Polymerization of ethene with zirconocene catalysts: An experimental and quantum chemical study of the influence of para-substituent in benzyl in bis{η5-(1-benzyl)indenyl}zirconium dichlorides

Article abstract of DOI:10.1016/j.jorganchem.2004.09.089

The meso- and rac-like isomers of bis{η⁵-(1-benzyl)indenyl} zirconium dichloride (5), bis{η⁵-(1-para-methoxybenzyl)indenyl} zirconium dichloride (6), bis{η⁵-(1-para-fluoro-benzyl)indenyl} zirconium dichloride (7) and bis{η⁵-(1-phenylethyl)indenyl} zirconium dichloride (8) were synthesized and isolated. Solid-state structures of meso- and rac-like 5 were determined by X-ray structure analysis. Polymerization properties of the methylaluminoxane (MAO) activated diastereomers of complexes 5-8 were studied in ethene polymerizations under different monomer concentrations. The rac-like isomer of 1-phenylethyl-substituted 8/MAO showed significantly higher activity than the 1-benzyl substituted analogs 5-7/MAO. In addition, rac-8/MAO behaves like a single center catalyst producing polyethene with narrow molar mass distribution (1.8-1.9), while diastereomers of 5-7/MAO produce polymers with molar mass distributions varying from 2.7 up to 10.3. The rac and meso-like isomers of 5-7/MAO have different response on the monomer concentration. Quantum chemical calculations suggest a strong interaction between the benzyl substituent and the electron deficient zirconium center. The phenyl metal coordination energies depend on the electronic properties of the para-substituent. In 8/MAO, due to the ethyl spacer, the coordination does not have a significant role and therefore much higher activity and single center polymerization behavior is observed.

Full text of DOI:10.1016/j.jorganchem.2004.09.089

Journal of Organometallic Chemistry 690 (2005) 773–783
www.elsevier.com/locate/jorganchem
Polymerization of ethene with zirconocene catalysts: an experimental
and quantum chemical study of the influence of para-substituent
in benzyl in bis{g5-(1-benzyl)indenyl}zirconium dichlorides
1
Erkki Aitola , Marina Surakka , Timo Repo , Mikko Linnolahti ,
Kristian Lappalainen, Kaisa Kervinen, Martti Klinga, Tapani Pakkanen ,
Markku Leskela¨
2
*
*
2
Laboratory of Inorganic Chemistry, Department of Chemistry, University of Helsinki, P.O. Box 55, FIN-00014 Helsinki, Finland
Received 17 February 2004; accepted 6 September 2004
Available online 8 December 2004
Abstract
The meso- and rac-like isomers of bis{g5-(1-benzyl)indenyl}zirconium dichloride (5), bis{g5-(1-para-methoxybenzyl)indenyl}zir-
conium dichloride (6), bis{g5-(1-para-fluoro-benzyl)indenyl}zirconium dichloride (7) and bis{g5-(1-phenylethyl)indenyl}zirconium
dichloride (8) were synthesized and isolated. Solid-state structures of meso- and rac-like 5 were determined by X-ray structure anal-
ysis. Polymerization properties of the methylaluminoxane (MAO) activated diastereomers of complexes 5–8 were studied in ethene
polymerizations under different monomer concentrations. The rac-like isomer of 1-phenylethyl-substituted 8/MAO showed signif-
icantly higher activity than the 1-benzyl substituted analogs 5–7/MAO. In addition, rac-8/MAO behaves like a single center catalyst
producing polyethene with narrow molar mass distribution (1.8–1.9), while diastereomers of 5–7/MAO produce polymers with
molar mass distributions varying from 2.7 up to 10.3. The rac and meso-like isomers of 5–7/MAO have different response on the
monomer concentration. Quantum chemical calculations suggest a strong interaction between the benzyl substituent and the elec-
tron deficient zirconium center. The phenyl metal coordination energies depend on the electronic properties of the para-substituent.
In 8/MAO, due to the ethyl spacer, the coordination does not have a significant role and therefore much higher activity and single
center polymerization behavior is observed.
ꢀ 2004 Elsevier B.V. All rights reserved.
1. Introduction
metallocenes provide precise control to polymer synthe-
ses, allowing the preparation of polymers with well-
defined microstructures. Olefin polymerization with
metallocene catalysts obeys well-defined kinetics and
polymers prepared have low molar mass distribution
[2]. The basic, sterically open unbridged metallocene,
bis(g5-cyclopentadienyl)zirconium-dichloride, Cp2ZrCl2,
polymerizes ethene to linear high molar mass crystalline
polyethene with good activity [3]. A variety of metallo-
cene dichloride precursors with different cyclopentadie-
nyl (Cp) and indenyl (Ind) ligand modifications have
been developed to improve the activity or to control
the monomer insertion process or both [4].
Metallocene dichlorides activated with methylalumi-
noxanes (MAO) are highly active catalyst for the polym-
erization of a-olefins [1]. Unlike traditional
heterogeneous Ziegler–Natta catalysts, single center
*
Corresponding authors. Tel.: +35 85 055 78048; fax: +35 89 191
50198.
E-mail address: erkki.aitola@helsinki.fi (E. Aitola).
Present address: Borealis Polymers Oy, P.O. Box 330, FIN-06101
1
Porvoo, Finland.
2
Department of Chemistry, University of Joensuu, P.O. Box, 111
FIN-80101 Joensuu, Finland.
0022-328X/$ - see front matter ꢀ 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2004.09.089

774
E. Aitola et al. / Journal of Organometallic Chemistry 690 (2005) 773–783
The function of different substitution patterns of the
the ligand precursors 1–4 in dichloromethane gave the
complexes 5–8 with 1:1 mixture of rac/meso isomers
(Scheme 1). Different solubilities of the rac and meso iso-
mers allowed separation of the diastereomers under
carefully chosen crystallization conditions. Crystals of
both isomers of complex 5 suitable for X-ray diffraction
measurements were obtained by this manner.
Cp- or Ind-ring is to create an electronically and steri-
cally optimized surrounding for the catalytic active me-
tal center. The size and symmetry of substituents linked
to the Cp or Ind moieties are essential variables for tun-
ing the polymerization efficiency of catalysts and poly-
mer properties. Benzyl substitution of the ligands is an
interesting way to sterically and electronically modify
the chemical structure of the catalyst as interactions be-
tween a phenyl ring and the transition metal center in
metallocenes can occur [5,6]. Computational studies of
1
In the H NMR spectrum the chemical shifts of the
indenyl five-member ring and the CH2 linkage of the
benzyl group are slightly affected by the different substit-
uents in the phenyl moiety. A methoxy substituent in
para-position (6) moves the signals to higher field while
the electron-withdrawing p-fluoro substituent (7) moves
them to lower field compared to the signals found for
benzyl-substituted
[(g5-Cp)(g5-CpCH2C6H5)]ZrMe+
have shown that g3- and g1-coordination modes of ring
carbons are energetically more favorable than an agostic
interaction via hydrogen [7].
