Zwitterionic guanidinium compounds serve as electroneutral anion hosts

Article abstract of DOI:10.1021/ja992028k

The complexation of anions in solvating phases by artificial receptors requires the total or at least partial replacement of their solvation shells by the mutual supramolecular interactions. Supplementing wellknown approaches that rely on net electrostatic charges or extensive hydrogen bonding for guest complexation the present work focuses on the alternative concept of charge separation in the host to create a cationic site suitable for strong interaction with negatively charged guests. An anionic moiety balancing the overall host charge to zero is incorporated in a way to prevent collapsing into an ion pair. Using bicyclic guanidinium anchor groups for direct guest binding and a closo-borane cluster as an anionic countercharge the ditopic zwitterionic host 17 was designed and prepared in a convergent synthetic scheme. Despite its electroneutrality 17 binds inorganic and organic oxoanions such as sulfate, oxalate, squarate and p-nitrophenyl phosphate with affinity constants reaching 3.1 x 10⁴ M^-1 in DMSO (squarate^2-) or 1.1 x 10⁵ M^-1 in acetonitrile (p-nitrophenyl phosphate^2-). Clean 1:1 host-guest stoichiometry is found in dilute solutions, whereas at higher concentrations different complex compositions and host dimerization are observed. Titration calorimetry reveals the major role of entropy in host-guest association. In essence, anion binding by 17 and its congeners in strongly solvating solvents such as DMSO is favored by massive positive association entropies that in the case of sulfate complexation to 17 override an unfavorable positive binding enthalpy constituting an exclusively entropy-driven process.

Full text of DOI:10.1021/ja992028k

9986
J. Am. Chem. Soc. 1999, 121, 9986-9993
Zwitterionic Guanidinium Compounds Serve as Electroneutral Anion
Hosts
Michael Berger and Franz P. Schmidtchen*
Contribution from the Institute fu¨r Organische Chemie und Biochemie, Technische UniVersita¨t Mu¨nchen,
D-85747 Garching, Germany
ReceiVed June 15, 1999
Abstract: The complexation of anions in solvating phases by artificial receptors requires the total or at least
partial replacement of their solvation shells by the mutual supramolecular interactions. Supplementing well-
known approaches that rely on net electrostatic charges or extensive hydrogen bonding for guest complexation
the present work focuses on the alternative concept of charge separation in the host to create a cationic site
suitable for strong interaction with negatively charged guests. An anionic moiety balancing the overall host
charge to zero is incorporated in a way to prevent collapsing into an ion pair. Using bicyclic guanidinium
anchor groups for direct guest binding and a closo-borane cluster as an anionic countercharge the ditopic
zwitterionic host 17 was designed and prepared in a convergent synthetic scheme. Despite its electroneutrality
17 binds inorganic and organic oxoanions such as sulfate, oxalate, squarate and p-nitrophenyl phosphate with
affinity constants reaching 3.1 × 104 M-1 in DMSO (squarate2-) or 1.1 × 105 M-1 in acetonitrile (p-nitrophenyl
phosphate2-). Clean 1:1 host-guest stoichiometry is found in dilute solutions, whereas at higher concentrations
different complex compositions and host dimerization are observed. Titration calorimetry reveals the major
role of entropy in host-guest association. In essence, anion binding by 17 and its congeners in strongly solvating
solvents such as DMSO is favored by massive positive association entropies that in the case of sulfate
complexation to 17 override an unfavorable positive binding enthalpy constituting an exclusively entropy-
driven process.
Host-guest binding of anions in condensed phases is
in host-guest interactions may be hard to achieve. Hosts of
high charge also suffer from the obligatory presence of
counteranions which unspecifically bind to the host compound
and furthermore contribute to the ionic strength. For these
reasons the desired complexation of the targeted guests is
diminished. Multiply charged hosts are also notoriously poorly
soluble in nonaqueous solvents restricting their use to the
aqueous environment. Specific anion hosts, however, may find
their use also as extractants in anion separation or membrane
transport4 as well as in ion-selective electrodes (ISE).5 The
successful only if the host can provide sufficient favorable
interaction modes to outmatch the solvation free energies of
both binding partners. A straightforward option to meet this
requirement is to rely on strong and far-reaching attractive
Coulombic forces, that is, to employ highly charged cationic
host species. Following this guideline a vast variety of structures
have been examined1 which, in general, will associate a guest
species of opposite charge rather readily even under competitive
solvation conditions. This ion pairing process thus is unavoidable
and ultimately limits the guest selectivity. Intuitively, one may
expect that the difference in the interaction free energies ∆∆G
of two competing guests causing the discrimination between
them should rise the more negative the free energy of binding
∆Gass becomes. As a corollary, high selectivity should be seen
with molecular hosts that bind their anionic guests rather
strongly. Although this is observed experimentally, for example,
in the series of polyaza macrocycles forming multiply charged
cations on protonation, this approach offers no general solution
to the problem of binding specific anionic guests. The construc-
tion of the corresponding hosts calls for the prudent implemen-
tation of structural moieties addressing the differences between
competing guests in order to support their discrimination.
Frequently, these differences are rather subtle and require
delicate balancing of the barriers for discrimination built into
the host structure. In the presence of overwhelmingly dominant
electrostatic interactions the tuning of low-energy contributions
(2) (a) Dietrich, B.; Fyles, T. M.; Lehn, J. M.; Pease, L. G.; Fyles, D. L.
J. Chem. Soc., Chem. Commun. 1978, 934-936. (b) Dietrich, B.; Hosseini,
M. W.; Lehn, J. M.; Sessions, R. B. HelV. Chim. Acta 1983, 66, 1262-
1278. (c) Hosseini, M. W.; Lehn, J. M. HelV. Chim. Acta 1987, 70, 1312-
1319. (d) Dietrich, B.; Hosseini, M. W.; Lehn, J. M.; Sessions, R. B. J.
Am. Chem. Soc. 1981, 103, 1282.
(3) (a) Bianchi, A.; Micheloni, M.; Orioli, P.; Paoletti, P.; Mangani, S.
Inorg. Chim. Acta 1988, 146, 153-159. (b) Bianchi, A.; Micheloni, M.;
Paoletti, P. Coord. Chem. ReV. 1991, 110, 17-113. (c) Bianchi, A.; Garcia-
Espan˜a, E.; Mangani, S.; Micheloni, M.; Orioli, P.; Paoletti, P. J. Chem.
Soc., Chem. Commun. 1987, 729-731. (d) Bencini, A.; Bianchi, A.;
Burguete, M. I.; Domenech, A.; Garcia-Espan˜a, E.; Luis, S. V.; Nino, M.
A.; Ramirez, J. A. J. Chem. Soc., Perkin Trans. 2 1991, 1445-1451. (e)
Bencini, A.; Bianchi, A.; Burguete, M. I.; Dapporto, P.; Domenech, A.;
Garcia-Espan˜a, E.; Luis, S. V.; Paoli, P.; Ramirez, J. A. J. Chem. Soc.,
Perkin Trans. 2 1994, 569-577. (f) Bencini, A.; Bianchi, A.; Burguete,
M. I.; Garcia-Espan˜a, E.; Luis, S. V.; Ramirez, J. A. J. Am. Chem. Soc.
1992, 114, 1919-1920.
(4) (a) Visser, H. C.; Reinhoudt, D. N.; de Jong, F. Chem. Soc. ReV.
