Design, synthesis, and conformational dynamics of a gated molecular basket

Article abstract of DOI:10.1021/ja060534l

We have developed a synthesis and examined the conformational behavior and recognition properties of dynamic molecular containers 1-3. As follows from the ¹H NMR dilution, diffusion NMR, and vapor pressure osmometry measurements, compound 1 has a low affinity for intermolecular aggregation and is mostly present in monomeric form in dilute chloroform solutions. Inspecting the O-H chemical shift resonances of 1, 3, and model compound 4 as a function of temperature afforded the Δδ/ΔT coefficients of 17.0, 17.3, and 4.7 ppb K^-1, respectively. In combination with the results from variable temperature ¹H NMR and IR measurements, the existence of conformers of 1 and 3 in equilibrium, each having a different extent of hydrogen bonding, was confirmed. Molecular mechanics calculations suggested 1 _a as the most favorable conformation, with three additional conformers, 1_b, 1_C, and 1_d, populating local energy minima. Further optimization of each of the four conformers using semiempirical PM3 and ab initio (HF/6-31G*) methods allowed a determination of their relative free energies and the corresponding Boltzmann population distributions which were heavily weighted toward 1_a. A computed composite IR spectrum of a fraction-weighted mixture of the conformers of 1 reproduced the experimentally observed IR spectrum in its structural features, leading to a conclusion that conformer 1_a indeed dominates the equilibrium. The egg-shaped cavity of 1 (136.6 A³) is complementary in size, shape, and electrostatic potential to chloroform (74.9 A³). A single-crystal X-ray study of 2 revealed a disordered chloroform molecule positioned inside the cavitand along its C₃, axis.

Full text of DOI:10.1021/ja060534l

Published on Web 04/06/2006
Design, Synthesis, and Conformational Dynamics of a Gated
Molecular Basket
Veselin Maslak, Zhiqing Yan, Shijing Xia, Judith Gallucci,
Christopher M. Hadad, and Jovica D. Badjic´*
Contribution from the Department of Chemistry, The Ohio State UniVersity,
100 West 18th AVenue, Columbus, Ohio 43210
Received January 23, 2006; E-mail: badjic@chemistry.ohio-state.edu
Abstract: We have developed a synthesis and examined the conformational behavior and recognition
properties of dynamic molecular containers 1-3. As follows from the 1H NMR dilution, diffusion NMR, and
vapor pressure osmometry measurements, compound 1 has a low affinity for intermolecular aggregation
and is mostly present in monomeric form in dilute chloroform solutions. Inspecting the O-H chemical shift
resonances of 1, 3, and model compound 4 as a function of temperature afforded the ∆δ/∆T coefficients
1
of 17.0, 17.3, and 4.7 ppb K-1, respectively. In combination with the results from variable temperature H
NMR and IR measurements, the existence of conformers of 1 and 3 in equilibrium, each having a different
extent of hydrogen bonding, was confirmed. Molecular mechanics calculations suggested 1a as the most
favorable conformation, with three additional conformers, 1b, 1c, and 1d, populating local energy minima.
Further optimization of each of the four conformers using semiempirical PM3 and ab initio (HF/6-31G*)
methods allowed a determination of their relative free energies and the corresponding Boltzmann population
distributions which were heavily weighted toward 1a. A computed composite IR spectrum of a fraction-
weighted mixture of the conformers of 1 reproduced the experimentally observed IR spectrum in its structural
features, leading to a conclusion that conformer 1a indeed dominates the equilibrium. The egg-shaped
cavity of 1 (136.6 A3) is complementary in size, shape, and electrostatic potential to chloroform (74.9 Å3).
A single-crystal X-ray study of 2 revealed a disordered chloroform molecule positioned inside the cavitand
along its C3 axis.
Introduction
assembled capsules, and metal-assembled cage molecules have
been constructed to surround a target molecule and selectively
The design and study of molecules that, apart from being
receptors, can also regulate kinetics and thermodynamics of the
recognition is essential for understanding and linking structure
with function in synthetic systems.1 The inspiration comes from
the natural world where traffic of substrates, products, or solutes
in and out of the active sites of many enzymes is regulated by
the rapid opening and closing of a flaplike “gate” or by
conformational adjustments of the macromolecule in response
to an external stimulus.2 Molecular vessels such as cavitands,
cyclotriveratrylenes, carcerands, velcrands, trinacrenes, self-
recognize it.3 Housing a guest in this manner allowed its
protection from external reactants4 and transport between phases5
and provided a molecular reaction domain.6 The controlled
release of small molecules by hemicarceplexes, metal-containing
self-assembled molecular cages, or self-assembled capsules has
(3) (a) Moran, J. R.; Karbach, S.; Cram, D. J. J. Am. Chem. Soc. 1982, 104,
5826-5828. (b) Moran, J. R.; Ericson, J. L.; Delcanale, E.; Bryant, J. A.;
Knobler, C. B.; Cram, D. J. J. Am. Chem. Soc. 1991, 113, 113, 5707-
5714. (c) Ashton, P. R.; Girreser, U.; Giuffrida, D.; Kohnke, F. H.; Mathias,
J. P.; Raymo, F. M.; Slawin, A. M. Z.; Stoddart, J. F. J. Am. Chem. Soc.
1993, 115, 5422-5429. (d) Collet, A.; Dataset, J. P.; Lozach, B.; Canceill,
J. Top. Curr. Chem. 1993, 165, 103-129. (e) Ikeda, A.; Shinkai, S. Chem.
ReV. 1997, 97, 1713-1734. (f) Jasat, A.; Sherman, J. C. Chem. ReV. 1999,
99, 931-967. (g) Caulder, D. L.; Raymond, K. N. Acc. Chem. Res. 1999,
32, 975-982. (h) Tucci, F.; Rudkevich, D. M.; Rebek, J., Jr. Chem.sEur.
J. 2000, 6, 1007-1016. (i) Fox, O. D.; Drew, M. G. B.; Beer, P. D. Angew.
Chem., Int. Ed. 2000, 39, 136-140. (j) Hof, F.; Craig, S. L.; Nuckolls, C.;
Rebek, J., Jr. Angew. Chem., Int. Ed. 2002, 41, 1488-1508. (k) Sun, W.
