Synthesis and decomposition of alkyl complexes of tungsten(IV) that contain a [(Me3SiNCH2CH2)3N]3- ligand

Article abstract of DOI:10.1021/om970670m

The synthesis of a variety of tungsten alkyl complexes of the type [N₃N]WR ([N₃N]^3- = [(Me₃SiNCH₂CH₂)₃N]^3-; R = Me, Et, Bu, CH₂Ph, CH₂SiMe₃, CH₂CMe₃, cyclopropyl, cyclobutyl, cyclopentyl, cyclohexyl, cyclopentenyl) was attempted by alkylation of [N₃N]WCl complexes. Only the methyl, phenyl, and cyclopentenyl complexes could be isolated. The susceptibility of the [N₃N]WR complexes is similar to that of [N₃N]WCl; the effective moment is depressed as a consequence of a combination of spin-orbit effects and low-symmetry ligand-field components that result in zero-field splitting of the d² ground-state spin triplet. No linear alkyl complexes could be observed as a consequence of α,α-dehydrogenation to give molecular hydrogen and alkylidyne complexes, [N₃N]W≡CR′. [N₃N]W(cyclopropyl) evolves ethylene in a first-order process to give [N₃N]W≡CH, while [N₃N]W(cyclobutyl) is converted into the 1-tungstacyclopentene complex, [N₃N]W(CHCH₂CH₂CH₂), as confirmed in an X-ray study. [N₃N]W(CHCH₂CH₂CH₂) decomposes readily in a first-order manner to give [N₃N]W≡CCH₂CH₂CH₃. An attempt to prepare [N₃N]W(cyclopentyl) led to formation of [N₃N]W(cyclopentylidene)(H), as confirmed in an X-ray study. NMR studies suggest that [N₃N]W(cyclopentylidene)(H) is in equilibrium with [N₃N]W(cyclopentyl) where ΔH°= 11.8-(6) kcal mol^-1 and ΔS° = 33(2) eu or K_eq ≈ 0.1 at 46 °C. From these and other data it is concluded that the rate constant for α elimination in [N₃N]M(cyclopentyl) is approximately the same for Mo and W. The [N₃N]W(cyclopentenyl) complex is unstable, decomposing to give a complex containing a "bent imido" ligand, [(Me₃SiNCH₂CH₂)₂(NCH ₂CH₂)N]W(1-(trimethylsilyl)cyclopentene), as confirmed in an X-ray study.

Full text of DOI:10.1021/om970670m

Organometallics 1997, 16, 5195-5208
5195
Syn th esis a n d Decom p osition of Alk yl Com p lexes of
Tu n gsten (IV) Th a t Con ta in a [(Me3SiNCH2CH2)3N]3-
Liga n d
Richard R. Schrock,* Scott W. Seidel, Nadia C. Mo¨sch-Zanetti, Daniel A. Dobbs,
Keng-Yu Shih, and William M. Davis
Department of Chemistry 6-331, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139
Received August 4, 1997X
The synthesis of a variety of tungsten alkyl complexes of the type [N3N]WR ([N3N]3-
[(Me3SiNCH2CH2)3N]3-; R ) Me, Et, Bu, CH2Ph, CH2SiMe3, CH2CMe3, cyclopropyl, cyclobu-
tyl, cyclopentyl, cyclohexyl, cyclopentenyl) was attempted by alkylation of [N3N]WCl
complexes. Only the methyl, phenyl, and cyclopentenyl complexes could be isolated. The
susceptibility of the [N3N]WR complexes is similar to that of [N3N]WCl; the effective moment
is depressed as a consequence of a combination of spin-orbit effects and low-symmetry ligand-
field components that result in zero-field splitting of the d2 ground-state spin triplet. No
linear alkyl complexes could be observed as a consequence of R,R-dehydrogenation to give
molecular hydrogen and alkylidyne complexes, [N3N]WtCR′. [N3N]W(cyclopropyl) evolves
ethylene in a first-order process to give [N3N]WtCH, while [N3N]W(cyclobutyl) is converted
into the 1-tungstacyclopentene complex, [N3N]W(CHCH2CH2CH2), as confirmed in an X-ray
study. [N3N]W(CHCH2CH2CH2) decomposes readily in a first-order manner to give
[N3N]WtCCH2CH2CH3. An attempt to prepare [N3N]W(cyclopentyl) led to formation of
[N3N]W(cyclopentylidene)(H), as confirmed in an X-ray study. NMR studies suggest that
[N3N]W(cyclopentylidene)(H) is in equilibrium with [N3N]W(cyclopentyl) where ∆H° ) 11.8-
(6) kcal mol-1 and ∆S° ) 33(2) eu or Keq ≈ 0.1 at 46 °C. From these and other data it is
concluded that the rate constant for R elimination in [N3N]M(cyclopentyl) is approximately
the same for Mo and W. The [N3N]W(cyclopentenyl) complex is unstable, decomposing to
give a complex containing a “bent imido” ligand, [(Me3SiNCH2CH2)2(NCH2CH2)N]W(1-
(trimethylsilyl)cyclopentene), as confirmed in an X-ray study.
)
In tr od u ction
In a recent paper,11 we reported the synthesis of a
variety of molybdenum alkyl complexes, [N3N]MoR
([N3N]3- ) [(Me3SiNCH2CH2)3N]3-; R ) Me, Et, Bu,
CH2Ph, CH2SiMe3, CH2CMe3, cyclopropyl, cyclobutyl,
cyclopentyl, cyclohexyl, cyclopentenyl). In general, such
species are relatively stable, even when â protons are
present in the alkyl. They are all Curie-Weiss S ) 1
paramagnets down to 50 K, but below ∼50 K the
effective moments undergo a sharp decrease as a
consequence of what are proposed to be a combination
of spin-orbit coupling and zero-field splitting effects.
NMR spectra are temperature dependent as a conse-
quence of a “locking” of the [N3N]3- ligand backbone in
one C3-symmetric conformation and as a consequence
of Curie-Weiss behavior. The cyclopentyl and cyclo-
hexyl complexes show another type of temperature-
dependent behavior that can be ascribed to a rapid and
reversible R-elimination process to give an unobservable
cycloalkylidene hydride intermediate. It was shown
that the rate of R-elimination for the cyclopentyl com-
plex is more than 106 times faster than the rate of â
elimination, perhaps in part as a consequence of steric
acceleration of the R-elimination process and concomi-
tant steric deceleration of the â-elimination process.
Triamidoamine ligands, [(RNCH2CH2)3N]3-, in which
R is a bulky substituent (usually SiMe3 or C6F52), can
1
bind to a variety of transition metals,3 especially in
oxidation states 3+ or higher. Triamidoamine ligands
usually bind to a transition metal in a tetradentate
manner, thereby creating a sterically protected, 3-fold-
symmetric “pocket” in which only three orbitals are
available to bond to additional ligands in that pocket,
two degenerate π orbitals (approximately dxz and dyz)
2
and a σ orbital (approximately dz ). We have been
interested in the chemistry of Mo and W complexes that
contain the [(Me3SiNCH2CH2)3N]3- {[N3N]3-} ligand
from several perspectives, including dinitrogen fixa-
tion,4,5 terminal chalcogenide complexes,6-8 and new
organometallic chemistry.4,9,10
X Abstract published in Advance ACS Abstracts, November 1, 1997.
(1) Verkade, J . G. Acc. Chem. Res. 1993, 26, 483.
(2) Kol, M.; Schrock, R. R.; Kempe, R.; Davis, W. M. J . Am. Chem.
Soc. 1994, 116, 4382.
(3) Schrock, R. R. Acc. Chem. Res. 1997, 90, 9.
(4) Shih, K.-Y.; Schrock, R. R.; Kempe, R. J . Am. Chem. Soc. 1994,
116, 8804.
(5) O’Donoghue, M. B.; Zanetti, N. C.; Schrock, R. R.; Davis, W. M.
J . Am. Chem. Soc. 1997, 119, 2753.
(6) Zanetti, N. C.; Schrock, R. R.; Davis, W. M. Angew. Chem., Int.
Ed. Engl. 1995, 34, 2044.
(7) J ohnson-Carr, J . A.; Zanetti, N. C.; Schrock, R. R.; Hopkins, M.
D. J . Am. Chem. Soc. 1996, 118, 11305.
(8) Wu, G.; Rovnyak, K.; J ohnson, M. J . A.; Zanetti, N. C.; Musaev,
D. G.; Morokuma, K.; Schrock, R. R.; Griffin, R. G.; Cummins, C. C. J .
Am. Chem. Soc. 1996, 118, 10654.
(9) Shih, K.-Y.; Totland, K.; Seidel, S. W.; Schrock, R. R. J . Am.
Chem. Soc. 1994, 116, 12103.
(10) Schrock, R. R.; Shih, K.-Y.; Dobbs, D.; Davis, W. M. J . Am.
Chem. Soc. 1995, 117, 6609.
(11) Schrock, R. R.; Seidel, S. W.; Mo¨sch-Zanetti, N. C.; Shih, K.-Y.;
O’Donoghue, M. B.; Davis, W. M. J . Am. Chem. Soc., in press.
S0276-7333(97)00670-5 CCC: $14.00 © 1997 American Chemical Society

