Mechanisms
of
Photoredox
Catalysis
Featuring
Nickel
−
Bipyridine
Complexes
David
A. Cagan,
Daniel
B
í
m,
Nathanael
P. Kazmierczak,
and Ryan
G. Hadt
*
Cite
This:
ACS
Catal.
2024,
14, 9055−9076
Read
Online
ACCESS
Metrics
& More
Article
Recommendations
ABSTRACT:
Metallaphotoredox
catalysis
can unlock useful path-
ways for transforming
organic
reactants
into desirable
products,
largely
due to the conversion
of photon
energy
into chemical
potential
to drive redox and bond transformation
processes.
Despite
the importance
of these processes
for cross-coupling
reactions
and
other transformations,
their mechanistic
details are only superficially
understood.
In this review,
we have provided
a detailed
summary
of
various
photoredox
mechanisms
that have been proposed
to date for
Ni
−
bipyridine
(bpy) complexes,
focusing
separately
on photo-
sensitized
and direct excitation
reaction
processes.
By highlighting
multiple
bond transformation
pathways
and key findings,
we depict
how photoredox
reaction
mechanisms,
which ultimately
define
substrate
scope, are themselves
defined
by the ground-
and excited-
state geometric
and electronic
structures
of key Ni-based
intermediates.
We further
identify
knowledge
gaps to motivate
future
mechanistic
studies and the development
of synergistic
research
approaches
spanning
the physical,
organic,
and inorganic
chemistry
communities.
KEYWORDS:
cross-coupling
reactions,
metallaphotoredox
catalysis,
organic
transformations,
photoredox
catalysis,
mechanism,
electronic
structure
1.
INTRODUCTION
Pd-catalyzed
cross-coupling
reactions
have transformed
organic
chemistry
with their synthetic
contributions
to drug discovery
and development.
1
−
3
Although
subtle
differences
emerge
between
reactions,
the majority
of Pd-catalyzed
couplings
leverage
a mechanism
featuring
dominantly
two-electron
processes:
oxidative
addition,
transmetalation,
and reductive
elimination.
4
Going beyond
Pd, a precious
metal and limited
resource,
significant
strides
have been made toward
more
sustainable
approaches
to catalysis.
These advances
feature
critical
contributions
from methodology-driven
research
into
homogeneous
cross-coupling
catalysis
by first-row
transition
metal complexes,
which are becoming
more widely adopted
for
enabling
the construction
of new C
−
X (X = C, O, N, F, etc.)
bonds.
5
−
7
Mechanistic
studies
highlight
the complexities
of
these ground-state
cross-coupling
reactions,
but also bring to
light new possibilities
stemming
from one-electron
redox
processes
and the variety
of intermediates
involved
in the
underlying
bond-formation
and bond-rupture
processes.
8
Ni-mediated
catalysis
has emerged
as a key alternative
to Pd,
as it can access a range of formal oxidation
and/or
spin states
(Figure
1) and facilitate
numerous
complex
substrate
trans-
formations.
9
−
11
In addition
to metal redox, ligand-based
redox
(i.e., ligand
noninnocence
and potential
multireference
character)
12
−
15
further
increases
reaction
complexity
by
providing
important,
yetpoorlyunderstood,
electronic
structure
contributions.
These can result in noble-metal-like
reactivity
in
base-metal
catalysts
and provide
a basis for transformative
structure/function
relationships.
7
Metallaphotoredox
catalysis
has had a profound
influence
on
many areas of organic
chemistry,
including
cross-coupling
reactions.
This approach
uses photosensitizers
to generate
metal-based
intermediates
thatcanbeactiveindarkcycles.
16
−
22
Theseintermediates
oftenformduetotheirpropensity
forsingle
electron
transfer
(SET).
23
Photosensitizers
can additionally
transfer
energy
to metal complexes
to form reactive
excited
states.
24,25
The merger
of photoredox
catalysis
with Ni
−
bipyridine
(bpy) complexes
has claimed
a prominent
place in
the organic,
inorganic,
and physical
chemistry
communities
owing to its wide synthetic
utility and rich photophysical
aspects.
20,26
−
32
In addition
to light absorption
by the photo-
Received:
April 5, 2024
Revised:
May 7, 2024
Accepted:
May 16, 2024
Published:
May 29,
2024
Review
pubs.acs.org/acscatalysis
© 2024
The Authors.
Published
by
American
Chemical
Society
9055
https://doi.org/10.1021/acscatal.4c02036
ACS
Catal.
