Immobilization of η5-cyclopentadienyltris(dimethylamido) zirconium polymerization catalysts on a chlorosilane- and HMDS-modified mesoporous silica surface: A new concept for supporting metallocene amides towards heterogeneous single-site-catalysts

Article abstract of DOI:10.1016/S1381-1169(01)00051-6

The modification of a mesoporous silica surface with Si(Ind)(CH₃)₂Cl and the immobilization of CpZr(NMe₂)₃ on this surface was studied via IR-spectroscopy. To reduce side reactions, the indenyl-modified silica was reacted with hexamethyldisilazane (HMDS) under IR-control before the CpZr(NMe₂)₃-immobilization. The role of the hydroxyl group protection with HMDS is discussed. The surface modifications have been repeated via Schlenk technique at the same conditions and the surface modifications were studied with ¹³C CP MAS-NMR, ¹H MAS-NMR, elemental-, SEM- and BET-analysis. The surface species of the resulting catalysts are discussed. The precatalysts have been treated with methylaluminoxane (MAO) (Al:Zr (mol:mol) = 500:1) and the resulting Zr contents (leaching-effect) are discussed. All catalysts have been tested in ethylene and propylene polymerization.

Full text of DOI:10.1016/S1381-1169(01)00051-6

Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
Immobilization of ␩5-cyclopentadienyltris(dimethylamido)
zirconium polymerization catalysts on a
chlorosilane- and HMDS-modified mesoporous silica surface:
a new concept for supporting metallocene amides towards
heterogeneous single-site-catalysts
Horst Schneidera, Gerd T. Puchtaa, Franz A.R. Kaula, Gabriele Raudaschl-Siebera,
Frédéric Lefebvreb, Guillaume Saggiob, Dimitrios Mihaliosa,
Wolfgang A. Herrmanna,∗, Jean Marie Bassetb
a
Lehrstuhl für Anorganische Chemie, Anorganisch-Chemisches Institut, Technische Universität München,
Lichtenbergstr. 4, 85747 Garching, Germany
Laboratoire de Chimie Organométallique de surface UMR CNRS-CPE 9986,
b
43 Bd du 11 novembre 1918, 691616 Villeurbanne Cedex, France
Received 10 November 2000; accepted 10 January 2001
Abstract
The modification of a mesoporous silica surface with Si(Ind)(CH3)2Cl and the immobilization of CpZr(NMe2)3 on this
surface was studied via IR-spectroscopy. To reduce side reactions, the indenyl-modified silica was reacted with hexamethyl-
disilazane (HMDS) under IR-control before the CpZr(NMe2)3-immobilization. The role of the hydroxyl group protection
with HMDS is discussed. The surface modifications have been repeated via Schlenk technique at the same conditions and
the surface modifications were studied with 13C CP MAS–NMR, 1H MAS–NMR, elemental-, SEM- and BET-analysis. The
surface species of the resulting catalysts are discussed. The precatalysts have been treated with methylaluminoxane (MAO)
(Al:Zr (mol:mol) = 500:1) and the resulting Zr contents (leaching-effect) are discussed. All catalysts have been tested in
ethylene and propylene polymerization. © 2001 Elsevier Science B.V. All rights reserved.
Keywords: Heterogeneous catalysts; Metallocene; Polymerization; Surface chemistry; Leaching
1. Introduction
tions in homogeneous systems have been done [1–5],
because these reactions are easier to understand in ho-
mogeneous than in heterogeneous systems [6].
The discovery that partially hydrolyzed alkyl alu-
minum compounds (MAO) increase the activity of
metallocenes [7,8] as shown by Sinn and Kaminsky
[9–17], and the introduction of ansa-zirconocenes
by Brintzinger and Ewen [6,18,19] as homogeneous
olefin polymerization catalysts activated with MAO,
To understand elementary steps in polymerization
of olefins by Ziegler-type catalysts many investiga-
∗ Corresponding author. Tel.: +49-89-2891-3080;
fax: +49-89-2891-3473.
E-mail address: lit@arthur.anorg.chemie.tu-muenchen.de (W.A.
Herrmann).
1381-1169/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S1381-1169(01)00051-6

128
H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
started a wide exploration of these systems [20–24].
For an industrial use in existing plants, metallocene
catalysts have to be heterogenized due to the fact that
current technology is based on gas phase and slurry
processes [25]. On the other hand, the heterogeniza-
tion of metallocene catalysts allows an improvement
of polymer morphology, whereas reactor fouling and
also the amount of MAO can be reduced [26–32].
As supports a wide range of materials like silica,
alumina, zeolites, MgCl2 [26,32–38] or organic car-
riers like cyclodextrins, polymers and starch [39–43]
are used.
model system on silica, where simple surface modi-
fying reactions and their analytical measurements can
be done at every step, from the initial surface modifi-
cation to polymerization.
We herein report the preparation of a chlorosilane
compound as spacer ligand, of a zirconiumamide com-
pound as Zr-precursor and of his anchoring on the
modified mesoporous silica surface.
At first, the immobilization reactions were done on
a pellet and examined via infrared spectroscopy. In
the second part, silica samples have been prepared in
toluene via Schlenk technique under the same con-
1
There are various immobilization techniques and
strategies, the most common being impregnation tech-
niques (simply removing the solvent from catalyst
suspension in vacuo) or grafting techniques (wash-
ing and filtering the catalyst before drying). Either
the metallocene is immobilized or the cocatalyst or
both, in several steps or as reaction mixture. The an-
alytical studies of such systems remain very compli-
cated. If in case of silica, the support is treated with
MAO at first, direct interactions between carrier and
metallocene are excluded and the results in propylene
polymerization are similar to those produced by homo-
geneous catalysts as found by Kaminsky and Renner
[32]. The complexity of heterogenized systems after
MAO-activation during the polymerization of olefins
can be shown with electron paramagnetic resonance
measurements. About 40% ZrIII could be detected
in the Cp2ZrCl2/MAO system, during the ethylene
polymerization [44]. Olivé and Olivé found approxi-
mately 70% TiIII for the same homogeneous Ti system
[45,46].
To prevent loss of activity, the fixed active catalytic
centers and the vacant coordination sites have to be
preserved for the olefins. But side reactions between
the catalytic precursor and the surface lower the cat-
alytic activity [47].
To examine the basic reactions during the immo-
bilization, a better approach is to modify the surface
with spacer-ligands at first and than, in a second step,
to fix the metallocene precursor, with analytical analy-
sis on every step as reported by Iiskola and coworkers
[48–50].
ditions and examined with 13C CP MAS–NMR, H
MAS–NMR, N2-porosimetry (BET-measurements),
elemental analysis and SEM-technique. The results
of the spectroscopic examinations are discussed. Also
polymerizations and reactions with MAO to exam-
ine the zirconium content after MAO-treatment, and
polymerizations with these pretreated samples have
been done.
2. Experimental
2.1. General procedures
All reactions were carried out under argon at-
mosphere using glove box and Schlenk techniques.
THF and Et2O were dried and deoxygenated by
distillation over CaH2 and sodium-benzophenone.
Pentane and toluene were dried as described in [57].
Indene (Fluka, 80%) was distilled prior to use. Di-
cyclopentadiene (Merck Schucharth, dimer, >93%)
was cracked and freshly distilled prior to use in syn-
thesis. ZrCl4 (Merck Schucharth, >98%), LiNMe2
(Aldrich, 95%), dichlorodimethylsilane (Merck
Schucharth, >98%), chlorotrimethylsilane (Merck
Schucharth, >99%), hexamethyldisilazane (HMDS)
(Merck Schucharth, >98%), Na2SO4 (Aldrich,
99.99%, activated at 120◦C) and NaNH2 (Merck
Schucharth, >97%) were used without further purifi-
cation.
Aeroperl 300/40 silica (Degussa, 300 m2 g−1, pore
volume = 1.9 cm3 g−1, particle size = 40–120 ␮m)
was used as support. Methylaluminoxane (Witco, 30%
in toluene) was used as cocatalyst. Ethylene (AGA Gas
GmbH, grade 3.5) and propylene (Linde AG, grade
2.8) were purified by passing them through two purifi-
There have been many affords to modify the sur-
face (mainly of silica) with spacer ligands and gener-
ate the catalyst precursor on the surface [51–56]. The
complexity of such systems prompted us to prepare a

