Enhanced reactivity of cationic hafnocene complexes toward σ-bond metathesis reactions. Si-H and Si-C bond activations in stoichiometric and catalytic organosilane conversions

Article abstract of DOI:10.1021/om0210107

Interest in developing highly active catalysts for processes involving σ-bond metathesis steps led to a study directly comparing neutral, zwitterionic, and cationic derivatives of CpCp*Hf- in reactions with silanes. In stoichiometric transformations, comparisons of the reactivities of CpCp*HfMe₂ (1), CpCp*HfMe(OTf) (2), CpCp*HfMe(μ-Me)B(C₆F₅)₃ (3), and [CpCp*HfMe] [B(C₆F₅)₄] (4) with organosilanes indicate that the cationic character in the complex influences the activity of the Hf-Me bond toward reactions with Si-H bonds. Enhanced activity in catalytic σ-bond metathesis reactions of PhSiH₃ was also observed with zwitterionic hafnium hydride complexes. Neutral CpCp*HfHCl (7) reacts slowly with PhSiH₃ to give oligomers, while zwitterionic CpCp*HfH(μ-H)B(C₆F₅) ₃ (6) reacts ca. 10 times faster, activating both Si-C and Si-H bonds of PhSiH₃, to give redistribution and dehydrocoupling products. A mechanism based on σ-bond metathesis is evidenced by identical experimental rate laws (rate ∝ [catalyst] [PhSiH₃]₂) for both the neutral and zwitterionic systems. The cationic complex 4 is an extremely reactive catalyst precursor and undergoes rapid redistribution and dehydropolymerization of PhSiH₃ to give highly cross-linked insoluble materials of the type H-(PhSiH)_n-(SiH)_m-H.

Full text of DOI:10.1021/om0210107

Organometallics 2003, 22, 3577-3585
3577
En h a n ced Rea ctivity of Ca tion ic Ha fn ocen e Com p lexes
tow a r d σ-Bon d Meta th esis Rea ction s. Si-H a n d Si-C
Bon d Activa tion s in Stoich iom etr ic a n d Ca ta lytic
Or ga n osila n e Con ver sion s
Aaron D. Sadow and T. Don Tilley*
Department of Chemistry, University of California, Berkeley, Berkeley, California 94720-1460
Received December 17, 2002
Interest in developing highly active catalysts for processes involving σ-bond metathesis
steps led to a study directly comparing neutral, zwitterionic, and cationic derivatives of
CpCp*Hf- in reactions with silanes. In stoichiometric transformations, comparisons of the
reactivities of CpCp*HfMe2 (1), CpCp*HfMe(OTf) (2), CpCp*HfMe(µ-Me)B(C6F5)3 (3), and
[CpCp*HfMe][B(C6F5)4] (4) with organosilanes indicate that the cationic character in the
complex influences the activity of the Hf-Me bond toward reactions with Si-H bonds.
Enhanced activity in catalytic σ-bond metathesis reactions of PhSiH3 was also observed with
zwitterionic hafnium hydride complexes. Neutral CpCp*HfHCl (7) reacts slowly with PhSiH3
to give oligomers, while zwitterionic CpCp*HfH(µ-H)B(C6F5)3 (6) reacts ca. 10 times faster,
activating both Si-C and Si-H bonds of PhSiH3, to give redistribution and dehydrocoupling
products. A mechanism based on σ-bond metathesis is evidenced by identical experimental
rate laws (rate
[catalyst][PhSiH3]2) for both the neutral and zwitterionic systems. The
cationic complex 4 is an extremely reactive catalyst precursor and undergoes rapid
redistribution and dehydropolymerization of PhSiH3 to give highly cross-linked insoluble
materials of the type H-(PhSiH)n-(SiH)m-H.
In tr od u ction
bonds (E ) H, Si, B, Sn, P) with high-valent organo-
metallic species containing reactive M-R σ-bonds (R )
H, C, Si, P, Sn). For olefin hydrogenations,7 hydrosila-
tions,8 and hydroborations9 mediated by early (d0)-
transition-metal and lanthanide complexes, the key
carbon-element bond formation proceeds via the inter-
action of an M-C bond with dihydrogen, a silane, or a
borane, respectively (eq 1). In silane, stannane, and
Facile σ-bond activations by electrophilic, high-valent
organometallic complexes hold considerable promise for
the development of catalytic processes. These σ-bond
metathesis reactions are typically associated with metal
centers in their highest oxidation states and involve
concerted steps and four-center electrocyclic transition
states.1 Such transition states have been postulated for
intramolecular ligand metalations2 and in C-H bond
activations of hydrocarbons such as methane and ben-
zene, by Cp*2MR (Cp* ) η5-C5Me5; M ) Sc, Lu, Y; R )
hydride, alkyl) complexes.3 These reactive complexes
possess several features that are required for σ-bond
metathesis transformations, including electrophilicity,
coordinative unsaturation, and a reactive M-R bond.
However, the utilization of these C-H bond activation
steps for catalytic hydrocarbon conversions has devel-
oped slowly.4-6
phosphine dehydropolymerizations,10-12 two sequential
E-H (E ) Si, Sn, P) bond activations produce dihydro-
gen and an element-element bond. Mechanistic studies
(4) (a) J ordan, R. F.; Taylor, D. F. J . Am. Chem. Soc. 1989, 111,
778. (b) J ordan, R. F.; Taylor, D. F.; Baenziger, N. C. Organometallics
1990, 9, 1546. (c) Rodewald, S.; J ordan, R. F. J . Am. Chem. Soc. 1994,
116, 4491.
(5) Sadow, A. D.; Tilley, T. D. Angew. Chem., Int. Ed. 2003, 42, 803.
(6) Sadow, A. D.; Tilley, T. D. J . Am. Chem. Soc. 2003, 125, 7971.
(7) (a) J eske, G.; Lauke, H.; Mauermann, H.; Schumann, H.; Marks,
T. J . J . Am. Chem. Soc. 1985, 107, 8111. (b) Giardello, M. A.; Conticello,
V. P.; Brard, L.; Gagne´, M. R.; Marks, T. J . J . Am. Chem. Soc. 1994,
116, 10241.
Several catalytic processes appear to involve σ-bond
metathesis, and these are based on reactions of E-H
(1) (a) Watson, P. L.; Parshall, G. W. Acc. Chem. Res. 1985, 18, 51.
(b) Rothwell, I. P. In Selective Hydrocarbon Activation; Davies, J . A.,
Watson, P. L., Liebman, J . F., Greenberg, A., Eds.; VCH: New York,
1990; p 43. (c) Watson, P. L. In Selective Hydrocarbon Activation;
Davies, J . A., Watson, P. L., Liebman, J . F., Greenberg, A., Eds.;
VCH: New York, 1990; p 79.
(2) (a) Simpson, S. J .; Turner, H. W.; Andersen, R. A. Inorg. Chem.
1981, 20, 2991. (b) Rothwell, I. P. Polyhedron 1985, 4, 177.
(3) (a) Watson, P. L. J . Chem. Soc., Chem. Commun. 1983, 276. (b)
Watson, P. L. J . Am. Chem. Soc. 1983, 105, 6491. (c) Thompson, M.
E.; Bercaw, J . E. Pure Appl. Chem. 1984, 56, 1. (d) Thompson, M. E.;
Baxter, S. M.; Bulls, A. R.; Burger, B. J .; Nolan, M. C.; Santarsiero, B.
D.; Schaefer, W. P.; Bercaw, J . E. J . Am. Chem. Soc. 1987, 109, 203.
(e) Fendrick, C. M.; Marks, T. J . J . Am. Chem. Soc. 1986, 108, 425.
(8) (a) Molander, G. A.; J ulius, M. J . Am. Chem. Soc. 1995, 117,
4415. (b) Fu, P.-F.; Brard, L.; Li, Y.; Marks, T. J . J . Am. Chem. Soc.
1995, 117, 7157. (c) Gountchev, T. I.; Tilley, T. D. Organometallics
1999, 18, 2896. (d) Voskoboynikov, A. Z.; Shestakova, A. K.; Beletskaya,
I. P. Organometallics 2001, 20, 2794. (e) Dash, A. K.; Gourevich, I.;
Wang, J . Q.; Wang, J .; Kapon, M.; Eisen, M. S. Organometallics 2001,
20, 5084.
