Mechanistic aspects of samarium-mediated σ-bond activations of arene C-H and arylsilane Si-C bonds

Article abstract of DOI:10.1021/ja011472w

To investigate the potential role of Sm-Ph species as intermediates in the samarium-catalyzed redistribution of PhSiH₃ to Ph₂SiH₂ and SiH₄, the samarium phenyl complex [Cp*₂SmPh]₂ (1) was prepared by oxidation of Cp*₂Sm (2) with HgPh₂. Compound 1 thermally decomposes to yield benzene and the phenylene-bridged disamarium complex Cp*₂Sm(μ-1,4-C₆H₄)SmCp* ₂ (3). This decomposition reaction appears to proceed through dissociation of 1 into monomeric Cp*₂SmPh species which then react via unimolecular and bimolecular pathways, involving rate-limiting Cp* metalation and direct C-H activation, respectively. The observed rate law for this process is of the form: rate = k₁[1] + k₂[1]². Complex 1 efficiently transfers its phenyl group to PhSiH₃, with formation of Ph₂SiH₂ and [Cp*₂Sm(μ-H)]₂ (4). Quantitative Si-C bond cleavage of C₆F₅SiH₃ is effected by the samarium hydride complex 4, yielding silane and [Cp*₂Sm(μ-C₆F₅)]₂ (5). In contrast, Si-H activation takes place upon reaction of 4 with o-MeOC₆H₄SiH₃, affording the samarium silyl species I Cp*₂SmSiH₂(o-MeOC₆H₄) (7). Complex 7 rapidly decomposes to [Cp*₂Sm(μ-o-MeOC₆H₄)]₂ (6) and other samarium-containing products. Compounds 5 and 6 were prepared independently by oxidation of 2 with Hg(C₆F₅)₂ and Hg(o-MeOC₆H₄)₂, respectively. The mechanism of samarium-mediated redistribution at silicon, and chemoselectivity in σ-bond metathesis reactions, are discussed.

Full text of DOI:10.1021/ja011472w

10526
J. Am. Chem. Soc. 2001, 123, 10526-10534
Mechanistic Aspects of Samarium-Mediated σ-Bond Activations of
Arene C-H and Arylsilane Si-C Bonds
Ivan Castillo and T. Don Tilley*
Contribution from the Department of Chemistry, UniVersity of California at Berkeley,
Berkeley, California 94720-1460
ReceiVed June 15, 2001
Abstract: To investigate the potential role of Sm-Ph species as intermediates in the samarium-catalyzed
redistribution of PhSiH3 to Ph2SiH2 and SiH4, the samarium phenyl complex [Cp*2SmPh]2 (1) was prepared
by oxidation of Cp*2Sm (2) with HgPh2. Compound 1 thermally decomposes to yield benzene and the phenylene-
bridged disamarium complex Cp*2Sm(µ-1,4-C6H4)SmCp*2 (3). This decomposition reaction appears to proceed
through dissociation of 1 into monomeric Cp*2SmPh species which then react via unimolecular and bimolecular
pathways, involving rate-limiting Cp* metalation and direct C-H activation, respectively. The observed rate
law for this process is of the form: rate ) k1[1] + k2[1]2. Complex 1 efficiently transfers its phenyl group to
PhSiH3, with formation of Ph2SiH2 and [Cp*2Sm(µ-H)]2 (4). Quantitative Si-C bond cleavage of C6F5SiH3
is effected by the samarium hydride complex 4, yielding silane and [Cp*2Sm(µ-C6F5)]2 (5). In contrast, Si-H
activation takes place upon reaction of 4 with o-MeOC6H4SiH3, affording the samarium silyl species
Cp*2SmSiH2(o-MeOC6H4) (7). Complex 7 rapidly decomposes to [Cp*2Sm(µ-o-MeOC6H4)]2 (6) and other
samarium-containing products. Compounds 5 and 6 were prepared independently by oxidation of 2 with
Hg(C6F5)2 and Hg(o-MeOC6H4)2, respectively. The mechanism of samarium-mediated redistribution at silicon,
and chemoselectivity in σ-bond metathesis reactions, are discussed.
Introduction
bond activation by early transition metal centers have been
reported by the group of Basset.7 Such heterogeneous systems,
which involve highly electrophilic, silica-supported early-
transition metal centers, allow chemical transformations of
hydrocarbons under mild conditions.
The activation of C-H1,2 and Si-H3 σ-bonds by f-element
complexes is well established. On the other hand, related
activations of C-C2a,4 and Si-C3f,5 bonds appear to be more
difficult and have been observed far less often, with the former
being limited to â-alkyl-transfer reactions. Several kinetic factors
seem to favor C-H over C-C activation, including the
inherently more hindered nature of C-C bonds, the statistical
abundance of C-H bonds in most hydrocarbons, and the higher
barrier for C-C activation due to its more directional bonding.6
Nevertheless, remarkable examples of metal-mediated C-C
(1) For example: (a) Watson, P. L. J. Am. Chem. Soc. 1983, 105, 6491.
(b) Jeske, G.; Lauke, H.; Mauermann, H.; Schumann, H.; Marks, T. J. J.
Am. Chem. Soc. 1985, 107, 8111. (c) Bruno, J. W.; Smith, G. M.; Marks,
T. J.; Fair, C. K.; Schultz, A. J.; Williams, J. M. J. Am. Chem. Soc. 1986,
108, 40. (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.
(2) For reviews see: (a) Watson, P. L.; Parshall, G. W. Acc. Chem. Res.
1985, 18, 51. (b) Bercaw, J. E. Pure Appl. Chem. 1990, 62, 1151. (c) Davis,
J. A., Watson, P. L., Liebman, J. F., Greenberg, A., Eds.; SelectiVe
Hydrocarbon ActiVation; VCH Publishers: New York, 1990. (d) Marks,
T. J. Acc. Chem. Res. 1992, 25, 57.
(3) For example: (a) Forsyth, C. M.; Nolan, S. P.; Marks, T. J.
Organometallics 1991, 10, 2543. (b) Sakakura, T.; Lautenschlager, H. J.;
Nakajima, M.; Tanaka, M. Chem. Lett. 1991, 913. (c) Tilley, T. D.; Radu,
N. S.; Walzer, J. F.; Woo, H.-G. Polym. Prepr. (Am. Chem. Soc., DiV. Polym.
Chem.) 1992, 33(1), 1237. (d) Molander, G. A.; Julius, M. J. Org. Chem.
1992, 57, 3266. (e) Radu, N. S.; Tilley, T. D. J. Am. Chem. Soc. 1995,
117, 5863. (f) Radu, N. S.; Hollander, F. J.; Tilley, T. D.; Rheingold, A. L.
J. Chem. Soc., Chem. Commun. 1996, 2459.
In metal-silicon chemistry, the analogous preference for
Si-H over Si-C activation may be explained by the same
factors recognized to account for chemoselectivity in C-H
versus C-C activation. Thus, in reactions of d0fn metal hydrides
with hydrosilanes, the usual reaction pathway involves de-
hydrocoupling via four-center transition state A and selective
formation of a metal silyl derivative (Scheme 1). Such species
have been invoked as intermediates in the dehydrocoupling of
silanes, by way of reaction with more hydrosilane via transition
state B to produce a Si-Si bond and regenerate the metal
hydride. Thus, the first two reactions of Scheme 1 can account
for the coordination-polymerization of organosilanes to poly-
silanes by early-transition metal and f-element complexes.8
The activation of Si-C bonds is potentially important in the
development of new processes in organosilicon chemistry. In
addition, studies of Si-C bond activations may provide
important insights into designing comparable chemistry for C-C
bonds, which is expected to be more difficult. Within this
context, we have observed Si-C bond activation in samarium-
(7) For example: (a) Lecuyer, C.; Quignard, F.; Choplin, A.; Olivier,
D.; Basset, J.-M. Angew. Chem., Int. Ed. 1991, 30, 1660. (b) Rosier, C.;
Niccolai, G. P.; Basset, J.-M. J. Am. Chem. Soc. 1997, 119, 12408. (c)
Maury, O.; Lefort, L.; Vidal, V.; Thivolle-Cazat, J.; Basset, J.-M. Angew.
Chem., Intl. Ed. 1999, 38, 1952. (d) Chabanas, M.; Vidal, V.; Cope´ret, C.;
Thivolle-Cazat, J.; Basset, J.-M. Angew. Chem., Int. Ed. 2000, 39, 1962.
