Chiral molecular recognition in a tripeptide benzylviologen cyclophane host

Article abstract of DOI:10.1021/jo980352c

A cationic chiral cyclophane was synthesized and studied as a host for chiral and racemic π-donor molecules. The cyclophane host has a rigid binding cavity flanked by (S)-(valine-leucine-alanine) and N,N'-dibenzyl- 4,4'-bipyridinium subunits, which allow for hydrogen-bonding and π-stacking interactions with included aromatic guest molecules. ¹H NMR binding titrations were performed with several different pharmaceutically interesting guest molecules including β-blockers, NSAIDs, and amino acids and amino acid derivatives. The host-guest complexation constants were generally small for neutral and cationic guests (0-39 M^-1 at 20 °C in water/acetone mixtures. However, a (R)/(S) enantioselectivity ratio of 13 ± 5 was found for DOPA, a strongly π-donating cationic guest. Two-dimensional NOESY ¹H NMR spectra confirm that (R)-DOPA binds inside the cavity of the host and that there is no measurable interaction of the cavity with (S)-DOPA under the same conditions.

Full text of DOI:10.1021/jo980352c

J . Org. Chem. 1998, 63, 7663-7669
7663
Ch ir a l Molecu la r Recogn ition in a Tr ip ep tid e Ben zylviologen
Cyclop h a n e Host
J ulia A. Gavin, Maurie E. Garcia,† Alan J . Benesi, and Thomas E. Mallouk*
Department of Chemistry, The Pennsylvania State University, University Park, Pennsylvania 16802
Received February 24, 1998
A cationic chiral cyclophane was synthesized and studied as a host for chiral and racemic π-donor
molecules. The cyclophane host has a rigid binding cavity flanked by (S)-(valine-leucine-alanine)
and N,N′-dibenzyl-4,4′-bipyridinium subunits, which allow for hydrogen-bonding and π-stacking
interactions with included aromatic guest molecules. 1H NMR binding titrations were performed
with several different pharmaceutically interesting guest molecules including â-blockers, NSAIDs,
and amino acids and amino acid derivatives. The host-guest complexation constants were generally
small for neutral and cationic guests (0-39 M-1 at 20 °C in water/acetone mixtures). However, a
(R)/(S) enantioselectivity ratio of 13 ( 5 was found for DOPA, a strongly π-donating cationic guest.
1
Two-dimensional NOESY H NMR spectra confirm that (R)-DOPA binds inside the cavity of the
host and that there is no measurable interaction of the cavity with (S)-DOPA under the same
conditions.
In tr od u ction
macrocycles onto silica and have used these materials
in the analysis of a wide variety of enantiomeric and
diastereomeric guests. Recent work in our laboratory has
shown that the intercalation of chiral cationic host
molecules into R-zirconium phosphate, a lamellar cation
exchanger, provides a useful medium for batchwise
resolution of racemic mixtures.8 The scale of this process
is more than an order of magnitude higher than it is with
conventional “brush”-type chiral stationary phases, but
expansion and contraction of the solid and concomitant
host preorganization effects have precluded its use in
chromatographic applications. Replacing the linear chiral
host that was used in those experiments with a cyclo-
phane-type host containing a rigid, preorganized binding
cavity could solve the problem of host preorganization
and motivates the study reported in this paper.
Several groups9-11 have now incorporated chiral amino
acids into cyclophane hosts. There are three significant
advantages to this approach. First, the synthesis is
modular and employs well-established peptide coupling
methods. Second, each amide bond potentially provides
two hydrogen-bonding contacts to the guest, in close
proximity to an asymmetric carbon atom. Third, because
there is a large inventory of natural and unnatural amino
acids from which to choose, a very large number of
structurally similar host molecules can be prepared. For
hosts containing more than one amino acid, the synthesis
can be done in combinatorial fashion, and so in principle
one can make and test a diverse library of host molecules.
Cyclophanes, which are cyclic molecules containing
aromatic groups in the ring, have interesting molecular
recognition properties.1 Because they are macrocycles,
these hosts have built into them a measure of preorga-
nization that enhances their affinity for guest molecules
of the appropriate size and shape.2 The hydrophobicity
and π-stacking interactions of their aromatic groups also
contribute to host-guest affinity, which can be very high.
For these reasons, they have been widely studied as
synthetic receptors3 and as components of supramolecu-
lar assemblies.4
Chiral cyclophanes are of particular interest as the
active components of stationary phases for chiral separa-
tions. Cram and co-workers prepared the first cyclo-
phane of this type, a crown ether incorporating a chiral
binaphthol unit.5 This host was later tethered to chro-
matographic silica to produce a commercially available
stationary phase that has been used for the analysis of
chiral hydrogen bond donors, including amines, amino
alcohols, amino acids, and amino esters.6 More recently,
Armstrong et al.7 have immobilized several antibiotic
† Current address: Department of Chemistry, Texas A&M Univerisi-
ty, College Station, TX 77843.
(1) For recent examples and reviews, see: (a) Wilcox, C. S.; Glagov-
ich, N. M.; Webb, T. H. ACS Symp. Ser. 1994, 568, 282-90. (b) Webb,
T. H.; Wilcox, C. S. Chem Soc. Rev. 1993, 22, 383-395. (c) Diederich,
F.; Cyclophanes; The Royal Society of Chemistry: Cambridge, U.K.
1991. (d) Torneiro, M.; Still, W. C. J . Am. Chem. Soc. 1995, 117, 5887.
(e) Helgeson, R. C.; Knobler, C. B.; Cram, D. J . J . Am. Chem. Soc.
1997, 119, 3229.
(2) Cram, D. J . Angew. Chem., Int. Ed. Engl. 1988, 27, 1009.
(3) Murakami, Y.; Hayashida, O. Proc. Natl. Acad. Sci. U.S.A. 1993,
90, 1140.
(7) (a) Armstrong, D. W.; Tang, Y.; Chen, S.; Zhou, Y.; Bagwill, C.;
Chen, J .-R. Anal. Chem. 1994, 66, 1473. (b) Hilton, M.; Armstrong, D.
W. J . Liquid Chromatogr. 1991, 14, 3673.
(8) (a) Cao, G.; Garcia, M. E.; Alcala, M.; Burgess, L. F.; Mallouk,
T. E. J . Am. Chem. Soc. 1992, 114, 7574. (b) Garcia, M. E.; Naffin, J .
L.; Deng, N. Chem. Mater. 1995, 7, 1968.
(9) (a) Erickson, S. D.; Simon, J . A.; Still, W. C. J . Org. Chem. 1993,
58, 1305. (b) Gasparrini, F.; Misiti, D.; Villanni, C.; Borchardt, A.;
Burger, M. T.; Still, W. C. J . Org. Chem. 1995, 60, 4314.
