Pentamethylcyclopentadienylmethyltitanium silsesquioxanes and their zwitterionic complexes with tris(pentafluorophenyl)borane

Article abstract of DOI:10.1021/om9007345

Reaction of [(η⁵-C₅Me₅)TiMe ₃] with [(c-C₅H₉)₇Si ₈O₁₂(OH)] (SIPOSS) or [(c-C₅H₉) ₇Si7O₉(OSiMe₃)(OH)₂)] (DIPOSS') affords stoichiometrically the half-sandwich titanium-disiloxy-methyl complexes [(η⁵-C₅Me₅){(c-C₅H ₉)₇Si₈O₁₂O)₂TiMe] (1) and [(η⁵-C₅Me₅){(c-C₅H ₉)₇Si₇O₉(OSiMe₃)O ₂)TiMe] (2), respectively. Compound 2 consists of the two stereoisomers possessing the (η⁵-C₅Me₅) ligand and OSiMe₃ group in syn- (2a) and anti-position (2b). The more abundant 2a (2a/2b ≈2:1) was isolated by fractional crystallization. Compounds 1 and 2a reacted rapidly with B(C₆F₅) ₃ to give zwitterionic complexes [(η⁵-C ₅Me₅){(c-C₅H₉)₇Si ₈O₁₂O}₂Ti]⁽⁺⁾[(μ-Me)B(C ₆F₅)₃]^(-) (3) and syn-[(η⁵C₅Me₅){(c-C₅H ₉)₇Si₇O₉(OSiMe₃)O ₂}Ti]^{+)[(μ-Me)B(C₆F₅) ₃]^(-) (4a), respectively. Infrared spectra proved that 3 and 4a were thermally stable in the solid state and the structure of 4a was determined by X-ray single-crystal analysis. In contrast, ¹H, ¹³C and ¹⁹F NMR spectra of 3 in toluene C ₇D₈ revealed that it was completely dissociated into initial components at 25 °C; however, the features of 3 were dominating already at -35 °C. On the other hand, 4a was dissociated only slightly at 25 °C. Stable ionic complexes 7 and 8 were prepared by reacting [PhNHMe ₂]⁺[B(C₆F₅)₄] ^- with 1 and 2, respectively. Both 7 and 8 contained the molecule PhNMe₂ coordinated to the titanium cation and 8 consisted of the syn- and anti-isomers 8a and 8b in abundances found for 2a/2b. Analogous reaction of 1 and 2 with [Ph₃C]⁺[B(C₆F₅) ₄]^- afforded impure ionic complexes 9 and 10. When dissolved in the presence of PhNMe₂, the amine coordinated to the titanium cations, thus reproducing complexes 7 and 8a/8b, identified by ¹H NMR spectra. All compounds 1-10 were inactive in polymerization of styrene to syndiotactic polymer. A low activity was achieved by adding of 2 equiv of AlMe₃ to 3 or 4.

Full text of DOI:10.1021/om9007345

6944 Organometallics 2009, 28, 6944–6956
DOI: 10.1021/om9007345
Pentamethylcyclopentadienylmethyltitanium Silsesquioxanes and Their
Zwitterionic Complexes with Tris(pentafluorophenyl)borane
†
§
,§
Pinkas,§ Ivana Cı
ꢀ ꢀ
ꢁ ‡
´ ´
Vojtech Varga, Jirı sarova, Michal Horacek, and Karel Mach*
ꢁꢀ
†
ꢁ
ꢀ ´
Research Institute of Inorganic Chemistry, v.v.o., Revolucnı 84, 400 01 Ustı nad Labem, Czech Republic,
´
‡Department of Inorganic Chemistry, Charles University, Hlavova 2030, 128 40 Prague 2, Czech Republic,
and J. Heyrovskyꢁ Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, v.v.i.,
ꢀ
Dolejskova 3, 182 23 Prague 8, Czech Republic
§
Received August 21, 2009
Reaction of [(η5-C5Me5)TiMe3] with [(c-C5H9)7Si8O12(OH)] (SIPOSS) or [(c-C5H9)7Si7O9-
(OSiMe3)(OH)2)] (DIPOSS0) affords stoichiometrically the half-sandwich titanium-disiloxy-methyl
complexes [(η5-C5Me5){(c-C5H9)7Si8O12O}2TiMe] (1) and [(η5-C5Me5){(c-C5H9)7Si7O9(OSiMe3)-
O2}TiMe] (2), respectively. Compound 2 consists of the two stereoisomers possessing the (η5-C5Me5)
ligand and OSiMe3 group in syn- (2a) and anti-position (2b). The more abundant 2a (2a/2b ≈ 2:1) was
isolated by fractional crystallization. Compounds 1 and 2a reacted rapidly with B(C6F5)3 to give
zwitterionic complexes [(η5-C5Me5){(c-C5H9)7Si8O12O}2Ti](þ)[(μ-Me)B(C6F5)3](-) (3) and syn-[(η5-
C5Me5){(c-C5H9)7Si7O9(OSiMe3)O2}Ti](þ)[(μ-Me)B(C6F5)3](-) (4a), respectively. Infrared spectra
proved that 3 and 4a were thermally stable in the solid state, and the structure of 4a was determined by
1
X-ray single-crystal analysis. In contrast, H, 13C, and 19F NMR spectra of 3 in toluene C7D8
revealed that it was completely dissociated into initial components at 25 °C; however, the features of 3
were dominating already at -35 °C. On the other hand, 4a was dissociated only slightly at 25 °C.
Stable ionic complexes 7 and 8 were prepared by reacting [PhNHMe2]þ[B(C6F5)4]- with 1 and 2,
respectively. Both 7 and 8 contained the molecule PhNMe2 coordinated to the titanium cation, and
8 consisted of the syn- and anti-isomers 8a and 8b in abundances found for 2a/2b. Analogous reaction
of 1 and 2 with [Ph3C]þ[B(C6F5)4]- afforded impure ionic complexes 9 and 10. When dissolved in the
presence of PhNMe2, the amine coordinated to the titanium cations, thus reproducing complexes 7
1
and 8a/8b, identified by H NMR spectra. All compounds 1-10 were inactive in polymerization
of styrene to syndiotactic polymer. A low activity was achieved by adding of 2 equiv of AlMe3 to 3
or 4.
Introduction
silanols offer the opportunity to study the stability and
catalytic properties of well-defined single-site catalytic
centers with the siloxy environment that occurs in silica-
anchored catalytic species.3 The use of POSS silanols for
the preparation of metal silanolates was reviewed,4 and
among early transition metal (Ti, Zr, Hf) POSS silanolates5
Polyhedral oligomeric silsesquioxanes (POSS) containing
monosilanol, e.g., [(c-C5H9)7Si8O12(OH)] (SIPOSS), disila-
nol, e.g., endo-[(c-C5H9)8Si8O11(OH)2] (DIPOSS), and trisi-
lanol functionalities, e.g., [(c-C5H9)7Si7O9(OH)3] (TRIPOSS)1
have been used to mimic the behavior of silica surface
hydroxyl groups acting as an anchor for homogeneous
catalysts. Compared with the silica surface where a single-
site catalyst is difficult to obtain due to a variety of Si-O-H
groups acting in isolated or cooperative modes,2 the POSS
(3) (a) Dijkstra, T. W.; Duchateau, R.; van Santen, R. A.; Meetsma, A.;
Yap, G. P. A. J. Am. Chem. Soc. 2002, 124, 9856–9864. (b) Fraile,
J. M.; García, J. I.; Mayoral, J. A.; Vispe, E. J. Catal. 2005, 233, 90–99.
(c) Dragonetti, C.; Carlucci, L.; D'Alfonso, G.; Lucenti, E.; Macchi, P.;
Roberto, D.; Sironi, A.; Ugo, R. Organometallics 2009, 28, 2668–
2676.
(4) (a) Duchateau, R. Chem. Rev. 2002, 102, 3525–3542. (b) Lorenz, V.;
Edelman, F. T. Adv. Organomet. Chem. 2005, 53, 101–151. (c) Marciniec,
B.; Maciejewski, H. Coord. Chem. Rev. 2001, 223, 301–335.
(5) (a) Feher, F. J.; Gonzales, S. L.; Ziller, J. W. Inorg. Chem. 1988,
27, 3440–3442. (b) Feher, F. J.; Walzer, J. F. Inorg. Chem. 1990, 29,
1604–1611. (c) Feher, F. J.; Budzichowski, T. A.; Rahimian, K.; Ziller,
K. J. J. Am. Chem. Soc. 1992, 114, 3859–3866. (d) Crocker, M.; Herold,
R. H. M.; Orpen, A. G. Chem. Commun. 1997, 2411–2412. (e) Crocker, M.;
Herold, R. H. M.; Orpen, A. G.; Overgaag, M. T. A. J. Chem. Soc., Dalton
Trans. 1999, 3791–3804. (f) Edelmann, F. T.; Giessmann, S.; Fischer, A.
J. Organomet. Chem. 2001, 620, 80–89. (g) Wada, K.; Itayama, N.;
Watanabe, N.; Bundo, M.; Kondo, T.; Mitsudo, T. Organometallics 2004,
23, 5824–5832.
*To whom correspondence should be addressed. E-mail: mach@
jh-inst.cas.cz.
(1) (a) Brown, J. F.; Vogt, L. H. J. Am. Chem. Soc. 1965, 87, 4313–
4317. (b) Feher, F. J.; Newman, D. A.; Walzer, J. F. J. Am. Chem. Soc. 1989,
111, 1741–1748. (c) Feher, F. J.; Newman, D. A. J. Am. Chem. Soc. 1990,
112, 1931–1936. (d) Feher, F. J.; Weller, K. J. Organometallics 1990, 9,
2638–2640. (e) Edelmann, F. T. Angew. Chem., Int. Ed. Engl. 1992, 31,
586–587. (f) Feher, F. J.; Soulivong, D.; Eklund, A. G. Chem. Commun.
1998, 399–400. (g) Liu, H.; Kondo, S.-I.; Tanaka, R.; Oku, H.; Unno, M.
J. Organomet. Chem. 2008, 693, 1301–1308.
(2) (a) Gauvin, R. M.; Buch, F.; Delevoye, L.; Harder, S. Chem.;
Eur. J. 2009, 15, 4382–4393. (b) Guzman, J.; Gates, B. C. Dalton Trans.
2003, 3303–3318. (c) Severn, J. R.; Chadwick, J. C.; Duchateau, R.;
Friederichs, N. Chem. Rev. 2005, 105, 4073–4147.
r
pubs.acs.org/Organometallics
Published on Web 12/03/2009
2009 American Chemical Society

