Reductive alkylation of β-alkoxy aziridines: New route to substituted allylic amines

Article abstract of DOI:10.1021/ol048201l

(Chemical Equation Presented) A new route to substituted cyclic allylic amines via the reductive alkylation of β-alkoxy aziridines using excess alkyllithium reagents is described.

Full text of DOI:10.1021/ol048201l

ORGANIC
LETTERS
2004
Vol. 6, No. 26
4817-4819
Reductive Alkylation of
Aziridines: New Route to Substituted
Allylic Amines
â-Alkoxy
Clare M. Rosser,† Susannah C. Coote,† Jonathan P. Kirby,† Peter O’Brien,*,† and
Darren Caine‡
Department of Chemistry, UniVersity of York, Heslington, York YO10 5DD, U.K., and
GlaxoSmithKline Pharmaceuticals, Medicines Research Centre, Gunnels Wood Road,
SteVenage SG1 2NY, U.K.
paob1@york.ac.uk
Received September 6, 2004
ABSTRACT
A new route to substituted cyclic allylic amines via the reductive alkylation of â-alkoxy aziridines using excess alkyllithium reagents is described.
Lithiated epoxides (oxiranyl anions) are now firmly estab-
lished as useful synthetic intermediates.1,2 In contrast,
harnessing the synthetic utility of the corresponding lithiated
aziridines has proved somewhat more problematic,1a,d and
this is particularly true for lithiated aziridines generated by
R-lithiation/deprotonation of aziridines that lack an anion-
stabilizing group.3 As part of our ongoing studies into
reactions of lithiated aziridines,4 we recently reported the
sec-butyllithium-mediated reductive alkylation of N-tosyl
aziridine cis-1 to give cyclopentene 2 in 76% yield (together
with p-toluenesulfonamide, TsNH2) (Scheme 1).5
from 4, itself generated from carbenoid insertion into the
C-Li bond of sec-butyllithium. Although well-known for
epoxides,6,7 we were surprised to find that this was the first
report of reductive alkylation of aziridines. To extend the
synthetic potential of lithiated aziridines and to further exploit
the reductive alkylation of aziridines, we now report the
conversion of â-methoxy aziridines 5 to substituted allylic
amines 7 (via 6) (Scheme 2). On the basis of the precedent
with epoxides,8,9 we incorporated a â-alkoxy group into the
aziridine 5 so that it could undergo alkoxide elimination from
6 and not elimination of TsNLi2 as in 4 (Scheme 1). In this
(1) (a) Satoh, T. Chem. ReV. 1996, 96, 3303. (b) Doris, E.; Dechoux, L.;
Mioskowski, C. Synlett 1998, 337. (c) Hodgson, D. M. Gras, E. Synthesis
2002, 1625. (d) For a special issue of Tetrahedron on oxiranyl and aziridinyl
anions, see: Tetrahedron 2003, 9693-9847.
(2) For recent examples, see: (a) Capriati, V.; Florio, S.; Luisi, R.; Nuzzo,
I. J. Org. Chem. 2004, 69, 3330. (b) Hodgson, D. M.; Paruch, E. Tetrahedron
2004, 60, 5185.
Scheme 1. Reductive Alkylation of an Aziridine
(3) (a) Beak, P.; Wu, S.; Yum, E. K.; Jun, Y. M. J. Org. Chem. 1994,
59, 276. (b) Yamauchi, Y.; Kawate, T.; Katagiri, T.; Uneyama, K.
Tetrahedron 2003, 59, 9839. (c) Vedejs, E.; Bhanu Prasad, A. S.; Kendall,
J. T.; Russel, J. S. Tetrahedron 2003, 59, 9849. (d) Mu¨ller, P.; Riegert, D.;
Bernardinelli, G. HelV. Chim. Acta 2004, 87, 227.
(4) O’Brien, P.; Rosser, C. M.; Caine, D. Tetrahedron Lett. 2003, 44,
6613.
(5) O’Brien, P.; Rosser, C. M.; Caine, D. Tetrahedron 2003, 59, 9779.
(6) (a) Crandall, J. K.; Lin, L.-H. C. J. Am. Chem. Soc. 1967, 89, 4526.
(b) Crandall, J. K.; Lin, L.-H. C. J. Am. Chem. Soc. 1967, 89, 4527.
(7) Doris, E.; Dechoux, L.; Mioskowski, C. Tetrahedron Lett. 1994, 35,
7943.
The formation of cyclopentene 2 presumably proceeds via
lithiated aziridine 3 and subsequent elimination of TsNLi2
† University of York.
‡ GlaxoSmithKline.
(8) Dechoux, L.; Doris, E.; Mioskowski, C. Chem. Commun. 1996, 549.
10.1021/ol048201l CCC: $27.50
© 2004 American Chemical Society
Published on Web 11/19/2004

