A short and convenient synthesis of chiral heterocyclic β-enamino esters from halogeno acetylenic esters

Article abstract of DOI:10.1016/S0040-4039(02)00577-4

An efficient synthesis of chiral pyrrolidine and piperidine enamino esters is described in one step from methyl halogeno alcynoate and (S)-1-phenylethylamine or (S)-phenylglycinol.

Full text of DOI:10.1016/S0040-4039(02)00577-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 3471–3474
A short and convenient synthesis of chiral heterocyclic b-enamino
esters from halogeno acetylenic esters
Olivier David, Marie-Claude Fargeau-Bellassoued and Ge´rard Lhommet*
Universite´ P. & M. Curie, Laboratoire de Chimie des He´te´rocycles, associe´ au CNRS (UMR 7611), 4 place Jussieu,
F-75252 Paris Cedex 05, France
Accepted 22 March 2002
Abstract—An efficient synthesis of chiral pyrrolidine and piperidine enamino esters is described in one step from methyl halogeno
alcynoate and (S)-1-phenylethylamine or (S)-phenylglycinol. © 2002 Elsevier Science Ltd. All rights reserved.
Heterocyclic trisubstituted enamino esters have found
broad interest in the synthesis of various natural prod-
ucts such as indolizidine alkaloids,1 ( )-prosopinine,2
( )-lythrancepines II and III,3 (+) and (−)-anatoxin a,4
and ( )-gephyrotoxin.5 Enamino ester intermediates are
next acylated or alkylated, but more often directly
reduced. For our part, in short earlier papers, we
In order to obtain N-substituted pyrroles, Bryson13
reported a relatively simple preparation of pyrrolidine
enamino esters by concomitant Michael addition–alkyl-
ation of primary aliphatic or aromatic amines with
methyl 6-chlorohexynoate 3a, in the presence of sodium
iodide and sodium carbonate. Thereby, we thought that
chiral enamino esters 1a and 1b could be easily pre-
pared by this method directly from the appropriate
methyl halo alcynoates 3 and chiral amine 2 (Scheme 1,
route b).
described
a
versatile access to pyrrolizidine,6
indolizidine7 and quinolizidine6 alkaloids using a reduc-
tion–alkylation sequence on the enamino esters 1a and
1b.8,9 In all cases, the strategy adopted to prepare the
appropriate enamino esters was the elegant Eschen-
moser sulfide contraction procedure.10
Recently, Ma and Zhu described a total synthesis of
(−)-lasubine II14 following this same strategy to build
their appropriate enamino esters from halo alcynoates
and various amines. These results prompt us to publish
in preliminary form our findings about the reactivity of
halogeno acetylenic carboxylates with chiral amines. In
this communication we describe a new synthesis of
pyrrolidine and piperidine enamino esters by reaction
of halogeno acetylenic esters with enantiopure
phenylethylamine and phenylglycinol.
However, difficulties, probably due to phosphorus and
sulfur residues, were encountered by us and others11
during either the catalytic hydrogenation of the trisub-
stituted double bond of the enamino ester, or the
debenzylation step. In order to obtain reproducible
results, several chromatographies and recrystallization
were sometimes necessary, considerably reducing the
yields.
( )n
In relation to our synthetic investigations utilizing the
chiral enamino esters 1a and 1b prepared through the
Eschenmoser reaction12 (Scheme 1, route a), we sought
an alternative process by way of which the phosphorus-
containing reagents could be avoided.
COOMe
( )n
S
N
+
a
COOMe
Br
N
Ph
H
CH3
X
Ph
H
( )n
CH3
NH2
b
+
COOMe
1a: n= 1
1b: n= 2
Ph
H
CH3
3a: n= 1, X= Cl
3b: n= 2, X= Cl
3c: n= 2, X= I
2
Keywords: pyrrolidine; piperidine; enamino ester; acetylenic ester;
chiral amine.
