Synthesis and reactivity of silyl ruthenium complexes: The importance of trans effects in C-H activation, Si-C bond formation, and dehydrogenative coupling of silanes

Article abstract of DOI:10.1021/ja035131p

A series of octahedral ruthenium silyl hydride complexes, cis-(PMe₃)₄Ru(SiR₃)H (SiR₃ = SiMe₃, 1a; SiMe₂CH₂SiMe₃, 1b; SiEt₃, 1c; SiMe₂H, 1d), has been synthesized by the reaction of hydrosilanes with (PMe₃)₃Ru(η²-CH₂ PMe₂)H (5), cis-(PMe₃)₄RuMe₂ (6), or (PMe₃)₄RuH₂ (9). Reaction with 6 proceeds via an intermediate product, cis-(PMe₃)₄Ru(SiR₃)Me(SiR₃ = SiMe₃, 7a; SiMe₂CH₂SiMe₃, 7b). Alternatively, 1 and 7 have been synthesized via a fast hydrosilane exchange with another cis-(PMe₃)₄Ru(SiR₃)H or cis-(PMe₃)₄-Ru(SiR₃)Me, which occurs at a rate approaching the NMR time scale. Compounds 1a, 1b, 1d, and 7a adopt octahedral geometries in solution and the solid state with mutually cis silyl and hydride (or silyl and methyl) ligands. The longest Ru - P distance within a complex is always trans to Si, reflecting the strong trans influence of silicon. The aptitude of phosphine dissociation in these complexes has been probed in reactions of 1a, 1c, and 7a with PMe₃-d₉ and CO. The dissociation is regioselective in the position trans to a silyl ligand (trans effect of Si), and the rate approaches the NMR time scale. A slower secondary process introduces PMe3-d₉ and CO in the other octahedral positions, most likely via nondissociative isomerization. The trans effect and trans influence in 7a are so strong that an equilibrium concentration of dissociated phosphine is detectable (～5%) in solution of pure 7a. Compounds 1a-c also react with dihydrogen via regioselective dissociation of phosphine from the site trans to Si, but the final product, fac-(PMe₃)₃Ru(SiR₃)H₃ (SiR₃ = SiMe₃, 4a; SiMe₂CH₂SiMe₃, 4b; SiEt₃, 4c), features hydrides cis to Si. Alternatively, 4a-c have been synthesized by photolysis of (PMe₃)₄RuH₂ in the presence of a hydrosilane or by exchange of fac-(PMe₃)₃Ru(SiR₃)H₃ with another HSiR₃. The reverse manifold - HH elimination from 4a and trapping with PMe₃ or PMe₃-d₉ - is also regioselective (1a-d₉ is predominantly produced with PMe₃-d₉ trans to Si), but is very unfavorable. At 70 °C, a slower but irreversible SiH elimination also occurs and furnishes (PMe₃)₄RuH₂. The structure of 4a exhibits a tetrahedral P₃Si environment around the metal with the three hydrides adjacent to silicon and capping the P₂Si faces. Although strong Si...HRu interactions are not indicated in the structure or by IR, the HSi distances (2.13-2.23(5) A) suggest some degree of nonclassical SiH bonding in the H₃SiR₃ fragment. Thermolysis of 1a in C₆D₆ at 45-55 °C leads to an intermolecular CD activation of C₆D₆. Extensive H/D exchange into the hydride, SiMe₃, and PMe₃ ligands is observed, followed by much slower formation of cis-(PMe₃)₄Ru(D)(Ph-d₅). In an even slower intramolecular CH activation process, (PMe₃)₃Ru(η²-CH₂ PMe₂)H (5) is also produced. The structure of intermediates, mechanisms, and aptitudes for PMe₃ dissociation and addition/elimination of H-H, Si-H, C-Si, and C-H bonds in these systems are discussed with a special emphasis on the trans effect and trans influence of silicon and ramifications for SiC coupling catalysis.

Full text of DOI:10.1021/ja035131p

Published on Web 06/10/2003
Synthesis and Reactivity of Silyl Ruthenium Complexes: The
Importance of Trans Effects in C-H Activation, Si-C Bond
Formation, and Dehydrogenative Coupling of Silanes
Vladimir K. Dioumaev, Leo J. Procopio, Patrick J. Carroll, and Donald H. Berry*
Contribution from the Department of Chemistry and Laboratory for Research on the Structure of
Matter, UniVersity of PennsylVania, Philadelphia, PennsylVania 19104-6323
Received March 13, 2003; E-mail: dberry@sas.upenn.edu
Abstract: A series of octahedral ruthenium silyl hydride complexes, cis-(PMe3)4Ru(SiR3)H (SiR3 ) SiMe3,
1a; SiMe2CH2SiMe3, 1b; SiEt3, 1c; SiMe2H, 1d), has been synthesized by the reaction of hydrosilanes with
(PMe3)3Ru(η2-CH2PMe2)H (5), cis-(PMe3)4RuMe2 (6), or (PMe3)4RuH2 (9). Reaction with 6 proceeds via an
intermediate product, cis-(PMe3)4Ru(SiR3)Me (SiR3 ) SiMe3, 7a; SiMe2CH2SiMe3, 7b). Alternatively, 1 and
7 have been synthesized via a fast hydrosilane exchange with another cis-(PMe3)4Ru(SiR3)H or cis-(PMe3)4-
Ru(SiR3)Me, which occurs at a rate approaching the NMR time scale. Compounds 1a, 1b, 1d, and 7a
adopt octahedral geometries in solution and the solid state with mutually cis silyl and hydride (or silyl and
methyl) ligands. The longest RusP distance within a complex is always trans to Si, reflecting the strong
trans influence of silicon. The aptitude of phosphine dissociation in these complexes has been probed in
reactions of 1a, 1c, and 7a with PMe3-d9 and CO. The dissociation is regioselective in the position trans
to a silyl ligand (trans effect of Si), and the rate approaches the NMR time scale. A slower secondary
process introduces PMe3-d9 and CO in the other octahedral positions, most likely via nondissociative
isomerization. The trans effect and trans influence in 7a are so strong that an equilibrium concentration of
dissociated phosphine is detectable (∼5%) in solution of pure 7a. Compounds 1a-c also react with
dihydrogen via regioselective dissociation of phosphine from the site trans to Si, but the final product,
fac-(PMe3)3Ru(SiR3)H3 (SiR3 ) SiMe3, 4a; SiMe2CH2SiMe3, 4b; SiEt3, 4c), features hydrides cis to Si.
Alternatively, 4a-c have been synthesized by photolysis of (PMe3)4RuH2 in the presence of a hydrosilane
or by exchange of fac-(PMe3)3Ru(SiR3)H3 with another HSiR3. The reverse manifold - HH elimination
from 4a and trapping with PMe3 or PMe3-d9 - is also regioselective (1a-d9 is predominantly produced with
PMe3-d9 trans to Si), but is very unfavorable. At 70 °C, a slower but irreversible SiH elimination also occurs
and furnishes (PMe3)4RuH2. The structure of 4a exhibits a tetrahedral P3Si environment around the metal
with the three hydrides adjacent to silicon and capping the P2Si faces. Although strong Si‚‚‚HRu interactions
are not indicated in the structure or by IR, the HSi distances (2.13-2.23(5) Å) suggest some degree of
nonclassical SiH bonding in the H3SiR3 fragment. Thermolysis of 1a in C6D6 at 45-55 °C leads to an
intermolecular CD activation of C6D6. Extensive H/D exchange into the hydride, SiMe3, and PMe3 ligands
is observed, followed by much slower formation of cis-(PMe3)4Ru(D)(Ph-d5). In an even slower intramolecular
CH activation process, (PMe3)3Ru(η2-CH2PMe2)H (5) is also produced. The structure of intermediates,
mechanisms, and aptitudes for PMe3 dissociation and addition/elimination of H-H, Si-H, C-Si, and C-H
bonds in these systems are discussed with a special emphasis on the trans effect and trans influence of
silicon and ramifications for SiC coupling catalysis.
Introduction
a sacrificial olefin). Many of these reactions were reported for
18e- complexes of ruthenium, (PMe3)4Ru(SiR3)H (1) or (PMe3)3-
Catalytic formation of SiC bonds from CH- and SiH-
containing substrates are rare examples of efficient catalytic CH
bond functionalization.1-4 The side product of these reactions
is either dihydrogen (coupling of hydrosilanes with aromatic
substrates and dehydrogenative coupling of alkylsilanes to
carbosilanes) or alkane (transfer dehydrogenative coupling with
Ru(SiR3)H3 (4). It has been long recognized that the true cata-
lytic species in these processes are 16e- (PMe3)3Ru(SiR3)R and
18e- (PMe3)3Ru(SiR3)a(R)bHc (a + b + c ) 4, R ) H, alkyl,
or aryl), which interconvert via addition/elimination of HH, SiH,
CH, and SiC bonds. The aptitude and selectivity for the addition
of SiH and CH bonds versus elimination of SiC and HH (or
another CH) directly impact the rate of the dehydrogenative
(or dealkanative) catalysis. These issues are particularly acute
in the case of (PMe3)3Ru(SiR3)a(R)bHc intermediates, which can
conceivably eliminate any combination of HH, CH, SiH, SiC,
(1) Procopio, L. J.; Berry, D. H. J. Am. Chem. Soc. 1991, 113, 4039-4040.
(2) Procopio, L. J.; Mayer, B.; Plo¨ssl, K.; Berry, D. H. Polym. Prepr. (Am.
Chem. Soc., DiV. Polym. Chem.) 1992, 33, 1241-1242.
(3) Djurovich, P. I.; Dolich, A. R.; Berry, D. H. J. Chem. Soc., Chem. Commun.
1994, 1897-1898.
(4) Ezbiansky, K.; Djurovich, P. I.; LaForest, M.; Sinning, D. J.; Zayes, R.;
Berry, D. H. Organometallics 1998, 17, 1455-1457.
9
10.1021/ja035131p CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 8043-8058
8043

A R T I C L E S
Dioumaev et al.
CC, or SiSi bonds. Indeed, each of the above elimination pro-
cesses has been separately observed, and there are precedents
for competitive eliminations in M(SiR3)(R)H species.5-10
The aptitude for addition/elimination is readily understood
in terms of bond disruption enthalpies (BDE) and the size
and directionality of the orbitals involved. The usual order -
HH ≈ SiH (fast) > CH > SiC (slow) - is not particularly
favorable for the CSi bond formation catalysis; however,
interesting exceptions exist. For example, an unusual accelera-
tion has been reported for hydrosilylation reactions with
chelating hydrosilanes.11-13 The effect correlates with the length
of chelating linkage, and the best rates are achieved with a 2-3
carbon atom bridge. The catalytic species have been studied
by NMR and concluded to be bis(silyl) trihydrides, (PPh3)2Rh-
(R2Si∼SiR2)(H)3.13 In contrast, only monosilyl species were
detected with nonchelating silanes under identical conditions.13
It has been suggested that the presence of the second silyl on
the metal and the geometrical restrictions imposed by the chelate
facilitate addition/elimination processes.
The ability of silicon to labilize other ligands, especially those
trans to Si, has been noted previously.14 Particularly relevant
for this paper are examples of the trans effect in six-15-18 and
seven-coordinate complexes.19,20 In addition to the kinetic
consequences, silyl ligands are also known for a ground-state
phenomenon of weakening and elongation of the metal ligand
bonds trans to silicon. This thermodynamic trans influence has
been used to explain bond elongation in the structures of
six-7,8,18,19,21-27 and seven-coordinate complexes20 as well as the
stability of five-coordinate 16e- species with an empty site trans
to Si.28-30 Because silyl species are inherently present in all
catalytic SiC coupling processes, it is of interest to study the
role of the silicon trans effects and influences on the chemistry
of ruthenium silyl compounds and exploit this information to
design optimal catalysts.
We now report the synthesis of a series of silyl ruthenium
complexes, precursors to SiC coupling catalysts. Dissociation
of PMe3 and addition/elimination of H-H, Si-H, C-Si, and
C-H bonds in these systems are studied with a special emphasis
on the trans effect/influence of silicon and are discussed in the
context of SiC coupling catalysis.
Results
Synthesis and Characterization of cis-(PMe3)4Ru(SiR3)H
Complexes. Ruthenium silyl hydride complexes cis-(PMe3)4-
Ru(SiR3)H (1) have been prepared by three main methods. Our
original synthesis involved photolysis of (PMe3)3Ru(η2-CH2-
PMe2)H, 5,31,32 in the presence of excess trimethylsilane leading
to isolation of cis-(PMe3)4Ru(SiMe3)H (1a, eq 1) in 59% yield.
The triethylsilyl derivative cis-(PMe3)4Ru(SiEt3)H (1c) was
prepared similarly from HSiEt3 in 37% yield. cis-(PMe3)4Ru-
(SiMe2H)H (1d) was prepared in 80% yield by photolysis of a
different precursor, (PMe3)4RuH2 (9), in excess H2SiMe2. Note
that this approach depends on the unfavorable equilibrium
between the initially formed (PMe3)3Ru(H)3(SiMe2H) with 1d
(vide infra) and is not a viable synthetic reaction for larger silyl
groups.
(5) Thorn, D. L.; Harlow, R. L. Inorg. Chem. 1990, 29, 2017-2019.
(6) Cleary, B. P.; Mehta, R.; Eisenberg, R. Organometallics 1995, 14, 2297-
2305.
(7) Aizenberg, M.; Milstein, D. Angew. Chem. 1994, 106, 344-346 (See
also: Angew. Chem., Int. Ed. Engl. 1994, 1933, 1317-1319).
(8) Aizenberg, M.; Milstein, D. J. Am. Chem. Soc. 1995, 117, 6456-6464.
(9) Mitchell, G. P.; Tilley, T. D. Organometallics 1998, 17, 2912-2916.
(10) Goikhman, R.; Aizenberg, M.; Ben-David, Y.; Shimon, L. J. W.; Milstein,
D. Organometallics 2002, 21, 5060-5065.
However, these methods require rather long reaction times
(ca. 10 days), and it was subsequently found that silyls can be
more conveniently prepared by the reaction of (PMe3)4RuMe2
(6) and excess hydridosilane at 60 °C (eq 2). The major
byproducts of this reaction are CH4, MeSiR3, and a carbosilane
produced by dehydrocoupling of the starting silane. In many
cases, these materials are volatile and easily separated from the
ruthenium silyl product. Prolonged reaction times at higher
temperatures result in further dehydrocoupling to less volatile
silane products and contamination of the products with trihydride
complexes, 4, both of which complicate isolation of the silyl
hydrides, 1.
(11) Nagashima, H.; Tatebe, K.; Itoh, K. J. Chem. Soc., Perkin Trans. 1 1989,
1707-1708.
(12) Nagashima, H.; Tatebe, K.; Ishibashi, T.; Sakakibara, J.; Itoh, K. Orga-
nometallics 1989, 8, 2495-2496.
(13) Nagashima, H.; Tatebe, K.; Ishibashi, T.; Nakaoka, A.; Sakakibara, J.; Itoh,
K. Organometallics 1995, 14, 2868-2879.
(14) Chatt, J.; Eaborn, C.; Ibekwe, S. D.; Kapoor, P. N. J. Chem. Soc. A 1970,
1343.
(15) (a) Pomeroy, R. K.; Gay, R. S.; Evans, G. O.; Graham, W. A. G. J. Am.
Chem. Soc. 1972, 94, 272-274. (b) Pomeroy, R. K. J. Organomet. Chem.
1979, 177, C27.
(16) Pomeroy, R. K.; Wijesekera, K. S. Inorg. Chem. 1980, 19, 3729-3735.
(17) Pomeroy, R. K.; Hu, X. Can. J. Chem. 1982, 60, 1279-1285.
(18) Haszeldine, R. N.; Parish, R. V.; Setchfield, J. H. J. Organomet. Chem.
1973, 57, 279-285.
(19) Okazaki, M.; Ohshitanai, S.; Iwata, M.; Tobita, H.; Ogino, H. Coord. Chem.
ReV. 2002, 226, 167-178.
(20) Dioumaev, V. K.; Yoo, B. R.; Procopio, L. J.; Carroll, P. J.; Berry, D. H.
J. Am. Chem. Soc., in press.
(21) Crozat, M. M.; Watkins, S. F. J. Chem. Soc., Dalton Trans. 1972, 2512.
(22) Holmes-Smith, R. D.; Stobart, S. R.; Vefghi, R.; Zaworotko, M. J.; Jochem,
K.; Cameron, T. S. J. Chem. Soc., Dalton Trans. 1987, 969.
(23) Zlota, A. A.; Frolow, F.; Milstein, D. J. Chem. Soc., Chem. Commun. 1989,
1826-1827.
(24) Levy, C. J.; Puddephatt, R. J.; Vittal, J. J. Organometallics 1994, 13, 1559-
1560.
(25) Levy, C. J.; Vittal, J. J.; Puddephatt, R. J. Organometallics 1996, 15, 2108-
2117.
In many instances, the most convenient route to analogous
ruthenium silyl complexes is the equilibrium exchange of the
rather hindered and labile cis-(PMe3)4Ru(SiMe3)H (1a) with
other hydridosilanes (eq 3). The reaction is rapid at room
temperature, and the volatility of HSiMe3 facilitates its removal.
(26) Coe, B. J.; Glenwright, S. J. Coord. Chem. ReV. 2000, 203, 5-80.
(27) Reinartz, S.; White, P. S.; Brookhart, M.; Templeton, J. L. Organometallics
2000, 19, 3748-3750.
(28) Heyn, R. H.; Macgregor, S. A.; Nadasdi, T. T.; Ogasawara, M.; Eisenstein,
O.; Caulton, K. G. Inorg. Chim. Acta 1997, 259, 5-26.
(29) Dioumaev, V. K.; Plo¨ssl, K.; Carroll, P. J.; Berry, D. H. J. Am. Chem.
Soc. 1999, 121, 8391-8392.
(30) Dioumaev, V. K.; Ploessl, K.; Carroll, P. J.; Berry, D. H. Organometallics
2000, 19, 3374-3378.
(31) Werner, H.; Werner, R. J. Organomet. Chem. 1981, 209, C60.
(32) Gotzig, J.; Werner, R.; Werner, H. J. Organomet. Chem. 1985, 290, 99.
9
8044 J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003

