Synthesis of chiral 2,5-bis(oxymethyl)-functionalized bis(phospholanes) and their application in Rh- and Ru-catalyzed enantioselective hydrogenations

Article abstract of DOI:10.1002/1099-0690(200112)2001:24<4615::AID-EJOC4615>3.0.CO;2-N

The synthesis of a series of chiral 2,5-bis(oxymethyl)-substituted bis(phospholanes) 13a-c and 15a,b (BASPHOS) is described, representing functionalized derivatives of the prominent DuPHOS or BPE ligands. D-Mannitol was used as the starting material for these ligands. New bisphospholanes were used as ligands in the enantioselective rhodium(I)-catalyzed hydrogenation of functionalized olefins like unsaturated α- and β-amino acid derivatives, itaconates, and an unsaturated phosphonate. A relevant ruthenium(II) catalyst was used for the reduction of prochiral β-oxo esters. The enantioselectivities, ranging from 8-99% ee, were strongly dependent on the type of the substituent on the oxymethyl group as well on the bridge connecting the phospholane units. Wiley-VCH Verlag GmbH, 2001.

Full text of DOI:10.1002/1099-0690(200112)2001:24<4615::AID-EJOC4615>3.0.CO;2-N

FULL PAPER
Synthesis of Chiral 2,5-Bis(oxymethyl)-Functionalized Bis(phospholanes) and
Their Application in Rh- and Ru-Catalyzed Enantioselective Hydrogenations
Jens Holz,*[a] Rainer Stürmer,[b] Ute Schmidt,[a] Hans-Joachim Drexler,[a] Detlef Heller,[a]
Hans-Peter Krimmer,[c] and Armin Börner*[a]
Keywords: Asymmetric synthesis / Hydrogenations / Phosphanes / Rhodium
The synthesis of a series of chiral 2,5-bis(oxymethyl)-substi-
ated α- and β-amino acid derivatives, itaconates, and an un-
tuted bis(phospholanes) 13a−c and 15a,b (BASPHOS) is de- saturated phosphonate. A relevant ruthenium(II) catalyst was
scribed, representing functionalized derivatives of the prom- used for the reduction of prochiral β-oxo esters. The enan-
inent DuPHOS or BPE ligands.
starting material for these ligands. New bisphospholanes
were used as ligands in the enantioselective rhodium(I)-cata- group as well on the bridge connecting the phospholane un-
D
-Mannitol was used as the tioselectivities, ranging from 8−99% ee, were strongly de-
pendent on the type of the substituent on the oxymethyl
lyzed hydrogenation of functionalized olefins like unsatur-
its.
Introduction
The electron-rich diphosphanes DuPHOS (A ϭ 1,2-
phenylene) and BPE (A ϭ 1,2-ethylene) originally disco-
vered by Burk belong to one of the most powerful family
of ligands currently used in RhI- and RuII-catalyzed asym-
metric hydrogenations.[9] There is some evidence that the
high enantiofacial discriminating ability of these and other
C2-symmetric bis(phospholanes),[10a] as well as that of re-
lated bis(phosphetanes),[10b] originates from the nature of
the alkyl groups in the 2,5-positions.[9c,11,12] Several pieces
of evidence have been accumulated that show that for each
class of substrates the appropriate alkyl groups have to be
selected in order to achieve maximal enantioselectivity.[9c,9d]
Understanding the mechanism of chirality transfer in en-
antioselective transition metal catalyzed reactions is a chal-
lenging task in current synthetic chemistry.[1] However, in
spite of literally thousands of chiral ligands that have been
reported in the past,[2] there is no unique rationale for the
efficiency of all catalysts not even for those that have been
applied in the same reaction. Therefore, the design of new
catalysts is mainly based on trial and error. Economic
reasons, like the actual patent situation or easily available
starting materials, as well as the aim of profiting from a
certain synthetic methodology still dominate the disclosure
of new ligands. Moreover, most ligands have been tested
only for special reactions or selected substrates (standard
substrates) and their suitability for other applications re-
mains vague.
For some years we have been interested in elucidating
the effect of additional functional groups like hydroxy and
alkoxy groups in conventional diphosphane ligands on
asymmetric hydrogenations with RhI catalysts.[3] In addi-
tion to their purely steric effect, such ‘‘hard’’ oxy func-
tionalities can act as ‘‘hemilabile’’ ligands.[4,5] Moreover, a
high potential exists for the establishment of hydrogen
bonds between the ligand and parts of a suitably func-
tionalized prochiral substrate.[6] In general, an alteration of
the performance of the parent (nonfunctionalized) catalyst
can be expected from these effects.[7,8]
Interestingly, in a recent work Brunner et al. achieved
some degree of enantioselection with a 3,4-dimethoxy-sub-
stituted bis(phospholane) ligand.[13] This report prompted
us and the groups of Zhang and RajanBabu to utilize oxy
groups in the 3,4-positions for the fine-tuning of 2,5-di-
methyl-substituted phospholanes (RoPHOS).[14] Indeed, in
the asymmetric hydrogenation, differences in the ee by up
to 7% were observed, depending on the nature of the oxy
substituent (BnO, MeO, HO).[14a,14b,15]
A closer proximity of the oxy groups to the catalytic
center as in complexes of RoPHOS ligands should enhance
their effect on the asymmetric reaction. In order to confirm
this hypothesis we envisaged the synthesis of 2,5-bis(oxyme-
thyl)-substituted bis(phospholanes) that we named
BASPHOS.[16] In a preliminary note we gave evidence that
[a]
Institut für Organische Katalyseforschung an der Universität
Rostock e.V.,
Buchbinderstraße 5/6, 18055 Rostock, Germany
E-mail: jens.holz@ifok.uni-rostock.de;
armin.boerner@ifok.uni-rostock.de
[b]
BASF Aktiengesellschaft, Hauptlaboratorium
Feinchemikalien,
Ϫ ZHF/D,
67056 Ludwigshafen, Germany
[c]
Degussa AG, Fine Chemicals Division,
Postfach 13 51, 63403 Hanau, Germany
Eur. J. Org. Chem. 2001, 4615Ϫ4624
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/1224Ϫ4615 $ 17.50ϩ.50/0
4615

J. Holz, A. Börner et al.
FULL PAPER
the relevant tetrahydroxy-substituted ligand can be advant-
By regioselective reaction of the primary hydroxy groups,
ageously used for the Rh-catalyzed hydrogenation in water, alkyl groups of different sizes could be incorporated in the
where excellent enantioselectivities could be achieved.[16]
1,6-positions (Scheme 2). Among different alkylating re-
In general, phospholanes with two oxymethyl groups in agents tried, we found that selective O-benzylation by em-
the 2,5-position combine structural motifs of alkyl and oxy ployment of dibutyltin oxide and benzyl bromide was the
groups and therefore the comparison with related DuPHOS most promising method.[22] The relevant 1,6-di-O-benzyl
and BPE ligands will give answers about the effect of the ether 7a was obtained in 60% yield. Unfortunately, all at-
oxymethyl groups. As substrates for the hydrogenation with tempts to achieve regioselective O-methylation only af-
corresponding RhI and RuII catalysts we have tested stand- forded mixtures of regioisomers. Therefore, the interme-
ard substrates such as itaconic acid and α-(acetylamino)cin- diate protection of the secondary hydroxy groups was
namic acid derivatives, as well as an unsaturated β-amino necessary. This was possible by prior protection of the
acid precursor and β-oxo acid derivatives.
primary hydroxy groups with a bulky silylating agent such
as TBDPS-Cl (tert-butyldiphenylsilyl chloride), affording
the silylated compound 7b.[22] In turn, the 2,5-dihydroxy
groups were treated with 2-methoxypropene in the presence
of toluenesulfonic acid to give 1,3-dioxepane 8.[22] After lib-
eration of the primary hydroxy groups with fluoride af-
fording 9,[22] and subsequent alkylation with methyl iodide,
1,6-di-O-methyl ether 10 was obtained. Cleavage of the ace-
tal provided the desired selectively protected compound 7c.
Diols 7a and 7c, in turn, were treated with thionyl chloride
according to the procedure of Burk et al.[9c] Subsequent in
situ ruthenium-catalyzed oxidation of cyclic sulfites gave
rise to the sulfates 11a and 11b.