* Corresponding author. E-mail: lhommet@ccr.jussieu.fr
Scheme 1.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00577-4

3472
O. Da6id et al. / Tetrahedron Letters 43 (2002) 3471–3474
Results and discussion
Route A involves the acyclic enamine intermediate 5
firstly formed by Michael addition of the (S)-1-
phenylethylamine to activated acetylenic bond of 3b.
Enamine 5 could then react according to two different
pathways: N-alkylation led to exocyclic enamino ester
1b while C-alkylation provided 4.
We first examined the preparation of the pyrrolidine
enamino ester 1a through this attractive methodology.
In a typical procedure (Scheme 2), an equimolar mix-
ture of (S)-1-phenylethylamine 2 and methyl 6-chloro-
hexynoate 3a,15 in the presence of sodium iodide (1
equiv.) and sodium carbonate (2 equiv.) was heated in
refluxing THF for 3 days. Simple acid–base workup
gave quantitatively 1a16 in almost pure form, as a sole
(E)-isomer; thus, it can be used directly in the following
reduction step.
In route B: halide substitution of alcynoate 3b by chiral
amine could lead to compound 6 which then evolves
only towards enamino ester 1b through intramolecular
Michael addition.
We then reconsidered the conditions described by
Bryson in order to decrease or eliminate the formation
of 4. The various conditions examined were reported in
Table 1.
Encouraged by these results, we then extended our
study to the synthesis of the piperidine enamino ester
1b in order to improve the scope and synthetic utility of
this strategy. However, when (S)-1-phenylethylamine 2
was condensed with methyl 7-chloroheptynoate 3b,17 in
the conditions described above (Scheme 3 and Table 1,
entry 2), the undesired cyclohexene 4 was obtained
along with the expected enamino ester 1b in 55/45 ratio.
The reaction was first performed in different solvents,
without adjunction of NaI. Non-polar solvents (toluene
or xylene) were not suitable because a complex mixture
was produced, or no reaction occurred in these cases
(entries 4 and 5). The use of a polar solvent such as
DMF led to a complete consumption of starting mate-
rials, but 4 was still the major product in the mixture
(entry 6) while in THF or CH3CN the reaction was not
even complete after 15 days under reflux (entries 1 and
7); however CH3CN led to a better result, since the
yield of enamino ester increased up to 70%. In a higher
boiling point solvent such as n-BuCN (entry 9), the
reaction was complete in 6 days but furnished an
equimolar mixture of 1b and 4. It is noteworthy that
This result could be explained considering two different
mechanistic routes.18
NH2
COOMe
Cl
N
NaI, Na2CO3
THF
+
Ph
H
CH3
COOMe
Ph
1a
H
3a
CH3
2
Scheme 2.
MeOOC
C-alkylation
N-alkylation
COOMe
HN
X
HN
Cl
5a: X= Cl
5b: X= I
H
H
Ph
4
Ph
CH3
3b
CH3
COOMe
A
NaI
+
Na2CO3
NH2
B
HN
COOMe
N
Ph
H
H
CH3
Ph
COOMe
CH3
H
Ph
CH3
1b
2
6
Scheme 3.
Table 1. Reaction of methyl 7-chloroheptynoate 3b with (S)-1-phenylethylamine 2a
Entry
Refluxing solvent
NaI (equiv.)
Time (days)
Ratio (%)b 1b/4
1
2
3
4
5
6
7
8
9
THF
THF
THF
Toluene
Xylene
DMF
CH3CN
CH3CN
n-BuCN
None
–
1
3
–
–
–
–
1
–
–
15
7
3
3
4
2
15
6
6
–
50
45
40
50c
55
60
No reaction
Complex mixture
30
70
65
50
0
70
30c
35
50
100
d
10
a All experiments were realized in the presence of Na2CO3 as base, until complete consumption of 3b.
b The ratio 1b/4 was determined by GC and 1H NMR spectrum of crude product.
c Unreacted amine: ꢀ30%.
d Reaction conditions: T=120°C, 30 min.

