Dinuclear dicyclopentadienyl titanium complexes with bridging cyclopentadienylsiloxo ligands

Article abstract of DOI:10.1021/om900938m

Addition of 1 equiv of T1Cp to compound [(TiCl₂) ₂(μ-{(η⁵-C₅Me₄SiMeO) ₂(μ-O)})] (A) at 80°C gave a mixture with different molar ratios of the two possible isomers of [(TiClC_p)(TiCl ₂)(μ-{(η⁵C₅Me₄SiMeO) ₂(μ-O)})] (1). Reaction of compound A with 2 equiv of TlC _p at 80°C afforded [(TiClC_p)₂(μ- {(η⁵-C₅Me₄SiMeO)₂(μ-O)})] (2as) as the unique reaction product. Each Cp ligand of 2as is located in different positions anti and syn with respect to the Si-O-Si bridge. However, a mixture of two isomers of [(TiClCp)₂(μ-{(η⁵-C ₅Me₄SiMeO)₂(μ-O)})], 2as ( > 95% by NMR) and 2aa ( < 5% by NMR), was obtained when the reaction was carried out at 60 °C; the two Cp ligands of 2aa were located at anti positions. Alkylation of isomer 2as with LiMe gave a mixture of the three possible isomers [(TiMeCp) ₂(μ-{(η⁵-C₅Me₄SiMeO) ₂(μ-O)})] (3as, 3aa, 3ss). The proportion of each was dependent on the reaction temperature. Isomer 2as reacted with Lewis acids E(C ₆Fs)₃ (E = B, Al) and with Li[B(C₆F ₅)₄] to give the chloro-bridged dititanium compounds [(TiCp)₂(μ-{(η⁵-C₅Me₄SiMeO) ₂(μ-O)})μ-Cl)][Q] (Q = ClB(C₆F₅) ₃, 4B; C1A1(C₆F₅)₃, 4Al; Q = B(C₆F₅)₄, 4C) as the unique reaction products. Addition of [Ph₃C][B(CgF₅)₄] to the mixture of isomers in 3 gave a mixture of complexes [(TiCpMe)(TiCp)(μ- {(η⁵-C₅Me₄SiMeO)₂(μ-O)})][Q] (5aC and 5sC; Q = B(C₆F₅)₄), with the remaining methyl ligand located at anti or syn positions depending on the methyl group being abstracted. DFT studies were carried out to determine the stability of the isomers of 2 and 3, to determine which chlorine atom in compound 2as was more easily eliminated, and also to clarify the transformation of isomer 2as into the mixture of isomers 3as, 3aa, and 3ss during the alkylation reaction.

Full text of DOI:10.1021/om900938m

642 Organometallics 2010, 29, 642–655
DOI: 10.1021/om900938m
Dinuclear Dicyclopentadienyl Titanium Complexes with Bridging
Cyclopentadienylsiloxo Ligands
†
Lorena Postigo, Luca Bellarosa, Javier Sanchez-Nieves, Pascual Royo,*
‡
†
,†
ꢀ
,‡
†,§
ꢀ
Agustı Lledos,* and Marta E. G. Mosquera
´
†
´
Departamento de Quımica Inorganica, Universidad de Alcala, Campus Universitario, E-28871 Alcala de
Henares (Madrid), Spain and Departament de Quımica, Universitat Autonoma de Barcelona, E-08193
ꢀ
ꢀ
ꢀ
‡
´
ꢁ
Bellaterra (Barcelona), Spain. §X-ray diffraction studies.
Received October 26, 2009
Addition of 1 equiv of TlCp to compound [(TiCl2)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (A) at 80 ꢀC
gave a mixture with different molar ratios of the two possible isomers of [(TiClCp)(TiCl2)(μ-{(η5-
C5Me4SiMeO)2(μ-O)})] (1). Reaction of compound A with 2 equiv of TlCp at 80 ꢀC afforded
[(TiClCp)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (2as) as the unique reaction product. Each Cp ligand of
2as is located in different positions anti and syn with respect to the Si-O-Si bridge. However, a
mixture of two isomers of [(TiClCp)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})], 2as (>95% by NMR) and 2aa
(<5% by NMR), was obtained when the reaction was carried out at 60 ꢀC; the two Cp ligands of 2aa
were located at anti positions. Alkylation of isomer 2as with LiMe gave a mixture of the three possible
isomers [(TiMeCp)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (3as, 3aa, 3ss). The proportion of each was
dependent on the reaction temperature. Isomer 2as reacted with Lewis acids E(C6F5)3 (E=B, Al) and
with Li[B(C6F5)4] to give the chloro-bridged dititanium compounds [(TiCp)2(μ-{(η5-C5Me4SiMeO)2-
(μ-O)})(μ-Cl)][Q] (Q=ClB(C6F5)3, 4B; ClAl(C6F5)3, 4Al; Q=B(C6F5)4, 4C) as the unique reaction
products. Addition of [Ph3C][B(C6F5)4] to the mixture of isomers in 3 gave a mixture of complexes
[(TiCpMe)(TiCp)(μ-{(η5-C5Me4SiMeO)2(μ-O)})][Q] (5aC and 5sC; Q=B(C6F5)4), with the remain-
ing methyl ligand located at anti or syn positions depending on the methyl group being abstracted.
DFT studies were carried out to determine the stability of the isomers of 2 and 3, to determine which
chlorine atom in compound 2as was more easily eliminated, and also to clarify the transformation of
isomer 2as into the mixture of isomers 3as, 3aa, and 3ss during the alkylation reaction.
Introduction
behavior with respect to related mononuclear systems.2-18
Although this cooperative effect is well documented, dicatio-
nic dinuclear group 4 complexes are mainly obtained from
cationic mononuclear derivatives.19-22
Functionalization of cyclopentadienyl rings with chlorosilyl
groups23-30 has provided a particularly useful procedure to
The design of ligands to generate bimetallic systems has
been developed with the aim of finding reactivity patterns
different from those observed for similar monometallic
complexes.1 In this regard, the cooperative effect between
the metal atoms in group 4 dinuclear complexes has been
shown to induce important modifications in polymerization
(13) Alt, H. G.; Ernst, R. Inorg. Chim. Acta 2003, 350, 1.
(14) Li, L. T.; Metz, M. V.; Li, H. B.; Chen, M. C.; Marks, T. J.;
Liable-Sands, L.; Rheingold, A. L. J. Am. Chem. Soc. 2002, 124, 12725.
(15) Alt, H. G.; Ernst, R.; Bohmer, I. K. J. Organomet. Chem. 2002,
658, 259.
*Corresponding authors. Fax: 00 34 93 581 2920. E-mail: pascual.royo@
uah.es; agusti@klingon.uab.es.
(1) Gavrilova, A. L.; Bosnich, B. Chem. Rev. 2004, 104, 349.
(2) Salata, M. R.; Marks, T. J. Macromolecules 2009, 42, 1920.
(3) Motta, A.; Fragala, I. L.; Marks, T. J. J. Am. Chem. Soc. 2009,
131, 3974.
(4) Salata, M. R.; Marks, T. J. J. Am. Chem. Soc. 2008, 130, 12.
(5) Guo, N.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 2008, 130,
2246.
(6) Lee, S. H.; Wu, C. J.; Joung, U. G.; Lee, B. Y.; Park, J. Dalton
Trans. 2007, 4608.
(7) Lee, M. H.; Kim, S. K.; Do, Y. Organometallics 2005, 24, 3618.
(8) Stojcevic, G.; Kim, H.; Taylor, N. J.; Marder, T. B.; Collins, S.
Angew. Chem., Int. Ed. 2004, 43, 5523.
(9) Guo, N.; Li, L. T.; Marks, T. J. J. Am. Chem. Soc. 2004, 126, 6542.
(10) Sierra, J. C.; Huerlander, D.; Hill, M.; Kehr, G.; Erker, G.;
(16) Noh, S. K.; Kim, J.; Jung, J.; Ra, C. S.; Lee, D.-h.; Lee, H. B.;
Lee, S. W.; Huh, W. S. J. Organomet. Chem. 1999, 580, 90.
(17) Chen, Y. X.; Metz, M. V.; Li, L. T.; Stern, C. L.; Marks, T. J.
J. Am. Chem. Soc. 1998, 120, 6287.
(18) Li, Y. F.; Ward, D. G.; Reddy, S. S.; Collins, S. Macromolecules
1997, 30, 1875.
(19) Chen, M. C.; Roberts, J. A. S.; Seyam, A. M.; Li, L.; Zuccaccia,
C.; Stahl, N. G.; Marks, T. J. Organometallics 2006, 25, 2833.
(20) Lian, B.; Toupet, L.; Carpentier, J.-F. Chem.;Eur. J. 2004, 10,
4301.
(21) Sadow, A. D.; Tilley, T. D. J. Am. Chem. Soc. 2003, 125, 9462.
(22) Yang, X.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1994, 116,
10015.
ꢀ
(23) Postigo, L.; Sanchez-Nieves, J.; Royo, P. Inorg. Chim. Acta
€
Frohlich, R. Chem.;Eur. J. 2003, 9, 3618.
2007, 360, 1305.
(24) Santamarıa, D.; Cano, J.; Royo, P.; Mosquera, M. E. G.;
´
(11) Noh, S. K.; Lee, J.; Lee, D. H. J. Organomet. Chem. 2003, 667,
~
53.
Cuenca, T.; Frutos, L. M.; Castano, O. Angew. Chem., Int. Ed. 2005,
44, 5828.
(12) Alt, H. G.; Ernst, R. J. Mol. Catal. A: Chem. 2003, 195, 11.
r
pubs.acs.org/Organometallics
Published on Web 01/07/2010
2010 American Chemical Society

