Modular chiral ligands: The profiling of the Mandyphos and Taniaphos ligand families

Article abstract of DOI:10.1016/j.tetasy.2004.06.033

A set of 11 ferrocenyl based diphosphine ligands (eight Mandyphos and three Taniaphos) was tested in more than 150 experiments using 20 test reactions. For the assessment of new ligands, a two-pronged strategy was developed consisting of a basic and an extended profiling. The basic profiling showed that the choice of the substituents at the P atoms has a significant effect on the catalyst performance. In the extended profiling it was confirmed that the Mandyphos ligands, in particular M4 with two bis(3,5-dimethyl-4-methoxyphenyl)phosphino groups, and the Taniaphos ligands, especially the all-phenyl derivative T1, showed good to outstanding performances in the hydrogenation of selected α- and β-enamides, acrylic acid derivatives, itaconates, β-ketoesters and 1,3-diketones yielding the corresponding products with up to 99% ee and at substrate/catalyst ratios up to 25,000.

Full text of DOI:10.1016/j.tetasy.2004.06.033

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 2299–2306
Modular chiral ligands: the profiling of the Mandyphos and
Taniaphos ligand families
a
a
a
Felix Spindler,a, Christophe Malan, Matthias Lotz, Martin Kesselgruber,
*
Ulrich Pittelkow,a Andreas Rivas-Nass,b Oliver Brielb and Hans-Ulrich Blasera
aSolvias AG, PO Box I, CH-4002 Basel, Switzerland
bUmicore AG &Co KG, Rodenbacherchaussee 4, 63403 Hanau, Germany
Received 5 May 2004; accepted 9 June 2004
Abstract—A set of 11 ferrocenyl based diphosphine ligands (eight Mandyphos and three Taniaphos) was tested in more than 150
experiments using 20 test reactions. For the assessment of new ligands, a two-pronged strategy was developed consisting of a basic
and an extended profiling. The basic profiling showed that the choice of the substituents at the P atoms has a significant effect on the
catalyst performance. In the extended profiling it was confirmed that the Mandyphos ligands, in particular M4 with two bis(3,5-di-
methyl-4-methoxyphenyl)phosphino groups, and the Taniaphos ligands, especially the all-phenyl derivative T1, showed good to
outstanding performances in the hydrogenation of selected a- and b-enamides, acrylic acid derivatives, itaconates, b-ketoesters
and 1,3-diketones yielding the corresponding products with up to 99% ee and at substrate/catalyst ratios up to 25,000.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
If we want to make catalytic enantioselective hydrogen-
ation a valuable tool for the synthetic chemist, these
observations have two consequences: (i) a large variety
of ligands has to be designed and prepared, and (ii) their
catalytic profiles, that is, the suitability to catalyze spe-
cific transformations, should be well known. Even if
one applies certain restricting design principles there is
still an almost unlimited number of potential ligands,
which should be tested for interesting reactions. Obvi-
ously, a compromise has to be reached between the ben-
efits of structural diversity and the costs and time
involved preparing and testing a large number of lig-
ands. Solvias decided on the following concept: Concen-
trate on a relatively small number of privileged ligands
families, which are highly modular and easily tuned both
sterically and electronically with a reasonable amount of
synthetic effort. For each ligand family we synthesized
various members with a sufficient electronic and steric
diversity and tested each in a number of test reactions
to obtain a good performance profile for each ligand.
This concept has already been implemented successfully
for the Josiphos ligand family, albeit with a limited num-
ber of test reactions.4
The most effective enantioselective catalysts are com-
plexes consisting of a central metal ion and one or more
chiral ligands (see Fig. 1).1 There is no doubt that the
choice of a specific metal–ligand combination essentially
determines what transformation a complex is able to
catalyze selectively. For hydrogenation reactions, the
most active and versatile catalysts are low valent Rh,
Ru and Ir complexes with tertiary phosphorus ligands
in which the chiral element is either the backbone or
sometimes the coordinating phosphine.2,3 An analysis
of the results reported so far shows two somewhat con-
tradictory tendencies. On the one hand, relatively high
catalyst specificity is observed, that is, a small change
in either ligand or substrate structure often has a strong
effect on the catalytic performance (ee, ton, tof). On the
other hand, there are a number of ligand families, which
are more likely to give better performances than others.
