Pure C-H hydrogen bonding to chloride ions: A preorganized and rigid macrocyclic receptor

Article abstract of DOI:10.1002/anie.200704717

Strength in numbers: When traditionally weak C-H hydrogenbonds are preorganized in a shape-persistent macrocycle, a strong affinity for chloride ions emerges. This discovery arises from the dual use of click chemistry for efficient macrocyclizationan d the favorable steric properties of the resulting 1,2,3-triazoles. (Figure Presented)

Full text of DOI:10.1002/anie.200704717

Angewandte
Chemie
DOI: 10.1002/anie.200704717
Anion Receptors
À
Pure C H Hydrogen Bonding to Chloride Ions: A Preorganized and
Rigid Macrocyclic Receptor**
Yongjun Li and Amar H. Flood*
Anions play important chemical and biological roles,[1] and
the supramolecular chemistry involving them[2] has grown
accordingly.[3] These developments range from fundamental
studies on noncovalent bonding[4] to the design of anion
receptors[2,5,6] and anion assistance to asymmetric catalysis.[7]
A substantial number of artificial organic hosts for anions
macrocycle displays a high affinity for chloride ions supported
À
À
solely by C H···Cl H-bonds.
In a structural context, the planar 1,2,3-triazole ring
system defines a nonlinear interconnection geometry. The
triazole bears one sterically active C5-H hydrogen. With these
features and the hypothesized coplanarity in mind, a com-
puter-aided approach using molecular mechanics (MM2) was
employed in the design of a planar macrocycle 1 (Scheme 1).
Eight ring systems are linked together, wherein four triazoles
alternate with four phenyl rings. This design allows for
minimal deviations from ideal bond angles and minor steric
interactions between hydrogens on adjacent ring systems. The
À
incorporate strong hydrogen bonds (H-bonds), usually N
À
H···XÀ,[8] e.g., calixpyrrole, or if C H···X H bonds are used,
À
the carbon atom is adjacent to a cationic center, e.g.,
imidazolium.[9] The strength and the relative significance of
À
À
weak C H···X H bonds originating from neutral ring systems
has been examined recently.[10] Although stronger than
previously thought, these H-bonds were only considered as
“additional binding sites within a host cavity.”[10b] It was
surprising, therefore, to discover a break from this dogma
when strong chloride binding was displayed by a novel
macrocycle described herein, in which there are only aromatic
À
C H···Cl H-bonds.[11] This example appears to strengthen
À
further Cramꢀs idea[12] that a rigid host in which each binding
site is preorganized[13] should show the strongest guest
binding.
The design for this receptor was originally motivated from
the ongoing synthetic developments[14] in shape-persistent
macrocycles together with the recognized efficiency of the
copper-catalyzed[15] 1,3-dipolar cycloaddition of terminal
azides and acetylenes to generate the five-membered 1,2,3-
triazole ring system. In recent studies,[16] these triazoles were
distinguished from other ring systems by the relief of steric
strain provided by the sp2 nitrogen atoms in the ring system.
This observation stimulated the prediction, elaborated herein,
that 1,4-disubstitution of triazoles with phenyl rings will
produce inter-ring coplanarity that would establish a favor-
able building block for the preparation of planar macrocycles.
To that end, we developed the efficient synthesis (27% yield,
seven steps), using click chemistry, of a shape persistent
macrocycle incorporating phenylene subunits. It was on the
basis of this work that the discovery was made that the
[*] Dr. Y. Li, Prof. A. H. Flood
Chemistry Department
Indiana University
800 East Kirkwood Avenue, Bloomington, IN 47405 (USA)
Fax: (+1)812-855-8300
E-mail: aflood@indiana.edu
Scheme 1. Synthesis and partial retrosynthetic analysis (inset) of 1: a)
ꢀ
TMSC CH, [PdCl2(PPh3)2], CuI, iPr2NH, THF, Ar, 8 h; b) NaN3, CuI,
[**] We acknowledge support from NIH 1S10RR016657-01 for the
Indiana University Mass Spectrometry Facility.
