Synthesis and assembly of new molecular hosts: Solvation and the energetics of encapsulation

Article abstract of DOI:10.1021/ja962991f

Experimental details are given for the preparation of 'softballs', large self-complementary molecules capable of assembly into pseudo-spherical capsules. Evidence is presented for their existence as hydrogen bonded dimers in organic solvents, and binding affinities for the reversible encapsulation of smaller molecules of suitable size and shape are given. Studies at various temperatures result in calculated enthalpies and entropies of encapsulation that are positive; accordingly, the process is entropy driven. It is proposed that the hosts in their resting states contain two molecules of solvent such as benzene, and the encapsulation of a single large guest-the hostage-liberates the two solvents. The resulting increase in the number of free molecules gives rise to the increase in entropy observed for the exchange process. Experiments involving solvent mixtures are consistent with this rationale. Calculation of the capsule's interior volume and molecular dynamics simulations support the experimental observations, and hint at unexpected phenomena dealing with the occupancy factors of these systems.

Full text of DOI:10.1021/ja962991f

J. Am. Chem. Soc. 1997, 119, 77-85
77
Synthesis and Assembly of New Molecular Hosts: Solvation
and the Energetics of Encapsulation
R. Meissner, X. Garcias, S. Mecozzi, and J. Rebek, Jr.*
Contribution from the Department of Chemistry, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139, and The Skaggs Institute for Chemical Biology, The Scripps
Research Institute, 10550 North Torrey Pines Road, La Jolla, California 92037
ReceiVed August 26, 1996X
Abstract: Experimental details are given for the preparation of “softballs”, large self-complementary molecules capable
of assembly into pseudo-spherical capsules. Evidence is presented for their existence as hydrogen bonded dimers in
organic solvents, and binding affinities for the reversible encapsulation of smaller molecules of suitable size and
shape are given. Studies at various temperatures result in calculated enthalpies and entropies of encapsulation that
are positive; accordingly, the process is entropy driven. It is proposed that the hosts in their resting states contain
two molecules of solvent such as benzene, and the encapsulation of a single large guest-the hostage-liberates the
two solvents. The resulting increase in the number of free molecules gives rise to the increase in entropy observed
for the exchange process. Experiments involving solvent mixtures are consistent with this rationale. Calculation of
the capsule’s interior volume and molecular dynamics simulations support the experimental observations, and hint
at unexpected phenomena dealing with the occupancy factors of these systems.
Introduction
two identical self-complementary subunits that fit together as a
dimer, in a way reminiscient of the way the two pieces of a
tennis ball fit together into a hollow sphere. These dimers
maintain the broad attributes of the carcerands and cryptophanes,
but they are formed reversibly, and their dynamic qualities
impart special merits. They form and dissipate on the time scale
that varies from milliseconds to hours, intervals long enough
for many types of interactions, even chemical reactions, to occur
within them. We term the process encapsulation and describe
here some unexpected behavior that is expressed when the larger
host capsules take and exchange their hostages.11
Elsewhere we have dwelt on the virtues of hydrogen bonding
as a vehicle for molecular recognition,12 replication,13 and self-
assembly;14 we have discussed in detail the application of the
glycoluril module in the latter context: its advantages and
limitations, the considerations of curvature, spacing, geometry,
and solubility;15 the failed early prototypes and the subsequent
successes,16 the spectroscopic characteristics of homodimeric
and heterodimeric capsules in solution and the solid state,15,17
and their ability to encapsulate smaller molecules.18 We cannot
repeat them here, rather, we proffer the picture of 1 and its
dimeric form as a spur to the reader’s attention.
“Molecules within molecules”1 is a phrase that characterizes
an unusual, perhaps unique arrangement of matter. Introduced
by Cram2 and Collet3 nearly a decade ago, these systems
featured covalently assembled structuresscarcerands and
cryptophanesswith cavities capable of completely surrounding
smaller molecules. The guest species could be found inside
under two conditions: when they act as templates during the
synthesis of the hosts,4,5 or when they enter through openings
created by structural distortions in the host molecules.6 High
activation barriers are generally associated with the latter
process, a feature that permits the isolation of reactive molecules
such as cyclobutadiene within the cavities.7 A new form of
stereoisomerism that arises from the restricted motion of
encarcerated species has recently been discovered.8
With the premise that further new phenomena await the
exploration of these systems, we have used weak intermolecular
forces rather than covalent bonds as a means of assembling
molecule within molecule complexes.9,10 Our systems involve
* Address correspondence to this author at The Scripps Research Institute.
X Abstract published in AdVance ACS Abstracts, December 15, 1996.
(1) This phrase first appears as the title of a lecture by D. J. Cram at the
C. David Gutsche Symposium, Washington University, St. Louis, Missouri,
May 5, 1990; for examples, see: Cram, D. J. Nature 1992, 356, 29-36.
(2) Cram, D. J.; Choi, H-J.; Bryant, J. A.; Knobler; C. B. J. Am. Chem.
Soc. 1992, 114, 7748-7765. Cram, D. J.; Blanda, M. T.; Paek, K.; Knobler,
C. B. J. Am. Chem. Soc. 1992, 114, 7765-7773.
The questionshow to make a sizable capsulesfinds an
answer in the gentle curvature and reliable rigidity of the
structure’s 13 fused and 1 bridged ring systems, the four basic
carbonyls of the centerpiece and their acidic glycoluril hydrogen
(3) Collet, A. Tetrahedron 1987, 43, 5725. Canceill, J.; Cesario, M.;
Collet, A.; Guilhem, J.; Lacombe, L.; Lozach, B.; Pascard, C. Angew. Chem.,
Int. Ed. Engl. 1989, 28, 1246-1248.
(10) Shimizu, K. D.; Rebek, J., Jr. Proc. Natl. Acad. of Sci. U.S.A. 1995,
92, 12403-12407.
(11) Meissner, R. S.; de Mendoza, J.; Rebek, J., Jr. Science 1995, 270,
1485-1488.
(4) Chapman, R. G.; Chopra, N.; Cochien, E. D.; Sherman J. C. J. Am.
Chem. Soc. 1994, 116, 369-370.
(12) Rebek, J., Jr. Angew. Chem., Int. Ed. Engl. 1990, 29, 245-255.
(13) Wintner, E.; Rebek, J., Jr. Acta Chem. Scand. 1996, 50, 469-485.
(14) Rebek, J. Jr. Chem. Soc. ReV. In press.
(5) Sherman J. C.; Cram, D. J. J. Am. Chem. Soc. 1989, 111, 4527-
4528.
(6) Quan, M. L. C.; Knobler, C. B.; Cram, D. J. J. Chem. Soc., Chem.
