Syntheses, structures, and anion-binding properties of two novel calix[2]benzo[4]pyrroles

Article abstract of DOI:10.1002/chem.200600452

Calix[2]benzo[4]pyrrole m-6 and p-6, each containing two dipyrromethane moieties and two m-phenylene or p-phenylene units, respectively, were readily synthesised from pyrrole, 1,3- and 1,4-bis(1,1′-dimethylhydroxymethyl) benzene, (m-4 and p-4, respectivel

Full text of DOI:10.1002/chem.200600452

FULL PAPER
DOI: 10.1002/chem.200600452
Syntheses, Structures, and Anion-BindingProperties of Two Novel
Calix[2]benzo[4]pyrroles
Grazia Cafeo,*[a] Franz H. Kohnke,*[a] Andrew J. P. White,[b] Domenico Garozzo,[c] and
Angela Messina[c]
Abstract: Calix[2]benzo[4]pyrrole m-6 and p-6, each containing two dipyrro-
methane moieties and two m-phenylene or p-phenylene units, respectively, were
readily synthesised from pyrrole, 1,3- and 1,4-bis(1,1’-dimethylhydroxymethyl)benzene,
(m-4 and p-4, respectively) and acetone.Macrocycles m-6 and p-6 were tested as
receptors for a selection of anions, such as acetate, dihydrogenphosphate and fluo-
ride.The X-ray structures of m-6 and p-6 and those of the complexes m-6·Fꢀ, m-
6·Clꢀ and m-6·CH3COOꢀ (with an nBu4N+ counterion) were also determined.
Keywords: anions · calixpyrroles ·
host–guest systems · macrocycles ·
structure elucidation
Introduction
with furan and/or thiophene proved to be less-effective
anion ligands overall than the “purely” pyrrole parent com-
pound, although they did show interesting and varied selec-
tivities.[5k] Similar results were reported for a comparison be-
tween calix[2]furan[4]pyrrole 2 and calix[6]pyrrole 3.[5h,e]
Calixpyrroles[1] are macrocycles made of pyrrole units linked
at their 2,5-positions with quaternary carbon-atom bridges.
Although structurally related to porphyrins, they have no
hydrogen atoms at their meso positions and, therefore, they
do not exhibit the typical chemistry of porphyrins.[2] Current
interest in calixpyrroles is mostly related to their ability to
act as molecular receptors for anions.[3] The initial discovery
that calix[4]pyrrole 1 is capable of binding anions by means
of multiple hydrogen bonds[1a] promoted considerable re-
search on the use of pyrroles as building blocks for the as-
sembly of both cyclic and acyclic anion-binding receptors,[4]
such as a number of “hybrid” and “expanded” (i.e., contain-
ing more than four aromatic units) calixpyrroles.[5] Calix[4]-
pyrroles in which some of the pyrrole units were replaced
However, benzo rings have the potential to form CH···anion
hydrogen bonds,[6] and their inclusion in pyrrole-containing
heterocalixarenes can be less detrimental to anion binding
than the inclusion of either furan or thiophene.Although a
study dealing with the replacement of some pyrrole rings
with benzo units has been reported for the calix[4] sys-
tems,[5k] to the best of our knowledge there are no data for
“hybrid” and “expanded” calixpyrroles containing benzo
units.Here we report our initial studies on two representa-
tives of these compounds.
[a] Dr.G.Cafeo, Prof.F.H.Kohnke
Dipartimento di Chimica Organica e Biologica
Università di Messina
Salita Sperone 31, 98166 Messina (Italy)
Fax : (+90)393-895
E-mail: ella@isengard.umime.it
franz@unime.it
[b] Dr.A.J.P.White
Chemical Crystallography Laboratory
Department of Chemistry, Imperial College London
South Kensington, London, SW7 2AZ (UK)
[c] Dr.D.Garozzo, Dr.A.Messina
CNR, ICTP Catania
Results and Discussion
Viale Regina Margherita 6, I-95125 Catania (Italy)
Calix[2]-m-benzo[4]pyrrole (m-6) containing two dipyrro-
Supporting information for this article is available on the WWW
under http://www.wiley-vch.de/home/chemistry/ or from the author.
methane moieties and two m-phenylene units and its ana-
Chem. Eur. J. 2007, 13, 649 – 656
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
649

logue p-6, in which the benzo rings have p-regiochemistry,
were readily synthesised as outlined in Scheme 1.The bis-
pyrrole intermediate m-5 was prepared as described in the
litera
ture,[5g] and p-5 was obtained by this same method.
N
The [2+2] macrocyclisation of p-5 with acetone was opti-
mised by testing various stoichiometric proportions of ace-
tone and TFA (trifluoroacetic acid) in various solvents.The
best yield of p-6 (77% isolated by crystallisation from the
crude reaction mixture with EtOAc/hexanes) was obtained
by treating a 0.006m solution of p-5 in acetone/acetonitrile
[1:12 (v/v)] with TFA (5 equiv).Because these conditions
with m-5 gave m-6 in 50% yield, this reaction was not opti-
mised further.The macrocyclic structures of m-6 and p-6
were evident from their 1H and 13C NMR spectra, which
were consistent with time-averaged planar conformations.
Both compounds gave single crystals by slow cooling of
EtOAc/hexane (4:1) solutions.
Figure 1.Molecular structure of the Ci-symmetric macrocycle m-6.
