Anion-π interaction augments halide binding in solution

Article abstract of DOI:10.1039/b511570a

¹H NMR spectroscopic data and complementary theoretical predictions suggest that a designed receptor exhibits the anion-π interaction in solution. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b511570a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Anion–p interaction augments halide binding in solution{
a
Orion B. Berryman, Fraser Hof, Michael J. Hynes and Darren W. Johnson*
b
c
a
Received (in Columbia, MO, USA) 12th August 2005, Accepted 10th November 2005
First published as an Advance Article on the web 13th December 2005
DOI: 10.1039/b511570a
1
H NMR spectroscopic data and complementary theoretical
predictions suggest that a designed receptor exhibits the anion–
p interaction in solution.
In an attempt to probe the efficacy of the anion–p interaction to
bind anions in solution, two receptor molecules were prepared (1
and 2, Scheme 1). Design of receptor 1 focused on a two point
recognition motif utilizing both a hydrogen bond and an electron-
17
Molecular receptors designed to target anions utilize a variety of
interactions to accomplish their goal. Some of the more common
reversible bonds employed include hydrogen bonding, electrostatic
interactions, hydrophobic effects, and coordination to a metal
deficient aromatic ring. In contrast to 1, control receptor 2 lacked
the electron-deficient aromatic substituent required for the anion–p
interaction. Any enhanced association for anions that receptor 1
exhibits over receptor 2 should be a result of the favorable anion–p
interaction present in the 1?anion complex. To the best of our
knowledge, a neutral receptor molecule designed to incorporate the
anion–p interaction to bind anions in solution is unknown. Herein
we report solution data illustrating the enhanced association for
anions that designed receptor 1 shows over control receptor 2.
Receptor 1 was synthesized by converting o-iodoaniline to the
1–4
ion. A promising binding strategy to target anions that has
recently garnered much attention in the literature is the anion–p
5–9
interaction. Currently, numerous computational studies
6,10–14
single crystal X-ray structures
and
support the viability of using
this noncovalent interaction as a design strategy to target anions.
Several of these reports compare the anion–p interaction to the
5,15
familiar cation–p interaction
18
where a positively charged ion
corresponding p-toluenesulfonamide followed by a palladium-
19,20 1
16
attractively interacts with an electron-rich aromatic ring. The
anion–p interaction is similarly proposed to arise from a negatively
charged species having a coulombic attraction to an area of low
electron density in an electron-deficient aromatic ring. Despite the
numerous solid state examples and theoretical treatments,
surprisingly few solution phase examples recognize the anion–p
mediated Ullmann coupling (Scheme 1).
H NMR spectro-
scopy and single crystal X-ray diffraction confirmed the structure
of receptor 1 (see ESI). A similar procedure provides 2 in 87% yield
starting from 2-aminobiphenyl.
Single crystals of 1 suitable for X-ray diffraction were grown by
diffusing pentane into a chloroform solution of the receptor.{
Receptor 1 crystallizes as a hydrogen bonded dimer in spacegroup
P-1 with two molecules of 1 per unit cell. It is interesting to note
that one sulfonamide oxygen from each receptor molecule is
13
interaction.
˚
located 3.1 A from the electron-deficient aromatic ring of an
21,22
adjacent molecule.
In the crystalline state 1 is preorganized in
the optimal conformation to interact with an anion through both a
hydrogen bond and an anion–p interaction.
Electrostatic potential surfaces (EPS) of molecules have been
used to illustrate areas of low electron density that can interact
7,15
with electron-rich anions.
