Design, Synthesis, and Evaluation of Bile Acid-Based Molecular Tweezers

Article abstract of DOI:10.1021/jo9607525

A family of bile acid-based molecular tweezers (7-9) has been constructed readily from simple precursors.Binding experiments with various electron deficient aromatic compounds showed that tweezer 8 binds trinitrofluorenone 10e with an association constant of 220 M^-1 in CDCl3.Single-crystal X-ray analysis of compound 8 shows aromatic-aromatic interactions producing a two-dimensional lattice of pyrene units.Tweezer 8 was immobilized on Merrifield resin, and binding studies have shown that these data compare well with those of the solution state studies.

Full text of DOI:10.1021/jo9607525

9494
J . Org. Chem. 1996, 61, 9494-9502
Design , Syn th esis, a n d Eva lu a tion of Bile Acid -Ba sed Molecu la r
Tw eezer s
Lawrence J . D’Souza and Uday Maitra*
Department of Organic Chemistry, Indian Institute of Science, Bangalore 560 012, India
Received April 24, 1996 (Revised Manuscript Received September 30, 1996X)
A family of bile acid-based molecular tweezers (7-9) has been constructed readily from simple
precursors. Binding experiments with various electron deficient aromatic compounds showed that
tweezer 8 binds trinitrofluorenone 10e with an association constant of 220 M-1 in CDCl3. Single-
crystal X-ray analysis of compound 8 shows aromatic-aromatic interactions producing a two-
dimensional lattice of pyrene units. Tweezer 8 was immobilized on Merrifield resin, and binding
studies have shown that these data compare well with those of the solution state studies.
In tr od u ction
properties, which can possibly be tailor-made through
proper and optimized molecular design.
The past decade has witnessed an explosive growth in
research involving various aspects of supramolecular
chemistry.1 A large number of molecular receptors
(including chiral receptors)2-4 of varying sizes, shapes,
and functionalities have been synthesized, and their
interaction with guests has been assessed. The design
of all such receptors is based on the fundamental molec-
ular interactions exhibited by biological systems, par-
ticularly enzymes.5 Since the binding of a substrate by
an enzyme is the first step in catalysis, the primary goal
in molecular recognition research has been toward the
development of selective synthetic receptors. Such stud-
ies have not only advanced our knowledge in understand-
ing fundamental molecular interactions but have also led
to the design of novel molecular devices, including
sensors.6 There has also been considerable interest in
designing molecular units which self-assemble in solution
and in the solid state.7 Many organized solid state
structures are expected to have interesting material
Although diverse types of molecular hosts have been
designed during the past decade, one of the simplest
concepts has been that of a molecular tweezer. First
developed by Whitlock8 and Zimmerman,9 this class of
molecules was shown to form sandwich complexes with
aromatic guests through π-π interaction. Whitlock's
tweezers were flexible, and apart from π-π interaction,
hydrophobic interaction also played an important role in
their tight binding to aromatic (bis-phenol)carboxylates
in water. The tweezers constructed by Zimmerman,
however, were more rigid and showed exceptionally high
association constants with guests such as polynitroaro-
matics and 9-alkylated adenines in chloroform.
Among the variety of molecular scaffolds which have
been employed in supramolecular design, bile acids have
recently attracted the attention of a number of research
groups for their ready availability and unique structural
features.4 The array of hydroxyl groups lining one
surface of bile acids provides opportunities for the at-
tachment of appropriate molecular units to generate
structures with properties determined by the nature and
arrangement of these units. Most of the interesting
chemistry of the bile acids before the 1990s have come
from their ability to form inclusion complexes and clath-
rates with organic guests.10 We have been interested in
building novel structures with bile acids and decided to
utilize the hydroxyl groups to design a series of hosts in
which structural and/or electronic perturbations could be
made easily. Although a variety of host structures could
be envisaged, we decided to construct a series molecular
tweezers as the first step.11
X Abstract published in Advance ACS Abstracts, December 1, 1996.
(1) (a) Lehn, J .-M. Angew. Chem., Int. Ed. Engl. 1990, 29, 1304. (b)
Cram, D. J . Angew. Chem., Int. Ed. Engl. 1991, 30, 1417. (c) Rebek,
J ., J r. Angew. Chem., Int. Ed. Engl. 1990, 29, 245. (d) Frontiers in
Supramolecular Organic Chemistry; Schneider, H.-J ., Durr, H., Eds.;
VCH: Weinheim, 1991 and references therein.
(2) Wilcox, C. S. Chem. Soc. Rev. 1993, 383 and references therein.
(3) (a) Maitra, U.; Bag, B. G. J . Org. Chem. 1994, 59, 6114. (b) Voyer,
N.; Lamothe, J . Tetrahedron 1995, 34, 9241. (c) Kikuchi, Y.; Kobayashi,
K.; Aoyama, Y. J . Am. Chem. Soc. 1992, 114, 1351. (d) Breslow, R.
Science 1982, 218, 532. (e) Schneider, H.-J . Angew. Chem., Int. Ed.
Engl. 1993, 32, 848. (f) Rebek, J ., J r.; Askew, B.; Nemeth, D.; Parris,
K. J . Am. Chem. Soc. 1987, 109, 2432. (g) Garcia-Tellado, F.; Albert,
J .; Hamilton, A. D. J . Chem. Soc., Chem. Commun. 1991, 1761.
(4) For representative examples, see: (a) Davis, A. P. Chem. Soc.
Rev. 1993, 243. (b) Bonar-Law, R. P.; Sanders, J . K. M. J . Am. Chem.
Soc. 1995, 117, 259. (c) Bonar-Law, R. P.; Mackay, L. G.; Walter, C.
J .; Marvaud, V.; Sanders, J . K. M. Pure Appl. Chem. 1994, 66, 803.
(d) Maitra, U.; Balasubramanian, S. J . Chem. Soc., Perkin Trans. 1
1995, 83. (e) Bonar-Law, R. P.; Sanders, J . K. M. J . Chem. Soc., Perkin
Trans. 1 1995, 3085.
(7) Sessler, J . S. New J . Chem. 1991, 15, 153.
(8) (a) Chen, C. W.; Whitlock, H. W. J . Am. Chem. Soc. 1978, 100,
4921. (b) Nedar, K. M.; Whitlock, H. W. J . Am. Chem. Soc. 1990, 112,
7269.
(9) (a) Zimmerman, S. C.; Zeng, Z.; Wu, W.; Reichert, D. E. J . Am.
Chem. Soc. 1991, 113, 183. (b) Zimmerman, S. C.; Wu, W.; Zeng, Z. J .
Am. Chem. Soc. 1991, 113, 196. (c) For further references see:
Zimmerman, S. C. Bioorganic Chemistry Frontiers; Springer-Verlag:
New York, 1991; Vol. 2, p 33. (d) Blacker, A. J .; J azwinski, J .; Lehn,
J .-M. Helv. Chem. Acta 1987, 70, 1.
(10) (a) Biro, R. P.; Chang, H. C.; Tang, T. P.; Shochet, N. R.; Lahav,
M.; Leiserowitz, L. Pure Appl. Chem. 1980, 52, 2693. (b) Giglio, E.
Inclusion Compounds; Atwood, J . L., Davies, J . E. D., MacNicol, D.
D., Eds.; Academic Press: London, 1984; Vol. 2, p 207. (c) Mik, K.;
Kasai, N.; Shibakami, M.; Takemoto, K.; Mujata, M. J . Chem. Soc.,
Chem. Commun. 1991, 1757. (d) Caira, M. R.; Nassimbeni, L. R.; Scott,
J . L. J . Chem. Soc., Chem. Commun. 1993, 612.
(5) (a) Schneider, H.-J . Angew. Chem., Int. Ed. Engl. 1991, 30, 1417.
(b) Fersht, A. Enzyme Structure and Mechanism, 2nd ed.; Freeman
and Co.: New York, 1985; p 293.
