Neutral and cationic [bis(η1-amidosilyl)-η5- cyclopentadienyl]titanium and -zirconium complexes: Synthesis, x-ray molecular structures and DFT calculations

Article abstract of DOI:10.1002/ejic.200200698

Treatment of LiNHtBu with THF solutions of C₅H₄(SiMe₂Cl)₂ gave C₅H₄(SiMe₂NHtBU)₂ (1). Deprotonation of 1 with M(NMe₂)₄ (M = Ti, Zr) under different conditions provided the monocyclopentadienyl complexes [M{η⁵-C₅H₃- [SiMe₂(NHtBu)]₂] (NMe₂)₃] [M = Ti (2), Zr (3)] and the single (η-amidosilyl)cyclopentadienyl compounds [M{η⁵-C₅H₃ [SiMe₂(NHtBu)][SiMe₂(η¹-NtBu)]} M = Ti (4), Zr (5)]. The related dibenzyl compounds [M{η⁵-C₅H₃ [SiMe₂(NHtBu)][SiMe₂ (η¹-NtBu)]}(CH₂Ph)₂] [M = Ti (6), Zr (7)] resulted from treatment of 1 with M(CH₂C₆H₅)₄ (M = Ti, Zr). Further deprotonation of the amido complexes 4 and 5 and the benzyl complexes 6 and 7 by heating in toluene solution gave the bis(η-amidosilyl)cyclopentadienyl complexes [M{η⁵-C₅H₃ [SiMe₂(η¹-NtBu)]₂}(NMe₂)] [M = Ti (8), Zr (9)] and [M{η⁵-C₅H₃ [SiMe₂(η¹-NtBu)]₂}(CH2Ph)] [M = Ti (10), Zr (11)], respectively. Treatment of the monobenzyl complexes 10 and 11 with B(C₆F₅)₃ yielded the cationic compounds [M{η⁵-C₅H₃ [SiMe₂(η¹-NtBu)]₂}]⁺ as [(CH₂Ph)B(C₆F₅)₃]^- [M = Ti (12), Zr (13)] salts. All new compounds were characterized by NMR spectroscopy, and the crystal structures of 10 and 13 were studied by diffraction methods. DFT calculations for the neutral and cationic species are described and provide an explanation for the unusual η¹ coordination of a phenyl ring to a group-4 metal cation. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejic.200200698

FULL PAPER
Neutral and Cationic [Bis(η1-amidosilyl)-η5-cyclopentadienyl]titanium and
-zirconium Complexes: Synthesis, X-ray Molecular Structures and DFT
Calculations
[a]
[a]
[b][‡]
Olivier Blacque,[b] Heinz Berke,[b] and
´
Jesus Cano, Pascual Royo,* Heiko Jacobsen,
Eberhardt Herdtweck[c]
Keywords: Cations / Cyclopentadienyl ligands / Density functional calculations / Nitrogen ligands / Titanium / Zirconium
Treatment of LiNHtBu with THF solutions of C5H4(SiMe2Cl)2
gave C5H4(SiMe2NHtBu)2 (1). Deprotonation of with
M(NMe2)4 (M = Ti, Zr) under different conditions provided
the monocyclopentadienyl complexes
[M{η5-C5H3-
[M{η5-C5H3[SiMe2(η1-NtBu)]2}(NMe2)] [M = Ti (8), Zr (9)]
and [M{η5-C5H3[SiMe2(η1-NtBu)]2}(CH2Ph)] [M = Ti (10), Zr
(11)], respectively. Treatment of the monobenzyl complexes
10 and 11 with B(C6F5)3 yielded the cationic compounds
[M{η5-C5H3[SiMe2(η1-NtBu)]2}]+ as [(CH2Ph)B(C6F5)3]− [M =
Ti (12), Zr (13)] salts. All new compounds were characterized
1
[SiMe2(NHtBu)]2}(NMe2)3] [M = Ti (2), Zr (3)] and the single
(η-amidosilyl)cyclopentadienyl compounds [M{η5-C5H3[Si-
Me2(NHtBu)][SiMe2(η1-NtBu)]}(NMe2)2] [M = Ti (4), Zr (5)]. by NMR spectroscopy, and the crystal structures of 10 and 13
The related dibenzyl compounds [M{η5-C5H3[SiMe2-
(NHtBu)][SiMe2(η1-NtBu)]}(CH2Ph)2] [M = Ti (6), Zr (7)] re-
sulted from treatment of 1 with M(CH2C6H5)4 (M = Ti, Zr).
were studied by diffraction methods. DFT calculations for the
neutral and cationic species are described and provide an
explanation for the unusual η1 coordination of a phenyl ring
Further deprotonation of the amido complexes 4 and 5 and to a group-4 metal cation.
the benzyl complexes 6 and 7 by heating in toluene solution
gave the bis(η-amidosilyl)cyclopentadienyl complexes
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
The technological applications[5] of these ‘‘constrained-ge-
ometry’’ catalysts are based not only on their electronic un-
saturation, but also on their sterically open reactive sites,
combined with rigidity of the molecular framework and
thermal stability. These features facilitate the coordination
of bulkier olefins, allowing the synthesis of various copoly-
mers of ethylene with 1-hexene, 1-octene, styrene and other
larger olefins.
The group-4 metallocenes [MCp2X2][1] and cylopentadi-
enylamido complexes [M(η5-C5R4-E-η1-NRЈ)X2][2] are the
two essential types of catalyst precursors for insertion poly-
merization of olefins based on the use of cyclopentadienyl
ligands. After alkyl abstraction by a Lewis acid {MAO,
B(C6F5)3, [CPh3][B(C6F5)4]} the metallocene compounds
generate the cationic, coordinatively unsaturated, 14-elec-
tron d0 [MCp2R]ϩ species,[3] whereas the cyclopentadienyl-
amido complexes provide the electronically more unsatu-
rated 12-electron d0 [M(η5-C5R4-E-η1-NRЈ)RЈЈ]ϩ cations.[4]
[a]
´
´
´
Departamento de Quımica Inorganica, Facultad de Quımica,
´
Universidad de Alcala, Campus Universitario,
´
28871 Alcala de Henares, Spain
Fax: (internat.) ϩ 34-91/8854683
Scheme 1. Types of metallocene and (cyclopentadienylsilyl)amido
cations
E-mail: pascual.royo@uah.es
[b]
Anorganisch-chemisches Institut der Universität Zürich
Winterthurerstrasse 190, 8057 Zürich, Switzerland
Fax: (internat.) ϩ 41-1/635-6802
E-mail: hberke@aci.unizh.ch
Anorganisch-chemisches Institut der Technischen Universität
München
Lichtenbergstrasse 4, 85747 Garching bei München, Germany
Present address: KemKom, 1864 Burfield Ave., Ottawa, Ontario
K1J 6T1, Canada, E-mail: jacobsen@kemkom.com
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
Amido ligands are π-donors and in this respect resemble
Cp ligands. Even if π-electrons are included, however, the
formal electron count of a single group never reaches that
of the Cp moiety. Replacement of one Cp by an amido
group, as demonstrated in structures I and II, establishes a
12-electron cyclopentadienylamido complex[6] on the basis
[c]
[‡]
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474
DOI: 10.1002/ejic.200200698
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2463

J. Cano, P. Royo, H. Jacobsen, O. Blacque, H. Berke, E. Herdtweck
FULL PAPER
of a 14-electron metallocene species I. Structures of type III isomers interconvert, probably through successive 1,2-shifts
with two amido ligands would indeed have the formal elec- of the silyl groups,[10] and increasing temperatures increase
tron count of I. Despite this, complexes III are special, since the a/b ratio to 4:1 at 50 °C, 8:1 at 70 °C and more than
they do not bear an alkyl group ready for ZieglerϪNatta- 10:1 at temperatures higher than 80 °C. The broad signals
type insertion polymerizations. Nevertheless, we have re- observed for 1a at 20 °C appeared as four singlets when the
ported[7] that they still have potential for polarized 1H NMR spectrum was recorded at Ϫ40 °C.
[11]
metalϪolefin binding similar to, for instance, the 18-elec-
The
tetraamide
M(NMe2)4
and
tetrabenzyl
[12]
tron Cp3Mϩ series (M ϭ Ti, Zr).[8] For this reason we de- M(CH2Ph)4
compounds (M ϭ Ti, Zr) were used to de-
cided to study the synthesis and chemical behaviour of protonate the cyclopentadiene compound 1, which pos-
[bis(amidosilyl)cyclopentadienyl]titanium and -zirconium sesses three acidic protons, to generate the metal-coordi-
complexes of type III.
nated cyclopentadienyl ligand.
Here we report the syntheses of monocyclopentadienyl
[M{η5-C5H3[SiMe2(NHtBu)]2}X3], (amidosilyl)cyclopenta-
dienyl [M{η5-C5H3[SiMe2(NHtBu)][SiMe2(η1-NtBu)]}X2]
and bis(amidosilyl)cyclopentadienyl [M{η5-C5H3(SiMe2-
η1-NtBu)2}X] group-4 metal compounds and their cationic
[M{η5-C5H3(SiMe2-η1-NtBu)2}]ϩ derivatives. The X-ray
molecular structures of the neutral [Ti{η5-C5H3[SiMe2(η1-
NtBu)]2}(CH2Ph)] and the cationic [Zr{η5-C5H3[SiMe2(η1-
NtBu)]2}]ϩ[(CH2C6H5)B(C6F5)3]Ϫ complexes were deter-
mined by diffraction methods and DFT calculations were
carried out for the neutral and cationic species.
Monocyclopentadienyl Compounds
Immediate deprotonation of the most acidic cyclopen-
tadiene proton took place when a toluene solution of 1 was
treated at room temperature with 1 equiv. of Zr(NMe2)4 to
give the monocyclopentadienyl compound 3 (Scheme 3) as
the unique reaction product after 1 h (100% NMR yield).
When Ti(NMe2)4 was used, heating at 50 °C was required
to complete the reaction, the analogous titanium derivative
2 being isolated. Compound 2 was isolated as a dark red
oil in quantitative yield after removal of the toluene and
extraction into hexanes. The related tribenzyltitanium and -
zirconium derivatives could not be isolated when
M(CH2Ph)4 was used[13] because further deprotonation to
give the mono- and bis(silylamido) benzyl derivatives took
place very easily at room temperature.
