Effects of substituents in cyclopentadienyltitanium trichlorides on electronic absorption and 47,49Ti NMR spectra and styrene polymerization activated by methylalumoxane

Article abstract of DOI:10.1016/j.molcata.2006.05.016

In the series of the methylated compounds (η⁵-C₅H_5-nMe_n)TiCl₃ (n = 0-5) 1-6, linear correlations of λ_max of the absorption band and δ ^49,47Ti chemical shift with the number of Me groups were found. It is due to the dependence of both the magnitudes on the HOMO-LUMO energy gap which is influenced by electron donating methyl substituents. The catalysts made by combining 1-6 with MAO (molar ratio Al/Ti 500) showed only a gross dependence of decreasing activity with the number of methyl groups. For compounds (η⁵-C₅Me₄R)TiCl₃ 7-14 and (η⁵-C₅H(1,2,3-)Me₃R)TiCl₃ 15-20 where R = alkyl, silyl or phenyl no correlation between λ_max and δ was found. Limited DFT calculations were performed to assign HOMO/LUMO orbitals involved. Crystal structures of 16 and 20 were determined and some other inspected to prove the absence of steric hindrance. The poor correlation of the spectroscopic properties with the catalytic activity is not surprising because steric effects of cyclopentadienyl ligands at cationic polymerization centers are generally anticipated.

Full text of DOI:10.1016/j.molcata.2006.05.016

Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
Effects of substituents in cyclopentadienyltitanium trichlorides
on electronic absorption and 47,49Ti NMR spectra and styrene
polymerization activated by methylalumoxane
a
b
d
e
ˇ
ˇ´
´
ˇ
ˇ
´ˇc
´ˇ
´
Jirı Pinkas , Antonın Lycka , Pavel Sindelar , Robert Gyepes , Vojtech Varga ,
a
a
a,∗
ˇ´
Jirı Kubista , Michal Horacek , Karel Mach
a J. Heyrovsky´ Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, Dolejsˇkova 3, 182 23 Prague 8, Czech Republic
b Research Institute for Organic Syntheses, Rybitv´ı 296, 532 18 Pardubice 20, Czech Republic
c Polymer Institute Brno, Tkalcovska´ 2, 656 49 Brno, Czech Republic
d Department of Inorganic Chemistry, Charles University, Hlavova 2030, 128 40 Prague 2, Czech Republic
e
´
Research Institute of Inorganic Chemistry, Revolucˇn´ı 84, 400 01 Ust´ı nad Labem, Czech Republic
Available online 14 June 2006
Dedicated to Professor Bogdan Marciniec on the occasion of his 65th birthday.
Abstract
In the series of the methylated compounds (␩5-C5H5−nMen)TiCl3 (n = 0–5) 1–6, linear correlations of λmax of the absorption band and δ
49,47Ti chemical shift with the number of Me groups were found. It is due to the dependence of both the magnitudes on the HOMO–LUMO
energy gap which is influenced by electron donating methyl substituents. The catalysts made by combining 1–6 with MAO (molar ratio Al/Ti
500) showed only a gross dependence of decreasing activity with the number of methyl groups. For compounds (␩5-C5Me4R)TiCl3 7–14 and
(η5-C5H(1,2,3-)Me3R)TiCl3 15–20 where R = alkyl, silyl or phenyl no correlation between λmax and δ was found. Limited DFT calculations were
performed to assign HOMO/LUMO orbitals involved. Crystal structures of 16 and 20 were determined and some other inspected to prove the
absence of steric hindrance. The poor correlation of the spectroscopic properties with the catalytic activity is not surprising because steric effects
of cyclopentadienyl ligands at cationic polymerization centers are generally anticipated.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Titanium; Half-sandwich titanocenes; Substituent effects; 47Ti and 49Ti NMR; Electronic absorption spectra; Styrene polymerization; Syndiotactic
polystyrene
1. Introduction
substituents attached to the cyclopentadienyl ligands [1]. Partic-
ularly, the effect of methyl group(s), which increases electron
density at the metal center, has been investigated by various
spectroscopic, physico-chemical and chemical methods with
the aim to effectively design new catalyst systems based on
sandwich titanocene compounds. ESCA (electron spectroscopy
for chemical analysis) measurements revealed that replacement
of cyclopentadienyl by permethylcyclopentadienyl ligands in
the metallocene dihalide series (C5H5−nMen)2MX2 (n = 0,5)
(M = Ti, Zr, Hf) (X = F, Cl, Br) results in the decrease of bind-
ing energies of the inner shell metal (2p3/2) electrons by ca.
0.8 eV [2]. Similar dependences for valence electron energies
were obtained by UPS (ultraviolet photoelectron spectroscopy)
in the series of (C5H5−nMen)2TiCl2 [3] and (C5H5−nMen)2TiCl
[4] (n = 0–5) compounds. Chemical shifts in the 49Ti and 47Ti
Bent sandwich titanocene dichlorides, constrained geometry
titanium complexes (CGC), or half-sandwich titanium trichlo-
rides combined with aluminium alkyls or methylalumoxane
(MAO) are frequently used homogeneous catalysts for ␣-olefin
polymerization or copolymerization. Due to the outstanding
steric and electronic versatility of the cyclopentadienyl-based
ligands, such catalysts can produce polyolefins with a broad
range of molecular weights and their distributions, composi-
tions and microstructures. The catalyst properties can be finely
tuned by electron donating/withdrawing and steric effects of the
∗
Corresponding author. Tel.: +420 2 6605 3735; fax: +420 2 8658 2307.
E-mail address: mach@jh-inst.cas.cz (K. Mach).
1381-1169/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2006.05.016

J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
15
Cp < C5H4Me = C5H2Me3 < Cp* in Ref. [16d]. Indeed, a higher
activity for Cp* > Cp was reported for polymerizations at 90 ◦C
and Al/Ti = 1000 [17] contrary to ca. five times higher activity
for Cp > Cp* at 50 ◦C and Al/Ti = 300 [18], and even ca. 70 times
higher activity for Cp > Cp* at 50 ◦C and Al/Ti = 300 [19]. Sur-
prisingly, partly methyl-substituted CpꢀTiCl3 compounds were
tested in catalytic systems with MAO only scarcely [16a].
In this paper we report the electronic absorption spec-
tra and 49Ti and 47Ti NMR data for a series of the (␩5-
C5H5−nMen)TiCl3 (n = 0–5) compounds and highly methyl-
substituted compounds CpꢀTiCl3 where Cpꢀ is (␩5-C5Me4R)
or (␩5-C5H(1,2,3-)Me3R) with R being an alkyl, phenyl or
silyl substituent. The nature of the electronic absorption band
is elucidated and the NMR shifts are correlated with DFT
calculations for representative complexes. The spectroscopic
data are compared with results of styrene polymerization using
the CpꢀTiCl3/MAO catalysts with Cpꢀ = C5H5−nMen (n = 0–5),
C5Me4R (R = but-3-en-1-yl, Bu, Ph, SiMe3), and C5H(1,2,3-
)Me3(4-)R (R = Ph, SiMe3) with the aim to evaluate applicability
of spectroscopic data for tuning the catalysts performance.
NMR also closely correlate with the electron density at the metal
nucleus [5,6]. Moreover, the linewidth of signals brings infor-
mation on the ligand field symmetry around the metal atom
which applies in structure/reactivity studies [7]. Similar studies
on the 91Zr NMR spectra/structure correlations are also numer-
ous [8]. The standard first reduction potentials (E10) for the
(C5H5−nMen)2TiCl2 (n = 0–5) series showed a negative incre-
ment of 0.093 V per methyl group for n = 0–3, a decline to a
lower value for n = 4, and a positive shift for n = 5 and for bulkier
fifth substituents. This was accounted for by an increase in the
Cg–Ti–Cg (Cg, centroid of the cyclopentadienyl ring) angle due
to a steric hindrance, diminishing the HOMO–LUMO energy
gap [9]. Also, the chemical behavior as expressed in the affin-
ity to coordinate THF in the series of (C5H5−nMen)2TiCl [10a]
and (C5H5−nMen)2Zr(Me3SiC CSiMe3) (n = 0–5) compounds
[10b] decreasing with increasing n and becoming negligible
for the permethylated compounds can be due to combination
of electronic and steric effects. In the half-sandwich series of
(C5H5−nMen)TiCl3 (n = 0–5) compounds the electronic absorp-
tion spectra showed a shift of the visible absorption band to
longer wavelength by ca. 11 nm per methyl group band. Accord-
ingly, the compounds showed a decrease of their reducibility by
an excess of Et2AlCl expressed by a decrease in the reduction
rate constant (1.1 × 10−3 to 6.1 × 10−5 s−1) [11]. The 49Ti and
47Ti NMR spectra of a number of CpꢀTiCl3 compounds proved
that methyl and trimethylsilyl substituents move chemical shifts
down-field, however, the latter in a much lower extent. Simi-
lar bathochromic trends were also observed for the electronic
absorption band, the shifts per SiMe3 group being again smaller
than for a Me group [12].
The cyclopentadienyl substituent effects in polymerizations
were widely explored in metallocene or ansa-metallocene cat-
alysts [1]. Among CGC catalysts, replacement of one methyl
group in ␣-position to the dimethylsilylene bridge in the par-
ent CGC complex by other substituents exerted large impacts
on copolymerization of styrene with ethene in optimum afford-
ing an alternating copolymer [13], however, no syndiotactic
enrichment of atactic polypropene was detected with the same
catalysts [14]. The half-sandwich catalysts turned out to be
high-performance catalysts for the polymerization of styrene
to syndiotactic polystyrene (s-PS) when activated with excess
methylalumoxane (MAO). The auxiliary cyclopentadienyl lig-
and in initially used CpTiCl3 (Cp = ␩5-C5H5) [15] was modified
both by electron releasing as well as electron withdrawing sub-
stituents, or replaced by annelated cyclopentadienyl ligands (␩5-
indenyl, ␩5-fluorenyl, and their derivatives) whereas the leaving
chloride ligands are advantageously replaced by fluorides or var-
ious alkyl- or aryloxides [16]. In addition, the performance of
the catalysts is dependent on the molar Al/Ti ratio, the quality
of MAO (namely content of free trimethylaluminum), tem-
perature of polymerization, nature of solvents, and procedure
of the catalyst preparation. Although the above conditions are
apparently controlled, different laboratories report varying data
about catalyst activity and molecular weight of s-PS. This can
lead to a misleading generalization for the fundamental cata-
lysts like CpTiCl3/MAO and Cp*TiCl3/MAO quoting increas-
ing polymerization activity for the cyclopentadienyl ligands
2. Results and discussion
2.1. Synthesis of half-sandwich complexes and their
electronic absorption spectra
The investigated half-sandwich complexes consist of three
series of compounds, methyl-substituted cyclopentadienylti-
tanium trichlorides of general formula (␩5-C5H5−nMen)TiCl3
(n = 0–5) 1-6, pentasubstituted compounds containing one other
substituent in addition to four Me groups (␩5-C5Me4R)TiCl3
7–14, and tetrasubstituted compounds containing one other sub-
stituent in addition to three vicinal Me groups (␩5-C5H(1,2,3-)
Me3R)TiCl3 15–20. The compounds are listed in Table 1
togetherwiththereferencetotheirorigin, thelongestwavelength
absorption band λmax, and chemical shifts of 49Ti and 47Ti reso-
nances and their line widths. Compounds 1–14 were prepared by
synproportionation reaction between titanocene dichloride and
TiCl4 (Eq. (1)) in nearly quantitative yields [22]. Compounds
15–20 were obtained by transmetallation reaction (Eq. (2)) in
rather moderate yields [23]. All compounds were characterized
by EI-MS, IR, 1H and 13C NMR spectra, and melting points in
sealed capillaries under nitrogen. Structures of compounds 16
and 20, representatives of the new, tetrasubstituted compounds
series, were proved by X-ray diffraction single crystal analysis.
Cp2TiCl2 + TiCl4 → 2CpTiCl3
(1)
(2)
The electronic absorption bands for all the measured com-
plexes (Table 1) consist of a broad, mostly asymmetrical band
whose λmax occurs in the range 384–450 nm and its approxi-
mate halfwidth is 70–100 nm. The largest changes in λmax were
observed in the series of the (␩5-C5H5−nMen)TiCl3 (n = 0–5)
compounds in agreement with previous measurements [11], the

