Ethylene, styrene, and α-methylstyrene polymerization by mono(pentamethylcyclopentadienyl) (Cp*) complexes of titanium, zirconium, and hafnium: Roles of cationic complexes of the type [Cp*MR2]+ (R = alkyl) as both coordination polymerization catalysts and carbocationic polymerization initiators

Article abstract of DOI:10.1021/om9501945

A variety of monocyclopentadienyl and mono(pentamethylcyclopentadienyl) complexes of titanium, zirconium, and hafnium are assessed for abilities to initiate polymerization of ethylene, styrene, and, in part, α-methylstyrene. In general, little or no activity was found for either neutral species of the types CpMMe₃ and CpMMe₂OR or for cationic 12- and 14-electron species of the types [CpMR₂L]⁺ and [CpMR₂L₂]⁺, respectively (Cp = η⁵-cyclopentadienyl; R = alkyl; L = amine, phosphine ligands). In contrast, much better olefin polymerization initiators result from abstraction of a methyl carbanion from Cp*MMe₃ (Cp* = η⁵-pentamethylcyclopentadienyl) by B(C₆F₅)₃, a reaction which gives cationic, 10-electron species of the type [Cp*MMe2][BMe(C₆F₅)₃] . Of these, the complex [Cp*TiMe₂][BMe(C₆F₅)₃] (A) is an excellent initiator or initiator precursor for the polymerization of ethylene and styrene, resulting in high yields respectively of high molecular weight polyethylene and atactic (a-PS) and/or syndiotactic polystyrene (s-PS), depending on conditions; the tacticity of purified s-PS, as judged by ¹³C{¹H} NMR spectroscopy, approaches 100%. While the polymerization of ethylene probably involves a classical Ziegler-Natta process, polymerization of styrene to s-PS and a-PS apparently involves respectively a Ziegler-Natta process and carbocationic initiation. High yields of essentially syndiotactic poly(α-methylstyrene) are obtained by utilizing the same initiator system, also apparently via a carbocationic process.

Full text of DOI:10.1021/om9501945

Organometallics 1996, 15, 693-703
693
Eth ylen e, Styr en e, a n d r-Meth ylstyr en e P olym er iza tion
by Mon o(p en ta m eth ylcyclop en ta d ien yl) (Cp *) Com p lexes
of Tita n iu m , Zir con iu m , a n d Ha fn iu m : Roles of Ca tion ic
Com p lexes of th e Typ e [Cp *MR2]+ (R ) Alk yl) a s both
Coor d in a tion P olym er iza tion Ca ta lysts a n d
Ca r boca tion ic P olym er iza tion In itia tor s
Qinyan Wang, Ruhksana Quyoum,† Daniel J . Gillis,‡ Marie-J ose´ Tudoret,
Dusan J eremic, Brian K. Hunter, and Michael C. Baird*
Department of Chemistry, Queen’s University, Kingston, Ontario, Canada K7L 3N6
Received March 14, 1995X
A variety of monocyclopentadienyl and mono(pentamethylcyclopentadienyl) complexes of
titanium, zirconium, and hafnium are assessed for abilities to initiate polymerization of
ethylene, styrene, and, in part, R-methylstyrene. In general, little or no activity was found
for either neutral species of the types CpMMe3 and CpMMe2OR or for cationic 12- and 14-
electron species of the types [CpMR2L]+ and [CpMR2L2]+, respectively (Cp ) η5-cyclopen-
tadienyl; R ) alkyl; L ) amine, phosphine ligands). In contrast, much better olefin
polymerization initiators result from abstraction of a methyl carbanion from Cp*MMe3 (Cp*
) η5-pentamethylcyclopentadienyl) by B(C6F5)3, a reaction which gives cationic, 10-electron
species of the type “[Cp*MMe2][BMe(C6F5)3]”. Of these, the complex [Cp*TiMe2][BMe(C6F5)3]
(A) is an excellent initiator or initiator precursor for the polymerization of ethylene and
styrene, resulting in high yields respectively of high molecular weight polyethylene and
atactic (a-PS) and/or syndiotactic polystyrene (s-PS), depending on conditions; the tacticity
of purified s-PS, as judged by 13C{1H} NMR spectroscopy, approaches 100%. While the
polymerization of ethylene probably involves a classical Ziegler-Natta process, polymeri-
zation of styrene to s-PS and a-PS apparently involves respectively a Ziegler-Natta process
and carbocationic initiation. High yields of essentially syndiotactic poly(R-methylstyrene)
are obtained by utilizing the same initiator system, also apparently via a carbocationic
process.
In tr od u ction
studied class of homogeneous catalysts for the poly-
merization of olefins.
There has in recent years been extensive research into
the utilization of titanocene, zirconocene, and hafnocene
derivatives as homogeneous catalysts for olefin polym-
erization,1 and it is now realized that the best catalysts
incorporate the following structural features: the pres-
ence of a coordinated alkyl ligand, of a vacant site, and
of a positive charge. For instance, 16-electron complexes
of the type [Cp′2MR(L)]+ (M ) Ti, Zr, Hf; Cp′ )
substituted η5-cyclopentadienyl group; R ) alkyl group;
L ) labile ligand) form probably the most extensively
In contrast, although it might be anticipated that
electronically less saturated, sterically less hindered
monocyclopentadienyl complexes of the general stoichi-
ometries [Cp′MRL2]2+ or [Cp′MR2L]+ might behave as
even more reactive catalysts, in fact until recently
relatively few publications have described the utilization
of monocyclopentadienyl compounds of titanium, zirco-
nium, and hafnium as olefin polymerization initiators.2
(2) (a) Skupkinski, W.; Cieslowska-Glinska, I.; Wasielewski, A. J .
Mol. Catal. 1985, 33, 129. (b) Ishihara, N.; Kuramoto, M.; Uoi, M.
Macromolecules 1988, 21, 3356. (c) Zambelli, A.; Oliva, L.; Pellecchia,
C. Macromolecules 1989, 22, 2129. (d) Chien, J . C. W.; Wang, B.-P. J .
Polym. Sci.: Part A 1989, 27, 1539. (e) Chien, J . C. W.; Wang, B.-P. J .
Polym. Sci.: Part A 1990, 28, 15. (f) Rieger, B. J . Organomet Chem.
1991, 420, C17. (g) Soga, K.; Park, J . R.; Shiono, T. Polym. Commun.
1991, 32, 310. (h) Park, J . R.; Shiono, T.; Soga, K. Macromolecules
1992, 25, 521. (i) Pellecchia, C.; Longo, P.; Proto, A.; Zambelli, A.
Makromol. Chem., Rapid Commun. 1992, 13, 265. (j) Pellecchia, C.;
Longo, P.; Proto, A.; Zambelli, A. Makromol. Chem., Rapid Commun.
1992, 13, 277. (k) Chien, J . C. W.; Salajka, Z.; Dong, S. Macromolecules
1992, 25, 3199. (l) Kucht, A.; Kucht, H.; Barry, S.; Chien, J . C. W.;
Rausch, M. D. Organometallics 1993, 12, 3075. (m) Soga, K.; Koide,
R.; Uozumi, T. Macromol. Chem., Rapid Commun. 1993, 14, 511. (n)
Ready, T. E.; Day, R. O.; Chien, J . C. W.; Rausch, M. D. Macromolecules
1993, 26, 5822. (o) Longo, P.; Proto, A.; Oliva, L. Macromol. Rapid
Commun. 1994, 15, 151. (p) Zambelli, A.; Pellecchia, C.; Oliva, L.;
Longo, P.; Grassi, A. Makromol. Chem. 1991, 192, 223. (q) Kucht, H.;
Kucht, A.; Chien, J . C. W.; Rausch, M. D. Appl. Organomet. Chem.
1994, 8, 393. (r) Pellecchia, C.; Immirzi, A.; Grassi, A.; Zambelli, A.
Organometallics 1993, 12, 4473. (s) Flores, J . C.; Chien, J . C. W.;
Rausch, M. D. Organometallics 1994, 13, 4140.
† Present Address: Department of Chemistry, The University of
Sheffield, Sheffield S3 7HF, U.K.
‡ Present address: Uniroyal Chemical Co., 280 Elm St., Building
S310, Naugatuack, CT 06770.
X Abstract published in Advance ACS Abstracts, December 1, 1995.
(1) See, for instance: (a) Kaminsky, W., Sinn, H., Eds. Transition
Metals and Organometallics as Catalysts for Olefin Polymerization;
Springer-Verlag: Berlin, 1988. (b) Quirk, R. P., Ed. Transition Metal
Catalyzed Polymerizations; Ziegler-Natta and Metathesis Polymeriza-
tions; Cambridge University Press: Cambridge, U.K., 1988. (c)
Mohring, P. C.; Coville, N. J . J . Organomet. Chem. 1994, 479, 1. (d)
Gupta, V. K.; Satish, S.; Bhardway, I. S. J . Macromol. Sci. 1994, C34,
439. (e) Brintzinger, H. H.; Fischer, D.; Mu¨lhaupt, R.; Rieger, B.;
Waymouth, R. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 1143. For
recent theoretical considerations at the density functional level, see:
(f) Woo, T. K.; Fan, L.; Ziegler, T. Organometallics 1994, 13, 2252. (g)
Weiss, H.; Ehrig, M.; Ahlrichs, R. J . Am. Chem. Soc. 1994, 116, 4919.
For counterion effects, see: (h) Chien, J . C. W.; Song, Y.; Rausch, M.
D. J . Polym. Sci.: Part A 1994, 32, 2387.
0276-7333/96/2315-0693$12.00/0 © 1996 American Chemical Society

694 Organometallics, Vol. 15, No. 2, 1996
Wang et al.
Interestingly, while very few reports concerning polym-
erization of styrene by metallocene systems have ap-
peared, a majority of the reports2b,c,j,l-r describing
monocyclopentadienylmetal catalysts, as well as a num-
ber of patents,3 are concerned largely with polymeriza-
tion of styrene to syndiotactic polystyrene (s-PS), a
material which is currently receiving considerable at-
tention4 because of its high melting point and low
solubility in organic solvents.4f
same borate anion to transition metal ions in a similar
fashion, have been reported.6
In aromatic solvents, on the other hand, arene com-
plexes of the type [Cp*MMe2(η6-arene)][MeB(C6F5)3] are
formed.5a Since the MeB(C6F5)3-, CH2Cl2, and arene
ligands are readily displaced, these complexes all behave
in solution as sources of the cationic, 10-electron species
[Cp*MMe2]+, which might be expected to behave as
excellent homogeneous olefin polymerization catalysts.