Results and Discussion
The cyclopentadiene C5H4(SiMe2NHtBu)2 (1) was iso-
lated as a yellow liquid by treatment of the reported
bis(chlorosilyl)cyclopentadiene[9] with 2 equiv. of LiNHtBu
The different thermal conditions required to deprotonate
the tribasic diaminocyclopentadiene 1 show that the depro-
tonation with irreversible toluene elimination by the tetra-
benzyl compounds M(CH2Ph)4 is much easier than the re-
versible elimination of amine by the amido derivatives
M(NMe2)4, and that the zirconium compounds are more
easily deprotonated than the more crowded titanium ana-
logues, in spite of their weaker TiϪC and TiϪN bonds.[14]
In all reactions the first deprotonation involves the most
acidic cyclopentadiene proton.
1
in THF and was characterized by H and 13C NMR spec-
troscopy as a mixture of isomers (Scheme 2). The 1H NMR
spectrum of 1 recorded at 20 °C indicated a mixture of iso-
mers a, b, c and d in a molar ratio of 8:4:1.3:1.
The 1H NMR spectra of 2 and 3 in C6D6 each show
equivalent silylamido groups with NH (δ
ϭ 1.04,
1.08 ppm), tBu (δ ϭ 1.16, 1.15 ppm), and diastereotopic
SiMe2 (δ ϭ 0.41, 0.43 ppm; δ ϭ 0.42, 0.44 ppm) resonances
close to those observed for 1. Additionally, they each show
an A2B spin system for the ring protons and one singlet
for all three equivalent NMe2 groups. The expected spectral
features were also observed in the 13C NMR spectra, which
each show one downfield shifted ring Cipso resonance (δ ϭ
127.8, 127.2 ppm) and ∆δ (δ Ctert Ϫ δ CMe) values[15]
(15.3 ppm and 15.4 ppm) consistent with the presence of
uncoordinated SiMe2ϪNHtBu groups.
Scheme 2. Isomers of C5H4(SiMe2NHtBu)2 observed by NMR
spectroscopy
In comparison with other previously reported disilyl-sub-
stituted compounds C5H4(SiR3)2,[10] in which the 1,1-iso-
mer (b; R ϭ Me, Cl) is by far the major component, the
steric hindrance caused by the bulkier substituents (R ϭ
NHtBu) means that the 1,3-isomer (a; two broad SiMe2 sin-
Single (Amidosilyl)cyclopentadienyl Compounds
When toluene or THF solutions of complex 3, prepared
glets) is the major component in the isomeric mixture of 1, at room temperature, were heated at 65 °C, complete trans-
being more favourable than the 1,1-isomer (b; one broad formation of 3 into 5 was observed after 3 h. Compound 5
SiMe2 singlet), as shown in Scheme 2. The other two iso- was the only NMR-detectable product under these con-
mers (2,4- and 2,5-), with both silyl groups bound to sp2 ditions, and it could be isolated as a brown solid and
carbon atoms, were less abundant components. All of these characterized by elemental analysis and NMR spec-
2464
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474

Neutral and Cationic [Bis(η1-amidosilyl)-η5-cyclopentadienyl]titanium and -zirconium Complexes
FULL PAPER
Scheme 3. Synthesis of cyclopentadienyl- and [(amidosilyl)- and bis(amidosilyl)cyclopentadienyl]titanium and -zirconium amide and
benzyl complexes
troscopy. A similar transformation of 2 was achieved when the SiMe2 moieties and two for the tBu groups (1H and
a toluene solution of previously isolated 2 was heated at 13C), along with two 13C resonances for their quaternary
reflux for 5 h. The deprotonation of only one of the Si- carbon atoms, are also observed for the two non-equivalent
Me2NHtBu substituents with elimination of amine NHMe2 SiMe2NtBu and SiMe2NHtBu groups. Both can be easily
occurred in a selective way to give pure 4, which was iso- distinguished by the different ∆δ values[15] observed in the
lated as a dark red oil and characterized by elemental analy- 13C NMR spectra, which are smaller [15.5 (4), 15.7 (5), 16.0
sis and NMR spectroscopy.
(6, 7) ppm] for uncoordinated NHtBu groups than for
The related dibenzyl[ansa-(cyclopentadienylsilyl)amido]- bridging NtBu groups [26.4 (4), 21.7 (5), 27.3 (6), 23.7 (7)
titanium compound 6 was produced when a toluene solu- ppm]. The amido complexes 4 and 5 also show two signals
tion of 1 and Ti(CH2Ph)4 was heated (below 70 °C) for (1H and 13C) for non-equivalent NMe2 ligands, whereas the
1
5 h. Higher temperatures could not be used, since further benzyl derivatives 6 and 7 show four H doublets and two
deprotonation would then occur, yielding small amounts of 13C resonances for diastereotopic methylene protons and
8. Under these conditions, 6 was the only reaction product carbon atoms of two non-equivalent benzyl ligands.
1
that could be isolated, as a red-brown solid, and identified
The chemical shifts observed in the H and 13C NMR
by elemental analysis and NMR spectroscopy. The third spectra for all of the ring substituents of the analogous ti-
deprotonation could not be avoided when Zr(CH2Ph)4 was tanium and zirconium complexes show very slight differ-
used, because deprotonation then occurs even at low tem- ences, whereas resonances for the additional NMe2 and
perature to give a mixture of 7 and other minor compo- CH2Ph ligands appear at a lower field for the titanium than
nents. Because the products could not be separated, 7 was for the zirconium analogues. The tert-butyl 1H and 13C res-
only characterized by NMR spectroscopy.
onances of the coordinated amido ligand are shifted down-
Coordination of the η5-cyclopentadienyl ligand increases field with respect to a free NHtBu group as a consequence
the acidity of the NHtBu amino protons [δ ϭ 1.04 (2), 1.08 of the π-donation to the metal centre, which is also consist-
(3) ppm], and this is further increased after abstraction of ent with a higher value of ∆δ.
the first amino hydrogen atom [δ ϭ 0.86 (4), 0.85 (5) ppm].
Bis(amidosilyl)cyclopentadienyl Compounds
For these reasons the reactions with the titanium reagents
could be carried out selectively to give pure amido and
benzyl compounds, whereas the tribenzyl compounds could SiMe2ϪNHtBu groups in the dibenzyl complexes 6 and 7
not be isolated at room temperature. required heating of their toluene solutions under reflux or
Complete deprotonation of the uncoordinated
1
The H NMR spectra of the silylamido compounds 4Ϫ7 at 70 °C, respectively, to give the bis(silylamido) compounds
in C6D6 each show an ABC spin system for the three ring 10, isolated as orange crystals and 11, isolated as a brown
protons while the 13C NMR spectra exhibit five resonances solid. Both complexes were identified by elemental analyses
for the inequivalent ring carbon atoms, as expected for and NMR spectroscopy. Temperatures higher than reflux
asymmetric molecules. Four resonances (1H and 13C) for in toluene were required to achieve similar deprotonation
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2465

J. Cano, P. Royo, H. Jacobsen, O. Blacque, H. Berke, E. Herdtweck
FULL PAPER
of the amido complexes 4 and 5. The reaction was moni-
1
tored by H NMR spectroscopy with a C6D6 solution of 5
in a Teflon-valved NMR tube. Complete deprotonation was
observed after 12 h of heating at 120 °C, giving the bis(silyl-
amido) compound 9 as the only reaction product. This re-
action was reversible, however, and when the solution was
cooled to room temperature, 9 reacted in solution with the
eliminated amine NHMe2 to give the starting product 5
quantitatively, and so could not be isolated in this way.
Nevertheless, pure 9 could be obtained at a preparative level
when the same thermal treatment of 5 was carried out in a
Teflon-valved Schlenk tube after addition of 1 equiv. of the
monobenzylzirconium compound 11, which reacts with am-
ine in an irreversible reaction to give 9 with elimination of
toluene (see Scheme 3). Similar deprotonation of the ti-
tanium compound 4 required heating at temperatures
higher than 160 °C to afford the bis(silylamido)titanium de-
rivative 8. This transformation was not quantitative, al-
Figure 1. ORTEP drawing of the molecular structure of the major
part of 10 in the solid state; thermal ellipsoids are drawn at the
though pure 8 could be easily recovered because the reac- 50% probability level; the hydrogen atoms are omitted for clarity
tion was not reversible on cooling to room temperature.
Compounds 8 and 9 were isolated as brown and yellow
solids, respectively, and identified by NMR spectroscopy.
Formation of the bis(silylamido) compounds is more
˚
selective, as it requires heating at temperatures above 100
Table 1. Selected interatomic distances [A] and angles [°] for com-
°C, except in the case of the benzylzirconium compound 11 plexes [M ϭ Ti (10), Zr (13)]; Cg denotes the centroid of the Cp li-
gand
(70 °C). For the same reason the formation of the amidozir-
conium compound 9 is reversible in the presence of free am-
10
13
ine.
1
The H and 13C NMR spectra of these diamido com-
MϪN1
MϪN2
1.9785(15)
1.9808(15)
2.3323(19)
2.2915(18)
2.3432(18)
2.3885(18)
2.381(2)
2.0858(17)
2.0812(17)
2.4430(19)
2.406(2)
2.437(2)
2.467(2)
pounds 8Ϫ11 in C6D6 are consistent with the presence of a
plane of symmetry, making the two silylamido groups MϪC1
equivalent. They show one singlet (1H) and two resonances
MϪC2
MϪC3
(13C) for the tBu groups and two signals (1H and 13C) for
MϪC4
the non-equivalent methyl groups of the SiMe2 moiety. Ad-
MϪC5
2.466(2)
ditionally, they show an A2B spin system for the ring pro-
tons and three signals for the ring carbon atoms, along with
one resonance for the methyl (8, 9) and methylene (10, 11)
protons and carbon atoms of the NMe2 and CH2Ph li-
gands, respectively. The ∆δ[15] values are larger than those
observed for non-coordinated SiMe2NHtBu groups in Si1ϪC1
monocyclopentadienyl (2, 3) and single silylamido (4Ϫ7)
MϪC6
2.1615(19)
M···C32
MϪC33
M···C34
MϪCg
3.019(2)
2.571(3)
3.057(3)
2.127
1.7579(17)
1.865(2)
1.7610(18)
1.870(2)
2.014
Si1ϪN1
1.7593(15)
1.859(2)
1.7491(15)
1.862(2)
Si2ϪN2
Si2ϪC3
compounds, but lower than those observed for the metal-
coordinated SiMe2-η1-NtBu groups in silylamido com-
N1ϪMϪN2
128.05(6)
104.63(7)
126.89(7)
pounds (4Ϫ7).
N1ϪMϪC6
N1ϪMϪC33
N1ϪMϪCg
N2ϪMϪC6
N2ϪMϪC33
N2ϪMϪCg
105.86(7)
101.0
Orange crystals of 10 suitable for X-ray diffraction stud-
ies were isolated after extraction into hexane and slow cool-
ing. The molecular structure of 10 is shown in Figure 1, and
selected bond lengths and angles are listed in Table 1.
