Thermal activation of hydrocarbon C - H bonds by tungsten alkylidene complexes

Article abstract of DOI:10.1021/ja002457e

Thermal activation of Cp*W(NO)(CH₂CMe₃)₂ (1) in neat hydrocarbon solutions transiently generates the neopentylidene complex, Cp*W(NO)( = CHCMe₃) (A), which subsequently activates solvent C - H bonds. For example, the thermolysis of 1 in tetramethylsilane and perdeuteriotetramethylsilane results in the clean formation of Cp*W(NO)(CH₂CMe₃) (CH₂SiMe₃) (2) and Cp*W(NO)(CHDCMe₃) [CD₂Si(CD₃)₃] (2-d₁₂), respectively, in virtually quantitative yields. The neopentylidene intermediate A can be trapped by PMe₃ to obtain Cp*W(NO)( = CHCMe₃)(PMe₃) in two isomeric forms (4a-b), and in benzene, 1 cleanly forms the phenyl complex Cp*W(NO)(CH₂CMe₃)(C₆H₅) (5). Kinetic and mechanistic studies indicate that the C - H activation chemistry derived from 1 proceeds through two distinct steps, namely, (1) rate-determining intramolecular α-H elimination of neopentane from 1 to form A and (2) 1,2-cis addition of a substrate C - H bond across the W = C linkage in A. The thermolysis of 1 in cyclohexane in the presence of PMe₃ yields 4a-b as well as the olefin complex Cp*W(NO)(η²-cyclohexene)(PMe₃) (6). In contrast, methylcyclohexane and ethylcyclohexane afford principally the allyl hydride complexes Cp*W(NO)(η³-C₇H₁₁)(H) (7a-b) and Cp*W(NO)(η³-C₈H₁₃)(H) (8a-b), respectively, under identical experimental conditions. The thermolysis of 1 in toluene affords a surprisingly complex mixture of six products. The two major products are the neopentyl aryl complexes, Cp*W(NO)(CH₂CMe₃) (C₆H₄-3-Me) (9a) and Cp*W(NO)(CH₂CMe₃) (C₆H₄-4-Me) (9b), in approximately 47 and 33% yields. Of the other four products, one is the aryl isomer of 9a-b, namely, Cp*W(NO)(CH₂CMe₃) (C₆H₄-2-Me) (9c) (～1%). The remaining three products all arise from the incorporation of two molecules of toluene; namely, Cp*W(NO)(CH₂C₆H₅) (C₆H₄-3-Me) (11a; ～12%), Cp*W(NO)(CH₂C₆H₅)- (C₆H₄-4-Me) (11b; ～6%), and Cp*W(NO)(CH₂C₆H₅)₂ (10; ～1%). It has been demonstrated that the formation of complexes 10 and 11a-b involves the transient formation of Cp*W(NO)(CH₂CMe₃)(CH₂ C₆H₅) (12), the product of toluene activation at the methyl position, which reductively eliminates neopentane to generate the C - H activating benzylidene complex Cp*W(NO)(= CHC₆H₅) (B). Consistently, the thermolysis of independently prepared 12 in benzene and benzene-d₆ affords Cp*W(NO)(CH₂C₆H₅) (C₆H₅) (13) and Cp*W(NO)-(CHDC₆H₅)(C₆D₅) (13-d₆), respectively, in addition to free neopentane. Intermediate B can also be trapped by PMe₃ to obtain the adducts Cp*W(NO)( = CHC₆H₅)(PMe₃) (14a - b) in two rotameric forms. From their reactions with toluene, it can be deduced that both alkylidene intermediates A and B exhibit a preference for activating the stronger aryl sp² C - H bonds. The C - H activating ability of B also encompasses aliphatic substrates as well as it reacts with tetramethylsilane and cyclohexanes in a manner similar to that summarized above for A. All new complexes have been characterized by conventional spectroscopic methods, and the solid-state molecular structures of 4a, 6, 7a, 8a, and 14a have been established by X-ray diffraction methods.

Full text of DOI:10.1021/ja002457e

612
J. Am. Chem. Soc. 2001, 123, 612-624
Thermal Activation of Hydrocarbon C-H Bonds by Tungsten
Alkylidene Complexes
Craig S. Adams, Peter Legzdins,* and Elizabeth Tran
Contribution from the Department of Chemistry, The UniVersity of British Columbia,
VancouVer, British Columbia, Canada V6T 1Z1
ReceiVed July 6, 2000
Abstract: Thermal activation of Cp*W(NO)(CH2CMe3)2 (1) in neat hydrocarbon solutions transiently generates
the neopentylidene complex, Cp*W(NO)(dCHCMe3) (A), which subsequently activates solvent C-H bonds.
For example, the thermolysis of 1 in tetramethylsilane and perdeuteriotetramethylsilane results in the clean
formation of Cp*W(NO)(CH2CMe3)(CH2SiMe3) (2) and Cp*W(NO)(CHDCMe3)[CD2Si(CD3)3] (2-d12),
respectively, in virtually quantitative yields. The neopentylidene intermediate A can be trapped by PMe3 to
obtain Cp*W(NO)(dCHCMe3)(PMe3) in two isomeric forms (4a-b), and in benzene, 1 cleanly forms the
phenyl complex Cp*W(NO)(CH2CMe3)(C6H5) (5). Kinetic and mechanistic studies indicate that the C-H
activation chemistry derived from 1 proceeds through two distinct steps, namely, (1) rate-determining
intramolecular R-H elimination of neopentane from 1 to form A and (2) 1,2-cis addition of a substrate C-H
bond across the WdC linkage in A. The thermolysis of 1 in cyclohexane in the presence of PMe3 yields 4a-b
as well as the olefin complex Cp*W(NO)(η2-cyclohexene)(PMe3) (6). In contrast, methylcyclohexane and
ethylcyclohexane afford principally the allyl hydride complexes Cp*W(NO)(η3-C7H11)(H) (7a-b) and
Cp*W(NO)(η3-C8H13)(H) (8a-b), respectively, under identical experimental conditions. The thermolysis of 1
in toluene affords a surprisingly complex mixture of six products. The two major products are the neopentyl
aryl complexes, Cp*W(NO)(CH2CMe3)(C6H4-3-Me) (9a) and Cp*W(NO)(CH2CMe3)(C6H4-4-Me) (9b), in
approximately 47 and 33% yields. Of the other four products, one is the aryl isomer of 9a-b, namely,
Cp*W(NO)(CH2CMe3)(C6H4-2-Me) (9c) (∼1%). The remaining three products all arise from the incorporation
of two molecules of toluene; namely, Cp*W(NO)(CH2C6H5)(C6H4-3-Me) (11a; ∼12%), Cp*W(NO)(CH2C6H5)-
(C6H4-4-Me) (11b; ∼6%), and Cp*W(NO)(CH2C6H5)2 (10; ∼1%). It has been demonstrated that the formation
of complexes 10 and 11a-b involves the transient formation of Cp*W(NO)(CH2CMe3)(CH2C6H5) (12), the
product of toluene activation at the methyl position, which reductively eliminates neopentane to generate the
C-H activating benzylidene complex Cp*W(NO)(dCHC6H5) (B). Consistently, the thermolysis of independ-
ently prepared 12 in benzene and benzene-d6 affords Cp*W(NO)(CH2C6H5)(C6H5) (13) and Cp*W(NO)-
(CHDC6H5)(C6D5) (13-d6), respectively, in addition to free neopentane. Intermediate B can also be trapped by
PMe3 to obtain the adducts Cp*W(NO)(dCHC6H5)(PMe3) (14a-b) in two rotameric forms. From their reactions
with toluene, it can be deduced that both alkylidene intermediates A and B exhibit a preference for activating
the stronger aryl sp2 C-H bonds. The C-H activating ability of B also encompasses aliphatic substrates as
well as it reacts with tetramethylsilane and cyclohexanes in a manner similar to that summarized above for A.
All new complexes have been characterized by conventional spectroscopic methods, and the solid-state molecular
structures of 4a, 6, 7a, 8a, and 14a have been established by X-ray diffraction methods.
Introduction
oxidatively add C-H bonds to the metal center,3 transition-
metal, lanthanide, and actinide complexes that facilitate C-H
activation via M-C σ-bond metathesis,4 and early- to mid-
transition-metal complexes that add C-H bonds across a
MdNR linkage.5 A fundamental understanding of the discrete
C-H activation processes has been acquired for many of these
The development of efficient, selective, and catalytic methods
for the direct conversion of hydrocarbon feedstocks into
functionalized organic compounds is an important ongoing
challenge in modern chemistry.1 Considerable progress in this
regard has been made in the field of organometallic chemistry
where numerous metal-containing complexes have been found
to effect selective intermolecular C-H bond activations.1,2
Notable examples include late-transition-metal complexes that
(3) (a) Janowicz, A. H.; Bergman, R. G. J. Am. Chem. Soc. 1983, 105,
3929-3939. (b) Hoyano, J. K.; Graham, W. A. G. J. Am. Chem. Soc. 1982,
104, 3723-3725. (c) Jones, W. D.; Feher, F. J. Acc. Chem. Res. 1989, 22,
91-100. (d) Arndsten, B. A.; Bergman, R. G. Science 1995, 270, 1970-
1973. (e) Wick, D. D.; Goldberg, K. I. J. Am. Chem. Soc. 1997, 119, 10235-
10236 and references therein.
(4) (a) Erker, G. J. Organomet. Chem. 1977, 134, 189-202. (b) Watson,
P. L.; Parshall, G. W. Acc. Chem. Res. 1985, 18, 51-56. (c) Fendrick, C.
M.; Marks, T. J. J. Am. Chem. Soc. 1986, 108, 425-437. (d) Thompson,
M. E.; Baxter, S. M.; Bulls, R.; Burger, B. J.; Nolan, M. C.; Santarsiero,
B. D.; Schaefer, W. P.; Bercaw, J. E. J. Am. Chem. Soc. 1987, 109, 203-
219. (e) Debad, J. D.; Legzdins, P.; Lumb, S. A.; Rettig, S. J.; Batchelor,
R. J.; Einstein, F. W. B. Organometallics 1999, 18, 3414-3428.
(1) (a) ActiVation and Functionalization of Alkanes; Hill, C. L., Ed.;
Wiley-Interscience: New York, 1989. (b) Crabtree, R. H. Chem. ReV. 1995,
95, 987-1007.
(2) For leading reviews, see: (a) Shilov, A. E.; Shul’pin, G. B. Chem.
ReV. 1997, 97, 2879-2932. (b) Arndtsen, B. A.; Bergman, R. G.; Mobley,
T. A.; Peterson, T. H. Acc. Chem. Res. 1995, 28, 154-162. (c) Stahl, S. S.;
Labinger, J. A.; Bercaw, J. E. Angew. Chem., Intl. Ed. Engl. 1998, 37, 2180-
2192. (d) Crabtree, R. H. Chem. ReV. 1985, 85, 245-269.
10.1021/ja002457e CCC: $20.00 © 2001 American Chemical Society
Published on Web 01/09/2001

