Manganese- and rhodium-catalyzed phenylsilane hydrosilation-deoxygenation of iron acyl complexes Cp(L)(CO)FeC(O)R (L = CO, PPh3, P(OMe)3, P(OPh)3; R = CH3, Ph, CHMe2, CMe3)

Article abstract of DOI:10.1021/om970659m

The manganese carbonyl acyl complexes L(CO)₄MnC(O)R (L = CO, R = CH₃ (2a); L = CO, R = Ph (2b); L = PPh₃, R = CH₃ (2c)) are precatalysts for the PhSiH₃ hydrosilation-deoxygenation of Cp(CO)₂FeC(O)CH₃ (1a) to Cp(CO)₂FeCH₂CH₃ (6a). Thus, 2a (4%) and 1.1 equiv of PhSiH₃ reduce 1a initially to a mixture of [Cp(CO)₂FeCH(CH₃)O]_3-xH_xSiPh (x = 2 (3b), 1 (4b), and 0 (5b)) (7-8 h), which transforms to a mixture of 5b (62%) and 6a (35%) (12 h). Similarly, 2b and PhSiH₃ transform 1a to a mixture containing 4% 3b, 22% 4b, 72% 5b, and 2% 6a (30 min), then up to 15-20% 6a (12 h). Use of 2c as the precatalyst selectively yields 71% 4b and 10% 6a (30 min), then up to 19% 6a (5 h). Both 4b and 5b were isolated by size-exclusion chromatography and characterized as mixtures of four stereoisomers, 4b as a pair of enantiomers having a prochiral center and two meso diastereoisomers with pseudoasymmetric centers and 5b as a diastereomeric pair of enantiomers. The PhSiH₃ (1.6 equiv)/RhCl(PPh₃)₃ (3%) system conveniently transforms Cp-(L)(CO)FeC(O)R to their alkyl derivatives Cp(L)(CO)FeCH₂R. Seven iron acyl complexes were studied (Cp(CO)₂FeC(O)R, R = Me, Ph, ⁱPr, ^tBu; Cp(L)(CO)FeC(O)CH₃, L = PPh₃, P(OMe)₃, and P(OPh)₃), which gave 49-87% isolated yields of Cp(L)(CO)FeCH₂R. Their modest yields correspond to the presence of byproducts, Cp(CO)₂FeCH=CH₂ (32%) and 4b (14%) with 1a and 10-15% of the vinyl complexes with L = PPh₃ and P(OPh)₃.

Full text of DOI:10.1021/om970659m

Organometallics 1998, 17, 1993-2002
1993
Ma n ga n ese- a n d Rh od iu m -Ca ta lyzed P h en ylsila n e
Hyd r osila tion -Deoxygen a tion of Ir on Acyl Com p lexes
Cp (L)(CO)F eC(O)R (L ) CO, P P h 3, P (OMe)3, P (OP h )3;
R ) CH3, P h , CHMe2, CMe3)
Zhibiao Mao, Brian T. Gregg, and Alan R. Cutler*
Department of Chemistry, Rensselaer Polytechnic Institute, Troy, New York 12180
Received J uly 30, 1997
The manganese carbonyl acyl complexes L(CO)4MnC(O)R (L ) CO, R ) CH3 (2a ); L )
CO, R ) Ph (2b); L ) PPh3, R ) CH3 (2c)) are precatalysts for the PhSiH3 hydrosilation-
deoxygenation of Cp(CO)2FeC(O)CH3 (1a ) to Cp(CO)2FeCH2CH3 (6a ). Thus, 2a (4%) and
1.1 equiv of PhSiH3 reduce 1a initially to a mixture of [Cp(CO)2FeCH(CH3)O]3-xHxSiPh (x
) 2 (3b), 1 (4b), and 0 (5b)) (7-8 h), which transforms to a mixture of 5b (62%) and 6a
(35%) (12 h). Similarly, 2b and PhSiH3 transform 1a to a mixture containing 4% 3b, 22%
4b, 72% 5b, and 2% 6a (30 min), then up to 15-20% 6a (12 h). Use of 2c as the precatalyst
selectively yields 71% 4b and 10% 6a (30 min), then up to 19% 6a (5 h). Both 4b and 5b
were isolated by size-exclusion chromatography and characterized as mixtures of four
stereoisomers, 4b as a pair of enantiomers having a prochiral center and two meso
diastereoisomers with pseudoasymmetric centers and 5b as a diastereomeric pair of
enantiomers. The PhSiH3 (1.6 equiv)/RhCl(PPh3)3 (3%) system conveniently transforms Cp-
(L)(CO)FeC(O)R to their alkyl derivatives Cp(L)(CO)FeCH2R. Seven iron acyl complexes
i
t
were studied (Cp(CO)2FeC(O)R, R ) Me, Ph, Pr, Bu; Cp(L)(CO)FeC(O)CH3, L ) PPh3,
P(OMe)3, and P(OPh)3), which gave 49-87% isolated yields of Cp(L)(CO)FeCH2R. Their
modest yields correspond to the presence of byproducts, Cp(CO)2FeCHdCH2 (32%) and 4b
(14%) with 1a and 10-15% of the vinyl complexes with L ) PPh3 and P(OPh)3.
In tr od u ction
C(O)CH3 (2a ) or (CO)5MnC(O)Ph (2b) as the precatalyst
with R2SiH2 (R ) Et, Ph) gave separable mixtures of
mono- and bis(Cp(CO)2Fe-R-siloxyethyl) complexes, 3a
and 4a , respectively, whereas use of (PPh3)(CO)4MnC-
(O)CH3 (2c) selectively furnished 3a . In the presence
of the more reactive PhSiH3, 2a catalytically trans-
formed 1a to mixtures of [Cp(CO)2FeCH(CH3)O]3SiPh
(5b) and Cp(CO)2FeCH2CH3 (6a ).1b
Rhodium(I) compounds also catalyze the hydrosilation
of CpFe acetyl complexes 1a and 1b, although the extent
of reaction and types of products often differ.2,3 For
example, RhCl(PPh3)3 does not catalyze the hydrosila-
tion of 1a with monohydrosilanes.2a With excess of the
more reactive dihydrosilanes,6 RhCl(PPh3)3 converted
1a to 3a in the presence of competing Rh(I)-promoted
Organotransition-metal acyl complexes exhibit a rich
and varied reaction chemistry with hydrosilanes that
depends on the metallic moiety and the choice of
hydrosilane and catalyst.1 The cyclopentadienyliron
acetyl complexes Cp(L)(CO)FeC(O)CH3 (L ) CO (1a ),
PPh3 (1b), P(OR)3), for example, undergo catalytic
hydrosilation with mono- or dihydrosilanes to give1b,2,3
isolable R-siloxyethyl complexes, Cp(CO)2FeCH(OSiR3)-
CH3 (Scheme 1).1b,e To date, the most active iron acetyl
hydrosilation catalysts are the degradation product(s)
that result from treating the manganese acyl complexes
L(CO)4MnC(O)R (2) with excess hydrosilane.4
An interesting selectivity was noted for the hydrosi-
lation of 1a with 1 equiv of a di- or trihydrosilane and
2-3% of a manganese acyl 2a -c. Using (CO)5Mn-
(4) (a) Gregg, B. T.; Cutler, A. R. Organometallics 1993, 12, 2006.
(b) Manganese alkyl and halide complexes (L)(CO)4MnX (X ) CH3,
Ph, CH(OSiR3)CH3, Br; L ) CO, PPh3, PEt3) also serve as effective
precatalysts for the hydrosilation of a variety of organic5 and organo-
metallic acyl compounds. Under comparable reaction conditions,
neither (CO)5MnSiR3 (R3Si ) Me2Ph, SiHPh2) nor Mn2(CO)10 are
effective precatalysts.
(5) (a) Gregg, B. T.; Cutler, A. R. Organometallics 1994, 13, 1039.
(b) Mao, Z.; Gregg, B. T.; Cutler, A. R. J . Am. Chem. Soc. 1995, 117,
10139. (c) DiBiase-Cavanaugh, M.; Gregg, B. T.; Cutler, A. R. Orga-
nometallics 1996, 15, 2764.
(6) The reactivity order of the hydrosilanes, RSiH3 . R2SiH2 . R3-
SiH, has been established for the Rh(PPh3)3Cl-catalyzed hydrosilation
of ketones. Reviews on catalytic hydrosilation of organic substrates:
(a) Ojima, I. In The Chemistry of Organic Silicon Compounds; Patai,
S., Rappoport, Z., Eds.; Wiley: New York, 1989; Chapter 25. (b)
Chaloner, P. A. Handbook of Coordination Catalysis in Organic
Chemistry; Butterworth: Boston, 1986; Chapter 7.2. (c) Marciniec, B.;
Gulin˜ski, J . J . Organomet. Chem. 1993, 446, 15.
(1) (a) Gregg, B. T.; Hanna, P. K.; Crawford, E. J .; Cutler, A. R. J .
