Organic Letters
Letter
direct C(sp3)−H bond functionalization.10 These reactions
were mediated by the generation of carbon-,11 nitrogen-,12 and
oxygen-13 centered radicals through photocatalysis.
toluene) were investigated, but no improvement of the
reaction yield was observed (entries 9−12). In addition, the
catalyst loading could be decreased to 0.1 mol % without much
influence on the reaction yield, albeit an increased reaction
time (48 h) was needed (entry 13).
With the optimized reaction in hand, we first tested the
scope of amides 1a−1n (Scheme 2). Amides bearing a δ-
Knowles,14 Rovis,15 and Meggers16 successfully applied this
methodology to amide substrates coupling to alkenes. These
reactions generally underwent a photoredox-catalyzed oxida-
tion of the pendent amide moiety and a 1,5-HAT process
followed by C-functionalization with Michael acceptors
(Scheme 1B). Despite the notable achievement, the formation
of an amidyl radical from the amide C−H bonds and
applications in the intermolecular transformation of C(sp3)−
H bonds are still scarce.17 Herein we develop a direct gem-
difluoroallylation of inert C(sp3)−H bonds in amides with α-
trifluoromethyl alkenes (Scheme 1C). Mediated by a photo-
catalysis-triggered 1,5-HAT process, gem-difluoroalkenyl-
group-functionalized amides are synthesized in good yields
with high functional group tolerance and high δ-selectivity.
At the outset of this project, we adopted the trifluor-
oacetamide (1a) and α-trifluoromethyl alkene (2a) as the
standard substrates (Table 1). More detailed reaction
a
Scheme 2. Scope with Respect to the Amide
a
Table 1. Variation of Reaction Parameters
b
entry
variation from standard conditions
yield (%)
c
1
none
89 (86)
2
without light
0
3
4
air instead of N2
without 4CzIPN
56
0
5
6
7
Ru(bpy)3Cl2 instead of 4CzIPN
eosin Y instead of 4CzIPN
0
0
0
MesAcr+CIO4 instead of 4CzIPN
−
8
9
10
11
12
13
Ir[dF(CF3)ppy]2(dtbbpy)PF6 instead of 4CzIPN
K2CO3 instead of K3PO4
CS2CO3 instead of K3PO4
DCM instead of MeCN
toluene instead of MeCN
45
54
37
61
13
80
a
Reaction conditions: 1a (0.20 mmol, 1.0 equiv), 2a (0.20 mmol, 1.0
equiv), 4CzIPN (1 mol %), K3PO4 (0.40 mmol, 2.0 equiv), and
0.1 mol % of catalyst for 48 h
MeCN (2.0 mL) under N2 and stirred at rt for 24 h under 30 W blue
LED irradiation. Isolated yields.
a
b
Reaction conditions: 1a (0.10 mmol, 1.0 equiv), 2a (0.10 mmol, 1.0
equiv), 4CzIPN (1 mol %), K3PO4 (0.20 mmol, 2.0 equiv), and
MeCN (1.0 mL) under N2 and stirred at rt for 24 h under 30 W blue
LED irradiation. Yields determined by GC using n-tetradecane as the
b
c
internal standard Isolated yield.
methine carbon were examined at the beginning, and
satisfactory results were obtained (3a−3d). Additionally,
amides bearing a methylene δ-carbon were also functionalized
using this approach in moderate yields, and the δ-selectivity
was retained (3e−3k, 3m, and 3n). For the substrate with
multiple functionalizable sites, only the δ-functionalized
product was selectively obtained (3g). Substrate 1k, with an
oxygen atom adjacent to the δ-C−H bond, reacted smoothly in
this reaction and generated 3k. With one more carbon
substrate 1m, both 1,5- and 1,6-HAT products could be
obtained. What should be of concern is that the primary δ-C−
H bond in substrate 1l that bears an α-substituent to nitrogen
was also competent in the reaction, albeit in slightly lower yield
(3l). The alkene also could be tolerated under the standard
reaction conditions to give the product 3n. Last but not the
least, to demonstrate the synthetic potential of this trans-
formation, the amide substrate derived from steroid 1o was
tested, and the desired product (3o) was obtained in satisfying
yield with high selectivity.
Tables S1−S4. After much effort, the standard conditions were
successfully established: A mixture of 1a (0.1 mmol, 1 equiv)
and 2a (0.1 mmol, 1.0 equiv) in MeCN was irradiated with a
30 W blue LED in the presence of 1 mol % of 2,4,5,6-
tetrakis(carbazol-9-yl)-1,3-dicyanobenzene (4CzIPN); K3PO4
(0.2 mmol, 2.0 equiv) was used as the base, and the reaction
was carried out at room temperature. The desired gem-
difluoroallylation product (3a) was isolated in 86% yield as a
single regioisomer (Table 1, entry 1). Control experiments
indicated that the base, light, and inert atmosphere are vital for
this reaction (Table 1, entries 2−4). Photoredox catalysts, such
as Ru(bpy)3Cl2 (entry 5), eosin Y (entry 6), and
MesAcr+ClO4 (entry 7), were unreactive in this reaction
−
system. Ir[dF(CF3)ppy]2(dtbbpy)PF6 showed good catalytic
reactivity but afforded the product in lower yield (entry 8).
Other bases (K2CO3 and Cs2CO3) and solvents (DCM and
4458
Org. Lett. 2021, 23, 4457−4462