α-Hydrogen migration reactions in tungsten(VI) cyclopentadienyl alkylidyne complexes

Article abstract of DOI:10.1016/S0022-328X(98)00781-5

W(CAd)(OCMe3)3 (Ad = 1-adamantyl) can be prepared by treating W2(OCMe3)6 with 1-adamantanecarbonitrile and converted into W(CAd)(triflate)(OCMe3)2(dme) by treating it with Me3Si(triflate) (dme = 1,2-dimehoxyethane).W(CAd)(triflate)(OCMe3)2(dme) reacts with NaCp to yield CpW(CAd)(OCMe3)2.CpW(CAd)(OCMe3)2 can be converted into CpW(CAd)Cl2 in dichloromethane by treating it with Me3SiCl.Alkylation of CpW(CAd)Cl2 yields CpW(CAd)(CH2CMe3)Cl or CpW(CAd)R2 complexes (R = CH2CMe3, CH2Ph, or CH3).Both CpW(CAd)(CH2CMe3)Cl and CpW(CAd)(CH2CMe3)2 tautomerize to give mixtures containing CpW(CCMe3)(CH2Ad)Cl and CpW(CCMe3)(CH2CMe3)(CH2Ad), respectively.An X-ray study of CpW(CAd)(CH2CMe3)2 is consistent with some α agostic interaction of a neopentyl proton in each neopentyl ligand with the metal.In contrast, CpW(CAd)(NMe2)(CH2CMe3) shows no evidence of tautomerizing to CpW(CCMe3)(NMe2)(CH2Ad) at room temperature over the course of two weeks, while CpW(CAd)(NHCMe3)Cl tautomerizes to two rotameric forms of CpW(NCMe3)(CHAd)Cl. - Keywords: Migration reactions; α-Hydrogen; Tungsten

Full text of DOI:10.1016/S0022-328X(98)00781-5

Journal of Organometallic Chemistry 569 (1998) 125–137
h-Hydrogen migration reactions in tungsten(VI) cyclopentadienyl
alkylidyne complexes
Timothy H. Warren, Richard R. Schrock *, William M. Davis
Department of Chemistry 6-331, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
Received 31 March 1998
Abstract
W(CAd)(OCMe3)3 (Ad=1-adamantyl) can be prepared by treating W2(OCMe3)6 with 1-adamantanecarbonitrile and converted
into W(CAd)(triflate)(OCMe3)2(dme) by treating it with Me3Si(triflate) (dme=1,2-dimethoxyethane). W(CAd)(triflate)
(OCMe3)2(dme) reacts with NaCp to yield CpW(CAd)(OCMe3)2. CpW(CAd)(OCMe3)2 can be converted into CpW(CAd)Cl2 in
dichloromethane by treating it with Me3SiCl. Alkylation of CpW(CAd)Cl2 yields CpW(CAd)(CH2CMe3)Cl or CpW(CAd)R2
complexes (R=CH2CMe3, CH2Ph, or CH3). Both CpW(CAd)(CH2CMe3)Cl and CpW(CAd)(CH2CMe3)2 tautomerize to give
mixtures containing CpW(CCMe3)(CH2Ad)Cl and CpW(CCMe3)(CH2CMe3)(CH2Ad), respectively. An X-ray study of CpW(-
CAd)(CH2CMe3)2 is consistent with some h agostic interaction of a neopentyl proton in each neopentyl ligand with the metal. In
contrast, CpW(CAd)(NMe2)(CH2CMe3) shows no evidence of tautomerizing to CpW(CCMe3)(NMe2)(CH2Ad) at room tempera-
ture over the course of two weeks, while CpW(CAd)(NHCMe3)Cl tautomerizes to two rotameric forms of Cp-
W(NCMe3)(CHAd)Cl. © 1998 Elsevier Science S.A. All rights reserved.
Keywords: Migration reactions; h-Hydrogen; Tungsten
1. Introduction
triamidoamine complexes [4], and that h abstraction to
give an alkylidene can be forced to be virtually the only
sterically tenable process by increasing the size of the
silyl substituent in [(R3SiNCH2CH2)3]Ta(alkyl)2 species
from R=Me to R=Et [5]. The evidence suggests that
in a more crowded ‘pocket’, the i agostic interaction
that precedes i abstraction in the alkyl becomes steri-
cally disfavored, while the h agostic interaction that
precedes h abstraction at the same time is encouraged.
(A lack of ‘steric pressure’ of this type is part of the
explanation as to why formation of methylene com-
plexes from methyl complexes is generally slow [2].) The
process of h abstraction in a dialkyl complex to give an
alkylidene complex can be viewed as one that involves
activation of an h-hydrogen through some agostic in-
teraction [6] with the metal, making it susceptible to
migration as a ‘proton’ to a neighboring nucleophilic h
carbon atom. Abstraction of an alkylidene’s h-hydro-
gen by an alkyl group to give an alkylidyne ligand is
In a d0 complex containing at least two alkyl ligands
an h-hydrogen in one alkyl ligand can sometimes be
‘abstracted’ by a neighboring alkyl to give an alkane
and an alkylidene complex [1–3]. It has long been
known that neopentyl complexes are most prone to h
abstraction, with trimethylsilylmethyl, benzyl, and (es-
pecially) methyl being progressively less so. It was also
noted in early studies that h abstraction is faster in
sterically crowded environments, e.g. coordination of
additional ligands such as phosphines can induce or
accelerate h-hydrogen abstraction reactions. Recently it
has been shown that h abstraction and i abstraction
can compete in [(Me3SiNCH2CH2)3]Ta(alkyl)2 species
[4], that the rate of h abstraction can be enhanced by
increasing the size of the alkyls in the apical ‘pocket’ of
* Corresponding author. Tel.: +1 617 753 7670.
0022-328X/98/$ - see front matter © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(98)00781-5

126
T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
2. Results
also relatively well-known [7] and most likely involves
an activation of the alkylidene’s h-hydrogen through an
h agostic interaction with the metal. Formation of
alkane has always been thought to be irreversible,
although there is now evidence for CH activation by a
non-d0 tungsten alkylidene complex [8] and by
‘(Me3CCH2)2Ti=CHCMe3’ [9].
2.1. Synthesis of CpW(CR)X2
The most facile synthesis of ‘d0’ tungsten alkylidyne
complexes is the reaction between internal alkynes or
nitriles [29,30] and readily prepared W2(OCMe3)6 [31–
33]. When
a
nitrile is employed the resulting
In contrast, there are few documented examples of
migration of an h-hydrogen in a d0 alkyl complex to an
alkylidene ligand (a literally or approximately degener-
ate process) or to an alkylidyne ligand (to give a
bis(alkylidene) complex). In the first category is h H/D
scrambling in Ta(CDCMe3)(CH2CMe3)3 [10], which
takes place with a rate constant of ca. 3.5×10−5 s−1
at 348 K. In the latter category is the reaction in which
W(CSiMe3)(CH2CMe3)3 produces an equilibrium mix-
ture of W(CSiMe3)(CH2CMe3)3 and W(CCMe3)
(CH2CMe3)2(CH2SiMe3) (Eq. (1)) upon heating in solu-
tion [11]; at 352 K the average rate constant (forward
(Me3CO)3WꢀCR and (Me3CO)3WꢀN products usually
can be separated readily, as the alkylidyne is soluble in
pentane, while the nitride product (a polymer in the
solid state [34]) is not. W2(OCMe3)6 reacts smoothly
with 1-adamantanecarbonitrile in pentane to give a
mixture of W(CAd)(OCMe3)3 (Ad=1-adamantyl) (1)
and W(N)(OCMe3)3. W(CAd)(OCMe3)3 is readily sepa-
rated from W(N)(OCMe3)3 by filtration and crystallizes
from the reaction mixture as colorless cubes in good
yield.
Attempts to displace a t-butoxide ligand in 1 with a
cyclopentadienyl ligand using NaCp under a variety of
conditions were not successful. However, a much more
reactive monotriflate species (2) can be prepared as
shown in Eq. (2). Complex 2 precipitates from the
reaction mixture as yellow microcrystals and can be
isolated in 85–95% yield by filtration.
and reverse reactions) is ca. 1.7×10−5
s
−1. The
‘bis(alkylidene)’ complex, W(CHSiMe3)(CHCMe3)
(CH2CMe3)2, is the proposed intermediate in this reac-
tion. Stable ‘d0’ bisalkylidene complexes are extremely
rare, the only examples being tantalum complexes such
as Ta(CHCMe3)2(PMe3)2Cl or CpTa(CHCMe3)2(PMe3)
[12–14].
D
W(CSiMe3)(CH2CMe3)3 X W(CCMe3)(CH2SiMe3)
(CH2CMe3)2
(1)
We became interested in the possibility of observing
migration of an h proton from an alkyl to an
adamantylidyne ligand in tungsten ‘d0’ complexes. An
attractive feature of the adamantyl group is that it is
electronically and sterically more similar to a t-butyl
group than is a trimethylsilyl group. (In order for h
proton scrambling to be observed in W(C-
SiMe3)(CH2CMe3)3 an h proton must migrate to a
silyl-substituted alkylidyne ligand and from a silyl-sub-
stituted alkyl ligand.) As in the case of h-hydrogen
migration in W(CSiMe3)(CH2CMe3)3, the consequence
of h-hydrogen migration to an adamantylidyne ligand
should be readily observable. In order to explore the
possibility of observing a bisalkylidene intermediate
related to CpTa(CHCMe3)2(PMe3), we chose to prepare
monocyclopentadienyl alkylidyne complexes of the type
CpW(CR)(CH2R%)2. Another interest in complexes of
this type is that the Cp(RC)W core is analogous in
electron count and structure to Cp2M (M in group 4)
[15,16], Cp(RN)M (M in group 5) [17–21], and
(RN)2M (M in group 6) [22–27] cores. (An extensive
discussion of these analogies and some of the structural
consequences thereof has been published by Gibson
[28].) The results of these investigations are reported
here.
Complex 2 exhibits broad, inequivalent t-butoxide reso-
nances in room temperature H-NMR spectra (in ben-
1
zene-d6) as
a
consequence of dissociation of
1
dimethoxyethane on the H-NMR time scale. The reac-
tion between an alkylidyne complex and trimethylsilyl
triflate should be compared with reactions between
W(CCMe3)(OCMe3)3 and two equivalents of acids HX
(e.g. X=Br, acetate, OC6F5) to give complexes of the
type W(CHCMe3)(OCMe3)2X2 [35].
Sodium cyclopentadienide reacts with 2 in THF at
−40°C to give CpW(CAd)(OCMe3)2 (3), which can be
isolated in high yield as yellow crystals from pentane
(Eq. (3)). In contrast, reactions between cyclopentadi-
enyl
sources
and
W(CCMe3)(dme)Cl3
or
[NEt4][W(CCMe3)Cl4] do not afford known CpW(C-
CMe3)Cl2 [36,37]. The presence of the two t-butoxide
ligands prevents side reactions such as reduction of the
metal or formation of dimeric species.

