Synthesis and assembly of self-complementary cavitands

Article abstract of DOI:10.1021/ja993922e

Cavitands with self-complementary shapes (3 and 4) were prepared by the covalent attachment of adamantane guest molecules to the upper rim of the host structures. Relatives of the 'self-folding' cavitands 2, these new structures possess a seam of intramolecular hydrogen bonds that stabilize the folded conformation. Their self-complementary shapes result in the formation of noncovalent dimers of considerable kinetic and thermodynamic stability (- ΔAG²⁹⁵ = 4.5 kcal/mol for 3a and 6.5 kcal/mol for 4a in p-xylene-d₁₀). The dimerization of cavitands 3 and 4 is reversible and subject to control by solvent and temperature. The dimerization process is enthalpically favored and entropy opposed and occurs with significant enthalpy-entropy compensation.

Full text of DOI:10.1021/ja993922e

J. Am. Chem. Soc. 2000, 122, 4573-4582
4573
Synthesis and Assembly of Self-Complementary Cavitands
Adam R. Renslo, Fabio C. Tucci, Dmitry M. Rudkevich, and Julius Rebek, Jr.*
Contribution from The Skaggs Institute for Chemical Biology and the Department of Chemistry,
The Scripps Research Institute, 10550 North Torrey Pines Rd., La Jolla, California 92037
ReceiVed NoVember 5, 1999
Abstract: Cavitands with self-complementary shapes (3 and 4) were prepared by the covalent attachment of
adamantane guest molecules to the upper rim of the host structures. Relatives of the “self-folding” cavitands
2, these new structures possess a seam of intramolecular hydrogen bonds that stabilize the folded conformation.
Their self-complementary shapes result in the formation of noncovalent dimers of considerable kinetic and
thermodynamic stability (-∆G295 ) 4.5 kcal/mol for 3a and 6.5 kcal/mol for 4a in p-xylene-d10). The
dimerization of cavitands 3 and 4 is reversible and subject to control by solvent and temperature. The dimerization
process is enthalpically favored and entropy opposed and occurs with significant enthalpy-entropy
compensation.
Introduction
directly as hosts by Aoyama9 and modified by Cram,10 to
become modules for the construction of carcerands11 and
velcrands.12 The latter work established the conformational
preferences and interconversion dynamics of the resorcinarene-
derived cavitands as subject to control by temperature and
solvent. Subsequently, we installed a hydrogen-bonded seam
along the upper rim of the structure, to hold the conformation
in the vase-like shape shown as in 2. This gave rise to “self-
folding” cavitands.13
Appropriately convex guest molecules for these cavitand hosts
were found in adamantane derivatives. The complementarity
of size is obvious from modeling but the complementarity of
shape is not. The cavitand has a 4-fold axis of symmetry while
the 1-substituted adamantanes have a 3-fold axis. But most
space-filling renderings of adamantane show it to be nearly
spherical whereas the appropriate cross section of the host is
nearly circular. Somewhat less obvious is the complementarity
of chemical surfaces. Almost all the atoms that line the cavity
are sp2 hybridized with their p orbitals directed inward, so the
concave surface may be thought to present a thin layer of
negative charge. For sp3 hybridized C-H bonds this environ-
ment offers little intrinsic appeal although a multitude of
attractions can be hidden within C-H/π interactions, van der
Waals forces, and solvophobicity. Cylcohexanes were also
welcome guests, even though their shapes were less comple-
mentary to the host. Cubane would seem ideal since its C-H
bonds of higher s character present a thin layer of positive charge
Small molecule guests can be surrounded by larger host
structures either reversibly or irreversibly. Within the host-
guest complex molecular recognition is expressed and it has
consequences for both host and guest: molecular surfaces in
contact with each other are protected from solvent and reagents
in the bulk solution, and complexation increases the components’
stabilities.1 While the complex exists it may also express new
forms of stereochemistry2 and stereoselectivity,3 reactive inter-
mediates can be generated with protracted lifetimes,4 and even
catalysis of reactions can take place.5 If the reaction involves
covalent bonding between host and guest, a self-complementary
structure results, and the new molecule may catalyze its own
formation: it may replicate.6 Complementarity of the host
and guest is the germ of these possibilities and we examine
here hosts and guests of the most primitive sort of shape
complementarity: concave with convex.
Cavitandssmolecular cavities with one open endsare among
the simplest of concave molecules and they hold a prominent
place in the history of synthetic host structures.7 Specifically,
the resorcinarenes 1 (Scheme 1) present a bowl-shaped inner
surface and are accessible by Ho¨gberg’s8 high-yielding con-
densation of resorcinol with various aldehydes. These were used
(1) (a) Cram, D. J.; Cram, J. M. Container Molecules and Their Guests;
Royal Society of Chemistry: Cambridge, 1994. (b) Conn, M. M.; Rebek,
J., Jr. Chem. ReV. 1997, 97, 1647. (c) Rebek, J., Jr. Acc. Chem. Res. 1999,
32, 278-286.
(2) van Wageningen, A. M. A.; Timmerman, P.; van Duynhoven, J. P.
M.; Verboom, W.; van Veggel, F. C. J. M.; Reinhoudt, D. N. Chem. Eur.
J. 1997, 3, 639-654.
(9) (a) Kikuchi, Y.; Kato, Y.; Tanaka, Y.; Toi, H.; Aoyama, Y. J. Am.
Chem. Soc. 1991, 113, 1349-1354. (b) Kobayashi, K.; Asakawa, Y.;
Kikuchi, Y.; Toi, H.; Aoyama, Y. J. Am. Chem. Soc. 1993, 115, 2648-2654.
(10) (a) Moran, J. R.; Karbach, S.; Cram, D. J. J. Am. Chem. Soc. 1982,
104, 5826. (b) Cram, D. J.; Karbach, S.; Kim, H.-E.; Knobler, C. B.;
Maverick, E. F.; Ericson, J. L.; Helgeson, R. C. J. Am. Chem. Soc. 1988,
110, 2229. (c) Reference 1a, pp 85-130.
(11) For reviews, see: (a) Jasat, A.; Sherman, J. C. Chem. ReV. 1999,
99, 931-967. (b) Reference 1a, pp 131-216.
(12) (a) Moran, J. R.; Ericson, J. L.; Dalcanale, E.; Bryant, J. A.; Knobler,
C. B.; Cram, D. J. J. Am. Chem. Soc. 1991, 113, 5707-5714. (b) Cram, D.
J.; Choi, H.-J.; Bryant, J. A.; Knobler, C. B. J. Am. Chem. Soc. 1992, 114,
7748-7765.
(3) Rivera, J. M.; Mart´ın, T.; Rebek, J., Jr. Science 1998, 279, 1021-
1023.
(4) (a) Cram, D. J.; Tanner, M. E.; Thomas, R. Angew. Chem., Int. Ed.
Engl. 1991, 30, 1024-1027. (b) Warmuth, R. Angew. Chem., Int. Ed. Engl.
1997, 36, 1347-1350. (c) Beno, B. R.; Sheu, C.; Houk, K. N.; Warmuth,
R.; Cram, D. J. Chem. Commun. 1998, 301-302.
(5) Kang, J.; Santamaria, J.; Hilmersson, G.; Rebek, J., Jr. J. Am. Chem.
Soc. 1998, 120, 7389-7390.
(6) Wintner, E. A.; Conn, M. M.; Rebek, J., Jr. Acc. Chem. Res. 1994,
27, 198-203.
(7) (a) Cram, D. J. Science 1983, 219, 1177-1183. (b) Rudkevich, D.
M.; Rebek, J., Jr. Eur. J. Org. Chem. 1999, 64, 1991-2005.
(8) (a) Ho¨gberg, A. G. S. J. Am. Chem. Soc. 1980, 102, 6046-6050. (b)
Ho¨gberg, A. G. S. J. Org. Chem. 1980, 45, 4498-4500.
(13) (a) Rudkevich, D. M.; Hilmersson, G., Rebek, J., Jr. J. Am. Chem.
Soc. 1997, 119, 9911-9912. (b) Rudkevich, D. M.; Hilmersson, G.; Rebek,
J., Jr. J. Am. Chem. Soc. 1998, 120, 12216-12225. (c) Ma, S.; Rudkevich,
D. M.; Rebek, J., Jr. Angew. Chem., Int. Ed. Engl. 1999, 38, 2600-2602.
10.1021/ja993922e CCC: $19.00 © 2000 American Chemical Society
Published on Web 04/26/2000

4574 J. Am. Chem. Soc., Vol. 122, No. 19, 2000
Renslo et al.
