Thermal chemistry of a tungsten trimethylsilylallyl complex in benzene and fluorobenzenes

Article abstract of DOI:10.1021/om300765e

The thermolysis of Cp*W(NO)(Npt)(η³-CH ₂CHCHSiMe₃) (1; Cp* = η⁵-C ₅Me₅; Npt = CH₂CMe₃) in benzene at 55 °C generates three isomeric products having the composition Cp*

Full text of DOI:10.1021/om300765e

Article
pubs.acs.org/Organometallics
Thermal Chemistry of a Tungsten Trimethylsilylallyl Complex in
Benzene and Fluorobenzenes
Catherine Chow, Brian O. Patrick, and Peter Legzdins*
Department of Chemistry, The University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z1
S
* Supporting Information
ABSTRACT: The thermolysis of Cp*W(NO)(Npt)(η3-CH2CHCHSiMe3) (1; Cp* = η5-C5Me5; Npt = CH2CMe3) in benzene
at 55 °C generates three isomeric products having the composition Cp*W(NO)(H)(η3-Me3SiCHCHCHPh) (2). These are
isomers of the expected Cp*W(NO)(Ph)(η3-Me3SiCHCHCH2) compound and result from an intramolecular Ph/allyl H
exchange. Thermolysis of 2 in the presence of pyridine produces the η2-olefin pyridine adduct Cp*W(NO)(η2-Me3SiCH2CH
CHPh)(C5H5N) (3). However, when the same reaction is carried out in deuterobenzene with 10 equiv of pyridine, NMR
spectroscopic data suggest that the meso hydrogen of the allyl ligand is exchanged for a deuterium atom before pyridine trapping
occurs. The activation of fluorobenzenes (i.e., pentafluorobenzene, p-difluorobenzene, and o-difluorobenzene) by Cp*W(NO)-
(Npt)(η3-CH2CHCHSiMe3) has also been studied, and for these substrates, C−H bond activation occurs exclusively. Selectivity
for the activation of these C−H bonds appears to be determined by sterics. Intramolecular migration of the newly formed
fluoroaryl ligands onto the allyl ligands does not occur when there is a fluorine atom in the position ortho to the newly formed
W−C bond. This behavior is probably a manifestation of the fact that metal−o-fluoroaryl bonds tend to be stronger than metal−
aryl linkages. All new complexes have been characterized by conventional spectroscopic and analytical methods, and the solid-
state molecular structures of most of them have been established by single-crystal X-ray crystallographic analyses.
presence of benzene, since we had previously established that
different members of the Cp*W(NO)(alkyl)(η3-allyl) family of
complexes react differently with this substrate, depending on
the substituent on the allyl ligand. For instance, thermolysis of
Cp*W(NO)(Npt)(η3-CH2CHCMe2) in benzene at 50 °C
results in the clean production of the phenyl complex
Cp*W(NO)(Ph)(η3-CH2CHCMe2).6 In contrast, a benzene
solution of Cp*W(NO)(Npt)(η3-CH2CHCHMe) maintained
at 26 °C for 4 h generates a complex mixture of at least five new
organometallic compounds resulting from isomerization of the
expected phenyl complex.7 We quickly discovered that
maintaining a benzene solution of 1 at 55 °C does not afford
the corresponding phenyl complex Cp*W(NO)(Ph)(η3-
CH2CHCHSiMe3) as a tractable final product. Instead, three
isomeric hydrido allyl complexes of tungsten are formed. In
order to gain some mechanistic insights into these trans-
formations, we extended our studies to encompass some
fluorobenzenes as well. In this article we present the results of
our investigations.
INTRODUCTION
■
The study of complexes that selectively break hydrocarbon C−
H bonds in unfunctionalized starting materials and transform
them into industrially useful and value-added products
continues unabated, motivated in part by the desire to utilize
the world’s finite petroleum resources more efficiently.1,2 Our
recent contributions to this area of chemistry have involved the
family of Cp*W(NO)(alkyl)(η3-allyl) (Cp* = η5-C5Me5)
complexes,3 which initiate alkane C−H activations via η2-
diene or η2-allene intermediates.4 In this connection we
recently investigated the scope and mechanism of the C−H
activation chemistry initiated by Cp*W(NO)(Npt)(η3-
CH2CHCHSiMe3) (1; Npt = CH2CMe3) in order to ascertain
the effects of increasing the steric bulk of the allyl ligand and
modifying its electronic properties by introducing the SiMe3
group.5 We discovered that thermolysis of 1 at 55 °C leads to
the loss of neopentane and the generation of the 16-electron η2-
allene intermediate complex Cp*W(NO)(η2-CH2C
CHSiMe3), which selectively activates hydrocarbons at its
methyl groups (Scheme 1).
In the case of linear alkanes, only terminal C−H activation
occurs. This selectivity persists in the presence of an ether
functionality, but not with other oxygen-containing substrates
such as aldehydes and alcohols.5 In light of these results, we
then decided to investigate the thermal chemistry of 1 in the
Received: August 8, 2012
Published: November 6, 2012
© 2012 American Chemical Society
8159
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
Scheme 1
Scheme 2
hydrogen at the meso position.8 Finally, the signal arising from
the hydride integrates to 1H relative to the Cp* Me peak. The
original source of the hydrogen atom for the hydride must
therefore be the allyl ligand, since there is no other plausible
RESULTS AND DISCUSSION
■
Thermal Reactions of 1 in Benzene. Thermolysis of 1 in
benzene at 55 °C generates three isomeric products, 2a−c
(Scheme 2), via migration of a phenyl group onto the allyl
ligand. These products are fairly air-stable and can be handled
without recourse to inert atmosphere for short periods of time
without noticeable degradation.
The expected initial product of these transformations,
namely Cp*W(NO)(Ph)(η3-CH2CHCHSiMe3), has not been
isolated, nor has it been directly observed by 1H NMR
spectroscopy. However, evidence for its existence is provided
1
source of H in the reaction mixture.
Isomerization to the 1,3-disubstituted allyl ligand appears to
be very favored. Since C−H activation by 1 presumably
proceeds via an η2-allene intermediate, it is expected that the
products of thermolysis in benzene should be similar to those
of Cp*W(NO)(Npt)(η3-CH2CHCHPh).9 However, because
the trimethylsilyl substituent on the allyl ligand causes the
resulting 1,3-disubstituted allyl ligand to be nonsymmetric, the
complex can exist in one of two conformations: the
trimethylsilyl substituent can be placed either proximal (2a)
or distal (2b) to the NO ligand.
1
by the thermolysis of 1 in benzene-d6. An expansion of the H
NMR spectrum of the final reaction mixture of this experiment
is presented in Figure 1. The existence, however transient, of
Since Cp*W(NO)(Npt)(η3-CH2CHCHSiMe3) (1) exists as
a single isomer with the trimethylsilyl substituent proximal to
the NO ligand,5 the initial product is expected to be the
1
corresponding mono isomer. Monitoring the reaction by H
NMR spectroscopy at 75 °C (Figure 2) confirms this
Figure 1. 1H NMR spectrum (300 MHz) of the final reaction mixture
of 1 thermolyzed in C6D6. Assignments (ppm): δ −1.08, W−H; 0.31,
SiMe3; 0.68, CHSiMe3; 1.71, Cp* Me; 2.25, CHPh. Inset:
magnification of the two signals arising from the non-meso protons
on the allyl ligand.
Figure 2. Product distribution during the thermolysis of 1 in C6D6 at
75 °C, as monitored by H NMR spectroscopy.
1
the phenyl complex can be inferred by three key pieces of
evidence. First, the absence of the signal at 5.22 ppm in the 1H
NMR spectrum supports deuterium incorporation at the meso
position of the allyl ligand through C−D activation by the
allene complex. Second, the signals at 0.68 and 2.25 ppm for
the other hydrogen atoms on the allyl ligand are singlets rather
than the doublets observed for the same reaction carried out in
hypothesis: the isomer with the SiMe3 proximal to the NO
ligand (2a) is the first hydride product observed. As heating is
continued, the distal product (2b) begins to appear in greater
quantities, concurrent with further formation of the proximal
product. Gradually, the proportion of 2b increases relative to
that of 2a. Thus, while the proximal product is the kinetic
product due to the arrangement of the ligands around the W
center, the distal isomer is the more thermodynamically stable
3
protiobenzene; this is due to loss of the JHH coupling to the
8160
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
form, possibly due to the more sterically favorable environment
resulting from placing the SiMe3 group and hydride in a cisoidal
arrangement. Isomerization from the distal back to the proximal
isomer has not been observed. The formation of 2c (the
complex containing the 1,1-disubstituted allyl ligand) is
considered later in this report.
These shifts in product distribution appear to be dictated by
sterics. Specifically, Cp*W(NO)(Npt)(η3-CH2CHCHMe) ex-
ists as two isomers in solution, whereas complexes containing
bulkier allyl ligands, e.g., 1 and Cp*W(NO)(Npt)(η3-
CH2CHCHPh), do not. Therefore, the isomerization from 2a
to 2b is probably facilitated by the small size of the hydride
ligand, which relieves steric congestion such that the necessary
haptotropic shifts for the allyl ligand to rotate about the W−C
bond become possible. The stereochemical flexibility of allyl
ligands has been implicated in “alkene flipping” mechanisms,
which interconvert substituents on metal-coordinated al-
kenes.10
Interestingly, the proximal and distal isomers can be
separated by fractional recrystallization from the same solution:
the proximal isomer crystallizes as an orange solid, while the
crystalline form of the distal isomer is yellow. Single-crystal X-
ray diffraction analyses have been carried out on both samples,
and the solid-state molecular structures of 2a,b are presented in
Figures 3 and 4, respectively.