(10) Garcia, M. E.; Gavin, J . A.; Deng, N.; Andrievsky, A. A.;
Mallouk, T. M. Tetrahedron Lett. 1996, 37, 8313.
(4) Fyfe, M. C. T.; Stoddart, J . F. Acc. Chem. Res. 1997, 30, 393.
(5) (a) Kyba, E. P.; Gokel, G. W.; De J ong, F.; Koga, K.; Sousa, L.
R.; Siegel, M. G.; Kaplan, L.; Sogah, G. D. Y.; Cram, D. J . J . Org. Chem.
1977, 42, 4173. (b) Cram, D. J .; Helgeson, R. C.; Peacock, S. C.; Kaplan,
L. J .; Domeier, L. A.; Moreau, P.; Koga, K.; Mayer, J . M.; Chao, Y.;
Siegel, M. G.; Hoffman, D. H.; Sogah, G. D. Y. J . Org. Chem. 1978, 43,
1930. (c) Lingenfelter, D. S.; Helgeson, R. C.; Cram, D. J . J . Org. Chem.
1981, 46, 393.
(6) (a) Shinbo, T.; Yamaguchi, T.; Nishimura, K.; Suguira, M. J .
Chromatogr. 1987, 405, 145. (b) Applications Guide for Chiral Column
Selection, 2nd ed.; Chiral Technologies Inc.: Exton, PA 1993.
(11) (a) Hayashida, O.; Ono, K.; Hisaeda, Y.; Murakami, Y. Tetra-
hedron 1995, 51, 8423. (b) Hayashida, O.; Tanaka, A.; Fujiyoshi, S.;
Hisaeda, Y.; Murakami, Y. Tetrahedron Lett. 1997, 38, 1219.
10.1021/jo980352c CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/09/1998

7664 J . Org. Chem., Vol. 63, No. 22, 1998
Gavin et al.
of 4-(chloromethyl)benzoic acid with 4,4′-bipyridine in
acetonitrile to give the chloride salt of N-[(p-carboxyphe-
nyl)methyl]4,4′-bipyridine. The amine end of the bipy-
ridinium unit derives from 4-(bromomethyl)benzylamine
hydrobromide, which was made in three steps from
p-bromotolunitrile, in an overall yield of ca. 37%. Briefly,
bromotolunitrile was refluxed in water with barium
carbonate to give p-cyanobenzyl alcohol, which was
reduced with lithium alumium hydride in refluxing ether
to give p-(hydroxymethyl)benzylamine hydrochloride,
followed by reflux in hydrogen bromide to give 4-(bro-
momethyl)benzylamine hydrobromide. N-[(p-Carboxy-
phenyl)methyl]4,4′-bipyridine]1+(PF6-) and 4-(bromom-
ethyl)benzylamine hydrobromide were combined and
refluxed in acetonitrile to give the complete bipyridinium
unit, {N-[(p-carboxyphenyl)methyl]-N′-[(p-(aminomethyl)-
phenyl)methyl]-4,4′-bipyridine}2+ (2Br-).
The tripeptide unit was synthesized by standard solid-
phase methods. The cesium salt of the first amino acid
was obtained by reacting the N-t-Boc-protected amino
acid with cesium carbonate. The cesium salt of the N-t-
BOC-amino acid was coupled to Merrifield resin and was
deprotected with trifluoroacetic acid in chloroform. Sub-
sequent coupling of N-t-BOC-amino acids was performed
using dicyclohexylcarbodiimide in dichloromethane and
dimethylformamide. Finally, the tripeptide was cleaved
from the resin using a phase-transfer catalyst and
isolated with the N-t-BOC protecting group intact.
F igu r e 1. Synthetic scheme for the (S)-valine-leucine-alanine
cyclophane 1.
There have been several papers that describe this kind
of combinatorial approach, although in practice it has not
yet been used to find the best host molecule for a given
enantioseparation. The closest examples are in the work
of Still and co-workers, who screened a library of pepti-
dosteroid hosts against chiral enkephalin guests12 and a
library of peptide guests against a single cyclophane
host.9 In related work, we recently prepared a small
combinatorial library of dipeptide-containing cyclo-
phanes. The affinity of those cyclophanes for chiral
guests was unfortunately too low for combinatorial
screening of their host-guest interactions.10
The cationic chiral cyclophane 1 that is reported here
was designed to contain the powerfully π-accepting
benzylviologen group13 along with a tripeptide chiral
barrier for enantioselective interactions. The aromatic
portion of 1 contains an amine and a carboxy terminus
for ease of inclusion of the tripeptide, (S)-valine-leucine-
alanine, which can be made by standard solid-phase
coupling methods. The synthesis is modular, and in
principle any tripeptide can be substituted for (S)-valine-
leucine-alanine. In this paper, we report the complex-
ation of this chiral cyclophane host in aqueous media
with NSAIDS, â-blockers, and amino acids and amino
acid derivatives, which are interesting analytes for
enantioseparations.
The bipyridinium unit was then coupled to the pro-
tected tripeptide using N-methylmorpholine, triethy-
lamine, and isobutyl chlorocarbonate in dry dimethyl-
formamide. Deprotection with trifluoroacetic acid in
chloroform and a second coupling with N-methylmorpho-
line, triethylamine, and isobutyl chlorocarbonate in dim-
ethylformamide resulted in the product cyclophane 1.
Despite the large size of the cyclophane ring, the final
coupling reaction proceeds in 40% yield under conditions
of moderate dilution (6 mM concentration). Compound
1
1 and the precursors to 1 were characterized by H and
13C NMR, mass spectrometry, and elemental analysis.
Thermal analysis showed that most of the salts and
peptide-containing compounds decomposed significantly
before melting.
Compound 1 was also characterized by amino acid
analysis. During the hydrolysis step of the amino acid
analysis, the cyclophane was cleaved at each of the amide
bonds. The bipyridinium portion of the cyclophane did
not interfere with the HPLC method for determining the
amino acids present in the mixture. The three amino
acids in the cyclophane, alanine, leucine, and valine were
detected in approximately equimolar quantities (1428,
1496, and 1497 nmol, respectively) in hydrolyzed samples
prepared from 1500 nmol of cyclophane 1.
Deter m in a tion of Bin d in g Con sta n ts. 1H NMR
titrations were performed in mixed water/acetone sol-
vents using the bromide salt of 1 at room temperature.