(9) Harrison, K. N.; Marks, T. J . J . Am. Chem. Soc. 1992, 114, 9220.
10.1021/om0210107 CCC: $25.00 © 2003 American Chemical Society
Publication on Web 07/26/2003

3578 Organometallics, Vol. 22, No. 17, 2003
Sadow and Tilley
indicate that these transformations are possible be-
cause, unlike carbon, certain main-group elements can
occupy the â-position of a four-centered transition state
(eq 1).7-13 Thus, the accessibility of such transition
states allows for various bond-forming reactions involv-
ing main-group elements.
σ-Bond metathesis reactions are potentially useful in
the development of new catalytic processes, but a wider
application of this chemistry in catalysis requires a more
thorough understanding of the factors that influence
reactivity and selectivity and the development of more
reactive metal complexes. For example, studies of C-H
and Si-H bond activations in hydrocarbons and hy-
drosilanes, mediated by the reactive Cp*2ScR deriva-
tives (R ) H, Me, CH2CMe3), have led to the discovery
of the first examples of catalytic methane conversions
based on σ-bond metathesis.5,6
A strategy for the design of new bond activation
catalysts is suggested by the similarities between olefin
polymerization and σ-bond metathesis. Both mecha-
nisms proceed via four-centered electrocyclic transition
states (2σ + 2σ vs 2σ + 2π) and require electrophilic metal
centers containing an open coordination site and a
reactive σ-bond to the metal.14 However, the most
efficient dehydropolymerization catalysts are neutral,
16-electron, group 4 metal hydride complexes containing
a single, low-lying, vacant d orbital, while effective
group 4 metallocene-based olefin polymerization cata-
lysts are cationic 14-electron complexes with two avail-
able d orbitals.15 This comparison suggests that cationic
group 4 metallocene derivatives might be highly reactive
in σ-bond metathesis.
The rich C-H bond activation chemistry for neutral
scandium group and lanthanide complexes of the type
Cp*2LnR has not been observed for isoelectronic, cat-
ionic group 4 complexes of the type [Cp′2MR]+. For the
cationic complex [Cp*2ZrH][HB(C6F5)3], C-H bond ac-
tivation is apparently involved in rapid deuterium
exchange between the zirconium hydride position and
benzene-d6.17 In addition, the catalytic alkylation of
pyridine with R-olefins has been reported for the cationic
complex [Cp2ZrH(THF)][BPh4].18 These examples sug-
gest that cationic centers should also be reactive toward
silanes and Si-H bond activations.
Interestingly, previous attempts to employ cationic
catalysts in the dehydropolymerization of organosilanes
did not yield increased molecular weights for the pol-
ysilane products.14a,18 As we have learned in studies
with active, lanthanide-based catalysts, low-molecular-
weight polysilanes may result from the action of highly
active catalysts which promote competing redistribution
processes.19 Thus, the molecular weights of polysilanes
produced in a dehydropolymerization are not a good
measure of a catalyst’s activity toward σ-bond metath-
esis processes. For this reason, we have attempted to
make direct comparisons between the activities of
analogous neutral and cationic complexes in well-
defined σ-bond metathesis steps. The study presented
here addresses this fundamental question by comparing
the reactivity of neutral and cationic complexes toward
silanes in stoichiometric Si-C bond-forming reactions
and in catalytic organosilane dehydropolymerizations.
These findings, which have been communicated previ-
ously,20 indicate that cationic complexes show enhanced
reactivity toward silanes in σ-bond metathesis reactions.
Resu lts a n d Discu ssion
Gen er a tion of Neu tr a l a n d Ca tion ic Der iva tives
of Cp Cp *HfMe2. A series of hafnium methyl com-
plexes, CpCp*HfMe2 (1),21 CpCp*HfMe(OTf) (2),
CpCp*HfMe(µ-Me)B(C6F5)3 (3), and [CpCp*HfMe]-
[B(C6F5)4] (4), were synthesized to evaluate the effect
of cationic character on stoichiometric σ-bond meta-
thesis reactions with silanes. Compound 1 was prepared
by the literature method,21 and 2 was obtained by an
adaptation of the reported synthesis of [Cp2ZrMe(THF)-
[BPh4].21 Thus, reaction of 1 with 1 equiv of AgOTf in
benzene-d6 at room temperature rapidly yielded 2,
1
ethane, and Ag0 (eq 2). The H NMR spectrum of 2 in
benzene-d6
CpCp*HfMe2 + AgOTf
8
(10) (a) Harrod, J . F.; Mu, Y.; Samuel, E. Polyhedron 1991, 10, 1239.
(b) Tilley, T. D. Acc. Chem. Res. 1992, 26, 22. (c) Gauvin, F.; Harrod,
J . F.; Woo, H.-G. Adv. Organomet. Chem. 1998, 42, 363. (d) Aitken,
C.; Harrod, J . F.; Samuel, E. J . Organomet. Chem. 1985, 279, C11.
(11) (a) Aitken, C. T.; Harrod, J . F.; Samuel, E. J . Am. Chem. Soc.
1986, 108, 4059. (b) Woo, H.-G.; Tilley, T. D. J . Am. Chem. Soc. 1989,
111, 8043. (c) Corey, J . Y.; Zhu, X.-H.; Bedard, T. C.; Lange, L. D.
Organometallics 1991, 10, 924. (d) Woo, H.-G.; Walzer, J . F.; Tilley, T.
D. J . Am. Chem. Soc. 1992, 114, 7047.
(12) (a) Imori, T.; Tilley, T. D. J . Chem. Soc., Chem. Commun. 1993,
1607. (b) Imori, T.; Lu, V.; Cai, H.; Tilley, T. D. J . Am. Chem. Soc.
1995, 117, 9931. (c) Fermin, M. C.; Stephan, D. W. J . Am. Chem. Soc.
1995, 117, 12645. (d) Etkin, N.; Fermin, M. C.; Stephan, D. W. J . Am.
Chem. Soc. 1997, 119, 2954. (e) Xin, S.; Woo, H.-G.; Harrod, J . F.;
Samuel, E.; Lebuis, J .-M. J . Am. Chem. Soc. 1997, 119, 5307.
(13) Piers, W. E.; Shapiro, P. J .; Bunel, E. E.; Bercaw, J . E. Synlett
1990, 74.
1
CpCp*HfMe(OTf) + C2H6 + Ag0 (2)
2
benzene-d6 contains singlets corresponding to the Cp,
Cp*, and HfMe ligands in the expected ratio of 5:15:3.
The HfMe resonance of 2 is downfield-shifted relative
to that of the hafnium dimethyl, from -0.48 to 0.18
ppm. The relatively high solubility of 2 in nonpolar
solvents such as pentane indicates that the OTf- ligand
is covalently bonded to the hafnium center.
(14) (a) Imori, T.; Tilley, T. D. Polyhedron 1994, 13, 2231. (b)
Dioumaev, V. K.; Harrod, J . F. Organometallics 1994, 13, 1548.
(15) (a) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.
Principles and Applications of Organometallic Chemistry; University
Science Books: Mill Valley, CA, 1987; p 577. (b) Grubbs, R. H.; Coates,
G. W. Acc. Chem. Res. 1996, 29, 225 and references therein.
(16) (a) Yang, X.; Stern, C. L.; Marks, T. J . J . Am. Chem. Soc. 1991,
113, 3623. (b) Yang, X.; Stern, C. L.; Marks, T. J . J . Am. Chem. Soc.
1994, 116, 10015.
(17) (a) J ordan, R. F.; Dasher, W. E.; Echols, S. F. J . Am. Chem.
Soc. 1986, 108, 1718. (b) J ordan, R. F.; LaPointe, R. E.; Bajgur, C. S.;
Echols, S. F.; Willett, R. J . Am. Chem. Soc. 1987, 109, 4111. (c) J ordan,
R. F.; LaPointe, R. E.; Bradley, P. K.; Baenziger, N. Organometallics
1989, 8, 2892.
