Synthesis and reactivity of heterodinuclear Fe-Ni complexes with a bridging alkoxysilyl ligand. Crystal structure of [(OC)3Fe{μ-Si(OMe)2(OMe)}(μ-dppm)NiMe]

Article abstract of DOI:10.1021/om010619p

The new heterobimetallic complex [(OC)₃Fe- {μ-Si(OMe)₂(OMe)}(μ-dppm)NiCl] (1) was obtained in 95% yield by reaction of [NiCl₂(PPh₃)₂] in THF with K[Fe{Si(OMe)₃}(CO)₃(dppm-P)] at -78°C. The analogous bromo and iodo complexes were also obtained; the latter is, however, less stable and could not be isolated pure. They display the first examples of a bridging alkoxysilyl ligand between Fe and Ni, and the μ₂-η²-SiO bridge is also present in the methyl complex [(OC)₃Fe{μ-Si(OMe)₂(OMe)}(μ-dppm)NiMe] (4) and its phenyl analogue 5. The presence of the Fe-Si-O→Ni four-membered rings was confirmed by a crystal structure determination of 4. Treatment of 1 with excess (allyl)MgCl led to the expected bimetallic allyl complex [(OC)₃{(MeO)₃Si}Fe(μ-dppm)Ni(η³- C₃H₅)] (6). The rapid η³-allyl → η¹-allyl → η³-allyl rearrangement is potentially assisted through stabilization of the coordinatively unsaturated Ni center by a SiO→Ni interaction. The bimetallic benzyl derivative 7 was also isolated. Purging a THF or benzene solution of 4 at room temperature with CO yielded after a few seconds the acyl complex [(OC)₃Fe{μ-Si(OMe)₂(OMe)}(μ-dppm)-NiC(O)Me] (8), which readily decarbonylates. Its reaction with norbornadiene leads to the insertion product, which is thought to exist as an isomeric mixture with terminal or chelating acyl group (11 ? 11′). When complex 4 was reacted with ^tBuNC, rapid insertion occurred and further coordination of a terminal ^tBuNC ligand to Ni led to the iminoacyl complex [(OC)₃{(MeO)₃Si}Fe(μ-dppm)Ni{C(N^tBu)Me} (CN^tBu)] (12). Complex 1 proved to be a more efficient catalyst (TON = 4100) for the dehydrogenative coupling of Ph₃SnH than its mononuclear counterpart [NiCl₂(PPh₃)₂] (TON = 1050). The maximum turnover frequency (TOF) was ca. 9800 h^-1.

Full text of DOI:10.1021/om010619p

5036
Organometallics 2001, 20, 5036-5043
Syn th esis a n d Rea ctivity of Heter od in u clea r F e-Ni
Com p lexes w ith a Br id gin g Alk oxysilyl Liga n d . Cr ysta l
Str u ctu r e of [(OC)3F e{µ-Si(OMe)2(OMe)}(µ-d p p m )NiMe]†
Pierre Braunstein,* Guislaine Clerc, and Xavier Morise
Laboratoire de Chimie de Coordination, UMR CNRS 7513, Universite´ Louis Pasteur,
4 rue Blaise Pascal, F-67070 Strasbourg Cedex, France
Received J uly 10, 2001
The new heterobimetallic complex [(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm)NiCl] (1) was
obtained in 95% yield by reaction of [NiCl2(PPh3)2] in THF with K[Fe{Si(OMe)3}(CO)3(dppm-
P)] at -78 °C. The analogous bromo and iodo complexes were also obtained; the latter is,
however, less stable and could not be isolated pure. They display the first examples of a
bridging alkoxysilyl ligand between Fe and Ni, and the µ2-η2-SiO bridge is also present in
the methyl complex [(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm)NiMe] (4) and its phenyl analogue
5. The presence of the Fe-Si-OfNi four-membered rings was confirmed by a crystal
structure determination of 4. Treatment of 1 with excess (allyl)MgCl led to the expected
bimetallic allyl complex [(OC)3{(ΜeÃ)3Si}Fe(µ-dppm)Ni(η3-C3H5)] (6). The rapid η3-allyl f
η1-allyl f η3-allyl rearrangement is potentially assisted through stabilization of the
coordinatively unsaturated Ni center by a SiOfNi interaction. The bimetallic benzyl
derivative 7 was also isolated. Purging a THF or benzene solution of 4 at room temperature
with CO yielded after a few seconds the acyl complex [(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm)-
NiC(O)Me] (8), which readily decarbonylates. Its reaction with norbornadiene leads to the
insertion product, which is thought to exist as an isomeric mixture with terminal or chelating
t
acyl group (11 h 11′). When complex 4 was reacted with BuNC, rapid insertion occurred
t
and further coordination of a terminal BuNC ligand to Ni led to the iminoacyl complex
[(OC)3{(ΜeÃ)3Si}Fe(µ-dppm)Ni{C(NtBu)Me}(CNtBu)] (12). Complex 1 proved to be a more
efficient catalyst (TON ) 4100) for the dehydrogenative coupling of Ph3SnH than its
mononuclear counterpart [NiCl2(PPh3)2] (TON ) 1050). The maximum turnover frequency
(TOF) was ca. 9800 h-1
.
In tr od u ction
of molecules.5-7 We have concentrated on iron-contain-
ing bimetallics, in part because the required iron-silyl
precursor is readily available.6,8 The discovery that a
trialkoxysilyl ligand can bridge between two metal
centers has opened the way to reactivity studies based
on the often-observed lability of the dative OfM bond
The growing interest in both heterometallic chem-
istry1-4 and organosilicon chemistry has led us to
investigate new aspects of heterobimetallic silicon chem-
istry, and a unique interplay between the bimetallic core
and the silicon-based ligand was observed in a number
in the resulting four-membered Fe-Si-OfM ring
(Scheme 1).6,9 The resulting hemilability10 of the -Si-
(OR)3 ligand is most pronounced in the case where M
) Pd, and we have investigated with such compounds
the multiinsertion of organic isonitriles, the sequential
† Dedicated to Professor F. Mathey, on the occasion of his 60th
birthday, with our most sincere congratulations and best wishes.
(1) Chetcuti, M. J . In Comprehensive Organometallic Chemistry II;
Abel, E. W., Stone, F. G. A., Wilkinson, G., Eds.; Pergamon: Oxford,
U.K., 1995; Vol. 10, p 23.
(2) Braunstein, P.; Rose´, J . Catalysis and Related Reactions with
Compounds containing Heteronuclear Metal-Metal Bonds. In Com-
prehensive Organometallic Chemistry II; Abel, E. W., Stone, F. G. A.,
Wilkinson, G., Eds.; Pergamon: Oxford, U.K., 1995; Vol. 10, p 351.
