Diastereoselective synthesis of 2-aryl-3-aminoazepanes via a novel ring- enlargement process

Article abstract of DOI:10.1016/S0040-4039(99)02256-X

Syntheses of optically pure 2-aryl-3-aminoazepanes derived from 2-cyano 6-oxazolopiperidine are described. The key step involves a one-pot reduction and ring-enlargement process occurring in a highly regio- and stereoselective way. (C) 2000 Elsevier Scien

Full text of DOI:10.1016/S0040-4039(99)02256-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 1179–1182
Diastereoselective synthesis of 2-aryl-3-aminoazepanes via a
novel ring-enlargement process
Sabrina Cutri,a Martine Bonin,a Laurent Micouin,a Olivier Froelich,a Jean-Charles Quirion b
and Henri-Philippe Husson a,∗
aLaboratoire de Chimie Thérapeutique associé au CNRS et à l’Université René Descartes (UMR 8638), Faculté des Sciences
Pharmaceutiques et Biologiques, 4 Avenue de l’Observatoire, 75270 Paris cedex 06, France
bLaboratoire d’Hétérochimie Organique associé au CNRS, IRCOF, rue Tesnière, 76821 Mont-Saint-Aignan cedex, France
Received 15 October 1999; accepted 2 December 1999
Abstract
Syntheses of optically pure 2-aryl-3-aminoazepanes derived from 2-cyano 6-oxazolopiperidine are described.
The key step involves a one-pot reduction and ring-enlargement process occurring in a highly regio- and stereo-
selective way. © 2000 Elsevier Science Ltd. All rights reserved.
Enantiomerically pure 1,2-diamines are of great interest, playing key roles in coordination chemistry,
in asymmetric catalysis and as medicinal agents.1 In these domains and more particularly in peptide
research, constrained diamine systems have found application as biological active compounds. Among
the seven-membered nitrogen containing rings, 3-aminocaprolactam is a particularly attractive scaffold
which has been widely used during the past decade. More recently, original and significant biological
activities were discovered in the 3-aminoazepane series. For instance, the arylether 1 is a dopamine
reuptake inhibitor,2 benzamides 2 are D2 or 5HT3 receptor antagonists,3 and the platinum complexes of
3 are employed as antitumor agents4 (Fig. 1).
Fig. 1.
∗
Corresponding author. Tel: 33 (0)1 53 73 97 54; fax: 33 (0)1 43 29 14 03; e-mail: husson@pharmacie.univ-paris5.fr (H.-P.
Husson)
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(99)02256-X
tetl 16184

1180
However, few asymmetric syntheses described the preparation of diversely C-substituted 3-
aminoazepane compounds. Whereas 3-amino derivatives 1–3 were simply derived from 3-
aminocaprolactam through LiAlH4 reduction, di-substituted azepinones related to peptidomimetic
compound 4 were prepared using various approaches, via Beckmann rearrangement, cyclization of
aminoester or lysine derivatives.5 Surprisingly, to our knowledge, no asymmetric syntheses have been
yet reported in the 2-substituted 3-amino series, in contrast to the numerous methods developed for the
preparation of corresponding piperidines.6
These observations prompted us to develop a novel and general synthetic route giving access to this
type of constrained diamines. As part of our on-going programme on asymmetric synthesis we recently
described a rapid and convenient route to optically pure 2-aminomethylpiperidines 8 or 9 starting with
2-cyano 6-oxazolopiperidine 57 (Scheme 1).
Scheme 1.
The excellent diastereoselectivities observed were explained by a preferential attack of the alcohol
function on the si face of the protonated form of imines 6, followed by regioselective cleavage of the
C-6/N-8 bond of the resulting aminal intermediates 7.
Considering the non-classical bridged-system of the bicyclic imines 6, we were interested in testing
other reduction methods allowing the C-6/N-1 bond cleavage giving a rapid access to the 3-aminoazepane
ring system substituted in the C-2 position. We report herein our preliminary results in this novel diamine
series with the synthesis of 2-substituted 3-aminoazepanes 12.
Since previous protic conditions lead exclusively to the piperidine derivatives 8 or 9 via the tricyclic
intermediates 7, we were interested in testing mixed Lewis acid–hydride reagents in aprotic solvents.
Indeed, we assumed that in such conditions a partial chelation of the tertiary amine (structures 10) could
enhance the cleaving ability of the C-6/N-1 aminal bond.
Initial experiments were conducted on the known n-butylimine 6a.7 Treatment of this compound
with an excess of LiAlH4 in Et2O at 20°C afforded a 3/2 inseparable mixture of the desired azepane
derivative 11a8 and piperidine 9a (Scheme 2). Adding AlCl3 as Lewis acid did not change the selectivity
significantly giving a 1:1 mixture of the same products. Replacement of Et2O by THF afforded the same
results. Finally, the best conditions were found in adding LAH carefully to the imine etheral solution
(20°C, 3 h; 11a/9a 9:1). Interestingly, when LAH in Et2O at −78°C (3 h) or BH3·THF (20°C, 3 h) were
used, the morpholine 8a was produced exclusively.
These results clearly show that morpholines 8 are always intermediates toward the formation of