Synthesis of the Ligands
As already briefly described in the synthesis of the hydro-
xymethyl-substituted bis(phospholane),[16] our synthetic
pathway starts from -mannitol (1, Scheme 1) and follows
general strategies for the synthesis of hydroxyphos-
phanes.[17] This approach is similar to the protocol giving
access to RoPHOS-type ligands,[14] however, in contrast to
the synthesis of the latter the hydroxy groups in 3,4-position
of -mannitol have to be reductively removed. The first step
of this conversion could be conveniently achieved by select-
ive protection of the hydroxy groups in 1,2;5,6-position as
isopropylidene acetals, affording the diacetonide 2.[18] Reac-
tion of the remaining hydroxy groups with thiophosgene in
turn gave the thiocarbonate 3.[19] An elimination reaction
with triethyl phosphite according to the procedure of
Haines yielded the (E)-olefin 4 in 90% yield.[20] By hydro-
genation of the double bond with heterogeneous Rh[21] or
Pt catalyst at 1 bar of hydrogen pressure the protected tetrol
5 was obtained. It is necessary to mention that careful puri-
fication of 4 was important in order to obtain a sufficient
yield in the subsequent hydrogenation of the olefin. Cleav-
age of the acetal groups with aq. HCl afforded the key com-
pound 6 bearing four hydroxy groups.[21]
1
Scheme 2. Reagents and conditions: a: for 7a: 1. Bu2SnϭO, toluene,
reflux, 4 h; 2. Bu4NBr, BnBr, toluene, reflux, 16 h; for 7b: TBDPS-
Cl, imidazole, DMF, 0 °C to room temp., 24 h; b: 2-methoxy pro-
pene, TsOH, 0 °C, 1 h; c: TBAF, THF, room temp., 2 h; d: NaH,
MeI, THF, room temp., 16 h; e: 1  HCl, THF, room temp., 1 h;
f: SOCl2, CCl4, reflux, 1.5 h; 2. RuCl3, NaIO4, CCl4, CH3CN, H2O,
0 °C to room temp., 1 h
The dibenzyl ether 11a could also be used for the syn-
thesis of the 1,6-dihydroxy derivative 12 (Scheme 3). This
compound can be advantageously employed for incorpora-
tion of other O-protective groups. As an example, the reac-
tion to the THP-protected diol 11c is given in the scheme.
Due to the formation of a new stereogenic carbon atom
in each THP ring three diastereomers were found in the
reaction mixture.
Scheme 1. Reagents and conditions: a: acetone, ZnCl2, 25 °C, 2 h;
b: SϭCCl2, DMAP, 0 °C, 1 h; c: P(OEt)3, 160 °C, 20 h; d: H2, Rh/
Al2O3, THF, room temp. or H2, Pt/C, MeOH, room temp.; e: 2 
HCl, 80 °C, 2 h
Cyclic sulfates 11aϪc are direct precursors of the desired
phospholanes (Scheme 4). Thus, treatment of 1,2-diphos-
phanylbenzene with nBuLi, followed by addition of the sul-
4616
Eur. J. Org. Chem. 2001, 4615Ϫ4624

Rh- and Ru-Catalyzed Enantioselective Hydrogenation
FULL PAPER
pholanes) 13aϪc and 15a and 15b with [Rh(COD)2]BF4,
affording
cationic
complexes
of
the
type
[Rh(COD){bis(phospholane)}]BF4 (Scheme 5). In Table 1
the relevant 31P NMR spectra of the ligands and their cor-
responding Rh complexes are detailed.
Scheme 3. Reagents and conditions: a: 1 bar H2, Pd/C, MeOH,
room temp.; b: 3,4-dihydro-2H-pyran, PPTS, CH2Cl2, room
temp., 2 h
Scheme 5. Reagents and conditions: a: [Rh(COD)2]BF4, THF, 0 °C
to room temp., 1 h; b: aq. HBF4, MeOH, room temp., 24 h
Table 1. 31P NMR spectra of functionalized bis(phospholanes) and
their [Rh(COD){bis(phospholane)}]BF4 complexes recorded in
CDCl3
Ligand
δ values of ligand
δ/J(103Rh-31P) [Hz]
values of Rh complex
Scheme 4. Reagents and conditions: a: 1. BuLi, THF, Ϫ78 °C to
room temp., 2 h; 2. 1,2-(H2P)2C6H4, Ϫ78 °C to room temp.; 3.
BuLi, Ϫ78 °C to room temp., 12 h; b: 1. BuLi, THF, Ϫ78 °C to
room temp., 2 h; 2. H2PϪC2H4ϪPH2, Ϫ78 °C to room temp.; 3.
BuLi, Ϫ78 °C to room temp., 12 h; 4. BH3ϪTHF, Ϫ10 °C to room
temp., 2 h; c: DABCO, toluene, 40 °C
13a
13b
13c
16
15a
15b
Ϫ11.5
ϩ64.3/149
Ϫ11.7
ϩ65.1/150
Ϫ11.5 to Ϫ12.3[a]
ϩ58.5 to ϩ73.6[a]
ϩ64.1/148
Ϫ
Ϫ6.9
Ϫ7.0
ϩ67.3/147
ϩ67.3/150
fates and a second portion of nBuLi gave, under complete
inversion of the stereogenic carbon atoms, the C2-symmet-
ric phospholanes 13aϪc in 32Ϫ62% yield. The THP-pro-
tected sulfate 11c was used as its diastereomeric mixture
and gave rise to the expected mixture of diastereomeric bis-
(phospholanes) 13c. Removal of the THP groups in this
compound should give the interesting tetrakis(hydroxyme-
thyl)-substituted bis(phospholane) 16. However, owing to
the high polarity and air sensibility of this compound, all
attempts to isolate the polyhydroxyphosphane from the
acidic reaction mixture failed.
Ethylene-bridged bis(phospholanes) 15a and 15b were
obtained under the same reaction conditions employing
1,2-diphosphanylethane as a nucleophile. However, for the
purification of the bis(phospholanes) prone to oxidation,
prior conversion into their BH3 adducts 14a and 14b was
advantageous.[23] Final cleavage of the borane adducts with
DABCO delivered the bis(phospholanes) 15a and 15b.
[a]
Mixture of diastereomers.
As mentioned above, removal of the THP groups in the
bis(phospholane) 13c in order to obtain access to the tetra-
hydroxyphosphane 16 was not successful. A more advant-
ageous route was revealed by the prior ‘‘protection’’ of the
trivalent phosphorus atoms by complexation of 13c to rho-
dium(I) by reaction with [Rh(COD)2]BF4 (Scheme 5).[16] In
Figure 1 (a) the 31P NMR spectrum of [Rh(COD)(13c)]BF4
is depicted. The numerous resonances observed are due to
the stereogenic THP groups in ligand 13c. By treatment of
this mixture with catalytic amounts of aqueous HBF4 in
methanol, all THP groups were removed by transacetaliz-
ation, and the desired precatalyst [Rh(COD)(16)]BF4 was
obtained in 72% yield. In Figure 1 (b) the 31P NMR spec-
trum of the product after hydrolysis is shown, which is char-
acterized by a single doublet due to the 103RhϪ31P coupling
of 148 Hz.
Crystals of [Rh(COD)(13b)]BF4 suitable for X-ray ana-
lysis could be obtained by slow diffusion of diethyl ether
Synthesis of Precatalysts
Rhodium(I) precatalysts suitable for the enantioselective into a concentrated MeOH solution of the rhodium com-
hydrogenation were produced by reaction of the bis(phos- plex. A single-crystal X-ray structural analysis established
Eur. J. Org. Chem. 2001, 4615Ϫ4624
4617

J. Holz, A. Börner et al.
FULL PAPER
Figure 1. 31P NMR spectra of [Rh(COD)(13c)]BF4 (a) and
[Rh(COD)(16)]BF4 in CDCl3 (b)
the structure of the cation as shown in Figure 2 (a), along
with the selected bond lengths and intramolecular angles.
All RhϪP and RhϪC bond lengths and angles of the exam-
ined complex are closely related to other bis(phospholanyl)-
benzene precatalysts.[9c,24] The phospholane rings adopt a
half-chair conformation wherein the substituents are ad-
justed in equatorial positions.