5196 Organometallics, Vol. 16, No. 24, 1997
Schrock et al.
Among the linear alkyl complexes, only [N3N]Mo(CH2-
CMe3) decomposes cleanly (albeit slowly) to give [N3N]-
MotCCMe3 and molecular hydrogen. Other decompo-
sitions of linear alkyl complexes are complicated by
competing reactions, including â-hydride elimination to
give the olefin and [N3N]MoH. â-hydride elimination
is the sole mode of decomposition of the cyclopentyl and
cyclohexyl complexes, and [N3N]MoH can be isolated in
high yield in each case. Other unusual decomposition
reactions that involve C-C bond cleavages were also
observed. For example, [N3N]Mo(cyclopropyl) evolved
ethylene upon being heated to give [N3N]MotCH, while
[N3N]Mo(cyclobutyl) was converted into [N3N]MotCCH2-
CH2CH3, both quantitatively.
F igu r e 1. Plot of the molar susceptibility versus temper-
ature for [N3N]WCl and [N3N]WH.
Attempts to prepare [N3N]WCH2R complexes led to
the surprising result that only [N3N]WtCR complexes
were formed when R was not a proton.9 Moreover,
attempts to prepare a cyclopentyl complex resulted in
formation of the cyclopentylidene hydride complex,
[N3N]W(C5H8)(H). The selective formation of [N3N]W-
(C5H8)(D) from Li-R-DC5H8 and [N3N]WCl at low tem-
peratures suggested that R elimination is faster than â
elimination.10 This paper contains the full report
concerning tungsten alkyl complexes that contain the
[N3N]3- ligand. These results are contrasted with those
obtained for the analogous Mo complexes, and an
attempt is made to rationalize any significant differ-
ences.
Resu lts
Tu n gsten Ha lid e Com p lexes. Addition of WCl4-
(dimethoxyethane) to Li3[N3N]12 in THF yielded para-
magnetic [N3N]WCl (1) in a maximum yield of ∼25%.
More typical yields are between 15 and 20%. Variations
of this reaction so far have failed to increase the yield
beyond 25%. As in the synthesis of [N3N]MoCl, we
suspect that loss of Me3SiCl and metal reduction are
persistent problems, in part because related reactions
between (C6F5NHCH2CH2)3N and WCl4 in the presence
of NEt3 give [(C6F5NCH2CH2)3N]WCl in high yield.2 The
reaction between [N3N]WCl and TMSI yields [N3N]WI,
so other derivatives (Br, triflate, etc.) most likely could
be prepared similarly. (It should be noted that [N3N]-
Mo(triflate) is a known compound that has been struc-
turally characterized.11)
A plot of the susceptibility of [N3N]WCl (along with
that for [N3N]WH, to be described later) versus T is
shown in Figure 1. The susceptibility of [N3N]WCl is
much smaller at a given temperature than the suscep-
tibility of [N3N]MoCl,11 especially at low temperatures.
The effective moment, therefore, is lower than the
expected spin-only value (2.83 µB) near room tempera-
ture and decreases smoothly as T approaches 5 K
(Figure 2). In contrast, [N3N]MoCl is a Curie-Weiss S
) 1 paramagnet down to 50 K, at which point the
effective moment decreases sharply. The decrease in
the effective magnetic moment was attributed to a
combination of spin-orbit effects and low-symmetry
ligand-field components that result in zero-field splitting
of the d2 ground-state spin triplet.11,13 The effect is
much larger for W than for Mo as a consequence of the
larger spin-orbit coupling constant for W4+ (λ(W4+) ≈
F igu r e 2. Plot of µeff versus temperature for [N3N]WCl
and [N3N]WH.
1050 cm-1, while λ(Mo4+) ≈ 475 cm-1).13 The two
electrons in the paramagnetic form of [N3N]WCl are
believed to be in the degenerate dxz and dyz orbitals. Two
measurements of the magnetic moment for [N3N]WCl
in solution at 22 °C by the Evan’s method gave values
of 2.4 and 2.5 µB, consistent with a magnetic moment
for [N3N]WCl that is reduced from the spin-only value.
1H NMR spectra of 1 show two characteristic upfield-
shifted resonances for the methylene groups on the
backbone of the triamidoamine ligand (at -29.83 and
-75.44 ppm at 22 °C) and a TMS resonance at 8.30 ppm
(at 22 °C). In general, 1H NMR resonances for tungsten
complexes are narrower than those for molybdenum
complexes, possibly as a consequence of the reduced
magnetic moment for tungsten species in general.
These spectra are temperature dependent in a manner
analogous to that described for [N3N]MoCl.11 (See also
[N3N]WMe below.) Two basic temperature-dependent
processes are observed. The [N3N]3- ligand’s methylene
and TMS resonances (to a lesser degree) shift to higher
field and lower field, respectively, as the temperature
is lowered. Second, at low temperature, four methylene
backbone resonances are observed as a consequence of
a process in which the WNCCN five-membered rings
within the [N3N]3- ligand system become locked in some
“puckered” configuration that gives the molecule an
overall C3 symmetry.
An X-ray structure of [N3N]WCl14 revealed that it has
a C3-symmetric structure that is virtually identical to
(12) Cummins, C. C.; Schrock, R. R.; Davis, W. M. Organometallics
1992, 11, 1452.
(13) Figgis, B. N.; Lewis, J . Prog. Inorg. Chem. 1964, 6, 37.
(14) Scheer, M.; Mu¨ller, J .; Ha¨ser, M. Angew. Chem., Int. Ed. Engl.
1996, 35, 2492.

Alkyl Complexes of Tungsten(IV)
Organometallics, Vol. 16, No. 24, 1997 5197
Ta ble 1. ∆Hq a n d ∆Sq Va lu es for Decom p osition
Rea ction s
∆Hq
(kcal
∆Sq 104k (s-1
(eu)b at 341 Kc
)
mol-1
)
a
[N3N]W(CH3) (2) in toluened
[N3N]Mo(CH2CMe3) in toluenee
[N3N]W(CHCH2CH2CH2) in toluenef
[N3N]Mo(cyclobutyl) in tolueneg
[N3N]W(C5H8)(H)h
19.9
25.9
20.7
24.3
22.5
-14
-7
-13
0
10.0
0.048
4.5
16.4
4.8
-8
a
Error ≈ 0.2 kcal. b Error ≈ 1 eu. c Calculated from exact values
for ∆Hq and ∆Sq given in the Experimental Section. R ) 0.995
d
for 13 runs between 300 and 341 K, 9 runs by 1H NMR, and 4
F igu r e 3. Variable-temperature 1H NMR spectra of the
ligand backbone in [N3N]WMe (∼0.055M; toluene-d8).
runs by visible spectroscopy. e See ref 11. f R ) 0.9998 for 4 runs
g
h
by NMR between 313 and 343 K. See ref 11. R ) 0.9997 for 4
runs by NMR between 318 and 353 K.
that of [N3N]MoCl.15 For example, W-Neq ) 1.985(11)
Å, Neq-W-Neq ) 117.66(6)°, and W-Nax ) 2.182(6) Å.
The trimethylsilyl groups are virtually “upright” with
Nax-W-Neq-Si angles near 180°, which is typical of
[N3N]MX complexes in which the X group in the trigonal
pocket has minimal steric demands.3,11 X-ray studies
of [N3N]MoX (X ) triflate, CH3, cyclohexyl)11 reveal that
the degree of “tipping” of the trimethylsilyl substituents,
as measured by the Nax-Mo-Neq-Si dihedral angle,
depends on the steric bulk of X, i.e., the Nax-Mo-Neq-
Si angles are smallest in [N3N]Mo(cyclohexyl), in which
the steric hindrance within the trigonal coordination
pocket is most severe.
spectroscopy by following the decrease in the intensity
of the TMS resonance.9 The values for ∆Hq and ∆Sq
obtained by combining these data (nine runs by NMR,
four runs by visible spectroscopy) are listed in Table 1.
(See Experimental Section for individual rate constants
at temperature T.) A resonance at 4.5 ppm for molecular
Attem p ted Syn th eses of Tu n gsten Lin ea r Alk yl
Com p lexes. Addition of methyllithium to 1 in ether
or THF yielded paramagnetic, orange [N3N]W(CH3) (2)
in high yield. A variable-temperature 1H NMR spec-
trum of 2 is shown in Figure 3. The observed temper-
ature-dependent processes are qualitatively the same
as those observed for [N3N]Mo(CH3)11 and described
above for [N3N]WCl. The activation energy (∆Gq) for
the WN2C2 “ring-flipping” process that produces an
effectively C3v-symmetric species on the NMR time scale
can be obtained from coalescence to give the upfield
resonance at -50 °C (9.0(2) kcal/mol) or the lower field
1
hydrogen could be observed in the H NMR spectrum
during conversion of 2 to 3a , but the low solubility of
dihydrogen in toluene-d8 did not allow the amount to
1
be quantified accurately. By H NMR at 47 °C kH/kD
for [N3N]W(CD3) was determined to be 5.6 while at 68
°C kH/kD ) 5.3.9 Isotope effects of this magnitude have
been documented for R-hydrogen abstraction reactions
in high oxidation state tantalum complexes.16,17 Loss
of dihydrogen was judged to be essentially irreversible
at 25 °C; no [N3N]WtCH was observed over a period of
14 days in a sample of [N3N]WtCD in C6D6 that had
been placed under 2 atm of dihydrogen at 25 °C. The
calculated value for the loss of dihydrogen from [N3N]W-
(CH3) at 120 °C (kW ) 8 × 10-2 s-1) should be compared
with the (estimated) value for the loss of dihydrogen
from [N3N]Mo(CH3) (kMo < 8 × 10-6 s-1) at 120 °C.11
(Although [N3N]MotCH is a known species that is
formed upon decomposition of [N3N]Mo(cyclopropyl),11
no [N3N]MotCH was observed to form from [N3N]Mo-
(CH3) at 120 °C over a period of 24 h; [N3N]Mo(CH3)
simply slowly disappeared with time, and no decomposi-
tion products could be identified.11) Therefore, we can
safely conclude that the rate of decomposition of [N3N]W-
(CH3) to [N3N]WtCH is at least 4 orders of magnitude
greater than the rate of decomposition of [N3N]Mo(CH3)
to [N3N]MotCH. [N3N]WtCH is also the product of
the attempted synthesis of [N3N]W(cyclopropyl) (see
below).
resonance at -30 °C (9.1(2) kcal/mol). A value of ∆Gq
W
) 9.0 kcal mol-1 is comparable to that estimated for
[N3N]Mo(CH3) (∆Gq ) 8.4 kcal mol-1).11 We believe
Mo
that the processes in which the ligand backbone begins
to flip from one C3 conformation to another have
approximately the same activation energy for W and Mo
because the process involves only conformational changes
in the MN2C2 rings, i.e., Nax remains bound during the
flipping process. One might expect that if dissociation
of Nax from the metal is required, the barrier for a W
complex would be significantly higher than for the
analogous Mo complex as a consequence of a higher
W-Nax bond strength. The effective magnetic moment
of [N3N]W(CH3) in the solid state was found to decrease
smoothly from ∼2 near room temperature to ∼0 at 5 K
in a manner similar to that shown in Figure 2, and the
magnetic moment at 25 °C in C6D6 was found to be 2.5
µB by the Evan’s method.
[N3N]W(CH3) is not very stable. Upon heating 2 at
temperatures between 25 °C and 80 °C it is converted
quantitatively into pale yellow [N3N]WtCH (3a ; eq 1).
The dehydrogenation reaction could be monitored by
UV-Vis spectroscopy in toluene by observing the disap-
[N3N]WCl reacts with LiCH2R′ reagents (R′ ) Me,
Prn, SiMe3, But) or KCH2Ph in ether or THF at room
temperature to yield the alkylidyne complexes, [N3N]-
WtCR′, in high yield (eq 2). We presume that the alkyl
1
(16) Wood, C. D.; McLain, S. J .; Schrock, R. R. J . Am. Chem. Soc.
1979, 101, 3210.
pearance of the absorption at 462 nm or by H NMR
(17) Schrock, R. R. In Reactions of Coordinated Ligands; Braterman,
P. R., Ed.; Plenum: New York, 1986.
(15) Duan, Z.; Verkade, J . G. Inorg. Chem. 1995, 34, 1576.

