A combined NMR, DFT, and X-ray investigation of some cinchona alkaloid O-ethers

Article abstract of DOI:10.1021/jo8008462

(Chemical Equation Presented) Structures and conformational behavior of several cinchona alkaloid O-ethers in the solid state (X-ray), in solution (NMR and DFT), and in the gas phase (DFT) were investigated. In the crystal, O-phenylcinchonidine adopts the Open(3) conformation similar to cinchonidine, whereas the O-methyl ether derivatives of both cinchonidine and cinchonine are packed in the Closed(1) conformation. Dynamic equilibria in solutions of the alkaloids were revealed by combined experimental-theoretical spin simulation/iteration techniques for the first time. In the ¹H NMR spectra in CDCl₃ and toluene-d₈ at room temperature, Closed(1) conformation was observed for the O-silyl ethers as a separate set of signals. For O-methyl ether derivatives Closed(1) could be separated only at -30°C in CDCl₃ or toluene-d8 and for O-phenylcinchonidine at -70°C in CDCl₃/CD₂Cl₂. The ratio between the Closed(2) and Open(3) conformers was estimated by analyzing the vicinal coupling constant ³J_H9,H8 at ambient and low temperatures. The observed conformational equilibria of O-(tert-butyldimethylsilyl) cinchonidine in CDCl₃ and toluene-d₈ are in good agreement with the theoretically estimated equilibrium populations of the conformations according to Boltzmann statistics. The conformational equilibria of four cinchona alkaloid O-ether solutes in CDCl₃ and toluene-d₈ are discussed in the light of their relevance to the mechanism of 1-phenyl-1,2-propanedione (PPD) hydrogenation over cinchona alkaloid modified heterogeneous platinum catalysts. It was demonstrated that the conformation found to be abundant in the liquid phase has no direct correlation with the enantioselectivity of the PPD hydrogenation reaction.

Full text of DOI:10.1021/jo8008462

A Combined NMR, DFT, and X-ray Investigation of Some Cinchona
Alkaloid O-Ethers
Igor Busygin,† Ville Nieminen,‡ Antti Taskinen,‡ Jari Sinkkonen,§ Esa Toukoniitty,‡
Reijo Sillanpa¨a¨, Dmitry Yu. Murzin,‡,* and Reko Leino†,*
Laboratories of Organic Chemistry and Industrial Chemistry and Reaction Engineering, Åbo Akademi
UniVersity, FI-20500 Turku, Finland, Department of Chemistry, UniVersity of Turku, FI-20014 Turku,
Finland, and Department of Chemistry, UniVersity of JyVa¨skyla¨, FI-40351 JyVa¨skyla¨, Finland
dmitry.murzin@abo.fi; reko.leino@abo.fi
ReceiVed April 18, 2008
Structures and conformational behavior of several cinchona alkaloid O-ethers in the solid state (X-ray), in solution
(NMR and DFT), and in the gas phase (DFT) were investigated. In the crystal, O-phenylcinchonidine adopts the
Open(3) conformation similar to cinchonidine, whereas the O-methyl ether derivatives of both cinchonidine and
cinchonine are packed in the Closed(1) conformation. Dynamic equilibria in solutions of the alkaloids were revealed
by combined experimental-theoretical spin simulation/iteration techniques for the first time. In the 1H NMR spectra
in CDCl3 and toluene-d8 at room temperature, Closed(1) conformation was observed for the O-silyl ethers as a
separate set of signals. For O-methyl ether derivatives Closed(1) could be separated only at -30 °C in CDCl3 or
toluene-d8 and for O-phenylcinchonidine at -70 °C in CDCl3/CD2Cl2. The ratio between the Closed(2) and Open(3)
conformers was estimated by analyzing the vicinal coupling constant 3JH9,H8 at ambient and low temperatures. The
observed conformational equilibria of O-(tert-butyldimethylsilyl)cinchonidine in CDCl3 and toluene-d8 are in good
agreement with the theoretically estimated equilibrium populations of the conformations according to Boltzmann
statistics. The conformational equilibria of four cinchona alkaloid O-ether solutes in CDCl3 and toluene-d8 are
discussed in the light of their relevance to the mechanism of 1-phenyl-1,2-propanedione (PPD) hydrogenation over
cinchona alkaloid modified heterogeneous platinum catalysts. It was demonstrated that the conformation found to
be abundant in the liquid phase has no direct correlation with the enantioselectivity of the PPD hydrogenation
reaction.
1. Introduction
and in catalytic applications as catalysts and as chiral auxiliaries.
To the most studied applications of cinchona alkaloids belong
their use as chiral modifiers in heterogeneous catalysis1 and as
organocatalysts for asymmetric homogeneous transformations.2
Cinchona alkaloids are an important class of natural chiral
compounds that have found wide use, e.g., as pharmaceuticals
and chiral resolving agents for separation of optical isomers
(1) (a) Mallat, T.; Orglmeister, E.; Baiker, A. Chem. ReV. 2007, 107, 4863–
4890. (b) Barto´k, M. Curr. Org. Chem. 2006, 10, 1533–1567. (c) Murzin, D.
Yu.; Ma¨ki-Arvela, P.; Toukoniitty, E.; Salmi, T. Catal. ReV. 2005, 47, 175–
256. (d) Hutchings, G. J. Annu. ReV. Mater. Res. 2005, 35, 143–166. (e) Studer,
M.; Blaser, H.-U; Exner, C. AdV. Synth. Catal. 2003, 345, 45–65. (f) Baddeley,
C. J. Top. Catal. 2003, 25, 17–28.
† Laboratory of Organic Chemistry, Åbo Akademi University.
‡ Laboratory of Industrial Chemistry and Reaction Engineering, Åbo Akademi
University.
§ University of Turku.
University of Jyva¨skyla¨.
10.1021/jo8008462 CCC: $40.75
Published on Web 08/07/2008
2008 American Chemical Society
J. Org. Chem. 2008, 73, 6559–6569 6559

Busygin et al.
SCHEME 1. Enantioselective Hydrogenation of 1 over
Pt/Al2O3 Modified by Cinchona Alkaloid Derivatives
the conformational diversity of the alkaloid. Several open and
closed conformations are known for cinchonidine,6 as studied
by both NMR6a–c and molecular modeling.6a,d,e It is well-known
that in apolar solvents cinchonidine preferentially adopts the
so-called Open(3) conformation with the population of ap-
proximately 60-70%.6a A substituent at the C(9)-O- position,
however, hinders the rotation around the C4′-C9 and C9-C8
bonds (for atom numbering see Supporting Information) and
alters the conformational equilibria.6b The modified conforma-
tions are illustrated in Figure 1 as Newman projections.
In most mechanistic proposals for chirally modified hydro-
genation catalysts, cinchonidine is assumed to interact with the
carbonyl substrate in Open(3) conformation,7a–f its most abun-
dant conformer in liquid-phase.6 A correlation between the
population of Open(3) conformation in different solvents and
the experimental enantiomeric excess (ee) observed in the
hydrogenation of ketopantolactone has been reported.6a Modifier
rigidity has also been taken as a proof of the crucial role of the
Open(3) conformation.8 The phenomenon of switching the
enantiodifferentiating properties of the catalyst by changing the
modifier structure has stimulated an interest for studying the
conformational properties of cinchona alkaloid O-ethers. Baiker
and co-workers have studied the adsorption of O-phenylcin-
chonidine (PhOCD),9a,b O-(trimethylsilyl)cinchonidine,9c and
O-methylcinchonidine9c on platinum by DFT and ATR-IR, as
well as the conformational equilibria of PhOCD in different
solvents by vibrational circular dichroism (VCD) spectroscopy.9c
Numerous derivatives of naturally occurring (-)-cinchonidine
(CD) and (+)-cinchonine (CN) have been synthesized and
utilized in both heterogeneous1,3 and homogeneous2,4 enanti-
oselective reactions in order to approach the mechanism of chiral
induction.
Recently, O-ether derivatives of cinchona alkaloids have
attracted considerable attention, as in several hydrogenation
reactions over chirally modified surfaces, they are capable of
inverting the sense of enantioselectivity when compared to the
reaction over the parent alkaloid modified surface.5 In the present
study, we have investigated the substituent effect on the
conformational equilibria of some cinchona alkaloid O-ethers
as well as on their structural behavior upon crystallization. In
an earlier related investigation, we employed nine derivatives
of cinchonidine and cinchonine possessing substituents with
different nature and bulkiness as chiral modifiers in the
enantioselective hydrogenation of 1-phenyl-1,2-propanedione
(PPD, 1) (Scheme 1).5e Here, we present the synthesis of these
modifiers, together with their detailed structural characterization
and conformational studies by X-ray crystallographic structure
determinations, by NMR spectroscopic analysis, and by DFT
calculations.
