Modular P-chirogenic aminophosphane-phosphinite ligands for Rh-catalyzed asymmetric hydrogenation: A new model for prediction of enantioselectivity

Article abstract of DOI:10.1002/ejoc.200600966

An original series of P-chirogenic aminophosphane-phosphinite (AMPP) ligands has been synthesized from (+)- or (-)-ephedrine in 23 to 61 % overall yields by a versatile three-step methodology. The AMPP ligands, bearing either one or two P-chirogenic cente

Full text of DOI:10.1002/ejoc.200600966

FULL PAPER
DOI: 10.1002/ejoc.200600966
Modular P-Chirogenic Aminophosphane-Phosphinite Ligands for Rh-Catalyzed
Asymmetric Hydrogenation: A New Model for Prediction of Enantioselectivity
Christophe Darcel,*[a] Dominique Moulin,[b] Jean-Christophe Henry,[c] Michaël Lagrelette,[a]
Philippe Richard,[a] Pierre D. Harvey,[d] and Sylvain Jugé*[a]
Keywords: Rhodium / P ligands / Aminophosphane-phosphinites / Hydrogenation / Asymmetric catalysis
An original series of P-chirogenic aminophosphane-phos-
phinite (AMPP) ligands has been synthesized from (+)- or
(–)-ephedrine in 23 to 61% overall yields by a versatile three-
step methodology. The AMPP ligands, bearing either one or
two P-chirogenic centers, were used in the form of rhodium
rhodium-substrate complex, has been proposed. This model
offers an alternative to the quadrant rule, well adapted to
the C2-symmetry ligands and the chair conformation of their
complex derivatives. In this work, the model, which schema-
tizes the front side of the complex as a sextant in the direction
complexes for the catalyzed hydrogenation of α-acet- of the cardinal points, fits with coordination of the substrate
amidocinnamate as a test reaction. Notably, even with AMPP
ligands all derived from (+)-ephedrine, variation of the sub-
stituent on a P-center allowed the phenylalanine derivatives
to be obtained in either (S) or (R) absolute configurations,
with ee values ranging from 99% (S) to 88% (R). The asym-
metric induction was analyzed with the aid of X-ray struc-
tures of AMPP complexes, and a new model for the enantio-
selectivity, taking into consideration the boat conformation
and the steric and electronic dissymmetries at the dihydride
by the acetamido and the cinnamyl groups in the north and
east (or west) parts, respectively. The enantioselectivity origi-
nates from the ligand residues located at the south-east or
south-west parts of the dihydride rhodium intermediate.
Computer modeling on several AMPP-rhodium complexes
with PCModel confirms the proposed predicting model.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
properties.[2,3,6] To improve an asymmetric catalytic reac-
tion for a new substrate is not trivial, however, possible ap-
proaches being mainly restricted either by the availability
of a library of chiral ligands,[3,6] or to the easy structure
modification of an efficient lead series such as BINAP,[16,17]
DUPHOS,[18] PHOX,[19] monophos,[20] etc. In particular,
the phosphorus and phosphinous derivatives usually
prepared by treatment of PCl3 or chlorophosphanes with
available chiral diols,[21,22] binaphthol,[23–25] or amino
alcohols[26–31] offer an efficient and easy route to new chiral
ligands. Notably, the phosphorus ligands are by far the
most commonly used ones,[16-31] because they offer far more
Introduction
Chiral transition metal complexes have given consider-
able impetus to asymmetric catalysis, which is today one of
the most efficient methods for the synthesis of enantiomer-
ically enriched compounds.[1] Indeed, several reactions in-
volving carbon–hydrogen,[2–6] carbon–carbon,[7–10] or car-
bon–heteroatom[11–15] bond formation provide highly asym-
metric induction for various substrates, and some of them
are applied on industrial scales for the production of sub-
stances with useful agrochemical, flavor, or pharmaceutical
[a] Laboratoire de Synthèse et d’Électrosynthèse Organométal- possibilities for modification of the structure (mono-, di-,
liques (LSEO), UMR CNRS 5188, Université de Bourgogne,
multidendate, hemilabile, hybrids, etc.), the symmetry (C2,
dissymmetric, etc.), the basicity, and the chirality at the
phosphorus atom.
9 Avenue A. Savary, BP 47870, 21078 Dijon, France
Fax: +33-3-80396113
E-mail: Christophe.Darcel@u-bourgogne.fr
Sylvain.Juge@u-bourgogne.fr
Prediction of the enantioselectivity of the hydrogena-
tion – catalyzed by rhodium complexes – of a substrate such
as 1 has previously been possible with the aid of the quad-
rant diagram rule[32,33] initially proposed by Knowles, but
recently reformulated by other research groups using nu-
merous other C2-symmetry ligands.[33,34] The quadrant dia-
gram describes the chiral environment of the rhodium atom
as four areas, in order to take account of the relative bulki-
ness of the ligand substituents, which may be placed at axial
or equatorial positions as a result of the twisted-chair con-
[b] BASF-Aktiengesellschaft, GCI/B-M 311,
67056 Ludwigshafen, Germany
Fax: +49-621-6059144
E-mail: dominique.moulin@basf-ag.de
[c] SYNTHELOR SAS,
102 impasse H. Becquerel, Dynapôle de Ludres-Fléville, 54510
Ludres, France
Fax: +33-3-83156735
E-mail: jc.henry@synthelor.com
[d] Département de Chimie de l’Université de Sherbrooke,
Sherbrooke J1K2R1 Québec, Canada,
Fax: +819-821-8017
E-mail: Pierre.Harvey@USherbrooke.ca
2078
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2078–2090

Ligands for Rh-Catalyzed Hydrogenation
FULL PAPER
Scheme 1.
formation in the complex (Scheme 1). It is therefore pos- hydrogenation have previously mainly been studied with P-
sible to predict the absolute configuration of the hydroge- chirogenic C2 symmetry diphosphanes,[33,47–54] with a few
nated product 2[34] from the relative bulks of the substitu- examples of dissymmetric ligands having been men-
ents at the corners (Scheme 1).
tioned.[55]
Recently, Vogt’s team and our own group have indepen-
While the quadrant rule helps to explain the enantio-
selectivity, prediction of the architecture of the rhodium dently described a series of modified P-chirogenic EPHOS
complex and its stereocontrol in the catalytic hydrogenation ligands 4 (Scheme 3) for catalytic asymmetric hydrogena-
of a given substrate is currently still highly speculative. Nat- tion and hydroformylation reactions involving rhodium
urally, much work has been focused on the origin of the complexes.[56–58] These ligands, preparable from ephedrine
enantioselectivity of such reactions, and for a long time it as starting material by a versatile methodology, offer the
was considered that the selectivity does not stem from the potential for modification of the substituents on one or two
relative stabilities of the substrate–rhodium complexes, but P-centers. We now wish to report novel modified P-chiro-
from their rate constants for the oxidative addition of dihy- genic EPHOS ligands 4, together with their use in rhodium-
drogen (unsaturated route).[36–44] However, recent experi- catalyzed hydrogenations of methyl α-acetamidocinnamate
mental and computational data have provided support for 1 (R = Ph; RЈ,RЈЈ = Me) as a test reaction. Although the
a second reaction pathway involving a fast equilibrium of ligands all originate from the same (+)-ephedrine backbone,
the dihydride–substrate complexes 3, followed by the stereo- here we demonstrate for the first time that it is possible to
determining migration insertion step[35,45–52] (Scheme 2). obtain the hydrogenated product 2 in either the (R) or the
Finally, the stereoselectivities of the unsaturated and the di- (S) configuration with excellent ee values, simply by chang-
hydride pathways are related, even though in the first, sub- ing the substituent on a P-center. The asymmetric induc-
strate coordination occurs before the oxidative addition of tions have been analyzed with the aid of X-ray data for
dihydrogen, and in the second, substrate coordination related AMPP 4 complexes, allowing us to propose a new
occurs after dihydrogen addition.
prediction model for the enantioselectivity, based on varia-
tion of the P-substituents.
Results and Discussion
The modified EPHOS ligands (AMPP) 4 were prepared
from the oxazaphospholidine borane 5 (Scheme 3), which
was itself obtained in a one-pot reaction from (+)-eph-
edrine.[59] Compound 5 was treated with organolithium rea-
gents to induce ring-opening reactions to afford products 6.
These were subsequently trapped in situ with achiral chlo-
Scheme 2.
rophosphanes R2 PCl, and then with borane, to provide the
2
From a mechanistic point of view, P-chirogenic ligands AMPP diborane complexes 7a–l in 30 to 79% yields
have attracted much interest for understanding of the inter- (Scheme 3, a). This methodology allows the introduction of
actions occurring between the prochiral substrate and the various R1 substituents – such as alkyl, cycloalkyl, aryl, or
catalyst, thanks to the potential offered by the presence of ferrocenyl – onto the aminophosphane residue (Table 1, En-
different substituents on the phosphorus atoms. Thus, the tries 1–12). The AMPP diborane complexes 7 are easily
P-chirogenic center allows a more sterically and electroni- purified and stored, while their decomplexation can be ac-
cally defined architecture about the metal center. As the complished by heating them in toluene at 50–60 °C in the
dihydride substrate complex 3 also possesses a source of presence of DABCO, to afford the corresponding dia-
chirality at the metal center, it is thus interesting to investi- stereomerically pure AMPP ligands 4a–l with retention of
gate whether the diastereoselectivity observed at this step configuration on the chiral phosphorus atom (Table 1). The
can be explained in terms of host–guest interactions (guest free AMPP 4a–l ligands were readily purified by filtration
= hydrogen or substrate), in order to correlate the ligand through neutral alumina, and their diastereomeric purities
with the enantioselectivity. Mechanistic aspects of catalyzed were checked by 31P NMR spectroscopy.
Eur. J. Org. Chem. 2007, 2078–2090
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2079