Article
Organometallics, Vol. 28, No. 24, 2009 6945
numerous representatives were investigated as catalyst pre-
cursors for polymerization of olefins.6
Similar decomposition reactions were also acting in a
recently attempted synthesis of dialkoxo-bridged cyclopen-
tadienylsiloxo titanium cationic complexes.7,10 Surprisingly,
the lowest stability for complexes of this type was noticed for
the expected [Cp*Ti(OtBu)(Me)](þ)[MeB(C6F5)3](-) com-
plex, which decomposed immediately after mixing its pre-
cursors.11 The complete elimination of isobutene however
indicated a different decomposition pathway.12
Activation of the half-sandwich precursors with B(C6F5)3 has
been investigated in detail by R. Duchateau et al.6a,c in an effort
to address the thermal stability of cationic catalytic species
grafted on a silica surface. The catalyst precursors of the type
Cp0Ti(OR0)(CH2Ph)2 were generated from Cp0Ti(CH2Ph)3
(Cp0 = 1,3-C5H3(SiMe3)2) and silanols (R0OH = SIPOSS or
Ph3SiOH), and the relation between the acidity of the silanols
and the stability of ionic products arising from interaction of
Cp0Ti(OR0)(CH2Ph)2 with B(C6F5)3 was studied. The pKa
measurements and DFT calculations of deprotonation energies
In the absence of the OR0 ligand, the [Cp*TiMe3]/
[B(C6F5)3] system formed a highly active catalyst for either
polymerization of ethene to a high molecular weight polymer
or styrene to syndiotactic polystyrene (syn-PS) when its
components were mixed in the presence of monomer using
nonpolar solvents.13a When no monomers were added, the
reaction led to the formation of the anticipated zwitterionic
complex [Cp*TiMe2](þ)[MeB(C6F5)3](-) in an NMR tube
experiment at low temperature; however, this product
rapidly decomposed at room temperature and was then
inactive toward olefins.13b Replacement of one Me group
with an alkoxy group led to a decrease in catalytic activity,
which was attributed to the decreased electrophilicity of the
metal resulting from π-electron donation of the alkoxy
oxygen lone pair to the empty d orbital on the metal.13c
The application of the C6F5 or OC6F5 ligands as poor π-
electron donor OR0 moieties yielded thermally unstable
zwitterionic complexes with B(C6F5)3.13d The only half-
sandwich alkoxytitanium complexes of this type character-
ized in the solid state were the catalytically inactive zwitter-
ionic dialkoxy complexes [Cp*Ti(OR0)2](þ)[MeB(C6F5)3](-)
showed that Ph3SiOH is significantly less Bronsted acidic than
e
SIPOSS, and correspondingly, 19F NMR investigation estab-
lished that the equilibrium (eq 1) is shifted more to the product
side for the less electron-withdrawing OSiPh3 ligand.
Cp0TiðOR0ÞðCH2PhÞ2 þ BðC6F5Þ3
a ½Cp0TiðOR0ÞðCH2PhÞꢀðþÞ½ðCH2PhÞBðC6F5Þ ꢀð -Þ
3
ð1Þ
Ethene and 1-hexene polymerization tests revealed that both
systems afforded atactic poly-1-hexene with a narrow distribu-
tion of molecular weight (D = 2.0), pointing to a single-site
catalytic center, and the system with OSiPh3 was twice as active
as that with the SIPOSS anion in polymerization of ethene,
apparently in accord with a higher concentration of ionic
catalytic species in equilibrium 1.6c
Investigations of half-sandwich titanium systems Cp0Ti-
(OR0)Me2/B(C6F5)3 containing other organic siloxy or
alkoxy ligands OR0 revealed that their thermal stability is
generally rather low. A recent study of the system for
Cp0 = η5-C5Me5 (Cp*) and OR0 = OSiiPr3 proved that
zwitterionic complexes [Cp*Ti(OSiiPr3)(Me)](þ)[MeB(C6-
F5)3](-) were formed and were observable in solution for
several days.7 The systems for Cp0=η5-C5H5 (Cp) and OR0=
OC6H3Me2-2,6 or similar ligands generated zwitterionic
complexes [CpTi(OR0)(Me)](þ)[MeB(C6F5)3](-); however,
these were unstable in solution even at -20 °C, decomposing
in two concurrent ways while eliminating HC6F5 and MeB-
(C6F5)2.8 The system for Cp0=Cp* and OR0=OSi(CH2CH2-
SiMePh2)3 also generated the zwitterionic [Cp*Ti(OR0)-
(Me)](þ)[MeB(C6F5)3](-) complex in solution, which decom-
posed at room temperature with t1/2 ≈ 48 h to give two
decomposition products, the same as in the preceding case.9
i
where R0 = tBu11 or Pr.14 Their X-ray crystal structures
showed that the methyl carbon atom was σ-bonded to boron,
and hydrogen atoms of the methyl group exerted an agostic
interaction15 with the titanium atom, similarly to that known
for zirconocene zwitterionic complexes.16
Recently, the effect of π-donation of the lone electron pair
from oxygen into the Ti-O bond was evaluated for a series of
Cp*2Ti(III)OR derivatives using the energy of the 1a1 f b2
transition in their electronic absorption spectra as an inher-
ent measure of the effect.17a The spectroscopic data revealed
ꢁ
(10) Postigo, L.; Sanchez-Nieves, J.; Royo, P.; Mosquera, M. E. G.
Dalton Trans. 2009, 3756–3765.
(11) Pinkas, J.; Varga, V.; Cı
ꢀ ꢁ ꢁꢀ
sarova, I.; Gyepes, R.; Horacek, M.;
´
Mach, K. J. Organomet. Chem. 2007, 692, 2064–2070.
(12) (a) Stoebenau, E. J.; Jordan, R. F. J. Am. Chem. Soc. 2006, 128,
8162–8175. (b) Jordan, A. Y.; Meyer, T. Y. J. Organomet. Chem. 1999, 591,
104–113.
(13) (a) Gillis, D. J.; Tudoret, M.-J.; Baird, M. C. J. Am. Chem. Soc.
1993, 115, 2543–2545. (b) Wang, Q.; Gillis, D. J.; Quyoum, R.; Jeremic, D.;
Tudoret, M.-J.; Baird, M. C. J. Organomet. Chem. 1997, 527, 7–14.
(c) Wang, Q.; Quyoum, R.; Gillis, D. J.; Tudoret, M.-J.; Jeremic, D.; Hunter,
B. K.; Baird, M. C. Organometallics 1996, 15, 693–703. (d) Sarsfield, M. J.;
Ewart, S. W.; Tremblay, T. L.; Roszak, A. W.; Baird, M. C. J. Chem. Soc.,
Dalton Trans. 1997, 3097–3104.
(6) (a) Duchateau, R.; Abbenhuis, H. C. L.; van Santen, R. A.; Thiele,
S. K.-H.; van Tol, M. F. H. Organometallics 1998, 17, 5222–5224.
(b) Duchateau, R.; Abbenhuis, H. C. L.; van Santen, R. A.; Meetsma, A.;
Thiele, S. K.-H.; van Tol, M. F. H. Organometallics 1998, 17, 5663–5673.
(c) Duchateau, R.; Cremer, U.; Harmsen, R. J.; Mohamud, S. I.; Abbenhuis,
H. C. L.; van Santen, R. A.; Meetsma, A.; Thiele, S. K.-H.; van Tol, M. F. H.;
Kranenburg, M. Organometallics 1999, 18, 5447–5459. (d) Kim, Y.; Han,
Y.; Lee, M. H.; Yoon, S. W.; Choi, K. H.; Song, B. G.; Do, Y. Macromol.
Rapid Commun. 2001, 22, 573–578. (e) Severn, J. R.; Duchateau, R.; van
Santen, R. A.; Ellis, D. D.; Spek, A. L. Organometallics 2002, 21, 4–6.
(f) Metcalfe, R. A.; Kreller, D. I.; Tian, J.; Kim, H.; Taylor, N. J.; Corrigan,
J. F.; Collins, S. Organometallics 2002, 21, 1719–1726.
ꢀ ꢁ
(14) Varga, V.; Pinkas, J.; Cısarova, I.; Gyepes, R.; Mach, K.;
´
ꢀ
ꢁꢀ
Kubista, J.; Horacek, M. Collect. Czech. Chem. Commun. 2008, 73,
1161–1176.
(15) (a) Brookhart, M.; Green, M. L. H. J. Organomet. Chem. 1983,
250, 395–408. (b) Braga, D.; Grepioni, F.; Tedesco, E.; Biradha, K.; Desiraju,
G. R. Organometallics 1997, 16, 1846–56. (c) Thakur, T. S.; Desiraju, G. R.
Chem. Commun. 2006, 552–555. (d) Mitoraj, M. P.; Michalak, A.; Ziegler, T.
Organometallics 2009, 28, 3727–3733.
ꢁ
ꢁ
(7) Postigo, L.;Vazquez, A. B.; Sanchez-Nieves, J.; Royo, P.; Herdtweck,
E. Organometallics 2008, 27, 5588–5597.
(8) (a) Phomphrai, K.; Fenwick, A. E.; Sharma, S.; Fanwick, P. E.;
Caruthers, J. M.; Delgass, W. N.; Abu-Omar, M. M.; Rothwell, I. P.
Organometallics 2006, 25, 214–220. (b) Fenwick, A. E.; Phomphrai, K.;
Thorn, M. G.; Vilardo, J. S.; Trefun, C. A.; Hanna, B.; Fanwick, P. E.;
Rothwell, I. P. Organometallics 2004, 23, 2146–2156.
(16) (a) Beck, S.; Prosenc, M. H.; Brintzinger, H.-H.; Goretzki, R.;
Herfert, N.; Fink, G. J. Mol. Catal. A 1996, 111, 67–79. (b) Beck, S.;
Lieber, S.; Schaper, F.; Geyer, A.; Brintzinger, H.-H. J. Am. Chem. Soc.
2001, 123, 1483–1489. (c) Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem.
Soc. 1994, 116, 10015–10031.
ꢁ
ꢁ
(9) Amo, V.; Andres, R.; de Jesus, E.; de la Mata, F. J.; Flores, J. C.;
(17) (a) Lukens, W. W., Jr.; Smith, M. R.III; Andersen, R. A. J. Am.
ꢁ
ꢁ
ꢀ ꢁ
Gomez, R.; Gomez-Sal, M. P.; Turner, J. F. Organometallics 2005, 24,
2331–2338.
Chem. Soc. 1996, 118, 1719–1728. (b) Varga, V.; Císarova, I.; Gyepes, R.;
ꢁꢀ
ꢀ
Horacek, M.; Kubista, J.; Mach, K. Organometallics 2009, 28, 1748–1757.

6946 Organometallics, Vol. 28, No. 24, 2009
Varga et al.
Scheme 1
that the oxygen π-donation effect is increasing in the order of
OR groups (c-C5H9)7Si8O12O (SIPOSS) < (tBuO)3SiO <
Ph3SiO < iPr3SiO < PhO < HO < tBuO < MeO, and DFT
calculations showed that the effect is inversely proportional
to polarity of the Ti-O bond.17b This implies that POSS
ligands will afford the least stable complexes in both
vacuum. The composition of 1, 2a, and 2b was determined
1
from H, 13C, and 29Si NMR spectra by integrals of reso-
nances for the C5Me5, cyclopentyl groups of the POSS
ligands, and the Ti-Me group. The NMR spectra of the
crude product 2 revealed two sets of signals in the intensity
ratio 2:1 attributed to 2a and 2b. Compound 2a was sepa-
rated from 2b by crystallization from hexane solution, and its
structure with the Cp* ligand in syn-position to the OSiMe3
group was deduced from a 1D NOESY spectrum, where the
irradiation of the OSiMe3 signal at 0.28 ppm gave rise to the
signal at 2.06 ppm of the η5-C5Me5 ligand. A fraction
crystallizing from the ultimate mother liquor was strongly
enriched in 2b (90%), showing interaction between reso-
nances of OSiMe3 at 0.28 ppm and TiMe at 1.16 ppm in the
1D NOESY experiment.
[Cp*Ti(OR)(Me)](þ)[MeB(C6F5)3](-) and [Cp*Ti(OR)2)](þ)
-
[MeB(C6F5)3](-) series. The low thermal stability of the
former complexes did not allow their isolation and investiga-
tion in the solid state;6-11,13 however, the presence of the
second OR group highly increased the stability of the latter
complexes for R=tBuO and iPrO ligands.11,14 This gives a
good prospect for investigation of properties of analogous
complexes with OR being a POSS ligand. Although they are
expected to be catalytically inactive, their subsequent activa-
tion, e.g., by methylation with AlMe3, should be applicable
for the design of single-site cationic centers grafted on the
silica surface without using methylaluminoxane.
Compounds 1 and 2a formed transparent crystals in
hexane; however, when the hexane was evaporated under vac-
uum or in argon, they became opaque, as they were losing the
hexane of crystallization. This property hampered the X-ray
single-crystal investigation because the crystals lost some
hexane before they were fixed to the measuring rod.
The structure of 1 was solved, confirming its formula in
Scheme 1; however, the data refinement was unsuccessful.
The infrared spectra of 1 and 2 were dominated by very
strong absorption bands due to valence Si-O-Si vibrations
of the silsesquioxane cages at νmax=1095 cm-1, and the next
intense absorption band due to the valence Si-O(Ti) vibra-
tion was displayed at 984 cm-1 for 1 and at 988 cm-1 for both
2a and 2b (90%). These wavenumbers are comparable with
those for the Si-O(Ti) vibration in Cp*2Ti(III)(SIPOSS)17b
at 1038 cm-1 on one side and in Ti(IV) titanasilsesquioxanes
in the range 918-970 cm-1 5e,18a,b or in Ti(OSiR3)4 com-
Here we report the synthesis of the zwitterionic [Cp*Ti-
(OR)2](þ)[(μ-Me)B(C6F5)3](-) and similar ion-pair com-
plexes [Cp*Ti(OR)2)]þ[B(C6F5)4]-, with OR being derived
from SIPOSS or [(c-C5H9)7Si7O9(OSiMe3)(OH)2] (DIPO-
SS0),1b and activation of the zwitterionic complexes with
AlMe3 for polymerization of styrene.
Results and Discussion
Half-Sandwich Methyltitanium Complexes with Silses-
quioxanes. An easy access to the titanium-monomethyl
complexes [Cp*{(c-C5H9)7Si8O12O}2TiMe] (1) and a mix-
ture of syn-[Cp*{(c-C5H9)7Si7O9(OSiMe3)O2}TiMe] (2a)
and anti-[Cp*{(c-C5H9)7Si7O9(OSiMe3)O2}TiMe] (2b) com-
plexes was found in the reaction of commercial [Cp*TiMe3]
with [(c-C5H9)7Si8O12(OH)] (SIPOSS) and [(c-C5H9)7Si7O9-
(OSiMe3)(OH)2)] (DIPOSS0), respectively (Scheme 1).
The reaction proceeded in hexane in a stoichiometric way
at room temperature and was indicated by evolution of
methane and by slight intensification of the pale yellow color
of the [Cp*TiMe3] solution. This was surprising since the
Cp00Ti(CH2Ph)3 (Cp00=1,3-C5H3(SiMe3)2) did not react with
more than one SIPOSS molecule.6c Actually, a 5% excess of
[Cp*TiMe3] was used to ascertain the complete consumption
of the POSS component because its unreacted residue was
easily removed from the reaction mixture as for the most
soluble component. On the other hand, the Ti-Me bond in 1
and 2 (i.e., 2a þ 2b) appeared to be extremely unreactive, as it
did not react with even one molar excess of the POSS
reagents at 60 °C. Both 1 and 2 proved to be thermally
robust, as they did not change after heating to 160 °C under
plexes close to 918 cm-1 18c
Conformers 2a and 2b (90%)
.
differed in the shape of the broad absorption band at 1095
cm-1, indicating a slight difference in Si-O-Si vibrations of
the silsesquioxane cages; however, the absorption bands due
to the OSiMe3 group at 840 and 753 cm-1 did not differ for
2a and 2b (see Supporting Information).
Formation of Zwitterionic Complexes 3 and 4a from 1 and
2a, Respectively. Mixing of 1 or 2a in a 5% excess over the
equimolar ratio with B(C6F5)3 in toluene gave immediately
an orange solution, which after solvent evaporation at 60 °C
under vacuum afforded an orange solid. This was dissolved
(18) (a) Ricchiardi, G.; Damin, A.; Bordiga, S.; Lamberti, C.; Spano,
G.; Rivetti, F.; Zecchina, A. J. Am. Chem. Soc. 2001, 123, 11409–11419.
(b) Bornhauser, P.; Calzaferri, G. J. Phys. Chem. 1996, 100, 2035–2044.
(c) Barraclough, C. G.; Bradley, D. C.; Lewis, J.; Thomas, I. M. J. Chem.
Soc. 1961, 2601–2605.

