Rhodium and palladium complexes from 1,1′ and 1,2 ferrocenylphosphine as bidentate ligands. Versatile coordination

Article abstract of DOI:10.1016/S0022-328X(00)00501-5

The complexation of the mixed bidentate ligands 1-diphenylphosphino-1′-diphenylthiophosphinoferrocenyl and 1,2-bis-(diphenylphosphino)ferrocenyl with rhodium(I) and palladium(II) species yield a range of mono- and dirhodium or palladium complexes. Their interest as possible catalysts for alkene hydroformylation and alkoxycarbonylation and Heck coupling reactions has been assessed. Fe[C₅Me₄P(S)Ph₂][C₅Me ₄PPh₂]PdCl₂ and Fe[C₅H₂-1,2-(PPh₂)₂-4- ^tBu][C₅H₅]PdCl₂ have been characterized by single-crystal X-ray diffraction studies.

Full text of DOI:10.1016/S0022-328X(00)00501-5

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 613 (2000) 77–85
Rhodium and palladium complexes from 1,1% and 1,2
ferrocenylphosphine as bidentate ligands.
Versatile coordination
Roland Broussier a,*, Emmanuelle Bentabet a, Myriam Laly a, Philippe Richard a,
Lyudmila G. Kuz’mina c, Philippe Serp b, Nigel Wheatley b, Philippe Kalck b,
Bernard Gautheron a
a Laboratoire de Synthe`se et d’Electrosynthe`se Organome´talliques (UMR 5632), Uni6ersite´ de Bourgogne, 6, boule6ard Gabriel,
21000 Dijon, France
b Laboratoire de Catalyse et Chimie fine, Ecole Nationale Supe´rieure de Chimie de Toulouse,118, route de Narbonne,
31077 Toulouse cedex, France
c N.S. Kursako6 Institute of General Inorganic Chemistry, RAS, Leninskiiprosp. 31, 117907 Moscow, Russia
Received 16 June 2000; accepted 11 July 2000
Abstract
The complexation of the mixed bidentate ligands 1-diphenylphosphino-1%-diphenylthiophosphinoferrocenyl and 1,2-bis-
(diphenylphosphino)ferrocenyl with rhodium(I) and palladium(II) species yield a range of mono- and dirhodium or palladium
complexes. Their interest as possible catalysts for alkene hydroformylation and alkoxycarbonylation and Heck coupling reactions
has been assessed. Fe[C5Me4P(S)Ph2][C5Me4PPh2]PdCl2 and Fe[C5H2-1,2-(PPh2)2-4-tBu][C5H5]PdCl2 have been characterized by
single-crystal X-ray diffraction studies. © 2000 Elsevier Science B.V. All rights reserved.
Keywords: Ferrocene; Phosphine; Rhodium; Palladium; Catalysis
more sensitive than palladium to the influence of the
ligands.
1. Introduction
We report here the reactions of [RhCl(CO)2]2 with
the metalloligands 1 and 2 (Scheme 1) and of
[Rh(StBu)(CO)2]2 with 2, which supplements the study
already published [1] with 3. These ligands give a range
of di- and polynuclear complexes with Rh(I), whereas
the reaction with PdCl2 led in each case to dinuclear
derivatives.
These compounds have been screened as possible
homogeneous catalysts for carbonylation and CꢀC cou-
pling reactions.
The on-going interest in ferrocene chemistry stems
from an important advantage of the ferrocenyl architec-
ture: for chelating substituents, their coordination abil-
ity may be fine tuned by the choice of the number and
the relative positioning of these substituents and by the
possible presence of other groups.
In the context of our research into rhodium and
palladium compounds of catalytic utility, we were inter-
ested in the effect of chelation on the stoechiometry and
coordination modes in complexes of ferrocenylphosphi-
nes with rhodium(I) and palladium(II).
We have indeed observed differences in the behavior
of the two platinum group metal with rhodium being
* Corresponding author. Tel.: +33-380-396100/03; fax: +33-380-
396100.
E-mail address: roland.broussier@u-bourgogne.fr (R. Broussier).
Scheme 1.
0022-328X/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(00)00501-5

78
R. Broussier et al. / Journal of Organometallic Chemistry 613 (2000) 77–85
2. Results and discussion
2.1. Rhodium complexes
two isomers 9a and 9b, with Cl− as counterion, can
also be obtained from 6, by the addition of an excess of
ligand.
It is known that [Rh(StBu)(CO)2]2 retains its dinu-
clear framework on reaction with diphosphines [3,4]:
the diphosphine acts as a supplementary bridging lig-
and between the two rhodium atoms. In contrast, on
reaction of the ligand 2 with [Rh(StBu)(CO)2]2, we
obtained a novel species 7, in which the diphosphine is
bounded to only one rhodium atom. This can be at-
tributed to the rigidity of the ligand with a short
distance between the phosphorus atoms and the global
planar geometry of the substituted cyclopentadienyl
ring.
2.1.1. Preparation of the complexes
The complexes obtained from the reaction of ligands
1 and 2 with [RhCl(CO)2]2 and of ligand 2 with
[Rh(StBu)(CO)2]2 are shown in Scheme 2.
