Nickelation of PCP- and POCOP-type pincer ligands: Kinetics and mechanism

Article abstract of DOI:10.1021/om3003784

This report describes the results of a combined experimental and computational investigation on the kinetics and mechanism of the C-H metalation step involved in the formation of PCP- and POCOP-type complexes of nickel. The kinetics of the C-H nickelation reaction was probed through competition studies involving two ligands reacting with a substoicheometric quantity of {(i-PrCN)NiBr₂}_n. These experiments have confirmed that metalation is more facile for aromatic ligands 1,3-(i-Pr₂PE) ₂C₆H₄ vs their aliphatic counterparts 1,3-(i-Pr₂PECH₂)₂CH₂ (sp² C-H > sp³ C-H; E = O, CH₂), ligands bearing phosphine moieties vs those with phosphinite moieties (PCP > POCOP), ligands bearing P substituents i-Pr₂P vs t-Bu₂P and Ph₂P, and POC sp²OP ligands 1,3-(i-Pr₂PO)₂C ₆R_nH_4-n bearing electron-donating vs electron-withdrawing substituents (p-OMe < m-OMe > p-Me > m-CO ₂Me > p-CO₂Me > m,m-Cl₂). Among the latter, there is a 6-fold difference in C-H metalation rate between ligands bearing p-OMe and p-COOMe, whereas the most readily metalating ligand, 1,3-(i-Pr₂PCH₂)₂C₆H₄, is metalated ca. 270 times more readily relative to the least reactive ligand, 1,3-(i-Pr₂POCH₂)₂CH₂. Density functional calculations indicate that PCP- or POCOP-type pincer ligands react with NiBr₂ to generate nonmetalated intermediates that form the corresponding pincer complexes via a two-step mechanism involving an ionic dissociation of the bromide to give a tight ion pair intermediate, followed by bromide-assisted deprotonation of the C-H bond. The type of structure adopted by the nonmetalated intermediates (mono- or dinuclear; tetrahedral, cis or trans square planar) and the energy barriers for the metalation transition states depend on the steric properties of the PR₂ moiety. The presence of a base that can neutralize the HBr generated in the metalation step is crucial for rendering the metalation process exergonic. One rationale for the more facile metalation of PCP ligands in comparison to their POCOP counterparts is the greater donor character of phosphine moieties, which allows a more effective stabilization of the coordination and metalation transition states wherein the strongly donor halide ligand is displaced by a much weaker C-H bond donor. The aromatic ligands metalate more readily than their aliphatic analogues for multiple reasons, including the higher ground state energy of the nonmetalated intermediates formed with aromatic ligands, the stronger C sp²-Ni bond formed via metalation, and the more stabilized anionic charge on the C atom being metalated.

Full text of DOI:10.1021/om3003784

Article
pubs.acs.org/Organometallics
Nickelation of PCP- and POCOP-Type Pincer Ligands: Kinetics and
Mechanism†
Boris Vabre,‡ Melinda L. Lambert,§ Alban Petit,§ Daniel H. Ess,*,§ and Davit Zargarian*,‡
‡Dep
́ ́ ́ ́ ́
artement de chimie, Universite de Montreal, Montreal (Quebec), Canada H3C 3J7
§Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84602, United States
S
* Supporting Information
ABSTRACT: This report describes the results of a combined experimental and computational investigation on the kinetics and
mechanism of the C−H metalation step involved in the formation of PCP- and POCOP-type complexes of nickel. The kinetics
of the C−H nickelation reaction was probed through competition studies involving two ligands reacting with a substoicheometric
quantity of {(i-PrCN)NiBr2}n. These experiments have confirmed that metalation is more facile for aromatic ligands 1,3-(i-
Pr2PE)2C6H4 vs their aliphatic counterparts 1,3-(i-Pr2PECH2)2CH2 (sp2 C−H > sp3 C−H; E = O, CH2), ligands bearing
phosphine moieties vs those with phosphinite moieties (PCP > POCOP), ligands bearing P substituents i-Pr2P vs t-Bu2P and
2
Ph2P, and POCsp OP ligands 1,3-(i-Pr2PO)2C6RnH4−n bearing electron-donating vs electron-withdrawing substituents (p-OMe ≈
m-OMe > p-Me > m-CO2Me > p-CO2Me > m,m-Cl2). Among the latter, there is a 6-fold difference in C−H metalation rate
between ligands bearing p-OMe and p-COOMe, whereas the most readily metalating ligand, 1,3-(i-Pr2PCH2)2C6H4, is metalated
ca. 270 times more readily relative to the least reactive ligand, 1,3-(i-Pr2POCH2)2CH2. Density functional calculations indicate
that PCP- or POCOP-type pincer ligands react with NiBr2 to generate nonmetalated intermediates that form the corresponding
pincer complexes via a two-step mechanism involving an ionic dissociation of the bromide to give a tight ion pair intermediate,
followed by bromide-assisted deprotonation of the C−H bond. The type of structure adopted by the nonmetalated intermediates
(mono- or dinuclear; tetrahedral, cis or trans square planar) and the energy barriers for the metalation transition states depend
on the steric properties of the PR2 moiety. The presence of a base that can neutralize the HBr generated in the metalation step is
crucial for rendering the metalation process exergonic. One rationale for the more facile metalation of PCP ligands in comparison
to their POCOP counterparts is the greater donor character of phosphine moieties, which allows a more effective stabilization of
the coordination and metalation transition states wherein the strongly donor halide ligand is displaced by a much weaker C−H
bond donor. The aromatic ligands metalate more readily than their aliphatic analogues for multiple reasons, including the higher
2
ground state energy of the nonmetalated intermediates formed with aromatic ligands, the stronger Csp −Ni bond formed via
metalation, and the more stabilized anionic charge on the C atom being metalated.
Bu2PCH2)2-C6H4 in EtOH at 80 °C for 5 min gave the
corresponding PCP-type complex A, shown in Chart 1.5
Shaw’s early reports have inspired many other groups to use
a C−H nickelation strategy for preparing a variety of nickel
complexes featuring pincer ligands based on both aromatic and
aliphatic skeletons (Chart 1);6 there is also increasing reliance
on direct C−H metalation reactions for preparation of pincer
complexes of other 3d metals.7,8 While formation of a pincer
complex is an overall intermolecular metalation process, the
C−H metalation step involved in this process is considered to
be intramolecular in nature, because it is thought to occur in
the intermediate generated following the initial binding of the
INTRODUCTION
■
Historically, the bulk of the experimental efforts dedicated to
promoting transition-metal-mediated C−H activation reactions
has involved complexes of 4d and 5d transition metals.1,2 Their
strong M−H and M−C bonds have contributed to our
perception of these metals as being much more adept at
promoting C−H activation reactions relative to their lighter
counterparts. Nevertheless, many reports of C−H activation/
metalation by 3d metals have appeared over the past decades
and lately are appearing more frequently.3 Indeed, Kleiman and
Dubeck reported an early instance of chelate-assisted C−H
nickelation nearly 50 years ago,4 while pincer-type complexes of
nickel were among the first such complexes reported. Thus,
Shaw showed that heating [Ni(H2O)6]Cl2 and 1,3-(t-
Received: May 5, 2012
Published: August 23, 2012
© 2012 American Chemical Society
6041
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
experimental and computational study to probe the rates and
mechanisms of these reactions. The experimental work has
sought to quantify the relative nickelation rates for PCP- and
POCOP-type pincer ligands as a function of donor moiety (i.e.,
P substituents), hybridization of the central carbon atom (sp2 vs
sp3), and ring substituents R′ in 1,3-(R2PO)2C6HnR′4−n. The
computational studies, on the other hand, have focused on
identifying the reaction pathways and structures of inter-
mediates and/or transition states, with the goal of identifying
the factors governing C−H nickelation reactions and shedding
light on their mechanisms.20 The overall objective is to
determine whether the C−H nickelation reactions involved in
the synthesis of PCP- and POCOP-type pincer nickel
complexes are best viewed as proceeding by oxidative
addition/reductive elimination steps, being electrophilic or
nucleophilic in nature, or otherwise belonging to other reaction
schemes under consideration by the community of research-
ers.21
Chart 1
ligand moieties (double chelate-assisted activation). In some
instances, intermediates en route to a pincer complex featuring
a C−H moiety “poised” for metalation have been detected or
isolated.9
A broad survey of the literature on pincer complexes of
nickel reveals that they can be prepared via C−H nickelation in
the case of bis(phosphine) (PCP),6,10 bis(phosphinite)
(POCOP),11,12 and amino(phosphinite) (POCN)13 ligands,
whereas virtually all reported cases of complexes featuring
bis(amino) moieties (NCN) are prepared via alternative
synthetic routes,14,15 C−H nickelation being generally inacces-
sible for this family of ligands.16 Indeed, the success of the C−
H nickelation step depends not only on the nature of ligand
donor moieties but also on the hybridization of the C atom
being nickelated (sp2 vs sp3), the type of nickel precursor used,
and the reaction conditions (temperature, solvent, etc.). The
importance of the “linker” moiety is evident from the
observation that aromatic ligands of the type 1,3-Q2-C6H4
(Q2 = (R2PCH2)2, (R2PO)2, R2PO/R2NCH2) nickelate much
more readily in most cases without detectable intermediates,17
whereas the analogous ligands based on aliphatic backbones
generate mixtures of nickelated and non-nickelated products,
some of which are illustrated in Scheme 1.10a,b,18,19
RESULTS AND DISCUSSION
■
Experimental Studies. The ligands required for the kinetic
studies are shown in Chart 2. Ligands 1−4 were selected to
quantify how nickelation rates are influenced by the type of
donor moiety (PCP vs POCOP) and linker skeleton (aromatic
vs aliphatic), while 1a−j were selected to evaluate the impact of
ring substituents on the nickelation of this family of ligands.