Article
Organometallics, Vol. 29, No. 3, 2010 643
derivatives were isolated. However, polymerization of
ε-caprolactone with [{Ti(OiPr)2}2(μ-{(η5-C5Me4SiMeO)2-
(μ-O)})] was influenced by the cooperative effect of both
metal atoms, obtaining PCL with double the Mw of that
obtained with a related mononuclear compound.38
In this paper we report our ongoing research on dinuclear
compounds with bridging cyclopentadienylsiloxo ligands
derived from compound A (Figure 1). New dinuclear di-
cyclopentadienyl complexes have been synthesized, and their
reactivity with different molar ratios of Lewis acids has been
explored, as the dinuclear nature of these new compounds
could aid the synthesis of dinuclear dicationic derivatives.
Figure 1. Diagram of complex A labeling the two different anti
(Cla) and syn (Cls) positions at the Ti atom with respect to the
orientation of the Si-O-Si bridge.
Results and Discussion
synthesize dinuclear compounds through hydrolysis of the
Si-Cl bonds.25,30-36 Following this procedure, the hydrolysis
of the dichlorosilyl-cyclopentadienyl titanium compound
[Ti{η5-C5Me4(SiMeCl2)}Cl3]26 allowed us to obtain the diti-
tanium derivative [(TiCl2)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})]
(A) with both Si-O-Si and Si-O-Ti bridges (Figure 1).37
The intrinsically diastereoselective formation of the
Si-O-Si bridge is responsible for the generation of only
one C2 symmetry diastereoisomer of A despite the presence
of two diastereogenic Si atoms. Furthermore, the Si-O-Si
bridge causes the two chlorine atoms to occupy different
positions, one toward the bridge (syn position) and the other
away from it (anti position).
Complex A has two potentially active metal centers for
olefin polymerization. However, we observed in previous
works that the ease of formation of bridges caused complex
A to be inactive for ethylene polymerization, although the
methyl derivative [(TiMe2)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})]
was active for MMA polymerization.37 Incomplete abstrac-
tion of two alkyl groups was possible only in the reaction of
the benzyl derivative [(TiBz2)2(μ-{(η5-C5Me4SiMeO)2(μ-
O)})] with excess B(C6F5)3 for several days.37 We synthesized
dialkoxo-bridged complexes [(TiX)2(μ-O2L)(μ-{(η5-C5Me4-
SiMeO)2(μ-O)})] (X=Cl, Me, Bz) to avoid the formation of
halide or alkyl bridges between the metal centers, thus
favoring the generation of dicationic dinuclear com-
pounds.38 Once again, we were unable to characterize any
dicationic compounds, and only monocationic dinuclear
Reactions of A with TlCp. Reaction of compound A with 1
equiv of TlCp in toluene at 80 ꢀC for 48 h afforded a mixture
containing unreacted complex A (25% by 1H NMR) together
with a very small amount of the dicyclopentadienyl derivative
1
2as (8% by H NMR) and a mixture of two components
corresponding to the two possible isomers of the monosub-
stitued dinuclear compound [(TiClCp)(TiCl2)(μ-{(η5-C5Me4-
SiMeO)2(μ-O)})] (1 anti and syn) (Scheme 1). These two
components containing the Cp ring located at the syn or anti
positions with respect to the orientation of the Si-O-Si bridge
are present in significantly different molar ratios (major isomer
50% by 1H NMR; minor isomer 17% by 1H NMR). Although
unfortunately we could not assign the syn and anti isomers to
the major and minor isomer components of this mixture, it is
clear that the introduction of the first cyclopentadienyl ligand
in compound A is a ca. 75% regioselective reaction for one
preferred position. The 1H and 13C NMR spectra of this
mixture are consistent with the presence of two compounds
that are asymmetric molecules, and each shows eight reso-
nances for two C5Me4Si rings, two resonances for both SiMe
groups, and one for the Cp ligand.
Reaction of [(TiCl2)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (A)
with 2 equiv of TlCp in toluene at 80 ꢀC gave [(TiClCp)2-
(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (2as) as the only titanium
product in high yield (Scheme 1). The Cp ligands of 2as are
located at opposite anti-syn positions with respect to the
Si-O-Si bridge, corresponding to the more stable isomer, as
this orientation probably minimizes steric repulsions (vide
infra). Compound 2as is an asymmetric molecule, and thus,
its 1H and 13C NMR spectra showed eight resonances for two
C5Me4Si rings and two resonances for both SiMe groups and
also for both Cp ligands. The relative position anti-syn for
each Cp ligand was determined by NOE NMR experiments
(Figure 2), taking into account that the structures of com-
plexes A and 2aa show that the ligand at a syn position is
closer to the SiMe group of the Ti-O-Si moiety to which
this Cp ligand is bound.
€
ꢀ
(25) Arteaga-Muller, R.; Sanchez-Nieves, J.; Royo, P.; Mosquera,
M. E. G. Polyhedron 2005, 24, 1274.
ꢀ
(26) Vazquez, A. B.; Royo, P.; Herdtweck, E. J. Organomet. Chem.
2003, 683, 155.
ꢀ
(27) Vazquez, A. B.; Royo, P. J. Organomet. Chem. 2003, 671, 172.
(28) Cano, J.; Royo, P.; Lanfranchi, M.; Pellinghelli, M. A.;
Tiripicchio, A. Angew. Chem., Int. Ed. 2001, 40, 2495.
ꢀ
(29) Alcalde, M. I.; Gomez-Sal, M. P.; Royo, P. Organometallics
1999, 18, 546.
(30) Alcalde, M. I.; Gomez-Sal, P.; Martın, A.; Royo, P. Organome-
´
ꢀ
tallics 1998, 17, 1144.
(31) Buitrago, O.; Jimenez, G.; Cuenca, T. J. Organomet. Chem. 2003,
ꢀ
Therefore, introduction of the second cyclopentadienyl
ligand at 80 ꢀC is regiospecific, as this second ligand is always
located opposite the first Cp ligand, regardless of its previous
position with respect to the Si-O-Si bridge.
683, 70.
ꢀ
(32) Alcalde, M. I.; Gomez-Sal, M. P.; Royo, P. Organometallics
2001, 20, 4623.
(33) Park, J. T.; Yoon, S. C.; Bae, B. J.; Seo, W. S.; Suh, I. H.; Han,
T. K.; Park, J. R. Organometallics 2000, 19, 1269.
(34) de la Mata, F. J.; Giner, P.; Royo, P. J. Organomet. Chem. 1999,
572, 155.
(35) Ciruelos, S.; Cuenca, T.; Gomez, R.; Gomez-Sal, P.; Manzanero,
A.; Royo, P. Organometallics 1996, 15, 5577.
ꢀ
(36) Ciruelos, S.; Cuenca, T.; Gomez-Sal, P.; Manzanero, A.; Royo,
P. Organometallics 1995, 14, 177.
(37) Postigo, L.; Vazquez, A. B.; Sanchez-Nieves, J.; Royo, P.;
Herdtweck, E. Organometallics 2008, 27, 5588.
(38) Postigo, L.; Sanchez-Nieves, J.; Royo, P.; Mosquera, M. E. G.
Dalton Trans. 2009, 3756.
However, a mixture of two isomers of [(TiClCp)2(μ-{(η5-
C5Me4SiMeO)2(μ-O)})] was obtained when the reaction of A
with 2 equiv of TlCp was carried out at 60 ꢀC. This mixture
contains the isomer 2as as the major component (>95% by
NMR) along with a very small amount (<5% by NMR) of a
second component identified as the 2aa isomer. Compound
2aa showed both Cp ligands located at the anti positions.
The 1H NMR spectrum of this mixture showed, apart from
ꢀ
ꢀ
ꢀ

644 Organometallics, Vol. 29, No. 3, 2010
Postigo et al.
Figure 2. 2D NOE NMR spectrum of compound 2as showing the interactions between Cp ligands and SiMe groups.
Scheme 1a
a (i) TlCp, 80 ꢀC ; (ii) 2 TlCp, 80 ꢀC; (iii) 2 TlCp, 60 ꢀC.
the resonances belonging to 2as, new resonances correspond-
ing to the new C2 symmetry isomer 2aa, which contains two
equivalent asymmetric titanium units. Thus, four resonances
for two equivalent C5Me4Si rings, one for both SiMe groups,
and one for both equivalent Cp ligands were observed.
Fortunately, single crystals of 2aa suitable for X-ray diffrac-
tion studies were obtained from a toluene solution of the
mixture cooled at -40 ꢀC, whereas appropriate crystals of
2as could not be isolated and formation of the remaining
possible isomer 2ss was not be detected.
molecule formed by two Ti atoms connected by two bridging
[μ-(η5-C5R4SiMeO)] fragments, with each of Ti atoms bound
to one bridging η5-C5Me4Si ring and to the oxygen atom of a
similar second bridging fragment. One additional oxygen
atom also links both bridges through the Si atoms, retaining
the initial geometry of compound A.37 Furthermore, both Ti
atoms are also bound to one Cp and one Cl ligand. Both Cp
and Cl ligands occupy equivalent positions in the molecule.
The environment about each Ti atom corresponds to the
typical pseudotetrahedral dicyclopentadienyl geometry.39,40
The molecular structure of 2aa is illustrated in Figure 3,
and selected bond lengths and angles are listed in Table 1, as
well as those obtained by theoretical calculations (see below).
The molecular structure of 2aa consists of a dinuclear
(39) Thiele, K.-H.; Gerstner, P.; Somoza, F. Z. Anorg. Allg. Chem.
2001, 627, 28.
(40) Grimmond, B. J.; Corey, J. Y.; Rath, N. P. Acta Crystallogr.,
Sect. C: Cryst. Struct. Commun. 2000, 56, 53.