5198 Organometallics, Vol. 16, No. 24, 1997
Schrock et al.
shown in eq 4 is unlikely to be an intermediate in the
ring opening, or if it is, its lifetime is too short for the
configuration at the radical center to invert.
complexes, [N3N]W(CH2R′), are intermediates in these
reactions, although they could not be observed. Reac-
tion of [N3N]WCl with LiCD2CH2CH2CH3 resulted in
the production of [N3N]WtCCH2CH2CH3 only, while
the reaction between [N3N]WCl and Li13CH2CH2CH2-
CH3 gave only [N3N]Wt13CCH2CH2CH3. The value for
We propose that the cyclopropyl complexes decompose
via C-C bond cleavage in the cyclopropyl ring to yield
a metallacyclobutene complex (eq 5). One would expect
J C W (242 Hz) in [N3N]Wt13CCH2CH2CH3 could be
2
obtained easily from the 13C NMR spectrum and is in
the range normally observed in tungsten alkylidyne
complexes.18 These results suggest that dihydrogen is
rapidly lost from the R-carbon atom in the intermediate
butyl complex compared to the rate of any process that
would lead to the metal “walking” to the other end of
the C4 chain or producing n-butyl groups in which H
and D have been partially scrambled along the chain.
For [N3N]Mo(butyl), the overall rates of â elimination
to give [N3N]MoH and R,R-dehydrogenation to give
[N3N]MotCCH2CH2CH3 are comparable and relatively
slow,11 although there is still no significant degree of
migration of the Mo along the butyl chain before the
complex decomposes.
that a metallacyclobutene complex of this type would
be unstable with respect to loss of ethylene to give the
methylidyne complex, since reactions between olefins
and high oxidation state alkylidyne complexes18 are
extremely rare.19 This mechanism of formation of
[N3N]WtCH is the same as that proposed for the
formation of [N3N]MotCH.11
Addition of cyclopropyllithium to [N3N]WCl at ca. -30
°C yields [N3N]WtCH quantitatively in less than ∼1
min as the solution warms to room temperature (eq 3).
Addition of cyclobutyllithium to [N3N]WCl yields a
1
diamagnetic complex whose H and 13C NMR spectra
are consistent with it being the “1-tungstacyclopentenyl”
complex, [N3N]W(CHCH2CH2CH3), shown in eq 6. The
The reaction turns from the orange color characteristic
of [N3N]WCl to purple quickly upon addition of cyclo-
propyllithium, consistent with formation of [N3N]W-
(cyclopropyl). However, the purple color then fades
rapidly as pale yellow [N3N]WtCH is formed. Although
1
the reaction could be monitored by H NMR at -70 °C
alkylidene proton in 4 appears in the 1H NMR spectrum
and evidence obtained for the formation of the cyclo-
propyl species in the form of characteristically high-
field-shifted [N3N]3- ligand resonances (comparable to
the resonances obtained in the 1H NMR spectrum of
[N3N]Mo(cyclopropyl)11), accurate kinetic measurements
were not possible. If we estimate the half-life for the
decomposition of intermediate [N3N]W(cyclopropyl) to
be ∼5 s at 243 K, then k ) 7 × 10-3 s-1 at 243 K. The
rate constant for decomposition of [N3N]Mo(cyclopropyl)
can be calculated to be 4 × 10-9 at 243 K.11 Therefore,
at 243 K, [N3N]W(cyclopropyl) can be estimated to
decompose roughly 6 orders of magnitude more rapidly
than [N3N]Mo(cyclopropyl).
as a triplet at 11.3 ppm, while in the 13C NMR spectrum
1
1
CR appears at 265 ppm with J CH ) 130 Hz. The H
NMR spectrum of 4 at room temperature is consistent
with C3v symmetry, so the tungstacyclopentene ring
must effectively “rotate” about the z axis with respect
to the ligand framework on the NMR time scale at 25
°C. We also can say that 4 is not in rapid equilibrium
with [N3N]W(cyclobutyl), since in that case only two sets
of methylene protons in a ratio of 4:2 would be observed
in the diamagnetic region. (Three methylene proton
sets are observed in the 1H NMR spectrum of 4.) It
should be noted that an analogous reaction between
[N3N]MoCl and cyclobutyllithium yielded purple, para-
magnetic [N3N]Mo(cyclobutyl).11 Therefore, the d0 “alkyl-
[N3N]WCl reacts with trans-2,3-dimethylcyclopropyl-
lithium in ether at room temperature to give trans-2-
butene and [N3N]WtCH quantitatively. No cis-2-
butene could be observed by 1H NMR. The stereo-
specificity of this experiment suggests that the radical
(18) Murdzek, J . S.; Schrock, R. R. In Carbyne Complexes; VCH:
New York, 1988.
(19) Weiss, K. Angew. Chem., Int. Ed. Engl. 1986, 25, 359.

Alkyl Complexes of Tungsten(IV)
Organometallics, Vol. 16, No. 24, 1997 5199
one would expect. The structure of 4 should be com-
pared with that for [N3N]W(cyclopentylidene)(H) dis-
cussed below.
Compound 4 is converted quantitatively into [N3N]-
WtCCH2CH2CH3 in a first-order reaction according to
kinetic studies carried out by visible spectroscopy (Table
1; eq 7) at a rate that is approximately the same as the
rate of conversion of [N3N]Mo(cyclobutyl) into [N3N]-
MotCCH2CH2CH3 (Table 1).11 (The Mo analog of 4 was
proposed to be an intermediate in the reaction in which
[N3N]Mo(cyclobutyl) is converted into [N3N]MotCCH2-
CH2CH3.11) At this stage we can only presume that the
alkylidene carbon atom in 4 becomes the alkylidyne CR
carbon atom in 3c. In that case the reaction could be
viewed as a type of R-hydrogen abstraction reaction from
a d0 alkyl/alkylidene complex to give a d0 alkylidyne
complex, a type of reaction that has been documented
in tantalum alkylidene/alkylidyne chemistry,17 and that
is the origin of many high oxidation state alkylidyne
complexes.18 However, it is not clear how a concerted
transfer of a proton from the alkylidene carbon to the
alkyl carbon could take place given the 18-electron
nature of 4 and the fact that the alkylidene proton
points 180° away from the alkyl carbon atom to which
it migrates. Therefore, we must consider the possibility
that 3c does not arise from 4 directly, but from some as
yet unspecified intermediate, or that the rearrangement
that gives 3c involves the [N3N]3- ligand in some
manner. However, at this stage we cannot justify
proposing any alternatives to a concerted transforma-
tion of 4 to 3c.
F igu r e 4. View of the structure of [N3N]W(CHCH2CH2-
CH2) (4).
Ta ble 2. Selected Dista n ces (Å) a n d An gles (d eg)
for [N3N]W(CHCH2CH2CH2) (4)
Distances
W-N(1)
W-N(2)
W-N(3)
W-N(4)
W-C(7)
2.015(5)
2.033(5)
2.004(5)
2.395(5)
2.216(6)
W-C(10)
1.972(6)
1.527(8)
1.508(8)
1.509(8)
C(7)-C(8)
C(8)-C(9)
C(9)-C(10)
Angles
134.0(3)
W-N(1)-Si(1)
N(1)-W-C(7)
N(2)-W-C(7)
N(3)-W-C(7)
N(1)-W-N(2)
N(2)-W-N(3)
N(1)-W-N(3)
C(10)-W-C(7)
C(10)-W-N(1)
C(10)-W-N(3)
C(10)-W-N(2)
C(10)-W-N(4)
151.0(2)
77.9(2)
91.1(2)
100.9(2)
133.5(2)
109.0(2)
72.7(3)
83.8(2)
97.4(2)
120.9(2)
155.3(2)
W-N(2)-Si(2)
133.1(3)
136.3(3)
142a
W-N(3)-Si(3)
N(4)-W-N(1)-Si(1)
N(4)-W-N(2)-Si(2)
N(4)-W-N(3)-Si(3)
W-C(7)-C(8)
170a
149a
113.0(5)
104.8(5)
105.2(5)
127.1(5)
130.3(2)
C(7)-C(8)-C(9)
C(8)-C(9)-C(10)
W-C(10)-C(9)
C(7)-W-N(4)
a
Obtained from a Chem 3D model. The error is estimated to
be (1°.
The reaction of [N3N]WCl with cyclopentyllithium in
ether at room temperature produces yellow, crystalline
[N3N]W(C5H8)(H) (5) in high yield (eq 8). An X-ray
idene/alkyl” tautomer (4) is lower in energy than
[N3N]W(cyclobutyl), whereas for Mo the reverse was
proposed to be the case.11
An X-ray structure of 4 confirms its proposed nature
(Figure 4, Table 2). The complex has a trigonal-
monopyramidal [N3N]W core typical of triamidoamine
complexes of this general type.3 The W-N(4) distance
is relatively long, and two of the three trimethylsilyl
substituents are twisted away from the “upright” ori-
entation (N(4)-W-Neq-Si ) 180°), typical of complexes
in which the trigonal coordination pocket is occupied by
a relatively bulky group,3,11 here the 1-tungstacyclopen-
tene ring. (Similar effects have been observed in [N3N]-
Mo(CD3) versus [N3N]Mo(cyclohexyl).11) The N(2)-W-
N(3) angle (135.5(2)°) is opened up significantly compared
to the other two as a consequence of the 1-tungstacy-
clopentene ring being oriented so that it lies roughly in
the N(4)-W-N(1) plane with C(10) (the alkylidene
carbon atom) pointing toward N(1) and C(7) pointing
between N(2) and N(3). The W-C(10) and W-C(7)
distances are typical of W(VI) WdC and W-C bonds,
respectively. The W-C(10)-C(9) angle is larger than
than the W-C(7)-C(8) angle, and the N(4)-W-C(10)
angle is much larger than the N(4)-W-C(7) angle, as
structural determination of 5 revealed the structure
shown in Figure 5. (See Table 3 for selected distances
and angles.) The W-C(1) bond length (1.97(1) Å) is
statistically somewhat longer than that found in
W(cyclopentylidene)(OCH2CMe3)2Br2 (1.890(5) Å) and
its gallium tribromide adduct (1.896(9) Å),20 perhaps
because of the lower coordination number of, or greater
electrophilicity of, the metal in W(cyclopentylidene)-
(OCH2CMe3)2Br2. Nonetheless, it is in the range ex-
pected for a “d0” alkylidene complex17,21 and is identical
(20) Youinou, M. T.; Kress, J .; Fischer, J .; Aguero, A.; Osborn, J . A.
J . Am. Chem. Soc. 1988, 110, 1488.