1994, 75-81. (b) Andreu, C.; Gala´n, A.; Kobiro, K.; de Mendoza, J.; Park,
T. K.; Rebek, J., Jr.; Salmer, N. A.; Usman, N. J. Am. Chem. Soc. 1994,
116, 5501-5502. (c) Stephan, H.; Gloe, K.; Schiessl, P.; Schmidtchen, F.
P. Supramol. Chem. 1995, 5, 273-280.
(5) (a) Bakker, E.; Bu¨hlmann, P.; Pretsch, E. Chem. ReV. 1997, 97, 3083-
3132. (b) Bu¨hlmann, P.; Pretsch, E.; Bakker, E. Chem. ReV. 1998, 98, 1593-
1687. (c) Antonisse, M. M. G.; Snellink-Runl, B. H. M.; Ygit, I.; Engbersen,
J. F. J.; Reinhoudt, D. N. J. Org. Chem. 1997, 62, 9034-9038.
(1) For recent reviews, see: (a) Schmidtchen, F. P.; Berger, M. Chem.
ReV. 1997, 97, 1609-1646. (b) Scheerder, J.; Engbersen, J. F. J.; Reinhoudt,
D. N. Recl. TraV. Chim. Pays-Bas 1996, 115, 307-320. (c) Supramolecular
Chemistry of Anions; Bianchi, A., Bowman-James, K., Garcia-Espan˜a, E.,
Eds.; Wiley-VCH: Weinheim, 1997.
10.1021/ja992028k CCC: $18.00 © 1999 American Chemical Society
Published on Web 10/16/1999

Guanidinium Compounds as Anion Hosts
J. Am. Chem. Soc., Vol. 121, No. 43, 1999 9987
Scheme 1
specific recognition event then must happen in an organic phase
and also has to avoid extremely strong guest binding for kinetic
reasons.
host before by attachment of carboxylate moieties to an anion-
binding macrotricyclic cage compound.11
These zwitterionic compounds 1 have good water solubility
and are fully functional anion hosts over a broad pH-regime
but require a laborious preparation and do not lend themselves
to nonaqueous applications. Here we report on an alternative
approach resting on the same basic concept of charge separation,
however, in a synthetically more accessible open-chain receptor
design.
All of these arguments may be taken to support an alternative
host design: Molecular recognition of anions ought to be
brought about by electroneutral host compounds relying on ion-
dipol forces as in the plethora of hydrogen bonding hosts6 or
specifically designed cage compounds.7 While this concept
succeeds in solvents of low polarity, it is generally not applicable
in the more polar organic solvents (DMSO, methanol, etc.)
which also solvate host and guest species by strong dipolar
interactions. The inspection of biological molecular hosts reveals
still another principle to selectively complex anions even under
the most competitive conditions. They manage to desolvate and
transfer negatively charged guests to peculiar sites in the interior
of their protein skeleton by virtue of unbalanced positive charges
and an array of preorganized hydrogen bond donor groups. In
many cases, exemplified for example by phosphate binding
protein (PBP),8 a guanidinium moiety in the side chain of an
arginine residue plays the role of a primary anchor function
which is supplemented by an ensemble of H-bond donors
furnishing a binding site of unique complementarity and
preorganization. As a result sulfate ion having the same charge
as phosphate at physiological pH is discriminated by a factor
of 105 in water.9 Despite their polyionic nature the net charge
of proteins is rather small so that most single charged groups
are counterbalanced by some function of opposite sign leading
overall to almost electroneutrality. The trick played by proteins
to make them successful anion hosts is the segregation of
charges enforced by a particular folding of the covalently
connected backbone. The prevention of collapse into ion pairing
of unlike charged functions opens the possibility of using the
strong Coulombic attraction of cationic guanidinium groups
toward oxoanions10 while minimizing the adverse effects of the
counterions. The same concept has been realized in an abiotic
Results and Discussion
Receptor Design. The strategic idea underlying host con-
struction has been reported in a previous paper12 and consists
of exploiting the inherent difference in binding affinity of the
bicyclic guanidinium moiety 213-15 to hydrogen-bond accepting
oxoanions on one hand versus the so-called noncoordinating
anions16,17 on the other. The latter do not contain lone electron
pairs and their charge is well delocalized to avoid any hydrogen-
bonding properties. This is in contrast to most oxoanions which
typically are quite basic (phosphate, carboxylate, sulfate, etc.)
and interact strongly with guanidinium groups such as 2. The
binding motif 3 forms the basis in the design of a number of
polytopic receptors for a variety of biologically important
anions,18-20 where A and B represent additional anchor groups.
Within this frame we embarked on the preparation of ditopic
(10) Hannon, C. L.; Anslyn, E. V. The Guanidinium Group: Its
Biological Role and Synthetic Analogues. In Bioorganic Chemistry
Frontiers; Dugas, H., Schmidtchen, F. P., Eds.; Springer: Heidelberg, 1993;
Vol. 3, pp 193-255.
(11) Worm, K.; Schmidtchen, F. P. Angew. Chem., Int. Ed. Engl. 1995,
34, 65-66.
(12) Berger, M.; Schmidtchen, F. P. J. Am. Chem. Soc. 1996, 118, 8947-
8948.
(13) (a) Schmidtchen, F. P. Chem. Ber. 1980, 113, 2175-2182. (b)
Schmidtchen, F. P. Tetrahedron Lett. 1990, 31, 2269-2272. (c) Kurzmeier,
H.; Schmidtchen, F. P. J. Org. Chem. 1990, 55, 3749-3755. (d) Gleich,
A.; Schmidtchen, F. P. Chem. Ber. 1990, 123, 907-915. (e) Schmidtchen,
F. P.; Oswald, H.; Schummer, A. Liebigs Ann. Chem. 1991, 539-543. (f)
Metzger, A.; Peschke, W.; Schmidtchen, F. P. Synthesis 1995, 566-570.
(14) (a) Echavarren, A.; Gala´n, A.; de Mendoza, J.; Salmero´n, A.; Lehn,
J. M. HelV. Chim. Acta 1988, 71, 685. (b) Segura, M.; Alca´zar, V.; Prados,
P.; de Mendoza, J. Tetrahedron 1997, 53, 13119-13128. (c) Segura, M.;
Alca´zar, V.; Prados, P.; de Mendoza, J. Tetrahedron Lett. 1998, 39, 1033-
1036.
(15) (a) Mu¨nster, I.; Rolle, U.; Madder, A.; DeClercq, P. J. Tetrahedron
Asymmetry 1995, 6, 2673-2674. (b) Davis, A. P.; Dempsey, K. J.
Tetrahedron Asymmetry 1995, 6, 2829-2840.
(16) Strauss, S. H. Chem. ReV. 1993, 93, 927-942.
(17) Reed, C. A. Acc. Chem. Res. 1998, 31, 133-139.
(18) For binding of nucleotides, see: (a) Gala´n, A.; Pueyo, E.; Salmero´n,
A.; de Mendoza, J. Tetrahedron Lett. 1991, 32, 1827-1830. (b) Deslong-
champs, G.; Gala´n, A.; de Mendoza, J.; Rebek, J., Jr. Angew. Chem., Int.
Ed. Engl. 1992, 31, 61-63. (c) Kato, Y.; Conn, M. M.; Rebek, J., Jr. J.
Am. Chem. Soc. 1994, 116, 3279-3284. (d) Gala´n, A.; de Mendoza, J.;
Toiron, C.; Bruix, M.; Deslongchamps, G.; Rebek, J., Jr. J. Am. Chem.
Soc. 1991, 113, 3, 9424-9425.