H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
129
cation columns containing activated BTS-catalyst and
molecular sieve (4 Å) before feeding the reactor.
and the slightly yellow solid purified by subli-
mation (60◦C/0.2 mbar). About 6.18 g (77%) of
tetrakis(dimethylamido)zirconium were obtained.
1H NMR (C6D6): δ (ppm), 3.0 (3H, –CH3). 13C
NMR (C6D6): δ (ppm), 39 (–CH3). Anal. Calcd. for
C8H24N4Zr (%): C, 35.92; H, 9.04; N, 20.94; Zr,
34.10. Found (%): C, 35.32; H, 9.02; N, 20.78; Zr,
34.22.
2.2. Synthesis procedures
2.2.1. Synthesis of chlorodimethylinden-3-ylsilane
A solution of indene (10.04 ml, 109 mmol) in THF
(20 ml) was added dropwise at room temperature
under vigorous stirring to a suspension of NaNH2
(4.68 g, 109 mmol) in THF (100 ml). The mixture
was stirred overnight. The dark red solution was
filtered, the solvent was removed in vacuo, and
the remaining solid (NaInd) washed with hexane
(2 × 25 ml). To a solution of dichlorodimethylsilane
(7.9 ml, 65 mmol, 3 eq.) in 50 ml THF, a solution
of NaInd (3 g, 22 mmol) in 25 ml THF was added
dropwise at −30◦C under vigorous stirring. The mix-
ture was slowly warmed up to room temperature
and stirred overnight. The brownish oily solution
was separated, the solvent removed in vacuo and
the solid washed with THF (3 × 20 ml). The brown
oil, a mixture of chlorodimethylinden-3-ylsilane, di-
indenyldimethylsilane, indene and higher oligomers,
was distilled in vacuo. About 2 g (44%) of color-
less chlorodimethylinden-3-ylsilane (boiling range:
80–95◦C/0.2 mbar) could be isolated. 1H NMR
(CDCl3): δ (ppm), 0.0 (3H, –CH3); 0.05 (3H,
–CH3); 3.5 (1H, C3–H); 6.5 (1H, C2–H); 6.8 (1H,
C1–H); 7.0 (1H, C5–H); 7.1 (1H, C6–H); 7.3 (1H,
C4–H); 7.4 (1H, C7–H). 13C NMR (CDCl3): δ
(ppm), −0.1 (–CH3); 0.0 (–CH3); 47 (C3); 121
(C4); 123 (C7); 124 (C6); 125 (C5); 131 (C1);
133 (C2); 142 (C7a); 144 (C3a). Anal. Calcd. for
C11H13ClSi (%): C, 63.29; H, 6.28; Cl, 16.98;
Si, 13.45. Found (%): C, 63.55; H, 6.30; Cl,
16.32.
2.2.3. Synthesis of η5-cyclopentadienyltris(dimethy-
lamido)zirconium
About 1.01 eq. of freshly distilled cyclopentadiene
(0.75 g, 11.33 mmol) was added dropwise to a col-
orless solution of tetrakis(dimethylamido)zirconium
(3 g, 11.22 mmol) in 30 ml toluene at −10◦C. Af-
ter refluxing the solution for 2 h, the solvent was
removed in vacuo, to give the deep yellow liquid
␩5-cyclopentadienyltris(dimethylamido)zirconium in
quantitative yield. 1H NMR (C6D6): δ (ppm), 2.9
(18H, –CH3); 6.0 (5H, –C5H5). 13C NMR (C6D6):
δ (ppm), 41.9 (–CH3); 107.5 (–C5H5). Anal. Calcd.
for C11H23N3Zr (%): C, 45.79; H, 8.03; N, 14.56; Zr,
31.61. Found (%): C, 46.59; H, 7.82; N, 14.61; Zr,
31.54.
2.3. Synthesis of the precatalysts
2.3.1. Dehydratation of silica and preparation of the
indenyl-modified support
Silica was first heated at 500◦C in a nitrogen flow
for 5 h and then kept under vacuum (2 × 10−4 mbar)
during 12 h at the same temperature, in order to obtain
the dehydroxylated silica.
Chlorodimethylinden-3-ylsilane (1.06 g, 5 mmol)
was added dropwise at room temperature to a sus-
pension of 4 g of dehydroxylated silica in 40 ml of
toluene. The mixture was heated at 100◦C for 8 h
and then at 200◦C for 4 h under refluxion. The liquid
phase was separated by filtration and the solid washed
with toluene (3 × 10 ml) and dried 30 h in vacuo at
room temperature.
For the infrared experiments, the silane was intro-
duced on the pellet via its vapor at 60◦C. After re-
action at 100 (8 h) and 200◦C (4 h), the excess of
chlorodimethylinden-3-ylsilane was condensed in a
liquid nitrogen trap and removed. Finally, the pellet
was treated under dynamic vacuum at 100◦C during
2 h in order to remove all volatile compounds.
2.2.2. Synthesis of tetrakis(dimethylamido)zirconium
About 6.99 g (30 mmol) of zirconium tetrachlo-
ride and 6.45 g (120 mmol) of lithiumdimethy-
lamide were combined in the glove box. The re-
action mixture was cooled to −196◦C and 150 ml
pentane were added slowly via cannula. The re-
sulting suspension was slowly heated up to room
temperature under vigorous stirring. After 4 days
of stirring at room temperature, the solution was
filtered off and the solid extracted with pentane
(3 × 50 ml). The solvent was removed in vacuo