Despite significant progress in the experimental techniques
and theoretical methods, the mechanism of enantiodifferentiation
in the hydrogenation of activated ketones over cinchona alkaloid
modified metal catalysts remains unclear in many respects.1
Several models have been proposed in order to describe the
origin of enantioselectivity over chirally modified metal cata-
lysts.7 The majority of the earlier mechanistic models have
utilized the chiral modifier with its quinoline aromatic system
adsorbed parallel to the metal surface. It is commonly assumed,
although not experimentally confirmed, that the π-bonded
modifier species interacts with the substrate on the metal surface,
thus steering the production of one enantiomer in excess over
the other.1,10 Recently, tilting of the quinoline ring relative to
the metal surface was reported for the adsorption of cinchona
alkaloids at saturation coverage.11 A potential involvement of
Assessing the dynamics of cinchona alkaloids is important
in developing a comprehensive understanding of the mechanisms
of enantiodifferentiation in catalytic reactions. Cinchona alka-
loids consist of two rigid moieties, a quinoline ring and a
quinuclidine ring, coupled together by two carbon-carbon bonds
(Scheme 1). The relative position of these two parts determines
(6) (a) Bu¨rgi, T.; Baiker, A. J. Am. Chem. Soc. 1998, 120, 12920–12926.
(b) 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. (c) Dijkstra, G. D. H.;
Kellogg, R. M.; Wynberg, H. J. Org. Chem. 1990, 55, 6121–6231. (d) Lindholm,
A.; Ma¨ki-Arvela, P.; Toukoniitty, E.; Pakkanen, T. A.; Hirvi, J. T.; Salmi, T.;
Murzin, D. Yu.; Sjo¨holm, R.; Leino, R. J. Chem. Soc. Perkin Trans. 1 2002,
2605–2612. (e) Schu¨rch, M.; Schwalm, O.; Mallat, T.; Weber, J.; Baiker, A. J.
Catal. 1997, 169, 275–286. (f) Martinek, T. A.; Varga, T.; Fu¨lo¨p, F.; Barto´k,
M. J. Catal. 2007, 246, 266–276.
(7) (a) Sutherland, I. M.; Ibbotson, A.; Moyes, R. B.; Wells, P. B. J. Catal.
1990, 125, 77–88. (b) Augustine, R. L.; Taneilyan, S. K.; Doyle, L. K.
Tetrahedron: Asymmetry 1993, 4, 1803–1827. (c) Vayner, G.; Houk, K. N.; Sun,
Y.-K. J. Am. Chem. Soc. 2004, 126, 199–203. (d) Schwalm, O.; Minder, B.;
Weber, J.; Baiker, A. Catal. Lett. 1994, 23, 271–279. (e) Baiker, A. J. Mol.
Catal. A: Chem. 2000, 163, 205–220. (f) Lavoie, S.; Laliberte´, M.-A.; McBreen,
P. H. J. Am. Chem. Soc. 2006, 128, 7588–7593. (g) Margitfalvi, J. L.; Hegedu¨s,
M. J. Mol. Catal. A: Chem. 1996, 107, 281–289.
(8) (a) Barto´k, M.; Sutyinszki, M.; Felfo¨ldi, K.; Szo¨llo¨si, G. Chem. Commun.
2002, 1130–1131. (b) Margitfalvi, J. L.; Ta´las, E. Appl. Catal., A 2006, 301,
187–195.
(9) (a) Bonalumi, N.; Vargas, A.; Ferri, D.; Bu¨rgi, T.; Mallat, T.; Baiker, A.
J. Am. Chem. Soc. 2005, 127, 8467–8477. (b) Vargas, A.; Ferri, D.; Bonalumi,
N.; Mallat, T.; Baiker, A. Angew. Chem., Int. Ed. 2007, 46, 3905–3908. (c)
Bonalumi, N.; Vargas, A.; Ferri, D.; Baiker, A. Chem. Eur. J. 2007, 13, 9236–
9244. (d) Vargas, A.; Bonalumi, N.; Ferri, D.; Baiker, A. J. Phys. Chem. A
2006, 110, 1118–1127.
(2) (a) France, S.; Guerin, D. J.; Miller, S. J.; Letcka, T. Chem. ReV. 2003,
103, 2985–3012. (b) Tian, S.-K.; Chen, Y.; Hang, J.; Tang, L.; McDaid, P.;
Deng, L. Acc. Chem. Res. 2004, 37, 621–631. Kacprzak, K.; Gawron´ski, J.
Synthesis 2001, 961–998. (d) EnantioselectiVe Organocatalysis: Reactions and
Experimental Procedures; Dalko, P. I., Ed.; Wiley-VCH: Weinheim, 2007. (e)
New Frontiers in Asymmetric Catalysis; Mikami, K., Lautens, M., Eds.; Wiley-
VCH: Weinheim, 2007. (f) Berkessel, A.; Gro¨ger, H. Asymmetric Organoca-
talysis; Wiley-VCH: Weinheim, 2005.
(3) (a) Blaser, H. U.; Jalett, H. P.; Lottenbach, W.; Studer, M. J. Am. Chem.
Soc. 2000, 122, 12675–12682. (b) Exner, C.; Pfaltz, A.; Studer, M.; Blaser, H.-
U. AdV. Synth. Catal. 2003, 345, 1253–1260. (c) Busygin, I.; Toukoniitty, E.;
Sillanpa¨a¨, R.; Murzin, D. Yu.; Leino, R. Eur. J. Org. Chem. 2005, 2811–2821.
(d) Busygin, I.; Toukoniitty, E.; Murzin, D. Yu.; Leino, R. J. Mol. Catal. A:
Chem. 2005, 236, 227–235.
(4) (a) Seitz, T.; Baudoux, J.; Bekolo, H.; Cahard, D.; Plaquevent, J.-C.;
Lasne, M.-C.; Rouden, J. Tetrahedron 2006, 62, 6155–6165. (b) Merschaert,
A.; Delbeke, P.; Daloze, D.; Dive, G. Tetrahedron Lett. 2004, 45, 4697–4701.
(5) (a) Diezi, S.; Szabo, A.; Mallat, T.; Baiker, A. Tetrahedron: Asymmetry
2003, 14, 2573–2577. (b) Sonderegger, O. J.; Ho, G. M. W.; Bu¨rgi, T.; Baiker,
A. J. Mol. Catal. A: Chem. 2005, 229, 19–24. (c) Csere´nyi, S.; Felfo¨ldi, K.;
Bala´zsik, K.; Szo¨llo˜si, G.; Bucsi, I.; Barto´k, M. J. Mol. Catal. A: Chem. 2006,
247, 108–115. (d) Toukoniitty, E.; Busygin, I.; Leino, R.; Murzin, D. Yu. J.
Catal. 2004, 227, 210–216. (e) Busygin, I.; Wa¨rnå, J.; Toukoniitty, E.; Murzin,
D. Yu.; Leino, R. J. Catal. 2008, 254, 339–348.
6560 J. Org. Chem. Vol. 73, No. 17, 2008

InVestigation of Some Cinchona Alkaloid O-Ethers
TABLE 1. Enantiomeric Excess (ee) Observed in the
Hydrogenation of 1-Phenyl-1,2-propanedione (PPD, 1) over Pt/Al2O3
Catalyst Modified with Cinchona Alkaloid O-Ethers
modifier
eea (%)
MeOCD (4)
PhOCD (5)
TMSOCD (6)
TBDMSOCD (7)
ADMSOCD (8)
DPMSOCD (9)
MeOCN (10)
PhOCN (11)
TMSOCN (12)
PDMSODHCD (13)
10 (S)-2b
28 (S)-2b
0
27 (S)-2b
1 (S)-2b
3 (S)-2b
32 (R)-2b
33 (R)-2b
30 (R)-2b
1 (S)-2b c
,
FIGURE 1. Newman projections of conformations of cinchonidine
derivatives (Q ) quinoline moiety). Open(3) conformation is formed
by combining the staggered conformation (a) along the C8-C9 bond
with the conformation (c) along the C9-C4′ bond. Analogously,
Closed(1) ) (b + d); Closed(2) ) (b + c); Open(4) ) (a + d).
a ee ) {[(R)-2] - [(S)-2]}/{[(R)-2] + [(S)-2]} at 50% conversion of
PPD. b Configuration of the major product enatiomer. c Same as
obtained for ADMSOCD (8). ADMSOCD (8) under hydrogenation
conditions is rapidly converted to PDMSODHCD (13).
the tilted modifier species in the enantioselection step has also
been discussed.12
a substantial inversion of enantioselectivity was observed and
the (1S)-hydroxyketone 2 was produced after the first hydro-
genation step with enantiomeric excess of approximately 28%.5e
In the case of 8, enantioselection was not observed. Compound
5, although structurally unrelated to 7, exhibits a very similar
behavior as chiral modifier in the hydrogenation of PPD,
whereas 4 favors the production of (1S)-hydroxyketone with
enantiomeric excess of approximately 10%.5e
Another elementary feature of the modifier-catalyst system
is that the stereochemistry of the main product isomer is
determined by modifier configuration.1 Thus, in the enantiose-
lective hydrogenation of PPD, (R)-1-hydroxy-1-phenyl-2-pro-
panone 2 is obtained in excess when cinchonidine is used as
the chiral catalyst modifier (Scheme 1).13 By use of cinchonine,
the opposite product enantiomer is formed in excess. When
O-ether derivatives of cinchona alkaloids are employed as
surface modifiers in the hydrogenation of PPD, either the
enantioselectivity is inverted or, alternatively, no chiral induction
is observed.5e Motivated by these structure-selectivity relation-
ships, we decided to investigate in detail the structures and
conformational behavior of the cinchona alkaloid O-ethers that
were earlier evaluated as chiral modifiers in the enantioselective
hydrogenation of PPD (Scheme 1).5e We demonstrate here that
the conformation found to be abundant in the liquid phase has
no direct correlation with the enantioselectivity of the PPD
hydrogenation reaction.