Figure 4. Solid-state molecular structure of 2b with 50% probability
thermal ellipsoids. Selected interatomic bond distances (Å) and angles
(deg): C(11)−W(1) = 2.302(4); C(12)−W(1) = 2.296(4); C(13)−
W(1) = 2.377(4); C(11)−C(12) = 1.411(6); C(12)−C(13) =
1.404(6); N(1)−O(1) = 1.229(4); N(1)−W(1) = 1.783(3); W(1)−
H(1) = 1.55(5); C(13)−C(12)−C(11) = 120.2(4); O(1)−N(1)−
W(1) = 169.1(3).
Table 1. Comparison of the Solid-State Structural and
Solution Spectroscopic Properties of 2a,b
proximal isomer
distal isomer
2a
2b
shorter allyl C−C bond length (Å)
longer allyl C−C bond length (Å)
N−O bond length (Å)
1.384(5)
1.458(4)
1.223(3)
1557
1.404(6)
1.411(6)
1.229(4)
1547
νNO (cm−1
λmax (nm)
)
244, 296, 355
44.7
240, 280, 343
61.7
CHSiMe3, 13C NMR chem shift (ppm)
meso carbon, 13C NMR chem shift
(ppm)
106.4
110.0
CHPh, 13C NMR chem shift (ppm)
86.7
72.6
structure. The signal due to the meso proton of 2a is a doublet
of doublets having the appearance of a triplet, with a 3JHH value
of 12.4 Hz to both non-meso protons. Conversely, the
resonance of the meso proton of 2b is a doublet of doublets,
3
with JHH values of 14.6 and 10.1 Hz to the CHSiMe3 and
CHPh protons, respectively. The general magnitudes of these
Figure 3. Solid-state molecular structure of 2a with 50% probability
thermal ellipsoids. Selected interatomic bond distances (Å) and angles
(deg): C(11)−W(1) = 2.420(3); C(12)−W(1) = 2.323(3); C(13)−
W(1) = 2.263(3); C(11)−C(12) = 1.384(5); C(12)−C(13) =
1.458(4); N(1)−O(1) = 1.223(3); N(1)−W(1) = 1.779(2); W(1)−
H(1) = 1.63(4); C(11)−C(12)−C(13) = 119.1(3); O(1)−N(1)−
W(1) = 169.4(2).
3
3JHH values are consistent with typical values for Jtrans
.
However, why the meso proton in one isomer should have
the same coupling constant to both allyl protons, while the
other should show distinct coupling constants, is unclear. The
meso hydrogen in 2b might be expected to have the same
coupling to the other allyl hydrogens, which appear to be in
similar chemical environments on the basis of the similarity in
C−C bond lengths and 13C chemical shifts. However, 2b is the
isomer in which the coupling constants are different. Analysis of
the torsion angles in the solid-state structure of 2a does not
resolve this conundrum either. In this isomer, the torsion angle
from the meso proton to CHSiMe3 is 163.3° and the torsion
angle to CHPh is 171.2°, and yet the coupling constants to
these two protons are identical.11
The solid-state molecular structures of the isomers display
some interesting structural features, which are compared in
Table 1.
The σ−π distortion of the allyl ligand exists only in the
kinetic product, 2a. Its C−C bond lengths are 1.384 and 1.458
Å, while in the thermodynamic product (2b) they are 1.411 and
1.404 Å. This feature is likely a manifestation of the different
electronic effects of the SiMe3 and Ph substituents and their
response to the electronic asymmetry arising from the presence
of the NO ligand.
Formation of Cp*W(NO)(H)(η3-CH2CHC(SiMe3)Ph). The
presence of Cp*W(NO)(H)(η3-CH2CHC(SiMe3)Ph) (2c) in
the final product mixture is somewhat surprising, given that 1,1-
disubstituted allyl products have only been observed previously
when the substituent on the allyl ligand is the less sterically
The coupling constants between the NMR signals arising
from the protons on the allyl ligands also reflect a difference in
8161
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
bulky methyl group.7 The formation of 2c is likely assisted by
conformational changes in the allyl ligand during the initial C−
H activation, but since the formation of this product requires
more shifts in the allyl ligand, its rate of formation is slower
than for the 1,3-disubstituted isomers. On the basis of the plots
in Figure 2, the proportion of this product in the reaction
mixture remains constant once the majority of the starting
material has been consumed. While 2c appears to undergo no
significant changes under these conditions, thermolysis of
protio-2c in C6D6 leads to deuterium incorporation (vide
infra). The temperature of the reaction appears to have only a
slight effect on the amount of 2c produced in comparison to
the 1,3-disubsituted isomers; at 75 °C roughly 15% of the
product mixture contains the 1,1-disubstituted allyl ligand,
while at 65 and 55 °C the proportion is closer to 20%.
Monitoring the Thermal Behavior of 1 in C6D6. When it
is thermolyzed in C6 D6 , Cp*W(NO)(Npt)(η3 -
CH2CHCHSiMe3) (1) is consumed according to first-order
kinetics, which is consistent with the rate-determining step in
C−H activation being unimolecular (Scheme 1). The rate
constant for this process has been measured at several different
temperatures, and it increases with increasing temperature
(Table 2).
contrast, Cp*W(NO)(Npt)(η3-CH2CHCHMe) is unstable at
ambient temperature.
This experimental evidence (provided as Supporting
Information) supports the presumed mechanism for C−H
activation by 1, depicted in Scheme 3 . Specifically, the initial
(and probably rate-determining) step is the loss of neopentane
from the metal’s coordination sphere via hydrogen abstraction
by the neopentyl ligand from the allyl ligand to form the η2-
allene intermediate. This is followed by oxidative addition of a
substrate C−H bond onto the metal center. The putative
hydrido ligand thus formed can then migrate onto the central
carbon of the allene in a reversal of the first step to regenerate
the allyl ligand.
Reactivity of Cp*W(NO)(η3-Me3SiCHCHCHPh)(H).
(a). With PMe3. Complexes 2a,b have been thermolyzed in
the presence of Lewis bases to ascertain if any 16e
intermediates, e.g., those formed by the intramolecular loss of
dihydrogen, can be trapped. Thermolysis of a mixture of 2a and
2b in neat PMe3 produces a final reaction mixture that consists
of five organometallic products containing W−P bonds: this is
evident from the 31P{1H} NMR spectrum, which exhibits five
signals with similar chemical shifts (within 14 ppm of one
another). Each of these signals also has the customary 183W
satellites confirming a W−P linkage (Table 3). The
Table 2. Rate Constants for the Loss of 1 during
Thermolysis in C6D6
Table 3. 31P NMR Spectroscopic Data for the Product
Mixture Arising from Thermolysis of 2a,b in PMe3
temp (°C)
rate constant (s−1
)
chem shift (ppm)
1JWP (Hz)
45.5
55.3
65.0
75.1
(5.57 0.01) × 10−6
(2.81 0.01) × 10−5
(9.93 0.07) × 10−5
(3.11 0.04) × 10−4
−7.2
−9.3
−11.0
−14.4
−21.5
373
376
369
357
453
An Eyring plot of this data affords values for the enthalpy
(ΔH⧧) and entropy (ΔS⧧) of activation of +122.6 kJ mol−1 and
+39.4 J K−1 mol−1, respectively. Finally, an Arrhenius plot
enables calculation of the activation energy (Ea) for the process
to be 125.4 kJ mol−1. This is significantly higher than the 79.1
kJ mol−1 reported for Cp*W(NO)(Npt)(η3-CH2CHCHMe),7
a fact consistent with the room-temperature stability of 1. In
coordination of phosphorus to W is also confirmed by
comparing the 1H and 1H{31P} NMR spectra, in which
2
doublets in the former arising from JPH collapse to singlets
upon decoupling. Unfortunately, column chromatography is
ineffective in separating the organometallic products, and
recrystallization by slow evaporation of a pentane solution
Scheme 3
8162
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
Figure 5. Possible products arising from the thermolysis of 2a,b in PMe3.
produces only starting material, a testament to the apparent
instability of the adducts formed. Mass spectrometric data,
however, are consistent with simple adduct formation with
PMe3 (m/z 615). Possible formulations for 18e complexes
consistent with these data are the η1-allyl PMe3 adduct or the
η2-olefin PMe3 adduct (through hydride migration onto the
allyl ligand), and their possible structural isomers (Figure 5).
(b). With Pyridine. Thermolysis of a mixture of 2a and 2b in
the presence of pyridine at 55−95 °C for 5 days also presents
an interesting case. Mass spectrometric and NMR spectroscopic
data suggest the formation of an η2-olefin pyridine adduct,
whose identity is presumed to be Cp*W(NO)(η2-
Me3SiCH2CHCHPh)(C5H5N) (3) by analogy with Cp*Mo-
(NO)(η2-CH2CHCH2CH2tBu)(C5H5N) (Scheme 4).12
Scheme 5
Scheme 4
signal. It thus appears that this position on the allyl ligand has
also become partially deuterated.
After reaction for 2 days at 75 °C, the signal for the meso
proton of 2b-d1 (Figure 6C) integrates to 0.5 H against the
Cp* Me resonance at 1.65 ppm. However, the signals due to
the non-meso protons (Figure 6A,B) on the allyl ligand still
integrate to one hydrogen, but their appearance has changed.
