ACS Catalysis
Research Article
fresh catalyst (Figure S16). The abundance of nitrogen atoms
renders the material as an effective polydentate ligand to
coordinate nickel atoms. Coordination with the covalent
carbon nitride mpg-CN gives rise to a 16-electron Ni(II)
chelate complex compared to that of a more reactive 14-
electron Ni(II)-amide complex in the case of ionic carbon
nitrides (Figure S26). These observations indicate that the
problem of nickel black formation in dual Ni/photoredox
catalysis (in addition to adjusting the rates of oxidative
addition and reductive elimination by tuning the energy of
incident light and the concentration of reagents)39 could also
be eliminated by using a robust carbon nitride photocatalyst
that stabilizes the low valent nickel species without altering the
overall reaction rates. The recovered mpg-CN showed a slight
shift of the absorption onset in the DRUV−vis spectrum and
an expansion of the optical band gap by ∼0.05 eV (Figure
S20). In steady-state photoluminescence (PL), surface
modification of mpg-CN is observed as a blue shift in
The morphology of the recovered mpg-CN particles
adopted a more rounded shape compared to that of a rougher
surface of freshly prepared mpg-CN (Figure 2d). Taken
together, the postcharacterization data of the recovered mpg-
CN clearly shows robustness, stability, and durability of this
heterogeneous organic semiconductor as a photocatalyst.
The aforementioned results provide evidence for catalytic
generation of HAT agent on the mpg-CN semiconductor
surface, which operates in synergy with nickel catalysis to
enable C(sp3)−H bond functionalization. However, the
photoactivity of mpg-CN/nickel dual catalytic system for C−
H bond cleavage at a molecular level remains to be elucidated.
Based on prior mechanistic investigations of dual Ni/
photocatalyzed cross-coupling reactions in homogeneous
systems,41 we postulate two plausible mechanistic scenarios
as depicted in Scheme 1. Initially, the Ni(0) complex I
undergoes oxidative addition with an aryl halide delivering
Ni(II) oxidative addition complex II. Concurrently, light
absorption by the mpg-CN semiconductor photocatalyst
triggers the charge separation producing two-dimensional
surface redox centers as electron−hole pairs. In pathway A,
SET oxidation of complex II by the photogenerated hole
(VBM located at +1.2 V vs SCE, E1/2 (NiII/NiIII) = +0.85 V vs
SCE) affords species III, which may undergo Ni(III)−X
homolysis to give a halogen radical and Ni(II) species IV.28
The resulting halogen radical can rapidly abstract a hydrogen
atom from DMA (H−Br BDE ∼88 kcal/mol, H−Cl BDE
∼102 kcal/mol, α-amino C−H BDE ∼89−94 kcal/mol),27
which immediately recombines with species IV to form V.
Subsequent reductive elimination of V results in the desired
product and Ni(I) species VI. Finally, reduction of VI by the
electron located on the semiconductor surface (CBM located
at −1.5 V vs SCE, E1/2 (NiI/Ni0) = −1.42 V vs SCE)41
regenerates Ni(0) and completes the catalytic cycle. In
pathway B, mpg-CN serves as a light-absorbing antenna
undergoing an energy-transfer process (EnT) (singlet−triplet
band gap ca. 0.39 eV)42 to produce electronically excited
Ni(II) species VII. Homolysis of the Ni(II)−X bond and HAT
followed by a rebound of the resulting carbon-centered radical
with VIII generates Ni(II) species IX. Reductive elimination
from the electronically excited species X, promoted by EnT
with mpg-CN, provides the final product and regenerates
Ni(0) species, thus completing the catalytic cycle.
Control experiments with a catalytic amount of preformed
[(bpy)NiII(o-tolyl)Br] complex II in our standard reaction
conditions (Table 5) show that both complex II and mpg-CN
Table 5. Catalytic Experiments with Preformed
a b
,
[(bpy)NiII(o-tolyl)Br] Complex II
entry
deviation from the above condition
none
yield (%)
1.
2.
5
56
with mpg-CN (10 mg/mL)
a
b
With 5 mol % II. Yields were determined by GC-FID using 1,4-
dimethoxybenzene as an internal standard.
are required in the productive reaction pathway. A major
distinction between these two pathways, as shown in Scheme
1, is that EnT process involves the excited state of Ni(II)
species, which should be directly accessible via visible light
excitation in the absence of photocatalyst (Figure S27).
Indeed, a stoichiometric experiment with [(bpy)NiII(o-tolyl)-
Br] complex II via direct photoexcitation at 450 nm (i.e.,
without mpg-CN) revealed the formation of the desired
product in appreciable yield (Table 6). An additional control
Table 6. Stoichiometric Experiments with Preformed
a b
,
[(bpy)NiII(o-tolyl)Br] Complex II
entry
light source
yield
1.
2.
3.
340 nm
450 nm
dark
17%
20%
ND
a
b
by GC-FID using 1,4-dimethoxybenzene as an internal standard.
experiment performed under identical conditions in the
absence of light failed to produce any detectable product.
These experiments are indicative toward pathway B where the
formation of an electronically excited Ni(II) complex is a
prerequisite for bond formation.
The reaction kinetics were monitored at different concen-
trations of the nickel complex as well as aryl bromide. In all
cases, we observed a zero-order kinetic profile (Figure 3) at
room temperature (Table S14), which is a typical behavior of
heterogeneous catalytic reactions in which active sites are
saturated by adsorbed reactant molecules. Similar observations
have been reported previously for carbon nitride/nickel dual
photocatalysis.43 Interestingly, increasing the [Ni] loading by
three times increased the relative amount of debrominated
product and halved the reaction rate (Figure 3a vs b). This
could be ascribed to saturating the mpg-CN active sites while
having more [Ni] complexes in solution. As a result, the [Ni]
complex, which is involved in the rate-determining step, has to
compete with other off-cycle or non-rate-determining on-cycle
[Ni] complexes for mpg-CN active sites leading to the overall
1600
ACS Catal. 2021, 11, 1593−1603