A molecular flytrap for the selective binding of citrate and other tricarboxylates in water

Article abstract of DOI:10.1021/ja0433469

The synthesis and binding properties of a new tricationic guanidiniocarbonyl pyrrole receptor 7 are described. Receptor 7 binds citrate 9 and other tricarboxylates such as trimesic acid tricarboxylate 8 with unprecedented high association constants of K

Full text of DOI:10.1021/ja0433469

Published on Web 02/17/2005
A Molecular Flytrap for the Selective Binding of Citrate and
Other Tricarboxylates in Water
Carsten Schmuck* and Michael Schwegmann
Contribution from the Institut fu¨r Organische Chemie, UniVersita¨t Wu¨rzburg,
Am Hubland, 97074 Wu¨rzburg, Germany
Received November 4, 2004; E-mail: schmuck@chemie.uni-wuerzburg.de
Abstract: The synthesis and binding properties of a new tricationic guanidiniocarbonyl pyrrole receptor 7
are described. Receptor 7 binds citrate 9 and other tricarboxylates such as trimesic acid tricarboxylate 8
with unprecedented high association constants of Kassoc > 105 M-1 in water as determined by UV and
fluorescence tritration studies. According to NOESY experiments and molecular modeling calculations,
the tricarboxylates are bound within the inner cavity of receptor 7 by ion pairing between the carboxylate
groups and the guanidiniocarbonyl pyrrole moieties, favored by the nonpolar microenvironment of the cavity.
Hence, receptor 7 can be regarded as a molecular flytrap. In the case of the aromatic tricarboxylate 8,
additional aromatic interactions further strengthen the complex. The complexes with the tricarboxylates
are so strong that even the presence of a large excess of competing anions or buffer salts does not
significantly affect the association constant. For example, the association constant for citrate changes only
from Kassoc ) 1.6 × 105 M-1 in pure water to Kassoc ) 8.6 × 104 M-1 in the presence of a 170-fold excess
of bis-tris buffer and a 1000-fold excess of chloride. This makes 7 one of the most efficient receptors for
the binding of citrate in aqueous solvents reported thus far.
Introduction
new receptor units with improved binding characteristics are
needed.8 We wish to report here the synthesis and binding
The detection of a given analyte by a chemosensor requires
a receptor unit that selectively interacts with the substrate of
choice and a method to read out the binding using a change in
a physical signal.1 In most cases, an additional reporter unit
(e.g., a fluorescent chromophore) is covalently attached to the
receptor for this purpose. Another possibility is to use a “silent”
receptor without an attached read-out device and to employ the
indicator displacement method.2,3 In any case, however, efficient
and strong binding of the analyte by the receptor is a necessary
prerequisite. For applications in aqueous solvents (e.g., under
physiological conditions) this still presents a challenging task.4
For example, pure hydrogen bond-based receptors work only
in organic solvents of low polarity as the strength of hydrogen
bonds decreases rapidly with increasing polarity of the solvent.5
In polar solutions, supramolecular aggregation can even be
endothermic6 and therefore entropy-driven because of the
reorganization of the solvent upon complexation.7 Therefore,
properties of a new tripodal guanidininocarbonyl pyrrole recep-
tor 7, which binds citrate and other tricarboxylates with
(4) Selected recent examples of supramolecular aggregates that are stable in
water: (a) Molt, O.; Ru¨beling, R.; Scha¨fer, G.; Schrader, T. Chem.-Eur.
J. 2004, 10, 4225-4232. (b) Wiskur, S. L.; Lavigne, J. J.; Metzger, A.;
Tobey, S. L.; Lynch, V.; Anslyn, E. V. Chem.-Eur. J. 2004, 10, 3792-
3804. (c) Schmuck, C.; Wienand, W. J. Am. Chem. Soc. 2003, 125, 452-
459. (d) Corbellini, F.; Di Costanzo, L.; Crego-Calama, M.; Geremia, S.;
Reinhoudt, D. N. J. Am. Chem. Soc. 2003, 125, 9946-9947. (e) Rekharsky,
M.; Inoue, Y.; Tobey, S.; Metzger, A.; Anslyn, E. J. Am. Chem. Soc. 2002,
124, 14959-14967. (f) Grawe, T.; Schrader, T.; Zadmard, R.; Kraft, A. J.
Org. Chem. 2002, 67, 3755-3763. (g) Fiammengo, R.; Timmerman, P.;
De Jong, F.; Reinhoudt, D. N. Chem. Commun. 2000, 2313-2314. (h)
Hamilin, B.; Jullien, L.; Derouet, C.; Herve´ du Penhoat, C.; Berthault, P.
J. Am. Chem. Soc. 1998, 120, 8438-8447. (i) Bok Lee, S.; Hong, J.-I.
Tetrahedron Lett. 1996, 37, 8501-8504.