The host concentration was kept in the range of 5-10
mM for 1:1 binding experiments with analytes. Self-
association of the cationic cyclophane does not occur to
any measurable extent from 1 to 13 mM. The concentra-
tion of 1 was kept constant in the titrations, and the guest
concentration was varied in the range of 1-500 mM
depending on solubility. Equilibrium concentrations of
the guest molecules, association constants, and errors
Resu lts a n d Discu ssion
Syn th esis. The synthesis of the (S)-valine-leucine-
alanine cyclophane 1 is represented in Figure 1. The
synthesis is convergent, in the sense that the bipyri-
dinium and tripeptide fragments are made separately
and then combined by peptide coupling. The bipyri-
dinium fragment is itself an ω-amino acid, which is made
in two separate high-yield quaternization steps from 4,4′-
bipyridine and the appropriate benzyl halide derivatives.
The bipyridinium part of the molecule is made by reflux
(12) (a) Boyce, R.; Li, G.; Nestler, H. P.; Suenaga, T.; Still, W. C. J .
Am. Chem. Soc. 1994, 116, 7955. (b) Still, W. C. Acc. Chem. Res. 1996,
29, 155.
(13) (a) Anelli, P. L.; Spencer, N.; Stoddart, J . F. J . Am. Chem. Soc.
1991, 113, 5131. (b) Goodnow, T.; Reddington, M. V.; Stoddart, J . F.
J . Am. Chem. Soc. 1991, 113, 4335.

Chiral Molecular Recognition
J . Org. Chem., Vol. 63, No. 22, 1998 7665
Ta ble 1. Resu lts of NMR Titr a tion s Usin g th e
Dibr om id e Sa lt of 1. Associa tion Con sta n ts Rep r esen t
th e Aver a ge of Tw o or Mor e P r oton Ch em ica l Sh ifts
F igu r e 2. (A) Changes in chemical shift of cyclophane protons
(-CH2 groups between bipyridinium and phenyl) in 1H NMR
spectra as a function of (R)-DOPA concentration; [R] ) (a) 0,
(b) 0.111, and (c) 0.272 M. (B) Change in chemical shift plotted
against the analytical concentration of (R)- and (S)-DOPA. The
solid line is calculated for 1:1 host:guest complexation.
*Determined by titration of racemic DOPA.
were calculated using the HOSTEST (v.5.0) program.14
A 1:1 stoichiometry was found for all complexes in which
there was detectable host-guest binding. The results of
NMR titrations with several guest molecules are collected
in Table 1. Weak binding (1-39 M-1) was found between
the guest molecules and 1 in aqueous solutions. This
observation is consistent with previous studies of simple
acceptors such as methylviologen or 1-methylnicotina-
dide,15 as well as â-cyclodextrin derivatives,16 and pyro-
gallol or resorcinol cyclic tetramers17 with amino acids
in aqueous media. Apparently, the free energy penalty
incurred in desolvating the ionic host is almost equal to
the attractive (hydrogen bonding and π-stacking) interac-
tions in water. Stoddart et al. have shown that stronger
affinity for amino acids in aqueous media can be achieved
using stronger π-accepting hosts such as a box-like
cyclophane containing two bipyridinium acceptor sub-
units instead of the single bipyridinium group in 1.18
Of the guest molecules studied, racemic nadolol and
DOPA showed significant association with 1. The com-
plexation of 1 with the individual enantiomers of DOPA
was measured. However, the enantioselectivity of bind-
ing to nadolol was not measured, because pure enanti-
omers were not available. Figure 2a shows two cyclo-
phane 1H NMR peaks (from the -CH2 groups which
reside between the viologen and benzyl units of the
cyclophane) in an expanded region in the spectrum of the
cyclophane/(R)-DOPA complex. While the shifts in these
and other cyclophane peaks is small (∆δ ≈ 0.015 ppm
between the free and fully complexed host), the spectra
are sufficiently well resolved that they can be accurately
measured. Figure 2b shows binding isotherms, in which
the change in chemical shift is plotted against the
analytical concentration of (R)- and (S)-DOPA. The solid
line is calculated for 1:1 host:guest complexation. Only
a slight shift in the cyclophane 1H NMR peaks was
observed at the highest (S)-DOPA concentration, 0.338
M. No shifts were observed at lower concentrations.
UV-vis spectra showed a negative deviation from Beer’s
law in the absorption peak at 302 nm for DOPA concen-
trations above 0.3 M, implying DOPA self-association at
1
high concentration. Therefore, H NMR data obtained
at concentrations above 0.3 M were not used in the
determination of the DOPA association constants. Self-
association was not significant for the other guest
molecules in the concentration ranges used. The (R)-
enantiomer binds to the cyclophane with an association
constant of 39 ( 6 M-1. Because there were no observ-
able NMR shifts below 0.3 M concentration with (S)-
DOPA, its association constant with 1 could not be
determined directly. However, it was possible to estimate
this association constant through an indirect measure-
ment, by using racemic DOPA.
(14) (a) Wilcox, C. S.; Cowart, M. D. Tetrahedron Lett. 1986, 27,
5563. (b) Cowart, M. D.; Sucholeiki, I.; Bukownik, R. R.; Wilcox, C. S.
J . Am. Chem. Soc. 1988, 110, 6204.
(15) (a) Verhoeven, J . W.; Verhoeven-Schoff, A. M. A.; Masson, A.;
Schwyzer, R. Helv. Chim. Acta 1974, 57, 2503. (b) Deranleau, D. A.;
Schwyzer, R. Biochemistry 1970, 9, 126. (c) Nakano, Y.; Komiyama,
J .; Iijima, T. Colloid Polym. Sci. 1987, 265, 139.
(16) Tabushi, I.; Kuroda, Y.; Mizutani, T.; J . Am. Chem. Soc. 1986,
108, 4514.
(17) Kobayashi, K.; Tominaga, M.; Asakawa, Y.; Aoyama, Y. Tet-
rahedron Lett. 1993, 34, 5121.