In solutions, cinchona alkaloids exist as mixtures of confor-
mations that are in rapid equilibria at room temperature.
Therefore, averaged signals are observed by NMR spectroscopy.
Conclusions concerning the conformational behavior of CD have
previously been drawn from spectral data acquired at room
temperature.6a–c In this case, the calculations of conformer
populations are limited to two-site exchange,14 whereas for
cinchonidine in solution at least three conformations were
observed by NOE. Nevertheless, some approximations allow
estimating the contribution of Open(3) conformation while not
describing the other counterparts.6a In order to extract the
complete data on conformational behavior, at least one confor-
mation should be separated, e.g., by freezing the equilibria at
low temperature. Unfortunately this approach could not be
applied to parent cinchona alkaloids in nonpolar solvents due
to their poor solubilities15 at ambient and in particular at low
temperatures. Solubilities of the cinchona alkaloid O-ether
derivatives are, however, remarkably improved and enable the
discrimination of signals arising from the individual conformers.
Accordingly, separation of the average signals has been observed
earlier for 4 in CD2Cl2/CDCl3 at low temperature.6a
2. Results and Discussion
2.1. NMR Analysis. In order to correlate the ee’s observed
in the catalytic experiments (Table 1) with the relative popula-
tion of any modifier conformation found to be abundant in the
liquid phase, four cinchona alkaloid derivatives were analyzed
in CDCl3 and toluene-d8 solutions by dynamic NMR and NOE
techniques. For these experiments, the O-ether derivatives 4,
5, 7, and O-(propyldimethylsilyl)-10,11-dihydrocinchonidine
(PDMSODHCD, 13) were selected. The last mentioned modifier
is obtained by hydrogenation of both the allyl and vinyl groups
of O-(allyldimethylsilyl)cinchonidine (8). The structures of these
two O-silyl ether modifiers (7 and 8) resemble each other and,
after hydrogenation of the CdC double bonds during the first
minutes of the reaction, differ only by the alkyl chains attached
to the silicon atom. Enantioselectivity of the PPD hydrogenation
over the catalyst modified by 7 is different from that over the
catalyst modified by 8. When the former was used as modifier,
2.1.1. O-Methylcinchonidine (4). The conformational equi-
libria of 4 were studied in toluene-d8 and CDCl3 using NMR
spectroscopy. When the temperature of the sample was gradually
decreased, the signals were first broadened and coalescence was
achieved at approximately -30 °C (Figure 2). When the
temperature was lowered below the coalescence, two sets of
signals could be observed. The signals from these conformations
were most sharpened at -70 °C in a ratio of 85:15 in toluene-
3
d8. A coupling constant JH8,H9 of approximately 9.8 Hz was
observed for the minor conformation. Assignment of the
conformations was based on NOE analysis at -70 °C. A
(10) (a) Nieminen, V.; Taskinen, A; Toukoniitty, E.; Hotokka, M.; Murzin,
D. Yu. J. Catal. 2006, 237, 131–142. (b) Taskinen, A.; Nieminen, V.; Hotokka,
M.; Murzin, D. Yu. J. Phys. Chem. C 2007, 111, 5128–5140.
(11) Ma, Z.; Lee, I.; Zaera, F. J. Am. Chem. Soc. 2007, 129, 16083–16090.
(12) (a) Jeffery, E. L.; Mann, R. K.; Hutchings, G. J.; Taylor, S. H.; Willock,
D. J. Catal. Today 2005, 105, 85–92. (b) Busygin, I.; Tkachenko, O. P.;
Nieminen, V.; Borovkov, V. Yu.; Sillanpa¨a¨, R.; Toukoniitty, E.; Murzin, D.
Yu.; Kustov, L. M.; Leino, R J. Phys. Chem. C 2007, 111, 9374–9383.
(13) Toukoniitty, E.; Ma¨ki-Arvela, P.; Kuzma, M.; Villela, A.; Neyestanaki,
A. K.; Salmi, T.; Sjo¨holm, R.; Leino, R.; Laine, E.; Murzin, D. Yu. J. Catal.
2001, 204, 281–291.
(14) Sandstrom, J. Dynamic NMR Spectroscopy; Academic Press: London,
1982; p 19.
J. Org. Chem. Vol. 73, No. 17, 2008 6561

Busygin et al.
FIGURE 2. 1H NMR spectra of 4 at various temperatures in CDCl3.
FIGURE 4. 1H NMR spectra of 13 at different temperatures. From
top to bottom: 40, 25, 15 and 5 °C.
the closed conformations involved in the equilibria. When the
signal at 5.74 ppm was irradiated in an NOE difference
experiment at room temperature, two negative peaks, at 5.74
and 4.81 ppm, were observed indicating transfer of magnetiza-
tion to the minor conformer, which is in rapid equilibrium with
the major one. The 1D selective NOESY experiments were
performed at -50 °C in CDCl3 solution in order to freeze the
equilibria and to assign the individual conformers. The following
NOEs were observed by irradiation of H-9 of the major
conformation: a strong NOE with H-5′, a weak NOE with H-8
and a weak NOE with H-6b (proton designations and the
proton-proton distances for different conformers of 7 are
available as Supporting Information). When the H-5′ of the
major conformation was irradiated, a strong NOE with H-9 and
a weak NOE with H-8 were observed. The irradiation of H-3′
revealed NOE between H-3′ and H-7b and a weak correlation
between H-3′ and H-8. These NOEs are indicative of the major
FIGURE 3. Expansion of the 1H NMR spectra of 7 in CDCl3 at
different temperatures. From top to bottom: 50, 40, 25, and 5 °C.
characteristic correlation between the H-5′ and H-8 protons
(Figure S1, Supporting Information) was observed for the major
conformation allowing to assign it as Open(3). The minor
conformation could be assigned to Closed(1), as confirmed by
the strong NOE between H-3′ and H-9. A weak correlation was
also found between H-3′ and H-8 for the major set of the signals,
indicating that Closed(2) might have some contribution in the
equilibrium. In CDCl3, O-methylcinchonidine 4 exhibits a
similar conformational behavior as in toluene-d8. The ratio
between the major and minor forms was 88:12.
1
set of the H NMR signals being a mixture of Open(3) and
Closed(2) conformers. The following NOEs were observed for
the minor form: a strong NOE from H-9 to H-3′, a moderate
from H-5′ to H-8, a weak NOE from H-9 to H-6b and also
from H-3′ to H-6b. Hence, the minor conformation was
identified as Closed(1).
Remarkably, the complete assignment of the minor group of
the signals was performed here for the first time and is provided
in the Supporting Information.
1
2.1.3. O-(Propyldimethylsilyl)-10,11-dihydrocinchonidine (13).
The 1H NMR spectra of 13 in CDCl3 at different temperatures
are represented in Figure 4. The H-9 proton occurs as a broad
signal at 5.94-5.75 ppm and as a small hump between 4.8 and
4.0 ppm at room temperature. Upon heating the sample, the
H-9 signals, in contrast to the rest of the spectrum, are broadened
to a higher extent. The coalescence point was reached at 40
°C. When the sample was cooled down to 5 °C, two sets of
signals could clearly be observed. The H-9 signal from the major
conformer is sharpening upon cooling. It appears as a broad
singlet centered at 5.88 ppm at 5 °C. The signal has the smallest
line width at -15 °C and starts to broaden below this
temperature. When the sample was cooled to -30 °C in toluene-
d8 solution, H-9 proton of the minor conformer appears as a
2.1.2. O-(tert-Butyldimethylsilyl)cinchonidine (7). The H
NMR spectra of 7 at different temperatures reveal conforma-
tional equilibria in solution (Figure 3). Two sets of signals were
observed already at room temperature in the ratio of 80:20 in
CDCl3 and 71:29 in toluene-d8. The coalescence point could
be reached by heating the sample to 50 °C, whereas for 4
separation of the average signals was observed at -30 °C.
Hence, the interconversional barrier for 7 with a bulkier group
at C(9) is higher than for 4. Two sharp signals from H-9 could
clearly be separated at 5 °C: a major broad singlet at 5.74 ppm
and16 a minor doublet at 4.81 ppm with JH8,H9 ) 9.6 Hz. The
3
broad singlet can be assigned to the Open(3) conformation in
which the dihedral angle H-C(9)-C(8)-H is close to -90°
and the proton-proton coupling is small as suggested by the
Karplus equation. The broadening of this signal results from
3
doublet at 4.76 ppm with JH8,H9 ) 9.8 Hz. Irradiation of the
H-9 of the major conformer in NOE difference experiments at
room temperature turns that of the minor form in a negative
(15) Ma, Z.; Zaera, F. J. Phys. Chem. B 2005, 109, 406–414.
6562 J. Org. Chem. Vol. 73, No. 17, 2008

InVestigation of Some Cinchona Alkaloid O-Ethers
SCHEME 2. Conformational Equilibria of 7 in Solution
phase indicating rapid exchange between the conformers. The
ratio between the two sets of signals is 90:10 in CDCl3 and
85:15 in toluene-d8. The assignment of the conformers was
carried out by measuring the NOE at -70 °C in toluene-d8.
