Mechanisms of stereocontrol for doubly silylene-bridged Cs- and Cl-symmetric zirconocene catalysts for propylene polymerization. Synthesis and molecular structure of Li2[(1,2-Me2Si)2{C5H <s

Article abstract of DOI:10.1021/ja982868j

Doubly [SiMe₂]-bridged metallocenes (1,2-SiMe₂)₂{η⁵-C₅H ₂-4-R}{η⁵-C₅H-3,5-(CHMe₂) ₂}ZrCl₂ (R = H (1a), CHMe₂ (1b), SiMe

Full text of DOI:10.1021/ja982868j

564
J. Am. Chem. Soc. 1999, 121, 564-573
Mechanisms of Stereocontrol for Doubly Silylene-Bridged Cs- and
C1-Symmetric Zirconocene Catalysts for Propylene Polymerization.
Synthesis and Molecular Structure of
Li2[(1,2-Me2Si)2{C5H2-4-(1R,2S,5R-menthyl)}{C5H-3,5-(CHMe2)2)}]‚3THF
and
[(1,2-Me2Si)2{η5-C5H2-4-(1R,2S,5R-menthyl)}{η5-C5H-3,5-(CHMe2)2}]ZrCl2
Dario Veghini,† Lawrence M. Henling,† Terry J. Burkhardt,‡ and John E. Bercaw*,†
Contribution from the Arnold and Mabel Beckman Laboratories of Chemical Synthesis, California Institute
of Technology, Pasadena, California 91125, and Baytown Polymer Center, Exxon Chemical Company,
Baytown, Texas 77520-2101
ReceiVed August 10, 1998
Abstract: Doubly [SiMe2]-bridged metallocenes (1,2-SiMe2)2{η5-C5H2-4-R}{η5-C5H-3,5-(CHMe2)2}ZrCl2 (R
) H (1a), CHMe2 (1b), SiMe3 (1c), CHMe(CMe3) (1d), (+)-menthyl (1e)), when activated by methylalumi-
noxane (MAO), catalyze propylene polymerization with high activities. The preparations and X-ray structures
of the dilithio salt of an enantiopure, doubly silylene-bridged bis(cyclopentadienyl) ligand, Li2[(1,2-Me2Si)2-
{C5H2-4-(1R,2S,5R-menthyl)}{(C5H-3,5-(CHMe2)2)}]‚3THF, as well as the corresponding zirconocene dichlo-
ride, [(1,2-Me2Si)2{η5-C5H2-4-(1R,2S,5R-menthyl)}{η5-C5H-3,5-(CHMe2)2}]ZrCl2 (1e), are reported. The Cs-
symmetric systems 1a-c are highly regiospecific and syndiospecific (>99.5%) in neat propylene. At lower
propylene concentrations, polymers with lower molecular weights and tacticity (mostly m-type stereoerrors)
are obtained. The microstructures of polymers produced under differing reaction conditions are consistent
with stereocontrol dominated by a site epimerization process, an inversion of configuration at zirconium resulting
from the polymer chain swinging from one side of the metallocene wedge to the other without monomer
insertion. The relative importance of chain epimerization (at the â carbon) has been established by parallel
polymerization of 2-d1-propylene and d0-propylene with 1b/MAO at low propylene concentrations. The C1-
symmetric systems 1d,e/MAO display an unusual dependence of stereospecificity on propylene concentration,
switching from isospecific to syndiospecific with increasing propylene pressure, consistent with a competitive
unimolecular site epimerization process and a bimolecular chain propagation. The microstructures of the
polypropylenes produced by 1d/MAO and 1e/MAO with [r] ≈ 50% resemble the hemiisotactic microstructure
produced by Me2C(η5-C5H3-3-Me)(η5-C13H8)ZrCl2 (2b)/MAO. Contrastingly, the hemiisotactic polypropylene
microstructure obtained with 2b/MAO is found to be maintained at all propylene concentrations examined.
Introduction
Cs-Symmetric precatalysts such as 1a-c and 2a enchain
R-olefins with regularly alternating enantiofaces and conse-
quently produce syndiotactic polymer.3,4 Methyl-substituted 2b
generates polypropylene, whose microstructure closely ap-
proximates hemiisotactic, one in which every other propylene
has been enchained with the same enantioface and the interven-
ing one stereorandomly. Syndiospecificity is believed to arise
from three key features: (1) the Cs-symmetric ansa-metallocene
structure has one cyclopentadienyl possessing bulky substituents
flanking the center of the metallocene wedge to direct the
Polymerization of R-olefins by metallocene systems is one
of the most efficient and selective catalytic reactions, commonly
displaying activities and stereoselectivities approaching those
of enzymes.1 Intense efforts have been undertaken to elucidate
the mechanistic details of the process in order to correlate
polymer properties (tacticity, molecular weight, etc.) with the
molecular structure of the catalyst counterpart. C2-symmetric
metallocene catalysts enchain selectively one of the R-olefin
enantiofaces to produce highly isotactic polymers. C1-Symmetric
metallocenes may also be isospecific but usually to a lesser
extent.2,4c
(3) (a) Herzog, T. A.; Zubris, D. L.; Bercaw, J. E. J. Am. Chem. Soc.
1996, 118, 11988. (b) Henling, L. M.; Herzog, T. A.; Bercaw, J. E. Acta
Crystallogr., submitted. (c) Herzog, T. A. Ph.D. Thesis, California Institute
of Technology, 1997. (d) Veghini, D.; Henling, L. M.; Bercaw, J. E. Inorg.
Chim. Acta 1998, 280, 226.
† California Institute of Technology.
‡ Exxon Chemical Co.
(1) Reviews: (a) Brintzinger, H.-H.; Fischer, D.; Mu¨lhaupt, R.; Rieger,
B.; Waymouth, R. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 1143. (b)
Horton, A. Trends Polym. Sci. 1994, 2 (5), 158. (c) Bochmann, M. J. Chem.
Soc. Dalton Trans. 1996, 255. (d) Fink, G., Mu¨lhaupt R., Brintzinger H.
H., Eds. Ziegler Catalysis, Recent Scientific InnoVation and Technological
ImproVements; Springer-Verlag: Berlin, 1995.
(4) (a) Ewen, J. A.; Jones, R. L.; Razavi, A.; Ferrara, J. D. J. Am. Chem.
Soc. 1988, 110, 6255. (b) Ewen, J. A.; Elder, M. J.; Jones, R. L.; Curtis,
S.; Cheng, H. N. Stud. Surf. Sci. Catal. 1990, 56, 439. (c) Ewen, J. A.;
Elder, M. J.; Jones, L.; Haspeslagh, L.; Atwood, J. L.; Bott, S. G.; Robinson,
K. Makromol. Chem., Makromol. Symp. 1991, 48/49, 253. (d) Ewen, J. A.;
Elder, M. J. Makromol. Chem. Makromol Symp. 1993, 66. (e) Ewen, J. A.;
Elder, M. J. Eur. Pat. Appl. EP-A 0 537 130, 1993. (f) Ewen, J. A.
Makromol. Chem., Makromol. Symp. 1995, 181.
(2) Giardello, M. A.; Eisen, M. S.; Stern, C. L.; Marks, T. J. J. Am.
Chem. Soc. 1993, 115, 3326.
10.1021/ja982868j CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/31/1998

Doubly Silylene-Bridged Zirconocene Catalysts
J. Am. Chem. Soc., Vol. 121, No. 3, 1999 565
Scheme 1
Scheme 2
polymer chain segment (up) toward the less substituted cyclo-
pentadienyl, whether it resides on the left or right side in the
transition structure for propagation (Scheme 1); (2) an open
region between the bulky substituents on the lower cyclopen-
tadienyl to accommodate the R-olefin alkyl (methyl for propy-
lene), which is directed by the polymer segment trans (down),
the dominant steric interaction in the transition structure;5,6 and
(3) migratory insertions that result in regular alternation of the
monomer approach from the left and right side of the metal-
locene wedge.
Our laboratory recently reported the preparation of Cs- and
C1-symmetric catalysts 1a-d along with some preliminary data
on their polymerization behavior.3a The Cs-symmetric precata-
lysts 1a-c, when activated with methylaluminoxane, display
high activity and syndioselectivity and produce high-molecular-
weight polymers. On the other hand, C1-symmetric systems
display a remarkable dependence of stereospecificity with
increasing propylene concentration, switching from isospecific
to syndiospecific. In this contribution, we report the results of
some recent investigations on the mechanisms of stereocontrol
operating in these catalytic systems.