Am. Chem. Soc. 1991, 113, 384. (b) Hanna, P. K.; Gregg, B. T.; Cutler,
A. R. Organometallics 1991, 10, 31. (c) Hanna, P. K.; Gregg, B. T.;
Tarazano, D. L.; Pinkes, J . R.; Cutler, A. R. In Homogeneous Transition
Metal Catalyzed Reactions; Advances in Chemistry Series No. 230;
Moser, W. R., Slocum, D. W., Eds.; American Chemical Society:
Washington, DC, 1992; p 491. (d) Gregg, B. T.; Cutler, A. R. Organo-
metallics 1992, 11, 4276. (e) Gregg, B. T.; Cutler, A. R. J . Am. Chem.
Soc. 1996, 118, 10069. (f) DiBiase-Cavanaugh, M.; Gregg, B. T.; Chiulli,
R. J .; Cutler, A. R. J . Organomet. Chem. 1997, 547, 173. (g) Gregg, B.
T.; Cutler, A. R. Manuscript in preparation.
(2) (a) Crawford, E. J .; Hanna, P. K.; Cutler, A. R. J . Am. Chem.
Soc. 1989, 111, 6891. (b) Pinkes, J . R.; Mao, Z.; Cutler, A. R. Manuscript
in preparation.
(3) (a) Akita, M.; Mitani, O.; Moro-Oka, Y. J . Chem. Soc., Chem.
Commun. 1989, 527. (b) Akita, M.; Mitani, O.; Sayama, M.; Moro-Oka,
Y. Organometallics 1991, 10, 1394.
S0276-7333(97)00659-6 CCC: $15.00 © 1998 American Chemical Society
Publication on Web 04/17/1998

1994 Organometallics, Vol. 17, No. 10, 1998
Mao et al.
Sch em e 1
silane dehydrogenative coupling.7 Although treatment
of 1b with monohydrosilanes and 2a -c yielded Cp-
(PPh3)(CO)FeCH(OSiR3)CH3, RhCl(PPh3)3 additionally
transformed these R-siloxyethyl products to the vinyl
complex Cp(PPh3)(CO)FeCHdCH2 (7b). With either
type of catalyst, dihydrosilanes additionally produce
varying amounts of the fully reduced Cp(PPh3)-
(CO)FeCH2CH3 (6b).2,3b It is this catalytic hydrosila-
tion-then-deoxygenation of the acyl ligand on LxMC-
(dO)R to an R-methylene group LxMCH2R that is of
interest to the present study.
surveyed included the manganese acyls 2a -c and RhCl-
(PPh3)3. These manganese acyls and PhSiH3 are effec-
tive for the hydrosilation-then-deoxygenation of organic
esters RC(dO)OR′ to give, successively, silyl acetals
[RCH(OR′)O]3-xHxSiPh and ethers RCH2OR′.5b In the
present study, RhCl(PPh3)3 proved to be the optimal
precatalyst for the deoxygenation of 1a and the scope
of this reaction was extended to Cp(CO)(L)FeC(O)CH3
(L ) CO, PPh3, P(OMe)3, P(OPh)3) and Cp(CO)2FeC-
(O)R (R ) Et, n-Pr, i-Pr, t-Bu). Portions of this study
have been communicated previously.1b,2a
The only other procedure that we are aware of for
directly reducing LxMC(dO)R to its alkyl derivative Lx-
MCH2R entails treatment with excess B2H6/C6D6 or
BH3‚THF.8 When carried out with 1b, however, this
procedure also gave 25% Cp(PPh3)(CO)FeH (8b) relative
to the major product 6b. Direct catalytic hydrogenation
(e.g., with RhCl(PPh3)3)8 of these acyl complexes has not
been documented.
In initiating this study, we wanted to find catalytic
conditions for the PhSiH3 hydrosilation-then-deoxygen-
ation of 1a to Cp(CO)2FeCH2CH3 (6a ) (eq 1) while
minimizing the accumulation of R-siloxyethylCp(CO)2-
Fe intermediates [Cp(CO)2FeCH(CH3)O]3-xHxSiPh (x )
2 (3b ), 1 (4b ), 0 (5b )). The precatalysts that we
Exp er im en ta l Section
Synthetic manipulations were performed using a combina-
tion of standard Schlenk-line, glovebox, and vacuum-line
procedures.9 Infrared spectra of benzene solutions were
recorded on a Perkin-Elmer FT spectrophotometer, model
1600, over the carbonyl ν(CO) frequency range (2200-1600
cm-1). NMR spectral data were obtained in C6D6 and were
reported as δ values relative to residual C6H6 (1H, 7.15 ppm)
and C6D6 (13C, 128.00 ppm) using Varian model XL-200 and
Unity 500 spectrometers. 1H NMR spectrometer parameters
had been optimized with a 38° pulse angle during quantitative
analyses of known concentrations of 1 and anisole. As a result,
integration traces are believed to be accurate within (5%.
Combustion microanalyses were done by Quantitative Tech-
nologies, Bound Brook, NJ .
Organic and inorganic reagents were obtained commercially
and used as received; silanes and C6D6 were stored in a
glovebox under nitrogen. THF and benzene were distilled from
sodium benzophenone ketyl. Wilkinson’s compound, Rh-
(PPh3)3Cl, was purchased or prepared and purified according
to the literature procedure;10 its purity (activity) was assayed
by 1H NMR spectral monitoring of the catalyzed H2SiPh2
hydrosilation of Cp(CO)2FeC(O)CH3.2a The manganese acyl
complexes, (CO)5MnC(O)CH3 (2a ),11 (CO)5MnC(O)Ph (2b),12
and (PPh3)(CO)4MnC(O)CH3 (2c),13 and the iron acyl com-
plexes, Cp(CO)(L)FeC(O)CH3 (L ) CO (1a ), PPh3 (1b),14
(9) Shriver, D. F.; Drezdon, M. A. The Manipulation of Air-Sensitive
Compounds, 2nd ed., Wiley-Interscience: New York, 1986.
(10) (a) Osborn, J . A.; Wilkinson, G. Inorg. Synth. 1990, 28, 77. (b)
Evans, D. A.; Fu, G. C.; Anderson, B. A. J . Am. Chem. Soc. 1992, 114,
6679. (c) Burgess, K.; van der Donk, W. A.; Weastcott, S. A.; Marder,
T. B.; Baker, R. T.; Calabrese, J . C. J . Am. Chem. Soc. 1992, 114, 9350.
(11) (a) Closson, R. D.; Kozikowski, J .; Coffield, T. H. J . Org. Chem.
1957, 22, 598. (b) King, R. B. Organometallic Syntheses; Academic
Press: New York, 1965; Vol. 1, Part I, p 147. (c) Mckinney, R. J .;
Crawford, S. S. Inorg. Synth. 1989, 26, 155. (d) Anderson, J . M.; Moss,
J . R. Organometallics 1994, 13, 5013.
(7) (a) Ojima, I.; Inaba, S. I.; Kogure, T.; Nagai, Y. J . Organomet.
Chem. 1973, 55, C7. (b) Chang, L. S.; Corey, J . Y. Organometallics
1989, 8, 1885.
(8) Van Doorn, J . A.; Masters, C.; Volger, H. C. J . Organomet. Chem.
1983, 243, 87.

Deoxygenation of Iron Acyl Complexes
Organometallics, Vol. 17, No. 10, 1998 1995
P(OMe)3 (1c), P(OPh)3 (1d )15) and Cp(CO)2FeC(O)R (R ) Ph16
(9), i-Pr17 (10), t-Bu18 (11)), were prepared by literature
procedures and judged pure by IR and H NMR spectroscopy.
was extracted by hexane (2 × 1 mL). This extractant was
chromatographed on a column (10 × 2 cm) of flash-grade silica
gel in hexane. Elution with hexane and then 2:1 hexane-
benzene brought down light orange bands that produced small
amounts of Cp(CO)2FeCH2CH3 admixed with unidentified
organic residues. Continued development of the column with
2:1 hexane-benzene removed fraction 3, a dark yellow band,
which provided 133 mg of a yellow gum. This gum cor-
responded to [Cp(CO)2FeCH(CH3)O]3SiPh (5b) (43% yield by
relative integration) that was contaminated (total 30 mg) with
6a (6%), <4% [Cp(CO)2FeCH(CH3)O]2SiHPh (4b), and uini-
dentified organic materials. Prolonged washing of the column
with 2:1 hexane-benzene and then benzene gave smaller
amounts of 5b that were increasingly contaminated with 4b,
Cp(CO)2Fe2, and other unidentified residues.
Hyd r osila tion of 1a w ith (P P h 3)(CO)4Mn C(O)CH3 (2c)
a s th e P r eca ta lyst. 1. To a 5-mL vial in the glovebox was
added 1a (44 mg, 0.20 mmol), C6D6 (600 mg), 2c (3 mg, 3.2%),
PhSiH3 (11 or 26 mg, 0.50 and 1.20 equiv respectively), and
toluene (10 mg, 0.11 mmol). The yellow solution darkened over
5 min. It was transferred to a 5-mm NMR tube that was fitted
with a rubber septum, and 1H NMR spectra were recorded
within 10 min. Within 0.5 h, all of the substrate 1a was
consumed.
