1:1 (v/v)):
d
¼8.64 (s, 2H), 8.26 (d,
3J¼3.9 Hz, 2H), 8.19 (s, 2H), 8.08 (d, 3J¼8.3 Hz, 2H), 7.22 (dd, 3J¼8.3,
3J¼3.9 Hz, 2H), 2.37 (t, 3J¼7.2 Hz, 4H), 1.73 (m, 4H), 1.43 (m, 4H);
13C{1H} NMR (50 MHz, CDOD):
d
¼174.6, 144.6, 141.3, 137.1, 128.5,
124.9, 37.3, 29.6, 26.1; IR (solid) 1670, 1535, 1417, 703 cmꢁ1; UV–vis
(MeOH) /nm (log
/Mꢁ1 cmꢁ1): 240 (4.3), 278 (3.8); HRESIMS m/z
l
3
327.1820, calcd for C18H23N4O2 [M]þ 327.1821, 349.1650, calcd for
solid (0.4 g, 1.3 mmol). 1H NMR (300 MHz, CDCl3):
d
¼8.51 (d,
C
18H23N4O2Na [MþNa]þ 349.1640.
3J¼2.1 Hz,1H), 8.31 (d, 3J¼4.2 Hz,1H), 8.16 (d, 3J¼8.1 Hz,1H), 7.35 (s,
1H), 7.24 (m, 1H), 3.39 (t, 3J¼6.9 Hz, 2H), 2.39 (t, 3J¼7.5 Hz, 2H), 1.86
(m, 2H), 1.82 (m, 2H), 1.35 (m, 8H).
4.3. Synthesis of VV
A mixture of 1,10-dibromodecane (1.0 g, 3.3 mmol) and 4,40-
bipyridine (7.4 g, 47 mmol) in CH3CN (50 mL) was refluxed over-
night under nitrogen. After cooling, the resulting precipitate was
filtered off and washed with Et2O and CH3CN. The residue was
dissolved in water and after addition of a saturated aqueous NH4PF6
solution a precipitate was formed, which was filtered off, washed
with water and dried in vacuo. A solution of the resulting white
solid (0.99 g) and iodomethane (0.33 mL, 5.3 mmol) in CH3CN
(35 mL) was refluxed under nitrogen for 64 h. The precipitate was
filtered off and dissolved in water, and this solution was added
slowly to an aqueous saturated solution of NH4PF6 resulting in the
formation of a precipitate. The suspension was heated until the
entire solid redissolved again. Upon cooling, a white solid was
formed, which was filtered off, washed with water and recrystal-
lised by the allowing Et2O to diffuse into a saturated CH3CN solu-
tion of the compound yielding the title compound as a white solid
(0.45 g, 0.4 mmol, 28%), dp¼250 ꢂC. 1H NMR (400.15 MHz, CD3CN):
4.4.3. Synthesis of VPy
A
solution of 9-bromo-(N-3-pyridyl)nonamide (0.1 g,
0.32 mmol) and 1-methyl-4-(40-pyridyl)-pyridinium hexafluoro-
phosphate (0.5 g, 1.58 mmol) in CH3CN (500 mL) was stirred for 2
weeks. The precipitate was filtered off and washed CH3CN (3ꢀ),
dissolved in water and this solution was added dropwise to a sat-
urated aqueous NH4PF6 solution. The resulting precipitate was
isolated via centrifugation and recrystallised by allowing CHCl3 to
diffuse into a saturated CH3CN solution of the compound yielding
a white solid (8 mg, 0.01 mmol, 4%). 1H NMR (400.15 MHz, CD3CN):
d
¼8.85 (m, 4H), 8.70 (s, 1H), 8.35 (m, 5H), 8.24 (d, 3J¼6.6 Hz, 1H),
8.00 (d, 3J¼11 Hz, 1H), 7.30 (dd, J¼6.6, 3J¼11 Hz, 1H) 4.60 (t,
3J¼10 Hz, 2H), 4.39 (s, 3H), 2.34 (t, 3J¼9.6 Hz, 2H), 2.02 (m, 2H), 1.65
(m, 2H), 1.40 (m, 8H); 13C{1H} NMR (50 MHz, CD3CN):
d¼171.6,
145.0, 143.6, 140.2, 126.7, 126.3, 126.2, 123.2, 77.7, 61.6, 36.0, 30.4,
28.2, 28.1, 27.8, 24.9, 24.4 (not all peaks could be assigned due to
broadening); IR (solid) 1674, 1536, 820, 555 cmꢁ1; HRESIMS m/z
717.1799, calcd for C25H32F12N4ONaP2 [MþNa]þ 717.1757, 549.2241,
calcd for C25H32F6N4OP [MꢁPF6]þ 549.2217.
