Ansa-zirconocenes based on N-substituted 2-methylcyclopenta[b]indoles: Synthesis and catalyst evaluation in liquid propylene polymerization

Article abstract of DOI:10.1021/om020505t

Cyclopenta[b]indole-based ansa-zirconocenes were prepared and studied as catalysts for propylene polymerization. Substituted 2-methyl-4-R-1,4-dihydrocyclopenta[b]indoles (R = Ph, 7a; R = o-Tol, 7b; R = Me, 7c) were synthesized and converted into the racemic forms of the corresponding ansa-zirconocenes Me₂Si(2-Me-4-R-cyclopenta[b]indolyl)₂ZrCl₂ (R = Ph, 10a; R = o-Tol, 10b; R = Me, 10c). The structure of 10a was determined by X-ray analysis. The effectiveness in polymerization catalysis of the metallocenes compared to benchmark metallocenes rac-Me₂Si(2-Me-4-Ph-Ind)₂ZrCl₂ (11) and rac-Me₂Si(2-Me-4,5-BzInd)₂ZrCl₂ (12) was investigated in liquid propylene at 70°C. Both 10a and 10b with MAO cocatalyst afforded highly isospecific (mmmm 95.9, 95.6%, respectively) and highly regisospecific polypropylene (0.17, 0.34% imbedded regioerrors), which exhibited a high melting point (T_m2 = 154.2, 156.3°C). The polymer properties were intermediate between those of 11 and 12, with the molecular weight, stereospecificity, and melting point increasing in the order 12 < 10a,b < 11. Although activities 1-2 orders of magnitude lower than for 11 and 12 were found at high MAO/Zr ratios, at low cocatalyst levels activities were similar. 10a,b were also successfully activated with alternative aluminoxanes based on branched alkylaluminums (e.g. Al(CH₂CHMePh)₃, obtained from α-methylstyrene), with measured activities up to 16.6 TON/((g of Zr) h) (10a/AKO). Surprisingly, no consistent effect on polymer properties of varying either the 4-aryl substituent of the cyclopenta[b]indole group or the cocatalyst was found.

Full text of DOI:10.1021/om020505t

Organometallics 2003, 22, 2711-2722
2711
a n sa -Zir con ocen es Ba sed on N-Su bstitu ted
2-Meth ylcyclop en ta [b]In d oles: Syn th esis a n d Ca ta lyst
Eva lu a tion in Liqu id P r op ylen e P olym er iza tion
J an F. van Baar,† Andrew D. Horton,*,† Kees P. de Kloe,† Eric Kragtwijk,†
Shaen G. Mkoyan,‡ Ilya E. Nifant’ev,*,§ Peter A. Schut,† and Ilya V. Taidakov§
Shell Research and Technology Centre, Amsterdam, Shell International Chemicals B.V.,
P.O. Box 38000, 1030 BN Amsterdam, The Netherlands, Department of Chemistry,
Moscow State University, Moscow 119899, Russia, and Institute of the Chemical Physics
Problems, 142432 Chernogolovka, Russia
Received J une 26, 2002
Cyclopenta[b]indole-based ansa-zirconocenes were prepared and studied as catalysts for
propylene polymerization. Substituted 2-methyl-4-R-1,4-dihydrocyclopenta[b]indoles (R )
Ph, 7a ; R ) o-Tol, 7b; R ) Me, 7c) were synthesized and converted into the racemic forms
of the corresponding ansa-zirconocenes Me2Si(2-Me-4-R-cyclopenta[b]indolyl)2ZrCl2 (R ) Ph,
10a ; R ) o-Tol, 10b; R ) Me, 10c). The structure of 10a was determined by X-ray analysis.
The effectiveness in polymerization catalysis of the metallocenes compared to benchmark
metallocenes rac-Me2Si(2-Me-4-Ph-Ind)2ZrCl2 (11) and rac-Me2Si(2-Me-4,5-BzInd)2ZrCl2 (12)
was investigated in liquid propylene at 70 °C. Both 10a and 10b with MAO cocatalyst
afforded highly isospecific (mmmm 95.9, 95.6%, respectively) and highly regisospecific
polypropylene (0.17, 0.34% imbedded regioerrors), which exhibited a high melting point (Tm2
) 154.2, 156.3 °C). The polymer properties were intermediate between those of 11 and 12,
with the molecular weight, stereospecificity, and melting point increasing in the order 12 <
10a ,b < 11. Although activities 1-2 orders of magnitude lower than for 11 and 12 were
found at high MAO/Zr ratios, at low cocatalyst levels activities were similar. 10a ,b were
also successfully activated with alternative aluminoxanes based on branched alkylaluminums
(e.g. Al(CH2CHMePh)3, obtained from R-methylstyrene), with measured activities up to 16.6
TON/((g of Zr) h) (10a /AKO). Surprisingly, no consistent effect on polymer properties of
varying either the 4-aryl substituent of the cyclopenta[b]indole group or the cocatalyst was
found.
1. In tr od u ction
C2-symmetric metallocenes (heterocenes) based on
cyclopentadienyls ring-fused to thiophenes and pyrroles
have been investigated in depth as procatalysts for
isotactic polypropylene (i-PP) formation.1-5 The cata-
lysts share structural features similar to those of the
best C2-symmetric metallocene catalysts for i-PP, such
as rac-Me2Si(2-Me-4-Ph-Ind)2ZrCl2 (11).6 The presence
of 2-methyl substituents (next to the silicon bridge) and
a phenyl group on the adjacent ring (protruding at the
* To whom correspondence should be addressed. E-mail: A.D.H.,
andrew.horton@shell.com; I.E.N., inif@org.chem.msu.ru.
F igu r e 1.
† Shell International Chemicals B. V., Shell Research and Technol-
ogy Centre, Amsterdam.
‡ Institute of the Chemical Physics Problems.
front of the complex) result in high polymer molecular
weight and high isotacticity. Zirconocenes 10a ,b (Figure
1), based on the 2-methyl-4-phenylcyclopenta[b]indolyl
fragment, are closely related to systems based on the
2-methyl-1-phenylcyclopenteno[1,2-b]pyrrole fragment
previously reported.1,2 These complexes differ in the
substitution pattern of the front-facing phenyl group,
with 10b expected on the basis of previous work with
metallocenes based on thiopentalenyl ligands3,4 to ex-
hibit higher polymer molecular weight and stereo-
regularity. We report here the preparation and perfor-
§ Moscow State University.
(1) Ewen, J . A.; Elder, M. J .; J ones, R. L., J r.; Dubitsky, Y. A. World
Patent Application WO 98/22486, 1998 (Montell).
(2) Ewen, J . A.; J ones, R. L.; Elder, M. J .; Rheingod, A. L.; Liable-
Sands, L. M. J . Am. Chem. Soc. 1998, 120, 10786.
(3) Ewen, J . A.; Elder, M. J .; J ones, R. L.; Rheingod, A. L.; Liable-
Sands, L. M.; Sommer, R. D. J . Am. Chem. Soc. 2001, 123, 4763, 6964.
(4) Ewen, J . A.; Elder, M. J .; J ones, R. L. World Patent Application
WO 2001044318, 2001 (Basell).
(5) Schottek, J .; Kratzer, R.; Winter, A.; Fraaije, V.; Brekener, M.-
J .; Oberhoff, M. World Patent Application WO 200044799, 2000
(Targor).
(6) Spaleck, W.; Ku¨ber, F.; Winter, A.; Rohrmann, J .; Bachmann,
B.; Antberg, B.; Dolle, V.; Paulus, E. F. Organometallics 1994, 13, 954.
10.1021/om020505t CCC: $25.00 © 2003 American Chemical Society
Publication on Web 05/29/2003

2712 Organometallics, Vol. 22, No. 13, 2003
van Baar et al.
F igu r e 2. Preparation of dihydrocyclopenta[b]indoles.
mance in isospecific propylene polymerization of this
new class of heterocene catalysts. The synthesis of
analogue 10c containing a 4-methyl group is also
described.
reaction14 with ethyl 2-bromopropionate (in the latter
case, the products yields are lower):
2. Syn th esis of Zir con ocen es
The synthesis of dihydrocyclopenta[b]indoles has not
been reported as yet. On the other hand, tetrahydro
derivatives of cyclopenta[b]indoles have been much
studied. These compounds are readily obtained by
cyclization of the corresponding cyclopentenone phenyl-
hydrazones according to Fischer.7,8 Attempts at using
this procedure to obtain dihydrocyclopenta[b]indoles
have failed.9 Therefore, we decided to assemble the
cyclopentadienone ring in a step-by-step manner, simi-
lar to that proposed for the synthesis of a number of
aza- and thiapentalenes.10 Figure 2 summarizes the
synthetic procedure applied for the formation of 2-meth-
yl-4-aryl-1,4-dihydrocyclopenta[b]indoles 7a ,b.
The initial compounds, N-substituted indoles 1, can
be obtained by alkylation or arylation of indole following
known procedures.11,12 Formylation of indoles under
conditions of the Vilsmeier reaction is known to proceed
with ease13 and results in the indole-3-carboxaldehydes
2 as exclusive products in high yields (80-95%).
There are two synthetic routes to unsaturated acids
4, namely, the Wadsworth-Emmons reaction with
triethyl 2-phosphonopropionate10 or the Reformatsky
Ethyl esters 3 were saponified under the action of a
methanol solution of KOH without isolation as indi-
vidual compounds. The overall yield calculated with
respect to free acid for the reaction with triethyl
2-phosphonopropionate was as high as 80%.
The facility of hydrogenation of 4, to give saturated
acids 5, depends on the substituent R, and the reaction
does not always proceed with ease. Because of this,
different methods have been used, such as reduction
of a solution of a sodium salt of the acid by hydrogen
under a slightly increased pressure (3 atm) on Pd
catalyst,10 electrochemical reduction at a mercury cath-
ode,15 hydrogenation under the action of sodium amal-
gam in alkali solution,16 and finally, reduction by
hydrazine hydrate in the presence of Raney nickel.17
The highest yields of unsaturated acids (up to 90%)
were obtained using electrochemical techniques, espe-
cially for large-scale charges. In small-scale experi-
ments, it is also convenient to use reduction by hydra-
zine hydrate.
(7) Perkin, S. G. P.; Plant, J . Chem. Soc. Perkin 1923, 123, 3242.
(8) Rodrigues, J . G.; Temprano, F.; Esteban-Calderon, C.; Martinez-
Ripoll, M.; Garcia-Blanco, S. Tetrahedron 1985, 41, 3813.
(9) Kelly, A. H.; McLeod, D. H.; Parrick, J . Can. J . Chem 1965, 43,
296.
(10) Ewen, J . A.; J ones, R. L.; Elder, M. J . In Metalorganic Catalysts
for Synthesis and Polymerization; Kaminsky, W., Ed.; Springer-
Verlag: Berlin, 1999, pp 150-169.
(11) Sundberg, R. S. Indoles; Academic Press: London, 1996.
(12) Khan, M. A.; Rocha, E. K. Chem. Pharm. Bull. 1979, 27,
529.
Cyclization of substituted propionic acids into ketones
6 can occur under the action of polyphosphoric acid
(PPA)18 or a solution of P2O5 in methanesulfonic acid
(14) Organic Reactions; Adams, R., Ed.; Vol 1, 11, New York, 1942;
Vol. 1, p 11.
(15) Marie, J . C.R. Hebd. Seances Acad. Sci. 1903, 136, 1331.
(16) Gattermann, L.; Wieland, H. Die Praxis des organischen
Chemikers; Walter de Cruyter: Berlin, 1947.
(13) Organic Synthesis; Wiley: New York, 1967; Collect Vol. 4, p
539.
(17) Balcom, D.; Furst, A. J . Am. Chem. Soc. 1953, 75, 4334.

