A supramolecular system for quantifying aromatic stacking interactions

Article abstract of DOI:10.1002/1521-3765(20011119)7:22<4863::AID-CHEM4863>3.0.CO;2-3

A supramolecular complex for investigating the thermodynamic properties of intermolecular aromatic stacking interactions has been developed. The conformation of the complex is locked in a single well-defined conformation by an array of H-bonding interacti

Full text of DOI:10.1002/1521-3765(20011119)7:22<4863::AID-CHEM4863>3.0.CO;2-3

FULL PAPER
A Supramolecular System for Quantifying Aromatic Stacking Interactions
Harry Adams,[a] Christopher A. Hunter,*[a] Kevin R. Lawson,[b] Julie Perkins,[a]
Sharon E. Spey,[a] Christopher J. Urch,[b] and John M. Sanderson[a]
Abstract: A supramolecular complex
for investigating the thermodynamic
properties of intermolecular aromatic
stacking interactions has been devel-
oped. The conformation of the complex
is locked in a single well-defined con-
formation by an array of H-bonding
interactions that force two aromatic
rings on one end of the complex into a
stacked geometry. Chemical double-mu-
tant cycles have been used to measure
phenyl ± aniline interaction
(
0.4 Æ
interactions. The nitropyrrole subunits
used to control the conformation of
these complexes lead to problems of
aggregation and multiple conformation-
al equilibria. The implications for the
thermodynamic analysis are examined
in detail, and the double-mutant-cycle
approach is found to be remarkably
robust with respect to such effects, since
systematic errors in individual experi-
ments are removed in a pair-wise fashion
when the cycle is constructed.
1
0.9 kJmol ) in this system. Although
the interactions are very weak, the
pentafluorophenyl interaction is attrac-
tive, whereas the anthracene interaction
is repulsive; this is consistent with the
dominance of p-electron electrostatic
Keywords: aromatic stacking
´
double-mutant cycle hydrogen
´
bonds ´ molecular recognition ´ pi
interactions
an
anthracene ± aniline
interaction
1
(‡0.6 Æ 0.8 kJmol ) and a pentafluoro-
have been various attempts to quantify the interactions by
using synthetic supramolecular systems.[33±38] Experimental
data on the thermodynamics of noncovalent interactions are
essential if we are to develop reliable computational methods
for predicting the behaviour of complex molecular sys-
tems.[39±41] We have developed an experimental method for
measuring intermolecular interactions using chemical double-
mutant cycles, and this has been successfully applied to edge-
to-face aromatic interactions.[42, 43] We present here the design
and realisation of a supramolecular system for measuring
aromatic stacking interactions.
Introduction
The process of molecular recognition is governed by non-
covalent interactions, and the development of simple model
systems to investigate the properties of these interactions has
been a major thrust of research in supramolecular chemistry.
One important class of intermolecular interactions is aromatic
stacking.[1] The classic example is base stacking in DNA,[2±4]
but stacking interactions between aromatic side chains are
also found in proteins and play an important role in
determining the fold of a protein and the substrate recog-
nition properties.[5±7] Supramolecular chemists have exploited
aromatic stacking interactions for substrate recognition by
synthetic receptors and in template-directed synthesis of
complex molecules.[9±20] The packing of aromatic molecules in
the solid state, and hence the physical properties of these
materials, is to some extent determined by stacking inter-
actions.[21±24] Theoretical calculations have yielded some in-
sight into the nature of aromatic interactions,[25±32] and there
Results and Discussion
Design and Synthesis: The starting point for the development
of a system to measure stacking interactions was the 1 ´ 2
complex previously used to measure edge-to-face interactions
(Scheme 1).[43±46] To introduce stacking interactions into this
system, the benzoyl rings must be rotated through 908, so that
they lie parallel to the aniline ring as shown in Scheme 2.
Substituents ortho to the carbonyl group force such a
conformation, so anthracene was chosen as it has such a large
aromatic surface that the aniline ring will be forced to sit over
it, even if the stacking interaction is repulsive.
[a] Prof. C. A. Hunter, H. Adams, J. Perkins, S. E. Spey, Dr. J. M. Sanderson
Centre for Chemical Biology
Krebs Institute for Biomolecular Science
Department of Chemistry, University of Sheffield
Sheffield S3 7HF (UK)
Fax : (‡44)114-273-8673
E-mail: c.hunter@sheffield.ac.uk
We have previously used X-ray crystal structures of simple
model compounds to probe the geometry of the terminal
aromatic interaction in edge-to-face zipper complexes. For
[b] Dr. K. R. Lawson, Dr. C. J. Urch
Zeneca Agrochemicals, Jealottꢀs Hill Research Station
Bracknell, Berkshire, RG42 6ET (UK)
Chem. Eur. J. 2001, 7, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4863 $ 17.50+.50/0
4863

C. A. Hunter et al.
FULL PAPER
chains as expected, but there is
a
mixture of head-to-tail
and head-to-head orientations
(Scheme 4a). The aromatic
rings are all approximately par-
allel, but the only true stacking
interaction is an anthracene ±
anthracene interaction in the
head-to-head pairs. The steric
bulk of the isopropyl substitu-
ents prevents close parallel
stacking of the anthracene with
the aniline rings. Replacement
of the isopropyl substituents
with
methyl
groups
(9,
Scheme 3) removes the steric
bulk, but in the crystal structure
of this compound, the mole-
cules all sit in a head-to-head
arrangement with only anthra-
cene ± anthracene and aniline ±
aniline stacking interactions
(Scheme 4b).
Scheme 1. a) Zipper complex 1 ´ 2 used to measure the terminal edge-to-face aromatic interaction in solution.
b) Model compound 3 used to probe the geometry of this interaction in the solid state. Three molecules from the
X-ray crystal structure of 3 are shown. The a and b interaction geometries are labelled.
In an attempt to favour a
head-to-tail arrangement, com-
pound 10 was synthesised
(Scheme 3), because the litera-
ture suggests that perfluorinat-
ed rings stack exceptionally
well
with
hydrocarbon
head-to-tail
rings.[23, 24, 47, 48]
A
arrangement was indeed ob-
tained, but the crystal structure
showed three different aromat-
ic interactions (Scheme 4c). By
analogy with the structure of 3
in Scheme 1b, these interac-
tions are labelled a, b1 and b2.
The a interaction is the desired
pentafluorophenyl ± anthracene
stacking interaction and in-
volves a H-bond between the
Scheme 2. Candidate complexes for measuring an aromatic stacking interaction.
example, 3 forms a H-bonded polymer with head-to-tail
packing; this gives edge-to-face aromatic interactions
throughout the structure (Scheme 1b). Closer examination
of this structure reveals that there are actually two different
aromatic interactions present: one in which the benzoyl
carbonyl oxygen is a H-bond acceptor (a) and one in which
the benzoyl amide is a H-bond donor (b). This difference is
also manifest in different interactions on the two ends of the
zipper complex in Scheme 1a.[43] Although the difference in
the geometry of the edge-to-face interactions is very subtle, it
becomes much more significant for stacking interactions.
X-ray crystal structures of model compounds were used to
probe the feasibility of the proposed aromatic stacking
interactions in the complexes in Scheme 2. The coupling
product 8 of 9-anthranoyl chloride and 2,6-diisopropylaniline
was synthesised (Scheme 3) and crystallised. In the X-ray
crystal structure, the molecules are arranged in H-bonded
aniline amide NH and the anthracene carbonyl oxygen. In the
two b interactions, the anthracene amide NH is the H-bond
donor, and the aromatic rings are arranged in an edge-to-face
geometry. In the head-to-tail packing arrangement, the b
H-bond geometry does not allow aromatic stacking to take
place: the aromatic rings must either be parallel and separated
by about 5 Š, or they can be in close contact but not parallel.
In contrast, the a H-bond geometry is perfectly adapted
for aromatic stacking. This has important implications for the
way in which these interactions are translated into the
complexes in Scheme 2. Only the a H-bond geometry will
result in a stacking interaction, and therefore we require
unsymmetrical complexes, in which it is possible to control
and distinguish the interactions on the a and b ends of the
complex.
We have previously found that replacing the benzoyl group
in the 1 ´ 2 complex with pyrrole gives rise to two bifurcated
4864
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4864 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 22

Stacking Interactions
4863±4878
1. (COCl)2
2.
NH2
CO2H
H
O
N
8
28%
1. (COCl)2
2.
NH2
CO2H
H
O
N
9
56%
1. (COCl)2
2.
NH2
F
F
F
CO2H
F
F
H
O
F
N
F
40%
F
F
10
F
Scheme 3. Synthesis of the variously substituted compounds 8, 9 and 10 in
an attempt to regulate stacking in H-bonded chains.
H-bonds to the adjacent carbonyl oxygen, and this approach
appears ideal for locking the conformation of the complex at
the b end.[49] The principle is illustrated for the 11 ´ 7 complex
in Scheme 5. When the anthracene ± aniline interaction of
interest is in the a geometry, there are two bifurcated H-bonds
between the pyrrole and amide NH groups and the adjacent
carbonyl oxygen; when the anthracene ± aniline interaction is
in the b geometry, there is only one H-bond between the
aniline amide NH and the pyrrole carbonyl oxygen. This
should significantly favour the conformation with the a
stacking interaction over the b geometry in which the relevant
aromatic rings are remote. We decided to bias the system
further by incorporating a nitro group in the 4-position of the
pyrrole to increase the H-bond donor strength of the pyrrole
and amide NHs and stabilise the a conformation. Compounds
11 and 7 were synthesised according to Schemes 6 and 7,
below. 4-Nitropyrrole-2-carbonyl chloride 15 was prepared
according to a procedure described by Morgan and Morrey.[50]
Mucobromic acid 12 was treated with sodium nitrite to form
sodium malondialdehyde 13[51] (Scheme 6). Condensation of
13 with glycine ethyl ester resulted in ethyl 4-nitropyrrole-2-
carboxylate (14) which was subsequently hydrolysed to the
Scheme 4. a) Model compound 8 and the arrangement of three molecules
in the X-ray crystal structure. Both head-to-tail and head-to-head arrange-
ments of the aromatic substituents are observed. b) Model compound 9 and
the arrangement of three molecules in the X-ray crystal structure. Only
head-to-head interactions are found for the aromatic substituents. c) Model
compound 10 with the arrangement of three molecules in the X-ray crystal
structure. The molecules pack in a head-to-tail arrangement, giving the
desired anthracene ± aniline stacking interaction in the a geometry and
edge-to-face interactions in the two b geometries.
corresponding acid. This was converted to the acid chloride 15
by treatment with oxalyl chloride. The synthesis of 16 has
been previously described.[52] This was coupled with 0.2 equiv-
alents of anthranoyl chloride to give the monosubstituted
product 17, which was coupled with 4-nitro-pyrrole-2-carbonyl
chloride 15 to give 11.
The isophthaloyl derivative 7 was synthesised by treating
isophthaloyl dichloride with a large excess of pentafluoroani-
line over 6 days (Scheme 7). Compound 7 was scarcely soluble
in CDCl3. Compound 11 has a solubility of about 0.5mm in
Scheme 5. Two possible conformations for complex 11 ´ 7. The a geometry anthracene ± aniline stacking interaction is favoured by the two bifurcated
H-bonds formed by the nitropyrrole amide.
Chem. Eur. J. 2001, 7, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4865 $ 17.50+.50/0
4865

