Preparation of axially chiral quinolinium salts related to NAD+ models: New investigations of these biomimetic models as 'chiral amide-transferring agents'

Article abstract of DOI:10.1016/j.tetasy.2004.11.004

The general purpose of this work is to investigate the potential of biomimetic NAD⁺ models as 'nucleophile-transferring agents' with the ultimate motivation to develop new synthetic tools. This first report focuses on the preparation of an axially chiral quinolinium salt 8. A preliminary investigation of these NAD⁺ analogues as 'chiral amide-transferring agents' is reported herein. The synthesis of the desired quinolinium salt 8 was first attempted via a Friedlaender approach. Given the poor reproducibility of this first synthetic route, a second strategy making use of an intramolecular nickel-catalyzed coupling was developed with success, furnishing the quinolinium salt 8 in 12% overall yield. The potential of the quinolinium salt 8 as a 'chiral amide-transferring agent' was then investigated. Regioselective 1,4-addition of benzylamine and piperidine produced, respectively, adducts 18a and 18b with high diastereoselectivity (de >95%). The resulting 'chiral masked-amide' 18b was reacted with various activated aryl esters affording the corresponding atropisomeric amide 20 with modest atropenantioselectivity (ee = 2-20%).

Full text of DOI:10.1016/j.tetasy.2004.11.004

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 3919–3928
Preparation of axially chiral quinolinium salts related to
NAD+ models: new investigations of these biomimetic models as
ꢀchiral amide-transferring agentsꢁ
*
´
Stephane Leleu, Cyril Papamicael, Francis Marsais, Georges Dupas and Vincent Levacher
¨
´ ´ ´
Laboratoire de Chimie Organique Fine et Heterocyclique de IRCOF, UMR 6014, CNRS. Universite et INSA de Rouen.
´
B.P. 08 F-76131 Mont Saint Aignan Cedex, France
Received 28 September 2004; accepted 3 November 2004
Available online 24 November 2004
Abstract—The general purpose of this work is to investigate the potential of biomimetic NAD+ models as ꢀnucleophile-transferring
agentsꢁ with the ultimate motivation to develop new synthetic tools. This first report focuses on the preparation of an axially chiral
quinolinium salt 8. A preliminary investigation of these NAD+ analogues as ꢀchiral amide-transferring agentsꢁ is reported herein.
The synthesis of the desired quinolinium salt 8 was first attempted via a Friedla¨nder approach. Given the poor reproducibility
of this first synthetic route, a second strategy making use of an intramolecular nickel-catalyzed coupling was developed with success,
furnishing the quinolinium salt 8 in 12% overall yield. The potential of the quinolinium salt 8 as a ꢀchiral amide-transferring agentꢁ
was then investigated. Regioselective 1,4-addition of benzylamine and piperidine produced, respectively, adducts 18a and 18b with
high diastereoselectivity (de >95%). The resulting ꢀchiral masked-amideꢁ 18b was reacted with various activated aryl esters affording
the corresponding atropisomeric amide 20 with modest atropenantioselectivity (ee = 2–20%).
ꢀ 2004 Elsevier Ltd. All rights reserved.
COR1
1. Introduction
R*
Nu-E
In the course of these three decades, a large number of
papers have been reported dealing with the NAD+/
NADH redox couple. Many research groups have
focused their efforts on the design of chiral biomimetic
NADH models,1a–e with the main goal to throw light
on the high stereospecificity of the hydride transfer
observed during biochemical processes. In parallel, this
biomimetic approach led to the development of new syn-
thetic tools, almost exclusively in the field of enantiose-
lective reduction. To extend further the potential of
these biomimetic tools, one can explore their aptitude
for transferring groups other than ꢀhydrideꢁ to a sub-
strate. Diverse nucleophiles such as cyanides,2a–k thio-
lates,3 alcoholates,4a–d amides5a–f and enolates6 have
been reported to add to a variety of pyridinium salts re-
lated to NAD+ models, yielding the corresponding 1,4-
adducts (Fig. 1). These resulting ꢀmasked nucleophilesꢁ
are expected to react with an electrophile, providing
the desired products (Nu–E) along with the recovered
N
X-
"nucleophile shuttle"
+
Nu-
"releasing step"
E+, X-
"anchoring step"
H
Nu
COR1
R*
Nu- = CN-, RS-, RO-, R2N-, enolate
N
"masked nucleophile"
Figure 1. New investigations of NAD+ models as ꢀnucleophile-trans-
ferring agentsꢁ.
pyridinium salt. The former may be considered in this
process as a ꢀnucleophile shuttleꢁ. These ꢀmasked nucleo-
philesꢁ might provide new synthetic tools in diverse areas
of synthetic organic chemistry. In particular, in the field
of asymmetric synthesis, chiral NAD+ models should
offer the opportunity to investigate the potential of
ꢀchiral masked nucleophilesꢁ in enantioselective transfor-
mations (Fig. 1).
*
One must briefly mention the pioneering work of Kel-
logg, reported some 20 years ago and related to the
Corresponding author. Tel.: +33 02 35 52 24 85; fax: +33 02 35 52 29
62; e-mail: Vincent.Levacher@insa-rouen.fr
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.11.004

3920
S. Leleu et al. / Tetrahedron: Asymmetry 15 (2004) 3919–3928
development of these NAD+ models as ꢀnucleophile-
transferring agentsꢁ.7 Their first paper deals with the
addition of thiolates to esters or amides via 1-methylpy-
ridinium-3,5-dicarboxylic acid salts. The resulting
ꢀmasked thiolateꢁ reacts with activated acid-derivatives
to produce the thioesters in excellent yields, and this,
under neutral conditions (Fig. 2a). Depending on the
nature of the thiols, the authors noticed the formation
of 1,2-adduct in addition to the expected 1,4-regio-
isomer. In a second paper, Kellogg reports attempts to
use a chiral bridged pyridinium salt6 as a chiral template
to realize an enantioselective aldol condensation (Fig.
2b). Although the condensation occurs smoothly to pro-
duce the corresponding aldols in moderate to good
yields, no enantioselectivity was observed (Fig. 2b). A
second problem arises from the regioselectivity of the
initial enolate addition. As observed with thiolates,
significant amounts of 1,2-adduct are formed in many
cases, which may complicate seriously the rational
design of these pyridinium salts as new synthetic tools.
develop these resulting ꢀchiral masked amidesꢁ as new
tools in stereoselective synthesis. To this end, it appears
attractive to evaluate their potential during atropenan-
tioselective amidification of aryl esters (Fig. 3).
O
O
Me
N
R1
Me
Me
N
R2
R1
N
Me
R2
Me
N
H
8
N
ee ?
Me
R2
R1
O
O
N
N
OR
Me
N
N
Me
de ?
"Chiral masked amide"
Figure 3. Design of a new axially chiral quinolinium salt as an ꢀamide-
transferring agentꢁ. Application to the atropenantioselective amidifi-
cation of aryl esters.
ROC
COR
R2COSR1
R1S-
N
Yield = 37-90%
2. Results and discussion
R1S
ROC
H
2.1. Design and preparation of the quinolinium salt 8
COR
R2COX
N
Quinolinium salt 8 was selected as the first candidate for
these investigations. This choice was initially motivated
by our previous work on the development of axially chi-
ral NADH models.8a–c The guiding principle behind the
design of these biomimetic models has been the configu-
rational control of the chiral axis C3-C@O, insured by
the presence of an additional configurational element
(chiral inducer in Fig. 4) installed on the lactam moiety.
In the course of this work, it was found that this chiral
axis C3-C@O is the main configurational element
responsible for the highly diastereoselective 1,4-reduc-
tion of these quinolinium salts. Accordingly, this config-
urational element is expected to play a key role, in terms
of diastereoselectivity, during the ꢀanchoring stepꢁ of the
amine on the quinolinium salt 8. It is noteworthy that
depending on the nature of the hydride, a complete
opposite diastereofacial selectivity was previously ob-
served during the reduction of these axially chiral quino-
linium salts. One could ascribe this result to a pre-
complexation of diborane on the carbonyl lactam, while
sodium borohydride would react on the opposite face,
as a result of a steric control (Fig. 4).
(a) : "masked thiolate"
Bridge
H
R
R
H
NH
HN
OH
R2
O
O
O
O
R1
N
R1
Yield = 37-90%
racemic
Bridge
O
R1
HN
H
R
R
H
NH
H
O
O
R2CHO
N
(b) : "chiral masked enolate"
Figure 2. (a) Thiolate-transferring NAD+ models used in the forma-
tion of thioesters. (b) Enolate-transferring NAD+ models used in aldol
condensation.
