Quantitative evaluation of anion-π interactions in solution

Article abstract of DOI:10.1002/anie.200800636

(Figure Presented) Anions included: A series of meso-tetraaryl calix[4]pyrrole receptors have been used as a model system to quantify chloride-π interactions in solution (see picture; green balls are chloride ions). The free energy values are generally indicative of a repulsive interaction; their magnitude depends on the substituent on the aromatic ring.

Full text of DOI:10.1002/anie.200800636

Communications
DOI: 10.1002/anie.200800636
Anion–p Interactions
Quantitative Evaluation of Anion–p Interactions in Solution**
Guzmµn Gil-Ramírez, Eduardo C. Escudero-Adµn, Jordi Benet-Buchholz, and Pablo Ballester*
Chemical intuition assigns a repulsive force to the interaction
of anions with the p system of aromatic compounds. Con-
sequently, for many years anion–p interactions have been
completely neglected as noncovalent interactions suitable for
the construction of efficient anion receptors.[1] In 2002 four
computational studies suggested the existence, in the gas
phase, of strong attractive interactions between highly p-
acidic aromatic systems and anions lying above the plane of
the p system.[2] These seminal studies triggered an extensive
investigation of the physical nature of anion–p interactions
using electronic structure methods.[3] Moreover, since 2004,
the reporting of several supramolecular complexes obtained
in the solid state, in which anions are nestled up next to
aromatic rings, has increased interest in the subject.[4]
Scheme 1. Chloride binding promotes a change from the 1,3-alternate
to cone conformation of 1. Substitution of the “axial” methyl groups of
1 by p-substituted phenyl groups produces a calix[4]pyrrole receptor 2
with a deep aromatic cavity surrounding the halide ion.
Models and estimates of the factors that govern the
binding geometry, the binding strength, and the nature of
anion–p interactions are available mainly from computational
studies.[3] In solution, only the use of highly electron-deficient
arenes has allowed the unambiguous quantification of the
halide–p interaction in the absence of other intermolecular
attractive interactions.[5] Accordingly, more experimental
quantitative data on the magnitudes of the interaction of
anions and p systems of aromatic rings in solution are needed,
as well as further understanding of the influence of the
substituents on the aromatic ring.[6]
Weak interactions may be detected as modulations of
stronger binding forces by using synthetic receptors in
combination with the “enforced proximity” approach.[7]
Octamethyl calix[4]pyrrole (1; Scheme 1) is an effective
receptor for anions,[8] and the binding of halide anions
organizes the receptor into the cone conformation.
anions.[9,10] The formation of four hydrogen bonds between
a halide ion and the NH groups of the calix[4]pyrrole scaffold
constitutes a reliable interaction which positions the anion
above the planes of the p systems of the meso aryl substitu-
ents (enforced proximity). Previous studies have already
shown that chemical modification of the substituents of the
aromatic rings alters the intrinsic anion-binding selectivity
and affinity of the basic calix[4]pyrrole skeleton.[11] Herein,
we report a systematic study of the binding of chloride to a
series of meso-tetraaryl calix[4]pyrrole receptors 2 displaying
deep and electronically tunable aromatic cavities. The recep-
tors are used as model systems to obtain quantitative data on
the strength of the anion–p interaction in solution and also to
measure the effect of the aromatic substituents.
