Consideration of molecular arrangements in regio- and enantioselective reduction of an NAD model compound controlled by carbonyl oxygen orientation

Article abstract of DOI:10.1039/b710780c

The regio- and enantioselectivity of the reduction of an NAD model compound having axial chirality with respect to the C₃(quinolinium)- C(carbonyl) bond, 3-piperidinylcarbonyl-1,2,4-trimethylquinolinium ion (1), by using several reducing agents is described. Reaction of 1 with sodium hydrosulfite affords the 1,4-reduced product, 3-piperidinylcarbonyl-1,2,4- trimethyl-1,4-dihydroquinoline (2), with low enantioselectivity, whereas sodium borohydride promotes 1,2-reduction, affording 3-piperidinylcarbonyl-1,2,4- trimethyl-1,2-dihydroquinoline (3) as the sole product in a moderate enantioselectivity. When 1 was reduced by the chiral NADH model compound, 2,4-dimethyl-3-(N-α-methylbenzylcarbamoyl)-1-propyl-1,4-dihydropyridine (Me₂PNPH (4)), the regioselectivity and enantioselectivity of the reaction were significantly altered by the stereochemistry of 1 and 4. An achiral NADH model compound, 1-propyl-1,4-dihydronicotinamide (PNAH (5)) exhibited both high regio- and enantioselectivities. The product selectivity reflects the change in molecular arrangement in the transition state of the reaction and reveals the relative importance of the parameters governing the molecular arrangement in the reaction. The Royal Society of Chemistry 2007.

Full text of DOI:10.1039/b710780c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Consideration of molecular arrangements in regio- and enantioselective
reduction of an NAD model compound controlled by carbonyl oxygen
orientation†‡
Yuji Mikata,*a Shiho Aida,b Yoko Inabab and Shigenobu Yanob
Received 13th July 2007, Accepted 28th September 2007
First published as an Advance Article on the web 15th October 2007
DOI: 10.1039/b710780c
The regio- and enantioselectivity of the reduction of an NAD model compound having axial chirality
with respect to the C3(quinolinium)–C(carbonyl) bond, 3-piperidinylcarbonyl-1,2,4-trimethyl-
quinolinium ion (1), by using several reducing agents is described. Reaction of 1 with sodium
hydrosulfite affords the 1,4-reduced product, 3-piperidinylcarbonyl-1,2,4-trimethyl-1,4-dihydro-
quinoline (2), with low enantioselectivity, whereas sodium borohydride promotes 1,2-reduction,
affording 3-piperidinylcarbonyl-1,2,4-trimethyl-1,2-dihydroquinoline (3) as the sole product in a
moderate enantioselectivity. When 1 was reduced by the chiral NADH model compound,
2,4-dimethyl-3-(N-a-methylbenzylcarbamoyl)-1-propyl-1,4-dihydropyridine (Me2PNPH (4)), the
regioselectivity and enantioselectivity of the reaction were significantly altered by the stereochemistry of
1 and 4. An achiral NADH model compound, 1-propyl-1,4-dihydronicotinamide (PNAH (5)) exhibited
both high regio- and enantioselectivities. The product selectivity reflects the change in molecular
arrangement in the transition state of the reaction and reveals the relative importance of the parameters
governing the molecular arrangement in the reaction.
In this article, we analyze the role of the carbonyl oxygen
orientation in determining the regio- and stereoselectivity of the re-
duction of non-annulated NAD model compound, 3-piperidinyl-
Introduction
The asymmetric reduction of activated ketones or olefins with
chiral NADH model compounds has been studied extensively.1–11
carbonyl-1,2,4-trimethylquinolinium ion (1), which has axial
This biomimetic molecular transformation including the non-
chirality with respect to the C3(quinolinium)–C(carbonyl) bond.
asymmetric version of the reaction with various substrates,12–17 has
attracted considerable interest in relation to transition metal-free
organic catalysts.18–20 On the other hand, the reverse reactions,
namely reductions of NAD model compounds exhibiting face-
selective reduction, have been reported in a limited number.6,8,21–26
The control of regioselectivity of the reduction is an additional
requirement for this type of reaction because it often affords
mixtures of 1,2-, 1,4- and 1,6-isomers.27–29
Several strategies for the differentiation of the nicotinamide
ring face in NAD model compounds have been developed. The
C3–C6,21 C3–C5,30 or C2–C526 strapped NAD model compounds
undergo enantiospecific reduction from the unhindered face.
