Different enantioselective interaction pathways induced by derivatized quinines

Article abstract of DOI:10.1021/jo991661l

The stereochemistries in solution of the diastereoisomeric complexes formed by quinines modified at the hydroxyl site (9-O-acetylquinine; 9-O- (3,5-dimethoxyphenylcarbamate)quinine) or quinuclidine nitrogen (N- benzylquininium chloride) and each enantiomer of 2-(3',5'-dinitrobenzamido)- 1-phenylethanol have been compared to those of the free compounds by ¹H NMR investigations. Completely different interaction models, also involving changes of the free state conformations, have been obtained.

Full text of DOI:10.1021/jo991661l

3596
J . Org. Chem. 2000, 65, 3596-3602
Differ en t En a n tioselective In ter a ction P a th w a ys In d u ced by
Der iva tized Qu in in es
Gloria Uccello-Barretta, Federica Balzano, Cristiana Quintavalli, and Piero Salvadori*
Centro di Studio del CNR per le Macromolecole Stereordinate ed Otticamente Attive,
Dipartimento di Chimica e Chimica Industriale, via Risorgimento 35, 56126 Pisa, Italy
Received October 21, 1999
The stereochemistries in solution of the diastereoisomeric complexes formed by quinines modified
at the hydroxyl site (9-O-acetylquinine; 9-O-(3,5-dimethoxyphenylcarbamate)quinine) or quinuclidine
nitrogen (N-benzylquininium chloride) and each enantiomer of 2-(3′,5′-dinitrobenzamido)-1-
1
phenylethanol have been compared to those of the free compounds by H NMR investigations.
Completely different interaction models, also involving changes of the free state conformations,
have been obtained.
In tr od u ction
by virtue of the presence of several functionalities. With
the aim of understanding the molecular basis of their
strong performances, accurate conformational investiga-
tions have been carried out on them using both compu-
tational and NMR methods.6 The multifunctionality of
Cinchona alkaloids has also given an important aim to
be pursued: the modulation and optimization of interac-
tions responsible for the chiral discrimination by selective
modification of functional groups. Regarding this, suc-
cessful modifications of Cinchona alkaloids have given
efficient ligands for the catalytic asymmetric osmylation
reaction2 and new chiral selectors to be immobilized onto
silica for preparing chiral HPLC stationary phases.7
Cinchona alkaloids analogues containing 9,9′-spirobi-
fluorene moiety, instead of the quinoline ring, have also
been prepared.8
Chiral molecular recognition is currently the subject
of increasing attention and intense research. In this field,
Cinchona alkaloids have demonstrated widespread po-
tentiality, being successfully employed as chiral resolving
agents,1 chiral auxiliaries, catalysts in asymmetric proc-
esses,2,3 new chiral stationary phases for HPLC,4 and as
chiral solvating agents for NMR spectroscopy.5
The versatility of Cinchona alkaloids is commonly
attributed to their ability to act as multisite receptors,
* Ph: +39-50-918203. Fax: +39-50-918409. E-mail: psalva@
dcci.unipi.it.
(1) (a) Newman, P. Optical Resolution Procedures for Chemical
Compounds; Riverdale: New York, 1981. (b) Duke, C. C.; Carter, S.
R.; Hambley, T. W.; Holder, G. M.; Cutler, D. J . Chirality 1994, 6, 670-
680. (c) Chung, J . Y. L.; Hughes, D. L.; Zhao, D.; Song, Z.; Mathre, D.
J .; Ho, G. J .; McNamara, J . M.; Douglas, A. W.; Reamer, R. A.; Tsay,
F. R.; Varsolona, R.; McCauley, J .; Grabowski, E. J . J .; Reider, P. J . J .
Org. Chem. 1996, 61, 215-222.
To define the stereochemical basis of enantiodiscrimi-
nation by quinine derivatives and to understand how
derivatization can be used to enhance the enantioselec-
(2) (a) Nelson, D. W.; Gypser, A.; Tong Ho, P.; Kolb, H. C.; Kondo,
T.; Kwong, H.-L.; McGrath, D. V.; Rubin, A. E.; Norrby, P.-O.; Gable,
K. P.; Sharpless, K. B. J . Am. Chem. Soc. 1997, 119, 1840-1858. (b)
Corey, E. J .; Noe, M. C. J . Am. Chem. Soc. 1996, 118, 11038-11053.
(c) Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B. Chem. Rev.
1994, 94, 2483-2547. (d) Ungva´ry, F. Coord. Chem. Rev. 1995, 141,
371-493. (e) Salvadori, P.; Superchi, S.; Minutolo, F. J . Org. Chem.
1996, 61, 4190-4191. (f) Superchi, S.; Minutolo, F.; Pini, D.; Salvadori,
P. J . Org. Chem. 1996, 61, 3183-3186. (g) Petri, A.; Pini, D.; Salvadori,
P. Tetrahedron Lett. 1995, 36, 1549-1552. (h) Pini, D.; Petri, A.;
Salvadori, P. Tetrahedron 1994, 50, 11321-11328. (i) Petri, A.; Pini,
D.; Rapaccini, S.; Salvadori, P. Chirality 1995, 7, 580-585.
(3) (a) Faber, W. S.; Kok, J .; de Lange, B.; Feringa, B. L. Tetrahedron
1994, 50, 4775-4794. (b) Wynberg, H. Top. Stereochem. 1986, 16, 87-
129. (c) Lee, T. B. K.; Wong, G. S. K. J . Org. Chem. 1991, 56, 872-
875. (d) Gasparski, C. M.; Miller, M. J . Tetrahedron 1991, 47, 5367-
5378. (e) Kobayashi, S.; Tsuchiya, Y.; Mukaiyama, T. Chem. Lett. 1991,
541-544.