Article
Organometallics, Vol. 28, No. 24, 2009 6947
in hexane and purified by crystallization driven by a slow
evaporation of the solvent under vacuum. Excess 1 and 2a
were removed with mother liquors, and orange crystalline
products 3 and 4a were washed with cold condensing hexane
vapor. The orange crystals of both 3 and 4a remained
transparent as long as they were wet with the solvent;
however, when dried, they turned opaque due to the loss of
hexane of crystallization. Therefore, the samples were
thoroughly dried by overnight evacuation to 2 ꢁ 10-4 Torr at
60 °C for analytical and spectroscopic investigation, while
hexane-wet crystals were submitted for X-ray single-crystal
analysis. The solid-state IR-ATR spectra of solid 3 and 4a
showed the absence of free B(C6F5)3 and 1 or 2a, and the
solids obtained by evaporation of the respective mother
liquors showed in addition to the above spectra of 3 or 4a
only very weak absorption bands of excess 1 (984 cm-1) or 2a
(988 cm-1). Hence, it could be assumed that the adducts of 1
and 2a with B(C6F5)3 were formed quantitatively (Schemes 2
and 3). Further investigation of compounds 3 and 4a by
NMR and IR methods in solution, however, revealed sub-
stantial differences in stability of the both products.
Scheme 2
Scheme 3
The 1H, 13C, and 19F NMR spectra of 3 and 4a in C6D6 or
toluene C7D8 solutions revealed that complexes 3 and 4a in
these aromatic solvents dissociate to give equilibria 2 and 3
with equilibrium constants K4a > K3.
K3
s 3
1 þ BðC6F5Þ3
ð2Þ
s
K4a
3 s
(R=tBu or Pr).11,14 In addition, the resonance in the 11B
NMR spectrum of 4a at -13 ppm is indicative of the
[(μ-Me)B(C6F5)3](-) species.21 The 1D NOESY NMR ex-
periments indicate that the syn-configuration of 2a was
preserved in 4a. This was further confirmed by the X-ray
crystal structure of 4a (see below). Alerting was, however,
that weak signals of free B(C6F5)3 were also observed besides
those of 4a, indicating a slight complex dissociation. The
thermally induced dissociation of 4a in C7D8 was particu-
larly well observed in the 19F NMR spectra. The overwhelm-
ing signals attributable to 4a at 25 °C disappeared at ca. 65
°C, and only signals belonging to B(C6F5)3 were observed
above this temperature. The quantitative assessment of the
spectra components was, however, rather difficult because
the chemical equilibria were accompanied by varying rates of
exchange processes, changing the shape and chemical shifts
of resonance signals (see Supporting Information).
i
2a þ BðC6F5Þ
4a
ð3Þ
s
The thermal stability of all components of the equilibria
toward toluene was proved by thermal cycling of the C7D8
solutions of 3 and 4a in flame-sealed NMR tubes three times
in the range 25-90 °C, giving reproducible spectra. The
solution of 4a in C6D6 at 25 °C displayed 1H NMR spectra
that differed only marginally in chemical shifts from those of
2a. The presence of the zwitterionic complex was deduced
from broadening of the signal for the bridging methyl group
at 0.85 ppm with a half-width of ca. 6 Hz. Unfortunately, the
MeB group was not detected in the 13C NMR spectra, which
1
did not allow the determination of the JCH coupling con-
stant, whose reduced value would be indicative of the pre-
sence of agostic interaction.19 Stronger evidence for the
presence of 4a came from 19F NMR spectra, showing a set
of three new signals having a difference between the meta-
and para-fluorine resonances (Δδ(m,p-F)=5.2 ppm) that is
typical for a cation-anion association occurring in zwitter-
ionic complexes,20a defined also as inner-sphere ion pairs
(ISIP).20b Both features were also previously found for
zwitterionic complexes [Cp*Ti(OR)2)](þ)[MeB(C6F5)3](-)
Contrary to 4a, a solution of 3 in C6D6 at 25 °C showed no
evidence for the presence of the zwitterionic complex, and
only weak and broad signals belonging to 1 and B(C6F5)3
were found in both H and 19F NMR spectra. However,
1
strong features of only 3 were observed at -35 °C in C7H8 in
the 19F NMR spectra, showing a new set of signals with
Δδ(m,p-F)=4.9 ppm. These signals broadened with increas-
ing temperature and disappeared at 5 °C. The broad signals
of B(C6F5)3 were first observed at 25 °C and narrowed
further with increasing temperature (to 75 °C). A practical
absence of signals of both 3 and B(C6F5)3 in the range
€
(19) (a) Erker, G.; Fromberg, W.; Angermund, K.; Schlund, R.;
€
Kruger, C. J. Chem. Soc., Chem. Commun. 1986, 372–374. (b) Brookhart,
M.; Green, M. L. H.; Wong, L.-L. Prog. Inorg. Chem. 1988, 36, 1–124.
(c) Hyla-Kryspin, I.; Gleiter, R.; Kr€uger, C.; Zwettler, R.; Erker, G. Organo-
metallics 1990, 9, 517–523. (d) Green, M. L. H.; Hughes, A. K.; Popham,
N. A.; Stephens, A. H. H.; Wong, L.-L. J. Chem. Soc., Dalton Trans. 1992,
3077–3082.
(20) (a) Horton, A. D. Organometallics 1996, 15, 2675–2677.
(b) Zuccaccia, C.; Stahl, N. G.; Macchioni, A.; Chen, M.-C.; Roberts,
J. A.; Marks, T. J. J. Am. Chem. Soc. 2004, 126, 1448-1464, and references
therein.
(21) Maggioni, D.; Beringhelli, T.; D’Alfonso, G.; Resconi, L.
J. Organomet. Chem. 2005, 690, 640–646.