For these, Jacobsen has coined the term Ôprivileged lig-
andsÕ (for selected examples see Fig. 1).
Herein we report on a systematic investigation of the
catalytic properties of the Mandyphos and Taniaphos
ligand families (see Fig. 2) first prepared by Knochel
*
Corresponding author. Tel.: +41-61-6866161; fax: +41-61-
6866311; e-mail: felix.spindler@solvias.com
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.06.033

2300
F. Spindler et al. / Tetrahedron: Asymmetry 15 (2004) 2299–2306
PR2
R1
R2
R1
O
O
R'
R'
R2
(chiral)
backbone
PR2
P
P
free
coordination
sites
R1
PR'2
R2
PR'2
Z
M
Z
R4
R3
X
M
X
Z
type, oxidation state
anion
coordinating atoms
CH3
R
R
H
Y
X = PPh2, H
Y = NHR, PR2
P
P
PR'2
X
Fe
R
R
Figure 1. Design elements for chiral metal complexes and selected privileged ligand backbones for enantioselective hydrogenation.
For all investigated derivatives, R = NMe2
OMe
CF3
CF3
R''
R
H
R''2P
R'
Fe
M2
T2
H
R
M6
M5
PR'2
OMe
Mandyphos
R'2P
M4
T3
M1
T1
M7
M8
R''2P
R
Fe
CF3
CF3
H
M3
Taniaphos
Figure 2. Generic structures of the MandyphosÒ M and TaniaphosÒ T ligand families and abbreviations used for tested ligands. As a general rule,
ligand basicity decreases going to the right and down.
in collaboration with Degussa5,6 and now being com-
mercialized by Solvias for Umicore. The Mandyphos di-
phosphines5 have a C2-symmetrical backbone with
structural motifs related to ppfa7 (see Fig. 3). Herein,
several C1 (or pseudoC2)-symmetrical derivatives
(R0„R00) were also prepared (M5–M8). The Taniaphos
family6 has a C1-symmetrical backbone and forms
eight-membered metallacycles similar to the Walphos
ligands (Fig. 3), independently developed around the
same time by Weissensteiner et al.8 In the original pub-
lications, it was shown that for both ligand types, vary-
ing the Rsubstituent on the side chain had a strong
effect on the ee, whereas the effect of R0 and R00 on phos-
phorus was not investigated systematically. Herein we
show that the nature of R0 and R00 has also a very strong
effect on the catalytic performance and that this can be
used to tailor the ligand to a particular transformation
with relatively little synthetic effort.
2. Results and discussion
2.1. Basic ligand profiling
The profiling of a new ligand is the most important as-
pect when assessing its usefulness for practical applica-
tions. To do this as effectively as possible we have
developed a two-pronged strategy. For the first assess-
ment of a ligandÕs potential and for comparisons with
literature results, we carried out several standard test
reactions with C@C, C@N and C@O double bonds2
(see Fig. 4) under standard reaction conditions (see
Table 1). Depending on the results of this so-called basic
profiling some or all ligands would then be tested either
on additional, more demanding test substrates or under
more demanding reaction conditions in order to assess
the potential of the ligand for industrial problems (this
we call extended profiling).
PR2
PR2
PR'2
CH3
N
CH3
CH3
R2P
R2P
Fe
H
Fe
Fe
H
H
Josiphos
ppfa
Walphos
Figure 3. Ferrocene based phosphine ligands with similar structural motifs.