(6) (a) Sandanayake, K. R. A. S.; Imazu, S.; J ames, T. D.; Mikami,
M.; Shinkai, S. Chem. Lett. 1995, 139. (b) Zhao, S.; Luong, J . H. T. J .
Chem. Soc., Chem. Commun. 1995, 663. (c) Rojas, M. T.; Kaifer, A. E.
J . Am. Chem. Soc. 1995, 117, 5883. (d) Sandanayake, K. R. A. S.;
Shinkai, S. Angew. Chem., Int. Ed. Engl. 1994, 33, 2207. (e) Fabbrizzi,
L.; Licchelli, M.; Pallavicini, P.; Perotti, A.; Sacchi, D. Angew. Chem.,
Int. Ed. Engl. 1994, 33, 1975. (f) Vance, D. H.; Czarnik, A. W. J . Am.
Chem. Soc. 1994, 116, 9397. (g) Chawla, H. M.; Srinivas, K. J . Chem.
Soc., Chem. Commun. 1994, 2593. (h) Hamasaki, K.; Ikeda, H.;
Nakamura, A.; Ueno, A.; Toda, F.; Suzuki, I.; Osa, T. J . Am. Chem.
Soc. 1993, 115, 5035. (i) Teixidor, F.; Flores, M. A.; Escriche, L.; Vinas,
C.; Casabo, J . J . Chem. Soc., Chem. Commun. 1994, 963. (j) Marsella,
M. J .; Swager, T. M. J . Am. Chem. Soc. 1993, 115, 12214.
(11) Preliminary communication: Maitra, U.; D’Souza, L. J . J .
Chem. Soc., Chem. Commun. 1994, 2793.
S0022-3263(96)00752-9 CCC: $12.00 © 1996 American Chemical Society

Bile Acid-Based Molecular Tweezers
J . Org. Chem., Vol. 61, No. 26, 1996 9495
F igu r e 1. (a) Schematic representations of a bile acid-based molecular tweezer (b) O3-O12 distances in bile acids 1-3, from
PCMODEL.
Objectives
are not exactly parallel; rather they diverge away from
the steroid. This suggested that the attachment of two
large and flat aromatic surfaces to these two hydroxyl
groups will lead to the formation of a deep cleft, thereby
generating a new class of molecular tweezers in which
structural variations could be accomplished in a straight-
forward manner.
The construction of bile acid-based molecular tweezers,
involving the linkage of the rigid bile acid backbone with
rigid π-surfaces, presents a general strategy by which a
variety of semirigid receptors can be synthesized easily.
Since a small degree of flexibility is present in these
tweezers, detailed studies on a family of such molecules
can lead to a better understanding of the geometrical
requirements of binding. In spite of the success of
Zimmerman’s approach, variation of electron density of
the aromatic arms required considerable synthetic effort.
Our modular approach for the construction of semirigid
systems, in which “sticky” arms are attached to a
template through spacers by a simple reaction, is syn-
thetically much straightforward. We initially planned
to synthesize molecular tweezers based on π-π interac-
tions.8,11,12
Bile acids have a rigid backbone: the C-3 and C-12
OH groups are ca. 6-6.5 Å apart, and they are ap-
proximately parallel. We reasoned that these would meet
the geometric requirements for synthesizing bile acid-
based tweezers which can be easily made by attaching
flat π-surfaces such as pyrene units on the 3- and 12-
hydroxyl groups. We also felt that the 7-OH position can
be used subsequently, after appropriate functionalization,
as an additional handle to interact with a bound guest
(Figure 1a). We have used both cholic and deoxycholic
acids to synthesize a family of molecular tweezers, which
are described in the following section.
Syn th esis. 1-Acetylpyrene (5a ) was prepared from
pyrene (4) using Ac2O/AcOH and anhyd ZnCl2 in 83%
yield.14 Compound 5a was oxidized with NaOCl in aq
pyridine to acid 5b in 75% yield, which was converted to
acid chloride 5c using either thionyl chloride or oxalyl
chloride in benzene (Scheme 1).
Methyl deoxycholate 2a was converted to 6 in 62%
yield by careful esterification using Oppenauer conditions
(CaH2, toluene, BnEt3N+Cl-, reflux) with pyrene-1-car-
bonyl chloride (5c) (Scheme 1). With the use of an excess
of 5c, and at a higher temperature, compound 7 was
obtained in 86% yield from methyl deoxycholate 2a . The
C-7 hydroxyl group of 1 was selectively oxidized with
NBS/aq NaHCO3 to the corresponding ketone,15 which
upon esterification with methanolic HCl afforded methyl
7-ketocholate (3a ) in 77% yield. Following the same
esterification procedure as above, compound 8 was
prepared from 3a in 86% yield. The C-7 hydroxy deriva-
tive 9 was prepared in 89% yield by the reduction of the
C-7 keto group of 8 with NaBH4 in MeOH/THF (Scheme
2). A mixture of steroidal ester 6, along with 5d , was
used as a control for all spectroscopic studies, including
association constant measurements.
F lu or escen ce Sp ectr a . Pyrene and its derivatives
show intense fluorescence emission even at very low
concentrations.16 Our systems with two pyrenes placed
at a distance of ca. 6-7 Å clearly are good substrates for
studying their fluorescence properties. As the distance
between the two pyrenes was close enough to sandwich
a guest (which was apparent from InsightII calcula-
Resu lts a n d Discu ssion
Mod elin g Stu d ies. Analysis of the PCMODEL13
-minimized structures of cholic (1), 7-deoxycholic (2), and
7-ketocholic (3) acids revealed that the hydroxyl groups
at the 3- and 12-positions are ca. 5.9-6.2 Å apart (O-O
distance) (Figure 1b). The two C-O vectors, however,
(12) (a) Heinemann, U.; Saenger, W. Nature 1982, 229, 27. (b) Ibid.
Pure Appl. Chem. 1985, 57, 417. (c) Saenger, W. Principles of Nucleic
Acid Structure; Springer-Verlag: New York, 1984; p 132. (d) Dahl, T.
Acta Chem. Scand. 1994, 48, 95. (e) Stynes, D. V. Inorg. Chem. 1994,
33, 5022. (f) Burly, S. K.; Petsko, G. A. Adv. Protein Chem. 1988, 39,
125. (g) Burly, S. K.; Petsko, G. A. Science 1985, 229, 23. (h) Burly, S.
K.; Petsko, G. A. J . Am. Chem. Soc. 1986, 108, 7995.
(13) PCMODEL version 4.0 and PCMODEL for Windows v. 5.13
were purchased from Serena Software, Bloomington. In general the
minimizations were first carried out without hydrogens. Subsequently
hydrogen atoms were put, and the resulting structure was minimized.
The entire operation was repeated several times with different initial
geometries until consistent results were obtained.
(14) Vollmann, H.; Becker, H.; Corell, M.; Streeck, H. Liebigs Ann.
Chem. 1937, 1, 531, 108.
(15) Fieser, L. F.; Rajagopalan, S. J . Am. Chem. Soc. 1950, 72, 5530;
1951, 73, 4133.
(16) (a) Iyoda, T.; Morimoto, M.; Kawasaki, N.; Shimadzu, T. J .
Chem. Soc., Chem. Commun. 1991, 1480. (b) J in, T.; Ichikawa, K.;
Koyama, T. J . Chem. Soc., Chem. Commun. 1992, 499. (c) Aoki, I.;
Sakaki, T.; Shinkai, S. J . Chem. Soc., Chem. Commun. 1992, 730. (d)
Ueno, A.; Suzuki, I.; Osa, T. J . Chem. Soc., Chem. Commun. 1988,
1373. (e) Sandanayake, K. R. A. S.; J ames, T. D.; Shinkai, S. Chem.
Lett. 1995, 503. (f) Aoki, I.; Harada, T.; Sakaki, T.; Kawahara, Y.;
Shinkai, S. J . Chem. Soc., Chem. Commun. 1992, 1341.