(3) Braunstein, P.; Rose´, J . Heterometallic Clusters for Heteroge-
neous Catalysis. In Catalysis by Di- and Polynuclear Metal Cluster
Complexes; Adams, R. D., Cotton, F. A., Eds.; Wiley-VCH: New York,
1998; p 443.
(5) Braunstein, P.; Knorr, M. J . Organomet. Chem. 1995, 500, 21.
(6) Braunstein, P.; Knorr, M.; Stern, C. Coord. Chem. Rev. 1998,
178-180, 903.
(7) Braunstein, P.; Knorr, M.; Reinhard, G.; Schubert, U.; Sta¨hrfeldt,
T. Chem. Eur. J . 2000, 6, 4265.
(8) Braunstein, P.; Knorr, M.; Schubert, U.; Lanfranchi, M.; Tirip-
icchio, A. J . Chem. Soc., Dalton Trans. 1991, 1507.
(9) Braunstein, P.; Knorr, M.; Tiripicchio, A.; Tiripicchio-Camellini,
M. Angew. Chem., Int. Ed. Engl. 1989, 28, 1361.
(4) Braunstein, P.; Rose´, J . Heterometallic Clusters in Catalysis.
In Metal Clusters in Chemistry; Braunstein, P., Oro, L. A., Raithby,
P. R., Eds.; Wiley-VCH: Weinheim, Germany, 1999; Vol. 2, p 616.
(10) Braunstein, P.; Naud, F. Angew. Chem., Int. Ed. 2001, 40, 680.
10.1021/om010619p CCC: $20.00 © 2001 American Chemical Society
Publication on Web 10/20/2001

Heterodinuclear Fe-Ni Complexes
Organometallics, Vol. 20, No. 24, 2001 5037
Sch em e 1
CO/olefin coupling which leads to perfectly alternating
polyketones,11,12 and the catalytic dehydrogenative cou-
pling of stannanes.13-16 These studies were facilitated
by the fact that complexes such as [(OC)3Fe{µ-Si(ÃΜe)2-
(ÃΜe)}(µ-dppm)PdX] (dppm ) Ph2PCH2PPh2; X ) Me,
Cl, ...) are both stable, largely because of the presence
of the bridging diphosphine ligand, and reactive, owing
to the kinetic lability of the dative SiOfPd interaction
(Scheme 1).
We have also prepared related Fe-Pt complexes and
ever, it formed when 1 was reacted with NaI (eq 1).
Unfortunately, complex 3 could not be isolated pure,
owing to the occurrence of decomposition products
during halide metathesis. The spectroscopic and ana-
lytical data for 1-3 are in agreement with the proposed
structure (see Experimental Section and Table 1). In
particular, the room-temperature 1H NMR spectra
contain for the methoxy protons two singlets in the
range δ 3.70-3.68 and 3.26-3.54, in a 2:1 ratio, indicat-
ing the presence of a µ2-η2-SiO bridge between the Fe
and Ni atoms and a static situation on the NMR time
scale at room temperature for the silyl ligand. This is
in contrast to the observations made with the analogous
Fe-Pd complex, where a dynamic situation is observed
at room temperature which exchanges the OMe groups.9
Increasing the temperature led to decomposition of the
Fe-Ni complex 1 before coalescence of these resonances
could be reached. The 31P{1H} NMR spectra of 1-3
observed that the lability of the Fe-Si-OfPt four-
membered ring is considerably reduced.17-20 It became
then particularly interesting to investigate the synthesis
of heterobimetallic Fe-Ni complexes containing a bridg-
ing alkoxysilyl ligand and compare their reactivity with
that of their Pd and Pt analogues.
Resu lts a n d Discu ssion
1. Syn th esis of F e-Ni Com p lexes. 1.1. Ha lid e
Com p lexes. The new complex [(OC)3Fe{µ-Si(OMe)2-
(OMe)}(µ-dppm)NiCl] (1) was prepared by reaction of
[NiCl2(PPh3)2] in THF with the potassium salt of the
silyliron metalate [Fe{Si(OMe)3}(CO)3(dppm-P)]- 8 (in
a 1:1 molar ratio) (eq 1). Although 1 is stable for weeks
in the solid state under an inert atmosphere, it decom-
poses upon exposure to air or in solution at room
temperature. Thus, to minimize the formation of side
products, such as [Ni(CO)2(dppm-P,P′)],21 the reaction
was carried out at -78 °C and the temperature main-
tained below 0 °C during workup (see Experimental
Section). Under these conditions 1 could be isolated as
a green solid in 95% yield.
show the expected AX pattern for the Fe- and Ni-bound
2+3
P atoms with a
J
value of ca. 70 Hz (Table 1). This
PP
value is in the range of those observed for other Fe(µ-
dppm)Pd complexes presenting a SiOfM interaction (M
) Pd, Pt, Rh, etc.), whereas in its absence, 2+3J PP values
larger than 100 Hz are generally observed.6,22,23 Al-
though many parameters contribute to the 31P NMR
chemical shift, we note that changing the nature of the
halide ligand on the Ni center induces an important
variation of the chemical shift value of the Ni-bound P
atom, whereas that of the Fe-bound P atom is unaf-
fected. Thus, on going from X ) Cl to X ) I, a
deshielding is observed (δ 9.4 for 1 vs δ 20.9 for 3). An
opposite but less pronounced trend was noticed for the
Pd-bound P atom in related Fe-Pd complexes (δ 34.9
for Fe-Pd-Cl9 vs δ 32.4 for Fe-Pd-I24). In general,
complexes 1-3 appear less stable than their Pd and Pt
analogues, both in the solid state and in solution. This
is likely to be due to a weaker metal-metal bond
The analogous bromo complex 2 was similarly pre-
pared from [NiBr2(PPh3)2], but in only 40% isolated
yield, whereas the iodo derivative 3 was not obtained
by reacting the iron metalate with [NiI2(PPh3)2]. How-
(11) Braunstein, P.; Knorr, M.; Sta¨hrfeldt, T. J . Chem. Soc., Chem.
Commun. 1994, 1913.
(12) Braunstein, P.; Cossy, J .; Knorr, M.; Strohmann, C.; Vogel, P.
New J . Chem. 1999, 23, 1215.
(13) Braunstein, P.; Morise, X.; Blin, J . J . Chem. Soc., Chem.
Commun. 1995, 14, 1455.
(14) Braunstein, P.; Morise, X. Organometallics 1998, 17, 540.
(15) Braunstein, P.; Durand, J .; Morise, X.; Tiripicchio, A.; Ugozzoli,
F. Organometallics 2000, 19, 444.
(16) Braunstein, P.; Morise, X. Chem. Rev. 2000, 100, 3541.
(17) Knorr, M.; Faure, T.; Braunstein, P. J . Organomet. Chem. 1993,
447, C4.