O. Da6id et al. / Tetrahedron Letters 43 (2002) 3471–3474
3473
COOMe
Na2CO3, NaI
CH3CN
N
NH2
Cl
+
OH
OH
8
Ph
3a
Ph
COOMe
7
COOMe
Na2CO3, TBAI
CH3CN
COOMe
+
N
7
N
I
O
OH
Ph
3c
COOMe
Ph
9
10
Scheme 4.
the sole product 419 was obtained when the reactants
were rapidly heated at 120°C without solvent (entry
10).
This method was successfully applied to (S)-phenylgly-
cinol 7. Thus methyl 6-chlorohexynoate 3a in refluxing
CH3CN provided the enamino ester 8 (E-isomer) in
95% yield.8a,22
Next, we studied the influence of various quantities of
sodium iodide in order to convert, in situ, methyl
7-chloroheptynoate 3b into the iodo derivative 3c. We
attempted to increase the attack of primary amine at
the halo substituted carbon of the iodo alcynoate 3c,
leading to the amino alcynoate 6 and thus favoring
route B. However, entries 1–3 (in THF) and entries 7
and 8 (in CH3CN) showed that the adjunction of 1 to
3 equiv. of NaI reduced the reaction time, but favored
the formation of the cyclohexene 4, which is not in
agreement with this hypothesis.
When methyl 7-iodoheptynoate 3c was put in reaction
with (S)-phenylglycinol in the conditions used for 1b,
the expected enamino ester 9 was not observed (Scheme
4). In place, we isolated the bicyclic derivative 10 in
90% yield.23 This compound obtained as a mixture of
two diastereomers in a 90:10 ratio, is issued from the
intramolecular addition of the hydroxymethyl moiety
of 9 to its enamine double bond.
In conclusion, this procedure for preparing chiral hete-
rocyclic enamino esters proceeds in a high overall yield
and provides an alternative to the excellent procedure
of Eschenmoser with the operational advantages that
phosphines need not be employed and that only one
step is required from the chiral amines to enamine
targets. Therefore, this approach minimized the loss of
chiral starting material.12 Thus, our preparation of 1
would improve the total syntheses of pyrrolizidine,
indolizidine and quinolizidine alkaloids we recently
realized.6,7
In view of this paradox, we checked that the halogen
exchange reaction of 3b into 3c by means of NaI was
slow (>3 days) with regard to the amine addition to
acetylenic ester bearing a simple alkyl chain (24 h).
Thus, we assumed that there was a rapid accumulation
of the Michael adduct 5a which then underwent the
halogen exchange, leading to the corresponding iodo
enamine 5b. Cyclization of this preferentially furnished
the cyclohexene 4. This can be explained by considering
the relative hardness and softness of the enamine reac-
tive centers versus the halide moiety in the intramolecu-
lar substitution. The nitrogen atom of enamine is
known to be a hard center while its carbon atom is
considered as a soft one. So, in the case of the enamine
5a which possesses a chlorine atom (hard), N-alkylation
is favored while the corresponding compound 5b bear-
ing a iodine atom (soft), gives rise preferentially to
C-alkylation.
We are currently investigating this approach to further
highly functionalized enamines synthesis, as well as
applications to biologically significant polycyclic
alkaloids.
Acknowledgements
The authors are grateful to the DSM Company for a
generous gift of (S)-phenylglycine. O. David thanks the
Ministe`re de l’Education Nationale, de la Recherche et
de la Technologie for its financial support.
This study by varying experimental conditions did not
allow us to form the enamino ester 1b from methyl
7-chloroheptynoate with a good selectivity. So we
decided to engage in reaction directly methyl 7-iodo-
heptynoate 3c.20 This compound was prepared from
methyl 7-chloroheptynoate 3b in boiling acetone in the
presence of 4 equiv. of sodium iodide. The reaction
between methyl 7-iodoheptynoate 3c and (S)-1-
phenylethylamine was complete in 4 days (Na2CO3,
refluxing CH3CN) and afforded the expected enamino
ester 1b accompanied by only 5% of the cyclohexene 4.