We now describe our findings on the use of [Cp*MMe2]+
and related species as catalysts/initiators for the po-
lymerization of ethylene, styrene, and R-methylstyrene,
and we provide evidence that this family of complexes
can operate as carbocationic polymerization initiators
as well as precursors of conventional Ziegler-Natta
polymerization catalysts. In particular, we show that
compound A is an excellent initiator or initiator precur-
sor for the polymerization of ethylene and styrene,
resulting in high yields respectively of high molecular
weight polyethylene and atactic (a-PS) and/or highly
stereoregular syndiotactic polystyrene (s-PS), depending
on conditions. While the polymerizatio of ethylene
probably involves a classical Ziegler-Natta process,
polymerization of styrene to s-PS and a-PS apparently
involves respectively a Ziegler-Natta process and car-
bocationic initiation.
We have previously demonstrated that treatment of
the compounds Cp*MMe3 (M ) Ti, Zr, Hf; Cp* ) η5-
pentamethylcyclopentadienyl) with B(C6F5)3 results in
methyl carbanion abstraction and formation of the
complexes [Cp*MMe2][MeB(C6F5)3], as shown in eq 1
for M ) Ti.5a
Cp*TiMe3 + B(C6F5)3 f [Cp*TiMe2][MeB(C6F5)3]
(1)
Similar results have been reported by others.2j When
M ) Ti, at least, the product appears to exist in
methylene chloride solution as the labile, methyl-
bridged species Cp*MMe2(µ-Me)B(C6F5)3 (A), in equi-
To our knowledge, A is the first soluble, metallocene-
like compound which can initiate polymerization of an
olefin via two fundamentally different mechanisms, all
others being considered (albeit not always demon-
strated) to behave only as single-site initiator species.
Aspects of this work have appeared previously as a
communication,5b but we now retract the earlier con-
clusion5b that s-PS is formed via a carbocationic process.
librium probably with the solvent-separated, ionic com-
plex [Cp*MMe2(CH2Cl2)][MeB(C6F5)3].5a Analogous
metallocene compounds, with methyl groups linking this
Exp er im en ta l Section
All experiments were carried out under nitrogen using
standard Schlenk line techniques, a Vacuum Atmospheres
glovebox and dried, thoroughly deoxygenated solvents. 1H,
13C{1H}, and 19F NMR spectra were run using a Bruker AM
400 spectrometer operating at 400.14, 100.6, and 376.5 MHz,
respectively; 1H and 13C{1H} spectra are referenced with
respect to internal TMS using residual proton resonances or
carbon resonances, respectively, of the solvents, and 19F spectra
to external CFCl3. Spectra of polymers at high temperatures
were also run in 1,1,2,2-tetrachloroethane-d2 on a Bruker CXP-
200 NMR spectrometer. Differential scanning calorimetric
(DSC) measurements were recorded using a Mettler TA 3000
(3) For representative, recent patents describing catalyst systems
similar to those described here: (a) Takeuchi, M.; Kuramoto, M. U.S.
Patent 5,023,304, J un 11, 1991 (to Idemitsu Kosan Co., Tokyo, J apan).
(b) Campbell, R. E. U.S. Patent 5,066,741, Nov 19, 1991 (to The Dow
Chemical Co., Midland, MI). (c) Canich, J . A. M. U.S. Patent 5,096,
867, Mar 17, 1992 (to Exxon Chemical Patents, Inc., Linden, NJ ). (d)
Stevens, J . C.; Neithamer, D. R. U.S. Patent 5,132,380, J ul 21, 1992
(to The Dow Chemical Co., Midland, MI). (e) Campbell, R. E. European
Patent 597961, May 25, 1994 (to The Dow Chemical Co., Midland, MI).
(4) (a) Ishihara, N.; Seimiya, T.; Kuramoto, M.; Uoi, M. Macromol-
ecules 1986, 19, 2464. (b) Zambelli, A.; Longo, P.; Pellecchia, C.; Grassi,
A. Macromolecules 1987, 20, 2035. (c) Pellecchia, C.; Longo, P.; Grassi,
A.; Ammendola, P.; Zambelli, A. Makromol. Chem., Rapid Commun.
1987, 8, 277. (d) Grassi, A.; Longo, P.; Proto, A.; Zambelli, A.
Macromolecules 1989, 22, 104. (e) Kawamura, T.; Toshima, N.;
Matzusaki, K. Macromol. Rapid Commun. 1994, 15, 479. (f) Pasquon,
I.; Porri, L.; Giannini, U. Encycl. Polym. Sci. Technol. 1989, 15, 660.
(g) Longo, P.; Grassi, A.; Proto, A.; Ammendola, P. Macromolecules
1988, 21, 24. (h) Pellecchia, C.; Pappalardo, D.; Oliva, L.; Zambelli,
A. J . Am. Chem. Soc. 1995, 117, 6593. (i) Grassi, A.; Pellecchia, C.;
Oliva, L. Macromol. Chem. Phys. 1995, 196, 1093.
(5) (a) Gillis, D. J .; Tudoret, M.-J .; Baird, M. C. J . Am. Chem. Soc.
1993, 115, 2543. (b) Quyoum, R.; Wang, Q.; Tudoret, M.-J .; Baird, M.
C.; Gillis, D. J . J . Am. Chem. Soc. 1994, 116, 6435. (c) The PS reaction
products obtained in ref 5b were all extracted with acetone to separate
a-PS and s-PS, since it was not then realized that small amounts of
very high molecular weight a-PS present in several cases were not
soluble in acetone. The a-PS and s-PS in these materials were
therefore not completely separated but were detected as separate bands
in GPC measurements on supposed s-PS samples and mistakenly
taken to be relatively low molecular weight s-PS. (d) Barsan, F.; Baird,
M. C. J . Chem. Soc., Chem. Commun. 1995, 1065. (e) Wang, Q.; Baird,
M. C. Unpublished results. (f) J eremic, D.; Baird, M. C., Unpublished
results. (g) While the NMR tubes were baked in an oven at 110 °C,
we also find that bubbling water-saturated nitrogen through a solution
of Cp*TiMe3 in CD2Cl2 results in the appearances of new Ti-Me
resonances at δ 0.29, 0.20, and 0.08.
system or
a DuPont Instruments Model 912 differential
scanning calorimeter. Gel permeation chromatograms (GPC)
of the polymers were obtained at 145° C in 1,2,4-trichloroben-
zene using a Waters Model 150-C GPC or in THF at room
temperature using a DuPont Instruments Model 850 or a
Waters Model 440 liquid chromatograph; data were analyzed
using polystyrene calibration curves. Elemental analyses were
performed by Canadian Microanalytical Services or at E. I.
DuPont de Nemours, Central Research and Development,
Wilmington, DE.
The following compounds were prepared as described pre-
viously: Cp*MMe3 (M ) Ti,7a Zr,7b Hf7c), CpTiMe3,7d Cp*ZrPh3,7b
B(C6F5)3,7e Cp*Ti(CH2Ph)3,7a [(C5H4Me)2Fe][BPh4],7g [Cp*ZrMe2-
(THF)2][BPh4].7h
(6) (a) Yang, X.; Stern, C. L.; Marks, T. J . J . Am. Chem. Soc. 1991,
113, 3623. (b) Bochmann, M.; Lancaster, S. J .; Hursthouse, M. B.;
Abdul Malik, K. M. Organometallics 1994, 13, 2235. (c) Bochmann,
M.; Lancaster, S. J . Angew. Chem., Int. Ed. Engl. 1994, 33, 1634. (d)
Yang, X.; Sern, C. L.; Marks, T. J . J . Am. Chem. Soc. 1994, 116, 10015.

[Cp*MR2]+ Complexes
Organometallics, Vol. 15, No. 2, 1996 695
[HNEt3][B(C6F 5)4]. A solution of 0.35 g of triethylammo-
nium chloride (2.50 mmol) in distilled water was added to a
stirred solution of 1.62 g of Li[B(C6F5)4]7e (2.36 mmol) in water
at 0 °C. The reaction mixture was stirred for 1 h at room
temperature, and then the white precipitate was collected by
filtration, washed with water, and dried in vacuo for 20 h.
Yield: 1.57 g (85%). 1H NMR (CD2Cl2): δ 4.93 (br t, 1H, J NH
52.8 Hz, NH), 3.27 (dq, 6H, J 2.28, 7.32 Hz, CH2), 1.36 (t, 9H,
Me). 13C{1H} NMR (CD2Cl2): δ 148.7 (d, J CF 245.3 Hz), 138.8
(d, J CF 249.4 Hz), 136.9 (d, J CF 247.6 Hz), 128.2 (br, ipso C),
48.6 (CH2), 9.2 (Me). 19F NMR (CD2Cl2): δ -133 (d, o-F), -165
(t, m-F), -168 (t, p-F).
[P h 3C][MeB(C6F 5)3]. This compound was synthesized by
utilizing a procedure provided by Bochmann.7j To a solution
of 0.75 g of Li[MeB(C6F5)3] (1.5 mmol), prepared by reacting
equimolar amounts of MeLi and B(C6F5)3, in 20 mL of CH2Cl2
was added 0.45 g of Ph3CCl (1.5 mmol) in 10 mL of CH2Cl2.
After the mixture was stirred overnight at room temperature,
the resulting orange precipitate was collected by filtration,
washed with hexane, and dried in vacuo.
C, 61.03; H, 9.19. 1H NMR (benzene-d6): δ 5.91 (s, 5H, Cp),
1.14 (s, 9H, CMe), 0.63 (s, 6H, TiMe).
Cp TiMe(OBu t)2. CpTi(OBut)2Cl was prepared by treating
a solution of CpTiCl3 (2.58 g, 11.8 mmol) in 150 mL of ethyl
ether with tert-butanol (1.74 g, 23.6 mmol) and triethylamine
(2.38 mL, 23.6 mmol). The reaction mixture was stirred for
16 h and filtered through Celite, and the solvent was removed
from the filtrate to give a dark yellow oil. The latter was
sublimed at 45 °C onto a cold finger at -78 °C to give CpTi-
(OBut)2Cl as a yellow, crystalline solid. Yield: 2.71 g (78%).
1H NMR (benzene-d6): δ 6.21 (s, 5H, Cp), 1.18 (s, 18H, Me). A
solution of CpTi(OBut)2Cl (0.78 g, 2.65 mmol) in 25 mL of ethyl
ether was treated with 2.65 mmol of MeMgBr (0.88 mL of 3.0
M ethyl ether solution) at 0 °C. The reaction mixture was
stirred for 3 h and filtered through Celite, and the solvent was
then removed from the filtrate to give CpTi(OBut)2Me as a dark
green oil. Yield: 0.30 g (41%). Anal. Calcd for C14H26O2Ti:
C, 61.31; H, 9.55. Found: C, 61.58; H, 9.66. 1H NMR
(benzene-d6): δ 6.01 (s, 5H, Cp), 1.20 (s, 18H, CMe), 0.72 (s,
3H, TiMe).