(4) (a) Rosini, C.; Bertucci, C.; Pini, D.; Altemura, P.; Salvadori, P.
Tetrahedron Lett. 1985, 26, 3361-3364. (b) Rosini, C.; Altemura, P.;
Pini, D.; Bertucci, C.; Zullino, G.; Salvadori, P. J . Chromatogr. 1985,
348, 79-87. (c) Pini, D.; Rosini, C.; Bertucci, C.; Altemura, P.;
Salvadori, P. Gazz. Chim. Ital. 1986, 116, 603-606. (d) Salvadori, P.;
Rosini, C.; Pini, D.; Bertucci, C.; Uccello-Barretta, G. Chirality 1989,
1, 161-166. (e) Karlsson, A.; Pettersson, C. Chirality 1992, 4, 323-
332. (f) Nesterenko, P. N.; Krotov, V. V.; Storoverov, S. M. J .
Chromatogr., A 1994, 667, 19-28.
1
tivity, we have compared by H NMR spectroscopy the
enantiodiscriminanting capabilities of 9-O-acetylquinine
(Qu iOAc), 9-O-(3,5-dimethoxyphenylcarbamate)quinine
(Qu ica r b), and N-benzylquininium chloride (Qu ibec)
toward the multifunctional chiral analyte 2-(3′,5′-dini-
trobenzamido)-1-phenylethanol (1) (Chart 1). Since the
conformational changes of chiral selectors upon substrate
binding are fundamental in the understanding of mech-
anisms of chiral recognition, we have compared the
(6) (a) Dijkstra, G. D. H.; Kellogg, R. M.; Wynberg, H. J . Org. Chem.
1990, 55, 6121-6131. (b) Dijkstra, G. D. H.; Kellogg, R. M.; Wynberg,
H. Recl. Trav. Chim. Pays-Bas 1989, 108, 195-204. (c) Dijkstra, G. D.
H.; Kellogg, R. M.; Wynberg, H.; Svendsen, J . S.; Marko, I.; Sharpless,
K. B. J . Am. Chem. Soc. 1989, 111, 8069-8076. (d) Oleksyn, B. J .;
Suszko-Purzycka, A.; Dive, G.; Lamotte-Brasseur, J . J . Pharm. Sci.
1992, 81, 122-127. (e) Berg, U.; Aune, M.; Matsson, O. Tetrahedron
Lett. 1995, 36, 2137-2140. (f) Bu¨rgi, T.; Baiker, A. J . Am. Chem. Soc.
1998, 120, 12920-12926.
(7) (a) Bertucci, C.; Rosini, C.; Pini, D.; Salvadori, P. J . Pharm.
Biomed. Anal. 1987, 5, 171-176. (b) Rosini, C.; Bertucci, C.; Pini, D.;
Altemura, P.; Salvadori, P. Chromatographia 1987, 24, 671-676. (c)
Pettersson, C.; Gioeli, C. J . Chromatogr. 1987, 398, 247-254. (d)
Pettersson, C.; Gioeli, C. J . Chromatogr. 1988, 435, 225-228. (e)
La¨mmerhofer, M.; Lindner, W. J . Chromatogr., A 1996, 741, 33-48.
(f) Piette, V.; La¨mmerhofer, M.; Bischoff, K.; Lindner, W. Chirality
1997, 9, 157-161. (g) Kellner, K.-H.; Blasch, A.; Chmiel, H.; La¨mmer-
hofer, M.; Lindner, W. Chirality 1997, 9, 268-273. (h) La¨mmerhofer,
M.; Di Eugenio, P.; Molnar, I.; Lindner, W. J . Chromatogr., B 1997,
689, 123-135.
(5) (a) Rosini, C.; Uccello-Barretta, G.; Pini, D.; Abete, C.; Salvadori,
P. J . Org. Chem. 1988, 53, 4579-4581. (b) Van Oeveren, A.; Menge,
W.; Feringa, B. L. Tetrahedron Lett. 1989, 30, 6427-6430. (c) Salvadori,
P.; Pini, D.; Rosini, C.; Bertucci, C.; Uccello-Barretta, G. Chirality 1992,
4, 43-49. (d) Klein, J .; Borsdorf, R. Fresenius J . Anal. Chem. 1994,
350, 644-646. (e) Klein, J .; Hartenstein, H.; Sicker, D. Magn. Reson.
Chem. 1994, 32, 727-731. (f) Uccello-Barretta, G.; Pini, D.; Mastan-
tuono, A.; Salvadori, P. Tetrahedron: Asymmetry 1995, 6, 1965-1972.
(g) Zymanczyk-Duda, E.; Skwarczynski, M.; Lejczak, B.; Kafarski, P.
Tetrahedron: Asymmetry 1996, 7, 1277-1280.
(8) Winter-Werner, B.; Diederich, F.; Gramlich, V. Helv. Chim. Acta
1996, 79, 1338-1360.
10.1021/jo991661l CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/23/2000

Pathways Induced by Derivatized Quinines
J . Org. Chem., Vol. 65, No. 12, 2000 3597
Ch a r t 1
tion of the chemical shifts of 1 (Table 1). Only Qu iOAc
and Qu ica r b double the NMR signals of the two enan-
tiomers of the analyte 1 remarkably. In these cases, the
relative positions of the signals originated by the two
enantiomers are different. Therefore all three quinines
interact with the substrate 1, but only Qu iOAc and
Qu ica r b display significant enantiodiscrimination, prob-
ably by means of different interaction mechanisms.
