Intermolecular methyl group exchange and reversible P-Me bond cleavage at cobalt(III) dimethyl halide species

Article abstract of DOI:10.1021/om3009537

The cobalt(III) dimethyl halide complexes cis,mer-(PMe₃) ₃Co(CH₃)₂X (X = Cl, I) were found to undergo a degenerate cobalt-to-cobalt transfer of the methyl ligands during isotopic labeling experiments. Extensive mechanistic studies exclude radical, methyl iodide elimination, and disproportionation/comproportionation pathways for exchange of the methyl groups between metals. A related cobalt(III) dimethyl complex supported by the tridentate phosphine ligand MeP(CH₂CH ₂PMe₂)₂ showed dramatically slower methyl ligand transfer, indicative of a mechanism for intermetallic exchange with a requisite phosphine dissociation. Crossover experiments between cobalt(III) dimethyl halide complexes supported by PMe₃ and MeP(CH ₂CH₂PMe₂)₂ are consistent with a dicobalt transition structure in which only one cobalt center requires phosphine dissociation prior to methyl transfer. An additional methyl group scrambling process between cis,mer-(PMe₃)₃Co(CH₃) ₂I and free PMe₃ was also identified during the investigation and originates from reversible P-CH₃ bond cleavage.

Full text of DOI:10.1021/om3009537

Article
pubs.acs.org/Organometallics
Intermolecular Methyl Group Exchange and Reversible P−Me Bond
Cleavage at Cobalt(III) Dimethyl Halide Species
Hongwei Xu, Paul G. Williard, and Wesley H. Bernskoetter*
Department of Chemistry, Brown University, Providence, Rhode Island 02912, United States
*
S Supporting Information
ABSTRACT: The cobalt(III) dimethyl halide complexes
cis,mer-(PMe ) Co(CH ) X (X = Cl, I) were found to undergo
3
3
3 2
a degenerate cobalt-to-cobalt transfer of the methyl ligands
during isotopic labeling experiments. Extensive mechanistic
studies exclude radical, methyl iodide elimination, and
disproportionation/comproportionation pathways for ex-
change of the methyl groups between metals. A related
cobalt(III) dimethyl complex supported by the tridentate phosphine ligand MeP(CH CH PMe ) showed dramatically slower
2
2
2 2
methyl ligand transfer, indicative of a mechanism for intermetallic exchange with a requisite phosphine dissociation. Crossover
experiments between cobalt(III) dimethyl halide complexes supported by PMe and MeP(CH CH PMe ) are consistent with a
3
2
2
2 2
dicobalt transition structure in which only one cobalt center requires phosphine dissociation prior to methyl transfer. An
additional methyl group scrambling process between cis,mer-(PMe ) Co(CH ) I and free PMe was also identified during the
3
3
3 2
3
investigation and originates from reversible P−CH bond cleavage.
3
INTRODUCTION
RESULTS AND DISCUSSION
■
■
The transfer of alkyl ligands between metal centers has been
observed in a variety of transition metal complexes and may
occur via a wide range of potential pathways depending on the
coordination, steric, and electronic characteristics of a particular
A diverse range of potential pathways for the intermolecular
methyl group exchange observed at 1-Me may be proposed on
2
the basis of common transformations for first-row transition
4
metals, related exchanges with heavier transition metal
1
1c−e,2
species. The intermolecular exchange of alkyl groups from one
species,
and previous examinations of methyl ligand
1a,f
metal to another has drawn interest for its relationship to
synthetic techniques using main group alkylating agents, the
biosynthesis of acetyl coenzyme A, alkyl/aryl scrambling during
palladium-catalyzed cross-coupling, and more recently as a
proposed intermediate step for oxidative methane oligomeriza-
transfer at cobalt.
Guided by these studies, a series of
general mechanistic routes for cobalt−methyl exchange were
considered, including: a radical crossover mechanism (path A),
a methyl iodide elimination route (path B), a cobalt(I)-
catalyzed disproportionation pathway (path C), a bimetallic
exchange derived from iodide loss (path D), and a bimetallic
exchange derived from phosphine loss (path E). These
mechanisms for cobalt-to-cobalt methyl group transfer in 1-
1
,2
tion methods. Metal-to-metal transfer of methyl groups
recently garnered the attention of our laboratory over the
course of investigations into the mechanism for C−C bond
3
formation at strong field cobalt centers. During these studies
the cobalt(III) complex cis,mer-(PMe ) Co(CH ) I (1-Me )
was found to scramble isotopicially labeled methyl ligands
between cobalt centers prior to selective reductive carbon−
carbon bond elimination (Figure 1). Similar intermolecular
methyl ligand exchange processes have also been reported for
Me were then probed by experimental methods to affirm or
discount the probability of each reaction pathway.
2
3
3
3 2
2
Path A: Radical Crossover. The relatively low bond
dissociation enthalpy of cobalt−methyl bonds has been
5
demonstrated in multiple studies and frequently gives rise to
6
the extrusion of methyl radicals. This feature accounts for
1
,2
several other late transition metal organometallic species.
Preliminary experiments with 1-Me suggested that methyl
much of the reactivity of the cobalamine complex in vitamin B12
and could also account for the intermolecular exchange of the
2
7
methyl ligands in 1-Me (Figure 2; path A). In fact, it was
group exchange persisted even under conditions where C−C
reductive elimination was extremely slow, leading to a
hypothesis that the two processes may not be intimately linked
2
during isotopic crossover experiments to probe for an outer-
sphere methyl radical mechanism in reductive C−C bond
3
elimination that our laboratory first observed intermolecular
in mechanism. Herein we report the results of a more
3
methyl transfer in 1-Me . Upon heating a mixture of 1-d Me
complete experimental examination of the mechanism for
cobalt-to-cobalt methyl group transfer at 1-Me , the observa-
tion of an unusual phosphorus−methyl bond scission reaction
at cobalt, and the synthesis and characterization of a closely
related cobalt(III) species, cis,mer-(PMe ) Co(CH ) (SO CF )
2
3
2
isotopomers to 50 °C (the cobalt methyl ligands undergo rapid
intramolecular site exchange at ambient temperature), sub-
2
Received: October 11, 2012
Published: January 22, 2013
3
3
3 2
3
3
(
2-Me2).
©
2013 American Chemical Society
798
dx.doi.org/10.1021/om3009537 | Organometallics 2013, 32, 798−806

Organometallics
Article
Figure 1. Intermolecular methyl ligand scrambling and ethane reductive elimination from 1-d Me .