(8) (a) Woo, H.-G.; Tilley, T. D. J. Am. Chem. Soc. 1989, 111, 8043.
(b) Woo, H.-G.; Heyn, R. H.; Tilley, T. D. J. Am. Chem. Soc. 1992, 114,
5698. (c) Woo, H.-G.; Walzer, J. F.; Tilley, T. D. J. Am. Chem. Soc. 1992,
114, 7047. (d) Tilley, T. D. Acc. Chem. Res. 1993, 26, 22.
(4) (a) Watson, P. L.; Roe, D. C. J. Am. Chem. Soc. 1982, 104, 6471.
(b) Bunel, E.; Burger, B. J.; Bercaw, J. E. J. Am. Chem. Soc. 1988, 110,
976.
(5) (a) Radu, N. S.; Tilley, T. D.; Rheingold, A. L. J. Organomet. Chem.
1996, 516, 41. (b) Castillo, I.; Tilley, T. D. Organometallics 2000, 19, 4733.
(6) (a) Crabtree, R. H. Chem. ReV. 1985, 85, 245 and references therein.
(b) Rybtchinski, B.; Milstein, D. Angew. Chem., Int. Ed. 1999, 38, 870.
10.1021/ja011472w CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/05/2001

Samarium-Mediated σ-Bond ActiVations
J. Am. Chem. Soc., Vol. 123, No. 43, 2001 10527
Scheme 1
A primary objective has been to identify fundamental processes
for the activation of Si-C (and other relatively inert) bonds.
Furthermore, it is of interest to establish factors that control
selectivity in reactions of electrophilic lanthanide complexes
with organosilanes (e.g., Si-C vs Si-H bond activation), and
their potential relevance to analogous processes involving
hydrocarbons.
Results and Discussion
In the samarium-catalyzed redistribution of PhSiH3 to SiH4,
Ph2SiH2, and Ph3SiH, a possible intermediate is the phenyl
derivative Cp*2SmPh (1, Scheme 2). To explore the possible
role of 1 in this chemistry, we sought to prepare and isolate
1
this species. The characterization of 1 in solution by H NMR
spectroscopy was reported by Evans and co-workers, and this
complex was isolated as its THF adduct Cp*2SmPh(THF).10
A
synthetic procedure that circumvents the use of ethereal solvents
that could potentially coordinate to the samarium center com-
prises oxidation of divalent Cp*2Sm (2) by diphenylmercury.11
In cyclohexane-d12, orange-red solutions of 1 were obtained in
Scheme 2
1
quantitative yield, as indicated by H NMR spectroscopy (eq
1). Attempts to isolate 1 that had been prepared in pentane or
benzene, however, led to low and variable yields (0-37%).
When the synthesis of 1 was attempted in toluene, reaction with
the solvent led to the isolation of Cp*2SmCH2Ph.12 Predominant
activation of the benzylic C-H bond of toluene has previously
been observed with related Y, La, and Ce systems.13
Cp*2Sm + 1/2HgPh2
8 1/2[Cp*2SmPh]2
(1)
-1/2Hg0
1
Monitoring the reaction in eq 1 in cyclohexane-d12 by H
NMR spectroscopy revealed the slow and quantitative decom-
position of 1 to benzene (0.5 equiv) and a samarium-containing
compound giving rise to a single Cp* resonance at δ 1.16.
Infrared spectroscopy revealed the presence of a low-energy
νCH stretching band at 2712 cm-1, suggesting an agostic
interaction.14 Storing samples of 1 in the dark did not inhibit
this clean decomposition. Manipulation of 1 proved virtually
impossible due to the fact that the thermal decomposition
reaction occurs in the solid state as well, even at -35 °C, such
that orange-red, crystalline 1 decomposes overnight under
nitrogen to a tan microcrystalline material (3). Combustion
analysis of the decomposition product of 1 is consistent with
the formula C23H32Sm. On the basis of the spectroscopic
evidence and the known analogous behavior of Cp*2LnPh
complexes (Ln ) Sc,1d Lu15), it was assumed that decomposition
product 3 is the p-phenylene-bridged dimer Cp*2Sm(µ-1,4-
C6H4)SmCp*2, which does not give rise to aromatic resonances
in the 1H NMR spectrum due to the proximity of all phenylene
protons to the paramagnetic Sm center. The tan complex 3 is
less soluble than 1 in aliphatic hydrocarbons, allowing for the
formation of X-ray quality crystals from concentrated cyclo-
hexane-d12 solutions.
mediated conversions of phenylsilane to silane, benzene, di-
phenylsilane, and triphenylsilane, which strongly compete with
the expected dehydrocoupling chemistry.3f,5b The Si-C activa-
tion could in principle occur via transfer of the phenyl group
of PhSiH3 to the metal-bound silyl fragment through a similar
four-centered transition state (C, Scheme 1). The pathway
involving transition state C, however, seems unlikely for both
steric and electronic reasons. Addition of a M-Si bond to the
Si-C bond of an organosilane produces an inherently crowded
transition state.3e In addition, theoretical studies indicate that
transition states with a carbon atom in the â-position are
disfavored.9
An alternative mechanism for Si-C bond activation involves
initial cleavage of the Si-C bond of phenylsilane via transition
state D (Scheme 2), yielding silane and a metal-phenyl species.
Arylation of PhSiH3 by M-Ph (E, Scheme 2) could account
for Ph2SiH2 formation, whereas protonation of M-Ph (F,
Scheme 2) could result in the production of benzene.
(10) (a) Evans, W. J.; Bloom, I.; Hunter, W. E.; Atwood, J. L.
Organometallics 1985, 4, 112. (b) Evans, W. J.; Gonzales, S. L.; Ziller, J.
W. J. Am. Chem. Soc. 1991, 113, 9880.
(11) We wish to thank Professor William J. Evans for suggesting the
synthesis of 1 by the oxidation method described.
(12) Evans, W. J.; Ulibarri, T. A.; Ziller, J. W. Organometallics 1991,
10, 134.
(13) (a) Booij, M.; Meetsma, A.; Teuben, J. H. Organometallics 1991,
10, 3246. (b) Booij, M.; Deelman, B.-J.; Duchateau, R.; Postma, D. S.;
Meetsma, A.; Teuben, J. H. Organometallics 1993, 12, 3531.
(14) Brookhart, M.; Green, M. L. H.; Wong, L. L. Prog. Inorg. Chem.
1988, 26, 1.
In light of the scarcity of well-defined examples of Si-C
σ-bond activation, and the potential relevance of this process
to related C-C bond activation with electrophilic early-transition
and f-element complexes, we have investigated mechanistic
aspects of the Cp*2SmR/PhSiH3 (R ) H, alkyl) reaction system.
(9) (a) Steigerwald, M. L.; Goddard, W. A., III. J. Am. Chem. Soc. 1984,
106, 308. (b) Rappe´, A. K. Organometallics 1990, 9, 466. (c) Folga, E.;
Ziegler, T. Can. J. Chem. 1992, 70, 333. (d) Ustynyuk, Y. A.; Ustynyuk,
L. Y.; Laikov, D. N.; Lunin, V. V. J. Organomet. Chem. 2000, 597, 182.
(15) Watson, P. L. J. Chem. Soc., Chem. Commun. 1983, 276.

10528 J. Am. Chem. Soc., Vol. 123, No. 43, 2001
Castillo and Tilley
Table 1. Summary of Crystallographic Data
compound
formula
3
5
C46H64Sm2
917.82
C52H60F10Sm2
1175.82
MW
cryst color, habit
cryst dimens, mm
cryst system
cell determination
(2θ range)
lattice parameters
a (Å)
tan, rodlike
red, rhomboidal
0.15 × 0.08 × 0.05
triclinic
0.31 × 0.05 × 0.06
monoclinic
2847 (4.0° - 45.0°)
3490 (4.0° - 45.0°)
8.6216(7
21.143(1)
11.9073(9)
10.065(1)
16.139(2)
17.323(2)
102.979(2)
97.128(2)
96.103(2)
2694.5(5)
P1h(No. 2)
4
b (Å)
c (Å)
Figure 1. ORTEP view of Cp*2Sm(µ-C6H4)SmCp*2 (3). The hollow
lines represent the agostic interactions.