Article
Organometallics, Vol. 29, No. 3, 2010 645
Scheme 2. Reaction of 2as with LiMe
Figure 3. ORTEP plot of the molecular structure of
[(TiCpCl)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (2aa) in the solid
state. Thermal ellipsoids are drawn at the 30% probability level.
Hydrogen atoms are omitted for clarity.
˚
Table 1. Selected Bond Distances (A) and Angles (deg) for
Compound 2aaa
experimental
Ti 1
calculated
bond/angle
Ti 2
Ti 1
Ti 2
Ti-O
1.903(3)
2.397(2)
2.257
1.898(4)
2.420(2)
2.247
2.114
1.661(4)
1.870
2.392
2.291
2.126
1.674
5.626
2.814
94.0
121.6
104.7
102.7
104.0
123.9
141.6
114.4
1.864
2.391
2.295
2.132
1.674
Ti-Cl
Ti-Cp0
Ti-Cp
Si-O(3)
Table 2. Total Yield and Relative Proportiona of Resulting
Isomers from Methylation of 2as at Variable Temperature
2.109
1.653(4)
5.914
2.809
Ti Ti
3 3 3
T (ꢀC)
t (h)
3as
3aa
3ss
yield (%)
Si Si
3 3 3
O-Ti-Cl
93.84(13)
119.65
104.93
107.83
103.97
121.92
142.4(2)
115.9(2)
94.77(13)
120.09
104.67
106.84
103.19
122.44
144.0(2)
94.4
-78
0
25
50
5
5
5
2
50
45
75
66
50
45
15
0
0
10
10
34
70
70
50
30
O-Ti-Cp0
O-Ti-Cp
Cl-Ti-Cp0
Cl-Ti-Cp
Cp0-Ti-Cp
Ti-O-Si
121.6
104.4
101.7
104.3
124.4
143.8
a Determined by 1H NMR spectroscopy.
Si-O(3)-Si
Table 2), depending on the temperature conditions. The yield
was also dependent on the temperature (Table 2). Higher total
yields were obtained at lower temperatures, indicating that
thermal decomposition of any reaction intermediates probably
takes place at higher temperatures; the starting complex 2as and
the mixture of isomers of 3 were stable when heated for several
hours in toluene. The isomer 3as with both Cp groups located at
opposite anti-syn positions with respect to the Si-O-Si bridge
is always the major component of the mixture, regardless of the
reaction temperature, although at -78 ꢀC it is formed in the
same proportion as isomer 3aa with both Cp substituents
located at anti positions. The proportion of isomer 3aa decreases
at increasing temperatures and is absent at 50 ꢀC, whereas
simultaneously the proportion of isomer 3ss, with both methyl
groups located at anti positions, increases and is absent at
-78 ꢀC. Alkylation of 2as with RMgCl (R=Me, Bz) or LiBz
failed to provide the corresponding dialkyl complexes, with
decomposition products always obtained.
All three isomers have been identified by NMR spectro-
scopy. Isomer 3as is an asymmetric molecule that shows a
NMR pattern similar to that described for isomer 2as,
containing eight resonances for the two C5Me4 groups, two
for SiMe groups, two for the Cp ligands, and two for the new
TiMe ligands, whereas the NMR spectra for either isomer
3aa or 3ss correspond to C2 symmetry compounds, showing
only four resonances for both C5Me4Si groups, one for both
SiMe groups, one for both Cp ligands, and one for the new
TiMe ligands. The relative disposition of the Me and Cp
ligands with respect to the Si-O-Si bridge were identified by
a “Experimental” stands for X-ray-determined parameters, whereas
“calculated” stands for the DFT-optimized structure. Cp0 stands for
C5Me4Si.
The bond distances and angles within the [Ti2(μ-{η5-
C5Me4SiMeO}2)] moiety are very similar to those found
for related Si-O-Ti bridged binuclear compounds,33,36,41
with or without an additional Si-O-Si bridge, with the
exception of the Ti(1)-O-Si(1) angle, which is smaller, and
the Ti-OSi bond distance, which is ca. 0.1 A longer for
compounds with this type of Si-O-Si bridging system or
with an additional single atom bridging both Ti atoms. The
presence of the cyclopentadienyl Cp ligands causes the
coplanarity of the η5-C5Me4 rings, as in [(TiCl2)2(μ-{η5-
C5R4SiMe2O}2)] (R=H,36 Me33), and also the elongation
of the Ti-Ti distance, which is the longest found in com-
plexes with a [Ti2(μ-{η5-C5Me4SiMeO}2)] moiety and also
longer than in the analogous compound [(TiCl2)2(μ-{(η5-
C5H4B{NHMe2}O)2(μ-O)})].42
Reaction of 2as with LiMe. Alkylation of the isolated pure
isomer 2as with LiMe in toluene did not give a unique com-
pound but instead gave a mixture containing variable amounts
of the three possible isomers of the dimethyl complexes
[(TiMeCp)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (3) (Scheme 2;
ꢀ
(41) Buitrago, O.; de Arellano, C. R.; Jimenez, G.; Cuenca, T.
Organometallics 2004, 23, 5873.
(42) Braunschweig, H.; Breitling, F. M.; Burschka, C.; Seeler, F.
J. Organomet. Chem. 2006, 691, 702.

646 Organometallics, Vol. 29, No. 3, 2010
Postigo et al.
considering that both Ti-Cl and Ti-O bonds of compound
2as are susceptible to alkylation50 (Scheme 3).
Stability of the Isomers of 2 and 3. In order to explain the
different regioselectivity observed for the products of the
reaction of A with 2 equiv of TlCp and the alkylation of 2as
with LiMe to form compounds 3, we theoretically studied the
relative stabilities of the three isomers of compounds 2 and 3.
The computational study requires calculating the energy of
the optimized structures of the isomers of the series 2 and 3
(as, aa, and ss) in order to understand the stability trends.
The calculated structural parameters of 2aa are compared in
Table 1 with those determined by X-ray diffraction for this
compound. The good agreement found does support the
methodology employed for the theoretical description of the
systems. Of note is the concord found even in the nonbond-
ing parameters, such as the Si Si and Ti Ti distances.
The relative energies in toluene of the calculated structures
are gathered in Table 3. Isomer 2as is considerably more
stable than 2aa and 2ss. In contrast, the isomers of 3 have
very similar energies. The relative energies in toluene of the
isomers of 2 and 3 agree with the experimental trends: 2as
was obtained as the major (>95%) or the unique product,
whereas a mixture of the three possible isomers of 3 was
found. To analyze the influence of the Cp ligand on the
relative stabilities, we have calculated the isomers of a
complex (B) derived from 2 by substituting the Cp ligand
by a less bulky methyl group (Figure 5). In B the nomen-
clature a-s refers to the anti and syn position of Me with
respect to Si-O-Si bridge. Results are very similar to those
obtained for the 3 set, and a clearly favored isomer was not
obtained. From these calculations it can be concluded that
the regioselectivity is introduced by the combination of
chloride and cyclopentadienyl ligands, vanishing when one
of these ligands is changed for a methyl group.
Aiming to understand the origin of the preference for the
2as isomer, we have carried out additional calculations in
two series of compounds related to 2. These compounds have
the same coordination sphere as in 2 for the two titanium
atoms, but the methyl groups attached to the C5Me4Si ring
(complexes C) or to the C5Me4Si ligands and silicon atoms
(complexes D) have been replaced by hydrogen atoms
(Figure 5). Results for the C set (Table 3) show that the
methyl substituents on the Cp ring have a very minor
influence on the stability of the isomers. Isomer Caa remains
at 2.8 kcal mol-1 above the most stable Cas; the only change
is a slight stabilization of Css, which lies 1.6 kcal mol-1 above
Cas. Thus, steric repulsions between methyl substituents on
the C5Me4Si ring and Si atom are partially responsible for the
destabilization of 2ss. It is interesting to note that, even at
this level, without the methyl substituents, the energy trend
describes greater stability for isomer 2as with respect to 2aa
and 2ss. Very similar results were obtained for complexes D
(Table 3); Daa and Dss are found 3.6 and 1.2 kcal mol-1
above the most stable Das. Removing the steric repulsion
between methyl substituents on the silicon atom and the
cyclopentadienyl ligands slightly stabilizes the syn-syn iso-
mer, where this repulsion was more important (two methyl
and Cp relatively close).
Figure 4. 2D NOE NMR spectrum of 3as and 3aa isomers
showing interaction between TiMe ligands and SiMe groups.
3 3 3
3 3 3
NOE NMR experiments (Figure 4), which showed the Cp
and Me ligands located in the syn position are closer to the
SiMe group of the Ti-O-Si moiety to which they are bound.
Finally, the reactivity of isomers 3 with Lewis acids described
below is consistent with this assignment.
Formation of isomers 3aa and 3ss in the alkylation reac-
tion of 2as may result either from isomerization of the
starting complex 2as or from isomerization of the alkyl
derivative 3as, which is the expected product for this reac-
tion, or alternatively from isomerization during the alkyl-
ation process. Several experiments have been realized in an
attempt to determine how this isomerization occurs, employ-
ing solutions of complex 2as and mixtures containing iso-
mers 3as and 3aa obtained at -78 ꢀC.
Thermal and photochemical isomerization43,44 of com-
plexes 2as and 3 was discounted after monitoring solutions of
2as and 3 by NMR in the range between -60 and 80 ꢀC and
irradiating these solutions with visible and UV light. No
modification was observed when the alkylation reaction was
carried out in the presence or absence of visible light.
The influence of different ions present in solution (Liþ,
Cl-) was also investigated.45-49 Addition of LiCl or
[Bu4N]Cl to toluene and THF solutions of 2as and 3 did
not show any change in the range between -60 and 80 ꢀC.
However, addition of the lithium salt Li[B(C6F5)4], as a more
soluble source of Liþ, to 2as gave a cationic compound,
which will be described later. Conversely, no change was
observed when Li[B(C6F5)4] was added to a THF solution of
isomers 3as and 3aa, although decomposition occurred in
toluene upon heating at 60 ꢀC, contrasting with the thermal
stability observed for these isomers.
These results suggest that isomerization occurred during
the alkylation process. To clarify this proposal, theoretical
calculations on the reaction mechanism were carried out by
(43) Schmidt, K.; Reinmuth, A.; Rief, U.; Diebold, J.; Brintzinger,
H. H. Organometallics 1997, 16, 1724.
(44) Collins, S.; Hong, Y. P.; Taylor, N. J. Organometallics 1990, 9,
2695.
(45) Buck, R. M.; Vinayavekhin, N.; Jordan, R. F. J. Am. Chem. Soc.
2007, 129, 3468.
(46) Curnow, O. J.; Fern, G. M.; Hamilton, M. L.; Zahl, A.;
van Eldik, R. Organometallics 2004, 23, 906.
(47) Hollis, T. K.; Wang, L. S.; Tham, F. J. Am. Chem. Soc. 2000, 122,
11737.
(48) Yoder, J. C.; Day, M. W.; Bercaw, J. E. Organometallics 1998,
17, 4946.
(49) Meppelder, G. J. M.; Fan, H. T.; Spaniol, T. P.; Okuda, J. Inorg.
These calculations prove that the presence of methyl
substituents on cyclopentadienyl and silicon is not the main
influence responsible for regioselectivity in 2 isomers.
(50) Nomura, K.; Naga, N.; Miki, M.; Yanagi, K.; Imai, A.
Organometallics 1998, 17, 2152.
Chem. 2009, 48, 7378.