T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
127
Complex 3 can be converted into CpW(CAd)Cl2 (4)
in dichloromethane in the presence of five equivalents
of Me3SiCl and a catalytic amount of 2,6-lutidinium
chloride (Eq. (4)). The reaction proceeds in 80–85%
yield and 4 crystallizes readily from ether/pentane solu-
tions as brilliant purple shards. Compound 4 also could
be prepared by dissolving 3 in neat trimethylsilyl chlo-
ride, but this procedure proved irreproducible. Addi-
tion of two equivalents of HCl to 3 in ether also gave 4
in good yield, but an unidentified impurity thwarted
ready isolation of 4 in this case.
If we assume that 24 h is at least five half-lives, then t1/2
is ca. 4 h, or the rate constant for scrambling is ca.
5×10−5 s−1 for 5a. The final equilibrium value in
each case is what one would expect on a statistical basis
(2 and 1, respectively) if one assumes there is no
significant difference between an adamantyl and a t-
butyl group in alkyl or alkylidyne ligands in systems of
this type. Addition of a slight excess of 2,6-lutidinium
chloride in dichloromethane to the equilibrium mix-
tures produces corresponding mixtures of CpW(-
1
CAd)Cl2 and CpW(CCMe3)Cl2, according to H-NMR
spectroscopy. In contrast, 5b and 5c decompose to give
many products over a period of 36 h, according to
NMR spectra, probably as a consequence of inter-
molecular processes that compete with intramolecular
processes when the alkyl (CH2C6H5 or CH3) is small
relative to neopentyl or adamantyl.
NMR data for compounds 5a, 5a%, 6, and 6% are
dramatically different from NMR data for 5b and 5c
(Table 1). In 5b, for example, although two different
methylene protons are observed at ca. 1.5 and 2.4 ppm,
the values for JWH and JCH are similar for the two
methylene protons. In contrast, each symmetry inde-
pendent WCH2R group in compounds 5a, 5a%, 6, and 6%
2.2. Synthesis of alkyl deri6ati6es containing the
CpW(CAd) core
Addition of Grignard reagents to 4 in ether at
−40°C smoothly yields the dialkyl complexes shown in
Eq. (5). Careful addition of one equivalent of neopentyl
Grignard to 4 produced the mononeopentyl complex,
WCp(CAd)(CH2CMe3)Cl (6). Complex 5b is isolated as
bright yellow crystals from a mixture of ether and
dichloromethane while complexes 5a, 5c, and 6 are
isolated as deep red crystals from pentane at −40°C
(70–80% yield).
1
displays two sets of H resonances due to chemically
inequivalent methylene protons centered around l 3.5
and −1.0 ppm, and the differences in the magnitude of
2JWH within a pair of geminal CH resonances is sub-
stantial (Fig. 1). The second order nature of the
methylene proton resonances in 5a is apparent (AA%BB%
pattern overlaying an AA%BB%X pattern where X=
183W, ca. 14% abundant.) In each case the upfield
resonance appears to be most strongly coupled to 183W
2
with JWH ranging from 12.0 to 13.0 Hz, whereas each
corresponding downfield resonance exhibits a 2JWH esti-
mated to be less than ca. 4 Hz. The 13C resonance for
the Ch carbon atoms in the neopentyl and adamantyl
complexes also are consistently ca. 50 ppm further
downfield from the Ch carbon resonances in 5b and 5c
and the geminal h-hydrogens in each symmetry inde-
pendent WCH2R group are coupled to their corre-
sponding alkyl h-C atom to a different extent. Each
WCH2R 13C resonance shows a four line pattern result-
Compounds 5a and 6 partially tautomerize over a
period of days in benzene-d6 at room temperature by a
double h-H shift (Eqs. (6–7)). At room temperature
equilibrium (Keq=2.1(1)) is reached between 5a and 5a%
after ca. 24 h, whereas the analogous equilibration of 6
with 6% occurs considerably more slowly, reaching com-
pletion (Keq=1.05(4)) only after 1–2 weeks at room
temperature.
1
1
ing from one JCH of ca. 120 Hz and one JCH of ca.
100 Hz. Selective 13C{1H}-NMR decoupling experi-
ments allow assignment of the high and low 1JCH values
to the downfield and upfield 1H resonances, respec-
tively, in several cases.
4
The JHH (2.6 Hz) in the upfield WCH2CMe3 and
WCH2Ad resonances (Fig. 1) in the mixed alkyl tau-
tomer 5a% is unusually large. For comparison, a related
4JHH of 3.1 Hz has been observed in the zirconocene
silylamide complex Cp2Zr(H)[N(tBu)(SiMe2H)] that
has been shown crystallographically to possess a i
agostic Si–H bond [38]. In organic molecules proton-
proton couplings through four |-bonds generally are