The C2 symmetry is disturbed in the solid state by differ-
ent conformations of the methoxymethyl groups. In general,
however, no intramolecular interactions of ether groups
with the rhodium center are visible [Figure 2 (b)]. As ex-
pected, only a very small dihedral angle (5.0 °) between the
planar 1,2-phenylene unit and the PϪRhϪP plane adopting
λ configuration is observed. The dihedral angle between the
planes P,Rh,P and X,Rh,X (X ϭ centroid of the double
bond) is 21.8° and matches well the value found with the
corresponding Et-DuPHOS complex (21.1°).[24] The large
Figure 2. a: top view, showing the 1,2-phenylene unit and the ori-
entation of the methyl oxymethylene group as well as the num-
bering scheme of [Rh(COD)(13b)]ϩ; all hydrogen atoms except
those that are linked to stereogenic carbon atoms have been omit-
˚
ted for clarity; selected interatomic distances [A]: Rh1ϪC1
2.283(4), Rh1ϪC2 2.207(5), Rh1ϪC5 2.250(4), Rh1ϪC6 2.211(4),
Rh1ϪP1 2.285(1), Rh1ϪP2 2.265(1); selected intramolecular
angles (°): P1ϪRh1ϪP2 84.36(4), C1ϪRh1ϪC2 35.5(3),
C5ϪRh1ϪC6 35.7(2); b: front view, showing phospholane rings
and the dihedral angle between the PϪRhϪP plane and the
centroidϪRhϪcentroid plane
Enantioselective Hydrogenations
The catalytic properties of the complexes derived from
deviation from the square-planar coordination geometry the new functionalized bis(phospholanes) were first proven
around the metal center observed with these in the hydrogenation of standard substrates at 1 bar of H2
(COD){bis(phospholane)}Rh complexes cannot be the re- pressure and at 25 °C in MeOH as solvent. Enantioselectivi-
sult of steric interactions of the ligands as a space-filling ties achieved in the hydrogenation of (Z)-α-(acetylamino)-
model indicated. The reason for this deviation is not yet cinnamic acid and itaconic acid as well as their methyl es-
clear.
ters are listed in Table 2. Also those results obtained in the
4618
Eur. J. Org. Chem. 2001, 4615Ϫ4624

Rh- and Ru-Catalyzed Enantioselective Hydrogenation
FULL PAPER
Table 2. Results of the hydrogenation of prochiral olefins with [Rh(P-P)(COD)]BF4
Ligand
No.
Substrate[a]
R1
ee
A
R
R2
H
R3
R4
[%][b]
13a
13b
16
15a
15b
13a
13b
13c
16
15a
15b
13a
13b
16
13b
13c
13a
13b
16
15a
15b
13a
13b
16
15a
15b
13b
1,2-phenylene
1,2-phenylene
1,2-phenylene
1,2-ethylene
1,2-ethylene
1,2-phenylene
1,2-phenylene
1,2-phenylene
1,2-phenylene
1,2-ethylene
1,2-ethylene
1,2-phenylene
1,2-phenylene
1,2-phenylene
1,2-ethylene
1,2-ethylene
1,2-phenylene
1,2-phenylene
1,2-phenylene
1,2-ethylene
1,2-ethylene
1,2-phenylene
1,2-phenylene
1,2-phenylene
1,2-ethylene
1,2-ethylene
1,2-phenylene
Bn
Me
H
Bn
Me
Bn
Me
THP
H
Bn
Me
Bn
Me
H
Bn
Me
Bn
Me
H
Bn
Me
Bn
Me
H
Bn
Me
Me
Ph
NHAc
COOH
86.6 (S)
94.8 (S)
79.5 (S)
91.5 (S)
81.1 (S)
95.7 (S)
98.9 (S)
96.6 (S)
95.8 (S)
95.6 (S)
87.3 (S)
95.9 (R)
96.9 (R)
37.1 (R)
81.5 (R)
21.1 (R)
70.3 (R)
97.9 (R)
5.7 (R)
Ph
H
NHAc
COOMe
H
H
CH2COOH
CH2COOMe
NHBz
COOH
COOMe
P(O)(OMe)2
H
H
H
73.1 (R)
8.1 (R)
Ph
H
20.8 (R)
78.8 (R)
76.4 (R)
71.9 (R)
56.2 (R)
68.2 (S)
NHAc
Me
COOMe
[a]
[b]
Catalyst/substrate (1:100), 1 atm total pressure above the reaction mixture, 25 °C; 1 mmol substrate in 15 mL of methanol. Deter-
mined after 100% conversion. Measured on the crude product, GC with a chiral column: for N-acetylphenylalanine and dimethyl succinate
after esterification with diazomethane and for the methyl ester of N-acetylphenylalanine and dimethyl succinate, respectively, with XE
60--valine tert-butylamide, 150 °C. The conversion into the saturated phosphonate and its ee were determined by HPLC: stationary
phase Chiralcel OD-H, eluent n-hexane/ethanol. The conversion into methyl 3-acetylaminobutanoate and its ee were determined by GC:
Chiraldex β-PM 50 m ϫ 0.25 mm (astec), 130 °C.
hydrogenation of a pharmaceutically important α,β-unsat- times dramatic influence upon the stereo-discriminating
urated phosphonate and methyl (Z)-3-(acetylamino)buteno- ability of the catalyst. Several tendencies become obvious.
ate are given. Measurements of the ee were performed after In general, superior enantioselectivities were observed with
completion of the consumption of hydrogen. Times for con- the catalyst based on the ligand 13b characterized by four
version of the prochiral substrate are not detailed, due to methoxy groups and the phenylene bridge. When the methyl
individually varying periods necessary for the generation of groups were replaced by the larger benzyl groups (13a) di-
the catalytically active species from the precatalysts.[24,25] minished ee values were obtained. Still more expressed is
This known feature did not allow a reliable differentiation the decrease of the ee by application of the relevant tetrahy-
and comparison between the time necessary for the prehy- droxy-substituted bis(phospholane) 16 as ligand. It should
drogenation of the diolefin (1,5-cyclooctadiene) in the pre- be reminded that the relevant catalyst gave Ͼ 99% ee in the
catalyst and the time for the hydrogenation of the prochiral hydrogenation of α-(acetylamino)acrylic acid in water.[16]
substrate proceeding in parallel. In general, no clear cor- Noteworthy is the excellent enantioselectivity produced in
relation in activity was found in dependency on the sub- the hydrogenation of methyl α-(acetylamino)cinnamate
stituent R in the ROCH2 groups, the type of the bridge A with the complex of THP-protected bis(phospholane) 13c.
connecting the phospholane units and the substrate.
Although the catalyst consists of a mixture of several dia-
Table 2 clearly shows, that the nature of the bridge be- stereomers [see Figure 1 (a)] an enantioselectivity of 96.6%
tween both phospholanes (ethylene or phenylene) as well as was reached. Either the configuration in the THP groups
the type of the oxy substituent exert a considerable, some- has no significant influence or only individual complexes
Eur. J. Org. Chem. 2001, 4615Ϫ4624
4619

J. Holz, A. Börner et al.
FULL PAPER
among the mixture of diastereomeric Rh complexes are act-
The tetramethoxy ligand 13b was also tested in the Ru-
ive in the enantioselective hydrogenation. Since the isolation catalyzed hydrogenation of pharmaceutically important β-
of single complexes failed, the answer remains speculative.
Seriously diminished enantioselectivities resulted from
the replacement of the rigid 1,2-phenylene bridge by the
conformationally more flexible ethylene bridge (ligands 15a
and 15b). This tendency is in agreement with some observa-
tions of Burk with BPE catalysts in hand,[9c,26] but more
expressed in the BASPHOS series (see in particular ligand
15b). At least for the substrates considered herein, their is
clear evidence that the oxy functionalization in the 2,5-posi-
tion of the phospholane rings makes bis(phospholane) cata-
lyst more sensitive to structural changes of the substrate.
With all catalysts tried inferior ee values were observed in
the hydrogenation of the α,β-unsaturated phosphonate in
comparison with RoPHOS catalysts.[14a]
oxo esters (Table 3). The catalyst was prepared in situ ac-
[28]
ˆ
cording to the protocol of Genet et al.
Hydrogenations
were carried out at 30 or 90 bar of hydrogen pressure. Elev-
ated temperature was required in order to obtain sufficient
conversion. As Table 3 shows, (S)-configured β-hydroxycar-
boxylates were obtained with excellent enantioselectivities.
Increasing of the reaction temperature only slightly influ-
enced the ee. However, raising the hydrogen pressure from
30 to 90 bar decreased the ee significantly, although not to
the extent reported for a BPE-Ru catalyst.[9e]
Table 3. Ruthenium-catalyzed hydrogenation of β-oxocarboxylates
In order to broaden the scope of the substrates we also
tested the precatalyst based on the MeO ligand (13b) in the
enantioselective hydrogenation of methyl (Z)-3-(acetylami-
no)butenoate. In a parallel work we could show that highly
enantioselective hydrogenation of this particular substrate
is possible by application of an Rh-DuPHOS catalyst on
condition that the reaction was run in a polar solvent like
methanol under application of normal hydrogen pres-
sures.[27] This protocol is contrary to the conditions usually
recommended in the past for these type of (Z)-configured
substrates. In Figure 3 the dependency of the enantioselec-
tivity upon the H2 pressure by application of
[Rh(COD)(13b)]BF4 is depicted. For comparison also rele-
vant curves registered with the corresponding Me-BPE and
Et-DuPHOS-Rh catalyst are shown. At elevated pressures
the MeO catalyst is superior. However, surprisingly, the en-
antioselectivity is rather independent on the H2 pressure,
which is in contrast to the behaviour of DuPHOS catalysts.