Scheme 1a
a Conditions: (a) 3.0 equiv of 1,2-difluoro-4,5-dinitrobenzene, 12 equiv of Et3N, DMF, 70 °C, 18 h, (45-56%); (b) H2, RaNi, toluene, 40 °C,
15 h; (c) 6 equiv of C7H15COCl, 12 equiv of K2CO3, EtOAc-H2O; (d) 10 equiv of H2NNH2-H2O, EtOH-toluene, reflux, 3 h, (25-30% overall
from 8); (e) 1.1 equiv of 14, DMF, 6.0 equiv of Et3N, 85 °C, 2 d, 55%; (f) H2, Pd/C, EtOAc, room temperature, 2 h; (g) 1.2 equiv of C6F5OH, 2.0
equiv of EDC-MeI, cat. DMAP, THF, room temperature, 20 h, 80% from 10.
on the surface, with shape complementarity to boot. Indeed
cubane proved an excellent guest for the related calixarene
capsules, but simple substituted derivatives are much less
available for screening than adamantanes or cyclohexanes. The
most appropriate complementssmolecules with a thin layer of
positive chargeswere found in quinuclidinium salts, and we
have made much use of these as guests for characterization of
complexes in the gas phase by electrospray mass spectrometry.14
These guests lack sufficient solubility for the solution studies
contemplated here.
Finally, the hydrogen bonding characteristics of the guest’s
functions also contribute to binding. Substituents that feature
hydrogen bond donors or acceptors can interact with the seam
of amides that hold the host together; substituents that feature
both, such as secondary amides, are especially welcome.13 The
sum of these recognition forces is modest; typically e2 kcal/
mol in binding affinity is observed for complexation in organic
solvents; their thermodynamic stability is low. What makes these
systems special is their unprecedented kinetic stability. The
energetic barriers for exchange of guests in and out of the
cavities are among the highest observed for open-ended
vessels.7b For example, adamantane exchange in xylene occurs
with a free energy of activation, ∆Gq of 17 kcal/mol.13a
Accordingly, the rate is slow on the NMR time scale (600 MHz,
ambient temperatures) and separate signals are observed for free
species and the complexes. The resonances of the complexed
guests appear far upfield, between 0 and -3 ppm. This open
window of the spectra allows easy determination of concentra-
tions, from which equilibrium constants and thermodynamic
parameters can be calculated. The lifetimes of the complexes
are long enough to expect that many types of reaction processes
could occur between host and guest.
The self-complementary cavitands (SCCs) 3 and 4 (Chart 1)
described here are cousins of cavitand 2 to which adamantane
guests were covalently attached on their upper rims. The
resulting structures could, from their self-complementarity, form
cyclic oligomers of any size or even polymerize, but they form
dimers in solution. The cooperative nature of the process gives
high thermodynamic and kinetic stability to the dimeric
complexes.
These are not the first self-including cavitand dimers. Efforts
directed at the deepening of cavitands to accommodate increas-
ingly larger guests earlier produced two other examples. The
tetraester 6 prepared by Cram15 and co-workers formed a dimeric
species (Chart 2) in the solid state and in solution at high
concentrations (50 mM). Likewise, the tetraamide 7 forms a
hydrogen bonded dimer in which one upper-rim alkyl chain from
each monomer is bound within the cavity of the dimer.16 The
SCCs described here were engineered to maximize the efficacy
of a self-complementary binding motif. We now provide a full
account of our studies directed at the synthesis and assembly
of these self-complementary molecules.17
Results and Discussion
Synthesis. The synthesis of cavitands featuring a unique site
required the selective tri-arylation18 of Ho¨gberg’s resorcinarene
1. This was accomplished as illustrated in Scheme 1. Reaction
of the resorcinarene 1 with 3 equiv of 1,2-difluoro-4,5-dinitro-
(15) von dem Bussche-Hu¨nnefeld, C.; Helgeson, R. C.; Bu¨hring, D.;
Knobler, C. B.; Cram, D. J. Croat. Chem. Acta 1996, 69, 447-458.
(16) Ma, S.; Rudkevich, D. M.; Rebek, J., Jr. J. Am. Chem. Soc. 1998,
120, 4977-4981.
(17) For the preliminary communication see: Renslo, A. R.; Rudkevich,
D. M.; Rebek, J., Jr. J. Am. Chem. Soc. 1999, 121, 7459-7460.
(18) Selective tri-arylation of a resorcinarene with quinoxaline dichlorides
was reported previously, see: (a) Soncini, P.; Bonsignore, S.; Dalcanale,
E.; Ugozzoli, F. J. Org. Chem. 1992, 57, 4608-4612. (b) Reference 10a.
(14) Schalley, C. A.; Castellano, R. K.; Brody, M. S.; Rudkevich, D.
M.; Siuzdak, G.; Rebek, J., Jr. J. Am. Chem. Soc. 1999, 121, 4568-4579.

Synthesis of Self-Complementary CaVitands
J. Am. Chem. Soc., Vol. 122, No. 19, 2000 4575
Chart 1
Chart 2
Scheme 2a
benzene in DMF/Et3N produced the desired hexanitro cavitand
8 (45-56%) which exists in equilibrium between the “kite” and
“vase” conformations. A smaller quantity (<20%) of the fully
bridged, octanitro compound is also formed but could be easily
separated from 8 by column chromatography. Installation of
the upper-rim seam of head-to-tail amides was initiated by
reduction of the hexanitro compound with Raney Ni and H2 in
toluene. The resulting hexaamino cavitand was immediately
acylated with octanoyl chloride under Schotten-Baumann
conditions (K2CO3, EtOAc-H2O). The free phenolic groups
were acylated simultaneously, but the resulting esters were
cleaved by brief treatment with H2NNH2 in toluene-EtOH. This
provided the hexaamide cavitand 9 (R′ ) C7H15) in 25-30%
overall yield from the hexanitro cavitand 8. Although the
hydrogen-bonded seam in hexaamide 9 is incomplete, the
cavitand still adopts the desired folded conformation in solution
as evidenced by NMR (vide infra).
a Conditions: (a) 1.3 equiv of 1-aminoadamantane, 4.0 equiv of Et3N,
DMF, room temperature, 16h (79% 3a) or 1.3 equiv of 1-methyl-
aminoadamantane, 4.0 equiv of Et3N, CH2Cl2, room temperature, 24 h
(70% 3b).
At this juncture, the syntheses of cavitands 3 and 4 diverge.
The SCCs 3 are most conveniently and reliably prepared from
the activated ester 12 (Scheme 1). Hence, coupling of the
quinoxaline dichloride 14 with hexaamide 9 in DMF at 85 °C
produces the desired benzyl ester 10 in 55% yield. Hydro-
genolysis of 10 (H2, Pd/C) and activation of the resulting acid
11 (C6F5OH, EDC-MeI, cat. DMAP, THF) provides the
pentafluorophenyl ester 12 in 80% overall yield. This activated
ester is then employed in the synthesis of SCCs 3a and 3b by
reaction with adamantane nucleophiles (1-amino or 1-amino-
methyl) under mild conditions (DMF or CH2Cl2 at room
temperature, Scheme 2). The formation of cavitands 4 in which
only one position on the quinoxaline heterocycle has undergone
reaction was unexpected and is, to the best of our knowledge,
without literature precedent. It took us sufficiently by surprise
that we mis-assigned their structures in our recent communica-
tion17 and here we correct them. To be sure, reaction of
hexaamide 9 with quinoxalines 13-15 under standard coupling
conditions (DMF-Et3N, 65 °C), as in Scheme 3, gives SCCs 3
as reported. It is, however, the minor product and not the one

4576 J. Am. Chem. Soc., Vol. 122, No. 19, 2000
Renslo et al.
Scheme 3a
functions (H2, RaNi, toluene) produced the diamine 17. Con-
densation of this diamine with 1,2-diketopyracene produced
cavitand 5.