Synthesis and Reactivity of Silyl Ruthenium Complexes
A R T I C L E S
1
The equilibrium substitution with a silyl group smaller than
SiMe3 (e.g., SiMe2H) is quite facile, whereas exchange with
larger silyls such as SiMe2CH2SiMe3 is not favored (Keq(300)
≈ 6.5 × 10-3) and requires excess silane and periodic removal
of HSiMe3 during the reaction. At temperatures below ∼60 °C,
no decomposition or polymerization products are produced, and
workup is straightforward. In the case of low boiling silanes,
removal of all volatiles under vacuum yields the pure ruthenium
silyl product quantitatively. The rate of exchange with 1a with
free silane approaches the NMR time scale (500 MHz, 30 °C),
as is evident from the drift of the RuSiMe3 1H NMR resonance
(∆δ ) 0.10) in the presence of ca. 10 equiv of HSiMe3.
Degenerate exchange is chemically verified by the rapid
incorporation of deuterium in the hydride position of 1a in
reactions with DSiMe3.
time scale obscures much of the scalar coupling in the H and
31P NMR spectra, and the structural assignment is most reliably
confirmed by the single-crystal X-ray diffraction study (vide
infra). In a separate experiment, treatment of 7a with a large
excess of HSiMe2CH2SiMe3 leads to equilibrium amounts of
(PMe3)4Ru(SiMe2CH2SiMe3)Me, 7b (Keq(300) ) 5.9 × 10-3).
Compound 7b was not isolated and was identified by 1H NMR
spectroscopy. As in the case of 1a, 7a undergoes rapid exchange
with free silane as indicated by a shift of the RuSiMe3 1H NMR
resonance in the presence of HSiMe3.
Phosphine Lability in (PMe3)4Ru(SiR3)X Complexes: Re-
actions with PMe3-d9. Many of the silyl complexes described
above exhibit much greater phosphine lability than is found in
other known cis-(PMe3)4Ru(X)(Y) complexes (X, Y ) H, Me,
Cl). For example, treatment of 1a with excess PMe3-d9 at 25
°C is fast (<5 min) and highly regioselective: 1 equiv of
unlabeled phosphine is released, and the labeled ligand is
incorporated into only one position in the complex (eq 5). The
31P NMR resonance for the exchanged phosphine site shifts by
ca. 2.4 ppm upfield from δ -17 in 1a to -19.4 in 1a-d9, whereas
the other resonances are unperturbed. Significantly, this same
phosphine resonance shows slight broadening in the presence
of free PMe3, suggesting the rate of the exchange process is
approaching the NMR time scale (25 °C, 80 MHz). For
comparison, phosphine dissociation from the dihydride, P4RuH2
(9), requires hours at 120 °C.
The cis-octahedral structures of the silyl hydride complexes
are clearly established by three distinct phosphine environments
1
in a 2:1:1 ratio in H and 31P NMR spectra and strong virtual
1
coupling between mutually trans phosphines in the H NMR
spectra. The hydride and silyl ligands are readily identified by
characteristic chemical shifts and multiplicities: 1H NMR ca.
δ -11 (dtd or dq) for RuH and ca. δ 0.7-1.0 for RuSiMen.
The silyl resonances are found between δ 11 and -1 ppm in
the 29Si NMR. The classical nature of the hydride and silyl
ligands is strongly suggested by the ruthenium-hydride stretch-
ing frequencies in the IR spectrum (ν(RuH) ) ca. 1820-1790
cm-1); significant agostic Si‚‚‚H interactions would decrease
this value below ca. 1650 cm-1 20
Further confirmation of the
.
classical nature of the silyl hydride complexes is found in the
solid-state structures, vide infra.
Products and Processes in the Reaction of HSiMe3 with
6. The reaction of 6 with HSiMe3 at 60 °C is a very convenient
synthetic route to 1a, but the stoichiometry of the reaction
suggests a fairly complicated mechanism involving competitive
C-H and C-Si elimination pathways which warranted closer
scrutiny. The thermal reaction of 6 with ca. 10 equiv of HSiMe3
(18 h, 60 °C) yields 1a (1 equiv), SiMe4 (0.25 equiv), HSiMe2-
CH2SiMe3 (0.75 equiv), and CH4 (not quantified.) No other
ruthenium complexes are observed, except for traces of 1b,
which is in equilibrium with 1a.
However, treatment of 6 with HSiMe3 at room temperature
permits observation of an intermediate complex, cis-(PMe3)4-
Ru(SiMe3)Me (7a). For example, after 18 h, the reaction mixture
consists of 40% 6, 50% 7a, and ∼10% silyl hydride species 1a
and 1b. Further reaction leads to a decrease in the concentration
of 7a, and increased amounts of 1a and 1b, which are in
equilibrium with each other and the respective hydridosilanes
(eq 4). A pure sample of 7a was isolated from 1a, 1b, and 6
with some difficulty via fractional crystallization (ca. 13% yield).
The spectral features of the silyl and phosphine ligands in 7a
are similar to those of 1a, 1b, or 1c, and a multiplet at δ -0.68
in the 1H NMR can be assigned to the methyl group on
ruthenium. However, rapid phosphine dissociation on the NMR
The specific site of PMe3 exchange in 1a can be assigned to
that trans to the silyl ligand by selective {1H}31P experiments:
decoupling only the methyl groups of 1a reveals strong coupling
between only one phosphine and the ruthenium hydride (δ
-15.2, JPH ≈ 61 Hz). This resonance is thus due to the
phosphine trans to the hydride. As the equivalent, mutually trans,
phosphines can be conclusively assigned on the basis of the
normal {1H}31P spectrum, the remaining resonance (δ -17) -
the extremely labile position - can be, therefore, attributed to
the PMe3 ligand trans to the trimethylsilyl group. Although the
initial phosphine exchange is site selective, reaction of 1a with
PMe3-d9 eventually leads to incorporation of labeled phosphine
into all of the sites. However, the process is fairly slow, and a
statistical distribution is reached only after weeks at 25 °C in
the presence of ca. 12 equiv of PMe3-d9. Reactions of the more
9
J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003 8045

A R T I C L E S
Dioumaev et al.
sterically crowded 1c with PMe3-d9 are also initially selective,
but secondary scrambling is faster than in 1a. Thus, selective
incorporation of labeled phosphine into the position trans to
the silyl in 1c also occurs in <5 min, but scrambling into all
sites is complete in <1 h (1H NMR).
Selective exchange with PMe3-d9 with the methyl silyl
complex 7a occurs rapidly on the NMR time scale at room
temperature. The phosphine presumed trans to silicon (in the
absence of exchange: 1H NMR δ 0.98, 31P NMR δ -21.7) is
in coalescence with free PMe3 (1H NMR δ 0.78, 31P NMR δ
-62.5) or with PMe3-d9 (31P NMR δ -65.5), and coupling to
the other phosphines in 7a is lost. In this case, further exchange
of labeled ligand into the other positions occurs within minutes
(ca. 20% in 3 min). In contrast, 5, which does not bear silyl
ligands, is considerably less prone to phosphine dissociation
and exchange (ca. 50% exchange after 102 h at 65 °C or ca.
20% exchange after 24 h of photolysis at 350 nm).
Furthermore, the equilibrium concentration of the 16e-
species generated from 7a (eq 6) is sufficiently high to be
detected. At 300 K, an isolated sample of 7a exhibits a very
broad 31P NMR resonance at δ -22.80 (ν1/2 ) 700 Hz), and
no free PMe3 is detected. However, at 240 K, two signals
decoalesce: δ -20.28 for the phosphine trans to Si in 7a and
-62.5 corresponding to ca. 7% free PMe3. The amount of free
PMe3 decreases to <5% at 190 K. Discrete 31P signals for 7a
and the 16e- (PMe3)3Ru(SiMe3)Me are not resolved, presumably
due to much smaller frequency differences of the coordinated
phosphines in these exchanging species as compared to ∆δ ≈
42 ppm for coordinated and free ligand. Additional evidence
for appreciable concentrations of the unsaturated ruthenium
complex is found in the dependence of the SiMe3 chemical shift
for 7a on temperature (e.g., 1H NMR (C7D8): δ 0.49 at 300 K
vs δ 0.81 at 190 K), and upon phosphine concentration (δ
(C6D12, 338 K) δ 0.12 without phosphine and δ 0.15 in the
presence of 5 equiv of added PMe3).
At room temperature, a benzene-d6 solution of mer-8 converts
to fac-8 with a half-life of ca. 35 min, ultimately yielding an
equilibrium ratio of 95% fac-8 to 5% mer-8, and recrystallized
product redissolves to yield the same 95:5 mixture. The mer/
fac isomerization appears to be nondissociative, as isomerization
of mer-8 proceeds in the presence of free PMe3-d9 or excess
CO to initially yield only fac-8-d0, that is, without incorporation
of labeled phosphine or a second CO on the time scale of
isomerization.
The unstable mer-8 exhibits two distinct phosphine environ-
ments in a 2:1 ratio that are characteristic of a cis,cis,trans (mer)
arrangement (1H NMR δ 1.29 (t) and 1.10 (d); 31P NMR δ -7.0
(d) and 15.7 (t)). The larger coupling of the Ru-H resonance
(δ -9.07, dt, JPH ) 73.8 and 29.1 Hz) establishes the mutually
trans relationship between the hydride and a phosphine and,
therefore, mutually trans position of CO and Si ligands. The
other isomer, fac-8, exhibits three phosphine environments (1H
NMR δ 1.123, 1.117, and 1.09; 31P NMR δ -11.2, -15.9, and
-22.0) consistent with the facial geometry. The multiplicity of
the NMR resonances of the hydride (1H NMR: ddd, JPH ) 67.1,
29.9, and 21.4 Hz), carbonyl (13C NMR: dt, JPC ) 79.3 and
9.0 Hz), and the silyl (29Si NMR: ddd, JPSi ) 74.8, 21.0, 11.3
Hz) ligands confirms facial geometry - each nonphosphine
ligand exhibits a single large scalar coupling constant due to a
trans phosphine. The IR spectrum exhibits strong bands for
classical RuH and terminal carbonyl ligands (ν(CO) ) 1932
cm-1; ν(RuH) ) 1858 cm-1).
Reactions with Dihydrogen and Synthesis of (PMe3)3Ru-
(SiR3)H3 Complexes. The ease of phosphine dissociation from
(PMe3)4Ru(SiR3)H complexes also manifests itself in rapid
reactions with dihydrogen under mild conditions. Thus, 1a and
1c react with dihydrogen to produce free PMe3 and (PMe3)3-
Ru(SiMe3)H3 (4a) or (PMe3)3Ru(SiEt3)H3 (4c) in quantitative
yield in <5 min at room temperature (eq 8). It is noteworthy
that reaction of 1a-d9 (labeled phosphine trans to silicon) with
hydrogen leads to formation of only 4a-d0 and free PMe3-d9 as
determined by 1H NMR. This regiospecificity strongly suggests
the reaction proceeds via the same intermediate as does the
phosphine exchange process.
The 18e-/16e- equilibrium also explains the coloration of
7a in solution. Saturated ruthenium(II) complexes such as 1a
and 6 are colorless as solids and in solution at room temperature,
whereas isolated 16e- (PMe3)3Ru(SiMe3)H, 2a, is brown-
red.29,30 However, solutions of the ostensibly 18e- 7a are red-
brown at room temperature and bleach to colorless in the
presence of 5 equiv of PMe3. Unfortunately, attempts to isolate
16e- (PMe3)3Ru(SiMe3)Me employing BPh3 and other phos-
phine sponges as described previously for 1a29,30 were not
successful and led to a plethora of decomposition products.
Reactions with CO and Synthesis of (CO)(PMe3)3Ru-
(SiMe3)H. Given the phosphine exchange reactions described
above, it is not surprising that 1a reacts rapidly with carbon
monoxide to yield 1 equiv of free PMe3 and mer-(CO)(PMe3)3-
Ru(SiMe3)H (mer-8), the monocarbonyl with CO trans to silicon
1
(eq 7). The reaction is quantitative by H NMR, but isolation
of mer-8 was thwarted by rapid isomerization at room temper-
ature to the facial isomer, fac-(CO)(PMe3)3Ru(SiMe3)H (fac-
8, eq 7).
In contrast, the dimethylsilyl derivative, 1d, is apparently
more stable than the corresponding trihydride complex, which
9
8046 J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003