9496 J . Org. Chem., Vol. 61, No. 26, 1996
D’Souza and Maitra
Sch em e 1
Sch em e 2
tions),17 fluorescence experiments were expected to give
additional information in support of the proximity of the
pyrenes in our tweezers. A strong intramolecular exci-
mer emission was observed for all the tweezers (1.04 µM
in 3% CHCl3/CH3CN), whereas, with the control run
using 6 and 5d each at 1.04 µM, only monomer emission
was observed (Figure 2). This variation in emission
behavior results in an important noticeable difference:
Hosts 7-9 in CHCl3 appear greenish, while a mixture of
6 and 5d (control) at identical concentrations appears
bluish when held in sunlight. These experiments sug-
gested that the long-wave emission resulted from the
interaction of the pyrene units from the same steroid unit
and not through intermolecular interaction.
F lu or escen ce Sp ectr a in th e P r esen ce of a Gu est.
Synergistic interaction between the two pyrenes in the
hosts was further evident from fluorescence-quenching
experiments. Figure 3 shows the fluorescence spectrum
of host 7 (0.67 mM) with octafluoronaphthalene (10b; 33
mM) in CHCl3. In this case the guest fluorescence was
completely quenched in the presence of the host, and
there was an increase in the monomer emission; however,
there was no significant change in the intensity of
excimer emission. This provides evidence that the guest
sandwiches between the two pyrenes thereby increasing
the monomer emission. Figure 3 also shows the fluores-
(17) InsightII minimization calculations showed that all three
tweezers have a distance of separation of 5-6 Å (data not shown).

Bile Acid-Based Molecular Tweezers
J . Org. Chem., Vol. 61, No. 26, 1996 9497
moieties of compounds 7-9 did not interact in the ground
state, since the spectral pattern of these compounds
roughly matched the absorption of the control (i.e., a
mixture of 6 and 5d ). When a CHCl3 solution of any of
the hosts was mixed with trinitrofluorenone 10e, trini-
trobenzene 10d , and 3,5-dinitrobenzonitrile (10c), deep
red, red, and yellowish-red solutions were obtained,
respectively. Tweezer 8 in the presence of trinitrofluo-
renone 10e as the guest showed a well-defined charge-
transfer band, as observed in the difference spectrum.18
These observations support the existence of donor-
acceptor interactions in our systems.
NMR Titr a tion Exp er im en ts. Encouraged by the
modeling work and fluorescence/UV experiments, we
proceeded to use NMR spectroscopy as the tool to assess
the binding strengths of our tweezers with different
guests in CDCl3. Before detailed NMR titration studies
were initiated with 7-9, an important control experiment
was done. A mixture of 5d and 6 (each at 0.031 M in
CDCl3) and 1,3,5-trinitrobenzene (10d ; 0.06 M) was
prepared in CDCl3, and the NMR spectrum was recorded.
This spectrum, when compared with that of a mixture
of 5d and 6, showed that under these conditions there
were no significant upfield shifts of the aromatic signals,
suggesting the lack of any interaction under these
conditions. Interestingly, NMR titration experiments with
trinitrofluorenone 10e with the controls under conditions
as stated above could not be carried out because of the
poor solubility of the former in CDCl3. But this 'problem'
was not encountered with tweezers 7-9 since all the
guest (up to a maximum of 1.5 equiv) went into solution.
This observation confirms that the hosts indeed increased
the solubility of the guest (upon binding).
F igu r e 2. Fluorescence spectra (λex ) 355 nm) in 3% chloro-
form-acetonitrile: (a) 7, (b) 8, (c) 9, and (d) 6, each at 1.04
µmol dm-3, and (e) 5d plus 6, each at 1.04 µmol dm-3
.
All NMR titration experiments were performed in
CDCl3 at 25 °C. Guests were usually titrated up to 70%
saturation limits. Almost all aromatic 1H NMR reso-
nances of the hosts moved upfield (by varying degrees),
but the aromatic region between δ 7 and 8.3 was quite
complicated for analysis. However, one of the pyrene
doublets (δ 8.91 in host 8) moved considerably upfield
and was well-separated from all other resonances, and
this signal was monitored during the titration. The
chemical shifts (∆δ) were analyzed using a nonlinear
curve-fitting program.19 All experiments were repeated
at least twice. When an NMR titration experiment was
carried on methyl pyrene-1-carboxylate (5d ) with guest
10c, the estimated association constant was <5 M-1
.
Thus, the order of magnitude difference with that of the
host (Ka ) 43 M-1) was completely due to the presence
of the other pyrene in a preorganized manner (Table 1).
A variety of other electroneutral aromatics (pyrene,
fluorene, anthracene, naphthalene, phenanthrene etc) or
electron-rich aromatics (phenol, anisole, 1,4-dimethoxy-
benzene, etc.) showed no detectable binding.
Binding measurements were also attempted in other
solvent systems, such as DMSO-d6 and acetone-d6. When
compared to CDCl3, similar shifts of the aromatic protons
of the host 7 in DMSO-d6 in the presence of guest 10b
were observed, but this study could not be pursued
further because of solubility problems. The experiments
F igu r e 3. Fluorescence spectra (λex ) 355 nm) in chloro-
form: (a) 7 at 0.67 mmol dm-3, (b) 7 at 0.67 mmol dm-3 plus
10b at 33 mmol dm-3, (c) 10b at 33 mmol dm-3, (d) 5d plus 6,
each at 0.67 mmol dm-3, and (e) 5d plus 6, each at 0.67 mmol
dm-3 plus 10b at 33 mmol dm-3
.
cence spectra for the controls (each at identical concen-
tration as 7) in the presence of guest 10b. Although the
fluorescence of the guest was quenched, there was no
significant increase in the monomer emission. This
clearly shows that the controls act as independent
entities, and the quenching observed is due to the
stacking of donor-acceptor groups which is a predomi-
nant interaction in organic solvents.
(18) Spectrum of (8 + 10e together in solution) - (computer sum of
the spectra of 8 and 10e taken separately). A similar difference
spectrum was recorded with the controls.
(19) Macomber, R. S. J . Chem. Ed. 1992, 69, 375. The reliability of
our data was assessed by the incorporation of the rms deviation of ∆δ
in the original program. rms deviation of <0.005 ppm indicated good
fit of the data.
Absor p tion Sp ectr a . UV-visible spectroscopic stud-
ies on these compounds indicated that the two pyrene

9498 J . Org. Chem., Vol. 61, No. 26, 1996
D’Souza and Maitra
Ta ble 1. Associa tion Con sta n ts in CDCl3 a t 25 °Ca
guests, Ka (dm3 mol-1
)
host
10a
10b
10c
10d
10e
10f
9
7
8
6
7
8
11
15
19
27
35
43
47
65
83
155
189
220
nd
nd
171
a
nd: not determined.
done in acetone-d6 with 7 and 10b did not show any
change in the chemical shifts, either of the host or of the
guest.20
F igu r e 4. X-ray structure of 8. Labeling diagram showing
molecule and solvent. Hydrogen atoms are numbered accord-
ing to the atom to which they are attached but are omitted
from the figure for clarity. See ref 23 for information on the
solvent molecule in the crystal.
According to the donor-acceptor concept, more electron
deficient is a group, tighter is the binding.21 Our systems
did not show this trend, as the Ka for tetranitrofluorenone
10f was found to be less than that for trinitrofluorenone
10e. These results possibly indicate that in addition to
the interaction forces, the cleft dimension may also play
an important role in binding.
NMR titration results mentioned earlier were analyzed
by the curve-fitting method assuming a 1:1 stoichiometry
of the complex formed between the host and the guest.
This was proved by a separate titration of host 8 with
guest 10e to get a J ob plot which showed a maximum at
a mole ratio of 1.0.22
Cr ysta l Str u ctu r e of 8. In spite of repeated at-
tempts, we were unable to grow crystals of host-guest
complexes. Host 8, however, was crystallized from a
mixture of ethyl acetate and hexanes to yield needle-
shaped crystals suitable for X-ray analysis.