The reaction of the phosphine–thiophosphine ligand
1 with 0.5 equivalent [RhCl(CO)2]2 (equimolar quanti-
ties of Fe and Rh) in toluene gave two products: the
expected dinuclear species 4 and a trinuclear species 5
in which rhodium(I) centers are linked by a single
m-chloro bridge. Compound 5 is neutral, as opposed to
the cationic tetranuclear species formed in the
analogous reaction with the ring-permethylated ana-
logue of 1 [1]. In the present case, the less sterically
hindered ligand shows fewer propensities to substitute
the carbonyl ligands, indicating the importance of the
electronic effects of ring substitution in these systems.
When the reaction is carried out in dichloromethane,
only the 1:1 complex 4 is produced. In order to obtain
5 in a pure form, it was necessary to use an excess of
[RhCl(CO)2]2.
The reaction of 2 with [RhCl(CO)2]2 gave a more
complex reaction mixture and three products 6, 8a and
8b were characterized. The formation of a cationic
rhodium complex 8 stabilized by four phosphorus
atoms has also been observed with dppf [2], but here
the disymmetrical substitution of the ferrocene leads to
cis–trans isomerism. The 31P-NMR spectrum indicates
an isomer ratio of 7:3, but we have been unable to
separate them, and consequently we did not try to
assign the signals to one or the other. Working under a
carbon monoxide atmosphere allows 6 to be cleanly
synthesized separately from 8a and 8b. The isomer
mixture 8a and 8b, obtained together with 6 in toluene,
is easily isolated as a precipitate associated with the
counterion [RhCl2(CO)2]− (wCO: 1990, 2068 cm−1). The
With regard to the use of the spectroscopic methods
to establish the molecular structures, the form of the
31P-NMR spectrum for 6 and 7 and the presence of
several absorption bands in the infrared spectrum of 5
should be mentioned. For 6 and 7 the two phosphorus
atoms are nonequivalent and the 31P signal (dd) results
from mutual coupling (2JPꢀP) and vicinal coupling
(JPꢀRh) with the rhodium atom. The IR absorptions in
the CO region for compound 5 (1992, 2003 and 2078
cm−1) are very similar to those of (CO)2ClRh(m-
Cl)Rh(CO){C6H5P[OC(CH3)2CH2]2NH} [5] (1998, 2008
and 2074 cm−1), which exhibits the same geometry and
coordination as 5 Table 1.
Table 1
NMR and IR data for 4, 5, 6, 7, 8a and 8b.
Compound
NMR 31P l, ppm (J, Hz)
IR wCO, cm−1
4
5
19.1 (142) and 39.7
36.6 (136) and 44.8
1971
1992, 2003,
2078
6
31.0 dd (108/36) and 54.0 dd
(127/36)
2011
7
31.0 dd (130/40) and 44.0 dd
(126/40)
1977
8a,b
37.0 or 39.0 (135)
Scheme 2.

R. Broussier et al. / Journal of Organometallic Chemistry 613 (2000) 77–85
79
Table 2
Hydroformylation of oct-1-ene by complexes 4, 13, 6 and 7 a
Precursor
t (h)
Pressure (bar)
P(OPh)3/Rh
Yield in
Selectivities
aldehyde (%)
Nonanal 2-Methyl octanal 2-Ethyl heptanal 2-Propyl hexanal
4 b
4 b
13 b
13 b
6 c
16
16
16
16
16
16
4
10
10
10
10
10
50
50
10
0
6
0
6
0
0
0
0
0
0
0
61
34
96
86.9
94
84.5
69
45.7
49.5
52
15.5
30.5
35.5
35.7
48
0.5
10.6
9.1
6 b
6 b
7 c
8.2
5.7
16
a Temperature: 80°C.
b Alkene/catalyst=200.
c Alkene/catalyst=215.
2.1.2. Catalytic hydroformylation
plex could activate the internal double bond of a-
pinene under the relatively mild conditions employed
(20 bar, 85°C, 18 h). However, both showed a certain
activity for the hydroformylation of b-pinene, with
10-formylpinane yields of 2.5% (6) and 11% (7) and
diastereoisomeric excesses of 56% (6) and 48% (7) for
the cis diastereoisomer. Isomerization was a competing
reaction (5% yield of a-pinene) in the case of 7.
Table 2 shows the results of the hydroformylation of
oct-1-ene catalyzed by compounds 4, 6, 7 and 13
(Fe[C5Me4P(S)Ph2][C5Me4PPh2]RhCl(CO) [1]). At 10
bar, the presence of a sulfur atom on one phosphorus
atom seems to inhibit the reaction. However, the addi-
tion of six molar equivalents of P(OPh)3 to complex 13,
but not to complex 4, leads to a 61% yield of aldehydes.
The difference of reactivity may come from the lower
basicity of the metal center for 13 (wCO=1987 cm−1
)
2.2. Palladium coordination
than for 4 (1971 cm−1), resulting from the electronic
influence of the methyl substituents on the cyclopenta-
diene rings. An intermediate species of general formula
RhH(CO)(P,PS) is expected.
