Organic Letters
Letter
a
Table 1. Optimization of the Reaction Condition
b
entry
base
additive
yield (%)
1
2
3
4
5
6
7
8
NaOAc
NaOAc
LiOAc
KOAc
CsOAc
Li2CO3
k2CO3
Cs2CO3
NaHCO3
KHCO3
CsOPiv
K2HPO4
KOAc
−
55%
63%
73%
75%
58%
50%
49%
28%
48%
44%
61%
63%
K2S2O8
K2S2O8
K2S2O8
K2S2O8
K2S2O8
K2S2O8
K2S2O8
K2S2O8
K2S2O8
K2S2O8
K2S2O8
9
10
11
12
13
14
15
Figure 1. Transition metal catalyzed synthesis of the maleimide.
d
K2S2O8, without Bu4NBr
K2S2O8, without KOAc
K2S2O8, without Pd(OAc)2
<5%
<7%
nd
d
−
KOAc
avoided by taking advantage of formic acid23 and DIAD24 as an
organic CO source (Figure 1b). The currently developed
protocol involves palladium-catalyzed C−C activation of
cyclopropenone followed by corbonylative amination to form
the N-heterocyclic maleimide scaffolds. During this process, 0.5
equiv of substrate is used up for the in situ generation of carbon
monoxide. This reaction involves elimination and insertion of
CO onto another palladacycle which is very unique and first of
its kind in the C−C bond activation process. Herein for the first
time, cyclopropenone has been used as a nontoxic CO source in
the presence of a palladium catalyst and can act as a sole carbon
monoxide source which is further confirmed by 18O isotope
studies.
As mentioned earlier our goal was to synthesize azetinones
through a C−C bond activation pathway; accordingly, we chose
diphenylcyclopropenone 1a and aniline 2a as the model
substrates. Initially, the reaction was carried out in the presence
of palladium acetate as the catalyst, tetrabutylammonium
bromide as an additive, and sodium acetate as the base at 120
°C for 12 h in DMF (0.25 M). Surprisingly we obtained the
maleimide product 3aa in 55% yield instead of azetinone
(Table 1, entry 1). As maleimides are also biologically
important molecules we decided to optimize the conditions
further to obtain a good yield of maleimide 3aa. Interestingly,
under the same reaction conditions, the addition of K2S2O8
further improved the yield of the desired product to 63% within
a short reaction time (Table 1, entry 2). Encouraged by this
intriguing observation, several other additives were screened
was found to be the most effective additive increasing the
catalytic efficiency of the reaction. Aiming to enhance the
product yield further, we screened various other acetate bases.
Fruitful results were observed in the case of LiOAc and KOAc
which gave 73% and 75% of 3aa, respectively (Table 1, entries 3
and 4), whereas cesium acetate decreased the yield of 3aa to
58% (Table 1, entry 5). Likewise, carbonate and bicarbonate
bases failed to enhance the yield of the C−C activated product
(Table 1, entries 6−10). To confirm the importance of the base
we performed analogous experiments with CsOPiv and
K2HPO4 (Table 1, entries 11 and 12). In both these cases,
no increment in the product yield was observed. Thus, KOAc
was found to be the best base producing a 75% yield of the
a
Conditions: 1a (1 equiv), 2a (2.5 equiv), Pd(OAc)2 (15 mol %),
b
Bu4NBr (1 equiv), additive (1.5 equiv), DMF as a solvent. Isolated
yields. Reaction time 12 h. GC yield (dodecane was taken as
internal standard for GC).
c
d
desired product which implies that optimum basicity is required
for this pathway. Then we performed the control experiments
in the absence of tetrabutylammonium bromide (TBAB) and
base (KOAc) and observed an insignificant GC yield (Table 1,
entries 13 and 14) which implies that both TBAB and KOAc
are required for the reaction to proceed catalytically. Also it has
been reported in the literature that TBAB helps in the reduction
of Pd(II) to Pd(0).25 Besides, to understand the influence of
the palladium acetate on the titled reaction, we carried out the
reaction in the absence of the catalyst which failed to give the
product 3aa (Table 1, entry 15).
After establishing the optimal reaction conditions, we
evaluated the substrate scope of this Pd-catalyzed cascade C−
C activation and carbonylative amination reaction. At first, we
treated diphenylcyclopropenone 1a with a variety of para-
substituted anilines. We found that with the gradual increase or
decrease in the electron density of the substituent in the para
position, the yields tend toward the lower side. While, p-
toluidine gave a 72% yield of the corresponding maleimide
product (Scheme 1, 3ab), relatively more electron-rich p-OMe
aniline delivered only 41% of the desired product (Scheme 1,
3ac). Similarly, while p-NO2 aniline gave a moderate yield
(65%) of the corresponding product (Scheme 1, 3ad),
relatively less electron-withdrawing substrates such p-CN and
p-CF3 anilines gave better yields, 80% and 71%, respectively
(Scheme 1, 3ae, 3af). Between p-Cl and p-F aniline, the p-Cl
aniline worked well giving a 70% yield (Scheme 1, 3ag) whereas
p-F aniline gave only 64% of the corresponding adduct (Scheme
1, 3ah). Further, we subjected meta substituted anilines to our
reaction conditions and obtained the desired product in 55%−
68% yield (Scheme 1, 3ai−3al). Furthermore, the compatibility
of the ortho substituted anilines has also been tested. A series of
mono- and dialkyl substituted substrates were found to be
compatible indicating the robustness of the reaction toward
steric hindrance (Scheme 1, 3am−3ao). Substrates bearing a
B
Org. Lett. XXXX, XXX, XXX−XXX