Dual Photoredox/Cobaloxime Catalysis for Cross-Dehydrogenative α-Heteroarylation of Amines

Article abstract of DOI:10.1021/acs.orglett.1c01703

We report a dual-catalytic platform for the cross-dehydrogenative-coupling between (benzo-)thiazoles and amines which combines low loadings of an iridium photoredox catalyst and a cobaloxime catalyst under blue light irradiation. This transformation occurs without stoichiometric oxidants, giving products in moderate to excellent yields. DFT calculations support the key role of Co(II) for rearomatization of the radical-addition intermediate to generate the product.

Full text of DOI:10.1021/acs.orglett.1c01703

pubs.acs.org/OrgLett
Letter
Dual Photoredox/Cobaloxime Catalysis for Cross-Dehydrogenative
α‑Heteroarylation of Amines
Enrico Bergamaschi, Christopher Weike, Victor J. Mayerhofer, Ignacio Funes-Ardoiz,*
and Christopher J. Teskey*
Cite This: Org. Lett. 2021, 23, 5378−5382
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: We report a dual-catalytic platform for the cross-
dehydrogenative-coupling between (benzo-)thiazoles and amines
which combines low loadings of an iridium photoredox catalyst and
a cobaloxime catalyst under blue light irradiation. This trans-
formation occurs without stoichiometric oxidants, giving products
in moderate to excellent yields. DFT calculations support the key
role of Co(II) for rearomatization of the radical-addition
intermediate to generate the product.
he direct functionalization of C−H bonds is a bedrock of
Scheme 1. Context of This Work
T
modern synthetic chemistry. By forgoing the need for
functional handles such as halides and organometallics, it offers
a route to atom-economical and sustainable synthesis.1 Indeed,
selectively reacting together two C−H bonds is potentially one
of the most efficient ways to form a new bond between carbon
atoms. However, though formally only loss of H2 gas occurs,
waste is frequently generated in other ways. For instance,
dehydrogenative couplings often rely on superstoichiometric
quantities of an oxidant which is undesirable, as these can be
unsafe to use, generate waste themselves, and can reduce the
functional-group tolerance of the methodology.
The combination of earth abundant transition metal
catalysis with photoredox catalysis for proton reduction has
its roots in bioinorganic chemistry;2 however, it has also
proven to be a powerful approach in organic methodology
development.3,4 For instance, the combination of cobaloxime
and photoredox catalysis has allowed the development of
efficient strategies for decarboxylative dehydrogenation of
carboxylic acids to alkenes,5,6 direct functionalization of arenes
with simple nucleophiles,7−9 dehydrogenative synthesis of
anilines from in situ formed enamines,10 and several other
transformations.11−13
A number of methods that are able to functionalize the α-
position of nitrogen have also been reported using this
approach. However, the vast majority involve oxidation of the
α-amino radical species to give an iminium or imine species
which makes the amino-substrate susceptible to attack by
electron-rich, nucleophilic coupling partners (Scheme 1,
top).14−17
would then be able to reduce the cobalt catalyst and return to
its ground state. Subsequent oxidation and rearomatization of
the dearomatized intermediate would then proceed via action
of reduced cobalt species to generate the product.
We began our investigations with N,N-dimethyl-p-toluidine
in combination with benzothiazole using [Ir(dF(CF3)-
Received: May 20, 2021
Published: July 1, 2021
We considered however that it may be possible instead to
trap the nucleophilic α-amino radicals, generated via oxidative
quenching of the excited photocatalyst, with electron-deficient
(hetero)arenes, opening up new reactivity patterns for this area
of research (Scheme 1, bottom). The reduced photocatalyst
https://doi.org/10.1021/acs.orglett.1c01703
© 2021 American Chemical Society
Org. Lett. 2021, 23, 5378−5382
5378

Organic Letters
pubs.acs.org/OrgLett
Letter
ppy)2dtbbpy]PF6 (5 mol %) and Co(dmgH)2PyCl (8 mol %)
with a variety of bases in DMF. We observed traces (<10%) of
product in combination with 1.0 equiv of DABCO (Table 1,
Scheme 2. Substrate Scope
a
Table 1. Reaction Discovery and Selected Optimization
entry
Variation from “optimized conditions”
Yield (2a)
1
2
3
4
None
8 mol % [Co]
4CzIPN instead of [Ir]
16 h
Co(dmgH)2Cl(4-CNPy)
NEt3 or pyridine as base
PhCOOK instead of DABCO
DABCO (0.1 equiv)
Without [Ir]
75%
<10%
10%
40%
80%
<5%
8%
50%
−
−
12%
b
5
6
7
c
8
9
c
c
10
11
Without light
Without [Co]
b
a
1
Yields were determined by H NMR of the crude reaction mixtures
b
c
with an internal standard. 24 h reaction time. 16 h reaction time.