Article
Organometallics, Vol. 29, No. 3, 2010 647
Scheme 3. Possible Alkylation Mechanism for Transformation of Isomer 2as into Isomer 3ss with Substitution of the Chloro Ligand
Located at the syn Positiona
a The same process on the chloro ligand located at the anti positions would result in formation of isomer 3aa.
Table 3. Relative Energies in Toluene (ΔEtol in kcal mol-1) of
the Optimized Structures of the Isomers of Compounds 2, 3, B, C,
and Da
The presence of a methyl ligand allows a relaxation of the
local geometry around the titanium centers, thus reducing
the steric strain. Methyl is a ligand with electrons localized in
the C-H bonds, whereas Cl and Cp ligands are electronically
more cumbersome, having wide clouds of electrons diffusing
all around and interacting with the electron clouds of the
other ligands. In the case of series 3, the methyl groups are
smaller than Cl, Me-Ti-L angles can be shorter, and so the
substituents around Ti can be better accommodated, result-
ing in more relaxed geometries with similar energies. Sub-
stituting the chloride ligand of series 2 with the methyl of
series 3, the ligands around Ti relax and the angles O-Ti-Cp
and O-Ti-C5Me4Si increase, decreasing interligand repul-
sion and ring strain and, in this way, decreasing the energy
difference between the compounds of the series. A detailed
comparison of the coordination spheres around titanium
atoms in compounds of series 2 and 3 and 2 and B is given in
the Supporting Information.
isomer ΔEtol isomer ΔEtol isomer ΔEtol isomer ΔEtol isomer ΔEtol
2as
2aa
2ss
0.0 3as
3.1 3aa
3.1 3ss
0.0 Bas
0.1 Baa
0.2 Bss
0.0 Cas
0.1 Caa
0.6 Css
0.0 Das
2.8 Daa
1.6 Dss
0.0
3.6
1.2
a For each series the energy of isomer as has been taken as a zero.
Repulsive interactions between cyclopentadienyl ligands and
the Si-O-Si bridge can also be discarded from the different
relative energies of parent compounds in series 2 and 3,
despite the same disposition of cyclopentadienyl ligands in 2
and 3 related compounds. Moreover distances between the
cyclopentadienyl ring centroid and the Si-O-Si bridging
oxygen are long enough (between 5.4 A for syn-cyclopenta-
dienyl and 5.8 A for anti-cyclopentadienyl) to preclude
significant interactions.
Theoretical Study of 2 f 3 Isomerization. We have com-
putationally examined the attack of methyllithium on the
two titanium centers of 2as. The reaction, carried out in
toluene, yields three possible products, as shown previously
in Scheme 2. Whereas the computational study of the
stability of 2 and 3’s isomers involved only geometry opti-
mizations of minimum structures, in this case a complete
exploration of the potential energy surface is needed, in order
to locate transition states and determine reaction barriers.
Given the dimensions of our system, a model system of the
compound 2as was employed instead of the real system used
above (Figure 7). We modeled complex 2as introducing the
simplification of examining only one center at a time, with
the hypothesis that the reaction on one center does not affect
the other. This implies a sort of “commutativity” of attacks
during the reaction (attacking Tia first and Tis in a second
moment is the same as attacking Tis first and then Tia; from
now on this will be implicitly assumed). Actually, given that
the two Ti centers are far away from each other in 2as (5.376
A), this can be considered quite a good approximation.
Nevertheless, the attacks on the 2as Ti centers are quite
different from each other and have to be studied separately;
that is, each reaction center has to be simulated. In light of
The interligand angles reported in Table 1 for 2aa show
significant deviations from a tetrahedral geometry. Hence,
we have further explored this behavior optimizing the mono-
nuclear titanium complexes E and F (Figure 6).
In Table 4 we have collected the angles describing the
environment about titanium atoms in the mononuclear tita-
nium complexes and in the as isomers of compounds 2, 3, and
B. When only one cyclopentadienyl ring is present (E), the
complex adopts the normal three-ledged piano-stool geometry
with the angles involving the Cp* (Cp*=C5Me5) ligand larger
than those between the legs (115ꢀ and 103ꢀ, respectively). The
presence of a second cyclopentadienyl ligand (F) markedly
modifies the angles, with the two bulky cyclopentadienyl rings
placed as far as possible apart (Cp*-Ti-Cp=134ꢀ). In the
bimetallic compounds analyzed, the strain caused by the
presence of two bridges connecting the two Ti centers provokes
notable changes in the angles, as can be appreciated by
comparing the angles in F and 2as. In addition to the
Cp*-Ti-Cp angles, the Cp*-Ti-O angles are also larger
than 120ꢀ. When the Cp ligand is replaced by a methyl group in
compound B, the C5Me4Si-Ti-O angle becomes the largest
(about 130ꢀ), with the other angles approaching the usual
tetrahedral values.

648 Organometallics, Vol. 29, No. 3, 2010
Postigo et al.
Figure 5. Structure drawing of modified complexes (B, C, D) designed for theoretical calculations.
Figure 6. Structure drawing of mononuclear complexes (E, F)
designed for theoretical calculations.
Figure 7. Model complexes and reagent considered in the
theoretical study of the alkylation reaction.
Table 4. Angles (deg) Describing the Environment of the Titanium
Centers in the Calculated Mono- and Dinuclear Complexesa
Table 5. Relative Energies in Toluene (in kcal mol-1) of the Six
Different Possible Attacks Analyzeda
E
F
2as
3as
Bas
[d(¥)]b energy barrier configuration
(X1 = Cl (X1 = Cp (X1 = Cp (X1 = Cp (X1 = Me
X2 = Cl) X2 = Cl) X2 = Cl) X2 = Me) X2 = Cl)
attack
I
TS
F
angle
LiO(A)
LiCl(A)
-3.4 7.0 -1.7 -34.6
-3.9 17.4 -41.7 -34.6
10.4
21.3
28.3
AfS
AfA
AfA
Cp0-Ti-O
Cp0-Ti-X1
Cp0-Ti-X2
X1-Ti-O
X2-Ti-O
X1-Ti-X2
114.8
114.9
115.0
103.4
103.4
103.8
102.1
134.2
106.9
107.5
96.5
121.4/120.0 123.4/122.2 131.3/130.3
124.4/ 122.9 123.5/122.3 107.7/103.6
102.3/105.2 102.5/104.5 108.0/112.6
104.6/104.6 106.1/105.5 99.2/99.8
LiOb(A) -6.1 22.2 -18.9 -34.6
LiO(S)
LiCl(S)
LiOb(S)
-10.1 15.3 -26.3 -34.8
-5.2 13.4 -34.8 -34.8
-2.5 13.3 -11.7 -34.8
25.4
18.6
15.8
SfA
SfS
SfA
93.3/96.8 90.7/94.1 106.2/106.1
104.4/102.8 102.3/102.4 100.0/99.0
103.6
a The letter in parentheses, in each attack, designates the center under
attack. The energy of the reactants at infinite distance is taken as zero.
Configuration refers to the change of the configuration of the Ti center;
for example, AfS means that the reaction happens to a Tia center that
changes its configuration, in the final product, to a Tis. b The final
product and LiCl(tol) at infinite distance.
a Cp0 stands for Cp* (Cp* = C5Me5) in complexes E and F and for
C5Me4Si in complexes 2, 3, and B.
the previous considerations, compound 2as was simplified
using two model systems, named Gs and Ga, differing in the
syn and anti position of the Cp ligand with respect to the
Si-O-Si bridge (Figure 7). In these models, the methyl
groups on each C5Me4Si as well as the methyl substituent
on the Si linked to the O-Ti have been considered. In
addition, to keep the Si-O-Si bridge fixed, a CH2-CH2
chain has been added in order to simulate the rigidity of
compound 2as. However, the description of the organo-
lithium reagent required some test calculations. Solvent
effects are very important in organolithium chemistry, and
they dramatically influence the aggregation state and reac-
tivity of alkyllithium compounds.51-58 Test calculations
proved than only one molecule of toluene is coordinated to
the lithium atom of LiMe. For this reason LiMe(tol) (H) was
chosen as reagent (see Supporting Information for further
discussion on the choice of the model system). Thus, the
solvent (toluene) was considered in two ways in the compu-
tational study of the reaction: explicitly, by coordinating one
molecule of toluene to LiMe, and implicitly by doing refine-
ment calculations in a uniform dielectric solvent as before,
using the CPCM model (solvent=toluene, ε=2.379).
Given the simplifications introduced in the system, due to
computational limitations, and the subtle differences that
could account for the temperature-dependent relative pro-
portion of isomers, a quantitative agreement with the pro-
portions of isomers of 3 obtained is beyond the scope of the
investigation. The goal of the calculations is to assess the
existence of low-energy pathways able to transform 2as in the
three isomers of 3 and to analyze the reaction mechanism for
the alkylation of 2as with LiMe. The reaction can be
portrayed as the approach of a LiMe compound to one of
the two Ti centers of 2as, the donation of a Me- anionic
group, and the extraction of a Cl- anionic group, with
subsequent elimination of LiCl. In 2as there are three basic
centers available to anchor the lithium prior to methyl
transfer: two oxygens and a chloride. We have considered
the three possible approaches: (1) attack of H on the Ti center
after a previous coordination to the oxygen atom bonded to
C5Me4Si (LiO); (2) attack of H on the Ti center after a
previous coordination of the Liþ cation to the Cl atom linked
to Ti (LiCl); (3) attack of H on the Ti center after a previous
coordination to the oxygen atom of the Si-O-Si bridge
(LiOb; Ob=O3; see Figure 3).
(51) Pratt, L. M.; Truhlar, D. G.; Cramer, C. J.; Kass, S. R.;
Thompson, J. D.; Xidos, J. D. J. Org. Chem. 2007, 72, 2962.
(52) Pratt, L. M.; Ramachandran, B.; Xidos, J. D.; Cramer, C. J.;
Truhlar, D. G. J. Org. Chem. 2002, 67, 7607.
(53) Leroy, B.; Marko, I. E. J. Org. Chem. 2002, 67, 8744.
(54) Katritzky, A. R.; Xu, Y.-J.; Jian, R. J. Org. Chem. 2002, 67, 8234.
(55) Mogali, S.; Darville, K.; Pratt, L. M. J. Org. Chem. 2001, 66,
2368.
(56) Streitwieser, A.; Juaristi, E.; Kim, Y.-J.; Pugh, J. Org. Lett. 2000,
2, 3739.
(57) Fraenkel, G.; Duncan, J. H.; Martin, K.; Wang, J. J. Am. Chem.
Soc. 1999, 121, 10538.
(58) Hoffmann, D.; Collum, D. B. J. Am. Chem. Soc. 1998, 120, 5810.