(5) (a) Jeffrey, G. A. An Introduction to Hydrogen Bonding; Oxford University
Press: New York, 1997. (b) Israelachvili, J. Intermolecular and Surface
Forces, 2nd ed.; Academic Press: London, 1992.
(6) (a) Corbellini, F.; Fiammengo, R.; Timmerman, P.; Crego-Calama, M.;
Versluis, K.; Heck, A. J. R.; Luyten, I.; Reinhoudt, D. N. J. Am. Chem.
Soc. 2002, 124, 6569-6575. (b) Schmuck, C. Tetrahedron 2001, 57, 3063-
3067. (c) Prohens, R.; Rotger, M. C.; Pin˜a, M. N.; Deya´, P. M.; Morey, J.;
Ballester, P.; Costa, A. Tetrahedron Lett. 2001, 42, 4933-4936. (d) Hauser,
S. L.; Johanson, E. W.; Green, H. P.; Smith, P. J. Org. Lett. 2000, 2, 3575-
3578. (e) Sebo, L.; Schweizer, B.; Diederich, F. HelV. Chim. Acta 2000,
83, 80-92. (f) Linton, B.; Hamilton, A. D. Tetrahedron 1999, 55, 6027-
6038. (g) Sto¨derman, M.; Dhar, N. J. Chem. Soc., Faraday Trans. 1998,
94, 899-903. (h) Kano, K.; Kitae, T.; Takashimaq, H.; Shimofuri, Y. Chem.
Lett. 1997, 899-900. (i) Meissner, R.; Garcias, X.; Mecozzi, S.; Rebek,
J., Jr. J. Am. Chem. Soc. 1997, 119, 77-85. (j) Kang, J.; Rebek, J., Jr.
Nature 1996, 239-241. (k) Cram, D. J.; Choi, H. J.; Bryant, J. A.; Knobler,
C. B. J. Am. Chem. Soc. 1992, 114, 7748-7765.
(7) In polar solutions, supramolecular aggregation can even be endothermic
and therefore entropy-driven because of the reorganization of the solvent
upon complexation: (a) Schmidtchen, F. P. Org. Lett. 2002, 3, 431-434.
(b) Haj-Zaroubi, M.; Mitzel, N. W.; Schmidtchen, F. P. Angew. Chem.,
Int. Ed. 2002, 41, 104-107. (c) Berger, M.; Schmidtchen, F. P. J. Am.
Chem. Soc. 1999, 121, 9986-9993. (d) Berger, M.; Schmidtchen, F. P.
Angew. Chem., Int. Ed. 1998, 37, 2694-2696. (e) Schiessl, P.; Schmidtchen,
F. P. Tetrahedron Lett. 1993, 34, 2449-2452.
(1) Reviews on chemosensors: (a) Martinez-Manez, R.; Sancenon, F. Chem.
ReV. 2003, 103, 4419-4476. (b) Czarnik, A W.; Yoon, J. Perspect.
Supramol. Chem. 1999, 4, 177-191. (c) Chemosensors of Ion and
Molecular Recognition; Desvergne, J. P., Czarnik, A. W., Eds.; Kluwer:
Dordrecht, The Netherlands, 1997. (d) De Silva, A. P.; Gunaratne, H. Q.
N.; Gunnlaugsson, T.; Huxley, A. J. M.; McCoy, C. P.; Rademacher, J.
T.; Rice, T. E. Chem. ReV. 1997, 97, 1515-1566. (e) Czarnik, A. W. Chem.
Biol. 1995, 2, 423-428. (f) Czarnik, A. W. Acc. Chem. Res. 1994, 27,
302-308.
(2) Wiskur, S. L.; Ait-Haddou, H.; Lavinge, J. J.; Anslyn, E. V. Acc. Chem.
Res. 2001, 34, 963-972.
(3) Recent examples: (a) Nguyen, B. T.; Wiskur, S. L.; Anslyn, E. V. Org.
Lett. 2004, 6, 2499-2501. (b) Piatek, A. M.; Bomble, Y. J.; Wiskur, S. L.;
Anslyn, E. V. J. Am. Chem. Soc. 2004, 126, 6072-6077. (c) Hu, L.; Anslyn,
E. V. J. Am. Chem. Soc. 2004, 126, 3676-3677. (d) Buryak, A.; Severin,
K. Angew. Chem, Int. Ed. 2004, 43, 4771-4774. (e) Tobey, S. L.; Anslyn,
E. V. Org. Lett. 2003, 5, 2029-2031.
9
10.1021/ja0433469 CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 3373-3379
3373

A R T I C L E S
Schmuck and Schwegmann
Scheme 1. Improved Synthesis of the Triamine Template 3
unprecedented high association constants of Kassoc g 105 M-1
in water. Because of the built-in pyrrole chromophore, the
receptor can be used as an optical chemosensor for citrate9
without the need for additional reporter units or the use of a
competitive indicator displacement assay.
much stronger than those of simple guanidinium cations.14
Hence, receptor 7 became a promising candidate to achieve
strong complexation of citrate in water even in the presence of
other competing anions and buffer salts.