F igu r e 5. Stereoview of the X-ray structure of 8 (packing
diagram). View down the b axis showing one of the aliphatic-
aromatic-aliphatic bilayers in the structure. The packing of
1
the aromatic-aromatic rings are centered on z ) 0, /2, 1, etc.
are almost parallel to each other and to the c axis.24 The
primary interaction is that of pyrene rings interacting
in a “herringbone”25 fashion (edge to face) around the
planes at z ) 0, 1/2, 1, etc., with rings entering the
interaction region from aliphatic backbones located in the
positive and negative c direction alternately. Each
pyrene unit is interacting with the adjacent four pyrenes
with a edge to face manner.
The two-dimensional supramolecular organization of
the pyrene moieties suggests that the bile acid backbone
may possibly be used to organize other flat aromatic units
to create a specific arrangement of such systems.26 Such
organization will allow one to orient aromatic residues
in a predictable way. It will also be of interest to
engineer such crystals and study their material proper-
ties, particularly for host-guest types of systems.
The O3-O12 distance in the solid state was found to
be 5.78 Å (Figure 4). Although the distance is somewhat
less than the ideal value, we believe that binding occurs
in solution because of the flexibility of the pyrene esters
to self-adjust the cleft dimensions in solution.
The packing diagram from the X-ray structure shown
in Figure 5 consists of molecules of the compound packed
in the unit cell.23 There are no abnormally close contacts
between the molecules. The distances and angles within
the molecule appear to be normal. The two pyrene planes
(20) At this time we are not sure of the reasons for such solvent
effects.
(21) Otsuki, J .; Chiang, L.-C.; Lee, S.-H.; Araki, K.; Seno, M.
Supramol. Chem. 1993, 2, 25.
(22) An attempted “intracomplex” NOE measurement between 8 and
10e was not successful.
(23) Given the crystallization conditions, the composition of this
solvent molecule almost certainly varies from crystal to crystal. The
solvent in this case seems to be a mixture of about 65% ethyl acetate
and 35% n-hexane. The fractional occupancy of atom O-9 is 0.65 and
of atom C-105 is 0.35; O-8 is given full occupancy as an oxygen atom
even though it should theoretically be 65% O and 35% C. There might
also be some contribution from isohexane, which has the same basic
shape as ethyl acetate. The empirical formula is based on full
occupancy by ethyl acetate.
(24) (a) Desiraju, G. R.; Gavezzotti, A. J . Chem. Soc., Chem.
Commun. 1989, 621. (b) Kai, Y.; Hama, F.; Yasuoka, N.; Kasai, N.
Acta Crystallogr. 1978, B34, 1263. (c) Munakata, M.; Dai, T.; Maekawa,
M.; Takayoshi K.-S.; Fukui, J .-T. J . Chem. Soc., Chem. Commun. 1994,
2331.
(25) Burley, S. K.; Petsko, G. A. J . Am. Chem. Soc. 1986, 108, 7995.
(26) Otsuki, J .; Oya, T.; Lee, S.-H.; Ariki, K. J . Chem. Soc., Chem.
Commun. 1995, 2193.

Bile Acid-Based Molecular Tweezers
J . Org. Chem., Vol. 61, No. 26, 1996 9499
Ta ble 2. Gu est Bin d in g by 12 a n d 13
Sch em e 3
% guest bound
guests
12
13
10d
10e
10f
52
75
68
8.2
37
37
the magnitude of binding with respect to that of host is
important, which means that the pyrene units in 13 are
acting cooperatively. The polymer-bound control 13 did
not bind trinitrobenzene 10d efficiently.
The results obtained with 10e,f are interesting, and it
was found that 10e binds to tweezer 12 more tightly than
10f. The relative degree of “removal” of the three guests
follows their relative association constants observed in
CDCl3 by NMR titration experiments. Estimated binding
constant from HPLC data for 12 and 10e was 97 M-1
.
Con clu sion
We have been able to readily construct a family of bile
acid-based molecular tweezers which bind polynitroaro-
matic compounds with moderate association constants
in CDCl3. Analysis of the solid state structure of 8
indicated that this structure represents a novel organiza-
tion of the bile acid backbone which has not yet been
observed in other bile acid crystals. Binding of the guests
with a polymer-bound host was demonstrated, and this
is likely to be of practical interest when more efficient
hosts are developed.
We believe that our simple approach toward the design
of steroidal molecular tweezers has provided sufficiently
interesting results to warrant further investigations, and
our attractive bile acid-based molecular tweezers are
likely to generate more interests, both academically as
well as in applied areas.
Im m obiliza tion of Tw eezer 8. After demonstrating
the binding of polynitro aromatic hydrocarbons (PAH’s)
by the steroidal tweezers, it was of interest to further
exploit and develop these compounds as immobilized
hosts.27 The side chain of the steroid provides a built-in
handle for attaching the tweezers to a polymeric support.
Compound 8a was obtained from 8 by treating it with
LiOH/THF/MeOH in 89% yield. This acid was converted
into its cesium salt using aq Cs2CO3/THF (Scheme 3).
The salt obtained was directly reacted with 1% cross-
linked chloromethylated polystyrene 11 in dry DMF at
80-90 °C for 24 h to yield the polymer-bound host 12,
which was washed thoroughly with DMF, water, and
EtOH to remove any unreacted starting materials present.
Similarly, the same reactions done on pyrene-1-carboxylic
acid (5b) yielded 13 which was used as the control.
The pyrene loading in the polymer was estimated by
hydrolyzing known weights of 12 and 13 using NaOMe/
MeOH/H2O and weighing the pyrene-1-carboxylic acid.
It was found that 12 contained 0.54 mmol of steroid
units/g of functionalized polymer, and the control (13)
contained 0.83 mmol of pyrene units/g of the function-
alized polymer.
We examined the interaction of the polymer-bound
tweezer with the same guests (viz., 10d -f), which were
used for solution studies as follows. A solution of the
three guests (0.5 mL of 1 × 10-4 M each) in chloroform
was stirred with a known weight of the polymer-bound
host and the control for 24 h, taking care to prevent the
evaporation of the solvent. After filtration, a known
volume of the filtrate was analyzed by HPLC (85:15
EtOAc/hexanes), and the areas under the three guest
peaks were measured. The original guest solution (10-
20 µL) was also analyzed by HPLC under identical
conditions. The decrease in any peak area relative to the
original guest solution implies its “removal” by the
polymer. The percentage of guest bound to 12 and 13 is
shown in Table 2.28
Exp er im en ta l Section
Gen er a l. All reactions were conducted under dry nitrogen
and stirred magnetically unless otherwise stated. Cholic acid,
7-deoxycholic acid, and pyrene were purchased from Fluka.
The polynitroaromatic compounds were commercial samples
or prepared according to literature procedures. All solvents
were purified and distilled before use.29 Toluene, benzene, and
tetrahydrofuran were distilled from sodium/benzophenone
ketyl. Methanol was distilled from magnesium methoxide.
Thin layer chromatography was performed on precoated plates
(silica gel 60F-254) purchased from Sigma. These plates were
stained either with iodine vapor or with 5-10% phosphomo-
lybdic acid in ethanol. Usually purification of the products
was done using gravity columns. Melting points were recorded
in open capillaries and are uncorrected. Proton NMR spectra
were recorded on 90, 200, 270, and 400 MHz spectrometers.
1
Unless otherwise stated H NMR spectra were taken in CDCl3
using CHCl3 as the internal standard (δ ) 7.270). For 13C
NMR spectra the peak at 77.0 ppm arising from CDCl3 was
used as the internal reference. All chemical shift values shown
are in δ scales. Multiplicity of NMR signals is designated as
s (singlet), d (doublet), t (triplet), etc. Broad unresolved lines
are designated as br m (broad multiplet).