130
H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
2.3.2. Reaction of HMDS with the indenyl-modified
support
nuclear magic-angle spinning probe-head. The 13C
CP MAS–NMR spectra were obtained at 75.47 MHz
with a 5 ms contact time and a 1 s delay between
HMDS (0.32 g, 2 mmol) was added dropwise to
a suspension of 1.5 g of the indenyl-modified sup-
port in 20 ml of toluene under stirring at room
temperature. After stirring for 13.5 h at room tem-
perature, 1 h at 60◦C and 0.5 h at 70◦C, the solid
was filtered off, washed with toluene (3 × 10 ml)
and dried 30 h in vacuo at room temperature. For
the infrared experiments, HMDS was introduced as
vapor at room temperature while the heating proce-
dure was the same as above. The excess of HMDS
was removed by use of a liquid nitrogen trap and
the pellet was treated at 70◦C in a dynamic vacuum
for 20 h.
1
the scans (ca. 5000 transients were accumulated). H
MAS–NMR spectra were obtained at 300.18 MHz
(32 transients, recycle delay 1 s). In all cases, the
spinning rate was 8 kHz.
Elemental analysis for carbon, hydrogen and nitro-
gen were determined on a Vario EL analyzer. The
zirconium contents were determined on a Schimadzu
UV-160 photometer and the aluminum contents on a
Varian Spectr-AA 400 spectrometer.
The nitrogen adsorption/desorption isotherms were
carried out on a Micromeritics ASAP 2000 equipment.
The scanning electron micrographs were obtained on
a Jeol, JSM-5900 LV Scanning Electron Microscope
at 5 and 15 kV.
2.3.3. Preparation of precatalyst A
About 2 g of the indenyl-modified support were
suspended in 20 ml of toluene. After addition of 0.73 g
(2.5 mmol) of ␩5-cyclopentadienyltris(dimethyla-
mido)zirconium, the suspension was heated to 120◦C
under stirring for 3 h. The resulting solid, called
precatalyst A, was filtered, washed with toluene
(3 × 10 ml) and dried at room temperature for 30 h in
vacuo.
2.5. Polymerization and polymer-analytical
procedures
2.5.1. Polymerization procedure
In a glove box 10 mg (200 mg) of the precatalyst
were dissolved/suspended in 50 ml (30 ml) of toluene
in a Schlenk tube. The appropriate amount of MAO
(30% in toluene, Al/Zr = 500) was then added slowly
under stirring upon a color change from yellow to
orange occurred. The reaction mixture was stirred
overnight at room temperature.
2.3.4. Preparation of precatalyst B
About 1.5 g of the indenyl- and HMDS-modified
support were suspended in 20 ml of toluene. After ad-
dition of 0.55 g (1.9 mmol) of ␩5-cyclopentadienyltris-
(dimethylamido)zirconium, the suspension was heated
to 120◦C for 3 h under stirring. The resulting solid,
called precatalyst B, was filtered, washed with toluene
(3 × 10 ml) and dried in vacuo at room temperature
for 30 h.
A 0.5 l Buechi glass-autoclave, equipped with a
pressflow gas controller was filled with approxi-
mately 230 ml of dry and oxygen-free toluene. The
autoclave was pressurized with ethene or propene
and thermostated to the reaction temperature. The
desired amount of catalyst/MAO solution/suspension
was transferred into the injection system via sy-
ringe or cannula, the injection system closed, the
solution/suspension injected into the autoclave and
the injection system washed with an additional 5 ml
of toluene. The pressure was kept constant dur-
ing the polymerization and the amount of ethene
consumed by the catalyst during the reaction was
measured. The polymerization was quenched af-
ter 1 h by adding 10 ml of acidic methanol and
the resulting polymer-suspension was poured into
250 ml of methanol–HCl mixture (3:1) and stirred
for 16 h to precipitate the polymer. After filtration,
the polymer was dried at 60◦C in vacuo. If no
2.4. Physicochemical characterization of the
precatalysts
All solids were characterized by various meth-
ods. The infrared spectra were recorded on a Nicolet
Magma IR 550 and a Perkin-Elmer FT-IR 1600. For
studies of the reactions of the indenyl-modified silane
and HMDS an infrared cell with CaF2-windows
analogous to [58] and allowing in situ studies was
used.
MAS–NMR experiments were recorded on a
Bruker DSX-300 spectrometer equipped with a multi-

H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
131
polymer/oligomer precipitated, the solution was
3. Results and discussion
treated with dilute HCl, water and methanol. The
organic layer was separated and the solvent removed
in vacuo. The molecular weight determination of
the polymers was performed on a combination of
a Waters 150 CV–high-temperature GPC at 135◦C
with 1,2,4-trichlorobenzene as solvent and a modified
Whyatt Mini Dawn light-scattering instrument. The
calibration was performed with polystyrene standards
and conversion into PE- and PP-calibration curves
using Mark-Houwink parameters. Melting points
determinations (Tm) were carried out on a Netzsch
TASC 414/2 differential scanning calorimeter (DSC)
upon reheating the polymer sample to 220◦C at a
3.1. Modification of the silica surface
3.1.1. Reaction with Si(Ind)(CH3)2Cl
The infrared spectrum of silica, dehydroxylated un-
der vacuum at 500◦C, shows in the ν(O–H) range
a strong band at 3750 cm−1 together with a small
broad shoulder near 3650 cm−1 (Fig. 1a), attributed to
isolated and H-bonded hydroxyl groups, respectively.
When chlorodimethylinden-3-ylsilane is adsorbed on
silica at room temperature, the infrared spectrum is
greatly modified, the ν(O–H) band at 3750 cm−1 be-
ing shifted to ca. 3600 cm−1 and new bands appear-
ing in the ν(C–H) range (Fig. 1b). No gas evolution
is observed in agreement with literature data, which
indicate that at room temperature just physical ad-
sorption of the chlorosilanes takes place, and that it
is necessary to increase the reaction temperature to
150◦C in order to observe a reaction with the support
[59]. After heating to 200◦C, HCl is observed in the
gas phase showing that the well known reaction be-
tween chlorosilanes and silica [59–66] has occurred
(Scheme 1).
heating rate of 3◦C min−1
.
2.5.2. Leaching of the catalysts
The precatalysts, suspended in 5 ml toluene, were
reacted with a MAO/toluene solution (30%, ratio
Zr:Al (mol:mol) = 1:500) at room temperature under
stirring in the glove box for 24 h. After the reaction,
the solids were separated by centrifugation, washed
with toluene (3 × 5 ml) and dried for 30 h in vacuo at
room temperature.
Fig. 1. Infrared spectra (in the 4000–2700 cm−1 range) of: (a) silica dehydroxylated at 500◦C; (b) physisorption of Si(Ind)(CH3)2Cl on
(a) at room temperature; (c) after grafting of Si(Ind)(CH3)2Cl on (a) at 200◦C and removal of the excess of chlorosilane; (d) sample (c)
after reaction with HMDS; (e) sample (c) after reaction with ZrCp(NMe2)3; (f) sample (d) after reaction with ZrCp(NMe2)3.