1
the unsubstituted benzyl unit (5). In the H NMR spec-
In our earlier studies, rac- and meso-like [8] isomers
of the unbridged bis{(para-fluoro-indenyl)}ZrCl2 (7)
showed moderate activity in ethene polymerization but
lacked the capability to polymerize propene [9]. In aug-
mentation for this study we investigated various ligand
systems where the para-substituent attached to the ben-
zyl group was varied. Here we report synthesis of a ser-
ies of benzyl-substituted bis(1-indenyl) zirconocene
dichlorides, use of their separated rac and meso forms
in ethene polymerization and quantum chemical studies
to illustrate interactions between various benzyl moieties
and the cationic metal center.
tra the difference in chemical shifts of the rac and meso
isomers of 5–8 can be clearly distinguished by the
appearing of the protons of indenyl five-member ring.
In complex 6, for example, these peaks appear at 5.22
and 6.20 ppm for the meso form and at 5.73 and 5.97
ppm for the rac isomer. There are also slight differences
in the shifts of the CH2 linkage protons for two isomers;
the signals appear at slightly higher field for the rac
isomers.
2.2. Solid-state structures of meso- and rac-5
The Ortep illustrations of the crystal structures of
meso-7, meso-5 and rac-5 are presented in Fig. 1. In
the solid-state meso-5 is C1-symmetric and the geometry
around the Zr center differs only slightly from that of
meso-7 [9]. meso-5 crystallizes in a central/lateral
gauche-type orientation just like the meso-7 while rac-5
crystallizes in a bis-lateral anti-type orientation similar
as reported for bis{(1-cyclohexyl)tetrahydroinde-
nyl}ZrCl2 [11]. In rac-5 the ligands are symmetry equiv-
alent and the complex is C2-symmetric. The bond
distances between Zr–Cl and Zr–Cp both in meso- and
rac-5 (Table 1) correspond to the distances reported
2. Results
2.1. Synthesis of ligands and complexes
Indene derivatives having a benzyl substituent in
1-position are easily synthesized by reacting indenyl lith-
ium with appropriate benzyl halides in polar solvents
[9,10]. After the reaction, ligand precursors were purified
by column chromatography and isolated with moderate
to good yields. Treatment of ZrCl4 with lithio salts of
Scheme 1. Synthesis of ligands 1–4 and complexes 5–8.

E. Aitola et al. / Journal of Organometallic Chemistry 690 (2005) 773–783
775
Fig. 1. Crystal structures of complexes meso-7, meso-5 and rac-5.
Table 1
Selected bonding parameters for rac- and meso-5
2.3. Ethene polymerization and characterization of
polyethene
Parameter
meso-5
rac-5
˚
˚
The catalytic behavior of 5–8 was studied in ethene
polymerization at 50 ꢁC at three different monomer
pressures (2.8, 3.8 and 4.6 bar), after activation with
MAO. The results are presented in Table 2. Both meso
and rac forms of the catalysts were examined (except
complex 8) and generally the two isomers behaved dif-
ferently. The dependence of the activity on the ethene
pressure is presented in Fig. 2.
Rac and meso-6/MAO, having a MeO group in para-
position were least active of the catalysts studied and
their activity was almost pressure independent, while
the molar mass of the polymers increased steadily with
monomer pressure. The molar mass distribution,
MMD, of the polymer produced with meso-6/MAO in-
creases from 4.2 to 4.5 with the pressure. The rac form
behaved differently; the molar mass distribution in-
creased from 6.0 to 10.3 when pressure was raised from
2.8 to 4.6 bars.
Both rac and meso isomers of unsubstituted 5/MAO
were clearly more active than the corresponding isomers
of 6. The activity of rac 5/MAO was almost independent
of pressure, 25–29 kgPE/mmolcat h, while the activity of
meso-5/MAO increased from 13 to 41 kgPE/mmolcat h.
The molar mass of the polymers made with the meso-
5/MAO increased up to pressure of 3.8 bar and then it
stayed constant. In the same pressure range the molar
mass of the polymers prepared with the corresponding
rac form increased from 480 to 700 kg/mol. With both
meso and rac forms, molar mass distributions were max-
imum at pressure of 3.8 bar.
Both isomers of the para-fluoro-substituted catalyst
7/MAO were the most active in the series of the ben-
zyl–indenyl zirconium dichlorides. Activity increased
with pressure for both meso and rac catalyst: with the
meso form from 21 to 55 kgPE/mmolcat h and the rac
form 39 to 47 kgPE/mmolcat h. The molar mass of the
polymers produced with rac- and meso-7/MAO in-
creased with ethene pressure: rac-7/MAO from 475 to
Zr–Cl1
Zr–Cl2
2.417(3) A
2.4358(12) A
˚
2.431(4) A
˚
2.4358(12) A
˚
˚
Zr–Cp1
Zr–Cp2
2.223 A
2.224 A
˚
2.237 A
˚
2.224 A
Cl1–Zr–Cl2
/, Cp1–Zr–Cp2a
C15–Cp1–Cp2–C26b
C10–Cp1–Cp2–C10Ab
b, Cp1–Zrc
b, Cp2–Zrc
hd
95.55(14)ꢁ
128.5ꢁ
96.22(6)ꢁ
128.1ꢁ
78.2ꢁ
0.0ꢁ
3.5ꢁ
3.5ꢁ
58.9(1)ꢁ
3.2
5.6ꢁ
55.2(6)ꢁ
Cp1 and Cp2 refer to the centers of the cyclopentadienyl ring plane.
a
Bending angle.
Dihedral angle.
b
c
Deviation from perpendicularity to the Cp–Zr vector.
Angle between the two Cp planes.
d
for Ind2ZrCl2 in the solid state [12]. Also the Cp1–Zr–
Cp2 angle (meso 128.5ꢁ and rac 128.1ꢁ) correlates with
the values for 7 (129.0), Ind2ZrCl2 (128.3ꢁ) and bis{(1-
cyclohexyl)tetrahydroindenyl}ZrCl2 (meso 130.5ꢁ and
rac 130.8ꢁ) [9,11]. These values are typical for unbridged
group 4 metallocenes, being a few degrees larger than
those for the corresponding ethyl-bridged ansa-com-
plexes [13]. The distances between zirconium and the
Cp carbon atoms in meso-5 vary from 2.417(11) to
2.631(11). The corresponding distances of rac 5 range
˚
from 2.469(4) to 2.594 A, and can be considered as typ-
ical values for g5:g5 coordination distances for bent
metallocenes. The angle between the two Cp planes, h,
is reduced in meso-5, as expected when indenyl is substi-
tuted in 1-position. The values of deviations from per-
pendicularity to the Zr–Cp vectors b are comparable
to those of meso-7. In the crystal structure of meso-5
the benzyl substituents point towards the same direction
like in meso-7 but the dihedral angle (C15–Cp1–Cp2–
C26, 78.2ꢁ) is diminished from that of meso-7. In the
structure of rac-5 the corresponding dihedral angle is
0ꢁ, indicating that the benzyl substituents are oriented
in opposite directions.

