Selectivity in propylene polymerization with group 4 Cp-amido catalysts

Article abstract of DOI:10.1021/om9609628

Mono-Cp-amido complexes [η⁵:η¹-Cp′SiMe₂(NR)]MCl ₂ (M=Ti, Zr) have been synthesized by metathesis and amine elimination routes in low to moderate yields (5-50%). Two optically active examples based on (S)-(-)-(α)-methylbenzylamine are presented, and the molecular structure of (+)-(R)-[η⁵:η¹-(Ind)SiMe ₂-(S)-NCHMePh]TiCl₂ (4) has been determined by X-ray crystallography. These complexes are active in the presence of methylaluminoxane for propylene polymerization and yield high molecular weight atactic polymers with slight syndiotactic enrichment. Productivities, molecular weights, and tacticities are dependent on Cp′ and NR. Incorporation of a chiral amine into the ligand framework has little effect on stereospecificity. Polypropylenes derived from the titanium catalysts show significant amounts of 2,1-monomer insertion (2-5%).

Full text of DOI:10.1021/om9609628

Organometallics 1997, 16, 2879-2885
2879
Selectivity in P r op ylen e P olym er iza tion w ith Gr ou p 4
Cp -Am id o Ca ta lysts
Andrew L. McKnight, Md. Athar Masood, and Robert M. Waymouth*
Department of Chemistry, Stanford University, Stanford, California 94305
Daniel A. Straus
Department of Chemistry, San J ose State University, San J ose, California
Received November 13, 1996X
Mono-Cp-amido complexes [η5:η1-Cp′SiMe2(NR)]MCl2 (M ) Ti, Zr) have been synthesized
by metathesis and amine elimination routes in low to moderate yields (5-50%). Two optically
active examples based on (S)-(-)-(R)-methylbenzylamine are presented, and the molecular
structure of (+)-(R)-[η5: η1-(Ind)SiMe2-(S)-NCHMePh]TiCl2 (4) has been determined by X-ray
crystallography. These complexes are active in the presence of methylaluminoxane for
propylene polymerization and yield high molecular weight atactic polymers with slight
syndiotactic enrichment. Productivities, molecular weights, and tacticities are dependent
on Cp′ and NR. Incorporation of a chiral amine into the ligand framework has little effect
on stereospecificity. Polypropylenes derived from the titanium catalysts show significant
amounts of 2,1-monomer insertion (2-5%).
The use of well-defined single-site metallocene cata-
lysts in olefin polymerizations is well-established.1
Isotactic, syndiotactic, and high molecular weight atactic
polypropylenes can be produced through the appropriate
choice of metallocene precursor.1,2 Recently, a new class
of homogeneous Ziegler-Natta catalysts were developed
by researchers at Dow and Exxon.3-5 These group 4
mono-cyclopentadienyl-amido (CpA) catalysts are based
on the ligand system first prepared by Bercaw6 for
organoscandium complexes and are distinguished by a
sterically accessible catalyst active site.7,8 They differ
from bis(cyclopentadienyl) metallocenes in their ability
to readily incorporate R-olefins, notably styrene,5,9,10 in
copolymerizations with ethylene. These catalysts are
remarkably stable up to polymerization temperatures
of 160 °C, and the titanium derivatives are less easily
reduced by methylaluminoxane (MAO) than the corre-
sponding titanocenes.
literature3 details several non-C2 symmetric catalysts
which yield isotactic-enriched polymers ([mmmm] >
75%) in the presence of MAO. Other reports indicate
that propylene polymerization with cationic catalysts
can yield either syndiotactic ([r] ) 95%)14 or isotactic
([mmmm] ) 95%)15 polymers depending on the nature
of the initiating system. Additionally, little is known13
about the regiospecificity of CpA catalysts in propylene
polymerization.
In view of the extraordinary potential of these cata-
lysts and the scattered and conflicting information in
the patent literature regarding the mechanism of poly-
merization and origin of stereocontrol with CpA cata-
lysts, we undertook a study of their propylene poly-
merization behavior. In particular, we sought to exploit
chiral amino auxilliaries as either stereocontrol ele-
ments or resolving agents in the CpA ligand system.
Herein, we report propylene polymerization results
using several CpA catalysts of varying steric and
electronic demands, including examples of optically
active CpA catalysts.
Polypropylenes (PP) produced by CpA catalysts are
generally atactic, with varying degrees of syndio-
tacticity.5,10-13 However, one report in the patent
X Abstract published in Advance ACS Abstracts, May 15, 1997.
(1) Brintzinger, H. H.; Fischer, D.; Mulhaupt, R.; Rieger, B.;
Waymouth, R. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 1143-1170.
(2) Ewen, J . A.; Elder, M. J . Makromol. Chem., Macromol. Symp.
1993, 66, 179-190.
(3) Canich, J . A. M. (Exxon) U.S. Patent 5,026,798, 1991.
(4) Devore, D. D.; Timmers, F. J .; Hasha, D. L.; Rosen, R. K.; Marks,
T. J .; Deck, P. A.; Stern, C. L. Organometallics 1995, 14, 3132-3134.
(5) Stevens, J . C.; Timmers, F. J .; Wilson, D. R.; Schmidt, G. F.;
Nickias, P. N.; Rosen, R. K.; Knight, G. W.; Lai, S.-y. (Dow Chemical
Company) Eur. Pat. Appl. 0416 815 A2, 1990.
Resu lts
P r ep a r a t ion of Mon o-Cp -Am id o Com p lexes.
Ligands were prepared according to published proce-
dures. Treatment of (Cp′H)SiMe2Cl (Cp′ ) tetramethyl-
cyclopentadienyl (Me4C5),5 1-indenyl (Ind),5 and 9-fluo-
renyl (Flu)16,17 with excess amine in THF yielded
(Cp′H)SiMe2NHR (R ) tBu, (S)-(-)-R-methylbenzyl,
C6H11), which can be purified by distillation.18 Scheme
(6) Shapiro, P. J .; Bunel, E.; Schaefer, W. P.; Bercaw, J . E.
Organometallics 1990, 9, 867-9.
(7) Okuda, J . Chem. Ber. 1990, 123, 1649.
(14) Turner, H. W.; Hlatky, G. G.; Canich, J . A. M. (Exxon) PCT
Int. Appl. WO 9,319,103, 1993.
