Dehydrogenative oligomerization of PhSiH3 catalyzed by (1-Me-Indenyl)Ni(PR3)(Me)

Article abstract of DOI:10.1021/om010757e

The complexes (1-Me-Ind)Ni(PR₃)Me (Ind = indenyl; R = Ph, 1; Cy, 2; Me, 3) can act as single-component precatalysts for the dehydrogenative oligomerization of PhSiH₃ to (PhSiH)_n with M_w up to 1.6 × 10₃ in the case of 3. The catalytic Si-Si bond formation is initiated by a second-order, concerted reaction between the Ni-Me precursor and PhSiH₃ which releases CH₄. Kinetic studies of this Si-H bond activation step gave the following activation parameters for the reaction of 3 with PhSiH₃: ΔH^? = 10.7 ± 0.7 kcal·mol^-1; ΔS? = -42 = 2 eu; k_H/k_n (for reaction with PhSiD₃ at 313 K) = 9.8 ± 0.5. The Ni-SiPhH₂ species presumed to form in this reaction was not detected directly, but the presence of 1-SiPhH₂-3-Me-Ind in the reaction mixtures, coupled with the fact that Ni(PMe₃)₄ is an active precatalyst for the oligomerization of PhSiH₃, suggests that the silyl intermediate undergoes a reductive elimination to generate Ni(0) species which might be involved in the catalysis.

Full text of DOI:10.1021/om010757e

Organometallics 2002, 21, 401-408
401
Deh yd r ogen a tive Oligom er iza tion of P h SiH3 Ca ta lyzed
by (1-Me-In d en yl)Ni(P R3)(Me)
Fre´de´ric-Georges Fontaine and Davit Zargarian*
De´partement de Chimie, Universite´ de Montre´al, Montre´al (Que´bec), Canada, H3C 3J 7
Received August 17, 2001
The complexes (1-Me-Ind)Ni(PR3)Me (Ind ) indenyl; R ) Ph, 1; Cy, 2; Me, 3) can act as
single-component precatalysts for the dehydrogenative oligomerization of PhSiH3 to (PhSiH)n
with Mw up to 1.6 × 103 in the case of 3. The catalytic Si-Si bond formation is initiated by
a second-order, concerted reaction between the Ni-Me precursor and PhSiH3 which releases
CH4. Kinetic studies of this Si-H bond activation step gave the following activation
parameters for the reaction of 3 with PhSiH3: ∆Hq ) 10.7 ( 0.7 kcal‚mol-1; ∆Sq ) -42 (
2 eu; kH/kD (for reaction with PhSiD3 at 313 K) ) 9.8 ( 0.5. The Ni-SiPhH2 species presumed
to form in this reaction was not detected directly, but the presence of 1-SiPhH2-3-Me-Ind in
the reaction mixtures, coupled with the fact that Ni(PMe3)4 is an active precatalyst for the
oligomerization of PhSiH3, suggests that the silyl intermediate undergoes a reductive
elimination to generate Ni(0) species which might be involved in the catalysis.
In tr od u ction
via a cationic mechanism involving an intermediate of
the type [(1-Me-Ind)Ni(PPh3)]+; later studies indicated,
however, that the catalysis involves neutral Ni-Me
derivatives formed in situ by the methylation of the Ni-
Cl precursor with MAO. This finding prompted us to
examine the reaction of PhSiH3 with a number of
preformed Ni-Me derivatives in an attempt to better
understand their role in the Ni/MAO-catalyzed polym-
erization of PhSiH3. Herein we report the results of our
mechanistic studies on the dehydrocoupling of PhSiH3
by the complexes (1-Me-Ind)Ni(PR3)Me (R ) Ph, 1; Cy,
2; Me, 3).
The search for new synthetic routes to polysilanes (R-
Si-R′)n is driven by the potential of these materials in
microlithography, photoconduction, electrolumines-
cence, the fabrication of optical devices, and as precur-
sors to silicon carbide.1 The discovery2 and development3
of effective, transition-metal-based routes for the dehy-
drogenative coupling of silanes has provided an inter-
esting alternative to the traditional Wurtz coupling of
halosilanes for the synthesis of these catenated poly-
mers. The most promising catalysts reported to date are
some derivatives of group 4 metallocenes,3 but recent
reports have shown that a number of other systems can
be effective for silane polymerization. For instance, Pt-
(COD)2 can polymerize4 secondary silanes such as Et2-
SiH2 to Si-containing polymers of relatively high Mw,
while M(CO)6 (M ) Cr, Mo, W) can dehydrocouple the
tertiary Si-H bonds in poly(methylsilene), (-Si(H)(Me)-
CH2-)n, to produce cross-linked preceramic materials.5
In this context, we have reported that PhSiH3 can be
polymerized to (PhSiH)n having Mw in the range of (2-
7) × 103 by a system consisting of the complex (1-Me-
Ind)Ni(PPh3)Cl (Ind ) indenyl) and methylaluminoxane
(MAO).6 This system was initially thought to operate
Exp er im en ta l Section
Gen er a l Com m en ts. All manipulations were performed
under an inert atmosphere of N2 or argon using standard
Schlenk techniques and a drybox. Dry, oxygen-free solvents
were employed throughout. The preparation and complete
characterization of complex 17 and the Ni-Cl precursors8 to 2
and 3 have been reported previously. PhSiH3 and all other
reagents used in the experiments were obtained from com-
mercial sources and used as received. The elemental analyses
were performed by the Laboratoire d’analyse e´le´mentaire
(Universite´ de Montre´al). The NMR spectra were recorded
1
using the following spectrometers: Bruker DMX600 (2D H-
29Si), Bruker AMX400 (1H at 400 MHz, 13C{1H} at 100.56 MHz,
and 31P{1H} at 161.92 MHz), and Bruker AV300 (1H at 300
MHz for TOCSY).
(1) (a) West, R. J . Organomet. Chem. 1986, 300, 327. (b) Miller, R.
D.; Michl, J . Chem. Rev. 1989, 89, 1359.
(2) Aitken, C.; Harrod, J . F.; Samuel, E. J . Organomet. Chem. 1985,
279, C11.
(3) For a few leading references see: (a) Tilley, T. D. Acc. Chem.
Res. 1993, 26, 22. (b) Harrod, J . F.; Mu, Y.; Samuel, E. Polyhedron
1991, 10, 1239. (c) Shaltout, R. M.; Corey, J . Y. Organometallics 1996,
15, 2866. (d) Campbell, W. H.; Hilty, T. K.; Yurga, L. Organometallics
1989, 8, 2615. (e) Banovetz, J . P.; Stein, K. M.; Waymouth, R. M.