C. A. Hunter et al.
FULL PAPER
A new target complex 18 ´ 6 was therefore designed
(Scheme 8). A solubilising group 21 was added to the
cyclohexyl ring of 11 and the pentafluoroaniline in 7 was
replaced by 2,6-dimethylaniline. The X-ray structure ana-
logues suggest that the aniline ortho methyl groups will keep
H2N
NH2
Br
Br
CHO
12
CO2H
16
9-anthranoyl chloride
NEt3
55%
H3C(H2C)13
O
NaNO2
44%
O(CH2)13CH3
O(CH2)13CH3
O
N
Na+
–
O2N
CHO
13
CHO
b4
n2
18
b3
nh
b2
b1
H
a1
a2
O
H
N
N
a3
a4
a5
H2N
p3
glycine ethyl ester
NaOH
N
n1
t
O
H
O
43%
17
O2N
N
d
H
p2
p1
H
O
N
s
O2N
O2N
N
1. KOH
2. (COCl)2
H
O
aa
O
O
me
pyridine
80%
76%
N
H
N
H
6
O
Cl
Scheme 8. Redesigned complex 18 ´ 6 for measuring aromatic stacking
interactions. The proton labelling scheme is shown.
14
15
O
the rings in a stacked conformation, but the problems of the
steric bulk associated with the isopropyl groups are almost
eliminated. Schemes 9 and 10 show the relevant synthetic
routes. Compound 21 was prepared from propyl gallate 19
according to Scheme 9. Compound 22 was prepared by
heating 2,6-dimethylaniline and N-acetyl-4-piperidone in
concentrated HCl under reflux. Protection of the aliphatic
amine of 22 with benzylchloroformate in methanol gave 23,
which was then sequentially coupled with the anthracene and
nitropyrrole acid chlorides. The protecting group was re-
moved by using trimethylsilyl iodide according to a procedure
described by Stammer.[53] The intermediate piperidine was
extremely insoluble and was used without purification in a
carbodiimide coupling with 21
H
N
H
N
O2N
N
O
H
11
Scheme 6. Convergent synthesis of compound 11.
CDCl3, and even then precipitates after being in solution for
1 ± 2 h. Thus complex 11 ´ 7 proved unsuitable for NMR
titration experiments, and more soluble analogues were
required.
to give the desired product 18.
NH2
The isophthaloyl derivative 6
F
F
F
F
H
N
F
has previously been described,
but the synthesis is outlined in
Scheme 7.[46]
O
F
F
F
F
7
F
F
N
F
O
F
O
Cl
H
F
F
80%
Cl
O
1
Binding studies: H NMR dilu-
NH2
tion experiments indicate that 6
H
N
O
has
a
dimerisation constant
6
1
Kd ꢀ 1m , but since 6 was used
N
O
H
as the host at millimolar con-
1
centrations in the H NMR ti-
tration experiments, the effects
of dimerisation are negligi-
ble.[46] However, 18 was used
as the guest at much higher
concentrations, and 1H NMR
dilution experiments with 18
showed very large changes in
chemical shift. The dilution da-
ta were fitted to both a dimer-
isation model and an aggrega-
H
N
MeNH2
90%
O
27
Me
N
N
O
O
Me
H
H
H
N
NH2
O
28
Scheme 7. Synthesis of compounds 6, 7, 27 and 28.
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4866 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 22
4866

Stacking Interactions
4863±4878
HO
HO
H3C(H2C)13
O
O
H3C(H2C)13
O
O
1-bromotetradecane
K2CO3
O
O
O
O
O
KOH
H3C(H2C)13
O
H3C(H2C)13
O
56% overall
OH
HO
H3C(H2C)13
H3C(H2C)13
19
Scheme 9. Synthesis of compound 21 from 19.
21
Scheme 10. Synthesis of compound 18.
Table 1. Dimerisation constants Kd [m 1] and limiting dimerisation-induced changes in chemical shift (ppm) for compounds 18, 26 and 33 in CDCl3 at
300 K.[a]
Signal
Compound
Kd
p1
p2
1.1
0.7
1.1
p3
n1
n2
b1
b2
b3
b4
0.2
0.2
0.2
a1
a2
a3
a4
a5
m1
18
26
33
13 Æ 2
20 Æ 2
11 Æ 2
3.0
2.9
6.5
0.1
nd[b]
nd[b]
2.4
2.0
2.7
0.5
1.1
1.7
0.4
nd[b]
0.2
0.3
0.1
0.1
0.1
0.1
0.1
0.3
±
±
0.2
±
±
0.2
±
±
0.2
±
±
0.3
±
±
±
0.1
±
[a] See Schemes 8, 11 and 12 for proton labelling schemes. [b] Not determined due to signal overlap throughout the experiment. There are no significant
changes in chemical shift for the signals due to the protons on the piperidine ring and solubilising group. Dilution experiments were repeated at least twice
and Kd is the weighted mean based on the observed change in chemical shift for all the signals monitored. The error is twice the standard error.
tion model, but there is little difference in the results because
a)
b)
O2N
O2N
O
N
H
at the concentrations of interest, the extent of interaction is
small. The maximum percentage bound is 30%; this means
that even if aggregation is the mode of association, the
N
N
N
H
H
H
O
O
H
N
O
predominant species is dimer. From the dimerisation model,
N
1
Kd ˆ 13 Æ 2m
for 18. The limiting dimerisation-induced
N
H
H
NO2
changes in chemical shift (Dd) are shown in Table 1. The
magnitudes of the Dd values imply that the main site of
dimerisation is the nitropyrrole moiety. The Dd values suggest
that a bifurcated hydrogen bond is present, because there are
large downfield shifts for the pyrrole and pyrrole amide NHs
(Scheme 11a). In contrast, the X-ray crystal structure of the
precursor 25 shows a different conformation for the nitro-
Scheme 11. a) The complexation-induced changes in 1H NMR chemical
shift indicate that the bifurcated H-bond motif is the principle mode of
dimerisation for 18 in CDCl3 solution. This stabilises the conformation in
which the pyrrole NH is anti to the amide carbonyl oxygen. b) The
H-bonding motif observed in the X-ray crystal structure of 25. The
conformation in which the pyrrole NH is syn to the carbonyl oxygen is
stabilised by intramolecular H-bonding interactions.
Chem. Eur. J. 2001, 7, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4867 $ 17.50+.50/0
4867

C. A. Hunter et al.
FULL PAPER
pyrrole (Figure 1): the pyrrole NH is oriented cis to the amide
carbonyl oxygen, and dimerisation occurs through two
intermolecular pyrrole NH-amide H-bonds (Scheme 11b
and Figure 1). This is almost certainly the lowest energy
conformation of the nitropyrrole amide moiety in the mono-
mer, because of intramolecular interactions between the
NO2
H
N
H
O
N
N
O
N
O2N
H
H
syn-trans
syn-cis
O2N
O2N
N
O
N
H
O
N
N
H
H
H
anti-cis
anti-trans
Scheme 12. Four possible conformations for the nitropyrrole amide
moiety. The amide group can be cis or trans, and the pyrrole NH can be
syn or anti with respect to the carbonyl oxygen.
Figure 1. Three molecules from the X-ray crystal structure of 25. The
intermolecular stacking interaction between the anthracene and aniline is
highlighted. The dashed lines denote intramolecular H-bonds, and the
dotted lines denote intermolecular H-bonds.
and the arrangement of the pyrrole NH relative to the amide
carbonyl group (syn or anti). The syn-trans conformation is
probably the most stable and is the conformation found in the
X-ray crystal structure of 25 (Figure 1). The anti-trans
conformation appears to be the predominant species involved
in dimerisation in solution. The syn-cis conformation is the
pyrrole NH and amide carbonyl oxygen. In the solution
dimer, however, it appears that the high energy conformation,
that is when the pyrrole NH is anti relative to the amide
carbonyl oxygen, is stabilised by the bifurcated H-bonding
motif in Scheme 11a relative to the syn conformation. It is
encouraging to note that in the crystal structure of 25 there is
also an anthracene ± aniline stacking interaction in precisely
the arrangement we have designed into the 18 ´ 6 complex
(Figure 1).
The 1H NMR spectrum of 18 in deuterochloroform solution
is further complicated by the fact that we can detect
approximately 10% of a second conformer that is in slow
exchange with the major species. The second conformer
differs in the chemical shifts of the signals due to p1 and p3,
which appear at d ˆ 10.5 and 5.3, respectively, as compared
with d ˆ 10.7 and 7.3 in the major conformer. These signals
were readily assigned based on chemical exchange cross-
peaks observed in a ROESY experiment. The rest of the
1H NMR spectrum does not appear to be affected by the
conformational change, and no other exchange cross-peaks
were observed. The minor conformer disappears in
[D6]DMSO, either due to an increase in the rate of exchange
or a decrease in the stability of the minor conformer. The
large upfield shift observed for p3 suggests that it sits over the
face of an aromatic ring in the minor conformer, and the syn-
cis conformation shown in Scheme 12 is the most likely
explanation. The edge-to-face interaction involving the very
polar p3 proton presumably compensates for the energy
required to isomerise the amide bond.
1
second minor conformer observed in the H NMR spectrum
(molecular-mechanics calculations suggest that this is actually
the lowest energy conformer). Exchange between syn and anti
is fast on the NMR timescale, whereas exchange between cis
and trans is slow, so it is not possible to rule out the presence
of the anti-cis conformer, but the chemical shifts of the minor
conformer suggest the syn-cis conformer is significantly more
populated (p3 is highly shielded and p1 is not). In the dilution
experiment, the ratio of the cis and trans conformers does not
change significantly, but the extent of dimerisation is probably
not enough to perturb the equilibrium. The signals due to p1
and p3 were not used in the determination of Kd for 18, so the
value measured in the dilution experiment is a weighted
average for all the species present.
The binding constant of the 6 ´ 18 complex was determined
by 1H NMR titration experiments. Compound 18 was titrated
into 6, and the host signals were fitted to a 1:1 binding
isotherm (Ka ˆ 85 Æ 7m 1). The data were also fitted to a
model that allowed for dimerisation of the guest, and this gave
1
a binding constant Ka ˆ 63 Æ 6m . Thus dimerisation of the
guest makes a detectable difference in the titration results,
and the significance of this in the double-mutant-cycle
analysis will be discussed later. The limiting complexation-
induced changes in chemical shift for formation of the 6 ´ 18
complex are shown in Figure 2 and Table 3. There are large
downfield shifts of the signals due to the pyrrole NH and the
pyrrole amide (p1 and n1) as well as for the signal due to the
amide of 6 (nh); these are indicative of H-bonding inter-
actions. The change in chemical shift for the signal due to the
anthracene amide (n2) is much smaller; this suggests that it is
not H-bonded. The large upfield shift for the signal due to p2
Thus there are four possible conformations for the pyrrole
amide moiety, which are illustrated in Scheme 12. The con-
formers differ in the geometry of the amide bond (cis or trans)
4868
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4868 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 22