Article
Considering that each of these attacks can happen both to
Organometallics, Vol. 29, No. 3, 2010 649
more detailed information is reported about the changes in
the most important distances describing the process: the
Ti-Me distance (the bond that is being formed), the Ti-Cl
and Li-Me distances (the bonds that are being broken), and
Li-X (X=Ob, O, or Cl, according to the attack).
Important conclusions about the reaction mechanism can
be deduced from these results. First, despite the similarity of
Tia and Tis, centers, differing only in the position of two
substituents, each can sometimes react in a quite different
way. Second, whereas several of the attacks take place with
retention of stereochemical configuration, others happen with
inversion of stereochemical configuration. This aspect is in-
dicated in Table 5 as AfS or SfA (inversion) and AfA or
SfS (retention). Calculations confirm that the formation of
the three isomers of 3 in the alkylation reaction of 2as is
produced during the alkylation process.
Tia and to Tis, we have six possible pathways. All of which
have been explored. In each pathway an initial intermediate
(I), a transition state describing the methyl transfer (TS), and
a final intermediate (F) have been located. The relative
energies of these structures in toluene, taking as a zero energy
that of Gs (or Ga) together with H at infinite distance (for the
attacks on the Tis and Tia centers, respectively), are collected
in Table 5. Figure 8 shows the energy profiles. In Table 6
The lowest energy pathway for methylation of Tia of 2as
with retention of configuration (thus keeping a Tia stereo-
chemistry in 3) is the LiCl(A) attack, whereas the LiO(A)
attack is the lowest with inversion of configuration. The
main steps of these two attacks to Tia are depicted in
Scheme 4. In the TS of the LiCl(A) pathway there is a
slippage of the C5Me4Si ring, which changes from an initial
η5 coordination to a η2 coordination in this transition state.
Accordingly, a rather high energy barrier is found for this
step. In the LiO(A) pathway Me attacks on the opposite side
Figure 8. Plot of the different energies (as reported in Table 4)
computed for the four steps of the six different attacks of
LiMe(tol) to Ga and Gs in toluene.
a
˚
Table 6. Most Relevant Distances (A) in All the Attacks
dI(Li-Me)/
dTS(Li-Me)/
dF(Li-Me)
dI(Ti-Me)/
dI(Li-X)/
dTS(Li-X)/
dF(Li-X)
dI(Ti-Cl)/
dTS(Ti-Cl)/
dF(Ti-Cl)
dTS(Ti-Me)/
attack
dF(Ti-Me)
LiO(A)
LiCl(A)
LiOb(A)
2.011/2.003/2.177
2.023/2.004/2.540
2.010/2.009/3.251
4.795/3.157/2.303
5.021/3.309/2.171
7.687/3.189/2.150
1.900/1.784/1.755
2.387/2.199/2.142
1.956/1.935/1.898
2.365/2.569/2.821
2.486/2.529/4.724
2.389/2.520/5.326
LiO(S)
LiCl(S)
LiOb(S)
2.033/1.991/2.498
2.022/1.987/2.266
1.997/2.010/2.255
5.026/3.313/2.169
4.993/3.176/2.181
4.884/3.148/2.216
1.990/1.786/1.733
2.390/2.136/2.122
3.493/2.131/1.863 1.912/1.903/1.932
2.429/2.470/2.352
2.473/2.792/5.928
2.360/2.564/4.334
a dI is the distance at the I geometry, dTS is the distance measured at the transition state geometry, and dF is the corresponding distance in the final
intermediate. X corresponds either to Ob, O, or Cl, depending on the initial site of coordination of compound H.
Scheme 4. Lowest Energy Pathway for Methylation of Tia of 2as with Inversion of Configuration (LiO(A) attack, upper) and with
Retention of Configuration (LiCl(A) attack, bottom)

650 Organometallics, Vol. 29, No. 3, 2010
Postigo et al.
Scheme 5. Lowest Energy Pathway for Methylation of Tis of 2as with Retention of Configuration (LiCl(S) attack, upper) and with
Inversion of Configuration (LiOb(S) attack, bottom)
Scheme 6a
a (i) E(C6F5)3 (E = B, Al) or Li[B(C6F5)4].
of the Cl atom (the angle Me-Ti-Cl is about 150ꢀ), making
this attack a sort of SN2 reaction. The SN2 character of the
reaction entails a change in the configuration of the Ti center.
The small change in geometry, going from the initial inter-
mediate to the TS, is the cause of the lowest energy barrier of
all six attacks studied. The LiOb(A) happens with retention
of configuration, but its transition state is quite high
in energy, due to its stretched geometry (see Supporting
Information).
(after LiO(A)), and its activation energy is quite low. This
double coordination is the reason that, in Table 5, two rows
of numbers for the Li-X value are reported. The first refers
to the Li-Ob distance, while the second refers to the Li-O
bond (where the oxygen atom is linked to Ti). As in the case
of LiO(A), given the opposite direction from which the Me
ligand approaches the Ti center with respect to the Cl ligand,
this can also be regarded as an SN2 mechanism (angle
Me-Ti-Cl in TS 151.3ꢀ). The product is characterized by
the migration of the Cl- anion to the closest Si atom, creating
a bipyramid whose center is the Si atom and the opposite
vertices are Ob and Cl. This structure, quite high in energy, is
more stable than those characterized by a separation of
charges. This attack happens with inversion of stereochemical
configuration (SfA). The low activation barrier makes this
attack the favorite to create a Tia center in 3 from a Tis in 2.
The LiO(S) attack also entails inversion of configuration
but has a notably higher energy barrier (see Supporting
Information).
The three isomers of 3 can be obtained from 2as by
combining sequentially two attacks. Isomer 3as can be
formed by a LiO(A) attack (which transforms a Tia center
in 2as into a Tis center in 3as) followed by a LiOb(S) attack
(which transforms a Tis center in 2as into a Tia center in 3as),
isomer 3aa can be obtained by a LiCl(A) attack and then
LiOb(S) attack, and finally, isomer 3ss can be obtained by
LiO(A) and LiCl(S) attacks. In the model systems considered
(Gs and Ga), the highest energy barriers for such transfor-
mations are 15.8 kcal mol-1 (2as f 3as), 21.3 kcal mol-1
The lowest energy pathways for methylation of Tis with
retention and inversion of configuration are shown in
Scheme 5. The LiCl(S) attack is the lowest energy pathway
for methylation of Tis of 2as with retention of configuration
(thus keeping a Tis stereochemistry in 3), as the LiCl(A)
attack was for methylation of Tia of 2as with retention of
configuration. The transition state is basically an exchange
of two groups (Me and Cl) between the reactant and the
substrate. The fact that it does not involve strong distortions
is the reason for its relatively low barrier. This attack is
similar to LiCl(A). LiOb(S) is the lowest energy pathway for
methylation of Tis of 2as with inversion of configuration.
Table 4 and Scheme 5 demonstrate how the approach should
be more properly called Li-O rather than Li-Ob, as in the
initial state the Li atom is much closer to the oxygen atom
bonded to Ti than to the oxygen atom of the Si-O-Si
bridge. During the transition state the Li atom coordinates
to both available oxygen atoms (the O bonded to Ti and Ob)
and starts a weak interaction with toluene. All these inter-
actions make this transition state the second most stable

Article
Organometallics, Vol. 29, No. 3, 2010 651
a
Table 7. Distances (A) between the Remaining Cl Atom (Cl for
˚
product J and Cls for product K) and the Ti Centersa
d(Cl-Tis)
d(Cl-Tia)
ΔE
J
K
2.589
2.358
2.589
6.957
0.0
18.3
a Tis is the Ti linked to Cls (in compound 2as), and Tia is the Ti linked
to Cla (in 2as). Energies were obtained in toluene, and ΔE (kcal mol-1
refers to the difference in energy with product J.
)
such as py or thf. Moreover, addition of these donor ligands
to solutions of 4B and 4Al selectively regenerates the starting
complex 2as with formation of adducts L E(C6F5)3.
3
Theoretical Study of Chloride Abstraction. The experimen-
tal study shows that during the electrophilic attack of
compounds of the type E(C6F5)3 (E=B, Al) and Li[B(C6F5)4]
to the dichloro compound 2as only one of the two Cl ligands
is lost, forming cation 4þ. In order to evaluate possible
differences between both chloride ligands in 2as, we per-
formed a natural bond orbitals (NBO) analysis of the
stabilization energy due to orbital interactions for both
chlorides.62,63 Overall, the summation of all the contribu-
tions to the stabilization of the chloro ligands of compound
2as points out a small difference between Cls and Cla (see
Supporting Information). The stabilization of Cla to the 2as
compound is 4.4 kcal mol-1 greater than that of Cls, suggest-
ing that its removal would destabilize the molecule more than
losing Cls. The analysis of the stability of the two different
products once a chloride has been removed gives the reason
for the preferred abstraction of the chloro ligand located at
the syn position. The relative energies of the two possible
products obtained by optimization after the removal of one
chloride are collected in Table 7 (DFT/BS1 calculations).
Product J is obtained from 2as removing Cls (in syn position
with respect to Si-O-Si), while product K is obtained from
2as removing Cla (in anti position).
Figure 9. ORTEP plot of the molecular structure of cation
[(TiCp)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})(μ-Cl)]þ (4þ) in the solid
state. Thermal ellipsoids are drawn at the 30% probability level.
Hydrogen atoms are omitted for clarity.
(2as f 3aa), and 18.6 kcal mol-1 (2as f 3ss). Although these
values cannot account for the proportion of isomers of 3
obtained experimentally, they do support the possible for-
mation of the three isomers of 3 from the alkylation of 2as.
Reactions of 2as with Lewis Acids. The dichloro compound
2as reacted with the Lewis acids E(C6F5)3 (E=B, Al) and the
lithium salt Li[B(C6F5)4] to give the chloro-bridged com-
pounds [(TiCp)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})(μ-Cl)][Q] (4;
Q=ClB(C6F5)3, 4B; ClAl(C6F5)3, 4Al; Q=B(C6F5)4, 4C)
(Scheme 6) as the only reaction products. The cation 4þ is
formed by abstraction of the chloro ligand located at the
syn position, because only the chloro ligand in the anti
position can close a bridge between both Ti atoms. Theore-
tical calculations indicate that 4þ is the thermodynamic and
kinetic product of the reaction (see below). The 1H and 13
C
From the distances collected in Tables 7 an important
difference between J and K products can be noticed.
Whereas in product K the removal of Cla causes a small
local rearrangement of the Tia ligands, but leaves the other
part of the molecule (the one with Cls) basically unchanged,
the removal of the Cls atom in product J causes a huge
rearrangement of the geometry of the whole molecule, end-
ing with the Cla atom bridging between the two Ti centers, as
evidenced by the distances between the remaining chlorine
and the Ti centers. The structure of J agrees with the X-ray
determined structure of 4þ. The rearrangement of the Cla
atom, bridging between the two Ti centers, stabilizes the
molecule by about 20 kcal mol-1 with respect to product K.
In this way, the dissociation of the Cls atom is thermodyna-
mically preferred. This fact, combined with the easier
extraction of the Cls atom, makes this chloro ligand the
better candidate for an electrophilic reaction, as proved by
experiments.
NMR spectra are consistent with formation of these C2
symmetry compounds, for which only four resonances for
both C5Me4 ligands, one for both SiMe groups, and for
both Cp ligands were observed. Furthermore, the 19F NMR
spectra of these species confirm the formation of the corre-
sponding anions [ClB(C6F5)3]-, [ClAl(C6F5)3]-, and [B(C6-
F5)4]-.19,59 An X-ray diffraction study for compound 4C was
carried out, confirming the proposed structure for this
complex (Figure 9),60,61 with a Cl atom bridging both Ti
atoms.
Addition of excess E(C6F5)3 (E=B, Al) or Li[B(C6F5)4] to
complexes 2as or 4þ did not cause the abstraction of the
second chloro ligand even in the presence of donor ligands
(59) Zhou, J.; Lancaster, S. J.; Walker, D. A.; Beck, S.; Thornton-
Pett, M.; Bochmann, M. J. Am. Chem. Soc. 2001, 123, 223.
(60) Although many efforts have been done to obtain suitable crystals
of 4C for X-ray analysis, only crystals of poor quality have been so far
obtained, so that although the identity of the complex was unambigu-
ously established, the anion B(C6F5)4- is very disordered and the
distances and angles may not be very reliable.
Reactions of 3 with Lewis Acids. As discussed above,
methylation of 2as under different thermal conditions always
afforded mixtures containing variable amounts of the three
possible isomers of the dimethyl compound 3, from which
none could be isolated as a pure component. Therefore,
(61) Crystal data for 4C: C54H34BClF20O3Si2Ti2 (C6H6), Mr =
3
1387.16, crystal size 0.42 ꢀ 0.32 ꢀ 0.25 mm3, monoclinic, space group
P21/c, a=15.199(4) A, b=18.274(5) A, c=24.262(5) A, β=90.60(3)ꢀ, V=
6738(3) A3, Z=4, Fcalcd =1.367 mg m-3, μ =0.408 mm-1, Mo KR
radiation (λ=0.71073 A). Data collection was performed at 200(2) K on
a Nonius Kappa CCD single-crystal diffractometer. Reflections col-
lected/unique 43 274/11 831. The final cycle of full matrix least-squares
refinement based on 17 03 reflections and 827 parameters converged to
final values of R1(F2 > 2σ(F2))=0.1765, wR2(F2 > 2σ(F2))=0.4612,
R1(F2) = 0.2085, wR2(F2) = 0.5007). Largest diff peak/hole 2.686/
(62) Weinhold, F.; Landis, C., Valency and Bonding. A Natural Bond
Orbital Donor-Acceptor Perspective; Cambridge University Press:
Cambridge, U.K., 2005.
(63) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. Rev. 1988, 88,
899.
1.007 e A-3
.