(18) (a) Dioumaev, V. K.; Harrod, J . F. J . Organomet. Chem. 1996,
521, 133. (b) Dioumaev, V. K.; Harrod, J . F. Organometallics 1996,
15, 3859. (c) Dioumaev, V. K.; Harrod, J . F. Organometallics 1997,
16, 2798. (d) Dioumaev, V. K.; Rahimian, K.; Gauvin, F.; Harrod, J . F.
Organometallics 1999, 18, 2249.
(19) (a) Radu, N. S.; Tilley, T. D. J . Am. Chem. Soc. 1995, 117, 5863.
(b) Radu, N. S.; Tilley, T. D.; Rheingold, A. L. J . Organomet. Chem.
1996, 516, 41. (c) Castillo, I.; Tilley, T. D. Organometallics 2001, 20,
5598.
(20) Sadow, A. D.; Tilley, T. D. Organometallics 2001, 20, 4457.
(21) (a) Rogers, R. D.; Benning, M. M.; Kurihara, L. K.; Morairty,
K. J .; Rausch, D. J . Organomet. Chem. 1985, 225, 1. (b) J ordan, R. F.;
Bajgur, C. S.; Dasher, W. E.; Rheingold, A. L. Organometallics 1987,
6, 1041.

Cationic Hafnocene Complexes
Organometallics, Vol. 22, No. 17, 2003 3579
By analogy to the literature synthesis of Cp*2ZrMe-
(µ-Me)B(C6F5)3,16 the reaction of CpCp*HfMe2 (1) with
the strong Lewis acid B(C6F5)3 at room temperature
rapidly (<5 min) and quantitatively produced the zwit-
terionic complex CpCp*HfMe(µ-Me)B(C6F5)3 (3; eq 3).
CpCp*HfMe2 + B(C6F5)3 f
1
CpCp*HfMe(µ-Me)B(C6F5)3
(3)
3
Compound 3 is soluble in benzene-d6 and bromoben-
zene-d5 but insoluble in aliphatic hydrocarbon solvents
such as pentane and hexamethyldisiloxane (HMDS).
Although both 1 and B(C6F5)3 are extremely soluble in
pentane at room temperature, when the two reagents
were mixed in that solvent, complex 3 precipitated as
an analytically pure, colorless powder. The 1H NMR
spectrum of 3 in benzene-d6 reveals singlets correspond-
ing to Cp, Cp*, and HfMe at 5.35 (5 H), 1.33 (15 H),
and 0.09 ppm (3 H). A broad resonance at 0.15 ppm (3
H) corresponds to the methylborate counterion. Both
methyl resonances are shifted downfield in comparison
to the dimethyl peak in 1. A singlet in the 11B NMR
spectrum (δ -17.5) and a 1H-11B HMQC experiment
confirm the formation of the methylborate anion. With
less than 1 equiv of B(C6F5)3, mixtures of 3 and the
dimeric, diastereomeric hafnium monocationic com-
plexes [(CpCp*HfMe)2(µ-Me)][MeB(C6F5)3] were formed.
The 1H NMR spectrum of the two diastereomers consists
of two sets of peaks corresponding to Cp and Cp*
ligands, bridging methyl groups (-1.81 and -1.85 ppm;
3 H each), and terminal methyl groups (-0.40 and
-0.44; 6 H each). Upon addition of a full 1 equiv of
B(C6F5)3 to this mixture, 3 was formed quantitatively.
Attempts to grow X-ray-quality crystals of 3 by cooling
concentrated toluene solutions or by slow diffusion of
pentane into toluene solutions were unsuccessful.
Reaction of colorless 1 with a benzene-d6 solution of
[Ph3C][B(C6F5)4] induced a color change from the in-
tense orange (for the Ph3C+ solution) to pale yellow. This
reaction produced Ph3CMe, but the desired [CpCp*HfMe]-
[B(C6F5)4] was not formed. Instead, mixtures containing
large amounts of unreacted 1 and two other CpCp*-
containing products were observed by 1H NMR spec-
troscopy. In THF-d8, Ph3CMe and a single complex
containing peaks corresponding to the Cp, Cp*, and
methyl ligands in a ratio of 5:15:3 was formed. However,
after 12 h the solution solidified, apparently due to the
formation of polymeric THF.
and anion occurs in that solvent. The reaction of this
pair of diastereomers with 0.5 equiv of [Ph3C][B(C6F5)4]
then converted the mixture to 4.
1
The H NMR spectrum of 4 contained the expected
three singlets, and as in 2 and 3, the shift for the HfMe
group is downfield relative to that for neutral 1. At room
temperature, no interaction between the cation and
anion was detected by 19F NMR spectroscopy. In fact,
the 19F NMR spectra of [Ph3C][B(C6F5)4] and 4 are
identical, and both contain three resonances at -132
(6 F), -162 (3 F), and -166 ppm (6 F). Unfortunately,
all attempts to isolate 4 were unsuccessful. Reaction of
1 with [Ph3C][B(C6F5)4] in pentane induced rapid pre-
cipitation of the insoluble dimeric monocations, thus
inhibiting complete conversion to 4. Therefore, the
studies of σ-bond metathesis reactions involving 4 and
silanes required in situ generation of the starting
hafnium compound.
Interestingly, the nature of the cyclopentadienyl
ancillary ligands affects the rate of formation of the
monomeric cationic complexes. A survey of related
abstraction reactions indicates that the rates of conver-
sion to the monomeric cationic complexes [Cp′2HfMe]-
[B(C6F5)4] from the dimeric intermediate follow the
trend Cp2Hf- < CpCp*Hf- < Cp*2Hf-. This trend
appears to reflect the destabilization of the dimeric
intermediates by the sterically demanding Cp* ligand.
The Cp*2Hf derivative further reacted in bromobenzene-
d5 (over ca. 12 h) to form two unidentified compounds
1
(by H NMR spectroscopy). Compound 4 was stable for
at least 1 week in bromobenzene-d5 at room tempera-
ture under an N2 atmosphere, but exposure of this
solution to air resulted in rapid decomposition to a black
material.
Cationic complex 4 reacted cleanly with PMe3 to form
the colorless adduct [CpCp*HfMe(PMe3)][B(C6F5)4] (5),
which was isolated by crystallization from bromoben-
zene-d5 or fluorobenzene solution layered with pentane.
1
The H NMR shift of the HfMe group in the 16-electron
The use of haloarene solvents proved critical in the
synthesis of the desired cationic complex. Thus, addition
of bromobenzene-d5 or fluorobenzene solutions of [Ph3C]-
[B(C6F5)4] to 1 quantitatively produced deep red solu-
tions of [CpCp*HfMe][B(C6F5)4] (4) and Ph3CMe after
1 h (eq 4). Initially, both diastereomers of [(CpCp*HfMe)2-
(µ-Me)][B(C6F5)4] formed, but within 1 h the dimeric
products were quantitatively converted to 4. Addition
of 0.5 equiv of [Ph3C][B(C6F5)4] to 1 in bromobenzene-
d5 resulted in quantitative formation of the two dimeric
diastereomers. Interestingly, the 1H NMR spectra of the
cationic hafnium fragment of the diastereomeric
complexes [(CpCp*HfMe)2(µ-Me)][MeB(C6F5)3] and
[(CpCp*HfMe)2(µ-Me)][B(C6F5)4] are identical in bro-
mobenzene-d5, suggesting that separation of the cation
PMe3 adduct is upfield (-0.47 ppm) relative to that of
the formally 14-electron, zwitterionic, cationic hafnium
methyl complexes 3 and 4. An attempt to remove the
datively bound phosphine ligand by exposure of crystals
of 5 to dynamic vacuum (ca. 1.3 Pa) for 6 h was
unsuccessful.
σ-Bon d Meta th esis Rea ction s of Cp Cp *HfMe(X)
(X ) Me, OTf, MeB(C6F 5)3, B(C6F 5)4) w ith Sila n es.
Phenylsilane was added to compounds 1-4 to evaluate
the relative reactivity of their respective Hf-C bonds
toward σ-bond metathesis. No reaction occurred be-
tween the neutral hafnocene 1 and 1 equiv of PhSiH3
at room temperature in benzene-d6 over 1 week. The
hafnium triflate 2 also did not react with PhSiH3 in
benzene-d6 after 5 days at 25 °C or over 12 h at 70 °C.

