Asymmetric catalytic hydroformylation of styrene: Stereoselective ferrocenylethyldiphosphine rhodium catalysts

Article abstract of DOI:10.1016/S0022-328X(00)00046-2

The use of a variety of chiral ferrocenylethyl diphosphines and a rhodium(I) precursor for the asymmetric hydroformylation of styrene is described. Some of these catalysts yield the chiral 2-phenylpropionic aldehyde with high enantioselectivity. The selectivity and activity of the catalysts are influenced by the substitution pattern of the phosphines. High enantioselectivity was achieved even at high reaction temperatures with a dialkylaryl-alkyldiaryl diphosphine ligand. o-Anisyl substituents at the side-chain phosphorus yield optical inductions of up to 76% ee, but at low conversion rates.

Full text of DOI:10.1016/S0022-328X(00)00046-2

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 601 (2000) 138–141
Asymmetric catalytic hydroformylation of styrene: stereoselective
ferrocenylethyldiphosphine rhodium catalysts
Florian A. Rampf, Wolfgang A. Herrmann *
Anorganisch-Chemisches Institut, Technische Uni6ersita¨t Mu¨nchen, Lichtenbergstraße 4, D-85747 Garching, Germany
Received 31 December 1999; accepted 16 January 2000
Abstract
The use of a variety of chiral ferrocenylethyl diphosphines and a rhodium(I) precursor for the asymmetric hydroformylation of
styrene is described. Some of these catalysts yield the chiral 2-phenylpropionic aldehyde with high enantioselectivity. The
selectivity and activity of the catalysts are influenced by the substitution pattern of the phosphines. High enantioselectivity was
achieved even at high reaction temperatures with a dialkylaryl–alkyldiaryl diphosphine ligand. o-Anisyl substituents at the
side-chain phosphorus yield optical inductions of up to 76% ee, but at low conversion rates. © 2000 Elsevier Science S.A. All
rights reserved.
Keywords: Rhodium; Hydroformylation; Asymmetric catalysis; Diphosphines; Ferrocenylphosphines
1. Introduction
amino function with a secondary phosphine [7]. These
ligands possess two elements of chirality: the chiral
center in the side chain and the planar chirality of the
disubstituted ferrocene. They have been used for cata-
lytic asymmetric reactions and proved to be efficient in
asymmetric hydrogenation, but also in hydroboration,
Kumada coupling and allylic alkylation reactions [8].
Despite their widespread application in asymmetric
catalysis, their use for enantioselective hydroformyla-
tion has not yet been reported. We now describe asym-
metric hydroformylation using chiral ferrocenylethyl
diphosphines with rhodium as catalyst metal.
Hydroformylation is the largest-volume technical
process of organometallic catalysis [1,2]. Beyond that,
asymmetric hydroformylation of styrene derivatives
promises a shortcut access to anti-inflammatory drugs,
yielding chiral 2-arylpropionic aldehydes, which can
easily be oxidized to pharmaceutically active 2-arylpro-
pionic acids like Naproxen or Ibuprofen [3]. However,
efficient catalysts for this reaction are still rare, and
asymmetric hydroformylation is still a challenging topic
in enantioselective catalysis [4]. Remarkable results with
rhodium-based catalyst systems were obtained by
Takaya and co-workers using the phosphine–phosphite
ligand BINAPHOS [5] and by DuPont researchers em-
ploying a diphosphite ligand [6]. However, simple phos-
phines yield only little enantioselectivity [4].
2. Results and discussion
Several in situ rhodium catalysts formed from
Rh(CO)2(acac) and two equivalents of various ferro-
cenylethyl diphosphines were tested for the enantiose-
lective hydroformylation of styrene. Styrene was chosen
as model olefin, because vinylic aromatics are the most
interesting substrates for an application of this reaction
in the synthesis of fine chemicals (Scheme 1). The
reaction yields the branched main product 2-phenylpro-
pionic aldehyde 1 and a small amount of the linear
3-phenylpropionaldehyde 2.
Diastereomerically pure ferrocenylethyl diphosphines
are conveniently synthesized from enantiopure 1-ferro-
cenylethylamine after diastereoselective ortho-lithiation
of the ferrocenyl backbone, reaction with chlorophos-
phines and subsequent nucleophilic displacement of the
* Corresponding author. Tel.: +49-89-28913080; fax: +49-89-
28913473.
E-mail address: lit@arthur.anorg.chemie.tu-muenchen.de (W.A.
Herrmann)
0022-328X/00/$ - see front matter © 2000 Elsevier Science S.A. All rights reserved.
PII: S 0 0 2 2 - 3 2 8 X ( 0 0 ) 0 0 0 4 6 - 2

