Organometallics
Article
by inversion symmetry with Ni···Ni separations of 2.949 and
2.998 Å, respectively (Figures 7a,b). The Ni−Namido distances
in 6 are 1.918(3) and 1.922(3) Å, whereas 7 shows Ni−Namido
bond lengths of 1.9462(13) and 1.9551(13) Å. In both
complexes 6 and 7, the Ni centers adopt a twisted square
planar geometry with angles of 20.02 and 22.14° between the
Nβ‑dik−Ni−Nβ‑dik and opposing N3−Ni−N3′ planes, respec-
1
tively. In solution, 7 is diamagnetic and possesses a H NMR
spectrum in C6D6 at room temperature (Figure S15) that
features a β-diketiminato C−H backbone resonance at δ 4.51
ppm with two different signals for the β-diketiminato N-aryl
meta-Ar−H sites at δ 6.880 and 6.464 ppm. Thus, 7 maintains
its dimeric, square-planar structure in solution at room
temperature with one anilido NArMe2 group above and below
the N3−Ni−N3′ plane.
A particularly novel complex, the parent amido species
{[Me3NN]Ni}2(μ-NH2)2 (8), was isolated in 91% yield as
orange crystals by slow diffusion of NH3 into a solution of
{[Me3NN]Ni}2(μ-OtBu)2 (2) (Figure 8c). As compared to
dinuclear amido complexes 6 and 7, parent amide 8 possesses
the shortest Ni−Namide bond distances of 1.9001(11) and
1.9017(11) Å that result in a shortened Ni···Ni separation of
2.919 Å (Figure 7c). The twist angle between the Nβ‑dik−Ni−
Nβ‑dik and Namide−Ni−Namide planes of 18.42° deviates least
from idealized square planar geometry, likely a result of the
minimal steric demands of the amido ligands. Unfortunately, 8
is not soluble in organic solvents even at elevated temper-
atures, which hampers its solution characterization. FT-IR
spectrum of 8 shows two peaks at 3430 and 3413 cm−1 as a
result of two N−H stretching modes of the amido ligand.
Synthesis and Characterization of Dinuclear and
Mononuclear Nickel(II) Phenolates 9 and 10 and
Mononuclear Anilide 11. Nickel(II) complexes [NiII]−FG
that bear weaker donors give rise to dinuclear or mononuclear
species that are paramagnetic. For instance, addition of 2 equiv
of phenol to {[Me3NN]Ni}2(μ-OBut)2 (2) in ether at RT gives
the dinuclear phenoxide complex {[Me3NN]Ni}2(μ-OPh)2
(9) as dark green crystals (Figure 8a). The X-ray structure
of 9 reveals a distorted tetrahedral environment around the
NiII centers with Ni−O distances of 1.966(3) and 1.967(3) Å
along with a Ni···Ni separation of 3.115 Å, which is the longest
in the series of dinuclear complexes 6, 7, and 8 (Figures 8a−c).
This complex is paramagnetic in solution with a μeff of 2.7(2)
B.M. in benzene-d6 by the Evans method.16
Figure 9. PhCH(•)Me radical capture by [NiII]−FG complexes.
from the aliphatic azo compound (E/Z)-azobis(α-phenyl-
ethane) at 100 °C (Figure 9).19 Reaction of complexes 3−7
and 9 with (E/Z)-azobis(α-phenylethane) at 100 °C in
fluorobenzene solvent gave 25−67% yields of the correspond-
ing radical capture products PhCH(FG)Me. Thus, radical
capture at [NiII]−FG complexes results in the formation of C−
C (59% for 3 and 61% for 4), C−N (67% for 5, 55% for 6, and
25% for 7), and C−O (48% for 9) bonds. In the case of
complex 7, concomitant formation of diazene ArNNAr (Ar
= 3,5-Me2C6H3) was observed in 49% GC yield along with
25% yield of the amination product PhCH(NHAr)Me. For the
parent amide complex {[Me3NN]Ni}2(μ-NH2) (8) only a
trace amount of the C−H amination product PhCH(NH2)Me
was detected by GC-MS analysis, attributed to its poor
solubility, even at elevated temperatures.
In contrast, mononuclear complexes [Me3NN]Ni−FG with
bulky, electron-poor functional groups (FG = OArCl3 (10) or
NHArCl3 (11)) did not result in radical capture. Rather, only
erythro and threo isomers of the ethylbenzene radical dimer
PhCH(Me)−CH(Me)Ph are observed. In fact, this product of
PhCH(•)Me radical dimerization occurs in comparable
amounts across all reactions. Control experiments show no
radical capture at room temperature after overnight stirring of
the complexes with (E/Z)-azobis(α-phenylethane) in PhF.
Computational Analysis of Radical Capture Pathways
at Nickel(II) Nitromethanoate 3 and Benzamide 5. Most
mechanistic proposals for Ni-catalyzed Csp2-Csp3 bond
formation involve radical capture at NiII to form a NiIII
organometallic complex, followed by reductive elimination
(Figures 1d and 10, route a).4,20 A less prominent mechanism
involves concerted bond formation between the PhCH(•)Me
radical at the functional group (Figure 10, route b).21 When
radical capture at the functional group involves bond formation
with an atom that is bound to the metal center, both radical
capture at the metal center and functional group could be
competitive. To illustrate this competition between radical
capture pathways, we computationally considered capture of
Addition of the bulky, electron-poor 2,4,6-trichlorophenol or
2,4,6-trichloroaniline to 2 provided deep blue [Me3NN]Ni−
OPh2,4,6‑Cl (10) or teal [Me3NN]Ni−NHPh2,4,6‑Cl (11)
isolated in 49 and 40% yields, respectively (Figures 8b,c).
These structures feature much shorter Ni−N and Ni−O
distances to the anilido or phenoxy donors, 1.8765(19) and
1.8812(12) Å, respectively, than observed in the dinuclear
structures of 7 and 9. Interestingly, these structures are best
described as pseudotetrahedral due to coordination of one of
the ortho-Cl atoms with Ni−Cl distances of 2.4811(6) and
2.4595(5) Å in anilido 10 and phenoxy 11, respectively.
Accordingly, each is paramagnetic in solution with magnetic
susceptibilities (benzene-d6) of 2.8(1) and 2.6(2) B.M.,16
respectively. This is consistent with S = 1, high-spin d8
electronic structures.
3
3
Capture of the PhCH(•)Me Radical by [NiII]−FG
Complexes. We explored the radical capture ability of each
[Me3NN]Ni−FG complex 3−11 through the in situ
generation of the secondary benzylic radical PhCH(•)Me
E
Organometallics XXXX, XXX, XXX−XXX