C. Darcel, S. Jugé, et al.
FULL PAPER
Scheme 3.
Table 1. P-chirogenic aminophosphane-phosphinites (AMPP) 4 and their diborane complexes 7, synthesized from either (+)- or (–)-
ephedrine.
Entry
Ephedrine
R1
R2
R3
AMPP (BH3)2 7
AMPP 4 (or 4Ј)
yield[a] (%)
yield[b] (%)
1
2
3
4
5
6
7
8
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(–)
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
q
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
C6H11
OPh
Ph
Ph
o-An
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
63
62
66
62
62
78
40
64
79
38
67
30
18[c]
55
55
65
40
89
91
88
86
99
93
97
88
42
97
o-An
o-MEMPh
1-Np
Fc
2-Np
o-biPh
t-Bu
Me
C6H11
o-An
o-An
Me
9
Ph
Ph
10
11
12
13
14
15
16
17
[d]
C6H11
OPh
Me
o-An
Ph
o-An
Ph
–
90
[d]
–
Ph
Ph
o-An
o-An
94
99
99
[d]
o-An
–
[a] Isolated chemical yield from starting complex 5. [b] Isolated yield. [c] Not optimized. [d] Air-sensitive ligand used without purification.
When the ring-opened products 6 are treated with a P- Crystal Structure of the AMPP Diborane 7j
chirogenic chlorophosphane borane 8[60] and then decom-
plexed with DABCO, the corresponding AMPP ligands
4m–p, featuring additional chirality at the phosphinite frag-
ment, are obtained (Scheme 3, b; Table 1, Entries 13–16).
Interestingly, the epimeric P-chirogenic AMPP ligands 4n
and 4o are obtained by use of the two enantiomers of the
o-anisylchlorophenylphosphane borane 8, which are also
prepared independently from (+)- or (–)-ephedrine (Table 1,
The solid-state structure of the AMPP diborane 7j was
established by X-ray analysis. The compound presents as an
unfolded conformation with a flattened nitrogen pyramidal
structure and the P–B bonds disposed anti from one an-
other (Figure 1). The P1 unit cell contains two independent
molecules with similar conformations. The (S) absolute
configuration of the cyclohexylaminophosphane borane
fragment is consistent with the retention of configuration
Entries 14, 15).
In addition, the AMPP 4Јq (R1,R2 = o-An; R3 = Ph),
previously reported for the ring-opening cleavage of the
featuring both P-chirogenic aminophosphane and phos-
starting complex 5[56,60] on treatment with an organo-
phinite residues, was obtained by the same methodology,
lithium reagent. Interestingly, the C34 and C30 methylene
but starting from (–)-ephedrine, followed by the trapping of
groups point towards the same side of the P–B2 bond,
the ring-opened intermediate 6Ј with the (S)-o-anisylchloro-
which can be explained by a weaker interaction of the C29–
H bond with the methyl C22 located by the nitrogen (Fig-
phenylphosphane borane 8[56,60] (Table 1, Entry 17).
ure 1). Evidence that this methylamino group plays a key
role in determining the conformation of the alkyl R1 sub-
stituent in the coordinated AMPP framework is provided
below.
2080
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2078–2090

Ligands for Rh-Catalyzed Hydrogenation
FULL PAPER
Table 2. Rhodium-catalyzed hydrogenation of methyl α-acet-
amidocinnamate (1) in the presence of AMPP ligands 4a–j bearing
P-chirogenic aminophosphane fragments.
Entry
AMPP
Conditions[a]
Hydrogenated product 2
R1
Solvent Time Yield
ee (%) Absol.
[c]
(h)
(%)[b]
conf.
1
2
3
4
5
6
7
8
4a
4a
4b
4b
4b
4b
4b
4c
4d
4d
4e
4f
Ph
Ph
C6H6
22
95
98
98
99
91
95
94
94
98
99
98
96
95
95
95
95
46
11
99
89
88
97
92
99
95
88
87
16
16
2
(S)
(S)
(S)
(S)
(S)
(S)
(S)
(S)
(S)
(S)
(S)
(S)
(S)
–
CH2Cl2 18
C6H6 20
CH2Cl2 10.5
MeOH
EtOH
iPrOH
o-An
o-An
o-An
o-An
o-An
62.5
41
40
36
17
4
Figure 1. ORTEP representation of AMPP-diborane complex 7j.
For clarity, only one independent molecule is shown. Selected bond
lengths [Å], angles [°], and dihedral angles [°] for one of the two
molecules: P1–B1 1.904 (2), P2–B2 1.915 (3), P2–N1 1.6697 (17),
P1–O1 1.6111 (14); C20–N1–P2 121.83 (13), C22–N1–P2 121.21
(14); C22–N1–C20 116.58 (16), B2–P2–N1–C22 173.99 (18), B2–
P2–C23–C28 3.58 (24), O1–C13–C20–N1 172.87 (15), B2–P2–C29–
C30 58.63 (18), B2–P2–C29–C34 65.82 (17). The thermal ellipsoids
are drawn at 50% probability. The H atoms are not shown for
clarity.
o-MEMPh C6H6
9
1-Np
1-Np
Fc
2-Np
o-biPh
t-Bu
Me
C6H6
CH2Cl2
C6H6
CH2Cl2 4.5
C6H6
CH2Cl2
CH2Cl2
C6H6
10
11
12
13
14
15
16
47
4g
4h
4i
60
4
3
22
80
(R)
(R)
4j
C6H11
63
[a] 0.5 mmol substrate, 3% catalyst, H2 pressure: 15–20 bar. [b] Iso-
lated yield. [c] Determined by HPLC with a Chiralcel OD column.
Enantioselective Rhodium-Catalyzed Hydrogenations
The AMPP ligands 4a–4Јq were used for rhodium-cata-
lyzed asymmetric hydrogenation of the methyl α-acet-
amidocinnamate 1 (R = Ph; RЈ, RЈЈ = Me; Scheme 4), and
the results are reported in Table 2 and Table 3.
4f, 4g), the catalysis gave lower asymmetric induction, the
(S)-phenylalanine derivative 2 being obtained with only
16% ee (Entries 12, 13). On the other hand, when the sub-
stituent R1 was an alkyl group, such as tert-butyl, methyl or
cyclohexyl, the corresponding AMPP ligands 4h–j afforded
either the racemic product or the (R) product 2 with ee
values of 22 and 80%, respectively (Entries 14–16).
Interestingly, in the cases of AMPP ligands 4a–j, each
bearing an alkyl or an aryl R1 substituent on the amino-
phosphane residue, the catalyzed hydrogenation provided
the phenylalanine derivative 2 with either (R) or (S) abso-
Scheme 4.
When the catalysis was performed in the presence of the lute configuration.
EPHOS 4a in benzene or in CH2Cl2, the (S)-phenylalanine
In addition, the catalysis was also performed with AMPP
derivative 2 was obtained with ee values of 46 or 11%, ligands bearing phosphite or P-chirogenic phosphinite frag-
respectively (Table 2, Entries 1, 2). Under similar catalytic ments (Table 3). If the diphenylphosphinite residue of the
conditions and in various polar or nonpolar solvents, the ligand 4a was changed for a diphenylphosphite, the hydro-
ligand AMPP 4b, bearing an o-anisyl group as R1 substitu- genation reaction again afforded the product 2 with the (S)
ent, provided the product (S)-2 with 88 to 99% ee (En- configuration, but with 71% ee (ligand 4l; Table 3, Entry 1).
tries 3–7). If the substituent R1 was o-MEMPh, 1-naphthyl, In the case in which the AMPP ligand featured a P-chiro-
or ferrocenyl (4c–e), the asymmetric hydrogenation pro- genic phosphinite group with o-anisyl as R3 substituent (li-
ceeded with ee values of 87 to 99% (Entries 8–11), while in gand 4n), the catalysis gave the hydrogenation product (S)-
the case of the 2-naphthyl or o-biphenyl substituents (i.e., 2 with only 35% ee (Entry 2), but the epimeric ligand 4o,
Table 3. Rhodium-catalyzed hydrogenation of methyl α-acetamidocinnamate (1) in the presence of AMPP ligands 4l–4Јq bearing phos-
phite or P-chirogenic phosphinite groups.
Entry
AMPP
Conditions[a]
Solvent
Hydrogenated product
Yield (%)[b] ee (%)[c]
R1
R2
R3
Time (h)
Absol. conf.
1
2
3
4
5
6
4l[d]
4n
o-An
Ph
Ph
Ph
o-An
o-An
OPh
Ph
o-An
o-An
Ph
OPh
o-An
Ph
Ph
o-An
Ph
C6H6
C6H6
C6H6
CH2Cl2
C6H6
24
23
24
85
21
12
60
96
95
95
95
94
71
35
62
88
75
1
(S)
(S)
(R)
(R)
(S)
–
4o
4o
4p
4Јq[e]
o-An
CH2Cl2
[a] 0.5 mmol substrate, 3 mol-% catalyst, H2 pressure: 15–20 bar. [b] Isolated yield. [c] Determined by HPLC with a Chiralcel OD column.
[d] Presence of minor impurity. [e] Prepared from (–)-ephedrine.[56]
Eur. J. Org. Chem. 2007, 2078–2090
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2081

C. Darcel, S. Jugé, et al.
FULL PAPER
with o-anisyl as R2 substituent, provided the product 2 with double bonds. The dihedral angles between the planes de-
the (R) configuration and with 62 to 88% ee, depending on fined by P1–Rh1–P2 and either C2–Rh1–C6 or C1–Rh1–
the solvent used (Entries 3, 4). As observed for the ligands C5 are 11.02 (12)° and 40.41 (15)°, respectively.
4a, 4b, and 4d, this result confirms increasing induction in
Interestingly, the seven-membered Rh-AMPP chelate
favor of either the (S) or the (R) enantiomer of 2 in benzene ring adopts a boat-like conformation, with a slight distor-
or dichloromethane, respectively, as reaction solvents tion of –21.83 (15)° for the O1–P1–P2–N1 dihedral angle,
(Table 2, Entries 1–4, 9–10; Table 3, Entries 3, 4). Finally, with the substituents of the ephedrine backbone (i.e., Ph,
the ligands 4p and 4Јq, possessing two P-chirogenic centers Me) in the equatorial positions. As a consequence, the H–
with o-anisyl as R1 and R3 (or R2) substituents, afforded C28 bond points in the direction of the rhodium atom with
the hydrogenated product 2 either with (S) configuration a H···Rh distance of 2.641 Å. It should be noted that the
with 75% ee, or in racemic form (Table 3, Entries 5, 6).
sum of the angles around the N-atom (i.e., 356.6°) indicates
a flattened geometry, which places the C30 methyl group
on a side face of the tetrahedral phosphorus center, and
in the anti position in relation to the P2–Rh bond. This
conformation features the C15 and C38 phenyl phosphorus
substituents in axial positions, but the distortion of the che-
lated ring induces an obvious facial dissymmetry for the
complex. In addition, the presence of the C30 methyl group
and cyclooctadiene ligand forces the methoxy group to be
situated at the anti position in relation to the P2–C38 axial
bond. Finally, the planes of the o-anisyl and the C9-phenyl
groups form an angle of 31.43°, with both faces oriented
toward the metal center.
Crystal Structure of [Rh(COD)AMPP 4b)]BF4(9)
The two (+)-ephedrine-derived AMPP ligands 4b and 4o
afforded both the (S) and the (R) forms of product 2 with
ee values of 99 and 88%, respectively (Table 2, Entry 3;
Table 3, Entry 4). This result illustrates the importance of
the P-center chirality with respect to the ephedrine back-
bone.
In order to interpret the influence of the chiral rhodium
environment on the enantioselectivity, complex 9 was pre-
pared from [Rh(COD)]2BF4 and the ligand 4b and then
characterized by X-ray crystallography. The crystal struc-
ture revealed the existence of two independent molecules
with very similar conformations in the unit cell. Figure 2
shows a perspective view of one of the two molecules, to-
gether with the numbering scheme. The coordination sphere
about the rhodium atom is a distorted square planar struc-
ture, with a P1–Rh1–P2 bite angle of 88.54 (4)° and two
vertices occupied by the 1,5-cyclooctadiene η2-bonded
Model for the Catalyzed Asymmetric Hydrogenation
Metal-chelate rings forming “quasi”-boat structures have
[62]
also been observed in the cases of RuCpCl[61] and PdCl2
complexes of EPHOS 4a and its modified P-chirogenic de-
rivatives 4e and 4g, respectively. This observed conforma-
tion was explained by several features. Firstly, the methyl
and the phenyl groups situated on the ephedrine backbone
were both found in equatorial positions. Secondly, the R1
group located on the aminophosphane residue is sterically
larger than a phenyl one, which favors the equatorial posi-
tions. Lastly, the phenyl groups positioned on the two phos-
phorus atoms do not exhibit intersubstituent interactions at
the axial positions, owing to the presence of oxygen and
nitrogen atoms on the chelating ring. Simple computer
modeling (PC-Model)[63] performed for the model com-
pound Rh(AMPP 4b)Cl2 (in the gas phase) has confirmed
that this “quasi”-boat structure is more stable than the
“quasi”-chair one (by about 10 kJmol–1).
Although the origin of the catalytic asymmetric induc-
tion needs to be carefully explained with the aid of crystal
structures,[37,42,54,64] the observation of the same boat con-
formation regardless of the metal and other ligands (Cl or
Cp) used suggests a remarkable stability of this geometry
during the catalytic process. Consequently, it may reason-
ably be assumed that this boat conformation was also re-
tained during the dihydride stereodetermination step.
In order to explain the enantioselectivity of the hydro-
genation, a proposed model of the dihydride Rh complex
10 showing the front view as a sextant has been designed
(Scheme 5). This model takes account of the boat confor-
mation and the steric and the electronic dissymmetries of
the complex in the direction of the cardinal points.
Figure 2. ORTEP representation of the [(COD)Rh(AMPP 4b)]BF4
complex 9. For clarity, only one molecule is shown. Selected bond
lengths [Å], angles, and dihedral angles [°] for one of the two mole-
cules: P1–O1 1.619 (3), P2–N1 1.666 (3), P1–Rh1 2.2773 (10), P2–
Rh1 2.3229 (10), Rh1–HC28 2.641; P1–Rh2–P2 88.54 (4); O1–P1–
P2–N1 21.83 (15), P1–O1–C21–C22 177.96 (23), P1–O1–C28–C29
139.87 (45), phenyl planes C15–C20 and C38–C43 14.00 (21),
phenyl–anisyl planes C9–C14 and C31–C36 31.43 (21). The ther-
mal ellipsoids are drawn at 50% probability. The H atoms are not
shown for clarity, except for those on the C21 and C28 carbon
atoms.
2082
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2078–2090