DMEA, sodium ascorbate, EtOH/H2O/PhMe (7:3:1), reflux, Ar, 1 h; c)
KF, MeOH, THF, 8 h; d) CuSO4, sodium ascorbate, EtOH/H2O/PhMe
(7:3:1), Ar; e) dropwise addition of 3 and 6b over 10 h to CuI, DBU,
PhMe, Ar, stirring, 4 h. TMS=trimethylsilyl; DMEA=N,N’-dimethylen-
thylenediamine; DBU=1,8-diazabicyclo[5.4.0]undec-7-ene.
Supporting information for this article, including details of the
experimental procedures, is available on the WWW under http://
www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2008, 47, 2649 –2652
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2649

Communications
structure is two-fold symmetric, with one pair of facing phenyl
added dropwise over 10 h to a heated solution (708C) of CuI
(0.03 mmol) with the base DBU (1,8-diazo-
rings having N1-bound triazoles, whereas the other pair are
C4-bound. The calculated size of the central cavity of 1 (HF/3-
21G*) is determined by the 5.38, 5.90, and 5.96-Š internuclear
distances between facing pairs of hydrogen atoms on triazole
groups, and on the nitrogen-linked and carbon-linked phenyl
groups, respectively. These endocyclic hydrogen atoms define
a space-filling cavity with a diameter of about 3.8 Š, sufficient
to contain a chloride ion (ionic radius = 1.81 Š).[2,17] Based on
the shorter of the three interhydrogen distances, the calcu-
lated distance from the triazole hydrogen to the center of the
[5.4.0]bicycloundec-7-ene) in PhMe. During the addition of
each drop (ca. 200 mL) to the reaction mixture, there is an
approximate 200-fold excess of CuI catalyst. This situation is
believed to favor the formation of a copper-acetylide[20] of 6b
that can react with 3 under the pseudo-high-dilution con-
ditions to generate, by an intermolecular coupling, the 7/8
premacrocyclic intermediate followed by the final unimolec-
ular cyclization. Initially, the product was isolated in a 50%
yield after chromatographic separation (SiO2) as a white
solid. An improved yield, 70%, which was obtained after
exhaustive degassing of the reaction mixture with argon,
indicates that the macrocyclization does in fact occur faster
than the reaction of a terminal copper acetylide with another
incoming (di)azide.
À
cavity is 2.69 Š. This is long enough to support a linear C
À
À
H···Cl interaction based on an analysis of 8537 observed C
H···ClÀ interactions[18] from which circa 11% display similar
or shorter interatomic distances. Finally, a range of substitu-
ents on the phenyl ring were screened computationally from
which it was hypothesized that meta substitution with tert-
butyl groups would enhance solubility and inhibit p-stacking
while retaining planarity.[14]
The identity of the macrocycle was confirmed from a
single peak in the MALDI-TOF mass spectrum with m/z
1
797.6. The H NMR spectrum (CD2Cl2) displays five well-
A convergent synthesis by a “5/8” pre-macrocyclic
oligomer, wherein five of the eight ring systems are present,
was considered as the initial cyclization strategy.[14] Macro-
cyclization will be facilitated if the immediate precursor to 1 is
preorganized. To investigate the energy landscape leading to
1, a model of this likely precursor was obtained by discon-
necting the macrocycle through one of the triazole linkages to
generate a “7/8” oligomer with a terminal alkyne and azide
functionality (Scheme 1, inset). A Monte Carlo simulation
(MM2) of the possible conformations returned 66 low-energy
conformers, each showing little deviation from inter-ring
coplanarity. The single conformer (50th based on relative
energy) that is preorganized to resemble the macrocycle is
seen at + 2.3 kcalmolÀ1. Although the statistical population of
this conformer will be relatively low, we considered employ-
ing pseudo-high-dilution reaction conditions to mitigate this
nonideal situation. Furthermore, the energy profile (HF/3-
21G*) generated by rotating one of the N1- or C4-bound
triazole groups past the central phenylene unit in both N1- and
C4-bound 1,3-bis(triazole)benzene models produces torsion
barriers of 6 and 4 kcalmolÀ1, respectively (see Supporting
Information). These analyses suggested that, for the hypo-
thetical 7/8 premacrocyclic intermediate, this potential energy