132
H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
Scheme 1. Modification of the silica surface at 100 and 200◦C with chlorodimethylinden-3-ylsilane.
After removal of the excess of chlorosilane and
after the reaction shows that less than 1% of the used
chlorodimethylinden-3-ylsilane reacts according to
this pathway, if all detected chlorine originates from
this side reaction. Nevertheless, about 0.9 indenyl
spacers per nm2 (4.76 wt.% carbon) can be fixed on
the surface at this conditions.
of the hydrochloric acid, the infrared spectra shows
the following features (Fig. 1c): (i) a non-negligible
amount of silanol groups at 3750 cm−1 have been re-
stored, showing that the grafting reaction is not com-
plete; (ii) a new ν(O–H) band appears at 3630 cm−1
,
attributed to silanol groups in interaction with the
chemisorbed chlorosilane; (iii) ν(C–H) bands, char-
acteristic of the indenyl and methyl moieties, appear
near 3000 cm−1. All these features are in agreement
with the reaction depicted in Scheme 1.
3.1.2. Reaction of the modified silica with HMDS
As mentioned above, a non-negligible amount
of silanol groups did not react with the chlorosi-
lane. These silanol groups can be responsible for
side-reactions during the grafting of zirconium. As
a consequence, it is necessary to remove them. The
reaction of the chlorosilane compound can be one
way with silica at higher temperatures. This had been
achieved at 300–400◦C with methylchlorosilanes [62]
but Si(Ind)(CH3)2Cl is not stable at these reaction
conditions for a longer period of time.
Another way, commonly used in catalysis, chro-
matography and tailoring of silica surfaces [64,72–77],
is the silylation of isolated silanol groups with
HMDS, leading to the formation of hydrophobic
Si(CH3)3 groups on the silica surface. The synthesis
and properties of HMDS towards chemical reac-
tions are well known and were already described by
Sauer and coworkers in the 40s [78,79]. Hertl and
Hair determined that the reaction of free silanols of
chemically unmodified silica and HMDS followed a
second-order kinetic law with an activation energy
of 18.5 kcal mol−1 [76]. Only a small fraction of
H-bonded hydroxyl groups react with HMDS [76].
Further, advantages of HMDS towards reaction with
silica is the high reaction rate already at 30◦C [76].
Infrared spectroscopy shows that after reaction
with HMDS (Scheme 2) at 70◦C nearly all isolated
silanol groups (characterized by the ν(O–H) band at
3750 cm−1) have disappeared (Fig. 1d). In addition,
the increase of the C–H streching bands at 2962
The solid-state NMR examinations are also in
agreement with these conclusions (Figs. 2 and 3).
After reaction of the dehydroxylated support with
the indenylsilane at 200◦C one signal at ca. 2 ppm
(attributed to the protons of isolated silanol groups,
Fig. 2a) can be observed. In addition to this signal,
three other signals at ca. 0, 4 and 7 ppm, attributed
to the Si–CH3, C3–H of the indenyl group and aro-
matic and olefinic protons of the indenyl moiety are
detected. Similarly, the 13C CP MAS–NMR spectrum
shows the presence of the Si(Ind)(CH3)2 fragment
on the surface (Fig. 3a): signals at 0 (Si–CH3), 47
(aliphatic carbon atom of the five-membered ring
of the indenyl spacer), 125 (aromatic carbons of
indenyl), 129 and 130 (olefinic carbon atoms) and
144 ppm (quaternary carbons).
During the reaction of chlorosilanes with silica, no
side reactions resulting in the reopening of siloxane
bridges, as reported by Borello et al. with alcohols
[67] or from the reaction of silica with the reaction
product HCl, as reported from Hertl during the reac-
tion of silica with methanol [68] could be observed
[69,70]. During the reaction of chlorosilanes with sil-
ica, an initial fast reaction of about 10–15% of direct
replacement of hydroxyl groups by a Cl group has
been observed from Hair and Hertel [71]. Our exam-
ination concerning the chlorine amount before and

H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
133
Fig. 2. 1H MAS–NMR of: (a) silica dehydroxylated at 500◦C after reaction with Si(Ind)(CH3)2Cl at 200◦C and removal of the physisorbed
species; (b) solid (a) after reaction with HMDS; (c) precatalyst A (reaction of zirconium complex with (a)); (d) precatalyst B (reaction of
zirconium complex with (c)).
and 2904 cm−1 clearly displays the progress of the
reaction via Scheme 2.
1,3-cyclopentadiene or indene to be deprotonated
with HNMe2 formation. The use of amido instead
of chlorozirconium precursors avoids the neces-
sity of metallating the fixed ligand on the surface
[48,49,85,86]. Also salt formation and the difficult
removal from the surface can be avoided by this way
[87]. In addition, the reaction product dimethylamine
produced via Scheme 3 is gaseous at room temper-
ature and can be easily removed in vacuo from the
surface.
The solid-state NMR results are also in agreement
1
with these results: both the H and 13C NMR spec-
tra (Figs. 2b and 3b) show the increase of the signal
intensity of methyl groups bound to silicon, while mi-
croanalysis gives a carbon content of ca. 5.32 wt.%.
3.1.3. Immobilization of η5-cyclopentadienyltris-
(dimethyl)amidozirconium on the indenyl-modified
silica surface
Zirconium amide complexes are well known and
suitable compounds for the preparation of metal-
locenes [50,80–84]. The basicity of dimethylamine
is high, allowing molecules with acidic protons like
When ␩5-cyclopentadienyltris(dimethyl)amidozir-
conium is adsorbed on the indenyl-modified silica at
room temperature, the band at 3750 cm−1, charac-
teristic of isolated silanol groups, totally disappears.
Only a small broad band at 3680 cm−1, resulting