C-H Bond ActiVation by W Alkylidene Complexes
J. Am. Chem. Soc., Vol. 123, No. 4, 2001 613
Scheme 1
on the nature of the solvent employed. Observed modes of
reactivity include the spontaneous functionalization of the
activated substrates by dehydrogenation and the formation of a
new benzylidene complex that is also capable of selectively
activating both alkane and arene C-H bonds. To the best of
our knowledge, this constitutes the first detailed study of the
activation of C-H bonds by alkylidene complexes.
complexes, and the chemistry of select systems has been
significantly advanced toward efficient, catalytic functionaliza-
tions.6
In recent years, organometallic C-H activation research has
expanded to include a small class of complexes that intermo-
lecularly activate C-H bonds via their addition across the MdC
bond of an alkylidene intermediate.7 In each case, the reactive
species is a transient neopentylidene complex generated ther-
mally in situ from a cis-bis(neopentyl) precursor by the
microscopic reverse of the C-H activation reaction, namely,
R-Η elimination of neopentane (Scheme 1). The neopentylidene
Results and Discussion
A. Generation of Cp*W(NO)(dCHCMe3) (A) and the
Mechanism by Which It Activates C-H Bonds. Activation
of Tetramethysilane. The thermolysis of 1 for 40 h at 70 °C
in neat tetramethylsilane results in the clean formation (∼97%
as judged by 1H NMR spectroscopy) of the known12 mixed alkyl
complex, Cp*W(NO)(CH2CMe3)(CH2SiMe3) (2) (eq 1). Com-
8
intermediates generated from (2,6-iPr2C6H3N)2Cr(CH2CMe3)2
and Cp′2Ti(CH2CMe3)2 [Cp′ ) C5H5 (Cp) or C5Me5 (Cp*)]9 in
this manner activate an aryl sp2 C-H bond of benzene to afford
the corresponding phenyl neopentyl complexes. There is also
strong evidence for the activation of alkane and arene C-H
bonds by a neopentylidene intermediate during the thermal
decomposition of Ti(CH2CMe3)4.10 Finally, we have observed
that the thermolysis of Cp*W(NO)(CH2CMe3)2 (1) leads to the
formation of a neopentylidene complex, namely, Cp*W(NO)-
(dCHCMe3) (A), which activates the primary and secondary
C-H bonds of tetramethylsilane and cyclohexane, respec-
tively.11 To date it appears that neopentylidene complex A is
unique in its ability to activate aliphatic C-H bonds to yield
isolable, well-defined products.
With the intention of advancing the understanding of alkyl-
idene-mediated C-H activation processes, we now report the
results of our survey of the thermal chemistry of 1 in various
solvents. Included in this report are details concerning the
previously communicated activation of tetramethylsilane and
cyclohexane, as well as additional mechanistic information
concerning the formation and reactivity of intermediate A. We
have also extended the scope of this chemistry to encompass
select alkane and arene substrates in an attempt to assess the
relative intramolecular selectivities of A for aliphatic and
aromatic C-H bonds. During these studies, we have discovered
that additional C-H activation chemistry can occur depending
plex 2 is clearly derived from 1 via the activation of an sp3
aliphatic C-H bond of a solvent molecule. Several plausible
mechanisms could account for this transformation. These include
direct σ-bond metathesis or homolytic cleavage mechanisms,
as well as pathways involving metallacyclic or alkylidene
intermediates via intramolecular R-H, γ-H, or Cp*-H abstrac-
tions. Our mechanistic studies are only consistent with the
formation of the alkylidene intermediate.
Monitoring by 1H NMR spectroscopy indicates that the
thermolysis of 1 in perdeuteriotetramethylsilane at 70 °C for
40 h produces free neopentane (1 equiv) and Cp*W(NO)-
(CHDCMe3)[CD2Si(CD3)3] (2-d12). Complex 2-d12 can be
isolated in high yield (93%) by crystallization from pentane. It
1
exhibits a singlet in its H NMR spectrum for the anticlinal
methylene hydrogen atom (with respect to the new M-C
bond)13 and a 1:1:1 triplet in its 13C{1H} NMR spectrum for
the methylene carbon of the neopentyl ligand. In addition, there
is a broad singlet in its 2H{1H} NMR spectrum that corresponds
to the deuteron at the synclinal neopentyl methylene position.
(5) (a) Schaller, C. P.; Wolczanski, P. T. Inorg. Chem. 1993, 32, 131-
144. (b) Schaller, C. P.; Cummins, C. C.; Wolczanski, P. T. J. Am. Chem.
Soc. 1996, 118, 591-611. (c) Bennett, J. L.; Wolczanski, P. T. J. Am. Chem.
Soc. 1997, 119, 10696-10719. (d) Schaefer, D. F., II; Wolczanski, P. T. J.
Am. Chem. Soc. 1998, 120, 4881-4882. (e) With, J.; Horton, A. D. Angew.
Chem., Intl. Ed. Engl. 1993, 32, 903-905. (f) Walsh, P. J.; Hollander, F.
J.; Bergman, R. G. Organometallics 1993, 12, 3705-3723. (g) Lee, S. Y.;
Bergman, R. G. J. Am. Chem. Soc. 1995, 117, 5877-5878.
(6) For example, see: (a) Chen, H.; Schlecht, S.; Semple, T. C.; Hartwig,
J. F. Science 2000, 287, 1995-1997. (b) Liu, F.; Pak, E. B.; Singh, B.;
Jensen, C. M.; Goldman, A. S. J. Am. Chem. Soc. 1999, 121, 4086-4087.
(c) Periana, R. A.; Taube, D. J.; Gamble, S.; Taube, H.; Satoh, T.; Fujii, H.
Science 1998, 280, 560-564.
2
There are no other resonances in the H{1H} NMR spectrum
in the regions for the Cp* ligand or the tert-butyl or anticlinal
methylene positions of the neopentyl ligand. This exclusive and
stereospecific deuterium incorporation is consistent with the 1,2-
cis addition of C-D across the MdCHR linkage of a neo-
pentylidene intermediate. The stereospecificity also indicates
that C-H(D) activation occurs from only one of the two possible
alkylidene isomers,14 namely, the isomer with anticlinal Cp*
t
and Bu groups.15 The C-H(D) activation step appears to be
irreversible since isolated 2-d12 does not convert to 2 after
heating in protiotetramethylsilane (70 °C, 40 h). However, 2
and 2-d12 do react further with tetramethylsilane under the
(7) There are several examples of intramolecular C-H activation by
alkylidene complexes: (a) Vilardo, J. S.; Lockwood, M. A.; Hanson, L.
G.; Clark, J. R.; Parkin, B. C.; Fanwick, P. E.; Rothwell, I. P. J. Chem.
Soc., Dalton Trans. 1997, 3353-3362 and references therein. (b) McDade,
C.; Green, J. C.; Bercaw, J. E. Organometallics 1982, 1, 1629-1634. (c)
Bulls, A. R.; Schaefer, W. P.; Serfas, M.; Bercaw, J. E. Organometallics
1987, 6, 1219-1226. (d) van Doorn, J. A.; van der Heijden, H.; Orpen, A.
G. Organometallics 1994, 13, 4271-4277. (e) Duncalf, D. J.; Harrison, R.
J.; McCamley, A.; Royan, B. W. J. Chem. Soc., Chem. Commun. 1995,
2421-2422.
(12) Debad, J. D.; Legzdins, P.; Batchelor, R. J.; Einstein, F. W. B.
Organometallics 1993, 12, 2094-2102.
(13) The relative stereochemistry of the methylene protons for the alkyl
ligands has been ascertained by 1H-1H NOE difference spectroscopy (see
Experimental Section).
(14) Recent theoretical calculations have shown that the ground-state
conformation of the model compound CpW(NO)dCH2 has a pyramidal
geometry similar to CpW(NO)(PMe3)(dCH2), with syn and anticlinal
orientations of the methylidene H’s: Smith, K. M.; Poli, R.; Legzdins, P.
Chem. Eur. J. 1999, 5, 1598-1608.
(8) Coles, M. P.; Gibson, V. C.; Clegg, W.; Elsegood, M. R. J.; Porrelli,
P. A. J. Chem. Soc., Chem. Commun. 1996, 1963-1964.
(9) van der Heijden, H.; Hessen, B. J. Chem. Soc., Chem. Commun. 1995,
145-146.
(10) Cheon, J.; Rogers, D. M.; Girolami, G. S. J. Am. Chem. Soc. 1997,
119, 6804-6813.
(15) If C-H activation was also occurring concomitantly from the other
t
isomer with synclinal Bu/Cp* geometry, then deuterium incorporation
(11) Tran, E.; Legzdins, P. J. Am. Chem. Soc. 1997, 119, 5071-5072.
would be observed at the anticlinal methylene position.

614 J. Am. Chem. Soc., Vol. 123, No. 4, 2001
Adams et al.
Table 1. Rate Constants for the Thermolysis of 1 and 1-d4
a
compd
solvent
temp (°C)
av kobs (s-1
)
1
1
1
1
1-d4
1
cyclohexane/ PMe3
cyclohexane/ PMe3
cyclohexane/ PMe3
cyclohexane/ PMe3
cyclohexane/ PMe3
benzene-d6
71
81
91
98
91
72
5.3(1) × 10-5
1.6(1) × 10-4
4.7(1) × 10-4
1.1(1) × 10-3
2.0(1) × 10-4
4.6(1) × 10-5
a Errors reported to 3σ.
ligands.14,18 In solution, 4a exhibits typical alkylidene NMR
resonances at δ 11.25 (3JHP ) 3.6 Hz) and 282.8 (1JCH ) 111
2
Hz, JCP ) 8.9 Hz) for the R-proton and the R-carbon of the
neopentylidene ligand, respectively.19 It also retains the anticlinal
t
orientation of the Bu and Cp* groups as indicated by NOE
enhancements between the alkylidene R-H atom and the methyl
H atoms of the Cp* ring.
Figure 1. ORTEP plot of the solid-state molecular structure of Cp*W-
(NO)dCHCMe3)(PMe3) (4a) with 50% probability ellipsoids. Selected
bond lengths (Å) and angles (deg): W(1)-C(14) ) 1.961(9), N(1)-
O(1) ) 1.22(2), W(1)-N(1)-O(1) ) 163.3(14), W(1)-C(14)-C(15)
) 144.7(11), and N(1)-W(1)-C(14)-C(15) ) -7(3).
The second alkylidene product 4b can be identified by
spectroscopic analysis of the final reaction mixture. It is the
t
geometric isomer of 4a with a synclinal orientation of the Bu
and Cp* groups, as indicated by the diagnostic resonances at δ
1
12.87 (3JHP ) 4.4 Hz) and δ 282.9 (2JCP ) 9.5 Hz) in the H
thermolysis conditions, albeit very slowly. For example, a trace
amount (∼3%) of a second product, Cp*W(NO)(CH2SiMe3)2
(3), is detectable in the mixture obtained from the thermolysis
of 1 in protiotetramethylsilane after 40 h (eq 1). The relative
amount of 3 gradually increases upon prolonged heating of this
mixture (to ∼8% after an additional 40 h), thereby indicating
that complex 3 is derived from 2 via the activation of a second
solvent molecule. The implications of these results are discussed
in due course (vide infra).
and 13C NMR spectra and NOE enhancements between the tBu
and Cp* moieties. The isomers 4a-b do not interconvert at
room temperature on the NMR time scale, but do interconvert
upon prolonged heating at elevated temperatures (vide infra)
presumably by rotation around the MdC bond.14,20
Kinetic Studies of the Formation of 4a-b. The conversion
of 1 to 4a-b can be conveniently monitored by UV-visible
spectroscopy in cyclohexane solutions containing an excess of
PMe3 (300 equiv). The reaction exhibits first-order kinetics for
the loss of 1 over the temperature range of 71-98 °C (Table
1). An Eyring plot of the data affords the activation parameters
∆H‡ ) 114(7) kJ mol-1 and ∆S‡ ) 3(12) J mol-1 K-1 (reported
errors are 3σ). Both parameters lie within the range reported
for other complexes that undergo rate-determining intramolecu-
lar R-H eliminations of alkane.7b,c A primary deuterium kinetic
isotope effect of 2.4(2) is observed when the PMe3-trapping
reaction is conducted with Cp*W(NO)(CD2CMe3)2 (1-d4) at 91
°C, thereby confirming that the rate-determining step involves
intramolecular cleavage of a methylene C-H(D) bond.21
Thermolysis of 1 in Benzene and Benzene-d6. Additional
mechanistic information can be gleaned from the thermolysis
of 1 in benzene and benzene-d6. In benzene, 1 cleanly forms
the known phenyl complex Cp*W(NO)(CH2CMe3)(C6H5) (5)12
Trapping of the Neopentylidene Intermediate. The inter-
mediacy of the neopentylidene complex A during the transfor-
mation of 1 to 2 is substantiated by its isolation in base-stabilized
form from the thermolysis of 1 in neat trimethylphosphine. Two
isomers of composition Cp*W(NO)(dCHCMe3)(PMe3) (4a-
b) occur in the final reaction mixture in ∼4:1 ratio (eq 2).
The major isomer 4a can be isolated in high yield (75%) as
yellow plates from ether, and its solid-state molecular structure
has been established by a single-crystal X-ray crystallographic
analysis. The resulting ORTEP plot is shown in Figure 1.
Notable metrical features of 4a are its short W(1)-C(14) bond
distance of 1.961(9) Å and the obtuse W-C(14)-C(15) bond
angle of 144.7(11)°, both of which are consistent with the
geometric parameters of neopentylidene ligands in other tungsten
complexes.16,17 The neopentylidene ligand is oriented anticlinal
to the Cp* group, and the N(1)-W(1)-C(14)-C(15) torsion
angle is -7(3)°. This configuration minimizes steric interactions
in the metal’s coordination sphere and maximizes the synergic
π-bonding between the π-donor alkylidene and π-acceptor NO
1
(∼95% as judged by H NMR spectroscopy) by activation of
an aryl sp2 C-H bond.22 The thermolysis of 1 in benzene-d6
affords free neopentane and Cp*W(NO)(CHDCMe3)(C6D5) (5-
(18) (a) Schilling, B. E. R.; Hoffman, R.; Faller, J. W. J. Am. Chem.
Soc. 1979, 101, 592-598. (b) Gibson, V. C. J. Chem. Soc., Dalton Trans.
1994, 1607-1618.
(19) Reference 16a; Chapter 4.
(20) Schofield, M. H.; Schrock, R. R.; Park, L. Y. Organometallics 1991,
10, 1844-1851.
(21) The deuterium isotope effect for several other rate-determining R-H
alkane elimination reactions lies in the range of 2-3: (a) Schrock, R. R.;
Fellmann, J. D. J. Am. Chem. Soc. 1978, 100, 3359-3370 (b) Malatesta,
V.; Ingold, K. U.; Schrock, R. R. J. Organomet. Chem. 1978, 152, C53-
C56. (c) Fryzuk, M. D.; Duval, P. B.; Mao, S. S. S. H.; Zaworotko, M. J.;
MacGillivray, L. R. J. Am. Chem. Soc. 1999, 121, 2478-2487. Also see
refs 7b,c.
(16) (a) Nugent, W. A.; Mayer, J. M. Metal-Ligand Multiple Bonds;
Wiley: New York, 1988; Chapter 3. (b) Labinger, J. A.; Winter, M. J. In
ComprehensiVe Organometallic Chemistry II; Abel, E. W., Stone, F. G.
A., Wilkinson, G., Eds.; Elsevier Science Ltd: New York, 1995; Vol. 5,
Chapter 5, pp 311-315.
(17) For the characterizations of the isostructural CpMo(NO)(PMe3)(d
CHCMe3) alkylidene complexes, see: Legzdins, P.; Rettig, S. J.; Veltheer,
J. E. Organometallics 1993, 12, 3575-3585.
(22) Like complex 2, complex 5 does react further with the solvent,
although the reaction does not proceed at a significant rate under the
thermolytic conditions employed for the conversion of 1 to 5. Previous
studies have shown that the thermolysis of pure 5 in benzene at 110 °C
generates Cp*W(NO)(C6H5)2, which decomposes thermally but can be
trapped and isolated as the PMe3 adduct in reasonable yield: Debad, J. D.
Ph.D. Dissertation, University of British Columbia, 1989.

