Synthesis and application of phosphinoferrocenylaminophosphine ligands for asymmetric catalysis

Article abstract of DOI:10.1021/jo048312y

(Chemical Equation Presented) A new class of bidentate ligands utilizing a phosphine-aminophosphine structure has been prepared on a ferrocenylethyl backbone in a straightforward and scalable fashion from acetylferrocene. The unique property of the α-ferrocenyl carbonium ion that allows the replacement of a variety of "leaving groups" with retention of configuration greatly facilitates the synthesis, and a number of ligands have been prepared by varying the nitrogen and phosphorus substituents on the aminophosphine. These readily prepared phosphinoferrocenylaminophosphines, known as BoPhoz ligands, show surprising hydrolytic and air stability, with no degradation after 3 years open to the air. The rhodium complexes of these ligands show exceedingly high enantioselectivities (generally > 95% ee) and activities often in excess of 50 000 catalyst turnovers per hour for the asymmetric hydrogenation of a wide variety of dehydro-α-amino acid and itaconic acid derivatives. They also show high activity and good to excellent enantioselectivity for the hydrogenation of a number of α-ketoesters.

Full text of DOI:10.1021/jo048312y

Synthesis and Application of
Phosphinoferrocenylaminophosphine Ligands for Asymmetric
Catalysis
Neil W. Boaz,* Elaine B. Mackenzie, Sheryl D. Debenham, Shannon E. Large, and
James A. Ponasik, Jr.
Research Laboratories, Eastman Chemical Company, P.O. Box 1972, Kingsport, Tennessee 37662
nwboaz@eastman.com
Received September 22, 2004
A new class of bidentate ligands utilizing a phosphine-aminophosphine structure has been prepared
on a ferrocenylethyl backbone in a straightforward and scalable fashion from acetylferrocene. The
unique property of the R-ferrocenyl carbonium ion that allows the replacement of a variety of “leaving
groups” with retention of configuration greatly facilitates the synthesis, and a number of ligands
have been prepared by varying the nitrogen and phosphorus substituents on the aminophosphine.
These readily prepared phosphinoferrocenylaminophosphines, known as BoPhoz ligands, show
surprising hydrolytic and air stability, with no degradation after 3 years open to the air. The rhodium
complexes of these ligands show exceedingly high enantioselectivities (generally >95% ee) and
activities often in excess of 50 000 catalyst turnovers per hour for the asymmetric hydrogenation
of a wide variety of dehydro-R-amino acid and itaconic acid derivatives. They also show high activity
and good to excellent enantioselectivity for the hydrogenation of a number of R-ketoesters.
Introduction
A large number of complexes have been prepared and
examined for asymmetric catalysis. The great majority
of phosphorus-based ligands are C -symmetric carbon-
2
1
Asymmetric catalysis is at the forefront of organic
chemistry research due to the immaturity of the science
and its unparalleled potential for the generation of single
enantiomer materials. The ever-increasing importance of
single enantiomer pharmaceuticals and the lack of
general and effective alternative methodology for the
preparation of these materials and their building blocks
have fueled intense interest in finding useful asymmetric
catalysts.
linked phosphines (including triaryl, trialkyl, and aryl-
alkyl species). Significantly lesser in number are other
2
arrangements such as bidentate bis-aminophosphines,
3
4
5
bis-phosphites, bis-phosphinites, bis-phosphonites, and
mixed ligands such as phosphine-phosphite, amino-
phosphine-phosphinite, phosphine-phosphonite, phos-
6
7
8
Asymmetric catalysts are usually discrete metal com-
plexes containing one or more chiral ligands. These
ligands are usually enantiomerically pure and are the
agents inducing the asymmetry, generally through steric
interactions, although electronic effects can play a role.
The complexes can either be isolated materials or can
be prepared in situ using the appropriate amount of
ligand and a metal catalyst precursor.
(
1) For recent reviews, see: (a) Ohkuma, T.; Kitamura, M.; Noyori,
R. In Catalytic Asymmetric Synthesis, 2nd ed.; Ojima, I., Ed.; Wiley-
VCH: New York, 2000; pp 1-110. (b) Brown, J. M. in Comprehensive
Asymmetric Catalysis; Jacobsen, E. N., Pfalz, A., Yamamoto, H., Eds.;
Springer-Verlag: Berlin, 1999; Vol. I, pp 121-182. (c) Tang, W.; Zhang,
X. Chem. Rev. 2003, 103, 3029-3069.
(
2) Fiorini, M.; Giongo, G. M. J. Mol. Catal. 1980, 7, 411 and
references therein.
3) Reetz, M. T.; Neiberger, T.Angew. Chem., Int. Ed. 1999, 38, 179
and references therein.