Instead of the simple doublets observed in the spectrum of 2b,
they both appear to be overlapping with a singlet having the
same chemical shift. The singlet is the result of deuterium
The hydride migrates exclusively onto the SiMe3-substituted
end of the allyl ligand rather than the Ph-substituted end, as
1
evidenced by a correlation in the H−13C HMBC spectrum
between a methine signal at 2.55 ppm and the signal of the ipso
C of the phenyl group at 150.0 ppm. In addition, monitoring by
1H NMR spectroscopy indicates that adduct formation only
occurs once 2a (the kinetic product) has isomerized to 2b (the
thermodynamic product). This accounts for the regioselectivity
of the reaction, since the hydride in 2b is cisoid to the SiMe3
group. This reaction is formally a reductive coupling of the allyl
and hydride ligands to produce an olefin; such couplings
require that the two ligands in question have a cis orientation.
The preference for the hydride to migrate onto the SiMe3-
substituted end of the allyl ligand is contrary to what might be
expected on the basis of sterics, since the SiMe3 group is more
sterically demanding than the Ph group.13 In this case, the
metrical parameters of the solid-state molecular structures can
provide some insight: the W−C bond length of the allyl carbon
closest to the hydride is shorter in 2b than in 2a (2.377 vs 2.420
Å, respectively). In other words, the allyl ligand in 2b is
therefore better situated for hydride migration.
3
incorporation at the meso position, leading to the loss of JHH
coupling, similar to the behavior exhibited during the synthesis
of 2a,b from 1 (vide supra).
The hydride signal at −1.11 ppm also merits mention, since
it integrates to one full hydrogen against the Cp* signal. The
retention of this signal implies that the hydride is not involved
in the incorporation of deuterium on the allyl ligand. Most
notably, it precludes H2 elimination from the metal’s
coordination sphere via abstraction of the allyl meso hydrogen
by the hydride. Furthermore, there is no evidence for the
1
evolution of either H2 or HD in the H NMR spectra of these
sealed-tube reaction mixtures.
Mass spectrometric data of the final reaction mixture (Figure
7B) suggest that 2 has reacted further and has incorporated
various numbers of deuterium atoms into the allyl ligand. This
can be seen by comparing the results of thermolyzing 2b in
neat pyridine (Figure 7A) with those obtained when 2b is
thermolyzed in pyridine and deuterobenzene (Figure 7B). The
intensity pattern of the two clusters of signals from the latter
reaction at m/z 619 and 540 (corresponding to monodeu-
terated 3 and monodeuterated 2, respectively) is broad and
does not show the usual W isotope pattern. However, the
cluster at m/z 428 corresponding to the Cp*W(NO) fragment
(c). With Deuterobenzene and Pyridine. When the same
reaction is carried out in deuterobenzene with 10 equiv of
pyridine, the 1H NMR spectroscopic data suggest that the meso
hydrogen of the allyl ligand is exchanged for a deuterium atom
before the pyridine trapping occurs, forming 2b-d1 (Scheme 5
1
and Figure 6). Analysis of the H NMR spectrum of the final
product mixture, however, shows that the methine signal α to
the Ph group integrates to less than 1H relative to the Cp*
8163
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
Figure 6. Expansion (−1.5 to +6.0 ppm) of the 1H NMR spectrum (300 MHz) of the mixture formed by the thermolysis of 2 in C6D6 with 10 equiv
of pyridine, demonstrating deuterium incorporation at the meso carbon of the allyl ligand. Inset: expansions of the signals due to the protons (A)
CHSiMe3, (B) CHPh, and (C) meso, on the allyl ligand, showing partial deuteration at the meso position. Signals marked by blue italics are due to
the formation of 3-d.
The retention of the hydride is a surprising feature of this
reaction, since H/D exchange on transition-metal complexes is
usually mediated by a metal−hydride linkage.14 This results in
replacement of the hydride by deuterium, in contrast to what
has been observed with these allyl complexes. However, in
1999, Jones et al. reported that CpRe(PPh3)2H2 catalyzes H/D
exchange without deuterium incorporation into the hydride
ligands.15 They proposed that the active intermediate in this
transformation is (η3-C5H7)Re(PPh3)2, in which both hydride
ligands have migrated onto the Cp ligand. In principle, a similar
mechanism could be operative in our system, but hydride
migration onto the Cp* ring would open a coordination slot on
the W center, and such an intermediate would probably be
trapped by the Lewis base present in the reaction mixture.
An alternative mechanism could involve hydride abstraction
from the meso position on the allyl ligand to form a dihydride
η2-allene intermediate (Scheme 6). One of the W−H bonds
could then undergo σ-bond metathesis with a C−D bond of
C6D6 to generate C6D5H and a W−D bond. This deuteride
could then migrate back to the meso position of the allyl ligand,
forming the observed 2b-d1 complex. The reaction’s selectivity
would then be the result of the steric bulk of the allene
substituents preventing the original hydride from reacting with
the solvent. The intervening allene ligand also prevents
reductive elimination of H2 from the metal’s coordination
sphere. A similar η3-allyl to η2-allene transformation has been
previously invoked for Cp*Ir(allyl)(alkyl) complexes.16
Figure 7. Comparison of the M+ peaks for the products of the
reactions of 2 in (A) neat pyridine, (B) pyridine and C6D6, and (C)
neat C6D6. The M+ peak of 2 occurs at m/z 539 (184W).
does show the expected W isotope pattern, implying that
deuterium is not incorporated into the Cp* Me groups.
Although first observed in the presence of a Lewis base, the
H/D exchange also occurs when 2 is thermolyzed in neat
deuterobenzene, forming 2b-d1. In this case, however, only
deuteration at a single sitethe meso positionis observed
and is corroborated by both MS (Figure 7C) and NMR data.
This deuteration of the allyl ligand is reversible; thermolysis of
2b-d1 in C6H6 regenerates 2b. Because this exchange happens
even in the presence of a Lewis base, it is unlikely that the
mechanism of this transformation involves an open coordina-
tion slot on the metal center. Additionally, there is neither
NMR nor MS evidence for the incorporation of a C6D5 group
into the metal complex; therefore, any mechanistic speculation
must account for the fact that this fragment does not remain in
the metal’s coordination sphere.
Finally, it is worth noting that the small amount of 2c present
in these reactions also incorporates deuterium into its allyl
8164
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
Scheme 6
ligand at the meso position to give 2c-d. Unfortunately, it is not
clear whether the hydride ligand persists in this complex, since
1
the region around 0.00 ppm where this H NMR signal is
expected to appear is obscured by other signals.
Figure 8. Solid-state molecular structure of 4 with 50% probability
thermal ellipsoids. Selected interatomic distances (Å) and angles
(deg): C(11)−W(1) = 2.345(4); C(12)−W(1) = 2.301(3); C(13)−
W(1) = 2.411(3); C(11)−C(12) = 1.391(6); C(12)−C(13) =
1.389(5); C(17)−W(1) = 2.216(3); N(1)−W(1) = 1.781(3);
N(1)−O(1) = 1.217(3); C(13)−C(12)−C(11) = 121.3(3); O(1)−
N(1)−W(1) = 166.2(3).
Reactions of Cp*W(NO)(Npt)(η3-CH2CHCHSiMe3) with
Fluorobenzenes. (a). Pentafluorobenzene. Consistent with
Cp*W(NO)(Ph)(η3-CH2CHCHSiMe3) being the initial prod-
uct formed during the reaction of 1 with benzene (Scheme 2),
the C−H bond of pentafluorobenzene is activated by
thermolysis of 1 in neat C6F5H at 55 °C for 2 days (Scheme
7), affording Cp*W(NO)(C6F5)(η3-CH2CHCHSiMe3) (4) as
the sole product.
distinct signals in the 19F{1H} spectrum of 4 indicates that the
C6F5 ligand is undergoing minimal bond rotation about its W−
C axis. If this were not the case, then only three signals should
be observed in the spectrum, one for each set of chemically
equivalent fluorine atoms.
Scheme 7
1
Monitoring of the thermolysis of 4 in C6D6 by H NMR
spectroscopy reveals a second set of signals growing into the
spectrum with chemical shifts and coupling constants similar to
those of 4. This observation is more consistent with 4
undergoing partial isomerization by changing the mode of
attachment of its allyl ligand (e.g., exo to endo) rather than by
effecting intramolecular migration of the C6F5 group to form a
1,3-disubstituted allyl ligand. Continued thermolysis eventually
leads to a thermodynamic distribution of approximately 3:1
between the major (4) and minor isomers, but there is no
evidence for activation of the solvent C−D bonds.
The lack of aryl migration onto the allyl ligand might be
expected due to the undistorted allyl ligand in 4 being more
electron rich and therefore less prone to nucleophilic attack
than σ−π distorted allyl ligands. This view is corroborated by
comparing the 13C chemical shifts of the signals due to the allyl
CH2 carbons of the product of pentafluorobenzene activation,
4, with the parent neopentyl complex, 1. For 1 this value is 84.6
ppm, while for 4 it is 56.1 ppm. These observations are
consistent with a better shielded carbon atom in 4 that likely
also has a greater degree of sp3 character than the analogous
atom in 1. In addition, the electronegativity of the fluorine
atoms on the C6F5 group also reduces the nucleophilic
character of the migrating group, and the presence of ortho
fluorine substituents results in a relatively strong W−C bond
(vide infra). Both factors further inhibit the C6F5 ligand’s
propensity to migrate.10
No C−F bond activation is observed, since no signals due to
aromatic protons appear in the 1H NMR spectrum of 4.