(18) (a) Goodnow, T. T.; Reddington, M. V.; Stoddart, J . F.; Kaifer,
A. E. J . Am. Chem. Soc. 1991, 113, 4335. (b) Asakawa, M.; Brown, C.
L.; Pasini, D.; Stoddart, J . F.; Wyatt, P. G. J . Org. Chem. 1996, 61,
7234. (c) Asakawa, M.; Ashton, P. R.; Boyd, S. E.; Brown, C. L.; Menzer,
S.; Pasini, D.; Stoddart, J . F.; Tolley, M. S.; White, A. J . P.; Williams,
D. J .; Wyatt, P. G. Chem. Eur. J . 1997, 3, 463.
Complexation with racemic DOPA gave a single chemi-
cal shift for each of the protons of 1, implying rapid

7666 J . Org. Chem., Vol. 63, No. 22, 1998
Gavin et al.
exchange between free 1 and its complexes (1‚R and 1‚
S) with (R)-DOPA and (S)-DOPA. The concentration of
free 1 ([1]) and the total concentration of 1 ([1]T) are
related to the observed chemical shift (δobs), the chemical
shift at saturation (δsat), and the chemical shift of free 1
(δo) by eq 1. This approximation is valid as long as [1‚
resonances in (S)-DOPA, (R)-DOPA, and 1, weak inter-
molecular NOE’s were observed for the solution contain-
ing (R)-DOPA and 1, confirming that the cyclophane
binds (R)-DOPA with rapid exchange between bound and
free forms. In Figure 1 of the Supporting Information,
horizontal rows from the 2-D NOESY NMR spectra show
weak NOE’s among the (R)-DOPA CH2 resonance at 3.0
ppm and cyclophane aromatic protons (9.0, 8.4, 7.2),
cyclophane CH2’s next to the viologen rings (5.9), the
amino acid chiral centers (4.3-3.8, 3.6), and the aliphatic
side chains of the amino acids (0.8 ppm); and weak NOE’s
from the aromatic protons of (R)-DOPA to cyclophane
resonances at 8.4, 7.2-7.9, 5.9, 5.8, 3.6, and 0.7-1.2 ppm.
From the intensity of the NOE peaks, it appears that the
aromatic region of 1 is closer to the aromatic region of
(R)-DOPA compared to the (R)-DOPA CH2 protons.
Because of the fast exchange of the NH3+ protons of the
(R)-DOPA, monitoring by NMR is not possible. There-
fore, no information about the location of the positive
charge relative to the positive charges on 1 can be implied
from the 2-D NOESY NMR experiments, although the
δ - δsat
δo - δsat
[1] ) [1]T
(1)
(
)
R] is large compared to [1‚S], which is the case here. KS,
the association constant of (S)-DOPA with 1, can be
calculated for each observed chemical shift in the titra-
tion curve with (R/S)-DOPA, using the experimentally
derived value of 39 ( 6 M-1 for KR, along with δobs, δsat
,
δo, [1]T, and the total concentrations of the two DOPA
enantiomers ([R]T and [S]T). The equilibrium expressions
needed to calculate KS are given in eqs 2 and 3. Five
[1]T - [1] - [1‚R]
KS )
(2)
(3)
[1] ([S]T - [1]T + [1] + [1‚R])
+
principles of electrostatics suggest that the NH3 group
[1]KR[R]T
[1‚R] )
of the (R)-DOPA zwitterion should be as far as possible
from the viologen dication region of 1. By contrast, the
equivalent rows from the 2-D NOESY spectrum of an (S)-
DOPA/cyclophane solution show under identical condi-
tions one extremely weak NOE from the (S)-DOPA CH2
resonances to a cyclophane resonance at 1.0 ppm and two
extremely weak NOE’s from the (S)-DOPA aromatic
peaks to cyclophane resonances at 5.9 and 5.8 ppm
(Figure 2a,b of the Supporting Information). The near
absence of intermolecular NOE’s for the latter solution
suggests that only incidental contact occurs between (S)-
DOPA and the exterior of the cyclophane ring. The
presence of many intermolecular NOE’s between (R)-
DOPA and the cyclophane supports the hypothesis that
there is relatively strong and enantiomerically selective
binding of (R)-DOPA within the cyclophane cavity.
([1]KR + 1)
titration points with (R/S)-DOPA in the concentration
range of 0.028-0.212 M gave KS ) 3 ( 1 M-1, resulting
in an enantioselectivity ratio of 13 ( 5 for the association
of 1 with the DOPA enantiomers.
Variable-temperature NMR was used to determine
thermodynamic parameters for the complexation of (R)-
DOPA with 1. Binding experiments were performed at
5, 15, and 25 °C. Although there were slight shifts in
the NMR peaks of the cyclophane-only solution with
temperature (downfield shift at lower temperature), no
variations in the ∆ppm values for each concentation of
(R)-DOPA were found with temperature. Therefore, ∆H°
) 0.0 ( 0.1 kcal/mol, and ∆S° ) +7.3 ( 1.2 cal/mol K.
It is interesting to compare cyclophane 1 with cage-
like cationic hosts synthesized by Murakami et al., which
also contain (S)-alanine, -leucine, and -valine groups.11
Murakami et al. observed much higher association
constants (on the order of 105-106 M-1), which may be
attributed to the more hydrophobic nature of the cage
and the fact that the guest molecules were anions. While
Murakami et al. did not directly measure enantioselec-
tivity ratios for chiral guests, they did determine that
the chiral cage imposes its helicity on “hinged” aromatic
guests.11,19 This behavior is consistent with the high
enantioselectivity we observe with 1. The enantioselec-
tivity ratio of DOPA with 1 compares favorably with
bowl-shaped cyclophane hosts that also contain three
amino acid groups9 and with commercially available
chiral separations media.
In ter ca la tion of 1 in to r-Zir con iu m P h osp h a te.
Because the cyclophane is cationic, it was easily inter-
calated in the lamellar solid acid R-zirconium phosphate
by cation exchange for tetra(n-butyl)ammonium ions, as
previously described.8
Attem p ts to Sep a r a te DOP A in 1-In ter ca la ted
Solid . Enantioseparation of DOPA using the cyclo-
phane-intercalated R-zirconium phosphate was attempted
by means of a batch-mode procedure as previously
described.23 Briefly, 100 mM racemic DOPA, in the same
solution mixture that was used for the NMR binding
experimnents, was added to the cyclophane-intercalated
solid in a 4:1 molar guest:host ratio. By chiral HPLC
analysis, there was no enantiomeric excess in the super-
natant after 12 h of stirring and filtration of the solid.
UV-vis spectroscopy of the supernatant showed that
there was no change in DOPA concentration despite this
long equilibration period. Thus, no DOPA intercalated
into the solid to complex with the cyclophane.
2-D NMR Exp er im en ts. To further characterize the
cyclophane/(R)-DOPA host-guest interaction, two-di-
mensional NOESY 1H NMR experiments were performed
on solutions containing both compounds. In addition to
the expected strong intramolecular NOE’s for the 1H
Related work in this laboratory has shown that the
association constants of other benzylviologen-cyclophane
intercalation compounds with π-donors are about a factor
of 5 lower than they are in solution.20 A possible
(19) Murakami, Y.; Hayashida, O.; Nagai, Y. J . Am. Chem. Soc.
1994, 116, 2611.
(20) Gavin, J . A.; Deng, N.; Alcala, M.; Mallouk, T. E. Chem. Mater.
1998, 10, 1937.
(21) Bardsley, W. G.; Ashford, J . S.; Hill, C. M. Biochem. J . 1971,
122, 557.