6948 Organometallics, Vol. 28, No. 24, 2009
Varga et al.
5-20 °C was due to extensive exchange broadening for all
components of equilibrium 2.
The solid-state IR-ATR spectra of 3 and 4a were character-
ized by shifts and splitting of valence Si-O(Ti) vibrations from
984 cm-1 for 1to 932 and 919 cm-1 for 3and from 988 cm-1 for
2 to 964 and 953 cm-1 in 4a. A larger shift for 3 compared to 4a
reflects a higher polarity of the Si-O(Ti) bond because this
silicon atom is involved in three Si-O-Si bonds, while in 4 the
silicon atom takes part in only two Si-O-Si bonds and binds
one rather electron-donating cyclopentyl group.5e,18 The kine-
matic and steric factors due to bonding of the DIPOSS0 ligand
by two Si-O(Ti) bonds should also favor higher wave-
numbers for ν(Si-O(Ti)) vibrations in 4. The vibrations of
the Si-O-SiMe3 group in free DIPOSS0, 2, and 4a moved only
slightly and were assigned to the bands at 847 and 868 cm-1 for
4a. The IR spectra of 3 in hexane solution revealed its dissocia-
tion into 1 and B(C6F5)3 with about 25% of 3 retained, as
deduced from the low intensity of the bands at 919 and
932 cm-1. It is apparent that some concentration of 3 was also
present in the C6D6 solution at 25 °C according to equilibrium
2, but not observed due to large NMR line broadening. The IR
spectrum of 4a in hexane did not reveal even a trace of free
B(C6F5)3, in contrast from the 19F NMR spectra in C6D6 or
C7D8, which reflects a poorer solvation ability of hexane with
regard to the aromatic hydrocarbons.
Crystal Structure of 4a. A simplified PLATON drawing of
4a is shown in Figure 1, and selected geometric parameters
are listed in Table 1. The ligands around the titanium atom
are arranged roughly in such a way that the plane defined by
the Ti and B atoms and the centroid of the cyclopentadienyl
ring (Cg) approximately bisects the cage of the DIPOSS0
ligand. The latter is, however, distorted, as seen from its
attachment to titanium with virtually equal Ti-O bonds of
1.80 A but with the Ti-O-Si angles differing by more than
11° (see Table 1). The structure of the DIPOSS0 ligand in 4a is
similar to those found in free DIPOSS0, its bis-
(diphenylphosphino)ethane platinum complex,22a chromate
complex,22b or various titanium complexes.5f The DIPOSS0
cage is formed of Si-O bonds with the lengths in the range
1.608-1.625(2) A, and the mouth to the cage is defined by
nonbonding distances between O(1) and O(2) of 2.872(3) A
and between Si(1) and Si(5) of 5.391(1) A. It is worthwhile to
compare Ti-O, Si-O(Ti), and Ti-Cg bond lengths of 4a
with those found for the Cp*Ti complex with two cyclohexyl-
containing DIPOSS0 ligands, syn,syn-[(η5-C5Me5){(c-C6-
H11)7Si7O9(OSiMe3)O2}{(c-C6H11)7Si7O9(OSiMe3)O(OH)}
Ti] (6) (Chart 1).5f On the other hand, to discuss the
zwitterionic bonding in 4a, its geometric data can be com-
pared with those for [Cp*Ti(OtBu)2](þ)[(μ-Me)B(C6F5)3](-)
(5) (Chart 1),11 whose catalytic activity induced by addition
of AlMe3 will be compared with those of 3 and 4a (see below).
The titanium-ligand bond lengths for the above com-
pounds and the [Cp*Ti(OtBu)3] complex23 related to 5 are
gathered in Table 2. It is surprising that replacement of the
methylborate anion with the singly bonded DIPOSS0 ligand
causes elongation of the Ti-Cg distance by as much as 0.3 A.
On the other side, a distinctly shorter Ti-C(μ) distance in 4a
Figure 1. PLATON drawing of [Cp*{(c-C5H9)7Si7O9(OSi-
Me3)O2}Ti](þ)[(μ-Me)B(C6F5)3](-) (4a) at the 30% probability
level with atom-labeling scheme. Cyclopentyl groups and hy-
drogen atoms except those on the (μ-Me) C(49) atom are
omitted for clarity.
ꢁ
˚
Table 1. Selected Bond Lengths (A) and Bond Angles (deg) for
Compound 4a
Ti-Cga
2.0301(15)
1.796(2)
1.639(2)
1.608(2)
1.680(4)
1.661(4)
2.24(4)
Ti-C(49)
Ti-O(2)
Si(5)-O(2)
O(12)-Si(8)
B-C(50)
B-C(62)
Ti-H(49B)
2.356(3)
1.803(2)
1.640(2)
1.639(2)
1.643(4)
1.645(5)
2.17(4)
Ti-O(1)
Si(1)-O(1)
Si(3)-O(12)
B-C(49)
B-C(56)
Ti-H(49A)
Ti-H(49C)
2.26(4)
Cg-Ti-O(1)
116.39(9)
119.13(9)
99.01(10)
161.35(14)
172.9(2)
Cg-Ti-O(2)
114.85(8)
105.87(9)
98.95(10)
145.68(13)
109.7(2)
106.7(2)
105.8(18)
70.9(19)
72.2(19)
Cg-Ti-C(49)
O(1)-Ti-O(2)
O(1)-Ti-C(49)
Ti-O(1)-Si(1)
Ti-C(49)-B
O(2)-Ti-C(49)
Ti-O(2)-Si(5)
C(49)-B-C(50)
C(49)-B-C(62)
B-C(49)-H(49B)
Ti-C(49)-H(49A)
Ti-C(49)-H(49C)
C(49)-B-C(56)
B-C(49)-H(49A)
B-C(49)-H(49C)
Ti-C(49)-H(49B)
107.9(2)
112.5(19)
111.3(19)
67.1(18)
a Cg denotes the centroid of the C(1-5) cyclopentadienyl ring.
with respect to 5 can reflect a lower electron density on the
titanium atom due to a lower oxygen π-donation effect.17 This
effect should be enhanced for [Cp*Ti(OtBu)3], resulting in
elongation of the Ti-Cg distance; however, the observed
simultaneous elongation of the Ti-O bonds can indicate that
the steric congestion at the titanium atom can take part. The
bonding of [(μ-Me)B(C6F5)3](-) to titanium is similar to that in
5. One of the bridging methyl group hydrogens, H(49B),
embedded in the Ti-C(49)-B angle (172.9(2)°) showed the
shortest distance to titanium (2.17(3) A) and the smallest
Ti-C(49)-H(49B) angle of 67.1(2)°; however, all bridging
methyl C-H bonds fulfill conditions for agostic bonding
interaction to titanium since the d0 titanium atom has empty
orbitals to interact with the σ-C-H bonding orbitals, the Ti-C
and Ti-H distances are comparable, and the Ti-C-H angle is
smaller than 100°.15b,c In solution, they all contribute equally to
a general electrostatic bonding interaction between the titanium
and boron atoms. A similar orientation of the bridging methyl
group was previously found in zirconocene complexes [Me4C2-
(22) Abbenhuis, H. C. L.; Burrows, A. D.; Kooijman, H.; Lutz, M.;
Palmer, M. T.; van Santen, R. A.; Spek, A. L. Chem. Commun. 1998,
2627–2628. (b) Feher, F. J.; Blanski, R. L. J. Chem. Soc., Chem. Commun.
1990, 1614–1616.
(η5-C5H4)2ZrMe](þ)[MeB(C6F5)3](-) 16a
(2-Me-4-tBu)}2ZrMe](þ)[MeB(C6F5)3](-) 16b
(1,2-Me2)}2ZrMe](þ)[MeB(C6F5)3](-) 16c
,
rac-[Me2Si{η5-C5H2-
,
(23) The crystal structure of [(Cp*)Ti(OtBu)3] will be published; unit
cell characterization: P2/1c, a 15.2964(3) A, b 8.6293(1) A, c 18.1119(4)
A, β 89.8924(11)°.
and [{η5-C5H3-
.

Article
Organometallics, Vol. 28, No. 24, 2009 6949
Chart 1
ꢁ
˚
Table 2. Titanium-Ligand Bond Lengths (A) for (Cp*)Ti Tripodal Compounds
compound
Ti-O
Si-O(Ti)
Ti-Cg
Ti-C(μ)
ref
6
4a
5
1.789-1.805(3)
1.796, 1.803(2)
1.766, 1.756(1)
1.785-1.806(2)
1.571-1.580(3)
1.639, 1.640(2)
2.317(5)
5f
this work
11
23
2.0301(15)
2.0464(7)
2.1007(12)
2.356(3)
2.436(1)
[Cp*Ti(OtBu)3]
Scheme 4
Scheme 5
Reactions of 1 and 2 with N,N-Dimethylanilinium and Trityl
Tetrakis(pentafluorophenyl)borates. Stirring of toluene solu-
tions of 1 or 2 in a slight excess over the equimolar ratio with
the low-soluble solid [PhNHMe2]þ[B(C6F5)4]- resulted in
dissolution of the latter to give cleanly ionic complexes
[Cp*{(c-C5H9)7Si8O12O}2Ti(κ-N-PhNMe2)]þ[B(C6F5)4]- (7)
(Scheme 4) or a mixture of syn- and anti-stereoisomers [Cp*
{(c-C5H9)7Si7O9(OSiMe3)O2}Ti(κ-N-PhNMe2)]þ[B(C6F5)4]-
(8a and 8b) (Scheme 5). The NMR probe reaction of 1 with
[PhNHMe2]þ[B(C6F5)4]- confirmed the formation of
methane (0.15 ppm in the 1H NMR spectrum) accompanied
with a gradual diminishing of the TiMe signal of 1 while
raising a new signal for the C5Me5 group at 1.83 ppm and a
broad signal at 2.71 ppm for the NMe2 group of N,N-
dimethylaniline coordinated to the titanium cation in 7. The
1H NMR spectrum of a mixture of8a and 8b possessed similar
characteristic features to that of 7. The C5Me5 signal occurred
at1.81 ppm for 8a andat1.68 ppm for 8b anda broad signal of

6950 Organometallics, Vol. 28, No. 24, 2009
Varga et al.
Scheme 6
NMe2 of the coordinated dimethylaniline at 2.55 ppm for 8a
and at 2.97 ppm for 8b. In addition, the signal for the OSiMe3
group was displayed at 0.21 ppm for 8a and at -0.18 ppm for
8b. This assignment was based on a 1D NOESY experiment,
where irradiation of the C5Me5 signal at 1.81 ppm enhanced
the OSiMe3 signal at 0.21 ppm and the NMe2 signal at 2.55
ppm. The enhancement of the latter signal also proved that
N,N-dimethylaniline was coordinated to the titanium center.
The N,N-dimethylaniline coordinated to cationic zirconium
centers gave either a similarly broad signal or two signals for
the two diastereotopic Me groups usually high-field shifted
from the signal of free N,N-dimethylaniline.24 The chemical
shifts for NMe2 in 7, 8a, and 8b were not shifted far from
that of the free ligand (δ 2.2-2.5 ppm depending on both
temperature and concentration),24e which would indicate
that N,N-dimethylaniline was only weakly bound. However,
an exhaustive evacuation of the solid complexes did not result
in its loss (see below).
under vacuum at 60 °C the 1H NMR spectrum of the residue
in C6D6 reproduced the signals obtained above for 7 in
intensity corresponding to complete conversion of 9 to 7
(eq 4). In addition, the appearance of signals of
9 þ PhNMe2 f 7
ð4Þ
free dimethylaniline with the sharp NMe2 resonance at 2.43
ppm indicated that some amount of the amine remained
entrapped in the solid 7.
A mixture of 2a and 2b also did not react cleanly with
[Ph3C]þ[B(C6F5)4]- to give syn- and anti-[Cp*{(c-C5H9)7Si7O9-
(OSiMe3)O2}Ti]þ[B(C6F5)4]- (10a and 10b) (Scheme 7).
The solid product after washing with hexane displayed a
1H NMR spectrum showing a complex signal pattern for
both OSiMe3 and C5Me5 protons with two dominating sets
of signals at 0.19/1.83 ppm and 0.09/1.94 ppm, which were
tentatively assigned to 10a and 10b, respectively. It is note-
worthy that the signal of the SiMe3 protons at 0.09 ppm was
surprisingly broad (ν1/2=8.5 Hz). This apparently indicates
an intramolecular coordination of the oxygen atom of the
OSiMe3 group to the titanium atom in 10b. This coordina-
tion bond is not, however, strong since addition of excess
N,N-dimethylaniline to the above product mixture followed
by its evaporation under vacuum resulted in the disappear-
ance of the above signals for both 10a and 10b. Instead, the
above signals for a mixture of 8a and 8b (eq 5) in intensities
corresponding
The reaction of [Ph3C]þ[B(C6F5)4]- with 1 to give [Cp*{(c-
C5H9)7Si8O12O}2Ti]þ[B(C6F5)4]- (9) (Scheme 6) was less
unambiguous, yielding a number of minor byproducts. The
NMR probe experiment showed a gradual diminution of the
TiMe signal at 1.02 ppm concurrent with a rise of the signal
at 2.02 ppm due to the formed Ph3CMe.25 In a region typical
for the C5Me5 resonances (1.6-2.4 ppm) the main signal
remained virtually at the same position as in 1 (2.04 ppm);
however, it was somewhat decreasing in intensity while four
new peaks were increasing, finally reaching only minor
intensities. In a preparative experiment, a solid product
was isolated and washed with hexane to remove Ph3CMe,
whose signal was nearly coincident with that of the main
10 þ PhNMe2 f 8
ð5Þ
1
product. The H NMR spectrum of the purified product
to the ratio of 2a/2b in 2 were observed in addition to minor
unidentified resonances. The trace of free dimethylaniline
was also observed with its NMe2 group resonating at 2.21
ppm. None of compounds 7-10 crystallized, which pre-
cluded their effective purification and separation of syn-
and anti-stereoisomers 8a/8b and 10a/10b.
The solid-state IR-ATR spectra of 7-10 showed absorp-
tion bands of the [B(C6F5)4]- anion, which are not sensitive
to the nature of the cation. Among them the strong band at
978 cm-1 falls into the range where the valence Si-O(Ti)
vibrations should be sought. The band is easily recognized
because its position and intensity relative to other absorption
bands of the [B(C6F5)4]- anion are constant. In DIPOSS0
complexes it competesinintensitywiththe most intensebroad
band of Si-O-Si vibrations, and in SIPOSS compounds it is
showed, besides the absence of the TiMe group of 1, one
dominant singlet signal at 2.03 ppm, which was tentatively
assigned to C5Me5 of 9, and a minor unidentified singlet
signal at 2.39 ppm. Due to uncertainty in the assignment of 9,
the solid product was reacted with an excess of N,N-dimethyl-
aniline. After removal of most of the free N,N-dimethylaniline
(24) (a) Horton, A. D.; Orpen, A. G. Organometallics 1991, 10, 3910–
3918. (b) Horton, A. D.; de With, J. Organometallics 1997, 16, 5424–5436.
(c) Schrock, R. R.; Casado, A. L.; Goodman, J. T.; Liang, L.-C.; Bonitatebus,
P. J., Jr.; Davis, W. M. Organometallics 2000, 19, 5325–5341. (d) Hollink,
E.; Wei, P.; Stephan, D. W. Organometallics 2004, 23, 1562–1569.
(e) Wilson, P. A.; Wright, J. A.; Oganesyan, V. S.; Lancaster, S. J.; Bochmann,
M. Organometallics 2008, 27, 6371–6374.
(25) Obrey, S. J.; Bott, S. G.; Barron, A. R. Organometallics 2001, 20,
5162–5170.