(
10.1021/jo048312y CCC: $30.25 © 2005 American Chemical Society
1872
J. Org. Chem. 2005, 70, 1872-1880
Published on Web 02/08/2005

Phosphinoferrocenylaminophosphine Ligands for Asymmetric Catalysis
SCHEME 1. BoPhoz Ligand Synthesis
FIGURE 1. Phosphine-aminophosphine ligands.
phine-phosphoramidite, and phosphonite-phosphite10
9
systems.
Although many of these ligands afford good asym-
metric induction for a narrow set of substrates, there are
only a few that consistently afford superior enantiose-
lectivity, even for a single reaction type. Unfortunately,
these latter ligands tend to be very dear as they are often
structurally complex, challenging to prepare, and not
particularly air-stable. Thus, a highly enantioselective
readily prepared cost-effective ligand or series of ligands
for asymmetric catalysis would be of great interest. This
paper deals with the preparation and utility of just such
a family of ligands.11
1
4
syntheses. The preparation of ligands such as BPPFA,
JOSIPHOS,15 and several others are based on this
1-ferrocenylethyl backbone. The phosphine-aminophos-
phine ligands shown in structures R,S- and S,R-1 (Figure
16
1
b) were designed to take advantage of the synthetic
Results and Discussion
simplicity and strong steric environment of the ferroc-
enylethyl system while introducing a new phosphine-
aminophosphine ligand class for asymmetric catalysis.
The synthesis of 1 is predicated on the preparation of
either enantiomer of N,N-dimethyl 2-diphenylphosphi-
noferrocenylethylamine (2) from acetylferrocene. The
initial side-chain chirality can be introduced in high
enantiomeric purity (generally >90% ee) via a number
Ligand Design and Synthesis. Although there are
a wide variety of reported ligand substitution patterns,
there are no examples of phosphine-aminophosphine
species as ligands for asymmetric catalysis. The lone
proline-derived example of this type of compound was
only examined structurally and not utilized for any
1
2
reactions.
The simplest approach to the phosphine-aminophos-
phine ligation motif involves linking the phosphine with
the aminophosphine through a chiral backbone (Figure
13a,17
of methods including classical resolution,
enzymatic
18
19
resolution, and CBS-catalyzed borane reduction. Dia-
stereoselective lithiation of R- or S-N,N-dimethyl ferro-
1
a). The ferrocenylethyl moiety was chosen as the chiral
13a
cenylethylamine
followed by reaction with chlorodi-
backbone as it has a very well-defined three-dimensional
structure with a significant steric size, which affords an
enhanced possibility for high asymmetric induction (par-
14a
phenylphosphine as described by Kumada and Hayashi
affords phosphine 2 as a 96:4 ratio of diastereomers
(
R,S-2:R,R-2 or the enantiomeric pair). A single crystal-
2
ticularly important for ligands without C -symmetry).
lization of this material affords absolute diastereomeric
and enantiomeric purity, even when starting with enan-
tiomerically impure precursor (e.g., 90% ee).
Also, the 1-ferrocenylethyl species is readily generated
in high enantiomeric purity, and substitution at the R
center is particularly facile and occurs with absolute
retention of configuration,13 greatly simplifying the ligand
Utilizing the unique reactivity of the R-ferrocenylethyl
carbonium ion, the dimethylamino compound 2 was
converted to acetate 3 by reaction with acetic anhydride
at 90 °C for 3.5-6 h (Scheme 1). Although acetate 3 can
be precipitated in over 90% yield by addition of the
reaction solution to a mixture of 2-propanol and water,
the reaction mixture containing 3 was usually added
(4) (a) Cullen, W. R.; Sugi, Y. Tetrahedron Lett. 1978, 19, 1635. (b)
Selke, R.; Pracejus, H. J. Mol. Catal. 1986, 37, 213. (c) Rajanbabu, T.
V.; Ayers, T. A.; Halliday, G. A.; You, K. K.; Calabrese, J. C. J. Org.
Chem. 1997, 62, 6012. (d) Chan, A. S. C.; Hu, W.; Pai, C. C.; Lau, C.-
P.; Jiang, Y.; Mi, A.; Yan, M.; Sun, J.; Lou, R.; Deng, J. J. Am. Chem.
Soc. 1997, 119, 9570.
(
5) Reetz, M. T.; Gosberg, A.; Goddard, R.; Kyung, S.-H. Chem.
Commun. 1998, 2077 and references therein.
6) Sakai, N.; Mano, S.; Nozaki, K.; Takaya, H. J. Am. Chem. Soc.
993, 115, 7033.