Thus, finally at 1 bar the Et-DuPHOS catalyst gave higher
ee values. The BASPHOS complex provided the β-amino
acid ester in only 68.2% ee (Table 2).
R
RЈ S/C T [°C] p [bar] t [h] Conversion [%] ee [%]
MeO2C(CH2)2 Me 200 35
MeO2C(CH2)2 Me 200 35
MeO2C(CH2)2 Me 300 25
MeO2C(CH2)2 Me 300 60
Me
Me
30
30
90
30
30
30
24 70
72 85
72 15
72 85
24 11
24 57
98.8[a]
98.7[a]
90.8[a]
95.8[a]
96.2[b]
97.2[b]
Me 300 25
Et 300 35
[a]
Determined after prior separation from nonconverted starting
material by HPLC: Chiralcel OD-H, eluent n-hexane/ethanol
(98:2).
umn: 50 m Lipodex A, 50 °C.
[b]
Measured on the crude product, GC with a chiral col-
Conclusion
The synthesis of a series of new 2,5-bis(oxymethyl)-func-
tionalized bis(phospholanes) (BASPHOS) is described
starting from -mannitol. Key intermediates of the ap-
proach allow the incorporation of different O-alkyl groups.
The new phosphanes represent functionalized ligands of the
highly efficient DuPHOS/BPE-type established by Burk
and co-workers. In the enantioselective hydrogenation of
standard substrates like (Z)-α-(acetylamino)cinnamic acid
or itaconic acid with cationic RhI complexes of the new
diphosphanes, enantioselectivities ranging from 8 to 99%
were observed. This means that in comparison to bis(phos-
pholanes) of the DuPHOS/BPE-type the 2,5-dioxy func-
tionalization affects more the enantiofacial discriminating
properties of the catalyst than simple alkyl groups. In other
words, the additional functional groups make the catalyst
more sensible to changes in the structure of the prochiral
substrate. The degree of the deviation from the catalytic
performance of the DuPHOS/BPE ligands was dependent
on the substrate used. Particularly large differences due to
the type of the oxy groups and the bridge between both
phospholane units were observed with itaconic acid and its
dimethyl ester. Effects found are much stronger than those
observed with corresponding RoPHOS catalysts bearing
(remote) oxy groups in 3,4-position of the phospholane
rings. In contrast, in the hydrogenation of an unsaturated
Figure 3. Dependency of the ee upon the H2 pressure with
[Rh(COD)(P-P)]BF4 (PϪP: 13b, Et-DuPHOS, Me-DuPHOS) in
the hydrogenation of methyl (Z)-3-(acetylamino)butenoate
4620
Eur. J. Org. Chem. 2001, 4615Ϫ4624

Rh- and Ru-Catalyzed Enantioselective Hydrogenation
FULL PAPER
CH3), 1.34 (m, 2 H, HaϪCH2), 1.67 (m, 2 H, HbϪCH2), 3.23 (dd,
β-amino acid derivative the known pressure dependency of
the ee with DUPHOS catalysts was less pronounced with
an oxy-functionalized bis(phospholane) as ligand. In the
hydrogenation of prochiral β-oxo esters up to 98.8% ee were
achieved with the RuII(13b) catalyst, depending upon the
pressure of hydrogen.
3
2Ja,b ϭ 9.9, JH,Hϭ 5.3 Hz, 2 H, HaϪCH2O), 3.30 (s, 6 H, CH3),
2
3
3.33 (dd, Ja,b ϭ 9.9, JH,Hϭ 6.3 Hz, 2 H, HbϪCH2O), 3.92 (m, 2
H, CH). 13C NMR (CDCl3): δ ϭ 25.6 (CH3), 31.1 (C-3/4), 59.1
˜
(OCH3), 70.4 (C-2/5), 76.2 (C-1/6), 100.5 [C(O)2]. IR (neat): ν ϭ
2988, 2939, 2876, 2812, 1458, 1381, 1219, 1139, 1098, 1073, 987,
829 cmϪ1 Ϫ MS (EI): m/z (%) ϭ 218 (1) [M], 203 (3) [M Ϫ Me],
173 (28) [M Ϫ CH2OMe]. C11H22O4 (218.3): calcd. C 60.52, H
10.16; found C 60.38, H 10.07.
Experimental Section
(2S,5S)-1,6-Dimethoxyhexan-2,5-diol (7c): The dimethyl ether 10
(4.0 g, 18.3 mmol) was hydrolyzed in a mixture of THF (60 mL)
and 1  HCl (60 mL) within 30 min. After evaporation of the solv-
ents, the residue was purified by column chromatography (EtOH/
EtOAc, 1:3; Rf ϭ 0.45) to yield a pale yellow syrup in nearly quant-
itative yield (3.14 g, 96% yield). [α]2D3ϭ Ϫ7.2 (c ϭ 1.09, CH3OH).
1H NMR (CD3OD): δ ϭ 1.56 (m, 4 H, CH2), 3.23 (m, 4 H, CH2O),
3.37 (s, 6 H, CH3), 3.72 (m, 2 H, CH). 13C NMR (CD3OD): δ ϭ
General Remarks
All reagents were obtained from Aldrich and Merck. Solvents were
dried and freshly distilled under argon before use. Reactions using
phosphanes and organometallic compounds were performed under
Ar by using standard Schlenk techniques. Thin-layer chromato-
graphy was performed on pre-coated TLC plates (silica gel 60 F254
,
Merck). Flash chromatography was carried out with silica gel 60
(particle size 0.040Ϫ0.063 mm, Merck). Melting points are cor-
rected. NMR spectra were recorded at the following frequencies:
400.13 MHz (1H), 100.63 MHz (13C), 161.98 MHz (31P). Chemical
˜
30.6 (C-3/4), 59.2 (OCH3), 71.1 (C-2/5), 78.2 (C-1/6). IR (neat): ν ϭ
3405, 2926, 2883, 2824, 1453, 1197, 1126, 965 cmϪ1. C8H18O4
(178.2): calcd. C 53.91, H 10.18; found C 53.38, H 10.01.
General Procedure for the Synthesis of (4S,7S)-4,7-Bis(alkoxyme-
thyl)[1,3,2]dioxathiepane 2,2-Dioxide (11a) and (11b): To a solution
of the diols 7a or 7c (10 mmol) in CCl4 (70 mL) was slowly added
thionyl chloride (1.43 g, 12 mmol) and the mixture heated under
reflux for 90 min. After evaporation of the solvent, the residue was
dissolved in a mixture of CCl4 (45 mL), acetonitrile (45 mL) and
water (65 mL) at 0 °C. Then, RuCl3 trihydrate (15 mg, 0.06 mmol)
and sodium periodate (4.28 g, 20 mmol) were added and the reac-
tion mixture was stirred at ambient temperature for 1 h. After addi-
tion of water (60 mL), the solution was extracted with diethyl ether
(3 ϫ 75 mL) and the combined phases were washed with brine (1
ϫ 50 mL). After drying with Na2SO4, the solution was filtered
through SiO2 and concentrated under vacuo. The nearly 5 mL of
residue left was treated with n-hexane and the pure cyclic sulfates
11a and 11b precipitated in 80% yields. If the precipitation failed,
the products could also be isolated by flash chromatography [11a:
n-hexane/EtOAc (2:1), Rf ϭ 0.20; 11b: n-hexane/EtOAc (1:2),
Rf ϭ 0.40].
1
shifts of H and 13C NMR spectra are reported in ppm downfield
from TMS as internal standard. Chemical shifts of 31P NMR spec-
tra are referred to H3PO4 as external standard. Signals are quoted
as s (singlet), d (doublet), br (broad) and m (multiplet). Elemental
analyses of the phospholanes are not detailed. Owing to their high
sensibility to oxidation values found were beyond the limit of 0.3%.