Organometallics 1991, 10, 3430. (f) Hengge, E.; Gspaltl P.; Pinter, E.
J . Organomet. Chem. 1996, 521, 145. (g) Choi, N.; Onozawa, S.;
Sakakura, T.; Tanaka, M. Organometallics 1997, 16, 2765.
(4) Chaunan, B. P. S.; Shimizu, T.; Tanaka, M. Chem. Lett. 1997,
785.
(1-Me-In d )Ni(P Cy3)Me (2). MeLi (1.94 mL of a 1.4 M
solution in Et2O, 2.72 mmol) was added dropwise to a solution
of (1-Me-Ind)Ni(PCy3)Cl (912 mg, 1.81 mmol) in Et2O (100 mL)
and stirred for 2 h. Deoxygenated water (0.3 mL) was then
added and stirred for 5 min. Drying over MgSO4 followed by
filtration and evaporation gave a dark powder (655 mg, crude
yield 75%), which was recrystallized from cold Et2O to give
red crystals suitable for X-ray diffraction studies. 1H NMR
(5) Yang, S. Y.; Park, J . M.; Woo, H. G.; Kim, W. G.; Kim, I. S.; Kim,
D. P.; Hwang, T.S. Bull. Korean. Chem. Soc. 1997, 12, 1264.
(6) Fontaine, F.-G.; Kadkhodazadeh, T.; Zargarian, D. J . Chem. Soc.,
Chem. Commun. 1998, 1253.
(7) Huber, T. A.; Bayrakdarian, M.; Dion, S.; Dubuc, I.; Be´langer-
Garie´py, F.; Zargarian, D. Organometallics 1997, 16, 5811.
(8) Fontaine, F.-G.; Dubois, M.-A.; Zargarian, D. Organometallics
2001, 20, 5156.
10.1021/om010757e CCC: $22.00 © 2002 American Chemical Society
Publication on Web 12/21/2001

402 Organometallics, Vol. 21, No. 2, 2002
Fontaine and Zargarian
Ta ble 1. Cr ysta l Da ta , Da ta Collection , a n d
Str u ctu r e Refin em en t of 2 a n d 3
(C6D6): δ 7.20-7.16 and 7.01-6.95 (m, aromatic protons of
3
4
Ind), 6.26 (d, J H-H ) 3.0, H2), 4.73 (b, H3), 1.90 (d, J P-H
)
3
3.5, 1-CH3-Ind), 1.80-1.00 (m, PCy3), -0.86 (d, J P-H ) 4.5,
2
3
Ni-CH3). 13C{1H} NMR (C6D6): 124.1 (s, C3a or C7a), 123.5 (s,
formula
mol wt
C29H45Ni P
483.330
clear red
plate
C14H21NiP
278.988
dark red
block
C3a or C7a), 122.3 (s, C5 or C6), 121.4 (s, C5 or C6), 118.0 (s,
2
C4 or C7), 116.1 (s, C4 or C7), 102.6 (s, C2), 87.0 (d, J P-C
)
cryst color
cryst habit
cryst dimens, mm
symmetry
space group
a, Å
1
11.2, C1), 69.3 (s, C3), 35.8 (d, J P-C ) 20.1, P-CH), 30.1 (d,
2J P-C ) 24.8, PCH(CH2CH2)2CH2), 27.9 (d, 3J P-C ) 11.2, PCH-
0.08 × 0.24 × 0.20 0.23 × 0.19 × 0.16
3
(CH2CH2)2CH2), 26.8 (s, PCH(CH2CH2)2CH2), 10.7 (d, J P-C
)
monoclinic
P21/n
orthorhombic
P212121
7.685(3)
11.990(5)
15.230(5)
90
1.6, CH3-Ind), -22.7 (d, J P-C ) 23.3, Ni-CH3). 31P{1H} NMR
(C6D6): 48.29 (s). Anal. Calcd for C29H45Ni1P1: C, 72.06; H, 9.38.
Found: C, 71.64; H, 9.73.
2
13.052(7)
14.345(6)
14.620(11)
90
106.98(5)
90
2618(3)
4
1.2263
b, Å
c, Å
(1-Me-In d )Ni(P Me3)Me (3). MeLi (1.07 mL of a 1.4 M
solution in Et2O, 1.5 mmol) was added dropwise to a solution
of (1-Me-Ind)Ni(PMe3)Cl (300 mg, 1.00 mmol) in Et2O (30 mL).
The solution was stirred for 30 min, and then deoxygenated
water (1 mL) was added and stirred for 5 min. Drying over
MgSO4 followed by filtration and evaporation gave a dark red
powder (128 mg, 53% yield). Recrystallization from cold
hexanes gave suitable crystals (dark red) for X-ray diffraction
R, deg
â, deg
90
90
γ, deg
volume, Å3
Z
1403.3(9)
4
1.3205
Nonius CAD-4
220(2)
D(calcd), g cm-1
diffractometer
temp, K
Nonius CAD-4
220(2)
1
3
λ (Cu KR), Å
µ, mm-1
scan type
1.54056
1.723
ω/2θ scan
69.84
1.54056
2.825
ω/2θ scan
69.87
studies. H NMR (C6D6): δ 7.16 (br d, J H-H ) 8.0, H4 or H7),
3
4
7.06 (pseudo td, J H-H ) 7.8, J H-H ) 1.6, H5 or H6), 7.00
3
4
(pseudo td, J H-H ) 7.7, J H-H ) 1.2, H5 or H6), 7.16 (br d,
θ
max, deg
3J H-H ) 7.6, H4 or H7), 6.07 (d, 3J H-H ) 2.0, H2), 4.44 (b, H3),
h,k,l range
-15 e h e 15
-17 e k e 17
0 e l e 17
4916
-9 e h e 9
-14 e k e 14
-18 e l e 18
2668
4
2
1.95 (d, J P-H ) 3.7, CH3-Ind), 0.63 (d, J P-H ) 9.1, P-CH3),
-0.78 (d, 3J P-H ) 6.2, Ni-CH3). 13C{1H} NMR (C6D6): 121.9 (s,
C5 or C6), 121.8 (s, C5 or C6), 121.5 (s, C3a or C7a), 120.6 (s,
C3a or C7a), 116.2 (s, C4 or C7), 116.1 (s, C4 or C7), 101.7 (s,
no. of reflns used
(I > 2σ(I))
abs corr
2
1
integration
0.89 and 0.69
integration
0.72 and 0.55
0.0346, 0.0754
0.915
C2), 88.3 (d, J P-C ) 12.0, C1), 69.4 (s, C3), 16.1 (d, J P-C
)
)
T (max, min)
3
2
27.3, P-CH3), 10.8 (d, J P-C ) 1.6, CH3-Ind), -20.7 (d, J P-C
R[F2>2σ(F2)], wR(F2) 0.0558, 0.1365
24.1, Ni-CH3). 31P{1H} NMR (C6D6): -3.72 (s). Anal. Calcd for
GOF
0.937
C
14H21Ni1P1: C, 60.27; H, 7.59. Found: C, 60.14; H, 7.72.