The experimental work began by synthesizing the PCP and
POCOP using modified versions of previously reported
procedures.22 In general, the POCOP ligands can be prepared
by reacting the requisite diol preligands with ClPR2 in the
presence of base.23 This methodology is much more
straightforward and often gives better yields in comparison to
the synthesis of PCP ligands, which are obtained from the
reaction of suitable dihalides with [R2P]− or R2PH/base.24,25
Previous reports have indicated that the choice of Ni
precursor and presence of a suitable base are important factors
for success in the synthesis of PCP- and POCOP-type pincer
complexes.10,11b,c,13a,18 In our experience, the best precursors
In an effort to understand the different aptitudes of pincer
ligands for C−H nickelation, we have undertaken a combined
26
for C−H nickelation of pincer ligands are (MeCN)nNiBr2
Scheme 1
6042
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
Chart 2
and {(i-PrCN)NiBr2}n,27 the latter being more suitable for
conducting kinetic studies owing to its well-defined molar
mass.28 The importance of base for the nickelation of PCP
ligands has been demonstrated by the low yields of (PCP)NiBr
(<50%) and recovery of the protonated (unmetalated) ligand
as a precipitate.10a Similarly, the complexes (POCN)NiBr can
be prepared in 80−90% yields when NEt3 is added to the
reaction mixture, whereas much lower yields are obtained
(∼35%) in the absence of added base.13a,b On the other hand,
Morales-Morales’s group has reported that refluxing PO-
CHOPPh and NiCl2 in toluene in the absence of any added
base gives (POCOPPh)NiCl with 80% yield.11a
NMR chemical shifts of various ligands and complexes were
free of overlap for the most part, making this the most suitable
method for kinetic studies. Preliminary measurements showed
that, under the pseudo-first-order conditions, nickelation of the
aromatic ligands 1 and 2 is too rapid for monitoring by NMR:
most reactions were essentially finished in the time of mixing.
Employing low reaction temperatures or very dilute samples
slowed down these reactions somewhat (2−5 min), but not
sufficiently to allow convenient rate measurements.29 As a
result, we have opted to measure relative rates of nickelation via
competition experiments, as described below.
Relative Nickelation Rates. The competition experiments
To further probe the importance of base for nickelation of
POCOP ligands, we used 31P NMR to monitor the reaction of
ligand 1a with {(i-PrCN)NiBr2}n in the absence of an added
base. The signal for 1a (148.1 ppm) disappeared over time, and
we observed a singlet resonance at 188.6 ppm due to the pincer
complex (POCOP)NiBr (1a′), in addition to a new peak at
134.7 ppm. Addition of NEt3 to the reaction mixture led to the
formation of a white precipitate and caused the suppression of
the latter peak and the reappearance of the original signal for
ligand 1a. An independent test established that reaction of 1a
with excess HX generates the diprotic species shown in Scheme
2 (no hydrolysis back to resorcinol); we conclude, therefore,
were conducted as follows. Mixtures of 1.0−1.1 equiv of each of
the two ligands under study were prepared and analyzed by 31
P
NMR to register the initial molar ratio. THF was used as
solvent for mixtures of aromatic ligands 1 and 2, whereas
toluene was used in the case of aliphatic ligands 3 and 4. Each
mixture was then added to a homogeneous solution containing
{(i-PrCN)NiBr2}n and NEt3 with stirring at room temperature;
the initial Ni:L:L′:NEt3 molar ratio was approximately
0.75:1.0:1.0:2.0 (eq 1). Stirring the resulting mixture at room
temperature for a few minutes was sufficient to nickelate the
aromatic ligands 1 and 2, as signaled by the nearly
instantaneous color change from blue to brown and then
yellow, followed by the gradual precipitation of a white solid
(HNEt3Br). It is worth noting that these mixtures remain
homogeneous throughout the reaction, except for the
precipitation of the ammonium salt at the end.
Scheme 2
Nickelation of the aliphatic ligands 3 and 4 followed a
different path: stirring the mixture at room temperature caused
a color change from blue to deep purple, which changed to
yellow only after the reaction mixture was heated to ca. 70−90
°C for 2−3 days. The 31P NMR spectra of these mixtures have
established that the yellow color corresponds to the target
pincer complexes, whereas the sparingly soluble deep purple
species is an NMR-silent nonmetalated species described
previously.10a,b,18
that the HX generated during the formation of our pincer
complexes must be quenched to prevent protonation of
unreacted ligands. Thus, the competition nickelation reactions
were conducted in the presence of NEt3 to ensure that these
reactions are driven to completion.
Kinetic Studies. Initial tests showed that UV−vis spectros-
copy could not be used for monitoring the metalation reactions,
due to significant overlap in the absorption energies of some
ligands and their respective complexes; in contrast, the 31P{1H}
Integration of the 31P signals due to the pincer complexes
present in the final reaction mixtures allowed us to establish the
relative nickelation aptitudes of the ligands studied. That the
nickelation rates can be considered invariable under the
conditions of these kinetic studies was confirmed by the
6043
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
a
Table 1. Relative Nickelation Rates for Various POCOP and PCP Ligands
a
For details of the procedure used for conducting these competition studies, see the main text or the Experimental Section. Footnotes in the body of
the table are as follows. (†) The relative rates of nickelation for these ligands could not be determined directly, because the 31P NMR signals of their
complexes overlapped. We integrated, instead, the signals for free ligands remaining after the nickelation and used their ratio for deducing the relative
rates of nickelation. (‡) In this case, addition of {(i-PrCN)NiBr2}n caused the reaction mixture to turn green initially and the nickelation was fairly
slow: the mixture was analyzed 1 h later after the anticipated white precipitate (HNEt3Br) and yellow color appeared.
Scheme 3
observation that no reaction occurs after stirring for 24 h
mixtures of complex 1e′ and the free ligand 1a (1 day at 55 °C)
or complex 3′ and ligand 4 (70 °C).30 Most experiments were
carried out in triplicate and showed fairly reproducible results,
the relative rates varying by ca. 10% on average. The results of
these competition experiments are presented in Table 1 and
discussed below.
observations that nickelation is much more facile for the
aromatic ligands. Monitoring these experiments by 31P NMR
also revealed important insights about how the nickelation
proceeds. For instance, holding the reaction mixture containing
1a/3/{(i-PrCN)NiBr2}n/NEt3 at room temperature over
several hours led to none of the anticipated nickelated product,
1a′: the 31P NMR spectra of this mixture showed only the
signals for 1a (major) and 3 (trace), but none for 1a′, which
was unexpected because nickelation of 1a to 1a′ normally
The first two entries of Table 1 show the results of
competition experiments that served to quantify our previous
6044
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
occurs in less than 2 min at room temperature in the absence of
3. On the basis of our previous observation that 3 reacts with
Br2Ni(NCMe)n to give a nonmetalated, NMR-silent species
that does not undergo metalation at ambient temperature,31 we
infer that at room temperature the nickel precursor is ligated by
the aliphatic ligand 3 and remains trapped in the form of a
nonmetalated (and NMR-silent) species. Heating the mixture
for several hours at 70 °C in the presence of 3 led to the
appearance of the pincer complex 1a′ and free ligand 3,
presumably because heating releases the nickel precursor and
allows it to react with 1a. These observations can be
rationalized by invoking a reaction scheme under Curtin−
Hammett control: the nonmetalated intermediates interconvert
rapidly via a binding equilibrium that favors ligation of Ni(II)
by 3, and the subsequent nickelation steps are slower and
irreversible (Scheme 3).