Optical rotations were measured in appropriate solvents
using sodium D light. Microanalyses were done on an auto-
mated CHNS analyzer. LR mass spectral data are given as
m/z (% abundance). FAB mass spectra were recorded using
argon/xenon (6 kV, 10 mA) as the FAB gas using m-nitrobenzyl
It is interesting to see that even the controls were
removing the nitroaromatics to a small extent (practically
no binding was detected by NMR, vide infra). However,
(27) (a) Zimmerman, S. C.; Saionz, K. W. J . Am. Chem. Soc. 1995,
117, 1175. (b) Smith, P. W.; Chang, G.; Still, W. C. J . Org. Chem. 1988,
53, 1587. (c) Borchardt, A.; Still, W. C. J . Am. Chem. Soc. 1994, 16,
373. (d) Yoon, S. S.; Still, W. C. J . Am. Chem. Soc. 1993, 115, 823.
(28) No significant (>5% loss of signal) binding was observed with
resin 11.
(29) (a) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals, 3rd ed.; Pergamon Press: New York, 1988. (b) Vogel, A. I.
Textbook of Practical Organic Chemistry, 4th ed.; Longman Group,
1978.

9500 J . Org. Chem., Vol. 61, No. 26, 1996
D’Souza and Maitra
alcohol as a matrix. Infrared spectra were taken in CHCl3,
as a thin film on NaCl plates (neat),30 or in Nujol. HPLC
analysis was carried out on either (a) an ODS column (25 cm
× 4.6 mm i.d.) using H2O/MeOH or (b) a silica gel column using
EtOAC/n-hexane as the mobile phase. Samples were detected
using a variable wavelength UV detector and the resulting
data analyzed on a data processor.
Meth yl 7-Deoxych ola te (2a ). This was prepared using
deoxycholic acid and methanolic HCl following a literature
procedure in 90% yield.31
Met h yl 3r-((1-P yr en oyl)oxy)-12r-h yd r oxy-5â-ch ola n -
24-oa te (6). To a stirred solution of methyl deoxycholate (0.3
g, 0.737 mmol) in dry toluene (4 mL) were added CaH2 (0.046
g, 1.1 mmol), BnEt3N+Cl- (0.041 g, 0.18 mmol), and pyrene-
1-carbonyl chloride (0.254 g, 0.96 mmol). This mixture was
heated in an oil bath at 60 °C for 14 h. The mixture was
filtered through a bed of Celite, and the residue was washed
with ethyl acetate. The combined organic layer was washed
with 7% NaHCO3 solution, water, and brine, finally dried over
anhyd Na2SO4, and filtered. The solvent was removed in vacuo
to yield the crude product, which was purified by column
chromatography on silica gel (100-200 mesh, 13 g, 23 × 1.4
cm) using 10-20% EtOAc/hexanes as the eluent. The pure
product was crystallized from EtOAc/hexane (1:1) and weighed
0.25 g (62%): mp 200-201 °C; IR (neat) 3010 (m), 2940 (s),
1710 (s), 1690 (m), 1440 (s), 840 (s), 950 (s) cm-1; [R]D21 +53 (c
5.34, CHCl3); UV (λmax, log ꢀ) (6% CHCl3/CH3CN, v/v) 350
(4.31), 280 (4.36), 243 (4.61); 1H NMR (CDCl3, 270 MHz) δ 0.75
(s, 3H), 0.99 (m, 6H), 1.01-2.40 (m, steroidal CH and CH2),
3.60 (s, 3H), 4.02 (s, 1H), 5.19 (br m, 1H), 8.28-8.03 (m, 7H),
8.63 (d, 1H, J ) 9.5 Hz), 9.24 (d, 1H, J ) 9.5 Hz); 13C NMR
(CDCl3, 100.5 MHz) δ 12.5, 17.1, 23.0, 23.4, 25.9, 26.7, 26.8,
27.2, 28.6, 29.5, 30.7, 30.8, 32.3, 33.6, 34.0, 34.8, 41.8, 46.3,
47.1, 48.1, 51.3, 72.9, 75.0, 123.9, 124.0, 124.2, 124.7, 125.8,
125.9, 126.0, 127.0, 128.1, 129.1, 129.2, 130.2, 130.7, 130.8,
133.9, 167.4, 174.4; MS: 634 (M+), 291, 246 (100), 57 (100),
43; HRMS calcd for C42H50O5 634.3658, found 634.3651.
Meth yl 3r,12r-Bis((1-p yr en oyl)oxy)-5â-ch ola n -24-oa te
(7). To a stirred solution of methyl deoxycholate (0.40 g, 0.98
mmol) in dry toluene (5 mL) were added CaH2 (0.124 g, 2.95
mmol), BnEt3N+Cl- (0.055 g, 0.24 mmol), and pyrene-1-
carbonyl chloride (0.573 g, 2.16 mmol). The resulting mixture
was heated in an oil bath at 110 °C for 24 h. The reaction
mixture was filtered through a pad of Celite which was
subsequently washed with ethyl acetate. The combined
organic layer was washed with 7% NaHCO3 solution, water,
and brine, finally dried over anhyd Na2SO4, and filtered. The
solvent was removed in vacuo to yield the crude product which
was purified on silica gel (100-200 mesh, 13 g, 13 cm × 1.8
cm) using 10-20% EtOAc/hexanes as the eluent. The pure
product was crystallized from EtOAc/hexane (1:1) and weighed
Met h yl 3r,12r-Dih yd r oxy-7-k et o-5â-ch ola n -24-oa t e
(3a ).32 To a warm (ca. 60 °C) mixture of NaHCO3 (2.5 g, 29.8
mmol) in water (80 mL) was added cholic acid (2.5 g, 6.12
mmol) till it dissolved completely. The solution was cooled to
about 15 °C, and N-bromosuccinimide (2.18 g, 12 mmol) was
added slowly over a period of 30 min. The mixture was stirred
for 14 h at rt when the NBS dissolved completely. It was then
heated in an oil bath (80-100 °C) for 1 h. The solution was
cooled and acidified with 5 M HCl. The white precipitate was
filtered off, washed thoroughly with water, and dried under
high vacuum. The crude product (2.4 g) was dissolved in dry
methanol (15 mL), and 2% HCl in CH3OH (4 mL) was added.
The mixture was stirred at rt for 22 h. Volatiles were removed
under reduced pressure, and the crude product was purified
by column chromatography (silica gel 100-200 mesh, 42 g,
12 cm × 3 cm) using 50-70% EtOAc/hexanes as the eluent.
The ester was isolated in 77% yield (1.91 g): IR (neat) 3390
1
(s), 1710 (s), 1440 (m), 1254 (m), 1060 (m) cm-1; H NMR (90
MHz, CDCl3) δ 0.69 (s, 3H), 1.01 (d, 3H, J ) 6.4 Hz), 1.19 (s,
3H), 1.3-2.5 (m, steroidal CH and CH2), 2.90 (m, 1H), 3.64
(br m, 1H), 3.70 (s, 3H), 4.00 (s, 1H).
1-Acetylp yr en e (5a ). A homogeneous solution of pyrene
(2 g, 9.9 mmol) in acetic anhydride (7.5 mL, 73.5 mmol) was
added to a solution of zinc chloride (2.5 g, 18 mmol) in acetic
acid (7.5 mL, 125 mmol) at 85 °C. The mixture was stirred at
80-90 °C for 1 h, and the resulting thick red precipitate was
filtered off. This complex was decomposed by adding crushed
ice and was left in the refrigerator for 4 h. The clear
supernatant was filtered, and the residue was dried under high
vacuum and then crystallized from hot MeOH. The pure
product was obtained in 83% yield (2 g): mp 88-89 °C (lit.14
mp 90 °C).
