Mixed substituted zirconocene dichloride complexes as catalyst precursors for homogeneous ethylene polymerization

Article abstract of DOI:10.1016/S1381-1169(00)00205-3

The synthesis and characterization of 15 new, unbridged metallocene dichloride complexes of the type (L)(Ind)ZrCl₂ (L = ω-phenylalkyl substituted cyclopentadienyl or 1-phenylsilyl substituted indenyl; Ind = indenyl) and (L')(Flu)ZrCl₂

Full text of DOI:10.1016/S1381-1169(00)00205-3

Ž
.
Journal of Molecular Catalysis A: Chemical 159 2000 273–283
www.elsevier.comrlocatermolcata
Mixed substituted zirconocene dichloride complexes as
catalyst precursors for homogeneous ethylene
polymerizationq
Erik H. Licht, Helmut G. Alt), M. Manzurul Karim1
Laboratorium f ur Anorganische Chemie, UniÕersitat Bayreuth, D-95440 Bayreuth, Germany
¨
¨
Received 7 December 1999; accepted 3 March 2000
Abstract
Ž .Ž
.
.
The synthesis and characterization of 15 new, unbridged metallocene dichloride complexes of the type L Ind ZrCl2
X
Ž
.
Ž
X.Ž
Lsv-phenylalkyl substituted cyclopentadienyl or 1-phenylsilyl substituted indenyl; Indsindenyl and L Flu ZrCl2 L s
Ž
.
1-benzyl substituted indenyl or 2-benzyl substituted indenyl; Flusfluorenyl are reported. After activation with methyl
Ž
.
aluminoxane MAO , these complexes are highly active catalysts for homogeneous ethylene polymerization. The influence
of various substituents on the catalyst activities and the polymer properties are discussed. q 2000 Elsevier Science B.V. All
rights reserved.
Keywords: Zirconium; Catalysis; Ethylene polymerization; Metallocene complexes; Substituent effect
w
x
1. Introduction
polymerization 5–24 . These complexes offer
an immense array of possibilities for varying the
ligand sphere and thus, allow specific changes
in catalytic and subsequent resin properties.
Therefore, metallocene catalysts are well suited
for studying the catalystrpolymer relationship.
The steric environment of the catalytically ac-
tive metallocene monoalkyl cation has a signifi-
cant influence on the observed polymerization
behavior.
Zirconocene dichloride complexes in combi-
nation with methyl aluminoxane MAO are
highly active catalysts for the homogeneous
polymerization of a-olefins 1–4 . In recent
years, a large number of unbridged and bridged
zirconocene dichloride complexes has been syn-
thesized as catalyst precursors, activated with a
co-catalyst and used for homogeneous olefin
Ž
.
w
x
Recently, we reported on the influence of
alkyl and silyl substituents, containing a termi-
nal phenyl group, on the polymerization behav-
q
Dedicated to Professor Herbert Schumann on the occasion of
w
x
ior of symmetric zirconocene complexes 25 .
The following question of interest remained: if
only one of the two p-ligands contained a
substituent, how would the length and type of
this substituent influence the polymerization be-
Ž
.
his 65th birthday August 06, 2000 .
)
Corresponding author. Tel.: q49-921-55-2555; fax: q49-921-
55-2157.
Ž
.
E-mail address: helmut.alt@uni-bayreuth.de H.G. Alt .
1 Alexander von Humboldt awardee from Jahangirnagar Uni-
versity, Dhaka, Bangladesh.
1381-1169r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S1381-1169 00 00205-3

