Toward the homologous N-benzyl-N-pent-4-enylform-
amide (5), the tosylate of 1 was treated with NaCN in
DMSO. Reduction of the resulting nitrile 4 with LiAlH4
provided 4-pentenylamine, which was quantitatively con-
verted to the corresponding formamide by reaction with ethyl
formate. Deprotonation of the amide and benzylation with
benzyl bromide led to 5 in 19% overall yield (Scheme 1).
albeit with yields of only 23 and 32%, respectively.9 Since,
however, the primary attack of the first titanium intermediate
most probably occurs at the terminal double bond, the
trajectory for intramolecular carbonyl attack is rather dis-
favored, and this may be one reason for the poor yield.10
Surprisingly, the analogous treatment of N-benzyl-N-hepta-
4,6-dienylformamide (10) did not yield the expected 2-azabi-
cyclo[4.1.0]heptane derivative but yielded only 15% of exo-
N-benzyl-6-propenyl-2-azabicyclo[3.1.0]hexane (11) (Scheme
2).
Scheme 1a
Scheme 2a
a Reaction conditions: (a) TsCl, Py, 0 °C, 3 days; (b) BnNH2
(2.5 equiv), MeCN, 60 °C, 3 h. (c) NaCN, DMSO, 25 °C, 24 h.
(d) EtOCHO, 65 °C, 24 h. (e) LiAlH4, Et2O, 25 °C, 36 h. (f) (1)
NaH, DMF, 60 °C, 1 h; (2) BnBr, DMF, 25 °C, 24 h. (g) Ti(OiPr)4
(1.1 equiv), cHexMgBr (3.3 equiv), THF, 25 °C, 24 h.
a Reaction conditions: (a) Ti(OiPr)4 (1.1 equiv), cHexMgBr (3.3
equiv), THF, 25 °C, 24 h.
This product most probably arises by initial isomerization
of the precursor 10 under the conditions employed.
In another context,5 we came upon a new synthesis of
2-substituted 3-butenylamines, which were quite welcome
as precursors to 4-substituted 2-azabicyclo[3.1.0]hexanes 14.
Upon treatment of N-benzylpyrroline (12) with 2.5 equiv
of an alkylmagnesium halide in the presence of 1.2 equiv of
titanium tetraisopropoxide, N-benzyl-N-(2-alkylbut-3-enyl)-
amines (alkyl ) ethyl, isopropyl, sec-butyl, cyclopentyl) were
obtained. This transformation must arise from a titanium-
mediated carbomagnesiation of N-benzylpyrroline (12) with
ring opening. It resembles the zirconium-catalyzed carbo-
magnesiations discovered by Dzhemilev et al.11 and further
developed by Hoveyda et al. as well as others.12 However,
in those reactions, only isopropyl- and ethylmagnesium
halides were used.
The intramolecular reductive cyclopropanations of 3 and
5 were carried out by adding 3.3 equiv of cyclohexylmag-
nesium bromide to a THF solution containing 1.1 equiv of
titanium(IV) tetraisopropoxide and 1.0 equiv of the respective
formamide. Conversion of the N-benzyl-N-alkenylform-
amides 3 and 5, after acid/base workup, gave the expected
N-benzyl-2-azabicycloalkanes 6 and 7 in good yields (84 and
67% respectively, Scheme 1).6
In view of this result and the fact that 1,3-dienes are
particularly good ligands on titanium,5,7 the analogous
N-benzyl-N-alkadienylformamides8 were also subjected to
the established cyclization conditions. In accordance with
previous observations that substituted 1,3-dienes with one
terminal vinyl group always undergo aminocyclopropanation
at the more highly substituted double bond, both the (E)-
and (Z)-isomer of N-benzyl-N-hexa-3,5-dienylformamide
(E,Z)-8 gave exo-N-benzyl-6-vinyl-2-azabicyclo[3.1.0]hexane
(9) derived from supposed attack on the internal double bond,
Upon reaction of the crude mixtures of the homoallyl-
amines with ethyl formate, the corresponding formamides
(9) An attempted analogous cyclization of N-benzyl-N-(3-methylenepent-
4-enyl)formamide gave a minute amount (9%) of 2-benzyl-5-methylene-
2-azabicyclo[4.1.0]heptane and traces of a second cyclization product,
possibly the expected exo-N-benzyl-5-vinyl-2-azabicyclo[3.1.0]hexane.
(10) In contrast to the unsubstituted azabicycloalkanes 6 and 7, com-
pounds 9 and 11 upon attempted purification by acid/base workup or
distillation underwent rapid polymerization. Therefore, they had to be
purified by column chromatography, which probably also contributed to
the decreased yields.
(11) (a) Dzhemilev, U. M.; Vostrikova, O. S. J. Organomet. Chem. 1985,
285, 43. (b) Lehmkuhl, H. Bull. Soc. Chim. Fr. II 1967, 87.
(12) (a) Hoveyda, A. H.; Xu, Z. J. Am. Chem. Soc. 1991, 113, 5079. (b)
Houri, A. F.; Didiuk, M. T.; Xu, Z.; Horan, N. R.; Hoveyda, A. H. J. Am.
Chem. Soc. 1993, 115, 6614. (c) Lewis, D. P.; Whitby, R. J. Tetrahedron
1995, 51, 4541. (d) Uesaka, N.; Mori, M.; Okamura, K.; Date, T. J. Org.
Chem. 1994, 59, 4542.
(6) A single example of a similar reaction has been reported by Cha et
al. leading to 1-methyl-N-phenyl-2-azabicyclo[3.1.0]hexane: Lee, J.; Cha,
J. K. J. Org. Chem. 1997, 62, 1584.
(7) Williams, C. M.; Chaplinski, V.; Schreiner, P. R.; de Meijere, A.
Tetrahedron Lett. 1998, 39, 7695 and references therein.
(8) N-Benzyl-N-alkadienylformamides 8 and 10 were synthesized in an
analogous manner as the monoene precursors by using hexa-3,5-dienol as
starting material instead of but-3-enol (1). Preparation of (E)-hexa-3,5-
dienol: Martin, S. F.; Tu, C.-y.; Chou, T.-s. J. Am. Chem. Soc. 1980, 102,
5274. (Z)-Hexa-3,5-dienol: Aerrsens, M. H. P. J.; van der Heiden, R.; Heus,
S.; Brandsma, L. Synth. Commun. 1990, 20, 3421.
484
Org. Lett., Vol. 5, No. 4, 2003