Ligands for Rh-Catalyzed Hydrogenation
FULL PAPER
Scheme 5. Only the groups relevant to the Rh–substrate interactions are shown.
From Figure 2 and Scheme 5 it is also easy to see that
the coordination of the C=C fragment of the substrate at
the north or south positions of the complex 9 produces se-
vere steric hindrance (as verified by computer modeling).
Instead, coordination of the C=C on the east or west side
of the complex is preferred in this model.
The north-east and north-west parts are thus sterically
hindered by the presence of the two axial phenyl phospho-
Figure 3. Computer model for the [Rh(AMPP 4b)H2 (substrate 1)]
intermediates 10a-re (left) and 10b-re (right) with the C=C
branched on the west side, showing plausible hydride nucleophilic
attacks on the re face of the substrate, both resulting in the same
observed (S) product 2.
rus substituents, almost parallel between them and perpen-
dicular to the complex front view. As the mean distance
between these two axial phenyl planes (as crystallographi-
cally determined on the rhodium complex 9) is 3.42 Å, it is
likely that the north position could be coordinated only by
one hydride or the substrate acetamido group. Computer
modeling shows that these axial phenyl groups should easily
As many monohydride rhodium-substrate complexes –
“open” thanks to the flexibility of the skeleton backbone unlike dihydride rhodium species[38,51] – have been reported
of the seven-membered ring chelate, to allow coordination in the literature, it appears reasonable to postulate that the
of these latter groups onto the Rh metal.
intramolecular hydrogen transfer in the intermediates 10
Computer modeling was performed on all four may occur as a fast step. In order to corroborate this hy-
[Rh(AMPP 4b)H2 (substrate 1)] intermediates, with the pothesis, the interactions between the substrate and the rho-
C=C donor of the substrate coordinated either in the west dium dihydride as host complex were analyzed with respect
or in the east positions of the complex (for a total of eight to the observed asymmetric induction.
intermediates; the methoxy group of the anisyl substituent
Thus, when R1 is a larger aryl group (i.e., than a phenyl),
was left as is in the structure 9 shown in Figure 2). From the dissymmetry in the east-west region of the rhodium
molecular distortion arguments (bond lengths and com- complex is amplified, which must promote the coordination
puted energies), two intermediates turned out to be reason- of the cinnamyl group on the west part and on its re face
able (10a-re and 10b-re with the re face of the C=C (Scheme 5). A facile hydrogen insertion could therefore oc-
branched on the west side) and these are shown in Figure 3. cur between the two aryl planes (i.e., the phenyl group of
The most likely nucleophilic attack of the hydride onto the the cinnamyl and the R1 group) to provide the intermediate
activated substrate should come from that situated parallel 10b-re as major product. This model explains why all the
to the C=C axis. In such a case the same (S) product 2 results obtained for the EPHOS 4a and its derivatives 4b–g
should be obtained. However, because the computed Rh–C involve the (S)-phenylalanine derivative 2 (Table 2, En-
distance in 10b-re is shorter than that of 10a-re (due to ste- tries 2–13). On this basis, when the steric hindrance of the
ric hindrance), 10b-re is sterically the most likely intermedi- R1 aryl group is increased, the coordination of the cinnamyl
ate, explaining the observed product.
in the west region of the intermediate 10b-re, and conse-
Eur. J. Org. Chem. 2007, 2078–2090
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2083

C. Darcel, S. Jugé, et al.
FULL PAPER
quently the asymmetric induction towards the (S) enantio-
mer, is still more favored. Whereas good to excellent ee
values were obtained with the AMPP ligands 4b–e (R1 = o-
An, o-MEMPh, 1-Np, Fc), the lower asymmetric induction
observed in the case of ligands 4f and 4g could be explained
in terms of similar plane conformation of the 2-naphthyl
or biphenyl substituents with regard to the phenyl group,
producing results comparable to those obtained with
EPHOS 4a (Table 2, Entries 1, 12, 13).
On the other hand, when the AMPP ligand 4 bears an
alkyl group as the R1 substituent, the insertion of the hy-
drogen between the cinnamyl and this group is now disfa-
vored for steric reasons, resulting in the coordination of the
substrate on the east region and its si face, as illustrated in
the case of the intermediate 10b-si (Scheme 5, Scheme 6).
Figure 4. Computer model for the [Rh(AMPP 4i)H2 (substrate 1)]
intermediates 10b-re (left) and 10b-si (right) with C=C branched
on the east and west sides, respectively, showing plausible hydride
nucleophilic attacks on the C=C bond, both resulting in the (S)
and the (R) product 2, respectively.
In this case, the steric interactions with the CH vinylic product 2 (Scheme 5; Table 3, Entry 1). If the AMPP ligand
group must be weaker. It should be noted that the methyl now bears an o-anisyl group as R3 substituent, the cata-
C30 group located on the amino residue, which occupies lyzed hydrogenation again affords the (S) product 2, but
the anti position in relation to the P2–Rh bond, probably with only 35% ee (ligand 4n, Table 3, Entry 2). The weak
contributes to the steric hindrance by forcing the alkyl R1 influence of the substitution at this R3 position could be
group to be located on the front side of the complex (Fig- explained by the o-anisyl conformation, the methoxy group
ure 2, Scheme 6). These steric effects explain why the meth- being positioned at the back side of the phosphorus-metal
yl- and cyclohexyl-containing AMPP ligands 4i and 4j pro- bond, thus affording steric hindrance comparable with that
vide the phenylalanine derivative (R)-2 with ee values of 22 of a phenyl group at the front side of the catalytic site. This
and 88%, respectively (Table 2, Entries 15, 16). Computer result is in good agreement with those obtained with the
modeling using the R1 = Me group (instead of anisyl), for ligand EPHOS 4a, also bearing two phenyl groups at the
example, confirms this trend. Indeed, the two most prob- R1 and R3 positions (46% ee; Table 2, Entry 1). Moreover,
able intermediate structures (from the lowest computed en- this result also compares favorably with those obtained with
ergy) promote both the (S) (10 b-re; most stable) and (R) the ligands 4b (R1 = o-An, R3 = Ph) and 4p (R1 = o-An,
(10b-si; second most stable) products (Figure 4). However, R3 = o-An), which provide the (S) product 2 with 99% and
the relative orientation of the hydride with respect to the 75% ee, respectively (Table 2, Entry 3; Table 3, Entry 5). Fi-
olefin is a little better for the intermediate (10b-si) that pro- nally, all these results demonstrate the strong influence of
vides the (R) product. All in all, these findings corroborate the substitution at the R1 position on the enantioselectivity,
the experimental results.
whereas the R3 substituents (at least in the case of an o-
When R1 was a tert-butyl group, however, no asymmetric anisyl substituent) play a secondary role. It should be noted
induction was observed (Table 2, Entry 14). In this last case, that the difference in the influence of R1 and R3 would be
it is plausible that the two dihydride rhodium intermediates more difficult to explain with the quadrant rule.
10b-re and 10b-si are both equally disfavored, owing to ste-
ric interactions of the tert-butyl group with either the ester o-anisyl groups, the front view dissymmetry of the rhodium
or the cinnamyl moiety in the east region (Scheme 6). complex is then lost, resulting in equal coordination of the
In the case of the AMPP 4Јq, in which R1 and R2 are two
In addition, when the ligand is substituted with phenoxy substrate to the re and si faces affording no asymmetric
groups in the R2 and R3 positions (ligand 4l), the east-west induction (Table 3, Entry 6). However, if the o-anisyl group
dissymmetry of the complex remains similar to that of the occupies only the R2 position, as in the ligand 4o, the dis-
AMPP ligands 4a–g, because the steric hindrance of the symmetry of the complex east-west region is then opposite.
diphenylphosphonite and diphenylphosphinite are closed. This then results in the coordination of the cinnamate on
Consequently, substrate coordination in the west region is the east side and its si-face, as shown by the intermediate
again favored, to afford the intermediate 10b-re and the (S) 10b-si (Scheme 5), the phenylalanine derivative 2 then being
Scheme 6.
2084
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2078–2090