2.2.1. Preparation of the complexes
The monomeric palladium complexes L2PdCl2 10, 11
and 12 have been obtained (Scheme 3) by reaction of 1,
When using ligand 2 (complexes 6 and 7), a different
reactivity is observed. Reaction occurs without the ad-
dition of auxiliary P(OPh)3, with a yield in aldehydes of
34 and 94% for 6 and 7, respectively. At the end of
these experiments, catalytic solutions were clear (no
rhodium precipitation) and the complexes were recov-
ered without transformation. The differences in reactiv-
ity between 6 and 7 could stem from the ease of
production of an active hydride species. This step
should be enhanced on 7, which presents a more basic
metal center. The unexpected formation of small
amounts of 2-ethylheptanal using precursor 6 prompted
us to repeat the experiment at a higher pressure (50
bar): under these conditions, all the isomeric aldehydes
(nonanal/2-methyloctanal/2-ethylheptanal/2-propylhex-
anal) were detected, in a ratio of: 46/36/10/8. This
unusual activity for the hydroformylation of internal
olefins was confirmed by the use of cyclohexene as
substrate: under the same conditions a 44% yield of
cyclohexanecarboxaldehyde was obtained.
2
and 3, respectively with either PdCl2 and
(PhCN)2PdCl2. The route appears to be specific and the
formation of mixtures of products is not observed
unlike in the case of rhodium. Complexation produces
a large downfield shift of the 31P(III)-NMR resonance
compared to that of the starting ligand (1 l−19.6, s
and 10 l+12.6, s; 2 l−24.3, s and 11 l+42.0 s; 3
l−24.9 s and 12 l+28.0 s).
2.2.2. X-ray structures of 11 and 12
Single-crystals of 11 and 12 which were suitable for
diffraction analysis were obtained by recrystallization in
CH2Cl2.
The solid-state structures of the compounds are
shown in Figs. 1 and 2, respectively.
As part of an on-going interest in the catalytic trans-
formation of natural products, we also examined the
activity of 6 and 7 for the hydroformylation of two
monoterpenes, namely a- and b-pinene. Neither com-
Scheme 3.

80
R. Broussier et al. / Journal of Organometallic Chemistry 613 (2000) 77–85
Fig. 1. ORTEP view [16] of 11 drawn at the 30% probability level. Hydrogen atoms and the CH2Cl2 solvate molecule are omitted for clarity.
Pertinent bond distances and angles are listed in
Table 3. The parameters describing the geometrical
environment of the palladium atom are compared in
Table 4 with those of complexes of similar chelating
ligands: dppf [6,7] and (R)(S)BPPFA [8]. Data collec-
tion and refinement parameters are given in Table 5.
For 11 the square planar coordination of the Pd
atom is very slightly distorted. The chelated PꢀPdꢀP
angle of 88.97(4)° leads to an opening of the ClꢀPdꢀCl
angle to 92.53(5)°. The phosphorus atoms are displaced
,
from the plane of the Cp ring by 0.25 A for P(1) and
,
0.17 A for P(2) (exo-type) and the coordination plane
of the palladium is tilted from the plane of the Cp ring
by 11.8(2)°. The ferrocenyl part shows a usual tilt of Cp
planes (3.1°) and an eclipsed conformation (twist angle
3.0°), whereas it is clearly staggered for the free ligand
2 (twist angle: 27.0° [11]). As one would expect, the
Fe-centroid (CNT) distance is shorter for the substi-
,
tuted ring (1.652 A) than for the non-substituted (1.673
,
A) [11].
The square planar geometry in molecule 12 is signifi-
cantly distorted by the difference in the trans influences
of the phosphorous center and the phosphine sulfide
center, with the PdꢀCl bond trans to phosphorous
,
being 0.05 A longer than that trans to sulfur. The
length and the zigzag geometry of the [4]-ferro-
cenophane bridge involves the shortest bite angle com-
Fig. 2. ORTEP view [16] of 12 drawn at the 30% probability level.
Hydrogen atoms and the CH2Cl2 solvate molecule are omitted for
clarity.

R. Broussier et al. / Journal of Organometallic Chemistry 613 (2000) 77–85
81
Table 3
Bond lengths (A) and angles (°) for 11 and 12.
The methoxycarbonylation of oct-1-ene gave disap-
pointing results. Under the usual experimental condi-
tions (95°C, 40 bar) in the presence of 2.5 equivalents of
SnCl2 per palladium [9] only 10 presents any activity
(esters 8%: methyl nonanoate 84%, methyl2-methyloc-
tanoate 16%). Further studies are continuing focussing
on the nature of hydride complexes formed under these
conditions.
Boyes and co-workers [10] reported that certain pal-
ladium complexes of ferrocenyl diphosphine ligands,
particularly PdCl2(diisoppf) (diisoppf=1,1%-bis(diiso-
propylphosphino)ferrocene), were active catalysts for
the Heck coupling reaction. We tested compounds 10–
12 under the same conditions (reaction of phenyl iodide
and methyl acrylate to give methyl trans-cinnamate),
and found very promising activity for 10 (yield 90%)
and 12 (yield 99%; c.f. PdCl2(diisoppf), 96%). The poor
yield (5%) obtained with 11 is comparable to that
observed with PdCl2(dppf) (7%). These results seem to
indicate that the reaction is primarily dependent upon
the electronic effects given by the ligand and can be
viewed as the result of the hemilabilisation of diisoppf
as well as that of 10 and 12.
