Effect of metal on the stereospecificity of 2-arylindene catalysts of elastomeric polypropylene

Article abstract of DOI:10.1021/ja971895r

Polymerization of propylene with bis(2-phenylindenyl)zirconium dichloride and bis(2-[3,5-bistrifluoromethyl)zirconium dichloride produces elastomeric polypropylene. The elastomeric properties of these polymers have been interpreted in terms of a stereoblock microstructure. Analysis of the microstructure by ¹³C NMR reveals isotactic pentad contents [mmmm] ranging from 6 to 74%. The hafnium derivatives were investigated to probe the influence of the transition metal on the polymerization behavior. The hafnium-based catalysts yield polypropylenes that are significantly less isotactic than the corresponding zirconium compounds, although molecular weights and productivities were similar for hafnium and zirconium derivatives. The X-ray crystal studies of these catalysts show nearly identical structures for corresponding zirconium and hafnium compounds. Variable temperature NMR of the metallocene dibenzyl analogues showed behavior consistent with rotation of indenyl ligands, where rotation of the indenyl ligand of bis(2-phenylidenyl)zirconium dibenzyl was 6800 Hz and that of hafnium was 6700 Hz at 20°C. Based on our proposed mechanism of polymerization, the origin of the different microstructures can be ascribed to a faster propagation of the isospecific zirconium site relative to that of the isospecific hafnium site. Polymerization of propylene with bis(2-phenylindenyl)zirconium dichloride and bis(2-(3,5-bistrifluoromethyl)indenyl]zirconium dichloride produces elastomeric polypropylene. The elastomeric properties of these polymers have been interpreted in terms of a stereoblock microstructure. Analysis of the microstructure by ¹³C NMR reveals isotactic pentad contents [mmmm] ranging from 6 to 74%. The hafnium derivatives were investigated to probe the influence of the transition metal on the polymerization behavoir. The hafnium-based catalysts yield polypropylenes. that are significantly less isotactic than the corresponding zirconium compounds, although molecular weights and productivities were similar for hafnium and zirconium derivatives. The X-ray crystal studies of these catalysts show nearly identical structures for corresponding zirconium and hafnium compounds. Variable temperature NMR of the metallocene dibenzyl analogues showed behavior consistent with rotation of indenyl ligands, where rotation of the indenyl ligand of bis(2-phenylindenyl)zirconium dibenzyl was 6800 Hz and that of hafnium was 6700 Hz at 20°C. Based on:our proposed mechanism of polymerization, the origin of the different microstructures can be ascribed to a faster propagation of the isospecific zirconium site relative to that of the isospecific hafnium site.

Full text of DOI:10.1021/ja971895r

11174
J. Am. Chem. Soc. 1997, 119, 11174-11182
Effect of Metal on the Stereospecificity of 2-Arylindene
Catalysts for Elastomeric Polypropylene
Michael D. Bruce,† Geoffrey W. Coates,† Elisabeth Hauptman,†
Robert M. Waymouth,*,† and Joseph W. Ziller‡
Contribution from the Departments of Chemistry, Stanford UniVersity,
Stanford, California 94305, and UniVersity of California, IrVine, California 92717
ReceiVed June 9, 1997X
Abstract: Polymerization of propylene with bis(2-phenylindenyl)zirconium dichloride and bis(2-[3,5-bistrifluoro-
methyl)indenyl]zirconium dichloride produces elastomeric polypropylene. The elastomeric properties of these polymers
have been interpreted in terms of a stereoblock microstructure. Analysis of the microstructure by 13C NMR reveals
isotactic pentad contents [mmmm] ranging from 6 to 74%. The hafnium derivatives were investigated to probe the
influence of the transition metal on the polymerization behavior. The hafnium-based catalysts yield polypropylenes
that are significantly less isotactic than the corresponding zirconium compounds, although molecular weights and
productivities were similar for hafnium and zirconium derivatives. The X-ray crystal studies of these catalysts show
nearly identical structures for corresponding zirconium and hafnium compounds. Variable temperature NMR of the
metallocene dibenzyl analogues showed behavior consistent with rotation of indenyl ligands, where rotation of the
indenyl ligand of bis(2-phenylindenyl)zirconium dibenzyl was 6800 Hz and that of hafnium was 6700 Hz at 20 °C.
Based on our proposed mechanism of polymerization, the origin of the different microstructures can be ascribed to
a faster propagation of the isospecific zirconium site relative to that of the isospecific hafnium site.
Introduction
stereospecificity of the catalyst; substitutions on the phenyl ring
yield polypropylenes that span the range from [mmmm] ) 6%
to [mmmm] ) 75%.4
We now report the influence of the transition metal center
on the polymerization behavior of these catalysts. Substitution
of zirconium with hafnium yields complexes that are isostruc-
tural with the original zirconium catalysts but which produce
polypropylenes of lower isotacticities than those from parent
zirconium compounds.
Soluble metallocene complexes provide well-defined homo-
geneous Ziegler-Natta catalysts for olefin polymerization.1
Correlation of the catalyst structure with the catalyst productiv-
ity, polymer molecular weight, molecular weight distribution,
and polymer microstructure can reveal important insights into
the mechanism of polymerization. The ligand environment has
an important influence on microstructures of polymers produced
by these catalysts: stereorigid chiral racemic ansa-metallocenes
have been used to produce isotactic polypropylenes,2,3 while
achiral catalysts have been employed to produce syndiotactic
and atactic polypropylenes.1
We have recently reported that the unbridged indenyl
metallocene bis(2-phenylindenyl)zirconium dichloride (1), when
combined with methylaluminoxane under a propylene atmo-
sphere generates rubbery, elastomeric polypropylenes.4,5 The
elastomeric properties of these polymers have been interpreted
in terms of a stereoblock atactic/isotactic microstructure. This
structure was, in turn, proposed to arise through a mechanism
involving fluctuation in the catalyst geometry from achiral to
chiral during the course of the polymerization reaction (Figure
1).4
Results
(PhInd)2ZrCl2 (1, (PhInd ) 2-phenylindenyl), ((CF3)2-
PhInd)2ZrCl2 (3, (CF3)2PhInd ) 2-[3,5-bis(triflouromethyl)-
phenyl]) and ((CF3)2PhInd)2HfCl2 (4) were prepared as previ-
ously described.4 The hafnium derivative of 1, (2-PhInd)2HfCl2
(2), was prepared analogously by reaction of lithium 2-phenyl-
indene with hafnium tetrachloride in toluene at room temper-
ature.
Crystallization of the PhInd hafnium complex 2 from toluene
at -18 °C gave crystals suitable for X-ray crystal analysis. The
crystal structure of 2 (Figure 2) is nearly identical to that of the
PhInd zirconium compound 1 (Figure 3). Both structures show
syn and anti isomers in the unit cell. The bond lengths of the
corresponding metal-cyclopentadienyl bonds are nearly identi-
cal (Table 1); the hafnium complex displays average metal-
carbon bond lengths that are 0.005-0.016 Å shorter than those
for the zirconium complex.