128
T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
Table 1
Ch- and Hh-NMR parameters or complexes 5–7 (in benzene-d6 unless noted otherwise)
Complex
l
1H (ppm)
2JWH (Hz)
l
13C (ppm)
1JCH (Hz)
CpW(CAd)(CH2)Ph)a2 (5b)
2.422
1.484
4.8
8.1
32.37
134.1b
137.1b
CpW(CAd)(CH3)2 (5c)
1.133
5.8
40.25
90.67
121.9
CpW(CAd)(CH2CMe3)2 (5a)
3.528c
1.147c
B4
12.6
121.0
98.8
CpW(CCMe3)(CH2CMe3)(CH2Ad) (5a%)
3.557d
B4
12.0
12.9
B4
94.98g
87.78g
120.2
99.6
101.2
119.3
−1.502d,e
0.886e,f
3.299f
CpW(CAd)(CH2CMe3)Cl (6)
4.790
−0.454
4.1
13.0
98.08
125.2
98.4
CpW(CCMe3)(CH2Ad)Cl (6%)
4.715
−0.631
1.69
4.4
12.9
11.7
8.7
100.58
48.60
125.0
97.1
CpW(CAd)(CH2CMe3)(NMe2) (7)
113.6h
115.9h
1.53
a Recorded in dichloromethane-d2.
b Arbitrary assignment.
c 4JHH=2.1 Hz.
d 2JHH=11.7 Hz.
e 4JHH=2.6 Hz.
f 2JHH=12.4 Hz.
1
g Arbitrary correlation of 13C resonance with 1H resonances. JCH values assigned to downfield and upfield 1H resonances in analogy with
compounds 5a and 5a%.
1
h Arbitrary assignment of JCH values.
might be viewed as an incipient alkylidyne hydride
complex. We would expect h agostic interactions in 5b
and 5c to be minimal, as h agostic interactions are
believed to be encouraged (in part) by steric crowding
in the alkyl complex by the alkyl itself, which leads to
an opening of the M–Ch–Ci angle and (presumably) a
closing of the M–Ch–Hh angle.
large only in bicyclic systems in which the H–C–C–C–
H linkage is locked in a ‘W’ configuration[39–41] such
as those shown below [42–44].
The second interpretation is that the bulky alkyl
ligands in 5a, 5a%, 6, and 6% have a preferred orientation
(as a consequence of steric crowding) that circumstan-
tially leads to the observed characteristics for the
methylene group in proton and carbon NMR spectra.
In contrast, the benzyl groups in 5b would be expected
to be more freely rotating and the two protons there-
fore more similar than in neopentyl or adamantyl com-
plexes. However, the steric hindrance that would lead
to limited rotation probably would also lead to an
opening of the W–Ch–Ci angle and to a bending of
one or both h protons toward the metal. Limited
rotation and h agostic interactions in fact should be
synergistic and inseparable. Therefore we can only say
that the steric demands in the system lead to limited
rotation, or to an h agostic interaction, or (most likely)
to both simultaneously.
The NMR data for 5a, 5a%, 6, and 6% are consistent
with two interpretations. One is that one or more
protons in the neopentyl or adamantyl complexes is
(are) interacting with the metal in an agostic fashion.
Upfield shifted h-CH resonances have been correlated
with agostic structures [6], although the chemical shift
anisotropy of the W-alkylidyne triple bond may also
contribute to the disparity in chemical shifts between
1
the geminal hydrogen resonances [45]. A low JChHh
coupling constant is not in itself a rigorous criterion for
the identification of h agostic interactions, but the low
1
and high ranges for JCH values is at least consistent
with h agostic interactions [46–48]. Agostic interactions
between CHh and the metal in alkylidene complexes has
also been documented [1,3]. In general the Ch resonance
shifts downfield and the Hh resonance shifts upfield as
the alkylidene becomes more distorted toward what
An X-ray study of 5a (Fig. 2; Tables 2 and 3) reveals
an overall structure that is consistent with a crowded
coordination sphere, a limited ability of the neopentyl

T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
129
Fig. 1. A partial 1H-NMR (300 MHz) spectrum of an equilibrium mixture of 5a and 5a% showing the downfield (top) and upfield (bottom)
methylene resonances.
support an argument concerning an h agostic interac-
tion [19]. Although the uncertainty in these distances is
understandably large, the resulting W…H(3A) and
ligands to rotate, and (we will argue) an h agostic
interaction of some magnitude. The two neopentyl lig-
ands are related by a crystallographic mirror plane
passing through the alkylidyne Ch–W–Cp(centroid)
plane. The C(1)–W–Cp(centroid) angle (121.5(5)°;
Table 2) is considerably larger than the angle between
the two neopentyl ligands (C(3)–W–C(3%)=106.2(4)°),
as might be expected for a complex that is formally
analogous to a metallocene in bonding parameters.
However, the disparity is not as great as found in a
typical metallocene (ca. 135° and 95°, respectively). The
W-alkylidyne linkage is nearly linear (BW–C(1)–
C(2)=166.2(6)°) with a normal [7] WꢀC distance of
˚
W…H(3B) contacts of 2.42(7) and 2.61(7) A suggest
that one of the neopentyl h-H atoms in 5a (H(3A)) lies
closer to the tungsten center than the other. The h-H
atom (H(3A)) nearest to W also makes the smallest
W–C(3)–H(3) angle (93(4)° versus 109(4)°) and rests
just above the C(3)–W–C(3%) plane. As located, pro-
˚
tons H(3A) and H(3B) are 3.19 A from the adamantyli-
dyne carbon atom.
The W–C(3) distance (2.159(7) A) lies toward the
low end of the range (2.096(5)–2.258(8) A) found in
˚
˚
˚
other crystallographically characterized four- and five-
coordinate tungsten neopentyl complexes [11,49–51].
Of note, however, is the wide W–C(3)–C(4) angle of
134.8(4)°, which is outside of the range in the neopentyl
complexes referenced above (122–131°). Similar Nb–
Ch–Ci angles (131.2(2), 132.5(3)°), however, have been
found in CpNb(NAr)(CH2CMe3)2 (Ar=2,6-diiso-
propylphenyl) in which an h agostic interaction was
proposed [19]. The t-butyl group is directed toward the
alkylidyne ligand by 38.5° (the dihedral angle between
the C(3)–W–C(3%) and W–C(3)–C(4) planes). There-
fore H(3A) should be located nearest to tungsten,
should be directed there by orientation of the neopentyl
group, and should lie almost in the C(3)–W–C(3%)
plane. If one takes the analogy between the d0 CpW(-
CAd) and Cp2M (M in group 4) fragments seriously
1.746(9) A. The deviation from linearity can be ascribed
to repulsion between the adamantyl group and the
neopentyl ligands. Although the cyclopentadienyl ring
is clearly bound in an p5 manner, some slippage is
˚
apparent with W–C distances ranging from 2.328(7) A
˚
(W–C(5)) to 2.512(10) A (W–C(7)). This asymmetry
may reflect the influence of the triply bound alkylidyne
ligand, as the longest distance is to the Cp carbon atom
(C(7)) roughly trans to the W-alkylidyne linkage (C(1)–
W–C(7)=149°).
Although the h-H atoms were not located in a differ-
ence map, they were refined by placing them in ideal-
ized positions, fixing the C(3)–H(3A) and C(3)–H(3B)
˚
distances at 0.96 A, and then allowing the orientations
of these C–H bonds to refine freely, a technique that
has been employed in similar circumstances in order to

130
T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
Fig. 2. Two views of the structure of 5a. (Only the methylene protons in the two neopentyl ligands are shown.)
[15], an unoccupied orbital with A1 symmetry is avail-
able approximately in the C(3)–W–C(3%) plane and is
the only one that could accept electron density from the
two suitably oriented h-CH bonds. The resulting ‘dou-
ble h agostic’ interaction would be consistent with the
X-ray structure and with all NMR parameters as dis-
cussed above. In the simple MO description shown in
Fig. 3, the lowest energy MO would be totally bonding,
while the next highest occupied MO would contain a
node at W. The highest energy MO that has antibond-
ing character would not be occupied.
Table 3
Crystallographic data, collection parameters, and refinements for
CpW(CAd)(CH2CMe3)2 (5a)
Empirical formula
Formula weight
Diffractometer
C26H42
538.47
W
2.3. The preparation of amido and imido complexes
Siemens SMART/CCD
Red, plate
0.18×0.24×0.24
Monoclinic
15.7061(8)
12.4619(7)
12.1803(6)
90
97.5680(10)
90
2363.3(2)
C2
Crystal color, morphology
Crystal dimensions (mm)
Crystal system
The
reaction
between
CpW(CAd)Cl2
and
Me3SiNMe2 produces yellow crystalline CpW(-
CAd)(NMe2)Cl (7a) in good yield (Eq. (8)).
˚
a (A)
˚
b (A)
˚
c (A)
Table 2
h (°)
i (°)
k (°)
˚
Selected bond lengths (A) and angles (°) for CpW(CAd)(CH2CMe3)2
(5a)
3
˚
V (A )
Distances
Space group
W–C(1)
W–C(5)
W–C(6)
W–C(7)
W-centroid
1.746(9)
2.328(7)
2.444(7)
2.512(10)
2.098(8)
W–C(3)
2.159(7)
0.98(7)
0.93(7)
2.42(7)
2.61(7)
Z
2
C(3)–H(3A)
C(3)–H(3B)
W…H(3A)
W…H(3B)
Dcalc (g cm−3
)
1.513
1088
4.896
ꢀ scans
293(2)
F000
v(Mo–Kh) (mm−1
Scan type
)
Angles
Temperature (K)
W–C(1)–C(2)
C(3)–W–C(3%)
C(1)–W-centroid
C(1)–W–C(3)
C(3)–W-centroid
166.2(6)
106.2(4)
121.5(5)
102.1(2)
111.8(5)
Total number of unique reflections
No. of variables
R [I\2|(I)]
Rw [I\2|(I)]
Goodness-of-fit
1797
140
0.0375
0.0906
W–C(3)–C(4)
W–C(3)–H(3A)
W–C(3)–H(3B)
134.8(4)
93(4)
109(4)
1.011