2024,
14, 9055
−
9076
This article is licensed under CC-BY 4.0
sensitizers
present
in reaction
mixtures
(often cyclometalated
Ir(III) heteroleptic
complexes
33
−
35
), these Ni
−
bpy
cocatalysts
also absorb strongly
across the UV
−
vis
region and can directly
harvest
light to access key excited
states.
22,36
−
40
In principle,
ultrafast
spectroscopic
methods
should be critical to studying
the photophysical
processes
that undergird
the overall chemical
bondtransformations.
41
However,
asdiscussed
below, thereare
oftenstrongly
competing
intramolecular
excited-state
relaxation
pathways,
and care needs to be taken to account
for low
quantum
yield processes
that can be difficult
to probe directly
usingtime-resolved
spectral
methods.
Overall,
theelucidation
of
mechanistic
routes requires
the knowledge
of both light- and
thermally
driven components
and the interplay
between
them.
Asdiscussed
below,thishasproventobeadifficult
taskforlight-
driven,
Ni-mediated
catalytic
cycles, and our overall
under-
standing
ofhowphoton
energydrivesorganic
transformations
is
still superficial.
The aforementioned
progress
motivates
further
efforts to
elucidate
the geometric
and electronic
structures
of critical
inorganic
species
and photoinduced
states that are involved
in
metallaphotoredox
cross-coupling
reactions.
We believe
these
knowledge
gaps can be addressed
by a synergistic
combination
ofsynthesis,
spectroscopy,
andcomputation
todefineelectronic
structure
contributions
to reactivity,
and we hold that there is
significant
general
potential
linked to leveraging
these complex-
ities for cross-coupling
catalysis.
To do so, however,
significant
strides
need to be made toward
detailed
and fundamental
studies of discrete
light and dark reaction
steps that constitute
photoredox
catalytic
cycles.
Ultimately,
in concert
with
additional
methodological
studies,
this understanding
will help
informchemists
howtoleverage
theinherent
properties
offirst-
row transition
metals and, thus, guide academic
and industrial
research
toward sustainable
approaches
for bond constructions
in organic
synthesis.
While previous
reviews
have highlighted
the tremendous
advancements
made in the development
of new photoredox-
enabled
transformations,
19,20,31
this review seeks to compare
and evaluate
mechanisms
that have been proposed
in the
literature,
with a focus on Ni
−
bpy
complexes.
We note that
additives
can influence
the catalytic
pathway.
However,
mechanistic
analysis
of their contributions
is quite limited.
Thus, while potentially
important
to consider,
this review does
Figure
1.
Qualitative
molecular
orbitalcorrelation
diagram
offourNi(II)
−
bpy
speciesofpotential
relevance
inphotocatalytic
pathways;
eachfeature
distinct
geometric
andelectronic
structures,
ligandfieldsplitting
energies
(
Δ
), and
σ
*
effects.TheNi(II)
−
bpy
aryl halide(A) adoptsasquareplanar
(
D
4h
) geometry,
leading
to a diamagnetic
S
= 0 ground
state. The high spin
S
= 1 geometry
(B) is observed
as a relaxed
excited-state
intermediate;
population
of the 3
d
(x
2
−
y
2
) orbital induces
a rotation
into a pseudo-
T
d
geometry.
Ni(II)
−
bpy
dihalide
(C) is stable as a
T
d
triplet ground
state. For
completeness,
wealsoshowthiscomplex
inasquareplanargeometry
(D).Thissingletstateisenergetically
disfavored
andyettobeidentified
todate.
Selectmolecular
orbitals
(computed
withDFTattheB3LYP/def2-TZVP
48
−
50
leveloftheory)
aredepicted
atthetopofthefigureforillustration
of
σ
*
effects.
ACS
Catalysis
pubs.acs.org/acscatalysis
Review
https://doi.org/10.1021/acscatal.4c02036
ACS
Catal.
2024,
14, 9055
−
9076
9056
not provide
a complete
picture
of their potential
mechanistic
roles. Given the growing
importance
of ground-
and excited-
state processes
in metallaphotoredox
catalysis,
the review first
features
a brief electronic
structure
primer,
which discusses
key
aspectsofdifferent
electronic
statesofNiatabroadly
accessible
level. We subsequently
provide
a summary
and comparison
of
proposed
photoredox
mechanisms.
Divided
into two main
sections,
we first summarize
mechanisms
featuring
key photo-
sensitization
steps. Secondly,
we discuss
mechanisms
that
feature
direct excitation
of Ni-based
species
for bond-
homolysis-driven
dark cycle initiation
or excited-state
bond
formation
reactions.