The microstructures of polymers produced by 1 are interme-
diate between atactic and isotactic. Analysis of the polymers
by 13C NMR reveals that the stereospecificity of the catalyst as
measured by the isotacticity index, [mmmm], is in the range of
6-40%. The ligand environment has a profound effect on the
Similarly, the X-ray crystal structures of (trifluoromethyl)-
phenyl-substituted complexes of zirconium (3,4b Figure 4) and
hafnium (4, Figure 5) are nearly identical (Table 2). Corre-
sponding metal-carbon bond lengths are within 0.020 Å, where
† Stanford University.
‡ University of California, Irvine.
X Abstract published in AdVance ACS Abstracts, November 1, 1997.
(1) Brintzinger, H. H.; Fischer, D.; Mulhaupt, R.; Rieger, B.; Waymouth,
R. M. Angew. Chem., Int. Ed. Eng. 1995, 34, 1143-1170.
(2) Ewen, J. A. J. Am. Chem. Soc. 1984, 106, 6355-6364.
(3) Kaminsky, W.; Ku¨lper, K.; Brintzinger, H. H.; Wild, F. R. W. P.
Angew. Chem., Int. Ed. Engl. 1985, 24, 507-508.
(4) (a) Coates, G. W.; Waymouth, R. M. Science 1995, 267, 217-219.
(b) Hauptman, E.; Waymouth, R. M.; Ziller, J. W. J. Am. Chem. Soc. 1995,
117, 11586-11587.
(5) For related work on elastomeric polypropylene see: (a) Collette, J.
W.; Tullock, C. W.; MacDonald, R. N.; Buck, W. H.; Su, A. C. L.; Harrel,
J. R.; Mulhaupt, R.; Anderson, B. C. Macromolecules 1989, 22, 3851-
3858. (b) Mallin, D. T.; Rausch, M. D.; Lin, Y. G.; Dong, S.; Chien, J. C.
W. J. Am. Chem. Soc. 1990, 112, 2030-2031. (c) Gauthier, W. J.; Corrigan,
J. F.; Taylor, N. J.; Collins, S. Macromolecules 1995, 28, 3771-3778.
S0002-7863(97)01895-7 CCC: $14.00 © 1997 American Chemical Society

2-Arylindene Catalysts for Elastomeric Polypropylene
J. Am. Chem. Soc., Vol. 119, No. 46, 1997 11175
Figure 1. Production of stereoblock polypropylenes with (2-PhInd)2ZrCl2.
Figure 3. ORTEPs of anti- and syn-bis(2-phenylindenyl)zirconium
dichloride (1).
Figure 2. ORTEPs of anti- and syn-bis(2-phenylindenyl)hafnium
dichloride (2).
At room temperature, the benzylic protons Ha and Hb (Figure
6) are equivalent and appear as sharp singlets at 0.11 (6) and
-0.19 ppm (7). The cyclopentadienyl protons Hc and Hd appear
as broad singlets at 5.92 (6) and 5.87 ppm (7). At -70 °C, the
benzylic protons become diastereotopic and appear as a doublet
of doublets (6, ∆ν, ) 123 Hz, J ) 11.16 Hz, Figure 7; 7, ∆ν
) 238 Hz, J ) 10.94 Hz, Figure 8), while the cyclopentadienyl
protons split into two singlets (6, ∆ν ) 502 Hz; 7, ∆ν ) 472
Hz). As the temperature is varied, the signals for the benzylic
protons exhibit behavior consistent with coupled two-site
exchange (HaTHb), while those for the cyclopentadienyl protons
exhibit behavior consistent with uncoupled two-site exchange
(HcTHd).6
The dynamics were investigated by comparing the experi-
mental spectrum with a calculated spectrum at each tempera-
ture.9,10 For the cyclopentadienyl protons, simulated spectra
were calculated from the modified Bloch equations for un-
coupled two-site exchange,11,12 while simulations corresponding
to the benzylic protons were calculated from Alexander’s and
Johnson’s equations for coupled two-site exchange.7,13 Simula-
the hafnium-carbon bond lengths are shorter than those of the
zirconium derivative. The ligand environment has only a small
effect on the bond lengths; the (CF3)2PhInd-substituted com-
plexes have average metal-carbon bond lengths that are
approximately 0.005 Å shorter than those of the corresponding
2-PhInd complexes.
The solution structure of these metallocenes was investigated
by variable temperature NMR spectroscopy.6,7 At room tem-
1
perature, the H NMR spectra at 400 MHz of the zirconocene
dichloride 1 and the hafnocene dichloride 2 exhibit time-
averaged C2 symmetry consistent with rapid rotation of the
indenyl ligands.8,9 Coalescence was not observed at any
temperature down to -100 °C. The same was observed for
the dimethyl derivative of 1, bis(2-phenylindenyl)zirconium
dimethyl (5). For this reason, the dibenzyl derivatives of 1 and
2, bis(2-phenylindenyl)zirconium dibenzyl (6) and bis(2-
phenylindenyl)hafnium dibenzyl (7) were synthesized by reac-
tion of 1 and 2 with benzylmagnesium chloride.
(6) Sandstrom, J. Dynamic NMR Spectroscopy; Academic Press: New
York, 1982.
(7) Johnson, C. S. AdV. Magn. Reson. 1965, 1, 33-102.
(8) Luke, W. D.; Streitwieser, A. J. Am. Chem. Soc. 1981, 103, 3241-
3243.
(9) (a) Kru¨ger, C.; Nolte, M.; Erker, G.; Thiele, S. Z. Naturforsch. 1992,
47b, 995-999. (b) Knickmeier, M.; Erker, G.; Fox, T. J. Am. Chem. Soc.
1996, 118, 9623-9630.
(10) Binsch, G. In Dynamic Nuclear Magnetic Resonance Spectroscopy;
Cotton, F. A., Jackman, L. M., Eds.; Academic Press: New York, 1975;
pp 45-90.
(11) Rogers, M. T.; Woodbry, J. C. J. Phys. Chem. 1962, 66, 540-
546.
(12) McConnell, H. M. J. Chem. Phys. 1958, 28, 430-431.

11176 J. Am. Chem. Soc., Vol. 119, No. 46, 1997
Bruce et al.
Table 1. Representative M-C Bond Lengths (Å) for
(2-PhInd)2MCl2
Anti Isomer
Zr(1)-C(1)
Zr(1)-C(2)
Zr(1)-C(3)
Zr(1)-C(4)
Zr(1)-C(9)
average
2.482(5)
2.561(3)
2.558(3)
2.612(4)
2.531(5)
2.548
Hf(1)-C(1)
Hf(1)-C(7)
Hf(1)-C(9)
Hf(1)-C(17)
Hf(1)-C(23)
average
2.538(6)
2.517(10)
2.542(6)
2.618(6)
2.454(6)
2.533
Zr(1)-C(16)
Zr(1)-C(17)
Zr(1)-C(18)
Zr(1)-C(19)
Zr(1)-C(24)
average
2.465(3)
2.551(3)
2.564(3)
2.622(3)
2.527(3)
2.545
Hf(1)-C(2)
Hf(1)-C(8)
Hf(1)-C(16)
Hf(1)-C(22)
Hf(1)-C(24)
average
2.591(7)
2.461(8)
2.547(5)
2.516(5)
2.542(6)
2.531
Syn Isomer
Zr(2)-C(31)
Zr(2)-C(32)
Zr(2)-C(33)
Zr(2)-C(34)
Zr(2)-C(39)
average
2.476(4)
2.504(4)
2.523(3)
2.624(3)
2.570(3)
2.539
Hf(2)-C(32)
Hf(2)-C(38)
Hf(2)-C(46)
Hf(2)-C(52)
Hf(2)-C(54)
average
2.582(6)
2.459(9)
2.528(6)
2.563(6)
2.485(7)
2.523
Figure 5. ORTEPs of bis(2-[3,5-bis(trifluoromethyl)phenyl]indenyl)-
hafnium dichloride (2).