F. Spindler et al. / Tetrahedron: Asymmetry 15 (2004) 2299–2306
2301
O
O
O
COOCH3
COOCH3
DMI
N
COOCH3
Ph
NH
O
O
COOH
O
Ph
Ph
COOEt
NHCOCH3
O
O
MAC
MCA
EAC
MEA Imin
PhCOCOOMe
Oxo-Val
Figure 4. Structures and abbreviations of standard test substrates used in the basic profiling.
Table 1. Standard reaction conditions for basic profiling
Substrate (mmol)
Precursor
Solvent (mL)
s/c
p (bar)
T (°C)
t (h)
Additive
MAC, MCA, DMI (2.53)
EAC (1.265)
[Rh(nbd)2]BF4
[Rh(nbd)2]BF4
[Ir(COD)Cl]2
MeOH (10)
EtOH (9.5)
Toluene (10)
Toluene (10)
EtOH (10)
200
100
100
200
200
1a
1
25
25
25
25
80
1
1
CF3CH2OH (0.5mL)
TBAI (4.8mg), CF3COOH (30lL)
MEA-Imine (12.65)
PhCOCOOMe (2.53)
Oxo-Val (2.53)
80
80
80
16
16
16
[Rh(nbd)Cl]2
[RuI2(p-cymene)]2
1M HCl (60lL)
a MCA at 5bar.
The results for the basic profiling of the Mandyphos and
Taniaphos families are summarized in Figures 5–7. As
already observed in the initial publications,5,6 the two
ligand families have quite different performance profiles.
Without going into detail, the following trends merit
comments:
formance for MAC were expected because of already
known results,5 it is not clear why DMI is hydrogen-
ated with such low ees.
ꢀ Taniaphos exhibited good to very good ees for MAC,
DMI, EAC and Oxo-Val and low to mediocre ees for
all other test substrates. T3 deserves special mention
with ees between 92% and >99.5%. Activities were
high for the first two and medium to low for the other
substrates. While good ees for MAC and DMI have
already been described,6 the good enantioselectivity
of Taniaphos for EAC and especially for the Ru cat-
alyzed hydrogenation of Oxo-Val was unexpected.
ꢀ Most Mandyphos ligands showed good to excellent
ees and low to good tofs for MAC and MCA but
low to mediocre ees for all other test substrates. In
particular, catalyst Rh-M4 hydrogenated MAC with
>99.5% ee and a very high rate. While the good per-
ee (%)
100
tof (1/h)
200
MAC
MCA
DMI
EAC
ee
80
60
40
20
0
tof
150
100
50
0
M2 M6 M5 M4 M7 M8 M1 M3
MEA imine
M2 M6 M5 M4 M7 M8 M1 M3
PhCOCOOMe
M2 M6 M5 M4 M7 M8 M1 M3
M2 M6 M5 M4 M7 M8 M1 M3
ee (%)
100
tof (1/h)
200
OxO-Val
ee
tof
80
60
40
20
0
150
100
50
0
M2 M6 M5 M4 M7 M8 M1 M3
M2 M6 M5 M4 M7 M8 M1 M3
M2 M6 M5 M4 M7 M8 M1 M3
Figure 5. Basic profiling of Mandyphos; ee and tof for eight ligands and eight test substrates. Ligands ordered according to decreasing basicity
(Cy>MeOXy>Xyl>Ph>CF3Xyl).

2302
F. Spindler et al. / Tetrahedron: Asymmetry 15 (2004) 2299–2306
ee (%)
tof (1/h)
200
Oxo-Val
MAC
MCA
DMI
EAC
MEA imine
PhCOCOOMe
100
80
60
40
20
0
150
100
50
ee
tof
0
T2 T3 T1
T2 T3 T1
T2 T3 T1
T2 T3 T1
T2 T3 T1
T2 T3 T1
T2 T3 T1
Figure 6. Basic profiling of Taniaphos; ee and tof for three ligands and eight test substrates. Ligands ordered according to decreasing basicity
(Cy>MeOXyl>Ph).
ee (%)
100
tof (1/h)
200
MA
DMI
MAC
MCA
MCA
Oxo-Val
ee
80
60
40
20
0
tof
150
100
50
0
M1 M6 M2 M8 M7 M5 M4 M3
M1 M6 M2 M8 M7 M5 M4 M3
M1 M6 M2 M8 M7 M5 M4 M3
T1 T2 T3
T1 T2 T3
T1 T2 T3
Figure 7. Ee and tof for Mandyphos and Taniaphos for selected test substrates. Ligands ordered according to increasing size (Ph<Cy<XylꢁMeO-
XylꢁCF3Xyl).