The synthesis of 7 is described in Schemes 1 and 2. Template
3 was synthesized in analogy to the literature procedure starting
from the tribromide 1.9b However, to improve the sometimes
troublesome isolation and purification of the corresponding
triamine 3, the literature synthesis was modified (Scheme 1).
The tribromide 1 was treated with concentrated ammonia in
THF/ethanol,15 and the resulting amine was trapped in situ as
the tBoc-carbamate by addition of Boc2O to give 2, which can
be easily isolated by chromatography and stored without prob-
lems. Deprotection with trifluoroacetic acid and anion exchange
by lyophilization with HCl then provided the desired triamine
3 as the hydrochloride salt in 63% yield starting from 1.
This triamine 3 was then reacted with the pyrrole dicarboxylic
acid monobenzyl ester 4 using PyBOP in DMF as the coupling
reagent (Scheme 2). The benzyl ester groups were cleaved off
by hydrogenolysis to yield the triacid 5 in 83% overall yield.
The guanidine group was introduced in form of the mono-tBoc-
guanidine 6 again using PyBOP in DMF as the coupling reagent.
Acidic cleavage of the tBoc-protecting group with TFA gave
the desired receptor 7 as the trifluoracetate salt, which was
lyophilized several times from aqueous HCl to obtain the
chloride salt of the receptor 7 in 72% yield.
NMR Binding Studies. Receptor 7 was designed to bind
citrate 9 in water through a combination of ion pairing and
hydrogen bonds. To probe the effciency of 7 for anion binding,
we first chose trimesic acid tricarboxylate 8 as a substrate, as
both substrate and host have the same threefold symmetry giving
rise to an optimal geometric complementarity between guest
and host binding groups. Upon the addition of trimesic acid
tricarboxylate 8 to a solution of 7 (1 mM, chloride salt) in 10%
H2O in DMSO-d6, significant complexation-induced shift
changes of the amide NH and the pyrrole CHs are observed,
indicating a strong molecular interaction between 7 and 8 (Figure
1). However, only the protons in the binding arms of the receptor
are shifted. The ethyl groups attached to the central benzene
ring do not show any shift change, demonstrating that they do
not participate in the binding of the substrate. A quantitative
analysis of these complexation-induced shift changes shows a
linear decrease until a molar ratio of 1:1 is reached (Figure 2),
Results and Discussion
Design and Synthesis of 7. The design of receptor 7 was
based on the triamine template 3 advertised by Anslyn in his
elegant work on anion sensors.10 Because of the alternate
arrangement of the ethyl and the aminomethylene groups, this
template provides a rather rigid scaffold that presents the three
amino groups all at the same side of the aromatic ring.11 When
suitable receptor units are attached to these amino groups, a
binding cavity with convergent binding sites is formed. In the
past, mainly guanidinium cations and/or boronic acid groups
were used to achieve complexation of various substrates such
as phosphate, citrate, tartrate, or heparine.4b,11 For example,
citrate is bound by a triguanidinium cation based on this template
with an association constant of Kassoc ) 7 × 103 M-1 in pure
water. However, in the presence of a buffer the binding affinity
dropped by nearly 2 orders of magnitude. Therefore, even
stronger binding motifs are needed to allow citrate binding under
physiological conditions () buffered water). We reasoned that
the binding affinity for citrate can be improved by using even
more efficient carboxylate binding sites than simple guandinium
cations.12 For this purpose, we recently introduced guanidinio-
carbonyl pyrroles.13 Because of the increased acidity of the
acylated guanidinium group and additional binding sites in form
of the pyrrole amide, their complexes with carboxylates are
(8) The importance of entropy effects in the design of artificial receptors has
long been overlooked, but probably will become more important in the
future: (a) Calderone, C. T.; Williams, D. H. J. Am. Chem Soc. 2001,
123, 6262-6267. (b) Williams, D. H.; Westwell, M. S. Chem. Soc. ReV.
1998, 27, 57-64.
(9) (a) Metzger, A.; Anslyn, E. V. Angew. Chem., Int. Ed. 1998, 37, 649-
652. (b) Metzger, A.; Lynch, V. M.; Anslyn, E. V. Angew. Chem., Int. Ed.
Engl. 1997, 36, 862-865.
(10) (a) Tobey, S. L.; Anslyn, E. V. Org. Lett. 2003, 5, 2029-2031. (b) Best,
M. D.; Anslyn, E. V. Chem.-Eur. J. 2003, 9, 51-57. (c) Zhong, Z.; Anslyn,
E. V. J. Am. Chem. Soc. 2002, 124, 9014-9015. (d) Wiskur, S. L.; Anslyn,
E. V. J. Am. Chem. Soc. 2001, 123, 10109-10110. (e) Ait-Haddou, H.;
Wiskur, S. L.; Lynch, V. M.; Anslyn, E. V. J. Am. Chem. Soc. 2001, 123,
11296-11297. (f) Niikura, K.; Bisson, A. P.; Anslyn, E. V. J. Chem. Soc.,
Perkin Trans. 2 1999, 2, 1111-1114. (g) Lavinge, J. J.; Anslyn, E. V.