Hydrogenation experiments of functionalized olefins have been car-
ried out under normal pressure and isobaric conditions with an
automatically recording gas measuring device (1.0 atm overall pres-
sure over the solution). The experiments were performed with
0.01 mmol of precatalyst, 1.0 mmol of prochiral olefin in 15.0 mL
of solvent at 25.0 °C. The conversion of the prochiral unsaturated
amino acids and itaconic acid as well as the ee of the products were
determined by GC. The acids were esterified with trimethylsilyl-
diazomethane before the GC-measurements [FID, carrier gas Ar
(1 mL/min); methyl N-acetylphenylalaninate: fused silica, 10 m,
XE-60--valin-tert-butylamide, ID 0.2 mm, oven temperature 150
°C; dimethyl methylsuccinate: fused silica, Lipodex E (Machery
and Nagel), 25 m, ID 0.25 mm, oven temperature 85 °C]. The con-
version into the saturated phosphonate and its ee were determined
by HPLC (stationary phase Chiralcel OD-H; eluent n-hexane/eth-
anol). The conversion into methyl 3-(acetylamino)butanoate and
its ee were determined by GC [Chiraldex β-PM 50 m ϫ 0.25 mm
(astec), 130 °C]. Hydrogenation experiments of ketones were car-
ried out as detailed in Table 3. The ee values of the chiral alcohols
obtained were determined either by HPLC (Chiralcel OD-H; eluent
n-hexane/ethanol, 98:2) or by GC with a chiral column (50 m Lipo-
dex A, 50 °C).
1
11a: M.p. 57Ϫ59 °C. [α]2D6 ϭ Ϫ37.2 (c ϭ 1.01; CHCl3). H NMR
(CDCl3): δ ϭ 2.01 (m, 4 H, CH2), 3.56 (dd, Ja,b ϭ 10.8, JH,Hϭ
4.9 Hz, 2 H, HaϪCH2O), 3.65 (dd, Ja,b ϭ 10.8, JH,Hϭ 5.4 Hz, 2
H, HbϪCH2O), 4.57 (AB, 2Ja,b ϭ 12 Hz, 4 H, OCH2Ph,), 4.78 (m,
2 H, CH), 7.27Ϫ7.37 (m, 10 H, arom. H). 13C NMR (CDCl3): δ ϭ
28.9 (CH2), 70.8 (CH2Ph), 73.4 (CH2O), 82.6 (CH), 127.7, 127.9,
2
3
2
3
˜
128.4, 137.3 (arom. C). IR (KBr): ν ϭ 3089, 3060, 3035, 2956,
2931, 2899, 2866, 1498, 1449, 1386, 1369, 1197, 1142, 1107, 1077,
1064, 987, 949, 913, 813, 751, 733, 697 cmϪ1. MS (CI): m/z (%) ϭ
301 (4) [M Ϫ C7H7], 271 (1) [M Ϫ CH2OBn], 91 (100) [C7H7].
C20H24O6S (392.5): calcd. C 61.21, H 6.16, S 8.17; found C 61.03,
H 6.23, S 8.07.
3,4-Dideoxy-2,5-O-isopropylidene-1,6-di-O-methyl-D-threo-hexitol
(10): To a suspension of sodium hydride (1.06 g, 44 mmol) in THF
(60 mL) was added at 0 °C a solution of the diol 9 (3.80 g,
20 mmol) in THF (30 mL). The suspension was stirred at room
temperature for 2 h and then methyl iodide (6.21 g, 44 mmol) was
added slowly. The stirring was continued for about 12 h at ambient
temperature and then the excess sodium hydride was destroyed by
carefully addition of water (30 mL). THF was removed using a
rotary evaporator and the resulting aqueous residue was extracted
with dichloromethane (3 ϫ 50 mL). The combined organic wash-
ings were dried (Na2SO4), and after evaporation of the solvent the
residue was purified by column chromatography (n-hexane/EtOAc,
2:1; Rf ϭ 0.40) to yield a colorless syrup (3.68 g, 84% yield). [α]2D3ϭ
1
11b: M.p. 75Ϫ78 °C. [α]2D3 ϭ Ϫ44.1 (c ϭ 1.01; CHCl3). H NMR
(CDCl3): δ ϭ 1.92Ϫ2.04 (m, 4 H, CH2), 3.37 (s, 6 H, OCH3), 3.47
(dd, Ja,b ϭ 10.8, 3JH,Hϭ 4.7 Hz, 2 H, HaϪCH2O,), 3.57 (dd,
2
2Ja,b ϭ 10.8, JH,Hϭ 5.4 Hz, 2 H, HbϪCH2O), 4.72 (m, 2 H, CH).
3
13C NMR (CDCl3): δ ϭ 28.8 (CH2), 59.3 (OCH3), 73.4 (CH2O),
˜
82.5 (CH). IR (KBr): ν ϭ 2998, 2973, 2951, 2933, 2859, 2839, 2823,
1473, 1457, 1384, 1350, 1207, 1188, 1142, 1126, 1107, 1097, 958,
939, 913, 857, 815 cmϪ1. MS (EI): m/z (%) ϭ 240 (1) [M], 195 (1)
[M Ϫ CH2OMe], 45 (100) [CH2OCH3]. C8H16O6S (240.3) calcd. C
39.99, H 6.71, S 13.34; found C 40.06, H 6.75, S 13.27.
(4S,7S)-4,7-Bis(hydroxymethyl)[1,3,2]dioxathiepane
2,2-Dioxide
Ϫ32.8 (c ϭ 1.01, CHCl3). 1H NMR (CDCl3): δ ϭ 1.31 (s, 6 H, (12): The O-benzyl-protected cyclic sulfate 11a (4.80 g, 12.2 mmol)
Eur. J. Org. Chem. 2001, 4615Ϫ4624 4621

J. Holz, A. Börner et al.
was dissolved in methanol (40 mL) and stirred under 1 bar of hy- HbϪCH2Ph), 7.10Ϫ7.45 (m, 24 H, arom. H). 13C NMR (CDCl3):
FULL PAPER
drogen in the presence of palladium on charcoal (5%, 0.2 g) until
the consumption of hydrogen ceased. After filtration of the suspen-
sion, the pure product 12 could be obtained by crystallization in
methanol (2.40 g, 94%). In most runs the raw product was suffi-
ciently pure and was used for the next step. M.p. 122Ϫ125 °C.
δ ϭ 30.4 (CH2), 30.9 (CH2), 38.9 (m, CHϪP), 39.5 (CHϪP), 72.5
(CH2Ph), 73.0 (CH2Ph), 74.1 (m, CH2O), 127.1Ϫ128.4 and 131.8
(arom. C), 138.5 and 138.6 (ipso-C), 141.8 (dd, 1JC,P ϭ 4.7, 2JC,P ϭ
4.7 Hz, CarϪP). 31P NMR (CDCl3): δ ϭ Ϫ11.5.
1
[α]2D3 ϭ Ϫ37.8 (c ϭ 1.01; CH3OH). H NMR (CD3OD): δ ϭ 1.97
13b: 1.05 g (62% yield) of a colorless syrup by column chromato-
1
3
graphy (n-hexane/EtOAc, 2:1; Rf ϭ 0.20). H NMR (CDCl3): δ ϭ
(m, 4 H, CH2), 3.63 (d, JH,Hϭ 5.2 Hz, 4 H, CH2O), 4.62 (m, 2 H,
CH). 13C NMR (CD3OD): δ ϭ 29.5 (CH2), 64.6 (CH2O), 86.8
(CH). IR (KBr): ν˜ ϭ 3391, 2978, 2955, 2942, 1456, 1436, 1400,
1374, 1353, 1232, 1199, 1087, 1070, 1035, 1007, 983, 941, 905, 857,
813 cmϪ1. MS (EI): m/z (%) ϭ 181 (2) [M Ϫ CH2OH], 121 (7), 87
(100), 57 (53) [C3H5O], 43 (20) [C2H3O], 31 (33) [CH2OH].
C6H12O6S (212.2) calcd. C 33.96, H 5.70, S 15.11; found C 34.13,
H 5.67, S 14.97.