P yr en e-1-ca r boxylic Acid (5b). A mixture of 1-acetyl-
pyrene (5a ) (2.3 g, 9.38 mmol) and pyridine (1.95 mL, 24.4
mmol) was stirred at 85-90 °C. After 5 min 4% NaOCl
solution (70 mL, 37.6 mmol) was added, and the mixture was
heated at the same temperature for 1 h. Volatiles were
distilled off; the mixture was concentrated and filtered. The
white voluminous precipitate was dissolved in hot water (150
mL) and filtered through a fluted filter paper. The filtrate
was cooled and acidified with dilute HCl, and the pale yellow
mass was filtered, dried, and sublimed at reduced pressure.
The pure product was obtained in 75% yield (1.73 g) and
melted at 260-261 °C (lit.14 mp 261-263 °C).
Meth yl P yr en e-1-ca r boxyla te (5d ). Methyl ester of
pyrene-1-carboxylic acid was prepared using 5b and MeOH/
concd H2SO4. The crude product was crystallized from
MeOH: mp 75-76 °C; UV (λmax, log ꢀ) (3% CHCl3/CH3CN, v/v)
350 (4.33), 280 (4.42), 243 (4.64); 1H NMR (90 MHz, CDCl3) δ
4.15 (s, 3H), 8.0-8.35 (m, 7H), 8.61 (d, 1H, J ) 8.1 Hz), 9.24
(d, 1H, J ) 9.9 Hz).
P yr en e-1-ca r bon yl Ch lor id e (5c). Pyrene-1-carboxylic
acid (5b) (1 mmol) was refluxed with an excess of thionyl
chloride (3-4 mmol) for 3 h at 80 °C. Excess thionyl chloride
and volatiles were distilled off, and the residue was dried
under reduced pressure for several hours. The crude acid
chloride was used directly for all esterification reactions.
0.73 g (86% yield): mp 183-185 °C; [R]21 +137 (c 1.90,
D
CHCl3); UV (λmax, log ꢀ) (6% CHCl3/CH3CN, v/v) 349 (4.63),
279 (4.70), 243 (4.93); IR (neat) 2910 (s), 1720 (s), 1695 (s),
1
1440 (s), 1250 (s), 840 (s), 750 (s) cm-1; H NMR (CDCl3, 400
MHz) δ 0.95 (s, 3H), 1.0 (d, 3H, J ) 6.5 Hz), 1.07 (s, 3H), 1.14-
2.43 (m, steroidal CH and CH2), 3.61 (s, 3H), 5.15 (br m, 1H),
5.63 (s, 1H), 7.47 (d, 1H, J ) 8.1 Hz), 7.60 (t, 1H, J ) 7.6 Hz),
7.68 (d, 1H, J ) 8.9), 7.75 (d, 1H, J ) 7.5 Hz), 7.81 (d, 1H, J
) 9.4 Hz), 7.89 (t, 2H, J ) 8.8 Hz), 7.97 (d, 2H, J ) 4.8 Hz),
8.04 (d, 1H, J ) 8.1 Hz), 8.07 (d, 1H, J ) 8.1), 8.12 (d, 1H, J
) 7.0), 8.14 (d, 1H, J ) 7.5 Hz), 8.20 (t, 2H, J ) 4.4 Hz), 8.58
(d, 1H, J ) 8.0 Hz), 8.95 (d, 1H, J ) 9.4 Hz), 9.21 (d, 1H, J )
9.4 Hz); 13C NMR (CDCl3, 100.5 MHz) δ 12.7, 17.8, 23.1, 23.6,
25.8, 26.3, 26.9, 27.4, 29.8, 30.9, 31.1, 32.5, 34.3, 34.9, 35.1,
35.9, 41.9, 45.6, 48.0, 50.1, 51.5, 74.9, 76.8, 123.9, 124.1, 124.4,
124.7, 125.4, 126.0, 126.1, 126.2, 127.0, 127.1, 127.2, 127.7,
128.0, 129.0, 129.3, 129.3, 129,4, 130.3, 130.5, 130.6, 130.8,
131.0, 133.9, 133.9, 167.7, 168.0, 174.6; MS 863 (M+, 16), 220,
154 (100); HRMS calcd for C59H58O6 862.4233, found 862.4226.
Meth yl 3r,12r-Bis((1-p yr en oyl)oxy)-7-k eto-5â-ch ola n -
24-oa te (8). To a stirred solution of methyl 3R,12R-dihydroxy-
7-keto-5â-cholan-24-oate (0.216 g, 0.515 mmol) in dry toluene
(2.6 mL) were added CaH2 (0.0864 g, 2 mmol), BnEt3N+Cl-
(0.03 g, 0.13 mmol), and pyrene-1-carbonyl chloride (0.3 g, 1.13
mmol). This was stirred in an oil bath at 110 °C for 24 h. The
reaction mixture was cooled, filtered through a bed of Celite,
and washed with ethyl acetate. The combined organic layer
was washed with 7% NaHCO3 solution, water, and brine,
finally dried over anhyd Na2SO4, and filtered. The solvent was
removed in vacuo to yield the crude product. The crude
product was purified by column chromatography on silica gel
(100-200 mesh, 16 g, 15 × 2.4 cm) using 15-30% EtOAc/
hexanes as the eluent. The pure product was crystallized from
EtOAc/hexane (1:1) and weighed 0.41 g (86%): mp 161-162
°C; IR (neat) 3001 (m), 2920 (s), 2850 (s), 1740 (s), 1690 (s),
(30) Even with solids with high melting points the IR data were
sometimes taken before crystallization, when the material was still
in a glassy form.
(31) (a) Fieser, L.; Rajagopalan, S. J . Am. Chem. Soc. 1949, 71, 3935.
(b) See also: Reigel, B.; Moffett, R. B.; Mcintosh, A. V. Organic
Syntheses; Wiley: New York, 1955; Collect. Vol. 3, p 237.
(32) Fieser, L. F.; Rajagopalan, S. J . Am. Chem. Soc. 1950, 72, 5530;
1951, 73, 4133.

Bile Acid-Based Molecular Tweezers
J . Org. Chem., Vol. 61, No. 26, 1996 9501
1590 (s), 1440 (s), 840 (s) and 740 (s) cm-1; [R]22D: +131 (c
1.902, CHCl3); 1H NMR (CDCl3, 400 MHz) δ 0.90 (s, 3H), 0.98
(d, 3H, J ) 6.5 Hz), 1.03-2.3 (m, steroidal CH and CH2), 2.57
(m, 2H), 2.99 (m, 1H), 3.59 (s, 3H), 5.08 (br m, 1H), 5.63 (s,
1H), 7.48 (d, 1H, J ) 8.1 Hz), 8.19-7.56 (m, 14H), 8.55 (d,
1H, J ) 8.0 Hz), 8.90 (d, 1H, J ) 9.5 Hz), 9.10 (d, 1H, J ) 9.5
Hz); 13C NMR (CDCl3, 100.5 MHz) δ 13.0, 14.3, 18.0, 21.1, 22.8,
24.5, 26.3, 26.5, 27.6, 31.0, 31.2, 33.5, 33.9, 34.8, 34.9, 37.4,
42.9, 45.4, 45.8, 47.3, 49.6, 51.5, 60.5, 73.4, 76.0, 123.7, 123.9,
124.1, 124.2, 124.4, 124.5, 124.5, 124.7, 124.9, 126.1, 126.1,
126.2, 126.2, 127.0, 127.2, 127.9, 128.1, 128.5, 129.2, 129.4,
129.6, 130.2, 130.3, 130.8, 130.8, 130.9, 134.0, 134.2, 167.4,
167.7, 174.5, 211.6; UV (λmax, log ꢀ) (6% CHCl3/CH3CN, v/v)
383 (4.05), 350 (4.65), 280 (4.65), 243 (4.91); MS (FAB) 877
(100), 632 (15), 460 (22). Anal. Calcd for C59H56O7: C, 80.79;
H, 6.44. Found: C, 80.75; H, 6.44.
standard solutions of host and guest were mixed in different
proportions, starting from 0.125 to 0.875 mole fraction so that
the [H] + [G] was kept constant while varying [H]/[G]. The
total volume of CDCl3 in the NMR tube was 400 µL. 1H NMR
spectra were recorded for each NMR tube, and ∆δ values were
calculated by subtracting the chemical shift of the doublet of
one of the pyrenes in the spectrum of the mixtures (δx) from
the same resonance of the pure host (δo) (δ 8.91). Using host
molarities and ∆δmax, the actual concentration of the complex
was calculated. A graph of the concentration of the complex
vs mole fraction was plotted. The maximum corresponded to
0.5 mole fraction, confirming a 1:1 stoichiometry.