Commun. 1991, 660-662. Quan, M. L. C.; Cram, D. J. J. Am. Chem. Soc.
1991, 113, 2754-2755.
(15) Valde´s, C.; Spitz, U. P.; Toledo, L.; Kubik, S.; Rebek, J., Jr. J. Am.
Chem. Soc. 1995, 117, 12733-12745.
(16) Rebek, J., Jr. Acta Chem. Scand. 1996, 50, 707-716.
(17) Toledo, L.; Valdes, C.; Rebek, J., Jr. Chem. Eur. J. 1996, 2, 989-
991.
(7) Cram, D. J.; Tanner, M. E. Angew. Chem., Int. Ed. Engl. 1991, 30,
1024-1027.
(8) Timmerman, P.; Verboom, W.; VanVeggel, F. C. J. M.; Duynhoven,
J. P. M.; Reinhoudt, D. N. Angew. Chem., Int. Ed. Engl. 1994, 33, 2345-
2348.
(18) Branda, N.; Wyler, R.; Rebek, J., Jr. Science 1994, 263, 1267-
1268. Branda, N.; Grotzfeld, R. M.; Valde´s, C.; Rebek, J., Jr. J. Am. Chem.
Soc. 1995, 117, 85-88. For methane complexation inside a cryptophane
see: Garel, L.; Dutasta, J.-P.; Collet, A. Angew. Chem., Int. Ed. Engl. 1993,
32, 1169.
(9) Wyler, R.; de Mendoza, J.; Rebek, J., Jr. Angew. Chem., Int. Ed.
Engl. 1993, 32, 1699-1701.
S0002-7863(96)02991-5 CCC: $14.00 © 1997 American Chemical Society

78 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Meissner et al.
deprotected and activated with oxalyl chloride to give the
centerpiece module, the tetraacid chloride 8.
Two versions of the softball were prepared. The one (1b)
derived from diphenylglycouril was prepared as in Figure 4,
followed by reaction of 22 with 8. However, it showed limited
solubility in organic media. The softball (1a) derived from the
isopentyl ester of dihydroxy tartrate, as described earlier,15
proved tractable in several solvents.
Characterization as Dimeric Structures
The proton NMR signals of molecule 1b are sharp and well-
defined in benzene-d6, and considerable downfield shifts (δ >
8 ppm) were observed for the N-H resonances. This is
characteristic of an ordered, extensively hydrogen bonded
system (Figure 5A). However, in CDCl3, the molecule dis-
solved with difficulty and formed a gel-like phase after a few
minutes. The spectrum showed broad unresolved peaks (Figure
5B) and the suspension in the NMR tube could be turned upside
down without loss of its contents. Evidently, in this solvent an
aggregate of relatively low order was formed. The very same
sample in a solvent such as benzene or toluene which, as we
shall see, are good guests for the interior showed very different
NMR spectra: the spectra of a well-filled dimeric capsule.
The more soluble derivative 1a showed a spectrum similar
to that of 1b in benzene-d6, but in p-xylene-d10 produced broad
signals in the NMR spectrum (Figure 6A). On the addition of
suitable guests, e.g. excess [2.2] paracyclophane (27), sharp and
clearly resolved spectra were obtained (Figure 6B).
Figure 1. Structure of monomeric 1 and its energy minimized
(MacroModel20 3.5X with MM2 force field) dimeric form. The
peripheral R groups and some hydrogens have been omitted for viewing
clarity.
complements on the molecule’s ends, and the well-filled surface.
The absence of large holes closes the avenues of escape of
hostages from the dimeric, capsular host. Relative to the size
of the original tennis ball, the structure becomes a notional
“softball”.19
Synthesis
Other sizable additives such as adamantane or ferrocene
derivatives also led to the characteristic resonances of the
capsule. Specifically, 1-adamantanecarboxylic acid (23a), the
1,3-dicarboxylic acid 24, 1-adamantaneamine (23b), and the
ferrocene acid 25 were able to induce this change in the
spectrum, and even the nonpolar tetramethyladamantane 26,
when present in large excess, was able to generate the capsular
form (the spectra are not shown here but appear in ref 11). In
short, these guests autoencapsulate themselves, and in doing
so, they become hostages, unable to leave until they are replaced.
The autoencapsulation process results in the breakup of the
rather loose aggregates and leads to the emergence of ordered,
occupied capsular forms.
The synthesis of the softball takes advantage of the same
starting material used for the tennis ballsthe tetrabromo durene
2 (Figure 2). This was used first to dialkylate a BOC-protected
hydrazine derivative 3 to give structure 4, then the remaining
two halide sites were used to alkylate the appropriate glycouril
5 (R ) CO2-isopentyl). The resulting structure 6 represents
the ends of the self-complementary pieces. Removal of the
BOC groups gave the hydrazine salt 7, the unstable free base
of which was acylated by the centerpiece, the tetraacid chloride
8, in good yield. The two other isomers were isolated by
chromatography. These were identified by the (spectroscopi-
cally) different ends featured by the S-shaped isomer and the
failure of the W-shaped isomer to assemble.
The ability of these guests to induce the capsular form at
room temperature varied, and the values within this limited set
are recorded in Table 1. Both adamantane and the ferrocene-
carboxylic acids were suitable guests. For the adamantanes,
signals appearing upfield of 0 ppm indicated that the guests
were in an environment shielded by aromatic ring currents. For
the ferrocene acid, the upfield shifts were comparable: 3.5 to
4.5 ppm free versus 2 to 2.5 ppm encapsulated. While it is
barely conceivable that these changes in chemical shift involve
some specific interactions between the solutes and the mono-
meric forms of 1 in an aromatic solvent, the facts leave little
doubt that true encapsulation occurs.
Mere integration of the spectra established the stoichiometries
of encapsuled guests in the softball. The apparent association
constants are derived from the equation below and are only
approximate, since the total concentration of the aggregate was
involved in the calculations. Accordingly, the values are
The synthesis of the centerpiece 8 (Figure 3) was somewhat
more convoluted but followed a strategy developed earlier for
such structures.21 It involved a protected acetylene dicarboxylic
acid 9 as the repeating module. Diels-Alder reaction with
furan, then reduction and ether cleavage to afford the dihydro-
phthalate 10, was followed by the second reaction with the
acetylene derivative. The bicylic framework 11 was then
(19) Garcias, X.; Rebek, J., Jr. Angew. Chem., Int. Ed. Engl. 1996, 35,
1225-1228.
(20) MacroModel V:5.5: Mohamadi, F.; Richards, N. G. J.; Guidia, W.
C.; Liskamp, R.; Caulfield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J.