The correlations detected for 13 were similar to those observed
for 7 with the exception of the NOE between H-9 and H-6b,
which was more pronounced in the case of 13. This very likely
originates from the higher population of Closed(2) conformer
in the solution of 13 as compared to 7.
2.1.4. O-Phenylcinchonidine (2). The 1H NMR spectra of 2
in both CDCl3 and toluene-d8 are represented by averaged
signals at room temperature. The H-9 proton appears as a broad
singlet at 6.10 ppm in CDCl3 and as a doublet at 6.06 ppm in
toluene-d8. Low-temperature experiments at -30 °C in CDCl3
1
did not alter the H NMR spectra, and averaged spectra were
again observed. In toluene at -70 °C, the signals were
broadened and assignment of the conformers was complicated
by spin diffusion. The NOESY spectrum displayed all possible
correlations between all protons. Correspondingly, the confor-
mations were assigned in CDCl3/CD2Cl2 solution. When the
temperature was lowered to -70 °C, two clear sets of signals
were observed. The H-9 signal from the minor conformer
appeared as a doublet at 5.32 ppm with a coupling constant
3JH8,H9 ) 10.3 Hz. The population of the minor set was 6%
and was assigned as Closed(1) conformation on the basis of a
strong NOE from H-9 to H-3′. NOE difference spectra in
toluene-d8 and CDCl3 at room temperature revealed a strong
NOE between H-9 and H-5′, a moderate NOE between H-3′
and H-7b and a weak NOE between H-9 and H-6b as well as
between H-9 and H-8. These NOEs are characteristic of the
Closed(2) and Open(3) conformations. A very weak interaction
between H-9 and H-3′, being indicative of Closed(1) conforma-
tions, was also observed. Since the Closed(1) conformer was
not separated for 2 in CDCl3 at -30 °C and in toluene-d8 at
-70 °C, quantification of the results is based on the data
obtained at room temperature. Exact quantification of the NOE
difference results requires information on the ratio of the
Closed(2) and Open(3) conformers in solution as the distances
between H-9 and H-5′ are different for the two. Nevertheless,
it is possible to estimate the interval in which the population of
Closed(1) conformation can vary depending on the ratio
mentioned above. Taking into account the NOE peak integrals
and the distances raised to the sixth power, we obtained the
population of Closed(1) as being between 8% and 10% in
toluene-d8 and between 13% and 17% in CDCl3.
2.1.5. Conformational Equilibria by Spin Simulation/
Iteration. Since the coalescence point for three cinchona
alkaloid O-ether derivatives decreases in the series 7 > 4 > 5,
the interconversion between the Closed(1) and Closed(2) should
be less hindered for the phenyl derivative 5 than for the methyl
derivative 4 and silyl derivative 7. The interconversion barriers
will be assessed theoretically in the next section.
The conformational equilibria of 7 in CDCl3 solution are
represented in Scheme 2. Because of their low interconversional
barrier,6a,c it is not possible to separate the signals of the
Closed(2) and Open(3) conformers, even at low temperatures,
and only by resorting to spin simulation/iteration techniques
could the ratio of Closed(2) and Open(3) conformers be reliably
extracted. Consequently, the corresponding coupling constants
were analyzed by the PERCH software.17 The results are
presented in Table 2. A very good agreement between the
experimentally recorded and calculated spectra was obtained
for 4, 5, and 7, whereas for 13 the reliability of the fit was lower
due to signal broadening. Thus, for 13 in toluene-d8, the coupling
constants could not be easily resolved. The torsion angles
H-C(9)-C(8)-H were calculated from the Altona equation18
using the MestRe-J graphical tool.19 The populations were
calculated by confrontation of the derived angle with the angles
in optimized structures of Closed(2) and Open(3) conformers
(for details, see Supporting Information).
By summarizing the results, it can be concluded that the
Open(3) conformation is the most populated in solutions of 7,
5, and 4 in toluene-d8 and CDCl3, whereas for 13 in CDCl3
Closed(2) is the predominant conformation.
2.2. Computational Study. The molecular structures and
energies of the methyl O-ether derivatives 4 and 10, and the
silyl O-ether derivatives 7 and 13 were investigated in the gas
phase and in solvents (CHCl3, toluene) using density functional
theory (DFT) at the B3LYP/T(ON)DZP level. Detailed informa-
tion on the computational methods employed can be found in
Supporting Information.
2.2.1. O-Methyl Derivatives of Cinchonidine (4) and
Cinchonine (10). The results for 4 and 10 are presented in Table
3. Optimized geometries for the conformations of 4 are
visualized in Figure 5. On the basis of the computational
electronic energies, Open(3) conformation was the most stable
one, followed by Closed(1) and Closed(2) for both O-methyl
derivatives 4 and 10. The Closed(1) conformers were 4.8 and
4.4 kJ mol-1 less stable than the Open(3) conformers of 4
and 10, respectively. The relative Gibbs energies (∆G) at 25
°C and 1 bar are rather similar to the relative electronic energies.
Only slight stabilization of all conformations (except for Open(5)
of 4) with respect to Open(3) was observed. Thus, Closed(1)
was stabilized by 1.2 and 2.6 kJ mol-1 for 4 and 10, respectively.
When the solvent effects of toluene were taken into account by
the conductor-like screening model (COSMO),20 the Closed(1),
Closed(2), Open(4), and Open(5) conformations were stabilized
further relative to Open(3). According to Boltzmann statistics,
the populations of the Closed(1) and Open(3) conformations
of 10 were almost the same, whereas in the case of 4, the
(17) Laatikainen, R; Niemitz, M; Weber, U; Sundelin, J.; Hassinen, T.;
Vepsa¨la¨inen, J. J. Magn. Reson. Ser. A 1996, 120, 1–10.
(18) Altona, C. In Encyclopedia of NMR; Grant, D. M., Morris, R., Eds.;
Wiley: New York, 1996; pp 4909-4923.
(19) Navarro-Vazquez, A.; Cobas, J.; C.; Sardina, F. J.; Casanueva, J.; Diez,
E. J. Chem. Inf. Comput. Sci. 2004, 44, 1680–1685.
(16) (a) Karplus, M. J. Am. Chem. Soc. 1963, 85, 2870–2871. (b) Karplus,
M. J. Phys. Chem. 1959, 30, 11–15.
J. Org. Chem. Vol. 73, No. 17, 2008 6563

Busygin et al.
TABLE 2. Populations of Conformers in the Solutions of Cinchona Alkaloid O-Ethers and the Efficiencies of the Compounds as Chiral
Modifiers in the Hydrogenation of PPD
Closed(1)
3JH9,H8, Hz
Closed(2) and Open(3)
PClosed(2)c, %
modifier
eea
solvent
T, °C
Pb, %
15
10
15
3JH9,H8, Hz
n.a.e
POpen(3)c, %
13
0d
toluene-d8
CDCl3
toluene-d8
CDCl3
toluene-d8
CDCl3
toluene-d8
CDCl3
-30
5
-50
-50
0
5
25
0
9.85
n.a.
9.84
n.a.
n.a.
9.67
n.a.
n.a.
n.a.
n.a.
4.50 ( 0.5
2.50 ( 0.3
2.70 ( 0.1
3.40 ( 0.1
2.70 ( 0.1
3.05 ( 0.1
2.27 ( 0.1
48 (3
29 (2
32
29
29
33 (1
26 (2
42 (3
56 (2
56
42
51
58 (1
60 (2
4
7
5
10 (S)-2
27 (S)-2
28 (S)-2
12
29
20
9 ( 1f
15 ( 2
f
a Enantiomeric excess in the hydrogenation of PPD in toluene, taken from ref 5e. b Population of the conformer as studied by 1H NMR. c Population
of the conformer derived from the coupling constant JH9,H8; ( 1% if not specified. d Data for 8. e Not available. f Calculated from NOE at room
3
temperature.
TABLE 3. Relative Electronic Energies (∆E), Gibbs Energies
(∆G), and Equilibrium Populations (PB) of the Conformations of 4
and 10
of freedom due to the substituent at the C(9)-O position. Starting
from the most stable Open(3) conformation of 7, rotation around
the O-Si bond yields a shorter distance between H-9 and
SiC(CH3), increasing the electronic energy by 7.5 kJ mol-1
.
4
4 in toluene
10
10 in toluene
∆Ea ∆Ga ∆Ga
PB
∆Ea ∆Ga
∆Ga
PB
This indicates increased steric hindrance due to the bulky tert-
butyl group. The equilibrium population of the Closed(2)
conformation (11% proportion) is approximately the same as
that of Closed(1) (19% proportion). Only traces of two other
studied conformations, namely, Open(4) and Open(5), are
expected to be involved in the equilibria for 7 as they are less
stable than the dominating Open(3) conformation. Accordingly,
the three most abundant conformations of 13 were studied. The
results obtained for 13 are similar to those for 7. Open(3) (73%
proportion) is the most populated conformation for 13 according
to the electronic energies. However, the ee’s obtained in the
hydrogenation of 1 over the catalysts modified by these two
silyl O-ether derivatives are different (Tables 1 and 2).5e
Therefore, it can be concluded that the inversion of enantiose-
lectivity in the specific case studied in this work cannot be
rationalized by considering the electronic energies of the
conformers in the gas phase only.
b
b
Closed(1)
Closed(2)
Open(3)
Open(4)
Open(5)
4.8
7.2
0.0
3.6
4.9
0.0
9.0
2.5
3.3
0.0
8.8
22
16
60
2
4.4
10.9
0.0
15.2 13.6
20.9 20.4
1.8
9.1
0.0
0.3
6.9
0.0
13.5
17.8
46
3
51
0
10.4
22.1 24.1 21.7
0
0
a In kJ mol-1 b Equilibrium population of conformation according to
.