Chiral axis (C3-C=O):
active configurational element
To the best of our knowledge, no steps forward have
been accomplished in this area since this pioneering
work. We report herein the use of quinolinium salts as
ꢀamide-transferring agentsꢁ. We speculated that the use
of quinolinium salts would be more appropriate to de-
velop such applications by avoiding the formation of
1,2 adducts. We essentially concentrate here on the de-
sign and preparation of a new axially chiral quinolinium
8 (Fig. 3). Additionally, we demonstrated the diastereo-
selective 1,4-addition of amines with the perspective to
BH3
R2
R2
N
O
O
N
Chiral inducer
Me
Me
N
R1
+
R1
+
H
H
3
NaBH4
N
TfO-
TfO-
Figure 4. Diastereoselective 1,4-reduction of axially chiral quinolinium
salt.

S. Leleu et al. / Tetrahedron: Asymmetry 15 (2004) 3919–3928
3921
Our initial retrosynthetic analysis for the synthesis of 8
CHO
NO2
Tol
Tol
(a)
N
is inspired by our previous work.8a–c This approach is
based on a Friedla¨nder condensation between benzaze-
pin-3,5-dione 4 and aniline 6 (Scheme 1).
(c)
80%
(b)
N
42%
80%
NO2
NH2
5
6
O
O
O
N
N
(d)
O
Me
Me
Tol
N
N
95%
N
N
7
N
Me
Me
Me
+
TfO
N
NH2
O
8
Me
6
8
4
Scheme 3. Synthesis of 8 via Friedla¨nder approach. Reagents and
conditions: (a) p-toluidine/EtOH/reflux/2h; (b) Na2SÆH2O/EtOH/
reflux/10min; (c) benzazepin-3,5-dione 4/EtOH/piperidine/reflux/48h;
(d) TfOMe/CH2Cl2/20ꢁC/2h.
Scheme 1. Retrosynthetic analysis of 8 via a Friedla¨nder approach.
The key intermediate benzazepin-3,5-dione 4 was pre-
pared within three steps from (S)-1-phenylethylamine.
First reductive amination in the presence of acetone
afforded 1 in 90% yield, which in turn was condensed
with ethyl malonyl chloride to give b-amide ester 2 in
79% yield. Hydrolysis of 2 furnished carboxylic acid 3
in 96% yield, which was subsequently converted into
its acid chloride, which was submitted to intramolecular
Friedel–Crafts acylation to provide the desired benzaze-
pin-3,5-dione 4, however, in only 30% yield (Scheme 2).
Amide bond
formation
O
O
N
N
Me
N
Cl
N
I
TfO
12
8
intramolecular
cross-coupling reaction
HN
COOH
(b)
(a)
N
NH2
O
N
H
+
I
90%
79%
N
Cl
O
1
CO2R
2 : R=Et
3 : R=OH
9
10
N
(c) 96%
Scheme 4. Retrosynthetic analysis of 8 via an intramolecular coupling
reaction.
Me
(d)
O
30%
4
The first key intermediate 9 was prepared by lithiation
of 2-chloroquinoline with n-BuLi in the presence of
TMEDA according to a known procedure.11 The lithi-
ated intermediate was trapped with carbon dioxide to
afford acid 9 in 65% yield. The second key intermediate
10 was envisioned to arise from an ortho-lithiation of 1.
This approach was inspired by previous investigations
on ortho-lithiation of benzylamine derivatives.12 After
many experimentations using various metallating agents
(n-BuLi, tert-BuLi, sec-BuLi) and solvents (Et2O, THF,
cyclohexane/THF), n-BuLi in the presence of TMEDA
in THF was found to offer the best conditions, affording
after treatment of the lithiated intermediate with iodine,
compound 10 in 50% yield. The latter compound was
converted into its acetyl derivative 11 and subsequently
subjected to chiral HPLC analysis, demonstrating that
no significant racemization occurred during the lithia-
tion of 1. With both key intermediates in hand, quino-
line 9 was converted into its acid chloride prior to
reacting with the amine 10 to supply 12 in 60% yield.
We next investigated the crucial intramolecular coupling
reaction by using first the Ullmann coupling conditions.
However, all attempts to cyclize 12 in the presence of
copper powder in refluxing DMF failed, providing 7 in
only 10% isolated yield. In search of more efficient con-
ditions, we turned our attention to a nickel-catalyzed
coupling reaction, which is reported to proceed under
milder conditions as compared with the classical
Scheme 2. Synthesis of benzazepin-3,5-dione 4. Reagents and condi-
tions: (a) acetone/reflux/12h then NaBH4/20ꢁC/2h; (b) ClCO-
CH2CO2Et/CH2Cl2/NEt3/12h/20ꢁC; (c) KOH/EtOH/0ꢁC/5h; (d)
(COCl)2/two drops of DMF/CH2Cl2 then AlCl3/CH2Cl2/1h/20ꢁC.
The classical Friedla¨nder conditions commonly make
use of an ortho-aminobenzaldehyde.9 Nevertheless, the
poor stability of the former, which may severely limit
the yield of the reaction, prompted us to undertake the
Borsche modification.10 The stable amino imine 6 was
straightforwardly prepared in two steps on a multigram
scale from 2-nitrobenzaldehyde (42% and 80% yields).
Friedla¨nder cyclization was carried out in EtOH in the
presence of piperidine to produce the quinoline 7 in
80% yield. The later product was subsequently quatern-
ized in a quantitative yield to give the final product 8
(Scheme 3). This first synthetic route suffers from a
low overall yield and a lack of reproducibility in certain
steps, making difficult to scale up this Friedla¨nder
approach.
An intramolecular metal-catalyzed coupling reaction of
12 was envisaged as an alternative route for the prepara-
tion of 8. This new strategy required ready access to 2-
chloro-3-quinolinecarboxylic acid 9 and (S)-2-iodo-a-
methyl-N-isopropylbenzylamine 10 as key intermediates
(Scheme 4).

3922
S. Leleu et al. / Tetrahedron: Asymmetry 15 (2004) 3919–3928
COOH
(a)
O
65%
N
N
CI
Cl
N
(d)
9
Me
60%
N
Cl
Me
Me
I
12
i-Pr
i-Pr
N
(b)
N
H
H
64% (e)
50%
I
O
1
10 : R=H
11 : R = Ac
N
(c)
50%
Me
N
7
Scheme 5. Preparation of 7 via an intramolecular nickel-catalyzed coupling reaction. Reagents and conditions: (a) n-BuLi/TMEDA/THF/78ꢁC/
20min then CO2; (b) n-BuLi/TMEDA/THF/ꢀ78ꢁC/2h then I2 in THF/12h/20ꢁC; (c) MeCOCl/NEt3/CH2Cl2/20ꢁC/2h; (d) (COCl)2/two drops of
DMF/CH2Cl2 then 10/12h; (e) NiBr2(PPh3)2/Zn/Et4N+Iꢀ/THF/2h/50ꢁC.
Ullmann conditions.13a–d The reductive coupling of 12
was carried out using NiBr2(PPh3)2 and zinc in the pres-
ence of Et4NI, affording 7 with a satisfactory yield
(64%). This new approach presents significant advan-
tages over the former Friedla¨nder synthetic route, in
particular it circumvents the poorly reproducible intra-
molecular Friedel–Crafts step (Scheme 5).
ium salt 14 could be isolated in excellent yield and re-
used with success in a second experiment without
erosion of the yield. This amide transfer may present a
synthetic interest, as it results in the formation of an
amide bond by means of an ꢀamide anion equivalentꢁ
under mild and neutral conditions (Scheme 7).
COOEt
2.2. Investigation of the chiral quinolinium salt 8as an
‘amide-transferring agent’
R1
R2
MeCONR1R2
N
Me
Ph
14
R1 R2
N
(a)
N
H
17a : R1=H ; R2=Bn
TfO
We first focused our attention on the optimization of the
two fundamental steps of this amide transfer process,
that is the ꢀanchoring stepꢁ of the amine followed by
the ꢀreleasing stepꢁ of the resulting ꢀmasked amideꢁ. This
optimization was achieved with the achiral quinolinium
salt 14, chosen as a model. It was prepared in two steps
by condensation of 6 with ethyl benzoylacetate to give
quinoline 13 in 72% yield. Quaternization of 13 afforded
quinolinium salt 14 in 95% yield (Scheme 6).
17b : R1=R2= -(CH2)4-
"anchoring step"
(b)
"releasing step"
MeCOCl
COOEt
Ph
15a : R1=H ; R2=Bn
15b : R1=R2= -(CH2)5-
N
Me
16a : R1=H ; R2=Bn (90%)
16b : R1=R2= -(CH2)4- (95%)
Scheme 7. Achiral quinolinium salt 14 as ꢀamide transferring-agentꢁ.