Comput. Chem. 1990, 11, 440-467. Although appropriate force field
parameters for all the functional groups in molecule 1 do not exist, we
believe that the use of the force field Amber* of MacroModel 5.5 allows
a reasonable approximation for the calculation and comparison of volumes.
(21) Beerli, R.; Rebek, J., Jr. Tetrahedron Lett. 1995, 36, 1813-1816.
(22) C. Wilcox, personal communication. We thank Prof. Wilcox for
providing this information prior to publication.

Synthesis and Assembly of New Molecular Hosts
J. Am. Chem. Soc., Vol. 119, No. 1, 1997 79
t
Figure 2. Synthesis of monomer 1a. Abbreviations: BOC ) tert-butoxycarbonyl, DMF ) N,N-dimethylformamide, BuOK ) potassium tert-
butoxide, DMSO ) dimethyl sulfoxide, NEt3 ) triethylamine.
Figure 3. Synthesis of the central tetraacyl chloride. Abbreviations: TFA ) trifluoroacetic acid, PMB ) p-methoxybenzyl.
denoted as apparent association constants, Ka(app), because of
the simplification of the several dynamic processes present into
one represented by eq 1.
to 60 °C, 59% is occupied. Encapsulation increases with
increasing temperature. Figure 8 shows the van’t Hoff plot
for the encapsulation, the most important feature of which is
the uncharacteristic downward slope. The usual treatment of
the data gives ∆S ) 35 ( 1.7 eu and ∆H ) 7.3 ( 0.4 kcal. As
both enthalpy and entropy are positive for the process, it must
be an entropy-driven process.
1b(aggregae) + guest h 1b‚guest‚1b
[1b‚guest‚1b]
(1)
Ka(app)
)
[1b(aggregate)][guest]
How can the phenomena be interpreted? The hydrophobic
effect is the classical example, but in studies in organic solvents,
this type of behavior is rarely seen. But why not? As two
molecules come together to form a noncovalent complex, the
liberation of solvent is a universal feature of molecular recogni-
tion phenomena and generally more than one solvent molecule
is released. For example, the “activated water”23 molecules
inside cyclodextrins are displaced by organic guests, but
conventional thermodynamics (enthalpy driven and entropy
resisted) are, nonetheless, observed. Wilcox22 has observed this
behavior in water using macrocyclic hosts and organic guests.
In solvents that gave indications that the capsular form was
the dominant species, it was possible to do more detailed studies.
These were executed at various temperatures and some unusual
results were encountered. For the specific case of 1-adaman-
tanecarboxamide (23a) in toluene-d8, the series of spectra are
shown in Figure 7.
The new signal for the encapsulated adamantane derivative
appears upfield from 0 ppm and two different softball resonances
can be observed between 7.8 and 8 ppm. Their ratios change
as the temperature changes. Specifically, with approximately
3 equiv of 23a at 45 °C, 47% of the dimeric hosts were occupied
by the hostage (Figure 7B) but when the temperature is raised
(23) Saenger, W. Angew. Chem., Int. Ed. Engl. 1980, 19, 344-362 and
references therein.

80 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Meissner et al.
Figure 4. Synthesis of components for monomer 1b.
Figure 5. 1H NMR spectra of 1b. (A) Benzene-d6 solvent; signals
between 7.2 and 7.0 ppm and between 2.2 and 0 ppm are from the
solvent and its impurities. (B) Chloroform-d solvent; the single peak
at 7.26 ppm is from the solvent.
Figure 6. 1H NMR spectra of 1a. (A) Compound 1a in 8.84 mM
p-xylene-d10 solvent at 25 C; signals between 0.1 and 0.3 and 7.1 and
6.6 ppm, and 2.3 and 1.8 ppm are from the solvent and its impurities.
(B) As in part A with 19.2 mM [2.2]paracyclophane (27) added. The
signals of the included and free 27 are designated.
The velcrands described by Cram24 show unprecedented (and
unpredictable) ranges of ∆H from +6 to -8 kcal/mol and ∆S
from-6 to +40 eu in organic media. There must be something
special about the guest solvent in these capsules. Compared to
the loosely bound surface solvents that are featured in most
cases, the encapsulated solvents must enjoy less freedom.
Indeed, much has been written about the compensating effects
of entropy and enthalpy in complex formation,25-27 but for the
softball the inclusion of guests involves an enthalpic cost
compensated by a larger entropic gain. This behavior is
reminiscent of the classical hydrophobic effect, wherein release
of bound water to the bulk solvent compensates for the
association of solutes. In organic media such behavior is rarely
encountered and generally unpredictable.24,28
Table 1. Apparent Association Constants Derived from Eq 1)
b
guest
Ka(app) (M-1
)
1,3,5,7-tetramethyladamantane (26)
adamantaneamine (23b)
1-adamantanecarboxamide (23c)
1-adamantanecarboxylic acid (23a)
1,3-adamantanedicarboxylic acid (24)
1-ferrocenecarboxylic acid (25)
6.7
190
310
780
a
280
a Value not calculated due to the insolubility of the free guest.
(24) Cram, D. J.; Choi, H-J.; Bryant, J. A.; Knobler, C. B. J. Am. Chem.
Soc. 1992, 114, 7748-7765. Cram, D. J.; Blanda, M. T.; Paek, K.; Knobler,
C. B. J. Am. Chem. Soc 1992, 114, 7765-7773.
Schematically, Figure 9 shows the case if two molecules of
solvent occupy the interior of the softball in its resting state in
benzene. The left hand side of the equation involves two
particles but the liberation of the encapsulated solvent molecules
(25) For a recent discussion with leading references, see: Dunitz, J. D.
Chem. Biol. 1995, 2, 709-712.
(26) Peterson, B. R.; Wallimann, P.; Carcanague, D. R.; Diederich, F.
Tetrahedron 1995, 51, 401-421.
(27) Searle, M. S.; Westwell, M. S.; Williams, D. H. J. Chem. Soc.,
Perkin Trans. 2 1995, 141-151.
(28) Canceill, J.; Cesario, M.; Collet, A.; Guilhem, J.; Lacombe, L.;
Lozach, B.; Pascard, C. Angew. Chem., Int. Ed. Engl. 1989, 28, 1246-
1248.

Synthesis and Assembly of New Molecular Hosts
J. Am. Chem. Soc., Vol. 119, No. 1, 1997 81
Figure 10.