d
0.1 equiv of DABCO was used.
entry 2). However, no significant increase in yield was
observed until the loading of cobalt catalyst was decreased,
with 1 mol % proving sufficient for a good yield in combination
with 1 mol % of iridium photocatalyst. This may be a result of
the reaction mixtures being insufficiently transparent for light
penetration at higher catalyst loadings.
A range of organometallic and organic photocatalysts were
screened under the reaction conditions; however, only
moderate yields were obtained for noniridium based species
(Table 1, entry 3). Shortening the reaction time resulted in a
decreased yield (Table 1, entry 4); however, this was
counteracted by switching to a cobaloxime containing an
electron-withdrawing group on the axial pyridine ligand (Table
1, entry 5). Although reducing the quantity of DABCO from
1.0 equiv to 0.1 equiv did not significantly impact the yield, few
other bases were effective in promoting the reaction. Notably,
besides quinuclidine derivatives, other amine bases such as
triethylamine and pyridine resulted in significantly lower
quantities of product (Table 1, entry 6) and carboxylates
were ineffective (Table 1, entry 7).18 The reaction also
proceeded with substoichiometric DABCO albeit in slightly
lower efficiency (Table 1, entry 8).
Besides DMF, other solvents such as DMA and NMP could
be used without significant loss of yield and DMSO was also
moderately successful (see Supporting Information for further
optimization data). Without light, photocatalyst or cobalt
catalyst, no significant quantities of product were observed,
confirming the necessity of all three for the reaction to proceed
efficiently (Table 1, entries 9−11).19
Reactions carried out on 0.2 mmol scale. Both isolated yields and
those with the yield determined by internal standard from the crude
a
reaction mixture (in brackets) are given. Reaction time was 64 h.
b
c
Reaction time was 24 h. Co(dmgH)2Cl(3,5-diBrPy) (1 mol %) and
PhCOOK (1 equiv) were used.
has been reported to selectively generate α-amino radicals at
the methyl position,20 and indeed we observed reactivity
consistent with this, with cyclohexyl-containing substrate
giving product 2d. Indeed, methyl groups are even preferred
to reaction at the benzylic position, as shown in the synthesis
of 2e. Turning to heterocycle derivatives, N-phenyl pyrrolidine
worked well under the reaction conditions to give product 2f
in good yield. N-Phenyl indoline was less reactive under the
conditions, however, giving product 2g in lower yield. Finally,
N,N-diethyl aniline was unreactive under our previously
optimized reaction conditions, but interestingly, after reoptim-
ization, we were able to obtain the product from cross-coupling
in good yield using PhCOOK in place of DABCO. Notably,
PhCOOK was ineffective for most other substrates that we
tested.
With these optimized conditions in hand, our attention
turned to the scope of this reaction (Scheme 2). Beginning
with the aniline component, compared to the parent aniline
which gave product (2b) in good yield, para-bromo
substitution decreased the yield although the functional handle
remained in place in the product (2c). Next we sought to
probe the selectivity of the process, having noted that DABCO
Next, we explored the range of heterocycles that could be
used as starting materials. A number of different functional
5379
https://doi.org/10.1021/acs.orglett.1c01703
Org. Lett. 2021, 23, 5378−5382

Organic Letters
pubs.acs.org/OrgLett
Letter
groups, both electron-donating and electron-withdrawing,
could be appended to the benzothiazole ring resulting in
moderate to good yields of products (2i−2l); however, under
our developed conditions, benzoxazole was unreactive.