Annulation in C2 and C3 results in inclination of the carbamoyl
As reported in our previous communication,32 when (aS)-1 was
reduced by a chiral NADH model compound, (4R,9R)-2,4-
dimethyl-3-(N-a-methylbenzylcarbamoyl)-1-propyl-1,4-dihydro-
pyridine ((4R)-Me2PNPH ((4R)-4), Chart 1), the main product
was 1,4-dihydroquinoline (4R)-2, which is the compound where
the hydride was introduced at the C4 position from the carbonyl
oxygen face direction. The axial chirality was lost upon reduction,
thus, axial chirality was converted to central chirality in this
system. On the other hand, the present paper reveals that in
the reaction of corresponding enantiomer of 1 ((aR)-1) with
the same reducing agent ((4R)-4) affords 1,2-dihydroquinoline
(2S)-3 as a major product, which is the compound where the
hydride was introduced at the C2 position from the carbonyl
oxygen face direction (Scheme 1). The only difference between
these two reactions is the orientation of the carbonyl group in
starting material 1; thus, the carbonyl dipole controls the regio-
and stereoselectivity of the reaction. The details in determination
of the absolute configuration of the product 3, regio- and
=
C O bond from the plane of the nicotinamide ring, mostly
activating the syn-face with the carbonyl oxygen in the 1,4-
reduction.6,8,23,25 The role of out-of-plane orientations of the
carbonyl group in NAD/NADH model compounds in controlling
the stereoselectivity has been studied extensively.22,23,25,31
aKYOUSEI Science Center, Nara Women’s University, Nara 630-8506,
Japan. E-mail: mikata@cc.nara-wu.ac.jp
bDivision of Functional Material Science, Nara Women’s University, Nara
630-8506, Japan
† Electronic supplementary information (ESI) available: General experi-
mental procedures, Fig. S1, Table S1 for crystallographic data for (2R)-
and (2S)-6 and cif files. See DOI: 10.1039/b710780c
‡ CCDC reference numbers 654141 and 654142. For crystallographic data
in CIF or other electronic format see DOI: 10.1039/b710780c
Chart 1
3834 | Org. Biomol. Chem., 2007, 5, 3834–3841
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Scheme 1
enantioselectivity of the reduction with several reducing agents,
and molecular arrangements at the transition states of the
reactions are discussed.
Fig. 1 CD spectra of resolved enantiomers of 3 in EtOH. The solid line
represents the fast-eluting enantiomer (first-3) in HPLC and the broken
line represents the slow-eluting enantiomer (second-3).
Results
X-Ray crystallography of (2R,13S)- and (2S,13S)-1,2,4-tri-
methyl-3-(N-a-methylbenzylcarbamoyl)-1,2-dihydroquinoline ((2R)-
and (2S)-1,2-Me2MQPH (6)). Since the CD spectra for chiral 2-
methyl-1,2-dihydroquinolines or corresponding pyridine deriva-
tives structurally related to 3 with known configurations are
unavailable, it is necessary to prepare a single diastereomer
of a 1,2-dihydroquinoline derivative with an authentic chiral
auxiliary group. Considering the crystallizing ability, optical
properties and synthetic availability of the molecule, (2R,13S)- and
(2S,13S)-1,2,4-trimethyl-3-(N-a-methylbenzylcarbamoyl)-1,2-
dihydroquinoline ((2R)- and (2S)-1,2-Me2MQPH (6)) were chosen
as target molecules. Sodium borohydride reduction of (13S)-
1,2,4-trimethyl-3-(N-a-methylbenzylcarbamoyl)quinolinium ion
((13S)-Me2MQP+)31 is expected to afford a diastereomeric mixture
of (2R)- and (2S)-6 (Scheme 3).
Optical resolution and CD spectra of 3
The optical resolution and determination of the absolute configu-
ration of each enantiomer of 1 and 2 have already been reported.10
The CD spectra of 1 (Fig. S1†) further support the assignment.
HPLC equipped with a chiral column (Daicel CHIRALCEL OD)
is effective to separate the enantiomers of 2, and optically active
1 was obtained by the stereospecific oxidation of chiral 2 with
methyl benzoylformate in the presence of magnesium perchlorate
(Scheme 2).10,33 Due to the carbonyl rotation, it was very difficult
to isolate 1 in enantiomerically pure form; however, ee > 98% was
confirmed by HPLC analysis for both enantiomers of 1 used for
the reaction in this paper.
Scheme 2
Scheme 3
The enantiomers of 3 had remained unresolved under any
HPLC conditions in which (4R)-2 and (4S)-2 are completely
resolved.10 However, the recently developed HPLC column, Daicel
CHIRALPAK IA, made it possible to separate the enantiomers
of 3 as well as those of 2 (eluent: hexane–iPrOH–diethylamine =
95 : 5 : 0.1). Fig. 1 shows the mirror-image CD spectra of resolved
enantiomers of 3, demonstrating the validity of the resolution and
from which the absolute configuration has been determined as
described below.