Synthesis and Reactivity of Silyl Ruthenium Complexes
A R T I C L E S
could not be isolated. Treatment of 1d with 3 atm H2 in an
NMR tube generated only ca. 10% free PMe3. A new hydride
resonance at δ -9.7 was also observed, but reversion to 1d
occurred upon removal of hydrogen pressure.
1500 ms for 4a and 1025 ms for 4c), and has been confirmed
by a neutron diffraction study on 4a.42
Reactions of (PMe3)3Ru(SiR3)H3 with PMe3-d9 and D2. In
contrast to the rapid, regiospecific incorporation of labeled
phosphine into octahedral silyl complexes such as 7a and 1a,
no exchange was detected in the reaction of 4a with 12 equiv
of PMe3-d9 after 30 days at 25 °C. Incorporation of PMe3-d9
was observed following photolysis of 4a (350 nm, 10 °C.)
Although 4a does not undergo net exchange with labeled
phosphine, hydrogen dissociation does occur at room temper-
ature, but the equilibrium strongly favors the seven-coordinate
trihydride complex. Thus, treatment of 4a in neat PMe3 for 20
h at room temperature and removal of all volatiles (including
hydrogen) yields a mixture containing mainly unreacted 4a
(80%) and 20% of the tetrakis phosphine complex 1a (eq 10).
Furthermore, introduction of H2 onto neat PMe3 solutions of
1a generates 4a quantitatively very rapidly at room temperature.
A more general method for the synthesis of the Ru(IV)
trihydride silyls (4a-c) is the photolytic reaction (350 nm) of
an appropriate hydridosilane with the more readily available
cis-(PMe3)4RuH2, 9 (eq 9).1,33 Alternatively, 4b, 4c, and other
derivatives can be conveniently prepared in quantitative yield
by exchange of 4a with an excess of the appropriate hydridosi-
lane HSiR3′ at room temperature. As in the case of silyl
exchange with 1a, this equilibrium is sensitive to the relative
steric demands of the silyl ligands, but use of excess HSiR3′
and removal of HSiMe3 permits quantitative conversion in the
cases examined. For example, the reaction of 4a with HSiMe2-
CH2SiMe3 is complete in minutes, generating 4b and HSiMe3.20
The same equilibrium mixture is obtained by reaction of 4b
with HSiMe3. The reaction clearly favors the smaller silyl ligand
bound to ruthenium (Keq(298) ) 1.2 × 10-2). The exchange of
the bulkier HSiEt3 (ca. 1 equiv) with 4a is very slow and yields
a 4:1 mixture of 4a:4c after 2 weeks at room temperature.
Interestingly, the complementary reaction - the bulky silyl
trihydride 4c with 1 equiv of HSiMe3 at 25 °C - is also
extremely slow. The less hindered 4a was not detected after 2
days (<2%), and the equilibrium ratio was not achieved after 2
weeks (4a:4c ) 1:4).
Repeating this experiment in neat PMe3-d9 reveals that the
preponderance of labeled phosphine in 1a is found trans to the
SiMe3 group; <20% label is observed in the other three sites
combined after 20 h in neat PMe3-d9. Recall that slow secondary
scrambling of labeled phosphine was found in the studies of
1a described above.
1
The ruthenium hydride resonances in 4a-c in the H NMR
The ease of H2 elimination from 4a is clearly illustrated by
the reaction with D2, where the rate is not masked by an
unfavorable equilibrium. Treatment of 4a with ca. 3 atm D2 at
room temperature for <5 min yields H2 and 4a-d2 and <5%
HD, consistent with a mechanism requiring the initial reductive
elimination of H2. Naturally, HD is observed after longer
reaction times, produced by subsequent elimination from 4a-
d2. Compound 4b also exchanges Ru-H positions with D2
rapidly at room temperature, but H/D exchange occurs only very
slowly with the triethylsilyl complex, 4c. Only trace amounts
appear as a complicated pattern with two sharp peaks centered
within a broad multiplet at ca. δ -10. Although the line shape
may suggest a dynamic process, complexes 4a and 4b are not
fluxional on the NMR time scale and show only a slight
broadening of the RuH resonances as the temperature is lowered
to 190 K. The line shape is in fact due to the three-fold
symmetry of the fac-RuH3P3 fragment and resultant AA′A′′XX′X′′
spin system. Overall, the NMR features closely resemble those
of previously described L3M(ER3)H3 complexes.20,34-41 The
spectroscopic assignment was confirmed by single-crystal X-ray
diffraction analysis of 4a, vide infra, and 4b.20 The absence of
appreciable nonclassical H‚‚‚H interactions is suggested by the
IR spectra (Nujol: ν(RuH) ) 1890 cm-1 for 4a and 1899 cm-1
for 4c), and relaxation constants (200 MHz, C6D6, 25 °C, T1 )
1
of H2 and HD are detected by H NMR after days at 25 °C.
Another possible reaction of 4a - Me3SiH elimination - is
extremely slow. Reactions of 4a in neat PMe3 at room
temperature do not yield detectable quantities of the cis-
dihydride complex (9). However, thermolysis of 4a at 70 °C in
the presence of ∼2 equiv of PMe3 results in the slow, but
quantitative, conversion to 9 (115 h, eq 10). Compound 9 is
virtually inert toward reactions with Me3SiH or phosphine
exchange below ca. 100 °C. Thus, Si-H elimination from 4a
cannot be occurring to a significant extent at 25 °C, even though
H-H elimination is rapid.
(33) Montiel-Palma, V.; Perutz, R. N.; George, M. W.; Jina, O. S.; Sabo-Etienne,
S. Chem. Commun. 2000, 1175-1176.
(34) Gilbert, S.; Knorr, M.; Mock, S.; Schubert, U. J. Organomet. Chem. 1994,
480, 241-254.
(35) Hubler, K.; Hubler, U.; Roper, W. R.; Schwerdtfeger, P.; Wright, L. J.
Chem.-Eur. J. 1997, 3, 1608-1616.
(36) Knorr, M.; Gilbert, S.; Schubert, U. J. Organomet. Chem. 1988, 347, C17-
C20.
(37) Mohlen, M.; Rickard, C. E. F.; Roper, W. R.; Salter, D. M.; Wright, L. J.
J. Organomet. Chem. 2000, 593-594, 458-464.
CH Activation and SiH Elimination in the (PMe3)4Ru-
(SiR3)H Complexes. Another important aspect of the chemistry
of these silyl ruthenium complexes is the ability to activate CH
(38) Rickard, C. E. F.; Roper, W. R.; Woodgate, S. D.; Wright, L. J. J.
Organomet. Chem. 2000, 609, 177-183.
(39) Schubert, U.; Gilbert, S.; Mock, S. Chem. Ber. 1992, 125, 835-837.
(40) Burn, M. J.; Bergman, R. G. J. Organomet. Chem. 1994, 472, 43-54.
(41) Feldman, J. D.; Peters, J. C.; Tilley, T. D. Organometallics 2002, 21, 4065-
4075.
(42) Koetzle, T.; Wu, P., unpublished results.
9
J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003 8047

A R T I C L E S
Dioumaev et al.
bonds. Activation of aryl CH bonds by ruthenium(0) complexes
is not that unusual, but CH bond addition to a 16e- ruthenium-
(II) species such as (PMe3)3Ru(SiR3)H is of particular interest
as the resultant species can potentially undergo elimination of
CH-, SiC-, SiH-, or HH-bonds. The course of these elimination
reactions has a direct bearing on catalytic processes such as
functionalization of CH bonds, hydrosilylation, dehydrogenative
coupling, and transfer dehydrogenative coupling of silanes to
oligocarbosilanes, and, therefore, deserves closer scrutiny.
Thermolysis of 1a in benzene at 65 °C leads to CH bond
activation of the solvent and formation of cis-(PMe3)4Ru(H)-
(Ph), 10 (eq 11). The trimethylsilane formed must be removed
periodically to drive the equilibrium and to avoid subsequent
dehydrocoupling to carbosilane (HSiMe2CH2SiMe3) and dihy-
drogen, both of which lead to additional ruthenium products.
The phenyl derivative 10 was isolated in 79% yield after ∼14
days at 65 °C, and the spectroscopic parameters match those
reported by Bergman and co-workers.43
Figure 1. An ORTEP drawing of (PMe3)4Ru(SiMe3)H, 1a (30% thermal
ellipsoids).
Thermolysis of 1a in C6D6 at 45 °C for 15 h leads to extensive
H/D exchange between C6D6 and 1a, and only trace amounts
of deuterated 10. Approximately 18% deuteration of the SiMe3
group and 8% of the PMe3 ligands was observed. After 16 h at
55 °C, <10% 10 was produced, but the 1H NMR signal for the
SiMe3 ligand of 1a had decreased by 80%, and those for the
PMe3 ligands had decreased by 40%. Corresponding increases
in intensity were observed in the 2H NMR spectra. In a slower
process, 5 (∼10% after 90 h at 55 °C) is also produced by
intramolecular CH activation. Addition of free PMe3 (3 equiv)
completely inhibits both the H/D exchange and the formation
of 10, but does not affect production of 5 (∼8% in 90 h at 55
°C). Compound 5 is the only new ruthenium product observed
when thermolysis of 1a is performed in cyclohexane-d12, and
no H/D scrambling is observed, consistent with a clear prefer-
ence for an intramolecular activation of a primary CH bond
over intermolecular activation of the secondary aliphatic CD.
Thermolysis of the triethylsilyl complex 1c in benzene also
produces the phenyl complex 10, but the reaction is 2-3× faster
than that for the less hindered 1a (eq 11). This presumably
reflects both greater phosphine lability in 1c and a more
favorable equilibrium for formation of 10 from 1c.
Solid-State Structures. The structures of 1a, 1b, 1d, and 7a
(Figures 1-4, Tables 1-5) were determined by single crystal
X-ray diffraction to be approximately octahedral. The steric
congestion causes the two ostensibly trans PMe3 ligands to tilt
toward the smallest groups (RuH in 1a and 1b, RuH and
RuSiMe2H in 1d, and RuMe in 7a; PsRusP ) 153.58(3)-
167.3(1)°). The other two ligands cis to the smallest groups do
not distort significantly in 1a, 1b, and 7a; that is, the silyl and
trans phosphine remain in essentially unperturbed octahedral
positions ( PsRusSi ) 176.7(5)-177.63(3)°). The geometry
Figure 2. An ORTEP drawing of (PMe3)4Ru(SiMe2CH2SiMe3)H, 1b (30%
thermal ellipsoids).
can be described as a distortion of the octahedron toward trigonal
bipyramidal, tbp (ignoring the hydride or methyl ligands), with
axial positions occupied by silyl and phosphine ligands. The
structure of 1d, on the other hand, is less crowded, and the steric
pressure is relieved by a more uniform distortion of all ligands
from the ideal octahedral positions. It is somewhat surprising
that the relief of steric crowding in the hydrides 1a, 1b, and 1d
is not achieved by a closer approach of the hydride and silyl
ligands and formation of a σ-SiH complex, as is found in many
higher valent ruthenium silyl hydride complexes. This may be
disfavored by the very electron-rich, formally Ru(0) centers that
would result in the hypothetical (PMe3)4Ru(η2-HSiR3). In any
event, the structures of 1a, 1b, 1d, and 7a can be described as
classical with all bond distances and angles within normal
ranges. In all cases, the Ru-P distances trans to Si, H, and CH3
ligands are the longest in a given complex. Such ground-state
elongation of metal-ligand bonds trans to a strong σ-donor such
as a silyl, hydride, and alkyl (trans influence) is well docu-
mented, and Nolan has correlated Ru-P bond length with
(43) Hartwig, J. F.; Andersen, R. A.; Bergman, R. G. J. Am. Chem. Soc. 1991,
113, 6492-6498.
9
8048 J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003

Synthesis and Reactivity of Silyl Ruthenium Complexes
A R T I C L E S
Table 1. Crystal Data for 1a, 1b, 1d, 4a, and 7a
compound
1a
1b
1d
4a
7a
formula
formula weight
C15H46P4RuSi
479.68
C18H54P4RuSi2
551.74
C14H44P4RuSi
465.56
C12H39P3RuSi
405.52
C16H48P4RuSi
493.58
crystal system
space group
color
Z
monoclinic
C2/c (No. 15)
colorless
8
monoclinic
P21/n (No. 14)
colorless
4
monoclinic
P21/n (No. 14)
colorless
4
monoclinic
P21/n (No. 14)
colorless
4
monoclinic
C2/c (No. 15)
colorless
8
a, Å
b, Å
c, Å
â, deg
V, Å3
T, K
R
wR
15.5859(2)
10.8807(1)
29.6653(4)
103.260(4)
4896.68(10)
210
12.3876(1)
23.6973(3)
10.2102(1)
97.801(1)
2969.49(5)
227
9.113(1)
29.260(5)
9.353(2)
106.13(1)
2396(1)
227
14.220(1)
9.802(3)
16.265(4)
96.67(1)
2252(1)
296
15.2642(2)
10.93880(10)
31.0714(4)
103.9820(10)
5034.34(10)
200
0.0444a
0.0512a
0.049b
0.034b
0.0653a
0.1024a
0.1035a
0.060b
0.039b
0.1295a
GOF
1.125
1.182
1.807
1.063
1.215
2
2
a All data used; R ) ∑(||Fo| - |Fc||)/∑|Fo|; wR ) {∑w(F02 - Fc )2/∑w(F0 )2}1/2
.
b F 2 > 3.0σ(F 2) data used; R ) ∑||Fo| - |Fc||/∑|Fo|; wR ) {∑w(|Fo|
2
- |Fc|)2/∑w|Fo| }1/2
.
Figure 3. An ORTEP drawing of (PMe3)4Ru(SiMe2H)H, 1d (30% thermal
ellipsoids).
Table 2. Selected Bond Distances and Nonbonding Contacts (Å)
and Angles (deg) in 1a
Figure 4. An ORTEP drawing of (PMe3)4Ru(SiMe3)Me, 7a (30% thermal
ellipsoids).
Ru-P1
Ru-P2
2.3663(9)
2.3398(8)
Ru-P3
Ru-P4
2.3561(9)
2.3060(8)
Ru-Si1
Ru-H1
2.4630(9)
1.59(4)
Table 3. Selected Bond Distances and Nonbonding Contacts (Å)
and Angles (deg) in 1b
P2-Ru-P1
90.76(3)
92.91(3)
93.18(3)
177.63(3)
88.08(3)
89.38(3)
86.95(3)
87.5(10)
P2-Ru-P3
103.81(3)
153.58(3)
102.07(3)
90.2(10)
75.0(14)
176.7(10)
78.8(14)
Ru-P1
Ru-P2
2.3683(9)
2.3198(8)
Ru-P3
Ru-P4
2.3591(8)
2.3222(8)
Ru-Si1
Ru-H1
2.4796(9)
1.61(4)
P3-Ru-P1
P4-Ru-P1
P1-Ru-Si1
P2-Ru-Si1
P3-Ru-Si1
P4-Ru-Si1
Si1-Ru-H1
P2-Ru-P4
P4-Ru-P3
P1-Ru-H1
P2-Ru-H1
P3-Ru-H1
P4-Ru-H1
P2-Ru-P1
90.61(3)
92.08(3)
93.44(3)
177.36(3)
88.47(3)
90.54(3)
86.31(3)
86.0(12)
P2-Ru-P3
102.74(3)
153.86(3)
102.90(3)
91.4(12)
77.8(12)
176.5(12)
76.3(12)
P3-Ru-P1
P4-Ru-P1
P1-Ru-Si1
P2-Ru-Si1
P3-Ru-Si1
P4-Ru-Si1
Si1-Ru-H1
P2-Ru-P4
P4-Ru-P3
P1-Ru-H1
P2-Ru-H1
P3-Ru-H1
P4-Ru-H1
dissociation enthalpies in other complexes. It is, therefore, not
surprising that the Ru-P distances in 1a, 1b, and 7a are
consistent with the qualitative trends for the kinetic and
thermodynamic lability of the phosphines in these complexes.
The structure of the trihydride silyl 4a was determined by
the single-crystal X-ray diffraction method to be a seven-
coordinate complex composed of a pseudo octahedral fac-
(PMe3)3RuH3 unit with the silyl group capping the face defined
by the three hydride ligands (Figure 5, Tables 1 and 6). The
hydride ligands were located and refined with isotropic thermal
parameters. The Si-C and Ru-P bonds are eclipsed, and the
hydride ligands are staggered with respect to the Si methyls
and P atoms, yielding C3 molecular symmetry. This ligand
arrangement is typical for such L3M(ER3)H3 (E ) Si or Sn)
complexes.20,34-39,41,44 All heavy atom distances and angles are
within the expected ranges. The formally nonbonded contacts
(44) Procopio, L. J. Ph.D. Thesis, University of Pennsylvania, 1991.
9
J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003 8049