We found that the addition of a catalytic amount of
tetrabutylammonium iodide reduced the reaction time
(2 days) and led to the exclusive formation of 1b
(E-isomer) in 95% yield.21
References
1. (a) Michael, J. P.; Gravestock, D. Eur. J. Org. Chem.
1998, 865–870; (b) Michael, J. P.; Gravestock, D. Synlett
1996, 981–982.
2. Cook, G. R.; Beholz, L. G.; Stille, J. R. Tetrahedron Lett.
1994, 35, 1669–1672.
3. Hart, D. J.; Hong, W.-P.; Hsu, L.-Y. J. Org. Chem. 1987,
52, 4665–4673.

3474
O. Da6id et al. / Tetrahedron Letters 43 (2002) 3471–3474
4. Petersen, J. S.; Fels, G.; Rapoport, H. J. Am. Chem. Soc.
1.72–1.80 and 1.85–1.94 (2m, 4H); 2.40 (t, 2H, J=7.0
Hz); 3.56 (t, 2H, J=6.2 Hz); 3.76 (s, 3H). 13C NMR 62.9
MHz (CDCl3) l (ppm): 17.4; 24.4; 31.0; 43.8; 52.0; 72.9;
88.0; 153.4.
1984, 106, 4539–4547.
5. Hart, D. J.; Kanai, K. J. Am. Chem. Soc. 1983, 105,
1255–1263.
6. Ledoux, S.; Marchalant, E.; Ce´le´rier, J.-P.; Lhommet, G.
Tetrahedron Lett. 2001, 42, 5397–5399.
7. Bardou, A.; Ce´le´rier, J.-P.; Lhommet, G. Tetrahedron
Lett. 1998, 39, 5189–5192.
8. (a) Nikiforov, T.; Stanchev, S.; Milenkov, B.; Dimitrov,
V. Heterocycles 1986, 24, 1825–1829; (b) Bardou, A.;
Ce´le´rier, J.-P.; Lhommet, G. Tetrahedron Lett. 1997, 38,
8507–8510.
18. (a) Bunce, R. A.; Peeples, C. J.; Jones, P. B. J. Org.
Chem. 1992, 57, 1727–1733; (b) Bunce, R. A.; Allison, J.
C. Synth. Commun. 1999, 29, 2175–2186.
19. For spectral data of 4, see: Cave´, C.; Daley, V.; d’Angelo,
J.; Guingant, A. Tetrahedron: Asymmetry 1995, 6, 79–82.
20. For spectral data of 3c, see: Kita, Y.; Okunaka, R.;
Honda, T.; Shindo, M.; Taniguchi, M.; Kondo, M.;
Sasho, M. J. Org. Chem. 1991, 56, 119–125.
9. Ledoux, S. Ph.D. Thesis, Universite´ P. and M. Curie,
Paris, 2000.
10. (a) Roth, M.; Dubs, P.; Go¨tschi, E.; Eschenmoser, A.
Helv. Chim. Acta 1971, 54, 710–734; (b) Shiosaki, K.
Comprehensive Organic Synthesis; Pergamon Press:
Oxford, 1991; pp. 865–892.
21. For the (E)-stereochemistry of the double bond, see Ref.
10b, p. 872. Selected data for 1b: 1H NMR 250 MHz
(CDCl3) l (ppm): 1.52 (d, 3H, J=7 Hz); 1.61–1.65 (m,
4H); 2.78–2.9 (m, 1H); 2.92–3.08 (m, 1H); 3.12–3.31 (m,
2H); 3.60 (s, 3H); 4.87 (s, 1H); 5.12 (q, 1H, J=7 Hz);
7.23–7.34 (m, 5H). 13C NMR 62.9 MHz (CDCl3) l
(ppm): 15.6; 19.7; 23.3; 26.7; 42.3; 50.3; 55.6; 81.8; 127.3;
127.7; 128.9; 142.1; 165.1; 171.6. [h]2D0 −121 (c 1.10,
CH2Cl2), mp 52°C. Anal. calcd for C16H21NO2: C, 74.10;
H, 8.16; N, 5.40. Found: C, 73.99; H, 8.00; N, 5.27.