T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
131
1H-NMR spectra of 7a at room temperature in toluene-
d8 display two well-separated methyl resonances for the
dimethylamido group (Me and Me%), as expected if the
lone pair on the amido ligand is donated into an orbital
with B2 symmetry (in the C2v symmetric Cp2HfR2 rela-
tive [15]) that lies in the Cl–W–N plane [28]. Heating
an NMR sample of 7a to 110°C caused the Me and Me%
resonances to broaden only slightly. We would not
expect the dimethylamido ligand to rotate about the
W–N bond rapidly on the NMR time scale, since a y
bond effectively would have to be fully broken in the
process (if the cyclopentadienyl ligand remains p5), i.e.
there is no orbital that can stabilize the rotated NMe2
ligand in the absence of significant distortion of the
(p5-C5H5)WꢀCAd framework.
ppm in the proton NMR spectrum and an N–H stretch
is observed at 3249 cm–1 in the IR spectrum. In theory
two rotameric forms of the amido complex could be
formed, since formation of the W–N pseudo double
bond employing an orbital that lies in the N–W–Cl
plane, and therefore the t-butyl group would be ori-
ented either toward the cyclopentadienyl ring or away
from it. Only one rotamer was observed and we cannot
say which rotamer that is.
Solutions of 8 are not stable towards migration of the
amido h-H to the alkylidyne h-carbon atom. After
standing a solution of 8 for 24 h at room temperature
in dichloromethane-d2 the major species that is present
is what we propose to be CpW(NCMe3)(CHAd)Cl (8%).
A second product (8%%) is also observed at this point and
becomes the sole product after 9 days at room temper-
ature. It is also an alkylidene, but its NMR characteristics
differ from those for 8% (Table 4). We propose that 8% and
8%% are WꢁC rotamers that differ in the relative orienta-
tion of the 1-adamantyl alkylidene substituent (Eq. (10)).
Treatment of 7a with neopentyllithium in ether yields
orange-yellow crystalline CpW(CAd)(NMe2)(CH2-
CMe3) (7b). In 7b two sharp and well-separated methyl
resonances also are observed for the NMe2 ligand,
They are analogous to the known compound, Cp-
W(NPh)(CH–t-Bu)Cl [52], only one rotamer of which
was observed. Formation of rotamers is best viewed in
terms of the analogy between the Cp(RN)M core and the
Cp2M core [28], and in terms of the bonding found in
Cp2Ta(CHR)(CH2R) complexes [1]. Formation of the
first rotamer formed by transfer of the amido Hh to the
adamantylidyne Ch would place the 1-adamantyl group
of the alkylidene syn to the Cp ring. Therefore we believe
that 8% is the syn rotamer, as shown in Eq. (10). This is
not likely to be the favored orientation of the alkylidene,
as the 1-adamantyl group is directed toward the cy-
clopentadienyl ring. Therefore the final product (8%%) is
believed to be that in which the 1-adamantyl group points
as expected. We also noticed that characteristics of h
agostic interactions are not present in 1H- and 13C-
NMR spectra of 7b (Table 1). Most informative are the
1JCH values of 113 and 115 Hz recorded for the
diastereotopic methylene protons. These similar values
are in the expected range for normal sp3 CH bonds, and
1
contrast with the JCH values centered around 100 and
120 Hz for complexes 5a, 5a%, 6, and 6%. This result is
sensible from an electronic point of view since one
might expect h agostic interactions to be ‘blocked’ in
7b, i.e. y donation from the dimethylamido ligand
involves the orbital that we proposed is used for an h
agostic interaction in 5a, 5a%, 6, or 6%. Compound 7b
shows no evidence of tautomerizing to CpW(C-
CMe3)(NMe2)(CH2Ad) at room temperature over the
course of 2 weeks (Eq. (9)), consistent with the 18
electron count if amido y donation is included.
Reaction of CpW(CAd)Cl2 with N-t-buty-
laminotrimethylsilane in dichloromethane gave CpW(-
CAd)(NH-t-Bu)Cl (8) as yellow crystals in high yield.
The NH proton is observed as a broad singlet at 9.40
Fig. 3. A simple MO description of the ‘double agostic’ interaction.

132
T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
Table 4
Selected NMR parameters for complexes 8, 8% and 8%% (in CD2Cl2)
Complex
l
1H (Cp)
l
1H (h-H)
l
13C (h-C)
1JCH (Hz)
CpW(CAd)(NHCMe3)Cl (8)
syn-CpW(NCMe3)(CHAd)Cl (8%)
anti-CpW(NCMe3)(CHAd)Cl (8%%)
6.08
6.12
6.05
9.39
303.5
264.7
274.1
11.15a
10.40b
131.3
123.5
a 2JWH=14.1 Hz.
b 2
J
B4 Hz.
WH
CpW(CAd)(CH2CMe3)X compounds occurs consider-
ably faster when X=CH2CMe3 than when X=Cl,
even after considering also the expected rate enhance-
ment due to a statistical factor of two, primarily (we
propose) because the chloride is so much less demand-
ing sterically than the neopentyl ligand. However, a
constrained orientation of an alkyl group in a crowded
environment is an inseparable part of the phenomenon,
and the fraction of its contribution to the whole cannot
be quantified. The lack of tautomerization in 5b and 5c
is consistent with the lack of any h agostic interactions
and with observed rates of h abstraction in d0 alkyl
complexes to give alkylidenes that rapidly decline in the
order CH2CMe3\CH2PhꢀMe. The extent to which
the agostic interaction could exist in the absence of
accompanying steric factors that reinforce it in general
should depend heavily upon the energy of the acceptor
orbital and its ability to accept electron density from a
CH bond. In fact, at this stage no definitive evidence
has been obtained for an h agostic interaction in a
methyl ligand. Even in [W(p5-C5Me5)Me4]+, which is
known to be deprotonated readily to give intermediate,
unstable W(p5-C5Me5)Me3(CH2) [56], a solid state
NMR study has revealed no evidence for an h agostic
interaction [57].
The reason why bisalkylidene intermediates such as
CpW(CHAd)(CHCMe3)Cl are not lower in energy than
the observed alkyl alkylidyne complexes is not obvious.
Both hypothetical CpW(CHAd)(CHCMe3)Cl and
known CpTa(CHCMe3)2(PMe3) are 16 electron species,
as are CpW(CAd)(CH2CMe3)Cl and hypothetical Cp-
Ta(CCMe3)(CH2CMe3)(PMe3). It should be noted,
however, that bisneopentylidene complexes such as Cp-
Ta(CHCMe3)2(PMe3) have not been thoroughly stud-
ied. In particular it was noted that the two
neopentylidene ligands in CpTa(CHCMe3)2(PMe3) be-
come equivalent at 60–80°C in proton and carbon
NMR spectra [12], but the nature of the equilibration
was not examined in detail. At that time equilibration
was ascribed to rotation of the neopentylidene ligands
to give an intermediate with mirror symmetry. Al-
though CpTa(CCMe3)(PMe3)2Cl was known, and it
was known to react with LiCH2CMe3 to give Cp-
away from the cyclopentadienyl ring (anti rotamer).
Consistent with this assignment is the fact that irradia-
tion of the Cp resonance in the proton NMR spectrum
of 8%% results in an NOE enhancement of the alkylide-
ne’s Hh resonance. Note that the syn and anti notations
here are different than those for tungsten and molybde-
num complexes of the type M(CHR)(NR%)(OR%%)2,
where the imido ligand is employed as the reference for
assigning syn and anti alkylidene orientations [53,54].
Since the imido ligand is pseudo triply bound to tung-
sten, and the CpW(N–t-Bu) fragment is metallocene-
like, one would not expect any agostic interaction
involving the h proton of the alkylidene ligand, and no
evidence for an h agostic interaction is observed. Simi-
lar arguments were put forward in order to explain the
lack of agostic interactions in 18 electron compounds of
the type Cp2Ta(CHR)X [55], which are related to 8 and
8% via the analogy between Cp2M and Cp(RN)M cores.
Migration of an amido ligand’s Ha to a neopentyli-
dyne Ca atom was the first method developed for
synthesizing imido neopentylidene complexes [2].
Therefore it is not surprising to find that 8 is converted
into 8% and ultimately 8%% by a ‘proton migration’ reac-
tion. However, it was noted at that time that the
process appeared to be catalyzed by external base (e.g.
NEt3). Therefore, in the absence of detailed studies, we
cannot exclude the possibility that this particular ‘mi-
gration’ reaction actually is catalyzed by external base,
e.g. by traces of t-BuNH2. All evidence in the literature
so far suggests that y donation from a primary amido
ligand takes precedence over any h agostic C–H inter-
action a primary amido ligand, and that proton migra-
tion from
a
nitrogen to
a
carbon ligand is
fundamentally different from proton migration from a
carbon to a carbon ligand.
3. Discussion and conclusions
The data presented here suggest that there is a strong
correlation between an agostic CHh interaction and the
tendency for a neopentyl or adamantyl proton to mi-
grate to a neighboring alkylidyne ligand to give what
we must presume is a bisalkylidene intermediate and
then a new alkylidyne complex. Proton migration in
Ta(CHCMe3)2(PMe3)
CMe3)(PMe3)2(CH2CMe3)’
equilibration of the neopentyl groups via intermediate
via
proposed
‘CpTa(C-
the possibility
of

