Synthesis of Group 4 [(RN-o-C6H4)2O]2- complexes where R is SiMe3 or 0.5 Me2SiCH2CH2SiMe2

Article abstract of DOI:10.1016/S0022-328X(99)00450-7

Complexes that contain the [(Me₃SiN-o-C₆H₄)₂O]^2- ligand ([1]^2-) of the type [1]M(NMe₂)₂, [1]MCl₂, and [1]MMe₂ have been prepared where M=Ti, Zr, or Hf. Although cations prepared by addition of [Ph₃C][B(C₆F₅)₄] or [PhNMe₂H][B(C₆F₅)₄] to [1]ZrMe₂ or [1]HfMe₂ could not be observed in NMR studies, addition of [(η⁵-C₅H₄Me)₂Fe][B(C ₆H₅)₄] to [1]HfMe₂ in the presence of THF led to isolation of {[1]HfMe(THF)₂}[B(C₆H₅)₄]. An X-ray study showed the cation to be a distorted octahedron in which the [1]^2- ligand is in the mer arrangement and is significantly twisted from a planar NC₂OC₂N arrangement. The THF ligands are trans to one another. No well-behaved activity for the polymerization of 1-hexene could be observed with activated [1]ZrMe₂, while {[1]HfMe(THF)₂}[B(C₆H₅)₄] was inactive. The reaction between Li₂[O(o-C₆H₄NH)₂] and Me₂ClSiCH₂CH₂SiMe₂Cl in THF produced a cyclic diamido/ether ligand H₂[2]. The reaction between H₂[2] and Zr(NMe₂)₄ or ZrR₄ (R=CH₂Ph, CH₂SiMe₃) gave [2]Zr(NMe₂)₂(HNMe₂) and Zr[2]₂, respectively. The dimethylamine in [2]Zr(NMe₂)₂(HNMe₂) could be replaced with pyridine or 2,4-lutidine to give [2]Zr(NMe₂)₂(L) (L=pyridine or 2,4-lutidine), which then could be converted into [2]ZrCl₂(L) with excess Me₃SiCl. The reaction between [2]ZrCl₂(py) and two equivalents of Me₃SiCH₂MgCl gave a bimetallic complex in which one of the trimethylsilyl methyl groups has been doubly C-H activated, as confirmed by X-ray crystallography.

Full text of DOI:10.1016/S0022-328X(99)00450-7

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 591 (1999) 163–173
Synthesis of Group 4 [(RN-o-C6H4)2O]2− complexes where R is
SiMe3 or 0.5 Me2SiCH2CH2SiMe2
Richard R. Schrock *, Lan-Chang Liang, Robert Baumann, William M. Davis
Department of Chemistry 6-331, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
Received 17 June 1999; accepted 29 July 1999
Abstract
Complexes that contain the [(Me3SiN-o-C6H4)2O]2− ligand ([1]2−) of the type [1]M(NMe2)2, [1]MCl2, and [1]MMe2 have been
prepared where M=Ti, Zr, or Hf. Although cations prepared by addition of [Ph3C][B(C6F5)4] or [PhNMe2H][B(C6F5)4] to
[1]ZrMe2 or [1]HfMe2 could not be observed in NMR studies, addition of [(h5-C5H4Me)2Fe][B(C6H5)4] to [1]HfMe2 in the
presence of THF led to isolation of {[1]HfMe(THF)2}[B(C6H5)4]. An X-ray study showed the cation to be a distorted octahedron
in which the [1]2− ligand is in the mer arrangement and is significantly twisted from a planar NC2OC2N arrangement. The THF
ligands are trans to one another. No well-behaved activity for the polymerization of 1-hexene could be observed with activated
[1]ZrMe2, while {[1]HfMe(THF)2}[B(C6H5)4] was inactive. The reaction between Li2[O(o-C6H4NH)2] and
Me2ClSiCH2CH2SiMe2Cl in THF produced a cyclic diamido/ether ligand H2[2]. The reaction between H2[2] and Zr(NMe2)4 or
ZrR4 (R=CH2Ph, CH2SiMe3) gave [2]Zr(NMe2)2(HNMe2) and Zr[2]2, respectively. The dimethylamine in [2]Zr(NMe2)2(HNMe2)
could be replaced with pyridine or 2,4-lutidine to give [2]Zr(NMe2)2(L) (L=pyridine or 2,4-lutidine), which then could be
converted into [2]ZrCl2(L) with excess Me3SiCl. The reaction between [2]ZrCl2(py) and two equivalents of Me3SiCH2MgCl gave
a bimetallic complex in which one of the trimethylsilyl methyl groups has been doubly CꢀH activated, as confirmed by X-ray
crystallography. © 1999 Elsevier Science S.A. All rights reserved.
Keywords: Titanium; Zirconium; Hafnium; Diamido; Polymerisation of hexene
1. Introduction
12]. Therefore, we thought it would be informative to
explore complexes that contain a trimethylsilyl analog
of the most successful of our diamido/donor ligands,
namely [(t-BuN-o-C6H4)2O]2− [1–3] , as we could then
directly compare the activity of Group 4 complexes that
contain the [(Me3SiN-o-C6H4)2O]2− ligand ([1]2−) with
catalysts that contain the [(Me3CN-o-C6H4)2O]2−
ligand. We have postulated that an important feature
of the successful living polymerization of 1-hexene
using activated [(t-Bu-d6-N-o-C6H4)2O]ZrMe2 is stabili-
zation of crowded tetrahedral cationic intermediates
in which the olefin inserts into the metal–carbon bond
virtually exclusively in a 1,2 manner to give inter-
mediates in which b elimination is slow. In order to
encourage the formation of tetrahedral cationic zirco-
nium alkyls we also have begun to explore diamido/
donor ligands in which the donor cannot invert when
bound to zirconium, namely those that contain N [8]
or P [7] , and those in which the geometry of the
We recently reported the synthesis of zirconium di-
alkyl complexes that contain a variety of diamido/
donor ligands such as [(RN-o-C6H4)2O]2− (R=t-Bu
[1–3] , i-Pr [4], or cyclohexyl [4]), [(t-Bu-d6-N-o-
C6H4)2S]2− [5], [(ArylNCH2CH2)2O]2− [6], [(Aryl-
NCH2CH2)2S]2− [6] , [(ArylNSiMe2CH2)2PPh]2− [7],
or [(ArylNCH2CH2)2NR]2− (R=H or Me) [8], and the
behavior of activated dialkyl complexes for the poly-
merization of 1-hexene. The activity of zirconium
dimethyl complexes activated with [PhNMe2H]-
[B(C6F5)4] or [Ph3C][B(C6F5)4] varied from living (at
0°C) in the case of [(t-Bu-d6-N-o-C6H4)2O]ZrMe2 to
little sustained activity in the case of complexes that
contain sulfur or phosphorus donors. At the time we
began this work zirconium complexes containing
[(Me3SiNCH2CH2)2N(SiMe3)]2− had been reported [9–
* Corresponding author. Fax: +1-617-2537670.
diamido/donor complex is restricted as a conse-
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(99)00450-7

164
R.R. Schrock et al. / Journal of Organometallic Chemistry 591 (1999) 163–173
quence of tying the two arms together as part of a
tetrahydrofuran ring [13]. A third potential method of
encouraging formation of tetrahedral cationic zirco-
nium alkyls is to tie the two amido substituents to-
gether. That is relatively easy in the case of silicon, as
bis(chlorosilanes) are readily available. We report here
the synthesis of Group 4 complexes that contain the
[(Me3SiN-o-C6H4)2O]2− ligand ([1]2−), as well as sev-
eral zirconium complexes that contain an analogous
ligand ([2]2−) in which the two nitrogen centers are
linked by a Me2SiCH2CH2SiMe2 chain.
show a single resonance for the methyl groups and a
single resonance for the TMS groups, consistent with
C26 or C2 symmetry in solution on the NMR time-scale.
Attempts to prepare [1]ZrEt2 by addition of an ethyl
Grignard reagent to ether solutions of [1]ZrCl2 failed;
the mixture darkened within minutes and H-NMR
spectra of reaction aliquots at 25°C could not be
interpreted.