A R T I C L E S
Dioumaev et al.
Table 6. Selected Bond Distances and Nonbonding Contacts (Å)
and Angles (deg) in 4a
Ru-P1
Ru-P2
Ru-P3
2.317(1)
2.320(1)
2.323(1)
Ru-H1
Ru-H2
Ru-H3
1.488(43) Si‚‚‚H1
1.637(50) Si‚‚‚H2
1.431(45) Si‚‚‚H3
2.228(42)
2.179(48)
2.128(48)
H1‚‚‚H2 2.362(65) H2‚‚‚H3 2.287(66) H1‚‚‚H3 2.387(54)
Ru-Si 2.376(1)
P1-Ru-P2
98.8(1)
98.8(1)
99.8(1)
P1-Ru-Si
P2-Ru-Si
P3-Ru-Si
P1-Ru-H1
P1-Ru-H2
P1-Ru-H3
P3-Ru-H1
P3-Ru-H2
P3-Ru-H3
H2-Ru-H3
118.5(1)
118.0(1)
119.0(1)
75.2(19)
79.3(16)
173.5(16)
82.7(17)
178.5(15)
84.9(18)
95.9(24)
P2-Ru-P3
P1-Ru-P3
Si-Ru-H1
Si-Ru-H2
Si-Ru-H3
P2-Ru-H1
P2-Ru-H2
P2-Ru-H3
H1-Ru-H2
H1-Ru-H3
65.7(18)
62.5(15)
62.1(17)
174.0(18)
80.2(14)
75.8(16)
98.3(23)
110.1(25)
Discussion
Phosphine Lability in (PMe3)4Ru(SiR3)H Complexes.
Reactions of 1a with PMe3-d9, CO, and H2 yield kinetically
regiospecific trapping products, which are probably due to
regiospecific phosphine dissociation from the site trans to Si
and formation of the previously described five-coordinate 16e-
intermediate with an empty site trans to Si.29,30 It is possible
that ligand dissociation might occur from one site to yield a
fluxional intermediate, which is preferentially trapped at a
different site. Fortunately, in the case of 1a, this complication
can be definitively excluded. Reaction of 1a-d9 (labeled phos-
phine trans to silicon) with H2 yields only 4a-d0, proving that
phosphine dissociation is regiospecific. Furthermore, the reaction
of 1a with CO yields a kinetic product, mer-8 (CO trans to
silicon), proving that trapping with CO is also selective.
Therefore, the regioselectivity of phosphine exchange in the site
trans to silicon can indeed be taken at face value. A facile
phosphine exchange in the position trans to silicon is also
observed for the structurally related complexes 7a and 1c. It is
reasonable to assume that the selective trapping is also due to
selective dissociation, as is the case with the more thoroughly
studied 1a.
Figure 5. An ORTEP drawing of (PMe3)3Ru(SiMe3)H3, 4a (30% thermal
ellipsoids).
Table 4. Selected Bond Distances and Nonbonding Contacts (Å)
and Angles (deg) in 1d
Ru-P1
Ru-P2
2.338(2)
2.328(2)
Ru-P3
Ru-P4
2.347(1)
2.316(2)
Ru-Si1 2.426(1)
Ru-H1
1.524(58)
Si1-H2 1.523(79) Si1‚‚‚H1 2.675(50)
P2-Ru-P1
P3-Ru-P1
P4-Ru-P1
P1-Ru-Si1
P2-Ru-Si1
P3-Ru-Si1
P4-Ru-Si1
Si1sRu-H1
93.4(1)
100.2(1)
95.1(1)
104.1(1)
86.0(1)
155.5(1)
82.8(1)
81.7(18)
P2-Ru-P3
95.7(1)
167.3(1)
92.0(0)
173.9(19)
85.2(22)
74.1(18)
87.3(23)
P2-Ru-P4
P4-Ru-P3
P1-Ru-H1
P2-Ru-H1
P3-Ru-H1
P4-Ru-H1
Table 5. Selected Bond Distances and Nonbonding Contacts (Å)
and Angles (deg) in 7a
Ru1-C16 2.215(6)
Ru1-P1 2.3680(13) Ru1-P3 2.3181(12)
Ru1-Si1 2.4681(14) Ru1-P2 2.3400(13) Ru1-P4 2.3813(13)
P1-Ru1-P2
P2-Ru1-P3
P1-Ru1-P3
P1-Ru1-P4
P2-Ru1-P4
P3-Ru1-P4
C16-Ru1-P1
C16-Ru1-P2
102.00(5)
97.49(5)
159.74(5)
90.47(5)
91.63(5)
94.47(5)
80.2(2)
C16-Ru1-P3
C16-Ru1-P4
C16-Ru1-Si1
P1-Ru1-Si1
P2-Ru1-Si1
P3-Ru1-Si1
P4-Ru1-Si1
80.3(2)
89.4(2)
89.9(2)
86.90(5)
89.11(5)
87.94(5)
177.36(5)
Exchange of the other phosphine sites in 1a, 1c, and 7a with
PMe3-d9 is always slower for a given complex than the exchange
of the site trans to Si. The absolute rates, however, vary greatly
between compounds and increase with steric congestion at the
Ru center. Thus, the rate of exchange with PMe3-d9 of the other
phosphines (i.e., not trans to silicon) follows this trend: 1a
(days) < 1c < 7a (minutes). Possible mechanisms for these
exchanges include: (1) the slow dissociation of the less labile
phosphines and immediate trapping; (2) trapping of minor
geometrical isomers of a fluxional five-coordinate 16e- inter-
mediate (e.g., 2a); and (3) slow intramolecular rearrangement
of the initial octahedral product. Although intramolecular
isomerization is not common in octahedral geometries, it has
been observed in some related complexes (PR3)4MH2 (M ) Fe,
Ru, Os)46-49 and (CO)4M(EMe3)2 (M ) Fe, Ru, Os; E ) Si,
177.6(2)
between silicon and the ruthenium hydrides (2.13-2.23(5) Å)
may indicate some delocalized bonding in the Ru(Si)H3 frag-
ment, albeit not as strong as typical agostic interactions (ca.
1.7-1.8 Å). The H‚‚‚H separations (2.29-2.39(7) Å) are much
longer than those found in η2-H2 and related complexes.45 This
is in accord with the solution spectroscopic data (T1(25 °C) )
1
1500 ms at 200 MHz; IR (Nujol): ν(RuH) ) 1890 cm-1; H-
{31P} NMR (C6D6): JSiH < 25 Hz). Evidence for nonclassical
EH (E ) Si or Sn) bonding, but not dihydrogen complexation,
is also observed in other L3M(ER3)H3 complexes, both in the
solid state (X-ray) and in solution (JEH and T1 relaxation
times).35,37-39
(46) Meakin, P.; Muetterties, E. L.; Tebbe, F. N.; Jesson, J. P. J. Am. Chem.
Soc. 1971, 93, 4701.
(47) Tebbe, F. N.; Meakin, P.; Jesson, J. P.; Muetterties, E. L. J. Am. Chem.
Soc. 1970, 92, 1068.
(48) Meakin, P.; Guggenberger, L. J.; Jesson, J. P.; Gerlach, D. H.; Tebbe, F.
N.; Peet, W. G.; Muetterties, E. L. J. Am. Chem. Soc. 1970, 92, 3482.
(49) Muetterties, E. L. Acc. Chem. Res. 1970, 3, 266.
(45) Heinekey, D. M.; Oldham, W. J., Jr. Chem. ReV. 1993, 93, 913-926.
9
8050 J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003

Synthesis and Reactivity of Silyl Ruthenium Complexes
A R T I C L E S
Ge, Sn, Pb),50 as well as in the mer to fac rearrangement of the
carbonyl compound mer-8 to fac-8 in the presence of PMe3-d9
(half-life of ca. 35 min at room temperature, vide supra).
However, the increase in scrambling rates with increasing steric
crowding in 1a, 1c, and 7a would be also consistent with a
dissociative mechanism.
Figure 6. Possible high-energy meridianal isomers of compound 4.
The ability of silicon ligands to exert large trans effects has
been noted previously in metal silyls,51,52 including carbonyl
complexes of ruthenium and osmium.15-17,21 It is, therefore, not
surprising that the ligands trans to Si in (PMe3)4Ru(SiR3)X
complexes are considerably more labile than those adjacent to
silicon. The magnitude of the effect is impressive, with most
of the substitution reactions in the present studies approaching
the fast regime on the NMR time scale. Alkyl and hydride
groups are also strong trans directing ligands, but the phosphine
trans to silicon in 1a, 1c, and 7a is always the most labile by a
substantial margin. The magnitude of silyl group trans effect
can be put into perspective by considering two related complexes
that do not contain silyl ligands: the cis dihydride 9 and the
cis dimethyl complex 6. Although hydride and methyl are
generally considered strong trans directing ligands in their own
right, 9 is quite inert to phosphine exchange below ca. 100 °C,
and 6 exchanges with labeled phosphine only slowly at room
temperature (t1/2 ≈ 10 h). Thus, it is clear that steric and
electronic effects combine to produce a dramatic, regioselective
labilization of phosphines trans to silicon in octahedral silyl
complexes.
Scheme 1
In addition to the kinetic consequences, silyl ligands also
induce weakening and elongation of the ruthenium-phosphine
bonds trans to silicon in the ground-state structures. This “trans
influence” is well established for metal silyl complexes,
including those of Ru and Os.21,22 The trans effect and trans
influence are closely connected. Analysis of this relation is
particularly instructive for series of structurally similar com-
plexes such as 1a, 1b, and 7a. In this series, the longest Ru-P
bond within a complex is always trans to Si, and the second
longest is trans to H. The most sterically crowded, 7a, however,
offers an interesting exception. The Ru-P bond trans to Si bond
in 7a is still the longest, and the bond trans to Me is also
elongated, but not as much as one of the mutually trans Ru-P
bonds. It is noteworthy that the relative elongation of Ru-P
bonds in the ground state (trans influence) clearly parallels the
lability of the ligands in question (trans effect). Thus, 1a has
only one Ru-P distance longer than 2.36 Å and exhibits
exchange with PMe3-d9 selectively in that position. Indeed,
exchange of this phosphine approaches the fast exchange regime
on the NMR time scale at room temperature. Dissociation of
the phosphine trans to silicon in 7a is even more rapid, and
this ruthenium-phosphorus distance is extremely long in the
solid state (2.3813(13) Å). In this case, however, exchange of
the other sites is relatively fast (minutes), consistent with another
long Ru-P bond (2.3680(13) Å) observed in the solid-state
structure. As a thermodynamic effect, the trans influence is also
manifest in the position of dissociation equilibria and, therefore,
in the abundance of the 16e- species produced by ligand
dissociation. Indeed, the longest Ru-P bond (trans to silicon
in 7a) correlates with the greatest concentration of (PMe3)3Ru-
(SiR3)X in solution.
Lability of Dihydrogen in (PMe3)3Ru(SiR3)H3 Complexes.
Elimination of dihydrogen from 4a (and 4b) is fast as is evident
from the reactions with D2, but the equilibrium lies strongly
toward the trihydride complex in neat PMe3, and it is difficult
to increase the concentration of 1a (and 1b) even with the
dynamic removal of the H2 produced. Significantly, the traces
of 1a-d9 produced by reaction of PMe3-d9 with 4a exhibit the
label in the position trans to silyl, even though the silyl ligand
and all three hydrides are mutually cis in 4a. Clearly, a
rearrangement must occur, either prior to HH elimination from
the seven-coordinate 4a or subsequently in the five-coordinate
intermediate, 2a. The fact that HH elimination is very slow from
the bulkier triethylsilyl complex 4c is not consistent with a rate-
limiting dissociative process, but rather suggests limiting
isomerization to a sterically less favorable seven-coordinate
species from which HH elimination occurs. One such species
that can be envisioned would resemble the 16e- intermediate
2a, but with two classical hydride ligands, or a single coordi-
nated dihydrogen molecule filling the open coordination site
(Figure 6). Presumably, the strong trans effect/influence of the
silyl ligand would enhance HH bond formation and favor
dissociation.
HH, SiH, CH, and SiC Reductive Eliminations from
Seven-Coordinate Species. As shown above, 16e-, five-
coordinate species (PMe3)3Ru(SiR3)X are readily accessed from
silyl complexes such as 1, 4, and 7. The 16e- species, in turn,
reversibly add SiH bonds to form 18e-, seven-coordinate
(PMe3)3Ru(SiR3)2(X)H complexes (X ) H20 or Me, Scheme
1). Subsequent H-Si elimination and reassociation of phosphine
or dihydrogen results in silyl group exchange (e.g., 1a and 1b,
7a and 7b, or 4a and 4b; Scheme 1). Silyl exchange is quite
(50) Vancea, L.; Pomeroy, R. K.; Graham, W. A. G. J. Am. Chem. Soc. 1976,
98, 1407.
(51) Azizian, H.; Dixon, K. R.; Eaborn, C.; Pidcock, A.; Shuaib, N. M.; Vinaixa,
J. J. Chem. Soc., Chem. Commun. 1982, 1020.
(52) Auburn, M. J.; Stobart, S. R. Inorg. Chem. 1985, 24, 318.
9
J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003 8051