Zr(2)-C(46)
Zr(2)-C(47)
Zr(2)-C(48)
Zr(2)-C(49)
Zr(2)-C(54)
average
2.485(5)
2.553(4)
2.529(3)
2.591(3)
2.559(4)
2.543
Hf(2)-C(31)
Hf(2)-C(37)
Hf(2)-C(39)
Hf(2)-C(47)
Hf(2)-C(53)
average
2.511(2)
2.551(7)
2.543(7)
2.614(6)
2.471(7)
2.538
Table 2. Representative M-C Bond Lengths for
(2-[3,5-Bis(trifluoromethyl)phenyl]indenyl)2MCl2
Zr(1)-C(1)
Zr(1)-C(7)
Zr(1)-C(9)
Zr(1)-C(2)
Zr(1)-C(8)
average
2.486(3)
2.587(4)
2.551(3)
2.562(4)
2.538(4)
2.545(4)
Hf(1)-C(1)
Hf(1)-C(7)
Hf(1)-C(9)
Hf(1)-C(2)
Hf(1)-C(8)
average
2.474(12)
2.561(14)
2.541(11)
2.543(12)
2.550(13)
2.534(13)
Zr(1)-C(18)
Zr(1)-C(24)
Zr(1)-C(26)
Zr(1)-C(19)
Zr(1)-C(25)
average
2.517(3)
2.524(4)
2.565(3)
2.588(4)
2.508(4)
2.540(4)
Hf(1)-C(18)
Hf(1)-C(24)
Hf(1)-C(26)
Hf(1)-C(19)
Hf(1)-C(25)
average
2.503(12)
2.487(14)
2.558(12)
2.552(13)
2.502(13)
2.520(13)
Figure 4. ORTEPs of bis(2-[3,5-bistrifluoromethyl)phenyl]indenyl)-
zirconium dichloride (3).
Figure 6. Bis-(2-phenylindenyl)2M(CH2Ph)2.
The metallocene dichlorides 1-4 are active propylene poly-
merization catalysts when combined with methylaluminoxane
(MAO) in toluene (Table 4).4,5 Polymerizations were carried
out under a variety of pressure/temperature combinations, using
two methods of addition of the catalyst to the polymerization
reactor. In method A, 200 psig argon gas was used to inject
20 mL of aged catalyst/cocatalyst toluene solution from a 50
mL single-ended injection tube into a 300 mL stirred Parr
reactor. Method B employed propylene pressure and a double-
ended injection tube to inject the catalyst solution.
The productivities of these catalysts range from 200 to 3000
kg of PP/(mol of M‚h), depending on the reaction conditions.
The productivities of the hafnium complexes are approximately
equal to those the zirconium complexes. The trends for the
different ligands are less clear, but the zirconium bis(triflouro-
methyl) derivative 3 appears to be slightly less productive than
the 2-phenylindene derivative 1. As previously observed, there
is generally an increase in productivity with increasing propylene
pressures for each catalyst system.4
tions were carried out on a locally written routine. These studies
revealed that the rate of exchange of the cyclopentadienyl
protons was equal to that of the benzylic protons, strongly
implicating indenyl rotation as the dynamic process resulting
in the observed exchanges.
The exchange rates are presented in Table 3. The values of
∆Gq were calculated at each temperature using the Eyring
equation (eq 1). A plot of ∆Gq versus temperature14 yielded
k ) κ(kBT/h) exp(-∆Gq/RT)
(1)
the following activation parameters: 6, ∆Hq ) 7.8((2.3) kcal/
mol, ∆Sq ) -15.4((10.1) cal/mol; 7, ∆Hq ) 7.1((3.0) kcal/
mol, ∆Sq ) -17.3((11.0) cal/mol.9
Comparison of the rates and activation parameters for the
zirconium derivative 6 and the hafnium derivative 7 indicate
that the exchange rates for the two different metals were the
same within experimental error.
Molecular weights produced by all complexes in solution are
in the range of 200 000 < Mw < 1 000 000. Molecular weights
(13) Alexander, S. J. Chem. Phys. 1962, 37, 974-980.
(14) Binsch, G. Topics Stereochem. 1968, 3, 97-192.

2-Arylindene Catalysts for Elastomeric Polypropylene
J. Am. Chem. Soc., Vol. 119, No. 46, 1997 11177
1 yields elastomeric polypropylenes with isotactic pentad
contents in the range of [mmmm] ) 28-36%. The value of
[mmmm] increases with increasing propylene pressure. For
solution polymerizations, a decrease in isotacticity is observed
with increasing temperature.
Polymers produced with the PhInd hafnium catalyst 2 are
tacky atactic materials that exhibit no melting transition. The
microstructure of polymer varies only slightly with propylene
pressure; the isotacticity index ranges from [mmmm] ) 6% to
10% upon increasing propylene pressure from 25 to 100 psig.
Molecular weights exhibit the same trends as those for 1.
Productivity of 2 is approximately the same as that of 1 and
similarly increases slightly with increasing pressure and tem-
perature.
The influence of the ligand on the stereospecificity, as
previously described,4b can be significant: the (CF3)2PhInd
zirconium derivative 3 yields polypropylenes with higher
isotacticities than those of the unsubstituted 2-PhInd catalyst 1.
A similar trend is observed for the hafnium derivatives 4 and
2, although the differences in these two cases are small. The
nature of the metal also has a large effect on the stereospeci-
ficity: for both ligands, the zirconium derivatives 1 and 3
produce more isotactic polymers than their hafnium congeners
2 and 4. Productivities of 3 and 4 are about half those of the
corresponding 2-PhInd derivatives 1 and 2, and again the
hafnium catalyst exhibits productivity similar to that of the
zirconium catalyst. Molecular weights and polydispersities are
comparable.
Figure 7. Variable temperature spectra and simulations of 6.
Discussion
Unbridged 2-arylindene catalysts produce elastomeric poly-
propylenes with a wide range of isotacticities.4 The nature of
the ligand has a dramatic effect on the stereospecificity of the
catalyst.4 In this paper, we report that the stereospecificity of
this class of catalysts is also quite sensitive to the nature of the
metal atom (Zr vs Hf).
Zirconium and hafnium are similar group 4 metals separated
by only one row on the periodic table of the elements.