Complexation in Open Containers. The octaamide 2 has
been characterized in detail13 and, as mentioned above, ada-
mantane or cyclohexane derivatives and some lactams are bound
with association constants generally e200 M-1. In the present
study, we investigated the thermodynamics of binding in
octaamide cavitand 2 and also cavitand 5 using N-(1-adamantyl)-
acetamide as guest in p-xylene-d10 and toluene-d8 solvent,21
respectively. Although cavitand 5 lacks a complete hydrogen
bonding seam (as in 2), guest exchange remains slow on the
NMR time scale. Variable temperature NMR was used to
determine equilibrium constants and van’t Hoff plots provided
the following thermodynamic parameters. Guest binding in 2
and 5 is weak (∆G295 ) -2.2 kcal/mol and ∆G270 ) -1.5 kcal/
mol, respectively) and enthalpically driven and entropy opposed
(∆H ) -5.5 and -2.1 kcal/mol; ∆S ) -11 and -2.2 eu,
respectively). The favorable binding enthalpy for 2 and 5 is
attributed to van der Waals and CH-π contacts between the
aromatic walls of the cavity and the aliphatic guests, and some
hydrogen bonding interactions where possible. The entropy of
binding in 2 is unexceptional for bimolecular host-guest
complexation in organic solvents. For 5 both the enthalpy and
entropy are unusually small. Since the same guest is involved
and the complexes are expected to be similar in structure, the
peculiarity of 5 must arise from its resting (ground) state in the
toluene solvent.
a Conditions: (a) 1.1 equiv of quinoxaline 13-15, 6.0 equiv of Et3N,
DMF, 65 °C, 20-40 h.
whose NMR spectra were published. Those spectra correspond
to the major products, the “half-bridged” cavitands 4.19 The mis-
assignment was discovered only through the independent
synthesis of 3 via the activated ester 12, as outlined in Scheme
2. The use of more forcing reaction conditions did not lead to
The spectroscopic properties and complexation behavior of
the partially substituted hexaamide cavitand 9 were also
examined. Compared to 2, hexaamide 9 lacks both a complete
hydrogen bonding seam and also one of the four aromatic
residues that make up the cavitand “walls”. Downfield signals
in the 1H NMR spectrum of hexaamide 9 and an N-H stretching
band at 3234 cm-1 (FTIR) indicate the formation of upper-rim
hydrogen bonding. At the same time, the presence of methine
signals at 6.2 (3 H) and 4.7 ppm (1 H) is characteristic of a
fully folded “vase”12 conformation. Accordingly, the presence
of four aromatic “walls” and a complete seam of hydrogen bonds
(as in octaamide 2) is not a prerequisite of the folded
conformation. Most surprisingly, the partially formed cavity in
hexaamide 9 binds guest molecules with an affinity comparable
to the cavity in octaamide 2 (Figure 1). Association constants
of Kass ) ∼40 and ∼60 M-1 (295 K, p-xylene-d10) were
calculated for the binding of N-(1-adamantyl)acetamide in
octaamide 2 and hexaamide 9, respectively. A broadening of
signals for guests bound in hexaamide 9 (Figure 1B) probably
reflects a higher exchange rate constant, but exchange remains
slow on the NMR time scale at room temperature!
a significant increase in the proportion of the fully bridged SCCs
3 except in reactions with the benzyl ester-substituted quinoxa-
line 14. These results may reflect conformational constraints
that prevent the nucleophilic and electrophilic sites in SCCs 4a
and 4b from reacting to form the fully bridged systems.20
Fortunately, the dimerization behavior of fully bridged and
partially bridged SCCs is a matter of degree rather than kind.
A second approach to the synthetic elaboration of 9 is
illustrated in Scheme 4 for the synthesis of the model compound
5. This cavitand lacks a complete hydrogen-bonding seam and
bears a large acenaphthene surface in lieu of the quinoxalines
of 3 and 4. Hence, reaction of the hexaamide 9 with 1,2-difluoro-
4,5-dinitrobenzene (DMF-Et3N, 70 °C) provided the corre-
sponding dinitro cavitand (16) which upon reduction of the nitro
Characterization. Self-complementary cavitands 3 and 4
were characterized by 1H NMR, FTIR, and MALDI-TOF
spectrometry. The FTIR spectra of 3 and 4 in CHCl3 show
hydrogen-bonded N-H signals at 3238-3245 cm-1 and less
intense signals for free N-H stretching at 3407-3416 cm-1
.
1
Figure 2 shows portions of the H NMR spectra of SCC 3a in
CDCl3, benzene-d6, and p-xylene-d10 solution. In CDCl3 (Figure
2A), three hydrogen-bonded N-H signals (out of six) are
observed downfield of 9.5 ppm while in p-xylene-d10 solution
(Figure 2C) five of the six N-H signals are located in this
position. In both solvents, one amide N-H is not hydrogen
bonded and its resonance appears upfield near 6 ppm. The
(19) A single regioisomer is formed. We presume that nucleophilic attack
occurs at the more electrophilic position on quinoxaline dichlorides to give
the regioisomer shown (i.e., SCCs 4).
(20) Dalcanale and Cram report the successful coupling of an unactiVated
quinoxaline dichloride with a trisubstituted quinoxaline cavitand (see ref
18). The partial coupling we observe might be a consequence of the
additional functionality present in our partially substituted cavitand (i.e.,
the C7H15 amide groups around the upper rim of the structure).
(21) The weak binding of guests in cavitand 5 required study at
temperatures below the melting point of p-xylene-d10 and so toluene-d8
solvent was used with 5.

Synthesis of Self-Complementary CaVitands
J. Am. Chem. Soc., Vol. 122, No. 19, 2000 4577
Scheme 4a
a Conditions: (a) 2.0 equiv of 1,2-difluoro-4,5-dinitrobenzene, 10 equiv of Et3N, DMF, 70 °C, 14 h (70-85% for 16); (b) H2, RaNi, toluene, 40
°C, 3 h, (65% for 17); (c) 3.0 equiv of 1,2-diketopyracene, THF-AcOH, reflux, 12-16 h, (50% for 5).
1
Figure 1. Portions of the H NMR spectra (600 MHz, p-xylene-d10, 295 K) of N-(1-adamantyl)acetamide in the presence of cavitand 2 (A) and
cavitand 9 (B). Signals for residual protonated solvent and other solvent impurities are indicated by filled and open circles, respectively.
1
Figure 2. Portions of the H NMR spectra (600 MHz, 295 K) of cavitand 3a in CDCl3 (A), benzene-d6 (B), and p-xylene-d10 (C). Solvent peaks
are denoted as before.
spectrum of 3a is considerably more complex in benzene-d6
(Figure 2B) than in CDCl3 or p-xylene-d10, and close inspection
reveals the presence of two species in roughly equal quantities.
The methine signals of the cavitand are located near 6 ppm in
all three solvents, as expected for the folded conformation. The
most striking difference between the spectra is found in the far
upfield region between 0 and -2 ppm. Whereas no upfield
signals are observed in CDCl3, in benzene-d6 and p-xylene-d10
complex than that of 3a in the same solvent whereas in CDCl3
both SCCs show relatively simple and similar resonances.
1
Figure 3 presents the H NMR spectra of the half-bridged
SCCs 4a (Figure 3A) and 4b (Figure 3B) in p-xylene-d10
solution published earlier as 3a and 3b. Methine signals near 6
ppm are indicative of the folded “vase” conformation. The
aromatic and far downfield portions of the spectra are similar
and, as with SCCs 3, both spectra feature adamantane signals
in the upfield window. The adamantane and cavitand signals
in 4b are more complex than are those in 4a, much as the
spectrum for 3b is more complex than that of 3a. Compared to
fully bridged SCC 3a, the adamantane signals in 4a are shifted
downfield by ∼0.5-0.7 ppm (compare C in Figure 2 and A in
1
a full set of adamantane signals are present. H NMR spectra
of the homologous SCC 3b are similar to those of 3a: both
possess downfield N-H resonances and upfield adamantane
signals (in p-xylene-d10 and benzene-d6 but not in CDCl3). The
1H NMR spectrum of 3b in p-xylene-d10 is significantly more

4578 J. Am. Chem. Soc., Vol. 122, No. 19, 2000
Renslo et al.
Figure 3. Portions of the 1H NMR spectra (600 MHz, p-xylene-d10, 295 K) of cavitands 4a (A) and 4b (B). Solvent peaks are denoted as before.
1H NMR data for the half-bridged SCCs 4 suggest that they
too form strong assemblies in noncompetitive solvents. This is
perhaps not surprising since these cavitands also possess self-
complementary shapes. The significantly different chemical
shifts observed for bound adamantane in SCCs 3 and 4 suggest
a structurally distinct binding cavity in the two classes of SCCs.
