Photoinduced Trifluoromethylation of Arenes and Heteroarenes Catalyzed by High-Valent Nickel Complexes

Article abstract of DOI:10.1002/anie.202109953

We describe a series of air-stable Ni^III complexes supported by a simple, robust naphthyridine-based ligand. Access to the high-valent oxidation state is enabled by the CF₃ ligands on the nickel, while the naphthyridine exhibits either a monodentate or bidentate coordination mode that depends on the oxidation state and sterics, and enables facile aerobic oxidation of Ni^II to Ni^III. These Ni^III complexes act as efficient catalysts for photoinduced C(sp²)?H bond trifluoromethylation reactions of (hetero)arenes using versatile synthetic protocols. This blue LED light-mediated catalytic protocol proceeds via a radical pathway and demonstrates potential in the late-stage functionalization of drug analogs.

Full text of DOI:10.1002/anie.202109953

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Photoinduced Trifluoromethylation of Arenes and Heteroarenes
Catalyzed by High-Valent Nickel Complexes
Authors: Shubham Deolka, Ramadoss Govindarajan, Eugene Khaskin,
Robert R. Fayzullin, Michael C. Roy, and Julia Khusnutdinova
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202109953
Link to VoR: https://doi.org/10.1002/anie.202109953

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
Photoinduced Trifluoromethylation of Arenes and Heteroarenes
Catalyzed by High-Valent Nickel Complexes
Shubham Deolka,[a] Ramadoss Govindarajan,[a] Eugene Khaskin,[a] Robert R. Fayzullin,[b] Michael C.
Roy,[a] and Julia R. Khusnutdinova*[a]
[a]
[b]
S. Deolka, R. Govindarajan, E. Khaskin, M. C. Roy, J. R. Khusnutdinova
Okinawa Institute of Science and Technology Graduate University
Onna-son, Kunigami-gun, Okinawa, Japan, 904-0495
E-mail: juliak@oist.jp
R. R. Fayzullin
Abuzov Institute of Organic and Physical Chemistry, FRC Kazan Scientific Center of RAS
8 Arbuzov Street, Kazan 420088, Russian Federation
Supporting information for this article is given via a link at the end of the document.
Abstract: We describe a series of air-stable NiIII complexes supported
by a simple, robust naphthyridine-based ligand. Access to the high
valent oxidation state is enabled by the CF3 ligands on the nickel,
while the naphthyridine exhibits either a monodentate or bidentate
coordination mode that depends on the oxidation state and sterics,
and enables facile aerobic oxidation of NiII to NiIII. These NiIII
complexes act as efficient photoredox catalysts for C(sp2)–H bond
trifluoromethylation reactions of (hetero)arenes using versatile
synthetic protocols. This blue LED light-mediated catalytic protocol
proceeds via a radical pathway and demonstrates potential in the late-
stage functionalization of drug analogs.
There are several recent reports where Ni complexes were
utilized for trifluoromethylation, all suggesting that a higher Ni
oxidation state is required for CF3 modification. Vicic and
coworkers showed that ArNiIICF3 was unfeasible for reductive
elimination (RE)[17]. Later, Nebra and coworkers designed a high
valent NiIV complex where stoichiometric trifluoromethylation was
shown to occur via reductive elimination.[18] There, the Ni complex
was used as a stoichiometric reagent and a CF3 source in the
presence of excess substrate. Finally, the Sanford group has
recently reported that a pyrazolyl borate supported complex
catalyzed trifluoromethylation of arenes via a NiIV intermediate
and a fluoro-modified Umemoto reagent.[16a] Similar catalytic
activity was reported recently by the Vicic group using solvated Ni
complexes.[19] In the Sanford and Vicic protocols, the Umemoto
reagent was used as the limiting reagent in the presence of the
excess substrate. This is not ideal in the case of an expensive
heterocycle substrate as it results in low conversion.
Introduction
Recently, nickel has emerged as
a significantly cheaper
alternative to palladium in a wide array of catalytic reactions.[1]
Many reactions that were first catalyzed by expensive noble
metals catalysts, such as cross-coupling,[2] nucleophilic
allylation,[3] and fluorination/trifluoromethylation reactions among
others,[4] can now all be done via nickel, albeit normally at greater
catalyst loading.[5] These catalytic reactions are often proposed to
proceed via high valent NiIII/NiIV intermediates.[6] In the last few
years, several reports have suggested that complexes that have
access to high valent oxidation states of nickel, can end up being
competent catalysts in C–O,[7] C–C[8] bond formation and C–H
bond functionalization.[9] However, a general approach for the
accessibility/formation of high valent nickel complexes is still
lacking due to their low stability.[10]
Although feasibility in catalysis was demonstrated, Ni-
catalyzed trifluoromethylation examples currently lack the
versatility of precious metal protocols and do not attempt to utilize
a photochemical approach such as that demonstrated by
MacMillan. In particular, despite the recent surge of interest in the
photochemistry of organometallic nickel complexes, examples of
Ni-catalyzed, photoinduced C-H bond functionalization reactions
remain scarce,[20] and no examples of photo-induced, base-metal,
trifluoromethylation catalysts are currently known.
Arene trifluoromethylation has recently emerged as
a
[11]
powerful strategy for modifying advanced intermediates,
and
even active pharmaceuticals,[12] with a functional group that can
significantly change the pharmacokinetics of a known drug[13] and
assists in its passage through the blood/brain barrier.[14] A seminal
report by MacMillan showed that a second-row metal Ru
photosensitizer was able to catalytically install the CF3 group at
good efficiency using a cheap CF3 source (Figure 1).[15] However,
there are only a few examples of inexpensive first-row transition
metals being used as catalysts for arene trifluoromethylation. [10c,
16]
An important consideration for developing a practically
applicable catalyst is the use of a simple and inexpensive ligand
and trifluoromethyl group source.
Figure 1. Catalytic trifluoromethylation of (hetero)arenes.