Interestingly, both the acetyl Qu iOAc and carbamoyl
Qu ica r b quinine derivatives produce a change in the
coupling pattern of the methylene-methine fragment of
(S)-1 but not (R)-1. The N-benzyl quinine, Qu ibec, does
not change this coupling pattern for (S)- or (R)-1 (Figure
1). These results suggest that the two quinines modified
at the hydroxyl site may change the conformation of the
(S)-enantiomer.
Con for m a tion a l An a lysis by 1H NMR. To under-
stand why the three chiral auxiliaries behave in such a
different way and to identify the interactions that
stabilize the diastereoisomeric species in solution, we
have determined the stereochemistries of the quinine
derivatives and chiral analyte and compared these
conformations to those of the diastereoisomeric species
formed by each chiral auxiliary and (R)- or (S)-1.
Ster eoch em istr ies of F r ee Com p ou n d s. The con-
formation of 1, as represented in Figure 2, has been
established on the basis of the relative intensities of the
Ha-Hb and Ha-Hb′ NOEs, indicating that the CHa proton
is nearer proton Hb and further from proton Hb′. By using
the Altona equation,10 the values of 14° and 113° have
been calculated for the dihedral angles Ha-C-C-Hb and
F igu r e 1. 1H NMR (300 MHz, CDCl3, ppm referred to TMS
as external standard, 25 °C) spectral regions corresponding
to the 3,5-dinitrobenzoyl (Ho and Hp) and methylene (CHbHb′)
proton resonances of 1 (120 mM) in (a) free (R)(S)-1, (b) (R)(S)-
1/Qu iOAc, (c) (R)(S)-1/Qu ica r b, and (d) (R)(S)-1/Qu ibec.
conformations of the quinine derivatives Qu iOAc,
Qu ica r b, and Qu ibec and analyte 1 in the free state
with the conformations of the diastereoisomeric adducts
formed in solution by each quinine and the two enanti-
omers of 1. This investigation has been carried out by
analyzing the intra- and intermolecular dipolar interac-
tions, in CDCl3, by means of 1D NOE and 2D ROESY
techniques.9
Ha-C-C-Hb′ from the vicinal coupling constants (J ab
)
8.5 Hz, J ab′ ) 3.4 Hz). The lack of significant NOEs
between the hydroxyl proton and both methylene protons
indicates that the OH group is directed toward the N-H
proton, probably as a result of the formation of an
intramolecular hydrogen bond NH-OH. Accordingly, no
concentration dependence (100-1 mM) of proton chemi-
cal shifts of 1 has been found.
Resu lts
The conformations of the three quinine derivatives
have been determined on the basis of NOEs, indicated
in Figure 3, and on the basis of the dihedral angle H8-
C9-C8-H9, obtained from the vicinal coupling constant10
between the protons H8 and H9 (J 89 is 7.3, 7.3, and 2 Hz
for Qu iOAc, Qu ica r b, and Qu ibec, respectively).
Ch em ica l Sh ift Non equ iva len ces Mea su r ed in
Mixtu r es Con ta in in g Qu in in e Der iva tives a n d 1.
The comparison of the H NMR spectra in CDCl3 (Figure
1) of the pure chiral analyte (R,S)-1 and its equimolar
mixtures with each quinine derivative clearly shows that
all three chiral auxiliaries produce a remarkable varia-
1
(9) Neuhaus, D.; Willianson, M. The Nuclear Overhauser Effect;
WCH Publishers: New York, 1989.
(10) Haasnoot, C. A. G.; de Leeuw, F. A. A. M.; Altona, C.
Tetrahedron 1980, 36, 2783-2792.

3598 J . Org. Chem., Vol. 65, No. 12, 2000
Uccello-Barretta et al.
Ta ble 1. Non equ iva len ces (∆∆δ,a 300 MHz, CDCl3) a n d Com p lexa tion Sh ifts (∆δ,b 300 MHz, CDCl3) In d u ced in th e
P r oton s of 1 (120 m M) by Equ im ola r Am ou n ts of Qu in in es
∆δb (Hz)
∆∆δa (Hz)
Qu iOAc
Qu ica r b
Qu ibec
(S)-1
proton
Qu iOAc
Qu ica r b
Qu ibec
(S)-1
(R)-1
(S)-1
(R)-1
(R)-1
Ho
Hp
NH
CHa
CH-H
CH-H
29.2
45.5
24.0
8.5
66.2
80.4
23.6
49.1
28.5
9.3
49.6
82.8
6.9
3.0
12.5
3.9
-42.0
-75.7
145.4
-30.7
-69.0
50.6
-12.8
-30.2
121.4
-22.2
-2.8
-37.1
-91.5
186.4
-9.0
-63.8
55.8
-13.5
-42.4
157.9
-18.3
-14.2
-27.0
41.2
-93.6
527.8
-17.4
-54.2
-38.4
34.3
-90.6
515.3
-21.3
-50.1
4.1
-29.8
a
∆δδ ) |δS - δR|, Hz: δS, chemical shift of the (S)-enantiomer of 1 in the presence of the chiral auxiliary; δR, chemical shift of the
b
(R)-enantiomer of 1 in the presence of the chiral auxiliary. ∆δ ) δmixture - δfree, Hz: δmixture, chemical shift measured for each enantiomer
of 1 in the presence of the chiral auxiliary; δfree, chemical shift measured for free 1.
F igu r e 2. Conformations from NMR data of (R)- and (S)-1 in
the free states and in the complexes formed with Qu iOAc,
Qu ica r b, and Qu ibec.