(22) Gisin, B. F. Helv. Chim. Acta 1973, 56, 1476.
(23) (a) Gutte, B.; Merrifield, R. B. J . Biol. Chem. 1971, 246, 1922.
(b) Kaiser, E.; Colescott, R. L.; Bossinger, C. D.; Cook, P. I. Anal.
Biochem. 1970, 34, 595.

Chiral Molecular Recognition
J . Org. Chem., Vol. 63, No. 22, 1998 7667
cipitated with an excess of ammonium hexafluorophosphate
followed by addition of hexafluorophosphoric acid (60 wt %
solution in deionized water, dropwise to pH 4). After cooling
the solution at 3 °C for several hours, the white solid was
vacuum-filtered and washed with deionized water to yield the
product (7.1 g, 98%): mp 166 °C; 1H NMR (CD3CN) δ 8.85 (d,
J ) 6.5 Hz, 4H), 8.33 (d, J ) 6.5 Hz, 2H), 8.08 (d, J ) 6.5 Hz,
explanation for this effect is that electron donation by
the negatively charged R-ZrP sheets reduces the π-acidity
of the cyclophane host. A reduction of the association
constant of 1 with DOPA by the same factor of 5 would
explain the fact that it does not intercalate into 1/R-ZrP
at 100 mM concentration. Interestingly, this effect was
not observed in the complexation/intercalation reactions
of a “Pirkle phase” chiral host in R-ZrP, because in that
case the center of positive charge on the molecule is
relatively remote from the π-accepting dinitrobenzoyl
group.8
2H), 7.80 (d, 7.0 Hz, 2H), 7.55 (d, 7.0 Hz, 2H), 6.80 (2H, s); 13
C
NMR ((CD3)2SO) δ 166.26 (carbonyl), [150.53, 144.95, 140.211,
130.92, 125.99, 121.80] (dipyridyl carbons), [150.32, 131.46,
130.30, 130.00] (phenyl carbons), 63.795 (methylene carbon);
positive ion FABMS M+ ) 291 (M without PF6 counterions).
Anal. Calcd (found): C, 49.54% (49.41%); H, 3.44% (3.57%);
N, 6.42% (6.40%).
Su m m a r y a n d Con clu sion s
4-(Br om om et h yl)b en zyla m in e H yd r ob r om id e. The
preparation of this compound is described in the literature.22
Modifications described below were made to the synthesis to
simplify as well as cut the time necessary for the purification
in half while increasing the overall yield of the reaction (21%
vs 37% overall yield with modifications). Bromotolunitrile
(10.0 g, 51.0 mmol) was refluxed in deionized water (300 mL)
with barium carbonate (20.0 g) for 2 h. The solid byproduct
was removed by filtration. Chloroform extraction of the filtrate
and removal of solvent by rotary evaporation yielded a white
solid, p-cyanobenzyl alcohol (5.10 g, 75.2%). p-Cyanobenzyl
alcohol (6.54 g, 49.3 mmol) was dissolved in anhydrous ether
and added dropwise to LiAlH4 (3.27 g) in ether over 15 min at
0 °C under argon. After the addition, the reaction was refluxed
for 3 h. Sodium hydroxide solution (0.10 M) was added
followed by chloroform extraction. The chloroform solution
was then extracted with 0.1 M HCl. Solvent was removed
using rotary evaporation to yield a white solid, p-(hydroxym-
ethyl)benzylamine (3.50 g, 51.8%). p-(Hydroxymethyl)benzy-
lamine (3.50 g, 25.5 mmol) was refluxed in HBr (100 mL of a
solution prepared from 60 mL of concentrated HBr andwater)
for 3 h. Solvent was removed by rotary evaporation and the
residue was washed with acetone to yield the product as a
The new tripeptide-containing cyclophane reported
here is the first member of a potentially diverse group of
hosts which might be prepared by substituting other
amino acids for Val-Leu-Ala and other π-acceptors for
bipyridinium. Because the binding pocket contains the
relatively hydrophilic bipyridinium dication, it binds
π-donor analytes weakly in water, and even more weakly
when it is intercalated into (anionic) R-zirconium phos-
phate. Although amino acids with small aliphatic side
groups were used to define the binding cavity, the
complexation of this host with DOPA is quite enantiose-
lective. Efforts are currently underway to redesign this
cyclophane to contain a more hydrophobic and more
powerfully π-accepting binding pocket, for more effective
complexation of hydrophobic π-donor guests. The fact
that this cyclophane, which represents only a single
member of a potentially large library, shows good enan-
tioselectivity suggests that a combinatorial synthesis/
screening approach could be effective in identifying useful
chiral selectors. Solid-phase synthesis and screening
techniques of these and related hosts are currently being
developed in this laboratory to evaluate the utility of this
approach.
1
white solid (6.80 g, 95.6%): mp above 300 °C (dec); H NMR
((CD3)2SO) δ 8.18 (3H, broad s), 7.45 (ABq, J AB ) 8.15 Hz, ∆υAB
) 13.8 Hz, 4H), 4.71 (s, 2H), 4.03 (m, 2H); 13C NMR ((CD3)2-
SO) δ 138.67, 134.22, 129.81, 129.75, 42.15, 34.42; electron
ionization MS M+ ) 199 (M - HBr). (Caution: product may
cause allergic reaction.)
Exp er im en ta l Section
{N-[(p-Ca r boxyp h en yl)m eth yl]-N′-[(p-(a m in om eth yl)-
p h en yl)m eth yl]-4,4′-bip yr id in e}2+(2P F 6-). {N-[(p-Carboxy-
phenyl)methyl]4,4′-bipyridine]}+PF6- (4.25 g, 9.74 mmol) and
4-(bromomethyl)benzylamine hydrobromide (2.72 g, 9.74 mmol)
were combined and refluxed in 500 mL of acetonitrile for 48
h. The product, which precipitated as a yellow solid, was
vacuum-filtered while hot and washed with acetonitrile and
diethyl ether. The solid was then dissolved in distilled water,
filtered, and precipitated with ammonium hexafluorophos-
phate. After cooling the solution at 3 °C for several hours,
the white solid was vacuum-filtered and washed with deionized
water, methanol, and diethyl ether to yield the product (4.93
g, 50.9%): decomposes without melting above 175 °C; 1H NMR
(CD3CN) δ 9.05 (d, J ) 6.5 Hz, 4H), 8.45 (d, J ) 6.5 Hz, 4H),
8.13 (d, J ) 7.0 Hz, 2H), 7.60 (m, 6H), 5.93 (s, 2H), 5.87 (s,
2H), 4.18 (s, 2H); 13C NMR ((CD3)2SO) δ 166.26 (carbonyl),
[150.24, 150.19, 145.50, 145.38, 127.31, 127.27] (dipyridyl
carbons), [150.32, 136.95, 133.80, 131.46, 130.30, 130.00,
129.48, 129.07] (phenyl carbons), 63.80 (two methylene car-
bons), 42.81 (C-NH2); positive ion FABMS M+ ) 702 (M
without PF6 counterions). Anal. Calcd (found): C, 36.8%
(37.24%); H, 3.07% (3.16%); N, 4.96% (4.97%).