2. Phenylsilane (43 mg, 0.40 mmol) was added to a benzene
solution (1 g) of 1a (176 mg, 0.80 mmol) and 2c (11 mg, 0.023
mmol, 2.9%). After 0.5 h, the darkened yellow solution was
chromatographed on Biorad SX-12 (size exclusion) polystyrene
beads in benzene. Four 15-mL yellow fractions were collected
and evaporated. Fraction 2 offered 180 mg of a brown gum
that was spectroscopically characterized as 4b (82% yield) plus
small amounts of organic residues. Further attempts to purify
this fraction proved unsuccessful. Fractions 1 and 3 also
contained 4b admixed with unidentified organic residues.
Tr ea tm en t of [Cp (CO)2F eCH(CH3)O]2SiHP h (4b) w ith
Rh Cl(P P h 3)3 a n d P h SiH3. A yellow solution containing 1a
(44 mg, 0.20 mmol), 2c (3 mg, 3.2%), and toluene (10.3 mg,
0.11 mmol) in 500 mg of C6D6 was treated with PhSiH3 (11
mg, 0.10 mmol). 1H and 13C NMR spectral monitoring of the
reaction mixture indicated that all of 1a and PhSiH3 were
consumed within 0.5 h, leaving a mixture of 4b (81%) and
Cp(CO)2FeCH2CH3 (6a ) (4%). The NMR tube was brought
back to the glovebox, and RhCl(PPh3)3 (3 mg) and excess
PhSiH3 (35 mg) were added. The results of 1H NMR spectral
monitoring over 3 h were consistent with a reduction in
concentration of 4b (70%) commensurate with an increase in
6a (14%).
Rea ction of [Cp (CO)2F eCH(CH3)O]3SiP h (5b) w ith
Rh Cl(P P h 3)3 a n d P h SiH3. To a 0.5 mL benzene solution of
1a (200 mg, 0.91 mmol) was added 1b (12 mg, 4.4%) and
PhSiH3 (110 mg, 1.01 mmol). The yellow solution turned red
over 1 h, and after 1.5 h it was chromatographed on Biorad
SX-12 polystyrene beads in benzene. Five 15-mL yellow
fractions were collected and evaporated, of which fraction 2
contained most of the initial yellow band. This band provided
172 mg of a yellow oil that was identified as [Cp(CO)2FeCH-
(CH3)O]3SiPh (5b) (0.22 mmol, 74% yield), albeit contaminated
by small amounts of organic residues. This material in 600
mg of C6D6 was treated with RhCl(PPh3)3 (21 mg, 0.022 mmol)
and PhSiH3 (130 mg, 1.19 mmol). 1H NMR spectral monitor-
ing over 8 h indicated that none of the substrate 5b was
consumed.
1
Samples of Cp(CO)2FeCHdCH2 (7a )16b and its spectral data
were available from previous studies:19 1H NMR (C6D6) δ 7.17
(dd, J ) 17.1, 9.0 Hz, Fe-CH), 6.26 (dd, J ) 9.0, 0.9 Hz, E
dCH), 5.76 (dd, J ) 17.1, 0.9 Hz, Z dCH), 4.02 (Cp); 13C NMR
(C6D6) δ 216.53 (CO), 85.41 (Cp), 139.76 (dCH2), 130.81 (Fe-
CHd).
Hyd r osila tion of Cp (CO)2F eC(O)CH3 (1a ) w ith P h SiH3
a n d (CO)5Mn C(O)CH3 (2a ) a s th e P r eca ta lyst. A solution
containing 200 mg of 1a (0.91 mmol), 10 mg of manganese
acetyl 2a 0.042 mmol, 4.6 mol %), and 33 mg of C6H5OCH3
(0.31 mmol) in 600 mg of C6D6 was treated with PhSiH3 (118
mg, 1.09 mmol). The resulting light yellow-orange solution
was transferred in the glovebox to a 5-mm NMR tube that was
fitted with a new rubber septum, and the first 1H NMR
spectrum was recorded within 10 min. Within 1 h, the reaction
solution had turned dark red; no other physical changes were
noted over 12 h. Since virtually all NMR absorptions were
1
assigned, the reaction progress was monitored by H and 13C-
{1H} NMR spectroscopy.
The reaction was worked up by size-exclusion chromatog-
raphy on polystyrene beads (Biorad SX-12, molecular weight
range 0-400) that had been packed into a 40 × 350 mm glass
column in benzene. A guard column (21 × 130 mm) also was
part of the high-performance, low- to medium-pressure chro-
matography20 apparatus, which was connected via 4-mm
Teflon tubing and fittings. The UV-vis detector was set at
463 nm. After 12 h, the reaction mixture was injected onto
the apparatus through a three-way valve and was pumped
through the columns. Six orange to yellow-orange fractions,
10-15 mL each, were collected and evaporated. When com-
bined after NMR analysis, fractions 2-4 afforded 57 mg of an
orange oil that was characterized as [Cp(CO)2FeCH(CH3)O]3-
SiPh (5b) (24% yield). Anal. Calcd for C33H32O9SiFe3: C,
51.59; H, 4.20. Found: C, 52.10; H, 4.11. Fractions 5 and 6
were evaporated under a stream of nitrogen to yield 104 mg
of spectroscopically pure Cp(CO)2FeCH2CH3 (6a ),21 an amber
oil, in 51% yield. 1H NMR (C6D6) δ 4.02 (Cp), 1.58 (q, J ) 7.6
Hz, FeCH2), 1.36 (t, CH3); 13C NMR δ 218.25 (CO), 85.32 (Cp),
22.53 (CH3), -2.81 (FeCH2).
Hyd r osila tion of Cp (CO)2F eC(O)CH3 (1a ) w ith (CO)5-
Mn C(O)P h (2b) a s th e P r eca ta lyst. A yellow C6D6 solution
(0.5 mL) containing 200 mg of 1a (0.91 mmol) and 12 mg of
2b (4.4 mol %) was treated with 118 mg of PhSiH3 (1.09 mmol).
The reaction solution turned yellow-orange after 15 min, and
after 2 h, the red solution was evaporated to a red oil, which
(12) (a) Lukehart, C. M.; Torrence, G. P.; Zeile, J . V. Inorg. Synth.
1978, 18, 56. (b) Motz, P. L.; Sheeran, D. J .; Orchin, M. J . Organomet.
Chem. 1990, 383, 201.
(13) (a) Mawby, R. J .; Basolo, F.; Pearson, R. G. J . Am. Chem. Soc.
1964, 86, 5043. (b) Kraihanzel, C. S.; Maples, P. K. Inorg. Chem. 1968,
7, 1806. (c) Noack, K.; Ruch, M.; Calderazzo, F. Inorg. Chem. 1968, 7,
345.
(14) (a) Reger, D. L.; Fauth, D. J .; Dukes, M. D. Synth. React. Inorg.
Metal Org. Chem. 1977, 7, 151. (b) Forschner, T. C.; Cutler, A. R. Inorg.
Synth. 1989, 26, 231.
(15) (a) Crawford, E J .; Lambert, C.; Menard, K. P.; Cutler, A. R. J .
Am. Chem. Soc. 1985, 107, 3130. (b) Bibler, J . B.; Wojcicki, A. Inorg.
Chem. 1966, 5, 889.
(16) (a) Brookhart, M.; Studabaker, W. B.; Humphrey, M. B.; Husk,
G. R. Organometallics 1989, 8, 132. (b) King, R. B.; Bisnette, M. B. J .
Organomet. Chem. 1964, 2, 15.
(17) Lukehart, C. M.; Zeile, J . V. J . Am. Chem. Soc. 1977, 99, 4368.
(18) Afzal, D.; Lenhert, P. G.; Lukehart, C. M.; Srinivasan, R. Inorg.
Chem. 1986, 25, 710.
(19) (a) Tarazano, D. L.; Cutler, A. R. J . Organomet. Chem. 1993,
448, 139. (b) Nesmeyanov, A. N.; Rybinskaya, M. I.; Rybin, L. v.;
Kaganovich, V. S.; Petrovskii, P. V. J . Organomet. Chem. 1971, 31,
257.
(20) Components for low- to medium-pressure chromatography
apparatus are available from Ace Glass Inc., Vineland, NJ .
(21) (a) Piper, T. S.; Wilkinson, G. J . Inorg. Nucl. Chem. 1956, 3,
104. (b) Green, M. L. H.; Nagy, P. J . Organomet. Chem. 1963, 1, 58.