A simple model for the encapsulation of two molecules of
solvent can be constructed by considering that two solvated
hemispheres come together to form a well-filled capsule. The
microscopic reverse of the process, the separation of the fully-
loaded softball into its halves, is also straightforward: as the
seam of hydrogen bonds unravels, the two solvents need only
to be positioned in a way that one solvent remains in the concave
cavity of each subunit (Figure 10). This may be likened to the
ease of evenly dividing a peach when the pit has already
separated in the plane of the knife cut: little vacuum is created
as the halves are pulled apart. It would appear, then, that two
solvent guests are encapsulated smoothly since the softball is,
after all, a dimer. For a larger solvent as xylene, the
hemispheres should still be well solvated but as they come
together the capsule is destabilized. Calculations reported
elsewhere indicate that a constellation of two xylenes can be
best accommodated inside when they are forced to a face-to-
face distance of <3.3 Å.
Figure 7. 1H NMR spectra of 1a in p-xylene-d10 solvent with 23a as
a function of temperature. Solvent resonances occur between 2.0 and
2.2 ppm and between 6.8 and 7.2 ppm. In spectra B through E,
resonances for the encapsulated portion of 23a occur between 0.2 and
-0.4 ppm, and those for chloroform as a singlet which moves between
6.0 and 6.2 ppm. The resonance of the Ar-H protons are labeled with
(T) for the host dimer assembly containing toluene-d8 and with (AA)
for that containing 23a; these peaks are the best resolved for the two
species, and were integrated for measurement of Keq. The resonances
for the encapsulated guest are labeled (E). (A) Toluene-d8 solvent; 9.9
mM in 1a, temperature ) 25 °C. (B) 7.3 mM in 1a, 20 mM in
1-adamantanecarboxylic acid (23a), temperature ) 25 °C. (C) as in
part B, temperature ) 35 °C; (D) as in part B, temperature ) 45 °C;
(E) as in part B, temperature ) 60 °C.
How only one large guest becomes encapsulated is somewhat
more complicated, but Figure10 provides a schematic proposal
which is scarcely more than a restatement of the facts: the guest
sequentially replaces solvent within each hemisphere.
We were able to find some confirmation, albeit indirect, of
the occupancy of the softball by benzene through a two solvent
experiment. In either deuterated benzene or deuterated fluo-
robenzene, sharply defined NMR signals are observed for the
softball with the characteristic downfield shifts that indicate its
assembly (Figure 11). When a mixture of the two solvents was
used, the spectrum becomes somewhat broadened, but cooling
the sample led to an increased resolution in which three species
of the capsule can be observed. In changing the relative
amounts of the two solvents, the corresponding redistribution
of the three signals occurs. If indeed two molecules of solvent
occupy each capsule, then the predicted result would be three
different species.
The positive enthalpy observed for encapsulation does not
give a comparably simple or satisfying picture. Adamantane-
carboxylic acid alone in a nonpolar solvent such as toulene is
expected to be largely in the form of a hydrogen bonded dimer.
In the presence of the softball, groups rich in hydrogen bond
donors and acceptors abound and the carboxylic acid must be
hydrogen bonded to the outer surface. We cannot predict how
the enthalpy of those hydrogen bonds, many of which are of
the bifurcated sort, are changed on the adamantane’s moving
to the interior. Nor can we predict whether the van der Waal’s
contacts between encapsulated solvents and the concave surface
of the capsule are more or less favorable than those interactions
of the hydrocarbon portion of the adamantane guest. We know
that aromatics such as benzene or p-xylene are less polarizable,
along the axis going through the centroid of the molecule, than
Figure 8. van’t Hoff plot of data from Figure 7.
Figure 9.
leads to three particles on the right hand side. For the host, the
process involves a hostage upgrade. Any solvent on the surface
of the guest is also freed to the medium as encapsulation occurs.
This should be an entropy driven process.

82 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Meissner et al.
Table 2. Molecular Volumes and Occupancy Factors for Some
Guests in the Softball 1b
guest
23a
24
23b
26
27
23c
benzene
p-xylene
molecular vol (Å3)
occupancy factor
155
185
137
181
188
157
77
0.52
0.62
0.46
0.60
0.63
0.52
0.26
0.35
105
conformation of the softball in the gas phase by using the force
field Amber* of MacroModel 5.5.20 Next a molecular dynamics
simulation33 in chloroform was run using the same force field.
The resulting averaged structure was then analyzed with the
program GRASP.34 A molecular surface was built using a
spherical probe of 1.4 Å radius, in such a way that open surface
areas were artificially closed to allow measurement of the
internal volume.35 We calculated an average internal volume
of 300 Å3, and used the same program to calculate the guest
volumes (Table 2).
It can be seen that the occupancy factor36 of the relevant
guests ranges from 0.26 for one molecule of benzene to 0.63
for the biggest guest, [2.2] paracyclophane. The latter value
corresponds to the occupancy factor of a typical solid. For
benzene, two molecules inside correspond to a concentration
of about 10 M, approximately the value in pure benzene (11
M). Alternatively stated, the occupancy, 0.52, for two benzenes
inside is also close to the value (0.58) calculated for pure
benzene. It should be possible to select solute combinations
merely on the basis of occupancy for encapsulation in dilute
solutions, and pseudoconcentrate them to high effective mo-
larities within these capsules. For the case at hand, a combina-
tion of, say, C4 with C8 should better fill the capsule at ∼5 M
each inside than two molecules each of C4 or C8.
In order to evaluate how the size and shape of the cavity is
affected by a bound guest we ran a similar molecular dynamics
simulation37 on the complex between molecule 1b and the diacid
24. We chose this large guest in order to maximize the effect
of a massive, asymmetric, and polar guest on the cavity. One
can imagine this guest molecule, inside the softball, as a lumpy
sphere that frequently collides with the cavity walls. After the
simulation, the guest was removed and the volume of the cavity
was recalculated. It amounted to 290 Å3. Apparently, the
presence of this guest inside the ball does not substantially
modify its internal volume, but some distortion of the cavity
occurs. It is clear that among guests 23-27, only one guest
molecule at a time can be encapsulated. What can be said about
smaller guests such as benzene and p-xylene? We ran molecular
dynamics simulations37 for these guests, and we encountered
two completely different situations. In the case of benzene,
two molecules can fit into the cavity and the simulation gave
Figure 11. 1H NMR spectra of compound 1a: encapsulation of
benzene-d6 and fluorobenzene-d5. The resonances of the N-H protons
for the dimeric host assemblies occur between 8.6 and 8.8 ppm; these
peaks are among the best resolved, clearly showing that three species
are present. Incompletely deuterated benzene-d6 resonances occur
between 6.8 and 7.2 ppm. (A) Benzene-d6 solvent (400 µL); 15.8 mM
in 1a, temperature ) 25 °C; (B) benzene-d6 solvent (400 µL) +
fluorobenzene-d5 (100 µL), 12.6 mM in 1a, temperature ) 25 °C; (C)
as in part B at temperature ) 0 °C; (D) benzene-d6 solvent (400 µL)
+ fluorobenzene-d5 (200 µL); 10.5 mM in 1a, temperature ) 0 °C.
saturated hydrocarbons such as cyclohexane.29,30 However, it
is uncertain how these molecules can fit into the cavity and
there is no reliable way of predicting what the enthalpy of the
overall process depicted in the equation should be.