Complexes of zirconium and hafnium with the same ligand
environments are closely isostructural and possess similar
chemical properties.16 In fact, the solid-state structures of
zirconium and hafnium metallocenes derived from 2-arylindene
ligands are nearly identical. Comparison of the crystal structures
of the (PhInd) zirconium and hafnium complexes 1 and 2 as
well as the ((CF3)2PhInd) complexes 3 and 4 reveals close
structural similarity. For each hafnium-zirconium pair, cor-
responding metal-cyclopentadienyl bond lengths are nearly
identical, with the hafnium-cyclopentadienyl bonds approxi-
mately 0.01 Å shorter, consistent with the lanthanide contraction.
The solution structure and dynamics of hafnium and zirco-
nium complexes also appear very similar. Solution NMR
studies of the zirconium dichloride 1 and the corresponding
dimethyl derivative (2-PhInd)2ZrMe2 (5) from -80 to 25 °C
reveal time-averaged C2 symmetry consistent with rapid rotation
of the indenyl ligands in solution. Substitution of the choride
ligands of 1 and 2 for benzyl ligands yields the dibenzyl
Figure 8. Variable temperature spectra and simulations of 7.
increase with increasing propylene pressure of polymerization
as well as with decreasing temperature. The Mw of polymers
produced by hafnium catalysts are approximately the same as
those produced by the zirconium catalysts. The polydispersities
(Mw/Mn) of polymers produced with (2-PhInd)2HfCl2 (2) are
close to Mw/Mn ) 2; polydispersities for polymers obtained from
zirconium 3 are as high as Mw/Mn ) 4. There appears to be a
correlation between isotacticity ([mmmm]) and the polydisper-
sity. Polydispersities appear to be higher for catalysts that
produce more highly isotactic polymer: 2 < 4 < 1 < 3.
The microstructure of polypropylene is measured by examin-
ing the methyl region of the polymer 13C NMR spectrum, which
gives the relative stereochemistry to the pentad level of
stereosequences.15 Two measures of isotacticity are the percent-
age of meso dyads ([m]) and the percentage of isotactic pentads
([mmmm]). Polymerization of propylene with the zirconocene
1
derivatives 6 and 7. The low-temperature (-70 °C) H NMR
spectra of (2-PhInd)2Zr(CH2Ph)2 (6) and (2-PhInd)2Zr(CH2Ph)2
(7) show two cyclopentadienyl resonances and two doublets
corresponding to diastereotopic benzylic protons. These spectra
imply that the observed exchange at -70 °C is between the
(16) Cardin, D. J.; Lappert, M. F.; Raston, C. L. Chemistry of Organo-
Zirconium and -Hafnium Compounds; Ellis Horwood: New York, 1986;
pp 1-451.
(15) Zambelli, A.; Locatelli, P.; Bajo, G.; Bovey, F. A. Macromolecules
1975, 8, 687-689.

11178 J. Am. Chem. Soc., Vol. 119, No. 46, 1997
Bruce et al.
Table 3. Dynamic NMR Data for (2-PhInd)2M(CH2Ph)2 Complexes 6 and 7
M ) zirconium (6)
M ) hafnium (7)
k/Hz
T/°C
k/Hz
∆Gq/kcal
T/°C
∆Gq/kcal
19.0 ( 0.2
12.5
1.6
6700 ( 300
3300 ( 200
1600 ( 100
775 ( 50
400 ( 25
150 ( 20
75 ( 12
12.0 ( 0.1
12.0
12.0
11.9
11.8
11.7
11.5
11.3
11.1
20.0 ( 0.2
10.0
0.2
6800 ( 300
3500 ( 200
1800 ( 100
850 ( 50
500 ( 25
290 ( 20
140 ( 12
60 ( 5
12.0 ( 0.1
11.9
11.9
11.8
11.6
11.4
11.1
11.0
10.7
-8.2
-9.8
-18.1
-29.7
-40.3
-50.8
-59.4
-70.1
-20.8
-30.7
-42.7
-52.7
-62.7
-71.9
32 ( 5
18 ( 3
13 ( 2
32 ( 3
10.7
20 ( 2
10.4
Table 4. Polypropylenes Produced with Catalysts 1-4
b
b
expt
cat.
method
conditionsd
prodvtya
Mw
Mw/Mn
mc (%)
mmmmc (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
1
1
1
1
1
2
2
2
2
2
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
Bd
B
B
B
B
B
B
B
B
B
A
A
A
A
B
B
B
B
A
A
A
A
B
B
B
B
35 psig, 20 °C
50 psig, 20 °C
75 psig, 20 °C
100 psig, 20 °C
75 psig, 0 °C
35 psig, 20 °C
50 psig, 20 °C
75 psig, 20 °C
100 psig, 20 °C
50 psig, 0 °C
25 psig, 25 °C
35 psig, 25 °C
50 psig, 25 °C
75 psig, 25 °C
35 psig, 20 °C
50 psig, 20 °C
75 psig, 20 °C
100 psig, 20 °C
35 psig, 25 °C
50 psig, 25 °C
75 psig, 25 °C
90 psig, 25 °C
35 psig, 20 °C
50 psig, 20 °C
75 psig, 20 °C
100 psig, 20 °C
952
832
1480
960
1544
1080
1460
1440
1180
1160
250
500
730
1370
206
582
624
638
1130
1350
2190
3050
506
360
660
700
232 000
250 000
399 000
435 000
973 000
345 000
198 000
375 000
486 000
812 000
3.8
3.3
3.8
3.0
3.2
2.3
2.0
2.2
2.1
2.1
63
66
68
70
73
55
57
58
57
56
75
78
80
86
85
85
88
87
55
64
58
57
54
56
56
57
23
27
29
31
35
8
9
9
9
10
45
51
58
73
56
59
64
67
12
18
15
21
10
11
11
12
243 000
296 000
332 000
454 000
470 000
631 000
608 000
285 000
330 000
415 000
483 000
292 000
376 000
474 000
531 000
3.2
3.4
3.7
4.0
3.9
4.3
5.6
2.9
2.4
2.4
2.5
2.4
2.2
2.4
2.3
a In kg of PP/(mol of Zr‚h). b Determined by GPC vs polypropylene. c Determined by 13C NMR spectroscopy. d [M] ) 5.0 × 10-5 M, [Al]/[Zr]
) 1000, t ) 1 h. Method A, [M]/MAO in 20 mL of toluene injected under 200 psig Ar; method B, [M]/MAO in 20 mL of toluene injected under
P + 5 psig propylene.
two enantiomers of the chiral anti conformation. The syn (or
meso) conformer is not observed as an energetic minimum,
although it cannot be ruled out as an intermediate in the
exchange process.
Line shape analysis was used to determine the rates of
exchange. The benzylic protons were modeled as a coupled
two-site exchange, and the cyclopentadienyl protons were
modeled as an uncoupled two-site exchange.6 These model
spectra were compared to the observed spectra to determine the
rate of exchange at each temperature. At any given temperature,
the cyclopentadienyl protons exhibit the same rate of exchange
as the benzylic protons, strongly implicating indenyl rotation
as the mechanism for the observed exchange.