F.A. Rampf, W.A. Herrmann / Journal of Organometallic Chemistry 601 (2000) 138–141
139
alkylaryl and one alkyldiaryl phosphino group. When
this ligand is used, the reaction rate is smaller, whereas
the enantioselectivity of the catalyst is higher, yielding
61% ee, notably of the opposite (S)-enantiomer. In-
creasing the electronic difference between the two phos-
phine atoms of the ligand by the introduction of
o-trifluoromethyl substituents in the aryl substituents
does not influence the reaction rate but reduces the
selectivity severely (entry 4). However, the additional
steric bulk of this group could also be a reason for the
observed effect.
Takaya and co-workers have observed that in cata-
lytically active RhH(CO)2(PSP%) complexes (PSP%:
chelating ligand) the C1-symmetric ligand BINAPHOS
shows a preferred coordination of its phosphite arm in
an apical position opposite to the hydride ligand [9]. A
similar behavior of the ferrocenylethyl diphosphines,
which are also C1-symmetric, is made plausible by the
change of selectivity observed when the positions of the
different phosphines are switched. If one phosphine
unit favors coordination in the apical position, the
ligand must turn round upon exchange of the different
phosphine substituents. This could lead to the observed
change of the preferred product enantiomer.
Scheme 1. Hydroformylation of styrene yields chiral 2-phenylpropi-
onic aldehyde and 3-phenylpropionic aldehyde as a side product.
Initially, all catalysts were tested at 60°C. However,
catalysts formed from ligands with alkyl-substituted
phosphorus showed very little conversion at that tem-
perature and were subsequently tested at a temperature
of 80°C (Table 1, entries 1–4).
The first four entries of Table 1 show a strong
influence of the substitution pattern of the phosphorus
atoms on the activity and selectivity of the rhodium
catalysts. When the basic phosphorus moiety bearing
two cyclohexyl groups is located in the side chain of the
ferrocenyl backbone (entry 1), the ligand has one tri-
alkyl and one triaryl function coordinating to the
rhodium metal. In this case a conversion of 32% and an
excess of 34% of the (R)-enantiomer is observed. t-
Butyl substituents at the side-chain phosphorus cause a
comparable selectivity but a reduced activity of the
catalyst (entry 2). The exchange of the phosphorus
substituents (entry 3) results in a ligand with one di-
Entries 5–9 show the influence of the substitution
pattern of aromatic substituents at the phosphine in the
chiral side chain. Entries 5 and 6 demonstrate that the
steric bulk of ortho-substituents influences the selectiv-
Table 1
Catalytic hydroformylation of styrene with different ferrocenylethyl diphosphines a
Entry
R
R%
Conversion (%)
ee (%)
1
2
3
4
5
6
7
8
Pcy2
PPh2 (80°C)
PPh2 (80°C)
PCy2 (80°C)
P[2-(CF3)ꢀC6H4]2 (80°C)
PPh2
PPh2
PPh2
PPh2
PPh2
PPh2
PPh2
PPh2
PPh2
PPh2
32
4
34 (S)
30 (R)
60 (S)
4 (S)
5 (R)
36 (S)
51 (S)
30 (S)
3 (S)
–
P(tBu)2
PPh2
Pcy2
P(o-Tol)2
P(m-Tol)2
P(o-Anis)2
P(m-Anis)2
P(p-Anis)2
P[2-(CF3)ꢀC6H4]2
P[3,5-(CF3)2ꢀC6H3]2
P[4-(tBu)ꢀC6H4]2
P[3,5-(CH3)2ꢀC6H3]2
P[3,5-(CH3)2-4-NBu2ꢀC6H3]2
13
29
54
66
26
16
11
100
82
51
46
72
9
10
11
12
13
14
2 (R)
3 (S)
3 (S)
4 (S)
a Reaction conditions: Rh(CO)2(acac)–ligand–styrene ratio: 1:2:300; 10 MPa H2–CO (1:1 mixture). Temperature: 60°C unless otherwise stated.
Reaction time: 16 h. Branched/linear-ratios were between 9 and 15 in all cases. (Cy=C6H11
Anis=CH3OꢀC6H4).
,
tBu=t-C4H9, Ph=C6H5, Tol=CH3ꢀC6H4,