Scheme 3
Results and Discussion
Dependence of Stereoselectivity on Monomer Concentra-
tion in Cs-Symmetric 1a-c/MAO. Although the stereospeci-
ficities of catalysts derived from 1a-c and 2a are high at low
temperatures in liquid propylene, under other conditions pro-
pylene enchainment does not occur with perfectly alternating
olefin enantiofaces, and isolated m or mm stereoerrors are
produced. Three types of stereoerror mechanisms are generally
considered for these Cs-symmetric systems (Schemes 2 and 3):
(1) enantiofacial misinsertion, giving an [mm] triad, the number
of which is independent of the concentration of propylene
(Scheme 2); (2) site epimerization (chain migration without
insertion), giving an isolated m stereoerror (Scheme 2); and (3)
chain epimerization,7-9 giving either m or mm stereoerrors,
depending on the whether stereochemistry at the metal center
is inverted or retained (Scheme 3). Since the epimerization
processes are unimolecular and compete with bimolecular
propagation, the frequency of stereoerrors produced by processes
2 and 3 should depend on propylene concentration.
There are some curious differences in the behavior of the
singly linked fluorenyl/cyclopentadienyl catalyst system 2. The
numbers of isolated m stereoerrors and [mm] triads produced
by catalyst 2a/MAO have been reported to increase with
decreasing propylene concentration. Some residual [mm] triads
were observed at all propylene concentrations examined.4 By
contrast, the stereoselectivity displayed by C1-symmetric 2b/
MAO was found to be relatively constant under all conditions
studied,10 and the observed polymer microstructure, closely
resembling hemiisotactic, was rationalized by invoking a regular
alternation of stereospecific olefin insertion (with propylene
adding to the more hindered side of the metallocene wedge),
followed by aspecific insertion (with propylene adding to the
less hindered side).11,12 Competitive site epimerization and chain
propagation were invoked in the other cases.13 Thus, the relative
rates of propagation vs site epimerization for catalyst systems
(5) (a) Corradini, P.; Guerra, G.; Vacatello, M.; Villani, V. Gazz. Chim.
Ital. 1988, 118, 173. (b) Corradini, P.; Guerra, G.; Cavallo, L.; Moscardi,
G.; Vacatello, M. Ziegler Catalysts. In Ziegler Catalysis, Recent Scientific
InnoVation and Technological ImproVements; Fink, G., Mu¨lhaupt R.,
Brintzinger H. H., Eds.; Springer-Verlag: Berlin, 1995; p 237. (c) Guerra,
G.; Corradini, P.; Cavallo, L.; Vacatello, M. Makromol. Chem., Makromol.
Symp. 1995, 89, 77.
(6) Gilchrist, J. H.; Bercaw, J. E. J. Am. Chem. Soc. 1996, 118, 12021.
(7) (a) Busico, V.; Cipullo, R. J. Am. Chem. Soc. 1994, 116, 9329. (b)
Busico, V.; Caporaso, L.; Cipullo, R.; Landriani, L. J. Am. Chem. Soc.
1996, 118, 2105. (c) Busico, V.; Brita, D.; Caporaso, L.; Cipullo, R.;
Vacatello, M. Macromolecules 1997, 30, 3971.
(10) Herfert, N.; Fink, G. Makromol. Chem., Makromol. Symp. 1993,
66, 157.
(8) Resconi, L.; Fait, A.; Piemontesi, F.; Colonnesi, M.; Rychlicki, H.;
Ziegler, R. Macromolecules 1995, 28, 6667.
(9) Leclerc, M. K.; Brintzinger, H.-H. J. Am. Chem. Soc. 1995, 117,
1651; 1996, 118, 9024.
(11) Farina, M.; DiSilvestro, G.; Sozzani, P. Macromolecules 1993, 26,
946.
(12) Cavallo, L.; Guerra, G.; Vacatello, M.; Corradini, P. Macromolecules
1991, 24, 1784.

566 J. Am. Chem. Soc., Vol. 121, No. 3, 1999
Veghini et al.
Table 1. Propylene Polymerization Data with 1a-c/MAO at 24 (
1 °C in Toluene Solution (Entry Numbers Are Correspondingly the
Same as for Table 2)
stereoerror mechanism. The strong dependence of the isolated
m stereoerrors with [propylene] further suggests that site
epimerization and olefin insertion occur at competitive rates over
this range of monomer concentrations, tentatively permitting
us to assign the major stereoerror-producing process to site
epimerization (Scheme 2). The remaining errors would then be
attributable to a less frequent chain epimerization, giving rise
to a few percent mm-type, and becoming apparent only at the
lowest propylene concentrations. Nonetheless, these results do
not allow us to exclude another possibility: that the isolated m
stereoerrors arise from the combination of chain epimerization
and site epimerization (Scheme 3), i.e., that chain epimerization
occurs mostly with simultaneous site epimerization (giving an
isolated m) and only occasionally without site epimerization
(giving rise to an mm stereoerror).
We therefore further examined the importance of chain
epimerization by performing a labeling study analogous to the
those described by Brintzinger and Busico.7,9 The number of
stereoerrors (isolated m and mm types, Scheme 3) that arise from
chain epimerization should decrease in proportion to the kinetic
deuterium isotope effect for â-H elimination, (k1)L (L ) H, D)
or, alternatively, in proportion to the preequilibrium deuterium
isotope effect for â-H elimination, (K1)L (Scheme 4). Parallel
polymerization reactions using d0-propylene and 2-d1-propylene
as a substrates were thus carried out, and the microstructures
of the poly-d0-propylene and poly-2-d1-propylenes so obtained
were compared.
The good precatalyst candidate for the experiments comparing
unlabeled and deuterium-labeled propylenes appeared to be 1b,
since at [propylene] ) 0.8 M it produces a large number of
mm and isolated m stereoerrors. The reactions were performed
in toluene solution saturated with propylene (1 atm, [propylene]
) 0.8 M) at room temperature. As can be seen from the results
presented in Table 3, although the molecular weights differ,
the poly-2-d1-propylene and poly-d0-propylene samples so
obtained have essentially the same microstructures within
experimental error ({1H}13C NMR). The number-average mo-
lecular weight of poly-d0-propylene (32 800) vs poly-2-d1-
propylene (56 800) implies a kinetic deuterium isotope effect
for chain transfer k(â-H)/k(â-D) ) 1.6, assuming â-H transfer
as the predominant chain transfer pathway.
[C3H6]
entry catalyst (M)
[r]
Tm
Mw/
Mn
activitya
(%) (°C)
Mw
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1a
1a
1a
1a
1a
1b
1b
1b
1b
1b
1c
1c
1c
1c
1c
0.8
1.5
2.1
2.8
3.4
0.8
1.5
2.1
2.9
3.4
0.8
1.5
2.1
2.8
3.4
3 × 103 b
90.6 109 2.9 × 104 1.9
1.3 × 104 b 95.0 125 ndc
nd
2. 9 × 104 b 94.7 131 6.1 × 104 3.0
3.1 × 104 b 95.9 139 9.0 × 104 4.2
5.3 × 104 b 97.1 140 1.01 × 105 1.5
1.9 × 104
1.9 × 104
3.5 × 104
5.9 × 104
80.2 nd
3.4 × 104 2.5
5.9 × 104 2.4
1.46 × 105 1.7
91.8
92.7
d
d
95.1 118 1.15 × 105 3.0
1.04 × 105 97.1 127 1.66 × 105 1.8
7 × 103
81.5
86.7
d
d
3.8 × 104 1.9
4.9 × 104 2.2
2.4 × 104
4.3 × 104
5.4 × 104
5.7 × 104
90.7 102 7.7 × 104 1.7
93.6 119 nd
nd
96.5 124 1.33 × 105 1.8
a Defined as grams of polymer per gram of Zr per hour. b A different
sample of methylalumoxane cocatalyst was employed for entries 1-5.
The activity data are therefore not comparable with entries 6-15. c Not
determined. d No melting point observed by DSC.
2a/MAO and 2b/MAO appear to depend explicitly on the nature
of the substituent R.