Cr ysta l Str u ctu r e Deter m in a tion s. Suitable crystals of
Ta ble 2. Selected Bon d Dista n ces (Å) a n d An gles
2 and 3 were attached to glass fibers and transferred rapidly
and under a cold stream of nitrogen to an Enraf-Nonius CAD-4
diffractometer equipped with a low-temperature gas stream
cryostat for data collection at 220(2) K. The data were collected
with graphite-monochromated Cu KR radiation; the refine-
ment of the cell parameters was done using the CAD-4
software9 on 25 reflections, while NRC-2 and NRC-2A were
used for the data reduction.10 Both structures were solved by
direct methods using SHELXS96,11 and the refinements were
done on F2 by full-matrix least squares. All non-hydrogen
atoms were refined anisotropically, while the hydrogens
(isotropic) were constrained to the parent atom using a riding
model. Crystal data and experimental details for both com-
pounds are listed in Table 1 and selected bond distances and
angles are listed in Table 2. The complete crystallographic
reports are included in the Supporting Information.
Ca ta lytic Stu d ies. The catalytic experiments were carried
out by stirring 3-5 mg of each precatalyst in ca. 200 mg of
PhSiH3 (ca. 200 equiv) under an inert atmosphere. The
resulting oil or solid product was dissolved in THF, passed
through a microfilter, and analyzed by a Waters GPC system
equipped with a refractive index detector calibrated with
polystyrene standards in THF. (Note: Care should be taken
not to expose these samples to air prior to analysis.)12 As in
(d eg) for 1-3
1
2
3
Ni-P
2.1213(13)
1.991(3)
2.089(4)
2.072(4)
2.082(4)
2.275(4)
2.273(4)
1.407(5)
1.404(5)
1.430(5)
1.424(5)
1.438(3)
0.19
2.1710(13)
1.965(4)
2.104(4)
2.071(4)
2.110(4)
2.321(3)
2.320(3)
1.410(6)
1.411(5)
1.457(5)
1.421(5)
1.460(5)
0.21
2.1239(11)
1.955(3)
2.097(3)
2.076(3)
2.095(3)
2.298(3)
2.288(3)
1.409(5)
1.410(4)
1.445(5)
1.435(4)
1.451(5)
0.20
93.30(10)
163.19(12)
103.25(9)
97.40(13)
168.34(10)
66.37(13)
Ni-C(9)
Ni-C(1)
Ni-C(2)
Ni-C(3)
Ni-C(3A)
Ni-C(7A)
C(1)-C(2)
C(2)-C(3)
C(3)-C(3A)
C(3A)-C(7A)
C(1)-C(7A)
∆M-Ca
P-Ni-C(9)
C(3)-Ni-C(9)
C(3)-Ni-P
C(1)-Ni-C(9)
C(1)-Ni-P
C(1)-Ni-C(3)
99.51(10)
159.84(16)
104.04(13)
94.26(16)
170.22(13)
66.19(16)
96.05(11)
158.05(15)
105.76(11)
92.05(16)
171.31(11)
66.04(15)
a
See ref 17 for the definition of ∆M-C and literature values
for other indenyl complexes.
the case of most (PhSiH)n samples obtained from transition-
metal-catalyzed reactions, the products of these catalytic
reactions displayed broad 1H NMR signals characteristic of
linear (4.4-4.8 ppm) and cyclic (5.2-6.0 ppm) oligomers.13 The
gel permeation chromatograms of these samples also display
two distinct, and often overlapping, peaks that represent the
linear and cyclic components of the mixture. The ratios of the
two Si-H signals in the 1H NMR spectra are similar to the
ratio of the peak areas in the GPC runs. It should be noted
that according to Harrod et al. the molecular weights of
(PhSiH)n determined from the GPC runs versus polystyrene
standards are often underestimated.14 The values presented
(9) CAD-4 Software, version 5.0; Enraf-Nonius: Delft, The Neth-
erlands, 1989.
(10) Gabe, E. J .; Le Page, Y.; Charlant, J .-P.; Lee, F. L.; White, P.
S. J . Appl. Crystallogr. 1989, 22, 384.
(11) Sheldrick, G. M. SHELXL96, Program for the Solution of
Crystal Structures; University of Gottingen: Germany, 1990.
(12) (a) According to a recent study (ref 12b), samples of (PhSiH)n
are only partially oxidized even after being exposed to air for 12 h in
refluxing xylene. We have found, however, that the samples obtained
from our Ni-catalyzed reactions are very air-sensitive even at ambient
temperature. It is not clear whether this apparently increased air-
sensitivity of our polysilane samples is due to the presence of trace
amounts of Ni particles left over from the catalytic reactions. (b)
Chatgilialoglu, C.; Guerrini, A.; Lucarini, M.; Pedulli, G. F.; Carrozza,
P.; Da Riot, G.; Borzatta, V.; Lucchini, V. Organometallics 1998, 17,
2169.
(13) (a) Woo, H.-G.; Walzer, J . F.; Tilley, T. D. J . Am. Chem. Soc.
1992, 114, 7047. (b) Gauvin, F. Ph.D. Thesis, McGill University, 1992.

Dehydrogenative Oligomerization of PhSiH3
Organometallics, Vol. 21, No. 2, 2002 403
in this report have not been corrected for such a systematic
error.