ansa-Zirconocenes
Organometallics, Vol. 22, No. 13, 2003 2713
(the Eaton reagent).19 In the latter case, the yields of
ketones 6 are higher.
Reduction of ketone 6b followed by dehydration of the
alcohol formed results in kinetic product 7bI as the only
product; however, an equimolar mixture of 7bI and 7bII
can be obtained by hydrolysis of the lithium salt (8b).
1
The H NMR spectrum (CDCl3, 25 °C) of isomer 7bII
exhibits a characteristic signal for the AB system of
diastereotopic -CH2- protons at δ 3.15 and 3.07 (J )
19 Hz). This AB pattern remains unchanged in the
temperature range 20-100 °C (in toluene-d8), consistent
with the absence of free rotation of the tolyl substituent
1
We investigated an alternative method for the syn-
thesis of ketones 6c from substituted indole 1c, consist-
ing of heating a mixture of indole and methacrylic acid
in the presence of PPA or the Eaton reagent. This
about the C-N bond under these conditions. In the H
NMR spectrum of compound 7bI (CDCl3, 25 °C), the
signal of allylic protons is a singlet at δ 3.36, which can
be due to either a free rotation of the tolyl group or the
fact that the difference in the chemical shifts of dias-
tereotopic protons is so small that it cannot be reliably
measured.
By and large, methods for the synthesis of organo-
metallic derivatives of 2-methylcyclopenta[b]indoles do
not differ from those used to obtain, for example, indenyl
or cyclopentadienyl complexes. Reactions of compounds
7 with a solution of butyllithium in ether results in
crystalline lithium salts 8 (compound 8c was isolated
in pure form, yield - 71%):
procedure appears to be of a general character and can
also be used to obtain substituted indanones20 and their
heteroanalogues.10 Some examples of the use of this
reaction for the synthesis of cyclic thiophene ketones
have been reported.21
Only in the case of 1-methylindole (1c) was the yield
of the desired ketone sufficiently high. In the case of
1a ,b, strong resinification was observed.
Reduction of ketones into alcohols by LiAlH4 in ether
occurs readily. The reducing agent should not be used
in large excess, since the alcohol formed in the reaction
can undergo further transformation into a tetrahydro-
cyclopenta[b]indole derivative.22
Dehydration of alcohols occurs with ease in the
presence of acid catalysts. The use of sufficiently dilute
solutions is a necessary condition for obtaining of
compounds 7 in high yields (80-90%).
It should be mentioned that no free rotation of the
tolyl substituent around the plane of the indole frag-
ment occurs in the molecule of 2-methyl-4-(2-methyl-
phenyl)-1,4-dihydrocyclopenta[b]indol-3(2H)-one (6b).
This leads to the appearance of two rotamers in a 1/1
ratio, which in turn causes doubling of the number of
all signals of the cyclopentenone ring in the NMR
spectrum. An analogous effect can also be observed for
2-methyl-4-(2-methylphenyl)-1,4-dihydrocyclopenta[b]-
indole (7), which can have two isomers, 7bI and 7bII,
differing in the position of the double bond.
Lithium salts 8a -c interact with (CH3)2SiCl2 to give
1
high yields (conversion 100%, determined by H NMR)
of reaction products only when using hexane/THF or
toluene/THF mixtures as solvents. In each case, the
reaction results in a mixture of nearly equal amounts
of racemic and meso forms of compounds 9. Performing
this reaction in ether resulted in the formation of a
mixture of products; however, our attempts to identify
them failed. Ligands 9 were then transformed into
corresponding zirconocenes using standard methodol-
ogy.
(18) Acheson, R. M.; MacPhee, K. E.; Philpott, P. G.; Barltrop, A. J .
Chem. Soc. 1956, 698.
(19) Cho, H.; Matsuki, Sh. Heterocycles 1996, 43, 127.
(20) Sunaga, T. J apanese Patent Application J P 10059873, 1998;
Chem. Abstr. 1998; 128, P 217124.
(21) Meth-Cohn, O.; Gronowitz, S. Acta Chem. Scand. 1966, 20,
1577.
The pure racemic forms of compounds 10a (yield
17%), 10b (6%), and 10c (32%) as well as the meso form
of compound 10a (23%) were isolated by recrystalliza-
(22) Bergman, J .; Backvall, J . E. Tetrahedron 1975, 31, 2063.

2714 Organometallics, Vol. 22, No. 13, 2003
van Baar et al.
3. Ca ta lytic In vestiga tion
To allow comparison to state-of-the-art metallocene
catalysts, the propylene polymerization activity of the
new complexes and reference zirconocenes 11 (rac-Me2-
Si(2-Me-4-Ph-Ind)2ZrCl2)6 and 12 (rac-Me2Si(2-Me-4,5-
BzInd)2ZrCl2)23 has been examined under the same
conditions using methylaluminoxane (MAO) as cocata-
lyst. Polymerizations were performed in liquid propy-
lene (LIPP) at 70 °C in the presence of hydrogen
(generally ca. 1% in the gas cap at 70 °C), which resulted
in a lowering of the polymer molecular weight, facilitat-
ing polymer analysis (NMR, GPC) and affording poly-
mers of greater industrial relevance. Catalyst precursor
11, and to a lesser extent 10a ,b, gave very high
molecular weight polymer in the absence of hydrogen.
A cold start procedure was applied to prevent reactor
fouling (deposition of polymer on the autoclave walls,
stirrer, and other surfaces), which can lead to poor
control of process conditions (temperature, pressure).
Catalyst injection into the autoclave containing propy-
lene at 30 °C was followed by rapid warming (6-7 min)
to the polymerization temperature. At the end of the
polymerization (generally about 60 min) the catalyst
was deactivated by injection of methanol and the
polymer isolated as a free-flowing powder. The time
quoted in Tables 2 and 3 and used to calculate activity
refers to the period from initiation of heating to catalyst
quenching.
Procatalysts 10a ,b were also activated with alterna-
tive cocatalysts based on the hydrolysis products of
alkylaluminums, in which the alkyl groups are â,γ-
branched (AK, AO, Q in Figure 5). These alkylalumi-
noxanes (particularly AKO and AOO) have been shown
to be effective cocatalysts for a range of isospecific
metallocenes.24-27 The optimum activity was obtained
with a water/Al molar ratio of ca. 1/2. Alkylaluminox-
anes with a â-phenyl substituent were particularly
effective at high temperatures,26 often affording catalyst
activity superior to that of MAO at 70 °C in LIPP.
Com p a r ison to Ben ch m a r k Meta llocen es w ith
Meth yla lu m in oxa n e a s Coca ta lyst. The activities of
procatalysts 10a ,b with MAO as cocatalyst in LIPP are
compared with those for the isospecific benchmarks 11
and 12 in Table 2. All catalysts afforded highly isotactic
polypropylene, and no significant reactor fouling was
observed under cold start conditions with added hydro-
gen. The relative performance of the catalysts depended
strongly on the Al/Zr molar ratio. At high MAO intakes,
12 and 11 were 1 and 2 orders of magnitude (respec-
tively) more active than 10a ,b, with, for example, a
remarkably high activity of 170 Ton/((g of Zr) h) found
for 11 (experiment 6).6 However, with a low aluminox-
F igu r e 3. Molecular structure of [µ-dimethylsilanediylbis-
(η5-2-methyl-4-phenylcyclopenta[b]indolyl)]zirconium(IV)
dichloride (10a ).
Ta ble 1. Essen tia l Str u ctu r a l P a r a m eter s of 10a ,
Com p a r ed w ith Th ose of Rela ted Zir con ocen es
param
116
133
143
163
10a
Zr-Cl (Å)
2.419(1) 2.437(3) 2.421(5) 2.443(4) 2.4413a
Cl-Zr-Cl (deg) 96.8(1) 98.00(11) 99.44(3) 97.86(3) 99.90(4)
Zr-Cpb (Å)
2.243(4) 2.251(6) 2.256(5) 2.230(4) 2.151a
Zr-C(Cp) min 2.478(3) 2.484(10) 2.466(2) 2.463(3) 2.275(3)
(Å)
Zr-C(Cp) max 2.640(4) 2.663(9) 2.665(2) 2.650(3) 2.635(3)
(Å)
Cp-Zr-Cp (deg) 128.5(3) 128.4(6) 127.5
128.7(4) 126.51(1)
59.9 58.1
141.2(8) 139.1(9)
30.7(55) 55.9(67) 35.5(34) 56.1
E1-E2c
59.2
131.4
44.4
59.8
139.3(9) 139.4
61.3
θd
Φe
a
b
Mean value. Cp ring centroid. c Angle between least-squares
d
planes of the Cp rings. Exo angle between condensed 5/5- and
5/6-angles. e Torsion angle between the benzene and Cp rings
least-squares planes.
tion from appropriate solvents. Single crystals of 10a
suitable for an X-ray diffraction study were obtained by
storing a saturated solution in an ether-hexane mix-
ture at 20 °C.
The structure of complex 10a is shown in Figure 3.
Key structural parameters for 10a are compared to
those for related crowded metallocene 11 and het-
erocenes 13, 14, and 16 (Figure 4) in Table 1. In the
molecule 10a , the Zr atom is coordinated by the cyclo-
pentadienyl moieties in an η5 fashion, the molecule
having typical distortions characteristic of ansa-metal-
locene complexes with short bridges. There is no contact
between the Zr and N atoms, which is also a general
characteristic of heteroaromatic zirconocenes.
(23) Stehling, U.; Disbold, J .; Kirsten, R.; Roll, W.; Brintzinger, H.
H. Organometallics 1994, 13, 964.
It is worth noting that 10a has the lowest Zr-Cp
(centroid) distance of the compounds considered and
exhibits a large Φ angle between the phenyl ring and
the Cp ring planes (similar to that in heterocene 14).
Other workers have suggested that the size of the
dihedral angle involving this front-side phenyl group is
highly sensitive to crystal-packing forces or the presence
of cocrystallized solvent molecules.3 Other structural
characteristics are practically the same and are typical
of these complexes.
(24) Horton, A. D.; van Baar, J . F.; Schut, P. A.; van Kessel, G. M.
M.; von Hebel, K. L. World Patent Appl. WO 9921899, 1999 (Montell).
(25) van Baar, J . F.; Galimberti, M.; von Hebel, K. L.; Horton, A.
D.; van Kessel, G. M. M.; Schut, P. A.; Dall’Occo, T. World Patent Appl.
9921896, 1999 (Montell).
(26) van Baar, J . F.; Horton, A. D.; van Kessel, G. M. M.; Kragtwijk,
E.; van de Pas, M. P.; Schut, P. A.; Stapersma, J World Patent Appl.
WO 2001021674, 2001 (Basell).
(27) van Baar, J . F.; Schut, P. A.; van Kessel, G. M. M.; von Hebel,
K. L.; Kragtwijk, E.; van de Pas, M. P.; Horton, A. D. Abstracts of
Papers, 221st National Meeting of the American Chemical Society, San
Diego, CA, Apr 1-5, 2001; American Chemical Society: Washington,
DC, 2001; INOR-64.