surface could be rapidly traversed, allowing the reaction to
funnel towards the final copper-catalyzed macrocyclization
rather than generating oligomeric byproducts.[14c]
separated singlets in the aromatic region (Figure 1). The
single downfield peak at d = 9.59 ppm corresponds to the four
A stepwise approach (Scheme 1) was employed to pre-
pare the macrocycle 1. The diiodo compound 2[19] served as an
excellent starting material to access building blocks 4a,b,
using Sonogashira conditions, as well as the diazide 3. The
latter diazide is itself utilized in two different steps. The first,
using click chemistry conditions, with the trimethylsilyl-
protected monoethynyl 4b leads to the diiodo 5/8 oligomeric
intermediate 5, which was converted into the diethynyl 5/8
premacrocycle using standard Sonogashira conditions and
deprotected to afford 6b in excellent yields. The second use of
diazide 3 is in the final one-pot intermolecular coupling
followed by unimolecular cyclization to generate 1. Pseudo-
high-dilution conditions are employed for the macrocycliza-
tion, wherein a 1:1 mixture of 3 and 6b (0.2 mmol, PhMe) is
Figure 1. 1H NMR spectra (aromatic region) of 1 upon titration with
tetrabutylammonium chloride (TBACl); CD2Cl2, 298 K.
H5 triazole hydrogen atoms residing in identical chemical
environments. This position is significantly shifted 1.27 ppm
downfield compared to 6b (see Supporting Information). This
shift arises when the triazole hydrogen atoms are located
within the greater deshielding environment generated by the
ring currents of all four phenyl rings. The 1H NMR data were
assigned with the aid of 2D experiments (see Supporting
Information), and are consistent with the two-fold symmetry.
The through-space couplings measured using 2D NMR
2650 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 2649 –2652

Angewandte
Chemie
(ROESY) spectroscopy are consistent with the predicted
planar structure, although deviations from complete planarity
cannot be excluded at this stage.
The host–guest chemistry between the macrocycle and the
chloride anion is immediately detected by a dramatic increase
in the solubility of the macrocycle beyond 1 mm (CH2Cl2)
upon the addition of tetrabutylammonium chloride (TBACl).
More quantitatively, a solution of 1 (2 mm, CD2Cl2) was
titrated with up to 10 equivalents of TBACl with concomitant
downfield shifts (Figure 1) in all aromatic hydrogen signals.
The triazole hydrogen signal shifts downfield by 0.73 ppm, the
endocyclic hydrogen atoms of the N1- and C4-linked phenyl
units shift 0.92 and 0.87 ppm, respectively. The exocyclic
hydrogen atoms of the phenyl rings distal from the central
cavity shift only half as much (ca. 0.4 ppm). The retention of
the simple peak pattern and the similar magnitude of the
shifts in all the endocyclic hydrogen atoms are consistent with
the macrocycle maintaining its shape[14] and with the local-
ization of the chloride ion inside the cavity. A geometry
optimization (HF/3-21G*) supports this interpretation and
shows that the chloride ion is held within the plane of the
macrocycle, which is consistent with the circa 3.7-Š cavity
diameter. The observation of extensive line broadening after
the addition of one equivalent of the chloride ion is consistent
with binding dynamics occurring on the same timescale as the
1H NMR experiment. A preliminary variable temperature
analysis (see Supporting Information) places the barrier for
this process at around 14.5kcalmol À1
.
The binding strength of 1 for chloride ions was deter-
mined by recording the changes (Figure 2a) in the UV
absorption (265nm) as a function of chloride-ion concen-
tration. Based on the Job-plot determination of 1:1 binding
stiochiometry, the data were fitted to a binding isotherm
(Figure 2b) providing the calculated association constant of
Figure 2. a) UV titration of 1 with TBACl (CH2Cl2, 100 mm) and b) the
fitting analysis (Job Plot: inset) obtained from UV spectroscopy.
chloride ions. This discovery highlights the broader strategy of
bringing efficient reaction chemistries into synergy with the
intrinsic properties of the resulting chemical functionality, in
this case, the 1,2,3-triazole ring system.