(8) Carpenetti, D. W.; Kloppenburg, L.; Kupec, J .; Peterson, J . L.
Organometallics 1996, 15, 1572-1581.
(15) Shiomura, T.; Asanuma, T.; Inoue, N. Macromol. Rapid Com-
mun. 1996, 17, 9-14.
(9) Sernetz, F. G.; Mulhaupt, R.; Waymouth, R. M. Macromol. Chem.
Phys. 1996, 197, 1071-1083.
(16) Okuda, J .; Schattenmann, F. J .; Wocadlo, S.; Massa, W.
Organometallics 1995, 14, 789-795.
(10) Pannell, R. B.; Canich, J . A. M.; Hlatky, G. G. (Exxon) PCT
Int. Appl. WO 94/00500, 1994.
(17) We found, like Okuda, that FluSiMe2Cl cannot be prepared
without some contamination by Flu2SiMe2. However, sublimation at
60 mTorr/100 °C gave pure FluSiMe2Cl.
(11) Canich, J . A. M. (Exxon) U.S. Patent 5,096,867, 1992.
(12) LaPointe, R. E.; Stevents, J . C.; Nickias, P. N.; McAdon, M. H.
(Dow Chemical Company) Eur. Pat. Appl. 0 520 732, 1992.
(13) Canich, J . A. M. (Exxon) U.S. Patent 5,504,169, 1996.
(18) As Herrmann showed, Ind-amino ligands are isolated as
mixtures of regioisomers.
S0276-7333(96)00962-4 CCC: $14.00 © 1997 American Chemical Society

2880 Organometallics, Vol. 16, No. 13, 1997
McKnight et al.
Sch em e 1. Syn th etic Ap p r oa ch es to Mon o-Cp -Am id o Com p lexes
Sch em e 2. Syn th esis of In d en yl Com p lexes by
Am in e Elim in a tion
1 summarizes the general synthetic routes. When Cp′
) Me4C5, double deprotonation with isopropylmagne-
sium chloride5 in THF/toluene followed by reaction with
TiCl3(THF)3 and PbCl2 oxidation19 gave [η5:η1-Me4C5-
SiMe2NR]TiCl2 in yields of 52% (1) and 16% (2).
Several routes were explored to prepare the indenyl-
amido complexes. Reaction of the indenyl dimagnesio
dichloride salts with TiCl3(THF)3 followed by PbCl2
oxidation or with TiCl4(THF)2 was unsuccessful, as was
the reaction of the dilithio salts with TiCl4. The only
viable metathesis route toward preparation of indenyl-
amido complexes proved to be reaction of the dilithio
salts with TiCl4(THF)2. Yields for this route were quite
low (<20%).
In the synthesis of [FluSiMe2NtBu]ZrCl2, it was
noted16 that preparations in THF or Et2O yielded the
Lewis base adducts [FluSiMe2NtBu]ZrCl2‚L (L ) THF,
Et2O). Earlier syntheses3 of [η5:η1-FluSiMe2NtBu]ZrCl2
utilized Et2O as a reaction solvent but made no mention
of base adducts. We therefore sought to prepare base-
free [FluSiMe2NtBu]ZrCl2 by treatment of the ligand
dilithio salt with ZrCl4 in toluene. The product is
obtained as bright yellow, air- and moisture-sensitive
crystals in 35% yield.
Optically active complexes 220,21 and 4 based on the
commercially available chiral (S)-(-)-(R)-methylbenzyl-
amine were prepared by the methods mentioned above
for Me4C5 and indene derivatives. In the case of 2, only
one isomer is possible and it gives a complex with an
optical rotation of [R]589 ) -190.5° (c ) 1.3 mg/mL,
CH2Cl2). On the other hand, two diastereomers are
possible for 4 because of the two possible orientations
of the indenyl ring. A single diastereomer was isolated
by preferential crystallization from Et2O/CH2Cl2 to give
(+)-(R)-[η5:η1-(Ind)SiMe2-(S)-NCHMePh]TiCl2 (4), with
an optical rotation of [R]589 ) +402° (c ) 0.95mg/mL,
CH2Cl2).
route of Herrmann22 for preparation of indenyl-amido
complexes via amine elimination from Ti(NR2)4. Double
protonation of Ti(NMe2)4 with (Ind-H)SiMe2NHR (R )
(S)-(-)-R-methylbenzyl, C6H11) in refluxing toluene gave
the expected diamide (Scheme 2). The (S)-(-)-R-methyl-
benzylamine derivative yields two diastereomers in a
ratio of 1.33/1 by 1H NMR. Subsequent treatment with
excess Me3SiCl8,23 in CH2Cl2 cleanly gave the desired
dichlorides in overall yields of 39% for the single
diastereomer of 4 and 45% for 5.
Cr ysta l Str u ctu r e Da ta for 4. To gain information
on the steric demands of the indenyl complexes and the
potential influence of metal chirality on reaction stereo-
specificity, a single-crystal X-ray diffraction analysis of
4 was performed. The crystal structure is illustrated
in Figure 1, and relevant bond distances and angles are
listed in Table 1.
The structural features of 4 are similar to those for
1.5,8 In particular, the “bite angle” (θ) of the ligand,
defined as Cp(cent)-Ti-N, is 106.3°. The metal to
ligand bond distances are also similar to those for 1,
with Ti(1)-N(1) ) 1.915 Å (vs 1.907 Å) and Ti(1)-
Ind(cent) ) 2.045 Å (vs 2.030 Å). The structure reveals
a pseudo-C2 symmetry when considering the aromatic
rings of the indene and the phenyl group of methyl-
Low yields for the metathesis routes, especially for
the Ind complexes, prompted us to follow the recent
(19) Luinstra, G. A.; Teuben, J . H. J . Chem. Soc., Chem. Commun.
1990, 1470.
(20) During the preparation of this manuscript, Okuda reported the
synthesis of this complex by similar methods.
(21) Okuda, J .; Verch, S.; Spaniol, T. P.; Stuermer, R. Chem. Ber.
1996, 129, 1429-1431.
(22) Herrmann, W. A.; Morawietz, M. J . A. J . Organomet. Chem.
1994, 482, 169-181.