Cp *TiMe2(OCP h 3). A solution of Cp*TiMe3 (0.114 g, 0.50
mmol) in 50 mL of pentanes was treated dropwise with Ph3-
COH (0.13 g, 0.50 mmol in 25 mL of 10:1 pentanes/toluene)
at -78 °C. The reaction mixture was allowed to warm to room
temperature over 2 h, and the solvent was removed under
reduced pressure to give a solid residue which was dissolved
in 5 mL of warm pentanes. On cooling of the solution to -15
°C for 16 h, Cp*TiMe2(OCPh3) was obtained as light yellow
crystals. Yield: 0.17 g (70%). Anal. Calcd for C31H36OTi: C,
78.80; H, 7.68. Found: C, 79.16; H, 7.75. 1H NMR (CD2Cl2):
δ 7.0-7.5 (m, 15H, Ph), 1.76 (s, 15H, CMe), 0.25 (s, 6H, TiMe).
Cp *Zr Me2(OCP h 3). A solution of Cp*ZrMe3 (0.272 g, 1.00
mmol) in 50 mL of pentanes was treated dropwise with Ph3-
COH (0.26 g, 1.00 mmol in 25 mL of 10:1 pentanes/toluene)
at -78 °C. The reaction mixture was allowed to warm to room
temperature over 4 h, and the solvent was removed under
reduced pressure to give a solid residue which was dissolved
in 5 mL of warm pentanes. On cooling of the solution to -15
°C for 16 h, Cp*ZrMe2(OCPh3) was obtained as off-white
crystals. Yield: 0.45 g (88%). Anal. Calcd for C31H36OZr: C,
72.18; H, 7.03. Found: C, 72.36; H, 7.13. 1H NMR (toluene-
d8): δ 7.0-7.5 (m, 15H, Ph), 1.74 (s, 15H, CMe), 0.14 (s, 6H,
ZrMe).
Cp *Zr Me2(OCMeP h 2). A solution of Cp*ZrMe3 (0.272 g,
1.00 mmol) in 50 mL of hexanes was treated dropwise with
Ph2CO (0.182 g, 1.00 mmol in 25 mL of hexanes) at -78 °C.
The reaction mixture was allowed to warm to room temper-
ature over 3 h, and the solvent was removed under reduced
pressure to give a solid residue which was dissolved in 10 mL
of hexanes. On cooling of the solution to -15 °C for 16 h,
Cp*ZrMe2(OCMePh2) was obtained as white crystals. Yield:
0.41 g (91%). Anal. Calcd for C26H34OZr: C, 68.82; H, 7.55.
Found: C, 68.96; H, 7.53. 1H NMR (benzene-d6): δ 7.0-7.5
(m, 10H, Ph), 1.91 (s, 3H, OCMe), 1.72 (s, 15H, CMe), 0.23 (s,
6H, ZrMe).
Cp *Zr Me(OCMeP h 2)2. A solution of Cp*ZrMe3 (0.272 g,
1.00 mmol) in 50 mL of hexanes was treated dropwise with
Ph2CO (0.364 g, 2.00 mmol in 25 mL of hexanes) at -78 °C.
The reaction mixture was allowed to warm to room temper-
ature over 3 h, and the solvent was removed under reduced
pressure to give a solid residue which was dissolved in 10 mL
hexanes. On cooling of the solution to -15 °C for 16 h,
Cp*ZrMe(OCMePh2)2 was obtained as white crystals. Yield:
0.59 g (93%). Anal. Calcd for C39H44O2Zr: C, 73.66; H, 6.97.
Found: C, 73.35; H, 6.78. 1H NMR (benzene-d6): δ 7.0-7.4
(m, 20H, Ph), 1.89 (s, 6H, OCMe), 1.69 (s, 15H, Cp*), 0.58 (s,
3H, ZrMe).
[Cp *Zr Me2(P Me3)2][MeB(C6F 5)3]. A stirred solution of
0.272 g of Cp*ZrMe3 (1.00 mmol) in 10 mL of toluene at -78
°C was treated first with a solution of 0.511 g of B(C6F5)3 (1.00
mmol) in CH2Cl2 cooled to -78 °C and then with a solution of
0.38 g of PMe3 (5.0 mmol; added dropwise over 45 min) in 200
Cp TiMe2(OP r i). CpTiCl3 (1.12 g, 5.11 mmol) in 100 m L
of ethyl ether was treated dropwise with 2-propanol (0.39 mL;
5.11 mmol) and triethylamine (0.71 mL; 5.11 mmol) in 25 mL
of ethyl ether. The reaction mixture was stirred for 12 h, after
which the resulting yellow suspension was filtered off. The
solvent was then removed from the filtrate, and the resulting
product, CpTiCl2(OPri), was purified by sublimation. Yield:
1.03 g (83%). 1H NMR (toluene-d8): δ 6.07 (s, 5H, Cp), 4.32
(septet, J HH 6.2, 1H, CH), 0.96 (d, 6H, Me).7f A solution of
CpTiCl2(OPri) (0.83 g, 3.42 mmol) in 25 mL of ethyl ether,
cooled to 0 °C, was treated with 6.84 mmol of MeMgBr (2.30
mL of 3.0 M solution). The reaction mixture was filtered
through Celite, and the filtrate was taken to dryness. Anal.
Calcd for C10H18OTi: C, 59.42; H, 8.97. Found: C, 59.57; H,
9.04. 1H NMR (benzene-d6): δ 5.92 (s, 5H, Cp), 4.51 (septet,
J HH 6.1, 1H, CH), 1.17 (d, 6H, CMe), 0.64 (s, 6H, TiMe).
Cp TiMe2(OBu t). CpTi(OBut)3 was formed by treating a
solution of CpTiCl3 (0.915 g, 4.17 mmol) in 150 mL of ethyl
ether with tert-butanol (0.93 g, 12.5 mmol) and triethylamine
(1.74 mL, 12.5 mmol). The reaction mixture was stirred for
16 h and filtered through Celite, and the solvent was removed
from the filtrate to give a lemon-yellow oil. The latter was
sublimed at 35 °C onto a cold finger at -78 °C to give
CpTi(OBut)3 as a colorless oil. Yield: 0.54 g (39%). 1H NMR
(toluene-d8): δ 6.18 (s, 5H, Cp), 1.22 (s, 27H, Me). Reaction
of CpTiCl3 (0.172 g, 0.78 mmol) with CpTi(OBut)3 (0.13 g, 0.39
mmol) in 100 mL of ethyl ether for 16 h, followed by solvent
removal, yielded a yellow-green solid which, on sublimation,
gave CpTi(OBut)Cl2 as a bright yellow, crystalline solid.
Yield: 0.18 g (88%). 1H NMR (benzene-d6): δ 6.14 (s, 5H, Cp),
1.09 (s, 9H, Me). A solution of CpTi(OBut)Cl2 (0.11 g, 0.43
mmol) in 25 mL of ethyl ether at 0 °C was treated with 0.86
mmol of MeMgBr (0.29 mL of 3.0 M ethyl ether solution). The
reaction mixture was stirred for 3 h and filtered through
Celite, after which the solvent was removed from the filtrate
to give CpTiMe2(OBut) as a lime green oil. Yield: 0.051 g
(55%). Anal. Calcd for C11H20OTi: C, 61.12; H, 9.32. Found:
(7) (a) Mena, M.; Royo, P.; Serrano, R.; Pellinghelli, M. A.; Tiripic-
chio, A. Organometallics 1989, 8, 476. (b) Wolczanski, P. T.; Bercaw,
J . E. Organometallics 1982, 1, 793. (c) Schock, L. E.; Marks, T. J . J .
Am. Chem. Soc. 1988, 110, 7701. (d) Giannini, U.; Cesca, S. Tetrahe-
dron Lett. 1960, 14, 19. (e) Massey, A. G.; Park, A. J . J . Organomet.
Chem. 1964, 2, 245. (f) Nesmeyanov, A. N.; Fedin, E. I.; Nogina, O.
V.; Kochetkova, N. S.; Dubovitsky, V. A.; Petrovsky, P. V. Tetrahedron
1966, Suppl. 8, Part II, 389. (g) J olly, W. L. The Synthesis and
Characterization of Inorganic Compounds; Prentice-Hall: Toronto,
1970; p 487 (describes the synthesis of [Cp2Fe][BPh4] in a procedure
which was used without modification). (h) Crowther, D. J .; J ordan,
R. F.; Baenziger, N. C.; Verma, A. Organometallics 1990, 9, 2574. (i)
An extensive series of alkoxy species may be synthesized by treating
Cp*MMe3 with ketones, aldehydes, and esters: Gillis, D. J .; Krupka,
C. A.; Lumb, S.; Baird, M. C. Unpublished results. (j) Bochmann, M.
Private communication.

696 Organometallics, Vol. 15, No. 2, 1996
Wang et al.
mL of hexanes, also cooled to -78 °C. The resulting yellow
precipitate was stirred at -78 °C for 30 min, and the product
was collected by filtration, washed with hexanes, and dried
in vacuo. Yield: 0.73 g (78%). Anal. Calcd for C37H42BF15P2-
Zr: C, 47.49; H, 4.52. Found: C, 47.94; H, 4.56. 1H NMR
(CD2Cl2 at 0 °C): δ 2.02 (s, 15H, Cp*), 1.38 (deceptively simple
triplet, 18H, PMe), 0.41 (br s, 3H, BMe), 0.07 (t, J 5.4 Hz, 6H,
ZrMe). 13C{1H} NMR (CD2Cl2 at 0 °C): δ 123.9 (Cp* ring C),
54.5 (t, J PC 10.4 Hz, ZrMe), 12.6 (d, J PC 10.2 Hz, PMe), 11.9
(br, BMe), 11.2 (Cp*Me).
[Cp *Zr Me2(NEt3)][MeB(C6F 5)3]. This compound was pre-
pared as was the trimethylphosphine analogue but decom-
posed at room temperature. It was therefore generated by
treating the previously described5a arene complex [Cp*ZrMe2-
(η6-toluene)2][MeB(C6F5)3] with triethylamine at low temper-
ature and characterized spectroscopically. 1H NMR (CD2Cl2
at -50 °C): δ 3.06 (q, J HH 7.2 Hz, 6H, CH2), 2.03 (s, 15H, Cp*),
1.04 (t, 9H, amine Me), 0.61 (s, 6H, ZrMe), 0.38 (br s, 3H, BMe).
[Cp *TiMe2(NEt3)][MeB(C6F 5)3]. Thiscompoundwasformed
by treating [Cp*TiMe2][MeB(C6F5)3] with triethylamine in CD2-
Cl2 at -50 °C, but it also could not be isolated. 1H NMR (CD2-
Cl2 at 0 °C): δ 3.32 (q, J HH 7.3 Hz, 6H, CH2), δ 2.01 (s, 15H,
Cp*), 1.35 (t, 9H, amine Me), 0.98 (s, 6H, TiMe), 0.45 (br s,
3H, BMe).
