Room-temperature hydrosilylation of the CF bond of vinyl fluoride catalyzed by osmium hydrides

Article abstract of DOI:10.1016/j.molcata.2004.07.028

The reactions of OsH(Ph)(CO)L₂ (L = P^tBu ₂Me) with HSiMe₃, H₂SiPh₂ or H ₃SiPh produce benzene and OsH₃(silyl)(CO)L₂, which is characterized as a fac-H₃OsP₃ shape capped on the H₃ face by the silyl group. For H₂SiPh₂ and H₃SiPh, these reactions are shown proceed through an intermediate Os(H)₂(silylene)(CO)L₂ species. OsHF(CO)L₂ reacts with these silanes to give the same OsH₃ product (and silyl-F). These reactions can be combined to effect osmium catalysis of conversion of SiH + H₂CCHF to fluorosilane and H₂CCH ₂ from OsH₃(SiMe₃)(CO)L₂, or from the catalyst precursor OsHF(CO)L₂.

Full text of DOI:10.1016/j.molcata.2004.07.028

Journal of Molecular Catalysis A: Chemical 224 (2004) 125–131
Room-temperature hydrosilylation of the C F bond of vinyl fluoride
catalyzed by osmium hydrides
Kenton B. Renkema, Ulrike Werner-Zwanziger, Mark D. Pagel, Kenneth G. Caulton∗
Department of Chemistry, Indiana University, Bloomington, IN 47405-7102, USA
Received 1 March 2004; received in revised form 21 July 2004; accepted 24 July 2004
Available online 19 October 2004
Dedicated to Prof. J.J. Zio´łkowski, on the occasion of his 70th birthday, for the international character he brought to chemistry in Wrocław.
Abstract
The reactions of OsH(Ph)(CO)L2 (L = PtBu2Me) with HSiMe3, H2SiPh2 or H3SiPh produce benzene and OsH3(silyl)(CO)L2, which is
characterized as a fac-H3OsP3 shape capped on the H3 face by the silyl group. For H2SiPh2 and H3SiPh, these reactions are shown proceed
through an intermediate Os(H)2(silylene)(CO)L2 species. OsHF(CO)L2 reacts with these silanes to give the same OsH3 product (and silyl-
F). These reactions can be combined to effect osmium catalysis of conversion of Si H + H2C=CHF to fluorosilane and H2C=CH2 from
OsH3(SiMe3)(CO)L2, or from the catalyst precursor OsHF(CO)L2.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Silanolysis; Fluorocarbon; Osmium; Hydride
1. Introduction
2. Results
Defluorination of fluorocarbons [1–12] is an important
current challenge. Transition metal complexes have demon-
strated activity in catalysis with silanes [13–17]. The conver-
sion of C F with M H is one approach, and would form C H
and M F. If M F could then be transformed back into M H,
a catalytic cycle could be achieved (Scheme 1) for the overall
reaction in Eq. (1). An attractive choice of E is silicon, since
the thermodynamic stability of the Si F bond could provide
2.1. Background
In separate work [19], we found that OsH(Ph)(CO)L2
reacted cleanly with vinyl fluoride at 25 ◦C to give ethy-
lene and OsF(Ph)(CO)L2. Thus, in this molecule, the phenyl
functions as a “spectator” and the observed reaction is sim-
ply metathesis between C F and Os H bonds. This accom-
plishes step a in Scheme 1. What must next be accomplished
is the transfer of this abstracted fluoride from osmium to
silicon.
“M”
C F+E H −→ C H+E F
(1)
the driving force for this reaction [5,18].
We explore here the implementation of this idea for an
F C(sp2) bond, that of vinyl fluoride. This will include a
comparative survey of the utility of several silanes, as well as
a study of catalytic intermediates and several catalyst precur-
sors, all based on a catalyst core furnished by the Os(CO)L2
fragment, where L = PtBu2Me.
2.2. Regeneration of Os H
ThefluorineinOsF(Ph)(CO)L2 (L = PtBu2Me)isreplaced
by H upon treatment with 23 equiv. of HSiMe3 (Eq. (2)).
Formation of OsH(Ph)(CO)L2 is complete in
OsF(Ph)(CO)L2 + HSiMe3 → OsH(Ph)(CO)L2+ Me3SiF
∗
Corresponding author. Tel.: +1 812 855 4798; fax: +1 812 855 8300.
E-mail address: caulton@indiana.edu (K.G. Caulton).
(2)
1381-1169/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2004.07.028

126
K.B. Renkema et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 125–131
Scheme 1.
3 h at 25 ◦C in benzene. Because of the excess HSiMe3,
the reaction proceeds farther (over a 9.5 h period) to give
OsH3(SiMe3)(CO)L2 (Eq. (3)). The facility of this reaction
is surprising since
Scheme 2.