J.; Yoshizawa, M.; Kusukawa, T.; Fujita, M. Curr. Opin. Chem. Biol. 2002,
6, 757-764. (l) Kryschenko, Y. K.; Seidel, S. R.; Arif, A. M.; Stang, P. J.
J. Am. Chem. Soc. 2003, 125, 5193-5198. (m) Casnati, A.; Sansone, F.;
Ungaro, R. Acc. Chem. Res. 2003, 36, 246-254. (n) Ryu, E.-H.; Zhao, Y.
Org. Lett. 2004, 6, 3187-3189. (o) Van, L.; Fijs, W. B.; Beijleveld, H.;
Kooijman, H.; Spek, A. L.; Verboom, W.; Reinhoudt, D. N. J. Org. Chem.
2004, 69, 3928-3936. (p) Dalgarno, S. J.; Tucker, S. A.; Bassil, D. B.;
Atwood, J. L. Science 2005, 309, 2037-2039. (q) Isaacs, L.; Park, S.-K.;
Liu, S.; Ko, Y. H.; Selvapalam, N.; Kim, Y.; Kim, H.; Zavalij, P. Y.; Kim,
G.-H.; Lee, H.-S.; Kim, K. J. Am. Chem. Soc. 2005, 127, 18000-18001.
(r) Rudkevich, D. M.; Kang, Y.; Leontiev, A. V.; Organo, V. G.; Zyryanov,
G. V. Supramol. Chem. 2005, 17, 93-99. (s) Azov, V. A.; Schlegel, A.;
Diederich, F. Angew. Chem., Int. Ed. 2005, 44, 4635-4638.
(1) (a) Mogck, O.; Pons, M.; Boehmer, V.; Vogt, W. J. Am. Chem. Soc. 1997,
119, 5706-5712. (b) Vysotsky, M. O.; Thondorf, I.; Bohmer, V. Angew.
Chem., Int. Ed. 2000, 39, 1264-1267. (c) Davis, A. V.; Yeh, R. M.;
Raymond, K. N. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 4793-4796. (d)
Zhong, Z.; Anslyn, E. V. Angew. Chem., Int. Ed. 2003, 42, 3005-3008.
(e) Davis, J. T.; Kaucher, M. S.; Kotch, F. W.; Iezzi, M. A.; Clover, B. C.;
Mullaugh, K. M. Org. Lett. 2004, 6, 4265-4268. (f) Heemstra, J. M.;
Moore, J. S. J. Org. Chem. 2004, 69, 9234-9237. (g) Gianneschi, N. C.;
Cho, S.-H.; Nguyen, S. B. T.; Mirkin, C. A. Angew. Chem., Int. Ed. 2004,
43, 5503-5507. (h) Zuccaccia, D.; Pirondini, L.; Pinalli, R.; Dalcanale,
E.; Macchioni, A. J. Am. Chem. Soc. 2005, 127, 7025-7032. (i) Rebek,
J., Jr. Angew. Chem., Int. Ed. 2005, 44, 2068-2078.
(2) (a) Nolte, H.-J.; Rosenberry, T. L.; Neumann, E. Biochemistry 1980, 19,
3705. (b) McCammon, J. A.; Northrup, S. H. Nature 1981, 293, 316. (c)
Sussman, J. L.; Harel, M.; Frolow, F.; Oefner, C.; Goldman, A.; Toker,
L.; Silman, I. Science 1991, 253, 872-879. (d) Wlodek, S. T.; Clark, T.
W.; Scott, L. R.; McCammon, J. A. J. Am. Chem. Soc. 1997, 119, 9513-
9522. (e) Zhou, H.-X.; Wlodek, S. T.; McCammon, J. A. Proc. Natl. Acad.
Sci. 1998, 95, 9280.
9
10.1021/ja060534l CCC: $33.50 © 2006 American Chemical Society
J. AM. CHEM. SOC. 2006, 128, 5887-5894
5887

A R T I C L E S
Maslak et al.
been investigated.7 However, the regulation of the kinetics of
molecular recognition, i.e., guest transport in/out of the host,
in synthetic receptors is still a challenging task. Rebek et al.
have recently recognized8 that the energy barrier for guest
exchange in resorcinarene-based cavitands can be controlled by
restricting their conformational dynamics. A set of amide groups
installed at the rim of the resorcinarene led to the formation of
a net of hydrogen bonds at the cavity opening that rigidified
the cavitand’s conformation. Remarkably, upon accommodating
a target guest molecule, these cavitands formed complexes with
high kinetic stabilities despite their open-ended structures and
low binding affinities. It is plausible that gaining control over
the kinetics of in/out molecular shuttling for artificial receptors
will allow the following: (a) regulation of useful and encap-
sulated reaction chemistry, (b) development of novel and
efficient molecular devices for delivery purposes, and (c) direct
engineering of controllable ion and molecular channels. In that
perspective, our efforts are focused on the design of molecular
containers having a set of stimuli-responsiVe gates, to enclose
space and thus regulate the incarceration of target molecules,
as depicted in Figure 1. In the study described here, we report
on the synthesis, conformational analysis, and preliminary
recognition properties of molecular containers 1-3 (Figure 1).
These compounds have a flat aromatic base which is fused to
three bicyclo[2.2.1]heptane rings to form a curved unit. Three
phthalimides extend this curvature into a rigid bowl-shaped
cavitand, to which three aromatic gates, each containing a single
CH2 rotor and, in 1 and 3, a pH responsive OH group, are
appended. We envision that (a) these three aromatics will
assemble on top of the cavitand, by way of directed intramo-
lecular O-H‚‚‚O hydrogen bonds, and thereby serve as gates
to regulate in/out exchange of guests and (b) the rate of opening
and closing of the gates, and therefore the kinetics of the guest
exchange, will be controlled by pH to regulate their degree of
protonation. Phenolic host molecules such as hydroquinone,
Figure 1. Energy minimized (Hartee-Fock, HF/6-31G*) conformations 1a,
1b, 1c, and 1d and their calculated relative standard free energies ∆G°;
based on the calculation, the conformer 1a dominates the equilibrium state,
94%. Chemical structures of dynamic receptors 1-3, with their H NMR
spectroscopic assignment. An external base/acid input is envisioned to drive
1
the opening and closing of 1 (see Figure 6).
phenol, and Dianin’s compound are known to form hydrogen-
bonded clathrates containing hexameric units of hosts stacked
on top of each other to encapsulate guests between the sextets.9
Integrated into dynamic receptors, but only in a multivalent
fashion, phenols are herein designed to assemble in a comparable
manner.