Stacking Interactions
4863±4878
conformer is approximately 10%, and the intensity of these
signals decreases during the titration until they are barely
detectable at the end. The results are summarised in
Scheme 13.
R
N
0.1
0.2
0.2
0.0
18
O
0.4
0.1
H
N
H
N
-0.1
-0.1
-0.1
-0.3
O2N
N
-1.5
2.7
-0.3
O
-1.0
NO2
H
2.3
H
N
H
-0.2
K
free = 0.1
O
O
N
N
H
1.3
O
N
N
N
0.2
O2N
H
H
H
6
O
-0.3
-0.3
free trans
free cis
Figure 2. Limiting 1H NMR complexation-induced changes in chemical
shift (ppm) and intermolecular NOEs observed for complex 6 ´ 18 in CDCl3
at 300 K.
–1
–1
Kcis = 12 M
Ktrans = 70 M
of the pyrrole indicates that it sits over the face of an aromatic
ring, so the complex clearly adopts the conformation shown in
Figure 2 with 18 in the anti-trans conformation. The large
upfield shift observed for the isophthaloyl triplet (t) and
doublet (d) indicate that these protons lie over the face of
another aromatic ring. The small downfield shifts observed for
the signals due to the diarylmethane subunit (b1 ± b4) indicate
that the face of this ring interacts with the edge of another and
confirm that the expected edge-to-face interactions are
present in the core of the complex.[54] Finally, the upfield
shifts observed for all of the signals due to the anthracene of
18 and the aniline groups of 6 indicate that they are involved
in a face-to-face stacking interaction. Five intermolecular
NOEs were observed (Figure 2 and Table 2). The two b ± d
NOEs confirm that the isophthaloyl group is docked into the
NO2
H
N
Kbound = 0.02
O2N
O
N
O
N
N
H
H
H
O
O
N
N
H
H
bound trans
bound cis
Scheme 13. Summary of conformational equilibria for 18 and the complex
28 ´ 18. Kfree is the cis-trans equilibrium constant for uncomplexed 18. Ktrans
is the binding constant for the trans conformer of 18 with 28. Kcis is the
binding constant for the cis conformer of 18 with 28. Kbound is the
equilibrium constant for the cis-trans isomerism in the complex 18 ´ 28.
Table 2. Intermolecular NOEs observed in two-dimensional ROESY
experiments.[a]
Integration of the two sets of signals in the 1H NMR
spectrum of uncomplexed 18 gives the cis-trans equilibrium
constant Kfree ˆ 0.1 (the cis/trans ratio) and, using this value
with the two binding constants, we can calculate the cis-trans
equilibrium constant in the 18 ´ 28 complex, Kbound ˆ 0.02 (cis/
trans ratio). Thus in the complex, the trans conformer is fifty
times more stable than the cis conformer, and, because
equilibration is fast on the timescale of the titration, the cis
conformer cannot be observed at the end of the titration when
saturation is reached. When 28 is the host, what we actually
measure is the weighted average of the Ktrans and Kcis
equilibrium constants. The cis-trans ratio does not change in
this experiment because 18 is always present in a large excess.
Thus, using the results in Scheme 13, we can predict an
expected value for the binding constant in the reverse
titration:
Isophthaloyl
compound
Bisaniline compound
18
28
33
6
p2-me, b1-d, b4-d,
a1-me, a2-me
b1-d
b1-d, b4-d
28
b1-d, b4-d, a1-nh
b1-d, b4-d
b1-d, b4-d
[a] Experiments were carried out on 1:1 mixtures of the two components at
the maximum concentration possible (3 ± 28mm). See Schemes 7, 8, 15 and
19 for the proton labelling schemes.
bisaniline pocket in the core of the complex. The remaining
NOEs confirm that the nitropyrrole and anthracene both lie
over the aniline rings of 6.
In order to determine the effect of the minor conformer on
the titration results, two titrations were carried out with 18
and the more soluble isophthaloyl compound 28. In one
titration 28 was the host and in the other it was the guest. With
1
Ka ˆ (70  0.9) ‡ (12  0.1) ˆ 63m
1
28 as the host, Ka ˆ 49 Æ 8m . With 18 as the host, the signals
for both the major and minor conformer could be followed
throughout the titration and separately fitted to a binding
isotherm that allowed for dimerisation of the host, but ignored
the fact that the proportions of cis and trans conformers
change during the titration. However, the guest was always
present in a very large excess, and so this approximation does
This explains why the association constant obtained from
the reverse titration is lower. The predicted value is based on
only one signal and, given the size of the errors involved, it is
in reasonable agreement with the value measured experi-
1
mentally, Ka ˆ 49 Æ 8m .
not affect the analysis. For the major trans conformer Ktrans
70m and for the minor cis conformer Kcis ˆ 12m . At the
beginning of the experiment, the proportion of the cis
ˆ
Double-mutant cycles: Despite the conformational complex-
ity of uncomplexed 18, these experiments show that this
molecule is suitable for use in a double-mutant cycle to
1
1
Chem. Eur. J. 2001, 7, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4869 $ 17.50+.50/0
4869

C. A. Hunter et al.
FULL PAPER
measure the terminal aromatic
stacking interaction. The com-
plex adopts a single well-de-
fined geometry, and the nitro-
pyrrole provides a good b con-
formational lock forcing the a
conformation required for aro-
matic stacking interactions at
the other end of the complex.
The nitropyrrole component is
present in all four complexes in
the double-mutant cycle, and so
the thermodynamic consequen-
ces of the conformational equi-
libria in the uncomplexed state
discussed above will cancel out
in the analysis. In principle, the
magnitude of the terminal aro-
matic interaction in complex A
(6 ´ 18) in Scheme 14 can be
estimated by chemical muta-
tions which remove it: by com-
paring the stability of complex
Awith complexes B or C, which
do not have the interaction
R
N
R
N
B
A
GA - GB
H
N
H
O
O
N
N
N
Me
O
O
H
H
O2N
O2N
N
N
H
H
H
N
H
N
O
O
N
N
H
H
O
O
GB - GD
G
A - GC
G = GA - GB - GC + GD
R
N
R
N
C
D
H
H
N
O
O
N
N
H
N
Me
O
O
H
O2N
O2N
N
H
N
H
H
N
H
N
O
O
GC - GD
N
N
H
H
O
O
Scheme 14. Chemical double-mutant cycle to measure the anthracene ± aniline stacking interaction in
complex A (R ˆ solubilising group, see Scheme 9).
present. However, this assumes that the aromatic rings in
question make no other intermolecular interactions in com-
plex A and that the strength of the hydrogen bonds remains
constant when the aromatic to alkyl mutation is made. These
effects are quantified by using the double mutant, complex D.
For example, if we compare complexes C and D, the differ-
ence DGC DGD is a direct measure of the sum of the change
in hydrogen bond strength and secondary interactions made
by the anthracene group between complexes A and C. Thus
the free energy difference of the two horizontal mutations in
Scheme 14 allows us to dissect the interaction of interest from
the complicated array of weak interactions present in complex
A.
The N,N'-dimethylisophthaloyl diamide 27 was synthesised
according to Scheme 7 but was unsuitable for these experi-
ments due to its low solubility, so the dihexyl derivative 28 was
used instead. The synthesis of 28 has been described
previously and is also outlined in Scheme 7.[46] The synthesis
of 26 was carried out according to Scheme 15. Compound 23
was coupled with 4-nitropyrrole-2-carbonyl chloride (15). The
O
O
O
O
N
N
15, pyridine
O
75%
H2N
NH2
N
H
NH2
O2N
N
H
23
29
glacial acetic acid
EDC
HOBt
76%
OR
s1
b3
O
O
OR
OR
O
N
N
1. TMSI
2. 21 / EDC / HOBt / NEt3
b2
b4
b1
O
O
N
p3
n2
H
H
60%
N
N
N
O2N
O2N
N
H
N
H
O
Me
m1
n1
O
Me
H
H
p2
p1
26
Scheme 15. Synthesis of compound 26 from 23.
30
4870
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4870 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 22

Stacking Interactions
4863±4878
1
acetyl group was added by a carbodiimide coupling reaction.
The product 30 was treated with TMSI to remove the
protecting group and then reacted with 21 by a carbodiimide
coupling to yield 26 in reasonable yield.
associated with a second signal due to n2 in the H NMR
spectrum. The n2 signal also collapses to one signal in DMSO,
so we attribute the third conformer to the cis form of the
acetamide. This conformational equilibrium does not affect
any other signals in the spectrum, although it is surprising that
it has an effect on p1. The conformer with both amides cis
cannot be detected, but it must only be present at a level of
about 2%.
In order to determine the effect of these minor conformers,
two titrations were carried out with 26 and 28, one in which 28
was the host and one in which it was the guest. With 28 as the
The dimerisation constant for 26 was determined by a
1H NMR dilution experiment (Kd ˆ 20 Æ 2m 1). The limiting
dimerisation-induced changes in chemical shift are shown in
Table 1. The results are very similar to those obtained for 18.
The largest changes in chemical shift are associated with the
pyrrole NH (p1) and the pyrrole amide (n1); this suggests the
same mode of dimerisation in solution: the bifurcated H-bond
motif shown in Scheme 11a. An X-ray crystal structure of the
cyclohexyl analogue 31 (Figure 3), which was synthesised
according to Scheme 16, shows that dimerisation in the solid
state again occurs through a doubly H-bonded dimer of the
pyrrole (cf Scheme 11b and Figure 1).
1
host, Ka ˆ 46 Æ 6m . With 26 as the host, the signals for all
three conformers could be followed and separately fitted to a
binding isotherm that allowed for dimerisation of the host,
again ignoring the changes in the populations of the three
conformers, since the guest is present in a large excess. For the
1
major trans conformer p1 Ktrans ˆ 64m , for the pyrrole cis
1
conformer Kcis-Pyr ˆ 14m and for the acetamide cis con-
1
former Kcis-Me ˆ 6m . At the beginning of the experiment, the
proportion of the pyrrole cis conformer is 10% and the
proportion of acetamide cis conformer is 20%, but both
decrease during the titration and cannot be detected at the
end of the experiment. The results are summarised in
Scheme 17.
Figure 3. Three molecules from the X-ray crystal structure of 31. The
dashed lines denote intramolecular H-bonds, and the dotted lines denote
intermolecular H-bonds.
Scheme 17. Summary of conformational equilibria for 26 and the complex
28 ´ 26. KPf yr is the pyrrole amide cis-trans equilibrium constant. KMf e is the
acetamide cis-trans equilibrium constant. Ktrans is the binding constant for
the all trans conformer of 26 with 28. Kcis-Pyr is the binding constant for the
pyrrole amide cis conformer of 26 with 28. Kcis-Me is the binding constant for
the acetamide cis conformer of 26 with 28. KPbyr is the equilibrium constant
for the pyrrole amide cis-trans equilibrium in the complex 26 ´ 28. KMb e is the
equilibrium constant for the acetamide cis-trans equilibrium in the complex
26 ´ 28.
15, pyridine
O
N
H
NH2
H2N
NH2
O2N
65%
N
H
32
16
glacial acetic acid
EDC
59%
HOBt
From the pyrrole and acetamide cis-trans equilibrium
constants for uncomplexed 26 and the three binding constants,
we can calculate the cis-trans equilibrium constants in the 28 ´
26 complex (KPbyr ˆ 0.02 and KMb e ˆ 0.03). This means that the
trans conformer is substantially stabilised in the complex, and
the results for the pyrrole end of the complex are identical to
those obtained for compound 18. When 28 is the host, we
actually measure the weighted average of Ktrans, Kcis-Pyr and
Kcis-Me equilibrium constants. Thus, using the results in
Scheme 17, we can predict the value of the binding constant
in the reverse titration:
O
N
H
N
O2N
N
H
O
Me
H
31
Scheme 16. Synthesis of compound 31 from 16.
For 26 in CDCl3, there are three different signals due to p1
in the 1H NMR spectrum, these collapse to one signal in
DMSO. ROESY experiments show that all three p1 signals
are in chemical exchange, so it appears that there are three
conformers of 26 in CDCl3. p3 shows a second minor signal at
d ˆ 5.5, which we attribute to the syn-cis geometry of the
pyrrole subunit (Scheme 12), and the population of this
conformer is 10% as observed previously for 18. The new
third conformer is more highly populated (20%) and is
1
Ka ˆ (6  0.2) ‡ (64  0.7) ‡ (14  0.1) ˆ 47m
This value is based on only one signal from each conformer
but is in excellent agreement with the value measured
1
experimentally, Ka ˆ 46 Æ 6m .
Chem. Eur. J. 2001, 7, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4871 $ 17.50+.50/0
4871

