Alkyl-η2-alkene niobocene and tantalocene complexes with the allyldimethylsilyl-η-5-cyclopentadienyl ligand: Synthesis, NMR studies and DFT calculations

Article abstract of DOI:10.1039/b406747a

Group 5 metal complexes [M(η⁵-C₅H ₅){η⁵-C₅H₄SiMe ₂(CH₂-η²-CH = CH₂)}X] (M = Nb, X = Me, CH₂Ph, CH₂SiMe₃; M = Ta, X = Me, CH₂Ph) and [Ta(η⁵-C₅Me ₅){η⁵-C₅H₄SiMe ₂(CH₂-η²-CH=CH₂] (X = Cl, Me, CH₂Ph, CH₂SiMe₃) containing a chelating alkene ligand tethered to a cyclopentadienyl ring have been synthesized in high yields by reduction with Na/Hg (X = Cl) and alkylation with reductive elimination (X = alkyl) of the corresponding metal(IV) dichlorides [M(η ⁵-Cp){η⁵-C₅H₄SiMe ₂(CH₂CH=CH₂)}Cl₂] (Cp = C ₅H₅, M = Nb, Ta, Cp = C₅Me₅, M = Ta). These chloro- and alkylalkene coordinated complexes react with CO and isocyanides [CNtBu, CN(2,6-Me₂C₆H₃)] to give the ligand-substituted metal(III) compounds [M(η⁵-Cp){η ⁵-C₅H₄SiMe₂(CH₂CH=CH ₂)}XL] (X = Cl, Me, CH₂Ph, CH₂SiMe ₃). Reaction of the chloro-alkene tantalum complex with LiNHtBu results in formation of the imido hydride derivative [Ta(η ⁵-C₅Me₅){η⁵-C₅H ₄SiMe₂(CH₂CH=CH₂)}H(NtBu)]. NMR studies for all of the new compounds and DFT calculations for the alkenecoordinated metal complexes are compared with those known for related group 4 metal cations.

Full text of DOI:10.1039/b406747a

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Alkyl–2-alkene niobocene and tantalocene complexes with the
allyldimethylsilyl–5-cyclopentadienyl ligand: synthesis, NMR
studies and DFT calculations†
Pilar Nicolás,a Pascual Royo,*a Mikhail V. Galakhov,b Olivier Blacque,c Heiko Jacobsenc and
Heinz Berkec
a
Departamento de Química Inorgánica, Facultad de Química, Universidad de Alcalá,
Campus Universitario, E-28871 Alcalá de Henares, Spain. E-mail: pascual.royo@uah.es;
Fax: +34.918854683
Centro de Espectroscopía de RMN, Universidad de Alcalá, E-28871 Alcalá de Henares, Spain
b
c
Anorganisch-Chemisches Institut der Universität Zürich, Winterthurerstrasse 190,
CH-8057 Zürich, Switzerland
Received 5th May 2004, Accepted 21st July 2004
First published as an Advance Article on the web 10th August 2004
Group 5 metal complexes [M(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}X] (M = Nb, X = Me, CH2Ph, CH2SiMe3;
M = Ta, X = Me, CH2Ph) and [Ta(5-C5Me5){5-C5H4SiMe2(CH2-2-CHCH2)}X] (X = Cl, Me, CH2Ph, CH2SiMe3)
containing a chelating alkene ligand tethered to a cyclopentadienyl ring have been synthesized in high yields by reduction
with Na/Hg (X = Cl) and alkylation with reductive elimination (X = alkyl) of the corresponding metal(IV) dichlorides
[M(5-Cp){5-C5H4SiMe2(CH2CHCH2)}Cl2] (Cp = C5H5, M = Nb, Ta, Cp = C5Me5, M = Ta). These chloro– and alkyl–
alkene coordinated complexes react with CO and isocyanides [CNtBu, CN(2,6-Me2C6H3)] to give the ligand-substituted
metal(III) compounds [M(5-Cp){5-C5H4SiMe2(CH2CHCH2)}XL] (X = Cl, Me, CH2Ph, CH2SiMe3). Reaction of the
chloro–alkene tantalum complex with LiNHtBu results in formation of the imido hydride derivative [Ta(5-C5Me5){5-
C5H4SiMe2(CH2CHCH2)}H(NtBu)]. NMR studies for all of the new compounds and DFT calculations for the alkene-
coordinated metal complexes are compared with those known for related group 4 metal cations.
bond followed by -hydrogen elimination. This leads to a dynamic
Introduction
behaviour, which can be studied by variable-temperature NMR
spectroscopy. The structure of various hydrido–alkene niobocene
and tantalocene complexes and the kinetics of their insertion reac-
tions have been extensively studied5–7 and the behaviour of ansa-
metallocenes in comparison with their non-bridged systems has
been reported recently.8
Alkene coordination to a group 4 d0 transition metal alkyl cation
has been proposed1 to precede the migratory insertion for a Ziegler–
Natta polymerization process.2 The resulting 16e− alkyl–alkene
species has a very low thermal stability, because the metal–alkene
interaction is usually only due to the poor -donor capacity of the
alkene. More stable alkyl–alkene chelates have still to be studied at
temperatures lower than −40 °C to prevent their decomposition or
reactions with the solvent.
We reported3,4 previously the synthesis of group 4 metal dialkyls
containing an allyldimethylsilyl–cyclopentadienyl ligand, which
after treatment with a Lewis acid B(C6F5)3, coordinate the olefinic
system of the allyl moiety to give alkyl cations with a chelated 5-
Cp–2-alkene ligand. Exchange between two enantiomeric struc-
tures was observed and studied by DNMR spectroscopy3 between
193 and 253 K.
The corresponding neutral d2 group 5 metal(III) 18e− compounds
are structurally similar, but it was expected that they would be more
stable, because the metal–alkene bond is now composed of -donor
and -acceptor interactions. These compounds may therefore be
easily accessible models to compare some aspects of their structural
behaviour with those found for d0 metal complexes.
Many hydrido–alkene group 5 metallocenes have been isolated5
by reaction of the alkene with the metal trihydrides [MCp2H3].
Similar compounds were also isolated by alkylation of the
metallocene dichlorides [MCp2Cl2] with -hydrogen containing
alkyl groups and simultaneous reduction to give [MCp2R], which
are transformed into the corresponding hydrido–alkene deriva-
tives by -hydrogen elimination.5b,d,6 It has been observed that all
these hydrido–alkene complexes show easy hydrogen exchange
processes, which are due to alkene insertion into the metal–hydride
Related chloro–alkene group 5 metal(III) compounds have
occasionally been singularly isolated by reduction of dichloro-
metallocenes in the presence of olefin,9 but relatively few alkyl–
alkene compounds have been reported. Ligand promoted olefin
insertion into the niobium–hydrido bond of permethylniobocene
[NbCp*2(olefin)H] complexes to give the ligand-trapped alkyl
compounds has been reported only for small ligands (CO,
CNMe),6d whereas the related niobocene complexes have been
shown to afford either ligand-trapped alkyl compounds,5a,6b
or the olefin substitution hydrido products.5d Severe restric-
tions were also found for the permethyltantalocene complex5c
whereas the related tantalocene [TaCp2(C2H4)H] compound
reacted with excess ethylene to give the trapped ethyl–ethylene
complex at elevated temperatures5d and similar CO and CNR
promoted insertions have been shown to afford a group of
ligand-trapped tantalocene alkyl [TaCp2RL] complexes.10
Attempts to induce similar olefin insertion and to trap the
resulting alkyl compounds failed when ansa-metallocene olefin
hydrides were used.6e
In this study we report the synthesis of the chloro- [Ta(5-
C5Me5){5-C5H4SiMe2(CH2-2-CHCH2)}Cl] and alkyl–alkene
coordinated niobium and tantalum derivatives [M(5-C5R5){5-
C5H4SiMe2(CH2-2-CHCH2)}R′] (M = Nb, Ta; R = H, Me;
R′ = Me, CH2Ph, CH2SiMe3) obtained from the dichloro metal(IV)
compounds [M(5-C5R5){5-C5H4SiMe2(CH2CHCH2)}Cl2] (M =
Nb, Ta; R = H, Me) prepared previously.11 The NMR behaviour of
these compounds is compared to that known for related group 4
metal cations. DFT calculations were carried out on model com-
plexes to support the structures of these complexes and the mecha-
nisms for their dynamic behavior.
† Electronic supplementary information (ESI) available: Full synthetic
and structural data for compounds 13a–d and 14a–g. 1H NMR data of the
dimethylsilylallyl moiety for complexes 4–12 in C6D6 at 20 °C. DFT Calcu-
lations. See http://www.rsc.org/suppdata/dt/b4/b406747a/
T h i s j o u r n a l i s
©
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
D a l t o n T r a n s . , 2 0 0 4 , 2 9 4 3 – 2 9 5 1
2 9 4 3

View Article Online
reduction of the chloro–alkyl compounds and lead to the chloro–
alkene complex 4, which is then further alkylated. Consequently,
these mixtures were transformed into the pure alkyl compounds
11 and 12 by addition of excess alkylating agent, or even better,
by direct alkylation of 3 with 3 equiv. of MgRCl. The tantalum
complexes 10–12 may also be prepared treating the isolated chloro
complex 4 with 1 equiv. of the corresponding alkylating agent.