(
)
274
E.H. Licht et al.rJournal of Molecular Catalysis A: Chemical 159 2000 273–283
w
x
indenyl or fluorenyl anions 26 as shown in
Fig. 1.
w
x
Cyclopentadienylzirconium trichloride 27 ,
w
x
indenylzirconium trichloride 28 and substi-
w
x
tuted indenylzirconium trichloride 29 half-
sandwich complexes are well suited for the type
of reaction shown in Fig. 1. The metallocene
dichloride complexes 1–15 were synthesized by
the reaction of phenyllithium with 6,6-dimethyl-
w
Fig. 1. Synthesis of the unbridged, mixed substituted 1-phenyl-1-
5
5
Ž
.xŽ
.
methylethyl h -cyclopentadienyl h -cyclopentadienyl zirco-
nium dichloride complex 1.
w
x
fulvene 30 or by the lithiation of v-phenylal-
w
x
kyl substituted ligand precursors 25 followed
by the reaction with the appropriate half-sand-
wich complex.
havior of the resulting catalyst and the proper-
ties of the resin produced?
2. Results and discussion
2.2. Spectroscopic characterization of the met-
(
)
allocene dichloride complexes 1–15 Fig. 2
2.1. Synthesis of the metallocene dichloride
complexes 1–15
The metallocene dichloride complexes were
characterized by H, 13C, and 29Si NMR spec-
1
Mixed substituted metallocene dichloride
complexes can be synthesized by the reaction of
unsubstituted half-sandwich complexes with the
corresponding substituted cyclopentadienyl,
troscopy. The spectroscopic data are listed in
Table 1. The mass spectroscopy data are found
in the experimental section.
Fig. 2. Overview of the synthesized metallocene dichloride complexes 1–15.

(
)
E.H. Licht et al.rJournal of Molecular Catalysis A: Chemical 159 2000 273–283
275
Table 1
1H, 13C,and 29Si NMR data of the zirconocene dichloride complexes 1–15
(
)
continued on next page

(
)
276
Table 1 continued
E.H. Licht et al.rJournal of Molecular Catalysis A: Chemical 159 2000 273–283
Ž
.

(
)
E.H. Licht et al.rJournal of Molecular Catalysis A: Chemical 159 2000 273–283
277
Ž
.
Table 1 continued
a 258C, in chloroform-d1
b258C, in chloroform-d1
c 258C, in chloroform-d1
d
d
d
ppm rel. chloroform 7.24 .
w
w
w
x
Ž
.
x
Ž
.
ppm rel. chloroform-d1 77.0 .
x
Ž
.
ppm rel. TMSext. 0.0 .
Complexes with doubly substituted cyclopen-
fusion enthalpies, D Hm, the melting tempera-
tures, Tm and the degrees of crystallinity, a, of
the resulting polymers are listed in Table 2.
Ž
.
tadienyl ligands 9–11 or mono substituted in-
Ž
.
denyl ligands substituted in position 1: 13, 14
Ž
.
can appear as isomers a, b because of the
prochirality of these ligands. In addition, the
indenyl derivatives can form rac and meso
isomers. As a consequence, the 1H and 13C
NMR spectra of these complexes can show
more than the expected number of signals.
2.4. Discussion of the polymerization results:
effect of the catalyst structure on the catalyst
properties
Principally, it is difficult to predict precise
catalyst features for different metallocene cata-
lysts. Too many parameters can determine the
activity of the catalysts and the properties of the
polymers, such as the nature of the organic
2.3. Polymerization of ethylene
The synthesized metallocene complexes, acti-
vated with MAO, were used for the homoge-
neous polymerization of ethylene. The com-
plexes were activated by adding a 3000-fold
molar excess of MAO to the toluene solution of
the corresponding complex. The formation of an
active catalyst system was indicated by a color
change.
In the following paragraphs, the experimental
polymerization results of one complex type or
family are discussed. The polymerization activi-
ties of the activated metallocene complexes, the
viscosity average molecular weights, Mh, the
Ž
.
ligands cyclopentadienyl, indenyl, fluorenyl ,
Ž
the influence of ligand substituents alkyl or
aryl groups, with or without hetero atoms , the
metal Ti, Zr, Hf , the co-catalysts MAO, PHT
31 , boranes, borates, etc. , the solvent a.o.
.
Ž
.
Ž
w
x
.
Some parameters can be contra-productive and
a compromise has to be found for the best
catalyst performance. In this investigation, we
made an empiric approach and compared the
activities of various unbridged metallocene cata-
Ž
.
lysts for ethylene polymerization Table 2 . All

