Organic Letters
Letter
product. In this reaction, triethylamine plays a crucial role as a
ligand in accelerating the initial copper-catalyzed ring-opening
reaction to produce these key intermediates in higher
concentrations, enabling the facile second self-cyclization
reaction to produce pyridines. We were intrigued by these in
situ-generated skipped yne−imine and yne−enamine inter-
mediates in this reaction and speculated that if the second self-
cyclization reaction between these two tautomers could be
interrupted, it might be possible to trap these intermediates
with a suitable third component and thereby develop an
unprecedented interrupted azirine−alkyne ring-expansion
strategy for the synthesis of useful N-heterocycles other than
pyrroles.
On this basis and inspired by the alkyne oxygenation
reactions,8 we envisaged that an intermolecular oxygenation of
the alkyne unit of these tautomers and intramolecular
cyclization cascade strategy could facilitate the synthesis of
valued pyrrolone compounds (Figure 1c). Notably, this would
represent a first example of a ring-expansion reaction of
azirines with alkynes for the synthesis of pyrrolone structures.
However, in order to accomplish this goal, we would have to
solve the challenging chemoselectivity issue of achieving the
complete azirine−alkyne ring-opening reaction for the
generation of the skipped yne−imine and yne-enamine
intermediates while fully suppressing the self-cyclization
reaction between these intermediates. In this report, we
describe how we were able to address these challenges to
develop a new three-component interrupted ring-expansion
reaction of azirines with alkynes and molecular oxygen for the
synthesis of highly valued 3-pyrrolones enabled by copper
catalysis.
We started our study by investigating 3-phenyl-2H-azirine
and phenylacetylene (2a) as model substrates with molecular
oxygen as the third component and acetonitrile as the solvent
was tested under these modified oxidative conditions, in
addition to the alkyne homocoupling product, alkynylated
pyridine was isolated as the major product (Table S1, entry 1).
In contrast, when inorganic bases such as Na2CO3, K2CO3,
Cs2CO3, and NaOH were employed, as expected, the pyridine
product was completely suppressed, but the desired oxy-
genated product was not observed (Table S1, entries 2−5). In
all of these cases, the starting azirine was completely
consumed, and homocoupling products were isolated.
Reactions performed at a low temperature of 5 °C resulted
only in poor conversion of the azirine, and the desired product
was not observed (Table S1, entry 6). The same result was
observed when less base was used (Table S1, entry 7).
We next moved on to investigate the more hindered diaryl
azirine 1a under these modified oxidative conditions (Table 1).
When TEA was tested as the base, 1a underwent complete
conversion and delivered the desired oxygenated product, 3-
pyrrolone 3a, in 84% yield (Table 1, entry 1). Consistent with
our mechanistic hypothesis, the presence of an additional
aromatic ring makes the self-cyclization reactions between the
skipped yne−imine and yne−enamine intermediates very
difficult and allows for the second radical oxygenation step
to form the desired pyrrolone products. The structure of 3a
was determined by NMR spectroscopic analysis and X-ray
crystallography of its derivative 3s (Figure S1), which
confirmed that oxygen is incorporated into the product.
Lower yields were observed when DIPEA and DBU were used
instead of TEA (Table 1, entries 2 and 3). Especially with
Table 1. Ring Expansion Reaction of Diaryl-Substituted
Azirine
a
b
b
entry
base
solvent
ACN
ACN
yield of 3a (%) yield of 1,3-diyne (%)
1
2
3
4
5
6
7
8
TEA
DIPEA
DBU
TEA
TEA
TEA
TEA
TEA
TEA
84
45
7
10
53
45
12
55
58
6
15
60
38
31
15
18
4
ACN
toluene
1,4-dioxane
DCE
DCM
DMSO
ACN
c
9
12
a
Conditions: 1a (0.52 mmol), 2a (0.6 mmol), base (1.04 mmol), CuI
b
(0.05 mmol), ACN (3 mL), O2 atmosphere, 25 °C, 24 h. Isolated
yields. The reaction was performed under air.
c
DBU, the alkyne homocoupling product was predominant and
was obtained in 60% yield. A quick survey of solvents revealed
that ACN is best for this transformation (Table 1, entries 4−
8). When the reaction was performed under open air
atmosphere, product 3a was obtained in a lower yield of
58% (entry 9). Other copper(I) sources like CuOAc, CuBr,
and CuCl were also tested under the present reaction
conditions. Surprisingly, all of the above-mentioned copper(I)
salts failed to catalyze the reaction (Table S3, entries 1−3). To
check the role of the anion associated with copper, reactions
were performed with a stoichiometric amount of KI as an
additive, but no improvement was noticed (Table S3, entries
4−6). Also, when Cu(II) salts such as Cu(OAc)2 and CuCl2
were tested in place of CuI, no conversion of the azirine to the
desired product was observed (Table S3, entries 7 and 8).
Thus, the optimal conditions were found to be the stirring of
1a with 2a (1.1 equiv), CuI (0.1 equiv), and TEA (2 equiv)
with an O2 balloon in ACN at room temperature (25 °C) for
24 h (Table 1, entry 1).
With the optimized conditions in hand, initially the scope of
the reaction was explored with a variety of 2,3-diaryl azirines
and the corresponding pyrrolone products 3a−g were isolated
in 56−85% yield (Figure 2). Azirines with phenyl rings
containing halogen substituents generally gave better yields
(3d−g, 80−83%). However, with the dimethyl-substituted
phenyl ring, a significant drop in yield was observed (3c, 56%).
Notably, when an azirine with a strong electron-donating
group, namely, 3-(4-methoxyphenyl)-2-phenyl-2H-azirine, was
employed with the optimized reaction conditions, the reaction
did not proceed, and more than 80% of the starting azirine was
recovered after the reaction (Scheme S7). These results
indicate that the electronics of the azirine aryl rings plays a
huge role in the reaction outcome. Next, aryl alkynes having
various substituents on the aromatic ring, including electron-
withdrawing and electron-donating functional groups, were
tested in this copper-catalyzed reaction cascade. The reaction
efficiency was not very sensitive to the electronic properties of
the substituents, and the corresponding 3-pyrrolones were
formed in good to high yields. Notably, substrates with a bulky
tert-butyl group or n-pentyl substituent were also well-tolerated
and afforded the corresponding products 3j and 3l in high
yields (84% and 82%, respectively). Aryl alkynes with fluorine
4220
Org. Lett. 2021, 23, 4219−4223