The fully formed cavity in SCCs 3 provides a more shielded
microenvironment for adamantane and its signals are shifted
further upfield. The more exposed cavity of the half-bridged
SCCs 4 is less shielding and the corresponding upfield shift is
smaller in magnitude. This analysis is undoubtedly an over-
simplification since the specific orientation of adamantane in
the binding cavity is also important and may well contribute to
the different chemical shifts observed in these two families of
Figure 4. Upfield region of the 1H NMR spectra (600 MHz, p-xylene-
d10, 295 K) of cavitands 4a (A), 4b (B), and a mixture of 4a and 4b
(C). Signals attributed to cavitand 4a in heterodimer 4a‚4b are denoted
with asterisks.
SCCs. Indeed, the adamantane signals observed for half-bridged
cavitands 4a and 4b are quite different, the signals in 4b being
less well separated and shifted less upfield (Figure 3). Since
the cavities in 4a and 4b are identical, adamantane orientation
within the cavity must be quite different in the two dimers, thus
producing the dissimilar signals observed by NMR spectroscopy.
Figure 3). When two different SCCs (4a and 4b) were mixed
The half-bridged SCCs 4 form significantly stronger as-
in p-xylene-d10 and an NMR spectrum was recorded, a new set
semblies than their fully bridged brethren. In p-xylene-d10 at
of bound adamantane signals appeared along with the original
signals for 4a and 4b (Figure 4).
mM concentrations, only dimeric species are observed for SCCs
4a and 4b, suggesting dimerization constants Kdim g 105 M-1
.
In contrast to SCCs 3, significant dimerization of half-bridged
SCCs 4 is observed in CDCl3, a solvent that competes well for
the cavity. Molecular modeling provides a possible explanation
for the increased affinity. The single-bond connectivity between
cavity and quinoxaline ring in SCCs 4 allows for the formation
of a more extended dimeric structure (Figure 7) in which the
upper-rim amide substituents are completely free of conforma-
tional constraints. This more open structure also permits
favorable aromatic stacking interactions between the quinoxaline
rings of the dimer.
The unsymmetrical quinoxaline wall in SCCs 3 and 4 renders
the cavitands inherently chiral. The presence of a directional
hydrogen bonding seam (clockwise or counterclockwise) is
expected to result in conformational diastereomers. However,
NMR spectra of 3a and 4a (Figures 2 and 3) reveal only six
amide NH resonances, suggesting that the chirality of the
cavitand leads to a preference for one of the two possible
cyclodiastereomers. Rapid interconversion of cyclodiastereomers
seems a less likely explanation considering the nonpolar solvents
employed and our experience with octaamide 2 (in which the
interconversion is slow on the NMR time scale).
Discussion
The NMR and FTIR data for SCCs 3 and 4 confirm the
presence of hydrogen bonding along the upper rim of the
structures. The data also suggest stronger hydrogen bonding in
less competitive solvents such as p-xylene-d10 and somewhat
weaker hydrogen bonding in CDCl3. The observed chemical
shifts of the methine protons provide further evidence in support
of the desired “vase” conformation, as does the presence of
adamantane signals in the far upfield window. This evidence
of adamantane inclusion is strongly suggestive of dimerization
both from an entropic standpoint and from the inspection of
molecular models. The possibility of intramolecular self-
inclusion can be discounted because the tether between cavity
and adamantane is simply too short. Models of higher-order
assemblies reveal significant conformational constraints on the
upper-rim amide substituents (R′ ) C7H15). Minimized struc-
tures of dimeric species (Figures 6 and 7) reveal less significant
conformational constraints. Integration of the signals for bound
adamantane and total methine protons allowed the determination
of a dimerization constant (Kdim). In the case of SCC 3a, a value
of Kdim ) 3000 M-1 was calculated. This value is roughly the
square of that for the binding of adamantane guests in open-
ended cavitands such as 2 (Kass ) ∼ 40 M-1). The stronger
binding observed in SCCs is likely a consequence of their self-
complementary shape and a corresponding cooperative action
of binding sites.
Even assuming a preference for one monomeric cyclodias-
tereomer, the assembly process itself is expected to involve the
formation of two diastereomeric dimerssas enantiomer pairs
(denoted here “R‚R/S‚S” and “R‚S/S‚R”). Modeling the “R‚R/
S‚S” diastereomer reveals a C2 rotational axis of symmetry
whereas the “R‚S/S‚R” diastereomer possesses an inversion

Synthesis of Self-Complementary CaVitands
J. Am. Chem. Soc., Vol. 122, No. 19, 2000 4579
Figure 5. Portions of the 1H NMR spectrum (600 MHz, benzene-d6, 295 K) of cavitand 5. Signals for the acenaphthene residue are denoted with
filled circles (assembled state) and open circles (monomeric state).
Figure 6. Stereoview of dimer 3a‚3a from the minimized structure obtained using the Amber* force field in MacroModel 6.5. The C7H15 and
C11H23 chains were replaced by CH3 to expedite the computation. One aromatic wall of each monomeric species was removed (after minimization)
for viewing clarity.
Figure 7. Stereoview of dimer 4b‚4b from the minimized structure obtained using the Amber* force field in MacroModel 6.5. The C7H15 and
C11H23 chains were replaced by CH3 to expedite the computation. One aromatic wall of each monomeric species was removed (after minimization)
for viewing clarity.
Table 1. Dimerization Constants (Kdim) for Self-Complementary
center. Hence, a single set of cavitand resonances would be
Cavitands in Various Solvents at 295 K
expected for each diastereomeric dimer (i.e., the two halves of
p-xylene-d10
benzene-d6
CDCl3
nd
nd
each dimer are magnetically equivalent). In fact, the presence
of diastereomeric dimers is evident22 in the complex NMR
spectra observed for SCCs 3b and 4b in p-xylene-d10 solvent,
and moreover, modest diastereoselectivity (ca. 3:1) is apparent
in the case of 4b. Much to our surprise, NMR spectra of dimeric
3a and 4a possess a single set of cavitand signals (Figures 2C
and 3A). That SCCs 3a and 4a should display such a high degree
of diastereoselectivity is not readily explained by our modeling
studies to date. It is unclear from modeling if the quinoxaline
amides participate in hydrogen bonding or if the alkyl groups
of the dimer are interdigitated. Lacking structural information
of crystallographic detail, we hesitate to speculate at this time
3a
3b
10
4a
4b
4c
4d
3000 M-1
1800 M-1
nda
325 M-1
50 M-1
nd
-
-
-
nd
nd
g105 M-1
g105 M-1
50 M-1
nd
250 M-1
1000 M-1
nd
nd
a nd ) no dimer.
on the identity of the preferred diastereomer or the source of
the high diastereoselectivity observed.
Table 1 presents dimerization constants for SCCs 3 and 4 in
various solvents. As mentioned previously, the assembly process
is highly solvent dependent. Not surprisingly, the best solvents
(22) This is supported by the observation of chemical exchange cross-
peaks between adamantane signals in the ROESY spectrum of 4b‚4b.

4580 J. Am. Chem. Soc., Vol. 122, No. 19, 2000
Renslo et al.
Table 2. Thermodynamic Parameters for the Dimerization of
Self-Complementary Cavitands
dimerization is enthalpically favorable due to a multitude of
weak binding interactions and a shape complementarity that
favors a stronger, more well-organized hydrogen bonding seam
(as compared to the solvent-filled monomeric species).
cavitand
solvent
∆H, kcal/mol
∆S, cal/(mol K)
3a
3b
4a
p-xylene-d10
p-xylene-d10
CDCl3
-17.2
-29.9
-10.6
-41
-85
-25
The free energies of dimerization in 3 and 4 are generally
small (∆G ≈ -2.5 to -6.7 kcal/mol), but the compensating
enthalpic and entropic contributions can be very significant
(Table 2). Whereas the open-ended cavitand 2 binds adamantane
derivatives with values of ∆H ) -5.5 kcal/mol and ∆S ) -11
eu, the corresponding values for dimerization of 3 and 4 are
between two and six times larger. Since dimerization involves
two binding events it might be reasonable to expect ∆H values
twice as high (as is the case for dimer 4a‚4a), but what is the
source of the additional binding enthalpy for dimer 3‚3? We
suggest that interactions between the upper-rim C7H15 chains
in dimer 3‚3 may be partially responsible. The more compact
structure of dimer 3‚3 (as compared to 4‚4) forces several amide
C7H15 chains into close proximity (Figures 6 and 7). To avoid
occupying the same space, some of these chains might
interdigitatesan entropically costly arrangement that might
provide compensating enthalpic benefit24 through a multitude
of van der Waals contacts. In contrast, the amide substituents
of the more extended dimer 4‚4 are free of such interactions
and smaller ∆H and ∆S values are observed (Table 2).25 It
is interesting to note that the formation of self-included dimer
7‚7 (Chart 2) occurs with similarly large enthalpy-entropy
compensation (∆H ) -21 kcal/mol, ∆S ) -58 eu).16
for assembly are those that are themselves poor guests. Hence,
p-xylene-d10, a flat aromatic that is a poor match for the nearly
cylindrical cavities in 3 and 4, generally produces the strongest
assemblies. Somewhat weaker assemblies are observed in other
aromatics such as benzene-d6 and toluene-d8. The roughly
spherical CDCl3 competes well for the cavity. The acidity of
its C-D function evidently overcomes the effect of the many
nonbonded electrons of the solvent and may also be involved
in weakening the intramolecular hydrogen bonds. The upshot
is that dimerization is less favored, when it occurs at all. The
polar aprotic solvent DMF-d7 is expected to be a good guest
for the cavity. Moreover, it can disrupt the hydrogen bonds and
thereby compromise the structural integrity of the cavitysno
dimerization is observed.