,
11
12
FeꢀCNT(1)
FeꢀCNT(2)
PdꢀP(1)
1.652
1.673
FeꢀCNT(1,5)
FeꢀCNT(6,9)
PdꢀS
1.675
1.683
2.2457(11)
2.2405(12)
2.3518(13)
2.3440(12)
1.800(4)
1.797(4)
2.3278(7)
2.2445(7)
2.3696(7)
2.3160(7)
1.795(3)
1.815(3)
2.0237(9)
PdꢀP(2)
PdꢀP(2)
PdꢀCl(1)
PdꢀCl(2)
P(1)ꢀC(4)
P(2)ꢀC(3)
PdꢀCl(1)
PdꢀCl(2)
P(1)ꢀC(10)
P(2)ꢀC(1)
P(1)ꢀS
CNT(1)ꢀFeꢀCNT(2) 176.4
CNTꢀFeꢀCNT 173.5
SꢀPdꢀP(2)
Cl(1)ꢀPdꢀCl(2)
SꢀPdꢀCl(1)
SꢀPdꢀCl(2)
P(2)ꢀPdꢀCl(1)
P(2)ꢀPdꢀCl(2)
C(1)ꢀP(2)ꢀPd
C(10)ꢀP(1)ꢀS
P(1)ꢀSꢀPd
P(1)ꢀPdꢀP(2)
Cl(1)ꢀPdꢀCl(2)
P(1)ꢀPdꢀCl(1)
P(1)ꢀPdꢀCl(2)
P(2)ꢀPdꢀCl(1)
P(2)ꢀPdꢀCl(2)
C(4)ꢀP(1)ꢀPd
C(3)ꢀP(2)ꢀPd
88.97(4)
85.75(2)
89.41(3)
92.80(3)
176.32(2)
176.29(3)
92.22(2)
114.40(9)
118.08(9)
107.13(3)
92.53(5)
90.64(5)
174.88(6)
179.41(5)
87.84(5)
107.1(2)
107.0(2)
pared to those reported for [3]-ferrocenophanes (Table
4). In the [4]-ferrocenophane 12, the Cp rings are in a
staggered conformation as in the [3]-ferrocenophane
3. Experimental
,
(12: twist angle 32.4°; P···P distance 4.361 A and
,
PꢀPdꢀSꢀP length 6.596 A compared to dppf, CHCl3 [6]:
,
,
P···P 3.487 A and PꢀPdꢀP 4.564 A). At the same time,
the two Cp rings are nearly parallel (tilt angle 4.6°) and
the P atoms are displaced from the plane of the corre-
3.1. General considerations
All manipulations were performed under Argon us-
ing standard Schlenk tube techniques. Mass spectra
(electronic ionization 70 eV) were recorded on a Kratos
concept IS machine. 1H- and 31P{1H}-NMR spectra
were recorded on a Bruker AC 200 spectrometer.
The starting compounds 2 [11] and 3 [1] were pre-
pared according to the literature.
,
sponding ring, away from the Fe atom by 0.265 A
,
(P(2)) and 0.175 A (P(1)).
2.2.3. Catalytic acti6ity in methoxycarbonylation and
Heck reactions
Among the numerous palladium-catalyzed CꢀC bond
formation reactions, we investigated two examples with
different potential applications: methoxycarbonylation
which transforms an alkene into a methyl ester in the
presence of carbon monoxide and an alcohol, and the
Heck reaction, in which an alkene is substituted.
3.2. Preparation of 1
A suspension of dppf (6.34 g, 11.4 mmol) in 120 ml
toluene was added, at 0°C, to a suspension of sulfur
(0.36 g, 11.4 mmol) in 30 ml toluene. The mixture was
Table 4
Comparison between X-ray parameters of 11, 12, dppf and (R)(S)BPPFA.
,
Ligand bite angle (°)
ClꢀPdꢀCl (°)
PdꢀP (ou PdꢀS)
PdꢀCl (A)
PdCl2[(R)(S)BPPFA] a
98.79(4)
99.07(5)
97.98(4)
88.97(4)
87.83(4)
87.8(1)
89.96(4)
92.53(5)
2.302(1) 2.296(1)
2.283(1) 2.301(1)
2.278(1) 2.289(1)
2.2457(11)
2.351(1) 2.334(1)
2.347(1) 2.348(1)
2.340(1) 2.358(1)
2.3518(13)
b
PdCl2(dppf),CHCl3
PdCl2(dppf),CH2Cl2
b
11
2.2405(12)
2.3440(12)
12
85.75(2)
89.41(3)
PdꢀP 2.2445(7)
PdꢀS 2.3278(7)
2.3160(7)
2.3696(7)
a (R)(S)BPPFA=(R)-N,N-dimethyl-1-[(S)-1%,2-bis-(diphenylphosphino)ferrocenyl]ethylamine.
b dppf=1,1%-bis-(diphenylphosphino)ferrocene.

82
R. Broussier et al. / Journal of Organometallic Chemistry 613 (2000) 77–85
Table 5
Three successive orange fractions were obtained: the
first one contained dppf, the second contained the
expected product and the last one the dithiophosphino
compound.