(
)
278
E.H. Licht et al.rJournal of Molecular Catalysis A: Chemical 159 2000 273–283
Table 2
Overview of the polymerization experimentsa and polymer analytic results

(
)
E.H. Licht et al.rJournal of Molecular Catalysis A: Chemical 159 2000 273–283
279
Ž
.
Table 2 continued
aTp s608C; solvent: 500-ml n-pentane; 10 bar ethylene pressure.
bw x w
x
M r Al s1:3000.
c Intrinsic viscosity.
d Degree of crystallinity relative to the fusion enthalpy of 100% crystalline polyethylene.
parameters were kept constant except the struc-
ture of the catalyst precursors 1–15.
Catalyst precursors 4–7 possess an alkyl sub-
stituent of a variable chain length and a terminal
phenyl group.
The activities of the corresponding catalysts
4–7rMAO are, in all cases, higher than the
activity of the parent catalyst Cp2 ZrCl2rMAO
Ž
Ž
.
.
1490 kg PErg Zr h under the same reaction
conditions. An explanation for this behavior
Fig. 3 shows the drastic increase of activity
when the length of such an alkylidene spacer
could be a better separation of the contact ion
Ž
pair in the activated form of the catalyst metal-
Ž
.
w
x.
–CH2– increases from ns2 to ns4.
locene monomethyl cationrMAO anion 31,32
n

(
)
280
E.H. Licht et al.rJournal of Molecular Catalysis A: Chemical 159 2000 273–283
Fig. 3. Comparison of the polymerization activities for the metallocene complexes 4–7 and molecular weights of the produced
polyethylenes.
induced by an increased spacer effect of the
bulkier alkyl substituent. Zirconocene dichloride
complexes with two v-phenylalkylcyclopenta-
dienyl ligands do not show such a strong trend
The activity of catalysts 4–7rMAO only
slightly improves when an additional methyl
substituent is fixed in position 3 of the cy-
clopentadienyl ligand to give the diastereomers
9a,b–11a,b. This effect could be due to an even
better separation of the corresponding catalyst
cation and MAO anion compared with the par-
ent complexes.
w
x
with increasing spacer chain length 25 . This
positive effect on the activity becomes negative
as soon as methyl substituents are fixed on a
Ž
Ž
.
C1-spacer 2887.8 kg PErg Zr h in 4rMAO
Ž
.
.
vs. 15.2 kg PErg Zr h in 2rMAO . There is
no obvious explanation for such a drastic drop
in activity except when it is assumed that the
terminal phenyl group can undergo a better
interaction with the metal center and thus, com-
pete with the coordination of the ethylene in the
first polymerization step. A similar explanation
can be given for the drop of activity of 12rMAO
due to the electron pushing effect of the methyl
substituent in the para position of the phenyl
ring that enhances the interaction of the aro-
matic ring system with the Lewis acidic center
at the metal. Such an intramolecular interaction
of a phenyl group with the metal center is
The substitution of the silyl substituted cy-
clopentadienyl ligand of 3 for an indenyl ligand
to give the diastereomers 13a,b has only a small
Ž
positive influence on the catalyst activity 289
Ž
.
Ž
.
kg PErg Zr h in 3rMAO vs. 300 kg PErg Zr
.
h in 13a,brMAO and is, by far, lower than the
activity of the parent catalyst Ind2 ZrCl2rMAO
Ž
Ž
.
.
3200 kg PErg Zr h under identical reaction
conditions.
Ž
The low activity of 15rMAO 85.4 kg
Ž
.
.
PErg Zr h clearly results from the thermal
instability of the unbridged fluorenyl complex
15. It is known that fluorenyl complexes readily
5
w
x
undergo ring slippage reactions 36 from h
h3 h1 bonding modes and then lose the lig-
and in the final step. For this reason, unbridged
fluorenyl ligands are not attractive for catalyst
molecules.
w
x
known from similar complexes 33–35 .
This interpretation would also explain the
remarkable improvement of activity when the
C1-spacer of 2 is substituted by a Si1-spacer to
Ž
Ž
.
give 3 15.2 kg PErg Zr h in 2rMAO vs. 289
The viscosity average molecular weights of
the polyethylenes produced with the metal-
Ž
.
.
kg PErg Zr h in 3rMAO .