Some other features of the binding cavity and the guest
species can be identified (Table 1). Adamantane derivatives are
most effective at stabilizing the self-complementary binding
motif. The connection between cavitand and adamantane is
important with amide linkages being much preferred to esters
(compare 4b and 4c).23 Aromatics (e.g., the benzyl ester 10 and
4d) are no better guests for the cavity than solvent and hence
no assembly is observed. Finally, the increased conformational
freedom and surface contact present in half-bridged dimers 4‚4
produce more stable assemblies than the fully bridged SCCs 3.
Our experience to date with cavitands 3-5 and related
derivatives suggests that their aggregation in solution may be a
general phenomena. For example, we were surprised to find
that NMR spectra of cavitand 5 contain peculiarities suggestive
Conclusions
Cavitands with self-complementary surfaces (convex and
concave) associate in solution to form complexes of high
thermodynamic and kinetic stabilities. The cooperation of
binding sites in these host-guest hybrids leads to much stronger
association than is observed for the component parts alone
(adamantane and cavitand). The intermolecular interactions
involved are subtle and assembly occurs with significant
enthalpy-entropy compensation. Because they are self-comple-
mentary structures, a question can be raised as to whether the
systems are capable of self-replication. We are currently
exploring this possibility and will report on our findings in due
course.
1
of an assembled state. As shown in Figure 5, the H NMR
spectrum of 5 in benzene-d6 solution possesses a separate set
of acenaphthene signals that are shifted upfield by ca. 3 ppm
(filled circles, Figure 5). The observation of acenaphthene
inclusion cannot be reconciled with a strictly monomeric species,
particularly since the intensity of the upfield acenaphthene
signals is temperature and concentration dependent. Despite
expending considerable experimental effort, we cannot be sure
of the specific nature of the aggregation, but it appears not to
be a simple symmetrical dimer as is observed with 3 and 4.
We are currently exploring the unusual behavior of cavitand 5
and will report on it more fully in a subsequent report.
Thermodynamics of Assembly. The thermodynamic param-
eters of assembly in SCCs 3 and 4 were examined using
variable-temperature NMR. The slow exchange between mono-
meric and dimeric species allows concentrations and hence
association constants to be easily determined. A van’t Hoff plot
was then used to derive ∆H and ∆S values for the assembly
process. The dimerization was studied using p-xylene-d10 as
solvent whenever possible. In the case of SCC 4, only dimeric
species are observed in p-xylene-d10 so CDCl3 solvent was used
in this case. The results are summarized in Table 2.
Experimental Section
General. Melting points were determined on a Thomas-Hoover
1
capillary melting point apparatus and are uncorrected. H NMR and
13C NMR spectra were recorded on Bruker AM-300 and DRX-600
spectrometers. The chemical shifts were measured relative to residual
nondeuterated solvent resonances. Matrix-assisted laser desorption/
ionization (MALDI) mass spectrometry experiments were performed
on a PerSeptive Biosystems Voyager-Elite mass spectrometer with
delayed extraction, using 2,5-dihydroxybenzoic acid (DHB) as a matrix.
Electrospray ionization (ESI) mass spectra were recorded on an API
III Perkin-Elmer SCIEX triple quadrupole mass spectrometer. HRMS
values generally agreed to e5 ppm for compounds with molecular
weight e500 and e10 ppm for compounds with higher molecular
weight g2000.26 FTIR spectra were recorded on a Perkin-Elmer Paragon
1000 PC FT-IR spectrometer. Column chromatography was performed
with Silica Gel 60 (EM Science or Bodman, 230-400 mesh). All
experiments with moisture- or air-sensitive compounds were performed
The dimerization of cavitands 3 and 4 is enthalpically favored
and entropy opposed. The inclusion of adamantane upon
(24) The source of the increased binding enthalpy and entropy in 3b (as
compared to 3a) is not readily apparent. Without detailed (crystallographic)
information regarding the structure of these dimers, this issue cannot be
meaningfully addressed.
(25) This analysis necessarily ignores solvent effects on the thermo-
dynamics of assembly. A direct comparison of SCCs 3 and 4 in the same
solVent was not possible due to the strictly dimeric structure of 4 in p-xylene-
d10 solvent.
(23) The surprisingly large effect of this change suggests that the amide
plays a crucial role in stabilizing the dimeric structure, probably as a
hydrogen bond donor/acceptor. The substitution of amide for ester is
destabilizing to a greater extent than simple loss of a hydrogen bond donor
and so it seems likely that lone pair-lone pair repulsion is additionally
destabilizing in the case of 4c.

Synthesis of Self-Complementary CaVitands
J. Am. Chem. Soc., Vol. 122, No. 19, 2000 4581
in anhydrous solvents under a nitrogen atmosphere. Compounds 1,27
1,2-difluoro-4,5-dinitrobenzene,28 1,2-diketopyracene,29 and 2,3-di-
hydroxy-6-quinoxalinecarboxylic acid30 were synthesized in accord
with the literature protocols. Molecular modeling was performed using
the Amber* force field in the MacroModel 5.5 program.31
Benzyl Ester 14. A suspension of 2,3-dihydroxy-6-quinoxaline-
carboxylic acid (0.207 g, 1.0 mmol) in SOCl2 (4 mL) was treated with
a catalytic amount of DMF and heated at 80 °C for 3 h. After being
cooled to room temperature the clear yellow solution was concentrated
to give 2,3-dichloro-6-quinoxalinecarbonyl chloride as a tan solid. This
material was then dissolved in 5 mL of CH2Cl2 and cooled at 0 °C.
The solution was treated with triethylamine (0.404 g, 0.57 mL, 4.0
mmol) and then benzyl alcohol (0.108 g, 0.103 mL, 1.0 mmol) and
allowed to warm to room temperature. After being stirred for 16 h, the
reaction mixture was poured into a mixture of 10% K2CO3 and CH2-
Cl2. The layers were separated and the organic phase washed with brine,
dried over MgSO4, filtered, and concentrated to give ca. 0.30 g of the
crude product. Column chromatography (0-20% EtOAc-hexane)
provided 0.261 g of 14 as a tan solid: mp 111-112 °C; FTIR (CHCl3,
Hexanitrocavitand 8. Triethylamine (2.23 mL, 16.0 mmol) was
added dropwise to a stirred solution of resorcinarene 1 (1.10 g, 1.0
mmol) and 1,2-difluoro-4,5-dinitrobenzene (0.612 g, 3.0 mmol) in
anhydrous DMF (50 mL). The resulting mixture was then heated to 70
°C and kept at that temperature for 16 h. The reaction mixture was
cooled to room temperature and poured into dilute aqueous HCl (pH
∼1, 400 mL). The bright yellow solids were filtered off, washed with
large amount of water, and dried in vacuo. Column chromatography
(CH2Cl2, then 2% MeOH-CH2Cl2) provided first the octanitro
compound (0.370 g; 0.21 mmol; 21%) and then the desired hexanitro
cavitand 8 (0.910 g; 0.57 mmol; 57%) as a yellow solid after trituration
with MeOH: mp >250 °C; FTIR (CHCl3, cm-1) υ 3578, 2919, 2851,
1590, 1543, 1490, 1343, 1286, 1195; 1H NMR (600 MHz, CDCl3,
330K) δ 7.63 (s, 2 H), 7.61 (s, 2 H), 7.58 (s, 2 H), 6.95 (s, 2 H), 6.84
(s, 2 H), 6.70 (s, 2 H), 6.60 (s, 2 H), 5.85 (s, 2 H), 4.31 (t, J ) 7 Hz,
1 H), 4.00-3.97 (m, 3 H), 2.10-2.01 (m, 8 H), 1.36-1.28 (m, 72 H),
0.90 (t, J ) 7 Hz, 12 H); MS-ESI- m/z 1597 ([M - H + 13C]- calcd
for C89H111N6O2013C 1597).