Boltzmann statistics, %.
Solvent effects on the conformational equilibria of 7 were
also studied by the COSMO model.20 Stabilization of the closed
conformations relative to Open(3) was observed when the effect
of apolar solvents chloroform and toluene was considered.
Boltzmann equilibrium populations obtained from the DFT
calculations are supported by the NMR data (cf. Tables 2 and
4). The small deviations between the experimental and theoreti-
cal values can be attributed to the inaccuracy of COSMO when
treating the screening effects of nonpolar solvents. The absolute
error in the energies resulting from the evaluation of the
screening effects is typically less than 4 kJ mol-1 when the
dielectric constant of the solvent is low (ꢀ ≈ 2). Nevertheless,
the computational results are well in line with the experimental
values and demonstrate the applicability of the approach utilized
in this work for studying such complex systems.
FIGURE 5. Conformations of 4 optimized at the B3LYP/T(ON)DZP
level of theory.
population of Closed(1) was about one-third of the population
of Open(3). Accordingly, Closed(1) was just 0.3 kJ mol-1 less
stable than Open(3) for 10 in toluene. Closed(2) of 10 was
always at least 5 kJ mol-1 less stable than Closed(1).
2.2.2. O-Silyl Derivatives of Cinchonidine. The optimized
structures of the conformations of 7 (Figure S3) and the most
relevant NOE proton-proton distances (Table S2) are provided
as Supporting Information. The stabilities of the conformers and
their equilibrium populations according to Boltzmann statistics
are given in Table 4. On the basis of the computational electronic
energies, Open(3) is the most abundant conformation (69%
proportion) followed by Closed(1) and Closed(2). The silyl ether
derivatives of cinchonidine have additional rotational degree
The Gibbs free energies including the solvent effects of
toluene and chloroform were also computed for the three most
stable conformations of 7. Relative stabilities of the conformers
with respect to the stabilities based on gas phase electronic
energies changed notably. In fact, the results were rather
surprising. According to the Gibbs energies, the most stable
conformation in the gas phase was Closed(2). The Open(3) and
Closed(1) conformations were less stable than Closed(2) by 1.6
and 2.7 kJ mol-1, respectively. In toluene and chloroform,
Closed(1) was the most stable conformation while the Closed(2)
and Open(3) conformers were less stable by 1.1-3.4 kJ mol-1
.
6564 J. Org. Chem. Vol. 73, No. 17, 2008

InVestigation of Some Cinchona Alkaloid O-Ethers
TABLE 4. Relative Electronic Energies (∆E), Gibbs Energies (∆G), and Equilibrium Population (PB) of the Conformations of 7 and 13
7
7 in CHCl3
7 in toluene
13
b
b
b
b
b
b
b
∆Ea
PB
∆Ga
PB
∆Ea
PB
∆Ga
PB
∆Ea
PB
∆Ga
PB
∆Ea
PB
Closed(1)
Closed(2)
Open(3)
Open(4)
Open(5)
3.2
4.5
0.0
11.0
20.2
19
11
69
1
2.7
0.0
1.6
18
54
29
1.8
3.1
0.0
10.5
20.1
27
16
56
1
0.0
3.4
3.4
66
17
17
2.4
3.6
0.0
10.7
20.2
24
0.0
2.1
1.1
48
21
31
3.7
4.8
0.0
17
11
73
14
61
1
0
0
0
a In kJ mol-1 b Equilibrium population of conformation according to Boltzmann statistics, %.
.
Instead, Open(3) was the most stable conformer in the gas phase
and in solution according to the electronic energies. Such a large
difference between the relative Gibbs energies and electronic
energies, not observed in the cases of 4 and 10, could be
explained by the low-frequency vibrations of 7 which are most
poorly described by the harmonic oscillator approximation.
Small errors in very small nonzero frequencies can lead to very
large errors in entropies and, consequently, in the Gibbs
energies.21 Therefore, it may be better to restrict oneself to
discuss only enthalpy or internal energy. Consequently, the
enthalpy energies for 7 are presented in Supporting Information
(Table S3). The results show that the computational values of
the relative enthalpies are similar to those of the relative
electronic energies.
2.2.3. Rotational Barriers. The potential energy map for 6
was earlier scanned at the HF/3-21G level of theory.6d It was
found that the Closed(1) conformation has practically the same
energy as Open(3). The interconversion between Open(3) and
Open(4) does not take place as a result of the high energy
barrier. The interconversion between Closed(1) and Closed(2)
was also hampered when compared to cinchonidine. It is
interesting that this path has a saddle point energy of 44 kJ
mol-1. Baiker an co-workers have studied the conformational
behavior of O-phenylcinchonidine 5 by scanning the potential
energy surface using the semiempirical PM3 Hamiltonian and
optimizing the minimum energy conformations at the DFT-
B3LYP/6-31G(d,p) level.9d It was found that the structural and
energetic data for 5 is rather similar to that for the parent
compound cinchonidine. Here, we aimed to investigate the
influence of the substituent bulkiness on the rotational barriers
by DFT. The interconversions Closed(1)-Closed(2) and
Open(3)-Open(4) for 4, 7, and 5 as well as Closed(2)-Open(3)
for 4 were studied.
The energies of the optimized Closed(1) and Closed(2)
conformations together with the transition state conformations
are represented in Figure 6 as a function of the C3′-C4′-C9-C8
torsional angle (τ1, for definition and direction of the rotation
see Supporting Information). Interconversion between the
Closed(1) and Closed(2) conformations has the lowest energy
barrier for 5 (30 kJ mol-1) and the substituents could be arranged
by the rotational hindrance in the following order: TBDMS (7)
> Me (4) > Ph (5). The lowest energy barrier observed for
compound 5 is very well in line with the NMR results and could
be explained by the flat geometry of the phenyl ring. Indeed, in
the transition state structure of 5, the distances between the
quinuclidine moiety and the O-substituent and between the
aromatic moiety and the O-substituent are 2.69 and 2.43 Å,
respectively, whereas in the case of 4, both distances are shorter
(ca. 1.90 Å). Therefore, the rotation around the C4′-C9 bond
is more hindered for 4 than for 5.
FIGURE 6. Energy of Closed(1)-Closed(2) interconversion for 4, 7,
and 5.
The influence of direction for the quinoline moiety rotation
was also investigated. It was found that “the substituent site”
rotation (in the direction of τ1 increase for Closed(1) f
Closed(2) interconversion and in the direction of τ1 decrease
for the reversed process) is energetically preferred over “the
quinuclidine site” rotation for 4 and 5, whereas for 7, both
directions are rather equal. The energy of “the quinuclidine site”
barrier did not depend on the size of the tetragonal O-substituent
(58 kJ mol-1 for both 4 and 7). However, this barrier was
slightly lower for the flat phenyl ring. Most likely, it was
originated from the minor repulsion between H-3′ and O-
substituent, which was lower for 5 in the same manner as
reasoned earlier for the H5′-R-H7b repulsion.
During the optimization of the molecular geometries with τ1
fixed to 120°, it was found that energies of the obtained
structures are lower than energies of the Closed(2) conformers
for 4 and 5 (Figure 6). The close inspection of the geometries
revealed steric repulsions between the quinuclidine-N and H-5′
which force the structures to change their τ2 angles by
approximately 80° in the direction toward open conformation.
When the τ1 angle was further increased to 150°, the steric
repulsion between H-5′ and H-8 forced backward change of τ2
to the values characteristic for closed conformations.
The energies of Open(3)TOpen(4) interconversion barriers
are represented in Figure 7. The correlation for Open(3) T
Open(4) interconversion was similar to that for Closed(1) T
Closed(2) interconversion. Compound 5 had the lowest and
compound 7 had the highest energy barriers for both Open(3)
f Open(4) and the reversed processes. Generally, the Open(3)
T Open(4) interconversions were less favorable than Closed(1)
T Closed(2) interconversions. Therefore, equilibria in the
solutions of 7 could be described as represented in Scheme 2,
whereas in solutions of 5, the Open(3) T Open(4) intercon-
version is rather fast in the NMR time scale.
(20) Klamt, A.; Schu¨u¨rmann, G. J. Chem. Soc., Perkin Trans. 2 1993, 799–
805.
J. Org. Chem. Vol. 73, No. 17, 2008 6565

Busygin et al.
FIGURE 8. Molecular structure of 4.
TABLE 5. (a) Calculated Dihedral Angles for Several Conformers
of Cinchonidine and (b) Experimental Dihedral Angles of Modifiers
in the Solid State
FIGURE 7. Energy barrier of Open(3)-Open(4) interconversion for
4, 7, and 5.
(a) CD6a
Closed(1) Closed(2) Open(3) Open(4) Open(5)
It is noteworthy that the NMR results are very much in line
with the data on the rotational barriers as studied by DFT.