C-H Bond ActiVation by W Alkylidene Complexes
J. Am. Chem. Soc., Vol. 123, No. 4, 2001 615
Scheme 2
Figure 2. ORTEP plot of the solid-state molecular structure of
Cp*(W(NO)(η2-cyclohexene)(PMe3) (6) with 50% probability ellip-
soids. Selected bond lengths (Å) and angles (deg): W(1)-C(14) )
2.246(9), W(1)-C(15) ) 2.193(9), C(14)-C(15) ) 1.447(12), N(1)-
W(1)-C(14) ) 105.8(3), N(1)-W(1)-C(15) ) 96.3(3), W(1)-N(1)-
O(1) ) 172.6(6), and N(1)-W(1)-C(14)-C(15) ) 79.8(5).
d6) with exclusive stereospecific deuterium incorporation at the
synclinal methylene position of the neopentyl ligand (eq 3). The
and isolated as pale yellow needles in moderate yield (38%)
by crystallization from ether/hexanes.
A single-crystal X-ray crystallographic analysis of 6 reveals
a solid-state molecular structure (Figure 2) that is similar to
those of the isoelectronic [CpRe(NO)(alkene)PPh3]+ cationic
complexes previously studied by Gladysz et al.23 The η2-
cyclohexene ligand possesses some metallacyclopropane char-
acter, as indicated by the C(14)-C(15) bond length of 1.447(12)
Å that is intermediate between a single and a double bond.24
The cyclohexene ring is pointing away from the Cp* ligand
and has a C(14)-C(15)-W(1)-N(1) torsion angle of 79.8(5)°.
This orientation is reflective of minimal ligand steric interactions
and maximum back-donation of electron density from the metal
to the alkene π* orbitals.14,18 In solution, the olefinic carbon
13C NMR resonances of 6 (i.e., δ 39.6 and 40.0) are diagnostic
of coordinated alkenes, being shifted upfield by ∼100 ppm as
compared to free cyclohexene.25 A NOE interaction between
the methyl H atoms of the Cp* ring and the olefinic H atoms
indicates that the solid-state molecular geometry is maintained
in solution.
Complex 6 could conceivably be formed by either associative
or dissociative mechanisms from 4a-b. However, the inde-
pendent thermolysis of pure 4a in cyclohexane (70 °C, 40 h)
merely results in equilibration with 4b, thus indicating that the
pathways to formation of 4a-b and 6 from 1 are different. It is
therefore likely that the C-H activation of cyclohexane by A
is in direct competition with trapping of A by PMe3 (Scheme
3). The resulting cyclohexyl intermediate undergoes a second
C-H activation at the â-H position of the cyclohexyl ring to
eliminate neopentane and form a 16e alkene intermediate, which
is subsequently trapped by residual PMe3 to form the 18e adduct
6. The competitive C-H activation by an unsaturated intermedi-
ate in the presence of a strong Lewis base is remarkable, yet it
has been observed before26 and is clearly aided by low
1
first-order rate constant obtained by H NMR spectroscopy at
72.0 °C is of the same magnitude as that obtained in cyclo-
hexane/PMe3 solutions under comparable conditions (Table 1).
This observation indicates that the presence of phosphine and
the nature of the solvent do not significantly affect the rate of
reaction, as expected for a process having an intramolecular
rate-determining step.
Taken together, the labeling, trapping, and kinetic results
verify that the C-H activation chemistry derived from 1
proceeds through two distinct steps: (1) formation of the
neopentylidene complex A via rate-determining intramolecular
R-H elimination of neopentane and (2) a 1,2-cis addition of
R-H across the MdCHR linkage of A. A simple two-step R-H
abstraction mechanism is thus proposed, as illustrated in Scheme
2 for the conversion of 1 to 2-d12. However, other more complex
mechanisms for the formation of A, such as synclinal R-H
elimination to a discrete metal hydride/alkyl/alkylidene inter-
mediate followed by neopentane reductive elimination,7b cannot
be ruled out at this time.
B. Activation of Cyclohexanes by Neopentylidene Complex
A. Activation and Functionalization of Cyclohexane. The
thermolysis of 1 in cyclohexane in the presence of a large excess
of PMe3 (i.e., >50 equiv) yields the alkylidene complexes 4a-b
as expected. However, as the relative amount of PMe3 is
decreased, the presence of a third product, namely, Cp*W(NO)-
(η2-cyclohexene)(PMe3) (6), can be detected in the H NMR
spectrum of the crude reaction mixture (eq 4). Complex 6 is
1
(23) Gladsyz, J. A.; Boone, B. J. Angew. Chem., Intl. Ed. Engl. 1997,
36, 550-583. (b) Kowalczyk, J. J.; Arif, A. M.; Gladysz, J. A. Chem. Ber.
1991, 124, 729-742.
(24) The strong π-donor ability of the Cp*W(NO)(PR3) fragment has
been recently demonstrated; see: Burkey, D. J.; Debad, J. D.; Legzdins, P.
J. Am. Chem. Soc. 1997, 119, 1139-1140.
(25) Silverstein, R. M.; Bassler, G. C.; Morrill, T. C. Spectrometric
Identification of Organic Compounds; Wiley and Sons: Toronto, 1991; p
238.
(26) (a) Buchanan, J. M.; Stryker, J. M.; Bergman, R. G. J. Am. Chem.
Soc. 1986, 108, 1537-1550. (b) McGhee, W. D.; Bergman, R. G. J. Am.
Chem. Soc. 1988, 110, 4246-4262.
formed preferentially under “dilute” conditions (i.e., e12 equiv
of PMe3 for 1.0 mL of a 0.020 M solution of 1) and can be
separated from 4a-b by chromatography on neutral alumina

616 J. Am. Chem. Soc., Vol. 123, No. 4, 2001
Adams et al.
Scheme 3
concentrations of the trapping agent relative to the cyclohexane
solvent. The presence of at least 1 equiv of PMe3 appears to be
necessary in order to obtain an isolable organometallic product
since the thermolysis of 1 in neat cyclohexane leads to a plethora
of intractable Cp*-containing compounds in the final reaction
mixture.
The overall transformation of 1 into 6 can be regarded as
involving a functionalization of cyclohexane to cyclohexene by
dehydrogenation. The spontaneous dehydrogenation of alkanes
to alkenes is not ubiquitous to all metal-based C-H activating
systems since the second â-H activation step requires electronic
and coordinative unsaturation at the metal center.2d The initial
C-H activation event in most systems regenerates a stable,
coordinatively saturated complex at which no further chemistry
can occur. In our system, however, the resulting C-H activation
product is a coordinatively unsaturated 16e species, thereby
facilitating multiple C-H activation events.
Activation and Functionalization of Methyl- and Ethyl-
cyclohexane. It is not unreasonable to expect that the ther-
molysis of 1 in methylcyclohexane under dilute conditions with
controlled amounts of PMe3 should likewise result in the
formation of PMe3-trapped alkene complexes in relative propor-
tions governed by the selectivity of A for primary, secondary,
and tertiary sp3 aliphatic C-H bonds. However, in addition to
trace amounts of 4a-b, the thermolysis of a dilute solution of
1 in methylcyclohexane (6 mM) in the presence of PMe3 (∼10
equiv) unexpectedly yields two products of solvent activation,
neither of which contains a phosphine ligand. Specifically, the
products are the allyl hydride complexes, Cp*W(NO)(η3-
C7H11)(H) (7a-b) in ∼9:1 ratio (eq 5).
Figure 3. ORTEP plot of the solid-state molecular structure of
Cp*W(NO)(η3-C7H11)(H) (7a) with 50% probability ellipsoids. Selected
bond lengths (Å) and angles (deg): W(1)-C(7) ) 2.234(11), W-C(1)
) 2.369(9), W(1)-C(2) ) 2.397(11), C(1)-C(2) ) 1.366(14), C(1)-
C(7) ) 1.39(2), C(2)-C(1)-C(7) ) 118.7(10), C(7)-C(1)-C(2)-
C(3) ) 172.9(11), and N(1)-W(1)-C(1)-C(6) ) -10.9(8).
nitrosyl ligand. The allyl ligand is distorted from a conventional
η3-endo conformation in a manner typical of chiral Cp′M(NO)L-
(allyl) (M ) Mo, W) complexes.27 It is canted toward the
nitrosyl ligand, as indicated by the N(1)-W(1)-C(1)-C(6)
dihedral angle of -10.9(8)°. Also, the exocyclic cis W-C bond
length (2.234(11) Å) is significantly shorter than the other two
W-C bonds (W-C(1) ) 2.369(9) Å, W-C(2) ) 2.397(11)
Å), while the C(1)-C(2) bond length (1.366(14) Å) is margin-
ally shorter than the C(1)-C(7) bond length (1.39(2) Å). Such
η3 f σ-π distortions are believed to be induced by the synergic
π-bonding between the mutually perpendicular π-systems of
the acceptor NO ligand and the donor π2 fragment of the allyl
ligand.18a The moderate “cis-exocyclic CH2” η3 f σ-π
distortion observed in this instance presumably places the more
electron-rich endocyclic π orbitals orthogonal to the NO ligand.
Spectroscopically, complex 7a exhibits a weak absorption at
1905 cm-1 in its Nujol mull IR spectrum in the range expected
1
for terminal M-H stretches28 and a hydridic signal in its H
NMR (C6D6) spectrum. Its 13C NMR spectrum in the same
solvent exhibits a high-field, sp3-like resonance for the exocyclic
allyl carbon (δ 41.3 (t, 1JCH ) 155 Hz)), and low-field, sp2-like
1
resonances for the endocyclic allyl carbons (δ 80.0 (d, JCH
)
148 Hz), 121.1 (s)), thereby indicating that the η3 f σ-π
distortion of this ligand is maintained in solution. The observa-
tion of NOE enhancements between the hydride ligand and the
allyl CH atom likewise indicates that the exocyclic CH2 moiety
remains cis to the NO ligand in solution.
The minor product of the thermolysis, 7b, can be partially
characterized by spectroscopic analysis of the final reaction
mixture. It exhibits signals in the 13C NMR spectrum consistent
with a hydride ligand and a η3 f σ-π distorted exocyclic allyl
ligand. The observation of an NOE enhancement between the
allyl CH2 group and the hydride ligand suggests that complex
7b is an isomer of 7a with a trans arrangement of exocyclic
Thermolysis of 1 in neat methylcyclohexane produces 7a-b
exclusively. The major product 7a can be isolated by crystal-
lization from ether/hexanes as yellow needles in reasonable yield
(57%). An X-ray crystallographic analysis reveals that 7a
possesses an exocyclic endo, cis -η3 allyl ligand (Figure 3). The
allyl carbon atoms are all within bonding distance of the metal
center, while the C-C bond distances are intermediate between
double and single bonds. The central allyl carbon is pointing
away from, rather than toward, the Cp* ligand, thereby
minimizing steric interactions between the cyclohexyl ring and
the Cp* ligand, while the exocyclic CH2 group is cis to the
(27) (a) Greenhough, T. J.; Legzdins, P.; Martin, D. T.; Trotter, J. Inorg.
Chem. 1979, 11, 3268-3270. (b) Adams, R. D.; Chodosh, D. F.; Faller, J.
W.; Rosan, A. M. J. Am. Chem. Soc. 1979, 101, 2570-2578. (c) Faller, J.
W.; Nguyen, J. T.; Ellis, W.; Mazzieri, M. R. Organometallics 1993, 12,
1434-1438. (d) Ipaktschi, J.; Mirzaei, F.; Demuth-Eberle, G.; Beck, J.;
Serafin, M. Organometallics 1997, 16, 3965-3972.
(28) Cotton, F. A.; Wilkinson, G.; Murillo, C. A.; Bochmann, M.
AdVanced Inorganic Chemistry, 6th ed.; Wiley & Sons: New York, 1998;
pp 79-80.