Ka = (130000 Æ 30000)mÀ1
;
DG = À7.0 kcalmolÀ1 (CH2Cl2,
298 K). Potential aggregation effects[14] were excluded on
the basis of a UV dilution study (see Supporting Information),
substantiating 1:1 binding. The strength of the binding is
comparable to a wide range of receptors.[5,6,8,9] What is
Received: October 12, 2007
Revised: December 16, 2007
Published online: February 25, 2008
remarkable, however, is that this association occurs solely
À
À
with the use of C H···Cl H-bonds, and arises from within a
5
À
neutral receptor. Triazole-based C H H-bonds or close
Keywords: anions · coordination chemistry · hydrogen bonds ·
contacts have been seen in many solid-state structures.[21] In
those instances, and presumably here, the H-bonds are aided
.
macrocyclic ligands · molecular modeling
by the relatively large 5-Debye dipole,[15c,22] with its positive
5
À
end directed almost in line with C H bond. Furthermore, on
[1] G. R. Desiraju, T. Steiner, The Weak Hydrogen Bond in
Structural Chemistry and Biology, Oxford University Press,
Oxford, 1999.
[2] a) Supramolecular Chemistry of Anions (Eds.: A. Bianchi, K.
Bowman-James, E. García-Espaæa), Wiley-VCH, New York,
1997; b) Special issue on 35Years of Synthetic Anion Receptor
Chemistry 1968–2003, see: Coord. Chem. Rev. 2003, 240, 1 – 221.
[3] a) K. Bowman-James, Acc. Chem. Res. 2005, 38, 671 – 678;
b) P. A. Gale, R. Quesada, Coord. Chem. Rev. 2006, 250, 3219 –
3244.
[4] C. A. Hunter, Angew. Chem. 2004, 116, 5424 – 5439; Angew.
Chem. Int. Ed. 2004, 43, 5310 – 5324.
[5] K.-J. Chang, D. Moon, M. S. Lah, K.-S. Jeong, Angew. Chem.
2005, 117, 8140 – 8143; Angew. Chem. Int. Ed. 2005, 44, 7926 –
7929.
account of the shape persistence, there is a negligible energy
cost that is often associated with a change in the conformation
of the receptor.[12] The dominant role of preorganization was
confirmed using a model compound, produced by click
chemistry between 6b and phenylazide, which contains the
four triazole units but delivers significantly reduced binding
affinity (Ka = 7mÀ1).
In summary, good yields of a shape-persistent macrocycle
have been established based upon a sequence of Sonogashira
and click-chemistry reactions. The collective effect of the
preorganized shape of the macrocycle, the ideal size of the
À
inner cavity, and the cumulative result of multiple weak C H
hydrogen bonds determine a relatively strong affinity for
Angew. Chem. Int. Ed. 2008, 47, 2649 –2652
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2651

Communications
[6] I. Hisaki, S.-I. Sasaki, K. Hirose, Y. Tobe, Eur. J. Org. Chem.
[14] a) C. Grave, A. D. Schlꢁter, Eur. J. Org. Chem. 2002, 3075– 3098;
b) S. Hꢂger, Chem. Eur. J. 2004, 10, 1320 – 1329; c) W. Zhang,
J. S. Moore, Angew. Chem. 2006, 118, 4524 – 4548; Angew. Chem.
Int. Ed. 2006, 45, 4416 – 4439.
[15] a) V. V. Rostovtsev, L. G. Green, V. V. Fokin, K. B. Sharpless,
Angew. Chem. 2002, 114, 2708 – 2711; Angew. Chem. Int. Ed.
2002, 41, 2596 – 2599; b) C. W. Tornoe, C. Christensen, M.
Meldal, J. Org. Chem. 2002, 67, 3057 – 3064; c) W. S. Horne,
M. K. Yadav, C. D. Stout, M. R. Ghadiri, J. Am. Chem. Soc. 2004,
126, 15366 – 15367; d) K. D. Bodine, D. Y. Gin, M. S. Gin, J. Am.