F lu or escen ce Exp er im en ts w ith Gu est. Standard solu-
tions of host 8 (0.66 mM) and guest 10b (33 mM) in CHCl3
were prepared. Fluorescence spectra were recorded for host,
guest, and the host-guest mixture. The excitation wavelength
and the measured emission ranges were 355 and 375-600 nm,
respectively. A similar procedure was followed for the controls
also using monopyrene 6 and methyl pyrene-1-carboxylate (5d )
each at 0.66 mM concentration.
Deter m in a tion of th e Ch a r ge-Tr a n sfer Ba n d s by th e
UV Meth od .35 The UV spectrum of host 8 (3.52 × 10-4 M in
CHCl3) and guest 10e (0.031 M CHCl3) were taken separately
in the 420-600 nm range using a 1 mm quartz cuvette. The
sum of these two were done by the computer to give the
spectrum of (H + G). Similarly, the UV spectrum of the
mixture (HG) was recorded. The difference spectrum [(HG)
- (H + G)] gives the charge-transfer spectrum which is usually
seen in the longer wavelength region than the normal absorp-
tions. A similar experiment was also done for the controls
taking 6 and 5d each at identical concentrations as above.
Met h yl 3r,12r-Bis((1-p yr en oyl)oxy)-7r-h yd r oxy-5â-
ch ola n -24-oa te (9). Methyl 3R,12R-bis((1-pyrenoyl)oxy)-7-
keto-5â-cholan-24-oate (8) (0.1 g, 0.11 mmol) was dissolved in
THF (0.2 mL), and methanol (1 mL) was added. The reaction
mixture was cooled in an ice bath, and NaBH4 (6 mg, 0.17
mmol) was added. The reaction mixture was stirred at rt
(ca. 24 °C) for 4 h. Volatiles were removed in vacuo, and the
residue was taken up in ethyl acetate. The organic layer was
washed with 7% NaHCO3 solution, water, and brine, finally
dried over anhyd Na2SO4, and filtered. The solvent was
removed in vacuo to yield the crude product, which was
purified by column chromatography on silica gel (100-200
mesh, 8 g, 18 cm × 1.4 cm) using 30-40% EtOAc/hexanes as
the eluent. The pure product was crystallized from EtOAc/
hexane (1:1) and weighed 0.09 g (89% yield): mp 172-174 °C;
[R]22D: +105 (c 4.904, CHCl3); IR (neat) 3100 (s), 2920 (s), 2840
(m), 1690 (s), 1720 (s), 1590 (m), 750 (s) cm-1; UV (λmax, log ꢀ)
(6% CHCl3/CH3CN, v/v) 383 (4.02), 350 (4.66), 280 (4.66), 243
(4.94); 1H NMR (400 MHz, CDCl3) δ 0.91 (s, 3H), 0.99 (d, 3H,
J ) 6.6 Hz), 1.01 (s, 3H), 1.10-2.30 (m, steroidal CH and CH2),
2.74 (m, 2H), 3.56 (s, 3H), 5.04 (br m, 1H), 5.67 (s, 1H), 7.44
(d, 1H, J ) 8.0 Hz), 7.61 (t, 1H, J ) 7.7 Hz), 8.09-7.69 (m,
12H), 8.20 (d, 1H, J ) 8.7 Hz), 8.60 (d, 1H, J ) 8.0 Hz), 8.98
(d, 1H, J ) 9.5 Hz), 9.18 (d, 1H, J ) 9.5 Hz); 13C NMR (100.5
MHz, CDCl3) δ 12.4, 14.1, 17.7, 22.3, 23.0, 25.5, 27.0, 27.2,
28.2, 29.2, 29.6, 30.7, 30.9, 34.4, 34.7, 34.8, 35.0, 35.4, 39.6,
41.2, 43.6, 45.3, 47.7, 51.3, 68.1, 74.8, 76.8, 123.8, 124.0, 124.0,
124.1, 124.3, 124.4, 124.6, 124.7, 125.2, 125.8, 125.9, 126.5,
126.8, 127.0, 127.9, 128.0, 128.8, 129.1, 129.2, 129.5, 130.1,
130.1, 130.4, 130.5, 130.7, 130.8, 133.7, 133.9, 167.6, 168.1,
174.4; MS (FAB) 879 (M+, 50), 229 (100). Anal. Calcd for
C59H58O7: C, 80.61; H, 6.65. Found: C, 80.60; H, 6.68.
Exp er im en ta l P r oced u r e for NMR Titr a tion . As a
specific example, titration of methyl 3R,12R-bis((1-pyrenoyl)-
oxy)-7-keto-5â-cholan-24-oate (8) with octafluoronaphthalene
(10b) is described here. A 0.0758 M stock solution of host
(CDCl3) and a 0.187 M stock solution of guest (CDCl3) were
prepared. In 8 separate NMR tubes, 100 µL of the of host
solution and 0, 20, 30, 40, 60, 80, 120, and 200 µL of the guest
solution were added.33 The total volume in each NMR tube
was adjusted to 400 µL by adding CDCl3. 1H NMR spectra
were recorded for each tube, and ∆δ values were calculated
by subtracting the chemical shift of interest (δ 8.91 doublet)
in the spectrum of the mixture (δx) from the same resonance
in the spectrum of pure host (δo). Thus, a titration curve of
∆δ vs mole ratio of guest/host was plotted. The data were
analyzed using a nonlinear curve-fitting program, and the
output file gave the values of Ka (19 M-1) and ∆δmax (0.5 ppm).
We have modified the program to give the value of the rms
P olym er -Bou n d Molecu la r Tw eezer : 3r,12r-Bis((1-
p yr en oyl)oxy)-7-k eto-5â-ch ola n -24-oic Acid (8a ). Methyl
3R,12R-bis((1-pyrenoyl)oxy)-7-keto-5â-cholan-24-oate (8) (0.08
g, 0.093 mmol) was dissolved in THF/MeOH (0.5:1, v/v). An
aqueous solution of LiOH (0.3 mL of 1.25 M, 0.38 mmol) was
added, and the mixture was stirred at room temperature for
5 h. The reaction mixture was acidified with dilute HCl, and
the volatiles were removed on a rotavapor. The residue was
taken up in ethyl acetate (5 mL), and the organic layer was
washed with water and brine, finally dried over anhyd Na2-
SO4, and filtered. The solvent was removed in vacuo to yield
the crude product, which was purified by column chromatog-
raphy on silica gel (100-200 mesh, 1.2 cm × 12 cm) using
5-10% EtOAC/CHCl3 as the eluent. The pure product was
obtained in 89% yield (0.07 g): mp 178-180 °C; IR (neat) 2940
(s), 1730 (s), 1580 (m), 1450 (m), 1250 (m), 850 (s), 750 (s) cm-1
;
1H NMR (270 MHz, CDCl3) δ 0.90 (s, 3H), 0.98 (d, 3H, J ) 6.3
Hz), 1.0-2.40 (m, steroidal CH and CH2), 2.58 (m, 2H), 3.0
(m, 1H), 5.2 (br m, 1H), 5.61 (s, 1H), 7.55 (d, 3H, J ) 6.5 Hz),
8.27-7.27 (m, 15H), 8.55 (d, 1H, J ) 8.1 Hz), 8.91 (d, 1H, J )
9.4 Hz), 9.10 (d, 1H, J ) 9.4 Hz); MS (FAB) 863 (M+ + H, 82),
229 (100). Anal. Calcd for C58H54O7: C, 80.72; H, 6.31.