For the two chiral auxiliaries derivatized at the hy-
droxyl group, Qu iOAc and Qu ica r b , we have found
closed conformations for the free state (NOE H9-H5
=
H9-H1, Figure 4a), with the N-quinuclidine nitrogen
pointed toward the quinoline ring. On the other hand,
Qu ibec, with the OH group underivatized, has an open
conformation (NOE H9-H5 > H9-H1) (Figure 3), in which
the quinuclidine nitrogen is oriented away from the
quinoline ring. In all cases the proton H8 is mainly
directed toward the quinoline proton H5 (NOE H8-H5 .
H8-H1, Figure 4a).
Analysis of the chemical shift dependence on total
concentration has demonstrated that the acetylated
quinine, Qu iOAc, is monomeric in solution, whereas the
two quinines having a C9 hydrogen bond donor, Qu ica r b
and Qu ibec, exist as dimers in equilibrium with mono-
mers, with signals that are in fast exchange on the NMR
time scale (Kautoassociation ) 1.8 M-1 for Qu ica r b and 7.7
M-1 for Qu ibec as calculated by fitting dilution NMR
data).
The stereochemistries of the dimers (Figure 5) have
been determined by 2D ROESY analyses in highly
concentrated solutions (80-200 mM). The H3-H5 NOEs
for both quinine derivatives, as well as additional NH-
H2 and NH-H3 NOEs for Qu ica r b or H2-OH and H3-
F igu r e 3. Conformations from NMR data of quinine deriva-
tives in the free states and in the complexes formed with (S)-1
and (R)-1.
OH NOEs for Qu ibec, including the absence of H1-OMe
NOEs, confirmed a head-to-tail autoassociation of the
quinoline rings. The quinuclidine moieties are external
to the dimers, as the H2 and H3 protons have not
produced any NOEs with the quinuclidine nuclei. The
monomeric quinine derivatives have retained the same
conformation that has been found in highly diluted
solution, where the monomer prevails.

Pathways Induced by Derivatized Quinines
J . Org. Chem., Vol. 65, No. 12, 2000 3599
F igu r e 4. 2D ROESY analysis (300 MHz, CDCl3, 25 °C, τm
)
0.6 s) of (a) pure Qu iOAc, (b) mixture (S)-1/Qu iOAc, traces
of H8 and H9 protons.
Ster eoch em istr ies of th e Dia ster eoisom er ic Com -
p lexes F or m ed by (R)- or (S)-1 a n d Qu in in e Der iva -
tives. Some very interesting conformational features
have been observed in the mixtures containing each
quinine derivative and the (S)- or (R)-enantiomer of 1.
In the mixture containing Qu iOAc and (S)-1, both
components change their conformation relative to the free
state. For (S)-1, the methine proton has become equi-
distant from the two methylene protons (Figure 2). The
two diastereotopic methylene protons of (S)-1 have
equivalent coupling constants in the Qu iOAc/1 mixture,
J ab ) J ab′ ) 6.1 Hz, corresponding to the same dihedral
angles Ha-C-C-Hb and Ha-C-C-Hb′ (about 130°). The
NH group is also close to the methine and far away from
the phenyl group given that only a NH-methine NOE has
been detected. Therefore, the intramolecular NH-OH
hydrogen bond present in the free state must be broken
in the complex Qu iOAc/1. The acetylated quinine,
Qu iOAc, assumes an open conformation (Figure 3) in
which the quinuclidine nitrogen undergoes a clockwise
rotation, withdrawing it from the quinoline plane, ac-
cording to changes of the relative intensities of the H9-
H5 and H9-H1 NOEs (Figure 4b).
F igu r e 5. Conformations of the dimers of Qu ica r b and
Qu ibec from NMR data.
produce small but detectable NOEs with H2 and H3
adjacent to the quinoline nitrogen. Similar dipolar in-
teractions are also generated by one methylene proton.
Finally, an intermolecular NOE between the NH proton
of (R)-1 and the H8 proton of Qu iOAc has also been
detected. No NOEs between (R)-1 and quinuclidine
protons have been observed. Therefore, the Qu iOAc-
(R)-1 interaction takes place at the quinoline plane at
the opposite side of the quinuclidine nitrogen. The (R)-
enantiomer of the analyte maintains its intramolecular
hydrogen bond NH-OH and probably interacts with the
quinoline nitrogen or acetyl function. These interactions
can be assisted by the π-π attraction between the
aromatic moieties of the analyte and the quinoline plane.
The different type and strength of the interactions for
the two diastereoisomeric complexes, each having 1:1
stoichiometry (see Experimental Section), are well re-
flected in their respective association constants, which
are K ) 41.2 M-1 for (S)-1/Qu iOAc and K ) 13.4 M-1
for (R)-1/Qu iOAc.
In the diastereomeric mixture containing Qu iOAc and
the (R)-1 both species have retained their free state
conformations (Figures 2 and 3).
The origin of the different spectra for the diastereo-
meric complexes Qu iOAc/(S)-1 and Qu iOAc/(R)-1 is
understandable on the basis of the different intermolecu-
lar NOEs (Figure 6). The (S)-enantiomer acts as a
bidentate ligand interacting by means of its NH group
with the quinuclidine nitrogen and by its OH group with
the acetyl function of Qu iOAc. These interactions are
indicated by the NOEs NH-H8, CHa-H8, CHa-H15, 3,5-
dinitrobenzoyl-OMe. The two intermolecular H-bond
interactions probably allow (S)-1 to change its conforma-
tion and to break the intramolecular hydrogen bond
present in its free state. The hydrogen bond between the
NH group of (S)-1 and the quinuclidine nitrogen also
justifies the “opening” of the conformation of quinine.