The mechanistic
summaries
are further
bolstered
by“KeyConsideration”
sections
designed
tohighlight
the importance
of Ni-based
intermediates
and their electronic
structures.
By doing so, we hope to 1) demonstrate
the
importance
and need for further
mechanistic
studies
of
metallaphotoredox
reactions,
even beyond
Ni, and 2) highlight
the interdisciplinary
nature
of this growing
area, hopefully
motivating
future synergistic
contributions
that will span the
physical,
organic,
and inorganic
chemistry
communities.
2.
NICKEL
ELECTRONIC
STRUCTURE
PRIMER
Prior to embarking
on our review of light-activated
catalytic
cycles featuring
Ni complexes,
it is valuable
to consider
the
distinct
electronic
structures
of the commonly
invoked
Ni
intermediates.
Even within a given oxidation
state, such as
Ni(II),
disparate
geometries,
spin states,
and ligand field
strengths
can lead to unique
properties
for different
species.
22
These changes
have direct implications
for evaluating
the
plausibility
of ground-
and excited-state
reactivity,
including
mechanistic
steps such as light harvesting,
energy/electron
transfer,
and electrophile
activation.
Nickel is most stable in the 2+ oxidation
state with a
d
8
electron
configuration.
42
Many stable four-coordinate
Ni(II)
species
are known,
and several feature
prominent
roles in the
mechanisms
outlined
below. Given the importance
of these
species
and their reactivity,
we consider
their geometric
and
electronic
structures
atlength.Asthisreviewfocuses
onNi
−
bpy
complexes,
weassume
thattwoofthefourcoordination
sitesare
occupied
by the bpy ligand.
Charge
balance
requires
the
remaining
two ligands
be anionic.
Common
options
in the
context
of cross-coupling
include
aryl and halide ligands,
which
couldpotentially
bearranged
ineitherasquareplanar(
D
4h
)ora
pseudotetrahedral
(
T
d
) geometry
(Figure
1). These options
are
not independent
of the ligand character,
as described
below.
In the square planar geometry,
the vast majority
of the
σ
*
character
is concentrated
in the 3
d
(x
2
−
y
2
) orbital, resulting
in a
largeligandfieldsplitting
energy,
Δ
(Figure
1).
43
Thissplitting
is
greater
than the electron
−
electron
repulsion
(i.e., spin pairing
energy)
incurred
by having the seventh
and eighth electrons
occupying
the same orbital (the 3
d
(z
2
) orbital here). Thus, it is
more energetically
favorable
to adopt a low-spin,
S
= 0
configuration
with a doubly
unoccupied
3
d
(x
2
−
y
2
) orbital.
If
the ligands
are rotated
into a pseudo-
T
d
geometry,
multiple
nearly degenerate
orbitals
share the
σ
*
character,
leading
to a
small
Δ
relativetothesquareplanarcaseandahigh-spin,
S
=1
d
8
configuration.
(Note that calculations
of molecular
orbital
energies
for related
pseudo-
T
d
Ni(II) complexes
suggest
two
main
σ
*
orbitals,
as opposed
to the three
σ
*
-orbitals
found in a
perfect
tetrahedron.)
44,45
Accordingly,
population
of the
strongly
antibonding
3
d
(x
2
−
y
2
) orbital in the
D
4h
geometry
(such as through
metal-centered
photoexcitation)
induces
a
geometric
rotation
to the pseudo-
T
d
geometry
to minimize
the
σ
*
overlap.
Thechoicebetween
asquareplanarandpseudo-
T
d
geometry
can thus be understood
as a competition
between
electron
repulsion
(spin pairing
energy)
and the ligand field splitting
energy.
46
The
D
4h
S
= 0 state pays the energetic
penalty
for
pairingelectrons,
butitavoidspopulating
thehigh-lying
3
d
(x
2
−
y
2
) orbital and is therefore
unaffected
by larger values of
Δ
(Figure
1,
A
and
D
). On the other hand, the pseudo-
T
d
S
= 1
state avoids the energetic
penalty
for pairing
electrons
in the
same orbital,
yet it populates
both the 3
d
(z
2
) and 3
d
(x
2
−
y
2
)
orbitals,
which each experience
an energetic
disadvantage
according
to the magnitude
of
Δ
(Figure
1,
B
and
C
).