776
E. Aitola et al. / Journal of Organometallic Chemistry 690 (2005) 773–783
Table 2
Ethene polymerization results with rac andmeso isomers of catalysts 5–8/MAOa
Catalyst Pressure Activity
(bar)
Mw (kg/mol) Mn (kg/mol)
MWD
Tm (ꢁC) Vinyl% trans-Vinylene% Vinylidene%
(kgPE/mmolcat h)
GPC
GPC
FT-IR (Mw/Mn) DSC
FT-IR
FT-IR
FT-IR
rac-5
rac-5
rac-5
2.8
3.8
4.6
2.8
3.8
4.6
2.8
3.8
4.6
2.8
3.8
4.6
2.8
3.8
4.6
2.8
3.8
4.6
2.8
3.8
4.6
29.0
25.2
26.7
12.8
31.4
40.7
18.3
17.2
16.0
8.0
487.2
629.8
710.3
586.2
680.5
677.6
442.8
465.8
539.2
229.8
240.3
270.4
487.9
608.6
724.5
456.3
558.2
606.8
471.0
493.3
444.7
169.2
185.9
223.4
207.1
182.4
218.5
73.80
53.23
52.60
54.79
53.99
52.60
167.9
169.7
156.1
156.5
203.5
222.7
242.8
273.9
249.8
189.0
200.6
226.7
233.1
213.3
266.0
57.33
45.11
47.22
39.52
39.99
43.60
130.9
124.0
144.3
150.0
196.2
201.1
292.1
276.5
268.0
2.88
3.39
3.18
2.83
3.73
3.10
6.00
8.75
10.25
4.19
4.45
4.54
2.92
3.59
4.64
2.92
2.74
2.73
1.94
1.80
1.78
133.7
133.8
133.5
134.0
133.3
135.5
134.0
134.0
135.0
133.3
134.4
135.0
133.5
134.8
134.1
133.9
133.0
133.9
135.7
134.2
134.7
83.9
86.3
91.3
87.7
86.9
85.8
93.8
96.4
96.6
96.2
97.0
97.4
86.4
89.5
90.4
87.4
87.4
88.2
86.9
87.3
90.0
11.8
9.6
4.3
4.1
3.7
4.1
4.3
1.6
1.8
1.4
1.4
1.8
1.6
1.8
2.9
3.3
2.0
3.1
3.2
3.8
2.9
2.4
3.1
5.0
8.1
meso-5
meso-5
meso-5
rac-6
8.8
12.5
4.4
rac-6
rac-6
2.1
2.0
meso-6
meso-6
meso-6
rac-7
2.0
1.4
8.4
8.9
0.75
10.6
7.2
39.2
39.2
46.8
20.8
38.0
55.0
105.6
178.7
205.7
rac-7
rac-7
7.8
9.5
meso-7
meso-7
meso-7
rac-8
9.4
8.1
10.2
10.4
6.9
rac-8
rac-8
a
Polymerization conditions: ethene pressures: 2.8, 3.8 and 4.6 bar, reactor temperature 50 ꢁC, Al:Zr = 9000 (rac and meso 5–7), 27000 (rac-8),
polymerization time ca. 30 min. Amount of catalyst: 0.3 lmol (rac and meso 5–7), 0.1 lmol (rac-8).
coordination of the benzyl ring to the metal center oc-
curs from o-carbon as illustrated in Fig. 3. In general
the coordination mode agrees with the earlier results
of Saßmannshausen et al. [7] calculated for [(g5-
Cp)(g5-CpCH2C6H5)]ZrMe+. The bond distances and
intramolecular phenyl metal coordination energies are
presented in Table 3. The coordination energies are esti-
mated by presenting the stabilities of the coordinating
isomers (meso-a and rac-c) relative to the corresponding
non-coordinating rac-b isomers presented in Fig. 3. The
metal–phenyl coordination energies are of the same
magnitude with metal–ethene coordination energies,
which in the case of the rac-b isomer of 5 are 55 and
53 kJ/mol by HF/3-21G(*) and B3LYP/6-31G* methods,
respectively. The HF/3-21G(*) method somewhat overes-
timates both metal–phenyl coordination energies and
Zr–C distances, but is capable of producing the stability
trends (OMe > H > F) in line with B3LYP/6-31G* cal-
culations. It should be noted that complexes are subject
to basis-set superposition error. BSSE corrections, cal-
culated by the counterpoise method, decrease the inter-
action energies by approximately 20 kJ/mol. The
influence of BSSE is systematic, therefore not affecting
the qualitative trends.
220
200
180
160
140
120
100
rac-5
meso-5
rac-6
meso-6
rac-7
meso-7
rac-8
50
40
30
20
10
0
2.5
3.0
3.5
4.0
4.5
5.0
Polymerization pressure, bar
Fig. 2. Ethene polymerization activities of catalysts rac and meso 5–7/
MAO and rac-8/MAO.
700 kg/mol and meso-7/MAO from 450 to 600 kg/mol.
In the related pressure range the MMD of the polye-
thene prepared with meso-7/MAO stayed almost con-
stant at about 2.8 and with rac-7/MAO it increased
from 2.8 to 4.5.
2.4. Correlation between the catalyst structures and the
polymerization behavior
We assume that the phenyl–metal interaction is
probably the main reason for of the low polymerization
activity of the benzyl catalysts 5–7/MAO; the coordina-
tion of the benzyl moiety to the vacant coordination
site interferes the monomer coordination. In the
Ab initio calculations indicated that an intramolecu-
lar interaction between the benzyl moiety and zirconium
center is relevant in cationic methyl complexes and the

E. Aitola et al. / Journal of Organometallic Chemistry 690 (2005) 773–783
777
mass indicates decreased rate of chain termination,
which may also explain the high trans-vinylene end
group concentrations in polymers made with the cata-
lysts (see below). The corresponding diastereomers of
5/MAO and 7/MAO follow same trends in activity, mo-
lar mass and polydispersity. The meso isomers of cata-
lysts 5/MAO and 7/MAO were more responsive to
increasing monomer pressure than the rac forms, which
produce PE with rather constant activity regardless the
applied monomer concentration (Fig. 2). The stronger
coordination of the benzyl group in the rac-isomers
could be an explanation for the observed polymerization
behavior (Table 3).
