W. D. Wulff et al.
action of the a-methylbenzyl imine 26a in the matched case
is 2.3 times faster than the same reaction in the mismatched
case. This is consistent with the slightly greater reaction
progress noted for the matched reaction when the reaction
of the imine (R)-26a was stopped after 1 h (Table 1, en-
tries 6 and 7).
the stereoselectivity in the aziridination reaction. This result
was a little unexpected, since it was observed by X-ray anal-
ysis that the binding of the imine substrate in the VAPOL
catalyst 9 results from different types of interactions for the
two phenyl groups.[3b] However, it should be noted that in
computational studies on the transition states for the aziridi-
nation reaction that these types of interactions of the phenyl
groups with the catalyst do not exist.[6]
As indicated by the data in Scheme 5, chiral benzhydryl
substituents provide absolutely no resistance to the catalyst
Substrate scope for cis-aziridines with a-methylbenzyl imine
26: With regard to the selection of a chiral auxiliary, it was
decided to move forward with the a-methylbenzyl amine 19
(Scheme 2). The a-tert-butylbenzyl amine 21 was shown to
give slightly higher diastereoselectivities in reactions of its
imine from benzaldehyde (Table 3) than the corresponding
imine from amine 19 (Table 1); however, the a-tert-butyl-
benzyl amine 21 is not commercially available. Both enan-
tiomers of the a-methylbenzyl amine 19 are commercially
available and, furthermore, both enantiomers are roughly
the same price as the benzhydryl amine 10. The diastereose-
lectivity for the aziridination of the imine 26a are very high
and the two diastereomers 24a and 25a are easy to separate
by column chromatography (Table 1). The isolated yields of
the major diastereomer 24a are also very high (86%). Even
though the yields and diastereoselectivities are slightly
higher in CCl4 than in toluene, it was decided to examine
the scope of the reaction with imines from other aldehydes
in toluene as solvent for reasons of cost and safety. With the
VANOL catalyst in toluene the diastereomer 24a was the
exclusive product with no detectable amount of the diaste-
reomer 25a (Table 1, entry 10).
Scheme 5.
that dominates these reactions with complete catalyst con-
trol. The reaction of the imine (S)-43a (>99% ee) with
a phenyl and a p-bromophenyl substituent gave a 97:3 mix-
ture of 44a to 45a in favor of 44a with the (S)-VAPOL cata-
lyst and a 3:97 mixture with the (R)-VAPOL catalyst. The
fact that the p-bromo substituent has no effect at all and
that the matched and mismatched reactions both give a 97:3
selectivity is consistent with the fact that the benzhydryl
imine 41a will react with ethyl diazoacetate to give aziridine
42a in 93% ee (96.5:3.5 e.r.) in carbon tetrachloride as sol-
vent.[1c] On this basis the relative stereochemistry of 44a and
45a were assigned such that the 2R,3R isomer is produced
by the S enantiomer of the ligand, as is the case for the reac-
tion of imine 41a. An alternative method for probing for
matched and mismatched effects is to look for their manifes-
tation in kinetic resolution experiments. This was examined
for the racemic benzhydryl imines 43a and 43b, bearing p-
bromo and p-methoxy substituents, respectively. The reac-
tion of the racemic p-bromophenyl imine 43a with
0.5 equivalents of ethyl diazoacetate gave a 50:50 mixture of
the diastereomers 44 and 45. This is consistent with the reac-
tions of the optically pure imine 43a with catalysts from (S)-
and (R)-VAPOL. Exactly the same results were obtained
with the racemic p-methoxy imine 43b and thus it is clear
that electronic differences in the two phenyl groups of the
benzhydryl unit on the imine do not have any influence on
The scope of the aziridination of imines from a-methyl-
benzyl amine 19 was explored with five additional aromatic
aldehydes and with three aliphatic aldehydes and the results
are summarized in Table 4. The imines 26c and 26d are
solids and were purified by crystallization prior to use. All
of the rest of the imines in Table 4 are oils and were used
without purification directly after they were prepared. The
optimal results for the para-nitrophenyl imine (R)-26b is
with (S)-VANOL, which gives the diastereomeric aziridine
24b in 90% isolated yield along with 4% of the diastereo-
mer 25b, which could be easily be separated by chromatog-
raphy on silica gel (Table 4, entry 3). A similar situation was
found with the para-bromophenyl imine (R)-26c for which
the strongest matched case was found with the (S)-VANOL
ligand giving a 97:3 mixture of aziridines 24c and 25c, which
were easily separated to give 24c in 77% yield (Table 4,
entry 8). Both ligands were equally effective in the aziridina-
tion of the para-methylphenyl imine (R)-26d with the isola-
tion of the matched imine 24d in 70–71% yield with no de-
tectable amount of the diastereomer 25d formed in either
reaction (Table 4, entries 11 and 13). A dramatic difference
between the two ligands was observed with the para-me-
thoxyphenyl imine (R)-26e, for which even for the matched
case the (S)-VAPOL catalyst only goes to 46% completion,
while the (S)-VANOL catalyst goes to completion and pro-
vides a 63% isolated yield of the pure aziridine 24e
5306
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 5302 – 5313