rides are often unstable and highly reactive. The analogous
sulfonyl cyanides are easier to manipulate but are quite
expensive and generate toxic byproducts.
Table 1. Ligand Effectsa
Because of the limitations of the above transformations,
the development of a new method for the synthesis of
sulfonyl ketimines under mild conditions would be desirable.
An ideal method would be general, atom economical,
catalytic, and tolerant of a wide range of functional groups
and would not require toxic or otherwise hazardous reagents.
During the course of our studies on the reactivity of
transition metal complexes toward aziridines13 we found that
treatment of 2-butyl-N-tosylaziridine (1) with a stoichiometric
mixture of Pd2(dba)3/PCy3 in C6D6 at 70 °C led to the clean
formation of N-tosylketimine 2 (eq 1).14,15 The transformation
entry
metal complex
ligand
P(o-tol)3
Cy2P(o-biphenyl)
PCy3
reaction timeb
1
2
3
4
5
6
7
8
9
Pd2(dba)3
Pd2(dba)3
Pd2(dba)3
Pd(OAc)2
Pd[P(t-Bu)3]2
Pd[PCy)3]2
Pd[P(t-Bu)2Me]2
Pd[P(t-Bu)2Cy]2
Pd[P(t-Bu)Cy2]2
no reaction
no reaction
8 h
no reaction
no reaction
1 h
25 min
no reaction
>72 hc
PCy3
a Conditions: 1.0 equiv of substrate, 1.0 equiv of metal complex, C6D6
1
(0.015 M). All reactions provided a single product as judged by H NMR
analysis. b Time required for the complete conversion of substrate to product.
c The reaction was stopped after 72 h and was found to have proceeded to
66% completion.
of 1 to 2 was found to proceed effectively with only a
catalytic amount (5 mol %) of palladium. Given the mild
nature of these reaction conditions and the essentially
quantitative conversion of aziridine to imine, we reasoned
that the palladium-catalyzed isomerization of aziridines could
potentially be employed as a very mild synthesis of N-tosyl
ketimines that would tolerate a wide variety of functional
groups. Our preliminary studies on the efficacy of this
transformation are described herein.
In contrast to 2-vinyl-N-tosylaziridines, which readily
undergo oxidative addition to Pd(0),16 the oxidative addition
of 2-alkyl or -aryl aziridines to Pd(0) has not been previously
described.13 Thus, to maximize the kinetic efficiency of this
process, we examined the stoichiometric reaction of 1 with
a number of palladium/phosphine complexes. Mixtures of
Pd2(dba)3 and triarylphosphines such as PPh3, P(o-tol)3,
BINAP, and dppf did not provide active catalysts for this
reaction. In contrast, palladium complexes supported by
electron-rich trialkylphosphines promoted the isomerization
(Table 1). In general, preformed L2Pd complexes were more
reactive than catalysts generated in situ from Pd2(dba)3 or
Pd(OAc)2.17 The steric properties of the ligand had a dramatic
impact on reactivity. For example, both Pd[PCy3]2 (3) and
Pd[P(t-Bu)2Me]2 (4) were sufficiently reactive to promote
the isomerization at room temperature, whereas Pd[P(t-Bu)3]2
did not catalyze the reaction even at 80 °C. Although 4 was
more reactive than 3, most reactions were conducted with
commercially available 3.
As shown in Table 2, a number of N-tosylaziridines are
isomerized to N-tosylketimines in good yield in the presence
of 2-4 mol % 3. The reaction conditions are sufficiently
mild to tolerate a variety of functional groups including esters
(entry 3), olefins (entry 3), acetals (entry 4), and ketones
(entry 5). The isomerization of these functionalized substrates
is particularly noteworthy, as highly functionalized N-
tosylketimines would be difficult to prepare using existing
methods and have not been previously described. In contrast
to the related isomerization of styrene oxide to 2-phenylacetal-
dehyde,14b 2-phenyl N-tosylaziridine 5 reacts at the less
hindered carbon rather than the activated benzylic position,
to afford sulfonyl ketimine 6 (entry 2). We also observed
that N-acylaziridine 8 was converted to N-acylenamine 9 in
47% yield (entry 7).18 At present, this method is limited to
the isomerization of terminal aziridines; 1,2-disubstituted
aziridines did not react under our standard conditions (entry
6). We are currently pursuing the development of other
catalysts that may be capable of transforming more highly
substituted substrates.
(12) Boger, D. L.; Corbett, W. L. J. Org. Chem. 1992, 57, 4777-4780.
(13) (a) Hillhouse has recently reported the synthesis of an azanickela-
cyclobutane via oxidiative addition of 2-butyl-N-tosylaziridine to (bipy)-
NiEt2 or (bipy)Ni(COD). See: Lin, B. L.; Clough, C. R.; Hillhouse, G. L.
J. Am. Chem. Soc. 2002, 124, 2890-2891. (b) The oxidative addition of
ethyleneimine to Pt(II) under acidic conditions has been previously
described. See: Mitchenko, S. A.; Slinkin, S. M.; Vdovichenko, A. N.;
Zamashchikov, V. V. Metallorg. Khim. 1991, 4, 1031-1035.
(14) The palladium-catalyzed isomerization of epoxides to ketones has
been previously reported. See: (a) Suzuki, M.; Watanabe, A.; Noyori, R.
J. Am. Chem. Soc. 1980, 102, 2095-2096. (b) Kulasegaram, S.; Kulawiec,
R. J. J. Org. Chem. 1997, 62, 6547-6561 and references therein.
(15) Control experiments demonstrated that the aziridine did not isomer-
ize when heated for 15 h at 70 °C in C6D6 with 1 equiv of tricyclohexyl-
phosphine in the absence of Pd.
(16) For recent examples in catalytic reactions, see: (a) Butler, D. C.
D.; Inman, G. A.; Alper, H. J. Org. Chem. 2000, 65, 5887-5890. (b)
Fugami, K.; Morizawa, Y.; Ishima, K.; Nozaki, H. Tetrahedron Lett. 1985,
26, 857-860. (c) Aoyagi, K.; Nakamura, H.; Yamamoto, Y. J. Org. Chem.
2002, 67, 5977-5980. (d) Trost, B. M.; Fandrick, D. R. J. Am. Chem. Soc.
2003, 125, 11836-11837.
In most cases the isolation of analytically pure N-
tosylimines was achieved by rapid chromatography on silica
gel. In some cases (Table 2, entries 4 and 5) the isolated
products were contaminated with small amounts (ca. 5-15%)
of byproducts resulting from partial hydrolysis of the imine
by silica gel. However, we found the metal-catalyzed
(17) Mixtures of Pd2(dba)3/PCy3 did effectively catalyze the isomerization
of 1. However, chromatographic separation of the imine product from dba
was very difficult.
(18) Aziridines bearing N-benzyl or N-diphenylphosphoryl groups did
not react under our standard conditions. The reaction of an N-boc aziridine
proceeded very slowly, requiring 1 equiv 3 and >24 h at 80 °C to reach
completion.
4608
Org. Lett., Vol. 5, No. 24, 2003