7) (a) Krause, H. W.; Kruezfeld, H.-J.; Schmidt, U.; Dobler, Ch.;
(13) (a) Marquarding, D.; Klusacek, H.; Gokel, G.; Hoffmann, P.;
Ugi, I. J. Am. Chem. Soc. 1970, 92, 5389-5393. (b) Gokel, G. W.;
Marquarding, D.; Ugi, I. K. J. Org. Chem. 1972, 37, 3052-3058.
(14) (a) Hayashi, T.; Mise, T.; Fukushima, M.; Kogatani, M.;
Nagashima, N.; Hamada, Y.; Matsumoto, A.; Kawakami, S.; Konishi,
M.; Yamamoto, K.; Kumada, M. Bull. Chem. Soc. Jpn. 1980, 53, 1138-
1151. (b) Yamamoto, K.; Wakatsuki, J.; Sugimoto, R. Bull. Chem. Soc.
Jpn. 1980, 53, 1132-1137.
(15) Togni, A.; Bruetel, C.; Schnyder, A.; Spindler, F.; Landert, H.;
Tijani, A. J. Am. Chem. Soc. 1994, 116, 4062-4066.
(16) For reviews, see: (a) Richards, C. J. Tetrahedron: Asymmetry
1998, 9, 2377-2407. (b) Colcacot, T. J. Chem. Rev. 2003, 103, 3101-
3118.
(
1
(
Michalik, M.; Taudien, S.; Fischer, C. H. Chirality 1996, 8, 173. (b)
Mi, A.; Lou, R.; Jiang, Y.; Deng, J.; Qin, Y.; Fu, F.; Li, Z.; Hu, W.;
Chan, A. S. C. Synlett 1988, 847.
(
8) Reetz, M. T.; Gosberg, A. Tetrahedron: Asymmetry 1999, 10,
2
129.
(
9) Francio, G.; Faraone, F.; Leitner, W. Angew. Chem., Int. Ed.
2
000, 39, 1428.
(
(
10) Reetz, M. T.; Pasto, M. Tetrahedron Lett. 2000, 41, 3315.
11) A preliminary report of this work appeared in: Boaz, N. W.;
Debenham, S. D.; Mackenzie, E. B.; Large, S. E. Org. Lett. 2002, 4,
(17) Gokel, G. W.; Ugi, I. K. J. Chem. Educ. 1972, 49, 294-296.
(18) Boaz, N. W. Tetrahedron Lett. 1989, 30, 2061-2064.
(19) (a) Wright, J.; Frambes, L.; Reeves, P. J. Organomet. Chem.
1994, 476, 215-217. (b) Brieden, W. US Patent No. 5760264, 1998.
2
421-2424.
12) M u¨ nzenberg, R.; Rademacher, P.; Boese, R. J. Mol. Struct. 1998,
44, 77-90.
(
4
J. Org. Chem, Vol. 70, No. 5, 2005 1873

Boaz et al.
directly to the desired amine in aqueous 2-propanol, and
the resulting mixture was heated to 50 °C for 12-18 h
to afford the secondary amine 4. The isolation was
optimized for the reaction with methylamine, in which
case dilution of the reaction mixture with water precipi-
tated the desired amine 4b in 95% overall yield from 2.
Derivatization of the amine with chlorodiphenylphos-
phine in the presence of an acid scavenger proceeded
smoothly when the R group of 4 was hydrogen, methyl,
ethyl, or n-propyl. In addition, the phosphine could be
varied by using commercially available chlorophosphines
TABLE 1. Stability of Ligand 1b and Its Rhodium
Complex in Alcohol Solvents at Room Temperature
half-life (h)
solvent
methanol
ethanol
-propanol
2-methyl-2-propanol
ligand 1b
(COD)Rh-1b
3
8
48
61
nd
nd
nd
2
>1 month
(
e.g., chlorodiethylphosphine, chlorodicyclohexylphos-
phine). Simple removal of the triethylammonium chloride
at the end of the reaction by filtration and a solvent strip
afforded the desired ligand 1 as a brittle foam. The
enantiomeric purity of the final ligand was assumed to
be unchanged from 2 due to the retention of configuration
observed with nucleophilic substitution at the R-ferroce-
nylethyl group and was verified by chiral HPLC for the
cases where R was hydrogen (1a) or methyl (1b).
The nature of the ligands 1 as brittle foams posed
operational difficulties, particularly for larger scale reac-
tions. It was found that ligand 1b can be isolated as a
melt instead of a foam by concentration from a 2-methyl-
2
-propanol solution and then crystallized from hot 2-meth-
yl-2-propanol by cooling to ambient temperature. The
recovery from 2-methyl-2-propanol was moderate (ca.
FIGURE 2. Air stability of R,S-1b.