R (deg)
â (deg)
â ) 101.063(1)°
γ (deg)
V (Å3)
2130.2(2)
P21/n (No. 14)
2
Compound 3 crystallizes in the monoclinic space group P21/
n. Four C23H32Sm fragments are contained in the unit cell, such
that the asymmetric unit consists of one-half of the molecule
including two Cp* rings, one samarium atom, and three carbon
atoms from the bridging phenylene ligand. The other half of
the molecule is related by an inversion center. One of the Cp*
rings presents rotational disorder about the centroid-samarium
vector, which was modeled as two superimposed Cp* groups
(each with 50% occupancy) related by a rotation of ap-
proximately 36°.
space group
Z value
D
calc (g/cm3)
1.431
1.449
µ(Mo KR)
diffractometer
radiation
temperature (°C)
scan type
no. of reflns measd
no. of reflns obsd
solution
27.64 cm-1
Siemens SMART
Mo-KR
22.28 cm-1
Siemens SMART
Mo-KR
-123.0
ω (0.3° per frame)
10330
1767 (I > 3.00σ(I))
direct methods (SIR92) direct methods (SIR92)
full-matrix least-squares full-matrix least-squares
-115.0
ω (0.3° per frame)
12182
4685 (I > 3.00σ(I))
refinement
A view of the molecular structure of 3 is provided in Figure
1. The Sm-Cipso distance of 2.42 Å is within the range of
reported Sm-C σ-bonds for metallocene derivatives.16 The Sm-
Cp* centroid distance of 2.40 Å is also within the expected
range for Sm(III) metallocenes,16 although it is among the
shortest, perhaps due to the electron deficient nature of 3. The
coordinative unsaturation of 3 is also evidenced by the presence
of a â-agostic interaction involving ortho C-H bonds of the
phenylene ligand. The Sm-C(28) distance of 2.79 Å involved
in the agostic interaction is shorter than the sum of the van der
Waals radii,17 but significantly longer than the Sm-Cipso
distance. The phenylene fragment is distorted by the â-agostic
interaction, such that the Cipso-C′ipso distance of 2.87 Å is
elongated relative to the C(28)-C′(28) distance of 2.68 Å, and
the Cipso-C′ipso vector is at a 30.7° angle with respect to the
Sm-Cipso bond vector. This type of â-H agostic interaction was
observed for the analogous Lu complex, but not for the Sc one.18
Finally, the wide centroid-Sm-centroid angle of 143.8°
probably reflects the small steric demand of the phenylene unit.
Table 1 summarizes the crystallographic data, and Table 2
provides selected bond distances and angles.
Reversibility of the Thermal Decomposition of 1 in
Benzene. In cyclohexane-d12 solution, the formation of 3 is
reversible, with the equilibrium lying toward the µ-phenylene
complex (eq 2). Appreciable amounts of 1 were detected only
when excess benzene was added to cyclohexane-d12 solutions
of 3 (Keq ) 8.7 at 25 °C). In benzene-d6 solution, 3 is
quantitatively transformed into Cp*2SmC6D5 (t1/2 ) 0.25 h at
50 °C). Solutions of Cp*2SmC6D5 can also be prepared by
dissolving 1 in benzene-d6. This conversion is indicated by
disappearance of the 1H NMR resonances for the phenyl group
R; Rw
0.044; 0.050
0.040; 0.060
max peak in diff. map 0.95 e-/Å3
2.86 e-/Å3
min peak in diff. map -0.68 e-/Å3
-0.53 e-/Å3
Table 2. Selected Bond Distances (Å) and Angles (deg) for 3a
(a) Bond Distances
2.4030(7)
2.3965(7)
Sm
Sm
Cp*(1)
Cp*(2)
Sm
Sm
C(27)
C(28)
2.42(2)
2.79(2)
(b) Bond Angles
Cp*(2) 143.76(3) Sm
C(27) C(28) 91(1)
Cp*(1) Sm
Sm
C(27) C(29) 152(1)
C(28) C(27) C(29) 117(1)
a Cp* denotes the centroid of the ring.
Although compound 1 decomposes in pentane or toluene
solution, as well as in the solid state, it is stable in benzene
solution over a period of weeks. The stability of 1 in benzene
allowed us to determine its solution molecular weight. By the
Signer method,19 1 exists as a dimer in benzene. This result is
consistent with the electrophilicity of the Sm center in 1, and
the fact that known metallocene hydrides and hydrocarbyls of
the lanthanides are often dimeric with a bridging hydride or
hydrocarbyl ligand.1b,16,20 Unfortunately, the instability of 1 has
precluded analysis of its solid-state structure by single-crystal
X-ray diffraction studies.
Mechanism of the Decomposition of [Cp*2SmPh]2 (1). To
gain insight into the mechanism of C-H σ-bond activation in
the decomposition of 1, we monitored the reaction by 1H NMR
spectroscopy at 78 °C. Due to the instability of 1, samples were
prepared in methylcyclohexane-d14 immediately prior to each
kinetic run. The experimental error in determining the concen-
tration was estimated to be j10%.
of 1, with concomitant formation of protiated benzene (t1/2
0.50 h at 50 °C).
)
At concentrations below 6.0 mM, plots of ln[1]/[1]0 versus
time reflect the first-order disappearance of 1 (Figure 2). In all
reactions, the rate of the formation of benzene was identical to
[Cp*2SmPh]
Cp* Sm(µ-C H )SmCp*
2 + PhH (2)
2 6 4
2 H
(18) Jahns, V.; Ko¨stlmeier, S.; Kotzian, M.; Ro¨sch, N.; Watson, P. L.
Int. J. Quantum Chem. 1992, 44, 853.
1
3
(19) Zoellner, R. W. J. Chem. Educ. 1990, 67, 714.
(20) For example: (a) Watson, P. L.; Herskovitz, T. ACS Symp. Ser.
1983, 212, 459. (b) Stern, D.; Sabat, M.; Marks, T. J. J. Am. Chem. Soc.
1990, 112, 9558. (c) Haar, C. M.; Stern, C. L.; Marks, T. J. Organometallics
1996, 15, 1765.
(16) (a) Evans, W. J.; Foster, S. E. J. Organomet. Chem. 1992, 433, 79.
(b) Schumann, H.; Meese-Marktscheffel, J. A.; Esser, L. Chem. ReV. 1995,
95, 865, and references therein.
(17) Emsley, J. The Elements; Oxford University Press: New York, 1990.

Samarium-Mediated σ-Bond ActiVations
J. Am. Chem. Soc., Vol. 123, No. 43, 2001 10529
of [Cp*2SmC6D5]2. A small isotope effect kH/kD ) 2.1(3) was
obtained from the low-concentration runs, whereas a much larger
value of kH/kD ) 5.3(7) was obtained at high concentrations.
The isotope effect obtained at low [1]0 corresponds to the ratio
of kobs (with kobs ) k1 + k2[1], as described above) for the
decomposition of [Cp*2SmC6H5]2 and [Cp*2SmC6D5]2, re-
spectively. The isotope effect determined from the high [1]0
kinetic runs, on the other hand, corresponds to the ratio of k2’s
for the second-order disappearances of [Cp*2SmC6H5]2 and
[Cp*2SmC6D5]2, respectively, since the bimolecular process
k2[1]2 predominates at high initial concentrations of 1. The
observed isotope effect is therefore a function of the concentra-
tion of [Cp*2SmC6D5]2.
Figure 2. Kinetic plots for the apparent first-order disappearance of 1
(rate ) kobs[1]) at initial concentrations of 3.0 and 6.0 mM.
Competing first- and second-order kinetic behavior was
previously observed in the related activations of benzene and
benzene-d6 by [Cp*2LuMe]2, which is a methyl-bridged dimer
in equilibrium with the monomeric species Cp*2LuMe.1a The
equilibrium amount of monomeric Cp*2LuMe in solution is
apparently responsible for the activation of benzene. At low
initial concentrations of benzene, a unimolecular reaction
pathway involving rate-limiting metalation of the C-H bond
of the Cp* ligand is associated with no kinetic isotope effect in
the activation of benzene-d6. This metalated species reacts
rapidly with both benzene and benzene-d6, resulting in no
isotope effect. At high initial concentrations of benzene, a
competing bimolecular reaction pathway was proposed to
involve direct activation of a C-H or C-D bond of benzene
or benzene-d6 by Cp*2LuMe in the transition state, resulting in
a large isotope effect at higher concentrations of benzene-d6
(where the bimolecular process predominates). The combination
of the two processes gives rise to a concentration-dependent
kinetic isotope effect.