Comparison of the data obtained from line shape analysis of
the VT spectra of zirconium dibenzyl 6 and hafnium dibenzyl
7 leads to the conclusion that exchange rates of 6 and 7 are the
same within experimental error at each temperature. The Eyring
equation yields ∆Gq values that are nearly the same for hafnium
as zirconium. ∆Hq and ∆Sq were calculated from a plot of ∆Gq
vs temperature. The values differ slightly from zirconium to
hafnium but are within the wider experimental error for this
method. These data indicate that hafnium and zirconium exhibit
nearly identical solution behavior, consistent with the nearly
identical X-ray crystal structures.
All catalysts 1-4 produce polypropylene when combined
with MAO under a propylene atmosphere. In each case, the
productivities of the hafnium and zirconium catalysts are
approximately the same. The similar productivities of the Hf
and Zr catalysts could be a consequence of similar rate constants;
however, hafnium complexes are typically less reactive than
the zirconium congeners.17 The other possibility is that, for
hafnium, the average number of active sites over the course of
the polymerization is greater than that for zirconium, but the
number of active centers is balanced by a lower reactivity of
hafnium.
Zirconium and hafnium catalysts with identical ligand
environments produce polymers with approximately the same
weight-average molecular weights, 200 000 < Mw < 1 000 000.
As the hafnium catalysts produce more monodisperse polymer
than zirconium, the number-average molecular weights are larger
for Hf than for Zr, although this difference is small for the
triflouromethyl-substituted ligands. The values of Mw/Mn for
atactic polymers produced by the hafnium catalysts are ap-
proximately 2.1, consistent with the value of 2 observed for
homogeneous Ziegler-Natta systems. The higher Mw/Mn for
(17) Ewen, J. A.; Haspeslach, L.; Atwood, J. L.; Zhang, H. J. Am. Chem.
Soc. 1987, 109, 6544-6545.

2-Arylindene Catalysts for Elastomeric Polypropylene
J. Am. Chem. Soc., Vol. 119, No. 46, 1997 11179
Scheme 1. Proposed Mechanism of Polymerization
the more isotactic polymers might be a consequence of a two-
site switching process, where each site is predisposed to give
different molecular weight polymers.18
against either the equilibrium constant or the rate of intercon-
version as the origin of the lower stereoselectivity for hafnium
relative to zirconium. Thus, we believe the most reasonable
explanation is that hafnium has a slower rate of propagation at
the isospecific site than zirconium (kp[Hf] < kp[Zr]). If the rate
of propagation for hafnium is slower, such that it is similar to
the rate of inconversion of the two sites, then the polymer
produced would have very short isotactic block lengths and have
a lower [mmmm].
The conclusion that rate of propagation at the isospecific site
is lower for hafnium is what might be expected, given that
organohafnium compounds are typically less reactive than
organozirconium compounds.16 However, this conclusion is
seemingly inconsistent with similar productivities of zirconium
and hafnium catalysts. However, as stated previously, produc-
tivities are not necessarily indicative of the rate constants of
propagation.19 Similar productivities for hafnium could be a
consequence of a higher number of active centers for hafnium
relative to those for zirconium but a lower rate of propagation
at hafnium centers than at zirconium centers.20 Assuming that
the two effects balance each other, it may be concluded that
hafnium produces atactic polymer by slow propagation at many
sites, and zirconium produces stereoblock polymer by fast
propagation at a few sites.
In spite of the close structural similarities and similar
rotational rate for the indenyl ligands, the zirconium and hafnium
complexes produce quite different polypropylene microstruc-
tures. For both 2-PhInd and (CF3)2PhInd systems, the zirconium
derivatives yield much higher isotacticities ([mmmm] by 13C
NMR) than the corresponding hafnium complexes. These
differences are difficult to rationalize on purely structural
grounds. However, we have previously proposed a mechanism
that might account for the large differences in stereospecificity
for these 2-arylindene zirconium and hafnium metallocenes. The
mechanistic model invokes the interconversion of different
catalyst states during the polymerization:4 a chiral rac-like state
(A) and an achiral meso-like state (B) (Scheme 1). State A is
predicted to produce isotactic polypropylene, while state B
should yield atactic polypropylene. The exchange of sites is
postulated to occur by rotation of the unbridged indenyl ligands.
As long as the rate of site switching is slower than the rate of
polymerization, the catalyst will produce blocks of isotactic
stereosequences and blocks of atactic stereosequences; if the
rate of rotation is faster than the rate of propagation, block
lengths will become negligible, and polymer stereochemistry
will be random (atactic).
A further point with regard to the proposed mechanism is
the rate of rotation of the dibenzyl compounds relative to the
rate of polymerization. The measured rates of rotation for the
dibenzyl derivatives are faster than reasonable estimates for the
propagation rates. A turnover frequency for propylene insertion
of 10 s-1 can be estimated for catalyst 1/MAO at 75 psig and
20 °C.21 This is several orders of magnitude slower than the
rate of rotation of 6700 Hz observed for the dibenzyl complexes.
(This is true even if less than 100% active centers is assumed.)19
These rates are seemingly inconsistent with the proposed
mechanism, since, for a block structure to be produced, the rate
of polymerization should be faster than the rate of interconver-
sion of the two sites. However, the rate of interconversion for
the active catalyst is not known, and the rates measured for the
dibenzyls may be poorly representative of the rate for the actual
catalyst. The fact that the rate of ligand rotation for the dibenzyl
derivatives 6 and 7 is slower than that for either the dichlorides
1 and 2 or the dimethyl 5 demonstrates that the rate of ligand
rotation is quite sensitive to the nature of the ligands bound to
the transition metal center. Given that the active catalysts
contain a growing polymer chain and are likely cationic with
some associated anion of ill-defined structure, it is likely that
the rate of ligand rotation will be different from that of the
dibenzyl derivatives. Current efforts are underway to devise
more realistic models for the active catalyst to investigate the
ligand dynamics.
According to this mechanism, the rate of conversion of state
A to state B is given by RA ) kf[A], and that of state B to site
A is given by RB ) kr[B]. There is also an equilibrium constant
for this exchange, given by Keq ) kr/kf. The rate of polymer-
ization is assumed to be second order: RpA ) kpA[C3H6][A] for
the isospecific site and RpB ) kpB[C3H6][B] for the aspecific
site. According to these definitions, the block lengths of the
polymers are defined as the ratio of propagation to rotation:
BLA ) RpA/RA and BLB ) RpB/RB. Therefore, only when RA
< RpA will blocks of isotactic polymer will be produced. As
RA approaches RpA, the blocks will grow progressively shorter.
As the length of an isotactic stereosequence approaches unity,
placement of stereocenters becomes random, and atactic polymer
is produced.
As the polymers made from the hafnium catalysts have the
lowest isotacticities ([mmmm] and [m]), it follows that hafnium-
produced polymers have shorter average isotactic sequence
lengths than polymers made from the zirconium catalysts. In
the context of this mechanism, there are three possible explana-
tions for the differences between hafnium and zirconium
catalysts: a difference in the steady-state concentration of
isospecific and aspecific states (Keq), a difference in the rate of
interconversion of the two states, and/or a different rate of
propagation for the isospecific site. The structural similarities
of the zirconium and hafnium complexes and the similar rates
of rotation for the dibenzyl derivatives of the metallocenes argue
(19) Chien, J. C. W. J. Polym. Sci. A.: Polym. Chem. 1991, 29, 459-
470.