[Cp *TiMe2(d m p e)][MeB(C6F 5)3]. A stirred solution of
0.114 g of Cp*TiMe3 (0.50 mmol) in 20 mL of CH2Cl2 at -78
°C was treated first with a solution of 0.256 g of B(C6F5)3 (0.50
mmol) in 15 mL of CH2Cl2 cooled to -78 °C and then with
0.15 g of dmpe (1.0 mmol). The resulting yellow solution was
stirred at -78 °C for 30 min, and the product was precipitated
with 200 mL of hexanes cooled to -78 °C, collected by
filtration, washed with hexanes, and dried in vacuo. Yield:
0.39 g (88%). Anal. Calcd for C37H40BF15P2Ti: C, 49.91; H,
4.52. Found: C, 50.03; H, 4.58. 1H NMR (CD2Cl2 at 0 °C): δ
1.97 (s, 15H, Cp*), 1.58 (deceptively simple triplet, 6H, PMe),
1.40 (deceptively simple triplet, 6H, PMe), 0.96 (br s, 4 H, CH2)
0.72 (t, J PH 10.4 Hz, 6H, PMe), 0.38 (br s, 3H, BMe).
P olym er iza tion of Eth ylen e. In a typical procedure, 0.10
mmol of Cp*TiMe3 was dissolved in 50 mL of solvent which
was saturated with and blanketed by 1 atm of monomer. The
solution was then treated with 0.10 mmol of B(C6F5)3 in 10
mL of solvent, and the reaction mixture was stirred for 30 min
under 1 atm of monomer. The polymerization was then
quenched by adding 150 mL of methanol, and the precipitated
polymer was collected by filtration, washed with methanol, and
dried under reduced pressure at 50 °C for 16 h.
In an experiment involving efficient mixing, to 0.007 g of
Cp*TiMe3 (0.03 mmol) in 190 mL of vigorously stirred toluene
under nitrogen was added 0.015 g of B(C6F5)3 (0.03 mmol) in
10 mL of toluene at 35 °C. The solution initially turned bright
yellow and then dark; the reaction vessel was pressurized with
3 atm of ethylene and maintained at 35 °C. The reaction was
quenched, as above, after 1 h to give 280 g of polyethylene. In
a separate experiment, polymerization was observed to con-
tinue at least 10 h.
P olym er iza tion of Styr en e. Styrene (typically 2 mL,
purified of inhibitor by passage through an Aldrich inhibitor
removal column) and Cp*TiMe3 (0.03 mmol) were introduced
into a flask, either at room temperature or thermostated at
some other temperature. Polymerization was initiated by the
rapid addition of B(C6F5)3 (0.03 mmol) dissolved in 2 mL of
solvent (which may be styrene), at which point the tempera-
ture increased rapidly unless held constant. Reactions in CH2-
Cl2 were stirred for 0.5 h (20 °C) or 1.5 h (0 °C, -78 °C), and
polymerizations were then terminated by the addition of 1%
HCl in methanol; the resulting polymer was washed with
methanol and dried to constant weight at 100 °C.
complished by dissolving them in 1,2,4-trichlorobenzene at 130
°C followed by precipitation in 1% HCl in methanol. The
precipitates were then washed three times with 50 mL aliquots
of methanol, dried in vacuo at 100 °C, and then extracted with
methyl ethyl ketone in a Soxhlet extractor to obtain a-PS
(soluble) and s-PS fractions. Reactions in toluene at -15, -47,
and -78 °C did not result in precipitation of polymer and were
treated as were reactions in CH2Cl2.
P olym er iza tion of Styr en e in th e P r esen ce of Eth yl-
en e. Ethylene was bubbled through a solution of 40 mg of
Cp*TiMe3 (0.175 mmol) and 9.09 g of styrene (87.3 mmol) in
40 mL CH2Cl2 at 20 °C for 10 min. At this point, 90 mg of
B(C6F5)3 (0.175 mmol) dissolved in 10 mL of CH2Cl2 was added.
Bubbling of ethylene was continued for 1 h, the reaction
turning dark brown and a precipitate appearing. The reaction
was quenched with 100 mL of methanol, and the solid polymer
was filtered off, washed with methanol, and dried at room
temperature in vacuo. Yield: 9.1 g, The 1H NMR spectrum
indicated the presence of a-PS and polyethylene. The material
was only partially soluble in toluene, THF, and CH2Cl2 and
was extracted overnight in a Soxhlet extractor with CH2Cl2.
Only about 100 mg remained unextracted; this was identified
(NMR) as polyethylene rather than as a copolymer.
A reaction run similarly but at -30 °C turned orange and
became viscous within 1 h, but no precipitate formed. The
polymeric product precipitated as above with methanol was
pure a-PS. Yield: 4.8 g.
P olym er iza tion of r-Meth ylstyr en e. The monomer was
passed through an inhibitor remover column, stirred overnight
over CaH2, and then filtered, degassed, and distilled under
reduced pressure. Polymerization was initiated by the rapid
addition of 15 mg of B(C6F5)3 (0.03 mmol) dissolved in 2 mL of
a solvent (toluene or dichloromethane) to 7 mg of Cp*TiMe3
(0.03 mmol) dissolved in R-methylstyrene monomer (15 mmol).
Polymerizations were carried out at room temperature and
-78 °C, reaction mixtures turning orange. After
a few
minutes, the polymerization solution was quenched in metha-
nol (1% HCl). The precipitated polymer was then washed with
methanol and dried in vacuo at room temperature.
In a separate experiment, a solution of 30 mg of B(C6F5)3
(0.06 mmol) dissolved in 1.5 mL of toluene was added to a
solution of 14 mg (0.06 mmol) of Cp*TiMe3 dissolved in 1.5
mL of toluene and maintained at -78 °C. A solution of 2 mL
of R-methylstyrene in 8 mL of toluene was then added slowly
in three aliquots (3, 3, 4 mL) such that the ratio of R-methyl-
styrene:titanium after the three additions was 77:1. 154:1,
and 256:1, respectively. After each addition, a small aliquot
of the reaction mixture was withdrawn and treated as above,
and the resulting poly(R-methylstyrene) was characterized by
GPC.
Syn th esis of a Ra n d om Cop olym er of Eth yl Vin yl
Eth er a n d Styr en e. To 14 mg (0.06 mmol) of Cp*TiMe3
dissolved in 1.5 mL of CH2Cl2 at -78 °C were added 30 mg
(0.06 mmol) of B(C6F5)3 dissolved in 1.5 mL of CH2Cl2. The
solution turned orange, and a solution containing 1.5 mL (15
mmol) of ethyl vinyl ether and 3.5 mL (30 mmol) of styrene in
4 mL of CH2Cl was added slowly. After 5 min, the reaction
was terminated by the addition of methanol to give a sus-
pended solid, which was concentrated by centrifugation,
separated from the liquid phase, and dried in vacuum. The
solvent of the supernatant was removed under reduced pres-
sure to give a yellow, sticky material. Both materials were
characterized by NMR spectroscopy and GPC.
In Situ NMR In vest iga t ion s. A mixture of Cp*Ti-
5f
(13CH3)3 (10 mg, 0.042 mmol) and B(C6F5)3 (23 mg 0.042
mmol) was placed in an NMR tube sealed with a septum. The
sample was cooled to -78 °C, and 0.4 mL of CD2Cl2 was added
to dissolve the solids. The solution turned orange immediately
as A formed. A solution containing 0.10 mL of styrene-d8 (1.26
mmol) in 0.32 mL of CH2Cl was added slowly with good
mixing, and the NMR tube was placed in the probe of an
Polymerizations in aromatic solvents at temperatures g0
°C were generally terminated in a similar manner within a
few minutes, although some samples solidified within seconds
(see below). In the latter cases, purification of the solids to
remove unreacted monomer and entrained catalyst was ac-
1
AM400 NMR spectrometer at -50 °C; H and 13C{1H} NMR

[Cp*MR2]+ Complexes
Organometallics, Vol. 15, No. 2, 1996 697
spectra were obtained over 30-60 min. The temperature was
then raised to -25 and 0 °C, at each of which temperatures
the reaction mixture was held for 15 min before 1H and 13C-
{1H} NMR spectra were obtained.
amount of B(C6F5)3 at -20 °C in CD2Cl2, exhibits
spectroscopic properties consistent with the structure
shown as A.5a The formation of A clearly involves initial
abstraction of a methyl carbanion from the titanium by
the strong Lewis acid B(C6F5)3, followed by binding of
the resulting borate anion to the electrophilic metal
center via the somewhat hydridic methyl hydrogen
atoms.6 The borate anion, as anticipated, is readily
displaced by a variety of ligands, and thus A provides a
source of the 10-electron, cationic species [Cp*TiMe2]+,
which should be an excellent catalyst for olefin polym-
erization.1
Resu lts a n d Discu ssion
Syn th eses of P oten tia l P olym er iza tion In itia -
tor s. The new compounds CpTiMe2(OR) (R ) Pri, But,
CPh3) CpTiMe(OBut)2, Cp*ZrMe2(OR) (R ) CPh3, CMe-
Ph2), and Cp*ZrMe(OCMePh2)2 were synthesized in
order to test these types of species as either polymeri-
zation initiators or as precursors to possible polymeri-
zation initiators. The compounds were prepared in
several ways, by treating the corresponding chloro
compounds with MeMgBr or by treating Cp*MMe3 (M
) Ti, Zr) with 1 equiv of either an alcohol or a ketone.
The latter two processes involve protonation of a methyl
group and insertion of a ketone into a methyl-metal
bond, respectively, and are quite general.7i
The compounds [Cp*ZrMe2(THF)2][BPh4], [Cp*ZrMe2-
(PMe3)2][MeB(C6F5)3], [Cp*ZrMe2(NEt3)][MeB(C6F5)3],
and [Cp*TiMe2(dmpe)][MeB(C6F5)3] were prepared as
representatives of a type of complex similar to the
established polymerization initiators [Cp*MMe2][MeB-
(C6F5)3] (M ) Ti, Zr, Hf)5a but with coordination sites
at least partially occupied by ligands which should not
be readily displaced by olefins. A variety of such
compounds may be synthesized, but many are unstable
and we report here only a few which are relevant to the
theme of this paper.