Accordingly,
strong-field
ligands
favor the square
planar
geometry,
while weak-field
ligands
favor the pseudo-
T
d
geometry.
47
Herein arises the essential
difference
between
aryl and halide
ligands.
From the perspective
of ligand field theory,
the aryl is
considered
a strong-field
ligand,
while halides
are weak-field
ligands.
51
As such, the gap between
the
σ
*
orbital(s)
and the
remaining,
lower-lying
3
d
-orbitals
willbelargeforaNi(II)
−
bpy
aryl halide species,
but comparatively
small for Ni(II)
−
bpy
dihalides.
For this reason,
Ni(II)
−
bpy
aryl halides
feature
singlet, square planar ground
states, whileNi(II)
−
bpy
dihalides
feature pseudo-
T
d
triplet ground
states. Note that pseudohalide
ligands(suchasalkoxides
oracetates)
resultinsimilarelectronic
structures
ashalides;
alkylligandsbehaveasaryls,butwithlarger
values for
Δ
, as they are stronger
σ
-donors.
For Ni(II)
−
bpy
aryl halides
vs Ni(II)
−
bpy
dihalides,
their
distinct
geometries
and spin states have significant
implications
for electron
transfer
in catalysis
owing to the divergent
energies
of the redox-active
molecular
orbital (RAMO).
For the ground-
state Ni(II)
−
bpy
aryl halide, the lowest unoccupied
molecular
orbital (LUMO)
is not metal-based.
The strong
σ
*
overlap
of
3
d
(x
2
−
y
2
) orbital raises its energy
above the bpy
π
*
orbital
manifold.
As such, the first reduction
event for Ni(II)
−
bpy
aryl
halide is observed
on the bpy ligand rather than the metal.
52,54
Reduction
of the complex
results in an anionic
[Ni(II)
−
bpy
•−
aryl halide]
−
, which slowly decomposes
to a three-coordinate
Ni(I)
−
bpy
aryl species.
52,53
When the aryl ligand is replaced
by
a halide, the reduction
in
σ
-donation
strength
and associated
antibonding
character
leads to a significant
decrease
in the
3
d
(x
2
−
y
2
) orbital
energy.
Furthermore,
as the ground-state
Ni(II)
−
bpy
dihalide
adopts
a pseudo-
T
d
geometry,
an addi-
tional stabilization
in the Ni-based
RAMO
is expected.
One-
electron
reduction
ofthiscomplex
affordsadoublyoccupied
σ
*
3
d
(z
2
) orbital,
resulting
in ejection
of a halide to give a Ni(I)
−
bpy halide complex.
55,56
Thereduction
potential
ofthecomplex
trendswiththeenergy
of the LUMO.
In Ni(II)(
t
‑
Bu
bpy)(
o
-tolyl)Cl,
the first reduction
event is found to be
−
1.6 V vs SCE, corresponding
to
electrochemically
reversible
bpy reduction.
Irreversible
Ni-
based reduction
appears
at
∼ −
1.8 V vs SCE.
57
By contrast,
the
first reduction
event for Ni(II)(
t
‑
Bu
bpy)Cl
2
is at
−
1.3 V vs SCE
(Ni-based
and irreversible).
55,56,58
The activity
of a Ni(II)
complex
toward
reductive
steps in a catalytic
cycle is
dramatically
influenced
byligandfieldstrength
andcoordination
geometry;
54,59,60
similar considerations
were also demonstrated
for
S
= 1 chiral enantioselective
Ni(II)
−
diimine
dihalide
cross-
coupling
catalysts.
44,61,62
Ligand
field analysis
of molecular
orbital
energies
indicates
the relative
plausibility
of various
catalytic
reduction
events.
ACS
Catalysis
pubs.acs.org/acscatalysis
Review
https://doi.org/10.1021/acscatal.4c02036
ACS
Catal.
2024,
14, 9055
−
9076
9057
In addition
to redox potentials,
the geometric
and electronic
structures
ofNi
−
bpy
complexes
determine
theirlightharvesting
ability through
the molar absorption
coefficients
of the UV
−
vis
transitions.
As a ligand with a significant
π
-conjugation,
the bpy
possesses
low-lying
π
*
-orbitals
capable
ofbackbonding
withthe
metal center.Thesebpy orbitals
serveasacceptors
for metal-to-
ligand
charge
transfer
(MLCT)
transitions
in the visible
absorption
spectrum
andpossess
significant
electron
delocaliza-
tion, leading
to a large transition
dipole moment.