(18) Knorr, M.; Braunstein, P.; Tiripicchio, A.; Ugozzoli, F. Orga-
nometallics 1995, 14, 4910.
(19) Knorr, M.; Braunstein, P.; DeCian, A.; Fischer, J . Organome-
tallics 1995, 14, 1302.
(22) Braunstein, P.; Knorr, M.; Piana, H.; Schubert, U. Organome-
tallics 1991, 10, 828.
(23) Braunstein, P.; Faure, T.; Knorr, M.; Balegroune, F.; Grandjean,
D. J . Organomet. Chem. 1993, 462, 271.
(20) Braunstein, P.; Faure, T.; Knorr, M.; Sta¨hrfeldt, T.; DeCian,
A.; Fischer, J . Gazz. Chim. Ital. 1995, 125, 35.
(21) van Hecke, G. R.; Horrocks, W. D. Inorg. Chem. 1966, 5, 1960.
(24) Balegroune, F.; Braunstein, P.; Durand, J .; Faure, T.; Grand-
jean, D.; Knorr, M.; Lanfranchi, M.; Massera, C.; Morise, X. Tiripichhio,
A. Monatsh. Chem. 2001, 132, 885.

5038 Organometallics, Vol. 20, No. 24, 2001
Braunstein et al.
Ta ble 1. Selected 31P {1H} NMR a n d IR Da ta of
Com p lexes 1-12
δ(31P) (ppm)a
2+3
b
complex
PFe
PNi
J
(Hz)
ν(CO), ν(CN) (cm-1
)
pp
1
2
3
4
55.3
53.8
52.1
64.9
9.4
13.0
20.9
32.9
72
1982 s, 1931 vs, 1895 s
1983 s, 1932 vs, 1897 s
1986 vs, 1936 vs, 1895 sd
1955 s, 1892 vs, 1857 vs
71
70
65
65
66
117
64
67
65.0c 33.9c
5
6
7
8
62.5
26.3
1959 s, 1896 s, 1855 vs
1951 s, 1881 s, 1849 vs
1955 s, 1895 vs, 1858 s
1957 m, 1891 s, 1854 s,
1660 md
1955 s, 1893 vs, 1857 vs,
1717 mw, 1653 mw
2148 m,f 1987 s, 1868 m,
1832 m, 1675 wd,g
71.6c 29.7c
62.1
65.8
27.3
28.3
11
12
60.5e 28.5e
70.5 37.6
67
102
a
b
Recorded in C6D6 unless otherwise stated. Recorded in THF
unless otherwise stated. c Recorded in acetone-d6. Recorded in
d
CH2Cl2. e Recorded in toluene-d8. f ν(CtN). ν(CdN).
g
strength, as usually observed for metal-metal bonds
between first-row elements, which is not fully compen-
sated by the SiOfNi interaction, expected to be stronger
than the SiOfPd and SiOfPt interactions, owing to
the harder character of Ni(II) vs Pd(II) and Pt(II).
1.2. Hyd r oca r byl Com p lexes. To access Fe-Ni
complexes containing a Ni-carbon bond, which would
allow further reactivity studies, we reacted 1 with MeLi
F igu r e 1. ORTEP view of the structure of [(OC)3Fe{µ-Si-
(OMe)2(OMe)}(µ-dppm)NiMe] (4).
Ta ble 2. Selected Bon d Dista n ces (Å) a n d An gles
(d eg) in 4 (Molecu le 1)
Fe1-Ni1
Fe1-C1
Fe1-C2
Fe1-C3
Fe1-P1
Fe1-Si1
Ni1-P2
Ni1-C29
Ni1-O4
2.5597(5)
1.756 (3)
1.767(3)
1.759(3)
2.2098(7)
2.2558(8)
2.1020(7)
1.953(3)
1.978(2)
C1-O1
C3-O3
P1-C16
P2-C16
O4-C30
Si1-O4
Si1-O5
Si1-O6
O5-C31
O6-C32
1.163 (3)
1.153(4)
1.850(2)
1.844(2)
1.440(3)
1.687(2)
1.633(2)
1.636(2)
1.409(4)
1.394(4)
in THF at -78 °C and isolated [(OC)3Fe{µ-Si(OMe)2-
1
(OMe)}(µ-dppm)NiMe] (4) in 40% yield (eq 2). The H
Ni1-Fe1-C1
Ni1-Fe1-C2
Ni1-Fe1-C3
Ni1-Fe1-P1
Ni1-Fe1-Si1
C1-Fe1-C2
C1-Fe1-C3
C2-Fe1-C3
C1-Fe1-Si1
C1-Fe1-P1
C2-Fe1-P1
C2-Fe1-S1
C3-Fe1-P1
C3-Fe1-Sil
P1-Fe1-Si1
64.83(9)
71.35(9)
162.6(1)
97.48(2)
72.75(2)
136.1(1)
113.0(1)
109.8(1)
85.58(9)
90.69(8)
91.87(9)
84.62(9)
99.8(1)
Fe1-Ni1-P2
Fe1-Ni1-C29
Fe1-Ni1-O4
C1-Ni1-P2
P2-Ni1-C29
P2-Ni1-O4
C29-Ni1-O4
Fe1-C1-O1
Fe1-C2-O2
P1-C16-P2
Ni1-O4-C30
Ni1-O4-Si1
C30-O4-Si1
Fe1-Si1-O4
97.29(2)
175.38(9)
83.63(6)
94.99(6)
86.92(9)
178.50(6)
92.2(1)
176.6(3)
178.9(3)
115.1(1)
132.7(2)
102.61(9)
124.4(2)
88.282
NMR spectrum of 4 contains a resonance at δ -0.22
(3J PH ) 3.8 Hz) for the Ni methyl protons and two
resonances at δ 3.70 and 3.27 for the methoxy protons,
in a 2:1 ratio, respectively. This is indicative of the
persistence of the SiO bridge present in 1. The 31P{1H}
90.0(1)
170.22(3)
NMR spectrum shows two doublets at δ 64.9 and 32.9
2+3
(
J
) 65 Hz), ascribed to the P(Fe) and P(Ni) atoms,
PP
containing this bond, such as [(OC)3(Me3P)Fe(µ-PtBu2)-
respectively. Although crystals of 4 are stable for a few
days at -20 °C, noncrystalline solid 4 rapidly decom-
posed, even at low temperature.
Ni(PMe3)Cl] (Fe-Ni ) 2.521 Å).25 In both molecules the
five-membered ring involving the dppm ligand and the
Since complexes 1-4 contain the first examples of
Fe and Ni atoms are almost coplanar with the Fe-Si-
Fe-Si-OfNi four-membered rings, a crystal structure
determination was performed on 4. An ORTEP view is
shown in Figure 1.