Interestingly, thiazoles containing either one or two ester
groups were also effective coupling partners (2n and 2p). It
seems, however, that 5-substitution was essential for the
reaction to proceed with the 4-substituted thiazole proving
unreactive. Notably, where the reactivity was low, the majority
of both starting materials could be recovered with the only
byproducts observed being the monodemethylated aniline,
which is consistent with previous reports, or small quantities of
the N-phenyl pyrrole from oxidation of the N-phenyl
pyrrolidine.13,21
Scheme 4. (a) Free Energy Profile for the α-N Carbon
Centered Radical Addition to Benzothiazole Followed by
Co(II) Mediated Rearomatization (Energies in kcal/mol
and Bond Lengths in Å); (b) Spin Density Analysis of
Co(II) Mediated Hydrogen Atom Abstraction Transition
State (Bond Lengths in Å)
On the basis of the reported literature, we outlined a
plausible mechanism (Scheme 3).22 Upon light irradiation, the
Scheme 3. Proposed Reaction Mechanism
iridium photocatalyst becomes excited and can subsequently
undergo single-electron transfer (SET), oxidizing either the
aniline or DABCO (both have been reported to quench the
excited state and they have similar oxidation potentials).10 The
photocatalyst itself simultaneously becomes reduced to Ir(II).
Next, DABCO can play the role of the base, deprotonating the
radical cation of the aniline, or the radical cation of DABCO
could directly abstract a hydrogen atom from the aniline.
Notably, in the formation of substrate 2h, where DABCO is
exchanged for PhCOOK, it is likely that the aniline is directly
oxidized due to the significantly higher oxidation potential of
the carboxylate.23 At this point, we hypothesized that the α-
amino radical would attack the most electrophilic C2 position
of the heteroarene and a Co(II) species, generated via SET
from the reduced Ir(II), might be responsible for rear-
omatization of the intermediate.
We therefore decided to analyze these potential steps of the
proposed mechanism via DFT calculations (Scheme 4).24 After
generating the α-amino radical via the photoredox catalytic
cycle, the addition of the nucleophilic C-centered radical to the
electrophilic C2 carbon of benzothiazole occurs via a low-
energy transition state (TSRad‑Add) of 11.7 kcal/mol (Scheme
4a). The polarity match in the transition state as well as the
π−π dispersion interaction between the two units facilitates the
formation of the radical addition intermediate 2·, which is
more stable than the separated reactants by 4.1 kcal/mol.
Then, rearomatization of 2· is enabled by the photochemically
generated Co(II) in the doublet state through a barrier of 18.2
kcal/mol, forming the final product 2a through an exergonic
process.
Due to the complex character of TSBSCo‑HAT, we further
analyzed the spin density evolution during the rearomatization
step (Scheme 4b). The reaction starts with the Co(II) radical
that is able to abstract a hydrogen atom from intermediate 2·
via a broken-symmetry open-shell singlet transition state. At
the same time, homolytic cleavage of the C−H bond facilitates
recombination of the β-unpaired electron in the benzothiazole
unit to form the singlet-aromatized product plus the singlet
Co(III)−H species, which will be reintroduced in the catalytic
cycle after reacting with the protonated base to form hydrogen
gas.25 As shown in the scheme, this concerted transition state
shows a partial reduction of the α spin density of Co(II) (0.52)
(i.e., a partial cobalt oxidation from II to III) and, at the same
time, reduction of the highly delocalized β spin density in the
benzothiazole radical adduct 2· to finally recover the
aromaticity in the product.
In conclusion, we have developed a novel dual catalytic
system, combining photoredox and cobaloxime catalysis, which
5380
https://doi.org/10.1021/acs.orglett.1c01703
Org. Lett. 2021, 23, 5378−5382

Organic Letters
pubs.acs.org/OrgLett
Letter
(5) Cartwright, K. C.; Tunge, J. A. Decarboxylative Elimination of
N-Acyl Amino Acids via Photoredox/Cobalt Dual Catalysis. ACS
Catal. 2018, 8 (12), 11801−11806.
(6) Sun, X.; Chen, J.; Ritter, T. Catalytic Dehydrogenative
Decarboxyolefination of Carboxylic Acids. Nat. Chem. 2018, 10
(12), 1229−1233.
enables the cross-dehydrogenative-coupling between the α-
position of amines and (benzo-)thiazoles without stoichio-
metric oxidants. Further work developing this chemistry is
underway in our laboratories which we will report in due
course.