To highlight both the pre-organiza-
tion of receptor 1 for binding anions and the predictive power that
electrostatic potential surfaces have for the anion–p interaction, the
6 5
Scheme 1 Reagents and conditions: (a) dry pyridine, rt, 4 h; (b) C BrF ,
0
3 4
dry DMSO, Pd(PPh ) , Cu , 105 uC, 5.5 h.
a
Department of Chemistry, 1253 University of Oregon, Eugene, Oregon
7403, USA and the Oregon Nanoscience and Microtechnologies
9
Institute (ONAMI). E-mail: dwj@uoregon.edu; Fax: 541 346 0487;
Tel: 541 346 1695
Department of Chemistry, University of Victoria, P.O. Box 3065,
b
Victoria, British Columbia, Canada V8W 3V6. E-mail: fhof@uvic.ca;
Fax: 250 721 7147; Tel: 250 721 7193
Department of Chemistry, National University of Ireland, Galway,
c
Fig. 1 (a) ORTEP representation of the single crystal X-ray structure of
Ireland. E-mail: michael.j.hynes@nuigalway.ie; Fax: +353 (0)91 525700;
Tel: +353 (0)91 492488
1. Ellipsoids are at the 50% probability level with sulfur yellow, oxygen
red, carbon gray, nitrogen blue, fluorine green and hydrogen light gray. (b)
Calculated EPS plot of receptor 1, scaling areas of highest electron density
(blue) to lowest (white).
{
synthetic procedures, titration experiments and pK
Electronic supplementary information (ESI) available: details of the
a
determinations. See
DOI: 10.1039/b511570a
5
06 | Chem. Commun., 2006, 506–508
This journal is ß The Royal Society of Chemistry 2006

View Article Online
crystal structure of 1 is shown alongside a minimized (CAChe 5.0,
EHT) electrostatic potential surface plot of receptor 1 in the
explanation for this enhanced binding observed with iodide could
be that the more polarizable iodide anion provides a stronger
interaction with the electron-deficient aromatic ring, which
23
optimal conformation for complexing an anion (Fig. 1). The
center of the electron-deficient aromatic ring of receptor 1 (Fig. 1)
exhibits a surface of low electron density (white) that is optimal for
interacting with an anion.
16,27
supplements the weaker hydrogen bond.
An alternative explanation for the differences in measured Ka
values between receptors 1 and 2 could be their variation in
sulfonamide N–H acidities. It is apparent that the electronic
differences between receptors 1 and 2 result in sulfonamide N–H
1
Following the synthesis of receptors 1 and 2, H NMR
spectroscopic titration experiments were performed for each
receptor with the tetra-n-butylammonium salts of chloride,
bromide and iodide. The downfield shifts of the N–H resonances
of 1 and 2 were monitored as aliquots from a stock solution of the
2
8
protons with different pK values. Abraham et al. have shown by
a
comparing a family of receptors, the change in association
constant based on hydrogen bonding acidity is only a fraction of
2
9
3
corresponding salt in CDCl were added to the receptors in
the pK difference between the H-bond donors. Receptors 1 and
a
2
4
CDCl . WinEQNMR was used to fit the raw data to a 1 : 1
3
2 exhibit at least a two order of magnitude difference in anion
association model (see ESI). Iterative calculations using
WinEQNMR yielded the stability constants of the receptors with
binding affinity, while the pK
a
values are only 2.5 log units
of the sulfonamide N–H
different (see ESI). Therefore, the pK
a
25
each anion (Table 1); each reported K for 1 represents the
a
cannot alone explain the difference in association constants for
average of three titrations.
anions. The difference in pK values is an issue with any receptor
a
The titration experiments depict a stark contrast between the
association constants of receptors 1 and 2 with a given halide.
Receptor 1 binds all the halides screened (iodide, bromide and
chloride) with a measurable, albeit modest association constant.
However, in the case of receptor 2—where an electron-deficient
aromatic ring is not present—there is no measurable association
with any of the halides tested. Comparison of the association
constants for the nearly isosteric receptors 1 and 2 allows for an
initial assessment of the anion–p interaction in solution. The
association constants measured for receptor 1 and the series of
halides iodide, bromide and chloride fall in the range of 20–
that utilizes both a hydrogen bond and an anion–p interaction,
therefore new receptors are being synthesized without this feature
to investigate the anion–p interaction further.