ꢀ For all test substrates and for both ligand families,
the nature of the PR2 groups had a significant influ-
ence on both the ee and tof. While there is no un-
equivocal correlation, a comparison of the selectivity
patterns depicted in Figures 5–7 shows that it is the
electronic property of the phosphine group (see Figs.
5 and 6) rather than its size (see Fig. 7), which has a
systematic effect on the catalytic properties of the
resulting complexes.
Scope: Study the effect of structural variations of
MAC and MCA and investigate the hydrogenation
of selected a-keto esters (see Fig. 8 and Table 2).
T1 Scope: Study the effect of structural variations of
MAC, EAC and Oxo-Val and investigate the
hydrogenation of selected a-keto esters, unsatu-
rated acidsandahydrazone (see Fig. 8andTable3).
ꢀ With very few exceptions, relatively electron rich all-
aryl substituted derivatives (Ph, Xyl, MeOXyl) gave
the best performances for both ligand families. Lig-
ands containing PCy2 and P(CF3Xyl)2 groups in gen-
eral gave moderate to very low ees and tofs. Except
for MCA, EAC and Oxo-Val (low ees), symmetrically
substituted Mandyphos derivatives usually showed a
somewhat better performance than comparable un-
symmetrical derivatives.
Furthermore, in some cases, pressure effects were inves-
tigated and for several substrates, both Rh and Ru cat-
alysts were tested.
Comments on the performance of M4 listed in Table 2
ꢀ
MAC can be hydrogenated with industrially useful
results: At 1bar ee>98%, tofs>7500hÀ1 and
tons>20,000 were achieved. The ee slightly dropped
at higher s/c ratios and values >20,000 led to incom-
plete conversions (here the purity of the starting
material starts becoming a critical issue). Increasing
the temperature to 35°C or the pressure to 5bar led
to a decrease in ee of ca. 1%. Compared to parent lig-
and M1, modified M4 clearly has superior enantio-
selectivity and activity.
2.2. Extended ligand profiling
Based on these screening results, it was decided to select
M4 and T1 for extended profiling with the following
goals:
ꢀ
The hydrogenation of dehydroamino acid deriva-
tives with Rh-M4 turned out to be quite sensitive
to substrate structure. While ACA, the correspond-
M4 Technical performance: Optimize the catalytic per-
formance (ee, ton, tof) for MAC.

F. Spindler et al. / Tetrahedron: Asymmetry 15 (2004) 2299–2306
2303
O
COOH
COOCH3
OH
COOCH3
OH
COOH
NHCOCH3
NHCOCH3
NHCOCH3
ACA
MAA
O
Cy=(NHAc)COOMe
Tiglic acid
3,3-diMe-2-Ph-acrylic acid
AllylOH
O
O
COOEt
COOMe
COOEt
O
O
N
NH
Ph(CH2)2COCOOEt
Etpy
PhCOCOOMe
O
O
O
O
O
O
O
OEt
COOEt
F3C
O
CF3COAcOiPr
AcAcH
cyCOAcOEt
PhCOAcOEt
Hydrazone
Figure 8. Structures and abbreviations of test substrates used in the extended profiling: dehydroamino acid derivatives, unsaturated acids, allylic
alcohol, a-keto esters, b-functionalized ketones, hydrazone.