Upon conducting the reaction depicted in Scheme 3, one of the
two diastereomers of 6 precipitated from the reaction solution. The
single crystals of this compound suitable for X-ray crystallography
were obtained by recrystallization from methanol. The remaining
reaction solution containing both diastereomers of 6 was subjected
to HPLC separation to obtain the other isomer. Single crystals
suitable for X-ray crystallography for the second diastereomer
were also obtained from methanol.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3834–3841 | 3835
©

View Article Online
Fig. 2 and 3 show the ORTEP drawing of the former
(precipitated from reaction solution) isomer and the separated
other isomer of 6, respectively. Since both compounds have no
heavy atoms, these X-ray analyses only establish the relative
stereochemistry. For the auxiliary (S)-phenylethylamine group, the
absolute configuration of the former isomer was determined to be
the 2R isomer and the latter was determined to be the 2S isomer. As
found in the previous report for 2-methyl-1,2-dihydropyridine,34
the methyl substituent on the C2 position (C11 in Fig. 2 and
3) is oriented axially to the 1,2-dihydroquinoline ring to avoid
steric hindrance from the alkyl substituents on the neighboring
C3 (C2 in Fig. 2 and 3) and ring nitrogen atoms. It is interesting
to note that in these crystal structures, the carbonyl orientation is
perpendicular to the dihydroquinoline ring and points in the same
direction as the hydrogen atom from C2 (C1 in Fig. 2 and 3).
Fig. 4 CD spectra of (2S)-6 (solid line) and (2R)-6 (broken line) in EtOH.
but diastereomers of each other. However, compounds (2R)- and
(2S)-7 (Chart 2), which were obtained by methylation of the amide
nitrogen atom in the side chain of (2R)- and (2S)-6, exhibited
nearly mirror image CD spectra with closely related Cotton
effect patterns with compound 3 (Fig. 5). Thus, the absolute
configuration of both enantiomers of 3 was unambiguously
determined as follows: the first-eluting isomer with a positive
Cotton effect at 235 nm is (2S)-3, and the second peak on the
HPLC with a negative Cotton effect at 235 nm corresponds to
(2R)-3.
Fig. 2 ORTEP representation (50% probability) of molecular structure
of (2R)-6.
Chart 2
Fig. 3 ORTEP representation (50% probability) of molecular structure
of (2S)-6.
Fig. 5 CD spectra of (2S)-7 (solid line) and (2R)-7 (broken line) in EtOH.
CD spectra of 6 and 1,2,4-trimethyl-3-(N-methyl-N-a-methyl-
benzylcarbamoyl)-1,2-dihydroquinoline (1,2-Me3MQPH (7)): de-
termination of absolute configuration of 3. The CD spectra of
(2R)- and (2S)-6 shown in Fig. 4 did not exhibit distinct mirror
images. This is because (2R)-6 and (2S)-6 are not enantiomers
Reduction of 1 with inorganic reducing agents
Reduction of optically active 1 was performed by sodium hydro-
sulfite in aqueous 0.5 M Na2CO3. Since a considerable extent
3836 | Org. Biomol. Chem., 2007, 5, 3834–3841
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Table 1 Regio- and enantioselectivity in 1,2-/1,4-reduction of 1 with several reducing agents
Product ratiob
Reductant
Na2S2O4
NaBH4
(4R)-4
(4S)-4
5
Configuration of 1a
2–3
4R–4S in 2
2R–2S in 3
Yield (%)
aR
aS
aR
aS
aR
aS
aR
aS
aR
aS
95 : 5 (0.5)
95 : 5 (1.0)
0 : 100
39 : 61 (5.8)
52 : 48 (1.4)
—
39 : 61 (7.1)
50 : 50 (1.4)
81 : 19 (2.4)
21 : 79 (0.7)
0.5 : 99.5 (0.25)
7 : 93 (1.2)
89 : 11 (1.3)
99 : 1 (0.3)
2 : 98 (1.6)
1–17c
9–40c
87
77
72
0 : 100
—
30 : 70 (1.6)
83 : 17 (2.5)
82 : 18 (1.2)
33 : 67 (2.8)
88 : 12 (1.2)
88 : 12 (0.7)
58 : 42 (2.7)
96 : 4 (1.5)
4 : 96 (0.7)
38 : 62 (4.0)
7 : 93 (1.9)
92 : 8 (2.0)
82
83
79
7–41c
9–53c
96 : 4 (3.3)
a Enantiomeric purity of the starting materials was >98%. b Errors in parentheses are one standard deviations derived from at least three independent
experiments. c Conversions determined by HPLC.
of racemization of 1 occurred during the reaction under the
present conditions, we decided to stop the reaction at 15 min,
at which point the racemization of 1 was found to be less than
10%. Table 1 shows the distribution of the reduction products
and yield/conversion of the reaction. Although the conversion of
the reaction varied from 1% to 40%, product ratios were fairly
consistent. Overall, sodium hydrosulfite promotes 1,4-reduction
predominantly with low enantioselectivity.
Reduction of optically active 1 with sodium borohydride was
examined in methanol at 0 ◦C. The reaction was completed
in two hours and the racemization of 1 under these reaction
conditions was negligible. Table 1 shows that sodium borohydride
reacts with 1 in a manner of 1,2-reduction with anti–syn = 4 : 1
enantioselectivity.
should be noted that the product selectivity in the reaction with 4
was not changed when the reaction was stopped at 1 h.