According to the calculations, an electronegative
para-fluoro substituent in the benzyl group in complex
7 decreases electron density in the phenyl ring and,
weakens its interaction with the metal center (Table 3)
and enhances the polymerization activity of rac- and
meso-7/MAO compared to rac- or meso-5/MAO. Corre-
spondingly the activities were lowest for the methoxy-
substituted catalysts rac- and meso-6/MAO. This can
be explained by the inductive effect of the electron-
donating methoxy group which increases electron den-
sity in the phenyl ring and the attraction between the
phenyl ring and the electron-deficient metal cation be-
comes stronger.
Molar mass distribution of the polymers prepared
with rac- and meso-6/MAO are significantly broad
(4.2–10.25) and this indicates the presence of several
active species. According to the study of Piccolrovazzi
et al. MAO has propensity to coordinate with MeO-
groups in [4,7-(MeO)2Ind]2ZrCl2 which cause broaden-
ing of the MMD. In addition, MAO coordination
causes further congestion around the catalytic active
metal center and leads to reduced activity. Besides the
strong phenyl interaction, another reason for the low
activity might be the coordination of Lewis-acidic alum-
inum centers of the MAO to the methoxy group [14].
Together with strong metal–phenyl interaction the
above-mentioned coordination of MAO with MeO-
groups offer a rational explanation for the observed
polymerization behavior of rac and meso isomers of 6/
MAO.
Fig. 3. Hartree–Fock optimized structures of dichloride and cationic
monomethyl forms of studied benzyl-substituted {g5-(1-benzyl)inde-
nyl}zirconocenes. (5) X = H; (6) X = OMe; (7) X = F.
Table 3
Calculated relative stabilities (DE; kJ/mol) of intramolecular benzyl
coordination in the cationic monomethyl forms of the rac and meso
isomers of 5–7 and closest Zr–CPh distances
Isomer
Substituent
HF/3-21G(*)
B3LYP/6-31G*
DE
(kJ/mol)
Zr–C
˚
(A)
DE
(kJ/mol)
Zr–CPh
˚
(A)
Rac-5a
Meso-5c
Rac-6a
Meso-6c
Rac-7a
Meso-7c
H
À62
À62
À67
À70
À62
À55
2.75
2.76
2.77
2.73
2.78
2.76
À52
À47
À52
À52
À50
À42
2.71
2.68
2.73
2.67
2.71
2.70
H
OMe
OMe
F
F
phenylethyl-substituted complex rac-8, the ethyl bridge
elongates the distance between the phenyl ring and the
Zr center and the coordination of phenyl is not energet-
ically favorable because such a coordination would re-
quire the transformation of the carbon chain from
trans to gauche conformation [7]. As a result, rac-8/
MAO is much more active than the benzyl-substituted
complexes studied in this work and acts like a typical
single site metallocene catalyst producing polyethene
with narrow molar mass distribution close to numeric
value two. This finding is in accordance with Licht
et al. [10c], who have shown that activity increases with
chain length of the carbon spacer between the Cp moiety
and the phenyl group.
2.5. Chain termination reactions
To have a deeper insight for the polymerization reac-
tion, unsaturated chain end structures formed by chain
termination reactions were measured by FT-IR. Absor-
bances of vinyl (908.8 cmÀ1), vinylidene (888.2 cmÀ1
)
and trans-vinylene (966 cmÀ1) end groups were deter-
mined [15] and the corresponding calculated end group
distributions are presented in Table 2. In general, num-
ber average molar mass, Mn, values obtained by FT-IR
and GPC followed each other well for the polymers pre-
pared at different pressures (Figs. 4 and 5) in spite that
The moderately active diastereomers of 5/MAO and
7/MAO produced high molar mass polyethene with
slightly broadened molar mass distribution. High molar

778
E. Aitola et al. / Journal of Organometallic Chemistry 690 (2005) 773–783
Fig. 4. Weight average molar mass, Mw, and number average molar mass, Mn, obtained with GPC and FT-IR of the polyethenes prepared in
pressures of 2.8, 3.8 and 4.6 bar with catalysts 5/MAO: rac (a) and meso (b) and catalyst 6/MAO: rac (c) and meso (d).
Fig. 5. Weight average molar mass, Mw, and number average molar mass, Mn, obtained with GPC and FT-IR of the polyethenes prepared in
pressures of 2.8, 3.8 and 4.6 bar with catalysts 7/MAO: rac (a) and meso (b) and catalyst 8/MAO: rac (c).
the difference between these two values was substantial
in many cases.
With catalyst meso-5/MAO the Mn values calculated
from FT-IR data are systematically higher than results
from GPC data (Fig. 4(b)). This indicates the presence
of saturated end groups in the polymers probably
formed by the termination via chain transfer to the
cocatalyst [16]. Similar observations were made for poly-
ethenes synthesized with the catalysts 7/MAO and
8/MAO and correlations between Mn values obtained

E. Aitola et al. / Journal of Organometallic Chemistry 690 (2005) 773–783
779
by FT-IR and GPC are presented in Fig. 5. In polyethe-
nes made with meso-6/MAO the difference between val-
ues obtained by FT-IR and GPC was almost constant
and Mn-values obtained by FT-IR were systematically
lower than those obtained by GPC (Fig. 5(b)). This is
probably a result of the formation of internal double
bonds to polymer chains by an allylic activation mecha-
nism [17].
According to the theoretical calculations of Zielger
et al. [18] thermodynamically favorable chain termina-
tion path in olefin polymerizations with metallocene cat-
alysts is the b-hydride transfer to a coordinated
monomer. This process leads to a vinyl chain end group
to the polymer. An alternating chain termination mech-
anism, which leads also to formation of vinyl chain ends
is intramolecular b-hydride elimination [20]. In all poly-
ethenes of this study the vinyl end group was domi-
nantly most common and its abundance varied from
86% to 97% of total amount of the unsaturated end
groups.
tant chain termination reaction at higher monomer con-
centrations. The observations above and the fact that
the activity of the rac-5 did not increase at higher mono-
mer concentrations are indications of the substantial
benzyl coordination. On the contrary the molar mass
of the polyethene prepared with meso-5/MAO was al-
most constant with increasing monomer concentration
and the activity was clearly enhanced. As the amount
of vinyl end groups was decreased and trans-vinylene
end groups were increased, the termination reactions
with meso-5 were dominantly b-H transfer to a coordi-
nated monomer and chain isomerization. Similar trends
can be observed also for rac- and meso-7/MAO, while
with the catalyst 6/MAO the presence of several active
species complicates the interpretation of the results.