Compound 11 was isolated as a green solid in 75% yield whereas
12 resulted as a brown solid in 65% yield after purification and both
were characterized by elemental analysis and NMR spectroscopy.
All compounds 4–12 were very soluble in alkanes and stable under
inert atmosphere into a dry-box.
All these features observed in the synthesis may be explained
assuming that formation of the final alkyl–alkene metal compounds
may take place through the initial formation of the chloro–alkyl
derivative, which may then be followed by its reduction to the
chloro metal(III) compound. These chloro metal(III) complexes
allow facile coordination of the tethered alkene, particularly when
bulkier alkyl (CH2Ph and CH2SiMe3) and ring (C5Me5) substituents
are present, and finally the chloro–alkene derivatives are easily
alkylated. However formation of intermediate dialkyl complexes
cannot be excluded.13
Results and discussion
Synthesis of chloro– and alkyl–alkene complexes
The dichloro metal(IV) complexes [M(5-Cp){5-C5H4SiMe2-
(CH2CHCH2)}Cl2] (Cp = C5H5, M = Nb 1, Ta 2, Cp = C5Me5,
M = Ta 3) were prepared previously by the reaction of the mono-
cyclopentadienyl compounds with the corresponding lithium cyclo-
pentadienides and simultaneous reduction with Na/Hg.11 Related
reduction of toluene solutions of 1 and 2 with Na/Hg gave brown
solids, presumably chlorine bridged12 species, which did not contain
the coordinated olefin and were not further studied. However, in an
analogous reaction, reduction of the chloro tantalum complex 3 bear-
ing the bulkier and more donating permethylated cyclopentadienyl
ligand takes place with coordination of the alkene system to give
[Ta(5-C5Me5){5-C5H4SiMe2(CH2-2-CHCH2)}Cl] 4 isolated
in 65% yield as a yellow–green solid characterized by elemental
analysis and NMR spectroscopy (see Scheme 1). Complex 4 is
soluble in pentane and reacts with chlorinated solvents, being stable
at room temperature for months under inert atmosphere.
Reactivity
All of the chloro-2-alkene 4 and alkyl-2-alkene complexes 5–12
described above react with CO and isocyanides CNtBu and CN(2,6-
Me2C6H3) to give the olefin substitution products as a new group of
the very well known14 alkyl metal(III) complexes [M(5-C5R5){5-
C5H4SiMe2(CH2–CHCH2)}XL] (M = Nb, R = H, X = CH2Ph,
L = CO 13a, CNtBu 13b; M = Ta, R = Me, X = Cl, L = CO
14a, CNtBu 14b, CN(2,6-Me2C6H3) 14c; X = Me, L = CO 14d;
X = CH2Ph, L = CO 14e, CN(2,6-Me2C6H3) 14f; X = CH2SiMe3,
L = CN(2,6-Me2C6H3) 14g) as illustrated in Scheme 2. Reactions of
complexes 10–11 with the bulkier permethylated ring required heat-
ing at 90 °C for 24 h whereas complexes 5–9 with the unsubstituted
ring took place in 16 h at room temperature. The same tantalum
complexes 14a–14c and related chloro niobium(III) derivatives
(M = Nb, L = CO 13c, CNtBu 13d) can also be obtained by reduc-
tion of the corresponding metal(IV) dichloro compounds with Na/Hg
in the presence of the ligand. These compounds were characterized
by elemental analysis and IR and 1H, 13C NMR spectroscopies. Full
synthetic and structural data for compounds 13a–d and 14a–g are
given in the ESI.†
Scheme 1
A more convenient method to prepare this type of alkyl–alkene
metal compounds was based on the alkylation of the dichloro com-
plexes 1–3, which, in the presence of excess alkylating agent induced
simultaneous reductive elimination of alkane and immediate coordi-
nation of the alkene tethered to the cyclopentadienyl ring. As shown
in Scheme 1, addition of 1 equiv. of MgRCl (R = Me, CH2Ph) or
LiCH2SiMe3 to toluene solutions of complexes 1–3 apparently did
not allow isolation of the monoalkyl species with the free non-
coordinated allyl group, nor could these species be observed spectro-
scopically. This reaction always resulted in formation of inseparable
mixtures of products. This behaviour could indicate that coordina-
tion of the tethered alkene moiety has a significantly higher thermo-
dynamic stability. When the same reaction was carried out using 2
equiv. of the alkylating agents, pure alkyl–alkene metal complexes
[M(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}R] (M = Nb, R =
Me 5, CH2Ph 6, CH2SiMe3 7; M = Ta, R = Me 8, CH2Ph 9) and
[Ta(5-C5Me5){5-C5H4SiMe2(CH2-2-CHCH2)}Me] 10 could be
obtained. Compounds 5–10 were isolated as brown (5–9) and green
(10) solids in 65–75% yields after purification. All of them were
characterized by NMR spectroscopy and elemental analyses. How-
ever, reactions of 3 with 2 equiv. of Mg(CH2Ph)Cl and LiCH2SiMe3
led to the formation of mixtures containing variable proportions of
the chloro–alkene 4 and the corresponding alkyl–alkene complexes
[Ta(5-C5Me5){5-C5H4SiMe2(CH2-2-CHCH2)}R] (R = CH2Ph
11, CH2SiMe3 12), indicating that these reactions occur through the
Scheme 2
Ligand promoted migration of the alkyl group to the alkene
moiety was not observed for any of these reactions, probably due
to the kinetic stability of the low-valent 18e− metal alkyl–alkene
compounds. As expected, neither insertion of CO nor of CNR into
the metal–alkyl bond was observed.
An interesting behaviour was observed when the reaction of the
chloro tantalum complex 4 with 1 equiv. of LiNHtBu was moni-
tored by 1H NMR spectroscopy in a Teflon valved NMR tube. As
shown in Scheme 3, formation of the intermediate amido–alkene
derivative was not observed, because rapid transfer of a hydrogen
atom to the metal with simultaneous dissociation of the alkene
moiety took place to give the imido hydride complex [Ta(5-
C5Me5){5-C5H4SiMe2(CH2CHCH2)}H(NtBu)] 15. The H and
1
13C NMR spectra of complex 15 are consistent with those expected
2 9 4 4
D a l t o n T r a n s . , 2 0 0 4 , 2 9 4 3 – 2 9 5 1

View Article Online
for an asymmetric molecule, showing typical resonances for the
non-coordinated allyl moiety, two singlets (1H and 13C{1H}) for
Me–Si groups and four (1H) and five (13C) signals for the silylated
cyclopentadienyl ring. The hydride ligand is significantly shifted
low field being observed as a singlet at  5.95 (see Experimental
section). Although this is the region where the bridging Ta(-H)
hydride signals are observed for monocyclopentadienyl complexes
containing bulky imido ligands,15 it was reported that this is also
the region where the hydride signal of terminal Nb–H16 and Ta–H17
bonds are observed for this type of dicyclopentadienyl imido com-
plexes. The difference between the 13C quaternary and methyl
carbon resonances of the tertbutylimido group is 30.8 ppm suggest-
ing a linear disposition of the Nb–N–tBu ligand in complex 15.
-CHCH2 and Si–CH2 multiplets which appear at  < 2.5 and  1–4
respectively of a ABCDE spin system. The very large shielding of
the vinyl resonances is consistent with the change of carbon atoms
hybridization from sp2 in 13, 14 to sp3 in 4–12 metal complexes.
Similar changes of chemical shifts ( < 55 ppm) were observed
in the 13C NMR spectra (see Experimental section). This behaviour
is consistent with the metallacyclopropane character reported5d,6e,18
for this type of compounds. Moreover, the direct C1–H2 and C2–H3
(1J ≈ 143–150 Hz) spin coupling constants found for the 13C signals
of the terminal and internal carbon atoms are also consistent with
their sp3 hybridization, being smaller than the value reported19 for
cyclopropane (161 Hz) due to the influence of the metal, assuming
the positive sign of these spin–spin coupling constants.20
In contrast, coordination of the vinyl moiety to the cationic group
4 metal complexes3,4 produce similar high-field shifts of the terminal
CH1H2 olefinic protons whereas the internal CH3 proton is
remarkably shifted low-field to  ~7.30 ppm, demonstrating the
high polarization of the olefinic system when it is coordinated to
a d0 metal center.
The dichloro niobium(IV) and tantalum(IV) 1–3 and the dialkyl
group
4
metal complexes [M(5-C5R5){5-C5H4SiMe2(CH2–
Scheme 3
CHCH2)}R′2] used as precursors for the preparation of the olefin
coordinated compounds are symmetric molecules for which both
right and left equatorial positions are equivalent. Therefore, treat-
ment with either, a reducing agent (Na/Hg for 4), an alkylating agent
(5–12) or B(C6F5)3 (cationic group 4 metal compounds) may elimi-
nate any of the two symmetric chloro or alkyl ligands. After coordi-
nation of the terminal alkene moiety all the resulting compounds are
characterized by two chiral principles, namely the enantioface of the
coordinated vinyl moiety and the metal atom. They should therefore
lead to the formation of any of the endo or exo diastereomers with
their pairs of enantiomers represented in Scheme 5.