3
2
Figure 2. Methyl radical extrusion pathway for methyl group exchange.
•
5
stantial quantities of the crossover products, CD CD and
abstraction by CH with a rate constant on the order of 10
3
3
3
−1 −1
CH CH , were observed (Figure 1). However, the deviation
M
s
(27 °C), and thus should make a reasonable probe for
the intermediacy of outer-sphere methyl radicals.
3
3
9
,10
from a statistically expected mixture of isotopologue products
1:2:1 neglecting secondary isotope effects) and their changing
Signifi-
(
cantly, no substantial quantities of methane were detected by
NMR spectroscopy during the trapping experiment (<4%) over
two days at ambient temperature, and the presence of CHD did
not alter the relative half-life of intermolecular methyl exchange
compared to parallel control samples without the methyl radical
probe. These observations are taken to be inconsistent with an
outer-sphere methyl radical crossover pathway (path A).
Path B: Methyl Iodide Elimination. The reductive
elimination of carbon−halide bonds from transition metals
has drawn increasing interest as the microscopic reverse of C−
X bond oxidative addition in catalytic cross-coupling reactions
ratio over the course of the reaction both hinted that a standard
radical or bimetallic elimination pathway was not manifesting in
3
C−C bond formation. Careful examination of the crossover
experiment at low conversion to ethane revealed intermolecular
scrambling of the cobalt−methyl ligands in the reactant prior to
elimination, evident from a slight isotopic perturbation of the
chemical shifts of the Co−CH signals. In subsequent studies,
3
this intermolecular methyl exchange has also been detected in
samples of 1-d Me in benzene-d at ambient temperature over
3
2
6
8
the course of a day (Figure 3; scrambling after 24 h at 23 °C).
11
as well as for its own utility in catalytic applications. Notably,
several examples of methyl iodide reductive elimination from
heavy metal complexes related to 1-Me have been reported.
1
2
2
A process invoking a similar iodomethane elimination from the
dimethyl iodide cobalt(III) species could result in intermo-
lecular scrambling of the cobalt−methyl sites (path B; Figure
4
). In this case CH I eliminated from one metal center would
3
react with the organometallic product of a second methyl
iodide elimination to afford the net exchange. Such methyl
iodide elimination reactions could presumably occur via either
−
sequential I loss followed by nucleophilic attack on Co−CH
3
or concerted C−I bond formation.
The reversible methyl iodide elimination pathway was first
explored by attempting to observe incorporation of isotopically
labeled CD I into 1-Me at ambient temperature over
1
31
Figure 3. Partial H{ P} NMR spectrum showing isotopic
perturbation of chemical shifts between the Co−CH of 1-Me and
3
2
3
2
1
-d Me at 10 °C in benzene-d .
approximately 24 h (conditions where intermolecular methyl
group exchange is easily detected by NMR spectroscopy).
3
2
6
In order to investigate the intermediacy of methyl radicals in
Unfortunately, addition of 0.5 to 2 equiv of CD I to benzene
3
the scrambling of cobalt−alkyl groups in 1-d Me , the exchange
solutions of 1-Me resulted in degradation of the organo-
3
2
2
reaction was monitored in the presence of excess (0.25 M) 1,4-
cyclohexadiene (CHD). This diene undergoes H atom
metallic species within 30 min, prior to observation of any
13
isotopic exchange. As a substitute for methyl iodide, a related
Figure 4. Reversible iodomethane elimination pathway for methyl group exchange.
7
99
dx.doi.org/10.1021/om3009537 | Organometallics 2013, 32, 798−806

Organometallics
Article
experiment was performed by treating 1-d Me with 1 equiv of
methyl trifluoromethanesulfonate (MeOTf) (eq 1). This
Table 1. Comparison of Selected Bond Distances (Å) and
6
2
Angles (deg) for cis,mer-(PMe ) Co(CH ) X (X = OTf, I,
3
3
3 2
CN)
a
b
OTf
I
CN
Co(1)−C(1)
Co(1)−C(5)
Co(1)−P(1)
Co(1)−P(2)
Co(1)−P(3)
P(1)−Co(1)−P(3)
C(1)−Co(1)−P(2)
2.038(3)
2.008(3)
2.238(2)
2.292(2)
2.234(2)
163.51(4)
168.25(9)
175.3(1)
2.033(10)
2.038(7)
2.249(2)
2.263(3)
2.237(2)
166.45(10)
167.9(3)
172.9(4)
2.062(5)
2.054(5)
2.207(2)
2.229(5)
2.205(1)
162.07(5)
167.47(11)
179.0(2)
reaction resulted in exchange of the iodide and triflate anions,
affording cis,mer-(PMe ) Co(CD ) OTf (2-d Me ) and free
3
3
3 2
6
2
iodomethane over one hour. Significantly, no isotopic exchange
of the methyl groups was observed in the iodomethane or
C(5)−Co(1)−X
1
2
a
b
cobalt product by H or H NMR spectroscopy. The cobalt(III)
dimethyl triflate complex may alternatively be prepared by
addition of MeOTf to (PMe ) CoCH (eq 2), with complete
Reference 16. Reference 17.
3
4
3
2-Me demonstrates the weaker trans influence of the triflate
2
ligand compared to iodide and cyanide.