16
J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
Table 1
Electronic absorption band and 49,47Ti chemical shifts of (␩5-Cpꢀ)TiCl3 compounds
Cpd
Cpꢀ
λ
δ(49Ti)
w1/2
δ(47Ti)
w1/2
Synthesis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
C5H5
C5H4Me
C5H3(1,3-)Me2
C5H2(1,2,4-)Me3
C5HMe4
C5Me5
C5Me4Et
C5Me4Bu
C5Me4(CH2CH2CH CH2)
C5Me4(CHMeCH CH2)
C5Me4Ph
C5Me4tBu
384
393
405
417
428
439
439
439
440
441
445
444
432
430
450
425
430
430
435
432
−387.0
−334.7
−280.3
−229.4
−164.7
−98.2
26
23
16
20
9
3
3
3
3
−653.2
−601.0
−546.3
−495.5
−431.1
−364.5
−378.7
−380.3
−381.8
−379.2
−379.9
−372.9
−399.0
n.d.a
−426.3
−464.7
−441.0
−438.5b; −435.6b
−440.0
−440
84
81
53
62
28
8
9
9
8
22
56
53
66
–
176
141
48
n.d.a; –
73
73
[9]
[9]
[9]
[9]
[9]
[9]
[9]
[10]
[10]
[10]
[10]
[12]
This work
This work
This work
This work
This work
This work
This work
This work
−112.4
−114.0
−115.5
−113.0
−113.8
−106.7
−132.1
−142.8
−159.0
−197.8
−174.5
−172.2b; −169.4b
−171.5
−173.4
7
19
15
21
42
35
27
15
18;23
16
20
C5Me4(SiMe3)
C5Me4(SiMe2CH2CH2CF3)
C5H(1,2,3–)Me3Ph
C5H(1,2,3–)Me3(SiMe3)
C5H(1,2,3–)Me3(iPr)
C5H(1,2,3–)Me3(sBu)
C5H(1,2,3-)Me3(tBu)
C5H(1,2,3-)Me3(cyclohexyl)
a
Not determined.
Two diastereomers due to chiral center in sec-butyl group.
b
Fig. 1. The dependence of λmax and δ 49Ti on the number of methyl groups in
compounds 1–6.
Fig. 2. Thedependenceof λmax and δ49TionthenatureofgroupRincompounds
6–14.
parent complex 1 showing the highest energy of the transition
(λmax 384 nm) and the smallest band half-width ca. 70 nm. The
band wavelength was increasing with increasing number of Me
substituents by ca. 11 nm per Me group to give 439 nm for 6 in
practical linear dependence (Fig. 1). This value was not chang-
ing further for compounds with the fifth substituent being Et
(7), Bu (8) or pendant alkenyl groups (9, 10). A distinct red
t
shift to 445 nm was observed for the bulky Bu group in com-
pound 12. On the other hand, the bulky silyl groups SiMe3 and
SiMe2CH2CH2CF3 in compounds 13 and 14 induced remark-
able blue shifts (see Fig. 2).
Fig. 3. Thedependenceof λmax and δ49TionthenatureofgroupRincompounds
5 and 15–20.
Analogously, replacement of one methyl group vicinal to
the proton in tetramethylated compound 5 (λmax 428 nm) by
alkyl groups resulted in small red shifts, the largest shift for
group show that the electron releasing effect of the SiMe3 group
is weaker than the effect of Me group. This is in line with the
electronic absorption spectra of (␩5-C5H5−n(SiMe3)n)TiCl3
(n = 1–3) which displayed smaller red shifts than the methyl-
substituted compounds [12a]. On the other hand, the red shifts
t
the Bu group in compound 19 to 435 nm. In this series of
compounds the SiMe3 group in 16 shifted λmax slightly to
shorter wavelengths (425 nm) (Fig. 3). These blue shifts for
compounds with one methyl group being substituted by SiMe3