Chart 3. Relative Nickelation Rates for PCP and POCOP
Ligands
nickelation. Our efforts have focused on identifying possible
ground-state structures for the nonmetalated (LH)NiBr2
complexes and the corresponding pincer species LNiBr (L =
PCP- and POCOP-type pincer ligands), in addition to any
plausible intermediates and transition states involved in the
metalation pathway. All stationary point structures presented
were verified as minima or saddle points with full calculation of
the Hessian using Gaussian 0936 with the M0637 density
functional method using an ultrafine integration grid. The
choice of M06, which is a hybrid meta functional that was
designed for broad applicability including transition metals,38
was based on several recent computational benchmark and
experimental comparative studies showing that M06 is the most
accurate functional currently available for organometallic
systems.39 In addition, the M06 functional is known to
accurately reproduce Ni−P type bond energies, whereas
many other common density functionals fail.40 Nevertheless,
comparison is also made throughout the present study to
results obtained from computations based on B3LYP, BP86,
and MPWPW91 functionals.
Next, we probed the question of which donor moiety,
phosphine or phosphinite, would be more adept at facilitating
the nickelation of pincer ligands. The results shown in entries 3
and 4 establish that nickelation is more than 5 times faster for
the PCP ligands 2 and 4 vs their POCOP analogues 1 and 3.
These observations, combined with the fact that NCN-type
ligands 1,3-(CH2NR2)2-C6H4 do not undergo nickel-
ation,14−16,18 demonstrate the importance of the donor moiety
character (PCP > POCOP ≫ NCN). P substituents were also
found to have a major influence on the nickelation rates of
2
POCsp OP ligands (entries 5 and 6: i-Pr2P ≫ t-Bu2P, Ph2P);
previous studies have also shown that nickelation proceeds
more readily with the PCP ligand 4 in comparison to its t-Bu2P
analogue.10a,b
The impact of ring substituents on the nickelation rates was
2
studied next. Among the variously substituted POCsp OP
ligands, those bearing the electron-donating substituents Me
and OMe accelerated the nickelation rates relative to the
unsubstituted ligand 1a (entries 7−9), while ligands bearing
electron-withdrawing substituents showed significantly slower
nickelation rates (entries 10−12; see the Supporting
Information for a plot of the Hammett parameters). It appears,
therefore, that nickelation of these aromatic POCOP ligands is
akin to an electrophilic aromatic substitution,32 whereas the
opposite has been observed for direct palladation of aromatic
C−H bonds. For instance, the groups of Maseras and
Echavaren have demonstrated a Pd-catalyzed C−H activa-
tion/cyclization reaction that proceeds preferentially on
aromatic rings bearing electron-withdrawing groups,33 and
Fagnou’s group has reported extensively on the much more
facile palladation of electron-poor aromatic rings.34
All optimizations were carried out in implicit THF solvent
using the CPCM model combined with UAKS radii. The basis
set employed consisted of LANL2TZ(f) with the polarizing f
exponent set to 3.130 for Ni, LANL2DZdp for Br, and 6-
311G(2d,p) for all other atoms. Unless otherwise noted, all
reported energy values are free energies in THF at 298 K.
Singlet−triplet spin-crossing (SC) points, also called minimum
energy crossing points (MECP), were located using the
algorithm developed by Harvey et al. in conjunction with
Gaussian 09.41 The 3D ball and stick structures were rendered
using CYLview.42
Ground State Ligand Coordination in (PCsp3HP)NiBr2. To
begin, we have explored tetrahedral as well as cis and trans
square-planar coordination geometries for the mononuclear
In the case of MeO-substituted ligands, the substitution
pattern (meta vs para) seems to have little or no effect on
metalation (entry 13), which is very different than in the case of
MeO2C-substituted ligands: a p-COOMe group appears to be
more effective in hindering the nickelation than a m-COOMe
group (entry 4).35 Finally, comparison of the results shown in
entry 15 vs entries 3 and 4 demonstrates that judicious choice
of ring substituents can induce as significant an impact on the
nickelation rate as induced by the character of the donor
moiety. Chart 3 shows the relative rates of metalation for all the
nonmetalated species (PCsp HP)NiBr2 1 (Figure 1). As
3
expected, structures featuring square-planar geometries
1
1
(1trans-1, trans-1-twist, and cis-1) are singlet species, while
3
3
the tetrahedral species tetra-1 is a triplet. Interestingly, tetra-
1 was found to be the lowest energy structure at the M06 level
1
of theory, being 4.7, 4.4, and 3.7 kcal/mol lower than cis-1,
1trans-1, and 1trans-1-twist, respectively; this difference in
stability was found to be even greater (by ca. 10 kcal/mol)
when the B3LYP functional was used. The observed order of
stabilities, which is contrary to the usual preference for square-
planar structures in (PR3)2NiX2, arises mostly from the
minimization of CH2−CH2 eclipsing interactions in the highly
3
pincer ligands studied here, normalized to the POCsp OP ligand
3.
Density Functional Investigation. Nickelation of the
PCP and POCOP pincer ligands has been investigated by
quantum mechanical calculations to shed light on reaction
mechanisms and interpret the observed relative rates of
twisted configuration adopted in 3tetra-1 by the PCsp HP ligand
3
backbone. Among the square-planar structures, the eclipsing
1
1
interactions seem to be highest in cis-1, followed by trans-1
6045
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
nearly optimal square-planar geometry with P−Ni−P angles of
ca. 176°. In contrast, B3LYP predicted that conversion of
1
3tetra-1 to trans,trans-1 should be endergonic by 1.3 kcal/
mol. However, this endergonic estimate is likely the result of
B3LYP overestimating the high-spin state. When i-Pr2P groups
3
were modeled instead of Me2P, tetra-1 was found to be more
1
stable than trans,trans-1 even by M06 (ΔG = 2.1 kcal/mol),
indicating that the steric bulk plays a significant role in whether
dimeric or monomeric nickel species are favored. It is worth
noting that if 1trans,trans-1 is formed, either as a kinetic
product or in equilibrium, it must convert to a mononuclear
species prior to metalation. In other words, factors leading to
stabilization of dimeric structures will likely hinder the
metalation step.44
Mechanism of (PCsp HP)NiBr2 Metalation. The first
3
mechanism we explored for the metalation of (PCsp HP)NiBr2
3
involves oxidative addition of the central C−H bond of the
coordinated ligand onto the NiII metal center to give an
octahedral, tetravalent hydrido-pincer intermediate (Scheme
4a). Oxidative addition is typically a closed-shell process and is
unlikely to proceed from the triplet species 3tetra-1.
Accordingly, the lowest energy oxidative addition transition-
state structure located for the bromo precursor (X = Br) is the
Figure 1. Lowest energy square-planar and tetrahedral ground-state
complexes for the nonmetalated (PCsp HP)NiBr2. Free energies are
3
3
relative to tetra-1 (kcal/mol).
and 1trans-1-twist. Another important factor is the smaller than
1
1
1
ideal P−Ni−P angles in trans-1 (157°) and trans-1-twist
(148°) structures in comparison to the nearly ideal P−Ni−P
angle of ca. 106° in 3tetra-1. The closest C−H···Ni distance (at
2.367 Å in 1trans-1) does not induce any C−H bond
elongation, implying that none of these nonmetalated
structures involves η2 C−H agostic interactions.
closed-shell singlet trans-OXadd, which leads to the hydrido-
1
pincer intermediate trans-INT (Figure 3). The free energy
barrier for this process is 20.7 kcal/mol relative to 3tetra-1, and
the hydrido-pincer intermediate is endergonic by 19.7 kcal/mol
relative to 3tetra-1. It is also worth noting that 1trans-INT faces
a reductive elimination barrier of only 1.4 kcal/mol; such
shallow M−H intermediates have also been detected in other
oxidative addition reactions analyzed by DFT.45
The structures depicted in Figure 1 are for 1:1 PCsp HP:NiBr2
3
species. Exploring the energetics of 2:2 species has led us to the
structure trans,trans-(PCsp HP)2Ni2Br4 (1trans,trans-1) shown
3
The reaction pathway leading to 1trans-OXadd goes through
the singlet species 1trans-2 (Figure 4), which resembles 1trans-
1 except for the proximity of the nickel center to the methylene
group that eventually undergoes metalation. At 1.958 Å, the C−
H···Ni distance in 1trans-2 is typical of an agostic type
interaction, but the relatively nonperturbed C−H bond (1.108
Å) and the higher energy of 1trans-2 relative to 3tetra-1 (by 9.9
kcal/mol) indicate that the C−H bond does not participate in
an agostic interaction. This is a reasonable conclusion, because
there is no low-lying and geometrically accessible vacant orbital
in Figure 2.43 The conversion of 2 equiv of tetra-1 into
3
1trans,trans-1 is exergonic by 7.5 kcal/mol at the M06 level of
theory (eq 2), because this dinuclear structure can adopt a
1
on the nickel center. Unexpectedly, trans-2 was found to be a
spin-contaminated species (⟨S2⟩ = 0.17), and its reoptimization
3
as a triplet gave the nearly equally energetic structure trans-2
(10.0 kcal/mol, Figure 4). The possible pathways leading to
3
1trans-2 and trans-2 are shown in Scheme 5.