C-H Bond ActiVation by W Alkylidene Complexes
J. Am. Chem. Soc., Vol. 123, No. 4, 2001 617
chemical shifts for the allyl carbons of 8a in solution are more
reminiscent of a terminal η3 f σ-π distortion (δ 37.2 (t, 1JCH
) 152 Hz, allyl CH2), 93.2 (d, 1JCH ) 150 Hz, allyl CH), 109.0
(s, allyl C)), thereby suggesting that the η3 f η2 distortion may
just be a solid-state phenomenon.31
The 1H and 13C NMR spectroscopic properties of the minor
product, 8b, are consistent with it being a terminal allyl isomer
of 8a. NOE enhancements between the terminal CH2 group and
the hydride ligand indicate a trans arrangement of the allyl CH2
moiety and the nitrosyl ligand, and a small amount of spin
saturation transfer can be observed between the respective allyl
resonances of 8a and 8b. Given the steric preference for the
exo over the endo configuration in 8a, 8b is probably the exo,
trans isomer.
The mechanisms for the transformations of 1 into 7a-b and
8a-b likely mirror that of cyclohexane, where the activation
of a solvent C-H bond by A is followed by â-H activation to
release the second neopentyl ligand as neopentane and form a
16e alkene complex. Unlike the cyclohexene case, however,
the alkene complex evidently does not persist long enough to
be trapped by PMe3. Rather, there is a third γ-H activation by
the metal center to intramolecularly trap the alkene complex as
the 18e allyl hydride complex.32 Formally this represents a triple
dehydrogenation, where both the neopentylidene ligand and the
metal center function as hydrogen acceptors during the conver-
sion of the alkane to a coordinated allyl. Intriguingly, the allyl
ligands of related complexes [CpM(NO)(L)(η3-allyl)]+ (M )
Mo, W; L ) CO, X-, OH2) have been regioselectively, and in
some cases stereoselectively, transformed into a wide variety
of liberated, functionalized alkenes in high yields by reactions
with both nucleophiles and electrophiles.27b,c,33 It is therefore
conceivable that the allyl hydride complexes that are obtained
from C-H activation by A will react in the same manner,
thereby facilitating the functionalization of the alkanes into
complex alkene derivatives.
Selectivity of Alkane C-H Bond Activation by Neopen-
tylidene Complex A. The original purpose of the thermolysis
of 1 in methylcylohexane was to establish the intramolecular
selectivity of A for aliphatic C-H bonds. The formation of
exocyclic allyl hydride complexes in methylcyclohexane and
in ethylcyclohexane suggests that initial activation by A occurs
at the terminal primary methyl C-H bond. The preferential
activation of the stronger primary C-H bond would be
consistent with the bond selectivities exhibited by all other
metal-based systems operating via nonradical mechanisms under
both kinetic and thermodynamic control.2b,34 However, it is
Figure 4. ORTEP plot of the solid-state molecular structure of
Cp*W(NO(η3-C8H13)(H) (8a) with 50% probability ellipsoids. Selected
bond lengths (Å) and angles (deg): W(1)-C(11) ) 2.267(8), W-C(12)
) 2.275(7), W(1)-C(13) ) 2.476(7), C(11)-C(12) ) 1.390(11),
C(12)-C(13) ) 1.38(10), C(11)-C(12)-C(13) ) 125.7(7), and
C(11)-C(12)-C(13)-C(14) ) 173.3(7).
CH2 and NO groups. The exo or endo configuration of 7b could
not be conclusively established by NOE experiments, but given
the steric preference for the endo form in 7a, 7b is likely the
endo, trans isomer. A small amount of spin saturation transfer
is observed between the allyl CH atoms of 7a and 7b in the
NOE difference spectra, thus indicating that 7a and 7b inter-
convert on the NMR time scale at room temperature, likely by
rotation around the terminal C-C bond of an unobserved η1-
allyl intermediate.29
Similar results are obtained for the thermolysis of 1 in neat
ethylcyclohexane. Two isomeric allyl hydride complexes,
Cp*W(NO)(η3-C8H13)(H) (8a-b), are detectable in the final
thermolysis residue in ∼10:1 ratio (eq 6), and the major isomer
8a can be isolated as a yellow crystalline solid (37% yield).
An X-ray crystallographic analysis of 8a reveals an exocyclic
η3-allyl moiety in the exo configuration, again consistent with
the complex adopting the conformation that minimizes steric
interactions between the Cp* ligand and the cyclohexyl ring
(Figure 4). The allyl ligand is also canted toward the NO ligand
with the terminal CH2 moiety cis to the nitrosyl ligand.
Interestingly, the allyl ligand is distorted in a fashion alternative
to the σ-π distortion extant in 7a. Both the terminal and the
central allyl carbons have short W-C bonds (W-C(11) )
2.267(8) Å, W-C(12) ) 2.275(7) Å) as compared to the
endocyclic allyl carbon (W-C(13) ) 2.476(7) Å). Similar η3
f η2 distortions have been observed previously in other
complexes bearing asymmetric η3-allyl ligands, such as the
[TpMo(NO)(CO)(η3-allyl)]+ cations, and they have been at-
tributed to nonbonding steric effects.30 Indeed, the 13C NMR
(31) Spectroscopically, complex 8a also exhibits a weak terminal W-H
stretch at 1935 cm-1 in its IR spectrum and a hydride signal at δ - 0.92
1
(1JHW ) 125) in its H NMR spectrum.
(32) It is curious to note that similar stable allyl hydride complexes are
not observed as the final products in the thermolysis of 1 in neat
cyclohexane. There are two possible explanations. One is that the third
activation does not occur for the endocyclic cyclohexenyl complex, leading
to decomposition of the unsaturated 16e species unless trapped by PMe3.
The other explanation is that the allyl hydride complexes do form by
intramolecular trapping but are thermally unstable under the experimental
conditions, leading to the observed decomposition products. Attempts to
resolve the issue experimentally have so far been unsuccessful. Spectroscopic
analysis of the reaction mixtures from the thermolysis of 1 at lower
temperatures or at partial conversion do reveal trace amounts of resonances
that could be attributable to allyl hydride complexes, but decomposition
products dominate the spectra in either case.
(33) (a) Kocienski, P. J.; Brown, R. C. D.; Pommier, A.; Procter, M.;
Schmidt, B. J. Chem. Soc., Perkin Trans. 1 1998, 1, 9-40. (b) Faller, J.
W.; Chao, K. J. Am. Chem. Soc. 1983, 105, 3893-3898. (c) VanArsdale,
W. E.; Winter, R. E. K.; Kochi, J. K. Organometallics 1986, 5, 645-655.
(d) Chen, C. C.; Fan, J.; Shieh, S.; Lee, G.; Peng, S.; Wang, S.; Lui, R. J.
Am. Chem. Soc. 1996, 118, 9279-9287. (e) Faller, J. W.; Linebarrier, D.
L. Organometallics 1990, 9, 3182-3184.
(29) Reference 28; pp 24-25.
(30) (a) Villanueva, L. A.; Ward, Y. D.; Lachiocotte, R.; Liebeskind, L.
S. Organometallics 1996, 15, 4190-4200. (b) Ward, Y. D.; Villanueva, L.
A.; Allred, G. D.; Payne, S. C.; Semones, M. A.; Liebeskind, L. S.
Organometallics 1995, 14, 4132-4156.

618 J. Am. Chem. Soc., Vol. 123, No. 4, 2001
Adams et al.
possible that the complexes 7a-b and 8a-b arise from initial
activation at one of the other allyl positions, followed by
dehydrogenation at the methyl position. Alternatively, since the
C-H activation steps are all post-rate-determining, other allyl
hydride isomers could be derived from activation by A at
nonallyl endocyclic positions but undergo rapid isomerization
(e.g., by reversible γ-H activation) to the more thermodynami-
cally stable isomers 7a-b and 8a-b. In other words, the
observed product distributions may not be kinetically controlled
and therefore may not correlate to the intramolecular selectivity
of A for alkane C-H bonds. Since no scenario can be discounted
at present, the intramolecular C-H bond selectivity of A for
alkanes cannot conclusively be determined from the activation
of these substrates.
Scheme 4
C. Activation of Toluene by Neopentylidene Complex A.
The intramolecular C-H bond selectivity of A for arene C-H
bonds can, in theory, be deduced from the activation of toluene.
Specifically, the relative preference of A for aryl sp2 versus
benzylic sp3 C-H bonds should be reflected in the distribution
of aryl and benzylic activated products. Likewise, the regio-
selectivity of aryl activation (ortho, meta, para) should be
reflected in the relative distribution of aryl isomers. The ther-
molysis of 1 in toluene, however, yields a surprisingly complex
mixture of six products (eq 7). The two major products are the
(12). However, 12 disappears as the thermolysis progresses and
is replaced by the benzyl-containing products 10 and 11a-b.
This observation suggests that complex 12 is thermally unstable
under the experimental conditions employed and mediates the
formation of 10 and 11a-b. In support of this hypothesis is
the fact that thermolysis of independently prepared 12 in toluene
for 40 h at 70 °C yields 10 and 11a-b in the same relative
distributions as from 1 (i.e., 11a:11b:10 ) 63:32:5). Complex
11a can be isolated as a crystalline solid from this reaction
mixture in adequate yield (25%) and fully characterized, but
11b can only be further characterized spectroscopically in the
crude reaction mixture. Both 11a and 11b exhibit spectral
features typical of Cp′M(NO)(CH2C6H5)(aryl) (M ) Mo, W)
complexes, most notably η2-benzyl ligands characterized by the
ipso carbon resonances at ∼114 ppm in their 13C NMR
spectra.36,37
The transformation of 12 into 10 and 11a-b could proceed
via any of the mechanisms previously discussed for 1 (vide
supra), but the most likely mechanism is the R-H elimination
of neopentane from 12 to generate a C-H activating benzylidene
complex, B (Scheme 4).
neopentyl aryl complexes, namely, Cp*W(NO)(CH2CMe3)-
(C6H4-3-Me) (9a) and the previously reported Cp*W(NO)-
(CH2CMe3)(C6H4-4-Me)35 (9b) in approximately 47 and 33%
yields, respectively. Complexes 9a-b can be isolated in
crystalline form as a mixture (56% yield) and are readily
The proposed mechanism does have literature precedents. The
complex Cp2Ti(CH2C6H4-4-Me)2 has been obtained from the
thermolysis of Cp2Ti(CH2CMe3)2 in p-xylene, and the related
PMe3-trapped intermediate, Cp2Ti(dCHC6H5), has been ob-
served spectroscopically during the thermolysis of Cp2Ti-
(CH2CMe3)(CH2C6H5).9 The mechanism is also supported by
the following experimental observations. The thermolysis of 12
in benzene and benzene-d6 cleanly affords Cp*W(NO)(CH2C6H5)-
(C6H5) (13) and Cp*W(NO)(CHDC6H5)(C6D5) (13-d6), respec-
tively, in addition to free neopentane. Complex 13-d6 exhibits
stereospecific syn incorportation of deuterium at the methylene
carbon, consistent with concerted 1,2-cis addition of C-H across
the MdC linkage of an anticlinal benzylidene intermediate.
3
distinguished by the distinct aromatic JH-H coupling patterns
in the 1H NMR spectrum of the mixture. The other spectroscopic
features of 9a-b are unremarkable. The other four products
1
resulting from this thermolysis can be identified from the H
NMR spectrum of the final reaction mixture. One is the
previously reported aryl isomer of 9a-b, namely Cp*W(NO)-
(CH2CMe3)(C6H4-2-Me) (9c;12 ∼1%). The remaining three
products all arise from the incorporation of two molecules of
toluene; namely, Cp*W(NO)(CH2C6H5)(C6H4-3-Me) (11a;
∼12%), Cp*W(NO)(CH2C6H5)(C6H4-4-Me) (11b; ∼6%), and
the previously reported Cp*W(NO)(CH2C6H5)2 (10;36 ∼1%).
Evidence for the Formation of the C-H Activating
Benzylidene Complex Cp*W(NO)(dCHC6H5). To determine
the origin of products 9-11, the thermolysis of 1 in toluene at
70 °C was monitored by periodic 1H NMR spectroscopic
analysis of the reaction mixture. Over the course of the first 15
h, all four products of toluene C-H activation by A are observed
to form, specifically 9a-c and the product of activation at the
tolyl methyl position, namely, Cp*W(NO)(CH2CMe3)(CH2C6H5)
1
Monitoring the rate of reaction of 12 in benzene by H NMR
spectroscopy at 72 °C affords a first-order rate constant of the
same order of magnitude as 1 (i.e., k12 ) 4.9(1) × 10-5 s-1
;
k12/k1 ) 1.07), thus indicating a similarity in the energetics of
the rate-determining steps.
Finally, the thermolysis of 12 in THF in the presence of
excess PMe3 (∼130 equiv) results in the clean formation of
Cp*W(NO)(dCHC6H5)(PMe3) 14a-b in two rotameric forms
(∼5:1 ratio). The major isomer 14a is isolable as an orange
crystalline solid (52% yield) and has been subjected to an X-ray
crystallographic analysis (Figure 5). The solid-state molecular
structure of 14a is similar to that of 4a, with a short WdC(1)
(34) For a detailed discussion of the origin of this phenomenon, see:
Bryzndza, H. E.; Fong, L. K.; Paciello, R. A.; Tam, W.; Bercaw, J. E. J.
Am. Chem. Soc. 1987, 109, 1444-1456.
(35) Debad, J. D.; Legzdins, P.; Rettig, S. J.; Veltheer, J. E. Organo-
metallics 1993, 12, 2714-2725.
(36) Legzdins, P.; Jones, R. H.; Phillips, E. C.; Yee, V. C.; Trotter, J.;
Einstein, F. W. B. Organometallics 1991, 10, 986-1002.
(37) Dryden, N. H.; Legzdins; Trotter, J.; Yee, V. C. Organometallics
1991, 10, 2857-2870.