A R T I C L E S
Dioumaev et al.
Scheme 2
selective at room temperature, and ruthenium products arising
from H2, CH4, Me-SiMe2CH2SiMe3, or Me-SiMe3 elimination
pathways are not produced as readily. Again, bulkier silyl
ligands inhibit the exchange rate, and reaction of 4c with
HSiMe3 to produce the less hindered 4a is very slow. This can
be attributed to both the slow loss of H2 from 4c (vide supra)
as well as the steric congestion in the requisite bis(silyl)
dihydride intermediate, (PMe3)3Ru(SiMe3)(SiEt3)(H)2. The latter
factor is responsible for the sluggishness of the reverse reaction,
4a with HSiEt3, as H2 loss from 4a is facile.
Elimination of H2 from (PMe3)3Ru(SiR3)(SiR3′)H2 has also
been observed and leads to (PMe3)3Ru(SiR3)(SiR3′) species.20
However, H2 loss is much less favorable than silane loss, except
in the case of small silanes such as PhSiH3. Thus, productive
dehydrogenative chemistry in these systems requires fairly high
temperatures, or the removal of H2 with hydrogen acceptors
such as an olefin or even other 16e- metal complex species.
The other possible dissociative process from (PMe3)3Ru(SiR3)-
(SiR3′)H2 - SiSi elimination to produce R3SiSiR3′ - has not
been observed in these systems.
An interesting situation arises in the reaction of a hydridosi-
lane with a complex containing an alkyl ligand, for example,
7a + HSiMe3, as the intermediate (PMe3)3Ru(SiMe3)2(Me)H
species can also undergo elimination of C-H and Si-C bonds.
The SiH elimination is the fastest process, as evidenced by
exchange with free silane with 7 on the NMR time scale at
room temperature. At higher temperatures, however, both CH
and SiC eliminations occur irreversibly, producing CH4 and
SiMe4. In the case of the reaction of dimethyl complex 6 with
silane, two methyl groups are ultimately extruded, and the CH4
and SiMe4 ratio of ca. 7:1 indicates a kinetic preference for
CH elimination.
sively studied and is actually slightly more stable than the 16e-
2a, although neither are stable in the presence of PMe3 and revert
to 1a.29,30
Although 5 could hypothetically arise from Me3SiH loss from
(PMe3)2Ru(η2-CH2PMe2)(SiMe3)(H)2 and association of phos-
phine, this cannot be the primary path, as formation of 5 is not
inhibited by added PMe3, whereas H/D scrambling is. Thus,
consistent with the extensive studies by Bergman,43 Flood,53,54
and their co-workers, 5 appears to arise from intramolecular
C-H activation in (PMe3)4Ru(0), whereas intermolecular C-H
activation of benzene involves a Ru(II)/Ru(IV) reaction mani-
fold. Formation of the phenyl complex 10 proceeds by the
elimination of HSiMe3 from (PMe3)3Ru(SiMe3)(Ph)(H)2 and
trapping with phosphine, a process that is relatively slow as
compared to benzene C-H (C-D) elimination and resultant
H/D exchange. This is analogous to the H/D exchange between
10 and C6D6 reported by Bergman and shown to proceed via
(PMe3)3Ru(Ph)(C6D5)(D)H. Similar pathways were described
earlier by Flood and co-workers in the thermal conversion of
(PMe3)4Os(CH2CMe3)H into Os analogues of 5 and 10.
Finally, it is worth noting that the faster rate for formation
of 10 observed with the more hindered 1c likely reflects both
greater phosphine lability as compared to 1a and a greater
preference for HSiEt3 loss from (PMe3)3Ru(SiEt3)(Ph)(H)2.
Relevance to Catalytic Dehydrocoupling and CH Func-
tionalization. The reactivity trends reported above have direct
implications for the SiC bond forming catalysis and allow for
a number of conclusions. One of the key elements for any
successful catalytic dehydrogenative (or dealkanative) SiC bond
making process is the aptitude for the elimination of HH (or
CH) and SiC bonds from the metal center. The present study
illustrates that this aptitude can be greatly enhanced when the
eliminating fragment is positioned trans to a silyl group, for
example, a strong trans directing ligand. In octahedral or capped
octahedral seven-coordinate complexes, simple geometrical
Activation of CH Bonds and H/D Exchange. The formation
of the phenyl hydride complex 10 from 1a and benzene could
proceed via C-H activation by at least two different 16 e-
intermediates: the Ru(0) complex (PMe3)4Ru formed by H-Si
elimination or the Ru(II) complex 2a formed by phosphine
dissociation. The facts that deuterium exchange into 1a is much
faster than formation of 10 and that H/D exchange and formation
of 10 are inhibited by added PMe3 strongly suggest that
phosphine dissociation is required. However, the formation of
the intramolecular C-H activation product 5 is independent of
phosphine concentration, and this species is produced by a
separate (and minor) pathway involving the (PMe3)4Ru species
(Scheme 2, Path A).
Reaction of 2a with benzene yields the 18e- (PMe3)3Ru-
(SiMe3)(Ph)(H)2, as shown in Scheme 2, Path B. Note that
analogues of this formally Ru(IV) intermediate, bis(silyl)-
dihydride complexes (PMe3)3Ru(SiMe3)2(H)2, have been isolated
and structurally characterized.20 The intermolecular CH oxida-
tive addition of benzene to 2a is rapidly reversible, which
accounts for the faster rate of deuterium incorporation than
formation of 10 in the case of reactions run in C6D6. Isotopic
exchange would initially occur at the Ru-H position of 2a (and
hence 1a), but subsequent and fast intramolecular C-H activa-
tion would lead to scrambling into the SiMe3 and PMe3 ligands
via (PMe3)3Ru(η2-CH2SiMe2)(D)(H) (3a-d1) and (PMe3)2Ru-
(η2-CH2PMe2)(SiMe3)(D)(H). The former (3a) has been exten-
(53) Desrosiers, P. J.; Shinomoto, R. S.; Flood, T. C. J. Am. Chem. Soc. 1986,
108, 7964-7970.
(54) Desrosiers, P. J.; Shinomoto, R. S.; Flood, T. C. J. Am. Chem. Soc. 1986,
108, 1346-1347.
9
8052 J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003

Synthesis and Reactivity of Silyl Ruthenium Complexes
A R T I C L E S
preferred geometry for a given complex. For example, the very
stable ground-state structures of 4a-c feature fac-phosphines,
and this very stability inhibits catalytic chemistry in the system.
It is, however, possible to enforce a meridianal structural motif
by using chelating “pincer” phosphine ligands to avoid formation
of excessively stable resting states for the catalyst. Another
advantage of a mer-tris-phosphine ligand is that phosphine
dissociation from such a rigid chelate would be very unlikely.
We have recently observed facile phosphine dissociation from
2a and formation of a plethora of bisphosphine intermediates.20
The relevance of these species to the catalytic cycle is not
known, but can be probed by the use of tridentate ligands. It
would, therefore, appear promising for a number of reasons to
attempt the use of “pincer” trisphosphine (or other tris-donor)
ligands in dehydrogenative catalysis. These studies are currently
underway and will be reported in the future.
Figure 7. Possible high-energy meridianal isomer of (PMe3)3Ru(η2-CH2-
SiMe2)(SiMe3)H with mutually trans silyl and agostic SiH ligands to
promote SiC elimination.
arguments dictate that the two eliminating ligands and the trans
directing silyl will be meridianal; thus, the remaining ligands
must also be roughly meridianal. Indeed, such an orientation is
found in the previously studied mer-(PMe3)3Ru(SiMe3)2(H)2,20
which is the true intermediate of the facile intermolecular SiH
exchange observed for 1a. This geometry is shared by those
(PMe3)3Ru(SiR3)2(H)2 (SiR3 ) SiH2Ph and SiPh2H) complexes
that exhibit fast SiH elimination.20 When the meridianal structure
is not accessible - such as in the chelating bis(silyl) complex
(PMe3)3Ru(SiMe2CH2CH2SiMe2)(H)2 - the SiH elimination is
a very slow process.20 It is reasonable to assume that SiH
exchange in 1b-c, 4a-c, and 7a-c (and CH/SiC eliminations
in a reaction of 7a-c with HSiR3) also follows a path via mer-
(PMe3)3Ru(SiR3)(SiR3′)(X)H intermediates (Scheme 1, X ) H
or Me). Note that the structures of these seven-coordinate
intermediates are such that pairs of SiR3 ligands are inequivalent,
and elimination of only one type of SiR3 group would be
facilitated by a trans silyl. However, the previously reported
dynamic exchange within pairs of H and SiR3 ligands in these
seven-coordinate species20 yields accessible geometries for
elimination of any SiH or SiC combination from a position trans
to another silicon. The relative rate of SiC elimination remains
slower than SiH elimination, due to the less favorable orbital
overlap between the eliminating ligands (i.e., directional p-orbital
of carbon vs spherical s-orbital of hydrogen). HH elimination
from 4a-c also proceeds via the mer isomer with HH trans to
Si (Figure 6), and similar hypothetical structures can be drawn
for HH and CH eliminations from (PMe3)3Ru(SiR3)(SiR3′)(X)H
(X ) H or Me).
It is noteworthy that the meridianal phosphine arrangement
is also favorable for the 16e- 2a,29,30 a true intermediate in the
SiC bond forming catalysis. The impact of meridianal phos-
phines is less obvious for the silyl silene species (PMe3)3Ru-
(η2-CH2SiMe2)(SiMe3)H, a long-postulated but never observed
catalytic intermediate.1,55 A facial phosphine arrangement was
found for the ground state of the chemically related dihydride
3a;29,30 however, elimination could well occur from a slightly
higher energy mer isomer. Indeed, both fac and mer isomers
(95:5) were found for the hydrido chloride analogue, (PMe3)3-
Ru(η2-CH2SiMe2-H)Cl.56 The fine aspects of bonding might
be somewhat different between the latter and 3a, as the chloride
derivative is better described as a â-agostic SiH complex rather
than silene hydride. However, in a more general sense, existence
of a mer-(PMe3)3Ru(η2-CH2SiMe2-H)Cl suggests a possibility
of a mer isomer for the hydride derivative (PMe3)3Ru(η2-CH2-
SiMe2)(SiMe3)H as well. For example, an isomer of (PMe3)3-
Ru(η2-CH2SiMe2)(SiMe3)H with a mutually trans arrangement
of silyl and “agostic SiH” ligands appears very reasonable on
steric grounds (Figure 7) and should favor SiC elimination.
Although a meridianal tris-phosphine geometry is associated
with faster rates of reductive elimination, this is not always the
Conclusions
Silyl complexes cis-(PMe3)4Ru(SiR3)H (SiR3 ) SiMe3, 1a;
SiMe2CH2SiMe3, 1b; SiEt3, 1c; SiMe2H, 1d) and cis-(PMe3)4-
Ru(SiR3)Me (SiR3 ) SiMe3, 7a; SiMe2CH2SiMe3, 7b) adopt
octahedral geometries in solution and the solid state with
mutually cis silyl and hydride (or silyl and methyl) ligands. The
longest Ru-P distance within each of the structurally character-
ized complexes (1a, 1c, 1d, and 7a) is always trans to Si,
reflecting the strong trans influence of silicon. Such phosphine
ligands positioned trans to Si exhibit regioselective dissociation
at a rate approaching the NMR time scale (trans effect of Si).
In 7a, the trans effect and trans influence are so strong that an
equilibrium concentration of dissociated phosphine is detectable
(∼5%) in solution of pure 7a. Although dissociation of phos-
phine in 1a-c is regioselective from the site trans to Si, the
final products often result from intramolecular rearrangement
and feature new ligands not trans, but cis to Si. Thus, oxidative
addition of dihydrogen to 1a-c furnishes hydrides cis to Si in
the very stable fac-(PMe3)3Ru(SiR3)H3 (SiR3 ) SiMe3, 4a;
SiMe2CH2SiMe3, 4b; SiEt3, 4c). The reverse manifold - HH
elimination from 4a and trapping with PMe3 or PMe3-d9 - is
also regioselective, but is very unfavorable. It appears to occur
via a putative isomer of 4a with two hydrides trans to a silyl
and meridianal phosphine ligands (1a-d9 is predominantly
produced with PMe3-d9 trans to Si and mer-phosphine ligands).
Slower, but irreversible, SiH elimination also occurs and
ultimately furnishes (PMe3)4RuH2 (9). The structure of 4a
exhibits a tetrahedral P3Si environment around the metal with
the three hydrides adjacent to silicon and capping the P2Si faces.
Although strong Si‚‚‚HRu interactions are not indicated in the
structure or by IR, the HSi distances (2.13-2.23(5) Å) suggest
some degree of nonclassical SiH bonding in the H3SiR3
fragment. Thermolysis of 1a in C6D6 at 45 °C leads to reversible
intermolecular CD activation of C6D6 via an 18e- (PMe3)3Ru-
(SiMe3)(Ph-d5)(H)(D) intermediate. Extensive H/D exchange
into the hydride, SiMe3, and PMe3 ligands is observed, followed
by much slower formation of cis-(PMe3)4Ru(D)(Ph-d5). The
extensive H/D exchange into SiMe3 and PMe3 ligands occurs
via (PMe3)3Ru(η2-CH2SiMe2)(D)(H) (3a-d1) and (PMe3)2Ru-
(η2-CH2PMe2)(SiMe3)(D)(H) intermediates. In an even slower
intramolecular CH activation process, (PMe3)3Ru(η2-CH2-
PMe2)H is also produced by a separate (and minor) pathway
involving the (PMe3)4Ru species. The reactivity trends reported
above illustrate that the HH, CH, and SiC elimination chemistry
(55) Berry, D. H.; Procopio, L. J. J. Am. Chem. Soc. 1989, 111, 4099-4100.
(56) Dioumaev, V. K.; Carroll, P. J.; Berry, D. H. Angew. Chem., in press.
9
J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003 8053