Kin etic Stu d ies. The kinetic studies were carried out as
follows. A Wilmad 528-TR-7 screwcap NMR tube containing
0.7 mL of a 0.02 M C6D6 solution of the precatalysts was heated
to the desired temperature inside the NMR instrument’s probe
prior to the injection (via the septum) of PhSiH3 (50, 70, or 90
equiv) and a precisely measured amount of toluene as internal
standard. The progress of each reaction was monitored by
following the disappearance of the Ni-CH3 and the Ind-CH3
1
signals of the precatalyst by H NMR spectroscopy. It should
be noted that we used the data obtained up to one half-life
because the data obtained after prolonged reaction times
caused deviations from linearity. The nonlinear kinetic be-
havior is attributed to the secondary reactions of PhSiH3 with
the intermediates arising from the initial reaction with the
precatalysts (vide infra).
Resu lts
F igu r e 1. ORTEP diagram for complex 2 showing 30%
probability thermal ellipsoids and the atom-numbering
scheme.
Syn th esis a n d Ch a r a cter iza tion of th e Com -
p lexes. Complexes 2 and 3 were obtained in 50-75%
yields by reacting the Ni-Cl precursors with MeLi in
analogy with the synthesis of 1.7 In solution, these
complexes are air sensitive but thermally stable; in the
solid state they are stable to air for a few days. The
spectroscopic characterization of these complexes was
quite straightforward: the 31P{1H} NMR spectra showed
the expected7,8,15 singlet resonances (48.3 and -3.7 ppm
for 2 and 3, respectively), while the 1H and 13C{1H}
NMR spectra contained the upfield signals character-
3
istic of the Ni-Me moieties (1H: ca. -0.8 ppm, J P-H
)
4.5 for 2 and 6.2 for 3; 13C{1H}: -22.7 and -20.7 ppm,
2J P-C ) 23.3 and 24.1 Hz for 2 and 3, respectively).
1
Significantly, the H NMR signals for Ind-CH3 and H3
protons appeared downfield of the corresponding signals
in the Ni-Cl analogues, thus confirming the increased
η5 hapticity of the Ind ligand upon methylation (vide
infra).
The results of the X-ray studies (ORTEP diagrams
in Figures 1 and 2) showed that the overall geometry
around the Ni atom in complexes 2 and 3 is ap-
proximately square planar, with the largest distortion
arising from the small C1-Ni-C3 angle of ca. 66°. The
Ni-CH3 distance is similar in both complexes (ca. 1.956-
(4) Å) and somewhat shorter than the corresponding
distance in complex 1 (1.991(3) Å); on the other hand,
the Ni-P distance is longer in complex 2 (2.1710(13)
Å) relative to both 3 (2.1239(11) Å) and 1 (2.1213(13)
Å) presumably because of the much larger cone angle
of PCy3. The large steric bulk of PCy3 is also reflected
in the slightly longer Ni-C(Ind) bonds.
As observed in the structures of the previously
reported examples of this family of complexes,7,8,15 the
1-Me-Ind ligand in both 2 and 3 is bound to the Ni
center primarily through the C1, C2, and C3 atoms (ca.
2.07-2.10 Å), whereas the Ni-C3a and Ni-C7a dis-
tances are significantly longer (ca. 2.29 and 2.32 Å). This
F igu r e 2. ORTEP diagram for complex 3 showing 30%
probability thermal ellipsoids and the atom-numbering
scheme.
kind of η5 T η3 “slippage” can be attributed to the
tendency of a d8 center to avoid forming an 18-electron
complex.16 The parameter ∆M-C ) {M-Cav (for C3a
and C7a) - M-Cav (for C1 and C3)} is often used to
measure the extent to which Ind ligands have slipped
away from the idealized η5 hapticity wherein ∆M-C
should be nearly zero.17 The ∆Ni-C values for 2 (0.21
Å) and 3 (0.20 Å) are slightly larger than those observed
for complex 1 (0.19 Å), probably reflecting the stronger
nucleophilicity of tri(alkyl) phosphines. Comparison of
the bonding parameters in these Ni-Me complexes and
their Ni-Cl analogues7,8 reveals that (a) the degree of
slippage is lower in the Ni-Me derivatives (∆Ni-C )
0.25-0.30 for Ni-Cl derivatives), and (b) the coordina-
tion of the 1-Me-Ind to the Ni center is symmetrical in
(14) Dioumaev V. K.; Harrod J . F. Organometallics 1994, 13, 1548.
(15) (a) Huber, T. A.; Be´langer-Garie´py, F.; Zargarian, D. Organo-
metallics 1995, 14, 4997. (b) Bayrakdarian, M.; Davis, M. J .; Reber,
C.; Zargarian, D. Can. J . Chem. 1996, 74, 2194. (c) Vollmerhaus, R.;
Be´langer-Garie´py, F.; Zargarian, D. Organometallics 1997, 16, 4762.
(d) Dubuc, I.; Dubois, M.-A.; Be´langer-Garie´py, F.; Zargarian, D.
Organometallics 1999, 18, 30. (e) Wang, R.; Be´langer-Garie´py, F.;
Zargarian, D. Organometallics 1999, 18, 5548. (f) Groux, L. F.;
Be´langer-Garie´py, F.; Zargarian, D.; Vollmerhaus, R. Organometallics
2000, 19, 1507.
(16) For a detailed discussion of hapticity issues in the corresponding
CpNi complexes see: Holland, P. L.; Smith, M. E.; Andersen, R. A.;
Bergman, R. G. J . Am. Chem. Soc. 1997, 119, 12815.
(17) For a discussion of the correlations between the solid-state and
solution hapticities for indenyl complexes, on one hand, and the ∆M-C
values and the 13C NMR data, on the other, see: (a) Baker, R. T.; Tulip,
T. H. Organometallics 1986, 5, 839. (b) Westcott, S. A.; Kakkar, A. K.;
Stringer, G.; Taylor, N. J .; Marder, T. B. J . Organomet. Chem. 1990,
394, 777.