1.55 (m, 2 H, HaϪCH2), 1.65 (m, 2 H, HaϪCH2), 2.16 (m, 2 H,
HaϪCH2), 2.31 (m, 2 H, HbϪCH2), 2.63 (m, 2 H, CHϪP), 2.78
(m, 2 H, CHϪP, CH2), 2.91 (m, 2 H, CH2O), 3.10 (s, 6 H, OCH3),
3.35 (s, 3 H, OCH3), 3.36 (m, 2 H, CH2O), 3.55 (m, 4 H, CH2O),
7.30 (m, 2 H, arom H), 7.45 (m, 2 H, arom. H). 13C NMR (CDCl3):
δ ϭ 30.3 (CH2), 30.9 (CH2), 39.0 (m, CHϪP), 39.6 (CHϪP), 58.2
(OCH3), 58.8 (OCH3), 74.5 (m, CH2O), 76.6 (m, CH2O), 128.4 and
2
131.8 (arom. C), 141.8 (dd, 1JC,P ϭ 4.0, JC,P ϭ 4.0 Hz, CarϪP), Ϫ
31P NMR (CDCl3): δ ϭ Ϫ11.7.
(4S,7S)-4,7-Bis(rac-tetrahydropyran-2-yloxymethyl)[1,3,2]dioxa-
thiepane 2,2-Dioxide (11c): To the diol 12 (3.00 g, 14.1 mmol), dis-
solved in dichloromethane (150 mL), were added pyridinium tosyl-
ate (0.59 g, 2.35 mmol) and 3,4-dihydro-2H-pyran (3.56 g,
42.3 mmol). After stirring for 2 h at ambient temperature, the solu-
tion was diluted with diethyl ether (250 mL) and washed with brine
(2 ϫ 75 mL). After drying and evaporation of the solvents, the
residue was purified by column chromatography (n-hexane/EtOAc,
1:1; Rf ϭ 0.20) to yield the diastereomeric mixture as a pale yellow
oil (4.80 g, 89%). 1H NMR (CDCl3): δ ϭ 1.45Ϫ2.10 (m, 16 H,
CH2), 3.55 (m, 4 H, CH2O), 3.81 (m, 4 H, CH2O), 4.62, 4.77 and
4.92 (m, 4 H, CHϪO). 13C NMR (CDCl3): δ ϭ 19.0 and 19.7
(CH2), 25.3 (CH2), 28.9 and 29.0 (CH2), 30.2 (CH2) 62.1, 62.2,
(CH2O), 68.0 and 68.2 (CH2O), 82.5, 82.5, 82.9 and 83.0 (CH),
98.6 and 99.1 [C(O)2]. C16H28O8S (380.5) calcd. C 50.51, H 7.42,
S 8.83; found C 50.12, H 7.77, S 8.56.
13c: 0.91 g (32% yield) of a pale yellow syrup by column chromato-
1
graphy (n-hexane/EtOAc, 2:1; Rf ϭ 0.30). H NMR (CDCl3): δ ϭ
1.32Ϫ1.85 (m, 28 H, CH2), 2.19 (m, 2 H, CH2), 2.35 (m, 2 H, CH2),
2.66 (m, 2 H, CHϪP), 2.91 (m, 2 H, CHϪP), 3.10Ϫ4.00 (m, 16 H,
CH2O), 4.12Ϫ4.62 (m, 4 H, CHϪO), 7.23 (m, 2 H, arom. H), 7.46
(m, 2 H, arom. H). 13C NMR (CDCl3) δϭ 19.4Ϫ19.6 (CH2),
25.4Ϫ25.5 (CH2), 30.2Ϫ30.9 (CH2), 38.7 (m, CHϪP), 39.5 (m,
CHϪP), 61.6Ϫ62.3 (CH2O), 69.3Ϫ71.1 (CH2O), 98.0Ϫ99.5
(CHO), 128.3 and 131.8 (arom. C), 141.8 (m, CarϪP). 31P NMR
(CDCl3): δ ϭ Ϫ11.5 to Ϫ12.3 (diastereomeric mixture).
14a: 0.42 g (15% yield) of a viscous syrup by column chromato-
1
graphy (n-hexane/EtOAc, 4:1; Rf ϭ 0.25). H NMR (CDCl3): δ ϭ
0.10Ϫ1.15 (br, 6 H, BH3), 1.20Ϫ1.41 (m, 4 H, HaϪCH2),
1.80Ϫ2.08 (m, 10 H, HbϪCH2, CH2, CHϪP), 2.36 (m, 2 H,
CHϪP), 3.43 (m, 4 H, CH2O), 3.38 (m, 4 H, CH2O), 4.37 (AB,
General Procedure for the Synthesis of the Bis(phospholanes) 13a؊c
and 15a,b: To a solution of 1,2-bis(phosphanyl)benzene (0.57 g,
4.0 mmol) and 1,2-bis(phosphanyl)ethane (0.38 g, 4.0 mmol), re-
spectively, in THF (60 mL) were added at a temperature of Ϫ78 °C
2 equiv. of nBuLi (1.6  solution in n-hexane, 5.00 mL). The re-
sulting yellow solution was stirred at 25 °C for 2 h. After cooling
again at Ϫ78 °C, the sulfates 11aϪc, dissolved in THF (10 mL),
were added dropwise. Stirring was continued at room temperature
for 3 h. Then a second portion of nBuLi (5.5 mL, 8.8 mmol) was
added at Ϫ78 °C and the mixture was stirred overnight at ambient
temperature. For the syntheses of ethylene-bridged bis(phosphol-
anes) 15a and 15b the reaction mixtures were cooled again to Ϫ10
°C and 3 equiv. of BH3·THF complex (1  solution in THF, 12 mL)
were added and the mixtures were warmed to room temperature
over a period of 2 h. The solvents were removed by concentration
under vacuo and the residue was taken up with water (20 mL) and
extracted under anaerobic conditions with dichloromethane
(100 mL). After separation of the organic phase, the solvent was
removed and the residue was purified by column chromatography
to yield the bis(phospholane). The borane complexes 14a and 14b
were dissolved in a solution of toluene (10 mL) and treated with 4
equiv. of DABCO by stirring at 40 °C under argon to yield the
unprotected bis(phospholanes) 15a and 15b, respectively. The com-
pletion of reaction was monitored by TLC.
2
2Ja,b ϭ 11 Hz, 2 H, HaϪCH2Ph), 4.40 (AB, Ja,b ϭ 12 Hz, 2 H,
HaϪCH2Ph), 4.43 (AB, 2Ja,b ϭ 12 Hz, 2 H, HbϪCH2Ph), 4.48 (AB,
2Ja,b ϭ 11 Hz, 2 H, HbϪCH2Ph), 7.22Ϫ7.40 (m, 20 H, arom. H).
13C NMR (CDCl3): δ ϭ 15.9 (m, CH2P), 28.6 (CH2), 29.1 (CH2),
39.5 (m, CHϪP), 68.4 (CH2Ph), 69.4 (CH2Ph), 72.7 (CH2O), 73.2
(CH2O), 127.4Ϫ128.3, 137.9 and 138.1 (arom. C). 31P NMR
(CDCl3): δ ϭ 40.2.
14b: 0.68 g (42% yield) of a white solid by column chromatography
(n-hexane/EtOAc, 4:1; Rf ϭ 0.20). m.p. 45Ϫ48 °C. [α]2D5 ϭ 21.9 (c ϭ
1; CHCl3). 1H NMR (CDCl3): δ ϭ 0.0Ϫ1.20 (br, 6 H, BH3),
1.35Ϫ1.56 (m, 4 H, CH2), 1.90Ϫ2.24 (m, 10 H, CH2, CHϪP), 2.36
(m, 2 H, CHϪP), 3.32 (s, 6 H, OCH3), 3.33 (s, 6 H, OCH3), 3.51
(m, 8 H, CH2O). 13C NMR (CDCl3): δ ϭ 15.8 (m, CH2P), 28.9
(CH2), 29.1 (CH2), 39.5 (m, CHϪP), 58.7 (OCH3), 58.7 (OCH3),
70.7 (CH2O), 71.6 (m, CH2O). 31P NMR (CDCl3): δ ϭ 40.4. IR
˜
(KBr): ν ϭ 3240, 2947, 2927, 2902, 2872, 2827, 2811, 2739, 2388,
2332, 2252, 1453, 1410, 1384, 1192, 1111, 1062, 958, 924, 773, 748,
701 cmϪ1
.