Further evidence for exclusive C−H activation is afforded by
the 19F{1H} NMR spectrum, which shows five highly complex
signals assignable to Caryl−F bonds, with no signals attributable
to either W−F17 or Callyl−F. The solid-state molecular structure
of 4 is presented in Figure 8.
The presence of the pentafluorophenyl ligand has a
pronounced effect on the allyl ligand in 4. The C−C bond
lengths are equal within the limits of experimental error (1.391
and 1.389 Å); thus, it is no longer σ−π distorted, especially in
comparison to the parent neopentyl complex 1. Therefore,
replacing the Npt ligand with the C6F5 ligand removes the
electronic asymmetry at the metal center. Both the NO and the
C6F5 ligands withdraw electron density from the metal center,
but via different mechanisms: NO through π back-bonding and
C6F5 through the inductive effect of the F atoms. The
trimethylsilylallyl ligand in 4a exists in an exo,anti bonding
mode18 with the SiMe3 group distal to the NO ligand rather
than proximal, as observed in the solid-state molecular structure
of 1.
Comparison of the IR νNO stretching frequencies of 1 and 4
reveals that 1 has a slightly higher νNO stretching frequency
(1589 cm−1) than 4 (1581 cm−1), implying that more W−NO
back-bonding occurs in 4. Furthermore, the presence of five
If migration of the aryl group onto the ligand in 4 is
prevented by the low nucleophilicity of the pentafluorophenyl
group and the lack of σ−π distortion in the allyl ligand, then a
8165
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
less-substituted fluorobenzene should mediate these effects
somewhat. The first such fluoroarene investigated was p-
difluorobenzene.
(b). p-Difluorobenzene. Since all the C−H bonds in p-
difluorobenzene are chemically equivalent, the thermolysis of 1
in this solvent might be expected to lead to a single product
resulting from simple activation of a C−H bond. Instead, three
1
organometallic products are detectable by H NMR spectros-
copy of the final reaction mixture after thermolysis and workup
of 1 at 55 °C in neat p-C6H4F2 (Figure 9). However, there are
1
no signals in the upfield region of the H NMR spectrum;
therefore, the formation of hydrides can be excluded from the
possible reaction pathways.
Figure 10. Solid-state molecular structure of one of the isomers of
complex 5 with 50% probability thermal ellipsoids. Selected
interatomic distances (Å) and angles (deg): C(11)−W(1) =
2.400(4); C(12)−W(1) = 2.314(3); C(13)−W(1) = 2.343(3);
C(11)−C(12) = 1.381(5); C(12)−C(13) = 1.422(5); C(17)−W(1)
= 2.207(3); N(1)−O(1) = 1.226(3); N(1)−W(1) = 1.777(3);
C(11)−C(12)−C(13) = 122.1(3); O(1)−N(1)−W(1) = 167.3(3).
Figure 9. Expansion from +4.1 to +5.1 ppm of the 1H NMR spectrum
(300 MHz, C6D6) of the products resulting from the thermolysis of 1
in p-difluorobenzene.
asymmetry at the metal center. This conclusion can be
confirmed by comparing the 13C chemical shifts of the
resonances due to the allyl CH2 carbons in 1, 4, and 5: the
value for 5 (80.8 ppm) more closely resembles that of 1 (84.6
ppm) than that of 4 (56.1 ppm).
Each meso proton signal is split into a doublet of doublets of
doublets, implying coupling of the meso proton to three other
protons in the same spin system. The allyl ligands must
therefore still be monosubstituted. Given the splitting pattern
of each signal, as well as the similarities in the chemical shifts
and the coupling constants of these signals, this appears to be a
straightforward case of allyl isomerization between the various
endo/exo and syn/anti combinations. Consequently, the
reaction of 1 with p-difluorobenzene proceeds as depicted in
Scheme 8 to produce 5, the complex resulting from activation
of a single C−H bond.
Since the σ−π distortion in the allyl ligand appears to be one
factor favoring intramolecular migration of the aryl ligand, it
might be expected that 5 could form the corresponding hydride
under thermolytic conditions. However, thermolysis of a
sample of 5 enriched in a single isomer in C6D6 leads to its
partial conversion to one of the isomers found after
recrystallization. After prolonged heating at 60 °C, the system
appears to reach equilibrium, with the two products present in
about a 5:4 ratio. Hydride formation resulting from migration
of the fluoroaryl ligand onto the allyl ligand is not observed.
Thus, while a distorted allyl ligand may be a necessary
condition for aryl migration, it alone is not sufficient; the aryl
group itself must also have sufficient nucleophilic character, and
the W−C bond must be prone to cleave.
Scheme 8
1
(c). o-Difluorobenzene. The H NMR spectrum of the final
reaction mixture arising from the reaction of 1 with o-
difluorobenzene is quite complex, but nevertheless shows the
presence of allyl meso proton signals corresponding to both
mono- and disubstituted allyl ligands, which can be
distinguished on the basis of their splitting patterns (Table
A single-crystal X-ray diffraction analysis of one of the
isomers of 5 has revealed its solid-state molecular structure to
be as shown in Figure 10.
1
The particular isomer shown in Figure 10 again contains the
allyl ligand in the exo,anti configuration, similar to the case for
4. The allyl ligand in 5 also displays the σ−π distortion
(C(11)−C(12) = 1.381; C(12)−C(13) = 1.422 Å) character-
istic of these complexes; in fact, the C−C bond lengths are very
similar to those in the starting material, 1. Despite the
electronegativity of the fluorine atoms, they appear to have only
a minor effect on the structure of 5. Therefore, decreasing the
number of fluorine atoms on the fluoroaryl ligand relative to
that in 4 has succeeded in restoring some of the electronic
4). In addition, the upfield region of the H NMR spectrum
contains two signals attributable to tungsten hydride species.
By analogy to the reaction of 1 with benzene (vide supra), it
is reasonable to expect that the two hydrides are the result of
aryl migration onto the allyl ligand, since the signals due to the
meso protons in these products only show coupling to two
other protons. In addition, if the reactivity is truly analogous,
then these two products are likely isomers differing in the
orientations of their allyl ligands. The products which are not
hydrides contain monosubstituted allyl ligands, as evidenced by
8166
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
1
Table 4. H NMR Chemical Shifts and Coupling Constants
of Selected Signals Exhibited by the Mixture of Products
Arising from the Thermolysis of 1 in o-Difluorobenzene
δCp*
(ppm)
δhydride
(ppm)
δmeso (ppm),
multiplicity
meso JHH
(Hz)
isomers of 6
isomers of 7
1.59
1.69
1.57
−1.13
−1.24
n/a
4.99, dd
5.13, t
14.8, 9.8
12.3
4.38, ddd
15.6, 10.6,
7.9
1.65
n/a
4.15, ddd
13.9, 11.4,
7.3
the ddd splitting pattern of the meso resonances. Taken
together, these features suggest that the C−H activation of o-
difluorobenzene proceeds via the two pathways depicted in
Scheme 9.
This hypothesis is further supported by the solid-state
molecular structure of one of the hydride products (6), which
has been isolated by recrystallization from pentane.
Figure 11. Solid-state molecular structure of 6 with 50% probability
ellipsoids. F2 is disordered over C21 and C19 in approximately a 3:1
ratio. Selected interatomic distances (Å) and angles (deg): C(11)−
W(1) = 2.292(5); C(12)−W(1) = 2.307(5); C(13)−W(1) =
2.372(5); C(11)−C(12) = 1.414(7); C(12)−C(13) = 1.399(7);
W(1)−H(1W) = 1.51(4); N(1)−O(1) = 1.238(5); N(1)−W(1) =
1.774(4); C(13)−C(12)−C(11) = 118.9(5); O(1)−N(1)−W(1) =
170.5(4).
In this structure (Figure 11), the SiMe3 group is distal to the
NO ligand, and the complex has metrical parameters similar to
those of the analogous complex 2b. In particular, the C−C
bond lengths in the allyl ligand are nearly equal. It seems then
that the presence of fluorine atoms at the 3- and 4-positions on
the aryl ring in 6 does not have a significant effect on the
structure of the disubstituted allyl ligand in comparison to 2b.
One might also infer that their effect on the nucleophilicity of
the aryl ligand in the presumed fluoroaryl intermediate is
decreased, since migration onto the allyl ligand does occur in
this case. Compound 6 probably arises via the first reaction
sequence in Scheme 8. First, C−H activation at position 4 on o-
difluorobenzene is effected; then, in a manner that parallels the
reactivity of the phenyl ligand (vide supra), the fluoroaryl group
migrates onto the allyl ligand.
As for its unfluorinated counterpart, 2b, the thermolysis of 6
in C6D6 leads to the incorporation of deuterium into the meso
position of the allyl ligand while the hydride ligand remains
intact. A brief comparison of the extent of deuterium
incorporation during these two reactions, as measured by the
integration of the signal due to the meso proton against that of
the Cp* signal, reveals that the incorporation of deuterium into
6 occurs more quickly than for 2b.
the 1H NMR signals due to the products which are not
hydrides are similar to those of the two major products
resulting from the thermolysis of 1 in p-difluorobenzene (vide
supra). It thus seems probable that these signals arise from
complexes resulting from the activation of a C−H bond
adjacent to a fluorine atom on the benzene ring (Scheme 8).