A mechanism that is consistent with the observed rate
behavior and the kinetic isotope effect in our system involves
dimeric [Cp*2SmPh]2 in equilibrium with a monomeric species
Cp*2SmPh. The total concentration [1] therefore refers to the
equilibrium mixture of monomer and dimer. Although we did
not see evidence for more than one species in solution even at
-80 °C, this can be expected if the monomer/dimer intercon-
version is fast even at low temperature. Rapid monomer/dimer
equilibria in related f-element metallocene hydride and hydro-
carbyl complexes have been described for a number of
cases.1a,b,20 In those systems, the monomeric species display
enhanced σ-bond metathesis reactivity relative to their dimeric
counterparts, as in the case of Cp*2LuMe.1a,20b,c,22 Therefore, it
seems reasonable to assume that the equilibrium amount of
Cp*2SmPh present in solution might be more reactive toward
σ-bond metathesis than [Cp*2SmPh]2. Decomposition of
Cp*2SmPh via activation of a C-H bond of one of the Cp*
rings in the rate-limiting step would lead to loss of benzene
and a metalated species which could rapidly react with another
monomer of Cp*2SmPh to produce 3 (G in Scheme 3). Such
metalated species have been postulated as intermediates in
hydrocarbon activation by early-transition and f-block
metals.1d,12,13,23 Decomposition of Cp*2SmC6D5 by C-H activa-
tion of the Cp* ring in transition state G (Scheme 3) is consistent
with the small kinetic isotope effect measured at low concentra-
tions. At such low concentrations, the observed rate behavior
can be modeled as a first-order process, associated with the
unimolecular term k1[1] in the rate equation.
Figure 3. Plot of the values of kobs with increasing [1]0.
the rate of the disappearance of 1. With an initial concentration
of 6.0 mM, the decomposition reaction was kinetically well-
behaved for 2 half-lives, indicating a rate law given by k1[1].
Although the reactions involving lower initial concentrations
of 1 (1.5-6.0 mM) seemed to also agree with first-order
behavior, such kinetic runs revealed a dependence of the
observed rate constant (kobs) on [1]0 (Figure 3). Thus, the
reaction does not exhibit simple first-order behavior, and the
rate behavior appears to be consistent with competing first- and
second-order pathways (rate ) k1[1] + k2[1]2) with the first-
order rate constant given by the y-intercept of Figure 3 (where
[1]0 ) 0),21 4(1) × 10-6 s-1. At the low initial concentrations
of 1 examined, the second-order term is sufficiently small that
the rearranged rate equation, rate ) (k1 + k2[1])[1], can be
described as kobs[1], with kobs ) (k1 + k2[1]). The expression
k
obs[1] gives rise to the apparent first-order behavior at such
initial concentrations of 1 (1.5-6.0 mM).
When the initial concentration of 1 was raised to 12 mM,
the kinetic behavior could be modeled reasonably by a second-
order process, particularly early in the reaction, before [1] falls
into the “first-order regime”. At the high limit of initial
concentration of 1 before insolubility becomes problematic (18
mM), the disappearance of 1 gives reasonable second-order plots
to 50% conversion. Beyond 50% conversion, more complex rate
behavior results from competing first- and second-order path-
ways. From the value of k1 determined at low [1]0 (Figure 3),
it is possible to derive a value for k2 by plotting exp[k1t + ln([1]/
[1]0)] as a function of [1].21 This method gives a linear plot
with a slope ) [k2/(k1 + k2[1]0)], from which a value of k2 )
1(1) × 10-3 M-1 s-1 was extracted.
For determination of a kinetic isotope effect for the conversion
of 1 to 3, [Cp*2SmC6D5]2 was prepared from 2 and diphenyl-
mercury-d10. This allowed a determination of the kinetic isotope
effect at low (6.0 mM) and high (18 mM) initial concentrations
(22) 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.
(21) Espenson, J. H. Chemical Kinetics and Reaction Mechanisms; Series
in Advanced Chemistry; McGraw-Hill: New York, 1981.
(23) Bulls, A. R.; Schaefer, W. P.; Serfas, M.; Bercaw, J. E. Organo-
metallics 1987, 6, 1219.

10530 J. Am. Chem. Soc., Vol. 123, No. 43, 2001
Castillo and Tilley
Scheme 3
Scheme 4
proposed in the redistribution of Scheme 2 involves cleavage
of a Si-C(phenyl) bond by a samarium hydride. In the
samarium-catalyzed redistribution of PhSiH3, the phenyl deriva-
tive 1 was not observed as an intermediate. However, this is
expected on the basis of the results described above, since any
1 produced during the reaction of 4 with PhSiH3 would react
rapidly with any phenylsilane still present in solution. Therefore,
to directly observe a Si-C cleavage reaction (involving aryl
transfer from silicon to samarium) we focused on reactions of
4 with arylsilanes that might produce a more stable (and
observable) Sm-aryl product.
Previously, Andersen and Burns described the preparation
of a perfluorophenyl derivative of ytterbocene which is quite
stable.24 Its enhanced stability undoubtedly results from the
inductive effect of the fluorine atoms, which increase the anionic
character of the Yb-C bond. This suggested that the silane
C6F5SiH3 might allow the direct observation of aryl transfer
from silicon to samarium, with formation of a stable Sm-C6F5
complex. A second type of stabilized Sm-aryl derivative is
suggested by the work of Teuben and co-workers in the
synthesis of ortho-substituted yttrium-aryl complexes by C-H
activation of the substituted arenes.13b Preference for activation
of the ortho position was attributed to internal stabilization
provided by coordination of a heteroatom lone pair to the metal
center. Since activation of anisole proved to be the most facile,13b
we also targeted the o-MeOC6H4- group for studies on aryl
transfers.
A decomposition route involving two Cp*2SmPh species
could occur by activation of the C-H bond of one of the phenyl
groups at the para position, with concomitant loss of benzene
in the rate-limiting step (H, Scheme 3). This would give rise to
second-order rate behavior at high concentrations of 1 with a
bimolecular rate law k2[1]2. The large kinetic isotope effect
observed for the decomposition of Cp*2SmC6D5 at high initial
concentration of 1 is consistent with C-D activation in transition
state H (Scheme 3). At this time, we cannot completely rule
out mechanisms involving reaction of monomeric Cp*2SmPh
with the dimer, although such possibilities seem less consistent
with the observed kinetic data.
Reaction of 1 with Phenylsilane. A key reaction in the
postulated mechanism for the Sm-mediated redistribution of
PhSiH3 (Scheme 2) involves phenyl group transfer from
samarium to silicon. To test for this possibility, we examined
reactions of 1 with PhSiH3. Samples of 1 were generated in
cyclohexane-d12 solution. Addition of 1 equiv of PhSiH3 to
solutions of 1 at room temperature resulted in rapid conversion
1
to products (phenylsilane was consumed before the H NMR
spectrum could be acquired). The observed products (80%
1
Ph2SiH2 and 20% benzene as determined by H NMR spec-
A preparative procedure for C6F5SiH3 was recently reported
by the group of Molander.25 We could not find reports on the
synthesis of o-MeOC6H4SiH3, but reduction of o-MeOC6H4-
troscopy and GC-MS) are consistent with phenyl transfer to
silicon. The phenylation of silicon by 1 results in an 80%
conversion of the samarium species to [Cp*2Sm(µ-H)]2 (4),
while the remaining 20% of the samarium-based products could
26
SiCl3 with LiAlH4 provided the desired silane o-MeOC6H4-
SiH3 as a colorless liquid after distillation.
1
not be accounted for by H NMR spectroscopy.
To assess the stability of the targeted Sm-aryl complexes,
they were prepared and isolated by independent syntheses that
employed an oxidation method identical to that used for the
synthesis of 1. Thus, treatment of pentane slurries of 2 with
0.5 equiv of the corresponding diaryl mercury compounds
afforded [Cp*2Sm(µ-C6F5)]2 (5) and [Cp*2Sm(µ-o-MeOC6H4)]2
(6) in good yields (eq 3). Compounds 5 and 6 are air- and
moisture-sensitive but are stable in the solid state and in
cyclohexane-d12 or benzene-d6 solutions for at least one week.