C-H Bond ActiVation by W Alkylidene Complexes
J. Am. Chem. Soc., Vol. 123, No. 4, 2001 619
Scheme 5
Figure 5. ORTEP plot of the solid-state molecular structure Cp*W-
(NO)(dCHC6H5)(PMe3) (14a) with 50% probability ellipsoids. Selected
bond lengths (Å) and angles (deg): W(1)-C(1) ) 1.975(3), N(1)-
O(1) ) 1.244(6), W(1)-N(1)-O(1) ) 178.1(3), W(1)-C(1)-C(2) )
137.2(2), W(1)-C(1)-C(2)-C(3) ) -171.3(3), and N(1)-W(1)-
C(1)-C(2) ) -8.3(4).
however, 9c does convert quantitatively to the same relative
distribution of aryl isomers 9a-c as observed during the
thermolysis of 1 in toluene (∼47:33:1). The thermolysis of 9c
in benzene-d6 at 70 °C reveals that the isomerization is a rapid
intramolecular process (t1/2 ) 18 min), with 9a and 9b forming
concomitantly to attain the equilibrium distribution after 120
min.40 Similar experiments with the third representative com-
plex, namely, the previously unobserved benzyl o-tolyl isomer,
Cp*W(NO)(CH2C6H5)(C6H4-2-Me) (11c), also reveal a con-
comitant intramolecular isomerization to a 2:1 mixture of 11a
and 11b (t1/2 ) 61 min in benzene-d6).41 The rapid isomeriza-
tions of both 9c and 11c indicate that the relative distributions
of aryl-activation isomers 9a-c and 11a-b do not represent
the regioselectivity of A and B for the ortho, meta and para
aryl bonds of toluene but instead reflect the relative thermo-
dynamic stabilities of the respective aryl complexes. An
important consequence of the isomerizations is that unobserved
intermediates must be present in solution to mediate the
migration of the metal carbon bond of the aryl ligand. Similar
cases of aryl isomerizations in other C-H activating systems
have been attributed to the reversible formation of π-3c or
σ-4darene complexes, or benzyne complexes,4a via intramolecular
H atom transfer. Various such intermediates could be formed
in this case by H-transfer from the alkyl and aryl ligands and/
or from a methyl group of the Cp* ligand. Investigations into
the mechanism of the isomerization process and its relation to
the C-H activation of arenes are currently underway. The
relationship between all of the complexes derived from the
activation of toluene from 1 is thus shown in Scheme 5.
Fortuitously, the lack of interconversion of the respective aryl
and benzylic C-H activation products under thermolytic
conditions indicates that the aryl and benzylic C-H bond
activation pathways from A and B are effectively independent
and irreversible. In other words, the observed aryl versus
benzylic product distributions do in fact reflect the intramo-
lecular kinetic selectivities of A and B for the aryl sp2 and
benzylic sp3 C-H bonds in toluene. The respective selectivities
of A and B can also be independently quantified since the ratio
of the products 9a-c to 10 and 11a-b represents the aryl versus
benzylic kinetic selectivity of A, while the ratio of 10 to 11a-b
represents that of B.42 The ratios of 4.3(3):1 for A and 19(2):1
for B clearly indicate a preferences for activation of the stronger
aryl sp2 C-H bond. Thus, both alkylidene intermediates do, in
bond of 1.975(3) Å and a nearly planar N(1)-W(1)-C(1)-
C(2) grouping (-8.3(4)°). An additional structural feature is
the planar alignment of the phenyl and WdC moieties (W(1)-
C(1)-C(2)-C(3) ) 171.3(3)°), which is indicative of conjuga-
tion of the alkylidene and phenyl π-systems.38 In solution, 14a
exhibits diagnostic high-field NMR resonances for the ben-
zylidene protons (δ 12.01, 1JHP ) 3.8 Hz) and the benzylidene
carbon (δ 261.2, 2JCP ) 9 Hz). The NOE enhancements between
the Cp* ring and the benzylidene H atom reveal that the
anticlinal orientation of the Cp* and Ph groups is maintained
in solution. Not surprisingly, the spectroscopic features of the
minor isomer 14b are consistent with it being the synclinal
geometric isomer. The trapping reaction appears to be irrevers-
ible since heating pure 14a in C6D6 (70 °C, 24 h) results in
intramolecular isomerization to 14b, with no evidence for the
formation of 13-d6.
Selectivity of Aryl versus Benzylic C-H Bond Activation
by Complexes A and B. Monitoring the thermolysis of 1 in
toluene for an additional 24 h past the end point of typical
thermolysis experiments (64 h total) reveals no significant
1
changes in the H NMR spectrum.39 This indicates that, once
formed, the neopentyl-containing products 9a-c do not convert
to the benzyl-containing products 10 and 11a-b (e.g., via
isomerization to 12). In addition, the relative ratios of aryl
isomers 9a-c and 11a-b remain constant during their formation
from 1 and 12, respectively. These results suggest that the
product distributions observed after the consumption of 12 (i.e.,
at 40 h) may indeed be a kinetic distribution of products and
thus correlate to the intramolecular C-H bond selectivities of
A and B.
To test this hypothesis, three representative complexes were
synthesized by metathetical methods and thermolyzed inde-
pendently in toluene at 70 °C for 24 h. As anticipated, the bis-
(benzyl) complex 10 is thermally robust and does not convert
to 11a-b. Likewise, the neopentyl o-tolyl complex 9c does not
convert to the benzyl complexes 10 or 11a-b. Surprisingly,
(38) For other examples of isostructural benzylidene complexes, see: (a)
Chan, M. C. W.; Cole, J. M.; Gibson, V. C.; Howard, J. A. K.; Lehmann,
C.; Poole, A. D.; Siemeling, U. J. Chem. Soc., Dalton Trans. 1998, 103-
111. (b) Mashima, K.; Tanaka, Y.; Kaidzu, M.; Nakamura, A. Organo-
metallics 1996, 15, 2431-2433. (c) Kiel, W. A.; Lin, G.; Constable, A.
G.; McCormick, F. B.; Strouse, C. E.; Eisenstein, O.; Gladysz, J. A. J. Am.
Chem. Soc. 1982, 104, 4865-4878.
(40) No resonances attributable to free toluene, 10, or 11a-b are
observed, even after prolonged heating for 40 h. Rather, 9a-c slowly
decompose to release free neopentane, presumably via reaction with benzene
in the same manner as 5.
(41) No resonances attributable to free toluene or 10 are observed, even
after prolonged heating for 40 h. Rather, 11a-b slowly decompose to a
plethora of unidentified products.
(39) A trace amount of unidentified decomposition products is observed
at 64 h and the amount gradually increases upon prolonged heating.

620 J. Am. Chem. Soc., Vol. 123, No. 4, 2001
Adams et al.
fact, possess the same qualitative bond selectivity as the majority
of reported metal-based C-H activation systems.2b,43 Further-
more, this fact supports the inference that the activation and
functionalization of alkanes such as methylcyclohexane does
indeed proceed by initial C-H activation of the stronger primary
C-H bond.
activation by the alkylidene complexes A and B that reveal them
to be versatile and useful additions to this area of synthetic
chemistry. First, the activation of C-H bonds occurs via the
1,2-R-H addition to the WdC double bond of one alkylidene
isomer, with a modest kinetic preference for the strongest C-H
bond. Second, spontaneous functionalization can be effected for
alkane substrates due to the inherent unsaturated, 16e nature of
the resulting C-H activation products. In particular, alkyl-
cyclohexanes are converted to allyl ligands that have the
potential for further elaboration and liberation as functionalized
alkenes. Finally, the C-H activating ability of Cp*W(NO)-
(dCHR) intermediates appears to be quite general as alkylidenes
A (R ) tBu), B (R ) Ph), and possibly C (R ) SiMe3) activate
alkane and arene C-H bonds. Presumably then, R-H elimination
of neopentane by the thermolysis of other congeners of the large
family of available Cp*W(NO)(CH2CMe3)(CH2R) complexes
can provide a whole host of alkylidene complexes for exploita-
tion as alkane functionalization reagents. Our studies with these
and related alkylidene complexes designed to address the
features outlined above are currently in progress.
The relative intramolecular kinetic selectivities of A and B
can be readily compared to those of other representative
complexes that activate toluene by alternative mechanisms under
kinetic conditions. Both alkylidene complexes exhibit selectivi-
ties that are higher than that exhibited by Tp*Rh(PMe3) (3.5:1
at 25 °C) for oxidative addition of tolyl C-H bonds.44 They
bracket that of Cp*2ScMe (6.1:1 at 80 °C) for σ-bond metathesis
of tolyl C-H4d but are considerably less than that exhibited by
Cp*Rh(PMe3) (>99:1 at -45 °C)45 and Cp*Ir(PMe3) (>99:1
at 25 °C)46 that operate by oxidative-addition mechanisms.
Similarly large (99:1 or greater) kinetic selectivities have been
reported for the activation of toluene by 1,2 C-H addition to
(HNR)2ZrdNR (R ) tBu3Si) at 97 °C5b and by C-H addition
to photochemically activated TpRe(O)(I)(Cl) at 25 °C.47 Thus,
it appears that the intramolecular kinetic selectivities of alkyl-
idene complexes A and B for arene C-H bonds are modest at
best. In comparison to each other, however, B appears to be
significantly more selective than A, thereby indicating that the
alkylidene R substituent can influence the C-H activation
process.
D. Activation of Aliphatic C-H Bonds by Benzylidene
Complex B. The C-H activation ability of the benzylidene
complex B encompasses aliphatic substrates as well. Thus, the
thermolysis of 12 in neat tetramethylsilane at 70 °C for 40 h
cleanly yields the previously reported Cp*W(NO)(CH2C6H5)-
(CH2SiMe3) (15;37 ∼95% as judged by 1H NMR spectroscopy)
via primary sp3 C-H bond activation. Likewise, the thermolysis
of 12 in cyclohexane in the presence of PMe3 (∼10 equiv)
generates 6 via secondary sp3 C-H bond activation, in addition
to the alkylidene complexes 14a-b. Finally, the thermolysis
of 12 in methylcyclohexane and ethylcyclohexane forms the
same distributions of allyl hydride complexes 7a-b and 8a-b
as observed during the corresponding thermolyses of 1. These
results clearly indicate that B is as effective as A in activating
the various types of C-H bonds of both alkane and arene
substrates. This is in sharp contrast to the related Cp2TidNR
system, where the replacement of an alkyl R substituent with
an aryl substituent renders the complex unreactive toward
hydrocarbons.5f The confirmed activation of alkanes by B also
supports the notion that a third alkylidene complex, namely,
Cp*W(NO)(dCHSiMe3), likely mediates the observed slow
conversion of Cp*W(NO)(CH2CMe3)(CH2SiMe3) (2) to Cp*W-
(NO)(CH2SiMe3)2 (3) by aliphatic C-H bond activation of
tetramethysilane (eq 1).
Experimental Section
General Methods. All reactions and subsequent manipulations
involving organometallic reagents were performed under anaerobic and
anhydrous conditions under either high vacuum or an atmosphere of
prepurified argon or dinitrogen. General procedures routinely employed
in these laboratories have been described in detail elsewhere.48 Unless
otherwise specified, all reagents were purchased from commercial
suppliers and used as received. Hydrocarbon solvents and trimethyl-
phosphine were freshly distilled or vacuum transferred from sodium
or sodium/benzophenone ketyl. Tetrahydrofuran was distilled from
molten potassium. Tetramethylsilane-d12 (99.8%, CDN Isotopes),
benzene-d6, and dioxane-d8 were dried over sodium and vacuum-
transferred as required. Methylene chloride-d2 and chloroform-d1 were
dried over activated 4-Å molecular sieves and vacuum-transferred as
required. The R2Mg‚X(dioxane) (R ) CX2CMe3 (X ) H, D), CH2C6H5,
C6H4-2-Me) alkylating reagents17,49 and the complexes Cp*W(NO)-
(R)(R′) (R) CH2CMe3, R′ ) CH2CMe3, CH2C6H5, C6H4-2-Me; R )
R′ ) CD2CMe3;2,35 R ) CH2C6H5, R′ ) C6H4-2-Me)37 were prepared
according to published procedures. Pertinent synthetic details and
characterization data for the new compounds Cp*W(NO)(CH2C6H5)-
(C6H4-2-Me) (11c) and Cp*W(NO)(CH2C6H5)(CH2CMe3) (12) are
reported below.
All IR samples were prepared as Nujol mulls sandwiched between
NaCl plates. NMR spectra were recorded at room temperature on Bruker
AC 200, Varian XL 300, Bruker WH-400 or Bruker AMX 500
instruments. All chemical shifts are reported in ppm and all coupling
1
constants are reported in hertz. H NMR spectra are referenced to the
residual protioisotopomer present in a particular solvent. 13C NMR
spectra are referenced to the natural-abundance carbon signal of the
solvent employed. 2H and 31P NMR spectra are referenced to external
C6H5D (7.15 ppm) and P(OMe)3 in C6D6 (141.0 ppm), respectively.
1
1
1
Where appropriate, H-1H COSY, H-1H NOEDS, H-13C HMQC,
1H-13C HMBC, 13C APT, homonuclear decoupling, and gated 13C{1H}
experiments were carried out to correlate and assign 1H and 13C signals.
Epilogue
Our initial exploration of the thermal chemistry of 1 and 12
has uncovered several important characteristics of C-H bond
Thermolyses of 1 and 12: The preparation of 2-11 and 13-15
by thermolyses of 1 or 12 was performed in a similar manner unless
otherwise stated. The thermolysis of 1 in tetramethylsilane is described
below as a representative example. Major products were isolated by
crystallization from the final thermolysis mixture using appropriate
solvents. The reported yields are not optimized. Minor products that
could not be isolated were characterized spectroscopically from the
final reaction mixture. Average product ratios were obtained by
(42) Calculation for A: (9a-c):(10 and 11a-b) ) (46 + 34 + 1):(19)
or 4.3(3):1. Calculation for B: (11a-b):10 ) 63 + 32:5 or 19(2), with an
error of 7% in the relative integrations of aryl and benzylic products.
(43) For a detailed discussion on the origin of this kinetic selectivity for
C-H activation by a 1,2 concerted R-H addition process, see refs 5b,c.
(44) Jones, W. D.; Hessell, E. T. J. Am. Chem. Soc. 1993, 115, 554-
562.
(45) Jones, W. D.; Feher, F. J. J. Am. Chem. Soc. 1984, 106, 1650-
1663.
(46) Burger, P.; Bergman, R. G. J. Am. Chem. Soc. 1993, 115, 10462-
10463.
(47) Brown, S. N.; Myers, A. W.; Fulton, J. R.; Mayer, J. M.
Organometallics 1998, 17, 3364-3374.
(48) Legzdins, P.; Rettig, S. J.; Ross, K. J.; Batchelor, R. J.; Einstein, F.
W. B. Organometallics 1995, 14, 5579.
(49) Dryden, N. H.; Legzdins, P.; Rettig, S. J.; Veltheer, J. E. Organo-
metallics 1992, 11, 2583-2590.