Reagents and conditions: (a) amines 15/Na2CO3/H2O/CH2Cl2/20ꢁC;
(b) CH2Cl2/20ꢁC.
COOEt
Ph
COOEt
(b)
95%
Tol
N
(a)
72%
N
N
13
NH2
Ph
We finally evaluated the potential of these new ꢀamide
transferring-agentsꢁ during the atropenantioselective
amidification of aryl esters by means of the chiral quino-
linium salt 8. While a large amount of work deals with
these atropoisomeric amides, only a few stereoselective
methods have been investigated to have access to this
class of atropisomers.14 This new approach would sup-
ply a straightforward and elegant access to enantiomer-
ically enriched atropisomeric benzamides. We first
examined the diastereoselective addition of both amines
15a and 15b. Under the experimental conditions previ-
ously optimized, we were pleased to find out that ad-
ducts 18a and 18b were formed quantitatively as a
single diastereoisomer in both cases. The relative config-
uration at C-4 was assigned with the aid of 2D-NOESY
experiments, showing dipolar coupling between the
methyl group of the lactam moiety and the methylene
protons adjacent to the nitrogen of the amine (Scheme
8). This information clearly indicates that both amines
15a,b react on the opposite face with respect to the
6
14
TfO
Scheme 6. Preparation of the achiral quinolinium salt 14. Reagents
and conditions: (a) ethyl benzoylacetate/EtOH/piperidine/reflux/48h;
(b) TfOMe/CH2Cl2/20ꢁC/2h.
The ꢀanchoring stepꢁ of both benzylamine 15a and pip-
eridine 15b on the quinolinium salt 14 was conducted
under biphasic conditions (CH2Cl2/aqueous Na2CO3
solution). Under these conditions, adducts 16a and 16b
could be isolated in 90% and 95% yields, respectively.
As expected, and in contrast to what has been previously
observed with pyridinium salts, addition of the nucleo-
phile takes place exclusively at the C-4 position of the
quinolinium salt 14. These resulting ꢀmasked amidesꢁ
16a and 16b could be released, under neutral reaction
conditions, by treatment with acetyl chloride in CH2Cl2,
to produce the corresponding carboxamides 17a and 17b
in nearly quantitative yields. Importantly, the quinolin-

S. Leleu et al. / Tetrahedron: Asymmetry 15 (2004) 3919–3928
3923
C@O lactam. The excellent diastereofacial selectivity
observed is assumed to originate from a steric control
exerted by the carbonyl lactam. This analysis is fully
consistent with our previous explanation which accounts
for the sense of the stereoselection observed during the
reduction of these axially chiral NAD+ models (Fig.
4). Finally, these results suggest that the chiral axis de-
fined by the C3-C@O bond is the main configurational
element governing the stereoselectivity of these addi-
tions, as already observed in the course of the reduction
of these biomimetic models (Scheme 8).
conducted at higher temperature (Table 1, entry 2).
Assuming that the former ester is too poorly reactive,
we envisaged to activate carboxylic acid 19a using
EDCI/HOBt coupling conditions. These new conditions
provided the quinoline amide 20 in good yield, however,
with insignificant enantioselectivity. (Table 1, entry 3).
When the reaction is carried out at lower temperature,
the yield is considerably affected, and this without
observing any improvement of the enantioselectivity
(Table 1, entry 4). Among the large variety of coupling
reagents at our disposal, a 2-halopyridinium type cou-
pling reagent was examined. Thus, 1-ethyl-2-fluoropyr-
idinium salt (FEP)18 was reacted with quinoline
carboxylic acid 19a, according to standard procedures,
to furnish the presumed reactive (2-acyloxy)pyridinium
salt. This resulting intermediate was then reacted with
18b at various temperatures (Table 1, entries 5–7). A
notable atropenantioselectivity was obtained at ꢀ60ꢁC
(Table 1, entry 7). Although the degree of stereoselectiv-
ity remains moderate, this result constitutes the first
atropenantioselective amidification of activated aryl
esters providing a proof of concept of this approach.
O
H
N
i-Pr
de >95%
N
15b
Me
N
H
O
N
18b
i-Pr
(a)
Me
N
H
H
TfO-
O
+
Ph
H
N
8
i-Pr
N
15a
de >95%
Me
H
N
noe observed
3. Conclusion
18a
In the course of this work, we have demonstrated that
quinolinium salts, structurally related to NAD+ models,
may be used as ꢀamide-transferring agentsꢁ. Indeed,
amines were found to add regioselectively at C4 of a bio-
mimetic NAD+ model 14 in quantitative yields. The
resulting adducts 16a and 16b could react with acetyl
chloride under neutral conditions to give the corre-
sponding amides 17a and 17b and the recovered quino-
linium salt 14 in excellent yields. We then turned our
interest in the preparation of an axially chiral quinolin-
ium salt 8 to investigate its potential as ꢀchiral amide-
transferring agentsꢁ. This new chiral synthetic tool was
assessed in an ꢀatropenantioselective amidificationꢁ of
aryl esters with moderate success (ee = 2–20%).
Although these preliminary results are presently modest
in term of enantioselectivity, the atropenantioselectivity
observed is encouraging and laid down the basis of
future developments.
Scheme 8. Diastereoselective formation of ꢀchiral masked amidesꢁ 18a
and 18b. Reagents and conditions: (a) CH2Cl2/Na2CO3/20ꢁC.
We next examined the reaction of the ꢀchiral masked
amideꢁ 18b with various activated 2,4-dimethylquino-
line-3-carboxylic esters, obtained from carboxylic acid
19a.17 The main results of this atropenantioselective
amidification leading to the atropisomeric amide 2015
are summarized in Table 1. The racemic mixture of 20
was prepared in 80% yield from carboxylic acid 19a
under standard EDCI/HOBt/CH2Cl2 conditions. Both
enantiomers of 20 could be separated by conventional
chiral HPLC on chiracel OD. In a first experiment, sub-
strate 19a was converted into ester 19b prior to reacting
with the ꢀchiral masked amideꢁ 18b in CH2Cl2 at ꢀ10ꢁC
(Table 1, entry 1). Under these conditions, no transfer of
the piperidine occurred and only the starting materials
were recovered. The same is true when the reaction is
Table 1. Atropenantioselective amidification of quinoline carboxylic acid 19a
O
Me
Me
Me
N
O
COOH
Step A
X
N
O
18b
Step B
N
Me
N
Me
Me
19a
20
Entry
Step A
X
Step B (ꢁC)
Yield (%)
Ee (%)
1
2
3
4
SOCl2 then
19b:PNP
19b:PNP
Bt
ꢀ10
25
0
0
/
p-Nitrophenol
EDCl/HOBt
EDCl/HOBt
/
25
ꢀ60
85
20
3
2
Bt
5
6
7
FEP
FEP
FEP
25
ꢀ10
ꢀ60
80
85
60
5
12
20
N
O
+
-
BF4
Et

3924
S. Leleu et al. / Tetrahedron: Asymmetry 15 (2004) 3919–3928
4. Experimental
13.9, 17.7, 18.0, 19.6, 20.5, 21.4, 21.9, 43.3, 47.7, 48.8,
51.8, 55.2, 61.2, 126.4, 126.7, 126.9, 127.5, 127.9,
141.4, 165.3, 167.9. Anal. Calcd for C16H23NO3: C,
69.29; H, 8.36; N, 5.05. Found: C, 69.45; H, 8.24; N,
4.95.
4.1. General methods
Infrared spectra were recorded on a Beckmann IR4250
1
spectrometer. H and 13C spectra were recorded on a
200 and 300MHz Bruker apparatus and calibrated with
the residual undeuterated solvent. Spectra were recorded
in deuteriochloroform (CDCl3), in hexadeuteriodimeth-
ylsulfoxide (DMSO-d6) or pentadeuteriomethanol
(MeOD-d4). The following compounds were prepared
according to literature procedures: ethyl malonyl chlo-
ride,16 2,4-dimethylquinoline-3-carboxylic acid 19a,17
FEP.18 All other reagents were commercially available
and, unless otherwise stated, were used without purifica-
tion. Flash chromatography was performed with silica
gel 60 (70–230 mesh from Merck) and monitored by thin
layer chromatography (TLC) with silica plates (Merck,
Kieselgel 60 F254).
4.2.3. (S)-2-[N-(1-Phenylethyl)-N-isopropyl]amino carbon-
ylacetic acid 3.
A
solution of KOH (930mg,
16.6mmol) in ethanol was slowly added to a solution
of 2 (919mg, 3.32mmol) in ethanol (10mL) at 0ꢁC.
The resulting solution was stirred for 5h and ethanol
was removed in vacuo. The residue was treated with
water (10mL) and 1N aqueous HCl (15mL). The result-
ing aqueous phase was extracted with CH2Cl2
(3 · 15mL). Drying (MgSO4) and evaporation of
CH2Cl2 under vacuum afforded acid 3 in 96% yield.