The design and study of 1-3 as gated molecular baskets as
well as their related analogues are directed at expanding our
fundamental knowledge of noncovalent forces which can drive
molecular self-assembly and at developing the important
principles for creating dynamic receptors. Moreover, elucidating
the mechanism by which these dynamic hosts operate will aid
in the predictable incorporation of their dynamic behavior
toward building more efficient receptors for various applications.
(4) (a) Cram, D. J.; Tanner, M. E.; Thomas, R. Angew. Chem., Int. Ed. Engl.
1991, 30, 1024-1027. (b) Kurdistani, S. K.; Helgeson, R. C.; Cram, D. J.
J. Am. Chem. Soc. 1995, 117, 1659-1960. (c) Warmuth, R.; Marvel, M.
A. Angew. Chem., Int. Ed. 2000, 39, 1117-1119. (d) Marquez, C.; Nau,
W. M. Angew. Chem., Int. Ed. 2001, 40, 4387-4390. (e) Ziegler, M.;
Brumaghim, J. L.; Raymond, K. N. Angew. Chem., Int. Ed. 2000, 39, 4119-
4121. (f) Makeiff, D. A.; Vishnumurthy, K.; Sherman, J. C. J. Am. Chem.
Soc. 2003, 125, 9558-9559. (g) Kaanumalle, L. S.; Gibb, C. L. D.; Gibb,
B. C.; Ramamurthy, V. J. Am. Chem. Soc. 2004, 126, 14366-14367. (h)
Kirmse, W. Angew. Chem., Int. Ed. 2005, 44, 2476-2479. (j) Hooley, R.
J.; Rebek, J., Jr. J. Am. Chem. Soc. 2005, 127, 11904-11905.
(5) Rudkevich D. M. Bull. Chem. Soc. Jpn. 2002, 75, 393-413.
(6) (a) Beno, B. R.; Sheu, C.; Houk, K. N.; Warmuth, R.; Cram, D. J. Chem.
Commun. 1998, 3, 301-302. (b) Yoshizawa, M.; Takeyama, Y.; Kusukawa,
T.; Fujita, M. Angew. Chem., Int. Ed. 2002, 41, 1347-1349. (c) Fiedler,
D.; Bergman, R. G.; Raymond, K. N. Angew. Chem., Int. Ed. 2004, 43,
6748-6751. (d) Vriezema, D. M.; Aragones, M. C.; Elemans, J. A. A.;
Cornelissen, J. J. L. M.; Rowan, A. E.; Nolte, R. J. M. Chem. ReV. 2005,
105, 1445-1489. (e) Purse, B. W.; Rebek, J., Jr. Proc. Natl. Acad. Sci.
U.S.A. 2005, 102, 10777-10782. (f) Purse, B. W.; Gissot, A.; Rebek, J.,
Jr. J. Am. Chem. Soc. 2005, 127, 11222-11223. (g) Cacciapaglia, R.;
Casnati, A.; Mandolini, L.; Reinhoudt, D. N.; Salvio, R.; Sartori, A.; Ungaro,
R. J. Org. Chem. 2005, 70, 5398-5402.
Results and Discussion
Synthesis. The synthesis of 1-3 is outlined in Scheme 1.
The originally reported methodology for the preparation of 8
has been modified by us.10 Diels-Alder cycloaddition of
cyclopentadiene to cis-1,4-dichloro-2-butene, in refluxing ben-
zene, gave compound 5 in 71% yield. The bromination of 5 to
yield 6 was entailed with some difficulties.
We noticed that the reaction temperature had a dramatic effect
on the product distribution. At room temperature, the reaction
gave the products of Wagner-Meerwein rearrangement, which
could be rationalized by an ionic addition of Br2 to norbornene
mediated by a σ-bridged nonclassical 2-norbornyl cation
intermediate.11 In boiling decalin at 150 °C, the addition of
bromine to 5 was not accompanied with significant skeletal
rearrangements and yielded 6 as the major product, through, as
(7) (a) Robbins, T. A.; Cram, D. J. J. Chem. Soc., Chem. Commun. 1995, 15,
1515-1516. (b) Sherman, J. C. Tetrahedron 1995, 51, 3395-422. (c)
Branda, N.; Grotzfeld, R. M.; Valdes, C.; Rebek, J., Jr. J. Am. Chem. Soc.
1995, 117, 85-88. (d) Houk, K. N.; Nakamura, K.; Sheu, C.; Keating, A.
E. Science 1996, 273, 627-629. (e) Piatnitski, E. L.; Deshayes, K. D.
Angew. Chem., Int. Ed. 1998, 37, 970-973. (f) Place, D.; Brown, J.;
Deshayes, K. D. Tetrahedron Lett. 1998, 39, 5915-5918. (g) Kerckhoffs,
J. M. C. A.; van Leeuwen, F. W. B.; Spek, A. L.; Kooijman, K.; Crego-
Calama, M.; Reinhoudt, D. N. Angew. Chem., Int. Ed. 2003, 42, 5717-
5722. (h) Hiraoka, S.; Harano, K.; Shiro, M.; Shionoya, M. Angew. Chem.,
Int. Ed. 2005, 44, 2727-2731. (i) Davis, A. V.; Raymond, K. N. J. Am.
Chem. Soc. 2005, 127, 7912-7919.
(9) Atwood, J. L., Davies, J. E. D., MacNicol, D. D., Eds. Inclusion
Compounds; Academic Press: London, U.K., 1984; Vol 2.