652 Organometallics, Vol. 29, No. 3, 2010
Postigo et al.
Figure 10. Structure drawing of mononuclear complexes (J, K) designed for theoretical calculations.
Scheme 7a
a (i) [Ph3C][B(C6F5)4], -70 ꢀC; (ii) XC6D5 (X = F, Cl, Br) or CD2Cl2, rt.
studies of their reactivity with Lewis acids with the aim of
generating cationic complexes had to be carried out using
mixtures of isomers. We have focused our studies on the
products obtained at -78 ꢀC (see Table 4), with a 3as:3aa
molar ratio of 1:1 due to the lack of purity of mixtures
obtained at higher temperatures. It is important to note that
isomer 3as shows two different methyl groups located at syn
and anti positions, respectively, whereas both methyl groups
of isomer 3aa are located at syn positions.
rings, two resonances for both SiMe groups and two for both
Cp ligands, and finally one resonance for the remaining
TiMe group for each isomer.
When the temperature was increased, isomer 5aþ was
transformed into the chloro-bridged cationic complex 4þ,
and after several hours at room temperature isomer 5sþ,
which cannot be transformed into the bridging species, was
completely decomposed. The same behavior was observed
when XC6D5 (X = F, Cl, Br) was employed as solvent,
affording the respective cationic compounds [(TiCp)2(μ-
{(η5-C5Me4SiMeO)2(μ-O)})(μ-X)]þ (X =F, 4Fþ; Cl, 4Clþ;
Br, 4Brþ) at room temperature, which showed very similar
1H and 13C NMR spectra. The 19F NMR spectrum of the
4Fþ cation showed a broad resonance at negative values
(δ -53.3), corresponding to the Ti-F-Ti bridge.19,64
Finally, as we have reported previously with other di-
nuclear complexes of this type, addition of an excess of any
Lewis acid, E(C6F5)3 (E=B, Al) or [Ph3C][B(C6F5)4], did not
afford dicationic complexes, even in the presence of donor
ligands.37,38
Reactions with the Lewis acids E(C6F5)3 (E=B, Al) of the
equimolar mixture of 3as and 3aa unfortunately failed to give
any identifiable product even at low temperature. However,
addition of [Ph3C][B(C6F5)4] to a CD2Cl2 solution of 3as and
3aa at -70 ꢀC gave a mixture containing a 1:1 molar ratio of
two isomers [(TiCpMe)(TiCp)(μ-{(η5-C5Me4SiMeO)2(μ-
O)})]þ (5aþ and 5sþ), which resulted from the abstraction
of one of the methyl groups to leave the remaining methyl
ligand at anti and syn positions, respectively (Scheme 7).
The 1:1 molar ratio of the resulting mixture of isomers
corresponds exactly with the initial molar ratio of 3as:3aa,
demonstrating that 5sþ was formed from 3aa only and not
from 3as. Consequently, methyl abstraction from 3as must
be regiospecific, as only the methyl group located at the syn
position of the two different methyl groups present can be
abstracted; this is the same behavior observed for the reac-
tion of 2as with Lewis acids.
Conclusions
The introduction of only one Cp ligand in the dinuclear
compound [(TiCl2)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (A),
with two different chloro ligands on each Ti atom with
The 1H and 13C NMR spectra of this mixture are consis-
tent with the presence of two compounds that are asym-
metric molecules and show eight resonances for two C5Me4Si
(64) Perdih, F.; Pevec, A.; Petricek, S.; Petric, A.; Lah, N.; Kogej, K.;
Demsar, A. Inorg. Chem. 2006, 45, 7915.