Figure 2. Possible nonmetalated trans-dinuclear species.
6046
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
Scheme 4
Modeling the transition structure with i-Pr2P moieties leads
to a greater predicted activation free energy for the two-step
ionic pathway (17.4 vs 21.2 kcal/mol), which is not surprising
given the congested transition state. It should be noted,
however, that oxidative addition remains the higher energy
pathway (29.4 vs 21.2 kcal/mol), even with i-Pr2P moieties.
1
The closed-shell structure trans-TS does not have a lower
energy unrestricted energy solution; a similar transition
structure was also identified on the triplet energy surface, but
this triplet structure (3trans-TS, Figure 5) is 24.9 kcal/mol
higher in free energy than the singlet transition structure.
Figure 3. Oxidative addition transition structure and nickel hydride
intermediate for sp3 PCP metalation. Free energies are expressed in
1
3
The imaginary vibrational frequency in trans-TS suggests
kcal/mol relative to tetra-1.
that this structure connects an ion pair intermediate to the
metalated product and liberated HBr. However, calculation of
the forward and reverse intrinsic reaction coordinates (IRC)
only showed product formation, while optimization of a slightly
1
1
perturbed trans-TS structure gave trans-2. Since the IRC
calculations were not definitive, we explored the homolytic and
heterolytic Ni−Br bond dissociation energies (BDE). The
unreasonably large BDE value for homolytic cleavage of the
1
Ni−Br bond (53.0 kcal/mol relative to trans-2) indicates that
all scenarios requiring such a step can be discounted.
Heterolytic cleavage of the Ni−Br bond (24.7 kcal/mol relative
1
Figure 4. Singlet and triplet intermediates prior to sp3 PCP ligand
metalation. Free energies are expressed in kcal/mol relative to 3tetra-1.
to trans-2) is significantly lower than homolytic cleavage but
higher than 1trans-TS, suggesting that the bromide anion likely
1
undergoes an incomplete dissociation prior to trans-TS to
form a tight ion pair intermediate between 1trans-2 and 1trans-
TS. To explore this possibility, we carried out Ni−Br bond
The next mechanism we explored was σ-bond metathesis
1
scans on trans-2 from 2.40 to 4.00 Å, which revealed that a
3
that would directly generate the metalated product (PCsp P)-
closed-shell minimum energy structure ion-pair (Scheme 5)
exists at a Ni−Br bond length of 3.90 Å.
NiBr and HBr (Scheme 4b). All attempts to locate a transition
state for the bromo precursor traversing this pathway
1
The tight ion-pair intermediate is generated from an
associative type transition structure where the incoming C−H
bond begins to form an agostic interaction with concomitant
displacement of the bromide ion via 1trans-TSassoc (Figure 5).
In this transition structure, the C−H bond is lengthened to
1.135 Å from 1.108 Å in 1trans-2, and the Ni−Br bond
increases by 0.3 Å to an elongated bond length of 2.674 Å. IRC
converged on the structure trans-TS (Figure 5), for which
1
the free energy was found to be lower than that of trans-
OXadd by 3.3 kcal/mol (M06) or 10.6 kcal/mol (B3LYP). The
most significant feature of 1trans-TS is the completely
dissociated character of the bromide ion during the
deprotonation of the C−H bond. This aspect differentiates
the C−H nickelation scheme operating in the present system
from analogous mechanistic schemes in which the base remains
coordinated to the metal center throughout the metalation
process (cf. concerted metalation deprotonation46 and
ambiphilic metal−ligand activation21b,47).
1
calculations for trans-TSassoc indeed confirm the connection
1
between trans-2 and ion-pair. The activation free energy of
1trans-TSassoc is 0.4 kcal/mol higher than the activation free
1
energy of trans-TS.
6047
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
a
3H
3
Scheme 5. Proposed Free Energy Surface for Conversion of (PCsp P)NiBr2 to the Pincer Species (PCsp P)NiBr
a
Grey lines indicate the triplet surface, and blue lines indicate the singlet surface. SC = singlet−triplet energy surface crossings. All energies are
expressed in kcal/mol.
3
Figure 5. Singlet and triplet transition structures for sp3 PCP ligand metalation. Free energies are relative to tetra-1 (kcal/mol).
Scheme 5 outlines a computed free energy surface for
discussed above, the lower energy pathway goes through a tight
ion pair intermediate (ion-pair) that forms via an associative
type transition state featuring a rupturing Ni−Br bond. The C−
metalation of (PCsp HP)NiBr2 starting with 3tetra-1. The
3
structural transformation of this species from a tetrahedral to
a square-planar geometry prior to C−H bond cleavage can
proceed via two possible triplet−singlet spin crossing points.
1
H bond is then cleaved via trans-TS to liberate HBr and
generate the pincer species. Interestingly, the overall free
3
1
3
The first crossing can occur as tetra-1 converts into trans-1-
twist (SC-1, Scheme 2). Alternatively, it is also possible that the
structures 1trans-1-twist and 1trans-1 are bypassed and 3tetra-1
energy of the nickelation reaction to give (PCsp P)NiBr and
HBr is endergonic by 12.1 kcal/mol relative to 3tetra-1;
moreover, this free energy decreases by only 0.5 kcal/mol when
i-Pr2P groups are modeled. All other density functionals tested
also indicate that this metalation reaction is endergonic: B3LYP
predicts a slightly higher reaction free energy, while BP86 and
MPWPW91 functionals predict the reaction free energy to be
endergonic by only 2−3 kcal/mol. These results indicate that
the added base likely plays an important thermodynamic role in
quenching the 1 equiv of HBr generated in the metalation
reaction. In THF, the acid−base reaction equilibrium between
NEt3 and HBr to give Et3NH+Br− is computed to be −15.6
1
is directly converted into trans-2 via SC-2. In this mechanism
the spin crossing likely occurs near the geometry of trans-2,
since the singlet and triplet energies are degenerate at this
geometry. Indeed, the located minimum energy crossing point
SC-2 and its geometry and energy are nearly identical with
those of 1trans-2 and 3trans-2 (see the Supporting Information
for the structure).
1
The trans-2 structure is then a branching point for the
oxidative addition and the stepwise ion pair pathways. As
6048
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
kcal/mol, which would render the overall process exergonic.
Alternatively, it is feasible that excess phosphine ligand plays a
role as the base required to quench the generated HBr.
reaction pathway for the sp3 POCOP ligand in order to allow a
comparison to the nickelation of PCsp HP. The geometrics of
3
the triplet tetrahedral ground state for (κP,P′-POCsp HOP)NiBr2
3
Comparing Metalation of (PCsp HP)NiBr2 and (PCsp HP)-
NiBr2. Investigating the nickelation of arene-based pincer
ligands revealed a similar mechanism for C−H activation in
comparison to the reaction of their saturated hydrocarbon
analogues, the major difference being the barrier heights for C−
H bond cleavage. The lower barrier for metalation of these
(3tetra-4, Figure 7) are nearly identical with those of its PCP
3
2
aromatic ligands is in part due to the fact that PCsp HP also
2
coordinates less favorably with NiBr2 to give either a tetrahedral
triplet complex or a trans singlet complex, both of which are
less stable than the corresponding nonmetalated intermediates
in the metalation of PCsp HP. The greater ground state energy
3
of (κP,P′-PCsp HP)NiBr2 might explain the lower barrier for the
2
metalation of these aromatic ligands. Despite the more
favorable Ni−C(sp2) bond present in the final product, M06
predicts the overall reaction free energy to be endergonic by
10.5 kcal/mol with Me2P and 5.4 kcal/mol with i-Pr2P moieties.
This implies that here, too, the base plays a crucial role in
rendering the C−H metalation thermodynamically favorable.
Figure 6 shows the monomeric tetrahedral triplet (κP,P′-
3
PCsp HP)NiBr2 complex tetra-3 and the singlet C−H bond
2
Figure 7. Triplet tetrahedral ground state and closed-shell ionic
transition structures and intermediate for nickelation of POCsp HOP.
3
3
analogue tetra-1. In THF solvent, coordination to tetrahedral
3
nickel is 3.3 kcal/mol more favorable in tetra-4 relative to
3tetra-1 (eq 3), whereas the opposite order of stabilities is
observed for the trans species (eq 4).