The product consisted of a mixture of two isomꢀer1s
owing to the presence of the amide function IR(cm
,
KBr): 3446, 3061, 3028, 2971, 2934, 2876, 1737, 1640.
1H NMR(300 MHz, CDCl 3) d 1.09 (2.25H, d,
J = 6.7Hz), 1.19 (0.75H, d, J = 6.9Hz), 1.40 (3H, J =
6.8Hz), 1.69 (2.25H, d, J = 6.8Hz), 1.71 (0.75H, d,
J = 6.7Hz), 3.30 (0.75H, d, J = 19.3Hz), 3.45 (0.5H, s),
3.61 (0.75H, d, J = 19.3Hz), 3.65 (0.75H, m), 3.91
(0.25H, m), 5.00 (0.75H, q, J = 6.9Hz), 5.24 (0.25H,
m), 7.30–7.40 (5H, m). HRMS (IE): calcd for
C14H19NO3 249.1365; found: 249.1370.
¨
4.2. Synthesis of the chiral quinolinium salt 8: Friedlander
approach
4.2.1. (S)-N-(1-Phenylethyl)-1-methylethylamine 1.
A
solution of (S)-a-methylbenzylamine in acetone
(400mL) was refluxed 12h. After evaporation of the sol-
vent under vacuum, the residue was dissolved in metha-
nol (300mL) and sodium tetrahydroborate (1.5g,
39.7mmol) was slowly added at 0ꢁC. The solution was
stirred 2h at 20ꢁC. After evaporation of the solvent in
vacuo, water (30mL) was added and the resulting aque-
ous phase was extracted with CH2Cl2 (3 · 30mL). The
combined organic layers were dried (MgSO4) and con-
centrated in vacuo. The desired amine 1 was purified
by distillation under reduced pressure to afford 1 in a
90% yield. Bp (20mm Hg): 104ꢁC. IR(cm ꢀ1, KBr):
4.2.4. (1S)-1,4-Dihydro-2-isopropyl-1-methyl-2-benzaze-
pin-3,5-dione 4. Oxalyl chloride (1.8mL, 11.9mmol)
was slowly added to a solution of acid 3 (2.69g,
10.8mmol) and DMF (100lL) in CH2Cl2 (10mL). After
1h stirring at 20ꢁC, the organic solvent and oxalyl chlo-
ride in excess were removed in vacuo. The reddish liquid
residue was dissolved in CH2Cl2 (15mL) and aluminium
chloride (3.33g, 27.4mmol) was added by small portions
at 0ꢁC. The reaction mixture was stirred 1h at room
temperature and water (20mL) was slowly added at
0ꢁC. The aqueous layer was separated and extracted
with CH2Cl2 (3 · 15mL). The organic layers were col-
lected, dried (MgSO4) and concentrated in vacuo. The
residue was purified by flash chromatography (SiO2;
cyclohexane/EtOAc: 50:50) to afford 4 as a white solid
in 38% yield. Mp: 148ꢁC (cyclohexane/EtOAc). 1H
NMR(300MHz, CDCl 3) d 0.99 (3H, d, J = 6.9Hz),
1.25 (3H, d, J = 6.9Hz), 1.80 (3H, J = 7.5Hz), 3.78
(1H, d, J = 15.5Hz), 4.48 (1H, d, J = 15.5Hz), 4.67
(1H, q, J = 7.5Hz), 5.00 (1H, m), 7.25 (1H, d,
J = 7.54Hz), 7.36 (1H, t, J = 7.54Hz), 7.51 (1H, t,
J = 7.5Hz), 8.03 (1H, d, J = 7.5Hz). 13C NMR
(75MHz, CDCl3) d 20.4, 20.6, 24.3, 46.7, 53.2, 54.6,
128.6, 129.1, 130.8, 134.1, 144.9, 165.8, 191.3. Anal.
Calcd for C14H17NO2: C, 72.70; H, 7.41; N, 6.06.
Found: C, 72.75; H, 7.52; N, 5.97.
1
3315, 2962, 1451, 1169, 761, 701. H NMR(300MHz,
CDCl3) d 1.02 (6H, t, J = 6.5Hz), 1.35 (3H, d,
J = 6.6Hz), 2.63 (1H, sept, J = 6.2Hz), 3.90 (1H, q,
J = 6.6Hz), 7.23–7.38 (5H, m). 13C NMR(75MHz,
CDCl3) d 20.6, 23.9, 24.7, 45.3, 54.9, 126.6, 126.5,
128.2, 145.9.
4.2.2. Ethyl 1-[N-(1(S)-phenylethyl)-N-isopropyl]-amino-
carbonylacetate 2. A solution of amine 1 (1.08g,
6.64mmol) and triethylamine (0.8g, 7.37mmol) in
CH2Cl2 (10mL) was added dropwise to a solution of
ethyl malonyl chloride (1.11g, 7.37mmol) in CH2Cl2
(10mL). The mixture was stirred 12h at 20ꢁC followed
by addition of water (10mL). After phase separation,
the organic layer was washed with a 10% aqueous solu-
tion of Na2CO3, dried (MgSO4) and concentrated in
vacuo. The residue was purified by flash chromatography
(SiO2; cyclohexane/EtOAc: 90:10) to afford 2 as a yellow
oil in 79% yield. The product consisted of a mixture of
two isomers owing to the presence of the amide func-
tion. IR(cm ꢀ1, KBr): 3454, 3060, 2977, 2937, 2360,
1736, 1647. 1H NMR(300MHz, CDCl 3) d 1.08
(1.8H, d, J = 6.7Hz), 1.12 (1.2H, d, J = 6.7Hz), 1.29
(6H, m), 1.65 (1.8H, d, J = 6.9Hz), 1.71 (1.2H, d,
J = 6.9Hz), 3.48 (2.6H, m), 3.83 (0.4H, m), 4.21 (2H,
q, J = 7.1Hz), 4.10 (0.6H, q, J = 6.9Hz), 5.04 (0.4H,
4.2.5. N-(2-Nitrobenzylidene)-p-toluidine 5. A solution
of 2-nitrobenzaldehyde (10.0g, 66mmol) and p-toluidine
(7.10g, 66mmol) in ethanol (250mL) was refluxed for
2h. After cooling to 0ꢁC, the orange precipitate was
filtered and washed with cooled ethanol. After dry-
ing, the desired compound 5 was obtained as an
orange solid in 42% yield. Mp: 70–71ꢁC (ethanol).
m), 7.25–7.35 (5H, m). 13C NMR(75MHz, CDCl ) d
IR(cm
, KBr): 3077, 3012, 2920, 1890, 1605.
ꢀ1
3

S. Leleu et al. / Tetrahedron: Asymmetry 15 (2004) 3919–3928
3925
1H NMR(200MHz, CDCl 3) d 2.41 (s, 3H), 7.24 (d, 4H,
J = 7.5Hz), 7.61 (t, 1H, J = 7.5Hz), 7.74 (t, 1H, J =
7.5Hz), 8.07 (d, 1H, J = 7.9Hz), 8.32 (d, 1H, J =
7.9Hz), 8.96 (s, 1H). 13C NMR(50MHz, CDCl 3) d
21.5, 121.6, 124.9, 130.0, 130.2, 131.3, 131.6, 133.8,
137.8, 148.8, 149.7, 155.2. Anal. Calcd for
C14H12N2O2: C, 69.99; H, 5.03; N, 11.66. Found: C,
69.95; H, 4.93; N, 11.86.
(3H, s), 4.73 (1H, q, J = 7.5Hz), 5.06 (1H, sept,
J = 6.8Hz), 7.48 (1H, d, J = 7.5Hz), 7.63 (2H, m),
7.76 (1H, d, J = 6.4Hz), 7.96 (1H, t, J = 7.5Hz),
8.23 (1H, d, J = 8.3Hz), 8.30 (1H, t, J = 7.5Hz),
8.31(1H, d, J = 9.0Hz), 9.36 (1H, s). 13C NMR
(75MHz, CDCl3) d 20.2, 20.6, 21.4, 45.8, 48.8, 52.6,
120.6, 127.4, 128.3, 129.3, 129.4, 131.1, 131.5, 132.7,
134.2, 134.4, 138.8, 141.6, 145.4, 149.1, 156.1, 162.5.
HRMS (FAB+): calcd for C22H23N2O: 331.1810; found:
331.1829.