(20) A higher average active site concentration for hafnium might result
from a slower rate of inactivation of hafnium catalysts sites over the course
of the polymerization reaction.
(21) At 75 psig, 20 °C, the productivity of 1 is 1480 kg of PP/mol of Zr.
This corresponds to 9.78 insertions/s.
(18) For example, if the isospecific site is predisposed to give Mw
)
500 000 with PDI ) 2.0, and the aspecific site gives Mw ) 100 000 with
PDI ) 2.0, a switching site catalyst would give an intermediate Mw with a
much broader PDI in order to encompass both the isotactic and atactic only
distributions. See: Coleman, B. D.; Fox, T. G. J. Am. Chem. Soc., 1963,
85, 1241-1244.

11180 J. Am. Chem. Soc., Vol. 119, No. 46, 1997
Bruce et al.
1H NMR (400 MHz, 20 °C, CDCl3): δ 7.65 (s, 2H), 7.51 (s, 4H),
6.7-7.3 (m, 8H), 5.63 (s, 4H). 13C{1H} NMR (100 MHz, 20 °C,
C6D6): δ 135.8, 132.9, 131.6, 127.2, 126.3, 126.0, 125.6, 123.8, 121.7,
100.1. Anal. Calcd for C34H18F2HfCl2: C, 45.10; H, 1.87. Found:
C, 45.18; H, 2.01.
In summary, we have provided an unusual example of
isostructural metallocene catalysts which have quite different
stereospecificities in propylene polymerization. This is remark-
able in that, in stereospecific catalysis, it is generally held that
the coordination geometry of the catalyst site is a key element
in stereodifferentiation. The large differences between the
isostructural hafnium and zirconium catalysts are most readily
rationalized in terms of the different dynamics of propagation
relative to isomerization in these dynamic catalyst systems.
Bis(2-phenylindenyl)zirconium Dibenzyl (6). To a solution of 205
mg (0.376 mmol) of bis(2-phenylindenyl)zirconium dichloride in 20
mL of THF was added 0.451 mL (0.903 mmol) of benzylmagnesium
chloride (1.0 M in THF) at room temperature under Ar. After 24 h,
the THF was removed in Vacuo and replaced with toluene. The orange
solution was filtered and dried in Vacuo. The compound was isolated
by recrystallization from toluene to give 102 mg (41%) of orange
crystals.
1H NMR (400 MHz, -60 °C, CD2Cl2): δ 7.57 (t, 4H, J ) 7 Hz),
7.43 (m, 6H), 7.21 (t, 4H, J ) 7 Hz), 7.02 (t, 4H, J ) 8 Hz), 6.86 (t,
2H, J ) 7 Hz), 6.78 (d, 4H, J ) 8 Hz), 6.51 (dd, 4H, J ) 62, 7 Hz),
5.78 (d, 4H, J ) 502 Hz), 0.15 (dd, 4H, J ) 123, 11 Hz) ppm. 1H
NMR (400 MHz, 17 °C, CD2Cl2): δ 7.57 (t, 4H, J ) 7 Hz), 7.43 (m,
6H), 7.21 (t, 4H, J ) 7 Hz), 7.02 (t, 4H, J ) 8 Hz), 6.86 (t, 2H, J )
7 Hz), 6.78 (d, 4H, J ) 8 Hz), 6.51 (s, 4H), 5.78 (s, 4H), 0.15 (s, 4H)
ppm. Anal. Calcd for C44H36Zr: C, 80.56; H, 5.53. Found: C, 80.53;
H, 5.40.
Experimental Section
General Comments. All organometallic reactions were performed
under an inert argon or nitrogen atmosphere using standard Schlenk or
drybox techniques. 2-Phenylindene, 2-[3,5-bis(trifluoromethyl)phenyl]-
indene, bis(2-phenylindenyl)zirconium dichloride, and bis(2-[3,5-bis-
(trifluoromethyl)phenyl]indenyl)zirconium dichloride were prepared as
previously described.4 Benzylmagnesium chloride (1.0 M in THF),
phenylmagnesium bromide (3.0 M in diethyl ether), and n-butyllithium
(2.5 M in hexanes) were purchased from Aldrich and used without
further purification. Zirconium tetrachloride was purchased from Fluka
and used without further purification. Hafnium tetrachloride (99.9%,
0.14% Zr) was purchased from Cerac and used without further
purification. MAO (type IV, 7.4% in toluene) was purchased from
Akzo and dried in Vacuo prior to use.
THF, diethyl ether, and benzene were each distilled from sodium-
benzophenone ketyls prior to use. Propylene and toluene were passed
through two columns: one containing alumina (Kaiser, dried by passing
N2 through the column at 350 °C for 5 h) and one containing Q-5
reactant (Englehard, prepared by passing N2 through the column at 300
°C for 3 h, followed by 5% H2 for 3 h, followed by N2 for 3 h).22
1H and 13C NMR spectra were recorded on a Varian XL-400
spectrometer operating at 400 MHz for protons and 100 MHz for 13C.
All polymer 13C spectra were deconvoluted on an Silicon Graphics
Indigo workstation using Felix. GPC data were recorded in trichloro-
benzene on a GPC operating at 135 °C and referenced to a polypro-
pylene standard. Productivities of the catalysts were calculated from
the weight of polymer produced.
Bis(2-phenylindenyl)hafnium Dibenzyl (7). To a solution of 98
mg (0.155 mmol) of bis(2-phenylindenyl)hafnium dichloride in 20 mL
of THF was added 0.170 mL (0.341 mmol) of benzylmagnesium
chloride (1.0 M in THF) at room temperature under Ar. After 24 h,
the THF was removed in Vacuo and replaced with toluene. The orange
solution was filtered and dried in Vacuo. The product was isolated by
recrystallization from toluene to give 52 mg (42%) of yellow crystals.
1H NMR (400 MHz, -70 °C, CD2Cl2): δ 7.60 (t, 4H, J ) 7 Hz),
7.44 (m, 6H), 7.22 (t, 4H, J ) 7 Hz), 7.12 (t, 4H, J ) 8 Hz), 6.84 (t,
2H, J ) 7 Hz), 6.76 (d, 4H, J ) 8 Hz), 6.47 (dd, 4H, J ) 38, 9 Hz),
5.82 (d, 4H, J ) 445 Hz), -0.09 (dd, 4H, J ) 237, 12 Hz) ppm. 1H
NMR (400 MHz, 17 °C, CD2Cl2): δ 7.60 (t, 4H, J ) 7 Hz), 7.44 (m,
6H), 7.22 (t, 4H, J ) 7 Hz), 7.12 (t, 4H, J ) 8 Hz), 6.84 (t, 2H, J )
7 Hz), 6.76 (d, 4H, J ) 8 Hz), 6.47 (s, 4H), 5.82 (s, 4H), -0.09 (s,
4H) ppm. Anal. Calcd for C44H36Hf: C, 71.10; H, 4.88. Found: C,
71.27; H, 5.10.
Dynamic NMR simulations were performed on a Macintosh
computer, using locally written software, DYNMR (see Supporting
Information for full details of dynamic NMR simulations).