3. Conclusions
Despite the freely rotating indenyl ligands, the rac
and meso isomers of studied metallocenes had different
polymerization properties and by changing the para-
substituents the catalytic behavior of 5–7/MAO can be
fine tuned. Computational results of methylated cationic
rac and meso isomers of complexes 5–7 showed that the
benzyl substituent has a possibility for coordination
with the cationic zirconium center. This coordination
orientates the phenyl moiety against the metal center
in all rac and meso benzyl–indenyl complexes. Calcula-
tions show a g1-coordination between the metal and
an o-carbon of the phenyl ring and the intramolecular
coordination energies are of the same magnitude as of
the ethene monomer. This creates a steric block to the
incoming monomer, disturbs the coordination equilib-
rium, and creates catalytic centers with different proper-
ties as shown in the broadening of MMD. According to
the polymerization results the overall polymerization
activity was decreased but the molar mass of the poly-
ethenes remained high. In rac-8/MAO, the interaction
of the phenyl ring does not have a notable role in polym-
erization due to the elongated distance between metal
and phenyl ring created by the ethyl spacer. Catalyst
rac-8/MAO was significantly more active than the other
catalysts and had a typical single-site polymerization
character producing polyethene with polydispersity
close to 2.0. All the catalysts produced polyethenes with
predominantly vinyl end groups and the competing
chain isomerization termination process produced ca.
10% of trans-vinylene units to the polymers (with the
exceptions of the isomers of 6/MAO).
The steric and electronic properties of the catalysts as
well as polymerization conditions affect on the activa-
tion energies of the different termination reactions and
can vary the ratio of the different end groups in the syn-
thesized polymers. When the activation energy barrier
for chain isomerization reaction is close to the b-hydride
transfer, the probability to the chain termination via
chain isomerization is increased. This termination reac-
tion leads to a formation of trans-vinylene chain end
groups and it is strongly affected by the steric properties
of the catalyst [19]. Relatively high contents of trans-
vinylene groups were detected in polyethenes prepared
with both isomers of the catalysts 5/MAO and 7/MAO
and rac-8/MAO the percentage of trans-vinylenes varied
from 5.0% to 11.8% while in polymers made with both
rac- and meso-6/MAO trans-vinylene contents only
0.8–4.4%. Decreased free space around the active sites
of rac- and meso-6 caused by the tightly coordinating
benzyl substituent hinders rearrangement of the coordi-
nated polymer and prevents formation of the bis-b-ago-
stic structure, which is an essential intermediate in the
chain isomerization process.
Vinylidene end groups in polyethenes have been ex-
plained in terms of the long chain branching mechanism.
In situ formed polyethene macromonomers with a viny-
lic double bond at the chain end can make a reinsertion
to the growing polymer [21]. If b-hydride transfer takes
place after a 1,2-insertion of the macromonomer, a viny-
lidene end group is formed. Contents of vinylidene end
groups were very low in all polymers (1.4–4.3%) and
rac and meso isomers of 6/MAO produced polymers
with lowest content of vinylidene groups.
3.1. General methods and materials
With increasing monomer pressure catalyst rac-5/
MAO produced polymers with higher molar mass and
vinyl content and lower amount of trans-vinylene chain
ends. This illustrates that b-H elimination is an impor-
All reactions were carried out under argon atmos-
phere using standard Schlenk techniques. Diethyl ether,
toluene, and hexane were purified by distillation on

780
E. Aitola et al. / Journal of Organometallic Chemistry 690 (2005) 773–783
sodium. Dichloromethane was distilled from CaH2. The
cocatalyst, methylaluminumoxane, MAO, 30 wt% solu-
tion in toluene, was received from Borealis Polymers.
Compounds 7–8 were prepared according to the proce-
dures presented in the literature [9,10].
888 cmÀ1, the vinyl peak at 908 cmÀ1 and the trans-vin-
ylene peak at 966 cmÀ1 were used to calculate the concen-
trations of the corresponding terminal groups [15].
Melting temperatures were measured with a Perkin–
Elmer DSC-2 equipped with a digital data acquisition
device. Thermogramms were measured from 30 to 230
ꢁC using a temperature scanning rate of 10 K/min. Peak
melting temperature was determined from the second
heating curve.
3.2. Polymerization procedure
Polymerizations were performed in a thermostatic
1000 mL Buchi glass autoclave. The autoclave was
¨
loaded with 300 mL of toluene and required amount of
30% MAO in toluene. After this ethene was fed to the
autoclave until saturated ethene/toluene solution was
achieved. A constant ethene pressure was maintained
with an automated pressure controller. The polymeriza-
tions were started by injecting 20 mL of zirconocene
dichloride precursor solutions into the autoclave. Ethene
consumption was monitored by a flow controller con-
nected to a computer. The polymerizations were stopped
by unloading the autoclave of the ethene and leading the
polymerization solutions to HCl–methanol solutions.
The precipitated polymers were washed with methanol,
filtered and dried overnight at 60 ꢁC.
In the polymerizations with both isomers of the cata-
lysts 5, 6 and 7 the amount of catalyst was 0.3 lmol and
the calculated Al:Zr was 9000. With highly active cata-
lyst, rac-8/MAO, only 0.1 lmol of the catalyst was used
in order to hold the polymerization process controllable.
As the same amount of MAO was added in all the pol-
ymerizations to utilize stable scavenging effect of free
TMA, the Al:Zr ratio with catalyst rac-8/MAO was
27000.
3.4. Computational details
All geometry optimizations were carried out with the
GAUSSIAN 03 program package [22], using HF/3-21G(*)
and hybrid density functional B3LYP/6-31G* methods
[23]. Since the standard 6-31G* basis set is not available
for zirconium, Zr was described by HuzinagaÕs extra ba-
sis (Zr, 433321/433/421) [24]. Theoretical treatment of
transition metal complexes generally requires the use
of more sophisticated methods than Hartree–Fock,
mainly due to near-degeneracy and relativistic effects
[25]. However, zirconium is located at the beginning of
the second transition row in the periodic table, which
means that near-degeneracy and relativistic effects are
weak and the Hartree–Fock method at the HF/3-
21G(*) level is capable of predicting the structures of
zirconocenes with a considerable accuracy. This was
recently demonstrated with a set of 62 zirconocenes
characterized by crystallographic method in which the
crystal structures were compared with the corresponding
optimized structures [26]. Furthermore, the performance
of the HF/3-21G(*) method was found to be comparable
or superior to the more expensive no local density func-
tional BLYP [27], hybrid density functional B3LYP and
MP2 calculations with larger basis sets.