Ma ter ia ls. All starting materials were used as received
from Aldrich Chemical Co., J . T. Baker, Sigma, or VWR
Scientific. Solvents for high-performance liquid chromatog-
raphy (HPLC) experiments were of HPLC grade and were
degassed with He for 30 min before use.
In str u m en ta tion . 1H NMR spectroscopy was carried out
using either a Bruker AM-300 or a Bruker AC-E-200 spec-
trometer. 13C NMR spectroscopy was carried out using a
Bruker AMX-360 spectrometer. 2-D NOESY NMR spectra
were acquired on a Bruker AMX-2-500 spectrometer. Ana-
lytical HPLC experiments were performed on a Waters 600E
system with a Waters 991 photodiode array detector using a
Crownpak CR+ column purchased from Chiral Technologies
Inc. UV-vis was carried out on a Hewlett-Packard 8452A
Diode Array Spectrophotometer for solution samples and on
a Varian DMS 300 fitted with a diffuse reflectance attachment
for solid samples. Powder X-ray diffraction was performed on
a Philips X′Pert MPD diffractometer. Elemental analysis
(CHN) was performed by Atlantic Microlabs, Inc., P.O. Box
2288, Norcross, GA 30091. Amino acid analysis was performed
by Commonwealth Biotechnologies, Inc. 911 E. Leigh St.,
Richmond, VA 23219.
Syn th esis. {N-[(p-Ca r boxyp h en yl)m eth yl]4,4′-bip yr i-
d in e]}+P F 6-. 4,4′-Bipyridine (4.00 g, 25.6 mmol) and 4-(chlo-
romethyl)benzoic acid (2.90 g, 17.1 mmol) were combined and
heated to reflux in 400 mL of acetonitrile for 24 h. The
product, which precipitated as a white solid, was vacuum-
filtered and washed with acetonitrile and diethyl ether. The
solid was then dissolved in water with the addition of
saturated potassium carbonate (dropwise to pH 8) and pre-
Cesiu m Sa lts of N-t-BOC-a m in o Acid s.23 The cesium
salt of the first amino acid to be coupled to the resin was
obtained by dissolving 2 g of the N-t-BOC-protected amino acid
in 15 mL ethanol/deionized water 2:1 (v/v) and adding aqueous
cesium carbonate to pH 7. After rotary evaporation, the
cesium salt was obtained as a white solid.
Solid -P h a se P ep tid e Syn th esis.24 In a typical coupling
reaction, 5.0 g of Merrifield peptide resin (1.2 mequiv/g), 6.0

7668 J . Org. Chem., Vol. 63, No. 22, 1998
Gavin et al.
mmol of N-t-BOC-amino acid, Cs+ salt, and 40 mL dimethyl-
formamide were combined and stirred at 50 °C overnight. The
resin was filtered and washed with dimethylformamide, di-
methylformamide/deionized water 9:1 (v/v), dimethylforma-
mide, and ethanol. Amino acid loading was determined by
elemental analysis. Values for nitrogen content ranged be-
tween 0.98 and 1.08% depending on the specific amino acid
being coupled. Deprotection of the first amino acid was
accomplished by stirring the resin in 25% trifluoroacetic acid
in chloroform for 30 min. After filtration, the resin was
washed with chloroform, neutralized with 10% triethylamine
in chloroform for 10 min, and washed again with chloroform.
In subsequent coupling steps, a 3-fold excess of N-t-BOC-amino
acid and dichyclohexylcarbodiimide (caution: DCC may cause
allergic reaction.) were added to the peptide resin in 100 mL
of dimethylformamide/dichloromethane 1:1 (v/v). After the
mixture was stirred for several hours, the peptide resin was
filtered and washed with dimethylformamide, dichloromethane,
and ethanol. A standard ninhydrin test verified complete
coupling after each amino acid addition.25
(100 mL), cooled to -15 °C under argon, and neutralized with
N-methylmorpholine (0.122 mL, 1.11 mmol). {N-[(p-Carboxy-
phenyl)methyl]-N′-[(p-(aminomethyl)phenyl)methyl]-4,4′-bi-
pyridine}2+(PF6-)2 (0.765 g, 1.11 mmol) was dissolved in dry
dimethylformamide (100 mL) and treated with triethylamine
(0.155 mL, 1.11 mmol). Isobutyl chlorocarbonate (0.144 mL,
1.11 mmol) was added to the peptide solution, followed, about
a minute later, by the bipyridine solution. The mixture was
allowed to warm to room temperature under argon. The argon
purge was removed, and the solution was stirred in air
overnight. The solvent was removed by slow evaporation at
50 °C at atmospheric pressure.28 The residue was dissolved
in acetonitrile (100 mL), filtered, and precipitated with tetra-
(n-hexyl)ammonium bromide. After cooling the solution at 3
°C for several hours, the solid was vacuum-filtered and washed
with acetonitrile and diethyl ether. The N-t-BOC protecting
group was removed by addition of 25% trifluoroacetic acid in
chloroform (100 mL). Solvent was removed by evaporation at
room temperature and atmospheric pressure. The residue was
then dissolved in acetonitrile, filtered, and precipitated with
tetrahexylammonium bromide. After cooling the solution at
3 °C for several hours, the solid was vacuum-filtered and
washed with acetonitrile and then diethyl ether to yield the
Clea va ge of P r otected P ep tid e fr om Mer r ifield Resin .
Cleavage of the N-t-BOC-protected amino acids was performed
as described previously with changes as noted below.26 In a
typical cleavage reaction, peptide resin (1.00 g) was suspended
in tetrahydrofuran (50 mL). Saturated aqueous potassium
carbonate (5.0 mL) and tetra(n-butyl)ammonium hydrogen
sulfate (0.68 g) were added, and the mixture was stirred for
2-3 h. Deionized water (20 mL) was added, and the resin
was vacuum filtered and washed with deionized water. The
combined filtrates were transferred to a round-bottom flask,
and the organic phase was removed by rotary evaporation. The
aqueous residue was filtered through a 0.2 µm filter and then
acidified to pH 1 with aqueous potassium hydrogen sulfate.