134
H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
Fig. 3. 13C CP MAS–NMR of: (a) silica dehydroxylated at 500◦C after reaction with Si(Ind)(CH3)2Cl at 200◦C and removal of the
physisorbed species; (b) solid (a) after reaction with HMDS; (c) catalyst A (reaction of zirconium complex with (a)); (d) catalyst B
(reaction of zirconium complex with (b)).
from inner and sterically hindered silanols remains
(Fig. 1e). Dimethylamine, formed during the reaction
of ZrCp(NMe2)3 with the silanol groups is detected
by a small infrared band at 3270 cm−1, arising from
its physisorption on the silica pellet. Infrared spec-
troscopy cannot indicate if, in addition to the reaction
with the remaining hydroxyl groups, the expected
reaction with the indenyl fragment occurred. How-
ever, such informations can only be obtained from
10 ppm to lower field and the aromatic indenyl car-
bons nearly 10 ppm to higher field, showing that the
indenyl spacer has been deprotonated. In addition to
these peaks, the cyclopentadienyl moiety of the zir-
conium complex gives a signal at 113 ppm and the
methyl group of the amido ligand gives a signal at
34 ppm.
The zirconium content of the solid (4.7 wt.%, cor-
responding to 1.5 Zr nm−2) is in agreement with a re-
action with both the hydroxyl groups and the indenyl
spacer (whose loading was 0.9 nm−2). The nitrogen
content of the sample is 0.43 0.2 wt.% correspond-
ing to a N/Zr ratio of 0.3–1. This value clearly indi-
cates that, in the solid, the zirconium complex has re-
tained zero or one amido group. The loss of the three
amido groups of the ZrCp(NMe2)3 complex should
be accompanied by the creation of three bonds with
1
solid-state NMR. As H MAS–NMR shows the ap-
pearance of the signal of the methyl group of the
amido group at 2.4 ppm (Fig. 2c), it is difficult to
determine, from this spectrum, if the indenyl group
has lost one proton. 13C CP MAS–NMR is more in-
formative: the aliphatic carbon (at 47 ppm) and the
olefinic carbon (at 130 ppm) signals have completely
disappeared, while the quartenary carbons are shifted

H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
135
Scheme 2. Silylation of the indenyl-modified silica surface with HMDS.
the surface. Due to the high dehydroxylation temper-
ature and the fact that almost all hydroxyl groups are
isolated, the most probable structure for this triply
surface-bonded species is the one depicted as species
IV in Scheme 3. A complex doubly bonded to the sur-
face could be either species II (reaction with one hy-
droxyl group and one indenyl spacer) or species III
(reaction with two hydroxyl groups). A structure like
species I or V where zirconium is bound to the sur-
face via only one bond is improbable, due to the low
nitrogen loading. This solid, which contains at least
three different zirconium species, will be denoted as
precatalyst A in the following.
spectrum. On the other hand, chemical analysis gives
some interesting features: (i) the zirconium loading is
now 3.11 wt.% (corresponding to 1.1 Zr nm−2) a value
which is 30% lower than that achieved on the support
without treatment with HMDS; (ii) the N/Zr molar ra-
tio of the solid is now 1.3 0.3, showing that there is
one or two amido groups in the coordination sphere of
zirconium and that, in this case, the structure IV is less
probable. 1H MAS–NMR and 13C CP MAS–NMR
experiments show that most of the indenyl fragments
have been deprotonated, allowing to reasonably con-
clude that a structure like V is of minor importance. In
the solid, the resulting zirconium-species should have
the structures I or II (note that the zirconium content,
1.1 nm−2, is quite the same than the indenyl spacer,
0.9 nm−2). This solid is called precatalyst B in the fol-
lowing.
3.1.4. Immobilization of η5-cyclopentadienyltris-
(dimethyl)amidozirconium on the indenyl- and
HMDS-modified silica surface
As mentioned above, ZrCp(NMe2)3 can react with
both the indenyl spacer and the remaining hydroxyl
groups on the silica surface. We then decided to study
its reactivity with the modified silica where most of
the remaining hydroxyl had been removed by reaction
with HMDS. The resulting infrared spectrum is shown
on Fig. 1f. No conclusion can be drawn from this
3.1.5. Morphology of precatalysts A and B
The SEM-pictures of the support and of the pre-
catalysts (Fig. 4) show a fragmentation of the support
during the modification reactions. Additional exami-
nations showed that this fragmentation originated from
the stirring of the samples during the reactions and

136
H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
Scheme 3. Reaction of ␩5-cyclopentadienyltris(dimethylamido)zirconium with the indenyl-modified silica surface and proposed Zr-species.
especially during the drying of the catalysts in vacuo.
The mean diameter of the particles varies between 40
and 120 ␮m. However, this macroscopic modification
of the material has only a slight influence on the ad-
sorption characteristics of the carriers (Table 1).
These data show that the three supports are related
to mesoporous morphology-materials with a surface
area around 200 m2 g−1. The slight decrease of the
surface area and of the pore volume after the grafting
reactions is probably related to: (i) the fact that in
1 g of precatalyst there is less than 1 g of support,
due to the presence of spacer, zirconium complex and
HMDS; (ii) the spacer and the zirconium complex are
grafted in the mesopores, resulting in a decrease of
the accessible volume of the pores.
3.2. Polymerization studies
3.2.1. Polymerization of ethylene
The ethylene polymerization results with precata-
lysts A, B and the precursor CpZr(NMe2)3 are listed
in Table 2.