As shown in Scheme 1, the conelike conformation of the
a,a,a,a stereoisomer of calix[4]pyrrole 2 bearing axially
substituted aryl groups on each of the four meso carbons
contains
a
deep aromatic cavity capable of including
We have prepared a series of neutral a,a,a,a-tetraaryl
calix[4]pyrroles 2 containing deep aromatic cavities with fixed
walls (Scheme 2). Different para substituents (R) are used to
modify the charge distribution of the aromatic ring. Com-
pounds 2a–e were synthesized by the acid-catalyzed con-
densation of pyrrole and the corresponding methyl aryl
ketone under standard reaction conditions. Careful purifica-
tion (crystallization and/or chromatography techniques)
allowed the isolation of the corresponding a,a,a,a isomers
in yields ranging from 6 to 20%. Compounds 2 f and 2g are
simple derivatives of 2e and were readily obtained through
described methodologies.[9a,b]
[*] G. Gil-Ramírez, E. C. Escudero-Adµn, J. Benet-Buchholz,
Prof. P. Ballester
Institute of Chemical Research of Catalonia (ICIQ)
Avda. Països Catalans 16, 43007 Tarragona (Spain)
Fax: (+34)977-920-221
E-mail: pballester@iciq.es
Homepage: http://www.iciq.es/english/research/grups_eng/
ballester/entrada.htm
Prof. P. Ballester
Catalan Institution for Research and Advanced Studies (ICREA)
Pg. Lluís Companys 23, 08010 Barcelona (Spain)
Configurational assignment was performed by a combi-
nation of 1H NMR spectroscopy and single-crystal X-ray
crystallographic analysis.[12] All the a,a,a,a isomers 2a–g
display only one type of b-pyrrole resonance and exhibit, in
acetonitrile solution, a sharp and well-resolved proton
spectrum indicative of a time-averaged C4 symmetry. The
[**] We thank the Spanish MEC for generous grants (CTQ2005-08989-
C01-C02/BQU and Consolider Ingenio 2010, Grant CSD2006-0003).
We also thank ICIQ Foundation and Generalitat de Catalunya,
DURSI (2005SGR00108) for financial support.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
4114
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 4114 –4118

Angewandte
Chemie
Figure 2. Regions of the 1H NMR spectra of free receptors 2a and 2c
and their chloride inclusion complexes (CD3CN, 298 K, [2]=1.3 mm,
500 MHz). See Scheme 2 for proton assignment. Numbers with
primes indicate the signals of complexed 2; “5” represents the N-CH2
protons of TBA. a) 2c; b) 2c+0.55 equiv TBACl; c) 2a+5 equiv
TBACl; d) 2a.
Scheme 2. Chemical structures of the meso-tetraaryl calix[4]pyrroles
2a–g.
single-crystal X-ray structures of free a,a,a,a isomers grown
from coordinating solvents revealed in most cases that the
calixpyrrole core is in a cone conformation, with one solvent
molecule included in the aromatic cavity and hydrogen
bonded to the four NH groups (Figure 1).
Consequently, the new signals were assigned to protons of
the receptor in the ClÀ@2 complex.
Upon chloride binding, the peaks of the pyrrole NH
groups were consistently shifted downfield (Dd = 3.03 to
3.48 ppm) while the signals for the ortho and meta aromatic
protons of the complex moved upfield (Dd = À0.16 to À0.05).
The values of the downfield shift experienced by the NH
groups are consistent with the formation of four hydrogen
bonds with the chloride, whereas the observed upfield shifts
of the aromatic protons (shielding effect) indicate that the
receptors 2 do not interact with the chloride ion by means of
À
C H···Cl hydrogen bonds but instead through ClÀ–p inter-
À
actions. Moreover, the existence of separate signals for free
and bound calix[4]pyrrole receptor 2 indicates that the
exchange of the bound chloride by a solvent molecule is
slow on the NMR timescale. Most likely, the release of the
included chloride ion requires the conversion of the low-
energy cone conformer assigned to the anionic complex into a
higher-energy half-cone or alternate conformer.
Taken together, these results constitute a clear indication
that in solution the ClÀ@2 complexes adopt a binding
geometry featuring the calixpyrrole core in a cone conforma-
tion and the chloride deeply included in the aromatic cavity
formed by the axially oriented phenyl groups.[16] The included
chloride is hydrogen bonded to the four pyrrolic NH groups
and experiences anion–p interactions with the p system of the
aromatic rings.
