Organometallics
Article
saturated THF and toluene solutions, respectively. The solid-
state structures, depicted in Figure 3, indicate they are square
planar and exhibit characteristic Ni−C bonds (1.929(3) Å for
7·PPh3 and 1.935(3) Å for 7·PiBu3), similar to known
complexes.62,63,95−97 The different steric profiles of phosphine
ligands does not have a significant influence on the bond
angles, with a N(2)−Ni(1)−C(1) angle of 168.5(1)° for 7·
PPh3 and 166.9(1)° for 7·PiBu3].
Complexes 7·PR3 are all diamagnetic and feature character-
istic coupling of NiCH2 to phosphorus in the 1H NMR spectra
of complexes 7·PPh3 (3JHP = 10.5 Hz) and 7·PiBu3 (3JHP = 6.5
Hz), consistent with coordination of the PR3 in solution. The
31P{1H} NMR also features sharp singlets for 7·PPh3 at δ 42.2
and 7·PiBu3 at δ 5.2; these are observed downfield from the
free PR3 by approximately 50 ppm. The addition of free PPh3
to 7·PPh3 converts the 1H NMR CH2 doublet to a singlet, and
the phosphorus peak broadens in the 31P{1H} NMR,
consistent with rapid phosphine ligand exchange. In contrast,
the addition of excess PiBu3 to 7·PiBu3 did not cause any
evidence of fluxional exchange, even upon heating to 373 K in
toluene-d8.
the formation of diarylated products. In the initial stage (<30
min) of the reaction the rate can be conveniently determined
from the formation of 1-R. The reaction was found to be first-
order with respect to both 7·PiBu3 and iodobenzene but
inverse first-order for PiBu3. The inhibition by PiBu3 suggests a
mechanism where a pre-equilibrium phosphine dissociation
occurs prior to reaction with aryl iodide. Iodobenzene was
used in excess to give pseudo-first-order kinetics, and PiBu3 was
added to ensure a constant concentration for all reactions
studied.
The electronic effect of para-substituents on the reaction
rate was examined using iodoarenes bearing functional groups
at the para-position. A Hammett plot was constructed, and a
linear correlation with a ρ of −0.74 was observed for
iodoarenes bearing p-NMe2, p-OMe, p-H, and p-CF3
substituents, as shown in Scheme 9b. In contrast to the
positive ρ values often observed for oxidative addition to Ni(0)
or Pd(0),98,99 the negative ρ could be appropriate for oxidative
addition to a Ni(II) center. Studies of the oxidative addition of
aryl iodides to Pd(II) show ρ values of −0.77100 and −0.85,101
respectively, as electron-donating substituents are expected to
stabilize Pd(IV) centers. Thus it is reasonable to think the
more electrophilic Ni(II) will react faster with electron-rich
arenes; however, Hammett studies of reactions involving
concerted oxidative addition to Ni(II) are currently lacking for
comparison. The small negative value of ρ rules out a rate-
determining electron transfer, because ρ values for this
mechanism would be expected to be near 2.0.98 Addition of
a radical scavenger such as TEMPO gave no decrease in the
rate of formation of 1b, which also suggests a radical pathway
is unlikely. These experimental results support the results of
computational studies by Liu61 and Sunoj.60
C−H Functionalization. After C−H activation, the next
step in the catalytic cycle is reaction of 7· PR3 with aryl halides.
In an attempt to observe intermediates, the pseudocatalytic
condition reaction between 7·PiBu3 and iodobenzene in the
presence of 1a in benzene-d6 at 80 °C was monitored by NMR
spectroscopy, as shown in Scheme 8a. No Ni(IV) complex
from oxidative addition was observed; if involved, the Ni(IV)
complex is likely a high energy intermediate. In addition to the
formation of the organic arylated product 1b, the distinctive
paramagnetic peaks of complex 3 were observable early in the
reaction. As increased conversion of 1a to 1b occurred, new
peaks were observed for the complexes 8 and 9, as shown in
Scheme 8. Complexes 8 and 9 are analogues of 3 with arylated
The p-CO2Et substituted arene is an obvious outlier in
Scheme 9b, with a much faster rate of reaction than anticipated
from the Hammett ρ value. To further understand this
exception and whether a concerted mechanism is still
operational with this substituent, DFT studies were carried
out using Gaussian 16 to analyze both phosphine dissociation
and oxidative addition steps with respect to 7·PiBu3 that
generates Ni(IV) intermediate INT-2 through transition state
TS-1 (Scheme 10a). The reaction with 4-iodo-N,N-dimethy-
laniline (p-NMe2) is calculated to have the lowest activation
1
ligands, and feature many H NMR chemical environments
proximal to those for 3. Complexes 3, 8, and 9 likely arise from
ligand redistribution from putative intermediate 4b shown
central in Scheme 8b, similar to the chemistry of the
unobserved complex 4 shown in Scheme 2. In the 31P{1H}
NMR, the broad singlet at δ 5.0 for NiI2(PiBu3)2 was also
observed. Complexes 3, 8, and 9 are in equilibrium under the
reaction conditions, with the relative proportion of each
determined by the ratio of 1a and 1b in solution.
energy barrier with a ΔG⧧ that is 0.6 kcal/mol lower than
OA
To confirm the identity of species 8 and 9, initially assigned
by 1H NMR and analysis of equilibria, complex 8 was
synthesized from NiCl2(PiBu3)2 and ligand 1b, as shown in
Scheme 8c. The solid-state X-ray structure of 8 shows it is
structurally similar to 3, with a disphenoidal geometry at Ni.
iodobenzene at 298 K. Calculation also revealed the reaction
with ethyl 4-iodobenzoate also has a lower ΔG⧧
than
OA
iodobenzene (ΔΔG⧧ = −0.1 kcal mol−1) and is slightly
OA
higher than 4-iodoanisole (ΔΔG⧧ = 0.1 kcal mol−1), which
OA
1
is in agreement with our experimental result and a higher rate
of reaction is expected. A plot of calculated ΔG⧧OA against the
natural log of the observed rate of each reaction was also
constructed and a linear correlation was observed, suggesting
the same mechanism was operative throughout the reactions
with a series of substituted iodoarenes (Scheme 10b).
Catalyst Resting State and the Deleterious Effect of
Na2CO3. Intuitively, one might anticipate either the C−H
activation step or the oxidative addition to give Ni(IV) would
be rate-limiting in the Ni-catalyzed C−H functionalization of
1a; however, we have shown both these steps proceed at
temperatures as low as 80 °C under stoichiometric conditions.
In contrast, under catalytic conditions, much higher temper-
atures of 140−160 °C are required. We examined solutions
during catalysis by 1H NMR to try to observe the resting state
The H NMR of 8 exhibits characteristic broad singlets for a
paramagnetic Ni(II) species. Although the asymmetric
complex 9 could not be isolated in a pure form, it could be
prepared in equilibrium by the addition of 1a to 8, which
demonstrates the validity of the equilibrium shown in Scheme
8b. Alternatively, a solution of 9 could be prepared by the
equilibrium reaction of equimolar amounts of 3 and 8 via
ligand redistribution. At 298 K this reaction proceeds over the
course of hours.
To support the proposed Ni(II) to Ni(IV) oxidative
addition mechanism for the functionalization step, the
influence of aryl iodide on reaction rate was studied, as
shown in Scheme 9a. Initial rate studies were carried out for
these reactions of 7·PiBu3 with aryl iodides to generate 1-R.
The reaction was done in the presence of excess 1a to avoid
H
Organometallics XXXX, XXX, XXX−XXX
Liu, Junyang
