Rhodium and ruthenium complexes of 1,1′-bis(phosphetano)ferrocenes: Structural characterisation and catalytic behaviour

Article abstract of DOI:10.1002/ejic.200300051

The first structural characterisation of ruthenium and rhodium complexes of the 1,1′-bis(phosphetano)ferrocenes 1 (FerroTANEs) - namely [{(S,S)-iPr-1}RuCl₂Py₂] and [{(S,S)iPr-1}Rh(COD)OTf] - is reported, X-ray data show that the ruthenium complex and the rhodium chelate complex both adopt δ conformations, as a result of the (S,S) configuration of the phosphetane moiety. The ruthenium complexes promote the catalytic hydrogenation of β-keto esters with moderate to high enantioselectivities. The behaviour of the rhodium complexes of 1 is compared with that of other phosphetane-based catalysts in the hydrogenation of methyl N-acetamidocinnamate. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejic.200300051

FULL PAPER
Rhodium and Ruthenium Complexes of 1,1Ј-Bis(phosphetano)ferrocenes:
Structural Characterisation and Catalytic Behaviour
[a]
[a]
[a]
[a]
[b]
´ ´
´
Angela Marinetti,* Francis Labrue, Benedicte Pons, Sebastien Jus, Louis Ricard,
[a]
ˆ
and Jean-Pierre Genet
Keywords: Asymmetric catalysis / Hydrogenations / P ligands / Phosphetane / Rhodium / Ruthenium
The first structural characterisation of ruthenium and
rhodium complexes of the 1,1Ј-bis(phosphetano)ferrocenes 1
(FerroTANEs) − namely [{(S,S)-iPr-1}RuCl2Py2] and [{(S,S)-
mote the catalytic hydrogenation of β-keto esters with mod-
erate to high enantioselectivities. The behaviour of the
rhodium complexes of 1 is compared with that of other phos-
iPr-1}Rh(COD)OTf] − is reported. X-ray data show that the phetane-based catalysts in the hydrogenation of methyl
ruthenium complex and the rhodium chelate complex both
adopt δ conformations, as a result of the (S,S) configuration
of the phosphetane moiety. The ruthenium complexes pro-
N-acetamidocinnamate.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
The 1,1Ј-bis(phosphetano)ferrocenes 1 (FerroTANEs)
rank among those ligands that induce excellent enantiose-
lectivities in rhodium-promoted hydrogenations.[1] In par-
ticular, their practical utility has been demonstrated in the
synthesis of amido succinates by asymmetric hydrogenation
of itaconate derivatives.[2] In these reactions, ligands 1 per-
form better, in terms of efficiency and enantioselectivity,
than other well known diphosphanes such as DuPHOS, DI-
PAMP or PHANEPHOS. Recently, their high efficiency in
the catalytic hydrogenation of (E)-β-amino acid derivatives
has also been established.[3] It thus seems that the combi-
nation of a conformationally constrained phosphetane ring
with the relatively flexible ferrocene backbone represents a
favourable structural feature for high catalytic efficiency.
For comparison, the structurally analogous 1,1Ј-bis(phos-
pholanyl)ferrocenes give only moderate to low enantioselec-
tivities in hydrogenation reactions.[4]
Following our initial report on the synthesis and prelimi-
nary catalytic screening of these ligands, we wish to disclose
here the preparation and first structural characterisation of
ruthenium and rhodium complexes of 1, which are efficient
catalyst precursors for hydrogenation reactions. Additional
data on their catalytic properties is also reported.
Results and Discussion
Besides their good catalytic properties, the potential util-
ity of diphosphanes 1 is also founded on their easy avail-
ability: diphosphanes 1 are easily accessible from fairly in-
expensive chiral 1,3-diols and their structures and catalytic
properties are easily and finely tuned by variation of the R
groups at the α-positions of the phosphetane rings.
Crystal Structures of the [(S,S-1c)RuCl2(Py)2] and
[(COD)Rh(S,S-1c)OTf] Complexes
With regard to the excellent catalytic properties of the
1,1Ј-bis(phosphetano)ferrocenes 1, it is highly interesting to
investigate the coordinating behaviour of these new ligands
towards catalytically active transition metals, as well as the
structural features of their complexes. With this aim, we
prepared the [{(S,S)-iPr-1}RuCl2(Py)2] complex 2, which is
a potential precursor for hydrogenation catalysts. The syn-
thesis of 2 was performed according to the Bergens pro-
cedure:[5] the reaction between 1 and the [(NBD)RuCl2Py2]
complex (NBD ϭ norbornadiene, Py ϭ pyridine) at room
temperature. Facile and quantitative displacement of the
[a]
`
´
Laboratoire de Synthese Selective Organique et Produits
Naturels Ϫ ENSCP,
11, rue Pierre et Marie Curie, 75231 Paris Cedex 05, France
Fax: (internat.) ϩ 33-1/4407 1062
E-mail: marinet@ext.jussieu.fr
[b]
´ ´ ´ ´
Laboratoire Heteroelements et Coordination
Ϫ
Ecole
Polytechnique,
91128 Palaiseau Cedex, France
Eur. J. Inorg. Chem. 2003, 2583Ϫ2590
DOI: 10.1002/ejic.200300051
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2583

ˆ
A. Marinetti, F. Labrue, B. Pons, S. Jus, L. Ricard, J.-P. Genet
FULL PAPER
˚
unsaturated ligand by the chelating diphosphane is ob-
served, as shown in Equation (1).
Table 1. Selected interatomic distances [A] and angles [deg] for
complex 2c
Ru(1)ϪP(1)
Ru(1)ϪP(2)
Ru(1)ϪN(1)
Ru(1)ϪN(2)
Ru(1)ϪCl(1)
2.321(1) Ru(1)ϪCl(2)
2.319(1) P(1)ϪC(1)
2.231(3) P(1)ϪC(3)
2.170(3) P(2)ϪC(11)
2.425(1) P(2)ϪC(9)
2.437(1)
1.892(3)
1.887(3)
1.908(3)
1.884(3)
P(1)ϪRu(1)ϪP(2)
N(1)ϪRu(1)ϪN(2) 81.7(1)
P(1)ϪRu(1)ϪN(1) 90.20(7)
98.57(3)
C(1)ϪP(1)ϪC(3)
C(4)ϪP(1)ϪRu(1)
C(11)ϪP(2)ϪC(12)
76.3(1)
119.9(1)
77.3(2)
The ruthenium complexes 2 are obtained in moderate to
high yields after crystallisation from dichloromethane/pen- planar and octahedral complexes of dppf[6] and other 1,1Ј-
tane mixtures, as air-stable, orange, crystalline solids. NMR bis(phosphanyl)ferrocenes.[7] The conformation of the Cp
spectroscopy shows that complexes 2 have C2-symmetrical rings is staggered, with a torsion angle of 45.7°. The ro-
structures, with equivalent phosphetane and also pyridine tation of the cyclopentadienyl rings against each other, to-
moieties. The 31P NMR resonance of complexes 2 is shifted gether with their tilt, produces a fairly large PϪP distance
˚
downfield with ∆δ of 60Ϫ70 ppm relative to 1, the less hin- of 3.51 A (as compared with the CpϪCpЈ distances of
˚
dered complex 1a undergoing the larger shift upon coordi- 3.28 A).
nation.
The structural parameters of the phosphetane rings show
The first structural characterisation of any 1,1Ј- standard values, with small intracyclic CϪPϪC angles of
bis(phosphetano)ferroceneϪtransition metal derivative has 76.3(1) and 77.3(2)°, respectively, and ring bending angles
been performed on complex 2c. An ORTEP drawing is of 25.5 and 17.6°, respectively.
shown in Figure 1, and selected bond lengths and angles
are reported in Table 1.
A significant structural feature of the complex is the δ
conformation[8] adopted by the ferrocene backbone, which
Complex 2c crystallises in the orthorhombic system, with enforces a dissymmetric coordination of the phosphetane
one dichloromethane molecule per molecular unit. The ru- rings (see Figure 2): the C(3) and C(9) ring atoms, and the
thenium atom is approximately octahedral. The bis-equa- corresponding isopropyl substituents, are shifted away from
torially coordinated diphosphane gives a PϪRuϪP bite an- the ruthenium coordination plane, in pseudoaxial positions,
gle of 98.57(3)°, in good agreement with the expected value, while the C(1) and C(11) atoms occupy pseudoequatorial
as large deflections from the ideal 90° (metal natural bite positions. The corresponding dihedral angles between the
angle) up to about 100° are usually observed in square- P(1)ϪRuϪP(2) plane and the PϪC axis are Ϫ93°
Figure 1. ORTEP drawing for the ruthenium complex 2c
2584
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2583Ϫ2590

Rhodium and Ruthenium Complexes of 1,1Ј-Bis(phosphetano)ferrocenes
FULL PAPER
[P(1)ϪC(3)], 163° [P(1)ϪC(1)], Ϫ128° [P(2)ϪC(9)] and 141° Ruthenium-Promoted Hydrogenations of β-Keto Esters
[P(2)ϪC(11)]. The δ conformation of the ferrocene back-
As shown by Bergens, many [(diphosphane)RuCl2Py2]
complexes afford efficient hydrogenation catalysts after
treatment with acids. This is also the case for complexes 2:
when activated by addition of four equivalents of HBr,
these complexes catalyse the asymmetric hydrogenation of
carbonyl derivatives. Methyl acetoacetate and ethyl isobu-
tyrylacetate were selected as model substrates for the hydro-
genation reactions (Table 3).
bone therefore tends to minimise the steric hindrance be-
tween the isopropyl groups, which are oriented toward the
metal, and the other ligands in the equatorial coordination
plane of ruthenium. In a sense, this also balances the dis-
symmetry of the ruthenium environment created by the chi-
ral phosphetane units and could decrease the asymmetric
discrimination from these ligands somewhat.
Under the conditions given (see Table 3), ligands 1aϪc
display moderate to high enantioselectivities as a function
of the phosphetane α-substituents: the enantiomeric ex-
cesses of the final β-hydroxy esters increase significantly
with increasing steric hindrance of the ligand. Thus, for in-
stance, ees of 44%, 56%, and 80% were obtained in the
hydrogenation of methyl acetoacetate with use of ligands
1a, 1b, and 1c, respectively. The observed trend contrasts
with the behaviour of the same ligands in rhodium-pro-
moted hydrogenations of olefinic substrates, where the best
enantioselectivities are attained with the 2,4-diethyl-substi-
tuted ligand.[2] This once again confirms the need for highly
specific tuning of the ligand substituents for any given reac-
tion.
Figure 2. Coordination mode of the bis(phosphetano)ferrocene
moiety in the ruthenium complex 2c
A similar coordination mode is observed in the rhodium
complex 3c, containing the same ligand 1c. An ORTEP
drawing and selected bond angles and distances for com-
plex 3c are shown in Figure 3 and Table 2, respectively.
Two other ruthenium catalyst precursors were evaluated
under the above hydrogenation reactions: catalysts were
formed in situ from 1aϪc and [(C6H6)RuCl2]2 or [(COD)-
Ru(2-methylallyle)2],[9] by known procedures. Selected re-
sults are given in Table 3. Both catalytic systems give activi-
ties and enantioselectivities either comparable to or some-
what lower than those obtained with complexes 2. Thus,
in our hands, complexes 2 were the most suitable catalyst
precursors, being well defined, bench-stable and easy to
handle.
The few catalytic tests above show that the 1,1Ј-bis(phos-
phetano)ferrocenes 1 afford moderate catalytic activities
and quite high, albeit not magnificent, enantioselectivities
in ruthenium-promoted hydrogenations of carbonyl deriva-
tives. At first sight it seems that, in these reactions, ligands
1 will be not very competitive with other chiral ligands,[10]
and especially with atropisomeric diphosphanes.[11] Of
course, this cannot be regarded as a final settlement, as very
few substrates, catalytic systems, and conditions have been
evaluated. Fine tuning of the phosphetane substituents and/
or of the reaction conditions could allow better achieve-
ments.
Figure 3. ORTEP drawing of the [(COD)Rh(1c)OTf] complex 3c
Table 2. Selected interatomic distances [A°] and angles [deg] for
complex 3c
Rh(1)ϪP(1)
Rh(1)ϪP(2)
P(1)ϪC(1)
P(1)ϪC(3)
P(1)ϪC(19)
2.324(1) C(1)ϪC(2)
2.317(1) C(2)ϪC(3)
1.877(4) P(2)ϪC(10)
1.884(4) P(2)ϪC(12)
1.810(3) P(2)ϪC(24)
1.570(5)
1.554(6)
1.885(3)
1.869(4)
1.813(3)
Rhodium-promoted Hydrogenations of Functionalised
Olefins
The catalytic efficiency of 1,1Ј-bis(phosphetano)ferro-
cenes 1 in rhodium()-promoted hydrogenations of func-
tionalised olefins has already been established.[1Ϫ3] The
main purpose of this study was to gain insight into their
P(1)ϪRh(1)ϪP(2)
C(19)ϪP(1)ϪRh(1) 121.5(1)
96.90(3)
C(1)ϪP(1)ϪC(3)
C(10)ϪP(2)ϪC(12) 77.9(2)
78.1(2)
Phosphetane 1c thus generally seems to favour δ confor- behaviour in these reactions, in comparison with that of
mations in its transition metal complexes, as a result of its other phosphetane-based ligands, namely CnrPHOS[12] and
(S,S) stereochemistry.
BPE-4.[13]
Eur. J. Inorg. Chem. 2003, 2583Ϫ2590
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2585

ˆ
A. Marinetti, F. Labrue, B. Pons, S. Jus, L. Ricard, J.-P. Genet
FULL PAPER
Table 3. Asymmetric hydrogenation of β-keto esters
Substrate
Ligand
Ru precatalyst
Conditions[a]
Conv. (%)
ee (config.)
4
1a
1a
1b
1c
1a
1a
1b
1c
1c
1c
(1a)RuCl2Py2 ϩ HBr
(COD)Ru(C4H7)2 ϩ HBr
(1b)RuCl2Py2 ϩ HBr
(1c)RuCl2Py2 ϩ HBr
(1a)RuCl2Py2 ϩ HBr
(COD)Ru(C4H7)2 ϩ HBr
(1b)RuCl2Py2 ϩ HBr
(1c)RuCl2Py2 ϩ HBr
[(C6H6)RuCl2]2
(a)
(b)
(a)
(a)
(a)
(b)
(b)
(a)
(a)
(c)
100
100
100
100
31
36
50
40
40
100
44 (R)
11 (R)
56 (R)
80 (R)
53 (S)
58 (S)
80 (S)
84 (S)
64 (S)
63 (S)
5
[(C6H6)RuCl2]2
[a]
Reaction conditions: 1 mmol substrate, S/C ϭ 100, MeOH (1 mL) as solvent, reaction time 24 hϪ48 h; (a) 50 °C; 10 bar H2, (b) 50
°C; 80 bar H2, (c) 80 °C, 10 bar.
More precisely, we examine here the enantioselectivity of
the catalytic hydrogenation of dehydro amino acid deriva-
tives, as a function of several experimental parameters. An
analogous study on the CnrPHOS and BPE ligands high-
lighted, among other things, an unusually strong effect of
the hydrogen pressure on the enantioselectivity, with higher
ees being obtained at higher pressures.[14] The preliminary
data on this phenomenon were confirmed by the additional
Figure 4. Plotting of data from Entries 2 and 3 in Table 4
experiments shown in Table 4 and Figure 4. Methyl α-aceta-
midocinnamate 6 was hydrogenated, at various hydrogen
pressures, in the presence of a rhodium catalyst generated in
situ from [Rh(COD)2OTf] and (S,S)-CyBPE-4. Mechanical
stirring was applied to ensure efficient adsorption of H2
into the solution. For comparison, the ees obtained with
magnetic stirring are reported in entry 1 (from ref.[14]):
mechanical stirring improves the enantioselectivity of reac-
tions performed under otherwise identical conditions. This
is probably related to more efficient and homogeneous dihy-
drogen exchange between gas and liquid phases.
Two series of experiments were performed, in CH2Cl2
and in a C6H6/CH2Cl2 mixture (4:1). The expected signifi-
cant effect of hydrogen pressure on stereoselectivity was ob-
served in each case: increased hydrogen pressure clearly fav-
ours formation of the R enantiomer, while the S enantiomer
predominates at low pressures. The enantioselectivity levels
depend on the solvent used, higher ees being attained in the
benzene/dichloromethane mixtures. Mechanistic impli-
cations of analogous results have already been discussed in
our preliminary communication. Most probably, the rever-
sal of stereoselectivity should be related to some form of
‘‘tilting’’ over to a new reaction pathway which operates at
Table 4. Enantiomeric excesses of the N-acetylphenylalanine methyl
ester obtained from 6 by Rh/(S,S)-CyBPE-4-promoted hydrogen-
ations
[a]
From ref.[14] 1% catalyst, substrate concentration: 0.1 , room
temperature.
2586
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2583Ϫ2590

Rhodium and Ruthenium Complexes of 1,1Ј-Bis(phosphetano)ferrocenes
FULL PAPER
higher pressures. According to the most recent studies on
rhodium-promoted enamide hydrogenation with electron-
rich phosphanes,[15] competition between the ‘‘olefin’’ and
the ‘‘hydride’’ mechanisms, operating at high pressure,
seems to be a reasonable hypothesis.
Table 5. Enantiomeric excesses of the catalytic hydrogenations pro-
moted by complexes 3a and 3c; all experiments give the R-config-
ured β-amino ester
After the above studies, we wondered whether the ferro-
cenyl-substituted phosphetanes 1 would or would not dis-
play analogous behaviour in rhodium-promoted hydrogen-
ations. We therefore examined the pressure effect in the
hydrogenation of the dehydroamino esters 6 with the pre-
formed complexes 3 as the catalyst precursors.
creased H2 pressure improves the enantioselectivity from
87%, under 1 bar pressure, to 94% under 5 bar (at room
temperature, in MeOH; R enantiomer).
The electron-rich character of P-ferrocenyl-substituted
phosphetanes was ascertained by IR spectroscopy of the
[Cl(CO)Rh{(R,R)-1-ferrocenyl-2,4-dimethylphosphetane}]
complex 7a.
The ν(CO) value for the ferrocenyl-substituted phosphet-
ane complex 7a appears at lower frequency than in the anal-
ogous P-phenylphosphetane complex 7c, within the range
expected for electron-rich phosphanes. For comparison, the
ν(CO) values for analogous rhodium complexes of cyclic
and acyclic phosphanes are reported in Table 6. The data in
Table 6 also suggest that phosphetanes are poorer donating
ligands than the corresponding phospholanes, as an effect
of ring size.
From the data in Table 4 it appears that the pressure/
enantioselectivity relationship for phosphetanes 1a and 1c
does not follow the same trend as for Cy-BPE-4 and
CnrPHOS: a slight decrease in the enantioselectivity is ob-
served at higher pressure, both in dichloromethane/benzene
mixtures and in methanol as solvents (see entries 1, 2, 5,
and 6). This suggests that the pressure effect observed with
Cy-BPE-4 and CnrPHOS cannot be assigned specifically to
their phosphetane-type structures; their catalytic behaviour
should be influenced by their whole structures. In particu-
lar, their electronic properties should be crucial, as it has
been demonstrated that the stereochemical issue of the
hydrogenation of dehydroamino acid derivatives is mainly
governed by the electron-donating power of the diphos-
phane ligands.[14]
FerroTANEs 1 simply behave like other electron-rich
diphosphanes such as DuPHOS[16] or BisP,[17] the (S,S)-
configured ligand 1c giving the expected (R)-configured
product in good enantiomeric excesses. The enantiomeric
excesses are then slightly modulated by hydrogen pressure,
as well as by any other experimental conditions, in the
usual manner.
Table 6. ν(CO) values for the [Cl(CO)RhL2] complexes 7
Thus far, the best conditions for the hydrogenation of 6
are the following: phosphetane 1a as the ligand, 1 bar of
H2 pressure, MeOH as the solvent, and a reaction tempera-
ture of about 50 °C. An enantiomeric excess of 96% is ob-
tained.
The results in Table 5 also suggest that hydrogenation
reactions with ligands 1a and 1c require slightly different
optimal conditions. Fine tuning of the reaction conditions
is thus required for each ligand-substrate couple in order to
attain the highest selectivities, even within a single series
of ligands.
In summary, the hydrogenation reactions of dehydro
amino acid derivatives promoted by rhodium complexes of
the 1,1Ј-bis(phosphetano)ferrocenes 1 do not display the
unusual pressure/enantioselectivity relationship shown by
other C2-symmetric bis(phosphetane) ligands. Ligands 1 be-
have like electron-rich diphosphanes and afford signifi-
cantly high enantioselectivities under optimised experimen-
tal conditions.
When a different substrate is considered, optimisation is
once again necessary, as variation of the experimental pa-
rameters has unpredictable effects on enantiomeric excesses.
In the hydrogenation of methyl acetamidoacrylate with
Experimental Section
All reactions were performed under inert atmosphere (Ar) by use of
complex 3c, for instance, we have observed that an in- Schlenk techniques. Solvents were purified according to standard
Eur. J. Inorg. Chem. 2003, 2583Ϫ2590
www.eurjic.org  2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 2587

ˆ
A. Marinetti, F. Labrue, B. Pons, S. Jus, L. Ricard, J.-P. Genet
FULL PAPER
procedures. 1H, 13C, and 31P NMR spectra were recorded on a
Bruker AM 400 at 400.13 Hz, 100.6 Hz, and 162 Hz, respectively.
HPLC was performed on a Waters 600 chromatograph equipped
with a variable wavelength detector and Daicel Chiracel OD-H col-
umn. Optical rotations were measured on a PerkinϪElmer 241
polarimeter, with a 1 dm cell. IR spectra were recorded on a Bruker
IFS 48 instrument.
Table 7. Crystal structure determination for compound 2c
Empirical formula
Formula mass
Crystal habit
Crystal dimensions [mm]
Crystal system
C39H56Cl4FeN2P2Ru
913.52
tan needles
0.20 ϫ 0.14 ϫ 0.14
orthorhombic
P212121
9.651(5)
19.366(5)
22.000(5)
90.000(5)
90.000(5)
90.000(5)
4112(3)
Space group
˚
a [A]
General Procedure for the Synthesis of trans-[RuCl2{1,1Ј-
Bis(dialkylphosphetanyl)ferrocene}Py2] (2): A solution of trans-
[RuCl2(NBD)Py2][5] (100 mg, 0.23 mmol) and ligand 1 (0.23 mmol)
in CH2Cl2 (13 mL) was stirred overnight at room temperature. The
solvent was removed under vacuum and the residue was recrys-
tallized from a dichloromethane/pentane mixture to afford com-
plexes 2 as yellow, crystalline solids.
˚
b [A]
˚
c [A]
α [°]
β [°]
γ [°]
3
˚
V [A ]
Z
4
d [g·cmϪ3
F000
]
1.476
1888
1.088
µ [cmϪ1
]
trans-[RuCl2{1,1Ј-Bis[(R,R)-2,4-dimethyl-1-phosphetanyl]ferro-
cene}Py2] (2a): Yield 120 mg, 73%. 31P NMR (CDCl3): δ ϭ 87. 1H
NMR (CDCl3): δ ϭ 9.6 (br., 4 H), 7.60 (m, 2 H), 7.3 (br., 4 H),
5.14 (br. s, 2 H, CHcp), 4.50 (m, J ഠ 1 Hz, 4 H, CHcp), 4.34 (m,
J ഠ 1 Hz, 2 H, CHcp), 2.66 (m, 2 H), 2.40 (m, 4 H), 1.79 (m, 2 H),
Absorption corrections
0.8626
0.8118 min
max
KappaCCD
Mo-Kα
Diffractometer
X-ray source
˚
λ [A]
0.71069
3
3
1.28 (dd, JH,H ϭ 7.7, JP-H ϭ 16.4 Hz, 6 H, CH3), 0.92 (dd,
3JH,H ϭ 6.3, 3JP-H ϭ 14.4 Hz, 6 H, CH3) ppm. 13C NMR (CDCl3,
100 MHz): δ ϭ 152.6 (br., CHPy), 152.1 (br., CHPy), 136.0 (CHPy),
123.6 (br., CHPy), 123.4 (br., CHPy), 77.6 (t, J ϭ 7.8 Hz, CHCp),
74.0 (t, J ϭ 3.4 Hz, CHCp),73.0 (m, AXXЈ, CCpP), 72.0 (CHCp),
71.0 (t, J ϭ 1.8 Hz, CHCp), 40.1 (t, J ϭ 7.6 Hz, CH2), 32.0 (m,
Monochromator
T [K]
Scan mode
graphite
150.0(10)
phi
Maximum q
hkl ranges
Reflections measured
30.02
Ϫ13 13; Ϫ27 27; Ϫ30 30
11960
AXXЈ, PCH), 30.2 (m, AXXЈ, PCH), 18.8 (CH3), 17.5 (t, J ϭ Independent reflections
11960
0.0000
11161
Ͼ 2σ(I)
Fsqd
mixed
450
24
20.0965
0.0353
0.006(17)
0.0380; 5.3251
1.072
Rint
2.3 Hz, CH3) ppm. [α]D ϭ Ϫ782 (c ϭ 0.2, CHCl3). C30H38Cl2FeN2-
P2Ru (716.028): calcd. C 50.30, H 5.35, N 3.91; found C 50.66, H
5.96, N 3.82.
Reflections used
Criterion
Refinement type
Hydrogen atoms
Parameters refined
Reflections / parameter
wR
trans-[RuCl2{1,1Ј-Bis[(R,R)-2,4-diethyl-1-phosphetanyl]ferro-
cene}Py2] (2b): Yield 124 mg, 70%. 31P NMR (CDCl3): δ ϭ
82 ppm. 1H NMR (CDCl3): δ ϭ 9.6 (br., 4 H), 7.60 (t, JH,H
ϭ
R1
7.4 Hz, 2 H,CH), 7.3Ϫ7.0 (br., 4 H, CH), 5.10 (s, 2 H, CHcp), 4.50
(s br, 4 H, CHcp), 4.32 (m, 2 H, CHcp), 2.77 (br., 2 H), 2.40 (br., 2
H), 2.22 (m, 2 H), 2.04Ϫ1.95 (m, 4 H), 1.70 (m, 2 H), 0.87 (m, 4
Flack’s parameter
Weighs a, b1
GoF
3
3
Ϫ3
H), 0.79 (t, JH,H ϭ 6.9 Hz, 6 H, CH3), 0.57 (t, JH,H ϭ 7.4 Hz, 6
H, CH3) ppm. 13C NMR (CDCl3, 100 MHz): δ ϭ 152.4 (br.,
CHPy), 136.0 (CHPy), 123.4 (br., CHPy), 77.5 (t, J ϭ 7.8 Hz, CHCp),
74.0 (t, J ϭ 3.4 Hz, CHCp), 73.5 (m, AXXЈ, CCpP), 72.0 (CHCp),
70.8 (CHCp), 39.2 (m, AXXЈ, PCH), 37.6 (m, AXXЈ, PCH), 34.0
(t, J ϭ 7.3 Hz, CH2), 25.7 (CH2CH3), 23.8 (CH2CH3), 13.0 (t, J ϭ
7.5 Hz, CH3), 12.0 (t, JP,C ϭ 5.9 Hz, CH3) ppm. [α]D ϭ Ϫ907 (c ϭ
0.5, CHCl3). C34H46Cl2FeN2P2Ru (772.091): calcd. C 52.86, H
6.00, N 3.63; found C 52.41, H 6.32, N 3.76.
˚
difference peak / hole [e A
]
0.767(0.090) / Ϫ0.944(0.090)
General Procedure for Hydrogenations Carried out with Complexes
2 as Catalyst Precursors: Hydrogenation experiments were carried
out at a 1 mmol scale. The ruthenium complex 2 (1·10Ϫ2 mmol)
was weighed into a small glass reactor fitted with a rubber septum
and flushed with argon. Degassed MeOH (1 mL) and then a solu-
trans-[RuCl2{1,1Ј-Bis[(S,S)-2,4-diisopropyl-1-phosphetanyl]ferro- tion of HBraq in MeOH (0.16 , 0.25 mL, 4 equivalents) were ad-
cene}Py2] (2c): Yield 105 mg (55%). 31P NMR (CDCl3): δ ϭ 71.
ded with stirring at room temperature. After a few minutes, the
1H NMR (CDCl3): δ ϭ 9.68 (br., 2 H, CHPy), 9.61 (br., 2 H, CHPy), substrate was introduced by syringe. The reaction vessel was placed
7.57 (t, JH,H ϭ 7.2 Hz, 2 H, CHPy), 7.19 (br., 2 H, CHPy), 7.09 (br.,
2 H, CHPy), 5.29 (s, 2 H, CHCp), 4.51 (s, 2 H, CHCp), 4.45 (s, 2 H,
in a stainless steel autoclave under argon. The argon atmosphere
was replaced by hydrogen at the given pressure. The reaction was
CHCp), 4.28 (m, Jഠ2 Hz, 2 H, CHCp), 3.10 (m, 2 H), 2.77 (m, 2 allowed to proceed for 24Ϫ48 h at the temperature given in Table 3.
3
1
H), 2.0 (m, 4 H), 1.75 (m, 2 H), 1.44 (m, 2 H), 1.05 (d, JH,H
ϭ
Conversion rates were determined by H NMR spectroscopy. The
3
6.5 Hz, 6 H, CH3), 0.79 (d, JH,H ϭ 6.6 Hz, 6 H, CH3), 0.59 (d, enantiomeric excesses of the final alcohols were determined by chi-
3JH,H ϭ 6.7 Hz, 6 H, CH3), 0.56 (d, JH,H ϭ 6.6 Hz, 6 H, CH3). ral CG and the absolute configurations were assigned from the GC
3
13C NMR (CDCl3, 100 MHz): δ ϭ 153.7 (CHPy), 153.0 (CHPy), retention times by comparison with known samples. Methyl 3-
135.9 (CHPy), 123.8 (CHPy), 123.1 (CHPy), 77.9 (CHCp), 73.7 hydroxybutyrate: Lipodex A, flow 1 mL·minϪ1, initial temperature
(CHCp), 73.2 (CCpP), 72.4 (CHCp), 70.3 (CHCp), 46.4 (m, PCH), 35 °C (30 min), rate 1 °C·minϪ1, final temperature 70 °C; retention
44.0 (m, PCH), 29.2 (CH), 27.8 (CH2), 24.5 (CH3), 23.5 (CH3), times 45 (S) and 48 (R) min. Ethyl 4-methyl-3-hydroxypentanoate:
19.8 (CH3), 19.2 (CH3) ppm. For crystal data see Table 7.
Lipodex A, flow 1 mL·minϪ1, initial temperature 50 °C (5 min),
2588
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2583Ϫ2590

Rhodium and Ruthenium Complexes of 1,1Ј-Bis(phosphetano)ferrocenes
rate 0.5 °C·minϪ1, final temperature 70 °C; retention times 48.6 (S)
FULL PAPER
Table 8. Crystal structure determination for compound 3c
and 49.2 (R) min.
Preparation of other Catalyst Precursors: (a) According to ref.[9]
[(cod)Ru(2-methylallyl)2] (3.2 mg, 1·10Ϫ2 mmol) and FerroTANE 1
(1.1·10Ϫ2 mmol) were dissolved in acetone (1 mL) under argon. An
HBraq solution (0.16  in MeOH, 137 µL, 2.2 equivalents) was
then added at room temperature, and the mixture was stirred for
30 min. The solvent was evaporated and the crude residue was
taken up in degassed MeOH and used as the catalyst for the hydro-
genation reactions (see Table 3). (b) The catalyst precursor was pre-
pared at room temperature by mixing the [(benzene)RuCl2]2 com-
plex (2.5 mg) with 1 in a 1:2 ratio, in CH2Cl2 at room temperature.
After evaporation of the solvent, the crude residue was used as the
catalyst precursor.
Molecular formula
Molecular weight
Crystal habit
Crystal dimensions [mm]
Crystal system
C37H56F3FeO3P2RhS
858.58
orange plate
0.22 ϫ 0.08 ϫ 0.06
orthorhombic
P212121
11.9840(10)
17.1000(10)
18.8990(10)
90.00
90.00
90.00
Space group
˚
a [A]
˚
b [A]
˚
c [A]
α [°]
β [°]
γ [°]
3
˚
V [A ]
3872.9(4)
Z
4
General Procedure for the Synthesis of the [(COD)Rh(1)OTf] Com-
plexes 3: A dichloromethane solution of diphosphane 1 (0.2 mmol,
1 mL CH2Cl2) was added dropwise, at room temperature, to a solu-
tion of (COD)2RhOTf (93 mg, 0.2 mmol) in the same solvent
(1 mL). After the mixture had been stirred for about 15 min, the
solvent was removed and the residue was recrystallised either from
a dichloromethane/ether mixture (for 3c) or from a dichlorometh-
ane/pentane mixture (for 3a) to afford a yellow-orange solid.
d [g-cmϪ3
F000
]
1.472
1784
0.986
µ [cmϪ1
Absorption corrections
]
multiple scans; 0.8123 min,
0.9432 max
KappaCCD
Mo-Kα
0.71069
graphite
150.0(10)
phi and omega scans
Diffractometer
X-ray source
λ [A]
Monochromator
T [K]
Scan mode
Maximum θ
hkl ranges
Reflections measured
Unique data
Rint
Reflections used
Criterion
Refinement type
Hydrogen atoms
Parameters refined
Reflections/parameter
wR
˚
[(COD)Rh{1,1Ј-Bis[(R,R)-2,4-dimethyl-1-phosphetanyl]ferro-
cene}OTf] (3a): Yield 97 mg, 65%. 31P NMR (CDCl3): δ ϭ 65 (d,
JP-Rh ϭ 147 Hz) ppm. 1H NMR (CDCl3): δ ϭ 5.8 (br. m, 2 H,
CHCOD), 5.15 (br. m, 2 H, CHCOD), 4.70 (s, 2 H, CHCp), 4.66 (s, 2
H, CHCp), 4.62 (s, 2 H, CHCp), 4.37 (s, 2 H, CHCp), 1.65 (dd,
30.03
Ϫ16 16; Ϫ23 23; Ϫ26 26
11222
11222
0.0000
10128
Ͼ 2σ(I)
Fsqd
mixed
382
26
20.1184
0.0416
0.050(18)
0.0702; 1.9976
1.040
3
3
3JH-P ϭ 19.7, JH,H ϭ 7.6 Hz, 6 H, CH3), 0.99 (dd, JH-P
ϭ
3
16.3, JH,H ϭ 7.0 Hz, 6 H, CH3) ppm.
[(COD)Rh{1,1Ј-Bis[(S,S)-2,4-diisopropyl-1-phosphetanyl]ferro-
ceneOTf} (3c): Yield 120 mg, 70%. 31P NMR (CDCl3): δ ϭ 47 (d,
J
P-Rh ϭ 146 Hz). 1H NMR (CDCl3): δ ϭ 5.56 (br. m, 2 H, CHCOD),
4.90 (br. m, 2 H, CHCOD), 4.80 (s, 2 H, CHCp), 4.60 (s, 4 H, CHCp),
4.37 (s, 2 H, CHCp), 2.2Ϫ2.8 (18H), 2.05 (q, J ϭ 10.1 Hz, 2 H),
3
3
1.84 (d, JH,H ϭ 6.5 Hz, 6 H, CH3), 1.15 (d, JH,H ϭ 6.1 Hz, 6 H,
CH3), 0.70 (d, 3JH,H ϭ 6.3 Hz, 6 H, CH3), 0.69 (d, 3JH,H ϭ 5.5 Hz,
6 H, CH3) ppm. C37H56F3FeO3P2RhS·CH2Cl2 (942.132): calcd. C
48.37, H 6.20; found C 49.80, H 6.56. For crystal data see Table 8.
R1
Flack’s parameter
Weights a, b
GoF
Ϫ3
˚
difference peak/hole [e·A
]
0.825(0.104)/Ϫ0.546(0.104)
The triflate anion (near Ϫ0.37, 0.10, 0.31) was highly disordered.
Attempts to generate disorder (Shelxl FRAG and PART) all re-
sulted in some non-positive thermal displacement parameters.
Consequently, it was decided to account for its electron density by
use of the Platon SQUEEZE function.
The same experimental procedure was applied to the hydrogenation
reactions reported in Table 5. The preformed rhodium complexes
3a and 3c were used as the catalysts. Experiments were carried out
at a 1 mmol scale, with a catalyst:substrate ratio of 5·10Ϫ3
.
Rhodium-Promoted Hydrogenations of Methyl α-Acetamidocinna-
mate: The hydrogenation experiments described in Table 4 were
performed in a Teflon-coated autoclave, with mechanical stirring.
The rhodium catalyst was prepared at room temperature by mixing
[(COD)2RhOTf] (11.2 mg, 2.4·10Ϫ2 mmol) and (S,S)-Cy-BPE-4
(14.4 mg, 2.8·10Ϫ2 mmol) in CH2Cl2 (2 mL) for about 10 min. A
degassed solution of the substrate (525 mg, 2.4 mmol) in 18 mL of
solvent (CH2Cl2 or CH2Cl2/C6H6, 1:4 mixture) was then added and
the system was pressurised with dihydrogen. The reaction time was
about 24 h. The catalyst was removed by filtration through a short
alumina column, with a cyclohexane/ethyl acetate mixture (1:1) as
the eluent. Enantiomeric excesses were determined either by chiral
HPLC or by chiral GC. The absolute configuration was assigned
from the retention times by comparison with known samples.
HPLC: Chiracel OD-H column, hexane/2-propanol (90:10), reten-
tion times: 11.5 min (R) and 14.8 min (S). GC: Chirasil--Val col-
umn, flow ϭ 1 mL/min, T ϭ 140 °C, retention times: 37 min (R),
39 min (S).
Synthesis of the [Cl(CO)RhL2] Complexes 7: 2,4-Dimethyl-1-ferro-
cenylphosphetane,[18] 2,4-dimethyl-1-phenylphosphetane,[19] and
2,5-dimethyl-1-phenylphospholane[20] were prepared as described
previously. The rhodium complex [ClRh(CO)2]2 (20 mg, 5·10Ϫ2
mmol) and the phosphetane (or phospholane) ligand (20·10Ϫ2
mmol) were allowed to react in CH2Cl2 (2 mL) at room tempera-
ture for 15 min. 31P NMR analysis of the crude reaction mixture
in C6D6 showed quantitative formation of a single rhodium-phos-
phane complex. Complexes 7aϪc were characterised by NMR
spectroscopy, without further purification. IR spectra of the crude
complexes were recorded in CH2Cl2.
[ClRh(CO)(2,4-Dimethyl-1-ferrocenylphosphetane)] (7a): 31P NMR
1
(C6D6): δ ϭ 63.6 (d, JP-Rh ϭ 117 Hz). H NMR (C6D6): δ ϭ 5.32
(m, 1 H, CHCp), 4.38 (s, 5 H, CHCpЈ), 4.26 (m, 2 H, CHCp), 4.19
(m, 1 H, CHCp), 3.32 (m, 1 H), 2.52 (m, 1 H), 2.3Ϫ2.0 (m, 2 H),
1.69 (dt, J ϭ 10.0, J ϭ 7.7 Hz, CH3), 1.15 (q, J ϭ 7.8 Hz, CH3).
Eur. J. Inorg. Chem. 2003, 2583Ϫ2590
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2589

ˆ
A. Marinetti, F. Labrue, B. Pons, S. Jus, L. Ricard, J.-P. Genet
FULL PAPER
13C NMR (C6D6): δ ϭ 188.7 (dt, JCRh ϭ 75, JC,P ϭ 17 Hz, CO),
78.6 (t, J ϭ 8.9 Hz, CHCp), 71.8 (CHCp), 71.6 (t, J ϭ 14.1 Hz, CCp),
71.0 (CCp), 70.8 (CHCp), 69.6 (CHCpЈ), 39.5 (t, J ϭ 5.3 Hz, CH2),
31.3 (t, J ϭ 18.4 Hz, PCH), 30.0 (t, J ϭ 18.9 Hz, PCH), 20.1 (CH3),
17.5 (CH3) ppm.
ics 1999, 18, 1041Ϫ1049. U. Nettekoven, M. Widhalm, P. C. J.
Kamer, P. W. N. M. van Leeuwen, K. Mereiter, M. Lutz, A. L.
Spek, Organometallics 2000, 19, 2299Ϫ2309. A. R. Elasgir, F.
Gassner, H. Görls, E. Dinjus, J. Organomet. Chem. 2000, 597,
139Ϫ145.
[8]
[9]
Inorg. Chem. 1970, 9, 1Ϫ5.
[ClRh(CO)(2,5-Dimethyl-1-phenylphospholane)] (7b): 31P NMR
ˆ
J.-P. Genet, C. Pinel, V. Ratovelomanana-Vidal, S. Mallart, X.
1
Pfister, L. Bischoff, M. C. Cano De Andrade, S. Darses, C.
Galopin, A. Laffitte, J. A, Tetrahedron: Asymmetry 1994, 5,
675Ϫ690.
M. J. Burk, T. G. P. Harper, C. S. Kalberg, J. Am. Chem. Soc.
1995, 117, 4423Ϫ4424.
(C6D6): δ ϭ 49.5 (d, JP-Rh ϭ 123 Hz). H NMR (C6D6): δ ϭ 7.7
3
3
(m, 2 H), 7.2 (m, 3 H), 1.28 (dd, JH-P ϭ 14.6, JH,H ϭ 7.0 Hz, 3
H, CH3), 0.75 (dd, 3JH-P ϭ 16.0, 3JH,H ϭ 7.2 Hz, 3 H, CH3) ppm.
[10]
[11]
[ClRh(CO)(2,4-Dimethyl-1-phenylphosphetane)] (7c): 31P NMR
1
For recent reviews see: H. Kumobayashi, T. Miura, N. Sayo, T.
Saito, X. Zhang, Synlett 2001, 1055Ϫ1064. M. McCarthy, P. J.
Guiry, Tetrahedron 2001, 57, 3809Ϫ3844.
(C6D6): δ ϭ 66.1 (d, JP-Rh ϭ 116 Hz). H NMR (C6D6): δ ϭ 7.7
(m, 2 H), 7.1Ϫ7.2 (m, 3 H), 3.6 (m, J ϭ 7.5 Hz, 1 H), 2.7 (m, 1
H), 2.2 (m, 1 H), 2.0 (m, 1 H), 1.61 (dt, J ϭ 10.1, J ϭ 7.5 Hz,
CH3), 0.88 (q, J ϭ 7.4 Hz, CH3) ppm.
[12]
[13]
[14]
[15]
ˆ
A. Marinetti, J.-P. Genet, S. Jus, D. Blanc, V. Ratovelomanana-
Vidal, Chem. Eur. J. 1999, 5, 1160Ϫ1165.
ˆ
CCDC-211829 (2c) and -211830 (3c) contain the supplementary
crystallographic data for this paper. These data can be obtained
free of charge at www.ccdc.cam.ac.uk/conts/retrieving.html [or
from the Cambridge Crystallographic Data Centre, 12, Union
Road, Cambridge CB2 1EZ, UK; Fax: (internat.) ϩ44-1223/336-
033; E-mail: deposit@ccdc.cam.ac.uk].
A. Marinetti, S. Jus, J.-P. Genet, L. Ricard, J. Organomet.
Chem. 2001, 624, 162Ϫ166.
ˆ
A. Marinetti, S. Jus, J.-P. Genet, Tetrahedron Lett. 1999, 40,
8365Ϫ8368.
I. D. Gridnev, N. Higashi, K. Asakura, T. Imamoto, J. Am.
Chem. Soc. 2000, 122, 7183Ϫ7194. I. D. Gridnev, M. Yasutake,
N. Higashi, T. Imamoto, J. Am. Chem. Soc. 2001, 123,
5268Ϫ5276. S. Feldgus, C. R. Landis, Organometallics 2001,
20, 2374Ϫ2386.
M. J. Burk, J. E. Feaster, W. A. Nugent, R. L. Harlow, J. Am.
Chem. Soc. 1993, 115, 10125Ϫ10138.
[1]
ˆ
A. Marinetti, F. Labrue, J.-P. Genet, Synlett 1999, 1975Ϫ1977.
[16]
[17]
[2]
[3]
[4]
[5]
[6]
U. Berens, M. J. Burk, A. Gerlach, W. Hems, Angew. Chem.
Int. Ed. 2000, 39, 1981Ϫ1984.
T. Imamoto, J. Watanabe, Y. Wada, H. Masuda, H. Yamada,
H. Tsuruta, S. Matsukawa, K. Yamaguchi, J. Am. Chem. Soc.
1998, 120, 1635Ϫ1636.
D. Heller, H. J. Drexler, J. You, W. Baumann, K. Drauz, H.-P.
Krimmer, A. Börner, Chem. Eur. J. 2002, 8, 5196Ϫ5203.
M. J. Burk, M. F. Gross, Tetrahedron Lett. 1994, 35,
9363Ϫ9366.
O. M. Akotsi, K. Metera, R. D. Reid, R. McDonald, S. H.
Bergens, Chirality 2000, 12, 514Ϫ521.
[18]
[19]
[20]
ˆ
A. Marinetti, S. Jus, F. Labrue, A. Lemarchand, J.-P. Genet, L.
Ricard, Synthesis 2001, 2095Ϫ2104.
A. Marinetti, V. Kruger, F.-X. Buzin, Tetrahedron Lett. 1997,
38, 2947Ϫ2950.
An average value of 95.60° for the bite angle of dppf complexes
has been calculated: P. Dierkes, P. W. N. M. van Leeuwen, J.
Chem. Soc., Dalton Trans. 1999, 1519Ϫ1529.
M. J. Burk, J. E. Feaster, R. L. Harlow, Tetrahedron: Asym-
metry 1991, 2, 569Ϫ592.
[7]
F. Maienza, M. Wörle, P. Steffanut, A. Mezzetti, Organometall-
Received January 24, 2003
2590
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2583Ϫ2590