Organic Letters
Letter
a
Scheme 2. Strategies for Synthesis of α-CF2R-Substituted
Table 1. Optimization of the Reaction Conditions
Ketones
b
entry
photocatalyst
fac-Ir(ppy)3
2a (equiv)
solvent
yield (%)
1
2
3
4
5
6
7
8
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
2
3
2
2
2
2
2
2
MeCN
DCM
DCE
toluene
DMF
DMAC
THF
DMSO
NMP
NMP
NMP
NMP
NMP
NMP
NMP
NMP
NMP
trace
trace
trace
trace
46
64
74
86
88
89
85
trace
trace
0
fac-Ir(ppy)3
fac-Ir(ppy)3
fac-Ir(ppy)3
fac-Ir(ppy)3
fac-Ir(ppy)3
fac-Ir(ppy)3
fac-Ir(ppy)3
fac-Ir(ppy)3
fac-Ir(ppy)3
fac-Ir(ppy)3
eosin Y disodium salt
4CzIPN
9
10
11
12
13
14
15
16
Ru(bpy)3Cl2
solvent Red 43
0
0
0
Previous reports presented various methods to synthesize α-
CF2R-substituted ketones (R = CO2Et),13f,g which need
multistep processes to form α-CF2H-substituted ketones.
Therefore, we visualized that photoredox catalysts can promote
radical difluoromethylation under mild conditions. Herein, we
report a new and highly efficient synthesis of α-CF2H-
substituted ketones using [Ph3PCF2H]+Br− (difluoromethyl-
triphenylphosphonium bromide) as a radical difluoromethylat-
ing reagent under visible-light photoredox catalysis. Qing’s
group first used fluoroalkylphosphonium salts as the
fluoroalkylating sources.9l−n The fluoroalkylating reagents
could be synthesized simply on a multigram scale from readily
available BrCF2CO2Et.14
We selected enol acetate (1a) as the model substrate and
phosphonium salt (2a) as the difluoromethylating source with
fac-Ir(ppy)3 under blue LED. As shown in Table 1, the solvent
presented a profound effect on the result of the procedure.
Only trace of α-CF2H-substituted ketone 3a was identified
after screening of acetonitrile, DCM (dichloromethane), DCE
(1,2-dichloroethane), and toluene (Table 1, entries 1−4).
Reactions in DMF (N,N-dimethylformamide), DMAC (N,N-
dimethylacetamide), and THF (tetrahydrofuran) provided 3a
in 46%, 64%, and 74% yields, respectively (Table 1, entries 5−
7). To our surprise, the yield of α-CF2H-substituted ketone 3a
was significantly increased upon switching to DMSO (dimethyl
sulfoxide) and NMP (N-methylpyrrolidone) (Table 1, entries
8 and 9). Then we investigated the effect of the amount of 2a.
It was found that 2 equiv of 2a showed the best result, giving
3a in 89% yield (Table 1, entry 10). Subsequently, we
examined four other photocatalysts (entries 12−15); photo-
catalyst eosin Y disodium salt was less effective since only trace
amount of α-CF2H-substituted ketone could be detected and
other photocatalysts could not provide the desired compound.
This reaction could not work either without fac-Ir(ppy)3 or in
the dark (entries 16 and 17). More detailed data refer to the
c
17
fac-Ir(ppy)3
a
Reaction conditions: 1a (0.2 mmol, 1 equiv), 2a, photocatalyst
(0.006 mmol, 3 mol %), solvent (2.0 mL), room temperature, 36 h,
b
c
argon atmosphere, 15 W blue LED. Isolated yield. In the dark.
Reactions of electron-donating groups (OMe, OBn, OAc,
OPh) were compatible with the present reaction system, which
gave the α-CF2H-substituted ketones without significant
decreased yields (3e: 71%, 3g: 76%, 3h: 65%, 3l: 64%). In
contrast, aliphatic enol acetates (e.g., cyclohept-1-en-1-yl
acetate) failed in this reaction conditions (3y). We found
that the substrate cyclohept-1-en-1-yl acetate was still left in
the reaction. Notably, o-phenyl-substituted product could not
be isolated because of self-dehydrofluorination in silica gel
column chromatography.15 Furthermore, dual-substituted
phenyl enol acetates (3m−3p) were all smoothly difluor-
omethylated to afford the desired products in 51−76% yields.
Weaker electron-withdrawing groups (3q−3s), such as F, Cl,
and Br, were all suitable substrates for this process and
generated α-CF2H-substituted ketones in 47%, 55%, and 43%
yields, respectively. However, a strongly electron-withdrawing
NO2 group at the para-position was not tolerated under the
reaction conditions (3t). These results indicated that electron-
withdrawing groups made the efficiency of the reaction
decreased. In addition, cyclic and branched enol acetates
were explored under the reaction conditions. Cyclic enol
acetate could work well, achieving 3u in 61% yield. α-Branched
enol acetate generated the corresponding product 3v in 62%
yield. The scope of this difluoromethylation method was then
tested on heteroaromatic substrates. We found that these
reactions were sensitive to the heteroaryl enol acetates (3w and
3x), giving 50% and 35% yields, respectively.
To examine the utility of difluoromethylation, we performed
a gram-scale reaction and some synthetic transformations of α-
CF2H-substituted ketones (Scheme 4). As shown in Scheme
4a, the gram-scale reaction of 1z and [Ph3PCF2H]+Br−
successfully afforded 1.045 g of 3z in 66% yield. Subsequently,
synthetic applications of 3b were carried out. Scheme 4b shows
the reduction of 3b with NaBH4 in ethanol to give the
corresponding CF2H-containing alcohol 4 in 95% yield. As
With the optimized conditions in hand (Table 1, entry 10),
we then investigated the generality of this radical difluor-
omethylation with numerous aromatic enol acetates (Scheme
3). Initially, various para- and meta-substituted phenyl enol
acetates were subjected to this new catalytic protocol and
furnished α-CF2H-substituted ketones in good yields (3b−3l).
509
Org. Lett. 2021, 23, 508−513