404 Organometallics, Vol. 21, No. 2, 2002
Fontaine and Zargarian
Ta ble 3. Deh yd r op olym er iza tion of P h SiH3 Usin g (1-Me-In d )Ni(P R3)Me
total mixtureb linear portionb
run
cat.
solvent
time (days)
T (°C)
% conva
Mw
Mn
Mw/Mn
Mw
Mn
Mw/Mn
% linc
1
2
3
4
5
6
7
8
9
1
2
3
3
3
3
3
3
4
4
neat
neat
neat
toluened
toluened
toluened
toluened
toluened
toluened
toluened
6
6
6
6
6
25
25
25
25
-35
55
-35
-35
25
83
12
530
372
950
715
769
1011
1207
1638
1780
1586
426
320
750
623
668
1.2
1.2
1.3
1.1
1.2
1.4
1.2
1.3
1.2
1.1
100
100
62
74
86
62
74
64
38
>99
>99
>99
>99
>99
>99
>99
>99
702
583
676
477
477
548
529
781
816
843
811
1.5
1.2
1.2
1.5
1.4
1.5
1.3
1.2
6
797
704
15
30
6
1081
1243
1095
993
1013
1281
1521
1399
10
15
25
38
a
Conversion of PhSiH3, determined by the relative intensities of the 1H NMR signals for the Si-H peaks of the oligomer and the
b
starting PhSiH3. Molecular weights of the total oligomeric mixture and the linear component, determined by GPC against polystyrene
standards. Dividing the Mw and Mn values by 100 gives a rough estimate of the number of PhSiH units in the oligomers. c Percent content
of the linear portion as determined by 1H NMR and GPC. d5 M toluene solution.
the Ni-Me complexes (Ni-C(1) ≈ Ni-C(3)) but nonsym-
metrical in the Ni-Cl counterparts (Ni-C(1) > Ni-
C(3)), reflecting the trans influence order of PR3 ∼ Me
> Cl.
To sum up, the spectroscopic and solid-state data
indicate that the structures of complexes 1-3 are very
similar, the main differences being the slightly longer
Ni-P and Ni-Ind distances in complex 2, which are
likely due to the steric bulk of PCy3. Despite their subtle
structural differences, however, these complexes possess
very different reactivities toward PhSiH3, as discussed
in the following sections.
Ca ta lytic Resu lts. Addition of catalytic quantities
of the Ni-Me complexes to solutions of PhSiH3 caused
an evolution of gas and gave various oligomeric prod-
ucts. The evolved gas was identified as H2 on the basis
of the Raman spectrum of the head gas, which displayed
a strong band at 4156 cm-1 corresponding to the
literature value for ν(H-H).18 The consumption of
PhSiH3 and the emergence of the oligomers was con-
F igu r e 3. Si-H region of the 1H NMR spectra for the
products of the catalytic runs at -35 °C (a: Table 3, run
5), 25 °C (b: Table 3, run 4), and 55 °C (c: Table 3, run 6).
1
veniently monitored from the H NMR spectra of these
following results. Somewhat higher proportions of the
linear oligomers can be obtained by running the cataly-
sis in concentrated toluene solutions (ca. 5 M) instead
of neat PhSiH3, but the molecular weights of the
resultant oligomers are slightly lower (compare runs 3
and 4). Decreasing the reaction temperature to -35 °C
did not affect the chain length significantly, but gave
more of the linear component (compare runs 4 and 5),
whereas increasing the reaction temperature to 55 °C
gave longer linear chains and a somewhat diminished
linear:cyclic ratio (compare runs 4 and 6; Figure 3).
Finally, running the catalysis in concentrated and cold
toluene solutions for longer times resulted in much
longer linear chains but smaller linear:cyclic ratios
(compare runs 7 and 8 to run 5).
reactions in C6D6 (PhSiH3: s, 4.21 ppm; dimer: s, 4.49
ppm; linear oligomers: br, 4.4-4.8 ppm; cyclic (Ph-
SiH)n: br, 5.2-6.0 ppm).13 Studying these reactions
under various conditions showed that their rates and
final products depend on the precatalyst used and the
monomer concentration. Thus, reacting 1 or 2 with 50
equiv of PhSiH3 (1 M in C6D6) led to the slow evolution
of H2 and a gradual formation of the dimer (PhH2Si)2
in <5% yields. In contrast, the reaction of 3 with PhSiH3
under the same conditions resulted in a vigorous ef-
fervescence of H2 and the formation of linear (PhH2Si)n
oligomers (conversion ca. 95%).
When the catalysis with 1 or 2 was carried out over
6 days in neat PhSiH3 (ca. 200 equiv), we obtained linear
(PhSiH)n with n ) 3-6 (Table 3, runs 1 and 2). The
corresponding reaction with complex 3 was much more
rapid, giving a virtually quantitative conversion of
PhSiH3 over a few hours to a viscous material that
solidified completely over ca. 2-3 days. Analysis of this
solid by NMR and GPC showed that it consisted of an
oligomeric mixture with a 38:62 ratio of cyclic (n ) 5-7)
and linear (n ) 7-10) components (run 3).19
Summing up, the oligosilanes obtained from the
reaction of 3 display narrow polydispersities (Mw/Mn ca.
1.1-1.5) but lower molecular weights in comparison to
samples obtained from some of the best early metal
systems such as (Cp)(Cp*)ZrCl2/n-BuLi/B(C6F5)3 (Mw
≈
(19) To determine whether the mixtures of the oligomeric products
contained any poly(hydrophenylsiloxane), which has 1H NMR chemical
shifts similar to those of (PhSiH)n (ca. 5 ppm in C6D6), we recorded
the IR spectra of the products. The absence of strong vibrations at
2170 and 1100 cm-1 characteristic of (OSiPhH)n eliminated this
possibility. We have also confirmed (by HMQC 29Si-1H) that the
1H NMR signals between 4.4 and 6.0 ppm are correlated to the 29Si
NMR resonances in the region of -55 to -65 ppm, which is typical of
(PhSiH)n (ref 13).
The influence of concentration and temperature on
the catalysis by complex 3 was studied further with the
(18) Herzberg, G. Molecular spectra and molecular structure of
diatomic molecules; V. I. D. Van Norstand Company Inc.: Toronto,
1950.

Dehydrogenative Oligomerization of PhSiH3
Organometallics, Vol. 21, No. 2, 2002 405
3000-14000, Mw/Mn > 2 even after multiple fractional
precipitations).20 Thus, our results show that the Ni
complex 3 and its analogues have the potential to
produce high molecular weight polysilanes in contrast
to the majority of the previously reported late metal
systems, which produce only short oligomers (mainly
dimers and trimers).21
Mech a n ist ic St u d ies. A number of interesting
mechanistic details on the dehydrogenative Si-Si bond
formation catalyzed by our indenyl nickel complexes
have emerged from the following experiments.
(1) Analysis of the gas evolved during the early stages
of the reactions of the Ni-Me complexes with PhSiH3
confirmed the presence of CH4,22 while no trace of
PhSiH2Me was found in the 1H NMR spectra. This
implies the initial formation of a Ni-SiPhH2 (as opposed
to a Ni-H) intermediate (eq 1).