More significant to mechanistic considerations, the absence
of isotopic crossover upon addition of MeOTf to 1-d Me
6
2
coupled with the successful preparation of 2-Me2 from
PMe ) CoCH indicates that path B is likely not operative
(
3
4
3
for methyl group exchange. As depicted in Figure 4, methyl
iodide reductive elimination from the cobalt(III) dimethyl
species would generate (PMe ) CoCH , which must be formed
3
3
3
conversion occurring within two hours. The cobalt dimethyl
triflate species was characterized in its perprotio isotopologue
in low concentrations, as it remains undetected during the
experiment and is reported to be unstable in the absence of
(
2-Me ) by variable-temperature NMR spectroscopy, combus-
20
2
additional phosphine. The alternative synthesis of 2-Me (eq
14
2
tion analysis, and single-crystal X-ray diffraction. At ambient
temperature the H and P NMR spectra of 2-Me exhibit
2
) from methyl triflate and (PMe ) CoCH occurs much faster
1
31
3 4 3
2
than bimetallic methyl group exchange and likely proceeds via
reaction of MeOTf with transiently generated, four-coordinate
broad resonances indicative of a dynamic molecular process
15
close to the time scale of the NMR experiment. Cooling the
(
PMe ) CoCH . Under these conditions the concentration of
3 3 3
sample to −40 °C in toluene-d afforded sharp spectral features,
8
MeOTf is high relative to (PMe ) CoCH and should thus
3
3
3
including two Co−CH resonances at 0.72 and −1.65 ppm in
3
provide a rapid trap for the unobserved (PMe ) CoCH to
1
3 3
3
the H NMR spectrum as well as a doublet and triplet at 9.09
and −5.95 ppm in the P NMR spectrum. The molecular
structure of 2-Me was confirmed by X-ray diffraction and is
yield significant isotopic incorporation into the dimethyl cobalt
triflate product if path B were operable. In addition, reactions
between the more stable four-coordinate cobalt(I) species
3
1
2
presented in Figure 5. A comparison of selected metrical
(
PMe ) CoI (1) and methyl iodide produced a mixture of
3 3
parameters for 2-Me and analogues bearing iodide and cyanide
2
products including 1-Me , (PMe ) CoI , and insoluble cobalt
2
3 3
2
ligands has been compiled in Table 1. The octahedral geometry
and structural features of the three congeners are quite similar;
however, the modestly shortened Co(1)−C(5) bond length for
material. Finally, the likelihood of a second-order process
requiring two cobalt intermediates of very low concentration
also augers poorly for path B.
Path C: Cobalt(I)-Catalyzed Disproportionation. The
reversible disproportionation of transition metal−methyl
ligands has recently been highlighted in palladium systems as
a possible intermediate step for methane homologation and
2
oxidatively induced C−C bond elimination. Analogous
disproportionation events from 1-Me could account for the
2
intermolecular methyl exchange (eq 3), but would require
III
2
(PMe ) Co (CH ) I
3
3
3 2
II
IV
⇄
(PMe ) Co (CH )I + (PMe ) Co (CH ) I
3
3
3
3 2
3 3
+
PMe3
(3)
formation of a rather uncommon organocobalt(IV) species and
18
likely necessitate the loss of phosphine to avoid hypervalency.
This equilibrium formation was deemed unlikely for 1-Me2
given the unusual oxidation state for organocobalt and an
absence of accelerated scrambling at higher cobalt concen-
trations (vida infra). Alternatively, a cobalt(I)-catalyzed
disproportionation−comproportionation sequence could affect
the same net exchange of methyl ligands via the more common
Figure 5. Molecular structure of 2-Me with 30% ellipsoids. Hydrogen
atoms have been removed for clarity.
2
8
00
dx.doi.org/10.1021/om3009537 | Organometallics 2013, 32, 798−806

Organometallics
Article
Figure 6. Cobalt(I)-catalyzed pathway for methyl group exchange.
Figure 7. Bimetallic iodide dissociation pathway for methyl group exchange.
Figure 8. Bimetallic phosphine dissociation pathway for methyl group exchange.
1
9
3
+, 2+, and 1+ oxidation states of cobalt (Figure 6; path C).
d Me acting as a nucleophile in a concerted interchange
3
2
22
In path C, a trace quantity of tris(trimethylphosphine)cobalt(I)
halide, formed by reductive elimination of 1-Me or persisting
substitution-type transition structure.
A direct method to examine path D is the inhibition of
2
3
,20
n
from the synthesis of 1-Me2,
d3Me2 to transiently afford two isotopologues of tris-
trimethylphosphine)cobalt(II) methyl halide. A productive
comproportionates with 1-
crossover by added iodide salts such as ( Bu N)I. Unfortu-
4
nately, this salt has only modest solubility in arene solvent, and
the use of more polar coordinating solvents could block path D
by saturating the cationic intermediate. After testing the
(
reversion of this step would disproportionate a pair of matched
cobalt(II) centers to yield 1-Me and 1-d Me as well as
regenerate the cobalt(I) catalyst. However, the viability of the
cobalt(I)-catalyzed pathway was easily discounted by inten-
tional addition of (PMe ) CoI and (PMe ) CoCl to samples of
stability of 1-Me at ambient temperature in a variety of poorly
2
2
6
2
coordinating polar solvents, chlorobenzene was selected as the
23
most suitable solvent for testing iodide inhibition. Interest-
n
ingly, addition of 5 equiv of ( Bu N)I to C D Cl solutions of 1-
4
6
5
3
3
3 3
d Me produced no detectable inhibition in the formation of 1-
3
2
1
-d Me , which afforded no observed acceleration in
3 2
Me compared to parallel samples without the iodide source
2
intermetallic exchange.
over two days at ambient temperature. This observation is
difficult to reconcile with methyl exchange occurring via path
D.
Path D: Iodide Loss−Bimetallic Exchange. The failure
to detect methyl group ejection from 1-Me as either a methyl
radical or methyl cation equivalent in earlier experiments
implies that alkyl ligand exchange most likely occurs via
formation of a dicobalt intermediate or transition structure.
Since 1-Me is an electronically saturated complex, formation of
such a bimetallic structure would probably require loss of at
least one ligand (either I or PMe ) from the coordination
sphere. While the reversible loss of PMe from 1-Me has
2
Path E: Phosphine Loss−Bimetallic Exchange. An
alternative bimetallic route for methyl ligand exchange is the
loss of PMe from 1-Me to obtain a neutral five-coordinate
3
2
2
cobalt species (Figure 8; path E). Similar to path D, a
coordinately unsaturated intermediate would exchange the
cobalt−methyl positions by forming a transient bridging alkyl
structure. An analogous mechanism has previously been
implicated in the cobalt-to-cobalt methyl group exchange at
cyclopentadienyl cobalt dimethyl phosphine complexes studied
−
3
3
2
previously been implicated as a step preceding reductive ethane
elimination, loss of the halide ligand would generate a more
electrophilic cobalt cation complex. In addition, use of polar
3
1a
by Bergman and co-workers (eq 4).
solvents with 1-Me2 and related cobalt complexes has
demonstrated the potential lability of the Co−I bond and in
some cases resulted in substitution of the iodide with the
solvent (e.g., acetonitrile). The enhanced electrophilicity of
the cation could then facilitate formation of a transient bridging
methyl structure (Figure 7; path D). In this path dissociation of
an iodide ligand would afford a 16-electron cationic
intermediate, which is trapped by a second equivalent of 1-
In this elegant work, the authors employed labeling studies,
ligand inhibition experiments, and the simulation of exper-
imental reaction kinetics to detect and study the crossover
reaction.