Angew. Chem., Int. Ed. 1999, 38, 3666-3669.
(11) (a) Hennrich, G.; Anslyn, E. V. Chem.-Eur. J. 2002, 8, 2218-2224. (b)
Kilway, K. V.; Siegel, J. S. J. Am. Chem. Soc. 1992, 114, 255-261.
(12) (a) Linton, B. R.; Goodman, M. S.; Fan, E.; Van Arman, S. A.; Hamilton,
A. D. J. Org. Chem. 2001, 66, 7313-7319. (b) Berger, M.; Schmidtchen,
F. P. J. Am. Chem. Soc. 1999, 121, 9986-9993. (c) Linton, B.; Hamilton,
A. P. Tetrahedron 1999, 55, 6027-6038. (d) Dietrich, B.; Fyles, D. L.;
Fyles, T. M.; Lehn, J.-M. HelV. Chim. Acta 1979, 62, 2763-2787. (e)
Dietrich, B.; Fyles, T. M.; Lehn, J.-M.; Pease, L. G.; Fyles, D. L. J. Chem.
Soc., Chem. Comm. 1978, 934-936.
(13) (a) Schmuck, C.; Geiger, L. J. Am. Chem. Soc. 2004, 126, 8898-8899.
(b) Schmuck, C.; Bickert, V. Org. Lett. 2003, 5, 4579-4581. (c) Schmuck,
C.; Geiger, L. Curr. Org. Chem. 2003, 7, 1485-1502. (d) Schmuck, C.
Chem.-Eur. J. 2000, 6, 709-718.
(14) For comprehensive reviews of anion recognition, see the following: (a)
Best, M. D.; Tobey, S. L.; Anslyn, E. V. Coord. Chem. ReV. 2003, 240,
3-15. (b) Gale, P. A. Coord. Chem. ReV. 2003, 240, 191-221. (c)
Fitzmaurice, R. J.; Kyne, G. M.; Douheret, D.; Kilburn, J. D. J. Chem.
Soc., Perkin Trans 1 2002, 841-864. (d) Snowden, T. S.; Anslyn, E. V.
Curr. Opin. Chem. Biol. 1999, 3, 740-746. (e) Beer, P. D.; Schmitt, P.
Curr. Opin. Chem. Biol. 1997, 1, 475-482. (f) Bianchi, A.; Bowman-
James, K.; Garcia-Espan˜a, E. Supramolecular Chemistry of Anions; Wiley-
VCH: New York, 1997. (g) Schmitchen, F. P.; Berger, M. Chem. ReV.
1997, 97, 1609-1646. (h) Seel, C.; Gala´n, A.; deMendoza, J. Top. Curr.
Chem. 1995, 175, 101-132.
(15) Findeis, M. A.; Kaiser, E. T. J. Org. Chem. 1989, 54, 3478-3482.
9
3374 J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005

Selective Binding of Tricarboxylates in Water
A R T I C L E S
Scheme 2. Synthesis of Receptor 7
proving the formation of a 1:1 complex. For example, upon
the addition of 1 equiv of tricarboxylate 8, the pyrrole H7 shifts
from δ ) 7.24 to 6.60, whereas the H6 shifts from δ ) 6.90 to
6.73, respectively (assignment based on NOESY experiments).
The amide NH4 shows a highfield shift from δ ) 8.41 to 8.29
followed by a small downfield shift if more than 1 equiv of
substrate is added, which could reflect the formation of weak
complexes of stoichiometry higher than 1:1.4d
Figure 2. Binding isotherm for the titration of 7 with 8 in 90% DMSO/
Wasser.
These NMR shift changes not only prove the formation of a
1:1 complex between the tripodal receptor 7 and tricarboxylate
8, but the linearity of the binding isotherm also shows that under
Figure 1. NMR shifts for the titration of 7 with 8 in 90% DMSO/Wasser.
9
J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005 3375

A R T I C L E S
Schmuck and Schwegmann
Figure 4. Main binding interactions within the complex between 7 and 8.
The carboxylate interacts with the guanidiniocarbonyl pyrrole cation (dashed
green lines), whereas the two aromatic rings π-stack (black arrow). Observed
NOE signals are shown in blue.
Figure 3. Calculated energy-minimized structure of the complex between
receptor 7 (gray) and tricarboxylate 8 (yellow). Nonpolar hydrogens are
omitted for clarity.
these conditions (10% water in DMSO) the association is too
strong to be measured by NMR techniques, suggesting a binding
constant Kassoc > 105 M-1 16
Unfortunately, at millimolar
.
concentrations as needed for NMR studies, increasing the water
content to >40% led to a precipitation of the complex during
the titration, preventing NMR studies in more polar solvent
mixtures.