106.9
99.22(7)
102.94(7)
100.8
106.4
111.0
The titanium atom is in a pseudo-tetrahedral environ- C6ϪMϪCg
C33ϪMϪCg
121.1
150.1
150.5
93.65(8)
92.94(9)
105.56(8)
105.68(8)
ment defined by the methylene carbon atom of the benzyl
CgϪC1ϪSi1
ligand, the centroid of the cyclopentadienyl ring, and the
CgϪC3ϪSi2
147.7
146.9
93.70(8)
93.92(8)
103.19(7)
103.53(8)
124.25(13)
two coordinated nitrogen atoms of the silylamido groups.
C1ϪSi1ϪN1
This is the first example of a neutral compound with a tri-
dentate cyclopentadienyl ligand tethered by two silylamido
groups.
Complex 10 shows TiϪN distances slightly longer than
those reported for related [Ti{η5-C5R4SiMe2(η1-NRЈ)}X2]
compounds with only one silylamido ligand (1.907Ϫ1.919
C3ϪSi2ϪN2
MϪN1ϪSi1
MϪN2ϪSi2
MϪC6ϪC61
MϪC33ϪC32
MϪC33ϪC34
94.56(15)
96.60(17)
2466
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474

Neutral and Cationic [Bis(η1-amidosilyl)-η5-cyclopentadienyl]titanium and -zirconium Complexes
FULL PAPER
[2h,4g,16,17,18]
[19]
˚
˚
A)
but similar to the distance (1.971A)
re- low). A similar orientation has also recently been observed
ported for R ϭ Me, RЈ ϭ tBu, X2 ϭ 2 NMe2 (see Table 2). for [Ti{1,3-[CH2(2-C4H3N)]}2(η5-C5H3)(NMe2)].[20]
This behaviour would be consistent with a weaker π-bond-
Cationic Compounds
ing contribution from the interaction between the nitrogen
pπ orbital with the appropriate vacant metal dπ orbital due
to the presence of a second amido group, which simul-
taneously increases the steric crowding and strain of the
bicyclic structure. In contrast, the SiϪN and SiϪC dis-
tances observed for 10 are very close to those observed for
all single silylamido compounds. The TiϪCg distance is in
the lower range found for singly bridged silylamido com-
pounds. There is no deviation of the disilyl-substituted
cyclopentadienyl ligand from η5-coordination, as deduced
Addition of C6D6 by vacuum transfer to a Teflon-valved
NMR tube containing a mixture of B(C6F5)3 and the
benzyl compounds 10 or 11 in a 1:1 molar ratio immedi-
ately gave the partially soluble cationic species [M(η5-
C5H3(SiMe2-η1-NtBu)2]ϩ, isolated as [B(CH2Ph)(C6F5)3]Ϫ
salts 12 and 13 as orange solids, which were identified by
NMR spectroscopy and X-ray diffraction studies.
˚
from the small differences in the TiϪC(ring) (2.347 A aver-
˚
age) and the CϪC ring distances (1.419 A average).
The amido nitrogen atom exhibits the expected planar
disposition with sp2 hybridization (the sum of bond angles
around N is 360°) known for all structurally characterized
silylamido complexes. The close CgϪCringϪSi angles
(147.3° average) account for the strain in these bicyclic com-
pounds. In addition, the CgϪTiϪN angles are similar,
whereas the TiϪNϪSi angles are significantly narrower and
the C1ϪSi1ϪN angles more open than those found in sin-
gle silylamido compounds. This is also consistent with the
differences observed in the bond lengths for a system in
Scheme 4. Formation of cationic bis(amidosilyl)cyclopentadienyl
complexes
The structures of the cations in C6D6 solutions were es-
which the SiϪN distances do not change whereas the TiϪN tablished by NMR studies. Their 1H and 13C NMR spectra
distances are longer. The N1ϪTiϪN2 angle is very open. are consistent with the Cs symmetry also observed for the
The TiϪC6 distance is in the known range for benzylti- neutral benzyl precursor compounds 10 and 11. They each
tanium compounds.
show one singlet (1H) and two resonances (13C) for the tBu
The most significant structural feature of 10 is the orien- groups, two signals (1H and 13C) for non-equivalent methyl
tation of the benzyl ligand, with its phenyl ring directed groups of the SiMe2 groups, an A2B spin system for the
toward the cyclopentadienyl ring (see DFT calculations be- ring protons and three signals for the ring carbon atoms.
˚
Table 2. Selected bond lengths [A] and bond angles [°] for related complexes
Compound
MϪN
MϪCg[a]
CgϪMϪN
MϪC
SiϪN
Ref.
[16]
Ti[(C5Me4)SiMe2(η1-NBz)]Bz2
1.919
2.045
107.4
2.131
1.731
2.158
2.106(5)
[17]
[2h]
[4g]
[19]
[19]
Ti[(C5Me4)(SiMe2{η1-NCHMePh})] Cl(CH2SiMe3)
Ti[(C5H4)SiMe2(η1-NAr)]Cl2
1.909(3)
1.914(3)
1.911(4)
1.907(4)
1.972(4)
1.924(5)
1.906(4)
1.919(9)
1.9785(15)
1.9808(15)
1.959(4)
1.946(5)
2.108(4)[b]
2.060(5)
2.064(4)
2.078
2.030
2.016
2.033
2.030
2.083
107.2
105.1
105.3
107.6
105.5
1.755(3)
1.749(4)
Ti[(C5H4)SiMe2(η1-NAr){CH2B(C6F5)2} (C6F5)]
Ti[(C5Me4)SiMe2(η1-NtBu)]Cl2
2.169(5)
Ti[(C5Me4)SiMe2(η1-NtBu)](NMe2)2
1.722(4)
[18]
Ti(C5Me5)(NMe2)3
2.132
2.014
119.4
106.9
106.4
106.7
106.6
100.2
Ti[C5H3{SiMe2(η1-NtBu)}2](CH2Ph)
2.1615(19)
2.447(3)
1.7593(15)
1.7491(15)
1.773(5)
1.788(5)
1.730(4)
[7]
[Ti{C5H3[SiMe2(η1-NtBu)]2}]ϩ [(CH2Ph)B(C6F5)3]Ϫ
Zr[(C5Me4)SiMe2(η1-NtBu)](NMe2)2
1.997(5)
2.233
[19]
[21]
[22]
Zr[(C5H3){SiMe2(η1-NtBu)}{SiMe2(NMe2)}]Cl(CH3)
Zr[(C13H8)SiMe2(η1-NtBu)](CH2SiMe3)2
2.198
2.127
101.0
2.311
1.742
1.738
2.061(2)
2.232(3)
2.248(3)
2.571(3)
[Zr{C5H3[SiMe2(η1-NtBu)]2}]ϩ [(CH2Ph)B(C6F5)3]Ϫ
2.0858(17)
2.0812(17)
101.0
100.8
1.7579(17)
1.7610(18)
[a]
[b]
Cg denotes the centroid of the cyclopentadienyl ring. The MϪN bond of the silylamido bridge.
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2467

J. Cano, P. Royo, H. Jacobsen, O. Blacque, H. Berke, E. Herdtweck
FULL PAPER
The ∆δ[15] values for the tBu 13C signals are slightly larger tral complexes (see Table 2), indicating higher electron do-
than those observed for corresponding neutral compounds. nation from the cyclopentadienyl ring.
1
A surprising effect is observed when the H and 13C NMR
The amido nitrogen atom exhibits the expected planar
spectra of the neutral benzyl precursor compounds 10Ϫ11 disposition with sp2 hybridization (the sum of bond angles
are compared with those of the corresponding cationic ti- around N is 360°) known for all reported silylamido com-
tanium and zirconium derivatives 12Ϫ13: all of the NMR plexes.
The
constrained
geometry
of
the
resonances for the cationic species are shifted upfield. It is CgϪCringϪSiϪNϪZr cyclic systems is similar to that found
also remarkable that this effect is particularly important for for 12. The ring C1 and C3 atoms, bearing the amidosilyl
the ring CH proton located between the two silyl substitu- arms, are pyramidally distorted almost to the same extent,
ents. Furthermore, it is interesting to note that the ring Cipso as shown by the sums of their bond angles (349.4° average)
resonance is the signal at highest field for the neutral com- and by the CgϪCringϪSi angles (150.3° average). The angles
pounds whereas for the cationic derivatives it appears be- at Si and N are in the range found for 12, whereas the
tween the other two ring C resonances. This unexpected CgϪZrϪN angles (100.9° average) are significantly nar-
behaviour seems to be associated with the anisotropy of the rower than those observed in the related titanium cation 12
benzyl ring exerted in the particular structure of these cat- (106.65° average). This is consistent with the increased
ionic species.
atomic radius of the Zr centre and is responsible for the
Orange crystals of 13, appropriate for a single-crystal X- slightly more constrained geometry shown by 13.
˚
˚
ray structure determination, were isolated from a C6D6
The ZrϪC33 bond [2.571(3) A] is about 0.30 A longer,
solution in the NMR tube. A perspective view of the molec- whereas the distances to the neighbouring C32 and C34
˚
ular structure of 13 is shown in Figure 2, along with the atoms are about 0.8 A longer than those observed for coval-
non-hydrogen labelling scheme, and selected bond lengths ent ZrϪC bonds, indicating an interaction of the empty hy-
and angles are listed in Table 1.
brid metal orbital with the pπ orbital of the meta-Cphenyl
atom of the borate anion.[23] The most significant feature
of this structure is that the phenyl ring is oriented upward
in the direction of the cyclopentadienyl ring, as also ob-
served for the benzyl substituent in complex 10.
Structure and Bonding of [Ti{η5-C5H3(SiMe2-η1-NtBu)2}]؉
Cations
In this section we present density functional calculations
performed on the model compound [Ti{η5-C5H3(SiH2-η1-
NH)2}]ϩ as well as on the real molecule [Ti{η5-
C5H3(SiMe2-η1-NtBu)2}]ϩ. The calculations for the model
system serve to establish characteristic patterns in chemical
bonding for bis(amidosilyl)cyclopentadienyl compounds of
Ti and Zr. As mentioned above, these compounds can
reasonably be regarded as 14-electron species, due to in-
clusion of the amido π-electrons. In the light of our pre-
vious analysis,[8c] we can expect that the bonding of [Ti{η5-
C5H3(SiR2-η1-NR)2}]ϩ cations should be governed by elec-
trostatic interaction and ligand-to-metal donation, whereas
back-bonding should play an only minor role.