Tr eatm en t of Cp(CO)2FeCH(OSiMe2P h )CH3 with Rh Cl-
(P P h 3)3 a n d P h SiH3. PhMe2SiH (138 mg, 1.01 mmol) was
added to a benzene solution (0.50 mL) containing 1a (200 mg,
0.91 mmol) and (CO)5MnC(O)CH3 (2a ) (12 mg, 4.4%). After
50 min, the resulting orange solution was chromatographed
on 4 × 1 cm of silica gel-pentane. The resulting yellow
pentane eluate (40 mL) was evaporated to yield a 188 mg of a
yellow oil, which was identified as spectroscopically pure Cp-
(CO)2FeCH(OSiMe2Ph)CH3 (0.53 mmol): 1H NMR (C6D6) δ
1b

1996 Organometallics, Vol. 17, No. 10, 1998
Mao et al.
5.72 (q, J ) 6.0 Hz, FeCH), 4.15 (Cp), 1.69 (d, CHCH3), 0.39
(s, SiMe2). To this material in C6D6 (500 mg) was added RhCl-
(PPh3)3 (15 mg, 0.016 mmol or 3.1%) and PhSiH3 (57 mg, 0.53
mmol). Although gas evolution occurred during the first 0.5
h, the solution remained orange and 1H NMR spectral moni-
toring over 8 h indicated that none of the Cp(CO)2FeCH-
(OSiMe2Ph)CH3 was consumed.
reaction was complete. The reaction mixture was poured into
10 mL of pentane and centrifuged. The centrifugate plus
pentane washings (2 × 5 mL) were chromatographed on
activity 2 neutral alumina-pentane (2 × 5 cm column).
Development with 5:1 pentane-CH2Cl2 eluted an orange band
that gave spectroscopically pure 6b (137 mg, 62% yield) as an
orange solid. 1H NMR (C6D6) δ 7.52 (m, 6H, Ph), 7.02 (m, 9H,
Ph), 4.12 (d, J ) 1.0 Hz, Cp), 1.85 (m, 1H, FeCHH), 1.56 (dt,
J ) 1.9, 7.3 Hz, FeCH2CH3), 1.04 (m, 1H, FeCHH); 13C NMR
(C6D6) δ 223.52 (d, J ) 32.3 Hz, CO), 137.71 (d, J ) 39.1 Hz,
ipso C), 133.40 (d, J ) 9.6 Hz, o-C), 129.45 (s, p-C), 128.21 (d,
J ) 6.9 Hz, m-C), 84.95 (Cp), 23.71 (d, J ) 4.1 Hz, FeCH2CH3),
-2.19 (d, J ) 18.5 Hz, FeCH2).
Similar reactions, 0.20 and 0.50 mmol scale, were performed
with the following iron acyls: Cp(CO)2FeC(O)Ph (9), Cp-
(CO)2FeC(O)CH(CH3)2 (10), Cp(CO)2FeC(O)C(CH3)3 (11), Cp-
[P(OMe)3](CO)FeC(O)CH3 (1c), and Cp[P(OPh)3](CO)FeC(O)-
CH3 (1d ). Reactions were chromatographed using the above
procedures; both NMR spectroscopic and isolated yields of the
resulting alkyl complexes are recorded in Table 2. Salient
NMR spectroscopic data, Cp(CO)2FeCH2Ph:26 1H NMR (C6D6)
δ 3.91 (Cp), 2.67 (CH2); 13C NMR (C6D6) δ 217.5 (CO), 86.01
(Cp), 5.42 (CH2). Cp(CO)2FeCH2CH(CH3)2:27 1H NMR (C6D6)
δ 4.05 (Cp), 4.74 (m, FeCH2CH), 1.59 (t, J ) 1.2 Hz, FeCH2),
1.09 (d, J ) 5.9 Hz, CH3); 13C NMR (C6D6) δ 218.39 (CO), 85.45
(Cp), 35.24 (Cp(CO)2FeCH2CH), 26.66 (Cp(CO)2FeCH2), 13.71
(CH3). Cp(CO)2FeCH2C(CH3)3:28 1H NMR (C6D6) δ 4.06 (Cp),
1.73 (s, FeCH2), 1.12 (s, CH3); 13C NMR (C6D6) δ 219.11 (CO),
85.62 (Cp), 35.40 (CMe3), 33.40 (FeCH2), 19.25 (CH3). Cp-
[P(OMe)3](CO)FeCH2CH3:29 1H NMR (C6D6) δ 4.31 (Cp), 3.26
(d, J ) 11.3 Hz, P(OMe)3), 1.58 (m, FeCH2CH3), 1.27 (m,
FeCH2CH3); 13C NMR (C6D6) δ 221.4 (d, J ) 47.6 Hz, CO),
83.32 (Cp), 51.13 (P(OMe)3), 23.27 (d, J ) 2.7 Hz, FeCH2CH3),
-6.01 (d, J ) 29.3 Hz, FeCH2). Cp[P(OPh)3](CO)FeCH2CH3:
24a,30 1H NMR (C6D6) δ 7.25 (m, 6H, Ph), 7.02 (m, 6H, Ph), 6.85
(m, 3H, Ph), 3.90 (Cp), 1.62 (m, 5H, CH2CH3).
Hyd r osila tion of Cp (CO)2F eC(O)CH3 (1a ) w ith P h SiH3
a n d Rh Cl(P P h 3)3. The reaction was carried out in a 1 dram
vial in the glovebox by adding 6 mg of RhCl(PPh3)3 (3.2%) to
a C6D6 solution (500 mg) containing 1a (44 mg, 0.20 mmol),
PhSiH3 (35 mg, 1.6 equiv), and PhMe internal standard (9 mg,
0.10 mmol). After 10 min, the yellow solution turned yellow-
ish-green (commensurate with gas evolution), and it was
transferred to a NMR tube. Within 40-60 min all of the 1a
was consumed, as judged by 1H and 13C NMR spectral
monitoring. The presence of Cp 13C NMR absorptions was
especially informative: Cp(CO)2FeCH2CH3 (6a ) (δ 85.32),
Cp(CO)2FeCHdCH2 (7a ) (δ 85.41), 1a (δ 86.22), Cp(CO)2Fe2
(δ 88.52). Product yields were determined via integration of
1
the H NMR spectra vs PhMe: 6a δ 1.58 (quart, J ) 7.6 Hz,
CH3), 1.36 (t, CH2), (49%); 7a δ 6.26 (E dCH), 5.76 (Z dCH)
(32%); [Cp(CO)2FeCH(CH3)O]2SiHPh (4b) δ 6.12-6.06 (Fe-
CH) (12-14%); 1a δ 2.40 (CH3); Cp(CO)2FeH22 (8a ) δ -11.8
(FeH).
When the reaction was repeated with 1.1 equiv of PhSiH3,
23% of 1a remained along with 6a (49%), 7a (32%), and 4b
(10-12%). With 0.56 equiv of H3SiPh, the reaction had 29%
1a remaining after 3 h and only trace quantities of 7a .
The reaction mixture (1.6 equiv of PhSiH3) and a 1 mL of
pentane rinse were poured into 5 mL of pentane and centri-
fuged. The supernatant solution was chromatographed on
activity 2 neutral alumina (5 × 90 mm) in pentane, from which
a yellow band was removed with 5:1 pentane-CH2Cl2. This
afforded 22 mg of a yellow oil that was identified as spectro-
scopically pure 6a (53% yield); NMR spectral yield vs added
(Me3Si)2O, 52%.
The reaction was carried out on a 0.50 mmol scale: 110 mg
of 1a (0.50 mmol), 86 mg of PhSiH3 (1.6 equiv), and 14 mg of
RhCl(PPh3)3 (3.2%) were mixed in 1.0 g of C6D6. An NMR
spectrum of the reaction mixture after 2 h indicated that all
of 1a was consumed; chromatographic yield of 6a (4 × 1 cm
alumina-pentane, eluted with pentane then 5:1 pentane-CH2-
Cl2), 52 mg, 49%.
Resu lts a n d Discu ssion
1. Ma n ga n ese-Acyl-Ca ta lyzed P h SiH3 Hyd r osi-
la tion of Cp (CO)2F eC(O)CH3 (1a ). The (CO)5MnC-
(O)CH3-catalyzed PhSiH3 hydrosilation of Cp(CO)2FeC-
(O)CH3 (1a ) exhibited three characteristics. First, these
sluggish reactions had variable reaction times under
comparable conditions. With 1.1-1.2 equiv of PhSiH3
and 4.4-4.6% of precatalyst 2a , typically 20% of 1a was
depleted over 4-5 h and then 70-85% over 7-8 h.
These values, averaged over a dozen runs, however, do
not convey the wide rate fluctuations with which
substrate 1a was consumed. The rate of consumption
of 1a varied from 85% within 2 h to an initial 30% after
6 h followed by jumping to 85% after 7 h.