Article
Organometallics, Vol. 28, No. 24, 2009 6951
Scheme 7
the second most intense band in the spectra. The Si-O(Ti)
vibrations for SIPOSS compounds 7 and 9 are assigned to
bands at 919/938 and 917/934 cm-1, and in DIPOSS0 com-
pounds 8 and 10 to bands at 948 and 951 cm-1, respectively.
Styrene Polymerization. All complexes 3-10 were inactive
for polymerization of styrene in toluene at 50 °C, like
the dialkoxy zwitterionic complexes [Cp*Ti(OtBu)2](þ)[(μ-
Me)B(C6F5)3](-) (5)11 and [Cp*Ti(OiPr)2](þ)[(μ-Me)B(C6-
Table 3. Polymerization of Styrene with Complexes 3, 4, and 5
a
Combined with 2 equiv of AlMe3
conver-
syn-
Tm
(°C) D (Mw/Mn)
catalyst sion (%) PS (%) activityb
Mw
3
4
5
1.8
2.5
2.6
75
91
92
368
520
528
254 600 271
323 750 269
523 400 269
3.93
3.37
3.10
a Polymerization conditions: [Ti] 1.25 ꢁ 10-3 M, [styrene] = 2.178 M,
F5)3](-) 14
The siloxy(alkoxy)-methyl(benzyl) zwitterionic
.
[AlMe3] = 2.50 ꢁ 10-3 M, toluene 6.0 mL, styrene 2.0 mL, temperature
50 °C, time 30 min. b Activity: kg of polymer/[mol(Ti) ꢁ mol(styrene) ꢁ h].
and ionic complexes were reported to have low thermal
stability;6a,c,7-10,13 however, it was also suggested that the
zwitterionic complexes [Cp*Ti(OR)Me](þ)[(μ-Me)B(C6-
F5)3](-) should be active in styrene polymerization to syn-
diotactic polystyrene (syn-PS), and perhaps, they could be
more convenient than the system made by mixing Cp*TiMe3
with B(C6F5)3, which is superior in activity but very rapidly
deactivated.13b-d Hence, an attempt was made to methylate
the titanium center in zwitterionic complexes 3, 4, and 5 via
metathesis reaction with AlMe3 (eq 6). Since the above
zwitterionic complexes are formed nearly
Table 4. Polymerization of Styrene with Complexes 1, 2, and
Cp*Ti(OtBu)2Me Combined with MAOa
catalyst
precursor
conversion
(%)
yield
(g)
syn-PS
(%)
activityb
1
2
35.9
24.2
24.5
0.664
0.443
0.446
98
99
99
7483
5035
5104
Cp*Ti(OtBu)2Me
a Polymerization conditions: [Ti] 1.25 ꢁ 10-3 M, [styrene] = 2.178 M,
[Al]/[Ti] = 250, toluene 6.0 mL, styrene 2.0 mL, temperature 50 °C, time
30 min. b Activity: kg of polymer/[mol(Ti) ꢁ mol(styrene) ꢁ h].
ðþÞ
½Cp TiðORÞ2ꢀ ½ðμ-MeÞBðC6F5Þ ꢀð -Þ
ꢂ
3
þ AlMe3 a ½Cp TiðORÞMeꢀ ½ðμ-MeÞBðC6F5Þ ꢀð -Þ
ðþÞ
ꢂ
the trimethylaluminum is known to exchange methyl group-
(s) with the pentafluorophenyl groups of the borane, afford-
ing products of lower Lewis acidity than that of B(C6F5)3.26
The reaction with the borane is apparently the reason for the
necessity to use a molar AlMe3/Ti ratio of 2. A low efficiency
of the formation of catalytic centers was further confirmed
by polymerization tests where the titanium precursors 1, 2,
and Cp*Ti(OtBu)2Me were combined with MAO under the
same experimental conditions. Activities of these systems
were 1 order higher and ordered 1 > 2 ∼ Cp*Ti(OtBu)2Me
(see Table 4).
3
þ AlMe2ðORÞ
ð6Þ
immediately, the catalytic complexes were prepared in the
polymerization reactor by mixing of toluene solutions of
their components in styrene/toluene, addition of 1, 2, or
3 molar equiv of AlMe3 being the last. Only addition of
2 molar equiv of AlMe3 to 3, 4, and 5 triggered the poly-
merization of styrene to syn-PS, albeit with low activity
(Table 3). The tert-butoxy and DIPOSS0 complexes exerted
virtually the same activity and produced a more than 90%
syndiotactic polystyrene (syn-PS). In contrast, the SIPOSS
complex showed a lower activity, and the polymer contained
only 75% of syn-PS. This can be accounted for by the above-
noted dissociation of 3 in solution, which led to a decrease in
concentration of the catalytic species, on one hand, and to a
B(C6F5)3-induced cationic polymerization to give a 2-buta-
none-soluble atactic PS.13c The molecular weights Mw of
syn-PS follow the trend of decreasing stability of the com-
plexes in toluene solution (5 > 4 > 3, as determined by
NMR methods above and in ref 11). The D values above 3.0
indicate the presence of likely two catalytic centers, which
can arise from a competitive interaction of AlMe3 with the
cationic and anionic part of the zwitterionic complexes. In
addition to the titanium methylation, as anticipated by eq 6,
Investigations of the composition of the 3-5/2 AlMe3
systems by means of NMR spectroscopy were abandoned
due to an expected large variety of products, which, more-
over, could not be identical with those obtained in the
presence of styrene. However, some insight into the above
systems was obtained from the metathesis reaction of com-
pounds 1, 2, and Cp*Ti(OtBu)2Me with 2 equiv of AlMe3,
yielding various amounts of Cp*TiMe3 and Cp*Ti(OR)Me2
1
compounds as determined by H and 13C NMR spectra.27
(26) (a) Klosin, J.; Roof, G. R.; Chen, E. Y.-X. Organometallics 2000,
19, 4684–4686. (b) Bochmann, M.; Sarsfield, M. J. Organometallics 1998,
17, 5908–5912.
(27) Mena, M.; Royo, P.; Serrano, R.; Pellinghelli, M. A.; Tiripicchio,
A. Organometallics 1989, 8, 476–482.

6952 Organometallics, Vol. 28, No. 24, 2009
Varga et al.
Surprisingly, under comparable conditions (60 °C) the
largest amount (at least 32%) of Cp*TiMe3 was obtained
for 1, somewhat less for 2 (14%), and Cp*TiMe3 was absent
for Cp*Ti(OtBu)2Me where Cp*Ti(OtBu)Me2 roughly
amounted to 70% and the initial compound 30%.
The shift of equilibria 7 to the right in the order 1 > 2 .
Cp*Ti(OtBu)2Me following from
K3 ≈ K4a was found to be about 50 °C lower for 3. A higher
stability of 4a with regard to 3 is in agreement with the acidity
measurement for SIPOSS and DIPOSS0 using an overlap-
ping indicator method, which put DIPOSS0 as being less
acidic (pKip=9.5 ( 0.1) than SIPOSS (pKip=8.9 ( 0.4).3a
The difference in the stability of DIPOSS0 and SIPOSS
zwitterionic complexes was smeared out for the ion-pair
complexes 7-10; however, higher frequencies of Si-O(Ti)
vibrations in DIPOSS0 compounds 4a, 8, and 10 compared to
SIPOSS compounds 3, 7, and 9 are in qualitative agreement
with K4a > K3. The X-ray crystal structure of 4a gave Ti-Cg
and Ti-C(μ) bond lengths close to those of the di-tert-
butoxy complex 5, although the latter is more stable in
solution. Complexes 3, 4, and 5 were activated by addition
of 2 molar equiv of AlMe3 to catalyze the styrene polymer-
ization to syn-PS. However, the catalysts were of low activ-
ity, as could be expected from the low stability of the borane-
generated [Cp*Ti(OR)Me](þ) species in solution.7-10,13
Also, the generation of these species from inactive
[Cp*Ti(OR)2](þ) by metathesis with AlMe3 is not efficient
since AlMe3 is known to competitively react with borate
anions to give products of low Lewis acidity.26 These results
indicate a poor prospect for obtaining a highly active borane-
activated polymerization catalyst with the titanium atom
grafted to a silica surface except that Cp*TiMe3 is reacted
with isolated Si-O-H groups to generate after addition of
B(C6F5)3 the [Cp*Ti(OR)Me](þ)[(μ-Me)B(C6F5)3](-) species
(R=surface Si atom).
ꢂ
ꢂ
Cp TiðORÞ2Me þ AlMe3 a Cp TiðORÞMe2
ꢂ
þ AlðORÞMe2 a Cp TiMe3 þ AlðORÞ2Me ð7Þ
these experiments did not respond to the above polymeriza-
tion activities of the 3-5/2 AlMe3 systems (Table 3); how-
ever, it conforms to activities found for the MAO systems
(Table 4). The ease of the Ti-O(Si) bond metathesis with the
Al-Me bond is apparently an important factor for account-
ing for the extremely high activities of some Cp*Ti(OSiR3)3/
MAO systems ([Al]/[Ti] = 500, 70 °C).28 Among them the
TRIPOSS complex [Cp*Ti{(c-C5H9)7Si7O9(O)3}] exerted
activity (6900 kg of polymer/[mol(Ti) ꢁ mol(styrene) ꢁ h)
comparable to that of 1/MAO (Table 4). These results show
that although the OSit ligands do not participate in the
construction of the active center, which is generally proposed
to be, for example, the Cp*Ti(III)(η2-styrene)(η1-styrene)
cation,29 the silica-grafted half-sandwich titanocene is easily
replaced with the MAO molecule(s) playing the role of silica-
grafted counteranion.
From the other point of view, the silica-grafted Cp*TiCl3/
MAO system ([Al]/[Ti]=600, 50 °C) exerted yet lower activities
than those of Table 3; however, the activity increased by more
than 1 order at the polymerization temperature of 110 °C,30 a
temperature that has never been used for homogeneous
catalysts since a drop in activity was reported at temperatures
above 70 °C.29 This invokes the idea that an active Ti(III)
cation is formed under action of styrene at high temperature,
and this can be stable enough in the presence of the silica-
grafted MAO counteranion. Following this, there is a good
prospect for the use of Cp*TiMe3 in grafting to silica and
preparation of surface catalytic complexes thereof using bor-
ane. The above observed dissociation of complex 3 in nonpolar
solvents does not imply that the surface, catalytic complex
[Cp*Ti(OR)Me](þ)[(μ-Me)B(C6F5)3](-) (R = Si from silica
SiOH group) cannot be stable in the absence of solvents.
Attempts to model such systems using the SIPOSS ligand are
under way.
Experimental Section
General Considerations. Reactions of POSS monosilanol
[(c-C5H9)7Si8O12OH] (SIPOSS) and disilanol [(c-C5H9)7Si7O9-
(OSiMe3)(OH)2] (DIPOSS0) with [Cp*TiMe3] (1) to give pro-
ducts 1 and 2, respectively, and reactions of 1 and 2 with
[B(C6F6)3] or with the solid salts [PhNMe2H]þ[B(C6F5)4]- and
[Ph3C]þ[B(C6F5)4]- were carried out on a vacuum/argon line in
sealed all-glass devices equipped with breakable seals. All air-
sensitive solid compounds were handled in a Labmaster 130
(mBraun) glovebox under purified nitrogen (concentrations of
oxygen and water were lower than 2.0 ppm) to weigh the
reagents, to prepare KBr pellets for IR spectra, or to fix single
crystals into Lindemann glass capillaries for X-ray diffraction
analysis. IR spectra were measured on a Nicolet Avatar FT IR
spectrometer in the range 400-4000 cm-1 and/or on a Nicolet
380 FT IR with an attenuated total reflectance (ATR) cell,
equipped with a silicon crystal in the range 650-4000 cm-1
(resolution 4 cm-1, 64 scans). A special air-protected cell for the
IR-ATR measurement was charged with the solid sample in a
glovebag under argon. KBr pellets were measured under nitro-
gen in an air-protected compartment equipped with KBr win-
dows. The IR-ATR data for neat samples were generally in
qualitative agreement with those obtained from KBr pellets.
The former data are presented for all compounds since they
eliminate effects arising from interaction of ionic samples with
the KBr matrix. Hexane solutions were filled into KBr cells (d=
Conclusions
The zwitterionic complexes 3 and 4a are formed easily
from the half-sandwich titanium disiloxymonomethyl com-
pounds 1 and 2a, respectively, by adding an equimolar
amount of B(C6F5)3. The compounds are thermally stable
at least to 60 °C in the solid state; however, they dissociate in
solutions, releasing the initial reagents. The equilibrium
constants for formation of 3 and 4a (K3 and K4a) could
not be determined from NMR data because the thermally
induced concentration changes were accompanied with
varying exchange rates; however, the temperature at which
0.025-0.1 cm) under argon in a glovebag. H (300 MHz), 13C
1
(75 MHz), 29Si (59.6 MHz), and 19F (282.2 MHz) NMR spectra
were measured on a Varian Mercury 300 spectrometer in C6D6
or C7D8.11 B NMR (160.4 MHz) spectra were acquired on a
UNITY Inova 500 MHz at 273 K in C7D8. Sample tubes were
filled on a vacuum line and sealed off with flame. Monitoring of
reactions by NMR spectra was done using NMR sample tubes
equipped with Teflon screw-caps. They were charged under
argon in a glovebag. Elemental analyses were carried out on a
FLASH EA1112 CHN/O automatic elemental analyzer
(Thermo Scientific).
(28) Lyu, Y.-Y.; Byun, Y.; Yim, J.-H.; Chang, S.; Lee, S.-Y.; Pu,
L. S.; Lee, I.-M. Eur. Polym. J. 2004, 40, 1051–1056.
(29) Schellenberg, J. Prog. Polym. Sci. 2009, 34, 688-718, and
references therein.
(30) Yim, J.-H.; Ihm, S.-K. J. Appl. Polym. Sci. 2006, 102, 2293–2298.