The X-ray structure of m-6 shows the macrocycle to have
adopted a centrosymmetric rectangular box conformation
(Figure 1), similar to that seen for the related calix[2]fur-
an[4]pyrrole analogue.[5h] The C(6), C(14), C(6A) and C-
ꢀ
dence for a pair of weak intramolecular C H···p contacts
between a C(25) methyl proton on one of the long edges of
the rectangle and the aryl ring on the opposite edge, and
ꢀ
A
vice versa, with a H···p distance of 3.12 Š and a C H···p
angle of 1268.The two unique pyrrole rings are each ap-
ꢀ
A
proached by a pyrrole C H proton from a symmetry-related
counterpart of the other ring.The N(2) ring is approached
ꢀ
A
by the C(13) H proton in an adjacent molecule (H···p=
ꢀ
methyl groups.The N(10) and aryl rings are steeply inclined
to the plane of the macrocycle (by ꢁ81 and 708, respective-
ly); the N(10) and C(18) atoms are inclined inwards (to-
wards the centre of the macrocycle) in each case.In con-
trast, the N(2) pyrrole ring is much less tilted: it is inclined
(14A) plane with
N(2) oriented towards the centre of the macrocycle.Al-
though the N H hydrogens do not appear to be involved in
any significant intra- or intermolecular contacts, there is evi-
3.02 Š, C H···p=1598), and the N(10) ring is approached
ꢀ
by the C(4) H proton in a different proximal macrocycle
ꢀ
(H···p=3.29 Š, C H···p=1738).The aryl ring is stacked
with its centrosymmetrically related counterpart with mean
interplanar and centroid···centroid separations of ꢁ3.48 and
4.01 Š, respectively, the two rings being perfectly parallel.
The X-ray analysis of crystals of p-6 revealed the C2-sym-
metric macrocycle to have a conformation resembling a
twisted tennis-ball seam (Figure 2) similar to that of the re-
by only ꢁ388 to the C(6)-C(14)-C(6A)-C
G
ꢀ
Scheme 1.
650
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2007, 13, 649 – 656

Anion Binding of Calixbenzopyrroles
FULL PAPER
Table 1.Association constants Ka
C
m-6
p-6
1
2
Fꢀ
ꢁ20000[b,c]
4975ꢂ372
296ꢂ10
2246ꢂ132 17170ꢂ900[d] 57000ꢂ9000[e]
Clꢀ
Brꢀ
Iꢀ
no CIS
no CIS
no CIS
597ꢂ236
350[d]
1600ꢂ570[f]
61ꢂ3[e]
no CIS
34ꢂ8
10ꢂ0.5[d]
no CIS
<10[d]
CH3COOꢀ 755ꢂ33[g]
547ꢂ30[h]
97ꢂ4[d]
<10[d]
ꢀ
H2PO4
1711ꢂ120[i] no CIS
no CIS
no CIS
no CIS
ꢀ
HSO4
NO3
no CIS
no CIS
ꢀ
<10
16ꢂ0.4[c]
<10[j]
[a] Unless indicated otherwise, data were obtained by 1H NMR titration
and from the complexation-induced shifts (CIS) of the NH resonances in
“anhydrous” CD2Cl2 at 208C.[b] NH resonances disappeared at 20 8C
upon addition of the salt.[c] Ka value was estimated by competitive bind-
ing with p-6 and 1.[d] Data obtained by 1H NMR titration in “anhy-
drous” CD2Cl2 reported in ref.[1a].[e] Data from ref.[5h].[f] From titra-
tion data based on the CIS observed for the pyrrole CH resonances (NH
signals disappeared upon addition of the salt).[g] Ka value (95% confi-
dence interval) was calculated for an “ideal unique” 1:1 complex from ti-
tration data obtained in competitive binding with 1 (see main text and
Supporting Information).[h] Value previously reported (see ref.[1d]) is
668mꢀ1.[i] Kinetically slow on the NMR timescale at 20 8C.[j] Data from
ref.[4i].
Figure 2.Molecular structure of the C2-symmetric macrocycle p-6 show-
ing hydrogen bonding to the included ethyl acetate solvent molecules.
Geometric data for the hydrogen bond (a) are N···O=3.060(2) Š,
ꢀ
H···O=2.17 Š and N H···O=1688.
lated calix[6]pyrrole species[5h] (the C2 axis runs vertically in
Figure 2, bisecting the N(2)···N(2A) and N(10)···N(10A) vec-
G
ꢀ
tors).Whilst the N H group of the N(10) pyrrole ring is ori-
ented inwards and does not participate in any noteworthy
plex (1:2:1 intensities with different couplings to fluoride).
Only one of the two “isolated” benzo protons was shifted
significantly (Figure 3).These features are consistent with
the fluoride being off-centre within the macroring, thus, pro-
ducing a dissymmetric structure.
ꢀ
interactions, the N(2) H group is oriented outwards and
forms a hydrogen bond to the included ethyl acetate solvent
molecule (N···O=3.060(2) Š, H···O=2.17 Š and N H···O=
1688).Unlike the structure of the calix[6]pyrrole, here there
is no guest species encapsulated within the macrocycle.The
closest approach to the N(2) pyrrole ring is from an aryl
proton on the neighbouring C6H4 ring with a H···p separa-
ꢀ
ꢀ
tion of 2.98 Š and a C H···p angle of 1238.The H··· p vector
is inclined by only ꢁ558 to the pyrrole ring plane.The
N(10) pyrrole ring is approached on one side by a methyl
proton from the ethyl acetate solvent molecule (H···p=
ꢀ
2.68 Š, C H···p=1378) and on the other by an aryl proton
in an adjacent macrocycle (H···p=2.93 Š, C-H···p=1438),
the two interactions subtending an angle of ꢁ1618 at the
N(10) ring centroid.The closest approach to the aryl ring is
from an isopropylidene methyl proton in a neighbouring
ꢀ
macrocycle with a H···p separation of ꢁ3.40 Š and a C
H···p angle of ꢁ1718.