F igu r e 4. Eyring plot for the reaction of 3 with PhSiH3.
The K2 values are (× 10-4 M-1 s-1): 0.474 ( 0.084 at 295
K; 1.68 ( 0.16 at 313 K; 2.85 ( 0.23 at 325 K; and 5.88 (
0.33 at 339 K. The activation parameters: ∆Hq ) 10.7 (
0.7 kcal/mol; ∆Sq ) -42 ( 2 eu.
(2) A number of observations have indicated that the
formation of oligosilanes begins after only a small
portion of the precursor Ni-Me complexes has reacted.
For example, the 1H NMR spectra recorded 30 min after
the start of the reaction of complex 3 with 200 equiv of
PhSiH3 showed that some dimer and trimer had already
formed, while the only discernible signal in the 31P{1H}
NMR spectrum was that of the starting material. For
the corresponding reaction with complex 1, we observed
the formation of dimer, while the main P-containing
species was the starting material. These results imply
that the rate of the initiation step (i.e., the reaction of
the Ni-Me precursors with PhSiH3) is similar to or
slower than the subsequent Si-Si bond forming reac-
tions.
consistent with the coordination of the Si-H bond
to the Ni center23 prior to the activation step (vide
infra).
(4) We have obtained a kH/kD ratio of 9.8 ( 0.5 for
the reaction of complex 3 with 40 equiv of PhSiD3 at
313 K.
(5) Monitoring the reaction of 3 with 200 equiv of
1
PhSiH3 at -30 °C by H NMR led to the detection of
small amounts of 3-methylindene24 in the reaction
mixture after 24 h; no trace of 1-phenylsilyl-3-methyl-
indene was found in this mixture. On the other hand,
running the reactions at 325 and 335 K allowed the
detection of both 1-phenylsilyl-3-methylindene25 (major)
1
and 3-methylindene (minor). After 24 h, the H NMR
spectrum of this mixture showed the formation of the
dimer, the trimer, and a number of higher linear
oligomers, while the 31P{1H} NMR spectrum showed
that the signal for 3 had been replaced by three broad
signals at ca. -7.0, -24.7, and -25.5 ppm. The first
signal is close to those of the complexes (1-Me-Ind)-
(PMe3)Ni-alkyl and might be due to an analogous Ni-
silyl or Ni-H species, whereas the more upfield signals
appear in the same region as the signals for the Ind-
free Ni(0) and Ni(II) complexes (e.g., -21 ppm for Ni-
(PMe3)4 and -22 ppm for (PMe3)2NiCl2).
(3) The rate of the initial Si-H bond activation
reaction between complexes 1-3 and a very large excess
of PhSiH3 (50-90 equiv) at 295 K was studied by
monitoring the disappearance of the Ni-CH3 and the
1
Ind-CH3 signals of the precatalysts by H NMR spec-
troscopy. Plots of kobs were linear for complexes 1 and
3, while the reaction of complex 2 was too slow for
convenient measurement. The kobs plots and tables of
data are provided in the Supporting Information; these
results show that the initial Si-H bond activation
proceeds ca. 15 times slower with 1 compared to 3 and
much slower still with 2. Studying the temperature
dependence of the reaction of complex 3 between 295
and 339 K allowed the determination of the activation
parameters from the Eyring plot (Figure 4): ∆Hq ) 10.7
( 0.7 kcal‚mol-1 and ∆Sq ) -42 ( 2 eu. The relatively
large negative value of ∆Sq implies a highly ordered
transition state for the reaction under study and is
The implications of the above observations for the
overall mechanism of the dehydrocoupling of PhSiH3
catalyzed by our complexes are discussed in the next
section.
(23) Note that the ligand exchange reaction between PCy3 and the
analogous complex (1-Me-indenyl)Ni(PPh3)Cl has been found to follow
an associative mechanism with the activation parameters ∆Hq ) 6.40
( 0.04 kcal‚mol-1 and ∆Sq ) -40 ( 4 eu (ref 8).
(24) From TOCSY and one-dimensional 1H NMR data (C6D6): 1H
NMR (C6D6): δ 5.99 (br, 1H, H2), 3.09 (m, 2H, H1), 2.01 (pseudo
quartet, J H-H ) 2.4, 3H, Me). The aromatic protons were obscured by
the much more intense signals due to excess PhSiH3.
(20) (a) Gauvin, F.; Harrod, J . F.; Woo, H.-G. Adv. Organomet. Chem.
1998, 42, 363. (b) Obora, Y.; Tanaka., M. J . Organomet. Chem. 2000,
595, 1.
(21) Corey, J . Y. Adv. Silicon Chem. 1991, 1, 327.
(25) From TOCSY and one-dimensional 1H NMR data (C6D6): δ 6.24
3
4
2
(22) Samples of the head gas for the reactions of 3 with large
excess of PhSiH3 were withdrawn using a gastight syringe and in-
jected into a Shimadzu GC-8A gas chromatograph equipped with an
Apiezon-L column and an FID detector. The chromatograms thus
obtained contained a peak at the same retention time as authentic
CH4 samples.
(dq, J H-H ) 3.7, J H-H ) 2.1, J Si-H (for the 29Si satellites) ) 160, 1
H, H2), 4.48 (dd, 2J H-H ) 7.8, 3J H-H ) 2.6, 1J Si-H (for the 29Si satellites)
2
3
1
) 204, 1 H, SiH2), 4.35 (dd, J H-H ) 7.8, J H-H ) 3.1, J Si-H (for the
29Si satellites) ) 204, 1 H, SiH2), 3.56 (m, 1 H, H1), 2.04 (dd, J H-H
)
4
5
2.1, J H-H ) 1.7, 3 H, Ind-CH3). The aromatic protons were obscured
by the much more intense signals due to excess PhSiH3.

406 Organometallics, Vol. 21, No. 2, 2002
Fontaine and Zargarian
Discu ssion
which they were measured: the higher temperatures
used for the studies involving the Hf systems can be
expected to result in higher zero-point energies, which
tend to reduce kinetic isotope effects. It appears, there-
fore, that the kinetic isotope effect for the reaction of
complex 3 with PhSiH3 is in the broad range of the
values reported for σ-bond metathesis processes.