Attempts to prepare cationic species by addition of
[Ph3C][B(C6F5)4] to [1]MMe2 complexes in chloroben-
zene at 0°C led to complex mixtures that could not be
identified. However, it was possible to isolate a cationic
Hf species that contained two equivalents of THF with
[B(C6H5)4]− as the counter ion. Addition of [(h5-
C5H4Me)2Fe][B(C6H5)4] to a THF solution of [1]HfMe2
at −25°C led to formation of (h5-C5H4Me)2Fe. Re-
crystallization of the crude mixture yielded a colorless
product that was shown to be {[1]HfMe-
(THF)2}[B(C6H5)4] in an X-ray study (see below). Un-
fortunately, crystallization of {[1]HfMe(THF)2}-
[B(C6H5)4] was difficult and not entirely reproducible.
Therefore, samples of pure {[1]HfMe(THF)2}[B(C6H5)4]
could not be obtained. The complex and temperature
dependent NMR spectra of the crude product mixture
also suggested that more than one product (perhaps
more than one isomer of {[1]HfMe(THF)2}[B(C6H5)4])
was present, and possibly also that some fluxional
process was taking place. Therefore we cannot be cer-
2. Results
2.1. Synthesis of [(Me3SiN-o-C6H4)2O]2− complexes of
Ti, Zr, and Hf
Addition of two equivalents of LiBu to an ether
solution of (H2N-o-C6H4)2O followed by two equiva-
lents of Me3SiCl yielded white crystalline [(Me3SiNH-o-
C6H4)2O] (H2[1]) in ꢀ75% yield on a 10 g scale.
Addition of two equivalents of LiBu to a solution of
H2[1] in ether, followed by (NMe2)2TiCl2 gave orange
crystalline [1]Ti(NMe2)2 in ꢀ55% yield (Eq. (1)).
Li [1]
excess Me SiCl
3
(NMe2)2TiCl2eꢀth2ꢀeꢁr[1]Ti(NMe2)2tꢀoꢀluꢀenꢀeꢀ, ꢀ11ꢀ0ꢀ°Cꢁ [1]TiCl2
(1)
tain that
a
mixture of (h5-C5H4Me)2Fe and
Subsequent cleavage of the dimethylamido ligands with
Me3SiCl gave purple–black [1]TiCl2 in 74% yield after
7 days at 110°C in a sealed Schlenk tube. We did not
find conditions where the reaction between Ti(NMe2)4
and H2[1] would produce [1]Ti(NMe2)2 readily, nor
conditions where the reaction between TiCl4 and Li2[1]
would produce [1]TiCl2 readily. The analogous [1]ZrCl2
and [1]HfCl2 complexes can be prepared by the se-
quence of reactions shown in Eq. (2).
{[1]HfMe(THF)2}[B(C6H5)4] is in fact produced ini-
tially. Complexes of the type {[(t-Bu-d6-N-o-
C6H4)2O]HfMe(THF)2}BPh4 have been prepared and
found to exist in the form of two isomers in solution,
while {[(t-Bu-d6-N-o-C6H4)2O]ZrMe(THF)2}[B(C6F5)4]
and
{[(t-Bu-d6-N-o-C6H4)2O]ZrMe(DME)][B(C6F5)4]
have been isolated and their structures determined in
X-ray studies [3].
+H [1]
excess Me SiCl
2
3
X-ray-quality crystals of {[1]HfMe(THF)2}[B(C6H5)4]
were obtained by allowing diethyl ether vapor to diffuse
into a concentrated chlorobenzene solution. Crystallo-
graphic details are listed in Table 1 and selected bond
lengths and angles are listed in Table 2, while drawings
are presented in Fig. 1 (a and b). (Only the cation is
shown in Fig. 1; the [B(C6H5)4]− anion is well separated
from the cation and entirely normal.) The cation is a
distorted octahedron in which the [1]2− ligand is found
in the mer arrangement and is significantly twisted from
a planar NC2OC2N arrangement (Fig. 1(b)). The
N(1)ꢀHfꢀN(2) angle is 141.23(12)°, which is characteris-
tic of the structure of diamido/donor complexes in
which the ligand has approximately the mer arrange-
ment [3,6,8]. The twisting of the diamido/donor ligand
also can be evaluated via the O/Hf/N/Si dihedral angles
of 141 and 164°, and via the angle between the N(1)/
Hf/O(1) and N(2)/Hf/O(1) planes (169°). The nitrogens
are planar, but the sum of the angles at the oxygen of
M(NMe2)4−ꢀꢀ2ꢀMꢀeꢀNꢀHꢁ [1]M(NMe2)2 ꢀꢀꢀꢀꢀꢀꢀꢀꢁ [1]MCl2 (2)
2
M=Zr or Hf
All reactions are complete in 2 days or less at 22°C,
those involving Zr being significantly faster than those
involving Hf. H-NMR spectra of all [1]M(NMe2)2 com-
plexes contain a single sharp NMe2 resonance, consis-
tent with the dimethylamido ligands being equivalent
and freely rotating about the MꢀNMe2 bond on the
NMR time scale in solution. Any [1]MCl2 complex
could be a dimer containing bridging chlorides, judging
from a single crystal X-ray study of {[(t-BuN-o-
C6H4)2O]ZrCl2}2 reported elsewhere [3].
The [1]MCl2 complexes are smoothly methylated by
two equivalents of methyl Grignard reagent to give
[1]MMe2 complexes. [1]TiMe2 is an orange microcrys-
talline solid, while the zirconium and hafnium analogs
are colorless. Room temperature (r.t.) H-NMR spectra

R.R. Schrock et al. / Journal of Organometallic Chemistry 591 (1999) 163–173
165
the [1]2− ligand is only 351.6°. The THF ligands are
Table 2
,
Selected bond lengths (A) and angles (°) for {[(Me3SiN-o-
nearly trans to one another (O(2)ꢀHfꢀO(3)=
176.25(9)°), while O(1) is approximately trans to C(13)
(167.31(13)°). The ring of each THF ligand is tipped
slightly away from each neighboring TMS group in
response to steric demands. The two HfꢀOTHF bond
C6H4)2O]Hf(CH3)(THF)2}+
Bond lengths
HfꢀN(1)
HfꢀO(3)
HfꢀC(13)
2.118(3)
2.165(2)
2.220(4)
HfꢀN(2)
HfꢀO(2)
HfꢀO(1)
2.122(3)
2.172(2)
2.289(3)
,
lengths are ꢀ0.1 A shorter than the HfꢀO(1) dative
Bond angles
bond, in part because O(1) is likely to be a poorer
s-donor and a poorer p-donor, and because the lig-
and’s relatively rigid conformation may not allow the
HfꢀO(1) bond length to adjust to what it otherwise
could be. (The ZrꢀO bond lengths in two six-coordinate
mer [(i-PrN-o-C6H4)2O]2− complexes range from 2.33
N(l)ꢀHfꢀN(2)
N(l)ꢀHfꢀO(1)
N(l)ꢀHfꢀO(2)
N(l)ꢀHfꢀO(3)
N(l)ꢀHfꢀC(13)
N(2)ꢀHfꢀO(1)
N(2)ꢀHfꢀO(2)
N(2)ꢀHfꢀO(3)
N(2)ꢀHfꢀC(13)
O(1)ꢀHfꢀO(2)
141.23(12)
71.77(10)
87.93(10)
89.65(10)
105.61(13)
70.99(11)
86.18(11)
93.93 (11) C(7)ꢀO(l)ꢀC(1)
113.08(14)
96.92(10)
O(1)ꢀHfꢀO(3)
O(1)ꢀHfꢀC(13)
O(2)ꢀHfꢀO(3)
O(2)ꢀHfꢀC(13)
O(3)ꢀHfꢀC(13)
Si(l)ꢀN(l)ꢀHf
Si(2)ꢀN(2)ꢀHf
79.59(10)
167.31(13)
176.25(9)
95.37(13)
88.05(13)
125.9(2)
130.4(2)
126.2(3)
164
,
,
to 2.37 A [4].) The HfꢀC(13) bond length (2.220(4) A)
O(1)/Hf/N(1)/Si(1)
O(1)/Hf/N(2)/Si(2)
N(1)/Hf/O(1)/N(2)
is not unusual.