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
A major theme of our group’s research consists of designing
bimetallic systems with a naphthyridine ligand core, resulting in
complexes that often reveal mechanistic insights into catalytic
processes.[21] We recently began to examine nickel as it is active
in a number of coupling reactions, including trifluoromethylation.
During an evaluation of a series of alkyl-substituted naphthyridine
ligands, we found that bis-trifluoromethyl nickel(II) complexes
chelated by simple naphthyridines are easily oxidized by air as
the sole oxidant to produce air-stable and isolable NiIII bis-
trifluoromethyl complexes. This is the first example of facile
aerobic oxidation of an organometallic Ni complex to give well-
defined, isolated NiIII species.
workers for their bipyridine-supported NiII(CF3)2 complex and was
attributed to sterics.[23] We hypothesized that the NMR broadening
of 1 likely arises due to the low interconversion barrier between
the square planar and distorted geometries, although other types
of ligand exchange cannot be excluded (vide infra). A solid state
and a solution (Evans method) magnetic moment measurement
both showed that the complex is diamagnetic.
Further exploring the chemistry of these unusual, bidentate-
mode coordinated naphthyridine complexes, we found that they
were capable of trifluoromethylating arenes even under air when
irradiated with blue LED light. Inspired by the above reports on
high-valent nickel catalysis and photoredox mediated Ru
trifluoromethylation, we were able to extend their chemistry to
achieve efficient photocatalytic trifluoromethylation via two
alternative protocols, using either a cheap CF3SO2Na, or the more
expensive but also commercially available Umemoto (non-
modified) reagent as a CF3 source (Figure 1).
This result is to the best of our knowledge, the first report of a
Ni-catalyzed photoinduced trifluoromethylation. Under the
optimized procedure, our TONs are also an order of magnitude
higher than those achieved in the earlier Ni protocol.
Thus, we developed a system with a cheap and readily
available ligand and CF3 reagent set, shows photoinduced
reactivity, and does not require a 2nd/3rd-row transition metal. Our
preliminary mechanistic studies show that CF3 radical formation
occurs during catalysis, and that it is promoted by blue LED light
irradiation (465 nm) under mild conditions.
Scheme 1. Synthesis of NiII(CF3)2 complexes 1-3 supported by naphthyridine
ligands.
Results and Discussion
Ligand design and synthesis of NiII/NiIII bis-trifluoromethyl
complexes. We focused our attention on Ni(MeCN)2(CF3)2 as a
stable and easily accessible precursor that could potentially
stabilize higher oxidation states of Ni when combined with N-
donor ligands. Treating Ni(MeCN)2(CF3)2 with 2 equiv. of 1,8-
naphthyridine ligand (L1) (Scheme 1) formed the yellow color
complex 1, which was isolated in 71% yield and characterized by
1H NMR, UV-vis, FT-IR, high-resolution mass spectrometry
1
(HRMS), and single-crystal X-ray diffraction (SC-XRD). The H
NMR spectrum of 1 displayed broadened resonances in the
aromatic region that could not be fully resolved even at –80 °C,
suggesting fluxional behavior (see below).
The X-ray structure of 1 revealed that there are two independent
halves of two molecules in the asymmetric cell (Z' = ½ + ½), where
each Ni atom is situated on the second-order rotational axis and
supported by two monodentate L1 ligands, with one Ni center
featuring almost ideal square planar geometry with τ4 = τ4' = 0.03
(Figure 2a) and another Ni center exhibiting slight distortion from
square planar geometry with τ4 = τ4' = 0.16.[22] This is also
manifested in the different trans-C–Ni–N angles in both
structures: while the square planar geometry is characterized by
an almost ideal C1–Ni1–N11i angle of 178.13(6)°, the analogous
angle in the distorted square planar complex is 168.77(6)°,
showing significant deviation from planarity. Interestingly,
deviation from planarity was also reported by Vicic and co-
Figure 2. ORTEP of 1 (a), 2 (b), 2’ (c), and 3 (d) at the 60% (a, b, and d) or 30%
(c) probability level. The second independent molecule for 1 and 2’, solvent
molecule for 2, and hydrogen atoms are omitted for clarity.[24]
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
To induce bidentate ligand coordination to the NiII metal center, performed in presence of 1 equiv of L3, the isolated yield
we then utilized monomethyl and dimethyl-substituted
naphthyridine ligands, L2 and L3. Reacting L2 with
Ni(MeCN)2(CF3)2 at room temperature (RT) gave yellow color
complex 2. Two types of crystals were obtained, 2 and 2’, by
crystallization from an acetone/pentane solvent mixture and m-
xylene, respectively (Figure 2b, c). Complex 2, crystallized from
acetone/pentane as a 1:1 acetone solvate, showed an almost
ideal square planar NiII atom bound to two monodentate L2
ligands (τ4 = 0.04, τ4' = 0.03), while N-atoms next to the two ortho-
Me groups in L2 remained non-coordinated. The XRD analysis for
crystals of 2’ obtained from a non-polar m-xylene allowed to
increased to 75%. Complex 3a was characterized by electrospray
ionization high-resolution mass spectrometry (ESI-HRMS), UV-
vis, FT-IR, and EPR spectroscopy, and elemental analysis. We
found that 3a is stable under air for months without apparent
decomposition.
establish atom connectivity and confirmed
a
bidentate
coordination mode of only one L2 ligand per Ni center, showing a
1
distorted square planar geometry (τ4 = τ4' = 0.21-0.22). The H
and 19F NMR spectra in CD2Cl2 solution showed broadened
signals at RT, suggesting that chemical exchange likely occurs in
solution. At –80 °C, signals of CF3 in the 19F NMR spectrum and
1
singlets of Me groups of L2 in the H NMR spectrum partially
resolved, confirming that the complex was present as a mixture of
two main components, presumably the 2:1 and 1:1 complexes 2
and 2’.