Formation of 20% benzene in the reaction between 1 and
PhSiH3 is consistent with the pathway that goes through
transition state F in Scheme 2. This result is also in agreement
with our previous identification of benzene as a coproduct in
the samarium-mediated redistribution at silicon.5b The mecha-
nism for the production of Ph2SiH2 and benzene from samarium
phenyl and PhSiH3 is outlined in Scheme 4. Although the mech-
anism shown requires the formation of 20% Cp*2SmSiH2Ph, it
is reasonable that this samarium silyl complex would readily
decompose to insoluble trisamarium clusters.3f An altenative
explanation for the production of benzene from 1 involves direct
hydrogenolysis with H2. It is possible that dihydrogen could
form by dehydrocoupling of PhSiH3 with Cp*2SmSiH2Ph, but
on the basis of the small amount of Si-H relative to Si-C
activation, we believe that this is a minor reaction pathway.
Reactions of [Cp*2Sm(µ-H)]2 (4) with Arylsilanes. A
second, previously unobserved σ-bond metathesis step that is
1/2[Cp*2SmAr]2
Ar ) C6F5; 5
1
Cp* Sm
+ /2HgAr2
8
(3)
2
-1/2Hg0
2
o-MeOC6H4; 6
(24) Burns, C. J.; Andersen, R. A. J. Chem. Soc., Chem. Commun. 1989,
136.
(25) Molander, G. A.; Corrette, C. P. Organometallics 1998, 17, 5504.

Samarium-Mediated σ-Bond ActiVations
J. Am. Chem. Soc., Vol. 123, No. 43, 2001 10531
transfer from silicon to samarium, with the formation of a stable
Sm-aryl complex. Apparently, the strength of the Sm-C bond
of 5 precludes further reaction, and no aryl transfer was observed
from 5 to the silicon center of C6F5SiH3 at room temperature
after 24 h. Heating solutions of 5 and C6F5SiH3 to 65 °C for 12
h led to formation of pentafluorobenzene and complex mixtures
of samarium-containing products.
1/2[Cp*2Sm(µ-H)]2
+ C6F5SiH3 f
4
Figure 4. ORTEP diagram of [Cp*2Sm(µ-C6F5)]2 (5).
1/2[Cp*2Sm(µ-C6F5)]2 + SiH4
(4)
5
Table 3. Selected Bond Distances (Å) and Angles (deg) for 5
(a) Bond Distances
Sm(1)
Sm(2)
Sm(1)
Sm(2)
C(1)
C(27)
F(1)
2.60(1)
2.60(1)
2.531(8)
2.539(7)
Sm(1)
Sm(1)
Sm(2)
Sm(2)
Cp*(1)
Cp*(2)
Cp*(3)
Cp*(4)
2.4329(7)
2.4094(8)
2.4069(8)
2.4379(7)
Addition of o-MeOC6H4SiH3 (1 equiv) to benzene-d6 solu-
tions of 4 resulted in a different type of reaction. A new
samarium-containing product was transiently formed, giving
rise to a Cp* resonance in the H NMR spectrum at δ 1.21,
as well as dihydrogen. We believe that the samarium species
F(6)
1
(b) Bond Angles
Sm(1) C(1) 77.9(4) Cp*(1) Sm(1) Cp*(2) 137.26(3)
Sm(2) C(27) 77.3(4) Cp*(3) Sm(2) Cp*(4) 136.42(3)
F(1)
F(6)
is a â-methoxy-stabilized samarium-silyl complex Cp*2-
Sm(1) C(1) C(2) 102.1(9) Sm(2) C(27) C(28) 100.1(9)
Sm(1) C(1) C(6) 145(1) Sm(2) C(27) C(32) 146(1)
SmSiH2(o-MeOC6H4) (7, eq 5), on the basis of its solution IR
spectrum which contains a νSiH band at 2015 cm-1 (for
o-MeOC6H4SiH3, νSiH ) 2159 cm-1). A shift to lower frequen-
cies for the Si-H stretching band is associated with bonding
of silicon to an electropositive center (in this case samarium).3f,5b,27
a Cp* denotes the centroid of the ring.
1
The H NMR spectrum of 5 displays a single resonance for
the equivalent Cp* groups at δ 0.19, while that for 6 displays
one Cp* resonance at δ 0.94 (30 H), four aromatic resonances
(1 H each), and a paramagnetically shifted MeO resonance at δ
-4.23 (3 H), indicative of the proximity of the methyl protons
to the Sm center. Unfortunately, 6 crystallizes as very thin plates
that were not suitable for single-crystal X-ray analysis. Thus,
the mode of interaction of the o-methoxy group with the
samarium atom could not be firmly established. A dimeric
structure for 6 is proposed on the basis of the solution molecular
weight obtained by the Signer method.
The 19F NMR spectrum of 5 consists of three resonances in
a 2:1:2 ratio. The resonance assigned to the ortho fluorine atoms
is considerably broadened, indicating the presence of a Sm-F
interaction. This was confirmed by analyzing the solid-state
structure of 5.
X-ray quality crystals of 5 were obtained by cooling
concentrated pentane solutions to -35 °C. Compound 5
crystallizes in the triclinic space group P1h with two independent
monomers in the asymmetric unit that are related by an inversion
center. The C6F5-groups bridge two samarium atoms through
one ortho fluorine atom as shown in the ORTEP diagram of
Figure 4. The two samarocene subunits are virtually identical,
with average Cp* centroid-Sm distances of 2.42 Å and an
average Cp*-Sm-Cp* angle of 136.9°. The Sm-Cipso and
Sm-F contacts are also identical for the two subunits, with
values of 2.60 and 2.54 Å, respectively. Another feature of the
solid-state structure is the proximity of a second set of ortho
fluorine atoms to the Sm center, the distances being 2.85 and
2.82 Å between Sm(1)-F(2) and Sm(2)-F(10), respectively.
A weak interaction between such atoms could provide a
mechanism for the fluxional behavior of 5 in solution that gives
rise to averaged fluorine resonances for the ortho and meta
positions. The crystallographic data is summarized in Table 1,
and selected bond distances and angles are provided in Table
3.
Unfortunately this Sm-Si species could not be isolated due
to its decomposition in cyclohexane-d12 or benzene-d6 solution,
which leads to a complex mixture of samarium-containing
products after 1 h. However, samarium-aryl 6 was identified
as one of the products of this decomposition (24%), along with
anisole (76%). Thus, the expected product of Si-C bond
activation was observed, but only as the minor product of a
complex reaction. The reaction of C6F5SiH3 and 4, on the other
hand, provided a good model for the Si-C bond cleavage
proposed in the pathway that proceeds through transition state
D (Scheme 2).
The reaction of phenylsilane with 1 is not chemoselective,
in that phenyl transfer (80%) and benzene formation (20%) both
occur (Scheme 4). To investigate electronic effects in this
transformation, we examined the reactions of 1 with C6F5SiH3
and o-MeOC6H4SiH3. When 1 equiv of C6F5SiH3 was added
to a cyclohexane-d12 solution of 1, the main samarium-
containing product obtained after 10 min was the robust
compound 5 (85% by 1H NMR spectroscopy), along with trace
amounts of [Cp*2SmSiH3]3.5b This transformation was ac-
companied by transfer of phenyl groups to silicon, such that
0.23 equiv of PhSiH3 and 0.10 equiv of Ph2SiH2 were also
detected in solution (eq 6). Most of the phenyl groups of 1 were
converted to benzene (0.42 equiv), while the remaining phenyl
groups (0.15 equiv) produce Ph(C6F5)SiH2, which gives rise to
1
a Si-H triplet (JHF ) 8.8 Hz) in the H NMR spectrum at δ
4.99 (the latter silane was also identified by GC-MS). Produc-
tion of benzene implies initial cleavage of the Si-H bond of
C6F5SiH3 and formation of a Sm-SiH2(C6F5) species, which
was not detected. However, such a silyl complex could react
Reaction of the samarium hydride 4 with 1 equiv of C6F5SiH3
in benzene-d6 leads to the rapid and quantitative formation (by
1H NMR spectroscopy) of 5 and SiH4 (eq 4). Therefore,
activation of the Si-C σ-bond of C6F5SiH3 results in aryl-group
(26) Motsarev, G. V.; Inshakova, V. T.; Kolbasov, V. I.; Rozenberg, V.
R. Zh. Obshch. Khim. 1974, 44, 1053.
(27) Radu, N. S.; Tilley, T. D. J. Am. Chem. Soc. 1992, 114, 8293.