The anion-binding properties of m-6 and p-6 were ex-
plored by means of NMR titration experiments in DCM and
compared with those of calix[2]furan[4]pyrrole 2.The data
obtained are summarised in Table 1, which includes the rele-
vant binding constants (Ka) for calix[4]pyrrole 1.This is in-
cluded because it contains the same number of NH units as
m-6, p-6 and 2.All anions were used as nBu4N+ salts.
The Ka value for the complexation of m-6 with fluoride
could not be determined at room temperature because the
NH resonances disappeared upon addition of the salt.How-
ever, mixtures of m-6 with nBu4NF (various proportions)
Figure 3. 1H NMR spectra (CD2Cl2, ꢀ658C) of: i) m-6, 0.005m, ii) m-6+
nBu4NF (1:0.77) and iii) m-6+nBu4NF (1:1.81). The intensities and
J(H,F) of the NH resonances of the complex (a, b and c) are 1:2:1 and
A
52, 22 and 27 Hz, respectively.ArH* accounts for only one of the “isolat-
ed” ArH protons of the complex.No free m-6 is visible in iii).
1
gave resolved H NMR spectra at ꢀ658C consistent with the
The relative affinities of m-6 and p-6 for fluoride were in-
vestigated by direct competition experiments.Titration of a
1:1 mixture of p-6 and nBu4NF with m-6 resulted in the dis-
placement of fluoride from p-6·Fꢀ, as indicated by the re-
formation of a complex that was kinetically slow on the
NMR timescale (Ka was too high to be calculated reliably).
Three different NH resonances were observed for the com-
Chem. Eur. J. 2007, 13, 649 – 656
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
651

F.H.Kohnke, G.Cafeo et al.
gression of the complexation-induced shifts (CIS) of the NH
protons of p-6.This regression was complete after the addi-
tion of just over one equivalent of m-6.Therefore, we esti-
mate that the m-6·Fꢀ complex is not less than ten times
more stable than p-6·Fꢀ.This conclusion is also supported
by a competition experiment between m-6 and 1.In both ex-
periments, the NH resonances remained too broad to deter-
mine reliably the ratios of the Ka values until nearly one
equivalent of salt was added.However, it was evident that
m-6 is a marginally better fluoride ligand than 1, hence, we
can estimate Ka for m-6·Fꢀ to be ꢁ20000mꢀ1.
m-6 is the best ligand for Clꢀ, whereas p-6 does not show
any interaction with Clꢀ.
The X-ray structure of m-6·Clꢀ reveals that the encapsula-
tion of a chloride anion results in the macrocycle adopting a
conformation that resembles a tennis-ball seam (Figure 5)
The solid-state structure of the fluoride complex m-6·Fꢀ
revealed two independent complexes (I and II) with the
ꢀ
fluoride anion sitting in the plane of the four pyrrole N H
A
nitrogen atoms in each case; a geometry that differs from
the perching arrangement seen for m-6·Clꢀ (see below).The
{N4F} atoms are coplanar to within ꢁ0.02 Š in complex I
and ꢁ0.01 Š in complex II (complex I is shown in Figure 4,
Figure 5.Molecular structure of the macrocycle and the encapsulated
anion present in the crystals of m-6·Clꢀ.Hydrogen-bonding geometries
ꢀ
N···Cl, H···Cl[Š] and N H···Cl[8], respectively, are a) 3.3816(15), 2.49,
172; b) 3.3254(14), 2.43, 173; c) 3.4216(14), 2.52, 178; d) 3.4241(14), 2.53,
174.
rather than the rectangular box conformation seen for the
empty macrocycle (Figure 1).The chloride is held in place
ꢀ
by hydrogen bonds to all four of the pyrrole N H atoms
(the N···Cl distances range between 3.3254(14) and
3.4241(14) Š). That the chloride ion perches (by ꢁ1.09 Š)
on top of the plane of the four pyrrole nitrogen atoms
(which are coplanar to better than 0.01 Š) is reminiscent of
the encapsulation of chloride and bromide by calix[6]pyrrole
3.[5h] The adjacent aryl protons are not in close proximity to
the chloride (H···Cl distances of ꢁ3.09 and 3.14 Š) with no-
Figure 4.Molecular structure of one ( I) of the two independent “macro-
cycle and encapsulated anion” complexes present in the crystals of m-
ꢀ
ꢀ
6·F .Hydrogen-bonding geometries N···F, H···F[Š] and N H···F[8], re-
ꢀ
ticeably nonlinear C H···Cl angles of ꢁ117 and 1168, re-
spectively, are: a) 2.917(3), 2.03, 170; b) 2.922(4), 2.04, 166; c) 2.890(3),
2.00, 169; d) 2.918(3), 2.03, 171.
spectively.[10] Adjacent molecules are held together by a
ꢀ
series of four cooperative C H···p hydrogen bonds and a p–
p stacking interaction (see Figure S7 in the Supporting In-
formation).The C(18) aryl ring of one molecule stacks with
the C(43) aryl ring in a symmetry-related counterpart with
mean interplanar and centroid···centroid separations of
ꢁ3.16 and 3.92 Š, respectively; the two rings being inclined
by ꢁ58.The stacking is such that two of the aryl protons of
and complex II in Figure S3 in the Supporting Information).
The complexes, which have essentially the same geometry,[7]
each have molecular CS symmetry about a plane that in-
cludes the fluorine atom and is perpendicular to the N4F
plane, bisecting the N(2)···N(10) and N(27)···N(35) vectors.
The only intermolecular macrocycle···macrocycle contact of
note[8] is a centrosymmetric p–p stacking interaction involv-
ing the C6 ring located between the N(2) and N(35) pyrrole
rings of complex I (mean interplanar and centroid···centroid
separations of ꢁ3.78 and 4.06 Š, respectively).