1181
Scheme 2.
piperidines 9. We therefore studied ring enlargements with aromatic imines, since these derivatives
should be less proned to intramolecular attack by the alcoholate. Thus, for phenylimine 6b,7 as well
as for furylimine 6c, a total conversion to the derivatives 11b and 11c was achieved in good yield (91
and 65%, respectively).9 Careful spectra analyses revealed that in each case, a single diastereomer was
formed. A 2,3-trans configuration was proposed on the basis of the large 9 Hz coupling constant observed
for the H-2 proton signals.
In order to explain the excellent regio and diastereoselectivities observed in these transformations, we
propose a mechanism involving a chelated process (Scheme 2).
The formation of alcoxyaluminum derivative 13 could allow an intramolecular attack of the hydride
nucleophile from the si face of the imine. Then a N,O-bimetallated species 14 can be considered, which
could favor the cleavage of the C-6/N-1 bond. The second reduction can then occur on the resulting
imine intermediate 15 leading to the 2,6-trans disubstituted azepanes 11b and 11c. For the butylimine
6a, the lack of selectivity (11a/9a 9:1) is probably due to the more electrophilic character of imine carbon
compared with the more stable arylimines. Thus piperidine 9a can be formed via the second pathway in
which imine carbon of 13 is preferentially attacked by the alcoholate rather than a hydride species.
To achieve our goal, we then had to cleave the chiral appendage of the secondary amine function.
Thus, treatment of phenylderivative 11b under classical hydrogenolysis conditions furnished the desired
diamine 12b in almost quantitative yield10 (60% overall yield from cyanopiperidine 5) (Scheme 3).
Scheme 3. Reagents and conditions: Ar=Phenyl: H2, Pd/C, MeOH–HCl, rt, 48 h, 96%; Ar=20-Furyl: (i) H5IO6 (2.6 equiv.),
MeNH2, H2O, MeOH, rt, 3 h; (ii) MeOH–HCl, rt, 3 h, 80%
In contrast, applying the same reaction procedure to aminoalcohol 11c provided a mixture of the dia-
mine 12c and its corresponding tetrahydrofuryl derivative.11 In order to avoid such undesired reduction,
we decided to apply oxidative conditions with periodic acid.12 In two successive steps, oxidative cleavage
and methanolysis of the imine intermediate gave 20-furyldiamine 12c in 74% yield (41% overall yield
from 5).13
In conclusion, we have developed an efficient diastereoselective synthesis of new 2-aryl-3-
aminoazepanes, which also allows differentiation of amino group reactivity. As a key step, we have
shown that LAH reduction can convert directly bicyclic piperidinylimines into 3-aminoazepane