Figure 2. ORTEP drawing of the molecular structure of 13 in the
solid state; thermal ellipsoids are drawn at the 10% probability level
and hydrogen atoms are omitted for clarity
The pseudo-tetrahedral geometry about the Zr atom of
13 is defined by the centroid of the cyclopentadienyl ring
and the two N-donors of the appended (dimethylsilyl)am-
ido arms, with the remaining coordination site occupied by
C(33) of the phenyl ring of the benzylborate anion. This
Bonding of the Model System [Ti{η5-C5H3(SiH2-η1-
NH)2}]؉
We first performed DFT calculations on the species with
disposition is very close to that observed for complex 12[7] aryl ligands, [Ti{η5-C5H3(SiH2-η1-NH)2}]ϩ[ηx-C6H5R]
with the expected longer ZrϪN and similar SiϪN [1.773(5) (R ϭ H, CH3, CH2BF3Ϫ; x ϭ 1, 2). For benzene, we con-
˚
˚
A; 1.788(5) A for 12] bonds. In comparison with those in sidered different ‘‘up’’ and ‘‘down’’ coordination modes, as
the related neutral [Zr{(η5-C5R4)SiMe2(η1-NRЈ)}X2] com- indicated in Scheme 5. For η1-coordination, ‘‘up’’ and
pounds containing only one silylamido ligand, the ZrϪN ‘‘down’’ coordination modes are characterized by dihedral
distances in 13 are similar,[21,22] but slightly shorter than the angles (C4ϪC1ϪMϪCg) close to 0 and 180°, respectively,
distances reported[19] for R ϭ Me, RЈ ϭ tBu, X2 ϭ 2 NMe2. C1 being the benzene carbon atom coordinating to the me-
This behaviour is consistent with a similar π-bonding con- tal centre, and C4 being the benzene carbon atom in the
tribution from the two amido ligands to the more acidic position para to C1. Similarly, for η2-coordination, dihedral
zirconium cation. The CgϪZr distance is significantly angles (X4ϪX1ϪMϪCg) close to 0 and 180° characterize
shorter than those observed in related single silylamido neu- ‘‘up’’ and ‘‘down’’ coordination modes, X1 being the mid-
2468
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474

Neutral and Cationic [Bis(η1-amidosilyl)-η5-cyclopentadienyl]titanium and -zirconium Complexes
FULL PAPER
point between the two coordinating benzene carbon atoms, ring carbon atom in the position para to the methyl sub-
and X4 being the midpoint of the two corresponding ben- stituent.
zene carbon atoms in the para position.
The relative energies for various investigated complex
geometries, together with optimized TiϪL bond lengths, are
collected in Table 3. Relative energies are determined for
groups of isomeric compounds, in which the energy of the
isomer with the largest absolute value in total bonding en-
ergy is set to zero, and energies of isomers are reported with
reference to the energy of the isomer at zero energy.
The energy differences are small, but significant. In par-
ticular, we note that η1-coordination is favoured for the cat-
ionic complexes by about 10 kJ/mol. The small energy dif-
ferences are accompanied by a fairly large change in the
˚
TiϪC bond length, of about 0.20 A, indicating the subtle
Scheme 5. The ‘‘up’’ and ‘‘down’’ coordination modes of the sol-
vated model cation [Ti{C5H3(SiH2NH)2}]ϩ
nature of this bonding interaction. The bond energies (ΒΕs)
of benzene and toluene amount to 89 kJ/mol and 100 kJ/
mol, respectively.
Finally, complexes of the borate anion C6H5CH2BF3
are characterized by the TiϪC bonds, which are about 0.20
The optimized geometry of the naked cation [Ti{η5-
C5H3(SiH2-η1-NH)2}]ϩ and its main empty frontier molec-
ular orbital are displayed in Figure 3. The HOMO of the
aryl ligand, dominates the orbital interaction through do-
Ϫ
˚
A shorter than those in the related benzene or toluene com-
2
nation to the empty dz orbital at the metal centre.
plexes. Electrostatic interactions between the anionic ligand
and the cationic metal fragment have a pronounced influ-
ence on the TiϪL bond strength. Again, η1-coordination is
favoured by 16 kJ/mol.
Complexes with the Cation [Ti{η5-C5H3(SiMe2-η1-
NtBu)2}]؉
We begin with the molecular geometry of [Ti{η5-
C5H3(SiMe2-η1-NtBu)2}{η1-C6H5CH2B(C6F5)3}],
which
has also been established by crystal structure determi-
nation.[7] By our analysis, we can expect the following fea-
tures for this compound. Firstly, the [C6H5CH2B(C6F5)3]Ϫ
borate ligand should coordinate in η1-fashion, being ori-
ented upwards toward the site of the cyclopentadienyl ring.
This structural feature is observed in the crystal structure of
this compound. Furthermore, from simple Hückel orbital
considerations, we expect that the coordinating ring carbon
atom should be located para to the boron substituent. How-
Figure 3. Optimized geometry for the model cation
[Ti{C5H3(SiH2NH)2}]ϩ and a main empty frontier MO
We recall from simple Hückel considerations that the ever, the solid-state structure reveals coordination through
HOMO of benzene consists of a set of degenerate e1g or- a meta carbon atom rather than one in the para position.
bitals, one of which is more likely to undergo η1-type do- We therefore optimized two representative geometries for
nation, whereas the other would induce a preference for η2- both meta and para coordination, as shown in Figure 4.
coordination. For toluene, the degeneracy of the e1g set is
For the meta coordination (Figure 4, a), inspection of the
lifted, due to the antibonding hyperconjugation of the optimized geometries reveals the presence of short F···H
methyl group. In this case, the HOMO is most suited to η1- distances (d ഠ 2.40 A), between the pentafluorophenyl rings
˚
coordination, having the largest orbital coefficient at the and methyl groups of the metal fragment. For para coordi-
˚
Table 3. Relative energies [kJ/mol] and metal-to-ligand bond lengths [A] for coordination of the aryl double bond for various complexes
of the type [{C5H3(SiH2NH)2}Ti]ϩ[ηx-C6H5R] (R ϭ H, CH3, CH2BF3Ϫ; x ϭ 1, 2)
Ϫ
R ϭ H
R ϭ Me
R ϭ CH2BF3
η1
η2
η1
‘‘up’’
η2
‘‘up’’
η1
η2
‘‘up’’
‘‘up’’[c]
‘‘down’’
‘‘up’’
‘‘down’’
‘‘up’’
[a]
Erel
d(TiϪC)[b]
0
2.46
6
2.54
9
2.69
8
2.66
0
2.43
12
2.67, 2.69
0
2.28
16
2.41, 2.66
[a] The most stable isomer is set to zero. [b] Bond lengths between the C6H5R ligand and the metal centre. [c] See Scheme 5 for a definition
of ‘‘up’’ and ‘‘down’’ coordination.
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2469

J. Cano, P. Royo, H. Jacobsen, O. Blacque, H. Berke, E. Herdtweck
FULL PAPER
Figure 4. Relative energies and selected bond lengths for meta (a) and para coordination (b) of [Ti{η5-C5H3(SiMe2NtBu)2}{η1-
C6H5CH2B(C6F5)3}]; comparison with X-ray analysis
nation (Figure 4, b), in contrast, no such short contacts can lending kinetic stability to the unusual η1-adduct is the en-
be found. The presence of such intramolecular hydrogen ergy of preparation of the B(C6F5)3 fragment, associated
bonds might be the key point that explains why the meta with the pyramidalization of the trigonal-planar-coordi-
geometry is energetically favoured by 30 kJ/mol. To pro- nated boron atom, which substantially increases the value
duce further evidence for our argument, we also optimized for the CϪB bond strength by the order of 80 kJ/mol.[8c]
meta and para geometries for [Ti{η5-C5H3(SiMe2-η1-
NtBu)2}(η1-C6H5CH2BF3)]. As was to be expected, since
no short F···H contacts can be achieved in this case, orbital
interactions now favour the latter coordination mode by
13 kJ/mol.
Conclusion
To conclude, we have synthesized a new type of [bis(η1-
amidosilyl)cyclopentadienyl]titanium and -zirconium com-
plexes by direct deprotonation of the corresponding bis(am-
inosilyl)cyclopentadiene with metal tetraamide and tetra-
benzyl derivatives under varying thermal conditions. The
complexes obtained in this way may be classified into three
categories: monocyclopentadienyl, ansa-(amidosilyl)cyclo-
pentadienyl and bis[ansa-(amidosilyl)cyclopentadienyl] com-
plexes. The first two categories show the structures and
chemical behaviour expected for these well-known types of
compounds. However, the bis[ansa-bis(amidosilyl)cyclopen-
tadienyl] complexes are isolobal with previously reported
tris(cyclopentadienyl) compounds and present reactivity
pathways determined by the strong electrostatic character
of the interactions, additionally facilitated by the unique
accessible metal orbital. Treatment of the benzyl complexes
with B(C6F5)3 gave the cationic species, which were charac-
terized by X-ray diffraction methods, indicating that the
counter-ion is coordinated to the metal atom through a sin-
gle interaction with the phenyl m-carbon atom of the benzyl
borate anion.
We then optimized the molecular geometry of [Ti{η5-
C5H3(SiMe2-η1-NtBu)2}(CH2C6H5)], which is displayed in
Figure 5. The geometry of the neutral benzyl complex again
exhibits ‘‘up’’ coordination, with a typical TiϪC bond
˚
length of 2.19 A. The optimized geometry is in good agree-
ment with the crystal structure of this compound. The reac-
tion energy (∆E) for the formation of the benzyl complex
from
[Ti{η5-C5H3(SiMe2-η1-NtBu)2}{η1-C6H5CH2B-
(C6F5)3}], which involves the cleavage of a CϪB bond, fol-
lowed by a reorientation of the benzyl fragment, amounts
to Ϫ15 kJ/mol. The latter compound is therefore stable for
kinetic, but not for thermodynamic, reasons. One factor
Density functional calculations developed for the cationic
complexes provide an explanation for their remarkable
structural features, particularly the ‘‘up’’ (oriented toward
the cyclopentadienyl ring) coordination of the benzyl group
in the neutral benzyltitanium complex and of the ligand
(solvent or counter-anion) bound in the cationic species.
Experimental Section
General Remarks: All experiments were carried out under argon by
use either of a Vacuum Atmospheres glove box or of standard
Schlenk techniques. Hydrocarbon solvents and THF were distilled
from Na/benzophenone and stored under argon prior to use.