2.4. Structure and dynamics of OsH3(SiMe3)(CO)L2
and its relation to catalysis
OsH(Ph)(CO)L2 + 3HSiMe3 → OsH3(SiMe3)(CO)L2
+ C6H6 + (Me3Si)2 (3)
it would appear to require some intermediate with two bulky
SiMe3 ligands attached to an osmium which also carries two
bulky phosphines [20].
Since OsH3(SiMe3)(CO)L2 is an 18-electron complex,
we sought full characterization in order to better under-
stand its reactivity. At 20 ◦C, the 1H NMR spectrum of
OsH3(SiMe3)(CO)L2 shows two hydride signals (2:1 inten-
sity ratio) at −9.5 ppm (broad singlet) and −10.3 ppm (broad
triplet). At −70 ◦C, the singlet decoalesced into a non-first-
orderfive-linepatternandthetripletwasresolvedintoatriplet
of triplets (JPH = 19.5 Hz; JHH = 4.5 Hz), as shown in Fig. 1.
At −70 ◦C, the selectively hydride-coupled 31P NMR spec-
trum (Fig. 1) showed a 10-line pattern that was essentially a
doubling of the five-line 1H NMR pattern seen at −9.5 ppm.
These NMR patterns identify the H3P2 part of the molecule as
an AA‘MXXꢀ spin system, which implies mirror symmetry
for the molecule. Among 22 possible line assignments for the
observed spectra [21], one was distinctly better than all oth-
ers, and yielded the coupling constants shown in Scheme 2.
These J values are structurally diagnostic. Based on the gen-
eral trends of JPP and JHP with their mutual angles, we can
conclude:
2.3. Catalytic reactivity of OsH3(SiMe3)(CO)L2
Fluorine abstraction from vinyl fluoride is possible with
OsH3(SiMe3)(CO)L2. This saturated species is thus a com-
petent catalyst precursor. When excess (36 equiv.) vinyl flu-
oride is added to a benzene solution of OsH3(SiMe3)(CO)L2
and 24 equiv. HSiMe3, catalytic formation of C2H4 and
Me3SiF occurs at a rate of 24 turnovers within 0.5 h. Defluori-
nation of H2C CF2 is also effected by OsH3(SiMe3)(CO)L2
in the presence of excess (35 equiv.) HSiMe3 in benzene at
25 ◦C. After 1 h, the major product was ethylene, although
vinyl fluoride was also detected. For comparison, under
these conditions in the absence of the Os complex, HSiMe3
was shown to be unreactive towards either vinyl fluoride or
H2CCF2 over 24 h.
(a) the phosphines are mutually cis (30 Hz is a very small
J
PP),
Fig. 1. Observed (upper) and calculated variable-temperature 300 MHz 1H (left) and selectively hydride-coupled 31P (122 MHz) NMR spectra of
OsH3(SiMe3)(CO)(PtBu2Me)2 in d8-toluene.

K.B. Renkema et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 125–131
127
ꢀ
ꢀ
(b) neither HA nor HA is approximately trans to PX or PX ,
catalytic cycle is thus OsH2(CO)L2. This species was also
invoked recently based on observations of the PPh3 analog
[29].
(c) hydrogens A, Aꢀ and M are cisoid; no two are mutually
trans.
We propose two mirror symmetric shapes, 1 and 2, that can
accommodate these conditions. 1 is based on a face-capped
ꢀ
octahedron [22,23], and the angle HA Os HA must deviate
2.5. SiH2Ph2
much from 90◦ to avoid A trans to Xꢀ, etc. 2 is a triangular
face-capped trigonal prism, and thus
The benzene elimination reaction with HSiMe3 in Eq.
(3) effectively converts OsH(Ph)(CO)L2 into the synthetic
equivalent of “Os(CO)L2”. Converting this into a trihydride
monosilyl product requires three moles of a tertiary silane (cf.
Eq. (3)), and so some intermediate osmium complex is impli-
cated by this 1:3 stoichiometry. To investigate this point, the
secondary silane SiH2Ph2 was reacted with OsH(Ph)(CO)L2
in a 1:1 mole ratio in toluene at 20 ◦C (Eq. (5))
naturally has no trans relationships. None of our observa-
tions allows any conclusion about incipient Si/HM bonding
[24,25].
The temperature dependence of the coalescence of the A
and Aꢀ hydrogens from AA‘MXXꢀ to A2MX2 at five tem-
peratures in the range −70 to −20 ◦C (Fig. 1) was simulated
using the program GNMR. An Eyring plot of ln(k/T) versus
T−1 gave ꢀH‡ = 7.6 kcal/mol and ꢀS‡ = −16.6 cal/deg mol.
This entropy change is inconsistent with a dissociative pro-
cess [26,27].