Ligands for Rh-Catalyzed Hydrogenation
FULL PAPER
were prepared from the appropriate (+)- or (–)-ephedrine as de-
obtained with the (R) absolute configuration and with ee
values reaching 88% (Table 3, Entries 3, 4).
scribed previously.[56]
HPLC analyses were performed on a Gilson and Shimadzu appara-
tus fitted with a UV detector. The enantiomeric excesses of the
optically active derivatives were determined on Chiralcel OD or
OK columns (Daicel), with hexane/iPrOH mixtures as the mobile
phase, a flow rate of 1 mLmin–1 and UV detection at λ = 254 and
280 nm. Thin-layer chromatography was performed on silica chro-
magel (60 F254; SDS) and results were viewed by UV, potassium
permanganate, or iodine treatment. Flash chromatography was
performed on silica gel (60ACC, 6–35 microns and 35–70 microns;
SDS). All NMR spectra data were recorded on Bruker DPX 250
or Advance 300–500 spectrometers with TMS as internal reference
Conclusions
A new series of P-chirogenic aminophosphane-phosphin-
ite ligands (AMPP) 4 has been synthesized from (+)- or
(–)-ephedrine by a versatile three-step methodology. The
AMPP ligands 4a–4Јq were used in the forms of their rho-
dium complexes for catalyzed hydrogenations of methyl α-
acetamidocinnamate 1 as a test reaction. Notably, although
all the AMPP ligands 4a–4p derive from the (+)-ephedrine
backbone, variation of the substituent on a P-center could
result in the catalyzed hydrogenation product 2 with either
(R) or (S) absolute configurations, with ee values ranging
from 99% (S) to 88% (R). Indeed, it is particularly interest-
ing to optimize an asymmetric catalysis of a designed target
by use of a modular approach such as those presented for
the EPHOS AMPP ligand 4a. The asymmetric induction
was analyzed in the light of X-ray structures of the AMPP
4 complexes, and a new model for the enantioselectivity,
taking into consideration both the boat conformation and
the steric and the electronic dissymmetries of the complex,
has been proposed. This model offers an alternative to the
quadrant rule well adapted to the C2-symmetry ligands and
the twisted-chair conformations of their complexes. In this
work, the model – which schematizes the front side of the
complex as a sextant in the direction of the cardinal
points – is consistent with complexation of the substrate by
the acetamido and the cinnamyl groups, in the north and
east (or west) parts, respectively. The enantioselectivity is
due to the nature of the ligand substituents in the south-
east or south-west parts of the key substrate-dihydride in-
termediate. Further developments to extend the catalyzed
asymmetric hydrogenation for pharmaceutical targets and
DFT calculations on the stereochemical course of the dihy-
dride-Rh-substrate complex are in progress.
1
for H and 13C NMR and 85% phosphoric acid as external refer-
ence for 31P NMR spectroscopy. Melting points were measured on
a Büchi 530 melting point apparatus and are uncorrected. Optical
rotations were determined at 20 °C on Perkin–Elmer 241 and 341
polarimeters. Infrared spectra were recorded on Bruker Equinox 55
and Vector 22 instruments. Mass spectral analyses were performed
on NERMAG R10-10C and on JEOL MS 700 and KRATOS Con-
cept S instruments at the ENSCP (Paris), ENS (Paris), and Bur-
gundy University (Dijon), respectively. The major peak m/z was
mentioned with the intensity as a percentage of the base peak in
brackets. Elemental analyses were measured with a precision supe-
rior to 0.3% at the Microanalysis Laboratories of P. & M. Curie
(Paris) and Burgundy (Dijon) Universities.
Preparation of 1-Bromo-2-{[(2-methoxy)ethoxy]methoxy}benzene:
This compound was prepared by a modified procedure described
previously.[65] NaH (60 % in mineral oil, 1.40 g, 35 mmol) was
placed in a three-necked flask fitted with a magnetic stirrer and
washed twice with pentane. Dry THF (10 mL) was then added, the
mixture was cooled to 0 °C, and 2-bromophenol (3.0 mL, 26 mmol)
was slowly added with stirring. The mixture was stirred for 2 h at
0 °C, and chloro{[(2-methoxy)ethoxy]methane (3.0 mL, 26 mmol)
was slowly added. The mixture was stirred for 30 min at 0 °C and
allowed to warm to room temperature overnight. THF was evapo-
rated under reduced pressure and the residue was hydrolyzed at
room temperature and extracted with CH2Cl2. The organic extracts
were dried with anhydrous MgSO4 and the solvent was removed
under vacuum. The residue was purified by chromatography on
silica gel with diethyl ether/petroleum ether (1:3) as eluent, yielding
the product (5.83 g, 86%) as a colorless oil. Rf = 0.3 (diethyl ether/
1
petroleum ether, 1:3). H NMR (CDCl3): δ = 3.37 (s, 3 H, CH3O),
Experimental Section
3.57 (m, 2 H, CH2O), 3.88 (m, 2 H, CH2O), 5.34 (s, 2 H, OCH2O),
6.89 (dt, J = 1.6, J = 7.6 Hz, 1 H, H arom.), 7.19–7.27 (m, 2 H, H
arom.), 7.54 (dd, J = 1.6, J = 7.9 Hz, 1 H, H arom.) ppm. 13C
NMR (CDCl3): δ = 58.80 (CH3), 67.86 (CH2), 71.34 (CH2), 93.91
(CH2), 112.64 (CBr), 116.13 (C arom.), 122.96 (C arom.), 128.33
(C arom.), 133.17 (C arom.), 153.60 (CO arom.) ppm. IR (neat,
General: All reactions were carried out under argon in dried glass-
ware. Solvents were dried and freshly distilled under argon and over
sodium/benzophenone for THF, diethyl ether, toluene, and ben-
zene, over P2O5 for CH2Cl2, and over sodium ethoxide for EtOH.
Hexane and propan-2-ol for HPLC were of chromatography grade
and were used without further purification. Methyllithium, sec-bu-
tyllithium, tert-butyllithium, ferrocene, chlorodiphenylphosphane,
and chlorodicyclohexylphosphane were purchased from Aldrich,
Acros, and Avocado and used as received. Commercially available
2-bromoanisole, and bromobenzene were distilled before use,
whereas 1-bromonaphthalene, 2-bromonaphthalene, 2-bromobi-
phenyl, and bromocyclohexane were used without purification. The
HCl in toluene solution was obtained by bubbling HCl gas and
titration of the resulting solution by colorimetry. The (2S,4R,5S)-
(–)-3,4-dimethyl-2,5-diphenyl-1,3,2-oxazaphospholidine-2-borane
(5) and its enantiomer,[59] (S)-(–)- and (R)-(+)-o-anisyl(chloro)-
phenylphosphane–borane 8, the AMPP ligands 4a, 4b, 4d, 4f, 4h,
4i, 4k, 4m, and 4q, and their corresponding diborane complexes 7
ν): 3069–2883, 1585, 1474, 1275, 1230, 1159, 1101, 1037, 976, 848,
˜
750 cm–1. MS (EI): m/z (%) = 262 (40) [M]+, 260 (40) [M]+, 187
(35), 185 (35), 174 (45), 172 (45), 157 (35), 155 (35), 145 (20), 143
(20), 89 (85), 59 (100). HRMS (EI) calcd. for C10H13O381Br [M]+:
262.00276; found: 262.00552. C10H13O3Br (261.116): C 46.00, H
5.02; found: C 46.33, H 5.16.
Transmetalation of 1-Bromo-2-{[(2-methoxy)ethoxy]methoxy}-
benzene: The bromo derivative (1 equiv.) was placed in a two-
necked flask fitted with a magnetic stirrer. The mixture was cooled
to 0 °C and sec-butyllithium (1 equiv.) was slowly added by syringe
and with stirring. After the formation of a white precipitate, the
mixture was stirred for 1 h at 0 °C. The organolithium reagent was
dissolved with a minimum of THF before use.
Eur. J. Org. Chem. 2007, 2078–2090
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2085