(6) (a) Raposo, C.; Perez, N.; Almarez, M.; Mussons, M. L.; Caballero,
M. C.; Mora´n, J. R. Tetrahedron Lett. 1995, 36, 3255-3258. (b) Ishida,
H.; Suga, M.; Donowaki, K.; Ohkubo, K. J. Org. Chem. 1995, 60, 5374-
5375. (c) Morzherin, Y.; Rudkevic, D. M.; Verboom, W.; Reinhoudt, D.
N. J. Org. Chem. 1993, 58, 7602-7605. (d) Casnati, A.; Fochi, M.; Minari,
P.; Pochini, A.; Reggiani, M.; Ungaro, R.; Reinhoudt, D. N. Gazz. Chim.
Ital. 1996, 126, 99-106. (e) Scheerder, J.; Engbersen, J. F.; Casnati, A.;
Ungaro, R.; Reinhoudt, D. N. J. Org. Chem. 1995, 60, 6448. (f) Gale, P.
A.; Sessler, J. L.; Kra´l, V. Chem. Commun. 1998, 1-8. (g) Kra´l, V.; Furuta,
H.; Shreder, K.; Lynch, V.; Sessler, J. L. J. Am. Chem. Soc. 1996, 118,
1595-1607.
(7) Worm, K.; Schmidtchen, F. P.; Schier, A.; Scha¨fer, A.; Hesse, M.
Angew. Chem., Int. Ed. Engl. 1994, 33, 327-329.
(8) (a) Luecke, H.; Quiocho, F. A. Nature, 1990, 347, 402-406. (b)
Shimoni, L.; Glusker, J. P. Protein Sci. 1995, 4, 65-74
(9) (a) Yao, N.; Ledvina, P. S.; Choudhary, A.; Quiocho, F. A.
Biochemistry 1996, 35, 2079-2085. (b) Kanyo, Z. F.; Christianson, D. W.
J. Biol. Chem. 1991, 266, 4264-4268.

9988 J. Am. Chem. Soc., Vol. 121, No. 43, 1999
Berger and Schmidtchen
receptors containing two bicyclic guanidinium moieties linked
by an aromatic spacer unit (Scheme 1). This design was directed
toward binding of doubly charged oxoanions mandating coop-
erative action of both anchor functions.
to simplify the analysis of host-guest-binding by NMR the
symmetrical arrangement of substituents at the spacer hub was
chosen.
Host Synthesis. Earlier host-guest-binding studies involving
the guanidinium building-block 2 had established that connecting
the anchor group to the spacer unit via sec-amide junctions is
profitable for anion binding.21 In addition to the aminomethyl
guanidinium module 525 the isophthalic acid derivative 10 was
required and obtained by lithium aluminum hydride reduction
of methyl trimesitoate and subsequent hydrolysis.26 After initial
attempts to couple both modules using Schotten-Baumann
conditions on the acid chlorides 8 or 9 had failed because of
undesired side reactions the attachment proceeded smoothly
employing HBTU/TEA at ambient temperature. This peptide
coupling reagent proved superior to DCC or some other standard
amidation reagents but required RP-chromatographic purification
of the product 10 to furnish a 72% yield. For connecting the
dianionic borocaptate cluster 4, the benzyl alcohol 10 was
converted to the mesylate 11 at low temperature as prolonged
reaction times or higher temperatures gave the corresponding
benzyl chloride 12. The subsequent substitution reaction with
the boron cluster 4 in the presence of strong base has some
analogy in the literature.23,27 Using a variety of conditions two
products were obtained in various ratios. While the reaction in
acetonitrile with triaza-bicyclodecene (TBD) as a base almost
exclusively furnished the undesired sulfonium salt 13 (cf. ref
23) switching the solvent to DMF gave preferentially the target
compound 17. The yield of 17 was decreased when 12 was used
as the alkylating agent of 4. In many reactions an intense blue
coloration appeared which, however, had no bearing on the yield
and may be attributed to a stable sulfur radical.28 For comparison
the thioether 14 was prepared by the same procedure as for 17
to give a ditopic host containing the same primary interaction
site for the guest anion as in 17 but possessing a 2-fold cationic
net charge. Irrespective of the charge difference borocaptate 4
and thiophenol have very similar sizes.22 Though the spectral
data of 14 and 17 are fully consistent with the assigned structures
The prime virtues of this design reside in the building-block
approach that enables a convergent strategy and allows ready
modification. The open-chain branched layout of the receptor
ensures adaptability of the host structure comforting the needs
of guest solvation shell displacement. Even though a price in
terms of the entropically demanding folding of the host must
be paid, experience tells that this path is preferable to attempting
total preorganization of the host. The incorporation of two
cationic guanidinium units required the attachment of a sub-
structure containing two negative charges to arive at an
electroneutral receptor. Among the various candidates the
dianionic icosahedral boron cluster compound 4 appeared
optimal with respect to its proven chemical stability and marked
lipophilicity22 and its well worked out chemistry of attachment
to organic structures23 which results from its use in the field of
boron neutron capture therapy BNCT.24 Being a strong nucleo-
phile the thiolate of 4 required an electrophilic point of
attachment to the spacer unit. Thus, our concept contained a
benzyclic halide moiety in the spacer in addition to two
carboxylic acid functions for anchor group attachment. To
maximize the mutual distance between the charged groups and
1
we notice a marked difference in the H NMR signal of the
methylene group bearing the thio substituent: Compound 14
displays a singlet at 4.15 ppm in accord with a freely rotating
CH2 group. In contrast 17 shows 2 mutually coupled doublets
indicating conformational locking and a well-defined folded
structure. Experiments discussed below provide arguments in
favor of intermolecular association to cause this observation.
(19) For binding of amino acids, see: (a) Gala´n, A.; Andreu, D.;
Echavarren, A. M.; Prados, P.; de Mendoza, J. J. Am. Chem. Soc. 1992,
114, 4, 1511-1512. (b) Metzger, A.; Gloe, K.; Stephan, H.; Schmidtchen,
F. P. J. Org. Chem. 1996, 61, 2051-2055.
(20) For binding of small, inorganic anions, see: (a) Hutchins, R. S.;
Molina, P.; Alajar´ın, M.; Vidal, A.; Bachas, L. G. Anal. Chem. 1994, 66,
3188-3192. (b) Schiessl, P.; Schmidtchen, F. P. J. Org. Chem. 1994, 59,
509-511.
(21) Mu¨ller, G.; Riede, J.; Schmidtchen, F. P. Angew. Chem. Int. Ed.
Engl. 1988, 27, 1516-1518.
(22) Plesek, J. Chem. ReV. 1992, 92, 269-278.
(23) Gabel, D.; Moller, D.; Harfst, S.; Ro¨sler, J.; Ketz, H. Inorg. Chem.
1993, 32, 2276-2278.
(24) Soloway, A. H.; Tjarks, W.; Barnum, B. A.; Rong, F.-G.; Barth, R.
F.; Codogni, I. M.; Wilson, J. G. Chem. ReV. 1998, 98, 1515-1562.