The three hydride signals coalesce to a single line
∼0.8 ppmbroadat60 ◦C. WhiletheseNMRresultshelptoes-
tablish the structure and dynamics of OsH3(SiMe3)(CO)L2,
they alone fail to reveal whether a 16-electron species re-
active towards vinyl fluoride is formed by reductive elimi-
nation of H2 or of HSiMe3. The latter is favored by steric
effects, but the former would produce OsH(SiMe3)(CO)L2,
whose ruthenium analog is known: RuH(SiHPh2)(CO)L2
[28].
After 1 h, there was 90% conversion to a species identified
as 3. Although hydride (two chemical shifts) and 31P{1H}
NMR signals were broad at 20 ◦C, by −60 ◦C the 31P{1H}
NMR signal was a sharp singlet and the hydride region was
resolved into two triplets of doublets; these two showed
equal intensity. Both doublet couplings were 3.9 Hz, sup-
porting the intensity conclusion that these hydrides are in
the same molecule (i.e. mutually coupled), and mutually
cis. This was confirmed by observing a selectively hydride
coupled 31P NMR spectrum (at −50 ◦C); the spectrum was
a doublet of doublets with J values (25 and 19 Hz) which
agreed well with the triplet splitting in the hydride signals.
The 29Si NMR spectrum at −40 ◦C is a broadened (due to
unresolved coupling) line at 328 ppm, which identifies this
as a silylene ligand. For comparison [30], SiR2 (R = Me or
Ph) ligands on Ru or Ir have 29Si chemical shifts in the range
250–350 ppm. A 31P{1H} NMR spectrum collected for high
S/N ratio at −40 ◦C showed satellites revealing JPSi = 20 Hz
and JPOs = 82 Hz, thus confirming connectivity and compo-
sition consistent with formulation Os(H)2(SiPh2)(CO)L2. It
is especially diagnostic of the reducing power of osmium in
this molecule that the ruthenium analog [28] is the valence
isomer RuH(SiHPh2)(CO)(PtBu2Me)2, where the Ph2Si/H
bond has not oxidatively added to the metal. However, the
dynamic process evident in 3 by NMR at 20 ◦C may be H
migration from Os to Si.
To test for Eq. (4) as the source of an unsaturated catalytic
reactant, the 1H NMR spectra
OsH3(SiMe3)(CO)L2 ꢀ OsH2(CO)L2 + HSiMe3
(4)
of a 1:3 mixture of OsH3(SiMe3)(CO)L2 and HSiMe3
was recorded in the temperature range 20–70 ◦C.
By 60 ◦C, when the two osmium hydride signals of
OsH3(SiMe3)(CO)L2 have just coalesced (thus, their
exchange rate is ∼3 × 102 s−1), the line width of the SiH
signal of free HSiMe3 shows exchange broadening with
a rate of ∼3 × 101 s−1). This suggests that OsH3 site
exchange has a lower barrier (which governs the dynamic
NMR behavior of OsH3(SiMe3)(CO)L2 in the absence of
added HSiMe3), but that the (slower, more energetically
demanding) HSiMe3 dynamic process is that of Eq. (4).
The intramolecular character of the OsH3 site exchange
is established by the retention of H3 Os P coupling
above the OsH3 coalescence temperature for the SiHPh2
and SiH2Ph analogs (see below). A viable species on a
This implied ␣-H migration of Os SiHPh2 to HOs SiPh2
in Eq. (5) is caused by reaction conditions where there is
a deficiency of (silane-derived) ligands (e.g., hydride). In-
deed, when OsH(Ph)(CO)L2 is reacted with SiH2Ph2 in a
1:2.5 mole ratio, the final product, OsH3(SiHPh2)(CO)L2 ap-
pears within 1 h (Eq. (6)), although Os(H)2(SiPh2)(CO)L2 is
present at earlier observation times, and is thus an interme-

128
K.B. Renkema et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 125–131
diate in formation of the trihydride. The resulting
bulky silane molecules
Os(H)2(SiPh2)(CO)L2 + 2SiH2Ph2
→ OsH3(SiHPh2)(CO)L2 + (SiHPh2)2
(6)
OsH3(SiHPh2)(CO)L2 has a lower barrier to hydride site ex-
change than does the SiMe3 analog, since all three hydrides
are coalesced to a sharp triplet at 20 ◦C, and they broaden,
but do not decoalesce at −80 ◦C.
2.8. Confirmation of catalytic conversion of vinyl
fluoride to ethylene with OsHF(CO)L2 as catalyst
precursor
2.6. SiH3Ph
Reaction of OsH(Ph)(CO)L2 with two moles of SiH3Ph
at 25 ◦C completely consumes the Os reagent in 1 h to give
two products, which are assigned as Os(H)2(SiHPh)(CO)L2
and OsH3(SiH2Ph)(CO)L2. This follows in part from their
similar 31P chemical shifts to their analogs from SiH2Ph2.
