Angewandte
Communications
Chemie
cheaper. Clearly, a new nickel-promoted formylation strategy,
which would enable aryl chlorides to serve as coupling
partners, would have substantial synthetic utility within both
academic and industrial settings. In this context, we recently
uncovered a formylation process under synergistic photo-
redox catalysis with nickel and organic catalysts. 2,2-Dieth-
oxyacetic acid served as the formylation reagent, thus
avoiding the need for CO. The process combines an
organic-dye-mediated photoredox-catalyzed formyl-radical-
forming reaction with a nickel-promoted radical coupling
process. Moreover, readily available aryl chlorides can serve
as substrates (Scheme 1b). The transformation proceeds
under straightforward and mild conditions that tolerate
a plethora of functional groups, and does not produce
abundant amounts of chemical waste.
Although nickel-catalyzed reactions of aryl halides with
[
3]
CO/H fail to generate formylation products, the capacity of
2
[6]
Ni catalysts to activate CÀCl bonds of aryl chlorides to
produce highly reactive organic radicals in cross-coupling
[8]
reactions is highly attractive. We reasoned that the gener-
ation of a formyl-radical equivalent in the presence of an aryl
chloride and a Ni catalyst might lead to a cross-coupling
process corresponding to the long-sought-after nickel-pro-
moted formylation reaction. Accordingly, the key to the
successful development of this process was the identification
of a new reagent that would efficiently produce a formyl-
radical equivalent under conditions that are compatible with
nickel catalysis. The results of studies of photochemical
decarboxylation reactions of glyoxylic acid and its acetals by
Scheme 2. Proposed nickel–photoredox-catalyzed formylation of aryl
halides and triflates.
harvesters in organic light-emitting diodes, we were attracted
to readily available 1,2,3,5-tetrakis(carbazol-9-yl)-4,6-dicya-
[
14,15]
nobenzene (4CzIPN).
The high reduction potential of the
*
red
photoexcited state of 4CzIPN (E =+ 1.35 V vs. SCE)
augurs well for the use of this substance to promote photo-
oxidative decarboxylation reactions.
[
9]
us, and related recent efforts by the research groups of
In initial studies to probe the feasibility of the new
formylation reaction, we used glyoxylic acid as the formyl-
radical source. Unfortunately, none of the desired aldehyde
product 3a was produced when a mixture of 4-bromobenzo-
[10]
MacMillan, Doyle, and Overman,
indicated that these
inexpensive and abundant substances might be ideal precur-
sors of formyl-radical equivalents. This consideration sug-
gested that the merging of visible-light photoredox catalysis
with nickel catalysis, in a manner earlier demonstrated by the
nitrile (1a), glyoxylic acid monohydrate, NiCl ·glyme
2
(glyme = ethylene glycol dimethyl ether), 4CzIPN, 4,4’-di-
tert-butyl-2,2’-bipyridine (dtbbpy), and Cs CO in N,N-dime-
[
11,12]
research groups of MacMillan, Doyle, and Molander,
2
3
could serve as the foundation for the new transformation.
Specifically, we hypothesized that photoredox-mediated
single-electron-transfer (SET) oxidation of glyoxylic acid or
thylformamide (DMF) at room temperature was irradiated
with blue-light-emitting diodes (LEDs) for 24 h (Table 1,
entry 1). We reasoned that the failure of this reaction might
be associated with the difficulty of promoting SET oxidation
its acetals, followed by the loss of CO , would form a formyl-
2
ox
radical equivalent 4 (Scheme 2, with a diacetal as an
of glyoxylic acid because of its high oxidation potential (E =
0
example). Reaction of the Ni complex with radical 4 would
+ 1.33 V vs. SCE, CH CN; see Figure S1 in the Supporting
3
[12d]
[16]
form the diacetal–nickel(I) intermediate 5.
Oxidative
Information) and lability. A careful search uncovered the
activation of aryl halide 1 to 5 would then generate the
putative Ni complex 6, which should undergo reductive
elimination to produce the desired aryl diacetal 7, a substance
that can be readily converted into aryl aldehyde 3 through
workup with an aqueous acid.
fact that the readily available and inexpensive diethyl acetal
III
derivative of glyoxylic acid 2a, generated by the hydrolysis of
À1 [17]
ethyl diethoxyacetate ($0.56g ), has a much lower oxida-
ox
tion potential [E (2a Cs salt) =+ 0.95 V vs. SCE, CH CN
3
(see Figure S2). Thus, this acetal should be more readily
oxidized by SET to the excited state of 4CzIPN. Significantly,
Thus far, ruthenium and iridium complexes have typically
been used as photocatalysts in photoredox nickel-catalyzed
reactions. It is somewhat surprising that far fewer reports exist
describing reactions of this type in which organic dyes serve as
photocatalysts, despite the lower cost, wider availability,
higher stability, and superior properties of these substances as
compared to those of their inorganic and organometallic
the irradiation of a mixture containing 1a, 2a, NiCl ·glyme,
2
4CzIPN, dtbbpy, and Cs CO in DMF at room temperature
2
3
with blue-light-emitting diodes (LEDs) led to smooth for-
mation of the formylation product 3a in 72% yield (Table 1,
entry 2).
Encouraged by this preliminary result, we evaluated the
effects of several parameters, including the Ni salt, photo-
sensitizer, base, solvent, ligand, and light source on this
transformation (Table 1; see also Table S1 in the Supporting
Information). Slightly higher efficiency was observed when
[
13]
counterparts in many cases. Because of this comparison, we
decided to exploit organic dyes as photoredox catalysts for the
new formylation reaction. Among the family of carbazoyl
dicyanobenzenes reported by Adachi and co-workers as light
2
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 7
These are not the final page numbers!