NMR spectroscopic studies
Despite the fact that crystals of good quality suitable for X-ray
diffraction studies could not be obtained of any of these com-
pounds, all of the chloro–alkene (4), alkyl–alkene (5–12) and
chloro- and alkyl-ligand (13, 14) niobocene and tantalocene com-
plexes were fully characterized by NMR spectroscopy in solution.
All of them are asymmetric molecules with a chiral metal center
and their NMR spectra show the following common features. All
of the NMR spectra exhibit two singlets for the Me2Si group and
four multiplets (1H) and five signals (13C) for the ring system of the
(5-C5H4SiMe2(allyl)) ligand. They also show one signal (1H, 13C)
for the (5-C5H5) (5–9, 13) and one singlet (1H) and two signals
(13C) for the (5-C5Me5) ligands (4, 10–12, 14). In addition the
expected singlets are observed for the Me–M (5, 8, 10, 14a), Me3Si
(7, 12, 14d), tBu (13b and 13d) and 2,6-Me2–C6H3 (14c, 14d, 14g)
groups. The typical C2v local symmetry multiplets for C6H5 and 2,6-
Me2–C6H3 and theAB spin systems for the diastereotopic methylene
protons of CH2Ph and CH2SiMe3 substituents were also observed in
the 1H NMR spectra.
NMR studies related to the allylic Si–CH2–CHCH2 frag-
ment. The most significant structural data are associated to the
NMR behaviour observed for the Si–CH2–CHCH2 system. The
1H NMR spectra in C6D6 for all complexes containing the free non-
coordinated allylic moiety (Scheme 4, A) exhibit one multiplet at
 ~5.6–5.9 (internal CH3), two multiplets at  ~4.9 (Jcis ≈ 10 Hz,
Jtrans ≈ 16 Hz, Jgem ≈ 4 Hz) (terminal CH1H2) and one doublet at
 ~1.5–1.8 (J ≈ 8 Hz) (SiCH4H5). This is consistent with the three
13C signals observed at  ~134–136 (internal CH),  ~113–114
(terminal CH2) and  ~25–26 (SiCH2). The chemical shifts and
coupling constants of these signals are only very slightly affected
by the remaining substituents being almost the same for group 4
metal(IV)3,4 and for group 5 metal(III) complexes 13 and 14.
Scheme 5
Anew remarkable difference between the group 4 (d0) and group 5
(d2) metal complexes was observed when their NMR spectroscopic
behaviour was studied at variable temperature.Acommon feature to
all of the benzyl cationic group 4 metal derivatives is that one single
set of resonances is observed at 193 K. Taking into account that
values of kinetic barriers for alkene dissociation should be similar
to those found for other related zirconium d0 species 1c,e,3 this set of
resonances has to be assigned to the averaged spectrum in the fast
exchange regime occuring between the two exo/endo diastereomers
represented in Scheme 5. An additional reversible exchange can be
detected in the NMR spectra. In fact, formation of related methyl
cationic species at 193 K always led to spectra with mixtures of
exo/endo diastereomers, for which, however, the complexity of the
NMR spectra observed between 193 and 253 K prevented detailed
studies of their dynamic behaviour.4
Coordination of the terminal vinyl moiety to the metal center in
the chloro- and alkyl–alkene metal compounds 4–12 (Scheme 4, B)
is clearly demonstrated by the high-field shift of the typical vinyl
In contrast, the H NMR spectra of most of the d2 niobium(III)
1
and tantalum(III) 4–12 complexes show the presence of two sets of
signals corresponding to the endo/exo diastereoisomers which are
not temperature dependent and do not show spin exchange between
193 and 343 K. This behaviour demonstrates that both endo/exo
isomers are simultaneously formed and their interconversion by
alkene dissociation/association processes require higher activation
energies. This feature is also significantly different from that found
for related hydrido–alkene complexes reported previously, which
Scheme 4
D a l t o n T r a n s . , 2 0 0 4 , 2 9 4 3 – 2 9 5 1
2 9 4 5

View Article Online
Table 1 Relative energies (kcal mol−1) for different conformations of
complexes [M(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}CH2Ph] with
M = Sc, Zr+, Hf+, Nb and Ta
show isomerism due to the restricted rotation of the alkene, which
is not only hindered by their high metallacyclopropane character5b,d
but also prevented by formation of a rigid cyclic system with the
cyclopentadienyl-pendant alkene. In addition a high barrier of acti-
vation21 prevents the insertion of the alkene into the metal–alkyl
bond.
Conformation
M = Sc
Zr+
Hf+
Nb (6)
Ta (9)
I(Bn)-exo
0.0
−0.1
0.0
5.4
0.0
5.8
0.0
1.2
0.0
1.0
Two sets of signals of the same ABCDE spin system were
II(Bn)-endo
III(Bn)-open-exo
1-CH2Ph
1
clearly observed in the H and 13C NMR spectra of the tantalum
a
—
20.1
4.9
14.9
6.6
22.1
15.2
21.8
18.6
chloro–alkene 4 and methyl–alkene 10 complexes (3:2 and 3:1
molar ratio, respectively), for which we propose the exo and endo
structural conformations represented in Fig. 1, as metallacyclo-
propane compounds.
3-CH2Ph
0.7
IV(Bn)-open-endo
3-CH2Ph
−1.1
3.6
5.5
14.9
18.5
a No local minimum could be found.
transition metal derivatives [M(5-C5H5){5-C5H4SiMe2(CH2-2-
CHCH2)}R] for M = Sc, Zr+, Hf+, Nb and Ta, and R = Me and
CH2Ph. The calculations presented here utilize fairly large basis
sets, but do not include solvation effects, as well as influences of
counterions. Before we discuss some of the structural data in more
detail, we will briefly comment on the bonding interactions between
a CC double bond and a transition metal center. The classical
Dewar–Chatt–Duncanson22,23 model of the LnM–olefin interaction
involves -donation from the  HOMO of the olefin to an empty
metal orbital, as well as back donation of electrons from filled
d-orbitals, usually the HOMO, into the olefin’s empty * LUMO.
For d0 systems -donation is the only possible orbital interaction,
whereas for d2-systems backdonation may additionally exist.
Besides orbital interaction, there are also electrostatic forces pres-
ent, which become of major importance when charged species are
involved. The experimentally studied alkyl–alkene niobocene and
tantalocene d2 complexes with coordination of the terminal olefin to
the metal center showed NMR spectra containing two diastereomers
with endo and exo coordination of the olefinic moiety in variable
proportions. In all these compounds we have two chiral principles:
the metal center and the enantiotopic face of the olefin and therefore
two enantiomeric pairs of diastereomers could be involved (endo
and exo). We first studied both exo and endo conformations for the
five selected group 3, 4 and 5 transition metal complexes [M(5-
C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}CH2Ph].
The calculated conformations I(Sc,Bn)-exo and II(Sc,Bn)-endo
of the scandium complex (Fig. 2) are very close in energy with an
energy difference E of about 0.1 kcal mol−1 (Table 1). In both
cases, the olefinic unit is relatively far away from the metal center.
The pendant olefin group in I(Sc,Bn)-exo is coordinated to the
Sc in an unsymmetrical fashion, primarily through the terminal
carbon atom C1 (Sc–C1 2.856 Å, Sc–C2 3.252 Å). In II(Sc,Bn)-
endo, the Sc–C separations are still larger and the main part of the
coordination consists of the interaction with the terminal carbon
atom C1 (Sc–C1 2.903 Å, Sc–C2 3.327 Å). The chelating alkene is
tipped significantly from the equatorial plane containing the metal
center and perpendicular to that defined by the metal center and the
centroids of the two Cp rings, in such a way that the dihedral angle
C3(Me)–Sc–C1–C2 is 56.8°.
Fig. 1 Labelling of the exo- and endo-isomers.
Assignments of the signals observed in the 1H spectrum of com-
plex 10 to the exo and endo stereoisomers was accomplished by
NOE spectroscopy. Selective excitation (PFG WFG NOESY1D
pulse sequence) of the Me–Ta resonance at  0.14 for the major
stereoisomer (see Experimental section) afforded an enhance-
ment in the two ring proton resonances at  4.48 and 5.11 of the
(5-C5H4SiMe2CH2CHCH2) ligand and in one of the two terminal
CH2 protons at  1.25. Similar excitation of the Me–Ta resonance
at  0.33 for the minor component produced the enhancement in the
resonances at  5.48 for only one of the ring protons, at  2.58 for
one of the CH2–Si protons and an almost insignificant enhancement
in the resonance at  1.04 for one of the terminal CH2 protons.
In agreement with these results we suggest that the endo isomer
represented in Fig. 1 corresponds to the minor species whereas the
major component is the exo isomer. Following this assignment all
the remaining signals observed in the 1H NMR spectra of 10 can be
unequivocally assigned (see Experimental section).