(23) J ordan, R. F.; Diamond, G. M.; Christopher, J . N.; Kim, I.
Polym. Prepr. 1996, 37, 256.

Group 4 Cp-Amido Complexes
Organometallics, Vol. 16, No. 13, 1997 2881
Incorporation of (S)-(-)-(R)-methylbenzylamine or cy-
clohexylamine into the ligand framework (catalysts 2,
4, and 5) seems to have a slight effect in lowering the
stereospecificity of polymerization (compare catalysts 3
vs 4, entries 5-8).
Regioselectivity. All of the Me4C5- and Ind-based
catalysts (1-5) give polymers with a significant amount
of regioirregularities, while the fluorenylzirconium cata-
lyst 6 gives no misinsertions. Figure 2 shows this
difference between entries 8 and 12, respectively. Cheng
and Ewen26 have assigned the 13C NMR chemical shifts
for regioirregularities in atactic PP by comparison with
calculated spectra. They observed peaks due to 2,1- and
1,3-misinsertions (Figure 3) which were sensitive to the
overall tacticity of the polymer, as they were greatly
broadened relative to those for isotactic PP.25,27 13C
NMR spectra of entries 2, 4, 6, and 8 were obtained to
identify the type of misinsertions and to quantify the
degree of regioirregularities. According to the method
of Resconi25 and Mizuno,27 the percentages of 2,1- and
1,3-misinsertions can be calculated from the following
equations
F igu r e 1. ORTEP diagram of (+)-(R)-[η5:η1-(Ind)SiMe2-
(S)-NCHMePh]TiCl2 (4).
Ta ble 1. Selected Bon d Len gth s (Å) a n d An gles
(d eg) for 4
Ti(1)-Cl(1)
Ti(1)-N(1)
Si(1)-N(1)
2.257(2)
1.915(5)
1.739(6)
Ti(1)-Cl(2)
Ti(1)-Ind(cent)
Si(1)-C(3)
2.258(2)
2.045
1.876(6)
Cl(1)-Ti(1)-Cl(2)
Ti(1)-N(1)-C(12)
Si(1)-N(1)-C(12)
102.54(8)
120.4(4)
132.7(4)
Ind-Ti(1)-N(1)
Ti(1)-N(1)-Si(1)
N(1)-Si(1)-C(3)
106.3
106.9(3)
90.0(3)
%2,1 ) C3/(CH2(main) + 3C3 + C3′)
%1,3 ) C3′/(CH2(main) + 3C3 + C3′)
(1)
(2)
benzylamine. The orientation of the phenyl group is
such that it is pointed away from the underside of the
indene ring, while the methyl group lies under the ring.
This effectively directs the R-H toward the metal center.
The lessened steric demands of this amino group rela-
tive to the more sterically demanding t-butylamine of
1 may explain the 6° smaller Ti(1)-N(1)-C(12) angle
(120.4°) relative to that for 1 (Ti-N-C ) 132.7°).
P r op ylen e P olym er iza tion s. Propylene polymer-
izations were conducted for catalysts 1-6 in toluene
solution at 50 psig or in liquid propylene at 30 °C ((1
°C) for 1 h. Conditions for catalysts 5 and 6 were chosen
to match those used by Canich;3 similar conditions were
used for catalysts 1-4. The results are summarized in
Table 2.
where C3 and C3′ represent the intensities of carbons 3
and 3′ of the regioirregular units (Figure 3) and
CH2(main) is the intensity of all methylene units arising
from 1,2-insertions. We were unable to identify reso-
nances corresponding to 1,3-misinsertions (C3′ ) 0) for
these samples. However, using the the peak assign-
ments shown in Figure 2, we calculate the amount of
2,1-insertion to be 2.2% for catalyst 1 (entry 2), 3.6%
for catalyst 2 (entry 4), 2.8% for catalyst 3 (entry 6),
and 4.7% for catalyst 4 (entry 8).
Discu ssion
Productivities for entries 1-14 are low relative to
most metallocene catalysts for propylene polymeriza-
tion, but are similar to those reported by Canich.3 The
low productivities are likely a consequence of low Al/
M, as indicated by the increased productivities for
entries 13 and 14 relative to entries 5 and 6. The
Me4C5-based catalysts (entries 1-4) are on average
more active than the Ind (entries 5-10) and Flu (entries
11,12) analogues.
The appropriate choice of cyclopentadienyl ligand
appears to have the most influence on the resultant
polymer molecular weight. Me4C5-based catalysts 1 and
2 and Flu-based 6 give polymers with higher molecular
weights (>250 000) than Ind-based catalysts. The
amino group appears to have no clear effect on Mw
(entries 1 and 2 vs 3 and 4, 5 and 6 vs 7 and 8).
Polymer microstructures for entries 1-14 were de-
termined by 13C NMR. In all cases, except for entries
5 and 6, the polymers were almost completely atactic,
with a slight syndiotactic preference.24 The similarity
in tacticity for polymerizations at 50 psig vs those in
the bulk monomer suggest that there is no strong
tacticity dependence on monomer concentration.25
The CpA catalyst has proven to be very useful due to
its ability to copolymerize ethylene with R-olefins.
These catalysts are especially attractive because they
are able to operate at high temperatures without
significant drops in productivity and molecular weight.
For example, Stevens reports5 that an ethylene/1-octene
copolymer produced at 160 °C with [(Me4C5)SiMe2NtBu]-
TiCl2 and MAO had Mw ) 53 000 and productivity )
27 000 kg/mol × Ti. One important commercial mate-
rial made from these catalysts is a linear low-density
polyethylene (LLDPE) with the processing character-
istics of low-density polyethylene (LDPE).28 The ease
with which these catalysts incorporate R-olefins appears
to be a consequence of the sterically accessible nature
of the coordination site.
One might predict that the open nature of these
catalysts would limit stereospecificity in R-olefin poly-
merization. We found that catalysts 5 and 6 (entries
11-14), reported to be isospecific3 under these condi-
tions, gave polymers which were almost completely
atactic with a bias toward syndiotactic. Cs-symmetric
(26) Cheng, H. N.; Ewen, J . A. Makromol. Chem. 1989, 190, 1931-
1943.
(24) In addition, DSC analysis showed a complete lack of crystal-
linity for entries 1-4, 11, and 12.