C. A. Hunter et al.
FULL PAPER
The implications of these equilibria for the double-mutant
cycle must now be considered. In all the double-mutant cycle
titrations, the bisaniline compound was used as the guest.
Since several signals can be followed and they all behave in
the same way, this avoids the complications of following
signals from different conformations and allows a more
accurate determination of the weighted-average binding
constant for all of the conformers. The binding constants
measured in this way are:
cycles, and we can use the observed binding constants, which
are the weighted average of all species present. H NMR
1
titrations were carried out for all four complexes in the
double-mutant cycle shown in Scheme 14, and the data were
fitted to a simple 1:1 binding isotherm as well as a model that
allows for dimerisation of the guest (Table 3). Although the
two models gave quite different association constants, the
1
overall DDG values are similar: ‡0.6 Æ 0.8 kJmol for the
1
1:1 model and ‡0.5 Æ 0.7 kJmol for the 1:1 ‡ dimerisation
model. The reason is that the effects of dimerisation cancel
out in the cycle in the same way as the effects of the
conformational equilibria. Thus the double-mutant cycle
provides a robust method for removing systematic errors
from the titration measurements.
for 18
for 26
Kobs ˆ 0.1Kcis ‡ 0.9Ktrans
(1)
(2)
Kobs ˆ 0.1Kcis-Pyr ‡ 0.2Kcis-Me ‡ 0.7Ktrans
Since Ktrans is more than five times larger than the other
binding constants, and the errors in the measurements are
about 10%:
The limiting complexation-induced changes in chemical
shift for all complexes are shown in Table 3. Although the
magnitudes of the complexation-induced changes in chemical
shift vary slightly, the pattern remains the same throughout
the double-mutant cycle for all four complexes, so we can be
confident that the structure of the core of the complex is not
affected by the chemical mutations. The intermolecular NOEs
detected in the two-dimensional ROESY spectra are shown in
Table 2, and again the pattern of NOEs is consistent for all of
the complexes. The stacking interaction measured for the
anthracene ± aniline system is slightly repulsive, but small. The
result might be interpreted as the absence of any interaction
between the two aromatic rings, since the value is zero within
experimental error. However, the complexation-induced
changes in chemical shift and NOEs show that the aromatic
rings are stacked in close contact, and the measurement
therefore represents a genuine evaluation of the stacking
interaction rather than the absence of any interaction due to a
problem with the geometry of the complex. The slightly
repulsive interaction can be rationalised as the result of
competition between favourable van der Waals interactions
and repulsive electrostatic interactions between the p-elec-
tron systems.
for 18
for 26
Kobs ꢀ 0.9Ktrans
Kobs ꢀ 0.7Ktrans
(3)
(4)
1
In the case of 26, Ktrans is 64m and Equation (4) gives
1
Kobs ꢀ 45m , which is very close to the true weighted average
1
of 47m . Ideally, we would like to construct double-mutant
cycles using association constants for a single well-defined
conformer, the trans conformer. However, using Equa-
tions (3) and (4), we can show that this is not required.
ꢀ
ꢁ
KA KD
KB KC
DDG ˆ RTLn
(
(
)
Ktransꢁ6 Á 18† Ktransꢁ28 Á 26†
Ktransꢁ6 Á 26† Ktransꢁ28 Á 18†
0:9 Ktransꢁ6 Á 18†  0:7 Ktransꢁ28 Á 26†
0:7 Ktransꢁ6 Á 26†  0:9 Ktransꢁ28 Á 18†
Kobsꢁ6 Á 18† Kobsꢁ28 Á 26†
Kobsꢁ6 Á 26† Kobsꢁ28 Á 18†
ˆ
ˆ
ꢀ
RTLn
RTLn
)
(5)
(
)
RTLn
As we have seen in the X-ray crystal-structure analysis of
10, above, perfluorination of an aromatic ring makes aromat-
ic-perfluoroaromatic stacking interactions favourable because
the quadrupole moment is reversed in a fluorinated ring.[55] In
order to quantify this effect by using our system, the double-
Although the absolute values of the observed binding
constants are clearly affected by the mixture of conformers, in
the double-mutant cycles, it is only the relative values of Ktrans
we are interested in. In other words, the weakly binding minor
conformers do not affect the results of the double-mutant
Table 3. Ka [m 1] and DG [kJmol 1] values and limiting complexation-induced changes in chemical shifts (ppm) measured in CDCl3 at 300 K for complexes used
in the double-mutant cycles.[a]
Com-
plex
1:1 Model
1:1 ‡ Dimer-
Isophthaloyl compound
nh aa me
Bisaniline compound
isation Model
Ka
DG
Ka
DG
s
d
t
mu
p1
p2
p3 n1 n2
b1
b2
b3
b4
A
6 ´ 18
85 Æ 7
11.1 Æ 0.2 63 Æ 6
12.8 Æ 0.4 134 Æ 22
10.3 Æ 0.2 0.2
12.2 Æ 0.4 0.1
0.3
0.5
1.1 1.8
1.2 1.3
0.3
0.1
0.4
0.2
±
±
2.3
2.3
1.5 0.1 2.7 0.4
1.5 0.3 2.5 2.1
0.2
0.1
0.0
0.1
0.2
0.1
0.1
0.0
6 ´ 33 168 Æ 26
B
6 ´ 26 100 Æ 23
C
11.5 Æ 0.6 73 Æ 17
10.7 Æ 0.5 0.2
0.4
1.0 1.6
0.0
0.1
±
2.1
1.4 0.1 3.0 0.8
0.1
0.2
0.1
0.0
28 ´ 18 70 Æ 9
28 ´ 33 102 Æ 19
D
28 ´ 26 65 Æ 6
10.6 Æ 0.3 49 Æ 8
11.5 Æ 0.5 72 Æ 16
9.7 Æ 0.4 0.0
10.7 Æ 0.6 0.0
0.4
0.3
0.9 1.0
0.5 0.6
±
±
±
±
0.5 2.3
0.2 3.2
0.5 0.6 3.0 0.6
0.4 0.4 2.3 1.9
0.2
0.2
0.1
0.2
0.1
0.3
0.1
0.1
10.4 Æ 0.3 46 Æ 6
9.5 Æ 0.3 0.1
0.4
0.8 1.0
±
±
0.1 2.5
0.5 0.6 3.1 0.9
0.2
0.1
0.1
[a] See Schemes 7, 8, 15 and 19 for the proton labelling schemes. There are no significant changes in chemical shift for the signals due to the protons on the piperidine ring
and solubilising group R. Titration experiments were repeated at least twice and Ka is a weighted mean based on the observed change in chemical shift for all the signals
monitored. The error is twice the standard error.
4872
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4872 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 22

Stacking Interactions
4863±4878
the complexes in the pentafluorophen-
yl double-mutant cycle are shown in
Table 3. The pattern is similar to the
anthracene ± aniline experiment. The
complexation-induced changes in
chemical shift of the fluorine atoms
were also determined. For complex 6 ´
33, the changes in chemical shift are:
R
N
R
N
B
A
GA - GB
H
H
O
O
N
N
N
F
F
N
Me
O
O
H
H
O2N
O2N
F
N
N
H
F
H
F
H
N
H
N
O
O
N
N
H
H
O
O
Dd ˆ 1.0 ppm (f1),
0.3 ppm (f2)
and 0.1 ppm (f3). For complex 28 ´
33, the changes in chemical shift are:
GB - GD
GA - GC
G = GA - GB - GC + GD
Dd ˆ 0.1 ppm (f1),
0.4 ppm (f2)
R
R
N
N
and 0.2 ppm (f3). These were deter-
mined in the same way as the proton
shifts, but it is difficult to interpret
these changes. Using the results of the
1H NMR titrations in a double-mutant
cycle, we find that the pentafluoro-
phenyl ± aniline interaction is slightly
C
D
H
H
O
N
O
N
N
F
F
N
Me
O
O
H
H
O2N
O2N
F
N
N
H
F
H
F
H
N
H
O
O
N
GC - GD
N
N
1
H
H
attractive, DDG ˆ 0.4 Æ 0.9 kJmol .
O
O
Although this value is within exper-
imental error of the anthracene ± ani-
line interaction, the pentafluorophen-
yl ± aniline interaction is more attrac-
tive, and this can be rationalised based
Scheme 18. Chemical double-mutant cycle to measure the pentafluorophenyl ± aniline stacking inter-
action in complex A (R ˆ solubilising group, see Scheme 9).
Scheme 19. Synthesis of compound 33 from 23.
mutant cycle in Scheme 18 was designed. Compound 33 was
prepared according to Scheme 19. The solubility of 33 is lower
than the anthracene and methyl derivatives used previously
(9mm versus 50mm), and this limits the accuracy of titration
experiments with this system. A 1H NMR dilution experiment
on the change in the quadrupole moment of the p-system
which leads to more favourable electrostatic interactions for
the pentafluorophenyl interaction.
1
gave Kd ˆ 11 Æ 2m , and the
Conclusion
dimerisation-induced changes
in chemical shift are shown in
Table 1. As before, the main
site of dimerisation is the pyr-
role NH and pyrrole amide
(Scheme 11a).
We have developed a system to measure aromatic stacking
interactions in which we can dictate the geometry of the
complexes by using a combination of noncovalent interac-
tions. The system is capable of measuring both attractive and
repulsive aromatic interactions while maintaining the same
geometry throughout the experiments. The effects of dimer-
isation and other conformational equilibria have been taken
into account, but have been shown to cancel out in the double-
mutant cycle. X-ray crystallography is a powerful tool in the
design of molecular complexes and reveals the geometry of
aromatic interactions that can be expected in solution. This
system has the potential to be used to probe other aromatic
interactions by changing the size and electronic nature of the
interacting rings to fully investigate structure-activity rela-
tionships in aromatic stacking.
a1
a2
a3
a4
a5
m1
0.3
0.1
±
0.1
±
0.1
±
0.2
±
±
±
Compound 33 has minor con-
formers similar to those in 18,
±
±
±
±
±
±
0.4 but because we have shown that
they have an insignificant effect
0.2
±
0.2
±
0.2
±
0.1
±
0.2
±
±
on the overall result of the
±
double-mutant cycle, they were
ignored in this experiment. The
±
±
±
±
±
0.0
results of 1H NMR titrations
and the complexation-induced
changes in chemical shift for
Chem. Eur. J. 2001, 7, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4873 $ 17.50+.50/0
4873