To quantify the relative contributions of [propylene]-depend-
ent and [propylene]-independent stereoerrors for catalyst systems
1a-c/MAO, polymerizations were carried out at 24 ( 1 °C in
toluene solution at various propylene pressures to effect a
propylene concentration range 0.8-3.4 M. The reproducibilities
of the polymer yield and molecular weight, as well as the
influence of [MAO]/[Zr] ratio on activity and polymer tacticity,
were investigated. Polymerization reactions carried out under
the conditions described gave reproducible r diad contents
({1H}13C NMR) with composite errors amounting to ap-
proximately (2%. The [MAO]/[Zr] ratio significantly influences
the catalyst activity, but the polymer tacticity is unaffected, at
least over the range examined.14 No regioerrors (“2,1” or “3,1”
errors) with any of the catalysts under any conditions investi-
gated were detected by 13C NMR spectroscopy, as was reported
earlier for catalyst 2a.3a,4
The variations of [r] diad contents with varying propylene
concentration are given in Table 1 and Figure 1. The stereo-
selectivities for all three catalysts 1a-c/MAO are comparable
at [propylene] > 3.5 M but differ significantly at lower
concentrations. As the [propylene] increases, the [r] diad content
rises moderately for 1a and more significantly for 1b and 1c.
The mm and (substantially more) m stereoerrors follow a similar
trend (Figure 2), decreasing with increasing [propylene] over
the entire range of concentrations examined. At low propylene
concentrations, 2-3% mm stereoerrors are detected for all
catalysts, while at higher concentration these are no longer
detectable for 1b and 1c, accounting for less than 1% for 1a.
Thus, stereoselectivity increases with propylene concentration,
indicating that enantiofacial misinsertion is not the major
The {1H}2H NMR spectrum of poly-2-d1-propylene displays
two broad signals at 1.4 and 0.5 ppm in an approximately 50:1
ratio.15 We tentatively assign the former to -CD(CH3)- groups
and the latter to ca. 2% -CH(CH2D)- groups arising from chain
epimerization (Scheme 4). This assignment is in agreement with
the finding of 2-3% mm stereoerrors for 1b/MAO under similar
conditions, implicating chain epimerization as the major source
of the mm stereoerrors. More importantly, the observation that
poly-2-d1-propylene has essentially the same microstructure (in
fact, slightly more (19.6% vs 17.3%), not fewer [rmrr] pentads,
as would be expected from a normal deuterium isotope effect
for â-H elimination), together with the observed dependence
of stereospecificity on propylene concentration, allows us to
confidently assign site epimerization as the principal process
for stereoerror formation with the catalyst system 1a-c/MAO.
Temperature Effects on the Polymerization Behavior of
1a-c/MAO Catalyst Systems. A significant temperature influ-
ence on stereoselectivity is expected for syndiospecific catalyst
systems 1a-c/MAO, where a first-order stereoerror process (site
(13) (a) Rieger, B.; Jany, G.; Fawzi, R.; Steimann, M. Organometallics
1994, 13, 647. (b) Gauthier, W. J.; Collins, S. Macromolecules 1995, 28,
3779. (c) Jany, G.; Repo, T.; Gustafsson, M.; Leskela¨, M.; Polamo, M.;
Kliga, M.; Dietrich, U.; Rieger, B. Chem. Ber./Recl. 1997, 130, 747. (d)
Waymouth, R. Oral presentation at the Meeting “Advances in Polyolefin”,
Napa Valley, CA, 1997. (e) For a related example in a heterogeneous system,
see: Busico, V.; Cipullo, R.; Talarico, G.; Segre, A. L.; Chadwick, J. C.
Macromolecules 1997, 30, 4786.
(14) [MAO]/[Zr] ratios in the range from 50:1 to 15 000:1 were
investigated. The activity reaches a maximum about 7000:1, but the [r]
content is found to be essentially independent ((3%) of the Al/Zr molar
ratio.
(15) The {1H}2H NMR spectra were recorded at 80 °C in C6H3Cl3/C6H6
solution containing a small amount of benzene-d6, externally lock-shimmed
in the 1H channel on a separate C6H3Cl3/C6D6 sample having no polymer.
The 2H chemical shifts (100 transients) are referenced to the small amount
of benzene-d6.

Doubly Silylene-Bridged Zirconocene Catalysts
J. Am. Chem. Soc., Vol. 121, No. 3, 1999 567
Table 2. Pentad Distributions (%) for the Polypropylene Samples Produced by 1a-c/MAO at 24 ( 1 °C in Toluene Solution (Entry Numbers
Are Correspondingly the Same as for Table 1)
entry
catalyst
[mmmm]
[mmmr]
[rmmr]
[mmrr]
[mrmm]+ [rmrr]
[mrmr]
[rrrr]
[mrrr]
[mrrm]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1a
1a
1a
1a
1a
1b
1b
1b
1b
1b
1c
1c
1c
1c
1c
1.7
<1
<1
<1
<1
3.0
3.6
1.5
2.1
2.0
1.9
6.0
2.3
1.4
1.3
11.6
5.1
4.7
3.5
2.7
19.9
12.6
10.3
8.6
68.6
79.5
83.6
86.1
90.7
41.5
68.2
71.6
82.6
89.0
47.2
59.0
68.9
75.8
81.6
14.4
13.5
9.0
7.7
4.2
20.4
17.0
13.3
7.4
1.4
1.2
5.1
<1
2.5
2.6
<1
<1
5.5
5.5
<1
2.4
1.7
<1
5.3
3.6
2.1
2.4
18.8
16.7
13.0
7.8
4.2
3.4
1.3
18.0
15.0
13.5
12.7
11.3
1.7
1.2
7
Table 3. Polymer Analyses for Poly-d0-propylene and
Poly-2-d1-propylene Produced by 1b/MAO in Toluene Solution at 1
atm Propylene, 25 °C
[r]
[rmrr]
[rmmr]
(%)
Mw/
Mn
polymer
(%) ([mrmm]+) (%)
Mn
poly-d0-propylene
poly-2-d1-propylene 84.2
83.3
17.3
19.6
2.2
1.6
32 840 1.9
56 830 1.8
Table 4. Propylene Polymerization Data with 1a-e and
(Ph2C)[(η5-C5H4)(η5-C13H8)]ZrCl2 (3)/MAO in Liquid Propylene
(Entry Numbers Are Correspondingly the Same as for Table 5)
T
[r]
Tm
Mw/
Mn
entry catalyst (°C)
activitya
(%) (°C)
Mw
Figure 1. Dependence of [r] diads on propylene concentration for Cs-
symmetric catalysts 1a-c/MAO.
16
17
18
19
20
21
22
23
24
25
26
27
28c
29c
1a
1a
1a
1b
1b
1b
1c
1c
1c
1d
1e
3
20
50
2.8 × 106 97.1 151 1.25 × 106 1.9
7.9 × 106 96.6 140 3.3 × 105
2.3
3.2
2.0
2.0
2.3
1.8
2.1
2.2
2.2
1.9
2.3
2.0
2.0
70 16.9 × 106 94.1 119 1.6 × 105
20
50
70
20
50
70
20
20
20
40
60
3.2 × 105 99.4 151 9.8 × 105
1.4 × 106 94.5 123 2.9 × 105
9.2 × 105 90.6
b
1.3 × 105
3.5 × 105 97.9 152 7.9 × 105
8.3 × 105 96.0 124 2.6 × 105
7.3 × 105 91.3
b
1.4 × 105
4.7 × 105
2.5 × 105 75.4 102 1.9 × 105
4.9 × 105 78.7
b
1.3 × 105 99.2 147 8.4 × 105
2.0 × 104 c 95.6 137 3.8 × 105
9.0 × 104 c 94.4 132 2.7 × 105
3
3
a Defined as grams of polymer per gram of Zr per hour. b No melting
point observed by DSC. c These two runs were carried out earlier with
a differing catalyst:MAO ratio (1:325) and with propylene of generally
lower purity that for the other runs.
Figure 2. Dependence of [rmrr] and [rmmr] pentads on propylene
concentration for Cs-symmetric catalysts 1a-c/MAO.
typically characterized by a much more negative ∆S‡ and
commensurably smaller ∆H‡ as compared with the first-order
process. Thus, we expect that site epimerization (larger ∆H‡)
would become more important (relative to propagation with
smaller ∆H‡) at higher temperatures, and a strong temperature
dependence of stereospecificity would obtain.
Scheme 4
To establish the influence of temperature on activity and
stereospecificity for catalysts 1a-c/MAO, polymerization reac-
tions were performed in neat propylene at 20, 50, and 70 °C
(Tables 4 and 5). A singly bridged cyclopentadienyl/fluorenyl
catalyst system,16 Ph2C(η5-C5H4)(η5-C13H8)ZrCl2 (3)/MAO,16b
was also examined for comparison. Catalyst 1a is the most active
system of the series, producing about 1.6 × 107 g of polymer
g of polymer g-1 Zr h-1, as compared with less than 1 × 106
g of polymer g-1 Zr h-1 for the others at 70 °C. Moreover,
epimerization) competes with a second-order chain propagation
process. Under conditions where a first-order process and a
second-order process have comparable rates, and hence com-
parable activation free energies, the second-order process is
(16) (a) Razavi, A.; Atwood, J. L. J. Organomet. Chem. 1993, 459, 117.