(10) Borsato, G.; De Lucchi, O.; Fabris, F.; Lucchini, V.; Pasqualotti, M.;
Zambon, A. Tetrahedron Lett. 2002, 44, 561-563.
(11) (a) Colter, A.; Friedrich, E. C.; Holness, N. J.; Winstein, S. J. Am. Chem.
Soc. 1965, 87, 378-379. (b) Walling, C. Acc. Chem. Res. 1983, 16, 448-
454.
(8) (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.
9
5888 J. AM. CHEM. SOC. VOL. 128, NO. 17, 2006

Gated Molecular Baskets
A R T I C L E S
Scheme 1. Synthesis of Dynamic Molecular Containers 1-3 and Model Compound 4
we believe, a radical process.12 The preferred stereochemistry
of the cis addition, occurring from the exo side, has been
confirmed by 2D-NOESY measurements. The compound 6,
upon successive halogen eliminations using t-BuOK and then
stanannylation, afforded the bromo(trimethylstannyl)triene 7.
Cyclotrimerization of 7, promoted by copper(I) thiophenylcar-
boxylate (Cu(I)Tc), yielded 8 as a 4:1 mixture of anti and syn
isomers in 65% isolated yield. The desired syn isomer was
successfully separated from the anti using column chromatog-
raphy. We initially anticipated that 8 would be highly reactive
in Diels-Alder reactions.10 Interestingly, the cycloaddition of
a reactive dienophile, dimethyl acetylenedicarboxylate (DMAD),
to 8 had to be performed at a high pressure in order to yield 9,
after the DDQ aromatization step. Hexaester 9 was then
converted to a stable tris-anhydride 10, which, in a reaction with
14, benzylamine, or 18, followed by subsequent hydrogenolysis,
yielded 1, 2, and 3, in 76, 61, and 10% yield, respectively. For
the preparation of 14, commercially available 4-hydroxyben-
zaldehyde 11 was converted to 13 that in a reaction with LiAlH4
yielded 14 as a major product. Similarly, 4-benzyloxy-2-
methoxybenzylamine 18 was synthesized starting from 2,4-
dihydroxybenzaldehyde 15. The model compound 4 was
obtained by reacting phthalanhydride with 4-benzyloxybenzy-
lamine 14, followed by hydrogenolysis promoted with Pd/C
catalyst.
which was then readily soluble in chloroform or dichlo-
romethane. This observation provided qualitative evidence that
both of these two solvents efficiently solvate the interior of 1
and 3, and as such, render them soluble. It is likely that
methanol, as a polar hydrogen bonding solvent, aided the
solvation process by breaking the crystal lattice of 1 and 3.
Interestingly, 2 was directly solubilized in nonpolar organic
solvents, without any pretreatment of its solutions with methanol.
1H NMR Spectroscopic Studies. Compounds 1-3 were
characterized by 1H NMR spectroscopy, including 1H-1H
COSY and TROESY two-dimensional experiments in order to
achieve full structural assignment. The 1H NMR spectrum of 1
in dry CDCl3 is assigned to a molecule with averaged C3V
symmetry, Figure 2a. The signal for its O-H proton shifted
upfield (from 6.88 to 6.66 ppm) as the CDCl3 solution was
diluted (from 3.8 to 0.5 mM) and then stayed constant after
1
further dilution. The H NMR signal for the O-H proton of
the model compound 4 (Figure 2b) also shifted to higher field,
from 4.79 to 4.63 ppm with 30.0 to 1.0 mM dilution, and then
stayed constant at lower concentrations. The observed depen-
dence of the O-H resonance shift with the concentration of
both 1 and 4 provides some evidence for intermolecular
hydrogen-bonding interactions which decreased upon dilution.
The small magnitude of the shift, however, signified nonex-
tensive aggregation.
Solubility and Sample Preparation. The solubility of 1 and
3 was very low in a range of neat organic solvents. We
discovered that the addition of methanol to a suspension of 1
or 3 in chloroform afforded a homogeneous solution. The
evaporation of such a prepared solution to dryness, using
elevated temperature in vacuo, afforded the desired compound
The model compound 4 self-associates via intermolecular
hydrogen bonding. Assuming a linear aggregation of 4 in
solution, the nonlinear curve fitting of its dilution 1H NMR data
to an EK isodesmic mathematical model yielded an apparent
association constant of 53 ( 10 M-1 13
. From this analysis, the
normalized weight fraction for the oligomeric distribution
(12) Tutar, A.; Taskesenligil, Y.; Cakmak, O.; Abbasoglu, R.; Balci, M. J. Org.
Chem. 1996, 61, 8297-8300.
(13) Martin, R. B. Chem. ReV. 1996, 96, 3943-3063.
9
J. AM. CHEM. SOC. VOL. 128, NO. 17, 2006 5889

A R T I C L E S
Maslak et al.
Figure 2. 1H NMR spectra (400 MHz, 298 K) of (a) 1 (0.5 mM) and (b) 4 (1.0 mM), in dry CDCl3. Insets: segments of 1H NMR spectra, showing the O-H
resonance shifts of 1 (top) and 4 (bottom) obtained in dilution experiments.
indicates that 82% of the nonassociated and 15% of dimeric 4
are present in the 5.0 mM solution. A low propensity of 4 for
having intermolecular association is particularly evident. The
trivalent 1, however, could form both inter- and intramolecular
hydrogen bonds. A nearly complete absence of intermolecular
association or aggregation is suggested by the invariant O-H
chemical shift observed in 0.5 mM or less concentrated solutions
of 1 (see inset in Figure 2a).