Molecular Modeling. The target structure 17 is zwitterionic
and though efforts were taken to minimize the specific interac-
tions of the substructures carrying unlike charges the Coulombic

Guanidinium Compounds as Anion Hosts
J. Am. Chem. Soc., Vol. 121, No. 43, 1999 9989
Figure 1. Energy-minimized structure of the dimer of 17 obtained by molecular modeling i.vac. using the MM+ force field. H-atoms and the
silyl-protecting groups were omitted for clarity.
attraction between them cannot be annihilated due to its
universal character. To learn about the preferred modes in this
interaction we undertook molecular modeling calculations in
vacuo without adjustment of the dielectric constant. First the
distribution of partial charges in 17 was obtained from semiem-
pirical AM1 calculations. In a second step molecular dynamics
used the MM+ force field on an ensemble of two separated
molecules followed by optimization of the ensemble geometry
by energy minimization at intervals. The low-energy structure
depicted in Figure 1 shows a dimer in which the substructures
of complementary charge of both monomers associate mutually
to form what we call a yin-yang dimer. The guanidinium units
approach the boron cluster anions side-on, indicating the total
absence of hydrogen bonds between them. Although this picture
may not reflect the true structure of 17 in solution, it is taken
as a good suggestion that the intermolecular association of two
molecules of 17 satisfying its unspecific electrostatic needs is
more favorable than the intramolecular folding to reach the same
goal.
can be cleanly modeled by a 1:1 dimerization process. The
Hostest 5.0 fitting program29 gives an apparent dimerization
constant of 250 M-1. The self-association is diminished on
addition of 50 mM tetrabutylammonium perchlorate (Kdim
)
100 M-1) as an inert electrolyte. Increasing the ionic strength
weakens electrostatic interactions in general, and this testifies
to the origin of the dimerization process.
The addition of bromide or nitrate salts to solutions of 17 in
DMSO did not cause any change in the NMR spectra. When
oxalate dianion, hydrogenphosphate, or adenosine monophos-
phate was added, dramatic shifts of all N-H signals were
observed initially in support of strong host-guest associations.
However, these interactions could not be quantified because
rapid deuterium exchange rendered these signals soon invisible,
and we were unable to resolve the shifts of other signals due to
severe overlap. Feeling that the high basicity of the guest anions
in DMSO were responsible we added up to 10 vol % of water
but without improving the analytical problem. Less basic anions
such as sulfate or the quasi-aromatic dianion squarate 18,
however, under the same conditions gave titration curves that
were readily analysable by curve fitting employing a mixed
model of 1:1 host-guest association and host dimerization.29
In this way Kass(SO42-, DMSO) ) 356 M-1 was obtained (see
Table 1) with a correlation coefficient of R2 ) 99.94. A small
but systematic deviation from the fit curve might indicate that
complexes of different stoichiometry might be present in
solution as well. If this assumption holds the 1:1 association
constant Kass would be diminished, but the data at hand do not
allow the meaningful and reliable interpretation in this sense.
The considerably larger squarate dianion 18 caused intense shifts
1
NMR and UV studies. The H NMR spectrum of 17 in
DMSO is concentration-dependent. This is particularly obvious
in the shift of the guanidinium NH signals which only
moderately move to lower field (∆δmax ) 0.35 ppm) on
increasing the concentration, indicating that these functions are
not primarily involved in supramolecular interactions. The shifts
(25) Peschke, W.; Schiessl, P.; Schmidtchen, F. P.; Bissinger, P.; Schier,
A. J. Org. Chem. 1995, 60, 1039-1043.
(26) Benning, A.; Grosskinsky, O. Chem. Ber. 1954, 54, 4-57.
(27) (a) Holmberg, A.; Meurling, L. Bioconjugate Chem., 1993, 4, 570-
573. (b) Nagasawa, K.; Narisada, M. Tetrahedron Lett. 1990, 31, 4029-
4032.
(28) Tolpin, E., I.; Wellum, G. R.; Berley, S. A. Inorg. Chem. 1978, 17,
2867-2873.
(29) Wilcox, C. S. HOSTEST 5.1; University of Pittsburgh, 1993.

9990 J. Am. Chem. Soc., Vol. 121, No. 43, 1999
Berger and Schmidtchen
Table 1. 1:1 Host-Guest Association of Zwitterionic Host 17 with
Anionic Guests in DMSO at 25 °C
Scheme 2
titration
method
anion
Kass (M-1
)
p-range31
NO3-, Br-
no effect
350
NMR
NMR
UV
UV
UV
2-
SO4
0.43 - 0.89
0.33 - 0.61
0.22 - 0.83
0.41 - 0.76
0.35 - 0.73
p-nitrophenyl phosphate2-
squarate2- 18
7 × 103
3.1 × 104
2.5 × 104
1.8 × 104
croconate2- 19
rhodizonate2- 20
UV
1
in the H NMR spectrum in DMSO, but the complex was too
stable to reliably derive an association constant. Instead, we
observed that this guest experienced an intensity enhancement
of its λ ) 313 nm UV-absorption band on complexation.
Employing Benesi-Hildebrand conditions ([17] g 10 × [18])
or curve-fitting procedures both revealed the formation of a
strong 1:1 complex giving Kass ) 3.1 × 104 M-1. In a similar
manner the homologous oxoanions croconate 19 and rhodizonate
20 were evaluated and showed binding constants somewhat
smaller but within a factor of 2 from the one found for squarate.
Applying UV titration to the binding of p-nitrophenyl phosphate
which particularly lends itself to this method owing to the long
wavelength n f π* transition at λ ) 436 nm30 required
solubilization of this guest in DMSO. Adding cryptand[2.2.2]
to a slurry of the sodium salt formed the stable alkaline cation
cryptand encapsulation complex and weakened the ion pairing
with the guest anion. Using this auxiliary, a 1:1 association
constant of 7 × 103 M-1 was deduced from the titration data.
All of these UV titrations were performed with 10-100 µM
concentrations of the host 17. Host dimerization in this regime
ready dissection of association free energies into its enthalpic
and entropic components. Since both are mutually compensating
in weak supramolecular interactions,33 the knowledge of the
individual enthalpy and entropy parts rather than their composite,
the free energy ∆G°, which is readily recalculated from Kass, is
required to gain insight into the intimacies of host-guest
binding. Because calorimetric measurements are truly integral
reflections of all processes occurring in solution, it is of utmost
importance to design very simple host-guest reactions to arrive
at interpretable results. A good example is given by the host-
guest binding of the prototypical oxoanion acetate to the bicyclic
guanidinium anchor group 6 (Scheme 2). In the concentration
range used for collecting the caloric data the salts on the left-
hand side of the equation in Scheme 2 as well as tetraethylam-
monium bromide are strong electrolytes and almost completely
dissociated in polar solvents like methanol, DMSO, or aceto-
nitrile. The only process to happen along with all changes in
the solvation of the species involved is the host-guest complex
formation of the guanidinium cation and acetate anion. The
corresponding ITC measurements and the titration curves
produced therefrom are shown in Figure 2. As is immediately
apparent from the sign of the heat pulses, host-guest binding
of these partners is exothermic in acetonitrile and DMSO
solvents, whereas the heat effects in methanol are minute and
do not allow us to decide whether there is no interaction at all
in this solvent or the reaction enthalpy adventitiously is near
zero. From the slope of the titration curves, itself being obtained
by plotting the integrals of the heat pulses, the standard free
energy ∆G° can be obtained. With ∆G° and ∆H° (the step
height of the titration curve) the entropy ∆S° can be calculated
from the Gibbs-Helmholtz equation (see Table 2). It comes as
a surprising result that in the former two cases the stability of
host-guest binding is not only due to the strong enthalpic
attraction but also to a favorable positive entropic component.
This association process combines two molecular species to form
one complex, yet the overall entropy of the system increases,
owing to the release of bound solvent molecules. Similar
entropic effects are widespread in aqueous systems34 but have
also been observed with supramolecular associations in organic
is negligible and need not be accounted for in the Kass
-
calculation.