(b) Winter, J.; Rohrmann, M.; Dolle, V.; Spaleck, W. Eur. Pat. Appls. 0,-
387,690 and 0.387,691, 1991.

568 J. Am. Chem. Soc., Vol. 121, No. 3, 1999
Veghini et al.
Table 5. Pentad Distribution (%) for the Polypropylene Produced by 1a-e and (Ph2C)[(η5-C5H4)(η5-C13H8)]ZrCl2 (3)/MAO in Liquid
Propylene (Entry Numbers Are Correspondingly the Same as for Table 4)
entry
catalyst
[mmmm]
[mmmr]
[rmmr]
[mmrr]
[mrmm] + [rmrr]
[mrmr]
[rrrr]
[mrrr]
[mrrm]
16
17
18
19
20
21
22
23
24
25
26
27
28a
29a
1a
1a
1a
1b
1b
1b
1c
1c
1c
1d
1e
3
0.8
<1
<1
-
1.6
1.8
0.3
-
1.5
2.0
3.0
-
1.9
2.4
0.6
1.2
3
14.4
12.9
1.2
3.1
3.4
0.9
2.4
7.3
1.2
4.2
12.9
1.6
4.8
13.6
6.6
11.4
93.4
90.6
79.5
97.5
81.0
68.4
94.2
83.8
65.8
49.5
45.4
97.5
83.6
81.1
3.5
4.9
10.0
1.2
11.1
14.3
3.3
10.1
17.1
17.5
13.7
1.2
<1
<1
<1
<1
4.0
6.0
1.3
3.3
4.4
4.5
2.3
2.8
3
3
1.5
1.7
2.0
2.6
0.8
0.9
9.1
9.3
0.5
0.6
a These two runs were carried out earlier with a differing catalyst:MAO ratio (1:325) and with propylene of generally lower purity than that for
the other runs.
Scheme 5
while the activities for 1b and 1c decrease on going from 50 to
70 °C (probably due to partial catalyst decomposition at elevated
temperatures), 1a reaches its maximum at 70 °C. No regioerrors
are detected at any of the temperatures examined.
Stereocontrol is generally high at 20 °C ([r] ) >99%), but
a moderate decrease in tacticity at higher temperatures for 1a
is more pronounced for 1b and 1c. In all cases, at T ) 70 °C,
isolated [m] diads constitute the main type of stereoerrors present
in the polymer microstructure ([rmrr] ) 7-12% for all
catalysts), while [rmmr] pentads account for less than 3% of
the stereoerrors. At 20 °C, the stereocontrol displayed by 1b is
comparable to that found for 3, slightly better than those for 1a
and 1c. Assuming an equal number of active sites for all
catalysts, a considerably higher rate of propagation for 1a is
also indicated.
The polymerization experiments in neat propylene at 20 °C
structure20 was determined by an X-ray diffraction study21
(Figure 3). The very soluble dilithio salt 7 may be metalated
with 1a-c/MAO and 3/MAO produced syndiotactic polymers
with high melting points (Tm > 150 °C for 1a-c/MAO, Tm
)
successfully using ZrCl4 in situ (see Experimental Section),
147 °C for 3/MAO). The molecular weights are in the order of
1 000 000 decreasing to about 100 000 at 70 °C (PDIs ) 1.8-
3.2).
1
affording 1e (eq 1). The H NMR spectral data and the X-ray
crystal structure data for 1e22 (Figure 4) confirmed the molecular
C1-Symmetric Systems. In a preliminary report,3a we noted
the exceptional behavior of C1-symmetric 1d, which appeared
to switch from isospecific to syndiospecific with increasing
propylene concentrations.17 Another C1-symmetric system was
therefore prepared to examine the scope of this “ambispecificity”
for systems bearing a chiral substituent at the central position
of the (upper) cyclopentadienyl ligand. Enantiopure (+)-menthyl
was chosen for the preparation of doubly [SiMe2]-bridged
zirconocene dichloride 1e. Although we anticipated no differ-
ence in the performance of a catalyst having an enantiopure
substitutent rather than a racemic one, the availability of the
precursor to 1e only in enantiopure form led us to incorporate
(+)-menthyl.
geometry shown in eq 1.
(1R,2S,5R)-Menthylcyclopentadiene was prepared according
to a procedure closely analogous to that reported by Marks and
co-workers.18 The steps shown in Scheme 5 proceeded in high
yields, giving 6 as a high-boiling oil.19 The dilithio salt Li2[(1,2-
Me2Si)2{η5-C5H2-4-(1R,2S,5R-menthyl)}{η5-C5H-3,5-(CHMe2)2}]‚
3THF (7) was crystallized from THF/petroleum ether, and its
Dependence of Stereoselectivity on Monomer Concentra-
tion with the 1d,e/MAO Catalyst Systems. Polymerization
reactions were performed in toluene solution in the propylene
(19) Because of the rapid fluxionality of neutral, silyl-substituted
cyclopentadienyl compounds 5 and 7, the NMR spectra are complex and
thus are not diagnostic (see Jutzi, P. Chem. ReV. 1988, 86, 983). More
informative NMR analyses could be performed on the lithium salts of these
derivatives.
(20) (a) Hiemeier, J.; Ko¨hler, F. H.; Mu¨ller, G. Organometallics 1991,
10, 11787. (b) Siemeling, U.; Jutzi, P.; Neuman, B.; Stammler, H.-G.
Organometallics 1992, 11, 1328.
(17) For a catalyst system that switches from moderately isospecific to
moderately syndiospecific as a function of temperature, see: Erker, G.;
Fritze C. Angew. Chem., Int. Ed. Engl. 1992, 31, 199.
(18) (a) Giardello, M. A.; Conticello, V. P.; Brard, L.; Sabat, M.;
Rheingold, A. L.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1994, 116,
10212. (b) Gagne´, M. R.; Brard, L.; Conticello, V. D.; Giardello, M. A.;
Stern, C.; Marks, T. J. Organometallics 1992, 11, 2003.

Doubly Silylene-Bridged Zirconocene Catalysts
J. Am. Chem. Soc., Vol. 121, No. 3, 1999 569
Figure 3. Molecular structure of [Li]2[(Me2Si)2{η5-C5H2-4-(1R,2S,5R-
menthyl)(η5-C5H-3,5-(CHMe2)2)]‚3THF (8). One of the molecules
contained in the asymmetric unit is shown (molecule A), the other being
essentially isostructural. Hydrogen atoms and some carbon atom
labeling is omitted for clarity.
Figure 4. Molecular structure of (Me2Si)2[η5-C5H2-4-(1R,2S,5R-
menthyl)] [η5-C5H-3,5-(CHMe2)2]ZrCl2 (1e). Hydrogen atoms and
carbon atom labeling are omitted for clarity.
Table 6. Propylene Polymerization Data with 1d,e/MAO in
Toluene Solution (Entry Numbers Are Correspondingly the Same as
for Table 7)
concentration range 0.5-4.6 M at 25 °C. The results are given
in Tables 6 and 7. As can be seen in Figure 5, there is a nearly
linear increase of [r] diad content with increasing propylene
concentration over this range for both 1d and 1e, the former
displaying a slightly stronger response than the latter. Contrast-
ingly, the tacticity for polypropylene prepared with the 2b/MAO
catalyst system remains essentially constant over the entire
concentration range.23 The comparatively large changes in the
pentad distribution for 1d are further illustrated in Figure 6.
For 1e, the limit in syndiospecificity at [propylene] ) 4.6 M is
close to what is observed in neat propylene at 25 °C ([r] )
74.5% vs [r] ) 78.7%, respectively). At 0 °C in neat propylene,
[C3H6]
(M)
[r]
Tm
Mw/
Mn
entry catalyst
activitya (%) (°C)
Mw
28
29
30
31
32
33
34
35
36
37
1d
1d
1d
1d
1d
1e
1e
1e
1e
1e
0.5
0.8
2.1
3.4
4.6
0.5
0.8
2.1
3.4
4.6
2.0 × 103 11.5 ndb nd
nd
2.0 × 103 11.3 108 1.9 × 104 1.8
2.5 × 103 38.6
5.3 × 103 53.7
1.0 × 104 63.3
2.0 × 103 40.3
b
b
c
b
3.5 × 104 2.0
nd
b
4.8 × 104 2.8
nd
b
4.0 × 103 44.2 80
4.0 × 104 1.9
8.5 × 103 54.6 106 6.2 × 104 1.6
1.8 × 104 67.7
b
c
nd
b
2.5 × 104 74.5
1.5 × 105 1.6
(21) Suitable crystals for X-ray structure analysis of 7 were obtained
from a petroleum ether/Et2O solution containing a few drops of THF.