10532 J. Am. Chem. Soc., Vol. 123, No. 43, 2001
Castillo and Tilley
rapidly with PhSiH3 or C6F5SiH3 present in solution. The
complexity of the reaction allowed only partial quantification
of the silicon-containing products, and 50% of these products
are unaccounted for due to the formation of insoluble
[Cp*2SmSiH3]3 and untractable materials. Analysis of the
observed reaction products, however, allowed for a comparison
of the extent of Ph-transfer to silicon versus H-transfer to the
phenyl group, which are roughly equal. Thus, 0.48 equiv of
the phenyl groups of 1 are converted to PhSiH3, Ph2SiH2, and
Ph(C6F5)SiH2, while 0.42 equiv of the phenyl groups give rise
to benzene (eq 6).
pounds provide further evidence that Sm-aryl complexes are
viable intermediates in lanthanide-mediated transformations of
arylsilanes.
The arylsilane σ-bond activation chemistry is very sensitive
to the electronic effects of the substituents on the aryl groups,
such that reaction of the samarium hydride 4 with C6F5SiH3
resulted in clean Si-C bond cleavage. This transformation
represents a remarkable example of a system with a Si-C bond
that is more reactive than Si-H bonds. Thus, polarized Siδ+
-
Cδ- bonds such as that of C6F5SiH3 are particularly susceptible
to activation by metal hydride complexes. The electron-
withdrawing ability of the C6F5- group gives rise to an
electrophilic silicon center that is readily attacked by samarium
hydride 4. In addition, Si-C bond cleavage results in formation
of a robust Sm-C bond in complex 5. In contrast, reaction of
4 with o-MeOC6H4SiH3 resulted in quantitative Si-H acti-
vation. Exclusive activation of an Si-H bond in o-MeOC6H4-
SiH3 is favored by formation of a five-membered metallacycle
1/2[Cp*2SmPh]2
+ C6F5SiH3 f
1
(6)
0.42[Cp*2Sm(µ-C6F5)]2 + 0.42 PhH
5
0.23PhSiH3 + 0.10Ph2SiH2 + 0.15Ph(C6F5)SiH3
In the case of o-MeOC6H4SiH3, addition of 1 equiv to
cyclohexane-d12 solutions of 1 resulted primarily in protonation
of the phenyl ligand (71% by H NMR spectroscopy) with
Cp*2SmSiH2(o-MeOC6H4) (7), which is stabilized by the
samarium-oxygen interaction. In addition, the electron-rich
o-MeOC6H4- group makes the silicon center of o-MeOC6H4-
SiH3 the least electrophilic in the series C6F5SiH3 > PhSiH3 >
o-MeOC6H4SiH3. The intermediate electrophilicity of PhSiH3
results in competitive Si-C (redistribution) and Si-H (dehydro-
coupling) bond activation in reactions with 4. The electron-
rich silicon center of o-MeOC6H4SiH3 is apparently not readily
attacked by 4, resulting in quantitative Si-H activation and no
Si-C activation.
Phenyl-group transfer from 1 to arylsilanes appears to also
be sensitive to electronic effects, as reactions of 1 with C6F5SiH3
led to competing phenyl transfer and phenyl protonation
(approximately 1:1). The lack of selectivity for Si-C versus
Si-H activation in this process likely arises from the presumably
weaker Sm-Ph bond relative to the Sm-H bond of 4,28 and
the lower relative nucleophilicty of Sm-Ph. Phenyl-group
transfer from 1 to the silicon center of o-MeOC6H4SiH3 was
minimal, accounting for only 10% of the identified silicon-
containing products. Predominant Si-H activation in this
reaction is apparently favored by the directing effect of the
methoxy-substituted arylsilane. The greater extent of phenyl
transfer from 1 to C6F5SiH3 and PhSiH3 versus o-MeOC6H4-
SiH3 is also in agreement with the greater electrophilicity of
the Si center of the former two, which results in more facile
nucleophilic transfer of the phenyl group.
1
concomitant formation of the Sm-silyl complex 7 shown in
eq 5 (45%) after 10 min. Among the products identified from
the complex reaction mixture were the Sm-aryl complex 6
(37%), Ph(o-MeOC6H4)SiH2 (0.1 equiv by 1H NMR and GC-
MS) and anisole (0.1 equiv). The presence of the methoxy group
seems to direct reaction toward the â-methoxy-stabilized silyl
complex, and a very small amount of phenyl transfer to the
silicon center is observed.
Conclusions
In synthesizing [Cp*2SmPh]2 (1), we discovered an intra-
molecular C-H activation that leads to the formation of
Cp*2Sm(µ-1,4-C6H4)SmCp*2 (3) and benzene in chemistry that
is analogous to that observed with related Sc1d and Lu15 phenyl
systems. Two pathways involving the C-H activation of a Cp*
ring and a phenyl group, respectively, were identified for the
thermal decomposition of 1. This behavior is analogous to that
observed for benzene activation by [Cp*2LuMe]2.1a In both cases
an equilibrium between monomeric and dimeric species seems
to be present, with the former being responsible for the σ-bond
metathesis chemistry. Such monomeric lanthanide hydrocarbyls
can react following two pathways: a unimolecular one involving
rate-limiting C-H activation of a Cp* ligand, and a bimolecular
one involving rate-limiting C-H activation of the substrate
(benzene or Cp*2SmPh). The first pathway exhibits no kinetic
isotope effect, but the second pathway is characterized by a large
isotope effect, as expected for C-H activation of benzene or
Cp*2SmPh in the rate-limiting step. The two competing
pathways lead to a concentration-dependent kinetic isotope effect
arising from the relative contributions of the two reaction
manifolds at different concentrations. In Watson’s Lu system,
the bimolecular mechanism dominates the σ-bond activation
chemistry, whereas in our Sm system the two components are
of comparable magnitude at the initial concentrations of 1
investigated.
This work emphasizes the importance of electronic factors
that control Si-C versus Si-H activation in organo-f-element
systems. With a better understanding of these factors, synthetic
efforts should allow development of more selective catalysts.
Likewise, development of metal complexes with enhanced
electrophilicity might lead to more active systems, as recently
demonstrated in hydrocarbon-metal chemistry.7 Future work
will address additional aspects of the role of electronic and steric
factors in metal-mediated Si-C and Si-H bond activations.
Experimental Section
General Considerations. Unless otherwise specified, all manipula-
tions were performed under a nitrogen or argon atmosphere using
standard Schlenk techniques or an inert atmosphere drybox. Dry,
oxygen-free solvents were employed throughout. Olefin-free pentane
was obtained by treatment with concentrated H2SO4, then 0.5 N KMnO4
in 3 M H2SO4, followed by NaHCO3, and finally MgSO4. Thiophene-
free benzene and toluene were obtained by pretreating the solvents with
In further studies, it was possible to show that phenyl-group
transfer from 1 to phenylsilane is preferred over hydrogen
transfer from phenylsilane (or dihydrogen) to the phenyl group
of 1 (4:1 ratio) and that 1 is a feasible intermediate in the
[Cp*2Sm(µ-H)]2-catalyzed redistribution of phenyl groups at
silicon. Since 1 proved to be unstable, the isolable samarium-
aryl complexes [Cp*2Sm(µ-C6F5)]2 (5) and [Cp*2Sm(µ-o-
MeOC6H4)]2 (6) were independently prepared. These com-
(28) Nolan, S. P.; Stern, D.; Marks, T. J. J. Am. Chem. Soc. 1989, 111,
7844.

Samarium-Mediated σ-Bond ActiVations
J. Am. Chem. Soc., Vol. 123, No. 43, 2001 10533
concentrated H2SO4, followed by Na2CO3, and CaCl2. Pentane, benzene,
toluene, and diethyl ether were distilled from sodium/benzophenone
and stored under nitrogen prior to use, whereas benzene-d6 and toluene-
d8 were vacuum-distilled from Na/K alloy. Cyclohexane-d12 and
methylcyclohexane-d14 were vacuum-distilled from Na and stored under
nitrogen. Reagents were purchased from commercial suppliers and used
without further purification unless otherwise specified. Cp*2Sm,29
[Cp*2Sm(µ-H)]2,30 and C6F5SiH325 were prepared by literature methods.
NMR spectra were recorded on Bruker AMX-300, AMX-400, or DRX-
500 spectrometers at ambient temperature unless otherwise noted.
Elemental analyses were performed by the Microanalytical Laboratory
in the College of Chemistry at the University of California, Berkeley.