OfNi four-membered ring; the dihedral angles between
their mean planes is 176.36(3)° (molecule 1) and
175.14(3)° (molecule 2). The Si-O bond involved in the
µ2-η2-SiO interaction is slightly longer than the other
two (Tables 2 and 3). A similar lengthening has been
observed in all the previous structural determinations
There are two crystallographically different but very
similar molecules in the unit cell, whose selected bond
distances and angles are given in Tables 2 and 3.
Crystal data and data collection parameters are given
in Table 4. The Fe and Ni atoms are connected via a
direct metal-metal bond, and the Fe-Ni bond lengths
of 2.5597(5) Å (in molecule 1) and 2.5688(4) Å (in
molecule 2) are comparable to those in other complexes
of Fe-Si-OfM systems.6 The distorted-trigonal-bipy-
(25) Arif, A. M.; Chandler, D. J .; J ames, R. A. J . Coord. Chem. 1988,
17, 45.

Heterodinuclear Fe-Ni Complexes
Organometallics, Vol. 20, No. 24, 2001 5039
(PMe3)2] (1.948(1) Å).27 The Ni-O bond lengths of
1.978(2) Å (in molecule 1) and 1.982(2) Å (in molecule
2) are within the range of those observed in Na3(µ3-I)-
{Ni[µ3-OSi(OtBu)3]3I}‚0.5THF‚0.5C5H12 (average 1.95
Å)28 or in Ni(II) cyclohexasiloxanolate sandwich com-
plexes (average value 2.03 Å)29 and are close to that
found in NiO (2.1 Å). They are, however, shorter than
the Pd-O and Pt-O bond lengths observed in the
Ta ble 3. Selected Bon d Dista n ces (Å) a n d An gles
(d eg) in 4 (Molecu le 2)
Fe2-Ni2
Fe2-C33
Fe2-C34
Fe2-C35
Fe2-P3
2.5688(4)
1.766(3)
1.777(3)
1.767(3)
2.2077(7)
2.2438(7)
2.1005(7)
1.954(3)
1.982(2)
C33-O7
C35-O9
P3-C48
P4-C49
O10-C62
Si2-O10
Si2-O11
Si2-O12
O11-C63
O12-C64
1.164(3)
1.151(4)
1.846(3)
1.832(3)
1.437(3)
1.687(2)
1.647(2)
1.638(2)
1.410(3)
1.433(4)
Fe2-Si2
Ni2-P4
Ni2-C61
Ni2-O10
related complexes [(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm)-
PdCl] (2.100(4) Å)9 and [(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-
Ni2-Fe2-C33
Ni2-Fe2-C34
Ni2-Fe2-C35
Ni2-Fe2-P3
Ni2-Fe2-Si2
C33-Fe2-C34
C33-Fe2-C35
C34-Fe2-C35
C33-Fe2-Si2
C34-Fe2-P3
C35-Fe2-P3
C34-Fe2-Si2
C33-Fe2-P3
C35-Fe2-Si2
P3-Fe2-Si2
65.06(8)
72.90(9)
161.12(9)
97.62(2)
72.15(2)
137.9(1)
110.8(1)
109.8(1)
85.73(8)
91.51(8)
100.94(9)
83.86(9)
91.58(8)
89.38(9)(1)
169.62(3)
Fe2-Ni2-P4
Fe2-Ni2-C61
Fe2-Ni2-O10
C33-Ni2-P4
P4-Ni2-C61
P4-Ni2-O10
C33-Ni2-O10
Fe2-C33-O7
Fe2-C34-O8
P3-C48-P4
Ni2-O10-C62
Ni2-O10-Si2
C62-O10-Si2
Fe2-Si2-O10
97.12(2)
175.65(7)
83.64(5)
96.33(6)
87.23(7)
178.20(6)
83.16(8)
176.3(2)
178.5(3)
115.8(1)
132.9(2)
101.41(8)
125.1(2)
88.272
dppm)Pt(norbornyl)] (2.217(4) Å),20 respectively, which
is consistent with the smaller radius of Ni(II).
The phenyl analogue of 4, [(OC)3Fe{µ-Si(OMe)2(OMe)}-
(µ-dppm)NiPh] (5), was obtained by reaction of 1 with
PhLi (eq 2), and its spectroscopic data (see Experimental
Section) are similar to those of 4. It is notable that,
although it is stable for several hours in a THF solution,
complex 5 could not be isolated pure, owing to rapid
decomposition once the solvent is evaporated.
Ni-alkylation reactions using Grignard reagents were
also attempted. Thus, treatment of 1 with (allyl)MgCl
led to the expected bimetallic allyl complex [(OC)3-
Ta ble 4. Cr ysta l Da ta a n d Da ta Collection
P a r a m eter s for 4
formula
mol wt
color
C32H34FeNiO6P2Si
719.22
orange
{MeO3Si}Fe(µ-dppm)Ni(η3-C3H5)] (6) (Scheme 2). It was
characterized by comparison of its NMR data with those
of an authentic sample prepared by the better reaction
of K[Fe{Si(OMe)3}(CO)3(dppm-P)] with [Ni(η3-allyl)(µ-
Cl)]2.23 Use of an excess of (allyl)MgCl was, however,
required in order to achieve complete transformation
(31P NMR monitoring). This is probably due to Ni-
cryst syst
a (Å)
monoclinic
15.4860(3)
19.9810(3)
102.620(1)
21.2070(3)
6403.5(3)
8
b (Å)
c (Å)
â (deg)
V (Å3)
Z
Dcalcd (g cm-3
)
1.49
catalyzed C-C coupling side reactions of allylic frag-
2+3
wavelength (Å)
0.710 73
1.221
P21/n
173
ments.30 On the basis of the high
J
value of 117
PP
µ (mm-1
)
Hz in the 31P{1H} NMR spectrum of 6 (see above) and
by analogy with the structure of the Fe-Pd and Fe-Pt
analogues of 6,22 it is proposed that the allyl ligand
adopts an η3 coordination mode, resulting in the absence
of a SiO bridge between the Fe and Ni centers (Scheme
space group
temp (K)
radiation
Mo KR, graphite
monochromated
KappaCCD
0-23, 0-29, -32 to +30
2.5-33.69
2976
diffractometer
hkl limits
θ limits (deg)
F(000)
1
2). However, the presence of broad H NMR resonances
for the allylic protons suggests the existence in solution
of a rapid η3-allyl f η1-allyl f η3-allyl rearrangement
potentially assisted through stabilization of the coordi-
natively unsaturated Ni center by a SiOfNi interaction
(equilibrium 6 h 6′, Scheme 2). Low-temperature stud-
ies were hampered by poor resolution and solubility
problems.23 Such an allyl η3 f η1 f η3 rearrangement
has been investigated in the related Fe-Pd-allyl
system.22 Interestingly, when 1 was treated with 1 equiv
of (allyl)MgBr, a mixture of complexes 2 (major product)
and 6 was obtained, whereas 2 was the sole product if
long reaction times were applied (2 h). This transforma-
tion is faster when 2 equiv of (allyl)MgBr is used.