In the other diastereoisomeric adduct, Qu iOAc/(R)-1
(see Figure 6), the 3,5-dinitrobenzoyl and NH protons

3600 J . Org. Chem., Vol. 65, No. 12, 2000
Uccello-Barretta et al.
F igu r e 7. Representation of the diastereoisomeric complexes
formed by Qu ica r b and (S)- or (R)-1.
F igu r e 6. Representation of the diastereoisomeric complexes
formed by Qu iOAc and (S)- or (R)-1.
hydrogen bonded moiety directed toward the quinoline
nitrogen (Figure 7).
It is noteworthy that, even if Qu ica r b is present in
solution in a monomer-dimer equilibrium, the interac-
tion with both enantiomers of the analyte only involves
the monomer.
In the case of the carbamoyl derivative of quinine,
Qu ica r b, a similar behavior for the two diastereoiso-
meric adducts has been found. The interaction between
(S)-1 and Qu ica r b causes a change in the conformation
of both components (Figures 2 and 3), analogous to that
observed for Qu iOAc/(S)-1, whereas there is little inter-
action of Qu ica r b with (R)-1, as the conformations of
both compounds remain unchanged from the free state.
However, the intermolecular interactions involved in
the formation of the diastereoisomeric adducts are rather
different (Figure 7). For the Qu ica r b/(S)-1 pair, one of
the quinine sites interacting with the chiral analyte is
once again the quinuclidine nitrogen. This explains why
the quinine derivative changes its conformation to open.
However, it is the hydroxyl of (S)-1 that interacts with
the Qu ica r b quinuclidine nitrogen and not the NH
group, as has been found for the acetyl quinine, Qu iOAc.
In fact, the NOE determinations indicate a proximity
between the 3,5-dimethoxyphenyl group of Qu ica r b and
the 3,5-dinitrophenyl group of the analyte. Moreover, the
protons NH, CHa, Ho, and Hp of (S)-1 all produce NOEs
with the proton H8. It is noteworthy that we have
measured a reproducible CHa-H15 NOE that indicates
the proximity of proton CHa to the quinuclidine moiety.
Therefore, it can be concluded that three interactions
are important: (1) a hydrogen bond between the OH
group of (S)-1 and the quinuclidine nitrogen of Qu ica r b;
(2) a hydrogen bond between the NH group of (S)-1 and,
probably, the CdO group of the Qu ica r b carbamate; (3)
a π-π interaction between the 3,5-dinitrophenyl and the
3,5-dimethoxyphenyl groups.
Finally, no conformational changes of both components
are observed for mixtures containing the quinine deriva-
tized at the quinuclidine nitrogen (Qu ibec) and either
(R)-1 or (S)-1. Very similar intermolecular NOEs have
been detected between protons of Qu ibec and protons
of (R)-1 or (S)-1, so as to indicate that the stereochem-
istries of the two diastereoisomeric adducts must be very
similar. This result is in keeping with the rather low
extent of enantiodiscrimination found in the mixtures
containing the chiral auxiliary and the racemate of the
chiral analyte.
Furthermore, the protons Ho and Hp of the 3,5-
dinitrobenzoyl group of (R)-1 originate dipolar interac-
tions with all of the quinoline protons H1-H5, but the
ortho protons Ho have also given NOEs with H20 and H10
quinuclidine protons. Therefore, it seems reasonable to
conclude that (R)-1 faces the more hindered side of
Qu ibec, the side containing the quinuclidine ring. How-
ever, an apparently incongruous NOE between the ortho
protons of the 3,5-dinitrobenzoyl group of (R)-1 and the
OH proton of Qu ibec has also been detected, according
to a structure in which the substrate approaches the
opposite side of Qu ibec. Furthermore, the proton H2 of
Qu ibec gives remarkable NOEs to the CH, CH2, and the
phenyl protons of (R)-1. Conversely, the CH proton of
(R)-1 shows clear dipolar interactions only with the
quinoline protons H2 and H3. Taking into account that it
is quite difficult to explain why (R)-1 should approach
the more hindered quinoline side, it seems reasonable
to suppose an interaction model as shown in Figure 8:
(R)-1 interacts directly with the dimer of Qu ibec. In this
way (R)-1 can bring its 3,5-dinitrobenzoyl moiety simul-
For the other enantiomer, (R)-1, the interaction with
Qu ica r b takes place at the less hindered face of the
quinine derivative. On the basis of the NOE data, the
(R)-enantiomer, in its free state conformation, faces the
quinoline plane at the side opposite to that of the
quinuclidine nitrogen, with the 3,5-dinitrophenyl moiety
bent at the 3,5-dimethoxyphenyl and with its NH-OH

Pathways Induced by Derivatized Quinines
J . Org. Chem., Vol. 65, No. 12, 2000 3601
acid-type substrates and in governing the enantiodis-
criminating pathways, but also great importance seems
to rely on the ground-state stereochemistry, as the free
state open conformation can determine a greater acces-
sibility to the same site.