On comparison to zwitterionic 17 the biscationic pendant host
14 shows very little change in its N-H NMR signals on dilution
(∆δ e 0.05 ppm, 0 f 12 mM). It thus appears safe to assume
that there is no aggregation in this concentration range which
is also plausible, viewing the net charge of this host. The NMR-
titration of 14 taking sulfate as a guest yields Kass ) 175 M-1
in DMSO. Contrary to the naive expectation, the host having
precisely the same primary guest binding site but a +2 net
charge exhibits a weaker affinity by 2-fold than its counterpart
with a formal charge of zero. We conclude there is apparently
no screening of the guest binding affinity from the anionic
portion in the zwitterionic host 17. The weaker binding power
of cationic 14 may be due to the ionic strength of this receptor,
which cuts down on the unspecific component of electrostatic
guest binding.
Isothermal Titration Calorimetry. Although NMR titrations
are indispensable tools in the wide-range collection of informa-
tion on supramolecular associations and can provide clues on
the structural mode of host-guest relationships, the elucidation
of the thermodynamic parameters ∆G°, ∆H°, and ∆S° by this
instrumental method is laborious, insensitive, and error-prone.
A more direct access to those important data is offered by
modern isothermal titration calorimetry (ITC),32 allowing the
(30) Ishida, H.; Suga, M.; Donowaki, K.; Ohkubo, K. J. Org. Chem.
1995, 60, 5374-5375.
(31) Wilcox, C. S. In Frontiers in Supramolecular Organic Chemistry
and Photochemistry; Schneider, H. J., Du¨rr, H., Eds.; VCH: Weihnheim,
1991; pp 123-143.
(32) (a) Ladbury, J. E.; Chowdhry, B. Z. Chem. Biol. 1996, 3, 791-
801. (b) Wadso¨, I. Chem. Soc. ReV. 1997, 79-86. (c) Janin, J. Structure
1997, 5, 473-479.
(34) (a) Gelb, R. I.; Lee, B. T.; Zompa, L. J. J. Am. Chem. Soc. 1985,
107, 909-916. (b) Labadi, I.; Jenei, E.; Lahti, R.; Lo¨nnberg, H. Acta Chem.
Scan. 1991, 45, 1055-1059.
(33) Dunitz, J. D. Chem. Biol. 1995, 2, 709-712. (b) Gallicchio, E.;
Kubo, M. M.; Levy, R. M. J. Am. Chem. Soc. 1998, 120, 4526-4527.

Guanidinium Compounds as Anion Hosts
J. Am. Chem. Soc., Vol. 121, No. 43, 1999 9991
Figure 2. ITC analysis of the guanidinium-acetate host-guest binding at 30 °C according to Scheme 2. (a) In acetonitrile, [6] ) 1.92 mM; (b) in
DMSO, [6] ) 2.5 mM; the step in the plot of heat pulses is due to a change in the titrand volume added and is accounted for by the evaluation
software; (c) in methanol.
Table 2. Thermodynamic Parameters Evaluated by ITC of
Tetraethylammonium Acetate Binding to 6 at 30 °C
agreement with the value obtained by NMR-titration. The close
inspection, however, reveals an exclusively entropy-driven
endothermic process having ∆H° ) +10.5 kJ mol-1 and ∆S°
) +78 J K-1 mol-1. Most surprisingly the nonlinear regression
gives n ) 0.5 as an independent fit parameter meaning that
there are two binding sites of equal affinity on each host 17.
Obviously, there is no cooperative binding of both guanidinium
anchor groups to sulfate when the anionic boron cluster resides
in the molecule translating into a 10 000-fold reduction in guest
affinity.
CH3CN
DMSO
K
ass (M-1
)
2.0 × 105
6.5 × 103
∆G° (kJ mol-1
∆H° (kJ mol-1
)
)
-30.7
-22.1
-15.5
+50.2
-14.2
+26.1
∆S° (J mol-1 K-1
)
Table 3. Host-Guest Complexation of {Na+cryptand[2.2.2]}2
p-Nitrophenyl Phosphate by Compounds 17, 15, and 16 in
Acetonitrile at 30 °C
Taking squarate as a guest anion the calorimetric technique
unravelled a more complicated host-guest binding process with
17 than anticipated on the basis of the UV titration. Unlike the
clear 1:1 stoichiometric complexation observed in the micro-
molar concentration range (see above) the ITC-measurements
at 100-fold higher concentration showed a stepwise binding
event first forming a complex of 2:1 host: guest stoichiometry
(inflection an n ) 0.5) which on further addition of the guest
solution gives rise to a step having 1:1 (or maybe 2:2)
stoichiometry. With the software at hand the second step was
evaluated giving Kass ) 8.9 × 105 M-1 in an exothermic and
entropically driven process. The observation of stepwise binding
indicates that squarate 18 does not bind with full complemen-
tarity to host 17 and satisfies all of its coordination needs in
DMSO. It is also not possible to decide from these data on the
sequence of binding events, that is, whether squarate binds to
the host dimer or rather a 1:1 host-guest complex is formed
that associates another host molecule in a subsequent step.
Viewing the simple structures we tend to speculate that there
is a random binding order. The boron cluster, however, is not
innocent in this binding since the analogous host compound 14
on binding squarate 18 shows very similar thermodynamic
parameters (Table 4) but lacks the stepwise binding feature.
Kass
∆G°
∆H°
∆S°
(M-1
)
(kJ mol-1
)
(kJ mol-1
)
(J K-1 mol-1
)
17
15
16
1.1 × 105
1.6 × 105
1.1 × 105
-29.2
-30.2
-29.2
-33.5
-37.7
-28.4
-14.1
-24.7
+2.4
solvents.35 The difference in stability of the guanidinium acetate
complex (Scheme 2) in CH3CN or DMSO is primarily attribut-
able to the less positive ∆S° value in the latter case, probably
reflecting a smaller number of the bigger DMSO molecules set
free on host-guest binding.
Caloric data may also provide structural information if a
collection of hosts of similar structure is compared. For instance,
the binding of p-nitrophenyl phosphate in acetonitrile to the
zwitterionic host 17 and two analogues 15 and 16 having exactly
the same primary anion binding site but differing at a remote
location having none or a neutral substituent at this position all
produce the same binding constant within experimental error
(Table 3). However, the immediately obvious conclusion
denying any influence of the remote substituent on the guest
binding process is false, because the unfolding of the binding
enthalpies and entropies clearly demonstrate its participation.
ITC measurements also offer an explanation for the surpris-
ingly weak binding of sulfate dianion to the zwitterionic host
17 in DMSO. While analogous dicationic guanidinium hosts
bind this guest very tightly (Kass > 106 M-1 36) the ITC analysis
gives Kass (SO42-, DMSO) ) 200 M-1 which is in reasonable
Conclusions
The contradicting goals of high affinity for anionic guests
under competitive solvation conditions, yet good lipophilicity
were addressed in a new receptor design. A zwitterionic host
17 was constructed in a building-block approach and proved
active in oxoanion binding in DMSO. The binding studies in
comparison to those of charged hosts of very similar structure
showed a minor influence of the net charge only. Instead the
(35) Cram, D. J.; Choi, H. J.; Bryant, J. A.; Knobler, C. B. J. Am. Chem.
Soc. 1992, 114, 7748-7765. (b) Linton, B.; Hamilton, A. D. Tetrahedron
1999, 55, 6027-6038.
(36) Berger, M.; Schmidtchen, F. P. Angew. Chem., Int. Ed. 1998, 37,
2694-2696.