The latter compound showed a sharp hydride triplet at
20 ◦C, consistent with rapid intramolecular fluxionality.
Os(H)2(SiHPh)(CO)L2 showed, at 20 ◦C, two broad hydride
peaks and one broad 31P NMR signal.
The catalyst is most conveniently prepared within 1 h by
reacting OsHF(CO)L2 with excess silane in C6D6 at 25 ◦C.
The production of OsH3(silyl)(CO)L2 was in every case con-
firmed by 31P, H and 19F NMR spectroscopy. Following
1
addition of vinyl fluoride (vinyl fluoride:Os ≥ 36:1), the ap-
pearance of products and disappearance of reactants (Eq. (9))
was monitored (NMR) over time. The reaction with HSiMe3
converted 24 equiv. silane
[Os]
H2C CHF + HSi −→ H2CCH2 + FSi
(9)
2.7. Alternative synthesis of OsH3(silyl)(CO)L2
within 30 min. Consumption of 6 equiv. HSiPh3 required
3 h, while SiH2Ph2 and SiH3Ph were slower still. Heat-
ing to 60 ◦C consumed 9.4 equiv. SiH2Ph2 in 30 min. The
final osmium-containing product after consumption of all
the silane (HSiMe3 or SiH2Ph2) was always OsHF(CO)(␩2-
H2 CCHF)L2, and this was the only species seen through-
out the catalysis using SiHPh3. Catalyst decomposition was
insignificant except for the SiH3Ph example at elevated tem-
perature.
Since the phenyl group of OsH(Ph)(CO)L2 is not inte-
gral to the catalysis of Eq. (1), we sought a simpler access
to the catalytic cycle. When OsHF(CO)L2 is reacted with
the silanes SiHPh3, SiH2Ph2 or SiH3Ph (1:2–8 mole ratio),
at 20 ◦C in benzene, there is rapid (<30 min) conversion to
OsH3(silyl)(CO)L2. The replaced fluoride is established by
19F NMR spectroscopy to be on silicon, and the released pri-
mary and secondary fluorosilanes show evidence for F/H re-
distribution to generate all possible multi-fluoro species (e.g.,
SiHnF3−nPh). The reaction applies equally to OsHF(CO)L2
and HSiMe3, and all OsH3(silyl)(CO)L2 analogs show 31P
NMR chemical shifts within a narrow (4 ppm) range, and hy-
dride chemical shifts within a 0.5 ppm range. What differs
is the rate of hydride scrambling in OsH3(silyl)(CO)L2 at
20 ◦C: the SiPh3 and SiMe3 examples show two signals (2:1
intensity) while the SiHPh2 and SiH2Ph examples show a
coalesced spectrum for three H. Because the coalesced spec-
tra show triplet coupling to phosphorus, this process is in-
tramolecular, and involves no ligand loss.
3. Discussion
The structure of OsH3(silyl)(CO)(PtBu2Me)2 deduced
here shares only certain features with the structure for
OsH3(SiHPh2)(CO)(PiPr3)2 derived from X-ray diffraction
and from DFT calculations [31]. Central among these is
1
the pyramidal Os(CO)P2 substructure. The H NMR spec-
tral data presented by Buil et al. show 2:1 intensity for
hydride ligands, but incipient 31P NMR decoalescence
was detected by 193 ◦K. This contrasts to what we ob-
serve for OsH3(SiMe3)(CO)(PtBu2Me)2. Buil et al. re-
port two inequivalent Si Os P angles, and thus no mirror
symmetry to the structure. While our observations might
suffer a yet-undetected (i.e., low barrier) fluxional pro-
cess, giving the illusion of mirror symmetry, the fact that
OsH3(SiMe3)(CO)(PtBu2Me)2 has a higher barrier for hy-
dride site exchange than OsH3(SiHPh2)(CO)(PiPr3)2 makes
this less probable. The mirror symmetry deduced here for
OsH3(SiMe3)(CO)(PtBu2Me)2 is in full agreement with
that reported [29] for the PPh3 analog as well as for sev-
eral RuH3(SiR3)(PR3)3 cases [20,22,23]. Thus, we suggest
that the lack of mirror symmetry in the X-ray structure of
The conversion of OsHX(CO)L2 to OsH3(silyl)(CO)L2
by an Si H reagent is faster for X = F than for X = Ph. This
acceleration probably involves incipient F/Si bonding in the
mechanism. Also consistent with our interpretation is the lack
of silylene intermediate for the OsHF(CO)L2 route. There
(Eq. (7)) the needed intermediate (A, Eq. (8)) for silylene
formation is absent and final product derives from
onemorestep. Thesilylenecomplexcanaccumulatesincethe
final product in Eq. (8) requires reaction with two additional

K.B. Renkema et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 125–131
129
Scheme 3.