Table 2. Best results for enamides, activated olefins and a-keto esters with Mandyphos M4 in the extended profiling
Substrate
Metala
p (bar)
t (h)
Ee (%)
Tof (hÀ1
)
Comments
MAC
MAC
Rh
Rh
Rh
Rh
Rh
Rh
1
1
1
5
1
1
20
1.3
7
97.6
98
50
7692
2857
11
ligand M1, s/c 1000
s/c 10,000, 35°C; ee 99% at 25°C
s/c 20,000
MAC
ACA
98.6
95
17.5
1
Ee 59% for Ru
MAA
Cy@(NHAc)COOMe
95
37
200
10
20
MCA
Tiglic acid
Rh
Ru
Rh
Rh
50
5
94
18
17
18
80
97
58
72
2
11
9b
10b
Ee 72% for Ru
Ee 76% for Rh
DiMe–Ph–acrylic acid
AllylOH
5
No reaction with Ru
No reaction with Ru
5
Etpy
Ph(CH2)2COCOOEt
PhCOCOOMe
Ru
Ru
Ru
50
50
50
21
18
22
69
88
66
5.2b
10b
5.4b
Solvent EtOH
Solvent EtOH
Reaction conditions: s/c 200, 25°C.
a Rh [Rh(nbd)2]BF4; Ru [RuI2(p-cymene)]2.
b Incomplete conversion.
Table 3. Best results for enamides, activated olefins, a- and b-keto esters and a hydrazone with Taniaphos T1 in the extended profiling
Substrate
Metala
p (bar)
t (h)
Ee (%)
Tof (hÀ1
)
Comments
MAC
MAA
Rh
Rh
Rh
5
1
5
22
20
20
96
98
0.5
9
10
10
Ee 97% at 1bar
Cy@(NHAc)COOMe
MCA
Tiglic acid
Rh
Ru
Rh
Rh
50
50
5
20
18
17
18
10
38
66
55
10
8b
Ee 32% at 5bar
DiMe–Ph–acrylic acid
AllylOH
8b
5
10b
Ph(CH2)2COCOOEt
AcAcH
Ru
Ru
Ru
Ru
Ru
Ru
50
80
10
80
80
80
20
16
20
18
16
20
41
9
13
Solvent EtOH
99.5
84
Additive 1M HCl; dl/meso 97
Solvent iPrOH, 1M HCl
Solvent EtOH, 1M HCl
Solvent EtOH, 1M HCl
Solvent EtOH, 1M HCl
CF3 COAcOiPr
cyCOAcOEt
PhCOAcOEt
AcAcOEtc
10
11
81
94
13
1200
79
Hydrazone
Rh
5
20
33
1
Conversion 11%
Reaction conditions: MeOH, s/c 200, 25°C.
a Rh [Rh(nbd)2]BF4; Ru [RuI2(p-cymene)]2.
b Incomplete conversion.
c s/c 25,000.

2304
F. Spindler et al. / Tetrahedron: Asymmetry 15 (2004) 2299–2306
ing free acid, and MAA (no phenyl group) were
ꢀ
Other b-keto esters were also hydrogenated with
hydrogenated with significantly lower but still very
good enantioselectivities, the tetra substituted
Cy@(NHAc)COOMe only gave 37% ee. This is in
contrast to the Rh-Duphos catalysts, which are less
sensitive to the degree of substitution and configura-
tion of the C@C bond.2
An even stronger structure sensitivity was found for
a,b-unsaturated olefins. An unprecedented perform-
ance was observed for the Ru catalyzed hydrogena-
tion of tiglic acid (ee 97%, full conversion). The best
results for MCA and 3,3-dimethyl- 2-phenyl-acrylic
acid were obtained with Rh catalysts with ees of
80% and 58%, respectively, and allylOH reduced
with 72% ee.
good to high ees, whereas a-keto esters or hydr-
azones seemed to be unsuitable substrates for this
Ru catalyst.
A broad selection of alkenes was studied but only
di and trisubstituted a-enamides gave high enantio-
selectivities. All other tested C@C substrates were
hydrogenated with medium to very low ees. The per-
formance of this catalyst for a-enamides is strongly
dependent on the degree of substitution at the
C@C bond.