9992 J. Am. Chem. Soc., Vol. 121, No. 43, 1999
Berger and Schmidtchen
Table 4. ITC Evaluation of Host-Guest Binding in DMSO at 30
°Ca
fractions containing 10 were pooled, concentrated under reduced
pressure, and extracted with CH2Cl2 (3×). The combined extracts were
washed with brine (2×) and dried over Na2SO4. After removal of the
solvent in vacuo, 10 was obtained as the chloride salt (366 mg, 72%).
1H NMR (CDCl3, 360 MHz) δ 8.46 (s), 8.38 (s), 8.07 (s), 7.93 (s),
7.73 (s), 7.70 (s), 7.56-7.59 (m), 7.31-7.40 (m), 4.46 (s), 3.84 (s),
3.21-3.84 (m), 2.01 (m), 1.87-1.89 (m), 1.03 (s). 13C NMR (CDCl3,
90 MHz), δ 167.3 (CO), 151,1 (C+), 142 (ArCH), 135.5 (silyl), 134.8
(ArC), 132.9 (ArCH), 132.6 (silyl), 129.8 (silyl), 127.6 (silyl), 124.0
(ArCH), 65.6 (CH2O), 64.0 (CH2O), 49.7 (CHN), 48.0 (CHN), 45.2
(CH2N), 45.0 (CH2N), 43.8 (CH2N), 26.7 (CH3), 23.7 (CH2), 22.6 (CH2),
19.1 (C). MS (FAB) m/e 1070 (M‚Cl+, 40%), 1034 (100%), 186 (95%).
Guanidinium-mesylate 11. The benzyl alcohol 10 (291 mg, 0.263
mmol) was dissolved in 5 mL of dry CH2Cl2, 110 µl of triethylamine
(TEA, 3 equiv) were added, and the resulting mixture was cooled to
-10 °C. Then 0.5 mL of a freshly prepared 1 M solution of
methanesulfonyl chloride in CH2Cl2 was added dropwise via syringe,
and stirring was continued at -10 °C for 30 min after which HPLC
analysis indicated complete conversion to a single product. The reaction
mixture was washed with 0.1 M HCl and 1 M NaClO4 and dried over
Na2SO4. After filtration and removal of the solvent under reduced
pressure, 11 (perchlorate salt) was obtained as a colorless, viscous
residue and used directly for further conversion. 1H NMR (CDCl3, 360
MHz) δ 7.98 (s), 7.92 (s), 7.75 (d), 7.70 (s), 7.56-7.59 (m), 7.31-
7.42 (m), 7.17 (s), 6.82 (s), 5.27 (dd), 3.84 (s), 3.21-3.67 (m), 1.88-
2.04 (m), 1.02 (s). 13C NMR (CDCl3, 90 MHz) δ 160.3 (CO), 150.9
(C+), 135.5 (silyl), 134.9 (ArC), 133.4 (ArC), 132.6 (silyl), 130.5
(ArCH), 129.8 (silyl), 127.6 (silyl), 125.5 (ArCH), 70.7 (CH2O), 65.7
(CH2O), 50.0 (CHN), 47.8 (CHN), 45.2 (CH2N), 44.8 (CH2N), 43.9
(CH2N), 37.6 (CH3, mesyl), 26.6 (CH3), 23.7 (CH2), 22.6 (CH2), 19.0
(C).
14
17
squarate 18 Kass (M-1
)
6.2 × 105
8.9 × 105
∆G° (kJ mol-1
∆H° (kJ mol-1
)
)
-33.6
-34.5
-7.4
-5.4
∆S° (J K-1 mol-1
n
)
+86.6
+96.0
0.85
1.10
sulfate
K
ass (M-1
)
not determined
200
∆G° (kJ mol-1
∆H° (kJ mol-1
)
)
-13.3
+10.5
+78
∆S° (J K-1 mol-1
n
)
0.5
a The data given for the 18 to 17 interaction refer to the step giving
the 1:1 complex.
role of the solvent in host-guest binding was apparent from
calorimetric measurements and materializes in positive reaction
entropies of respectable size. One is led to conclude that
solvation design is a profitable way to supplement host-guest
complementarity in the construction of selective artificial hosts.
Experimental Section
General methods. Proton-, 11B-, and 13C NMR spectra were recorded
on a Bruker AM 360 (360 MHz) instrument and are calibrated to
tetramethylsilane as internal standard (11B to BF3OEt2 as external
standard). Mass spectra were obtained on Varian MAT CH5 (FAB) or
Finnigan LQC (ESI) instruments at the TU-Mu¨nchen. MALDI-TOF
spectra were measured on a Micromass TofSpec at the University of
Bonn. IR spectra were measured on a Perkin-Elmer FTIR 1600
instrument. HPLC-analyses were performed on Merck-Hitachi instru-
ments (L6200A or L7100 pump connected to L4250 or L7400 UV-
detectors or to a Eurosep DDL-31 light-scattering detector). As columns,
Prontosil C8 SH (125 × 4 mm) or Nucleosil C18 (250 × 4 mm) were
used with mixtures of acetonitrile/water containing 30 mM of H3PO4
and 30 mM NaClO4 or 0.1% trifluoroacetic acid (for detection with
light scattering detector). Column chromatography (HPLC) was per-
formed in Michel-Miller columns packed with RP-modified silica
(Nucleoprep 100-30 C8) connected to a Knauer 364 HPLC pump, a
Phillips PU 8620 photometer and a LKB fraction collector.
Solvents were distilled before use except for DMF which was
purchased in anhydrous quality from Aldrich and acetonitrile which
was purchased in HPLC-quality from Zefa). All other chemicals were
purchased in reagent quality and used as received. Aqueous solutions
were prepared from deionized, glass-distilled water. All reactions were
carried out in an atmosphere of nitrogen and solvents (CH2Cl2, CH3-
CN) were passed through a small column of activated alumina directly
into the reaction vessel. Sodium borocaptate 4 (10B-enriched) was
purchased from Boron Biological Inc. Aminomethylguanidinium
chloride 513f and 5-hydroxymethylisophthalic acid26 were prepared as
previously described.