Our results reveal two important findings. At least for
the cases investigated, quinine derivatives having an
available quinuclidine nitrogen result in the highest
extent of enantiodiscrimination, even if these have a
closed conformation in the free state (Qu iOAc and
Qu ica r b). The quaternarization of this quinuclidine site,
as for Qu ibec, results in a completely different interac-
tion mechanism with the chiral analyte, resulting in a
lower degree of enantiodifferentiation but not labilization
of the diastereoisomeric species. Finally, both the strength
and the nature of the interactions with the analyte are
greatly affected by monomer-dimer self-assemby of the
quinines.13
The latter point gets to the root of chiral recognition
mechanisms, not only for quinines but also for other
classes of chiral selectors. The frequent hypothesis that
the free state conformation of a ligand is retained in the
diastereoisomeric adduct with a chiral selectand may be
misleading. In fact, in our cases the formation of the most
stable diastereoisomeric adducts (S)-1/Qu iOAc and (S)-
1/Qu ica r b involves significant conformational changes
for the adducts relative to the free compounds. Similarly,
misleading chiral discrimination mechanisms may also
originate from the hypothesis that the knowledge of the
stereochemistry for the components of only one dia-
stereoisomeric species, the most stable one, can be
assumed as a basis for extrapolating the stereochemistry
of the less stable diastereoisomer. In fact, completely
different interaction pathways have been found for the
two enantiomers of 1 in the presence of the same chiral
selectand.
F igu r e 8. Representation of the diastereoisomeric complex
formed by Qu ibec and (R)-1.
taneously close to the quinuclidine ring of one Qu ibec
monomer and to the OH group of the other Qu ibec
monomer. This model also explains the proximity be-
tween the ethanolamine portion of 1 and the H2 and H3
protons of Qu ibec. It is noteworthy that the 2D ROESY
map also reveals NOEs due to the presence of the Qu ibec
dimer.
Exp er im en ta l Section
All spectra were recorded using a spectrometer operating
at 300 MHz for H, and the temperature was controlled to (0.1
1
°C.
The 2D NMR spectra were obtained by using standard
sequences. The double-quantum-filtered (DQF) COSY experi-
ments were recorded with a spectral width of 3300 Hz; 512
increments of 8 scans and 2K data points were acquired. The
relaxation delay was 5 s. The data were zero-filled to 2K ×
1K, and a Gaussian function was applied for processing in both
dimension. The phase-sensitive ROESY9 spectra were acquired
with a spectral width of 3300 Hz in 2K data points using 8
scans for each of the 512 t1 increments, with a mixing time of
600 or 300 ms. A Gaussian function was applied for processing
in both dimension.
The 1H{1H}-NOE experiments were performed in the dif-
ference mode. The decoupler power used was the minimum
required to saturate the spin of interest. A waiting time of
5-10 s was used to allow the system to reach the equilibrium.
Each NOE experiment was repeated at least four times. All
of the solutions were accurately degassed by freeze-pump-
thaw cycles for 1D and 2D NOE experiments.
The stoichiometries were determined14 by measuring the
chemical shift of quinine in solutions prepared by mixing
different volumes of stock solutions of each component having
Con clu sion
For the naturally occurring quinine and quinidine
alkaloids an open conformation is preferred in solution.
In NMR investigations regarding chiral recognition
phenomena,11 this open conformation is believed to be
responsible for the enantiodiscriminating ability of the
alkaloids, as the unhindered quinuclidine nitrogen is able
to generate strong interactions with selectands, even if,
in some cases, the quinuclidine nitrogen has not dem-
onstrated involvement, probably because of the unneg-
ligible bulkiness of the quinuclidine moiety.12
Therefore, not only is the relevance of quinuclidine
nitrogen, the most basic site, clearly recognized in
determining the strength of alkaloid interactions with
(11) Reeder, J .; Castro, P. P.; Knobler, C. B.; Martinborough, E.;
Owens, L.; Diederich, F. J . Org. Chem. 1994, 59, 3151-3160.
(12) (a) Salvadori, P.; Rosini, C.; Pini, D.; Bertucci, C.; Altemura,
P.; Uccello-Barretta, G.; Raffaelli, A. Tetrahedron 1987, 43, 4969-4978.
(b) Uccello-Barretta, G.; Pini, D.; Rosini, C.; Salvadori, P. J . Chro-
matogr., A 1994, 666, 541-548.
(13) (a) Williams, T.; Pitcher, R. G.; Bommer, P.; Gutzwiller, J .;
Uskokovic, M. J . Am. Chem. Soc. 1969, 91, 1871-1872. (b) Uccello-
Barretta, G.; Di Bari, L.; Salvadori, P. Magn. Reson. Chem. 1992, 30,
1054-1063.
(14) (a) J ob, P. Ann. Chem. 1928, 9, 113-134. (b) Homer, J .; Perry,
M. C. J . Chem. Soc., Faraday Trans. 1 1986, 82, 533-543.

3602 J . Org. Chem., Vol. 65, No. 12, 2000
Uccello-Barretta et al.
the same molar concentration M to obtain a prefixed volume
V directly in the NMR tube. Owing to the solubility of 1, the
stoichiometry was determined on quite dilute solutions (M )
0.015).