3580 Organometallics, Vol. 22, No. 17, 2003
Sadow and Tilley
In contrast, the zwitterionic complex 3 reacted with
1 equiv of PhSiH3 (room temperature, 3 h, benzene-d6)
to quantitatively form PhMe2SiH and CpCp*HfH(µ-
H)B(C6F5)3 (6, eq 5). Thus, both the hafnium methyl and
cies were formed regardless of the silane reagent. The
major product contained resonances at 5.60 and 1.86
ppm for the Cp and Cp* ligands, respectively, and the
resonances for a minor product were observed at 5.84
and 1.64 ppm. Methane was detected in the reaction
mixture, which contrasts with the quantitative Si-C
bond formation observed in reactions of 3 with orga-
nosilanes. As with 3, compound 4 reacts with unhin-
dered organosilanes more rapidly than with bulky
CpCp*HfMe2(µ-Me)B(C6F5)3
+ PhSiH3 f
3
CpCp*HfH(µ-H)B(C6F5)3 + PhMe2SiH (5)
6
silanes: PhSiH3 ≈ CySiH3 ≈ HexSiH3 > MesSiH3
>
Ph2SiH2 > Mes2SiH2 ≈ Ph3SiH. In reactions with less
hindered primary silanes, 4 was completely consumed
before an initial spectrum could be obtained. Thus, this
cationic hafnium methyl complex appears to be ex-
tremely reactive toward organosilanes.
a methyl group from the borate of 3 were transferred
to PhSiH3, and methane was not detected in the 1H
NMR spectrum. Although the reaction required ap-
proximately 3 h to proceed to completion, the presumed
mixed hydride/methyl intermediates CpCp*HfH(µ-Me)B-
(C6F5)3 and CpCp*HfMe(µ-H)B(C6F5)3 were not detected
by 1H NMR spectroscopy. However, the intermediate
silane product PhMeSiH2 was observed. The reaction
of 3 with 0.5 equiv of PhSiH3 yielded PhMe2SiH and a
1:1 mixture of 3 and 6. Ligand exchange between 3 and
6, to give the mixed hydrido/methyl species, also did not
occur. Addition of excess PhSiH3 (ca. 5 equiv) to 3
rapidly produced 6 and 2 equiv of PhMeSiH2 (t1/2 < 5
min) at room temperature.
Many σ-bond metathesis reactions of silanes, includ-
ing dehydropolymerizations,10,22 are accelerated by light.
However, reactions of 3 proceed identically under ambi-
ent light or in the dark. Due to the crowded nature of
the concerted transition states, the reaction rate is
sensitive to steric effects imposed by substituents on the
silane.23 Thus, larger silanes reacted more slowly than
smaller silanes with 3 (1 equiv of silane in benzene-d6
The reaction of dihydrogen with 4 produced methane
and the same two organometallic products that are
produced in the presence of silanes. Therefore, the
species formed in the reactions described above are not
cationic hafnium silyl complexes.26 Although [CpCp*HfH]-
[B(C6F5)4] could not be spectroscopically identified, its
formation is suggested by the observed reaction chem-
istry (see below). A compound analogous to this putative
cationic hafnium hydride, [Cp*2ZrH][B(C6F5)4], was
previously reported.27 Attempts to isolate [CpCp*HfH]-
[B(C6F5)4] via the procedure used for the zirconium
species (H2 in toluene at -78 °C) were not successful.
The broad dihydrogen resonance (or the broad SiH
resonance observed upon reaction of 4 and hydrosilanes)
suggests that rapid hydrogen exchange processes may
obscure the hafnium hydride resonance of [CpCp*HfH]-
[B(C6F5)4]. Note that the hydride resonance in isoelec-
tronic Cp*2ScH could not be detected by 1H NMR
spectroscopy, due in part to similar rapid exchange with
reagents used to generate this hydride (H2 or hydro-
silanes).3,5 Interestingly, under an atmosphere of H2 this
mixture catalyzed H/D exchange among the meta posi-
tion of bromobenzene-d5, cyclooctane, and the Cp ligands.
Thus, deuterium was incorporated into the typically
inert secondary methylene groups of cyclooctane such
that 92% of the cyclooctane was deuterated after 24 h.
This H/D exchange is presumed to proceed via metala-
tion of bromobenzene-d5 and elimination of HD or D2,
followed by hydrogenolysis to form bromobenzene-d4
and a cationic hafnium deuteride complex. Surprisingly,
the ortho and para positions of bromobenzene-d5 and
Ph3CMe were unaffected.
at room temperature), following the trend PhSiH3
CySiH3 > MesSiH3 > Ph2SiH2 > Mes2SiH2 (Cy ) C6H13
>
,
Mes ) 2,4,6-trimethylphenyl). Thus, the initial buildup
of the PhMeSiH2 intermediate results from the relative
rates of the reactions of 3 with PhSiH3 vs PhMeSiH2.
Tertiary silanes, such as PhMe2SiH, Ph3SiH, and Et3-
SiH, did not react with 3 (1:1) in benzene-d6 at room
temperature over 1 week.
Note that the reaction of 3 with PhSiH3 models the
Si-C bond-forming step in the hydrosilation of olefins,
as catalyzed by early-transition-metal and lanthanide
complexes.8 Also, Cp*2ZrMe(µ-Me)B(C6F5)3 was reported
to be a diene cyclization-hydrosilation catalyst precur-
sor.24 Similar stoichiometric Si-C bond formations have
been observed in the reactions of lanthanide methyl
compounds [Cp′2LnMe]2 (Cp′ ) Cp*, Ln ) Lu, Y; Cp′ )
The rapid reactions of 4 with silanes clearly indicate
that this cationic complex is considerably more reactive
toward silanes than the zwitterionic compound 3. The
relative rates of reactions with hydrosilanes (conversion
of the hafnium methyl species) follow the trend 4 > 3
> 2 ≈ 1. Thus, cationic character appears to signifi-
cantly impact the reactivity of a complex toward σ-bond
metathesis.
C5H4 Bu, Ln ) Lu, Sm, Nd) with silanes.25
t
The cationic hafnium methyl complex 4 reacted
rapidly with primary and secondary organosilanes in
bromobenzene-d5 at room temperature, to give compli-
cated mixtures of products; however, some characteristic
features of the reaction mixtures could be identified. For
example, spectroscopically identical organometallic spe-
Syn th esis a n d Ch a r a cter iza tion of Cp Cp *HfH-
(µ-H)B(C6F 5)3 (6). The zwitterionic hafnium hydride
complex 6 was prepared by vigorously stirring a pentane
suspension of 3 with 1.5 equiv of PhSiH3 for 3 h at room
temperature. Alternatively, compound 6 could be syn-
thesized via reaction of 3 with H2 (1 atm, pentane). The
(22) Casty, G. L.; Lugmair, C. G.; Radu, N. S.; Tilley, T. D.; Walzer,
J . F.; Zargarian, D. Organometallics 1997, 16, 8.
(23) Woo, H.-G.; Heyn, R. H.; Tilley, T. D. J . Am. Chem. Soc. 1992,
114, 5698.
(24) Molander, G. A.; Corrette, C. P. Tetrahedron Lett. 1998, 39,
5011.
(25) (a) Radu, N. S.; Hollander, F. J .; Tilley, T. D.; Rheingold, A. L.
J . Chem. Soc., Chem. Commun. 1996, 2459. (b) Castillo, I.; Tilley, T.
D. Organometallics 2000, 19, 4733. (c) Castillo, I.; Tilley, T. D. J . Am.
Chem. Soc. 2001, 123, 10526. (d) Voskoboynikov, A. Z.; Parshina, I.
N.; Shestakova, A. K.; Butin, K. P.; Beletskaya, I. P.; Kuz’mina, L. G.;
Howard, J . A. K. Organometallics 1997, 16, 4041.
(26) Sadow, A. D.; Tilley, T. D. J . Am. Chem. Soc. 2002, 124, 6814.
(27) J ia, L.; Yang, X.; Stern, C. L.; Marks, T. J . Organometallics
1997, 16, 842.