ꢀ
the C(43) ring of the neighbouring molecule form C H···p
interactions with the N(2) and N(35) pyrrole rings (H···p=
ꢀ
ꢀ
3.00 Š, C H···p=1498, and H···p=2.91 Š, C H···p=1538,
respectively), and two of the aryl protons of the C(18) ring
ꢀ
form C H···p interactions with the N(10) and N(27) pyrrole
In the presence of Clꢀ, the 1H NMR spectrum of m-6
shows significant CISs not only for the NH resonances, but
rings of the neighbouring molecule (H···p=2.87 Š, C
ꢀ
ꢀ
H···p=1538, and H···p=2.77 Š, C H···p=1588 respective-
ly).This results in the formation of chains of molecules
ꢀ
also for the aryl C H units (from d=7.08 to d=8.10 ppm),
which appear to be involved in the binding of the anion.[9]
However, in the solid state (see below), this interaction
seems less relevant.Of those indicated in Table 1, receptor
along the direction of the crystallographic b axis.
1
The NH resonances in the H NMR of m-6 became very
broad on addition of acetate at room temperature.At
652
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2007, 13, 649 – 656

Anion Binding of Calixbenzopyrroles
FULL PAPER
ꢀ308C these signals became clearly visible as two pairs of
singlets each having 1:1 intensity (at d=12.27 and 11.40, and
d=10.37 and 8.10 ppm), labelled L and M, respectively, in
Figure 6.In the presence of 05. equivalents of acetate, M is
ꢁ4 times more intense than L.As the amount of acetate in-
resonances of 1 in the absence of competing receptors (1:1,
Ka value of 547ꢂ30mꢀ1, see Table 1).[13]
Titration data of receptor p-6 with acetate were consistent
with the kinetically fast formation of a 1:1 complex and a Ka
value that is equivalent to that for 1·CH3COOꢀ, but which is
also considerably higher than that of 2·CH3COOꢀ.Competi-
tive binding experiments with m-6, p-6 and acetate were
hampered by excessive broadening of the NH resonances of
both receptors.Acetate was complexed with equal strength
by p-6 and 1.
The X-ray structure of m-6·CH3COOꢀ (Figure 7) shows
the macrocycle to have adopted a CS-symmetric conforma-
tion with the encapsulated acetate anion sitting on the crys-
Figure 6.CIS in 1H NMR spectra of m-6 upon addition of
CH3COOnBu4N at ꢀ308C: i)–iv) 0, 0.5, 1.2 and 2.5 equivalents of added
salt, respectively; each pair of M and L resonances always have 1:1 inten-
sities, the M:L ratios in ii)–iv) are 3.8, 2.1 and 1.3, respectively.
creased, this proportion decreased (to M:Lꢁ1.3 with
2.5 equivalents of salt). These spectral features indicate the
presence of more than one complex.However, m-6 (0.005m)
appears to be nearly 100% complexed in the presence of
1.2 equivalents of acetate. This strong binding and the pres-
ence of various types of complexes hampered the evaluation
of a binding constant.However, we believe that the M reso-
nances refer to a dominant mode of complexation that is
compatible with the structural features in the X-ray crystal
structure of the host–guest complex formed between m-6
and acetate (see below).
To compare the overall affinities of m-6 and 1 for acetate,
we assumed m-6 forms a single ideal 1:1 complex with this
anion.Therefore, a 1:1 mixture of m-6 and 1 was used in a
1H NMR titration with acetate.Because the NH resonances
of m-6 disappeared upon addition of the salt, only the CISs
of 1 were used.Nonlinear least-square fitting of these data
for the system of equations comprising a) the expressions of
the 1:1 equilibrium constants of each receptor, b) the mass
balance relations and c) the well-known equation[11] that cor-
relates the observed d value to the degree of complexation
(used to determine the concentration of 1·CH3COOꢀ), gave
a Ka value of 755mꢀ1 for m-6 (Table 1).[12] In this calculation
we used the limiting chemical shift determined for the NH
Figure 7.Molecular structure of the CS-symmetric macrocycle and encap-
sulated anion present in the crystals of m-6·CH3COOꢀ.Hydrogen-bond-
ꢀ
ing geometries N···O, H···O[Š] and N H···O[8], respectively, are
a) 2.805(3), 1.91, 178; b) 2.764(3), 1.87, 173, and the p–p interactions (c)
have mean interplanar and centroid···C(31) separations of ꢁ3.55 and
3.58 Š, respectively (aryl and acetyl planes are inclined by ꢁ78).
tallographic mirror plane.The anion is held within the mac-
ꢀ
rocycle by N H···O hydrogen bonds to all four of the pyr-
ꢀ
role N H groups, each of the acetate oxygen atoms being
ꢀ
linked to two of the N H donors (the N···O distances are
2.805(3) and 2.764(3) Š).[14] The aryl rings are oriented ap-
proximately parallel to the plane of the acetate (inclined by
ꢁ78), such that there are a pair of p–p stacking interactions.
The delocalised COOꢀ moiety is sandwiched between the
two C6 rings with mean interplanar and centroid···C(31) sep-
arations of ꢁ3.55 and 3.58 Š, respectively. The two
aryl···C(31) vectors subtend an angle of ꢁ1738 at C(31).
There are no intercomplex interactions of note.
1
The H NMR spectrum of m-6 with dihydrogen phosphate
revealed the kinetically slow formation of a complex exhib-
iting two different NH resonances (at d=12.27 and
11.85 ppm), which were sufficiently sharp at room tempera-
ture (208C) to provide a means to calculate a Ka value of
Chem. Eur. J. 2007, 13, 649 – 656
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
653

F.H.Kohnke, G.Cafeo et al.