Crystallographic data: a ) 9.919(7) Å, b ) 15.117(9) Å, c ) 15.554(9)
Å, R ) 75.15(5)°, â ) 85.85(5)°, γ ) 74.81(5)°; V ) 2176(2) Å3; Z ) 2;
space group P1. Data (11 635 reflections, θmax ) 22.5°, GOFmerge ) 1.00)
were collected on a CAD-4 diffractometer with Mo KR (λ ) 0.7107) at
a Defined as grams of polymer per gram of Zr per hour). b Not
determined. c No melting point observed by DSC.
2
160 K. Final solution parameters: R ) 0.078 for 3057 reflections with Fo
2
> 3σ(Fo ); GOF ) 1.50 for 480 parameters and 5668 reflections; largest
excursions in final difference map were 0.76 e Å-3, -0.79 e Å-3. The two
molecules in the asymmetric unit are related by an approximate center of
symmetry and have similar conformations and geometric parameters. There
is slight disorder in a few of the THF ligands. The most relevant bonds
(averages) are given here: Li1-CCp ) 2.17 Å, Li2-CCp ) 2.29 Å, Li1-
Cpcentroid ) 1.80 Å, Li2-Cpcentroid ) 1.94 Å, Li1-O1 ) 1.88 Å, Li2-O )
2.02 Å, Cpcentroid-Li1-O1 ) 168°.
(22) The crystals used for the X-ray structure analysis of 1e were obtained
from a concentrated petroleum ether/(SiMe3)2O solution. Crystallographic
data: a ) 12.094(6) Å, b ) 15.285(9) Å, c ) 25.949(17) Å, R ) 90.52-
(5)°, â ) 98.47(5)°, γ ) 90.28(4)°; V ) 4744(5) Å3; Z ) 6; space group
P1. Data (22 659 reflections, θmax ) 22.5°, GOFmerge ) 1.20) were collected
on a CAD-4 diffractometer with Mo KR (λ ) 0.7107) at 160 K. The crystal
was non-merohedrally twinned by rotation of 180° about the b-axis. As a
result, all non-hydrogen atoms were refined isotropically, and several groups
of atoms were restrained to have similar geometries for all six molecules.
The six molecules are related pairwise by a pseudocenter; the three pairs
by approximate translations along [1 -1 1]. Final solution parameters: R
) 0.115 for 9453 reflections with Fo > 2σ(Fo ); GOF ) 2.96 for 692
parameters, 693 restraints, and 20 826 reflections; largest excursions in final
difference map were 9.1 e Å-3, -4.8 e Å-3. Relevant bond distances and
angles (average): Zr-Cp1 ) 2.22 Å, Zr-Cp2 ) 2.24 Å, Zr-Cl ) 2.43
Å, Cp1-Zr-Cp2 ) 122°, Cl-Zr-Cl ) 105°. Both 1e and 1b (ref 3b,c)
have similar conformations.
Figure 5. Dependence of the [r] diads on propylene concentration for
C1-symmetric catalysts 1d,e and 2b.
the stereocontrol improves considerably ([r] ) 90.4%), as
expected with a polymerization system where a unimolecular
site epimerization process occurs in competition with a bimo-
lecular chain propagation (vide supra). At [propylene] ) 0.5
M, the pentad relationship of [mmmr]:[mmrr]:[mrrm] ≈ 2:2:1
displayed by 1d/MAO is characteristic of isospecific catalysts
operating under enantiomorphic site control.
2
2
Assuming that, as for 1a-c/MAO, site epimerization also
competes with propagation for propylene polymerizations with
catalysts 1d/MAO and 1e/MAO, we can rationalize the observed
polymerization behavior for the latter two systems as follows:
at higher propylene concentrations, migratory insertions occur
(23) The pentad distributions (average, %) for polypropenes produced
by 2b in the propene concentration range 0.8-4.6 M with [r] ) 55.3 and
49.2, respectively, are [mmmm] ) 14, [mmmr] ) 12, [rmmr] ) 6, [mmrr]
) 25, [rmrr] + [mrmm] ) 2, [rmrm] absent, [rrrr] ) 21, [mrrr] ) 14,
[mrrm] ) 5. The polymer microstructure is essentially identical at 0 °C in
neat propene.

570 J. Am. Chem. Soc., Vol. 121, No. 3, 1999
Veghini et al.
Table 7. Pentad Distribution (%) for the Polypropylene Produced by 1d,e/MAO in Toluene Solution (Entry Numbers Are Correspondingly the
Same as for Table 6)
entry
catalyst
[mmmm]
[mmmr]
[rmmr]
[mmrr]
[mrmm] + [rmrr]
[mrmr]
[rrrr]
[mrrr]
[mrrm]
28
29
30
31
32
33
34
35
36
37
1d
1d
1d
1d
1d
1e
1e
1e
1e
1e
65.6
65.5
29.3
17.7
9.9
26.3
18.6
11.5
4.3
14.4
13.0
16.6
15.0
11.6
15.3
15.9
12.9
8.1
<1
<1
3.2
3.3
3.7
3.8
2.6
4.4
3.9
2.6
13.6
12.8
18.1
19.9
18.9
17.2
21.5
19.5
14.7
16.1
<1
<1
<1
<1
<1
<1
3.8
4.2
9.1
8.3
5.7
9.7
8.2
7.4
5.9
3.8
<1
<1
<1
7.3
9.1
10.3
8.9
9.4
12.1
11.4
11.4
7.2
8.8
14.6
23.9
7.0
11.2
18.8
34.4
38.3
12.0
16.2
8.5
11.8
13.3
17.2
19.3
3.0
<1
1.6
6.8
Figure 7. Methyl region of the 13C{1H} NMR spectra of polypropylene
produced by 1d,e/MAO and 2b/MAO having [r] ≈ 50%
epimerization becomes competitive, allowing the chain a greater
opportunity to migrate to the less hindered side of the metal-
locene wedge. In the limit of very low [propylene], the migratory
insertion occurs preferentially from the same side of the
metallocene wedge (with the same propylene enantioface,
Scheme 6); isotactic polypropylene is produced. Apparently,
2b operates with a different stereocontrol mechanism, since its
stereoselectivity is essentially independent of propylene con-
centration (Figure 6), in agreement with the original proposals
that its stereocontrol arises from regularly alternating insertions
from stereospecific and aspecific sides of the metallocene wedge
(without competitive site epimerization).1d,4d,11
Figure 6. Methyl region of the 13C{1H} NMR spectra of polypropylene
produced at various propylene concentrations by 1d/MAO.
Scheme 6
At intermediate propylene concentrations, catalyst systems
1d/MAO and 1e/MAO produce a polypropylene with [r] ) [m]
≈ 50% that is not, howeVer, atactic. Rather, the pentad
distributions (Figure 7) at [r] ) 54% and [r] ) 45% ([propylene]
) 3.4 and 0.8 M, respectively) appear to be, at first glance,
close to what is required for perfectly hemiisotactic polypro-
pylene ([mmmm]:[mmmr]:[rmmr]:[mmrr]:[mrmm] + [rmrr]:
[mrmr]:[rrrr]:[rrrm]:[mrrm] ) 3:2:1:4:0:0:3:2:1). In fact, the
{1H}13C NMR spectra of the polymers produced by 1d/MAO
and 1e/MAO under these conditions are quite similar to that
produced by 2b/MAO (Figure 7).
A closer examination of the microstructure of the hemiiso-
tactic-like polypropylenes produced by 1d/MAO and 1e/MAO
reveals some important differences from that of perfectly
hemiisotactic examples: whereas the [mrmr] pentad is es-
sentially absent (Figure 7, Table 7), that for [mrmm] + [rmrr],
also strictly required to be missing for a perfectly hemiisotactic
microstructure, is clearly present, comprising 9.1% and 9.4%
of the pentads, respectively. For polypropylene generated by
2b, these comprise only ca. 2%.23,24 Further consideration of
the consequences of the mechanism proposed in Scheme 6, in
the limiting situation where the rate for the disfavored direction
of site epimerization is zero, leads to the prediction that eight
of the nine possible pentads should be observed, including
[rmrr]. Only the [mrmr] pentad is predicted to be absent, as is
predominantly from regularly alternating sides of the metal-
locene wedge (and from alternating propylene enantiofaces,
Scheme 6), and thus the system behaves syndiospecifically.