3.3. Polyethene characterization
Molar mass and molar mass distributions were meas-
ured with a Polymer Labs high temperature GPC appa-
ratus equipped with refractive index and light scattering
detectors. Chromatographic runs were performed at 140
ꢁC with 1,2,4-trichlorobenzene as an eluent; the flow
rate was kept at 1.0 mL/min. Absolute Mw values were
computed from data of a 15ꢁ and 90ꢁ light scattering
detector.
End group analysis of polyethenes was done with a
Perkin–Elmer Fourier transform infrared spectrometer.
The spectra were measured from thin disks prepared
from polyethenes. PE sample powder (0.2 g) was pressed
onto a film at ambient temperature between two smooth
steel plates using a hydraulic press. A round disk of 13
mm of diameter was cut from the film and its thickness,
approximately 0.50 mm, was measured with a microme-
ter. The disk was placed into the IR sample holder and
measured with 16 scans from 4000 to 370 cmÀ1 with
0.5 cmÀ1 nominal resolution. Baseline correction of the
absorbance spectrum was done by an interactive method.
Corrected absorbance values of the vinylidene peak at
3.5. Synthesis of complexes
3.5.1. (1-Benzyl)indene (1)
n-Butyllithium (1.6 M in hexane, 26.6 ml, 42.6 mmol)
was added dropwise at 0 ꢁC to a solution of indene (5
ml, 42.6 mmol) and Et2O and the reaction was com-
pleted by the addition of benzyl bromide (5.0 ml, 42.6
mmol). After stirring overnight at ambient temperature,
an aqueous solution of NH4Cl was added and the mix-
ture was extracted with Et2O. The solvent was evapo-
rated and the crude products were purified by column
chromatography giving yellow oils.
1 (Crude product 7.2 g, 35.0 mmol, 82%) MS(EI) m/z
1
HRMS(EI) Calcd. for C16 H14. Found: H NMR (200
MHz, CDCl3) d = 2.63 (dd, 1H, J = 9.2/4.2/9.2 Hz,
CH2), 3.04 (dd, 1H, J = 6.8/6.8/6.8 Hz, CH2), 3.60–
3.68 (m, 1H, CHInd), 6.36 (d, 1H, J = 7.4 Hz, CHInd),
6.72 (d, 1H, J = 7.4, CHInd), 7.09–7.31 (m, 9H, aromatic
H). Anal. Calcd. for C16H14: C, 93.16; H, 6.84. Found:
C, 90.43; H, 6.85%.

E. Aitola et al. / Journal of Organometallic Chemistry 690 (2005) 773–783
781
3.5.2. 1-(p-Methoxy benzyl)indene (2)
C-11), 129.1 (CH, C-13,15), 128.8 (CH, C-12,16),
126.4, 126.3 (CH, C-6,7), 125.5 (CH, C-5,8), 123.7
(CH, C-14), 99.4 (CH, C-2,3), 70.8 (C, C-1), 30.0
(CH2, C-10).
This ligand was prepared in the same way as 1 except
that para-methoxy benzyl chloride (5.78 ml, 42.6 mmol)
was added to the indenyl lithium.
1
2 (Crude product 8.0 g, 33.9 mmol, 80%) H NMR
(200 MHz, CDCl3) d = 2.58 (dd, 1H, J = 8.8/4.6/9.2
Hz, CH2), 2.95 (dd, 1H, J = 7.0/6.6/7.0 Hz, CH2),
3.55–3.82 (4H, CH3 + CHInd), 6.35 (dd, 1H, J = 1.6/
3.8/1.8 Hz, CHInd), 6.68–7.27 (m, 9H, CHInd + aromatic
H). Anal. Calcd. for C17H16O: C, 86.40; H, 6.83; O,
6.77. Found: C, 85.16; H, 6.82%. MS(EI) m/z: 236 (M+).
3.5.6. Bis{(1-p-methoxy benzyl)indenyl}zirconium
dichloride (6)
The methoxy-substituted complex was prepared in a
similar manner to complex 5 (ligand 2 5.7 g, 24.2 mmol).
The rac isomer was obtained from toluene extraction
and the meso isomer from hexane extraction.
6 (Crude product 5.8 g, 9.2 mmol, 76%) Anal. Calcd.
for C34H30O2ZrCl2: C, 64.54; H, 4.78; O, 5.06. Found:
C, 64.34; H, 4.85%. MS(EI) m/z: 632–636 (M+), 397
(M+ À 2).
3.5.3. 1-(p-Fluoro benzyl)indene (3)
The ligand was prepared according to the literature
[14].
1
rac-6 H NMR (200 MHz, CDCl3) d = 3.69 (d, 6H,
3.5.4. 1-(Phenyl ethyl)indene (4)
J = 5.2 Hz, CH3), 3.96 (s, 2H, CH2), 4.06 (s, 2H,
CH2), 5.73 (d, 2H, J = 3.0 Hz, CHInd), 5.97 (d, 2H,
J = 3.2 Hz, CHInd), 6.69–6.73 (m, 3H, aromatic H),
6.96–7.25 (m, 10H, aromatic H), 7.49–7.56 (m, 3H, aro-
matic H). 13C NMR, CCl3D: d, 158.3 (C, C-14), 132.0
(C, C-11), 129.9 (CH, C-12,17) 126.8–126.2 (CH, C-
6,7), 125.8–125.5 (CH, C-5,8), 114.1 (CH, C-13,16)
99.3 (CH, C-2,3), 70.8 (C, C-1), 55.4 (CH3, C-15), 33.3
(CH2, C-10).
meso-6 1H NMR (200 MHz, CDCl3) d = 3.70 (d, 6H,
J = 5.2 Hz, CH3), 4.10 (d, 4H, J = 3.2 Hz, CH2), 5.22 (d,
2H, J = 2.8 Hz, CHInd), 6.20 (d, 2H, J = 3.0 Hz, CHInd),
6.74–6.79 (m, 3H, aromatic H), 7.05–7.33 (m, 10H, aro-
matic H), 7.54–7.57 (m, 3H, aromatic H). 13C NMR,
CCl3D: d, 158.4 (C, C-14), 131.6 (C, C-11), 130.1
(CH, C-12,17), 126.6–126.0 (CH, C-6,7), 126.0–125.2
(CH, C-5,8), 114.1 (CH, C-13,16), 99.2 (CH, C-2,3),
55.4 (CH3, C-15), 33.4 (CH2, C-10).
This ligand was obtained in the same manner as com-
pound 1 except for the addition of 2-phenyl ethyl bro-
mide (5.8 ml, 42.6 mmol) indenyl lithium salt.