Rh Cl(P P h 3)3-Ca ta lyzed P h SiH3 Hyd r osila tion of Cp -
(L)(CO)FeC(O)R: Cp(P P h 3)(CO)FeC(O)CH3 (1b). An NMR
tube was charged with a C6D6 solution (600 mg) of 1b (91 mg,
0.20 mmol), PhSiH3 (35 mg, 1.6 equiv), PhMe (14 mg, 0.15
mmol), and RhCl(PPh3)3 (6 mg, 3.2%). The reaction was
1
complete within 45 min, as judged by H NMR spectroscopy:
61% Cp(PPh3)(CO)FeCH2CH3 (6b),8,23 11% Cp(PPh3)(CO)-
FeCHdCH2 (7b)24 (quantitated by integration of the vinyl
absorptions at δ 6.31 (ddd, 1H), 5.86 (dt, 1H), 4.18 (d, J ) 1.2
Hz, Cp)), and 11% Cp(PPh3)(CO)FeH (8b )23a,25 (quantitated
The second observation concerns the fate of the
precatalyst 2a . Most of it remained intact (as deduced
from 1H NMR spectroscopy), up to at least 85-90%
consumption of the substrate 1a . This substrate inhibi-
tion of the required hydrosilation of the precatalyst1d is
typical of the manganese-acyl-catalyzed hydrosilation
of 1.2 As expected,4 a reactive precatalyst is required
for the hydrosilation of 1a ; neither (CO)5MnSi(CH3)3
nor Mn2(CO)10 function as viable hydrosilation catalysts
toward 1 under these conditions.
with the hydride and Cp absorptions at δ -12.85 (d, J PH
75.3 Hz) and 4.25 (d, J ) 1.0 Hz), respectively).
Cp(PPh3)(CO)FeC(O)CH3 (1b) (227 mg, 0.50 mmol), PhSiH3
(86 mg, 1.6 equiv), and RhCl(PPh3)3 (14 mg, 3.2%) were
dissolved in 1.2 g of C6D6. The orange solution turned green
within 15 min; NMR spectral monitoring confirmed that the
)
(22) (a) Ferguson, S. B.; Sanderson, L. J .; Skackleton, T. A.; Baird,
M. C. Inorg. Chim. Acta 1984, 83, L45. (b) Shackleton, T. A.; Mackie,
S. C.; Fergusson, S. B.; J ohnston, L. J .; Baird, M. C. Organometallics
1990, 9, 2248.
(23) (a) Su, S.; Wojcicki, A. J . Organomet. Chem. 1971, 27, 271. (b)
Reger, D. L.; Culbertson, E. C. J . Am. Chem. Soc. 1976, 98, 2789. (c)
Flood, T. C.; DiSanti, F. J .; Miles, D. L. Inorg. Chem. 1976, 15, 1910.
(24) (a) Bodnar, T.; Cutler, A. R. J . Organomet. Chem. 1981, 213,
C31. (b) Boland-Lusser, B. E.; Churchill, M. R.; Hughes, R. P.;
Rheingold, A. L. Organometallics 1982, 1, 628.
(26) Bibler, J . B.; Wojcicki, A. J . Am. Chem. Soc. 1966, 88, 4862.
(27) Green, M. L. H.; Smith, M. J . J . Chem. Soc. A 1971, 1943.
(28) J acobson, S. E.; Wojcicki, A. J . Am. Chem. Soc. 1973, 95, 6962.
(29) Alt, H. G.; Eichner, M. E.; J ansen, B. M.; Thewalt, U. Z.
Naturforsch., B: Anorg. Chem., Org. Chem. 1982, 37B, 1109.
(30) Reger, D. L.; Coleman, C. J . Inorg. Chem. 1979, 18, 3155.
(25) (a) Kalck, P.; Poilblanc, R. C. R. Acad. Sci. Paris 1972, 274, 66.
(b) Reger, D. L.; Culbertson, E. C. J . Am. Chem. Soc. 1976, 98, 2789.

Deoxygenation of Iron Acyl Complexes
Organometallics, Vol. 17, No. 10, 1998 1997
13C{1H} NMR (C6D6), δ
Ta ble 1. Sp ectr a l Da ta for r-Siloxyeth yl F e(CO)2Cp Com p lexes
complex (IR)
1H NMR (C6D6), δ
Cp(CO)2FeCH(CH3)OSiH2Ph, 3ba
5.73 (q, J ) 6.0 Hz, FeCH)
5.25 (s, 2H, SiH)
72.26 (CH)
34.88 (CH3)
4.24 (s, Cp)
1.78 (d, J ) 6.0 Hz, CH3)
[Cp(CO)2FeCH(CH3)O]2SiHPh, 4bb
(ν(CO) 2004, 1945 cm-1(br))
7.90 (m, 2H, SiPh)
217.85, 217.55, 217.49, 217.46 (CO)
134.30, 130.56, 127.49 (ipso-SiPh)
86.29, 86.27, 86.24 (br) (Cp)
70.20, 70.02, 69.41, 69.40 (CH)
35.73, 35.69, 35.49, 35.42 (CH3)
7.26 (m, 3H, SiPh)
6.07 (q, J ) 6.4 Hz, FeCHA)
6.05 (q, J ) 6.4 Hz, FeCHB)
6.04 (q, J ) 5.9 Hz, FeCHC)
6.01 (q, J ) 5.9 Hz, FeCHD)
5.47 (s, 1H, SiH)
5.42 (s, 3H, SiH)
4.24, 4.22, 4.18, 4.17 (s, CpA-D
)
1.96 (d, J ) 6.4 Hz, CH3(C)
1.95 (d, J ) 6.4 Hz, CH3(D)
1.88 (d, J ) 5.9 Hz, CH3(A)
1.87 (d, J ) 5.9 Hz, CH3(B)
8.10 (m, 2H, Ph)
)
)
)
)
[Cp(CO)2FeCH(CH3)O]3SiPh,
218.09, 218.08, 218.05, 218.03,
217.91, 217.87, 217.82, 217.77 (CO)
138.22 (ipso-C, Ph)
5b (ν(CO) 2008, 1951 cm-1(br))
7.36 (m, 2H, Ph)
7.26 (m, 1H, Ph)
6.39 (q, J ) 6.2 Hz, FeCHA)
6.38 (q, J ) 6.2 Hz, FeCHB)
6.32 (q, J ) 5.8 Hz, FeCHC)
6.23 (q, J ) 6.1 Hz, FeCHD)
4.39, 4.38, 4.31, 4.27 (s, Cp)
130.06, 128.53, 127.99 (Ph)
86.36, 86.30, 86.27, 86.21 (Cp)
69.37, 69.29, 69.27, 69.10 (FeC)
36.12, 36.02, 35.97, 35.93 (CH3)
2.06 (d, J ) 5.8 Hz, CH3(A)
2.04 (d, J ) 5.9 Hz, CH3(C)
2.01 (d, J ) 5.9 Hz, CH3(B)
)
)
)
1.98 (d, J ) 6.1 Hz, CH3(D)
)
a
Observed as a minor component during the early stages of hydrosilation of 1a . Carbonyl (IR, 13C NMR), phenyl (1H, 13C NMR), and
Cp (13C NMR) absorptions were not discerned. Methine absorption appears as an eight-line multiplet at δ 6.05, and two of the methyl
doublets appear as an apparent triplet (δ 1.99). Double irradiation at δ 1.9 reduced the δ 6.05 multiplet to four equally intense singlets;
reversing the decoupling experiment revealed four methyl resonances. Further correlation of these methine-methyl spin systems are
consistent with the results of the COSY experiments.33
b
Ta ble 2. Rh Cl(P P h 3)3-Ca ta lyzed P h SiH3 Hyd r osila tion of Cp (CO)(L)F eC(O)R
substrate
reaction scale (mmol)a
product
NMR yield (%)b
49d
isolated yield (%)c
1. Cp(CO)2FeC(O)CH3 (1a )
0.20
0.50
0.50
0.20
0.50
0.20
0.50
0.20
0.50
0.20
0.50
0.50
Cp(CO)2FeCH2CH3 (6a )
53
49
69
2. Cp(CO)2FeC(O)CH(CH3)2 (10)
3. Cp(CO)2FeC(O)C(CH3)3 (11)
Cp(CO)2FeCH2CH(CH3)2
Cp(CO)2FeCH2C(CH3)3
91
66
58
73
4. Cp(CO)2FeC(O)Ph (9)
Cp(CO)2FeCH2Ph
82
5. Cp(PPh3)(CO)FeC(O)CH3 (1b)
6. Cp[P(OMe)3](CO)FeC(O)CH3 (1c)
Cp(PPh3)(CO)FeCH2CH3 (6b)
Cp[P(OMe)3](CO)FeCH2CH3
Cp[P(OPh)3](CO)FeCH2CH3
61e
75e
62
74
87
7. Cp[P(OPh)3](CO)FeC(O)CH3 (1d )
a
Reaction carried out in 600 mg of C6D6 (0.20 mmol) or 1.5 mL of C6D6 or C6H6 (0.50 mmol) with 1.6 equiv of PhSiH3 and 3.2%
RhCl(PPh3)3. Quantified by 1H NMR vs PhMe internal standard ((5%). c Isolated as spectroscopically pure material via column
b
d
chromatography on activity
2
alumina using 5:1 pentane-CH2Cl2. Also detected 32% Cp(CO)2FeCHdCH2 (7a ) and 12-14%
[Cp(CO)2FeCH(CH3)O]2SiHPh (4b). e Also detected 10-15% of the vinyl complexes Cp(L)(CO)FeCHdCH2 (L ) PPh3 (7b) and P(OMe)3
(7c)).