Vapor Pressure Osmometry. The intermolecular association
of 1 and 4 was also examined using vapor pressure osmometry
(VPO). It is important to note that, in this technique, the
experimental molecular weight of the unknown is a weighted
average of the molecular weights of its free and aggregated
molecules.14 Calibration with a standard is required, and
consequently, the results are dependent on the choice of the
standard.15 The observed molecular weight of the model 4
(calculated Mw ) 253) varied between 256 ( 5 and 325 ( 7
for its 3.7-8.3 mM chloroform solutions, Figure 3a. A low
degree of intermolecular aggregation over this concentration
1
range is evident, and these results are in accord with the H
NMR chemical shifts and spectra. The VPO measurements of
1, using sucrose octaacetate and triacetyl-â-cyclodextrin as
standards, showed that the observed molecular weight deviates,
but not drastically, about the value calculated for the free 1 (Mw
Figure 3. Vapor pressure osmometry determined molecular weights of
) 946), Figure 3b. At lower concentrations, the observed
molecular weights were below (almost 48%!) the expected value
and is presumably a result of the large experimental uncertainty
at those concentrations. The difference in the voltage (∆V)
(a) model compound 4 and (b) dynamic receptor 1, in dry chloroform at
310 K, using sucrose octaacetate (b), triacetyl-â-cyclodextrine (2), and
benzil (9) as standards. The error bars correspond to the standard deviations
of the VPO measurements of the sample.
between the thermister beads in the osmometer is proportional
to the difference in the vapor pressures of the solvent and the
sample. The voltage change is smaller when the concentration
of the sample (C) is decreased so the measured ∆V/C, used to
(14) (a) Solie, T. N. Methods Enzymol. 1972, 26, 50. (b) Hiroki, S.; Takahira,
Y.; Yamaguchi, M. J. Org. Chem. 2005, 70, 5698-5708.
(15) Seto, C. T.; Whitesides, G. M. J. Am. Chem. Soc. 1993, 115, 905-916.
(b) Zimmerman, S. C.; Zeng, F.; Reichert, D. E. C.; Kolotuchin, S. V.
Science 1996, 271, 1095-98.
9
5890 J. AM. CHEM. SOC. VOL. 128, NO. 17, 2006

Gated Molecular Baskets
A R T I C L E S
determine the molecular weight, becomes uncertain in highly
dilute solutions. Nonetheless, these experiments unambiguously
suggested that, in the examined range of chloroform concentra-
tions, 1 is mostly present in its monomeric state.
Diffusion NMR Spectroscopy. HR-DOSY16 measurements
additionally confirmed the low propensity of 1 for intermolecular
association. The observed diffusion coefficient of 1, in its 0.3-
1.0 mM CDCl3 solutions, increased from 5.50 × 10-10 to 6.17
× 10-10 m2 s-1. Importantly, under the conditions of fast
chemical exchange, the observed diffusion coefficient D of the
unknown has the average value of the free Dfree and bound Dbound
states weighted by their fraction.17 The slight variation in the
experimentally determined diffusion coefficients with concentra-
tion, therefore, indicated insignificant alternations in the equi-
librium composition and, all together with the 1H NMR dilution
and VPO studies, confirmed that 1 is indeed “free” in dilute
chloroform solutions.
Figure 4. 1H NMR chemical shifts for the O-H in 1 (2, 0.5 mM), 3 (9,
0.5 mM), and 4 (b, 1.0 mM) in dry CDCl3 as a function of temperature.
Based on these data, temperature dependence chemical shift coefficients
∆δ/∆T, 223-323 K, were calculated to be 17.0 × 10-3 ppm K-1 for 1,
17.3 × 10-3 ppm K-1 for 3, and 4.7 × 10-3 ppm K-1 for 4.
1
Conformational Studies. H NMR Spectroscopy. Ruling
out the aggregation of 1 allowed us to investigate its confor-
mational behavior. The resonance shift for the O-H groups in
the 1H NMR spectrum of 1 is positioned downfield with respect
to the O-H signal for the model compound 4, ∆∆δ ) 2.03
ppm (see Figure 2). It is reasonable to ascribe this observation
to the existence of intramolecular hydrogen-bonding interactions
involving the three phenol gates of the capsule. Recall that the
1H NMR spectrum of 1 is assigned to a molecule with averaged
that is, the higher the chemical shift, the more extensive the
hydrogen-bonding becomes. A small, but positive, temperature
coefficient (4.7 × 10-3 ppm K-1) is calculated for 4 and implies
a somewhat intensified hydrogen-bonding interaction as the
temperature is lowered (Figure 4). Evidently, the model
compound aggregates to a small extent at lower temperatures.
In contrast, the temperature coefficient calculated for the O-H
resonance in 1 and 3 (17.0 × 10-3 and 17.3 × 10-3 ppm K-2
,
C
3V symmetry, and this indicates (a) the simultaneous presence
respectively) was significantly bigger. The existence of the
conformers of 1 and 3 in equilibrium, each having a different
extent of hydrogen bonding, could account for this result, as
the corresponding equilibrium constant is likely to be quite
receptive to a temperature change. To a first approximation,
extensive intermolecular aggregation of 1 at low temperatures
can be eliminated when compared to the behavior of model
compound 4.
of different conformers of 1 in rapid exchange on the NMR
time scale or (b) the exclusive presence of a C3V symmetrical
conformer (as shown in Figure 1). To examine these possibili-
1
ties, variable temperature H NMR spectra were recorded for
the dynamic containers 1 and 3 and the model compound 4 in
CDCl3. In the case of 1 and 3, as the CDCl3 solution was cooled
from 323 to 223 K, notable changes in the chemical shifts and
broadening of the signals for the Hg, Hh, and Hi protons as well
as He,f were observed (see Figure 1 for the hydrogen assign-
ments), while the signals for the other protons were unaffected.