1
4 (Crude product 6.8 g, 31.0 mmol, 75%) H NMR
(200 MHz, CDCl3) d = 1.66–1.84 (m, 1H, CH2), 2.00–
2.23 (m, 1H, CH2), 2.54–2.62 (m, 2H, CH2), 3.36–3.46
(m, 1H, CHInd), 6.48 (dd, 1H, J = 1.8/4.0/1.8 Hz,
CHInd), 6.74 (d, 1H, J = 1.8/3.8/1.8 Hz, CHInd), 7.01–
7.35 (m, 9H, aromatic H). Anal. Calcd. for C17H16: C,
92.68; H, 7.32. Found: C, 85.28; H, 6.88%.
3.5.5. Bis(1-benzyl)indenylzirconium dichloride (5)
The ligand 1 precursor (5.0 g, 24.2 mmol) was con-
verted to a lithio salt by using n-butyllithium (15.2 ml,
24.2 mmol) at 0 ꢁC in Et2O. The reaction mixture was
stirred 2 h at room temperature. After removal of the
solvent the solid residue was cooled down to À78 ꢁC.
Cooled CH2Cl2 was added, followed by an addition of
ZrCl4 (3.1 g, 13.3 mmol). The reaction mixture was al-
lowed to come to room temperature and stirred over-
night. After filtering through Celite the solvent was
distilled off and the solid residue was extracted with
hot hexane and toluene. meso and rac isomers crystal-
lized from the mother liquors, meso from toluene and
rac from hexane.
3.5.7. Bis{(1-phenyl ethyl)indenyl}zirconium dichloride
(8)
This complex was prepared in the same way as com-
plex 5 (ligand 5.3 g, 24.2 mmol). The rac isomer was ob-
tained from the toluene mother liquor.
8 (Crude product 5.7 g, 9.5 mmol, 78%) Anal. Calcd.
for C34H30ZrCl2: C, 67.98; H, 5.03. Found: C, 66.16; H,
5.02%.
rac-8 1H NMR (200 MHz, CDCl3) d = 2.67–3.20 (m,
8H, CH2), 5.74 (d, 2H, J = 2.4 Hz, CHInd), 5.85 (d, 2H,
J = 2.6 Hz, CHInd), 6.98–7.59 (m, 18H, aromatic H). 13
NMR (CCl3D): d, 141.5 (C, C-12), 128.7 (CH, C-14,16),
128.5 (CH, C-13,17), 126.0–126.8 (CH, C-6,7), 125.5–
126.0 (CH, C-5,8), 99.1 (CH, C-2,3), 70.8 (C, C-1),
36.1 (CH2, C-11), 30.1 (CH2, C-10).
5 (Crude product 5.5 g, 9.6 mmol, 79.0%) Anal.
Calcd. for C32H26ZrCl2: C, 67.11; H, 4.58. Found: C,
66.53; H, 4.68%.
1
rac-5 H NMR (200 MHz, CDCl3) d = 4.15 (q, 4H,
C
J = 8.2/7.4/16.0 Hz, CH2), 5.84 (d, 2H, J = 3.4 Hz,
CHInd), 6.07 (d, 2H, J = 3.0 Hz, CHInd), 7.12–7.67 (m,
18H, aromatic H). 13C NMR, CCl3D: 140.0 (C, C-11),
128.9 (CH, C-13,15), 128.7 (CH, C-12,16), 127.0–126.4
(CH, C-6,7), 126.2–125.8 (CH, C-5,8), 124.0 (CH, C-
14), 99.3 (CH, C-2,3), 52.0 (C, C-1), 34.2 (CH2, C-10).
meso-5 1H NMR (200 MHz, CDCl3) d = 4.18 (q, 4H,
J = 7.2/15.6/11.0 Hz, CH2), 5.31 (d, 2H, J = 4.0 Hz,
CHInd), 6.31 (d, 2H, J = 3.0 Hz, CHIns), 7.12–7.67 (m,
18H, aromatic H). 13C NMR, CCl3D: d, 139.7 (C,
3.6. X-ray crystallographic studies
The crystal data of meso-5 and rac-5 were collected
on a Rigaku AFC-7S single crystal diffractometer at

782
E. Aitola et al. / Journal of Organometallic Chemistry 690 (2005) 773–783
193(2) K using Mo Ka radiation (graphite monochro-
matized, scan type w/2h). Intensities were corrected for
Lorentz and polarization effects and for absorption
(0.89 < 1.00). Solution: Direct methods combined with
subsequent Fourier analysis. All non-hydrogen atoms
were refined anisotropically. Hydrogen atoms were re-
fined isotropically on calculated positions (riding mod-
el). Suitable crystals for X-ray determination were
obtained by recrystallization. The meso isomer crystal-
lized from hot hexane and the rac crystals from hot tol-
References
[1] (a) L. Resconi, L. Cavallo, A. Fait, F. Piemontesi, Chem. Rev.
100 (2000) 1253;
(b) H.H. Brintzinger, D. Fischer, R. Mulhaupt, B. Rieger, R.M.
¨
Waymouth, Angew. Chem., Int. Ed. Engl. 34 (1995) 1143;
(c) G.W. Coates, Chem. Rev. 100 (2000) 1223.
[2] H. Yiannoulakis, A. Yiagopoulos, P. Pladis, C. Kiparissides,
Macromolecules 33 (2000) 2757.
[3] W. Kaminsky, Macromol. Chem. Phys. 197 (1996) 3907.
[4] (a) P.C. Mo¨hring, N.J. Coville, J. Organomet. Chem. 479 (1994)
1;
uene. Calculations were carried out with SHELXTL
and SHELSL-93 programs.
/PC
(b) H.G. Alt, A. Ko¨ppl, Chem. Rev. 100 (2000) 1205, see also [1].
[5] M. Bochmann, M.L.H. Green, A.K. Powell, J. Saßmannshausen,
M.U. Triller, S. Wocadlo, J. Chem. Soc., Dalton Trans. (1999) 43.
[6] L. Doerrer, M.L.H. Green, D. Ha¨ußinger, J. Saßmannshausen, J.
Chem. Soc., Dalton Trans. (1999) 2111.
meso: Crystal dimensions 0.15 · 0.15 · 0.08 mm,
a = 10.380(4), b = 11.715(4), c = 20.972(6), a = b =
3
˚
c = 90ꢁ, V = 2550.2(15) A , crystal system orthorhom-
bic, space group P2(1)2(1)2(1), Z = 4, D(calcd) = 1.491
[7] M. Buhl, J. Saßmannshausen, J. Chem. Soc., Dalton Trans.
¨
(2000) 79.
3
˚
g/cm , k = 0.71073 A, data collection range h = 2.61–
[8] Diastereomers of unbridged 1-substituted metallocenes should be
denoted as rac-like and meso-like. Further in the text they are
denoted simply as rac and meso which is a common practice in the
referred papers.