(7) Zhang, G.; Liu, C.; Yi, H.; Meng, Q.; Bian, C.; Chen, H.; Jian, J.-
X.; Wu, L.-Z.; Lei, A. External Oxidant-Free Oxidative Cross-
Coupling: A Photoredox Cobalt-Catalyzed Aromatic C−H Thiolation
for Constructing C−S Bonds. J. Am. Chem. Soc. 2015, 137 (29),
9273−9280.
(8) Zheng, Y.-W.; Chen, B.; Ye, P.; Feng, K.; Wang, W.; Meng, Q.-
Y.; Wu, L.-Z.; Tung, C.-H. Photocatalytic Hydrogen-Evolution Cross-
Couplings: Benzene C−H Amination and Hydroxylation. J. Am.
Chem. Soc. 2016, 138 (32), 10080−10083.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.orglett.1c01703.
■
sı
Procedures and data for synthesized compounds;
computational details and xyz+energies of all the
calculations (PDF)
(9) Zhao, F.; Yang, Q.; Zhang, J.; Shi, W.; Hu, H.; Liang, F.; Wei,
W.; Zhou, S. Photocatalytic Hydrogen-Evolving Cross-Coupling of
Arenes with Primary Amines. Org. Lett. 2018, 20 (24), 7753−7757.
(10) Dighe, S. U.; Juliá, F.; Luridiana, A.; Douglas, J. J.; Leonori, D.
A Photochemical Dehydrogenative Strategy for Aniline Synthesis.
Nature 2020, 584 (7819), 75−81.
(11) Tian, W.-F.; Hu, C.-H.; He, K.-H.; He, X.-Y.; Li, Y. Visible-
Light Photoredox-Catalyzed Decarboxylative Alkylation of Hetero-
arenes Using Carboxylic Acids with Hydrogen Release. Org. Lett.
2019, 21 (17), 6930−6935.
FAIR data, including the primary NMR FID files, for
compounds 2a−2l, 2n, and 2p (ZIP)
AUTHOR INFORMATION
Corresponding Authors
■
Ignacio Funes-Ardoiz − Department of Chemistry, Centro de
Investigación en Síntesis Química (CISQ), Universidad de La
̃
Rioja, 26006 Logrono, Spain; orcid.org/0000-0002-
5843-9660; Email: ignacio.funesa@unirioja.es
Christopher J. Teskey − Institute of Organic Chemistry,
RWTH Aachen University, 52074 Aachen, Germany;
orcid.org/0000-0003-0776-5674;
(12) McManus, J. B.; Griffin, J. D.; White, A. R.; Nicewicz, D. A.
Homobenzylic Oxygenation Enabled by Dual Organic Photoredox
and Cobalt Catalysis. J. Am. Chem. Soc. 2020, 142 (23), 10325−
10330.
(13) Zhao, H.; Leonori, D. Minimization of Back-Electron Transfer
Enables the Elusive Sp3 C−H Functionalization of Secondary
Anilines. Angew. Chem., Int. Ed. 2021, 60 (14), 7669−7674.
(14) Zhong, J.-J.; Meng, Q.-Y.; Liu, B.; Li, X.-B.; Gao, X.-W.; Lei, T.;
Wu, C.-J.; Li, Z.-J.; Tung, C.-H.; Wu, L.-Z. Cross-Coupling Hydrogen
Evolution Reaction in Homogeneous Solution without Noble Metals.
Org. Lett. 2014, 16 (7), 1988−1991.
(15) Gao, X.-W.; Meng, Q.-Y.; Li, J.-X.; Zhong, J.-J.; Lei, T.; Li, X.-
B.; Tung, C.-H.; Wu, L.-Z. Visible Light Catalysis Assisted Site-
Specific Functionalization of Amino Acid Derivatives by C−H Bond
Activation without Oxidant: Cross-Coupling Hydrogen Evolution
Reaction. ACS Catal. 2015, 5 (4), 2391−2396.
(16) Niu, L.; Wang, S.; Liu, J.; Yi, H.; Liang, X.-A.; Liu, T.; Lei, A.
Visible Light-Mediated Oxidative C(Sp3)−H Phosphonylation for α-
Aminophosphonates under Oxidant-Free Conditions. Chem. Com-
mun. 2018, 54 (13), 1659−1662.
(17) Jia, Z.; Yang, Q.; Zhang, L.; Luo, S. Photoredox Mediated
Acceptorless Dehydrogenative Coupling of Saturated N-Heterocycles.
