Angewandte
Communications
Chemie
typically, the TFT salts gave higher yield (see Supporting
Information, Figure S1). Based on the higher reduction
in the sense that hydrodefunctionalized sideproducts are
formed with Cu O; in such cases copper(I) thiophene-2-
2
0
+
potential of arylthianthrenium salts [E (ArTT /ArTTC) ffi
carboxylate (CuTC) could be used successfully to circumvent
the problem to give the desired phenols (5, 10, 12, 14, 19, 23)
in higher yields.
We could generalize the transformation beyond phenol
synthesis to also include the formation of ethers with phenols
as well as primary and secondary alcohols (Table 2). The use
ꢀ
1.5 V vs. SCE in CH CN] when compared to aryl halides,
3
single electron reduction of the thianthrenium moiety is
[
17]
possible at a potential that copper(III) is feasible,
which facile CꢀO reductive elimination could occur.
from
[
18]
A
Stern–Volmer analysis revealed that reductive quenching of
the excited iridium(III) photocatalyst by thianthrene is faster
than oxidative quenching by the thianthrenium salt. The
subsequently formed iridium(II) has a suitable reduction
potential for the reduction of the arylthianthrenium salt,
which results in aryl radicals that can undergo an oxidative
ligation to afford copper(III) aryl complexes (Scheme 2a).
Consistent with the formation of aryl radicals, we could
observe the 2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO)
adduct (52) upon addition of TEMPO to the reaction mixture
of CuTC instead of Cu O was crucial to obtain high yields, as
2
was the use of Na CO3 as base. Otherwise, substantial
2
amounts of hydrodefunctionalized side-products were
observed. The use of Na CO3 additionally reduced the
2
formation of aryl thiophene-2-carboxylate (i.e. Ar-TC) that
was observed in the absence of base and was also beneficial to
increase the conversion of the transformation. In the absence
of Na CO , Ar-TC byproduct was observed in up to 42%
2
3
yield (see Supporting Information, Table S2).
(
Scheme 2b).
Primary alcohols including methanol, d -methanol, etha-
4
nol, and 2,2,2-trifluoroethanol were coupled in 49–94% yield.
Additionally, 4-membered cyclic alcohols such as cyclobuta-
nol, 3-hydroxyoxetane and N-Boc protected 3-hydroxyazeti-
dine were successfully employed. Coupling with phenol, 1,2-
diols (30, 37), as well as alcohols of higher complexity (35, 38,
3
9) also proceeded to give the corresponding ethers. Beta
hydride elimination of secondary alcohols, which is a common
challenge for transition-metal-catalyzed CꢀO bond formation
[
10e,g]
to access aryl ethers
was not observed. The site-selective
etherification exhibits a similar functional group tolerance as
the hydroxylation and can be performed on gram scale (46). A
range of functional groups including nitriles, ketones, alde-
hydes, esters, carbamates, amides, halides, sulfonamides (40)
and heterocycles (47), as well as substrates bearing an ortho-
substituent (45) are tolerated. We also demonstrated that
complex drug-like molecules can be alkoxylated in high site-
selectivity (43, 47). Similar to ether bond formation, thioether
bond formation could be accomplished under similar reaction
conditions, with tetramethylenediamine (TMEDA) as addi-
tional reaction component, presumably as ligand for copper
(Table 3).
Scheme 2. a) Reaction design. b) Radical trapping experiment: reaction
conditions: thianthrenium salt (1.0 equiv), TEMPO (2.5 equiv), [Ir[dF-
(
CF )ppy] (dtbpy)PF ] (1 mol%), CuTC (1.5 equiv), H O (20 equiv),
3 2 6 2
MeCN, blue LED (34 W), 308C, 24 h.
In conclusion, we have reported the first site-selective
late-stage aromatic CꢀO bond formation synthesizing phe-
nols and arylethers from arene CꢀH bonds in two steps via
The scope of the hydroxylation spans from electron-
deficient (2, 7) to electron-rich arenes (4, 6). A wide range of
functional groups are tolerated including cyclopropyls, alde-
hydes, esters, nitriles, ethers, ketones, protected anilines,
protected amino acids (5), amides, heterocycles (11, 26) and
arylthianthrenium intermediates. We envisage that the reac-
tion will be enabling for late-stage diversification, especially
in drug discovery.
carbamates. Importantly, substrates containing protic groups Acknowledgements
such as alcohols, halides and pseudohalides (24), which can be
incompatible with transition metal-catalyzed hydroxylation
We thank the MPI fꢀr Kohlenforschung for funding, Chris-
[
8,10]
reactions,
are tolerated in our process. The basic con-
tophe Farꢁs and Markus Leutzsch (MPI fꢀr Kohlenforschung)
for NMR analysis regarding the determination of the
stereochemistry of compound (20), and Samira Speicher for
providing tetrafluorothianthene-S-oxide. We are grateful for
the help of the analytical departments of the MPI fꢀr
Kohlenforschung.
ditions used for most other cross coupling protocols to
introduce hydroxide can lead to side reactions, such as
[
8j]
hydrolysis of esters like in compound 20, or epimerization.
In our protocol, the base-sensitive methyl ester in 20 was
tolerated and no epimerization of the stereocenter was
observed. Several pharmaceuticals were hydroxylated site-
selectively (1, 6, 11, 16, 21, 25, 26). Occasionally, the
hydroxylation of alkyl substituted arenes can be problematic
Angew. Chem. Int. Ed. 2019, 58, 1 – 7
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
These are not the final page numbers!