However, as the propylene concentration is lowered, the second-
order chain propagation process slows, and unimolecular site
(24) Since these pentads may arise from site epimerization, it may be
the case that site epimerization does occasionally occur, even for 2b/MAO.

Doubly Silylene-Bridged Zirconocene Catalysts
J. Am. Chem. Soc., Vol. 121, No. 3, 1999 571
Conclusions
Some of the doubly silylene-bridged zirconocene catalysts
investigated here have high syndiospecificity (>99.5%) for
insertion of R-olefins into zirconium-polymeryl bonds. Loss
of stereospecificity in these catalysts occurs mainly by stereo-
chemical inversion at the metal center and, to a lesser extent,
by a chain epimerization process which inverts the chirality of
the â-carbon of the polymer chain. The R substituents (on the
upper cyclopentadienyl ligand) of 1a-e influence migratory
olefin insertion and site and chain epimerization and transfer,
and thus affect catalyst activity as well as the tacticity and
molecular weight of the polypropylene. Although these effects
are presently difficult to quantify, the polymerization data
available suggest a significantly faster chain propagation for
1a (R ) H) than for the other catalysts, as shown by its
astounding activity, while retaining relatively high syndiotac-
ticity.
Figure 8. Dependence of molecular weights on propylene concentra-
tion for 1a-e/MAO.
observed.25 Thus, this model for stereocontrol does accom-
modate the observed polymerization behavior for C1-symmetric
catalyst systems rac-1d/MAO and enantiopure 1e/MAO.26
In comparison to Ewen/Razavi singly bridged fluorenyl/
cyclopentadienyl zirconocene catalysts (e.g., 2a/MAO) and aryl-
substituted, singly bridged cyclopentadienyl/fluorenyl (e.g.,
3/MAO), the doubly [SiMe2]-bridged catalysts described here
display a slightly higher stereospecificity, but a stereoselectivity
that appears to be more sensitive to changes in propylene
concentration. A major difference in stereocontrol is observed
for C1-symmetric, doubly bridged 1d,e/MAO as compared with
singly bridged 2b/MAO. The former catalyst systems, as a result
of competitive chain propagation and site epimerization, produce
polypropylenes of widely differing microstructures, from mod-
erately syndiospecific through a form resembling hemiisotactic
to moderately isotactic, as the propylene concentration decreases.
The latter produces a polypropylene more closely matching
hemiisotactic at all propylene concentrations examined, since
site epimerization does not compete with propagation. These
differences may be a result of the more open structures of the
doubly [SiMe2]-bridged zirconocene catalysts, permitting the
polymer chain to swing more freely from side to side of the
metallocene wedge.
Chain Transfer. Although there appears to be considerable
scatter, the molecular weights (GPC) of the polymers produced
by 1a-e/MAO increase with increasing propylene concentration
(Figure 8). A roughly linear increase with [propylene] is
apparent for 1a-c/MAO, suggesting that chain transfer is
predominantly a unimolecular process.27 On the other hand,
catalysts 1d,e/MAO display a less than first-order dependence
of MW with [propylene], possibly indicating that chain transfer
might depend on propylene concentration for these two systems.
1H NMR analysis of low-molecular-weight polypropylene
produced by 1d/MAO at [propylene] ) 0.5 M and 25 °C shows
vinylidene signals at δ 4.72 and 4.80 ppm, indicative of chain
transfer by â-H elimination. There are no signals attributable
to vinylic end groups that would arise from â-CH3 elimination.
A primary kinetic isotope effect k(â-H)/k(â-D) of about 1.6
derived from the difference in number-average molecular
weights of poly-d0-propylene (32 800) vs poly-2-d1-propylene
(56 800) (vide supra), further suggests that chain transfer occurs
primarily by â-H elimination. As is commonly found for
R-olefin polymerization catalysts, a decrease in molecular
weights with increasing temperature is also observed in liquid
propylene and in toluene solution.28
Experimental Section
General Methods. All compounds were manipulated using high-
vacuum line, Schlenk, cannula techniques or in a drybox under nitrogen
atmosphere.29 Argon and nitrogen gases were deoxygenated and dried
by passage over columns of MnO on vermiculite and activated
molecular sieves. Solvents were stored under vacuum over sodium
benzophenone ketyl (THF, Et2O, and (SiMe3)2O) or titanocene (toluene,
petroleum ether) and freshly distilled immediately prior to use.
Propylene was dried by passage through a Matheson 2110 drying system
equipped with an OXISORB column. For polymerization experiments
in toluene solutions, MAO (Albemarle) was used as a solid after
removal of toluene in vacuo and being dried under high vacuum at 25
°C for 2 days. Comparisons of the activities between runs employing
different MAO batches indicated strong deviations. 1H, 2H, and13C NMR
spectra were recorded on a Bruker AM500 or a AMX500 at 500.13,
76.77, and 125.77 MHz, respectively. Elemental analyses were
performed by Fenton Harvey of the Caltech Analytical Laboratory.
(25) A full analysis of the possible pathways to the “allowed” pentads
is given in the Supporting Information.
(26) The pentad distributions in the 13C NMR spectra for polypropylenes
produced by 1d/MAO in neat propylene and in toluene solutions having
differing [propylene] can be fit very well to a three-parameter model: one
parameter (R) represents the monomer enantiofacial selectivity when the
chain is on the less crowded side of the metallocene wedge, a second (â)
represents (the same) monomer enantiofacial preference with the chain on
the more crowded side of the metallocene wedge, and the third (ꢀ) represents
the probability of site epimerization wherein the chain moves from the less
selective side to the more selective side (in the favored direction of Scheme
6). As required by the model in Scheme 6, least-squares fits give R ≈ 0.99,
â ≈ 0.17, and ꢀ increases steadily as the [propylene] decreases from that
in liquid propylene (ꢀ ≈ 0.12) to 0.5 M (ꢀ ≈ 0.91). Details of this statistical
model will be presented in a forthcoming manuscript (Miller S. A.; Bercaw,
J. E., manuscript in preparation).
The compounds (Me2C)[(η5-C5H3Me)(η5-C13H8)]ZrCl2,30 (Ph2C)-
(27) The dependence of the degree of polymerization (PN) on propene
concentration is a function of the rate of propagation (νP) vs rate of
termination (νT) by â-H transfer to olefin (kTO/kP) or to metal (kTM/kP).
Since PN ) Mn/42 ) νP/νT ) C*kP[C3H6]/C*(kTM + kTO[C3H6]), one
obtains 1/PN ) (kTM/kP)(1/[C3H6]) + kTO/kP. A plot of 1/PN vs 1/[propene]
gives kTM/kP about 10-3 for 1a-e and kTO/kP about 10-3 for 1d and e10-4
for the other catalysts. These data suggest dominant monomolecular chain
termination for 1a-c and 1e and partial contribution of a bimolecular process
for 1d. For analogous discussions, see: Stehling, U.; Diebold, J.; Kirsten,
R.; Ro¨ll, W.; Brintzinger, H. H.; Ju¨ngling, S.; Mu¨lhaupt, R.; Langhauser,
F. Organometallics 1994, 13, 964 and ref 8.
[(η5-C5H4)(η5-C13H8)]ZrCl2,15a metallocenes 1a-d, and Me2ClSi[C5H3-
(28) Propene polymerizations at [propene] ) 3.4 M by 1b/MAO
produced a polymer of Mw ) 101 000 at 23 °C and 157 000 at 0 °C (PDI
) 1.5 for both).
(29) Burger, B. J.; Bercaw, J. E. In Experimental Organometallic
Chemistry; Wayda, A. L., Darensbourg, M. Y., Eds.; ACS Symposium
Series 357; American Chemical Society: Washington, DC, 1987: Chapter
4.
(30) Razavi, A.; Atwood, J. L. J. Organomet. Chem. 1995, 497, 105.

572 J. Am. Chem. Soc., Vol. 121, No. 3, 1999
Veghini et al.