Crystallography. Single crystals suitable for X-ray analysis were
prepared by recrystallization from toluene (1 and 2) or toluene/hexane
(4) at -25 °C.
Metallocene Synthesis. Bis(2-phenylindenyl)hafnium Dichloride
(2). To a stirred solution of 1.89 g (9.83 mmol) of 2-phenylindene
and 40 mL of THF was added dropwise 3.93 mL (9.83 mmol) of
n-butyllithium (2.5 M in hexanes) at -78 °C under Ar. After the
solution was warmed to room temperature, the THF was removed in
Vacuo and replaced with toluene. Next, 1.57 g (4.90 mmol) of hafnium
tetrachloride was added via cannula as a slurry in 20 mL of toluene at
room temperature. After 24 h, the yellow solution was filtered over
Celite. The compound was isolated by recrystallization from toluene
to give 0.62 g (20%) of yellow crystals.
1H NMR (400 MHz, 20 °C, CDCl3): δ 7.36 (d, 4H, J ) 7.2 Hz),
7.18 (m, 4H), 7.12 (m, 2H), 7.07 (dd, 4H, J ) 6.6, 3.1 Hz) 6.88 (dd,
4H, J ) 6.6, 3.1 Hz), 6.29 (s, 4H) ppm. 13C{1H} NMR (100 MHz, 20
°C, C6D6): δ 133.7, 132.1, 128.9, 128.5, 127.3, 126.5, 126.1, 125.2,
101.4 ppm. Anal. Calcd for C30H22HfCl2: C, 57.02; H, 3.51; Cl, 11.22.
Found: C, 56.64; H, 3.65; Cl, 10.91.
Bis(2-[3,5-bis(trifluoromethyl)phenyl]hafnium Dichloride (4). n-
Butyllithium (1.6 M in hexanes, 2 mL, 3.20 mmol) was added dropwise
at ambient temperature to a solution of 2-[3,5-bis(trifluoromethyl)-
phenyl]indene (1.03 g, 3.14 mmol) in 10 mL of diethyl ether. After
the solution was stirred for 30min, the solvent was removed in Vacuo,
leaving a green-yellow solid. In an N2 drybox, HfCl4 (510 mg, 1.59
mmol) was added to the lithium salt. The solids were then cooled to
-78 °C, at which temperature 45 mL of toluene was slowly added.
The flask was allowed to reach ambient temperature, and the suspension
was stirred at room temperature for 24 h. The solvent was removed
in Vacuo, and the residual solid was extracted with CH2Cl2. The
solution was filtered and purified by precipitation by hexanes from
CH2Cl2.
Data Collection and Reduction of Bis(2-phenylindenyl)zirconium
Dichloride (1). A yellow plate crystal of 1, C30H22Cl2Zr, with
approximate dimensions of 0.600 mm × 0.400 mm × 0.150 mm, was
mounted in paratone oil on a glass fiber. The specimen was placed in
a cold stream of nitrogen on an Enraf-Nonius CAD4 diffractometer
with graphite-monochromated Mo KR radiation. A total of 8658
reflections were collected, of which 8274 were unique (Rint ) 0.026).
Crystallographic details are summarized in Table 5.
The structure was solved by direct methods23 and expanded using
Fourier techniques.24 All non-hydrogen atoms were refined anisotro-
pically. Hydrogen atoms were located by difference Fourier maps but
included at idealized position 0.95 Å from their parent atoms. Hydrogen
atoms were included but not refined. The final cycle of full-matrix
least-squares refinement was based on 6362 observed reflections (I >
3σ(I)) and 613 variable parameters and converged (largest parameter
shift was 0.20 times its estimated standard deviation) with unweighted
and weighted agreement factors of R ) 0.032 and Rw ) 0.032.
Two rotational isomers of the complex were observed in the refined
structure. The complex displaying eclipsed ligands (syn) was deter-
mined to be slightly disordered and was refined at 96% occupancy.
Only the heavy atoms (Zr(3), Cl(5), and Cl(6)) of the disordered
molecule were observed and were modeled at 4%. All calculations
were performed using the teXsan software package (Molecular Structure
Corp.).
(23) SAPI91: Fan, H.-F. Structure Analysis Programs with Intelligent
Control; Rigaku Corp. Tokyo, Japan, 1991.
(24) DIRDIF92: Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman,
W. P.; Garcia-Granda, S.; Gould, R. O.; Smits, J. M. M.; Smykalla, C. The
DIRDIF program system; Technical Report of the Crystallography Labora-
tory, University of Nijmegen, The Netherlands, 1992.
(22) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518-1520.

2-Arylindene Catalysts for Elastomeric Polypropylene
J. Am. Chem. Soc., Vol. 119, No. 46, 1997 11181
Table 5. Crystallographic Details of 1, 2, and 4
1
2
4
empirical formula
formula weight
color; habit
crystal size (mm)
crystal system
space group
unit cell dimensions
a (Å)
C30H22Cl2Zr
544.63
yellow plate
0.60 × 0.40 × 0.15
triclinic
C30H22Cl2Hf
631.9
yellow plate
C34H18Cl2F12Hf
903.9
yellow prism
0.23 × 0.24 × 0.36
monoclinic
P2l/n
0.04 × 0.20 × 0.23
triclinic
P1h (No. 2)
P1h (No. 2)
13.538(2)
14.553(2)
15.365(1)
65.780(1)
64.950(1)
63.160(6)
2355.4(6)
13.4826(16)
14.5600(11)
15.3248(12)
65.740(7)
64.631(9)
63.116(9)
16.983(2)
8.3677(11)
22.495(3)
b (Å)
c (Å)
R (deg)
â (deg)
105.862(5)
γ (deg)
volume (Å3)
Z
2334.8(4)
3079.0(7)
4
4
4
density (calcd)
absorption coeff
F(000)
1.563 g/cm3
1.798 mg/m3
4.714 mm-1
1232
1.950 mg/m3
7.1
1104.00
3.660 mm-1
1744
diffactometer used
radiation
temperature (K)
monochromator
2θ range
scan type
scan speed
reflections collected
ind reflns
obsd reflns
Enraf-Nonius CAD-4
Mo KR(λ ) 0.710 73 Å)
203
Siemens P4RA
Mo KR (λ ) 0.710 73 Å)
158
Siemens P4
Mo KR (λ ) 0.710 73 Å)
163
highly-oriented graphite
highly oriented graphite crystal
4.0-50.0°
highly oriented graphite crystal
40° < 2θ < 42°
4.0-50.0°
ω
θ-2θ
θ-2θ
5.5 deg/min (θ)
const. (ω, 4.00 deg/min)
8513
const. (ω, 3.00 deg/min)
8658
6029
8274(Rint ) 2.60%)
6362
613
RF ) 3.2%, RwF ) 3.2%
10.4
0.62
-0.88
8138 (Rint ) 1.12%)
7050 (F > 3.0σ(F))
595
RF ) 3.6%, RwF ) 4.1%
11.8:1
5.08
-1.25
5131 (Rint ) 7.3%)
4834 (F > 2.0σ(F))
no. params refined
R indicies (obsd data)
data-to-param ratio
largest diff peak (e Å-3
272
RF ) 7.5%, RwF ) 8.9%
17.8:1
6.17
)
)
largest diff hole (e Å-3
-4.20
Data Collection and Reduction of Bis(2-phenylindenyl)hafnium
Dichloride (2). A yellow crystal of approximate dimensions 0.04 mm
× 0.20 mm × 0.23 mm was oil-mounted on a glass fiber and transferred
to the Siemens P4 rotating-anode diffractometer. Determination of Laue
symmetry, crystal class, unit cell parameters, and the crystal’s orienta-
tion matrix were carried out according to standard techniques.25 Low
temperature (158 K) intensity data were collected via a θ-2θ scan
technique with Mo KR radiation. Crystallographic details are given
in Table 5.
carried out according to standard techniques.25 Low-temperature (163
K) intensity data were collected via a θ-2θ scan technique with Mo
KR radiation. Crystallographic details are given in Table 5.