(
)
E.H. Licht et al.rJournal of Molecular Catalysis A: Chemical 159 2000 273–283
281
locene complexes 4–7 are in the range 500–640
kgrmol. For the metallocene complexes 5–7,
they are in a very narrow region 610–640 .
Obviously, the spacer chain length has only
little influence on the chain termination rates,
except in the case of the C1-spacer.
3.4. General synthesis procedure for the metal-
locene dichloride complexes 1 and 2
Ž
.
Ž
.
6,6-Dimethylfulvene, 1.2 ml 10.0 mmol ,
was dissolved in 150 ml diethylether at y788C
Ž
.
and mixed with 5.0 ml 10.0 mmol phenyl-
Ž
lithium 2 M solution in cyclohexanerdiethyl-
ether 70%r30% . The white suspension was
.
stirred for 4 h at room temperature. Then, the
reaction mixture was mixed with 10.0 mmol of
the corresponding half-sandwich complex at
y788C and stirred overnight at room tempera-
ture. The solvent was evaporated in vacuo, and
the residue washed twice with 50-ml n-pentane.
The residue was extracted with dichloromethane,
the solution filtered over sodium sulfate and the
3. Experimental part
3.1. NMR spectroscopy
The Bruker spectrometers ARX 250, AC 300,
and DRX 500 were available for recording the
NMR spectra. The organometallic samples were
placed into the NMR tubes under argon and
measured at 258C. The 1H NMR spectral chemi-
cal shifts are referred to the residual proton
solvent evaporated. The yields were 80–90%. 1:
q
Ž
.
MS: mres410
M
; 2: MS: mres461
q
Ž
.
.
M
Ž
signal of the solvent ds7.24 ppm for chloro-
form and in 13C NMR spectra to the solvent
.
3.5. General synthesis procedure for the metal-
locene dichloride complexes 3–15
Ž
.
signal ds77.0 ppm for chloroform-d1 .
The selected ligand precursor, 10.0 mmol,
was dissolved in 150 ml diethylether at –788C
3.2. Mass spectroscopy
Ž
.
and mixed with 6.3 ml 10.0 mmol n-butyl-
Routine measurements were performed using
Ž
.
lithium 1.6 M solution in n-hexane . The reac-
tion mixture was then stirred for 6 h at room
temperature. Then, 10.0 mmol of the corre-
sponding half-sandwich complex were added at
y788C, and the reaction mixture was stirred
overnight at room temperature. The solvent was
evaporated in vacuo, and the residue washed
twice with 50-ml n-pentane. The residue was
extracted with dichloromethane, the solution fil-
tered over sodium sulfate and the solvent evapo-
Ž
a VARIAN MAT CH7 instrument direct inlet,
.
electron impact ionization 70 eV . GCrMS
spectra were recorded using a VARIAN 3700
gas chromatograph in combination with a VAR-
IAN MAT 312 mass spectrometer.
3.3. Gas chromatography
Organic compounds were analyzed with a
Carlo-Erba HRGC gas chromatograph equipped
with a flame ionization detector. The J&W
fused silica column was 30 m long, had a
diameter of 0.32 mm, and a film thickness of
0.25 mm. Helium served as carrier gas. The
following temperature program was used:
Ž
rated in vacuo. The yields were 80–90% Table
.
3 .
3.6. ActiÕation of the metallocene complexes
with MAO
The selected metallocene complex, 10–15
Ž .
Zr:Als1:3000 . A volume of the catalyst solu-
tion, containing 0.5–1.5 mg metallocene com-
starting phase: 3 min at 508C;
mg, was activated with MAO 10% in toluene
Ž
.
Ž
.
heating phase: 58Crmin 15 min ; and
Ž
.
plateau phase: 3108C 15 min .