J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
17
induced by the tBu group in compounds 12 or 19 show that this
group is more electron releasing than all other applied alkyl
substituents. Surprisingly large red shifts for the phenyl group
in compounds 11 and 15 are not compatible with the generally
known electron-withdrawing effect of this group. The observed
red shifts are tentatively ascribed to conjugation of phenyl and
cyclopentadienyl aromatic systems apparently diminishing the
HOMO–LUMO energy differences.
The nature of the observed absorption band which is broad
and whose shape is more or less asymmetrical was semiquan-
titatively elucidated by DFT calculations (see below). It was
confirmed that the band consists of several charge transfer tran-
sitions from cyclopentadienyl ligand based HOMO orbitals to
LUMO Ti d-orbitals. The precision of the performed calcu-
lations does not allow the successful simulation of the band
position and its shape, however, the orbital assignment is cor-
rect as proved by an assignment for UPS spectrum of 1 [24] (see
below).
in 6 (Table 1). A slight downfield shift from these values to δ
−106.7 ppm was observed for the bulkiest Bu group in com-
t
pound12, however, anotherverybulkySiMe3 groupin13moved
δ to −132.1 ppm and the SiMe2CH2CH2CF3 group in 14 to
−142.8 ppm (Fig. 2). These observations are in accord with pre-
vious reports that the SiMe3 group exerts only a small downfield
shift with respect to proton [12]. A similar situation was found
also for the (␩5-C5H(1,2,3-)Me3R)TiCl3 compounds 17–20.
Their chemical shifts were observed upfield from δ −164.7 ppm
for 5 spanning −170 to −175 ppm region. Also here, the SiMe3
group in 16 caused a larger upfield shift. On the other hand, the
phenyl group in 15 exerted a large shielding effect shifting δ
below the value for 5 (Fig. 3).
Over the whole series of investigated compounds, it is dif-
ficult to correlate the resonance halfwidth w1/2 with the com-
pound structure. It was generally accepted [26] and confronted
in a number of papers [5–8], that increasing w1/2 follows an
increase in molecular asymmetry. In the present data it is, how-
ever, difficult to explain large values of w1/2 for compounds 1–3,
a sudden drop for compound 5 and a minimum value of 3 Hz
for 6 as well as for asymmetrical alkyl-substituted compounds
7–10. It seems that a large increase in w1/2 among compounds
7–20 occurs only when the linking atom of the substituent dif-
fers remarkably from sp3 carbon atom bearing 1–3 protons, i.e.
in Ph, tBu or silyl groups.
2.2. 49Ti and 47Ti NMR spectra of CpꢀTiCl3 compounds
A series of CpꢀTiCl3 compounds were measured in C6D6
solutions under comparable conditions, and their 49Ti and 47Ti
chemical shifts (relative to TiCl4) and resonance half-widths are
given in Table 1. The two 49Ti (I = 7/2) and 47Ti (I = 5/2) nuclei
have almost identical magnetogyric ratios, and hence, the reso-
nances of both are readily apparent in the NMR spectra of most
compounds with separation of ca. 266 ppm. The 49Ti nucleus,
although of lower natural abundance (5.51% against 7.75%
for 47Ti), is easier observable because of smaller quadrupolar
broadening of its resonance line [25]. Therefore, the further
discussion will confine largely to 49Ti data. The 49Ti chemi-
cal shifts for compounds 1 and 6 are in reasonable agreement
with literature data which were obtained in CDCl3 solvent: 1, δ
−390 ppm, w1/2 60 Hz [12a], δ −396 ppm, w1/2 49 Hz [12b];
6, δ −85 ppm, w1/2 10 Hz [12a], δ −94.7 ppm, w1/2 23 Hz
[12b]. Larger w1/2 values compared with the data in Table 1
can be due to outer sphere coordination of the CDCl3 solvent.
The chemical shifts for compounds 1–6 are moving downfield
with an increment of ca. 60 ppm per Me group. The plot of
49Ti chemical shifts against the number of Me groups in the
(␩5-C5H5−nMen)TiCl3 (n = 0–5) compounds is shown in Fig. 1
together with the λmax plot. Both the dependencies are close to
linear, and this implies that the shielding effect of methyl groups
mimics their electron-donating effect in electronic absorption
spectra. This is in line with the suggestion that the chemical
shifts of transition metal nuclei are dominated by the param-
agnetic shielding which correlates with the energy difference
between the ground state and low-lying electronic excited states
[25b]. Indeed, the very good correlation between δ 49Ti and λmax
determines an increase of 5.254 ppm/nm. A similar relationship
was found in incomplete series of CpꢀTiCl3 compounds [12]
and in substituted (␩5-indenyl)TiCl3 compounds [26]. This is,
however, not valid for compounds of the (␩5-C5Me4R)TiCl3
or (␩5-C5H(1,2,3-)Me3R)TiCl3 types. The groups bulkier than
methyl in compounds 7–11 move δ upfield to −112 to −115 ppm
whereas their λmax values indicate the same electronic effect as
The DFT GIAO calculation of 49Ti NMR chemical shifts
showed the downfield shift of ca. 300 ppm when going from 1
to 6, however, virtually no difference in δ between 6, 12 and 13.
2.3. DFT calculations
DFT calculations were performed using the B3LYP func-
tional. The Gaussian 98, Revision A.7 program [27] was used
in all cases. DFT studies were carried out for the CpꢀTiCl3 com-
pounds 1, 6, 12, and 13. All model structures were generated by
application of a mirror plane containing the Ti atom, one of the
chlorine atoms and bisecting the cyclopentadienyl ligand. The
structures were optimized using the 3-21G basis set; all molec-
ular properties were obtained employing the 6-311G(d) basis
set for the optimized structures. Of the calculated MO orbitals
listed in Table 2 only two HOMO frontier orbitals involving
the cyclopentadienyl ligand orbitals (Nos. 10 and 11) and three
LUMO frontier orbitals Nos. 12–14, which are predominantly Ti
d-orbitals, are involved in the generation of the observed absorp-
tion band. Obeying transition rules, only three symmetrically
allowed transitions 6aꢀ → 7aꢀ, 6aꢀ → 8aꢀ, and 5aꢀꢀ → 6aꢀꢀ should
be present, however, the other three were also observed (albeit
with smaller oscillator strengths) since the symmetry exclusion
rule is broken due to the presence of the metal. The calculation
of totally six transitions which constitute the absorption band for
compounds 1, 6, 12, and 13 followed the trend of λmax values,
however, all calculated values were systematically shifted by
ca. 20 nm to shorter wavelengths. This discrepancy is probably
the consequence of using a rather limited basis set for all atoms
and/or the insufficient treatment of relativistic effects caused by
the presence of the metallic atom. Larger basis sets and effec-
tive core potentials were not used in order to avoid excessive

18
J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
Table 2
Calculated frontier orbitals, their energies and OVGF energies for 1
No.
Orbital notation
Orbital energy (a.u.)
OVGF energy (eV)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1aꢀ
−0.37932
−0.37861
−0.37399
−0.36298
−0.35948
−0.35400
−0.34536
−0.34219
−0.31388
−0.29675
−0.29205
−0.12520
−0.12418
−0.11532
−0.07384
12.12
11.78
11.73
11.64
11.57
11.51
11.21
11.18
10.50
9.54
1aꢀꢀ
2aꢀꢀ
3a
2aꢀꢀ
4aꢀꢀ
5aꢀꢀ
3a
4aꢀꢀꢀ
6a
5aꢀꢀ (HOMO)
9.32
7aꢀ (LUMO)
6aꢀꢀ
8aꢀ
9aꢀ
Orbitals are classified according to Cs symmetry.
Fig. 4. PLATON drawing of compound 16 at the 30% probability level and atom
labeling scheme. Hydrogen atoms are omitted for clarity.
calculational time. The same is true for not using more elabo-
rate methods for treating the near-degeneracies present in these
types of molecules (the original degeneracy of the two HOMO
and three LUMO orbitals must be removed, since the Cs point
group does not allow degenerate representations). The essen-
tial correctness of assignment of the HOMO orbital energies
was proved by calculation of photoelectron spectrum (UPS) for
1 by means of Outer Valence Green’s Function (OVFG) and
using a larger basis set (6–311 + G(2d)) which reasonably fitted
to the experimental spectrum published by Terpstra et al. [24].
It became apparent that a broad valence band at vertical ion-
ization energy of 9.79 eV is due to electron ionization from the
two cyclopentadienyl-based 5aꢀꢀ and 6aꢀ orbitals (calc. values
9.32 and 9.54 eV). A sharp valence band at 10.77 eV is due to
the electron ionization from the 4aꢀꢀ orbital (calc. 10.50 eV) and
the next peak at 11.64 eV at the edge of a very broad unre-
solved valence band takes origin from the 3aꢀꢀ orbital (calc.
value 11.18 eV). The valence bands extending over 12–15 eV
are due to electron ionization from MO orbitals 1aꢀ–5aꢀ, 1aꢀꢀ,
and 2aꢀꢀ (Table 2). The reason for lower values of calculated ion-
ization potentials may be caused by relativistic effects. This is
also the reason that GIAO shifts of 49Ti NMR spectra do reflect
a change between compounds 1 and 6 but they do not distin-
guish substituents R in compounds 12 or 13 from the methyl
group.
Fig. 5. PLATON drawing of compound 20 at the 30% probability level and atom
labeling scheme. Hydrogen atoms are omitted for clarity.
toward the titanium atom by the carbon atom C(5) bearing the
hydrogen atom. This is probably caused by the vicinity of very
bulky cyclohexyl group and its crystal packing effect in the unit
cell. The cyclohexyl group adopts the chair structure, and the
dihedral angle τ of the least-squares plane of its ␤ and ␥ carbon
atoms and the cyclopentadienyl ring plane is 52.73(14)◦. The Si
atom in 16 and the C(9) atom of the cyclohexyl group in 20 are
declined from the least-squares plane of the Cp ring by 0.159(6)
2.4. Crystal structures of CpꢀTiCl3 compounds
˚
Of the investigated compounds X-ray diffraction crystal
structures are known for 1 [28], 5 [29], 7 [30], and 12 [21], and
the crystal structures for 16 and 20 were determined in this work.
The crystal structures of 16 and 20 are drawn in Figs. 4 and 5,
respectively, and the important geometric parameters are listed
in Table 3. The structures are in no respect exceptional having
the least-squares plane of the cyclopentadienyl ring virtually
parallel with the plane of chlorine atoms (angle ϕ). In 20, hav-
ing ϕ 1.15(11)◦ the cyclopentadienyl ring is slightly inclined
and 0.171(6) A farther away from the Ti atom. Such values are
common for the methyl substituents and do not indicate any
steric hindrance of the bulky groups with the Ti–Cl bonds. The
comparison of the main structure parameters for compounds
involved is shown in Table 4. The average Ti–Cl and Ti–C bond
lengths demonstrate that the main elongation of both the bonds
occurs between the non-methylated and tetramethylated com-
pound whereas the introduction of tBu group has only marginal
elongation effect. In none of the crystal structures of Table 4