Replacement
ofthebpyligandforaliphatic
N,N,N
′
,N
′
-tetramethylethylenedi-
amine (TMEDA)
exemplifies
this point, where only ligand field
bands become
possible,
leading
to reduced
values of
ε
(Figure
2).
38,63,64,66
Ni(II)
−
bpy
aryl halide complexes
exhibit MLCT
transitions
(350 nm
−
550
nm) that possess
molar extinction
coefficients
of comparable
magnitude
as iridium
photosensi-
tizers(
ε
=10
3
−
10
4
M
−
1
cm
−
1
),
36,38,65,67
rendering
theseNi(II)
species competitive
for photocatalytic
light harvesting.
Ni(II)
−
bpy dihalide
complexes
show orbitally
forbidden
ligand field
transitions
in the visible to near-infrared
region with
ε
= 10
1
−
10
2
M
−
1
cm
−
1
(Figure
2).
36
Similar analyses
may be conducted
for other oxidation
states.
Three-coordinate
Ni(I) complexes
adopt an approximately
planar
geometry;
while the 3
d
(x
2
−
y
2
)
σ
*
interaction
is
somewhat
lessened
due to the loss of 4-fold symmetry
and
consequent
orbital overlap,
there nonetheless
remains
a large
energetic
separation
between
the 3
d
(x
2
−
y
2
) orbital
and the
remainder
of the 3
d
-manifold
due to
σ
*
interactions
with the
bpy and
π
*
interactions
with the halide.
The
d
9
Ni(I)
configuration
implies
single occupation
of the high-energy
σ
*
-
orbital;
however,
this is tolerated,
and such Ni(I) compounds
have been characterized.
54,56,68
However,
further
reduction
of
Ni(I) to Ni(0) requires
the introduction
of an additional
electron
into the destabilized
σ
*
3
d
(x
2
−
y
2
) orbital.
The
reduction
potentials
for such an event are thought
to be high,
anditisunclear
whether
Ni(0)iscatalytically
accessible
69,70
(see
Reductive
SET mechanism
below).
Indeed,
Ni(0)
−
bpy
cyclo-
octadiene
(COD)
exhibits
a large degree of bpy ligand redox
noninnocence
and is proposed
to exist as Ni(I)(bpy
•−
)-
(COD).
71
Interestingly,
Ni(I)
−
bpy
halide complexes
exhibit
MLCT
transitions
across a wide wavelength
range (350 nm
−
1400 nm) and have molar extinction
coefficients
of equal or
greater
magnitude
than Ni(II)
−
bpy
aryl halides,
marking
yet
another
competitive
light-harvesting
species
in photocatalytic
cycles.
56
3.
SUMMARY
AND
COMPARISONS
OF
PROPOSED
PHOTOREDOX
MECHANISMS
Key consideration
sections
are provided
for each of the
mechanisms
summarized
herein,
with the goal of connecting
these considerations
to experimental
observations
that are
emphasized
across all Ni
−
bpy-based
photoredox
mechanisms,
both in terms of direct excitation
and photosensitization.
3.1.
Photosensitization.
3.1.1.
Reductive
SET.
The first
metallaphotoredox
reactions
using light-activated
nickel were
reported
independently
in 2014 by the groups of Molander
29
and Doyle and MacMillan,
30
where C(sp
2
)
−
C(sp
3
) cross-
couplings
were discovered
in reactions
combining
Ni(0)
−
bpy,
an Ir(III) photosensitizer,
and organic
coupling
partners.
The
reaction
scope was further
extended
to C(sp
2
)
−
C(sp
2
) and
C(sp
3
)
−
C(sp
3
) couplings
in 2015 and 2016, respectively,
72
−
74
thenfortheactivation
ofaliphatic
C
−
H bondsin2018,
75
andto
alkyl chloride
substrates
in 2019
76
and 2020;
77
enantioselective
cross-coupling
was seen a year later.
78
Based on a thermody-
namic redox potential
argument,
it was speculated
that the
iridium
excited
state,
*
Ir(III),
carried
out two separate
SET
events. This mechanism
is termed
“Reductive
SET” herein, as
the first (and only) proposed
interaction
between
iridium
and
nickel is a reduction
of Ni(I) to Ni(0) (Figure
3).
In the Reductive
SET mechanism,
the Ir(III) photosensitizer
is the sole excited-state
active species.
