Anion binding versus intramolecular hydrogen bonding in neutral macrocyclic amides

Article abstract of DOI:10.1002/chem.200501471

Although amide groups are important hydrogen-bond donors in natural and synthetic anion receptors, studies on structure-affinity relationships of amide-based macrocyclic receptors are still very limited. Therefore, we synthesized a series of macrocyclic t

Full text of DOI:10.1002/chem.200501471

DOI: 10.1002/chem.200501471
Anion Binding versus Intramolecular Hydrogen Bonding in Neutral
Macrocyclic Amides
Michał J. Chmielewski[a] and Janusz Jurczak*[a, b]
Abstract: Although amide groups are
important hydrogen-bond donors in
natural and synthetic anion receptors,
studies on structure–affinity relation-
ships of amide-based macrocyclic re-
ceptors are still very limited. Therefore,
we synthesized a series of macrocyclic
tetraamides 5–8 derived from 1,3-ben-
zenedicarboxylic (isophthalic) acid and
aliphatic a,w-diamines of different
lengths. 1H NMR titrations in DMSO
solution show that the anion affinity of
these receptors decreases with increas-
ing size of the macrocycle irrespective
of the anion, and this suggests a minor
role of geometric complementarity.
Comparison with their previously stud-
ied pyridine congeners reveals that the
isophthalic acid based macrocycles are
less potent, in contrast to what was
found for simple model diamides. Com-
bined theoretical and experimental
structural studies were carried out to
determine the reasons behind this be-
haviour. The results show that the un-
expectedly low anion binding ability of
the isophthalic acid-based receptors is
due to the self-complementary nature
of the isophthalic bis-amide fragments:
when two such moieties are present
within a sufficiently flexible macrocy-
cle, they adopt syn–anti conformations
and bind each other by two strong in-
tramolecular hydrogen bonds that close
the macrocyclic cavity. Nevertheless,
anion binding is able to break these hy-
drogen bonds and switch a macrocycle
into a convergent all-syn conformation.
Despite the ill-preorganized conforma-
tion, 20-membered receptor 6 is better
than either its open-chain analogue
(macrocyclic effect) and/or its isomer
having differently placed carbonyl
groups. The crystal structures of four
anion complexes of the macrocyclic re-
ceptors are reported. X-ray studies and
solution NMR data confirmed the in-
clusive nature of the complexes and
pointed to strong involvement of aro-
matic CH hydrogen atoms in anion
binding.
Keywords: amides · anion binding ·
hydrogen bonds
·
macrocyclic
ligands · receptors
Introduction
Although, macrocyclic anion receptors generally perform
better than their linear counterparts,[7] most efforts in this
area were devoted to simple acyclic receptors. As a result,
little is known about the relative influence of various factors
such as the size, shape and rigidity of the macrocycle or the
type and arrangement of building blocks on anion binding
by macrocyclic, amide-based receptors. Therefore, we under-
took systematic studies on anion recognition by this class of
receptors.[8–10] Previously, we studied the influence of the
size of a macrocycle on the stability of its anion complexes.[9]
As model receptors, a series of macrocyclic tetraamides 1–4
derived from 2,6-pyridinedicarboxylic acid and flexible, ali-
phatic a,w-diamines of various lengths was used.
To add another dimension to our study, we now also
varied the structure of the diacid component. According to
Crabtree et al.,[11] isophthalamides are much more potent
anion receptors than 2,6-pyridine diamides,[12] in which the
electron pair of the pyridine nitrogen atom competes with
anions for hydrogen bonding with the amide NH groups.
According to an alternative view, higher anion affinity of
Noncovalent interactions with negatively charged species
play an important role in many essential chemical and bio-
logical processes.[1] Hydrogen-bonding interactions between
anions and amide groups are an important example. Pro-
teins can use multiple such interactions to achieve strong
and selective binding of anions even in a highly competitive
aqueous medium.[2,3] Accordingly, the development of
amide-based anion receptors is of considerable interest.[4–7]
[a] Dr. M. J. Chmielewski, Prof. J. Jurczak
Institute of Organic Chemistry
Polish Academy of Sciences
Kasprzaka 44/52, 01-224 Warsaw (Poland)
Fax : (+48)22-632-66-81
E-mail: jurczak@icho.edu.pl
[b] Prof. J. Jurczak
Department of Chemistry, Warsaw University, Warsaw (Poland)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
7652
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7652– 7667

FULL PAPER
binding affinity. It was therefore natural to extend our stud-
ies to isophthalic acid based macrocycles. However, in con-
trast to pyridine diamides, the isophthalamides are not pre-
organized toward convergent binding. Thus, it was interest-
ing to find out which effect would prevail and which family
of macrocyclic receptors would show higher anion affinity.
Here we describe the synthesis and anion-binding properties
of macrocyclic isophthalamide receptors 5–8. Particular at-
tention was paid to conformational preferences of these re-
ceptors and their complexes and to anion-induced confor-
mational changes they undergo. Furthermore, these new re-
ceptors proved be good models for studying competition be-
tween intramolecular hydrogen bonding and anion binding.
Synthesis
Low solubility notoriously hampers both the synthesis and
study of macrocyclic amides. We already faced these prob-
lems during our previous work with pyridine derivatives.
Thus, we decided to introduce tert-butyl groups into the
structures to enhance their solubility in organic solvents,
taking advantage of the commercial availability of 5-tert-bu-
tylisophthalic acid. All the desired macrocyclic tetraamides
5–8 were synthesized by one-step [2+2] condensation of 5-
tert-butylisophthalic acid dichloride with an appropriate
a,w-diamine under high-dilution conditions (Scheme 1).[14]
Despite the presence of the
isophthalamides may be attributed to additional favorable
interaction with aromatic CH protons at the 2-position. This
second explanation was recently supported by a theoretical
study of Bryantsev and Hay,[13] who convincingly demon-
strated that even moderately acidic aromatic CH groups in
the receptor structure could considerably enhance its anion-
solubilising groups, the products
derived from 1,2-diaminoethane
turned out to be hardly soluble
in most solvents, including
DMSO, and therefore only a
small sample of 5 could be ob-
tained for analytical purposes.
The solubility of the products
of the remaining reactions, al-
though still low, allowed for
their separation, characteriza-
tion and evaluation of their
anion-binding properties. The
yields of the macrocyclic tetraa-
mides were moderate to low. In
no case were we able to isolate
the products of [1+1] condensa-
tion, the macrocyclic diamides,
although analogous products
were obtained earlier in similar
reactions.[15–18] However, consid-
erable amounts of larger mac-
rocyclic products were formed:
hexaamides, octaamides, and
even traces of decaamides and
dodecaamides. These large
macrocycles are also potentially
interesting as anion receptors;
Scheme 1. Synthesis of macrocyclic amides.
for example, a hexaamide simi-
Chem. Eur. J. 2006, 12, 7652– 7667
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7653

J. Jurczak et al.
À
lar to 9 having the same ring size is able to bind CaCl3 ions
The stability constants collected in Table 1 are averaged
values obtained from nonlinear fitting of a 1:1 binding
model to the shifts of NH and CHint protons. The low solu-
bility of receptor 5 in DMSO precluded the determination
of its stability constants.
Contrary to our initial assumptions, the isophthalamides
turned out to be weaker anionophores than the pyridine dia-
mides. Similar observations were made by Bowman-James
et al.[24] on the 24-membered macrocyclic receptors 15 and
16 and were ascribed to worse preorganization of isophtha-
lamide receptors. As will be shown below, there is also an-
other and probably more serious reason: competition from
intramolecular hydrogen bonds in the isophthalamide mac-
rocycles.
in the solid state,[19] whereas octaamides have been shown
by us[20] to bind two chloride ions simultaneously. For octaa-
mides 12, 13 and 14, there is uncertainty concerning their
structures: are they single macrocycles or catenanes built
from two tetraamide rings? The macrocyclic structure of 12
was unambiguously confirmed by single-crystal X-ray analy-
sis of its chloride complex, which also confirmed its ability
to bind two ClÀ ions simultaneously. This extraordinary phe-
nomenon will be described in detail elsewhere. The identity
of octaamides 13 and 14 was confirmed on the basis of frag-
mentation patterns in their mass spectra.
In summary, the reaction is unselective and the product
ratio resembles what might have been expected from the
statistical mixture. This is in contrast to what was found in
many condensations of isophthalic acid dichloride with
longer and more rigid amines, where specific hydrogen-
bonding interactions involving linear intermediates directed
the course of the reaction and led to the preferential forma-
tion of, for example, catenanes or rotaxanes.[21,22,33,34] The
above-described method of synthesis of the macrocyclic tet-
raamides suffers from low yields and problems with separa-
tion of the desired tetraamides from the larger macrocyclic
products. However, these difficulties are compensated by
the fact that this is a one-pot synthesis and that each such
reaction affords at least three interesting macrocycles that
are potential anion receptors, that is, tetraamides (6: 31.9%,
7: 15.0%, 8: 11.4%), hexaamides (9: 24.6%, 11: 1.8%) and
octaamides (12: 15.7%, 14: 2.8%). It is therefore good
enough for our exploratory studies.
Previously, we found a peak affinity of the pyridine recep-
tors as a function of ring size: the 20-membered receptor
was better than either its smaller or larger homologues. In
the present isophthalamide receptors, a similar maximum of
affinity could not be demonstrated, due to the lack of data
for the smallest, 18-membered macrocycle. However, the
right-hand part of the size-affinity curve is now supplement-
ed by the 22-membered receptor, which was unavailable in
the pyridine series. Monotonic decrease of the binding affin-
ity with increasing macrocycle size is clearly observed for all
anions tested, irrespective of their size. Thus, the size match
between an anion and the cavity of the receptor does not
determine the observed selectivity. For instance, the cavities
of the largest, 24-membered receptors 4 and 8 match the
size of phosphate or carboxylate anions better than recep-
tors 2 and 6,[25] but their respective complexes are weaker.
Apparently, anion binding abilities of our receptors result
from intrinsic properties inherent to their structure. This is
reminiscent of what is well-known for simple crown ethers
and alkali metal cations: no matter what the cation diameter
is, it is better complexed by [18]crown-6 than by either
smaller or larger crowns.[26]
Anion binding in solution: The stability constants of recep-
tors 6, 7 and 8 with various anions were determined by
1H NMR titration[23] in [D6]DMSO (Table 1). The addition
of anions in the form of tetrabutylammonium (TBA) salts to
the solutions of each receptor caused significant downfield
shifts of the amide NH and internal aromatic CHint signals
in the H NMR spectra that indicated fast equilibrium be-
tween complexed and free ligands and also suggested anion
encapsulation inside the macrocyclic cavity by N H···A
and C H···A hydrogen-bonding interactions. Typical titra-
1
À
À
À
À
tion curves are shown in the Supporting Information.
Table 1. Stability constants [mÀ1] for the formation of 1:1 complexes of 6,
7 and 8 with various anions in [D6]DMSO at 298 K.[a] Values for recep-
tors 2 and 4[9] are listed for comparison.
Anion
6
7
8
2
4
ClÀ
378
20
296
20
302
552
573
–
48
6.5
82
1930
150
2283
3240
7410
75
18
<5
30
310
450
<5
BrÀ
The selectivity trends displayed by all three receptors 6–8
are similar and common for hydrogen-bonding receptors:
AcOÀ >ClÀ >BrÀ, AcOÀ >PhCOOÀ. Although it is tempting
to link this order to the basicity of anions, such correlation
may be valid only within the limited data set and may break
when more basic anions such as p-nitrophenolate or cyanide
are taken into account.[27] Remarkably, chloride anion forms
more stable complexes than might be expected based on its
PhCOOÀ
AcOÀ
H2PO4
601
3130
205
À
[b,c]
[b]
–
–
À
HSO4
<5
–
[a] Errors are estimated to be <10%. Tetrabutylammonium salts were
used as the anion sources. [b] The data do not fit satisfactorily to a
simple 1:1 binding model. [c] K=1320 was obtained in DMSO+5%
H2O.[10]
7654
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7652– 7667