The metallacyclopropane system is characterized by the same
values of the vicinal proton–proton coupling constant (3JH1–H3 and
3JH2–H3), which mainly depend on the dihedral angles between the
corresponding planes, whereas the negative values of the geminal
2JH1–H2 coupling constants and their chemical shifts are controlled
by the electron-donor character of the (5-C5R5) and R′ ligands. The
short distances between the Me–Ta group and only one proton of
the terminal CH2 group observed for 10-exo in the NOESY spectra
suggest that the C1, C2, Me and Ta atoms are not exactly coplanar.
Taking into account the large (more than two times) difference
between the vicinal 3JH3–H4 coupling constants observed for exo and
endo isomers of 10, we suggest to assign the exo diastereomer to
the minor component of solutions of complex 4 (3JH3–H4 = 6.5 Hz)
whereas the endo diastereomer would correspond to the major com-
ponent (3JH3–H4 = 3.0 Hz). The methyl complexes 5 and 8 containing
the unsubstituted cyclopentadienyl ligand, which are more closely
related to complex 10 were however found as almost just one single
component with a very small amount of a minor one which could
only be detected by their 5-C5H5 singlets, whereas signals due to
their vinyl moiety could not be identified. Even smaller, almost
undetectable amounts of the minor component were observed for
complexes 6, 7, 9, 11 and 12. It has been reported5 previously that
Turning to the coordination geometries of the other d0 complexes
with group 4 cationic Zr and Hf centers, we observe that the olefinic
unit is now bound closer to the metal center, indicating a stronger
bondinginteraction(Fig. 3).Also, themetalboundCCbondisnow
slightly more elongated. This is a first indication of the enhanced
2
the geminal JH1–H2 coupling constants are larger for exo- than for
endo-isomers of related hydrido-alkene complexes. The analysis of
the spectral data observed for complexes 5–9 and 11, 12 allowed us
to make only tentative assignments of the exo/endo configurations
for these compounds. However, strong general support is given by
the following theoretical analysis.
Theoretical calculations. In order to obtain further informations
about the energetic preferences for the endo/exo conformations we
carried out density functional calculations (DFT) on different con-
formations of the d0 and d2 silyl-bridged alkyl–alkene coordinated
Fig. 2 Optimized geometries I(Sc,Bn)-exo and II(Sc,Bn)-endo of
the scandium complex [Sc(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}-
CH2Ph].
2 9 4 6
D a l t o n T r a n s . , 2 0 0 4 , 2 9 4 3 – 2 9 5 1

View Article Online
stabilizing influence of electrostatic effects on this particular bond-
ing interaction. Nevertheless, the almost symmetrical coordination
of the olefin observed for the zirconium compound II(Zr,Bn)-endo
(Zr–C1 2.796 Å, Zr–C2 2.865 Å) is disfavored by 5.4 kcal mol−1
over the I(Zr,Bn)-exo conformation and its unsymmetrical bond-
ing mode (Zr–C1 2.608 Å, Zr–C2 3.125 Å). This suggests that the
electronic stabilization between the olefin and the zirconium center
is less important than the overall steric influences.
(ii) rotation of the cyclopentadienyl ring, (iii) rotation of the pendant
group around the Si–C and C–C bonds, (iv) recoordination of the
alkene through the opposite enantioface. Thus further calculations
were carried out on different conformations of the pendant olefin.
To obtain reference energies for the olefin face exchange, we
optimized two different conformations III and IV for each complex
with the olefin not bound to the metal center. The fully optimized
geometries III and IV are interconverted by simple rotation of
the pendant group around the C(Cp)–Si bond starting from the
previous stable conformations I and II. Then, we assumed that
the olefin enantioface seen by the metal center is unchanged during
the processes I-exo → III-exo and II-endo → IV-endo, and that the
olefin face exchange occurs during III-exo → IV-endo by rotation
of the CH2–CHCH2 fragment around the Si–C bond (Scheme 6).
Fig. 3 Optimized geometries I(Zr,Bn)-exo and II(Zr,Bn)-endo of the
zirconium complex [Zr(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}-
CH2Ph]+.
For Nb and Ta, where backbonding is possible, we find in both
conformations C1 and C2 very close to the metal center with M–C
bond distances in the range of 2.297–2.431 Å and a substantially
elongated CC bond between 1.425 Å and 1.444 Å to form a sort of
metallacyclopropane system (Fig. 4). Compared to the relative high
energy differences between the exo and endo geometries of 5–6 kcal
mol−1 calculated for the group 4 transition metal complexes, the exo
conformations of the Nb and Ta complexes are only slightly favored
over the endo ones by only 1.2 and 1.0 kcal mol−1.
Scheme 6
The coordination geometry for the open complexes III(Bn)-exo
suggests that the benzyl ligand is bound to the metal center via
three carbon atoms (Figs. 6–8). The separations lie in the ranges
2.338–2.863, 2.358–2.636, 2.337–2.617, 2.330–2.601 and 2.324–
2.594 Å for the Sc, Zr+, Hf+, Nb, and Ta complexes, respectively.
Similar separations are also found for IV(Bn)-endo complexes in
the ranges 2.341–2.800 (Sc), 2.352–2.633 (Zr+), 2.332–2.616 (Hf+),
2.322–2.585 (Nb), and 2.309–2.574 (Ta).
Fig. 4 Optimized geometries I(Nb,Bn)-exo and II(Nb,Bn)-endo of the
niobium complex [Nb(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}-
CH2Ph].
Fig. 5 shows the energy levels of the frontier molecular orbitals
for the three complexes I(Sc,Bn)-, I(Zr,Bn)-, and I(Nb,Bn)-exo.
The three well known frontier metal orbitals 1a1, 1b2 and 2a1 of C2v
MCp2 complexes are hybrid metal orbitals interacting with high
p atomic orbital character of the benzyl group to form a  bond
(HOMO for I(Sc,Zr)), and with the  orbital of the olefin as the
dative bond (LUMO for I(Sc,Zr) and HOMO for I(Nb)). The back-
bonding from the d2 niobium center is established by the HOMO of
I(Nb) with the * orbital of the olefin.
Fig. 6 Optimized geometries III(Sc,Bn)-exo (3) and IV(Sc,Bn)-endo
(3) of the scandium complex [Sc(5-C5H5){5-C5H4SiMe2(CH2-2-
CHCH2)}CH2Ph].
The mechanism for the interconversion of the exo and endo dia-
stereomers was expected to involve (i) dissociation of the olefin,
Fig. 7 Optimized geometries III(Zr,Bn)-exo (1) and III(Zr,Bn)-exo
(3) of the zirconium complex [Zr(5-C5H5){5-C5H4SiMe2(CH2-2-
CHCH2)}CH2Ph]+.
The second highest filled molecular orbital of the niobium
complex III(Nb,Bn)-exo depicted in Fig. 9 clearly shows
the 3-coordination mode of the benzyl unit. Analyzing the
reaction energies I → III, we obtain values of 0.7, 4.9, 6.6, 15.2 and
Fig. 5 Energy levels of frontier molecular orbitals for the exo confor-
mations of the three complexes I(Sc–Bn), I(Zr,Bn) and I(Nb,Bn).
D a l t o n T r a n s . , 2 0 0 4 , 2 9 4 3 – 2 9 5 1
2 9 4 7

View Article Online
Table 2 Relative energies (kcal mol−1) for different conformations of com-
plexes [M(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}CH3] with M = Sc,
Zr+, Hf+, Nb and Ta
Conformation
MSc
Zr+
Hf+
Nb (5)
Ta (8)
I(Me)-exo
II(Me)-endo
III(Me)-open exo
IV(Me)-open endo
0.0
0.5
4.6
4.1
0.0
0.0
0.0
−0.4
22.7
21.6
0.0
−0.6
22.6
22.1
0.2
16.7
15.6
−0.1
17.6
16.7
or H2CPh groups. Consequently, 3 binding seems to be more appro-
priate for the estimation of the energetic barriers.
Fig. 8 Optimized geometries III(Nb,Bn)-exo (1) and III(Nb,Bn)-
exo (3) of the niobium complex [Nb(5-C5H5){5-C5H4SiMe2(CH2-2-
CHCH2)}CH2Ph]+.
When single sets of resonances were observed at 193 K for the
cationic group 4 metal derivatives low exo/endo interconversion
energies are expected and the calculated low energetic barriers of
4.9 and 6.6 kcal mol−1 for [M(5-C5H5){5-C5H4SiMe2(CH2-2-
CHCH2)}CH2Ph] with M = Zr+ and Hf+, respectively, support this
picture. The exo–endo interconversions are thus expected to be fast
on the NMR time scale, even at very low temperatures, and should
therefore show averaged signals in a wide temperature range. Con-
cerning the group 5 metal complexes a decoordination of the olefin
would involve breakage of the metallacyclopropane with strong
backbonding interactions. Furthermore the vacancy of the olefin
cannot fully be compensated with a 3-benzallylic rearrangement.
Consequently, higher energetic barriers are observed for the Nb and
Ta complexes (15–19 kcal mol−1) despite relatively small thermo-
dynamic differences between the exo and endo diastereomers. In
addition the calculated differences in energy between exo and endo
are actually too small (less than 1.2 kcal mol−1 for Bz compounds)
to predict a preference for any conformer.