OsH3(SiHPh2)(CO)(PiPr3)2 may be induced by the very dif-
acted with HSiMe3 (5.8 × 10−4
,
23 equiv.) conver-
ferent steric demands of the Ph and H substituents on Si. The
published 1H NMR data [31] showed no clear evidence for
three different hydride environments, but are fully consistent
with mirror symmetry.
A highly schematic portrayal of the observations reported
here appears in Scheme 3, which indicates why there are nu-
merous ways to enter the catalytic cycle. Both the relationship
to and the contrast to analogous ruthenium chemistry is ev-
ident since RuH3(SiHPh2)(CO)(PtBu2Me)2 has been shown
[32] to exchange with added free SiH2Ph2, but also to lose H2
under vacuum, to furnish RuH(SiHPh2)(CO)(PtBu2Me)2.
sion first to OsH(Ph)(CO)(PtBu2Me)2 and then to
OsH3(SiMe3)(CO)(PtBu2Me)2 was accomplished over
two days. 1H NMR (C6D6, 293 K): 1.28 (overlapping
vt, 18H each, PCCH3, JHP = 12 Hz); 1.41 (vt, 6H, PCH3,
JPH = 2.8 Hz); 0.59 (s, 9H, SiCH3); −9.41 (br s, 2H, OsH);
−10.17 (br t, 1, OsH, JHP = 19.9 Hz). 31P NMR (C6D6,
293 K): 29.9 (s). Cooling the sample in d8-toluene reduced
chemical exchange such that more coupling could be seen
1
in H and partially coupled 31P NMR. In addition to the
parameters given in the text, the signals for PCCH3 are
found at 1.17 and 1.14 as broad singlets and PCH3 at 1.03,
also as a broad singlet.
4. Experimental
4.3. OsH(Ph)(CO)(PtBu2Me)2 + HSiMe3
4.1. General
When OsH(Ph)(CO)(PtBu2Me)2 (9.4 mg, 1.52 ×
10−5 mol) in 500 ␮L d8-toluene was reacted with HSiMe3
(6.2 equiv.) at 20 ◦C for 12 h, greater than 90% conversion to
OsH3(SiMe3)(CO)(PtBu2Me)2 (vide supra) was observed.
31P NMR did show a small amount of an intermediate
species (at 42.8 ppm), which was present from beginning to
end; we suggest it is OsH2(SiMe3)2(CO)(PtBu2Me)2.
All manipulations were done under an atmosphere of dry,
O2-free Ar employing a vacuum atmospheres inert atmo-
sphere glovebox or standard Schlenk-line techniques. The
solvents were reagent grade, and distilled from the appropri-
ate drying agents under Ar. A lecture bottle of HSiMe3 and
solid HSiPh3 were obtained from Aldrich and used directly.
H2SiPh2 and H3SiPh were also obtained from Aldrich, but
were dried and distilled before use. Vinyl fluoride and gem-
difluoroethylene were purchased from Lancaster Chemicals.
All other reagents were used without further purification.
OsH(Ph)(CO)(PtBu2Me)2 [33] and OsHF(CO)(PtBu2Me)2
4.4. OsH(Ph)(CO)(PtBu2Me)2 + HSiPh3
When OsH(Ph)(CO)(PtBu2Me)2 (7.1 mg, 1.15 ×
10−5 mol) in 500 ␮L C6D6 was reacted with HSiPh3
(6.4 mg, 2.46 × 10−5 mol) at 20 ◦C for 12 h, greater than
90% conversion to OsH3(SiPh3)(CO)L2 (vide supra) was
observed. 31P NMR again showed an intermediate present
throughout the conversion process (at 30.8 ppm) which is
assigned as OsH2(SiPh3)2(CO)(PtBu2Me)2.
1
[34] were synthesized according to literature. The H, 31P,
19F, and 29Si NMR were collected on Varian Gemini 2000 or
Inova 400 spectrometers (1H: 300, 400 MHz; 13C: 100 MHz;
31P: 122, 162 MHz; 19F: 282, 376 MHz; 29Si: 79.5 MHz).
Proton NMR spectra were referenced to residual solvent
peaks as internal standards. 31P NMR was referenced to an
external standard of 85% H3PO4. 19F NMR was referenced
to trifluoroacetic acid in benzene.