5
0%), but dilution of the cooled crystallization mixture
with water prior to filtration afforded improved recovery
>90%) of ligand 1b with no observable hydrolysis of
appear to be more stable than anticipated to hydroxylic
solvents. Indeed, many of the best reactions under the
most demanding conditions of low catalyst loadings have
been obtained using methanol as solvent.
(
the N-P bond (presumably due to the insolubility of 1b
in the mixture). This crystallization protocol appears
to be specific for 2-methyl-2-propanol, as 2-propanol
affords poor recovery, 2-propanol/water affords little
solid, and 2-methyl-2-butanol affords 1b as a gum. In
addition, there is limited evidence that the crystallization
enhances the enantiomeric purity of ligand 1b (crystal-
lization has afforded both enantiomers with >99.9% ee).
This crystallization procedure has proven effective for
most ligands 1.
Thus ligands 1 are synthetically advantaged over many
other speciessthere are no exotic reagents, no temper-
ature extremes, no air-sensitive intermediates or prod-
ucts, and only one moisture-sensitive (n-butyllithium)
reaction. Indeed, many of the reactions use water as a
solvent! In addition, both enantiomers of ligand 1 are
equally available.
Air Stability. Many phosphine ligands are not par-
ticularly air-stable, and a large number of ligand inter-
mediates are very air-sensitive. This is due to the very
nature of phosphines, which are prone to oxidation,
particularly if the phosphines possess hydrogen or alkyl
substituents (as opposed to all aryl groups). It was
anticipated that the phosphine-aminophosphines 1 might
show enhanced air stability due to the combination of a
ferrocenyl phosphine (known to be very stable) and an
aminophosphine in which the electronegative nature of
the nitrogen could decrease the electron density of the
phosphorus and reduce its propensity toward oxidation.
The air stability of ligand R,S-1b was examined by
holding this material at ambient temperature open to the
air (in a bottle with the top off) and periodically testing
it as the rhodium complex for asymmetric hydrogenation.
Over more than a 3 year period, within experimental
error no degradation of enantioselectivity or activity for
the hydrogenation of 2-acetamidocinnamic acid (substrate-
to-catalyst ratio 10 000:1 as described below) was ob-
served (Figure 2).
Solvolytic Stability. The solvolytic stability of ligands
in alcoholic solvents (often used in asymmetric hydro-
1
genations) was a concern, as simple aminophosphines are
in general very reactive toward nucleophiles through
heterolytic cleavage of the nitrogen-phosphorus bond.4
c
This stability was investigated by stirring 1b with a large
excess of a variety of alcoholic solvents at ambient
N-Acyl Dehydroamino Acid Asymmetric Hydro-
genation Screening. Asymmetric hydrogenation is one
of the most powerful methods for inducing chirality into
an achiral substrate, as readily available starting materi-
als (olefins, ketones, imines) can be transformed using
an inexpensive reagent (hydrogen) into high-value high-
1
temperature with periodic examination by H NMR. The
results (Table 1) indicate moderate stability in methanol
with increasing stability upon increasing hindrance
around the alcohol. Complexation of 1b to rhodium had
an unexpected stabilizing effect, resulting in an ap-
proximate half-life in methanol of 61 h. This is a
minimum value and may result from oxidative instability
of the complex rather than hydrolytic instability, as no
measures were taken to exclude oxygen from the NMR
tube. Thus, the ligands and particularly the complexes
1
demand products. The area of asymmetric hydrogena-
tion that has received the most attention is the prepa-
ration of amino acid derivatives by the asymmetric
hydrogenation of N-acyl dehydroamino acid species.
N-Acyl dehydroamino acids are excellent substrates for
1874 J. Org. Chem., Vol. 70, No. 5, 2005

Phosphinoferrocenylaminophosphine Ligands for Asymmetric Catalysis
TABLE 2. Asymmetric Hydrogenation of Dehydroamino
synthetically useful carbamate substituents on the amine
a
Acid Derivatives
performed as well as amide groups, a result that is not
always the case for other ligands.22
There are certain ligands that afford high enantio-
selectivities for some aromatic amino acids but not for
alkyl-substituted analogues.5 The results in Table 2,
entries 18-22 indicate that the rhodium complex of R,S-
c
1
b is a truly general catalyst for amino acid preparation,
entry substrate
R
R′
R′′
ee (%)
as in addition to the superior results with the aryl amino
acids it affords high enantioselectivity for a large number
of alkyl-substituted amino acids with a variety of carboxyl
and amino substituents.
