Angewandte
Chemie
Table 2: Tolerance of epoxide formation to nitrone substitution.
and 1,10-phenanthroline was used as a catalyst (Table 1,
entry 4). The yield of this transformation was dependent upon
the choice of reaction medium, copper salt, and N-donor
ligand. Benzene, THF, and dimethylsulfoxide (DMSO) were
tolerated as solvents, but the most efficient transformation
was observed in acetonitrile (Table 1, entries 4–8). Several
copper(I) and copper(II) salts were observed to catalyze the
oxygen transfer in the presence of phenanthroline, but CuCl
was shown to be the optimal (Table 1 entries 8–12). The
copper catalyst required the presence of a N-donor ligand for
an efficient process and 1,10-phenanthroline was shown to
provide the highest yield (Table 1, entries 8 and 13–15). The
conversion of 1a into 2a can be achieved at 25–808C under
the conditions shown in Table 1, entry 8; however, 808C was
chosen as the preferred reaction temperature, as lower
temperatures required longer reaction times. Compound 2a
was consistently isolated as a mixture of imine isomers, but
the diastereoselectivity of the oxygen transfer reaction was
excellent and only the trans isomer of 2a was observed.
Once optimal conditions for the conversion of aryl nitrone
1a into a,b-epoxyketimine 2a were determined, the scope of
the transformation was investigated. The electronic and steric
preferences of the oxygen-transfer reaction were initially
tested by varying the styrenyl group of chalcone-derived
nitrones (Table 2, entries 1–7). As shown in Table 2, both
electron-rich and electron-poor styrenyl functional groups,
with ortho, meta, and para substitution patterns, were
tolerated by the reaction conditions and provided the desired
a,b-epoxyketimines in excellent yields.[7] Both electron-rich
and electron-poor aryl groups were also screened for the
imine substituent, and were observed to undergo equally
efficient transformations (Table 2, entries 8–11). Alkyl-sub-
stituted a,b-unsaturated nitrones such as 1m were unreactive
under the optimized reaction conditions, but this stark change
in reactivity does not appear to be due to the fact that these
substrates favor the opposite imine isomer to the chalcone-
derived nitrones.[8] A control experiment showed that a 1:1 E/
Z mixture of 1a gave the desired product in high yield. Finally,
dibenzylidene acetone substrates were shown to be tolerated
by the copper-catalyzed transformation, but unsymmetrical
examples did not exhibit an electronic preference for oxygen
transfer (Table 2, entries 13–14).
Entry
1
Product[b]
Entry
8
Product[b]
2b (97%)
2i (88%)
2
3
4
9
2c (99%)
2d (99%)
2e (82%)
2j (95%)
2k (93%)
2l (92%)
10
11
5
12
2 f (98%)
2m (0%)[c]
6
7
13
14
2g (88%)
2h (83%)
2n (96%)[d]
2o (90%, 1:1)[d]
[a] E nitrones were subjected to the reaction conditions. [b] Data in
parentheses are yields of isolated products. [c] 1m was used as a 3.5:1
mixture of E/Z nitrone isomers. [d] Cu(OAc)2 was used.
transfer from benzaldehyde-derived nitrones to o-alkynyl
functionalites for the preparation of enones.[9,10] Although the
intermediacy of an isoxazoline was not observed in the
copper-catalyzed transformations described above, spectral
evidence for this intermediate was observed in the product
mixtures of the less efficient palladium-catalyzed transforma-
tions discussed in Table 1.[11] Ring openings of 2-isoxazolines
of this type are unknown; however, related 4-isoxazoline
rearrangements to form a,b-aziridinyl ketones have been
previously reported.[12]
With a variety of a,b-epoxyketimines in hand, we decided
to investigate the synthetic utility of these compounds for
initial comparison to their a,b-epoxyaldimine analogues. As
shown in Scheme 3, 2a was shown to readily undergo a variety
of regioselective reductions. Treatment of 2a with LiAlH4
proceeded under steric control to give 1,3-aminoalcohol 5 as
a single diastereomer. In contrast, treatment of 2a with
diisobutylaluminium hydride (DIBAL-H) proceeded under
electronic control to give 1,2-aminoalcohol 6 as a 4:1 mixture
In addition to the carbon backbone of the nitrones, the N-
aryl substituent was varied to determine its effect on the
copper-catalyzed preparation of a,b-epoxyketimines from
a,b-unsaturated arylnitrones. Both electron-rich and elec-
tron-poor N-aryl groups with either meta or para substituents
provided the desired products in high yield. We were pleased
to observe that aryl groups with bromide, vinyl, and ester
functional groups were compatible with the reaction condi-
tions, as these substituents further enhanced the potential
synthetic utility of these densely functionalized products
(Table 3, entries 4, 7, and 10).
A proposed mechanism for the rearrangement of a,b-
unsaturated arylnitrones into a,b-epoxyketimines is illus-
trated in Scheme 2. An initial copper-catalyzed attack of the
nitrone at the styrenyl group could precipitate a subsequent
À
N O bond cleavage to form 2. This mechanism is similar to
a pathway proposed for gold-catalyzed intramolecular oxygen
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
These are not the final page numbers!