4.5. Os(H)2( SiPh2)(CO)(PtBu2Me)2
When OsH(Ph)(CO)(PtBu2Me)2 (8.4 mg, 1.36 ×
10−5 mol) dissolved in 500 ␮L d8-toluene was reacted
with H2SiPh2 (6.2 ␮L, 3.30 × 10−5 mol) at 20 ◦C for 1 h,
two products were formed. The kinetic product consti-
tuted about 80% after 1 h and was assigned to the title
4.2. Os(Ph)F(CO)(PtBu2Me)2 + HSiMe3
When a solution of Os(Ph)F(CO)(PtBu2Me)2 (15.8 mg,
2.5 × 10−5 mol) in 500 ␮L cyclohexane-d12 was re-

130
K.B. Renkema et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 125–131
4.10. OsH3(SiH2Ph)(CO)(PtBu2Me)2
compound using the following NMR. 1H NMR (C7D8,
233 K): selected aryl peaks (8.05, 7.32, 7.29, 7.25); 1.48
(br s, 6H, PCH3); 1.21 (vt, 18H, PCCH3, JHP = 5.9 Hz);
1.14 (vt, 18H, PCCH3, JHP = 5.9 Hz); −8.98 (dt, 1H,
OsH, JHP = 19.0 Hz, JHH = 3.9 Hz); −10.07 (dt, 1H, OsH,
When OsHF(CO)(PtBu2Me)2 (5.2 mg, 9.31 × 10−6 mol)
dissolved in 500 ␮L C6D6 was reacted with H3SiPh (2.5 ␮L,
2.02 × 10−5 mol) at 20 ◦C for 15 min, the title compound
1
J
HP = 25.2 Hz, JHH = 3.9 Hz). 31P NMR (C7D8, 233 K):
was produced in greater than 95% yield. H NMR (C6D6,
293 K): selected aryl peaks (8.12, 7.50, 7.31, 7.14, 7.08);
5.75 (t, 1H, SiH, JPH = 4.5 Hz); 1.17 (vt, 36H, PCCH3,
45.5 (s, Os and Si satellites: JOsP = 82 Hz; JSi P = 19.7 Hz);
31P{hydride coupled, 223 K} 45.1 (dd, 2P, JPH = 25.2 Hz,
JPH = 19.0 Hz). 29Si NMR (C7D8, 233 K): 328 (br).
J
HP = 6.1 Hz); 1.35 (vt, 6H, PCH3, JHP = 3.0 Hz); −8.9 (t,
3H, OsH, JHP = 15.8 Hz). 31P NMR (C6D6, 293 K): 29.4 (s).
4.6. OsH3(SiMe3)(CO)(PtBu2Me)2
4.11. Catalytic conversions of vinyl fluoride
4.11.1. OsH3(SiMe3)(CO)(PtBu2Me)2 + H2C CHF +
HSiMe3
When OsHF(CO)(PtBu2Me)2 (9.0 mg, 1.61 × 10−5 mol)
dissolved in 500 ␮L C6D6 was reacted with HSiMe3
(23.0 equiv.) at 20 ◦C for 1 h, conversion to the title com-
pound was accomplished.
When OsH3(SiMe3)(CO)(PtBu2Me)2 (9.0 mg, 1.69 ×
10−5 mol) in 500 ␮L C6D6 was reacted with H2C CHF
(36 equiv.) in the presence of 24 equiv. of HSiMe3
at 20 ◦C for 30 min, nearly all HSiMe3 had been
consumed. The production of C2H4, FSiMe3, and
OsHF(CO)(␩2-C2H3F)(PtBu2Me)2 (δ(OsH) = −3.0 (br t,
4.7. OsH(Ph)(CO)(PtBu2Me)2 + H3SiPh
When OsH(Ph)(CO)(PtBu2Me)2 (6.1 mg, 1.09 × 10−5
mol) in 500 ␮L C6D6 was reacted with H3SiPh for 1 h at
20 ◦C, two new products, OsH3(SiH2Ph)(CO)(PtBu2Me)2
(vide supra) and OsH2(SiHPh)(CO)(PtBu2Me)2 (δ(31P) =
42.5 ppm, br s; δ(Os H) = −9.6 ppm, br), free phosphine,
and the complete consumption of OsH(Ph)(CO)(PtBu2Me)2
were observed. Another hour of stirring caused the
conversion of OsH2(SiHPh)(CO)(PtBu2Me)2 to OsH3
(SiH2Ph)(CO)(PtBu2Me)2.
J = 28 Hz); δ(31P) = 22.4 (br d); δ(19F) = −171.3 (ddd,
1
J
HF = 73, 19, 21 Hz) were seen by H and 19F NMR spec-
troscopy.
4.11.2. OsH3(SiPh3)(CO)(PtBu2Me)2 + H2C CHF +
HSiPh3
When OsH3(SiPh3)(CO)(PtBu2Me)2 (7.1 mg, 1.27 ×
10−5 mol) in 500 ␮L C6D6 was reacted with HSiPh3
(41.0 mg, 12.4 equiv.) and H2C CHF (36 equiv.) for 3 h at
20 ◦C, nearly half of the silane had been consumed. The
observation of FSiPh3 and C2H4 confirmed the Si H/C F
metathesis. At all times during the process, the only os-
mium species that could be seen was OsHF(CO)(␩2-
H2C CHF)(PtBu2Me)2.