15a: 0.21 g (73% yield) of a syrup by column chromatography (n-
1
hexane/EtOAc, 4:1; Rf ϭ 0.30). H NMR (CDCl3): δ ϭ 1.25Ϫ1.51
(m, 8 H, CH2), 2.02Ϫ2.38 (m, 8 H, CH2, CHϪP), 3.40Ϫ3.60 (m,
8 H, CH2O), 4.40 (AB, 2Ja,b ϭ 12 Hz, 2 H, HaϪCH2Ph), 4.43 (AB,
2
13a: 1.35 g (46% yield) of a pale yellow oil by column chromato-
2Ja,b ϭ 12 Hz, 2 H, HbϪCH2Ph), 4.48 (AB, Ja,b ϭ 12 Hz, 2 H,
1
2
graphy (n-hexane/EtOAc, 4:1; Rf ϭ 0.35). H NMR (CDCl3): δ ϭ HaϪCH2Ph), 4.52 (AB, Ja,b
ϭ 12 Hz, 2 H, HbϪCH2Ph),
1.53Ϫ1.76 (m, 4 H, HaϪCH2), 2.15 (m, 2 H, HbϪCH2), 2.29 (m, 7.22Ϫ7.35 (m, 20 H, arom. H). 13C NMR (CDCl3): δ ϭ 19.1 (m,
2 H, HbϪCH2), 2.67 (m, 2 H, CHϪP), 2.80Ϫ2.97 (m, 4 H, CHϪP, CH2P), 31.3 (CH2), 31.4 (CH2), 40.0 (m, CHϪP), 43.7 (m, CHϪP),
2
CH2O), 3.45Ϫ3.65 (m, 6 H, CH2O), 4.01 (AB, Ja,b ϭ 12 Hz, 2 H,
70.2 (CH2Ph), 72.7 (CH2Ph), 72.9 (CH2O), 74.1 (CH2O),
HaϪCH2Ph), 4.15 (AB, 2Ja,b ϭ 12 Hz, 2 H, HbϪCH2Ph), 4.44 (AB, 127.4Ϫ128.3, 138. and 138.6 (arom. C). 31P NMR (CDCl3): δ ϭ
2Ja,b ϭ 12 Hz, 2 H, HaϪCH2Ph), 4.48 (AB, Ja,b ϭ 12 Hz, 2 H,
Ϫ6.9.
2
4622
Eur. J. Org. Chem. 2001, 4615Ϫ4624

Rh- and Ru-Catalyzed Enantioselective Hydrogenation
FULL PAPER
ganic Synthesis, Wiley, New York, 1994. Comprehensive Asym-
metric Catalysis (Eds.: E. N. Jacobsen, A. Pfaltz, H. Yamam-
oto), Springer, Heidelberg, 1999, vol. IϪIII. Catalytic Asym-
metric Synthesis (Ed.: I. Ojima), Wiley-VCH, New York, 2000.
See, e.g.: H. Brunner, W. Zettlmeier, Handbook of Enantioselec-
tive Catalysis with Transition Metal Compounds, vol. II (Li-
gands), VCH-Verlagsgesellschaft, Weinheim, 1993.
15b: 0.39 g (71% yield) of a syrup by column chromatography (n-
hexane/EtOAc, 3:2; Rf ϭ 0.35). H NMR (CDCl3): δ ϭ 1.13Ϫ2.46
1
(m, 16 H, CH2, CHϪP), 3.25(s, 6 H, OCH3), 3.27 (s, 6 H, OCH3),
3.41 (m, 8 H, CH2O). 13C NMR (CDCl3): δ ϭ 19.9 (m, CH2P),
31.1 (CH2), 31.2 (CH2), 39.7 (m, CHϪP), 43.3 (m, CHϪP), 58.6
(OCH3), 58.6 (OCH3), 70.6 (CH2O), 72.5 (CH2O). 31P NMR
(CDCl3): δ ϭ Ϫ7.0.
[2]
[3]
J. Ward, A. Börner, H. B. Kagan, Tetrahedron: Asymmetry
1992, 3, 849Ϫ852. A. Börner, J. Ward, W. Ruth, J. Holz, A.
Kless, D. Heller, H. B. Kagan, Tetrahedron 1994, 50,
10419Ϫ10430. A. Börner, J. Holz, A. Kless, D. Heller, U. Be-
rens, Tetrahedron Lett. 1994, 33, 6071Ϫ6074. J. Holz, A.
Börner, A. Kless, S. Borns, S. Trinkhaus, R. Selke, D. Heller,
Tetrahedron: Asymmetry 1995, 6, 1973Ϫ1988. D. Heller, J.
Holz, S. Borns, A. Spannenberg, R. Kempe, U. Schmidt, A.
Börner, Tetrahedron: Asymmetry 1997, 8, 213Ϫ222. R. Selke, J.
Holz, A. Riepe, A. Börner, Chem. Eur. J. 1998, 4, 769Ϫ771.
J. C. Jeffrey, T. B. Rauchfuss, Inorg. Chem. 1979, 18,
2658Ϫ2666. A. Bader, E. Lindner, Coord. Chem. Rev. 1991,
108, 27Ϫ110. J. J. Schneider, Nachr. Chem. 2000, 48, 614Ϫ620.
P. Braunstein, F. Naud, Angew. Chem. 2001, 113, 702Ϫ722;
Angew. Chem. Int. Ed. 2001, 40, 680Ϫ699.
S. Borns, R. Kadyrov, D. Heller, W. Baumann, A. Spannen-
berg, R. Kempe, J. Holz, A. Börner, Eur. J. Inorg. Chem. 1998,
1291Ϫ1295. S. Borns, R. Kadyrov, D. Heller, W. Baumann, J.
Holz, A. Börner, Tetrahedron: Asymmetry 1999, 10,
1425Ϫ1431.
J. Holz, R. Kadyrov, S. Borns, D. Heller, A. Börner, J. Or-
ganomet. Chem. 2000, 603, 61Ϫ68.
For a review on the effect of internal hydroxy groups on the
asymmetric hydrogenation with (phosphane)Rh catalysts, see:
A. Börner, Eur. J. Inorg. Chem. 2001, 327Ϫ337. For a more
general review about ambi-functional catalysts, see, e.g.: G. J.
Rowland, Tetrahedron 2001, 57, 1865Ϫ1882.
For a possible rationale on the role of the MeO group of the
chiral monodentate phosphane ligand MOP in several Pd-cata-
lyzed asymmetric reactions, compare: T. Hayashi, Acc. Chem.
Res. 2000, 33, 354Ϫ362.
General Procedure for the Preparation of the [{Bis-
(phospholane)}(COD)Rh]BF4 Complexes: To solutions of bis(phos-
pholanes) 13aϪc and 15a,b (1 mmol) in THF (3 mL) was added at
0 °C 1 equiv. of [Rh(COD)2]BF4 (406 mg) and the mixture was
stirred at ambient temperature for 1 h. Then the complex was pre-
cipitated by addition of ether (12 mL) and the solution was filtered.
The residue was washed twice with ether (4 mL) and dried under
vacuo to yield the rhodium complexes in yields of 40Ϫ70% as or-
ange-yellow powder. For the deprotection of the hydroxy groups in
complex of 13c the THP groups were removed by stirring the rhod-
ium complex (300 mg, 0.30 mmol) in MeOH (4 mL) with aq. HBF4
(0.1 mL) over a period of 24 h. After evaporation of the solvent
and addition of THF (2 mL), [Rh(COD)(16)]BF4 was precipitated
with ether (8 mL). Purification was carried out as described above.
Selected 31P NMR spectroscopic data are summarized in Table 1.
[4]
[5]
Crystal Structure Determinations for [Rh(COD)(13b)]BF4: Crystals
were obtained by slow diffusion of diethyl ether into a concentrated
MeOH solution of the rhodium complex. Diffraction data were
collected with an STOE-IPDS diffractometer using graphite-mon-
ochromated Mo-Kα radiation. The structure was solved by direct
methods (SHELXS-97)[30] and refined by full-matrix least-squares
techniques against F2 (SHELXL-97).[29] XP (Siemens Analytical X-
ray Instruments, Inc.) was used for structure representations. The
non-hydrogen atoms were refined anisotropically. The hydrogen
atoms were placed into theoretical positions and were refined by
using a riding model. The weighting scheme used in the last cycles
of refinement was w ϭ 1/[σ2(F2o) ϩ (0.0403P)2 ϩ 0.0000P]. Crystal
and refinement data of [Rh(COD)(13b)]BF4: empirical formula
C30H48BF4O4P2Rh, formula mass 724.34, crystal size 0.4 ϫ 0.3 ϫ
[6]
[7]
[8]
[9] [9a]
[9b]
M. J. Burk, J. Am. Chem. Soc. 1991, 113, 8518Ϫ8519.
M. J. Burk, J. E. Feaster, R. L. Harlow, Tetrahedron: Asym-
[9c]
metry 1991, 2, 569Ϫ592.
M. J. Burk, J. E. Feaster, W. A.
0.3 mm, space group P212121, orthorhombic, a ϭ 10.468(2), b ϭ
Nugent, R. L. Harlow, J. Am. Chem. Soc. 1993, 115,
[9d]
3
˚
˚
10125Ϫ10138.