141
169
Table 1
Crystallographic data, collection parameters, and refinement parame-
ters for {[1]Hf(CH3)(THF)2}[B(C6H5)4] and [2][2%]Zr2(CH2SiMe3)2
We were somewhat surprised to find that a cation
can be isolated in the presence of a [B(C6H5)4]− anion
(instead of a relatively poorly coordinating [B(C6F5)4]−
Empirical formula
Formula weight
Temperature (K)
C51H65BHfN2O3Si2
999.53
185(2)
C44H68N4O2Si6Zr2
1036.00
188(2)
,
Wavelength (A)
Crystal system
Space group
0.71073
Monoclinic
P21/c
15.849(3)
13.977(2)
22.837(2)
90
106.458(13)
90
4851.9(12)
4
0.71073
Triclinic
(
P1
,
a (A)
13.328(4)
14.964(4)
18.212(3)
100.217(12)
103.932(14)
114.90(2)
3032.9(13)
2
,
b (A)
,
c (A)
h (°)
i (°)
k (°)
Volume (A )
Z
3
,
Dcalc (Mg m−3
Absorption
)
1.368
2.242
1.134
0.494
coefficient (mm−1
)
F(000)
2056
1080
Crystal size (mm)
q Range for data
collection (°)
0.42×0.26×0.18
1.34–23.26
0.28×0.24×0.13
1.58–23.26
Limiting indices
−175h517,
−85k515,
−255l524
−145h514,
−135k516,
−205l516
12 202
Reflections collected 19 232
Independent
reflections
Absorption
correction
6953 (Rint=0.0348) 8370 (Rint=0.0357)
Semi-empirical
None
Refinement method Full-matrix least-
Full-matrix least-
squares on F2
8361/0/523
squares on F2
Data/restraints/
6937/0/542
parameters
Goodness-of-fit on F 1.121
0.936
Final R indices
[I\2|(I)]
R indices (all data)
R1=0.0254,
wR2=0.0591
R1=0.0328,
wR2=0.0759
0.525 and −0.499
R1=0.0431,
wR2=0.1174
R1=0.0548,
wR2=0.1415
0.546 and −0.455
Largest difference
Fig. 1. (a) An ORTEP drawing of the structure of {[(Me3SiN-o-
peak and hole
C6H4)2O]Hf(CH3)(THF)2}+. (b) A CHEM 3D drawing of the struc-
−3
,
(e A
)
ture of {[(Me3SiN-o-C6H4)2O]Hf(CH3)(THF)2}+
.

166
R.R. Schrock et al. / Journal of Organometallic Chemistry 591 (1999) 163–173
anion). However, the strongly coordinating THF lig-
ands and resulting octahedral coordination geometry
prevent binding of the [B(C6H5)4]− anion to the metal
and possible destructive reactions between the two [14].
The apparent instability of {[1]HfMe(THF)2}[B(C6H5)4]
and our inability to obtain it in pure form might be the
result of dissociation of one THF ligand and subse-
quent destructive reactions involving [B(C6H5)4]−.
similarly. In [2]Zr(NMe2)2(py) the two dimethylamido
resonances are found at 3.04 and 2.76 ppm at r.t., while
in [2]Zr(NMe2)2(2,4-lut) they are found at 3.19 and 2.71
ppm at r.t. All NMR data of [2]Zr(NMe2)2(L) species
indicate that at all temperatures the TMS groups are
equivalent, and that added L exchanges during the
process in which the two dimethylamido methyl reso-
nances exchange. Since the two amido nitrogens are
linked together, we propose that the core is a pseudooc-
tahedron in which the [2]2− ligand is in a fac geometry,
the two NMe2 ligands are cis to one another and do not
rotate about the ZrꢀN bonds on the NMR time scale,
and the ‘L’ ligand is in a position trans to the oxygen
donor, viz.
2.2. Synthesis of [(CH2Me2SiN-o-C6H4)2O]2−
complexes of Zr
The reaction between Li2[O(o-C6H4NH)2] and
ClMe2SiCH2CH2SiMe2Cl in THF in dilute solution
gave crystalline H2[2] in ꢀ70% yield (Eq. (3)). H2[2]
exhibits one CH2 and one SiMe resonance in both
proton and C-NMR spectra, consistent with the pres-
ence of two mirror planes on the NMR time scale, as
expected.
We also propose that the fluxional process consists of
loss of L and exchange and rotation of the dimethy-
lamido ligands about the MꢀN bond in the resulting
five-coordinate species. It is not possible to say whether
the dimethylamido ligands lie in the ZrN4 plane, or
whether the dimethylamido plane contains the LꢀZrꢀO
axis. In either case the methyl groups on each dimethy-
lamido ligand would be inequivalent.
Reactions between [2]ZrCl2(py) and one or two
equivalents of MeMgCl, PhCH2MgCl, or LiMe under a
variety of conditions produced magnesium or lithium
salts, but only mixtures of products, none of which
could be identified. However, addition of two equiva-
lents of Me3SiCH2MgCl to [2]ZrCl2(py) in diethyl ether
at −35°C yielded a pale yellow crystalline solid in
ꢀ50% yield. This crystalline product also could be
obtained in ꢀ70% yield in the reaction between Li2[2]
and ZrCl2(CH2SiMe3)2 in diethyl ether at −35°C. An
X-ray study (Tables 1 and 3, Fig. 2) reveals this
product to be a dimer that has the composition
[2][2%]Zr2(CH2SiMe3)2, in which the [2%]4− ligand is
formed from a [2]2− ligand by loss of two protons from
C(28), thereby creating a bridging CH ligand between
the two metal centers, as shown schematically below.
(The phenylene rings are not shown.)
(3)
Attempts to react Li2[2] with ZrCl4 or ZrCl4(THF)2
under a variety of conditions did not yield any isolable
complexes. However, the reaction between H2[2] and
Zr(NMe2)4 in pentane at r.t. yielded a dimethylamine
adduct, [2]Zr(NMe2)2(HNMe2), as colorless crystals in
ꢀ90% yield (Eq. (4)). The dimethylamine appears to
coordinate to zirconium strongly since [2]Zr(NMe2)2-
(HNMe2) is unchanged after being heated to 100°C for
1 day in vacuo (ꢀ30 mtorr). The dimethylamine can
be replaced by pyridine or 2,4-lutidine to give related
[2]Zr(NMe2)2(L) species (L=pyridine or 2,4-lutidine)
in quantitative yield. Several attempts to convert
[2]Zr(NMe2)2(HNMe2) into [2]ZrCl2 by treatment with
excess Me3SiCl led to formation of H2[2] (observed by
H-NMR) and apparent decomposition. However, the
analogous pyridine or 2,4-lutidine adducts can be con-
verted readily to dichloride complexes in high yield (Eq.
(5)).