Crystal data and structure refinement for 11 and 12
Compound
11
12
Formula
C39H38Cl4FeP2Pd
872.68
293(2)
Orthorhombic
P212121
12.065(2)
17.250(2)
18.230(2)
90
90
90
3794.1(9)
4
1768
C
43H46Cl4FeP2PdS
The solvent was removed from the second fraction
and gave 1 as a yellow powder (1 g, 1.71 mmol,
yield=15%), m.p. 183°C. Anal. Calc. for C34H28FeP2S:
C, 69.66; H, 4.77; S 5.47. Found: C, 69.78; H, 4.76; S
Formula weight
Temperature (K)
Crystal system
Space group
960.85
150.0(2)
Triclinic
(
P1
1
5.25%. H-NMR (CDCl3): l 4.01 (m, 2H, Cp), 4.35 (m,
,
a (A)
11.0490(1)
11.5799(1)
18.4383(2)
97.379(1)
104.510(1)
108.720(1)
2106.25(3)
2
980
1.515
Siemens SMART
ꢀ
,
b (A)
2H, Cp), 4.38 (m, 2H, Cp), 4.49 (m, 2H, Cp), 7.39 (m,
10 H, Ph), 7.65 (m, 10 H, Ph). 31P{1H}-NMR (CDCl3):
l −19.4 (s), 39.2 (s). EI-MS (200°C), m/z (%): 586
(100) [M+], 554 (96) [M+−S], 509 (20) [M+−Ph], 337
(85) [M+−C5H4PPh2].
,
c (A)
h (°)
i (°)
k (°)
3
,
V (A )
Z
Complex
analogous to that used to prepare 3 [1], using diphenyl-
phosphinocyclopentadienyllithium and diphenyl-
1 can be obtained by a procedure
F(000)
Dcalc (g cm−3
)
1.528
Nonius CAD-4
ꢀ
Diffractometer
Scan type
thiophosphinocyclopentadienyllithium. Yield=8%.
,
u (A)
0.71073
1.250
0.50×0.41×0.32
0.62
−155h50,
−215k50,
05l522
0.71073
1.181
v (mm−1
)
3.3. Preparation of 4
Crystal size (mm3)
sin(q)/umax (A
Index ranges
0.20×0.18×0.12
0.65
−125h514,
−155k512,
−235l523
Empirical
−1
,
)
A solution of 1 (77 mg, 0.13 mmol) and [RhCl(CO)2]2
(25 mg, 0.065 mmol) in 10 ml dichloromethane was
stirred, at r.t. for 2 h. The solvent was removed under
reduced pressure and the resulting oil was washed with
10 ml hexane to give an orange powder. The yield was
Absorption correction Psi-scan
(
SHELTXTL-PLUS)
Max transmission
Min transmission
Reflections collected
0.993
0.954
4285
0.783
0.628
15 555
essentially
quantitative.
Anal.
Calc.
for
C35H28ClFeOP2RhS: C, 55.86; H, 3.72; S 4.26. Found:
C, 56.27; H, 3.97; S 4.09%. H-NMR (CDCl3): l 4.30
1
Independent reflections 4284
9570 [Rint=0.0314]
Reflections
observed[I\2|(I)]
Refinement method
3807
8036
(m, 2 H, Cp), 4.40 (m, 2 H, Cp), 4.57 (m, 2 H, Cp), 4.83
(m, 2 H, Cp), 7.16–7.69 (m, 20 H, Ph). 31P{1H}-NMR
(CDCl3): l 19.1 (d, 142 Hz), 39.7 (s). MS (FAB), m/z
(%): 716 (95) [M+−HCl], 688 (100) [M+−HCl−
CO], 656 (50) [M+−HCl−CO−S]. IR (CH2Cl2,
cm−1): w(CO) 1971.
Full-matrix
least-squares
on F2
Full-matrix
least-squares
on F2
Data/restraints/
parameters
R for IRCGT
4284/0/425
8977/0/654
R1 a=0.0297
wR2 b=0.0771
R1 a=0.0412
wR2 b=0.0826
1.034
R1 a=0.0342
wR2 b=0.0767
R1 a=0.0483
wR2 b=0.0957
1.032
3.4. Preparation of 5
R for IRC
Goodness-of-fit c
Absolute structure
parameters
A
mixture of 1 (146 mg, 0.25 mmol) and
0.06(3)
[RhCl(CO)2]2 (97 mg, 0.25 mmol ) in 10 ml toluene was
stirred, at r.t. for 1 h. The solvent was removed under
reduced pressure and the resulting oil was washed twice
with 10 ml hexane to give an orange powder (200 mg,
0.19 mmol, yield=76%), m.p. 146°C (dec.). Anal. Calc.
for C37H28Cl2FeO3P2Rh2S: C, 46.93; H, 2.96; S, 3.38.
Largest difference
0.618 and −0.422 0.636 and −0.608
peak and hole
−3
,
(e A
)
a R1=S(ꢀFoꢀ−ꢀFcꢀ)/SꢀFoꢀ.
b w=1/[|2(F2o)+(0.047P)2+1.95P] for 11 where P=(max(F2o,0)+
2*F2c)/3 and wR2={S[w(Fo2−Fc2)2]/S[w(F2o)2}1/2 where w=1/
[|2(F2o)+(0.030P)2+3.01P] for 12.