Discussion
The 1,2-dihydroquinolines 3, 6 and 7 are stable compounds and the
isomerization of 1,2-dihydroquinoline into 1,4-dihydroquinoline is
clearly excluded in the present system in contrast to the previous
report.29 This is probably due to the axial methyl orientation of
the C2 carbon as shown in the elucidated structures from X-ray
crystallography,34 forcing the C2 hydrogen into a less reactive
equatorial position. Scrambling reactions between NAD and
NADH model compounds36 were also found to be negligible in
the present system. Incubation of optically active 2 or 3 in the
presence of racemic 1 promoted neither racemization nor 1,2-/1,4-
isomerization under the present experimental conditions. Thus,
the product distribution shown in Table 1 unambiguously reflects
the selectivity of the reaction.
Reduction of 1 with NADH model compounds
The chiral NADH model compound (Me2PNPH, 4, Chart 1)35
with a restricted reaction face exhibited characteristic regio- and
enantioselectivity in the reduction of optically active 1 depending
on the configuration of both reactant and reducing agent. When
(aS)-1 was reduced by (4R)-4, the main product was (4R)-2,
which is the compound where the hydride was introduced at
the C4 position from the carbonyl oxygen face direction. On
the other hand, in the reaction of (aR)-1 with (4R)-4, the major
product was (2S)-3, which is the compound where the hydride
was introduced at the C2 position from the carbonyl oxygen face
direction (Scheme 1). Preliminary results from this reaction have
been reported earlier with the exception of the enantioselectivity
in 3.32 In this paper, we re-examined the whole set of the reactions
and completed the evaluation of the selectivity as shown in Table 1.
1-Propyl-1,4-dihydronicotinamide (PNAH (5), Chart 1) was
also employed as an NADH model compound without restriction
on the reaction face of the reducing agent. This compound
provides experimental evidence for the intrinsic preference of 1,2-
/1,4-reduction in the reactions between NADH and NAD model
compounds. Since this reaction has been found to be very slow,
the reaction was stopped at 1 h to evaluate the initial selectivity
of the reaction. The product ratio shown in Table 1 demonstrates
that this reaction proceeds in the preference of 1,4-reduction with
quite high enantioselectivity for both 1,2- and 1,4-reductions. It
In the reaction of 1 with reducing agents, a net “hydride” reacts
in four possible ways shown in Fig. 6, and the population of each
transition state reflects the product distribution ((4R)-2, (4S)-2,
(2R)-3 and (2S)-3). In Table 1, the product ratio of 2–3 reveals the
regioselectivity of the 1,4-/1,2-reduction, and the enantiomer ratio
in 2 and 3 reflects the face selectivity of the 1,4- and 1,2-reduction,
respectively.
In general, hydride reagents such as sodium borohydride
afford 1,2- or 1,6-isomers, while sodium hydrosulfite or NADH
model compounds give 1,4-isomers predominantly,27,28 although
some exceptions have been reported.6 Since the 1,6-reduction is
unfavorable in the present system because of the disruption of the
aromaticity of the phenyl ring, sodium borohydride exclusively
afforded 1,2-dihydroquinoline 3. The face selectivity of this
reaction was anti with respect to the carbonyl oxygen direction.
The anti selectivity of NaBH4 was reported previously6 and
understood by an electrostatic repulsion between carbonyl oxygen
and the negatively-charged reducing agent in the syn-face attack.
Sodium hydrosulfite afforded 1,4-dihydroquinolines in a low syn
preference in the present reaction, which is in good agreement with
previous results for NAD model compounds with fixed carbonyl
rotations.8,25,31 Since this reaction proceeds with partial (<10%)
racemization of the starting material as described above, the actual
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3834–3841 | 3837
©

View Article Online
Fig. 6 Four possible methods of (aR)-1 leading to different products.
stereoselectivity in this reaction may be slightly higher than the
values reported in Table 1.
is inverted from that found in the reaction of (aR)-1 with (4R)-4.
As discussed previously,32 the syn–trans–parallel arrangement is
considered as the most stable transition state in the viewpoints of
both theoretical37–40 and experimental5,6,8,10,41 investigations. The
antiparallel preference found in the former reaction results from
the lack of this most stable arrangement in the transition state.
Since the syn and trans arrangements have strong preferences over
anti and cis arrangements, respectively, the parallel preference was
sacrificed in the former reaction. In order to confirm the parallel
preference over antiparallel arrangement in the transition state
in a single experiment, the reaction of 1 with an achiral NADH
model compound 5 was investigated.
In the reaction of (aR)-1 with 5, all the eight possible
arrangements in the transition state can be adopted (Fig. 9).