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
a
b
c
d
e
f
phenyl
phenyl
phenyl
phenyl
CH3
H
Ac
Ac
99.1
99.4
99.5
98.1
99.0
97.7
98.1
97.2
98.1
98.0
97.7
96.4
97.2
99.3
98.2
98.1
97.4
98.4
96.1
98.8
98.1
98.6
CH3
CH3
CH3
CH3
CH3
CH3
CH3
Boc
Bz
Ac
One requirement for these reactions is a Z configura-
tion of the olefin in substrate 5. Much lower enantiose-
lectivities (and rates) are observed for E olefins, a situa-
p-cyanophenyl
p-nitrophenyl
p-chlorophenyl
p-fluorophenyl
p-methoxyphenyl
m-methoxyphenyl CH3
o-methoxyphenyl
2-furyl
b
Ac
c,d
g
h
i
Ac
23
Ac
tion that has been observed with several other ligands.
c,e,f
b,c
Ac
Fortunately, most preparations of dehydroamino acids
afford the Z isomer selectively, often specifically.
There are a number of factors that can play a role in
asymmetric hydrogenation reactions. The optimal ligand-
to-metal ratio was determined to be equimolar (bidentate
binding), as deviating significantly above or below this
ratio resulted in markedly lower substrate conversion
over a standard time period (although the enantioselec-
tivity remained the same). It was also noted that prehy-
drogenation of the rhodium complex prior to substrate
introduction afforded a much less active catalyst.
Temperature effects were not highly pronounced, as
hydrogenations performed at 0 °C afforded only a slight
increase in enantioselectivity (but often within the error
of the measurement). Limited results at higher temper-
ature (30-35 °C) indicated little effect on enantioselec-
tivity, although significantly higher temperature would
be expected to erode the selectivity.
1
1
1
1
1
1
1
1
1
1
2
2
2
j
Ac
c,d
k
m
n
o
p
q
r
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
H
Ac
Bz
Boc
Ac
e
e
2-furyl
1-naphthyl
1-naphthyl
2-naphthyl
2-naphthyl
H
Boc
Ac
Boc
Ac
Ac
Cbz
Cbz
e
e
g
h
i
s
t
u
v
w
H
H
PhCH2
CH3
CH3
CH2Ph Boc
e
j
cyclopropyl
a
Reactions were run for 1 h at 10 psig hydrogen at ambient
temperature, conversion was >95% except where indicated dif-
b
c
ferently. Reaction run for 30 min. Reaction run with 2 mol %
catalyst. d Reaction run for 2 h. Reaction run for 6 h. Conversion
e
f
g
h
was 92%. Reaction run with 2.5 mol % catalyst. Reaction run
for 24 h. Reaction run in ethyl acetate. j Reaction run with 0.1
mol % catalyst.
i
Solvent can play a crucial role in asymmetric hydro-
genations, particularly as the resting state of the catalyst
is often depicted as a solvated complex. Standard solvents
(THF, toluene, ethyl acetate, acetone, methanol, 2-pro-
panol) afforded nearly identical high enantioselectivity
and high conversion in the reduction of substrate 5a with
the rhodium complex of R,S-1b. Strongly ligating solvents
such as diethoxymethane (potential bidentate donor),
DMF, DMSO, and NMP afforded poor results for this
reaction, likely due to stable nonproductive solvent-
ligated complexes.
asymmetric hydrogenation using rhodium catalysts due
to the bidentate binding of the substrate through the
olefin and the N-acyl carbonyl oxygen.1b Preliminary
results indicated that ligand 1b afforded the highest
enantioselectivities for the asymmetric hydrogenation of
a limited number of dehydroamino acid derivatives.11
Thus, the hydrogenations of a wide variety of N-acyl
dehydroamino acid derivatives were examined with the
rhodium complex of this ligand under screening condi-
tions (0.5 mmol of substrate at 0.1 M concentration, 100:1
substrate/catalyst ratio).
Hydrogen pressure can have a large effect upon the
enantioselectivity of rhodium-catalyzed asymmetric hy-
drogenation reactions. This is perhaps most pronounced
with regards to dehydroamino acid hydrogenations. The
The dehydroamino acid hydrogenation substrates were
generally prepared by either the classical Erlenmeyer
2
0
condensation or a Horner-Wadsworth-Emmons con-
densation of the desired aldehyde with an appropriately
substituted amido acid phosphonate reagent.2
rationale behind these effects is based on the accepted
1
1b,24
mechanism of these particular reactions,
wherein
The results for aromatic amino acids (Table 2, entries
-17) indicate that these species are hydrogenated with
substrate binding (through both the olefin and amide
carbonyl) results in the formation of two diastereomeric
complexes in dynamic equilibrium. In many cases the
minor diastereomer is sufficiently more reactive toward
hydrogen (or perhaps subsequent reductive elimination)
and diastereomer interconversion is sufficiently rapid at
low hydrogen pressures to overwhelm the diasteromeric
ratio and provide the major observed product enantiomer.