1
Third, results of H NMR spectroscopic monitoring
established that the 2a -catalyzed hydrosilation of 1a
occurs in two stages, Scheme 2. Within the first 7-8
h, 1a transformed to variable mixtures of Cp(CO)2FeCH-
(CH3)OSiH2Ph (3b), [Cp(CO)2FeCH(CH3)O]2SiHPh (4b),
and [Cp(CO)2FeCH(CH3)O]3SiPh (5b), typically 1:1:2 at
70% depletion of 1a . Little or no Cp(CO)2FeCH2CH3
(6a ) or Cp2(CO)4Fe2 was detected. Between 8 and 12
h, the reaction mixture of 3b-5b plus remaining
PhSiH3 produced 6a (along with traces of Cp2(CO)4Fe2)
and left 5b. Then 5b evidently underwent only very
slow reduction to 6a , which was not further studied.
Although the use of 2.2-2.4 equiv of PhSiH3 and the
same precatalyst concentration accelerated the reaction
(1a was consumed in under 6 h), the same ratio of 6a
and 5b formed as the final products.
Together, 6a and 5b constituted the final products
after reaction for 12 h with 1.2 equiv of PhSiH3 and
accounted for virtually all of the starting 1a . Their
NMR yields of 35% and 62%, respectively (based upon
1a ), were calculated via two procedures. Results of
direct quantitation of either their Cp resonances or their
upfield ethyl multiplets and methyl doubletssusing
anisole as an internal standardsagreed with the other
procedure of extrapolating the maximum yields of 6a
and 5b that are possible from the relative integrated
areas of their Cp resonances.31 We made no attempts
to account for the residual organosilane byproducts.
These byproducts left no definitive NMR spectral ab-
1
sorptions other than broad SiH H NMR absorptions at
5.2-5.4 ppm, which coincided with the buildup of
substantial concentrations of 6a .

1998 Organometallics, Vol. 17, No. 10, 1998
Mao et al.
Sch em e 2
We isolated [Cp(CO)2FeCH(CH3)O]3SiPh (5b) in 24%
yield by size-exclusion chromatography, as outlined in
the Experimental Section. This chromatography pro-
cedure, when used on 0.5-2.0 mmol scale reactions, had
been used to separate similar hydrosilation reaction
products involving dihydrosilanes. Both Cp(CO)2FeCH-
(OSiHR2)CH3 (3a ) and [Cp(CO)2FeCH(CH3)O]2SiR2 (4a )
(R ) Et, Ph), the R2SiH2 dihydrosilation products of 1a
(Scheme 1), accordingly had been isolated by size-
exclusion chromatography.1b,e Experimental details also
are outlined for working up the 1a /PhSiH3 hydrosilation
reactions using the more traditional column chroma-
tography on 40 µm (flash-grade) silica gel-hexane.
Under these chromatography conditions, however, con-
siderable decomposition occurred on the column and
produced unidentified residues admixed with both 5b
and 6a .
duced only fractions that contained varying ratios
1
(always <1:5) of 4b and 5b. Although both the H and
13C NMR spectra of these reaction mixtures are rich in
the number of absorptions corresponding to diastereo-
meric mixtures for 3b-5b, the isolation of 5b (and of
4b from the 2c-catalyzed reactions, vide infra) greatly
simplified the assignment of their NMR spectra and
spectral monitoring of these hydrosilation reactions.
NMR spectroscopic monitoring of the catalyzed Ph-
SiH3 hydrosilation of 1a also was facilitated by spread-
ing out of the corresponding absorptions for 3b-5b. In
particular, a characteristic downfield progression of the
methine quartets, Cp singlets, and methyl doublets
1
occurs in their H NMR spectra in going from 5b to 4b
to 3b. These chemical shift regions for each of 3b-5b
are consistent with the results of COSY NMR experi-
ments.33
The isolated 5b is a stable yellow oil that that was
fully characterized as an equimolar mixture of four
diastereomers. Its IR spectrum shows two broadened
carbonyl υ(CO) bands at 2008 and 1951 cm-1. Chemical
Ethoxyphenylsilanes HxSi(EtO)3-xPh (x ) 1-3) that
might have originated from continued oxidative addition
of PhSiH3 to 3b-5b and then reductive elimination
were not detected. Similar elimination of alkoxysilanes
from CpFe R-siloxyethyl complexes has been observed
with trialkylsilanes upon photolysis or heating.34 1H
and 13C NMR spectral data of authentic samples of these
ethoxysilanes, generated by 2-catalyzed dehydrogena-
tive silation of ethanol with PhSiH3,5a were obtained
independently. These ethoxyphenylsilanes exhibit di-
agnostic 1H NMR absorptions at δ 3.6-3.7 and 1.0-1.2
and 13C NMR resonances at δ 18-19 and 59-61 that
easily would have been detected during the hydrosila-
tion of 1a .
1
shifts for the H and 13C NMR spectra of 5b (Table 1)
correspond to those of the fully characterized Cp-
(CO)2FeCH(OSiHPh2)CH3 (3a ) and [Cp(CO)2FeCH-
1
(CH3)O]2SiPh2 (4a ). Both the H and 13C NMR spectra
of 5b are well resolved and further establish it as a
constitutionally symmetrical molecule that has three
stereogenic Cp(CO)2FeCHMeO groups branching off an
achiral silicon center. The resulting four stereoisomers
constitute a diastereomeric pair of enantiomers [RRR
(SSS) and RRS (SSR)]32 that accounts for the net 1:1:
1:1 relative intensities for the Cp (and other Cp-
(CO)2FeCHMeO) 1H and 13C NMR spectral resonances
of 5b.33
We were unable to isolate Cp(CO)2FeCH(CH3)OSiH2-
Ph (3b) and [Cp(CO)2FeCH(CH3)O]2SiHPh (4b) during
the earlier stages of these reactions catalyzed by 2a .
Attempted size-exclusion chromatographic workup pro-
Substitution of (CO)5MnC(O)Ph (2b) as the precata-
lyst (4.4%) under otherwise identical reaction conditions
considerably accelerated the PhSiH3 hydrosilation of
Cp(CO)2FeC(O)CH3 (1a )sthe substrate was depleted
within 30 min. In its place, a mixture containing 4%
3b , 22% 4b , 72% 5b , and approximately 2% Cp-
1
(CO)2FeCH2CH3 (6a ) was quantified by H NMR spec-
troscopy. Over 2 h, 3b was consumed, leaving 15% 4b,
81% 5b, but only 4% 6a . Prolonged sitting of the
reaction mixture up to 12 h only increased the yield of
(31) The ratio of the integral traces of the Cp resonances for 5b is
0.61 times that of Cp(CO)2FeEt (6a ) or n5b ) (0.61)n6a (n ) number of
moles), after accounting for the relative number of Cp rings involved.
Thus, n1a ) n6a + 3n5b, and n1a ) n6a + 3(0.61)n6a ) 2.83n6a . This
affords n6a /n1a ) 0.35 or 35% and 65% yields of 6a and 5b, respectively,
as deduced from the ratio of the Cp ligand integrations and the
observation that all of the substrate 1a had been converted to 6a and
5b.
1
6a to 15-20% (peak broadening of the H NMR spectra
was problematic). The higher yields of 5b when using
2b as the precatalyst facilitated the isolation of 5b by
size-exclusion chromatography on polystyrene beads. It
(32) Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic Compounds;
Wiley: New York, 1994; p 68.
(33) 1H NMR spectra of isolated 4b and 5b (with discussion of their
assignments) as well as of the reaction mixtures of 2a - and 2c-catalyzed
PhSiH3 hydrosilation of 1a are deposited in the Supporting Informa-
tion.
(34) (a) Akita, M.; Oku, T.; Moro-oka, Y. J . Chem. Soc., Chem.
Commun. 1989, 1790. Akita, M.; Oku, T.; Tanaka, M.; Moro-oka, Y.
Organometallics 1991, 10, 3080. (b) Brunner, H.; Fisch, K. Ang. Chem.,
Int. Ed. Engl. 1990, 29, 1131; J . Organomet. Chem. 1991, 412, C11.

Deoxygenation of Iron Acyl Complexes
Organometallics, Vol. 17, No. 10, 1998 1999
Sch em e 3
was very difficult to completely remove all of 4b,
although benzene easily eluted 100-150 mg samples of
5b that were contaminated with 5-10% 4b.
of 1a (0.20 mmol) with 1.6 equiv of PhSiH3 and 3.2%
1
RhCl(PPh3)3 in C6D6, as deduced by H and 13C NMR
spectroscopy. The catalyst and silane concentrations
were optimized for the rapid depletion of substrate 1a
(40 to 60 min), conditions under which neither Cp2-
Fe2(CO)4 (δ 4.22) nor Cp(CO)2FeH (δ -11.8) were
detected. Isolated yields of spectroscopically pure 6a
varied from 48% (0.50 mmol scale) to 51% (0.20 mmol
scale) after chromatography on alumina.