Concomitantly, the preference of one parameter, i.e., the cis–trans
arrangement, cannot be estimated from the product ratio. The
main product of this reaction was (4S)-2, indicating syn–parallel
preference. The quantitative measures of the preference for these
two parameters other than cis–trans arrangements were syn–anti =
94 : 6 and parallel–antiparallel = 89 : 11, showing the evident
parallel preference.
Interestingly, in the reaction of (aR)-1 with (4R)-4, the enantios-
electivity in 2 is very poor compared to the other face selectivities
(Table 1). The molecular arrangement leading to these compounds
is anti–trans–parallel for the (4R)-2 and syn–cis–parallel for the
(4S)-2 (Fig. 7). The observed small preference for the (4R)-
product indicates that trans preference for the arrangement of
two carboxamide groups would be the dominant factor over the
syn orientation of the carbonyl group in 1. As a consequence,
the relative order of importance between these three parameters
should be: trans–cis arrangement for the carboxamide groups >
syn–anti orientation of carbonyl oxygen > parallel–antiparallel
arrangements of heteroaromatic rings.
In the reaction of (aS)-1 with (4R)-4, 1,4-reduction occurs
predominantly, affording (4R)-2 in a 96 : 4 enantiomer ratio. On
the other hand, 1,2-reduction is the main reaction giving (2S)-3
in a 99.5 : 0.5 enantiomer ratio when (aR)-1 is employed with the
same reducing agent. The carbonyl dipole controls the regio- and
stereoselectivity of the reaction (Scheme 1). It should be noted
that the phenylethylamine auxiliary in the side chain of 4 exhibits
a very small effect to determine the reaction face of the substrate.35
As discussed in previous papers,22,24,25,32 the following three
factors in the molecular orientations in the transition state should
be considered:
(i) syn–anti orientation of the carbonyl group of 1 with respect
to the reaction face,
(ii) cis–trans arrangement between the carbamoyl side chains of
1 and 4,
(iii) parallel–antiparallel (some papers used endo–exo notations)
arrangement with respect to the ring nitrogen orientation between
1 and 4 in the p–p stacking interaction.
These parameters differentiate eight (23 = 8) possible molecular
arrangements, from which half of such arrangements are excluded
using 4 as a reducing agent with a restricted reaction face blocked
by the 4-methyl group. For instance, the possible molecular
arrangements in the reaction of (aR)-1 with (4R)-4 are (i) anti–
trans–parallel, (ii) syn–cis–parallel, (iii) anti–cis–antiparallel and
(iv) syn–trans–antiparallel (Fig. 7). Assuming that the parallel
arrangement in the transition state promotes 1,4-reduction and
the antiparallel arrangement leads to 1,2-reduction, the four
molecular arrangements (i)–(iv) correspond to the transition states
leading to (4R)-2, (4S)-2, (2R)-3 and (2S)-3, respectively. The
ratio in the products reveals the preferences of each parameter
in this reaction in the following manner: syn–anti = 82 : 18, trans–
cis = 86 : 14 and parallel–antiparallel = 30 : 70, indicating the
strong dominance of the syn–trans arrangement with moderate
antiparallel preference.
Conclusion
The reaction of (aS)-1 with (4R)-4 highlights the other set of
arrangements from eight possible transition states (Fig. 8). In this
case, the preferences of each parameter estimated from the product
ratio of the reaction are: syn–anti = 81 : 19; trans–cis = 96 : 4 and
parallel–antiparallel = 83 : 17. Here again, the strong syn–trans
arrangement is emphasized. The parallel–antiparallel preference
The absolute configuration assignment for separated isomers of
1,2-dihydroquinoline 3 has been achieved. In the reduction of
NAD model compound 1 with NADH model compounds 4
and 5, the syn–trans–parallel arrangement in the transition state
of the reaction was proven to be the most stable in energy.
This clear-cut conclusion was obtained by the combined use of
3838 | Org. Biomol. Chem., 2007, 5, 3834–3841
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Fig. 7 Molecular arrangements in the transition state for the reaction of (aR)-1 with (4R)-4.
Fig. 8 Molecular arrangements in the transition state for the reaction of (aS)-1 with (4R)-4.
The Royal Society of Chemistry 2007 Org. Biomol. Chem., 2007, 5, 3834–3841 | 3839
This journal is
©

View Article Online
Fig. 9 Molecular arrangements in the transition state for the reaction of (aR)-1 with 5.
reaction-face-restricted and unrestricted NADH model com-
pounds. Although the assumption that the 1,2-reduction proceeds
via antiparallel arrangement remains to be proven, the importance
of the entropy contribution of the NADH model compounds
may support this idea.11,42,43 The present results demonstrate the
important contribution of the carboxamide group of the NAD
model compounds for controlling the entire reaction pathway and
provide valuable information for future organic catalyst design for
highly stereoselective reactions.
6.51 (d, J = 10.8 Hz, 1H), 6.70 (t, J = 9.2 Hz, 1H), 7.14–7.36 (m,
7H).