ꢀ
ꢀ
ꢀ
3. Conclusions
Acceptable results were also obtained for a-keto
esters. The best enantiomeric excess was obtained
for the hydrogenation of the commercially interest-
ing HPB-keto ester (Ph(CH2)2COCOOEt, 88%
ee) while ees of up to 69% ee and 66% ee
were achieved for etpy and PhCOCOOMe, respec-
tively.
The results summarized in Table 4 show that Mandy-
phos and Taniaphos are versatile ligands with very good
catalytic performances for a variety of substrate classes.
For acrylic acid derivatives, itaconates and b-diketones,
the performance was equal to or better than that known
for leading commercial ligands. The basic profiling has
shown that the choice of the substituents at the P atoms
has a significant effect on the catalyst performance.
Interestingly, for both ligand families, relatively electron
rich aromatic substituents are most suitable for transfor-
mations with high enantioselectivity, whereas cyclohexyl
(exception MAC) and CF3-substituted aryl groups have
a negative influence on ee and often the activity as well.
This is in contrast to the Josiphos ligands with which
several cases are known where these substituents are
essential for good performance.4
Comments on the performance of T1 listed in Table 3
ꢀ
Taniaphos ligand T1 is suitable for the Rh catalyzed
hydrogenation of selected enamides and is very
efficient for the Ru catalyzed hydrogenation of
b-diketones and b-keto esters. The results for
the other tested substrate classes confirm the
initial tests and are not useful for preparative applica-
tion.
ꢀ
The most outstanding result was obtained in the
hydrogenation of acetylacetone yielding (2R,4R)-
pentanediol with 99.5% ee with an excellent
diastereoselectivity of 97.4:1 (dl:meso). At this
small scale, Ru-SL is an alternative for the
Ru-Binap or R–MeObiphep catalysts; the potential
for large scale application has yet to be
investigated.
4. Experimental
All ligands were prepared in analogy to the procedures
published by Knochel and co-workers5,6 and were fully
characterized. All ligands can be obtained from Solvias
and the derivatives listed in Table 5 are also available
from Strem.
ꢀ
AcAcOEt could be hydrogenated with Ru-T1 on a
154mmol scale with an s/c ratio of 25,000 with
complete conversion after 20h, but with only
78.8% ee; significantly lower than the 96% ee
obtained at s/c 100.5 The reasons for this drop in
enantioselectivity is not clear and should be studied
in more detail.
The hydrogenation experiments were carried out with
the test substrates listed in Figures 4 and 8 in analogy
as described in detail for two selected substrates below.
The screening experiments were carried out in parallel in
10mL glass vials. Selected single experiments were car-
ried in larger glass vessels or stainless steel autoclaves.
Table 4. Comparison of the best ees of the Mandyphos and Taniaphos ligands with other ligand classes
Substrate class
Umicore ligand (not optimized)
Other commercial ferrocenyl ligands
Josiphos (98%)
Other commercial state of the art ligands
Rophos, Duphos Butiphane (99%)
Disubstituted
a-Enamides
Trisubstituted
a-Enamides
b-Enamides
Acrylic acids
M4 (95%)
T1 (>98%)
M4 (>98%)
T1 (>96%)
Josiphos, Taniaphos6 (99%)
Rophos, Duphos Butiphane (99%)
T1, T3 (up to 92%)
M4 (97%)
Josiphos (>95%)
Walphos (>92%)
Josiphos (98%)
Duphos, Bicp, Tangphos (99%)