The autoassociation constants were determined by analyz-
ing the dependence of the chemical shifts from the total
concentration in the range 200-2 mM for Qu ica r b and 120-4
mM for Qu ibec.13b
m), 1.79 (1H, H10, m), 1.80 (1H, H12, m), 2.21 (1H, H17, m),
2.58 (1H, H18, m), 2.58 (1H, H16, m), 2.98 (1H, H19, m), 3.06
(1H, H15, m), 3.27 (1H, H9, m), 3.66 (6H, 2 OMec, s), 3.89 (3H,
OMe, s), 4.96 (1H, H22, d, J 22-20 ) 18.7 Hz), 4.97 (1H, H21, d,
J 21-20 ) 11.4 Hz), 5.76 (1H, H20, ddd, J 20-22 ) 18.7 Hz, J 20-21
) 11.4 Hz, J 20-17 ) 8.5 Hz), 6.13 (1H, Hpc, t, J p-o ) 1.6 Hz),
c
6.50 (1H, H8, d, J 8-9 ) 7.3 Hz), 6.57 (2H, Ho , d, J o-p ) 1.6
Hz), 7.28 (1H, H1, d, J 1-2 ) 4.9 Hz), 7.31 (1H, H4, dd, J 4-3
)
The heteroassociation constants for the two diastereoiso-
meric complexes (S)-1/Qu iOAc and (R)-1/Qu iOAc were de-
termined by nonlinear fitting15 of the experimental data
(concentration vs chemical shift) obtained from the analysis
of the proton NMR spectra acquired for two sets of solutions
progressively diluted from 122.5 to 0.4 mM, containing equimo-
lar amounts of 1 and Qu iOAc.
Melting points were determined using a Koffler hot-stage
apparatus. Optical rotations were measured using a Perkin-
Elmer 142 polarimeter.
9.4 Hz, J 4-5 ) 2.4 Hz), 7.45 (1H, H5, d, J 5-4 ) 2.4 Hz), 7.97
(1H, H3, d, J 3-4 ) 9.4 Hz), 8.65 (1H, H2, d, J 2-1 ) 4.9 Hz).
Anal. Calcd for C29H33O5N3: C, 69.17; H, 6.61; N, 8.34.
Found: C, 69.20; H, 6.58; N, 8.32.
N-Ben zylqu in in iu m Ch lor ide (Qu ibec). 1H NMR (CDCl3,
25 °C, 80 mM, ppm referred to TMS as external standard) 1.46
(1H, H10, m), 1.64 (1H, H13, m), 1.93 (1H, H12, m), 2.18 (1H,
H
11, m), 2.24 (1H, H14, m), 2.46 (1H, H17, m), 2.98 (1H, H16
m), 3.37 (1H, H19, m), 3.53 (1H, H18, m), 3.89 (1H, H9, m), 3.93
(3H, OMe, s), 4.76 (1H, CH2Bz, d, J ) 12.2 Hz), 4.88 (1H, H15
,
,
Ma ter ia ls. N-benzylquininium chloride (Qu ibec) was pur-
chased from Fluka. L-(+)-and D-(-)-mandelic acid were ob-
tained from Aldrich. All chemicals were purified prior to use
by standard methods.16 (R)- and (S)-2-(3′,5′-dinitrobenzamido)-
1-phenylethanol [(R)- and (S)-1] have been prepared starting
from (R)- and (S)-2-hydroxy-1-phenylethanol [(R)- and (S)-2],
respectively, as previously reported.17
m), 4.88 (1H, H21, d, J 21-20 ) 10.6 Hz), 5.02 (1H, H22, d, J 22-20
) 17.0 Hz), 5.52 (1H, H20, ddd, J 20-22 ) 17.0 Hz, J 20-21 ) 10.6
Hz, J 20-17 ) 6.9 Hz), 5.99 (1H, CH2Bz, d, J ) 12.2 Hz), 6.56
(1H, H8, d, J 8-OH ) 5.7 Hz), 7.23 (1H, H4, dd, J 4-3 ) 8.9 Hz,
J 4-5 ) 2.7 Hz), 7.27 (1H, Hp, m), 7.27 (1H, H5, d, J 5-4 ) 2.7
Hz), 7.29 (1H, d, J OH-8 ) 5.7 Hz), 7.33 (2H, Hm, m), 7.66 (1H,
H1, d, J 1-2 ) 4.9 Hz), 7.69 (2H, Ho, d, J o-m ) 6.9 Hz), 7.92
(1H, H3, d, J 3-4 ) 8.9 Hz), 8.62 (1H, H2, d, J 2-1 ) 4.9 Hz).
9-O-Acetylqu in in e (Qu iOAc).18 According to literature
methods, Qu iOAc was obtained starting from quinine in 92%
1
yield: mp 116-117 °C (lit.18 115-117 °C); H NMR (CDCl3,
(R)- a n d (S)-2-(3′,5′-Din itr oben za m id o)-1-p h en yleth a -
n ol [(R)- a n d (S)-1]. 1H NMR (CDCl3, 25 °C, ppm referred to
TMS as external standard) 2.92 (1H), 3.56 (1H), 4.01 (1H), 5.00
(1H), 6.88 (1H), 7.30-7.50 (5H), 8.94 (2H), 9.16 (1H).
25 °C, 120 mM, ppm referred to TMS as external standard)
1.45 (1H, H11, m), 1.49 (1H, H13, m), 1.64 (1H, H14, m), 1.80
(1H, H12, m), 1.80 (1H, H10, m), 2.07 (3H, OAc, s), 2.21 (1H,
H
17, m), 2.55 (1H, H18, m), 2.60 (1H, H16, m), 2.98 (1H, H19
m), 3.05 (1H, H15, m), 3.32 (1H, H9, m), 3.91 (3H, OMe, s), 4.96
(1H, H21, dd, J 21-20 ) 10.6 Hz, J 21-22 ) 1.6 Hz), 4.97 (1H, H22
,
(R)- a n d (S)-2-Hyd r oxy-1-p h en yleth a n ol [(R)- a n d (S)-
2].19 To a stirred suspension of LiAlH4 (66.0 mmol) in dry Et2O
(150 mL) was added D-(-)-mandelic acid (32.9 mmol). The
reaction mixture was refluxed for 6 h, allowed to stir at room
temperature for 2 h, treated with ice water, and then filtered.