A number of monocyclopentadienylmetal systems
have been shown to polymerize ethylene,2e,g,h,j,r,s and we
find that bubbling ethylene through solutions of A,
formed by combining equimolar amounts of Cp*TiMe3
and B(C6F5)3 in methylene chloride, 1,2-dichloroethane,
or toluene at room temperature, results in significant
heating and almost immediate polyethylene (PE) gel
formation. Treatment, after 30 min, of the reaction
mixtures with acidic methanol followed by drying at 80
°C results in white, low-solubility polymers with Tm
varying from ∼127 to ∼141 °C and measured Mw
)
(4.1-22.3) × 104; in other cases, the PE was insuf-
ficiently soluble for molecular weights to be made (see
below), and we estimate that Mw g 300 000 for these
1
samples. The PE appears largely to be linear, since H
NMR spectra of soluble samples exhibit only a singlet
at δ 1.39 (135 °C in 1,1,2,2-C2D2Cl4).
While slow monomer diffusion results in low total
conversion to PE under these conditions, that the
catalyst is highly active is attested to by the observation
that dissolution of Cp*TiMe3 and B(C6F5)3 in liquid
ethylene, at -197 °C, also results in PE formation; the
polymerization process continues at the boiling point of
ethylene (-104 °C). Indeed, in an experiment carried
out with vigorous stirring and careful control of tem-
perature, a much higher yield of PE was obtained.
Exposure of a toluene solution of [Cp*TiMe2(η6-toluene)]-
[BMe(C6F5)3] to 3 atm of ethylene at 35 °C, with a
controlled monomer feed in order to prevent runaway
conditions, resulted in polymerization which continued
for at least 10 h and which produced 280 g of PE within
1 h. The latter corresponds to unoptimized production
of 9.3 × 106 g of polymer/mol of Ti/h. The PE would
not dissolve in either 1,2,4-trichlorobenzene or R-chlo-
ronaphthalene (6 days at 150 °C), could not be forced
through a melt indexer (ASTM method D-1238, proce-
dure 8, condition E), and melted at 141 °C (DSC remelt).
Thus the material presumably has a very high weight
average molecular weight.
The 1H NMR spectrum of [Cp*TiMe2(dmpe)][MeB-
(C6F5)3] exhibits inter alia two distinct PMe environ-
ments and a BMe resonance at a chemical shift char-
acteristic of the free borate anion5a but little temperature
dependence. Therefore the dmpe does not dissociate
significantly or take part in any intramolecular ex-
change process; indeed, unlike [Cp*TiMe2][MeB(C6F5)3]
(A),5a [Cp*TiMe2(dmpe)][MeB(C6F5)3] behaves in solu-
tion as a single, solvent-separated, 14-electron species.
The 1H NMR spectrum of [Cp*ZrMe2(PMe3)2][MeB-
(C6F5)3] also exhibits little variation with temperature
or on addition of free PMe3 to the solution between -50
and 0 °C. Interestingly, however, spin saturation
transfer experiments at both temperatures indicated
that exchange between free and coordinated PMe3 does
occur. Similar results were obtained with [Cp*ZrMe2-
(NEt3)][MeB(C6F5)3]; addition of free NEt3 has no affect
1
on the H NMR spectrum, but spin saturation transfer
experiments demonstrate that exchange between free
and coordinated ligands does occur. The analogous
titanium complex [Cp*TiMe2(NEt3)][MeB(C6F5)3] could
be formed by treating A with triethylamine in CD2Cl2
at -50 °C and exhibited similar exchange properties but
could not be isolated. The 1H NMR spectra of all of
these complexes exhibit methyl resonances attributable
to the free borate anion, as expected.
P olym er iza tion In vestiga tion s. Neu tr a l Com -
p ou n d s. No appreciable activity for ethylene or styrene
polymerization was observed for Cp*MMe3 (M ) Ti, Zr,
Hf), CpTiMe2(OPri), CpTiMe2(OBut), CpTiMe(OBut)2,
Cp*TiMe2(OCPh3), Cp*ZrMe2(OCPh3), Cp*ZrMe2(OCMe-
Ph2), or Cp*ZrMe(OCMePh2)2.
The borane, B(C6F5)3, is a very poor catalyst for
ethylene polymerization, while no cationic catalysts for
ethylene polymerization could be obtained by treating
Cp*TiMe3 with B(n-Bu)3 or BPh3; methyl abstraction
does not occur in these cases. The cationic complex
[Cp*TiMe2(NEt3)][BMe(C6F5)3], formed by treating
Cp*TiMe3 with the proton source [HNEt3][B(C6F5)4],
was also much less active for ethylene polymerization,
presumably because ethylene coordination was severely
hindered by the presence of the amine or the counterion;
anion effects have also been reported for metallocene
systems.1h The dmpe complex, [Cp*TiMe2(dmpe)][BMe-
(C6F5)3], also exhibited very low activity, presumably for
the same reason.
Ca tion ic Com p ou n d s. Eth ylen e P olym er iza tion .
We have earlier shown that [Cp*TiMe2][MeB(C6F5)3],
generated by treating Cp*TiMe3 with an equimolar

698 Organometallics, Vol. 15, No. 2, 1996
Wang et al.
Catalysts of the type [Cp*MMe2][BMe(C6F5)3] (M )
Zr, Hf) were found to be good but not excellent ethylene
polymerization initiators in toluene, in which they exist
as the arene complexes [Cp*MMe2(η6-toluene)][BMe-
(C6F5)3],5a yielding PE with Tm > 140 °C. However,
coordination of good ligands, as in [Cp*ZrMe2(PMe3)2]-
[MeB(C6F5)3], [Cp*ZrMe2(THF)2][MeB(C6F5)3], and [Cp*-
ZrMe2(NEt3)][MeB(C6F5)3], resulted in very poor cata-
lyst systems. The complex [Cp*ZrMe2][BPh4], formed
by treating Cp*ZrMe3 with [(C5H4Me)2Fe][BPh4], was
also a relatively poor catalyst, suggesting that the
nature of the counteranion is important.1h
normally contained relatively small amounts of a-PS,
but as much as 15-20% by weight entrapped monomer
(by TGA), consistent with the general tendency of the
s-PS helices to trap aromatic molecules.8 The monomer
is only partially removed by heating in vacuo at 80 °C
for 24 h but is almost completely removed by heating
in vacuo at 80 °C for 48 h. Since we did not initially
appreciate the tenacity with which monomer is retained
by the solid, crude polymer, the yields claimed originally
are high by ∼15%.5b
The crude s-PS was generally purified by dissolution
in 1,2,4-trichlorobenzene at 150 °C, followed by precipi-
tation with acidified methanol to give white s-PS. The
tacticities were in all cases assessed by 13C{1H} NMR
spectroscopy on solutions in 1,1,2,2-tetrachloroethane-
d2 at 120 °C, the resonances of the ipso-carbon atoms
being well-resolved singlets at δ 145.36, characteristic
of the rrrr pentad of s-PS.4a,d,e In general, values of Mw
increased from ∼1 × 105 at 70 °C to ∼2 × 106 at 0 °C,
while Mw/Mn was essentially constant at ∼2.3. Glass
transition temperatures and melting points were gener-
ally ∼104 and ∼272 °C, respectively, comparable with
data from the literature,3,4 but the yields obtained here
and our observations of rapid solidification of the
product are not reported previously for this and similar
initiators.2b,c,i,k-o,3 However, we find that rigorous
exclusion of moisture is necessary for rapid and repro-
ducible polymerization to solid s-PS to occur. The
overall percentage conversion of styrene to s-PS appears
to be significantly higher than those reported elsewhere
for this and similar initiators.2,3
Good yields of and selectivities to s-PS are obtained
in n-hexanes, neat styrene, and some other aromatic
solvents such as mesitylene and m- and p-xylene, but
reactions run in benzene and anisole give lower yields.
While a number of the above-mentioned cationic com-
plexes which initiate ethylene polymerization also ini-
tiate styrene polymerization at room temperature, only
low yields of a-PS are formed. Interestingly, we find
that the cationic complex [Cp*Ti(CH2Ph)2][B(CH2Ph)-
(C6F5)3], formed by treating Cp*Ti(CH2Ph)3 with B(C6F5)3
in toluene, does initiate the formation of s-PS, and we
note recent reports that apparently analogous indenyl
and other variously substituted cyclopentadienyl sys-
tems are also effective in toluene.2l,n Indeed, toluene
seems generally to be the solvent of choice in almost all
cases for the polymerization of styrene to s-PS by
monocyclopentadienyltitanium systems,2b,c,k-m,p although
electron-withdrawing substituents on the aromatic sol-
vent ring can enhance s-PS yields.2p
The yields and weight average molecular weights of
the PE produced by initiator systems based on A are
comparable with those produced with many metallocene
catalysts,1 a finding consistent with the hypothesis,
proposed above, that 10-electron systems of the type
[Cp′MR2]+ or 12-electron complexes of the types [Cp-
′MRL2]2+ or [Cp′MR2L]+ might, for both steric and
electronic reasons, behave as more reactive catalysts
than their metallocene analogues. Our results are also
rather different from those reported earlier2i for the
same initiator system, the PE melting points reported
here being higher. The reasons for the discrepancy are
not clear, but we find that rigorous exclusion of moisture
is necessary for rapid and reproducible polymerization
to high molecular weight product to occur.
Styr en e a n d r-Meth ylstyr en e P olym er iza tion .
Treating solutions of styrene in CH2Cl2 with Cp*TiMe3
and B(C6F5)3 (1:1 molar ratio) at 20, 0, and -78 °C,
followed by addition of 1% HCl in methanol after 0.5-
1.5 h, yielded a-PS in yields of 97%, 99%, and 55%,
respectively. The values of Mw increased from 9.1 ×
103 to 5.7 × 104 as the temperature decreased in this
range, but the polydispersities varied little (2.5 ( 0.3).
Addition of solutions of B(C6F5)3 in toluene to solutions
of Cp*TiMe3 (1:1 molar ratio) in styrene at -15, -47,
and -78 °C resulted in darkening of the solutions and
increasing of the viscosities; frozen styrene monomer
(mp -31 °C) precipitated from the reaction mixture at
-78 °C. Addition of methanol after about 2 h resulted
in the precipitation of a-PS (yields 72 and 66% at -15
and -47 °C, respectively) which exhibited weight aver-
age molecular weights ∼104 and polydispersities ∼2.8
( 0.3. The a-PS formed at -78 °C exhibited a lower
weight average molecular weight, ∼5.3 × 103, presum-
ably because of the freezing out of the monomer before
polymerization was complete, but a significantly nar-
rower molecular weight distribution, 2.1. The a-PS
samples were readily soluble in many organic solvents,
including methyl ethyl ketone, and were identified by
1H NMR spectroscopy.4a No s-PS was formed in any of
the toluene polymerizations carried out at temperatures
<0 °C.5c
In vivid contrast to the reactions which result in the
formation of a-PS, addition of solutions of B(C6F5)3 in
toluene to solutions of Cp*TiMe3 (1:1 molar ratio) in
styrene under strictly anhydrous conditions in the
temperature range 0-70 °C resulted in significant
heating and darkening and in almost immediate forma-
tion of solid blocks of material that were predominantly
s-PS. Since B(C6F5)3 itself slowly polymerizes styrene
to a-PS, addition of more than 1 equiv of the borane to
the titanium complex resulted in an increase in the
proportion of a-PS formed. The crude, solid materials
Polymerizations of purified R-methylstyrene were
carried out in toluene and methylene chloride at room
temperature and at -78 °C; in most cases, the resulting
orange solutions were quenched in methanol (1% HCl)
to give white, polymeric materials which are soluble in
organic solvents (CH2Cl2, CHCl3, THF, and toluene).