pentane
H2[2]+Zr(NMe2)4−ꢀꢀHꢀNꢀMꢀeꢁ [2]Zr(NMe2)2(HNMe2)
(4)
2
[2]Zr(NMe2)2(L)+2Me3SiCl
“[2]ZrCl2(L)+2Me3SiNMe2
(5)
H-NMR spectra of [2]Zr(NMe2)2(HNMe2) exhibit a
broad resonance centered at ꢀ2.86 ppm that can be
ascribed to the two dimethylamide ligands. This broad
resonance becomes two resonances below −20°C and
sharpens at higher temperatures (Tc=30°C, Dwo=238
Hz). The pyridine and 2,4-lutidine complexes behave
Metal center Zr(2) is pseudotetrahedral with bonds to
the two trimethylsilyl ligands, the bridging methine

R.R. Schrock et al. / Journal of Organometallic Chemistry 591 (1999) 163–173
167
Table 3
Zr(1)/N(2)/Si(2)/C(28) ring. The ZrꢀO, ZrꢀNamide, and
ZrꢀC bond lengths are all typical of zirconium com-
plexes that contain diamido/ether ligands [1,6,15] . The
bond lengths between the bridging methine carbon and
,
Selected bond lengths (A) and angles (°) in [2][2%]Zr2(CH2SiMe3)2
Bond lengths
Zr(1)ꢀN(1)
Zr(1)ꢀN(3)
Zr(1)ꢀO(1)
Zr(1)ꢀSi(2)
Zr(2)ꢀC(28)
Zr(2)ꢀC(6)
2.109(3)
2.154(4)
2.365(3)
2.884(2)
2.141(4)
2.253(4)
Zr(1)ꢀN(2)
Zr(1)ꢀC(28)
Zr(1)ꢀO(2)
Zr(2)ꢀN(4)
Zr(2)ꢀC(5)
Zr(2)ꢀC(22)
2.114(3)
2.236(4)
2.374(3)
2.082(4)
2.229(4)
2.768(4)
,
the two zirconium centers (2.236(4) and 2.141(4) A) are
typical of ZrꢀC single bonds. The Zr(1)ꢀC(28)ꢀZr(2)
angle (116.0(2)°), however, is ꢀ22° larger than the
ZrꢀCꢀZr angle (93.9(5)°) in {cyclo-ZrCHSiMe2-
NSiMe3[N(SiMe3)2]}2, a complex that also contains a
bridging methine connecting two zirconium centers and
that is formed by double CꢀH activation upon thermol-
ysis of [(Me3Si)2N]2ZrR2 (R=Me, Et, CH2SiMe3) com-
plexes at 60°C 0.01 mm−1 [16]. The sum of bond angles
around three of the amido nitrogens (N(1)=357.2,
N(3)=358.7, N(4)=355.5°) suggests that they are pla-
nar. However, the sum of the angles around N(2) is
only 349.2°, which may result from the strain produced
in the Zr(1)/N(2)/Si(2)/C(28) ring. The distance between
Bond angles
N(1)ꢀZr(1)ꢀN(2)
N(2)ꢀZr(1)ꢀN(3)
N(2)ꢀZr(1)ꢀC(28)
N(1)ꢀZr(1)ꢀO(1)
N(3)ꢀZr(1)ꢀO(1)
N(1)ꢀZr(1)ꢀO(2)
N(3)ꢀZr(1)ꢀO(2)
O(1)ꢀZr(1)ꢀO(2)
N(2)ꢀZr(1)ꢀSi(2)
C(28)ꢀZr(1)ꢀSi(2)
O(2)ꢀZr(1)ꢀSi(2)
N(4)ꢀZr(2)ꢀC(5)
N(4)ꢀZr(2)ꢀC(6)
C(5)ꢀZr(2)ꢀC(6)
95.85(13) N(1)ꢀZr(1)ꢀN(3)
146.96(13) N(1)ꢀZr(1)ꢀC(28)
75.70(14) N(3)ꢀZr(1)ꢀC(28)
155.31(12) N(2)ꢀZr(1)ꢀO(1)
70.80(11) C(28)ꢀZr(1)ꢀO(1)
73.82(12) N(2)ꢀZr(1)ꢀO(2)
86.01(12) C(28)ꢀZr(1)ꢀO(2)
82.27(9)
36.66(9)
40.34(10) O(l)ꢀZr(1)ꢀSi(2)
100.97(14)
104.3(2)
125.99(14)
82.09(11)
99.01(13)
71.61(11)
146.80(12)
94.98(11)
162.05(10)
97.64(7)
105.7(2)
106.5(2)
107.9(2)
30.47(12)
143.38(14)
N(1)ꢀZr(1)ꢀSi(2)
N(3)ꢀZr(l)ꢀSi(2)
106.47(7)
114.1(2)
117.7(2)
104.4(2)
N(4)ꢀZr(2)ꢀC(28)
C(28)ꢀZr(2)ꢀC(5)
C(28)ꢀZr(2)ꢀC(6)
N(4)ꢀZr(2)ꢀC(22)
,
Si(2) and Zr(1) (2.884(2) A) is typical of compounds
that contain M/N/Si/C rings [16,17] . The oxygen atoms
O(1) and O(2) coordinate to Zr(1) with bond lengths of
C(28)ꢀZr(2)ꢀC(22) 96.83(14) C(5)ꢀZr(2)ꢀC(22)
C(6)ꢀZr(2)ꢀC(22) 94.47(14)
,
ꢀ2.36–2.37 A and an O(1)ꢀZr(1)ꢀO(2) bond angle of
82°.
The NMR spectrum of [2][2%]Zr2(CH2SiMe3)2 is fully
in accord with the X-ray structure. The most character-
istic feature of the NMR spectra of this unsymmetric
compound is a singlet at 5.42 ppm in the H-NMR
spectrum and a doublet at 163.59 ppm (1JCH=106 Hz)
in the C-NMR spectrum for the bridging CH group.
The analogous resonances in the thermolysis product
of [(Me3Si)2N]2ZrR2 are found at 7.08 and 201.4 ppm
[16] .
The reaction between [2]ZrCl2(py) and two equiva-
lents of PhMe2CCH2MgCl gave an orange crystalline
product in ꢀ50% yield whose NMR spectra are
analogous to those for [2][2%]Zr2(CH2SiMe3)2, in partic-
ular a singlet resonance at 4.88 ppm that can be as-
cribed to the proton in the bridging methine group.
Elemental analysis further supports the proposal that
this product can be formulated as an analog of
[2][2%]Zr2(CH2SiMe3)2, namely [2][2%]Zr2(CH2CPhMe2)2.
The failure to observe similar products when
[2]ZrCl2(py) is treated with methyl Grignard or methyl-
lithium or with benzyl Grignard might be attributed to
the relatively small size of these alkyls in comparison to
trimethylsilylmethyl and neophyl, and competing de-
composition reactions; details remain obscure.
carbon C(28), and N(4) of an intact [2]2− ligand that
spans the two metal centers. (Carbon atom C(22) could
be said to be weakly bound to the metal judging from
,
the ZrꢀC(22) distance of 2.768 (4) A).) Metal center
Zr(1) is pseudooctahedral and contains the N(2)/Zr(1)/
N(1)/O(2) core of the [2%]4− ligand in a fac arrange-
ment. The N(1)ꢀZrꢀO(2) and N(2)ꢀZrꢀO(2) angles are
73.82(12) and 71.61(11)°, similar to what is found in
{[1]HfMe(THF)2}[B(C6H5)4], but the N(2)ꢀZrꢀN(1) an-
gle is only 95.85(13)°, perhaps in large part as conse-
quence of the restrictions imposed by the SiCH2CH2Si
link between N(1) and N(2) and the formation of the
Attempts to prepare dialkyl complexes of the type
[2]ZrR2 by adding H2[2] to ZrR4 (R=CH2Ph,
CH2SiMe3) in toluene or benzene led to 1:1 mixtures of
ZrR4 and a compound whose NMR spectra and ele-
mental analyses are consistent with its formulation as
Zr[2]2. Addition of two equivalents of H2[2] to ZrR4
yielded Zr[2]2 quantitatively. There is no evidence for
formation of [2]ZrR2, even if the reactions are con-
Fig. 2. An ORTEP drawing of the structure of [2][2%]Zr2(CH2SiMe3)2.

168
R.R. Schrock et al. / Journal of Organometallic Chemistry 591 (1999) 163–173
ducted in diethyl ether or THF. Complex Zr[2]2 shows
four SiCH3 resonances in both proton and C-NMR
spectra plus twelve aromatic signals in the C-NMR
spectrum, which suggests that the [2]2− ligand is bound
to the metal in a fac manner with the oxygen donors cis
to one another, as shown schematically in Eq. (6). If
[2]ZrR2 complexes are intermediates in these reactions,
we could conclude that the volume of the [2]2− ligand
is significantly smaller than the volume occupied by
[1]2− in analogous [1]ZrR2 complexes, so that further
reaction of [2]ZrR2 complexes with H2[2] is fast. How-
ever, it is still possible that Zr[2]2 forms without first
forming [2]ZrR2 intermediates.
[2][2%]Zr2(CH2SiMe3)2 is compelling evidence that there
will be a significant difference in the NꢀMꢀN angle in
[1]2− and [2]2− complexes, since the metal in both of
these complexes is six-coordinate. A smaller value for
the NꢀZrꢀN angle should lead to a stabilization of a
fac-pseudooctahedral coordination geometry and (for
steric reasons) to low stability of five-coordinate species
toward intermolecular reactions. In contrast, a larger
NꢀZrꢀN angle for steric reasons also should discourage
intermolecular CH activation in dialkyl species. It
should be possible to connect the two amido sub-
stituents with a longer link, e.g. Me2Si(CH2)xSiMe2
where x=3, 4, or 5, but we suspect that cations
prepared from such species also will not be stable.
In this work we have been able to compare directly a
successful catalyst that contains a t-butyl group on the
amido nitrogen with an analogous complex that con-
tains a trimethylsilyl group on the amido nitrogen.