5200 Organometallics, Vol. 16, No. 24, 1997
Schrock et al.
the hydride ligand was found approximately in the
C(1)-W-N(1) plane 1.81 Å from the metal, but it would
not survive refinement. The largest W-N-Si angle is
actually W-N(1)-Si(1) (131.2(5)°), presumably as a
consequence of the hydride pointing directly toward Si-
(1). The W-N(4) distance (2.346(9) Å) is close to the
W-N(4) distance in 4. The position and observed
orientation of the cyclopentylidene ligand is consistent
with the nature of the three available bonding orbitals,
i.e., formation of two σ bonding hybrid orbitals in either
the dxz or dyz plane and a WdC π bond using the π
orbital perpendicular to that plane. Although the Mo
complex analogous to 5, [N3N]Mo(C5H8)(H), could not
be observed directly, its formation was the most expedi-
tious way to explain the interconversion of the exo and
endo protons in the cyclopentyl ring in [N3N]Mo-
(cyclopentyl).11
The 1H NMR spectrum of 5 at 22 °C displayed a broad
singlet at δ 0.22 ppm for the trimethylsilyl groups of
the [N3N]3- ligand, two triplets for the methylene
groups on the [N3N]3- ligand backbone, and a broad
resonance for W-H at δ 19.6 ppm (Figure 6a). When
the sample was cooled to -3 °C, the hydride resonance
sharpened and W-H coupling (90 Hz) could be ob-
served. At -40 °C, two singlets for the trimethylsilyl
groups could be resolved in a ratio of 2:1, the methylene
groups in the ligand become multiplets at δ 3.27 and
2.02 ppm, and the cyclopentylidene ligand displays three
resonances at δ 5.18, 3.98, and 1.90 ppm in a ratio of
1:1:2. All data are consistent with a structure having
mirror symmetry on the NMR time scale at -40 °C. (At
-40 °C the WN2C2 “ring-flipping” process is still rela-
tively fast on the NMR time scale). The resonances at
5.18, 3.98, and 1.90 ppm in a ratio of 1:1:2 are assigned
to the methylene resonances 1-4 (eq 9) in the cyclo-
pentylidene ring, two sets of which overlap. The low-
F igu r e 5. Two views of the structure of [N3N]W(cyclo-
pentylidene)(H) (5).
Ta ble 3. Selected Dista n ces (Å) a n d An gles (d eg)
for [N3N]W(cyclop en tylid en e)(H) (5)
Distances
W-C(1)
W-N(1)
W-N(2)
1.97(1)
2.046(9)
2.20(1)
W-N(3)
2.045(9)
2.346(9)
W-N(4)
temperature 13C{1H} NMR spectra of 5 are also consis-
tent with mirror symmetry at low temperatures and
reveal the cyclopentylidene R-carbon resonance at 268.4
ppm.
Angles
H-W-C(1)
80(2)a
N(2)-W-N(3)
N(1)-W-N(3)
N(4)-W-N(1)-Si(1)
N(4)-W-N(2)-Si(2)
N(4)-W-N(3)-Si(3)
125.7(4)
W-N(1)-Si(1)
W-N(2)-Si(2)
W-N(3)-Si(3)
N(1)-W-N(2)
131.2(5)
122.3(5)
128.7(5)
112.4(4)
105.4(4)
179.9(7)
165.0(7)
148.1(7)
The temperature-dependent 1H NMR spectra of 5
reveal that the resolved cyclopentylidene methylene
resonances at 5.18 and 3.98 ppm broaden and coalesce
at high temperatures (Figure 6b). Therefore, we pro-
pose that these resonances can be assigned to protons
1 and 4 (eq 9) and the overlapping methylene resonances
at 1.90 ppm (see also Figure 8) to protons on carbons 2
and 3; the overlapping resonances assigned to protons
on carbons 2 and 3 also sharpen to a single resonance
at higher temperatures. We propose that we are
observing the consequences of an interconversion of
[N3N]W(cyclopentylidene)(H) and [N3N]W(cyclopentyl)
(eq 9). In [N3N]W(cyclopentyl), which we presume
would be a high-spin paramagnetic species analogous
to [N3N]Mo(cyclopentyl)11 and [N3N]WMe, the â protons
a
Hydride does not survive refinement.
to the W-C(10) distance in 4. The cyclopentylidene ring
is oriented so that it lies approximately in a plane that
passes through W, N(1), N(4), and Si(1). The N(2)-W-
N(3) angle (125.7(4)°) is consequently slightly larger
than the other two Neq-W-Neq angles (cf. the N(2)-
W-N(3) angle of 133.5(2)° in 4). The cyclopentylidene
ring is leaning slightly toward Si(3), and Si(3) is
consequently tipped furthest from “vertical” (N(4)-W-
N(3)-Si(3) ) 148.1(7)°). Electron density ascribable to
(21) Feldman, J .; Schrock, R. R. Prog. Inorg. Chem. 1991, 39, 1.

Alkyl Complexes of Tungsten(IV)
Organometallics, Vol. 16, No. 24, 1997 5201
protons (on carbons 1 and 4) would become equivalent
when the rate of interconversion of [N3N]W(cyclo-
pentylidene)(H) and [N3N]W(cyclopentyl) becomes fast
on the NMR time scale, as would all γ protons (on
carbons 2 and 3). A similar process was proposed to
account for interconversion of the endo and exo protons
in [N3N]Mo(cyclopentyl), and the slower rate constant
for that reaction (the equivalent of kR in eq 9) was found
to be ∼103 at 22 °C.11 In the tungsten system, k1 is
smaller than kR and can be determined at 46 °C from
the coalescence point of the resonances for methylene
protons 1 and 4 (Figure 6b), assuming a chemical shift
difference that is relatively independent of temperature.
The value so obtained (at 46 °C) is k1 ) 1.3 × 103 s-1
.
The other important feature of the temperature-
dependent 1H NMR spectrum of 5 is that the hydride
resonance shifts downfield and broadens as the tem-
perature is increased (Figure 6a). Since paramagnetic
[N3N]W(cyclopentyl) is being accessed rapidly on the
NMR time scale and since HR could not be observed in
[N3N]Mo(cyclopentyl) or any similar alkyl complex,11 the
chemical shift and width of the HR resonance in [N3N]W-
(C4H8)(H) in some temperature range should be affected
accordingly, depending upon the magnitude of k1/kR.
(The hydride resonance should be affected the most,
since it becomes attached to the R carbon atom in the
cyclopentyl complex.) If we assume that the temperature
dependence of the chemical shifts of the resonances in
unobservable [N3N]W(cyclopentyl) follow a 1/T depen-
dence (as found for [N3N]Mo(cyclopentyl)11) and that
these resonances do not shift to any significant degree
for any other reason, then we can determine the ∆G°
for the equilibrium in eq 9 (i.e., the magnitude of the
equilibrium constant k1/kR) using the expression shown
in eq 10,22-24 where δ is the observed chemical shift of
the hydride resonance in the spectrum of [N3N]W-
(cyclopentylidene)(H) at temperature T, δdia is the
chemical shift of the hydride resonance in the spectrum
of [N3N]W(cyclopentylidene)(H) in the absence of any
contact shift caused by equilibration of [N3N]W(cyclo-
pentylidene)(H) with [N3N]W(cyclopentyl), and C is a
constant. From a plot of δ for the hydride ligand in
F igu r e 6. Variable-temperature 1H NMR spectra of
[N3N]W(cyclopentylidene)(H). (a) hydride resonance; (b) â
protons of the cyclopentylidene ring. (H1 and H4 assign-
ments are arbitrary.)
[N3N]W(cyclopentylidene)(H) versus T (Figure 7), we
extract values of ∆H° ) 11.8(6) kcal mol-1, ∆S° ) 33(2)
eu, and δdia ) 19.46 ppm. Therefore, at a temperature
of 46 °C, the equilibrium constant k1/kR is calculated to
be 0.13 and kRW ) 1.0 × 104 s-1. This value should be
compared with that determined in the analogous Mo
system,11 kRMo ) 103 s-1 at 22 °C. At 46 °C, kRMo could
be as much as 1 order of magnitude larger than that at
22 °C. Thus, kRW and kRMo are approximately the same
at ∼46 °C. Therefore, the fact that [N3N]W(cyclo-
pentylidene)(H) is the favored species for W while [N3N]-
Mo(cyclopentyl) is the favored species for Mo must be
ascribed to a significantly smaller k1W than k1Mo
.
F igu r e 7. Plot of the chemical shift of the hydride in
[N3N]W(cyclopentylidene)(H) versus temperature.
(22) Gu¨tlich, P.; McGarvey, B. R.; Kla¨ui, W. Inorg. Chem. 1980, 19,
3704.
(23) Smith, M. E.; Andersen, R. A. J . Am. Chem. Soc. 1996, 118,
1119.
(24) Kla¨ui, W.; Eberspach, W.; Gu¨tlich, P. Inorg. Chem. 1987, 26,
3977.
on the same side of the ring as the metal (endo) would
be equivalent and the â protons on the opposite side of
the ring (exo) would be equivalent. Therefore, all â

5202 Organometallics, Vol. 16, No. 24, 1997
Schrock et al.
2
F igu r e 8. Stacked H NMR spectra in toluene showing the incorporation of deuterium from [N3N]W(C5H8)(D) into the â
and γ ring positions over time (Peaks marked with * indicate the natural abundance of 2H in toluene; those marked with
& cyclopentane-d1 from trace hydrolysis; those marked with # cyclopentene-d1 from decomposition of [N3N]W(C5H8)(D) to
give [N3N]WH).
Unfortunately, this conclusion cannot be confirmed, as
there is no way to measure the magnitude of the
equilibrium constant between observable [N3N]Mo-
(cyclopentyl) and unobserved [N3N]Mo(cyclopentylidene)-
(H) and, therefore, k1Mo. If we assume that k1/kR in the
Mo system is >102, then we can estimate that k1Mo
>
105 s-1 at 22 °C. This value should be compared with
k1W ) 1.3 × 103 s-1 at 46 °C. The fact that the kR values
for W and Mo are roughly the same while the k1 values
for W and Mo differ by approximately 3 orders of
magnitude is somewhat surprising.
Addition of 1-d-cyclopentyllithium to [N3N]WCl at
-60 °C followed by warming the sample to -13 °C
resulted in formation of only [N3N]W(C5H8)(D), as
1
2
determined by H and H NMR spectroscopy (eq 11).10
F igu r e 9. Plot of ln(A/A0) versus time (A ) amount of
deuterium present in the hydride site) during the incor-
poration of deuterium from [N3N]W(C5H8)(D) into the â and
γ ring positions.
cyclopentene is lost from C at room temperature over a
period of 24 h.) It takes more than 2 h for D to be evenly
distributed between the â and γ positions since the one
deuterium becomes distributed throughout four â posi-
tions before it scrambles from the four â positions to
the four γ positions. We can obtain a measure of the
rate of scrambling of D into â positions via C by plotting
the decrease in intensity of the 2H resonance at 19.3
2
The H resonance in [N3N]W(C5H8)(D) disappears over
2
a period of 24 h, and H intensity begins to appear in
the cyclopentylidene â-methylene positions at ∼3.8 and
5 ppm first (Figure 8). Only after a much longer period
of time does D appear in the γ positions (2 and 3 in eq
9) at ∼2.0 ppm (Figure 8). We propose that D scrambles
into the â positions first as a consequence of reversible
â-hydride elimination from the the cyclopentyl complex
B (eq 11) to give C. (We show below that little
2
ppm in the H NMR spectrum of A at 22 °C (Figure 9).
If we assume that the initial decrease is first order (the
accuracy of the experiment is not great enough to
confirm that that is the case), then the first-order rate
constant at 22 °C is kobs ) 3.7 × 10-4 s-1 (t1/2 ) 31 min).