What is the relationship of the two encapsulated species to
each other? Confined at such close quarters, are their rotational
motions restricted? A clue is given by the behavior of
asymmetric guests within. To date, we have yet to observe a
loss of symmetry in the capsule’s NMR spectra even when, for
example, asymmetric guests or two different guests are inside.
Accordingly, the guests must be rotating rapidly within. For
the two guests encapsulated in the softball, the motion might
be similar to a binary star system, a microconstellation inside a
cavity that defines their universe. This is not the case in the
covalent systems of Reinhoudt,8 who discovered a new form
of stereoisomerism caused by restricted rotation of a guest inside
a carcerand. Even constrained internal rotations, such as the
increased barrier for rotation around the amide bond of
encarcerated dimethylacetamide have been observed.31 Whether
such phenomena can be detected in a capsule held together with
easily deformable hydrogen bonds remains to be seen, but there
is cause for optimism. Relevant to the case at hand is the
observation reported by Sherman.32 In a (heterodimeric) capsule
held together by only four hydrogen bonds, a desymmetrization
in the NMR spectrum of encapsulated pyrazine was interpreted
in terms of the guest’s restricted rotation.
(33) Temperature was kept constant at 300 K during the whole simulation.
Total time ) 100 ps, time steps ) 1 fs. GB/SA chloroform solvation was
used.
(34) Nicholls, A.; Sharp, K. A., Honig, B. Proteins 1991, 11, 281.
(35) This task is automatically accomplished by GRASP. During the
triangulation of the surfaces of the two softball halves for the calculation
of the interior volume, the program calculates the average position of all
the triangles at the edges of these surfaces. Joining of these edges generates
the surface of the internal cavity.
Additional insights can be gained by a quantitative evaluation
of the volume of the interior cavity of 1. In order to estimate
this volume in the most precise way, we first optimized the
(29) Craven, I. E.; Hesling, M. R.; Laver, D. R.; Lukins, P. B.; Ritchie,
G. L. D.; Vrbancich, J. J. Phys. Chem. 1989, 93, 627.
(30) Gentle, I. R.; Ritchie, G. L. D. J. Phys. Chem. 1989, 93, 7740.
(31) Helgeson, R. C.; Paek, K.; Knobler, C. B.; Maverick, E. B.; Cram,
D. J. J. Am. Chem. Soc. 1996, 118, 5590-5604.
(36) The occupancy factor of a guest molecule in the cavity of a host is
usually defined as the ratio of the van der Waals radius of the guest to the
volume of the cavity calculated on the van der Waals interior surface of
the host. We calculated this factor based on the volumes generated by the
program GRASP.
(37) T ) 300 K, total time ) 150 ps, time step ) 1fs. GB/SA chloroform
solvation was used.
(32) Chapman, R. G.; Sherman J. C. J. Am. Chem. Soc. 1995, 117, 9081-
9082.

Synthesis and Assembly of New Molecular Hosts
J. Am. Chem. Soc., Vol. 119, No. 1, 1997 83
1 L of water containing 15 mL of concentrated HCl. The mixture was
then filtered and the solids washed with water. The solids were slurried
in 300 mL of MeOH and filtered, which gave recovered excess glycoril
as a white solid. The filtrate was evaporated, and the residue slurried
in 200 mL of dichloromethane and filtered to give additional glycoluril.
The filtrate was evaporated, and the residue was chromatographed on
silica gel with 30% ethyl acetate/dichloromethane to give 1.87 g (56%)
of the product as a colorless foam. 1H NMR (CDCl3, mixture of
rotomers): δ 7.10 (br s, 2H), 5.80 (m, 2H), 5.18-4.75 (m, 4H), 4.50-
4.18 (m, 8H), 1.40-1.72 (m, 24H), 0.92 (m, 12H).
Tetraacid Chloride 8. To a suspension of 0.100 g (0.354 mmol)
of the tetraacid 12 (see below) in 1 mL of dichloromethane was added
1 mL of oxalyl chloride and 1 drop of a 10% v/v mixture of DMF in
dichloromethane. The mixture was stirred for 3 h or until complete
dissolution occured, then diluted with benzene and evaporated to give
0.120 g (96%) of the tetraacid chloride as a light brown solid. 1H
NMR (CDCl3): δ 4.77 (s, 4H), 1.92 (s, 4H).
Figure 12. Size and shape of the cavity in the softball. The cavity
inside the dimeric form of 1b: (a) in the free capsule; (b) in the presence
of 24. Both structures originated by molecular dynamic simulations
(see text for details).
Acetylenic Ester 9. To a mixture of 1.0 g (8.77 mmol) of acetylene
dicarboxylic acid in 30 mL of benzene at reflux was added 2.8 mL
(21.2 mmol) of 4-methoxybenzyl alcohol in 4 portions over 15 min.
The solution was then heated at reflux for 3 h with Dean-Stark removal
of water. Following evaporation of the benzene, the residue was
chromatographed on silica gel with 15% ethyl acetate/hexane to give
1.89 g (61%) of the ester as a white solid. 1H NMR (CDCl3): δ 7.30
(d, 4H, J ) 8.7 Hz), 6.88 (d, 4H, J ) 8.6 Hz), 5.17 (s, 4H), 3.81 (s,
6H). HRMS (EI): calculated for C20H18O6, 354.1103; found 354.1100.
Diester 9b. A mixture of 3.0 g (8.47 mmol) of the ester and 5 mL
of furan was heated in a sealed tube placed in an 85 °C oil bath for 4
h. Following evaporation of the furan, the residue was chromato-
graphed on silica gel with 30% ethyl acetate/hexane to give 2.33 g
(65%) of the ester as a white solid. 1H NMR (CDCl3): δ 7.31 (d, 4H,
J ) 8.4 Hz), 7.26 (s, 2H), 6.93 (d, 4H, J ) 8.6 Hz), 5.73 (s, 2H), 5.16
(dd, 4H, J ) 14.2, 26.5 Hz), 3.87 (s, 6H). HRMS (EI): calculated for
C24H22O7, 422.1366; found 422.1369.
an increased volume for the cavity (∼330 Å3), but the hydrogen-
bonding network remains intact; the ball is not destabilized. The
cavity shape after the simulation with two benzenes inside is
nearly spherical (Figure 12, left), while the space occupied by
the two benzenes is more elongated (shown on the right in
Figure 12). In the case of p-xylene, two molecules can fit into
the cavity only if the ball is deformed in such a way that at
least two hydrogen bonds are destroyed. This means that a
rupture is created in the softball and the assembly is destabilized.