Anion Binding versus Intramolecular Hydrogen Bonding in Neutral Macrocyclic Amides
FULL PAPER
very weak basicity; for example, it binds to 7 as strongly as
the much more basic benzoate anion.
contrast to these results, Malone et al.[29] have shown that
simple N,N’-diphenylisophthalamide crystallizes in a syn–syn
conformation. This is also the most stable conformation ac-
cording to AM1 calculations, although it is only 3 kJmolÀ1
more stable than the syn–anti conformation. In the context
of the present studies, we were interested in the conforma-
tional preferences of aliphatic isophthalamides, and there-
fore we performed ab initio calculations on the simple
model compound N,N’-dimethylisophthalamide.
Structural studies: To gain some insight into the structural
determinants of the binding abilities observed above, we un-
dertook systematic studies on the conformational preferen-
ces of isophthalamide receptors, starting with the simple
building blocks.
Building blocks: 2,6-Pyridine diamides are well known for
their strong preference for syn–syn conformation of both
amide NH groups, stemming from hydrogen bonding with
the pyridine nitrogen atom and lone-pair repulsion in the al-
ternative conformations (Figure 1).[28,29] As a result of these
interactions, the whole 2,6-bis(carbamoyl)pyridine fragment
is rigid, planar, exhibits a well-defined U-shape, and displays
In crystal structures, the amide groups are usually tilted
with respect to the adjacent benzene rings. Therefore, two
starting geometries were set for each of the three basic con-
formations depicted on Figure 1: one with both amide hy-
drogen atoms on the same side of the central ring, denoted
as (+,+), and the other with amide hydrogen atoms on op-
posite sides, labelled (+,À). The resulting six starting geo-
metries were optimized at both B3LYP/6-31G+(d,p) and
HF/6-31G+(d,p) levels of theory. For comparison, much
simpler semiempirical AM1 calculations were also per-
formed. The results are collected in Table 2. On optimisa-
two hydrogen-bond donating groups in
manner.
a convergent
tion, the syn–anti
(+,À) one, so there are only five entries in the table.
The results obtained by both ab initio methods are simi-
A
AHCTREUNG
lar. As anticipated, the amide groups are not coplanar with
the benzene ring and the respective torsion angles vary be-
tween 20 and 308, depending on the particular conformation
and the method applied.[31] This twist is a compromise be-
tween the tendency of the molecule to maximize amide–
phenyl conjugation and to minimize steric repulsion be-
tween the NH hydrogen atom and the ortho CH group of
the aromatic ring. The mutual arrangement of both amide
groups in the molecule is apparently governed by the inter-
action of their local dipole moments: in all cases, the (+,À)
conformation is more stable than the (+,+) one, and, more
generally, the energy of each conformation rises monotoni-
cally with increasing overall dipole moment of the molecule
Figure 1. Conformations of isophthalamide amide groups.
On the contrary, the isophthalamide moiety, lacking such
ordering interactions, is often perceived as having prefer-
ence towards syn–anti conformation of its amide groups
(Figure 1).[30] This conformation does not allow two conver-
gent hydrogen bonds to be formed and is therefore unsuita-
ble for anion binding. Such a view was gained from numer-
ous crystal structures of syn–anti isophthalamides and rein-
forced by calculations of Hunter and Purvis.[28] However,
these calculations explored the rather special case of 2,6-di-
methylaniline derivatives in which each aromatic substituent
on nitrogen is perpendicular to the plane of amide group. In
(Figure 2a). As a result, the syn–anti
A
the most stable according to both ab initio methods. Such a
preference is disadvantageous with respect to anion binding,
but, fortunately, it is not very strong: only 3–4 kJmolÀ1 rela-
tive to the syn–syn
A
syn–syn
syn–syn
ACHTREUNG
AHCTREUNG
amide groups of syn–syn
ACHTREUNG
Table 2. Relative energies [kJmolÀ1], dipole moments [D] and torsion angles between the amide groups and phenyl ring [8] of five different conforma-
tions of dimethyl isophthalamide as calculated at the B3LYP, HF and AM1 levels of theory.[a]
B3LYP/6-31G+(d,p)//6–311G++(3df,3pd)
HF/6-31G+(d,p)//6–311G++(3df,3pd)
AM1
Torsion
angles [8]
Relative
energy
Torsion
angles [8]
Dipole
moment
Relative
energy
Torsion
angles [8]
Dipole
moment
Relative
energy
Dipole
moment
[D]
[kJmolÀ1
]
A
[kJmolÀ1
]
A
A
syn–anti
syn–syn
(+,À)
syn–syn(+,+)
A
0
158.7, 20.2
27.6, 27.6
24.2, À24.2
152.4, 152.4
155.5, À155.5
3.32
3.77
5.21
5.89
7.00
0
155.5, 24.6
29.1, 29.1
27.0, À27.0
150.2, 150.2
À153.3
3.34
3.44
5.28
6.06
7.37
0
143.5, 37.4
41.4, 41.4
38.4, À38.5
137.2, 137.2
140.9, À140.9
2.70
2.04
4.77
4.80
6.72
G
3.77
6.30
7.07
8.39
3.16
6.09
6.70
1.34
3.79
3.52
5.89
G
anti–anti
anti–anti
ACHTREUNG
A
8.32153.4,
[a] Zero-point energy and thermal corrections have not been applied.
Chem. Eur. J. 2006, 12, 7652– 7667
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7655

J. Jurczak et al.
method to study the conformational preferences of recep-
tors 5–8, in order to investigate whether the macrocyclic
topology is able to preorganize the isophthalamide moieties
towards anion binding.
Macrocycles: An extensive conformational search shows a
huge variety of structures available for macrocyclic tetraa-
mides 5–8 within an energy span of 50 kJmolÀ1. Here we
discuss only selected conformations which are relevant to
experimental results. Further data are tabulated in the Sup-
porting Information.
Even the smallest, 18-membered macrocycle 5 is flexible
enough that its isophthalamide moieties can be either syn–
syn, syn–anti, or anti–anti. In the largest, 24-membered mac-
rocycle 8, both isophthalamide units can adopt all possible
arrangements almost independently of each other. Many of
the calculated conformations are stabilized by one or two in-
tramolecular hydrogen bonds.
However, the most stable conformations of each macrocy-
cle are very similar: both isophthalamide units adopt syn–
anti
A
NH···Oamide hydrogen bonds (Figure 3a and b). These two in-
tramolecular hydrogen bonds close the macrocyclic cavity so
that the receptor in such a conformation can form only
weak, external complexes with anions through single hydro-
gen bonds. This is why the macrocyclic isophthalamide re-
ceptors form weaker anion complexes than their pyridine
analogues: the energy gain from the formation of four
strong intermolecular hydrogen bonds is partially lost for
breaking of two strong intramolecular hydrogen bonds.
There are two major variants of these doubly syn–anti (or
SASA for short) conformations: U-shaped and roughly
planar (Figure 3a and b). Depending on the macrocycle,
either the former or the latter is preferred. Interestingly,
both possibilities were found in the crystal structure of free
ligand 6 (vide infra).
Figure 2. Correlation between dipole moment and energy of a) the five
conformers of dimethyl isophthalamide (squares: calculated using HF
method, diamonds: calculated using B3LYP method) and b) conformers
of 5 calculated using the AM1 method.
The SASA conformation is not the only one that allows
for the simultaneous formation of two intramolecular hydro-
gen bonds. The all-anti AA
N
N
different acceptors, as exemplified by the 2:2 complex with
fluoride anions observed in the solid state.[32] However, even
this larger energy difference is still rather small compared to
the intermolecular hydrogen-bonding interactions. Thus, the
isophthalamide moiety may be depicted as rigid but readily
adaptable. This paradoxical characteristic is the reason for
many fascinating phenomena, such as high-yield one-pot
syntheses of rotaxanes, catenanes and knotanes. The iso-
phthalamide fragments play different roles at various stages
of these syntheses by switching between their three basic
conformations to satisfy the needs of interacting part-
ners.[33,34]
Qualitatively, the semiempirical AM1 method gave results
similar to the previous higher level ab initio methods. The
method is very fast and gives a qualitatively acceptable de-
scription of hydrogen bonding. Thus, it can be applied to
modelling large systems dominated by strong hydrogen-
bonding interactions.[35] Hence, we employed the AM1
is an example, and is only 9.9 kJmolÀ1 less stable than 6.1.
Astonishingly, even the smallest, 18-membered macrocycle 5
can exist as an all-anti, doubly hydrogen bonded conformer
(AA
N
N
mum, see Supporting Information).
In the context of anion binding, the all-syn conformations
are most interesting. As they are not stabilized by intramo-
lecular hydrogen bonds,[36] they are ranked significantly
higher in energy than the most stable conformers:
À1
21 kJmolÀ1 for 5, 2 7 kJmol for 6, 13 kJmolÀ1 for 7 and
19 kJmolÀ1 for 8. The relatively low energy of the SSSS con-
formation of receptor 7 is noteworthy. As will be shown
below, this is the preferred conformation of this ligand both
in solution and in the solid state. Importantly, although the
above-mentioned structures must be formally described as
SSSS, they are not truly convergent: their NH bonds are not
oriented in the same direction (e.g., compare Figure 4a and
b).
7656
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7652– 7667