All 6029 data were corrected for Lorentz and polarization effects
and placed on an approximately absolute scale. An absorption
correction was not done because the orientation of the crystal was
inadvertently changed prior to collection of ψ-scans. All crystal-
lographic calculations were carried out using the UCI-modified ver-
sion of the UCLA Crystallographic Computing Package27 and the
SHELXTL PLUS program set.26 The quantity minimized during least-
squares analysis was ∑w(|F0| - |Fc)2, where w-1 ) σ2(|F0|) +
0.0002(|F0|)2.
All 8513 data were corrected for absorption26 and for Lorentz and
polarization effects and placed on an approximately absolute scale. Any
reflection with I(net) < 0 was assigned the value |F0| ) 0. There were
no systematic extinctions nor any diffraction symmetry other than the
Friedel condition. All crystallographic calculations were carried out
using the UCI-modified version of the UCLA Crystallographic
Computing Package27 and the SHELXTL PLUS program set.26 The
quantity minimized during least-squares analysis was ∑w(|F0| - |Fc)2,
where w-1 ) σ2(|F0|) + 0.0002(|F0|)2.
The structure was solved by direct methods (SHELXTL) and refined
by full-matrix least-squares techniques. There are two independent
molecules of the formula unit present. Hydrogen atoms were included
using a riding model with d(C-H) ) 0.96 Å and U(iso) ) 0.06 Å2.
Refinement of the model led to convergence with RF ) 7.5%, RwF
)
8.9%, and GOF ) 3.46 for 272 variables refined against those 4834
data with |F0| > 2.0σ(|F0|)). A final difference Fourier map yielded
F(max) ) 6.2 e Å-3 at distances of 0.93 and 1.05 Å from Hf(1). The
high difference peaks, the necessity for isotropic refinement of carbon
atoms, and the higher than expected residuals are most likely due to
the absence of an absorption correction.
The structure was solved by direct methods (SHELXTL) and refined
by full-matrix least-squares techniques. There are two independent
molecules of the formula unit present. Hydrogen atoms were included
using a riding model with d(C-H) ) 0.96 Å and U(iso) ) 0.06 Å2.
Refinement of the model led to convergence with RF ) 3.6%, RwF
)
4.1%, and GOF ) 1.78 for 595 variables refined against those 7050
data with |F0| > 3.0σ(|F0|)). A final difference Fourier map yielded
F(max) ) 5.08 e Å-3 at a distance of 1.873 from Hf(2) and 1.805 from
Cl(3).
Data Collection and Reduction of Bis(2-[3,5-bis(trifluoromethyl)-
phenyl]indenyl)hafnium Dichloride(4). A yellow crystal of ap-
proximate dimensions 0.23 mm × 0.24 mm × 0.36 mm was
oil-mounted on a glass fiber and transferred to the Siemens P4 rotating-
anode diffractometer. The determination of Laue symmetry, crystal
class, unit cell parameters, and the crystal’s orientation matrix was
Polymerization of Propylene. Method A. In a N2-filled glovebox,
a 100 mL two-ended SS (Wiley) injection tube was loaded with 80
mL of toluene. Then, 2.5 mL (5 × 10-6 mol from stock solution of
20 × 10-6 mol in 10.0 mL of toluene) of metallocene dichloride was
mixed with 17.5 mL of toluene and 289 mg (5 × 10-3mol) of MAO.
The solution was aged for 10 min and loaded into a 50 mL single-
ended SS (Wiley) injection tube.
After a 300 mL SS Parr reactor was cleaned and evacuated to 30
mTorr, it was pressurized to 150 psig Ar and vented three times.
Toluene was injected into the reactor under 100 psig Ar from the 100
mL injection tube. The reactor was vented and refilled twice with
propylene of the appropriate pressure. The reactor was allowed to
equilibrate for 10 min. The equilibrated catalyst solution in the 50
mL injection tube was injected into the reactor with 200 psig Ar.
(25) XSCANS Software Users Guide, Version 2.1; Siemens Industrial
Automation, Inc.: Madison, WI, 1994.
(26) SHELXTL.
(27) UCLA Crystallographic Computing Package.

11182 J. Am. Chem. Soc., Vol. 119, No. 46, 1997
Bruce et al.
The reactor pressure and temperature were maintained for 60 min.
The reaction was quenched by injecting 10 mL of methanol into the
reactor. The reactor was vented, and the polymer/toluene slurry was
stirred in 300 mL of methanol for 6 h. The solvent was decanted, and
the polymer was dried overnight in a vaccuum oven at 40 °C.
Method B. In a N2-filled glovebox, a 100 mL two-ended SS (Wiley)
injection tube was loaded with 80 mL of toluene. Then, 2.5 mL (5 ×
10-6 mol from stock solution of 20 × 10-6 mol in 10.0 mL of toluene)
of metallocene dichloride was mixed with 17.5 mL of toluene and 289
mg (5 × 10-3 mol) of MAO. The solution was aged for 10 min and
loaded into a 50 mL single-ended SS (Wiley) injection tube.
After a 300 mL SS Parr reactor was cleaned and evacuated to 30
mTorr, it was pressurized to 150 psig Ar and vented three times.
Toluene was injected into the reactor under 100 psig Ar from the 100
mL injection tube. The reactor was vented and refilled twice with
propylene of the appropriate pressure. The reactor was allowed to
equilibrate for 10 min, at which time the reactor was vented to 10 psig
below reaction pressure. The equilibrated catalyst solution in the 50
mL injection tube was injected into the reactor with propylene of
reaction pressure.
The reactor pressure and temperature were maintained for 60 min.
The reaction was quenched by injecting 10 mL of methanol into the
reactor. The reactor was vented, and the polymer/toluene slurry was
stirred in 300 mL of methanol for 6 h. The solvent was decanted, and
the polymer was dried overnight in a vaccuum oven at 40 °C.
Acknowledgment. We thank Amoco Chemical Co. and the
NSF for financial support (CHE-9615699 and DMR-9258324).
R.M.W. is the recipient of NSF’s Alan T. Waterman Award,
for which he is grateful. We thank Prof. Dan Stack for assistance
with the structural determination of 1.
Supporting Information Available: Crystal structure pa-
rameters and coordinates for 1, 2, and 4, and dynamic NMR
spectra and simulations for 6 and 7 (43 pages). See any current
masthead page for ordering and Internet access instructions.
JA971895R