Figure 5. Optimized geometry of the benzyl complex [Ti{η5-C5H3
(SiMe2NtBu)2}(CH2C6H5)]
2470
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474

Neutral and Cationic [Bis(η1-amidosilyl)-η5-cyclopentadienyl]titanium and -zirconium Complexes
FULL PAPER
1,1-Bis(chlorodimethylsilyl)cyclopentadiene,
Zr(NMe2)4,[11] Ti(CH2Ph)4,[12] Zr(CH2Ph)4
Ti(NMe2)4,[11]
6 H, NMe2), 6.12 (m, 1 H, C5H3), 6.31 (m, 1 H, C5H3), 6.50 (m, 1
and B(C6F5)3,[24] H, C5H3) ppm. 13C NMR (C6D6, 20 °C): δ ϭ 2.1 (SiMe2), 2.4
[12]
were prepared by previously reported methods. NMR spectra were (SiMe2), 2.4 (SiMe2), 3.7 (SiMe2), 34.1 (NHtBu), 34.4 (NtBu), 49.1
recorded with Varian Unity FT 300 and Varian FT 500 Unity Plus
instruments in Teflon-valved tubes at probe temperature (20 °C).
1H and 13C NMR chemical shifts were measured relative to the
residual resonances of C6D6 used as solvent, but the chemical shifts
are reported with respect to TMS. 19F NMR chemical shifts are
referenced to external CFCl3. Coupling constants are reported in
Hz. C, H and N analyses were carried out with a PerkinϪElmer
240 C analyzer.
(NMe2), 49.6 (NHtButert), 50.4 (NMe2), 60.8 (NtButert), 109.1
(C5H3ipso), 121.2 (C5H3), 122.5 (C5H3), 123.7 (C5H3), 130.7 (C5H3)
ppm. IR (Nujol): ν˜ ϭ 3385 (NϪH) cmϪ1. C21H46N4Si2Ti (458.67):
calcd. C 54.99, H 10.11, N 12.21; found C 54.67, H 9.97, N 12.04.
[Zr{η5-C5H3-1-[SiMe2(NHtBu)]-3-[SiMe2(η1-NtBu)]}(NMe2)2] (5):
A solution of Zr(NMe2)4 (3.37 g, 12.6 mmol) in toluene (70 mL)
was cooled to 0 °C, and 1 (4.09 g, 12.6 mmol) was added by syringe.
The resulting yellow solution was warmed to 65 °C for 5 h. When
the evolution of gas had stopped, the solvent was removed under
vacuum and the residue was extracted into pentane (70 mL). After
filtration and removal of the solvent, complex 5 was isolated as a
light brown solid (6.19 g, 12.3 mmol, 98%). IR (Nujol): ν˜ ϭ 3384
C5H4(SiMe2NHtBu)2 (1):
A solution of LiNHtBu (1.61 g,
20.5 mmol) in THF (40 mL) was added at 0 °C to a solution of
C5H4-1,1-(SiMe2Cl)2 (2.51 g, 10 mmol) in THF and the mixture
was stirred for 12 h. The volatiles were removed under vacuum and
the residue was extracted into hexane to give a mixture of isomers
of 1 as a yellow liquid after removal of the solvent (2.92 g,
(NϪH) cmϪ1 1H NMR (C6D6, 20 °C): δ ϭ 0.35 (s, 3 H, Si-
.
Me2NHtBu), 0.42 (s, 3 H, SiMe2NHtBu), 0.57 (s, 3 H, Si-
Me2NtBu), 0.59 (s, 3 H, SiMe2NtBu), 0.85 (s, 1 H, NHtBu), 1.11
(s, 9 H, NHtBu), 1.27 (s, 9 H, NtBu), 2.81 (s, 6 H, NMe2), 2.82 (s,
6 H, NMe2), 6.39 (m, 1 H, C5H3), 6.47 (m, 1 H, C5H3), 6.69 (m, 1
H, C5H3) ppm. 13C NMR (C6D6, 20 °C): δ ϭ 2.3 (SiMe2), 2.7
(SiMe2), 2.8 (SiMe2), 3.0 (SiMe2), 33.9 (NHtBu), 34.9 (NtBu), 44.2
(NMe2), 44.5 (NMe2), 49.6 (NHtButert), 56.6 (NtButert), 111.4
(C5H3ipso), 122.6 (C5H3), 122.9 (C5H3); 124.7 (C5H3), 126.7 (C5H3)
ppm. C21H46N4Si2Zr (502.01): calcd. C 50.24, H 9.24, N 11.16;
found C 49.73, H 9.05, N 11.45.
1
9.0 mmol, 90%). H NMR (C6D6, 20 °C) of the major component
(1a): δ ϭ Ϫ0.05 (br. s, 6 H, SiMe2), 0.21 (br. s, 6 H, SiMe2), 1.12
(s, 9 H, NHtBu), 1.17 (s, 9 H, NHtBu), 3, 57 (m, 1 H, CpH), 6.50
(m, 1 H, C5H3), 6.71 (m, 2 H, C5H3) ppm. Component 1b: δ ϭ
0.09 (s, 12 H, SiMe2), 1.15 (s, 18 H, NHtBu), 6.53 (m, 2 H, C5H4),
6.78 (m, 2 H, C5H4) ppm.
[Ti{η5-C5H3-1,3-[SiMe2(NHtBu)]2}(NMe2)3] (2):
A solution of
Ti(NMe2)4 (0.65 g, 2.9 mmol) in toluene (70 mL) was cooled to 0
°C, and 1 (0.94 g, 2.9 mmol) was added by syringe. The resulting
red solution was stirred at 50 °C for 5 h. When the evolution of
gas had stopped, the solvent was removed under vacuum and the
residue was extracted into pentane (70 mL). After filtration and
removal of the solvent, complex 2 was isolated as a deep red oil
(1.45 g, 2.8 mmol, 97%). 1H NMR (C6D6, 20 °C): δ ϭ 0.41 (s, 6
H, SiMe2), 0.43 (s, 6 H, SiMe2), 1.04 (s, 2 H, NHtBu), 1.16 (s, 18
H, NtBu), 3.13 (s, 18 H, NMe2), 6.31 (d, 2 H, C5H3), 6.60 (t, 1 H,
C5H3) ppm. 13C NMR (C6D6, 20 °C): δ ϭ 2.9 (SiMe2), 4.1 (SiMe2),
34.0 (NtBu), 49.3 (NtButert), 50.3 (NMe2), 119.9 (C5H3), 125.2
[Ti{η5-C5H3-1-[SiMe2(NHtBu)]-3-[SiMe2(η1-NtBu)]}(CH2Ph)2] (6):
A solution of Ti(CH2Ph)4 (2.14 g, 5.2 mmol) in toluene (70 mL)
was cooled to 0 °C, and 1 (1.69 gr, 5.2 mmol) was added by syringe.
The resulting yellow solution was warmed to 70 °C for 5 h. When
the evolution of gas had stopped, the solvent was removed under
vacuum and the residue was extracted into pentane (70 mL). After
filtration and removal of the solvent, complex 6 was isolated as a
1
red-brown solid (2.86 g, 5.1 mmol, 98%). H NMR (C6D6, 20 °C):
δ ϭ 0.21 (s, 3 H, SiMe2NHtBu), 0.27 (s, 3 H, SiMe2NHtBu), 0.37
(s, 3 H, SiMe2NtBu), 0.38 (s, 3 H, SiMe2NtBu), 0.72 (s, 1 H,
NHtBu), 1.06 (s, 9 H, NHtBu), 1.44 (s, 9 H, NtBu), 2.46 (d, J ϭ
10.5 Hz, 1 H, CH2Ph), 2.55 (d, J ϭ 10.5 Hz, 1 H, CH2Ph), 2.81 (d,
J ϭ 10.5 Hz, 1 H, CH2Ph), 2.97 (d, J ϭ 10.5 Hz, 1 H, CH2Ph),
5.83 (m, 1 H, C5H3), 6.14 (m, 1 H, C5H3), 6.83 (m, 1 H, C5H3),
6.87Ϫ7.20 (m, 10 H, CH2Ph) ppm. 13C NMR (C6D6, 20 °C): δ ϭ
0.6 (SiMe2), 1.5 (SiMe2), 2.4 (SiMe2), 2.9 (SiMe2), 33.8 (NHtBu),
34.2 (NtBu), 49.8 (NHtButert), 61.5 (NtButert), 79.6 (CH2Ph), 83.7
(CH2Ph), 110.2 (C5H3ipso), 122.1 (C5H3), 122.5 (C5H3), 122.8
(C5H3), 123.0 (C5H3), 125.9 (C6H5), 126.8 (C6H5), 127.4 (C6H5),
128.5 (C6H5), 128.6 (C6H5), 128.7 (C6H5), 128.9 (C6H5), 129.8
(C6H5), 132.6 (C6H5), 149.6 (C6H5ipso), 150.1 (C6H5ipso) ppm. IR
(nujol): ν˜ ϭ 3349 (NϪΗ) cmϪ1. C31H48N2Si2Ti (552.78): calcd.
C 67.36, H 8.75, N 5.07; found C 67.82, H 8.71, N 4.81.
(C5H3), 127.8 (C5H3ipso) ppm. IR (Nujol): ν˜ ϭ 3383 (NϪH) cmϪ1
.
C23H53N5Si2Ti (503.76): calcd. C 54.84, H 10.60, N 13.09; found
C 54.17, H 10.23, N 12.82.
[Zr{η5-C5H3-1,3-[SiMe2(NHtBu)]2}(NMe2)3] (3): Compound
1
(0.05 g, 0.1 mmol) was added by syringe to a solution of Zr(NMe2)4
(0.041 g, 0.1 mmol) in C6D6 (0.6 mL). Complex 3 was immediately
identified in the NMR spectrum as the unique product of the reac-
1
tion. H NMR (C6D6, 20 °C): δ ϭ 0.42 (s, 6 H, SiMe2), 0.44 (s, 6
H, SiMe2), 1.08 (s, 2 H, NHtBu), 1.15 (s, 18 H, NtBu), 2.98 (s, 18
H, NMe2), 6.50 (d, 2 H, C5H3), 6.78 (t, 1 H, C5H3) ppm. 13C NMR
(C6D6, 20 °C): δ ϭ 2.8 (SiMe2), 3.5 (SiMe2), 33.9 (NtBu), 45.6
(NMe2), 49.3 (NtButert), 120.5 (C5H3), 127.0 (C5H3), 127.2
(C5H3ipso) ppm. Pure 3 was only observed in NMR-tube experi-
ments, while solid samples were always contaminated by small
amounts of 5.