(via the keto aziridines13) can be utilized to produce single
diastereomers of cis-hydroxy aziridines 9a-d. For reduction
of the keto aziridines, NaBH4 in MeOH gave complete
stereocontrol with the cyclopentenes, but high stereoselec-
tivity with the cyclohexenes necessitated the use of L-
selectride in THF at -78 °C. The high cis-selectivity of these
reductions is mirrored in nucleophilic additions to the
corresponding keto-epoxides8,14 and results from attack on
the least hindered CdO face and anti to the adjacent C-O
Scheme 2. Reductive Alkylation of â-Alkoxy Aziridines
way, the sulfonamide group is retained in the product 7.10
Herein, we report the stereoselective synthesis of cyclic
â-methoxy aziridines cis-5 and their reductive alkylation to
substituted allylic amines 7.
Our route to cyclopentene and cyclohexene â-methoxy
aziridines cis-10a-d started with allylic alcohols 8a-d,
prepared by Luche reduction of the enones. Aziridination
of 8a-d using phenyltrimethylammonium tribromide (PTAB)
and Chloramine-T11 gave hydroxy aziridines 9a-d in high
yields (78-90%) and with a satisfactory degree of cis
selectivity (Scheme 3).
bond. Finally, methylation using Ag2O and MeI gave the
required â-methoxy aziridines cis-10a-d (typical NaH/MeI
conditions were much less successful).
The relative stereochemistry of hydroxy aziridines 9a and
9c was established unequivocally using an aza-Payne rear-
rangement.15 Only the trans-hydroxy aziridines will undergo
a base-mediated conversion into an amino epoxide. Thus,
upon reaction with KHMDS, trans-9a (n ) 0) gave epoxide
trans-11a and trans-9c (n ) 1) gave known16 epoxide trans-
11c (Scheme 4). The relative stereochemistry of hydroxy
aziridines 9b and 9d was assigned by analogy.
Scheme 4. Aza-Payne Rearrangement of trans-9a and -9c
Scheme 3. Synthesis of â-Methoxy Aziridines cis-10a-d
With â-methoxy aziridines cis-10a-d in hand, we were
now ready to study their reductive alkylation using excess
alkyllithiums. Our typical protocol involved reaction of the
â-methoxy aziridine with 2.5 equiv of alkyllithium in Et2O
at -78 °C for 5 min followed by warming to room
temperature over 1 h. In this way, â-methoxy aziridines cis-
10a-d were converted into substituted allylic amines 12-
15 in 6-67% yield (Scheme 5).
Two general conclusions can be made on the basis of these
preliminary results. First of all, cyclopentene aziridines
appear to be more susceptible to reductive alkylation and
thus they generally give higher yields of allylic amines (e.g.,
12a,b and 13a-c, 53-67% yield). This is consistent with
our previous findings on the rearrangement of aziridines to
allylic amines using sec-butyllithium and (-)-sparteine.4,5
The cis-selectivity of the aziridination presumably arises
from preferential bromination on the less hindered face of
the alkene (opposite to the hydroxyl group); similar stereo-
selectivity has been noted in an expanding number of related
examples.11,12 Aziridines cis- and trans-9a-d can be sepa-
rated by chromatography to give 44-68% isolated yields of
cis-9a-d. Alternatively, an oxidation-reduction sequence
(13) Related N-alkyl and N-CO2Et keto aziridines have been prepared
using other routes. See: (a) Fioravanti, S.; Pellacani, L.; Tabanella, S.;
Tardella, P. A. Tetrahedron 2003, 54, 14105. (b) Barros, M. T.; Maycock,
C. D.; Ventura, M. R. Tetrahedron Lett. 2002, 43, 4329.
(9) (a) Hodgson, D. M.; Stent, M. A. H.; Wilson, F. X. Org. Lett. 2001,
3, 3401. (b) Hodgson, D. M.; Stent, M. A. H.; Wilson, F. X. Synthesis
2002, 1445. (c) Hodgson, D. M.; Maxwell, C. R.; Miles, T. J.; Paruch, E.;
Matthews, I. R.; Witherington, J. Tetrahedron 2004, 60, 3611.
(10) A similar strategy using dihydrofuran aziridine and aziridines of
acyclic allylic ethers has recently been disclosed: Hodgson, D. M.; Stefane,
B.; Miles, T. J.; Witherington, J. Chem. Commun. 2004, 2234.
(11) Jeong, J. U.; Tao, B.; Sagasser, I.; Henniges, H.; Sharpless, K. B.
J. Am. Chem. Soc. 1998, 120, 6844.
(12) (a) Caine, D.; O′Brien, P.; Rosser, C. M. Org. Lett. 2002, 4, 1923.
(b) Schmitt, A. C.; Smith, C. M.; Voight, E. A.; O’Doherty, G. A.
Heterocycles 2003, 62, 635. (c) Armstrong, A.; Cumming, G. R.; Pike, K.
Chem. Commun. 2004, 812.
(14) (a) Sepu´lveda, J.; Soto, S.; Mestres, R. Bull. Soc. Chim. Fr. 1983,
233. (b) Sepu´lveda, J.; Soriano, C.; Mestres, R.; Sendra, J. Bull. Soc. Chim.
Fr. 1983, 241. (c) Sepu´lveda, J.; Soriano, C.; Roquet-Jalmar, J.; Mestres,
R.; Riego, J. Bull. Soc. Chim. Fr. 1987, 189.
(15) (a) Ibuka, T.; Nakai, K.; Habashita, H.; Hotta, Y.; Otaka, A.;
Tamamura, H.; Fujii, N.; Chounan, Y.; Yamamoto, Y. J. Org. Chem. 1995,
60, 2044. (b) Hudlicky, T.; Rinner, U.; Gonzalez, D.; Akgun, H.; Schilling,
S.; Siengalewicz, P.; Martinot, T. A.; Pettit, G. R. J. Org. Chem. 2002, 67,
8726.
(16) (a) O’Brien, P.; Childs, A. C.; Ensor, G.; Hill, C. L. Kirby, J. P.;
Dearden, M. J.; Oxenford, S.; Rosser, C. M. Org. Lett. 2003, 5, 4955. (b)
Ba¨ckvall, J.-E.; Oshima, K.; Palermo, R. E.; Sharpless, K. B. J. Org. Chem.
1979, 44, 1953.
4818
Org. Lett., Vol. 6, No. 26, 2004