C. A. Hunter et al.
FULL PAPER
chemical shift. The results are compared with the values obtained by the
fraction bound method in Table 4. Except for the value for p2, reasonable
agreement is obtained; this indicates that the method used to determine the
limiting complexation-induced changes in chemical shift is reliable.
Experimental Section
NMR Experiments: 1H, 19F and 13C spectra were recorded on either a
Bruker AC250 or AMX400 spectrometer with residual solvent as an
internal standard. Fluorine chemical shifts were referenced to an external
CFCl3 reference. Two-dimensional ROESY experiments were recorded on
a Bruker AMX2-400 with 300 ms mixing time and a 3 s relaxation delay
between pulses.
For the 1H NMR dilution experiments, a 2.0 mL sample of known
concentration was prepared in CDCl3. Aliquots of this solution were
sequentially added to 0.5 mL of CDCl3, and a 1H NMR spectrum recorded
after each addition. All dilutions were repeated twice. A two-dimensional
ROESY experiment was carried out on the final solution to assign
unambiguously all of the signals, and the data were analysed by using
nonlinear curve fitting for both dimerisation and aggregation isotherms
NMRDil Dimer and NMRDil Agg.[56] These estimate the dimerisation
constant and the limiting bound and free chemical shifts. The dimerisation
constant is quoted as the weighted mean based on the observed change in
chemical shift for all of the signals monitored. The error is twice the
standard error.
General synthetic procedures: The reactions were carried out according to
the following general procedures unless otherwise stated.
Acid chloride preparation: Acid chlorides were formed by suspending the
corresponding acid in dry CH2Cl2 under a nitrogen atmosphere. An excess
of oxalyl chloride and 1 ± 2 drops of catalytic DMF were added. The
reaction mixture was stirred until all the acid had gone into solution
(typically 0.5 ± 1.0 h), and the CH2Cl2 and excess oxalyl chloride were
removed under reduced pressure. The acid chloride was used without
further purification.
Amide coupling reactions: These reactions were carried out in CH2Cl2
under a nitrogen atmosphere. The acid chloride was prepared as above,
dissolved in CH2Cl2 and added drop-wise to a stirred solution of the amine
and base (where applicable) in CH2Cl2 (50 ± 200 mL). The reactions were
stirred for 16 h and the solvent was then removed under reduced pressure.
The product was isolated by column chromatography.
1H NMR titrations were carried out by preparing a 3 mL sample of the host
at known concentration (3 ± 4mm). 0.5 mL of this solution was removed,
and a H NMR spectrum was recorded. An accurately weighed sample of
Piperidine deprotection and addition of the solubilising group: The
benzylchloroformate-protected compound was dissolved in CH2Cl2. TMSI
was added, and the mixture was stirred until the reaction was completed as
determined by TLC (0.5 ± 16 hours). The solvent was removed under
reduced pressure, and the resulting solid was washed repeatedly with
diethyl ether and dried. Compound 21, 1-hydroxybenzotriazole (HOBt)
and 1,3-(dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC)
were dissolved in CH2Cl2 and stirred for 1 h, and then the free piperidine
was added portion-wise with a few drops of triethylamine. The reaction
mixture was stirred for 16 h. The reaction was quenched with methanol
(2 mL), and the solvent was removed under reduced pressure. The product
was isolated by medium-pressure column chromatography (CH2Cl2/meth-
anol 98:2). The product was dissolved in CH2Cl2, washed with 1m HCl, sat.
Na2HCO3 and water and dried over Na2SO4. The solvent removed under
reduced pressure.
1
the guest was then dissolved in 2 mL of the host solution. The solution was
saturated with guest (35 ± 40mm) to allow access to as much of the binding
isotherm as possible (50 ± 86% was achieved) and contained host, so that
the concentration of host remained constant during the titration. Aliquots
of guest solution were added successively to the NMR tube containing the
host solution, and the 1H NMR spectrum was recorded after each addition.
Signals that moved more than 0.01 ppm were analysed by using nonlinear
curve fitting, NMRTit HG and NMRTit HGHHGG.[57] NMRTit HG fits
the data to a 1:1 binding model to give the association constant and the
limiting bound chemical shift in the complex, and NMRTit HGHHGG fits
the data to a 1:1 binding isotherm, allowing for dimerisation of both binding
partners. The association constant for each experiment was evaluated as the
weighted mean based on the observed change in chemical shift for all
signals monitored. The error was taken as twice the standard error.
Titrations where the identities of the host and guest could be reversed were
used to determine the limiting complexation-induced shifts for both
binding partners. However, this was not possible in cases where the
solubility of the guest was not sufficient to achieve saturation. In these
cases, a 1:1 mixture of the two components was prepared, and the 1H NMR
spectrum was recorded. A ROESYexperiment was used to unambiguously
assign all of the signals. By using the limiting complexation-induced
changes in chemical shift for one of the
N-[(2,6-Diisopropyl)phenyl]-9-anthracene carboxamide (8): Compound 8
was prepared by the standard amide coupling procedure by using
9-anthracene carboxylic acid (1.5 g, 6.76 mmol) in CH2Cl2 (40 mL), 2,6-
diisopropyl aniline (1.0 g, 5.64 mmol) and triethylamine (0.57 g, 5.64 mmol)
in CH2Cl2 (50 mL). Purification by column chromatography yielded a pale
yellow solid. Yield: 0.9 g, 28%; m.p. decomposes >2068C; 1H NMR
3
(250 MHz, CDCl3, 258C, TMS): d ˆ 8.54 (s, 1H), 8.41 (d, J(H,H) ˆ 7 Hz,
2H), 8.07 (d, 3J(H,H) ˆ 7 Hz, 2H), 7.54 (m, 4H), 7.38 (t, 1H), 7.29 (m, 2H),
3
3
3.54 (sept, J(H,H) ˆ 7 Hz, 2H), 2.14 (d, J(H,H) ˆ 7 Hz, 12H); 13C NMR
components from the titration experi-
Table 4. Fraction bound calculations and limiting complexation-induced changes in chemical shift (ppm) for
complex 28 ´ 18.
ment (Ddbound) and the free chemical
shift from the dilution experiments,
the proportion of bound complex
could be determined:
[a]
[b]
dfree
dmixture
Ddobs
Ddbound
%B
Ddbound
Ddbound
Host 28
nh
d
mu
Guest 18
p1
p2
p3
n1
n2
b1
b2
b3
b4
a1
a2
a3
a4
6.292
7.893
3.447
6.616
7.740
3.275
0.324
0.153
0.172
0.77
0.30
0.34
42
50
49
fraction bound ˆ Ddobs/Ddbound
(6)
The observed change in chemical shift
of the other compound in the 1:1
mixture could then be used to deter-
mine the limiting complexation-in-
duced changes in chemical shift
(Ddbound).
9.85
7.79
7.28
7.14
7.23
2.25
7.04
7.07
5.52
8.35
7.55
7.55
8.07
8.55
10.35
7.47
7.30
7.71
7.32
2.21
7.04
7.11
2.53
8.29
7.53
7.53
8.04
8.50
0.50
0.32
0.02
0.57
0.09
0.04
0.00
0.05
0.01
0.06
0.02
0.02
0.03
0.05
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
±
2.3
0.5
0.6
3.0
0.6
0.2
0.1
0.1
0.1
0.2
0.2
0.2
0.1
0.2
1.3
0.04
nd[c]
2.2
0.7
0.1
0.0
0.1
0.0
0.1
nd
Ddbound ˆ Ddobs/fraction bound
(7)
a
In order to verify this method,
titration was carried out with the
identities of the host and guest re-
versed (18 ´ 28), and the signals were
analysed with NMRTit HG (18 was
used at low concentration, so there
should be an insignificant amount of
dimerisation) to determine the limit-
ing complexation-induced changes in
nd
0.1
0.1
a5
[a] Ddbound calculated by fraction bound method. [b] Ddbound determined by reverse titration method. [c] Not
determined due to signal overlap during the experiment.
4874
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4874 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 22