(25) Resconi, L.; Fait, A.; Piemontesi, F.; Colonnesi, M.; Rychlicki,
H.; Zeigler, R. Macromolecules 1995, 28, 6667-6676.
(27) Tsutsui, T.; Ishimaru, N.; Mizuno, A.; Toyota, A.; Kashiwa, N.
Polymer 1989, 30, 1350-1356.
(28) Lai, S.; Yaw; Wilson, J . R.; Knight, G. W.; Stevens, J . C. (Dow
Chemical) U.S. Patent 5,278,272, 1994.

2882 Organometallics, Vol. 16, No. 13, 1997
Ta ble 2. P r op ylen e P olym er iza tion u sin g Ca ta lysts 1-6a
McKnight et al.
d
d
entry
catalyst
[catalyst] (µM) P (psig) Al/M
yield (g) productivityb
mmmmc rrrrc Mw (×10-3
)
Mw/Mn
% 2,1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
1
1
2
2
3
3
4
4
5
5
6
6
3
3
38
38
34
34
39
39
34
34
36
36
46
46
39
39
50
100e
50
289
289
350
350
350
350
350
350
400
400
350
350
1000
1000
3.10
2.89
1.11
1.28
0.22
0.15
0.25
1.33
0.23
1.72
0.14
0.11
0.79
1.48
820
380
325
188
56
19
72
196
63
239
31
2.7
1.0
2.8
17.7
23.3
7.9
10.4
5.0
583
400
820
269
118
37
130
111
130
165
652
779
130
133
2.01
1.68
1.77
3.55
1.83
2.29
2.80
2.04
3.55
2.20
1.78
1.94
1.96
2.07
2.2
3.6
2.8
4.7
100e
50
4.3
16.3
14.5
6.2
5.5
5.7
4.6
1.3
3.4
12.1
12.2
100e
50
7.4
11.2
12.9
11.9
7.5
14.5
17.8
6.1
100e
50
100e
50
120f
50
17
203
190
0
100e
8.5
a
b
Polymerizations were carried out in toluene with MAO at 30 °C ((1 °C) for 1 h. Productivity in kg PP/mol M × h. c Determined by
13C NMR spectrum. Determined by gel-permeation chromatography. e Pressure of a mixture of 100 mL of toluene and 100 mL of propylene.
d
f Pressure of a mixture of 50 mL of toluene and 100 mL of propylene.
ments to be much less hindered for CpA complexes than
for Cp2M systems. The CpA complexes also have a
higher calculated barrier to monomer insertion (ca. 5
kcal) than their metallocene counterparts. Polymer
chain rotation about the metal center between inser-
tions could lead to atactic polymerization in a Cossee-
Arlman polymerization mechanism.31
Statistical analysis of these polymers revealed that
the Bernoullian model for chain-end control is the
dominant propagation mechanism for all catalysts
except 3. Polymers derived from Me4C5- and Flu-based
catalysts have Bernoullian indices (â ) 4[mm][rr]/[mr]2)
which are very close to 1.0, indicating chain-end control,
while the indenyl analogues show slightly higher â
values (see Supporting Information). Polymers from 3
have average â values equal to 2.32, indicating some
small influence of this ligand on the stereospecificity of
polymerization.
The accessible coordination site of these catalysts also
leads to lower regiospecificity in propylene polymeriza-
tion. A notable exception is the fluorenylzirconium
catalyst 6, which yields highly regioregular, atactic
polypropylene. At this time, it is not clear if the higher
regiospecificity of 6 is a consequence of the fluorenyl
F igu r e 2. 13C NMR spectra of polymer derived from (a)
catalyst 4 (entry 12) and (b) catalyst 6 (entry 8).
ligand or the nature of the metal atom. Attempts to
compare metal effects were complicated by our inability
to synthesize either indenylzirconium or fluorenyl-
titanium analogues. Also, it is known5 that a zirconium
analogue of 1 shows very low activity in propylene
polymerization.
In Figure 4, the effects of the Cp and amido ligand
on molecular weight, productivity, isospecificity, and
regiospecificity are summarized. It is apparent that
Me4C5-based catalysts (1, 2) generally have higher
molecular weights and productivities than their indenyl
analogues (3, 4) under these polymerization conditions.
F igu r e 3. Representation of regioerrors.
This is likely the effect of the increased electron-
1 and 6 could be expected on symmetry grounds to be
donating character of the Me4C5 ligand, which may lead
to a decrease in chain termination (i.e., â-H elimination)
and/or increased stabilization of the active species
toward reduction. The effect of the amido ligand on Mw
and productivity is inconclusive; however, in the indenyl
series, where the electronic effect of the Cp ligand may
be minimized, the less sterically encumbered secondary
amides (4, 5) tend to be more productive and give
syndiospecific much like Me2C(Flu)(Cp)ZrCl2.29 Of the
catalysts studied, 1 and 6 gave polymers with the
highest [rrrr] values, which are nevertheless quite
modest (14 < [rrrr] < 23, entries 1, 2, 11, 12). Ziegler30
has studied simulated M-CR rotation processes for
Ziegler-Natta catalysts, and he found these rearrange-
(29) Ewen, J . A.; J ones, R. L.; Razavi, A.; Ferrara, J . D. J . Am. Chem.
Soc. 1988, 110, 6255.
(30) Fan, L.; Harrison, D.; Woo, T. K.; Ziegler, T. Organometallics
1995, 14, 2018-2026.
(31) Arlman, E. J .; Cossee, P. J . Catal. 1964, 3, 99.

Group 4 Cp-Amido Complexes
Organometallics, Vol. 16, No. 13, 1997 2883
drofuran, benzene-d6, tetrahydrofuran-d8, and toluene-d8 were
distilled from sodium/benzophenone ketyl. Methylene chlo-
ride, chloroform, and chloroform-d were distilled from calcium
hydride. Deuterated solvents were obtained from Cambridge
Isotopes Labs.