Article
Chemicals. The solvents hexane and toluene were dried by
Organometallics, Vol. 28, No. 24, 2009 6953
Preparation of a Mixture of syn- and anti-[Cp*{(c-C5H9)7-
Si7O9(OSiMe3)O2}TiMe] (2a and 2b) and Isolation of syn-Iso-
mer (2a). [Cp*TiMe3] (0.19 g, 0.83 mmol) and DIPOSS0 (0.725 g,
0.76 mmol) were reacted in hexane in the same manner as
described above for 1. Crystallization of a crude solid product
from a saturated hexane solution (ca. 5 mL) at -28 °C afforded
a yellow crystalline mixture of syn-2a and anti-2b. Composition
from NMR data: 2a 66%, 2b 33%. After drying under vacuum
overnight a light yellow pseudocrystalline material was ob-
tained. Yield: 0.79 g (91%).
refluxing over LiAlH4 and stored as solutions of green dimeric
titanocene [(μ-η5:η5-C5H4C5H4)(μ-H)2{Ti(η5-C5H5)}2].31 Com-
mercial [Cp*TiMe3] was purchased from Strem Chemicals, Inc.,
and was not further purified. It was handled and weighed under
an argon atmosphere and transferred into reaction ampules on a
vacuum/argon line. The monosilanol POSS [(c-C5H9)7Si7O12-
SiOH] (SIPOSS) and trisilanol POSS [(c-C5H9)7Si7O9(OH)3]
(TRIPOSS) (all Aldrich) were degassed and used as such. [(c-
C5H9)7Si7O9(OSiMe3)(OH)2] (DIPOSS0) was prepared from
TRIPOSS according to the literature.1b B(C6F5)3 (Strem Che-
micals, Inc.) was analyzed by ATR under argon for the presence
of hydrolytic products (indicated by the absorption band at 1290
cm-1) and purified by reacting with an estimated amount of
[Cp*TiMe3] in hexane. A brown solution was separated from a
brown oily sediment, the hexane was distilled off under vacuum,
and the colorless B(C6F5)3 was obtained by sublimation of the
residue at 100 °C in high vacuum. The final purity was tested
using 19F NMR.32 Trityl tetrakis(pentafluorophenyl)borate,
[Ph3C]þ[B(C6F5)4]-, and N,N-dimethylanilinium tetrakis(pen-
tafluorophenyl)borate, [PhNMe2H]þ[B(C6F5)4]-, were pur-
chased from Aldrich and handled as such under argon. [Cp*Ti-
(OtBu)2Me] was obtained as described.11 Styrene (Aldrich) was
degassed and distilled under vacuum on solid green dimeric
titanocene. The green solution did not change its color after
warming to 60 °C for 30 min. The styrene was again degassed by
a freezing/thawing cycle and distributed by vacuum distillation
into ampules. Trimethylaluminum in toluene (1.0 M) was
diluted with toluene and distributed to make 10 μmol/2.0 mL
doses. N,N-Dimethylaniline (Aldrich) was distilled under argon
and degassed. The toluene solution of MAO (Aldrich) was used
as such.
Data for 2 (2a þ 2b) are as follows. IR (ATR Si, cm-1): 2948
(s), 2909 (m, sh), 2864 (m), 1449 (vw), 1249 (w), 1094 (vs, b), 1062
(s), 1042 (s, sh), 988 (s), 944 (m), 914 (m, b), 861 (w), 840 (m), 754
(w). Anal. Calcd for C49H90O12Si8Ti (1143.82): C, 51.45; H,
7.93. Found: C, 51.49; H, 7.98.
Isolation of 2a was performed by fractional crystallization
from a saturated hexane solution driven by a slow evaporation
of hexane under vacuum. The first crop of crystals contained
pure 2a. Drying under vacuum yielded 0.32 g (37%). The NMR
data for 2b were obtained from the content of the ultimate
mother liquor, consisting of 2b (90%) and 2a (10%).
Data for 2a are as follows. 1H NMR (298 K, C6D6): 0.28 (s,
9H, SiMe3), 0.84 (s, 3H, TiMe), 1.06-1.28 (m, 7H, CH,
cyclopentyl), 1.44-1.62 (m, 14H, CH2, cyclopentyl), 1.62-
1.86 (m, 28H, CH2, cyclopentyl), 1.86-2.03 (m, 14H, CH2,
cyclopentyl), 2.06 (s, 15H, C5Me5). 13C{1H} NMR (298 K,
C6D6): 2.33 (SiMe3), 11.95 (C5Me5), 23.10, 23.29, 23.93, 24.10,
25.84 (CH, cyclopentyl), 27.32, 27.47, 27.50, 27.64, 27.75,
28.07, 28.10, 28.18, 28.36, 28.54 (CH2, cyclopentyl), 47.15
(TiMe), 123.61 (C5Me5). 29Si (298 K, C6D6): -67.44 (2Si),
-66.70 (2Si), -65.74 (1Si), -65.57 (1Si), -64.96 (1Si), 7.72
(1Si, SiMe3).
Preparation of [Cp*{(c-C5H9)7Si8O12O}2TiMe] (1). The re-
agents [Cp*TiMe3] (0.228 g, 1.0 mmol) and SIPOSS (1.838 g,
1.96 mmol) were weighed under argon into a reaction ampule
attached to an argon/vacuum line, which was cooled to -70 °C.
After sealing off the filling arm, the ampule was degassed, and
hexane (15 mL) was distilled onto the reaction mixture. The
mixture was left to warm to room temperature under stirring
while the bubbling of liberated methane was observed. After
stirring for another 1 h the formed yellow solution was evapo-
rated under vacuum to dryness, and the solid residue was
dissolved in hexane. The saturated solution was cooled to
-28 °C, affording transparent yellow crystals. These were sepa-
rated from the mother liquor and dried by evacuating overnight
at 2 ꢁ 10-3 Torr to give a light yellow pseudocrystalline
material. The loss of hexane of crystallization during tranfer
of transparent crystals from hexane to Nujol oil and to the
cooled rod of the X-ray diffractometer was apparently respon-
sible for the poor diffraction data. These allowed solving the
structure and identifying all the ligands and the Ti-Me bond,
but precluded the anisotropic refinement. Yield: 1.75 g (88%).
Data for 1 are as follows. 1H NMR (298 K, C6D6): 1.04 (s, 3H,
TiMe), 1.10-1.40 (m, 14H, CH, cyclopentyl), 1.40-2.18 (m,
112H, CH2, cyclopentyl), 2.04 (s, 15H, C5Me5). 13C{1H} NMR
(298 K, C6D6): 11.69 (C5Me5), 22.66, 22.78 (CH, cyclopentyl),
27.50, 27.63, 27.86, 27.90 (CH2, cyclopentyl), 54.53 (TiMe),
124.24 (C5Me5). 29Si NMR (298 K, C6D6): -110.8 (1Si, SiO4);
-66.0 (3Si); -65.9 (1Si); -65.8 (3Si). IR (ATR Si, cm-1): 2947
(m), 2911 (w, sh), 2863 (m), 1451 (w), 1378 (vw), 1246 (w), 1097
(vs), 1043 (m), 1024 (m, sh), 984 (s), 917 (w), 731 (vw), 676 (vw).
Anal. Calcd for C81H144O26Si16Ti (2031.29): C, 47.89; H, 7.14.
Found: C, 47.94; H, 7.18.
Data for 2b are as follows. 1H NMR (298 K, C6D6): 0.28 (s,
9H, SiMe3); 1.06-1.28 (m, 7H, CH, cyclopentyl), 1.16 (s, 3H,
TiMe); 1.44-1.62 (m, 14H, CH2, cyclopentyl); 1.62-1.86 (m,
28H, CH2, cyclopentyl); 1.86-2.03 (m, 14H, CH2, cyclopentyl);
1.93 (s, 15H, C5Me5). 13C{1H} NMR (298 K, C6D6) selected
signals: 2.29 (SiMe3); 11.67 (C5Me5); 23.10, 23.20, 24.10, 24.63,
25.43 (CH, cyclopentyl); 27.3-28.6 (CH2, cyclopentyl); 55.10
(TiMe); 122.86 (C5Me5). 29Si NMR (298 K, C6D6): -67.90 (2Si);
-67.25 (2Si); -66.66 (1Si); -65.67 (1Si); -64.74 (1Si); 7.78 (1Si,
SiMe3).
Preparation of [Cp*{(c-C5H9)7Si8O12O}2Ti](þ)[(μ-Me)B(C6-
F5)3](-) (3). Compound 1 (0.426 g, 0.21 mmol) was dissolved in
3.0 mL of toluene, and this solution was mixed with a solution of
B(C6F5)3 in toluene (2.0 mL of a 0.1 M solution) at room
temperature. The product formation was indicated by an im-
mediate color change from yellow to brownish-orange. All the
solvent was evaporated under vacuum, and the product was
dissolved in hexane (10 mL), leaving traces of brown oil. Orange
crystals of 3 were obtained by a slow vacuum distillation of the
solvent into an attached ampule at room temperature. The
residual mother liquor (ca. 1 mL) containing excess 1 was also
removed. A bright orange crystalline product was washed out
with condensing hexane vapors. After drying under vacuum
of 2 ꢁ 10-4 Torr at 60 °C an opaque pseudocrystalline solid was
obtained. Yield: 0.41 g (80%).
Data for 3 are as follows. 1H NMR (238 K, C7D8): 0.92 (br s,
3H, BMe), 1.02-1.28 (m, 14H, CH, cyclopentyl), 1.38-2.16 (m,
112H, CH2, cyclopentyl), 1.99 (s, 15H, C5Me5). 19F NMR (238
K, C7D8): -133.0 (br s, 6F, o-F); -159.3 (br s, 3F, p-F); -164.2
(br s, 2F, m-F). IR (ATR Si, cm-1): 2950 (s), 2909 (m, sh), 2864
(m), 1640 (vw), 1510 (w), 1458 (m), 1382 (vw), 1247 (w), 1101
(vs), 1006 (m), 972 (m), 932 (s), 919 (s), 798 (w), 732 (w), 670 (w),
662 (w). Anal. Calcd for C99H144O26Si16F15BTi (2543.30): C,
46.75; H, 5.71. Found: C, 46.70; H, 5.68.
ꢁ
ꢁ
ꢀ
(31) Antropiusova, H.; Dosedlova, A.; Hanus, V.; Mach, K. Transi-
tion Met. Chem. (London) 1981, 6, 90–93.
Preparation of [Cp*{(c-C5H9)7Si7O9(OSiMe3)O2}Ti](þ)[(μ-
Me)B(C6F5)3](-) (4a). Compound 2a (0.240 g, 0.21 mmol) in
3.0 mL of toluene reacted with the solution of B(C6F5)3
in toluene (2.0 mL of 0.1 M solution) in the same way as for 3.
(32) (a) Kalamarides, H. A.; Iyer, S.; Lipona, I.; Rhodes, L. F.; Day,
C. Organometallics 2000, 19, 3983–3990. (b) Beringhelli, T.; Maggioni, D.;
D'Alfonso, G. Organometallics 2001, 20, 4927–4938. (c) Bergquist, C.;
Bridgewater, B. M.; Harlan, C. J.; Norton, J. R.; Friesner, R. A.; Parkin, G.
J. Am. Chem. Soc. 2000, 122, 10581–10590.