The precipitated solid was filtered and washed with deionized
water to yield the N-t-BOC-protected amino acid. In the
original synthetic procedure, the resin was washed with both
deionized water and ethyl acetate. It was found that the
smaller oligomers of the polystyrene resin were soluble in the
ethyl acetate and therefore were not filtered properly from the
solution prior to acidification. Also, a significant amount of
the resin was able to pass through a fine glass fritted filter.
Initial attempts at the cleavage procedure resulted in elemen-
tal analysis showing a very low nitrogen content, which was
attributed to the sample containing up to 80% resin with only
20% peptide. Filtering the solution through a 0.2 µm filter
prior to acidification removed any residual resin pieces,
resulting in a pure peptide product as shown by elemental
analysis.
1
off-white product (0.765 g, 80.8%): mp dec above 100 °C; H
NMR (C2D6OS) δ 9.55 (d, J ) 6.5 Hz, 4H), 8.79 (d, J ) 6.5 Hz,
4H), 8.45 (m, 2H), 8.22 (d, J ) 6.70 Hz, 1H), 8.13 (m, 1H),
8.04 (d, J ) 7.0 Hz, 2H), 7.71 (d, J ) 7.0 Hz, 2H), 7.60 (d, J )
7.0 Hz, 2H), 7.33 (d, J ) 7.0 Hz, 2H), 6.08 (s, 2H), 5.93 (s,
2H), 4.40 (m, 1H), 4.28 (m, 2H) 3.60 (m, 1H), 2.80 (m, 1H),
2.05 (m, 1H), 1.60 (m, 1H), 1.45 (m, 2H), 1.20 (d, J ) 7.01 Hz,
3H), 0.85 (m, 12H); 13C NMR ((CD3)2SO) δ [174.21, 171.69,
167.86] (carbonyl), 167.06 (carboxylic acid), [149.53, 146.20,
138.88, 131.96, 128.02, 127.66] (dipyridyl carbons), [149.60,
135.89, 130.324, 130.121, 129.51, 129.44, 128.47, 128.41]
(phenyl carbons), [62.80, 62.78] (methylene groups next to
dipyridyl), [57.41, 51.38, 47.90] (chiral center carbons), 41.03
(C-NH2), [24.24, 23.44, 21.98, 19.21, 18.31, 17.35] (amino acid
side chains); positive ion FABMS M+ ) 695 (M without Br
counterions). Anal. Calcd (found): C, 56.21% (56.30%); H,
5.85% (5.72%); N, 9.84% (10.01%).
(S)-Ala -Leu -Va l-Cyclop h a n e2+‚2Br -. {N-[(p-(S)-Ala-Leu-
Val-phenyl)methyl]-N′-[(p-(aminomethyl)phenyl)methyl]-4,4′-
bipyridine}2+(2Br-) (0.765 g, 0.777 mmol) was converted to the
-
2PF6 salt as described above and dissolved in dry dimethyl-
formamide (130 mL, 6 mM solution) at -15 °C under argon
with stirring. N-Methylmorpholine (0.095 mL, 0.855 mmol)
and isobutyl chlorocarbonate (0.111 mL, 0.855 mmol) were
added to the solution followed, about a minute later, by
triethylamine (0.119 mL, 0.855 mmol). The mixture was
allowed to warm to room temperature and stirred in air
overnight. The reaction was cooled to -15 °C again followed
by addition of N-methylmorpholine (0.095 mL, 0.855 mmol),
isobutyl chlorocarbonate (0.111 mL, 0.855 mmol), and triethy-
lamine (0.119 mL, 0.855 mmol) to ensure complete coupling.
After stirring the solution for an additional 48 h, the solvent
was removed by evaporation at 50 °C at atmospheric pressure.
The residue was dissolved in acetonitrile (100 mL), filtered,
and precipitated with tetra(n-hexyl)ammonium bromide. After
the solution was cooled at 3 °C for several hours, the solid was
vacuum-filtered and washed with acetonitrile and diethyl ether
to yield the pale yellow product (0.262 g, 40.4%): mp dec above
120 °C; 1H NMR (C2D6OS) δ [9.60 (d, J ) 6.5 Hz, 4H), 8.80 (d,
J ) 6.5 Hz, 4H)] (dipyridyl), [8.45 (2H, d), 8.20 (1H, d),7.92
(1H, d)] (four -NHCO), [8.02 (d, J ) 7.0 Hz, 2H), 7.77 (d, J )
7.0 Hz, 2H), 7.62 (d, J ) 7.0 Hz, 2H), 7.31 (d, J ) 7.0 Hz, 2H)]
(phenyl), [6.04 (s, 2H), 5.98 (s, 2H)] (two -N+CH2), 4.28 (m,
2H) (Ph-CH2-NH), [4.40 (m, 1H), 3.60 (m, 1H), 2.80 (m, 1H)]
(S)-Ala -Leu -Va l-N-t-BOC. The solid-phase synthesis and
cleavage procedures described above were utilized to yield the
product as a white solid (93% yield): mp dec above 158 °C; 1H
NMR (C2D6OS) δ 8.13 (d, J ) 6.71 Hz, 1H), 7.75 (d, J ) 7.94
Hz, 1H), 6.75 (d, J ) 8.24 Hz, 1H), 4.38 (m, 1H), 4.15 (m, 1H),
3.73 (m, 1H), 1.90 (m, 1H), 1.62 (m, 1H), 1.45 (m, 2H), 1.33 (s,
9H), 1.22 (d, J ) 7.01 Hz, 3H), 0.81 (m, 12H); 13C NMR ((CD3)2-
SO) δ [175.35, 171.97, 171.43, 155.753] (carbonyl), [78.37,
28.47] (BOC carbons), [60.31, 50.78, 46.92] (chiral center
carbons), [24.24, 23.462, 21.862, 19.561, 18.513, 17.462] (amino
acid side chains); positive ion FABMS M+ ) 401. Anal. Calcd
(found): C, 56.90% (56.90%); H, 8.70% (8.76%); N, 10.50%
(10.43%).
{N-[(p-(S)-Ala -Leu -Va l-p h en yl)m eth yl]-N′-[(p-(a m in o-
m eth yl)p h en yl)m eth yl]-4,4′-bip yr id in e}2+(2Br -). A stan-
dard peptide coupling procedure was used to connect {N-[(p-
carboxyphenyl)methyl]-N′-[(p-aminomethylphenyl)methyl]-
4,4′-bipyridine}2+ to the tripeptide.27 (S)-Ala-Leu-Val-N-t-BOC
(0.444 g, 1.11 mmol) was dissolved in dry dimethylformamide
(27) Solvents could not be removed by standard techniques, such
as rotary evaporation or vacuum distillation, because of autoreduction
of the viologen subunit under reduced pressures. Autoreduction was
evident as a loss of viologen protons in the NMR and a change in color
to deep blue.