Article
Organometallics, Vol. 29, No. 3, 2010 653
respect to the Si-O-Si bridge, is more favorable in one of
the two possible positions, although a mixture of both
possible isomers of [(TiClCp)(TiCl2)(μ-{(η5-C5Me4SiMeO)2-
(μ-O)})] were obtained in the reaction of A with 1 equiv of
TlCp at 80 ꢀC. However addition of 2 equiv of TlCp at 80 ꢀC
to compound A gave the disubstituted isomer [(TiCpCl)2-
(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (2as) regiospecifically, with
each Cp ligand in anti-syn positions. According to the DFT
studies, this is the lowest energy isomer of the three possible
isomers that this reaction could produce. The isomer 2aa, with
both Cp ligands located at anti positions with respect to the
Si-O-Si bridge, was also obtained as a minor product when
the same reaction was carried out at 60 ꢀC.
0.5(toluene) Al(C6F5)3,66 and [Ph3C][B(C6F5)4]67 were pre-
pared by literature methods. TlCp and Li[B(C6F5)4] were ob-
tained from commercial sources.
3
[(TiCl2){Ti(η5-C5H5)Cl}(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (1).
A suspension of [(TiCl2)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (A)
(0.20 g, 0.32 mmol) and TlCp (0.10 g, 0.38 mmol) in toluene
(20 mL) was heated at 80 ꢀC for 48 h. A 10 mL amount of hexane
was then added, and the solution was filtered. The volatiles were
pumped off to give a mixture of the two monosubstituted
1
isomers (50% of major isomer (1M) by H NMR and 17% of
minor isomer (1m) by 1H NMR), the disubstituted product 2as
1
1
(8% by H NMR), and the starting material A (25% by H
NMR). 1H NMR (C6D6): 0.29 (s, 3 H, SiMe, 1m), 0.33 (s, 3 H,
SiMe, 1m), 0.34 (s, 3 H, SiMe, 1M), 0.48 (s, 3 H, SiMe, 1M), 1.45
(s, 3 H, C5Me4, 1M), 1.58 (s, 3 H, C5Me4, 1m), 1.60 (s, 3 H,
C5Me4, 1M), 1.95 (s, 3 H, C5Me4, 1m), 2.03 (s, 3 H, C5Me4, 1M),
2.08 (s, 3 H, C5Me4, 1M), 2.10 (s, 3 H, C5Me4, 1m), 2.11 (s, 3 H,
C5Me4, 1M), 2.15 (s, 3 H, C5Me4, 1m), 2.18 (s, 3 H, C5Me4, 1m),
2.20 (s, 3 H, C5Me4, 1M), 2.20 (s, 3 H, C5Me4, 1m), 2.37 (s, 3 H,
C5Me4, 1M), 2.42 (s, 3 H, C5Me4, 1m), 2.48 (s, 3 H, C5Me4, 1M),
2.58 (s, 3 H, C5Me4, 1m), 5.90 (s, 5 H, C5H5, 1m), 6.07 (s, 5 H,
C5H5, 1M). 13C NMR (C6D6): -3.2 and -1.5 (SiMe, 1M),
-2.1 and -2.4 (SiMe, 1m), 11.9-15.4 (C5Me4, 1M and 1m),
119.0 (C5H5, 1m), 120.1 (C5H5, 1M), 119.2-148.3 (C5Me4, 1M
and 1m).
Reaction of isomer [(TiCpCl)2(μ-{(η5-C5Me4SiMeO)2(μ-
O)})] (2as) with LiMe in toluene gave the three possible
isomers of [(TiCpMe)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (3as,
3ss, 3aa). The relative proportion of isomers was dependent
on the reaction temperature, the total yield and purity being
higher at -78 ꢀC. A DFT study of the reaction mechanism
shows that despite the similarity of Tia and Tis centers, each
can sometimes react in a quite different ways. Moreover, the
attack of LiMe can take place both with retention of stere-
ochemical configuration and with inversion of stereochemical
configuration. Calculations do confirm that the formation of
the three isomers of 3 takes place during the alkylation
process of 2as.
Compound 2as reacts with E(C6F5)3 (E=B, Al) or Li[B-
(C6F5)4] to give the cationic chloro-bridged complex
[(TiCp)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})(μ-Cl)]þ (4þ) by ab-
straction of the Cl atom next to the Si-O-Si bridge. For-
mation of complex 4þ corresponds with the thermodynamic
product of the reaction.
[{Ti(η5-C5H5)Cl}2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (2as). A
suspension of [(TiCl2)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (A)
(0.90 g, 1.47 mmol) and TlCp (1.23 g, 4.41 mmol) in toluene
(50 mL) was heated at 80 ꢀC for 48 h, when 20 mL of hexane was
added and the solution was filtered. The red residue was
extracted again into a mixture of solvents toluene/hexane
(30 mL/20 mL). The volatiles were removed under vacuum,
leaving a red solid (0.88 g, 90%). 1H NMR (CDCl3): 0.11 (s, 3 H,
SiMe), 0.18 (s, 3 H, SiMe), 1.62 (s, 3 H, C5Me4), 1.69 (s, 3 H,
C5Me4), 2.16 (s, 3 H, C5Me4), 2.22 (s, 3 H, C5Me4), 2.23 (s, 3 H,
C5Me4), 2.31 (s, 3 H, C5Me4), 2.33 (s, 3 H, C5Me4), 2.40 (s, 3 H,
C5Me4), 6.05 (s, 5 H, C5H5), 6.26 (s, 5 H, C5H5). 13C NMR
(CDCl3): -1.7 and -0.1 (SiMe), 11.9, 12.8, 13.8, 14.3, 14.4, 15.2,
15.6, and 16.4 (C5Me4), 118.4 and 119.0 (C5H5), 119.1, 120.0,
122.4, 124.9, 127.0, 127.6, 131.9, 132.1, 139.0, and 147.1
(C5Me4). Anal. Calcd for C30H40O3Si2Ti2Cl2 (670.55): C,
53.69; H, 5.96. Found: C, 53.53; H, 5.84.
A
mixture of isomers 3as and 3aa reacted with
[Ph3C][B(C6F5)4] to give a mixture of the cationic complexes
[(TiCpMe)(TiCp)(μ-{(η5-C5Me4SiMeO)2(μ-O)})]þ (5aþ and
5sþ). Formation of 5aþ corresponds to the selective abstrac-
tion of the methyl group at the syn position with respect to
the Si-O-Si bridge of isomer 3as, whereas formation of
complex 5sþ occurs only with abstraction of the methyl
group syn to the Si-O-Si bridge of isomer 3aa. Thus, the
molar ratio 5aþ:5sþ is the same as the initial molar ratio of
3as:3aa. Complex 5aþ evolved at ambient temperature to the
halogen-bridged complexes [(TiCp)2(μ-{(η5-C5Me4SiMe-
O)2(μ-O)})(μ-X)]þ (X=F, Cl, Br) by methyl-halide exchange
with the solvent (XC6D5, X=F, Cl, Br; CD2Cl2).
[{Ti(η5-C5H5)Cl}2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (mixture
of 2aa and 2as). A suspension of [(TiCl2)2(μ-{(η5-C5Me4SiMeO)2-
(μ-O)})] (A) (0.45 g, 0.73 mmol) and TlCp (0.91 g, 3.31 mmol) in
toluene (50 mL) was heated at 60 ꢀC for 48 h. After addition of
hexane (20 mL), the solution was filtered. The remaining solution
was cooled to -40 ꢀC to give a mixture of the isomer 2as (95%)
and 2aa (5%) (total yield 0.40 g, 87%). 1H NMR (CDCl3): 0.14 (s,
6 H, SiMe), 2.15 (s, 6 H, C5Me4), 2.24 (s, 6 H, C5Me4), 2.31 (s, 6 H,
C5Me4), 2.33 (s, 6 H, C5Me4), 6.12 (s, 10 H, C5H5).
Experimental Section
[{Ti(η5-C5H5)Me}2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (3).
A
solution of 2as (0.30 g, 0.44 mmol) in toluene (30 mL) was
treated with 2 equiv of LiMe (0.60 mL, 0.90 mmol) at the
reaction temperature. After the corresponding reaction time,
hexane (10 mL) was added and the whole filtered. The volatiles
were removed under vacuum, yielding a mixture of the isomers
3as, 3aa, and 3ss in different proportions depending on the
temperature of the reaction (see Table 2). The yield decreases
with increasing temperature, and the elemental analysis re-
ported corresponds to the sample synthesized at -78 ꢀC. Data
for 3as: 1H NMR (C6D6): 0.24 (s, 3 H, SiMe), 0.31 (s, 3 H, SiMe),
0.60 (s, 3 H, TiMe), 0.70 (s, 3 H, TiMe), 1.20 (s, 3 H, C5Me4),
1.27 (s, 3 H, C5Me4), 1.56 (s, 3 H, C5Me4), 1.57 (s, 3 H, C5Me4),
2.10 (s, 3 H, C5Me4), 2.13 (s, 3 H, C5Me4), 2.30 (s, 3 H, C5Me4),
General Considerations. All manipulations were carried out
under an argon atmosphere, and solvents were purified from
appropriate drying agents. NMR spectra were recorded at
400.13 (1H), 376.70 (19F), and 100.60 (13C) MHz on a Bruker
AV400. Chemical shifts (δ) are given in ppm. 1H and 13C
resonances were measured relative to solvent peaks considering
TMS=0 ppm, while 19F resonances were measured relative to
external CFCl3. Assignment of resonances was made from
HSQC, HMBC, and NOESY NMR experiments. Elemental
analyses were performed on a Perkin-Elmer 240C. Compounds
[(TiCl2)2(μ-{(η5-C5Me4SiMeO)2(μ-O)})] (A),37 B(C6F5)3,65
2.39 (s, 3 H, C5Me4), 5.69 (s, 5 H, C5H5), 5.84 (s, 5 H, C5H5). 13
C
(65) Lancaster, S. In http://www.syntheticpages.org/pages/215,
2003, p 215.
(66) Feng, S. G.; Roof, G. R.; Chen, E. Y. X. Organometallics 2002,
(C6D6): 0.6 and 1.4 (SiMe), 40.3 and 43.6 (TiMe), 10.7, 10.9,
12.7, 12.8, 13.4, 14.1, 14.4, and 14.9 (C5Me4), 114.1 and 114.4
(C5H5), 117.1-135.5 (C5Me4). Data for 3ss: 1H NMR (C6D6):
0.27 (s, 6 H, SiMe), 0.68 (s, 6 H, TiMe), 1.63 (s, 6 H, C5Me4), 1.99
21, 832.
(67) Lancaster, S. In http://www.syntheticpages.org/pages/216,
2003, p 216.