Closed(1) conformation of the O-silyl ethers was observed by
1H NMR as a separate set of signals in CDCl3 and toluene-d8
at room temperature, whereas it was detected for 4 at -30 °C
and for 5 only at -70 °C. Computational study of the influence
of the substituent bulkiness on the rotational barriers showed
the increase of the hindrance for interconversion between
Closed(1) (the minor set of the signals in NMR spectra) and
Closed(2) (component of the major set of the signals in NMR
spectra) within the same series: 7 (55 kJ mol-1) > 4 (39 kJ
mol-1) > 5 (30 kJ mol-1).
a
b
τ1
τ2
-107.0
-176.8
57.5
80.4
-172.5
65.3
101.4
-78.3
153.6
-89.3
-77.5
150.1
85.7
76.2
-48.6
N(1)-C(8)-C(9)-C(4′)
(b)
CD22
101.5(7) -106.2(3) 88.9(5)
4
5
CN24
10
τ1
τ2
-99.82
102.5(2)
173.7
-70.8 175.6 -72.3
70.8
N(1)-C(8)- 158.0(6) 52.7(3)
C(9)-C(4′)
162.6(4) -160.35 -59.0(3)
a τ1: C(3′)-C(4′)-C(9)-C(8). b τ2: H(9)-C(9)-C(8)-H(8).
Since it was not possible to separate the Closed(2) and
Open(3) conformers at low temperature, we have also calculated
the energy barrier of this interconversion for compound 4. The
interconversion barrier between Closed(2) and Open(3) is
expected to be 13-17 kJ/mol as studied by Bu¨rgi and Baiker6a
with model compounds, and consequently, the conformational
equilibria between Closed(2) and Open(3) could not be frozen
at -90 °C. As illustrated in Supporting Information (Figure S4),
the energy barrier for Open(3) f Closed(2) transfer was 11 kJ
mol-1, whereas it was only 2.5 kJ mol-1 for the reversed
process. These data are in line with the results obtained earlier
by Bu¨rgi and Baiker.6a
FIGURE 9. Molecular structure of 10.
2.3. X-Ray Structures of the Modifiers. In a recent paper,
Zaera and coauthors11 studied the adsorption of cinchona
alkaloids onto the platinum surface from solutions of different
alkaloid concentrations. It was determined that, at saturation
coverage, the quinoline ring of cinchonidine tilts along its
longitudinal axis to optimize π-π intermolecular interaction.
Moreover, it was suggested that intermolecular π-π stacking
is a dominant force defining the adsorption geometry. Since a
similar stacking defines the crystal lattice in the solid,22 the
modifier crystal structures gain an additional value. Thus, in
the light of these results it would be interesting to compare the
potential intermolecular π-π stacking in different derivatives
of cinchona alkaloids.
O-ethers as chiral modifiers, the study of their structures in
crystal form may provide additional information relevant to the
structures of these molecules on the surface. Molecular structures
of these compounds are represented in Figures 8–10. By
analyzing the data in Table 5 and Table S1 (Supporting
Information), it is clear that the O-methyl ether derivatives adopt
the Closed(1)-like conformation, whereas 5 adopts the Open(3)-
like conformation in the crystal. Since the positions of hydrogen
atoms are generated by calculations from the positions of the
host carbon atoms, the estimated deviation of the dihedral angle
τ2 could be in the order of 2-3°. Therefore, we additionally
provide here the N(1)-C(8)-C(9)-C(4′) dihedral angles with
standard deviations for the parameter.
Crystal structures of three cinchona alkaloid derivatives,
O-methylcinchonidine (4), O-phenylcinchonidine (5), and O-
methylcinchonine (10), were determined in the present study.23
Regarding the interesting behavior of the cinchona alkaloid
It is noteworthy that geometries of the optimized conforma-
tions are slightly different from those in the crystal. Conse-
quently, the dihedral angles τ1 and τ2 for 5 in the crystal differ
by 12° and 8° from the minimum energy Open(3) conformation
(21) Cramer, C. J. Essentials of Computational Chemistry - Theories and
Models; Wiley: Chichester, 2002; p 340.
(22) Oleksyn, B. J. Acta Crystallogr. 1982, B38, 1832–1834.
(23) The X-ray structure of MeOCN has been reported earlier in an
independent study: Exner, C. Ph.D. Thesis, University of Basel, 2002.
6566 J. Org. Chem. Vol. 73, No. 17, 2008

InVestigation of Some Cinchona Alkaloid O-Ethers
Dijkstra and co-workers studied the conformational equilibria
of O-methyl-10,11-dihydroquinidine in different solvents.6b This
compound predominantly adopts the Open(3) conformation in
CDCl3 and to a lesser amount the Closed(2) conformation.
Closed(1) conformer was not detected. O-Methylcinchonidine
4 was studied in CDCl3/CD2Cl2 solution at low temperatures
by Baiker and co-workers.6a In their study, Closed(1) conformer
was separated at low temperature with a population of 6-7%.
Considering that the modification of the cinchona alkaloid
structure induces changes in the conformational equilibria,6b,c
it was challenging to investigate whether the enantioselectivity
over the cinchona alkaloid O-ether modified metal catalysts
correlates with the population of Open(3) or any other confor-
mation in the liquid phase.
FIGURE 10. Molecular structure of 5.
The crucial role of the C(9)-OH group of modifiers in
enantiodifferentiation over the metal catalyst is incontestable
in the hydrogenation of 1.5d,e Under optimal conditions, the main
product, (R)-2 (Scheme 1), can be obtained in 65% ee using
cinchonidine as the chiral catalyst modifier. When O-methyl,
O-phenyl or O-silyl ether derivatives of cinchonidine are used,
a loss (ee ) 0%) or inversion of enantioselectivity results with
10-28% ee of (S)-2 produced. The same trend is observed with
cinchonines as chiral modifier. With parent cinchonine, the main
product (S)-2 (Scheme 1) is obtained in 19% ee. When
O-methyl, O-phenyl or O-silyl ether derivatives of cinchonine
are used, an inversion of enantioselectivity results with (R)-2
being produced in 28-33% ee (Table 1).
After changing the chiral pocket by utilization of cinchona
alkaloid O-ethers as modifiers, the polar hydroxyl group is
replaced by an apolar and bulkier ether group. The modifier-
reactant interaction is generally composed of both attractive and
repulsive interactions which play a role in the enantiodifferen-
tiating interactions. It is interesting that the new chiral pocket
induces the production of the opposite product enantiomer in
excess. At first glance, since the attractive interactions (possible
H-bonding between the OH-group in the modifier and the
carbonyl group in the reactant)5d are now replaced by repulsive
ones, it would be natural to expect an inversion in the sense of
enantioselectivity. In this case one should expect an inversion
of enantioselectivity always when the parent cinchona alkaloid
is substituted by its O-ether. On the contrary, we observe the
inversion of enantioselectivity for 4, 5, and 7 in toluene, whereas
in the same solvent, the modification of the surface with 6, 8,
and 9 results in the production of a mixture of the product
enantiomers in nearly equal amounts.5e In the present work, thus,
an attempt to correlate the enantioselectivities observed with 8,
4, 7, and 5 (increase in the production of (S)-enantiomer within
this series) with their conformational equilibria in solution was
made. The hydroxyl group of all of these cinchona alkaloid
derivatives is blocked by substituents and, therefore, one would
not at first hand expect significant differences in their binding
to the heterogeneous catalyst surface, as is the case with the
parent cinchona alkaloids versus their O-ether derivatives.
optimized by Baiker and co-workers.9d However, the relative
position of the phenyl ring, which could be defined by the
C(9)-O-Cipso-Cortho′ dihedral angle differs only by 4° (DFT,
9.9°; X-ray, 5.6°). For O-methyl derivatives 4 and 10, the
difference between the dihedral angles in solid and in the gas
phase is less and varies from 1° to 2.5°.
Conclusively, due to the bulky O-ether substituent, none of
the cinchona alkaloid O-ether derivatives studied in this work
showed π-π stacking in the crystal. Only weak C· · ·C
interactions in two compounds were observed: one in compound
5 (the C · · ·C distance is 3.377(7) Å) and two in 10 (the C · · ·C
distances are 3.356(4) and 3.468(4) Å). Possibly, the effect of
lateral interactions between molecules could be an important
factor in their assembly on the metal surface at relatively high
coverages.
2.4. Conformational Equilibria versus Enantioselectivity. The
Open(3) conformation of cinchonidine has commonly been
accepted to have a crucial role in obtaining high enantioselec-
tivity in the hydrogenation of carbonyl compounds over
cinchona alkaloid modified platinum catalysts.6a The decrease
of ee for O-substituted cinchona alkaloids in enantioselective
hydrogenation of ethyl pyruvate in apolar solvents was related
to the population of the Open(3) conformation in the liquid phase
and consequently on the metal surface.6a As a result of modifier
adsorption-desorption processes, the conformational equilibria
in solution can provide some valuable information on the
distribution of the adsorbed species.