Chem. Soc. 2004, 126, 1638 – 1639; e) Y. Angell, K. Burgess, J.
Org. Chem. 2005, 70, 9595 – 9598.
[16] a) Y. Li, J. C. Huffman, A. H. Flood, Chem. Commun. 2007,
2692 – 2694; b) B. M. J. M. Suijkerbuijk, B. N. H. Aerts, H. P.
Dijkstra, M. Lutz, A. L. Spek, G. van Koten, R. J. M. K.
Gebbink, Dalton Trans. 2007, 1273 – 1276; c) R. M. Meudtner,
M. Ostermeier, R. Goddard, C. Limberg, S. Hecht, Chem. Eur. J.
2007, 13, 9834 – 9840.
[17] Halide selectivity will be the topic of future studies.
[18] L. Brammer, E. A. Bruton, P. Sherwood, Cryst. Growth Des.
2001, 1, 277 – 290.
2007, 607 – 615.
[7] a) M. S. Taylor, E. N. Jacobsen, Angew. Chem. 2006, 118, 15 5 0 –
1573; Angew. Chem. Int. Ed. 2006, 45, 1520 – 1543; b) G. L.
Hamilton, E. J. Kang, M. Mba, F. D. Toste, Science 2007, 317,
496 – 499.
[8] a) K. A. Nielsen, W.-S. Cho, J. Lyskawa, E. Levillain, V. M.
Lynch, J. L. Sessler, J. O. Jeppesen, J. Am. Chem. Soc. 2006, 128,
2444 – 2451; b) H. Maeda, Y. Kusunose, Y. Mihashi, T. Mizogu-
chi, J. Org. Chem. 2007, 72, 2612 – 2616.
[9] a) K. Chellappan, N. J. Singh, I.-C. Hwang, J. W. Lee, K. S. Kim,
Angew. Chem. 2005, 117, 2959 – 2963; Angew. Chem. Int. Ed.
2005, 44, 2899 – 2903; b) N. Alhashimy, D. J. Brougham, J.
Howarth, A. Farrell, B. Quilty, K. Nolan, Tetrahedron Lett.
2007, 48, 125– 128.
[10] a) G. R. Desiraju, Acc. Chem. Res. 2002, 35, 565 – 573; b) V. S.
Bryantsev, B. P. Hay, J. Am. Chem. Soc. 2005, 127, 8282 – 8283.
[11] For comparison, there are only two examples of wholly non-
À
aromatic C H···anion interactions, see: a) W. B. Farnham, D. C.
Roe, D. A. Dixon, J. C. Calabrese, R. L. Marlow, J. Am. Chem.
Soc. 1990, 112, 7707 – 7718; b) S. S. Zhu, H. Staats, K. Brand-
horst, J. Grunenberg, F. Gruppi, E. Dalcanale, A. Lꢁtzen, K.
Rissanen, C. A. Schalley, Angew. Chem. Int. Ed. 2008, 47, 788 –
792.
[19] Y. Miura, Y. Ushitani, K. Inui, Y. Teki, T. Takui, K. Itoh,
Macromolecules 1993, 26, 3698 – 3701.
[20] V. D. Bock, H. Hiemstra, J. H. van Maarseveen, Eur. J. Org.
Chem. 2006, 51 – 68.
[12] D. J. Cram, Angew. Chem. 1988, 100, 1041 – 1052; Angew. Chem.
Int. Ed. Engl. 1988, 27, 1009 – 1020.
[13] a) M. D. Best, S. L. Tobey, E. V. Anslyn, Coord. Chem. Rev.
2003, 240, 3 – 15; b) C. A. Ilioudis, D. A. Tocher, J. W. Steed, J.
Am. Chem. Soc. 2004, 126, 12395– 12402.
[21] A CCDC search (v. 5.28, Nov. 2006) returned 28 structures
containing 1,2,3-triazole, of which 19 showed close contacts to
5
À
the C H hydrogen.
[22] M. H. Palmer, R. H. Findlay, A. J. Gaskell, J. Chem. Soc. Perkin
Trans. 2 1974, 420 – 428.
2652
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 2649 –2652