By contrast, the even more sterically hindered bis-Me L3 was
able to react in a clean 1:1 ratio with the precursor to provide facile
access to the bidentate yellow complex 3, which displayed a well-
resolved NMR spectrum featuring sharp peaks even at room
temperature. Isolated complex 3 was characterized by NMR, FT-
IR, UV-vis, ESI-HRMS, and elemental analysis The strained
bidentate coordination of L3 forming a four-membered chelate
cycle is manifested in the small N1–Ni1–N1i angle of 67.09(13)°
(Figure 2d). The geometry around the Ni atom is distorted square
planar (τ4 = τ4' = 0.23) with a C1–Ni1–N1i angle of 163.59(12)°.
Next, we examined whether 1-3 can react with 1e–/2e–
oxidants to yield NiIII/NiIV products. Recent studies by Mirica, Vicic,
and Sanford have shown that NiII organometallic complexes can
react with outer-sphere 1e–/2e– oxidants to form high-valent NiIII
or NiIV complexes. The cyclic voltammograms (CV) for complexes
1, 2, and 3 showed chemically reversible oxidation waves at the
anodic peak potentials Epa 0.492, 0.468, and 0.448 V vs. Fc/Fc+,
respectively, with a large peak-to-peak separation between
forward and reverse waves (ΔEp in a range of 0.612-0.695 V)
indicative of significant structural reorganization during oxidation
(Table S1, Figure 3). Surprisingly, we found that solution of
complex 3 was air-sensitive even in the absence of added
oxidants leading to the formation of a purple solution containing
paramagnetic oxidation product as confirmed by NMR and EPR
spectroscopy. Upon further optimization, we found that simply
stirring complex 3 under air with a NaBF4 salt additive in MeOH
solution gave the deep-purple complex 3a, which was isolated in
42% yield and characterized by XRD (Figure 4b). According to
XRD, 3a is a monocationic complex showing a tetragonally
distorted octahedral geometry around the Ni atom, with two L3
ligands bound in a bidentate fashion and the two CF3 groups cis
to each other. The axial Ni–N distances (2.1665(12)-2.1937(12)
Å) are significantly elongated compared to the equatorial Ni–N
distances (2.0340(12)-2.0343(13) Å), consistent with a Jahn–
Teller distorted NiIII center. The Ni–C bond distances in 3a
(1.9306(16)-1.9364(15) Å) are slightly longer compared to the
corresponding NiII complex 3 (1.879(3) Å).
Figure 3. Cyclic voltammograms of complexes 1 (black), 2 (blue), and 3 (red).
Conditions: 0.1M nBu4NBF4/MeCN solutions at RT, scan rate 0.1 V s-1, 1.6 mm
Pt disk electrode, the arrow indicates the initial scan direction. Complex 2 is
obtained by reacting 2 equiv of L2 with Ni precursor.
Scheme 2. Synthesis of NiIII-CF3 complexes 1a-3a.
–
Figure 4. ORTEP of 1a (a) and 3a (b) at 60% probability level. Counterion BF4
and hydrogen atoms are omitted for clarity.[24]
The low yield is a direct result of requiring a sacrificial L3
ligand from another complex of 3. When the reaction was
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
The X-band EPR spectrum of 3a (nPrCN/EtCN glass, 94 K)
exhibits an axial signal with g⏊ =2.203 and g∥=2.020 featuring
superhyperfine splitting from two N-atoms along gz (AN = 17.8 G)
(Figure 5). Significant g anisotropy (Δg = g⏊ − g∥ = 0.183) and
relatively high average g value (gave = 2.142) suggest significant
localization of the spin density at the Ni center consistent with
DFT-calculated spin density distribution (Figure 6).[8, 25]
We next investigated aerobic oxidation of the two other
NiII(CF3)2 complexes 1 and 2 in the presence of NaBF4 (Scheme
2). As these complexes were obtained by reacting Ni precursor
with 2 equiv of ligand L1 or L2, no additional ligand was added (in
the case of complex 2, an undetermined amount of free ligand L2
was present in the reaction mixture before oxidation). We found
that these complexes also reacted with air to form NiIII complexes,
and reddish-purple complex 1a and purple complex 2a were
isolated from MeOH solutions in 74% and 57% yields respectively,
and characterized by EPR, FT-IR, UV-vis, ESI-HRMS, and
elemental analysis. Complex 1a was also characterized by SC-
XRD (Figure 4a), showing a similar distorted octahedral geometry
with an elongated axial Ni–N distances (2.1826(11) Å) as
compared to equatorial Ni–N distances (2.0248(12) Å). The Ni–C
distances in 1a (1.9373(14) Å) are also elongated as compared to
1 (Ni–C 1.9023(16) and 1.8977(16) Å for the first and second
molecules), reflecting a similar trend as in complexes 3a and 3.
The rest of the spectroscopic features were consistent with the
structure of 1a and 3a. The magnetic moments in CD3CN solution
measured by the Evans method for complexes 1a, 2a, and 3a
were determined to be 2.04, 1.87, and 1.95 μB, respectively,
corresponding to one unpaired electron. The EPR spectrum of
complex 1a revealed a nearly axial signal similar to 3a with gave of
2.141 (gx = 2.200, gy = 2.183, gz = 2.0202) and showing
superhyperfine splitting from two N-atoms in the gz component
(AN = 16.8 G) (Figure S34), while the signal for 2a was broadened
and poorly resolved probably due to low symmetry and possible
isomer formation (Figure S35).
Figure 5. Experimental (black line) and simulated (red line) EPR spectra of 3a
in nPrCN–EtCN glass (1:1) at 94 K. Simulation parameters: g⏊ =2.203, g∥
2.020, AN,zz = 17.86 G.
=
Figure 6. Calculated spin density (left; isovalue 0.004) and singly occupied
molecular orbital (right; isovalue 0.06) for 3a cation [(L3)2Ni(CF3)2]+ in the gas
phase. The geometry was optimized using ωB97XD functional[29] and def2tzvp
basis set[30] for all elements.