21
On the basis of the previous studies by Bergman and the3
prior observation of reversible PMe3 loss from 1-Me2,
8
01
dx.doi.org/10.1021/om3009537 | Organometallics 2013, 32, 798−806

Organometallics
Article
examinations of path E were in fact among the first mechanistic
experiments our laboratory performed on this system. Given
the similarities in structure and cobalt oxidation state, it seemed
reasonable to expect a resemblance between the exchange
pathways for 1-Me and CpCoMe PPh . Yet as mentioned
parallel NMR tube study was conducted. One NMR tube was
charged only with 1-d Me , while a second sample contained
6
2
an equimolar mixture of 1-d Me and 1-Me , holding the total
6
2
2
cobalt concentration the same for both samples. Over one day
at ambient temperature the first sample (which may exchange
2
2
3
briefly in an earlier report, preliminary experiments monitoring
only by PMe activation) exhibited only trace formation of 1-
3
1
the conversion of 1-d Me to 1-Me by H NMR spectroscopy
d Me , while the second tube contained more than one-third 1-
3
2
2
3
2
with 10−20 equiv of free PMe revealed no significant
d Me . This demonstrates that under identical conditions and
3
3 2
inhibition of exchange compared to parallel NMR tube samples
without added phosphine the intermetallic exchange of the
3
without added PMe . Having then proceeded to examine and
Co−CH ligands is dramatically faster than exchange from
3
3
discount a range of alternative pathways for intermolecular
methyl group exchange as described above, our mechanistic
studies appeared frustrated. However, upon repeating the PMe3
reversible P−CH cleavage.
3
The effect of free PMe on P−CH cleavage was further
3
3
explored by monitoring the conversion of 1-d Me to 1-d Me
6
2
3
2
inhibition experiments over longer time courses, a significant
and 1-Me with varying amounts of added phosphine (eq 5).
2
observation was made. Monitoring the scrambling of 1-d Me
3
2
Samples with excess PMe (4−10 equiv) produced mixtures of
3
in the presence of excess PMe over three days at ambient
3
1-d Me and 1-Me at considerably higher conversions (20−35
3
2
2
temperature appeared to produce nonequilibrium quantities of
times greater) than parallel samples without additional
1
-Me (i.e., a greater than 25% conversion to 1-Me ). This
2
2
phosphine. The dependence of reversible P−CH cleavage on
3
suggested that additional 1-Me may be formed by incorporat-
2
the concentration of free PMe suggests that the activation does
3
ing −CH groups from a source other than the cobalt-methyl
3
not occur via a classical P−C bond oxidative addition reaction
sites and that such a process could mask PMe inhibition of
3
as was observed by Wolczanski, Andersen, Wakatsuki, and
1
intermetallic exchange when solely H NMR spectroscopy was
25
others with earlier transition metals. A possible mechanism
used to quantify methyl transfer.
consistent with the limited data in hand is nucleophilic attack of
Reversible P−CH Bond Cleavage. The only significant
3
outer-sphere PMe on a Co−CH bond to reversibly form a
3
3
source for additional proton incorporation into the tris-
26
cobalt phosphorane intermediate (Figure 9). Similar metal-
lophosphorane complexes have been widely implicated in the
scrambling of ligand alkyl/aryl groups during palladium-
catalyzed cross-coupling reactions, though these typcially
form via an intramolecular nucleophilic reaction between
(
trimethylphosphine)cobalt dimethyl iodide species was
PMe , as these reactions were performed in deuterated solvents
3
under anhydrous conditions. Additionally, only d , d , and d
0
3
6
isotopologues of ethane (from reductive elimination) and 1-
Me were observed during the experiment, indicating that any
24b
2
phosphine and alkyl/aryl ligands.
scrambling process involving PMe3 must occur via P−C
cleavage as opposed a C−H bond activation. Cleavage of P−
C bonds at transition metals is fairly common for aryl-
Reexamination of Path E. The evidence of two significant
routes for conversion of 1-d Me to 1-Me in the presence of
3
2
2
excess PMe complicates the endeavor to detect inhibition of
24
3
substituted phosphines, but significantly more rare for
25
intermetallic methyl group exchange by added phosphine and
through these experiments provides evidence for the
mechanism in Figure 8 (path E). Alternative attempts to
inhibit methyl group exchange using other phosphine donors
were inconclusive, as these reactions produced either other
phosphorus−alkyl bonds.
The activation of trimethylphosphine was confirmed by
addition of 10 equiv of PMe to a sample of 1-d Me (eq 5),
3
6
2
cobalt species (e.g., PPh ) or no measurable effect for sterically
3
hindered ligands (e.g., PCy ). Additionally, attempts to
3
establish a second-order dependence on the concentration of
the cobalt complex showed little change between samples with
increasing metal concentrations. However this does not strictly
exclude path E if the initial phosphine dissociation is rate
limiting under the conditions of the experiment. Thus a new
investigation was pursued using a chelating phosphine with
which resulted in formation of 1-d Me , 1-Me , and PMe -d
3
3
2
2
3
(detected by GC-MS) at ambient temperature over two days. It
is important to consider that while exchange of a methyl
substituent between phosphine and cobalt is evident, this
process does not account for the substantial formation of 1-
d6Me2 and/or C D observed in previous crossover and
steric and electronic effects similar to PMe
A tridentate analogue of trimethylphosphine MeP-
(CH CH PMe (triphos) was prepared by the coupling of
dimethyl phosphine and divinylmethyl phosphine similar to the
.
3
)
2 2
2
6
2
2
reductive elimination experiments using 1-d Me . Instead two
3
2
27
separate pathways for cobalt−methyl isotopic exchange appear
procedure previously described by King and Cloyd. The
operative: an intermetallic exchange of the Co−CH groups
strong chelate effect of this ligand should dramatically slow the
3
and reversible P−CH bond cleavage. In order to gain
rate of cobalt-to-cobalt methyl transfer relative to PMe if path
3
3
perspective on the relative rates of these two reactions, a
E is operative. In order to test this hypothesis, the synthesis of
Figure 9. Possible pathway for methyl group exchange between PMe and 1-Me .