Anion Binding versus Intramolecular Hydrogen Bonding in Neutral Macrocyclic Amides
FULL PAPER
Figure 4. Two SS
N
N
AHCTREUNG
the same acceptor in the gas phase. The respective values
are probably lower in DMSO solution, because the two in-
tramolecular hydrogen bonds to be broken are weakened by
this highly polar solvent with pronounced hydrogen-bond
accepting ability.
The stabilizing hydrogen-bonding interactions together
with the destabilizing steric repulsions seem to be the most
important factors influencing the relative energies of the
conformers. The effect of remote dipole–dipole interactions,
observed in a model diamide, is completely masked here
(Figure 2b).
The number of unique structures grows very quickly with
increasing size of the macrocyclic ring (see Supporting Infor-
mation), due to rapid increase in flexibility of the aliphatic
linkers. Thus, although even tight macrocyclic confinement
is not able to force our tetraamide receptors into the conver-
gent, all-syn conformation, the smaller macrocycles 5 and 6
have a significantly reduced conformational space in com-
parison to the larger macrocycles 7 and 8, and therefore pre-
sumably pay a smaller entropic penalty due to anion bind-
ing. That is probably why the stability constants decrease
monotonically with increasing size of the macrocycle irre-
spective of the anion tested.
X-raystudies on free ligands : The X-ray crystal structures
of receptors 6, 7, and 8 were determined. Two different sol-
vates 6·0.5H2O and 6·CH3OH were obtained by slow diffu-
sion of pentane or diethyl ether, respectively, into a solution
of 6 in a CH2Cl2/CH3OH (1/1).
The quality of the first structure is poor due to severe dis-
order. Nevertheless, it is apparent that the isophthalic moiet-
ies are in a syn–anti conformation, similar to the lowest
energy structure 6.1 (Figure 5). However, there are no intra-
molecular hydrogen bonds in the crystal structure. Instead,
Figure 3. Geometries and relative energies of five selected conformers of
20-membered macrocycle 6 calculated using the AM1 method. Numbers
in parentheses refer to conformation numbers from Supporting Informa-
tion.
À
four intermolecular N H···O=C hydrogen bonds link two
Truly convergent conformations lie even higher on the en-
ergetic scale: 31 kJmolÀ1 for 5, 34 kJmolÀ1 for 6, 2 1 kJmoÀl 1
for 7 and 29 kJmolÀ1 for 8. This second set of values esti-
mates the energetic cost of the conformational change that
each ligand must undergo to form four hydrogen bonds with
neighbouring, symmetrically equivalent molecules to form a
dimer. Due to the flexibility of the aliphatic linkers, the
macrocycle adopts a highly bent, U-shaped conformation,
characterized by a very acute angle between its two phenyl
rings (ca. 308).
Chem. Eur. J. 2006, 12, 7652– 7667
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7657

J. Jurczak et al.
Figure 6. ORTEP view of the molecular structure of 6·CH3OH. a) Top
view, atom numbering is limited to symmetrically independent part.
b) Side view. (Displacement ellipsoids are scaled to the 30% probability
level. Dashed lines indicate hydrogen-bonding interactions.)
Figure 5. ORTEP view of the X-ray crystal structure of 6·0.5H2O. Part of
the disorder has been removed for clarity. Displacement ellipsoids are
scaled to the 50% probability level. Dashed lines indicate hydrogen-
bonding interactions.
In the second structure, the free ligand 6 adopts a roughly
planar, but highly twisted conformation similar to 6.2
(Figure 6). Both symmetrically equivalent isophthalic moiet-
ies are in a syn–anti(+,+) conformation and bind each other
G
in a complementary fashion by two NH···Oamide hydrogen
À
bonds. The externally directed N H and C=O groups form
four intermolecular hydrogen bonds with the neighbouring
macrocycles from the crystal lattice.
The syn–anti isophthalamide moieties in the above two
crystal structures are self-complementary and bind each
other by strong hydrogen bonds. It might therefore be ex-
pected that such a dimeric motif will also be found in the
crystal structures of other macrocyclic isophthalamides, as it
exists in their most stable conformations. However, 22-mem-
bered receptor 7 prefers an open conformation in the solid
state, with both isophthalamide halves in a convergent, syn–
Figure 7. ORTEP view of the X-ray crystal structure of 7. a Top view; car-
bonyl oxygen atoms of neighbouring macrocycles are shown inside the
cavity. b) Side view. (Displacement ellipsoids are scaled to the 50% prob-
ability level. Dashed lines indicate hydrogen-bonding interactions. Dis-
tances are given in angstrom.)
to the crystal packing very similar to that previously ob-
served in the crystal structure of its pyridine congener.[37]
X-ray-quality single crystals of 24-membered tetraamide 8
were grown by slow diffusion of diethyl ether into a solution
of the ligand in CH2Cl2/CH3OH (1/1). The structure of 8·
(C2H5)2O closely resembles that of 6·0.5H2O. The isophthal-
syn(+,+) conformation (Figure 7). Although such a confor-
A
mation is energetically unfavorable, as revealed by calcula-
tions, it allows both amide groups of the same isophthala-
mide moiety to bind to the same acceptor, that is, the car-
bonyl oxygen atom of the neighbouring molecule. This leads
7658
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7652– 7667

Anion Binding versus Intramolecular Hydrogen Bonding in Neutral Macrocyclic Amides
FULL PAPER
A
Owing to the many strong intermolecular interactions
present in the above crystal structures, the conformations
described above may be unrepresentative for the solution
structures. Nevertheless, the solution conformations of re-
ceptors 6 and 7 in DMSO are similar to those observed in
the solid state, as revealed by 2D NOESY experiments. A
nuclear Overhauser effect between the amide NH protons
and both internal and external CH protons of the benzene
ring (see Supporting Information) observed for solutions of
6 both in DMSO and DMF indicates considerable fractions
of both syn- and anti-oriented amide hydrogen atoms in an
averaged conformation.
(Figure 8). Two U-shaped macrocycles dimerize by means of
four intermolecular N H···O=C hydrogen bonds. However,
both the angle (438) and the distance between the opposite
phenyl rings in 8·(C2H5)2O are larger than in 6, and the re-
sulting empty space is filled with a solvent molecule (diethyl
ether).
The apparent symmetry observed in the 1D NMR spectra
must therefore be due to rapid interconversion between the
equivalent syn–anti conformations (Scheme 2). This dynamic
process is fast on the NMR timescale even at À708C (in
[D7]DMF).
Scheme 2. Fast conformational equilibrium between two equivalent struc-
1
tures of 6 leading to the high symmetry in the H NMR spectra of the re-
ceptor.
In contrast to 6, cross-peaks between amide NH protons
and only internal CH benzene protons are visible for recep-
tor 7 (see Supporting Information). This suggests that the
all-syn conformation of 7, similar to that in the crystal struc-
ture, also predominates in solution. According to AM1 cal-
culations, the all-syn conformation of 7 is the second most
stable and lies only about 13 kJmolÀ1 above the SASA
structure. Apparently, stabilization due to intramolecular hy-
drogen bonds is less effective in this compound, possibly
due to an uncomfortable arrangement of the aliphatic link-
ers.
We were unable to obtain good quality NOESY spectra
of 8 in [D6]DMSO, but the similarity of the spectrum re-
corded in [D7]DMF to that of 6 indicates the syn–anti con-
formation, in agreement with the crystal structure.
X-raystudies on anion complexes : To gain insight into the
binding modes and coordination geometries, X-ray analysis
of crystals of several anion complexes of 6 and 8 was per-
formed.
Figure 8. ORTEP view of the X-ray crystal structure of 8·(C2H5)2O.
A
a) Top view. b) Side view. (Part of the disorder has been removed for
clarity. Displacement ellipsoids are scaled to the 50% probability level.
Dashed lines indicate hydrogen-bonding interactions. The distance is
given in angstroms.)
Complexes of 20-membered ligand 6: Crystals of the chlor-
ide complex 6-PPh4Cl suitable for X-ray analysis were
Chem. Eur. J. 2006, 12, 7652– 7667
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7659