Considering that aerobic oxidation is often proposed to
involve superoxide intermediates, we attempted to detect
possible metal-superoxide species using DMPO as a spin trap
(Figure 7). Indeed, exposure of 3 to air in MeOH solution in the
presence of DMPO gave rise to
a new signal whose
superhyperfine splitting constants (AN = 13.57 G, AH = 8.21 G, g
= 2.008) resemble those reported for detected or proposed metal-
superoxide species in aerobic oxidation of first-row transition
metals such as Co,[26] Cu,[27] or Ni.[28]
Overall, these results show that simple naphthyridine-based
ligands induce facile aerobic oxidation and stabilization of NiIII
complexes. Complex multi-chelating tri- and tetradentate ligands
were not necessary for the formation of stable NiIII complexes,
contrary to the commonly used strategy reported in the literature
for stabilization of high valent Ni species, and even with a simple
naphthyridine ligand, selective aerobic oxidation was achieved.
We believe that the highly strained coordination and small bite
angle of the naphthyridine-based ligands, resulting in
unsymmetrical coordination with notably different Ni–N distances,
make them suitable for stabilization of Jahn–Teller distorted NiIII
centers leading to such unexpected oxidation behavior and the
trifluoromethylation reactivity presented below.
Figure 7. Experimental (black line) and simulated (red line) EPR spectra of the
solution of 3 exposed to air for 4 min at RT in MeOH in the presence of NaBF4
and DMPO spin trap. Simulation parameters: AN = 13.57 G; AH = 8.21 G, g =
2.008.
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
Stoichiometric trifluoromethylation. The above results
meant that we now had access to three air-stable NiIII–CF3
complexes. Based on literature analogs, we anticipated that these
complexes might show propensity to undergo homolysis of the
NiIII–CF3 bond, leading to CF3 radical formation, which could
ultimately be applied for C(sp2)–H bond trifluoromethylation. For
example, transient formation of unstable NiIII(CF3)2 complexes
with terpyridine ligand has been reported by Vicic upon oxidation
of the corresponding NiII(CF3)2 precursor with ferrocenium.[25] This
complex was shown to be unstable and highly reactive, leading to
immediate Ni–CF3 bond homolysis, and was not used in
stoichiometric or catalytic trifluoromethylation reactions. Transient
NiIII species were proposed in other trifluoromethylation protocols
photoirradiation improved the reaction outcome (Table 1, entry 2).
Ultimately, we found that the transformation was more effective in
the case of pyrroles, giving moderate yields of 40-58% in the
stoichiometric reaction for N-substituted pyrroles (Table 1, entries
3-4). Interestingly, adding an oxidant[33] to the typical pyrrole C–H
trifluoromethylation reaction (entries 5-6) gave much larger yields
and suggested that the reaction may not have been strictly
stoichiometric, which ultimately led us to develop a catalytic
version of this transformation (vide infra).
To confirm that light-induced CF3 radical formation is involved
in the stoichiometric reactivity, we exposed a DMSO solution of
complex 3a and the TEMPO radical trap to blue LED irradiation at
RT. NMR analysis showed the formation of the TEMPO-CF3
adduct after just several minutes of irradiation (Scheme 3, a).
Similarly, the TEMPO-CF3 adduct was also observed when a
solution of 3a and TEMPO was irradiated in the presence of a
substrate, N-phenylpyrrole, where at the same time, formation of
the trifluoromethylation product was suppressed (Scheme 3, b).
No TEMPO-CF3 radical was observed when 3a was reacted with
TEMPO in the absence of light showing that the bond homolysis
at RT is induced by irradiation (Scheme 3, c). The ESI-HRMS
analysis of complex 3a solution after irradiation with blue LED in
the absence of TEMPO showed the presence of a signal
corresponding to a [(L3)2NiIICF3]+ species (m/z 443.0981), which
is likely to result from a NiIII–CF3 bond homolyses (Scheme 3, d).
In addition, we performed spin trapping experiments using a
solution of 3a in the presence of PBN as a radical trap. The EPR
analysis of the reaction mixture confirmed the formation of a CF3-
using the Togni reagent or Langlois reagent, however, they were
16b]
not detected experimentally.[10c,
Soper and co-workers
reported light-induced CoIII–CF3 bond homolysis, however, only
stoichiometric trifluoromethylation of arene was observed and no
catalytic transformation was reported.[31] Having isolated
NiIII(CF3)2 complexes 1a-3a, we aimed to induce CF3 radical
formation in a controlled way, in particular, using light to trigger
bond homolysis, and utilize it for stoichiometric or catalytic C–H
bond trifluoromethylation.
Table 1. Stoichiometric C–H bond trifluoromethylation by 3a.[a]
Entry
1
Substrate
Conditions
Yield [%]
< 5
Ambient light
PBN adduct, which was detected by EPR spectroscopy (AN
=
14.28 G, AH = 2.11 G, AF = 1.54 G (CF3), g = 2.007) after several
minutes of irradiation with blue LED light (Figure 8 and Scheme 3,
e). The simulated spectrum and superhyperfine splitting
constants are in good agreement with literature data for the PBN-
CF3 radical adduct. [34],[25]
2
Blue LED
(465 nm)
16
Finally, we checked the reactivity of a parent NiII(CF3)2
complex in the absence of oxidant/air. However, exposure of NiII
complex 3 to blue LED light in the presence of TEMPO did not
lead to TEMPO-CF3 formation, suggesting that a higher oxidation
state is necessary for CF3 radical formation (Scheme 3, f).
Overall, these results show that light-induced homolysis of the
NiIII–CF3 bond in complex 3a results in the formation of a CF3
radical, which then participates in C(sp2)–H bond
trifluoromethylation. This reaction is enhanced by the addition of
oxidants.