Cationic Hafnocene Complexes
Sch em e 1
Organometallics, Vol. 22, No. 17, 2003 3581
Ch a r t 1
um derivatives of the type CpCp*HfRCl (R ) H, SiH2-
Ph).11b,23 However, the kinetics of the overall catalytic
process (i.e. monomer consumption) has not been re-
ported. To better understand the relationship between
σ-bond metathesis reactivity and the molecular weights
of polysilanes produced by a catalytic dehydropolymer-
ization, we investigated the kinetics of the dehydro-
polymerization of PhSiH3, as catalyzed by 6 and 7.
Addition of PhSiH3 to 7 (neat or in benzene-d6), which
is known to react with silanes via four-centered transi-
tion states,10 produced low-molecular-weight polysilanes
(neat, Mn ≈ 1000 by GPC analysis). The initial stages
of the reaction were monitored by 1H NMR spectroscopy
in benzene-d6, which revealed that the silane reacted
to form dehydrocoupling products. Thus, H2, PhH2Si-
SiH2Ph, and PhH2Si-(SiHPh)-SiH2Ph were detected,
and after 1 day mixtures of linear and cyclic silicon
chains were formed (identified by their characteristic
1H NMR chemical shifts).11d,29 Additionally, the hafnium
hydride 7 was partially converted to the corresponding
silyl compound CpCp*Hf(SiH2Ph)Cl.11d For reactions in
sealed NMR tubes, plots of 1/[PhSiH3] vs time were
linear for approximately 2 half-lives but indicated that
the reaction slowed after extended reaction times at
high PhSiH3 conversions (due to a buildup of the H2
concentration; vide infra). However, when the reaction
vessel was opened to a N2 atmosphere (i.e., H2 was
allowed to escape), the reaction exhibited pseudo-second-
order kinetics for greater than 3 half-lives. Plots of the
observed second-order rate constant kobs vs [7] demon-
strate first-order dependence on the hafnium hydride
catalyst, indicating that silicon-silicon bond formation
occurs at the metal center. A mechanism consistent with
these observations involves reversible, concurrent silane
metalation and H2 elimination with 7 (Scheme 1). As
expected, the addition of 1 atm of H2 inhibited the
polymerization. Overall second-order dependence on
substrate concentration is expected for a condensation
polymerization reaction;30 another 1 equiv of PhSiH3 is
required for a single turnover of the catalytic cycle. The
rate law for such a mechanism, rate ) kobs[7][PhSiH3]2
(kobs ) KAeqkA2), is consistent with the experimentally
determined rate law (kobs ) 3.3 × 10-3 s-1 M-2 at 51.2
°C). The buildup of CpCp*Hf(SiH2Ph)Cl indicates that
the silicon-silicon bond formation step of the cycle is
rate-limiting, at least during the early stages of the
polymerization. This conclusion is consistent with the
postulate that σ-bond metathesis steps involving hin-
dered transition states are slow (Chart 1). Thus, these
results are consistent with the previously proposed
catalytic cycle (Scheme 1).10b,11d
1H NMR spectrum of 6 (benzene-d6) contains a hydride
peak at 12.84 ppm (1 H), in addition to the Cp and Cp*
resonances. Due to the quadrupolar boron nucleus, the
hydridoborate signal was broadened into the baseline
1
and could not be detected directly by H NMR spectros-
copy, but it was identified using a 1H-11B HMQC
experiment (at 0.5 ppm). The 11B NMR resonance
1
(-16.91 ppm) and the H-11B HMQC data indicate the
presence of an anionic hydroborate counterion. More-
over, the 1J BH coupling constant (50 Hz) and the boron-
hydrogen stretching frequency (1950 cm-1) are consis-
tent with a hydride bridging the boron and hafnium
centers, Hf-(µ-H)-B. For comparison, the crystallo-
graphically characterized, cationic zirconocene hydride
[Cp*2ZrH][HB(C6F5)3] was shown to contain a terminal
boron hydride (ν(B-H) 2346 cm-1) and aryl fluorides
bridging to the unsaturated Zr center.28 Interestingly,
the compound [Cp*2ZrH][MeB(C6F5)3] was isolated by
the reaction of [Cp*2ZrMe][MeB(C6F5)3] with H2 in
pentane,16,28 while the mixed hafnium hydride/methyl
borate CpCp*HfH(µ-Me)B(C6F5)3 was never detected. In
contrast to [Cp*2ZrH][HB(C6F5)3], which reacts rapidly
with benzene-d6 to incorporate deuterium into its hy-
dride position, zwitterionic 6 does not react with benzene-
d6 at room temperature over 1 week.
Rea ction s of Cp Cp *HfHX (X ) Cl, HB(C6F 5)3)
w ith Sila n es. Or ga n osila n e Deh yd r op olym er iza -
tion a n d Red istr ibu tion . Dehydropolymerizations of
excess PhSiH3 by zwitterionic 6 (2 mol %, 2 days, neat
substrate) produced polysilanes of similar molecular
weight (Mn ≈ 1000) to those formed by neutral CpCp*Hf-
HCl (7) under analogous conditions. Although these low
molecular weights are consistent with previous findings
regarding PhSiH3 polymerizations by cationic group 4
complexes,14,18 the studies described above comparing
reactions of neutral vs cationic hafnium methyl com-
pounds with silanes indicate that cationic species should
be significantly more reactive toward σ-bond metathesis.
These seemingly contradictory observations prompted
a closer examination of neutral and cationic hafnocene
hydride complexes in silane dehydropolymerizations.
The catalytic cycle of Scheme 1 was proposed on the
basis of studies on observed stoichiometric reactions,
which represent the two steps of this cycle. These
mechanistic studies involved neutral, mixed-ring hafni-
(28) Yang, X.; Stern, C. L.; Marks, T. J . Angew. Chem., Int. Ed. Engl.
1992, 31, 1375.
(29) Aitken, C.; Barry, J .-P.; Gauvin, F.; Harrod, J . F.; Malek, A.;
Rousseau, D. Organometallics 1989, 8, 1732.
(30) Odian, G. Principles of Polymerization, 3rd ed.; Wiley-Inter-
science: New York, 1991.

3582 Organometallics, Vol. 22, No. 17, 2003
Sadow and Tilley
Sch em e 2
The reaction of neat PhSiH3 with zwitterionic 6
resulted in effervescence, and a gelatinous oil formed
after approximately 5 h (eq 6). Despite the visibly
5 mol % 6
nPhSiH3
8 H(SiPhH)n(SiH)mH +
neat or in benzene-d6
H2 + Ph2SiH2 + SiH4 (6)
vigorous reaction, gel-permeation chromatography (GPC)
traces of the resulting product indicated only short
oligosilane formation (Mn ≈ 1000). The 1H and 29Si NMR
spectra of the crude reaction revealed a mixture of
polysilanes, Ph2SiH2, and unreacted PhSiH3. Monitoring
the reaction by 1H NMR spectroscopy (benzene-d6,
sealed NMR tube) revealed that H2 and SiH4 formed as
well. The 29Si{1H} NMR spectra of the polysilane
produced by 6 exhibited two bands of resonances from
-55 to -65 ppm and from -95 to -120 ppm. The former
resonances are attributed to structures of the type
-(PhSiH)n-, on the basis of comparisons to resonances
for polysilanes produced by 7 and other neutral group
°C), takes the same form as that of the simple dehy-
dropolymerization reaction catalyzed by CpCp*HfHCl.
Notably, the observed rate constants are greater than
those for dehydropolymerization, as catalyzed by the
neutral catalyst 7, by more than an order of magnitude.
Unlike the CpCp*HfHCl-catalyzed dehydropolymeriza-
tion, the reaction could be performed in a sealed NMR
tube. However, at high conversions of PhSiH3 (<4 half-
lives), the observed rate constant decreased slightly due
to increased [H2] from the silane redistribution process.