Scheme 5. Reductive Alkylation of â-Methoxy Aziridines
Scheme 6. Suggested Mechanism for Formation of 16
shift (to give 16). Such a 1,2-alkyl shift is well-established
for alkoxy epoxides18 and, in Et2O, may have a more
concerted mechanism, i.e., from 17 f 16, without the
intermediacy of the carbene 18.18a However, the 1,2-alkyl
shift was not observed with cis-â-methoxy epoxides,8
analogous to the aziridines studied here. In addition, we did
not isolate any enamides from aziridines cis-10a-c.
In summary, a new route to cyclic substituted allylic
amines 12-15 via the reductive alkylation of â-methoxy
aziridines has been established. Furthermore, as shown by
the reactions of aziridine cis-10d and in contrast to the
corresponding â-methoxy epoxides, two different reaction
manifolds occur. Future work will focus on optimizing
reaction conditions so that products of type 15 (reductive
alkylation) and 16 (1,2-alkyl shift) can each be obtained in
high yields.
Acknowledgment. We thank the EPSRC and Glaxo-
SmithKline for a CASE award (to C.M.R.), the EPSRC for
a DTA award (to S.C.C.), The University of York Innovation
and Research Priming Fund for a grant (to J.P.K.), and Prof
D. Craig (Imperial College London) for discussions relating
to the aza-Payne rearrangement.
Second, the best alkyllithium reagent for reductive alkylation
is trimethylsilylmethyllithium (e.g., 12a, 67%; 13a, 67%;
14a, 67%; 15a, 47%), and the worst is methyllithium.
Unfortunately, we have been unable to prepare methoxy
aziridines trans-10a-d by methylation (under a range of
conditions), and this has so far precluded a study of their
reductive alkylation.
On careful inspection of the product mixtures generated
from â-methoxy aziridine cis-10d, we noticed the formation
of the same byproduct, enamide 16 (13-27% yield). The
regiochemistry of enamide 16 was confirmed by NOESY
experiments, and it must be formed by a skeletal rearrange-
ment process. Our suggested mechanism for the formation
of 1617 is outlined in Scheme 6. Thus, initial lithiation occurs
to give 17, but reductive alkylation of 17 by the second
alkyllithium (to give 6, see Scheme 2) is competitive with
R-elimination (to give carbene 18) and subsequent 1,2-methyl
Supporting Information Available: Representative ex-
perimental procedures for the preparation of cis-10d and
12a-d, 13a-c, 14a-d, and 15a-d/16 and full characteriza-
1
tion data and copies of H NMR and 13C NMR spectra of
cis-9d, cis-10d, 12a-d, 13a-c, 14a-d, 15a-d, and 16. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL048201L
(17) The isolation of an ene-sulfonamide from the reaction of N-tosyl
cyclohexeneaziridine with sec-butyllithium and (-)-sparteine has been
described previously. See: Mu¨ller, P.; Nury, P. HelV. Chim. Acta 2001,
84, 662. However, isolation of the ene-sulfonamide could not be reproduced
by the same group; see ref 3d.
(18) (a) Doris, E.; Dechoux, L.; Mioskowski, C. J. Am. Chem. Soc. 1995,
117, 12700. (b) Doris, E.; Mioskowski, C.; Dechoux, L.; Agami, C. J. Org.
Chem. 1998, 63, 3808.
Org. Lett., Vol. 6, No. 26, 2004
4819