654 Organometallics, Vol. 29, No. 3, 2010
Postigo et al.
Table 8. Crystallographic Data for [(TiClCp)2(μ-{(η5-
(s, 6 H, C5Me4), 2.00 (s, 6 H, C5Me4), 2.26 (s, 6 H, C5Me4), 5.76
(s, 10 H, C5H5). 13C NMR (C6D6): -0.6 (SiMe), 41.8 (TiMe),
10.6, 12.8, 14.2, and 15.3 (C5Me4), 114.5 (C5H5), 117.1-135.5
C5Me4SiMeO)2(μ-O)})] (2aa)
formula
fw
C30H40Cl2O3Si2Ti2
671.50
1
(C5Me4). Data for 3aa: H NMR (C6D6): 0.30 (s, 6 H, SiMe),
0.74 (s, 6 H, TiMe), 1.25 (s, 6 H, C5Me4), 1.52 (s, 6 H, C5Me4),
2.15 (s, 6 H, C5Me4), 2.22 (s, 6 H, C5Me4), 5.70 (s, 10 H, C5H5).
Anal. Calcd for C32H46O3Si2Ti2 (629.55): C, 60.99; H, 7.30.
Found: C, 60.79; H, 7.14.
color/habit
cryst dimens (mm3)
cryst syst
space group
a, A
orange/prism
0.26 ꢀ 0.22 ꢀ 0.16
monoclinic
P21/c
12.537(8)
15.206(8)
16.338(9)
101.59(5)
3051(3)
4
200
1.462
0.808
1400
[{Ti(η5-C5H5)}2(μ-Cl)(μ-{(η5-C5Me4SiMeO)2(μ-O)})][Q] (Q=
ClB(C6F5)3 4B,Q=ClAl(C6F5)3 4Al). Compounds 2as (0.10 g,
0.15 mmol) and E(C6F5)3 (B(C6F5)3 0.076 g, 0.15 mmol; (0.5
b, A
c, A
β, deg
C7H8) Al(C6F5)3 0.085 g, 0.15 mmol) were stirred in CH2Cl2
V, A3
3
(2 mL) for 5 min. The volatiles were removed under vacuum, and
the oil was washed with hexane (2 mL) to give 4B (0.150 g, 85%)
and 4Al (0.152 g, 85%) as a brownish oils. Anal. Calcd for
C48H40O3Si2Ti2F15AlCl2 (1198.42): C, 48.06; H, 3.33. Found: C,
Z
T, K
Dcalcd, g cm-3
μ, mm-1
F(000)
1
48.85; H, 3.80. Data for 4B: H NMR (CDCl3): 0.30 (s, 6 H,
θ range, deg
index ranges (h, k, l)
no. of rflns collected
3.52 to 27.58
(16, (19, (21
25 126
6987/0.1449
4262
6987/232/353
0.0885/0.2215
0.1407/0.2561
1.007
0.913 and -0.754
P
SiMe), 2.06 (s, 6 H, C5Me4), 2.12 (s, 6 H, C5Me4), 2.18 (s, 6 H,
C5Me4), 2.50 (s, 6 H, C5Me4), 6.22 (s, 10 H, C5H5). 13C NMR
(CDCl3): -0.1 (SiMe), 12.5, 13.8, 16.1, and 16.6 (C5Me4), 118.5,
120.5, 131.1, 141.2, and 145.8 (C5Me4), 121.3 (C5H5), 137.7 and
148.8 (m, C6F5). 19F NMR (CDCl3): -129.8 (o-C6F5), -163.5
(bs, m-C6F5 and p-C6F5). Data for 4Al: 1H NMR (CDCl3): 0.30
(s, 6 H, SiMe), 2.06 (s, 6 H, C5Me4), 2.12 (s, 6 H, C5Me4), 2.18 (s,
6 H, C5Me4), 2.50 (s, 6 H, C5Me4), 6.22 (s, 10 H, C5H5). 13C
NMR (CDCl3): -0.1 (SiMe), 12.5, 13.8, 16.1, and 16.6 (C5Me4),
118.5, 120.5, 131.1, 141.2, and 145.8 (C5Me4), 121.3 (C5H5),
136.6, 141.3, and 150.1 (m, C6F5). 19F NMR (CDCl3): -122.1
(o-C6F5), -156.5 (p-C6F5) and -163.2 (m-C6F5).
no. of indep rflns/Rint
no. of obsd rflns (I > 2σ(I))
no. of data/restraints/params
R1/wR2 (I > 2σ(I))a
R1/wR2 (all data)a
GOF (on F2)a
largest diff peak and hole, e A-3
P
P
P
a R1= ( Fo|-|Fc )/ |Fo|; wR2={ [w(Fo2 - Fc2)2]/ [w(Fo2)2]}1/2
;
P
GOF = { [w(Fo2 - Fc2)2]/(n - p)}1/2
.
[{Ti(η5-C5H5)}2(μ-Cl)(μ-{(η5-C5Me4SiMeO)2(μ-O)})][B(C6-
F5)4] (4C). Compounds 2as (0.10 g, 0.15 mmol) and LiB-
(C6F5)4 2.5C4H10O (0.13 g, 0.15 mmol) were stirred in toluene
formation of 5s-C and 4X-C (X=Cl, Br, F) in ca. 1:1. Data for 4F-
C: 1H NMR (FC6D5): 0.21 (s, 6 H, SiMe), 1.78 (s, 6 H, C5Me4),
1.80 (s, 6 H, C5Me4), 1.84 (s, 6 H, C5Me4), 1.86 (s, 6 H, C5Me4),
5.73 (s, 10 H, C5H5). 13C NMR (FC6D5): -0.1 (SiMe), 11.8, 12.1,
15.1, and 15.6 (C5Me4), 119.5-149.4 (C5Me4), 119.8 (C5H5),
135.4, 138.7, and 149.4 (m, C6F5). 19F NMR (FC6D5): -53.3
(Ti-F-Ti), -132.7 (o-C6F5), -162.8 (p-C6F5), and -166.9 (m-
C6F5). Data for 4Br-C: 1H NMR (BrC6D5): 0.21 (s, 6 H, SiMe),
1.75 (s, 6 H, C5Me4), 1.79 (s, 6 H, C5Me4), 1.80 (s, 6 H, C5Me4),
1.83 (s, 6 H, C5Me4), 5.98 (s, 10 H, C5H5). 13C NMR (BrC6D5):
-1.0 (SiMe), 10.8, 11.3, 13.7, and 14.6 (C5Me4), 119.5-149.4
(C5Me4), 120.2 (C5H5), 135.4, 138.7, and 149.4 (m, C6F5). 19F
NMR (BrC6D5): -134.2 (o-C6F5), -162.2 (p-C6F5), and -166.6
(m-C6F5).
3
(2 mL) for 5 min. The solution was filtered, and the remaining
solid was washed with hexane (2 ꢀ 2 mL) to give the compound
4-C (0.152 g, 0.12 mmol, 85%). 1H NMR (CDCl3): 0.35 (s, 6 H,
SiMe), 2.14 (s, 6 H, C5Me4), 2.16 (s, 6 H, C5Me4), 2.37 (s, 6 H,
C5Me4), 2.38 (s, 6 H, C5Me4), 6.17 (s, 10 H, C5H5). 13C NMR
(CDCl3): 0.3 (SiMe), 13.0, 13.3, 16.0, and 16.5 (C5Me4), 119.7,
133.6, 140.1, 143.2, and 149.7 (C5Me4), 120.1 (C5H5), 136.8,
139.7, and 149.4 (m, C6F5). 19F NMR (CDCl3): -133.5
(o-C6F5), -162.5 (p-C6F5) and -166.4 (m-C6F5). Anal. Calcd
for C54H40O3Si2Ti2F20BCl (1313.86): C, 49.32; H, 3.04. Found:
C, 49.16; H, 3.11.
[{Ti(η5-C5H5)}{Ti(η5-C5H5)Me}(μ-{(η5-C5Me4SiMeO)2(μ-
O)})][B(C6F5)4] (5aC, 5sC). A 0.5 mL sample of CD2Cl2 pre-
viously cooled at -78 ꢀC was added to a mixture of 3as and 3aa in
ca. 1:1 (0.015 g, 0.023 mmol) and 1 equiv of [Ph3C][B(C6F5)4]
(0.021 g, 0.023 mmol) in a NMR tube at -78 ꢀC. The reaction was
immediately monitored by NMR spectroscopy, showing a com-
plete transformation into a mixture of 5aC, 5sC, and 4C in ca.
1:2:1 after 5 min at -50 ꢀC. The isomer 5aþ evolved to the chloro-
bridged ionic compound 4þ at ambient temperature. Data for
5aC: 1H NMR (CD2Cl2): 0.14 (s, 3 H, SiMe), 0.33 (s, 3 H, SiMe),
0.69 (s, 3 H, TiMe), 1.40-2.50 (s, 8 ꢀ 3 H, C5Me4), 5.98 (s, 5 H,
Single-Crystal X-ray Structure Determination of Compound
2aa. Suitable single crystals for the X-ray diffraction study were
grown from toluene. A crystal was selected, covered with
perfluorinated ether, and mounted on a Bruker-Nonius Kappa
CCD single-crystal diffractometer. Data collection was per-
formed at 200(2) K. The structure was solved, using the
WINGX package,68 by direct methods (SHELXS-97) and re-
fined by using full-matrix least-squares against F2 (SHELXL-
97).69 All non-hydrogen atoms were anisotropically refined, and
hydrogen atoms were geometrically placed and left riding on
their parent atoms. SIMU and DELU restraints were applied in
C5H5), 6.15 (s, 5 H, C5H5). 13C NMR (CD2Cl2): 53.4 (TiMe). 19
F
this structure. Full-matrix least-squares refinements were car-
P
2
2 2
NMR (CD2Cl2): -133.5 (o-C6F5), -162.5 (p-C6F5), and -166.4
(m-C6F5). Data for 5sC: 1H NMR (CD2Cl2): 0.16 (s, 3 H, SiMe),
0.31 (s, 3 H, SiMe), 0.66 (s, 3 H, TiMe), 1.42 (s, 3 H, C5Me4), 1.83
(s, 3 H, C5Me4), 2.00 (s, 3 H, C5Me4), 2.12 (s, 3 H, C5Me4), 2.13 (s,
3 H, C5Me4), 2.25 (s, 3 H, C5Me4), 2.37 (s, 3 H, C5Me4), 2.40 (s, 3
H, C5Me4), 5.98 (s, 5 H, C5H5), 6.15 (s, 5 H, C5H5). 13C NMR
(CD2Cl2): 53.4 (TiMe). 19F NMR (CD2Cl2): -133.5 (o-C6F5),
-162.5 (p-C6F5), and -166.4 (m-C6F5).
ried out by minimizing w(Fo - Fc ) with the SHELXL-97
weighting scheme and stopped at shift/err <0.001. Crystal data
and details of the structure determination are presented in
Table 8. Crystallographic data (excluding structure factors)
for the structures reported in this paper have been deposited
with the Cambridge Crystallographic Data center as supple-
mentary publication no. CCDC-756896 (2aa). Copies of the
data can be obtained free of charge on application to CCDC, 12
Union Road, Cambridge CB2 1EZ, UK (fax: (þ44)1223-336-
033; e-mail: deposit@ccdc.cam.ac.uk).
[{Ti(η5-C5H5)}2(μ-X)(μ-{(η5-C5Me4SiMeO)2(μ-O)})][B(C6-
F5)4] (X=F, 4F-C; Cl, 4Cl-C; Br 4Br-C). A mixture of 3as and 3aa
in ca. 1:1 (0.015 g, 0.023 mmol) and 1 equiv of [Ph3C][B(C6F5)4)]
(0.021 g, 0.023 mmol) were loaded into a NMR tube, and 0.5 mL
of the corresponding solvent CD2Cl2 or XC6D5 (X=Br, F) was
added. Each sample was then shaken vigorously, and the reaction
was monitored by NMR spectroscopy at 25 ꢀC, showing the
(68) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837.
(69) Sheldrick, G. M. SHELXL-97; University of Gottingen: Gottin-
gen, 1998.

Article
Computational Details. All the theoretical calculations in this
Organometallics, Vol. 29, No. 3, 2010 655
methyllithium, the lithium atom and the methyl group were
described with a 6-31G(d) basis set, whereas the toluene mole-
cule explicitly introduced was described by the 6-31G basis set.
To approximately account for the effects of the surrounding
medium, solvent corrections were calculated in toluene (ε =
2.379) as single points at the B3LYP/6-31þþG(d,p) level using
the conductor-like polarizable continuum (CPCM)78 model, in
which the cavity is created via a series of overlapping spheres.79
Finally, during the studies, it was sometimes necessary to
elucidate how the electron density was distributed around the
atoms, both the metallic centers, and the substituents.
To evaluate this, natural bond orbital analysis (NBO) was
performed.62,63
article have been performed at the density functional (DFT)
level of theory by using the Gaussian 03 program package.70
Becke’s three-parameter exchange functional (B3)71,72 in con-
junction with the Lee-Yang-Parr correlation functional
(LYP)73,74 was employed as implemented in Gaussian 03.
The optimizations were performed with basis set BS1; in this
basis set, the lanl2dz effective core potential75 has been used to
describe the inner electrons of the metal centers (Ti), whereas
their associated double-ζ basis set, with an extra series of
f functions,76 has been employed for the remaining electrons;
Si, Cl, O, and C atoms were described with Pople’s 6-31G(d)
split valence basis set, and H atoms were described with the
6-31G basis set.77 In the study of the alkylation reaction with
Acknowledgment. We gratefully acknowledge the
Spanish MICINN (DGI) (projects MAT2007-60997,
CTQ2008-06866-C02-01), Consolider Ingenio 2010
(CSD2007-00006), and DGUI-Comunidad de Madrid
(programme S-0505/PPQ-0328, COMAL-CM) (Spain)
for financial support. L.P. acknowledges DGUI-CM
for a Fellowship. The use of computational facilities of
(70) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J., J. A.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Lyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson,
G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai,
H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo,
C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin,
A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Moroku-
ma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.;
Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.;
Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.;
Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challa-
combe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.;
Gonzalez, C.; Pople, J. A. Gaussian 03, Revision C.02; Gaussian, Inc.:
Pittsburgh, PA, 2004.
ꢀ
the Centre de Supercomputacio de Catalunya (CESCA) is
gratefully appreciated.
Supporting Information Available: Crystallographic data for
2aa and selected NMR spectra. Detailed comparison of the
coordination spheres around titanium atoms in compounds of
series 2 and 3 and 2 and B, discussion on the choice of the model
system, highest energy pathways for methylation of 2as, and
Cartesian coordinates and absolute energies of all optimized
species. This material is available free of charge via the Internet
at http://pubs.acs.org.
(71) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(72) Becke, A. D. J. Chem. Phys. 1993, 98, 1372.
(73) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J.
J. Phys. Chem. 1994, 98, 11623.
(74) Lee, C.; Parr, R. G.; Yang, W. Phys. Rev. B 1988, 37, 785.
(75) Wadt, W. R.; Hay, P., J. J. Chem. Phys. 1985, 82, 284.
(77) Henre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A., Ab Initio
Molecular Orbital Theory; Wiley: New York, 1986.
(78) Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995.
(79) Miertus, S.; Scrocco, E.; Tomasi, J. Chem. Phys. Lett. 1981, 55,
117.
€
(76) Ehlers, A. W.; Bohme, M.; Dapprich, S.; Gobbi, A.; Hollwarth,
A.; Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.
€
€
Chem. Phys. Lett. 1993, 208, 111.