18.6 kcal mol−1 for Sc, Zr+, Hf+, Nb and Ta, respectively (Fig. 10).
Comparing the values for the d0 complexes, we see that the coor-
dination of the olefin in I(Sc,Bn)-exo is really weak and not really
preferred over the 3-coordination of the benzyl, and that the elec-
trostatic contribution, which stabilizes the Zr+-complex is estimated
to about 5 kcal mol−1. For the neutral group 5 metal complexes,
where electrostatic interactions are of minor importance, we have
a larger stabilization of 15.2 and 18.6 kcal mol−1, mainly due to
backbonding into the * orbital.
The influence of the 3-coordination mode of the benzyl ligand
has been studied comparing the previous results with DFT calcula-
tions for the methyl complexes [M(5-C5H5){5-C5H4SiMe2(CH2-
2-CHCH2)}CH3] (M = Sc, Zr+, Hf+, Nb and Ta) (Table 2).
Except for the scandium complex, for which the energetic barrier
between exo and endo remains relatively low (4.6 kcal mol−1),
we observe higher relative energies between the coordinated and
noncoordinated olefin states. In fact, a stabilizing influence of the
3-coordinated benzyl group can not be found in comparison with
the methyl substituted system. Rather we find a preference for the
already seen agostic interactions between the metal centers and
hydrogens from H2CSiMe3 or H2CPh. Based on the very small
thermodynamic energy differences between the exo and endo
conformers (less than 0.6 kcal mol−1) together with the relatively
high kinetic barriers (16–23 kcal mol−1), we therefore expect for all
group 4 and 5 metal complexes the presence of two sets of signals in
the 1H NMR spectra corresponding to the endo/exo diastereomers,
irrespective of the averaged resonances for the benzallylic group 4
Fig. 9 MOLDEN plot of the second highest occupied molecular orbital
of the optimized geometry III(Nb,Bn)-exo (3) of the niobium complex
[Nb(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}CH2Ph].
1
(d0) metal complexes. The H NMR spectra of the related methyl
substituted cationic species showed at 193 K the expected mixtures
of exo–endo diastereomers, but unfortunately the complexity of the
spectra observed between 193 and 253 K prevented an assignment
to the respective structures.
Conclusions
The work presented herein outlines the synthesis of a new type
of chloro– and alkyl–dicyclopentadienyl niobium and tantalum
d2 complexes [M(5-Cp){5-C5H4SiMe2(CH2-2-CHCH2)}X]
(M = Nb, Ta; Cp = C5H5, X = Me, CH2Ph, CH2SiMe3 and M = Ta,
Cp = C5Me5, X = Cl, Me, CH2Ph, CH2SiMe3) containing the
alkene moiety of an allyldimethylsilyl group tethered to one of
the cyclopentadienyl rings. The success of the synthesis of all
of these alkyl compounds is based on the facile reduction of the
starting dichloro metal(IV) complexes [M(5-Cp){5-C5H4SiMe2-
(CH2CHCH2)}Cl2] (Cp = C5H5, M = Nb 1, Ta 2, Cp = C5Me5,
M = Ta 3) when they were treated with different alkylating agents.
Reduction with Na/Hg was required to isolate the chloro tantalum
derivative.
Fig. 10 Relative energies (kcal mol−1) calculated for different conforma-
tionsofcomplexes[M(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}CH2Ph]
with M = Sc, Zr+, Hf+, Nb and Ta.
We calculated also a different conformation of the non-olefin
coordinated complex having additionally an 1-coordinated benzyl
ligand. The optimized geometries led to relatively high energetic
barriers, depending on whether we take into account this conforma-
tion (about 15 to 20 kcal mol−1 for Hf and Zr) or the 3-benzallylic
intermediates (about 5 to 7 kcal mol−1). From our point of view the
1 conformations are too “artificial”, because in solution such mole-
cules would be stabilized by interactions with surrounding solvent
molecules. In the gas phase optimized 1-benzallylic geometries,
the transition metals try to compensate the vacancy left by the
olefin by agostic interactions with hydrogens from the H2CSiMe3
1H NMR studies demonstrate that two endo/exo diastereoisomers
are simultaneously formed for all of these compounds and their inter-
2 9 4 8
D a l t o n T r a n s . , 2 0 0 4 , 2 9 4 3 – 2 9 5 1

View Article Online
conversion by alkene dissociation/association processes does not
take place, as their H NMR spectra are temperature independent
the solvent was removed under vacuum from the resulting brown
solution and the residue was extracted into hexane (30 cm3). The
solution was concentrated to 10 cm3 and cooled at −40 °C to give 5
(0.60 g, 1.80 mmol, 70%) and 6 (0.61 g, 1.45 mmol, 60%) as brown
solids which were dried under vacuum.
(5): (Found: C, 56.93; H 6.73%. C16H23NbSi requires C, 57.14;
H 6.89%); H (C6D6) 0.10, 0.18 (2s, 2 × 3H, SiMe2), 0.27 (s, 3H,
NbMe) 1.14 (dd, 2J = 14.6 Hz, 3J = 8.2 Hz, 1H, SiCH2), 1.34
1
and do not show spin exchange between 193 and 343 K. This
behaviour is in agreement with the high energetic barriers required
for breaking the metallacyclopropane system with strong back-
bonding interactions through a 3-benzallylic rearrangement, as
evaluated by DFT theoretical calculations for Nb and Ta benzyl
complexes (15–19 kcal mol−1), in spite of the relatively small
thermodynamic differences (less than 1.2 kcal mol−1) between the
exo and endo diastereomers. These barriers are much higher than
those found for related Zr+ and Hf+ cationic derivatives (4.9 and
6.6 kcal mol−1) for which rapid spin exchange between exo/endo
diastereomers was observed at 193 K.
Neither ligand promoted migration of the alkyl group to the
alkene moiety nor insertion of CO and CNR into the metal–alkyl
bond was observed. Addition of these ligands always results in
displacement of the coordinated alkene with formation of neutral
[M(5-Cp){5-C5H4SiMe2(CH2–CHCH2)}XL] niobium(III) and
tantalum(III) compounds. However formation of the imido hydride
complex [Ta(5-C5Me5){5-C5H4SiMe2(CH2CHCH2)}H(NtBu)]
was observed when the corresponding chloro derivative was treated
with LiNHtBu.
2
3
2
(dd, J = 5.4 Hz, J = 11.8 Hz, 1H, CH2), 1.50 (dd, J = 5.4 Hz,
3J = 9.9 Hz, 1H, CH2), 2.15 (dd, J = 14.6 Hz, J = 7.5 Hz, 1H,
SiCH2), 2.33 (m, 1H, CH), 4.44 (s, 5H, C5H5) and 3.81, 4.61, 4.77,
5.21 (4m, 4 × 1H, C5H4); C (C6D6) −3.5, 1.6 (SiMe2), 1.3 (NbMe),
28.2 (SiCH2), 41.5 (CH2), 48.0 (CH), 91.7, 98.2, 100.9, 119.6
(C5H4), 98.6 (C5H5) and 106.0 (C5H4 ipso).
2
3
(6): (Found: C, 64.38; H, 6.81%. C22H27NbSi requires C,
64.07; H 6.60%); H (C6D6) 0.08, 0.10 (2s, 2 × 3H, SiMe2), 1.16
2
3
2
(dd, J = 14.0 Hz, J = 7.6 Hz, 1H, SiCH2), 1.30 (dd, J = 4.8 Hz,
3J = 12.0 Hz, 1H, CH2), 1.41 (dd, 2J = 4.8 Hz, 3J = 11.3 Hz, 1H,
CH2), 2.08 (dd, 2J = 14.0 Hz, 3J = 6.0 Hz, 1H, SiCH2), 2.27, 2.63
(2d, 2J = 9.2 Hz 2 × 1H, NbCH2Ph), 2.40 (m, 1H, CH), 4.39 (s,
5H, C5H5), 3.88, 4.53, 4.61, 5.62 (4m, 4 × 1H, C5H4) and 6.99–7.29
(m, 5H, NbCH2Ph); C (C6D6) −3.4, 1.3 (SiMe2), 28.1 (SiCH2), 38.1
(CH2), 42.2 (CH), 52.1 (NbCH2Ph), 92.6, 100.9, 101.5, 117.8
(C5H4), 98.8 (C5H5), 107.0 (C5H4 ipso), 121.5, 126.2, 129.2 and
129.5 (NbCH2Ph)
Experimental
General remarks
[Nb(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}(CH2SiMe3)]
7. A solution of Li(CH2SiMe3) (0.24 g, 5.2 mmol) in hexane
(20 cm3) was added to a solution of [Nb(5-C5H5){5-C5H4SiMe2
(CH2CHCH2)}Cl2] (1) (1.00 g, 2.55 mmol) in toluene (30 cm3)
at −78 °C. The mixture was stirred for 16 h at room temperature.