C. Darcel, S. Jugé, et al.
FULL PAPER
Preparation of the Aminophosphane-Phosphinite Diboranes 7. Gene-
ral Procedure for AMPP Ligands 7a–k: A 100 mL round-bottomed
flask fitted with a magnetic stirrer and an argon inlet was charged
with oxazaphospholidine borane 5 (1 equiv., 3.5 mmol) and THF
(6 mL). An organolithium reagent[66] (2 equiv., 7 mmol) was added
slowly at –78 °C with stirring, the reaction temperature was allowed
to rise slowly to 0 °C, and stirring was maintained until complex 5
had disappeared completely. Chlorophosphane (2 equiv., 7 mmol)
was added and the mixture was stirred for 2 h and allowed to warm
(t, J = 9.1 Hz, 1 H, CHO), 6.80–6.88 (m, 2 H, H arom.), 6.95–7.10
(m, 7 H, H arom.), 7.13–7.30 (m, 6 H, H arom.), 7.32–7.50 (m, 3 H,
H arom.), 7.60–7.70 (m, 2 H, H arom.) ppm. 13C NMR (CDCl3): δ
2
2
= 15.8 (CH3), 29.0 (d, JPNC = 4.0 Hz, NCH3), 56.9 (dd, JPNC
9.0, JPOCC = 10.2 Hz, NCH), 70.2 (Cp), 70.6 (Cp), 70.9 (d, JPC
=
=
=
3
6.2 Hz, Cp), 71.1 (d, JPC = 8.3 Hz, Cp), 71.5 (Cp), 72.0 (d, JPC
2
3
8.8 Hz, Cp), 72.1 (Cp), 83.4 (dd, JPOC = 2.8, JPNCC = 9.3 Hz,
OCH), 127.8 (d, JPC = 5.5 Hz, C arom.), 127.9 (d, JPC = 5.4 Hz,
C arom.), 128.0 (C arom.), 128.3 (d, JPC = 13.0 Hz, C arom.), 128.5
to room temperature. Borane–dimethyl sulfide (BMS, 10 equiv.) (C arom.), 128.6 (C arom.), 130.1 (d, JPC = 2.1 Hz, C arom.),
was added and the mixture was stirred overnight. BMS and THF
were evaporated under reduced pressure and the residue was hy-
drolyzed at room temperature and then extracted with CH2Cl2. The
organic extracts were dried with anhydrous MgSO4 and the solvent
was removed. The residue was purified by chromatography on silica
gel with toluene/petroleum ether (1:1) as eluent, yielding the AMPP
diborane 7, which was then recrystallized from a CH2Cl2/hexane
mixture.
130.9–131.6 (C arom.), 131.8 (C arom.), 132.1 (d, JPC = 6.2 Hz, C
arom.), 132.8 (C arom.), 138.2 (C arom.) ppm. 31P NMR (CDCl3):
1
δ = 71.8 (br, P-N), 107.8 (d, JPB = 58.0 Hz, P-O) ppm. IR (neat):
ν = 3053–2930, 2383, 2348, 1483, 1457, 1435, 1224, 1205, 1167,
˜
1131, 1108, 1064, 1027, 1006, 988, 963, 898, 866, 825, 762, 734,
715, 697 cm–1. MS (EI): m/z (%) = 669 (2) [M]+, 452 (10), 387 (5),
293 (100), 226 (5), 186 (25), 107 (5). C38H43B2NOP2Fe (669.183):
C 68.21, H 6.48, N 2.09; found: C 67.85, H 6.60, N 2.88.
Compound 7g: 0.89 g (40%); colorless crystals (CH2Cl2/hexane);
m.p. 159–160 °C. [α]2D0 = –79.2 (c = 1, CHCl3); Rf = 0.3 (toluene/
petroleum ether, 1:1). 1H NMR (CDCl3): δ = 0.20–1.60 (m, 6 H,
AMPP 7l: This AMPP diborane complex was prepared by a modi-
fication of the procedure described above with use of tri-
phenylphosphite instead of the chlorophosphane. After workup
and chromatography, the ligand was used without further purifica-
tion.
3
3
BH3), 1.10 (d, JHH = 6.6 Hz, 3 H, CH3), 2.30 (d, JHP = 7.0 Hz,
3 H, NCH3), 4.32 (m, 1 H, CHN), 5.33 (t, J = 8.6 Hz, 1 H, CHO),
6.43–6.78 (m, 3 H, H arom.), 6.87–7.02 (m, 12 H, H arom.), 7.02–
7.16 (m, 6 H, H arom.), 7.28–7.39 (m, 6 H, H arom.), 7.54–7.60
AMPP Ligands 7m–q: The AMPP diboranes 7m–q were obtained
by a modification of the procedure described above with use of the
corresponding P-chirogenic chlorophosphane borane 8[56,60] instead
of the chlorodiphenylphosphane.
3
(m, 2 H, H arom.) ppm. 13C NMR (CDCl3): δ = 14.6 (d, JPNCC
2
2
= 2.3 Hz, CH3), 29.9 (d, JPNC = 3.5 Hz, NCH3), 57 (dd, JPNC
=
3
8.4, JPOCC = 10.3 Hz, NCH), 82.5 (m, OCH), 126.8–127.1 (C
arom.) 127.7 (d, JPC = 2.1 Hz, C arom.), 127.7 (d, JPC = 26.3 Hz,
C arom.) 128.5 (d, JPC = 10.9 Hz, C arom.), 129.2 (C arom.), 129.8
(C arom.), 130.0 (d, JPC = 2.1 Hz, C arom.), 130.3 (d, JPC = 1.9 Hz,
C arom.), 130.5 (C arom.), 131.0 (d, JPC = 11.2 Hz, C arom.), 131.2
(Rp)-(–)-{[(1R,2S)-2-(Diphenylphosphinito–borane)-1-methyl-2-phen-
ylethyl](methyl)amino}{o-[(2-methoxyethoxy)methoxy]phenyl}-
phosphane–Borane (7c): 1.62 g (66%); white crystals (CH2Cl2/hex-
ane); m.p. 95 °C. [α]2D0 = –51 (c = 1, CHCl3); Rf = 0.2 (toluene/
AcOEt, 98:2). 1H NMR (CDCl3): δ = 0.20–1.60 (m, 6 H, BH3),
(d, JPC = 2.1 Hz, C arom.), 131.3 (C arom.), 131.5 (d, JPC
=
3
3
1.32 (d, JHH = 6.5 Hz, 3 H, CH3), 2.34 (d, JPNCH = 7.7 Hz, 3 H,
NCH3), 3.23 (s, 3 H, OCH3), 3.34 (m, 4 H, OCH2), 4.35–4.70 (m,
1 H, CHN), 4.77 (d, J = 7.1 Hz, 1 H, OCH2O), 4.84 (d, J = 7.1 Hz,
1 H, OCH2O), 5.23 (t, J = 9.2 Hz, 1 H, CHO), 6.54–6.63 (m, 2 H,
H arom.), 6.90–7.30 (m, 15 H, H arom.), 7.31–7.47 (m, 5 H, H
arom.), 7.60–7.68 (m, 2 H, H arom.) ppm. 13C NMR (CDCl3): δ =
15.9 (CH3), 29.8 (d, 2JPNC = 4.7 Hz, NCH3), 56.9 (dd, 2JPNC = 8.8,
3JPOCC = 11.7 Hz, NCH), 58.9 (OCH3), 67.7 (OCH2), 71.3 (OCH2),
83.8 (dd, JPC = 2.7, JPC = 8.6 Hz, OCH), 93.2 (OCH2O), 114.5 (d,
JPC = 4.4 Hz, C arom.), 118.8 (d, JPC = 56.1 Hz, C arom.), 121.9
(d, JPC = 10.5 Hz, C arom.), 127.7 (d, JPC = 10.9 Hz, C arom.),
127.8 (d, JPC = 10.6 Hz, C arom.), 128.0 (C arom.), 128.4 (d, JPC
= 7.7 Hz, C arom.), 128.6 (C arom.), 128.7 (C arom.), 129.5 (d, JPC
= 2.3 Hz, C arom.), 130.4 (d, JPC = 10.8 Hz, C arom.), 131.0 (d,
JPC = 11.1 Hz, C arom.), 131.9–132.3 (C arom.), 132.8 (C arom.),
133.4 (d, JPC = 1.5 Hz, C arom.), 134.9 (d, JPC = 11.0 Hz, C arom.),
137.9 (C arom.), 159.0 (d, JPC = 2.2 Hz, C arom.) ppm. 31P NMR
11.6 Hz, C arom.), 131.5 (d, JPC = 2.4 Hz, C arom.), 131.8 (d, JPC
= 15.2 Hz, C arom.), 132.0 (d, JPC = 22.9 Hz, C arom.) 132.7 (C
arom.), 132.8 (d, JPC = 8.2 Hz, C arom.), 133.3 (d, JPC = 9.4 Hz,
C arom.), 137.6 (C arom.), 140.7 (d, JPC = 2.7 Hz, C arom.), 146.9
(d, JPC = 10.4 Hz, C arom.) ppm. 31P NMR (CDCl3): δ = 72.4 (br,
1
P-N), 107.2 (d, JPB = 69.9 Hz, P-O) ppm. IR (neat): ν = 3056–
˜
2910, 2425, 2400, 2376, 2336, 1587, 1481, 1457, 1448, 1437, 1223,
1161, 1108, 1058, 1011, 999, 987, 965, 894, 860, 758, 736, 712, 694,
689, 621, 605 cm–1. MS (LSIMS): m/z (%) = 636.1 [M – H]+ (7),
408.1 (15), 318.1 (50), 261.0 (95), 183 (100). C40H43B2NOP2
(637.358): C 75.38, H 6.80, N 2.20; found: C 75.59, H 6.88, N 2.81.
Compound 7j: 0.75 g (38%); colorless crystals (CH2Cl2/hexane);
m.p. 127–128 °C. [α]2D0 = –42.7 (c = 0.4, CHCl3); Rf = 0.56 (toluene/
1
petroleum ether, 1:1). H NMR (CDCl3): δ = 0.20–1.20 (m, 9 H),
3
1.35 (d, JHH = 6.6 Hz, 3 H, CH3), 1.30–1.86 (m, 7 H), 2.18 (m, 1
H, PCH), 2.39 (d, 3JHP = 6.7 Hz, 3 H, NCH3), 4.27 (m, 1 H, CHN),
5.40 (dd, J = 9.2 Hz, 1 H, CHO), 6.75–7.15 (m, 11 H, H arom.),
7.22–7.30 (m, 4 H, H arom.), 7.40–7.60 (m, 3 H, H arom.), 7.68–
7.72 (m, 2 H, H arom.) ppm. 13C NMR (CDCl3): δ = 15.8 (d, JPNCC
= 1.4 Hz, CH3), 25.9 (d, JPC = 1.4 Hz, CH2), 26.0 (CH2), 26.6 (d,
(CDCl ): δ = 69.9 (br, P-N), 108.0 (br, P-O) ppm. IR (neat): ν =
˜
3
3065–2817, 2394, 2374, 2342, 1586, 1573, 1473, 1457, 1435, 1438,
1434, 1308, 1269, 1242, 1226, 1201, 1163, 1154, 1132, 1100, 1087,
1076, 1055, 1031, 1009, 999, 984, 965, 924, 895, 873, 854, 846, 797,
765, 743, 733, 727, 709, 695; 653, 620, 607 cm–1.
C38H47B2NO4P2·2H2O (701.397): C 65.07, H 7.33, N 2.00; found:
C 65.41, H 7.34, N 2.47.
JPC = 7.6 Hz, CH2), 26.7 (d, JPC = 1.9 Hz, CH2), 26.9 (d, JPC
=
=
2
1
11.9 Hz, CH2), 28.0 (d, JPNC = 3.4 Hz, NCH3), 32.7 (d, JPC
2
3
43.1 Hz, CH), 57.7 (dd, JPNC = 8.6 Hz, NCH), 82.8 (dd, JPNCC
2
= 4.3, JPOC = 3.1 Hz, OCH), 125.2 (C arom.), 127.6–128.2 (C
Compound 7e: 1.45 g (62%); orange crystals (CH2Cl2/hexane); m.p.
170–171 °C. [α]2D0 = +45.6 (c = 1, CHCl3); Rf = 0.2 (toluene/petro-
leum ether, 1:1). H NMR (CDCl3): δ = 0.20–1.60 (m, 6 H, BH3),
arom.), 128.5 (d, JPC = 10.9 Hz, C arom.), 129.0 (C arom.), 129.7
(C arom.), 130.0 (d, JCP = 2.0 Hz, C arom.), 130.4 (C arom.), 130.8
(d, JPC = 9.3 Hz, C arom.), 131.0 (d, JPC = 11.2 Hz, C arom.),
131.2–131.6 (C arom.), 131.9 (d, JPC = 23.7 Hz, C arom.), 132.7 (C
1
3
3
1.25 (d, JHH = 6.5 Hz, 3 H, CH3), 1.99 (d, JHP = 8.0 Hz, 3 H,
NCH3), 3.96 (br, 1 H, Cp), 4.19 (s, 5 H, Cp), 4.34 (br, 1 H, Cp), arom.), 137.8 (C arom.) ppm. 31P NMR (CDCl3): δ = 74.6 (d, 1JPB
1
4.40 (br, 1 H, Cp), 4.45 (m, 1 H, CHN), 4.50 (br, 1 H, Cp), 5.23 = 60.5 Hz, P-N), 107.7 (d, JPB = 64.9 Hz, P-O) ppm. IR (neat): ν
˜
2086
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2078–2090