J. Jurczak et al.
grown by vapour diffusion of pentane into a solution of 6
and PPh4Cl in 1,2-dichloroethane.
significance of the geometric complementarity in selectivity
displayed by the receptor. The whole structure is very simi-
lar to that of the chloride complex of receptor 2, a pyridine
analogue of 6.
In the crystal structure, the chloride anion resides inside
the macrocyclic cavity of the receptor 6, docked by six hy-
drogen bonds, four of which are donated by amide NH
groups and two by internal aromatic CH protons (Figure 9).
Thus, the receptor sacrifices its intramolecular hydrogen
bonds to adopt an SSSS conformation and to maximize the
interactions with the anion. In contrast to both structures of
the free ligand 6, and also to most known structures of ben-
zamides and isophthalamides, the amide groups are almost
coplanar with the adjacent benzene rings. A side view (Fig-
ure 9b) reveals that the chloride ion is only shallowly insert-
ed into the cavity of the V-shaped receptor (ca. 1.3 Š above
the mean plane of the amide hydrogen atoms). The fact that
even the small chloride ion is not truly surrounded by the
macrocycle is probably responsible for a rather secondary
Vapor diffusion of diethyl ether into a solution of 6 and
tetrabutylammonium fluoride (TBAF) trihydrate in 1,2-di-
chloroethane resulted in a crystalline complex of unexpected
stoichiometry 6·TBAPF6·TBAF·C2H4Cl2·1.8H2O, instead of
the fluoride complex of 6 (Figure 10). The hexafluorophos-
phate ions were not added deliberately, but were probably
present as an impurity in the commercial fluoride.[38] Both
À
PF6 and FÀ ions are present in the crystal structure. Where-
as the fluoride ion, due to its high charge density, is well
known for its ability to form very strong hydrogen bonds,
À
PF6 is an archetypal example of weakly interacting anion.
À
Surprisingly, however, it is the PF6 ion that is complexed by
the receptor. What is even more astonishing is that the
À
water molecules present in the structure also prefer the PF6
ion. In fact, FÀ forms only C H···F contacts, of which the
À
À
shortest is that with the N+CH2 methylene group of the
TBA+ ion (2.31 Š). It is therefore worthwhile to look closer
À
at the geometry of binding between the PF6 ion and 6. The
shape of the receptor molecule is very similar to that in the
chloride complex discussed above. The place of the chloride
ion is taken by axial fluoride F1, which is docked inside the
cavity by hydrogen bonds with all six donors. Unlike the
chloride complex, the F1 atom does not occupy the central
À
position; it is considerably shifted towards N7 and N11: H7
À
À
À
F1 2.06, H11 F1 2.14, H1 F1 2.76, H17 F1 2.93 Š. This
shift enables two vertical fluoride atoms F4 and F5 to ap-
proach closer to N1 and N17 on the opposite side of the
À
À
macrocycle (H1 F4 2.06, H17 F5 2.09 Š). As a result, the
vertical axis of symmetry of the PF6 ion is tilted with re-
À
spect to the mean plane of the amide nitrogen atoms. In
total, the anion forms 12hydrogen-bonding contacts with
the receptor. Furthermore, it is additionally coordinated by
two water molecules and two tetrabutylammonium cations,
increasing the number of hydrogen bonds to 23. In contrast,
À
À
the fluoride anion forms only five C H···F hydrogen
bonds: three from the TBA+ ion (H48B F7 2.31, H41A F7
2.93, H68B F7 2.73 Š), one from the dichloroethane mole-
cule (H70B F7 2.55 Š) and one from the aliphatic linker of
the receptor (H9A F7 2.76 Š). Multipoint binding of the
hexafluorophosphate ion does not translate into high stabili-
ty of the complex in solution: the stability constant in
DMSO is too small to be determined by H NMR titration
À
À
À
À
À
1
(ca. 1mÀ1).
Complexes of 24-membered ligand 8: The 24-membered
macrocycle of 8 is large enough to encapsulate two oxygen
atoms of carboxylate, phosphate or sulfate anions.[25,39] Un-
fortunately, all attempts to obtain X-ray quality crystals of
these complexes have been unsuccessful so far. However,
the structures of two different chloride complexes of recep-
tor 8 were solved. Although the 24-membered receptor 8 is
clearly too large for a single chloride ion, the study of such
geometrically mismatched systems may give interesting in-
Figure 9. ORTEP view of the molecular structure of 6-Ph4PCl. a) Top
view. b) Side view. (Part of disorder in b) has been removed for clarity.
Displacement ellipsoids are scaled to the 50% probability level. Dashed
lines indicate hydrogen-bonding interactions. The distances are given in
angstroms.)
7660
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7652– 7667

Anion Binding versus Intramolecular Hydrogen Bonding in Neutral Macrocyclic Amides
FULL PAPER
À
Figure 10. ORTEP view of the X-ray crystal structure of 6·TBAPF6·TBAF·C2H4Cl2·1.8H2O. a) Top view. Binding mode of the PF6 ion in the cavity of 6.
b) Side view. An array of hydrogen bonds. (Displacement ellipsoids are scaled to the 50% probability level. Dashed lines indicate hydrogen-bonding in-
teractions.)
sight into means by which they adapt to maximize the
number of noncovalent interactions.[40]
The second chloride complex was obtained during our ef-
forts to isolate the benzoate complex of 8 by diffusion of di-
ethyl ether vapour into a dichloromethane solution contain-
ing 8 and tetrabutylammonium benzoate. The chloride
anions in the solution arose from the reaction of benzoate
anions with the solvent. X-ray structural analysis revealed
the presence of two guests inside the macrocyclic cavity: a
chloride anion and a water molecule (Figure 12). Each guest
The first complex was obtained by slow diffusion of hex-
anes into a solution of 8 and TBACl in 1,2-dichloroethane
(Figure 11). Unlike in the structure of the free ligand, both
isophthalamide moieties are in the syn–syn conformation.
To form reasonably short hydrogen bonds with the central
anion, these two diamide units approach and twist with re-
spect to each other. As a result, the amide protons form a
distorted tetrahedron around the guest. The chloride ion
forms two further hydrogen bonds with internal aromatic
CHint protons and additional contacts with methylene
groups of aliphatic linkers.
is chelated by one isophthalamide moiety. These two guests
À
À
are connected by an almost linear O H···Cl hydrogen bond
(O H···Cl angle 1648, H···ClÀ distance 2.41 Š). The coordi-
nation sphere of the chloride anion is completed from the
À
À
Figure 11. ORTEP view of the X-ray crystal structure of 8·TBA+ClÀ.
(Displacement ellipsoids are scaled to the 50% probability level. Dashed
lines indicate hydrogen-bonding interactions. The distances are given in
angstroms.)
Figure 12. ORTEP view of the X-ray crystal structure of 8·TBA+
ClÀ·2H2O; top view. (Displacement ellipsoids are scaled to the 50%
probability level. Dashed lines indicate hydrogen-bonding interactions.
The distances are is given in angstroms.)
Chem. Eur. J. 2006, 12, 7652– 7667
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7661

J. Jurczak et al.
top by two hydrogen bonds to a tetrabutylammonium
ed was the chloride complex of 6. These observations can be
interpreted by bearing in mind the above-discussed X-ray
structure of this complex, in which the central methylene
group is disordered over two positions: one close to the
anion, and the other bent to the opposite side. Apparently,
most complexes in solution prefer the first possibility,
whereas 6-Cl chooses the second. Unfortunately, in most
cases, these signals overlap with signals of tetrabutylammo-
nium ion.
À
À
À
À
cation: H52B Cl =2.82 Š, H41A Cl =2.81 Š.
Previously we reported a very similar structure of the
chloride complex of pyridine-based macrocyclic tetraamide
4, which has the same size as 8.[9] These two structures, to-
gether with the other known from the literature,[41,42] serve
to highlight the role of the water molecules that fill the gaps
between anionic guests and excessively large hosts. Consid-
ering the strong interaction of water with anions, it seems
reasonable to deliberately construct hosts that can accom-
modate the target anion along with a fraction of its hydra-
tion sphere. Such receptors may be useful in applications
such as the transport of anions through lipophilic channels
or membranes, where stripping off the solvation shell gives
rise to a high energy barrier.[43] Recently, a strong impulse
to such studies was given by Burns et al.,[44] who found that
anion complexation in a competitive solvent was enhanced
when a solvent molecule was incorporated into the binding
motif. This strategy is also widely used in Nature, where en-
zymes bind water for use as a part of their substrate-recog-
nition domains.[44] Receptors 4 and 8 may serve as prototype
anion receptors of this class.
All the above observations point to a conclusion that, also
in solution, the anionic guests are bound inside the macrocy-
clic cavities of the receptors 6–8. Further evidence concern-
ing the conformation of ligands in complexes comes from
1
the 2D NOESY spectra. In the H NMR spectra of fluoride
and acetate complexes of 6, all aromatic and amide signals
are clearly distinguishable, and therefore NOE could be ob-
served between amide NH protons and internal aromatic
CH protons (see Supporting Information). At the same
time, no effect was detected between the amide protons and
the external aromatic protons. These spectra indicate that
the syn–syn conformation of the isophthalic moieties in 6
largely predominates in the presence of anions. It is thus evi-
dent that anion-induced conformational change of receptor
6 takes place and, as a result, all amide NH protons point
towards the cavity (Scheme 3).
Structure of the anion complexes in solution: Complexation-
induced shifts in the 1H NMR spectra give valuable hints
about the structure of complexes in solution and the nature
of interactions between host and guest. The most striking
anion-induced changes in the spectra of 6–8, apart from the
expected downfield shifts of amide signals, are the large
downfield shifts of internal aromatic protons CHint, indica-
À
À
tive of C Hint···A hydrogen-bonding interactions. Noticea-
bly, in many cases these shifts are larger than those of the
amide NH protons. In the extreme case of 6 and the bro-
mide anion, the NH signals even move slightly upfield on
anion binding. This upfield shift can be explained by assum-
ing that the amide protons are already engaged in hydrogen
bonding (either intramolecular or with a solvent) before
anion binding.
Signals of external aromatic CH protons also moved
downfield in several titrations, albeit to a much lesser
extent. This is opposite to what might have been expected,
since anion binding increases the electron density of phenyl
rings, and this should result in upfield shifts of phenyl pro-
tons. This effect is probably due to the interaction of these
protons with oxygen atoms of the amide groups. Anion
binding forces isophthalamide moieties into syn–syn confor-
mations, in which both external aromatic protons are in
close proximity to the amide oxygen atoms. Furthermore,
the angle between the phenyl ring and amide groups de-
creases significantly on anion binding, as revealed by X-ray
analyses, and consequently the carbonyl oxygen atoms ap-
proach CH groups even more closely.
Scheme 3. Anion-induced conformational change taking place in the
DMSO solution of 6.
This observation is interesting in view of the fact that, in
Nature, conformational changes on substrate binding, ampli-
fied and propagated through space, are one of the most
common pathways of transduction of biological signals. Arti-
ficial systems based on this principle utilize mostly cation
coordination to trigger molecular shape changes.[45–49] An
analogous and complementary approach based on anion
binding has only recently been pioneered, so there is still a
very limited number of well-defined anion-triggered shape-
switching systems.[50–56] Some of them have already been ap-
plied for construction of fluorescent and chromogenic sen-
sors[51–55] or for controlled translocation of a macrocyclic
ring in a rotaxane.[56] Amides derived from 2,6-pyridinedicar-
boxylic acid have been used for the construction of switch-
ing systems because they exist as a single syn–syn conforma-
In most cases, slight deshielding of methylene hydrogen
atoms of aliphatic linkers was observed, which suggests their
close proximity to the anion and points to additional partici-
pation of these groups in anion binding. The only case in
which upfield shift of central methylene protons was record-
7662
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7652– 7667