H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
137
Fig. 4. SEM pictures of: (a) initial support; (b) precatalyst A; (c) precatalyst B.
Table 1
BET-surface and pore characteristics of the support and precatalyst A and B
Support
Precatalyst A
Precatalyst B
Surface areaa (m2 g−1
)
255
1.64
254
211
1.29
215
183
1.07
198
Cumulative pore volume of poresb (cm3 g−1
Average pore diameter of poresb (Å)
)
a BET.
b BJH, desorption, 17–1000 Å diameter.
Fig. 5 shows the activity of the three catalysts as
a function of time. It can be seen that CpZr(NMe2)3
which is the most active catalyst, is rapidly deac-
tivated. One reason could be the observed reactor
fouling. The two supported catalysts are less rapidly
deactivated, probably due to a slighter reactor fouling,
the conversion being smaller. The comparison of the
initial activities shows that the homogeneous catalyst
is the more active one, due to the fact that the two sup-
ported catalysts can contain various zirconium-species
Table 2
a
Polymerization results with precatalyst A, B and CpZr(NMe2)3
Precatalyst A
Precatalyst B
CpZr(NMe2)3
Zr (wt.%)
4.70
3.11
1.60 × 103
0.70 × 103
3.46
0.83
4.17
31.61
4.50 × 103
1.65 × 103
4.60
2.10
2.19
Maximum activity (kg PE mol−1 Zr bar−1 h−1
)
1.50 × 103
1.36 × 103
3.88
Average activity (kg PE mol−1 Zr bar−1 h−1
)
Mw × 105 (g mol−1
)
)
Mn × 105 (g mol−1
0.93
3.35
134.20
D
mp (◦C)
134.10
134.20
a Cocatalyst MAO (Al/Zr (mol/mol) = 1000), polymerization temperature 40◦C, 2 bar, polymerization time 1 h.

138
H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
Fig. 5. Polymerization behavior of the MAO activated precatalyst A, B and ␩5-cyclopentadienyltris(dimethylamido)zirconium (Al/Zr
(mol/mol) = 1000, 40◦C, 2 bar, 1 h).
which are more or less active. The higher polydisper-
sities of the polymers obtained with the heterogeneous
catalyst compared to the polymers obtained with the
homogeneous one are in agreement with that results.
The higher polydispersities of the polymers ob-
tained with catalyst B indicate that, in consequence
of the HMDS-treatment before the immobilization of
CpZr(NMe2)3, more polymerization active Zr-species
are generated on the silica surface.
The higher degree of deactivation of catalyst B
(which contains monografted complexes) compared
to catalyst A (which contains more bigrafted species)
can be related to a leaching-effect, which should be
more easier for monografted complexes than for bi-
grafted ones. For this reason, we decided to study the
leaching process on the two catalysts.
results obtained with these “preactivated” catalysts are
more interesting (Fig. 6 and Table 4). Indeed, a de-
activation can be observed due to reactor fouling, as
in the case of the homogeneous catalyst. This can be
explained by the formation of ionic zirconium-species
during the treatment with the MAO, which are the ac-
tive sites for the polymerization and which are rapidly
removed from the solid. Table 4 shows that the preacti-
vated catalysts are less active than the non-preactivated
ones, probably due to a loss of zirconium during the
activation, even if it is not shown by the chemical
analysis. It has to be pointed out that the decrease of
the activity of precatalyst A upon treatment of MAO
is more significant than that of precatalyst B, show-
ing that the Si–O–Zr bridges are more easily broken
Table 3
3.2.2. Leaching experiments with MAO
BET and metal analysis after MAO-treatment
For this purpose, the catalysts were first treated with
a MAO/toluene solution (30%, Al/Zr (mol/mol) =
500); the resulting catalysts were then characterized
and used in ethylene polymerization. The results of
the analysis of the solids are given on Table 3.
The very high amount of aluminum on the solids,
due to the adsorption of MAO on the silica support
can explain the decrease of the zirconium loadings on
the two catalysts, even if it cannot be excluded that
some leaching occurred during this step. The catalytic
Catalyst A Catalyst B
Al (wt.%)
Zr (wt.%)
31.80
1.03
30.40
0.87
Surface areaa (m2 g−1
Cumulative pore
volume of poresb
)
249
0.21
244
0.18
(cm3 g−1
)
Average pore diameter of poresb (Å)
70
83
a BET.
b BJH, desorption, 17–1000 Å diameter.

H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
139
Fig. 6. Evolution of the activity as a function of time for the preactivated catalysts.
Table 4
Polymerization results and polymer weights with the MAO-treated precatalyst A (= C) and B (= D)a
MAO-modified precatalyst A = C
MAO-modified precatalyst B = D
0.70 × 103
0.47 × 103
6.04
1.71
3.53
Maximum activity (kg PE mol−1 Zr bar−1 h−1
)
0.90 × 103
0.38 × 103
4.36
1.47
2.96
Average activity (kg PE mol−1 Zr bar−1 h−1
)
Mw × 105
Mn × 105
D
mp (◦C)
135.00
139.20
a Cocatalyst MAO (Al/Zr (mol/mol) = 1000), polymerization temperature 40◦C, 2 bar, polymerization time 1 h.
than the bonds between the indenyl fragment and zir-
conium.
The molecular weights of the polymers obtained af-
ter preactivation of the catalysts are higher than those
obtained by ordinary activation. A decrease of the
polydispersities can also be observed.
produced polypropene were atactic and of low molec-
ular mass, resulting in oily liquids. No activity was
observed with the preactivated catalysts.
4. Conclusion
The polydispersities of the polymers obtained with
catalyst D (equal to precatalyst B) are higher in com-
parison to the polymers obtained with catalyst C (equal
to precatalyst A). We assume that the polymerization
active Zr-species in the case of catalyst D are mainly
of the types I and II (Scheme 3). In the case of cat-
alyst C, the polymerization active Zr-species mainly
consists of species III.
A
dimethylinden-3-ylsilane spacer has been
anchored on a thermally treated mesoporous silica-
surface. The low reactivity of chlorosilanes with
silica surfaces was shown by IR-spectroscopy.
CpZr(NMe2)3 was immobilized on this surface under
dimethylamine elimination. This unselective pathway
generates several Zr-surface species. In a better ap-
proach towards heterogeneous single-site-catalysts,
the unreacted hydroxyl groups have been silylated af-
ter anchoring the indenyl-spacer with HMDS. HMDS
The two catalysts were also used in propene poly-
merization, but they gave only poor polymerization
activity of 100–500 kg PP mol−1 Zr bar−1 h−1. The