Conversion of the monomer is almost 100% at -78 °C,
but only 20-40% at room temperture. Poly(R-methyl-
styrene) formed at room temperature in toluene was of
relatively low molecular weight (Mw < 103) and Tg (44
°C), but polymer formed at -78 °C in toluene exhibited
Mw ) 9 × 104, Mw/Mn ) 2.7, and Tg ) 158 °C. 13C{1H}
(8) (a) Roels, T.; Deberdt, F.; Berghmans, J . Macromolecules 1994,
27, 6216. (b) A suggestion4h that the sample contained residual acetone
from the extraction procedure seems quite unlikely in view of the
conditions under which the material was dried.

[Cp*MR2]+ Complexes
Organometallics, Vol. 15, No. 2, 1996 699
NMR spectroscopy indicated that the polymer formed
is predominantly syndiotactic in nature (84% rr
triads).9a-c Polymer formed in methylene chloride was
of lower molecular weight and higher polydispersity, but
the Mw values, the tacticity, and the glass transition
temperature reported here for material prepared in
toluene at -78 °C are comparable with those of poly-
(R-methylstyrene) prepared elsewhere using a variety
of initiators.9d-h
Interestingly, the growing poly(R-methylstyrene) ap-
pears to be partially living.10 Incremental additions of
monomer to a cooled (-78 °C) solution of initiator in
methylene chloride such that the monomer:initiator
ratios were ∼77:1, ∼154:1, and 256:1 resulted in the
formation of poly(R-methylstyrene) samples exhibiting
Mn ) 1.4 × 104 (Mw/Mn ) 8.1), 3.0 × 104 (Mw/Mn ) 4.1),
and 3.9 × 104 (Mw/Mn ) 2.9), respectively. The close
correlation between number average molecular weights
and amount of monomer added is consistent with the
existence of a living carbocationic polymer chain in this
system (see below).9f,10
in an NMR experiment involving A and carried out in
chlorobenzene-d5, in which A is stable for a few minutes,
at least, at room temperature.5e Interestingly, we also
find a remarkable cut-off point somewhat below 0 °C
for the formation of s-PS in toluene, such that only a-PS
is formed at lower temperatures.5c This observation
strongly suggests that at least two different types of
polymerization initiators, and hence titanium species,
are present at 0 °C and above in this solvent also.
We previously observed that ethylene and styrene are
seemingly polymerized by A faster in toluene at room
temperature than is propylene, and noted that these
relative rates appeared qualitatively to be inconsistent
with a generally applicable coordination polymerization
(Ziegler-Natta) mechanism.5b Although ethylene would
+
be expected to coordinate to the Cp*TiMe2 ion, as in
B, more readily than propylene11 and hence might well
Mech a n ism s of Olefin P olym er iza tion s by [Cp *-
TiMe2][MeB(C6F 5)3]. The enormous contrast in cata-
lytic activity between neutral compounds of the types
Cp*MMe3, CpMMe2(OR), CpMMe(OR)2 (M ) Ti, Zr, Hf),
and the various cationic derivatives described above is
consistent with the similar trends reported previously
for neutral and cationic metallocene catalyst systems1
and recently rationalized theoretically for Ziegler-Natta
processes in general.1f,g The same types of factors
presumably affect catalytic activity, namely the pres-
ence of both a positive charge to facilitate olefin coor-
dination and an alkyl group to initiate olefin insertion.1d,e
Consistent with this hypothesis, treatment of Cp*TiMe3
with [Ph3C][MeB(C6F5)3] in toluene at room tempera-
ture has also been found by us and others2q to result in
the formation of an excellent system for polymerization
of styrene to s-PS. This reaction is expected to proceed
as in eq 2 and, thus, should result simply in the
formation of the now well-characterized A.
polymerize faster, the much more sterically demending
olefins styrene and R-methylstyrene should coordinate
much less strongly than propylene and, in spite of their
higher solubilities, might therefore be expected to
undergo relatively slow Ziegler-Natta polymerization.
Since styrene and especially R-methylstyrene also un-
dergo very effective carbocationic polymerization,10
however, we suggested5b tht the strongly electrophilic
[Cp*TiMe2]+ may behave not only as a Ziegler-Natta
polymerization catalyst but as a carbocationic polym-
erization initiator10 with these monomers under some
conditions.
Indeed, consistent with a carbocationic mechanism10
for styrene polymerization in CH2Cl2 and at low tem-
peratures in toluene, we found that generally higher
molecular weights are obtained at lower temperatures
and with high monomer:initiator ratios. We also found,
for R-methylstyrene, that incremental addition of mono-
mer to a polymerization reaction in toluene at -78 °C
results in the formation of partially living poly(R-
methylstyrene), with increases of Mn correlating well
with the amounts of R-methylstyrene added. Similar
results have been reported using other carbocationic
initiators.9f
The interpretations drawn from these considerations,
apparently supported by complementary analyses of the
end groups formed when low-temperature polymeriza-
tions were quenched with alcohols, resulted in an
erroneous but ultimately serendipitous conclusion con-
cerning the applicability of a carbocationic mechanism
for the formation of s-PS.5b In brief, carbocationic
initiation of olefin polymerization by A would occur as
in Scheme 1. Here the olefin coordinates in C in a
nonclassical η1-fashion, the metal-olefin interaction
being stabilized by a complementary borate-olefin
Cp*TiMe3 + [Ph3C][MeB(C6F5)3] f
[Cp*TiMe2][MeB(C6F5)3] (A) + MeCPh3 (2)
The olefin polymerization initiator system described
here is complicated since, while A can confidently be
identified by its characteristic 1H NMR spectrum in CD2-
Cl2 in the temperature range 0 to -78 °C,5a warming a
solution to room temperature results in the appearance
1
of many new resonances in the methyl region of the H
NMR spectrum. While a similar experiment cannot be
carried out in toluene because the titanium complex
forms an insoluble oil, comparable results were obtained
(9) (a) Elgert, K.-F.; Wicke, R.; Stu¨tzel, B.; Ritter, W. Polymer 1975,
16, 465. (b) Malhotra, S. L. J . Macromol. Sci.-Chem. 1978, A12, 73.
(c) Wicke, R.; Elgert, K. F. Makromol. Chem. 1977, 178, 3075. (d) Ran,
R.-C.; J ia, X.-R.; Li, M.-Q.; J iang, S.-J . J . Macromol. Sci.-Chem. 1988,
A25, 907. (e) Toman, L.; Pokorny, S.; Spevacek, J . J . Polym. Sci., Part
A 1989, 27, 2217. (f) Toman, L.; Pokorny, S.; Spevacek, J . J . Polym.
Sci., Part A 1989, 27, 2229. (g) Da Silva, A.; Gomes, A. S.; Soares, B.
G. Polym. Bull. (Berlin) 1992, 29, 509. (h) Sawamoto, M.; Hasebe, T.;
Kamigaito, M.; Higashimura, T. J . M. S.-Pure Appl. Chem. 1994, A31,
937.
(10) (a) Kennedy, J . P.; Iva´n, B. Designed Polymers by Carbocationic
Macromolecular Engineering: Theory and Practice; Hanser Publish-
ers: Munich, Germany, 1991. (b) Sawamoto, M. Prog. Polym. Sci.
1991, 16, 111. (c) Matyjaszewski, K. In Comprehensive Polymer
Science; Eastmond, G. C., Ledwith, A., Sigwalt, P., Eds.; Pergamon
Press: New York, 1989; Vol. 3, p 639.
(11) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry;
University Science Books: Mill Valley, CA, 1987; sections 3.7, 11.2-
11.5.

700 Organometallics, Vol. 15, No. 2, 1996
Wang et al.
enriched -CHPhCH213CH3 end groups which would
result from the secondary insertion process of eq 3.
Sch em e 1
[Cp*Ti(13CH3)2]+ + PhCHdCH2 f
[Cp*Ti(13CH3)(CHPhCH213CH3)]+ (3)
Since a carbocationic process would not result in the
observed 13C-enriched end group, Pellecchia et al.
interpreted their results as providing unambiguous
evidence that only a Ziegler-Natta insertion process
applies in the formation of s-PS.
We have carried out a similar synthesis of s-PS by
utilizing purified Cp*Ti(13CH3)3, prepared via the reac-
tion of Cp*TiCl3 with Li(13CH3).5f The 13C{1H} NMR
spectrum of the resulting s-PS, prepared in toluene as
above, exhibited a sharp resonance at 11.70 ppm rela-
tive to TMS (1,1,2,2-C2D2Cl4 at 80 °C); assuming a
chemical shift of 1.5-2.0 ppm for hexamethyldisiloxane
relative to TMS,14a,b this result confirms that of Pellec-
chia et al. and indicates that their result was not unduly
influenced by the presence in solution of trimethylalu-
minum. We have, for other reasons, independently
come to the conclusion that the formation of s-PS does
not involve a carbocationic process (see below).
We have in the meantime, however, carried out a
series of experiments which strongly implicate a car-
bocationic mechanism for the polymerization of styrene
by A to a-PS in CH2Cl2. These experiments have
involved (a) NMR monitoring of polymerization reac-
tions of styrene-d0 and styrene-d8 by solutions of
Cp*Ti(13CH3)3 and B(C6F5)3 (1:1) in CD2Cl2 at -50, -30,
and 0 °C and (b) characterization by NMR spectroscopy
of the products of polymerization of styrene-d0 by
solutions of Cp*Ti(13CH3)3 and B(C6F5)3 (1:1) in CH2Cl2
at room temperature and -30 °C.
interaction. The structure of C resembles the structures
of several vinyl alcohol, vinyl ethyl ether, vinyl amine,
and methyleneimidazoline complexes of electrophilic
metals12 and also the probable transition state involved
in nucleophilic attack on olefins activated by coordina-
tion to electrophilic metals.13 The next step in the
polymerization process would involve the carbocationic
center of the metal ion-activated olefin molecule being
attacked by a second olefin monomer (D), in a manner
normally postulated for carbocationic polymerization
processes,10 followed by chain growth as in E.