Synthesis. Guanidinium-benzyl Alcohol 10. Aminomethyl guani-
dinium chloride 5 (500 mg, 0.982 mmol, HCl salt) was dissolved in 3
mL of DMF and added to a suspension of 5-hydroxymethylisophthalic
acid 7 (91.7 mg, 0.468 mmol) in 2 mL of acetonitrile. The minimum
amount of DMF was then added with stirring to furnish a clear solution
followed by N-methylmorpholine (0.5 mL, 5 mmol). To the resulting
suspension a solution of O-(benzotriazol-1-yl)-tetramethyluronium
hexafluorophosphate (HBTU, 426 mg) in a mixture of acetonitrile/
DMF (1:1 vol) was added dropwise with stirring. After 1 h, another
20 mg of solid 5 and 30 mg of solid HBTU were added, followed by
another 30 mg of 5 (solid) and 40 mg of HBTU (solid). After the final
addition, the HPLC analysis showed complete conversion to one main
product. The mixture was diluted with 30 mL of CH2Cl2, washed with
0.1 M HCl, 1 M Na2CO3 (2×), 0.1 M HCl and water (20 mL portions
each) and dried over Na2SO4. The residue obtained on filtration and
removal of the solvent under reduced pressure was purified by flash
chromatography (Nucleoprep 100-30 µm, C8, 35 × 4 cm, CH3OH/
H2O (85/15 vol) containing 30 mM H3PO4 and 30 mM NaClO4). The
Borocaptate-receptor 17. The residue of the mesylate 11 obtained
above was dissolved in dry DMF, and a stream of nitrogen was passed
through the solution for 10 min. Likewise, 169 mg of sodium
borocaptate 4 (0.8 mmol) was dissolved in 2 mL of DMF and was
degassed. Both solutions were combined to give a pale green solution
to which a solution of 1,5,7-triazabicyclo[4.4.0]dec-5-en (TBD) (21
mg, 0.14 mmol in 0.25 mL of DMF) was added slowly via syringe
pump over 90 min. HPLC analysis showed the disappearance of the
starting material, and the reaction mixture was diluted with 20 mL of
CH2Cl2 and washed with 0.1 M HCl (2×) and brine. After drying over
Na2SO4 and filtration and removal of the solvent under reduced pressure,
the residue was purified by column chromatography (Nucleoprep 100,
30 µm, C8, 30 × 2.1 cm, CH3CN/H2O (86/14 vol) containing 30 mM
H3PO4 and 30 mM NaClO4). The fractions containing 17 were pooled,
carefully neutralized with 0.2 M NaOH, concentrated under reduced
pressure and extracted with CH2Cl2 (3×). The combined extracts were
dried over Na2SO4. After filtration and removal of the solvent under
reduced pressure, 17 was obtained as a slightly yellow solid (161 mg,
1
52% from 10). H NMR (DMSO-d6, 360 MHz) δ 8.81 (s, NH), 8.23
(s, 1H, ArH), 7.86 (s, 2H, ArH), 7.81 (s, NH), 7.49-7.62 (m, 8H, ArH),
7.40-7.45 (m, 12H, ArH), 4.41 (dd, 2H, ArCH2), 3.20-3.63 (m, 20H);
1.70-2.05 (m, 8H, CH2), 1.00 (s, 18H, CH3). 13C NMR (CD3CN/CDCl3,
90 MHz) δ 165.4 (7, CO), 150.2 (C+), 134.5 (silyl), 132.7 (ArCH),
132.1 (ArC), 131.8 (silyl), 130.7 (ArCH), 129.1 (silyl), 128.1 (silyl),
124.5 (ArCH), 65.2 (CH2O), 49.3 (CHN), 47.8 (CHN), 45.2 (CH2S),
44.4 (CH2N), 44.1 (CH2N), 43.0 (CH2N), 25.5 (CH3), 23.0 (CH2), 21.5
(CH2), 18.9 (C). 11B NMR (CD3CN, 115,5 MHz) δ -12 (-16) (m).
IR (KBr, cm-1) 3431 (s), 2506 (w), 1624 (s), 1540 (w). MS (MALDI-
TOF, negative-mode, 2,5-dihydroxybenzoic acid as matrix) m/e 1180
(M-).
Guanidinium-benzyl Chloride 12. 5-tButyldimethylsilyloxymeth-
ylisophthalic acid (0.6 g, 2 mmol) were suspended in 20 mL of toluene,
1 mL of oxalic acid chloride was added, and the mixture was stirred
overnight. After the solvent was stripped off, CH2Cl2 (100 mL) was
added to the residue, and the resulting slurry was filtered. The filtrate
was concentrated to dryness to yield 530 mg of the acid chloride 9
(1.52 mmol). The residue of 9 was dissolved in 10 mL of CH2Cl2, the
amino compound 5 (1.58 g, 3.3 mmol, HCl salt) was added, and the
solution was cooled to 0 °C. TEA (1.7 mL, 12 mmol) was added
dropwise over a period of 60 min with stirring. The reaction mixture

Guanidinium Compounds as Anion Hosts
J. Am. Chem. Soc., Vol. 121, No. 43, 1999 9993
was diluted with 40 mL of CH2Cl2, washed with 1 M HCl (3×), and
dried over Na2SO4. The residue obtained after filtration and removal
of the solvent under reduced pressure was purified by column
chromatography (Nucleoprep 100, 30 µm, C8, CH3OH/H2O, step
gradient first 75/25, then 80/20 vol, containing 30 mM H3PO4 and 30
mM NaClO4). Employing the same workup as described for 11, 360
mg of 12 (0.32 mmol, 21% from 8), 250 mg of 10, and 250 mg of a
ylguanidine 5 (chloride salt). The solution was cooled to 0 °C, and
TEA (840 µl, 6 mmol) was added dropwise with stirring. Stirring was
continued for 30 min followed by washings with 0.1 M HCl (3×) and
brine. After drying over Na2SO4 and filtration, the solution was
concentrated under reduced pressure, and the product was precipitated
by addition of diethyl ether. After separation and drying in vacuo 16
1
was obtained as a white solid (chloride salt, 723 mg, 91%). H NMR
1
mixture of 10 and 12 were obtained. H NMR (360 MHz, CDCl3) δ
(360 MHz, CDCl3) δ 8.85 (NH), 8.59 (s, 1H, ArCH), 8.53 (s, 1H, NH),
8.47 (s, 1H, NH), 7.72 (s, 2H, ArCH), 7.59-7.63 (m, 8H, silyl), 7.3-
7.45 (m, 12H, silyl), 5.11 (s, 5H, Bn), 3.1-3.9 (m, 20H, CHN, CH2N),
1.95-2.10 (m, 4H, CH2), 1.75-1.95 (m, 4H, CH2), 1.00 (s, 18H, CH3).
13C NMR (90 MHz; CDCl3) δ 167.0 (CO), 151.0 (C+), 136.6 (Bn),
135.5 (silyl), 135.1 (ArC), 132.7 (silyl), 129.9 (silyl), 128.4 (Bn), 127.7
(silyl), 127.5 (Bn), 119.0 (ArCH), 117.6 (ArCH), 70.1 (Bn), 65.2
(CH2O), 49.1 (CHN), 47.6 (CHN), 45.0 (CH2N), 44.5 (CH2N), 43.5
(CH2N), 26.8 (CH3), 23.6 (CH2), 22.5 (CH2), 19.1 (C). MS (ESI) m/e
1224 (100%, [M ‚ TFA]+), 1145 (90%, [M ‚ Cl]+), 555.7 (85%, M2+).
Phenol 15. The benzyl ether 16 (646 mg, 0.546 mmol) was dissolved
in a mixture of 5 mL of ethanol and 2.5 mL of cyclohexene, 100 mg
Pd(OH)2/C (20%) was added, and the mixture was heated under reflux
for 4 h, after which another 100 mg of the catalyst were added. Heating
was continued for another 4 h, and then the suspension was filtered
through Celite. The residue was rinsed with CH3OH, and the filtrates
were combined and the solvents stripped off. Reprecipitation of the
residue from CH2Cl2 (1 mL) + diethyl ether was repeated twice. Finally,
the solid was separated and dried in vacuo, yielding 15 as a white solid
8.86 (s, 1H, ArCH), 8.68 (m, 2H, NH), 8.29 (s, 2H, NH), 8.09 (s, 2H,
ArCH), 8.02 (s, 2H, NH), 7.60-7.63 (m, 8H, silyl), 7.31-7.41 (m,
12H, silyl), 4.60 (s, 2H, CH2Cl), 3.90 (m, 2H, CH2O), 3.50-3.77 (m,
8H, CHN, CH2N), 3.11-3.47 (m, 10H, CHN, CH2N), 1.95-2.10 (m,
4H, CH2), 1.75-1.90 (m, 4H, CH2), 1.03 (s, 18H, CH3). 13C NMR (90
MHz, CDCl3) δ 167.0 (CO), 151.1 (C+), 138.2 (ArC), 135.5 (silyl),
134.3 (ArC), 132.7 (silyl), 131.4 (ArCH), 129.9 (silyl), 127.8 (silyl),
126.1 (ArCH), 65.4 (CH2O), 49.3 (CHN), 47.7 (CHN), 45.4 (CH2Cl),
45.0 (CH2N), 44.7 (CH2N), 43.5 (CH2N), 26.9 (CH3), 23.6 (CH2), 22.5
(CH2), 19.1 (C). MS (ESI) m/e 1165.2 ([M ‚ TFA]+, 100%), 1087.2
([M ‚ Cl]+, 7%), 526.8 (M2+, 22%).