Figure 1. X-ray crystal structures of the solvates of the receptors:
a) 2a+acetone, b) 2b+dichloromethane, c) 2c+acetonitrile,
d) 2d+acetonitrile, e) 2 f+DMF, and f) 2g+acetonitrile. DMF=N,N-
dimethylformamide. H white, C gray, O red, N blue, Cl green.
1H NMR spectroscopy was also used to probe the chloride
binding in solution and provide structural information on the
geometry of the complex (Figure 2). We used tetrabutylam-
monium chloride (TBACl) as the precursor of the chloride
guest. In acetonitrile and in the range of concentrations used,
this electrolyte is predominantly dissociated and the binding
process of calix[4]pyrroles with TBACl is expected to produce
simple ClÀ@1 anionic complexes.[13] The initial addition of
0.5 equivalents of TBACl to an acetonitrile solution (1 mm)
containing the aryl calix[4]pyrrole 2[14] induced the appear-
ance of a new set of proton signals except for 2e.[15] When
more than 0.5 equivalents of TBACl was added, the new
signals grew at the expense of the signals of free 2.
Quantitative assessments of the binding constants with
chloride and most receptors 2 were made by simple integra-
1
tion of the H NMR spectra and/or by means of isothermal
titration calorimetry (ITC) experiments. As a result of the low
solubility of 2b in acetonitrile, we used the solubility
technique in this particular instance.[17] In all cases we used
acetonitrile containing about 200 ppm of water as solvent.
The calculated association constants for the 1:1 complexes are
summarized in Table 1. The inclusion complexes ClÀ@2 can
Angew. Chem. Int. Ed. 2008, 47, 4114 –4118
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4115

Communications
Table 1: Association constants Ka [”103 mÀ1] measured in acetonitrile or CD3CN at 298 K for the endo 1:1
(chloride) and the electrostatic
potential surface of the aromatic
ring.[22]
complexes of 2 and chloride. Statistically corrected free energy values DDG [kcalmolÀ1] calculated for the
chloride–p interactions. 1H chemical shift values dNH [ppm] of the pyrrole NH groups and their
complexation-induced change in chemical shift Dd [ppm].
The maximum free energy value
measured for the chloride–p inter-
action with this model system is
clearly repulsive and can be esti-
mated as approximately 1.0 kcal
molÀ1 (R = OMe). An increase in
the electron-withdrawing character
of the R substituent induces a less
repulsive anion–p interaction. This
result could be simply caused by the
depletion of the electron density of
2 f
2a
2g
2b
2c
2d
1
R
Ka
MeO
0.13[a]
Æ0.03
2.9
1.0
7.9
H
MeCO2
1.10[a,b]
Æ0.20
4.1
0.7
8.0
Br
3.80[c]
Æ0.40
4.8
0.5
–
CN
33.0[b]
Æ3.50
6.2
0.1
7.9
NO2
–
0.25[a]
180[b]
Æ20.0
7.2
108[b]
Æ11.0
6.9
Æ0.02
ÀDG
3.1
0.9
7.9
3.5
DDG[d]
dNH
Dd
À0.1
–
7.4
3.6
8.1
3.5
3.2
–
3.1
3.0
[a] Ka value determined by 1H NMR spectroscopy. [b] Ka value determined by ITC. [c] Ka value determined
using solubility titration. [d] DDG=
(DGClÀ@2ÀDGClÀ@1)/4.
also be readily detected by negative-ion mass spectrometry
(MS). The relative binding affinities of chloride towards
calix[4]pyrrole receptors established by ESI-MS competitive
binding experiments are in line with the results obtained in
solution (see the Supporting Information).
In general, the association constant values for the ClÀ@2
complexes increase with the electron-withdrawing character
of the R substituent. The chemical shift data presented in
Table 1 show that the strength of the primary interaction, that
is, the hydrogen bonding between the pyrrolic NH groups and
the chloride anion, is not significantly perturbed by the
introduction of the meso phenyl groups and by the modifi-
cation of their para substituent. Thus, the change in chemical
shift experienced by the NH proton signal upon chloride
binding is quite similar in the entire receptor series.[18]
Figure 3. Experimental chloride–p interaction energies (DDG) correlate
with the Hammett constant sp values for the substituent R.
the p clouds of the aromatic rings owing to the influence of
the substituent. Nevertheless, this simple model also suggests
that eventually, when R = NO2, the interaction between the
anion and the aromatic surfaces could become slightly
attractive (ca. À0.1 kcalmolÀ1).