Furthermore, the appearance of new resonances, or the coales-
cence of existing ones, did not occur for 1 and 3. In the case of
the model compound 4, however, no changes were observed
when its CDCl3 solution was cooled across the same temperature
range. Since the protons of the aromatic gates, present in both
1 and 3, experienced chemical shift and line broadening at lower
temperatures, the spectral results are commensurate with a
change in the conformational distribution of the aryl gate that
is occurring rapidly on the NMR time scale. Inspecting the O-H
resonance shifts of 1, 3, and 4 with temperature allowed us to
calculate the corresponding temperature-dependent, chemical-
shift coefficients ∆δ/∆T for the 223 to 323 K interval as shown
in Figure 4. The magnitude of this value provides qualitative
information about the structure of the molecule of interest.18
Namely, as the hydroxyl proton is rapidly exchanging among
its hydrogen-bonded and non-hydrogen-bonded conformations,
the observed chemical-shift signal represents a weighted average
of the signals for the same proton in its alternative environments;
Infrared Spectroscopy and Theoretical Calculations. The
conformational behavior of 1 and 4 was also thoroughly
examined using infrared spectroscopy in combination with
theoretical calculations. First, let us analyze the model compound
4. As shown in Figure 5a, the FT-IR spectra of 0.4 to 4.8 mM
CHCl3 solutions of the model compound 4 revealed two O-H
stretching vibrations, one strong at 3597 cm-1 and one weak at
3466 cm-1, corresponding to free and hydrogen-bonded O-H
groups, respectively. HF/6-31G* calculations19 of the optimized
geometries and the resultant harmonic vibrational frequencies
of monomeric and dimeric forms of 4 confirmed this assignment
(Figure 5c). Moreover, the experimental spectrum was simulated
by a curve-fitting routine (IGOR20) using two Gaussian functions
centered at the calculated peak positions. In this simulation, the
broadening (fwhm) of each peak was kept the same for both
calculated peaks so as to simulate the experimental spectrum
of the monomeric and dimeric 4. The resulting simulated IR
spectrum clearly reproduced the features of its experimentally
recorded spectrum, and from this analysis, we determined that
the 4.8 mM CHCl3 spectrum of 4 was composed of the
monomeric (89%) and dimeric (11%) forms. Furthermore, this
(16) Cohen, Y.; Avram, L.; Frish, L. Angew. Chem., Int. Ed. 2005, 44, 520-
554.
(17) Cabrita, E. J.; Berger, S. Magn. Reson. Chem. 2002, 40, S122-127.
(18) For instance, see: (a) Gellman, S. H.; Adams, B. R.; Dado, G. P. J. Am.
Chem. Soc. 1990, 112, 460-461. (b) Winningham, M. J.; Sogah, D.; Sogah,
D. Y. J. Am. Chem. Soc. 1994, 116, 11173-11174.
(19) (a) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab initio
Molecular Orbital Theory; John Wiley & Sons: New York, 1986. (b)
Frisch, M. J. et al. Gaussian 03, revision C.02; Gaussian, Inc.: Wallingford,
CT, 2004.
(20) 1988-2004 WaveMetrics, Inc. version: 5.0.2.0. More information can be
found at http://www.wavemetrics.com.
9
J. AM. CHEM. SOC. VOL. 128, NO. 17, 2006 5891

A R T I C L E S
Maslak et al.
Figure 6. A series of partial 1H NMR spectra of 1 (400 MHz, 298 K)
recorded on addition of Et3N (a-l) and CF3CO2H (m-o) to a 1.0 mM
CDCl3 solution; (a) 0, (b) 0.24, (c) 0.48, (d) 0.96, (e) 1.44, (f) 1.92, (g)
2.64, (h) 3.36, (i) 4.32, (j) 5.52, (k) 7.92, (l) 9.12 mol equiv of Et3N, and
(m) 3.90, (n) 7.80, (o) 11.70 mol equiv of CF3CO2H.
local energy minima, as shown in Figure 1. Predicting the
hydrogen-bonded mediated folding pattern of small molecules
with molecular mechanics is indeed debatable and has already
been addressed in parametrizing various force fields.22 There-
fore, we further optimized each of the four conformers using
both semiempirical PM323 and ab initio Hartee-Fock (HF/6-
31G*) methods19 to (1) calculate their relative free energies,
(2) estimate their Boltzmann population distribution, and (3)
calculate harmonic vibrational frequencies so as to simulate their
composite infrared spectra. Based on the free energies (Figure
1), conformer 1a having all three phenols involved in a cyclic
array of the intramolecular hydrogen bonding is the most stable
and largely populates the equilibrium, 94%. Conformer 1b,
having three phenols interacting intramolecularly but forming
a linear array of only two hydrogen bonds, is less stable and
populates 3.5% of the equilibrium. The partially open 1d (one
H-bond) and fully open 1c (no H-bonds) conformers are
energetically unfavored and contribute only 2.5% to the equi-
librium. With this information in hand, we computed a
composite IR spectrum of a fraction-weighted mixture of the
conformers of 1, using computer-generated Gaussian-broadened
functions for each O-H vibrational frequency for each con-
former, Figure 5c. The simulated spectrum of 1 does indeed
reproduce the experimental spectrum in its features. All three
IR bands are visible, although shifted in wavenumber and with
the off scale relative intensities. A better agreement between
the theoretical and experimental results is possible if (a) the
solvation is accounted for in the calculations and (b) a higher
level of theory could be used.24 In this context, the νO-H signal
at 3402 cm-1 can be ascribed to intramolecularly hydrogen-
bonded O-H groups in 1a. The νO-H vibrations at 3581 and
3460 cm-1 are in fact a composite of three O-H vibrations
Figure 5. O-H stretch region of infrared spectra for (a) 4.8, 2.0, 1.0, and
0.4 mM CHCl3 solution of 4 and (b) 4.4, 2.1, 0.6, and 0.2 mM CHCl3
solution of 1. The spectra were obtained after subtraction of the spectrum
of pure CHCl3. (c) Calculated IR spectrum (Hartee-Fock, HF/6-31G*) of
a fraction-weighted mixture of free and dimeric 4 (dashed line). Calculated
IR spectrum of a fraction-weighted mixture of the conformers of 1 (solid
line). The standard free energies for each conformer of 1 were calculated
using Hartee-Fock theoretical method (HF/6-31G*). Vibrational frequencies
were calculated using the semiempirical PM3 theoretical method. Note that
the simulated spectra are deliberately shifted, 160 cm-1 for 4 and 245 cm-1
for 1, to visually overlap the experimentally obtained results.
analysis indicated a low degree of intermolecular association
which is in accord with the prior experiments.
The FT-IR spectra of 0.2 to 4.4 mM CHCl3 solutions of 1
revealed three concentration-invariant O-H stretching vibra-
tions, two strong and broad bands at 3581 and 3402 cm-1 and
a shoulder at 3460 cm-1 (Figure 5b). The signal at 3581 cm-1
in 1 is at a slightly lower wavenumber than the free O-H
stretching vibration at 3697 cm-1 in the model compound 4.