(
)
282
E.H. Licht et al.rJournal of Molecular Catalysis A: Chemical 159 2000 273–283
Table 3
Ž
.
from y408C to 2008C heating rate 208Crmin .
Melting points and fusion enthalpies were taken
from the second heating phase. The temperature
MS data of the compounds 3–15
Mq mre
w
x
Compound
3
4
5
6
7
8
526
270
284
298
312
326
284
498
312
284
393
533
533
Ž
Ž
was linearly corrected relative to indium Mp.
429.79 K . The fusion enthalpy of indium Hm
s28.45 Jrg was used for calibration.
.
.
9a,b
10a,b
11a,b
12
13a,b
14
3.8.2. Viscosimetry
The viscosity average molecular weight de-
termination of the polyethylene samples was
performed using an Ubbelohde precision capil-
lary viscometer in cisrtrans decalin at 135"
0.18C. For the measurements, 50 mg polymer
were completely dissolved in 45.0 ml decalin at
1308C within 3–4 h and insoluble ingredients
were filtered over glass wool. Mh was deter-
mined using a calibration curve that was avail-
able for the selected concentration.
15
plex, was used within 1 h for homopolymeriza-
tion.
3.7. Homopolymerization of ethylene
Acknowledgements
n-Pentane, 500 ml, was placed in a 1 l Buchi
laboratory autoclave, mixed with the catalyst
¨
solution and the autoclave thermostated at 608C.
We thank Deutsche Forschungsgemeinschaft
and Alexander von Humboldt-Foundation as
Ž
.
An ethylene pressure 99.98% ethylene of 10
bars was applied after an inside temperature of
508C was reached. The mixture was stirred for 1
Ž
well as Phillips Petroleum Co. Bartlesville, OK,
.
USA for the financial support and Witco for
Ž
.
h at 60 "2 8C, and subsequently, the reaction
was terminated by releasing the pressure in the
reactor. The obtained polymer was dried in
vacuo.
the supply of MAO.
References
3.8. Characterization of the polyethylene sam-
ples
w x
Ž .
M. Bochmann, J. Chem. Soc., Dalton Trans. 1996 255.
1
x
w
w
2a H.-H. Brintzinger, D. Fischer, R. Mulhaupt, B. Rieger, R.
¨
Ž
.
Waymouth, Angew. Chem. 107 1995 1255.
(
)
x
2b H.-H. Brintzinger, D. Fischer, R. Mulhaupt, B. Rieger, R.
¨
3.8.1. Differential scanning calorimetry DSC
Ž
.
Waymouth, Angew. Chem. Int. Ed. Engl. 34 1995 1143.
The polymer samples were investigated for
phase transitions with DSC using a Pyris DSC 7
instrument. For the measurements, 5–12 mg
dried polymer were fused into standard alu-
w x
Ž .
M. Kaulbach, F. Kuber, Chem. Unserer Zeit 28 1994 197.
¨
Ž .
W. Kaminsky, J. Chem. Soc., Dalton Trans. 1998 1413.
3
w x
4
w x
Ž .
P.C. Mohring, N.J. Coville, J. Mol. Catal. 96 1995 181.
¨
5
w x
6
P.C. Mohring, N. Vlachakis, N.E. Grimmer, N.J. Coville, J.
¨
Ž
.
Organomet. Chem. 483 1994 159.
Ž
.
minum pans 0 7 mm and measured under
nitrogen using the following temperature pro-
gram: first heating phase: from 258C to 2008C
w x
Ž .
7
R. Choukroun, F. Dahan, Organometallics 13 1994 2097.
w x
8
H. Sitzmann, P. Zhou, G. Wolmershauser, Chem. Ber. 127
¨
Ž
.
1994 3.
w x
9
R.D. Rogers, M.M. Benning, L.K. Kurihara, K.J. Moriarty,
Ž
.
Ž
heating rate 40 Krmin , isothermal phase 3
Ž
.
M.D. Rausch, J. Organomet. Chem. 293 1985 51.
.
min , cooling phase from 2008C to y408C
w
x
10 G. Schmid, U. Thewalt, M. Polasek, K. Mach, P. Sedmera, J.
Ž
.
Ž .
cooling rate 208Crmin , second heating phase
Organomet. Chem. 482 1994 231.