d
¼8.85 (m, 8H), 8.37 (m, 8H), 4.60 (t, 3J¼7.6 Hz, 4H), 4.40 (s, 6H),
2.00 (m, 4H), 1.37 (m, 12H); 13C{1H} NMR (50 MHz, CD3CN):
d
¼150.8, 150.6, 147.4, 146.5, 140.0, 63.1, 49.6, 32.0, 29.9, 29.6, 26.6;
IR (solid) 819, 560 cmꢁ1; HRESIMS m/z 1085.1888, calcd for
C32H42F24N4NaP4 [MþNa]þ 1085.1874, 917.2334, calcd for
4.5. Synthesis of V–C16–V
C
M
32H42F18N4P3 [MꢁPF6]þ 917.2334; UV–vis (CH3CN)
l/nm (log
3/
To a suspension of 1,16-dihydroxyhexadecane (1.0 g, 3.8 mmol)
in Et2O (20 mL) was added dropwise PBr3 (0.24 mL, 2.5 mmol), after
which the suspension was refluxed for 5 h, during which the solid
dissolved. After cooling, the solution was poured into water (50 mL)
and this suspension was extracted with CH2Cl2 (3ꢀ). The combined
organic layers were dried with Na2SO4, filtered and the solvent was
evaporated. The resulting solid was recrystallised from boiling
EtOH to yield 1,16-dibromohexadecane. A mixture of 1,16-dibromo-
hexadecane (0.556 g, 1.44 mmol) and 4,40-bipyridine (2.25 g,
14.4 mmol) in DMF (25 mL) was stirred at 90 ꢂC for 3 days under
nitrogen. After cooling, the resulting precipitate was filtered off and
washed with Et2O and CH3CN. The filtrate was evaporated and the
residue was dissolved in DMSO. After the addition of a saturated
aqueous NH4PF6 solution a precipitate was formed, which was fil-
tered off, washed with water and dried in vacuo. A solution of the
resulting white solid (65 mg, 0.12 mmol) and iodomethane (1 mL,
16 mmol) was refluxed in CH3CN (15 mL) for 14 h under nitrogen.
The precipitate was filtered off and added to a saturated aqueous
NH4PF6 solution. This mixture was heated to 100 ꢂC (3ꢀ) after
which a white precipitate was formed upon cooling. This solid was
filtered off, washed with water and Et2O, and dried in vacuo
ꢁ1 cmꢁ1): 265 (4.6).
4.4. Synthesis of VPy
4.4.1. Synthesis of 1-methyl-4-(40-pyridyl)-pyridinium
hexafluorophosphate
A solution of 4,40-bipyridine (1.2 g, 7.7 mmol) and methyl iodide
(0.57 mL, 9 mmol) in CH2Cl2 (5 mL) was refluxed overnight under
nitrogen. The resulting precipitate was filtered off, redissolved in
MeOH, precipitated by the addition of Et2O, filtered off and washed
with EtOH. To an aqueous solution of the resulting solid, a saturated
aqueous NH4PF6 solution was added, which resulted in the for-
mation of a precipitate (1.8 g, 5.7 mmol, 74%) that was filtered off.
Physical properties were in agreement with those previously
reported.19
4.4.2. Synthesis of 9-bromo-(N-3-pyridyl)nonamide
To a solution of chromium oxide (3.4 g, 35 mmol) in water
(5 mL) at 0 ꢂC was added dropwise concentrated sulfuric acid
(3 mL) and water (10 mL). This solution was slowly added at ꢁ5 ꢂC