Anion Binding versus Intramolecular Hydrogen Bonding in Neutral Macrocyclic Amides
FULL PAPER
tion and can be forced to adopt an anti–anti conformation
through cation binding[57,58] or protonation.[59,60] The second
method was used to reversibly switch oligoamides of this
type between linear and helical forms.[59,60] In contrast,
simple isophthalamides may exist as mixtures of the possible
conformers: syn–syn, syn–anti and anti–anti. Small energy
differences between these forms make it relatively easy to
amplify one of them by using an external factor. Such a
system, however, does not constitute a molecular switch,
due to undefined starting point. If two such units are incor-
porated within one molecule, the number of possibilities
rises dramatically. However, within an appropriate macrocy-
clic structure, like that of tetraamide 6, the self-complemen-
Scheme 4. Synthesis of the open-chain tetraamide 17.
tary character of the syn–anti conformation stabilizes one
particular structure to such an
extent that it dominates even in
Table 3. Binding constants [mÀ1] for the formation of 1:1 complexes of 6 and 17 with various anions in
a
highly competitive solvent
[D6]DMSO at 298 K.[a]
such as DMSO. We have dem-
onstrated here that anion bind-
ing provides sufficient energy
to break the two intramolecular
hydrogen bonds and switch
both isophthalic moieties into a
syn–syn conformation. It may
be speculated that other effec-
tors (cations or neutral mole-
cules) should be able to amplify
some other conformations from
the plethora of structures un-
covered by conformational
search.
Anion
Protons used for determination of K
6
17
Ddmax [ppm]
K [mÀ1
]
Ddmax [ppm]
K [mÀ1
14.0Æ0.3
]
ClÀ
amide (triplet)
“internal” arom.
amide (quartet)
amide (triplet)
“internal” arom.
amide (quartet)
amide (triplet)
“internal” arom.
amide (quartet)
amide (triplet)
“internal” arom.
amide (quartet)
371Æ13
385Æ8
–
22Æ2.1
0.18Æ0.001
0.63Æ0.006
Æ0.20.42 Æ0.004
0.55Æ0.00213.1
–
À0.03Æ0.001
Æ0..80022
–
8.4Æ0.3
2.3Æ0.20.39
Æ0.920.2
1.8Æ0.20.38
126Æ4
0.59Æ0.02
Æ0.03
BrÀ
19Æ0.20.38
Æ0.02
–
Æ0.02
AcOÀ
H2PO4
3202Æ179
3066Æ1320.43
–
0.55Æ0.003
Æ0.00271
–
1.04Æ0.01
Æ0.004
-
1.31Æ0.007
0.40Æ0.005
Æ0.04
Æ4
298Æ21.2
63Æ5
À
222Æ17
473Æ320.69
-
1.51Æ0.04
59Æ5
0.57Æ0.02
[b]
[b]
[a] Tetrabutylammonium salts were used as anion sources. [b] Value could not be determined due to severe
broadening of the signal.
Macrocyclic effect: Macrocyclic
topology clearly fails to preor-
ganize isophthalamide receptors towards anion binding.
Moreover, it can actually stabilize an array of intramolecular
hydrogen bonds that hamper anion binding. To probe the
macrocyclic effect in anion recognition by receptor 6, analo-
gous acyclic tetraamide 17 was prepared and titrated with
selected anions. Its one-pot synthesis consisted of the reac-
tion of 5-tert-butylisophthalic acid dichloride with half an
equivalent of 1,3-diaminopropane and subsequent quench-
ing of the reaction mixture with an excess of anhydrous
methylamine (Scheme 4).
the result of a simple 1:1 binding process.[62] For example, in
the previously studied acyclic tetraamide 18, in which both
aromatic halves are locked in a convergent syn–syn confor-
mation, we have found an excellent agreement between the
association constants determined from both amide groups in
the case of chloride and acetate, whereas large differences
were observed for dihydrogenphosphate, an anion which
very often gives complicated stoichiometries.
It is difficult to ascribe specific values to the stability con-
stants of receptor 17, because of considerable discrepancies
between the values obtained from different signals: the sta-
bility constants derived from amide groups joined by ali-
phatic linkers (triplet) are higher than those from internal
aromatic CH protons, which are in turn higher than the
values resulting from terminal amide protons (quartet;
Table 3). A similar phenomenon was observed by Gale et al.
for another acyclic tetraamide and interpreted as reflecting
the differences in the strength of interaction with the
anion.[61] However, discrepancies among the dissociation
constants determined by following the shift of different pro-
tons indicate that the observed titration curves may not be
Thus, despite a very good fit of a 1:1 binding model to the
titration data for 17, the values obtained should be inter-
preted with caution. Nevertheless, it is apparent that macro-
cyclic receptor 6 is much more effective than 17: even the
highest values obtained for 17 are still several times lower
than the respective values obtained for 6 (Table 3).
Chem. Eur. J. 2006, 12, 7652– 7667
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7663

J. Jurczak et al.
The influence of the position of
carbonyl groups: Thus, the mac-
rocyclic topology is crucial for
the high anion affinity of our
receptors, and the optimal mac-
rocycle size was estimated at
about 20. However, there are
still many possibilities for fur-
ther variation of the structure.
One of them consists of chang-
ing the positions of carbonyl
groups in the macrocycle. To
explore the potential of this ap-
proach, we studied receptor 21,
an isomer of 6. This compound,
derived from b-alanine, is an es-
pecially interesting model, be-
cause it would be relatively
easy to obtain its chiral ana-
Scheme 5. Synthesis of 21.
Table 4. Binding constants [mÀ1] for the formation of 1:1 complexes of 21
with various anions in [D6]DMSO at 298 K.[a]
logues by using enantiomerically pure b-amino acids.
Receptor 21 was obtained according to Scheme 5. Reac-
tion of 5-tert-butylisophthalic acid dichloride with b-alanine
gave ester 19, which was subjected to double amidation with
amine 20. Owing to the low reactivity of diester 19, sodium
methoxide was used as basic catalyst in this macrocycliza-
tion reaction.
Protons used for determi-
Anion
nation of K
This relatively small structural change resulted in a signifi-
cant deterioration of the anion-binding ability (Table 4). As
in 17, there are also two distinct kinds of amide groups in
this receptor. In this case, however, the association constants
determined from each pair are in reasonable agreement
(again with the exception of hydrogenphosphate), whereas
the asymptotic values of chemical shifts Ddmax differ signifi-
cantly. This is in agreement with common intuition: the
macrocyclic structure imposes severe constraints on the con-
formational space and, once a complex is formed, the anion
interacts with all hydrogen-bond donors inside the macrocy-
clic cavity. But these interactions are not equally strong: ar-
omatic amides are known to form stronger hydrogen bonds
than aliphatic amides,[63] and this is reflected by the dmax
values.
K [mÀ1
]
dmax [ppm]
ClÀ
amide “L”
amide “R”
29Æ1
27Æ20.18
27Æ1
0.54Æ0.005
Æ0.007
aromat. “internal L”
aromat. “internal R”
amide “L”
amide “R”
aromat. “internal L”
aromat. “internal R”
amide “L”
amide “R”
aromat. “internal L”
aromat. “internal R”
1.24Æ0.007
Æ0.006
528Æ20.2
160Æ7
AcOÀ
1.28Æ0.02
0.82Æ0.01
1.10Æ0.02
0.17Æ0.003
1.42Æ0.02
1.25Æ0.006
1.20Æ0.01
0.65Æ0.007
147Æ6
155Æ9
181Æ10
231Æ10
327Æ10
306Æ15
163Æ8
À
H2PO4
[a] Tetrabutylammonium salts were used as anion sources.
These observations suggest that the “right” (R) fragment
of macrocycle 21 is not a good building block for the con-
struction of strong anion receptors. However, the catalytic
activity does not always parallel strong substrate binding
and, on the other hand, the strength of interaction depends
strongly on the solvent. Thus, the applicability of such recep-
tors as organocatalysts in enantioselective synthesis is still to
be explored.
adopt numerous mutual arrangements and bind to each
other via one or two hydrogen bonds. In the case of tetraa-
mides 5–8, the most stable conformations are those in which
both isophthalamide moieties adopt a self-complementary
syn–anti conformation, form two strong intramolecular hy-
drogen bonds and thus close the macrocyclic cavity. Some of
these structures are stable enough that they can be observed
in condensed phases: in the crystal structures of 6 and 8,
and for 6 even in DMSO solution. On the other hand, 20-
membered macrocycle 7, for which the SASA conformation
results in particularly little energy gain, sacrifices it to maxi-
mize intermolecular contacts in the solid state and in
DMSO solution (with the solvent molecules). Anions are
sufficiently strong hydrogen-bond acceptors to force even
20-membered receptor 6 to switch to a convergent, all-syn
conformation. However, the energy required for such dra-
Conclusion
The isophthalamide moiety has only minor conformational
preferences and thus easily conforms to the needs of the in-
teracting partners. When two such moieties are incorporated
within a sufficiently flexible macrocyclic structure, they can
7664
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7652– 7667