Figure 5. Guest binding by receptor 7 resembles a “molecular flytrap”.
pyrrole NH and the amide NH (highlighted in Figure 4). This
binding motif is the same as that we previously observed for
the binding of monocarboxylates by simple monocationic
guanidiniocarbonyl pyrrole receptors.13 This overall binding
scheme is furthermore in good agreement with the observed
NOE signals and shift changes in the NMR.
Complex Structure. Complex formation could also be
proven by NOESY experiments (DMSO, room temperature),
which show cross-peaks between the CH of trimesic acid 8 and
both the pyrrole NH and the amide NH of the receptor 7. Hence,
these cross-peaks not only confirm that the substrate is bound
but also confirm that the guanidiniocarbonyl pyrrole moiety in
the binding arms of receptor 7 adopts a syn conformation with
regard to these NHs. This conformation is not the energetical
favored ground state of receptor 7 (even though in water the
differences in energy between the various conformers are
expected to be rather small according to modeling studies)13d,17
but the one needed for anion binding by a guanidiniocarbonyl
pyrrole. To obtain more information about the most likely
complex structure, we performed molecular modeling calcula-
tions (Macromodel V 8.0, Amber* force field, GB/SA water
solvation treatment).18 A Monte Carlo conformational search
was performed (10 000 steps), and the obtained energy-
minimized complex structure was further subjected to a MD
simulation at 300 Kassoc (10-ps simulation time, 1.5-fs time
steps). The resulting complex structure is shown in Figure 3.
The substrate (yellow) lies atop the benzene ring of the receptor
(gray) within van der Waals distance, probably allowing for an
attractive π-stacking between the electron-rich benzene ring of
the receptor 7 and the electron-poor trimesic acid tricarboxylate
8.19,20 Each carboxylate group is bound by one of the guani-
diniocarbonyl pyrrole arms by ion pair formation with the
guanidinium cation and additional hydrogen bonds from the
According to this calculated structure, receptor 7 can be seen
as a “molecular flytrap” (Figure 5). In the absence of a substrate,
the three arms are pointing away from each other because of
their mutual electrostatic repulsion, providing a more open form
of the receptor. Upon the binding of the substrate, the three
arms close up and completely surround the guest, locking it in
place.
This binding mode also explains the high stability of this
complex, which most likely stems from three factors. First, the
binding properties of the guanidiniocarbonyl pyrrole cation
relative to simple guanidinium or ammonium cations are
superior.14 Second, π-stacking interactions between the two
aromatic rings, one electron-rich and the other electron-poor,
further stabilize the complex. Third, the three binding arms of
7 provide an extensive hydrophobic shielding for the substrate
once it is bound within the inner cavity. Hence, the microen-
vironment around the binding sites is more hydrophobic than
the bulk solvent, favoring both ion pair interactions and
hydrogen bonds. Most parts of the bound guest molecule, and
therefore also its binding sites, are not accessible by the solvent
anymore (Figure 6), as can be seen also from the calculated
solvent accessible surface.
(16) Wilcox, C. S. In Frontiers in Supramolecular Chemistry and Photochem-
istry; Schneider, H. J., Du¨rr, H., Eds.; VCH: Weinheim, Germany, 1990;
pp 123-144.
Spectroscopic Titration Studies in Water. Because of high
complex stability in aqueous DMSO further binding studies were
performed in water using UV-titration studies. At the much
lower concentrations needed for UV studies (∼0.01 mM), the
solubility of both the receptor and the complex was sufficient
even in water. Aliquots of a stock solution of the carboxylate 8
(0.2 mM) were added to a solution of the receptor 7 (0.012
mM). The complexation was monitored by the decrease of the
UV absorbance of the pyrrole moiety at λ ) 300 nm. The
corresponding binding isotherm (Figure 7) was then analyzed
(17) Schmuck, C. Chem. Commun. 1999, 843-844.
(18) Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.; Lipton, M.;
Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J. Comput. Chem.
1990, 11, 440-467.
(19) For examples of recent work on the importance of π-stacking in supramo-
lecular aggregates see, for example: (a) Lahiri, S.; Thompson, J. L.; Moore,
J. S. J. Am. Chem. Soc. 2000, 122, 11315-11319. (b) Sirish, M.; Schneider,
H. J. J. Am. Chem. Soc. 2000, 122, 5881-5882. (c) Guckian, K. M.;
Schweitzer, B. A.; Ren, R. X.-F.; Sheils, C. J.; Tahmassebi, D. C.; Kool,
E. T. J. Am. Chem. Soc. 2000, 122, 2213-2222. (d) Isaacs, L.; Witt, D.;
Fettinger, J. C. Chem. Commun. 1999, 2549-2550.
(20) For a recent review on aromatic interactions, see: Hunter, C. A.; Lawson,
K. R.; Perkins, J.; Urch, C. J. J. Chem. Soc., Perkin Trans. 2 2001, 651-
669.