J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
19
Table 3
any steric hindrance between the cyclopentadienyl ligand and
chlorine atoms was noticed.
◦
˚
Selected bond lengths (A) and angles ( ) for 16 and 20
Compound
16
20
2.5. Polymerization of styrene
Bond lengths
Ti–Cga
Ti–C(1–5)av
Ti–Cl(1)
TiCl(2)
Ti–Cl(3)
Si–C(1)
C(1)–C(9)
C–C(Cp ring)
2.0149(17)
2.349(3)
2.2358(13)
2.2301(12)
2.2292(13)
1.889(3)
2.0087(15)
2.342(3)
2.2378(10)
2.2438(10)
2.2425(10)
–
A series of variously substituted cyclopentadienyltitanium
trichlorides (␩5-C5H5−nMen)TiCl3 (n = 0–5; 1–6) were com-
bined with MAO at molar ratio Al/Ti = 500 and used to catalyze
the polymerization of styrene in toluene at 50 ◦C under standard
conditions. The polymerization results given in Table 5 show
a sharp decrease of catalytic activity for 2/MAO followed by
a moderate decrease for 3/MAO and 4/MAO. The 5/MAO sys-
tem was, however, more active, showing the activity between
those for 2/MAO and 3/MAO. The lowest activity, ca. 20 times
lower than for 1/MAO, was found for 6/MAO. The molecular
weight (Mw) of essentially s-PS polymer was increasing with
decreasing catalyst activity, reaching maximum Mw 254 730 for
6/MAO. The s-PS Mw for 5/MAO was, however, higher than
for the less active 3/MAO catalyst. It was previously shown
that replacement of one methyl group in 6 by an alkyl or aryl
group increased the activity while the s-PS molecular weight
remained high [20]. Therefore, we have carried out polymer-
izations with 8/MAO, 9/MAO, and 11/MAO catalysts and have
found a good reproducibility of activities and s-PS molecular
weights with previous data (Table 5). In contrast to Ref. [19]
all these catalysts gave a higher activity than the 6/MAO sys-
tem. This was previously [20] accounted for the most difficult
reduction of 6 among the studied CpꢀTiCl3 compounds to give a
Ti(III) cationic species which is thought to be the active catalytic
center for the styrene polymerization to s-PS [31]. As a conse-
quence, it was proposed that the activity of 6/MAO catalyst will
grow faster with increasing Al/Ti molar ratio than for the other
persubstituted compounds 7–10 [20]. Indeed, a series of exper-
iments performed at the molar ratio Al/Ti = 1500 showed that
the activity of 6 approached activities of variously penta- and
tetra-substituted compounds (Table 5, activity in parentheses).
Polydispersity of all the s-PS polymers was close to 2, indicat-
–
1.506(4)
1.403–1.425
1.413–1.425(5)
Angles
Cl(1)–Ti–Cl(2)
Cl(2)–Ti–Cl(3)
Cl(1)–Ti–Cl(3)
ϕb
102.28(6)
102.15(4)
103.33(4)
104.04(4)
1.15(11)
102.28(6)
103.57(6)
0.11(11)
–
τc
52.73(14)
a
Cg denotes the centroid of the C(1–5) cyclopentadienyl ring.
Dihedral angle between the least-squares plane of the C(1–5) cyclopentadi-
b
enyl ring and the Cl(1–3) plane.
c
Dihedral angle between the least-squares plane of the C(1–5) cyclopentadi-
enyl ring and the least-squares plane of C(10), C(11), C(13), and C(14) atoms
of the cyclohexyl substituent.
Table 4
Molecular parameters of CpꢀTiCl3 complexes
Complex Cyclopentadienyl ligand
Ti–Clav Ti–Cav Referencesa
1
5
7
12
16
20
C5H5
C5HMe4
C5Me4Et
2.223
2.243
2.243
2.246
2.232
2.314
2.350
2.352
2.361
2.349
2.342
[28]
[29]
[30]
[21]
This work
This work
C5Me4(t-C4H9)
C5H(1,2,3-)Me3(SiMe3)
C5H(1,2,3-)Me3(cyclohexyl) 2.241
a
Structural data retrieved from Cambridge Crystallographic Data Centre.
Table 5
Styrene polymerization catalyzed by the [CpꢀTiCl3]/MAO systemsa
Cpꢀ
Ab × 10−2
Mw × 10−3
D Mw/Mn
c
Tm (◦C)
Cpd
1
2
3
4
5
6
8
9
11
13
15
16
C5H5
C5H4Me
C5H3Me2
C5H2Me3
C5HMe4
C5Me5
C5Me4(butyl)
C5Me4(3-butenyl)
C5Me4Ph
C5Me4(SiMe3)
C5HMe3Ph
C5HMe3(SiMe3)
81.46
12.34
10.20
9.71
11.43 (61.30)
3.95 (31.09)
32.43
23.33
34.42
72.65
102.84
144.92
184.41
168.01
254.73
193.10
181.80
304.90
–
1.9
2.1
1.9
2.1
1.6
1.9
1.9
2.1
1.9
–
255.6
265.4
268.5
267.7
269.6 (268.9)
271.1 (268.7)
255.6
265.4
268.5
(35.21)
(41.44)
(35.57)
(269.3)
(271.9)
(266.2)
–
–
–
–
a
Polymerization conditions: [Ti] = 5.8 ␮M, [styrene] = 4.36 × 10−2 M in toluene, Al/Ti = 500, temperature 50 ◦C, reaction time 1 h; in parentheses: [Ti] = 10.0 ␮M,
[styrene] = 4.54 × 10−2 M in toluene, Al/Ti = 1500, temperature 50 ◦C, reaction time 1 h.
b
Activity: kg of polymer/[mol(Ti) × mol(styrene) × h].
c
Melting point of polystyrene.

20
J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
ing their single site catalyst formation, syndiotacticity ranged
85–93%. Melting points of polymers (typically 265–271 ◦C)
reflected the polymer crystallinity rather than their molecular
weight. The catalysts based on 8, 9, 11, 13, 15, and 16 did not
show a dramatical increase in their performance, and since these
and other CpꢀTiCl3 compounds were not useful for correlation
with their spectroscopic properties they were not used for poly-
merization tests.
the 6/MAO system dramatically increased at the three times
higher Al/Ti molar ratio. The non-methyl substituents R in the
(␩5-C5Me4R)TiCl3 series increased the activity with respect to
6/MAO, however, the opposite effect was observed for the (η5-
C5H(1,2,3-)Me3R)TiCl3 representativeswithrespectto5/MAO.
These systems were not investigated in more detail because the
reproducibility of polymerization experiments should not meet
the precision in determination of e.g. δ shifts.
2.6. Conclusions
3. Experimental
The (␩5-C5H5−nMen)TiCl3 (n = 0–5) compounds 1–6 dis-
play red shifts of their lowest lengths absorption band (λmax) and
downfield shifts of their 49,47Ti NMR chemical shift (δ) propor-
tionally to the number of methyl groups n. The both magnitudes
linearly correlate (constant of proportionality 5.254 ppm/nm)
sincebotharedeterminedbytheenergygapbetweenHOMOand
LUMO orbitals. The DFT calculations proved that the absorp-
tion band is due to transitions from nearly degenerate HOMO
orbitals largely located on cyclopentadienyl ligand to predom-
inantly titanium d-orbitals. The above δ/λmax correlation, how-
ever, does not apply for compounds (␩5-C5Me4R)TiCl3 7–14,
and (␩5-C5H(1,2,3-)Me3R)TiCl3 15–20. The alkyl groups R
induce a slight upfield shift of δ relative to the methyl derivatives
6 and 5 whereas λmax is moved slightly to longer wavelengths.
This discrepancy can be caused by a pronounced asymmetry of
the absorption band extending toward a short wavelengths side.
This moves the center of gravity of the absorption band to blue
side (correlating with δ) while λmax moves very slightly to red
side. The silyl groups induce the blue shift of λmax as well as the
upfield shift of δ which indicates that they are weaker electron
donors than methyl group. The δ upfield shifts for phenyl group
in 11 and 15 correspond qualitatively with its electron with-
drawing property, however, the red shifts of λmax indicate the
conjugation of aromatic phenyl and cyclopentadienyl systems.
This will be treated experimentally as well as by more precise
DFT calculations in our next work.
The styrene polymerizations catalyzed by (1–6)/MAO sys-
tems showed a general decreasing trend of activities with
increasing number of Me groups, the difference between 1/MAO
and 6/MAO being ca. 20-fold. In contrast to linear depen-
dences of spectroscopic magnitudes, a large drop of activity
was found for 2/MAO whereas the 5/MAO system was more
active than 3/MAO and 4/MAO systems. Thus the correlation
of spectroscopic data with the catalytic activity is gross only. A
limited correlation between the catalytic activity and the pure
electronic effect (HOMO–LUMO energy gap) in the catalyst
precursor is not surprising when considering that the catalytic
center differs from the precursor. The catalytic cationic centers
in these systems are usually not accessible to structural inves-
tigations, however, it is generally accepted that steric effects of
the auxiliary cyclopentadienyl ligand participate in the control
of polymer chain growth, chain transfer, and termination reac-
tions [31]. A similar, not clear correlation between δ 49Ti and
polymerization activity in substituted (␩5-indenyl)TiCl3/MAO
catalysts was reported [26]. The activities are further sensitive
to changes in the Al/Ti molar ratio, particularly the activity of
3.1. General procedures
Allreactionswithmoistureandair-sensitivecompoundswere
carried out under argon atmosphere using standard Schlenk tech-
niques, or in vacuum using all-sealed glass devices equipped
with breakable seals. Samples of solid products for melting point
and EI MS measurements were sealed into capillaries and KBr
pellets were pressed in a glovebox Labmaster 130 (mBraun)
under purified nitrogen. 1H (300 MHz) and 13C (75 MHz) NMR
spectra were recorded on a Varian Mercury 300 spectrometer
in CDCl3 solutions at 25 ◦C. Chemical shifts (δ/ppm) are given
relative to the solvent signal (δH 7.15, δC 128.00). 49,47Ti NMR
spectra at 20.32 MHz were measured on a Bruker AMX 360
spectrometer equipped with 5 mm broadband probe at 300 K.
The samples were dissolved in C6D6 and sealed in 5 mm NMR
tubes. The chemical shifts were referred to the 49Ti NMR sig-
nal of external neat TiCl4 (δ(49Ti) = 0.0) in a co-axial capillary
placedinto5 mmNMRtubecontainingC6D6 usingabsoluteres-
onance frequency. Negative values of titanium chemical shifts
denote shifts to lower frequencies. Line broadening of 2–100 Hz
(depending on the signal half-widths) were applied prior Fourier
transformation. This broadening was subtracted when line half-
widths were determined. EI-MS spectra were measured on a
VG-7070E mass spectrometer at 70 eV. The crystalline samples
in sealed capillaries were opened and inserted into the direct
inlet under argon. UV–vis measurements were performed on a
Varian Cary 17 D spectrometer in the range 340–800 nm. Solu-
tions in hexane or toluene were measured in a pair of quartz cells
(0.1 and 1.0 cm) in the absence of air. Infrared spectra of KBr
pellets were recorded in an air-protective cuvette on a Nicolet
Avatar FT-IR spectrometer in the range 400–4000 cm−1. Melt-
ing points were determined in nitrogen-filled sealed capillaries
on a Kofler apparatus and are uncorrected. Syndiotacticity of
polystyrene was determined by the polymer extraction in boil-
ing ethylmethylketone for 8 h and drying in vacuum to constant
weight.
3.2. Chemicals
Solvents hexane, toluene, and m-xylene were refluxed with
LiAlH4, degassed, and stored as solutions of green dimeric
titanocene, [(␮-␩5:␩5-C10H8)(␮-H)2{Ti(␩5-C5H5)}2] [32] on a
vacuum line. Diethyl ether or tetrahydrofuran (THF) were dried
by potassium benzophenone or Na–K alloy, and freshly dis-
tilled prior to use. Phenylbromide, polyphosphoric acid (PPA),
n-butyllithium, dichlorodimethylsilane, and chlorotrimethyl-