13C NMR (CDCl3) d 15.5, 21.6, 36.0, 49.0, 56.7, 111.4, 111.6,
116.8, 123.2, 124.8, 126.1, 127.4, 128.7, 129.7, 131.2, 143.0, 144.5,
168.2.
Anal. calcd for C21H24N2O: H, 7.55; C, 78.72; N, 8.74. Found:
H, 7.62; C, 78.44; N, 8.70%.
Crystal data for (2R)-6: C21H24N2O, monoclinic, space group
˚
˚
P21 with a = 12.3636(10) A, b = 5.2242(4) A, c = 13.3704(11)
◦
3
˚
˚
A, b = 97.482(2) , V = 856.24(12) A , Z = 2, R = 0.051, Rw =
0.102.‡
Experimental
(2S)-6. 1H NMR (CDCl3) d 1.10 (d, J = 6.6 Hz, 3H), 1.58 (d,
J = 6.9 Hz, 3H), 2.12 (s, 3H), 2.89 (s, 3H), 4.16 (q, J = 6.0 Hz,
1H), 5.27 (m, 1H), 5.88 (d, J = 7.5 Hz, 1H), 6.52 (d, J = 8.1 Hz,
1H), 6.70 (t, J = 7.5 Hz, 1H), 7.14–7.38 (m, 7H).
13C NMR (CDCl3) d 15.5, 15.7, 21.6, 36.1, 49.0, 56.8, 111.5,
116.8, 123.2, 124.8, 126.3, 127.6, 128.8, 129.8, 131.1, 142.8, 144.6,
168.2.
Anal. calcd for C21H24N2O: H, 7.55; C, 78.72; N, 8.74. Found :
H, 7.37; C, 78.64; N, 8.74%.
Crystal data for (2S)-6: C21H24N2O, orthorhombic, space group
(2R,13S)- And (2S,13S)-1,2,4-trimethyl-3-(N-a-
methylbenzylcarbamoyl)-1,2-dihydroquinoline ((2R)- and
(2S)-1,2-Me2MQPH (6))
To an argon-filled flask containing (13S)-1,2,4-trimethyl-3-(N-a-
methylbenzylcarbamoyl)quinolinium iodide31 ((13S)-Me2MQP+
I−) (0.53 g, 1.19 mmol) and sodium borohydride (0.11 g,
2.85 mmol), methanol (23 mL) was injected and the reaction
mixture was stirred for two hours at room temperature in the
dark. The resulting yellow precipitate was separated by filtration
to give diastereomerically pure (2R)-6 (70 mg, 0.22 mmol) in 18%
yield. The other diastereomer was obtained from the filtrate by
HPLC purification (column: Daicel CHIRALPAK IA 0.46␾ ×
25 cm; eluent: hexane–iPrOH–diethylamine = 95 : 5 : 0.1) to
give 13 mg of (2S)-6 (from 20 mg of diastereomer mixture
containing (2R)-6–(2S)-6 = 25 : 75). Both compounds were
recrystallized from methanol to afford single crystals suitable for
X-ray crystallography.
˚
˚
˚
P212121 with a = 5.3478(10) A, b = 13.519(3) A, c = 23.990(5) A,
3
˚
V = 1734.4(6) A , Z = 4, R = 0.039, Rw = 0.115.‡
(2R,13S)-1,2,4-Trimethyl-3-(N-methyl-N-a-
methylbenzylcarbamoyl)-1,2-dihydroquinoline
((2R)-1,2-Me3MQPH ((2R)-7))
To an argon-filled flask containing (2R)-6 (30 mg, 0.094 mmol) and
patassium tert-butoxide (69 mg, 0.61 mmol), dry THF (4.7 mL)
was injected. After stirring for 20 min, methyl iodide (0.08 mL,
1.3 mmol) was added and the reaction mixture was stirred for
two days at room temperature in the dark. After the precipitate
(2R)-6. 1H NMR (CDCl3) d 1.05 (d, J = 8.4 Hz, 3H), 1.58 (d,
J = 9.6 Hz, 3H), 2.11 (s, 3H), 2.89 (s, 3H), 4.20 (q, J = 8.4 Hz,
1H), 5.26 (dq, J = 9.6, 10.4 Hz, 1H), 5.95 (d, J = 10.4 Hz, 1H),
3840 | Org. Biomol. Chem., 2007, 5, 3834–3841
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
was removed by filtration, the filtrate was evaporated, dissolved
in dichloromethane, washed with water, dried over Na2SO4, and
evaporated to give 12.2 mg of crude product. This material was
purified by HPLC (Daicel CHIRALCEL OD 2.0␾ × 25 cm;
eluent = hexane–iPrOH–diethylamine = 95 : 5 : 0.1) to give 7.5 mg
(0.022 mmol, 24% yield) of pure (2R)-7 as a white powder.
7 C. Virty, J. Be´dat, Y. Prigent, V. Levacher, G. Dupas, I. Salliot, G.
Que´guiner and J. Bourguignon, Tetrahedron, 2001, 57, 9101–9108.