3,5-(CHMe2)2]3a,31 were prepared as previously reported. (1S,2S,5R)-
(+)-neomenthol was used as received (Aldrich), while (1R,2S,5R)-
menthylcyclopentadiene was prepared by a slight modification of a
recently reported preparation.16 6 2-d1-Propylene was prepared as
previously reported9 and purified by distillation at -100 °C (cyclo-
hexane/LN2 slush) to a trap at -196 °C, passing through two traps at
-80 °C. Gas chromatography and 1H NMR analyses showed no traces
of residual Et2O. The product was stored at -80 °C in a Schlenk tube
equipped with a Teflon valve. The concentration of propylene in toluene
solutions was calculated according to literature data.32
Polymerization Reaction Procedures. The polymerization experi-
ments in toluene solutions were carried out in a 250-mL glass reactor
(Andrews Glass Co., maximum pressure 120 psi) equipped with a
septum port, a large stir bar, and a pressure gauge (0-200 psi).
CAUTION: All of these procedures should be performed behind a blast
shield. Stirring was kept at 700 rpm. In a typical procedure, the glass
reactor described previously was charged with solid MAO (180 mg
for runs 1-5 and 500 mg for all others) and assembled in the drybox.
The reactor was then connected to the propylene tank and purged with
propylene at atmospheric pressure for 10 min. After this time, 40 mL
of toluene was introduced in the reactor through the septum port. Under
vigorous stirring, the reactor was purged again for an additional 15
min and, if necessary, pressurized for 10 min. The freshly prepared
standard precatalyst solutions (1 mL) (0.42 mM, 1a; 1.05 mM, 1b;
0.98 mM, 1c; 2.52 mM, 2b, 1d, and 1e) were added by using a gastight
syringe through the septum port. The polymerization reactions were
quenched by slow addition of MeOH initially and then poured into
400 mL of HCl:MeOH (1:4) and stirred for about 4 h. Slow precipitation
of the polymeric material from the toluene/MeOH solution allows an
efficient elimination of aluminum residues from the polypropylene
samples.
Bulk polymerization reactions were carried out in a 2-L steel reactor
according to the following procedure. The reactor was purged with
dinitrogen for several hours, 0.2 mL of a 10% MAO toluene solution
was added, and condensed propylene was added. Under stirring (400
rpm), the temperature was then set at the desired value by use of an
external metal jacket with circulating water. In the drybox, 4 mg of
the metallocene was activated for 10 min with 9 mL of a 10% MAO
toluene solution ([MAO]/[Zr] ) 2000/1). An aliquot of this solution
was then introduced into a 5-mL pressure chamber, which was then
connected to the reactor and introduced into the reactor under N2
pressure. The stirring was kept constant at about 400 rpm with a
nitrogen flow controlled stirrer. After the given reaction time, the reactor
was depressurized and the polymer collected and vacuum-dried
overnight.
Polymerization of 2-d1/d0-Propylene. In a glovebox, a round-bottom
flask was charged with 500 mg of solid MAO, assembled to a 90°
needle valve, and the whole apparatus connected to a high-vacuum
line. Toluene (50 mL) was vacuum transferred, and the whole assembly
was saturated at atmospheric pressure with 2-d1-propylene or d0-
propylene at 23 °C. The Teflon valve was then replaced by a septum,
and 1 mL of a 2.52 mM standard toluene solution of 1b was added by
using a gastight syringe. After 10 min, the reaction mixture was
quenched and the polymer treated as described above. This procedure
was repeated 3 times for d0-propylene and gave consistent productivities,
which were comparable to the run with 2-d1-propylene.
present. The polymer microstructures were obtained by analyzing the
18-22-ppm region in the {1H}13C NMR spectra, as previously
described.33
Syntheses. C5H5-4-(1R,2S,5R-Menthyl) (4). To a solution of 5.0 g
(32 mmol) of (1S,2S,5R)-(+)-neomenthol (degassed) in 150 mL of THF
at -20 °C was added 20 mL (32 mmol) of a 1.6 M solution of n-BuLi
in hexanes over 15 min. The temperature was kept constant for 1 h,
and then a solution of 6.77 g (35 mmol) of tosyl chloride in 50 mL of
THF was added over 20 min. The reaction mixture was allowed to
warm to room temperature over 4 h and stirred for an additional 8 h.
A solution of 3.52 g (32 mmol) of Na(C5H5) in 150 mL of THF was
then slowly added to the reaction mixture at -10 °C under vigorous
stirring over 30 min. After 2 h at -10 °C, the solution was warmed to
room temperature and stirred for 14 h. The solution was then quenched
with 15 mL of H2O, the two phases were separated, and the water
phase was extracted with 3 × 50 mL of Et2O. The organic phases were
collected, the solvent was removed in vacuo, and the product was
Kugelrohr distilled at 70 °C/high vacuum, giving 4 in 72% yield (4.75
g, 23 mmol). The NMR analyses are consistent with the reported data.16a
Li[C5H4-4-(1R,2S,5R-menthyl)]. To a solution of 4.00 g (19.6
mmol) of 4 in 100 mL of Et2O at 0 °C was added 12.5 mL (20 mmol)
of a 1.6 M solution of n-BuLi in hexanes over 10 min. After the solution
was stirred at -10 °C for 10 min, the cold bath was removed. Stirring
was continued for 10 h at room temperature. The volume was then
reduced to ca. 20 mL, and 100 mL of petroleum ether was vacuum
transferred onto the reaction mixture. The product was filtered and dried
overnight under vacuum, giving Li[C5H4-4-(1R,2S,5R-menthyl)] in 80%
1
yield (3.3 g, 15.7 mmol). H NMR (THF-d8): δ 5.52-5.48 (m, 4H,
Cp), 2.28 (m, 1H, Cp-CH), 1.62 (m, 1H, CH Me), 1.38 (m, 1H, CH
3
Me2), 1.26-0.80 (m, -CH2-), 0.85 (d, JHH ) 6 Hz, 3H, Me), 0.71
(d, 3JHH ) 7 Hz, 3H, CHMe2), 0.65 (d, 3JHH ) 7 Hz, 3H, CHMe2). 13
C
NMR (THF-d8): δ 125.1, 101.9, 101.6, 51.1, 48.7, 41.3, 36.8, 34.8,
27.8, 26.0, 25.6, 23.4, 22.2, 16.1.
(Me2Si){C5H4-4-(1R,2S,5R-Menthyl)}{C5H3-3,5-(CHMe2)2} (5).
To a solution of 2.30 g (9.5 mmol) of Me2ClSi[C5H3-3,5-(CHMe2)2]
in 100 mL of THF was added 2.00 g (9.52 mmol) of Li[C5H4-4-
(1R,2S,5R-menthyl)] in 50 mL of THF at room temperature over 10
min. The resulting mixture was stirred overnight. Solvents were then
removed in vacuo, and the residue was extracted with petroleum ether.
Removal of the solvent gave 5 as a yellow oil in 92% yield (3.6 g, 8.7
mmol). 13C NMR (benzene-d6) (mixture of double bond isomers): δ
164.3, 161.3, 159.2, 157.0, 152.6, 152.3, 152.0, 150.7, 150.3, 150.1,
143.6, 143.4, 143.1, 142.9, 141.9, 134.3, 133.9, 133.7, 133.1, 132.4,
131.8, 131.3, 131.1, 130.8, 130.3, 127.4, 127.0, 126.8, 125.8, 125.1,
125.0, 124.6, 123.9, 123.8, 122.6, 122.3, 121.2, 52.2, 51.0, 50.9, 50.0,
49.3, 48.2, 48.0, 47.5, 47.2, 46.8, 46.3, 45.2, 45.0, 44.7, 44.5, 43.9,
42.9, 42.7, 42.3, 41.0, 40.2, 35.7, 35.6, 33.3, 33.1, 30.4, 30.2, 29.5,
28.2, 28.0, 25.3, 24.9, 24.8, 24.7, 23.8, 23.2, 22.8, 22.5, 22.2, 21.7,
21.1, 20.8, 15.7, 15.6, 2.0, 2.2, -3.9, -4.2, -4.7, -5.5. Elemental
Anal. Calcd for C28H46Si: C, 81.95; H, 11.21. Found: C, 81.18, 81.33;
H, 10.12, 10.42.