Ligands for Rh-Catalyzed Hydrogenation
FULL PAPER
= 3064, 2932, 2855, 2393, 2348, 1451, 1437, 1223, 1132, 1113, 1065,
H arom.), 7.28–7.40 (m, 4 H, H arom.), 7.45 (m, 2 H, H arom.),
1016, 967, 920, 895, 861, 768, 756, 734, 710, 687, 623 cm–1. 7.60 (m, 1 H, H arom.), 7.90 (m, 1 H, H arom.) ppm. 13C NMR
2
C34H45B2NOP2 (567.308): C 71.98, H 8.00, N 2.47; found: C 71.84, (CDCl3): δ = 15.7 (CH3), 29.9 (d, JPNC = 4.7 Hz, NCH3), 55.0
H 8.50, N 2.48.
3
2
(OCH3), 55.5 (OCH3), 57.8 (dd, JPOCC = 11.4, JPNC = 8.5 Hz,
NCH), 82.7 (dd, 3JPNCC = 9.0, JPOC = 3.0 Hz, OCH), 111.6 (d, JPC
= 4.3 Hz, C arom.), 111.9 (d, JPC = 5.3 Hz, C arom.), 118.7 (d, JPC
= 51.0 Hz, C arom.), 120.3 (d, JPC = 70.9 Hz, C arom.), 120.9 (t,
JPC = 10.0 Hz, C arom.) 127.4–128.9 (C arom.), 129.6 (C arom.),
130.1 (d, JPC = 10.6 Hz, C arom.), 130.7 (C arom.), 131.1 (d, JPC
= 12.0 Hz, C arom.), 131.9 (d, JPC = 70.2 Hz, C arom.), 132.8 (d,
Compound 7l: 0.65 g (30%); colorless, viscous oil; Rf = 0.45 (tolu-
1
3
ene). H NMR (CDCl3): δ = 0.20–1.20 (m, 6 H), 1.31 (d, JHH
=
3
6.6 Hz, 3 H, CH3), 2.52 (d, JPNCH = 8.0 Hz, 3 H, NCH3), 3.55 (s,
1 H, OCH3), 4.59 (m, 1 H, CHN), 5.72 (t, J = 8.0 Hz, 1 H, CHO),
6.75–7.65 (m, 24 H, H arom.) ppm. 31P NMR (CDCl3): δ = 71.6
(br, P-N), 113.3 (br, P-O) ppm. The presence of a minor impurity
was detected at δ = 69.5 ppm.
JPC = 57.5 Hz, C arom.), 133.2–134.0 (C arom.), 135.6 (d, JPC
=
12.2 Hz, C arom.), 138.5 (C arom.), 160.9 (d, JPC = 4.4 Hz, C
arom.) 161.0 (C arom.) ppm. 31P NMR (CDCl3): δ = 69.9 (br, P-
Compound 7n: 1.14 g (55%); white crystals (CH2Cl2/hexane); m.p.
165 °C. [α]2D0 = –71.1 (c = 1.3, CHCl3); Rf = 0.55 (toluene). 1H
N), 105.8 (br, P-O) ppm. IR (neat): ν = 3061–2896, 2383, 2349,
˜
3
NMR (CDCl3): δ = 0.30–1.80 (br, 6 H, BH3), 1.24 (d, JHH
=
1590, 1573, 1477, 1458, 1431, 1278, 1252, 1220, 1163, 1134, 1064,
1008, 991, 959, 888, 868, 802, 758, 744, 712, 698, 614 cm–1. MS
(EI): m/z (%) = 620 (10) [M+ -H], 620 (10), 606 (100) [M+ – H –
BH3], 589 (45), 573 (20), 561 (20), 545 (50), 500 (25), 485 (15), 471
(45), 453 (40), 438 (45), 390 (50), 374 (65), 363 (60), 317 (60), 286
(65), 272 (100), 256 (65), 245 (70), 228 (100), 215 (90), 199 (80),
183 (90), 170 (95), 151 (90), 137 (70), 123 (80), 107 (85), 91 (95),
77 (80). C36H43B2NO3P2 (621.308): C 69.59, H 6.98, N 2.25; found:
C 69.57, H 7.09, N 2.48.
3
6.8 Hz, 3 H,CH3), 2.21 (d, JPNCH = 7.6 Hz, 3 H, NCH3), 3.46 (s,
3 H, OCH3), 4.50 (m, 1 H, CHN), 5.33 (t, J = 9.4 Hz, 1 H, CHO),
6.50–6.58 (m, 2 H, H arom.), 6.76 (dd, JHH = 8.3, JPCCH
3
3
=
4.7 Hz, 1 H, H arom.), 6.90–7.50 (m, 20 H, H arom.), 7.75–7.85
(m, 1 H, H arom.) ppm. 13C NMR (CDCl3): δ = 16.1 (CH3), 29.5
2
3
(d, JPNC = 4.5 Hz, NCH3), 55.6 (OCH3), 57.7 (dd, JPOCC = 11.0,
2JPNC = 8.4 Hz, NCH), 82.5 (dd, JPNCC = 8.9, JPOC = 2.6 Hz,
3
2
OCH), 111.9 (d, JPC = 5.3 Hz, C arom.), 120.3 (d, JPC = 70.3 Hz,
C arom.), 120.9 (d, JPC = 10.5 Hz, C arom.), 127.5 (d, JPC
=
General Procedure for Decomplexation into Free AMPP 4: The
AMPP borane 7 (1 mmol) was placed in a three-necked flask fitted
with a reflux condenser, a magnetic stirrer, and an argon inlet. A
solution of DABCO (4 mmol) in toluene (6 mL) was added, and
the flask was purged with three cycles of argon. The mixture was
heated at 50 °C for 12 h and the crude product was rapidly filtered
off on a neutral alumina column (15 cm height, 2 cm diameter)
with use of toluene/AcOEt (9:1) as degassed eluent. After removal
of the solvent, the free AMPP 4 was obtained in excellent yields
and used without further purification.
10.9 Hz, C arom.), 128.0–129.4 (C arom.), 130.3–131.3 (C arom.),
131.8 (d, JPC = 10.3 Hz, C arom.), 132.6 (d, JPC = 10.3 Hz, C
arom.), 132.7 (d, JPC = 62.3 Hz, C arom.), 133.4 (d, JPC = 9.4 Hz,
C arom.), 134.0 (d, JPC = 2.5 Hz, C arom.), 138.5 (C arom.), 161.0
(d, JPC = 4.3 Hz, C arom.) ppm. 31P NMR (CDCl3): δ = 72.4 (br,
P-N), 106.6 (br, P-O) ppm. IR (neat): ν = 3093–2898, 2385, 2345,
˜
1478, 1455, 1436, 1253, 1159, 1134, 1111, 1060, 1009, 979, 893,
864, 755, 737, 724, 694 cm–1. MS (EI): m/z (%) = 590 (20)
[M – H]+, 586 (40), 576 (100), 563 (45), 529 (20), 506 (50), 487 (35),
471 (20), 458 (20), 441 (15), 431 (20), 417 (20), 400 (25), 390 (20),
360 (20), 348 (25), 332 (40), 276 (30), 242 (85), 181 (85), 141 (100),
91 (45). C35H41B2NO2P2 (591.238): C 71.10, H 6.99, N 2.37; found:
C 71.10, H 7.26, N 2.48.
Compound 7o: 1.14 g (55%); white crystals (CH2Cl2/hexane); m.p.
170 °C. [α]2D0 = –98.5 (c = 1.4, CHCl3); Rf = 0.55 (toluene). 1H
3
NMR (CDCl3): δ = 0.30–1.80 (br, 6 H, BH3), 1.41 (d, JHH
=
(Rp)-{[(1R,2S)-2-(Diphenylphosphinito)-1-methyl-2-phenylethyl]-
(methyl)amino}{o-[(2-methoxyethoxy)methoxy]phenyl}phosphane
(4c): 560 mg (88%); white solid. 1H NMR (CDCl3): δ = 1.43 (d,
3
6.5 Hz, 3 H, CH3), 2.31 (d, JPNCH = 7.6 Hz, 3 H, NCH3), 3.37 (s,
3 H, OCH3), 4.63 (m, 1 H, CHN), 5.37 (t, J = 9.2 Hz, 1 H, CHO),
6.40–6.85 (m, 4 H, H arom.), 7.00–7.15 (m, 5 H, H arom.), 7.20–
7.80 (m, 15 H, H arom.) ppm. 13C NMR (CDCl3): δ = 16.3 (CH3),
29.5 (NCH3), 54.8 (OCH3), 57.6 (m, NCH), 82.3 (m, OCH), 110.5
3
3JHH = 6.6 Hz, 3 H, CH3), 2.17 (d, JPNCH = 2.7 Hz, 3 H, NCH3),
3.24 (s, 3 H, OCH3), 3.27–3.39 (m, 4 H, CH2), 3.90 (m, 1 H, CHN),
4.76 (t, J = 8.9 Hz, 1 H, CHO), 4.97 (dd, J = 7.0, J = 26.7 Hz, 2
H, CH2O), 6.56 (t, J = 7.1 Hz, 2 H, H arom.), 6.84 (t, J = 7.3 Hz,
1 H, H arom.), 6.92–7.48 (m, 22 H, H arom.) ppm. 31P NMR
(CDCl3): δ = 55.1 (P-N), 111.7 (P-O) ppm.
(C arom.), 118.9 (d, JPC = 55.1 Hz, C arom.), 120.0 (d, JPC
=
11.3 Hz, C arom.), 128.1–133.3 (C arom.), 134.3 (C arom.), 135.4
(d, JPC = 16.6 Hz, C arom.), 138.0 (C arom.), 160.5 (C arom.) ppm.
1
31P NMR (CDCl3): δ = 72.2 (br, P-N), 105.1 (br, JPB = 52.4 Hz,
1
P-O) ppm. IR (neat): ν = 3090–2890, 2381, 2346, 1592, 1479, 1455,
˜
Compound 4e: 634 mg (99%); orange crystals. H NMR (CDCl3):
3
3
1436, 1281, 1135, 1106, 1068, 1011, 977, 893, 868, 764, 741, 717,
693, 621, 612 cm–1. MS (EI): m/z (%) = 590 (10) [M – H]+, 586
(30), 576 (100), 458 (10), 416 (30), 390 (20), 344 (30), 332 (40), 282
(45), 256 (50), 242 (95), 228 (50), 199 (50), 181 (100), 141 (80), 91
(55), 77 (45). C35H41B2NO2P2 (591.238): C 71.10, H 6.99, N 2.37;
found: C 71.04, H 7.14, N 2.51.
δ = 1.48 (d, JHH = 6.6 Hz, 3 H, CH3), 2.14 (d, JPNCH = 3.0 Hz,
3 H, NCH3), 3.91–4.20 (m, 2 H), 4.27 (s, 5 H, Cp), 4.32–4.35 (m,
2 H, Cp), 4.38–4.41 (m, 1 H, Cp), 4.79 (t, J = 8.7 Hz, 1 H, CHO),
6.85–6.90 (m, 2 H, H arom.), 7.08–7.43 (m, 16 H, H arom.), 7.57–
7.64 (m, 2 H, H arom.) ppm. 31P NMR (CDCl3): δ = 62.3 (P-N),
111.6 (P–O) ppm.
1
Compound 7p: 1.41 g (65%); white crystals (CH2Cl2/hexane); m.p.
Compound 4g: 590 mg (97%); white powder. H NMR (CDCl3): δ
3
3
210 °C. [α]2D0 = –52.4 (c = 1.5, CHCl3); Rf = 0.25 (toluene). 1H
= 1.31 (d, JHH = 6.6 Hz, 3 H CH3), 2.29 (d, JPNCH = 2.9 Hz, 3
H, NCH3), 3.74 (m, 1 H, CHN), 4.86 (t, J = 8.5 Hz, 1 H, CHO),
6.68–6.71 (m, 2 H, H arom.), 7.04–7.06 (m, 2 H, H arom.), 7.10–
7.18 (m, 1 H, H arom.), 7.20–7.35 (m, 18 H, H arom.), 7.40–7.50
3
NMR (CDCl3): δ = 0.30–1.80 (br, 6 H, BH3), 1.36 (d, JHH
=
3
6.5 Hz, 3 H, CH3), 2.39 (d, JPNCH = 7.9 Hz, 3 H, NCH3), 3.49 (s,
3 H, OCH3), 3.54 (s, 3 H, OCH3), 4.63 (m, 1 H, CHN), 5.43 (t, J
= 9.3 Hz, 1 H, CHO), 6.62 (m, 2 H, H arom.), 6.84 (m, 2 H, H (m, 4 H, H arom.), 7.57–7.62 (m, 2 H, H arom.) ppm. 31P NMR
arom.), 6.95–7.15 (m, 9 H, H arom.), 7.20 (d, JHP = 5.7 Hz, 2 H,
(CDCl3): δ = 57.3 (s, P-N), 111.95 (s, P-O) ppm.
Eur. J. Org. Chem. 2007, 2078–2090
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2087