1711ꢂ120mꢀ1 for the 1:1 complex.Unlike the analogous
acetate complex, there was no evidence of the formation of
different complexes as a function of the amount of added
salt.
aryl ring.[16] Finally, m-6 proved to be a better anion receptor
than 1 towards all of the anions tested, with the exception of
iodide and hydrogen sulfate.For this reason, it may be a
more convenient starting material than 1 for the construc-
tion of optical anion sensors by functionalisation with appro-
priate chromophores.[17]
These spectral features suggest a binding geometry similar
to that observed in the solid state for the corresponding ace-
tate complex.Molecular models indicate the potential for
additional stabilisation by means of OH··p interactions be-
tween the hydroxy groups of the dihydrogenphosphate
anion and the aromatic benzo rings.Unfortunately, attempts
Experimental Section
ꢀ
to obtain crystals of the m-6·H2PO4 complex were unsuc-
General methods and instrumentation: Acetonitrile was distilled from
CaH2.Pyrrole was distilled before use.All other chemicals were standard
reagent grade and were used without further purification.All air-sensi-
tive and/or moisture-sensitive reactions were conducted under an inert
atmosphere.Thin-layer chromatography was carried out by using Merck
SiO2 60F254 plastic plates.Compounds were visualised with vanillin or by
examination under UV light.Column chromatography was conducted by
using silica gel (Aldrich, 230–400 mesh, 60 Š). 1H NMR and 13C NMR
spectra were recorded in CDCl3 or CD2Cl2 by using a Varian Gemini300
spectrometer at 300 and 75 MHz, respectively.Melting points were deter-
mined by using a Kofler hot-stage apparatus, and are not corrected.
1H NMR tritations: The n-tetrabutylammonium salts were dried in a
vacuum oven for a least 24 h.Solvents were used as supplied in sealed
ampoules, and care was taken to minimise exposure to moisture.Howev-
er, trace amounts of water could not be removed owing to the extremely
hydrophilic nature of the salts.Because water has been demonstrated to
lower the values of the observed binding constants, the data given in
Table 1 should be considered to be the minimum observable value if op-
erating under strictly anhydrous conditions.The anions were added as
measured volumes of solution (ꢁ0.035m) in CD2Cl2 to a solution of the
macrocycle under investigation (0.005m) in the same solvent (0.7 mL),
and the total volume was kept constant by evaporation with anhydrous
nitrogen.After each addition, the stoichiometric ratios between the salt
and macrocycle were also redetermined from the resonance intensities of
the host pyrrole protons versus those of the n-tetrabutylammonium
cation.Quantitative 1H NMR integrations were obtained by the use of
appropriate pulse delays in all cases.The data were processed by using
the WinEQNMR[18] program and gave the reported Ka values for the 1:1
complexation model.For binding processes that were kinetically slow on
the NMR timescale, the Ka values were determined from the ratios of
the intensities of bound and free species.[11] No analysis of experimental
errors was conducted, and the errors quoted in Table 1 are those related
to the fitting calculation alone.However, measurements were reproduci-
ble within 15%.
cessful.
ꢀ
Receptor p-6 displayed no CIS with H2PO4 .No interac-
tion could be detected for HSO4 with either m-6 or p-6;
ꢀ
however, NO3ꢀ was weakly bound by both receptors.
The anion-binding properties of m-6, p-6, 1 and 2 were
also tested by means of ESI MS (negative mode).[15] Howev-
er, upon mixing solutions of the receptors in CH2Cl2/CH3CN
(1:1) with solutions (various stoichiometric proportions) of
the anions given in Table 1 (with a nBu4N+ counterion), the
only supramolecular complexes that could be detected were
those with chloride.This can be explained by the fact that
under the experimental ESI conditions, CH2Cl2 becomes a
good source of chloride and, because this anion is present in
excess, it displaces any other of the tested anions from the
receptors.[15c]
Conclusion
This study demonstrates that hybrid-expanded calixpyrroles
m-6 and p-6 can be synthesised and purified very easily
from readily available starting materials at low cost, which is
a key issue for future practical and commercial applications.
Although chemically very similar, these two macrocycles
have considerably different anion-binding properties.Over-
all, m-6 is a better anion ligand than p-6.In fact, among the
tested anions, p-6 binds only fluoride and acetate to an ap-
preciable extent.However, with the exclusion of fluoride,
the selectivity of p-6 towards acetate is remarkably higher
than that observed for m-6.For p-6 there is a “shape” mis-
match with spherical or tetragonal anions (fluoride is too
small to be substantially affected by this factor), thus, only
the planar Y-shaped acetate and nitrate are bound (the
latter weakly) by this receptor.The kinetically slow forma-
The new compounds gave ESI mass spectra that were consistent with the
proposed structures.ESI mass spectra were acquired by using a Mariner
ESI-TOF instrument (resolution >7000) (Applied Biosystems, Framing-
ham, MA, USA) with CH2Cl2/CH3CN (negative-ion mode).Accurate
mass measurements were recorded by performing ESI.In all cases, differ-
ences between measured and calculated masses were less than 10 ppm.
Comparisons between measured and calculated isotopic patterns were
also performed.
ꢀ
tion of the m-6·H2PO4 complex is intriguing because the
1,3-Bis
pound was prepared as described in ref.[5g].
1,4-Bis[1’-(pyrrol-2-yl)-1’,1’-(dimethyl)methyl]benzene (p-5): A mixture
A
corresponding association constant is relatively small.