The redistribution process accounts for much of the
PhSiH3 conversion (ca. 66%), and therefore relatively
large amounts of H2 (relative to the silane dehydro-
polymerization catalyzed by compound 7) are required
to inhibit the catalysis. In related work, Rosenberg
recently reported that the dehydrocoupling of Ph2SiH2
to Ph2SiSiH2, as catalyzed by (PPh3)3RhCl, could be
favored over redistribution to PhSiH3 and Ph3SiH by
rapid H2 removal.31
4 catalysts.10
A
29Si{1H} DEPT 90° spectra of the
polysilane produced by 6 indicated that the latter set
of resonances were due to HSi(Si3) centers, apparently
resulting from incorporation of SiH4 into the polymer
backbone. Notably, the formation of Ph2SiH2 and SiH4
demonstrate that organosilane redistribution, via Si-C
bond activations, occurred. As in the [Cp*2LnH]2 sys-
tems, the cleavage of Si-C bonds implies that the
cationic metal-hydride species are highly reactive. The
corresponding reaction between 6 and CySiH3 (25 equiv)
was slow (only 25% conversion after 3 days at room
temperature in benzene-d6), and the Si-C bond activa-
tion product Cy2SiH2 was formed. No reaction was
observed between 6 and Ph2SiH2 (23 equiv) in benzene-
d6 after 3 days.
In contrast to the reaction of 7 with PhSiH3, in which
CpCp*Hf(SiH2Ph)Cl was formed,11d the presumed cat-
ionic hafnium silyl intermediate CpCp*Hf(SiH2Ph)(µ-
H)B(C6F5)3 was not detected in reaction mixtures of 6
with PhSiH3 (by 1H NMR spectroscopy, benzene-d6, over
8 h at room temperature or 35 °C). Attempts to
independently synthesize the cationic hafnium hydro-
silyl complex CpCp*Hf(SiH2Ph)(µ-H)B(C6F5)3 were un-
successful, and the synthesis and σ-bond metathesis
chemistry of related cationic hafnium silyl complexes
is presented elsewhere.26 The cationic hafnium phenyl
complex CpCp*HfPh(µ-H)B(C6F5)3, which is the pre-
sumed intermediate in the redistribution of PhSiH3 to
Ph2SiH2 and SiH4, was also not detected in this reaction
mixture. Unfortunately, independent synthesis of the
putative phenyl intermediate CpCp*HfPh(µ-H)B(C6F5)3
was unsuccessful, as the reaction of CpCp*HfPhH with
B(C6F5)3 yielded a mixture of products. The only orga-
nometallic species observed in the reaction mixture was
6, indicating that this cationic hafnium hydride complex
represents the resting state of the catalyst.
Although the dehydropolymerization and redistribu-
tion of PhSiH3 are both catalyzed by 6, the observed
kinetic behavior of this reaction system is remarkably
simple. Second-order plots of 1/[PhSiH3] vs time were
linear for greater than 3 half-lives, and plots of kobs vs
[6] revealed first-order dependence on the catalyst
concentration. The experimentally determined rate law,
rate ) kobs[6][PhSiH3]2 (kobs ) 4.5 × 10-2 s-1 M-2 at 35
The significantly increased reaction rate for 6 appears
to reflect the enhanced reactivity of the cationic system
toward organosilanes. Furthermore, the similarity of the
two experimentally determined rate laws suggests that
the cationic hafnium catalyst mediates organosilane
dehydrocoupling via a mechanism analogous to that of
the neutral catalyst, involving σ-bond metathesis steps.
Analysis of the kinetic data and the observed reaction
products suggests that phenylsilane redistribution pro-
ceeds through a two-step mechanism involving revers-
ible Si-C bond cleavage by 6 followed by transfer of the
phenyl group from hafnium to silicon (Scheme 2). The
summation of the rate laws for redistribution and
dehydropolymerization (both first order in 6 and second
order in PhSiH3) yields an overall rate law (eq 7), for
KAeqkA
KBeqkB
2
2
rate )
+
[6][PhSiH3]2
(7)
(
)
[H2]
[SiH4]
which the second term, KBeqkB2/[SiH4], is derived from
the PhSiH3 redistribution reaction. Because the first
step of both pathways is an equilibrium, the relative
rates of the pathways should be dependent on the
concentrations of the H2 and SiH4 byproducts in the
reaction mixture. In accord with this mechanism, ad-
(31) Rosenberg, L.; Davis, C. W.; Yao, J . J . Am. Chem. Soc. 2001,
123, 5120.

Cationic Hafnocene Complexes
Organometallics, Vol. 22, No. 17, 2003 3583
dition of 1 atm of H2 inhibited dehydrocoupling to favor
redistribution (by a factor of 4:1 based on the amount
of Ph2SiH2 relative to the initial PhSiH3).
excess PhSiH3 to form the same dehydropolymerization
and redistribution products that were observed in
reactions of PhSiH3 with [CpCp*HfMe][B(C6F5)4] (gen-
erated from 1:1 mixtures of 1 and [Ph3C][B(C6F5)4]).
Additionally, the cationic hafnium complex [CpCp*HfMe]-
[B(C6F5)4] did not react with Ph3SiH (room temperature,
bromobenzene-d5, 1 day), whereas [Ph3C][B(C6F5)4]
rapidly reacted with Ph3SiH.18a,32 For example, a mix-
ture of 1 and 1.1 equiv of [Ph3C][B(C6F5)4] (bromoben-
zene-d5, 2 h) was allowed to react with 0.2 equiv of
Ph3SiH; after 1 day at room temperature, the cationic
hafnium methyl 4 had not reacted, but the excess [Ph3C]-
[B(C6F5)4] had been consumed (by 1H NMR spectros-
copy). This mixture retained the same activity toward
reaction with PhSiH3, and thus we conclude that [Ph3C]-
[B(C6F5)4] is not the active catalyst in these silane
dehydropolymerization and redistribution reactions.
Rapid Or gan osilan e Deh ydr opolym er ization an d
R ed ist r ib u t ion w it h [Cp Cp *H fMe][B(C6F 5)4] (4).
Although the cationic hydride complex [CpCp*HfH]-
[B(C6F5)4] could not be isolated and studied in direct
comparisons with 6 or 7, the cationic hafnium methyl
derivative 4 was extremely active as a catalyst precursor
for both organosilane dehydropolymerization and re-
distribution processes. Thus, addition of excess PhSiH3
(ca. 25 equiv) to bromobenzene-d5 solutions of 4 resulted
in vigorous gas evolution. After 5 min, a 1H NMR
spectrum of the reaction mixture indicated that ap-
proximately half of the PhSiH3 had been converted to
polysilanes and a mixture of Ph2SiH2 (ca. 40% of the
organosilane species present) and small amounts of Ph3-
SiH and SiH4 (<5% each). After 1 h, the relative
amounts of PhSiH3 and Ph2SiH2 had changed only
slightly (ca. 30% and 55%, respectively), but a large
amount of white solid had formed in the NMR tube.
After approximately 3 h, the PhSiH3 was completely
converted to Ph2SiH2 and the insoluble white precipi-
tate. Addition of more PhSiH3 to the mixture resulted
in further effervescence and production of more floc-
culent solid. Once formed, the resulting precipitate was
insoluble in all organic solvents tested, including THF
and DMSO. This insolubility hindered its characteriza-
tion by 29Si NMR spectroscopy or gel permeation chro-
matography. Given the observation of SiH4 in the
reaction mixture, it is likely that this species partici-
pates as a cross-linking agent to produce insoluble
polysilanes.
The concurrent redistribution and dehydropolymer-
ization of organosilanes was also observed upon addition
of [CpCp*HfMe][B(C6F5)4] (4) to a large excess of Hex-
SiH3, MesSiH3, or Ph2SiH2, as evidenced by formation
of large amounts of insoluble products for each reaction.
Interestingly, 4 reacted with more sterically hindered
silanes such as CySiH3 and Ph2SiH2. The polymer
resulting from reaction of 4 and Ph2SiH2 was also
insoluble in common organic solvents, making charac-
terization difficult. Interestingly, this polymer exhibits
a relatively sharp melting point, at 229-231 °C. Reac-
tion of 4 with CySiH3 did not result in precipitation of
a polymer from bromobenzene-d5. Instead, large amounts
of Cy2SiH2 were formed (by 1H NMR spectroscopy). The
hindered silanes Cy2SiH2, Ph3SiH, and Mes2SiH2 did
not react with 4 after 1 week at room temperature (by
1H NMR spectroscopy, bromobenzene-d5).