4.2.6. 2-(p-Tolylaminomethylidene)-aniline 6. Com-
pound 5 (8.26g, 22mmol) was dissolved in ethanol
(200mL) the solution was heated to reflux and
sodium sulfur monohydrate was added. The reflux was
maintained 10min after which the solution was
cooled to 0ꢁC and the suspension was stirred for an
additional 4h at this temperature. During this time, a
yellow precipitate was formed. This precipitate was
filtered, washed with water and dried to give 6 in
4.3. Preparation of the chiral quinoline 7: intramolecular
coupling approach
4.3.1. 2-Chloroquinoline-3-carboxylic acid 9. To a solu-
tion of diisopropylamine (5.3mL, 38mmol) in dry THF
(75mL) cooled at ꢀ78ꢁC was added, under nitrogen, a
solution of n-butyllithium in hexane (16.5mL, 2.5M,
41mmol). After being stirred 20min at ꢀ78ꢁC, a solu-
tion of 2-chloroquinoline (5g, 30mmol) in dry THF
(30mL) was slowly added. The reaction mixture was
stirred for an additional 20min and then poured into a
large excess of dry ice. The solution was concentrated
under vacuum before addition of water (25mL). The
aqueous solution was washed with Et2O (3 · 10mL)
and acidified (pH = 4–5) with 1N aqueous HCl. The
precipitate formed was collected by filtration, washed
with water (10mL) and dried. The crude product was
purified by recrystallization in ethanol to provide 9ꢀa1s
80% yield. Mp: 101ꢁC. IR(cm ꢀ1, KBr): 3447, 3257,
1
3018, 2890, 1624. H NMR(200MHz, CDCl ) d 2.4
3
(3H, s), 6.76 (2H, d, J = 7.5Hz), 7.12–7.37 (6H, m),
8.56 (1H, s). 13C NMR(50MHz, CDCl 3) d 21.3,
116.1, 116.6, 118.1, 121.1, 130.1, 132.0, 134.6, 135.7,
149.1, 149.7, 162.8. Anal. Calcd for C14H14N2: C,
79.97; H, 6.71; N, 13.32. Found: C, 79.92; H, 6.86; N,
13.04.
4.2.7. (S)-1H-2-Isopropyl-1-methyl-2-benzazepino-[5,4-b]-
quinolein-3-one 7. A solution of 2-(p-tolylaminoethyl-
idene)-aniline 6 (1.5g, 7.1mmol), a few drops of
triethylamine and 2-benzazepin-3,5-dione 4 (1.36g,
7.1mmol) in ethanol (200mL) was refluxed 24h, after
which the solution was cooled and concentrated in
vacuo. The residue was chromatographed on silica gel
a white solid in 65% yield. Mp: 220ꢁC. IR(cm
,
KBr): 3418, 3072, 2897, 2534, 2352, 1731. 1H NMR
(DMSO-d6, 300MHz) d 7.55 (1H, t, J = 7.5Hz), 7.75
(1H, t, J = 7.5Hz), 7.83 (1H, d, J = 8.3Hz), 8.01 (1H,
d, J = 8.3Hz,), 8.75 (1H, s). 13C NMR(75MHz, CDCl 3)
d 126.1, 128.0, 128.2, 128.4, 129.3, 132.9, 138.1, 140.4,
148.2, 166.6. HRMS (FAB+): calcd for C10H6ClNO2:
207.0087; found: 207.0082.
(cyclohexane/EtOAc: 90:10) to afford quinoline 7 as a
25
yellow oil in 80% yield. ½aꢁ = +40 (c 0.13, CH2Cl2).
D
IR(cm ꢀ1, KBr): 1615, 1493, 1235, 1201. 1H NMR
(300 MHz, CDCl3) d 1.15 (3H, d, J = 7.15Hz), 1.20
(3H, d, J = 7.16Hz), 1.31 (3H, d, J = 6.8Hz), 4.55
(1H, q, J = 7.2Hz), 5.20 (1H, sept, J = 6.8Hz), 7.20
(1H, d, J = 7.2Hz), 7.35 (1H, t, J = 7.2Hz), 7.45 (1H,
t, J = 7.5Hz), 7.55 (1H, t, J = 7.9Hz), 7.75 (1H, t,
J = 7.2Hz), 7.92 (1H, d, J = 8.3Hz), 8.10 (1H, d,
J = 8.6Hz), 8.25 (1H, d, J = 7.2Hz), 8.84 (1H, s). 13C
4.3.2.
(S)-N-[1-(2-Iodophenyl)-ethyl]isopropylamine
10. To a solution of TMEDA (3.61mL, 43mmol) in
dry THF (75mL) was added at ꢀ78ꢁC n-butyllithium
in hexane (17mL, 2,5M, 43mmol). The solution was
stirred for 20min before adding (S)-N-isopropyl-a-
methylbenzylamine 1 (3.6g, 22mmol). The reaction mix-
ture was stirred at 20ꢁC for a further 2h and then cooled
to ꢀ78ꢁC prior to adding stepwise a solution of iodine
(11g, 43mmol) in THF (75mL). The reaction mixture
was stirred at 20ꢁC for 12h. The reaction was treated
with a solution of sodium disulfite (10g) in water
(50mL). After evaporation of the organic solvent under
vacuum, the aqueous solution was extracted with
dichloromethane (3 · 50mL). The combined organic
layers were dried (MgSO4) and CH2Cl2 evaporated in
vacuo. Yield: 50% of 10 as a colourless oil after
flash chromatography (SiO2; cyclohexane/ethyl acetate:
NMR(75MHz, CDCl ) d 20.6, 21.2, 21.8, 47.3, 53.0,
3
127.1, 127.4, 127.5, 128.9, 129.0, 129.8, 130.1, 130.7,
131.6, 131.8, 137.6, 140.7, 143.5, 148.9, 153.6, 166.5;
HRMS (IE): calcd for C21H20N2O: 316.1575; found:
316.1582.
4.2.8.
(1S)-1H-2-Isopropyl-1,9-dimethyl-3-oxo-2-benz-
azepino[5,4-b]quinolinium triflate 8. Methyl trifluoro-
methanesulfonate (0.38mL, 3.3mmol) and anhydrous
Na2CO3 were added to a solution of quinoline 7
(870mg, 2.8mmol) in CH2Cl2 (15mL) under nitrogen
atmosphere. After being stirred at 20ꢁC for 2h, the
solution was filtrated to remove Na2CO3. The filtrate
was concentrated in vacuo to afford the desired quino-
25
D
ꢀ1
65:35) ½aꢁ = ꢀ36.5 (c 0.0043, CH2Cl2). IR(cm
,
KBr): 3325, 3057, 2962, 2925, 2865. 1H NMR
(300MHz, CDCl3) d 0.93 (3H, d, J = 6.4Hz), 0.97
(3H, d, J = 6.0Hz), 1.20 (3H, d, J = 6.8 Hz), 2.50 (1H,
sept, J = 6.4Hz), 4.10 (1H, q, J = 6.8Hz), 6.85 (1H, t,
J = 7.0Hz), 7.28 (1H, d, J = 7.0Hz), 7.32 (1H, t,
J = 7.9Hz), 7.73 (1H, d, J = 7.9Hz). 13C NMR
(75MHz, CDCl3) d 22.5, 23.9, 24.2, 46.4, 59.2, 100.0,
linium 8 as a white solid in 100% yield. Mp: 136ꢁC
25
(CH2Cl2/Et2O). ½aꢁ = ꢀ32 (c 0.45, CH2Cl2). IR
D
1
(cmꢀ1, KBr): 3505, 3069, 2983, 2937, 1630, 1588. H
NMR(300MHz, CDCl 3) d 1.12 (3H, d, J = 6.8Hz),
1.26 (3H, d, J = 7.5Hz), 1.27 (3H, d, J = 6.8Hz), 4.69

3926
S. Leleu et al. / Tetrahedron: Asymmetry 15 (2004) 3919–3928
127.4, 129.0, 139.9, 147.2. HRMS (IE): calcd for C11H16
IN 289.0327; found: 289.0320.
several times with nitrogen, dry THF (50mL) was added
and the solution became immediately reddish brown.