ansa-Zirconocenes
Organometallics, Vol. 22, No. 13, 2003 2715
F igu r e 4.
Ta ble 2. LIP P Activity of 10a ,b Com p a r ed to Ben ch m a r k Meta llocen es w ith Meth yla lu m in oxa n e
Coca ta lyst (Cold Sta r t P olym er iza tion , 70 °C)a
prealkylation
procatalyst
agent
aluminoxane
(reactor)
amt,
amt,
Al/Zr ratio
amt of TIBA, H2 gas cap, time, yield,
activity,
expt type µmol
type
mmol type amt, mmol (aluminoxane)
mmol
vol %
min
g
Ton/((g of Zr) h)
1
10a
1.44 MAO
0.75
MAO
MAO
4.5
4.5
3646
0.0
1.0
49
195
1.8
2
3
10b
10b
0.48 MAO
1.00 MAO
0.75
0.50
10938
500
0.0
1.0
1.1
1.0
60
60
260
320
5.9
3.5
4
5
12
12
0.05 MAO
0.75 MAO
0.75
0.375
MAO
MAO
4.5
4.5
105000
500
0.0
1.0
1.0
1.0
60
60
140
129
30.7
1.9
6
7
11
11
0.05 MAO
0.75 MAO
0.75
0.375
105000
500
0.0
1.0
1.0
0.8
34
60
438
275
169.5
4.0
a
Short cold start procedure, except long cold start for experiments 1 and 2. The time was measured from catalyst injection (short cold)
or injection + 10 min (long cold start). Negligible reactor fouling was found.
Ta ble 3. LIP P Activity of Heter ocen es w ith Va r iou s Alk yla lu m in oxa n e Coca ta lysts a t High a n d Low Al/Zr
Ra tios (Cold Sta r t P olym er iza tion , 70 °C)a
prealkylation
procatalyst
amt,
agent
aluminoxane
amt,
amt,
Al/Zr ratio
amt of TIBA, H2 gas cap,
activity,
expt
type
µmol
type
mmol
type
mmol
(aluminoxane)
mmol
vol %
time, min
Ton/((g of Zr) h)
8
9
10
11
10a
10a
10a
10a
0.40
0.48
0.24
0.24
MAO
TIOA
TIOA
TIOA
0.75
0.30
0.15
0.15
MAO
AKO
AOO
QO
4.5
4.5
2.3
2.2
13125
9375
9375
9167
0.0
0.0
0.0
0.0
1.0
1.0
1.0
0.9
60
43
52
50
1.8
16.6
6.3
7.4
3
4
10b
10b
0.48
1.00
MAO
MAO
0.75
0.50
MAO
4.5
10938
500
0.0
1.0
1.1
1.0b
60
60
5.9
3.5
12
13
10b
10b
0.48
1.00
TIOA
TIOA
0.30
0.10
AKO
AKO
4.5
0.5
9375
500
0.0
1.0
1.3b
1.0
60
60
9.1
0.2
14
15
10b
10b
0.48
1.00
TIOA
TIOA
0.30
0.10
AOO
AOO
4.5
0.5
9375
500
0.0
1.0
1.8
0.7
60
60
5.3
3.4
16
17
10b
10b
0.48
1.00
TIOA
TIOA
0.30
0.10
QO
QO
4.5
0.5
9375
500
0.0
1.0
1.5b
0.8
60
60
3.1
1.4
a
Long cold start, except short cold start for experiments 4 and 13. The time was measured from catalyst injection (short cold) or
b
injection + 10 min (long cold start). Negligible reactor fouling was found. Average H2 concentration (varied in range 0.5% above or
below).
ane/Zr molar ratio of about 500 in the premix (and
utilizing 1.0 mmol of TIBA as scavenger in the auto-
clave),28 the observed activities with 10b, 11, and 12
were similar, in the range 2-4 Ton/((g of Zr) h). Some
caution should also be exercised in comparing activities
at different procatalyst intake levels, as activity, in
(28) Experimental studies (with other metallocenes) of the effect of
scavenger intake on activity at low aluminoxane/Zr ratios indicated
that the level of 1.0 mmol of TIBA applied in this work is well above
that needed to effectively scavenge poisons, such as water, oxygen, and
other reactive contaminants in the 5 L reactor containing 1.6 kg of
propylene.
F igu r e 5. Trialkylaluminum precursors of branched alkyl-
aluminoxanes.

2716 Organometallics, Vol. 22, No. 13, 2003
van Baar et al.
Ta ble 4. LIP P Activity of 10a w ith Va r iou s Alk yla lu m in oxa n e Coca ta lysts a t High Al/Zr Ra tios u n d er Hot
Sta r t Con d ition s (70 °C)a
procatalyst
amt,
prealkylation
amt,
aluminoxane
amt,
Al/Zr ratio
(aluminoxane)
H2 gas cap,
vol %
time,
min
activity,
Ton/((g of Zr) h)
expt
type
µmol
type
mmol
type
mmol
fouling, %
18
19
10a
10a
1.44
1.44
MAO
MAO
0.75
0.75
MAO
MAO
4.5
4.5
3646
3646
0.01b
1.0
60
60
1.4
0.9
77
82
20
21
10a
10a
1.44
1.44
TIOA
TIOA
0.30
0.30
AKO
AKO
4.5
4.5
3125
3125
0.02b
0.7
60
60
1.6
2.4
87
92
22
23
10a
10a
1.44
1.44
TIOA
TIOA
0.30
0.30
AOO
AOO
4.5
4.5
3125
3125
0.02b
0.9
38
42
2.0
2.7
80
91
24
25
10a
10a
1.44
1.44
TIOA
TIOA
0.30
0.30
QO
QO
4.5
4.5
3125
3125
0.02b
0.9
27
17
2.5
7.1
96
94
a
b
Time measured from catalyst injection. No H2 added: maximum H2 concentration measured.
particular at low Al/Zr ratios, is strongly dependent on
the absolute concentration of the zirconocene. For
example, the activity of 11 at 70 °C (1 v % hydrogen,
MAO/Zr ca. 500, 1.0 mmol TIBA) at 0.25, 0.50 and 0.75
µmol Zr intake increased in the order, 0.4, 1.0, 4.0 Ton/
((g of Zr) h) (last entry, experiment 7).29
Br a n ch ed Alk yla lu m in oxa n e Coca ta lysts. The
influence of the cocatalyst on polymerization with 10a ,b
was investigated at both high and low aluminoxane/Zr
ratios. Unlike MAO, effective activation of metallocene
dihalides with branched alkylaluminoxanes requires
pretreatment with an alkylaluminum species such as
could be discerned. Similarly, polymerizations with 10a
and branched alkylaluminoxanes exhibited low catalyst
decay. In contrast, 10a /MAO (experiment 1) showed
significant decay over the polymerization period, partly
explaining the low activity found for this system.
Results obtained with 10b at very low aluminoxane/
Zr levels (aluminoxane/Zr ) 500) using TIBA (1.0 mmol)
as autoclave scavenger are detailed in Table 3; experi-
ment 4 with MAO cocatalyst is included for comparison.
In all cases activity was lower at low Al/Zr ratios, but
the effect was most dramatic for cocatalyst AKO (45-
fold decrease in activity). In general, kinetic profiles for
10b at low aluminoxane/Zr levels showed catalyst decay
during the initial period of polymerization (possibly
reflecting slow reaction with sub-equimolar amounts of
contaminants), followed by constant activity for the final
30 min (of a 60 min polymerization). Under these
conditions MAO was either equally effective as (10b/
AOO) or superior to the branched alkylaluminoxane
cocatalysts. The relative cocatalyst effectiveness differs
from that at high aluminoxane/10b ratios, where activ-
ity increased in the order QO < MAO, AOO < AKO.
Application of a “hot start” procedure, omitting the
low-temperature (pre-)polymerization step (injection
into the autoclave at 70 °C), resulted in severe reactor
fouling and generally lower activities, as shown by a
series of experiments with 10a at high aluminoxane/Zr
levels (Table 4). Most of the polymer product was
deposited on the walls, stirrer, and other surfaces.
Addition of hydrogen (0.7-1.0 v %) did not lead to
reduced reactor fouling, although it did afford higher
activity in three of four cases. The effect of hydrogen
on propylene polymerization with 10a is illustrated in
the cases of MAO and AKO by the kinetic data in Figure
6: despite the fouling, experiments performed under
hydrogen exhibited no-decay kinetics over a period of 1
h. Again, changing the cocatalyst had a clear effect on
activity (both without and with hydrogen), increasing
in the order MAO < AKO < AOO < QO, compared to
the order MAO < AOO, QO < AKO found under cold
start conditions at high Al/Zr ratio.
“tris(isooctyl)aluminum” (TIOA, actual constitution H0.15
-
Al(CH2CHMeCH2CMe3)2.85), as these aluminoxanes are
ineffective in prealkylation.25 With high aluminoxane
levels, the (TIOA or MAO) pretreated procatalyst was
introduced into the autoclave containing liquid propy-
lene and aluminoxane. With low aluminoxane levels, the
metallocene was successively treated with TIOA (100
equiv) and aluminoxane (500-750 equiv) before being
introduced into the reactor containing propylene and 1.0
mmol of TIBA.28 In experiments with MAO at low levels
(250-500 equiv), MAO was used as the sole (pre)-
activation agent.
In the experiments at high aluminoxane/Zr ratio
under H2 (ca. 1 vol %) in Table 3, the aluminoxane
intake was generally fixed at 4.5-5.25 mmol and the
zirconium intake varied (depending on the intrinsic
catalyst activity). Branched alkylaluminoxanes acti-
vated both 10a and 10b for isotactic polypropylene
formation, with activities in the range 3.1 Ton/((g of Zr)
h) (10b/QO, experiment 16) to 16.6 Ton/((g of Zr) h) (10a /
AKO, experiment 9), and no reactor fouling.
The kinetic profile of the polymerizations in LIPP
could be determined by continuous GC monitoring of
the ratio of propane (typically 0.3 mol % in propylene
feedstock) to propylene in the gas cap of the autoclave.
Typically, little catalyst decay was observed over the
polymerization period. For example, polymerizations
using MAO, AOO, and QO cocatalysts with 10b (long
cold start procedure) exhibited constant activity for the
last 50 min of a 60 min polymerization (constant
temperature, after warmup period), whereas with AKO
as cocatalyst, a gradual increase in activity with time
Polymerizations at 70 °C without added hydrogen
were associated with hydrogen formation in the reactor,
building up over 1 h to a constant level of 0.01-0.02
vol % (gas cap). Hydrogen production was presumably
associated with the formation of internal vinylidene
functionalities (P-CHMeCH2C(dCH2)CH2CHMe-P′),30,31
(29) The influence of Al/Zr ratio and Zr intake on activity is further
illustrated by the following activity data for 11 in LIPP at 70 °C (1 vol
% hydrogen, 2.1 mmol of TIBA scavenger): Al/Zr 2000, 0.25 µmol Zr,
2.5 Ton/((g of Zr) h); Al/Zr 990, 0.25 µmol Zr, 0.6 Ton/((g of Zr) h); Al/
Zr 500, 0.75 µmol Zr, 2.6 Ton/((g of Zr) h); Al/Zr 250, 0.75 µmol Zr, 0.3
Ton/((g of Zr) h); Al/Zr 100, 0.75 µmol Zr, 0.0 Ton/((g of Zr) h).
(30) Horton, A. D.; Schut, P. A.; van Baar, J . F. World Patent
Application WO 2001090205, 2001 (Basell).