Since the latter fails, we conclude that if the develop-
ment of active catalysts for a-olefin polymerization is
the goal, then silylamido groups should not be incorpo-
rated in the design of diamido/donor ligands. Similar
conclusions were reached in another recent paper in
which zirconium complexes were prepared that contain
diamido/phosphine ligands [7], although in that case no
complex that contains a diamido/phosphine ligand (and
no NꢀSi bond) has been shown to be an effective
catalyst for polymerization of ordinary olefins.
(6)
3. Discussion
We have shown that it is possible to prepare some
Group 4 dialkyl complexes that contain the [(Me3SiN-
o-C6H4)O]2− ligand, but monoalkyl cationic versions
apparently are not stable in the absence of coordinating
solvents such as THF. Nevertheless, we were somewhat
surprised that a crowded cationic monoalkyl pseudooc-
tahedral complex that contains two coordinated THF
ligands could be prepared, and that strong binding of
THF to the metal could force the [(Me3SiN-o-
C6H4)O]2− ligand to adopt the twisted mer arrange-
ment. Six-coordinate mer complexes are much more
readily accessible when the substituents on the amido
nitrogens are i-Pr [4], although an analogous zirconium
cation in which the amido substituent is t-Bu can also
be prepared [3], again presumably as a consequence of
strong binding of THF to the cationic Zr center.
4. Experimental
4.1. General procedures
All experiments were performed under a nitrogen
atmosphere in a Vacuum Atmospheres drybox or by
standard Schlenk techniques unless specified otherwise.
Tetrahydrofuran and diethyl ether were sparged with
nitrogen and passed through two columns of activated
alumina. Toluene was distilled from sodium benzophe-
none ketyl. Pentane was sparged with nitrogen and
passed through a column of activated alumina. All
Dialkyl complexes that contain the [2]2− ligand ap-
pear to be dramatically less stable than those that
contain the [1]2− ligand. We suggest that the lower
stability of hypothetical [2]ZrR2 species and (possibly)
the rapid reaction of intermediate [2]ZrR2 species with
more H2[2] to yield Zr[2]2 can be ascribed a significant
restriction of the NꢀZrꢀN bond angle in fac-[2]2−
complexes. However, we have assumed that the
NꢀZrꢀN angle in the fac coordinated ligand
(95.85(13)°) in [2][2%]Zr2(CH2SiMe3)2 is an indication of
what it might be in five-coordinate [2]ZrR2. Almost
certainly that will not be the case, although a compari-
son of the NꢀHfꢀN angle in {[1]HfMe(THF)2}-
[B(C6H5)4] with the NꢀZrꢀN angle around Zr(1) in
,
solvents were stored in the drybox over 4 A molecular
sieves. Molecular sieves and Celite were activated in
vacuo (10−3 torr) for 24 h at 175 and 125°C,
respectively.
NMR chemical shifts are listed as parts per million
downfield from tetramethylsilane. Routine coupling
constants are not reported. Spectra were obtained at
22°C in C6D6 unless otherwise noted. A standard vari-
able temperature unit was used to control the probe
temperature in variable temperature runs and tempera-
tures are considered accurate to 91°C. NMR solvents
,
were sparged with nitrogen and stored over 4 A molec-
ular sieves. Elemental analyses were performed by H.

R.R. Schrock et al. / Journal of Organometallic Chemistry 591 (1999) 163–173
169
4.4. [1]TiCl2
Kolbe Microanalytical Laboratory, Mu¨lheim an der
Ruhr, Germany. X-ray data were collected on a
Siemens SMART/CCD diffractometer with u(Mo–
A mixture of [1]Ti(NMe2)2 (1.00 g, 2.09 mmol) and
Me3SiCl (1.00 g, 9.22 mmol) in toluene (10 ml) was
heated in a sealed Schlenk tube to 100°C. After 7 days
the reaction mixture was filtered and all volatile compo-
nents were removed from the filtrate in vacuo. The
black residue was redissolved in methylene chloride
(ꢀ10 ml) and the solution was filtered and then concen-
trated in vacuo to ꢀ2 ml. The solution was layered
with pentane (ꢀ2 ml) and stored at −25°C to yield
595 mg of large deep purple–black crystals. Concentra-
tion of the mother liquor afforded a second crop of 117
,
Ka)=0.71073 A and solved using a full-matrix least-
squares refinement on F2. No absorption correction
was applied. High-resolution mass spectroscopy were
performed on a Finnigan MAT 8200 Sector Mass
Spectrometer.
O(o-C6H4NH2)2 [18], Zr(NMe2)4 [19], ZrCl2(CH2-
SiMe3)2 [20], Zr(CH2SiMe3)4 [21], Zr(CH2Ph)4 [22], and
TiCl(NMe2)2 [23] were prepared according to literature
procedures. All other chemicals were purchased from
commercial suppliers and used as received.
1
mg; total yield 712 mg (74%): H-NMR l 6.87 (d, 2),
6.79 (t, 2), 6.57 (t, 2), 6.36 (d, 2), 0.27 (s, 18, SiMe3);
13C-NMR l 148.5, 145.1, 126.9, 123.8, 119.7, 119.0, 1.47
(SiMe3). Anal. Calc. for C18H26Cl2N2OSi2Ti: C, 46.86;
H, 5.68; N, 6.07. Found: C, 46.75; H, 5.75; N, 6.03%.
4.2. (Me3SiNH-o-C6H4)2O (H2[1])
A solution of LiBu in hexane (51 ml, 1.6 M) was
added to a solution of (2-NH2C6H4)2O (8.08 g, 40.4
mmol) in THF (120 ml) at −25°C. The mixture was
allowed to warm to r.t. and was stirred for 4 h. Me3SiCl
(11.3 ml, 89 mmol) was then added at −25°C and the
solution was allowed to warm to r.t. After 8.5 h all
volatile components were removed in vacuo and the
residue was extracted with pentane (60 ml) over a
period of ꢀ15 min. A white solid was filtered off and
washed with pentane (20 ml). The solution was concen-
trated in vacuo and stored at −25°C overnight to yield
crystals of the colorless product; yield 10.61 g (76%):
1H-NMR l 6.88 (m, 6), 6.58 (m, 2), 4.21 (br s, 2, NH),
0.095 (s, 18, SiMe3); 13C-NMR l 146.1, 139.8, 124.9,
119.2, 118.6, 116.5, 0.3 (SiMe3). Anal. Calc. for
C18H28N2OSi2: C, 62.74; H, 8.19; N, 8.13. Found: C,
63.11; H, 8.52; N, 7.99%.
4.5. [1]TiMe2
[1]TiCl2 (261 mg, 566 mmol) was treated with MeMgI
(380 ml, 3.0 M in ether) at −25°C. The product was
isolated as described for [1]Ti(NMe2)2: yield 174 mg
1
(73%): H-NMR l 6.89–6.81 (m, 4), 6.76 (d, 2), 6.57 (t,
2), 1.63 (s, 6, TiMe2), 0.28 (s, 18, SiMe3); 13C-NMR l
149.0, 145.4, 126.4, 122.1, 121.7, 119.5, 66.6 (TiMe2),
1.8 (SiMe3). Anal. Calc. for C20H32N2Si2OTi: C, 57.12;
H, 7.67; N, 6.66. Found: C, 57.04; H, 7.65; N, 6.73%.
4.6. [1]ZrCl2 6ia [1]Zr(NMe2)2
H2[1] (1.29 g, 3.75 mmol) and Zr(NMe2)4 (1.00 g,
3.75 mmol) were dissolved in pentane (10 ml) at 25°C.
After 18 h all volatile components were removed in
vacuo. The off-white residue was identified as
[1]Zr(NMe2)2 on the basis of H- and C-NMR spectra:
1H-NMR l 6.91 (m, 6), 6.55 (m, 2), 2.93 (s, 12, NMe2),
0.24 (s, 18, SiMe3); 13C-NMR l 148.3, 146.4, 126.1,
123.3, 119.1, 117.6, 43.1, 2.4.