C-H Bond ActiVation by W Alkylidene Complexes
J. Am. Chem. Soc., Vol. 123, No. 4, 2001 621
integration of appropriate signals in the 1H NMR spectrum of the final
reaction mixture. Typically, the standard deviations in the averages were
<7%.
1.26 (s, 9H, CMe3), 1.54 (s, 15H, C5Me5), 4.37 (s, 1H, CHD); 2H{1H}
NMR (77 MHz, C6H6) δ -1.95 (br s, CHD), 7.26, 7.55, 7.79 (br m,
Ar D). Anal. Calcd for C21H25D6NOW:50 C, 50.12; H/D, 6.22, N, 2.82.
Found: C, 50.17; H/D, 6.32; N, 2.78.
Cp*W(NO)(CH2CMe3)(CH2SiMe3) (2). In the glovebox, a thick-
walled reaction bomb was charged with 1 (48 mg, 0.098 mmol) and a
stir bar. On a vacuum line, tetramethylsilane (∼4 mL) was added via
vacuum transfer at -196 °C. The resulting ruby-red solution was
allowed to warm to room temperature and then was placed in an oil
bath set at 70(2) °C for 40 h. During this time, the solution darkened
to wine-red. The organic volatiles were removed under vacuum, and
the residual purple-red solid was dissolved in C6D6 and analyzed by
1H NMR spectroscopy. The NMR solvent was removed in vacuo, and
the residue was dissolved in pentane and filtered through Celite (1 ×
0.7 cm) supported on a frit. The filtrate was reduced in volume and
stored at -30 °C for several days to induce the deposition of 2 (45
mg, 90%) as maroon crystals. The characterization data for this complex
Cp*W(NO)(cyclohexene)(PMe3) (6). Complex 6 was prepared by
the thermolysis of 1 (81 mg, 0.16 mmol) in cyclohexane (8 mL) and
PMe3 (∼1.6 mmol, ∼10 equiv). It was isolated as pale yellow needles
(32 mg, 38% yield) by extraction into 4:1 Et2O/hexanes and filtration
through neutral alumina I (1 × 0.7 cm): IR (cm-1) 1518 (s, νNO); MS
(LREI, m/z, probe temperature 150 °C) 507 [P+, 184W], 425 [ P+-
C6H10]; 1H NMR (500 MHz, C6D6) δ 0.8 (m, 1H, R′ CH), 1.22 (d, 2JHP
) 9H, PMe3), 1.5 (m, 1H, R CH), 1.6 (m, 2H, γ′ CHH), 1.67 (s, 15H,
C5Me5), 1.8 (m, 2H, γ CHH), 2.4 (m, 1H, â′ CHH), 2.8 (m, 2H, â
CHH), 3.1 (m, 1H, â′ CHH); 13C{1H} NMR (125 MHz, C6D6) δ 10.3
(C5Me5), 17.4 (d, 2JCP ) 29, PMe3), 24.2 (γ CH2), 25.3 (γ′ CH2), 28.3
(â CH2), 30.8 (d, 3JCP ) 4, â′ CH2), 39.6 (1JCW ) 39, R CH), 40.0 (d,
2JCP ) 12, R′ CH), 103.0 (C5Me5); 31P{1H} NMR (202 MHz, C6D6) δ
1
match that previously reported:35 IR (cm-1) 1551 (s, νNO); H NMR
2
4
1
(400 MHz, C6D6) δ -2.10 (dd, JHH ) 12.8, JHH ) 1.9, 1H,
CHsynHCMe3), -1.35 (dd, 2JHH ) 12.0, 4JHH ) 2.2, 1H, CHsynHSiMe3),
0.40 (s, 9H, SiMe3), 1.08 (d. 2JHH ) 12.0, 1H, CHantiHSiMe3), 1.35 (s,
-13.2 (s, JPW ) 373). Anal. Calcd for C19H34NOPW: C, 44.98; H,
6.76; N, 2.76. Found: C, 44.64; H, 6.77; N, 2.59.
Cp*W(NO)(η3-C7H11)(H) (7a-b). Compounds 7a-b were prepared
from the thermolysis of 1 in methylcyclohexane. Complex 7a was
isolated as yellow needles (28 mg, 70%) by crystallization from pentane.
7a: IR (cm-1) 1905 (w, νWH), 1564 (s, νNO); MS (LREI, m/z, probe
2
9H, CMe3), 1.51 (s, 15H, C5Me5), 3.25 (d, JHH ) 12.8, 1H,
CHantiHCMe3); NOEDS (400 MHz, C6D6) δ irrad at -2.10 ppm, NOEs
at δ -1.35, 3.25; irrad at -1.35, NOEs at δ -2.10, 1.08; irrad at 0.40
ppm, NOE at δ -1.35; irrad at 1.08 ppm, NOE at δ -1.35; irrad at
1.35 ppm, NOEs at δ -2.10, 0.40, 3.25; irrad at 3.25 ppm, NOEs at δ
-2.10, 1.35.
1
temperature 150 °C) 445 [P+, 184W]; H NMR (400 MHz, C6D6) δ
1
-0.73 (s, JHW ) 123, 1H, WH), 0.31 (br s, 1H, allyl CHsynH), 1.39
(m, 1H, CbHH), 1.54 (m, 1H, CaHH), 1.71 (m, 1H, CbHH), 1.76 (s,
Cp*W(NO)(CHDCMe3)(CD2Si(CD3)3 (2-d12). Complex 2-d12 was
prepared and analyzed by thermolysis of 1 in tetramethylsilane-d12 in
a J. Young NMR tube. It was subsequently isolated as mauve crystals
by crystallization from pentane (27 mg, 93%):IR (cm-1) 1554 (s, νNO);
15H, C5Me5), 1.84 (m, 1H, CaHH), 2.28 (br s, 1H, allyl CH,), 2.60 (m,
2
1H, allyl CHantiH), 2.68 (m, 2H, CcHH, CdHH), 2.91 (q, 1H, JHH
)
7.9, CcHH), 2.97 (m, 1H, CdHH); 13C{1H} (75 MHz, C6D6) δ 10.7
1
1
(C5Me5), 22.1 (t, JCH ) 124, CaH2), 22.5 (t, JCH ) 126, CbH2), 28.8
1
MS (LREI, m/z, probe temperature 150 °C) 519 [P+, 184W]; H NMR
(t, 1JCH ) 125, CcH2), 29.6 (t, 1JCH ) 126, CdH2), 41.3 (t, 1JCH ) 155,
1
(300 MHz, C6D6) δ 1.36 (s, 9H, CMe3), 1.51 (s, 15H, C5Me5), 3.19
allyl CH2), 80.0 (d, JCH ) 148, allyl CH), 104.9 (C5Me5), 121.1 (s,
2
(br s, 1H, CHantiD). H{1H} NMR (77 MHz, C6H6) δ -2.16 (s, 1D,
allyl C). NOEDS (400 MHz, C6D6) δ irrad at -0.73, NOEs at 2.28,
2.97; irrad at 0.31, NOEs at 2.28, 2.68; irrad at 2.28, NOEs at -0.73,
0.31, 2.97; irrad at 2.68, NOE at 0.31. Anal. Calcd. for C17H27NOW:
C, 45.86; H, 6.11; N, 3.15. Found: C, 45.99; H, 6.19; N, 2.97.
7b: 1H NMR (400 MHz, C6D6) δ -0.73 (s, 1H, WH), 0.48 (br s,
1H, allyl CHH), 1.75 (s, 15H, C5Me5), 2.61 (br s, 1H, allyl CH), 3.95
CHDsynCMe3), -1.38 (s, 1D, CDsynDSi Me3), 0.33 (s, 9D, Si(CD3)3),
1.06 (s, 1D, CDantiDSiMe3). Anal. Calcd for C19H25D12NOSiW:50 C,
43.94; H/D, 7.13; N, 2.70. Found: C, 44.26; H/D, 7.09; N, 2.74.
Cp*W(NO)(dCHCMe3)(PMe3) (4a-b). Complexes 4a-b were
prepared by the thermolysis of 1 (172 mg, 0.350 mmol) in tri-
methylphosphine (2 mL). Complex 4a was isolated as platelike yellow
crystals (130 mg, 75% yield) by crystallization from Et2O.
(m, 1H, allyl CHH), other resonances obscured; 13C{1H} (75 MHz,
1
C6D6) δ 10.3 (C5Me5), 22.5, 23.0, 27.3, 29.6 (CH2), 49.7 (t, JCH
)
4a: IR (cm-1) 1513 (s, νNO); MS (LREI, m/z, probe temperature
1
155, allyl CH2), 62.8 (d, JCH ) 146, allyl CH), 104.7 (C5Me5). The
13C resonance attributable to allyl C was not observed. NOEDS (400
MHz, C6D6) δ irrad at 3.95, NOEs at -0.73, 0.48.
200 °C) 495 [P+, 184W]; 1H NMR (400 MHz, C6D6) δ 1.06 (d, 2JHP
)
9.0 Hz, 9H, PMe3), 1.43 (s, 9H, CMe3), 1.87 (s, 15H, C5Me5), 11.25
3
(d, JHP ) 3.6 Hz, 1H, WdCH); 13C NMR (75, MHz, C6D6) δ 11.0
Cp*W(NO)(η3-C8H13)(H) (8a-b). Compounds 8a-b were prepared
from the thermolysis of 1 in ethylcyclohexane. Complex 8a was isolated
as yellow needles (21 mg, 37%) by crystallization from pentane.
8a: IR (cm-1) 1935 (m, νWH), 1575 (s, νNO); MS (LREI, m/z, probe
1
(C5Me5), 19.7 (d, JCP ) 32.4 Hz, PMe3), 32.0 (CMe3), 52.2 (CMe3),
106.4 (C5Me5), 282.8 (d, 2JCP ) 8.9 Hz, WdCH); 31P NMR (121 MHz,
1
C6D6) δ -7.9 (s, JPW ) 446 Hz). NOEDS (400 MHz, C6D6) δ irrad
temperature 150 °C) 457 [P+-H2, 184W]; H NMR (400 MHz, C6D6)
at 1.06, NOE at 11.25; irrad at 1.43; NOEs at 1.06, 11.25; irrad at
1.87, NOE at 11.25. Anal. Calcd for C18H34NOPW: C, 43.65; H, 6.92;
N, 2.83. Found: C, 43.62; H, 7.00; N, 2.78.
1
δ -0.92 (s, 1JHW ) 125, 1H, WH), 1.42 (m, 1H, CbHH), 1.58 (m, 3H,
CaHH, CcHH), 1.71 (s, 15H, C5Me5), 1.79 (m, 1H, CbHH), 1.90 (m,
1H, CdHH), 2.00 (m, 1H, CcHH), 2.15 (m, 3H, CdHH, allyl CHsynH,
CeHH), 2.44 (dd, 1H, 2JHH ) 3.1, 3JHH ) 13.3, allyl CHantiH), 2.51 (m,
1H, CeHH), 2.66 (dd, 3JHH ) 7.8, 3JHH ) 13.3, 1H, allyl CH); 13C{1H}
4b: 1H NMR (400 MHz, C6D6) δ 1.14 (d, 2JHP ) 9.4 Hz, 9H, PMe3),
3
1.15 (s, 9H, CMe3), 1.97 (s, 15H, C5Me5), 12.87 (d, JHP ) 4.4 Hz,
1H, WdCH); 13C NMR (75, MHz, C6D6) δ 11.5 (C5Me5), 20.2 (d,
1JCP ) 31.2 Hz, PMe3), 31.9 (CMe3), 48.3 (CMe3), 106.6 (C5Me5), 282.9
1
(75 MHz, C6D6) δ 10.7 (C5Me5), 27.1 (t, JCH ) 132, CH2), 30.6 (t,
2
(d, JCP ) 9.5 Hz,WdCH); 31P NMR (121 MHz, C6D6) δ - 4.0 (s).
1JCH ) 128, CH2), 32.6 (t, 1JCH ) 130, CH2), 35.1 (t, 1JCH ) 126, CH2),
37.2 (t, 1JCH ) 152, allyl CH2), 41.0 (t, 1JCH ) 129, CH2), 93.2 (d, 1JCH
) 150, allyl CH), 103.8 (C5Me5), 109.0 (allyl C). NOEDS (400 MHz,
C6D6) δ irrad at -0.92 ppm, NOEs at 2.00, 2.15, 2.51; irrad at 2.44
ppm, NOE at 2.15; irrad at 2.66 ppm, NOE at 2.15. Anal. Calcd for
C18H29NOW: C, 47.07; H, 6.36; N, 3.05. Found: C, 46.86; H, 6.42;
N, 2.96.
NOEDS (400 MHz, C6D6) δ irrad at 1.14, NOE at 12.87; irrad at 1.97,
NOE at 1.15.
Cp*W(NO)(CH2CMe3)(C6H5) (5) and Cp*W(NO)(CHDCMe3)-
(C6D5) (5-d6). Compounds 5 and 5-d6 were prepared from the
thermolysis of 1 in benzene and benzene-d6, respectively, and isolated
as wine-red rods by recrystallization from pentane. The characterization
data for complex 5 matched that previously reported.12
8b: 1H NMR (400 MHz, C6D6) δ -0.71 (s, 1JWH ) 126, 1H, WH),
1
2
5: 22 mg (33%); H NMR (300 MHz, C6D6) δ -2.05 (d, JHH
)
3
2
0.60 (dd, JHH ) 11.7, JHH ) 3.5, 1H, allyl CHantiH), 1.76 (s, 15H,
11.4, 1H, CHsynH), 1.26 (s, 9H, CMe3), 1.53 (s, 15H, C5Me5), 4.50 (d,
C5Me5), 2.87 (dd, 2JHH ) 3.5, 3JHH ) 7.4, 1H, allyl CHsynH), 4.69 (dd,
2JHH ) 11.4, 1H, CHantiH), 7.10 (m, 3H, Ar H), 7.70 (d, JHH ) 6.0,
3
3JHH )11.7, JHH ) 7.4, 1H, allyl CH), other resonances obscured;
3
2H, Ar H).
13C{1H} (75 MHz, C6D6) δ 10.5 (C5Me5), 27.0 (CH2), 30.5 (CH2), 32.5
(CH2), 33.0 (CH2), 40.9 (allyl CH2), 42.8 (CH2), 100.9 (d, 1JCH ) 151,
allyl CH), 104.8 (C5Me5). The 13C resonance attributable to allyl C
was not observed. NOEDS (400 MHz, C6D6) δ irrad at 2.87, NOEs at
-0.71, 0.60, 4.69; irrad at 0.60 ppm, NOE at 2.87; irrad at 4.69 ppm,
NOE at 2.87.
5-d6: 46 mg (65%); IR (cm-1) 1549 (s, νNO); MS (LREI, m/z, probe
temperature 150 °C) 503 [P+, 184W]; H NMR (400 MHz, C6D6) δ
1
(50) Since the detection method used in the elemental analysis cannot
distinguish between D2O and H2O, H/D abundances were calculated using
1 D ) 1 H.