3
2
8
02
dx.doi.org/10.1021/om3009537 | Organometallics 2013, 32, 798−806

Organometallics
triphos)Co(CH ) I was attempted by addition of the
Article
(
Control experiments monitoring equimolar mixtures of 3-
Me and 3-d Me at ambient temperature were performed first
3
2
tridentate ligand to 1-Me , a method our laboratory has
2
2
6
2
previously used to good effect in installing bidentate ligands for
other studies. However, upon stirring the reactants in diethyl
ether for 18 h, a light yellow precipitate was formed, which has
been tentatively assigned as two isomers of the cationic
and established that the chloride-substituted PMe complex
undergoes intermetallic methyl transfer, easily observable by
3
NMR spectroscopy after 4−6 h, comparable to the time scale
29
for the iodide analogue. More significantly, parallel NMR
studies indicated that the mixtures of PMe -bound 3-Me and
[
(triphos)Co(CH ) PMe ]I complex (eq 6). This product
3 2 3
3
2
3
-d Me exchanged methyl groups dramatically faster than
6 2
samples of triphos-bound 4-Me and 4-d Me (Figure 10).
2
6
2
mixture was sparingly soluble in benzene-d and exhibited three
6
1
Co−CH resonances in the H NMR spectrum. Notably, one
3
Co−CH resonance at −0.05 ppm exhibited an apparent 10-
3
3
line splitting pattern derived from J
coupling, indicating the
P−H
Figure 10. Comparison of methyl group exchange processes between
presence of more than three phosphorus nuclei. Additional
evidence that the isolated products resulted from iodide
dissociation was obtained from analogous experiments treating
PMe - and triphos-supported cobalt(III) dimethyl chloride complexes.
3
After 36 h the trimethylphosphine species had reached an
the triflate complex 2-Me with triphos ligand to produce
1
2
equilibrium ratio (1:2) of 3-Me to 3-d Me , as judged by H
2
3
2
samples with identical NMR spectra to those originating from
NMR spectroscopy, while the triphos complexes had achieved
1
-Me2.
minimal exchange with a 1:0.4 ratio of 4-Me to 4-d Me . This
2
3
2
The mechanism of the cationic [(triphos)Co(CH ) PMe ]I
3
2
3
observation is consistent with requisite phosphine dissociation
during cobalt-to-cobalt methyl group exchange and strongly
supports proposed path E as the intermetallic methyl exchange
mechanism. Additionally, crossover experiments between 4-
Me and 3-d Me showed substantial exchange over the 36 h
formation was not examined experimentally, but is likely to
begin via substitution between a triphos −PMe group and the
2
PMe of 1-Me bound trans to Co−CH (this PMe site has
3
2
3
3
3
proven the most labile in previous EXSY NMR studies). Such
a sequence could lead to a circumstance where the last triphos
phosphine group to bind is positioned opposite the remaining
Co−PMe3 bond, affording preferential substitution of the
2
6
2
period. In this case, the formation of 4-d Me and 3-d Me was
3
2
3
2
evident within 4 h of mixing, dramatically faster than the 24 h
required to observe 4-d Me by NMR spectroscopy in the
3
2
iodide over PMe . If this is true, use of a more strongly bound
3
triphos self-exchange experiment. This qualitative difference in
exchange rate suggests that only one of the two cobalt centers
must achieve coordinative unsaturation in order for exchange to
occur (Figure 11) and provides additional evidence for
proposed path E.
halide could slow anion dissociation, permit more time for
geometric rearrangements, and allow formation of the desired
neutral (triphos)Co(CH ) X (X = halide) species. Fortunately,
3
2
stirring the chloride congener cis,mer-(PMe ) Co(CH ) Cl (3-
3
3
3 2
2
8
Me2) with triphos in tetrahydrofuran over one day did indeed
afford good yields of cis,mer-(triphos)Co(CH ) Cl (4-Me ) (eq
3
2
2
CONCLUDING REMARKS
Having investigated multiple plausible mechanisms for cobalt-
to-cobalt methyl group exchange in cis,mer-(PMe ) Co-
■
7). Small quantities of the corresponding cationic product were
3
3
(
CH ) X, the available evidence most strongly supports a
3 2
route where phosphine dissociation at one cobalt center is
followed by attachment to a second equivalent of cobalt(III)
dimethyl halide to transiently form a high-energy dicobalt
structure with bridging methyl groups (Figure 8; path E). The
dissociation of the bridging methyl structure could then
produce a net exchange of the labeled alkyl groups. This
mechanism is supported by relative isotopic exchange rates and
also observed in the reaction, but were easily removed by
filtration from hydrocarbon solvent. Complex 4-Me2 was
prepared only in small quantities and characterized by one-
crossover experiments with PMe - and triphos-supported 3-
3
Me and 4-Me . Multiple plausible alternative pathways for
2
2
1
and two-dimensional NMR analysis. The H NMR spectrum of
intermetallic exchange, including radical crossover, reversible
methyl halide elimination, and cobalt(I)-catalyzed disproportio-
nation, were also considered but excluded on the basis of
trapping tests, isotopic labeling studies, and other comparative
rate experiments. During the course of these mechanistic
investigations, a second phosphine-dependent methyl exchange
4
-Me exhibited signals for the two Co−CH sites at −0.98
2
3
(
dt) and 1.22 (td) ppm. Significantly, two of the three
resonances for the PMe ligands appeared as virtual triplets in
n
1
the H NMR spectrum, indicating a trans orientation of the
triphos −PMe groups and establishing a meridional coordina-
2
tion for the chelate.
reaction was revealed to swap −CH substituents between
3
8
03
dx.doi.org/10.1021/om3009537 | Organometallics 2013, 32, 798−806

Organometallics
Article
Figure 11. Crossover methyl group exchange between 4-Me and 3-d Me .
2
6
2
Figure 12. Partial reaction pathway for ethane reductive elimination from 1-Me2.
cobalt and PMe . This second exchange reaction complicates
some mechanistic probes of the cobalt-to-cobalt exchange, but
calibrated using ethylene glycol and methanol as previously
15
3
described. IR spectra were recorded on Jasco 4100 FTIR or Metler
Toledo React IR spectrometers. X-ray crystallographic data were
collected on Bruker Smart Apex I or D8 Quest diffractometers. The
structures were solved from direct methods and Fourier syntheses and
refined by full-matrix least-squares procedures with anisotropic thermal
parameters for all non-hydrogen atoms. Crystallographic calculations
were carried out using SHELXTL. Elemental analyses were performed
at Robertson Microlit Laboratories, Inc., in Ledgewood, NJ, or Atlantic
Microlab, Inc., in Norcross, GA (USA).
occurs with negligible rates in the absence of added PMe . The
3
preliminary data are consistent with P−CH3 scission
proceeding through reversible PMe nucleophilic attack on a
3
cobalt−methyl group to produce a transient cobalt phosphor-
ane intermediate.