The signal at 3402 cm-1, however, may well be due to
intramolecularly associated O-H groups. In parallel with the
1H NMR spectroscopic measurements, these results point to the
existence of a conformational equilibrium in 1. Which conform-
ers then populate the equilibrium state of 1? As implemented
in MacroModel,21 the results of molecular mechanics (MMFF)
calculations suggested 1a as the most favorable conformation,
and with three additional conformers, 1b, 1c, and 1d, populating
(21) (a) Osawa, E.; Musso, H. Angew. Chem., Int. Ed. Engl. 1983, 22, 1. (b)
Halgren, T. A. J. Am. Chem. Soc. 1992, 114, 7827-7843. (c) Halgren, T.
A. J. Comput. Chem. 1996, 17, 616-641.
(22) (a) Gellman, S. H.; Dado, G. P. Tetrahedron Lett. 1991, 50, 7377-7380.
(b) McDonald, D. Q.; Still, W. C. Tetrahedron Lett. 1992, 33, 7743-7747.
(c) Dado G. P.; Gellman, S. H. J. Am. Chem. Soc. 1994, 116, 1054-1062.
(23) Stewart, J. J. P. Semiempirical Molecular Orbital Methods. In ReViews in
Computational Chemistry; Lipkowitz, K. B., Boyd, D. B., Eds.; VCH: New
York, 1990; Vol. 1, p 45.
(24) We have attempted, but unsuccessfully due to the molecular size, to
implement higher level calculations in a reasonable amount of time.
9
5892 J. AM. CHEM. SOC. VOL. 128, NO. 17, 2006

Gated Molecular Baskets
A R T I C L E S
Figure 7. (a) Electrostatic potential surface map of 1 and CHCl3, calculated using the AM1 method within Spartan Software. (b) ORTEP representation of
the structure of the 2‚CHCl3 complex in the solid state. (c) A π-π stacked dimer of 2.
and correspond to three hydroxyl groups in 1b. Comparing the
experimental and simulated IR spectra, it is likely that the
conformers 1b-d contribute more significantly to the equilibrium.
As the conformer 1a that encloses space dominates the equilib-
rium, the designed receptor could operate as intended. We are,
however, already examining various possibilities to completely
bias the conformational balance toward 1a.
supporting the contention that an acid/base input can be used
to reversibly affect the conformational equilibrium in 1 and
related molecules.
Recognition Studies. The egg-shaped cavity of 1 is 136.6
A3 in volume and approximately 10 Å tall and 9 Å wide, Figure
7.26 The interior of the cavity has walls built of electron-deficient
phthalimides. Interestingly, the electrostatic potential map of
the concave interior surface of 1 appears to have negative
electrostatic potentials, suggesting electron-deficient molecules
could be bound as potential guests, Figure 7a.27 The outer
convex surface is, in contrast, electron-deficient. A chloroform
molecule is 74.9 Å3 in volume,26 with both shape and
electrostatic potential that are complementary to the receptor’s
interior. If trapped inside of 1, chloroform would occupy 54.8%
of its interior volume, which is proportionate to the packing
coefficient of liquids and is a good sign for molecular recogni-
tion in the liquid state.26 We have attempted to observe the
encapsulation of various guests including chloroform, at ambient
and lower temperatures, in CDCl3 and CD2Cl2, using 1H NMR
spectroscopy. The highly dynamic nature of 1, in combination
with its limited solubility in solvents which are not capable of
occupying the interior of the cavity, prevented us from observing
and studying the guest exchange in solution. Further studies
are being conducted in our laboratory wherein the fine balance
between the dynamic nature of these newly designed receptors
and thermodynamics of the molecular encapsulation are under
examination. A single-crystal X-ray study of 2, however,
revealed a disordered chloroform molecule positioned inside
the cavitand along its C3 axis, as shown in Figure 7b.28 The
three nonfunctionalized aromatic gates are situated on the
opposite side of the cavity and arranged in a propeller fashion.
The unit cell is comprised of dimers of 2, organized along the
crystallographic c-axis, facing each other using base tris-
bicycloannelated benzenes at a 4.1 Å distance, indicating weak
Acid/Base Titrations. One of the reasons for studying gated
molecular receptors such as 1-3 is to develop molecular devices
that can be acid/base driven. We were therefore motivated to
subject 1 to deprotonation and subsequent reprotonation condi-
tions and demonstrate a reversible cycle. The chemical shifts
for Hg and Hh protons, in the 1H NMR spectrum of 1 are affected
by addition of triethylamine and practically restored by subse-
quent addition of trifluoroacetic acid, Figure 6. Specifically,
addition of Et3N to a CDCl3 solution of 1 resulted in a change
1
in its H NMR spectrum, Figure 6. The resonance for Hg was
shifted upfield from δ ) 7.04 to 6.96 ppm, whereas the
resonance for Hh changed from δ ) 6.64 to 6.56 ppm.
Interestingly, the resonances for other protons stayed unaffected.
A 9-fold excess of Et3N is evidently required to fully complex
all three phenol units in 1. It is therefore likely that hydrogen
bonding interaction [Et3N‚‚‚H-OAr] rather then proton transfer
dominates in our system. This is further vindicated by the fact
that for the proton transfer to occur in related aliphatic amine/
phenol complexes, sufficiently acidic phenols in more polar
solvents are required.25 The completely screened O-H groups
in the complexed 1 are incapable of interacting intramolecularly,
and the receptor gates are therefore prevented from forming a
cyclic array of hydrogen bonds as shown in 1a. Subsequent
addition of a strong acid CF3CO2H practically restored the
1
chemical shifts for Hh and Hg in the H NMR spectrum of 1,
(25) (a) Taft, R. W.; Gurka, D.; Joris, L.; Schleyer, P. Von R.; Rakshys, J. W.