J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
21
329 (15), 296 (10), 294 ([M − Cl − CH2CH2CF3]+; 13), 275
([Cpꢀ]+; 7), 197 (11), 178 ([Cpꢀ − CH2CH2CF3]+; 19), 177
(19), 164 (21), 135 (29), 133 (30), 120 (16), 119 (61), 105
silane were obtained from Aldrich and used as received.
Methylalumoxane (10% in toluene (w/v)) was purchased from
Crompton (Germany). TiCl4 (Aldrich) was degassed, refluxed
with copper wire in a sealed ampoule, and distributed by
distillation on a vacuum line. Its 0.1 M solution in toluene
and 0.2 M solution in m-xylene were made by mixing known
volumes in evacuated all-sealed glass distributing devices.
Cyclopentadienyltitanium trichlorides containing Cpꢀ C5H5
(1), C5H4Me (2), C5H3Me2(1,3-), C5H2Me3(1,2,4-) (4),
C5HMe4 (5), C5Me5 (6), and C5Me4Et (7) [9], C5Me4Bu (8),
C5Me4(CH2CH2CH CH2) (9), C5Me4(CHMeCH CH2) (10),
C5Me4Ph (11), [20], and C5Me4tert-Bu (12) [21] were prepared
by synproportionation of the appropriate Cpꢀ2TiCl2 compounds
with TiCl4. This method was also convenient for obtaining of the
(␩5-C5Me4(SiMe2R))TiCl3 compounds where R = Me (13) or
3,3,3-trifluoropropyl (14). The corresponding titanocene dichlo-
rides were prepared previously: [TiCl2{C5Me4(SiMe3)}2] [33]
and [TiCl2{C5Me4(SiMe2CH2CH2CF3)}2] [34]. Precur-
sors of CpꢀTiCl3 compounds containing Cpꢀ = 1-alkyl-2,3,4-
trimethylcyclopentadienyl (17–20), mixtures of (isopropyl-
trimethylcyclopentadienyl)trimethylsilanes, (sec-butyltrimeth-
ylcyclopentadienyl)chlorodimethylsilanes, (tert-butyltrimethy-
lcyclopentadienyl)trimethylsilanes, and chloro(cyclohexyltri-
methylcyclopentadienyl)dimethylsilanes were identical with
those recently reported [35].
1
(38), 91 (29), 81 (22), 77 (79), 59 ([SiMe2H]+; 49). H NMR
(C6D6): 0.22 (s, 6H, SiMe2); 0.76–0.90 (m, 2H, SiCH2);
1.58–1.76 (m, 2H, CH2CF3); 1.82 (s, 6H, ␤-Me, C5Me4);
2.11 (s, 6H, ␣-Me, C5Me4). 13C { H}(C6D6): −0.97 (SiMe2);
1
3
8.57 (q, JFC = 2 Hz, SiCH2); 13.71, 17.24 (C5Me4); 28.82 (q,
2JFC = 30 Hz, CH2CF3);128.09(q, 1JFC = 277 Hz, CF3);136.41,
141.76, 143.96 (Cq, Cipso and C-Me). 19F(C6D6): −68.60 (t,
3JFH = 10.9 Hz, CF3). 28Si { H}(C6D6): −2.86 (SiMe2CH2).
1
IR (KBr, cm−1): 2964 (m), 2950–2907 (w,b), 1476 (w), 1445
(m), 1383 (m), 1367 (m), 1334 (w), 1322 (w), 1261 (vs), 1206
(s), 1127 (s), 1068 (s), 1022 (m), 901 (m), 846 (m), 814 (m), 781
(w), 737 (w), 623 (vw), 549 (vw), 467 (s), 410 (vs).
3.4. Preparation of (η5-1,2,3-trimethylphenyl-
cyclopentadienyl)trichlorotitanium (15)
A hexane solution of n-butyllithium (7.9 ml, 1.6 M,
12.7 mmol) was drop wise added at 0 ◦C to a solution of mix-
ture trimethylphenylcyclopentadienes [36] (2.12 g, 11.5 mmol)
in diethyl ether (60 ml). The reaction mixture was warmed to
room temperature and stirred overnight. The resulting white
precipitate was filtered, washed three times with 20 ml of
diethyl ether and dried in vacuum to obtain 1.98 g of lithium
salt as white powder. The cyclopentadienyl lithium salt was
dissolved in THF (20 ml) and the resulting yellow THF solution
was slowly added to a mixture of excess chlorotrimethylsilane
in THF (15 ml) previously cooled to −78 ◦C. The reaction
mixture was warmed to room temperature and then stirred for
an additional 14 h. The volatiles were removed in vacuum, the
yellow-orange oily residue was dissolved in pentane (30 ml)
and filtered to remove LiCl. The pentane was removed in
vacuum to leave a slightly yellow liquid of (3,4,5-trimethyl-2-
phenylcyclopenta-2,4-dienyl)trimethylsilane.
3.3. Preparation of compounds 13 and 14
A typical synthesis is described for [{␩5-C5Me4(SiMe3)}
TiCl3] (13). Compound [TiCl2{C5Me4(SiMe3)}2] [33] (1.08 g,
2.0 mmol) was degassed, and 10 ml of m-xylene (distilled from
LiAlH4 onavacuumline)wasadded. Then, asolutionofTiCl4 in
m-xylene (0.2 M, 10 ml, 2.0 mmol) was added, and the reaction
ampule was heated to 130 ◦C for 2 h. After cooling to ambient
temperature, all volatiles were evaporated in vacuum at 100 ◦C
and discarded. The residue was extracted repeatedly by 20 ml
of hexane in a closed system until the extract was yellow. The
extracted solid was then dissolved in the hexane at 100 ◦C and
the product crystallized by slow cooling. Orange-yellow crystals
were washed with hexane and dried in vacuum.
Yield 2.57 g (87%). 1H NMR (C6D6): −0.25 (s, 9H,
Si(CH3)3); 1.82–1.86 (m, 3H, C(3)CH3); 1.97 (s, 3H, C(5)CH3);
5
1.99 (d, JHH = 1.8 Hz, 3H, C(4)CH3); 3.39 (br s, 1H, CH);
1
7.05–7.10(m, 1H, Ph);7.18–7.22(m, 4H, Ph). 13C{ H}(C6D6):
13: Yield 1.22 g (88%). m.p. 164 ◦C. EI-MS (direct inlet,
110 ◦C; m/z (%)): 348 (6), 346 (M•+; 6), 336 (9), 335 (40), 334
(26), 333 (98), 332 (33), 331 ([M − Me]+, 100), 330 (11), 329
(10), 315 (6), 312 (5), 310 ([M − HCl]+, 6), 298 (7), 297 (7), 296
([M − Me − Cl]+, 9), 295 ([M − Me − HCl]+, 8), 194 (7), 193
([Cpꢀ]+; 33), 178 (16), 177 (15), 163 (14), 133 (26), 119 (15),
105 (10), 73 (45), 59 (38). IR (KBr, cm−1): 2963 (m), 2919)
(w), 2854 (vw), 1477 (m), 1445 (w), 1409 (w), 1379 (m), 1335
(m), 1250 (s), 1128 (w), 1091 (w), 1021 (m), 845 (vs), 758 (w),
696 (vw), 636 (vw), 466 (s), 409 (vs).
Compound (14) was made analogously from
[TiCl2{C5Me4(SiMe2CH2CH2CF3)}2] [34] (1.34 g, 2.0 mmol).
14: Yield 1.34 g (78%). m.p. 102 ◦C. EI-MS (direct inlet,
90 ◦C; m/z (%)): 430 (4), 428 (M•+; 4), 395 (5), 393
([M − Cl]+; 5), 379 (6), 377 ([M − HCl − Me]+; 7), 359 (4),
357 ([M − Cl − HCl]+; 4), 336 (13), 335 (15), 334 (62), 333
(40), 332 (100), 331 ([M − CH2CH2CF3]+; 100), 330 (17),
−2.00 (Si(CH3)3); 11.27, 12.65, 14.89 (3× CH3); 54.16 (CH);
125.87, 128.30, 129.45 (CH, Ph); 136.12, 136.39, 136.87,
138.44, 139.32 (Cq; C-CH3, C-Ph and Ph). GC–MS (m/z (%)):
257 (23), 256 (M•+; 100), 241 ([M − Me]+; 10), 183 (21),
182 ([M − SiMe3H]+; 86), 167 ([M − SiMe3H − Me]+; 26), 165
(17), 152 (9), 141 (5), 135 (4), 128 (4), 115 (6), 91 (3), 73
([SiMe3]+; 85).
The silane (2.56 g, 10.0 mmol) in 30 ml CH2Cl2 was cooled
to 0 ◦C and TiCl4 solution (10.0 ml, 1 M, 10.0 mmol) in CH2Cl2
was added in one portion. The mixture was allowed to warm to
room temperature and then stirred for 14 h. The resulting dark
red mixture was transferred into a sublimator, volatiles were
removed and the solid residue was sublimed in vacuum. Pure
product sublimed at 120–125 ◦C/4 × 10−5 mbar as a red solid.
Yield 1.79 g (53%). m.p. 70 ◦C. EI-MS (direct inlet, 80 ◦C;
m/z (%)): 338 (18), 336 (M•+; 19), 301 ([M − Cl]+; 3), 184
(17), 183 ([C5HMe3Ph]+; 100), 168 (11), 167 (14), 165 (16),