In one SET,
*
Ir(III)
oxidizes
the alkyl coupling
partner,
affording
C(alkyl)
•
and
Ir(II). In another
SET, Ir(II) reduces
a Ni(I)
−
bpy
halide
complex
(top box, Figure 3) to Ni(0)
−
bpy,
which can undergo
oxidative
addition
withanarylhalidetogenerate
asquare-planar
(
S
=0)Ni(II)
−
bpy
arylhalidecomplex
(bottom
box,Figure3).
This Ni(II)
complex
captures
the
*
Ir(III)-generated
alkyl
radical,
and the resultant
pentacoordinate
Ni(III)
species
undergoes
reductive
elimination
to form a Ni(I)
−
bpy
halide
and the C(sp
2
)
−
C(sp
3
) cross-coupled
product.
The cycle
continues
upon further reduction
of Ni(I)
−
bpy
halide by Ir(II)
to Ni(0)
−
bpy
and Ir(III).
3.1.2.
Key
Considerations
for
the
Reductive
SET
Mecha-
nism.
3.1.2.1.
Ir(III)
Acts
as the
Sole
Light-Harvesting
Species.
Figure
2.
UV
−
vis
absorption
spectra
of a common
Ir(III) photo-
sensitizer
andvariousNicomplexes.
(A)Strongly
absorbing
complexes
with charge transfer
bands, [Ir(III)[
R
ppy]
2
(
t
‑
Bu
bpy)]PF
6
(green line, R
= 2-(2,4-difluorophenyl)-5-trifluoromethyl),
and Ni(II)(
t
‑
Bu
bpy)(
o
-
tolyl)Cl
(
S
= 0, blue line), are highlighted.
(B) An expanded
view of
complexes
with only ligand field transitions
in the visible region,
Ni(II)(TMEDA)(
o
-tolyl)Cl
(
S
= 0, orange line), Ni(II)(TMEDA)Cl
2
(
S
= 1, red line), and Ni(II)(
t
‑
Bu
bpy)Cl
2
(
S
= 1, black line). Solvent
=
THF. Spectra
were digitized
and scaled
with permission
from
references
36, (Copyright
2018 American
Chemical
Society)
38
(Copyright
2022 American
Chemical
Society),
65 (available
under a
CC-BY
NC 3.0 Deed license,
copyright
2024 Bryden
and Zysman-
Colman),
66 (Copyright
2016 John Wiley and Sons) and 67
(Copyright
2020 American
Chemical
Society).
ACS
Catalysis
pubs.acs.org/acscatalysis
Review
https://doi.org/10.1021/acscatal.4c02036
ACS
Catal.
2024,
14, 9055
−
9076
9058
This is a critical
point for any photoredox
cycle featuring
multiple
intermediates
that could absorb photons
with energies
matching
those of the irradiation
source. For example,
Ni(II)
−
bpy aryl halide complexes
(bottom
box, Figure 3) are now
known
to be photoactive
in C(sp
2
)
−
C(sp
3
) cross-coupling
upon direct excitation
via a Ni(II)
−
C(aryl)
to Ni(I) + C(aryl)
•
bond homolysis
step.
36
−
38,79
Even a small amount
of photo-
generated
Ni(I)through
thisalternative
stepmaybesufficient
to
catalyze
the reaction.
These examples
are discussed
in Section
3.1.9. Importantly,
both the Ir photosensitizer
and the Ni(II)
−
bpy aryl halide complexes
absorb light in the visible region with
molar extinction
coefficients
of 10
3
M
−
1
cm
−
1
(Figure
2). The
molar absorptivities
of the various
Ni intermediates
possible
in
the reaction
cycle are largely unknown.
3.1.2.2.
*
Ir(III)
Is Sufficiently
Oxidizing
to React
with
Alkyl
Substrates,
Doing
so
Preferentially.
Redox
interactions
between
*
Ir(III)
and substrate
can be probed
through
electrochemical
measurements
and the oxidation
state of the
Ir complex
tracked
by absorption
spectroscopy.
Interactions
between
*
Ir(III)
and species in solution
other than the organic
substate,
including
any Ni complexes
in the putative
cycle, are
possible
and should
be evaluated.
For example,
the alkyl
substrates
used in the above-mentioned
work have accessible
oxidation
potentials
of
∼
1 V versus SCE,
29,30,80
but these
neighbor
the oxidation
potential
of Ni(II)
−
bpy
aryl halide
(
∼
0.8
−
0.9
V versus SCE). As will be seen below,
related
interactions
between
*
Ir(III)
and Ni complexes
are invoked
in
the Oxidative
SET mechanism
(Section
3.1.3). Furthermore,
both SET and triplet energy transfer
(
3
EnT) are possible
from
*
Ir(III)
to Ni(II),
81
further complicating
analyses
(see Sections
3.1.3 and 3.1.7).