622 J. Am. Chem. Soc., Vol. 123, No. 4, 2001
Adams et al.
2
2
Cp*W(NO)(CH2CMe3)(C6H4-3-Me) (9a) and Cp*W(NO)(CH2-
CMe3)(C6H4-4-Me) (9b). Complexes 9a and 9b were prepared by
thermolysis of 1 in toluene and cocrystallized as mauve microcrystals
(35 mg, 56%) by recrystallization from pentane. The characterization
data for 9b match that previously reported.35 IR (cm-1) 1549 (s, νNO);
2.20 (d, JHH ) 13.4, 1H, CHantiHCMe3), 2.22 (d, JHH ) 8.5, 1H,
CHsynHPh), 2.98 (d, 2JHH ) 8.5, 1H, CHantiHPh), 6.89 (m, 3H, Bzl Hortho
,
3
H
meta), 7.31 (t, JHH ) 7.3, 1H, Bzl Hpara); 13C{1H} NMR (50 MHz,
CDCl3) δ 10.4 (C5Me5), 33.8 (CMe3), 38.4 (CMe3), 51.7 (CH2), 64.0
(CH2) 109.2 (C5Me5), 113.3 (Bzl Cipso), 127.3, 129.2, 131.8 (Bzl Caryl).
Anal. Calcd for C22H33NOW: C, 51.67; H, 6.50, N, 2.74. Found: C,
51.74; H, 6.50; N, 2.82.
1
MS (LREI, m/z, probe temperature 150 °C) 511 [P+, 184W]; H NMR
(400 MHz, C6D6). 9a: δ -1.91 (d, 2JHH ) 11.7, 1H, CHsynH), 1.26 (s,
2
9H, CMe3), 1.56 (s, 15H, C5Me5), 2.17 (s, 3H, Tol Me), 4.44 (d, JHH
Cp*W(NO)(CH2C6H5)(C6H5) (13) and Cp*W(NO)(CHDC6H5)-
(C6H5) (13-d6). Compounds 13 and 13-d6 were prepared via the
thermolysis of 12 in benzene and benzene-d6, respectively, and were
isolated as red needles by crystallization from 4:1 Et2O/hexanes.
13: 46 mg (59%); IR (cm-1) 1562 (s, νNO); MS (LREI, m/z, probe
) 11.7, 1H, CHantiH), 6.93 (d, 3JHH ) 7.4, 1H, Tol H), 7.16 (t, 3JHH
)
3
7.6, 1H, Tol Hmeta), 7.45 (d, JHH ) 7.5, 1H, Ar H), 7.76 (s, 1H, Ar
H
ortho). 9b: δ -1.83 (d, 2JHH ) 11.7, 1H, CHsynH), 1.27 (s, 9H, CMe3),
1.58 (s, 15H, C5Me5), 2.09 (s, 3H, Tol Me), 4.28 (d, 2JHH ) 11.7, 1H,
CHantiH), 7.03 (d, 3JHH ) 7.4, 2H, Tol H); 7.71 (d, 3JHH ) 7.4, 2H, Tol
H); 13C{1H} NMR (125 MHz, dioxane-d8). 9a: δ 10.1 (C5Me5), 21.4
(Tol Me), 33.9 (CMe3), 41.3 (CMe3), 111.45 (C5Me5), 121.0 (CH2),
127.7, 128.8, 137.2, 137.3 (Tol Caryl), 182.9 (Tol Cipso). 9b: δ 10.0
(C5Me5), 21.5 (Tol Me), 33.9 (CMe3), 41.0 (CMe3), 111.4 (C5Me5),
118.6 (CH2), 128.6, 133.6, 137.6 (Tol Caryl), 180.2 (Tol Cipso). Anal.
Calcd for C22H33NOW: C, 51.67; H, 6.50; N, 2.74. Found: C, 51.78;
H, 6.68; N, 2.68.
temperature 120 °C) 517 [P+, 184W], 487 [P+-NO]; H NMR (400
1
MHz, C6D6) δ 1.53 (s, 15H, C5Me5), 2.14 (d, 2JHH ) 6.4, 1H, CHsynH),
2
3
3.41 (d, JHH ) 6.4, 1H, CHantiH), 6.50 (t, JHH ) 7.8, 2H, Bzl Hmeta),
3
6.84 (d, JHH ) 7.6, 2H, Bzl Hortho), 6.9-7.3 (m, Bzl Hpara, Ph H);
13C{1H} NMR (100 MHz, CDCl3) δ 10.5 (C5Me5), 48.0 (CH2), 108.9
(C5Me5), 113.2 (Bzl Cipso), 124.1, 127.0 (Ph Caryl), 129.1, 132.3, 135.7
(Bzl Caryl), 138.6 (Ph Caryl), 172.0 (Ph Cipso). Anal. Calcd for C23H27-
NOW: C, 53.40; H, 5.26; N, 2.71. Found: C, 53.10; H, 5.36; N, 2.77.
13-d6: (39 mg, 48%); IR (cm-1) 1562 (s, νNO); MS (LREI, m/z, probe
Cp*W(NO)(CH2C6H5)(C6H4-3-Me) (11a) and Cp*W(NO)(CH2-
C6H5)(C6H4-4-Me) (11b). Complexes 11a-b were prepared by ther-
molysis of 12 in toluene. Complex 11a was crystallized as orange
microcrystals (19 mg, 25%) by crystallization from 4:1 Et2O/hexanes.
11a: IR (cm-1) 1562 (s, νNO); MS (LREI, m/z, probe temperature
150 °C) 531 [P+, 184W], 501 [P+-NO]; 1H NMR (400 MHz, C6D6) δ
temperature 120 °C) 523 [P+, 184W], 493 [P+-NO]; H NMR (400
1
MHz, C6D6) δ 1.53 (s, 15H, C5Me5), 3.39 (br s, 1H, CHantiD), 6.50 (t,
3JHH ) 7.8, 2H, Bzl Hmeta), 6.84 (d, 3JHH ) 7.6, 2H, Bzl Hortho), 7.31 (t,
3JHH ) 7.3, 1H, Bzl Hpara); H{1H} NMR (77 MHz, C6H6) δ 2.15 (br
2
s, CHDsyn), 7.1 (m, Ph D); 13C{1H} NMR (100 MHz, CDCl3) δ 10.5
2
1
1.59 (s, 15H, C5Me5), 2.19 (d, JHH ) 5.9, 1H, CHsynH), 2.20 (s, 3H,
(C5Me5), 47.7 (t, JCD ) 22.5 Hz, CHD), 108.9 (C5Me5), 113.2 (Bzl
2
3
Tol Me), 3.46 (d, JHH ) 5.9, 1H, CHantiH), 6.56 (t, JHH ) 7.8, 2H,
C
ipso), 129.1, 132.3, 135.7 (Bzl Caryl), 171.8 (Ph Cipso). 13C resonances
3
3
Bzl Hmeta), 6.82 (d, JHH ) 7.1, 1H, Tol H), 6.89 (d, JHH ) 7.4, 2H,
attributable to Ph-D were not observed. Anal. Calcd for C23H21D6-
NOW:50 C, 52.80; H/D, 5.20; N, 2.67. Found: C, 52.45; H/D, 4.99; N,
2.80.
3
3
Bzl Hortho), 6.96 (d, JHH ) 7.1, 1H, Tol H), 7.03 (t, JHH ) 7.1, 1H,
Tol Hmeta), 7.10 (t, 3JHH ) 7.4, 1H, Bzl Hpara), 7.19 (s, 1H, Tol Hortho);
13C{1H} NMR (75 MHz, CDCl3) δ 10.60 (C5Me5), 20.79 (Tol Me),
48.18 (CH2), 108.95 (C5Me5), 113.57 (Bzl Cipso), 129.11, 132.07, 135.71
(Bzl Caryl), 125.05, 126.67, 128.25, 135.37, 138.7 (Tol Caryl), 172.4
(Tol Cipso). Anal. Calcd for C24H29NOW: C, 54.25; H, 5.50; N, 2.64.
Found: C, 54.34; H, 5.65; N, 2.70.
Preparation of Cp*W(NO)(dCHC6H5)(PMe3) (14a-b). Com-
plexes 14a-b were prepared by the thermolysis of 12 (75 mg, 0.15
mmol) in THF (4 mL) and PMe3 (∼20 mmol). Complex 14a was
isolated as yellow-orange blocks (38 mg, 52% yield) by crystallization
from 1:2 THF/hexanes.
1
11b: H NMR (400 MHz, C6D6) δ 1.60 (s, 15H, C5Me5), 2.19 (d,
14a: IR (cm-1) 1527 (s, νNO); MS (LREI, m/z, probe temperature
2
2JHH ) 5.9, 1H, CHsynH), 2.23 (s, 3H, Tol Me), 3.46 (d, JHH ) 5.9,
150 °C) 515 [P+, 184W]; 1H NMR (400 MHz, C6D6) δ 0.94 (d, 2JHP
)
3
3
3
1H, CHantiH), 6.58 (t, JHH ) 7.8, 2H, Bzl Hmeta), 6.90 (d, JHH ) 7.4,
9.0, 9H, PMe3), 1.88 (s, 15H, C5Me5), 7.05 (t, JHH ) 7.3, 1H, Bzl
3
3
2H, Bzl Hortho), 6.92 (d, JHH ) 7.1, 2H, Tol H), 7.00 (d, JHH ) 7.1,
H
H
para), 7.37 (t, 3JHH ) 7.6, 2H, Bzl Hmeta), 8.06 (d, 3JHH ) 7.6, 2H, Bzl
3
3
2H, Tol H), 7.10 (t, JHH ) 7.4, 1H, Bzl Hpara); 13C{1H} NMR (75
ortho), 12.01 (d, JHP ) 3.8, 1H, WdCH); 13C{1H} NMR (75 MHz,
1
MHz, CDCl3) δ 10.68 (C5Me5), 21.26 (Tol Me), 48.22 (CH2), 108.95
(C5Me5), 113.59 (Bzl Cipso), 129.22, 132.09, 135.98 (Bzl Caryl), 128.08,
133.35, 139.2, (Tol Caryl), 168.4 (Tol Cipso).
CDCl3) δ 11.2 (C5Me5), 17.8 (d, JCP ) 34, PMe3), 108.4 (C5Me5),
3
125.6, 128.4, 128.6 (Bzl Caryl), 153.6 (d, JCP ) 3.3, Bzl Cipso), 262.5
(d, 2JCP ) 9.6, WdCH); 31P NMR (81 MHz, C6D6) δ -8.9 (s, 1JPW
)
445 Hz). NOEDS (400 MHz, C6D6) δ irrad at 0.94, NOEs at 8.06,
12.01; irrad at 1.88; NOE at 12.01. Anal. Calcd for C20H30NOPW: C,
46.62; H, 5.87; N, 2.72. Found: C, 46.42; H, 5.69; N, 2.95.
Cp*W(NO)(CH2C6H5)(C6H4-2-Me) (11c). Complex 11c was pre-
pared via the reaction of Cp*W(NO)(CH2C6H5)Cl (0.160 g, 0.34 mmol)
and (2-Me-C6H4CH2)2Mg‚x(dioxane) (0.063 g, 0.34 mmol) in THF (10
mL). The crude residue was extracted with 4:1 Et2O/hexanes (3 × 5
mL) and filtered through Celite (1 × 0.7 cm) supported on a frit.
Concentrating the solution followed by storage at -30 °C overnight
provided 11c as red blocks (55 mg, 32%): IR (cm-1) 1552 (s, νNO);
14b: 1H NMR (400 MHz, C6D6) δ 1.08 (d, 2JHP ) 9.0, 9H, PMe3),
1.80 (s, 15H, C5Me5), 7.05 (t, 3JHH ) 7.8, 1H, Bzl Hpara), 7.10 (t, 3JHH
3
) 7.8, 2H, Bzl Hmeta), 7.20 (d, JHH ) 7.8, 2H, Bzl Hortho), 13.70 (d,
3JHP ) 4.4, 1H, WdCH); 13C{1H} NMR (75 MHz, CDCl3) δ 11.0
(C5Me5), 19.4 (d, 1JCP ) 34, PMe3), 107.8 (C5Me5), 124.8, 126.7, 127.9
(Bzl Caryl), 153.5 (d, 3JCP ) 2.5, Bzl Cipso), 262.5 (d, 2JCP ) 10.1, Wd
1
MS (LREI, m/z, probe temperature 150 °C) 531 [P+, 184W]; H NMR
(400 MHz, C6D6) δ 1.92 (s, 15H, C5Me5), 2.62 (m, 4H, Tol Me,
1
CH); 31P NMR (81 MHz, C6D6) δ -7.2 (s, JPW ) 482 Hz). NOEDS
2
CHsynH), 3.46 (d, JHH ) 6.6, 1H, CHantiH), 5.77 (br d, 1H, Tol H),
3
3
(400 MHz, C6D6) δ irrad at 1.08, NOE at 13.70; irrad at 1.80, NOE at
7.20.
6.59 (t, JHH ) 7.5, 1H, Bzl Hpara), 6.79 (t, JHH ) 7.9, 1H, Tol H),
3
3
6.90 (t, JHH ) 7.9, 2H, Bzl Hmeta), 7.07 (d, JHH ) 7.9, 1H, Tol H),
3
3
7.16 (t, JHH ) 7.9, 1H, Tol H), 7.21 (d, JHH ) 7.5, 2H, Bzl Hortho);
13C{1H} NMR (75 MHz, CDCl3) δ 10.4 (C5Me5), 28.5 (Tol Me), 52.8
(br s, CH2), 109.6 (C5Me5), 119.9 (br, Bzl Cipso), 124.0, 124.8, 128.5,
129.1, 131.2, 134.3, 149.6 (Ar C), 177.7 (Tol Cipso). Anal. Calcd for
C24H29NOW: C, 54.25; H, 5.50; N, 2.64. Found: C, 54.16; H, 5.64;
N, 2.74.
Cp*W(NO)(CH2C6H5)(CH2SiMe3) (15). Complex 15 was prepared
by the thermolysis of 12 in tetramethylsilane. The characterization data
for this complex match that previously reported:37 MS (LREI, m/z, probe
1
temperature 150 °C) 527 [P+, 184W]; H NMR (400 MHz, C6D6) δ
-4.06 (d, 2JHH ) 13.5, 1H, CHsynHSiMe3), -0.73 (d, 2JHH ) 13.5, 1H,
CHantiHSiMe3), 0.32 (s, 9H, SiMe3), 1.55 (s, 15H, C5Me5), 2.10 (d,
2
2JHH ) 5.0, 1H, CHsynHPh), 3.25 (d, JHH ) 8.5, 1H, CHantiPh), 6.54
Cp*W(NO)(CH2C6H5)(CH2CMe3) (12). Complex 12 was prepared
via the reaction of Cp*W(NO)(CH2CMe3)Cl (1.0 g, 2.2 mmol) and
(C6H5CH2)2Mg‚x(dioxane) (0.34 g, 2.2 mmol) in Et2O (10 mL). The
crude residue was extracted with 4:1 Et2O/hexanes (3 × 15 mL) and
filtered through Celite (1 × 0.7 cm) supported on a frit. Storing the
filtrate at -30 °C overnight provided 10 as yellow-orange needles (890
mg, 79%): IR (cm-1) 1538 (s, νNO); MS (LREI, m/z, probe temperature
3
3
(d, 2H, JHH ) 7.5, 1H, Bzl Hortho), 6.62 (d, 2H, JHH ) 7.5, 2H, Bzl
3
H
meta),7.50 (t, JHH ) 7.3, 1H, Bzl Hpara); 13C{1H} NMR (75 MHz,
CDCl3) δ 3.0 (SiMe3), 10.5 (C5Me5), 20.0 (CH2SiMe3), 47.0 (CH2Ph),
108.3 (C5Me5), 115.5 (Bzl Cipso), 129.5, 130.4, 134.7 (Bzl Caryl).
Kinetic Studies. Kinetic studies of the trapping of 1 and 1-d4 by
PMe3 in cyclohexane were performed using a HP8452 UV-visible
spectrometer equipped with a thermostated cell holder connected to a
VWR1150 constant-temperature bath that is accurate to (0.05 °C.
1
2
150 °C) 511 [P+, 184W]; H NMR (400 MHz, C6D6) δ -2.35 (d, JHH
) 13.4, 1H, CHsynHCMe3), 1.16 (s, 9H, CMe3), 1.54 (s, 15H, C5Me5),