C. Darcel, S. Jugé, et al.
FULL PAPER
Compound 4j: 523 mg (97%); white powder. H NMR (CDCl3): δ
H arom.) ppm. 31P NMR (CDCl3): δ = 66.44 (s, P-N), 103.25 (s,
P-O) ppm.
1
3
= 1.10–1.25 (m, 3 H), 1.35 (d, JHH = 6.7 Hz, 3 H, CH3), 1.60–
3
2.15 (m, 7 H), 2.35 (d, JPNCH = 3.2 Hz, 3 H, NCH3), 3.84–3.89
1
Compound 4p: 587 mg (99%); white powder. H NMR (CDCl3): δ
(m, 1 H, CHN), 4.75 (t, J = 8.4 Hz, 1 H, CHO), 6.96–7.00 (m, 2
H, H arom.), 7.12–7.32 (m, 16 H, H arom.), 7.35–7.37 (m, 3 H, H
arom.), 7.50–7.53 (m, 2 H, H arom.) ppm. 31P NMR (CDCl3): δ =
70.2 (P-N), 111.3 (P-O) ppm.
3
3
= 1.55 (d, JHH = 6.6 Hz, 3 H, CH3), 2.24 (d, JPNCH = 2.8 Hz, 3
H, NCH3), 3.68 (s, 3 H, OCH3), 3.71 (s, 3 H, OCH3), 4.04 (m, 1
H, CHN), 4.87 (t, J = 8.6 Hz, 1 H, CHO), 6.55–7.80 (m, 23 H, H
arom.) ppm. 31P NMR (CDCl3): δ = 56.77 (s, P-N), 102.37 (s, P-
Compound 4l: 535 mg (90%). 1H NMR (CDCl3): δ = 1.46 (d, JHH O) ppm.
3
= 6.7 Hz, 3 H, CH3), 2.24 (d, 3JPNCH = 2.29 Hz, 3 H, NCH3), 3.71
Preparation of the Precatalyst AMPP-Rh Complexes. General Pro-
(s, 3 H, OCH3), 3.97 (m, 1 H, CHN), 5.51 (t, J = 9.0 Hz, 1 H,
CHO), 6.65–7.45 (m, 24 H, H arom.) ppm. 31P NMR (CDCl3): δ
= 56.5 (P-N), 134.5 (P-O). The presence of a minor impurity was
detected at 52.8 ppm.
cedure: In a Schlenk tube, a solution of AMPP (0.020 mmol) in
dry and degassed CH2Cl2 (1 mL) was added under argon at room
temperature to Rh(COD)2BF4 (0.197 mmol) in dry and degassed
CH2Cl2 (1 mL). The mixture was stirred for 1 h at room tempera-
1
Compound 4n: 529 mg (99%); white powder. H NMR (CDCl3): δ ture. The CH2Cl2 was evaporated under vacuum to a volume of
= 1.45 (d, JHH = 6.6 Hz, 3 H, CH3), 2.24 (d, JPNC = 3.1 Hz, 3 0.5 mL and dry ether (5 mL) was added to precipitate the cationic
H, NCH3), 3.71 (s, 3 H, OCH3), 4.00 (m, 1 H, CHN), 4.90 (t, J = complex. The precipitate was used for the asymmetric hydrogena-
3
3
9.0 Hz, 1 H, CHO), 6.60–7.75 (m, 24 H, H arom.) ppm. 31P NMR
(CDCl3): δ = 66.14 (s, P-N), 102.63 (s, P-O) ppm.
tion without further purification. In the case of the complex 9, it
was obtained as brown crystals by recrystallization from a CH2Cl2/
hexane mixture.
1
Compound 4o: 557 mg (99%); white powder. H NMR (CDCl3): δ
2
3
= 1.30 (d, JHP = 6.6 Hz, 3 H, CH3), 2.20 (d, JHP = 3.2 Hz, 3 H, Typical Procedure for the Hydrogenation of Methyl α-Acetamidocin-
NCH3), 3.56 (s, 3 H, OCH3), 4.00 (m, 1 H, CHN), 4.81 (t, J =
9.1 Hz, 1 H, CHO), 6.68–6.70 (m, 3 H, H arom.), 6.89 (t, 1 H, JHP
= 7.4, H arom.), 7.05–7.45 (m, 18 H, H arom.), 7.55–7.65 (m, 2 H,
namate (1): Methyl α-Acetamidocinnamate (130 mg, 0.593 mmol),
catalyst (prepared by the procedure described above, 3 mol%,
0.0197 mmol) and degassed dry solvent (8 mL) were introduced un-
Table 4. Crystallographic data for 7j and 9.
Compound
7j
9
Formula
M
Crystallization
C34H45B2NOP2
567.27
colorless crystal,
from CH2Cl2/hexane
triclinic
0.28ϫ0.20ϫ0.10
P1
9.9240(1)
12.6034(2)
15.0250(2)
68.793(1)
81.887(1)
71.235(1)
1658.23(4)
2
C43H47NO2P2Rh, BF4
861.48
brown crystal,
from CH2Cl2/hexane
triclinic
0.22ϫ0.2ϫ0.15
P1
10.7679(1)
11.2809(2)
19.0990(3)
96.842(1)
95.705(1)
118.467(1)
1992.27(5)
2
Crystal system
Crystal size [mm]
Space group
a [Å]
b [Å]
c [Å]
α [°]
β [°]
γ [°]
V [Å3]
Z
F (000)
608
1.136
888
1.436
Dcalcd. [gcm–3]
Scan type
mixture of Φ rotations and ω scans
0.157
µ [mm–1]
0.565
sin(θ)/λmax [Å–1]
Index ranges h/k/l
Absorption correction
Reflections collected (RC)
Independent RC (IRC)
IRCGT = IRC and [I Ͼ 2σ(I)]
Refinement method
Data/restraints/parameters
R for IRCGT
0.65
0.65
–12;12/–16;16/–19;19
Scalepack
14321
14321
–13;13/–14;14/–24;24
16811
16811
15239
12348
full-matrix, least-squares on F2
14321/3/738
16811/3/978
[a]
[a]
R1 = 0.0408,
R1 = 0.0399,
[b]
[c]
wR2 = 0.0864
wR2 = 0.0798
[a]
[a]
R for IRC
R1 = 0.053,
R1 = 0.0479,
[b]
[c]
wR2 = 0.0919
wR2 = 0.0834
Goodness-of-fit[d]
1.025
1.032
Absol. structure parameter
0.03(4)
0.279 and –0.271
0.000(15)
1.045 and –0.555
Largest diff. peak and hole; eÅ–3
[a] R1 = Σ(||Fo| – |Fc||)/Σ|Fo|. [b] wR2 = [Σw(Fo – Fc ) /Σ[w(Fo2)2]1/2 where w = 1/[σ2(Fo2) + 0.22P + (0.04P)2] where P = (max(Fo2, 0) +
2
2 2
2Fc2)/3. [c] wR2 = [Σw(Fo – Fc ) /Σ[w(Fo2)2]1/2 where w = 1/[σ2(Fo2) + 2.38P + (0.0171P)2] where P = (max(Fo2, 0) + 2Fc )/3. [d] [Σw
2
2
2 2
2
(Fo – Fc ) /(No – Nv)]1/2
.
2 2
2088
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2078–2090

Ligands for Rh-Catalyzed Hydrogenation
FULL PAPER
[18]
T. P. Clark, C. R. Landis, Tetrahedron: Asymmetry 2004, 15,
2123–2137.
H. A. McManus, P. J. Guiry, Chem. Rev. 2004, 104, 4151–4202.
M. van den Berg, A. J. Minnaard, E. P. Schudde, J. van Esch,
A. H. M. de Vries, J. G. de Vries, B. L. Feringa, J. Am. Chem.
Soc. 2000, 122, 11539–11540.
M. Dieguez, O. Pamies, C. Claver, Chem. Rev. 2004, 104, 3189–
3215.
M. Tanaka, I. Ogata, J. Chem. Soc., Chem. Commun. 1975,
735.
der argon into a 100 mL autoclave. Subsequently, the reactor was
connected to a hydrogen cylinder. The argon was replaced with
hydrogen by six pressurizing cycles and the hydrogenation was run
under the reaction conditions indicated in the text. The solvent was
removed under reduced pressure. The phenylalanine methyl ester 2
was purified by flash chromatography on silica gel with a toluene/
AcOEt (3:1) mixture as eluent.
[19]
[20]
[21]
[22]
[23]
1
Phenylalanine Methyl Ester (2): H NMR (CDCl3): δ = 1.90 (s, 3
H, COCH3), 3.04 (m, 2 H, CH2Ph), 3.64 (s, 3 H, COOCH3), 4.81
(m, 1 H, CHCO2CH3), 5.90 (br, 1 H, NHCOCH3), 6.90–7.3 (m, 5
H, H arom.) ppm. 13C NMR (CDCl3): δ = 22.9 (s, COCH3), 37.7
(s, CO2CH3), 52.1 (s, CH2Ph), 53.1 (s, CHCOOCH3), 127 (s, C
arom.), 128.4 (s, C arom.), 129.1 (s, C arom.), 135.8 (s, C arom.),
169.6 (s, CO), 172.0 (s, CO) ppm.
R. H. Grubbs, R. A. deVries, Tetrahedron Lett. 1977, 22, 1879–
1880.
[24]
[25]
M. T. Reetz, T. Sell, Tetrahedron Lett. 2000, 41, 6333–6336.
M. van den Berg, A. J. Minnaard, R. M. Haak, M. Leeman,
E. P. Schudde, A. Meetsma, B. L. Feringa, A. H. M. de Vries,
C. E. P. Maljaars, C. E. Willans, D. Hyett, J. A. F. Boogers,
H. J. W. Henderickx, J. G. de Vries, Adv. Synth. Catal. 2003,
345, 308–323.
X-ray Crystallographic Study of Complexes 7j and 9: The X-ray
crystallographic analysis were collected with a Nonius Kappa CCD
diffractometer at 110 K, with use of graphite-monochromated Mo-
Kα radiation. The structure was solved by direct method
(SHELXS-97)[67] and refined by full-matrix, least-squares methods
(SHELXL-97)[67] with the aid of the WINGX program.[68] Crystal-
lographic data are reported in Table 4.
[26]
[27]
[28]
[29]
[30]
H. J. Kreuzfeld, U. Schmidt, C. Döbler, H. W. Krause, Tetrahe-
dron: Asymmetry 1996, 7, 1011–1018.
F. Agbossou, J. F. Carpentier, F. Hapiot, I. Suisse, A. Mor-
treux, Coord. Chem. Rev. 1998, 178–180, 1615–1645.
E. A. Broger, W. Burkart, M. Hennig, M. Scalone, R. Schmidt,
Tetrahedron: Asymmetry 1998, 9, 4043–4054.
O. Lot, I. Suisse, A. Mortreux, F. Agbossou, J. Mol. Catal. A
2000, 164, 125–130.
G. Chen, X. Li, H. Zhang, L. Gong, A. Mi, X. Cui, Y. Jiang,
M. C. K. Choi, A. S. C. Chan, Tetrahedron: Asymmetry 2002,
13, 809–813.
CCDC-619377 and -619378 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[31]
[32]
N. V. Dubrovina, V. I. Tararov, Z. Kadyrova, A. Monsees, A.
Börner, Synthesis 2004, 12, 2047–2051.
B. D. Vineyard, W. S. Knowles, M. J. Sabacky, G. L. Bachman,
D. J. Weinkauff, J. Am. Chem. Soc. 1977, 99, 5946–5952.
W. S. Knowles, Acc. Chem. Res. 1983, 16, 106–112.
A. Marinetti, S. Jus, J. P. Genêt, Tetrahedron Lett. 1999, 40,
8365–8368.
I. D. Gridnev, T. Imamoto, Acc. Chem. Res. 2004, 37, 633–644.
J. M. Brown, P. A. Chaloner, J. Am. Chem. Soc. 1980, 102,
3040–3048.
Acknowledgments
This work was supported by the Ministry of research, France, the
CNRS, and Burgundy regional council. We would also like to
thank Messrs. N. Morin (ENS Paris), N. Sellier (ENSCP Paris), B.
Rebière, and E. Pousson (LSEO, Dijon) for the analysis and J. Rey
for his help in synthesis.
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[1] Comprehensive Asymmetric Catalysis (Eds: E. N. Jacobsen, A.
Pfaltz, H. Yamamoto), Springer, New York, 2004.
[2] R. Noyori, T. Ohkuma, Angew. Chem. Int. Ed. 2001, 40, 40–
73.
[3] H. U. Blaser, C. Malan, B. Pugin, F. Spindler, H. Steiner, M.
Studer, Adv. Synth. Catal. 2003, 345, 103–151.
[4] O. Riant, N. Mostefaï, J. Courmarcel, Synthesis 2004, 18,
2943–2958.
[5] X. Cui, K. Burgess, Chem. Rev. 2005, 105, 3272–3296.
[6] C. Jäkel, R. Paciello, Chem. Rev. 2006, 106, 2912–2942.
[7] L. F. Tietze, I. Hiriyakkanavar, H. P. Bell, Chem. Rev. 2004,
104, 3453–3516.
A. S. C. Chan, J. J. Pluth, J. Halpern, J. Am. Chem. Soc. 1980,
102, 5952–5954.
A. S. C. Chan, J. Halpern, J. Am. Chem. Soc. 1980, 102, 838–
840.
P. S. Chua, N. K. Roberts, B. Bosnich, S. J. Okrasinski, J. Halp-
ern, J. Chem. Soc., Chem. Commun. 1981, 102, 1278–1280.
C. R. Landis, J. Halpern, J. Am. Chem. Soc. 1987, 109, 1746–
1754.
U. Nagel, B. Rieger, Organometallics 1989, 8, 1534–1538.
J. S. Giovannetti, C. M. Kelly, C. R. Landis, J. Am. Chem. Soc.
1993, 115, 4040–4057.
[41]
[42]
[43]
[44]
[45]
[46]
[47]
C. R. Landis, P. Hilfenhaus, S. Feldgus, J. Am. Chem. Soc.
1999, 121, 8741–8754.
[8] H. M. L. Davies, R. E. J. Beckwith, Chem. Rev. 2003, 103,
2861–2903.
S. Feldgus, C. R. Landis, J. Am. Chem. Soc. 2000, 122, 12714–
12727.
R. Giernoth, H. Heinrich, N. J. Adams, R. J. Deeth, J. Bargon,
J. M. Brown, J. Am. Chem. Soc. 2000, 122, 12381–12382.
C. R. Landis, S. Feldgus, Angew. Chem. Int. Ed. 2000, 39,
2863–2866.
I. D. Gridnev, N. Higashi, K. Asakura, T. Imamoto, J. Am.
Chem. Soc. 2000, 123, 7183–7194.
I. D. Gridnev, T. Imamoto, Organometallics 2001, 20, 545–549.
H. Heinrich, R. Giernoth, J. Bargon, J. M. Brown, Chem. Com-
mun. 2001, 1296–1297.
I. D. Gridnev, N. Higashi, T. Imamoto, Organometallics 2001,
20, 4542–4553.
I. D. Gridnev, M. Yasutake, N. Higashi, T. Imamoto, J. Am.
Chem. Soc. 2001, 123, 5268–5376.
K. V. L. Crépy, T. Imamoto, Adv. Synth. Catal. 2003, 345, 79–
101.
[9] B. M. Trost, M. L. Crawley, Chem. Rev. 2003, 103, 2921–2923.
[10] M. Shibasaki, E. M. Vogl, T. Ohshima, Adv. Synth. Catal.
2004, 346, 1533–1552.
[11] A. M. Carroll, T. P. O’Sullivan, P. J. Guiry, Adv. Synth. Catal.
2005, 347, 609–631.
[12] H. Ibrahim, A. Togni, Chem. Commun. 2004, 1147–1155.
[13] K. B. Sharpless, Angew. Chem. Int. Ed. 2002, 41, 2024–2032.
[14] E. N. Jacobsen, M. H. Wu in Comprehensive Asymmetric Ca-
talysis, vol. II (Eds: E. N. Jacobsen, A. Pfaltz, H. Yamamoto),
Springer, Heidelberg, 1999, chapter 21, pp. 649–677.
[15] H. C. Kolb, K. B. Sharpless, Asymmetric dihydroxylation in
Transition Metals for Organic Synthesis, 2nd ed., vol. 2 (Eds.:
M. Beller, C. Bolm), Wiley-VCH, Weinheim, Germany, 2004,
pp. 275–298.
[48]
[49]
[50]
[51]
[52]
[16] W. Tang, X. Zang, Chem. Rev. 2003, 103, 3029–3069.
[17] M. Berthod, G. Migniani, G. Woodward, M. Lemaire, Chem.
Rev. 2005, 105, 1801–1836.
Eur. J. Org. Chem. 2007, 2078–2090
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2089

C. Darcel, S. Jugé, et al.
FULL PAPER
[53] Y. Wada, T. Imamoto, H. Tsuruta, K. Yamaguchi, I. D. Grid-
nev, Adv. Synth. Catal. 2004, 346, 777–788.
[54] T. Schmidt, W. Baumann, H.-J. Drexler, A. Arrieta, D. Heller,
H. Buschmann, Organometallics 2005, 24, 3842–3848.
[55] G. Hoge, H.-P. Wu, W. S. Kissel, D. A. Pflum, D. J. Greene, J.
Bao, J. Am. Chem. Soc. 2004, 126, 5966–5967.
[63] The computations were performed with PC-Model Version 7.0
from Serena Software (MMX and MM3 force field). For ap-
propriate modeling of the hydrides, these were replaced by F
atoms, which exhibit the same atomic radii and M–H bond
lengths.
[64] H.-J. Drexler, S. Zhang, A. Sun, A. Spannenberg, A. Arrieta,
A. Preetz, D. Heller, Tetrahedron: Asymmetry 2004, 15, 2139–
2150.
[56] D. Moulin, C. Darcel, S. Jugé, Tetrahedron: Asymmetry 1999,
10, 4729–4743.
[57] For modified P-chirogenic EPHOS ligands 4Ј derived from (_)-
ephedrine with R1 = Me, n-Bu, 1-Np and R2 = R3 = Ph, p-
Tol, 3,5-(CF3)2Ph, see: R. Ewalds, E. B. Eggeling, A. C. Hewat,
P. C. J. Kamer, P. W. N. M. van Leeuwen, D. Vogt, Chem. Eur.
[65] A. Kless, R. Kadyrov, A. Börner, J. Holz, H. B. Kagan, Tetra-
hedron Lett. 1995, 36, 4601–4602.
[66] The organolithium reagents are prepared by bromide metal ex-
change with sec-butyllithium as described previously.[56] In the
case of the ferrocenyllithium, it was prepared by deprotonation
of ferrocene with 1 equiv. of butyllithium in THF at –78 °C.
[67] G. M. Sheldrick, SHELX97 (includes SHELXS97 and
SHELXL97), Release 97-2, Programs for Crystals Structure
Analysis, University of Göttingen, Göttingen, Germany, 1998.
[68] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837–838.
Received: November 5, 2006
J. 2000, 6, 1496–1504 and Ref.[58]
.
[58] J. J. Carbo, A. Lledos, D. Vogt, C. Bo, Chem. Eur. J. 2006, 12,
1457–1467.
[59] E. B. Kaloun, R. Merdès, J. P. Genêt, J. Uziel, S. Jugé, J. Or-
ganomet. Chem. 1997, 529, 455–463.
[60] C. Bauduin, D. Moulin, E. B. Kaloun, C. Darcel, S. Jugé, J.
Org. Chem. 2003, 68, 4293–4301.
[61] P. Haquette, F. Lebideau, S. Dagorne, J. Marrot, G. Jaouen,
Acta Crystallogr., Sect. C 2002, 58, m551–m552.
[62] P-chirogenic AMPP ligands 4e and 4g were also investigated
in asymmetric allylation catalyzed by palladium complexes: C.
Darcel et al., manuscript in preparation.
Published Online: March 14, 2007
2090
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2078–2090