Monitoring the reaction by 31P NMR revealed that the
no. of data measd
no. of data (I > 3σ(I))
weighting scheme
Ra
50 780
13 449
4Fo2/(σ2(Fo2) + 0.0064Fo
)
4
0.043
0.065
1.212
0.614
a
Rw
GOF
largest peak in final diff map (e Å-3
)
a
2
R ) ∑||Fo| - |Fc||/∑|Fo|; Rw ) [∑w(|Fo| - |Fc|)2/∑w(|Fo| ]1/2
.
ramidal arrangement of the ligands around the Fe
center, when the metal-metal bond is omitted, is also
consistent with previous structures and suggests the
presence of a formally zerovalent Fe center in these
complexes. The Ni(1)-C(1) (molecule 1) and Ni(2)-C(33)
(molecule 2) distances of 2.411(3) and 2.427(3) Å are too
long for semibridging carbonyl ligands. The Ni-Me
distances of 1.953(3) Å (in molecule 1) and 1.954(3) Å
(in molecule 2), for which, to the best of our knowledge
(CSD search), there is no precedent in a heterobimetallic
environment, do not significantly differ from those
usually observed in mononuclear systems such as
[NiMe(acac)(PCy3)] (1.941(7) Å)26 and [NiMe(OPh)-
(26) Barnett, B. L.; Kruger, C. J . Organomet. Chem. 1972, 42, 169.
(27) Kim, Y.-J .; Osakada, K.; Takenaka, A.; Yamamoto, A. J . Am.
Chem. Soc. 1990, 112, 1096.
(28) McMullen, A. K.; Tilley, T. D.; Rheingold, A. L.; Geib, S. J . Inorg.
Chem. 1990, 29, 2228.
(29) Cornia, A.; Fabretti, A. C.; Gatteschi, D.; Pa´lyi, G.; Rentschler,
E.; Shchegolikhina, O. I.; Zhdanov, A. A. Inorg. Chem. 1995, 34, 5383.
(30) Chetcuti, M. J . Nickel-Carbon π-Bonded Complexes. In Com-
prehensive Organometallic Chemistry II; Abel, E. W., Stone, F. G. A.,
Wilkinson, G., Eds.; Elsevier: Oxford, U.K., 1995; Vol. 9, pp 107, and
references therein.

5040 Organometallics, Vol. 20, No. 24, 2001
Braunstein et al.
Sch em e 2
The Ni-benzyl complex 7 was obtained in a manner
similar to 6, from 1 and (PhCH2)MgCl (Scheme 2). Again
an excess of the Grignard reagent was necessary for the
reaction to reach completion. Like its Ni-phenyl ana-
logue 5, complex 7 appeared relatively stable in solution
(only ca. 25% decomposition after 24 h in THF, as
monitored by 31P NMR) but rapidly decomposed in the
absence of solvent, which precluded its isolation in pure
form. It was thus characterized in situ by 31P NMR and
2+3
IR spectroscopic methods. The
J
value of 64 Hz
PP
suggests the presence of a SiOfNi interaction and an
η1 coordination mode for the benzyl ligand (Scheme 2).
A similar situation was found for [(OC)3Fe{µ-Si(OMe)2-
(OMe)}(µ-dppm)PdCH2Ph].34
1.3. Sp ectr oscop ic Da ta . Selected spectroscopic data
are presented in Table 1, and some have been discussed
above. It is perhaps useful to comment briefly on the
IR data for the CO ligands. Whereas they are very
similar within each of the halide and hydrocarbyl series,
a significant shift toward lower wavenumbers is ob-
served on going from the former series to the latter. This
would be consistent with the hydrocarbyl ligands being
stronger donors than the halides, and we see no other
parameter to be invoked, since these molecules are
otherwise identical. In the analogous Fe-Pd complexes,
we also noticed a significant shift of the ν(CO) values,
desired product 6 initially forms, together with 2, and
that a rapid decrease of the intensities of its resonances
occurs, paralleled by an increase of those of 2. Since
treatment of 1 with (phenyl)MgBr selectively led to the
phenyl complex 5, whereas this latter could not be
isolated pure from the reaction of 1 with PhLi, owing
to decomposition in the solid state, formation of the
bromo complex 2 in the reaction of 1 and (allyl)MgBr
certainly results from a displacement of the allyl ligand
in 6 by bromide ions, rather than from a direct halide
metathesis between 1 and the Grignard reagent (Scheme
2). In agreement with this assumption, we observed the
formation of 2 when 1 was reacted with LiBr. In
addition, the reaction between 6 and (allyl)MgBr did not
afford 2 but the silyliron metalate [Fe{Si(OMe)3}(CO)3-
(dppm-P)]-, as evidenced by31P{1H} NMR spectroscopy
(δ 72.8 and -25.8, 2+3J PP ) 89 Hz), formally associated
with a MgBr+ cation (which may be solvated). The
nickel-containing fragment was not isolated but could
be Ni(allyl)2, whose progressive decomposition in solu-
tion could account for the formation of 1,5-hexadiene
(identified by GC).31-33 One could therefore envisage
that (allyl)MgBr transfers an allyl ligand to 6 with
displacement of the Fe-Ni and PfNi bonds according
to eq 3. Note that complex 2 appeared nonreactive
toward Grignard or RLi (R ) Me, Ph, ...) reagents.
going from 1995 s, 1940 vs, and 1923 s for [(OC)3Fe{µ-
Si(OMe)2(OMe)}(µ-dppm)PdCl] (in CH2Cl2) to 1956 m,
1892 s, and 1862 vs for the methyl derivative (in CH2-
Cl2).34 These relatively low ν(CO) values do not imply a
semibridging character for the CO ligands, and the
X-ray structures of [(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm)-
PdCl]9 and of 4 point toward a terminal bonding mode
for these ligands, although weak interactions cannot be
completely ruled out. The lower ν(CO) values found in
the allyl complex 6 compared to those for 5 or 7 are
consistent with the silyl ligand becoming terminal, more
electron density remaining with the Fe center. As
expected from the formation of the Fe-Ni bond, the ν-
(CO) values, even in the hydrocarbyl series, remain
much higher than in the potassium metalate K[Fe{Si-
(OMe)3}(CO)3(dppm-P)] (1928 w, 1847 vs, 1826 s).8
2. Rea ction s w ith CO, Nor bor n a d ien e, or CNtBu .
Bubling CO through a THF or benzene solution of 4 at
room temperature yielded after a few seconds the acyl
complex [(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm)NiC(O)-
(31) Wilke, G.; Bogdanovic, B.; Borner, P.; Breil, H.; Hardt, P.;
Heimbach, P.; Herrmann, G.; Kaminsky, H.-J .; Keim, W.; Kro¨ner, M.;
Mu¨ller, H.; Mu¨ller, E. W.; Oberkirch, W.; Schneider, J .; Tanaka, K.;
Weyer, K. Angew. Chem. 1963, 75, 10.