Just as in the case of p-difluorobenzene, migration onto the allyl
ligand does not occur when the activated C−H bond is ortho to
a fluorine atom. As before, the separate meso proton signals are
probably due to the different possible coordination modes of
the allyl ligands in these compounds. By comparison of the
relative integrations for signals due to the meso protons in each
of these complexes, it is clear that the favored reaction pathway
is activation at position 4 on o-difluorobenzene; only about 30%
of the complexes in the final product mixture contain a
monosubstituted allyl ligand. Here again sterics likely play a
role in determining the regioselectivity of the reaction, since a
similar preference has been observed for [TpOs-
(CH2CH2PiPr3)(η2-CH2CH2)2][BF4] (Tp = hydridotris-
(pyrazolyl)borate).19 In fact, for all substrates tested to date,
The inequivalence of the C−H bonds in o-difluorobenzene
clearly is a key factor in accounting for the observed product
distribution. Both the chemical shifts and coupling constants of
Scheme 9
8167
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
other chemicals were ordered from commercial suppliers and used as
received.
complex 1 has shown a preference for C−H activation at the
most sterically accessible site.5
All IR samples were prepared as Nujol mulls sandwiched between
NaCl plates, and their spectra were recorded on a Thermo Nicolet
Model 4700 FT-IR spectrometer. NMR spectra were recorded at room
temperature (except where noted) on Bruker AV-300, AV-400, and
AV-600 instruments, and all chemical shifts and coupling constants are
reported in ppm and in Hz, respectively. 1H NMR spectra were
referenced to the residual C6D5H present in C6D6 (7.16 ppm). 13C
NMR spectra were referenced to C6D6 (128.4 ppm). 19F NMR spectra
were externally referenced to CFCl3. 31P NMR spectra were externally
Activation at position 4 on o-difluorobenzene, farthest from
the fluorine atoms, is contrary to some of our previous research,
during which a preference for activation at the o-F position was
documented during the activation of fluorobenzene by
Cp*W(NO)(Npt)2.20 Indeed, in most cases of C−H activation
of fluorobenzenes, the C−H bond ortho to a C−F bond is the
favored site for activation.21−23 This preference has been
attributed to a thermodynamic stabilizing effect by the o-F,
since the resulting M−C bond is strengthened more by the
ortho C−F bond than the C−H bond in the substrate.24
Interestingly, Harman’s investigation of the activation of
fluoroarenes with TpW(NO)(PMe3) has shown that both C−
H and C−F bond activation can occur, with C−F activation
predominating in the case of mono- and disubstituted
fluorobenzenes.17 DFT calculations have revealed that, even
though the barriers to C−H and C−F activation for his system
are similar, the thermodynamic stability of a resultant tungsten
fluoride complex such as TpW(NO)(PMe3)(F)(Ph) leads to it
being the sole product isolated from the reaction. In our system
there is no evidence that C−F activation is a possible mode of
reactivity for 1; indeed, it is C−H activation that seems to be
irreversible, particularly in the case of 6.
referenced to 85% H3PO4. When necessary, H−1H COSY, H−13C
HSQC, and 1H−13C HMBC experiments were carried out to correlate
and assign 1H and 13C NMR signals. UV/vis spectra were recorded on
an Agilent Cary 5000 spectrophotometer in single-beam mode using a
quartz cuvette with a 1.0 cm path length. Low- and high-resolution
mass spectra (EI, 70 eV) were recorded by Mr. Marshall Lapawa of the
UBC mass spectrometry facility using a Kratos MS-50 spectrometer.
Elemental analyses were performed by Mr. Derek Smith of the UBC
microanalytical facility. X-ray crystallographic data collection, solution,
and refinement were performed at the UBC X-ray crystallography
facility.
1
1
Preparation of Cp*W(NO)(H)(η3-Me3SiCHCHCHPh) (2).
Cp*W(NO)(Npt)(η3-CH2CHCHSiMe3) (1; 146 mg, 0.274 mmol)
was dissolved in C6H6 (ca. 15 mL) to produce an orange solution. The
reaction mixture was heated at 55 °C for 17 h. The solvent was then
removed in vacuo, and the solid recrystallized from pentane contained
2a−c in approximately a 2:2:1 ratio, as judged by 1H NMR
spectroscopy (84 mg, 41% yield). 2a,b were individually crystallized
from a minimal amount of pentane to give yellow and orange crystals,
respectively. 2c was identified from NMR spectroscopic data.
Characterization Data for 2a. IR (cm−1): 1557 (s, νNO), 1933 (m,
EPILOGUE
■
In summary, we have demonstrated that the thermolysis of
Cp*W(NO)(Npt)(η3-CH2CHCHSiMe3) (1) in benzene at 55
°C generates three isomeric products having the composition
Cp*W(NO)(H)(η3-Me3SiCHCHCHPh) (2). These are iso-
mers of Cp*W(NO)(Ph)(η3-Me3SiCHCHCH2), the com-
pound presumed to be formed initially, and they result from
an intramolecular Ph/allyl H exchange. Interestingly, thermol-
ysis of 2 in the presence of pyridine produces the η2-olefin
pyridine adduct Cp*W(NO)(η2-Me3SiCH2CHCHPh)-
(C5H5N). We have also shown that the activation of
fluorobenzenes initiated by Cp*W(NO)(Npt)(η3-
CH2CHCHSiMe3) occurs exclusively via C−H bond activa-
tions. Selectivity for the activation of these C−H bonds appears
to be determined by steric factors. Intramolecular migration of
the newly formed fluoroaryl ligands onto the allyl ligands does
not occur when there is a fluorine atom in the position ortho to
the newly formed W−C bond, probably because this linkage is
stronger than a W−aryl bond.
1
1
νWH). H NMR (300 MHz, C6D6): δ −0.97 (s, 1H, JWH = 119, W−
H), 0.13 (d, 1H, 3JHH = 13.2, CHSiMe3), 0.41 (s, 9H, SiMe3), 1.71 (s,
15H, C5Me5), 2.80 (d, 1H, JHH = 12.5, CHPh), 5.34 (dd, 1H, JHH
3
3
=
3
12.4, 12.4, allyl meso), 7.00 (m, 1H, para CH), 7.12 (obscured t, JHH
= 7.8, meta CH), 7.28 (d, 2H, 3JHH = 7.3, ortho CH). 13C NMR (100
MHz, C6D6): δ 2.1 (SiMe3), 11.1 (C5Me5), 44.7 (CHSiMe3), 86.7
(CHPh), 104.3 (C5Me5), 106.4 (allyl meso), 126.7 (para CH), 127.0
(meta CH), 129.1 (ortho CH), 142.7 (ipso C). MS (LREI, m/z, probe
temperature 120 °C): 539 [M+, 184W]. Anal. Calcd for C22H33NOSiW:
C, 48.98; H, 6.17; N, 2.60. Found: C, 49.00; H, 6.24; N, 2.53.
Characterization Data for 2b. IR (cm−1): 1547 (s, νNO), 1926 (m,
1
1
νWH). H NMR (400 MHz, C6D6): δ −1.10 (s, 1H, JWH = 122, W−
H), 0.32 (s, 9H, SiMe3), 0.63 (d, 1H, 3JHH = 14.8, allyl CHSiMe3), 1.65
(s, 15H, C5Me5), 2.20 (d, 1H, 3JHH = 10.1, allyl CHPh), 5.22 (dd, 1H,
3JHH = 14.6, 10.1, allyl meso), 7.08 (m, 1H, para CH), 7.28 (t, 2H, 3JHH
= 7.7, meta CH), 7.49 (d, 2H, JHH = 7.6, ortho CH). 13C NMR (75
3
MHz, C6D6) δ 0.7 (SiMe3), 10.7 (C5Me5), 61.7 (allyl CHSiMe3), 72.6
(allyl CHPh), 105.0 (C5Me5), 110.0 (allyl meso), 125.9 (para CH),
127.4 (ortho CH), 129.0 (meta CH), 142.9 (ipso C). MS (LREI, m/z,
probe temperature 120 °C): 539 [M+, 184W]. MS (HREI, m/z, 182 W):
calcd 537.181 37, found 537.181 79.
EXPERIMENTAL SECTION
■
General Methods. Except where noted, all reactions and
subsequent manipulations involving organometallic reagents were
performed under anhydrous and anaerobic conditions, either on
double-manifold Schlenk lines or in Innovative Technologies Lab-
Master 100 and MS-130 BG dual-station gloveboxes equipped with
freezers maintained at −30 °C. Purification of dinitrogen and argon
was achieved by passing these gases first through a column containing
MnO and then a column of activated 4 Å molecular sieves. Small-scale
reactions were conducted in J. Young NMR tubes equipped with
Kontes greaseless stopcocks. Larger scale thermolyses were performed
in thick-walled glass reaction vessels equipped with Kontes greaseless
stopcocks. Pentane, diethyl ether (Et2O), benzene, and tetrahydrofur-
an (THF) were dried over sodium/benzophenone ketyl and freshly
distilled prior to use. All other solvents were dried according to
standard procedures.25 Cp*W(NO)(Npt)(η3-CH2CHCHSiMe3) was
prepared according to the published procedure utilizing penta-
methylcyclopentadiene obtained from Boulder Scientific Co.5 All
Characterization Data for 2c. 1H NMR (400 MHz, C6D6): δ 0.00
(s, 1H, 1JWH = 123, W−H), 0.29 (s, 9H, SiMe3), 1.52 (s, 15H, C5Me5),
3
2
2.66 (ddd, 1H, JHH = 8.6, J = 2.82, JHH = 2.05, allyl CH2), 3.28 (dd,
1H, 3JHH = 14.1, 2JHH = 1.47, allyl CH2), 3.84 (dd, 1H, 3JHH = 13.9, 8.7,
allyl meso), 6.94 (m, 1H, para CH), 7.11 (obscured t, 2H, 3JHH = 8.2,
meta CH), 7.57 (d, 2H, 3JHH = 6.8, ortho CH). 13C NMR (100 MHz,
C6D6): δ 2.4 (SiMe3), 10.5 (C5Me5), 50.4 (allyl CH2), 84.7
(C(SiMe3)Ph), 105.3 (C5Me5), 105.5 (allyl meso), 125.6 (para CH),
128.2 (meta CH), 133.4 (ortho CH), 147.62 (ipso C).
