Organic Letters
Letter
be a far more efficient quencher (see SI for more details). In
addition, a reaction between 1a and I in the presence of
TEMPO under otherwise identical conditions resulted after 24
h in the formation of only 5% of the α-difluoromethylated
ketone 2a along with 53% of the TEMPO−CHF2 adduct
(Figure 2B). These data strongly support an oxidative
quenching pathway, resulting in a one-electron reduction of I
and subsequent decomposition releasing the CHF2 radical.
We also studied the radical addition onto the silyl enol ether
by DFT using the PBE0 functional with Grimme D3
dispersion corrections, SMD solvent model for acetone as
implemented in Gaussian 16,25 and the trimethylsilyl analogue
of 1a, 1o, and 1q as a conformationally simplified model
substrate. Our model (Figure 2D) reflects a highly exergonic
addition (ΔG = −34.0 kcal/mol) mediated by an early, readily
accessible transition state (ΔG⧧ = 7.4 kcal/mol, 2.58 Å C−
CHF2 distance). These data are consistent with a rapid and
irreversible radical addition. Furthermore, our model correctly
anticipates the decrease in reactivity caused by the presence of
α-substituents (e.g., propiophenone- and isubutyrophenone-
derived silyl enol ethers 3o and 3q) due to less accessible
transition states. The effect of the substituents at the α position
seems to relate to an intramolecular steric repulsion that causes
a loss of planarity of the aromatic ring and consequent loss of
conjugation. This is consistent with the good yield obtained
experimentally with 1p, in which the α-substituent forms a
bicyclic system with the aromatic ring, thus maintaining a
coplanar arrangement with the enol. Finally, our calculations
predicted that the ketyl radical B resulting from the addition
was a reasonably good reductant, with a reduction potential for
its corresponding silyloxonium (C) of EC = −0.23 V (all
potentials reported vs SCE).26 This enables its easy oxidation
by the oxidized form of the catalyst Ir(ppy)3+ (EIr+ = +0.77 V),
closing the photoredox catalytic cycle. In sharp contrast, C is
not strong enough of a reductant to propagate a chain reaction
through single electron transfer with compound I (EI = −2.26
V). This was further supported by an intermittent irradiation
experiment, which showed the reaction only to occur during
irradiation and abruptly stop in the absence of light (see SI for
more details).
chemistry. A first experiment run with 1i (0.1 mmol scale)
under our previously optimized conditions [2 equiv of I, 2 mol
% of fac-Ir(ppy)3, 2 equiv of H2O, acetone, rt, 8 W blue LED
(460 nm), 10 mL PFA coil (Ø = 1 mm), 0.1 mL.min−1 flow
rate (residence time = 1.7 h)], afforded the corresponding α-
difluoromethyl ketones 2i in 34% NMR yield, which compared
favorably with the result obtained under the standard batch
conditions over 24 h. After fine-tuning the catalyst loading (5
mol % instead of 2 mol %) and the reaction concentration
(0.05 M instead of 0.1 M), we were able to maintain a low
reaction time and reach up to 91% yield. These conditions
were eventually applied to convert 1 mmol of 1i in only 5 h
and 90% isolated yield (Figure 2C).
In summary, we have developed a straightforward method
for the direct difluoromethylation of enol silanes through a
visible-light-driven photoredox catalytic process. The method
affords the corresponding α-CHF2 substituted ketones in good
to excellent yields under mild conditions. Moreover, the
method is practical and readily scalable under continuous flow
conditions. Most importantly, these α-difluoromethylated
ketones are useful building blocks, which have already been
shown to undergo various postfunctionalizations such as α-
brominations, asymmetric reductions, hydrations, acetal
dehydrofluorinations, halohydrin formations, and amida-
tions,18−20 thus offering great perspectives for medicinal
chemistry applications.
ASSOCIATED CONTENT
* Supporting Information
■
sı
The Supporting Information is available free of charge at
1
Details of experimental procedures, H, 13C, and 19F
NMR spectra, HPLC chromatograms, and detailed
AUTHOR INFORMATION
Corresponding Author
■
Stellios Arseniyadis − Department of Chemistry, Queen Mary
University of London, London E1 4NS, U.K.; orcid.org/
We therefore propose the following mechanism (Figure 2E),
where the excited *Ir(ppy)3 catalyst transfers an electron to I,
promoting its decomposition to the CHF2 radical in
conjunction with N-tosylsulfimate, which may be subsequently
protonated. Then, addition of the electrophilic CHF2 radical
onto the silyl enol ether affords the corresponding ketyl radical
Authors
Elias Selmi-Higashi − Department of Chemistry, Queen Mary
University of London, London E1 4NS, U.K.
Jinlei Zhang − Department of Chemistry, Queen Mary
University of London, London E1 4NS, U.K.
+
A, which is in turn oxidized by Ir(ppy)3 to generate the
silyloxonium species B. The latter ultimately undergoes
desilylation to afford the desired ketone product 2a, with the
base acting as a proton acceptor. Alternatively, α-deprotona-
tion can also occur to afford the difluoromethylated silyl enol
ether 2′a, which was observed in several cases and could be
isolated (see SI for more details). Nonetheless, 2′a can be
quantitatively hydrolyzed to the desired ketone upon workup
with TFA (4 equiv).
With the aim of scaling up the reaction, we immediately
became interested in developing a continuous flow process.27
Indeed, the narrow width of the flow reactor coil would ensure
a more uniform distribution of the light within the entire
reaction mixture, which would in turn increase the effective
concentration of active catalyst and therefore result in shorter
reaction times and improved scalability, not to mention the
reduced safety concerns that are usually associated with flow
Xacobe C. Cambeiro − School of Science, University of
Greenwich, Chatham Maritime ME4 4TB, U.K.;
Complete contact information is available at:
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge the China Scholarship Council
(CSC) and Queen Mary University of London for financial
support. This work is dedicated to Professor Christian Bruneau
(Rennes Institute of Chemical Sciences, Université de Rennes
1) for his outstanding contribution to catalysis.
4242
Org. Lett. 2021, 23, 4239−4243