(32) Wilke, G.; Bogdanovic´, B.; Hardt, P.; Heimbach, P.; Keim, W.;
Kro¨ner, M.; Oberkirch, W.; Tanaka, K.; Steinru¨cke, E.; Walter, D.;
Zimmermann, H. Angew. Chem. 1966, 78, 157-172.
(33) Henc, B.; J olly, P. W.; Wilke, G.; Benn, R.; Hoffmann, E. G.;
Mynott, R.; Schroth, G.; Seevogel, K.; Sekutowski, J . C.; Kru¨ger, C. J .
Organomet. Chem. 1980, 191, 425.
(34) Braunstein, P.; Durand, J .; Kickelbick, G.; Knorr, M.; Morise,
X.; Pugin, R.; Tiripicchio, A.; Ugozzoli, F. J . Chem. Soc., Dalton Trans.
1999, 4175.

Heterodinuclear Fe-Ni Complexes
Organometallics, Vol. 20, No. 24, 2001 5041
of the IR and 31P{1H} NMR spectra and by analogy with
previous studies on related Fe-Pd systems,12 we pro-
pose for this new complex the structures pictured in
Scheme 3, with the coexistence of a terminal and
chelating acyl group (11 h 11′). Attempts to produce
polyketones by further reacting either 11/11′ with NBD
and CO or 8 with an ethylene/CO mixture (1:1 ratio, 40
bar) were unsuccessful. The high affinity of the Ni
center for CO probably accounts for these observations,
although Ni(II) complexes have recently been used with
success in the copolymerization of ethylene and CO.35
The rupture of the metal-metal bond of the precursors
and formation of mononuclear fragments such as [Ni-
(CO)2(dppm-P,P′)]2 and [Fe(CO)3(dppm-P,P′)] were ob-
served instead.
Sch em e 3
When complex 4 was reacted with tBuNC, rapid
formation of the iminoacyl complex [(OC)3{(MeO)3Si}-
Fe(µ-dppm)Ni{C(NtBu)Me}(CNtBu)] (12) was evidenced
by the new IR νCN vibration at 1675 cm-1. An additional
vibration at 2148 cm-1 is indicative of the presence of a
t
terminal Ni-bound BuNC ligand. The 31P{1H} NMR
spectrum of the new complex consisted of two doublets
2+3
at δ 70.5 and 37.6, with a
J
coupling of 102 Hz,
PP
suggesting the absence of a -SiO bridge between the
two metal centers. These data are in favor of the
formulation given for complex 12 (Scheme 3). A pal-
ladium analogue of this complex has been character-
ized.11 Once again, the low stability of the product
during workup precluded its isolation in pure form and
better characterization.
3. Deh yd r ogen a tive Cou p lin g of P h 3Sn H. In
previous studies we have shown that heterobimetallic
alkoxy- and siloxysilyl Fe-Pd complexes were efficient
catalysts for the dehydrogenative coupling of stannanes
(eq 4).13-16 It was established that although the elemen-
Me] (8) (Scheme 3). Insertion of CO into the Ni-Me
bond of 4 was evidenced by the occurrence of a νCO
vibration at 1660 cm-1, by a 1H NMR resonance at δ
1.93 for the acyl protons, and by two new 31P{1H}
resonances at δ 65.8 and 28.3 ppm (2+3J PP ) 67 Hz),
which are assigned to the Fe- and Ni-bound P atoms,
respectively. Isolation of complex 8 in the solid state was
precluded by its tendency to decarbonylate and/or
decompose. Note that formation of 8 should be carefully
monitored, since it reacts further with CO, leading to
decomposition products such as [Ni(CO)2(dppm-P,P′)].21
Attempts to carbonylate the phenyl complex 5 did not
lead to the benzoyl derivative 9; however, when 7 was
treated with CO (in situ IR monitoring), the complex
2Ph3SnH [cat.]8 Ph3Sn-SnPh3 + H2
(4)
tary chemical transformations take place at the Pd
center, the silyl ligand plays a key role in the reaction
process, via formation of transient OfPd interactions.
Furthermore, the beneficial effect of the bimetallic
structure was demonstrated. Since we are not aware of
related catalytic studies with mono- or polynuclear
complexes containing nickel,16 it became of interest to
determine the catalytic properties of the Fe-Ni complex
1 for the dehydrogenative coupling of Ph3SnH. Thus,
when 1 was added to a Et2O solution of Ph3SnH (ca.
10 000-fold excess; similar reaction conditions were used
to determine the catalytic properties of the heterobi-
metallic silyl Fe-Pd complexes14-16), a vigorous evolu-
tion of H2 and the formation of Ph6Sn2 were observed.
Although its intensity rapidly and continuously de-
creased, the gas evolution lasted for ca. 2 h. The reaction
mixture was, however, stirred for a period of 48 h, to
make sure that the reaction went to completion. From
the amount of Ph6Sn2 isolated, a turnover number value
(TON) of ca. 4100 was determined, whereas unreacted
Ph3SnH was recovered. Further runs were carried out
during which the evolution of H2 was monitored, and a
[(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm)NiC(O)CH2Ph] (10)
was formed (νCO 1637 cm-1), which was only observable
for a short period of time (Scheme 3).
By analogy with studies on CO/olefin coupling reac-
tions, which led to the addition of the acyl group of the
Fe-Pd complex analogous to 8 across the exo face of
norbornadiene (NBD),12 we treated 8 with NBD in THF.
IR monitoring showed the disappearance of the νCO
vibration of the acyl group of 8 at 1660 cm-1 and the
occurrence of new vibrations at 1717 and 1653 cm-1. A
new set of two doublets was observed in the 31P{1H}
NMR spectrum at δ 60.5 and 28.5 (2+3J PP ) 67 Hz). The
product could not be isolated in pure form, since as
already mentionned above for similar complexes, de-
composition was observed when volatiles and solvent
were evaporated under reduced pressure. On the basis
(35) (a) Kla¨ui, W.; Bongards, J .; Reiâ, G. J . Angew. Chem., Int. Ed.
2000, 39, 3894. (b) Shultz, C. S.; DeSimone, J . M.; Brookhart, M. 2001,
20, 16.