3
4
5
6
Blue LED
Blue LED
40
58
71
80
Blue LED
K2S2O8 (1 equiv)
Blue LED
tBuOOH (1 equiv)
[a] The reactions were performed under air for 24 hours using 1 equiv of
substrate and 1 equiv of 3a in DMSO at RT under blue LED light (465 nm) unless
indicated otherwise. The yields were determined by 19F NMR based on
integration against α,α,α-trifluorotoluene as a standard.
Our first attempt, the reaction of 1,3,5-trimethoxybenzene
under ambient light, afforded only a trace amount of mono-
trifluoromethylated product. Inspired by previous reports of light-
induced metal-carbon bond homolysis in PdIII complexes[32] and
the synthetic utility of photochemical trifluoromethylation reported
by the MacMillan group,[15] we then attempted this transformation
under blue LED light (465 nm). Exposing 3a and 1,3,5-
trimethoxybenzene to blue LED light afforded the
monotrifluoromethylated product in 16% yield, showing that
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
Catalytic trifluoromethylation. After the observation that
added oxidants improved the reaction outcome, we believed that
we could develop a workable catalytic protocol for selective C–H
trifluoromethylation of heteroarenes with a cheap oxidant and a
readily available CF3-source. Considering the stability of our NiIII
complexes under air, we were also looking to possibly develop a
robust system for photoinduced trifluoromethylation that could
work under ambient atmosphere.
Using 10 mol% of 3a and N-phenylpyrrole as a substrate in
the presence of 3 equiv of sodium triflinite, CF3SO2Na, as a CF3
source, and 1 equiv of K2S2O8 as an oxidant under blue LED
irradiation, afforded the mono-trifluoromethylated pyrrole product
in 85% yield after 24 h at RT (Table 2, entry 1), by contrast to the
16% yield obtained in the presence of ambient light under
otherwise identical conditions. The yield did not change
dramatically in the presence of K2HPO4 base (87% yield) (entry
3). Catalyst 3a was the most efficient compared to 1a (40%) and
2a (67%) (entries 1, 4, and 5). As expected, under these
conditions
the
background,
non-photoinduced
radical
trifluoromethylation also occurred;[33] however, the reaction in the
absence of Ni complexes afforded only 30% of
trifluoromethylation product, indicating that the presence of 3a
leads to significantly higher yields and more efficient
trifluoromethylation (entry 6). Interestingly, when the reaction was
performed under an N2 atmosphere, the yield decreased to 62%
(entry 7). Considering that complex 3a was formed from 3 under
aerobic conditions, we hypothesize that air could play a role in
recycling the catalytically relevant NiIII species. Other solvents,
MeOH and MeCN, led to lower yields (entries 8 and 9).
Table 2. Optimization of reaction conditions for C-H trifluoromethylation using
CF3SO2Na/oxidant as a CF3-source.[a]
Scheme 3. Ni–CF3 bond homolysis in 3a. For experiments under light irradiation,
14 W blue LED lamp (465 nm) was used.
Entry
Catalyst
3a
Solvent, conditions
DMSO
Yield [%][b]
1
2
3
4
5
6
7
8
9
85
16
87
40
67
30
62
15
6
3a
Ambient light, DMSO
K2HPO4 (1 equiv), DMSO
DMSO
3a
1a
2a
DMSO
none
3a
DMSO
DMSO, under N2
MeOH
3a
3a
MeCN
[a] The reaction was performed under air unless indicated otherwise. [a] NMR
yield was based on integration against α,α,α-trifluorotoluene as an internal
standard.
Figure 8. Experimental (black line) and simulated (red line) EPR spectra of the
solution of 3a exposed blue LED irradiation for 2 min in the presence of PBN
spin trap in DMSO solution. Simulation parameters: AN = 14.28 G; AH = 2.11 G,
AF = 1.54, g = 2.0075.
Other oxidants were also tested in this reaction, however,
using KHSO5-K2SO4 led to a notably lower yield (32%), while
tBuOOH resulted in an exothermic reaction. Another
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
trifluoromethyl group source, CF3SO2Cl in the presence of K2S2O8, Umemoto reagent as a limiting reagent with 3 equiv of 3-
was also less efficient, giving a 40% NMR yield of the
trifluoromethylation product (see SI).
methylindole (Table S4, entry 10), similar to the protocol used by
Sanford et al. for Ni-catalyzed trifluoromethylation with 2,8-
difluoro-5-(trifluoromethyl)-5H-dibenzo[b,d]thiophen-5-ium
trifluoromethanesulfonate.[16a] We found that trifluoromethylation
also occurs in MeOH, MeCN, and acetone, albeit in lower yields
(56-72%) when compared to DMSO. The trifluoromethylation
under nickel-free conditions, in the presence of 10 mol% of L3,
showed no appreciable catalytic activity for the free ligand alone
(Table S5 in the SI).
Finally, our optimized protocol was applied to a range of
electron-rich arenes (m-t) and pharmaceutically active molecules
(Scheme 5). For example, tadalafil underwent trifluoromethylation
at the 7-position of the indole backbone to afford the same
fluorinated analog r earlier reported by the Sanford group using 5
equiv of substrate relative to the Umemoto reagent, however
using excess substrate was not required in our protocol leading to
higher arene conversion. Melatonin, a well-studied natural
hormone, reacted to afford a single trifluoromethylated product (q).
A natural product amino acid, Cbz-L-tryptophan, was also well-
tolerated (p), demonstrating our protocol’s potential utility in the
late-stage functionalization of pharmaceutically active molecules.
For several substrates such as resorcinol and indole, a minor
monotrifluoromethylation isomer was present (see t and u in
Scheme 5 with arrows indicating alternative trifuoromethylation
sites), while for other challenging substrates the amount of
alternative product did not generally exceed 5%.