24.99ꢁ, reflections collected 2144, independent reflec-
tions 2144, reflections refined 1418, goodness-of-fit on
F2 = 0.994, R1 = 0.0695 and wR2 = 0.0804 (with
I0 > 2dI0), R1 = 0.1171 and wR2 = 0.909 (all data). The
largest difference peak and hole in the final residual elec-
tron density map is situated in the vicinity of zirconium
[9] G. Jany, M. Gustafsson, T. Repo, E. Aitola, J.A. Dobado, M.
Klinga, M. Leskela¨, J. Organomet. Chem. 553 (1998) 173.
[10] (a) N.E. Grimmer, N.J. Coville, C.B. de Koning, J.M. Smith,
L.M. Cook, J. Organomet. Chem. 616 (2000) 112;
(b) N.E. Grimmer, N.J. Coville, C.B. de Koning, J. Organomet.
Chem. 642 (2002) 195;
(0.477 and À0.558 e AÀ3).
˚
rac: Crystal dimensions 0.70 · 0.25 · 0.08 mm,
(c) E.H. Licht, H.G. Alt, M.M. Karim, J. Organomet. Chem. 599
(2000) 261;
˚
a = 19.559(4), b = 6.6980(10), c = 20.551(4) A, a =
3
˚
c = 90ꢁ, b = 107.76(3)ꢁ, V = 2564.0(8) A , crystal system
monoclinic, space group C2/c, Z = 4, D(calcd) = 1.483
(d) E.H. Licht, H.G. Alt, M.M. Karim, J. Organomet. Chem. 599
(2000) 275.
[11] (a) G. Erker, M. Aulbach, M. Knickmeier, D. Wingbermuhle,
3
¨
˚
g/cm , k = 0.71073 A, data collection range h = 2.52–
C. Kruger, M. Nolte, S. Werner, J. Am. Chem. Soc. 115
¨
(1993) 4590;
22.49ꢁ, reflections collected 2872, independent reflec-
tions 1628, reflections refined 1367, goodness-of-fit on
F2 = 1.015, R1 = 0.0397 and wR2 = 0.0918 (I0 > 2dI0),
R1 = 0.0524 and wR2 = 0.0962 (all data). The largest dif-
ference peak and hole in the final residual electron den-
sity map is situated in the vicinity of zirconium (0.441
(b) C. Kruger, F. Lutz, M. Nolte, G. Erker, M. Aulbach, J.
¨
Organomet. Chem. 452 (1993) 79.
[12] T. Repo, M. Klinga, I. Mutikainen, Y. Su, M. Leskela¨, M.
Polamo, Acta Chem. Scand. 50 (1996) 1116.
[13] (a) S. Collins, B.A. Kuntz, N.J. Taylor, D.G. Ward, J. Organ-
omet. Chem. 342 (1988) 21;
and À0.604 e AÀ3).
˚
(b) W. Kaminsky, K. Kuelper, H.H. Brintzinger, F.R.W.B. Wild,
Angew. Chem. 97 (1985) 507.
[14] (a) N. Piccolrovazzi, P. Pino, G. Consiglio, A. Sironi, M. Moret,
Organometallics 9 (1990) 3098;
4. Supplementary material
(b) M. Linnolahti, T.A. Pakkanen, Macromolecules 33 (2000)
9205;
Crystallographic data for the structural analyses of
the complexes have been deposited with Cambridge
Crystallographic Data Centre, CCDC, No. 222747 for
rac-5 and 222746 for meso-5. Copies of this information
is available free of charge from: CCDC, 12 union Road,
Cambridge CB2 1EZ, UK; fax: +44-123-336-336033;
e-mail: deposit@ccdc.cam.ac.uk.
(c) P. Foster, M.D. Rausch, J.C.W. Chien, J. Organomet. Chem.
527 (1997) 71.
[15] J. Haslam, H.A. Willis, D.C.M. Squirrel, Identification and
Analysis of Plastics, second ed., Illiffe Books, London, 1972.
[16] (a) L. Resconi, S. Bossi, L. Abis, Macromolecules 23 (1990) 4489;
(b) L. Resconi, F. Piemontesi, G. Franciscono, L. Abis, T.
Fiorani, J. Am. Chem. Soc. 114 (1992) 1025;
(c) S. Lieber, H.H. Brintzinger, Macromolecules 33 (2000) 9192.
[17] G. Moscardi, L. Resconi, Organometallics 20 (2001) 1918.
[18] J.C.W. Lorenz, T.K. Woo, L. Fan, T. Ziegler, J. Organomet.
Chem. 449 (1995) 91.
Acknowledgments
[19] (a) V. Busico, R. Cipullo, J. Am. Chem. Soc. 116 (1994) 9329;
(b) K. Thosrhaug, J.A. Støvneng, E. Rytter, M. Ystenes,
Macromolecules 31 (1998) 7149;
Finnish Technology Development Center (TEKES)
and Academy of Finland (projects 204408 and 77317)
are acknowledged for funding of the research. Hedvig
Byman-Fagerholm is thanked for her expert assistance
in the FT-IR measurements.
(c) P. Lehmus, E. Kokko, R. Leino, H. Luttikhedde, B. Rieger,
J.V. Seppa¨la¨, Macromolecules 33 (2000) 8534;
(d) M.K. Leclerc, H.H. Brintzinger, J. Am. Chem. Soc. 118 (1996)
9024.

E. Aitola et al. / Journal of Organometallic Chemistry 690 (2005) 773–783
783
[20] (a) M.H. Prosenc, H.H. Brintzinger, Organometallics 16 (1997)
3889;
Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador,
J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels,
M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Ragh-
avachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S.
Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-
Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W.
Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, J.A. Pople,
Gaussian Inc., Pittsburgh, PA, 2003.
(b) P.M. Margl, T.K. Woo, T. Ziegler, Organometallics 17 (1998)
4997.
[21] (a) A. Malmberg, J. Liimatta, A. Lehtinen, B. Lo¨fgren, Macro-
molecules 32 (1999) 6689;
(b) T.K. Woo, P.M. Margl, T. Ziegler, Organometallics 16 (1997)
3454.
[22] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A.
Robb, J.R. Cheeseman, J.A. Montgomery, T. Vreven, Jr., K.N.
Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V.
Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A.
Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian,
J.B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Strat-
mann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W.
[23] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.
[24] S. Huzinaga, Gaussian Basis Sets for Molecular Calculations,
Elsevier, Amsterdam, 1984.
[25] (a) P.E.M Siegbahn, Adv. Chem. Phys. 93 (1996) 333;
(b) P. Pyykko¨, Chem. Rev. 88 (1988) 563.
[26] M. Linnolahti, P. Hirva, T.A. Pakkanen, J. Comput. Chem. 22
(2001) 51.
[27] D. Becke, Phys. Rev. A 38 (1988) 3098.