Monitoring the Conversion of 1 to 2a−c. Complex 1 (13 mg,
0.024 mmol) was dissolved in C6D6 (ca. 0.7 mL) and transferred to a J.
Young NMR tube. The NMR probe was heated to 75.1 °C, and the 1H
NMR spectrum of the reaction was recorded every 5 min for 10 h. The
area under the signals at 1.79 (2a, C5Me5), 1.72 (2b, C5Me5), 1.69 (2c,
C5Me5), and 1.23 ppm (1, Npt Me) was integrated against the area of
the signal at 7.16 ppm (C6D5H), which was referenced to 1.
8168
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
Thermolysis of 2 in PMe3. Complex 2 (46 mg, 0.084 mmol) was
thermolyzed in neat PMe3 at 55 °C for 24 h. The volatiles were then
splitting resulting from 19F. 19F{1H} NMR (282 MHz, C6D6) δ
−165.5, −162.3, −158.8, −105.7, −105.5. MS (LREI, m/z, probe
temperature 120 °C): 629 [M+, 184W]. MS (HREI, m/z, 182 W): calcd
627.134 26, found 627.134 39.
1
removed in vacuo to leave an orange oil (47 mg). H and 31P{1H}
NMR spectra of the final reaction mixture indicated the presence of
five phosphorus-containing compounds. Separation of these com-
pounds by column chromatography was unsuccessful.
1
Spectral Data for Minor Isomer. H NMR (300 MHz, C6D6): δ
0.20 (s, 9H, SiMe3), 1.39 (s, 15H, C5Me5), 1.60 (dd, 4JHH = 2.6, 3JHH
=
Preparation of Cp*W(NO)(η2-PhCHCHCH2SiMe3)(py) (3).
After thermolysis of 2 (18 mg, 0.033 mmol) in neat pyridine at 85 °C
for 2 days, removal of solvent in vacuo afforded Cp*W(NO)(η2-
PhCHCHCH2SiMe3)(C5H5N) (3) as the sole organometallic
product (20 mg, 96% yield).
3
10.6, allyl CH2), 3.78 (dd, 1H, JHH = 8.1, 2.5, CHSiMe3), 4.10 (dd,
2
3
3JHH = 15.3, JHH = 1.5, allyl CH2), 4.41 (ddd, JHH = 15.2, 10.6, 7.9,
allyl meso).
Preparation of Cp*W(NO)(2,5-F2C6H3)(η3-CH2CHCHSiMe3)
(5). Complex 1 (107 mg, 0.201 mmol) was dissolved in p-
difluorobenzene (ca. 4 mL) to produce an orange solution. The
reaction mixture was heated at 55 °C for 48 h, whereupon a color
change to yellow occurred. The solvent was then removed in vacuo to
give solid 5. Recrystallization of this solid from a pentane solution
maintained at −30 °C afforded one of the isomers of complex 5 as a
microcrystalline yellow-orange solid (87 mg, 75% yield).
Characterization Data for Isolated Isomer of 5. IR (cm−1): 1592
1
Characterization Data for 3. IR (cm−1): 1537 (s, νNO). H NMR
(300 MHz, C6D6): δ 0.21 (s, 9H, SiMe3), 1.57 (s, 15H, C5Me5), 1.93
(s, 2H, CH2SiMe3), 2.55 (td, 1H, 3JHH = 9.9, 3.7, CHPh), 3.40 (d, 1H,
3JHH = 10.0, CH), 6.17 (t, 2H, 3JHH = 6.9, pyridine CH), 6.32 (d, 2H,
3
3JHH = 7.3, aryl CH), 6.60 (t, 1H, JHH = 7.6, pyridine CH), 6.76 (m,
1H, aryl CH), 7.06 (t, 2H, 3JHH = 7.5, aryl CH), 8.53 (m, 2H, pyridine
CH). 13C NMR (75 MHz, C6D6): δ −0.47 (SiMe3), 9.8 (C5Me5), 29.7
(CH2), 52.7 (CHPh), 62.3 (CH), 105.0 (C5Me5), 121.6 (aryl CH),
123.2 (pyridine CH), 124.4 (aryl CH), 128.3 (aryl CH), 135.4
(pyridine CH), 150.0 (ipso C), 153.4 (pyridine CH). MS (LREI, m/z,
probe temperature 120 °C): 618 [M+, 184W]. MS (HREI, m/z, 184 W):
calcd 618.226 30, found 618.226 83. MS (HREI, m/z, 182W): calcd
616.223 57, found 616.224 00.
1
(s, vNO). H NMR (300 MHz, C6D6): δ 0.19 (s, 9H, SiMe3), 1.43 (s,
15H, Cp* Me), 1.70 (dd, 1H, 3JHH = 10.7, 4JHH = 2.5, allyl CHSiMe3),
3.91 (dd, 1H, 3JHH = 15.6, 4JHH = 2.2, allyl CH2), 4.05 (dd, 1H, 3JHH
=
4
3
8.1, JHH = 2.2, allyl CH2), 4.41 (ddd, 1H, JHH = 15.5, 10.8, 8.2, allyl
meso), 6.63 (m, 2H, aryl CH), 8.12 (dt, 1H, J = 9.1, 2.9, aryl CH). 13C
NMR (75 MHz, C6D6): δ 2.1 (SiMe3), 9.6 (C5Me5), 51.1 (allyl
CHSiMe3), 80.7 (allyl CH2), 107.3 (C5Me5). Due to splitting by F and
isomerization of the complex at room temperature, satisfactory spectra
for assigning the 13C resonances on the fluoroaryl ligand could not be
Thermolysis of 2 in C6D6. A mixture of the isomers of 2 (6 mg,
0.01 mmol) was dissolved in C6D6 (ca. 0.7 mL) and transferred to a J.
1
Young NMR tube. The sample was heated at 75 °C, and H NMR
obtained. MS (LREI, m/z, probe temperature 120 °C): 575 [M+,
spectra were taken periodically to assess the extent of reaction. 2a,b
were converted exclusively to 2b-d1. 2c was converted to 2c-d, a
partially deuterated product.
182
184W]. MS (HREI, m/z,
W): calcd 573.162 52, found 573.162 79.
Thermolysis of 5. A C6D6 solution of 5 in a J. Young NMR tube
was heated at 50 and 60 °C for 24 h. The progress of the isomerization
1
Characterization Data for 2b-d1. H NMR (300 MHz, C6D6): δ
1
−1.11 (s, 1H, JWH = 121, W−H), 0.32 (s, 9H, SiMe3), 0.63 (s, 1H,
1
was monitored by H NMR spectroscopy.
allyl CHSiMe3), 1.65 (s, 15H, C5Me5), 2.20 (s, 1H, allyl CHPh), 7.08
1
Spectral Data for Other Isomer. H NMR (300 MHz, C6D6): δ
3
(m, 1H, para CH), 7.28 (t, 2H, JHH = 7.7, meta CH), 7.49 (d, 2H,
3
−0.03 (s, 9H, SiMe3), 1.38 (s, 15H, Cp* Me), 2.38 (ddd, 1H, JHH
=
3JHH = 7.6, ortho CH). MS (LREI, m/z, probe temperature 120 °C):
540 [M+, 184W].
2
4
3
7.3, JHH = 2.6, JHH = 1.1, allyl CH2), 3.13 (d, 1H, JHH = 14.1, allyl
3
2
CHSiMe3), 3.51 (dd, JHH = 11.4, JHH = 2.6, allyl CH2), 4.17 (ddd,
1H, 3JHH = 13.9, 11.4, 7.4, allyl meso), 6.66 (m, 2H, aryl CH), 8.04 (dt,
1H, J = 9.2, 3.4, aryl CH).
Characterization Data for 2c-d. 1H NMR (300 MHz, C6D6): 0.29
(s, 9H, SiMe3), 1.52 (s, 15H, C5Me5), 2.66 (s, 1H, allyl CH2), 3.28 (s,
1H, allyl CH2).