Moreover, the ∆Sq value of -42 ( 2 eu found for our
system (vide supra) is also similar to those found for
the σ-bond metathesis reactions such as the reaction of
Cp*2ScMe with styrene27a (∆Sq ) -36 eu) or that of
PhSiH3 with Tilley’s Hf compounds mentioned above
(∆Sq ) -21 to -27 eu). Finally, the observation that
the rate of the Si-H activation step is very sensitive to
steric factors (3 > 1 . 2) is consistent with a highly
congested, late transition state typical of a concerted
process. Therefore, we propose that the initial Si-H
bond activation reactions between PhSiH3 and the
complexes under study proceed by a concerted σ-bond
metathesis process. This reaction is believed to generate
a Ni-SiPhH2 species that undergoes subsequent, rela-
tively rapid reactions to form PhH2Si-SiH2Ph and higher
oligomers; the various pathways that these reactions
can follow are considered next.
Since the silyl intermediate does not accumulate in
the reaction medium, it could not be detected directly;
therefore, we have relied on indirect evidence to deter-
mine the sequence of reactions involved in the Si-Si
bond formation process. Recall that NMR monitoring
of some catalytic reactions led to the detection of
1-phenylsilyl-3-methylindene and PMe3-containing spe-
cies which might be Ind-free Ni(0) and Ni(II) complexes
(vide supra); these observations hinted at the possibility
that the Ni-SiPhH2 intermediate might undergo a
reductive elimination reaction at some point during the
catalysis to produce Ni(0) species. To determine whether
Ind-free Ni(0) species might play a role in the catalysis,
we reacted Ni(PMe3)4, 4, with 200 equiv of PhSiH3 (5
M in toluene) and observed a vigorous bubbling of gas
that continued for a few hours and subsided gradually.
Analysis of the viscous oil formed over a few days
showed the formation of (PhSiH)n consisting of 60%
cyclic oligomers (Table 3, runs 9 and 10). Taken to-
gether, these observations suggest that (a) the Ni-
SiPhH2 intermediate arising from the initial Si-H bond
activation step may undergo reductive elimination to
give Ind-free Ni(0) species (eq 2), and (b) Ind-free Ni(0)
species might be involved in the Si-Si bond forming
reactions initiated by precatalysts such as 3 (eq 3).
Detailed studies reported by Tilley’s group on the
polymerization of silanes catalyzed by a family of
zirconocene complexes,3a and in particular on the reac-
13a
tion of PhSiH3 with CpCp*HfCl(SiR3) (SiR3 ) SiPhH2
and Si(SiMe3)326), have shown that the Si-H bond
activation and Si-Si bond formation steps involved in
these reactions proceed by a concerted, σ-bond metath-
esis mechanism. In contrast, little is known about the
mechanism of the polymerization of silanes by late
metals. In principle, low-valent late metal systems are
more likely to go through Si-H oxidative addition and
Si-Si reductive elimination sequences, whereas σ-bond
metathesis type processes commonly occur in d0 metal
systems wherein oxidative additions are not feasible.27
Nevertheless, it is possible that in some late metal
systems the oxidative addition, although feasible, might
be energetically less favored than a concerted process.
For example, Hartwig has reported that the elimination
of methane from the reaction of Cp(PPh3)2Ru-Me with
H-BR2 proceeds via σ-bond metathesis.28 We have relied
on these literature precedents and the mechanistic
findings outlined in the previous section to propose a
mechanistic picture for the system under study, as
described below.
We propose that the initial reaction between PhSiH3
and the Ni-Me complexes 1-3 (eq 1) proceeds by a
concerted σ-bond metathesis process. The main justifi-
cation for this proposal is the large kinetic isotope effect
found for the Si-H activation reaction by complex 3. A
number of literature reports29 have shown that the
oxidative addition of H-H and C-H bonds exhibits kH/
kD values in the range of 1-2; on the other hand, the
concerted, metathesis-type processes often display larger
kinetic isotope effects, typically in the range of 2-4,26
but values as high as 10-1730 have also been reported.31
The kH/kD value of 9.8 ( 0.5 found for the reaction of
complex 3 at 313 K is larger than the values of 2.5-2.7
reported by Tilley13a,25 for the reaction of PhSiH3 with
CpCp*Hf(SiR3)Cl at 343 K, but comparable to the value
of 9.7 reported by Bercaw et al.30a for the first-order
elimination (by σ-bond metathesis) of methane from
Cp*Ta(NMe2)Me3 at 307 K. Some of the discrepancy
between the kinetic isotope effects for the Si-H activa-
tion reactions by complex 3 and Tilley’s Hf complexes
is presumably due to the different temperatures at
(26) Woo, H.-G.; Heyn, R. H.; Tilley, T. D. J . Am. Chem. Soc. 1992,
114, 5698.
(27) For a discussion of this topic see: (a) 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.
(b) Ziegler, T.; Folga, E.; Berces, A. J . Am. Chem. Soc. 1993, 115,
636.
(28) Hartwig, J . F.; Bhandari, S.; Rablen, P. R. J . Am. Chem. Soc.
1994, 116, 1839.
(29) (a) Zhou, P.; Vitale, A. A.; San Filippo, J ., J r.; Saunders: W.
H., J r. J . Am. Chem. Soc. 1985, 107, 8049. (b) J anowicz, A. H.;
Bergman, R. G. J . Am. Chem. Soc. 1982, 104, 352. (c) Collman, J . P.;
Hegedus, L. S.; Norton, J . R.; Finke, R. G. Principles and Applications
of Organotransition Metal Chemistry; University Science Books: Mill
Valley, CA, 1987; pp 286-290.
(30) (a) Mayer, J . M.; Curtis, C. J .; Bercaw, J . E. J . Am. Chem. Soc.
1983, 105, 2651. (b) Hajela, S.; Schaefer, W. P.; Bercaw, J . E. J .
Organomet. Chem. 1997, 532, 45, and references therein.
(31) Unusually large kinetic isotope effects are thought to result
from the involvement of quantum mechanical tunneling effects in the
rate-determining step: Melander, L.; Saunders: W. H., J r. Reaction
Rates of Isotopic Molecules; Robert E. Krieger: Malabar, FL, 1987; p
140.