Thioether 14. The mixture of 10 and 12 obtained above (250 mg,
combined ∼0.18 mmol) was dissolved in 10 mL of CH2Cl2 and
subjected to standard mesylation: 0.4 mL of a freshly prepared 1 M
solution of TEA in CH2Cl2 was added, and the mixture was cooled to
0 °C. Then 0.4 mL of a freshly prepared 1 M solution of methane-
sulfonyl chloride in CH2Cl2 was added dropwise with stirring followed
by 0.4 mL of the TEA solution. After stirring at 0 °C for 60 min the
reaction mixture was washed with 0.1 M HCl (3×), 1 M NaClO4 and
dried over Na2SO4. The residue obtained after filtration and removal
of the solvent was divided into two aliquots, one-half was dissolved in
3 mL of CH2Cl2, 25 mg TBD was added, and the resulting mixture
was cooled to 0 °C. Thiophenol (18 µl, 0.18 mmol) was added dropwise,
and stirring was continued for 60 min. The reaction mixture was washed
with 0.1 M HCl (2×) and 1 M Na2CO3 (2×) and dried over Na2SO4.
After filtration and concentration of the solution to a volume of ∼0.5
mL, the crude product was precipitated by addition of diethyl ether.
The product was collected by centrifugation, separated from the
supernatant, and dissolved again in 0.5 mL of CH2Cl2. Precipitation
and redissolving was repeated 3 times. The crude product obtained in
this manner was distributed in a two-phase system (prepared from
HCCl3/heptanes/CH3OH/0.1 M NaCl in a volumetric ratio of 2/2/2/1),
the lower phase was separated and washed 3 times with fresh upper
phase. After drying of the lower phase over Na2SO4, filtration and
removal of the solvent under reduced pressure, gave 14 as a slightly
1
(chloride salt, 565 mg, 95%). H NMR (360 MHz, CDCl3) δ 9.43 (s,
NH), 8.66 (s, 1H, ArCH), 8.28 (s, NH), 8.07 (s, 2H, ArCH). 7.59-
7.63 (m, 8H, silyl), 7.3-7.45 (m, 12H, silyl), 3.1-3.9 (m, 24H, CH2O,
CHN, CH2N), 1.95-2.10 (m, 4H, CH2), 1.75-1.95 (m, 4H, CH2), 1.00
(s, 18H, CH3). 13C NMR (90 MHz; CDCl3) δ 167.5 (CO), 157.1 (ArC),
151.0 (C+), 135.3 (silyl), 134.6 (ArC), 132.5 (silyl), 129.6 (silyl), 127.6
(silyl), 117.8 (ArCH), 117.3 (ArCH), 65.3 (CH2O), 49.2 (CHN), 47.9
(CHN), 45.0 (CH2N), 44.5 (CH2N), 43.0 (CH2N), 26.6 (CH3), 23.4
(CH2), 22.4 (CH2), 18.9 (C). MS (ESI) m/e 1133.2 (100%, [M ‚ TFA]+),
1055.4 (55%, [M ‚ Cl]+), 510.7 (60%, M2+).
Binding Studies. ITC titrations were performed using a MCS-ITC
instrument (MicroCal Inc., Northampton, MA). All measurements were
performed at 303 K. Stock solutions were prepared by weighing the
substances directly in volumetric flasks. In general, the host solution
was filled into the cell of the ITC instrument and guest solutions were
added with the syringe. In each case control experiments with dilution
of guest solution in neat solvent were performed. The dilution of the
host was found to be negligible. Analysis and curve fitting was done
using the software Origin 2.9.
1
yellow solid (42 mg, 41%). H NMR (CDCl3; 360 MHz) δ 8.71 (s,
1H, ArCH), 8.47 (s, 2H, NH), 8.13 (s, 2H, NH), 8.03 (s, 2H, ArCH),
7.84 (s, 2H, NH), 7.49-7.62 (m, 8H, silyl), 7.40-7.45 (m, 12H, silyl),
7.14-7.29 (m, 5H, Bn), 4.15 (s, 2H, CH2S), 3.86 (m, 2H, CH2O), 3.17-
3.73 (m, 18H, CHN, CH2N), 2.03 (m; 4H, CH2), 1.85 (m, 4H, CH2),
1.03 (s; 18H, CH3). 13C NMR (CDCl3; 90 MHz) δ 167.2 (CO), 151.2
(C+), 138.6 (Bn), 135.8 (ArC); 135.5 (silyl), 133.9 (ArC), 132.7 (silyl),
131.7 (ArCH), 130.0 (silyl, Bn), 128.9 (Bn), 127.8 (silyl), 126.5 (Bn),
124.8 (ArCH), 65.4 (CH2O), 49.3 (CHN), 47.9 (CHN), 45.2 (CH2N),
44.8 (CH2N), 43.4 (CH2N), 38.7 (CH2S), 26.9 (CH3), 23.7 (CH2), 22.5
(CH2), 19.1 (C). MS (ESI) m/e 1239.2 (40%, [M ‚ TFA]+), 1125.7
(100%, M+), 563.8 (40%, M2+).
NMR titrations were performed on a Bruker AM 360 instrument at
298 K in DMSO-d6. The concentrations of guest and host were varied,
1
and the shift of the NH-protons was followed in the H NMR spectra
and analyzed with Wilcox’s HOSTEST program.29
UV titrations were performed using a Eppendorf 1101M photometer
equipped with a thermostated cuvette holder in 1 cm cuvettes at 298 K
in DMSO. The change in adsorption at λ ) 436 nm (for p-nitrophenyl
phosphate) or 313 nm (for the cyclic oxoanions) was followed and
analyzed by standard Benesi-Hildebrand plots39 (for [host] > 10 ×
[guest]) or curve fitting using the HOSTEST routines. Solutions of
p-nitrophenyl phosphate, croconate 19 and rhodizonate 20 in DMSO
were prepared by stirring the bis-alkaline salts with 2 equiv of cryptand-
[2.2.2] in volumetric flasks until a clear solution was obtained.
Benzyl ether 16. 5-Benzyloxyisophthalic acid37 (184 mg, 0.68 mmol)
was suspended in 3 mL of thionyl chloride, triphenylphosphine (2 mg)
was added, and the mixture was heated under reflux for 60 min.38 Excess
thionyl chloride was removed in a stream of nitrogen, and the residue
was dissolved in 5 mL of CH2Cl2 containing 707 mg aminometh-
Acknowledgment. We gratefully acknowledge the financial
support by Deutsche Forschungsgemeinschaft (project Schm
369/12-1/2), of the German ministry of Science and Technology
BMBF (project 13N7130) and Fonds der Chemischen Industrie.
(37) Diederich, F.; Schurmann, G.; Chao, I. J. Org. Chem. 1988, 53,
2744-2757.
(38) Rudolph, U.; Schmitt, M.; Freitag, D.; Bottenbruch, L. Eur. Pat.
Appl. EP 50,783, 1982; Chem. Abstr. 1982, 97, 92968.
(39) Connors, K. A. Binding Constants; Wiley: New York, 1987.
JA992028K