LiCl was separated by filtration and the solvent was removed
under vacuum from the resulting brown solution. The residue was
extracted into hexane (50 cm3) and the solution was evaporated to
obtain 7 (0.67 g, 1.65 mmol, 65%) as a foamy brown solid (Found:
C, 56.18; H, 7.72%. C19H31NbSi2 requires C, 55.86; H, 7.65%); H
(C6D6) −0.6, −0.15 (2d, 2J = 13.4 Hz, 2 × 1H, NbCH2SiMe3), 0.09,
0.14 (2s, 2 × 3H, SiMe2), 0.18 (s, 9H, SiMe3), 1.10 (dd, 2J = 14.7 Hz,
All experiments were carried out under argon using a Vacuum
Atmospheres glove box or standard Schlenk techniques. Hydro-
carbon solvents and THF were distilled from Na/benzophenone
and stored under argon prior to use. Complexes [M(5-Cp){5-
C5H4SiMe2(CH2CHCH2)}Cl2] (Cp = C5H5, M = Nb 1, Ta 2,
Cp = C5Me5, M = Ta 3) were prepared by methods reported11
previously. NMR spectra were recorded at 20 °C on Unity-300 and
Unity Plus 500 instruments in Teflon-valved tubes. 1H and 13C NMR
chemical shifts were measured relative to the resonances of C6D6
used as solvent. Coupling constants are reported in Hz. C, H, and N
analyses were carried out with a Perkin-Elmer 240 C analyzer.
2
3
3J = 8.7 Hz, 1H, SiCH2), 1.29 (dd, J = 5.1 Hz, J = 11.7 Hz, 1H,
[Ta(5-C5Me5){5-C5H4SiMe2(CH2-2-CHCH2)}Cl] (4). A
solution of [Ta(5-C5Me5){5-C5H4SiMe2(CH2CHCH2)}Cl2] (3)
(1.005 g, 1.82 mmol) in toluene (75 cm3) was added to 10% Na/Hg
(0.042 g, 1.82 mmol) and the mixture was stirred for 60 h at room
temperature. NaCl was separated by filtration and the solvent was
removed under vacuum from the resulting green solution. The
residue was extracted into hexane (50 cm3) and the solution was
concentrated to 10 cm3 and cooled at −40 °C to give a mixture of
(4-exo + 4-endo) (0.610 g, 1.18 mmol, 65%) as a yellow green solid
(Found: C, 46.30; H, 5.96%. C20H30ClSiTa requires C, 46.65; H,
5.87%); (4-endo): H (C6D6) 0.25, 0.27 (2s, 2 × 3H, SiMe2), 1.00
2
3
CH2), 1.42 (dd, J = 5.1 Hz, J = 9.1 Hz, 1H, CH2), 2.10 (dd,
2J = 14.7 Hz, 3J = 6.9 Hz, 1H, SiCH2), 2.45 (m, 1H, CH), 4.52 (s,
5H, C5H5), 3.70, 4.72(2) and 5.6 (4m, 4 × 1H, C5H4); C (C6D6) −5.9
(NbCH2SiMe3), −3.6, 2.0 (SiMe2), 4.6 (NbCH2SiMe3) 27.9 (SiCH2),
44.6 (CH2), 53.4 (CH), 91.2, 101.1, 105.5, 117.4 (C5H4), 97.5
(C5H5) and 100.2 (C5H4 ipso).
[Ta(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}R] (R = Me
8, CH2Ph 9). A solution of MgClR (R = Me 1.40 cm3, 4.16 mmol;
R = CH2Ph 2.08 cm3, 4.06 mmol) in THF was added at −78 °C to
a solution of [Ta(5-C5H5){5-C5H4SiMe2(CH2CHCH2)}Cl2] (2)
(1.00 g, 2.08 mmol) in toluene (50 cm3) The mixture was stirred for
16 h at room temperature. MgCl2 was separated by filtration and the
solvent was removed under vacuum from the resulting brown solu-
tion. The residue was extracted into hexane (30 cm3) and the solu-
tion was evaporated and cooled to give 8 (0.42 g, 1.24 mmol, 60%)
or 9 (0.51 g, 1.24 mmol, 60%) as foamy brown solids. (8): (Found:
C, 45.52; H, 5.51%. C16H23SiTa requires C, 45.28; H, 5.46%); H
(C6D6) 0.11, 0.17 (2s, 2 × 3H, SiMe2), 0.37 (s, 3H, TaMe) 1.15
2
3
(dd, J = 6.5 Hz, J = 11.0 Hz, 1H, CH2), 1.39 (m, 1H, CH),
2
3
1.49 (s, 15H, C5Me5), 1.59 (dd, J = 6.5 Hz, J = 11.5 Hz, 1H,
CH2), 3.12 (dd, 2J = 14.5 Hz, 3J = 4.5 Hz, 1H, SiCH2), 3.86 (dd,
2J = 14.5 Hz, J = 3.0 Hz, 1H, SiCH2) and 4.02, 4.65, 6.15, 6.35
3
(4m, 4 × 1H, C5H4); C (C6D6) 0.3, 0.5 (SiMe2), 11.6 (C5Me5), 32.2
(SiCH2), 38.3 (CH2), 49.4 (CH), 93.6, 97.1, 109.9, 135.6 (C5H4),
105.8 (C5H4 ipso) and 108.1 (C5Me5). (4-exo): H (C6D6) 0.04, 0.13
(2s, 2 × 3H, SiMe2), 0.90 (dd, 2J = 7.0 Hz, 3J = 10.5 Hz, 1H, CH2),
2
1.53 (s, 15H, C5Me5), 1.81 (m, 1H, CH), 1.91 (dd, J = 7.0 Hz,
3J = 12.0 Hz, 1H, CH2), 2.13 (dd, 2J = 15.0 Hz, 3J = 6.5 Hz, 1H,
SiCH2), 2.49 (dd, 2J = 15.0 Hz, 3J = 5.0 Hz, 1H, SiCH2) and 4.14,
5.03, 5.50, 5.82 (4m, 4 × 1H, C5H4); C (C6D6) −1.9, 0.1 (SiMe2),
11.9 (C5Me5), 30.1 (SiCH2), 47.3 (CH2), 55.0 (CH), 95.4, 108.6,
118.4, 120.4 (C5H4), 104.9 (C5H4 ipso) and 107.8 (C5Me5).
2
3
2
(dd, J = 14.7 Hz, J = 7.2 Hz, 1H, SiCH2), 0.93 (dd, J = 6.9 Hz,
3J = 10.8 Hz, 1H, CH2), 1.32 (dd, J = 6.9 Hz, J = 9.9 Hz, 1H,
2
3
2
3
CH2), 2.68 (dd, J = 14.7 Hz, J = 7.5 Hz, 1H, SiCH2), 1.97 (m,
1H, CH), 4.45 (s, 5H, C5H5) and 3.82, 4.58, 4.64, 5.28 (4m,
4 × 1H, C5H4); C (C6D6) −7.0 (TaMe), −3.3, 2.1 (SiMe2), 29.2
(SiCH2), 35.2 (CH2), 38.4 (CH), 91.2, 98.9, 99.5, 120.3 (C5H4),
97.7 (C5H5) and 106.4 (C5H4 ipso).
[Nb(5-C5H5){5-C5H4SiMe2(CH2-2-CHCH2)}R] (R = Me
5, CH2Ph 6). A solution of MgClR (R = Me 1.7 cm3, 5.1 mmol;
R = CH2Ph 2.55 cm3, 5.1 mmol) in THF was added at −78 °C to
a solution of [Nb(5-C5H5){5-C5H4SiMe2(CH2CHCH2)}Cl2] (1)
(1.00 g, 2.55 mmol) in toluene (50 cm3) The mixture was stirred
for 16 h at room temperature. MgCl2 was separated by filtration,
(9): (Found: C, 53.09; H 5.55%. C22H27SiTa requires C, 52.80;
H, 5.44%); H (C6D6) 0.09, 0.10 (2s, 2 × 3H, SiMe2), 0.91 (dd,
2J = 6.6 Hz, 3J = 10.8 Hz, 1H, CH2), 1.25 (dd, 2J = 6.6 Hz,
3J = 10.2 Hz, 1H, CH2), 1.36 (dd, 2J = 14.7 Hz, 3J = 6.0 Hz, 1H,
SiCH2), 2.08 (m, 1H, CH), 4.43 (s, 5H, C5H5), 2.34, 2.68 (2d,
D a l t o n T r a n s . , 2 0 0 4 , 2 9 4 3 – 2 9 5 1
2 9 4 9

View Article Online
2J = 11.5 Hz 2 × 1H, TaCH2Ph), 2.56 (dd, 2J = 14.7 Hz, 3J = 7.2 Hz,
1H, SiCH2), 3.88, 4.53(2), 5.67, (4m, 4 × 1H, C5H4) and 6.9–7.4
(m, 5H, TaCH2Ph); C (C6D6) −3.2, 1.6 (SiMe2), 29.2 (SiCH2), 35.7
(CH2), 38.1 (TaCH2Ph), 42.0 (CH), 92.3, 99.0, 100.1, 98.0
(C5H4), 108.1 (C5H4 ipso), 121.8 (C5H5), 126.1–129.2, (CH2Ph) and
141.8 (TaCH2Ph ipso).