T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
133
CpTa(CCMe3)(CH2CMe3)(PMe3) was not considered.
Neopentylidene ligands in complexes such a trigonal
bipyramidal Ta(CHCMe3)2(PMe3)2X (X=Cl, CH2-
CMe3, etc.) [13] also equilibrate readily, but on the
basis of the reaction of Ta(CHCMe3)2(PMe3)2(CD2-
CMe3) with acetone to give ca. two equivalents of
Me2CꢁCHCMe3-d0 alkylidene equilibration was as-
cribed to alkylidene ligand rotation. There would seem
to be a significant possibility that neopentylidene lig-
ands equilibrate in some circumstances, if not in all
alkyl alkylidyne complexes and tantalum bisneopentyli-
dene complexes will be necessary before firm conclu-
sions can be reached.
4. Experimental section
4.1. General procedures
All experiments were performed under nitrogen in a
Vacuum Atmospheres drybox or under argon using
standard Schlenk techniques. All solvents were purified
by standard techniques while deuterated NMR solvents
circumstances, via
h proton migration from one
neopentylidene ligand to another to form a neopentyl
neopentylidyne intermediate. Therefore the energy dif-
ference between CpTa(CCMe3)(CH2CMe3)(PMe3) and
CpTa(CHCMe3)2(PMe3) may not be large. The tanta-
lum systems clearly warrant further study before any
definitive conclusions can be drawn. In any case tanta-
lum bisneopentylidene complexes are the lower energy
species. We speculate that the efficiency of an h agostic
interaction in at least one of the neopentylidene ligands
in CpTa(CHCMe3)2(PMe3) is the reason why Cp-
Ta(CHCMe3)2(PMe3) is stabilized relative to CpTa(C-
CMe3)(CH2CMe3)(PMe3). The recently reported
rearrangement of a tantalum ethylene complex to its
ethylidene tautomer [4,58] also was rationalized in
terms of a strong h agostic interaction and the resulting
18 electron count. We further speculate that (for exam-
ple) CpW(CAd)(CH2CMe3)Cl is lower in energy than
CpW(CHAd)(CHCMe3)Cl because of the inherent
strength of the WꢀC bond but perhaps primarily be-
cause of a weaker h agostic interaction in one of the
alkylidene ligands in 16 electron CpW(CHAd)(CHC-
Me3)Cl than in a 16 electron tantalum bisalkylidene
complex.
The original goal was to observe a proton migration
analogous to that proposed for conversion of W(C-
SiMe3)(CH2CMe3)3 to W(CCMe3)(CH2CMe3)2(CH2Si-
Me3) [11]. In this context it is worth noting that h
proton scrambling in CpW(CAd)(CH2CMe3)2 (ca. 5×
10−5 s−1 at 298 K) appears to be significantly faster
than h proton migration in W(CSiMe3)(CH2CMe3)3
(ca. 1.7×10−5 s−1 at 352 K) or h proton migration in
Ta(CDCMe3)(CH2CMe3)3 (ca. 3.5×10−5 s−1 at 348
K). In view of these results and the uncertainty con-
cerning the mechanism of neopentylidene ligand equili-
bration in CpTa(CHCMe3)2(PMe3), we may not any
longer be justified in saying that h proton migration
from an alkyl ligand to an alkylidyne ligand in ‘d0’
systems is necessarily inherently ‘slow’. At the same
time it should be pointed out that h-hydrogen migra-
tion in the CpW(CAd)(CH2CMe3)X systems conceiv-
ably could involve migration of Hh first to the
cyclopentadienyl ligand and then to the alkylidyne’s Ch,
although this possibility is not thought to be very likely.
In any case detailed kinetic studies of both tungsten
˚
were dried and stored over activated 4 A molecular
sieves before use.
W2(OCMe3)6
was
prepared
by
adding
Na[W2Cl7(THF)5] [33] (prepared in situ in THF from
WCl4 and sodium [32]) to a THF solution of six
equivalents of LiOCMe3 (prepared from Li metal and
t-butanol in hexane). 1-Adamantanecarbonitrile,
trimethylsilyl triflate, N,N-dimethylaminotrimethylsi-
lane, and N-t-butylaminotrimethylsilane were pur-
chased commercially and used as received. Solutions of
benzylmagnesium chloride and methylmagnesium chlo-
ride were purchased from Aldrich and titrated before
use. Neopentylmagnesium chloride [59] and neopentyl-
lithium [60] were prepared as described in the literature.
Proton spectra were referenced internally by the
residual solvent proton signal relative to tetramethylsi-
lane. Carbon spectra were referenced internally relative
to the 13C signal of the NMR solvent relative to te-
tramethylsilane. All spectra were run in C6D6, except
where noted otherwise. All coupling constants are re-
ported in Hz; those found in Tables 1 and 4 are not
reported below. IR spectra were recorded as Nujol
mulls between KBr plates on a Perkin-Elmer 1600
FT-IR spectrometer. Elemental analyses (C, H, N) were
performed in our laboratories using a Perkin-Elmer
PE2400 microanalyzer or by Oneida Research Services,
Whitesboro, New York.
4.2. W(CAd)(OCMe3)3 (1)
1-Adamantanecarbonitrile (2.91 g, 18.08 mmol) was
added to a solution of W2(OCMe3)6 (14.58 g, 18.08
mmol) in pentane (175 ml. The deep red color of
W2(OCMe3)6 soon faded as a voluminous white precip-
itate formed. The mixture was stirred overnight and the
[W(N)(OCMe3)3]x collected by filtration. The filtrate
was concentrated in vacuo and cooled to −40°C to
afford 8.0 g (80%) of colorless W(CAd)(OCMe3)3 in
three crops. An analytical sample was obtained by
1
double recrystallization from pentane: H-NMR l 2.08
(br, 6, Ad), 2.00 (br, 3, Ad), 1.62 (br, q 6, Ad), 1.503 (s,
27, OCMe3); 13C{1H}-NMR l 272.1 (JCW=294.1,
CAd), 79.0 (OCMe3), 52.7 (CAd), 47.0 (Ad), 37.3 (Ad),