Butyllithium, dimethyldichlorosilane, isopropylmagnesium
chloride, fluorene, PbCl2, and TiCl4(THF)2 were obtained from
Aldrich and used as received. ZrCl4 was obtained from Fluka
and used as received. Indene (Wiley) was dried over 3 Å
Molecular sieves and distilled. tert-Butylamine, (S)-(-)-R-
methylbenzylamine, and cyclohexylamine were obtained from
Aldrich and distilled from calcium hydride. (Me4C5H)SiMe2-
Cl,5 [(Me4C5)SiMe2NtBu]Mg2Cl2(THF)2,5 (Ind-H)SiMe2Cl,5
[(Ind)SiMe2NtBu]Li2,5 [η5:η1-(Ind)SiMe2N(C6H11)]TiCl2 (5),3
37
[FluSiMe2NtBu]Li2,16 TiCl3(THF)3,36 and Ti(NMe2)4 were
prepared according to literature procedures. [η5:η1-(Ind)-
SiMe2N(C6H11)]TiCl2 (5) was also prepared by a modification
of the procedure by Herrmann22 (see below).
1H NMR were recorded on Varian Gemini 200, Gemini 300,
and XL-400 spectrometers and were referenced relative to
TMS, while 13C NMR were recorded at 100 MHz on a Varian
XL-400 NMR spectrometer. Inverse-gated decoupled 13C NMR
spectra of PP samples 2, 4 , 6, and 8 were recorded on a Varian
Unity Plus 500 spectrometer at 125.8 MHz. Optical rotations
were measured on a J ASCO DIP-360 digital polarimeter.
Elemental Analyses were performed by Desert Analytics or
E+R Microanalytics Laboratories.
P r op ylen e P olym er iza tion s. Polymerizations were car-
ried out in a 450 mL stainless steel Parr reactor equipped with
a magnetic stirrer. A cooling mixture (ethylene glycol/water)
was passed through the reactor in conjunction with a heating
mantle to maintain the temperature within (1 °C.
F igu r e 4. Summary of ligand effects on PP molecular
weight, productivities, isospecificities, and regiospecificities
of Cp-amido catalysts (Prod ) productivity in kg PP/mol
M × h).
slightly higher molecular weight PP than the tertiary
amido catalyst (3).
At the same time, the overall trend toward atactic
microstructures and the low mmmm values (< 6%)
indicate that neither Cp nor amido ligands have a
significant effect on the stereospecificity. While most
of the catalysts studied tend toward a slight syndiotactic
bias, only chiral indenyl complex 3 shows enhancements
in mm. Perhaps in this case, the combination of sterics
from a more encumbered amine and a less-demanding
indenyl ring gives rise to a slight energy difference
between the two sites available for monomer coordina-
tion or the polymer chain. Catalysts of a similar
symmetry have been found to give hemiisotactic32 or
stereoblock33-35 polypropylene.
Finally, while differences in the number of regioerrors
are small, there does appear to be a noticeable influence
of the ligand environment on the regiospecificity of these
titanium catalysts. Quantitative 13C NMR reveals that
the more sterically hindered Me4C5-based catalysts (1
and 2) produce polymers with about 0.5-1.0% less 2,1-
insertion than their indenyl analogues (3 and 4). The
amino group has a slightly greater effect, with NtBu-
based catalysts (1 and 3) leading to 1.5-2.0% less
misinsertion than the NCHMePh analogues (2 and 4).
Catalyst solutions were prepared by treating a toluene
solution of the desired metal complex with solid MAO (Akzo,
Type 4A), diluting to 25 mL with toluene, and stirring for 20
min. Meanwhile, toluene (75 mL) was injected into the reactor
with a propylene back-pressure and then warmed to 30 °C.
Propylene was then introduced as a gas for reactions at 50
psig or as a liquid (100 mL) for reactions at 100 psig. For entry
12, only 50 mL of toluene was used to give a final pressure of
120 psig. After equilibrating the reactor for 20 min, the
catalyst solution was injected with either a propylene back-
pressure (50 psig reactions) or an argon back-pressure (bulk
reactions). The reactor was stirred for 1 h with the temper-
ature maintained within (1 °C. Polymerizations were termi-
nated by injecting MeOH (20 mL) and venting for ca. 10 min.
Polymers were precipitated by pouring the slurry into acidic
MeOH (300 mL) and stirring overnight. Finally, the polymer
was filtered, washed with MeOH, and then dried at 50-60 °C
in a vacuum oven.
P olym er Ch a r a cter iza tion s. GPC measurements were
carried out on 0.2% w/v filtered samples using Waters R401
with a Polymer Laboratories Mixed C (PLGel 5 µm) × 2
column set at 40 °C in THF. 13C NMR spectra of polypropylene
samples (100 mg) were obtained in C2D2Cl4 (0.5 mL) at 100
°C. A spectral width of 80 ppm was used with a minimum of
6000 transients acquired for each spectrum. For the inverse-
gated decoupled 13C NMR spectra of PP 2 and 10, a minimum
of 6400 transients were obtained with pulse width ) 90° and
relaxation delay. Chemical shifts were referenced to solvent
(74.12 ppm).
Exp er im en ta l Section
Gen er a l Con d ition s. All experiments involving air-sensi-
tive compounds were performed under nitrogen in a Vacuum
Atmospheres drybox or under argon using standard Schlenk
line techniques. Hydrocarbon solvents, diethyl ether, tetrahy-
[η5:η1-(Me4C5)SiMe2Nt Bu ]TiCl2 (1).
A
mixture of
[(Me4C5)SiMe2NtBu]Mg2Cl2(THF)2 (2.00 g, 3.90 mmol) and
TiCl3(THF)3 (1.46 g, 3.94 mmol) were treated at -78 °C with
THF (60 mL) with shaking. The light blue suspension was
warmed to room temperature, darkening quickly to a deep red
solution. After the mixture was stirred for 10 min, PbCl2
(0.548 g, 1.97 mmol) was added as a solid and lead precipitated
immediately. THF was removed in vacuo after 8 h and the
(32) Farina, M.; Di Silvestro, G.; Sozzani, P. Macromolecules 1993,
26, 946.
(33) Chien, J . C. W.; Llinas, G. H.; Rausch, M. D.; Lin, G. Y.; Winter,
H. H.; Atwood, J . L.; Bott, S. G. J . Am. Chem. Soc. 1991, 113, 8569-
8570.
(34) Gauthier, W. J .; Corrigan, J . F.; Taylor, N. J .; Collins, S.
Macromolecules 1995, 28, 3771-3778.