The optimized protocol with 3a was used for a number of
pyrroles (Scheme 4, entries a-c), pyrimidines (entries i-j), and
some electron-rich simple arenes (entries d-h) successfully to
give C-H trifluoromethylation products in ca. 40-78% isolated
yields. However, the protocol was not successful in the case of
relatively more electron-poor pyridines and other electron-
deficient arenes, and non-substituted indole gave a mixture of 2-
and 3-trifluoromethylated isomers (see SI, Scheme S7). A further
list of substrates that we found worked poorly under this (and the
electrophilic CF3 source protocol vide infra) is given in the SI.
Scheme 4. Scope of radical trifluoromethylation catalyzed by 3a in the presence
of CF3SO2Na.
Considering the limited substrate scope and non-negligible
contribution from
trifluoromethylation of our protocol, our next target was to achieve
more selective and higher-yielding C–H bond
a
non-catalyzed background radical
a
trifluoromethylation of a wider range of substrates by utilizing the
Umemoto reagent A as both a CF3 source and oxidant. Although
Umemoto reagents are relatively expensive, the ability to
selectively trifluoromethylate C–H bonds using a simple and
accessible Ni catalyst and a versatile reaction protocol (i.e., using
either substrate or Umemoto reagent as a limiting reagent) may
provide additional benefits when compared to existing
trifluoromethylation methods where an excess of a substrate
relative to the Umemoto reagent is always required. The resulting
low conversion means that existing methods present a significant
disadvantage for expensive substrates and natural products that
may only be available only in small amounts.
After initial optimization (Table S4) using 3-methylindole as a
model substrate, we found that trifluoromethylation catalyzed by
5 mol% of 3a in the presence of 3 equiv of Umemoto reagent gives
the product in 99% NMR and 92% isolated yields. Alternatively,
the same quantitative 99% NMR yield is obtained when using the
Scheme 5. Substrate scope of trifluoromethylation using Umemoto reagent.
Radical trap experiments. Although the Umemoto reagents
are typically used as electrophilic trifluoromethylating reagents,
examples of radical trifluoromethylation induced by single-
electron transfer (SET) are also known. For example, Akita and
co-workers described in detail a SET processes induced by light
irradiation in the presence of common Ru or Ir photoredox
catalysts, which led to one-electron reduction of the Umemoto
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
reagent and generation of a CF3 radical, which could further react
with unsaturated substrates.[35]
We thus set out to investigate whether CF3 radical formation
occurs under typical conditions for 3a-catalyzed C–H
trifluoromethylation in the presence of Umemoto reagent A, by
using common radical traps (Scheme 6). The reaction between 3-
methylindole and A catalyzed by 3a under blue LED light was
performed in the presence of TEMPO as the radical trap, and led
to the formation of TEMPO-CF3, while product formation was
inhibited. The formation of the trifluoromethyl radical was further
confirmed via trapping it with 1,1-diphenylethylene (DPE) in 71%
yield when analyzed by 19F NMR spectroscopy and confirmed by
HRMS analysis. Similarly, EPR spectroscopy was used to detect
that PBN could be used to trap the CF3 radical under catalytic
conditions.
Scheme 7. Reactivity of 3 with the Umemoto reagent A. For experiments under
light irradiation, 14 W blue LED lamp (465 nm) was used.
Based on these preliminary observations, we propose the
following mechanism (Scheme 8). The initiation step involves NiIII-
CF3 homolysis from catalyst 3a to generate NiII complex B and a
CF3 radical; the latter then reacts with an aromatic substrate C to
form intermediate D. Oxidation/deprotonation of D by another
molecule of 3a gives the trifluoromethylation product E; a free
ligand equivalent released during reduction of 3a to 3 may assist
in deprotonation of D. One-electron reduction of A with 3 then
generates a CF3 radical which can further react directly with the
aromatic substrate, or alternatively may react to regenerate 3a
from B. The protonated ligand is likely able to recombine with the
complex during these 1e- redox transformations since the solvent,
DMSO, can assist with deprotonation.
Thus, we propose that 3a acts as an initiator, which might lead
to its partial decomposition, although further reactivity with
Umemoto reagent might lead to some catalyst recovery. However,
we have also shown that the catalytic cycle may be initiated
directly by using 3 in catalytic amounts (Scheme 7, a) leading to
a qualitatively similar result and confirming that the catalytic cycle
could operate via NiII/NiIII redox transformations. Further studies
are underway to clarify the role of the Ni complexes in different
oxidation states in photocatalyzed trifluoromethylation.
Scheme 6. Radical trap experiments under catalytic conditions. For
experiments under light irradiation, 14 W blue LED lamp (465 nm) was used.
With the Umemoto protocol (where A can also act as an
oxidant), we found that parent NiII complex 3 can also now act as
a catalyst for the trifluoromethylation of 3-methylindole in the
presence of 3 equiv of A under the standard conditions to give the
trifluoromethylated product in 64% yield (Scheme 7, a).
Notably, this reactivity also involved CF3 radical formation.
The PBN-CF3 adduct formed in the presence of PBN and
substrate (Scheme 7, b). In the absence of substrate, we were
also able to confirm the formation of CF3 radical in the presence
of 3, 3 equiv of A, and TEMPO or PBN radical trap under blue
LED irradiation, to give TEMPO-CF3 or PBN-CF3 adduct
respectively (Scheme 7, c and d). This reactivity required light and
no reaction was observed between 3 and A in the dark. The
formation of 3a by oxidation of 3 with A was also confirmed by the
appearance of the characteristic purple color, and the resultant
matching UV-vis absorption band, as well as ESI-HRMS, both in
the absence and in the presence of 5 equiv of HBF4.
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
mechanism and that a NiIV oxidation state may not be involved.
Considering
the
increased
interest
in
late-stage
trifluoromethylation of natural products, future work will
investigate the use of aliphatic substrates for trifluoromethylation
involving NiIII-CF3 complexes.
Acknowledgements
R. R. F. performed crystal structure determination within the
statements for Kazan Scientific Center of RAS. We thank
Instrumental Analysis Section and Engineering Support Section
for technical support and Scientific Computing and Data Analysis
Section for access to HPC Cluster. The authors acknowledge the
Okinawa Institute of Science and Technology Graduate University
for funding.