The original idea that the replacement of the furan rings
of calix[2]furan[4]pyrrole 2 with benzo rings would lead to
improved anion binding was confirmed for all of the anions
tested, with the exception of fluoride.On the basis of this
observation, we are currently pursuing the synthesis of ana-
logues of m-6 that have electron-withdrawing groups on the
benzo rings.These are expected to exhibit improved anion
binding relative to m-6 because of the increased acidity of
the CH unit between the two meta positions and/or by
means of anion···p interactions with the electron-deficient
ACHTREUNG
of pyrrole (14 mL) and 1,4-bis(1,1’-dimethylhydroxymethyl)benzene (2 g,
10.3 mmol) was degassed by bubbling with argon for 5 min, and
BF3·Et2O (1.27 mL, 10.3 mmol) was added. The mixture was stirred for
30 min at RT, diluted with CH2Cl2 (20 mL) and quenched by the addition
of aqueous NaOH (0.1 N, 20 mL). The organic layer was separated, ex-
tracted with water, dried (MgSO4) and concentrated.The resulting
brown oil was subjected to column chromatography (hexane/EtOAc 4:1)
to give a white solid as the major fraction, which was crystallised from
hexane/EtOAc (4:1) to give p-5 (30%).Mp.164–166
8C; 1H NMR
(CDCl3): d=1.67 (s, 12H, CH3), 6.10–6.12, 6.13–6.17 and 6.64–6.66 (3m,
3”2H, pyrrole-CH), 7.15 (s, 4H, benzo-CH), 7.68 ppm (brs, 2H, NH);
654
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2007, 13, 649 – 656

Anion Binding of Calixbenzopyrroles
FULL PAPER
Table 2.Crystal data, data collection and refinement parameters for compounds p-6, m-6, m-6·Fꢀ, m-6·Clꢀ and m-6·CH3COOꢀ.[a]
Data
p-6
m-6
m-6·Fꢀ
m-6·Clꢀ
m-6·CH3COOꢀ
(C16H36N+) [C46H56N4·C2H3OOꢀ](C16H36N+)
formula
solvent
formula weight
colour, habit
C46H56N4
2C4H8O2
841.16
colourless
blocks
C46H56N4
–
664.95
colourless
needles
[C46H56N4·Fꢀ]
0.75C6H14
991.04
colourless
blocks
G
R
N
CH2Cl2·C4H8O2·C6H14
C6H14
1052.62
colourless
blocks
0.36”0.31”0.21
173
1202.06
colourless
plates
0.28”0.13”0.03
173
crystal size
T[K]
[mm3]
0.17”0.13”0.08 0.27”0.07”0.04 0.11”0.08”0.06
173
173
173
crystal system
space group
a[Š]
b[Š]
c[Š]
monoclinic
C2/c (no.15)
25.076(3)
10.0222(10)
20.280(2)
–
monoclinic
P21/n (no.14)
7.0634(4)
26.3820(13)
10.0535(5)
–
monoclinic
P21/c (no.14)
23.8427(18)
13.9176(14)
39.047(3)
–
monoclinic
P21/n (no.14)
16.5509(17)
18.024(3)
24.900(4)
–
tetragonal
P421m (no.113)
20.051(2)
–
15.9085(7)
–
¯
a[8]
b[8]
109.320(9)
–
90.325(4)
–
103.884(6)
–
106.825(11)
–
7109.8(17)
4
1.123
CuKa
–
–
g[8]
V[Š3]
Z
4809.8(8)
1873.41(17)
12578.6(18)
6395.8(9)
4[b]
2[c]
8[d]
4[e]
1calcd
[gcmꢀ3
]
1.162
CuKa
0.566
142
1.179
CuKa
0.519
142
1.047
CuKa
0.471
130
1.093
MoKa
0.065
64
radiation
m[mm]
2q max[8]
1.510
142
unique reflns
4551
3547
3166
234
0.039, 0.099
20792
8352
1316
0.060, 0.136
13639
7860
711
0.041, 0.099
11004
6617
409
obsd reflns jFo j>4s
(jFoj) 3270
parameters
298
[f]
R1, wR2
0.051, 0.136
0.089, 0.212
[a] Details in common: graphite-monochromated radiation, refinement based on F2.[b] Molecule has crystallographic C2 symmetry.[c] Molecule has
crystallographic Ci symmetry.[d] Two independent molecules.[e] Molecule has crystallographic
CS symmetry.[f] R1 =ꢀjjFo jꢀjFc jj/ꢀjFo j ; wR2 =
{ꢀ[w(F2oꢀF2c)2]/ꢀ[w(F2o)2]}1/2; wꢀ1 =s2(F2o)+(aP)2 +bP.
13C NMR (CDCl3): d=30.0, 38.8, 104.2, 107.6, 116.6, 126.0, 140.4,
146.7 ppm.
CCDC 603345 (m-6), 603346 (p-6), 603347 (m-6·Clꢀ), 603348 (m-
6·CH3COOꢀ) and 603349 (m-6·Fꢀ) contain the supplementary crystallo-
graphic 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.
Calix[2]-m-benzo[4]pyrrole (m-6): TFA (778 mg, 6.82 mmol) was added
to a mixture of m-5 (400 mg, 1.36 mmol) and acetone (17 mL) in anhy-
drous acetonitrile (200 mL) at 08C.The mixture was stirred at RT for
1.5 h, neutralised with NaOH (0.1m) and then concentrated and extract-
ed with CH2Cl2 (50 mL).The organic phase was separated, washed with
aq.NaHCO (2”50 mL), dried (MgSO4) and concentrated.The residue
3
was crystallised from hexane/EtOAc (4:1) to give m-6 (225 mg, 50%).