Alkyl Complexes of Tungsten(IV)
Organometallics, Vol. 16, No. 24, 1997 5203
(A more accurate experiment would have to include the
sequential nature of the scrambling reaction.) This value
is much smaller than both k1 and kR (eq 11). Unfortu-
nately, we cannot say which is larger, kâ or k2, and
therefore cannot say which corresponds to kobs. The
observation that only [N3N]W(C5H8)(D) is formed upon
adding 1-d-cyclopentyllithium to [N3N]WCl at -60 °C
followed by warming to -13 °C would require only that
kâ < kR, and that seems assured in view of the large
values for k1 and kR relative to kobs. It should be noted
that in the Mo system kâMo was estimated to be ∼7 ×
10-4 s-1 at 22 °C. It also should be noted that after 24
h only ∼3% of the original D (by NMR integration) is
bound to the metal (Figure 8) instead of what one would
predict (1/9 or ∼11% of the original) if no equilibrium
isotope effect were operative and no other H/D exchange
were taking place that is slower than exchange into the
cyclopentylidene ring. A substantial normal equilibri-
um isotope effect should be observed, as a proton should
favor being bound to the metal versus carbon.25-30
Although the residual WD (∼3%) cannot be measured
accurately by integration, the amount is, in fact, ap-
proximately what one would expect using plausible
stretching frequencies for WH, WD, CH, and CD as a
means of estimating the energy difference between the
WD compound and the WH compound (∼750 cal).
Therefore, there is no need to invoke any other exchange
reaction that would further reduce the amount of WD
or potentially alter the linearity of the plot shown in
Figure 9.
which are similar to those of [N3N]W(cyclopentylidene)H.
However, [N3N]W(cyclohexylidene)(H) does not decom-
pose cleanly to [N3N]WH, although some [N3N]WH is
formed. The competitive side reaction or reactions have
not been elucidated. Addition of cyclopentenyllithium
to [N3N]WCl yielded [N3N]W(cyclopentenyl) (2b; eq 13),
while addition of phenyllithium yielded [N3N]W(phenyl).
[N3N]W(cyclopentylidene)(H) (5) decomposes readily
at temperatures above ∼45 °C to give cyclopentene and
paramagnetic [N3N]WH (6). Compound 6 has NMR
spectra and magnetic behavior (Figures 1 and 2) typical
of d2 complexes of the type [N3N]WX. Warming the
sample of 5-d1 discussed in the preceding paragraph
above 45 °C yielded cyclopentene-d1 in which deuterium
was scrambled throughout all positions in the ring
2
according to 1H and H NMR spectroscopy, as expected.
The behavior of the susceptibility and effective moment
of [N3N]W(phenyl) versus temperature are analogous
to the behavior of [N3N]WMe, [N3N]WCl, and [N3N]WH.
Compound 2b is also a paramagnetic species, judging
from its NMR spectra, although detailed susceptibility
studies on it were not carried out. The paramagnetism
of both [N3N]W(phenyl) and 2b suggests that conjuga-
tion of the CdC π system with one of the π orbitals
available for bonding in the trigonal pocket (dxz or dyz)
is not sufficient to break the degeneracy of the dxz and
dyz orbitals in either compound. [N3N]WMe, [N3N]WPh,
and 2b are the only five-coordinate species so far that
contain a W-C single bond.
Like [N3N]WMe, 2b was found to be unstable at room
temperature. It decomposes cleanly to produce a dia-
magnetic species that has no symmetry. An X-ray
structural study revealed that the product is that shown
in eq 14; a trimethylsilyl group has migrated from an
amido nitrogen to the cyclopentenyl ligand to form a
1-(trimethylsilyl)cyclopentene complex (7). Two views
of 7 are shown in Figure 10, and some relevant bond
distances and angles are listed in Table 4. The most
notable aspect of this structure is the absence of a TMS
group on one nitrogen, leading to a bent imido ligand
(W-N(3)-C(5) ) 135.7(6)°). (Bent imido ligands are
somewhat unusual,31,32 although they appear to be more
The decomposition of 5 is proposed to proceed via
formation and decomposition of the cyclopentyl complex
by â-hydride elimination and loss of cyclopentene, eq
12. The decomposition of [N3N]W(cyclopentylidene)(H)
was found by 1H NMR to follow a first-order dependence
through three or more half-lives at 45 (k ) 4.8 × 10-5
s-1), 60 (k ) 2.3 × 10-4s-1), and 80 °C (k ) 1.8 × 10-3
and 2.0 × 10-3 s-1). An Eyring plot yielded ∆Hq
)
22 463 cal/mol and ∆Sq ) -7.84 eu. The value for the
loss of cyclopentene at 22 °C, therefore, can be calculated
to be 2.3 × 10-6 s-1. This is approximately 2 orders of
magnitude slower than the rate of deuterium scram-
bling (kobs ) 3.7 × 10-4 s-1). Many attempts to
synthesize [N3N]WH directly from [N3N]WCl using
various hydride sources have failed, so that decomposi-
tion of 5 currently is the only practical method of
forming 6.
Addition of cyclohexyllithium to [N3N]WCl yielded
[N3N]W(cyclohexylidene)H, the NMR characteristics of
(25) Isotope Effects in Chemical Reactions; Collins, C. J ., Bowman,
N. S., Eds.; Van Nostrand Reinhold Co.: 1970.
(26) Calvert, R. B.; Shapley, J . R. J . Am. Chem. Soc. 1978, 100, 7726.
(27) Faller, J . W.; Murray, H. H.; M., S. J . Am. Chem. Soc. 1980,
102, 2306.
(28) Luo, X.-L.; Crabtree, R. H. J . Am. Chem. Soc. 1990, 112, 6912.
(29) Rabinovich, D.; Parkin, G. J . Am. Chem. Soc. 1994, 115, 353.
(30) Tanner, M. J .; Brookhart, M.; DeSimone, J . M. J . Am. Chem.
Soc. 1997, 119, 7617.
(31) Wigley, D. E. Prog. Inorg. Chem. 1994, 42, 2239.

5204 Organometallics, Vol. 16, No. 24, 1997
Schrock et al.
position, presumably again as a consequence of steric
pressure in the apical pocket. The W-C(41) and
W-C(45) distances are relatively short for W-C single
bonds, while C(41)-C(45) is not much shorter than
other C-C bonds in the cyclopentene ring, consistent
with a “tungstacyclopropane” description for the W(ole-
fin) bond. We presume at this stage that the TMS group
migrates directly to the unsubstituted carbon atom in
2b to give 7.
Discu ssion
In another paper,11 we discuss how little is known
concerning R elimination to give an alkylidene hydride35
relative to â-hydride elimination to give an olefin
hydride in transition metal alkyl chemistry. An equi-
librium between an alkyl complex and alkylidene hy-
dride has usually involved high oxidation state early
transition metal alkyl complexes, i.e., those of tung-
sten36-38 or tantalum39-43 complexes. It was noted43 in
a paper concerning the decomposition of Cp*2(H)-
TadCdCH2 that although “the generality of faster R-H
elimination versus â-H elimination is questionable,
since the transition state for â-H elimination in this
system is highly strained, whereas that for R-H elimi-
nation is much less so.” It has also been noted recently
that competition between R abstraction and â abstrac-
tion in dialkyltantalum species containing the [(Me3-
SiNCH2CH2)3N]3- ([N3N]3-) ligand can be controlled by
varying the size of the apical pocket, either by increasing
the size of the alkyl44 or by increasing the size of the
silyl substituent from SiMe3 to SiEt3.45 Most recently,
it has been shown that R elimination in two [N3N]Mo-
(cycloalkyl) complexes is 6-7 orders of magnitude faster
than â elimination.11 We now have the opportunity to
compare several [N3N]W(alkyl) complexes with the
analogous [N3N]Mo(alkyl) complexes.
F igu r e 10. View of the structure of [(Me3SiNCH2CH2)2-
NCH2CH2N)W(1-(trimethylsilyl)cyclopentene) (7).
Ta ble 4. Selected Dista n ces (Å) a n d An gles (d eg)
for [(Me3SiNCH2CH2)2NCH2CH2N]-
W(1-(tr im eth ylsilyl)cyclop en ten e) (7)
Distances
W-C(41)
2.198(8)
2.200(8)
2.005(7)
1.513(13)
1.544(12)
1.530(14)
W-N(2)
2.040(7)
1.765(7)
2.428(7)
1.548(13)
1.556(13)
W-C(45)
W-N(3)
W-N(1)
W-N(4)
C(43)-C(44)
C(44)-C(45)
C(41)-C(45)
C(41)-C(42)
C(42)-C(43)
Angles
W-N(3)-C(5)
W-N(2)-C(3)
W-N(1)-C(1)
W-N(1)-Si(1)
W-N(2)-Si(2)
135.7(6)
110.3(6)
112.2(5)
136.1(4)
127.2(4)
N(1)-W-N(2)
105.5(3)
105.8(3)
128.3(3)
126.1(4)
132.6(4)
N(2)-W-N(3)
N(1)-W-N(3)
N(4)-W-N(2)-Si(2)
N(4)-W-N(1)-Si(1)
The equilibrium shown in eq 15 lies to the right for
M ) Mo, with a magnitude of ∼102 or greater, and to
the left for M ) W with an estimated Keq ≈ 0.04 (Keq
)
[alkyl]/[alkylidene]) at 298 K. Differences in the mag-
nitude of Keq of several orders of magnitude can be easily
explained in terms of differences in MdC and M-H
bond energies relative to the M-C bond energy, with
greater differences being expected when M ) W. For
example, a change in ∆G°alkylidene - ∆G°alkyl of only ca.
-5 kcal mol-1 translates into an increase in Keq of ∼104.
Since the alkylidene ligand can be viewed as a dianion,
common when there is little alternative, i.e., when
incorporated into some metal-containing ring,33,34 as
found here. Under these circumstances, a MdN-C
value as low as 116.3(4)° has been observed.34) The
W-N(3)-C(5) angle is considerably larger than W-
N(2)-C(3) or W-N(1)-C(1), which are more typical for
[N3N]M systems. The N(3)‚‚‚Si(3) distance is 3.14 Å,
consistent with a fully broken N-Si bond. All carbon
atoms â to the equatorial nitrogens serve as the “flap”
in the MC2N2 rings, thereby leading to a pronounced
tilt to the TMS groups, presumably in order to relieve
steric interactions between the trimethylsilyl substitu-
ents and the cyclopentene ring. The W-N(4) bond
length is also longer than in typical molecules having
only a single cylindrically symmetric ligand in the axial
(35) Collman, J . P.; Hegedus, L. S. Principles and Applications of
Organotransition Metal Chemistry; University Science Books: Mill
Valley, CA, 1980.
(36) Cooper, N. J .; Green, M. L. H. J . Chem. Soc., Chem. Commun.
1974, 209.
(37) Cooper, N. J .; Green, M. L. H. J . Chem. Soc., Chem. Commun.
1974, 761.
(38) Cooper, N. J .; Green, M. L. H. J . Chem. Soc., Dalton Trans.
1979, 1121.
(39) Turner, H. W.; Schrock, R. R. J . Am. Chem. Soc. 1982, 104,
2331.
(40) Turner, H. W.; Schrock, R. R.; Fellmann, J . D.; Holmes, S. J .
J . Am. Chem. Soc. 1983, 105, 4942.
(41) Fellmann, J . D.; Schrock, R. R.; Traficante, D. D. Organome-
tallics 1982, 1, 481.
(42) van Asselt, A.; Burger, B. J .; Gibson, V. C.; Bercaw, J . E. J .
Am. Chem. Soc. 1986, 108, 5347.
(43) Parkin, G.; Bunel, E.; Burger, B. J .; Trimmer, M. S.; van Asselt,
A.; Bercaw, J . E. J . Mol. Catal. 1987, 41, 21.
(44) Freundlich, J . S.; Schrock, R. R.; Davis, W. M. J . Am. Chem.
Soc. 1996, 118, 3643.
(45) Fruendlich, J . S.; Schrock, R. R.; Davis, W. M. Organometallics
1996, 15, 2777.
(32) Haymore, B. L.; Maatta, E. A.; Wentworth, R. A. D. J . Am.
Chem. Soc. 1979, 101, 2063.
(33) Herrmann, W. A.; Marz, D. W.; Herdtweck, E. Z. Naturforsch.,
B 1991, 46, 747.
(34) Gountchev, T. I.; Tilley, T. D. J . Am. Chem. Soc., in press.