11. Campbell, J. A.; Rapoport, H. J. Org. Chem. 1996, 61,
6313–6325.
12. Synthesis of 1a8 and 1b9 through Eschenmoser
condensation:
( )n
( )n
Cl
( )n
P4S10
1/ BrCH2COOMe
2/ PPh3, NEt3
2
1a: n= 1
1b: n= 2
Cl
O
+
S
O
N
N
Ph
CH3
Ph
CH3
22. For the (E)-stereochemistry of the double bond, see Ref.
10b, p. 872. Selected data for 8: 1H NMR 250 MHz
(CDCl3) l (ppm): 1.67 (t, 1H); 1.93–2.04 (m, 2H); 3.22
(m, 2H); 3.32–341 (m, 1H); 3.51–3.61 (m, 1H); 3.58 (s,
3H); 4.05–4.11 (m, 2H); 4.66 (s, 1H); 4.79 (t, 1H, J=6.2
Hz); 7.19–7.39 (m, 5H). 13C NMR 62.9 MHz (CDCl3) l
(ppm): 21.1; 32.7; 48.3; 50.1; 60.3; 62.6; 78.6; 127.0; 127.9;
128.9; 136.7; 166.1; 170.1. [h]2D0 +224 (c 1.18, CHCl3), mp
61°C. Anal. calcd for C15H21NO3: C, 68.94; H, 7.33; N,
5.36. Found: C, 69.07; H, 7.39; N, 5.29.
13. Wilson, C. A., II; Bryson, T. A. J. Org. Chem. 1975, 40,
800–801.
14. Ma, D.; Zhu, W. Org. Lett. 2001, 3, 3927–3929.
15. Molander, G. A.; McKie, J. A. J. Org. Chem. 1993, 58,
7216–7227.
16. Capillary gas chromatography of the crude product
showed 98% of 1a. For the (E)-stereochemistry of the
double bond, see Ref. 10b, p. 872. Flash chromatography
1
gave pure 1a in 95% yield. Selected data: H NMR 250
MHz (CDCl3) l (ppm): 1.53 (d, 3H, J=7 Hz); 1.70–2.00
(m, 2H); 3.06–3.34 (m, 4H); 3.60 (s, 3H); 4.65 (s, 1H);
4.85 (q, 1H, J=7 Hz); 7.15–7.34 (m, 5H). 13C NMR 62.9
MHz (CDCl3) l (ppm): 16.8; 20.9; 32.8; 47.2; 49.9; 52.9;
77.9; 126.5; 127.4; 128.6; 140.4; 164.9; 169.9. [h]2D0 −256 (c
1.1, EtOH) mp 71°C. Anal. calcd for C15H19NO2: C,
73.44; H, 7.80; N, 5.70. Found: C, 73.56; H, 7.93; N,
5.57.
23. Absolute configurations of the two isomers of 10 could
not be assigned. Selected data for the major diastereomer
1
10: H NMR 250 MHz (CDCl3) l (ppm): 1.17–1.24 (m,
2H); 1.42–1.84 (m, 4H); 2.5–2.98 (m, 4H); 3.61–3.71 (m,
1H); 3.75 (s, 3H); 4.18–4.26 (m, 1H); 4.32–4.39 (m, 1H);
7.26–7.36 (m, 5H). 13C NMR 62.9 MHz (CDCl3) l
(ppm): 18.4; 21.9; 31.1; 41.4; 42.4; 51.3; 62.2; 72.9; 93.4;
127.5; 127.6; 128.0; 128.3; 139.4; 170.5. Anal. calcd for
C16H21NO3: C, 69.79; H, 7.68; N, 5.08. Found: C, 69.82;
H, 7.73; N, 5.02.
17. Compound 3b was prepared in quantitative yield from
6-chlorohex-1-yne as described for 3a in Ref. 15. Selected
data for 3b: 1H NMR 250 MHz (CDCl3) l (ppm):