A solution of quinoline 12 (2.1g, 4.2mmol) in dry
THF (50mL) was added dropwise. The resulting
solution was stirred at 50ꢁC for 2h under nitrogen
atmosphere and then poured into 2M aqueous ammo-
niac (50mL). The aqueous solution was extracted
with Et2O (3 · 50mL). The combined organic layers
were dried (MgSO4) and concentrated in vacuo. The
residue was chromatographed on silica gel (cyclohex-
4.3.3. ( )-N-[1-(2-Iodophenyl)-ethyl]-N-isopropyl acet-
amide 11. To a solution of ( )-10 (290mg, 1mmol)
and NEt3 (210lL, 2mmol) in CH2Cl2 (15mL) was
added dropwise acetyl chloride (143lL, 2mmol). The
solution was stirred at 20ꢁC for 2h and then washed
with water (3 · 15mL). The combined organic phases
were dried (MgSO4), evaporated and subjected to flash
chromatography (SiO2; cyclohexane/ethyl acetate:
80:20) to give 11 in 50% yield. The product consisted
of a mixture of two isomers owing to the presence of
ane/ethyl acetate: 90:10) to afford 7 as a yellow oil
25
in 64% yield. ½aꢁ = +40 (c 0.013, CH2Cl2). IR
D
(cmꢀ1, KBr): 1615, 1493, 1235, 1201. 1H NMR
(300MHz, CDCl3) d 1.15 (3H, d, J = 7.2Hz), 1.20
(3H, d, J = 7.2Hz), 1.31 (3H, d, J = 6.8Hz), 4.55 (1H,
q, J = 7.2Hz), 5.20 (1H, sept, J = 6.8Hz), 7.20 (1H, d,
J = 7.2Hz), 7.35 (1H, t, J = 7.2Hz), 7.45 (1H, t, J =
7.5Hz), 7.55 (1H, t, J = 7.9Hz), 7.75 (1H, t,
J = 7.15Hz), 7.92 (1H, d, J = 8.3Hz), 8.10 (1H, d,
J = 8.7Hz), 8.25 (1H, d, J = 7.15Hz), 8.84 (1H, s). 13C
1
the amide function. H NMR(300MHz, MeOD- d4) d
0.77 (3H, d, J = 6.5Hz), 1.25 (3H, d, J = 6.5Hz), 1.52
(3H, d, J = 6.5Hz), 2.11 (0.5H, s), 2.32 (1.5H, s), 3.06
(0.7H, m), 3.71 (0.3H, br s), 4.84 (0.65H, q,
J = 7.0Hz), 5.07 (0.35H, br s), 6.90 (0.3H, m), 6.99
(0.66H, t, J = 7.0Hz), 7.28–7.49 (2H, m), 7.78 (0.34H,
d, J = 6.5Hz), 7.85 (0.66H, d, J = 7.7Hz). 13C NMR
(75MHz, CDCl3) d 18.8, 19.9, 21.6, 24.9, 48.1, 61.1,
96.3, 128.7, 129.5, 130.0, 138.4, 140.8, 171.4. Anal Calcd
for C13H18INO: C, 47.14; H, 5.48; N, 4.23. Found: C,
47.07; H, 5.38; N, 4.13. Both enantiomers of 11 were
separated by HPLC (chiracel OJ, 1mL/min, heptane/iso-
propanol: 90:10); retention time = 6.03 and 7.36min.
NMR(75MHz, CDCl ) d 20.6, 21.2, 21.8, 47.3, 53.0,
3
127.1, 127.4, 127.5, 128.9, 129.0, 129.8, 130.1, 130.7,
131.6, 131.8, 137.6, 140.7, 143.5, 148.9, 153.6, 166.5.
HRMS (IE): calcd for C21H20N2O: 316.1575; found:
316.1582.
4.3.6. Ethyl 2-phenylquinoline-3-carboxylate 13.
A
solution of 2-(p-tolylaminoethylidene)-aniline 6 (1.5g,
7.1mmol), ethyl benzoyl acetate (1.36g, 7.1mmol) and
a few drops of triethylamine in ethanol (200mL) was
refluxed 24h. After cooling to room temperature, EtOH
was evaporated under vacuum and the resulting residue
was chromatographed on silica gel (cyclohexane/EtOAc:
90:10) affording 13 as a yellow oil in 72% yield.
4.3.4. N-[(1S)-1-(2-Iodophenyl)ethyl]-N-isopropyl-2-chlo-
roquinoline-3-carboxamide 12. To a solution of 2-chlo-
roquinoline-3-carboxylic acid (716mg, 3.46mmol) and
DMF (50lL) in CH2Cl2 (10mL) was added oxalyl chlo-
ride (606lL, 6.92mmol). The reaction mixture was stir-
red at room temperature for 2h. The organic solvent
and oxalyl chloride in excess were evaporated in vacuo.
The residue was dissolved in CH2Cl2 (20mL) and a solu-
tion of amine 10 (1g, 3.5mmol) and triethylamine
(0.484mL) in CH2Cl2 (10mL) was added dropwise.
The reaction mixture was stirred at 20ꢁC overnight.
The solution was washed with water (3 · 10mL). The
organic layer was dried (MgSO4) and concentrated
under vacuum. The residue was chromatographed on
IR(cm ꢀ1, KBr): 3423, 3059, 2981, 2936, 2902, 1721,
1
1619, 1594, 1555. H NMR(300MHz, CDCl ) d 1.10
3
(3H, d, J = 7.15Hz), 4.10 (2H, q, J = 7.15Hz), 7.4–7.3
(3H, m), 7.41 (1H, t, J = 7.9Hz), 7.53 (2H, m), 7.65
(1H, t, J = 7.15Hz), 7.75 (1H, d, J = 8.3Hz), 8.06 (1H,
d, J = 8.3Hz), 8.50 (1H, s). 13C NMR(75MHz, CDCl 3)
d 14.1, 61.9, 125.9, 126.2, 127.6, 128.5, 128.6, 128.9,
129.0, 129.9, 131.9, 139.4, 141.2, 148.7, 158.5, 168.4.
Anal Calcd for C18H15NO2: C, 77.96; H, 5.45; N, 5.05.
Found: C, 78.13; H, 5.44; N, 5.23.
silica gel (cyclohexane/ethyl acetate: 85:15) to afford 12
25
in 60% yield. Mp: 94ꢁC (cyclohexane). ½aꢁ = +7.6
D
(c 0.25, CH2Cl2). IR(cm ꢀ1, KBr): 3059, 2977, 2926,
2849, 1633. 1H NMR(300MHz, CDCl 3) d 0.90
(3H, d, J = 6.78Hz), 1.04 (3H, d, J = 6.8Hz), 1.88
(3H,d, J = 7.2Hz), 3.67 (1H, sept, J = 6.8Hz), 4.95
(1H, q, J = 7.2Hz), 6.91 (1H, m), 7.34 (1H, s), 7.55
(1H, m), 7.72 (1H, m), 7.77–7.81 (2H, m), 7.98–8.01
4.3.7. Ethyl 2-phenylquinolinium-3-carboxylate triflate
14. To a solution of quinoline 13 (332mg, 1.2mmol)
in dry CH2Cl2 was added methyl triflate (164mg,
1.7mmol). The resulting solution was stirred at room
temperature for 2h. Addition of Et2O (10mL) furnished
(2H, m), 8.06 (1H, s). 13C NMR(75MHz, CDCl ) d
3
20.1, 21.2, 21.5, 52.5, 59.2, 99.3, 128.9, 128.1, 128.2,
128.9, 129.0, 129.3, 129.6, 131.6, 132.6, 135.1, 139.9,
145.5, 147.0, 147.7, 167.7. Anal Calcd for C21H20ClI-
N2O: C, 52.68; H, 4.21; N, 5.85. Found: C, 52.43; H,
4.52; N, 5.62.
a white precipitate which was filtered to give the desired
ꢀ1
quinolinium salt 14 in a quantitative yield. IR(cm
,
KBr): 3065, 2991, 2364, 1720, 1624, 1591. 1H NMR
(300MHz, CDCl3) d 1.10 (3H, d, J = 7.15Hz), 4.10
(2H, q, J = 7.15Hz), 4.42 (3H, s), 7.63 (5H, m), 8.04
(1H, t, J = 7.5Hz), 8.28 (2H, t, J = 7.5Hz), 8.35 (1H,
d, J = 2.3Hz), 8.55 (1H, d, J = 9.05Hz), 9.41 (1H, s).
4.3.5. (S)-1H-2-Isopropyl-1-methyl-2-benzazepino [5,4-b]-
quinolein-3-one 7. A 250-mL, round-bottomed, three-
necked flask was charged with NiBr2(PPh3)2 (3.12g,
4.2mmol), Et4N+Iꢀ (1.86g, 4.2mmol) and zinc
powder (2.75g, 42mmol) activated by acetic acid prior
to use. After the material was evacuated and filled
13C NMR(75MHz, CDCl ) d 13.9, 43.5, 63.3, 120.6,
3
128.1, 128.5, 128.7, 129.6, 131.4, 131.7, 131.8, 132.2,
138.8, 141.0, 148.4, 159.4, 136.4. Anal Calcd for
C20H18F3NO5S: C, 54.42; H, 4.11; N, 3.17. Found: C,
54.22; H, 4.25; N, 3.12.