6954 Organometallics, Vol. 28, No. 24, 2009
Varga et al.
After replacement of toluene with hexane an orange crystalline
material of 4a was obtained by crystallization from the latter.
The orange crystalline material of 4a was dried overnight at 2 ꢁ
10-4 Torr to give a light orange product. Yield: 0.27 g (82%).
Data for 4a are as follows. 1H NMR (298 K, C6D6): 0.20 (s,
9H, SiMe3); 0.85 (br s, 3H, BMe); 0.94-1.20 (m, 7H, CH,
cyclopentyl), 1.40-1.76 (m, 42H, CH2, cyclopentyl); 1.76-1.92
(m, 14H, CH2, cyclopentyl); 2.04 (s, 15H, C5Me5). 13C{1H}
NMR (298 K, C6D6): 2.26 (SiMe3); 12.61 (C5Me5); 22.03, 22.91,
23.76, 25.54 (CH, cyclopentyl); 27.22, 27.30, 27.37, 27.40, 27.51,
27.66, 27.88, 27.94, 28.00, 28.43 (CH2, cyclopentyl); 133.62
3.0 mL) was added to a solution of [C6H5NHMe2]þ[B(C6F5)4]-
(0.135 g, 0.17 mmol) in 20 mL of toluene to give a yellow
solution. After stirring the mixture for 1 h the solution was
evaporated under vacuum. The oily residue was repeatedly
extracted with 15 mL of hexane to give a yellow oil. This was
rinsed with 10 mL of hexane and dried under vacuum. Yield:
0.29 g (76%).
Data for 8 are as follows. IR (ATR Si, cm-1): 2951 (s), 2913
(w, sh), 2866 (m), 1641 (w), 1513 (m), 1462 (s), 1410 (vw), 1381
(w), 1276 (w), 1252 (w), 1105 (s, sh), 1083 (vs), 1058 (s), 1040 (s,
sh), 978 (vs), 948 (s), 911 (m), 860 (m), 844 (m), 773 (m), 755 (m),
727 (w), 699 (w), 683 (m), 661 (m). Anal. Calcd for C80H98O12-
Si8F20NBTi (1929.04): C, 49.81; H, 5.12. Found: C, 49.77; H,
5.08.
1
(C5Me5); 137.52 (d of multiplets, JCF ≈ 255 Hz, m-CF);
1
139.92 (d of multiplets, JCF ≈ 245 Hz, p-CF); 148.73 (d of
1
multiplets, JCF ≈ 235 Hz, o-CF). 19F NMR (298 K, C6D6):
-133.0 (br s, 6F, o-F); -159.7 (br s, 3F, p-F); -164.9 (br s, 2F,
m-F). 29Si NMR (298 K, C6D6): -66.61 (2Si); -65.02 (1Si);
The NMR spectra were assigned to syn-8a and anti-8b on the
basis of 1D NOESY experiments. The ratio 8a/8b was close
to 2:1.
-64.92 (1Si); -64.74 (1Si); -64.05 (2Si); 8.16 (1Si, SiMe3). 11
B
NMR (273 K, C7D8): -13 (Δν1/2=46 Hz). IR (ATR Si, cm-1):
2951 (m), 2912 (w, sh), 2865 (m), 1640 (w), 1511 (m), 1458 (s),
1393 (vw), 1382 (w), 1366 (vw), 1322 (vw), 1269 (w, sh),
1252 (m), 1112 (s, sh), 1085 (vs), 1061 (s, sh), 1044 (s, sh),
973 (s,sh), 964 (vs), 953 (s, sh), 909 (w, sh), 868 (s), 847 (m),
800 (w), 758 (w), 734 (w), 660 (w). Anal. Calcd for C67H90-
O12Si8F15BTi (1655.83): C, 48.60; H, 5.48. Found: C, 48.56;
H, 5.52.
Data for 8a are as follows. 1H NMR (298 K, C6D6): 0.21 (s,
9H, SiMe3); 1.00-1.20 (m, 7H, CH, cyclopentyl); 1.42-1.77 (m,
42H, CH2, cyclopentyl); 1.81 (s, 15H, C5Me5); 1.80-1.98 (m,
14H, CH2, cyclopentyl); 2.55 (br s, 6H, NMe2); 7.08-7.26 (m,
5H, Ph). 13C{1H} NMR (298 K, C6D6): 2.10 (SiMe3); 12.76
(C5Me5); 22.73, 22.79, 23.31, 23.55, 25.38 (CH, cyclopentyl);
27.00-28.50 (CH2, cyclopentyl); 51.48 (NMe2Ph); 118.18,
120.95, 130.39 (CH, NMe2Ph); 135.94 (C5Me5); 137.06 (d of
Reaction of 1 with [C6H5NHMe2]þ[B(C6F5)4]- to Give
[Cp*{(c-C5H9)7Si8O12O}2Ti(NMe2Ph-κ-N)]þ[B(C6F5)4]- (7).
A solution of 1 (0.426 g, 0.21 mmol) in 5.0 mL of toluene under
argon atmosphere was added to the solid [C6H5NHMe2]þ-
[B(C6F5)4]- (0.160 g, 0.20 mmol), and the mixture was stirred
for 5 h. All volatiles were distilled off under vacuum, and the
orange residue was extracted repeatedly with 20 mL of hexane.
Evaporation of combined extracts afforded an amorphous
solid. This was rinsed with condensing hexane vapor and dried
under vacuum to give a yellow powder. Yield: 0.42 g (75%).
Data for 7 are as follows. 1H NMR (298 K, C6D6): 1.05-1.28
(m, 14H, CH, cyclopentyl), 1.40-2.00 (m, 56H, CH2,
cyclopentyl); 1.83 (s, 15H, C5Me5); 2.71 (br s, 6H, NMe2);
multiplets, 1JCF=243 Hz, m-CF); 138.93 (d of multiplets, 1JCF
=
246 Hz, p-CF); 146.96 (Cipso, NMe2Ph); 149.17 (d of multiplets,
1JCF=242 Hz, o-CF). 29Si{1H} NMR (99.3 MHz, 298 K, C6D6):
-65.89 (2Si); -65.32 (1Si); -64.76 (1Si); -64.38 (1Si); -63.77
(2Si); 8.98 (1Si, SiMe3).
Data for 8b are as follows. 1H NMR (300 MHz, 298 K, C6D6):
-0.18 (s, 9H, SiMe3); 1.00-1.20 (m, 7H, CH, cyclopentyl),
1.42-1.77 (m, 42H, CH2, cyclopentyl); 1.68 (s, 15H, C5Me5);
1.95-2.08 (m, 14H, CH2, cyclopentyl); 2.97 (br s, 6H, NMe2);
7.08-7.26 (m, 5H, Ph). 13C{1H} NMR (75 MHz, 298 K, C6D6):
1.55 (SiMe3); 12.23 (C5Me5); 22.63, 24.82, 25.19, remaining
signals are overlapped by signals of the more abundant stereo-
isomer (CH, cyclopentyl); 27.00-28.50 (CH2, cyclopentyl);
3
52.27 (NMe2Ph); 120.86, 129.13, one signal was not found
7.04 (t, JHH = 7.5 Hz, 1H, CHpara, Ph); 7.17-7.22 (m, 2H,
1
(CH, NMe2Ph); 135.05 (C5Me5); 137.06 (d of multiplets, JCF
Ph); 7.28-7.35 (m, 2H, Ph). 13C{1H} NMR (298 K, C6D6):
12.29 (C5Me5); 21.94, 22.38 (CH, cyclopentyl); 27.48, 27.70,
27.91 (CH2, cyclopentyl); 50.43 (NMe2Ph); 118.22, 121.00,
130.16 (CH, NMe2Ph); 137.01 (d of multiplets, 1JCF=236 Hz,
m-CF); 137.14 (C5Me5); 138.80 (d of multiplets, 1JCF=241 Hz,
=
243 Hz, m-CF); 138.93 (d of multiplets, JCF =246 Hz, p-CF);
1
1
146.62 (Cipso, NMe2Ph); 149.17 (d of multiplets, JCF =242 Hz,
o-CF). 29Si{1H} NMR (99.3 MHz, 298 K, C6D6): -66.26 (2Si);
-65.15 (1Si); -65.08 (1Si); -64.93 (2Si); -64.55 (1Si); 9.26 (1Si,
SiMe3).
p-CF); 146.05 (Cipso, NMe2Ph); 149.05 (d of multiplets, 1JCF
=
Reaction of 1 with [Ph3C]þ[B(C6F5)4]- to Give Impure [Cp*{(c-
C5H9)7Si8O12O}2Ti]þ[B(C6F5)4]- (9). A solution of 7 (0.426 g,
0.21 mmol) in 5.0 mL of toluene was added under an argon
atmosphere to solid [Ph3C]þ[B(C6F5)4]- (0.184 g, 0.20 mmol),
and the mixture was stirred for 5 h. All volatiles were distilled off
under vacuum, and the red-brown oily residue was extracted
repeatedly with 20 mL of hexane to collect a low-soluble red-
brown oil (a remaining brown oily solid residue insoluble in
hexane was discarded). Evaporation of combined extracts af-
forded an oil, which, upon warming to ca. 40 °C, suddenly
released the solvent, forming a solid foam. The foam disap-
peared after condensing hexane vapor on it to form an oily solid
sticking to the ampule walls. The hexane extract was separated,
and the residue dried under vacuum, forming a foam again. The
solid product was then mechanically scraped from the walls and
collected. Yield: 0.34 g (63%).
243 Hz, o-CF). 19F NMR (298 K, C6D6): -132.3 (br s, 6F, o-F);
-163.4 (m, 3F, p-F); -167.1 (br s, 2F, m-F). 29Si NMR (298 K,
C6D6): -109.47 (1Si, SiO4); -65.65 (3Si); -65.63 (1Si); -65.24
(3Si). IR (ATR Si, cm-1): 2949 (m), 2912 (w, sh), 2865 (m), 1642
(vw), 1512 (m), 1462 (s), 1452 (m, sh), 1274 (w), 1245 (w), 1100
(vs, b), 1028 (w), 1005 (m, b), 978 (s), 938 (m, b), 919 (m, b), 771
(w), 754 (w), 683 (w), 661 (w). Anal. Calcd for C112H152O26-
Si16F20NBTi (2816.51): C, 47.76; H, 5.44. Found: C, 47.72; H,
5.47.
Reaction of 1 with [C6H5NHMe2]þ[B(C6F5)4]- in an NMR
Tube. A solution of 1 (0.163 g, 0.08 mmol) in 0.8 mL of C6D6 was
added under argon atmosphere to solid [C6H5NHMe2]þ-
[B(C6F5)4]- (0.095 g, 0.12 mmol) in an NMR tube, and this
was closed with a screw-tightened Teflon cap. Measurements of
1H, 13C, and 19F spectra after 20 min intervals revealed the
formation of the ion-pair complex [Cp*{(c-C5H9)7Si8O12O}2Ti-
(C6H5NMe2)]þ[B(C6F5)4]- (7) at the expense of 1. Methane
Data for 9 are as follows. 1H NMR (298 K, C6D6): 1.08-1.36
(m, 14H, CH, cyclopentyl), 1.44-2.06 (m, 112H, CH2, cyclo-
pentyl), 2.03 (s, 15H, C5Me5). IR (ATR Si, cm-1): 2949 (s), 2913
(w, sh), 2866 (m), 1643 (vw), 1512 (w), 1463 (s), 1275 (vw), 1246
(w), 1103 (vs), 1028 (m), 1001 (m), 978 (s), 934 (m, sh), 917 (s, b),
775 (w), 755 (w), 700 (w), 682 (w), 660 (w).
Reaction of 9 with PhNMe2 to Give 7. Solid 9 (0.18 g) was
dissolved in toluene (3.0 mL), and PhNMe2 (0.3 mL) was added.
After stirring for 30 min at ambient temperature all volatiles
1
evolved during the reaction, giving the H NMR signal at δH
0.15 ppm. Consumption of 1 was complete after 90 min, while an
excess of the low-soluble [C6H5NHMe2]þ[B(C6F5)4]- remained
at the bottom of the tube and was not observed in the NMR
spectra.
Reaction of 2 with [C6H5NHMe2]þ[B(C6F5)4]- to Give [Cp*
{(c-C5H9)7Si7O9(OSiMe3)O2}Ti(NMe2Ph-κ-N)]þ[B(C6F5)4]- (8).
A solution of 2 (a mixture of 2a and 2b) in toluene (0.2 mmol in