Con clu d in g Rem a r k s
Previous to this work, it had been shown that neutral,
16-electron, mixed-ring zirconocene hydride or silyl
complexes were among the most reactive catalysts for
the dehydropolymerization of PhSiH3.10,11 Mechanistic
studies have revealed that these neutral catalysts
mediate the polymerization reaction through Si-H bond
activations and silicon-silicon bond formations via four-
centered transition states.11d Attempts to develop more
efficient Si-H bond activations and silane dehydropo-
lymerizations, for the production of high-molecular-
weight polysilanes, prompted studies of more electro-
philic metal centers as catalysts for these reactions. In
this context, some reports in the literature suggest that
cationic group 4 centers are more active catalysts for
the synthesis of high-molecular-weight polysilanes,14b,18
while other studies suggest that cationic centers are less
active in silane dehydropolymerizations.14a
We have addressed this issue by comparing the
activities of neutral and cationic hafnium complexes in
reactions with silanes, in terms of both the molecular
weights of the polysilanes produced and the rates at
which analogous neutral and cationic complexes react.
Polysilane molecular weights indicate that cationic
hafnium-catalyzed polymerizations of PhSiH3 do not
produce longer polysilane chains (in comparison to
analogous neutral catalysts). On the other hand, our
investigations indicate that cationic hafnium centers
react much more rapidly with silanes than comparable
neutral hafnium compounds.
One possible explanation for the lower molecular
weights (despite enhanced rates in σ-bond metathesis
reactions) is that σ-bond metathesis is not the mecha-
nism for polysilane formation. Harrod has suggested
that the mechanism for silane dehydropolymerization,
as catalyzed by cationic group 4 metal centers, involves
silyl radical coupling for Si-Si bond formation.18 Our
investigations of the kinetics of PhSiH3 dehydropoly-
merizations implicate a mechanism for dehydropolym-
erization that involves a metal-mediated condensation
polymerization.30 Also, the general rate expressions for
neutral and cationic hafnium-catalyzed monomer con-
sumption in the silane dehydropolymerization reactions
are identical {kobs[catalyst][PhSiH3]2}. These observa-
tions provide support for σ-bond metathesis as the
mechanism for silane dehydropolymerization, as cata-
lyzed by these cationic hafnium complexes.
Lewis acids, including [Ph3C][B(C6F5)4], are known to
promote organosilane redistributions.18a,32 Thus, the
possibility that trace [Ph3C][B(C6F5)4] contamination
could be responsible for the observed chemistry was
considered. Reaction of 1 with 1 equiv of [Ph3C]-
[B(C6F5)4] (bromobenzene-d5) produced a solution of 4
and Ph3CMe (1:1) in which the trityl reagent could not
1
be detected (by H NMR spectroscopy). To ensure that
no free [Ph3C][B(C6F5)4] was available in solution, 1 was
allowed to react with 0.8-0.99 equiv of [Ph3C][B(C6F5)4]
(in bromobenzene-d5) to produce mixtures of
[CpCp*HfMe][B(C6F5)4] and [(CpCp*HfMe)2(µ-Me)]-
[B(C6F5)4]. These Ph3C+-free mixtures reacted with
(32) Corey, J . Y. J . Am. Chem. Soc. 1975, 97, 3227.

3584 Organometallics, Vol. 22, No. 17, 2003
Sadow and Tilley
removed under reduced pressure to yield 3 (0.606 g, 0.658
mmol, 66.3%). 1H NMR (400 MHz): δ 5.35 (s, 5 H, C5H5), 1.33
(s, 15 H, C5Me5), 0.146 (br, 3 H, BMe), 0.088 (s, 3 H, HfMe).
13C{1H} NMR (100 MHz): δ 121.9 (C5Me5), 113.1 (C5H5), 42.5
(HfMe), 22.0 (BMe), 10.8 (C5Me5). 19F NMR (377 MHz): δ -134
(m, 6 F), -159 (m, 2 F), -165 (m, 6 F). 11B NMR (160 MHz):
δ -17.5 (s). IR (KBr, cm-1): 1645 (m), 1516 (s), 1462 (s), 1385
(m), 1281 (m), 1151 (w), 1097 (s), 1026 (m), 974 (s). Anal. Calcd
for BC35F15H26Hf: C, 45.65; H, 2.85. Found: C, 45.50; H, 2.97.
Mp: 65-68 °C dec.
The lower molecular weight products observed for
highly active (e.g., cationic) catalysts may instead be
attributed to the competitive organosilane redistribution
process, via Si-C bond activation, which interferes with
polysilane chain growth. The activation of Si-C bonds
via four-centered transition states is more difficult than
Si-H bond activation but has been observed in trans-
formations of silanes with highly reactive lanthanide
complexes (Cp*2LnH)2 (Ln ) Sm, Lu, Y).25 Thus, the
organosilane redistribution process further indicates
that the cationic hafnium centers are highly reactive
in σ-bond metathesis steps, more so than neutral
hafnium centers. As is also evident from the σ-bond
metathesis chemistry of f-element complexes, more
reactive metal centers may react with both Si-H bonds
and Si-C bonds. These results suggest that new cata-
lytic bond activation processes may be based on cationic
centers. However, successful strategies for bond activa-
tion catalysis based on σ-bond metathesis reactions
require consideration of selectivity as well as activity.
In Situ Gen er a tion of [Cp Cp *HfMe][B(C6F 5)4] (4). Bro-
mobenzene-d5 solutions of CpCp*HfMe2 (0.016 g, 0.038 mmol)
and [Ph3C][B(C6F5)4] (0.035 g, 0.038 mmol) were mixed at room
1
temperature. After 2 h, H NMR spectroscopy indicated that
the [Ph3C][B(C6F5)4] was consumed and that 4 and Ph3CMe
were formed in a 1:1 ratio. 1H NMR (500 MHz, bromobenzene-
d5): δ 5.54 (s, 5 H, C5H5), 1.73 (s, 15 H, C5Me5), 0.00 (s, 3 H,
HfMe). 13C{1H} NMR (100 MHz, bromobenzene-d5): δ 123.36
(C5Me5), 115.37 (C5H5), 50.76 (HfMe), 11.48 (C5Me5). 19F NMR
(377 MHz, bromobenzene-d5): δ -132 (m, 8 F), -162, (m, 4
F), -166 (m, 8 F). 11B NMR (160 MHz, bromobenzene-d5): δ
-16.85 (s).
[Cp Cp *HfMe(P Me3)][B(C6F 5)4] (5). Cp*CpHfMe2 (0.017
g, 0.041 mmol) and [Ph3C][B(C6F5)4] (0.038 g, 0.041 mmol) were
mixed in bromobenzene-d5. After 2 h, 1.2 equiv of PMe3 was
added, which resulted in an immediate color change from
orange to pale yellow. The bromobenzene-d5 solution was
layered with pentane in an NMR tube, and after 1 day crystals
of 5 formed (0.025 g, 0.022 mmol, 53.6%). 1H NMR (400 MHz,
bromobenzene-d5): δ 5.69 (s, 5 H, C5H5), 1.64 (s, 15 H, C5Me5)
0.84 (d, 9 H, PMe3), -0.47 (s, 3 H, HfMe). 13C{1H} NMR (100
MHz, bromobenzene-d5): δ 121.82 (C5Me5), 113.41 (C5H5),
55.65 (HfMe), 14 (br, PMe3), 11.59 (C5Me5). 19F NMR (377 MHz,
bromobenzene-d5): δ -132.1 (d, 8 F), -162.3 (2, 4 F), -166.1
(t, 8 F). 11B NMR (160 MHz, bromobenzene-d5): δ -16.84. 31P-
{1H} NMR (162 MHz, bromobenzene-d5): 36.70. IR (KBr,
cm-1): 2926 (m), 1644 (s), 1515 (s), 1462 (s), 1383 (m), 1089
Exp er im en ta l Section
Gen er a l Con sid er a tion s. All manipulations were per-
formed under an atmosphere of nitrogen using Schlenk
techniques or a glovebox. Dry, oxygen-free solvents were
employed throughout. Removal of thiophenes from benzene
and toluene was accomplished by washing each with H2SO4
and saturated NaHCO3 and treating with the drying agent
MgSO4. Olefin impurities were removed from pentanes by
treatment with concentrated H2SO4, 0.5 N KMnO4 in 3 M H2-
SO4, saturated NaHCO3, and the drying agent MgSO4. All
solvents were distilled from sodium benzophenone ketyl, with
the exception of benzene-d6, which was purified by vacuum
distillation from Na/K alloy. The compounds CpCp*HfMe2
(1),21a CpCp*Hf(H)Cl (7),10 B(C6F5)3,33 and [Ph3C][B(C6F5)4]33
were prepared according to literature procedures. All hydrosi-
lanes were prepared via reduction of the corresponding chlo-
rosilanes with LiAlH4. Elemental analyses were performed by
the microanalytical laboratory at the University of California,
Berkeley. Infrared spectra were recorded using a Mattson
FTIR instrument at a resolution of 4 cm-1. All NMR spectra
were recorded at room temperature in benzene-d6 unless
otherwise noted, using a Bruker AM-400 spectrometer or a
Bruker DRX-500 spectrometer.