S. Leleu et al. / Tetrahedron: Asymmetry 15 (2004) 3919–3928
3927
4.4. Preparation of the ‘masked amides’ 16a,b and 18a,b
4.4.1. General procedure
4.5. Atropenantioselective amidification of 2,4-dimethyl-
quinoline-3-carboxylic acid 19a
4.5.1. 4-Nitrophenyl 2,4-dimethylquinoline-3-carboxylate
19b. To a solution of 2,4-dimethylquinoline-3-carb-
oxylic acid 19a (0.201g, 1mmol) in CH2Cl2 (10mL)
was added oxalyl chloride (0.18mL, 1.19mmol) and
DMF (25lL). The resulting solution was stirred at
20ꢁC for 1h. The solvent was evaporated under vacuum
and the residue dissolved in CH2Cl2 (10mL). A solution
a 4-nitrophenol in CH2Cl2 (10mL) was added at 0ꢁC
and the solution was allowed to stir for an additional
12h at 20ꢁC. After adding water (20mL) and separation
phase, the aqueous layer was extracted twice with
CH2Cl2 (2 · 10mL). The combined organic layers were
dried (MgSO4), filtered and evaporated under vacuum
to afford ester 19b in 84%, as a yellow oil, after flash
chromatography (SiO2; cyclohexane/ethyl acetate:
4.4.1.1. Ethyl 1-methyl-2-phenyl-4-benzylamino-1,4-
dihydroquinoline-3-carboxylate 16a. To a solution of
quinolinium salt 14 (44mg, 0.104mmol) and Na2CO3
(11.0mg, 0.104mmol) in CH2Cl2 (5mL) was added
benzylamine (12lL, 0.114mmol) at room temperature.
The resulting biphasic solution was then stirred for
1.5h at 20ꢁC after which 5mL of water was added.
The biphasic solution was stirred for an additional
15min. The organic layer was collected, dried (MgSO4)
and concentrated in vacuo at room temperature.
The adduct product 16a was obtained in a quantitative
yield. 1H NMR(300MHz, CDCl 3) d 1.12 (3H, t,
J = 7.0Hz), 3.37 (2H, q, J = 7.0Hz), 3.75 (5H, br s),
4.75 (1H, s), 7.12–7.30 (14H, m). 13C NMR(75MHz,
CDCl3) d 14.0, 15.5, 46.5, 66.0, 123.0, 127.5, 127.7,
128.5, 128.7, 128.9, 129.3, 131.0, 134.0, 142.5, 157.0,
169.5.
90:10) IR(cm ꢀ1, KBr): 3111, 3071, 2930, 1926, 1730,
1
1615, 1581, 1568, 1524. H NMR(300MHz, CDCl ) d
3
2.75 (3H, s), 2.79 (3H, s), 7.43 (2H, d, J = 8.5Hz), 7.54
(1H, t, J = 7.7Hz), 7.72 (1H, t, J = 7.7Hz), 7.99 (2H,
d, J = 8.3Hz), 8.31 (1H, d, J = 8.5Hz). 13C NMR
(75MHz, CDCl3) d 16.5, 24.6, 122.7, 124.5, 125.9,
126.4, 127.2, 129.8, 131.2, 143.1, 146.1, 149.9, 154.3,
155.4, 158.9, 167.0. Anal. Calcd for C18H16N2O4: C,
67.07; H, 4.38; N, 8.69. Found: C, 67.15; H, 4.50; N,
8.71.
4.4.1.2. Ethyl 1-methyl-2-phenyl-4-(piperidin-1-yl)-1,4-
dihydroquinoline-3-carboxylate 16b. According to the
general procedure from 14 (44mg, 0.104mmol) and
piperidine (10lL, 0.117mmol) affording 16b in a
quantitative yield. 1H NMR(300MHz, CDCl 3) d
0.70 (3H, d, J = 7.0Hz), 1.25 (2H, br s), 1.40 (2H,
br s), 2.20 (2H, br s), 2.41 (2H, br s), 2.99 (3H,
s), 3.75 (2H, q, J = 7Hz), 4.90 (1H, s), 6.95–7.40 (9H,
4.5.2. rac-3-(Piperidinylcarbonyl)-2,4-dimethylquinoline
20. A solution of 2,4-dimethylquinoline-3-carboxylic
acid 19a (0.201g, 1.0mmol), 1-ethyl-3-(3-dimethylami-
nopropyl)-carbodiimide (0.191g, 1.0mmol), 1-hydroxy-
1H-benzotriazol (0.135g, 1.0mmol) and diisopropyleth-
ylamine (0.174mL, 1.0mmol) in dichloromethane
(10mL) was stirred for 15min. Piperidine (0.99mL,
1.0mmol) was added and the resulting solution was
stirred for a further 2h at 20ꢁC. The solvent was evapo-
rated under vacuum and the residue was chromato-
graphed on silica gel (cyclohexane/ethyl acetate: 80:20).
m). 13C NMR(75MHz, CDCl
) d 14.0, 25.0,
3
26.7, 36.8, 49.0, 59.5, 60.3, 98.0, 113.7, 123.4, 123.5,
127.7, 128.5, 128.9, 130.5, 130.8, 137.7, 141.7, 154.4,
168.5.
4.4.1.3. (1S,4R)-4-Benzylamino-4,9-dihydro-1,9-dime-
thylbenzazepino[5,4-b]quinolin-3-one 18a. According to
the general procedure from 8 (50mg, 0.104mmol) and
benzylamine (12lL, 0.114mmol) affording 18a in a
1
Yield = 80%. IR(cm ꢀ1, KBr): 3436, 3066, 2925, 2854,
quantitative yield. H NMR(300MHz, CDCl ) d 1.00
3
1
(3H, d, J = 7.0Hz), 1.24 (3H, d, J = 7.0Hz), 1.55 (1H,
s), 1.76 (3H, d, J = 7.15Hz), 3.17 (3H, s), 3.69 (2H, s),
4.56 (1H, q, J = 7.15Hz), 5.13 (1H, m), 5.35 (1H, s),
1628, 1588, 1566. H NMR(300MHz, CDCl ) d 1.38–
3
1.44 (2H, m), 1.61–1.65 (6H, m), 2.55 (3H, s), 2.60
(3H, s), 3.65–3.85 (2H, m), 7.48 (1H, t, J = 7.15 Hz),
7.63 (1H, t, J = 8.3Hz), 7.90 (1H, d, J = 7.5Hz), 7.94
(1H, d, J = 8.3Hz). 13C NMR(75MHz, CDCl 3) d
14.5, 22.4, 23.4, 24.7, 25.5, 41.2, 46.3, 122.7, 125.0,
125.2, 128.2, 128.6, 129.0, 138.8, 145.9, 153.4, 166.9.
Anal. Calcd for C17H20N2O: C, 76.09; H, 7.51; N,
10.44. Found: C, 75.95; H, 7.44; N, 10.28. Both enantio-
mers of 20 were separated by HPLC (chiracel OD,
1mL/min, heptane/isopropanol: 90:10); retention
time = 7.43 and 9.06min.
7.0–7.3 (13H, m). 13C NMR(75MHz, CDCl ) d 20.1,
3
20.8, 21.1, 39.2, 46.4, 51.0, 53.1, 55.4, 90.4, 114.4,
115.6, 122.3, 126.7, 127.1, 127.4, 127.8, 127.9, 128.4,
128.6, 128.9, 129.7, 130.7, 131.8, 141.4, 142.6, 142.9,
143.0, 168.0.
4.4.1.4.
(1S,4R)-4,9-Dihydro-1,9-dimethyl-4-(piper-
idin-1-yl)benzazepino[5,4-b]quinolin-3-one 18b. Accord-
ing to the general procedure from 8 (50mg, 0.104mmol)
and piperidine (10lL, 0.107mmol) affording 18b in a
quantitative yield. 1H NMR(300MHz, CDCl 3) d
0.98 (3H, d, J = 6.8Hz), 1.22 (3H, d, J = 6.8Hz), 1.35
(6H, s large), 1.71 (3H, d, J = 7.15Hz), 2.15 (2H, br s),
2.38 (2H, br s), 3.12 (3H, s), 4.53 (1H, q, J = 7.15Hz),
5.11 (1H, m), 5.37 (1H, s), 7.0–7.3 (8H, m). 13C NMR
(75MHz, CDCl3) d 19.7, 20.8, 21.1, 24.9, 26.8, 39.2,
46.5, 50.3, 52.8, 61.9, 90.4, 113.9, 115.4, 121.7, 122.5,
127.4, 127.9, 129.6, 130.3, 130.7, 131.8, 143.1, 143.5,
144.7, 167.9.