This result nicely fits the experimental observations of a
broadened NMR spectrum in deuterated xylene.
The complex calculated between 1b and two benzene
molecules features the benzene molecules in parallel planes at
an average distance of 3.4 Å. This is approximately the distance
between layers in graphite. Accordingly, this system can be
considered as a microenvironment where the guest molecules
experience conditions that are more typical of the solid state.
When proximity effects are also taken into account, it then
becomes apparent that the interiors of these self-assembled balls
represent very special chambers where reactions might take
place. We will report on these developments in the sequel.
Finally, a word on vocabulary. Because there are not yet
single words in the language precise enough to provide the
images we intend to evoke, we have usedsand unabashedly
mixedssports, fruits, political, and astrophysical metaphors.
After all, reversible encapsulation is a recent concept and
somewhat discontinuous with other recognition phenomena, and
if the promise of discovery with these systems is fulfilled, new
words may have to be added to an already jargon-ridden
vocabulary.
Dihydro Diester 9a. A vigorously stirred mixture of 4.5 g (10.7
mmol) of the diene and ∼0.5 g of 5% Pd/C in 60 mL of ethyl acetate
was hydrogenated at ambient pressure until 260 mL of hydrogen had
1
been consumed. Additional hydrogen was introduced until H NMR
analysis showed that the diene was consumed. The mixture was then
filtered and the solvent evaporated to give 4.39 g (97%) of the
cyclohexene as a white solid. 1H NMR (CDCl3): δ 7.21 (d, 4H, J )
9.0 Hz), 6.82 (d, 4H, J ) 9.0 Hz), 5.2 (m, 2H), 5.04 (dd, 4H, J ) 13.5,
22.0 Hz), 3.76 (s, 6H), 1.80 (m, 2H), 1.42 (m, 2H). HRMS (EI):
calculated for C24H24O7, 424.1522; found 424.1517.
Dihydrophthalate 10. To a solution of 15.31 mL (140 mmol) of
TiCl4 in 30 mL of hexane at OC was added 125 mL of THF over 20
min. A slurry of 1.87 g (49.4 mmol) of LAH in 15 mL of THF was
then added over 10 min. The mixture was then heated at reflux for 30
min and cooled to room temperature, and a solution of 5.70 g (13.4
mmol) of the cyclohexene in 15 mL of THF was added over 5 min.
After 2 h, the mixture was poured into 1 L of 10% K2CO3(aq). The
mixture was extracted with ether and the organic extracts were dried
over MgSO4 and evaporated. The residue was chromatographed on
silica gel with 25% ethyl acetate/hexane to give 4.32 g (79%) of the
diene as a colorless oil. 1H NMR (CDCl3): δ 7.23 (d, 4H, J ) 9.1
Hz), 6.85 (d, 4H, J ) 8.7 Hz), 5.02 (s, 4H), 3.79 (s, 6H), 2.25 (m,
4H). HRMS (EI): calculated for C24H24O6, 408.1573; found 408.1567.
Bridged Tetraester 11. A mixture of 1.75 g (4.28 mmol) of the
diene and 1.98 g (5.57 mmol) of the acetylene in 10 mL of benzene
was heated at reflux for 40 h. Following evaporation of the benzene,
the residue was chromatographed on silica gel with 35% ethyl acetate/
hexane to give 2.46 g (61%) of a mixture of the starting components
and 1.06 g (33%) of the product tetraester as a colorless oil.
Resubjection of the recovered starting components to heating and
chromatography gave an additional 2.37 g (41%) of the product after
two additional cycles, for a total of 2.43 g (74%). 1H NMR (CDCl3):
δ 7.17 (d, 8H, J ) 8.5 Hz), 6.84 (d, 8H, J ) 8.5 Hz), 4.97 (s, 8H),
4.42 (s, 2H), 3.79 (s, 12H), 1.58 (s, 4H). 13C NMR (CDCl3): δ 164.6,
159.6, 140.8, 130.1, 127.3, 113.8, 66.9, 55.2, 40.8, 24.4. HRMS
(FAB): calculated for C44H42O12 -C2H4 + H+ ) C42H39O12, 737.2598;
found 737.2611.
Experimental Section
Hydrazide 4. A mixture of 2.61 g (11.24 mmol) of the hydrazide
3 and 15.16 g (33.7 mmol) of the tetrakis(bromomethyl)benzene 2 in
200 mL of DMF was heated to 65 °C. Then 1.35 g of NaH (60% in
mineral oil) was added in 20 mL of DMF and the mixture was stirred
for 20 min and then evaporated to dryness. The residue was slurried
in 300 mL of ether, the solids were allowed to settle, and the supernatant
was decanted. The supernatant was washed with water, dried over
MgSO4, and evaporated. The resulting residue was chromatographed
on silica gel with 12% ethyl acetate/hexanes to give 2.40 g (41%) of
the hydrazide as a colorless foam. 1H NMR (CDCl3, mixture of
rotomers): δ 7.13 (s, 2H), 5.2-4.7 (m, 2H), 4.62 (s, 4H), 4.4-4.2 (m,
2H), 1.47 (s, 18H). HRMS (EI): calculated for C20H28O4N2Br2,
518.0416; found 518.0413.
Glycoluril Hydrazide 6. To a solution of 17.05 g (46 mmol) of
isopentyl ester glycouril 5 in 300 mL of DMSO at 50 °C was added
10.33 g (92.1 mmol) of KOtBu. After the mixture was stirred for 20
min, 2.395 g (4.6 mmol) of the dibromide in 10 mL of DMSO was
added dropwise, the heating bath was removed, and the mixture was
allowed to stir for 25 min. The reaction mixture was then poured into
Centerpiece Tetraacid 12. To a solution of 1.33 g (1.74 mmol) of

84 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Meissner et al.
the tetraester in 3 mL of dichloromethane and 3 mL of anisole was
added 25 mL of TFA in one portion. After being stirred for 15 min,
the mixture was diluted with benzene and evaporated. The residue
was then diluted with ether and filtered to give 0.355 g (72%) of the
tetraacid as a white powder. 1H NMR (DMSO-d6): δ 4.27 (s, 2H),
1.45 (s, 4H). 13C NMR (DMSO-d6): δ 166.2, 140.8, 40.7, 24.1.
HRMS (FAB): calculated for C12H10O8, 282.0376; found 282.0371.