Stacking Interactions
4863±4878
(250 MHz, CDCl3, 258C): d ˆ 169.0, 146.2, 131.7, 131.2, 130.5, 128.8, 128.7,
the pH brought to about 5 by the addition of 10% v/v formic acid. The
product was extracted with CH2Cl2 and dried over Na2SO4. The solvent was
removed under reduced pressure, and the resulting syrup was purified by
column chromatography directly through basic alumina (CH2Cl2) to yield
the intermediate ester 20.
‡
128.5, 125.1, 123.8, 29.0, 24.0; FAB MS: m/z: 381 [M] .
N-[(2,6-Dimethyl)phenyl]-9-anthracene carboxamide (9): Compound 9
was prepared by the standard amide coupling procedure with 9-anthracene
carboxylic acid (2.2 g, 9.90 mmol) in CH2Cl2 (30 mL), 2,6-dimethylaniline
(1.0 g, 8.25 mmol) and triethylamine (0.84 g, 8.25 mmol). Purification by
column chromatography yielded a pale yellow solid. Yield: 1.0 g, 38%; m.p.
decomposes >2258C; 1H NMR (250 MHz, CDCl3, 258C, TMS): d ˆ 8.50 (s,
1H), 8.37 (d, 3J(H,H) ˆ 9 Hz, 2H), 8.03 (d, 3J(H,H) ˆ 9 Hz, 2H), 7.65 ± 7.45
(m, 4H), 7.3 (brs, 1H), 6.55 (brs, 2H), 2.50 (s, 6H); 13C NMR (250 MHz,
CDCl3, 258C): d ˆ 168.0, 135.3, 133.8, 132.0, 131.3, 128.8, 128.5, 127.8, 127.0,
Compound 20 (26.2 mmol) was suspended in a solution of KOH (3.93 g,
66 mmol) in ethanol/water (19:1, 130 mL) and heated under reflux for 1 h.
The reaction was cooled and brought to pH 5 with 10% v/v formic acid.
Compound 21 precipitated upon further cooling, it was recrystallised from
hot ethanol (225 mL) before being dried in a vacuum oven to yield a
colourless waxy solid. Yield: 11 g, 56%; m.p. 43 ± 458C; 1H NMR
(250 MHz, CDCl3, 258C, TMS): d ˆ 7.31 (s, 2H), 4.10 ± 4.00 (m, 6H),
1.85 ± 1.70, 1.50 ± 1.20, 1.20 ± 1.0 (brm, 72H), 0.90 ± 0.80 (m, 6H); elemental
analysis calcd (%) for C50H90O5: C 77.87, H 11.76; found C 77.52, H 11.95.
‡
125.7, 125.3, 193.8; FAB MS: m/z: 325 [M] .
N-Pentafluorophenyl-9-anthracene carboxamide (10): Compound 10 was
prepared by the standard amide coupling procedure with 9-anthracene
carboxylic acid (0.25 g, 1.13 mmol) in CH2Cl2 (30 mL), pentafluoroaniline
(0.28 g, 1.24 mmol) and triethylamine (0.12 g, 1.24 mmol) in CH2Cl2
(50 mL). The reaction was stirred for 3 days. Purification by column
chromatography yielded a pale yellow solid. Yield: 0.18 g, 40%; m.p.
decomposes >2208C; 1H NMR (250 MHz, CDCl3, 258C, TMS): d ˆ 8.5 (s,
1H), 8.13 (d, 3J(H,H) ˆ 9 Hz, 2H), 8.04 (d, 3J(H,H) ˆ 9 Hz, 2H), 7.6 ± 7.5
(m, 4H), 7.34 (brs, 1H); 13C NMR (250 MHz, CDCl3, 258C): d ˆ 134.1,
130.9, 129.3, 128.6, 128.1, 127.4, 125.7, 124.4; 19F NMR (CDCl3) d ˆ 144,
[4,4-Bis(4-amino-3,5-dimethylphenyl)]piperidine (22): A mixture of 2,6-
dimethylaniline (117 g, 1236 mmol), N-acetyl-4-piperidone (63.5 g,
449 mmol) and conc. HCl (150 mL) was heated under reflux. On the
second day, 2,6-dimethylaniline (15 g, 120 mmol) was added to the
refluxing mixture. This was repeated on the fourth day of reflux. After
6 days, the reaction mixture was cooled to room temperature and diluted
with water (1200 mL). The reaction was neutralised by addition of solid
sodium carbonate, and the resulting precipitate was filtered off and washed
with water, diethyl ether and pentane. The resulting colourless powder was
dried to give a colourless solid. Yield: 87.5 g, 60%; m.p. 203 ± 2058C;
1H NMR (250 MHz, [D6]DMSO, 258C, TMS): d ˆ 8.85 (brs, 1H), 6.71 (s,
4H), 4.41 (brs, 4H), 2.95 ± 2.90 (brm, 4H), 2.45 ± 2.40 (brm, 4H), 2.09 (s,
12H); 13C NMR (250 MHz, [D6]DMSO, 258C): d ˆ 141.9, 134.9, 127.8,
‡
156, 162; FAB MS: m/z: 387 [M] .
N-(9-Anthranoylcarboxy)-{1,1-bis[(4-amino-3,5-dimethyl)phenyl]cyclo-
hexane} (17): Compound 17 was prepared by the standard amide coupling
procedure with 9-anthracene carboxylic acid (1.07 g, 4.8 mmol) in CH2Cl2
(50 mL), bisaniline 16 (7.74 g, 24.0 mmol) and triethylamine (0.48 g,
4.8 mmol). The reaction mixture was washed with 1m HCl to remove
excess 16. Purification by column chromatography gave a yellow solid.
Yield: 1.34 g, 55%; m.p. 175 ± 1788C; 1H NMR (250 MHz, CDCl3, 258C,
‡
120.6, 115.9, 41.4, 41.1, 32.8, 18.2; FAB MS: m/z: 324 [M‡H] .
(1-Benzyloxycarbonyl)-[4,4-bis(4-amino-3,5-dimethylphenyl)]piperidine
(23): Compound 22 (14.1 g, 44 mmol) and triethylamine (9.1 mL 66 mmol)
were dissolved in methanol (250 mL). Benzylchloroformate (6.9 mL
48 mmol) in methanol (30 mL) was added dropwise. The reaction mixture
was stirred for 16 h, and the reaction mixture was reduced to half the
original volume under reduced pressure. Water (200 mL) was added, and
the resulting precipitate was filtered and washed repeatedly with petroleum
ether. The product was isolated by column chromatography to afford a pink
solid. Yield: 13.6 g, 68%; m.p. 132 ± 1348C; 1H NMR (250 MHz, CDCl3,
258C, TMS): d ˆ 7.35 ± 7.30 (m, 5H), 6.68 (s, 4H), 5.03 (s, 2H), 3.32 (brs,
4H), 2.50 ± 2.45 (brm, 4H), 2.15 ± 2.10 (brm, 4H), 2.00 (s, 12H); 13C NMR
(250 MHz, [D6]DMSO, 258C): d ˆ 154.7, 141.7, 137.3, 134.9, 128.5, 127.9,
3
TMS): d ˆ 8.50 (s, 1H), 8.35 (d, J(H,H) ˆ 9 Hz), 8.02 (d, 3J(H,H) ˆ 7 Hz),
7.52 (m, 4H), 7.21 (s, 1H), 7.08 (s, 2H), 6.89 (s, 2H), 3.47 (brs, 2H), 2.47 (s,
6H), 3.44 (brm, 4H), 2.16 (s, 6H), 1.58 (brm, 6H); 13C NMR (250 MHz,
CDCl3, 258C): d ˆ 167.8, 148.9, 140.0, 137.5, 134.4, 132.0, 131.1, 130.7, 128.6,
128.5, 128.3, 127.3, 127.0, 126.8, 125.5, 125.2, 121.5, 44.9, 37.2, 26.4, 23.0, 20.1,
‡
18.1; FAB MS: m/z: 526 [M] ; elemental analysis calcd (%) for
C37H38N2O ´ 0.5H2O: C 82.95, H 7.33, N 5.22; found C 82.57, H 7.14, N 5.18.
N-(9-Anthranoylcarboxy)-N'-(4-nitropyrrole-2-carboxy){1,1-bis[(4-amino-
3,5-dimethyl)phenyl]cyclohexane} (11): Compound 11 was prepared by the
standard amide coupling procedure with 4-nitropyrrole-2-carboxylic acid
(1.74 g, 11.2 mmol) in CH2Cl2 (30 mL), 17 (4.51 g, 8.57 mmol) and pyridine
(1.35 g, 17.1 mmol). The reaction mixture was washed with 1m HCl and
water. Purification by column chromatography gave a pale yellow solid.
Yield: 4.5 g, 80%; m.p. >3008C; 1H NMR (250 MHz, [D6]DMSO, 258C,
TMS): d ˆ 12.93 (s, 1H), 10.16 (s, 1H), 9.63 (s, 1H), 8.75 (s, 1H), 8.25 (d,
‡
127.5, 126.1, 120.4, 66.1, 42.1, 41.2, 35.6, 18.2; FAB MS: m/z: 457 [M] ;
elemental analysis calcd (%) for C29H35N3O2: C 74.12, H 7.71, N 9.18; found
C 74.48, H 7.61, N 8.82.
(1-Benzyloxycarbonyl)-N-(9-anthranoylcarboxy)-[4,4-bis(4-amino-3,5-di-
methylphenyl)]piperidine (24): Compound 24 was prepared by the stand-
ard amide coupling procedure with 9-anthracene carboxylic acid (0.50 g,
2.24 mmol) in CH2Cl2 (50 mL), 23 (5.20 g, 11.2 mmol) and triethylamine
(1.13 g, 11.2 mmol). The reaction mixture was washed with 1m HCl to
remove excess 23, which was recovered. Purification by column chroma-
tography gave a pale yellow solid. Yield: 0.85 g, 60%; m.p. 150 ± 1528C;
1H NMR (250 MHz, CDCl3, 258C, TMS): d ˆ 8.48 (s, 1H), 8.30 (d,
3
3J(H,H) ˆ 8 Hz, 2H), 8.19 (d, J(H,H) ˆ 8 Hz, 2H), 7.98 (s, 1H), 7.67 ± 7.57
(m, 4H), 7.18 (s, 2H), 7.16 (s, 2H), 2.47 (s, 6H), 2.32 (brm, 4H), 2.18 (s, 6H),
1.52 (brm, 6H); 13C NMR (250 MHz, [D6]DMSO, 258C): d ˆ 168.6, 158.7,
147.8, 147.6, 137.2, 135.2, 134.8, 131.3, 131.1, 130.2, 128.6, 128.1, 127.4, 126.8,
125.6, 125.5, 124.8, 106.0, 45.2, 37.0, 31.1, 26.3, 22.9, 19.9, 18.6; FAB MS:
‡
m/z: 665 [M‡H] ; elemental analysis calcd (%) for C42H40N4O4 ´ H2O:
C 73.87, H 6.20, N 8.20; found C 74.13 H 6.03 N 8.20.
3
3J(H,H) ˆ 8 Hz, 1H), 8.02 (d, J(H,H) ˆ 7 Hz, 1H), 7.50 (m, 4H), 7.30 (m,
5H), 7.01 (s, 2H), 6.84 (s, 2H), 5.07 (s, 2H), 3.60 (brs, 2H), 3.45 ± 3.30 (brm,
4H), 2.46 (s, 6H), 2.30 ± 2.10 (brm, 4H), 2.16 (s, 6H); 13C NMR (250 MHz,
[D6]DMSO, 258C): d ˆ 167.8, 155.4, 147.0, 140.7, 136.9, 134.9, 134.7, 131.9,
131.4, 128.6, 128.4, 128.3, 127.9, 127.7, 127.1, 127.0, 126.8, 125.5, 125.1, 121.8,
N,N'-Bis-pentafluorphenyl isophthaldiamide (7): Compound 7 was pre-
pared by the standard amide coupling procedure with isophthaloyl
dichloride (0.30 g, 1.5 mmol) and pentafluoroaniline (1.08 g, 5.9 mmol) in
CH2Cl2 (50 mL). The reaction was stirred for 6 days, and a colourless
precipitate formed. The reaction mixture was diluted with CH2Cl2, washed
with 1m NaOH, 1m HCl and water and dried over Na2SO4. The solvent was
then removed under reduced pressure. The product was recrystallised form
ethanol/water to give a colourless solid. Yield: 0.59 g, 80%; m.p. >2508C;
1H NMR (250 MHz, [D6]DMSO, 258C, TMS): d ˆ 10.83 (brs, 2H), 8.63 (s,
1H), 8.26 (d, 3J(H,H) ˆ 8 Hz, 2H), 7.77 (t, 1H, 3J(H,H) ˆ 8 Hz); 19F NMR
‡
66.9, 43.5, 41.1, 36.0, 20.0, 18.1; FAB MS: m/z: 662 [M‡H] ; HRMS-FAB:
m/z: 662.3319 (calcd for C44H44N3O3 662.3382).
(1-Benzyloxycarbonyl)-N-(9-anthranoylcarboxy)-N'-(4-nitropyrrole-2-car-
boxy)-[4,4-bis(4-amino-3,5-dimethylphenyl)]piperidine (25): Compound
25 was prepared by the standard amide coupling procedure with 4-nitro-
pyrrole-2-carboxylic acid (0.28 g, 1.84 mmol) in CH2Cl2 (40 mL), 24 (1.1 g,
1.67 mmol) and pyridine (0.15 g, 1.90 mmol). Purification by medium
pressure column chromatography gave a pale yellow solid. Yield: 1.1 g,
85%; m.p. 204 ± 2068C; 1H NMR (250 MHz, [D6]DMSO, 258C, TMS): d ˆ
12.85 (s, 1H), 10.15 (s, 1H), 9.63 (s, 1H), 8.71 (s, 1H), 8.21 (d, 3J(H,H) ˆ
‡
(CDCl3/CD3OD) d ˆ 145, 153, 163; FAB MS: m/z: 497 [M‡H] ;
elemental analysis calcd (%) for C20H6F10N2O2 ´ 0.5H2O: C 47.54, H 1.40, N
5.54; found C 47.84, H 1.32, N 5.54.
(3,4,5-Tri-O-tetradecane)benzoic acid (21): n-Propyl gallate 19 (5.55 g,
26.2 mmol), 1-bromodecane (22 mL, 78.6 mmol) and anhydrous K2CO3
(32.2 g, 236 mmol) were suspended in a mixture of acetone/DMSO (9:1,
100 mL). The mixture was stirred for 30 minutes and then heated under
reflux for 10 h. The reaction mixture was poured into water (1500 mL), and
3
8 Hz, 2H), 8.16 (d, J(H,H) ˆ 8 Hz, 2H), 7.95 (s, 1H), 7.65 ± 7.55 (m, 5H),
7.35 ± 7.30 (m, 5H), 7.18 (s, 2H), 7.15 (s, 2H), 5.05 (s, 2H), 3.60 ± 3.40 (brm,
4H), 2.60 ± 2.30 (brm, 4H), 2.43 (s, 6H), 2.14 (s, 6H); 13C NMR (250 MHz,
[D6]DMSO, 258C): d ˆ 167.3, 155.0, 145.5, 145.3, 137.5, 136.8, 135.9, 135.5,
Chem. Eur. J. 2001, 7, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4875 $ 17.50+.50/0
4875