(35) Gauthier, W. J .; Collins, S. Macromolecules 1995, 28, 3779-
3786.
(36) Manzer, L. Inorg. Synth. 1982, 135.
(37) Bradley, D. C.; Thomas, I. M. J . Chem. Soc. 1960, 3857.

2884 Organometallics, Vol. 16, No. 13, 1997
McKnight et al.
residue was extracted with pentane (2 × 30 mL). Concentrat-
ing to 30 mL and cooling to -78 °C for 2 h gave a yellow
microcrystalline solid (4). Yield: 0.767 g (53.4%). 1H NMR
(C6D6): δ 2.00 (s, 6H, CH3), 1.99 (s, 6H, CH3), 1.42 (s, 9 H,
NC(CH3)3), 0.43 (s, 6H, Si(CH3)2). 13C NMR (C6D6): δ 140.8,
138.1, 121.6, 62.4, 32.9, 16.4, 13.3, 5.6. Anal. Calcd for
7.34 (d, 1H, Ind-H) 7.32-7.28 (m, 2H, Ar-H), 6.58 (d, 1H,
Ind-H), 1.39 (s, 9H, NC(CH3)3), 0.95 (s, 3H, SiCH3), 0.69 (s,
3H, SiCH3). 13C NMR (CDCl3): δ 135.9, 134.7, 129.1, 128.4,
128.3, 127.4, 126.3, 119.3, 98.4, 63.3, 32.2, 3.3, 0.9. Anal. Calcd
for C15H21Cl2NSiTi: C, 49.73; H, 5.84; N, 3.87. Found: C,
48.84; H, 5.99; N, 3.51.
C
15H27Cl2NSiTi: C, 48.92; H, 7.39; N, 3.80. Found: C, 49.34;
(In d -H )SiMe2-(S)-NH CH MeP h . A solution of (Ind-
H)SiMe2Cl (3.43 g, 16.4 mmol) in THF (50 mL) was treated at
room temperature with (S)-(-)-R-methylbenzylamine (5.0 mL,
39 mmol) over a period of 1 min. A dense white precipitate
formed immediately, and after adding additional THF (50 mL),
the slurry was stirred overnight. Removal of THF in vacuo
gave a white paste, which was extracted with hexane (3 × 25
mL). Concentrating the extract in vacuo and distillation (bp
130-140 °C/0.005 mmHg) gave a colorless oil, which was a
mixture of three regioisomers. Yield: 4.24 g (88%). 1H NMR
(CDCl3): δ 7.10-7.58 (m, 9H, Ar-H), 6.93 (d, 1H, Ind-H),
6.55-6.65 (m, 2H, Ind-H), 3.86, 3.94, 4.08 (m, 1H, NCH), 3.18,
3.49 (s, 1H, Ind-H), 1.20, 1.21, 1.29 (d, 3H, CH3), 0.95, 0.68
(m, 1H, NH), 0.32, -0.09, -0.14 (d, 6H, Si(CH3)). 13C NMR
(CDCl3): δ 148.7, 145.5, 144.9, 144.3, 135.8, 128.9, 128.2, 128.1,
126.4, 126.3, 126.2, 126.0, 125.9, 125.7, 124.6, 124.2, 123.6,
123.5, 122.8, 122.3, 120.9, 51.7, 51.6, 48.3, 40.7, 27.9, -0.5,
-2.5, -2.6, -2.9. Anal. Calcd for C19H23NSi: C, 77.76; H,
7.90; N, 4.77. Found: C, 78.02; H, 7.83; N, 5.77.
[(In d )SiMe2-(S)-NCHMeP h ]Li2. A solution of (S)-(Ind-
H)SiMe2NHCHMePh (4.24 g, 14.4 mmol) in Et2O (60 mL) was
slowly treated with butyllithium (20 mL, 1.6 M in hexane) with
stirring at room temperature. After the yellow-orange solution
was stirred overnight, solvents were removed in vacuo to give
a brittle glassy solid. Slurrying with hexane (50 mL) for 30
min and filtering gave an off-white solid, which was dried in
vacuo. Yield: 4.08 g (93%).
(+)-(R)-[η5:η1-(In d )SiMe2-(S)-NCHMeP h ]TiCl2 (4). A.
Meta th esis Rou te. A mixture of [(Ind)SiMe2-(S)-NCHMePh]-
Li2 (0.935 g, 3.06 mmol) and TiCl4(THF)2 (1.022 g, 3.06 mmol)
was treated at -78 °C with toluene (8 mL) precooled to -78
°C. The red suspension was slowly warmed to room temper-
ature, stirred overnight, and filtered through Celite. Removing
toluene in vacuo from the filtrate gave a red oil, which was
subsequently treated with Et2O (8 mL) at -20 °C. After the
mixture was stirred at -20 °C for 10 min, a red-brown solid
was isolated by decanting off the red supernatant and then
drying in vacuo. This solid was recrystallized by treating with
Et2O (5 mL) and CH2Cl2 (5 mL), heating to dissolve the solid,
and then slowly cooling to -40 °C overnight. This was further
cooled to -78 °C for 2 h, and deep red block crystals (4) were
isolated by decanting off the supernatant and drying in vacuo.
Yield: 0.065 g (5.2%).
H, 7.57; N, 3.78.
(S)-(Me4C5H)SiMe2NHCHMeP h . A solution of (S)-(-)-R-
methylbenzylamine (1.50 g, 12.4 mmol) in THF (10 mL) was
added over 2 min to a solution of (Me4C5H)SiMe2Cl (0.88 g,
4.11 mmol) in THF (20 mL) at room temperature. A dense
white precipitate formed immediately. After the mixture was
stirred for 24 h, THF was removed in vacuo and the resulting
residue was extracted with pentane (3 × 10 mL). Removal of
pentane in vacuo at 60 °C afforded 5, a pale yellow oil (1.18 g,
96%). 1H NMR (CDCl3): δ 7.4-7.2 (m, 5H, Ar-H), 4.05 (m,
1H, NCH(CH3), 2.00 (s, 3H, C5CH3), 1.94 (s, 3H, C5CH3), 1.83
(s, 3H, C5CH3), 1.82 (s, 3H, C5CH3), 1.36 (s, 3H, NCH(CH3)),
0.80 (d, 1H, NH), -0.05 (s, 3H, SiCH3), -0.06 (s, 3H, SiCH3).