Anion Binding versus Intramolecular Hydrogen Bonding in Neutral Macrocyclic Amides
FULL PAPER
150 mL) and than left for 24 h at room temperature. The solution was
then decanted and the precipitate was again washed with water in a simi-
lar manner and dried. The residue was purified by chromatography on
30 g of silica gel using 1 dm3 of 2% CH3OH in CH2Cl2, and then 2.5%
CH3OH in CH2Cl2 until macrocylic tetraamide 6 was completely eluted.
Further elution with 3% methanol gave hexaamide 9 and octaamide 12.
Tetraamide 6: Yield: 104 mg, 31.9%. M.p. 302–3048C. 1H NMR
(500 MHz, [D6]DMSO): d=8.58 (t, J=5.6 Hz, 4H), 8.27 (t, J=1.4 Hz,
2H), 7.89 (d, J=1.5 Hz, 4H), 3.47 (q, J=5.4 Hz, 8H), 1.89 (br, 4H),
1.26 ppm (s, 18H); 13C NMR (125 MHz, [D6]DMSO): d=165.81 (48),
150.63 (48), 133.88 (48), 126.6 (38), 123.16 (38), 39.15 (28), 34.592(4 8),
30.924 (18), 28.21 ppm (28); HRMS (ESI): m/z calcd for C30H40N4O4Na
[M+Na]+: 543.2942, found: 543.2947. Hexaamide 9: Yield: 64 mg,
matic conformational change lowers the anion-binding affin-
ity of isophthalamide receptors. That is why, although
simple isophthalamides are better anion receptors than pyri-
dinediamides, the contrary is true for the macrocyclic tetra-
amides. On the basis of these results, two further directions
of research may be suggested. First, the construction of
stronger anion receptors would lead to better conformation-
al control of isophthalamide building blocks and prevent
them from forming intramolecular hydrogen bonds. This can
be achieved in a number of ways, and the first, promising
step in this direction was reported by us recently.[10] The
second direction would focus on the anion-induced confor-
mational change and try to couple it to some signal-genera-
tion mechanism in order to construct supramolecular switch-
ing devices.
As in our previous studies on pyridine-based macrocyclic
tetraamides, the best anion affinities were obtained for the
20-membered macrocycle 6, irrespective of anion size. This
confirms a minor role of geometric complementarity in rec-
ognition by these receptors. One reason for this finding is
apparent from the crystal structures of anion complexes of
receptor 6: anions are not truly encapsulated by the recep-
tor; instead, they perch on the top of the A-shaped ligand.
The macrocyclic topology and the position of carbonyl
groups in the macrocycle are crucial for the effectiveness of
receptor 6, as revealed by the comparison of 6 with model
compounds 17 and 21. These structural elements should
remain unchanged in future modifications.
1
24.6%. M.p. 308–3108C. H NMR (200 MHz, [D6]DMSO): d=8.60 (t, J=
5.3 Hz, 6H), 8.09 (br, 3H), 7.97 (d, J=1.2 Hz, 6H), 3.42–3.22 (br, 12H),
1.81 (br, 6H), 1.34 Hz (s, 27H); 13C NMR (50 MHz, [D6]DMSO): d=
166.24 (48), 151.16 (48), 134.67 (48), 126.54 (38), 123.22 (38), 36.89 (28),
34.76 (48), 31.01 (18), 29.30 Hz (28). HRMS (ESI): m/z calcd for
C45H60N6O6Na [M+Na]+: 803.4467; found: 803.4493. Octaamide 12:
Yield: 51 mg, 15.7%. M.p. 198–2108C (decomp). 1H NMR (200 MHz,
[D6]DMSO): d=8.60 (t, J=5.3 Hz, 6H), 8.09 (br, 3H), 7.97 (d, J=
1.2 Hz, 6H), 3.42–3.22 (br, 12H), 1.81 (br, 6H), 1.34 Hz (s, 27H);
13C NMR (50 MHz, [D6]DMSO): d=166.24 (48), 151.16 (48), 134.67 (48),
126.54 (38), 123.22 (38), 36.89 (28), 34.76 (48), 31.01 (18), 29.30 ppm (28).
HRMS (ESI): m/z calcd for C60H80N8O8Na [M+Na]+: 1063.5991, found:
1063.6008.
Reaction of 5-tert-butylisophthalic acid chloride with 1,4-diaminobutane:
A solution of 5-tert-butylisophthalic acid chloride (2591 mg, 10.0 mmol)
in dry CH2Cl2 (20 mL) and a solution of 1,4-diaminobutane (881 mg,
10.0 mmol) in dry CH2Cl2 (20 mL) were slowly added by syringe pump
over several hours at the same rate to a mixture of triethylamine
(2.6 mL) and dry CH2Cl2 (250 mL). The next day, the mixture was fil-
tered and the filtrate evaporated. The dry residue was boiled with water
(ca. 150 mL) and then left for 24 h at room temperature. The solution
was then decanted and the precipitate was again washed with water in a
similar manner and dried. The residue was purified by chromatography
on 60 g of silica gel using 2–3% CH3OH in CH2Cl2 to yield tetraamide 7.
Further elution with 3% methanol gave a mixture of hexaamide 9 and
octaamide 12. We were unable to separate these two compounds due to
their very similar chromatographic properties. Hexaamide 7: Yield:
412mg, 15.0%. M.p. >3508C. 1H NMR (200 MHz, [D6]DMSO): d=8.37
(t, J=5.2Hz, 4H), 8.03 (br, 2H), 7.90 (d, J=1.1 Hz, 4H), 3.40–3.25 (br,
8H, overlapping with water), 1.57 (br, 8H), 1.32ppm (s, 18H); 13C NMR
(50 MHz, [D6]DMSO): d=166.34 (48), 151.03 (48), 134.89 (48), 126.34
(38), 123.12 (38), 38.59 (28), 34.64 (48), 30.96 (18), 26.45 ppm (28); HRMS
(ESI): m/z calcd for C32H44N4O4Na [M+Na]+: 571.3255, found: 571.3265.
Experimental Section
Instruments and methods: NMR spectra were measured on a Varian
Gemini 200 (1H: 200 MHz, 13C: 50 MHz) or a Bruker AM-500 (1H:
500 MHz, 13C: 125 MHz) spectrometer. The chemical shifts are given in
ppm relative to solvent signals as internal standards; the coupling con-
stants are in hertz. The ESI mass spectra were obtained on Mariner (ESI
TOF) and API 365 (ESI 3Q) mass spectrometers with methanol as a
spray solvent. The melting points are uncorrected.
Syntheses: All precursors for syntheses were obtained from Aldrich or
Fluka and were used as received. 5-tert-Butyl-isophthalic acid chloride
was obtained from 5-tert-butyl-isophthalic acid according to the literature
procedure.[64] TLC was carried out on Merck silica gel 60 F254 plates; for
column chromatography, Merck silica gel 60 (63–100 mm mesh size) was
used. The studied macrocycles gave unsatisfactory elemental analyses
due to solvent molecules encapsulated in the crystals. This phenomenon
is well known and, even after prolonged heating in high vacuum, the sol-
vents can still be detected by NMR spectroscopy. The isotope patterns
obtained by ESI-MS spectra are, however, consistent with those calculat-
ed on the basis of natural isotope abundances and thus confirm the ele-
mental composition.
Reaction of 5-tert-butylisophthalic acid chloride with 1,5-diaminopen-
tane: The reaction was carried out by the high-dilution technique as de-
scribed above. The mixture of products (free from triethylamine hydro-
chloride) was chromatographically separated. Tetraamide 8: Yield:
330 mg, 11.4%. M.p. >3508C. 1H NMR (200 MHz, [D6]DMSO): d=8.41
(t, J=5.2Hz, 4H), 8.05 (br 2H), 7.89 (d, J=1.2Hz, 4H), 3.40–3.25 (br,
8H) 1.55 (br, 8H), 1.42–1.04 (br, 4H), 1.27 ppm (s, 18H); 13C NMR
(50 MHz, [D6]DMSO): d=166.01 (48), 150.75 (48), 134.54 (48), 126.42
(38), 123.29 (38), 39.38 (28), 34.67 (48), 30.96 (18), 29.12 (28), 24.64 ppm
(28). HRMS (ESI): m/z calcd for C34H48N4O4Na [M+Na]+: 599.3568,
found: 599.3558. Hexaamide 11: Yield: 53 mg, 1.8%. M.p. 338–3428C
1
(with decomp). H NMR (200 MHz, [D6]DMSO): d=8.56 (brt, 6H), 8.10
Macrocyclic tetraamide 5: M.p. >3508C. 1H NMR (500 MHz,
[D6]DMSO): d=8.33 (br, 4H), 8.29 (br, 2H), 7.90 (d, J=1.5 Hz, 4H),
3.47 (br, 8H), 1.32(s, 18H); HRMS (ESI): m/z calcd for C28H36N4O4Na
[M+Na]+: 515.6133, found 515.6131.
(s, 3H), 7.93 (s, 6H), 3.28 (m, 12H, partially overlapping with water
signal), 1.58 (m, 12H), 1.46–1.20 (brm, 6H, overlapping with singlet at
1.31), 1.31 ppm (s, 27H); 13C NMR (50 MHz, [D6]DMSO): d=166.11
(48), 150.94 (48), 134.67 (48), 126.39 (38), 123.53 (38), 39.16 (28), 34.70
(48), 30.99 (18), 28.87 (28), 23.89 ppm (28). HRMS (ESI): m/z calcd for
C51H72N6O6Na [M+Na]+: 887.5406, found: 887.5437. Octaamide 14:
Yield: 80 mg, 2.8%. 1H NMR (200 MHz, [D6]DMSO): d=8.55 (t, J=
5.4 Hz, 8H), 8.09 (br, 4H), 7.92(d, J=1.0 Hz, 8H), 3.27 (br, 16H), 1.57
(br, 8H), 1.29 ppm (s, 36H); 13C NMR (50 MHz, [D6]DMSO): d=166.04
(48), 150.89 (48), 134.61 (48), 126.38 (38), 123.49 (38), 39.24 (28), 34.72
Reaction of 5-tert-butylisophthalic acid chloride with 1,3-diaminopro-
pane:
A solution of 5-tert-butylisophthalic acid chloride (324 mg,
1.25 mmol) in dry CH2Cl2 (1.5 mL) was added dropwise to a mixture of
1,3-diaminopropane (92mg, 1.24 mmol), triethylamine (0.5 mL) and dry
CH2Cl2 (250 mL) over 5 min. The next day, the mixture was filtered and
the filtrate evaporated. The dry residue was boiled with water (ca.
Chem. Eur. J. 2006, 12, 7652– 7667
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7665