Thermolysis of 1 in o-Difluorobenzene. Complex 1 (110 mg,
0.206 mmol) was thermolyzed in o-difluorobenzene for 2 days at 55
°C. The final reaction mixture contained products with both
monosubstituted (i.e., 6) and disubstituted (i.e., 7) allyl ligands in
Thermolysis of 2 in C6D6 and Pyridine. A sample of 2a,b (6 mg,
0.01 mmol) and pyridine was dissolved in C6D6 (ca. 0.7 mL) and
transferred to a J. Young NMR tube. The number of equivalents of
pyridine was determined by integration of the upfield hydride signals
at −0.98 and −1.11 ppm against the signal due to the hydrogens at the
2- and 6-positions of pyridine at 8.53 ppm.26 The sample was heated in
approximately
a
3:7 ratio. Cp*W(NO)(η3-(3,4-F2C6H3)-
CHCHCHSiMe3)(H), (6; 31 mg, 26% yield) was obtained by
recrystallization of the final reaction residue from a pentane solution
maintained at −30 °C. The isomeric complex Cp*W(NO)(η3-
CH2CHCHSiMe3)(2,3-F2C6H3) (7) remained in solution.
1
10 °C increments from 55 to 95 °C for 24 h at a time, and H NMR
spectra were acquired periodically to assess the extent of reaction.
Removal of the volatiles in vacuo from the final mixture yielded 3-d, a
partially deuterated product.
Characterization Data for 6. IR (cm−1): 1596 (s, νNO), 1928 (w,
1
1
Characterization Data for 3-d. 1H NMR (300 MHz, C6D6): δ 0.21
(s, 9H, SiMe3), 1.57 (s, 15H, C5Me5), 1.93 (m, 2H, CH2SiMe3), 2.55
(m, <1H, CHPh), 3.40 (s, <1H, CH), 6.17 (m, 2H, pyridine CH), 6.32
νWH). H NMR (300 MHz, C6D6): δ −1.12 (s, 1H, JWH = 120, W−
H), 0.30 (s, 9H, SiMe3), 0.62 (d, 1H, 3JHH = 14.7, CHSiMe3), 1.59 (s,
3
15H, C5Me5), 1.90 (d, 1H, JHH = 9.7, CH(3,4-F2C6H3)), 5.00 (dd,
1H, 3JHH = 14.7, 9.7, allyl meso), 6.84 (dd, 1H, 3JHF = 18.5, 3JHH = 8.5,
aryl CH). 6.96 (m, 1H, aryl CH), 7.18 (obscured, 1H, aryl CH). 13C
NMR (75 MHz, C6D6): δ 0.6 (SiMe3), 10.6 (C5Me5), 62.8
(CHSiMe3), 69.8 (CH(3,4-F2C6H3)), 105.1 (C5Me5), 109.8 (allyl
3
3
4
(d, 2H, JHH = 7.3, aryl CH), 6.58 (tt, 1H, JHH = 7.6, JHH = 1.5,
3
4
pyridine CH), 6.76 (tt, JHH = 7.2, JHH = 1.2, 1H, aryl CH), 7.06 (t,
2H, JHH = 7.5, aryl CH), 8.53 (br m, 2H, pyridine CH).
3
Preparation of Cp*W(NO)(C6F5)(η3-CH2CHCHSiMe3) (4). Com-
plex 1 (115 mg, 0.216 mmol) was dissolved in pentafluorobenzene (ca.
5 mL) to produce an orange solution. The reaction mixture was heated
at 55 °C for 48 h. Subsequently, the solvent was removed in vacuo to
give a brown oil. Slow diffusion of pentane into an Et2O solution of
this oil at −30 °C afforded 4 as a yellow microcrystalline solid (86 mg,
63% yield).
2
2
meso), 115.8 (aryl CH, JCF = 17.9), 117.7 (aryl CH, JCF = 16.2),
122.8 (aryl CH, 3JCF = 5.7), 140.5 (ipso C, 3JCF = 6.3, 4JCF = 4.0), 148.5
(aryl CF, 1JCF = 153.7, 2JCF = 12.7), 151.6 (aryl CF, 1JCF = 155.5, 2JCF
=
12.7). 19F{1H} NMR (282 MHz, C6D6): δ −143.0 (3JFF = 22.6),
−138.3 (3JFF = 21.1). MS (LREI, m/z, probe temperature 120 °C):
575 [M+, 184W]. MS (HREI, m/z, 182W): calcd 573.162 52, found
573.162 32.
Characterization Data for 4. IR (cm−1): 1507 (m, C6F5), 1581 (s,
1
1
νNO). H NMR (300 MHz, C6D6): δ −0.06 (s, 9H, SiMe3), 1.34 (s,
Selected H NMR Signals for the Minor Isomer of 6 (300 MHz,
1
3
4
3
C6D6): δ −1.24 (s, 1H, JWH = 115, W−H), 0.10, (d, 1H, JHH = 12.1,
allyl CHSiMe3), 0.38 (s, 9H, SiMe3), 1.69 (s, 15H, C5Me5), 2.50 (d,
1H, 3JHH = 12.6, allyl CH(3,4-F2C6H3)), 5.13 (t, 1H, 3JHH = 12.0, allyl
meso).
15H, C5Me5), 2.23 (ddd, 1H, JHH = 2.5, JHH = 7.3, 1.3, allyl CH2),
3
3
2.87 (d, 1H, JHH = 13.5, allyl CH2), 3.53 (dd, 1H, JHH = 11.4, 2.6,
CHSiMe3), 4.22 (ddd, 1H, JHH = 13.6, 11.3, 7.3, allyl meso). 13C
3
NMR (75 MHz, C6D6) δ 1.9 (SiMe3), 9.8 (C5Me5), 56.1 (allyl CH2),
72.8 (CHSiMe3), 107.5 (C5Me5), 109.8 (allyl meso). Signals due to
the carbon atoms on the C6F5 ring were not observed due to the
Selected 1H NMR Signals for 7 (300 MHz, C6D6): isomer 1, 1.57 (s,
15H, C5Me5), 4.38 (ddd, 1H, 3JHH = 15.6, 10.6, 7.9, allyl meso); isomer
8169
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
Table 5. X-ray Crystallographic Data for Complexes 2a,b and 4−6
2a
2b
4
5
6
Crystal Data
empirical formula
cryst habit, color
cryst size (mm)
cryst syst
C22H33NOSiW
prism, orange
0.20 × 0.10 × 0.09
monoclinic
C2/c
C22H33NOSiW
plate, yellow
0.09 × 0.02 × 0.02
monoclinic
P21/n
C22H28F5NOSiW
needle, yellow
0.40 × 0.10 × 0.04
monoclinic
P21/n
C22H31F2NOSiW
prism, yellow-orange
0.11 × 0.06 × 0.05
monoclinic
P21/c
C22H31F2NOSiW
prism, yellow
0.11 × 0.06 × 0.05
monoclinic
P21/n
space group
V (Å3)
4402.0(4)
29.8232(18)
10.1727(6)
15.8503(9)
90
2243.7(3)
8.4708(7)
13.6779(11)
19.5480(18)
90
2306.14(19)
8.6761(4)
16.4535(8)
16.2203(8)
90
2235.5(2)
11.4956(7)
11.1957(6)
17.5205(8)
90
2313.3(3)
8.5346(7)
13.7027(11)
19.8752(16)
90
a (Å)
b (Å)
c (Å)
α (deg)
β (deg)
113.733(2)
90
97.850(5)
90
95.150(2)
90
97.530(4)
90
95.591(4)
90
γ (deg)
Z
8
4
4
4
4
calcd density Mg/m3)
1.628
1.597
1.813
1.710
1.652
abs coeff (mm−1
)
5.313
5.212
5.115
5.249
5.073
F000
2144
1072
1232
1136
1136
Data Collection and Refinement
no. of measd rflns: total
40 913
5358
5358
21 710
6380
29 440
7395
20 920
5101
no. of measd rflns: unique 5076
a
final R Indices
R1 = 0.0158, wR2 =
0.0458
R1 = 0.0310, wR2 =
0.0636
R1 = 0.0262, wR2 =
0.0562
R1 = 0.0323, wR2 =
0.0508
R1 = 0.0337, wR2 =
0.0745
b
goodness of fit on F2
1.353
1.535 and −0.516
1.069
1.041
0.997
1.030
largest diff peak, hole (e−
2.082 and −1.502
2.007 and −1.434
1.156 and −1.394
3.003 and −0.958
Å−3
)
a
b
2
2
R1 on F = ∑|(|Fo| − |Fc|)|/∑|Fo|; wR2 = [(∑(Fo2 − Fc2)2)/∑w(Fo )2]1/2; w = [σ2Fo ]−1. GOF = [∑(w(|Fo| − |Fc|)2)/(degrees of freedom)]1/2
.
3
2, 1.65 (s, 15H, C5Me5), 4.15 (ddd, 1H, JHH = 13.9, 11.4, 7.3, allyl
meso).
included in fixed positions. The final cycle of full-matrix least-squares
analysis was based on 7395 observed reflections and 273 variable
parameters.
X-ray Crystallography. Data collection was carried out at −173
1 °C on a Bruker X8 diffractometer, using graphite-monochromated
Mo Kα radiation.
Data for 6 were collected to a maximum 2θ value of 54.3° in 0.5°
oscillations with 60 s exposure times. The structure was solved by
direct methods27 and expanded using Fourier techniques. All non-
hydrogen atoms were refined anisotropically, and hydrogens H1w,
H11, H12, and H13 were refined isotropically. The thermal parameter
of H1w was constrained to be 1.2 times that of the tungsten atom. All
other hydrogen atoms were included in fixed positions. Fluorine F2
was disordered over two sites on the aryl ring in approximately a 74:26
ratio. The final cycle of full-matrix least-squares analysis was based on
5101 observed reflections and 286 variable parameters.