4.5.3. 3-(Piperidinylcarbonyl)-2,4-dimethylquinoline 20
(EDCI/HOBt activation). A solution of 2,4-dimethyl-
quinoline-3-carboxylic acid 19a (100 mg, 0.5 mmol), 1-
ethyl-3-(3-dimethylaminopropyl)-carbodiimide (95 mg,
0.5 mmol), 1-hydroxy-1H-benzotriazol (70 mg, 0.5
mmol) and diisopropylethylamine (87 lL, 0.5 mmol)
in dichloromethane (3 mL) was stirred for 15 min. A
solution of 18b (207.8 mg, 0.5 mmol) in CH2Cl2 was
then added at 20 ꢁC and the resulting solution was

3928
S. Leleu et al. / Tetrahedron: Asymmetry 15 (2004) 3919–3928
1990, 55, 647; (c) Gundel, W. H. Liebigs Ann. Chem. 1980,
¨
1350; (d) Bunting, J. W.; Bolton, J. Tetrahedron 1986, 42,
1007.
stirred for 12 h. After evaporation of the solvent, the
residue was purified by flash chromatography (SiO2;
cyclohexane/ethyl acetate: 80/20) affording 20 in 80%
yield. The enantiomeric excess was determined by chiral
HPLC in the same conditions than those reported in
Section 4.5.2.
5. (a) Moracci, F. M.; Rienzo, B. D.; Tortorella, S.;
Liberatore, F. Tetrahedron 1980, 36, 785; (b) Angelino,
S. A. G. F.; Van Veldhuizen, A.; Buurman, D. J.; Van Der
Plas, H. C. Tetrahedron 1984, 40, 433; (c) Okada, E.;
Sakaemura, T.; Shimomura, N. Chem. Lett. 2000, 50; (d)
Zoltewicz, J. A.; Helmick, L. S.; OꢁHalloran, J. K. J. Org.
Chem. 1976, 41, 1303; (e) Van Der Plas, H. C.; Buurman,
D. J. Tetrahedron Lett. 1984, 25, 3763; (f) Angelino, S. A.
G. F.; Van Valkengoed, B. H.; Buurman, D. J.; Van Der
4.5.4. 3-(Piperidinylcarbonyl)-2,4-dimethylquinoline 20
(FEP activation). To a solution of 18b (41.5mg,
0,1mmol) in CH2Cl2 (1mL) was slowly added a solution
of 2,4-dimethylquinoline-3-carboxylic acid 19a (20.0mg,
0.1mmol), N-ethylpyridinium trifluoroborate (21mg,
0.1mmol) and diisopropylethylamine (18lL, 0.1mmol)
in dichloromethane (3mL). The mixture was stirred
overnight at ꢀ60ꢁC and then slowly warmed to room
temperature. The solvent was removed in vacuo and
the resulting residue was chromatographed on prepara-
tive TLC plate (SiO2; cyclohexane/ethyl acetate: 65:35).
The enantiomeric excess was determined by HPLC as re-
ported in Section 4.5.2.
Plas, H. C.; Muller, F. J. Heterocycl. Chem. 1984, 21, 107.
¨
6. Mashraqui, S. H.; Kellogg, R. M. J. Am. Chem. Soc. 1983,
105, 7792.
7. Kellogg, R. M. Angew. Chem., Int. Ed. Engl. 1984, 96, 769.
8. (a) Vasse, J.-L.; Levacher, V.; Bourguignon, J.; Dupas, G.
Chem. Commun. 2002, 2256; (b) Vasse, J.-L.; Dupas, G.;
´
Duflos, J.; Queguiner, G.; Bourguignon, J.; Levacher, V.
Tetrahedron Lett. 2001, 42, 3713; (c) Vasse, J.-L.; Dupas,
´
G.; Duflos, J.; Queguiner, G.; Bourguignon, J.; Levacher,
V. Tetrahedron Lett. 2001, 42, 4613.
9. Cheng, C. C.; Yan, S. J. The Friedlander synthesis of
quinolines. Org. React. 1982, 28, 37.
References
10. (a) Borsche, W.; Ried, W. Liebigs Ann. Chem. 1943, 554,
269; (b) Borsche, W.; Barthenheier, J. Liebigs Ann. Chem.
1941, 548, 50.
´
11. Marsais, F.; Godard, A.; Queguiner, G. J. Heterocycl.
Chem. 1989, 26, 1589.
12. (a) Narasimhan, N. S.; Mali, R. S. Synthesis 1983, 957; (b)
Hilmersson, G.; Davidsson, O. Organometallics 1995, 14,
912.
13. (a) Iyoda, M.; Otsuka, H.; Sato, K.; Nisato, N.; Oda, M.
Bull. Chem. Soc. Jpn. 1990, 63, 80; (b) Colon, I.; Kelsey,
D. R. J. Org. Chem. 1986, 51, 2627; (c) Ohno, A.;
Tsutsumi, A.; Yamazaki, N.; Okamura, M.; Mikata, Y.;
Fujii, M. Bull. Chem. Soc. Jpn. 1996, 69, 1679; (d)
Semmelhack, M. F.; Helquist, P.; Jones, L. D.; Keller, L.;
Mendelson, L.; Ryono, L. S.; Gorzynski Smith, J.;
Stauffer, R. D. J. Am. Chem. Soc. 1981, 103, 6460.
14. (a) Thayumanavan, S.; Beak, P.; Curran, D. P. Tetrahe-
dron Lett. 1996, 37, 2899; (b) Clayden, J.; Lai, L. W.
Tetrahedron Lett. 1996, 37, 3163; (c) Clayden, J.; Johnson,
P.; Pink, J. H.; Helliwell, M. J. Org. Chem. 2000, 65, 7033;
(d) Clayden, J. Chem. Commun. 2004, 127.
15. This atropenantioselective approach would lead to the first
enantioselective preparation of the atropisomeric amide
20, a precursor of a highly performant chiral NADH
model: Mikata, Y.; Hayashi, K.; Mizukami, K.; Matsu-
moto, S.; Yano, S.; Yamazaki, N.; Ohno, A. Tetrahedron
Lett. 2000, 41, 1035.
1. (a) Inouye, Y.; Oda, J.; Baba, N. Reductions with Chiral
Dihydropyridine Reagents In Asymmetric Synthesis; Mor-
rison, J. D., Ed.; Academic: New York, London, 1983;
Vol. 2, p 92; (b) Ohno, A.; Ushida, S. Mechanistic Models
of Asymmetric Reductions. In Lecture Notes in Bio-
Organic Chemistry; Baulieu, E., Jaenicke, L., Massey, V.,
Williams, R. J. P., Winnacker, E. L., Zerner, B., Eds.;
Springer: Berlin, 1986; p 1; (c) Burgess, V. A.; Davies, S.
G.; Skerj, R. T. Tetrahedron: Asymmetry 1991, 2, 299; (d)
´
Dupas, G.; Levacher, V.; Bourguignon, J.; Queguiner, G.
Heterocycles 1994, 39, 405; (e) Levacher, V.; Dupas, G.;
´
Queguiner, G.; Bourguignon, J. Trends Heterocycl. Chem.
1995, 4, 293.
2. (a) Ichinose, K.; Kodera, M.; Leeper, J.; Battersby, R. J.
Chem. Soc., Perkin Trans. 1 1999, 879; (b) Fife, W. K.
Heterocycles 1984, 22, 2375; (c) Goto, A.; Kishimoto, H.
J. Chem. Soc., Perkin Trans. 2 1990, 73; (d) Lyle, R. E.;
Gauthier, G. J. Tetrahedron Lett. 1965, 4615; (e) Clerici,
A.; Porta, O. Tetrahedron Lett. 1980, 1675; (f) Lindquist,
R. N.; Cordes, E. H. J. Am. Chem. Soc. 1968, 1269; (g)
Engbersen, J. F. J.; Koudijs, A.; Sleiderink, H. M.;
Franssen, M. C. R. J. Chem. Soc., Perkin Trans. 2 1990,
79; (h) Fife, W. K. J. Org. Chem. 1983, 48, 1375; (i) Foster,
R.; Fyfe, C. A. Tetrahedron 1969, 25, 1489; (j) Anderson,
A. G.; Berkelhammer, G. J. Org. Chem. 1958, 1109; (k)
Damji, S. W.; Fyfe, C. A. J. Org. Chem. 1979, 44, 1757.
3. Piepers, O.; Kellogg, R. M. J. Chem. Soc., Chem.
Commun. 1980, 1147.
16. Box, V. G. S.; Marinovic, N.; Yiannikouros, G. P.
Heterocycles 1991, 32, 245.
17. Ohno, A.; Kashiwagi, M.; Ishihara, Y.; Ushida, S.; Oka,
S. Tetrahedron 1986, 42, 961.
18. Li, P.; Xu, J.-C. Tetrahedron 2000, 56, 8119.
4. (a) Damji, S. W. H.; Fyfe, C. A. J. Org. Chem. 1979, 44,
1761; (b) Speelman, J. C.; Kellogg, R. M. J. Org. Chem.