Organic Letters
Letter
phosphine additives P(OEt)3 under the same reaction
conditions, but the results are not satisfactory for these cases
(Table 1, entries 21−24). Consequently, it could be concluded
that the best reaction conditions for the construction of
aposafranone 3a were Cs2CO3 as the base in DMSO at room
temperature.
Scheme 3. Substrate Scope of 4a
With the optimal conditions in hand, we then extended the
reaction to a range of readily available enaminones 1 (Scheme
2). First, we tested the substrate scope of N-substituted 3-
a
Scheme 2. Aposafranones 3 Synthesis
However, when the reaction temperature was enhanced at 40
°C, we found that the reaction completed quickly and obtained
the desirable aposafranone N-oxide 5a with complete
regiocontrol in low yield (7%).
To increase the yield, various reaction conditions were
screened. Initially, we continued to enhance the temperature of
the reaction from 50 °C to 110 °C, the yield increased from
20% to 37%, accompanied by a reduction in the yield of 6 from
35% to 20%. Next, we controlled the sequence of initial
substrate addition, with enaminone 4a and Cs2CO3 added first
and stirred for 1 h at 110 °C, and then 2,6-difluoronitro-
benzene 2b was added to the mixture. Excitingly, this manner
led to the formation of aposafranone N-oxides 5a at a 47%
yield with a trace amount of side product 6 (Scheme 3B).
Further increasing the reaction temperature to 120 °C and 140
°C, the desired product 5a gave yields of 47% and 44%,
respectively. These results therefore showed that the best
reaction temperature was 110 °C. According to the results
shown in Table 1, the better solvents (DMF, CH3CN, 1,4-
dioxane, toluene) and bases (K2CO3, KOH) were screened
next at 110 °C. The results suggest that DMSO and Cs2CO3
continued to be the optimal solvent and base. Thus, it can be
concluded that the optimal reaction conditions for the
preparation of aposafranone N-oxides were 1.2 equiv Cs2CO3
as the base at 110 °C in DMSO.
Under these optimal reaction conditions, we turned our
attention to investigating the cascade reaction for preparing the
other target molecules 5, and the series of aposafranone N-
oxides 5a−5r were synthesized with moderate yields (Scheme
4). Similar to the results of Scheme 2, the reaction was
tolerated when 4 aromatic rings bearing an electron-with-
drawing or electron-deficient group at different positions and
afforded comparable yields. It was observed that the
substituents (R′ = F) on substrates (2) had some influence
on the yields of 5, affording lower yields than those with no
substitutions.
a
The reaction was performed with 1 (0.5 mmol), 2 (0.6 mmol), and
Cs2CO3 (0.6 mmol), in DMSO (3 mL) at room temperature for
8−12 h; isolated yields are based on enaminones 1. One mmol scale.
b
aminocyclohex-2-enonephenyls 1a-1j. To our delight, sub-
strates bearing either electron-rich or electron deficient
substituents, such as halogen, methyl and methoxy, at different
positions (meta-, ortho-, para-) of the rings all generated
corresponding aposafranone products 3a−3j with moderate to
good yields. The results indicated that the electronic effects
and steric hindrance did not have much impact on the
reaction. Subsequently, we also evaluated the scope of the o-
fluoronitrobenzenenes. The 2,6-difluoronitrobenzene 2b was
used as the substrate under the optimized conditions and also
afforded the target products 3j, 3k, and 3l at yields of 46%,
45%, and 53%, respectively. Disappointingly, the target
molecule 3 was difficult to purify. In the process of column
chromatography, part of the products will be adsorbed on the
silica gel, which further converted to some larger polar and
difficult-separated materials, and consequently, the yield of 3
was only moderate.
We next extended the scope of the reaction by using N-aryl
3-amino-5,5-dimethylcyclohex-2-enones 4. Under the optimal
conditions in Table 1, we first tested the reaction of 3-amino-
5,5-dimethylcyclohex-2-enone 4a with 2,6-difluoronitroben-
zene 2b. Disappointingly, this reaction only produced N-aryl
enaminone 6 with 41% yield, because of the larger steric
hindrance on the C5-position of substrate 4 (Scheme 3A).
The chemical structures of 3 and 5 were identified from
1
their IR, H NMR, 13C NMR, and HRMS spectra, and were
supported by single crystal X-ray diffraction analysis of
compound 5l (Figure 2).
To support the proposed mechanism, intermediate 6 was
employed under the standard conditions; however, the
aposafranone N-oxides 5a was not generated and no other
transformation occurred in this reaction (Scheme 5). The
results of Schemes 5 and 3B verified that the key step for this
reaction should be the formation of the enamine-enol oxygen
3034
Org. Lett. 2021, 23, 3032−3037