3.1.2.3.
Ni(0)
−
bpy
Undergoes
Oxidative
Addition,
While
Ni(I)
−
bpy
Halide
Does
Not.
BothNi(0)andNi(I)canundergo
oxidative
addition
with aryl halides.
However,
Ni(I)
−
Ni(III)
oxidative
addition
would divert the proposed
Reductive
SET
mechanism
from Ni(0)
−
Ni(II)
oxidative
addition.
The
reactivity
of Ni(0) and Ni(II) vs Ni(I) and Ni(III)
are distinct.
Furthermore,
thepresence
ofNi(I)andNi(III)canleadtofacile
comproportionation
to
S
= 0 Ni(II)
−
bpy
aryl halide and
S
= 1
Ni(II)
−
bpy
dihalide,
82
another
chemically
distinct
species that
Figure
3.
Proposed
Reductive
SET mechanism.
C(sp
2
)
−
C(sp
3
) coupling
is presented
as a representative
example.
LG = leaving group.
ACS
Catalysis
pubs.acs.org/acscatalysis
Review
https://doi.org/10.1021/acscatal.4c02036
ACS
Catal.
2024,
14, 9055
−
9076
9059
is not considered
in this mechanism
but is important
for others
(Section
3.1.9).
Additionally,
Oderinde,
Johannes,
and co-
workers
notedthereduction
potential
ofIr(II)isscarcely
ableto
reduce various
Ni(I) complexes,
finding
their potentials
to be
similar ([Ir
III
/Ir
II
] =
−
1.37 V vs SCE, [Ni
I
/Ni
0
] =
−
1.41 V vs
SCE), and that Ni(0) is ineffective
to turn over the cycle.
69
Further
disfavoring
Ni(0), Gutierrez,
Martin
and co-workers
found that Ni(II)
−
bpy
dihalide
complexes
engage
in rapid,
facile comproportionation
with Ni(0)
−
bpy
species in solution,
affording
Ni(I)
−
bpy
halide
species.
83
However,
Plasson,
Fensterbank,
Grimaud
and co-workers
argued
that Ni(0) is
indeed
a vital source
of Ni(II)
−
bpy
aryl halide,
84
and
Bahamonde
and co-workers
argued that oxidative
addition
to
Ni(0) outcompeted
the comproportionation
reaction,
support-
ing an Oxidative
SET mechanism,
though
3
EnT pathways
were
notdiscarded
85
(seeSection
3.1.7). Altogether,
therequirement
of Ni(0) for catalytic
cycle turnover
is still debated.
3.1.2.4.
Alkyl
Radicals
Are
Preferentially
Captured
by
Ni(II),
Not
Ni(0).
GiventhatNi(0)andNi(II)complexes
arepresent
in
the proposed
mechanism,
a comparison
between
the relative
rates of radical capture
by both of these species
would help
confirm
the Ni(II) to Ni(III)
−
alkyl
hypothesis.
Computations
byMolander,
Kozlowski,
andco-workers
suggest
bothoxidation
states should
be productive
toward
radical
capture.
86
While
kinetic
analysis
for radical
capture
at Ni(II)
was recently
reported
(
k
=10
6
−
10
7
M
−
1
s
−
1
),
87
weareunaware
ofstudiesfor
C(alkyl)
•
capture
by Ni(0).
3.1.2.5.
Ir(II)
Is Sufficiently
Reducing
to
Regenerate
Ni(0)
and
Ir(III).
The presence
of Ir(II) presupposes
that Reductive
SET is indeed operative
(see point 2 above).
Given the highly
reducing
nature of Ir(II), one must also consider
its potential
interaction
with Ni(II) and Ni(III).
Reduction
of Ni(II) to
Ni(I) would
present
an alternative
mechanistic
route,
potentially
favoring
a Ni(I/III)
catalytic
cycle (see point 3).
Additionally,
Neurock,
Minteer,
Baran,andco-workers
reported
that pentacoordinate
Ni(III)
complexes
are readily reduced
to
Ni(II) via Ni
−
X heterolysis.
55
It is possible
the Ni(III)
species
couldbeintercepted
byIr(II)priortoreductive
elimination
and
thereby
be diverted
from the cycle making
C(sp
2
)
−
C(sp
3
)
bonds. Again, relative
reactivity
rates between
Ir(II) and the
relevant
Ni species
would prove invaluable
for mechanistic
insight.