ansa-Zirconocenes
Organometallics, Vol. 22, No. 13, 2003 2717
F igu r e 6. Influence of cocatalyst and hydrogen on 10a catalyst under hot start conditions (no added H2 or 0.7-1.0 vol %
H2; injection t ) 10 min; experiments 18 and 19 (MAO) and 20 and 21 (AKO) in Table 4).
Ta ble 5. Coca ta lyst Effects on P r op er ties of Isota ctic P olyp r op ylen e Obta in ed w ith Heter ocen e Ca ta lysts
aluminoxane
procat-
alyst
type
H2 gas
cap,
regioerrors
Al/Zr
ratio
activity,
2,1 and 1,3, mmmm, sum,a Tm2
,
∆H, XS,
J /g
expt
type
vol % Ton/((g of Zr) h) 10-3Mw 10-3Mn Mw/Mn
%
%
%
°C
%
1
9
10a
10a
MAO
3646
1.0
1.8
69.3
34.0
2.1
0.17
95.9
3.0
154.4
99.8 n.d.
154.6 100.4 n.d.
151.9 91.4 n.d.
AKO
AKO
9375
3125
1.0
0.7
16.6
2.4
81.0
64.4
42.1
26.9
1.9
2.4
0.23
0.32
95.8
92.2
2.6
4.4
21 10a
10 10a
23 10a
AOO
AOO
9375
3125
1.0
0.9
6.3
2.7
49.9
51.3
12.7
9.9
3.9
5.2
0.18
0.19
94.0
93.4
3.2
3.3
154.6 102.3 n.d.
152.2 100.6 n.d.
11 10a
25 10a
QO
QO
9167
3125
0.9
0.9
7.4
7.1
112.0
75.7
42.9
41.7
2.6
1.8
0.19
0.23
96.5
96.2
2.1
2.2
156.8 101.4 n.d.
158.0 108.1 n.d.
2
3
10b
10b
MAO
MAO
10938
500
1.1
1.0b
5.9
3.5
55.5
37.8
12.6
9.3
4.4
4.1
0.34
0.24
95.6
93.0
2.3
3.9
156.3 108.7 1.4
154.6 115.6 2.8
12 10b
13 10b
AKO
AKO
9375
500
1.3b
1.0
9.1
0.2
29.3
25.7
8.6
8.4
3.4
3.0
0.29
0.21
92.8
95.0
3.8
2.8
151.0 109.0 3.6
156.8 127.2 4.0
14 10b
15 10b
AOO
AOO
9375
500
1.8
0.7
5.3
3.4
46.1
35.2
12.2
12.6
3.8
2.8
0.26
0.24
93.8
93.8
3.1
3.0
154.0 104.8 3.6
155.3 112.6 2.6
16 10b
17 10b
QO
QO
9375
500
1.5b
0.8
3.1
1.4
40.8
35.6
10.9
9.1
3.7
3.9
0.29
0.26
94.8
94.4
2.8
3.4
153.3 109.9 2.0
154.9 114.6 3.6
4
5
12
12
MAO 105000
MAO 500
1.0
1.0
30.7
1.9
21.7
27.5
11.3
14.4
1.9
1.9
0.30
0.32
91.2
92.7
6.6
5.5
145.4 103.6 1.38
147.3
96.6 0.84
6
7
11
11
MAO 105000
MAO 500
1.0
0.8
169.5
4.0
115.0
88.6
57.2
39.4
2.0
2.2
0.37
0.31
97.7
88.9c
1.4
5.3
158.5 104.3 n.d.
160.1 102.2 3.5
a
b
“sum” refers to sum of mmmr, mmrr, and mrrm pentads. Average H2 concentration (varied in range 0.5% above or below). c Low
mmmm value and high Tm2 value reflect the presence of some atactic polymer (XS 3.5%); the bulk of the polymer is highly isotactic.
but this could not be confirmed by NMR due to the low
solubility and high molecular weight of the polymer
samples.
and melting points (DSC) were expected to be less
sensitive to variation in the hydrogen levels.
Com p a r ison to Ben ch m a r k Meta llocen es w ith
MAO Coca ta lyst. In general the properties of i-PP
obtained with 10a ,b were found to be intermediate
between those of benchmark systems 11 and 12 (Table
5). The i-PP molecular weight (MW) decreased in the
following order: 11 > 10a ,b > 12. Polymer obtained
with 10b generally exhibited a broader MW distribution
than that obtained with 10a . The steroregularity (re-
flected in mmmm and, partially, in Tm2) decreased in
the same order as the MW: 11 > 10a ,b > 12. The
relatively low mmmm value of 88.9% (experiment 7),
for polymer prepared with 11 at a low MAO/Zr ratio,
reflects the presence of some polymer of low stereoregu-
larity (XS ) 3.5%), the major (highly isotactic) compo-
Effect of Ca ta lyst a n d Coca ta lyst on P olym er
P r op er ties. The properties of isotactic polypropylenes
obtained with 10a ,b and benchmark systems 11 and 12
are compared in Table 5 to those prepared with bench-
mark metallocenes. Elucidation of the effect of catalyst
and cocatalyst on polymer properties is somewhat
hindered by differences in the hydrogen levels between
experiments. In particular, comparison of polymer mo-
lecular weight (from GPC) and xylene solubles (XS)
levels must be done with caution. Regio- and stereo-
regularity, determined using 13C NMR spectroscopy,
(31) Resconi, L.; Camurati, I.; Sudmeijer, O. Top. Catal. 1999, 7,
145.

2718 Organometallics, Vol. 22, No. 13, 2003
van Baar et al.
nent exhibiting a high Tm2 value of 160.1 °C. The levels
of imbedded regioerrors (primarily erythro-2,1-inser-
tions) are rather similar in all four systems, in the range
0.2-0.4%. No consistent effect of the Al/Zr ratio on
polymer properties could be found.
Catalysts 10a,b: In flu en ce of Cocatalyst on P oly-
m er P r op er ties. From the data in Table 5 on polymers
obtained with heterocenes 10a ,b, no general trends
concerning the effect of catalyst and cocatalyst on the
properties of i-PP could be ascertained. i-PP prepared
using 10b exhibited somewhat lower molecular weight
(Mw ) 29 300-55 500; GPC data) than that prepared
at similar H2 levels with 10a (Mw ) 49 900-112 000).
In contrast to previous results with other metallocenes
(e.g. lower polymer MW with cocatalyst AOO),27,30 no
consistent cocatalyst effect on MW could be discerned.
solid state is already large (similar to that in 14; see
above) and further substitution of the phenyl ring may
not result in further increase in this angle. The stereo-
selectivity of 10a is slightly lower and the regioselec-
tivity similar to that of 13, which has a similar substi-
tution pattern.
Comparison of polymer prepared with 10a with that
reported for the very closely related heterocene 172,3 (see
above) is also informative.33,34 Crudely, the annelated
benzene ring in 10a was expected to have a roughly
similar effect as the adjacent methyl group in 17,
increasing the angle between the aryl group and the
cyclopentadienyl ring plane. The measured polymer
melting point (155 °C) is within the range found for
polymers prepared with 10a . The level of regioerrors is
about twice as high for 10a as for 17 (erythro-2,1 ca.
0.2% and 0.1%, respectively), whereas the stereoselec-
tivity is much higher (mrrrm ca. 0.5% and 1.4%,
respectively). These two effects appear to work in
opposition, leading to similar Tm2 values. The higher
stereoselectivity and lower regiospecificity of 10a rela-
tive to 17 may reflect the more crowded nature of the
ligand in 10a , perhaps arising from interaction of the
annelated benzene ring with the bridge.
The high polymerization activities found with branched
alkylaluminoxanes and 10a ,b, particularly at high Al/
Zr ratios, highlight the promising nature of this family
of alternative cocatalysts. For the reference system, rac-
C2H4(Ind)2ZrCl2, for which a very wide range of alky-
laluminoxanes has been investigated,24-27 those based
on alkylaluminums in which the alkyl contains a
â-phenyl group (e.g. Al(CH2CHMePh)3) were previously
found to be the best performing cocatalysts in LIPP at
70 °C. Introduction of a 4-halo substituent in the phenyl
group (in AO) was particularly favorable.24-27 The
observation that QO and AKO with 10a ,b were more
effective cocatalysts (under some conditions) than the
other aluminoxanes is rather surprising.30 However,
similar positive results with QO as cocatalyst were
found with the thiopentalenes 13-15.4 This may, in
some way, result from the presence of heteroatoms in
the catalyst, although the exact details are unclear. The
limited cocatalyst influence on polymer properties is also
remarkable, in the light of previous studies with met-
allocenes, where cocatalyst AOO resulted in a dramatic
reduction in polymer MW relative to other branched
alkylaluminoxanes and MAO.26,27,34
Complexes 10a ,b afforded highly isotactic (mmmm
92.2-96.9%) and high-melting polypropylene (Tm2
)
150.4-159.4 °C). The major stereoerrors are formed as
a result of a single misinsertion of propylene, giving
mrrm pentads (0.5-0.8%); as expected, the sum of the
mmmr, mmrr, and mrrm pentads is approximately
equivalent to 5 times the mrrm pentad (Table 5).32 The
polymers contained low levels of imbedded regioerrors
in the range 0.16-0.23% for 10a and 0.21-0.34% for
10b. The regioerrors were again primarily erythro-2,1,
although trace levels of 1,3-propylene insertion were
found in four cases. The stereoregularity and melting
points showed a reasonable correlation, as expected,
given the small variation in regioregularity between
samples. The increased XS levels found with 10b may
partly reflect the rather low molecular weight of these
samples.
Discu ssion of P olym er iza tion Resu lts. It is clear
that the polymers obtained with heterocenes 10a ,b
exhibit properties (stereoselectivity, melting point, MW)
which are intermediate between those of the analogous
dimethysilyl-bridged ansa-metallocenes containing
2-methyl-4,5-benzindenyl and 2-methyl-4-phenyl-inde-
nyl groups. The polymerization activity of 10a ,b at high
aluminoxane/Zr levels is much lower than for the
benchmark metallocenes but is similar at low alumi-
noxane/Zr levels.
In the related thiopentalene series (Tp, 13-15;3,4 see
above) the substitution pattern of the phenyl group had
a significant effect on polymer properties: at 70 °C
(MAO, no H2) the molecular weight (Mw ) 445 000-
795 000), stereoselectivity (m 99.2-99.5%) and melting
point (Tm2 ) 156-160 °C) were all found to increase in
the series Ph < 2-Me-Ph < 2,5-Me2-Ph. The effect on
polymer properties was ascribed to an increase in the
torsion angle between the phenyl group and the cyclo-
pentadienyl ring plane as phenyl ring substitution
increased, resulting in higher effective bulk of the
protruding phenyl group. It is not clear why the same
trend in terms of polymer properties is not observed on
going from 10a to 10b; however, such subtle changes
may be masked by small changes in reaction conditions
(e.g. H2 level, start procedure). It should be noted that
the phenyl-Cp ring plane torsion angle in 10a in the
4. Exp er im en ta l Section
All operations with organometallic compounds were carried
out using Schlenk-type glassware or in a glovebox. Ether and
(33) Unfortunately only the rac/meso (1/1) mixture of 17 was tested
(MAO, 70 °C, no H2),4 leading to a mixture of isotactic and atactic
polymer (XS 7.0%). Data in the text refer to the isotactic polymer
fraction.
(34) The two systems differ in the relative activity of the rac and
meso isomers. Whereas the 1:1 rac/meso mixture of 17 afforded 7% of
“atactic-PP” with MAO, the 1:1 rac/meso mixture of 10a afforded about
30% of (presumably) atactic material (70 °C, no H2, MAO). With
cocatalyst AOO and added hydrogen, the meso isomer of 10a was more
active than the rac isomer. The following activities (Ton/((g of Zr) h))
were measured for the 1/1 rac/meso mixture of 10a at 70 °C (hot start,
0.8-1.0 vol % H2, aluminoxane/Zr ratio in parentheses): 0.58 (MAO,
3133), 0.71 (AKO, 3136), 7.10 (AOO, 18 805), 3.47 (QO, 2808). Analysis
of the total polymer samples afforded the following property informa-
tion for MAO, AOO experiments, respectively: Mw, 38 700, 16 500; Mn,
9140, 1170; 2,1-insertions, 0.19, 0.10%; mmmm, 71.4, 36.2%; Tm2, 153.7,
144.2 °C.
(32) A single (stereo) misinsertion of propylene affords polymer with
the ...mmmrrmmm... pentad structure. If this is the sole stereoerror
in the polypropylene, the sum of the mmmr, mmrr, and mrrm pentads
will be equal to 5 times the mrrm pentad.