140
H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
is a very versatile reagent to minimize side reactions
in the third step of the CpZr(NMe2)3-immobilization.
Without OH-silylation more Zr-surface species were
generated, but less polymerization active centers
could be generated by MAO activation compared to
the HMDS-silylated catalyst.
The first surface reaction in CpZr(NMe2)3-immobi-
lization is the reaction of the precursor with steri-
cally hindered surface hydroxyl groups, in a second
step the desired deprotonation of the anchored
indenyl-ligand takes place. The content of the de-
sired species increased after HMDS-treatment of the
indenyl-modified support.
and L. Veyre (Laboratoire de chimie organomet-
allique de surface UMR CNRS-CPE 9986, 43 Bd
du 11 novembre 1918, 691616 Villeurbanne Cedex,
France), for useful discussions and practical help. For
SEM-experiments, we thank Dr. J. Froh (Lehrstuhl
für Allgemeine, Angewandte und Ingenieur-Geologie,
Technische Universität München, Arcisstraße 21,
80290 München).
References
[1] A.D. Count, in: C. Kemball (Ed.), Catalysis, Vol. 1, Chemical
Society, London, 1977, p. 234 (Chapter 7).
In ethene polymerization, the non HMDS-treated
catalyst shows higher activity, produces higher molec-
ular weight polymers with lower polydispersities.
They only differ in the concentration of the polymer-
ization active species resulting in the lower activities
of the heterogeneous catalyst.
In propene polymerization, the catalysts only
show low activity of about 100–500 kg PP mol−1 Zr
bar−1 h−1 and only produce atactic polymer.
[2] J.C.W. Chien, Coordination Polymerization, Academic Press,
New York, 1975.
[3] F.S. Dyachovskyii, in: J.C.W. Chien (Ed.), Coordination
Polymerization, Academic Press, New York, 1975, p. 199.
[4] J. Boor, Ind. Eng. Chem. Prod. Res. Dev. 9 (1970) 437.
[5] J.A. Waters, G.A. Mortimer, J. Polym. Sci., Part A 1 (1972)
895.
[6] H.-J. Sinn, W. Kaminsky, Adv. Organomet. Chem. 18 (1980)
99.
[7] K.H. Reichert, K.R. Meyer, Makromol. Chem. 169 (1973)
163.
A MAO-treatment of the precatalysts (Al/Zr
(mol/mol) = 500) results in a dramatic decrease of
pore volume, Zr content and polymerization activity.
The BET surface of the precatalysts and the molecular
weights of the produced polymers have been increased
significantly in ethene polymerization (Altotal/Zr
(mol/mol) = 1000), also the melting points of the
polymers. After the MAO-treatment, the polymer dis-
persities have decreased. The HMDS-treated catalysts
now produce higher molecular weights and polymer
dispersity in comparison to the non HMDS-treated
catalyst. The polymerization activity is increased with
the HMDS-treated catalyst compared to the unmodi-
fied one. The MAO-treated catalysts are completely
unactive in propene polymerization. Nevertheless, we
could show that an approach towards heterogeneous
single-site-catalysts by the used step-by-step one-pot
synthesis of the precatalyst is possible.
[8] W.P. Long, D.S. Breslow, Justus Liebigs Ann. Chem. (1975)
463.
[9] A. Andresen, H.G. Cordes, J. Herwig, W. Kaminsky, A.
Merck, R. Mottweiler, J. Pein, H. Sinn, H.J. Vollmer, Angew.
Chem. 88 (1976) 689.
[10] A. Andresen, H.G. Cordes, J. Herwig, W. Kaminsky, A.
Merck, R. Mottweiler, J. Pein, H. Sinn, H.J. Vollmer, Angew.
Chem. Int. Ed. Engl. 15 (1976) 630.
[11] H. Sinn, W. Kaminsky, H.J. Vollmer, R. Woldt, Angew. Chem.
92 (1980) 396.
[12] H. Sinn, W. Kaminsky, H.J. Vollmer, R. Woldt, Angew. Chem.
Int. Ed. Engl. 19 (1980) 396.
[13] W. Kaminsky, Nachr. Chem. Tech. Lab. 29 (1981) 373.
[14] W. Kaminsky, MMI Press. Symp. Ser. 4 (1983) 225.
[15] J. Herwig, W. Kaminsky, Polym. Bull. (Berlin) 9 (1983) 464.
[16] W. Kaminsky, M. Miri, H. Sinn, R. Woldt, Makromol. Chem.
Rapid Commun. 4 (1983) 417.
[17] W. Kaminsky, Naturwissenschaften 71 (1984) 93.
[18] J.A. Ewen, J. Am. Chem. Soc. 106 (1984) 6355.
[19] W. Kaminsky, K. Külper, H.H. Brintzinger, F.R.W.P. Wild,
Angew. Chem. 97 (1985) 507.
[20] H.-H. Brintzinger, D. Fischer, R. Mülhaupt, B. Rieger, R.
Waymouth, Angew. Chem. 107 (1995) 1255.
[21] H.-H. Brintzinger, D. Fischer, R. Mülhaupt, B. Rieger, R.
Waymouth, Angew. Chem. Int. Ed. Engl. 34 (1995) 1143.
[22] J.C.W. Chien, W.-M. Tsai, M.D. Rausch, J. Am. Chem. Soc.
113 (1991) 8570.
[23] W. Spaleck, M. Antberg, J. Rohrmann, A. Winter, B.
Bachmann, P. Kiprof, J. Behm, W.A. Herrmann, Angew.
Chem. Int. Ed. Engl. 31 (1992) 1347.
Acknowledgements
We gratefully acknowledge M. Widenmeyer
(Lehrstuhl für Anorganische Chemie, Anorganisch-
chemisches Institut, Technische Universität München,
Lichtenbergstr. 4, 85747 Garching), M. Chabanas