In the case of a Ziegler-Natta catalyst system,11
alcoholysis of a growing polymer should result in
protonation of the titanium-carbon bond and formation
of a saturated hydrocarbon polymer backbone contain-
ing no alkoxy group. In contrast, alcoholysis of the
growing chain of Scheme 1 could involve nucleophilic
attack on the carbenium ion of E accompanied by
protonation of the titanium-carbon bond, resulting in
the formation of a polymer with methyl and alkoxy end
groups. Analysis by 1H and 13C{1H} NMR spectroscopy
of a sample of s-PS formed in toluene at 0 °C and
obtained after quenching with tert-butanol and drying
overnight at 80 °C initially gave seemingly promising
results, as indicated previously.5b However, although
the spectra exhibited tert-butoxy resonances with chemi-
cal shifts slightly different than those of the free alcohol,
we eventually found that drying samples in vacuo at
110 °C for 8 h results in complete removal of the alkoxy
resonances. Thus our original findings are now shown
to be in error, arising possibly because of the tenacity
with which molecules are trapped in the helical struc-
ture of s-PS.8a,b
The 1H NMR spectrum of [Cp*Ti(13CH3)2][(13CH3)-
B(C6F5)3] in CD2Cl2 in the temperature range -30 to
-60 °C exhibits resonances at δ 1.98 (s, 15H, Cp*
methyl), 1.53 (d, J 125 Hz, 6H, Ti-Me), and 1.19 (br d,
J 114 Hz, 3H, B-Me). Spectra also usually contain an
extremely weak resonance at δ 0.16 (d, J 126 Hz),
possibly attributable to partially hydrolyzed material.5g
The 13C{1H} NMR spectrum of [Cp*Ti(13CH3)2]-
[(13CH3)B(C6F5)3] in CD2Cl2 in the temperature range
-30 to -60 °C exhibits resonances at δ 131.1 (w, Cp*
ring C), 80.1 (vs, Ti-Me), ∼44.3 (br, m, B-Me), and 12.6
(w, Cp* methyl). Addition of styrene-d8 to solutions of
A (30:1 ratio) in CD2Cl2 at -78 °C, followed by warming
in a spectrometer probe to -50 °C, resulted initially in
the rapid disappearance of the 1H (residual) and 13C
resonances of the monomer and in the appearance of
broad resonances attributable to deuterated a-PS in the
A more definitive study of Pellecchia et al.4h has
presented strong evidence that polymerization of sty-
rene by A to s-PS proceeds via a Ziegler-Natta process
which involves initial secondary insertion of monomer
into a Ti-Me bond. Utilizing a mixture of Cp*TiMe3
and 13C-labeled AlMe3, in which there occurs rapid
methyl ligand exchange between titanium and alumi-
num, they obtained purified s-PS which exhibited in its
13C{1H} NMR spectrum a sharp resonance at 9.8 ppm
relative to hexamethyldisiloxane (in 1,1,2,2-C2D2Cl4 at
120 °C). This resonance was attributed to the 13C-
(14) (a) Randall, J . C. Polymer Sequence Determination Carbon-13
Method; Academic Press: New York, 1977; p 8. (b) Drake, J . E.;
Glavincevski, B. M.; Humphries, R.; Majid, A. Can. J . Chem. 1979,
57, 3253. (c) Ammendola, P.; Tancredi, T.; Zambelli, A.; Macromol-
ecules 1986, 19, 307. (d) Oliva, L.; Caporaso, L.; Pellecchia, C.;
Zambelli, A. Macromolecules 1995, 28, 4665. (e) The methyl reso-
nances of the compounds PhCHMe2, PhCH(Me)Et, and PhCH2CH2Me
all occur in the region δ 0.8-1.3. (f) Poly(EVE) exhibits 1H resonances
at δ ∼3.4-3.6 (br m, OCH + OCH2), ∼1.75-1.19, ∼1.5-1.7 (br m,
CCH2), 1.17 (t, CH3) (100 °C in 1,1,2,2-C2D2Cl4); a-PS exhibits 1H
resonances at δ ∼7.0, ∼6.55 (br m, Ph), ∼1.8-2.3 (br m, CH), ∼1.3-
1.6 (br, m, CH2) (100 °C in 1,1,2,2-C2D2Cl4). Poly(EVE) exhibits
13C{1H} resonances at δ 73.6-73.8 (m, OCH), ∼64.0-64.5 (m, OCH2),
∼39.4-41.4, (m, CCH2), ∼15.0 (CH3) (21 °C in CDCl3); a-PS exhibits
13C{1H} resonances at δ ∼125.6-128.3 (m, Ph), ∼40.3 (br s, CH), ∼43.3
(br m, CH2) (21 °C in CDCl3).
(12) (a) Cotton, F. A.; Francis, J . N.; Frenz, B. A.; Tsutsui, M. J .
Am. Chem. Soc. 1973, 95, 2483. (b) Chang, T. C. T.; Foxman, B. M.;
Rosenblum, M.; Stockman, C. J . Am. Chem. Soc. 1981, 103, 7361. (c)
Kuhn, N.; Bohnen, H.; Bla¨ser, D.; Boese, R. Chem. Ber. 1994, 127, 1405.
(13) Cameron, A. D.; Smith, V. H.; Baird, M. C. J . Chem. Soc., Dalton
Trans. 1988, 1037 and references therein.

[Cp*MR2]+ Complexes
Organometallics, Vol. 15, No. 2, 1996 701
1H and 13C{1H} spectra.14f There were slight (∼0.06
of [Cp*Ti(13CH3)2][(13CH3)B(C6F5)3] in CH2Cl2; the reac-
tions were initiated at -30 °C and at room temperature,
but both may have became warmer because of the
exothermicity of the reactions. The lower temperature
reaction mixture quickly turned dark brown, and the
reaction was quenched with methanol after 15 min. The
polymer was purified as above, and the 13C{1H} NMR
spectrum was found to exhibit a methyl end group
resonance at δ 12.9 (br) and a weaker resonance at δ
11.8 (br). The similar experiment carried out at room
temperature quickly turned dark blue, and the 13C{1H}
NMR spectrum of the purified a-PS exhibited the same
resonances at δ 12.9 and 11.8 as well as weaker methyl
resonances to slightlyl lower field.
ppm), apparently media-induced, changes in chemical
1
shifts of the H and 13C{1H} resonances of the titanium
species in solution and slight darkening of the color of
the solution on the addition of styrene. Although very
weak Ti-Me 13C resonances appeared at δ 94.2 and
102.6, there were no new resonances attributable to 13C-
enriched methyl end groups of any kind in either the 1H
or the 13C{H} NMR spectra. Thus we observed no
resonances in the 13C{1H} NMR spectrum at δ ∼11.70
or ∼22, which could be attributed to -CD(C6D5)CD2-
(13CH3) or -CD2CD(C6D5)(13CH3) end groups, respec-
tively,4b,h,14c,d or in the region δ 0.5-1.5 in the 1H
spectrum, where resonances of 13CH3 end groups would
appear.14e This experiment was found to be readily
reproducible.
In these cases, the onset of dark coloration, especially
blue in the room-temperature reaction, seems to imply
the presence of the type of Ti(III) species implicated
below for the formation of s-PS. If so, the onset of
Ziegler-Natta processes is possible, and although the
resonance at δ 11.8 suggests the presence of a -CH-
PhCH213CH3 end group, that at δ 12.9 cannot be
assigned with confidence. It does not correspond to the
any of the end group resonances reported elsewhere for
a-PS.14c The increases in viscosity in the NMR experi-
ments where the styrene8:Ti ratio was 30:1 may also
be indicative of reduced titanium species. The viscosity
increases could not be a result of increased monomer
incorporation, since no monomer was left, and instead
it seems likely that thermal decomposition of A involved
formation of methyl radicals, which in turn induced
crosslinking of the already formed polymer.
Although warming the reaction mixture to -25 °C
resulted in little discernible change in the 1H spectrum,
warming to 0 °C resulted in diminishing of the reso-
nances of the [Cp*Ti(13CH3)2][13CH3)B(C6F5)3] accom-
panied by the appearance of new Cp* methyl (δ 2-2.5)
and Ti-Me resonances (δ 0.217-0.16), of varying rela-
tive intensities, in the 1H spectrum. The 13C{1H}
spectrum exhibited very weak Cp* methyl resonances,
as well as the weak Ti-Me resonances noted above. The
new species have not been identified but may be
products of hydrolysis arising from traces of water
desorbing from the walls of the NMR tubes5g and/or,
more likely, products of thermal decomposition (the
relative intensities of these resonances increased with
time). On standing at 0 °C and above, this and similar
solutions darkened and became quite viscous. The
resonances noted above in the 1H spectrum shifted little
but became significantly weaker and broader; no new
methyl resonances of significant intensity appeared.
Desiring further evidence for the apparent carboca-
tionic mechanism for the formation of a-PS at low
temperatures, we treated ethylene-saturated solutions
of styrene in CH2Cl2, at 20 and -30 °C, with first
Cp*TiMe3 and then B(C6F5)3. Interestingly, the reaction
carried out at -30 °C produced only a-PS (Mw ) 8.3 ×
104, Mw/Mn ) 1.7), no polyethylene being found. This
result, obtained under conditions where A is thermally
stable and where the individual homopolymers are
readily formed, provides very strong evidence that the
active sites for the formation of polyethylene and
polystyrene are quite different. Indeed, if styrene were
polymerized to a-PS via intermediate η2-styrene com-
plexes analogous to B, then ethylene should compete
very effectively for metal coordination sites and at least
some polyethylene should be formed. On the other
hand, if polystyrene initiation and growth occur as in
Scheme 1, then the titanium complexes would become
neutral trialkyltitanium species, similar to Cp*TiMe3,
which is inactive as an ethylene polymerization initia-
tor. Although some titanium centers may in principle
coordinate ethylene initially, exchange with styrene
would remove these sites irreversibly from the polyeth-
ylene catalytic cycle.
Complementary experiments, with lower styrene-d8:
Ti ratios, exhibited similar NMR spectra at -50 °C.
However, different experiments with somewhat differ-
ent rates of warming resulted in the appearance of
C-Me resonances, of varying relative intensities, in the
regions δ ∼11.8-13.8 (Cp* Me region) and 15.8-16.2
(C-Me region) and Ti-Me resonances at δ ∼64.3, 94.2,
and 102.6 in the 13C{1H} spectra. In one case where
the polystyrene produced was precipitated with metha-
nol, twice dissolved in CH2Cl2 and reprecipitated with
methanol, and then dried, the 13C{1H} NMR spectrum
exhibited no enriched methyl end groups although the
spectrum of the reaction mixture had clearly exhibited
a resonance at δ 16.2.
Our interpretation of these observations is that the
deuterated polystyrene induced in the low-temperature
reaction by A is formed via other than a Ziegler-Natta
polymerization process. Since free radical initiated
processes seem under the circumstances quite unlikely,
carbocationic initiation seems indicated although no
species of type E could be detected. Apparently only a
small proportion of the total titanium in solution is
present as a very highly reactive E rather than as A.
