Organic Letters
Letter
a
a
Table 1. Reaction Optimization
Scheme 2. Scope and Limitation of N-Tosyl Vinylaziridines
entry
variation from the standard conditions
yield (%)
b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
standard conditions
88 (85 )
TMEDA instead of DIPEA
K2CO3 instead of DIPEA
K3PO4 instead of DIPEA
NaOAc instead of DIPEA
AgF instead of DIPEA
Ag2CO3 instead of DIPEA
Ru(bpy)3Cl2 instead of fac-Ir(ppy)3
eosin instead of fac-Ir(ppy)3
DMF instead of DMSO
MeCN instead of DMSO
DCE instead of DMSO
CHCl3 instead of DMSO
addition of 1 equiv of Hantzsch ester
addition of 1 equiv of sodium ascorbate
0.6 equiv DIPEA
42
46
55
60
53
46
50
c
ND
52
49
52
messy
cd
,
ND
15
d
40
ND
40
ce
,
without DIPEA
in air
without fac-Ir(ppy)3
in darkness
a
Standard conditions: 1 (0.2 mmol), 2a (0.24 mmol), fac-Ir(ppy)3 (1
mol %), DIPEA (0.24 mmol), DMSO (2 mL), blue LEDs, Ar, rt, 12 h.
Isolated yields after column chromatography.
trace
f
NR
a
Reaction conditions: 1a (0.2 mmol), 2a (0.24 mmol), fac-Ir(ppy)3
(1 mol %), DIPEA (0.24 mmol), solvent (2 mL), blue LEDs, Ar, rt,
1
12 h. Yields were determined by H NMR spectroscopy using 1,3,5-
well with 2a, affording the desired indolin-2-ones (3aa−3ga)
in yields of 65%−85%. However, the sterically hindered 1h
could also be employed as a substrate, only giving the desired
product 3ha in 35% yield. Notably, disubstituted vinyl-
aziridines also proved to be suitable for this transformation,
with 1h giving the desired 3ia product in 65% yield, while the
thienyl-substituted vinylaziridine 1j also worked well in the
reaction, affording the desired product 3ja in a good yield.
Expanding the scope of N-tosyl vinylaziridines from aryl to
alkyl groups (R1 = Me) only gave the corresponding product
3ka in 28% yield. Unfortunately, the use of 1-tosyl-2-
vinylaziridine 1l (R1 = H) resulted in no reaction product
being observed.22 Besides, N-tosyl vinylaziridine 1m (R3 = Ph)
and internal olefins (1n and 1o) were not suitable for this
system.
Encouraged by the above results, we investigated the scope
of difluoroalkyl halides with 2-(1-phenylvinyl)-1-tosylaziridine
1a. As shown in Scheme 3, this catalytic system showed a
broad substrate scope and high functional group tolerance
concerning difluoroalkyl halides. As expected, 2-iodide-2,2-
difluoroacetate and methyl 2-bromo-2,2-difluoroacetate also
worked well, yielding the desired 3aa and 3ab in 88% and 60%
yields, respectively. Various types of bromodifluoroacetamide
derivatives were examined, affording the corresponding
pyridines (3ac−3am) in yields of 41%−95%. Unfortunately,
the reactions of N-tosyl vinylaziridine 1a and difluoroalkyl
bromides (R = PhCO, EtCO) does not proceed under
standard conditions. Significantly, bromodifluoroacetamide
amino acids can serve as versatile building blocks for diversified
transformations, giving the desired pyridines (3an−3ar) in
yields of 44%−81%. Bromodifluorooxadiazole also worked
well, yielding the desired 3as in 82% yield, and using
dibromofluoromethane instead of difluoroalkyl halides, the
desired pyridine 3at was obtained in 45% yield. When N-aryl
b
trimethoxybenzene as the internal standard. Yield of the isolated
product after column chromatography. ND = none detected. Self-
c
d
e
cyclization product of 1a was observed. The intermediate 3aa′ was
f
1
observed in the crude H NMR. NR = no reaction.
DIPEA was the most effective adjuvant (Table 1, entries 1−7).
The photocatalyst choice was critical in the transformation.
For example, 3aa was obtained in 50% yield using Ru(bpy)3Cl2
instead of fac-Ir(ppy)3; however, no product 3aa was observed
using Eosin Y as the photocatalyst (Table 1, entries 8 and 9).
Solvent screening revealed that the reaction proceeded
smoothly in other solvents, such as dimethylfluoride (DMF),
acetonitrile (MeCN), dichloroethane (DCE) (Table 1, entries
10−12). However, a messy reaction was found using CHCl3 as
the solvent (Table 1, entry 13). The byproduct 3-phenyl-1-
tosyl-2,5-dihydro-1H-pyrrole, which is generated from the self-
cyclization of 1a, was observed obviously when the Hantzsch
ester or sodium ascorbate was added (Table 1, entries 14 and
15). Notably, the yield of 3aa declined to 40% when using 0.6
equiv of DIPEA (Table 1, entry 16). To our delight, 3aa′ was
1
observed in the crude H NMR without DIPEA (Table 1,
entry 17). In addition, conducting the reaction in air resulted
in a yield of only 40% for 3aa (Table 1, entry 18). As expected,
control experiments verified the necessity of light irradiation
and the photocatalyst for the current transformation (Table 1,
entries 19 and 20).
Upon optimizing the reaction conditions, the scope and
limitations of this reaction, with respect to N-tosyl vinyl-
aziridines, were investigated with ethyl 2-bromo-2,2-difluor-
oacetate 2a (Scheme 2). To our delight, a series of
representative aryl N-tosyl vinylaziridines (at the R1 site)
containing either electron-donating (OMe and Me) or
electron-withdrawing (F, Cl, Br, and Ph) groups all worked
B
Org. Lett. XXXX, XXX, XXX−XXX