Li2(Me2Si){C5H3-4-(1R,2S,5R-Menthyl)}{C5H2-3,5-(CHMe2)2}. To
a solution of 1.5 g (3.6 mmol) of 5 in 50 mL of Et2O was added 5 mL
(8 mmol) of a 1.6 M solution of n-BuLi in hexanes at room temperature
over 5 min. The mixture was stirred for 24 h, and the solvent was
removed in vacuo, leaving a white foam. Petroleum ether (3 × 20 mL)
was vacuum transferred onto the reaction mixture and removed three
times. The resulting powder was then washed with an additional 30
mL of petroleum ether and filtered, giving Li2(Me2Si){C5H3-4-
(1R,2S,5R-menthyl)}{C5H2-3,5-(CHMe2)2} in 85% yield (1.3 g, 3.0
mmol). 1H NMR (THF-d8): δ 5.81, 5.78, 5.64, 5.62, 5.60 (s, 5H, Cp),
3.12 (m, 1H, CHMe2), 2.75 (m, 1H, CHMe2), 2.31 (m, 1H, -CH-),
1.61 (m, 1H, -CH-), 1.37 (m, 1H, CHMe2), 1.26-1.03 (m, -CH2-
), 1.13 (d, 3JHH ) 6 Hz, 6H, CHMe2), 0.85 (d, 3JHH ) 6 Hz, 3H, Me),
0.71 (d, 3JHH ) 6 Hz, 3H, CHMe2), 0.65 (d, 3JHH ) 6 Hz, 3H, CHMe2),
0.26 (s, 6H, SiMe), 0.25 (s, 6H, SiMe). 13C NMR (THF-d8): δ 134.8,
128.1, 127.7, 122.1, 118.8, 110.5 (broad), 108.2, 107.2, 104.1 (broad),
101.9, 101.6, 100.5, 98.7, 94.2, 51.1, 50.8, 48.8, 43.5, 36.8, 34.7, 30.5,
Polymer Analyses. Solution {1H}13C NMR spectra were recorded
at 75.4 MHz on a Bruker AM500 or AMX500 spectrometer. Samples
from the runs in toluene solution were prepared by mixing ca. 20 mg
of the polymer in 0.6 mL of benzene-d6:C6H3Cl3:(SiMe3)2O (1:3:0.5),
preheating at 80 °C for few hours, and measuring at that temperature.
A total of 8000-10 000 transients were accumulated for each spectrum
with a delay adjusted to have a recycling time of 4 s (12 s for poly-
2-d1-propylene). The samples for the experiments in neat propylene
were prepared with ca. 5-7 mg of polymer in 0.6 mL of solvent. About
30 000 transients were acquired with 4 s recycling time. The chemical
shifts are referenced to residual benzene-d6 (128 ppm) or (SiMe3)2O
(0 ppm). Decoupling was always on during acquisition, so NOE was
(31) Yoder, J. C.; Day, M. W.; Bercaw, J. E. Organometallics 1998, 17,
4946.
(33) Busico, V.; Cipullo, R.; Corradini, P.; Landriani, L.; Vacatello, M.;
(32) Frank, H. P. O¨ ster. Chem. Zeit. 1967, 11, 360.
Segre, A. L. Macromolecules 1995, 28, 1887.

Doubly Silylene-Bridged Zirconocene Catalysts
J. Am. Chem. Soc., Vol. 121, No. 3, 1999 573
3
3
30.2, 29.7, 27.8, 27.3, 26.1, 25.9, 25.6, 25.4, 25.3, 25.1, 24.9, 23.4,
22.3, 22.2, 19.2, 16.1, 14.2, 3.1 (SiMe2). Elemental Anal. Calcd for
C28H44Si2Li2: C, 79.64; H, 10.42. Found: C, 77.32, 77.30; H, 11.50,
11.44.
(d, JHH ) 7 Hz, 3H, CHMe2), 0.73 (d, JHH ) 7 Hz, 3H, CHMe2),
0.57, 0.53, 0.50, 0.48 (s, 3H, SiMe2). 13C NMR (C6D6): δ 3.0 (SiMe2),
3.3 (SiMe2), 15.6, 20.5, 20.6, 21.7, 22.7, 24.9, 27.9, 28.3, 28.2, 29.2,
29.3, 33.3, 35.4, 41.3, 43.4, 50.6, 108.3, 108.5, 113.6, 114.8, 133.63,
138.5, 139.0, 163.8, 166.2. Elemental Anal. Calcd for C30H47Cl2Si2Zr:
C, 57.59; H, 7.51. Found: C, 56.38, 56.19; H, 7.73, 7.60. Li2[(1,2-
Me2Si)2{η5-C5H2-4-(1R,2S,5R-menthyl)}{η5-C5H-3,5-(CHMe2)2}] (7)
can be obtained as a THF adduct in crystalline form by cooling at -30
°C a concentrated Et2O/petroleum ether solution containing a few drops
of THF.
(1,2-Me2Si)2{C5H3-4-(1R,2S,5R-Menthyl)}{C5H2-3,5-(CHMe2)2}
(6). To a solution of 1.93 g (4.6 mmol) of Li2(Me2Si){C5H3-4-
(1R,2S,5R-menthyl)}{C5H2-3,5-(CHMe2)2} in 50 mL of THF at -78
°C was vacuum transferred 1.0 g (1.1 mL, 5 mmol) of SiMe2Cl2. The
resulting mixture was stirred at that temperature for 4 h and then
warmed to room temperature over 4 h. After the solution was stirred
for an additional 20 h at room temperature, the volatiles were removed,
and the product was extracted with 100 mL of petroleum ether. Removal
of the solvent gave a yellow oil, which was then Kugelrohr distilled
twice at 120 °C/high vacuum. White solid 6 was obtained in 83% yield
(1.8 g, 3.8 mmol). 13C NMR (benzene-d6): δ 162.7, 162.2, 151.3, 151.1,
146.5, 146.3, 140.2, 139.4, 138.9, 132.3, 131.9, 131.8, 131.4, 131.1,
130.9, 130.7, 123.3, 123.2, 56.8, 56.6, 47.5, 47.3, 45.6, 45.3, 42.8, 35.7,
33.4, 30.0, 29.9, 29.5, 27.9, 27.8, 24.9, 24.8, 23.2, 22.9, 22.0, 15.3,
0.6, 0.5, -3.2, -6.2. Elemental Anal. Calcd for C30H49Si2: C, 77.24;
H, 10.72. Found: C, 76.93; H, 10.98.
Acknowledgment. This work has been supported by the U.S.
DOE Office of Basic Energy Sciences (Grant No. DE-FG03-
85ER13431) and by Exxon Chemicals America. D.V. thanks
the Swiss National Science Foundation (Stipendium fu¨r ange-
hende Forscher) and Ciba-Geigy (Jubila¨um Stiftung) for post-
doctoral fellowships. Dr. Timothy Herzog, Deanna Zubris, and
Shigenobu Miyake (Japan Polyolefins) are gratefully acknowl-
edged for helpful discussions and for preparing some of the
starting material for the ligand syntheses. Japan Polyolefins is
gratefully acknowledged for performing some of the molecular
weight and DSC measurements. The experimental assistance
of Mr. William B. Brandley (Exxon Chemical Co.) is gratefully
acknowledged.
[(1,2-Me2Si)2{η5-C5H2-4-(1R,2S,5R-Menthyl)}{η5-C5H-3,5-
(CHMe2)2}]ZrCl2 (1e). To a solution of 600 mg (1.3 mmol) of 6 in
50 mL of Et2O at room temperature was added 1.8 mL (2.85 mmol) of
a 1.6 M solution of n-BuLi in hexanes. After the solution was stirred
at 25 °C for 20 h, a suspension of 350 mg (1.5 mmol) of ZrCl4 in 20
mL of Et2O was added via cannula to the solution of Li2[(1,2-Me2-
Si)2{η5-C5H2-4-(1R,2S,5R-menthyl)}{η5-C5H-3,5-(CHMe2)2}] (7), and
the reaction mixture was stirred for 22 h. The resulting precipitate was
filtered, and the filtrate was dried in vacuo. The resulting powder was
then washed with cold hexamethyldisiloxane. Yield: 290 mg (46%).
1H NMR (benzene-d6): δ 6.80 (s, 1H, Cp), 6.73 (s, 1H, Cp), 6.46 (s,
1H, Cp), 2.92-2.99 (m, 2H, Cp-CHMe2), 2.75-2.82 (m, 2H, -CH-
Supporting Information Available: Schemes detailing the
possible pathways to the “allowed” pentads with catalyst 1d/
MAO. Details of the structure determinations, including listing
of final atomic coordinates, thermal parameters and bond
distances and angles (56 pages, print/PDF). See any current
masthead page for ordering information and Web access
instructions.
3
), 1.79-1.52 (m, 4H, -CH2-), 1.36 (d, JHH ) 6 Hz, 3H, CHMe2),
3
1.21 (m, 2H, -CH2-), 1.01 (d, JHH ) 7 Hz, 3H, CHMe2), 0.96 (d,
3
3JHH ) 7 Hz, 3H, CHMe2), 0.95 (d, JHH ) 7 Hz, 3H, CHMe2), 0.76
JA982868J