The differences in the stability constants calculated for the
ClÀ@2 complexes can provide a direct measurement of the
relative interaction energy of chloride with different aromatic
p systems.[19] The maximum contribution of the chloride–p
interactions to the overall free energy of binding in the
receptor series 2 can be estimated as + 4.4 kcalmolÀ1 (about
1.1 kcalmolÀ1 per aromatic ring). This value represents the
difference in free binding energy between 2 f and 2d. Due to
the fact that repulsive chloride–p interactions are surely
operative in the complex ClÀ@2 f, the value of 1 kcalmolÀ1
represents an overestimation of the stabilization energy for
the attractive interaction between the chloride ion and the
p system of p-nitrophenyl in acetonitrile. The chloride com-
plex (ClÀ@1) of octamethyl calix[4]pyrrole (1) supplies an
ideal reference to better quantify the free energy values for
chloride–p interactions in solution.[20] The statistically cor-
rected chloride–p free energy values (DDG) for various
substituted aromatics are reported in Table 1.
The obtained data indicate that in all cases, except for R =
NO2, the free energy for the chloride–p interaction is
repulsive. The magnitude of the repulsion, however, is clearly
sensitive to the nature of the aromatic substituent R. The
calculated DDG data correlate well with the corresponding
Hammett constants for the para and meta substituents[21] (sp
R2 = 0.95 and sm R2 = 0.92, Figure 3), thus providing some
insight into the origin of the variation. A likely explanation of
the observed trend is consistent with the existence and
variation of an electrostatic interaction involving the anion
The binding geometry proposed in solution is also
supported by the formation of endo cavity complexes ClÀ@2
in the solid state (Figure 4).
In conclusion, we have shown that a series of meso-
tetraaryl calix[4]pyrrole receptors can be used as a model
system to quantify chloride–p interactions in solution. By
1
means of H NMR spectroscopy and X-ray crystallographic
studies we have demonstrated that the chloride–arene
interactions observed in these complexes are established
exclusively with the p-aromatic system. The quantitative
Hammett free-energy relationship derived demonstrates that
the detected chloride–p interactions are dominated by
electrostatic effects. The observation, in the solid state and
in solution, of chloride ions placed next to aromatic surfaces
does not relate directly to the existence of attractive
interactions between them. The values of the free energies
estimated for the interaction of chloride with the p-aromatic
systems are probably not transferable to other model systems.
We believe, however, that the observed trend and order of
magnitude should be of general applicability. Our study
clearly demonstrates that, in acetonitrile solution, equipping
the calix[4]pyrrole core with four additional chloride–p-
nitrophenyl interactions does not produce a chloride receptor
with much higher affinity than the simple octamethyl
calix[4]pyrrole (1).
4116
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 4114 –4118

Angewandte
Chemie
P. Ballester, A. Costa, P. M. Deyà, Chem. Phys. Lett. 2002, 359,
486 – 492.
[3] a) C. Garau, A. Frontera, D. Quiæonero, P. Ballester, A. Costa,
P. M. Deyà, ChemPhysChem 2003, 4, 1344 – 1348; b) D. Kim, P.
Tarakeshwar, K. S. Kim, J. Phys. Chem. A 2004, 108, 1250 – 1258;
c) A. Clements, M. Lewis, J. Phys. Chem. A 2006, 110, 12705 –
12710; d) B. L. Schottel, H. T. Chifotides, M. Shatruk, A.
Chouai, L. M. Perez, J. Bacsa, K. R. Dunbar, J. Am. Chem.