Keywords: nickel • trifluoromethylation • C-H bond
functionalization • radical formation • photoreactivity
[1]
a) V. P. Ananikov, ACS Catal. 2015, 5, 1964-1971; b) S.
Z. Tasker, E. A. Standley, T. F. Jamison, Nature 2014,
509, 299-309.
[2]
[3]
[4]
J. P. Stambuli, R. Kuwano, J. F. Hartwig, Angew. Chem.
Int. Ed. 2002, 41, 4746-4748.
W. Guo, A. Cai, J. Xie, A. W. Kleij, Angew. Chem. Int. Ed.
2017, 56, 11797-11801.
a) E. J. Cho, T. D. Senecal, T. Kinzel, Y. Zhang, D. A.
Watson, S. L. Buchwald, Science 2010, 328, 1679; b) A.
R. Mazzotti, M. G. Campbell, P. Tang, J. M. Murphy, T.
Ritter, J. Am. Chem. Soc. 2013, 135, 14012-14015; c) M.
H. Katcher, A. Sha, A. G. Doyle, J. Am. Chem. Soc. 2011,
133, 15902-15905.
Scheme 8. Proposed radical pathway for trifluoromethylation using Umemoto
reagent A.
[5]
[6]
a) X. Hu, Chem. Sci. 2011, 2, 1867-1886; b) V. B.
Phapale, D. J. Cárdenas, Chem. Soc. Rev. 2009, 38,
1598-1607; c) V. Snieckus, C. Allais, Synfacts 2017, 13,
0138-0138; d) X. Gao, Y. Geng, S. Han, A. Liang, J. Li, D.
Zou, Y. Wu, Y. Wu, Tetrahedron Lett. 2018, 59, 1551-
1554.
a) J. R. Bour, N. M. Camasso, M. S. Sanford, J. Am.
Chem. Soc. 2015, 137, 8034-8037; b) N. M. Camasso, M.
S. Sanford, Science 2015, 347, 1218; c) N. Nebra,
Molecules 2020, 25.
F. Le Vaillant, E. J. Reijerse, M. Leutzsch, J. Cornella, J.
Am. Chem. Soc. 2020, 142, 19540-19550.
B. Zheng, F. Tang, J. Luo, J. W. Schultz, N. P. Rath, L. M.
Mirica, J. Am. Chem. Soc. 2014, 136, 6499-6504.
E. Chong, J. W. Kampf, A. Ariafard, A. J. Canty, M. S.
Sanford, J. Am. Chem. Soc. 2017, 139, 6058-6061.
a) Y. Aihara, N. Chatani, J. Am. Chem. Soc. 2014, 136,
898-901; b) Z. Ruan, S. Lackner, L. Ackermann, Angew.
Chem. Int. Ed. 2016, 55, 3153-3157; c) J. Xu, L. Qiao, J.
Shen, K. Chai, C. Shen, P. Zhang, Org. Lett. 2017, 19,
5661-5664; d) S.-K. Zhang, R. C. Samanta, N.
Sauermann, L. Ackermann, Chem. Eur. J. 2018, 24,
19166-19170; e) T. T. Tsou, J. K. Kochi, J. Am. Chem.
Soc. 1978, 100, 1634-1635; f) M. Rovira, S. Roldán-
Gómez, V. Martin-Diaconescu, C. J. Whiteoak, A.
Company, J. M. Luis, X. Ribas, Chem. Eur. J. 2017, 23,
11662-11668.
The proposed mechanism in Scheme
8 implies that
acidification of the reaction mixture should occur since no base
was added. This was confirmed by pH measurements after the
reaction, showing a weakly acidic media due to acid build-up. A
similar effect could also be present in other reported examples of
Ni-catalyzed trifluoromethylation using Umemoto-type reagents.
Although this could eventually limit the overall TON, complex 3a
shows good stability even in the presence of a large excess of
strong acid (up to 5200 equiv) under ambient conditions.
Overall, although the mechanism of trifluoromethylation
catalyzed by 3a will need to be further investigated in greater
detail, these preliminary experiments show that the Umemoto
reagent likely serves as the CF3 radical source, rather than an
electrophilic reagent, and light-induced one-electron redox
transformations mediated by L3-supported Ni complexes are
involved in the catalytic turnover.
[7]
[8]
[9]
[10]
Conclusion
In summary, we developed three bench-stable NiIII-CF3
complexes with a simple naphthyridine ligand using air as an
oxidant. We showed that using blue light promotes catalytic C–H
trifluoromethylation of heterocycles using either a radical or an
electrophilic CF3 source. This method is easy to set up, does not
require a complicated or expensive protocol, and it is also efficient
and can be performed under mild conditions. Preliminary
mechanistic studies suggest that catalysis proceeds via a radical
[11]
[12]
C. Zhang, Org. Biomol. Chem. 2014, 12, 6580-6589.
L. Amini-Rentsch, E. Vanoli, S. Richard-Bildstein, R. Marti,
G. Vilé, Ind. Eng. 2019, 58, 10164-10171.
S. Purser, P. R. Moore, S. Swallow, V. Gouverneur,
Chem. Soc. Rev. 2008, 37, 320-330.
a) K. Müller, C. Faeh, F. Diederich, Science 2007, 317,
1881; b) W. K. Hagmann, J. Med. Chem. 2008, 51, 4359-
4369.
D. A. Nagib, D. W. C. MacMillan, Nature 2011, 480, 224-
228.
[13]
[14]
[15]
9
This article is protected by copyright. All rights reserved.

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
[16]
[17]
[18]
a) E. A. Meucci, S. N. Nguyen, N. M. Camasso, E. Chong,
A. Ariafard, A. J. Canty, M. S. Sanford, J. Am. Chem. Soc.
2019, 141, 12872-12879; b) X. Gao, Y. Geng, S. Han, A.