A R T I C L E S
Dioumaev et al.
190 K) δ 1.16 (br s, 18H, PMe3), 1.05 (br s, 9H, PMe3), 0.88 (d, JPH
) 4.5 Hz, 9H, PMe3), 0.81 (s, 9H, SiMe3), -0.32 (m, 3H, CH3). 31P
NMR (C7D8, 300 K): δ -4.10 (br s, 2P, mutually trans PMe3), -15.65
(br s, 1P, mutually cis PMe3), -22.80 (very br s, ν1/2 ) 700 Hz, 1P,
mutually cis PMe3); (C7D8, 240 K) δ -3.43 (dd, JPP ) 29.6 and 21.5
Hz, 2P, mutually trans PMe3), -15.13 (q, JPP ) 21.5 Hz, 1P, mutually
cis PMe3), -20.28 (td, JPP ) 29.6 and 22.3 Hz, 1P, mutually cis PMe3),
-62.5 (s, free PMe3); (C7D8, 190 K) δ -3.28 (dd, JPP ) 29.6 and 21.5
Hz, 2P, mutually trans PMe3), -14.92 (q, JPP ) 21.5 Hz, 1P, mutually
cis PMe3), -19.88 (td, JPP ) 29.6 and 22.3 Hz, 1P, mutually cis PMe3),
-62.5 (s, free PMe3).
can be strongly enhanced by a trans directing silyl ligand. In
octahedral or capped octahedral seven-coordinate complexes,
simple geometrical arguments dictate that the two eliminating
ligands and the trans directing silyl will be meridianal; thus,
the remaining ligands must also be roughly meridianal. It is
proposed to enforce a meridianal structural motif in the future
generations of catalysts by using chelating “pincer” phosphine
ligands to avoid the formation of excessively stable resting states
(e.g., 4a-c) and to enhance dehydrogenative and dealkanative
catalytic activity.
Synthesis of cis-(PMe3)4Ru(SiMe3)H, 1a. A cyclohexane solution
(3 mL) of (PMe3)4RuMe2 (2200 mg, 5.06 mmol) and HSiMe3 (30.3
mmol) was stirred for 14 h at 60 °C in a bomb. The volatiles were
removed in vacuo, and the residue was recrystallized from pentane to
yield 2150 mg (89% yield) of pure (PMe3)4Ru(SiMe3)H as a white
Experimental Section
All manipulations were performed in Schlenk-type glassware on a
dual-manifold Schlenk line or in a nitrogen-filled Vacuum Atmospheres
glovebox. NMR spectra were obtained at 200- and 500-MHz (for H)
1
1
crystalline powder. H NMR (C6D6): δ 1.33 (t, JPH ) 2.0 Hz, 18H,
on Bruker AF-200 and AM-500 FT NMR spectrometers, respectively.
All NMR spectra were recorded at 303 K unless stated otherwise.
mutually trans PMe3), 1.15 (d, JPH ) 4.2 Hz, 9H, mutually cis PMe3),
1.13 (d, JPH ) 4.9 Hz, 9H, mutually cis PMe3), 0.72 (d, JPH ) 1.4 Hz,
9H, SiMe3), -11.17 (dtd, JPH ) 66.5, 31.9, and 15.5 Hz, 1H, RuH);
(C6D12) δ 1.38 (m, 27H, PMe3), 1.25 (d, JPH ) 5.0 Hz, 9H, PMe3),
0.22 (s, 9H, SiMe3), -11.33 (dtd, JPH ) 67.4, 32.8, and 15.5 Hz, 1H,
RuH). 13C NMR (C6D6): δ 28.5 (m, PMe3), 28.0 (d, JPC ) 15.1 Hz,
PMe3), 16.6 (d, JPC ) 2.5 Hz, SiMe3). 29Si NMR (C6D6): δ 7.9 (dtd,
JPSi ) 97.6, 26.0, 17.8 Hz, SiMe3). 31P NMR (C6D6): δ -4.8 (dd, JPP
) 31 and 25 Hz, 2P, mutually trans PMe3), -15.2 (q, JPP ) 25 Hz,
1P, mutually cis PMe3), -17.0 (q, JPP ) 25 Hz, 1P, mutually cis PMe3).
31P NMR (C6D12): δ -4.6 (dd, JPP ) 31 and 24 Hz, 2P, mutually
trans PMe3), -15.0 (q, JPP ) 25 Hz, 1P, mutually cis PMe3), -16.6
(q, JPP ) 27 Hz, 1P, mutually cis PMe3). IR (Nujol): ν(RuH) ) 1821
cm-1. IR (C6H6): ν(RuH) ) 1827 cm-1. Anal. Calcd for C15H46Si1P4-
Ru1: C, 37.56; H, 9.67. Found: C, 37.82; H, 10.16.
Chemical shifts are reported relative to tetramethylsilane for H, 13C,
1
and 29Si spectra, and external 85% H3PO4 for 31P resonances. The
temperature of the NMR probe was calibrated against methanol
(estimated error 0.3 K). 13C and 31P NMR spectra were recorded with
1
broadband H decoupling. 29Si NMR spectra were obtained using a
DEPT-135 pulse sequence with 1H refocusing. Spin-lattice relaxation
times (T1) were measured by using the standard inversion-recovery
(180° - τ - 90°) pulse sequence. Infrared spectra were recorded on a
Perkin-Elmer model 1430 spectrometer. Elemental analyses were
performed by Robertson Laboratory, Inc. (Madison, NJ). Photolysis
reactions were carried out in a Rayonet photochemical reactor using
low-pressure Hg arc lamps (λ ) 350 nm).
Hydrocarbon solvents were dried over Na/K alloy-benzophenone.
Benzene-d6, cyclohexane-d12, and methylcyclohexane-d14 were dried
over Na/K alloy. H2, D2, and CO (Airco) were used as received. cis-
(PMe3)4RuMe2,57 (PMe3)3Ru(η2-CH2PMe2)H,31,32 PMe3,58 and HSiMe2-
Alternative Synthesis of cis-(PMe3)4Ru(SiMe3)H, 1a. A benzene
solution (15 mL) of (PMe3)3Ru(η2-CH2PMe2)H (550 mg, 1.36 mmol)
and HSiMe3 (2970 mg, 40.1 mmol) was photolyzed (350 nm) at 10 °C
for 80.5 h. The volatiles were removed in vacuo, and the residue was
recrystallized from petroleum ether, yielding 381 mg (59% yield) of
(PMe3)4Ru(SiMe3)H as a white crystalline powder.
59
CH2SiMe3 were synthesized according to the literature procedures.
HSiMe3 and DSiMe3 were prepared by the reaction of Me3SiCl and
LiAlH4 or LiAlD4 in nBu2O and purified by trap-to-trap vacuum
fractionation. Triethylsilane (Aldrich) was dried over molecular sieves
prior to use. PPh3-polystyrene beads (Aldrich) were dried in vacuo.
Triphenylborane (Aldrich) was recrystallized from hexanes/toluene
before use.
Synthesis of cis-(PMe3)4Ru(SiMe2CH2SiMe3)H, 1b. A cyclohexane
(2 mL) solution of (PMe3)4RuMe2 (1.192 g, 2.74 mmol) and HSiMe2-
CH2SiMe3 (1.6 mL, 13.7 mmol) was heated for 24 h at 60 °C. The
volatiles were removed in vacuo, and the residue was recrystallized
Synthesis of cis-(PMe3)4Ru(SiMe3)Me, 7a. A toluene (10 mL)
solution of (PMe3)4RuMe2 (1740 mg, 4.0 mmol) and HSiMe3 (4.0
mmol) was stirred for 24 h at room temperature. The volatiles were
removed in vacuo, and the residue was washed with cold pentane (-40
°C). Fractional recrystallization from toluene-pentane (1:10) yielded
250 mg (12.7% yield) of pure (PMe3)4Ru(SiMe3)Me as a white
1
from pentane to yield 0.76 g (50.3%) of colorless crystals. H NMR
(C6D6): δ 1.32 (t, JPH ) 2.4 Hz, 18H, mutually trans PMe3), 1.15 (d,
JPH ) 5.0 Hz, 9H, mutually cis PMe3), 1.13 (d, JPH ) 5.0 Hz, 9H,
mutually cis PMe3), 0.76 (d, JPH ) 1.1 Hz, 6H, SiMe2), 0.39 (s, 9H,
SiMe3), 0.33 (s, 2H, CH2), -11.21 (dtd, JPH ) 67.5, 32.8, and 16.4
Hz, 1H, RuH); (C6D12) δ 1.39 (t, JPH ) 2.3 Hz, 18H, PMe3), 1.38 (d,
JPH ) 1.8 Hz, 9H, PMe3), 1.25 (d, JPH ) 5.5 Hz, 9H, PMe3), 0.33 (s,
6H, SiMe2), 0.21 (d, JPH ) 1.8 Hz, 2H, CH2), -0.01 (s, 9H, SiMe3),
-11.34 (dtd, JPH ) 68.4, 32.4, and 15.7 Hz, 1H, RuH). 13C NMR
1
crystalline powder. H NMR (C6D6): δ 1.18 (br s, 27H, PMe3), 0.98
(d, JPH ) 4.0 Hz, 9H, PMe3), 0.59 (s, 9H, SiMe3), -0.37 (m, 3H, CH3);
(C6D12) δ 1.31 (br s, 27H, PMe3), 1.28 (d, JPH ) 4.4 Hz, 9H, PMe3),
0.14 (s, 9H, SiMe3), -0.68 (m, 3H, CH3). 13C NMR (C6D6): δ 29.4
(m, PMe3), 23.48 (m, PMe3), 24.16 (d, JPC ) 11.6 Hz, Me), 12.47 (q,
JPC ) 3.0 Hz, SiMe3). 31P NMR (C6D6): δ -3.0 (br m, 2P, mutually
trans PMe3), -14.5 (m, 1P, mutually cis PMe3), -21.7 (br m, 1P,
mutually cis PMe3). Anal. Calcd for C16H48Si1P4Ru1: C, 38.93; H, 9.80.
Found: C, 38.88; H, 9.97.
(C6D6): δ 28.4 (tt, JPC ) 13.2 and 4.0 Hz, PMe3), 28.0 (dm, JPC
)
16.0 Hz, PMe3), 17.8 (br s, CH2), 16.4 (m, JPC ) 3.2 Hz, SiMe2), 3.6
(s, SiMe3). 29Si NMR (C6D6): δ 11.05 (dm, JPSi ) 83 Hz, SiMe2), 0.65
(d, JPSi ) 5 Hz, SiMe3). 31P NMR (C6D6): δ -5.64 (q, JPP ) 29.2 Hz,
2P, mutually trans-PMe3), -15.74 (m, 1P, mutually cis-PMe3), -17.27
1
1
(br m, 1P, mutually cis-PMe3). IR (fluorolube): ν(RuH) ) 1790 cm-1
.
VT H NMR of (PMe3)4Ru(SiMe3)Me, 7a. H NMR (C7D8, 300
K): δ 1.19 (br s, 18H, PMe3), 1.14 (br s, 9H, PMe3), 1.01 (d, JPH
Anal. Calcd for C18H54Si2P4Ru1: C, 39.18; H, 9.86. Found: C, 38.98;
H, 9.85.
)
4.5 Hz, 9H, PMe3), 0.49 (s, 9H, SiMe3), -0.48 (m, 3H, CH3); (C7D8,
240 K) δ 1.17 (br s, 18H, PMe3), 1.11 (br s, 9H, PMe3), 0.95 (d, JPH
) 4.5 Hz, 9H, PMe3), 0.44 (s, 9H, SiMe3), -0.41 (m, 3H, CH3); (C7D8,
Synthesis of cis-(PMe3)4Ru(SiEt3)H, 1c. A benzene solution (20
mL) of (PMe3)3Ru(η2-CH2PMe2)H (0.515 g, 1.270 mmol) and HSiEt3
(3.02 g, 26.034 mmol) was photolyzed (350 nm) at 10 °C for 3 weeks.
Volatiles were removed in vacuo, and the residue was recrystallized
from petroleum ether, yielding 0.242 g of light brown (PMe3)4Ru-
(57) Statler, J. A.; Wilkinson, G.; Thornton-Pett, M.; Hursthouse, M. B. J. Chem.
Soc., Dalton Trans. 1984, 1731.
(58) Luetkens, M. L.; Sattelberger, A. P.; Murray, H. H.; Basil, J. D.; Fackler,
J. P. Inorg. Synth. 1989, 26, 7.
1
(SiEt3)H (37% yield). H NMR (C6D6): δ 1.42 (t, JHH ) 7.8 Hz, 6H,
(59) Sakurai, H.; Hosomi, A.; Kumada, M. Chem. Commun. 1968, 930.
CH2CH3), 1.32 (t, JPH ) 2.2 Hz, 18H, two mutually trans PMe3), 1.16
9
8054 J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003