M.p. 206–2088C; 1H NMR (CDCl3): d=1.52 (s, 12H, CH3), 1.53 (s, 24H,
CH3), 5.83–5.86 (m, 8H, pyrrole-CH), 7.01–7.11 (m, 8H, benzo-CH),
Acknowledgements
1
7.67 ppm (brs, 4H, NH); H NMR (CD2Cl2): d=1.41 (s, 12H, CH3), 1.50
We thank the University of Messina for financial support and the
EPSRC (SRIF2) in the UK for purchase of the X-ray diffractometers.
This work was sponsored by the University of Messina, Italy.
(s, 24H, CH3), 5.76–5.81 (m, 8H, pyrrole-CH), 6.88–6.98 (m, 6H, benzo-
CH), 7.08 (s, 2H, benzo CH), 7.48 ppm (brs, 4H, NH); 13C NMR
(CDCl3): d=29.8, 29.8, 35.6, 39.2, 103.0, 103.9, 123.6, 123.6, 124.1, 127.9,
138.5, 139.5, 148.9 ppm.
[1] a) PA. .Gale, JL. .Sessler, V.Krµl, V.Lynch,
J. Am. Chem. Soc.
Chem.
Calix[2]-p-benzo[4]pyrrole (p-6): TFA (973 mg, 8.53 mmol) was added to
a mixture of p-5 (500 mg, 1.70 mmol) and acetone (21 mL) in anhydrous
acetonitrile (250 mL) at 08C.The mixture was stirred at RT for 2 h and
then neutralised with NaOH (0.1m).The white solid was filtered and
crystallised from hexane/EtOAc (4:1) to give p-6 (510 mg, 77%).Mp..
2098C; 1H NMR (CDCl3): d=1.54 (s, 12H, CH3), 1.56 (s, 24H, CH3),
5.84–5.88 (m, 8H pyrrole-CH), 6.98 (s, 8H, benzo-CH), 7.25 ppm (brs,
1996, 118, 5140–5141; b) P.A. Gale, J.L. Sessler, V. Krµl,
Commun. 1998, 1–8, and references therein; c) P.A. Gale, P. An-
zenbacher, Jr., J. L. Sessler, Coord. Chem. Rev. 2001, 222, 57–102;
d) C.N.Warriner, P.A.Gale, M.E.Light, M.B.Hursthouse,
Commun. 2003, 1810–1811.
Chem.
[2] a) V.Krµl, J.L.Sessler, R.S.Zimmerman, D.Seidel, V.Lynch, B.
Andrioletti, Angew. Chem. 2000, 112, 1097–1100; Angew. Chem.
Int. Ed. 2000, 39, 1055–1058; b) J.L. Sessler, R.S. Zimmerman, C.
Bucher, V.Krµl, B.Andrioletti, Pure Appl. Chem. 2001, 73, 1041–
1057.
1
4H, NH); H NMR (CD2Cl2): d=1.50 (s, 12H, CH3), 1.52 (s, 24H, CH3),
5.79–5.84 (m, 8H, pyrrole-CH), 6.96 (s, 8H, benzo-CH), 7.26 ppm (brs,
4H, NH); 13C NMR (CDCl3): d=30.3, 30.3, 35.1, 35.1, 102.5, 103.8, 125.8,
137.9, 139.2, 146.8 ppm.
[3] Recent papers in this area include: a) P.K. Panda, C-.H. Lee, Org.
Lett. 2004, 6, 671–674; b) X.K.Ji, D.StC. Black, S.B.Colbran,
X-ray crystallography: Table 2 summarises the crystallographic data for
compounds m-6, p-6, m-6·Fꢀ, m-6·Clꢀ and m-6·CH3COOꢀ.Data were col-
lected by using Oxford Diffraction PXUltra (m-6, p-6, m-6·Fꢀ and m-
6·Clꢀ) and an Xcalibur3 diffractometers (m-6·CH3COOꢀ), and the struc-
tures were refined based on F2 by using the SHELXTL and SHELX-97
program systems.[19] The absolute structure of m-6·CH3COOꢀ could not
be unambiguously determined by either R-factor tests (R1+ =0.0889,
R1ꢀ =0.0889) or by the Flack parameter (x+ =+0.4(19), xꢀ =+0.6(19)).
D.C.Craig, K.M.Edbey, J.B.Harper, G.D.Willett,
2005, 61, 10705–10712; c) A.F.D. de Namor, M. Shehab, J. Phys.
Chem. B 2005, 109, 17440–17444; d) J.L. Sessler, W-.S.Cho, D.E.
Tetrahedron
Gross, J.A.Shriver, V.M.Lynch, M.Marquez,
J. Org. Chem. 2005,
70, 5982–5986; e) P.K. Panda, C-.H. Lee, J. Org. Chem. 2005, 70,
3148–3156; f) C.-H. Lee, J.-S. Lee, H.-K. Na, D.-W. Yoon, H.
Miyaji, W.-S. Cho, J. L. Sessler, J. Org. Chem. 2005, 70, 2067–2074.
Chem. Eur. J. 2007, 13, 649 – 656
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
655

F.H.Kohnke, G.Cafeo et al.
[4] a) JL. .Sessler, H.Maeda, T.Mizuno, VM. .Lynch, H.Furuta,
Chem. Commun. 2002, 862–863; b) J.L. Sessler, S. Camiolo, P.A.
Gale, Coord. Chem. Rev. 2003, 240, 17–55; c) J.M.Veauthier, W-.S.
Cho, V.M. Lynch, J.L. Sessler, Inorg. Chem. 2004, 43, 1220–1228;
[8] The disorder present in the nBu4N+ cations precludes detailed anal-
ysis of any potential intermolecular contacts involving these units.
[9] See, for example D.Yoon, H.Hwang, C-.H.Lee,
Angew. Chem.
2002, 114, 1835–1837; Angew. Chem. Int. Ed. 2002, 41, 1757–1759.