MeObiphep (>90%)
Itaconates
T3 (>99%)
Josiphos (99%)
Rophos (>95%)
Butiphane, Duphos, FerroTANE (>95%)
a-Keto esters
b-Keto esters
b-Diketones
M4 (88%) M1 (95%)9
T1 (95%)
T1 (>99%)
Josiphos (60%)
Taniaphos (99%), Walphos (96%)
Walphos (96%)
MeObiphep (94%)
Binap, MeObiphep, Segphos (>99%)
Binap, MeObiphep, Segphos (>99%)

F. Spindler et al. / Tetrahedron: Asymmetry 15 (2004) 2299–2306
Table 5. Mandyphos and Taniaphos ligands available from Strem
2305
20
½aꢂD
Abbreviation
Solvias no
Strem no
CAS Reg. no
Name of ligand
Acronym
Molecular formula, MW
Chemical purity
Enantiomeric excess
M1
SL-M001-1
(aR,aR)-2,20-Bis(a-N,N-dimethylaminophenylmethyl)-
(S,S)-1,10-bis(diphenylphosphino)ferrocene
(R)-(S)-NMe2–PPh2–Mandyphos
>À320
26-0252
[210842-74-3]
>97%
>99% ee
C
52H50FeN2P2, 820.28
M2
M3
M4
T1
SL-M002-1
(aR,aR)-2,20-Bis(a-N,N-dimethylaminophenylmethyl)-
(S,S)-1,10-bis(dicyclohexylphosphino)ferrocene
(R)-(S)-NMe2–PCy2–Mandyphos
>+54
26-0240
[494227-35-9]
>97%
>99% ee
C
52H74FeN2P2, 844.97
SL-M003-1
(aR,aR)-2,20-Bis(a-N,N-dimethylaminophenylmethyl)-(S,S)-1,10-
bis[di(bis-(3,5-trifluoromethyl)phenyl)phosphino]ferrocene
(R)-(S)-NMe2–P(3,5-CF3Ph)2–Mandyphos
>À250
26-0244
[494227-36-0]
>97%
>99% ee
C60H42F24FeN2P2, 1364.75
SL-M004-1
(aR,aR)-2,20-Bis(a-N,N-dimethylaminophenylmethyl)-(S,S)-1,10-
bis[di(3,5-dimethyl-4-methoxyphenyl)phosphino]ferrocene
(R)-(S)-NMe2–P(3,5-Me-4–MeOPh)2–Mandyphos
C64H74FeN2O4P2, 1053.09
>À61
26-0248
[494227-37-1]
>97%
>99% ee
SL-T001-1
(1S)-Diphenylphosphino-2-[(R)-a-(N,N-dimethylamino)-o-
diphenylphosphinophenyl)methyl]ferrocene
(R)-(S)-Ph2PPhCHNMe2–T–PPh2
>+312
26-1155
[255884-98-1]
>97%
>99% ee
C
43H39FeNP2, 687.56
T2
SL-T002-1
(1S)-Dicyclohexylphosphino-2-[(R)-a-(N,N-dimethylamino)-o-
dicyclohexylphosphinophenyl)methyl]ferrocene
(R)-(S)-Cy2PPhCHNMe2–T–PCy2
>+110
26-0955
[494227-38-2]
>97%
>99% ee
C43H63FeNP2, 711.79
The reaction mixtures were analyzed by GC and HPLC
methods using literature methods.
of 97.4:1.0 and an enantiomeric purity of 99.5% [deter-
mined after derivatization with trifluoroacetic anhydride
using GC (column: Lipodex E, 50m·0.25mm, 80°C,
isothermic; carrier gas: H2 120kPa)].
4.1. Ru catalyzed hydrogenation of acteylacetone (small
scale single experiment)
4.2. Rh catalyzed hydrogenation of Z-acetamidocinna-
mate (scale up experiment)
The autoclave was purged with argon by setting the
pressure to 40bar and releasing it. This operation was
repeated four times. The starting material acteylacetone
(0.254g, 2.53mmol) and 5mL of degassed methanol
were placed in a 10mL Schlenk flask with a stirring
bar and a sequence high vacuum/argon-filing repeated
six times. 60lL of 1M methanolic HCl was carefully
added. In a 10mL Schlenk flask (also under argon, same
procedure as above) [RuI2(p-cymene)]2 (6.2mg;
0.0063mmol) and T1 (9.1mg; 0.0133mmol) were placed,
and dissolved in methanol (5mL). These solutions were
stirred at room temperature for 10min and then trans-
ferred via cannula to the autoclave with a gentle argon
flow. The autoclave was purged with hydrogen (40bar,
four times), set at 60bar. The autoclave was then heated
up to 80°C, and after 30min, the hydrogen pressure set
to 80°C. After 16h reaction time, the pressure was re-
leased. The clear reaction mixture was carefully evapo-
rated to dryness under reduced pressure.