13C NMR (CDCl3): δ 151.0, 149.2, 135.5, 128.1, 126.2, 125.9,
56.6, 51.5, 28.0, 14.6, 11.2, -1.6, -1.7. Anal. Calcd for
C
15H27Cl2NSiTi: C, 76.18; H, 9.76; N, 4.68. Found: C, 75.83;
H, 9.59; N, 6.04.
[(S)-(Me4C5)SiMe2NCHMeP h ]Mg2Cl2(THF ). Isopropyl-
magnesium chloride (4.0 mL, 2.0 M in Et2O) was dried in vacuo
to remove Et2O. The Grignard reagent was treated with a
solution of (S)-(Me4C5H)SiMe2NHCHMePh (1.18 g, 3.93 mmol)
in THF (4 mL) and toluene (17 mL). The solution was refluxed
for 36 h, during which some solids formed. Solvents were
removed in vacuo to give a brittle glass, which was then
slurried with pentane (20 mL) for 30 min. After the mixture
was filtered, the white solid was dried under vacuum. Yield:
1.70 g (88%). THF coordination was confirmed by 1H NMR in
THF-d8.
[η5:η1-(Me4C5)SiMe2-(S)-NCHMeP h ]TiCl2 (2). A suspen-
sion of TiCl3(THF)3 (1.13 g, 3.04 mmol) in THF (20 mL) was
treated at -78 °C with
a solution of [(S)-(Me4C5)SiMe2-
NCHMePh]Mg2Cl2(THF) (1.50 g, 3.07 mmol) in THF (20 mL)
with shaking. The blue-green suspension was warmed to room
temperature, darkening quickly to a deep yellow-red color.
After the mixture was stirred for 15 min, PbCl2 (0.427 g, 1.54
mmol) was added as a solid. The solution lightened slightly
to yellow-brown, and lead precipitated over the course of 15
min. THF was removed in vacuo, and the brownish-yellow
residue was exhaustively extracted with warm hexane (150
mL total). The extract was concentrated to 15 mL to give a
yellow microcrystalline solid. Yield: 0.195 g (15%). 1H NMR
(CDCl3): δ 7.4-7.3 (m, 2H, Ar-H), 7.2-7.1 (m, 3H, Ar-H),
6.22 (q, 1H, J ) 7 Hz, NCH), 2.14 (s, 3H, CH3C5), 2.13 (s, 3H,
CH3C5), 2.07 (s, 3H, CH3C5), 1.99 (s, 3H, CH3C5), 1.67 (d, 3H,
CH3), 0.45 (s, 3H, SiCH3), -0.18 (s, 3H, SiCH3). 13C NMR
(CDCl3): δ 144.5, 140.8, 140.6, 136.1, 128.4, 128.2, 127.9, 127.5,
126.9, 103.5, 62.0, 19.5, 15.8, 12.9, 12.8, 4.5, 2.0. Anal. Calcd
for C19H27Cl2NSiTi: C, 54.81; H, 6.54; N, 3.37. Found: C,
53.99; H, 6.57; N, 2.86. [R]589 ) -190.5° (c ) 1.3 mg/mL,
CH2Cl2).
B. Am in e Elim in a tion /Me3SiCl Ch lor in a tion . A solu-
tion of Ti(NMe2)4 (3.79 g, 16.9 mmol) in toluene (100 mL) was
treated with a toluene solution (50 mL) of (S)-(Ind-H)SiMe2-
NHCHMePh (5.00g, 17.0 mmol) at -78 °C. The mixture was
slowly warmed to room temperature and then brought to reflux
with argon flow through an oil bubbler. The reaction was
1
monitored by H NMR and was completely converted after 3
days. Toluene was removed in vacuo to give a yellow-brown
residue. 1H NMR analysis revealed the presence of two
diastereomers (A, B) in approximately 1.33/1 ratio. 1H NMR
(C6D6): δ 7.83 (d, 1H, Ar-H, A), 7.74 (d, 1H, Ar-H, B), 7.43
(d, 2H, Ar-H, A), 7.31 (d, 2H, Ar-H, B), 7.22-6.82 (m, 2H,
Ar-H), 6.73 (d, 1H, Ind-H, B), 6.71 (d, 1H, Ind-H, A), 6.39
(d, 1H, Ind-H, A), 6.34 (d, 1H, Ind-H, B), 5.20 (q, 1H, NCH,
A), 5.08 (q, 1H, NCH, B), 3.17 (s, 6H, N(CH3)2, B), 2.87 (s, 6H,
N(CH3)2, A), 2.39 (s, 6H, N(CH3)2, A), 2.37 (s, 6H, N(CH3)2,
B), 1.45 (d, 3H, CH3, A), 1.41 (d, 3H, CH3, B), 0.71 (s, 3H,
SiCH3, A), 0.45 (s, 3H, SiCH3, B), 0.42 (s, 3H, SiCH3, A), 0.30
(s, 3H, SiCH3, B).
[η5:η1-(In d )SiMe2Nt Bu ]TiCl2 (3).38
A
mixture of
[(Ind)SiMe2NtBu]Li2 (0.462 g, 1.80 mmol) and TiCl4(THF)2
(0.603 g, 1.81 mmol) were treated at -78 °C with precooled
-78 °C toluene (25 mL). After slowly warming the mixture
to room temperature, the red suspension was stirred 24 h and
then filtered. Removal of toluene from the supernatant in
vacuo gave a sticky dark red solid. The solid was recrystallized
from a mixture of hexane (10 mL) and toluene (5 mL) by slowly
cooling to -50 °C overnight, followed by cooling to -78 °C for
2 days. The supernatant was decanted off at -78 °C, and the
resulting red microcrystalline solid (4) was washed with -78
1
°C hexane and then dried in vacuo. Yield: 0.087 g (13%). H
The yellow-brown residue was dissolved in CH2Cl2 (150 mL)
and then treated with excess trimethylsilyl chloride (6.43 mL,
50.7 mmol) at room temperature. The color changed to dark
red, and this was stirred overnight. All volatiles were then
NMR (CDCl3): δ 7.76 (m, 2H, Ar-H), 7.44 (m, 1H, Ar-H),
(38) Amor, F.; Okuda, J . J . Organomet. Chem. 1996, 520, 245.