1
exceedingly high S enantioselectivity (using R,S-1b)
regardless of the electronic or positional nature of the
substitutents on the aromatic ring or whether the car-
boxyl group was an ester or an acid. Finally, the more
(
20) See, for example: Herbst, R. M.; Shemin, D. In Organic
Synthesis; Blatt, A. H., Ed.; John Wiley: New York, 1943; Collect. Vol.
2
6
; pp 1-3.
21) (a) Schmidt, U.; Lieberknecht, A.; Wild, J. Synthesis 1984, 53-
(
(22) Kreuzfeld, H.-J.; D o¨ bler, C.; Krause, H. W.; Facklam, C.
Tetrahedron: Asymmetry 1993, 4, 2047-2051.
(23) Reference 1a, pp 11-12.
0. (b) Schmidt, U.; Griesser, H.; Leitenberger, V.; Lieberknecht, A.;
Mangold, R.; Meyer, R.; Riedl, B. Synthesis 1992, 487-490. (c) Schmidt,
U.; Lieberknecht, A.; Wild, J. Synthesis 1988, 159-172.
(24) See ref 1b, pp 124-132 and references therein.
J. Org. Chem, Vol. 70, No. 5, 2005 1875

Boaz et al.
TABLE 3. Pressure Effects on
TABLE 4. Catalyst Activity Comparison for the
Asymmetric Hydrogenation of 5aa
Rhodium-R,S-1b-Catalyzed Asymmetric Hydrogenation
of Dehydroamino Acid Derivatives 5a
chelate catalyst turnover
-
1
ee (%) at
ee (%) at
entry
ligand
size
rate (h
)
ee (%)
entry substrate
R
R′′ 10 psig H2 300 psig H2
b
1
2
3
4
5
CHIRAPHOS
5
5
5
6
7
30
440
4800
960
30 500
87
92
97
44
97
b
1
2
3
4
5
6
7
8
a
w
s
g
i
k
o
q
phenyl
cyclopropyl
H
p-chlorophenyl
p-methoxyphenyl Ac
o-methoxyphenyl Ac
1-naphthyl
2-naphthyl
Ac
99.1
98.6
98.5
98.8
98.1
97.7
99.3
98.1
96.8
96.6
95.2
96.2
97.3
96.7
97.4
97.6
DIPAMP
b
b
Boc
Ac
Methyl DuPHOS
c
JOSIPHOS
Ac
R,S-1b
a
Reactions were run at an S/C ratio of 10 000:1 in methanol at
ambient temperature under 40 psig hydrogen except where
b
b
Ac
Ac
b
indicated differently. Reaction run at S/C of 1000:1 to afford
c
appreciable rate. Reaction run at S/C of 5000:1 to afford ap-
a
Reactions were run with 1 mol % catalyst in THF at ambient
preciable rate.
temperature except where indicated differently. b Reaction run in
acetone as solvent.
CHART 1. Ligand Structures
At higher pressures, the enantioselectivity often erodes
due to enhancement of the rate of hydrogen addition to
both diastereomers such that this rate may begin to
compete with their rate of interconversion. In some
extreme cases the enantioselectivity can reverse (albeit
with low ee) at higher hydrogen pressure.25
The hydrogenation of a variety of dehydroamino acid
esters 5 using the rhodium complex of ligand 1b at an
elevated pressure of hydrogen (300 psig) showed little
pressure dependence, affording the corresponding amino
acid derivatives 6 with only slightly lower enantioselec-
tivities (0.5-4.5% ee lower, with the majority only 0.5-
of 1b has an exceedingly high turnover frequency. These
results validate the literature reports that Rh-methyl
DuPHOS (Chart 1) is a highly catalytically active com-
plex; however, Rh-1b is nearly an order of magnitude
faster for this transformation.
Part of the reason for this enhanced activity may lie
in the size of the metallacycle that is generated from the
ligand and the rhodium. Halpern has shown that ligands
that afford a seven-membered rhodium chelate (such as
27
3
% ee lower) compared to low pressure (10 psig) results
(Table 3). Indeed, some of these results (entries 7 and 8)
may be due to a solvent rather than a pressure effect.
Although the decrease may be due to any number of
factors, the limited pressure effect suggests that the
diastereomer interconversion rate (and thus the reaction
rate) using the rhodium complex of the ligand 1b might
be rapid.