H. Schneider et al. / Journal of Molecular Catalysis A: Chemical 170 (2001) 127–141
141
[24] G. Erker, M. Aulbach, M. Knickmeier, D. Wingbermühle, C.
Krüger, M. Nolte, S. Werner, J. Am. Chem. Soc. 115 (1993)
4590.
[25] W. Kaminsky, M. Arndt, Adv. Polym. Sci. 127 (1997) 143.
[26] T. Kamfjord, T.S. Wester, E. Rytter, Macromol. Rapid
Commun. 19 (1998) 505.
[54] N. Suzuki, H. Asami, T. Nakamura, T. Huhn, A. Fukuoka, M.
Ichikawa, M. Saburi, Y. Wakatsuki, Chem. Lett. (1999) 341.
[55] R. Spitz, N. Verdel, V. Pasquet, J. Dupuy, J.P. Broyer, T.
Saudemont, in: W. Kaminsky (Ed.), Metalorganic Catalysts
for Synthesis and Polymerisation, Springer, Berlin, 1999,
p. 347.
[27] H.C.L. Abbenhuis, Angew. Chem. 111 (1999) 1125.
[28] M. Antberg, V. Dolle, S. Hafka, J. Rohrmann, W. Spaleck,
A. Winter, H.J. Zimmermann, Makromol. Chem. Macromol.
Symp. 48/49 (1991) 333.
[29] K. Ziegler, E. Holzkamp, H. Breil, H. Martin, Angew. Chem.
67 (1955) 541.
[30] K. Ziegler, Angew. Chem. 76 (1964) 545.
[31] J. Tudor, D. O’Hare, Chem. Commun. 6 (1997) 603.
[32] W. Kaminsky, F. Renner, Makromol. Chem. Rapid Commun.
14 (1993) 239.
[33] J.C.W. Chien, D. He, J. Polym. Sci., Part A: Polym. Chem.
29 (1991) 1603.
[56] D.-H. Lee, in: W. Kaminsky (Ed.), Metalorganic Catalysts for
Synthesis and Polymerisation, Springer, Berlin, 1999, p. 397.
[57] A.B. Pangborn, M.A. Giardello, R.H. Grubbs, R.K. Rosen,
F.J. Timmers, Organometallics 15 (1996) 1518.
[58] G. Coudurier, F. Lefebvre, in: B. Imelik, J.C. Vedrine (Eds.),
Catalyst Characterisation — Physical Techniques for Solid
Materials, 3rd Edition, Plenum Press, New York, 1994, p. 11
(Chapter 2).
[59] M.L. Hair, W. Hertl, J. Phys. Chem. 73 (1969) 2372.
[60] B. Evans, T.E. White, J. Catal. 11 (1968) 336.
[61] M.L. Hair, W. Hertl, J. Catal. 15 (1969) 307.
[62] F. Buzek, J. Rathouský, J. Coll. Interf. Sci. 79 (1981) 47.
[63] H.W. Kohlschütter, P. Best, G. Wirzing, Z. Anorg. Allg. Chem.
285 (1956) 236.
[34] K. Soga, M. Kaminaka, Macromol. Chem. Rapid Commun.
13 (1992) 221.
[35] K.-H. Dahmen, D. Hedden, R.L.J. Burwell, T.J. Marks,
Langmuir 4 (1988) 1212.
[36] H. Rahiala, I. Beurroies, T. Eklund, K. Hakala, R. Gougeon,
p. Trens, J.B. Rosenholm, J. Catal. 188 (1999) 14.
[37] S.S. Sarma, G. Satyanarayana, S. Sivaram, Polym. Sci. 1
(1994) 315.
[38] K. Czaja, M. Bialek, Macromol. Rapid Commun. 17 (1996)
253.
[39] W. Kaminsky, US Patent l.4.4431.788.
[40] D.-h. Lee, K.-b. Yoon, Macromol. Rapid Commun. 15 (1994)
841.
[64] L.R. Snyder, J.W. Ward, J. Phys. Chem. 70 (1966) 3941.
[65] J.B. Peri, J. Phys. Chem. 70 (1966) 2937.
[66] J. Wartmann, H. Deuel, Helv. Chim. Acta 42 (1959) 1166.
[67] E. Borello, A. Zecchina, C. Morterra, J. Phys. Chem. 71
(1967) 2938.
[68] W. Hertl, J. Phys. Chem. 72 (1968) 1248.
[69] M.L. Hair, W. Hertl, J. Phys. Chem. 73 (1969) 2372.
[70] J.B. Peri, J. Phys. Chem. 70 (1966) 2937.
[71] M.L. Hair, W. Hertl, J. Catal. 15 (1969) 307.
[72] A.J. Van Roosmalen, M.C.G. Hartmann, J. Catal. 66 (1980)
112.
[41] D.-h. Lee, K.-b. Yoon, Macromol. Symp. 97 (1995) 185.
[42] H. Mori, K. Ohnishi, M. Terano, Macromol. Rapid Commun.
17 (1996) 25.
[43] T. Kitagawa, T. Uozumi, K. Soga, T. Takata, Polymer 38
(1997) 615.
[44] D. Cam, F. Sartori, Macromol. Chem. Phys. 195 (1994) 2817.
[45] G.H. Olivé, S. Olivé, Angew. Chem. 18/19 (1967) 764.
[46] G.H. Olivé, S. Olivé, Angew. Chem. Int. Ed. Engl. 6 (1967)
790.
[47] A. Munoz-Escalona, in: Proceedings of the 5th International
Congress on Metallocene Polymers, Metallocenes Europe’98,
Düsseldorf, 31 March–1 April 1998.
[48] E.I. Iiskola, S. Timonen, T.T. Pakkanen, O. Härkki, P. Lehmus,
J.V. Seppälä, Macromolecules 30 (1997) 2853.
[49] H. Juvaste, E.I. Iiskola, T.T. Pakkanen, J. Organomet. Chem.
587 (1999) 38.
[50] S. Timonen, T.T. Pakkanen, E.I. Iikola, J. Mol. Cat. A: Chem.
148 (1999) 235.
[73] S.P. Boudreau, W.T. Cooper, Anal. Chem. 61 (1989) 41.
[74] J. Nawrocki, Chromatographia 25 (1988) 404.
[75] D.W. Sindorf, G.E. Maciel, J. Phys. Chem. 86 (1982)
5208.
[76] W. Hertl, M.L. Hair, J. Phys. Chem. 75 (1971) 2181.
[77] S. Haukka, A. Root, J. Phys. 98 (1994) 1695.
[78] R.O. Sauer, J. Am. Chem. Soc. 66 (1944) 1707.
[79] R.O. Sauer, R.H. Hasek, J. Am. Chem. Soc. 68 (1946) 241.
[80] D.C. Bradley, I.M. Thomas, J. Chem. Soc. (1960) 3857.
[81] G. Chandra, M.F. Lappert, J. Chem. Soc. (A) (1968)
1940.
[82] W.A. Herrmann, M.J.A. Morawietz, H.F. Herrmann, F. Küber,
J. Organomet. Chem. 509 (1996) 115.
[83] W.A. Herrmann, M.J.A. Morawietz, T. Priermeier, J.
Organomet. Chem. 506 (1996) 351.
[84] J.M. Diamond, S. Rodewald, R.F. Jordan, Organometallics
14 (1995) 5.
[85] K. Soga, M. Kaminaka, H.J. Kim, T. Shiono, in: G. Fink,
R. Mülhaupt, H.-H. Brintzinger (Eds.), Ziegler Catalysts,
Springer, Berlin, 1995, p. 333.
[51] K. Soga, T. Shiono, H.J. Kim, Makromol. Chem. 194 (1993)
3499.
[52] K. Soga, H.J. Kim, T. Shiono, Macromol. Rapid Commun.
15 (1994) 139.
[86] D.-h. Lee, K.-b. Yoon, Macromol. Rapid Commun. 18 (1997)
427.
[53] D.-h. Lee, K.-b. Yoon, Macromol. Rapid Commun. 16 (1995)
223.
[87] J. Wartmann, H. Deuel, Helv. Chim. Acta 42 (1959)
1166.