1
cm-1) υ 2995, 1719, 1452, 1381, 1295, 1267, 1152, 1119, 1000; H
NMR (600 MHz, CDCl3) δ 8.78 (d, J ) 1.4 Hz, 1 H), 8.44 (dd, J )
1.4, 9 Hz, 1 H), 8.10 (d, 9 Hz, 1 H), 7.51-7.35 (m, 5 H), 5.46 (s, 2
H); 13C NMR (150 MHz, CDCl3) δ 164.8, 147.5, 146.6, 142.5, 139.9,
135.3, 132.5, 130.97, 130.7, 128.7, 128.6, 128.5, 128.4, 67.6; HRMS-
MALDI-FTMS m/z 333.0194 ([M + H]+, calcd for C16H11Cl2N2O2
333.0198, error 1.2 ppm). Analogously, 2,3-dichloroquinoxalines 13a,b
and 15 were synthesized from 2,3-dichloro-6-quinoxalinecarbonyl
chloride and 1-aminoadamantane, 1-aminomethyladamantane, and
1-adamantanol.
Hexaamide 9. A solution of hexanitro cavitand 8 (3.92 g, 2.45 mmol)
in toluene (150 mL) was treated with a catalytic amount of Raney
Nickel as a suspension in EtOH (commercial H2O suspension was pre-
washed with EtOH (2 × 5 mL)). The mixture was stirred for 16 h at
40 °C under a H2 atmosphere. After cooling, the mixture was filtered
through a pad of Celite with the aid of MeOH (2 × 50 mL). The filtrates
were combined and evaporated under vacuum to give a brown solid
(3.52 g) that was protected from air and used directly in the next step.
The crude hexaamine was dissolved in 150 mL of degassed (purged
with N2) EtOAc-water (1:1) and K2CO3 (4.01 g, 29.4 mmol) was added
with vigorous stirring. Octanoyl chloride (2.39 g, 2.50 mL, 14.7 mmol)
was added at once, and the resulting mixture was stirred at room
temperature for 2 h. The solution was then diluted with half-saturated
NaHCO3 and extracted with EtOAc. The combined organic extracts
were washed with brine, dried over MgSO4, filtered, and concentrated
to give 5.40 g of the crude amide. The crude amide product was
dissolved in 60 mL of toluene-EtOH (1:1) and treated with hydrazine
hydrate (0.613 g, 0.59 mL, 12.3 mmol). The mixture was stirred at 80
°C for 3 h, cooled to room temperature, and concentrated. Column
chromatography (0-1% MeOH in CH2Cl2) provided the product 9 as
a yellow solid after trituration with MeOH (1.44 g, 27% overall): mp
>250 °C; FTIR (CHCl3, cm-1) υ 3684, 3234, 2919, 2852, 1704, 1657,
General Procedure for the Coupling of Hexaamide 9 with 2,3-
Dichloroquinoxalines: Synthesis of Cavitands 4a-d. An oven-dried
10 mL flask was charged with hexaamide 9 (0.019 g, 0.0087 mmol)
and the quinoxaline 13a (3.6 mg, 0.0096 mmol). The reactants were
dissolved in dry DMF (1 mL) and the solution treated with triethylamine
(7 µL, 0.052 mmol). The reaction mixture was heated at 65 °C for 20
h and then allowed to cool to room temperature. The solution was
poured into a mixture of EtOAc (10 mL) and dilute aqueous HCl (10
mL). The layers were separated and the aqueous phase extracted with
two 10 mL portions of ethyl acetate. The combined organic extracts
were washed with brine, dried over Na2SO4, filtered, and concentrated
to provide 0.024 g of the crude product. Column chromatography (0-
20% EtOAc-hexane) provided 9 mg (42%) of cavitand 4a as a pale
yellow oil: FTIR (CHCl3, cm-1) υ 3244, 3014, 2918, 2852, 1657, 1514,
1
1481, 1400, 1324, 1276, 895 cm-1; H NMR (Figure 3A, 600 MHz,
p-xylene-d10) δ 10.3 (s, 1 H), 9.87 (s, 1H), 9.85 (s, 1 H), 9.79 (s, 1 H),
9.68 (s, 1 H), 9.54 (s, 1 H), 8.9 (s, 2 H), 8.05-7.15 (12 s, 15 H),
6.35-6.25 (m, 4 H), 6.01 (s, 1H), 5.90 (br tr, 1 H), 3.1-1.1 (m, 152
H), 1.0-0.8 (m, 30 H), 0.8-0.7 (m, 6 H), -0.58 (s, 3 H), -0.83 (d,
J ) 11 Hz, 3 H), -1.07 (d, J ) 11 Hz, 3 H); MALDI-MS m/z 2536
([M + Na]+ calcd for C157H226ClN9O15Na 2536).
1
Cavitand 4b: 11 mg of pale yellow oil, 51%; FTIR (CHCl3, cm-1
υ 3244, 3014, 2928, 2852, 1657, 1510, 1481, 1405, 1271, 890 cm-1
)
;
1600, 1514, 1481, 1404, 1276; H NMR (600 MHz, benzene-d6, 340
K) δ 9.85 (s, 2 H), 9.55 (s, 2 H), 8.70-8.80 (br s, 2 H), 7.74 (s, 2 H),
7.72 (s, 2 H), 7.62 (s, 2 H), 7.60 (s, 2 H), 7.39 (s, 2 H), 7.36 (s, 2 H),
7.07 (s, 2 H), 6.29 (t, J ) 8 Hz, 2 H), 6.23 (t, J ) 8 Hz, 1 H), 4.73 (t,
J ) 8 Hz, 1 H), 2.5-1.20 (m, 152 H), 0.94-0.90 (m, 30 H); 13C NMR
(150 MHz, benzene-d6, 330 K) δ 173.6, 173.4, 173.2, 156.4, 156.1,
155.8, 153.0, 151.3, 151.1, 137.3, 136.9, 132.5, 130.4, 129.5, 124.7,
124.5, 122.4, 121.3, 117.4, 111.7, 38.6, 37.6, 35.5, 34.8, 34.4, 34.3,
34.0, 33.5, 32.7, 32.5, 32.47, 32.4, 30.8, 30.63, 30.6, 30.55, 30.5, 30.3,
30.2, 29.9, 29.86, 29.6, 29.1, 29.0, 27.0, 26.95, 26.3, 23.4, 23.3, 14.6,
14.5; HRMS-MALDI-FTMS m/z 2196.5792 ([M + Na]+, calcd for
1H NMR (Figure 3B, 600 MHz, p-xylene-d10) δ 10.75 (s, 1 H, minor
diastereomer), 10.1-9.4 (9 s, 5 H, both diastereomers), 9.1-8.8 (3 s,
2 H, both), 8.4-7.0 (14 s, 17 H, both), 6.4-6.3 (m, 4 H, both), 5.9 (s,
1H, major diastereomer), 5.7 (s, 1 H, minor), 2.9-1.1 (m, 154 H),
1.1-0.8 (m, 30 H), 0 to -0.6 (m, 15 H); MALDI MS m/z 2551
([M + Na]+ calcd for C158H228ClN9O15Na 2550).
Cavitand 4c: 13 mg of pale yellow oil, 65%; FTIR (CHCl3, cm-1
)
υ 3407, 3244, 2919, 2851, 1714, 1662, 1600, 1509, 1481, 1405, 1267,
1195; 1H NMR (600 MHz, CDCl3) δ 9.56 (s, 1 H), 9.47 (s, 1 H), 9.37
(s, 1 H), 9.15 (s, 1 H), 8.72 (s, 1 H), 8.24 (d, J ) 9 Hz, 1 H), 8.1 (s,
1 H), 7.73 (s, 1 H), 7.63 (d, J ) 9 Hz, 1 H), 7.51-7.06 (8 s, 12 H),
6.70 (s, 1 H), 6.50 (s, 1 H), 5.86 (t, J ) 8 Hz, 1 H), 5.83, (t, J ) 8 Hz,
1 H), 5.73 (t, J ) 8 Hz, 1 H), 4.25 (t, J ) 8 Hz, 1 H), 3.99 (dd, J )
5 Hz, 5 Hz, 2 H), 2.5-1.0 (m, 167 H), 1.0-0.8 (m, 30 H); MALDI
MS m/z 2552 ([M + Na]+ calcd for C158H227ClN8O16Na 2551).