8 J.-L. Vasse, V. Levacher, J. Bourguignon and G. Dupas, Tetrahedron:
Asymmetry, 2002, 13, 227–232.
9 J.-L. Vasse, V. Levacher, J. Bourguignon and G. Dupas, Tetrahedron,
2003, 59, 4911–4921.
10 Y. Mikata, K. Hayashi, K. Mizukami, S. Matsumoto, S. Yano, N.
Yamazaki and A. Ohno, Tetrahedron Lett., 2000, 41, 1035–1038.
11 R. Saito, S. Naruse, K. Takano, K. Fukuda, A. Katoh and Y. Inoue,
Org. Lett., 2006, 8, 2067–2070.
(2R)-7. 1H NMR (CDCl3) d 1.01 (br.d, J = 6.0 Hz), 1.16 (d,
J = 6.6 Hz), 1.24 (br.), 1.57–1.67 (m, 3H), 1.93 (s), 2.00 (s), 2.08
(s), 2.70 (s), 2.73 (s), 2.78 (s), 2.88 (s), 2.91, (s), 3.79–3.86 (m), 4.46
(br.), 5.56 (q, J = 6.9 Hz), 6.16 (br.q), 6.26 (br.q), 6.52 (m), 6.74
(m), 7.15–7.39 (m).
13C NMR (CDCl3) d 15.0, 15.3, 15.8, 16.1, 18.6, 27.2, 29.7,
36.0, 49.6, 49.9, 56.7, 57.3, 111.0, 111.5, 116.3, 116.9, 123.4, 124.2,
124.3, 126.4, 127.4, 128.5, 128.7, 129.2, 130.2, 140.8, 144.2, 170.9.
ESI-MS: m/z 357.19353 (M + Na+. C22H26N2ONa requires
357.19428).
12 M. Fujii, K. Nakamura, S. Yasui, S. Oka and A. Ohno, Bull. Chem.
Soc. Jpn., 1987, 60, 2423–2427.
13 M. J. de Nie-Sarink and U. K. Pandit, Tetrahedron Lett., 1979, 2449–
2452.
14 T. J. van Bergen, D. M. Hedstrand, W. H. Kruizinga and R. M. Kellogg,
J. Org. Chem., 1979, 44, 4953–4962.
15 S. Yasui, K. Nakamura, M. Fujii and A. Ohno, J. Org. Chem., 1985,
50, 3283–3287.
16 N. Ono, R. Tamura and A. Kaji, J. Am. Chem. Soc., 1983, 105, 4017–
4022.
17 M. Fujii, K. Nakamura, H. Mekata, S. Oka and A. Ohno, Bull. Chem.
Soc. Jpn., 1988, 61, 495–500.
18 J. W. Yang, M. T. H. Fonseca and B. List, Angew. Chem., Int. Ed., 2004,
43, 6660–6662.
19 J. W. Yang, M. T. H. Fonseca, N. Vignola and B. List, Angew. Chem.,
Int. Ed., 2005, 44, 108–110.
(2S,13S)-1,2,4-Trimethyl-3-(N-methyl-N-a-
methylbenzylcarbamoyl)-1,2-dihydroquinoline
((2S)-1,2-Me3MQPH ((2S)-7))
20 S. G. Ouellet, J. B. Tuttle and D. W. C. MacMillan, J. Am. Chem. Soc.,
2005, 127, 32–33.
Starting from a diastereomeric mixture of 6 (60 mg (0.19 mmol),
(2R)-6–(2S)-6 = 20 : 80), a similar procedure to that above afforded
54.3 mg of crude product. Recrystallization of this material from
methanol afforded 17 mg (0.051 mmol) of diastereomerically pure
(2S)-7 (34% yield based on (2S)-6).
21 P. M. T. de Kok and H. M. Buck, J. Chem. Soc., Chem. Commun., 1985,
1009–1010.
22 A. Ohno, M. Ogawa and S. Oka, Tetrahedron Lett., 1988, 29, 1951–
1954.
23 A. Ohno, A. Tsutsumi, Y. Kawai, N. Yamazaki, Y. Mikata and M.
Okamura, J. Am. Chem. Soc., 1994, 116, 8133–8137.
24 A. Ohno, N. Yamazaki, M. Okamura, Y. Kawai, A. Tsutsumi, Y. Mikata
and M. Fujii, Bull. Chem. Soc. Jpn., 1996, 69, 1093–1098.
25 A. Ohno, A. Tsutsumi, N. Yamazaki, M. Okamura, Y. Mikata and M.
Fujii, Bull. Chem. Soc. Jpn., 1996, 69, 1679–1685.
26 N. Kanomata and T. Nakata, J. Am. Chem. Soc., 2000, 122, 4563–4568.
27 U. Eisner and J. Kuthan, Chem. Rev., 1972, 72, 1–42.
28 R. M. Acheson and G. Paglietti, J. Chem. Soc., Perkin Trans. 1, 1976,
45–48.