5042 Organometallics, Vol. 20, No. 24, 2001
Braunstein et al.
carried out in the presence of a radical scavenger, such
as galvinoxyl, the same catalytic performances as in its
absence were observed. This is in favor of a nonradical
pathway for this reaction. However, its mechanistic
details, which probably ressemble those proposed for the
Fe-Pd analogous complexes,14 have not been elucidated.
The fate of 1 could not be determined, mainly because
of the very small amounts of catalyst employed. The use
of 1 as a catalyst for the dehydrogenative coupling of
silanes has been unsuccessful so far.
Con clu sion
We have shown for the first time that the occurrence
of a µ2-η2-SiO bridging mode for an alkoxysilyl ligand
can be extended to the Fe-Ni couple, as established by
the X-ray structure of complex 4. The nickel halide
complexes 1-3 are more stable than their analogues
containing a Ni-carbon bond, which rapidly decom-
posed in the solid state and, except for crystalline 4,
were characterized in situ. Insertion reactions of CO,
F igu r e 2. Catalytic dehydrogenative coupling of Ph3-
SnH: plot of TON vs time for the complexes 1 and [NiCl2-
(PPh3)2] (b, Fe-Ni complex 1; O, [NiCl2(PPh3)2]).
t
norbornadiene, and BuNC into the Ni-C bond of some
Fe-Ni heterobimetallic complexes have been observed
and were much faster than with the related Fe-Pd
complexes. The potential use of these derivatives as
catalysts (precursors) for CO/olefins co-oligomerization
or polymerization is, however, limited by the formation
under CO of Ni-containing side products. However, the
dehydrogenative coupling of Ph3SnH is efficiently cata-
lyzed by complex 1. Interestingly, the latter proves to
be a better catalyst (precursor) than its Fe-Pd ana-
logue. It is also much more efficient than [NiCl2(PPh3)2],
which evidences the beneficial effect of the bimetallic
structure on the reactivity of the Ni center, through the
combined effect of the metal-metal bonding and the
assembling diphosphine and bridging alkoxysilyl ligands.
F igu r e 3. Catalytic dehydrogenative coupling of Ph3-
SnH: plot of TON vs time for the complexes Fe-Ni-Cl
(1) and [(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm)PdCl] (b, Fe-
Ni complex 1; 2, Fe-Pd-Cl complex).
Exp er im en ta l Section
parallel study (under similar reaction conditions) was
performed with [NiCl2(PPh3)2] as catalyst (precursor).
The comparative data are shown in Figure 2 (b, Fe-Ni
complex 1; O, [NiCl2(PPh3)2]).
The heterobimetallic complex 1 proved to be a more
efficient catalyst (TON ) 4100) than its mononuclear
counterpart [NiCl2(PPh3)2] (TON ) 1050). Similar ob-
servations have been made for Pd complexes.14 This
reflects a beneficial effect of the bimetallic structure and
a positive influence of the Fe fragment on the Ni center,
achieved through the metal-metal bonding, the bridg-
ing silyl ligand, and the assembling phosphine ligand.
Interestingly, when it is compared to its Pd analogue
All the reactions were performed using Schlenk tube
techniques under an inert atmosphere of purified nitrogen.
Solvents were freshly distilled under nitrogen from the usual
drying agent prior to use. Nitrogen was passed through BASF
R3-11 catalyst and molecular sieves columns to remove
1
residual oxygen and water. The H and 31P{1H} NMR spectra
were recorded at 300.13 and 121.5 MHz, respectively, on a FT
Bruker AC 300 instrument. Infrared spectra were recorded
in the 4000-400 cm-1 range on
a IFS-66 FTIR Bruker
spectrometer and in the 400-90 cm-1 range on a IFS-113 FTIR
Bruker spectrometer. Samples were prepared as KBr pellets
or in solutions using CaF2 cells.
Syn th esis of [(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm )NiCl]
(1). To a stirred THF solution (10 mL) solution of [NiCl2-
(PPh3)2] (5.24 g, 8.0 mmol), at -78 °C was added a filtered
(Celite pad) THF solution (30 mL) of K[Fe{Si(OMe)3}(CO)3-
(dppm-P)] prepared from [HFe{Si(OMe)3}(CO)3(dppm-P)] (5.18
g, 8.0 mmol) and excess KH.8 After the mixture was stirred
for 0.5 h at -78 °C, it was filtered and the volatiles were
removed under vacuum at 0 °C, giving a green residue. The
latter was washed with pentane (20 mL) and Et2O (25 mL),
affording 1 as a green solid which was dried under vacuum
(6.21 g, 95%). 1H NMR (C6D6, 298 K): δ 8.2-6.8 (m, 20H,
aromatics); 3.70 (s, 6H, Si(OMe)2); 3.26 (s, 3H, µ-SiOMe); 3.11
[(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm)PdCl] (see Figure
3), complex 1 appears to be a much more efficient
catalyst for the dehydrogenative coupling of Ph3SnH.
Not only was a TON value ca. 6 times higher than that
observed for obtained but also the time to reach maxi-
mum conversion is shortened from ca. 48 h for the Fe-
Pd-Cl complex to ca. 2 h for 1. As a consequence, the
maximum turnover frequency (TOF), determined for a
given catalyst and expressed in turnovers per hour,
found for 1 (ca. 9800 h-1) is much higher than that for
the Fe-Pd-Cl complex (ca. 100).13 Note that when the
dehydrogenative coupling of Ph3SnH catalyzed by 1 was
2
(t, 2H, J PH ) 10 Hz, PCH2P). IR (polyethylene): ν(NiCl) 336
cm-1. Anal. Calcd for C31H31ClFeNiO6P2Si: C, 50.34; H, 4.22.
Found: C, 50.02; H, 4.44.

Heterodinuclear Fe-Ni Complexes
Organometallics, Vol. 20, No. 24, 2001 5043
used, complete transformation to the bromo complex 2 was
observed. Complex 6 could not be separated from 1 or 2,
respectively, but it was characterized in situ by 31P{1H} NMR
and IR spectroscopy.
Syn th esis of [(OC)3F e{µ-Si(OMe)2(OMe)}(µ-d p p m )Ni-
Br ] (2). This compound was prepared in a manner similar to
1 from [HFe{Si(OMe)3}(CO)3(dppm-P)] (5.185 g, 8.0 mmol) and
[NiBr2(PPh3)2] (5.95 g, 8.0 mmol). Complex 2 was isolated as
1
F or m a tion of [(OC)3F e{µ-Si(OMe)2(OMe)}(µ-d p p m )Ni-
a green solid (2.51 g, 40%). H NMR (C6D6, 298 K): δ 8.2-6.7
(m, 20H, aromatics); 3.70 (s, 6H, Si(OMe)2); 3.33 (s, 3H,
CH2P h ] (7). This complex was obtained in a manner similar
to 6 from (PhCH2)MgCl (2 M in Et2O) (0.90 mL, 1.8 mmol)
and 1 (0.4 g, 0.59 mmol) in THF (30 mL). The red mixture
was filtered and concentrated to ca. 5 mL, and an aliquot was
analyzed by 31P{1H} NMR and IR spectroscopy (Table 1). This
complex is unstable in the solid state.