ansa-Zirconocenes
Organometallics, Vol. 22, No. 13, 2003 2719
THF were distilled over benzophenone sodium ketyl; hexane,
The mixture was allowed to stay at room temperature until
hydrogen evolution ceased (for ca. 40 min). Then the temper-
ature was reduced to -15-20 °C and a solution of indole-3-
carboxaldehyde (2; 0.18 mol) in THF (120 mL) was added
dropwise. The reaction mixture was warmed to room temper-
ature, 200 mL of saturated NH4Cl solution was added, and
the aqueous phase was extracted with ether. The combined
organic extracts were dried over MgSO4 and evaporated.
Eth yl (E,Z)-2-Meth yl-3-(1-p h en yl-1H-in d ol-3-yl)-2-p r o-
p en oa te (3a ). Light yellow fine crystals (from aqueous etha-
1
toluene, and benzene were distilled over metallic Na. H and
13C NMR spectra were recorded on a Varian VXR-400 spec-
trometer.
1H and 13C NMR spectra of i-PP were recorded using a
Bruker DRX 500 spectrometer at 130 °C, under proton noise
decoupling in FT mode, with 10 kHz spectral width, 60° pulse
angle, and 15 s pulse repetition (2800 scans). A 5 mm probe
was generally used due to the better resolution obtained. The
sample concentration was 10 wt % in 1,2-dichlorobenzene
solution. Pentad distributions and the proportion of 2,1- and
1,3-regioerrors were determined as described previously.35
Alkylaluminums TIBA (Aldrich, Akzo-Nobel, Witco) and
TIOA (Witco) were obtained commercially and used without
further purification. Methylaluminoxane (MAO) was pur-
chased from Akzo or Witco as solutions in toluene. Other
alkylaluminums were prepared from the reaction of TIBA or
DIBAH with an excess amount of the appropriate olefin, as
described previously.24-27 The alkylaluminoxanes were pre-
pared by the controlled hydrolysis of trialkylaluminums in
toluene solvent (temperature range 4-10 °C) using a water/
Al molar ratio of 0.5. Toluene (Aldrich, anhydrous grade) was
stripped with nitrogen overnight and dried over 4 Å molecular
sieves (H2O level ∼5-8 ppm). Nitrogen and propylene were
purified over a BTS column containing a copper catalyst (BASF
R3-11) as well as a column containing a mixed bed of molecular
sieves (3 Å (bottom), 4 Å (middle), and 13 Å (top)) to remove
oxygen and water, respectively (O2 level ∼0.3 ppm, H2O level
0.5 ppm).
1
nol). Mp: 96-97 °C. Yield: 50.2 g (91%). H NMR (δ, CDCl3,
25 °C): 8.15 (s, 1H, CH), 7.95 (m, 1H, CH arom), 7.66 (s, 1H,
CH indole), 7.6 (m, 5H, CH arom), 7.45 (m, 1H, CH arom),
7.35 (m, 2H, CH arom), 4.4 (q, J ) 5.9 Hz, 2H, -CH2), 2.3 (s,
3H, -CH3), 1.45 (t, J ) 5.9 Hz, 3H, -CH3). 13C NMR (δ, CDCl3,
25 °C): 168.7 (-CdO); 138.9, 135.6, 128.8, 123.6, 113.7 (>C<);
129.6, 129.21, 128.7, 127.1, 124.4, 123.3, 121.1, 119.1 110.5
(dCH-); 60.4 (-CH2-); 15.1 (-CH3); 14.3 (-CH3). Anal.
Found: C, 78.57; H, 6.31; N, 4.52. Calcd (C20H19NO2): C, 78.66;
H, 6.27; N, 4.59.
Eth yl (E,Z)-2-Meth yl-3-[1-(2-m eth ylp h en yl)-1H-in d ol-
3-yl]-2-p r op en oa te (3b). Dark yellow viscous oil. Yield: 51.2
g (89%). The specimen used for analytical measurements was
purified by chromatography on a column with silica gel (with
1
a benzene/ethyl acetate (4/1) mixture as eluent). H NMR (δ,
CDCl3, 25 °C): 8.2 (s, 1H, CH), 7.9 (m, 1H, CH arom), 7.55 (s,
1H, CH indole), 7.46 (m, 2H, CH arom), 7.38 (m, 2H, CH arom),
7.31 (m, 2H, CH arom), 7.08 (m, 1H, CH arom), 4.35 (q, J )
5.8 Hz, 2H, -CH2-), 2.25 (s, 3H, -CH3), 2.1 (s, 3H, -CH3),
1.4 (t, J ) 5.8 Hz, 3H, -CH3). 13C NMR (δ, CDCl3, 25 °C):
168.8 (-CdO); 137.3, 136.7, 135.5, 129.6, 127.9, 113.0 (>C<);
131.2, 129.7, 129.6, 128.8, 127.9, 126.9, 123.1, 120.9, 118.8,
110.6 (dCH-); 60.4(-CH2-); 17.4 (-CH3); 15.1 (-CH3); 14.3
1-Phenylindole (1c) and 1-(o-tolyl)indole (1b) were synthe-
sized by following the known procedure.12
Gen er a l P r oced u r e for F or m yla tion of In d oles.13 In a
100 mL round-bottom flask equipped with a reflux condenser
and a dropping funnel, indole (0.1 mol) was dissolved in DMF
(45 mL, 0.6 mol) with stirring. The mixture was cooled to 0
°C, and POCl3 (9 mL, 0.1 mol) was added dropwise. On
completion of the vigorous stage, the reaction mixture was
heated for 2 h at 80 °C using a water bath. The resulting red
solution was poured into 200 mL of a warm saturated solution
of Na2CO3 with stirring. The aldehyle was separated, washed
with water, and dried in vacuo.
(-CH3). Anal. Found: C, 78.82; H, 6.52; N, 4.41. Calcd (C21H21
NO2): C, 78.97; H, 6.63; N, 4.39.
-
Hyd r olysis a n d Red u ction . The nonpurified ester (0.16
mol) obtained in the previous stage was suspended in methanol
(350 mL), and a solution of KOH (23 g, 0.4 mol) in water (50
mL) was added. Refluxing the mixture for 1 h afforded a nearly
transparent solution that solidified on cooling to give a compact
aggregate of the potassium salt. The salt was dissolved in
water (200 mL), alcohol was removed by distilled in vacuo,
and the moist salt was used in the reduction.
1-P h en yl-1H-in d ole-3-ca r boxa ld eh yd e (2a ). Colorless
crystals. Mp: 45-46 °C. Yield: 20.6 g (93%). 1H NMR (δ,
CDCl3, 25 °C): 10.1 (s, 1H, CHO), 8.4 (m, 1H, CH arom), 7.9
(s, 1H, CH indole), 7.6 (m, 2H, CH arom), 7.55 (m, 4H, CH
arom), 7.4 (m, 2H, CH arom). 13C NMR (δ, CDCl3, 25 °C): 184.7
(-CHO); 137.99, 137.3, 125.3 (>C<); 137.93, 128.8, 128.1,
124.6, 124.4, 123.2, 122.1, 110.9 (dCH-). Anal. Found: C,
81.25; H, 5.20; N, 6.12. Calcd (C15H11NO): C, 81.43; H, 5.01;
N, 6.33.
1-(2-Meth ylp h en yl)-1H-in d ole-3-ca r boxa ld eh yd e (2b).
Viscous oil. Yield: 20.7 g (88%). 1H NMR (δ, CDCl3, 25 °C):
10.1 (s, 1H, CHO), 8.4 (d, J ) 8.1 Hz, 1H, CH arom), 7.8 (s,
1H, CH indole), 7.6-7.4 (m, 6H, arom), 7.1 (d, J ) 8.1 Hz, 1H,
CH arom), 2.1 (s, 3H, -CH3). 13C NMR (δ, CDCl3, 25 °C): 184.7
(-CHO); 138.1, 136.4, 135.3, 124.6, 119.0 (>C<); 138.6; 131.3,
129.3, 127.6, 127.0, 124.3, 122.9, 121.9, 110.94 (dCH-); 17.2
(a ) Red u ction by Hyd r a zin e (Meth od A). The potasium
salt of the appropriate 2-methyl-3-(1-aryl-1H-3-indolyl)-2-
propenic acid (0.16 mol) was dissolved in water (800 mL), and
a suspension of Raney nickel (2.5 g) was added. The solution
was heated to 90 °C with continuous stirring with a magnetic
stirrer. To the hot solution was added portionwise a 85%
hydrazine hydrate (2-3 mL). Complete reduction requires 60-
70 mL of hydrazine hydrate and takes 7-8 h. After completion
of the reaction, the catalyst was filtered off, 10% HCl was
added to pH 2, and the acid that precipitated was separated.
The residue was washed with water and dried in vacuo.
(b) Electr och em ica l Red u ction . (Meth od B). A suspen-
sion of the potassium salt of the appropriate 2-methyl-3-(1-
aryl-1H-3-indolyl)-2-propenic acid (0.16 mol) in 600 mL of 5%
aqueous KOH was electrolyzed in a diaphragm cell with a
mercury cathode and graphite anode. The cathode current
density was 0.15-0.16 A/cm2, and the temperature of the
electrolyte was 40-50 °C. During electrolysis, the salt gradu-
ally dissolved. The catholyte was filtered to remove suspended
particles, acidified with 10% hydrochloric acid, and isolated
as described above.
(-CH3). Anal. Found: C, 81.93; H. 6.02; N, 6.12. Calcd (C16H13
NO): C, 81.68; H, 5.57; N, 5.95.
-
Gen er a l P r oced u r e for th e Syn th esis of N-Su bstitu ted
3-(1H-3-In d olyl)-2-m eth ylp r op ion ic Acid s. Eth yl (Z)-2-
Met h yl-3-(1-a r yl-1H -in d ol-3-yl)-2-p r op en oa t es. In a 500
mL flask equipped with a magnetic stirrer and a dropping
funnel, sodium hydride (6.4 g, 0.27 mol) was suspended in THF
(100 mL) under an argon atmosphere. The flask was cooled to
0 °C, and a solution of triethyl 2-phosphonopropionate (43 mL,
0.19 mol) in THF (30 mL) was added slowly to the hydride.
2-Meth yl-3-(1-p h en yl-1H-in d ol-3-yl)p r op ion ic Acid (5a )
(Meth od A). Colorless crystals (from alcohol). Mp: 155-156
°C. Yield: 36.9 g (76%). 1H NMR (δ, CDCl3, 25 °C): 8.3 (s,
1H, CH arom), 7.9 (m, 1H, CH arom), 7.7s7.2 (m, 8 H, arom
and CH indole), 3.35 (m, 1H, CH), 2.9 (m, 2H, -CH2-), 1.35
(d, J ) 8.2 Hz, 3H, -CH3). 13C NMR (δ, CDCl3, 25 °C): 180.2
(-COOH); 139.9, 139.2, 127.2, 117.9 (>C<); 128.1, 124.3,
(35) Chadwick, J . C.; Morini, G.; Albizatti, E.; Balbontin, G.;
Mingozzi, I.; Cristofori, A.; Sudmeijer, O.; van Kessel, G. M. M.
Macromol. Chem. Phys. 1996, 197, 2501 and references therein.