Synthesis of Furan 13. To a solution of 10.0 g (47.1 mmol) of
3,4-bis(acetoxymethyl)furan in 200 mL of methanol was added 7 mL
of a saturated solution of K2CO3 in methanol. The flask was placed
on a rotory evaporator with a bath temperature of about 50 °C, and the
flask was rotated with heating but no vacuum for 15 min. The vacuum
was then applied, and half the methanol was evaporated to drive out
the methyl acetate byproduct. Following the addition of 100 mL of
methanol, the procedure was repeated, and the solution fully evaporated.
Two sequential additions and evaporations of 100 mL of benzene each
were used to remove any residual methanol. The resulting residue was
dissolved in 80 mL of DMF and cooled to 0 °C, and 4.51 g (60% in
oil, 113 mmol, 2.4 eq) of NaH was added in 3 portions over 10 min.
After 15 min, 12.35 mL (103 mmol, 2.2 equiv) of benzyl bromide was
added over 1 h. The mixture was then allowed to warm to room
temperature and stirred for 15 h, poured into 500 mL of water, and
extracted with ether. The organic extracts were dried over MgSO4 and
evaporated. The residue was chromatographed on silica gel with 8%
ethyl acetate/hexanes to give 12.7 g (88%) of the dibenzyl ether as a
colorless oil. 1H NMR (CDCl3): δ 7.41 (s, 2H), 7.3-7.2 (m, 10H),
4.52 (s, 4H), 4.47 (s, 4H). HRMS (EI): calculated for C20H20O3,
308.1412; found 308.1414.
Bridged Diester 14. A mixture of 14.33 g (46.6 mmol) of the
dibenzylfuran and 6.93 g (48.7 mmol, 1.05 equiv) of dimethyl
acetylenedicarboxylate was heated to 110 °C for 1 h, then cooled to
room temperature. The residue was chromatographed on silica gel with
30% ethyl acetate/hexanes to give 18.69 g (89%) of the oxa-diene as
a colorless oil. 1H NMR (CDCl3): δ 7.4-7.2 (m, 10H), 5.69 (s, 2H),
4.45 (dd, 4H, J ) 12.0 Hz), 4.28 (dd, 4H, J ) 12.5 Hz), 3.73 (s, 6H).
HRMS (EI): calculated for C26H26O7, 450.1678; found 450.1683.
Durene Diol 15. To 200 mL of vigorously stirring THF at 0 °C
was added 29.73 mL (269 mmol) of TiCl4 dropwise over 20 min. A
suspension of 3.64 g (95.2 mmol) of LAH in 100 mL of THF was
then added in 3 portions over 10 min. The mixture was then heated at
reflux for 30 min and cooled to room temperature, and a solution of
18.69 g (41.4 mmol) of the oxadiene in 40 mL of THF was added
over 5 min. After 15 min, the mixture was poured into 1.4 L of 10%
K2CO3 in water, and filtered. The blue gelatinous residue was extracted
with THF, and the THF washes were dried over MgSO4 and evaporated.
The oily residue (17.0 g) was dissolved in 150 mL of THF, and 1.92
g (538 mmol) of LAH was added in 4 portions over 5 min. After the
mixture was stirred for 15 min, 4 g of ice was added cautiously, then
5 mL of 15% NaOH, then 7 mL of water, then 30 g of MgSO4.
Filtration and evaporation afforded an oily residue which was chro-
matographed on silica gel with 70% ethyl acetate/hexanes to give 14.14
g (90%) of the diol as a white solid. 1H NMR (CDCl3): δ 7.41 (s,
2H), 7.4-7.1 (m, 10H), 4.65 (s, 4H), 4.46 (s, 4H), 4.41 (s, 4H). HRMS
(EI): calculated for C24H26O4, 378.1831; found 378.1830.
water and filtered. The solid residue was extracted with THF, and the
THF was dried over MgSO4 and evaporated. The resulting residue
was slurried in 30 mL of pyridine, 3 mL of acetic anhydride was added,
and the mixture was heated at reflux for 2 h, during which time the
residue dissolved. The mixture was then evaporated and the residue
chromatographed on silica gel with 40% ethyl acetate/dichloromethane
to give 2.60 g (96%) of the monoacetate as a white solid. 1H NMR
(CDCl3): δ 7.49 (s, 1H), 7.47 (s, 1H), 7.4-6.9 (m, 10H), 6.34 (s, 1H),
4.96 (d, 2H, J ) 20.0 Hz), 4.85 (d, 2H, J ) 18.5 Hz), 4.6-4.5 (m,
8H), 4.29 (d, 2H, J ) 19 Hz), 4.18 (d, 2H, J ) 20.0 Hz), 2.52 (s, 3H).
HRMS (EI): calculated for C42H38N4O5, 678.2842; found 678.2846.
Protected Dihalide 20. A stream of HBr(g) was bubbled through
a solution of 2 g (2.95 mmol) of the monoacetate in 100 mL of CHCl3
for 8 min. The flask was then capped, and the solution was stirred for
3.5 h. Benzene was then added and the solution evaporated. The
residue was chromatographed on silica gel with 7.5% ethyl acetate/
dichloromethane to give 1.64 g (91%) of the dibromide as a white solid.
1H NMR (CDCl3): δ 7.4-6.8 (m, 12H), 6.28 (s, 1H), 4.93 (d, 1H, J
) 15.6 Hz), 4.81 (d, 1H, J ) 15.9 Hz), 4.62 (s, 2H), 4.60 (s, 2H), 4.27
(d, 1H, J ) 15.0 Hz), 4.14 (d, 1H, J ) 15.6 Hz), 2.51 (s, 3H). HRMS
(EI): calculated for C28H24N4O3Br2, 622.0216; found 622.0215.
Glycoluril Hydrazide 21. A mixture of 1.70 g (7.32 mmol) of the
hydrazide and 0.35 g (14.6 mmol) of NaH in 20 mL of DMF was
stirred for 15 min. The dibromide was then added as a solid, and the
mixture was then stirred 20 min and poured into 300 mL of water.
The mixture was extracted with ether and the organic extracts were
dried over MgSO4 and evaporated. The resulting residue was dissolved
in 30 mL of THF and 6 mL of MeOH, and 5 drops of saturated LiOH-
(aq) was added. After 30 min, the mixture was evaporated to one-
fourth the original volume, diluted with dichloromethane, dried over
MgSO4, and evaporated. The resulting residue was chromatographed
on silica gel with 50% ethyl acetate/dichloromethane to give 1.46 g
(84%) of the hydrazide as a colorless oil. 1H NMR (CDCl3, mixture
of rotamers): δ 7.3-7.0 (m, 12H), 5.1-4.7 (m, 4H), 4.4-4.0 (m, 4H),
1.5-1.3 (m, 18H). HRMS (EI): calculated for C36H40N6O6, 652.3009;
found 652.3011.