[Zr{η5-C5H3-1-[SiMe2(NHtBu)]-3-[SiMe2(η1-NtBu)]}(CH2Ph)2]
(7): A solution of Zr(CH2Ph)4 (2.93 g, 6.43 mmol) in toluene
(70 mL) was cooled to 0 °C, and 1 (2.09 g, 6.43 mmol) was added
[Ti{η5-C5H3-1-[SiMe2(NHtBu)]-3-[SiMe2(η1-NtBu)}(NMe2)2] (4):
A solution of Ti(NMe2)4 (2.00 g, 8.9 mmol) in toluene (70 mL) was
cooled to 0 °C, and 1 (2.91 g, 8.9 mmol) was added by syringe. The by syringe. The resulting yellow solution was warmed to 40 °C for
resulting red solution was heated at reflux for 5 h. When the evol-
ution of gas had stopped, the solvent was removed under vacuum
and the residue was extracted into pentane (70 mL). After filtration
and removal of the solvent, complex 4 was isolated as a deep red
5 h. The solvent was removed under vacuum and the residue was
extracted into pentane (70 mL). After filtration and removal of the
solvent, a mixture of complexes 7 and 11 was isolated. Formation
of 7 was confirmed by IR and NMR spectroscopy, although pure
oil (4.10 g, 8.8 mmol, 98%). 1H NMR (C6D6, 20 °C): δ ϭ 0.35 7 free of 11 could not be isolated. Data for 7: IR (Nujol): ν˜ ϭ
(s, 3 H, SiMe2NHtBu), 0.43 (s, 3 H, SiMe2NHtBu), 0.50 (s, 3 H,
SiMe2NtBu), 0.54 (s, 3 H, SiMe2NtBu), 0.86 (s, 1 H, NHtBu), 1.12 SiMe2NHtBu), 0.34 (s, 3 H, SiMe2NHtBu), 0.35 (s, 3 H, Si-
(s, 9 H, NHtBu), 1.28 (s, 9 H, NtBu), 2.87 (s, 6 H, NMe2), 3.05 (s, Me2NtBu), 0.40 (s, 3 H, SiMe2NtBu), 0.69 (s, 1 H, NHtBu), 1.06
3353 (NϪΗ) cmϪ1 1H NMR (C6D6, 20 °C): δ ϭ 0.33 (s, 3 H,
.
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2471

J. Cano, P. Royo, H. Jacobsen, O. Blacque, H. Berke, E. Herdtweck
FULL PAPER
(s, 9 H, NHtBu), 1.13 (s, 9 H, NtBu), 1.66 (s, 1 H, CH2Ph), 1.72
(s, 1 H, CH2Ph), 2.12 (s, 1 H, CH2Ph), 2.17 (s, 1 H, CH2Ph), 6.21
(70 mL). After filtration and removal of the solvent, complex 11
was isolated as a brown solid. (3.20 g, 6.34 mmol, 99%). H NMR
1
(m, 1 H, C5H3), 6.35 (m, 1 H, C5H3), 6.60 (m, 1 H, C5H3), (C6D6, 20 °C): δ ϭ 0.40 (s, 6 H, SiMe2), 0.42 (s, 6 H, SiMe2), 1.27
6.90Ϫ7.30 (m, 10 H, C6H5) ppm. 13C NMR (C6D6, 20 °C): δ ϭ (s, 18 H, NtBu), 2.12 (s, 2 H, CH2Ph), 6.14 (d, 2 H, C5H3), 6.53 (t,
1.3 (SiMe2), 2.0 (SiMe2), 2.4 (SiMe2), 2.5 (SiMe2), 33.7 (NHtBu), 1 H, C5H3) 6.9Ϫ7.23 (m, 5 H, C6H5) ppm. 13C NMR (C6D6, 20
33.8 (NtBu), 49.7 (NHtButert), 54.9 (CH2Ph), 57.5 (NtButert), 57.9
(CH2Ph), 109.6 (C5H3ipso), 122.2 (C5H3), 122.5 (C5H3), 124.8 57.1 (NtButert), 116.5 (C5H3ipso), 120.7 (C5H3), 126.4 (C5H3), 128.5
(C5H3), 125.6 (C5H3), 125.9 (C6H5), 126.4 (C6H5), 126.6 (C6H5), (C6H5), 129.8 (C6H5), 132.7 (C6H5), 150.6 (C6H5ipso ppm.
127.8 (C6H5), 129.3 (C6H5), 129.6 (C6H5), 129.6 (C6H5), 129.8 C24H40N2Si2Zr (503.97): calcd. C 57.20, H 8.00, N 5.56; found C
°C): δ ϭ 2.4 (SiMe2), δ ϭ 2.4 (SiMe2), 35.8 (NtBu), 55.8 (CH2Ph),
)
(C6H5), 131.9 (C6H5), 145.8 (C6H5ipso), 146.2 (C6H5ipso) ppm.
57.01, H 7.65, N 5.48.
[Ti{η5-C5H3-1,3-[SiMe2(η1-NtBu)]2}]؉ [(CH2Ph)B(C6F5)3]؊ (12): A
toluene (20 mL) solution of the monobenzyl complex 10 (0.125 g,
0.25 mmol) was treated at room temperature with B(C6F5)3
(0.126 g, 0.25 mmol), and the mixture was stirred for 30 min and
cooled to Ϫ35 °C. The solvent was filtered off from the resulting
insoluble residue, which was then dried under vacuum to give 12
(0.13 g, 60% yield) as an orange, crystalline solid. 1H NMR (C6D6,
20 °C): δ ϭ 0.19 (s, 6 H, SiMe2), 0.38 (s, 6 H, SiMe2), 1.12 (s, 18
H, NtBu), 3.49 (s, 2 H, BCH2), 5.03 (d, 2 H, C5H3), 5.86 (t, 1 H,
C5H3), 6.21Ϫ7.10 (m, 5 H, C6H5) ppm. 13C NMR (C6D6, 20 °C):
δ ϭ 0.6 (SiMe2), 1.4 (SiMe2), 34.6 (NtBu), 59.3 (NtButert), 122.1
(C5H3ipso) 126.1 (C5H3), 126.2 (C5H3), 128.3 (C6H5), 128.7 (C6H5),
132.9 (C6H5), 135.1 (C6F5), 140.1 (C6F5), 145.8 (C6F5), 150.9
(C6F5) ppm. 19F NMR (300 MHz, C6D6, 20 °C, CCl3F): δ ϭ 132.1
(m, 2 F, o-C6F5), 163.7 (m, 1 F, p-C6F5), 167.3 (m, 2 F, m-C6F5)
ppm. C42H40BF15N2Si2Ti (972.63): calcd. C 51.87, H 4.15, N 2.88;
found C 51.98, H 4.04, N 3.08.
[Ti{η5-C5H3-1,3-[SiMe2(η1-NtBu)]2}(NMe2)] (8): When a C6D6
solution of the diamido complex 4 was heated at 160 °C for 12 h
in a sealed NMR tube, a slow reaction took place with elimination
of NHMe2 to give 8, identified by its 1H and 13C NMR spectra,
together with the starting complex 4. When the reaction mixture
was cooled to room temperature, the reaction proved to be irrevers-
1
ible. H NMR (C6D6, 20 °C): δ ϭ 0.50 (s, 6 H, SiMe2), 0.52 (s, 6
H, SiMe2), 1.38 (s, 18 H, NtBu), 2.83 (s, 6 H, NMe2), 6.17 (d, 2 H,
C5H3), 6.71 (t, 1 H, C5H3) ppm. 13C NMR (C6D6, 20 °C): δ ϭ 2.7
(SiMe2), 2.8 (SiMe2), 35.2 (NtBu), 51.3 (NMe2), 59.3 (NtButert),
117.7 (C5H3ipso), 121.3 (C5H3), 131.9 (C5H3) ppm.
[Zr{η5-C5H3-1,3-[SiMe2(η1-NtBu)]2}(NMe2)] (9). Method A: When
a C6D6 solution of the diamido complex 5 was heated at 120 °C
for 12 h in a sealed NMR tube, a slow reaction took place with
elimination of NHMe2 to give 9, identified by its 1H and 13C NMR
spectra. However, when the reaction mixture was cooled to room
temperature, a reversible reaction with the free NHMe2 present in
solution once more yielded the starting complex 5. Method B: A
1:1 molar ratio of the benzyl derivative 11 (1.46 g, 3.1 mmol) and
the diamido derivative 5 (1.45 g, 3.1 mmol) in toluene was heated
at 120 °C in a Teflon-valved Schlenk tube for 12 h. The solvent was
removed under vacuum and the residue was extracted into pentane
(50 mL). After filtration and removal of the solvent, complex 9
(1.32 g, 2.9 mmol, 94%) was isolated as a yellow solid. 1H NMR
(C6D6, 20 °C): δ ϭ 0.51 (s, 6 H, SiMe2), 0.55 (s, 6 H, SiMe2), 1.31
(s, 18 H, NtBu), 2.77 (s, 6 H, NMe2), 6.34 (d, 2 H, C5H3), 6.78 (t,
1 H, C5H3) ppm. 13C NMR (C6D6, 20 °C): δ ϭ 2.9 (SiMe2), 3.0
[Zr{η5-C5H3-1,3-[SiMe2(η1-NtBu)]2}]؉ [(CH2Ph)B(C6F5)3]؊ (13): A
toluene (20 mL) solution of the monobenzyl complex 11 (0.116 g,
0.25 mmol) was treated with B(C6F5)3 (0.126 g, 0.25 mmol) at room
temperature and the mixture was stirred for 30 min and cooled to
Ϫ35 °C. The solvent was filtered off from the resulting insoluble
residue, which was then dried under vacuum to give 13 (0.21 g, 83%
yield) as a dark brown oil. After recrystallization, a suitable orange
1
monocrystal of 13 was separated for X-ray diffraction studies. H
NMR (C6D6, 20 °C): δ ϭ 0.17 (s, 6 H, SiMe2), 0.36 (s, 6 H, SiMe2),
0.99 (s, 18 H, NtBu), 3.42 (s, 2 H, BCH2), 5.20 (d, 2 H, C5H3),
6.01 (t, 1 H, C5H3), 6.10 (m, 1 H, p-C6H5), 6.34 (m, 2 H, m-C6H5),
6.87 (m, 2 H, o-C6H5) ppm. 13C NMR (C6D6, 20 °C): δ ϭ 1.5
(SiMe2), 1.6 (SiMe2), 34.8 (NtBu), 58.6 (NtButert), 121.4 (C5H3),
123.5 (C5H3ipso), 127.4 (C5H3), 127.9 (C6H5), 128.1 (C6H5), 128.3
(C6H5), 136.5 (C6F5), 138.3 (C6F5), 147.8 (C6F5), 149.8 (C6F5),
162.0 (C6H5ipso) ppm. 19F NMR (300 MHz, C6D6, 20 °C, CCl3F):
δ ϭ 132.1 (m, 2 F, o-C6F5), 163.6 (m, 1 F, p-C6F5), 167.2 (m, 2 F,
m-C6F5) ppm. C42H40BF15N2Si2Zr (1015.97): calcd. C 49.65, H
3.97, N 2.76; found C 50.51, H 3.71, N 3.35.