140
F.A. Rampf, W.A. Herrmann / Journal of Organometallic Chemistry 601 (2000) 138–141
alkyldiaryl phosphine (Table 1, entry 3) was studied
because the conversion of styrene at 80°C was still very
low. Therefore, temperatures of 90, 100 and 120°C were
applied (Table 2).
In these experiments an increased activity of the
catalyst at higher temperatures was observed, accompa-
nied by an increasing loss of enantioselectivity. How-
ever, the enantiomeric excess of the iso-aldehyde
product remained nearly constant at about 60% ee up
to a temperature of 100°C.
ity and activity of the catalyst negatively: The reaction
with the ortho-tolyl-substituted ligand yields only an
enantiomeric excess of 5% (R), which is considerably
lower than the 36% ee (S) for the meta substituents. In
entries 7–9, the ring position of a methoxy substituent
at the side-chain phosphorus illustrates that the pres-
ence of a donor substituent in the close vicinity of the
phosphorus atom can increase the excess of the pre-
ferred enantiomer. Whereas the para-anisyl-substituted
ligand only yields 3% ee, the ligand with meta-anisyl
substituents generates an enantiomeric excess of 30%,
and the ortho-substituted ligand shows 51% ee for the
same enantiomer. As the steric influence of the ortho-
methoxy substituent on the rhodium atom is the
strongest among the three isomeric ligands, its negative
effects on the selectivity of the catalyst must be com-
pensated by another effect. A plausible explanation of
this behavior is that the oxygen of the ortho-methoxy
substituents helps to stabilize one of several possible
complex configurations in the catalytic cycle and there-
fore influences the product distribution of the overall
reaction.
The catalyst system used in Table 1, entry 7, showed
a comparable temperature dependence of the enantio-
selectivity (Table 3).
In this case the selectivity of the catalyst could even
be improved on running the reaction at lower tempera-
tures. At 40°C, the 2-phenylpropionic aldehyde was
formed with 76% ee, the highest enantiomeric excess
obtained with a rhodium diphosphine catalyst system.
At 75°C, however, the catalyst was considerably less
selective and yielded only 31% ee. This supports the
concept of a weak additional coordination of the oxy-
gen atom in the ortho-anisyl group to the rhodium
atom. As this interaction competes with the strongly
binding phosphine and carbonyl ligands, it will not be
observable in the resting state of the catalyst. However,
dissociation of one carbonyl ligand or one arm of the
diphosphine opens a free coordination site. This coordi-
natively unsaturated complex might then be stabilized
in one preferred configuration by the oxygen, before it
is attacked by the substrate styrene. Higher tempera-
tures would make an interaction of this kind more
difficult.
Other chiral ferrocenylethyl diphosphines with elec-
tronically different phosphorus moieties listed in entries
10–14 did not show enantioselectivities comparable to
those of the ligands discussed before.
The temperature dependence of the selectivity and
activity of the catalyst employing the dialkylaryl and
Table 2
Temperature dependence of the selectivity observed for R=Ph and
R%=Cy a
Entry
Temperature (°C)
Conversion (%)
ee (%)
15
16
17
18
19
60
80
90
100
120
3
13
31
47
84
63 (S)
60 (S)
57 (S)
57 (S)
45 (S)
3. Conclusions
The stepwise introduction of the phosphine moieties
in the synthesis of ferrocenylethyl phosphines permits
the design of bidentate ligands with two sterically and
electronically different phosphine atoms. This opens the
opportunity to study the influence of different elec-
tronic and steric properties of both phosphorus atoms
independent from one another.
a Reaction conditions: Rh(CO)2(acac)–ligand–styrene ratio:
1:2:300; 10 MPa H2–CO (1:1 mixture). Reaction time: 16 h.
Table 3
The use of these ligands for the asymmetric hydro-
formylation of styrene with a rhodium(I) precursor has
been described for the first time. Some of them show
remarkably high stereoselectivities. The substitution
pattern of the phosphine atoms strongly influences the
selectivity and activity of the catalyst. With one dialky-
laryl–alkyldiaryl diphosphine, high enantioselectivity
was achieved even at high reaction temperatures and
therefore increased conversion of the substrate. o-Ani-
syl substituents at the side-chain phosphorus allow up
Temperature dependence of the selectivity observed for R=o-Anis
and R%=Ph a
Entry
Temperature (°C)
Conversion (%)
ee (%)
20
21
22
23
40
50
60
75
3
8
26
78
76 (S)
75 (S)
50 (S)
31 (S)
a Reaction conditions: Rh(CO)2(acac)–ligand–styrene ratio:
1:2:300; 10 MPa H2–CO (1:1 mixture). Reaction time: 16 h.