1
Found C, 46.66; H, 2.93; S, 3.64%. H-NMR (CDCl3):
l 4.13 (m, 2 H, Cp), 4.47 (m, 2 H, Cp), 4.61 (m, 2 H,
Cp), 4.93 (m, 2 H, Cp), 7.25–7.70 (m, 20 H, Ph).
31P{1H}-NMR (CDCl3): l 36.6 (d, 136 Hz), 44.8 (s).
MS (FAB): fragmentation as 4. IR (CH2Cl2 cm−1):
w(CO) 2078, 2003, 1992.
c Goodness-of-fit={S[w(F2o−F2c)2]/(No−Nv)}1/2
.
allowed to warm to room temperature (r.t.) and stirred
for 18 h. The mixture was filtered and the filtrate
evaporated under reduce pressure. The resulting brown
residue was chromatographed on a silica column with
3:1 toluene–hexane as eluent.
3.5. Preparation of 6
To 2(300 mg, 0.49 mmol) in 20 ml toluene under CO
were added, at r.t. [RhCl(CO)2]2 (95.5 mg, 0.24 mmol)

R. Broussier et al. / Journal of Organometallic Chemistry 613 (2000) 77–85
83
in 10 ml toluene. After stirring under CO for 45 min the
solvent was removed under reduced pressure to give a
yellow solid, identified as 6. The yield was essentially
quantitative, m.p. 206°C (dec.). Anal. Calc. for
C39H36ClFeOP2Rh: C, 60.29; H 4.68. Found: C, 60.05;
and 2 in toluene. Anal. Calc. for C76H72ClFe2P4Rh: C,
68.94; H, 5.49. Found: C, 68.80; H, 5.47%.
3.9. Preparation of 10
1
t
H 4.66%. H-NMR (CDCl3): l 1.27 (s, 9H, Bu), 3.37
(s, 5H, Cp), 4.35 (s, 1H, Cp), 4,41 (s, 1H, Cp), 7–8.2
(m, 20H, Ph). 31P{1H}-NMR(CDCl3):d 54 (dd,
PꢀRhꢀCl, 2J(PꢀP)=36 Hz, 1J(PꢀRh)=164 Hz), 31 (dd,
PꢀRhꢀCO, 2J(PꢀP)=36 Hz, 1J(PꢀRh)=127 Hz). MS
(FAB), m/z (%): 777 (10) [M+], 747 (60) [(M+−CO)],
712 (70) [(M+−COCl)]. IR (CH2Cl2 cm−1): wCO 2011.
A suspension of 1 (116 mg, 0.19 mmol) and PdCl2
(35 mg, 0.19 mmol) in 15 ml toluene was stirred, at r.t.
for 72 h. The resulting solution was evaporated to
dryness giving 10 as an orange–yellow solid (130 mg,
0.17 mmol, yield=89%), m.p. 201°C. Anal. Calc. for
C34H28Cl2FeP2PdS: C, 53.48; H, 3.67; S, 4.20. Found:
1
C, 53.6; H, 3.66; S, 4.03%. H-NMR (CDCl3): l 4.36
(m, 2 H, Cp), 4.49 (m, 2 H, Cp), 4.59 (m, 2 H, Cp), 4.98
(m, 2 H, Cp), 7.12–7.68 (m, 20 H, Ph). 31P{1H}-NMR
(CDCl3): l 12.6 (s), 38.9 (s). MS (FAB), m/z (%): 726
(90) [M+−HCl], 691 (100) [M+−HCl2], 613 (90) [M+
−HCl2−C6H6].
3.6. Preparation of 7
To 2 (516 mg, 0.84 mmol ) in 15 ml toluene were
added, at r.t. [Rh(mStBu)(CO)2]2 (209 mg, 0.42 mmol) in
5 ml toluene. The solvent was immediately removed
under reduced pressure and the residue washed with 15
ml hexane to give an orange powder (273 mg, 0.33
mmol, yield=78%), m.p. 206°C. Anal. Calc. for
C43H45FeOP2RhS: C, 62.20; H 5.47; S, 3.86. Found: C,
3.10. Preparation of 11
A
mixture of 2 (341 mg, 0.56 mmol) and
(PhCN)2PdCl2 (214 mg, 0.56 mmol) in 20 ml
dichloromethane was stirred at r.t. for 48 h. The result-
ing solution was evaporated to dryness giving 11 as a
yellow–brown solid (411 mg, 0.52 mmol, yield=93%),
m.p.\260°C. Anal. Calc. for C38H36FeP2PdCl2: C,
1
61.78; H, 5.48; S, 3.69%. H-NMR (CDCl3): l 1.25 (s,
t
9H, Bu), 1.46 (s, 9H, StBu), 3.31 (s, 5H, Cp), 4.29 (s,
1H, CH), 4.35 (s, 1H, CH), 7–8, 11 (m, 20H, Ph).
31P{1H}-NMR(CDCl3): l 44 (dd, PꢀRhꢀStBu, J(PꢀP)
=
=
2
1
2
1
40 Hz, J(PꢀRh)=126 Hz), 31 (dd, PꢀRhꢀCO, J(PꢀP)
57.93; H, 4.62. Found: C, 57.89; H, 4.60%. H-NMR
1
t
40 Hz, J(PꢀRh)=130 Hz). MS (FAB), m/z (%): 830 (70)
[M+], 802 (50) [(M+−CO)], 741 (40) [(M+−StBu)],
712 (100) [(M+−StBuCO)]. IR (CH2Cl2 cm−1): wCO
1977.