22
J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
1
1
153 (10), 152 (11). H NMR (C6D6): 1.83, 1.96, 2.16 (3 s,
(16). H NMR (C6D6): 0.25 (s, 9 H, SiMe3), 1.75, 1.99, 2.18
3
1
3 3H, 3 CCH3); 6.50 (s, 1H, CpH); 7.04 (tt, JHH = 7.2 Hz,
(3× s, 3× 3H, C5Me3); 6.53 (s, 1 H, C5Me3H). 13C { H} NMR
4JHH = 1.5 Hz, 1H, Ph). 7.09–7.16 (m, 2H, Ph); 7.40–7.45 (m,
(C6D6): −0.5 (SiMe3), 12.8, 15.9, 17.0 (C5Me3); 129.5 (CH,
C5Me3H), 141.6, 142.3, 142.5, 144.6 (Cq; C-Me and Cipso). IR
(KBr, cm−1): 3080 (vw), 2959 (m), 2918 (w), 2851 (vw), 1477
(w), 1445 (w), 1404 (w), 1381 (m), 1348 (w), 1287 (vw), 1267
(w), 1252 (vs), 1150 (m), 1139 (m), 1026 (m), 931 (vw), 872
(m), 842 (vs), 753 (m), 697 (vw), 629 (w), 498 (vw), 464 (s),
410 (vs).
2H, Ph). 13C { H}(C6D6): 13.62, 15.20, 16.04 (3 CH3); 122.06
1
(CH, Cp); 128.74, 128.99, 129.10 (CH, Ph); 134.33, 134.86,
138.33, 139.05, 141.11 (Cq; C-CH3, C-Ph and Ph). IR (KBr,
cm−1): 3106 (w), 3053 (vw), 2986 (vw), 2963 (vw), 2919 (w),
2852 (vw), 1957 (vw), 1890 (vw), 1815 (vw), 1752 (vw), 1600
(vw), 1576 (w), 1484 (s), 1451 (m), 1381 (s), 1205 (vw), 1183
(w),1159 (w), 1127 (w), 1076 (w), 1021 (m), 1003 (w), 923
(vw), 877 (m), 772 (s), 764 (s), 699 (s), 659 (w), 643 (m), 617
(vw), 560 (w), 544 (vw), 503 (vw), 458 (vs), 412 (vs).
3.6. Preparation of (η5-1-isopropyl-2,3,4-
trimethylcyclopentadienyl)titanium trichloride (17)
3.5. Preparation of (η5-1,2,3-trimethyl(trimethylsilyl)-
cyclopentadienyl)trichlorotitanium (16)
A
mixture of (1-isopropyl-2,3,4-trimethylcyclopenta-
dienyl)trimethylsilanes [35] (0.75 g, 3.38 mmol) was degassed
on a vacuum line, and a solution of TiCl4 in toluene (0.1 M,
35 ml) was added under stirring. The mixture was heated to
90 ◦C for 30 min, cooled to room temperature, and all volatiles
including unreacted TiCl4 were distilled off in vacuum. The
residue was extracted repeatedly by 30 ml of hexane in a sealed
system. The extracted solid was dissolved in the mother liquor
by warming to 100 ◦C and crystallized by slow cooling. Yield
of orange crystals was 0.76 g (74%).
A mixture of 1,2,3-trimethylcyclopentadiene isomers [19]
(6.0 g, 55.5 mmol) in diethyl ether (100 ml) was mixed with
BuLi (1.6 M in hexane, 35 ml) and stirred overnight. Then, a
solution of chlorotrimethylsilane (6.2 g, 57 mmol) in 50 ml of
diethyl ether was added, and the mixture was refluxed for 2 h. At
room temperature, BuLi (1.6 M in hexane, 37 ml) was added and
the mixture stirred for 8 h. Then, a solution of chlorotrimethyl-
silane (7.0 g, 65 mmol) in 50 ml of diethyl ether was added,
the mixture was refluxed for 4 h, and then, the solvents were
distilled from a water bath (90 ◦C). The fraction distilling at
5 Torr and a lower pressure was collected. Yield of 1,2,3-
trimethyl-5,5-bis(trimethylsilyl)cyclopenta-1,3-diene was 9.1 g
(65%).
17: m.p. 111 ◦C. EI-MS (direct inlet, 70 ◦C; m/z (%)):
304 (14), 302 (M•+; 15), 269 (17), 268 (53), 267 (28), 266
([M − HCl]+; 70), 265 (11), 232 (39), 231 (24), 230 ([M − 2
HCl]+; 97), 150 (17), 149 ([C5HMe3(iPr)]+; 100), 135 (14),
134 (65), 133 (55), 119 (51), 115 (12), 105 (25), 91 (36), 79
1
(13), 77 (19), 41 (28). H NMR (C6D6): 0.74, 1.17 (2× d,
3
GC–MS (m/z, %): 252 (M•+; 57), 237 ([M − Me]+; 40),
179 (4), 178 ([M − Me − HSiMe2]+; 10), 165 (16), 164
([HC5Me3SiMe2]+; 100), 149 (22), 105 (5), 73 ([SiMe3]+; 70),
2× JHH = 7.0 Hz, 2× 3H, CHMe2); 1.89 (s, 6H, C5Me3); 1.92
3
(s, 3H, C5Me3); 3.07 (septuplet, JHH = 7.0 Hz, 1H, CHMe2);
6.28 (s, 1H, C5Me3H). 13C { H} (C6D6): 13.49, 13.80, 15.80
1
1
59 ([HSiMe2]+; 13), 45 (17). H NMR (CDCl3): 0.00 (s, 18
(C5Me3); 20.33, 23.88 (CHMe2); 29.66 (CHMe2); 119.26 (CH,
C5Me3); 136.39, 137.01, 139.47, 149.68 (Cq; C-Me and Cipso).
IR (KBr, cm−1): 3096 (vw), 2967 (s), 2925 (m), 2873 (w), 1498
(m), 1470 (m), 1447 (m), 1416 (w), 1383 (m), 1364 (m), 1349
(w), 1311 (vw), 1186 (w), 1144 (vw), 1098 (vw), 1070 (vw),
1047 (vw), 1021 (w), 880 (vw), 862 (m), 762 (m,b), 673 (vw),
617 (vw), 496 (w), 457 (s), 411 (vs).
H, SiMe3); 1.81 (bs, 3 H), 1.94 (d, J = 1.3 Hz, 3 H), 1.97 (s, 3
1
H) (C5Me3); 5.83 (s, 1 H, C5Me3H). 13C{ H} NMR (CDCl3):
δ −0.7 (SiMe3), 10.7, 14.0, 14.8 (C5Me3); 53.6 (C(SiMe3)2),
135.7, 138.9, 141.5 (Cq; C-Me and Cipso). IR (neat): 3047 (w),
2947 (s), 2900 (s), 2853 (s), 1440 (m), 1400 (w), 1377 (m), 1243
(vs), 1230 (m), 1117 (m), 1089 (m), 1047 (m), 980 (w), 957 (s),
927 (m), 870 (vs), 833 (vs), 777 (s), 753 (s), 680 (s), 667 (m),
640 (w), 627 (m), 615 (m), 537 (w), 498 (m).
3.7. Preparation of (1-sec-butyl-2,3,4-
TiCl4 (1.3 ml 14.5 mmol) was distilled in vacuum onto a
frozen solution of C5Me3(SiMe3)2H (3.78 g, 15.0 mmol) in
40 ml of toluene cooled by liquid nitrogen. The mixture was
then slowly warmed up with vigorous shaking. After subsequent
warming to 80 ◦C all volatiles were evaporated in vacuum and
the residue was extracted by 60 ml of hexane. The orange-yellow
solution was concentrated, and cooled in a refrigerator. Orange
crystals obtained overnight were separated, washed with hexane,
and dried in vacuum.
trimethylcyclopentadienyl)titanium trichloride (18)
Analogously to the synthesis of 16, a mixture of (1-sec-
butyl-2,3,4-trimethylcyclopentadienyl)chlorodimethylsilanes
[35] (0.85 g, 3.3 mmol) was reacted with the toluene solution of
TiCl4 (0.1 M, 35 ml). Orange crystalline material consisted of
two diastereomers (I and II) due to a chiral center in the sec-butyl
substituent and a planar chirality of the cyclopentadienyl ring
1
as indicated by two sets of resonances in H, 13C, and 49,47Ti
16: Yield 3.0 g (63%). m.p. 98 ◦C. EI-MS (direct inlet, 70 ◦C;
m/z (%)): 332 (M•+; 0.7), 323 (7), 322 (8), 321 (39), 320 (24),
319 (98), 317 ([M − Me]+; 100), 316 (10), 315 (10), 302 (6), 284
(7), 283 (9), 282 ([M − Me − Cl]+; 11), 281 ([M − Me − HCl]+;
11), 179 ([C5HMe3SiMe3]+; 15), 164 ([C5HMe3SiMe2]+; 12),
163 (9), 149 (11), 119 (12), 105 (10), 97 (6), 93 (6), 91 (7), 83
(8), 73 ([SiMe3]+; 35), 59 ([SiHMe2]+; 40), 58 (6), 45 (11), 43
NMR spectra. Yield 0.81 g (78%).
18: m.p. 92 ◦C. EI-MS (direct inlet, 80 ◦C; m/z (%)): 318 (9),
316 (M•+; 10), 283 (10), 282 (29), 281 (17), 280 ([M − HCl]+;
41), 246 (36), 245 (21), 244 ([M − 2HCl]+; 85), 243 (13), 165
(18), 164 (10), 163 ([C5HMe3(sBu)]+; 48), 135 (100), 134 (30),
133 (29), 121 (21), 119 (65), 107 (23), 105 (32), 91 (42), 43
(16), 41 (28). IR (KBr, cm−1): 3091 (vw), 2965 (s), 2924 (m),