Alkyl Complexes of Tungsten(IV)
Organometallics, Vol. 16, No. 24, 1997 5205
species compared to its Mo(VI) analog. The 4-6 orders
of magnitude difference in the rate of R,R dehydroge-
nation between analogous Mo and W complexes that we
have been able to estimate in several circumstances,
therefore, can be rationalized readily.
The chemistry of tungsten cyclopropyl and cyclobutyl
complexes is similar to that of analogous Mo complexes,
although the reactions (eqs 3, 4, and 5) are again simply
much faster for tungsten compared to molybdenum. If
we make the reasonable assumption that the cyclobutyl
reactions proceed analogously for W and Mo, then it is
interesting to note that [N3N]W(CHCH2CH2CH2) (4) is
formed rapidly from unobservable [N3N]W(cyclobutyl)
but [N3N]Mo(cyclobutyl) is observable and intermediate
[N3N]Mo(CHCH2CH2CH2) is not observed when [N3-
N]Mo(cyclobutyl) is converted into [N3N]MotCCH2CH2-
CH3. These results could be circumstantial, as we know
few details about how [N3N]M(CHCH2CH2CH2) com-
plexes are converted into [N3N]MtCCH2CH2CH3 com-
plexes. As shown in Table 1, the value of ∆Hq for
[N3N]Mo(cyclobutyl) is ∼4 kcal mol-1 higher than the
value of ∆Hq for the decomposition of 4. It seems likely
that intermediate [N3N]Mo(CHCH2CH2CH2) is higher
in energy than [N3N]Mo(cyclobutyl), while [N3N]W-
(CHCH2CH2CH2) is clearly lower in energy than [N3N]-
W(cyclobutyl), consistent with the observations dis-
cussed above with respect to the cyclopentyl complexes.
this finding also can be rationalized in terms of oxida-
tion-state stabilities (W(VI) > Mo(VI)). The similar kR
values for Mo and W may be circumstantial. Another
possibility, however, is that the transition state for R
elimination is relatively early and kR, therefore, is
relatively independent of the nature of the alkylidene
hydride product.
We have documented that formation of [N3N]MotCR
complexes by what is essentially irreversible R,R dehy-
drogenation (eq 16) and [N3N]MoH by loss of olefin are
slow and competitive, while R,R dehydrogenation in
[N3N]W(CH2R) complexes is many orders of magnitude
more rapid than in an analogous [N3N]Mo(CH2R)
complex. On the basis of the data obtained for Mo and
In the paper concerning molybdenum chemistry11 and
in an earlier communication,10 we raised the possibility
that the overall rate of a unimolecular reaction could
be slowed significantly as a consequence of a small
equilibrium constant for the equilibrium between the
observed high-spin ground state and a “required” low-
spin form with 1A symmetry. However, so far all
magnetic behavior is of a relatively classical type, and
1
we have no evidence that a more readily accessible A
state is what leads to some dramatically faster reactions
in the W systems compared to the analogous Mo system.
We also still have not been able to eliminate the
possibility that the [N3N] ligand is directly involved in
proton-migration reactions of the type described here
nor the possibility that the axial nitrogen atom does not
remain strongly bound in all reactions. We hope to
address some of these issues in the future through the
synthesis of triamido/donor ligands with amido substit-
uents different from those available so far and with
donor atoms other than nitrogen.
W cyclopentyl complexes we propose that the position
of the equilibrium between [N3N]M(CH2R) and [N3N]M-
(CHR)(H) is several orders of magnitude larger (toward
the latter) when M ) W than when M ) Mo. The
overall more facile R,R elimination when M ) W (in
analogous W and Mo complexes) could be ascribed
largely to the higher concentration of [N3N]W(CHR)-
(H) versus [N3N]W(CH2R), if we could say with some
confidence that the rate of loss of molecular hydrogen
from a given [N3N]W(CHR)(H) complex were at least
as great as the rate of loss of molecular hydrogen from
a given [N3N]Mo(CHR)(H) complex. Although we have
no independent data that would suggest that an R-ab-
straction reaction is at least as fast for a third-row metal
complex as it is for a second-row metal complex, it would
not be surprising to find that to be the case. If the rate
of loss of molecular hydrogen from a given [N3N]W-
(CHR)(H) complex were significantly greater than the
rate of loss of molecular hydrogen from a given [N3N]-
Mo(CHR)(H) complex, then the overall rate of loss of
molecular hydrogen from a given [N3N]W(CH2R) com-
plex relative to a [N3N]Mo(CH2R) complex would only
become that much faster. These arguments are all
consistent with the preference for tungsten (versus Mo)
to form a M(VI) species, which may ascribed in part to
the likely stronger metal-carbon bonds in a W(VI)
Exp er im en ta l Section
Gen er a l Deta ils. All experiments were conducted under
nitrogen in a Vacuum Atmospheres drybox, using standard
Schlenk techniques, or on a high-vacuum line (<10-4 Torr).
Pentane was washed with HNO3/H2SO4 (5/95 v/v), sodium
bicarbonate, and H2O, stored over CaCl2, and then distilled
from sodium benzophenone under nitrogen. Reagent grade
ether, tetrahydrofuran, and benzene were distilled from
sodium benzophenone under nitrogen. Toluene was distilled
from molten sodium. Methylene chloride was distilled from
CaH2. All solvents were stored in the drybox over activated 4
Å molecular sieves. Deuterated solvents were freeze-pump-
thaw degassed and vacuum transferred from an appropriate
drying agent. NMR spectra are recorded in C6D6 unless noted
otherwise. 1H and 13C data are listed in parts per million
(ppm) downfield from tetramethylsilane and were referenced
using the residual protonated solvent peak. 2H NMR spectra