(
)
E.H. Licht et al.rJournal of Molecular Catalysis A: Chemical 159 2000 273–283
283
w
x
w
w
x
11 M.A. Schmid, H.G. Alt, W. Milius, J. Organomet. Chem.
24 R. Fierro, Z. Yu, M.D. Rausch, S. Dong, D. Alvares, J.C.W.
Ž
.
Ž
.
501 1995 101.
Chien, J. Polym. Sci., A 32 1994 661.
w
w
x
x
Ž
.
x
12 A. Razavi, J. Ferrara, J. Organomet. Chem. 435 1992 299.
13 J.A. Ewen, R.L. Jones, A. Razavi, J. Ferrara, J. Am. Chem.
25 E.H. Licht, H.G. Alt, M.M. Karim, J. Organomet. Chem. 599
Ž
.
2000 275.
Ž
.
w
w
x
x
Soc. 110 1988 6255.
26 M. Schmid, Dissertation, University of Bayreuth, 1994.
27 G. Erker, K. Berg, L. Treschanke, K. Engel, Inorg. Chem. 21
w
x
Ž
14a J.A. Ewen, A. Razavi, Eur. Pat. Appl. EP 351, 392 17
.
Ž
.
January 1990 .
1982 1277.
w
w
x
Ž
.
w
w
w
x
14b J.A. Ewen, A. Razavi, Chem. Abstr. 112 1990 199346.
28 C. Schmid, Dissertation, University of Bayreuth, 1996.
x
x
15 A. Razavi, J.L. Atwood, A. Razavi, J. Ferrara, J. Organomet.
29 M. Deppner, unpublished results.
Ž
.
x
Ž
.
Chem. 459 1993 117.
30 K.J. Stone, R.D. Little, J. Organomet. Chem. 49 1984
1849.
w
x
16 W. Kaminsky, O. Rabe, A.-M. Schauwienold, G.U. Schupfer,
Ž
.
w
x
J. Hanss, J. Kopf, J. Organomet. Chem. 497 1995 181.
31 A. Koppl, H.G. Alt, M.D. Phillips, J. Appl. Polym. Sci.,
¨
w
w
x
x
Ž
.
17 J.A. Ewen, Macromol. Symp. 89 1995 181.
18 Y.-X. Chen, M.D. Rausch, J.C.W. Chien, J. Organomet.
submitted for publication.
w
x
Ž
.
32 P. Schertl, H.G. Alt, J. Organomet. Chem. 582 1999 328.
Ž
.
w x
33 L.H. Doerrer, M.L.H. Green, D. Haußinger, J. Saßmannshau-
¨
Chem. 497 1995 1.
w
w
w
x
Ž
.
Ž .
sen, J. Chem. Soc., Dalton Trans. 1999 2111.
19 K. Patsidis, H.G. Alt, J. Organomet. Chem. 501 1995 31.
20 H.G. Alt, R. Zenk, J. Organomet. Chem. 512 1996 51.
21 H.G. Alt, R. Zenk, W. Milius, J. Organomet. Chem. 514
x
Ž
.
w
x
34 E.H. Licht, H.G. Alt, M.M. Karim, J. Organomet. Chem. 599
x
Ž
.
2000 261.
Ž
.
w
x Ž .
x
1996 257.
35 E.H. Licht, H.G. Alt, J. Mol. Catal. A 153 2000 109.
w
w
w
x
Ž
.
22 G.S. Herrmann, H.G. Alt, M.D. Rausch, J. Organomet. Chem.
36 H.G. Alt, E. Samuel, Chem. Soc. Rev. 27 1998 323, and
references therein.
Ž
.
401 1991 C5.
x
Ž
.
23 H.G. Alt, J. Chem. Soc., Dalton Trans. 1999 1703.