(24) Sarin, V. K.; Kent, S. B. H.; Tam, J . P.; Merrifield, R. B. Anal.
Biochem. 1981, 117, 147.
(25) Anwer, M. K.; Spatola, A. F. Tetrahedron Lett. 1992, 33, 3121.
(26) The Practice of Peptide Synthesis; Bodansky, M.; Bodansky, A.;
Springer-Verlag: New York, 1984; p 109.
(28) (a) Kumar, A.; Ernst, R. R.; Wuethrich, K. Biochem. Biophys.
Res. Commun. 1980, 95, 1. (b) Bodenhausen, G.; Kogler, H.; Ernst, R.
R. J . Magn. Reson. 1984, 58, 370.

Chiral Molecular Recognition
J . Org. Chem., Vol. 63, No. 22, 1998 7669
(-CH amino acid chiral centers), 2.05 (m, 1H) (-CH valine),
1.60 (m, 1H) (-CH leucine), 1.42 (m, 2H) (-CH2 valine), 1.17
(d, J ) 7.01 Hz, 3H) (-CH3 alanine), 0.80 (m, 12H) (two -CH3
valine and two -CH3 leucine); 13C NMR (MeOD-d4) δ [172.32,
170.08, 169.83, 169.56] (carbonyl), [152.38, 151.91, 148.88,
148.46, 128.31, 127.97] (dipyridyl carbons), [145.43, 144.35,
134.92, 134.50, 132.08, 131.82, 129.87, 129.35] (phenyl car-
bons), [66.82, 62.48] (methylene groups next to dipyridyl),
[57.41, 51.38, 47.90] (chiral center carbons), 40.31 (C-NH2),
[31.92, 26.28, 23.14, 19.68, 19.03, 18.51] (amino acid side
chains); positive ion FABMS M+ ) 676.5 (M without Br
counterions). Anal. Calcd (found): C, 57.4% (57.0%); H, 5.74%
(5.83%); N, 10.0% (9.93%).
1H 1D spectra and phase sensitive 2D NOESY spectra were
also obtained (with presaturation of the residual HDO peak)
for solutions of 0.026 M cyclophane/0.026 M (R)-DOPA and
0.026 M cyclophane/0.026 M (S)-DOPA in the same deuterated
solvent mixture. A mixing time of 500 ms was used in the
NOESY experiments for each of the two solutions.
In ter ca la tion Exp er im en ts. R-Zirconium phosphate was
synthesized and intercalated with tetra(n-butyl)ammonium
hydroxide (TBA+OH-) as previously described.29 Elemental
analysis showed 0.80 mmol of TBA+ per gram of solid (C,
15.03%; H, 3.76%; N, 1.14%). In a typical intercalation
procedure, 1.0 g of 1‚2PF6 (1.0 mmol) in 20 mL of acetonitrile
was added to 1.3 g of TBA-intercalated solid and shaken for
24 h. The solid was then filtered and washed with acetonitrile
to ensure removal of excess 1. Because there is a 2+ charge
on 1, each molecule of 1 should displace two TBA+ ions.
Elemental analysis showed complete displacement of TBA+ by
1 to give 0.40 mmol 1 per gram of solid (C, 17.82%; H, 2.84%;
N, 3.04%). X-ray powder diffraction (XRD) and diffuse reflec-
tance UV-vis spectroscopy were also used to observe the
intercalation of 1 into the solid. XRD showed a layer spacing
of 14.5 Å with the intercalation compound of 1. UV-vis of
the cyclophane-intercalated solid showed a peak at 260 nm;
the same peak was observed in the UV-vis spectrum of 1 in
acetonitrile.
The counterion was changed from Br- to PF6- by dissolving
the solid in water and precipitating with ammonium hexafluo-
rophosphate. Anal. Calcd (found): C, 36.8% (37.23%); H,
3.07% (3.16%); N, 4.96% (4.97%).
Deter m in a tion of Associa tion Con sta n ts. The complex-
ation of each guest was monitored by measuring the change
in the chemical shift of the cyclophane protons upon mixing
of the two components. To determine the association con-
stants, a 5-10 mM stock solution of 1 was made in an
appropriate deuterated solvent or mixture. An amount of solid
guest was weighed out and dissolved in 1-5 mL of the
cyclophane stock solution, depending on solubility. The guest
solution was then titrated with aliquots from the cyclophane
stock solution in order to keep the concentration of 1 constant
while varying the guest concentration. Titrations were con-
tinued until no further changes were observed in the chemical
shifts. A constant volume of 1.0 mL was used in the NMR
tube for each titration. Reported association constants were
calculated from the average of at least two (often 4) indepen-
dent chemical shifts. The approximate equilibrium concentra-
tions of each guest were calculated by the HOSTEST (v.5.0)
program, and an error function was defined. The Newton
method was then used to optimize the calculated concentra-
tions and minimize the error function. Error limits for the
association constants are determined by the program using
Monte Carlo methods.
HP LC Sep a r a tion of DOP A. The enantiomers of DOPA
were separated on an analytical scale using a Crownpak CR-
(+) column at room temperature. The mobile phase was a
mixture of 30 mM perchloric acid and methanol (9:1 v/v) with
a detection wavelength of 200 nm and a flow rate of 0.5 mL/
min. The DOPA enantiomers were baseline resolved (6.61 min
and 10.54 min, R ) 1.59, Rs ) 1.98).
Ack n ow led gm en t. We thank Prof. Craig Wilcox of
the University of Pittsburgh for providing a copy of the
HOSTEST (5.0) program. This work was supported by
a grant from the National Institutes of Health (GM
43844).
NOESY Exp er im en ts.28 2-D NOESY 1H NMR experi-
ments were performed at 300 K on a Bruker AMX-2-500
spectrometer operating in the quadrature mode at a 1H
resonance frequency of 500.13 MHz. 1H 1-D reference spectra
were obtained for 0.026 M (R)-DOPA, 0.026 M 1, and 0.026 M
(S)-DOPA, each prepared in a solvent mixture of deuterated
acetone, DCl, and D2O in 2:1:17 mole ratio. In all cases, the
spectra were obtained with selective presaturation of the large
residual HDO peak. The residual 1H signal of deuterated
acetone (δ(1H) ) 2.04 ppm) was used as an internal reference.
Su p p or tin g In for m a tion Ava ila ble: 1H NOESY data (3
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 O980352C
(29) Garcia, M. E.; Naffin, J . L.; Deng, N.; Mallouk, T. E. Chem.
Mater. 1995, 7, 1968.