134
T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
33.0 (OCMe3), 29.9 (Ad). Anal. Calc. for C23H42O3W:
C, 50.19; H, 7.69. Found C, 50.03; H, 7.70.
removed in vacuo from the intensely purple solution,
and the resulting purple solid was extracted with a
50/50 mixture of ether and pentane (125 ml). Purple
crystals formed upon concentrating the solution to ca.
10 ml. The solution was stood overnight at −40°C and
4.3. W(CAd)(OCMe3)2(OTf )(dme) (2)
1
Trimethylsilyl triflate (2.92 g, 13.2 mmol) was added
dropwise to a solution of 1 (7.24 g, 3.87 mmol) and
dimethoxyethane (3.62 g) in pentane (125 ml) at −
40°C. Bright yellow microcrystals began to appear after
addition began. The mixture was stood for 1 h at
–40°C and the microcrystals were collected by filtration
and washed with pentane (3×15 ml); yield 9.03 g
(96%). The product is thermally sensitive and should be
stored below room temperature or used immediately:
1H-NMR l 3.98 (s, 3, MeOCH2), 3.57 (br, 1,
MeOCH2), 3.08 (s, 3, MeOCH2), 3.02 (br, 1,
MeOCH2), 2.96 (br, 1, MeOCH2), 2.04 (br, 9, OCMe3),
1.99 (br, 3, Ad), 1.60 (br, 9, OCMe3), 1.53 (br, 6, Ad),
1.49 (br, 6, Ad); 13C{1H}-NMR l 288.1 (JCW=273.5,
CAd), 120.6 (JCF=318.4, O3SCF3), 80.13 (br,
OCMe3), 80.08 (br, OCMe3), 75.2 (MeOCH2), 73.4
(MeOCH2), 69.5 (MeOCH2), 59.2 (MeOCH2), 52.2
(JCW=40.0, CAd), 46.6 (Ad), 37.1 (Ad), 33.1 (br,
OCMe3), 32.9 (br, OCMe3), 29.7 (Ad). An analytical
sample was recrystallized from toluene at −40°C.
Anal. Calc. for C24H43F3O7SW: C, 40.23; H, 6.05.
Found: C, 40.27; H, 6.27.
2.10 g (81%) of purple product was collected: H-NMR
l 5.72 (s, 5, Cp), 1.92 (br, 3, Ad), 1.62 (br d, 6, Ad),
1.47 (br t, 6, Ad); 13C{1H}-NMR l 328.6 (JCW=233.4,
CAd), 106.1 (Cp), 51.3 (JCW=32.9, CAd), 44.7 (Ad),
36.6 (Ad), 29.0 (Ad). An analytical sample was recrys-
tallized from ether/dichloromethane at −40°C. Anal.
Calc. for C16H20Cl2W: C, 41.14; H, 4.32. Found: C,
41.17; H, 4.42.
4.6. CpW(CAd)(CH2CMe3)2 (5a)
Neopentylmagnesium chloride (0.259 ml, 2.97 M in
ether, 0.768 mmol) was added to a solution of 4 (0.175
g, 0.375 mmol) in ether (10 ml) at −40°C. The solution
immediately turned cherry red and became turbid. The
mixture was stirred at room temperature for 1h and
dioxane (0.070 ml, 0.82 mmol) was added and the
suspension was filtered through Celite. The filtrate was
concentrated to dryness and the solid was extracted
with pentane (10 ml). The mixture was filtered again
and the volatile components were once again removed
in vacuo to afford a red powder which was recrystal-
lized from pentane at −40°C to afford 0.143 g (71%)
1
4.4. CpW(CAd)(OCMe3)2 (3)
of small red crystals in two crops: H-NMR l 5.40 (s,
5, Cp), 3.54 (JAB+JAB%=11.3, 2, CH2CMe3), 1.99 (br,
9, Ad), 1.64 (br, 6 Ad), 1.30 (s, 18, CMe3), −1.14
(JBB%=4JHH=2.1, 2JWH=12.6, 2, CH2CMe3); 13C-
NMR l 300.4 (CAd), 100.19 (Cp), 90.67 (JCW=94.9,
CH2CMe3), 51.98 (CAd), 45.18, 37.18 (Ad), 35.32
(CMe3), 34.37 (CMe3), 29.52 (Ad). Anal. Calc. for
C26H42W: C, 58.00; H, 7.86. Found: C, 57.74; H, 7.73.
A solution of NaCp in THF (4.6 ml, 2.0 M, 9.1
mmol) was added to a solution of 2 (6.50g, 9.11 mmol)
in THF (70 ml) at −40°C. The resulting light orange
solution was stirred overnight at room temperature.
The volatiles were removed in vacuo and the residue
was extracted with pentane (100 ml). The mixture was
filtered and the solvent was removed from the filtrate in
vacuo to give a yellow solid. This solid was recrystal-
lized from pentane to yielded 4.88 g (99%) of the
4.7. CpW(CAd)(CH2Ph)2 (5b)
1
product: H-NMR l 6.071 (s, 5, Cp), 1.97 (br, 3, Ad),
Benzylmagnesium chloride (0.593 ml, 1.11 M in
ether, 0.658 mmol) was added to a chilled (−40°C)
solution of 4 (0.150 g, 0.321 mmol) in ether (10 ml).
The solution immediately turned yellow. After 1 h
dioxane (0.057 ml, 0.66 mmol) was added and the
volatile components were removed in vacuo. The
residue was extracted with toluene (25 ml) and the
extract was filtered through Celite. The filtrate was
concentrated to ca. 1.5 ml and stood at −40°C to
afford 0.147 g (79%) of yellow crystals: 1H-NMR
(CD2Cl2) l 7.27 (t, 1, Php), 7.05 (d, 2, Pho), 6.96 (t, 2,
1.83 (d, 6, Ad), 1.60 (br, 6, Ad), 1.41 (s, 18, OCMe3);
13C{1H}-NMR l 285.2 (JCW=277.2, CAd), 105.7 (Cp),
76.7 (OCMe3), 51.8 (JCW=43.5, CAd), 46.6 (Ad), 37.4
(Ad), 32.8 (OCMe3), 29.9 (Ad). An analytical sample
was prepared by recrystallizing a sample from pentane.
Anal. Calc. for C24H38O2W: C, 53.15; H, 7.06. Found:
C, 53.42; H, 7.24.
4.5. CpW(CAd)Cl2 (4)
2
2
A solution of 2,6-lutidinium chloride (0.500 g, 3.48
mmol) and trimethysilyl chloride (3.00 g, 27.6 mmol) in
dichloromethane (20 ml) was added to 3 (3.00g, 5.53
mmol) in dichloromethane (15 ml). The solution imme-
diately turned deep red and slowly became purple over
a period of ca. 4 h. After 12 h the volatile solvents were
Phm), 5.51 (s, 5, Cp), 2.42 (d, JHH=7.5, JWH=4.8, 2,
CH2Ph,), 1.94 (br, 3, Ad), 1.69 (d, 6, Ad), 1.63 (t, 6,
2
2
Ad), 1.48 (d, JHH=7.5, JWH=8.1, 2, CH2Ph); 13C-
NMR l 294.5 (WCAd), 135.6 (Ci), 131.4, (Co), 127.5
(Cp), 125.6 (Cm), 97.7 (Cp), 53.2 (CAd), 42.9 (Ad), 37.2
(Ad), 32.4 (JCW=63.3, CH2Ph), 29.2 (Ad). An analyti-