2J = 15.2 Hz, 3J = 4.6 Hz, 1H, SiCH2), 4.62(2) and 5.27, 5.46 (4m,
4 × 1H, C5H4); C (C6D6) −1.5, −0.1 (SiMe2), −0.4 (TaCH2SiMe3),
5.5 (TaCH2SiMe3), 11.2 (C5Me5), 30.7 (SiCH2), 36.1 (CH2), 40.1
(CH), 97.4, 100.8, 102.4, 107.7 (C5H4), 105.1 (C5Me5) and 118.1
(C5H4 ipso).
Method 2. A solution of Li(CH2SiMe3) (0.091 g, 0.97 mmol)
in hexane (20 cm3) was added to a solution of [Ta(5-C5Me5){5-
C5H4SiMe2(CH2-2-CHCH2)}Cl] (4) (0.5 g, 0.97 mmol) in
hexane (30 cm3) at −78 °C. The mixture was stirred for 16 h at room
temperature. LiCl was separated by filtration and the solvent was
removed under vacuum from the resulting brown solution to obtain
12 (0.55 g, 0.73 mmol, 75%) as a foamy brown solid.
[Ta(5-C5Me5){5-C5H4SiMe2(CH2-2-CHCH2)}R]
(R = Me 10, CH2Ph 11). Method 1. A solution of MgClR (R = Me
1.82 cm3, 5.46 mmol; R = CH2Ph 2.73 cm3, 5.46 mmol) in THF
was added at −78 °C to a solution of [Ta(5-C5Me5){5-C5H4SiMe2-
(CH2CHCH2)}Cl2] (3) (1.00 g, 1.82 mmol) in toluene (50 cm3).
The mixture was stirred for 16 h at room temperature. MgCl2 was
separated by filtration and the solvent was removed under vacuum
from the resulting brown solution. The residue was extracted into
hexane (30 cm3) and the solution was evaporated and cooled to give
a mixture of (10-endo + 10-exo) (0.48 g, 1.36 mmol, 75%) or (11)
(0.78 g, 1.36 mmol, 75%) as green solids. (10): (Found: C, 51.16;
H, 6.80%. C21H33SiTa requires C, 51.00; H, 6.73%). (10-exo): H
(C6D6) 0.07 (dd, 2J = 7.7 Hz, 3J = 11.0 Hz, 1H, CH2), 0.10, 0.16
(2s, 2 × 3H, SiMe2), 0.14 (s, 3H, TaMe), 1.04 (m, 1H, CH), 1.42 (s,
15H, C5Me5), 1.25 (dd, 2J = 7.7 Hz, 3J = 11.2 Hz, 1H, CH2), 2.15
(dd, 2J = 15.1 Hz, 3J = 5.2 Hz, 1H, SiCH2), 2.40 (dd, 2J = 15.1 Hz,
3J = 7.7 Hz, 1H, SiCH2) and 4.48, 4.55, 4.65, 5.11 (4m, 4 × 1H,
C5H4); C (C6D6) −0.2 (TaMe), −0.1, 0.01 (SiMe2), 10.5 (C5Me5),
30.2 (SiCH2), 38.7 (CH), 40.7 (CH2), 96.5, 98.7, 103.6, 107.3
(C5H4), 103.0 (C5Me5) and 115.2 (C5H4 ipso).
[M(5-C5R5){5-C5H4SiMe2(CH2–CHCH2)}XL] 13a–13d,
14a–14g. The synthesis of all these compounds and full details of
their structural characterization are given in the ESI.†
[Ta(5-C5Me5){5-C5H4SiMe2(CH2–CHCH2)}H(NtBu)]
15. C6D6 (0.4 cm3) was added to a mixture of LiNHtBu (4.6 mg,
0.116 mmol)
and
[Ta(5-C5Me5){5-C5H4SiMe2(CH2-2-
CHCH2)}Cl] (4) (30 mg, 0.058 mmol) into a NMR tube. After
24 h at room temperature 15 was the unique product in solution. H
(C6D6) 0.45, 0.44 (2s, 2 × 3H, SiMe2), 1.14 (s, 9H, CMe3), 1.86 (d,
3J = 8.4 Hz, 2H, SiCH2), 1.95 (s, 15H, C5Me5), 4.73, 5.44, 5.75, 6.11
(4m, 4 × 1H, C5H4), 4.98 (m, 2H, CH2), 5.91 (m, 1H, CH) and
5.95 (s, 1H, TaH); C (C6D6) −1.4, −1.5 (SiMe2), 12.9 (C5Me5), 25.6
(SiCH2), 113.3 (CH2), 97.7, 106.1, 109.8, 110.8 (C5H4), 135.5
(CH), 113.4 (C5Me5) and 115.5 (C5H4 ipso).
(10-endo): H (C6D6) 0.18, 0.23 (2s, 2 × 3H, SiMe2), 0.26 (dd,
2J = 7.1 Hz, 3J = 11.3 Hz, 1H, CH2), 0.33 (s, 3H, TaMe), 0.45 (m,
1H, CH), 1.04 (dd, 2J = 7.1 Hz, 3J = 10.6 Hz, 1H, CH2), 1.39 (s,
15H, C5Me5), 2.58 (dd, 2J = 14.8 Hz, 3J = 4.9 Hz, 1H, SiCH2), 3.38
(dd, 2J = 14.8 Hz, 3J = 2.9 Hz, 1H, SiCH2) and 4.25, 4.38, 5.48, 5.98
(4m, 4 × 1H, C5H4); C (C6D6) −2.81 (TaMe), 0.43, 2.09 (SiMe2),
10.53 (C5Me5), 25.9 (SiCH2), 30.6 (CH2), 40.5 (CH), 92.1,
103.6, 112.7, 123.9 (C5H4), 103.4 (C5Me5) and 95.2 (C5H4 ipso).
(11): (Found: C 56.72, H 6.35%. C27H37SiTa requires C, 56.83;
H, 6.54%); H (C6D6) −0.10, 0.02 (2s, 2 × 3H, SiMe2), 0.36 (dd,
2J = 6.9 Hz, 3J = 10.7 Hz, 1H, CH2), 1.21 (dd, 2J = 6.9 Hz,
3J = 10.6 Hz, 1H, CH2), 1.17 (m, 1H, CH), 1.39 (s, 15H,
Computational details
Density functional theory (DFT) calculations were performed with
the TURBOMOLE program package, version 5.5.24 The Vosko–
Wilk–Nusair25 local density approximation (LDA) and the general-
ized gradient approximation (GGA) with corrections for exchange
and correlation according to Becke26 and Perdew27 (BP86) were
used for all calculations. The TURBOMOLE approach to DFT GGA
calculations is based on the use of Gaussian-type-orbitals (GTO) as
basis functions. Geometries were optimized within the framework
of the RI-J approximation28 using accurate triple- valence basis sets
augmented by one polarization function TZV(P)29 for all elements.
Relativistic effective core potentials were employed to model the
energetically deep-lying and chemically mostly inert 1s–3d core
electrons of the zirconium and niobium atoms, and 1s–4f core
electrons of the hafnium and tantalum atoms. Single-point calcula-
tions with the ESCF program30 of the same TURBOMOLE package
have followed the geometry optimizations to check the instability
of the models and no negative eigenvalues were found. Optimized
geometries and final energies are presented in the ESI.†
2
C5Me5), 1.69, 1.88 (2d, J = 7.8 Hz, 2 × 1H, TaCH2Ph), 1.75 (dd,
2J = 15.3 Hz, 3J = 4.9 Hz, 1H, SiCH2), 3.20 (dd, 2J = 15.3 Hz,
3J = 2.9 Hz, 1H, SiCH2), 4.11, 4.59, 4.82, 5.12 (4m, 4 × 1H, C5H4)
and 7.32–7.42 (m, 5H, TaCH2Ph); C (C6D6) −1.8, −0.5 (SiMe2),
10.8 (C5Me5), 25.3 (TaCH2Ph), 29.3 (SiCH2), 39.3 (CH2), 42.3
(CH), 95.1, 102.4, 104.9, 112.0 (C5H4), 104.7 (C5Me5), 123.0
(C5H4 ipso) and 121.4 (TaCH2Ph).
Method 2.Asolution of MgClCH2Ph (0.5 cm3, 1.0 mmol) in THF
was added to a solution of [Ta(5-C5Me5){5-C5H4SiMe2(CH2-2-
CHCH2)}Cl] (4) (0.50 g, 0.97 mmol) at −78 °C and the mixture
was stirred for 16 h at room temperature. MgCl2 was separated by
filtration and the solvent was removed under vacuum from the
resulting brown solution. The residue was extracted into hexane
(50 cm3) and the solution was concentrated to 10 cm3 and cooled
at −40 °C to give 11 (0.39 g, 0.70 mmol, 70%) as a green solid.
Acknowledgements
The authors acknowledge MCyT (project MAT2001-1309), the
Funds of the University of Zürich and COST (project COST-D12/
0016/98) for financial support.