9
3376 J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005

Selective Binding of Tricarboxylates in Water
A R T I C L E S
Figure 8. Job plot extracted from the UV titration indicating the 1:1
complex stoichiometry.
Figure 6. Solvent-accessible surfaces of receptor (green) and substrate
(yellow) within the complex. The substrate’s carboxylates are completely
shielded from the solvent, allowing for a strong interaction with the
guanidiniocarbonyl pyrrole moieties even in water.
Figure 9. Fluorescence titration of 7 (1.2 × 10-5 M) with citrate 9 (2 ×
10-4 M) in water.
Figure 7. Binding isotherm at λ ) 300 nm for the titration of 7 (1.2 ×
10-5 M) with trimesic acid tricarboxylate 8 (2 × 10-4 M) in water. The
solid line represents the calculated curve fit for the experimental data (9),
whereas the dotted line indicates the expected change in absorption due to
simple dilution of the sample during the titration.
using a nonlinear curve fitting with a 1:1 association model.16,21
This 1:1-complex stoichiometry was confirmed by a Job plot
(Figure 8).22
According to this UV titration, in water at pH ) 6.3 trimesic
acid tricarboxylate 8 is bound by receptor 7 with an association
constant of Kassoc ) 3.4 × 105 M-1! This makes 7 one of the
most efficient carboxylate receptors for aqueous solvents
reported thus far. This surprisingly high binding constant was
independently confirmed by a fluorescence titration study, in
which the decrease of the fluorescence signal at λ ) 335 nm
was monitored. The fluorescence titration provided an associa-
tion constant of Kassoc ) 4.4 × 105 M-1, which is in excellent
agreement with the value obtained from the UV titration. A
clean 1:1 binding stoichiometry was again confirmed by a job
plot.
Figure 10. Binding isotherm at λ ) 300 nm for the titration of 7 (1.2 ×
10-5 M) with citrate 9 (2 × 10-4 M) in water for the fluorescence titration.
The solid line represents the calculated curve fit for the experimental data
(9), whereas the dotted line indicates the expected change in absorption
due to simple dilution of the sample during the titration.
as well. The UV titration provided an association constant of
Kassoc ) 1.6 × 105 M-1, whereas the fluorescence titration
(Figures 9-11) gave Kassoc ) 2.3 × 105 M-1, respectively. To
the best of our knowledge, 7 is hence the most efficient receptor
for the binding of citrate in water reported thus far. With a
binding constant of Kassoc g 105 M-1 in water, the complex is
nearly 2 orders of magnitude more stable than Anslyns’ citrate
receptor (Kassoc ) 7 × 103 M-1). Even in the presence of a
large excess of bis-tris buffer and chloride anions, the binding
Next we tested citrate 9, a more flexible and less symmetric
tricarboxylate, as a substrate, which is however bound nearly
(21) (a) Connors, K. A. Binding Constants; Wiley: New York, 1987. (b)
Fielding, L. Tetrahedron 2000, 56, 6151-6170.
(22) (a) MacCarthy, P. Anal. Chem. 1978, 50, 2165. (b) Job, P. C. R. Acad. Sci.
1925, 180, 928; Ann. Chim. (Paris) (Serie 10) 1928, 9, 113-203; (Serie
11) 1936, 6, 97-144.
9
J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005 3377

A R T I C L E S
Schmuck and Schwegmann
Conclusion
We have presented here a new tripodal receptor 7 that binds
citrate 9 and two tricarboxylates 8 and 10 with unprecedented
high association constants of Kassoc > 105 M-1 in water as could
be shown by UV and fluorescence titration studies. In all cases
a clean 1:1 complex stoichiometry was found. The binding
interactions within the complex are mainly electrostatic, favored
by the hydrophobic environment formed by the cavity of the
receptor. Therefore, structure and flexibility of the carboxylate
substrate play only a minor role in determining complex
stability. More important is the charge complementarity.
Therefore, the binding of citrate is enormously selective relative
to anions such as trifluoroacetate or chloride. Even complexes
with tartrate or maleate are around 1 order of magnitude weaker.
A further variation of the sidearms of 7 by introduction of
additional binding sites should lead to more selective receptors
in the future.
Figure 11. Job plot for the complex formation between receptor 7 and
citrate 9 in pure water as obtained from the changes in fluorescence,
confirming the 1:1 binding stoichiometry.
constant for citrate is still Kassoc ) 8.6 × 104 M-1 (see below),
whereas for example Anslyn’s receptor shows only a rather weak
binding of citrate in buffer solution.