C. A. Hunter et al.
FULL PAPER
133.7, 133.2, 131.2, 128.2, 127.9, 127.0, 126.8, 126.5, 126.1, 66.6, 43.8, 20.0,
and the resultant colourless precipitate was filtered and washed repeatedly
with CH2Cl2 and petroleum ether to afford a colourless solid. Yield: 1.7 g,
90%; m.p. 169 ± 1718C; 1H NMR (250 MHz, [D6]DMSO, 258C, TMS): d ˆ
8.90 (s, 2H), 8.55 (s, 1H), 8.14 (d, 3J(H,H) ˆ 8 Hz, 2H), 8.09 (s, 4H), 7.65 (t,
3J(H,H) ˆ 8 Hz, 1H), 2.90 (s, 6H); 13C NMR (250 MHz, [D6]DMSO,
258C): d ˆ 166.2. 134.5. 129.6. 128.4. 126.1. 26.3. 24.1; FAB MS: m/z: 192
‡
19.1; FAB MS: m/z: 800 [M‡H] ; HRMS-FAB: m/z: 800.3460 (calcd for
C49H46N5O6 800.3448).
[1-(3,4,5-Tri-O-tetradecane)-benzoyl]-N-(9-anthranoylcarboxy)-N'-(4-ni-
tropyrrole-2-carboxy)-[4,4-bis(4-amino-3,5-dimethylphenyl)]piperidine
(18): Compound 18 was prepared by the standard piperidine deprotection
procedure with 25 (1.4 g, 1.75 mmol) in CH2Cl2 (300 mL) and TMSI
(0.6 mL). The solubilising group was added by using 21 (1.46 g, 1.91 mmol),
EDC (0.40 g, 2.10 mmol) and HOBt (0.28 g, 2.10 mmol). Purification by
column chromatography afforded a pale yellow waxy solid. Yield: 0.97 g,
40%; 1H NMR (250 MHz, CDCl3, 0.0095m, 258C, TMS): d ˆ 10.42 (s, 1H),
8.51 (s, 1H), 8.30 (d, 3J(H,H) ˆ 7 Hz, 2H), 8.04 (d, 3J(H,H) ˆ 7 Hz, 2H),
7.61 (s, 1H), 7.51 (m, 4H), 7.57 (s, 1H), 7.28 (s, 1H), 7.31 (s, 1H), 7.07 (s, 2H),
6.99 (s, 2H), 6.56 (s, 2H), 3.95 (m, 6H), 3.80 ± 3.40 (brm, 4H), 2.50 ± 2.20
(brm, 4H), 2.50 (s, 6H), 2.18 (s, 6H), 1.80 ± 1.50, 1.50 ± 1.0 (brm, 72H),
0.95 ± 0.80 (m, 6H); 13C NMR (250 MHz, [D6]DMSO, 258C): d ˆ 170.5,
168.2, 158.3, 145.4, 139.2, 137.5, 135.7, 135.3, 131.8, 131.3, 131.0, 128.7, 128.2,
127.1, 126.9, 126.6, 125.6, 125.5, 124.8, 105.3, 73.5, 69.2, 44.2, 32.9, 30.3, 29.7,
‡
[M] .
N-(4-Nitropyrrole-2-carboxy)-{1,1-bis[(4-amino-3,5-dimethyl)phenyl]cy-
clohexane} (31): Compound 31 was prepared by using the standard amide
coupling procedure with 4-nitro-pyrrole-2-carboxylic acid (1.0 g,
6.41 mmol) in CH2Cl2 (40 mL), bisaniline 16 (6.19 g, 19.2 mmol) and
pyridine (0.55 g, 7.05 mmol). Purification by column chromatography
afforded a yellow solid. Yield: 0.65g, 22%; m.p. 188 ± 1908C; 1H NMR
(250 MHz, [D6]DMSO, 258C, TMS): d ˆ 12.89 (s, 1H), 9.58 (s, 1H), 7.96 (s,
1H), 7.63 (s, 1H), 7.00 (s, 2H), 6.76 (s, 2H), 4.33 (brs, 2H), 2.30 ± 2.00 (m,
4H), 2.11 (s, 6H), 2.04 (s, 6H), 1.6 ± 1.4 (brm, 6H); 13C NMR (250 MHz,
[D6]DMSO, 258C): d ˆ 158.7, 149.5, 140.1, 137.5, 134.8, 129.4, 127.1, 127.0,
‡
125.6, 121.6, 105.1, 44.9, 37.1, 26.4, 22.9, 18.8, 18.0; FAB MS: m/z: 460 [M] ;
‡
29.6, 29.4, 29.3, 26.1, 22.6, 20.0, 18.7. 14.1; FAB MS: m/z: 1405 [M] ; HRMS-
elemental analysis calcd (%) for C27H32N4O3 ´ 0.25H2O: C 69.73, H 7.04, N
FAB : m/z: 1406.9712 (calcd for C90H128N5O8 1406.9762).
11.72; found C 69.95, H 7.10, N 12.05.
(1-Benzyloxycarbonyl)-N-(4-nitropyrrole-2-carboxy)-[4,4-bis(4-amino-3,5-
dimethylphenyl)]piperidine (29): Compound 29 was prepared by the
standard amide coupling procedure with 4-nitropyrrole-2-carboxylic acid
(0.75 g, 4.80 mmol) in CH2Cl2 (50 mL), 23 (10.99 g, 24.0 mmol) and
pyridine (0.57 g, 7.20 mmol). The product was isolated form the excess 23
by column chromatography followed by medium pressure column chro-
matography to give a colourless solid. Yield: 2.15 g, 75%; m.p. 158 ± 1608C;
1H NMR (250 MHz, [D6]DMSO, 258C, TMS): d ˆ 12.90 (s, 1H), 9.61 (s,
1H), 7.97 (s, 1H), 7.65 (s, 1H), 7.36 (m, 5H), 7.04 (s, 2H), 6.80 (s, 2H), 5.07(s,
2H), 3.48 (brs, 2H), 3.36 (brm, 4H), 2.31 (brm, 4H), 2.13 (s, 6H), 2.06 (s,
6H); 13C NMR (250 MHz, [D6]DMSO, 258C): d ˆ 158.2, 155.0, 147.1, 142.1,
137.5, 136.8, 135.5, 133.1, 132.1, 128.8, 128.2, 127.9, 127.0, 126.6, 126.3, 121.0,
N-(4-Nitropyrrole-2-carboxy)-N'-acetyl-{1,1-bis[(4-amino-3,5-dimethyl)-
phenyl]cyclohexane} (31): Compound 32 (1.0 g, 2.17 mmol) and glacial
acetic acid (0.13 g, 2.20 mmol) were suspended in CH2Cl2 (40 mL). EDC
(0.54 g, 2.82 mmol) was added, and the reaction mixture was stirred for
12 h. It was then washed with 1m HCl and water, and the solvent was
removed under reduced pressure. The product was isolated by column
chromatography to give a colourless solid. Yield: 0.64 g, 59%; m.p.
>2508C; 1H NMR (250 MHz, [D6]DMSO, 258C, TMS): d ˆ 12.90 (brs,
1H), 9.54 (s, 1H), 9.04 (s, 1H), 7.92 (s, 1H), 7.58 (s, 1H), 7.05 (s, 2H), 6.98 (s,
2H), 2.30 ± 2.00 (m, 4H), 2.20 (m, 4H), 2.11 (s, 6H), 2.07 (s, 6H), 1.99 (s,
3H), 1.6 ± 1.4 (brm, 6H); 13C NMR (250 MHz, [D6]DMSO, 258C): d ˆ
167.2, 147.2, 146.5, 136.8, 135.5, 135.0, 133.2, 126.6, 126.3, 45.1, 36.5, 23.0,
‡
‡
66.5, 43.1, 35.6, 18.9, 18.6; FAB MS: m/z: 595 [M] ; HRMS-FAB: m/z:
595.2835 (calcd for C34H37N5O5 595.2794).
19.0; FAB MS: m/z: 503 [M‡H] ; elemental analysis calcd (%) for
C29H34N4O4 ´ H2O: C 66.90, H 6.97, N 10.76; found C 66.89, H 6.82, N 10.65.
(1-Benzyloxycarbonyl)-N-acetyl-N'-(4-nitropyrrole-2-carboxy)-[4,4-bis(4-
amino-3,5-dimethylphenyl)]piperidine (30): Compound 29 (0.84 g,
1.41 mmol), glacial acetic acid (0.42 g, 7.0 mmol) and EDC (1.35 g,
7.0 mmol) were dissolved in CH2Cl2 (50 mL) and stirred for 16 h. The
reaction mixture was washed with 1m HCl to remove most of any excess 29,
and the solvent removed under reduced pressure. The product was isolated
by column chromatography to give a colourless solid. Yield: 0.68g, 76%;
m.p. 185 ± 1888C; 1H NMR (250 MHz, [D6]DMSO, 258C, TMS): d ˆ 12.92
(brs, 1H), 9.61 (s, 1H), 9.12 (s, 1H), 7.98 (s, 1H), 7.64 (s, 1H), 7.35 (m, 5H),
7.12 (s, 2H), 7.05 (s, 2H), 5.07 (s, 2H), 3.50 ± 3.35 (m, 4H), 2.40 ± 2.30 (m,
4H), 2.15 (s, 6H), 2.10 (s, 6H), 2.02 (s, 3H); 13C NMR (250 MHz,
[D6]DMSO, 258C): d ˆ 167.7, 127.7, 124.5, 144.9, 144.3, 137.0, 136.3, 135.4,
134.9, 133.2, 132.0, 127.7, 126.5, 126.0, 125.7, 125.6, 66.0, 22.5, 18.5, 18.4; FAB
(1-Benzyloxycarbonyl)-N-(4-nitropyrrole-2-carboxy)-N'-pentafluoroben-
zoyl-[4,4-bis(4-amino-3,5-dimethylphenyl)]piperidine (34): The monosub-
stituted bisaniline was synthesised by the standard amide coupling
procedure by reaction of pentafluorobenzoyl chloride (3.28 g, 14.2 mmol),
23 (13.8 g, 43.7 mmol) and triethylamine (2.16 g, 21.4 mmol), but was not
isolated as it contained a lot of the diaddition product and was difficult to
separate at this stage. Therefore the mixture was treated with 4-nitro-
pyrrole-2-carbonyl chloride 15 (0.53 g, 3.40 mmol) and pyridine (0.40 g,
5.1 mmol) by using the standard procedure and the products separated
after this step. The product was isolated by medium-pressure column
chromatography to give a colourless solid. Yield: 6.73 g, 60%; m.p. 187 ±
1898C; 1H NMR (250 MHz, [D6]DMSO, 258C, TMS): d ˆ 12.92 (s, 1H),
10.28 (s, 1H), 9.62 (s, 1H), 7.98 (s, 1H), 7.64 (s, 1H), 7.35 (m, 5H), 7.16 (s,
2H), 7.15 (s, 2H), 5.08 (s, 2H), 3.45 (brm, 4H), 2.40 (brm, 4H), 2.21 (s, 6H),
2.16 (s, 6H); 13C NMR (250 MHz, [D6]DMSO, 258C): d ˆ 145.2, 137.4,
136.8, 135.9, 135.2, 132.5, 131.8, 128.8, 128.2, 127.9, 126.9, 126.7, 66.5, 43.9,
18.9, 18.7; 19F NMR (CDCl3) d ˆ 141, 151, 160; FAB MS: m/z: 790
‡
MS: m/z: 638 [M‡H] ; HRMS-FAB: m/z: 638.3002 (calcd for C36H40N5O6
638.2978).
[1-(3,4,5-Tri-O-tetradecane)benzoyl]-N-acetyl-N'-(4-nitropyrrole-2-car-
boxy)-[4,4-bis(4-amino-3,5-dimethylphenyl)]piperidine (26): Compound
26 was prepared by the standard piperidine deprotection procedure with
30 (0.38 g, 0.59 mmol) in CH2Cl2 (70 mL) and TMSI (0.15 mL). The
solubilising group was added by using 21 (0.67 g, 0.89 mmol), EDC (0.22 g,
1.18 mmol) and HOBt (0.16 g, 1.20 mmol). Purification by column chro-
matography afforded a pale yellow waxy solid. Yield: 0.74 g, 60%; 1H NMR
(250 MHz, CDCl3, 0.0036m, 258C, TMS): d ˆ 10.53 (brs), 10.34 (brs, 1H),
7.70 (s, 1H), 7.32 (s, 1H), 7.30 (s, 1H), 6.98 (s, 2H), 6.93 (s, 2H), 6.65 (s, 1H),
6.53 (s, 2H), 3.95 ± 3.90 (m, 6H), 3.90 ± 3.60 (brm, 4H), 2.19, 2.18, 2.20 (s,
15H), 2.30 ± 2.20 (brm, 4H), 1.90 ± 1.70, 1.50 ± 1.35, 1.35 ± 1.20 (brm, 72H),
0.90 ± 0.80 (m, 6H); 13C NMR (250 MHz, CDCl3, 258C): d ˆ 170.4, 168.9,
158.3, 139.2, 137.5, 135.7, 135.6, 132.0, 130.8, 130.6, 126.6, 126.5, 125.6, 105.3,
73.5, 69.2, 44.1, 31.9, 30.2, 29.7, 29.7, 29.6, 29.4, 29.3, 26.0, 23.1, 22.6, 18.8,
‡
[M‡H] ; HRMS-FAB: m/z: 790.2708 (calcd for C41H37F5N5O6 790.2664).
1-(3,4,5-Tri-O-tetradecane)benzoyl)-N-(4-nitropyrrole-2-carboxy)-N'-pen-
tafluorobenzoyl-[4,4-bis(4-amino-3,5-dimethylphenyl)]piperidine
(33):
Compound 33 was prepared by the standard piperidine deprotection
procedure with 34 (0.99 g, 1.25 mmol) in CH2Cl2 (150 mL) and TMSI
(0.5 mL). The solubilising group was added by using 21 (1.14 g, 1.50 mmol),
EDC (0.28 g, 1.50 mmol) and HOBt (0.20 g, 1.50 mmol). Purification by
column chromatography gave a colourless waxy solid. Yield: 0.95 g, 70%;
1H NMR (250 MHz, CDCl3, 0.0014m, 258C, TMS): d ˆ 9.98 (brs, 1H), 9.83
(s), 7.45 (s, 1H), 7.18 (s, 1H), 7.09 (s, 1H), 7.00 (s, 2H), 6.99 (s, 2H), 6.54 (s,
2H), 5.33 (s), 3.95 ± 3,90 (m, 6H), 3.80 ± 3.40 (brm, 4H), 2.50 ± 2.30 (brm),
2.28 (s, 6H), 2.22 (s, 6H), 1.85 ± 1.60, 1.60 ± 1.10 (m, 72H), 0.95 ± 0.85 (m,
9H); 13C NMR (250 MHz, [D6]DMSO, 258C): d ˆ 170.4, 158.1, 155.9, 153.2,
146.5, 145.8, 139.3, 137.8, 135.7, 130.6, 126.9, 126.7, 125.5, 105.4, 73.5, 69.2,
44.3, 31.9. 30.3, 29.7, 29.6, 29.4, 29.3, 26.1, 22.6, 18.8, 14.0; 19F NMR (CDCl3)
‡
14.1; FAB MS: m/z: 1244 [M‡H] ; elemental analysis calcd (%) for
C77H121N5O8: C 74.29, H 9.79, N 5.25; found C 74.11, H 10.12, N 5.25.
N,N'-Bis-methyl isophthaldiamide (27): Methylamine (10.8 mL, 2m solu-
tion in tetrahydrofuran, 22 mmol) was added dropwise to isophthaloyl
dichloride (2.0 g, 9.85 mmol) in CH2Cl2 (100 mL), and the reaction was
stirred for 10 ± 15 minutes. After this time, a colourless precipitate began to
form. The reaction mixture was reduced in volume on a rotary evaporator,
‡
d ˆ 142, 150, 163; FAB MS: m/z: 1395 [M] ; HRMS-FAB: m/z:
1396.8863 (calcd for C82H119F5N5O8 1396.8978); elemental analysis calcd
(%) for C82H118F5N5O8: C 70.50, H 8.51, N 5.01; found C 70.27, H 8.61, N
4.95.
4876
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4876 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 22