C-H Bond ActiVation by W Alkylidene Complexes
J. Am. Chem. Soc., Vol. 123, No. 4, 2001 623
Table 2. X-ray Crystallographic Data for Complexes 4a, 6, 7a, 8a, and 14a
4a
6
7a
8a
14a
Crystal Data
empirical formula
cryst habit, color
cryst size (mm)
cryst syst
space group
vol (Å3)
C18H34NOW
irregular, yellow
0.13 × 0.11 × 0.09
monoclinic
P21
1054.83(6)
8.5740(3)
13.9660(5)
8.9140(3)
90
98.804
90
2
1.559
55.54
C19H34NOPW
plate, orange
0.50 × 0.35 × 0.02
monoclinic
C2/c
C17H27NOW
plate, yellow
0.28 × 0.20 × 0.20
monoclinic
P21/c
1655.88(7)
7.5179(2)
13.8169(3)
16.1151(4)
90
C18H29NOW
irregular, yellow
0.20 × 0.25 × 0.30
monoclinic
P21/n (No. 14)
1743.5(2)
8.4943(13)
13.8052(12)
14.9389(5)
90
95.5851
90
4
1.751
66.38
C20H30NOPW
block, yellow
0.40 × 0.20 × 0.20
monoclinic
Pc (No. 7)
1028.93(4)
9.5066(3)
8.8989(2)
12.5831(3)
90
104.8540(5)
90
2
1.663
57.04
4022.4(2)
30.6298(7)
8.2908(2)
16.6982(4)
90
108.4520(10)
90
a (Å)a
b (Å)
c (Å)
R (deg)
â (deg)
98.424(1)
90
4
1.786
69.73
γ (deg)
Z
8
density (calcd) (Mg/m3)
1.675
58.28
absorptn coeff (cm-1
)
F000
492
2016
872
904
508
Data Collection and Refinement
total measd reflctns
6896
9770
8282
16138
8241
no. of unique reflctns
3562 (Rint ) 0.0620)
RF ) 0.0404,
wRF ) 0.0916
1.059
3514 (Rint ) 0.0648)
RF ) 0.0424,
wRF ) 0.0760
0.963
2905 (Rint ) 0.0577)
RF ) 0.0535,
wRF ) 0.1414
1.012
4516 (Rint ) 0.048)
RF ) 0.046,
wRF )0.046
1.87
4072 (Rint ) 0.040)
RF ) 0.016,
wRF ) 0.025
0.89
final R indicesb
goodness-of-fit on F2c
largest diff peak and
1.270 and -1.714
1.755 and -0.932
3.601 and -2.645
2.78 and -2.69
1.29 and -1.12
hole (e Å-3
)
a Cell dimensions based on: 4a, 5466 reflections, 2.31 < 2θ < 25.29°; 5, 5497 reflections, 1.40 < 2θ < 25.06°; 6a, 5216 reflections, 1.95 <
2θ < 25.23°. b Number of observed reflections: 4a, 3092 (I > 2σ(I)); 5, 2571 (I > 2σ(I)); 6a, 2116 (I > 2σ(I)); 7a, 2795 (I > 3σ(I)); 14a, 3836
2
2
4
2
2
(I > 3σ(I)). RF on F ) ∑|(|Fo| - |Fc|)|/∑|Fo|; wRF ) [∑(w(|Fo| - |Fc| )2)/∑wFo ]1/2; w ) [σ2Fo + (AP)2 + (BP)]-1, where P ) (Fo + 2Fc2)/3
for 4a (A ) 0.0.284, B ) 0.0), 5 (A ) 0.0181, B ) 0.0) and 6a (A ) 0.094, B ) 0.0); w ) [σ2Fo ]-1 for 7a and 14a. c GOF ) [∑(w(|Fo| -
2
|Fc|)2)/(degrees of freedom)]1/2
.
Typical experiments were conducted using a cyclohexane solution of
1 (3 mL, 5.29 × 10-4 M) and excess PMe3 (50 µL, 300 equiv) in a
gastight quartz cuvette equipped with a Schlenk sidearm. Kinetic runs
monitored the starting material band at 486 nm for no less than 3.5
half-lives. The rate constants (kobs) were calculated using linear
regression methods from plots of ln(A(t) - A∞) versus time. Values of
∆H‡ and ∆S‡ were calculated using regression methods from plots of
ln(kobs/T) versus 1/T for the temperature range 344-371 K.
atom pairs in the Cp* group to be highly correlated. A total of 121
restraints were imposed to help the Cp* group become more regular
(SAME, DELU, SIMU, and FLAT) and prevent C(4) (ISOR) from
going nonpositive definite. The carbene, being off but close to the
pseudomirror, was also restrained with DELU. For 6, the anisotropic
displacement parameters of the ring carbon atoms of the Cp* ligand
were restrained to approximate rigid body motion. For 7a, the hydride
was not found and instead was placed in a physically reasonable location
and refined positionally while the Uiso was tied to 1.5 times the
equivalent of W(1). All calculations were performed using SGI INDY
R4400-SC or Pentium computers using the SHELXTL V5.0 suite of
programs.
Data collection and structure solution for complexes 8a and 14a
were conducted at the University of British Columbia by Steven J.
Rettig and Brian O. Patrick, respectively. All measurements were
recorded at -93(1) °C on a Rigaku/ADSC CCD area detector using
graphite-monochromated Mo KR radiation. For 8a, the data were
collected to a maximum 2θ value of 63.6° in 0.50° oscillations with a
15.0 s exposure. A sweep of data was collected using φ oscillations
from 0.0 to 190.0° at ø ) -90°, and a second sweep was performed
using ω oscillations between -22.0 and 18.0° at ø ) -90°. The crystal-
to-detector distance was 39.16 mm, and the detector swing angle was
-10.00°. For 14a, the data were collected to a maximum 2θ value of
55.8° in 0.50° oscillations with a 12.0-s exposure. A sweep of data
was collected using φ oscillations from 0.0 to 190.0° at ø ) 0°, and a
second sweep was performed using ω oscillations between -18.0 and
23.0° at ø ) 90°. The crystal-to-detector distance was 40.13(6) mm,
and the detector swing angle was -5°. Both data sets were corrected
for Lorentz and polarization effects. The solid-state structures were
solved by heavy-atom Patterson methods and expanded using Fourier
techniques. The non-hydrogen atoms were refined anisotropically. For
8a, the metal hydride was placed in a difference map position and was
not refined, and the carbon-bound hydrogen atoms were fixed in
calculated positions with d(C-H) ) 0.98 Å. For 14a, the hydrogen
atoms were included but not refined. The final cycles of full-matrix
least-squares refinements were based on 4130 unique reflections and
190 variable parameters for 8a and 3850 unique reflections and 215
variable parameters for 14a. All calculations were performed using the
teXsan crystallographic software package of Molecular Structure Corp.
Kinetic studies of the thermolysis of 1 and 12 in benzene-d6 were
performed using a J. Young NMR tube charged with 1 or 12 (15 mg,
30 mmol) and 1 mL of a benzene-d6 solution containing 0.5 equiv (15
mmol) of hexamethylsilane as an internal standard. The thermolysis
was conducted in the constant-temperature bath, and the samples were
removed at several intervals to record 1H NMR spectra. The monitoring
was performed for >3.5 half-lives, and rate constants were calculated
using linear regression methods from first-order plots of the loss of
starting material versus time. Studies of the isomerization of 9c and
1
11c were conducted in a similar fashion in neat benzene-d6. H NMR
spectra were recorded at several time intervals and the percent
conversions were calculated using integrations of 9a-b product signals
to the combined CMe3 signals for 9a-c and integrations of 11c signals
versus the combined Cp* signal for 11a-c.
X-ray Diffraction Studies. Data collection and structure solution
for complexes 4a, 6, and 7a were conducted at the X-ray Crystal-
lographic Laboratory at the University of Minnesota by Victor G.
Young. Data were recorded at 173(2) K with a Siemens SMART
System using graphite-monochromated Mo KR radiation. The data were
collected using the hemisphere technique to a resolution of 0.84 Å with
0.30° steps in ω. All structures were solved by direct methods, and all
non-hydrogen atoms were refined anisotropically unless stated other-
wise. All carbon-bound hydrogen atoms were placed in ideal positions
and refined as riding atoms with individual (or group if appropriate)
isotropic displacement parameters. For 4a, racemic twinning in the
crystal lattice prevented a direct solution in the space group P21. A
direct-methods solution was forced in P21/m, which provided a
contiguous molecule that was easily transformed into P21 and subse-
quently refined in P21. The presence of a twin caused serious problems
during the final stages of refinement. The nominal structure is
pseudosymmetric with its twin, thus causing the parameters of several

624 J. Am. Chem. Soc., Vol. 123, No. 4, 2001
Adams et al.
X-ray crystallographic data for all five complexes are collected in
Table 2, and full details of all crystallographic analyses are provided
as Supporting Information.
Young of the University of Minnesota for solving the crystal
structures of the representative complexes reported in this paper.
Supporting Information Available: ORTEPs and tables
listing crystallographic information, atomic coordinates and Beq
values, anisotropic thermal parameters, and intramolecular bond
distances, angles, and torsion angles. This material is available
free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. We are grateful to the Natural Sciences
and Engineering Research Council of Canada for support of
this work in the form of grants to P.L. and postgraduate
scholarships to C.S.A. and E.T. We also thank Drs. K. J. Ross
and W. S. McNeil for stimulating and helpful discussions.
Finally, we acknowledge Drs. S. J. Rettig (deceased October
27, 1998) and B. O. Patrick of this Department, and Dr. V. G.
JA002457E