Experimental Section
General Remarks. Reaction solvents were dried and distilled under
argon before use. All other reagents were used as obtained from either
To probe whether the better binding of the aromatic tricar-
boxylate 8 relative to citrate 9 is due to the additional aromatic
π-stacking or the rigidity and threefold symmetry of 8, which
matches the C3V symmetry of the receptor, we also tested Kemps
triacid 10. Kemps triacid is also a rigid tricarboxylate with C3V
symmetry, but it is not aromatic. If rigidity and symmetry are
important, 10 should be bound as well as 8. Kemps triacid 10
is, however, bound even less efficiently than citrate 9. In aqueous
buffer, the association constant for 10 is Kassoc ) 5.1 × 104
M-1 compared to Kassoc ) 8.6 × 104 M-1 for citrate. The better
binding of citrate in this case despite its larger flexibility might
be due to the R-OH group. It was already demonstrated for other
guanidinium hosts that R-OH carboxylates can be even better
guests than a carboxylate.10g
1
Aldrich or Fluka. H and 13C NMR shifts are reported relative to the
deuterated solvents. Peak assignments are based on either DEPT, 2D
NMR studies, and/or comparison with literature data. IR spectra were
recorded using samples prepared as tablets (NaBr). Melting points are
not corrected.
1,3,5-Tris(methylammonium Chloride)-2,4,6-triethylbenzene (3).
To a solution of the tribromide 1 (1.00 g, 2.3 mmol) in THF/ethanol
(80 mL, 1/1), concentrated ammonia (40 mL) was added, and the
reaction mixture was stirred overnight. The solvent was removed under
reduced pressure, and the resulting solid was suspended with NaOH
(0.40 g, 10.0 mmol) in dioxan/H2O (40 mL, 1/1). The suspension was
treated with Boc2O (2.18 g, 10 mmol) in dioxane at 0 °C for 1 h and
then stirred at room temperature for 3 h. The reaction mixture was
extracted with DCM (3 × 30 mL), and the solvent was removed under
reduced pressure. The resulting solid was purified by column chroma-
tography (SiO2, hexane/ethyl acetate/TEA ) 7/3/0.1), yielding a white
solid. This was dissolved in DCM (20 mL), trifluoroacetic (4 mL) acid
was added, and the solution was stirred overnight. After evaporation
to dryness the resulting oil was dissolved in HCl (40 mL, 1 M) and
lyophilized. This procedure was repeated twice to get the chloride salt
3 as a white solid. (0.50 g, 1.4 mmol, 63%); for spectroscopic data see
the literature.9b
1,3,5-Tris(5-ethylcarbamoyl-1H-pyrrole 2-carboxylic acid)-2,4,6-
triethylbenzene (5). To a solution of compound 3 (0.03 g, 0.09 mmol)
and NaOH (0.01 g, 0.25 mmol) in DMF/water (10 mL; 5/1), carboxylic
acid 4 (0.20 g, 0.77 mmol), PyBOP (0.4 g, 0.77 mmol), DMAP (0.04
g, 0.3 mmol), and n-methyl morpholine (0.5 mL) were added, and the
reaction mixture was stirred at room temperature for 3 days under
nitrogen. The brown solution was hydrolyzed with water (20 mL), and
the resulting solid was filtered off. The solution was extracted with
DCM (3 × 20 mL) and combined with the filtered solid, and the solvent
was removed under reduced pressure. The resulting brown oily residue
was purified by column chromatography (SiO2, DCM/methanol ) 20/
1), yielding the product as a white solid.
Substrate Selectivity. Receptor 7 binds citrate 10 with an
unprecedented high affinity. Whereas the two tricarboxylates 8
and 10 are of no relevance in terms of natural occurrence, other
anions such as tatrate, maleate, acetate, or chloride could
interfere with citrate binding. And indeed, chloride anions also
interact with the receptor as shown by the decrease of the UV
absorption of 7 upon addition of NaCl to a solution of the
receptor. However, even in the presence of a 1000-fold excess
of chloride anions, citrate still forms an extremely stable
complex with 7. For example, in 2 mM bis tris buffer (pH )
6.3) with 10 mM sodium chloride receptor 7 (12 µM, trifluo-
racetate salt) binds citrate 9 with Kassoc ) 8.6 × 104 M-1 relatiVe
to all other anions present in the buffer mixture. Trimesic acid
tricarboxylate 8 is bound with Kassoc ) 1.5 × 105 M-1 under
these conditions. Obviously, the tricarboxylate complexes are
so strong even in water, that despite the large excess chloride
binding cannot compete with their formation. This is a major
improvement compared to other citrate receptors for which the
complex stability decreases up to 2 orders of magnitude in the
presence of buffer or other competing anions. Even dicarboxy-
lates such as tartrate (Kassoc ) 7.0 × 103 M-1) or maleate (Kassoc
) 1.1 × 104 M-1) are bound much less efficiently than citrate
as determined by UV titration. Therefore, receptor 7 allows the
complexation of citrate 9 also in water in the presence of a large
excess of other competing anions.