After the usual workup, (R)-2 was recovered (70% yield) by
,
dd, J 22-20 ) 17.0 Hz, J 22-21 ) 1.6 Hz), 5.80 (1H, H20, ddd, J 20-22
) 17.0 Hz, J 20-21 ) 10.6 Hz, J 20-17 ) 7.7 Hz), 6.44 (1H, H8, d,
J 8-9 ) 7.3 Hz), 7.30 (1H, H1, d, J 1-2 ) 4.5 Hz), 7.32 (1H, H4,
dd, J 4-3 ) 8.9 Hz, J 4-5 ) 2.9 Hz), 7.40 (1H, H5, d, J 5-4 ) 2.9
Hz), 7.96 (1H, H3, d, J 3-4 ) 8.9 Hz), 8.69 (1H, H2, d, J 2-1 ) 4.5
Hz).
recrystallization from Et2O/pentane (1:2): mp 66 °C (lit.19 66-
1
67 °C); [R]21 -39.3 (c 1.26, EtOH); H NMR (CDCl3, 25 °C,
D
P r ep a r a tion of 9-O-(3,5-Dim eth oxyp h en ylca r ba m a te)-
qu in in e (Qu ica r b). 3,5-Dim eth oxyben zoyl Azid e. To a
stirred suspension of 3,5-dimethoxybenzoic acid (50 mmol) in
anhydrous toluene (60 mL) was added a solution of oxalyl
chloride (70 mmol) in anhydrous acetone (40 mL). The reaction
mixture, refluxed until no more gas development was observed,
was allowed to warm to 50 °C, and the solvents were removed.
Then to the crude residue, dissolved in anhydrous acetone (120
mL), was added a solution of sodium azide (184 mmol) in H2O
(50 mL), and the reaction mixture was stirred for 12 h. The
organic materials were extracted with CH2Cl2. After the usual
workup, 3,5-dimethoxybenzoyl azide was recovered (77%
yield): 1H NMR (CDCl3, 25 °C, ppm referred to TMS as
ppm referred to TMS as external standard) 2.30 (1H), 2.75
(1H), 3.60-3.80 (2H), 4.85 (1H), 7.20-7.50 (5H). Anal. Calcd
for C8H10O2: C, 69.54; H, 7.30. Found for (R)-2: C, 69.57; H,
7.24.
According to the above procedure, from L-(+)-mandelic acid
was obtained (S)-2: 70% yield, [R]21 +39.3 (c 1.26, EtOH)
D
according to literature.19 Anal. Calcd for C8H10O2: C, 69.54;
H, 7.30. Found for (S)-2: C, 69.52; H, 7.25.
Ack n ow led gm en t . This work was supported by
the Ministero della Ricerca Scientifica e Tecnologica
(MURST) and CNR, Italy.
external standard) 3.81 (6H, s, 2 OMe), 6.65 (1H, Hp, t, J p-o
)
2.0 Hz), 7.22 (2H, Ho, d, J o-p ) 2.0 Hz). Anal. Calcd for
C9H9O3N3: C, 52.17; H, 4.38; N, 20.28. Found: C, 52.19; H,
4.36; N, 20.32.
Su p p or tin g In for m a tion Ava ila ble: Tables of vicinal
coupling constants and calculated dihedral angles for (S)- or
(R)-1 and quinine in the free state and in equimolar mixtures.
1H{1H}-NOE difference spectra of 1 in CDCl3. Traces of 300
MHz ROESY spectra of each quinine in the free state and in
the equimolar mixtures containing (R)- or (S)-1 in CDCl3. Plot
of concentration against chemical shifts for the determination
of the autoassociation constants of Qu ica r b and Qu ibec. J ob
plot for the determination of the stoichiometry of (S)-1/
Qu iOAc and (R)-1/Qu iOAc. Plot of concentration against
chemical shifts for the determination of association constants
of (S)-1/Qu iOAc and (R)-1/Qu iOAc. This material is available
free of charge via the Internet at http://pubs.acs.org.
9-O-(3,5-Dim eth oxyph en ylcar bam ate)qu in in e (Qu icar b).
A solution of 3,5-dimethoxybenzoyl azide (39 mmol) in anhy-
drous toluene (170 mL) was stirred at 110 °C until no more
gas development was observed, and then quinine (33.2 mmol)
was added. The reaction mixture was refluxed for a further 3
h. After the usual workup, Qu ica r b was recovered (69% yield)
by recrystallization from THF/n-pentane: mp 119-121 °C; 1H
NMR (CDCl3, 25 °C, 200 mM, ppm referred to TMS as external
standard) 1.46 (1H, H13, m), 1.51 (1H, H11, m), 1.63 (1H, H14
,
(15) Uccello-Barretta, G.; Balzano, F.; Caporusso, A. M.; Iodice, A.;
Salvadori, P. J . Org. Chem. 1995, 60, 2227-2231.
(16) Perrin, D. D.; Armarego, W. L. F.; Perrin, D. R. Purification of
Laboratory Chemicals; Pergamon Press: Toronto, 1983.
(17) Uccello-Barretta, G.; Cuzzola, A.; Balzano, F.; Menicagli, R.;
Iuliano, A.; Salvadori, P. J . Org. Chem. 1997, 62, 827-835.
(18) J aeger, F. M. Verslag. Akad. Wetenschappen Amsterdam 1926,
35, 61-74.
J O991661L
(19) King, R. B.; Bakos, J .; Hoff, C. D.; Marko´, L. J . Org. Chem.
1979, 44, 1729-1731.