DOI: 10.1002/chem.201103155
N-Fused Indolines through Non-Carbonyl-Stabilized Rhodium Carbenoid
À
C H Insertion of N-Aziridinyl Imines
Stuart J. Mahoney and Eric Fillion*[a]
À
Metal-catalyzed methods of functionalizing C H bonds
have seen incredible advancements in recent times, allowing
for new retrosynthetic disconnections to otherwise unreac-
tive bonds and executing transformations with high chemo-,
regio-, and stereocontrol.[1] A more established area of func-
3
À
À
tionalizing C
H
(sp ) H bonds has been rhodium-catalyzed C
insertions from carbonyl-stabilized diazo substrates,
À
which has reached a level at which even intermolecular C
H insertions with high enantioselectivity have been ach-
ieved.[2] Key to the success of the intermolecular methodolo-
gy was shifting the focus from studying ligand alterations to
substrate design, specifically in moving to donor–acceptor
Scheme 1. General strategy.
carbenoids.[2a–b] Despite the progression, analogous C H in-
In this manuscript, we report a general catalytic protocol
À
À
sertions of carbenoids without an acceptor (primarily car-
bonyl functionalities) have remained elusive due to the in-
herent difficulties with controlling selectivity of the reactive
species. Alternatively, a rapidly developing redox-neutral
of non-carbonyl-stabilized rhodium carbenoid C H inser-
tions enabling rapid synthesis of N-fused indolines and com-
plex heterocycles. The ability of hydrazone 1a to cyclize to
3
À
tricycle 2a through C
G
3
À
method of functionalizing C
(sp ) H bonds has been cata-
first examined (Table 1). Upon screening Lewis and Brøn-
lyzed variants of the tert-amino effect,[3] which now includes
unactivated alkyne and allene acceptors,[4] tertiary aliphatic
hydride donors,[5] domino reactions,[6] and enantioselective
protocols.[7] Seeking to develop a methodology to give direct
access to the privileged N-fused indoline scaffold[8] through
ACHTUNGTRENsNUNG ted acids, only varying amounts of starting material and de-
composition were observed. However, when heating the re-
action (ꢀ708C) in the absence of a promoter, the carbene
pathway was evident by the formation of the desired prod-
À
uct 2a (by C H insertion) along with aldehyde 3a, cyclopro-
3
C
(sp ) H bond functionalization, we turned our attention to
panes 4a,[14] and dimerization products (azine 5a and al-
kenes 6a; Table 1, entry 1). It was then found that the cyclo-
propanes could be selectively formed (4a, trans/cis ratio of
1.6:1) by intermolecular scavenging of the carbene inter-
mediate upon addition of an excess of styrene (Table 1,
entry 2). Optimistic about the possibility of mediating the
carbene reaction[15] with rhodium,[16–17] a catalyst screen was
performed. It was gratifying to see that the product distribu-
À
N-aziridinyl imines 1 (Eschenmoser hydrazones),[9] which
potentially offered two distinct reactivity modes to achieve
the desired transformation (Scheme 1), namely, hydride ac-
ceptor and decomposition to a benzylic carbene.[10] By virtue
of the proposed [1,5] hydride shift/cyclization mechanism
(Scheme 1, path A), the benzylic carbon would act as a
geminal acceptor/donor (effectively a 1,1-dipole) instead of
the typical vicinal acceptor/donor; the net result would be
the formation of a five-membered ring as opposed to the
six-membered ring created with traditionally employed ac-
ceptors.[11–12] Also, cognizant of the ability of the N-aziridinyl
imine to function as a carbene precursor (Scheme 1, path B)
a competing pathway that could lead to N-fused indoline 2
had to be considered.[13]
À
tion changed significantly to predominantly form the C H
insertion product (Table 1, entry 3) in contrast to a recent
report of tosyl hydrazone decomposition, which exclusively
formed alkenes through dimerization.[13e] Steric effects of
dirhodium(II) carboxamidates proved to be beneficial
(Table 1, entry 4 versus entries 5 and 6). Additional experi-
ments with [Rh2ACHTNUTRGNEUNG(cap)4] (cap=caprolactamate) probing
higher dilution, increased catalyst loading, and slow addition
of the substrate resulted in negligible improvement in for-
[a] S. J. Mahoney, Prof. Dr. E. Fillion
Department of Chemistry, University of Waterloo
200 University Avenue West, Waterloo, Ontario
Canada N2L 3G1 (Canada)
À
mation of the C H insertion product 2a (Table 1, entries 7–
9). The catalyst of choice was determined to be [Rh
2A
MEPY)4]
(5S-MEPY=methyl-2-oxopyrrolidine-5(S)-car-
Fax : (+1)519-746-0435
boxylate) on the basis of its slight superiority in terms of se-
À
lectivity for C H insertion and yield (Table 1, entry 6,
Supporting information for this article is available on the WWW
51%), albeit affording racemic product.[18–19] Since this
68
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 68 – 71