Figure 6. Tetrahedral intermediate and transition structure for
(PCsp HP)NiBr2 nickelation.
2
1
cleavage transition structure trans-TS3. The M06 activation
free energy for 1trans-TS3 is 13.9 kcal/mol, 3.5 kcal/mol lower
1
than the trans-TS discussed above. As before, this reaction
pathway is favored over oxidative addition that has a free energy
barrier of 24.3 kcal/mol. The lower barrier for sp2 vs sp3 C−H
bond metalation likely results from forming a shorter and
stronger Ni−C bond (1.946 Å) and a more stabilized anionic
sp2 C as the reaction proceeds through the ionic pathway.
These transition effects combine with the above-mentioned
favorable ground state effects to facilitate the metalation of sp2
C−H bonds.
We note that, despite significant effort, we were unable to
locate either an associative transition-state structure or a tight
ion pair intermediate on the potential energy surface.
Nevertheless, by analogy to the free energy surface shown in
As before, the ionic pathway mechanism begins with an
associative transition structure, trans-TS4assoc, that induces
1
an agostic C−H bond interaction with the nickel metal center
Scheme 5 for metalation of PCsp HP, we assume that the free
and results in cleavage of the Ni−Br bond to give the tight ion
3
energy for an associative-like transition structure for PCsp HP
pair structure trans-4ion. The free energy barrier for trans-
TS4assoc relative to 3tetra-4 is 27.4 kcal/mol, and the ion pair
intermediate free energy is 23.5 kcal/mol. This activation free
energy is ∼10 kcal/mol higher than the free energy barrier for
1
1
2
1
would be very similar to trans-TS3, since the bromide is
dislodged from the nickel center.
Comparing Nickelation of PCP and POCOP Ligands. The
competition experiments discussed earlier have revealed that
PCP-type ligands are metalated more readily than their
POCOP counterparts. To explore the kinetic and thermody-
namic roots of this phenomenon, we have computed the ionic
the associative transition structure for nickelation of PCsp HP.
3
This substantially higher barrier can be understood in terms of
the significantly weaker donor character of phosphinite-type
donor moieties relative to phosphines, which would not allow
6049
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
an effective stabilization of a transition state involving
substitution of the bromide ligand by a weakly donating C−
H bond.
facilitating C−H nickelation reactions by using more electron-
rich aromatic substrates and Ni precursors featuring more easily
ionized Ni−X bonds.
The C−H bond cleavage transition structure 1trans-TS4
1
shows marked differences in comparison to trans-TS. The
EXPERIMENTAL SECTION
■
breaking C−H bond distance is shorter by 0.22 Å and the
forming Ni−C bond is also shorter by 0.03 Å, while the
forming H−Br bond is 0.09 Å longer. This comparison
indicates that the C−H bond cleavage transition structure for
General Procedures. Unless otherwise indicated, all manipu-
lations were carried out under nitrogen using standard Schlenk
procedures and a drybox. Solvents were dried by passage over
molecular sieves contained in MBRAUN systems. Triethylamine was
dried by distillation over CaH2. The reagents isobutyronitrile, nickel
powder, bromine, ClP(i-Pr)2, ClP(t-Bu)2, ClP(Ph)2, 5-methyl-1,3-
benzenediol, 4,6-dichloro-1,3-benzenediol, methyl 2,4-dihydroxyben-
zoate, methyl 3,5-dihydroxybenzoate, and NaH were purchased from
Aldrich and used without further purification. 5-Methoxyresorcinol is
purchased from Chemsavers. 4-Methoxyresorcinol has been synthe-
sized following a published procedure.48 The synthesis and character-
ization of the ligands and their Ni complexes used in the present study
have been described in previously published reports.22 The Ni−Cl
analogue of 1c′ and the t-Bu2P analogues of ligands 1e and 1i have
been reported previously.11d
POCsp HOP is “earlier” along the reaction coordinate than the
3
corresponding PCsp HP transition structure. The activation free
3
1
energy for trans-TS4 is 28.3 kcal/mol, 0.9 kcal/mol higher
than 1trans-TS4assoc. The barrier for C−H cleavage relative to
the 1trans-4ion intermediate is 4.8 kcal/mol, which is similar to
the 3.4 kcal/mol barrier for the analogous step in the
nickelation of PCsp HP; this confirms that the most energetically
3
important step is the Ni−Br bond rupture. As before, formation
of the pincer species (κP,C,P′-POCsp O)PNiBr is endergonic with
3
a reaction free energy of 14.2 and 7.8 kcal/mol for Me2P and i-
Pr2P moieties, respectively. Finally, we have also found that the
Most NMR spectra were recorded at 400 (1H) and 161.9 MHz
(31P) using a Bruker AV400rg spectrometer, or at 400 (1H) and
100.56 MHz (13C{1H}) using a Bruker ARX400 spectrometer. The
31P{1H} NMR spectrum of complex 1 h′ was recorded at 202.5 MHz
using a Bruker AV500 spectrometer. Chemical shift values are reported
in ppm (δ) and referenced internally to the residual solvent signals (1H
and 13C: 7.26 and 77.16 ppm for CDCl3; 7.16 and 128.06 ppm for
C6D6) or externally (31P, H3PO4 in D2O, δ= 0). Coupling constants
are reported in Hz. UV/vis spectra were measured on a Varian Cary
500i.
General Procedure for Competition Reactions. Ligands 1a−j.
A THF solution (650 μL) containing approximately equimolar
quantities of ligands 6 (0.175 mmol, 60 mg) and 1 (0.175 mmol, 65
mg) was analyzed by 31P NMR to confirm the molar ratio of the
ligands, and then added to a homogeneous THF solution (650 μL)
containing {(i-PrCN)NiBr2}n (0.132 mmol, 38 mg) and NEt3 (49 μL,
0.35 mmol). The initial blue color of the reaction mixture changed
instantly, turning to brown first and then yellow, and a white
precipitate formed over time. The mixture was stirred for 5 min at
room temperature and samples were withdrawn over regular intervals
for analysis by 31P NMR. Reaction progress was determined on the
basis of the relative intensities of the 31P signals for the ligands and the
in situ formed complexes. To confirm the irreversible nature of the
nickelation process, the samples were also analyzed after 24 h.
Ligands 3 and 4. Applying the above procedure to the aliphatic
ligands (but in Toluene instead of THF) results in formation of a
homogeneous, dark brown solution, which was heated over three days
at 100 °C before analysis by 31P NMR.
3
pincer complex featuring the PCsp P ligand is slightly more
3
stable (by 1.1 kcal/mol) in comparison to its POCsp OP
analogue.
CONCLUSION
■
The combined experimental and computational investigations
described above have shed some light on the relative rates and
mechanisms of the direct nickelation of the PCP- and POCOP-
type pincer ligands studied. We have confirmed and quantified
the more facile nickelation of PCP vs POCOP ligands and
aromatic vs aliphatic backbones. The more facile metalation of
the aromatic C−H bonds is related to their more effective
coordination to the electrophilic Ni center and better
stabilization of the developing Ni−C bond, as well as to the
absence of dinuclear species in the ground states of the
aromatic ligands. The preference for PCP vs POCOP
nickelation likely results from the better chelation of the
phosphine moieties; the importance of chelation is also
reflected in the preference for P(i-Pr)2 vs P(t-Bu)2 and PPh2
moieties.
The experimental finding that electron-releasing substituents
favor the nickelation of aromatic POCOP ligands is consistent
with a mechanism resembling electrophilic aromatic substitu-
tion, which is in stark contrast to what has been observed for
direct C−H palladation reactions of aromatic substrates.31,32
Our DFT calculations indicate that oxidative addition is higher
in energy than a two-step pathway characterized by a bromide
dissociation to give a tight-ion pair followed by bromide
deprotonation of the C−H bond. The crucial stabilization of
the electron-depleted metal center in such a pathway explains
the importance of Ni−C(H) interactions and hence the
preference for electron-rich aromatic ligands. These features,
namely the preference for activation of more electron rich C−H
bonds and the complete dissociation of the internal H acceptor
(Br−), distinguish the direct metalation process in our Ni-
promoted systems from the much more extensively investigated
Pd-based systems wherein electron-deficient C−H bonds are
metalated preferentially and the H transfer involves Pd-bound
acceptors (halides, acetate, carbonate, etc.).
ASSOCIATED CONTENT
■
S
* Supporting Information
Full citation for ref 36 and a plot showing the free energy
relationship between relative nickelation rates for substituted
aromatic POCOP ligands and Hammett σ parameters. Free of
charge at http://pubs.acs.org. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: dhe@chem.byu.edu (D.H.E.); zargarian.davit@
umontreal.ca (D.Z.).