Found: C, 80.28; H, 6.24.
P r ep a r a tion of P olym er -Bou n d Host 12.36 3R,12R-bis-
((1-pyrenoyl)oxy)-7-keto-5â-cholan-24-oic acid (8a ) (0.181 g,
0.21 mmol) was suspended in THF (1 mL), and Cs2CO3 (0.05
g in 0.3 mL of H2O, 0.15 mmol) was added. After stirring for
3 h, the solvents were evaporated and the residue was dried
in vacuum. To this residue (0.2 g, 0.3 mmol) were added DMF
(2 mL) and 1% cross-linked chloromethylated polystyrene (0.18
g, 0.18 mmol). The mixture was heated in an oil bath at 80-
100 °C for 24 h. The reaction mixture was filtered and washed
thoroughly with dilute HCl, DMF, EtOH, THF, and dry
acetone to remove any soluble material. After drying under
high vacuum, the polymer weighed 0.33 g (0.025 g of 8a was
recovered). For estimating the degree of pyrene loading, the
polymer-bound host (20 mg) was refluxed with an excess of
deviation of ∆δ (∆calc-∆δobs), and the calculated value (∆δrms
in this case was 0.005 33 ppm.
)
J ob P lot for Tr in itr oflu or en on e 10e a n d 8.34 This
experiment involved the preparation of a standard solution of
the guest (G ) 10e) (0.015 g, 0.03 M in CDCl3) and the host
(H ) 8) (0.05 g, 0.057 M in CDCl3). In nine NMR tubes
(35) Schwartz, M. H. J . Inclusion Phenom. Mol. Recog. 1990, 9, 1.
(36) Shrerrington, D. C.; Hodge, D. Synthesis and Separation using
Functional Polymers; Wiley: New York, 1988.
(33) For all NMR titration experiments the same gas tight microliter
syringe was used.
(34) (a) J ob, A. Ann. Chim. (Paris) 1928, 9, 113. (b) Gil, V. M. S.;
Olivira, N. C. J . Chem. Ed. 1990, 67, 473.
(37) The author has deposited atomic coordinates for 8 with the
Cambridge Crystallographic Data Centre. The coordinates can be
obtained, on request, from the Director, Cambridge Crystallographic
Data Centre, 12 Union Road, Cambridge, CB2 1EZ, U.K.

9502 J . Org. Chem., Vol. 61, No. 26, 1996
D’Souza and Maitra
sodium methoxide in MeOH/H2O. After 24 h, the reaction
mixture was acidified (dilute HCl) and stirred for 1 h. The
reaction mixture was filtered, and the residue was washed
thoroughly with EtOH. Volatiles were removed from the
filtrate, and the residue was thoroughly dried under vacuum.
The resulting pyrene-1-carboxylic acid (5b ) and the steroid
were separated by chromatography on a short silica gel column
using 50% EtOAc/hexanes as the eluent. Pure pyrene-1-
carboxylic acid obtained weighed 7.5 mg (0.03 mmol). The
Merrifield resin was reported to have 1 mequiv of active
chlorine/g of resin. Calculation based on the amount of the
pyrene-1-carboxylic acid obtained showed that 12 contained
0.54 mmol of steroid units/g of functionalized polymer.
P r ep a r a tion of P olym er -Bou n d Con tr ol 13. From
pyrene-1-carboxylic acid (5b) (0.1 g/1 mL of THF, 0.4 mmol)
and Cs2CO3 (0.08 g/0.3 mL of H2O, 0.24 mmol) the same
procedure as above was followed to prepare the cesium salt of
the acid. The crude cesium salt (0.15 g, 0.4 mmol) was
suspended in DMF (2 mL). To this was added 1% cross-linked
chloromethylated polystyrene (0.32 g, 0.32 mmol), and the
coupling and purification were carried out as for 12. The
product weighed 0.32 g. Estimation of the amount of pyrene-
1-carboxylic acid bound to the polymer was done by refluxing
the polymeric product (20 mg) with an excess of sodium
methoxide in MeOH and then hydrolyzing the methyl ester
with HCl. The pure acid obtained weighed 9.5 mg (0.039
mmol). Calculation based on the amount of the pyrene-1-
carboxylic acid obtained indicated that polymer 13 contained
0.83 mmol of pyrene units/g of the functionalized polymer.
Deter m in a tion of Gu est Selectivity of th e P olym er ic
Ma ter ia ls by HP LC. Solutions (0.1 mM) of different guests
(TNB-10d , TNF-10e, TENF-10f) in CHCl3 (10 mL) were
prepared. In three separate 2 mL volumetric flasks were
taken 20 mg each of polymeric host 12, control 13, and
Merrifield resin 11; 0.5 mL of the standard guest solution was
added to each, and the flasks were tightly stoppered. After
stirring for 24 h, the CHCl3 layer was filtered and analyzed
by HPLC on a 250 mm × 4.6 mm silica gel column using 85:
15 (v/v) hexane/EtOAc as a mobile phase at rt at a detector
wavelength of 254 nm. The retention time (tR) of each guest
was determined prior to the binding studies. Table 2 contain-
ing data for percent guest bound to 12 and 13 is given in the
Results and Discussion. From the ratios it is clear that 12
removes trinitrofluorenone most effectively. Merrifield resin
11 did not remove any of the guests under the identical
conditions.
X-r a y Str u ctu r e Deter m in a tion . Clear colorless colum-
nar crystals of the compound were obtained by slow crystal-
lization from petroleum ether/ethyl acetate: instrument used,
Enraf-Nonius CAD-4 diffractometer, equipped with a nitrogen-
flow low-temperature apparatus and controlled by a microVAX
II computer; radiation, Mo KR (λ ) 0.710 73 Å); monochroma-
tor, highly oriented graphite (2θ ) 12.2); detector, crystal
scintillation counter, with PHA; reflections measured, +h, +k,
+l; 2θ range, 3° f 45°; scan type, ω; scan width, ∆ω ) 0.60 +
0.35 tan θ; scan speed, 5.49 (ω, deg/min); background, mea-
sured over 0.25*(∆ω) added to each end of the scan; vertical
aperture ) 3.0 mm; horizontal aperture ) 2.0 + 1.0 tan θ mm;
no. of reflections collected, 3794; no. of unique reflections, 3763.
Crystal data of 8: Empirical formula, C63H64O9, Mr ) 965.2;
orthorhombic, space group P212121; a ) 8.299(2) Å, b ) 16.071-
(2) Å, c ) 38.037(6) Å, R ) 90.0°, â ) 90.0°, γ ) 90.0°; V )
5073.10(24) Å3, Z ) 4; Dc ) 1.26 g cm-3; Mo KR (λ ) 0.710 73
Å); crystal size, 0.22 × 0.28 × 0.50 mm.
Ack n ow led gm en t. We are grateful to Prof. Paul A.
Bartlett for help with the X-ray analysis. We thank Dr.
Frederick J . Hollander for performing the X-ray analysis
of 8 in the CHEXRAY facility at the University of
California at Berkeley. Thanks are due to the SERC
and the SIF of this institute for providing computing
and high-field NMR facilities, respectively. L.J .D.
thanks the UGC for a graduate fellowship. This work
was supported by a grant from the Department of
Science and Technology (SP/S1/G09/91), New Delhi.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
1
spectra for compounds 6-9, H NMR spectrum of 8a , repre-
sentative NMR titration data, two other views of the packing
diagram of 8, and charge-transfer spectrum and J ob plot for
the 8‚10e complex (13 pages). This material is contained in
libraries on microfiche, immediately follows this article in the
microfilm version of the journal, and can be ordered from the
ACS; see any current masthead page for ordering information.
J O9607525