Soc. 2006, 128, 5895 – 5912; e) P. U. Maheswari, B. Modec, A.
Pevec, B. Kozlevcar, C. Massera, P. Gamez, J. Reedijk, Inorg.
Chem. 2006, 45, 6637 – 6645; f) M. Mascal, Angew. Chem. 2006,
118, 2956 – 2959; Angew. Chem. Int. Ed. 2006, 45, 2890 – 2893;
g) O. B. Berryman, V. S. Bryantsev, D. P. Stay, D. W. Johnson,
B. P. Hay, J. Am. Chem. Soc. 2007, 129, 48 – 58.
[4] For recent reviews of computational and experimental work
related to anion interactions with aromatic rings, see: a) P.
Gamez, T. Mooibroek, S. Teat, J. Reedijk, Acc. Chem. Res. 2007,
40, 435 – 444; b) B. L. Schottel, H. T. Chifotides, K. R. Dunbar,
Chem. Soc. Rev. 2008, 37, 68 – 83.
[5] Y. S. Rosokha, S. V. Lindeman, S. V. Rosokha, J. K. Kochi,
Angew. Chem. 2004, 116, 4750 – 4752; Angew. Chem. Int. Ed.
2004, 43, 4650 – 4652.
[6] For other solution studies that highlight the existence of
attractive anion–p interactions, see: a) O. B. Berryman, F. Hof,
M. J. Hynes, D. W. Johnson, Chem. Commun. 2006, 506 – 508;
b) M. Mascal, I. Yakovlev, E. B. Nikitin, J. C. Fettinger, Angew.
Chem. 2007, 119, 8938 – 8940; Angew. Chem. Int. Ed. 2007, 46,
8782– 8784.
Figure 4. Side and top views (left and center, respectively) of the X-ray
structures of the chloride complexes with receptors 2: a) TEA·ClÀ@2d,
b) TBA·ClÀ@2d, and c) TMA·ClÀ@2g. Solvent molecules and the
organic cation are omitted in the top views. The right-hand column
illustrates the binding motif of the anion–p interaction with the shorter
distance to the ring centroid of the p system. The degree of displace-
ment of the chloride from the center of the arene (doffset) is also
indicated. TEA=tetraethylammonium, TMA=tetramethylammonium.
H white, C gray, O red, N blue, Cl,ClÀ green.
[7] J. Rebek, B. Askew, P. Ballester, C. Buhr, S. Jones, D. Nemeth, K.
Williams, J. Am. Chem. Soc. 1987, 109, 5033 – 5035.
[8] a) J. L. Sessler, D. E. Gross, W.-S. Cho, V. M. Lynch, F. P.
Schmidtchen, G. W. Bates, M. E. Light, P. A. Gale, J. Am.
Chem. Soc. 2006, 128, 12281 – 12288; b) P. A. Gale, J. L. Sessler,
V. Kral, V. Lynch, J. Am. Chem. Soc. 1996, 118, 5140 – 5141.
[9] a) C. J. Woods, S. Camiolo, M. E. Light, S. J. Coles, M. B.
Hursthouse, M. A. King, P. A. Gale, J. W. Essex, J. Am. Chem.
Soc. 2002, 124, 8644 – 8652; b) P. Anzenbacher, Jr., K. Jursikova,
V. M. Lynch, P. A. Gale, J. L. Sessler, J. Am. Chem. Soc. 1999,
121, 11020 – 11021; c) L. Bonomo, E. Solari, G. Toraman, R.
Scopelliti, C. Floriani, M. Latronico, Chem. Commun. 1999,
2413 – 2414.
[10] G. Gil-Ramirez, J. Benet-Buchholz, E. C. Escudero-Adan, P.
Ballester, J. Am. Chem. Soc. 2007, 129, 3820 – 3821.
[11] a) S. Camiolo, P. A. Gale, Chem. Commun. 2000, 1129 – 1130;
b) G. Bruno, G. Cafeo, F. H. Kohnke, F. Nicolo, Tetrahedron
2007, 63, 10003 – 10010.