Synthesis and Reactivity of Silyl Ruthenium Complexes
A R T I C L E S
(d, JPH ) 5.1 Hz, 9H, PMe3), 1.14 (d, IPH ) 4.6 Hz, 9H, PMe3), 1.04
(q, JHH ) 7.8 Hz, SiCH2), -11.04 (dq, JPH ) 69.5 and 23.8 Hz, 1H,
RuH). 13C NMR: δ 28.4 (m, PMe3), 15.0 (CH2CH3), 11.0 (s, SiCH2).
31P NMR: δ -3.9 (t, JPP ) 30.9 Hz, 2P, two mutually trans PMe3),
-16.5 (q, JPP ≈ 18 Hz, 1P, PMe3), -19.9 (q, JPP ≈ 21 Hz, 1P, PMe3).
IR (benzene): ν(RuH) ) 1813 cm-1. Anal. Calcd for C18H52P4Si1Ru1:
C, 41.44; H, 10.05. Found: C, 41.46; H, 9.61.
mutually trans PMe3-d9), -15.46 (t, JPP ) 21.6 Hz, PMe3 trans to H),
-17.77 (t, JPP ) 21.6 Hz, PMe3-d9 trans to H).
Reaction of cis-(PMe3)4Ru(SiEt3)H and PMe3-d9. An NMR tube
was loaded with a C6D6 solution (0.5 mL) of cis-(PMe3)4Ru(SiEt3)H
(6 mg, 0.012 mmol) and placed under vacuum. At -196 °C, PMe3-d9
(0.17 mmol) was added, and the tube was sealed. Upon thawing, the
1H NMR spectrum was recorded within 5 min, confirming that ca. 1
equiv of deuterated phosphine had been incorporated into cis-(PMe3)4-
Synthesis of cis-(PMe3)4Ru(SiMe2H)H, 1d. A cyclohexane solution
(50 mL) of (PMe3)4Ru(H)2 (9) (270 mg, 0.66 mmol) and Me2SiH2 (390
mg, 6.50 mmol) was photolyzed (350 nm) at 15 °C for 6 days. Volatiles
were removed in vacuo, and the residue was recrystallized from
petroleum ether at -40 °C to yield 240 mg (80%) of (PMe3)4Ru-
1
Ru(SiEt3)H. After 1 h at 25 °C, the H NMR spectrum showed that
complete exchange with all positions had occurred.
Photolytic Reaction of (PMe3)3Ru(η2-CH2PMe2)H and PMe3-d9.
An NMR tube was loaded with a C6D6 solution (0.5 mL) of (PMe3)3-
Ru(η2-CH2PMe2)H (17 mg, 0.042 mmol) and placed under vacuum.
PMe3-d9 (0.4 mmol) was added, and the tube was sealed. The tube
was photolyzed (λ ) 350 nm) at 10 °C, and the reaction was monitored
(SiMe2H)H. 1H NMR (C6D6): δ 4.60 (m, 1H, SiMe2H), 1.35 (t, JHP
)
2.2 Hz, 18H, mutually trans PMe3), 1.15 (d, JHP ) 3.6 Hz, 9H, mutually
cis PMe3), 1.13 (d, JHP ) 3.5 Hz, 9H, mutually cis PMe3), 0.83 (d, JHP
) 4.0 Hz, 6H, SiMe2H), -10.96 (dq, JHP ) 72 and 24 Hz, 1H, RuH).
13C NMR: δ 28.1 (d, JCP ) 15 Hz, PMe3), 27.2 (d, JCP ) 18 Hz, PMe3),
26.7 (t, JCP ) 14 Hz, PMe3), 9.2 (s, SiMe2H). 29Si NMR δ -1.1 (dtd,
JPSi ) 94, 28, 14 Hz, SiMe2H). 31P NMR δ -3.2 (t, JHP ) 29 Hz, 2P,
two mutually trans PMe3), -14.3 (m, 1P, PMe3), -16.5 (m, 1P, PMe3).
Anal. Calcd for C14H44SiP4Ru1: C, 36.12; H, 9.53. Found: C, 36.58;
H, 9.73.
1
by H NMR. After 24 h, ca. 20% exchange of the PMe3 ligands for
PMe3-d9 had occurred.
Thermal Reaction of (PMe3)3Ru(η2-CH2PMe2)H and PMe3-d9.
An NMR tube was loaded with a C6D12 solution (0.5 mL) of (PMe3)3-
Ru(η2-CH2PMe2)H (15 mg, 0.037 mmol) and C6Me6 (2 mg, internal
standard) and placed under vacuum. PMe3-d9 (0.44 mmol) was added,
and the tube was sealed. The tube was heated at 65 °C, and the reaction
1
Synthesis of cis-(PMe3)4RuH2, 9. A cyclohexane solution (10 mL)
of (PMe3)3Ru(η2-CH2PMe2)H (0.797 g, 1.966 mmol) was placed in a
thick-walled glass pressure flask, and PMe3 (1.45 mmol) was added
by vacuum transfer at -196 °C. The solution was placed under 3 atm
H2 and heated at 110 °C for 91 h. Volatiles were removed in vacuo,
and the pale yellow residue was sublimed (85 °C, 10-4 Torr), yielding
0.702 g of white cis-(PMe3)4RuH2 (88% yield). Spectroscopic data are
in agreement with literature values.60,61 1H NMR (C6D12): δ 1.32 (t,
JPH ) 2.5 Hz, 18H, mutually trans PMe3), 1.28 (d, JPH ) 4.6 Hz, 18H,
was monitored by H NMR. After 102 h, ca. 50% exchange of the
PMe3 ligands for PMe3-d9 had occurred.
Attempted Reaction of (PMe3)4Ru(SiMe3)Me with BPh3. BPh3
(19 mg, 0.04 mmol) and (PMe3)4Ru(SiMe3)Me (40 mg, 0.12 mmol)
were suspended in 1 mL of pentane. A color change from light orange
to dark red and formation of a white precipitate started within seconds.
The mixture was stirred for 10 min, filtered, and stirred with 40 mg of
polymer-supported PPh3 (cross-linked polystyrene beads, 3 mmol of
PPh3 per 1 g of polymer) for 10 min, decanted, treated again with 10
mg of PPh3-polystyrene beads, and decanted again. The volatiles were
removed in vacuo, the solids were redissolved in C6D12, and the sample
was sealed in an NMR tube under ca. 1 atm of N2.
Observation of mer-(CO)(PMe3)3Ru(SiMe3)H, mer-8. An NMR
tube was loaded with a C6D6 solution (0.5 mL) of (PMe3)4Ru(SiMe3)H
(17 mg, 0.035 mmol), and the solution was degassed in vacuo. At -196
°C, carbon monoxide (ca. 0.30 mmol) was added, and the tube was
sealed. The mixture was thawed to room temperature, and the 1H and
31P NMR spectra were recorded within 10 min. Initial spectra showed
only the presence of mer-(CO)(PMe3)3Ru(SiMe3)H and PMe3. The
isomerization to fac-(CO)(PMe3)3Ru(SiMe3)H (fac-8) was followed by
1H NMR spectroscopy. After 36 min at 25 °C, the tube contained a
1:1 mixture of fac and mer isomers. The fac/mer system came to a
PMe3), -10.09 (m, RuH2). IR (benzene): ν(RuH) ) 1820 cm-1
.
Reaction of (PMe3)4Ru(SiMe3)H and PMe3-d9. An NMR tube was
loaded with a C6D6 solution (0.5 mL) of (PMe3)4Ru(SiMe3)H (11 mg,
0.023 mmol), and the solution was degassed in vacuo. At -196 °C,
PMe3-d9 (0.35 mmol) was added, and the tube was sealed. Upon
1
thawing, H and 31P NMR spectra were recorded, showing complete
exchange of only the phosphine ligand in the position trans to silicon
within 2 min. The other phosphine ligands are exchanging with free
phosphine with a τ1/2 of ca. 2 days. 31P NMR (C6D6): δ -5.22 (dd, JPP
) 31 and 25 Hz, mutually trans PMe3), -5.88 (dd, JPP ) 31 and 25
Hz, PMe3 trans to PMe3-d9), -6.80 (dd, JPP ) 31 and 25 Hz, PMe3-d9
trans to PMe3), -7.40 (dd, JPP ) 31 and 25 Hz, mutually trans PMe3-
d9), -15.56 (m, PMe3 trans to H), -17.45 (m, PMe3-d9 trans to H),
-19.85 (m, PMe3-d9 trans to Si).
Reaction of (PMe3)4Ru(SiMe3)Me and PMe3-d9. An NMR tube
was loaded with a C6D6 solution (0.5 mL) of (PMe3)4Ru(SiMe3)Me
(10 mg, 0.02 mmol), and the sample was degassed in vacuo. At -196
°C, PMe3-d9 (0.24 mmol) was added, and the tube was sealed. Upon
thawing, the reaction was followed by 1H and 31P NMR spectroscopy.
After only 3 min, all phosphine positions showed nonselective
incorporation of labeled phosphine (1H NMR: 40% deuterium in the
mutually trans positions at δ 1.21 and 30% in the position at δ 1.11).
The 1H (δ 0.98) and 31P (δ -21.7) resonances for one of the mutually
cis phosphines disappeared due to coalescence with free phosphine
signal. After 1 day, statistical distribution of 76% deuterium is found
in all observable positions including free phosphine (1H and 31P NMR).
1H NMR (C6D6): δ 1.20 and 1.18 (br s, 18H, mutually trans PMe3 and
PMe3 trans to PMe3-d9), 1.11 (d, JPH ) 5.2 Hz, 9H, PMe3 trans to H),
0.79 (d, JPH ) 1.9 Hz, free PMe3), 0.59 (s, 9H, SiMe3), -0.41 (m, 3H,
CH3). 31P NMR (C6D6): δ -3.80 (d, JPP ) 21.5 Hz, mutually trans
PMe3), -4.60 (d, JPP ) 21.5 Hz, PMe3 trans to PMe3-d9), -5.45 (d,
JPP ) 21.9 Hz, PMe3-d9 trans to PMe3), -6.16 (d, JPP ) 21.5 Hz,
95:5 equilibrium within 12 h.
mer-(CO)(PMe3)3Ru(SiMe3)H. H NMR (C6D6): δ 1.29 (t, JPH
1
)
2.8 Hz, 18H, mutually trans PMe3), 1.10 (d, JPH ) 6.3 Hz, 9H, PMe3),
0.51 (s, 9H, SiMe3), -9.07 (dt, JPH ) 73.8 and 29.1 Hz, 1H, RuH,
trans to CO). 31P NMR (C6D6): δ -7.0 (d, JPP ) 23 Hz, 2P, mutually
trans PMe3), -15.7 (t, JPP ) 23 Hz, 1P, PMe3, trans to hydride).
Synthesis of fac-(CO)(PMe3)3Ru(SiMe3)H, fac-8. A C6H6 solution
(20 mL) of (PMe3)4Ru(SiMe3)H (385 mg, 0.80 mmol) was stirred under
1 atm of carbon monoxide for 2.5 h. Volatiles were removed in vacuo,
and the residue was recrystallized from petroleum ether, yielding 294
mg (85% yield) of a mixture of fac- and mer-(CO)(PMe3)3Ru(SiMe3)H
in a 95:5 ratio. 1H NMR (C6D6): δ 1.123 (d, JPH ) 6.3 Hz, 9H, PMe3),
1.117 (d, JPH ) 6.3 Hz, 9H, PMe3), 1.09 (d, JPH ) 7.0 Hz, 9H, PMe3),
0.71 (d, JPH ) 1.3 Hz, 9H, SiMe3), -9.11 (ddd, JPH ) 67.1, 29.9, and
21.4 Hz, 1H, RuH). 13C NMR (C6D6): δ 207.9 (dt, JPC ) 79.3 and 9.0
Hz, CO), 25.4 (d, JPC ) 19.5 Hz, PMe3), 24.5 (td, JPC ) 24.2 and 5.0
Hz, PMe3), 24.2 (d, JPC ) 21.3 Hz, PMe3), 12.9 (s, SiMe3). 29Si NMR
(C6D6): δ 4.6 (ddd, JPSi ) 74.8, 21.0, 11.3 Hz, SiMe3). 31P NMR
(C6D6): δ -11.2 (t, JPP ) 37 Hz, 1P, PMe3), -15.9 (dd, JPP ) 37 and
21 Hz, 1P, PMe3), -22.0 (dd, JPP ) 37 and 21 Hz, 1P, PMe3). IR
(C6H6): ν(CO) ) 1932 cm-1, ν(RuH) ) 1858 cm-1. Anal. Calcd for
C13H37O1Si1P3Ru1: C, 36.19; H, 8.64. Found: C, 36.10; H, 8.83.
(60) Jones, R. A.; Wilkinson, G.; Colquohoun, I. J.; McFarlane, W.; Galas, A.
M. R.; Hursthouse, M. B. J. Chem. Soc., Dalton Trans. 1980, 2480.
(61) Mainz, V. V.; Andersen, R. A. Organometallics 1984, 3, 675.
9
J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003 8055

A R T I C L E S
Dioumaev et al.
Isomerization of mer-(CO)(PMe3)3Ru(SiMe3)H in the Presence
of PMe3-d9. A C6H6 solution (1 mL) of (PMe3)4Ru(SiMe3)H (10 mg,
0.02 mmol) was stirred under ca. 3 atm of CO for 5 min. The excess
of CO was removed in vacuo, and PMe3-d9 (0.40 mmol) was added.
The solution was stirred in the dark for 1 h, and volatiles were removed
in vacuo. 1H and 31P spectra showed no incorporation of PMe3-d9 into
the final product, fac-(CO)(PMe3)3Ru(SiMe3)H.
Reaction of (PMe3)4Ru(SiMe3)H with H2. An NMR tube was
loaded with a C6D6 solution (0.5 mL) of (PMe3)4Ru(SiMe3)H (9 mg,
0.02 mmol), and the sample was degassed in vacuo. Hydrogen (ca. 3
atm) was added, and the tube was sealed. The 1H NMR spectrum was
recorded within 5 min and showed complete conversion to (PMe3)3-
Ru(SiMe3)H3, with the generation of ca. 1 equiv of free PMe3.
Synthesis of (PMe3)3Ru(SiMe3)H3, 4a. Method 1: A cyclohexane
solution (20 mL) of cis-(PMe3)4RuH2 (0.500 g, 1.23 mmol) and HSiMe3
(0.900 g, 12.16 mmol) was photolyzed (350 nm) at 10 °C for 87 h.
Volatiles were removed in vacuo, and the residue was recrystallized
from petroleum ether, yielding 0.442 g of colorless (PMe3)3Ru(SiMe3)-
H3 (89% yield).
Reaction of (PMe3)4Ru(SiMe3)Me with HSiMe2CH2SiMe3. An
NMR tube was loaded with a C6D6 solution (0.5 mL) of (PMe3)4Ru-
(SiMe3)Me (10 mg, 0.02 mmol). HSiMe2CH2SiMe3 (0.20 mmol) was
vacuum transferred into the tube, and the tube was flame sealed. The
1H NMR spectra were recorded after 0.2, 2, and 48 h to verify
establishment of the equilibrium. The equilibrium constant (Keq(300)
) 0.0059) was recalculated from the integral intensities of the 1H NMR
signals of (PMe3)4Ru(SiMe3)Me, (PMe3)4Ru(SiMe2CH2SiMe3)Me,
HSiMe3, and HSiMe2CH2SiMe3.
Method 2: A benzene solution (50 mL) of (PMe3)4Ru(SiMe3)H
(1.491 g, 3.11 mmol) was stirred under 3.5 atm hydrogen for 20 h.
Volatiles were removed in vacuo, and the residue was recrystallized
from petroleum ether, yielding 0.822 g of colorless (PMe3)3Ru(SiMe3)-
H3 (65% yield).
1
(PMe3)4Ru(SiMe2CH2SiMe3)Me. H NMR (C6D6): δ 1.17 (br s,
1H NMR (C6D6): δ 1.13 (m, 27H, PMe3), 0.90 (s, 9H, SiMe3),
-10.18 (br m, 3H, RuH3); (C6D12) δ 1.32 (m, 27H, PMe3), 0.30 (s,
9H, SiMe3), -10.34 (br m, 3H, RuH3). 13C{1H} NMR (C6D6): δ 26.7
(m, PMe3), 18.8 (d, 3JPC ) 3.3 Hz, SiMe3). 29Si NMR (C6D6): δ -10.8
27H, PMe3), 0.99 (d, JPH ) 2.5 Hz, 9H, PMe3), 0.54 (s, 6H, SiMe2),
0.30 (s, 9H, SiMe3), 0.20 (s, 2H, CH2), -0.15 (m, 3H, RuMe).
Reaction of (PMe3)4Ru(SiMe3)H with HSiMe2CH2SiMe3. An
NMR tube was loaded with a C6D6 solution (0.5 mL) of (PMe3)4Ru-
(SiMe3)H (10 mg, 0.02 mmol). HSiMe2CH2SiMe3 (0.20 mmol) was
vacuum transferred into the tube, and the tube was flame sealed. The
1H NMR spectra were recorded after 2 min and 48 h to verify
establishment of the equilibrium. The equilibrium constant (Keq(300)
) 0.0065) was recalculated from the integral intensities of the 1H NMR
signals of (PMe3)4Ru(SiMe3)H, (PMe3)4Ru(SiMe2CH2SiMe3)H, HSiMe3,
and HSiMe2CH2SiMe3.
2
(q, JPSi ) 7.6 Hz, SiMe3). 31P{1H} NMR (C6D6): δ -5.0 (s, PMe3).
31P{1H} NMR (C6D12): δ -4.5 (s, PMe3). IR (Nujol): ν(RuH) ) 1890
cm-1. IR (benzene): ν(RuH) ) 1887 cm-1. Anal. Calcd for C12H39P3-
SiRu: C, 35.54; H, 9.69. Found: C, 35.36; H, 10.02.
Synthesis of (PMe3)3Ru(SiMe2CH2SiMe3)H3, 4b. A benzene solu-
tion (25 mL) of cis-(PMe3)4RuH2 (0.500 g, 1.23 mmol) and HSiMe2-
CH2SiMe3 (0.843 g, 5.77 mmol) was photolyzed (350 nm) at 10 °C
for 167 h. Volatiles were removed in vacuo, and the residue was
recrystallized from petroleum ether, yielding 0.540 g of colorless
(PMe3)3Ru(SiMe2CH2SiMe3)H3 (92% yield).
Reaction of (PMe3)4Ru(SiMe3)H with DSiMe3. An NMR tube was
loaded with a C6D6 solution (0.5 mL) of (PMe3)4Ru(SiMe3)H (24 mg,
0.05 mmol). DSiMe3 (0.5 mmol) was vacuum transferred into the tube,
1
and the tube was flame sealed. The H NMR spectrum was recorded
Synthesis of (PMe3)3Ru(SiEt3)H3, 4c. A benzene solution (25 mL)
of cis-(PMe3)4RuH2 (0.313 g, 0.768 mmol) and HSiEt3 (0.880 g, 7.788
mmol) was photolyzed (350 nm) at 10 °C for 195 h. Volatiles were
removed in vacuo, and the residue was recrystallized from petroleum
ether, yielding 0.309 g of colorless (PMe3)3Ru(SiEt3)H3 (90% yield).
1H NMR: δ 1.32 (t, JHH ) 7.8 Hz, 9H, CH2CH3), 1.15 (d, JPH ) 5.0
Hz, 27H, PMe3), 1.07 (q, JHH ) 7.8 Hz, 6H, SiCH2), -10.53 (br m,
within 5 min and showed complete conversion to (PMe3)4Ru(SiMe3)D,
with the generation of ca. 1 equiv of free HSiMe3.
Thermolysis of (PMe3)4Ru(SiMe3)H in C6D12. An NMR tube was
loaded with a C6D12 solution (0.5 mL) of (PMe3)4Ru(SiMe3)H (5 mg,
0.01 mmol) and C6Me6 (2 mg, internal standard), the solution was
degassed in vacuo, and the tube was sealed. The sample was heated at
65 °C, and the reaction was monitored by 1H NMR spectroscopy. The
reaction proceeded to yield (PMe3)3Ru(η2-CH2PMe2)H and free HSiMe3
over a period of several days.
3
3H, RuH3). 13C NMR: δ 26.4 (m, PMe3), 19.2 (q, JPC ) 2.9 Hz,
SiCH2), 9.8 (s, CH2CH3). 29Si NMR: δ 12.7 (q, 2JPSi ) 7.7 Hz, SiEt3).
31P NMR: δ -5.4 (s, PMe3). IR (Nujol): 1899 cm-1 (νRuH). IR
(benzene): ν(RuH) ) 1897 cm-1. Anal. Calcd for C15H45P3Si1Ru1: C,
40.25; H, 10.13. Found: C, 40.53; H, 10.52.
Thermolysis of (PMe3)4Ru(SiMe3)H in C6H6 - Synthesis of cis-
(PMe3)4Ru(Ph)H. A benzene solution (30 mL) of (PMe3)4Ru(SiMe3)H
(275 mg, 0.57 mmol) was heated at 65 °C for a total of 348 h. To
prevent the buildup of HSiMe3, the volatiles were periodically removed
in vacuo, and fresh benzene was added. The reaction was monitored
Reaction of cis-(PMe3)4Ru(SiEt3)H with H2. An NMR tube was
loaded with a C6D6 solution (0.5 mL) of cis-(PMe3)4Ru(SiEt3)H (5 mg,
0.001 mmol) and placed under vacuum. Hydrogen (ca. 3 atm) was
1
1
added, and the tube was sealed. The H NMR spectrum was recorded
by H NMR spectroscopy. Volatiles were removed in vacuo, and the
within 5 min and showed complete conversion to (PMe3)3Ru(SiEt3)-
H3, with the generation of ca. 1 equiv of free PMe3.
residue was recrystallized from petroleum ether to yield 220 mg of
pale orange (PMe3)4Ru(Ph)H (79% yield). H NMR (C6D6): δ 7.99
1
and 7.11 (m, 5H, Ph), 1.20 (d, JPH ) 4.6 Hz, 9H, mutually cis PMe3),
1.16 (d, JPH ) 5.4 Hz, 9H, mutually cis PMe3), 1.08 (t, JPH ) 2.7 Hz,
18H, mutually trans PMe3), -9.41 (dq, JPH ) 94.0 and 26.3 Hz, 1H,
RuH). 13C NMR (C6D12): δ 125.5 (br s, Ph), 120.4 (br s, Ph), 29.0 (d,
JPC ) 16 Hz, PMe3), 25.6 (d, JPC ) 17 Hz, PMe3), 24.6 (m, mutually
trans PMe3). 31P NMR (C6D6): δ -1.9 (t, JPP ) 26 Hz, 2P, mutually
trans PMe3), -11.0 (dt, JPP ) 26 and 18 Hz, 1P, PMe3, trans to Ph),
-17.6 (dt, JPP ) 26 and 18 Hz, 1P, PMe3, trans to hydride). IR
(Nujol): ν(RuH) ) 1855 cm-1. Anal. Calcd for C18H42P4Ru1: C, 44.72;
H, 8.76. Found: C, 44.32; H, 8.98.
Reaction of (PMe3-d9)(PMe3)3Ru(SiMe3)H with H2. An NMR tube
was loaded with a C6D6 solution (0.5 mL) of (PMe3)4Ru(SiMe3)H (9
mg, 0.02 mmol), and the solution was degassed in vacuo. At -196
°C, PMe3-d9 (0.35 mmol) was added, and the tube was sealed. Upon
thawing, 1H and 31P showed that complete exchange had occurred,
forming (PMe3-d9)(PMe3)3Ru(SiMe3)H with the labeled phosphine in
the position trans to silicon. The tube was opened under inert
atmosphere, and the volatiles were removed in vacuo. The solid was
dissolved in fresh C6D6 (0.5 mL), placed in an NMR tube, and hydrogen
(ca. 3 atm) was added. The tube was sealed, and the 1
H NMR spectrum
was recorded within 5 min. The 1H and 31P NMR spectra showed that
the free phosphine was more than 95% PMe3-d9 and that the product
(PMe3)3Ru(SiMe3)H3 contained no labeled phosphine.
Thermolysis of cis-(PMe3)4Ru(SiEt3)H in C6H6. A benzene solution
(0.8 mL) of cis-(PMe3)4Ru(SiEt3)H (15 mg, 0.029 mmol) was heated
1
at 65 °C for 50 h, and the volatiles were removed in vacuo. The H
NMR spectrum of the solid residues showed complete conversion to
cis-(PMe3)4Ru(Ph)H (>85%) and several other minor products, includ-
ing (PMe3)3Ru(η2-CH2PMe2)H (ca. 5%).
Photolysis of (PMe3)3Ru(SiMe3)H3 with PMe3-d9. An NMR tube
was loaded with a cyclohexane-d12 solution (0.4 mL) of (PMe3)3Ru-
(SiMe3)H3 (6 mg, 0.015 mol) and C6Me6 (2 mg, internal standard).
9
8056 J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003