(SiMe2), 35.6 (NtBu), 45.1 (NMe2), 55.8 (NtButert), 117.7 (C5H3ipso
)
119.8 (C5H3), 133.2 (C5H3) ppm. C19H39N3Si2Zr (456.93): calcd. C
49.94, H 8.60, N 9.20; found C 50.28, H 8.31, N 8.86.
[Ti{η5-C5H3-1,3-[SiMe2(η1-NtBu)]2}(CH2Ph)] (10): A solution of
Ti(CH2Ph)4 (5.34 g, 12.9 mmol) in toluene (70 mL) was cooled to
0 °C, and 1 (4.21 g, 12.9 mmol) was added by syringe. The resulting
deep red solution was warmed at 70 °C for 5 h. The solvent was
removed under vacuum and the residue was extracted into pentane
(70 mL). After filtration and removal of the solvent, complex 10
was isolated as a waxy, red solid, which was recrystallized from
hexane to give orange crystals (5.88 g, 12.7 mmol, 98%). 1H NMR
(C6D6, 20 °C): δ ϭ 0.39 (s, 6 H, SiMe2), 0.40 (s, 6 H, SiMe2), 1.42
(s, 18 H, NtBu), 2.59 (s, 2 H, CH2Ph), 6.12 (d, 2 H, C5H3), 6.39 (t,
1 H, C5H3), 6.89 (m, 1 H, C6H5), 6.95 (m, 2 H, C6H5), 7.22 (m, 2
H, C6H5) ppm. 13C NMR (C6D6, 20 °C): δ ϭ 2.1 (SiMe2), 2.2
(SiMe2), 35.6 (NtBu), 59.3 (NtButert), 69.6 (CH2Ph), 117.7
(C5H3ipso), 121.5 (C5H3), 126.3 (C5H3), 128.5 (C6H5), 130.3 (C6H5),
132.6 (C6H5), 152.4 (C6H5ipso) ppm. C24H40N2Si2Ti (460.63): calcd.
C 62.58, H 8.75, N 6.08; found C 62.68, H 8.68, N 5.76.
Crystal Structure Determination for Compound 10: Crystals of 10
were grown from a hexane solution. An orange fragment in perflu-
orinated ether was selected and transferred to a glass capillary,
which was mounted in a cold N2 stream. Preliminary examination
and data collection were carried out with a Nonius KappaCCD
device at the window of a rotating anode X-ray generator with
˚
use of graphite-monochromated Mo-Kα radiation (λ ϭ 0.71073 A),
controlled by the Collect software package.[25] Collected images
were processed by use of Denzo. The unit cell parameters were
obtained by full-matrix, least-squares refinements of 4895 reflec-
tions.[26] A total number of 41477 reflections were integrated. After
merging, 4739 reflections remained, and these were used for all
further calculations. Absorption effects were corrected during the
scaling procedure. The structure was solved by direct methods[27]
and refined by standard difference Fourier techniques.[28] All non-
hydrogen atoms of the asymmetric unit were refined with aniso-
[Zr{η5-C5H3-1,3-[SiMe2(η1-NtBu)]2}(CH2Ph)] (11): A solution of
Zr(CH2Ph)4 (2.93 g, 6.43 mmol) in toluene (70 mL) was cooled to
0 °C, and 1 (2.09 g, 6.43 mmol) was added by syringe. The resulting
yellow solution was heated at reflux for 5 h. The solvent was re-
moved under vacuum and the residue was extracted into pentane
2472
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474

Neutral and Cationic [Bis(η1-amidosilyl)-η5-cyclopentadienyl]titanium and -zirconium Complexes
FULL PAPER
tropic thermal displacement parameters. All hydrogen atoms were
found in the difference Fourier map and refined freely with individ-
ual isotropic thermal displacement parameters, except for those lo-
cated at a disordered tert-butyl group, which were placed in calcu-
lated positions (riding model). Full-matrix, least-squares refine-
ments were carried out by minimization of Σw(Fo ϪFc ) by use of
the SHELXL-97 weighting scheme and stopped at a maximum
shift/err Ͻ 0.001.[28,29,30] One tert-butyl group in 10 appears to be
disordered over two positions (73:27).
Table 4. Crystallographic data for complexes 10 and 13
10
13
Empirical formula
Formula mass
Colour/shape
C24H40N2Si2Ti
460.63
orange/fragment
C42H40BF15N2Si2Zr
1015.97
orange/plate
2
2 2
Crystal size [mm]
Crystal system
Space group
0.36 ϫ 0.23 ϫ 0.20 0.56 ϫ 0.30 ϫ 0.10
monoclinic
P21/c (No. 14)
10.6962(1)
16.0412(2)
15.3339(2)
90
99.6928(7)
90
2593.43(5)
4
triclinic
¯
P1 (No. 2)
˚
a [A]
11.7349(1)
13.4253(1)
14.4024(1)
77.4374(3)
88.2620(4)
86.9316(5)
2211.12(3)
2
Crystal Structure Determination for Compound 13: Crystals of 13
were grown from a C6D6 solution. An orange plate in perfluori-
nated ether was selected and transferred to a glass capillary. Pre-
liminary examination and data collection were carried out with a
Nonius KappaCCD device at the window of a rotating anode X-
ray generator by use of graphite-monochromated Mo-Kα radiation
˚
b [A]
˚
c [A]
α [°]
β [°]
γ [°]
3
˚
V [A ]
Z
[26]
˚
(λ ϭ 0.71073 A), controlled by the Collect software package.
Collected images were processed by use of Denzo. The unit cell
parameters were obtained by full-matrix, least-squares refinements
of 8080 reflections.[27] A total number of 44162 reflections were
integrated. After merging, 8099 reflections remained and were used
for all further calculations. Absorption effects were corrected dur-
ing the scaling procedure. The structure was solved by direct meth-
ods[27] and refined by standard difference Fourier techniques.[28] All
non-hydrogen atoms of the asymmetric unit were refined with an-
isotropic thermal displacement parameters. All hydrogen atoms
were found in the difference Fourier map and refined freely with
individual isotropic thermal displacement parameters. Full-matrix,
least-squares refinements were carried out by minimization of
T [K]
123
1.180
0.435
293
1.526
0.399
ρcalcd. [g cmϪ3
]
µ [mmϪ1
]
F000
992
1028
θ range [°]
Data collected (h,k,l)
No. of reflns. collected 41477
1.85Ϫ25.34
Ϯ12, Ϯ19, Ϯ18
1.45Ϫ25.38
Ϯ14, Ϯ16, Ϯ17
44162
No. of independent
reflns./Rint
No. of obsd. reflns.
(I Ͼ 2σ(I)]
4739 (all)/0.033
4193 (obsd.)
8099 (all)/0.033
7177 (obsd.)
No. of params. refined 414
729
R1 (obsd./all)
wR2 (obsd./all)
GOF (obsd./all)
0.0333/0.0394
0.0855/0.0895
1.045/1.045
0.0306/0.0370
0.0752/0.0788
1.040/1.040
ϩ0.30/Ϫ0.27
2
2 2
Σw(Fo ϪFc ) by use of the SHELXL-97 weighting scheme and
stopped at a maximum shift/err Ͻ 0.001. A correction for extinc-
tion effects was applied.[28Ϫ30]
Ϫ3
˚
max./min. ∆ρ [e·A
]
ϩ0.32/Ϫ0.27
CCDC-200139 (10) and -200140 (13) contain the supplementary
crystallographic data for this paper. These data can be obtained
free of charge at www.ccdc.cam.ac.uk/conts/retrieving.html [or
from the Cambridge Crystallographic Data Centre, 12 Union
Road, Cambridge CB2 1EZ, UK; Fax: (internat.) ϩ 44-1223/336-
033; E-mail: deposit@ccdc.cam.ac.uk].
[2] [2a]
P. J. Shapiro, E. E. Bunel, W. E. Piers, J. E. Bercaw, Synlett
[2b]
1990, 2, 74Ϫ84.
P. J. Shapiro, E. Bunel, W. P. Schäfer, J. E.
Bercaw, Organometallics 1990, 9, 867Ϫ869. [2c] P. J. Shapiro, W.
E. Cotter, W. P. Schäfer, J. A. Labinger, J. E. Bercaw, J. Am.
Chem. Soc. 1994, 116, 4623Ϫ4640.
1990, 123, 1649Ϫ1651.
Teuben, Organometallics 1993, 12, 1936Ϫ1945.
Herrmann, M. J. A. Morawietz, J. Organomet. Chem. 1994,
482, 169Ϫ181.
mez-Sal, A. Manzanero, P. Royo, Organometallics 1996, 15,
[2d]
J. Okuda, Chem. Ber.
[2e]
A. K. Hughes, A. Meetsma, J. H.
Computational Details: Gradient corrected density functional cal-
culations were carried out, with corrections for exchange and corre-
lation according to Becke[31] and Perdew,[32] respectively (BP86).
Geometries were optimized by use of the TURBOMOLE[33] pro-
gram system within the framework of the RI-J approximation.[34]
For the model systems, a triple-ζ valence basis plus polarization
TZVP[35] was employed. For the geometry optimization of the real
molecules, the main group elements were described by a split-val-
ence basis set[36] with one set of polarization functions for the non-
H atoms, and Ti was treated with the TZVP basis. For the determi-
nation of bond energies, final energies were obtained in a single-
point calculation using the TZVP basis set for all atoms.
[2f]
W. A .
[2g]
´ ´
S. Ciruelos, T. Cuenca, R. Gomez, P. Go-
[2h]
´ ´ ´
´
R. Gomez, P. Gomez-Sal, A. Martın, A. Nu-
5577Ϫ5585.
n˜ez, P. A. del Real, P. Royo, J. Organomet. Chem. 1998, 564,
[2i]
93Ϫ100.
A. L. Mcknight, R. M. Waymouth, Chem. Rev.
1998, 98, 2587Ϫ2598. [2j] J. Klosin, W. J. Kruper, Jr., P. N. Nick-
ias, G. R. Roof, P. DeWaele, K. A. Abboud, Organometallics
[2k]
2001, 20, 2663Ϫ2665.