1182
derivatives in a totally regio- and diastereoselective way. Further application of this methodology to
synthesize biological compounds is actually in progress in our laboratory.
References
1. Lucet, D.; Le Gall, T.; Mioskowski, C. Angew. Chem., Int. Ed. Engl. 1998, 37, 2580–2627.
2. Matecka, D.; Rothman, R. B.; Radesca, L.; De Costa, B. R.; Dersch, C. M.; Partilla, J. S.; Pert, A.; Glowa, J. R.; Wojnicki,
H. E.; Rice, K. C. J. Med. Chem. 1996, 39, 4704–4716.
3. Hirokawa, Y.; Morie, T.; Yamazaki, H.; Yoshida, N.; Kato, S. Bioorg. Med. Chem. Lett. 1998, 8, 619–624.
4. Rezler, E. M.; Fenton, R. R.; Esdale, W. J.; McKeage, M. J.; Russell, P. J.; Hambley, T. W. J. Med. Chem. 1997, 40,
3508–3515.
5. Robl, J. A.; Sieber-McMaster, E.; Sulsky, R. Tetrahedron Lett. 1996, 37, 8985–8988.
6. (a) Martin-Martinez, M.; Bartolomé-Nebreda, J. M.; Gomez-Monterrey, I.; Gonzalez-Munich, R.; Garcia-Lopez, M. T.;
Ballaz, S.; Barber, A.; Fortuno, A.; Del Rio, J.; Herranz, R. J. Med. Chem. 1997, 40, 3402–3407. (b) Laschat, S.; Fröhlich,
R.; Wibbeling, B. J. Org. Chem. 1996, 61, 2829–2838. (c) Armour, D. R.; Chung, K. M. L.; Congreve, M.; Evans, B.;
Guntrip, S.; Hubbard, T.; Kay, C.; Middlemiss, D.; Mordaunt, J. E.; Pegg, N. A.; Vinader, M. V.; Ward, P.; Watson, S. P.
Bioorg. Med. Chem. Lett. 1996, 6, 1015–1020.
7. (a) Bonin, M.; Grierson, D. S.; Royer, J.; Husson, H.-P. Org. Synth. 1991, 70, 54–59. (b) Froelich, O.; Desos, P.; Bonin, M.;
Zhu, J.; Quirion, J.-C.; Husson, H.-P. J. Org. Chem. 1996, 61, 6700–6705.
8. Satisfactory spectral and analytical data were obtained for all new compounds.
9. Typical procedure for imine reduction. Preparation of 11b: To a cooled suspension of LiAlH4 (0.8 g, 20.9 mmol) in ether
(50 mL, −10°C) was carefully added a solution of phenylimine 6b (0.81 g, 2.67 mmol) in ether (4 mL). After stirring the
mixture for 3 h, NaOH (1N, 1.6 mL) was added dropwise, followed by H2O (2.4 mL). After filtration and removal of the
solvent under reduced pressure, the oily residue was purified by flash chromatography (SiO2, CH2Cl2/MeOH 9:1) to give
aminoalcohol 11b as a colorless oil in 91% yield (0.75 g, 2.42 mmol).
10. Diamine (+) 12b. Colorless oil. [α]D= +14 (c=1, MeOH). 1H NMR (300 MHz, CDCl3) (δ, ppm; J, Hz): 2.1–1.5 (6H, m),
2.83 (2H, m, 2H7), 3.03 (1H, ddd, J=9.2, 7.8, 4.1, H3), 3.21 (1H, d, J=9.2, H2), 7.1–7.4 (5H, m, Ph); 13C NMR (75.43 MHz,
CDCl3, δ, ppm): 21.8, 29.5, 35.6 (C4, C5, C6 not assigned), 49.2 (C7), 58.9 (C3), 74.1 (C2), 127.5, 128.3, 128.7, 144.3 (Ph).
Anal. calcd for C12H18N2·2HCl,0.5H2O: C 52.95, H 7.78, N 10.29; found C 52.65, H 7.49, N 10.11.
11. Williams, R. M.; Zhai, W.; Aldous, D. J. J. Org. Chem. 1992, 57, 6527–6532.
12. Chang, Z. Y.; Coates, R. M. J. Org. Chem. 1990, 55, 3475–3483.
13. Diamine (+) 12c. Pale yellow oil. [α]D= +7 (c=0.45, MeOH). 1H NMR (300 MHz, CDCl3) (δ, ppm; J, Hz): 1.80–1.28 (6H,
0
0
m), 2.47 (1H, ddd, J=13.7, 8.4, 3.8, H3), 3.64 (d, J=9, H2), 6.06 (1H, J=3.1, 1.9, H4 ), 6.12 (1H, J=3.2, H3 ), 7.16 (d, J=1.3,
0
H5 ); 13C NMR (75.43 MHz, CDCl3, δ, ppm): 21.6, 28.8, 30.7, (C4, C5, C6 not assigned), 45.6 (C7), 54.6 (C3), 58.5 (C2),
0
0
0
0
108.3, 110.3 (C3 , C4 ), 142.5 (C5 ), 151.1 (C2 ); HRMS calcd for C10H16N20: 180.1262, found: 180.1262.