Infrared spectra were recorded on a Mattson Infinity 60 FT IR
instrument. Samples were prepared as KBr pellets unless otherwise
extracted with 60 mL of pentane. This pentane extract was filtered and
concentrated to a viscous oil. The mixtures obtained consisted of
o-MeOC6H4SiCl3 in 65% yield (14.2 g, 59.1 mmol) contaminated with
ca. 8% of the diaryldichlorosilane (o-MeOC6H4)2SiCl2. 1H NMR (300
MHz, benzene-d6) δ 3.15 (s, 3 H, OMe), 6.26 (d, 1 H, Ar), 6.67 (m, 1
H, Ar), 7.05 (m, 1 H, Ar), 7.63 (m, 1 H, Ar). Samples obtained in this
manner were employed for the synthesis of o-MeOC6H4SiH3 without
further purification. Thus, a 1:1 mixture by volume of o-MeOC6H4-
SiCl3 (14.2 g, 59.1 mmol) and diethyl ether was added dropwise via
addition funnel to a stirred solution of LiAlH4 in 150 mL of diethyl
ether over a period of 45 min. After the addition was complete, the
mixture was heated to reflux and stirred for 2 h. Upon cooling to room
temperature, the products were quenched with 2-propanol and then
water (50 mL each), and the resulting mixture was neutralized with 3
N HCl solution. The organic phase was isolated with a separatory
funnel. The aqueous layer was washed with 2 × 25 mL of diethyl
ether, and the combined organic layers were dried over MgSO4, filtered,
and concentrated with a rotary evaporator. The clear liquid obtained
was dried over CaH2, and distilled at 124 °C under an atmosphere of
nitrogen for a yield of 68% (5.52 g, 40.0 mmol). IR (benzene-d6
solution) 3068 (w, br), 3007 (w), 2959 (m), 2937 (w), 2836 (w, sh),
2159 (s, νSiH), 1589 (m), 1573 (m), 1475 (m), 1462 (m), 1430 (m),
1297 (w), 1278 (m, sh), 1241 (s), 1181 (m), 1162 (w), 1132 (m, sh),
1086 (m), 1043 (m), 1024 (m), 940 (s), 920 (s), 798 (w), 759 (s), 661
(m), 647 (m), 504 (w, sh), 492 (w, sh). 1H NMR (300 MHz, benzene-
d6) δ 3.22 (s, 3 H, OMe), 4.41 (s, 3 H, SiH3), 6.40 (d, 1 H, Ar), 6.81
(m, 1 H, Ar), 7.16 (m, 1 H, Ar), 7.46 (m, 1 H, Ar). 13C{1H} NMR
(126 MHz) δ 55.24 (OMe), 109.91 (Ar), 117.67 (Ar), 121.49 (Ar),
132.87 (Ar), 138.59 (Ar), 165.13 (Ar). 29Si NMR (376 MHz) δ -64.68.
Anal. Calcd for C7H10OSi: C, 60.82; H, 7.29. Found: C, 60.92; H,
7.49.
noted, and data are reported in units of cm-1
.
Caution! All organomercurial compounds described are potentially
toxic and should be handled with caution. Manipulation with protective
gloves in a well-ventilated fume hood is recommended.
[Cp*2SmPh]2 (1). A Schlenk tube equipped with a magnetic stirbar
was charged with Cp*2Sm (2) (0.20 g, 0.48 mmol) and HgPh2 (0.08 g,
0.24 mmol). As benzene was added (ca. 20 mL), gray, metallic mercury
began to deposit on the bottom of the flask. After stirring the mixture
for 12 h under argon, the volatile materials were removed under
vacuum, and the orange-red mass was extracted with 15 mL of pentane,
cannula-filtered and concentrated to a volume of ca. 5 mL. Cooling to
-35 °C afforded orange-red crystals in 37% yield (0.09 g, 0.09 mmol).
Due to its thermal instability, compound 1 could only be characterized
in solution by NMR spectroscopy: 1H NMR (500 MHz, cyclohexane-
d12) δ 0.71 (s, 30 H, Cp*), 6.84 (d, 2 H, m-Ph), 7.71 (t, 1 H, p-Ph).
Solution MW: 1160 ( 120. Calcd for [Cp*2SmPh]2: 995.92.
Cp*2Sm(µ-C6H4)SmCp*2 (3). Orange-red crystals of 1 (0.09 g, 0.09
mmol) were left standing overnight in a vial inside an inert atmosphere
box. Yellow-tan microcrystals of 3 were recovered in quantitative yield
(0.08 g, 0.09 mmol): mp > 260 °C (190 °C dec). IR 2964 (s), 2903
(s), 2857 (s), 2712 (m, νagosticCH), 2628 (w), 2516 (w, br), 1480 (m),
1438 (m), 1380 (m), 1343 (w), 1212 (m), 1083 (w), 1060 (w), 1022
(m), 948 (w, br), 800 (w, sh), 728 (m, sh), 675 (w, sh), 606 (w), 590
Hg(o-MeOC6H4)2. A solution of o-methoxyphenyllithium was
prepared as described for the synthesis of o-MeOC6H4SiCl3 from
o-bromoanisole (5.00 g, 26.7 mmol). The cold solution (-40 °C) was
then added dropwise via cannula to a stirred slurry of HgBr2 (4.80 g,
13.4 mmol) in 50 mL of diethyl ether kept at 0 °C. The suspension
was allowed to warm to room temperature, and then it was stirred
vigorously for 12 h. At that point the reaction mixture was quenched
with 100 mL of water, and the organic phase was isolated with a
separatory funnel. The aqueous phase was extracted with 3 × 25 mL
of diethyl ether, and the combined organic layers were concentrated in
a rotary evaporator until a white crystalline solid was obtained.
Recrystallization from toluene afforded two crops of white crystals in
73% combined yield (4.03 g, 9.75 mmol). Hg(o-MeOC6H4)2 had been
prepared previously by different methods,32 but complete characteriza-
tion has not been reported. Mp 107 °C (lit.32 108 °C). IR 3065 (m),
3005 (m, sh), 2949 (s), 2921 (m), 2902 (m), 2827 (s, sh), 1573 (s),
1461 (s, br), 1424 (s, br), 1296 (m), 1280 (m), 1231 (s), 1179 (m),
1162 (m), 1119 (m), 1063 (s), 1026 (w), 935 (w, sh), 790 (m, sh), 756
1
(w), 451 (m, sh). H NMR (500 MHz, cyclohexane-d12) δ 1.16 (s, 60
H, Cp*). 13C{1H} NMR (126 MHz) δ 18.87 (C5Me5), 119.96 (C5Me5),
125.86 (ipso-Ph), 128.70 (o-Ph). Anal. Calcd for C46H64Sm2: C, 60.20;
H, 7.03. Found: C, 60.03; H, 7.01.
Hg(C6D5)2. To a -80 °C solution of bromobenzene-d5 (1.50 g, 9.26
mmol) in 30 mL of diethyl ether was added a hexanes solution of 1.6
M n-butyllithium (5.80 mL, 9.26 mmol) via syringe. The mixture was
stirred for 2 h at -40 °C and then added with a cannula to a diethyl
ether slurry of HgBr2 (1.67 g, 4.63 mmol) kept at 0 °C. The reaction
mixture was warmed to room temperature and stirred overnight. The
products were quenched with 20 mL of water, and the organic layer
was isolated with a separatory funnel. The aqueous layer was washed
twice with 25 mL of toluene, and the combined organic phases were
dried over MgSO4, filtered, and concentrated with a rotary evaporator.
Cooling the concentrated solution to -35 °C afforded 0.44 g of white
crystals. A second crop was obtained from the mother liquor for a
combined yield of 74% (1.25 g, 3.43 mmol). Hg(C6D5)2 was simply
analyzed by 1H NMR spectroscopy (benzene-d6) to confirm the absence
of resonances in the aromatic region, and its melting point was
compared to that of HgPh2: mp 123-125 °C (lit.31 122 °C).
o-MeOC6H4SiH3. To a diethyl ether solution of o-bromoanisole
(17.0 g, 90.9 mmol) kept at -80 °C was added a hexanes solution of
1.6 M n-butyllithium (56.8 mL, 90.9 mmol) dropwise with an addition
funnel. The mixture was warmed to -40 °C and stirred for an hour.