[1]Zr(NMe2)2 was dissolved in ether (20 ml) and
Me3SiCl (1.4 ml, 11.25 mmol) was added. After a few
minutes a solid began to precipitate. After 90 min the
volume of the mixture was reduced to ꢀ10 ml and
pentane (20 ml) was added. Copious amounts of pale
yellow powder precipitated. All volatile components
were removed and the yellow powder was washed with
pentane (10 ml) and then dried in vacuo; yield 1.85 g
(97%). An analytically pure sample was obtained by
recrystallization from hot toluene: 1H-NMR (CD2Cl2) l
7.19 (m, 4), 6.96 (m, 2), 6.81 (dd, 2), 0.24 (s, 18, SiMe3);
13C-NMR (CD2Cl2) l 147.85, 142.14, 127.98, 122.89,
122.54, 118.97, 1.07 (SiMe3). Anal. Calc. for
C18H26Cl2N2OSi2Zr: C, 42.84; H, 5.19; N, 5.55. Found:
C, 43.07; H, 5.15; N, 5.49%.
4.3. [1]Ti(NMe2)2
A solution of LiBu in hexane (6.1 ml, 1.6 M) was
added to a solution of H2[1] (1.678 g, 4.88 mmol) in
ether (30 ml) at −25°C. The solution was allowed to
warm to r.t. After 4 h Ti(NMe2)2Cl2 (1.01 g, 4.88
mmol) was added to this solution at −25°C. The
mixture was allowed to warm to r.t. and was stirred for
22 h, during which time a white precipitate formed. All
volatile components were removed in vacuo and the
residue was extracted with pentane (40 ml) for 30 min.
The extract was filtered and the pentane was removed
in vacuo. Recrystallization of the residue from ether at
−25°C produced orange crystals; yield 1.283 g (55%):
1H-NMR l 6.91 (m, 4), 6.80 (d, 2), 6.58 (t, 2), 3.13 (s,
12, NMe2), 0.21 (s, 18, SiMe3); 13C-NMR l 149.74,
148.23, 125.26, 122.26, 119.75, 117.97, 47.38 (NMe2),
3.03 (SiMe3). Anal. Calc. for C22H38N4OSi2Ti: C,
55.21; H, 8.00; N, 11.71. Found: C, 55.08; H, 8.11; N,
11.63%.

170
R.R. Schrock et al. / Journal of Organometallic Chemistry 591 (1999) 163–173
4.7. [1]HfCl2 6ia [1]Hf(NMe2)2
(SiMe3). Anal. Calc. for C20H32HfN2OSi2: C, 43.59; H,
5.85; N, 5.08. Found: C, 43.68; H, 5.79; N, 5.02%.
A solution of Hf(NMe2)4 (1.60 g, 4.51 mmol) and
H2[1] (1.55 g, 4.51 mmol) in pentane (30 ml) was
allowed to stand at r.t. for 3 days. All volatile compo-
nents were then removed in vacuo. The colorless
residue was identified as [1]Hf(NMe2)2 on the basis of
4.10. H2[2]
A 1 l three-necked round-bottom flask was equipped
with a magnetic stir-bar, two 120 ml additional funnels,
and a septum. The assembly was purged with nitrogen
for 1 h, and anhydrous THF (500 ml) was added by
syringe. Two solutions were prepared as follows and
transferred to the two additional funnels, respectively.
Solution 1: A 250 ml one-necked round-bottom flask
was charged with a magnetic stir-bar, O(o-C6H4NH2)2
(4.975 g, 0.025 mol) and THF (45 ml). The solution was
chilled in an acetone/dry ice cold bath. n-Butyllithium
(20.0 ml, 2.5 M in hexane, 0.050 mol) was added by
syringe. The solution turned green after about half the
n-butyllithium had been added, then orange eventually.
The solution was slowly warmed up to r.t. and stirred
at r.t. for an additional 3 h. The resulting solution was
transferred via a cannula to one of the additional
funnels to which extra THF was added to make up to
100 ml.
1
H- and C-NMR spectra: H-NMR l 6.92 (m, 6), 6.55
(t, 2), 3.00 (s, 12, NMe2), 0.24 (s, 18, SiMe3); 13C-NMR
l 148.3, 145.8, 126.3, 124.1, 119.3, 117.8, 42.9 (NMe2),
2.5 (SiMe3).
The [1]Hf(NMe2)2 was redissolved in ether (30 ml)
and Me3SiCl (1.5 ml, 11.8 mmol) was added to this
solution. The reaction mixture was stirred at r.t. for 2
days and the solvent was then removed in vacuo.
Recrystallization at −25°C of the residue from boiling
methylene chloride produced colorless microcrystalline
1
material; yield 2.09 g (78%): H-NMR (CD2Cl2) l 7.19
(m, 4), 6.93 (t, 2), 6.86 (d, 2), 0.22 (s, 18, SiMe3);
13C-NMR (CD2Cl2) l 147.8, 141.7, 128.1, 123.7, 122.0,
118.9, 1.3 (SiMe3). Anal. Calc. for C18H26Cl2HfN2OSi2:
C, 36.52; H, 4.43; N, 4.73. Found: C, 36.35; H, 4.39; N,
4.70%.
Solution 2:
A
THF solution (100 ml) of
4.8. [1]ZrMe2
ClMe2SiCH2CH2SiMe2Cl (5.348 g, 0.025 mol) was
placed in the other additional funnel.
A solution of MeMgI in ether (3.0 M, 710 ml) was
added to a suspension of [1]ZrCl2 (535 mg, 1.06 mmol)
in ether (10 ml) at −25°C. The reaction mixture was
allowed to warm to r.t. and was stirred for 20 min. All
volatile solvents were then removed in vacuo and the
residue was extracted with pentane (10 ml) for 15 min.
The extract was filtered and the pentane was removed
in vacuo. Recrystallization of the residue from a mix-
ture of ether and pentane at −25°C produced colorless
These two solutions were added at the same rate
(ꢀ2 drops s−1) to the 1 l three-necked round-bottom
flask containing 500 ml THF while the reaction solu-
tion was stirred vigorously. The reaction mixture was
stirred overnight and transferred to a 1 l one-neck
round bottom flask and solvent was removed on a
rotary evaporator. The residue thus obtained was ex-
tracted with pentane (300 ml) and the mixture was
filtered through Celite. Solvent was removed in vacuo
to give a brown oily residue which solidified after
several days. It was recrystallized from ether at −35°C
to give colorless crystals; yield (5 crops) 5.89 g (69%):
1H-NMR l 7.15 (d, 2, Ar), 6.90 (m, 4, Ar), 6.65 (m, 2,
Ar), 4.37 (br s, 2, NH), 0.70 (s, 4, CH2), 0.03 (s, 12,
SiMe); 13C-NMR (CDCl3) l 148.11 (Cipso), 139.47
(Cipso), 124.84 (CAr), 121.31 (CAr), 118.83 (CAr), 117.78
(CAr), 7.75 (CH2), 0.60 (SiMe). HRMS (EI, 70 eV):
342.15842. Calc. for C18H26N2OSi2: 342.15837.
1
crystals; yield 288 mg (55%): H-NMR l 6.85 (m, 6),
6.54 (m, 2), 0.82 (s, 6, ZrMe2), 0.26 (s, 18, SiMe3);
13C-NMR l 148.73, 143.94, 126.85, 123.16, 120.66,
118.94, 47.19 (ZrMe2), 1.53 (SiMe3). Anal. Calc. for
C20H32Si2N2OZr: C, 51.79; H, 6.95; N, 6.04. Found: C,
51.49; H, 7.20; N, 6.04%.
4.9. [1]HfMe2
A solution of MeMgI in ether (3.0 M, 1.7 ml) was
added to a suspension of [1]HfCl2 (1.51 g, 2.55 mmol)
in ether (30 ml) at −25°C. The reaction mixture was
stirred for 15 min. All volatile components were then
removed in vacuo and the residue was extracted with
pentane (30 ml) for 15 min. The extract was filtered and
the pentane was removed in vacuo. Recrystallization of
the residue from a mixture of ether and pentane at
−25°C produced colorless microcrystals; yield 1.056 g
4.11. [2]Zr(NMe2)2(HNMe2)
A solution of H2[2] (407 mg, 1.188 mmol) in pentane
(10 ml) was added to a solution of Zr(NMe2)4 (318 mg,
1.188 mmol) in pentane (10 ml) at r.t. The clear solu-
tion was shaken thoroughly then slowly poured into
another flask to initiate the formation of crystals. The
reaction solution was stood at r.t. overnight. The super
natant was decanted away from the colorless crystals,
which were then dried in vacuo. Repeated reduction of
1
(75%): H-NMR l 6.86 (m, 6), 6.54 (t, 2), 0.63 (s, 6,
HfMe2), 0.23 (s, 18, SiMe3); 13C-NMR l 148.56, 143.31,
127.08, 124.13, 120.69, 119.12, 58.06 (HfMe2), 1.57

R.R. Schrock et al. / Journal of Organometallic Chemistry 591 (1999) 163–173
171
the volume of the mother liquor afforded two more
crops; total yield (3 crops) 603 mg (90%): H-NMR l
CH2), 0.99 (m, 2, CH2), 0.71 (s, 6, SiMe), 0.13 (s, 6,
SiMe); 13C-NMR l 153.66 (Cipso), 150.58 (CAr), 145.28
(Cipso), 139.34 (CAr), 126.97 (CAr), 124.19 (CAr), 122.44
(CAr), 119.87 (CAr), 119.26 (CAr), 13.26 (CH2), 2.32
(SiMe), 0.54 (SiMe). Anal. Calc. for C23H29Cl2N3-
OSi2Zr: C, 47.48; H, 5.02; N, 7.22. Found: C, 47.59; H,
5.10; N, 7.16%.