2720 Organometallics, Vol. 22, No. 13, 2003
van Baar et al.
122.9, 122.7, 121.8, 120.6, 118.6, 110.9, 40.1 (dCH-); 34.2
(-CH2-); 14.8 (-CH3). Anal. Found: C, 77.36; H, 6.08; N, 5.10.
Calcd (C20H21NO2): C, 77.4; H, 6.13; N, 5.01.
78.85; H, 7.25; N, 6.91. Calcd (C13H13NO): C, 78.36; H, 6.58;
N, 7.03.
Gen er a l P r oced u r e for th e Syn th esis of 2-Meth yl-4-
a r yl-1,2,3,4-tetr a h yd r ocyclop en ta [b]in d oles. Lithium alu-
minum hydride (4.8 g, 0.13 mol) was suspended in 400 mL of
ether under an argon atmosphere. The suspension was cooled
to 0 °C, and a solution of 2-methyl-4-aryl-1,2,3,4-tetrahydro-
cyclopenta[b]indol-3-one (0.18 mol) in ether (150 mL) was
added slowly. The reaction mixture was stirred for an ad-
ditional 2 h at room temperature and then decomposed using
a solution of ammonium chloride. Aluminum hydroxide was
filtered off, the organic phase was separated, washed with
water, and dried over Na2SO4. After removal of the solvent
the alcohols obtained could be used without further purifica-
tion.
2-Met h yl-3-(1-(2-m et h ylp h en yl)-1H -in d ol-3-yl)p r op i-
on ic Acid (5b) (Meth od B). Fine crystals (from alcohol).
1
Mp: 187-188 °C. Yield: 44.2 g (86%). H NMR (δ, DMSO-d6,
25 °C): 7.7 (m, 1H, CH arom), 7.4 (m, 4 H, CH arom), 7.2 (s,
1H, CH indole), 7.1 (m, 2H, CH arom), 6.9 (m, 1H, CH arom),
3.1 (m, 1H, CH), 2.8 (m, 2H, -CH2-), 2.0 (s, 3H, -CH3), 1.2
(d, J ) 6.6 Hz, 3H, -CH3). 13C NMR (δ, DMSO-d6, 35 °C):
177.3 (-COOH); 137.6, 136.6, 134.9, 127.5, 113.3 (>C<); 131.2,
128.1, 127.6, 127.1, 126.9, 121.9, 119.2, 118.8, 109.9, 39.7 (d
CH); 28.5 (-CH2-); 17.1 (-CH3); 16.8 (-CH3). Anal. Found:
C, 77.45; H, 6.62; N, 5.03. Calcd (C21H23NO2): C, 77.79; H,
6.53; N, 4.77.
A corresponding alcohol (15 g) was dissolved in benzene (300
mL), TsOH (0.05 g) was added, and the mixture was refluxed
with a Dean-Stark trap until the evolution of water ceased.
The reaction mixture was cooled, and the organic layer was
washed with a solution of NaHCO3, dried over Na2SO4, and
concentrated. The residue was purified by chromatography on
a column with silica gel (with a hexane/benzene (3/2) mixture
as eluent).
Gen er a l P r oced u r e for th e Syn th esis of 2-Meth yl-4-
a r yl-1,2,3,4-tetr a h yd r ocyclop en ta [b]in d ol-3-on es. Phos-
phorus pentoxide (6 g, 0.042 mol) was dissolved in methane-
sulfonic acid (40 mL, 0.625 mol). To the hot solution was added
dry 2-methyl-3-(1-aryl-1H-3-indolyl)propionic acid (0.036 mol)
in one portion. The mixture was stirred for 10 min, and the
hot reaction mass was then poured onto 200 g of crushed ice.
The ketone was extracted with methylene chloride. The
organic phase was washed with water, a solution of KHCO3,
and again with water and then dried over MgSO4, and the
solvent was removed.
2-Meth yl-4-ph en yl-1,4-dih ydr ocyclopen ta[b]in dole (7a).
Yellow crystals. Mp: 121.5-122 °C. Yield (overall): 39.7 g
1
(89%). H NMR (δ, C6D6, 25 °C): 7.8 (d, J ) 12.1 Hz, 1H, CH
arom), 7.6 (d, J ) 11.9 Hz, 1H, CH arom), 7.4 (m, 3H, CH
arom), 7.2 (m, 3H, CH arom), 7.1 (m, 1H, CH arom), 6.3 (s,
1H, dCH), 3.1 (s, 2H, -CH2-), 1.9 (s, 3H, -CH3). 13C NMR
(δ, C6D6, 25 °C): 149.1, 148.9, 140.1, 139.3, 119.2, 118.3 (>C<);
129.6, 126.0, 124.9, 120.9, 120.5, 118.1, 111.5 (dCH-), 36.5
(-CH2-), 17.3 (-CH3). Anal. Found: C, 87.91; H, 6.26; N, 5.83.
Calcd (C18H15N): C, 88.13; H, 6.16; N, 5.71.
2-Meth yl-4-(2-m eth ylp h en yl)-1,4-d ih yd r ocyclop en ta [b]-
in d ole (7b). Viscous yellow oil. Yield (overall): 38.7 g (83%).
1H NMR (δ, CDCl3, 25 °C): 7.6 (d, J ) 12.1 Hz, 1H, CH arom),
7.4 (m, 5H, CH arom), 7.2 (m, 1H, CH arom), 7.0 (m, 1H, CH
arom), 6.2 (sext, J ) 1.5 Hz, 1H, dCH-), 3.3 (s, 2H, -CH2-),
2.3 (d, J ) 1.6 Hz, 3H, -CH3), 2.1 (s, 3H, -CH3). 13C NMR (δ,
CDCl3, 25 °C): 149.7, 149.4, 140.4, 137.2, 135.9, 124.5, 117.3
(>C<); 131.1, 128.1, 127.7, 126.6, 119.8, 119.4, 117.6, 117.3,
110.9 (dCH-); 36.6 (-CH2-); 17.6 (-CH3); 17.4 (-CH3). Anal.
Found: C, 87.73; H, 7.02; N, 5.25. Calcd (C19H17N): C, 87.99;
H, 6.61; N, 5.40.
2,4-Dim eth yl-1,4-d ih yd r ocyclop en ta [b]in d ole (7c). Col-
orless oil (after chromatographic purification with a hexane-
methylene chloride (1:1) mixture as eluent), which undergoes
transition into a glassy state on cooling. Yield (overall): 23.7
g (72%).1H NMR (δ, C6D6, 25 °C): 7.8 (d, J ) 10.1 Hz, 1H, CH
arom), 7.2 (m, 3H, CH arom), 6.1 (s, 1H, CH), 3.1 (s, 3H,
N-CH3), 3.0 (s, 2H, -CH2-), 2.0 (m, 3H, -CH3). 13C NMR
(δ,C6D6, 25 °C): 149.8, 148.7, 140.9, 125.3, 110.0 (>C<); 119.8,
119.5, 117.9, 117.1, 109.9 (dCH-); 53.2 (-CH2-); 30.2 (N-
CH3); 17.4 (-CH3). Anal. Found: C, 85.30; H, 7.33; N, 7.37.
Calcd (C13H13N): C, 85.21; H, 7.15; N, 7.64.
[µ-Dim et h ylsila n ed iylb is(η5-2-m et h yl-4-p h en ylcyclo-
p en ta [b]in d olyl)]zir con iu m (IV) Dich lor id e (10a ). 2-Meth-
yl-4-phenyl-1,4-dihydrocyclopenta[b]indole (7a ; 3.6 g, 14 mmol)
was dissolved in a mixture of toluene (40 mL) and THF (5 mL),
and a solution of butyllithium in hexane (8.8 mL, 14 mmol,
1.6 M) was added on cooling. Then, (CH3)2SiCl2 (0.9 mL, 7
mmol) was added to the solution of the lithium salt and the
reaction mixture was stirred at room temperature.
2-Meth yl-4-ph en yl-1,4-dih ydr ocyclopen ta[b]in dol-3(2H)-
on e (6a ). Fine yellow crystalline substance (after recrystal-
lization from
a hexane/ethyl acetate (1/1) mixture). Mp:
110.5-111 °C. Yield: 6.6 g (71%). 1H NMR (δ, C6D6, 25 °C):
7.8 (d, J ) 11.6 Hz, 1H, arom), 7.8s7.3 (m, 8H, CH arom),
3.44 (dd, J 1 ) 6.4 Hz, J 2 ) 16.1 Hz, 1H, -CH2-), 3.1(m, 1H,
-CH-), 2.76 (dd, J 1 ) 2.4 Hz, J 2 ) 16.1 Hz, 1H, -CH2-), 1.45
(d, J ) 7.5 Hz, 3H, -CH3). 13C NMR (δ, C6D6, 25 °C): 194.6
(>CdO); 144.9; 144.5, 137.9; 136.9; 124.4 (>C<); 129.0; 127;
121.7; 121.2; 112.6, 47.3 (dCH-); 28.0 (-CH2-); 17.1 (-CH3).
Anal. Found: C, 82.62; H, 5.58; N, 5.41. Calcd (C18H15NO):
C, 82.73; H, 5.79; N, 5.36.
2-Meth yl-4-(2-m eth ylp h en yl)-1,4-d ih yd r ocyclop en ta [b]-
in d ol-3(2H)-on e (6b). The substance was purified by chro-
matography on a column with silica gel (with a benzene/ethyl
acetate (4/1) mixture as eluent). Yellow-brown oil. Yield: 6.4
1
g (65%). H NMR (δ, CDCl3, 25 °C): 7.8 (d, J ) 12.1 Hz, 1H,
arom), 7.5-7.1 (m, 7H, CH arom), 3.52 and 3.48 (dd, J 1 ) 6.8
Hz, J 2 ) 16.8 Hz, 1H, -CH2-), 3.1(m, 1H, -CH-), 2.84 and
2.8 (dd, J 1 ) 2.2 Hz, J 2 ) 16.8 Hz, 1H, -CH2-), 2.13 and 2.11
(s, 3H, Ar-CH3), 1.47 and 1.45 (d, J ) 4.1 Hz, 3H, -CH3). 13
C
NMR (δ, C6D6, 25 °C): 196.07 and 197.02 (>CdO); 144.9,
144.28 and 144.22, 144.2, 138.3, 135.9 and 135.8, 123.05
(>C<); 130.9, 128.4, 127.9, 126.9, 126.5, 121.3, 120.6, 112.2,
47.1 and 47.0 (dCH-); 28.4 (-CH2-); 17.3 (-CH3); 17.15 and
17.07 (-CH3). Anal. Found: C, 82.73; H, 6.49; N, 5.14. Calcd
(C19H17NO): C, 82.88; H, 6.22; N, 5.09.
2,4-Dim et h yl-1,2,3,4-t et r a h yd r ocyclop en t a [b]in d ol-3-
on e (6c). In a flask equipped with a reflux condenser and a
dropping funnel, P2O5 (7.2 g, 0.05 mol) was dissolved with
stirring in CH3SO3H (60 mL, 0.9 mol) heated to 120 °C. Then,
a mixture of 1-methylindole (30 mL, 0.23 mol), methacrylic
acid (23 mL, 0.26 mol), and methylene chloride (30 mL) was
added dropwise to the hot solution. The reaction mixture was
stirred for 2.5 h at 70 °C, cooled, and poured into 200 mL of
water. The crude product was extracted with ether, and the
organic phase was washed with a solution of KHCO3 and dried
over MgSO4. After removal of the solvent the residue was
purified by chromatography on a column with silica gel (with
a benzene/ethyl acetate (4/1) mixture as eluent) to obtain 25.7
g of dark yellow oil. Yield: 51%. 1H NMR (δ, CDCl3, 25 °C):
7.7 (d, J ) 12.1 Hz, 1H, CH arom) 7.4-7.2 (m, 3H, arom), 3.9
(s, 3H, -CH3, indole), 3.4 (dd, J 1 ) 8.1 Hz, J 2 ) 17.5 Hz, 1H,
-CH2-), 3.1 (m, 1H, CH), 2.7 (dd, J 1 ) 4.2 Hz, J 2 ) 17.5 Hz,
1H, -CH2-) 1.4 (d, J ) 8.1 Hz, 3H, -CH3). Anal. Found: C,
The reaction was monitored by 1H NMR spectroscopy. After
24 h the 1H NMR spectrum (C6D6, 25 °C) of the reaction
mixture showed the complete conversion of starting compound
7a . Signals typical of a mixture of the rac and meso forms of
bridged compound 9a were observed at δ 6.48 and 6.45 (t, J )
1.4 Hz, 1H, dCH-), 4.50 (s, 1H, dCH-Si rac form), 4.04 (s,
1H, dCH-Si meso form), - 0.03 and -0.19 (s, 3H, Si-CH3
meso form), and -0.26 (s, 3H, Si-CH3 rac form). After