T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
135
cal
sample
was
recrystallized
from
ether/
chloride (0.030 g, 0.21 mmol) in dichloromethane (3
ml). The mixture was stood overnight and the volatile
components were removed in vacuo the next day. The
resulting residue was extracted with ether and filtered
through Celite to remove unreacted 2,6-lutidinium chlo-
ride. The sample was taken to dryness in vacuo and
dissolved in benzene-d6. A 2.0 to 1 ratio of CpW(C-
CMe3)Cl2 and CpW(CAd)Cl2 was measured by inte-
grating the Cp resonances at l 5.63 and 5.69 ppm,
respectively.
dichloromethane at −40°C. Anal. Calc. for C30H34W:
C, 62.29; H, 5.92. Found: C, 62.41; H, 5.86.
4.8. CpW(CAd)Me2 (5c)
A solution of methylmagnesium chloride (0.228 ml,
2.89 M in THF, 0.658 mmol) was added to a solution
of CpW(CAd)Cl2 (0.150 g, 0.321 mmol) in THF (3 ml)
at −40°C. After standing the solution for ca. 10 min at
−40°C, dioxane (0.070 ml, 0.82 mmol) was added and
the solvents were removed in vacuo. The residue was
was extracted with pentane (10 ml) and the extract was
filtered and taken to dryness in vacuo. The resulting red
powder was recrystallized from ether/pentane at −
40°C to yield 0.099 g (72%) of red crystals in two crops:
1H-NMR l 5.42 (s, 5, Cp), 2.03 (br, 3, Ad), 1.89 (d, 6,
Ad), 1.67 (br, 6, Ad), 1.13 (s, 6, Me,); 13C-NMR l
300.4 (JCW=240, CAd), 100.14 (Cp), 50.79 (CAd),
44.72 (Ad), 40.25 (JCW=93, W–CH3), 37.25, 29.37
(Ad). Anal. Calc. for C18H26W: C, 50.72; H, 6.14.
Found: C, 50.97; H, 5.97.
2
1H-NMR (5a%) l 5.37 (s, 5, Cp), 3.56 (d, JHH=11.7,
2
1, CH2), 3.28 (d, JHH=12.3, 1, CH2), 2.11 (m, 3, Ad),
(br d of t, 3, Ad), 1.77 (t, 6 Ad), 1.623 (br d of t, 3, Ad),
1.35 (CMe3), 1.273 (CMe3), −0.88 (dd, 2JHH=12.3
Hz, 4JHH=2.6, 1, CH2), −1.50 (dd, 2JHH=11.7,
4JHH=2.6, 1, CH2); 13C-NMR l 299.62 (CAd), 100.39
(Cp), 94.98 (JCW=93.5, CH2), 87.78 (JCW=95.6,
CH2), 49.74 (CAd), 47.21 (CAd), 37.84 (Ad), 35.57,
33.48 (CMe3), 30.42 (Ad).
4.11. Con6ersion of 6 to a mixture containing
CpW(CCMe3)(CH2Ad)Cl (6%)
4.9. CpW(CAd)(CH2CMe3)Cl (6)
6 (0.072 g, 0.093 mmol) was dissolved in benzene-d6
(0.60 ml) and a small amount of C6H6 was added as an
internal standard. The solution was transferred to a
Teflon-sealed NMR tube and monitored periodically
over 3 weeks while standing at room temperature. After
14 days an equilibrium mixture of 6 and 6% was reached
in the ratio of 1:1.05 with minimal sample decomposi-
tion. After 3 weeks the mixture was quenched with
2,6-lutidinium chloride as described for the mixture of
5a and 5a% to give a 1.0:1 ratio of CpW(CCMe3)Cl2 and
CpW(CAd)Cl2.
Neopentylmagnesium chloride (0.108 ml, 2.97 M in
ether, 0.321 mmol) was added to a stirred solution of 4
(0.150 g, 0.321 mmol) in ether (10 ml) at −40°C. The
solution turned red immediately and became turbid.
The solution was stood at room temperature for 15 min
and dioxane (0.028 ml, 0.32 mmol) was added. The
volatile components were removed in vacuo and the
residue was extracted with pentane (10 ml). THe extract
was filtered and concentrated to dryness. The resulting
red powder was recrystallized from pentane at −40°C
to yield 0.123 g (76%) of red microcrystals in two crops:
1H-NMR l 5.52 (s, 5, Cp), 4.79 (d, 2JHH=10.3,
2JWH=4.1, 1, CH2CMe3), 1.95 (br, 3, Ad), 1.81 (br, 6,
Ad), 1.56 (br, 6, Ad), 1.18 (s, 9, CH2CMe3), −0.45 (d,
2JHH=10.3, 2JWH=13.0, 1, CH2CMe3); 13C-NMR l
2
1H-NMR (6%) l 5.50 (s, 5, Cp), 4.72 (d, JHH=9.9
2
Hz, JWH=4.4 Hz, 1, CH2Ad), 2.05 (br, 3, Ad), 1.80
(br d of m, 6, Ad), 1.71 (br, 6, Ad), 1.41 (br d of m, 3
2
2
Ad), 1.23 (s, 9, CCMe3), −0.63 (d, JHH=9.9, JWH
=
12.8, 1, CH2Ad); 13C-NMR l 312.8 (JCW=242, CAd),
101.89 (Cp), 100.58 (CH2Ad), 49.35 (CAd), 46.15 (Ad),
38.62 (CCMe3), 37.37 (Ad), 32.99 (CH2CMe3), 29.96
(Ad).
313.6 (JCW=242, CAd), 101.80 (Cp), 98.08 (JCW
=
93.8, CH2CMe3), 51.74 (CAd), 45.04 (Ad), 36.88 (Ad),
36.49 (CCMe3), 33.51 (CH2CMe3), 29.25 (Ad). Anal.
Calc. for C21H31ClW: C, 50.17; H, 6.21. Found: C,
50.24; H, 6.01.
4.12. CpW(CAd)(NMe2)Cl (7a)
Me3SiNMe2 (60 ml, 0.37 mmol) was added to a
solution of 4 (0.160 g, 0.343 mmol) at −40°C in
dichloromethane (5 ml). The solution turned from bril-
liant purple to light yellow over a period of 15 min. The
mixture was stirred for 1h and the volatile components
were removed in vacuo. The residue was extracted with
ether (10 ml) and the extracts were filtered and the
filtrate was concentrated and cooling to −40°C to
afford 0.142 g (87%) of yellow crystals: 1H-NMR l 5.73
(s, 5, Cp), 4.25 (s, 3, NMe2), 3.04 (s, 3, NMe2), 1.93 (br,
3, Ad), 1.78 (d, 6, Ad), 1.57 (t, 6, Ad); 13C{1H}-NMR
4.10. Con6ersion of 5a to a mixture containing
CpW(CCMe3)(CH2CMe3)(CH2Ad) (5a%)
A solution of 5a (0.050 g, 0.093 mmol) in benzene-d6
(0.60 ml) and a small amount of C6H6 was added as an
internal standard. The solution was transferred to a
Teflon-sealed NMR tube and monitored periodically
over 2.5 days. After 48 h an equilibrium mixture of 5a
and 5a% was reached; the ratio was 1:2.1. After 55 h, the
solution was added to a solution of 2,6-lutidinium

136
T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
2
2
l 302.3 (JCW=257.7, CAd), 102.2 (Cp), 71.1 (NMe2),
57.1 (NMe2), 52.5 (JCW=36.4, CAd), 44.2 (Ad), 37.0
(Ad), 29.0 (Ad). An analytical sample was recrystallized
from ether at -40 °C. Anal. Calc. for C18H26NClW: C,
45.45; H, 5.51; N, 2.94. Found: C, 45.50; H, 5.37; N,
2.72.
5.50 (s, 5, Cp), 4.72 (d, JHH=9.9 Hz, JWH=4.4 Hz,
1, CH2Ad), 2.05 (br, 3, Ad), 1.80 (br d of m, 6, Ad),
1.71 (br, 6, Ad), 1.41 (br d of m, 3 Ad), 1.23 (s, 9,
CCMe3), −0.63 (d, 2JHH=9.9, 2JWH=12.8, 1,
CH2Ad); 13C–NMR l 274.1 (JCW=154.7 in CDCl3),
CAd), 103.82 (Cp), 69.81 (NCMe3), 46.65 (Ad), 45.82
(CAd), 37.40 (Ad), 31.77 (CMe3), 30.23 (Ad).
Complete NMR data for syn-CpW(NCMe3)
(CHAd)Cl (8%) could not be obtained. Selected data can
be found in Table 4.
4.13. CpW(CAd)(NMe2)(CH2CMe3) (7b)
Neopentyllithium (0.012 g, 0.15 mmol) in ether (1 ml)
was added with stirring to a solution of 7a (0.060 g,
0.13 mmol) in ether (3 ml) at −40°C. The solution
became cloudy immediately. The mixture was stood for
6 h at −40°C and then taken to dryness in vacuo.
Dichloromethane was added in order to destroy any
excess neopentyllithium. The residue was extracted with
pentane and the extract was filtered. The extracts were
taken to dryness in vacuo. The residue was recrystal-
lized from pentane at −40°C to afford 0.045 g (70%)
5. Supporting information available
Crystal data and structure refinement, atomic coordi-
nates, bond lengths and angles, and anisotropic dis-
placement parameters, for 5a.
1
of orange crystals: H-NMR (CD2Cl2) l 5.78 (s, 5, Cp),
Acknowledgements
4.10 (s, 3, NMe2), 3.16 (s, 3, NMe2), 1.96 (br, 3, Ad),
1.77 (d, 6, Ad), 1.69 (AB, 2JWH=11.7 Hz, 1,
We thank the National Science Foundation for sup-
port (CHE 91-22827 and 97-00736) and for funds to
help purchase a departmental Siemens SMART/CCD
diffractometer.
2
CH2CMe3), 1.63 (t, 6, Ad), 1.53 (AB, JWH=8.7 Hz, 1,
CH2CMe3), 1.02 (s, 9, CH2CMe3); 13C{1H}-NMR l
291.7 (JCW=254.7 Hz, CAd), 100.63 (Cp), 71.00
(NMe2), 56.27 (NMe2), 52.27 (CAd), 48.60 (JCW
=
118.6, CH2CMe3), 45.15, 37.46 (Ad), 35.52
(CH2CMe3), 35.11 (CH2CMe3), 29.63 (Ad). Anal. Calc.
for C23H37NW: C, 54.02; H, 7.29; N, 2.74. Found: C,
54.18; H, 7.51; N, 2.62.
References
[1] R.R. Schrock, in: P.R. Braterman (Ed.), Reactions of Coordi-
nated Ligands, Plenum, New York, 1986, p. 221.
[2] J. Feldman, R.R. Schrock, Prog. Inorg. Chem. 39 (1991) 1.
[3] W.A. Nugent, J.A. Mayer, Metal–Ligand Multiple Bonds, Wi-
ley, New York, 1988.
4.14. CpW(CAd)(NH–t-Bu)Cl (8)
Me3SiNH(t-Bu) (0.195 g, 1.46 mmol) was added to a
solution of 4 (0.500 g, 1.07 mmol) in dichloromethane
(15 ml). After 30 min the volatiles were removed from
the solution in vacuo. The residue was extracted with
ether, and the extracts were filtered, concentrated in
vacuo and cooled to yield 0.463 g (88%) of yellow
[4] J.S. Freundlich, R.R. Schrock, W.M. Davis, J. Am. Chem. Soc.
118 (1996) 3643.
[5] J.S. Freundlich, R.R. Schrock, W.M. Davis, Organometallics 15
(1996) 2777.
[6] M. Brookhart, M.L.H. Green, L.-L. Wong, Prog. Inorg. Chem.
36 (1988) 1.
[7] J.S. Murdzek, R.R. Schrock, in: Carbyne Complexes, VCH,
New York, 1988, p. 147.
1
crystals: H-NMR (CD2Cl2) l 9.40 (br, 1, NH), 6.077
[8] E. Tran, P. Legzdins, J. Am. Chem. Soc. 119 (1997) 5071.
[9] J. Cheon, D.M. Rogers, G.S. Girolami, J. Am. Chem. Soc. 119
(1997) 6804.
[10] R.R. Schrock, J.D. Fellmann, J. Am. Chem. Soc. 100 (1978)
3359.
[11] K.G. Caulton, M.H. Chisholm, W.E. Streib, Z. Hue, J. Am.
Chem. Soc. 113 (1991) 6082.
[12] J.D. Fellmann, G.A. Rupprecht, C.D. Wood, R.R. Schrock, J.
Am. Chem. Soc. 100 (1978) 5964.
(s, 5, Cp), 2.00 (br, 3, Ad), 1.79 (d, 6, Ad), 1.64 (t, 6,
Ad), 1.39 (s, 9, NH–t-Bu); 13C-NMR l 303.54 (CAd),
103.01 (Cp), 59.53 (NCMe3), 52.72 (CAd), 44.49 (Ad),
37.05 (CMe3), 33.97, 29.36 (Ad); IR (Nujol/KBr) 3249
w(NH) cm−1. Anal. Calc. for C20H30NClW: C, 47.68;
H, 6.00; N, 2.78. Found: C, 48.04; H, 6.05; N, 2.67.
4.15. anti-CpW(NCMe3)(CHAd)Cl (8%%)
[13] J.D. Fellmann, R.R. Schrock, G.A. Rupprecht, J. Am. Chem.
Soc. 103 (1981) 5752.
A saturated solution of 8 in dichloromethane-d2 (ca.
[14] M.R. Churchill, W.J. Youngs, Inorg. Chem. 18 (1979) 1930.
[15] J.W. Lauher, R. Hoffmann, J. Am. Chem. Soc. 98 (1976) 1729.
[16] G. Wilkinson, F.G.A. Stone, E.W. Abel, Comprehensive
Organometallic Chemistry, Pergamon Press, Oxford, 1982.
[17] D.D. Devore, J.D. Lichtenhan, F. Takusagawa, E.A. Maatta, J.
Am. Chem. Soc. 109 (1987) 7408.
0.020 g in 0.60 ml) was allowed to stand in a Teflon-
1
sealed NMR tube for 11 days. After 3 days H reso-
nances for 8 had completely disappeared and a mixture
of 8% and 8%% had formed. After 10 days only 8%% was
present and B5% of the sample had decomposed as
[18] F. Preuss, H. Becker, H.-J. Ha¨usler, Z. Naturforsch. 42b (1987)
881.
1
measured by a CH2Cl2 internal standard: H-NMR l