J. Jurczak et al.
(48), 31.01 (18), 28.95 (28), 24.09 ppm (28); MS (ESI): m/z: 1175
[M+Na]+.
Acknowledgements
We are grateful to Dr. Edyta Małolepsza for help with computer model-
ing and Dr. Michał Barbasiewicz for sharing computing resources. We
also thank Dr. Agnieszka Szumna for valuable suggestions and Mr.
Łukasz Dobrzycki for X-ray analyses.
Acyclic tetraamide 15: A solution of 1,3-diaminopropane (5 mmol) and
triethylamine (10 mmol) in dry CH2Cl2 was slowly added by syringe
pump over 2h to a stirred solution of 5- tert-butylisophthalic acid chloride
(5183 mg, 20 mmol) in dry CH2Cl2 (150 mL). After stirring for about 1 h,
the reaction mixture was quenched with methylamine (3 mL, 66 mmol,
cooled to about À708C). After an additional 1 h of mixing, the precipi-
tate of methylamine hydrochloride was filtered off, and the filtrate
evaporated to dryness. The dry residue was boiled with water (ca.
150 mL) and then left for 24 h at room temperature. The water solution
was then decanted and the precipitate dissolved in CH2Cl2/CH3OH (1:1),
evaporated and again washed with water in a similar manner as above.
The mixture of products, free from triethylamine hydrochloride, was puri-
fied by chromatography on 200 g of silica gel using 1% CH3OH in
CH2Cl2 until the diamide was completely eluted. Then the desired tetra-
amide was washed with 3% CH3OH in CH2Cl2. Dimethylamide of 5-tert-
butylisophthalic acid: Yield: 2057 mg, 41.4%. M.p. 126–1278C. 1H NMR
(200 MHz, CDCl3): d=7.87 (br, 3H), 6.69 (br, 2H), 2.99 (d, J=3.8 Hz,
6H), 1.31 (s, 9H); HRMS (ESI): m/z calcd for C14H20N2O2Na [M+Na]+:
271.1417, observed 271.1426. Tetraamide 15: Yield: 271 mg, 5.3%.
1H NMR (200 MHz, [D6]DMSO): d=8.67 (t, J=5.5 Hz, 2H), 8.54 (q, J=
4.6 Hz, 2H), 8.13 (br, 2H), 7.97 (br, 2H), 3.28–3.45 (m, 4H), 2.79 (d, J=
4.4 Hz, 6H), 1.81 (brm, 2H), 1.33 ppm (s, 18H); 13C NMR (50 MHz,
[D6]DMSO): d=166.50, 166.12, 151.02, 134.54, 154.51, 126.36, 123.55,
37.21, 34.75, 31.00, 29.33, 26.26 ppm; HRMS (ESI): m/z calcd for
C29H40N4O4Na [M+Na]+: 531.2942, observed 531.2968.
[1] A. Bianchi, K. Bowman-James, E. Garcia-EspaÇa, Supramolecular
Chemistry of Anions, Wiley-VCH, Weinheim, 1997.
[2] A. Bianchi, K. Bowman-James, E. Garcia-EspaÇa, Supramolecular
Chemistry of Anions, Wiley-VCH, Weinheim, 1997, Chap. 3.
[3] J. D. Watson, J. Milner-White, J. Mol. Biol. 2002, 315, 171–182.
[4] C. R. Bondy, S. J. Loeb, Coord. Chem. Rev. 2003, 240, 77–99.
[5] P. A. Gale, Chem. Commun. 2005, 3761–3772.
[6] K. Bowman-James, Acc. Chem. Res. 2005, 38, 671–678.
[7] For a review on macrocyclic anion receptors based on hydrogen-
bonding interactions, see: K. Choi, A. D. Hamilton, Coord. Chem.
Rev. 2003, 240, 101–110.
[8] M. J. Chmielewski, A. Szumna, J. Jurczak, Tetahedron Lett. 2004, 45,
8699–8703.
[9] M. J. Chmielewski, J. Jurczak, Chem. Eur. J. 2005, 11, 6080–6094.
[10] M. J. Chmielewski, J. Jurczak, Tetahedron Lett. 2005, 46, 3085–3088.
[11] K. Kavallieratos, C. M. Bertao, R. H. Crabtree, J. Org. Chem. 1999,
64, 1675–1683.
[12] Similar behaviour of bis-amide receptors was observed earlier by
Hamilton et al. with respect to uncharged hydrogen-bond acceptors:
S.-K. Chang, D. Van Engen, E. Fan, A. D. Hamilton, J. Am. Chem.
Soc. 1991, 113, 7640–7645.
[13] V. S. Bryantsev, B. P. Hay, J. Am. Chem. Soc. 2005, 127, 8282–8283.
[14] A 20-membered macrocyclic tetraamide devoid of tert-butyl groups
was already synthesized from isophtahlic acid chloride and 1,3-dia-
minopropane; however, its anion-binding properties were not stud-
ied. L. Wambach, F. Voegtle, Tetrahedron Lett. 1985, 26, 1483–1486.
[15] From isophthalic acid chloride and 1,4-diaminobutane: H. Stetter, L.
Marx-Moll, H. Rutzen, Chem. Ber. 1958, 1775–1780.
Amidoester 16: The hydrochloride salt of b-alanine methyl ester
(2792 mg, 20.0 mmol) was suspended in dry CH2Cl2 (200 mL) and cooled
to about À558C. To this mixture, triethylamine (5.6 mL) was added drop-
wise to liberate the ester from its salt. Immediately after the addition was
finished, a solution of 5-tert-butyl-isophthalic acid chloride (2591 mg,
10.0 mmol) in dry CH2Cl2 (10 mL) was slowly added. The mixture was
left in a cooling bath and allowed to reach slowly room temperature.
Then it was washed twice with 10% NaHSO4 (2”50 mL) and once with
water and dried over MgSO4. The solvent was removed and the residue
was purified by chromatography on 150 g of silica gel using 3% CH3OH
in CH2Cl2 as an eluent. Yield: 2925 mg, 74.5%. M.p. 88–898C. 1H NMR
(200 MHz, CDCl3, TMS): d=7.94 (d, J=1.6 Hz, 2H), 7.89 (s, J=1.5 Hz,
1H), 7.03 (t, J=5.8 Hz, 2H), 3.80–3.64 (m + s, 10H), 2.67 (t, J=6.0 Hz,
4H), 1.34 ppm (s, 9H); 13C NMR (50 MHz, CDCl3, TMS): d=173.02
(48), 167.07 (48), 152.42 (48), 134.47 (48), 127.28 (38), 122.14 (38), 51.84
(18), 35.51 (28), 35.01 (48), 33.68 (28), 31.13 ppm (18); elemental analysis
calcd (%) for C20H28N2O6 (392.45): C 61.21, H 7.19, N 7.14; found: C
[16] From 2,6-pyridinedicarboxylic acid chloride and 1,4-diaminobutane:
K. Krakowiak, Pol. J. Chem. 1988, 62, 231–233.
[17] From 2,6-pyridinedicarboxylic acid chloride and 1,5-diaminopen-
tane: K. Krakowiak, Pol. J. Chem. 1986, 60, 277–281.
[18] From 1,2-diaminobenzene and both isophthalic acid chloride and
2,6-pyridinedicarboxylic acid chloride: Z. Mo, W. Yang, J. Gao, H.
Chen, J. Kang, Synth. Commun. 1999, 12, 2147–2153.
[19] Y. H. Kim, J. Calabrese, C. McEwen, J. Am. Chem. Soc. 1996, 118,
1545–1546.
[20] A. Szumna, J. Jurczak, Helv. Chim. Acta 2001, 84, 3760–3765.
[21] F. J. Carver, C. A. Hunter, R. J. Shannon, Chem. Commun. 1994,
1277–1280.
[22] A. G. Johnston, D. A. Leigh, L. Nezhat, J. P. Smart, M. D. Deegan,
Angew. Chem. 1995, 107, 1327–1331; Angew. Chem. Int. Ed. Engl.
1995, 34, 1212–1216.
[23] L. Fielding, Tetrahedron 2000, 56, 6151–6170.
[24] M. A. Hossain, S. U. Kang, D. Powell, K. Bowman-James, Inorg.
Chem. 2003, 42, 1397–1399.
61.19,
H 7.12, N 7.08; HRMS (ESI): m/z calcd for C20H28N2O6Na
[M+Na]+: 415.1840, found: 415.1852.
Tetraamide 18: Sodium (0.25 g) was dissolved in anhydrous methanol
(100 mL), and amido ester 16 (1962mg, 5.00 mmol) and diamine 17
(962mg, 5.00 mmol) were dissolved in this solution of sodium methano-
late. The reaction flask was tightly sealed and left at room temperature
for about one month. After this time, the solvent was evaporated and the
desired product was isolated by column chromatography on silica gel
using 3% CH3OH in CH2Cl2 as eluent. Yield 205 mg, 7.9%. M.p.
>3508C. 1H NMR (200 MHz, [D6]DMSO): d=8.32–8.18 (m, 4H), 7.93
(m, 3H), 7.14 (d, J=1.2Hz, 2H), 7.08 (bs, 1H), 4.23 (d, J=5.0 Hz, 4H),
3.55 (brm, 4H), 2.48–2.35 (brm, 4H), 1.30 (s, 9H), 1.16 ppm (s, 9H);
13C NMR (50 MHz, [D6]DMSO): d=170.45, 166.14, 151.00, 150.64,
138.64, 134.17, 126.96, 125.12, 123.83, 122.13, 43.06, 35.86, 35.40, 34.60,
34.20, 31.08, 30.96 ppm; elemental analysis (%) calcd for C30H40N4O4: C
69.20, H 7.74, N 10.76; found: C 69.10, H 7.81, N 10.88; HRMS (ESI):
m/z calcd for C30H40N4O4Na [M+Na]+: 543.2942, found: 543.2945.
´
[25] M. J. Chmielewski, Ł. Dobrzycki, J. Jurczak, K. Wozniak, Cryst.
Growth Des. 2005, 5, 1339–1341.
[26] G. W. Gokel, D. M. Goli, C. Minganti, L. Echegoyen, J. Am. Chem.
Soc. 1983, 105, 6786–6788.
[27] See reference [9] and references therein.
[28] C. A. Hunter, D. H. Purvis, Angew. Chem. 1992, 104, 779–782;
Angew. Chem. Int. Ed. Engl. 1992, 31, 792–795.
CCDC-246610 (6·CH3OH), CCDC-246611 (6·Ph4Cl), CCDC-283441
(6·0.5H2O), CCDC-283442 (6·TBAPF6·TBAF·C2H4Cl2·1.8H2O), CCDC-
283443 (7), CCDC-283444 (8·TBA+ClÀ), CCDC-283445 (8·TBA+
[29] J. F. Malone, C. M. Murray, G. M. Dolan, R. Docherty, A. J. Lavery,
Chem. Mater. 1997, 9, 2983–2989.
[30] For recent example, see: T. Velasco, G. Lecollinet, T. Ryan, A. P.
Davis, Org. Biomol. Chem. 2004, 2, 645–647.
ClÀ·2H2O) and CCDC-283446 (8·
(C2H5)2O) contain the supplementary
A
crystallographic data for this paper. These data can be obtained free of
charge from the Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[31] A similar twist of the amide group was calculated earlier for benz-
ˇ
ˇ
amides: M. Cajan, I. Stibor, J. Koca, J. Phys. Chem. A 1999, 103,
3778–3782.
7666
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 7652– 7667