Experimental Section
Crystal data, a general X-ray diffraction experimental section, ESI-
1
MS experiments, ITC titrations, H NMR titrations, and the single-
crystal X-ray structure of the complex of ClÀ with 2e are available in
the Supporting Information. CCDC-677137, -677138, -677139,
-677140, -677141, -677142, -677143, -677144, and -677145 contain
the supplementary 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.
Received: February 8, 2008
[12] The configurations of 2a–d, 2 f, and 2g were confirmed
unambiguously by X-ray crystallographic analysis.
Published online: April 21, 2008
[13] A. F. Danil De Namor, M. Shehab, J. Phys. Chem. B 2003, 107,
6462– 6468.
[14] Free receptor 2b has a very low solubility in acetonitrile;
however, it is readily extracted in acetonitrile solutions contain-
ing TBACl.
Keywords: anions · calixarenes · hydrogen bonds ·
Pi interactions · receptors
.
[15] Incremental addition of TBACl to an acetonitrile solution
containing 2e produced significant downfield shifts of the proton
signal assigned to the hydroxyl groups, while the pyrrole NH
signals remained at a constant chemical shift. X-ray structure
determination of crystals grown from acetonitrile revealed exo
complexation of the halide with two adjacent OH groups in the
ClÀ·2e complex. A related exo geometry was also reported in
reference [9b] for the complex FÀ·2e.
[16] The proposed binding geometry in solution has been shown in
the solid state for several ClÀ@2 complexes with TBA, TEA, and
TMA as counterions.
[17] K. A. Connors, Binding Constants: The Measurement of Molec-
ular Complex Stability, Wiley, New York, 1987, p. 261.
[1] For the first report on anion–p interactions, see: a) H. J.
Schneider, F. Werner, T. Blatter, J. Phys. Org. Chem. 1993, 6,
590 – 594. The CH···XÀ hydrogen bonding established between
the CH groups of electron-deficient arenes and anions is not a
result of the interaction of the anion with the p-aromatic system:
b) H. Schneider, K. M. Vogelhuber, F. Schinle, J. M. Weber, J.
Am. Chem. Soc. 2007, 129, 13022 – 13026, and reference [3g].
[2] a) D. Quiæonero, C. Garau, C. Rotger, A. Frontera, P. Ballester,
A. Costa, P. M. Deyà, Angew. Chem. 2002, 114, 3539 – 3542;
Angew. Chem. Int. Ed. 2002, 41, 3389 – 3392; b) M. Mascal, A.
Armstrong, M. D. Bartberger, J. Am. Chem. Soc. 2002, 124,
6274 – 6276; c) I. Alkorta, I. Rozas, J. Elguero, J. Am. Chem. Soc.
2002, 124, 8593 – 8598; d) D. Quiæonero, C. Garau, A. Frontera,
Angew. Chem. Int. Ed. 2008, 47, 4114 –4118
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4117

Communications
[18] The average distance values for the N–ClÀ interaction measured
[20] Ka = 2.5 ꢁ 105 mÀ1 (DG = À7.3 kcalmolÀ1) has been reported in
reference [8] for the ClÀ@1 complex in anhydrous acetonitrile
(< 50 ppm H2O).
[21] F. A. Carey, R. J. Sundberg, Advanced Organic Chemistry,
Vol. A, 5th ed., Kluwer Academic/Plenum Publishers, New
York, 2001.
[22] The energies and geometries observed are indicative of an
interaction mainly controlled by electrostatics. No significant
color changes were observed upon anion binding.
in the X-ray structures of the ClÀ@2 complexes of 2d, 2 f
(preliminary data), and 2g are 3.26, 3.23, and 3.26 ꢀ, respec-
tively. A similar distance of 3.28 ꢀ is observed in the ClÀ@1
complex, see reference [9a].
[19] This difference intrinsically includes additional contributions
from desolvation.
4118
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 4114 –4118