While a crystal structure of a 1?anion complex in the proper
orientation remains elusive—presumably a result of the low
association constant of 1 with anions—Fig. 2 illustrates the only
solid state structure of receptor 1 and an anion observed to date.{
In this structure the sulfonamide N–H is drastically rotated out of
conjugation with the adjacent phenyl ring in order to form a very
weak hydrogen bond with a bromide ion (avg. Br–H–N angle =
129.7u). This conformation precludes formation of an anion–p
21
3
4 M . The measured association constants are similar in
interaction and is adopted to allow for a polymeric ion pair
between tetra-n-butylammonium counterions and the bromide
anion to form in the crystal state (Fig. 2, inset). This solid state
structure likely does not represent the solution structure of 1 when
bound to an anion: first, Hartree–Fock (6-31+G*) geometry
optimizations for the complex of each receptor 1 and 2 with
chloride showed two different energy minima for both receptors.
Receptor 1 exhibited one minimum with the chloride over the face
of the electron-deficient aromatic ring (Fig. 3a) and one minimum
with the chloride positioned to the side of the electron-deficient
aromatic ring. Receptor 2 on the other hand showed no reasonable
structure positioning the halide over the face of the aromatic ring
13
magnitude with reported anion–p binding constants. Receptor 2
on the other hand shows a significantly weaker binding to the
same halides. The change in chemical shift from the titration
experiments was so small for receptor 2 that no association
constant could be determined. The association constants for
receptors 1 and 2 provide strong support demonstrating the anion–
p interaction in solution, highlighting the possibility of utilizing the
anion–p interaction to bind anions by design.
The trend in association constant strength for receptor 1 and the
halides chloride, bromide and iodide cannot solely be explained by
the strength of the hydrogen bond with each anion. In a similar
system containing only a sulfonamide substituent (but no adjacent
aromatic ring), it was observed that chloride forms the strongest
association by an order of magnitude whereas bromide and iodide
(Fig. 3b). These calculations suggest that the chloride anion is
26
are significantly weaker and equal to each other. However,
receptor 1, which contains an electron-deficient aromatic ring
designed to interact with anions, shows a deviation from this trend
with the most polarizable anion (iodide) exhibiting a comparable
association to that of the more basic chloride ion. One plausible
2
1
a
Table 1
K
a
(M ) for receptors 1 and 2 with selected halides
b
b
Receptor 1
Receptor 2
2
c
Cl
Br
I
30 ¡ 3
20 ¡ 2
34 ¡ 6
, 1c
2
, 1c
, 1
b
2
a
4+
The NEt
salts of each halide were used. Initial receptor
Fig. 2 Stick representation of the X-ray crystal structure of receptor 1
and tetra-n-butylammonium bromide where the ion pairing in the solid
state is forcing the sulfonamide N–H away from the preferred
conformation. The inset shows the polymeric ion pair formed in the solid
state.
concentrations fall in the range of 9–25 mM; full details of the
c
titration experiments are contained in the ESI.
constants for receptor 2 were too small to be determined by
Association
1
H
NMR titration experiments.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 506–508 | 507

View Article Online
1
2
P. D. Beer and P. A. Gale, Angew. Chem., Int. Ed., 2001, 40, 486.
A. P. Bisson, V. M. Lynch, M.-K. C. Monahan and E. V. Anslyn,
Angew. Chem., Int. Ed., 1997, 36, 2340.
J. L. Sessler, S. Camiolo and P. A. Gale, Coord. Chem. Rev., 2003, 240,
3
4
5
6
1
7.
A. Bianchi, K. Bowman-James and E. Garcia-Espana, Editors,
Supramolecular Chemistry of Anions, Wiley-VCH, New York, 1997.
D. Quinonero, C. Garau, C. Rotger, A. Frontera, P. Ballester, A. Costa
and P. M. Deya, Angew. Chem., Int. Ed., 2002, 41, 3389.
A. Frontera, F. Saczewski, M. Gdaniec, E. Dziemidowicz-Borys,
A. Kurland, P. M. Deya, D. Quinonero and C. Garau, Chem. Eur. J.,
2
005, 11, 6560.