After re-cooling to -80 °C, the solution was slowly added to a stirred,
-80 °C solution of SiCl4 (15.4 g, 90.9 mmol) in 50 mL of diethyl
ether. After the addition was complete, the reaction mixture was allowed
to reach room temperature, and then it was stirred overnight. Volatile
materials were then evaporated under vacuum, and the products were
1
(s), 723 (m), 568 (w, sh). H NMR (500 MHz, benzene-d6) δ 3.28 (s,
3 H, OMe), 6.68 (d, 1 H, Ar), 7.04 (m, 1 H, Ar), 7.15 (m, 1 H, Ar),
7.19 (m, 1 H, Ar). 13C{1H} NMR (126 MHz) δ 54.99 (OMe), 110.62
(Ar), 122.01 (Ar), 129.52 (Ar), 138.21 (Ar), 158.79 (Ar), 164.97 (Ar).
Anal. Calcd for C14H14HgO2: C, 40.53; H, 3.40. Found: C, 40.78; H,
3.64.
[Cp*2Sm(µ-C6F5)]2 (5). A mixture of 2 (0.20 g, 0.48 mmol) and
Hg(C6F5)2 (0.13 g, 0.24 mmol) was suspended in 20 mL of pentane
and stirred vigorously for 2 h in a Schlenk tube. The resulting red
solution was cannula-filtered into another Schlenk tube, leaving behind
gray, metallic mercury. The solution was concentrated to a volume of
ca. 8 mL and cooled to -35 °C. Two crops of red crystalline 5 were
obtained for a total yield of 64% (0.18 g, 0.15 mmol): mp > 260 °C.
IR 2964 (s), 2908 (s), 2911 (s), 2860 (s), 2726 (w), 1634 (w, br), 1596
(w), 1532 (m, sh), 1512 (m), 1487 (m), 1424 (s), 1379 (m), 1355 (w),
1311 (w), 1222 (m, sh), 1179 (w), 1068 (m), 1027 (m), 954 (w), 921
(s), 803 (w), 774 (w), 717 (w, sh), 579 (m, br), 474 (w). 1H NMR (500
MHz, benzene-d6) δ 0.19 (s, 60 H, Cp*). 13C{1H} NMR (126 MHz) δ
19.88 (C5Me5), 121.81 (C5Me5). 19F NMR (376 MHz) δ -161.4 (d, 2
F, m-Ph), -152.0 (t, 1 F, p-Ph), -146.8 (s, 2 F, o-Ph). Anal. Calcd for
C52H60F10Sm2: C, 53.12; H, 5.14. Found: C, 53.27; H, 5.24.
(29) Burns, C. J. Ph.D. Thesis, Materials and Chemical Sciences Division,
Lawrence Berkeley Laboratory, University of California, Berkeley, 1987.
(30) Jeske, G.; Lauke, H.; Mauermann, H.; Swepston, P. N.; Schumann,
H.; Marks, T. J. J. Am. Chem. Soc. 1985, 107, 8091.
(31) Wade, R. C.; Seyferth, D. J. Organomet. Chem. 1970, 22, 265.
(32) Kozyrod, R. P.; Pinhey, J. T. Aust. J. Chem. 1985, 38, 1155.

10534 J. Am. Chem. Soc., Vol. 123, No. 43, 2001
Castillo and Tilley
[Cp*2Sm(µ-o-MeOC6H4)]2 (6). A Schlenk tube was charged with
2 (0.20 g, 0.48 mmol) and Hg(o-MeOC6H4)2 (0.10 g, 0.24 mmol). The
solids were slurried in 20 mL of pentane, and the suspension was stirred
vigorously for 2 h. A yellow solution was obtained, which was filtered
and concentrated to a volume of ca. 8 mL. Cooling to -35 °C afforded
yellow plates of 6 in 63% yield (0.16 g, 0.15 mmol): mp 209-212 °C
(205 °C dec). IR 2962 (s), 2911 (s), 2857 (s), 2726 (w), 1601 (w),
1498 (m), 1440 (m), 1413 (m), 1378 (m), 1247 (m), 1125 (w), 1101
(Rint ) 0.073). No correction for decay was necessary. The structure
was solved using direct methods (SIR92) and refined by full-matrix
least-squares methods using teXsan. The number of variable parameters
was 211, giving a data/parameter ratio of 8.37. The maximum and
minimum peaks on the final difference Fourier map correspond to 0.95
and -0.68 e-/Å3: R ) 0.044, Rw ) 0.050, GOF ) 1.36. The
crystallographic data are summarized in Table 1.
X-ray Crystal Structure Determination of 5. A red, rhomboidal
crystal of approximate dimensions 0.15 mm × 0.08 mm × 0.05 mm
was mounted on a glass capillary using Paratone N hydrocarbon oil
and placed under a stream of cold nitrogen on the same Siemens
SMART diffractometer. Preliminary orientation matrix and unit cell
parameters were determined by collecting 60 20-s frames. A hemisphere
of data was collected at a temperature of -115 ( 1 °C using ω scans
of 0.30° and a collection time of 20 s per frame. Frame data were
integrated using SAINT. An absorption correction was applied using
XPREP (Tmax ) 0.959, Tmin ) 0.828). The 12182 reflections integrated
were averaged in point group P1h to yield 8593 unique reflections (Rint
) 0.031). No correction for decay was necessary. The structure was
solved using direct methods (SIR92) and refined by full-matrix least-
squares methods using teXsan. The number of variable parameters was
577, giving a data/parameter ratio of 8.12. The maximum and minimum
peaks on the final difference Fourier map correspond to 2.86 and -0.53
e-/Å3: R ) 0.040, Rw ) 0.060, GOF ) 1.66. The crystallographic
data are summarized in Table 1.
1
(w), 1043 (w), 1014 (w), 755 (m, sh), 578 (m), 552 (m). H NMR
(500 MHz, cyclohexane-d12) δ -4.23 (s, 6 H, OMe), 0.94 (s, 60 H,
Cp*), 4.65 (s, 2 H, -Ph), 6.52 (d, 2 H, -Ph), 7.63 (d, 2 H, -Ph), 8.50
(t, 2 H, -Ph). 13C{1H} NMR (126 MHz) δ 17.29 (C5Me5), 107.88
(OMe), 118.83 (C5Me5), 129.97 (-Ph). Solution MW: 1250 ( 120.
Calcd for [Cp*2Sm(µ-o-MeOC6H4)]2: 1055.98. Anal. Calcd for
C54H74O2Sm2: C, 61.42; H, 7.06. Found: C, 61.43; H, 7.17.
Kinetic Studies. Samples of 1 for kinetic runs were prepared by
weighing the appropriate amounts of 2 and HgPh2 in the drybox,
dissolving in slightly less than 1 mL of methylcyclohexane-d14, and
filtering into a 1 mL volumetric flask. The sample was then brought to
a volume of exactly 1 mL and transferred to a J-Young tube. A
convenient rate of decomposition was observed at 78 ( 1 °C (as
determined by calibration with an ethylene glycol standard) at initial
concentrations of 1 ranging from 1.5 to 18 mM. The reaction mixtures
were heated in the DRX 500 spectrometer probe and monitored by 1H
NMR spectroscopy over a period of 8-10 h. The concentrations of 1
and benzene were obtained by integration of the Cp* and aromatic
resonances, respectively. The error in integration was estimated to be
5%.
X-ray Crystal Structure Determination of 3. A tan, rodlike crystal
of approximate dimensions 0.31 × 0.05 × 0.06 mm was mounted on
a glass capillary using Paratone N hydrocarbon oil and placed under a
stream of cold nitrogen on a Siemens SMART diffractometer with a
CCD area detector. Preliminary orientation matrix and unit cell
parameters were determined by collecting 60 20-s frames. A hemisphere
of data was collected at a temperature of -123 ( 1 °C using ω scans
of 0.30° and a collection time of 20 s per frame. Frame data were
integrated using SAINT. An absorption correction was applied using
XPREP (Tmax ) 0.855, Tmin ) 0.573). The 10330 reflections integrated
were averaged in point group P21/n to yield 3927 unique reflections
Acknowledgment is made to Professor Richard Andersen
for useful discussions and a generous gift of Hg(C6F5)2, and to
the National Science Foundation for their generous support of
this work.
Supporting Information Available: Tables of crystal, data
collection and refinement parameters, atomic coordinates, bond
distances, bond angles, and anisotropic displacement parameters
for 3 and 5; kinetic plots and data (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
JA011472W