The glass reactor was set under an atmosphere of argon
by repeated sequences of high vacuum/argon-filling
(four times). The starting material, methyl Z-acetamido-
cinnamate (13.18g, 63.0mmol), and 40mL of degassed
methanol were placed in a 100mL Schlenk flask with a
stirring bar and a sequence of high vacuum/argon-filling
repeated six times. In a 100mL Schlenk flask (also under
argon, same procedure as above) [Rh(nbd)2]BF4 (1.2mg;
0.0032mmol) and M4 (3.74mg; 0.0036mmol) were
placed, and dissolved in methanol (40mL). These solu-
tions were stirred at room temperature for 10min and
then transferred via cannula to the glass reactor with a
gentle argon flow. The reactor was purged with hydro-
gen/vacuum (1.2bar/0.4bar, six times), set at 1.05bar
hydrogen, the temperature maintained at 25°C and stir-
ring then started. After 7h reaction time, the hydrogen
was replaced by argon. The clear reaction mixture was
evaporated to dryness under reduced pressure.
2,4-Pentanediol was quantitatively obtained (deter-
mined by GC and H NMR). The predominant stereo-
isomer was (2R,4R)-pentanediol with a dl/meso ratio
1
Conversion and enantiomeric purity were determined by
using GC [Chirasil-L-Val: 30m; 150°C, 40, 10min,

2306
F. Spindler et al. / Tetrahedron: Asymmetry 15 (2004) 2299–2306
2. Blaser, H. U.; Malan, Ch.; Pugin, B.; Spindler, F.; Steiner,
H.; Studer, M. Adv. Synth. Catal. 2003, 345, 103.
3. Tang, W.; Zhang, X. Chem. Rev. 2003, 103, 3029.
4. Blaser, H. U.; Brieden, W.; Pugin, B.; Spindler, F.; Studer,
M.; Togni, A. Top. Catal. 2002, 19, 3.
190°C, 20min; carrier gas: He (200kPa)]. The conver-
sion was quantitative and the enantiomeric excess of
N-acetyl phenylalanine methyl ester determined to be
98.6% (R).
5. (a) Almena Perea, J.; Bo¨rner, A.; Knochel, P. Tetrahedron
Lett. 1998, 39, 8073; (b) Lotz, M.; Ireland, T.; Almena
Perea, J.; Knochel, P. Tetrahedron: Asymmetry 1999, 10,
1839; (c) Almena Perea, J.; Lotz, M.; Knochel, P. Tetra-
hedron: Asymmetry 1999, 10, 375.
6. Ireland, T.; Tappe, K.; Grossheimann, G.; Knochel, P.
Chem. Eur. J. 2002, 8, 843, and references cited therein.
7. Hayashi, T. In Ferrocenes; Togni, A., Hayashi, T., Eds.;
VCH: Weinheim, 1995; p 105.
Acknowledgements
The authors thank Bruno Casagrande, Andrea Holder-
er, Willi Lottenbach, Oanh Mai-Huynh and Thomas
Stebler for careful preparative work.
References and notes
8. Weissensteiner, W.; Sturm, T.; Spindler, F. Adv. Synth.
Catal. 2003, 345, 160.
9. Knochel, P.; Almena Perea, J.; Drautz, K.; Klement, I., EP
1. Catalytic Asymmetric Synthesis; Ojima, I., Ed.; Wiley–
VCH: New York, 2000.
965574 A2, 1999 (assigned to Degussa-Huls).
¨