The two pathways for methyl group exchange at cobalt are
each influenced by the presence of free phosphine, as is the rate
of reductive ethane elimination from 1-Me (Figure 12), which
Preparation of cis,mer-(PMe ) Co(CH ) (SO CF ) (2-Me ).
2
3 3
3 2
3
3
2
Method A. A 50 mL Schlenk flask was charged with 0.092 g (0.24
was the focus of earlier studies that led to the investigations
described here. The methyl exchange process between cobalt
and phosphine is unlikely to impact the rate of C−C bond
reductive elimination from cobalt(III), as the two reactions are
observable only under essentially mutually exclusive concen-
mmol) of (PMe ) Co(CH ) and approximately 10 mL of diethyl
3
4
3
ether. On a high-vacuum line, 2 equiv (89 Torr in 101 mL; 0.49
mmol) of MeOTf was transferred into the Schlenk flask at −196 °C
via calibrated gas bulb. The reaction mixture was then stirred at
ambient temperature for two hours. The reaction mixture was cannula
filtered into a new Schlenk flask to remove a dark precipitate. The
resulting orange solution was evaporated to afford 35 mg (31%) of 2-
trations of free PMe . Conversely, ethane elimination from 1-
3
Me and the cobalt-to-cobalt methyl group exchange appear to
2
share a common unobserved intermediate, (PMe ) Co(CH ) I,
Me as an orange powder. This compound can be recrystallized from
3
2
3 2
2
which means intermetallic methyl exchange could in principle
influence the rate and mechanistic investigations of reductive
eliminations from cobalt(III) species. However the cobalt-to-
cobalt methyl exchange proceeds on a time scale far faster than
C−C bond coupling, takes place without any other detectable
intermediates, and is a degenerate reaction that only reproduces
the reactants (though isotoptically scrambled if the species are
labeled). These features indicate that the intermetallic methyl
exchange reactions observed in cobalt(III) dimethyl halide
complexes are unlikely to have meaningful impacts on the
kinetics of C−C bond reductive elimination from the species,
which our laboratory is pursuing as part of ongoing efforts to
elucidate the pathways and influencing factors for selective C−
C bond formation at base metal centers.
diethyl ether at −35 °C. Similar procedures performed in a J. Young
tube indicate near quantitative conversion by NMR spectroscopy,
suggesting the modest isolated yield may be due to decomposition of
2
-Me2 during workup. Attempts to perform the synthesis in the
absence of light did not dramatically improve the yield.
Method B. A J. Young NMR tube was charge with 0.007g (0.019
mmol) of 1-Me and approximately 400 μL of benzene-d . On a high-
2
6
vacuum line, 1 equiv (27 Torr in 10.9 mL; 0.019 mmol) of MeOTf
was added at −196 °C via calibrated gas bulb. Full conversion of 1-
1
Me to (PMe ) Co(CH ) OTf and MeI was observed by H NMR
2
3
3
3 2
spectroscopy in less than one hour at ambient temperature. Anal.
Calcd for C H F O P SCo: C 30.91; H 7.13. Found: C 30.81; H
1
2
33
3
3 3
1
6.84. H NMR (23 °C; toluene-d ): δ −1.57 (bs, 9.6 Hz, 3H, Co-
8
3
1
1
CH ), 0.67 (brs, 3H, Co-CH ); 1.05 (brs, 27H, P-CH ). P{ H}
3
3
3
1
9
NMR (23 °C; toluene-d ): δ −7.71 (brs, 1P), 6.19 (brs, 2P). F NMR
8
1
(
(
23 °C; C D ): δ −79.07. H NMR (−40 °C, toluene-d ): δ −1.65
6
6
8
dt, 9.6 Hz, 9.5 Hz, 3H, Co-CH ), 0.72 (brt, 3H, Co-CH ); 0.96 (d,
3
3
EXPERIMENTAL SECTION
31
1
■
6.5 Hz, 9H, P-CH
3
), 1.03 (vt, 3.0 Hz, 18H, P-CH ). P{ H} NMR
3
General Considerations. All manipulations were carried out using
(−40 °C; toluene-d
8
): δ −5.95 (t, 27.9 Hz, 1P), 9.09 (d, 27.9 Hz, 2P).
13
1
standard vacuum, Schlenk, cannula, or glovebox techniques.
C{ H} NMR (−40 °C; toluene-d ): δ −22.80 (Co-CH ), 3.43 (Co-
8
3
1
6
28
20
1
Compounds 1-Me , 3-Me2, (PMe ) CoCH , and MeP-
CH ), 12.96 (P-CH ), 15.89 (P-CH ), 120.13 (CF , J = 320.3 Hz).
2
3
4
3
3
3
3
3
CF
27
−1
(
CH CH PMe ) were prepared according to previously described
IR (KBr): ν
= 1299 cm .
2
2
2
2
OSO
procedures. All other chemicals were purchased from Aldrich, VWR,
Strem, or Cambridge Isotope Laboratories. Volatile chemicals were
dried with 4 Å molecular sieves and distilled prior to use. Solvents
Preparation of cis,mer-[MeP(CH CH PMe ) ]Co(CH ) (Cl) (4-
2 2 2 2 3 2
Me
). A 20 mL scintillation vial was charged with 0.064 g (0.181
, 0.041 g (0.183 mmol) of MeP(CH CH PMe , and
2
mmol) of 3-Me
)
2 2
2
2
2
3
0
were dried and deoxygenated using literature procedures.
approximately 5 mL of tetrahydrofuran. The solution was stirred at
ambient temperature for approximately one day. Three times during
this period, the solvent and free trimethylphosphine were removed in
vacuo followed by redissolving the reaction mixture in tetrahydrofuran.
After one day the solvent was removed and the residue extracted using
1
2
13
19
31
H, H, C, F, and P NMR spectra were recorded on Bruker
1
13
DRX 400 Avance and 300 Avance MHz spectrometers. H and
C
P
31
chemical shifts are referenced to residual protio solvent signals;
and 19F chemical shifts are referenced to an external standard of
H PO and trifluorotoluene, respectively. Probe temperatures were
pentane and diethyl ether to yield 50 mg (83%) of 4-Me as an orange
3
4
2
8
04
dx.doi.org/10.1021/om3009537 | Organometallics 2013, 32, 798−806

Organometallics
Article
powder when dried under vacuum. Complex 4-Me is rather lipophilic
(4) (a) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry;
University Science Books: Sausalito, 1987. (b) Crabtree, R. H. The
Organometallic Chemistry of the Transition Metals; John Wiley & Sons:
Hoboken, 2005. (c) Hartwig, J. Organotransition Metal Chemistry:
From Bonding to Catalysis; University Science: Sausalito, 2010.