Future studies will be directed at examining the C−H
metalation of other pincer ligands (PCN, POCN, NCN, etc.).
If these investigations confirm the main findings of the current
study, it should be possible to develop successful strategies for
Notes
The authors declare no competing financial interest.
6050
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
Acta Crystallogr. 1985, C41, 893. (c) Fryzuk, M. D.; Leznoff, D. B.; Ma,
E. S. F.; Rettig, S. J.; Young, V. G., Jr. Organometallics 1998, 17, 2313.
(d) Ingleson, M. J.; Fullmer, B. C.; Buschhorn, D. T.; Fan, H.; Pink,
M.; Huffman, J. C.; Caulton, K. G. Inorg. Chem. 2008, 47, 407.
(e) Adhikari, D.; Basuli, F.; Fan, H.; Huffman, J. C.; Pink, M.;
Mindiola, D. J. Inorg. Chem. 2008, 47, 4439. (f) Xu, G.; Sun, H.; Li, X.
Organometallics 2009, 28, 6090. (g) Hosokawa, S.; Ito, J.; Nishiyama,
H. Organometallics 2010, 29, 5773. (h) Bhattacharya, P.; Krause, J. A.;
Guan, H. Organometallics 2011, 30, 4720.
ACKNOWLEDGMENTS
■
The authors are grateful to: Brigham Young University for
financial support to D. H. E., access to the Fulton Super-
computing Center, and undergraduate research award to M. L.
L.; NSERC of Canada for a Discovery Grant to D. Z.;
́ ́
Universite de Montreal and Centre in Green Chemistry and
Catalysis for graduate fellowships to B. V.; the Center for
Catalytic Hydrocarbon Functionalization, an Energy Frontier
Research Center funded by the U.S. Department of Energy,
Office of Science, Office of Basic Energy Sciences (Award
Number DE-SC0001298, D. H. E.); the American Chemical
Society Petroleum Research Fund (D. H. E.). We have also
benefited from valuable discussion with Dr. Laure Benhamou
and Dr. Denis Spasyuk.
(8) (PCP)Co complexes can be prepared via direct C−H
metalation,7f but preparation of the analogous POCOP complexes
via direct C−H metalation has proven difficult (see: Lefev
̀
re, X. M.Sc.
thesis, Universite de Montreal, 2010). A recent report has described
́
́
the preparation of (POCOP)Co complexes via salt metathesis using
lithiated ligands: Hebden, T. J.; St. John, A. J.; Gusev, D. G.; Kaminsky,
W.; Goldberg, K. I.; Heinekey, D. M. Angew. Chem., Int. Ed. 2011, 50,
1873.
(9) (a) Gusev, D. G.; Madott, M.; Dolgushin, F. M.; Lyssenko, K. A.;
Antipin, M. Yu. Organometallics 2000, 19, 1734. (b) Dani, P.;
Toorneman, M. A. M.; van Klink, G. P. M.; van Koten, G.
Organometallics 2000, 19, 5287. (c) Agapie, T.; Bercaw, J. E.
Organometallics 2007, 26, 2957.
DEDICATION
■
†This report is dedicated to Prof. Kenneth G. Caulton as a
tribute to his numerous and significant contributions to
organometallic chemistry.
(10) (a) Castonguay, A.; Sui-Seng, C.; Zargarian, D.; Beauchamp, A.
L. Organometallics 2006, 25, 602. (b) Castonguay, A.; Beauchamp, A.
L.; Zargarian, D. Organometallics 2008, 27, 5723. (c) Castonguay, A.;
Spasyuk, D. M.; Madern, N.; Beauchamp, A. L.; Zargarian, D.
Organometallics 2009, 28, 2134. (d) Castonguay, A.; Beauchamp, A. L.;
Zargarian, D. Inorg. Chem. 2009, 48, 3177.
(11) (a) Valente, G. B.; Oscar, B. P.; Cesar, H- A.; Toscano, R. A.;
Morales-Morales, D. Tetrahedron Lett. 2006, 47, 5059. (b) Pandarus,
V.; Zargarian, D. Chem. Commun. 2007, 978. (c) Pandarus, V.;
Zargarian, D. Organometallics 2007, 26, 4321. (d) Chakraborty, S.;
Krause, J. A.; Guan, H. Organometallics 2009, 28, 582. (e) Zhang, J.;
Medley; Krause, J. A.; Guan, H. Organometallics 2010, 29, 6393.
(f) Chakraborty, S.; Zhang, J.; Krause, J. A.; Guan, H. J. Am. Chem. Soc.
REFERENCES
■
(1) (a) Shilov, A. E.; Shul’pin, G. B. Chem. Rev. 1997, 97, 2879.
(b) Crabtree, R. H. J. Chem. Soc., Dalton Trans. 2001, 2437.
(c) Labinger, J. A.; Bercaw., J. E. Nature 2002, 417, 507. (d) Crabtree,
R. H. J. Organomet. Chem. 2004, 689, 4083. (e) Activation and
Functionalization of C-H Bonds; Goldman, A. S., Goldberg, K. I., Eds.;
American Chemical Society: Washington, DC, 2004; ACS Symposium
Series 885.
(2) (a) Kakiuchi, F.; Murai, S. Acc. Chem. Res. 2002, 35, 826.
(b) Ritleng, V.; Sirlin, C.; Pfeffer, M. Chem. Rev. 2002, 102, 1731.
(c) Dick, A. R.; Sanford, M. Tetrahedron 2006, 62, 2439. (d) Alberico,
D.; Scott, M. E.; Lautens, M. Chem. Rev. 2007, 107, 174. (e) Lewis, J.
C.; Bergman, R. G.; Ellman, J. A. Acc. Chem. Res. 2008, 41, 1013.
(f) Stuart, D. R.; Fagnou, K. Science 2007, 316, 1172.
2010, 132, 8872. (g) Lefev
Organomet. Chem. 2011, 864. (h) Lefev
̀
re, X.; Spasyuk, D. M.; Zargarian, D. J.
re, X.; Durieux, G.; Lesturgez,
̀
S.; Zargarian, D. J. Mol. Catal. A 2011, 335, 1. (i) Salah, A. B.;
Zargarian, D. Dalton Trans. 2011, 40, 8977. (j) Salah, A.; Offenstein,
C.; Zargarian, D. Organometallics 2011, 30, 5352. (k) Huang, F.;
Zhang, C.; Jiang, J.; Wang, Z.-X.; Guan, H. Inorg. Chem. 2011, 50,
3816. (l) Chakraborty, S.; Patel, Y. J.; Krause, J. A.; Guan, H.
Polyhedron 2012, 32, 30. (m) Xu, G.; Li, X.; Sun, H. J. Organomet.
Chem. 2011, 696, 3011.
(12) A recent report has shown that the closely related PNCNP-type
complexes of nickel can be prepared in the presence of a strong Lewis
acid such as B(C6F5)3: Gwynne, E. A.; Stephan, D. W. Organometallics
2011, 30, 4128−4135. This strong Lewis acid appears to be required
for converting a species formed initially via nickelation at the N−H
moiety of the ligand; in contrast, palladation of the central C−H
moiety of this PNCNP ligand proceeds readily in the absence of Lewis
acids.
(13) (a) Spasyuk, D. M.; van der Est, A.; Zargarian, D.
Organometallics 2009, 28, 6531. (b) Spasyuk, D. M.; Zargarian, D.
Inorg. Chem. 2010, 49, 6203. (c) Spasyuk, D. M.; Gorelsky, S. I.; van
der Est, A.; Zargarian, D. Inorg. Chem. 2011, 50, 2661. (d) Zhang, B.-
S.; Wang, W.; Shao, D.-D.; Hao, X.-Q.; Gong, J.-F.; Song, M.-P.
Organometallics 2010, 29, 2579. (e) Niu, J.-L.; Chen, Q.-T.; Hao, X.-
Q.; Zhao, Q.-X.; Gong, J.-F.; Song, M.-P. Organometallics 2010, 29,
2148. (f) Yang, M.-J.; Liu, Y.-J.; Gong, J.-F.; Song, M.-P. Organo-
metallics 2011, 30, 3793.
(14) Most (NCN)Ni complexes have been prepared via strategies
based on lithiation/nickelation (using Ni(II) precursors) or oxidative
addition of C-halide moieties (using Ni(0) precursors). For
representative examples see: (a) Grove, D. M.; van Koten, G.;
Ubbels, H. J. C.; Zoet, R. Organometallics 1984, 3, 1003. (b) Stol, M.;
Snelders, D. J. M.; Godbole, M. D.; Havenith, R. W. A.; Haddleton, D.;
Clarkson, G.; Lutz, M.; Spek, A. L.; van Klink, G.; P., M.; van Koten,
G. Organometallics 2007, 26, 3985. (c) Bugarin, A.; Connell, B. T.
(3) (a) Bolig, A. D.; Brookhart, M. J. Am. Chem. Soc. 2007, 129,
14544. (b) Bullock, R. M. Catalysis Without Precious Metals; Wiley-
VCH: Weinheim, Germany, 2010. (c) Kulkarni, A. A.; Daugulis, O.