1
b) undergo internal rearrangement (facilitating dia-
stereomer interconversion) much more rapidly than
28
smaller chelate rings, resulting in faster replenishment
of the minor diastereomer and thus reaction rate en-
hancement. Limited data suggests that other seven-
membered chelate ligands (e.g., chiral phosphinites) also
show rate enhancements over ligands with smaller
Catalytic Activity in Dehydroamino Acid Asym-
metric Hydrogenations. Although high enantioselec-
tivity is a prerequisite for an effective asymmetric
catalyst, high catalytic activity is also important, and
often overlooked. This activity has a direct effect on the
catalyst contribution cost per mole of product, as the cost
is a combination of the amount of catalyst used (turnover
number) and the reaction time (turnover frequency) of
the expensive precious metal complex, unless the catalyst
can be reused and/or recycled. The limited pressure
effects observed with the rhodium-1b complex suggested
that this catalyst may possess high inherent reactivity.
Thus the kinetics of asymmetric hydrogenation reactions
using the rhodium complex of 1b were investigated and
5c
chelates. The price exacted for this rate enhancement
is often a severe loss of enantioselectivity due to ligand
conformational flexibility. Ligand 1b is an anomaly in
this regard, as high enantioselectivity is achieved concur-
rent with high rates of reaction, perhaps due to confor-
2
mational restriction enforced by the sp -hybridized car-
bons of the ferrocene cyclopentadiene ring.
The high activity and high enantioselectivity observed
for the Rh-R,S-1b asymmetric hydrogenation of 5a was
mirrored in the asymmetric hydrogenations of a variety
of dehydroamino acids with this catalyst (Table 5). In
particular, even the challenging hydrogenation of the
highly hindered N-Boc dehydrocyclopropylalanine benzyl
ester with the rhodium complex of 1b exhibited catalytic
compared to the rhodium complexes of a number of other
ligands, as limited literature data is available.2
2,26
The
catalytic activities (turnover frequencies) were deter-
mined in a sealed pressure vessel by correlating pressure
drop (initial hydrogen pressure 40 psig) with conversion
vs time, utilizing the initial rates wherein the reactions
were zero order in substrate (in practice, up to at least
2
9
activity in excess of 9000 turnovers per hour.
These high catalyst turnover frequencies translate
directly to high turnover numbers, as these rapid rates
allow the use of vanishingly small quantities of catalyst
(
as low as 0.002 mol %) while maintaining short reaction
4
>
0% conversion) and all afforded excellent linearity (R2
0.995).
The results for the hydrogenation of substrate 5a
times, resulting in an advantageously low ligand and
(
27) Armstrong, S. K.; Brown, J. M. Tetrahedron Lett. 1993, 34,
79-882.
28) Landis, C. R.; Halpern, J. J. Organomet. Chem. 1983, 250, 485-
490.
(29) Boaz, N. W.; Debenham, S. D.; Large, S. E.; Moore, M. K.
Tetrahedron: Asymmetry 2003, 14, 3575-3580.
presented in Table 4 indicate that the rhodium complex
8
(
(
25) Nogradi, M. In Stereoselective Synthesis; VCH: Weinheim, 1995;
p 46 and references therein.
26) Oliver, J. D.; Riley, D. P. Organometallics 1983, 2, 1032-1038.
(
1876 J. Org. Chem., Vol. 70, No. 5, 2005

Phosphinoferrocenylaminophosphine Ligands for Asymmetric Catalysis
TABLE 5. Catalyst Activity of Rh-1b with Various
IT t Aa cB o Ln Ea t 7e .s 7Caatalyst Activity of Rh-1 with Various
Dehydroamino Acids 5a
R′ solvent TOF (h-
1 b
)
ee (%) conv (%)
catalyst
entry ligand
R
turnover ee
1
2
3
4
5
R,S-1b
R,S-1b
R,S-1b
R,S-1b
H
H
H
H
H
H
MeOH
EtOH
53 200
96
96
94
94
nd
96.1
95.5
93.8
94.0
18.0
-
1
entry substrate
R
R′
R′′ rate (h
)
(%)
46 200
53 800
77 700
1200
b
1
2
3
4
5
a
g
h
r
phenyl
4-chlorophenyl CH3
4-fluorophenyl CH3
2-naphthyl
cyclopropyl
H
Ac
Ac
Ac
Ac
68 100 98.0
25 300 95.3
40 100 96.8
13 900 98.0
9200 96.4
CH3 MeOH
CH3 EtOH
c
c
R,S-1a Ph
H
MeOH
c
CH3
PhCH2 Boc
a
d
Reactions were run at ambient temperature at 40 psig
hydrogen at an S/C ratio of 2500:1 except where indicated
differently. Turnover frequency. Reaction run at an S/C of
1000:1.
w
a
b
c
Reactions were run in methanol with 40 psig hydrogen at an
S/C ratio of 2500:1 except where indicated differently. All reactions
b
afforded >98% conversion. Reaction run at S/C of 10 000:1.
c
d
Reaction run in 9:1 toluene/methanol. Reaction run at S/C of
1
000:1.
complex of 1b also affords exceedingly high reaction rates
while maintaining high enantioselectivities. In particular,
the reaction of dimethyl itaconate with the rhodium
complex of R,S-1b in ethanol afforded a turnover fre-
quency in excess of 77 000 per hour, the highest rate
observed for a rhodium-1 catalyzed hydrogenation.