(5) (a) Tsou, T.-T.; Loots, M.; Halpern, J. J. Am. Chem. Soc. 1982,
104, 623. (b) Halpern, J. Acc. Chem. Res. 1982, 15, 238.
(6) (a) Mok, C. Y.; Endicott, J. F. J. Am. Chem. Soc. 1978, 100, 123.
(b) Riordan, C. G.; Wei, P. J. Am. Chem. Soc. 1994, 116, 2189.
(c) Hager, E.; Sivaramakrishna, A.; Clayton, H. S.; Mogorosi, M. M.;
Moss, J. R. Coord. Chem. Rev. 2008, 252, 1668. (d) Hojilla Atienza, C.
C.; Milsmann, C.; Lobkovsky, E.; Chirik, P. J. Angew. Chem., Int. Ed
2011, 50, 8143.
2
and requires hours of drying under vacuum at room temperature to
obtain powdered samples from solution. Elemental analysis was not
obtained for this species due to the small scale of synthesis, which was
necessitated by safety concerns in generating dimethyl phosphine for
the preparation of MeP(CH CH PMe ) . A description of safe
2
2
2 2
synthetic procedures for this ligand has been reported by King and
27
1
Cloyd. H NMR (C D ): δ −0.98 (dt, 8.4 Hz, 7.9 Hz, 3H, Co-CH ),
6
6
3
0
8
.67 (d, 7.2 Hz, 3H, P-CH ), 0.85 (vt, 3.4 Hz, 6H, P-CH ), 1.22 (td,
3
3
.0 Hz, 3.6 Hz, 3H, Co-CH ), 1.26−1.43 (m, 4H, PCH ), 1.51 (vt, 4.1
3
2
31
1
Hz, 6H, P-CH ), 2.08−2.46 (m, 4H, PCH ). P{ H} NMR (C D ): δ
3
2
6
6
13
1
5
0.41 (br, 2P), 111.40 (br, 1P). C{ H} NMR (C D ): δ −13.43
6
6
(
(
Co-CH ), −0.58 (Co-CH ), 10.86 (P-CH ), 11.64 (P-CH ), 12.26
3
3
3
3
P-CH ), 24.81 (PCH ), 33.77 (PCH ).
3
2
2
General Procedure for Monitoring Bimetallic Methyl Group
(7) Davies, J. A.; Watson, P. L.; Liebman, J. F.; Greenberg, A. In
1
Exchange by H NMR Spectroscopy. In a typical experiment a J.
Young NMR tube(s) was charged with a stock solution of 1-d Me in
Selective Hydrocarbon Activation: Principles and Progress; VCH Publish-
ers: New York, 1990.
3
2
benzene-d (0.015−0.020 M). The sample(s) was then allowed to
6
(8) The presence of ambient light does appear to accelerate this
stand at ambient temperature. At intervals the sample(s) was inserted
exchange modestly, though it may also be observed on comparable
time scales with the exclusion of light.
31
1
into a 400 MHz NMR probe, and P decoupled H NMR spectra
were acquired. The ratio of 1-d Me and 1-Me was determined by
3
2
2
(
9) Hawari, J. A.; Engel, P. S.; Griller, D. Int. J. Chem. Kinet. 1985, 17,
215.
10) Attempts to use isolable radicals as traps (e.g., TEMPO,
integration of singlet resonances at −0.743 (1-Me ) and −0.759 ppm
2
1
(
1-d Me ). When possible, samples were cooled to 10 °C to obtain
3 2
(
sharper line shapes and better separation of the peaks for the two
isotopologues. Analogous procedures can be used for monitoring
exchange between 1-Me and 1-d Me using equimolar solutions. This
Galvinoxyl) resulted in significant decomposition of 1-Me2 but
without stoichiometric observation of the methyl abstraction product.
See: Albeniz, A. C.; Espinet, P.; Lopez-Fernandez, R.; Sen, A. J. Am.
Chem. Soc. 2002, 124, 11278.
2
6
2
method was also employed to study exchange in 3-Me and 4-Me2
2
isotopologues. Relative rates of isotopic exchange were always
monitored using parallel NMR tube samples to obviate variations in
temperature, mixing, or exposure to ambient light (which produced
modest accelerations in isotopic exchange).
(11) (a) Roy, A. H.; Hartwig, J. F. J. Am. Chem. Soc. 2003, 125,
1
2
3944. (b) Yahv-Levi, A.; Goldberg, I.; Vigalok, A. J. Am. Chem. Soc.
006, 128, 8710. (c) Lui, H.; Li, C.; Qui, D.; Tong, X. J. Am. Chem.
Soc. 2011, 133, 6187. (d) Powers, D. C.; Ritter, T. Acc. Chem. Res.
012, 45, 840.
12) (a) Goldberg, K. I.; Yan, J. Y.; Winter, E. L. J. Am. Chem. Soc.
994, 116, 1573. (b) Goldberg, K. I.; Yan, J. Y.; Breitung, E. M. J. Am.
2
(
1
ASSOCIATED CONTENT
Supporting Information
Crystal Information File for 2-Me and 3-Me and NMR
■
*
S
2
2
Chem. Soc. 1995, 117, 6889. (c) Frech, C. M.; Milstein, D. J. Am.
Chem. Soc. 2006, 128, 12434. (d) Scott, V. J.; Labinger, J A.; Bercaw, J.
E. Organometallics 2010, 29, 4090.
spectrum of 4-Me . These data are available free of charge via
2
the Internet at http://pubs.acs.org.
(
13) Within two hours of CD I addition a copious amount of black
3
AUTHOR INFORMATION
Corresponding Author
E-mail: wb36@brown.edu.
material began to precipitate from solution.
■
(
(
14) See Experimental Section.
15) Sandstrom, J. Dynamic NMR Spectroscopy; Academic Press: New
̈
*
York, 1982.
Notes
(16) Beck, R.; Klein, H. F. Z. Anorg. Allg. Chem. 2008, 634, 1971.
(17) Xu, H.; Williard, P. G.; Bernskoetter, W. H. Organometallics
The authors declare no competing financial interest.
2
(
012, 31, 1588.