There have been limited
experimental
mechanistic
studies
conducted
onthisreaction,
butonenotable
example
isthework
by Lloyd-Jones
and co-workers
in 2022.
88
Careful
kinetic
analysis
using radiolabeled
substrates
and
13
C NMR identified
the Ni(II)
−
bpy
aryl halide as a genuine
intermediate.
From the
kineticmodeling,
threeplausible
mechanisms
wereproposed
for
the reaction,
including
one which is akin to the Reductive
SET
mechanism
illustrated
above.
Interestingly,
this mechanistic
possibility
was the only one of the three the researchers
were
able to rule out. The remaining
two mechanisms
proposed
by
Lloyd-Jones
and co-workers
centered
around
*
Ir(III)
promot-
ing a photoinduced
Ni
−
halide
bond homolysis
step, referred
to
here as “Photosensitization
for Homolysis”
(see Section
3.1.5).
However,
the three mechanisms
considered
therein
are not an
exhaustive
list, as noted by the authors.
88
Nonetheless,
based on
these considerations
and the recent
kinetics
study,
the initially
proposed
Reductive
SET mechanism
is unlikely
operative.
Addi-
tional detailed
experimental
studies
are necessary,
however,
particularly
addressing
the five points outlined
above.
3.1.3.
Oxidative
SET.
The expansion
of dual Ni/Ir metal-
laphotoredox
reactions
to C(sp
2
)
−
X coupling
led to an
additional
mechanistic
hypothesis,
Oxidative
SET, as proposed
for C(sp
2
)
−
N coupling
by Jamison
and co-workers
in 2015
89
and C(sp
2
)
−
O/N
coupling
by MacMillan
and Buchwald
and
co-workers
28,90
in2015and2016,respectively.
IntheReductive
SET mechanism
for C
−
C bond coupling,
Ir(II) interacted
with
a Ni(I)
−
bpy
halide complex,
reducing
it by one electron
in a
dark reaction.
Keeping
with the naming
convention
adopted
herein, the Oxidative
SET mechanism
features
a SET wherein
*
Ir(III)
oxidizes
aNi(II)
−
bpy
arylalkoxide
complex
(rightbox,
Figure4), leadingtoaNi(III)speciesandIr(II).AsinReductive
SET, the Ir(III) complex
acts as the sole excited-state
active
species
in Oxidative
SET. Ir(II) reduces
a Ni(I)
−
bpy
halide
species
to generate
Ni(0)
−
bpy,
which undergoes
oxidative
addition
ofanarylhalidecoupling
partner
toformaNi(II)
−
bpy
aryl halide species
(bottom
box, Figure 4). Ligand substitution
of the alcohol
(or amine)
via the assistance
of exogenous
base
generates
the aforementioned
four-coordinate,
square-planar
Ni(II)
−
bpy
aryl alkoxide
(right box, Figure 4). The critical
chemical
impetus
behind
this mechanism
is the Ni(III)-
promoted
reductive
elimination
of the C
−
X product,
akin to
the one-electron
oxidation
chemistry
developed
by Hillhouse
andco-workers.
91,92
Initialreportsfounded
thisreaction
scheme
on the basis of redox potentials
and reductive
elimination
thermodynamics
for Ni(II) vs Ni(III).
3.1.4.
Key
Considerations
for
the
Oxidative
SET
Mecha-
nism.
3.1.4.1.
Ni(II)
−
bpy
Aryl
Alkoxide
Is the
SET
Partner
with
*
Ir(III).
While oxidation
of the Ni(II)
−
bpy
aryl alkoxide
species
to formal
Ni(III)
may be necessary
to drive reductive
elimination,
there are additional
Ni species
present,
including
the Ni(II)
−
bpy
aryl halide complex.
It is currently
unclear
why
*
Ir(III)
would preferentially
oxidize
one and not the other.
Additionally,
if Ir(II) is competent
for the reduction
of Ni(I) to
Ni(0), why either of these Ni(II) species
is not also reduced
presents
an open question.
As demonstrated
by Diao and co-
Figure
4.
Proposed
Oxidative
SET mechanism.
C(sp
2
)
−
O coupling
(alcohols)
is shown as a representative
example.
ACS
Catalysis
pubs.acs.org/acscatalysis
Review
https://doi.org/10.1021/acscatal.4c02036
ACS
Catal.
2024,
14, 9055
−
9076
9060