Interestingly, although Cp*TiMe3 readily transfers a
methyl ligand to the tritylcarbenium ion (eq 2), the
results discussed here indicate that methyl transfer
from alkyltitanium species of type E to the carbenium
ion centers of the growing polymer chain does not occur.
In contrast, the reaction carried out at 20 °C turned
dark brown rather than orange and became warm;
precipitation of some polyethylene was observed within
1 min. The polymeric product isolated after 1 h at 20
°C contained a-PS (Mw ) 1.8 × 104, Mw/Mn ) 3.9) and
1
a small amount of polyethylene, identified by H NMR
spectroscopy. Clearly the thermal decomposition of A
above ∼0 °C, noted above, results in a degree, at least,
of Ziegler-Natta polymerization, perhaps by Ti(III)
species (see below).
In larger scale reactions, aliquots of styrene-d0 (0.5
mL in 2 mL of CH2Cl2) were polymerized in solutions

702 Organometallics, Vol. 15, No. 2, 1996
Wang et al.
We have obtained further support for the relevance
of Scheme 1 in the formation of a-PS by successfully
synthesizing random copolymers of styrene with ethyl
vinyl ether (EVE). Since EVE is polymerized via
carbocationic rather than Ziegler-Natta processes,10,15
it follows that any EVE-styrene copolymer formed
under the same conditions as is styrene hompolymer
would involve the styrene being able to participate in a
carbocationic polymerization process induced by the
titanium initiator under consideration here. In a typical
experiment, a 2:1 molar mixture of styrene and EVE in
CH2Cl2 was added slowly (30 min) to an orange solution
of A5a in CH2Cl2 at -78 °C. On completion of the
addition, methanol was added to the reaction mixture
to give a small amount of a fine, white suspension which
was concentrated by centrifugation, separated from the
supernatant solution, washed with methanol, and dried
at room temperature in vacuo for 2 h. This material
was quite different in appearance from a-PS prepared
in the same way but in the absence of EVE, which
precipitates cleanly to give a white solid and a clear
supernatant. Removal of solvent from the supernatant
resulted in a sticky, yellow oil which could be purified
of residual monomers by elution of a solution in CDCl3
through a Celite column prior to NMR analysis.
The 13C{1H} NMR spectrum (CDCl3) of the oil is
distinctly different from spectra of both the individual
homopolymers14f and of mixtures of homopolymers. In
particular, the CH2 13C multiplet resonance of a-PS at
∼δ 43 is absent, apparently because it has shifted in
the spectrum of the copolymer and is obscured by the
near-lying, overlapping EVE CH2 and a-PS CH 13C
multiplet resonances in the region δ 39-42. The change
in chemical shift presumably occurs because of a change
in environment on going from a-PS to a random copoly-
mer in which most of the styrene CH2 groups are bound
to an EVE OCH group, as in F .
Attempts to confirm that the oil contained styrene
bonded to EVE by obtaining a COSY spectrum16 of the
oil in CDCl3 at room temperature were of limited
success. The CH and CH2 1H resonances of the indi-
vidual atactic hompolymers are all closely spaced,
largely unresolved multiplets because of the variety of
microstructures possible,17 and the styrene CH and CH2
resonances of the copolymer are largely obscured by the
stronger EVE CCH2 resonance. However, a weak cross-
peak between the very broad OCH resonance(s) of the
EVE and a CH2 resonance of the styrene at δ ∼1.4
(shoulder, CDCl3) was observed, as anticipated for F .
Addition of alcohols to polystyrene reaction mixtures
in CH2Cl2 at low temperatures does not result in the
formation of alkoxy end groups, but this finding does
not in fact provide evidence against a carbocationic
mechanism. It has been found that even known, living
styrene polymers do not add a methoxy end group on
treatment with methanol.18
Two symmetric bands were observed in the GPC of
the solid suspension, with Mw ) 4.5 × 105 (minor
fraction) and 2.9 × 104 (major fraction). These are both
significantly higher than the value of Mw found for poly-
(EVE) (3.3 × 103), formed under the same conditions.
A single symmetric peak was observed in the GPC of
the oil, with Mw ) 1.9 × 104.
Returning to s-PS, the process producing this material
in toluene presumably involves coordination polymer-
ization,2b,c,k-m,3,4a-g but the nature of the catalytic
species is not at all clear since the originally formed A
is unstable above about 0 °C (see above). While a high-
temperature carbocationic process initiated by a cationic
thermal decomposition product of A might seem feasible
and was earlier advocated by us on the basis of argu-
ments outlined above,5b the carbenium ion end groups
of the growing polymer chain would be stabilized as in
Scheme 1 and a-PS should still be formed. We had not
yet investigated in detail the reactions in CH2Cl2 at the
time of our preliminary communication5b and did not
realize that a-PS is formed via the type of carbocationic
process under consideration. However, s-PS initiating
systems based on CpTi(OBu)3/MAO2k and Cp*Ti(CH2-
The 1H NMR spectra (CDCl3) of both materials
exhibited resonances of the two incorporated monomers
and are readily assigned on the basis of comparisons
with spectra of poly(EVE) and a-PS14f prepared under
the same conditions; there were essentially no differ-
ences in chemical shifts between spectra of the ho-
mopolymers and of the copolymers.14f Integrations of
the resonances of the solid copolymer suggested that the
ratio of incorporated styrene:EVE is ∼1.8:1, although
the possibility that the material contains some a-PS
homopolymer cannot be ruled out on the basis of the
chemical shifts. However, since any poly(EVE) formed
would be very soluble in methanol, the material which
precipitated clearly is not solely a mixture of homopoly-
mers but contains a significant amount of (presumably)
random copolymer, as in F . Integrations of the spectrum
4i
Ph)3/B(C6F5)3 exhibit ESR spectra attributable to Ti-
(III) species, and it has been sugested that the Ziegler-
Natta catalyst for s-PS in these cases are methyl-
titanium(III) compounds.2k,4i We have observed similar
ESR resonances on treatment of solutions of Cp*TiMe3
in toluene and chlorobenzene with equimolar amounts
of B(C6F5)3 under the conditions utilized in ref 4i, and
again a Ti(III) species seems possible although other
workers have failed to observe an ESR resonance using
Cp*TiMe3 activated with either B(C6F5)3 or [Ph3C]-
[B(C6F5)4].2q
of the oil suggested that the ratio of EVE:a-PS in this
material was ∼5.5:1, and while the possibility that this
fraction contains some EVE homopolymer cannot be
ruled out, a-PS is insoluble in methanol and it is clear
that the material must also contain a copolymer, also
incorporating both monomers as in F . The oil may be
lower melting and/or of greater solubility because of its
lower weight average molecular weight or, more likely,
because of a higher proportion of EVE.
In complementary experiments, we found that initiat-
ing the polymerization process in toluene at room
temperature in the presence of excess trimethylalumi-
(16) Sanders, J . K. M.; Hunter, B. K. Modern NMR Spectroscopy,
2nd ed.; Oxford University Press: New York, 1993; Chapter 4.
(17) Tonelli, A. E. NMR Spectroscopy and Polymer Microstructure;
VCH Publishers: New York, 1989.
(15) (a) Higashimura, T.; Sawamoto, M. Compr. Polym. Sci. 1989,
3, 673. (b) Mukul, B.; Mazumdar, A.; Mitra, P. Encycl. Polym. Sci.
Eng. 1989, 17, 446.
(18) Faust, R.; Kennedy, J . P. Polym. Bull. 1988, 19, 21.

[Cp*MR2]+ Complexes
Organometallics, Vol. 15, No. 2, 1996 703
num (TMA) resulted in the formation of s-PS with Mw
only ∼2.6 × 104. Since the probable role of TMA would
be to transfer methyl groups to titanium, it would seem
that the catalytically most active titanium species in
solution does indeed contain methyl groups, but possibly
fewer than three, the presumed maximum possible in
the presence of TMA.
radical-generating homolysis processes would be ac-
centuated at higher temperatures, free radical styrene
polymerization would produce a-PS.20 The fact that
relatively little a-PS is formed at higher temperatures
by A thus argues very effectively against free radical
initiation of polymerization being a significant factor.
In contrast, initiating the polymerization process in
toluene and at room temperature under 1 atm of
hydrogen resulted in very little change in the molecular
weight distribution of the product s-PS; Mw ) 5 × 105.
Hydrogen is well-known for its ability to cleave alkyl-
titanium bonds and may be used to control the molec-
ular weights of polymers formed via Ziegler-Natta
processes.19 While we find that Cp*TiMe3 and A are
relatively inert to hydrogen gas (1 atm) at low temper-
atures, we have no way of knowing how a Ti(III) species
might behave. In any case, however, the major reason
for the lack of effect of hydrogen during polymerization
to s-PS may be that polymerization is too fast for
effective diffusion of hydrogen to the Ti-polymer bonds
to occur. This suggestion is quite reasonable given the
low solubility of hydrogen in toluene under these
conditions and the rapidity with which the reaction
mixtures solidify.
Su m m a r y
Treatment of Cp*TiMe3 with the highly electrophilic
borane B(C6F5)3 results in the formation of the methyl-
bridged complex Cp*TiMe2(µ-Me)B(C6F5)3 (A), which is
a good source in solution of the cationic species
[Cp*TiMe2]+. The latter is a very effective carbocationic
initiator for polymerization of styrene to atactic poly-
styrene and R-methylstyrene to largely syndiotactic
polymer at low temperatures and is an excellent catalyst
or catalyst precursor for the Ziegler-Natta polymeri-
zation of ethylene and, in aromatic media at tempera-
tures g0 °C, of styrene to syndiotactic polystyrene.
Ack n ow led gm en t. We are indebted to the Natural
Sciences and Engineering Research Council of Canada
for financial support of this research (Research, Col-
laborative Research and Development, and Strategic
Grants to M.C.B., Graduate Fellowship to D.J .G.), to
the Government of Ontario for a graduate fellowship
to D.J .G., and to DuPont Canada Inc. for both generous
and financial support and for a leave of absence to
D.J .G. We also thank A. L. Natansohn, K. E. Russell,
and M. Bochmann for helpful discussions and T. M. Liu
for helpful discussions and TGA and DSC measurements.
Finally, we note that free radical initiation of styrene
polymerization, which might be expected to be induced
by homolysis of titanium-methyl bonds at the relatively
high temperatures at which s-PS is most readily formed,
cannot be a factor in the formation of s-PS. While
(19) Reference 11, section 1.1.2.
(20) Meister, B. J .; Malanga, M. T. Encycl. Polym. Sci. Eng. 1989,
16, 21.
OM9501945