Communication
these mild conditions. Notably, heterocyclic product 3h, which
bears a pyridine moiety, could be isolated in an excellent 94%
yield. Furthermore, the benzothiophene-substituted alkene 2p,
which does not bear an additional aryl group, also delivered
the corresponding product 3p in 58% yield in a 25:75 E/Z
ratio favoring the Z-isomer. Unfortunately, all alkyl-substitut-
ed alkenes, such as 2q and 2r, which are less capable of stabi-
lizing radical intermediates, were not suitable substrates for
this transformation.
Mechanistic experiments were also conducted to investigate
the initial visible-light-promoted step, in which the SCF radical
3
is generated. In our previous studies on the electrophilic tri-
fluoromethylthiolation, halide catalysts were found to activate
the reagent 1 and we envisaged a similar scenario occurring in
this process. Although Stern–Volmer luminescence quenching
studies showed that reagent 1 can act as a highly efficient
[16]
quencher of fac-[Ir(ppy) ], conducting the reaction without
3
[
17]
a halide source did not lead to product formation. Further-
1
19
In order to gain insight into the reaction mechanism,
a series of experiments were conducted. Firstly, addition of the
radical trapping agent TEMPO (2 equiv) led to a shutdown of
the reaction (3% NMR yield), supporting the proposed involve-
ment of radical intermediates. Observations made during the
reaction optimization also imply that benzylic cations may be
formed. For example, conducting the process with 1,1-diphe-
nylethylene (2a) using DMSO as solvent led to a mixture of
alkene 3a (22% NMR yield) and the corresponding alcohol 4
more, no change in the H or F NMR spectra was observed
upon irradiating reagent 1 under the reaction conditions with-
out nBu NBr or nBu NCl. These results imply that direct
4
4
quenching of the excited photocatalyst by 1 is not productive
and that the SCF –halide species C formed in small amounts
3
[
18]
may instead be the source of the SCF radical (Scheme 3).
3
Oxidative quenching with this species would deliver fac-
+
[Ir(ppy)3] and a radical anion amenable to mesolysis, to afford
the SCF radical and regenerate the halide (see the SI for de-
3
(
65% NMR yield; Figure 1). The latter compound, which was
tails of mechanistic experiments).
The implication that cationic intermediate B is generated
during these processes led us to consider whether alternative
reaction pathways could be accessed with suitably substituted
substrates. In particular, alkene derivatives bearing 1-hydroxy-
cyclobutyl groups have been shown to undergo a semi-pina-
col-type rearrangement process through a radical-polar cross-
[
19]
Figure 1. Structures of side products 4 and 5.
over mechanism under photoredox conditions. Such a trans-
formation would deliver novel cyclic ketone derivatives featur-
3
ing a quaternary center and a C(sp )ÀSCF bond. In order to
3
[
19]
not formed during a control reaction conducted under the
same conditions with 3a as substrate, likely results from nucle-
ophilic trapping of a benzylic cation by the solvent. Similarly,
the chlorinated compound 5, resulting from trapping by the
chloride, was observed when 2-vinylnaphthalene was em-
avoid competing epoxide formation, a one-pot procedure
was envisaged, in which substrates 6 would be first silyl-pro-
tected using trimethylsilyl trifluoromethanesulfonate (TMSOTf)
and subsequently trifluoromethylthiolated using our dual cata-
lytic system. An optimization study led to the set of reaction
conditions shown in Scheme 4a (see the SI for details). The
most notable change involved switching from bromide to
chloride as the halide co-catalyst due to competitive bromina-
ployed as substrate using nBu NCl as the halide source. These
4
observations point towards a radical-polar crossover mecha-
nism of the type shown in Scheme 3. After formation of the
SCF radical, addition to the alkene would deliver the alkyl radi-
tion of the double bond using nBu NBr. Under these condi-
3
4
cal species A, which is stabilized by the aryl or heteroaryl
group in substrates 2 seemingly required for the successful
tions, substrate 6a was transformed into the desired trifluoro-
methylthiolated cyclic ketone 7a in 62% yield after 24 h at
room temperature under irradiation from blue LEDs. As with
the previous reaction, a range of functional groups, including
ethers, and aliphatic and aromatic moieties, were tolerated de-
livering the corresponding products 7 in moderate to good
yields (up to 73%). These compounds could serve as building
blocks for the synthesis of a range of trifluoromethylthiolated
structures through further functionalization of the carbonyl
group or, in the case of 7d, cross-coupling reactions.
conversion. Oxidation by the oxidized photocatalyst fac-
+
[
Ir(ppy)3] would then generate the corresponding aryl-stabi-
lized cation B, affording the trifluoromethylthiolated alkene
product 3 upon deprotonation.
Finally, the ability of dual photoredox/halide catalysis to fa-
cilitate another radical-polar crossover trifluoromethylthiolation
process was tested. Reacting the methacrylamide derivative 8
under the standard conditions with 1 mol% fac-[Ir(ppy) ] and
3
nBu NBr in acetonitrile led smoothly to the oxindole product 9,
4
resulting from a trifluoromethylthiolation/cyclization sequence
in 81% yield (Scheme 4b). This reaction complements a previ-
ous report on the synthesis of these heterocyclic alkyl–SCF3
compounds using AgSCF and three equivalents of K S O in
Scheme 3. Mechanistic hypothesis. Phthal=phthalimide, counterions omit-
3
2
2
8
ted for clarity.
the presence of hexamethylphosphoramide (HMPA; 0.5 equiv)
Chem. Eur. J. 2016, 22, 4395 – 4399
4397 ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim