A highly tunable family of chiral bisphospholanes for Rh-catalyzed enantioselective hydrogenation reactions

Article abstract of DOI:10.1002/chem.200600033

A set of 16 new and closely related bisphospholane ligands have been prepared by using a highly flexible and convergent approach. Each synthesis can be performed on an industrially relevant scale. The bisphosphines differ in the nature of the bridge connecting both phospholane units. Bridges are formed by three-, four-, five- and six-membered heterocyclic or alicyclic rings. Bisphospholanes and their Rh-precatalysts have been investigated by using results of theoretical calculations (DFT) and analytic measurements ( ³¹P and ¹⁰³Rh NMR spectroscopy, X-ray structure analysis). The studies showed that catalysts based on ligands with maleic anhydride or maleimide bridges give constantly superior enantioselectivities in methanol as the solvent. This may account for optimised steric and electronic effects. However, by changing the solvent catalysts with other backbones can give rise to excellent results. This gives proof that simple correlations between steric and electronic properties and results in the enantioselective hydrogenation frequently claimed in literature are not general.

Full text of DOI:10.1002/chem.200600033

FULL PAPER
DOI: 10.1002/chem.200600033
A Highly Tunable Family of Chiral Bisphospholanes for Rh-Catalyzed
Enantioselective Hydrogenation Reactions
AHCTREUNG
Jens Holz,*[a] Odalys Zayas,[a] Haijun Jiao,[a] Wolfgang Baumann,[a]
Anke Spannenberg,[a] Axel Monsees,*[b] Thomas H. Riermeier,[b] Juan Almena,[b]
Renat Kadyrov,[b] and Armin Bçrner*[a, c]
Abstract: A set of 16 new and closely
related bisphospholane ligands have
been prepared by using a highly flexi-
ble and convergent approach. Each
synthesis can be performed on an in-
dustrially relevant scale. The bisphos-
phines differ in the nature of the
bridge connecting both phospholane
units. Bridges are formed by three-,
four-, five- and six-membered heterocy-
clic or alicyclic rings. Bisphospholanes
and their Rh-precatalysts have been in-
vestigated by using results of theoreti-
cal calculations (DFT) and analytic
measurements (31P and 103Rh NMR
maleimide bridges give constantly su-
G
perior enantioselectivities in methanol
as the solvent. This may account for
optimised steric and electronic effects.
However, by changing the solvent cata-
lysts with other backbones can give rise
to excellent results. This gives proof
that simple correlations between steric
and electronic properties and results in
the enantioselective hydrogenation fre-
quently claimed in literature are not
general.
spectroscopy, X-ray structure analysis).
A
The studies showed that catalysts based
on ligands with maleic anhydride or
Keywords: asymmetric catalysis
density functional calculations
·
·
hydrogenation · phosphane ligands ·
rhodium
Introduction
gands bearing a trivalent phosphorus ligating atom for
“soft” metals like rhodium(i), ruthenium(ii), or iridium(i)
play a pivotal role in this area.[2] Over the last 35 years a tre-
mendous number of phosphorus ligands have been de-
Transition-metal-catalyzed enantioselective hydrogenation
has been established as one of the most popular reactions
for the synthesis of optically active compounds in both aca-
demic and industrial environments.[1] In particular chiral li-
A
made in finding new ligand motifs, there remains a lack of a
fundamental understanding of how the enantioselectivity
can be controlled.
Enantioselectivity is the result of several factors. Besides
reaction conditions, geometric or/and electronic parameters
of ligands and catalysts play a pivotal role. A literature
survey revealed that different aspects have been addressed
in the last years. Achiwa was one of the first who suggested
that electron-rich phosphines could be an important requi-
site for achieving high enantioselectivity (respective control
concept).[3] In agreement with this assumption, RajanBabu
and co-workers observed enhanced enantioselectivities in
the Rh-catalyzed hydrogenation of a-acetamido acrylates
with sugar-derived diarylphosphinites as ligands bearing
electron-donating groups in meta or para position.[4] Howev-
er, in strong contrast, Bakos et al.[5a] and some of us[5b] found
with rhodium complexes derived from C2-symmetric 1,3-
bisphosphine ligands that the hydrogenation of itaconic acid
and its dimethyl ester takes benefit from catalysts bearing P
ligands with electron-withdrawing groups.
[a] Dr. J. Holz, Dr. O. Zayas, Dr. H. Jiao, Dr. W. Baumann,
Dr. A. Spannenberg, Prof. Dr. A. Bçrner
Leibniz-Institut für Katalyse e.V. an der Universität Rostock
A.-Einstein-Str. 29a, 18059 Rostock (Germany)
Fax : (+49)381-1281-5000
E-mail: jens.holz@ifok-rostock.de
armin.boerner@ifok-rostock.de
[b] Dr. A. Monsees, Dr. T. H. Riermeier, Dr. J. Almena, Dr. R. Kadyrov
Degussa AG, Degussa Homogeneous Catalysts
Rodenbacher Chaussee 4, Building 097
63457 Hanau-Wolfgang (Germany)
Fax : (+49)6181-59-2417
E-mail: axel.monsees@degussa.com
[c] Prof. Dr. A. Bçrner
Institut für Chemie der Universität Rostock
A.-Einstein-Str. 3a, 18059 Rostock (Germany)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author. It contains the
full experimental details and all hydrogenation results.
Chem. Eur. J. 2006, 12, 5001 – 5013
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5001

Several efforts were also made to find correlations be-
tween geometric features in the catalysts and enantioselec-
tivity. From the stereochemical point of view the quadrant
diagram model is one of the best known concepts.[6] In this
model, occupied and vacant quadrants indicate areas of
maximum and miminum repulsive interactions between
parts of the catalysts and the prochiral substrate. Provided
that the mechanism of the hydrogenation reaction is
known,[6c] this model may help to predict the dominant con-
figuration in the product. Derived from the theory of metal-
reactions.[16] These unique features have stimulated several
groups to construct other chiral phospholanes[17,18] or related
saturated P heterocycles with varying ring sizes, such as
phosphetanes,[19] and phosphinanes.[20]
From industrial as well from academic points of view
there are some serious problems connected with the em-
ployment of this ligand type. The main drawback is the tra-
ditional and commonly used synthesis first suggested by
Burk based on the double nucleophilic substitution of a
chiral cyclic sulfate by two primary phosphine groups
catalyzed regioselective C C coupling reactions, van Leeu-
(Scheme 1a).[13] This methodology is restricted to accessible
ꢀ
wen and co-workers stressed the importance of a natural
bite angle P-Rh-P of 85–958 for efficient enantiodiscrimina-
tion in hydrogenation reactions.[7] A practical example of
geometric fine-tuning of the catalyst was provided by Zhang
et al. with the discovery of the tunephos (tunephos=
(R)-(6,6’-O-(1,3-alkylene)oxylbiphenyl-2,2’-diyl)bis(diphenyl)-
phosphine) family.[8] This allowed the selection among six
electronically similar ligands, but with different dihedral
angles for the efficient hydrogenation of b-keto esters.[8a]
Later on GenÞt et al. gave proof that this transformation
may be also run by electronic effects.[9] Thus, a binap-type
(binap=(1,1’-binaphthalene)-2,2’-diylbis(diphenylphosphine))
ligand with enhanced p-acidic character on the phosphorus
atom, but constant dihedral angle (e.g., 4,4’-bis(2,2-difluoro-
1,3-benzodioxol)-5,5’-diyl]bis(diphenylphosphine)=difluor-
phos) was found to be superior. Recently, in extending the
quadrant diagram model Saito et al. developed a mathemati-
cal approach, in which atoms or groups that can impact
enantioselectivity were localized in a three-dimensional
space.[10] The idea was advocated that broad mapping of cat-
alysts and identification of so-called “sweet spots” could be
helpful to find most effective catalysts for the hydrogenation
of a particular substrate.
Currently, the realization of this demand is hampered by
some serious problems: thus, up to now the synthesis of
chiral P ligands is mainly governed by synthetic limitations.
Therefore, a vast number of strongly varying ligand motifs
has been described in the literature,[2] but families covering
a large set of closely related ligands are very scarce (excep-
tions are for example, permutations of binap[11] and 1-[2-(di-
phenylphosphino)ferrocenyl]ethyldicyclohexylphosphine (jo-
siphos).[12] Only a few such related ligands required for
broad testing are commercially available. In addition, only a
small number of P ligands and their metal complexes have
been comprehensively characterized in terms of steric and
electronic properties.
Scheme 1. Different pathways for the construction of chiral bisphospho-
lanes.
primary phosphines as starting material. Unfortunately,
most primary phosphines are quite expensive. In addition,
strong bases are required for the generation of the nucleo-
philic phosphide anion[21] that may interfere with other func-
tional groups in the molecule. Due to these reasons the
number of synthetically feasible permutations is rather limit-
ed.
Due to the restricted selection of required bisphosphines
a stepwise variation of the steric and electronic properties of
the bridge connecting both phospholane units is difficult to
achieve. In addition, structural modifications, for example,
in the bridge, typically 1,2-phenylene,[22] may break the C2
symmetry of the ligand, commonly considered as a precon-
dition for high stereodiscriminating ability of bisphospho-
lanes catalysts. The amount of structural modifications pub-
lished up to now has lead to dramatic differences in the
physical properties of these ligands (geometry, polarity).
Moreover, only a few of them are commercially available.[23]
Therefore, until now comprehensive testing and reliable
comparisons of catalytic properties were not possible.
Recently, we discovered an alternative process to access a
chiral bisphospholane ligand based on the reaction of 2,3-di-
chloromaleic anhydride with a trimethylsilylphospholane
(TMS-phospholane; Scheme 1b).[24] The new ligand, now al-
ready commercialized, gives enhanced selectivities over Me-
duphos (Me-duphos=1,2-bis(2,5-dimethylphospholano)ben-
zene) in several Rh-catalyzed enantioselective hydrogena-
tion applications.[25,26] Herein, we show that through certain
One of the most prominent bisphosphine ligands for
enantioselective hydrogenation is the bisphospholane
duphos (duphos=1,2-bis(phospholano)benzene), which was
developed by Burk based on Achiwaꢁs idea to create an
electron-rich bisphosphine.[13] This ligand and the related
bisphospholane BPE (BPE=1,2-bis(phospholano)ethane)
A
were successfully applied for the reduction of a broad range
of prochiral olefins and ketones.[13,14] Now, duphos is used
on an industrial scale.[15] Moreover, the ligand could also be
advantageously employed in other enantioselective catalytic
5002
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5001 – 5013

Asymmetric Catalysis
FULL PAPER
modifications this synthetic approach can be easily scaled
up. Moreover, it establishs one of the most comprehensive
sets of closely related P ligands.[27]
For the design of ligands we were inspired by the idea of
ꢀ
creating a selection of bisphospholanes with different C=C
P angles (g; type 1) and varying alkyl groups on the phos-
pholane ring (type 2), respectively (Figure 1).
ꢀ
Different C=C P angles
g
can be easily adjusted by the
variation of the angle b. The
size of the latter should be di-
rectly proportional to the size
of the ring. These parameters
determine the magnitude of the
bite angle a at the rhodium
center as well as the length of
Scheme 2. Small-scale synthesis of TMS-phospholanes 4a,b.
ꢀ
the Rh P bond (d). The latter
An amount of 150 g of P(SiMe3)3 was used for the further
N
should be also dependent on
the basicity of the ligating phos-
phorus atom. It is reasonable to
assume that an increase of the
bite angle and a shortening of
trials. This reacts easily with chiral sulfates 3a,b to afford
TMS-phospholanes 4a,b in good yields. Both compounds
proved to be remarkably stable reagents. The pyrophoric
phospholanes could be distilled and stored under argon for
several months without any tendency to decompose or epi-
merize. For safety issues the compounds were diluted with
THF to 50% and 20%.
Figure 1. Tunable properties of
Rh catalysts with bisphospho-
lane ligands of type 1 (varia-
tion of the bite angle a) and 2
(variation of R); (d=distance
ꢀ
the Rh P bond corresponds to
ꢀ
P Rh).
a decrease of the distance be-
tween the catalytic center and
ligand backbone. This should
Scale-up synthesis of TMS-phospholane 4a and preparation
of bisphospholanes: To provide the material on a commer-
cial scale we developed an alternative synthesis for 4a,
avoiding the hazardous sodium/potassium alloy and the use
of the chiral cyclic sulfate 3a, which is part of a patent
claim.[28] First we tried the synthesis of phenylphospholane 6
by the reaction of (S,S)-2,5-hexanediolbismesylate (5) and
phenylphosphine in the presence of a strong base as report-
ed by Burk[13a,b] and Wilson,[31] respectively (Scheme 3). This
therefore affect the strength of the repulsive interactions be-
tween catalyst and prochiral substrate. Further and more
subtle permutations of geometric and electronic parameters
can be achieved by replacement of single atoms in bridges
of the same ring size.
Results and Discussion
Synthesis of ligands: An important reaction in our conver-
gent approach is the coupling of 2,5-dialkyl-1-trimethylsilyl-
phospholanes with activated 1,2-dichlorides. We had to
modify several steps of our originally reported route[24,28] for
broader applicability, safety issues, and commercial scale-up.
Laboratory-scale synthesis of TMS-phospholanes 4a,b: Our
recently reported approach was based on the synthesis of
trimethylsilylphospholane 4a starting from tris(trimethylsi-
lyl)phosphine (Scheme 2).[24] The latter was synthesized by
Scheme 3. Alternative syntheses of 2,5-dimethyl-1-phenylphospholane 6.
[29]
the condensation of piperidine with PCl3
and a subse-
quent reaction of the product with chlorotrimethylsilane in
the presence of lithium as a strong base.[30a] The major draw-
back of this procedure is the availability of appropriate lithi-
um powder on a commercial scale. Unfortunately, we did
not succeed in an alternative approach by using granulated
procedure gave fast access to chiral 2,5-dimethyl-1-phenyl-
phospholane (6) in 76% yield and could also be used for
laboratory-scale experiments. However, due to the high raw-
material costs and the handling of flammable and highly vol-
atile phenylphosphine, the step was difficult to scale up. We
found dichlorophenylphosphine as a more suitable starting
material; however, the required halogen–lithium exchange
in PhPCl2 is a potential safety risk during scale up. The en-
thalpy (DH) of the strongly exothermic reaction was deter-
mined to be ꢀ195 kcalmolꢀ1. Therefore, the temperature
lithium. However, P(SiMe3)3 could be successfully synthe-
A
sized from a procedure noted in literature by using red
phosphorus and sodium/potassium alloy for the formation of
sodium/potassium phosphide followed by the addition of
chlorotrimethylsilane.[30b] The product was obtained in 60%
yield. This step could be scaled up to 1 mol.
Chem. Eur. J. 2006, 12, 5001 – 5013
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5003

J. Holz, A. Monsees, A. Bçrner et al.
during the addition of lithium should be carefully controlled
(<208C). Best results were obtained by exchanging the sol-
vent from THF to DME 1 h after the addition. Then the
mixture was refluxed for 5 h. The excess lithium was re-
moved by filtration. Subsequently, bismesylate 5 was added
to the solution at low temperature (DH=ꢀ160 kcalmolꢀ1!).
Pure (R,R)-2,5-dimethyl-1-phenylphospholane (6) was iso-
lated with 51% overall yield after distillation.
The next synthetic step relied on the ability to cleave the
ꢀ
P Ph bond, while avoiding epimerization of the chiral cen-
ters. Burk et al. found that the quality of lithium and the ep-
imerization rate were related.[13a,b] They observed the forma-
tion of up to 40–50% of the meso compound. In our pre-
liminary experiments we could obtain the TMS-phospholane
4a in a 4:1 chiral/meso ratio.[24] However, this mixture is not
suitable for an industrial approach. After careful modifica-
tion of reaction conditions, we were able to reduce the for-
mation of compound meso-4a to below 13% as well as the
formation of the undesired secondary phospholane
7
(Table 1). Stirring of lithium in THF for 1 h before starting
Scheme 4. Convergent approach to the synthesis of chiral bisphospho-
lanes.
ꢀ
Table 1. Cleavage of P Ph bond in 6 with lithium.
Synthesis of Rh complexes and characterization of ligands
and precatalysts: Cationic Rh complexes of the type [Rh-
AHCTREUNG
AHCTREUNG
Remarkably, ligand 1a did not give a uniform complex. In
the NMR spectra various species were observed. Due to our
unique construction principle, with the exception of ligands
of type 2, all variations are dependent on the nature of the
backbone bridge. To quantify these differences, the bisphos-
Li
T
t
6
4a
meso-4a
7
Other pro-
[equiv] [8C] [h] [%][a] [%][a] [%][a]
[%][a] ducts[b] [%][a]
1
2
4
8
4
8
3
3
4
4
4
23
23
23
50
50
50
23
50
50
4
2
4
2
4
3
4
3
3
54
12
0
25
52
65
65
72
71
5
3
12
6
6
8
11
1
8
9
9
9
15
18
10
18
11
15
14
13
3[c]
4[c]
5[c]
6[c]
7[c,d]
8[c,e]
9[c,f]
9
pholanes and corresponding Rh(cod) complexes were opti-
A
0
19
2
mized at the B3LYP level of density functional theory with
the LANL2DZ basis set as implemented in the Gaussian98
program.[32] As clearly shown in Figure 2, which details data
of most typical complexes based on ligands of type 1, a de-
crease of the ring size (following the order six-, five-, four-,
0
0
7
73
0
21
6
78
51
0
5
10
[a] Determined by 31P NMR spectroscopy in [D8]THF. [b] Phosphine
oxides, P P coupling products etc. [c] Li was stirred for 1 h in THF
before the reaction was initiated. [d] With two equivalents of DMPU.
[e] With two equivalents of TMEDA. [f] With two equivalents of NMP.
and three-membered rings) corresponds to a larger C=C P
ꢀ
ꢀ
the cleavage gave the best results. Under optimized condi-
tions 4a was obtained in 51% yield and >95% purity after
distillation on a multigram scale.
Subsequent reaction of 2,5-dimethyl- or 2,5-diethyl-TMS-
phospholanes 4a,b with a series of dichlorides 8a–n yielded
the desired bisphospholanes 1a–n and 2a,b bearing three-,
four-, five-, and six-membered hetero- or carbocycles as
bridges (Scheme 4). Most of the required dichlorides are
commercially available or could be synthesized in a few
steps by means of known procedures. Based on this ap-
proach, one of the most comprehensive library of chiral bis-
ꢀ
Figure 2. Calculated correlation between C=C P angle (g) and natural
bite angle (a) of selected bisphospholanes of type 1 and their cationic
A
could be prepared.
[Rh
(cod)(bisphospholane)]+ complexes with different backbones.
A
5004
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5001 – 5013

Asymmetric Catalysis
FULL PAPER
ꢀ
ꢀ
angle. Simultaneously, the natural bite angle P Rh P angle
increases. The largest difference is 3.88 (complexes derived
from 1a and 1n). Apparently, due to the extremely large C=
Recently a discussion on which substituents on 2,5-di-
methyl-phospholane rings are responsible for the high ste-
reodiscriminating ability of the corresponding Rh catalysts
ꢀ
C P angle of the bisphospholane with the three-membered
has been debated in the literature. In Rh–duphos complexes
this feature was attributed to the bulky alkyl groups.[13d,34,35]
In contrast, Orpen and Pringle found that with a related Pt
complex, bearing a ligand in which both phospholane units
were connected by a 1,2-trans-cyclopentane bridge, the ax-
ially placed hydrogen atom in a’-position to the phosphorus
atom rather than the equatorial a-methyl group dominates
the critical diastereofacial repulsive interaction with the pro-
chiral substrate.[36,37]
We have investigated this aspect in our Rh precatalysts.
Inspection of the X-ray structure analyses of Rh complexes
bearing ligands with planar five-membered-ring backbone
bridges (Figure 3) show that the phospholane units adopt a
half-chair conformation. All methyl groups are fixed in
equatorial positions. Density functional theory (DFT) calcu-
lations were performed to iden-
backbone ring in 1a, this ligand tends more to form bridged
dinuclear or polymeric Rh species instead of the desired
mononuclear complex (vide supra). Within the five-mem-
bered ring series, replacing the oxygen atom with nitrogen
and carbon decreases the bite angle. This is clearly due to
an increase in the radii of the concerned atom. The complex
derived from ligand 1l is unique and can be explained by
the large atomic radius of sulfur (element of the second row
of the periodic table).
Crystals suitable for X-ray analysis were obtained by slow
diffusion of diethyl ether into a concentrated solution of the
rhodium complexes in CH2Cl2.[33] X-ray single-crystal struc-
tural analyses established the structures of the cations as
shown in Figure 3; selected bond lengths and intramolecular
tify atomic groups or single
atoms that could be attributed
to maximal repulsive interac-
tions with a metal-coordinated
substrate. In particular, the dis-
tance between the highest lo-
cated hydrogen atom of the
methyl group and the plane
hosting both P atoms (Figure 4,
distance a) and the distance of
a relevant axially situated hy-
drogen atom connected with
the phospholane ring (distance
b) were compared. In all instan-
ces one of the hydrogen atoms
Figure 3. a)–c) Crystal structures of cationic complexes of the type [Rh(cod)(bisphospholane)]BF4 based on li-
gands 1d, 1e, and 1k, respectively. Hydrogen atoms as well as BF4 anions have been omitted for clarity. The
thermal ellipsoids correspond to 30% probability. The bite angles (a) are 87.04(5)8 for 1d, 86.30(5)8 for 1e,
A
of the (free-rotating) methyl
groups projects more strongly
out of the plane than the rele-
vant ring hydrogen atom
and 85.98(5)8 for 1k. Other interatomic distances (Š) and selected intramolecular angles (8) for a) [Rh(cod)-
N
+
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
A
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
C P=8.2. b) [Rh
A
N
ꢀ
ꢀ
ꢀ
ꢀ
dihedral angle P C=C P=4.0. c) [Rh
G
N
ꢀ
ꢀ
ꢀ
C=C P2=119.0(3); dihedral angle P C=C P=2.3.
(Table 2). Therefore, the quad-
rant model attributing vacant
angles are given in the legend. Experimentally determined
angles correlate well with the calculated values.
and occupied diagonal quadrants can be depicted as shown
in Figure 4c.
Figure 4. Methyl groups versus hydrogen atoms as stereodiscriminators a) top view (arrows indicate relevant H atoms); b) side view; c) quadrant model
for the C2-symmetric Rh complex.
Chem. Eur. J. 2006, 12, 5001 – 5013
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5005

J. Holz, A. Monsees, A. Bçrner et al.
Table 2. Computed structural parameters at the B3LYP/LANL2DZ(d)
level.
For the characterization of the s-donor properties of
phosphines, the comparison of phosphorus–selenium cou-
[b]
Rh C[a] [Š]
Rh P (d) [Š]
pling constants J
(77Se,31P) has been suggested.[38] In general,
A
ꢀ
ꢀ
Ligand
a [Š]
b [Š]
aꢀb
1a
1b
1c
1d
1e
1l
1.645
1.650
1.653
1.645
1.652
1.667
1.654
1.688
1.759
1.720
0.845
0.835
0.902
0.925
0.945
0.976
0.964
1.106
0.965
1.036
0.800
0.815
0.751
0.720
0.707
0.691
0.690
0.582
0.794
0.684
3.9703
3.9397
3.9445
3.8945
3.8843
3.8551
3.8679
3.7835
3.8743
3.7592
2.4283
2.4013
2.3969
2.3821
2.3813
2.3708
2.3762
2.3567
2.3643
2.3541
the higher the magnitude of the coupling constant the
higher the s character of the phosphorus lone-pair orbital
and therefore the lower the basicity. A high basicity of the
phosphine causes high electron density on the rhodium. This
effect will strengthen the interactions between the occupied
d orbital of the rhodium and the s* orbital of the hydrogen
molecule, which is important in the overall reaction steps.[3]
Moreover, it supports the complexation of the prochiral ole-
finic substrate due to enhanced transfer of electron density
from the metal d orbitals to the p* orbital of the olefin (p-
back-bonding). More tightly bound p substrates should lead
to shorter metal–substrate bond lengths and thus result in a
greater stereochemical communication between the chiral
ligand and the substrate.
1k
1m
1n
[a] Distance between Rh and CH3 group. [b] See Figure 1.
It is interesting to note, that the difference in length be-
tween a and b becomes smaller (in total by ca. 15%) in the
order of three-, four-, five-, and six-membered-ring back-
bones. In other words, based on purely steric reasons it is
justified to assume that the intrinsic stereodiscriminating
power of Rh complexes with small backbone rings should
be superior in comparison to catalysts with a larger back-
bone ring. The Rh complex derived from ligand (1m) bear-
ing the hexafluorocyclopentene bridge also fits in the quad-
rant model depicted in Figure 4c, albeit the partially saturat-
ed ring adopts an envelope confirmation. Evidently, all Rh–
bisphospholane complexes in which an ethylene bridge con-
nects both phospholane units owe their intrinsic stereodiscri-
minating ability to the a-methyl groups. Only a trans-substi-
tution pattern (e.g., caused by a fully saturated cyclopentane
backbone as investigated by Orpen and Pringle[36a]) may
give rise to a reversed steric arrangement. As a result, rele-
vant axial a’-H atoms can become dominant in this respect.
Our calculations also reveal that the distance between the
rhodium center and the C atoms of the methyl groups is
largest with complexes based on ligands bearing small back-
bone rings. This can be easily explained by a widening of
the bite angle. In terms of the quadrant model, our results
can be summarized in the following way: catalysts discussed
herein with a small bite angle offer the prochiral substrate
rather “narrow” repulsive interactions caused by both
methyl groups (Figure 5a). Their stereodifferentiating poten-
Selected monoselenides were prepared by treating the bis-
A
refluxing chloroform according to Equation (1).
Unfortunately, attempts to prepare the phosphorselenides
of bisphospholane 1l under standard conditions led to the
decomposition of the phosphine.
In Table 3, the values are in order of decreasing J
magnitudes. The largest difference measured for all the bis-
phospholane selenides is 68 Hz.[39] Apparently, the J
(P,Se)
(P,Se)
A
ACHTREUNG
Table 3. J
A
1a
1c
1b
1m
1d
1e
1n
jJ
G
780
771
763
760
750
741
726
[a] Measured in [D]chloroform.
ꢀ
coupling is significantly affected by the size of the C=C P
angle. Bisphospholane 1a is the poorest and 1n is the stron-
gest s donor. As the calculations of Table 2 show, strong
basic phosphines coordinate with a shorter distance d
(Figure 1) to the rhodium center than their electron-poor
analogues. As expected incorporation of fluorine atoms
(1c,n) increases the value of the coupling constants (ca.
10 Hz) in comparison to the parent dicarbonyl compounds
(1b,d). Replacement of the electron pushing NMe group by
Figure 5. Quadrant models of small a) and large b) bite angle catalysts.
tial may be low due to less pronounced differences in the
geometric position of competing atoms and atomic groups
(methyl groups vs. H atoms). In contrast, catalysts with a
large bite angle should have a larger stereodifferentiating
potential. The decisive stereodifferenting methyl groups are
remote, hence a maximal stereodifferentiating interaction
with a prochiral substrate should be expected (Figure 5b).
oxygen in the five-membered ring enhances the J(P,Se) cou-
A
pling (1e vs. 1d).
Summarizing the trends of electronic and steric effects on
the stereodiscriminating coordination of a prochiral sub-
strate, it can be concluded that these oppose each other.
Thus, Rh complexes bearing ligands with a large stereodif-
5006
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5001 – 5013

Asymmetric Catalysis
FULL PAPER
ferentiating potential bind the prochiral substrate less effec-
tively than Rh complexes with bisphosphine ligands that
DFT calculations considering the relative thermodynamic
stability of the interaction between chiral phospholane
ligand and the Rh center support this conclusion. The rela-
tive stability of the complexes has been estimated on the
ꢀ
have a smaller C=C P angle, but higher basicity.
For further characterization of the precatalysts, 103Rh
NMR spectra were recorded. It is known that the d
A
103Rh)
basis of the ligand exchange reaction, such as Rh
A
values for a given array of coordinating atoms are governed
primarily by the coordination geometry, bond angles, and
bond lengths to the rhodium center.[40,41] Thus, all NMR
1d!Rh
C
(1d)+1b!Rh
ACHTREUNG
shifts of [Rh
(cod)(PP)]+ complex ions, in which PP is a bi-
A
dentate ligand forming a five-membered metal chelate, fall
in a narrow range that is characterized by compounds such
as [Rh
C
G
cationic Rh
(1n) complex is computed to be 15.9 kcalmolꢀ1
A
more stable than the complex with the maleic anhydride
backbone (1d). The latter is 5.9 kcalmolꢀ1 more stable than
the Rh complex derived from the squaric acid ligand 1b
(Figure 6).
(dipamp=1,2-ethanediylbis[(2-methoxyphenyl)phenylphos-
phine]), d=ꢀ439 ppm.[42] Calculations for the model com-
pound [Rh
minimum for the 103Rh shift at an P-Rh-P angle of approxi-
mately 858 that evidently permits an optimal orbital over-
lap.[41a,b] Any variation of this angle a, in either direction,
immediately leads to an increase (i.e. shift toward higher
frequency) in the d
of 85.048 (see Figures 1 and 2) it can be concluded that [Rh-
(cod)
G
A
G
an energetic minimum among all investigated precatalysts.
In a series of geometrically closely related complexes the
one with the metal resonance at lowest frequency (lowest d
value) is usually the thermodynamically most stable.[41d]
ꢀ
The Rh P distance (d) was found to have an even more
pronounced effect on the metal NMR shift. A linear corre-
ꢀ1
lation with a coefficient larger than 10000 ppmŠ emerged
from Bühlꢁs investigations.[41a,b] In metal chelates, a and d
are strongly dependent on each other (from the calculated
structural data of the complexes presented in Table 2 fol-
lows a=46.909dꢀ25.35 with r2 =0.9194). In such terms, the
observed shifts (Table 4) are caused by the increasing ring
Figure 6. Calculated stabilities of [Rh
(cod)(bisphospholane)]+ complexes
based on ligands 1b, 1d, and 1n.
A
The differences in stability could be confirmed by a
ligand-exchange experiment (Scheme 5). Thus treatment of
[Rh
A
been prepared by reaction of [Rh(cod)2]BF4 with two equiv-
N
Table 4. 103Rh NMR spectra (d) of precatalysts derived from 1.
Rh-1a Rh-1b Rh-1l Rh-1d Rh-1k Rh-1e Rh-1n
alents of 1b in CH2Cl2, with one or two equimolar amounts
of bisphospholane 1n afforded a stepwise exchange of the
coordinated ligand. The addition of the first equivalent of
1n liberated one diphospholane 1b and gave rise to a mix-
ture consisting of the mixed ligand complex and the homo-
(1n)2 complex. The mixed ligand complex was characterized
in the 31P NMR spectrum in CD2Cl2 by two double doublets
103Rh [ppm][a]
[a] Measured in [D6]acetone.
–
ꢀ246
ꢀ355 ꢀ376
ꢀ377
ꢀ386
ꢀ467
strain in the ligand backbone (Figure 2). Consequently,
at d=56.0 (JRh-P =129 Hz,
(JRh-P=132 Hz). By the addition of a second equivalent of
1n, the bis-homoligand complex [Rh(1n)2]BF4 was formed
JP-P =22.9 Hz) and 98.0 ppm
ligand 1b (based on squaric acid), which forms the complex
with the highest d
chelate (vide infra).
Remarkably enough, the already discussed J
G
103Rh) value, is also the least stable[41d]
ACHTREUNG
which gave a doublet at d=94.7 ppm (JRh-P =130 Hz) in the
(P,Se) cou-
31P NMR spectrum.
pling constants are also predictive for the steric effects of
the rhodium complexes. The trend is clear-cut: an increasing
Results of the enantioselective hydrogenation reactions: The
new precatalysts were tested in the hydrogenation of bench-
mark substrates such as, methyl (Z)-N-acetamido cinna-
mate[43] and dimethyl itaconate, but also b-dehydroamino
acid precursors,[44] in order to assess its stereodiscriminating
ability for a broad range of chiral substituted olefins. Hydro-
ꢀ
jJ(P,Se)j is directly related to increasing P Rh distance d
A
and also bite angle a. Only the fluorinated system m be-
haves somewhat differently. In this case the electronegativity
of six fluorine atoms might override the otherwise dominant
steric influence.
Chem. Eur. J. 2006, 12, 5001 – 5013
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5007

J. Holz, A. Monsees, A. Bçrner et al.
Scheme 5. Ligand exchange reaction due to different complex stabilities.
genations were performed with a molar precatalyst/substrate
ratio of 1:100 at 1 bar hydrogen pressure at 258C in CH2Cl2,
THF, and MeOH as solvent.
In Figures 7–12 selected results are listed in order of in-
creasing s-donor capacity of the phospholanes and decreas-
ing bite angle in the precatalyst.[45] In several attempts the
sulfur-containing ligand 1l blocked the hydrogenation. The
results obtained with ligand 1a varied over a broad range.
Therefore, results were only reported when constant values
were observed within two runs.
In general, differences in the enantioselectivities by up to
Dee=90% were noted. With the same substrate all catalysts
induced the same configuration in the product. Deviations
of the enantioselectivity in the hydrogenation of methyl (Z)-
N-acetamido cinnamate are less pronounced than with di-
methyl itaconate as substrate (Figure 7). Apparently for the
assessment of the intrinsic stereodiscriminating ability of
this type of catalyst dimethyl itaconate as benchmark sub-
strate is better suited than the former. In MeOH or THF as
solvent for most substrates investigated herein, catalysts
with a maleic anhydride or maleimide backbone gave the
best results. Replacement of these ligands by more electron-
deficient or electron-rich phosphines may negatively affect
the enantioselectivity. In CH2Cl2 as solvent these sometimes
dramatic differences are leveled and in general superior
enantioselectivities result. The differences in the level of
enantioselectivity observed when (E)- or (Z)-b-dehydroami-
no acid esters were subjected to the hydrogenation have
been already described (Figures 8 and 9).[24,44]
In his pioneering work, Burk found that in some cases the
replacement of methyl groups in 2,5-position of the phos-
pholane unit by ethyl groups (Et-duphos=1,2-bis(2,5-dieth-
ylphospholano)benzene) can increase the enantioselectivity
of the hydrogenation.[13d] As shown in Figure 10 this effect
may hold sometimes also for hydrogenations with ligands
considered herein. However, a general tendency can not be
deduced.
As cited in the introduction, electronic fine tuning of li-
gands has been proposed several times as a powerful tool
for the increase of the enantioselectivity of chiral hydroge-
nation catalysts.[4,5,46] Our results observed with nonfluori-
nated and fluorinated ligands show that the reaction can
indeed be significantly influenced by electronic changes in
the backbone of the ligand. However, in general, no clear
correlation such as that found by RajanBabu,[4] Bakos,[5a]
and us[5b] exist in the series of ligands investigated herein
(Figures 11 and 12).
Figure 7. Results of the enantioselective hydrogenation of methyl (Z)-N-
acetamido cinnamate (top) and dimethyl itaconate (bottom) with ligands
of type 1. Values are listed in order of decreasing bite angle in the preca-
talysts.
Table 5 summarizes the best results of all 360 hydrogena-
tion runs. The data clearly show that our set of 16 ligands
allows each substrate to find the best conditions. In those in-
stances in which a ligand bearing a maleimide backbone was
identified as the most effective catalysts further fine-tuning
was possible by variation of the N-alkyl group. Further im-
provements could be achieved by variation of the chiral N-
phenylethyl unit (1i,j) taking benefit from the application of
the matched versus mismatched diastereomeric catalysts. In
general, our results fulfill the recent assertion of Saito et al.
who proposed that each substrate needs its own catalyst.[10]
However, it should be noted that in this model purely steric
reasons have been attributed to play a pivotal role. Our re-
5008
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5001 – 5013

Asymmetric Catalysis
FULL PAPER
Figure 8. Results of the enantioselective hydrogenation of methyl (E)-b-
acetamido acrylate (top) and methyl (Z)-b-acetamido acrylate (bottom)
with ligands of type 1. Values are listed in order of decreasing bite angle
in the precatalysts.
Figure 9. Results of the enantioselective hydrogenation of ethyl (E)-2-
acetamido-3-methylbutenoate (top) and ethyl (Z)-2-acetamido-3-methyl-
butenoate (bottom) with ligands of type 1. Values are listed in order of
decreasing bite angle in the precatalysts.
Figure 10. 2,5-Methyl- versus ethyl-substituted phospholanes as ligands (1d vs. 2a)
Chem. Eur. J. 2006, 12, 5001 – 5013
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5009

J. Holz, A. Monsees, A. Bçrner et al.
Figure 11. Influence of electronic effects on the enantioselective hydroge-
nations with bisphospholanes of type 1 bearing a four-membered back-
bone bridge.
Figure 12. Influence of electronic effects in enantioselective hydrogena-
tions with bisphospholanes of type 1 with a five-membered backbone
bridge.
Conclusion
Table 5. Best results of all hydrogenations.
Substrate
In conclusion, by a simple and signficantly improved conver-
gent strategy, one of the most comprehensive family of
chiral bisphosphine ligands have been synthesized. This
methodology, which can be easily scaled up for industrial re-
quirements, was used for the construction of closely related
bisphospholanes, differing in the size of the natural bite
angles and s-donor properties of the phosphine. These fea-
tures are caused by three-, four-, five-, and six-membered
heterocyclic or alicyclic rings connecting both phospholane
units. Replacement of methyl groups by ethyl groups in the
phospholane opened up avenues to further modifications.
Bisphospholanes and their Rh precatalysts have been inves-
tigated by means of density functional theory (DFT) calcu-
lations and analytic measurements (31P, 103Rh NMR spectro-
scopy, X-ray structure analysis). The catalytic studies
showed that ligands with maleic anhydride or maleimide
bridges give consistently superior enantioselectivities in
polar solvents. It is reasonable to assume that these Rh com-
R1
R2
R3
R4
Ligand
Solvent
ee [%]
COOMe
CH2COOMe
COOMe
H
COOBn
H
NHAc
COOMe
H
COOMe
H
COOBn
H
COOEt
H
H
Ph
H
2b
1e
1e
1b
1b
1d
1d
1j
THF
99.0
99.1
98.6
93.6
99.5
89.9
99.7
89.6
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
MeOH
CH2Cl2
CH2Cl2
Me
Me
Me
Me
iPr
iPr
NHAc
NHAc
NHAc
NHAc
NHAc
NHAc
COOEt
H
sults reveal, however, that also electronic effects may be of
significant importance. Sometimes dramatic differences
were noted only by a slight change of the bite angle and/or
the basicity of the ligand. Moreover, by a change of the sol-
vent other catalysts could give rise to superior results.
5010
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5001 – 5013

Asymmetric Catalysis
FULL PAPER
by distillation (b.p.1 80–858C) to give a colorless liquid of 1 (150.3 g,
60%). 1H NMR [D6]benzene: d=0.06 ppm (d, 27H); 31P NMR
[D6]benzene: d=ꢀ251.6 ppm (caution: pyrophoric and moisture sensi-
tive).
plexes represent unique structures among the whole family
of catalysts tested for which opposite trends of beneficial
geometric and electronic effects are optimum. Interestingly,
these precatalysts are not thermodynamically the most
stable ones. By changing the solvent, other catalysts could
provide superior results. It seems that simple steric or elec-
tronic models are limited in the prediction of enantioselec-
tivities. They are not able to describe the complexity of
enantioselective hydrogenation reactions. Evidently, correla-
tions between geometric or electronic parameters in ligands
and precatalysts, respectively, and hydrogenation results, fre-
quently quoted in the literature and derived from small sets
of ligands and reactions conditions, are not general. They
can be easily overridden by other effects (e.g. solvent ef-
fects). At the current level of investigations a combination
of a large family of related and scalable ligands with the fa-
cility of high through-put screening[47] seems to be the best
guarantee for the fast identification of most effective hydro-
genation conditions. This precondition is fulfilled by the new
and large family of tuneable bisphospholanes described
herein. Further work will address the functionalization of
appropriate ligands for reactions under multiphase condi-
tions and the ongoing search for relationships between the
structure of catalysts or catalytic intermediates and the re-
sults in enantioselective hydrogenation.
Large-scale synthesis of (S,S)-2,5-hexanediolbismethanesulfonate (5): In
a 2 L flask (S,S)-2,5-hexanediol (120 g, 1.015 mol) was suspended in tol-
uene (450 mL), and triethylamine (353.5 mL, 2.54 mol) was added. The
suspension was cooled to 08C and a solution of methanesulfonylchloride
(255.9 g, 2.23 mol) in toluene (50 mL) was slowly added. During the addi-
tion the temperature was kept below 258C. The suspension changed its
color to yellow/brownish. The solution was stirred for 1 h. The reaction
mixture was hydrolyzed by addition of water (250 mL). The organic layer
was separated and the aqueous layer extracted with toluene (200 mL).
The organic layers were washed with brine and dried with magnesium
sulfate. Toluene was removed until the solution contained about 30% of
5. This solution was used for subsequent reactions. 1H NMR
[D]chloroform: d=4.85 (m, 2H; CHO), 3.02 (s, 6H; CH3), 1.80 (m, 4H;
CH2), 1.44 (d, J=6.3 Hz, 6H; CH3). Data were in accordance with those
reported in reference [13b].
A
(44.4 g, 6.4 mol) and THF (400 mL) was stirred at 58C. A solution of di-
chlorophenylphosphine (143.2 g, 0.8 mol) in THF (220 mL) was added
dropwise over a period of 90 min, while maintaining the temperature be-
tween 5–178C. The resulting orange suspension was warmed up to 238C
and stirred for 1 h. The solvent was removed under reduced pressure,
and DME (600 mL) was added to the semisolid residue. The suspension
was refluxed for 5 h and cooled to 238C. The excess lithium was removed
by transferring the suspension into another flask. Lithium was washed
with DME (50 mL). The suspension was cooled to ꢀ158C and the solu-
tion of 5 (182.2 g, 0.66 mol) in toluene (200 mL) was added over a period
of 1 h. During the addition the temperature was kept between ꢀ15 and
ꢀ28C. The solution was allowed to warm up to 238C and stored for 16 h.
The solvent was removed under reduced pressure and heptane (500 mL)
was added. The suspension was filtered and the solid washed with hep-
tane (2”300 mL). The solvent was removed under reduced pressure and
the crude product was purified by distillation (b.p.=125–1308C/30 mbar)
to give 6 as a colorless liquid (65.2 g, 0.34 mol, 51%). NMR data were in
accordance with those reported in reference [13b].
Experimental Section
Solvents were dried and freshly distilled under argon before use. All re-
actions were performed under an argon atmosphere by using standard
Schlenk techniques.
NMR spectra were recorded at a Bruker ARX 400 spectrometer at the
following frequencies: 400.13 MHz (1H), 100.63 MHz (13C), 161.98 MHz
(31P), 282 MHz (19F). Chemical shifts of 1H and 13C NMR spectra are re-
ported in ppm downfield from TMS as internal standard. Chemical shifts
of 31P NMR spectra are referred to H3PO4 as external standard. Signals
are quoted as s (singlet), d (doublet), br (broad) and m (multiplet). 103Rh
NMR spectroscopy was performed with the same spectrometer (B0 =
9.4 T), equipped with a commercial triple-resonance probe (doubly-tuned
1H/31P coil with a surrounding broadband coil) with p/2 pulse width for
Large-scale synthesis of (R,R)-2,5-dimethyl-1-trimethylsilylphospholane
(4a): A suspension of lithium (76.0 g, 11.0 mol) was stirred in THF (2 L)
for 1 h. The suspension was cooled to ꢀ58C and a solution of 6 (531.0 g,
2.76 mol) in THF (400 mL) was added over a period of 1 h, while main-
taining the temperature between ꢀ5 and 108C. The reaction mixture was
warmed to 238C and stirred for additional 4 h. The remaining lithium
was removed by filtration and washed with THF (100 mL). The filtrate
was cooled to ꢀ108C and trimethylchlorosilane (764 mL, 6.07 mol) in
THF (300 mL) was added over a period of 1 h. The temperature was
held between ꢀ5 and 258C. After the complete addition the solution was
stirred for 1 h at 238C. About 60% of the solvent was removed under re-
duced pressure and the remaining salts were removed by filtration. Sol-
vents were removed and the crude product was purified by distillation. It
was important to remove the byproduct trimethylphenylsilane completely
from 4a (b.p. =938C/20 mbar). Compound 4a was obtained as a colorless
liquid in 51% yield (265.2 g).
103Rh of 29.5 ms. Chemical shifts d 103Rh) are given in ppm relative to X=
(
G
3.16 MHz,[40a] positive signs indicating low-field shifts, and were deter-
mined by inverse detected (four-pulse HMQC) triple-resonance experi-
ments 31P, 103Rh{1H};[40b,c] each determination was carried out at least twice
with variation of the 103Rh frequency and the t1 increment to ensure that
the signals in the F1 dimension were not folded. The temperature was
about 297 to 298 K.
Large-scale synthesis of tris(trimethylsilyl)phosphine: Na/K alloy was
prepared by carefully melting of potassium (51.0 g, 1.3 mol) and sodium
(39.0 g, 1.7 mol) in a 4 L three-necked round-bottomed flask equipped
with a mechanical stirrer, dropping funnel, and reflux condenser. Dry
DME (2 L) was added and the mixture was then heated under reflux for
1 h. The slightly purple mixture was cooled to 238C and red phosphorus
(31.0 g, 1 mol) was added. The mixture was heated to reflux for 18 h in
total (for safety reasons the mixture was cooled down to 238C). Chloro-
trimethylsilane (420 mL, 3.3 mol) in DME (500 mL) was added dropwise
to the black solution and the temperature was allowed to rise to gentle
reflux. The mixture was refluxed for an additional hour and cooled to
238C. The resulting gray suspension was filtered through dried Celite
under Ar. The residue was washed with pentane (2”100 mL). The sol-
vent was removed at ambient pressure. The crude product was purified
Large-scale synthesis of 3,4-bis-[(R,R)-(2,5-dimethylphospholan-1-yl]ma-
leic anhydride (1d):
A solution of TMS-phospholane 4a (51.1 g,
0.270 mol, 2 equiv) in absolute Et2O (50 mL) was added drop-wise to a
stirred solution of dichloromaleic anhydride (22.5 g, 0.135 mmol) in Et2O
(200 mL) over a period of 30 min at 08C under Ar (a dark-red color ap-
peared immediately after first few drops). After stirring for 15 min at
08C and an additional 30 min at room temperature, the reaction mixture
was cooled to ꢀ788C and left for 48 h to crystallize. The mother liquor
was removed and the first fraction of dark-brownish crystals of 1d were
isolated and dried under vacuum (25.0 g, 57%). The mother liquor was
evaporated again to dryness and a second fraction of 1d could be isolated
(16.4 g, 37%).
Chem. Eur. J. 2006, 12, 5001 – 5013
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5011

J. Holz, A. Monsees, A. Bçrner et al.
Chem. Int. Ed. 1998, 37, 1931–1933; M. J. Burk, G. Casy, N. B. John-
son, J. Org. Chem. 1998, 63, 6084–6085;
Acknowledgements
[15] M. J. Burk, M. F. Gross, T. G. P. Harper, C. S. Kalberg, J. R. Lee, J. P.
Martinez, Pure Appl. Chem. 1996, 68, 37–44; “The Application of
DuPHOS Rhodium(i) Catalysts for Commercial Scale Asymmetric
Hydrogenation”: C. J. Cobley, N. B. Johnson, I. C. Lennon, R.
McCague, J. A. Ramsden, A. Zanotti-Gerosa, in Asymmetric Cataly-
sis on Industrial Scale (Eds.: H.-U. Blaser, E. Schmidt) Wiley-VCH,
Weinheim, 2004.
The authors are grateful for the financial support provided by Degussa
AG (Hanau) and the Fonds der Chemischen Industrie. It is a pleasure
for us to acknowledge highly skilled technical assistance by G. Wenzel
and H. Borgwaldt. We are thankful for the valuable advice from Prof.
Dr. K. Drauz (Degussa AG) and Prof. Dr. M. Beller (IfOK Rostock).
We thank Dr. C. Fischer and S. Buchholz for the GC and HPLC analy-
ses.
[16] A. Yanagisawa, H. Nakashima, A. Ishiba, H. Yamamoto, J. Am.
Chem. Soc. 1996, 118, 4723–4724; R. W. Barnhart, D. A. McMorran,
B. B. Bosnich, Chem. Commun. 1997, 589–590; A. A. Boezio, J. Pyt-
kowicz, A. CôtØ, A. B. Charette, J. Am. Chem. Soc. 2003, 125,
14260–14261; A. CôtØ, A. A. Boezio, A. B. Charette, Angew.
Chem. 2004, 116, 6687–6690; Angew. Chem. Int. Ed. 2004, 43, 6525–
6528; C. Korff, G. Helmchen, Chem. Commun. 2004, 530–531.
[17] For an excellent review, see T. P. Clark, C. R. Landis, Tetrahedron:
Asymmetry 2004, 15, 2123–2137, and references therein.
[18] For more recent works, see: a) G. Hoge, J. Am. Chem. Soc. 2004,
126, 9920–9921; b) G. MalaisØ, S. Ramdeehul, J. A. Osborn, L.
Barloy, N. Kyrisakas, R. Graff, Eur. J. Inorg. Chem. 2004, 3987–
4001; c) A. Zhang, T. V. RajanBabu, Org. Lett. 2004, 6, 3159–3161;
d) S. Basra, J. G. de Vries, D. J. Hyett, G. Harrison, K. M. Heslop, A.
G Orpen, P. G. Pringle, K. von der Luhe, Dalton Trans. 2004, 1901–
1905; e) G. Hoge, B. Samas, Tetrahedron: Asymmetry 2004, 15,
2155–2157; f) for chiral 3,4-diazoaphospholanes, see: T. P. Clark,
C. R. Landis, S. L. Freed, J. Klosin, K. A. Abboud, J. Am. Chem.
Soc. 2005, 127, 5040–5042, and references therein.
[1] a) J. M. Brown in Comprehensive Asymmetric Catalysis,Vol. I (Eds.:
E. N. Jacobsen, A. Pfaltz, H. Yamamoto) Springer, Berlin, 1999,
pp. 121–182; b) T. Ohkuma, M. Kitamura, R. Noyori In Catalytic
Asymmetric Synthesis (Ed.: I. Ojima), Wiley-VCH, New York, 2000,
pp. 1–110; c) H.-U. Blaser, B. Pugin, F. Spindler, J. Mol. Catal. A
2005, 231, 1–20.
[2] For recent reviews see: a) M. DiØguez, O. Pàmies, A. Ruiz, Y. Diaz,
S. Castillón, C. Claver, Coord. Chem. Rev. 2004, 248, 2165–2192;
b) T. T.-L. Au-Yueng, A. S. C. Chan, Coord. Chem. Rev. 2004, 248,
2151–2164; c) P. Barbara, C. Bianchini, G. Giambastiani, S. L. Pari-
sel, Coord. Chem. Rev. 2004, 248, 2131–2150; d) W. Tang, X. Zhang,
Chem. Rev. 2003, 103, 3029–3069.
[3] Review: K. Inoguchi, S. Sakuraba, K. Achiwa, Synlett 1992, 169–
178.
[4] a) T. V. RajanBabu, T. A. Ayers, G. A. Halliday, K. K. You, J. C.
Calabrese, J. Org. Chem. 1997, 62, 6012–6028; b) T. V. RajanBabu,
T. A. Ayers, A. L. Casalnuovo, J. Am. Chem. Soc. 1994, 116, 4101–
4102; c) T. V. RajanBabu, B. Radetich, K. K. You, T. A. Ayers, A. L.
Casalnuovo, J. C. Calabrese, J. Org. Chem. 1999, 64, 3429–3447.
[5] a) Z. Herseczki, I. Gergely, C. Hegedüs, À. Szçllçsy, J. Bakos, Tetra-
hedron: Asymmetry 2004, 15, 1673–1676; b) N. V. Dubrovina, V. I.
Tararov, A. Monsees, A. Spannenberg, I. D. Kostas, A. Bçrner, Tet-
rahedron: Asymmetry 2005, 16, 3640–3649.
[19] A. Marinetti, D. Carmichael, Chem. Rev. 2002, 102, 201–230.
[20] a) M. Ostermeier, J. Priess, G. Helmchen, Angew. Chem. 2002, 114,
625–628; Angew. Chem. Int. Ed. 2002, 41, 612–614; b) Y. Yan, X.
Zhang, Tetrahedron Lett. 2006, 47, 1567–1569.
[21] M. J. Burk, (E. I. Du Pont de Nemours and Company), EP 0592552,
US 5329015 (02.07.1991).
[22] For some other “backbones” of bisphospholanes, see for example,
a) ferrocene: M. J. Burk, M. F. Gross, Tetrahedron Lett. 1994, 35,
9363–9366; b) xanthene: P. C. J. Kamer, P. W. N. M. van Leeuwen,
J. N. H. Reek, Acc. Chem. Res. 2001, 34, 895–904; c) phenoxathiin:
P. Dierkes, S. Ramdeehul, L. Barloy, A. De Cian, J. Fischer, P. C. J.
Kamer, P. W. N. M. van Leeuwen, J. A. Osborn, Angew. Chem. 1998,
110, 3299–3301; Angew. Chem. Int. Ed. 1998, 37, 3116–3118;
d) Benzo(b)thiophene: U. Berens, (Solvias AG) WO 03/031456 A2
2003; e) Thiophene: T. Benincori, T. Pilati, S. Rizzo, F. Sannicolo,
M. J. Burk, L. de Ferra, E. Ullucci, O. Piccolo, J. Org. Chem. 2005,
70, 5436–5441.
[23] DuPHOS (ligands are commercialized by DowPharma), TangPHOS
(W. Tang, X. Zhang, Angew. Chem. 2002, 114, 1682–1684; Angew.
Chem. Int. Ed. 2002, 41, 1612–1614; ligand is commercialized by
ChiralQuest), Butiphane (Ref. 22d; ligand is commercialized by Sol-
vias AG) and RoPHOS (J. Holz, M. Quirmbach, U. Schmidt, D.
Heller, R. Stürmer, A. Bçrner, J. Org. Chem. 1998, 63, 8031–8034;
ligands are commercialized by Solvias AG). Some other bisphospho-
lanes derived from d-mannitol (W. Li, Z. Zhang, D. Xiao, X. Zhang,
Tetrahedron Lett. 1999, 40, 6701–6704; W. Li, Z. Zhang, D. Xiao, X.
Zhang, J. Org. Chem. 2000, 65, 3489–3496; ligands are commercial-
ized by ChiralQuest).
[6] a) B. D. Vineyard, W. S. Knowles, M. J. Sabacky, G. L. Bachman,
D. J. Weinkauff, J. Am. Chem. Soc. 1977, 99, 5946–5952; b) W. S.
Knowles, Acc. Chem. Res. 1983, 16, 106–112; c) I. D. Gridnev, T.
Imamoto, Acc. Chem. Res. 2004, 37, 633–644, and references there-
in.
[7] a) P. W. N. M. van Leeuwen, P. C. J. Kamer, J. N. H. Reek, P. Dier-
kes, Chem. Rev. 2000, 100, 2741–2769; b) P. Dierkes, P. W. N. M. van
Leeuwen, J. Chem. Soc. Dalton Trans. 1999, 1519–1529.
[8] a) Z. Zhang, H. Qian, J. Longmire, X. Zhang, J. Org. Chem. 2000,
65, 6223–6226; b) C.-J. Wang, X. Sun, X. Zhang, Angew. Chem.
2005, 117, 5013–5015; Angew. Chem. Int. Ed. 2005, 44, 4933–4935.
[9] S. Jeulin, S. Duprat de Paule, V. Ratovelomanana-Vidal, J.-P. GenÞt,
N. Champion, P. Dellis, Angew. Chem. 2004, 116, 324–329; Angew.
Chem. Int. Ed. 2004, 43, 320–325.
[10] H. Shimizu, T. Ishizaki, T. Fujiwara, T. Saito, Tetrahedron: Asymme-
try 2004, 15, 2169–2172.
[11] A. Miyashita, A. Yasuda, H. Takaya, K. Toriumi, T. Ito, T. Souchi,
R. Noyori, J. Am. Chem. Soc. 1980, 102, 7932–7934; X. Wan, Y.
Sun, Y. Luo, D. Li, Z. Zhang, J. Org. Chem. 2005, 70, 1070–1072; H.
Shimizu, I. Nagasaki, T. Saito, Tetrahedron 2005, 61, 5402–5432, and
references therein.
[12] A. Togni, C. Breutel, A. Schnyder, F. Spindler, H. Landert, A.
Tijani, J. Am. Chem. Soc. 1994, 116, 4062–4066; H.-U. Blaser, Adv.
Synth. Catal. 2002, 344, 17–31; F. Spindler, C. Malan, M. Lotz, M.
Kesselgruber, U. Pittelkow, A. Rivas-Nass, O. Briel; H.-U. Blaser,
Tetrahedron: Asymmetry 2004, 15, 2299–2306.
[24] J. Holz, A. Monsees, H. Jiao, J. You, I. V. Komarov, C. Fischer, K.
Drauz, A. Bçrner, J. Org. Chem. 2003, 68, 1701–1707.
[25] J. Almena, A. Monsees, R. Kadyrov, T. H. Riermeier, B. Gotov, J.
Holz, A. Bçrner, Adv. Synth. Catal. 2004, 346, 1263–1266.
[26] For a first industrial application, see: F. Klingler, M. Steigerwald, R.
Ehlenz, (Boehringer Ingelheim Pharma GmbH & Co. KG) DE
10313118A1, 2004.
[27] The whole family of new bisphospholanes are now summarized
under the trade name catASiumM. Ligands and metal complexes
are available from Degussa Homogeneous Catalysts.
[28] The new route also circumvents the claims of a Du Pont de Nem-
ours and Co. patent: M. J. Burk, US 5171892 (15.12.1992), which re-
stricts the industrial application of several bisphospholane ligands.
[13] a) M. J. Burk, J. E. Feaster, R. L. Harlow, Organometallics 1990, 9,
2653–2655; b) M. J. Burk, J. E. Feaster, R. L. Harlow, Tetrahedron:
Asymmetry 1991, 2, 569–592; c) M. J. Burk, J. Am. Chem. Soc. 1991,
113, 8518–8519; d) M. J. Burk, J. E. Feaster, W. A. Nugent, R. L.
Harlow, J. Am. Chem. Soc. 1993, 115, 10125–10138.
[14] M. J. Burk, M. F. Gross, J. P. Martinez, J. Am. Chem. Soc. 1995, 117,
9375–9376; M. J. Burk, T. G. P. Harper, C. S. Kalberg, J. Am. Chem.
Soc. 1995, 117, 4423–4424; M. J. Burk, F. Bienewald, M. Harris, A.
Zanotti-Gerosa, Angew. Chem. 1998, 110, 2034–2037; Angew.
5012
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 5001 – 5013

Asymmetric Catalysis
FULL PAPER
[29] G. Märkl, B. Alig, J. Organomet. Chem. 1984, 273, 1–29.
[30] a) E. Niecke, H. Westermann, H. Synthesis 1987, 330; E. V. Popova,
I. I. Patsanovskii, M. V. Livantsov, E. A. Ishmaeva, Zh. Obshch.
Khim. 1996, 66, 1406; Russ. J. Gen. Chem. 1996, 66, 1369–1370;
b) H. H. Karsch, F. Bienlein, T. Rupprich, F. Uhlig, E. Herrmann,
M. Scheer in Synthetic Methods of Organometallic and Inorganic
Chemistry,Vol. 3 (Eds.: W. A. Herrmann, G. Brauer), Thieme, Stutt-
gart, 1996, pp. 58–65.
1986, 108, 5622–5624; c) R. Benn, A. Rufinska, Magn. Reson.
Chem. 1988, 26, 895–902.
[41] For systematic investigations in this area, see: a) W. Leitner, M.
Bühl, R. Fornika, C. Six, W. Baumann, E. Dinjus, M- Kessler, C.
Krüger, A. Rufinska, Organometallics 1999, 18, 1196–1296; b) J. G.
Donkervoort, M. Bühl, J. M. Ernsting, C. J. Elsevier, Eur. J. Inorg.
Chem. 1999, 27–33; c) J. M. Ernsting, C. J. Elsevier, W. G. J. de
Lange, K. Timmer, Magn. Reson. Chem. 1991, 29, S118–124; d) L.
Öhrstrçm, Comments Inorg. Chem. 1996, 18, 305–323.
[31] S. R. Wilson, A. Pasternak, Synlett 1990, 199–200.
[32] Gaussian98 (RevisionA.11.3), M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakr-
zewski, J. A. Montgomery, R. E. Stratmann, J. C. Burant, S. Dap-
prich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O.
Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C.
Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y.
Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Ra-
ghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov,
G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Na-
nayakkara, C. Gonzalez, M. Challacombe, P. M. W. Gill, B. G. John-
son, W. Chen, M. W. Wong, J. L. Andres, M. Head-Gordon, E. S.
Replogle, J. A. Pople, Gaussian, Inc., Pittsburgh, PA, 1998..
[33] CCDC-294536 (1d), CCDC-294537 (1e), and CCDC-294538 (1k)
contain the supplementary crystallographic data for this paper.
These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.uk/data_request/cif.
[34] For a discussion of this issue with Rh–bisphosphetane complexes,
see: A. Marinetti, J.-P. GenÞt, S. Jus, D. Blanc, V. Ratovelomanana-
Vidal, Chem. Eur. J. 1999, 5, 1160–1165; A. Marinetti, S. Jus, J.-P.
GenÞt, Tetrahedron Lett. 1999, 40, 8365–8368.
[35] S. Feldgus, C. R. Landis, J. Am. Chem. Soc. 2000, 122, 12714–12727;
Angew. Chem. 2000, 112, 2985–2988; Angew. Chem. Int. Ed. 2000,
39, 2863–2866; C. R. Landis, P. Hilfenhaus, S. Feldgus, J. Am. Chem.
Soc. 1999, 121, 8741–8754.
[36] a) E. Fernandez, A. Gillon, K. Heslop, E. Horwood, D. J. Hyett,
A. G. Orpen, P. G. Pringle, Chem. Commun. 2000, 1663–1664;
b) see also L. Dahlenburg, Eur. J. Inorg. Chem. 2003, 2733–2747.
[37] In enantioselective hydrogenation reactions, products with the oppo-
site configuration were formed in comparison to the reduction with
an analogue Rh–Me-duphos catalyst.
[42] W. Baumann, D. Heller, unpublished results. The complex [Rh(Me-
duphos)
(1n)]BF4 gives a resonance at d=ꢀ465 ppm (measured in metha-
nol).
[43] U. Nagel, J. Albrecht, Top. Catal. 1998, 5, 3–23.
A
G
AHCTREUNG
[44] H.-P. Wu, G. Hoge, Org. Lett. 2004, 6, 3645–3647; T. Jerphagnon, J.-
L. Renaud, P. Demonchaux, A. Ferreira, C. Bruneau, Adv. Synth.
Catal. 2004, 346, 33–36; N. V. Dubrovina, V. I. Tararov, A. Monsees,
R. Kadyrov, C. Fischer, A. Bçrner, Tetrahedron: Asymmetry 2003,
14, 2739–2745; J. Holz, A. Monsees, H. Jiao, J. You, I. V. Komarov,
C. Fischer, K. Drauz, A. Bçrner, J. Org. Chem. 2003, 68, 1701–1707;
W. Tang, W. Wang, Y. Chi, X. Zhang, Angew. Chem. 2003, 115,
3633–3635; Angew. Chem. Int. Ed. 2003, 42, 3509–3511; H. J. Drex-
ler, J. You, S. Zhang, C. Fischer, W. Baumann, A. Spannenberg, D.
Heller, Org. Process Res. Dev. 2003, 7, 355–361; S.-G. Lee, Y.
Zhang, Org. Lett. 2002, 4, 2429; D. PeÇa, A. J. Minnaard, J. G. de
Vries, B. L. Feringa, J. Am. Chem. Soc. 2002, 124, 14552–14553; W.
Tang, X. Zhang, Org. Lett. 2002, 4, 4159–4161; I. V. Komarov, A.
Monsees, R. Kadyrov, C. Fischer, U. Schmidt, A. Bçrner, Tetrahe-
dron: Asymmetry 2002, 13, 1615–1620; M. Yasutake, I. D. Gridnev,
N. Higashi, T. Imamoto, Org. Lett. 2001, 3, 1701–1704; G. Zhu, Z.
Chen, X. Zhang, J. Org. Chem. 1999, 64, 6907–6910.
[45] For all and more detailed results see the Supporting Information.
[46] K. Mashima, K.-h. Kusano, N. Sato, Y.-i. Matsumura, K. Nozaki, H.
Kumobayashi, N. Sayo, Y. Hori, T. Ishizaki, S. Akutagwa, H.
Takaya, J. Org. Chem. 1994, 59, 3064–3076; H.-W. Krause, U.
Schmidt, S. Taudien, B. Costisella, M. Michalik, J. Mol. Catal. A
1995, 104, 147–157; J. Wu, J.-X. Ji, R. Guo, C.-H. Yeung, A. C. S.
Chan, Chem. Eur. J. 2003, 9, 2693–2986; I. Gergely, C. Hegedüs, A.
Szçllçsy, A. Monsees, T. Riermeier, J. Bakos, Tetrahedron Lett. 2003,
44, 9025–9028.
[38] a) D. A. Allen, B. F. Taylor, J. Chem. Soc. Dalton 1982, 51–54; b) A.
Suµrez, M. A. MØndez-Rojas, A. Pizzano, Organometallics 2002, 21,
4611–4621.
[39] This difference is much larger than the difference of 11 Hz observed
in the binap series covering bisphosphines with strong (binap) and
poor donor properties (difluorphos); see reference [9].
[47] For recent reviews on high-throughput experimentation in homoge-
neous catalysis, see: J. G. de Vries, A. H. M. de Vries, Eur. J. Org.
Chem. 2003, 799–811; C. Gennari, U. Piarulli, Chem. Rev. 2003, 103,
3071–3100; T. Satyanarayana, H. B. Kagan, Adv. Synth. Catal. 2005,
347, 737–748.
[40] a) R. J. Goodfellow in Multinuclear NMR (Ed. J. Mason), Plenum,
New York, 1987, p. 521; b) R. Benn, C. Brevard, J. Am. Chem. Soc.
Received: January 9, 2006
Published online: May 4, 2006
Chem. Eur. J. 2006, 12, 5001 – 5013
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5013