8542
N. Veling et al. / Tetrahedron 64 (2008) 8535–8542
resulting in a white solid (30%). 1H NMR (400.15 MHz, CDCl3/CD3CN
3. Huang, Y. W.; Doyle, M. L.; Ackers, G. K. Biophys. J. 1996, 71, 2094–2105.
4. Johnson, M. L. Methods Enzymol. 2000, 323, 124–155.
5. Harman, J. G. Biochim. Biophys. Acta 2001, 1547, 1–17.
6. Ercolani, G. J. Am. Chem. Soc. 2003, 125, 16097–16103.
7. Ballester, P.; Costa, A.; Deya, P. M.; Frontera, A.; Gomila, R. M.; Oliva, A. I.;
Sanders, J. K. M.; Hunter, C. A. J. Org. Chem. 2005, 70, 6616–6622; Hunter, C. A.;
Meah, M. N.; Sanders, J. K. M. J. Am. Chem. Soc. 1990, 112, 5773–5780; Anderson,
H. L.; Hunter, C. A.; Meah, M. N.; Sanders, J. K. M. J. Am. Chem. Soc. 1990, 112,
5780–5789.
1:1 (v/v)):
d
¼8.86 (m, 8H), 8.38 (m, 8H), 4.59 (t, 3J¼7.6 Hz, 4H), 4.41
(s, 6H), 2.10 (m, 4H), 1.35 (m, 24H); 13C{1H} NMR (50 MHz, CD3CN):
d
¼149.9, 149.6, 146.5, 145.5, 127.2, 126.8, 62.1, 48.6, 31.0, 29.5, 29.4,
29.3, 29.1, 28.7, 25.6; IR (solid) 836, 555 cmꢁ1; HRESIMS m/z
1001.3276, calcd for C38H54F18N4P3 [MꢁPF6]þ 1001.3273.
8. Ubukata, T.; Seki, T.; Ichimura, K. Adv. Mater. 2000, 12, 1675–1678.
9. Fernandez, G.; Perez, E. M.; Sanchez, L.; Martin, N. Angew. Chem., Int. Ed. 2008,
47, 1094–1097.
Acknowledgements
This research was supported by an NWO Veni grant to J.A.A.W.E.,
an ARC-DP0666325 Australian Research Fellowship to P.T., a KNAW
professorship to R.J.M.N. and an NWO Vidi grant to A.E.R.
10. Monod, J.; Wyman, J.; Changeux, J. P. J. Mol. Biol. 1965, 12, 88–118.
11. Alderton, W. K.; Cooper, C. E.; Knowles, R. G. Biochem. J. 2001, 357, 593–615.
12. Hynes, R. O. Cell 2002, 110, 673–687.
13. Takeuchi, M.; Ikeda, M.; Sugasaki, A.; Shinkai, S. Acc. Chem. Res. 2001, 34,
865–873.
14. Elemans, J. A. A. W.; Bijsterveld, E. J. A.; Rowan, A. E.; Nolte, R. J. M. Eur. J. Org.
Chem. 2007, 751–757.
Supplementary data
15. Thordarson, P.; Bijsterveld, E. J. A.; Elemans, J. A. A. W.; Kasak, P.; Nolte, R. J. M.;
Rowan, A. E. J. Am. Chem. Soc. 2003, 125, 1186–1187.
16. Thordarson, P.; Bijsterveld, E. J. A.; Rowan, A. E.; Nolte, R. J. M. Nature 2003, 424,
915–918.
Supplementary data associated with this article can be found in
the online version, at doi:10.1016/j.tet.2008.05.139.
17. Thordarson, P.; Coumans, R. G. E.; Elemans, J. A. A. W.; Thomassen, P. J.; Visser, J.;
Rowan, A. E.; Nolte, R. J. M. Angew. Chem., Int. Ed. 2004, 43, 4755–4759.
18. Elemans, J. A. A. W.; Claase, M. B.; Aarts, P. P. M.; Rowan, A. E.; Schenning,
A. P. H. J.; Nolte, R. J. M. J. Org. Chem. 1999, 64, 7009–7016.
19. Park, Y. S.; Lee, E. J.; Chun, Y. S.; Yoon, Y. D.; Yoon, K. B. J. Am. Chem. Soc. 2002,
124, 7123–7135.
References and notes
1. Klug, A. Angew. Chem., Int. Ed. Engl. 1983, 22, 565–582.
2. Ackers, G. K.; Doyle, M. L.; Myers, D.; Daugherty, M. A. Science 1992, 255,
54–63.