Although the conformational equilibria of the parent cinchona
alkaloids in the liquid phase have been studied in detail
earlier,6a–c only a few studies have addressed the conformational
behavior of the ether derivatives.6b,9 The conformations of
O-phenylcinchonidine 5 have been analyzed by a theoretical-
experimental approach.9d It was found that Open(3) is the
predominant conformation in apolar solvents with a fraction of
Closed(2) conformation, and the population of the latter is in
direct proportion to the solvent polarity. Correlation between
the inversion of enantioselectivity in ketopantolactone hydro-
genation when 5 was used as a chiral modifier and the
conformational behavior of 5 and cinchonidine was not ob-
served, as for both alkaloids Open(3) is the dominant conforma-
tion in CHCl3. Closed(1) conformation was omitted from the
discussion on the basis of the calculated and experimental VCD
spectra confrontation. In the present work, we have demonstrated
the involvement of Closed(1) conformation of 5 in the liquid
phase equilibria by NMR spectroscopy.
One of the representatives from each group with similar
structure and different properties (7 and 8) as well as with
different nature of the substituent but similar properties (7 and
5) were selected and analyzed by NMR. The results presented
in Table 2 demonstrate that direct correlation between the liquid
phase populations of conformers in CDCl3 and toluene-d8 with
enantioselectivities in PPD hydrogenation cannot be established.
The only possible link to the catalytic data is that the Open(3)
conformer of 13, contrary to all the other modifiers studied by
(24) Oleksyn, B.; Lebioda, L.; Ciechanowicz-Rutkowska, M. Acta Crystal-
logr. 1979, 35, 440–444.
J. Org. Chem. Vol. 73, No. 17, 2008 6567

Busygin et al.
NMR, is not the most representative one in CDCl3 solution and
this compound does not induce enantioselectivity in the
hydrogenation of 1. It seems that the modifier adsorption
behavior is a midpoint in the enantioselection process. Particu-
larly, the modifier may change the conformation upon adsorption
onto the metal surface or, alternatively, the population of a
certain conformer on the metal surface is not necessarily
correlated with the same in the liquid phase.
3JH8,H9 coupling constants of the averaged Open(3) and
Closed(2) conformations were extracted by spin simulation/
iteration techniques, and the distribution between these confor-
mations was subsequently calculated from the Altona equation.
Closed(1) conformation of the O-silyl ethers was observed by
1H NMR as a separate set of signals in CDCl3 and toluene-d8
at room temperature, whereas it was detected for O-methylcin-
chonidine (4) and O-phenylcinchonidine (5) only at low
temperatures (-30 and -70 °C, respectively). Open(3) was
found to be the most populated conformation of 4, 10, 7, and 5
in both CDCl3 and toluene-d8 by NMR, whereas the population
of the Closed(2) conformation of 13 was slightly higher than
that of Open(3) in CDCl3. Populations of the conformers of
O-methyl derivatives 4 and 10 in toluene and those of 7 in
chloroform and toluene were calculated by DFT. The results
are in accordance with the experimental data. X-ray structures of
4 and 5 were reported here for the first time. It was observed that
the O-methyl ether derivatives of cinchonidine 4 and cinchonine
10 adopt the Closed(1) conformation in the crystal, whereas
O-phenylcinchonidine 5 attains the Open(3) conformation.
The enantioselectivity over the cinchona alkaloid O-ether
modified metal catalysts was confronted with the populations
of Closed(1), Closed(2), and Open(3) conformations in the liquid
phase. Thereby, we have demonstrated that the modifier
conformation in the liquid phase has a minor role in determining
the enantioselectivity in the hydrogenation of PPD over cinchona
alkaloid modified platinum catalysts. The driving force for
production of one of the enantiomers in excess over another is
deduced to result from a specific adsorption of the modifier on
the catalyst surface, a phenomenon that does not correlate with
the population of the conformers in the liquid phase as was
assumed earlier. In other words, the data presented here confirm
that the origin of chiral switch phenomenon in the hydrogenation
of PPD is related to depletion of the H-bonding interaction
between the modifier OH-group and the carbonyl group of the
reactant rather than to a decreased population of the Open(3)
conformation in the solutions of O-ether derivatives when
compared with the solution behavior of the parent alkaloid.
Additionally, we propose that the method applied here for
the complete assignment of conformational behavior of cinchona
alkaloid O-ethers in solutions might be successfully utilized for
correlating the catalyst structure effects with selectivities of
different cinchona alkaloid-catalyzed organic reactions.
Finally, as overviewed in the Introduction, conformational
equilibria in the solutions of cinchona alkaloids and their
derivatives are of interest not only for researchers working in
the field of chiral modification of metal catalysts but for broader
audiences as well working in the field of organocatalysis. A
variety of examples in the literature have demonstrated that
cinchona alkaloids are efficient catalysts for several asymmetric
reactions. Therefore, the findings of the present paper should
be of practical interest when developing models for stereocontrol
in cinchona alkaloid catalyzed asymmetric reactions in organic
media in general.
In several mechanistic models,7 it is commonly assumed,
while not yet experimentally confirmed, that the origin of
enantioselectivity in the Orito reaction is related to the specific
interaction between the surface Open(3) conformation of the
modifier and the reactant. In a paper by Bu¨rgi and Baiker,6e
a
correlation between the liquid phase fraction of Open(3)
conformation of cinchonidine in different solvents and the
enantioselectivities of ketopantolactone hydrogenation over
cinchonidine-modified Pt was reported. This observation gener-
ally confirms the validity of the Langmuir adsorption equation
in this particular case, i.e., that the coverage or adsorption of
molecules on a solid surface corresponds to the concentration
of these molecules in the liquid phase at a fixed temperature.
In the present study, it is demonstrated for the first time that
concentrations of any conformers in the liquid phase do not
correlate with the fractions of the corresponding surface
adsorbed species. The observed phenomenon suggests a more
complex dependence of the surface coverage on enthalpy of
adsorption.
Recently, Baiker and co-workers,9c by means of theoretical
calculations and infrared spectroscopic techniques, clearly
demonstrated that compound 6 preferentially adsorbs on the
platinum surface in surface-quinuclidine-bound mode. These
results generally overcome the earlier suggested correlation
between the population of conformer in the liquid phase and
the observed enantioselectivities in the hydrogenation of R-keto
esters6a or limit it to the cinchonidine case alone, and are in
line with our current findings. Yet compound 4, as reported,9c
exhibits an adsorption behavior similar to the parent cinchoni-
dine. These data well describe the features of R-keto ester
hydrogenation, while being unable to correlate with the results
observed in the hydrogenation of PPD.5e DFT calculations are
not able to correctly model the noncovalent part of the
interaction with the surface. Hence, very little emphasis should
be placed on literature conformational data obtained using DFT
calculations. Similarly, while vibrational studies of adsorbed
alkaloids can give reliable information on the quinoline orienta-
tion, it is very doubtful whether they can be used to establish
the conformation of the adsorbate. This remains an extraordinar-
ily difficult challenge for surface spectroscopy. A closer
inspection is therefore needed regarding the modifier adsorption
mode and conformation on the catalyst surface and their role
in the enantiodifferentiation process.
3. Conclusions
The complete assignment of the conformational equilibria of
the solute cinchona alkaloid O-ether derivatives was performed
for the first time. In the earlier studies of cinchonidine in
solutions, the population of the Open(3) conformation was based
on an NOE-derived value, whereas the data presented here is
based on a more accurate coupling constant analysis. Direct
measurements of the population of Closed(1) conformation were
performed by examination of the NMR spectra at temperatures
under conditions of slow exchange on the NMR time-scale. The
4. Experimental Section
Synthesis of the Modifiers. O-Phenylcinchonidine (5). The
procedure for the synthesis of O-phenyl derivatives of cinchona
alkaloids described here is a modification of that previously
published5a and is illustrated below for the preparation of PhOCD.