F.A. Rampf, W.A. Herrmann / Journal of Organometallic Chemistry 601 (2000) 138–141
141
to 76% ee; however, the conversion rates are still too
Acknowledgements
low for technical applications.
This work was supported by the Bundesministerium
fu¨r Bildung und Forschung (Bonn, Germany), the
Fonds der Chemischen Industrie (Frankfurt, Germany)
and Novartis Services AG (Basel, Switzerland). We are
grateful to Petra Ankenbauer for experimental
assistance.
4. Experimental
All reactions were carried out using standard Schlenk
techniques in an oxygen-free nitrogen atmosphere. Sol-
vents were dried with standard methods and distilled
under nitrogen. Chiral ferrocenylethyl diphosphines
were prepared following published methods [10]. Their
diastereomeric excess was 96% or higher. Hydroformy-
lation experiments were performed in a Parr 300 ml
stainless-steel autoclave, using dried and degassed
styrene and a 1:1 mixture of hydrogen and carbon
monoxide. The catalysts were formed in situ from
Rh(CO)2(acac) and two equivalents of ligand in
toluene, before the calculated amount of styrene was
added and the mixture transferred to the autoclave
under nitrogen. Modification of the commercially avail-
able autoclave allowed several catalytic reactions to be
run at the same time without cross-contamination of
the different samples. Through this modification the
volume of the reaction mixtures could be reduced to 1.5
ml and the amount of rhodium catalyst to 1 mmol.
Conversions were determined on an HP 5890 gas chro-
matograph with a 12 m HP-1 capillary column and
flame ionization detection using the method of internal
standardization. Enantiomeric excesses of the aldehydes
were measured on a Chrompack CP 9000 gas chro-
matograph (50 m Lipodex A column, carrier gas he-
lium, split injector, flame ionization detector) after
reduction to the corresponding alcohols with sodium
borohydride. Their absolute configuration was deter-
mined by comparison with an authentic sample of
(R)-(+)-2-phenylpropanol.
References
[1] (a) C.D. Frohning, C.W. Kohlpaintner, in: B. Cornils, W.A.
Herrmann (Eds.), Applied Homogeneous Catalysis with
Organometallic Compounds: A Comprehensive Handbook in
Two Volumes, vol. 1, VCH, Weinheim, 1996, pp. 27–104. (b)
W.A. Herrmann, C.W. Kohlpaintner, Angew. Chem. 105 (1993)
1588; Angew. Chem. Int. Ed. Engl. 32 (1993) 1524.
[2] W.A. Herrmann, J.A. Kulpe, J. Kellner, H. Riepl, H. Bahrmann,
W. Konkol, Angew. Chem. 102 (1990) 408; Angew. Chem. Int.
Ed. Engl. 29 (1990) 391.
[3] H. Smuda, GIT Fachztg. Labor 12 (1995) 1159.
[4] (a) F. Agbossou, J.-F. Carpentier, A. Mortreux, Chem. Rev. 95
(1995) 2485. (b) G. Consiglio, in: I. Ojima (Ed.), Catalytic
Asymmetric Synthesis, VCH, Weinheim, 1993, pp. 273–302.
[5] N. Sakai, S. Mano, K. Nozaki, H. Takaya, J. Am. Chem. Soc.
113 (1993) 7033.
[6] J.E. Babin, G.T. Whiteker (Union Carbide Chemicals & Plastics
Technology Corporation), US Patent No. 5 491 266, 1994; US
Patent No. 5 360 938, 1992.
[7] A. Togni, T. Hayashi, Ferrocenes. Homogeneous Catalysis, Or-
ganic Synthesis, Materials Science, VCH, Weinheim, 1994.
[8] (a) A. Togni, F. Spindler, N. Zanetti, A. Tijani (Ciba–Geigy
AG), EP 564 406, 1993. (b) T. Hayashi, N. Kawamura, Y. Ito, J.
Am. Chem. Soc. 109 (1987) 7876. (c) K. Tamao, A. Minato, T.
Matsuda, Y. Kiso, M. Kumada, Chem. Lett. (1975) 133.
[9] K. Nozaki, N. Sakai, T. Nanno, T. Higashijama, S. Mano, T.
Horiuchi, H. Takaya, J. Am. Chem. Soc. 119 (1997) 4413.
[10] (a) F. Spindler, A. Wirth-Tijani, H. Landert (Ciba–Geigy AG),
EP 612 758, 1994. (b) F. Spindler (Ciba–Geigy AG), EP 646 590,
1994.
.