J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
23
(12), 272 (54), 271 (35), 270 ([M − 2HCl]+; 100), 269 (18), 189
([C5HMe3(cyclohexyl)]+; 33), 134 (11), 133 (73), 121 (22), 119
(17), 117 (14), 115 (11), 109 (10), 107 (21), 105 (24), 91 (35), 81
(24), 79 (13), 77 (15), 55 (15), 53 (11), 41 (38). 1H NMR (C6D6):
0.60–0.80 (m, 1H, cyclohexyl); 0.94–1.32 (m, 4H, cyclohexyl);
1.42–1.72 (m, 4H, cyclohexyl); 1.90, 1.92, 1.93 (3× s, 3×
3H, C5Me3); 2.02–2.22 (m, 1H, cyclohexyl); 2.76–2.92 (m,
2876 (w), 1494 (w), 1455 (m), 1414 (vw), 1381 (m), 1349 (vw),
1260 (vw), 1182 (vw), 1019 (m), 877 (m), 764 (w,b), 673 (vw),
617 (vw), 497 (w), 458 (s), 414 (vs).
3.7.1. I (56%)
1H NMR (C6D6): 0.56 (t, JHH = 7.5 Hz, 3H, CH2Me);
3
0.96–1.28 (m, 2H, CH2Me); 1.18 (d, 3JHH = 6.5 Hz, 3H, CHMe);
1.88, 1.90, 1.93 (3× s, 3× 3H, C5Me3); 2.90–3.02 (m, 1H,
1
1H, C5Me3CH, cyclohexyl); 6.31 (s, 1H, C5Me3H). 13C { H}
CHMe); 6.30 (s, 1H, C5Me3H). 13C { H} (C6D6): 11.01
1
(C6D6): 13.48, 13.83, 15.80 (C5Me3); 26.15, 26.75, 30.64,
35.06 (CH2, cyclohexyl); 39.69 (CH, cyclohexyl); 119.66
(CH, C5Me3); 136.39, 136.78, 139.62, 148.84 (Cq; C-Me and
Cipso). IR (KBr, cm−1): 3093 (w), 2981 (vw), 2926 (vs), 2851
(s), 1496 (m), 1475 (w), 1447 (s), 1382 (m), 1375 (m), 1270
(vw), 1176 (vw), 1135 (vw), 1021 (m), 1000 (w), 891 (w),
762 (vw), 680 (vw), 617 (vw), 532 (vw), 490 (m), 457 (s),
409 (vs).
(CH2Me); 13.70, 13.87, 15.74 (C5Me3); 17.30 (CHMe); 31.23
(CH2Me);35.62(CHMe);119.65(CH, C5Me3);136.45, 136.72,
139.45, 149.04 (Cq; C-Me and Cipso).
3.7.2. II (44%)
1H NMR (C6D6): 0.77 (d, 3JHH = 7.0 Hz, 3H, CHMe); 0.87 (t,
3JHH = 7.5 Hz, 3H, CH2Me); 0.96–1.28 (m, 2H, CH2Me); 1.88,
1.89, 1.92 (3× s, 3× 3H, C5Me3); 2.79–2.91 (m, 1H, CHMe);
1
6.27 (s, 1H, C5Me3H). 13C { H} (C6D6): 12.13 (CH2Me);
3.10. Styrene polymerization
13.56, 13.80, 15.79 (C5Me3); 20.14 (CHMe); 27.61 (CH2Me);
36.49 (CHMe); 119.98 (CH, C5Me3); 136.26, 136.82, 139.49,
149.83 (Cq; C-Me and Cipso).
The polymerizations were carried out at 50 ◦C in a 70 ml
glass reactor. The reactor was equipped with a magnetic
stirring bar and connected to a nitrogen/high vacuum line. After
evacuation for 2 h, the reactor was filled with purified nitrogen
and charged by syringes in nitrogen stream subsequently with
purified toluene (30 ml), styrene (5.2 ml), the MAO solution
in toluene and a toluene solution of a titanium complex.
The mixture was then heated to 50 ◦C and the temperature
was kept constant during the polymerization (60 min). The
polymerization was quenched by addition of a mixture of
aqueous HCl (5 ml of 35% HCl) and methanol (50 ml). The
precipitated polymer was filtered off, washed with methanol
and dried in vacuum at 80 ◦C to constant weight. Syn-
diotacticity was determined by polymer extraction in boiling
2-butanone for 8 h followed by drying of the residue till constant
weight.
3.8. Preparation of (1-tert-butyl-2,3,4-
trimethylcyclopentadienyl)titanium trichloride (19)
Compound 19 was obtained from a mixture of isomers of
(1-tert-butyl-2,3,4-trimethylcyclopentadienyl)trimethylsilane
[35] (0.78 g, 3.3 mmol) reacted with the toluene solution of
TiCl4 (0.1 M, 35 ml). Orange crystalline product was obtained
from hexane. Yield 0.72 g (69%).
19: m.p. 98 ◦C. EI-MS (direct inlet, 80 ◦C; m/z (%)): 320
(12), 318 (31), 316 (M•+; 32), 305 (14), 303 (34), 302 (10),
301 ([M − Me]+, 37), 283 (10), 282 (22), 281 (14), 280
([M − HCl]+; 30), 269 (16), 268 (15), 267 (76), 266 (25), 265
([M − Me − HCl]+; 100), 246 (7), 244 ([M − 2 HCl]+; 15),
230 (12), 229 (8), 227 (13), 164 (9), 163 ([C5HMe3(tBu)]+;
69), 149 (7), 148 ([C5HMe2(tBu)]+; 41), 147 (15), 133 (23),
3.11. X-ray crystal structure of compounds 16 and 20
1
91 (11), 41 (10). H NMR (C6D6): 1.17 (s, 9H, CMe3); 1.74,
1.99, 2.18 (3× s, 3× 3H, C5Me3); 6.34 (s, 1H, C5Me3H). 13
C
Crystal fragments of 16 and 20 were inserted into a
Lindemann glass capillary under nitrogen atmosphere in a
glovebox and the capillary was sealed with wax. Diffraction
data were collected on an Enraf-Nonius CAD 4-MACH III
diffractometer using graphite monochromated Mo K␣ radiation
1
{ H} (C6D6): 13.49, 16.38, 16.44 (CpMe3); 30.11 (CMe3);
35.17 (CMe3); 121.24 (CH, C5Me3H); 134.58, 138.13, 140.26,
153.78 (Cq; C-Me and Cipso). IR (KBr, cm−1): 3113 (vw), 2970
(s), 2918 (m), 2873 (w), 1496 (w), 1479 (m), 1463 (m), 1370
(m), 1231 (m), 1018 (m), 869 (m), 756 (w,b), 674 (vw), 610
(vw), 495 (w), 456 (s), 407 (vs).
˚
(λ = 0.71073 A) for 16 and on a Nonius KappaCCD diffractome-
ter with CCD area detector for 20, both at ambient temperature.
The structures were solved by direct methods (SIR-92) [37] and
refined by full-matrix least squares on F2 (SHELXL-97) [38].
Crystallographic data, details on data collection and the structure
refinement are given in Table 6.
3.9. Preparation of (1-cyclohexyl-2,3,4-
trimethylcyclopentadienyl)titanium trichloride (20)
Following the previous receipts, a mixture of chloro(1-
cyclohexyl-2,3,4-trimethylcyclopentadienyl)dimethylsilanes
[35] (0.93 g, 3.3 mmol) was reacted with 35 ml of 0.1 M
toluene solution of TiCl4. Orange finely crystalline product was
obtained after workup and crystallization from hexane. Yield
0.82 g (73%).
4. Supplementary material
Relevant crystallographic data have been deposited with the
Cambridge Crystallographic Data Centre, CCDC 296372 for
compound 16 and CCDC 296371 for 20. Copies of the data
can be obtained free of charge on application to CCDC, e-mail:
deposit@ccdc.cam.ac.uk.
20: m.p. 108 ◦C. EI-MS (direct inlet, 80 ◦C; m/z (%)): 344
(9), 342 (M•+; 9), 308 (35), 307 (19), 306 ([M − HCl]+; 47), 273