ACS Catal. 2019, 9 (4), 3589−3594.
(18) For a study on the privileged role that DABCO can play in
metallophotoredox catalysis, see: Till, N. A.; Tian, L.; Dong, Z.;
Scholes, G. D.; MacMillan, D. W. C. Mechanistic Analysis of
Metallaphotoredox C−N Coupling: Photocatalysis Initiates and
Perpetuates Ni(I)/Ni(III) Coupling Activity. J. Am. Chem. Soc.
2020, 142 (37), 15830−15841.
(19) Further investigation is underway in our group to identify the
precise mechanism of product formation without the cobaloxime
catalyst.
(20) Barham, J. P.; John, M. P.; Murphy, J. A. Contra-
Thermodynamic Hydrogen Atom Abstraction in the Selective C−H
Functionalization of Trialkylamine N-CH3 Groups. J. Am. Chem. Soc.
2016, 138 (47), 15482−15487.
Email: christopher.teskey@rwth-aachen.de
Authors
Enrico Bergamaschi − Institute of Organic Chemistry, RWTH
Aachen University, 52074 Aachen, Germany; orcid.org/
0000-0001-6725-8426
Christopher Weike − Institute of Organic Chemistry, RWTH
Aachen University, 52074 Aachen, Germany
Victor J. Mayerhofer − Institute of Organic Chemistry,
RWTH Aachen University, 52074 Aachen, Germany
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.orglett.1c01703
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the FCI (Liebig Fellowship to C.J.T., PhD
scholarship to E.B.) for financial support of this project. Joel
Reuleaux (RWTH Aachen University) is acknowledged for
starting material synthesis. I.F.A. thanks the University of La
Rioja for financial support and “Beronia” cluster for computa-
tional resources. We are grateful to Prof. Dr. Franziska
Schoenebeck for her generous support and advice.
REFERENCES
■
(1) Dalton, T.; Faber, T.; Glorius, F. C−H Activation: Toward
Sustainability and Applications. ACS Cent. Sci. 2021, 7 (2), 245−261.
(2) Artero, V.; Chavarot-Kerlidou, M.; Fontecave, M. Splitting
Water with Cobalt. Angew. Chem., Int. Ed. 2011, 50 (32), 7238−7266.
(3) Cartwright, K. C.; Davies, A. M.; Tunge, J. A. Cobaloxime-
Catalyzed Hydrogen Evolution in Photoredox-Facilitated Small-
Molecule Functionalization. Eur. J. Org. Chem. 2020, 2020 (10),
1245−1258.
(21) Rueping, M.; Vila, C.; Szadkowska, A.; Koenigs, R. M.; Fronert,
J. Photoredox Catalysis as an Efficient Tool for the Aerobic Oxidation
of Amines and Alcohols: Bioinspired Demethylations and Con-
densations. ACS Catal. 2012, 2 (12), 2810−2815.
(22) We also considered alternative mechanistic pathways. The
formation of the benzothiazole radical via an HAT process was
considered energetically unviable and is detailed in the Supporting
(4) Kojima, M.; Matsunaga, S. The Merger of Photoredox and
Cobalt Catalysis. Trends Chem. 2020, 2 (5), 410−426.
5381
https://doi.org/10.1021/acs.orglett.1c01703
Org. Lett. 2021, 23, 5378−5382

Organic Letters
pubs.acs.org/OrgLett
Letter
Information. For the rearomatization, we cannot exclude electron
transfer to form Co(I) and subsequent proton abstraction.
(23) Roth, H.; Romero, N.; Nicewicz, D. Experimental and
Calculated Electrochemical Potentials of Common Organic Molecules
for Applications to Single-Electron Redox Chemistry. Synlett 2016, 27
(05), 714−723.
(24) DFT calculations were carried out at the SMD(DMF) M06/
Def2TZVP//ωB97xD/Def2SVP level of theory. See Supporting
Information for the full computational details.
(25) At this current time we have been unable to detect hydrogen
gas which unambiguously arises from product formation. Therefore it
is not possible to rule out that intermediates or starting materials
could reversibly trap hydrogen that is generated.
5382
https://doi.org/10.1021/acs.orglett.1c01703
Org. Lett. 2021, 23, 5378−5382