That the above-mentioned reductive elimination of
the Ni-SiPhH2 intermediate is not the only pathway
open to this species was inferred from two other pieces
of evidence. First, the reactions catalyzed by complex 3
give higher proportions of the linear oligomers compared
to those formed from the reaction of 4. In addition, the

Dehydrogenative Oligomerization of PhSiH3
Organometallics, Vol. 21, No. 2, 2002 407
Sch em e 1
bonds and a new Ni-H intermediate. In path D, the Ni-
silyl intermediate undergoes a reductive elimination
reaction to produce 1-PhSiH2-3-Me-indene and generate
a Ni(0) species. The Ni-H intermediate generated in
path B can then react with PhSiH3 (or, at a later stage
of the catalysis, with a higher oligomer) to extrude H2
and form another Ni-silyl species (path C). Alterna-
tively, the Ni-H species can eliminate 3-Me-Ind and
generate a Ni(0) species (path E).
Finally, the Ni(0) species formed in paths D and E is
thought to oxidatively add PhSiH3 (or a higher oligomer)
to generate a new Ni(II) hydrido(silyl) compound which
does not bear an indenyl ligand (path F); this species
can then react further with PhSiH3 (or a higher oligo-
mer) to give H2 and form Si-Si bonds (path G).
Although our results do not provide any information on
the mechanisms of paths B-E, we suspect that paths
B and C proceed via concerted reactions such as that
proposed for path A, because the structures of the Ni-
silyl and Ni-H species are similar to that of the Ni-Me
precursor. Moreover, paths D and E likely involve
slippage of the Ind ligand to form η1-Ind complexes that
should be prone to elimination.
formation of 3-Me-indene in some of the catalytic
reactions initiated by 3, especially those run at lower
temperatures, probably results from the reductive elimi-
nation of an intermediate such as (1-Me-Ind)(PMe3)Ni-
H;32 the latter might, in turn, form during an Si-Si
bond formation reaction between the (1-Me-Ind)(PMe3)-
Ni-SiPhH2 intermediate and PhSiH3 or some oligomeric
products (eq 4). Therefore, our observations are consis-
tent with a scenario in which the Ni-silyl intermediate
follows two pathways: one of these involves a reductive
elimination of the silyl and Me-Ind ligands to give a
Ni(0) species (eq 2), which catalyzes the dehydrogena-
tive oligomerization of PhSiH3 to (PhSiH)n products rich
in cyclic components (eq 3); in the second pathway, the
Ni-silyl intermediate reacts with PhSiH3 to give the
dimer and an analogous Ni-H species (eq 4), which can,
in turn, react with more PhSiH3 to extrude H2 and
reform the Ni-silyl intermediate (eq 5). The latter
pathway appears to be more dominant at lower tem-
peratures (see Table 3, run 5) and gives oligosilanes that
are rich in the linear component. The intermediate
complexes (1-Me-Ind)(PMe3)Ni-R (R ) H or SiPhH2)
involved in the second pathway eventually follow the
reductive elimination pathway to give R-Ind and the Ni-
(0) species.
The mechanism shown in Scheme 1 combines the
reductive elimination and oxidative addition pathways
common to late metal systems with the main elements
of Tilley’s σ-bond metathesis mechanism for Si-Si bond
formation by group 4 metallocenes. While some aspects
of our proposal are somewhat speculative at this stage,
we believe that the results of the mechanistic studies
presented here support the main outline of this mecha-
nistic picture.33 The next stage of our studies will be
focused on the preparation of model complexes for the
above-alluded Ni-silyl and Ni-H intermediates and
studying their reactivities with PhSiH3.34 We will also
study the analogous Cp complexes to determine if the
reductive elimination pathway observed with the Ind
complexes occurs to the same extent. The results of
these studies should advance our understanding of the
intimate mechanism of this catalytic process and lead
to the development of a new generation of precatalysts
with improved features.
Con clu sion s. The present study has demonstrated
that the complexes (Ind)Ni(PR3)(Me) can act as single-
component catalysts for the dehydrogenative oligomer-
ization of PhSiH3. Since these complexes are generated
in situ from the action of MAO on the Ni-Cl derivatives,
we conclude that the Ni-Me species are likely involved
in the catalysis promoted by the (Ind)Ni(PR3)(Cl)/MAO
system. It should be emphasized, however, that the role
of MAO in the latter system is not limited to methylat-
ing the Ni-Cl bond; indeed, since the (Ind)Ni(PPh3)-
(Cl)/MAO reactions are more rapid and give (PhSiH)n
with higher Mw values than those obtained from the
reactions of 1 or 3,6 we believe that MAO plays another,
as yet unknown, role in promoting the catalysis.
P r op osed Ca ta lytic Cycle. The above conclusions
can be regrouped to develop the tentative mechanistic
picture illustrated in Scheme 1. Thus, the Ni precata-
lysts are proposed to react with PhSiH3 through a
concerted process which releases methane and forms a
Ni-SiPhH2 intermediate (path A). This intermediate is
then partitioned into two different pathways, as follows.
In path B, the Ni-silyl intermediate reacts directly with
PhSiH3 (or, at a later stage of the catalysis, with a
higher oligomer), resulting in the formation of Si-Si
(33) A reviewer has suggested that the dependence of the Si-H
activation rate on the concentration of added phosphine be measured.
Unfortunately, it is not possible to carry out such studies because the
presence of excess phosphine (particularly PMe3) decomposes these
complexes, presumably via slippage of the Ind ligand to give (η1-Ind)-
Ni(PR3)n(Me) followed by reductive elimination.
(32) Since complexes (1-Me-Ind)Ni(PR3)(X) are stable to hydrolysis,
the formation of 3-Me-Ind is unlikely to be related to the residual
moisture present in the reaction medium.
(34) All attempts to prepare (1-Me-Ind)Ni(PR3)(X) where X ) H,
SiR′3 have resulted in decomposition, but PhSiH3 reacts with the
species formed in the mixture of these complexes with LiAlH4 (ref 6).

408 Organometallics, Vol. 21, No. 2, 2002
Fontaine and Zargarian
Su p p or tin g In for m a tion Ava ila ble: Complete details on
the X-ray analysis of 2 and 3, including tables of crystal data,
collection, and refinement parameters, bond distances and
angles, anisotropic thermal parameters, and hydrogen atom
coordinates; tables of the raw data for the kinetic studies. This
material is available free of charge via the Internet at
http://pubs.acs.org.
Ack n ow led gm en t. The authors gratefully acknowl-
edge NSERC (Canada) and Fonds FCAR (Que´bec) for
the financial support of this work; Professor J . F. Harrod
for valuable discussions; Prof. X. X. Zhu and his
research group for the use of their GPC instrument; and
Mr. Ralph Flachsbart from Wilmad Labglass for the gift
of the 528-TR-7 NMR tubes.
OM010757E