Under argon atmosphere, 457 mg of NaH (19 mmol, 761 mg of
60% dispersion in mineral oil washed with absolute hexane three
6568 J. Org. Chem. Vol. 73, No. 17, 2008

InVestigation of Some Cinchona Alkaloid O-Ethers
times) was suspended in 30 mL of absolute DMSO and cooled
with an ice-bath, whereupon cinchonidine (4 g, 13.6 mmol) was
added in small portions. The reaction mixture was stirred for 1 h
at 0 °C and for additional 2 h at 50 °C. To the cooled orange
solution were added absolute pyridine (2.2 mL, 27.2 mmol) and
CuI (2.59 g, 13.6 mmol), and the reaction mixture was stirred for
30 min. After addition of iodobenzene (1.515 mL, 13.59 mmol)
the mixture was kept at 100 °C for 72 h. Water (40 mL),
dichloromethane (80 mL), ethylenediaminetetraacetic acid (0.8 g),
and finally concentrated ammonia solution (10 mL) were added to
the reaction mixture cooled to room temperature earlier. The
resulting mixture was stirred at room temperature for 1 h. The
organic layer was separated, and the aqueous phase was extracted
with CH2Cl2 (3 × 25 mL). The combined organic layers were
washed with 5% ammonia solution (5 × 25 mL) until the aqueous
phase remained colorless and then with water (25 mL). The solvent
was removed by rotary evaporation. The residue was dissolved in
EtOAc (80 mL), and the resulting solution was extracted with 2 M
HCl solution (80 mL). The acidic solution was washed with EtOAc
(2 × 35 mL), neutralized with solid NaHCO3 until pH ) 9, and
extracted with EtOAc (3 × 40 mL). The combined organic layers
were washed with brine, dried over Na2SO4, and evaporated. The
product was purified over silica (120 g) using diethyl ether/
triethylamine 9:1 as eluent. The solvent evaporation yielded white
crystals which after recrystallization from hexane gave 2.1 g of a
yield 4.8 g (86%) of the product as colorless crystals. Mp )
1 3
168-173 °C (dec). H NMR (CDCl3, δ): 8.91 (d, J ) 4.4 Hz,
4
3
1H, H-2′), 8.63 (dd, 1H, J ) 1.4, J ) 7.9 Hz, 1H, H-5′), 8.13
4
3
(dd, J ) 1.4, J ) 7.9 Hz, 1H, H-8′), 7.75 (m, 2H, H-6′, H-7′),
7.54 (d, 1H, J ) 4.4 Hz, 1H, H-3′), 6.97 (s, 1H, H-9), 5.99 (m,
3
1H, H-10), 5.26 (m, 2H, H-11), 3.96 (m, 1H, H-2a), 3.43 (m, 2H,
H-2b, H-6a), 3.28 (m, 1H, H-8), 3.20 (m, 1H, H-6b), 2.65 (m, 1H,
H-3), 2.50 (m, 1H, H-7b), 2.40 (m, 1H, H-4), 1.93 (m, 1H, H-5b),
1.69 (m, 1H, H-5a), 1.15 (m, 1H, H-7a), 0.19 (s, 9H, Si-CH3). 13
C
NMR (CDCl3, δ): 149.6 (C-2′), 148.3 (C-8a′), 145.6 (C-4′), 136.3
(C-10), 130.2 (C-8′), 129.8 (C-7′), 128.3 (C-6′), 124.5 (C-4a′), 123.1
(C-5′), 118.4 (C-3′), 117.8 (C11), 68.1 (C-9), 60.6 (C-8), 49.1 (C-
6), 47.6 (C-2), 37.3 (C-3), 27.5 (C-4), 23.3 (C-5), 17.8 (C-7), 0.4
(3C, Si-CH3). [R]30D ) +149.9° (c 0.036, CHCl3). Analytical data
for C22H30N2OSi, HRMS (calcd/found): 366.2127/366.2126.
Hydrogenation of 8. A 100 mg portion of 8 was dissolved in
50 mL of toluene and stirred over the Pt/Al2O3 catalyst (100 mg)
under hydrogen pressure of 10 bar for 12-16 h. After the reaction
the catalyst was filtered, the solvent was evaporated, and the sample
was analyzed by NMR without further purification. O-(Propyldim-
ethylsilyl)-10,11-dihydrocinchonidine (13): C24H36N2OSi; 1H NMR
3
(CDCl3, δ): 8.87 (d, 1H, J ) 4.0 Hz, H-2′), 8.20-8.10 (br, 1H,
3
H-5′), 8.14 (d, 1H, J ) 8.4 Hz, H-8′), 7.72 (tr, 1H, J ) 7.7 Hz,
1H, H-7), 7.79 (tr, 1H, J ) 7.5 Hz, 1H, H-6′), 7.54-7.48 (br, 1H,
H-3′), 5.94-5.70 (br, 1H, H-9), 3.49 (br, 1H, H-6b), 3.10 (br, 1H,
H-2a), 2.98 (br, 1H, H-8), 2.70 (br, 1H, H-6a), 2.39 (br, 1H, H-2b),
1.80 (m, 1H, H-4), 1.77-1.70 (br, 1H, H-7b), 1.50-1.40 (br, 3H,
H-3, H-5b, H-5a), 1.35-1.29 (m, 2H, H-13), 1.28-1.24 (br, 1H,
H-7a), 1.24-1.15 (br, 2H, H-10), 0.90 (tr, 3J ) 7.2 Hz, 3H, H-14),
1
pure product. Yield: 42%; mp 125-127 °C. H NMR (500 MHz,
CDCl3, 273.1K): δ 8.85 (d, 3J ) 4.5 Hz, 1H, H-2′), 8.21 (dd, 4J )
3
4
3
1.3, J ) 8.5 Hz, 1H, H-8′), 8.19 (dd, J ) 1.3, J ) 8.4 Hz, 1H,
4
3
3
H-5′), 7.82 (ddd, J ) 1.3, J ) 6.9, J ) 8.5 Hz, 1H, H-7′), 7.70
(ddd, 4J ) 1.3, 3J ) 6.9, 3J ) 8.4 Hz, 1H, H-6′), 7.49 (d, 3J ) 4.5
Hz, 1H, H-3′), 7.18 (m, 2H, m-C6H5), 6.91 (dd, 4J ) 0.9, 3J ) 7.3
Hz, 1H, p-C6H5), 6.79 (m, 2H, o-C6H5), 6.10 (br s, 1H, H-9), 5.75
(ddd, JH3,H10 ) 7.8, JH11E,H10 ) 10.3, JH11Z,H10 ) 17.1 Hz, 1H,
H-10), 4.99 (ddd, J ) 1.2, JH11Z,H10 ) 17.1, J ) - 1.7 Hz, 1H,
3
0.79 (tr, J ) 7.3 Hz, 3H, H-11), 0.55 (m, 2H, H-12) 0.07 (s, 3H,
Si-CH3-a), -0.07 (s, 3H, Si-CH3-b). 13C NMR (CDCl3, δ): 149.9
(C-2′), 149.3 (br, C-8a′), 148.3 (br, C-4′), 130.4 (C-8′), 129.0 (C-
7′), 126.8 (br, C-6′), 125.4 (br, C-4a′), 122.9 (br, C-5′), 118.5 (br,
C-3′), 72.4 (br, C-9), 61.2 (C-8), 58.7 (C-2), 43.2 (br, C-6), 37.3
(C-3), 28.2 (br, 2C, C-5, C-7), 27.6 (C-10), 25.5 (C-4), 19.4 (C-
12), 18.1 (C-14), 16.7 (C-13), 12.0 (C-11), -1.3 (Si-CH3-b), -1.6
(Si-CH3-a).
3
3
3
4
3
2
4
3
2
H-11-Z), 4.93 (ddd, J ) 1.0, JH11E,H10 ) 10.3, J ) - 1.7 Hz,
1H, H-11-E), 3.42 (m, 1H, H-6b), 3.25 (m, 1H, H-8), 3.20 (dd, 3J
) 10.1, 2J ) - 13.8 Hz, 1H, H-2a), 2.75 (m, 1H, H-6a), 2.72 (m,
1H, H-2b), 2.34 (m, 1H, H-3), 2.08 (m, 1H, H-7b), 1.96 (m, 1H,
H-5b), 1.90 (m, 1H, H-4), 1.62 (m, 1H, H-5a), 1.56 (m, 1H, H-7a)
ppm. 13C NMR (150 MHz, CDCl3, 298.1K): δ 157.0 (ipso-C6H5),
150.3 (C-2′), 148.5 (C-8a′), 146.5 (C-4′), 141.7 (C-10), 130.8 (C-
8′), 129.6 (2C, m-C6H5), 129.3 (C-7′), 127.1 (C-6′), 125.4 (C-4a′),
122.7 (C-5′), 121.4 (p-C6H5), 118.2 (C-3′), 115.5 (2C, o-C6H5),
114.4 (C-11), 78.9 (C-9), 60.7 (C-8), 57.3 (C-2), 43.4 (C-6), 40.0
Additional experimental details can be found in Supporting
Information.
Acknowledgment. Financial support by the Magnus Ehrn-
rooth Foundation (I.B.) is gratefully acknowledged. Dr. Mattias
Roslund is acknowledged for fruitful discussions on the NMR
results. Dr. Angelo Vargas is acknowledged for providing the
information on the structures of optimized conformations for
O-phenylcinchonidine. Computer resources provided by CSC,
the Finnish IT center for science, are kindly acknowledged.
(C-3), 28.1 (C-5), 27.7 (C-4), 21.2 (C-7) ppm. [R]30 ) +106.5°
D
(c 0.048, CHCl3). Analytical data for C25H26N2O, HRMS (calcd/
found): 370.2045/370.2045.
O-(Trimethylsilyl)cinchonine (12). Trimethylsilyl chloride (2.16
mL, 16.81 mmol) was added dropwise to an ice-cooled solution of
cinchonine (4.5 g, 15.28 mmol) in DMF (1 L) containing triethy-
lamine (2.4 mL, 16.81 mmol) within 15 min. The reaction mixture
was stirred at room temperature for 48 h until the starting material
disappeared (TLC, CHCl3/MeOH/triethylamine 40:10:1). The sol-
vent was removed by rotary evaporation. The resulting oil was
dissolved in 100 mL of chloroform and washed with water (100,
50, 50 mL). The water layer was extracted with additional
chloroform (2 × 50 mL). Combined organic layers were washed
with brine and dried over Na2SO4. After solvent evaporation, a white
amorphous solid was obtained that recrystallized from hexane to
Supporting Information Available: Experimental details,
atom numbering in cinchonidine derivatives, proton-proton
distances, equilibrium population and torsion angles for selected
cinchona alkaloids, expansion of the recorded and simulated
1H NMR spectrum of TBDMSOCD (7), conformations of
TBDMSOCD (7), and 1H and 13C NMR spectra for the
synthesized compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
JO8008462
J. Org. Chem. Vol. 73, No. 17, 2008 6569