ansa-Zirconocenes
Organometallics, Vol. 22, No. 13, 2003 2721
Ta ble 6. Cr ysta l Da ta a n d Str u ctu r e Refin em en t
for 10a
completion of the reaction, the solution was decanted from the
LiCl precipitate and concentrated, and the oily residue was
dissolved in toluene (40 mL). After this solution was cooled to
-40 °C, a solution of BuLi in hexane (10 mL, 1.6 M) was added
and the mixture was stirred for 2 h at room temperature and
cooled again; ZrCl4 (1.6 g, 7 mmol) was added, and the mixture
was stirred for an additional 5 h at 20 °C. Then the solvent
empirical formula
fw
C38H32Cl2N2SiZr
706.87
temp
93(2) K
wavelength
cryst syst
space group
unit cell dimens
0.710 69 Å
triclinic
P1h
was removed. Recrystallization from ether resulted in
a
mixture of rac and meso forms of compound 10a as an orange
crystalline substance (3.5 g, 76%). Fractional crystallization
from an ether/hexane mixture gave the pure racemic form as
yellow crystals (0.76 g, 17%) and the pure meso form as orange
a
9.295(2) Å
b
12.914(3) Å
c
15.619(3) Å
R
111.88(3)°
81.03(3)°
1
crystals (1.04 g, 23%). H NMR (δ, CD2Cl2, 25 °C): rac form,
â
7.9 (d, J ) 6.1 Hz, 2H, dCH- arom), 7.7 (m, 4H, dCH- arom),
7.5 (m, 10 H, dCH- arom), 7.2 (m, 2H, dCH- arom), 6.35 (s,
2H, dCH-), 2.1 (s, 6H, -CH3), 1.45 (s, 6H, Si-CH3); meso
form, 8.1 (d, J ) 7 Hz, 2H, dCH- arom), 7.6 (m, 4H, dCH-
arom), 7.5 (m, 10 H, dCH- arom), 7.0 (m, 2H, dCH- arom),
6.25 (s, 2H, dCH-), 2.5 (s, 6H, -CH3), 1.65 (s, 3H, Si-CH3),
1.3 (s, 3H, Si-CH3). Anal. Found for r a c-10a : C, 64.41; H,
4.63; N, 4.11. Calcd (C38H32Cl2N2SiZr): C, 64.57; H, 4.56; N,
3.96.
γ
V
117.77(3)°
539.0(6) Å3
Z
1
density (calcd)
abs coeff
1.525 g/cm3
602 mm-1
F(000)
724
cryst size
θ range for data collection
index ranges
0.4 × 0.3 × 0.2 mm
1.41-29.60°
-11 e h e 11, -16 e k e 15,
0 e l e 12
[µ-Dim eth ylsila n ed iylbis(η5-2-m eth yl-4-(2-m eth ylp h e-
n yl)cyclopen ta[b]in dolyl)]zir con iu m (IV) Dich lor ide (10b).
2-Methyl-4-(2-methylphenyl)-1,4-dihydrocyclopenta[b]indole (7b;
2.6 g, 10 mmol) was dissolved in a mixture of toluene (50 mL)
and THF (5 mL), and a solution of butyllithium in hexane (6.5
mL, 10.4 mmol, 1.6 M) was added on cooling. Then (CH3)2-
SiCl2 (0.65 mL, 5.3 mmol) was added to the solution of the
lithium salt and the mixture maintained at room temperature
for 24 h. The solution was decanted and concentrated, the
semisolid residue was dissolved in 40 mL of toluene, and the
solution was cooled to -40 °C. A solution of BuLi in hexane
(6.5 mL, 1.6 M) was added, the mixture was stirred for 2 h at
room temperature and cooled again to -40 °C, ZrCl4 (1.6 g, 7
mmol) was added, the mixture was stirred for an additional
24 h, and then the solvent was removed. Recrystallization from
ether gave a mixture of rac and meso forms of compound 10b
as a dark orange crystalline substance (1.32 g, 33%). Repeated
recrystallization from hexane-ether mixture allowed isolation
no. of rflns collected
no. of indep rflns
refinement method
no. of data/restraints/params
goodness of fit on F2
final R indices (I > 2σ(I))
R indices (all data)
5252
5242 (R(int) ) 0.0128)
full-matrix least squares on F2
5242/0/526
0.879
R1 ) 0.0364, wR2 ) 0.0961
R1 ) 0.0438, wR2 ) 0.0991
0.0069(9)
extinction coeff
largest diff peak and hole
0.487 and -0.487 e/Å-3
of ether. The ethereal solution was cooled again, and a solution
of butyllithium in hexane (9.5 mL, 2.11 M) was added. After
30 min, ZrCl4 (2.2 g, 10 mmol) was added. The mixture was
stirred for 5 h at room temperature, and the solution was
decanted and concentrated. The product was isolated as a
yellow-brown crystalline substance (the pure rac form).
1
Yield: 1.72 g (32%). H NMR (δ, CD2Cl2, 25 °C): 7.6 (d, J )
8.8 Hz, 2 H, dCH- arom), 7.2 (m, 6H, dCH- arom), 6.05 (s,
2H, dCH-), 3.7 (s, 6H, N-CH3), 2.1 (s, 6H, - CH3), 1.4 (s,
6H, Si-CH3). Anal. Found: C, 57.48; H, 5.01; N, 4.55. Calcd
(C28H28Cl2N2SiZr): C, 57.71; H, 4.84; N, 4.81.
1
of the pure rac form (0.21 g, 6%). H NMR (δ, CD2Cl2, 25 °C):
rac form, 7.95 (m, 4H, dCH- arom), 7.3 (m, 6 H, dCH- arom),
7.25 (m, 6H, dCH- arom), 5.95 (s, 2H, dCH-), 2.1 (s, 6H,
CH3-), 1.9 (s, 6H, CH3-), 1.45 (s, 6H, CH3-Si); meso form,
8.1 (d, J ) 5.9 Hz, 2H, dCH- arom), 7.85 (d, J ) 6 Hz, 2H,
dCH- arom), 7.35 (m, 6 H, dCH- arom), 7.0 (m, 4H, dCH-
arom), 6.7 (d, J ) 7 Hz, 2H, dCH- arom), 6.0 (s, 2H, dCH-),
2.5 (s, 6H, CH3-), 1.8 (s, 6H, -CH3-), 1.65 (s, 3H, Si-CH3),
1.4 (s, 3H, Si-CH3). Anal. Found for r a c-10b: C, 64.93; H,
5.06; N, 3.52. Calcd (C40H36Cl2N2SiZr): C, 65.37; H, 4.94; N,
3.81.
Cr ysta l Str u ctu r e Deter m in a tion of 10a . Heavy-atom
positions were determined by a Patterson map using the
AREN-88 program,36 and the remaining hydrogen atoms were
located from successive difference Fourier map calculations.
Refinements were carried out with the SHELXL-93 program37
by using full-matrix least-squares techniques minimizing the
function ∑w(|Fo| - |Fc|).2 Crystal data and structure determi-
nation details are given in Table 6.
LIP P Test P r oced u r es. Polymerizations were carried out
in a 5 L batch reactor provided with a turbine stirrer. A steam/
water system was used for temperature control. Exotherms
were monitored by determining the temperature difference
between the reactor contents and the incoming cooling water
(thermocouple). A specially designed injection system, which
can withstand the high pressures in the reactor, was applied
to introduce the catalyst components into the reactor.
Procedures for aluminoxane preparation, metallocene pre-
alkylation, polymerization in LIPP, and workup of the polymer
products are outlined below. Gas-cap composition during
propylene polymerization was analyzed by means of a GC
apparatus from Interscience (propane/propylene ratio for yield
versus time measurements) and an Orbisphere hydrogen
analyzer.
Lith iu m Sa lt of 2,4-Dim eth yl-1,4-d ih yd r ocyclop en ta -
[b]in d ole (8c). 2,4-Dimethyl-1,4-dihydrocyclopenta[b]indole
(7c; 7.1 g, 0.04 mol) was dissolved in 200 mL of dehydrated
ether, and a solution of butyllithium in hexane (50 mL, 1.6
M) was added on cooling. The reaction mixture darkened and
a thin light-coffee precipitate began forming after some time.
The precipitate was decanted with ether (5 × 150 mL) and
dried in vacuo to give the desired product as a beige powder
1
(5.5 g, 71%). H NMR (δ, THF-d8, 25 °C): 7.2 (d, J ) 11.9 Hz,
1H, Ar H), 6.9 (d, J ) 11.9 Hz, 1H, Ar H), 6.7 (m, 2H, Ar H),
5.5 and 5.1 (both s, 1H each, Cp H), 3.4 (s, 3H, N-CH3), 2.3
(s, 3H, -CH3). 13C NMR (δ, THF-d8, 25 °C): 144.1, 139.2, 125.8,
117.3, 106.9 (>C<); 118.1, 116.7, 116.5, 109.3, 87.8, 78.2 (d
CH-); 31.3 (N-CH3); 17.3 (-CH3).
[µ-Dim eth ylsila n ed iylbis(η5-2,4-d im eth ylcyclop en ta [b]-
in d olyl)]zir con iu m (IV) Dich lor id e (10c). Lithium salt 8c
(3.75 g, 20 mmol) was dissolved in a mixture of toluene (50
mL) and THF (5 mL), and (CH3)2SiCl2 (1.2 mL, 10 mmol) was
added on cooling to -40 °C. The solution was decanted and
concentrated, and the viscous residue was dissolved in 30 mL
The autoclave was heated to 150-160 °C overnight, while
being purged with nitrogen, cooled, and then pickled at 70 °C
(36) Adrianov, V. I. Kristallografiya 1987, 32, 228.
(37) Sheldrick, G. M. SHELXL-93: Program for Crystal Structure
Refinement; University of Go¨ttingen: Go¨ttingen, Germany, 1993.