M. J. Burk, M. F. Gross, T. G. P. Harper, C.
17.379(3), c ϭ 18.040(4) A, V ϭ 3281.9(11) A , Z ϭ 4, ρcalcd.
ϭ
S. Kalberg, J. Am. Chem. Soc. 1995, 117, 4423Ϫ4424. [9e] M. J.
Burk, M. F. Gross, T. G. P. Harper, C. S. Kalberg, J. R. Lee, J.
P. Martinez, Pure Appl. Chem. 1996, 68, 37Ϫ44. [9f] J. Albrecht,
1.466 g/cm3, µ (Mo-Kα) ϭ 0.674 mmϪ1, T ϭ 200 K, F(000) ϭ 1504,
Θ range for data collection 2.25Ϫ22.50 °, index range (h,k,l) Ϫ11/
11, Ϫ18/18, Ϫ18/7, reflections collected 11290, observed reflections
3593, refined parameters 379, R1 [2σ(I)] ϭ 0.0273, R1 (all data) ϭ
0.0300. Crystallographic data (excluding structure factors) for the
structure reported in this paper have been deposited at the Cam-
bridge Crystallographic Data Centre as supplementary publication
no. CCDC-172512. Copies of the data can be obtained free of
charge on application to CCDC, 12 Union Road, Cambridge,
CB2 1EZ, UK [Fax: (internat.) ϩ 44-1223/336-033]; E-mail:
deposit@ccdc.cam.ac.uk].
U. Nagel, Angew. Chem. 1996, 108, 444Ϫ446; Angew. Chem.
Int. Ed. Engl. 1996, 35, 407Ϫ409.
[9g]
M. J. Burk, J. G. Allen,
[9h]
W. F. Kiesmann, J. Am. Chem. Soc. 1998, 120, 657Ϫ663.
M. J. Burk, C. S. Kalberg, A. Pizzano, J. Am. Chem. Soc. 1998,
120, 4345Ϫ4353. [9i] M. J. Burk, N. B. Johnson, J. R. Lee, Tetra-
hedron Lett. 1999, 40, 6685Ϫ6688.
[10] [10a] PennPHOS: Q. Jiang, Y. Jiang, D. Xiao, P. Cao, X. Zhang,
Angew. Chem. 1998, 110, 1203Ϫ1207; Angew. Chem. Int. Ed.
1998, 37, 1100Ϫ1103. [10b] CnrPHOS: A. Marinetti, J. P. Genet,
ˆ
S. Jus, D. Blanc, V. Ratovelomanana-Vidal, Chem. Eur. J. 1999,
5, 1160Ϫ1165. A. Marinetti, F. Labrue, J. P. Genet, Synlett
ˆ
ˆ
1999, 12, 1975Ϫ1977. A. Marinetti, S. Jus, J. P. Genet, Tetra-
hedron Lett. 1999, 40, 8365Ϫ8368. A. Marinetti, S. Jus, J. P.
Acknowledgments
ˆ
Genet, L. Ricard, Tetrahedron 2000, 56, 95Ϫ100.
[11]
The authors are grateful for the financial support provided by the
BASF AG (Ludwigshafen), the Degussa AG (Frankfurt) and the
Fonds der Chemischen Industrie. We appreciate the collaboration
of Dr. A. Spannenberg and Dr. C. Fischer. It is a pleasure for us
to acknowledge skilled technical assistance by Mrs. G. Wenzel,
Mrs. C. Pribbenow, and Mrs. K. Romeike.
M. J. Burk, A. Pizzano, J. A. Martin, Organometallics 2000,
19, 250Ϫ260. S. Feldgus, C. R. Landis, J. Am. Chem. Soc. 2000,
122, 12714Ϫ12727. C. R. Landis, S. Feldgus, Angew. Chem.
2000, 112, 2985Ϫ2988; Angew. Chem. Int. Ed. 2000, 39,
2863Ϫ2866.
See, however, the controversial discussion about the effect of
H-atoms in 2,5-position of the phospholane ring: E. Fernan-
dez, A. Gillon, K. Heslop, E. Horwood, D. J. Hyett, A. G.
Orpen, P. Pringle, Chem. Commun. 2000, 1663Ϫ1664.
H. Brunner, R. Sievi, J. Organomet. Chem. 1987, 328, 71Ϫ80.
[12]
[13]
[1]
Asymmetric Synthesis (Ed.: J. D. Morrison), Academic Press,
Orlando, 1985, vol. 5. R. Noyori, Asymmetric Catalysis in Or-
Eur. J. Org. Chem. 2001, 4615Ϫ4624
4623

J. Holz, A. Börner et al.
FULL PAPER
[21]
[22]
[23]
[24]
H. Brunner, G. Net, Z. Naturforsch., Teil
1210Ϫ1212.
B
1996, 51,
N. Machinaga, C. Kibayashi, J. Org. Chem. 1992, 57,
5178Ϫ5189.
A. Marzi, P. Minetti, D. Misiti, Tetrahedron 1992, 48,
10127Ϫ10132.
Review: M. Ohff, J. Holz, M. Quirmbach, A. Börner, Synthesis
1998, 1391Ϫ1415.
H.-J. Drexler, W. Baumann, A. Spannenberg, C. Fischer, D.
Heller, J. Organomet. Chem. 2001, 621, 89Ϫ102.
A. Börner, D. Heller, Tetrahedron Lett. 2001, 42, 223Ϫ225.
For β-branched unsaturated amino acid derivatives, however,
the BPE-Rh catalyst was found to be superior: M. J. Burk, M.
F. Gross, J. P. Martinez, J. Am. Chem. Soc. 1995, 117,
9375Ϫ9376.
[14] [14a] J. Holz, M. Quirmbach, U. Schmidt, D. Heller, R. Stürmer,
[14b]
A. Börner, J. Org. Chem. 1998, 63, 8031Ϫ8034.
W. Li, Z.
Zhang, D. Xiao, X. Zhang, Tetrahedron Lett. 1999, 40,
[14c]
6701Ϫ6704.
Y.-Y. Yan, T. V. RajanBabu, J. Org. Chem.
[14d]
2000, 65, 900Ϫ906.
W. Li, Z. Zhang, D. Xiao, X. Zhang,
J. Org. Chem. 2000, 65, 3489Ϫ3496. [14e] Y.-Y. Yan, T. V. Rajan-
Babu, Org. Lett. 2000, 2, 199Ϫ202.
[25]
[26]
[15]
Recently, Brown and co-workers gave proof, that DuPHOS-
DiPAMP hybrids with hydroxy or methoxy groups attached
to the phospholane in 3,4-position feature significantly higher
intrinsic stereoselectivity than the parent ligand: D. Carm-
ichael, H. Doucet, J. M. Brown, Chem. Commun. 1999,
261Ϫ262.
J. Holz, D. Heller, R. Stürmer, A. Börner, Tetrahedron Lett.
1999, 40, 7059Ϫ7062.
Reviews for the synthesis of hydroxyphosphanes, see: J. Holz,
M. Quirmbach, A. Börner, Synthesis 1997, 983Ϫ1006. J. Holz,
M. Quirmbach, L. Götze, A. Börner, Rec. Res. Develop. Org.
Chem. 1998, 2, 685Ϫ711.
[27]
[28]
D. Heller, J. Holz, H.-J. Drexler, J. Lang, H.-P. Krimmer, K.
Drauz, A. Börner, J. Org. Chem., 2001, 66, 6816Ϫ6817.
[16]
[17]
ˆ
J. P. Genet, C. Pinel, V. Ratovelomanana-Vidal, S. Mallart, X.
Pfister, L. Bischoff, M. C. Cano De Andrade, S. Darses, C.
Galopin, J. A. Laffitte, Tetrahedron: Asymmetry 1994, 5,
675Ϫ690.
[29]
[30]
G. M. Sheldrick, SHELXS-97, University of Göttingen, Ger-
many, 1997.
[18]
[19]
R. S. Tipson, A. Cohen, Carbohydr. Res. 1968, 7, 232Ϫ243.
E. J. Corey, P. B. Hopkins, Tetrahedron Lett. 1982, 23,
1979Ϫ1982.
G. M. Sheldrick, SHELXL-97, University of Göttingen, Ger-
many, 1997.
Received July 3, 2001
[O01325]
[20]
A. H. Haines, Carbohydr. Res. 1965, 1, 214Ϫ228.
4624
Eur. J. Org. Chem. 2001, 4615Ϫ4624