Synthesis 2009, 4087. (d) Chakraborty, S.; Guan, H. Dalton Trans.
2010, 39, 7427.
(4) Kleiman, J. P.; Dubeck, M. J. Am. Chem. Soc. 1963, 85, 1544.
(5) Moulton, C. J.; Shaw, B. L. J. Chem. Soc., Dalton Trans. 1976,
1020.
(6) (a) Rimml, H.; Venanzi, L. M. J. Organomet. Chem. 1983, 259,
C6−C7. (b) Rimml, H.; Venanzi, L. M. J. Organomet. Chem. 1984,
260, C52−C54. (c) Kennedy, A. R.; Cross, R. J.; Muir, K. W. Inorg.
Chim. Acta 1995, 231, 195. (d) Huck, W. T. S.; Snellink-Ruel, B.; van
̈
Veggel, F. C. J. M.; Reinhoudt, D. N. Organometallics 1997, 16, 4287.
(e) Kozhanov, K. A.; Bubnov, M. P.; Cherkasov, V. K.; Fukin, G. K.;
Abakumov, G. A. Chem. Commun. 2003, 2610. (f) van der Boom, M.
E.; Liou, S. Y.; Shimon, L. J. W.; Ben-David, Y.; Milstein, D. Inorg.
́ ́
Chim. Acta 2004, 357, 4015. (g) Groux, L. F.; Belanger-Gariepy, F.;
Zargarian, D. Can. J. Chem. 2005, 83, 634. (h) Castonguay, A.;
Charbonneau, F.; Beauchamp, A. L.; Zargarian, D. Acta Crystallogr.,
Sect. E 2005, E61, m2240. (i) Sui-Seng, C.; Castonguay, Chen, A. Y.;
Gareau, D.; Groux, L. F.; Zargarian, D. Top. Catal. 2006, 37, 81.
(j) Castonguay, A.; Beauchamp, A. L.; Zargarian, D. Acta Crystallogr.,
Sect. E 2007, E63, m196. (k) Boro, B. J.; Duesler, E. N.; Goldberg, K.
I.; Kemp, R. A. Inorg. Chem. 2009, 48, 5081. (l) Boro, B. J.; Dickie, D.
A.; Goldberg, K. I.; Kemp, R. A. Acta Crystallogr., Sect. E 2008, E64,
m1304. (m) Boro, B. J.; Dickie, D. A.; Duesler, E. N.; Goldberg, K. I.;
Kemp, R. A. Acta Crystallogr., Sect. E 2008, E64, m1402. (n) Schmeier,
T. J.; Hazari, N.; Incarvitoa, C. D.; Raskatovb, J. A. Chem. Commun.
2011, 47, 1824.
(7) For pincer complexes of Fe see: (a) Creaser, C. S.; Kaska, W. C.
Inorg. Chim. Acta 1978, 30, L325−L326. (b) de Koster, A.; Kanters, J.
A.; Spek, A. L.; van der Zeijden, A. A. H.; van Koten, G.; Vrieze, K.
6051
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
Organometallics 2008, 27, 4357. (d) Lee, D. H.; Sungwon, J. H.; Park,
S. Bull. Korean Chem. Soc. 2008, 29, 187. (e) Hurtado, J.; Ibanez, A.;
3786. (b) More recently, we have reported a more detailed and
optimized protocol.10a
(26) (a) For the original report on the preparation of this precursor
see: Hathaway, B. J.; Holah, D. G. J. Chem. Soc. 1964, 2400. (b) For
examples of its use in the preparation of pincer complexes of nickel see
refs 10c, 11b,c,h−j, and 13a,b.
Rojas, R.; Valderrama, M.; Frohlich, R. J. Braz. Chem. Soc. 2011, 22,
̈
1750. (f) For a review on this subject see: Gossage, R. A.; van de Kuil,
L. A.; van Koten, G. Acc. Chem. Res. 1998, 31, 423.
(15) van Koten’s group has reported an alternative synthetic path to
(NCN)Ni complexes using gold as transmetalation agent: Contel, M.;
Stol, M.; Casado, M. A.; van Klink, G. P. M.; Ellis, D. D.; Spek, A. L.;
van Koten, G. Organometallics 2002, 21, 4556.
(16) To our knowledge, there is only one reported example of a
(NCN)Ni complex prepared by direct C−H nickelation: Shao, D.-D.;
Niu, J.-L.; Hao, X.-Q.; Gong, J.-F.; Song, M.-P. Dalton Trans. 2011, 40,
9012.
(17) For rare examples of Ni complexes featuring nonmetalated
aromatic pincer ligands see: (a) Smith, R. C.; Protasiewicz, J. D.
Organometallics 2004, 23, 4215. (b) Johnson, M. T.; Wendt, O. F.
Inorg. Chim. Acta 2011, 367, 222. For an example of a cis Pd complex
featuring a chelating, nonmetalated aliphatic bis(phosphine) ligand
(27) Vabre, B.; Zargarian, D. Unpublished results.
(28) The molar mass of (MeCN)nNiBr2 is somewhat uncertain
because of the poorly defined acetonitrile content that varies
considerably as a function of the sample’s age and the specific
protocol used for the recrystallization and drying steps. This
uncertainty in the molar mass is not a major concern when conducting
preparative experiments wherein the precursor is often used in excess,
but this issue is crucial for the kinetic measurements reported in this
study.
(29) It should be noted that nickelation of the aliphatic ligands 3 and
4 proceeds sufficiently slowly to allow direct rate measurements, but
using large excess of Ni and NEt3, as required under pseudo-first-order
conditions, often led to variable amounts of unidentified side products
and gave irreproducible rates. In contrast, nickelation reactions using
small excesess of NEt3 and Ni do not generate side products.
(30) One of the reviewers of our paper urged us to explore the
possibility that in some instances the initially formed pincer complexes
might react with unreacted ligands to generate new pincer complexes;
these types of “trans-cyclometalation” reactions have been observed
earlier between (NCN)PtCl complexes and PCHP ligands to generate
(PCP)PtCl and NCHN ligands. See: Albrecht, M.; Dani, P.; Lutz, M.;
Spek, A. L.; van Koten, G. J. Am. Chem. Soc. 2000, 122, 11822. We can
confirm here that trans-cyclometalation reactions do not take place
with aliphatic ligands/complexes even under forcing conditions. For
instance, heating a mixture of preformed 3′ and excess 4 at 100 °C
over 15 h showed no sign of reaction. On the other hand, trans-
cyclometalation does occur with aromatic ligands under forcing
conditions. For instance, heating a mixture of preformed complex 2′
and 7 equiv of ligand 1a in toluene at 100 °C for 18 h showed the
formation of ligand 2 and complex 1a′ (final ratio 2′:1a ≈ 3:2 by 31P
NMR). It should be emphasized, however, that the nickelation ratios
reported here for the aromatic ligands/complexes are invariable under
the conditions of the kinetic experiments: i.e., equimolar quantities of
ligands reacting with a slight deficiency of {(i-PrCN)NiBr2}n at room
temperature.
see: (c) Sjovall, S.; Johansson, M. H.; Andersson, C. Eur. J. Inorg.
̈
Chem. 2001, 2907.
(18) Pandarus, V.; Castonguay, A.; Zargarian, D. Dalton Trans. 2008,
4756.
(19) Di- or trinuclear macrocyclic complexes featuring nonmetalated
pincer complexes have also been reported for Rh, Pd, Pt, and Au:
(a) Pryde, A. J.; Shaw, B. L.; Weeks, B. J. Chem. Soc., Chem. Commun.
1973, 947. (b) Al-Salem, N. A.; Empsall, H. D.; Markham, R.; Shaw, B.
L.; Weeks, B. J. Chem. Soc., Dalton Trans. 1979, 1972. (c) Al-Salem, N.
A.; McDonald, W. S.; Markham, R.; Norton, M. C.; Shaw, B. L. J.
Chem. Soc., Dalton Trans. 1980, 59. (d) Crocker, C.; Errington, R. J.;
Markham, R.; Moulton, C. J.; Odell, K. J.; Shaw, B. L. J. Am. Chem. Soc.