2722 Organometallics, Vol. 22, No. 13, 2003
van Baar et al.
using a mixture of TIBA (0.25 g), toluene (20 mL), and
propylene (500 g). The pickle mixture was removed, the
temperature reduced to 20 °C, and the reactor then charged
with 1650 g of liquid propylene, while the temperature was
increased from 20 to 30 °C. Then 4-5 vol % hydrogen was
added to the gas cap, aiming at 1-1.5 vol % hydrogen in the
gas cap at 70 °C.
procedure is identical with that described above. The polym-
erization time (as measured from initiation of warming) was
60 min.
When MAO was used as cocatalyst at low Al/Zr levels, the
above procedure was followed, but the metallocene dichloride
was simply mixed with MAO (500 equivalents) and, after 5
min, injected into the autoclave.
Typ ica l Lon g Cold Sta r t P r oced u r e a t High Al/Zr
Ra tio. (a ) AKO Coca ta lyst (Exp er im en t 9). A 1.73 g
amount of tris(2-phenylpropyl)aluminum, Al(CH2CHMePh)3
(4.5 mmol), was dissolved in 20 mL of toluene in a bottle with
a septum cap. The solution was cooled to 0-4 °C using an ice
bath, and 41 µL of water (2.28 mmol) was added in two shots
using a 100 µL syringe, while the temperature was maintained
below 15 °C. The resulting solution was injected into the
reactor using 20 mL of toluene. Meanwhile, 6.6 mg of 10a (9
µmol) was dissolved in 14.5 g of toluene, and 0.743 g of the
obtained solution containing 0.48 µmol of Zr was reacted with
0.11 g of TIOA (0.3 mmol), resulting in a color change from
yellow to light yellow. Ten minutes after the introduction of
the hydrolyzed alkylaluminum mixture containing 4.5 mmol
of aluminoxane into the reactor, the alkylated zirconocene
(aged for 5 min) was injected into the reactor at 30 °C using
20 mL of toluene. After 10 min the temperature was raised
over 6-7 min to 70 °C and polymerization was continued (43
min from initiation of warming), using 840-1100 rpm stirring,
keeping the hydrogen concentration at 1 vol % in the gas cap.
The polymerization was stopped by injection of 5-10 mL of
methanol. The heating was then discontinued, the propylene
was rapidly vented, and the powdered polypropylene was
collected. The polypropylene fractions were dried (70-80 °C,
200 mbar, nitrogen purge) and combined to give the total yield
of polypropylene.
(b) MAO Coca ta lyst (Exp er im en t 1). A 2.44 g amount of
a toluene solution of MAO (4.98 w % Al, 4.5 mmol) was injected
into the reactor using 20 mL of toluene. Meanwhile, 6.6 mg of
10a (9 µmol) was dissolved in 14.5 g of toluene, and 2.228 g of
the obtained solution containing 1.44 µmol of Zr was reacted
with 0.405 g of MAO solution (0.75 mmol), resulting in a color
change from yellow to red. Ten minutes after the introduction
of the MAO solution, the alkylated zirconocene (aged for 5 min)
was injected into the reactor (20 mL of toluene) at 30 °C. The
remaining procedure is identical with that described above.
The polymerization time (as measured from initiation of
warming) was 49 min.
Sh or t Cold Sta r t a n d Hot Sta r t P r oced u r es. In short
cold start experiments, the above procedures were followed,
but the temperature of the autoclave (under 4-5 vol % H2)
was raised immediately from 30 °C (to 70 °C) after injection
of the catalyst solution. In the hot start procedure the cata-
lyst was injected into the autoclave at 70 °C (for experi-
ments with hydrogen, under 1.0 vol % H2). Fouled material, if
present, was removed using hot xylene and precipitated with
methanol.
P olym er An a lysis P r oced u r es. GP C An a lysis. High-
temperature GPC analyses were carried out by Rapra Tech-
nology Ltd. A single solution of each sample was prepared by
adding 15 mL of solvent to a weighed quantity of sample and
refluxing gently for 20 min. The solutions were then filtered
through a fiber pad at 140 °C, and part of each filtered solution
was transferred into special glass sample vials. The vials were
then placed in a heated sample compartment, and after an
initial delay of 20 min to allow the samples to equilibrate
thermally, injection of part of the contents of each vial was
carried out automatically in series. The following chromato-
graphic conditions were used: column, PLgel 2 x mixed bed-
B, 30 cm, 10 µm; solvent, 1,2-dichlorobenzene with antioxidant;
flow rate, 1.0 mL/min; temperature, 140 °C; detector, refractive
index; calibration, polystyrene.
DSC Ch a r a cter iza tion . DSC analysis of the copolymers
was performed on a Perkin-Elmer DSC-4. The following
temperature program was used: (1) temperature 1 25 °C, time
1 1.0 min, rate 1 10.0 °C/min; (2) temperature 2 160 °C, time
2 0.1 min, rate 2 20.0 °C/min; (3) temperature 3 25 °C, time 3
1.0 min, rate 3 10.0 °C/min; (4) temperature 4 160 °C, time 4
0.1 min, rate 4 20.0 °C/min. (5) temperature 5 25 °C.
Xylen e Solu bles. The fraction of xylene solubles at 25 °C
was determined using ca. 2 g of polymer sample and 100 mL
of xylene or ca. 2.5 g in 250 mL of xylene.
Ack n ow led gm en t. This work was funded entirely
by Basell Polyolefins, for which we are very grateful.
The important contribution of I. Bolt-Westerhoff and S.
Kerkvliet (SRTCA) in performing the polymer NMR
analyses is acknowledged.
P olym er iza tion a t Low Al/Zr Ra tio (Exp er im en t 13).
A solution of 0.20 g of TIBA (1 mmol) in toluene was introduced
into the reactor with ca. 25 mL of toluene. Separately, a
solution of Al(CH2CHMePh)3 (AK) in 20 mL of toluene was
hydrolyzed using 1/2 equiv of water. Meanwhile, 1.935 g of a
toluene solution of 10b (1.0 µmol) was reacted with 0.038 g of
TIOA (0.10 mmol). After 5 min, 2.5 g of a solution of the
alkylaluminoxane containing 0.5 mmol of aluminoxane was
added to the alkylated zirconocene solution. Twenty minutes
after the introduction of the TIBA into the reactor, the
activated heterocene (aged for 5 min) was injected into the
reactor (at 30 °C) using 20 mL of toluene. The remaining
Su p p or tin g In for m a tion Ava ila ble: For 10a , tables of
bond lengths and angles, atomic coordinates and equivalent
isotropic displacement parameters, anisotropic displacement
parameters, hydrogen coordinates, and isotropic displacement
parameters. This material is available free of charge via the
Internet at http://pubs.acs.org.
OM020505T