5206 Organometallics, Vol. 16, No. 24, 1997
Schrock et al.
MHz in toluene-d8 through at least two half-lives at
were obtained at 46.0 MHz and referenced to external C6D6
(7.15 ppm). Coupling constants are given in Hertz, and routine
coupling constants are not listed. Elemental analyses (C, H,
N) were performed on a Perkin-Elmer 2400 CHN analyzer in
our own laboratory.
a
concentration of ∼0.020 M.9 The values for k (×10-4 s-1) at
temperature T (K) are as follows: 30.6 (351), 12.2, 11.8 (341),
4.35 (330), 1.56, 1.49, 1.50 (320), 0.644 (310), 0.251 (300).
The conversion was followed in toluene by UV-vis; the
values for k (×10-4 s-1) at temperature T (K) are as follows:
0.105 (298), 1.28 (320), 9.66 (341), 10.50 (341).
LiCH2CD3 was prepared by treatment of CD3CH2Br (Cam-
bridge Isotope Labs) with lithium metal. Li13CH2CH2CH2CH3
was prepared by (i) treating 13CO2 with PrnMgCl, (ii) reducing
the acid with LiAlH4 to the alcohol, (iii) chlorinating the alcohol
with SOCl2 in hexane, and (iv) treating the chloride with
lithium in the usual manner; overall yield 20% (one synthesis).
All organic compounds were received from commercial
suppliers and used as received. Triethylamine was distilled
from CaH2 and stored over 4 Å molecular sieves. WCl4(DME)46
was prepared according to a published procedure. Deuterated
solvents were dried by passage through alumina or stirring
over sodium benzophenone ketyl and storage over 4 Å molec-
ular sieves. 1-d-Cyclopentyl lithium was prepared by reduc-
tion of cyclopentanone with lithium aluminum deuteride
followed by treatment with Ph3Br2. Li3[N3N] was prepared
as reported in the literature.12
A plot of ln(k/T) versus 1/T for all data gave ∆Hq ) 19 916
cal/mol and ∆Sq ) -13.87 eu with R ) 0.995.
[N3N]W(CD3). This compound was prepared in a manner
analogous to that used to prepare 2a ; yield 0.22 g (75%): 1H
NMR (toluene-d8) δ 10.05 (SiMe3), -32.41 (NCH2, w1/2 ) 58),
-73.17 (NCH2, w1/2 ) 35). Anal. Calcd for C16H39N4D3Si3W:
C, 34.22; H, 7.00; N, 9.98. Found: C, 33.57; H, 6.69; N, 9.91.
In the preliminary communication,10 it was reported that a
2
resonance for CD3 could be observed in the H NMR spectrum
of [N3N]W(CD3) in C6H6 at 7.5 ppm. This is incorrect. The
resonance at 7.5 ppm is due to deuterium in C6H6 in natural
abundance. We have not been able to locate the resonance in
[N3N]W(CD3).
[N3N]W(p h en yl) (2c). Phenyllithium (12 mg, 0.14 mmol)
was added to a cold solution of 45 mg (0.08 mmol) of [N3N]-
WCl in 5 mL of diethyl ether. The color changed immediately
to red. The solution was stirred for 1 h at room temperature,
and the solvent was removed in vacuo. The residue was
extracted with pentane, and the volume of the solution was
reduced to 2 mL in vacuo. The solution was kept at -30 °C
for 2 h, and the red crystalline product was collected; yield 35
mg: 1H NMR δ 58.1 (s, Ph), 14.6 (s, TMS), -30.0 (s, CH2),
-59.8 (s, CH2). Anal. Calcd for C21H44N4WSi3: C, 40.64; H,
7.14; N, 9.03. Found: C, 40.28; H, 7.30; N, 8.89.
[N3N]WtCH (3a ). Meth od a : [N3N]W(CH3) (0.20 g, 0.36
mmol) was dissolved in 30 mL of freshly distilled ether, and
the solution was stirred at 25 °C for 3 days. The resulting
light yellow solution was evaporated to dryness in vacuo, and
the residue was dissolved in pentane (∼60 mL, freshly passed
through alumina). The pentane solution was filtered, and the
pentane was removed in vacuo to yield brownish, pale yellow
crystals; yield 0.18 g (90%).
In the UV-vis studies, the following equation was employed
to follow the disappearance of an alkyl species: ln[(λo - λ∞)/(λ
- λ∞)] ) kt, where λo ) absorbance at wavelength λ at time 0,
λ∞ ) absorbance at wavelength λ at infinite time, and λ )
absorbance at wavelength λ at time t.
[N3N]WCl (1). WCl4(DME) (3.84 g, 9.23 mmol) dissolved
in minimum DME was added to a THF (40 mL) solution of
Li3[N3N] (3.54 g, 9.23 mmol) at -40 °C. The reaction mixture
was allowed to warm slowly to 25 °C and was stirred for 3 h.
The resulting yellow-brown solution was evaporated to dryness
in vacuo, and pentane (∼200 mL) and ether (∼20 mL) were
added to the reaction residue. The mixture was stirred at
room temperature for 3 h and filtered through a pad of Celite.
The filtrate volume was reduced until a quantity of yellow solid
formed. The yellow solid was collected by filtration, washed
with cold pentane (2 × 3 mL), and dried in vacuo. The filtrate
and pentane washes were combined, evaporated to dryness,
and extracted with pentane (∼20 mL). The extract was chilled
to -40 °C to afford another crop of yellow product; total yield,
1.38 g (26%): 1H NMR δ 8.30 (br s, SiMe3), -29.83 (br s, NCH2,
w1/2 ≈ 55), -75.44 (br s, NCH2, w1/2 ≈ 87). Anal. Calcd for
Meth od b: A toluene (∼20 mL) solution of [N3N]W(CH3)
(0.40 g, 0.72 mmol) was freezed-pump-thawed four times on
a high-vacuum line in a thick Pyrex glass reaction vessel. The
reaction vessel was heated to 75 °C for 5 h, and the solvent
was then removed in vacuo. The reaction residue was dis-
solved in pentane, and the extract was filtered. The pentane
filtrate was concentrated to ∼3 mL and chilled to -40 °C for
1 h. The resulting white precipitate was filtered off and dried
in vacuo; yield 0.35 g (90%): 1H NMR δ 0.48 (SiMe3), 2.06
(NCH2), 3.46 (NCH2), 7.08 (WtCH, 2J HW ) 81); 13C{1H} NMR
(toluene-d8) δ 4.47 (SiMe3), 51.66 (NCH2), 52.50 (NCH2), 272.6
C
15H39N4ClSi3W: C, 31.11; H, 6.79; N, 9.68. Found: C, 30.88;
H, 6.64; N, 9.56. Two measurements of µeff by the Evan’s
method gave values of 2.4 and 2.5.
[N3N]WI. Into a 20 mL flask was added 74 mg (0.13 mmol)
of [N3N]WCl and 5 mL of benzene. To this solution was added
28 mg of TMSI (0.14 mmol). The mixture was stirred at room
temperature for 2 days. The volatile materials were removed
in vacuo and the residue extracted with minimum pentane (5
mL). This solution was filtered, reduced in volume, and cooled
to -40 °C to give yellow crystals; yield 32 mg (38%): 1H NMR
δ 17.7 (s, NSiMe3), -29.5 (s, NCH2CH2N), -92.5 (s, NCH2-
CH2N). Anal. Calcd for C15H39N4Si3WI: C, 26.87; H, 5.86; N,
8.36. Found: C, 26.50; H, 5.68; N, 8.47.
[N3N]W(CH3) (2a ). A solution of [N3N]WCl (0.30 g, 0.52
mmol) in ether (∼40 mL) was treated with LiMe (0.41 g, 0.78
mmol, 1.4 M in ether) at 25 °C. The reaction mixture was
stirred for 30 min, and the solvent was removed in vacuo. The
orange reaction residue was extracted with pentane (∼80 mL),
and the extract was filtered through a pad of Celite. The
volume of the filtrate was reduced to ∼5 mL in vacuo, and
the solution was chilled to -40 °C to yield orange crystals of
the product; yield 0.25 g (86%): 1H NMR δ 9.89 (SiMe3),
-31.43 (NCH2, w1/2 ) 40), -72.70 (NCH2, w1/2 ) 23); µ ) 2.5
BM at 25 °C (Evan’s method). Anal. Calcd for C16H42N4-
Si3W: C, 34.40; H, 7.58; N, 10.03. Found: C, 34.73; H, 7.19;
N, 9.84.
1
1
(WtC, J CW ) 244, J CH ) 138). Anal. Calcd for C16H40N4-
Si3W: C, 34.52; H, 7.24; N, 10.07. Found: C, 34.82 H, 6.64;
N, 10.00.
[N3N]WtCD. The reaction procedure is similar to that
used to prepare [N3N]WtCH, except [N3N]W(CD3) and a
longer reaction time (∼10 days with method a; ∼24 h with
method b) were used; 72% yield: 1H NMR δ 0.49 (SiMe3), 2.03
(NCH2), 3.48 (NCH2); 13C{1H} NMR δ 4.80 (SiMe3), 51.96
1
(NCH2), 52.25 (NCH2), 231.15 (WtC, J CD) 33); 2H NMR
(C6H6) δ 7.4 (WtCD).
The values for k (×10-4 s-1) at temperature T (K), deter-
mined as described for [N3N]WtCH, are as follows: 2.30 (341),
0.273 (320). (Taken from ref 10).
[N3N]WtCCH3 (3b). An ether (∼25 mL) solution of [N3N]-
WCl (0.20 g, 0.345 mmol) was treated with LiEt (0.019 g, 0.52
mmol) at 25 °C. The reaction mixture turned light yellow as
gas evolved. The resulting mixture was stirred for another 3
h and then evaporated to dryness in vacuo. The yellow
reaction residue was extracted with pentane (∼60 mL), and
the extract was filtered. The filtrate was concentrated to ∼5
mL in vacuo and chilled at -40 °C for several hours to yield
light yellow crystals; yield 0.16 g (81%): 1H NMR δ 0.49
The conversion of [N3N]W(CH3) was measured by following
the decrease in the intensity of the SiMe3 resonance at 500
(46) Persson, C.; Andersson, C. Inorg. Chim. Acta 1993, 203, 235.