1
7.33 (d, 2, Ar), 7.00–6.96 (m, 4, Ar), 6.59 (t, 2, Ar),
2.86 (br s, 12, ZrNMe2), 1.80 (d, 6, HNMe2), 1.43 (m,
2, CH2), 1.18 (hept, 1, HNMe2), 0.98 (m, 2, CH2), 0.49
(s, 6, MeSi), 0.15 (s, 6, MeSi); 13C-NMR l 153.73
(Cipso), 149.18 (Cipso), 126.07 (CAr), 123.74 (CAr), 120.02
(CAr), 117.49 (CAr), 44.04 (br s, ZrNMe2), 40.11
(HNMe2), 13.05 (CH2), 5.07 (SiMe), 0.79 (SiMe). Anal.
Calc. for C24H43N5OSi2Zr: C, 51.02; H, 7.67; N, 12.39.
Found: C, 50.88; H, 7.59; N, 12.32%.
4.15. [2][2%]Zr2(CH2SiMe3)2
4.15.1. Method (a)
An ether solution of Me3SiCH2MgCl (0.344 ml, 1 M
in diethyl ether, two equivalents) was added to a
prechilled solution (−30°C) of [2]ZrCl2(py) (100 mg,
0.172 mmol) in diethyl ether (4 ml). The reaction mix-
ture was stirred at r.t. for 20 min and the solvent was
removed in vacuo. The resulting solid residue was
extracted with pentane (8 ml) and the extract was
filtered through a bed of Celite. The filtrate was concen-
trated in vacuo to ꢀ1 ml and chilled to −30°C to give
pale yellow crystals of the product, which were isolated
by decanting the solution and removing all residual
solvent from the product in vacuo; yield 46 mg (52%).
4.12. [2]Zr(NMe2)2(py)
Neat pyridine (0.300 g, 3.793 mmol) was added to a
solution of [2]Zr(NMe2)2(HNMe2) (1.00 g, 1.77 mmol)
in ether (8 ml) at r.t. The solution turned yellow
immediately. After 4 h all volatile components were
removed in vacuo to give yellow crystalline solid; yield
1
1.050 g (99%): H-NMR l 8.53 (d, 2, Ar), 7.28 (d, 2,
Ar), 7.06 (d, 2, Ar), 6.99 (t, 2, Ar), 6.72 (t, 1, Ar), 6.57
(t, 2, Ar), 6.45 (t, 2, Ar), 3.04 (br s, 6, NMe2), 2.76 (br
s, 6, NMe2), 1.49 (m, 2, CH2), 1.06 (m, 2, CH2), 0.57 (s,
6, Me), 0.24 (s, 6, Me); 13C-NMR l 153.65 (Cipso),
150.78 (CAr), 149.25 (CAr), 138.31 (CAr), 125.88 (CAr),
124.44 (CAr), 123.40 (CAr), 119.98 (CAr), 117.54 (CAr),
45.03 (NMe), 44.22 (NMe), 12.73 (CH2), 4.90 (SiMe),
1.14 (SiMe). Anal. Calc. for C27H41N5OSi2Zr: C, 54.14;
H, 6.90; N, 11.69. Found: C, 53.94; H, 6.78; N, 11.56%.
4.15.2. Method (b)
Under reduced lighting, a cold solution (−30°C) of
LiCH2SiMe3 (90 mg, 0.961 mmol, two equivalents) in
diethyl ether (3 ml) was added to a vigorously stirred
cold solution (−30°C) of ZrCl4 (112 mg, 0.481 mmol)
in diethyl ether (3 ml). The reaction mixture was stirred
at r.t. for 30 min. Insoluble materials were filtered off
with Celite and the filtrate (containing ZrCl2(CH2-
SiMe3)2(Et2O)x) was chilled to −30°C. To this was
added solid Li2[2] (170 mg, 0.481 mmol). The reaction
mixture was stirred at r.t. overnight during which time
a white solid (LiCl), apparently precipitated. The solu-
tion was filtered through a bed of Celite and the filter
cake was washed with diethyl ether (2 ml). The com-
bined filtrate and washing were concentrated in vacuo
to ꢀ1 ml to form microcrystalline solid product. The
concentrated solution was chilled to −30°C and yellow
crystalline product was obtained by removing the solu-
tion portion with a pipet and drying in vacuo; yield 172
mg (69%).
4.13. [2]Zr(NMe2)2(2,4-lutidine)
Neat 2,4-lutidine (12 mg, 0.112 mmol) was added to
a solution of [2]Zr(NMe2)2(HNMe2) (12 mg, 0.021
mmol) in ether (1 ml) at r.t. The solution was stirred for
6.5 h and all volatile materials were removed in vacuo
to give yellow crystalline product; yield 13 mg (98%):
1H-NMR l 8.48 (d, 1, lut.), 7.25 (d, 2, Ar), 7.04 (d, 2,
Ar), 6.95 (t, 2, Ar), 6.54 (t, 2, Ar), 6.42 (m, 2, Ar), 3.19
(br s, 6, NMe2), 2.71 (br s, 6, NMe2), 2.42 (s, 3,
Me-lutidine), 1.73 (s, 3, Me-lutidine), 1.49 (m, 2, CH2),
1.04 (m, 2, CH2), 0.56 (s, 6, MeSi), 0.19 (s, 6, MeSi).
Anal. Calc. for C29H45N5OSi2Zr: C, 55.54; H, 7.23; N,
11.17. Found: C, 55.39; H, 7.20; N, 11.12%.
An X-ray quality crystal was obtained by recrystal-
lization from a concentrated diethyl ether solution at
−30°C. H-NMR l 7.24–6.36 (m, 16, Ar), 5.42 (s, 1,
4.14. [2]ZrCl2(py)
1
Neat Me3SiCl (1.4 ml, 10.764 mmol) was added to a
solution of [2]Zr(NMe2)2(py) (496 mg, 0.828 mmol) in
CH2Cl2 (10 ml) at r.t. The reaction solution was stirred
at r.t. overnight and solvent was removed in vacuo. The
resulting yellow solid was washed with pentane (3×5
ml) and the yellow powder was collected on a fine
fritted funnel dried in vacuo; yield 443 mg (92%):
1H-NMR l 8.88 (d, 2, Ar), 7.23 (d, 2, Ar), 6.80 (m, 2,
Ar), 6.68–6.50 (m, 5, Ar), 6.38 (t, 2, Ar), 1.84 (m, 2,
ZrꢀCHSiꢀZr), 1.43–0.75 (m, 8, diastereotopic Si-
(CH2)2Si), 1.01 (s, 3, NSiMe), 0.53 (s, 9, ZrCH2SiMe3),
0.40 (s, 3, NSiMe), 0.39 (s, 3, NSiMe), 0.32 (s, 9,
ZrCH2SiMe3), 0.24 (s, 3, NSiMe), 0.23 (s, 3, NSiMe),
−0.02 (s, 3, NSiMe), −0.60 (s, 2, ZrCH2SiMe3),
−0.91 (s, 2, ZrCH2SiMe3); 13C-NMR
l 163.59
1
(ZrCHZr, JCH=103), 157.55, 155.53, 152.23, 151.31,
147.92, 147.87, 146.65, 145.21, 144.66, 135.12, 134.80,
127.68, 127.21, 126.39, 126.30, 125.40, 124.26,

172
R.R. Schrock et al. / Journal of Organometallic Chemistry 591 (1999) 163–173
122.35, 120.95, 120.17, 118.75, 118.05, 117.20, 116.96,
71.29, 60.60, 34.79, 23.08, 15.79, 14.65, 9.38, 7.65, 5.04,
4.31, 4.15, 3.76, 0.39, −3.20, −5.13. Anal. Calc. for
C44H68N4O2Si6Zr2: C, 51.01; H, 6.62; N, 5.41. Found:
C, 51.19; H, 6.53; N, 5.37%.
standard hemisphere of data was obtained for both
compounds. Data reduction was carried out with SAINT
[24], while SHELXTL [24] was used to solve and refine
both structures. Patterson methods were employed to
locate the heavy atoms in each instance, while subse-
quent difference Fourier calculations revealed the posi-
tions of the remaining non-hydrogen atoms. Those
atoms were treated as anisotropic scatterers. Hydrogen
atoms were placed in calculated positions and were
allowed to ride upon their respective non-hydrogen
atoms. In the case of {[1]HfMe(THF)2}[B(C6H5)4] an
empirical absorption correction was applied.