The precatalyst (PPh3)(CO)4MnC(O)CH3 (2c) differed
in activity from 2a and 2b in that PhSiH3 hydrosilation
of 1a yielded [Cp(CO)2FeCH(CH3)O]2SiHPh (4b) as the
major product. The use of 2c (3.0 mol %) as the
precatalyst consumed 1a within 0.5 h and selectively
yielded 4b and 6a as the only NMR detectable products.
With 1.2 equiv of PhSiH3, the yields of 4b and 6a went
from 71% and 10% after 0.5 h to 66% and 19% after 5
h. With 0.50 equiv of PhSiH3 under otherwise identical
conditions, 1a transformed to mixtures containing 73%
4b plus 6% 6a .33
The 1H NMR spectrum of [Cp(CO)2FeCH(CH3)O]2-
SiHPh (4b) also exhibits four sets of NMR absorptions
for its Cp(CO)2FeCHMeO groups. These NMR absorp-
tions are anticipated for a constitutionally symmetrical
molecule that has two stereogenic Cp(CO)2FeCHMeO
groups separated by the SiHPh center. This structure
affords four stereoisomers, a pair of enantiomers having
a prochiral center [R‚R (S‚S)] and two meso diastere-
oisomers each with their pseudoasymmetric centers
[RrS and RsS]. The meso diastereoisomers account for
two sets of 1H and 13C NMR absorptions for their
Cp(CO)2FeCHMeO groups, whereas the enantiomeric
pair exhibits two sets of NMR absorptions due to the
presence of their diastereotopic and magnetically non-
equivalent Cp(CO)2FeCHMeO moieties. Detailed NMR
assignments for these sets of NMR absorptions of 4b,
however, were more difficult to make than those for 5b
because of overlapping of the NMR spectral absorp-
tions.33
2. Rh Cl(P P h 3)3-Ca ta lyzed P h SiH3 Hyd r osila tion
of Ir on Acyl Com p lexes Cp (CO)(L)F eC(O)R. The
RhCl(PPh3)3-catalyzed PhSiH3 hydrosilation of Cp(CO)2-
FeC(O)CH3 (1a ) is the most efficient procedure for
reducing it to Cp(CO)2FeCH2CH3 (6a ), albeit at the
expense of producing another organoiron productsthe
known16b,19 vinyl complex Cp(CO)2FeCHdCH2 (7a ), eq
2. A mixture of 7a (32%), 6a (49%), and [Cp(CO)2FeCH-
(CH3)O]2SiHPh (4b) (12-14%) results from the reaction
To probe the mechanism for the RhCl(PPh3)3-cata-
lyzed PhSiH3 hydrosilation of 1a , we looked at the
independent reactions of several Cp(CO)2Fe-R-siloxy-
ethyl complexes with RhCl(PPh3)3/PhSiH3 (Scheme 3).
The addition of catalytic quantities of RhCl(PPh3)3 had
1a
no effect on mixtures of Cp(CO)2FeCH(OSiMe2Ph)CH3
or [Cp(CO)2FeCH(CH3)O]3SiPh (5b) with PhSiH3 over
at least 3 h. We did note, however, the much slower
dehydrogenative silation and oligomerization of PhSiH3
to at least PhH2SiSiH2Ph (10%).7
In contrast, [Cp(CO)2FeCH(CH3)O]2SiHPh (4b) proved
to be a more responsive substrate toward RhCl(PPh3)3/
PhSiH3. The substrate 4b was generated conveniently
from 2c-catalyzed hydrosilation of 1a (0.50 equivalent
of PhSiH3) as a mixture containing 72% 4b and 2% 6a
but no residual PhSiH3. Although the addition of 3%
RhCl(PPh3)3 to this mixture had no effect, repeating the

2000 Organometallics, Vol. 17, No. 10, 1998
Mao et al.
Sch em e 4
1
same procedure and adding another 1.0 equiv of PhSiH3
yielded 12% 6a after 2 h.
(CH3)3 (11), Cp[P(OMe)3](CO)FeC(O)CH3 (1c), and Cp-
[P(OPh)3](CO)FeC(O)CH3 (1d )swere subjected to simi-
lar hydrosilation procedures. The resulting fully reduced
alkyl complexes are reported in Table 2 as both their
in situ NMR spectral and isolated yields after chroma-
tography. Both yields, in good agreement with one
another, ranged from 58% (for 9) to 87% (1d ). The
model substrate 1a under comparable reaction condi-
tions, in contrast, afforded somewhat lower yields of 6a ,
due to the presence of 4b and especially the vinyl
complex 7a . Of the five new iron acyl compounds, only
the trimethyl-phosphite-substituted acetyl complex 1c
provided detectable concentrations of a vinyl complex,
Wilkinson’s compound also catalyzed the PhSiH3
hydrosilation-deoxygenation of Cp(PPh3)(CO)FeC(O)-
CH3 (1b). The addition of 1.6 equiv of PhSiH3 and 3.2%
RhCl(PPh3)3 promptly transformed 1b to a mixture
containing 61% Cp(PPh3)(CO)FeCH2CH3 (6b), 11% Cp-
(PPh3)(CO)FeCHdCH2 (7b), and 11% Cp(PPh3)(CO)FeH
(8b) (eq 3). A slightly higher yield of the ethyl com-
2a
Cp[P(OMe)3](CO)FeCHdCH2 (7c). Catalytic hydro-
silationsdeoxygenation, using RhCl(PPh3)3 and Ph2SiH2
in CH2Cl2, also has been reported by Akita, Moro-oka,
and co-workers.3 They transformed 1b to 6b plus
varying amounts of 7b and Cp(CO)2FeC(O)Ph (9) to Cp-
(CO)2FeCH2Ph plus Cp(CO)2FeCH(OMe)Ph after meth-
anolysis.
A plausible pathway for the rhodium(I)-catalyzed
hydrosilation-reduction of iron acyl complexes to their
fully reduced alkyl compounds (Scheme 4) has as its first
step catalytic hydrosilation that gives an R-siloxyalkyl
intermediate, Cp(L)(CO)FeCH(OSiH2Ph)R (3). In this
first step, RhCl(PPh3)2 as the active catalyst promotes
hydrosilation of the acyl complex analogous to the
accepted mechanism for the Rh(I)-catalyzed hydrosila-
tion of ketones.6,35 We emphasize, however, that al-
though this latter reaction often provides vinyl silyl
ethers as byproducts (e.g., PhC(OSiR3)dCH2 from PhC-
(O)CH3),36 full reduction of the acyl functionality to a
methylene group is rare.6,37
pound 6b (62%) was realized after column chromatog-
raphy of a larger scale reaction (0.50 mmol). The
1
presence of 7b and 8b was signaled by diagnostic H
NMR spectral absorptions for their vinyl and iron
hydride spin systems, respectively.
Five other iron acyl compoundssCp(CO)2FeC(O)Ph
(9), Cp(CO)2FeC(O)CH(CH3)2 (10), Cp(CO)2FeC(O)C-

Deoxygenation of Iron Acyl Complexes
Organometallics, Vol. 17, No. 10, 1998 2001
The resulting R-siloxyalkyl intermediate 3 then con-
tinues to react with the active catalyst and PhSiH3 (or
Ph2SiH23), perhaps via an active catalyst‚PhSiH3 adduct
(PhH2Si)(H)RhCl(PPh3)2.6,38 This hydrido(silyl)rhod-
ium(III) transient would transfer hydride to 3 concur-
rent with abstracting the R-siloxy group as a disiloxane.
A potential driving force for loss of the R-siloxy group
relates to the established proclivity of organometallic
centers to ionize even poor leaving groups from a sp3
hybridized R-carbon.39 This ionization step is commonly
used to produce stable alkylidene derivatives such as
(PPh3)3 as the precatalyst and 1a as the substrate gave
<15% [Cp(CO)2FeCH(CH3)O]2SiHPh (4b) as the minor
byproduct to Cp(CO)2FeCH2CH3 (6a ) and Cp(CO)2-
FeCHdCH2 (7a ). Nevertheless, isolated samples of 4b
when subjected to RhCl(PPh3)3/PhSiH3 catalysis only
slowly reduced to Cp(CO)2FeCH2CH3 (6a ); in separate
reactions, 5b, as well as Cp(CO)2FeCH(OSiMe2Ph)CH3,
remained inert. It is of course possible that acyl
substrates with more electron-rich iron centers, espe-
cially 1b, may form (undetected) mono-, bis-, and tris-
R-siloxyethyl intermediates [Cp(L)(CO)FeCH(CH3)O]x-
SiH3-xPh (3, x ) 1; 4, x ) 2; 5, x ) 3) that undergo
subsequent catalytic deoxygenation or deprotonation
along with loss of disiloxane.