4.8. OsH3(SiPh3)(CO)(PtBu2Me)2
When OsHF(CO)(PtBu2Me)2 (7.1 mg, 1.27 × 10−5 mol)
dissolved in 500 ␮L C6D6 was reacted with HSiPh3 (5.9 ␮L,
2.27 × 10−5 mol) at 20 ◦C for 30 min, conversion to the ti-
4.11.3. OsH3(SiHPh2)(CO)(PtBu2Me)2 + H2C CHF +
H2SiPh2
1
tle compound was accomplished. H NMR (C6D6, 293 K):
When OsH3(SiHPh2)(CO)(PtBu2Me)2 (6.4 mg, 1.15 ×
10−5 mol) in 500 ␮L C6D6 was reacted with C2H3F
(36 equiv.) and H2SiPh2 (20 ␮L, 9.4 equiv.) for 30 min at
20 ◦C, almost no conversion was observed. However, af-
ter 1 h at 60 ◦C, the silane was fully consumed, pro-
ducing C2H4, FSiH2Ph, and (exclusively) OsHF(CO)(␩2-
C2H3F)(PtBu2Me)2.
8.06 (d, 6H, JHH = 6.9 Hz); 7.6 (d, 3H, JHH = 6.5 Hz); 7.3
(t, 6H, JHH = 6.9 Hz); 1.24 (vt, 6H, PCH3, JHP = 2.9 Hz);
1.10 (vt, 18H, PCCH3, JHP = 6.7 Hz); 1.00 (vt, 18H, PCCH3,
J
J
HP = 6.7 Hz); −8.59 (s, 2H, OsH); −9.7 (br t, 1H, OsH,
HP = 21 Hz). 31P NMR (C6D6, 293 K): 27.50 (s).
4.9. OsH3(SiHPh2)(CO)(PtBu2Me)2
4.11.4. OsH3(SiH2Ph)(CO)(PtBu2Me)2 + H2C CHF +
H3SiPh
When OsHF(CO)(PtBu2Me)2 (6.3 mg, 1.13 × 10−5 mol)
dissolved in 500 ␮L C6D6 was reacted with H2SiPh2 (20 ␮L,
1.07 × 10−4 mol) at 20 ◦C for 30 min, >95% conversion to
the title compound was observed. 1H NMR (C6D6, 293 K):
selected aryl peaks (8.00, 7.27, 7.08, 7.03); 6.67 (t, 1H,
SiH, JHP = 7.2 Hz); 1.16 (vt, 18H, PCCH3, JHP = 6.6 Hz);
1.14 (vt, 18H, PCCH3, JHP = 6.6 Hz); 1.06 (vt, 6H, PCH3,
JHP = 2.7 Hz); −9.1 (t, 3H, OsH, JHP = 15.6 Hz). 31P NMR
(C6D6, 293 K): 31.0 (s).
When OsH3(SiH2Ph)(CO)(PtBu2Me)2 (6.9 mg, 1.24 ×
10−5 mol) in 500 ␮L C6D6 was reacted with C2H3F
(40 equiv.) and H3SiPh (20 ␮L, 13.1 equiv.) at 20 ◦C for
3 h, almost no conversion (less than 5%) took place. When
the temperature was raised to 60 ◦C for 4 h, some of the
H2C CHF was converted to C2H4, but 31P NMR showed
significant degradation of the complex and 19F NMR showed
many signals.

K.B. Renkema et al. / Journal of Molecular Catalysis A: Chemical 224 (2004) 125–131
131
4.11.5. OsH3(SiMe3)(CO)(PtBu2Me)2 + HSiMe3 + 1,
1-C2H2F2
[5] M. Aizenberg, D. Milstein, Science 265 (1994) 359.
[6] M. Aizenberg, D. Milstein, J. Am. Chem. Soc. 117 (1995) 8674.
[7] B.L. Edelbach, W.D. Jones, J. Am. Chem. Soc. 119 (1997) 7734.
[8] Y. Ishii, C. Naoto, S. Yorimitsu, S. Murai, Chem. Lett. (1998)
157.
[9] R.P. Hughes, J.M. Smith, J. Am. Chem. Soc. 121 (1999) 6084.
[10] H. Yang, H.R. Gao, R.J. Angelici, Organometallics 18 (1999) 2285.
[11] R.J. Young, V.V. Grushin, Organometallics 18 (1999) 294.
[12] J.L. Kiplinger, T.G. Richmond, J. Am. Chem. Soc. 118 (1996) 1805.