[10] A comparison of the geometry of m-6·Clꢀ with that of the chloride
complex of calix[2]bis(pyrrol-2-yl)benzene (see ref.[4i]) is especially
interesting.
d) R.Li, L.S.Evans, D.S.Larsen, P.A.Gale, S.Brooker,
NewJ.
Chem. 2004, 28, 1340–1343; e) J.L.Sessler, E.Katayev, G.Dan Pan-
tos, Y.A.Ustynyuk, Chem. Commun. 2004, 1276–1277; f) J.L.Sess-
ler, D.An, W-.S.Cho, V.Lynch, M.Marquez,
2005, 540–542; g) P.Piatek, V.M.Lynch, J.L.Sessler,
Soc. 2004, 126, 16073–16076; h) E. C. Lee, Y.-K. Park, J.-H. Kim,
H. Hwang, Y.-R. Kim, C.-H. Lee, Tetrahedron Lett. 2002, 43, 9493–
9495; i) JL. .Sessler, D.An, W-S. .Cho, V.Lynch, M.Marquez,
Chem. Eur. J. 2005, 11, 2001–2011.
Chem. Commun.
J. Am. Chem.
[11] a) K.Hirose, J. Inclusion Phenom. Macrocyclic Chem. 2001, 39,
193–209; b) L.Fielding, Tetrahedron 2000, 56, 6151–6170.
[12] Additional details on the method and calculations used for the de-
termination of the “overall” 1:1 binding constant of m-6 with acetate
are available as Supporting Information.
[13] For a similar use of the limiting chemical shift, see: K.Williams, B.
[5] a) B.Turner, M.Botoshansky, Y.Eichen,
Angew. Chem. 1998, 110,
Askew, P.Ballester, C.Buhr, K.S.Jeong, S.Jones, J.Rebek, Jr,.
J.
2633–2637; Angew. Chem. Int. Ed. 1998, 37, 2475–2478; b) J.L.
Sessler, P. Anzenbacher, Jr., J.A. Shriver, K. Jursíkovµ, V.M.
Lynch, M.Marquez, J. Am. Chem. Soc. 2000, 122, 12061–12062;
Am. Chem. Soc. 1989, 111, 1090–1094.
[14] This mode of binding is reminiscent of the “cross-binding” geometry
described for the solid-state complexes of calix[2]bipyrrole[2]thio-
phene with acetate and benzoate (see ref.[5k]).
c) N.Arumugam, Y-.S.Jang, C-.H.Lee,
3117; d) Y.-S. Jang, H.-J. Kim, P.-H. Lee, C.-H. Lee, Tetrahedron
Lett. 2000, 41, 2919–2923; e) A.Nagarajan, J.W.Ka, C-.H.Lee, Tet-
rahedron 2001, 57, 7323–7330, and references therein; f) B.Turner,
A.Shterenberg, M.Kapon, Y.Eichen, K.Suwinska, Chem.
Org. Lett. 2000, 2, 3115–
[15] For some recent examples on the use of ESI MS for the characteri-
sation of supramolecular complexes, see: a) I.Alfonso, C.Astorga,
V.Gotor, J. Incl. Phenom. Macroc. Chem. 2005, 53, 131–137;
b) G.V.Oshovsky, W.Verboom, R.H.Fokkens, D.N.Reinhoudt,
Chem. Eur. J. 2004, 10, 2739–2748; c) Acetonitrile was a necessary
addition to obtain ESI MS spectra.We did not test solvents other
than DCM as we wanted to obtain data as comparable as possible
Commun. 2001, 13–14; g) J.L.Sessler, W-.S.Cho, V.Lynch, V.Krµl,
Chem. Eur. J. 2002, 8, 1134–1143; h) G.Cafeo, F.H.Kohnke, G.L.
La Torre, M.F.Parisi, R.Pistone Nascone, A.J.P.White, D.J.Wil-
liams, Chem. Eur. J. 2002, 8, 3148–3156; i) B.Turner, A.Shteren-
1
with those obtained by H NMR in this solvent.
berg, M.Kapon, M.Botoshansky, K.Suwinska, Y.Eichen,
Commun. 2002, 726–727; j) G.Cafeo, F.H.Kohnke, M.F.Parisi, R.
Pistone Nascone, G.L. La Torre, D.J. Williams, Org. Lett. 2002, 4,
2695–2697; k) 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.
[6] a) C.-H. Lee, H.-K. Na, D.-W. Yoon, D.-H. Won, W.-S. Cho, V. M.
Lynch, S.V. Shevchuk, J.L. Sessler, J. Am. Chem. Soc. 2003, 125,
7301–7306; b) J.Y. Kwon, Y.J. Jang, S.K. Kim, K-.H. Lee, J.S.
Kim, J.Yoon, J. Org. Chem. 2004, 69, 5155–5157.
Chem.
[16] For a discussion on this type of interaction, see: C.Garau, D.QuiÇo-
nero, A.Frontera, P.Ballester, A.Costa, P.M.Deyà,
NewJ. Chem.
2003, 27, 211–214.
[17] R.Nishiyabu, P.Anzenbacher, Jr,.
8270–8271.
J. Am. Chem. Soc. 2005, 127,
[18] M.J.Hynes, J. Chem. Soc. Dalton Trans. 1993, 311–312.
[19] SHELXTL PC version 5.1, BrukerAXS, Madison, WI, 1997; G.
Sheldrick, SHELX-97, Institut Anorg.Chemie, Tammannstr.4,
D37077 Gçttingen (Germany), 1998.
[7] The rms fit of the non-hydrogen atoms of complexes I and II is
ꢁ0.19 Š, with the largest deviation being ꢁ0.38 Š.
Received: June 28, 2006
Published online: September 25, 2006
656
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2007, 13, 649 – 656