A reaction pathway similar to that advanced in
Scheme 4 for the hydrosilation-deoxygenation of iron
acyl complexes using RhCl(PPh3)3 as the catalyst also
applies to the manganese acyl precatalysts 2a -c. These
precatalysts could afford coordinatively unsaturated
L(CO)3MnSiH2Ph (L ) CO, PPh3) as the presumed
active catalysts.4a Analogous manganese silyls, gener-
ated by hydrosilation of 2a -c with excess hydrosilane
and by photolysis of (CO)5MnSiR3, are postulated active
catalysts for the hydrosilation of ketones,5c esters,5b and
Cp(CO)2FeC(O)R.1b,e Indeed, treatment of (CO)5MnC-
(O)CH3 (1a ) with PhSiH3 yields R-siloxyethyl complexes
[(CO)5MnCH(CH3)O]xH3-xSiPh (3c-5c);33 their degra-
dation was reminescent of the hydrosilation chemistry
of 2a -c with other hydrosilanes to generate transient
concentrations of the purported hydrosilation active
catalysts.1a,b,d Studies in progress address the proposed
use of active catalyst-hydrosilane adducts, L(CO)3Mn-
(H)(SiR3)2, as reducing agents toward organic silyl
acetals and organometallic R-siloxyalkyl complexes.
+
39a,b
Cp(L)(CO)FedCHCH3
,
which are reduced by hy-
dride donors (giving 6) or are deprotonated by weak
bases (giving 7).24a Accordingly, we hypothesize that
hydrido(silyl)rhodium(III), while abstracting the R-si-
loxy group, also competitively deprotonates 3 to a vinyl
complex 7 concurrent with releasing disiloxane and H2-
RhCl(PPh3)2 (then H2).
Supporting evidence for the ionization-deoxygenation
step is available from the studies of Akita, Moro-oka,
and co-workers3 on the RhCl(PPh3)2-catalyzed reactions
of Ph2SiH2 with iron R-methoxyalkyl complexes. They
observed that RhCl(PPh3)3 in the presence of Ph2SiH2
catalytically deoxygenated Cp(PPh3)(CO)FeCH(OCH3)R
(R ) H, CH3) to give Cp(PPh3)(CO)FeCH2R, although
Cp(CO)2FeCH2OCH3 was inert. This diminution in
reactivity by switching to the comparatively electron-
deficient Cp(CO)2Fe moiety can be attributed to its
diminished capability of promoting the alkoxide ioniza-
tion or otherwise stabilizing an incipient positive charge
at the R-carbon.
We propose that the RhCl(PPh3)2‚PhSiH3 adduct
preferentially reduces the mono-R-siloxyethyl interme-
diates Cp(L)(CO)FeCH(OSiH2Ph)CH3 (3), as opposed to
their bis- and tris-counterparts 4 and 5. The intermedi-
ates 3-5 were detected only during the catalytic PhSiH3
hydrosilation of 1a . With this substrate, choice of a
manganese acyl precatalyst 2a-c gave variable amounts
of 3b-5b, although 3b never accumulated in more than
18% of the converted 1a and never appeared as part of
the final product array. These observations correspond
to using 2a -c as catalysts, whereas choice of RhCl-
Con clu sion s
The PhSiH3/RhCl(PPh3)3 catalytic system affords a
convenient synthetic procedure for the hydrosilation
then deoxygenation of organoiron acyl complexes Cp-
(L)(CO)FeC(O)R to yield their alkyl derivatives Cp(L)-
(CO)FeCH2R. We used the transformation of Cp(CO)2-
FeC(O)CH3 (1a ) to Cp(CO)2FeCH2CH3 (6a ) as the
control reaction for optimizing catalysis, although this
substrate of the seven iron acyl complexes studied
provided the lowest yields (49%) of its reduced alkyl.
With 3% RhCl(PPh3)3 and 1.6 equiv of PhSiH3, the
reduced yields of 6a corresponded to the presence of
Cp(CO)2FeCHdCH2 (7a ) (32%) and [Cp(CO)2FeCH-
(CH3)O]2SiHPh (4b) (12-14%). The other iron acyl
complexes provided their fully reduced alkyl derivatives
(0.50 mmol scale reactions) in 58-87% isolated yields
after chromatography, for which analogous vinyl byprod-
ucts were observed with only two other iron acyl
substrates.
The manganese carbonyl acyl complexes L(CO)4MnC-
(O)R (L ) CO, R ) CH3 (2a ); L ) CO, R ) Ph (2b); L )
PPh3, R ) CH3 (2c)) also served as viable PhSiH3
hydrosilation precatalysts for 1a . With 1 equiv of
PhSiH3 and 3% manganese acyl precatalyst, these
reactions produced Cp(CO)2FeCH(CH3)OSiH2Ph (3b),
[Cp(CO)2FeCH(CH3)O]2SiHPh (4b), and [Cp(CO)2FeCH-
(CH3)O]3SiPh (5b) plus variable amounts of 6a (maxi-
mum 35% with 2a ) but no 7a . Depending on the choice
of precatalyst, both 4b and 5b were isolated whereas
(35) (a) Ojima, I.; Kogure, T.; Kumagai, M.; Horiuchi, S.; Sato, T. J .
Organomet. Chem. 1976, 122, 83. Ojima, I.; Kogure, T.; Kumagai, M.
J . Org. Chem. 1977, 42, 1671. Ojima, I.; Kogure, T. Organometallics
1982, 1, 1390. (b) Hayashi, T.; Yamamoto, K.; Kumada, M. J .
Organomet. Chem. 1976, 113, 127. (c) Corriu, R. J . P.; Moreau, J . J .
E. Nouv. J . Chim. 1977, 1, 71. (d) Peyronel, J . F.; Kagan, H. B. Nouv.
J . Chim. 1978, 2, 211. Kagan, H. B.; Peyronel, J . F.; Yamagishi, T. In
Adv. Chem. Ser. 1979, 173, 50. (e) Kolb, I.; Hetflejs Collect. Czech.
Chem. Commun. 1980, 45, 2224. (f) Zheng, G. Z.; Chan, T. H.
Organometallics 1995, 14, 70 and references therein.
(36) (a) Benes, J .; Hetflejs, J . Collect. Czech. Chem. Commun. 1976,
41, 2264. (b) Sakurai, H.; Miyoshi, K.; Nakadaira, Y. Tetrahedron Lett.
1977, 2671. (c) Matsumoto, H.; Hoshino, Y.; Nagai, Y. Bull. Chem. Soc.
J pn. 1981, 54, 1279. (d) Payne, N. C.; Stephan, D. W. Inorg. Chem.
1982, 21, 182. (e) Lappert, M. F.; Maskell, R. K. J . Organomet. Chem.
1984, 264, 217. (f) Brunner, H.; Becker, R.; Riepl, G. Organometallics
1984, 3, 1354. Brunner, H.; Kurzinger, A. J . Organomet. Chem. 1988,
346, 413. (g) Ojima, I.; Donovan, R. J .; Clos, N. Organometallics 1991,
10, 2606. (h) Wright, M. E.; Cochran, B. B. J . Am. Chem. Soc. 1993,
115, 2059; Organometallics 1996, 15, 317.
(37) Cullen, W. R.; Evans, S. V.; Han, N. F.; Trotter, J . Inorg. Chem.
1987, 26, 514.
(38) Reviews on metal silyl complexes: (a) Tilley, T. D. In The
Chemistry of Organic Silicon Compounds; Patai, S., Rappoport, Z.,
Eds.; Wiley Interscience: New York, 1989; Part 2, Chapter 24. (b)
Tilley, T. D. In The Silicon-Heteroatom Bond; Patai, S., Rappoport,
Z., Eds.; Wiley: New York, 1992; Chapter 10.
(39) (a) Brookhart, M.; Studabaker, W. B. Chem. Rev. 1987, 87, 411.
(b) Cutler, A. R.; Hanna, P. K.; Vites, J . C. Chem. Rev. 1988, 88, 1363.
(c) Bruce, M. I. Chem. Rev. 1991, 91, 197.

2002 Organometallics, Vol. 17, No. 10, 1998
Mao et al.
3b was only detected by 1H and 13C NMR spectral
monitoring. We believe that 3b, as opposed to 4b or
5b, preferentially undergoes reduction to 6a with the
manganese acyls 2a -c (or RhCl(PPh3)3) as the precata-
lysts. Consistent with this hypothesis, samples of 4b
and 5b were poor substrates for PhSiH3/RhCl(PPh3)3
reduction to 6a . We surmise that similar intermediates
Cp(L)(CO)FeCH(OSiH2Ph)R (3) serve as precursors to
other alkyl products, Cp(L)(CO)FeCH2R).
of Energy, Office of Basic Energy Science. We also are
indebted to Dr. Paul Hanna and Edward Crawford (RPI)
for their preliminary observations.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of
isolated 4b and 5b (with discussion of their assignments) and
of the reaction mixtures of 2a - and 2c-catalyzed PhSiH3
hydrosilation of 1a , as well as spectroscopic data, and discus-
sion of the reaction of 2a and PhSiH3 (9 pages). Ordering
information is given on any current masthead page.
Ack n ow led gm en t. We gratefully acknowledge re-
ceiving generous financial support from the Department
OM970659M