[13] B. Marciniec, P. Krzyzanowski, E. Walczuk-Gusciora, W. Duczmal,
J. Mol. Catal. A: Chem. 144 (1999) 263.
When a solution of OsH3(SiMe3)(CO)(PtBu2Me)2
(10 mg, 1.6 × 10−5 mol) in 500 ␮L benzene-d6 was stirred
at room temperature with 35 equiv. of HSiMe3 and C2H2F2,
the products, C2H4, C2H3F, FSiMe3, and OsHF(CO)(␩2-
C2H2F2)(PtBu2Me)2 were seen after 1 h. The Os-containing
product was identified by a broad signal at 17 ppm in 31P
NMR. Conversion slowed with time (4 h).
[14] B. Marciniec, C. Pietraszuk, Organometallics 16 (1997) 4320.
[15] Y. Itami, B. Marciniec, M. Majchrzak, M. Kubicki, Organometallics
22 (2003) 1835.
4.12. Os-free control reactions
[16] B. Marciniec, C. Pietraszuk, M. Kujawa, J. Mol. Catal. A: Chem.
133 (1998) 41.
[17] J. Gulinski, B.R. James, B. Marciniec, J. Organomet. Chem. 499
(1995) 173.
[18] F. Stoehr, D. Sturmayr, U. Schubert, Chem. Commun. (2002) 2222.
[19] K.B. Renkema, K.G. Caulton, unpublished.
[20] V.K. Dioumaev, B.R. Yoo, L.J. Procopio, P.J. Carroll, D.H. Berry, J.
Am. Chem. Soc. 125 (2003) 8936.
[21] H. Guenter, Angew. Chem. 11 (1972) 861.
[22] V.K. Dioumaev, L.J. Procopio, P.J. Carroll, D.H. Berry, J. Am. Chem.
Soc. 125 (2003) 8043.
4.12.1. HSiMe3 + 1,1-C2H2F2
When a solution of HSiMe3 (3.8 × 10−4 mol) and 1,1-
C2H2F2 (3.8 × 10−4 mol) in 500 ␮L acetone-d6 was stirred
for a day at room temperature, H2C CHF, C2H4, and FSiMe3
were not observed.
4.12.2. HSiMe3 + H2C CHF
When a solution of HSiMe3 (3.8 × 10−4 mol) and
H2C CHF (3.8 × 10−4 mol) in 500 ␮L benzene-d6 was
stirred for a day at room temperature, no C2H4 or FSiMe3
were observed.
[23] N.M. Yardy, F.R. Lemke, L. Brammer, Organometallics 20 (2001)
5670.
[24] U. Schubert, Adv. Organomet. Chem. (1990) 151.
[25] J.Y. Corey, J. Braddock-Wilking, J. Chem. Rev. 99 (1999) 175.
[26] U. Schubert, S. Gilbert, S. Mock, Chem. Ber. 125 (1992) 835.
[27] M. Knorr, S. Gilbert, U. Schubert, J. Organomet. Chem. 347 (1988)
17.
Acknowledgment
[28] R.H. Heyn, J.C. Huffman, K.G. Caulton, New J. Chem. 17 (1993)
797.
[29] M. Mo¨hlen, C.E.F. Rickard, W.R. Roper, D.M. Salter, L.J. Wright,
J. Organomet. Chem. 593/594 (2000) 458.
This work was supported by the National Science Foun-
dation.
[30] P.W. Wanandi, P.B. Glaser, T.D. Tilley, J. Am. Chem. Soc. 122
(2000) 972.
[31] M.L. Buil, P. Espinet, M.A. Esteruelas, F.J. Lahoz, A. Lledo´s, J.M.
Mart´ınez-Iladuya, F. Maseras, J. Modrego, E. On˜ate, L.A. Oro, E.
Sola, C. Valero, Inorg. Chem. 35 (1996) 1250.
[32] D.G. Gusev, T.T. Nadasdi, K.G. Caulton, Inorg. Chem. 35 (1996)
6772.
[33] K.B. Renkema, R. Bosque, W.E. Streib, F. Maseras, O. Eisenstein,
K.G. Caulton, J. Am. Chem. Soc. 121 (1999) 10895.
[34] D.J. Huang, K.G. Caulton, J. Am. Chem. Soc. 119 (1997)
3185.
References
[1] J.L. Kiplinger, T.G. Richmond, G. Osterberg, Chem. Rev. 94 (1994)
373.
[2] J. Burdeniuc, R.H. Crabtree, Science 271 (1993) 340.
[3] Chemistry of Organic Fluorine Compounds, A Critical Review, ACS
Monograph 187, American Chemical Society, Washington, 1995.
[4] R.E. Banks, B.E. Smart, J.C. Tatlow, Organofluorine Chemistry:
Principles and Commercial Applications, Plenum Press, New York,
1994.