18) (a) Topich, J.; Halpern, J. Inorg. Chem. 1979, 18, 1339.
b) Vol’pin, M. E.; Levitin, I.; Sigan, A. I.; Nikitaev, A. T. J. Organomet.
ACKNOWLEDGMENTS
We gratefully acknowledge funding by Brown University and
the American Chemical Society Petroleum Research Fund
■
(
Chem. 1985, 279, 263. (c) Toscano, P. J.; Barren, E.; Seligson, A. L.
Organometallics 1989, 8, 2085. (d) Byrne, E. K.; Theopold, K. H. J.
Am. Chem. Soc. 1989, 111, 3887. (e) Harmer, J.; Van Doorslaer, S.;
Gromov, I.; Broring, M.; Jeschke, G.; Schweiger, A. J. Phys. Chem. B
(52066-DNI3).
REFERENCES
2
002, 106, 2801. (f) Ohkubo, K.; Fukuzumi, S. J. Phys. Chem. A 2005,
■
(
1) For examples including intermetallic methyl group exchange see:
a) Bryndza, H. E.; Evitt, E. R.; Bergman, R. B. J. Am. Chem. Soc. 1980,
109, 1105 and references therein.
(
(19) (a) Socol, S. M.; Verkade, J. G. Inorg. Chem. 1986, 25, 2658.
(b) Lui, W.; Hempstead, M. R.; Nevin, A. W.; Melnik, M.; Lever, A. B.
P.; Leznoff, C. C. J. Chem. Soc., Dalton Trans. 1987, 11, 2511.
(c) Andreev, A.; Ivanova, V.; Prahov, L.; Schopov, I. D. J. Mol. Catal. A
1995, 95, 197.
102, 4948 and references therein. (b) Pankowski, M.; Samuel, E. J.
Organomet. Chem. 1981, 221, C21. (c) Baxter, S. M; Ferguson, G. S.;
Wolczanski, P. T. J. Am. Chem. Soc. 1988, 110, 4231. (d) Hill, G. S.;
Puddephatt, R. J. Organometallics 1997, 16, 4522. (e) Hill, G. S.; Yap,
G. P. A.; Puddephatt, R. J. Organometallics 1999, 18, 1408. (f) Ram, M.
S.; Riordan, C. G.; Yap, G. P. A.; Liable-Sands, L.; Rheingold, A. L.;
Marchaj, A.; Norton, J. R. J. Am. Chem. Soc. 1997, 119, 1648.
(20) (PMe ) CoCl is a precursor employed in the synthesis of 1-Me
3
3
2
two steps prior to isolation of the dimethyl cobalt(III) complex: Klein,
H. F.; Karsch, H. H. Chem. Ber. 1975, 108, 944.
(
2) (a) Remy, M. S.; Cundari, T. R.; Sanford, M. S. Organometallics
(21) (a) Capelle, B.; Beuchamp, A.; Dartiguenave, M.; Dartiguenave,
Y.; Klein, H.-F. J. Am. Chem. Soc. 1982, 104, 3891. (b) Capelle, B.;
Dartiguenave, M.; Dartiguenave, Y.; Beuchamp, A. J. Am. Chem. Soc.
1983, 105, 4662.
(22) For select examples of first-row metal complexes with bridging
methyl ligands see: (a) Dawkins, G. M.; Green, M.; Orpen, A. G.;
Stone, F. G. A. J. Chem. Soc., Chem. Commun. 1982, 41. (b) Casey, C.
P.; Fagan, P. J.; Miles, W. H. J. Am. Chem. Soc. 1982, 104, 1134.
2
010, 29, 1522. (b) Tang, F.; Zhang, Y.; Rath, N. P.; Mirica, L. M.
Organometallics 2012, 31, 6690. (c) Khusnutdinova, J. R.; Rath, N. P.;
Mirica, L. M. J. Am. Chem. Soc. 2012, 134, 2414. (d) Morita, D. K.;
Stille, J. K.; Norton, J. R. J. Am. Chem. Soc. 1995, 117, 8576.
(
e) Goodson, F. E.; Wallow, T. I.; Novak, B. M. J. Am. Chem. Soc.
997, 119, 12441.
3) Xu, H.; Bernskoetter, W. H. J. Am. Chem. Soc. 2011, 133, 14956.
1
(
8
05
dx.doi.org/10.1021/om3009537 | Organometallics 2013, 32, 798−806

Organometallics
Article
(c) Niemeyer, M.; Power, P. P. J. Chem. Soc., Chem. Commun. 1996,
1573. (d) Chai, J.; Zhu, H.; Fan, H.; Roesky, H. W.; Magull, J.
Organometallics 2004, 23, 1177. (e) Ni, C.; Power, P. P. Organo-
metallics 2009, 28, 6541.
(
23) After 48 h at ambient temperature, chlorobenzene solutions of
1
-Me showed only small amounts of degradation.
2
(
24) (a) Garrou, P. Chem. Rev. 1985, 85, 171. (b) Macgregor, S. A.
Chem. Soc. Rev. 2007, 36, 67.
25) For examples of P−C cleavage at PMe see: (a) Hartwig, J. F.;
(
3
Bergman, R. B.; Andersen, R. A. J. Organomet. Chem. 1990, 394, 417.
(
b) Lin, W.; Wilson, S. R.; Girolami, G. S. Inorg. Chem. 1994, 33, 2265.
(
c) Nakajima, T.; Shimizu, I.; Kobayashi, K.; Koshino, H.; Wakatsuki,
Y. Inorg. Chem. 1997, 36, 6440. (d) Shin, J. H.; Parkin, G. Chem.
Commun. 1998, 1273. (e) Hirsekorn, K. F.; Veige, A. S.; Wolczanski, P.
T. J. Am. Chem. Soc. 2006, 128, 2192.
(
26) For a review on metallaphosphoranes see: Nakazawa, H.; Kubo,
K.; Miyoshi, K. Chem. Soc. Jpn. 2001, 74, 2255.
(
(
(
27) King, R. B.; Cloyd, J. C. J. Am. Chem. Soc. 1975, 97, 53.
28) Klein, H. F.; Karsch, H. H. Chem. Ber. 1975, 108, 956.
29) The scrambling studies were conducted between 3-Me and 3-
2
d Me owing to synthetic difficulties in obtaining 3-d Me .
6
2
3
2
(
30) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518.
8
06
dx.doi.org/10.1021/om3009537 | Organometallics 2013, 32, 798−806