Glycoluril Amine 22. A stream of HCl(g) was bubbled through a
solution of 0.060 g (0.0919 mmol) of the monoacetate in 3 mL of
nitromethane and 1 mL of CHCl3 for 1 min. The solvents were then
evaporated to give 0.044 g (98%) of the hydrochloride as a white
powder. 1H NMR (DMSO-d6): δ 9.9 (br s, 3H), 8.18 (s, 2H), 7.3-
7.1 (m, 10H), 4.79 (d, 2H, J ) 11 Hz), 4.38 (s, 4H), 4.18 (d, 2H, J )
12 Hz). HRMS (EI): calculated for C36H40N6O6, 652.3009; found
652.3011.
Softball 1a (R ) Phenyl). To a mixture of 0.10 g (0.205 mmol) of
the tetrahydrophthalazine hydrochloride in 2 mL of DMF was added
0.143 mL (1.02 mmol) of NEt3, and the mixture was stirred to effect
dissolution. After cooling to 0 °C, a solution of 0.0364 g (0.102 mmol)
of the tetraacid chloride in 0.4 mL of CHCl3 was added dropwise. The
mixture was allowed to warm to room temperature, stirred for 14 h,
then evaporated to dryness. The residue was slurried in 2 mL of MeOH,
then 5 mL of water was added, the mixture was filtered, and the solids
were washed with water. Drying of the solids under vacuum gave 0.101
g (89%) of the crude isomer mixture as a white solid. The residue
was chromatographed on silica gel with 10-30% MeOH/chloroform
to give the C-shaped isomer (middle fraction) as a colorless oil. 1H
NMR (DMSO-d6): δ 8.17 (s, 4H), 7.46 (s, 4H), 7.18-7.0 (m, 20H),
5.33 (d, 4H, J ) 15.6 Hz), 5.07 (d, 4H, J ) 15.9 Hz), 4.98 (s, 2H),
4.68 (d, 4H, J ) 11.6 Hz), 4.06 (d, 4H, J ) 15.9 Hz), 1.44 (s, 4H).
HRMS (CI): calculated for C64H50N12O8, 1114.3874; found 1114.3872.
Softball 1b. Through a solution of 1.0 g (1.37 mmol) of the
hydrazide in 50 mL of nitromethane and 10 mL of chloroform was
bubbled a gentle stream of HCl(g) with a pasteur pipet for 30 s. After
the mixture was stirred for 1 min, the solution was purged with argon
for 3 min, which caused a white precipitate to form. The solvent was
evaporated, and the residue was subjected to high vacuum, producing
0.734 g (95%) of the hydrazine hydrochloride as an off-white foam.
To a solution of this foam in 8 mL of DMF was added a trace of DMAP
and 0.956 mL (6.50 mmol) of NEt3, and the solution was cooled to 0
°C. A solution of 0.231 g (0.650 mmol) of the acid chloride in 2 mL
of chloroform was added dropwise, and then the solution was allowed
to warm to room temperature and stirred for 14 h. The solvents were
Durene Dihalide 16. To a mixture of 17.4 mL (149 mmol) of 2,6-
lutidine, 8.65 mL (112 mmol) of methanesulfonyl chloride, and ∼0.5
g of DMAP in 90 mL of DMF at 0 °C was added a solution of 14.1 g
(37.3 mmol) of the diol in 40 mL of DMF over 10 min. The mixture
was then allowed to warm to room temperature and stirred for 15 h.
Then 20 g of LiBr was added in 50 mL of THF, and the mixture was
stirred for 1 h, then poured into 500 mL of 1.5 N HCl and extracted
with ether. The organic extracts were dried over MgSO4 and
evaporated. The residue was chromatographed on silica gel with 12%
ethyl acetate/hexanes to give 14.37 g (76%) of the dibromide as a white
solid. 1H NMR (CDCl3): δ 7.43 (s, 2H), 7.4-7.2 (m, 10H), 4.70 (s,
4H), 4.52 (s, 4H), 4.49 (s, 4H). HRMS (EI): calculated for C24H24-
Br2O2, 502.0144; found 502.0139.
Acylated Glycoluril 19. To a solution of 11.67 g (39.7 mmol) of
diphenylglycouril in 280 mL of DMSO was added 4.45 g (79.3 mmol)
of powdered KOH. The mixture was heated to 50 °C, and after 10
min, 2.0 g (3.97 mmol) of the dibromide was added in 50 mL of DMSO.
The solution was stirred at 50 °C for 10 min, then poured into 1.2 L of

Synthesis and Assembly of New Molecular Hosts
J. Am. Chem. Soc., Vol. 119, No. 1, 1997 85
then evaporated to give a moist residue, which was dissolved in 4 mL
of MeOH and added to 30 mL of water. The mixture was filtered,
and the solids washed with water. To the solids were added 40 mL of
benzene, and the mixture was vigorously stirred for 15 min. Filtration
and evaporation of the filtrate gave a tan solid which was about 75%
pure C-shaped isomer by NMR. Chromatograph of this residue on
silica gel with 10% MeOH/25% ethyl acetate/65% chloroform gave
0.071 g (9%) of the product as an off-white glassy solid. Additional
C-shaped isomer can be obtained from chromatography of the remaining
crude residue. 1H NMR (DMSO-d6): δ 8.45 (s, 4H), 7.43 (s, 3H),
5.25 (d, 4H, J ) 16.8 Hz), 5.09 (d, 4H, J ) 16.9 Hz), 4.97 (s, 2H),
4.60 (d, 4H, J ) 15.0 Hz), 4.45 (d, 4H, J ) 15.0 Hz), 4.19 (t, 4H, J
) 6.6 Hz), 4.08 (t, 4H, J ) 6.6 Hz), 1.7-1.4 (m, 16H), 0.88 (d, 12H,
J ) 4.5 Hz), 0.86 (d, 12H, J ) 4.8 Hz). 13C NMR (benzene-d6): δ
167.6, 167.2, 157.9, 153.6, 144.0, 139.2, 129.8, 129.6, 126.0, 83.7,
74.9, 65.8, 65.5, 46.0, 45.1, 37.8, 37.6, 25.7, 25.5, 23.0, 22.9. HRMS
(FAB): calculated for C64H74O16N12, 1266.5346; found 1266.5334.
Acknowledgment. We thank the National Institutes of
Health and the Skaggs Foundation for financial support and Drs.
David Case and Doree Sitkoff for helpful discussions concerning
the computations. R.S.M. is grateful to the National Institutes
of Health for a postdoctoral fellowship.
JA962991F