1980, 102, 4373. (e) Hill, W. E.; Minahan, D. M. A.; Taylor, J. G.;
McAuliffe, C. A. J. Am. Chem. Soc. 1982, 104, 6001. (f) Al-Baker, S.;
Hill, W. E.; McAuliffe, C. A. J. Chem. Soc., Dalton Trans. 1985, 2655.
(g) van der Boom, M. E.; Gozin, M.; Ben-David, Y.; Shimon, L. J. W.;
Frolow, F.; Kraatz, H.-B.; Milstein, D. Inorg. Chem. 1996, 35, 7068.
(h) Dilworth, J. R.; Zheng, Y.; Griffiths, D. V. J. Chem. Soc., Dalton
Trans. 1999, 1877.
(20) For representative reviews on the contribution of computational
methods to delineating the mechanisms of C−-H activation reactions
see: (a) Dedieu, A. Chem. Rev. 2000, 100, 543. (b) Frenking, G.;
Frohlich, N. Chem. Rev. 2000, 100, 717. (c) Niu, S.; Hall., M. B. Chem.
̈
(31) We have proposed that this nonmetalated intermediate is the
dinuclear species {trans-(μ-3)NiBr2}2 featuring a 16-membered
metallacycle.18 Such complexes have also been isolated during the
Rev. 2000, 100, 353. (d) Ke, Z.; Cundari, T. R. Organometallics 2010,
29, 821.
(21) For recent reviews on C−H bond activation mechanisms see:
(a) Balcells, D.; Clot, E.; Eisenstein, O. Chem. Rev. 2010, 110, 749.
(b) Boutadla, Y.; Davies, D. L.; Macgregor, S. A.; Poblador-
Bahamonde, A. I. Dalton Trans. 2009, 5820. (c) Perutz, R. N.; Sabo-
Etienne, S. Angew. Chem., Int. Ed. 2007, 46, 2578. (d) Niu, S.; Hall, M.
3
attempted nickelation of the PCsp P ligand {i-Pr2P(CH2)2}2CH2 and
fully characterized.10b
(32) Plotting σp vs ln(relative rates) gave a Hammett reaction
constant of −2.87, which indicates a positive charge developing on the
p-C atom during nickelation (see the Supporting Information).
(33) (a) Garcia-Cuadrado, D.; Braga, A. A. C.; Maseras, F.;
Echavaren, A. M. J. Am. Chem. Soc. 2006, 128, 1066. (b) Garcia-
Cuadrado, D.; de Mendoza, P.; Braga, A. A. C.; Maseras, F.;
Echavaren, A. M. J. Am. Chem. Soc. 2007, 129, 6880. (c) Pascual, S.; de
Mendoza, P.; Braga, A. A. C.; Maseras, F.; Echavaren, A. M.
Tetrahedron 2008, 64, 6021.
̀
B. Chem. Rev. 2000, 100, 353. (e) Torrent, M.; Sola, M.; Frenking, G.
Chem. Rev. 2000, 100, 439. (f) Dedieu, A. Chem. Rev. 2000, 100, 543.
(22) The synthesis and characterization of these ligands and their
(pincer)NiX complexes have been reported previously, as follows: 1a
and 3;11b,c 1b;11a,i−k 1c;11d,e 1d;10c 1e−j;27 2;6l,10c 4.10b The nickel
pincer complexes (POCOPt‑Bu)NiX (X = Cl, H, SAr, etc.) have been
reported by Guan’s group.11d−f,k
(34) (a) Lafrance, M.; Rowley, C. N.; Woo, T. K.; Fagnou, K. J. Am.
Chem. Soc. 2006, 128, 8754. (b) Campeau, L. C.; Fagnou, K. Chem.
Commun. 2006, 1253.
(23) (a) Beedford, R. B.; Draper, S. M.; Scully, P. N.; Welch, S. L.
New J. Chem. 2000, 24, 745. (b) Morales-Morales, D.; Grause, C.;
Kasaoka, K.; Redon, R.; Cramer, R. E.; Jensen, C. M. Inorg. Chim. Acta
2000, 300−302, 958.
(35) A reviewer of our paper has suggested that substituents located
at the meta position with respect to the central C−H bond would be
expected to favor nickelation by sterically restricting the conforma-
tional freedom of the donor moiety in its proximity and hence
increasing its rate of coordination to the Ni center. This explanation is
consistent with the relative nickelation rates we have observed for
ligands 1g,h and hence should be taken into consideration when
discussing the impact of meta substituents on the nickelation of
aromatic POCOP ligands.
(24) (a) The synthesis of 1,5-(i-Pr2P)2-(CH2)5 has been reported in:
Geier, S.; Goddard, R.; Holle, S; Jolly, P. W.; Kruger, C.; Lutz, F.
̈
Organometallics 1997, 16, 1612. (b) An optimized protocol for the
synthesis of the t-Bu analogue of this ligand has been reported by
Gusev: Gusev, D. G.; Lough, A. J. Organometallics 2002, 21, 5091. (c)
We reported subsequently a more detailed and higher yielding
protocol for the preparation of 1,5-(i-Pr2P)2-(CH2)5.10b
(25) (a) For a report on the synthesis of 1,3-(i-Pr2PCH2)2-C6H4 see:
(36) Frisch, M. J., et al. Gaussian 09, Revision A.02; Gaussian, Inc.,
Rybtchinski, B.; Ben-David, Y.; Milstein, D. Organometallics 1997, 16,
Wallingford, CT, 2009.
6052
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053

Organometallics
Article
(37) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215.
(38) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157.
(39) (a) Sieffert, N.; Buhl, M. Inorg. Chem. 2009, 48, 4622.
̈
(b) Minenkov, Y.; Occhipinti, G.; Jensen, V. R. J. Phys. Chem. A 2009,
113, 11833. (c) Stewart, I. C.; Benitez, D.; O’leary, D. J.; Tkatchouk,
E.; Day, M. W.; Goddard, W. A., III; Grubbs, R. H. J. Am. Chem. Soc.
2009, 131, 1931. (d) Benitez, D.; Tkatchouk, E.; Goddard, W. A., III
Organometallics 2009, 28, 2643. (e) Fedorov, A.; Couzijn, E. P. A.;
Nagornova, N. S.; Boyarkin, O. V.; Rizzo, T. R.; Chen, P. J. Am. Chem.
Soc. 2010, 132, 13789. (f) Takatani, T.; Sears, J. S.; Sherrill, C. D. J.
Phys. Chem. A 2010, 114, 11714. (g) Pratt, L. M.; Voit, S.; Okeke, F.
N. J. Phys. Chem. A 2011, 115, 2281. (h) Kulkarni, A. D.; Truhlar, D.
G. J. Chem. Theory Comput. 2011, 7, 2325.
(40) Minenkov, Y.; Occhipinti, G.; Jensen, V. R. J. Phys. Chem. A
2009, 113, 11833.
(41) (a) Harvey, J. N.; Aschi, M.; Schwarz, H.; Koch, W. Theor.
Chem. Acc. 1998, 99, 95. (b) Harvey, J. N.; Aschi, M. Phys. Chem.
Chem. Phys. 1999, 1, 5555.
(42) C Legault, C. Y. YLview, version 1.0b; Universite de Sherbrooke,
́
2009 (http://www.cylview.org).
(43) We note that only dinuclear coordination has been explored
here. It is plausible that an oligomeric species similar to trans,trans-1
might be formed upon complexation of the PCsp HP ligand with NiBr2.
3
(44) We have obtained a calculated ΔG value of 34.7 kcal/mol for a
single C−H nickelation from the dinuclear intermediate, confirming
that nickelation pathways going through a prior dissociation of the
dinuclear intermediates into mononuclear species are significantly
more favorable. This preference for nickelation from the mononuclear
species can be ascribed to the significant contortion of the PCP ligand
and the formation of a highly distorted geometry around the Ni center
required for bringing the C−H bond into the coordination sphere of
the Ni center in the dinuclear species.
(45) See, for instance, the proposed mechanism for C−H activation
reactions of (tris(pyrazolyl)borate)Ru(PH3)CH3: Lam, W. H.; Jia, G.
C.; Lin, Z. Y.; Lau, C. P.; Eisenstein, O. Chem. Eur. J. 2003, 9, 2775.
(46) Gorelsky, S. I.; Lapointe, D.; Fagnou, K. J. Am. Chem. Soc. 2008,
130, 10848.
(47) Ess, D. H.; Bischof, S. M.; Oxgaard, J.; Periana, R. P.; Goddard,
W. A., III Organometallics 2008, 27, 6440.
(48) Foti, M. C.; Daquino, C.; Mackie, I. D.; DiLabio, G. A.; Ingold,
K. U. J. Org. Chem. 2008, 73, 9270.
6053
dx.doi.org/10.1021/om3003784 | Organometallics 2012, 31, 6041−6053