(2S)-7. 1H NMR (CDCl3) d 1.08 (br.), 1.19 (d, J = 6.0 Hz),
1.25 (br.d), 1.57 (s), 1.59 (s), 1.69 (s), 1.85 (s), 2.01 (s), 2.07 (s),
2.69 (br.), 2.77 (br.), 2.86 (s), 2.90 (s), 2.94 (s), 3.80 (br.q), 3.96 (q,
J = 6.3 Hz), 4.45 (br.), 5.63 (q, J = 6.9 Hz), 6.14 (m), 6.54 (m),
6.73 (m), 7.20–7.38 (m).
13C NMR (CDCl3) d 15.0, 16.1, 17.6, 17.7, 26.9, 29.9, 36.1, 36.4,
36.5, 49.4, 50.5, 56.3, 56.4, 57.3, 57.6,111.3, 111.5, 111.6, 116.8,
116.9, 123.5, 124.2, 124.4, 126.0, 126.8, 127.3, 128.5, 129.3, 130.0,
130.4, 139.8, 140.3, 144.0, 144.2, 170.2, 170.8.
ESI-MS: m/z 357.19287 (M + Na+. C22H26N2ONa requires
357.19428).
29 R. M. G. Roberts, D. Ostovic and M. M. Kreevoy, J. Org. Chem., 1983,
48, 2053–2056.
30 F. Rob, H. J. van Ramesdonk, W. van Gerresheim, P. Bosma, J. J. Scheele
and J. W. Verhoeven, J. Am. Chem. Soc., 1984, 106, 3826–3832.
31 A. Ohno, M. Kashiwagi, Y. Ishihara, S. Ushida and S. Oka, Tetrahe-
dron, 1986, 42, 961–973.
32 Y. Mikata, S. Aida and S. Yano, Org. Lett., 2004, 6, 2921–2924.
33 Y. Mikata, K. Mizukami, K. Hayashi, S. Matsumoto, S. Yano, N.
Yamazaki and A. Ohno, J. Org. Chem., 2001, 66, 1590–1599.
34 E. E. Knaus, H. Wynn, M. W. Wolowyk and R. G. Ball, Acta
Crystallogr., Sect. C: Cryst. Struct. Commun., 1987, C43, 1734–1737.
35 A. Ohno, M. Ikeguchi, T. Kimura and S. Oka, J. Am. Chem. Soc., 1979,
101, 7036–7040.
Acknowledgements
We thank Dr K. Murazumi of the Daicel Company for helpful
advice on the HPLC measurements.
36 M. F. Powell and T. C. Bruice, J. Am. Chem. Soc., 1983, 105, 7139–7149.
37 M. C. A. Donkersloot and H. M. Buck, J. Am. Chem. Soc., 1981, 103,
6554–6558.
References
1 V. A. Burgess, S. G. Davies and R. T. Skerlj, Tetrahedron: Asymmetry,
1991, 2, 299–328.
2 Y. Combret, J.-J. Torche´, N. Ple´, J. Duflos, G. Dupas, J. Bourguignon
and G. Que´guiner, Tetrahedron, 1991, 47, 9369–9382.
3 Y. Combret, J. Duflos, G. Dupas, J. Bourguignon and G. Que´guiner,
Tetrahedron: Asymmetry, 1993, 4, 1635–1644.
38 Y.-D. Wu and K. N. Houk, J. Org. Chem., 1993, 58, 2043–2045.
39 Y.-D. Wu, D. K. W. Lai and K. N. Houk, J. Am. Chem. Soc., 1995, 117,
4100–4108.
¨
40 O. Almarsson and T. C. Bruice, J. Am. Chem. Soc., 1993, 115, 2125–
2138.
4 Y. Combret, J. Duflos, G. Dupas, J. Bourguignon and G. Que´guiner,
Tetrahedron, 1993, 49, 5237–5246.
5 J.-L. Vasse, G. Dupas, J. Duflos, G. Que´guiner, J. Bourguignon and V.
Levacher, Tetrahedron Lett., 2001, 42, 3713–3716.
6 J.-L. Vasse, G. Dupas, J. Duflos, G. Que´guiner, J. Bourguignon and V.
Levacher, Tetrahedron Lett., 2001, 42, 4613–4616.
41 P. M. T. de Kok, L. A. M. Bastiaansen, P. M. van Lier, J. A. J. M.
Vekemans and H. M. Buck, J. Org. Chem., 1989, 54, 1313–1320.
42 A. Ohno, M. Goto, Y. Mikata, T. Kashiwagi and T. Maruyama, Bull.
Chem. Soc. Jpn., 1991, 64, 87–90.
43 M. Okamura, Y. Mikata, N. Yamazaki, A. Tsutsumi and A. Ohno,
Bull. Chem. Soc. Jpn., 1993, 66, 1191–1196.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3834–3841 | 3841
©