4.16. [2][2%]Zr2(CH2CMe2Ph)2
To a prechilled solution (−30°C) of [2]ZrCl2(py)
(106 mg, 0.182 mmol) in diethyl ether (4 ml) was added
PhMe2CCH2MgCl (0.33 ml, 1.12 M in diethyl ether,
two equivalents). The reaction mixture was stirred at
r.t. for 2.5 h. Solvent was removed in vacuo to dryness.
The solid residue was extracted with pentane (10 ml).
The extract was filtered through a bed of Celite and
filtrate was concentrated in vacuo to ꢀ1 ml. The
concentrated solution was chilled to −30°C to give the
orange crystalline product, which was isolated by de-
canting the solution and removing all residual solvent
from the sample in vacuo; yield 33 mg (47%). Anal.
Calc. for C56H72N4O2Si4Zr2: C, 59.63; H, 6.43; N, 4.97.
Found: C, 59.73; H, 6.56; N, 4.88%.
5. Supplementary information
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
Data Centre, CCDC no. 132502 for {[1]HfMe-
(THF)2}[B(C6H5)4] and 132501 for [2][2%]Zr2(CH2-
SiMe3)2. Copies of this information may be obtained
free of charge from The Director, CCDC, 12 Union
Road, Cambridge CB2 1EZ (Fax: +44-1223-336-033
or e-mail: deposit@ccdc.cam.ac.uk or http://www.ccdc.
cam.ac.uk).
4.17. Zr[2]2
4.17.1. Method (a)
In the absence of light a mixture of Zr(CH2SiMe3)4
(56 mg, 0.127 mmol) and H2[2] (87 mg, 0.255 mmol) in
toluene (3 ml) was stirred at r.t. for 4 days. All volatile
components were removed in vacuo to yield the
product as an orange solid; yield 98 mg (99%).
Acknowledgements
R.R.S. thanks the Department of Energy (DE-FG02-
86ER13564) for supporting this research, while R.B.
thanks the Studienstiftung des Deutschen Volkes for a
predoctoral fellowship.
4.17.2. Method (b)
In the absence of light a mixture of Zr(CH2Ph)4 (27
mg, 0.059 mmol) and H2[2] (40 mg, 0.059 mmol) in
toluene (3 ml) was heated to 80°C for 4 days. After
removal of all volatile components the product was
References
[1] R. Baumann, R.R. Schrock, W.M. Davis, J. Am. Chem. Soc.
119 (1997) 3830.
1
obtained as an orange solid; yield 43 mg (94%). H-
NMR l 7.06 (dd, 4, Ar), 6.97 (t, 2, Ar), 6.88 (t, 4, Ar),
6.57 (t, 4, Ar), 6.45 (t, 4, Ar), 1.64 (m, 4, CH2), 1.05 (m,
4, CH2), 0.46 (s, 6, SiMe), 0.09 (s, 6, SiMe), −0.03 (s,
6, SiMe), −0.04 (s, 6, SiMe); 13C-NMR l 153.84 (C,
Ar), 153.80 (C, Ar), 140.82 (C, Ar), 145.93 (C, Ar),
127.05 (CH, Ar), 126.44 (CH, Ar), 122.72 (CH, Ar),
121.08 (CH, Ar), 120.54 (CH, Ar), 119.69 (CH, Ar),
119.22 (CH, Ar), 118.03 (CH, Ar), 12.28 (CH2), 10.62
(CH2), 5.30 (SiMe), 3.61 (SiMe), 1.76 (SiMe), 0.19
(SiMe). Anal. Calc. for C36H48N4O2Si4Zr: C, 55.98; H,
6.26; N, 7.25. Found: C, 56.09; H, 6.21; N, 7.20%.
[2] R. Baumann, R.R. Schrock, J. Organomet. Chem. 557 (1998) 69.
[3] R.R. Schrock, R. Baumann, R. Stumpf, W.M. Davis, L.-C.
Liang, J.T. Goodman, Organometallics 18 (1999) 3649.
[4] R. Baumann, R. Stumpf, W.M. Davis, L.-C. Liang, R.R.
Schrock, J. Am. Chem. Soc. 121 (1999) 7822.
[5] D.G. Graf, R.R. Schrock, W.M. Davis, R. Stumpf,
Organometallics 18 (1999) 843.
[6] M. Aizenberg, L. Turculet, W.M. Davis, F. Schattenmann, R.R.
Schrock, Organometallics 17 (1998) 4795.
[7] R.R. Schrock, S.W. Seidel, Y. Schrodi, W.M. Davis,
Organometallics 18 (1999) 428.
[8] L.-C. Liang, R.R. Schrock, W.M. Davis, D.H. McConville, J.
Am. Chem. Soc. 121 (1999) 5797.
[9] F.G.N. Cloke, P.B. Hitchcock, J.B. Love, J. Chem. Soc. Dalton
Trans. (1995) 25.
4.18. X-ray studies
[10] H.C.S. Clark, F.G.N. Cloke, P.B. Hitchcock, J.B. Love, A.P.
Wainwright, J. Organomet. Chem. 501 (1995) 333.
[11] A.D. Horton, J. de With, A.J. van der Linden, H. van de Weg,
Organometallics 15 (1996) 2672.
[12] A.D. Horton, J. de With, Chem. Commun. (Cambridge) (1996)
1375.
Crystals
of
{[1]HfMe(THF)2}[B(C6H5)4]
and
[2][2%]Zr2(CH2SiMe3)2 were examined on a Bruker Plat-
form diffractometer equipped with a CCD area detec-
tor and driven by the SMART [24] suite of programs. A

R.R. Schrock et al. / Journal of Organometallic Chemistry 591 (1999) 163–173
173
[13] M.A. Flores, M.R. Manzoni, R. Baumann, W.M. Davis, R.R.
Schrock, Organometallics, 18 (1999) 3220.
[19] G.M. Diamond, S. Rodewald, R.F. Jordan, Organometallics 14
(1995) 5.
[14] R.F. Jordan, Adv. Organomet. Chem. 32 (1991) 325.
[15] R.R. Schrock, F. Schattenmann, M. Aizenberg, W.M. Davis,
Chem. Commun. (1998) 199.
[16] R.P. Planalp, R.A. Andersen, A. Zalkin, Organometallics 2
(1983) 16.
[17] R.R. Schrock, S.W. Seidel, N.C. Mo¨sch-Zanetti, K.-Y. Shih,
M.B. O’Donoghue, W.M. Davis, W.M. Reiff, J. Am. Chem.
Soc. 119 (1997) 11 876.
[18] J.J. Randall, C.E. Lewis, P.M. Slagan, J. Org. Chem. 27 (1962)
4098.
[20] H. Brand, J.A. Capriotti, J. Arnold, Organometallics 13 (1994)
4469.
[21] M.R. Collier, M.F. Lappert, R. Pearce, J. Chem. Soc. Dalton
Trans. (1973) 445.
[22] U. Zucchini, E. Albizzati, U. Giannini, J. Organomet. Chem. 26
(1971) 357.
[23] E. Benzing, W. Kornicker, Chem. Ber. 94 (1961) 2263.
[24] SMART
, SAINT, and SHELXTL are part of the Bruker Analytical
X-ray Systems Software, Madison, WI, 1995.
.