(CDCl3): l 1.27 (s, 9H, Bu), 3.40 (s, 5H, Cp), 4,41 (s,
2H, CH), 7,17–8,17 (m, 20H, Ph). 31P{1H}-
NMR(CDCl3):l 42,0 (s). EI–MS (200°C), m/z (%): 787
(15) [M+], 751 (100) [(M+−Cl)], 716 (20) [(M+−
2Cl)].
3.7. Preparation of 8a and 8b
3.11. Preparation of 12
To 2 (180 mg, 0.295 mmol) in 20 ml toluene were
added, at r.t. [RhCl(CO)2]2 (57.8 mg, 0.148 mmol ) in
10 ml toluene. The resulting precipitate (the filtrate
contained 6) was immediately isolated to give 8a, and
8b as an orange powder (108 mg, 0.088 mmol, yield=
59%). Anal. Calc. for C78H72Cl2Fe2O2P4Rh2: C, 60.29;
H, 4.63. Found: C, 60.36; H, 4.80%. 1H-NMR (CDCl3):
A suspension of 3 (800 mg, 15 mmol) and PdCl2 (200
mg, 1.15 mmol) in 30 ml toluene was stirred, at r.t. for
1 week. The resulting solution was evaporated to dry-
ness giving 12 as a brown solid (870 mg, 0.99 mmol,
yield=86%), m.p. 190°C (dec.). Anal. Calc. for
C42H44CH2Cl2Cl2FeP2PdS: C, 53.76; H, 4.79. Found C,
t
1
l 1.13 (s, 9H, Bu), 3.18 (s, 5H, Cp), 4.11 (s, 2H, Cp),
53.5; H, 4.79%. H-NMR (CDCl3): l 1.76 (s, 12 H,
7.0–7.5 (m, 20H, Ph). 31P{1H}-NMR(CDCl3):l 37.0 (d,
1J(PꢀRh)=135 Hz), 39.0 (d, 1J(PꢀRh)=135 Hz). MS
(FAB), m/z (%): 1324 [M+]. IR (CH2Cl2 cm−1): wCO
1990 and 2068.
Me), 1.80 (s, 6 H, Me), 1.85 (s, 6 H, Me), 7.20–7.64 (m,
20 H, Ph). 31P{1H}-NMR (CDCl3): l 28.0 (s), 40.9 (s).
MS (FAB), m/z (%): 838 (100) [M+−HCl], 803 (80)
[M+−HCl2].
3.8. Preparation of 9a and 9b
3.12. Hydroformylation of oct-1-ene and h- or i-pinene
To 2 (180 mg, 0.295 mmol) in 20 ml toluene were
added, at r.t. [RhCl(CO)2]2 (30 mg, 0.077 mmol) in 10
ml toluene. After stirring at r.t. for 30 min, the solvent
was removed under reduced pressure to give 9a and 9b
as an orange solid. The yield was essentially quantita-
tive. The same mixture can be obtained by mixing 6
The hydroformylation of these alkenes was per-
formed in a magnetically stirred 100 ml stainless steel
autoclave with magnetic stirring, charged at constant
pressure with an equimolecular mixture of H2 and CO
from a tank. The autoclave was charged with the
complex and evacuated. The liquid mixture contained

84
R. Broussier et al. / Journal of Organometallic Chemistry 613 (2000) 77–85
in a Schlenk tube (toluene, alkene, excess of ligand
when necessary) was introduced by aspiration. The
pressure was kept at 1 bar with nitrogen while the
autoclave was heated to 80°C. The H2:CO gas mixture
was then admitted to the desired pressure. After the
desired reaction time (usually 16–18 h), the stirring and
the heating were stopped and once the temperature had
returned to near 25°C the pressure was slowly released.
The reaction mixture was transferred under nitrogen in
a Schlenk tube for analyses by GC. For oct-1-ene,
anisole was used as internal standard.
4.2. Structural analysis of 12
Crystals for the X-ray structure analysis were grown
from a saturated CH2Cl2 solution of 12 at r.t. A
brown–red block of 0.20×0.18×0.12 mm [14] was
selected for X-ray. A total of 15 555 reflections were
collected at 153 K on a Siemens SMART CCD diffrac-
tometer. An empirical (SHELXTL-PLUS) absorption cor-
rection was applied. The structure was solved by direct
methods [14] and refined by full-matrix least squares
methods [14] based on 8977 unique ꢀF2ꢀ. All non-hydro-
gen atoms were refined with anisotropic thermal
parameters. Hydrogen atoms were found from differ-
ence Fourier synthesis and refined with isotropic ther-
mal parameters. At the end of this refinement the
agreement indices were wR2=0.0957 for all data and
R1=0.0342 for 8036 intensities with I\2|(I). The
final difference electron density is: Dz=0.636 and −
3.13. Methoxycarbonylation of oct-1-ene
The procedure was similar to the above, except that
a 200 ml autoclave with mechanical stirring was used,
at a pressure of 40 bar of CO and a temperature of
95°C. A typical reaction mixture contained 0.15 mmol
PdCl2L2, 0.375 mmol SnCl2, 15 mmol oct-1-ene, 30
mmol methanol, 30 ml toluene, and 1.2 ml dodecane
(internal standard).