Synthesis and Reactivity of Silyl Ruthenium Complexes
A R T I C L E S
PMe3-d9 (0.179 mmol) was added, and the tube was sealed. The tube
e h e 18, -12 e k e 12, -34 e l e 36. The intensity data were
corrected for Lorentz and polarization effects but not for absorption.
All 4272 unique reflections (Rint ) 0.0403) were used during subsequent
structure refinement (200 parameters refined.) The structure was solved
by direct methods (SIR92).64 Refinement was by full-matrix least
squares techniques based on F 2 using SHELXL-93.65 Non-hydrogen
atoms were refined anisotropically, and hydrogen atoms were included
as constant contributions to the structure factors and were not refined.
The maximum ∆/σ in the final cycle of least squares was -0.001, and
the two most prominent peaks in the final difference Fourier were
+0.504 and -0.583 e/Å3.
For (PMe3)4Ru(SiMe2CH2SiMe3)H, indexing was performed from
a series of 1° oscillation images with exposures of 100 s per frame. A
hemisphere of data was collected using 6° oscillation angles with
exposures of 100 s per frame and a crystal-to-detector distance of 82
mm. A total of 28 749 reflections were measured over the ranges 5.16
e 2θ e 54.96°, -16 e h e 16, -30 e k e 30, -12 e l e 13. The
intensity data were corrected for Lorentz and polarization effects but
not for absorption. All 6757 unique reflections (Rint ) 0.0382) were
used during subsequent structure refinement (292 parameters refined.)
The structure was solved by direct methods (SIR92).64 One of the PMe3
groups (P3, C7, C8, C9) was found to be rotationally disordered with
two contributing rotamers with a ratio of 3:1. Refinement was by full-
matrix least squares techniques based on F 2 using SHELXL-93.65 Non-
hydrogen atoms were refined anisotropically, H1 was refined isotro-
pically, and all other hydrogen atoms were refined using a “riding”
model. The maximum ∆/σ in the final cycle of least squares was 0.007,
and the two most prominent peaks in the final difference Fourier were
+0.521 and -0.613 e/Å3.
1
was kept at room temperature, and the reaction was monitored by H
1
NMR. After 6 days at 25 °C, the H NMR spectrum showed that no
reaction had occurred. The tube was then photolyzed (λ ) 350 nm) at
10 °C. After 24 h, the 1H NMR spectrum showed that ca. 21% exchange
had occurred between the phosphine ligands of (PMe3)3Ru(SiMe3)H3
and the free PMe3-d9. After 112.5 h, ca. 52% exchange had occurred.
Reaction of (PMe3)3Ru(SiMe3)H3 in Neat PMe3-d9. (PMe3)3Ru-
(SiMe3)H3 (2 mg, 0.005 mmol) was dissolved in 0.5 mL of PMe3-d9
and stirred at 25 °C for 21.5 h in the dark. The mixture was periodically
degassed by freeze-pump-thaw cycles to remove any evolved
hydrogen. The volatiles were removed in vacuo, and the solids were
analyzed by 1H and 31P NMR spectroscopy. The product was composed
of ca. 75% of (PMe3)4Ru(SiMe3)H and 25% starting (PMe3)3Ru(SiMe3)-
1
H3. Both the H and the 31P NMR spectra indicate that the mutually
trans PMe3 positions of (PMe3)4Ru(SiMe3)H do not contain labeled
phosphine and that the majority of labeled phosphine (>80%) has been
incorporated into the position trans to the silyl group.
Thermolysis of (PMe3)3Ru(SiMe3)H3 and PMe3 at 70 °C. A NMR
tube was loaded with a cyclohexane-d12 solution (0.4 mL) of (PMe3)3-
Ru(SiMe3)H3 (4 mg, 0.01 mmol) and placed under vacuum. PMe3 (0.02
mmol) was added, and the tube was sealed. The tube was heated at 70
°C, and the reaction was followed by 1H NMR. After 4 h at 70 °C, the
1H NMR spectrum showed ca. 50% conversion to a ca. 10:1 mixture
of (PMe3)4RuH2 and (PMe3)4Ru(SiMe3)H. Free trimethylsilane was also
observed. The reaction was complete after 115 h, yielding (PMe3)4-
RuH2 and HSiMe3 as the only products.
Reaction of (PMe3)3Ru(SiMe3)H3 with D2. An NMR tube was
loaded with a C6D6 solution (0.5 mL) of (PMe3)3Ru(SiMe3)H3 (20 mg,
0.049 mmol), and the sample was degassed in vacuo. Deuterium (ca.
3 atm) was added, and the tube was sealed. The reaction was monitored
For (PMe3)4Ru(SiMe3)H, indexing was performed from a series of
1° oscillation images with exposures of 180 s per frame. A hemisphere
of data was collected using 3° oscillation angles with exposures of 100
s per frame and a crystal-to-detector distance of 82 mm. A total of
17 931 reflections were measured over the ranges 5.0 e 2θ e 50.7°,
-18 e h e 18, -12 e k e 12, -34 e l e 35. The intensity data were
corrected for Lorentz and polarization effects but not for absorption.
All 4431 unique reflections (Rint ) 0.0267) were used during subsequent
structure refinement (195 parameters refined.) The structure was solved
by Patterson methods (DIRDIF94).66 Refinement was by full-matrix
least squares based on F 2 using SHELXL-93.65 The weighting scheme
1
by H NMR spectroscopy. H2 was detected after 5 min at 25 °C, and
HD was detected after 3 h at 25 °C. The ratio of H2:HD was ca. 1:1
after 24 h at 25 °C.
Reaction of (PMe3)3Ru(SiEt3)H3 with D2. An NMR tube was
loaded with a C6D6 solution (0.5 mL) of (PMe3)3Ru(SiEt3)H3 (20 mg,
0.049 mmol), and the sample was degassed in vacuo. Deuterium (ca.
3 atm) was added, and the tube was sealed. The reaction was monitored
1
by H NMR spectroscopy. Only trace amounts of H2 and HD were
detected after 13 days at room temperature.
used was w ) 1/[σ2(F0 ) + 0.0524P 2 + 7.3452P], where P ) (F0
+
2
2
Exchange Reactions of (PMe3)3Ru(SiR3)H3 with HSiR3′. The
general method is described here for the reaction of (PMe3)3Ru-
(SiMe3)H3 and HSiMe2CH2SiMe3: An NMR tube was loaded with a
C6D6 solution (0.5 mL) of (PMe3)3Ru(SiMe3)H3 (12 mg, 0.03 µmol)
and HSiMe2CH2SiMe3 (ca. 0.032 mmol) and sealed under vacuum. The
reaction was monitored by 1H NMR. Although no reaction was apparent
after 0.5 h at 25 °C, a ca. 9:1 ratio of (PMe3)3Ru(SiMe3)H3:(PMe3)3-
Ru(SiMe2CH2SiMe3)H3 was present after 24 h. This ratio did not change
at longer reaction times.
Single-Crystal X-ray Diffraction Analyses of (PMe3)4Ru(SiMe3)-
Me (7a), (PMe3)4Ru(SiMe3)H (1a), and (PMe3)4Ru(SiMe2CH2SiMe3)H
(1b). X-ray intensity data were collected on a Rigaku R-AXIS IIc area
detector employing graphite-monochromated Mo KR radiation (λ )
0.71069 Å). Oscillation images were processed using bioteX,62 produc-
ing a listing of unaveraged F2 and σ(F2) values which were then passed
to the teXsan63 program package for further processing and structure
solution on a Silicon Graphics Indigo R4000 computer.
2Fc2)/3. Non-hydrogen atoms were refined anisotropically, and hydrogen
atoms were included as constant contributions to the structure factors
and were not refined. The maximum ∆/σ in the final cycle of least
squares was -0.001, and the two most prominent peaks in the final
difference Fourier were +0.458 and -0.539 e/Å3.
Single-Crystal X-ray Diffraction Analyses of (PMe3)4Ru(SiMe2H)H
(1d) and (PMe3)3Ru(SiMe3)H3 (4a). X-ray intensity data were collected
on an Enraf-Nonius CAD4 diffractometer employing graphite-mono-
chromated Mo KR radiation (λ ) 0.71073 Å) and using the ω-2θ scan
technique. The cell constants were determined from a least-squares fit
of the setting angles for 25 accurately centered reflections. Three
standard reflections measured every 3500 s of X-ray exposure showed
no intensity decay over the course of data collection.
For (PMe3)4Ru(SiMe2H)H, a total of 5945 reflections were measured
over the ranges 4.0 e 2θ e 55.0°, 0 e h e 11, 0 e k e 37, -12 e
l e 12. The intensity data were corrected for Lorentz and polarization
effects, and an empirical absorption correction was applied. Of the 5482
unique reflections measured, a total of 3777 reflections (Rint ) 0.023)
with F 2 > 3σ(F 2) were used during subsequent structure refinement
For (PMe3)4Ru(SiMe3)Me, indexing was performed from a series
of four 1° oscillations with exposures of 200 s per frame. A hemisphere
of data was collected using 8° oscillations with exposures of 200 s per
frame and a crystal-to-detector distance of 82 mm. A total of 18 969
reflections were measured over the ranges: 5.02 e 2θ e 50.00°, -17
(64) SIR92: Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, M.; Gioco-
vazzo, C.; Guagliardi, A.; Polidoro, G. J. Appl. Crystallogr. 1994, 27, 435.
(65) Sheldrick, G. M. SHELXL-93: Program for the Refinement of Crystal
Structures; Go¨ttingen University: Go¨ttingen, Germany, 1993.
(66) Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman, W. P.; de Gelder,
R.; Israe¨l, R.; Smits, J. M. M. The DIRDIF-94 Program System; Crystal-
lography Laboratory, University of Nijmegen: The Netherlands, 1994.
(62) bioteX: A Suite of Programs for the Collection, Reduction and Interpretation
of Imaging Plate Data; Molecular Structure Corp., 1995.
(63) teXsan: Crystal Structure Analysis Package; Molecular Structure Corp.,
1985 and 1992.
9
J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003 8057

A R T I C L E S
Dioumaev et al.
(189 parameters refined). The structure was solved by standard heavy
atom Patterson techniques followed by weighted Fourier syntheses.
Hydrogen atoms were found from difference Fourier maps calculated
after anisotropic refinement. Refinement was by full-matrix least squares
techniques based on F to minimize the quantity ∑w(|Fo| - |Fc|)2 with
w ) 1/σ2(F). Non-hydrogen atoms were refined anisotropically, the
Ru and Si hydrides (H1 and H2) were refined isotropically, and all
other hydrogen atoms were included as constant contributions to the
structure factors and were not refined. The maximum ∆/σ in the final
cycle of least squares was 0.001.
For (PMe3)3Ru(SiMe3)H3, a total of 5658 reflections were measured
over the ranges 4 e 2θ e 55°, 0 e h e 18, -12 e k e 12, 0 e l e
21. The intensity data were corrected for Lorentz and polarization
effects, and an empirical absorption correction was applied. Of the 5165
unique reflections measured, a total of 2842 reflections (Rint ) 0.023
with F 2 > 3σ(F 2) were used during subsequent structure refinement
(166 parameters refined). The structure was solved by standard heavy
atom Patterson techniques followed by weighted Fourier syntheses.
Hydrogen atoms were found from difference Fourier maps calculated
after anisotropic refinement. Refinement was by full-matrix least squares
techniques based on F to minimize the quantity ∑w(|Fo| - |Fc|)2 with
w ) 1/σ2(F). Non-hydrogen atoms were refined anisotropically, the
Ru hydrides (H1, H2, and H3) were refined isotropically, and all other
hydrogen atoms were included as constant contributions to the structure
factors and were not refined. The maximum ∆/σ in the final cycle of
least squares was 0.001.
Acknowledgment. We are grateful to the National Science
Foundation (CHE-9904798) and Tokyo Electron Massachusetts
for support of this work. We also thank Dr. Bernd Mayer for
his assistance with the characterization of (PMe3)4Ru(SiMe2H)H.
Supporting Information Available: X-ray crystallographic
data for 1a, 1b, 1d, 4a, and 7a in CIF format. This material is
available free of charge via the Internet at http://pubs.acs.org.
JA035131P
9
8058 J. AM. CHEM. SOC. VOL. 125, NO. 26, 2003