Cp Cp *Hf(OTf)Me (2). Addition of a benzene solution of
AgOTf (0.449 g, 1.75 mmol) to a benzene solution of CpCp*Hf-
Me2 (0.711 g, 1.74 mmol) immediately induced formation of a
black solution. The resulting mixture of 2 and Ag0 was stirred
for 1 h. The Ag0 precipitate was removed by filtration of the
solution, and analytically pure 2 was isolated by evaporation
of the benzene solvent under reduced pressure (0.752 g, 1.39
mmol, 79.7%). 1H NMR (400 MHz): δ 5.793 (s, 5 H, C5H5),
1.588 (s, 15 H, C5Me5), 0.152 (s, 3 H, HfMe). 13C{1H} NMR
(100 MHz): δ 120.5 (C5Me5), 113.7 (C5H5), 38.0 (HfMe), 11.4
(C5Me5). 19F NMR (377 MHz): δ -77.4 (OTf). IR (KBr, cm-1):
2926 (w), 1350 (s), 1239 (m), 1200 (s), 834 (m), 632 (m). Anal.
Calcd for C17H23F3HfO3S: C, 37.61; H, 4.27. Found: C, 37.94;
H, 4.02. Mp: 142.5-143 °C.
(s), 979 (s), 824 (s), 774 (s), 662 (m). Anal. Calcd for BC43F20H32
-
HfP: C, 44.96; H, 2.81. Found: C, 45.06; H, 2.81. Mp: 165-
167 °C dec.
Cp Cp *HfH(µ-H)B(C6F 5)3 (6). A 100 mL Schlenk flask was
charged with 3 (0.781 g, 0.848 mmol), pentane (30 mL), and
PhSiH3 (0.1 g, 0.927 mmol). The flask was sealed, and the
resulting mixture was stirred overnight. The solvent and silane
were removed by filtration, and the resulting white powder
was dried under reduced pressure to yield analytically pure 6
(0.361 g, 0.404 mmol, 47.6%). 1H NMR (400 MHz): δ 12.84 (s,
1 H, HfH), 5.26 (s, 5 H, C5H5), 1.59 (s, 15 H, C5Me5). 13C{1H}
NMR (100 MHz): δ 120 (C5Me5), 110.1 (C5H5), 11.8 (C5Me5).
19F NMR (377 MHz): δ -138 (m, 6 F), -157 (m, 3 F), -164
1
(m, 6 F). 11B NMR (160 MHz): δ -16.91 (d, J BH ) 50 Hz). IR
(KBr, cm-1): 2362 (w), 1950 (m), 1645 (m), 1516 (s), 1467 (s),
1381 (m), 1281 (m), 1111 (s), 968 (s), 835 (m). Anal. Calcd for
BC33F15H22Hf: C, 44.39; H, 2.48. Found: C, 44.06; H, 2.34.
Mp: 86-89 °C dec.
Cp Cp *Hf(P h )Me. A 250 mL Schlenk flask was charged
with CpCp*Hf(OTf)Me (0.752 g, 1.38 mmol), PhLi (0.142 g,
1.69 mmol), and benzene (100 mL). The flask was sealed, and
the resulting mixture was stirred overnight. All volatile
materials were removed under reduced pressure, and the
resulting solid was extracted with pentane (3 × 30 mL). The
combined pentane extracts were concentrated to ca. 20 mL and
cooled to -78 °C. The resulting white crystals were collected
and dried under vacuum to give a yield of 58% (0.377 g, 0.798
Cp Cp *HfMe(µ-Me)B(C6F 5)3 (3). A 100 mL Schlenk flask
was charged with CpCp*HfMe2 (0.406 g, 0.993 mmol), B(C6F5)3
(0.509 g, 0.993 mmol), and pentane (ca. 35 mL). The resulting
suspension was stirred vigorously for 3 h. The pentane was
removed via filtration, and a white, powdery product was
washed with pentane (3 × 25 mL). Residual solvent was
1
mmol). H NMR (400 MHz): δ 7.25 (t, 2 H, Ph), 7.05 (t, 1H,
Ph), 6.95 (d, 2 H, Ph), 5.71 (s, 5 H, C5H5), 1.62 (s, 15 H, C5-
Me5), -0.05 (s, 3 H, HfMe). 13C{1H} NMR (100 MHz): δ 199.08
(Ph), 137.16 (Ph), 127.43 (Ph), 125.12 (Ph), 118.09 (C5Me5),
112.15 (C5H5), 42.69 (HfMe), 12.02 (C5Me5). IR (KBr, cm-1):
3046 (m), 2952 (s), 2922 (s), 2854 (s), 1455 (m), 1378 (m), 1143
(33) (a) Massey, A. G.; Park, A. J . J . Organomet. Chem. 1964, 2,
245. (b) Chien, J . C. W.; Tsai, W.-M.; Tausch, M. D. J . Am. Chem. Soc.
1991, 113, 8570.
(m), 1057 (m), 1017 (m), 807 (s), 705 (s). Anal. Calcd for C22H28
-

Cationic Hafnocene Complexes
Organometallics, Vol. 22, No. 17, 2003 3585
Hf: C, 56.37; H, 5.99. Found: C, 56.37; H, 5.91. Mp: 129.5-
130.5 °C dec.
ethylene glycol standard and monitored with a thermocouple.
Single-scan spectra were acquired automatically at preset time
intervals. The peaks were integrated relative to cyclooctane
as an internal standard. Rate constants were obtained by
nonweighted linear least-squares fit of the integrated second-
order rate law 1/[C] ) 1/C0 + kobst.
Cp Cp *HfP h (µ-Me)B(C6F 5)3. This complex was generated
in situ by reaction of CpCp*Hf(Ph)Me with 1 equiv of B(C6F5)3
in benzene-d6 and characterized in solution. 1H NMR (500
MHz): δ 7.15 (t, 2 H, Ph), 6.63 (t, 1 H, Ph), 6.63 (d, 2 H, Ph),
5.52 (s, 5 H, C5H5), 1.33 (s, 15 H, C5Me5), 0.712 (br, 3 H, MeB-
(C5F5)3). 13C{1H} NMR (100 MHz): δ 193.35 (Ph), 150.15
(C6F5), 148.20 (C6F5), 139.02 (C6F5), 136.98 (C6F5), 136.39 (Ph),
127.00 (Ph), 124.67 (C5Me5), 115.32 (C5H5), 20.50 (MeB(C6F5)3,
11.88 (C5Me5). 11B NMR (160 MHz): δ -13.75.
Ack n ow led gm en t is made to the National Science
Foundation (NSF) and Department of Energy (DOE) for
their generous support of this work and to Dr. J ack Kyte
(UC San Diego) for valuable discussions.
1
Kin etic Mea su r em en ts. Reactions were monitored by H
NMR spectroscopy with a Bruker DRX500 spectrometer using
5 mm Wilmad NMR tubes, equipped with J . Young Teflon
screw caps. The samples were frozen in liquid N2 immediately
after preparation and defrosted just before being placed in the
probe, which was preheated to the required temperature.
When CpCp*HfHCl was used as the catalyst, the headspace
of the NMR tube was evacuated while the benzene solution
was frozen. The probe temperature was calibrated using an
Su p p or tin g In for m a tion Ava ila ble: Figures giving rep-
resentative kinetics data for the reactions of 6 and 7 with
PhSiH3 and 29Si NMR spectra for poly(phenylsilanes). This
material is available free of charge via the Internet at
http://pubs.acs.org.
OM0210107