J. Am. Chem. Soc. 1969, 91, 4801-4808. (b) Hudson, R. A.; Scott, R. M.;
Vinogradov, S. N. J. Phys. Chem. 1972, 76, 1989-1993. (c) Ilczyszyn,
M.; Ratajczak, H. J. Mol. Liq. 1995, 67, 125-131. (d) Menger, F. M.;
Barthelemy, P. A. J. Org. Chem. 1996, 61, 2207-2209. (e) Hannon, C.
L.; Bell, D. A.; Kelly-Rowley, A. M.; Cabbel, L. A.; Anslyn, E. V. J.
Phys. Org. Chem. 1997, 10, 396-404.
(26) The inner volume of 1 was calculated following Rebek’s protocol. See:
Mecossi, S.; Rebek, J., Jr. Chem.sEur. J. 1998, 4, 1016-1022.
(27) Kamieth, M.; Kla¨rner, F.-G.; Diederich, F. Angew. Chem., Int. Ed. 1998,
37, 3303-3306.
9
J. AM. CHEM. SOC. VOL. 128, NO. 17, 2006 5893

A R T I C L E S
Maslak et al.
π-π aromatic attraction, Figure 7c. The observed stacking
geometry is generally not preferred in aromatic interactions.32
In the crystal structures of other tris-bicycloannelated benzenes
C-H/π interactions are usually seen.33 We reason that herein
observed stacking could be ascribed to the following: (1)
reduced aromaticity of the central benzenes due to their bond
alternation and distortion imposed by bicyclic annelation34 and
(2) tendency for a high packing coefficient in the crystal35 that
is attained by interdigitation of the CH2 bicyclic bridges.
Conclusions
The research described in this paper is projected toward
developing multivalent molecular receptors that may potentially
operate in a stimuli-responsive fashion to regulate molecular
encapsulation, i.e., guest transport in/out of the host. As a first
step toward that goal, we developed the synthesis of dynamic
molecular containers having a set of functionalized aromatic
rings installed at the rim to assemble on top of the cavitand via
hydrogen bonding. Indeed, experimental and theoretical studies
have, in unison, demonstrated that these compounds have a
tendency to associate intramolecularly and prefer a conformation
whereby all three aromatic rings are involved in a cyclic array
of intramolecular hydrogen bonding, thus forming a fully closed
receptor. The recognition properties have been preliminarily
examined and have evidenced the dynamic nature of the receptor
and its propensity toward hosting small guest molecules. Studies
in our laboratory are now focused on the next challenge, which
is to understand the operation of this and related molecular
devices in controlling the encapsulation of guests.
(28) Crystal data for 2‚CHCl3: C60H39N3O6 + CHCl3, formula weight )
1017.31, trigonal, space group R3h, colorless plate, a ) 16.326(1) Å, c )
32.400(4) Å, T ) 150 K, Z ) 6, R1(F) ) 0.118, and wR2(F2) ) 0.298 on
all the data. The structure was solved by the direct methods procedure in
SHELXS-97.29 Molecule 2 contains a crystallographic three-fold rotation
axis. There is a molecule of CHCl3 disordered on this three-fold axis, with
one chlorine atom and the carbon atom located on this axis and the
remaining two chlorine atoms and one hydrogen atom disordered over the
three sites related by the three-fold rotation. There is another region of
disorder which is located outside of the cavity of the main molecule, and
it is not within bonding range of it. Because the identity of the molecule
in this region cannot be determined, it is modeled as three carbon atoms,
C(1A), C(2A), and C(3A), with each assigned an occupancy factor of 0.5.
Full-matrix least-squares refinements based on F2 were performed in
SHELXL-97,30 as incorporated in the WinGX package.31 The hydrogen
atoms were added to molecule 2 at calculated positions using a riding model
with U(H) ) 1.2*Ueq (attached atom). The final refinement cycle was
based on 2939 intensities and 235 variables. The three largest peaks in the
final difference map, ranging from 1.43 to 0.93 e/Å3, are in the immediate
vicinity of the disordered CHCl3 molecule. The next three largest peaks,
ranging from 0.79 to 0.65 e/Å3, are located in the region of the disordered
“solvent” atoms of C(1A), C(2A), and C(3A). It may be possible that these
large final R factors are the result of insufficient modeling of the disordered
CHCl3 and “solvent” molecules.
(29) Sheldrick, G. M. SHELXS-97; Universitat Go¨ttingen: Go¨ttingen, Germany,
1997.
(30) Sheldrick, G. M. SHELXL-97; Universitat Go¨ttingen: Go¨ttingen, Germany,
1997.
(31) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837-838.
(32) (a) Hunter, C. A.; Sanders, J. K. M. J. Am. Chem. Soc. 1990, 112, 5525-
5534. (b) Hunter, C. A.; Lawson, K. R.; Perkins, J.; Urch, C. J. J. Chem.
Soc., Perkin Trans. 2001, 2, 651-669. (c) Keyer, E. A.; Castellano, R. K.;
Diederich, F. Angew. Chem., Int. Ed. 2003, 42, 1210-1250.
(33) Frank, N. L.; Baldridge, K. K.; Gantzel, P.; Siegel, J. S. Tetrahedron Lett.
1995, 36, 4389-4392 and references therein.
(34) (a) Siegel, J. Angew. Chem., Int. Ed. Engl. 1994, 33, 1721-1723. (b) Buergi,
H.-B.; Baldridge, K. K.; Hardcastle, K.; Frank, N. L.; Gantzel, P.; Siegel,
J. S.; Ziller, J. Angew. Chem., Int. Ed. Engl. 1995, 34, 1454-1456.
Acknowledgment. Financial support of this research was
supported by the Ohio State University. Generous computational
resources were provided by the Ohio Supercomputer Center.
Supporting Information Available: Detailed descriptions of
experimental methods, synthetic procedures, calculations, and
X-ray crystallographic data for 2. Complete list of authors for
ref 19b. This material is available free of charge via the Internet
at http://pubs.acs.org.
JA060534L
(35) Kitaigorodsky, A. I. Molecular Crystals and Molecules; Academic Press:
New York, 1973.
9
5894 J. AM. CHEM. SOC. VOL. 128, NO. 17, 2006