K. Kunz, G. Erker, S. Döring, R.
Fröhlich, G. Kehr, J. Am. Chem. Soc. 2001, 123, 6181Ϫ6182.
[3] [3a]
M. Bochmann, J. Chem. Soc., Dalton Trans. 1996,
[3b]
255Ϫ270.
C. P. Casey, D. W. Carpenetti II, H. Sakurai,
[3c]
Organometallics 2001, 20, 4262Ϫ4265.
C. G. Brandow, A.
Acknowledgments
The authors acknowledge the MCyT (project MAT2001-1309) for
financial support and the EC (project COST-D12/0016/98). J. C.
acknowledges CAM for a fellowship.
Mendiratta, J. E. Bercaw, Organometallics 2001, 20, 42534261.
[4] [4a]
Y.-X. Chen, P.-F. Fu, C. L. Stern, T. J. Marks, Organomet-
[4b]
allics 1997, 16, 5958Ϫ5963.
B. E. Bosch, G. Erker, R.
[4c]
Fröhlich, O. Meyer, Organometallics 1997, 16, 5449Ϫ5456.
Y.-X. Chen, T. J. Marks, Organometallics 1997, 16, 3649Ϫ3657.
[4d] A. Bertuleit, C. Fritze, G. Erker, R. Fröhlich, Organometall-
[1] [1a]
[4e]
`
G. Lanza, I. L. Fragala, T. J.
P. C. M öhring, N. J. Coville, J. Organomet. Chem. 1994,
ics 1997, 16, 2891Ϫ2899.
[1b]
[4f]
479, 1Ϫ29.
H. H. Brintzinger, D. Fischer, R. Mülhaupt, B.
Marks, J. Am. Chem. Soc. 1998, 120, 8257Ϫ8258.
F. Amor,
Rieger, R. M. Waymouth, Angew. Chem. Int. Ed. Engl. 1995,
A. Butt, K. E. Du Plooy, T. B. Spaniol, J. Okuda, Organomet-
34, 1143Ϫ1170. [1c] G. G. Hlatky, Coord. Chem. Rev. 1999, 181,
allics 1998, 17, 5836Ϫ5849. [4g] R. Gomez, P. Gomez-Sal, P. A.
´
´
[1d]
[4h]
243Ϫ296.
H. G. Alt, A. Köppl, Chem. Rev. 2000, 100,
L. Resconi, L. Cavallo, A. Fait, F. Piemontesi,
del Real, P. Royo, J. Organomet. Chem. 1999, 588, 22Ϫ27.
R. J. Keaton, K. C. Jayaratne, J. C. Fettinger, L. R. Sita, J. Am.
Chem. Soc. 2000, 122, 12909Ϫ12910.
[1e]
1205Ϫ1222.
[4i]
Chem. Rev. 2000, 100, 1253Ϫ1345.
J.-F. Carpentier, V. P.
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 2473

J. Cano, P. Royo, H. Jacobsen, O. Blacque, H. Berke, E. Herdtweck
FULL PAPER
[21]
Maryin, J. Luci, R. F. Jordan, J. Am. Chem. Soc. 2001, 123,
898Ϫ909.
Y. Mu, W. E. Piers, L. R. MacQuarrie, M. Zaworotko, Can. J.
Chem. 1996, 74, 1696Ϫ1703.
J. Okuda, F. J. Schattenmann, S. Wocadlo, W. Massa, Or-
[22]
[5] [5a]
J. C. Stevens, F. J. Timmers, D. R. Wilson, G. F. Schmidt,
ganometallics 1995, 14, 789Ϫ795.
P. N. Nickias, R. K. Rosen, G. W. Knight, S. Y. Lai, Eur. Patent
Appl. EP 1991, 416815 (Dow Chemical Co.) (Chem. Abstr.
1991, 115, 93163). [5b] J. M. Canich, Eur. Patent Appl. EP 1991,
420436 (Exxon Chemical Co) (Chem. Abstr. 1991, 115, 184145).
[23] [23a]
K. Shelly, D. C. Finster, Y. J. Lee, W. R. Scheidt, C. A.
[23b]
Reed, J. Am. Chem. Soc. 1985, 107, 5955Ϫ5959.
M. La-
guna, M. D. Villacampa, M. Contel, J. Garrido, Inorg. Chem.
1998, 37, 133Ϫ135. [23c] A. S. Batsanov, S. P. Crabtree, J. A. K.
Howard, C. W. Lehmann, M. Kilner, J. Organomet. Chem.
1998, 550, 59Ϫ61.
A. G. Massey, A. J. Park, J. Organomet. Chem. 1964, 2,
245Ϫ250.
COLLECT, Data Collection Software for Nonius Kappa CCD
Devices, Hooft, R.Nonius B. V., Delft, The Netherlands, 1998.
Z. Otwinowski, W. Minor, Processing of X-ray Diffraction Data
Collected in Oscillation Mode, Methods in Enzymology, vol.
276: Macromolecular Crystallography, part A (Eds.: C. W.
Carter, Jr., R. M. Sweet), Academic Press, New York, 1997,
p. 307Ϫ326.
A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M.
C. Burla, G. Polidori, M. Camalli, J. Appl. Crystallogr. 1994,
27, 435Ϫ436.
[6]
V. C. Gibson, J. Chem. Soc., Dalton Trans. 1994, 1607Ϫ1618.
[7a] J. Cano, P. Royo, A. Tiripicchio, M. A. Pellinghelli, M. Lan-
[7]
[7b]
franchi, Angew. Chem. Int. Ed. 2001, 40, 2495Ϫ2497.
P.
[24]
[25]
[26]
Royo, J. Cano, M. A. Flores, B. Pen˜a, EP 2001, 1225179 A1;
US 2002, 0115560 A1.
[8] [8a]
B. E. Bursten, L. F. Rhodes, J. Strittmatter, J. Am. Chem.
[8b]
Soc. 1989, 111, 2758Ϫ2766.
Fröhlich, Organometallics 1997, 16, 531Ϫ536.
H. Berke, S. Doring, G. Kehr, G. Erker, R. Fröhlich, O. Meyer,
Organometallics 1999, 18, 1724Ϫ1735.
Brakemeyer, R. Fröhlich, G. Erker, Organometallics 2001, 20,
T. Brackemeyer, G. Erker, R.
[8c]
H. Jacobsen,
[8d]
N. Kliegrewe, T.
1952Ϫ1955.
[27]
[9]
J. M. Rozell, P. R. Jones, Organometallics 1985, 4, 2206Ϫ2210.
[10] [10a]
Y. A. Ustynyuk, A. V. Kisin, J. M. Pribytkova, A. A. Zar-
kin, N. D. Antonova, J. Organomet. Chem. 1972, 42, 47Ϫ63.
[28]
[29]
G. M. Sheldrick, SHELXL-97, Universität Göttingen, Ger-
many 1998.
International Tables for Crystallography, vol. C, Tables 6.1.1.4
(pp. 500Ϫ502), 4.2.6.8 (pp. 219Ϫ222), and 4.2.4.2 (pp.
193Ϫ199) (Ed.: A. J. C. Wilson), Kluwer Academic Publishers,
Dordrecht, 1992.
[10b]
P. Jutzi, Chem. Rev. 1986, 86, 983Ϫ996.
[11]
[12]
[13]
[14]
G. M. Diamond, R. F. Jordan, J. L. Petersen, J. Am. Chem.
Soc. 1996, 118, 8024Ϫ8033.
U. Zucchini, E. Albizzati, U. Giannini, J. Organomet. Chem.
1971, 26, 357Ϫ372.
Y. Mu, W. E. Piers, L. R. MacQuarrie, M. Zaworotko, J.
Young, Organometallics 1996, 15, 2720Ϫ2726.
D. M. P. Mingos, Essential Trends in Inorganic Chemistry, Ox-
ford Univ. Press, Oxford, 1998, p. 368.
[30]
[31]
[32]
A. L. Spek, Acta Crystallogr., Sect. A 1990, 46, C34.
A. D. Becke, Phys. Rev. 1988, A38, 3098Ϫ3100.
J. P. Perdew, Phys. Rev. 1986, B33, 8822Ϫ8824.
[33] [33a] R. Ahlrichs, M. Bär, M. Häser, H. Horn, C. Kölmel, Chem.
[15] [15a]
W. A. Nugent, B. L. Haymore, Coord. Chem. Rev. 1980,
[33b]
Phys. Lett. 1989, 162, 165Ϫ169.
O. Treutler, R. Ahlrichs,
[15b]
31, 123Ϫ175.
D. E. Wigley, Prog. Inorg. Chem. 1994, 42,
[33c]
J. Chem. Phys. 1995, 102, 346Ϫ354.
R. Ahlrichs, M. von
239Ϫ482. [15c] M. J. Humphries, M. L. H. Green, M. A. Leech,
V. C. Gibson, M. Jolly, D. N. Williams, M. R. J. Elsegood, W.
Clegg, J. Chem. Soc., Dalton Trans. 2000, 4044Ϫ4051.
J. Okuda, T. Eberle, T. P. Spaniol, Chem. Ber./Recueil 1997,
130, 209Ϫ215.
Arnim, Methods and Techniques in Computational Chemistry:
METECC-95 (Eds.: E. Clementi, G. Corongiu), STEF, Cagli-
ari, 1995, chapter 13.[33d]R. Ahlrichs, M. von Arnim, J. Com-
put. Chem. 1998, 19, 1746Ϫ1757.
[16]
[17]
[18]
[19]
[20]
[34] [34a]
K. Eichkorn, O. Treutler, H. Öhm, M. Häser, R. Ahlrichs,
[34b]
J. Okuda, S. Verch, R. Sturmer, T. P. Spaniol, J. Organomet.
Chem. 2000, 605, 55Ϫ67.
Chem. Phys. Lett. 1995, 242, 652Ϫ660.
Weigand, O. Treutler, R. Ahlrichs, Theor. Chem. Acc. 1997,
K. Eichkorn, F.
´
A. Martın, M. Mena, C. Yelamos, R. Serrano, P. R. Raithby,
97, 119Ϫ124.
[35]
[36]
J. Organomet. Chem. 1994, 467, 79Ϫ84.
D. W. Carpenetti, L. Kloppenburg, J. T. Kupec, J. L. Petersen,
Organometallics 1996, 15, 1572Ϫ1581.
W. S. Seo, Y. J. Cho, S. C. Yoon, J. T. Park, Y. Park, J. Or-
ganomet. Chem. 2001, 640, 79Ϫ84.
A. Schäfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100,
5829Ϫ5835.
A. Schäfer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97,
2571Ϫ2577.
Received December 20, 2002
2474
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2463Ϫ2474