T.H. Warren et al. / Journal of Organometallic Chemistry 569 (1998) 125–137
[19] A.D. Poole, D.N. Williams, A.M. Kenwright, V.C. Gibson, W. [40] M. Barfield, J. Chem. Phys. 41 (1964) 3825.
137
Clegg, D.C.R. Hockless, P.A. O’Neil, Organometallics 12 (1993)
2549.
[20] D.N. Williams, J.P. Mitchell, A.D. Poole, U. Siemling, W.
Clegg, D.C.R. Hockless, P.A. O’Neil, V.C. Gibson, J. Chem.
Soc., Dalton Trans. (1992) 739.
[41] L.M. Jackman, S. Sternhell, Applications of Nuclear Magnetic
Resonance Spectroscopy in Organic Chemistry, Pergamon, Ox-
ford, 1969.
[42] K.B. Wiberg, B.R. Lowry, B.J. Nist, J. Am. Chem. Soc. 84
(1962) 1594.
[21] A.D. Poole, V.C. Gibson, W. Clegg, J. Chem. Soc., Chem.
Commun., (1992) 237.
[43] J. Meinwald, A. Lewis, J. Am. Chem. Soc. 83 (1961) 2769.
[44] J. Meinwald, Y.C. Meinwald, J. Am. Chem. Soc. 85 (1963) 2514.
[45] Z. Xue, L. Li, L.K. Hoyt, J.B. Diminnie, J.L. Pollitte, J. Am.
Chem. Soc. 116 (1994) 2169.
[46] M. Etienne, Organometallics 13 (1994) 410.
[47] M. Etienne, F. Biasotto, R. Mathieu, J.L. Templeton,
Organometallics 15 (1996) 1106.
[48] M. Etienne, R. Mathieu, B. Donnadieu, J. Am. Chem. Soc. 119
(1997) 3218.
[49] R.R. Schrock, R.T. DePue, J. Feldman, K.B. Yap, D.C. Yang,
W.M. Davis, L. Park, M. DiMare, M. Schofield, J. Anhaus, E.
Walborsky, E. Evitt, C. Kru¨ger, P. Betz, Organometallics 9
(1990) 2262.
[50] M.R. Churchill, W.J. Youngs, Inorg. Chem. 18 (1979) 2454.
[51] M.H. Chisholm, J.C. Huffmann, J.A. Klang, Polyhedron 9
(1990) 1271.
[52] S.F. Pedersen, R.R. Schrock, J. Am. Chem. Soc. 104 (1982)
7483.
[53] R.R. Schrock, Polyhedron 14 (1995) 3177.
[54] R.R. Schrock, in: D.J. Brunelle, Ring-Opening Polymerization,
Hanser, Munich, 1993, p. 129.
[55] R.R. Schrock, L.W. Messerle, C.D. Wood, L.J. Guggenberger,
J. Am. Chem. Soc. 100 (1978) 3793.
[22] M.P. Coles, V.C. Gibson, W. Clegg, M.R.J. Elsegood, P.A.
Porrelli, J. Chem. Soc., Chem. Commun. (1996) 1963.
[23] W.A. Nugent, Inorg. Chem. 22 (1983) 965.
[24] G. Schoettel, J. Kress, J.A. Osborn, J. Chem. Soc., Chem.
Commun. (1989) 1062.
[25] H.H. Fox, K.B. Yap, J. Robbins, S. Cai, R.R. Schrock, Inorg.
Chem. 31 (1992) 2287.
[26] D.S. Williams, M.H. Schofield, R.R. Schrock, Organometallics
12 (1993) 4560.
[27] P.W. Dyer, V.C. Gibson, J.A.K. Howard, B. Whittle, C. Wilson,
J. Chem. Soc., Chem. Commun. (1992) 1666.
[28] V.C. Gibson, J. Chem. Soc., Dalton Trans. (1994) 1607.
[29] R.R. Schrock, M.L. Listemann, L.G. Sturgeoff, J. Am. Chem.
Soc. 1982 (1982) 4291.
[30] M.L. Listemann, R.R. Schrock, Organometallics 4 (1985) 74.
[31] M. Akiyama, M.H. Chisholm, F.A. Cotton, M.W. Extine, D.A.
Haitko, D. Little, P.E. Fanwick, Inorg. Chem. 18 (1979) 2266.
[32] P.R. Sharp, R.R. Schrock, J. Am. Chem. Soc. 102 (1980) 1430.
[33] M.H. Chisholm, B.W. Eichhorn, K. Folting, J.C. Huffman, C.D.
Ontiveros, W.E. Streib, W.G. Van Der Sluys, Inorg. Chem. 26
(1987) 3182.
[34] D.M.T. Chan, M.H. Chisholm, K. Folting, J.C. Huffman, N.S.
Marchant, Inorganic Chemistry 25 (1986) 4170.
[35] J.H. Freudenberger, R.R. Schrock, Organometallics 4 (1985)
1937.
[36] S. Pedersen, Ph.D.Thesis, Massachusetts Institute of Technol-
ogy, 1983.
[56] A.H. Liu, R.C. Murray, J.C. Dewan, B.D. Santarsiero, R.R.
Schrock, J. Am. Chem. Soc. 109 (1987) 4282.
[57] D.C. Maus, V. Copie´, B. Sun, J.M. Griffiths, R.G. Griffin, S.
Luo, R.R. Schrock, A.H. Liu, S. Seidel, W.M. Davis, A.
Grohmann, J. Am. Chem. Soc. 118 (1996) 5665.
[58] J.S. Freundlich, R.R. Schrock, C.C. Cummins, W.M. Davis, J.
Am. Chem. Soc. 116 (1994) 6476.
[37] M.R. Churchill, J.W. Ziller, L. McCullough, S.F. Pedersen, R.R.
Schrock, Organometallics 2 (1983) 1046.
[59] R.R. Schrock, J. Sancho, S.F. Pedersen, Inorg. Syn. 26 (1989)
44.
[38] L.J. Procopio, P.J. Carroll, D.H. Berry, J. Am. Chem. Soc. 116
(1994) 177.
[60] R.R. Schrock, J.D. Fellmann, J. Am. Chem. Soc. 100 (1978)
3359.
[39] H. Friebolin, Basic One- and Two-Dimensional NMR Spec-
troscopy, VCH, New York, 1993.
.