M. Mascal, A. Armstrong and M. D. Bartberger, J. Am. Chem. Soc.,
002, 124, 6274.
D. Kim, P. Tarakeshwar and K. S. Kim, J. Phys. Chem. A, 2004, 108,
250.
C. Garau, A. Frontera, D. Quinonero, P. Ballester, A. Costa and
P. M. Deya, J. Phys. Chem. A, 2004, 108, 9423.
0 P. de Hoog, P. Gamez, W. L. Driessen and J. Reedijk, Tetrahedron
7
8
9
Fig. 3 HF geometry minimizations of receptors 1 and 2 with chloride
using a 6-31+G* basis set. (a) Energy minimization of receptor 1 with
chloride (green) showing the halide over the face of the electron-deficient
aromatic ring. (b) One of the representative minimizations of receptor 2
with chloride (green) not positioned over the face of the aromatic ring.
2
1
1
Lett., 2002, 43, 6783.
1 P. de Hoog, P. Gamez, I. Mutikainen, U. Turpeinen and J. Reedijk,
repulsed by the aromatic ring in receptor 2 while being less
repulsed by the electron-deficient aromatic ring of receptor 1.
Secondly, the polymeric ion pair observed in the crystalline state
1
Angew. Chem., Int. Ed., 2004, 43, 5815.
2 S. Demeshko, S. Dechert and F. Meyer, J. Am. Chem. Soc., 2004, 126,
1
is clearly not maintained in CDCl
3
solutions. Furthermore, this ion
4508.
3 Y. S. Rosokha, S. V. Lindeman, S. V. Rosokha and J. K. Kochi,
1
1
pair forming in solution would not explain the H NMR chemical
shift dependence of the N–H resonance of 1 on anion concentra-
tion (Table 1). Additionally, since receptors 1 and 2 are nearly
isosteric, 2 could equally-well adopt this twisted conformation to
bind anions in solution. If this were the case, 1 and 2 would have
very similar binding constants for anions, which they do not.
Despite this perplexing structure, the solution data are best
corroborated by invoking an attractive anion–p interaction when 1
binds anions.
Angew. Chem., Int. Ed., 2004, 43, 4650.
4 J. Pang, Y. Tao, S. Freiberg, X.-P. Yang, M. D’Iorio and S. Wang,
1
J. Mater. Chem., 2002, 12, 206.
15 C. Garau, A. Frontera, D. Quinonero, P. Ballester, A. Costa and
P. M. Deya, Chem. Phys. Chem., 2003, 4, 1344.
16 J. C. Ma and D. A. Dougherty, Chem. Rev., 1997, 97, 1303.
17 A hydrogen bonding interaction is included in our first receptor design
1
to enhance the association of the receptor for anions. H NMR and
UV-Vis spectroscopic studies of simple electron deficient aromatics such
as hexafluorobenzene, 1,3,5-trifluorobenzene, 1,3,5-tribromobenzene
and 1,3,5-trichlorobenzene showed no measurable binding with the
halides. This suggested that a single anion–p interaction using a
halogenated aromatic would not be strong enough to drive binding in
solution.
The solution data presented herein underscore the hypothesis
that electron-deficient aromatics can be used as a component of a
design strategy to target anions in solution. A pair of receptors that
differ only by the substituents on their aromatic rings were
designed and synthesized. Analysis of the association constants of
each receptor with an array of halides has demonstrated the
enhanced affinity for anions that receptor 1 shows in solution over
a control receptor lacking an electron-deficient aromatic ring. The
enhanced binding that receptor 1 displays results from an
attractive anion–p interaction. These experiments support the use
of the anion–p interaction as an emerging noncovalent interaction
for the selective targeting of anions in solution.
18 T. Sakamoto, Y. Kondo, S. Iwashita, T. Nagano and H. Yamanaka,
Chem. Pharm. Bull., 1988, 36, 1305.
9 N. Shimizu, T. Kitamura, K. Watanabe, T. Yamaguchi, H. Shigyo and
1
T. Ohta, Tetrahedron Lett., 1993, 34, 3421.
20 M. G. Banwell, B. D. Kelly, O. J. Kokas and D. W. Lupton, Org. Lett.,
2003, 5, 2497.