This was dissolved in a mixture of methanol/THF (20 mL, 1/1),
10% Pd/C (0.01 g) was added, and the suspension was stirred at 40 °C
for 3 h under hydrogen atmosphere. The catalyst was filtered off through
a Celite pad and washed with methanol. The filtrate was evaporated to
give the carboxylic acid 5 as a white solid (0.05 g, 0.07 mmol, 82%):
1
mp 209-210 °C dec; H NMR (400 MHz, [D6]DMSO) δ )1.12 (t,
3
9H, J ) 7.4 Hz, CH2-CH3), 2.79-2.81 (m, 6H, CH2-CH3), 4.52-
4.53 (m, 6H, CH2-NH), 6.69-6.70 (m, 3H, pyrrole-CH), 6.76-6.78
(m, 3H, pyrrole-CH), 8.22 (s, 3H, amide-NH), 12.07 (s, 3H, pyrrole-
9
3378 J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005

Selective Binding of Tricarboxylates in Water
A R T I C L E S
NH); 13C NMR (100 MHz, [D6]DMSO) δ ) 16.2 (CH2-CH3), 22.7
(CH2-CH3), 37.2 (CH2-NH), 113.6 (pyrrole-CH), 114.5 (pyrrole-CH),
125.6 (pyrrole-Cq), 130.1 (pyrrole-Cq), 132.4 (aryl-Cq), 143.8 (aryl-
Cq), 158.8, 161.6 (2 carbonyl CO); MS (pos. ESI): m/z ) 683 [M +
Na+]+; HR-MS (pos. ESI): m/z ) 683.244 (calcd. for C33H36N6O9 +
Na+: 683.2441); IR (KBr-pellet) ν˜ ) 3258 (m, br), 2971 (m, br), 1678
(s), 1627 (m), 1559 (s), 1475 (m), 1357 (w), 1270 (s), 1196 (m), 1046
(w), 817 (w), 756 (m) [cm-1].
1,3,5-Tris(5-ethylcarbamoyl-1H-pyrrole-2-carbonylguanidini-
um Chloride)-2,4,6-triethylbenzene (7). To a mixture of the tricar-
boxylic acid 5 (0.05 g, 0.07 mmol) and n-methyl morpholine (0.5 mL)
in DMF (10 mL), PyBOP (0.33 g, 0.63 mmol) and tBoc-guanidine 6
(0.1 g, 0.63 mmol) were added, and the solution was stirred overnight.
The brown reaction mixture was hydrolyzed with water (20 mL), and
the resulting brown solid was filtered off. The solution was extracted
with DCM (3 × 20 mL) and combined with the solid, and the
suspension was evaporated in vacuo. The resulting brown oily residue
was purified by column chromatography (SiO2, DCM/methanol/TEA
) 30/1/0.3), yielding the protected receptor as a white solid.
This was dissolved in DCM (10 mL), and trifluoroacetic acid (2
mL) was added to the suspension. The resulting solution was stirred at
room temperature overnight and then evaporated in vacuo. The pale
yellow oil was dissolved in HCl (20 mL, 1 M) and lyophilized to
remove any trifluoroacetic acid. This procedure was repeated twice to
obtain the chloride salt 7 as a white solid (0.05 mg, 0.05 mmol, 73%):
1
mp 248-249 °C dec; H NMR (400 MHz, [D6]DMSO) δ ) 1.14 (t,
9H, 3J ) 7.4 Hz, CH2-CH3), 2.82-2.84 (m, 6H, CH2-CH3), 4.56 (s,
6H, CH2-NH), 6.91 (s, 3H, pyrrole-CH), 7.48 (s, 3H, pyrrole-CH),
8.42 (bs, 12H, guanidinium-NH2), 8.57 (s, 3H, amide-NH), 11.92 (s,
3H, guanidinium-NH), 12.47 (s, 3H, pyrrole-NH); 13C NMR (100 MHz,
[D6]DMSO) δ ) 16.2 (CH2-CH3), 22.7 (CH2-CH3), 37.2 (CH2-NH),
113.6 (pyrrole-CH), 115.6 (pyrrole-CH), 125.5 (pyrrole-Cq), 131.9
(aryl-Cq), 132.4 (pyrrole-Cq), 143.8 (aryl-Cq), 155.2, 158.5, 159.7 (2
carbonyl CO, guanidinium CN); MS (pos. ESI): m/z ) 784 [M -
2H+]+, 393 [M - H+]2+, 262 [M]3+; HR-MS (pos. ESI): m/z ) 392.690
(calcd. for C36H45N15O6 + 2H+: 392.6917); IR (KBr-pellet) ν˜ ) 3258
(s, br), 2966 (w), 1699 (s), 1636 (m), 1558 (m), 1473 (m), 1274 (s),
1195 (m), 1092 (s), 755 (w), 615 (w), 482 (m) [cm-1].
Acknowledgment. Financial support from the Fonds der
Chemischen Industrie and the Deutsche Forschungsgemeinschaft
is gratefully acknowledged. We thank Dr. Mathias Scha¨fer
(University of Cologne) for measuring the ESI-MS spectra.
JA0433469
9
J. AM. CHEM. SOC. VOL. 127, NO. 10, 2005 3379