Data for 2a were collected to a maximum 2θ value of 55.1° in 0.5°
oscillations with 5 s exposure times. The structure was solved by direct
methods27 and expanded using Fourier techniques. All non-hydrogen
atoms were refined anisotropically, and hydrogens H1, H11, H12, and
H13 were refined isotropically. All other hydrogen atoms were
included in fixed positions. The final cycle of full-matrix least-squares
analysis was based on 5076 observed reflections and 259 variable
parameters.
For each structure, neutral-atom scattering factors were taken from
Cromer and Waber.31 Anomalous dispersion effects were included in
Fc;32 the values for Δf′ and Δf′′ were those of Creagh and McAuley.33
The values for mass attenuation coefficients are those of Creagh and
Hubbell.34 All calculations were performed using SHELXL-9735 via the
WinGX interface.36 X-ray crystallographic data for all five structures
are presented in Table 5.
Data for 2b were collected to a maximum 2θ value of 55.9° in 0.5°
oscillations with 60 s exposure times. The structure was solved by
direct methods27 and expanded using Fourier techniques. The
structure was a two-component twin; the two components were
related by a 2.2° rotation about the (0.654, 0.077, 1.000) reciprocal
axis. The twin components were separated using Cell_Now,28
SAINTPLUS,29 and TWINABS.30 Refinement was performed on a
deconvoluted HKLF4 data set. All non-hydrogen atoms were refined
anisotropically, and hydrogen H1 was refined isotropically. All other
hydrogen atoms were included in fixed positions. The final cycle of
full-matrix least-squares analysis was based on 5358 observed
reflections and 248 variable parameters.
Data for 4 were collected to a maximum 2θ value of 59.6° in 0.5°
oscillations. The structure was solved by direct methods27 and
expanded using Fourier techniques. All non-hydrogen atoms were
refined anisotropically, and hydrogens H11a and H11b were refined
isotropically. All other hydrogen atoms were included in fixed
positions. The final cycle of full-matrix least-squares analysis was
based on 6380 observed reflections and 296 variable parameters.
Data for 5 were collected to a maximum 2θ value of 62.9° in 0.5°
oscillations with 5 s exposure times. The structure was solved by direct
methods27 and expanded using Fourier techniques. All non-hydrogen
atoms were refined anisotropically, and hydrogens H11a, H11b, and
H13 were refined isotropically. All other hydrogen atoms were
ASSOCIATED CONTENT
■
S
* Supporting Information
Text, figures, and tables giving details of the experimental
studies of the thermal behavior of 1 in C6D6 and CIF files
providing full details of the crystallographic analyses of
complexes 2a,b and 4−6. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: legzdins@chem.ubc.ca.
Notes
The authors declare no competing financial interest.
8170
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171

Organometallics
Article
(25) Armarego, W. L. F.; Chai, C. L. L. Purification of Laboratory
Chemicals, 5th ed.; Elsevier: Amsterdam, 2003.
(26) Fulmer, G. R.; Miller, A. J. M.; Sherden, N. H.; Gottleib, H. E.;
Nudelman, A.; Stoltz, B. M.; Bercaw, J. E.; Goldberg, K. J.
Organometallics 2010, 29, 2176.
ACKNOWLEDGMENTS
■
We are grateful to The Dow Chemical Co. for continuing
financial support of this work and to our Dow colleagues, Drs.
Devon C. Rosenfeld and Brandon A. Rodriguez, for assistance
and helpful discussions. We also thank Mr. Rhett A. Baillie for
technical assistance.
(27) SIR-92: Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi,
A. J. Appl. Crystallogr. 1993, 26, 343.
(28) Cell_Now 2008/2; Bruker AXS Inc., Madison, WI.
(29) SAINTPLUS, Version 7.60A; Bruker AXS Inc., Madison,WI,
1997−2008.
(30) TWINABS, V2008/2; Bruker AXS Inc., Madison, WI, 2007.
(31) Cromer, D. T.; Waber, J. T. International Tables for X-ray
Crystallography; Kynoch Press: Birmingham, U.K., 1974; Vol. IV,
Table 2.2 A.
(32) Ibers, J. A.; Hamilton, W. C. Acta Crystallogr. 1964, 17, 781.
(33) Creagh, D. C.; McAuley, W. J. International Tables for X-ray
Crystallography; Kluwer Academic: Boston, 1992; Vol. C., Table
4.2.6.8.
(34) Creagh, D. C.; Hubbell, J. H. International Tables for X-ray
Crystallography; Kluwer Academic: Boston, 1992; Vol. C., Table
4.2.4.3.
REFERENCES
■
(1) See, for example: (a) Crabtree, R. H. Dalton Trans. 2001, 17,
2437. (b) Labinger, J. A.; Bercaw, J. E. Nature 2002, 417, 507.
(c) Goldman, A., Goldberg, K. I., Eds. Activation and Functionalization
of C−H Bonds; American Chemical Society: Washington, DC. 2004.
(d) Fekl, U.; Goldberg, K. I. Adv. Inorg. Chem. 2003, 54, 259.
(e) Hartwig, J. F. Organotransition Metal Chemistry: From Bonding to
Catalysis; University Science Books: Sausalito, CA, 2010; Chapter 18.
(2) For recent examples of the selective functionalization of C−H
bonds, see: Crabtree, R. H. Chem. Rev. 2010, 110, 575 and the reviews
in this special issue.
(3) Tsang, J. Y.K.; Buschhaus, M. S. A.; Fujita-Takayama, C.; Patrick,
B. O.; Legzdins, P. Organometallics 2008, 27, 1634.
(4) Baillie, R. A.; Tran, T.; Lalonde, K. M.; Tsang, J. Y. K.; Thibault,
M. E.; Patrick, B. O.; Legzdins, P. Organometallics 2012, 31, 1055 and
references cited therein.
(35) SHELXL-97: Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112.
(36) WinGX, V1.70: Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837.
(5) Chow, C.; Patrick, B. O.; Legzdins, P. Organometallics 2012, 31,
DOI: 10.1021/om300742y.
(6) Ng, S. H. K.; Adams, C. S.; Hayton, T. W.; Legzdins, P.; Patrick,
B. O. J. Am. Chem. Soc. 2003, 125, 15210.
(7) Baillie, R. A.; Tran, T.; Thibault, M. E.; Legzdins, P. J. Am. Chem.
Soc. 2010, 132, 15160.
(8) Dougherty, D. A.; Mislow, K.; Blount, J. F.; Wooten, J. B.;
Jacobus, J. J. Am. Chem. Soc. 1977, 99, 6149.
(9) Baillie, R. A.; Man, R. W. Y.; Shree, M. V.; Chow, C.; Thibault,
M. E.; McNeil, W. S.; Legzdins, P. Organometallics 2011, 30, 6201.
(10) Stoebenau, E. J., III; Jordan, R. F. J. Am. Chem. Soc. 2006, 128,
8638.
(11) A meaningful analysis of the torsion angles in 2b was not
possible, since the hydrogen atoms in question were included in
calculated positions rather than located from the difference map.
(12) Tran, T.; Chow, C.; Zimmerman, A. C.; Thibault, M. E.;
McNeil, W. S.; Legzdins, P. Organometallics 2011, 30, 738.
(13) Branchadell, V.; Moreno-Manas, M.; Pleixats, R. Organometallics
2002, 21, 2407.
(14) Atzrodt, J.; Derdau, V.; Fey, T.; Zimmermann, J. Angew. Chem.,
Int. Ed. 2007, 46, 7744.
(15) Jones, W. D.; Rosini, G. P.; Maguire, J. A. Organometallics 1999,
18, 1754.
(16) McGhee, W. D.; Bergman, R. G. J. Am. Chem. Soc. 1988, 110,
4246.
(17) Liu, W.; Welch, K.; Trindle, C. O.; Sabat, M.; Myers, W. H.;
Harman, W. D. Organometallics 2007, 26, 2589.
(18) The nomenclature outlined by Faller et al. for allyl complexes of
this type is used throughout; see: Faller, J. W.; Chodosh, D. F.;
Katahira, D. J. Organomet. Chem. 1980, 187, 227.
(19) Bajo, S.; Esteruelas, M. A.; Lopez, A. M.; Onate, E.
Organometallics 2011, 30, 5710.
(20) Tsang, J. Y. K.; Buschhaus, M. S. A.; Legzdins, P.; Patrick, B. O.
Organometallics 2006, 25, 4215.
(21) Lafrance, M.; Rowley, C. N.; Woo, T. K.; Fagnou, K. J. Am.
Chem. Soc. 2006, 128, 8754.
(22) Nakao, Y.; Kashihara, N.; Kanyiva, K. S.; Hiyama, T. J. Am.
Chem. Soc. 2008, 130, 16170.
(23) Selmeczy, A. D.; Jones, W. D.; Partridge, M. G.; Perutz, R. N.
Organometallics 1994, 13, 522.
(24) (a) Clot, E.; Megret, C.; Eisenstein, O.; Perutz, R. N. J. Am.
Chem. Soc. 2009, 131, 7817. (b) Evans, M. E.; Burke, C. L.; Yaibuathes,
S.; Clot, E.; Eisenstein, O.; Jones, W. D. J. Am. Chem. Soc. 2009, 131,
13464. (c) Tanabe, T.; Brennessel, W. W.; Clot, E.; Eisenstein, O.;
Jones, W. D. Dalton Trans. 2010, 10495.
8171
dx.doi.org/10.1021/om300765e | Organometallics 2012, 31, 8159−8171