2
µ-SiOMe); 3.18 (t, 2H, J
) 10 Hz, PCH2P). Anal. Calcd for
PH
C
31H31BrFeNiO6P2Si: C, 47.49; H, 3.99. Found: C, 47.27; H,
4.45.
F or m a tion of [(OC)3F e{µ-Si(OMe)2(OMe)}(µ-d p p m )NiI]
(3). To a stirred solution of 1 (0.123 g, 0.17 mmol) in THF (20
mL) was added NaI (0.027 g, 0.18 mmol) at room temperature.
After the reaction mixture was stirred for 1 h, it was filtered,
and the solvent was removed under reduced pressure. The
residue was then washed with 10 mL of hexane and 20 mL of
Et2O, affording a green-brown solid which was dried under
F or m a tion of [(OC)3F e{µ-Si(OMe)2(OMe)}(µ-d p p m )Ni-
C(O)Me] (8). CO was bubbled for a few seconds through a
C6D6 solution of 4 (0.06 g, 0.084 mmol) at room temperature,
until the solution turned deep red. It is then necessary to stop
the CO current to prevent decomposition of the acyl complex
8. 1H NMR (C6D6, 298 K): δ 8.0-6.6 (m, 20H, aromatics); 1.93
(s, 3H, NiC(O)CH3).
1
vacuum. H NMR (C6D6, 298 K): δ 8.2-6.7 (m, 20H, aromat-
ics); 3.68 (s, 6H, Si(OMe)2); 3.54 (s, 3H, µ-SiOMe); 3.28 (t, 2H,
2J
) 10.1 Hz, PCH2P). Owing to some decomposition, this
PH
F or m a tion of [(OC)3F e{Si(OMe)3}(µ-d p p m )Ni(CNtBu )-
{C(Me)dNtBu }] (12). To a stirred solution of 4 (0.103 g, 0.15
complex could not be isolated analytically pure.
t
Syn th esis of [(OC)3F e{µ-Si(OMe)2(OMe)}(µ-d p p m )Ni-
mmol) in CH2Cl2 was added 1 equiv of BuNC (0.0125 g) via
syringe. The orange solution turned red, and the corresponding
complex 12 could only be characterized in situ (see text).
Deh yd r ogen a tive Cou p lin g of P h 3Sn H Ca ta lyzed by
1 a n d [NiCl2(P P h 3)2]. A Schlenk flask equipped with a
stirring bar and a serum cap was charged with Ph3SnH (4.98
g, 14.2 mmol) in 15 mL of Et2O and was placed in a water
bath at 293 K. The volume of H2 released was monitored, using
the following procedure: the Schlenk was fitted onto a gas
buret, and the catalyst (14.2 × 10-4 mmol in 1 mL of Et2O)
was rapidly added to the reaction mixture via syringe through
the serum cap. The turnover numbers were calculated using
the following equation: TON ) (n mol of H2)/(n mol of catalyst),
with the number of moles of H2 being determined by applying
the gas equation: PV ) nRT. Each experiment was repeated
at least two times, and the mean values were used for the
plots. When the reaction was over, after decantation and
filtration the residue was washed with Et2O, affording Ph6-
Me] (4). To a stirred solution of 1 (1.515 g, 2.05 mmol) in THF
(50 mL) at -78 °C was added MeLi (1.6 M) in Et2O (1.34 mL,
2.15 mmol) via a syringe. After completion of the reaction (0.5
h, IR monitoring in the νCO region) the solvent was evaporated,
toluene was added (50 mL), and the mixture was filtered. The
filtrate was concentrated to 10 mL, and CH2Cl2 (10 mL) was
added slowly. The Schlenk flask was placed at -20 °C, and
orange crystals of 4 slowly formed (0.59 g, 40%). 1H NMR
(acetone-d6, 298 K): δ 8.1-6.8 (m, 20H, aromatics); 3.70 (s,
6H, Si(OMe)2); 3.66 (dd, 2H, 2J
) 11.4, 11 Hz, PCH2P); 3.27
PH
(s, 3H, µ-SiOMe); -0.22 (s, 3H, 3J (P-H) ) 3.8 Hz, NiCH3).
Anal. Calcd for C32H34FeNiO6P2SiCH2Cl2: C, 49.29; H, 4.51.
Found: C, 49.64, H, 4.55.
Syn th esis of [(OC)3Fe{µ-Si(OMe)2(OMe)}(µ-dppm )NiP h ]
(5). This complex was synthesized in a method similar to 4
using PhLi (1.8 M in cyclohexane/Et2O). Unfortunately, it could
not be isolated pure, owing to rapid decomposition in the solid
state. However, 5 was unambiguously characterized by spec-
toscopic methods in solution (THF) and by comparison of its
data with those of 4. 1H NMR (C6D6, 298 K): δ 8.1-6.6 (m,
25H, aromatics); 3.47 (s, 6H, Si(OMe)2); 3.33 (br, 2H, PCH2P);
3.10 (s, 3H, µ-SiOMe).
1
Sn2 as a white solid (mp 225-235 °C); the H NMR spectrum
only showed signals in the aromatic region. The turnover
numbers determined from the amount of Ph6Sn2 recovered
were in all cases in good accordance with those determined
from the volumes of H2 released.
F or m a tion of [(OC)3{(MeO)3Si}F e(µ-d p p m )Ni(η3-C3H5)]
(6). To a stirred solution of 1 (0.365 g, 0.49 mmol) in THF at
-78 °C was added a 2 M solution of (allyl)MgCl in THF (0.245
mL, 0.49 mmol) via a syringe. The solution immediately turned
red. After it was stirred for 2 h at -78 °C, the solution was
filtered and concentrated. The 31P{1H} NMR spectrum showed
a 1:1 mixture of 6 and unreacted 1. When more than 2 equiv
of (allyl)MgCl was used, only complex 6 was observed. The
preparation of 6 was also attempted from 1 and (allyl)MgBr.
Under stoichiometric conditions, a 1:1 mixture of 6 and 2 was
obtained, whereas when more than 2 equiv of (allyl)MgBr was
Ack n ow led gm en t. We are grateful to Dr. M. Knorr
for preliminary experiments and discussions and to the
CNRS and the Ministe`re de la Recherche (Paris) for
financial support.
Su p p or tin g In for m a tion Ava ila ble: Tables of atomic
coordinates, anisotropic displacement parameters, and all bond
lengths and angles for 4. This material is available free of
charge via the Internet at http://pubs.acs.org.
OM010619P