Liang, J. Li, D. Zou, Y. Wu, Y. Wu, Org. Lett. 2018, 20,
3732-3735.
a) G. G. Dubinina, W. W. Brennessel, J. L. Miller, D. A.
Vicic, Organometallics 2008, 27, 3933-3938; b) B. Vabre,
P. Petiot, R. Declercq, D. Zargarian, Organometallics
2014, 33, 5173-5184; c) J. Hao, B. Vabre, D. Zargarian,
Organometallics 2014, 33, 6568-6576.
F. D'Accriscio, P. Borja, N. Saffon-Merceron, M. Fustier-
Boutignon, N. Mézailles, N. Nebra, Angew. Chem. Int. Ed.
2017, 56, 12898-12902.
[19]
[20]
S. T. Shreiber, D. A. Vicic, Angew. Chem. Int. Ed. 2021,
https://doi.org/10.1002/anie.202104559.
a) S. K. Kariofillis, A. G. Doyle, Acc. Chem. Res. 2021, 54,
988-1000; b) B. J. Shields, A. G. Doyle, J. Am. Chem.
Soc. 2016, 138, 12719-12722; c) Z. Zuo, D. T. Ahneman,
L. Chu, J. A. Terrett, A. G. Doyle, D. W. C. MacMillan,
Science 2014, 345, 437; d) C. L. Joe, A. G. Doyle, Angew.
Chem. Int. Ed. 2016, 55, 4040-4043; e) D. T. Ahneman, A.
G. Doyle, Chem. Sci. 2016, 7, 7002-7006.
[21]
a) S. Deolka, O. Rivada-Wheelaghan, S. L. Aristizábal, R.
R. Fayzullin, S. Pal, K. Nozaki, E. Khaskin, J. R.
Khusnutdinova, Chem. Sci. 2020, 11, 5494-5502; b) O.
Rivada-Wheelaghan, S. L. Aristizábal, J. López-Serrano,
R. R. Fayzullin, J. R. Khusnutdinova, Angew. Chem. Int.
Ed. 2017, 56, 16267-16271; c) O. Rivada-Wheelaghan, A.
Comas-Vives, R. R. Fayzullin, A. Lledós, J. R.
Khusnutdinova, Chem. Eur. J. 2020, 26, 12168-12179.
a) A. Okuniewski, D. Rosiak, J. Chojnacki, B. Becker,
Polyhedron 2015, 90, 47-57; b) D. Rosiak, A. Okuniewski,
J. Chojnacki, Polyhedron 2018, 146, 35-41.
[22]
[23]
[24]
Y. Yamaguchi, H. Ichioka, A. Klein, W. W. Brennessel, D.
A. Vicic, Organometallics 2012, 31, 1477-1483.
Deposition Numbers 2092535-2092540 contain the
supplementary crystallographic data for this paper. These
data are provided free of charge by the joint Cambridge
Crystallographic Data Centre and
Fachinformationszentrum Karlsruhe Access Structures
service www.ccdc.cam.ac.uk/structures.
[25]
C.-P. Zhang, H. Wang, A. Klein, C. Biewer, K. Stirnat, Y.
Yamaguchi, L. Xu, V. Gomez-Benitez, D. A. Vicic, J. Am.
Chem. Soc. 2013, 135, 8141-8144.
[26]
[27]
[28]
D. E. Hamilton, R. S. Drago, J. Telser, J. Am. Chem. Soc.
1984, 106, 5353-5355.
E. Lamour, S. Routier, J.-L. Bernier, J.-P. Catteau, C.
Bailly, H. Vezin, J. Am. Chem. Soc. 1999, 121, 1862-1869.
S. Lapointe, E. Khaskin, R. R. Fayzullin, J. R.
Khusnutdinova, Organometallics 2019, 38, 4433-4447.
J.-D. Chai, M. Head-Gordon, PCCP 2008, 10, 6615-6620.
F. Weigend, R. Ahlrichs, PCCP 2005, 7, 3297-3305.
C. F. Harris, C. S. Kuehner, J. Bacsa, J. D. Soper, Angew.
Chem. Int. Ed. 2018, 57, 1311-1315.
[29]
[30]
[31]
[32]
[33]
J. R. Khusnutdinova, N. P. Rath, L. M. Mirica, J. Am.
Chem. Soc. 2010, 132, 7303-7305.
a) B. R. Langlois, E. Laurent, N. Roidot, Tetrahedron Lett.
1991, 32, 7525-7528; b) Y. Ji, T. Brueckl, R. D. Baxter, Y.
Fujiwara, I. B. Seiple, S. Su, D. G. Blackmond, P. S.
Baran, PNAS 2011, 108, 14411; c) Y. Fujiwara, J. A.
Dixon, F. O’Hara, E. D. Funder, D. D. Dixon, R. A.
Rodriguez, R. D. Baxter, B. Herlé, N. Sach, M. R. Collins,
Y. Ishihara, P. S. Baran, Nature 2012, 492, 95-99.
a) L. D. Haire, P. H. Krygsman, E. G. Janzen, U. M.
Oehler, J. Org. Chem. 1988, 53, 4535-4542; b) E. G.
Janzen, B. J. Blackburn, J. Amer. Chem. Soc. 1968, 90,
5909-5910; c) M. Danilczuk, F. D. Coms, S. Schlick, Fuel
Cells 2008, 8, 436-452.
[34]
[35]
T. Koike, M. Akita, Acc. Chem. Res. 2016, 49, 1937-1945.
10
This article is protected by copyright. All rights reserved.

10.1002/anie.202109953
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
Nickel complexes with simple naphthyridine ligands undergo facile aerobic oxidation to stable, isolated Ni(III) organometallic
complexes, in which Ni-CF3 bond homolysis is observed upon blue LED light irradiation. These complexes catalyze photoinduced C-
H bond trifluoromethylation in arenes and heteroarenes.
Institute and/or researcher Twitter usernames:@KhusnutdinovaU
11
This article is protected by copyright. All rights reserved.