21 The X-ray crystal structure of 1 is in agreement with recent theoretical
studies regarding the attractive interaction between the lone pair of a
heteroatom and an electron-deficient aromatic ring.
22 I. Alkorta, I. Rozas and J. Elguero, J. Org. Chem., 1997, 62, 4687.
2
2
2
3 CAChe, version 5.0, Fujitsu America, Beaverton, USA, 2002.
4 M. J. Hynes, J. Chem. Soc., Dalton Trans., 1993, 311.
5 The dimerization constants for receptors 1 and 2 were measured by H
We acknowledge Dr Elisabeth Rather Healey for help in solving
the crystal structures reported herein. Professors Michael M. Haley,
David R. Tyler and John F. W. Keana are acknowledged for
helpful advice. We thank the NSF for an IGERT fellowship
1
NMR titrations (see method described in: A. P. Bisson, C. A. Hunter,
J. C. Morales and K. Young, Chem. Eur. J., 1998, 4, 845). Receptor 1
21
exhibits weak dimerization (K
a
3
y 2 M ) in CDCl , whereas receptor 2
shows no measurable dimerization. Attempts to correct anion binding
constants for the dimerization of receptor 1 using either WinEQNMR
or the method reported by Hunter et al. failed to produce acceptable fits
to the data with reasonable precision. Presumably, the weak dimeriza-
tion in receptor 1 plays a negligible role in the binding of anions (see
ESI).
(
OBB) and the University of Oregon for generous support of this
research.
Notes and references
2
6 O. B. Berryman and D. W. Johnson, unpublished results, 2005.
{
Crystal data 1: C19
12 5 2
H F NO S, M = 413.36, triclinic, P-1, a = 9.0547(11),
˚
27 D. Quinonero, C. Garau, A. Frontera, P. Ballester, A. Costa and
P. M. Deya, Chem. Phys. Lett., 2002, 359, 486.
28 M. H. Abraham, P. L. Grellier, D. V. Prior, P. P. Duce, J. J. Morris and
P. J. Taylor, J. Chem. Soc., Perkin Trans. 2, 1989, 699.
b = 10.2933(12), c = 10.4106(13) A, a = 97.125(2)u, b = 112.783(2)u, c =
3
21
˚
101.424(2)u, V = 854.95(18) A , Z = 2, m(Mo-Ka) = 0.257 mm . Final
residuals (242 parameters) R1 = 0.0818 for 3444 reflections with I > 2s(I),
and R1 = 0.1570, wR2 = 0.2413, GooF = 1.030 for all 6901 data. CCDC
2
61862. 1?(n-Bu
4
N?Cl): C35
H
47BrF
5
N
2
O
˚
2.5S, M = 742.72, triclinic, P-1, a =
29 Reference 28 presents a scale of hydrogen bond acidities based on log K
values. While the sulfonamide functionality is not specifically addressed,
data for other acidic functionalities are presented with the equation log
1
0.863(4), b = 16.807(6), c = 21.000(7) A, a = 74.118(6)u, b = 83.011(6)u, c =
3
2
1
˚
89.974(6)u, V = 3658(2) A , Z = 4, m(Mo-Ka) = 1.240 mm . Final
residuals (498 parameters) R1 = 0.0916 for 4148 reflections with I > 2s(I),
and R1 = 0.1400, wR2 = 0.2477, GooF = 1.013 for all 6598 data. CCDC
286064. For crystallographic data in CIF or other electronic format see
DOI: 10.1039/b511570a
K = L
are constants specific for a given family of molecules. Estimates using
the upper limit of these constants result in predicted K values for
b
pK
a
b a a b
+ D relating pK and hydrogen bonding K , where L and
D
b
a
1
receptor 2 sufficient to be detected by H NMR spectroscopy.
This journal is ß The Royal Society of Chemistry 2006
5
08 | Chem. Commun., 2006, 506–508