0.608 e A−3. Crystal data are reported in Table 5.
,
5. Supplementary material
3.14. Reaction of phenyliodide with methylacrylate
under Heck coupling conditions
Crystallographic data (excluding structural factors)
for the structures has been deposited with the Cam-
bridge Crystallographic Data Center, CCDC no.
142833 for compound 11 and CCDC no. 142834 for
compound 12. Copies of the data can be obtained free
of charge from The Director, CCDC, 12 Union Road,
Cambridge CB2 1EZ, UK (Fax: +44-1223-336033;
e-mail: deposit@ccdc.cam.ac.uk or www: http://www.
ccdc.cam.ac.uk).
The experimental procedure reported by Boyes and
co-workers [10] was followed.
4. X-ray crystallography
4.1. Structural analysis of 11
Crystals for the X-ray structure analysis were grown
from a saturated CH2Cl2 solution of 11 at r.t. A red
prism of 0.50×0.41×0.32 mm3 was selected for X-ray
analysis. A total of 25 reflections was used for an
accurate orthorhombic cell determination. Three reflec-
tions and their equivalents were selected to check the
Acknowledgements
We thank S. Gourier for technical assistance. L.G.
Kuz’mina acknowledges Judith A.K. Howard, Univer-
sity of Durham, UK, for the structure solving of 12 in
her laboratory.
Laue group. A total of 4285 reflections were collected
−1
,
at r.t. up to sin(q)/u=0.623 A
on an Enraf–Nonius
CAD-4 diffractometer. The data were corrected for
Lorentz and polarization effects [12] and for absorption
(psi-scan method) [13]. No decay was observed. The
structure was solved via a Patterson search program
[14] and refined in the polar space group P212121 with
full-matrix least squares methods [14] based on ꢀF2ꢀ. All
non-hydrogen atoms were refined with anisotropic ther-
mal parameters. Hydrogen atoms were included in their
calculated positions and refined with a riding model. At
the end of this refinement the agreement indices were
wR2=0.0826 for all data and R1=0.0297 for 3807
intensities with I\2|(I), the absolute structure
References
[1] R. Broussier, M. Laly, P. Perron, B. Gautheron, I.E. Nifant’ev,
J.A.K. Howard, L.G. Kuz’mina, Ph. Kalck, J. Organomet.
Chem. 587 (1999) 104.
[2] U. Castellato, B. Corain, R. Graziani, B. Longato, G. Pilloni,
Inorg. Chem. 29 (1990) 1193.
[3] R. Choukroun, D. Gervais, J. Jaud, Ph. Kalck, F. Senocq,
Organometallics 5 (1986) 67.
[4] Ph. Kalck, C. Randrianalimana, M. Ridmy, A. Thorez, H.T.
Dieck, J. Ehlers, New J. Chem. 12 (1988) 679.
[5] J. Wachter, F. Jeanneaux, G. Le Borgne, J.G. Riess,
Organometallics 3 (1984) 1034.
[6] T. Hayashi, M. Konishi, Y. Kobori, M. Kumada, T. Higuchi, K.
Hirotsu, J. Am. Chem. Soc. 106 (1984) 158.
parameter [15] was x=0.06(3). The final difference
−3
,
electron density is: Dz=0.618 and −0.422 e A
.
Crystal data are reported in Table 5.

R. Broussier et al. / Journal of Organometallic Chemistry 613 (2000) 77–85
85
[7] I.A. Butler, W.R. Cullen, T.J. Kim, S.J. Rettig, J. Trotter,
Organometallics 4 (1985) 972.
[8] T. Hayashi, M. Kumada, T. Higuchi, K. Hirotsu, J. Organomet.
Chem. 334 (1987) 195.
[12] K. Harms, University of Marburg, 1996.
[13] (a) A.C.T. North, D.C. Phillips, F.S. Mathews, Acta Crystallogr.
Sect. A 24 (1968) 351. (b) A.L. Spek, Acta Crystallogr. Sect. A
46 (1990) C34.
[9] T. Chenal, R. Naigre, I. Cipres, Ph. Kalck, J.C. Daran, J.
Vaissermann, J. Chem. Soc. Chem. Commun. (1993) 747.
[10] A.L. Boyes, I.R. Butler, S.C. Quayle, Tetrahedron Lett. 39
(1998) 7763.
[11] R. Broussier, E. Bentabet, P. Mellet, O. Blacque, P. Boyer,
M.M. Kubicki, B. Gautheron, J. Organomet. Chem. 598 (2000)
365.
[14] G.M. Sheldrick, SHELXS-97 and SHELXL-97, University of Go¨t-
tingen, Germany, 1997.
[15] H.D. Flack, Acta Crystallogr. Sect. A 39 (1983) 876.
[16] (a) M.N. Burnett, C.K. Johnson, ORTEP-III, Oak Ridge Thermal
Ellipsoid Plot Program for Crystal Structure Illustrations, Oak
Ridge National Laboratory Report ORNL-6895, 1996. (b) L.J.
Farrugia, J. Appl. Crystallogr. 30 (1997) 565.
.