C138H208N6O14Na 2196.5645, error 6.7 ppm).
(26) For details on high-resolution mass spectrometry, see: (a) Rose,
M. E.; Johnstone, R. A. W. Mass Spectrometry for Chemists and
Biochemists; Cambridge University Press: Cambridge, 1982. (b) Jennings,
K. R.; Dolnikowsi, G. G. Methods in Enzymology; McCloskey, J. A., Ed.;
Academic Press: New York, 1990; p 37 and references therein.
(27) Tunstad, L. M.; Tucker, J. A.; Dalcanale, E.; Weiser, J.; Bryant, J.
A.; Sherman, J. C.; Helgeson, R. C.; Knobler, C. B.; Cram, D. J. J. Org.
Chem. 1989, 54, 1305-1312.
(28) Kazimierczuk, Z.; Dudycz, L.; Stolarski, R.; Shugar, D. Nucleosides,
Nucleotides 1981, 8, 101-117.
(29) Richter, H. J.; Stocker, F. B. J. Org. Chem. 1959, 24, 366-368.
(30) Kleb, K. G.; Siegel, E.; Sasse, K. Angew. Chem., Int. Ed. 1964, 3,
408.
Cavitand 4d: 0.104 g of pale yellow oil, 55%; FTIR (CHCl3, cm-1
)
1
υ 3407, 3244, 2928, 2852, 1719, 1657, 1514, 1486, 1404, 1267; H
NMR (600 MHz, CDCl3) δ 9.62 (s, 1 H), 9.54 (s, 1 H), 9.32 (s, 1 H),
9.13 (s, 1 H), 8.76 (s, 1 H), 8.24 (d, J ) 9 Hz, 1 H), 7.79 (s, 1 H), 7.69
(s, 1 H), 7.60 (d, J ) 9 Hz, 1 H), 7.50-7.04 (8 s, 18 H), 6.85 (s, 1 H),
6.70 (s, 1 H), 5.85 (t, J ) 8 Hz, 1 H), 5.82, (t, J ) 8 Hz, 1 H), 5.71
(t, J ) 8 Hz, 1 H), 5.42 (dd, J ) 8 Hz, 2 Hz, 2 H), 4.27 (t, J ) 8 Hz,
1 H), 2.5-1.0 (m, 152 H), 1.0-0.8 (m, 30 H); MALDI MS m/z 2509
([M + K]+ calcd for C154H217ClN8O16K 2509).
(31) Mohamadi, F.; Richards, N. G.; Guida, W. C.; Liskamp, R.; Lipton,
M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J. Comput. Chem.
1990, 11, 440-467.

4582 J. Am. Chem. Soc., Vol. 122, No. 19, 2000
Renslo et al.
Benzyl Ester 10. An oven-dried 10 mL flask was charged with
hexaamide 9 (0.103 g, 0.047 mmol) and the quinoxaline 14 (17.3 mg,
0.052 mmol). The reactants were dissolved in dry DMF (4 mL) and
the solution treated with triethylamine (39 µL, 0.282 mmol). The
reaction mixture was heated at 85 °C for 48 h and then allowed to
cool to room temperature. The solution was poured into a mixture of
EtOAc (20 mL) and dilute aqueous HCl (20 mL). The layers were
separated and the aqueous phase extracted with two 20 mL portions of
EtOAc. The combined organic extracts were washed with brine, dried
over MgSO4, filtered, and concentrated to provide the crude product.
Column chromatography (0-20% ethyl acetate-hexane) provided 59
(s, 1 H), 8.32 (s, 1 H), 8.21 (s, 2 H), 7.91 (s, 1 H), 7.79 (s, 2 H), 7.78
(s, 1 H), 7.64 (s, 2 H), 7.56 (s, 1 H), 7.30-7.28 (3 s, 4 H), 7.25 (s, 1
H), 7.24 (s, 1 H), 7.20 (s, 1 H), 7.18 (s, 1 H), 6.45 (s, 1 H), 5.82-5.75
(m, 3 H), 5.66 (t, J ) 8 Hz, 1 H), 3.3 (m, 1 H), 3.0 (m, 1 H), 2.5-1.0
(m, 167 H), 1.0-0.8 (m, 30 H); MALDI-FTMS m/z 2514.7289
([M + Na]+ calcd for C157H227N9O15Na13C 2514.7208, error 3.2 ppm).
Dinitrocavitand 16. A solution of hexaamide 9 (0.135 g, 0.062
mmol), 1,2-difluoro-4,5-dinitrobenzene (0.025 g 0.124 mmol), and
triethylamine (0.070 mL; 0.5 mmol) in anhydrous DMF (18 mL) was
stirred for 14 h at 70 °C. The reaction mixture was then poured into
dilute aqueous HCl (pH ∼1, 150 mL) and the resulting precipitate was
filtered and rinsed with water. After being dried in vacuo, the solids
were chromatographed (10% EtOAc-hexane). The product was
obtained as a yellow solid after trituration with MeOH (124 mg, 0.053
mg (55%) of cavitand 10 as a pale yellow oil: FTIR (CHCl3, cm-1
)
υ 3416, 3244, 2919, 2852, 1662, 1600, 1514, 1481, 1400, 1267, 1195;
1H NMR (600 MHz, CDCl3) δ 9.53 (s, 1 H), 9.43 (s, 1 H), 9.18 (s, 1
H), 8.52 (s, 2 H), 8.37 (d, J ) 8 Hz, 1 H), 8.3 (br s, 1 H), 8.02 (d, J
) 8 Hz, 1 H), 7.78 (s, 1 H), 7.74 (s, 2 H), 7.72 (s, 1 H), 7.66 (s, 1 H),
7.45-7.37 (m, 7 H), 7.28 (s, 1 H), 7.26 (s, 2 H), 7.24 (s, 1 H), 7.20 (s,
2 H), 7.19 (s, 2 H), 5.80 (t, J ) 8 Hz, 1 H), 5.75 (t, J ) 8 Hz, 2 H),
5.50 (t, J ) 8 Hz, 1 H), 5.42 (d, J ) 12 Hz, 1 H), 5.33 (d, J ) 12 Hz,
1 H), 2.5-1.0 (m, 152 H), 1.0-0.8 (m, 30 H); MALDI MS m/z 2556
([M + Na]+ calcd for C154H216N8O16Na 2556.6).
1
mmol, 85%): mp >250 °C; H NMR (600 MHz, benzene-d6, 295 K)
δ 9.79 (s, 2 H), 9.74 (s, 2 H), 8.53 (br s, 2 H), 7.75 (s, 4 H), 7.67 (s,
4 H), 7.56 (br s, 2 H), 7.45 (br s, 2 H), 7.34 (br s, 2 H), 7.28 (s, 2 H),
6.38 (t, J ) 8 Hz, 2 H), 6.15 (t, J ) 8 Hz, 1 H), 5.93 (t, J ) 8 Hz, 1
H), 2.56-2.37 (m, 20 H), 1.87-1.83 (m, 12 H), 1.55-1.28 (m, 120
H), 0.97-0.89 (m, 30 H); MALDI-MS m/z 2361 ([M + Na]+, calcd
for C144H208N8O18Na 2362).