Alkyl Complexes of Tungsten(IV)
Organometallics, Vol. 16, No. 24, 1997 5207
(SiMe3), 2.03 (NCH2), 3.48 (NCH2), 3.73 (WtCCH3, 3J WH ) 7.9);
13C{1H} NMR δ 4.91 (SiMe3), 35.58 (CCH3, 1J CH ) 124), 52.06
(NCH2), 52.31 (NCH2), 274.33 (WtCCH3). Anal. Calcd for C17-
to yield yellow crystals; yield 0.21 g (99%): 1H NMR δ 0.22
(SiMe3), 1.88 (m, 4, -CCH2(CH2)2CH2), 2.11 (NCH2), 3.34
(NCH2), 3.97 (br s, 2, -CCH2(CH2)2CH2), 5.11 (br s, 2, -CCH2-
(CH2)2CH2), 19.66 (br s, 1, W-H); 13C{1H} NMR (-40 °C;
toluene-d8) δ 3.95 (SiMe3), 5.16 (SiMe3), 28.40 (-CCH2(CH2)2-
CH2), 30.47 (-CCH2(CH2)2CH2), 52.33 (NCH2), 51.40 (NCH2),
54.20 (NCH2), 55.87 (NCH2), 268.37 (WdC). Anal. Calcd for
H
42N4Si3W: C, 35.78; H, 7.42; N, 9.87. Found: C, 35.57; H,
7.81; N, 9.67.
[N3N]WtC-P r n (3c). This compound was prepared in a
manner analogous to that used to prepare 3b from [N3N]WCl
(0.20 g, 0.345 mmol) and LiBun (0.30 mL, 0.483 mmol, 1.6 M
in hexane) at 25 °C; yield 0.15 g (73%): 1H NMR δ 0.53 (SiMe3),
0.94 (t, 3, -CH2CH2CH3), 2.00 (m, 2, -CH2CH2-), 2.04 (NCH2),
3.51 (NCH2), 4.23 (m, 2, -CH2CH2CH3); 13C{1H} NMR δ 281.9
(WtC), 53.55 (WtCCH2-), 52.23 (NCH2), 25.52 (-CH2CH2-
CH3), 14.82 (-CH2CH2CH3), 5.06 (SiMe3). Anal. Calcd for
C
20H48N4Si3W: C, 39.20; H, 7.90; N, 9.14. Found: C, 39.04;
H, 7.44; N, 9.14.
The decomposition of 5 was followed by 1H NMR through
at least three half-lives at three temperatures: at 45 °C, k )
4.8 × 10-5 s-1; at 60 °C, k ) 2.3 × 10-4s-1; at 80 °C, k ) 1.8 ×
10-3 and 2.0 × 10-3 s-1. An Eyring plot yielded ∆Hq ) 22 463
cal/mol and ∆Sq ) -7.84 eu with R ) 0.9997.
C
19H46N4Si3W: C, 38.12 H, 7.74; N, 9.36. Found: C, 38.34;
H, 8.03; N, 9.19.
[N3N]W(C6H10)(H). To a cold solution (-40 °C) of 138 mg
(0.24 mmol) of [N3N]WCl in 10 mL of diethyl ether was added
37 mg (0.41 mmol) of cyclohexyllithium. The orange suspen-
sion was stirred for 1.5 h at 22 °C. The solvent was removed,
and the residue was extracted with 10 mL of pentane and
filtered through Celite. The solvent was reduced to 1 mL and
cooled to -40 °C. After 12 h, orange needles were collected
by filtration; yield 110 mg (74%): 1H NMR δ 3.1 (br s, CH2),
1.92 (br s, CH2), 1.80 (br m, Cy), 1.63 (br m, Cy), 0.41 (br s,
TMS); 1H NMR (toluene-d8, -40 °C) 19.4 (s, 1J HW ) 102, WH),
4.53 (br s, 2, cy), 3.33 (br s, 2), 3.2 (br s, 2), 3.05 (br s, 2), 2.35
(br s, 2), 2.06 (br s, 4), 1.93 (br s, 2), 1.78 (br s, 4), 1.68 (br s,
2), 0.41 (s, 9, TMS), 0.18 (s, 18, TMS); 13C{1H} NMR (toluene-
d8, -40 °C) 252.0 (s, WC), 57.6 (NCH2), 54.2 (NCH2), 53.8
(NCH2), 51.1 (NCH2), 30.4 (cyclohexyl CH2), 27.4 (cyclohexyl
CH2), 24.5 (cyclohexyl CH2), 23.7 (cyclohexyl CH2), 4.1 (TMS),
3.8 (TMS). Anal. Calcd for C21H50N4Si3W: C, 40.88; H, 8.04;
N, 8.94. Found: C, 40.53; H, 8.00; N, 8.83.
[N3N]WtCP h (3d ). KCH2Ph (0.068 g, 0.52 mmol) was
added to a THF (∼20 mL) solution of [N3N]WCl (0.20 g, 0.345
mmol) at 25 °C, and the mixture was stirred for 5 h. The
resulting yellow-orange mixture was evaporated to dryness in
vacuo, and the residue was extracted with pentane (∼100 mL)
and ether (∼10 mL). The extract was filtered through a pad
of Celite, and the solvent was removed in vacuo to give a brown
oil. The oil was dissolved in 5 mL of pentane, and the solution
was chilled to -40 °C to obtain yellow crystals; yield 0.16 g
(73%): 1H NMR δ 0.51 (SiMe3), 2.10 (NCH2), 3.54 (NCH2), 6.88
(t, Hpara), 7.36(t, Hmeta), 7.44 (d, Hortho); 13C{1H} NMR δ 5.10
(SiMe3), 52.39 (NCH2), 52.63 (NCH2), 125.12 (Cpara), 126.82
(Cmeta), 134.58 (Cortho), 152.03 (Cipso), 277.18 (WtC).
[N3N]WtCSiMe3 (3e). The reaction procedure is similar
to the synthesis of [N3N]WtCCH3, except LiCH2SiMe3 was
employed instead of LiEt; yield 91%: 1H NMR δ 0.51 (CSiMe3),
0.52 (SiMe3), 1.98 (NCH2), 3.47 (NCH2); 13C{1H} NMR δ 319.43
(WtC), 52.57 (NCH2), 52.50 (NCH2), 5.97 (CSiMe3), 5.01
(SiMe3). Anal. Calcd for C19H48N4Si4W: C, 36.29; H, 7.69; N,
8.91. Found: C, 36.32; H, 7.52; N, 8.94.
[N3N]WtC-Bu t (3f). The reaction procedure is similar to
the synthesis of [N3N]WtCCH3, except LiCH2-But was em-
ployed; yield 88%: 1H NMR δ 0.46 (SiMe3), 1.64 (CMe3), 2.05
(NCH2), 3.35 (NCH2); 13C{1H} NMR δ 297.93 (WtC), 55.70
(NCH2), 51.59 (CMe3), 50.79 (NCH2), 35.51 (CMe3), 4.67
(SiMe3). Anal. Calcd for C20H48N4Si3W: C, 39.20; H, 7.90; N,
9.14. Found: C, 39.09; H, 8.18; N, 8.88.
[N3N]W(CHCH2CH2CH2) (4). A solution of 126 mg (0.22
mmol) of [N3N]WCl in 10 mL of diethyl ether was cooled to
-40 °C, and a cold (-40 °C) solution of 17 mg (0.27 mmol) of
cyclobutyllithium in 2 mL of diethyl ether was added. The
color changed from orange to red. The solution was stored at
-40 °C overnight. The solvent was removed in vacuo, and
the residue was extracted with pentane. The extract’s volume
[N3N]WH (6). [N3N]W(C5H8)(H) 653 mg (1.06 mmol) and
10 mL of toluene were added to a 100 mL glass bomb that
was then sealed and heated to 45 °C for 24 h. The volatile
components were removed in vacuo, and the resulting yellow
crystalline material was extracted with minimum pentane.
The extract was filtered through a 1 cm plug of Celite, and
the filtrate was reduced in volume by half and stored at -35
°C for 1 day. The resulting yellow-orange needles were
collected by filtration; yield 501 mg (0.920 mmol, 87%): 1H
NMR δ 17.7 (SiMe3), -29.5 (NCH2), -92.5 (NCH2); IR (Nujol)
cm-1 1766.3 (νMH). Anal. Calcd for C15H40N4Si3W: C, 33.08;
H, 7.40 N, 10.29 Found: C, 32.67; H, 7.44; N, 10.09.
[N3N]W(1-cyclop en ten yl). Cyclopentenyl lithium (714 µL
of a 1.7 M solution, 1.21 mmol) was added to 471 mg (0.809
mmol) of [N3N]WCl in 50 mL of ether. After 1 h, the volatile
components of the reaction were removed in vacuo and the
residue was extracted with 10 mL of pentane. The extract
was filtered through Celite and cooled to -40 °C to give deep
red crystals; yield 363 mg (0.550 mmol, 68%): 1H NMR δ 15.14
(TMS), 8.88 (CH2), -30.57 (NCH2), -54.81 (NCH2). Anal.
Calcd for C20H46N4Si3W: C, 39.33; H, 7.59; N, 9.17. Found:
C, 39.72; H, 7.60; N, 9.09.
was reduced to 3 mL, and after 2 h red needles were collected
3
by filtration; yield 86 mg (66%): 1H NMR δ 11.43 (t, J HH
)
4.8, WCH), 5.24 (q, 2, WCHCH2), 3.28 (t, 8, NCH2 and WC4-
CH2), 2.62 (t, 2, WC4CH2), 2.0 (NCH2), 0.38 (TMS); 13C NMR
1
1
265 (d, J CH ) 130, WCH), 92.79 (t, J CH ) 120, WC4 CH2),
58.06 (NCH2), 57.06 (t, 1J CH ) 124, WC4 CH2), 52.69 (t, 1J CH
)
)
1
1
135, NCH2), 36.49 (t, J CH ) 127, WC4 CH2), 3.38 (q, J CH
[(Me3SiNCH2CH2)2NCH2CH2N]W(1-(tr im eth ylsilyl)cy-
clop en ten e) (7). A solution of [N3N]W(1-pentenyl) (120 mg,
0.196 mmol) in 5 mL of benzene was stirred at room temper-
ature for 24 h. The volatile components were removed in vacuo
and the orange solid was extracted with minimum pentane
(∼5 mL). The extract was cooled to -40 °C to yield 104 mg
(0.171 mmol, 87%) of orange crystals: 1H NMR δ 5.05 (m, 1,
W(TMSC5H7)), 4.09 (dd, 1, W(TMSC5H7)), 3.84 (m, 1, W-
(TMSC5H7)), 3.53 (dt, 2), 3.25 (m, 4), 2.60 (m, 2), 2.47 (dd, 1,
W(TMSC5H7)), 2.15 (m, 4), 2.08 (d, 1), 1.66 (m, 2), 0.329 (s,
119, TMS). Anal. Calcd for C19H46N4Si3W: C, 38.12; H, 7.74;
N, 9.36. Found: C, 38.19; H, 7.83; N, 9.03.
The decomposition of [N3N]W(CHCH2CH2CH2) to give
[N3N]WtCCH2CH2CH3 was followed in toluene by UV-vis
spectroscopy at 480 nm. The values for k (×10-4 s-1) at
temperature T (K) that were obtained are 0.27 (313), 0.75
(323), 2.03 (333), 5.48 (343). A plot of ln(k/T) versus 1/T gave
∆Hq ) 20 731 cal/mol and ∆Sq ) -13.34 eu.
[N3N]W(C5H8)(H) (5). A solution of [N3N]WCl (0.20 g,
0.345 mmol) in ether (∼25 mL) was treated with LiC5H9 (0.05
g, 0.66 mmol) at 25 °C. The reaction turned light yellow
immediately. The resulting mixture was stirred for 3 h, and
the volatile components were removed in vacuo. The yellow
reaction residue was extracted with pentane (∼80 mL), and
the extract was filtered through Celite. The filtrate was
concentrated to ∼3 mL and chilled at -40 °C for several hours
18, TMS), 0.142 (s, 9, TMS); 13C NMR δ 83.8 (s, WCH, J CW
)
38, J CH ) 156), 70.4 (s, CTMS), 65.8 (s, CH2, J CW ) 18.3), 59.6
(s, CH2), 57.9 (s, CH2), 57.1 (s, CH2), 54.3 (s, CH2), 50.9 (s,
CH2), 40.5 (s, CH2), 36.0 (s, CH2), 34.9 (s, CH2), 4.47 (s, TMS),
1.52 (s, TMS), -0.046 (s, TMS). Anal. Calcd for C20H46N4-
Si3W: C, 39.33; H, 7.59; N, 9.17. Found: C, 39.68; H, 7.82;
N, 9.21.

5208 Organometallics, Vol. 16, No. 24, 1997
Schrock et al.
X-r a y Str u ctu r e of [N3N]W(CHCH2CH2CH2) (4). The
structure was solved at 183 K using a Siemens SMART/CCD
diffractometer, see Supporting Information. Empirical for-
mula C19H46N4Si3W, fw ) 598.71, space group Pbca, a )
17.494(3) Å, b ) 15.860(4) Å, c ) 19.350(4) Å, V ) 5369(2) Å3,
Z ) 8, Dcalc ) 1.474 Mg/m3.
X-r a y Str u ctu r e of [(Me3SiNCH2CH2)2(NCH2CH2)N]W-
(1-(tr im eth ylsilyl)cyclop en ten e) (7). The structure was
solved at 188 K using a Siemens SMART/CCD diffractometer,
See Supporting Information. Empirical formula C20H46N4-
Si3W, fw ) 610.71, space group P21/n, a ) 14.676(4) Å, b )
11.402(4) Å, c ) 16.758(3) Å, â ) 107.366(14), V ) 2676.3(12)
Å3, Z ) 4, Dcalc ) 1.513 Mg/m3.
Solid -Sta te Ma gn etic Su scep tibility Mea su r em en ts.
SQUID experiments were performed at 5 KG on a Quantum
Design 5.5 T instrument running MSRP2 software. Samples
were prepared in an N2-filled drybox. A gelatin capsule and
a 2.2 × 1.9 cm piece of parafilm were weighed. The capsule
was then loaded with the sample, and the parafilm was folded
and packed on top using plastic tongs. The capsule was closed
and weighed again to determine the sample mass. It was then
suspended in a straw. The straw was placed in a plastic bottle
with a screw cap, and the bottle was tightly sealed. At the
instrument, the straw was quickly attached to the sample rod
and transferred to the helium atmosphere. Measurements
were taken in 1° intervals from 5 to 10 K, 2° intervals from
12 to 20 K, 3° intervals from 23 to 50 K, 5° intervals from 55
to 100 K, 10° intervals from 110 to 200 K, and 20° intervals
from 220 to 300 K. A background measurement of an empty
gel capsule, parafilm square, and straw was taken over the
entire temperature range and subtracted from the experimen-
tal values at each temperature. The susceptibility at each
temperature also was corrected for the diamagnetic contribu-
tion by the ligands using Pascal’s constants.
Ack n ow led gm en t. We thank the National Science
Foundation for support (Grant No. CHE 91 22827).
N.C.M.-Z. is also grateful for a Ciba-Geigy J ubila¨ums
Stiftung. We also thank the NSF for funds to help
purchase a departmental Siemens SMART/CCD diffrac-
tometer, Lan-Chang Liang for determining the magnetic
moment of [N3N]WCl, and Professor William Reiff for
helpful discussions and advice.
Su p p or tin g In for m a tion Ava ila ble: Labeled ORTEP
diagrams and tables of crystal data and structure refinement,
atomic coordinates, bond lengths and angles, and anisotropic
displacement parameters for [N3N]W(CHCH2CH2CH2) and
[(Me3SiNCH2CH2)2NCH2CH2N]W(1-(trimethylsilyl)cyclopen-
tene) (12 pages). Ordering information is given on any current
masthead page. Supporting information for the structure of
[N3N]W(C5H8)(H) can be found in the previous communication.10
OM970670M