Stacking Interactions
4863±4878
Crystal structure data: Crystallographic data (excluding structure factors)
for the structures reported in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary publication
nos. CCDC-162594 to 162598. Copies of this data can be obtained free of
charge on application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK
(fax: (‡44)1223-336033; e-mail: deposit@ccdc.cam.ac.uk).
Compound 8: C27H25NO; M ˆ 379.48. Crystallises from petroleum ether/
CH2Cl2 as pale yellow blocks. Crystal dimensions 0.40 Â 0.18 Â 0.10 mm;
monoclinic, a ˆ 20.611(2), b ˆ 13.4661(14), c ˆ 26.000(3) Š, b ˆ 111.789(3)8
U ˆ 6700.8(12) Š3, Z ˆ 12, Dc ˆ 1.128 Mgm 3, space group P21/n (a non
standard setting of P21/c C52 h' No.14), Mo-Ka radiation (l ˆ 0.71073 Š),
m(Mo-Ka) ˆ 0.068 mm 1, F(000) ˆ 2424.
Compound 9: C23H19NO; M ˆ 325.39. Crystallises from CH2Cl2 as colour-
less blocks. Crystal dimensions 0.68  0.42  0.32 mm; monoclinic, a ˆ
16.604(3), b ˆ 4.6310(9), c ˆ 23.722(5) Š, b ˆ 108.08(3)8, U ˆ 1734.0(6) Š3,
Z ˆ 4, Dc ˆ 1.246 Mgm 3, space group P21/c (C25 h' No.14), Mo-Ka radiation
(l ˆ 0.71073 Š), m(Mo-Ka) ˆ 0.076 mm 1, F(000) ˆ 688.
Compound 10: C21H10F5NO; M ˆ 387.30. Crystallises from petroleum
ether/CH2Cl2 as colourless blocks. Crystal dimensions 0.4 Â 0.1 Â 0.1 mm;
monoclinic, a ˆ 14.100(3), b ˆ 13.585(3), c ˆ 18.372(4) Š, b ˆ 108.308(4)8,
U ˆ 3341.0(12) Š3, Z ˆ 8, Dc ˆ 1.540 Mgm 3, space group P21/n (a non
standard setting of P21/c C52 h' No.14), Mo-Ka radiation (l ˆ 0.71073 Š),
m(Mo-Ka) ˆ 0.133 mm 1, F(000) ˆ 1568.
[18] S. C. Zimmerman, W. S. Kwan, Angew. Chem. 1995, 107, 2404 ± 2406;
Angew. Chem. Int. Ed. Engl. 1995, 34, 2404 ± 2406.
[19] J. N. H. Reek, A. Kros, R. J. M. Nolte, Chem. Commun. 1996, 245 ±
247.
[20] J. C. Ma, D. A. Dougherty, Chem. Rev. 1997, 97, 1303 ± 1324.
[21] G. R. Desiraju, A. Gavezzotti, Acta. Cryst. B 1989, 45, 473 ± 482.
[22] G. R. Desiraju, A. Gavezzotti, Chem. Commun. 1989, 621 ± 623.
[23] G. W. Coates, A. R. Dunn, L. M. Henling, D. A. Dougherty, R. H.
Grubbs, Angew. Chem. 1997, 109, 290 ± 293; Angew. Chem. Int. Ed.
Engl. 1997, 36, 248 ± 251.
[24] G. W. Coates, A. R. Dunn, L. M. Henling, J. W. Ziller, E. B. Lobkov-
sky, R. H. Grubbs, J. Am. Chem. Soc. 1998, 120, 3642 ± 3649.
[25] P. Linse, J. Am. Chem. Soc. 1992, 114, 4366 ± 4373.
[26] W. L. Jorgensen, D. L. Severance, J. Am. Chem. Soc. 1990, 112, 4768 ±
4774.
[27] R. L. Jaffe, G. D. Smith, J. Chem. Phys. 1996, 105, 2780 ± 2788.
[28] R. L. Jaffe, G. D. Smith, J. Phys. Chem. 1996, 100, 9624 ± 9630.
[29] C. Chipot, R. Jaffe, B. Maigret, D. A. Pearlman, P. A. Kollman, J. Am.
Chem. Soc. 1996, 118, 11217 ± 11224.
[30] P. Hobza, H. L. Selzle, E. W. Schlag, J. Phys. Chem. 1993, 97, 3937 ±
3938.
[31] P. Hobza, H. L. Selzle, E. W. Schlag, J. Am. Chem. Soc. 1994, 116,
3500 ± 3506.
[32] S. L. Price, A. J. Stone, J. Chem. Phys. 1987, 86, 2859 ± 2868.
[33] F. Cozzi, F. Ponzini, R. Annunziata, M. Cinquini, J. S. Siegel, Angew.
Chem. 1995, 107, 1092 ± 1093; Angew. Chem. Int. Ed. Engl. 1995, 34,
1019 ± 1020.
[34] F. Cozzi, M. Cinquini, R. Annunziata, J. S. Siegel, J. Am. Chem. Soc.
1993, 115, 5330 ± 5331.
Compound 25: C53.5H45Cl9N5O7.5 ; M ˆ 1196.99. Crystallises from petroleum
ether/CH2Cl2 as colourless blocks. Crystal dimensions 0.39 Â 0.38 Â
0.36 mm; monoclinic, a ˆ 22.82(2), b ˆ 21.214(18), c ˆ 25.75(2) Š, b ˆ
106.49(13)8, U ˆ 11949(19) Š3, Z ˆ 8, Dc ˆ 1.331 Mgm 3, space group C2/
c, Mo-Ka radiation (l ˆ 0.71073 Š), m(Mo-Ka) ˆ 0.474 mm 1, F(000) ˆ
4912.
[35] S. Paliwal, S. Geib, C. S. Wilcox, J. Am. Chem. Soc. 1994, 116, 4497 ±
4498.
[36] E. Kim, S. Paliwal, C. S. Wilcox, J. Am. Chem. Soc. 1998, 120, 11192 ±
11193.
[37] L. F. Newcomb, S. H. Gellman, J. Am. Chem. Soc. 1994, 116, 4993 ±
4994.
[38] N. J. Heaton, P. Bello, B. Herrandon, A. del Campo, J. Jiminez-
Barbero, J. Am. Chem. Soc. 1998, 120, 12371 ± 12384.
[39] M. J. Packer, C. A. Hunter, J. Mol. Biol. 1998, 280, 407 ± 420.
[40] M. J. Packer, M. P. Dauncey, C. A. Hunter, J. Mol. Biol. 2000, 295, 71 ±
83.
Compound 31: C30H36Cl2N4O4; M ˆ 587.53. Crystallises from petroleum
ether/CH2Cl2 as colourless blocks. Crystal dimensions 0.45 Â 0.20 Â
0.20 mm; triclinic, a ˆ 9.368(2), b ˆ 11.533(3), c ˆ 27.828(7) Š, a ˆ
97.545(5)8 b ˆ 92.026(5)8, g ˆ 92.943(5)8, U ˆ 2974.0(12) Š3, Z ˆ 4, Dc ˆ
3
1
Å
1.312 Mgm
,
space group P1 (Ci ' No. 2), Mo-Ka radiation (l ˆ
0.71073 Š), m(Mo-Ka) ˆ 0.260 mm 1, F(000) ˆ 1240.
Acknowledgement
[41] M. J. Packer, M. P. Dauncey, C. A. Hunter, J. Mol. Biol. 2000, 295, 85 ±
103.
[42] L. Serrano, M. Bycroft, A. R. Fersht, J. Mol. Biol. 1991, 218, 465 ± 475.
[43] H. Adams, F. J. Carver, C. A. Hunter, J. C. Morales, E. M. Seward,
Angew. Chem. 1996, 108, 1628 ± 1631; Angew. Chem. Int. Ed. Engl.
1996, 35, 1542 ± 1544.
We thank the Lister Institute (C.A.H.), EPSRC (J.P.), BBSRC (J.M.S.) and
Zeneca Agrochemicals (J.P.) for financial support.
[1] C. A. Hunter, J. K. M. Sanders, J. Am. Chem. Soc. 1990, 112, 5525 ±
5534.
[2] J. D. Watson, F. H. Crick, Nature 1953, 171, 737 ± 738.
[3] W. Saenger, Principles of Nucleic Acid Structure Springer, New York,
1988.
[4] C. A. Hunter, J. Mol. Biol. 1993, 230, 1025 ± 1054.
[5] S. K. Burley, G. A. Petsko, Science 1985, 23 ± 28.
[6] C. A. Hunter, Chem. Soc. Rev. 1994, 23, 101 ± 109.
[7] L. S. Lerman, J. Mol. Biol. 1961, 3, 18 ± 30.
[8] P. L. Anelli, N. Spencer, J. F. Stoddart, J. Am. Chem. Soc. 1991, 113,
5131 ± 5133.
[9] A. D. Hamilton, D. Van Engen, J. Am. Chem. Soc. 1987, 109, 5035 ±
5036.
[10] S. B. Ferguson, F. Diederich, Angew. Chem. 1986, 98, 1127 ± 1129;
Angew. Chem. Int. Ed. Engl. 1986, 25, 1127 ± 1129.
[11] B. J. Whitlock, H. W. Whitlock, J. Am. Chem. Soc. 1990, 112, 3910 ±
3915.
[12] J. Rebek, D. Nemeth, J. Am. Chem. Soc. 1986, 108, 5637 ± 5638.
[13] H. J. Schneider, T. Blatter, S. Simova, I. Theis, Chem. Commun. 1989,
580 ± 581.
[14] A. W. Schwabacher, S. Zhang, W. Davy, J. Am. Chem. Soc. 1993, 115,
6995 ± 6996.
[44] A. P. Bisson, F. J. Carver, C. A. Hunter, J. P. Waltho, J. Am. Chem. Soc.
1994, 116, 10292 ± 10293.
[45] A. P. Bisson, C. A. Hunter, Chem. Commun. 1996, 1723 ± 1724.
[46] F. J. Carver, C. A. Hunter, P. S. Jones, D. J. Livingstone, J. F. McCabe,
E. M. Seward, P. Tiger, Chem. Eur. J. 2001, 7, 4854 ± 4862.
[47] C. R. Patrick, G. S. Prosser, Nature 1960, 1021.
[48] T. Dahl, Acta. Chem. Scan. A 1975, 29, 170 ± 174.
[49] H. Adams, K. D. M. Harris, G. A. Hembury, C. A. Hunter, D.
Livingstone, J. F. McCabe, Chem. Commun. 1996, 2531 ± 2532.
[50] K. J. Morgan, D. P. Morrey, Tetrahedron 1966, 22, 57 ± 62.
[51] P. E. Fanta, Org. Synth. 1952, 32, 95 ± 96.
[52] C. A. Hunter, J. Am. Chem. Soc. 1992, 114, 5303 ± 5311.
[53] R. S. Lott, V. S. Chauhan, C. H. Stammer, Chem. Commun. 1979,
495 ± 496.
[54] C. A. Hunter, M. J. Packer, Chem. Eur. J. 1999, 5, 1891 ± 1897.
[55] J. Vrbancich, G. L. D. Ritchie, J. Chem. Soc. Faraday Trans. II 1980, 76,
648 ± 659.
[56] M. Gardner, A. J. Guerin, C. A. Hunter, U. Michelsen, C. Rotger, New
J. Chem. 1999, 309 ± 316. Note in Equation (6) of this paper, the factor
of 2 should be absent.
[57] A. P. Bisson, C. A. Hunter, J. C. Morales, K. Young, Chem. Eur. J.
1998, 4, 845 ± 851.
[15] J. S. Zhang, J. S. Moore, J. Am. Chem. Soc. 1992, 114, 9701 ± 9702.
[16] S. C. Zimmerman, C. M. Vanzyl, J. Am. Chem. Soc. 1987, 109, 7894 ±
7896.
[17] S. C. Zimmerman, C. M. Vanzyl, G. S. Hamilton, J. Am. Chem. Soc.
1989, 111, 1373 ± 1381.
Received: May 7, 2001 [F3238]
Chem. Eur. J. 2001, 7, No. 22
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4877 $ 17.50+.50/0
4877

C. A. Hunter et al.
FULL PAPER
4878
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
0947-6539/01/0722-4878 $ 17.50+.50/0
Chem. Eur. J. 2001, 7, No. 22