Group 4 Cp-Amido Complexes
Organometallics, Vol. 16, No. 13, 1997 2885
removed in vacuo to give a copious dark red powder, which
was washed with Et2O (75 mL, then 3 × 30 mL) and dried in
vacuo. This powder was then stirred in CH2Cl2 (50 mL) for
30 min, and Et2O (75 mL) was added. After the mixture was
cooled at -20 °C overnight, deep red crystals (4) were isolated
by decanting off the supernatant and drying in vacuo. Yield:
2.73 g (39.3% based on Ti(NMe2)4, 68.9% based on intermediate
diastereomer A). 1H NMR (C6D6): δ 7.53 (d, 1H, Ind-H), 7.4-
6.7 (m, 9H, Ar-H), 6.31 (q, 1H, NCH), 6.18 (d, 1H, Ind-H),
1.41 (d, 3H, CH3), 0.42 (s, 3H, SiCH3), -0.36 (s, 3H, SiCH3).
13C NMR (C6D6): δ 144.3, 135.4, 134.7, 128.9, 128.8, 128.4,
127.3, 127.1, 126.6, 116.9, 98 (low intensity, tentative assign-
ment), 64.0, 19.0, 2.4, -2.2. Anal. Calcd for C19H21Cl2NSiTi:
C, 55.62; H, 5.16; N, 3.41. Found: C, 55.11; H, 5.18; N, 3.29.
[R]589 ) +402° (c ) 0.95 mg/mL, CH2Cl2).
127.1, 126.2, 116.7, 99 (low intensity, tentative assignment),
65.7, 34.3, 26.1, 25.6, 2.6, 0.4. Anal. Calcd for C17H23Cl2-
NSiTi: C, 52.59; H, 5.97; N, 3.61. Found: C, 52.14; H, 5.84;
N, 3.60.
[η5:η1-F lu SiMe2Nt Bu ]Zr Cl2 (6).
A
suspension of
[FluSiMe2NtBu]Li2 (1.00 g, 3.25 mmol) in toluene (100 mL)
was treated with a slurry of ZrCl4 (0.758 g, 3.25 mmol) in
toluene (50 mL) at room temperature. The resulting bright
yellow suspension darkened on stirring. After the mixture was
stirred overnight, the suspension was filtered and the solids
were washed with hot toluene (4 × 20 mL). The combined
orange filtrates were concentrated in vacuo to 20 mL, and then
hexane (75 mL) was added. Cooling to -20 °C overnight and
then to -78 °C for 1 day yielded a crop of yellow microcrys-
talline solid. Yield: 0.275 g (18.6%). 1H NMR (toluene-d8):
δ 7.82-7.69 (m, 4H, Ar-H), 7.18-7.10 (m, 2H, Ar-H), 7.00
(d, 2H, Ar-H), 1.27 (s, 9H, NC(CH3)3), 0.68 (s, 6H, Si(CH3)2).
13C NMR: the poor solubility in deuterated hydrocarbons and
decomposition in CDCl3 led to characterization as the mono-
THF adduct.16 Anal. Calcd for C19H23Cl2NSiZr: C, 50.09; H,
5.09; N, 3.07. Found: C, 50.09; H, 5.04; N, 2.83.
[η5:η1-(In d )SiMe2N(C6H11)]TiCl2 (5). Am in e Elim in a -
tion /Me3SiCl Ch lor in a tion . A solution of Ti(NMe2)4 (0.41
g, 1.84 mmol) in toluene (15 mL) was treated with (Ind-H)-
SiMe2NH(C6H11) (0.50 g, 1.84 mmol) at -78 °C. The mixture
was slowly warmed to room temperature and then brought to
reflux. The reaction was monitored by 1H NMR and was
completely converted after 5 days. Toluene was removed in
vacuo to give a red residue. 1H NMR (C6D6): δ 7.82 (d, 1H,
Ar-H), 7.43 (d, 1H, Ar-H), 6.98-6.85 (m, 2H, Ar-H), 6.71
(d, 1H, Ind-H), 6.33 (d, 1H, Ind-H), 3.80 (m, 1H, NCH), 3.17
(s, 6H, N(CH3)2), 2.41 (s, 6H, N(CH3)2), 1.85-1.00 (m, 11H,
C6H11), 0.77 (s, 3H, SiCH3), 0.52 (s, 3H, SiCH3).
The red residue was dissolved in CH2Cl2 (10 mL) and then
treated with excess trimethylsilyl chloride (0.51 mL, 4.0 mmol)
at room temperature. After the mixture was stirred overnight,
all volatiles were removed in vacuo to give an orange red
powder, which was washed with pentane (30 mL) and dried
in vacuo. Yield: 0.319 g (45% based on Ti(NMe2)4). 1H NMR
(C6D6): δ 7.53 (d, 1H, Ar-H), 7.27 (d, 1H, Ar-H), 7.01 (t, 1H,
Ar-H), 6.93 (t, 1H, Ar-H), 6.82 (d, 1H, Ind-H), 6.35 (d, 1H,
Ind-H), 4.97 (m, 1H, NCH), 2.00 (m, 1H, C6H11), 1.55 (m, 2H,
C6H11), 1.36 (m, 2H, C6H11), 1.16 (m, 2H, C6H11), 0.90 (m, 2H,
C6H11), 0.69 (m, 2H, C6H11), 0.47 (s, 3H, SiCH3), 0.29 (s, 3H,
SiCH3). 13C NMR (CDCl3): δ 135.0, 134.1, 128.7, 128.6, 128.2,
Ack n ow led gm en t. We gratefully acknowledge sup-
port from the NSF-NYI program (NSF-DMR 9258334-
003) and the Donors of the Petroleum Research Fund,
administered by the American Chemical Society. We
would also like to thank Professor Don Tilley and Dr.
Shane Mao for GPC analysis.
Su p p or tin g In for m a tion Ava ila ble: Text giving the
experimental details associated with data collection and
refinement and tables of crystal data, positional parameters,
anisotropic thermal factors, bond distances, bond angles, and
torsional angles for 4 and methyl pentad distributions for
entries 1-14 (12 pages). Ordering information is given on any
current masthead page.
OM9609628