Pentafluorophenyl Ester 12. To a solution of benzyl ester 10 (0.059
g, 0.025 mmol) in EtOAc (4 mL) was added 0.059 g of Pd/C. The
suspension was stirred under a H2 atmosphere for 2 h and then filtered
through Celite twice. The filtrate was concentrated to give 0.046 g of
the acid 11 which was converted directly to 12 without further
purification. The crude acid 11 (0.046 g, 0.020 mmol) was dissolved
in 3 mL of THF and treated with EDC-MeI (0.012 g, 0.040 mmol),
pentafluorophenol (4.0 mg, 0.022 mmol), and (dimethylamino)pyridine
(0.6 mg, 0.005 mmol). The suspension was sonicated briefly and then
stirred in the dark for 20 h. The reaction mixture was then passed
through a small column eluting with THF. The fractions containing
product were collected and concentrated to provide 0.050 g (78%
Diamino Cavitand 17. Dinitrocavitand 16 (0.124 g, 0.053 mmol)
was dissolved in toluene (40 mL). To this solution was added a catalytic
amount of Raney Nickel that had been previously washed with MeOH
(2 × 5 mL). The resulting suspension was stirred for 3 h at 40 °C
under a H2 atmosphere. After cooling, the Raney Nickel was decanted
and the supernatant liquid was pipetted off, rinsing the residual nickel
with MeOH (2 × 10 mL). The filtrates were combined and evaporated
under vacuum to give a yellowish solid (0.079 g, 0.0347 mmol, 65%)
that was used in the next step without further purification: 1H NMR
(600 MHz, benzene-d6, 295 K) δ 10.00 (s, 2 H), 9.52 (s, 2 H), 9.05 (s,
2 H), 7.77 (s, 2 H), 7.75 (s, 2 H), 7.59 (s, 4 H), 7.50 (s, 2 H), 7.43 (s,
2 H), 7.36 (s, 2 H), 6.95 (s, 2 H), 6.41 (t, J ) 8 Hz, 1 H), 6.34 (t, J )
8 Hz, 1 H), 6.27 (t, J ) 8 Hz, 2 H), 3.28 (br, 4 H), 2.46-2.22 (m, 20
H), 1.88-1.65 (m, 12 H), 1.54-1.25 (m, 120 H), 0.98-0.92 (m, 30
H); MALDI-MS m/z 2279 ([M + H]+, calcd for C144H212N8O14H
2278.6).
overall) of active ester 12 as a yellow film: FTIR (CHCl3, cm-1
)
υ 3416, 3244, 2919, 2852, 1757, 1657, 1519, 1481, 1400, 1338, 1190;
1H NMR (600 MHz, CDCl3, 295 K) δ 9.44 (s, 1 H), 9.36 (s, 1 H),
9.27 (s, 1 H), 9.05 (s, 1 H), 8.9 (s, 1 H), 8.78 (s, 1 H), 8.44 (s, 1 H),
8.20 (d, J ) 8 Hz, 1 H), 8.1-7.2 (m, 15 H), 5.8-5.75 (m, 3 H), 5.45
(br t, 1 H), 2.6-1.0 (m, 152 H), 1.0-0.8 (m, 30 H); 19F NMR (565
MHz, CDCl3) δ -151.8 (d, J ) 20 Hz, 2 F), -156.5 (t, J ) 20 Hz,
1 F), -161.5 (t, J ) 20 Hz, 2 F).
Cavitand 5. To a stirred solution of diamine 17 (0.024 g, 0.0105
mmol) and 1,2-diketopyracene (7 mg, 0.032 mmol) in THF (5 mL)
was added glacial acetic acid (30 µL) and the resulting mixture was
refluxed for 16 h.32 After cooling, the volatiles were removed under
vacuum and the residue was purified by preparative thin-layer chro-
matography (silica gel, 0.5 mm) eluting with 20% EtOAc-hexane. The
product was obtained as a yellow solid (13 mg, 0.0053 mmol, 50%)
Self-Complementary Cavitand 3a. A 10 mL oven-dried flask was
charged with active ester 12 (13.3 mg, 0.0052 mmol) and 1-aminoada-
mantane (1.0 mg, 0.0066 mmol) in DMF (0.5 mL). The solution was
treated with triethylamine (2.0 mg, 3.0 µL, 0.021 mmol) and stirred at
room temperature for 16 h. The solution was then poured into a mixture
of EtOAc (10 mL) and dilute HCl (10 mL). The layers were separated
and the aqueous phase extracted with two 10 mL portions of EtOAc.
The combined organic extracts were washed with brine, dried over Na2-
SO4, filtered, and concentrated to provide 0.013 g of the crude product.
Column chromatography (0-20% EtOAc-hexane) provided 10.2 mg
after trituration with MeOH: mp >250 °C; FTIR (CDCl3, cm-1
)
υ 3238, 3034, 2957, 2928, 2855, 1968, 1731, 1658, 1600, 1511, 1482,
1
1436, 1273, 1189, 1058, 850; H NMR (600 MHz, CDCl3, 340 K)
δ 9.25 (br, 4 H), 8.46 (br, 4 H), 7.68 (br, 4 H), 7.47 (br, 6 H), 7.32 (s,
2 H), 7.29 (s, 2 H), 7.22 (s, 2 H), 5.84-5.79 (m, 3 H), 5.64 (br, 1 H),
3.67 (dd, J ) 17.3, 22.2 Hz, 4 H), 2.50-2.02 (m, 20 H), 1.80-1.25
(m, 132 H), 0.93-0.84 (m, 30 H); 13C NMR (150 MHz, CDCl3, 295
K) δ 172.88, 172.45, 172.40, 163.10, 162.74, 155.04, 154.76, 154.46,
154.21, 153.56, 149.66, 149.40, 136.67, 135.90, 135.50, 135.35, 134.95,
128.97, 127.94, 127.79, 127.62, 124.35, 123.72, 123.48, 122.05, 121.21,
120.82, 116.22, 37.22, 33.79, 33.35, 33.30, 32.83, 32.66, 32.25, 31.95,
31.72, 31.64, 29.83, 29.82, 29.78, 29.76, 29.73, 29.42, 29.36, 29.25,
29.08, 28.99, 28.05, 26.08, 25.80, 22.70, 22.55, 14.13, 14.06, 13.99;
MALDI-MS m/z 2453 ([M + H]+, calcd for C158H216N8O14H 2452.5);
HRMS-MALDI-FTMS m/z 2473.6491 ([M + Na]+, calcd for
(79%) of the cavitand 3a as a pale yellow oil: FTIR (CHCl3, cm-1
)
1
υ 3684, 3244, 2919, 2852, 1657, 1600, 1514, 1481, 1405, 1229; H
NMR (600 MHz, CDCl3) δ 9.65 (s, 1 H), 9.54 (s, 1 H), 9.44 (s, 1 H),
9.12 (s, 1 H), 8.68 (s, 1 H), 8.47 (d, J ) 8 Hz, 1 H), 8.15 (s, 1 H), 8.04
(s, 1 H), 7.95 (s, 1 H), 7.82 (s, 1 H), 7.76-7.73 (3 s, 3 H), 7.60 (s, 1
H), 7.39 (d, J ) 8 Hz, 1 H), 7.21-7.10 (4 s, 8 H), 5.90 (s, 1 H), 5.81
(t, J ) 8 Hz, 1 H), 5.73 (m, 2 H), 5.59 (t, J ) 8 Hz, 1 H), 2.70 (br t,
2 H), 2.5-1.0 (m, 165 H), 1.0-0.8 (m, 30 H); MALDI-FTMS m/z
2500.7060 ([M + Na]+ calcd for C156H225N9O15Na13C 2500.7051, error
0.4 ppm).
C
157H216N8O14Na13C 2473.6367, error 5.0 ppm).
Acknowledgment. We are grateful to the Skaggs Founda-
Self-Complementary Cavitand 3b. A 10 mL oven-dried flask was
charged with active ester 12 (12 mg, 0.0048 mmol) and 1-amino-
methyladamantane (1.0 mg, 0.0062 mmol) in CH2Cl2 (0.5 mL). The
solution was treated with triethylamine (1.9 mg, 2.7 µL, 0.019 mmol)
and stirred at room temperature for 24 h. The reaction mixture was
then concentrated and the residue purified by column chromatography
(0-20% EtOAc-hexane) to provide 8.2 mg (70%) of the cavitand 3b
as a pale yellow oil: FTIR (CHCl3, cm-1) υ 3684, 3234, 2919, 2852,
1657, 1600, 1510, 1481, 1405, 1338, 1271; 1H NMR (600 MHz, CDCl3,
320K) δ 9.40 (s, 1 H), 9.31 (s, 1 H), 9.20 (s, 1 H), 8.90 (s, 1 H), 8.40
tion, the NASA Astrobiology Institute, and the National
Institutes of Health for support of this work. We also thank
Prof. T. Haino for experimental advice.
JA993922E
(32) This synthetic approach was employed in the synthesis of deepened
cavitands: Tucci, F. C.; Rudkevich, D. M.; Rebek, J., Jr. J. Org. Chem.
1999, 64, 4555-4559.