Organic Letters
Letter
a hurdle. To overcome these limitations and meet the high
demand of discovering new CF2-containing bioactive mole-
cules, a radical transformation would be an attractive strategy,
as the radical reactions usually are not sensitive to the steric
effect.11 For practical applications, we sought to use
inexpensive and widely available aliphatic carboxylic acids as
the starting materials, which can be easily transformed into N-
(acyloxy)phthalimides.12 These redox esters can efficiently
generate alkyl radicals via a single electron transfer (SET)
pathway in the presence of photocatalyst.13
We hypothesized that once the alkyl radical A was generated
between N-(acyloxy)phthalimide and the excited photo-
catalyst, the difluoroalkylated compound would be efficiently
accessed with a suitable radical-trapping difluoroalkylating
reagent, such as difluoroenoxysilane previously used for the
visible-light-initiated difluoroalkylation of arene diazonium
salts14 (Scheme 1C). In this process, the newly generated
alkyl radical B was oxidized by the photocatalyst to form a
carbon cation species C, which subsequently underwent
desilylation to produce the difluoroalkyl ketone. This radical
strategy would enable difluoroalkylation of an array of
unactivated primary, secondary, and tertiary aliphatic sub-
strates, thus providing a general approach to access
difluoroalkylated alkanes.
DMA and DMSO were comparable with DMF, and CH3CN
was slightly less effective (entry 5, for details see the
almost suppressed the reaction (entry 6). Decreasing the
reaction temperature to room temperature slightly diminished
the reaction efficiency with an 85% yield of 3a obtained (entry
7). Essentially, no reaction occurred in the absence of
photocatalyst or blue light irradiation (entries 8 and 9),
demonstrating that the [Ir]-catalyst and blue light play an
essential role in promoting the reaction.
With the viable reaction conditions in hand, we examined
the difluoroalkylation of a variety of redox-esters 1. Overall, the
reaction was not sensitive to the steric effect. Good to high
yields were obtained with primary, secondary, and tertiary alkyl
carboxylic esters. As illustrated in Scheme 2, cyclic secondary
Scheme 2. Photoredox-Catalyzed Decarboxylative
a
Alkylation of Difluoroenoxysilanes 2
Accordingly, we chose aryl-substituted difluoroenoxysilanes
2 as the difluoroalkylating reagents. Due to the electron-
withdrawing effect of the fluorine atom, these electron-
deficient alkenes are favorable to reacting with a nucleophilic
alkyl radical. We began our studies from the reaction of
secondary alkyl carboxylic redox-ester 1a with p-toyl
substituted difluoroenoxysilane 2a under photoredox reaction
conditions (Table 1). We found that the use of commercially
Table 1. Representative Results for Optimization of
Photoredox-Catalyzed Decarboxylative Alkylation of
a
Difluoroenoxysilane 2a with 1a
b
entry
reaction conditions
Standard conditions
[Ir(dtbbpy)(ppy)2]PF6 instead of fac-Ir(ppy)3
Ir[dF(CF3)ppy]2(dtbbpy)PF6 instead of fac-Ir(ppy)3
Ru(bpy)3(PF6)2 instead of fac-Ir(ppy)3
CH3CN instead of DMF
1,4-Dioxane instead of DMF
reaction run at room temperature
w/o blue LED
yield (%), 3a
1
2
3
4
5
6
7
8
9
94 (94)
76
26
8
87
10
85
0
a
Reaction conditions (unless otherwise specified): 1 (0.5 mmol, 1.0
equiv), 2 (1.5 mmol, 3.0 equiv), DMF (5 mL) at 60 °C for 48 h.
b
Gram-scale reaction: 1 (4 mmol, 1.0 equiv), 2 (12 mmol, 3.0 equiv),
DMF (40 mL) at 60 °C for 48 h.
alkyl redox esters furnished the corresponding products
efficiently (3a−3d). High yields were obtained with piper-
idine-, tetrahydropyran-, and cyclohexane-derived substrates
(3a−3c). A four-membered azetidine derivative was also
applicable to the reaction, and an even higher yield (98%) was
obtained by gram-scale synthesis (3d). Since CF2 and azetidine
are pharmaceutically relevant structural motifs, the current
process may have potential applications in medicinal
chemistry. The reaction was not restricted to the cyclic
substrates, as the acyclic secondary alkyl redox esters
underwent the difluoroalkylation smoothly (3e and 3f). An
unactivated secondary heptyl substituted ester provided the
corresponding product 3e with high efficiency. A phenyl-
alanine-derived substrate was also a competent coupling
partner with an 81% yield obtained (3f). The generality of
w/o fac-Ir(ppy)3
3
a
Reaction conditions (unless otherwise specified): 1a (0.5 mmol, 1.0
b
equiv), 2a (1.5 mmol, 3.0 equiv). Determined by 19F NMR using
fluorobenzene as an internal standard and number in parentheses is
isolated yield.
available photocatalyst fac-Ir(ppy)3 (1 mol %) in DMF under
blue light (12 W, 460−465 nm) irradiation could provide the
desired product 3a in 94% yield upon isolation (entry 1). A
slightly lower yield was obtained with [Ir(dtbbpy)(ppy)2]PF6
as the catalyst (entry 2). However, other photocatalysts,
Ir[dF(CF3)ppy]2(dtbbpy)PF6 and Ru(bpy)3(PF6)2, led to
poor yields (entries 3 and 4). Among the tested solvents,
B
Org. Lett. XXXX, XXX, XXX−XXX