24
J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
Table 6
[8] (a) R. Benn, A. Rufin´ska, J. Organomet. Chem. 273 (1984) C51;
(b) A.R. Siedle, R.A. Newmark, W.B. Gleason, W.M. Lamanna,
Organometallics 9 (1990) 1290;
Crystallographic data, data collection and structure refinement data for com-
pounds 16 and 20
(c) M. Bu¨hl, G. Hopp, W. von Philipsborn, S. Beck, M.H. Prosenc, U. Rief,
H.H. Brintzinger, Organometallics 15 (1996) 778.
Compound
[9] J. Langmaier, Z. Samec, V. Varga, M. Hora´cˇek, K. Mach, J. Organomet.
Chem. 579 (1999) 348.
[10] (a) K. Mach, J.B. Raynor, J. Chem. Soc. Dalton Trans. (1992) 683;
(b) J. Hiller, U. Thewalt, M. Pola´sˇek, L. Petrusova´, V. Varga, P. Sedmera,
K. Mach, Organometallics 15 (1996) 3752.
16
20
Chemical formula
Molecular weight
Crystal system
Space group
C
11H19Cl3SiTi
C14H21Cl3Ti
343.56
333.60
Monoclinic
P21/n (No. 14)
6.8139(9)
24.054(3)
9.7860(16)
90
97.921(13)
90
1588.7(4), 4
1.395
Monoclinic
P21/c (No. 14)
6.9420(5)
11.5560(6)
20.1620(16)
90
94.214(3)
90
1613.06(19), 4
1.415
[11] K. Mach, V. Varga, H. Antropiusova´, J. Pola´cˇek, J. Organomet. Chem. 333
(1987) 205.
˚
a (A)
˚
b (A)
[12] (a) A. Hafner, J. Okuda, Organometallics 12 (1993) 949;
(b) M. Erben, A. Ru˚zˇicˇka, M. Picka, I. Pavl´ık, Magn. Reson. Chem. 42
(2004) 414.
˚
c (A)
α (◦)
β (◦)
γ (◦)
ˇ
[13] R. Skeˇril, P. Sindela´ˇr, Z. Salajka, V. Varga, I. C´ısaˇrova´, J. Pinkas, M.
Hora´cˇek, K. Mach, J. Mol. Catal. A 224 (2004) 97.
3
˚
V (A ), Z
[14] O.K.B. Staal, D.J. Beetstra, A.P. Jekel, B. Hessen, J.H. Teuben, P.
dcalc (g cm−3
)
ˇ
Steˇpnicˇka, R. Gyepes, M. Hora´cˇek, J. Pinkas, K. Mach, Collect. Czech.
µ (Mo K␣) (mm−1
F(000)
)
1.093
688
1.009
712
Chem. Commun. 68 (2003) 1119.
[15] (a) N. Ishihara, T. Seimiya, M. Kuramoto, M. Uoi, Macromolecules 19
Crystal size (mm3)
T (K)
0.5 × 0.25 × 0.2
0.25 × 0.18 × 0.15
(1986) 2464;
293(2)
293(2)
(b) N. Ishihara, M. Kuramoto, M. Uoi, Macromolecules 21 (1988)
3356.
θ range (◦)
2.27–24.95
0/8; 0/28; −11/11
2769
1967
151
3.43–27.47
−8/8; −11/14; −26/26
5802
3641
166
h k l range
[16] (a) N. Tomotsu, N. Ishihara, T.H. Newman, M.T. Malanga, J. Mol. Catal.
Diffractions collected
Unique diffractions
Parameters
A 128 (1998) 167;
(b) K. Yokota, T. Inoue, S. Naganuma, H. Shozaki, N. Tomotsu,
M. Kuramoto, N. Ishihara, in: W. Kaminski (Ed.), Metaloorganic
R, wR [I > 2σ(I)]
R, wR (all data)
S
0.0389, 0.0952
0.0795, 0.1126
1.028
0.634, −0.266
0.0494, 0.0933
0.1049, 0.1126
1.037
0.353, −0.378
Catalysts for Synthesis and Polymerization, Springer, Berlin, 1999,
p. 435;
(c) J. Schellenberg, N. Tomotsu, Prog. Polym. Sci. 27 (2002)
1925;
−3
˚
ꢀρmax,min (eA
)
(d) Y. Qian, J. Huang, M.D. Bala, B. Lian, H. Zhang, H. Zhang, Chem.
Rev. 103 (2003) 2633.
[17] P. Longo, A. Proto, A. Zambelli, Macromol. Chem. Phys. 196 (1995)
3015.
Acknowledgement
[18] G. Xu, E. Ruckenstein, J. Polym. Sci., Part A: Polym. Chem. 37 (1999)
2481.
This work was supported by the Ministry of Industry and
Trade of the Czech Republic (project No. FT-TA3/078).
[19] W. Kaminsky, S. Lenk, V. Scholz, H.W. Roesky, A. Herzog, Macro-
molecules 30 (1997) 7647.
[20] J. Zema´nek, L. Fro¨hlichova´, P. Sindela´ˇr, P. Steˇpnicˇka, I. C´ısaˇrova´, V. Varga,
M. Hora´cˇek, K. Mach, Collect. Czech. Chem. Commun. 66 (2001) 1359.
[21] L. Lukesˇova´, R. Gyepes, J. Pinkas, M. Hora´cˇek, J. Kubisˇta, J. Cejka, K.
Mach, Collect. Czech. Chem. Commun. 70 (2005) 1589.
[22] R.D. Gorsich, J. Am. Chem. Soc. 82 (1960) 4211.
[23] G.H. Llinas, M. Mena, F. Palacios, P. Royo, R. Serrano, J. Organomet.
Chem. 340 (1988) 37.
[24] A. Terpstra, J.N. Louwen, A. Oskam, J.H. Teuben, J. Organomet. Chem.
260 (1984) 207.
ˇ
ˇ
ˇ
References
[1] (a) H.-H. Brintzinger, D. Fischer, R. Mu¨lhaupt, B. Rieger, R.M. Waymouth,
Angew. Chem., Int. Ed. Engl. 34 (1995) 1143;
(b) P.C. Mo¨hring, N.J. Coville, J. Organomet. Chem. 479 (1994) 1;
(c) M. Bochmann, J. Chem. Soc. Dalton Trans. (1996) 255.
[2] P.G. Gassman, D.W. Macomber, J.W. Hershberger, Organometallics 2
(1983) 1470.
[3] T. Vondra´k, K. Mach, V. Varga, J. Organomet. Chem. 367 (1989) 69.
[4] T. Vondra´k, K. Mach, V. Varga, Organometallics 11 (1992) 2030.
[5] (a) R.G. Kidd, R.W. Mathews, H.G. Spinney, J. Am. Chem. Soc. 94 (1972)
6686;
[25] (a) J. Mason, Polyhedron 8 (1989) 1657;
(b) J.J. Dechter, Prog. Inorg. Chem. 33 (1985) 393.
[26] Y. Kim, B.H. Koo, Y. Do, J. Organomet. Chem. 527 (1997) 155.
[27] Gaussian 98, Revision A.7. M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E.
Scuseria, M.A. Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery,
Jr., R.E. Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels,
K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R.
Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A.
Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D. Rabuck, K.
Raghavachari, J.B. Foresman, J. Cioslowski, J.V. Ortiz, A.G. Baboul, B.B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L.
Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara,
C. Gonzalez, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W.
Wong, J.L. Andres, C. Gonzalez, M. Head-Gordon, E.S. Replogle, J.A.
Pople, Gaussian, Inc., Pittsburgh, PA, 1998.
(b) N. Hao, B.G. Sayer, G. Denes, D.G. Bickley, M.J. McGlinchey, J. Magn.
Reson. 50 (1982) 50;
(c) A. Dormond, M. Fauconet, J.C. Leblanc, C. Moise, Polyhedron 3 (1984)
897.
[6] (a) P.G. Gassman, P.H. Campbell, D.W. Macomber, Organometallics 3
(1984) 385;
(b) W.C. Finch, E.V. Anslyn, R.H. Grubbs, J. Am. Chem. Soc. 110 (1988)
2406;
(c) M.J. Sarsfield, S.W. Ewart, T.L. Tremblay, A.W. Roszak, M.C. Baird,
J. Chem. Soc., Faraday Trans. (1997) 3097.
[28] L.M. Engelhardt, R.I. Papasergio, C.L. Raston, A.H. White,
Organometallics 3 (1984) 18.
[29] W.J. Evans, D.G. Giarikos, J.W. Ziller, J. Organomet. Chem. 688 (2003)
200.
[7] (a) A. Hafner, R.O. Duthaler, R. Marti, G. Rihs, P. Rothe-Streit, F.
Schwarzenbach, J. Am. Chem. Soc. 114 (1992) 2321;
(b) S. Berger, W. Bock, C.F. Marth, B. Raguse, M.T. Reetz, Magn. Reson.
Chem. 28 (1990) 559.

J. Pinkas et al. / Journal of Molecular Catalysis A: Chemical 257 (2006) 14–25
25
[30] N.W. Alcock, G.E. Toogood, M.G.H. Wallbridge, Acta Crystallogr., Sec.C
[33] M. Hora´cˇek, R. Gyepes, I. C´ısaˇrova´, M. Pola´sˇek, V. Varga, K. Mach, Col-
lect. Czech. Chem. Commun. 61 (1996) 1307.
[34] L. Lukesˇova´, M. Hora´cˇek, R. Gyepes, I. C´ısaˇrova´, P. Steˇpnicˇka, J. Kubisˇta,
40 (1984) 598.
ˇ
[31] (a) J.C.W. Chien, Z. Salajka, S. Dong, Macromolecules 25 (1992)
3199;
K. Mach, Collect. Czech. Chem. Commun. 70 (2005) 11.
[35] J. Pinkas, J. Kubisˇta, R. Gyepes, I. C´ısaˇrova´, N. Pirio, Ph. Meunier, K.
Mach, J. Organomet. Chem. 690 (2005) 731.
(b) G. Xu, S. Lin, Macromolecules 30 (1997) 685;
(c) T.E. Ready, R. Gurge, J.C.W. Chien, M.D. Rausch, Organometallics 17
(1998) 5236;
ˇ
[36] J. Zema´nek, P. Steˇpnicˇka, K. Fejfarova´, R. Gyepes, I. C´ısaˇrova´, M. Hora´cˇek,
(d) A. Grassi, C. Lamberti, A. Zambelli, I. Mingozzi, Macromolecules 30
(1997) 1884;
J. Kubisˇta, V. Varga, K. Mach, Collect. Czech. Chem. Commun. 66 (2001)
605.
(d) E.F. Williams, M.C. Murray, M.C. Baird, Macromolecules 33 (2000)
261.
[32] H. Antropiusova´, A. Dosedlova´, V. Hanusˇ, K. Mach, Transit. Met. Chem.
(London) 6 (1981) 90.
[37] A. Altomare, M.C. Burla, M. Camalli, G. Cascarano, C. Giacovazzo, A.
Guagliardi, G. Polidori, J. Appl. Crystallogr. 27 (1994) 435.
[38] G.M. Sheldrick, SHELXL-97. Program for Crystal Structure Refinement
from Diffraction Data, University of Go¨ttingen, Go¨ttingen, 1997.