References
[Ta(5-C5Me5){5-C5H4SiMe2(CH2-2-CHCH2)}CH2SiMe3]
12. Method 1. A solution of Li(CH2SiMe3) (0.25 g, 2.72 mmol)
in hexane (20 cm3) was added to a solution of [Ta(5-C5Me5){5-
C5H4SiMe2(CH2CHCH2)}Cl2] (3) (0.5 g, 0.908 mmol) in toluene
(30 cm3) at −78 °C. The mixture was stirred for 16 h at room
temperature. LiCl was separated by filtration and the solvent was
removed under vacuum from the resulting brown solution. The
residue was extracted into hexane (50 cm3) and the solution was
evaporated and cooled to obtain 12 (0.51 g, 0.59 mmol, 65%) as
a brown solid (Found: C, 51.13; H, 7.38%. C24H41Si2Ta requires
C, 50.87; H, 7.29%); H (C6D6) −1.12, 0.76 (2d, 2J = 11.0 Hz,
2 × 1H, TaCH2SiMe3), 0.00 (m, 2H, CH2), 0.16, 0.20 (2s, 2 × 3H,
SiMe2), 0.48 (s, 9H, SiMe3), 0.90 (m, 1H, CH), 1.37 (s, 15H,
C5Me5), 2.20 (dd, 2J = 15.2 Hz, 3J = 6.2 Hz, 1H, SiCH2), 2.54 (dd,
1 (a) Z. Wu, R. F. Jordan and J. L. Petersen, J. Am. Chem. Soc., 1995,
117, 5867; (b) J.-F. Carpentier, Z. Wu, C. W. Lee, S. Stroemberg,
J. N. Christopher and R. F. Jordan, J. Am. Chem. Soc., 2000, 122,
7750; (c) C. P. Casey and D. W. Carpenetti II, Organometallics, 2000,
19, 3970; (d) C. P. Casey, D. W. Carpenetti II and H. Sakurai, Organo-
metallics, 2001, 20, 4262; (e) C. G. Brandow, A. Mendiratta and
J. E. Bercaw, Organometallics, 2001, 20, 4253.
2 (a) R. F. Jordan, Adv. Organomet. Chem., 1991, 32, 325; (b) T. J. Marks,
Acc. Chem. Res., 1992, 25, 57; (c) H. H. Brintzinger, D. Fischer,
R. Mülhaupt, B. Rieger and R. M. Waymouth, Angew. Chem., Int.
Ed. Engl., 1995, 34, 1143; (d) M. Bochmann, J. Chem. Soc., Dalton
Trans., 1996, 255; (e) G. G. Hlatky, Coord. Chem. Rev., 1999, 181, 243;
(f) H. G. Alt andA. Köppl, Chem. Rev., 2000, 100, 1205; (g) L. Resconi,
L. Cavallo, A. Fait and F. Piemontesi, Chem. Rev., 2000, 100, 1253.
3 M. V. Galakhov, G. Heinz and P. Royo, Chem. Commun., 1998, 17.
2 9 5 0
D a l t o n T r a n s . , 2 0 0 4 , 2 9 4 3 – 2 9 5 1

View Article Online
4 J. Cano, P. Gómez-Sal, G. Heinz, G. Martínez and P. Royo, Inorg. Chim.
Acta, 2003, 345, 15.
15 U. Burckhardt, G. L. Casty, J. Gavenonis and T. Don Tilley, Organo-
metallics, 2002, 21, 3108.
5 (a) F. N. Tebbe and G. W. Parshall, J. Am. Chem. Soc., 1971, 93, 3793;
(b) N. M. Doherty and J. E. Bercaw, J. Am. Chem. Soc., 1985, 107,
2670; (c) V. C. Gibson, J. E. Bercaw, W. J. Bruton and R. D. Sanner,
Organometallics, 1986, 5, 976; (d) B. J. Burger, B. D. Santarsiero,
M. S. Trimmer and J. E. Bercaw, J. Am. Chem. Soc., 1988, 110, 3134;
(e) A. Antiñolo, F. Carrillo, S. García-Yuste and A. Otero, Organo-
metallics, 1994, 13, 2761.
16 A. N. Chernega, M. L. H. Green and A. G. Suárez, J. Chem. Soc.,
Dalton Trans., 1993, 3031.
17 G. Parkin, A. van Asselt, D. J. Leahy, L. Whinnery, N. G. Hua,
R. W. Quan, L. M. Henling, W. P. Schaefer, B. D. Santarsiero and
J. E. Bercaw, Inorg. Chem., 1992, 31, 82.
18 L. J. Guggenberger, P. Meakin and F. N. Tebbe, J. Am. Chem. Soc.,
1974, 96, 5420.
6 (a) A. H. Klazinga and J. H. Teuben, J. Organomet. Chem., 1978, 157,
413; (b) A. H. Klazinga and J. H. Teuben, J. Organomet. Chem., 1980,
194, 309; (c) M. E. Eichiner, H. G. Alt and M. D. Rausch, J. Organomet.
Chem., 1984, 264, 309; (d) A. Antiñolo, F. Carrillo-Hermosilla, I. del
Hierro, A. Otero, M. Fajardo and Y. Mugnier, Organometallics, 1997,
16, 4161; (e) P. J. Chirik, D. L. Zubris, L. J. Ackerman, L. M. Henling,
M. W. Day and J. E. Bercaw, Organometallics, 2003, 22, 172.
7 (a) D. L. Lichtenberger, G. P. Darsey, G. E. Kellogg, R. D. Sanner,
V. G. Young, Jr. and J. R. Clark, J. Am. Chem. Soc., 1989, 111, 5019;
(b) M. L. H. Green, A. Sella and L.-L. Wong, Organometallics, 1992,
11, 2650.
19 G. C. Levy and G. L. Nelson, Carbon-13 NMR for Organic Chemists,
Wiley, New York, 1972.
20 J. L. Marshall, Carbon–Carbon and Carbon-Proton NMR Coupling,
Verlag Chemie Inter., Berlin, 1983.
21 J. W. Lauher and R. Hoffmann, J. Am. Chem. Soc., 1976, 98, 1729.
22 M. J. S. Dewar, Bull. Chem. Soc. Fr., 1951, 18, C-79.
23 J. Chatt and L. A. Duncanson, J. Chem. Soc., 1953, 2939.
24 TURBOMOLE, Program package for ab initio electronic structure
calculations, Theoretical Chemistry, University of Karlsruhe, Germany,
http://www.turbomole.com; (a) R. Ahlrichs, M. Bär, M. Häser, H. Horn
and C. Kölmel, Chem. Phys. Lett., 1989, 162, 165; (b) O. Treutler and
R. Ahlrichs, J. Chem. Phys., 1995, 102, 346; (c) M. von Arnim and
R. Ahlrichs, J. Comput. Chem., 1998, 19, 1746.
8 L. J. Ackerman, M. L. H. Green, J. C. Green and J. E. Bercaw, Organo-
metallics, 2003, 22, 188.
9 S. Fredericks and J. L. Thomas, J. Am. Chem. Soc., 1978, 100, 350.
10 (a) A. H. Klazinga and J. H. Teuben, J. Organomet. Chem., 1979, 165,
31; (b) A. H. Klazinga and J. H. Teuben, J. Organomet. Chem., 1980,
192, 75.
11 P. Nicolás and P. Royo, Inorg. Chim. Acta, 2004, in press.
12 (a) D. A. Lemenovskii, T. V. Baukova and V. P. Fedin, J. Organomet.
Chem., 1977, 132, C14; (b) L. Acedo, A. Otero and P. Royo, J.
Organomet. Chem., 1983, 258, 181.
13 (a) I. H. Elson, J. K. Kochi, U. Klabunde, L. E. Manzer, G. W. Parshall
and F. N. Tebbe, J. Am. Chem. Soc., 1974, 96, 7374; (b) L. E. Manzer,
Inorg. Chem., 1977, 16, 525.
14 P. Royo and E. Ryan, Metallocenes, ed. A. Togni and R. L. Halterman,
Wiley-VCH, New York, 1998, vol. 1, p. 321, and references therein.
25 S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1980, 58, 1200.
26 A. D. Becke, Phys. Rev. A: At. Mol. Opt. Phys., 1988, 38, 3098.
27 (a) J. P. Perdew, Phys. Rev. B: Condens. Matter, 1985, 33, 8822;
(b) J. P. Perdew, Phys. Rev. B: Condens. Matter, 1986, 34, 7406.
28 (a) K. Eichkorn, O. Treutler, H. Öhm, M. Häser and R. Ahlrichs, Chem.
Phys. Lett., 1995, 240, 283; K. Eichkorn, O. Treutler, H. Öhm, M. Häser
and R. Ahlrichs, Chem. Phys. Lett., 1995, 242, 652; (b) K. Eichkorn,
F. Weigand, O. Treutler and R. Ahlrichs, Theor. Chem. Acc., 1997, 97,
119.
29 A. Schäfer, C. Huber and R. Ahlrichs, J. Chem. Phys., 1994, 100, 5829.
30 (a) R. Bauernschmitt and R. Ahlrichs, Chem. Phys. Lett., 1996, 256,
454; (b) H. Weiss, R. Ahlrichs and M. Haser, J. Chem. Phys., 1993, 99,
1262; (c) F. Furche, J. Chem. Phys., 2001, 114, 5982.
D a l t o n T r a n s . , 2 0 0 4 , 2 9 4 3 – 2 9 5 1
2 9 5 1