Anion Binding versus Intramolecular Hydrogen Bonding in Neutral Macrocyclic Amides
FULL PAPER
[32] S. J. Coles, J. G. Frey, P. A. Gale, M. B. Hursthouse, M. E. Light, K.
Navakhun, G. L. Thomas, Chem. Commun. 2003, 568–569.
[33] C. A. Schalley, W. Reckien, S. Peyerimhoff, B. Baytekin, F. Vçgtle,
Chem. Eur. J. 2004, 10, 4777–4789.
[49] M. Bꢁschel, M. Helldobler, J. Daub, Chem. Commun. 2003, 1338–
1339.
[50] H. Huang, L. Mu, J. He, J.-P. Cheng, Tetrahedron Lett. 2002, 43,
2255–2258.
[34] D. A. Leigh, A. Venturini, A. J. Wilson, J. K. Y. Wong, F. Zerbetto,
Chem. Eur. J. 2004, 10, 4960–4969.
[51] S. B. Suh, C. Cui, H. S. Son, J. S. U, Y. Won, K. S. Kim, J. Phys.
Chem. B 2002, 106, 2061–2064.
[35] The AM1 method has been recently successfully applied to much
[52] J. L. J. Blanco, J. M. Benito, C. O. Mellet, J. M. G. Fernandez, Org.
Lett. 1999, 1, 1217–1220.
larger systems containing isophthalamide units.[33]
[36] The only exceptions are found in the largest macrocycle 8, for exam-
[53] R. Martꢂnez-MꢃÇez, F. Sancenꢄn, Chem. Rev. 2003, 103, 4419–4476.
[54] F.-Y. Wu, Z. Li, Z.-C. Wen, N. Zhou, Y.-F. Zhao, Y.-B. Jiang, Org.
Lett. 2002, 4, 3203–3205.
ple, SS
N
N
ing Information.
[37] A. Szumna, D. T. Gryko, J. Jurczak, Heterocycles 2002, 56, 361–368.
[38] For another example of a PF6 complex grown in the presence of an
excess of tetrabutylammonium fluoride, see reference [32].
[39] X-ray structure of the receptor 19–sulfate complex: M. A. Hossain,
J. M. Llinares, D. Powell, K. Bowman-James, Inorg. Chem. 2001, 40,
2936–2937.
[55] C. Demuth, O. Zerbe, D. Rognan, R. Sçll, A. Beck-Sickinger, G.
Folkers, U. E. Spichinger, Biosens. Bioelectron. 2001, 16, 783–789.
[56] C. M. Keaveney, D. A. Leigh, Angew. Chem. 2004, 116, 1242–1244;
Angew. Chem. Int. Ed. 2004, 43, 1222–1224.
À
´
[57] D. T. Gryko, A. Pecak, W. Koz´minski, P. Piatek, J. Jurczak, Supra-
mol. Chem. 2000, 12, 229–235.
[40] Sterically mismatching cation complexes of crown ethers: J. W.
Steed, P. C. Junk, J. Chem. Soc. Dalton Trans. 1999, 2141–2146.
[41] M. A. Hossain, J. M. Linares, S. Mason, P. Morehouse, D. Powell, K.
Bowman-James, Angew. Chem. 2002, 114, 2441–2444; Angew.
Chem. Int. Ed. 2002, 41, 2335–2338.
[42] J. L. Sessler, D. An, W.-S. Cho, V. Lynch, D.-W Yoon, S.-J. Hong, C.-
H. Lee, J. Org. Chem. 2005, 70, 1511–1517.
[43] O. Beckstein, K. Tai, M. S. P. Sansom, J. Am. Chem. Soc. 2004, 126,
14694–14695.
[58] A. Szumna, D. T. Gryko, J. Jurczak, J. Chem. Soc. Perkin Trans. 2
2000, 1553–1558.
[59] E. Kolomiets, V. Berl, I. Odriozola, A.-M. Stadler, N. Kyritsakas, J.-
M. Lehn, Chem. Commun. 2003, 2868–2869.
[60] C. Dolain, V. Maurizot, I. Huc, Angew. Chem. 2003, 115, 2844–
2846; Angew. Chem. Int. Ed. 2003, 42, 2738–3740.
[61] K. Navakhun, P. A. Gale, S. Camiolo, M. E. Light, M. B. Hursthouse,
Chem. Commun. 2002, 2084–2085.
[62] M. D. Covart, I. Sucholeiki, R. R. Bukownik, C. S. Wilcox, J. Am.
Chem. Soc. 1988, 110, 6204–6210.
[44] D. H. Burns, K. Calderon-Kawasaki, S. Kularatne, J. Org. Chem.
2005, 70, 2803–2807.
ˇ
ˇ
[63] I. Stibor, D. S. M. Hafeed, P. Lhotꢃk, J. Hodacovꢃ, J. Koca, M.
ˇ
[45] V. Balzani, A. Credi, J. F. Stoddart, Angew. Chem. 2000, 112, 3484–
3530; Angew. Chem. Int. Ed. 2000, 39, 3348–3391.
[46] J. F. Stoddart, Acc. Chem. Res. 2001, 34, 409–522.
[47] A. Petitjean, R. G. Khoury, N. Kyritsakas, J.-M. Lehn, J. Am. Chem.
Soc. 2004, 126, 6637–6647.
Cajan, Gazz. Chim. Ital. 1997, 127, 673–685.
[64] C. Heim, A. Affeld, M. Nieger, F. Vçgtle, Helv. Chim. Acta 1999,
82, 746–768.
[48] M. Barboiu, G. Vaughan, N. Kyritsakas, J.-M. Lehn, Chem. Eur. J.
2003, 9, 763–769.
Received: November 28, 2005
Published online: July 5, 2006
Chem. Eur. J. 2006, 12, 7652– 7667
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7667