Article
Organometallics, Vol. 28, No. 24, 2009 6955
were evaporated under vacuum, at 2 ꢁ 10-4 Torr and 60 °C. A
dark orange solid was dissolved in C6D6. The 1H and 19F NMR
spectra of the product identified compound 7 to be by far the
main component (>85%).
Table 5. Crystal Data and Structure Refinement for Compound 4a
empirical formula
fw
temperature
C68.50H93.50BF15O12Si8Ti
1677.36
150(2) K
Reaction of 2 with [Ph3C]þ[B(C6F5)4]- to Give Impure [Cp*{(c-
C5H9)7Si7O9(OSiMe3)O2}Ti]þ[B(C6F5)4]- (10). A solution of
2 in toluene (0.21 mmol in 3.0 mL) was added to a suspension
of solid [Ph3C]þ[B(C6F5)4]- (0.19 g, 0.20 mmol) in 10 mL of
toluene, and the mixture was stirred at room temperature for 3 h
to give an orange-brown solution. All volatiles were distilled off
under vacuum, and the red-brown oily residue was extracted
repeatedly with 20 mL of hexane to collect a red-brown oil
poorly soluble in hexane (a remaining brown, oily solid residue
insoluble in hexane was discarded). The hexane from combined
extracts was partly evaporated under vacuum, and the mother
liquor (10 mL) was separated from an orange-brown oily
residue. The hexane mother liquor contained, according to IR
spectra, mainly Ph3CMe. The oily residue was dried under a
vacuum of 2 ꢁ 10-4 Torr, at 60 °C, to give a brown solid
containing, according to 1H NMR, mainly a mixture of isomers
10a and 10b in the ratio 2:1. The presence of minor singlet signals
in the 1H NMR spectra close to resonances for C5Me5 and
SiMe3 indicated the presence of impurities, which were not
identified. Yield: 0.30 g (83%).
cryst syst, space group
unit cell dimens
triclinic, P1 (No. 2)
a 14.5433(3) A, R 104.1142(11)°
b 15.8531(3) A, β 92.3741(9)°
c 19.9883(4) A, γ 113.0560(11)°
4063.61(14) A3
volume
Z, calcd density
absorp coeff μ
F(000)
cryst size
θmin; θmax
limiting indices
2, 1.371 Mg/m3
0.313 mm-1
1749
0.5 ꢁ 0.3 ꢁ 0.075 mm3
2.03°; 27.49°
-13 e h e 18, -20 e k e 20,
-25 e l e 25
68 319/18 362
reflns collected/unique
[R(int) = 0.0433]
98.6%
952
completeness to θ = 27.49
parameters
goodness-of-fit on F2
final R indices [I > 2σ(I)]
R indices (all data)
1.042
R1 = 0.0535, wR2 = 0.1382
R1 = 0.1203, wR2 = 0.1624
0.877 and -0.593 e A3
largest diff peak and hole
to activated charcoal cooled with liquid nitrogen. The residue
(yellow solid for 2, yellow oil for 1 or for Cp*Ti(OtBu)2Me) was
dissolved in 1.4 mL of C6D6 or C7D8, and the solution was
poured into an NMR sample tube, sealed with flame, and
analyzed by 1H and 13C NMR spectroscopy.
The 1/2AlMe3 system yielded Cp*TiMe3 (TiMe δH/δC 1.00/
61.16 ppm)27 and most probably Cp*Ti(SIPOSS)Me2, with
TiMe δH/δC 0.76/55.93 ppm, C5Me5 δH/δC 1.86/11.67 ppm,
and C5Me5 at δC 123.05 ppm. Integrals for TiMe resonances
indicated contents of these compounds as 32% and 60%,
whereas other minor signals observed using qHSQC indicated
the presence of trace amounts of 1 and two other products in the
total amount of 8%. In the AlMe resonance range -0.3 to 0.4
ppm, the characteristic signals of the AlMe2 group at δH/δC
-0.14/-7.10 ppm for (SIPOSSAlMe2)n were observed33 besides
other signals at δH/δC -0.29/-13.05, -0.24/-8.67, 0.17/0.50,
and 0.38/0.50, which were not identified.
Data for 10 are as follows. 1H NMR (298 K, C6D6) 10a: 0.19
(s, 9H, SiMe3), 0.94-1.14 (m, 7H, CH, cyclopentyl), 1.42-2.02
(m, 56H, CH2, cyclopentyl), 1.83 (s, 15H, C5Me5). 10b: 0.09 (br
s, ν1/2= 8.5 Hz, 9H, SiMe3), 0.94-1.14 (m, 7H, CH, cyclo-
pentyl), 1.42-2.02 m, 56H, CH2, cyclopentyl), 1.94 (s, 15H,
C5Me5). IR (ATR Si, cm-1): 2951 (s), 2913 (w, sh), 2865 (m),
1642 (w), 1512 (m), 1463 (s), 1410 (vw), 1382 (w), 1275 (w), 1254
(w), 1114 (s, sh), 1085 (vs), 1038 (m), 978 (vs), 951 (s, b), 909 (m,
sh), 860 (m, b), 774 (m), 755 (m), 727 (w), 683 (m), 661 (m).
Reaction of 10 with PhNMe2 to Give 8. Solid 10 (0.14 g) was
dissolved in toluene (3.0 mL), and N,N-dimethylaniline (0.3 mL)
was added. After stirring for 30 min at ambient temperature all
volatiles were evaporated under vacuum, at 2 ꢁ 10-4 Torr at
60 °C. A dark orange solid was dissolved in C6D6. The 1H and
19F NMR spectra of the product revealed that a mixture of
compounds 8a and 8b in about 2:1 abundance formed the main
component (>80%).
Styrene Polymerization. The reaction components were dosed
to the evacuated reactor equipped with a magnetic stirring bar in
the order styrene (2.0 mL), the titanium precursor (Cp*Ti-
(OtBu)2Me, 2, or 1, 10 μmol in 2.0 mL of toluene), B(C6F5)3
(10 μmol in 2.0 mL of toluene), and AlMe3 (20 μmol in 2.0 mL of
toluene). The polymerization mixture was cooled to about 5 °C
and sealed off with a flame. The reactor was immersed into a
thermostated water bath, and its contents were rapidly stirred.
After 30 min the reactor was cooled to 20 °C and opened to air,
and the contents were poured into ethanol (50 mL) acidified
with aqueous HCl. The reactor was washed with ethanol/HCl,
and washings with the solid polymer were combined with the
main product part. After stirring for 1 h the polymer was filtered
off, washed repeatedly with ethanol, and dried in a vacuum oven
at 120 °C to constant weight. The syndiotacticity was deter-
mined by polymer extraction with boiling 2-butanone for 6 h
followed by drying of the residue until constant weight. The
melting temperature, Tm, was determined by DSC analysis, and
Mw and Mn were determined by the GPC method. In prelimin-
ary polymerizations with MAO the above titanium precursors
were activated with 1.66 mL of a toluene solution of MAO ([Al]=
1.51 M).
The 2/2AlMe3 system afforded a wide range of products,
where only signals of Cp*TiMe3 and trace amounts of 2 were
identified. The Cp*TiMe3 content of 14% was calculated from
the integral of the peak at δH 1.00 ppm with respect to the sum of
integrals in the TiMe range 0.58-0.92 ppm.
The Cp*Ti(OtBu)2Me/2AlMe3 system afforded the following
titanium compounds: Cp*Ti(OtBu)Me2 (δH 1.44 (TiMe2) and
1.82 (C5Me5) ppm in integral ratio 6:15) (70%) and Cp*Ti-
(OtBu)2Me (δH 1.30 (TiMe) and 1.94 (C5Me5) ppm) (30%).11
The presence of Cp*TiMe3 was observed after this system was
heated to 100 °C for 1 h: Cp*TiMe3 5%, (Cp*Ti(OtBu)Me2
81%, and Cp*Ti(OtBu)2Me 14%.
X-ray Crystallography. Compound 4a was recrystallized from
hexane at room temperature to give thin orange plates. These
were transferred to a glovebag in a saturated hexane solution
under argon, immersed in Nujol oil, and mounted onto a
measuring rod of a diffractometer under a stream of cooled
nitrogen. Diffraction data were collected on a Nonius Kap-
paCCD diffractometer (Mo KR radiation, λ = 0.71073 A) at
150(2) K and processed by the HKL program package.34 The
phase problem was solved by direct methods (SIR97),35
followed by consecutive Fourier syntheses, and refined by
Reactions of 1, 2, and [Cp*Ti(OtBu)2Me] with 2 equiv of
AlMe3. A solution of titanium compound in toluene (50 μmol
in 1.0 mL) was mixed with a solution of AlMe3 in toluene (100
μmol in 1.0 mL) in a sealed ampule, and this was heated to 60 °C
in a water bath for 1 h. Then all volatiles including AlMe3 and
methylaluminum tert-butoxides were evaporated on a vacuum
line at 60 °C, then the very low-volatile products were condensed
(33) Skowronska-Ptasinska, M. D.; Duchateau, R.; van Santen,
R. A.; Yap, G. P. A. Organometallics 2001, 20, 3519–3530.
(34) Otwinowski, Z.; Minor, W. HKL Denzo and Scalepack program
package by Nonius. For a reference, see: Otwinowski, Z.; Minor, W.
Methods Enzymol. 1997, 276, 307–326.
(35) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G.;
Giacovazzo, C.; Guagliardi, A.; Polidori, G. J. Appl. Crystallogr.
1994, 27, 435–436.

6956 Organometallics, Vol. 28, No. 24, 2009
Varga et al.
full-matrix least-squares on F2 (SHELXL-97).36 Relevant crys-
tallographic data are given in Table 5. All non-hydrogen atoms
were refined anisotropically except those with partial occupancy
factors. The hydrogen atoms were recalculated into idealized
map indicate only the partial presence of the whole solvating
molecule in the unit cell. The hexane was partly released from
the crystal during its transfer on the measuring rod.
positions (riding model) and assigned temperature factors Hiso
-
Acknowledgment. This research was supported by the
Academy of Sciences of the Czech Republic (Grant No.
KAN100400701) and the Ministry of Education, Youth
and Sports (Project No. LC06070). I.C. thanks MSM-
0021620857.
(H)=(1.2-1.5)Ueq(pivot atom), except for the hydrogen atoms
on bridging methyl carbon atom C(49). These were found on
difference Fourier maps and refined isotropically without any
restraints. Several cyclopentyl moieties were disordered into two
positions with partial occupancy; the atoms involved in this
disorder were refined isotropically. The solvating hexane mole-
cule occupying the center of symmetry in the unit cell was
disordered due to flexibility of its linear chain. Moreover, the
low electron densities of corresponding maxima on the Fourier
Supporting Information Available: CIF file for the structure
4a (XRAY.CIF). Supporting Information and PLATON draw-
ing at 30% probability for structure 4a, IR (ATR-Si) spectra of
1
1, 3, 7, and 9 (Figure 1); 2, 4a, 8, and 10 (Figure 2), H NMR
spectra of reacting 1 with the solid [C6H5NHMe2]þ[B(C6F5)4]-
to give 7, and 19F NMR spectra of 8 at 25-95 °C. This material
is available free of charge via the Internet at http://pubs.acs.org.
(36) Sheldrick, G. M. SHELXL97, Program for Crystal Structure
€
€
Refinement from Diffraction Data; University of Gottingen: Gottingen,
1997.