Angewandte
Communications
Chemie
a benzene ring to construct an oxindole skeleton, we initiated
our screenings to generate azaoxyallyl cations from anilides
under mild reaction conditions: a KHCO3/1,1,1,3,3,3-hexa-
fluoro-2-propanol (HFIP) system. It was found that the
substituent on the anilide nitrogen atom is a key factor, as no
reaction occurred when the substituent was H, methyl,
methoxy, or benzyl (Table 1, entries 1–4). To our delight,
Table 1: Optimization of the reaction conditions.[a]
Entry
R
Base
Solvent
Yield [%][b]
1
2
3
4
5
6
7
8
9
H
Me
Bn
KHCO3
KHCO3
KHCO3
KHCO3
KHCO3
Et3N
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
HFIP
THF
n.r.
n.r.
n.r.
n.r.
90
83
52
no
no
no
no
OMe
OH
OH
OH
OH
OH
OH
OH
K2CO3
KHCO3
KHCO3
KHCO3
KHCO3
CH2Cl2
MeCN
toluene
10
11
[a] Reaction conditions: 7 (0.3 mmol, 1.0 equiv) and base (1.1 equiv) in
solvent (2 mL) at room temperature (RT) for 5 h. [b] The yield is that of
isolated product. n.r.=no reaction, no=no desired product 8,
THF=tetrahydrofuran.
Scheme 2. Substrate scope for the reaction. Yield is that of the isolated
product. X=Br. Reaction conditions: 10 (0.3 mmol, 1.0 equiv), and
KHCO3 (1.1 equiv) in HFIP (2 mL) at room temperature (RT) for 5 h.
[a] Reaction time: 12 h. [b] X=Cl. 12 h.
when a-bromo-N-hydroxy anilide (7, R=OH) was subjected
to the system,[20–22] the desired oxindole 8 (R =OH) was
obtained in 90% yield (entry 5). Thus, 7 (R=OH) was chosen
as the model substrate for the initial survey. Replacing KHCO3
with either K2CO3 or Et3N as the base resulted in lower yield
(entries 6 and 7). The solvent was so crucial that the reaction
conducted in common solvents gave the elimination product 9
as the main product without forming 8 (entries 8–11).
subsequent evaporation of the ether solvent and addition of
HFIP. The reaction mixture was then stirred for another
5 hours at room temperature, thus resulting in the desired
product 8a in 68% yield. Because various phenylhydroxyl-
amines and acyl halides are either commercially available or
can be prepared readily, this one-pot procedure is more
practical than the two-step method, albeit in lower, but
acceptable yield (11g, 11q, 11r, 11u–z).
Having identified the optimized reaction conditions, we
next explored the substrate scope. A variety of N-hydroxy
anilides were subjected to the KHCO3/HFIP reaction system
(Scheme 2). Functional groups such as methyl, methoxy,
halogens, ester, nitrile, or hydroxy were found to be well
tolerated in the system (11a–m). The position of substituents
at the para, meta, or ortho positions did not affect the
reaction. The substituted anilides with electron-withdrawing
groups tended to have poor yields (11i,j). Spiro-oxindoles
(11n,o) with five-membered or six-membered rings were
readily prepared in this method. Both 3-monosubstituted and
3,3-disubstituted substrates gave oxindoles in good yields
(11p–t), but the former (11p–s) reacted more slowly (12 h)
than did the latter (11t). Besides a-bromo-N-hydroxy ani-
lides, a-chloro-N-hydroxy anilide (11r) also reacted smoothly.
Since N-hydroxy anilides can be prepared readily by
reacting phenylhydroxylamine (12) with either acyl chloride
or bromide,[23] we turned our attention to testing the
possibility of a one-pot process for the synthesis of 11 and 8
(Scheme 3). In the one-pot process, 2-bromo-2-methylpropa-
noyl bromide (13; R2 = R3 = Me, X1 = X2 = Br) was added
dropwise to a mixture of 12 and KHCO3 in ether, with
Given the versatility of the one-pot process, it was applied
to the synthesis of the progesterone receptor antagonist
5
[24](Scheme 4). The synthesis began with treatment of N-(4-
bromophenyl)hydroxylamine (14) with the acyl bromide 15 in
a one-pot process to give the oxindole 11 f in 55% yield. The
oxindole 11 f was then reacted with the boronic acid 16 under
Suzuki coupling conditions to afford the precursor 17 in 72%
yield.[25] Cleavage of the N O bond of 17 afforded 5 in 81%
À
yield.[26]
It was found that when the substituent on the anilide
nitrogen atom was either H, methyl, benzyl, or methoxy, the
reaction did not afford the desired product 8. In addition, an
attempt with N-benzyl-2-bromo-N-hydroxy-2-methylpropa-
namide (18) under the standard reaction conditions to
prepare the six-membered ring compound 19 gave the
olefin 20 in almost quantitative yield without even trace
amount of 19 formed (Scheme 5). Thus, the reactivity of 7 can
possibly be ascribed to the nitrogen atom attached directly to
the benzene ring. Based on these results and previous
mechanistic studies of azaoxyally cations,[17–21] we hereby
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!