ACS Catalysis
Letter
To probe the viability of our mechanistic blueprint, we
performed a series of mechanistic studies. Conducting the
reaction in the presence of D2O led to deuterated products,
which point toward a rapid, reversible C−H metalation step
(Scheme 4A). This was not observed for acrylic acids, which
support a C−H alkenylation/oxa-Michael addition sequence,
and rules out a cross-dehydrogenative coupling route for the
acrylate cross-couplings. A negligible isotope effect for
deuterated substrate confirmed that neither of the two C−H
activation steps is rate determining (Scheme 4B). This explains
why it is so difficult to achieve selectivity in cross-couplings.
The reaction rates correlate with the electric current densities,
showing that the regeneration of the RhIII catalyst is rate-
determining (Scheme 4D).
When conducting the reaction in the absence of electricity,
the amount of product corresponded to the amount of Rh
catalyst (entry 22, Table 1). With Rh(OAc)3, the stoichio-
metric reaction gave 51% yield (Scheme 4C). This suggests
that the reaction starts from RhIII without an upfront oxidation
step. In cyclic voltammetry, two reduction peaks of RhCl3·
3H2O were observed at −0.50 and −1.10 V (curve c), which
confirms that the RhIII salt is easily reduced to RhII and/or RhI,
but that an oxidation to RhIV and/or RhV is difficult under the
Figure 1. Cyclic voltammograms using glassy carbon electrode, scan
rate: 100 mV/s, 10 mM 1a, 0.1 M Bu4NPF6, 0.1 M Bu4NOAc, 5
mM RhCl3·3H2O.
n
n
tolerance of the process is remarkable. It includes ether (2g,
2h, 2y), ester (2l, 2q, 2ad, 2ai), keto (2j, 2k, 2ab), cyano
(2ag), sulfonyl (2ae), amino (2o), fluoroalkyl (2m),
fluoroalkoxy (2i, 2z, 2aa), and fluoroalkylthio groups (2p).
Oxidation-sensitive formyl groups (2ac) and easily reduced
nitro groups (2r, 2af) are tolerated. Compounds bearing
chloro (2e), bromo (2f), iodo (2x), and even SO2F (2s)
groups were selectively coupled, demonstrating the orthogon-
ality of this process to traditional cross-couplings.
Competing double arylation was observed in significant
amounts only for unsubstituted benzoic acid (product 3t). Any
substituent seems to effectively suppress unwanted double
arylation (2n−2ag). 3-Substituted benzoic acids as well as 2-
naphthoic acid underwent selective coupling in the less-
hindered position. Multisubstituted and fused benzoic acids
were also successfully coupled (2ah−2at), as were several
heterocyclic acids (2au−2ax). In some cases, for example, 2h,
2r, and 2ag, less than 5% yields of final ortho-C−H
methoxylated products were observed, which probably result
from the ortho-C−H hydroxylation16 of benzoic acids with air
and subsequent etherification by reaction with methyl iodide.
The electrooxidative coupling was successfully performed on
multigram scale with 3% catalyst loading (Scheme 3). The
diacid 2a was obtained in high purity after recrystallization,
which underlines the scalability of the reaction.
We went on to investigate whether selective electrocatalytic
dehydrogenative cross-coupling could be achieved for mixtures
of two different aromatic carboxylates (Table 3). Since 1-
naphthoic acids had reacted rather slowly in homodimeriza-
tions, we investigated their cross-coupling with more reactive
acids. Indeed, moderate yields and reasonable selectivities were
achieved in cross-dimerizations with benzoic acids (4a−4h).
Attempts to achieve selectivity in the coupling of electron-rich
with electron-deficient benzoic acids gave close to statistical
product mixtures (4i). Cross-couplings with vinylcarboxylates
resulted in the formation of lactones. The lactonization is a
typical follow-up reaction in α-vinylations of benzoic acids
(5a−5c), which itself could result from a C−H alkenylation/
oxa-Michael addition sequence, or a cross-dehydrogenative
coupling. In this reaction, [Cp*RhCl2]2 gave higher yields than
RhCl3·3H2O, which points toward a shift in the mechanistic
pathway.
n
reaction conditions (Figure 1). Adding Bu4NOAc to RhCl3·
3H2O showed a broad reduction peak, which may result from
carboxylate-bridged oligonuclear species (curve d). This
system displayed a large oxidation peak at +1.35 V, which
may be explained with the oxidation of the acetate (curve b).
All findings support the pathway proposed in Scheme 2.
In conclusion, the newly developed electrocatalytic dehy-
drogenative C−H/C−H coupling allows generating various
biaryl dicarboxylates under mild conditions. The reaction is
operationally simple and easily scalable, has an excellent
functional group compatibility, and is orthogonal to common
cross-coupling strategies.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
■
sı
General information, experimental details, spectra data,
copies of NMR spectra for all final products (PDF)
AUTHOR INFORMATION
Corresponding Author
■
Lukas J. Gooßen − Fakultät Chemie und Biochemie, Ruhr
Universität Bochum, 44801 Bochum, Germany;
Authors
Zhongyi Zeng − Fakultät Chemie und Biochemie, Ruhr
Universität Bochum, 44801 Bochum, Germany
Jonas F. Goebel − Fakultät Chemie und Biochemie, Ruhr
Universität Bochum, 44801 Bochum, Germany
Xianming Liu − Fakultät Chemie und Biochemie, Ruhr
Universität Bochum, 44801 Bochum, Germany
Complete contact information is available at:
Notes
The authors declare no competing financial interest.
6630
ACS Catal. 2021, 11, 6626−6632