Surprisingly, the hydrogenation of phenylitaconic acid
TABLE 6. Asymmetric Hydrogenation of Itacontate
Derivatives
(
7c) with the rhodium complex of 1a had a disappoint-
ingly low reaction rate of about 1200 turnovers per hour
Table 7, entry 5). The literature suggests that the
(
addition of base affords a rate enhancement in the
hydrogenation of itaconic acids.30 The addition of even a
small amount of triethylamine (0.1 equiv) to this reaction
with the preferred rhodium-R,S-1a complex afforded a
large rate enhancement to 11 000 catalyst turnovers per
hour while maintaining exceedingly high enantioselec-
tivity. This effect may be due to the formation of the
distal carboxylate anion, which should be a much stron-
ger donor resulting in a stronger bond to the metal, with
the improved chelation ability accounting for the in-
creased rate. Surprisingly, there was no enhancement of
the rate of itaconic acid (7a) hydrogenation with added
amine.
entry
ligand
substrate
R
R′
ee (%)
conv (%)
1
2
3
R,S-1b
R,S-1b
R,S-1b
R,S-1a
R,S-1a
R,S-1a
a
b
c
a
c
d
H
H
Ph
H
Ph
Ph
H
CH3
H
H
H
CH3
97.6
94.0
90
94.0
99
80
99.8
99.9
99.5
99.8
>99
>95
b
4
5
6
b
a
Reactions run for 6 h at 300 psig hydrogen at ambient
b
temperature. Reaction run in acetone.
metal contribution cost for the transformation. The only
limitation then becomes trace impurities in the substrate
or solvent that may act as catalyst poisons.
Thus, the rhodium complex of ligand R,S-1b is a truly
general catalyst for the hydrogenation of dehydroamino
acid derivatives, affording exceedingly high enantiose-
lectivities independent of the carboxyl, amino, and ter-
minal substituents.
Itaconate Asymmetric Hydrogenations. Itaconic
acid derivatives are structurally and electronically simi-
lar to N-acyl dehydroamino acids, in that the olefin and
the distal carbonyl ligating groups of the itaconate are
positioned in an identical fashion to the olefin and amide
carbonyl of the dehydroamino acids. Thus many catalysts
that work well for dehydroamino acid hydrogenation also
work well for itaconates. The rhodium complexes of
ligands 1 were effective species for the asymmetric
hydrogenation of various itaconate derivatives 7 at 300
psig hydrogen to provide the corresponding 2-substituted
succinates 8, with the R,S enantiomer of the ligands (R,S-
R-Ketoester Asymmetric Hydrogenations. The
asymmetric hydrogenation of R-ketoesters has received
much less attention than the hydrogenation of dehy-
droamino acid derivatives and itaconates, and there are
few ligands that afford high enantioselectivies across this
substrate class.1 This is a different type of reaction
relative to dehydroamino acids or itaconates, since
similar substrate ligation is not geometrically possible.
The asymmetric hydrogenation of R-ketoesters in THF
with the rhodium complexes of ligands 1 was ineffective
at low pressure (10 psig hydrogen), but at high pressure
(300 psig hydrogen) the R enantiomer of the hydroxyester
10 was obtained in 80-90+% ee using Rh-R,S-1 (Table
8). The dicyclohexylphosphino-substituted ligand 1c in
particular afforded enhanced enantioselectivities (in some
cases significantly improved, see entry 5) as compared
to the diphenylphosphino species 1b for these particular
transformations. The high enantioselectivity and R con-
figuration observed with substrate 9e suggest that this
reaction probably is a direct ketone hydrogenation, and
does not proceed through the enol form, as 9e is prohib-
ited from enolization.
1
) invariably affording the R enantiomer of the succinate
8
(Table 6). For unsubstituted itaconates, the rhodium
complex of 1b afforded the best results (Table 6, compare
entries 1 and 4). However, for â-substituted itaconates,
the rhodium complex of 1a afforded decidedly superior
enantioselectivities (compare entries 3 and 5). In all cases
the diacid substrates performed much better than the
diesters.
Unlike the hydrogenation of dehydroamino acid de-
rivatives, the hydrogenation of R-ketoesters showed
remarkable solvent effects (Table 9). An absolute reversal
The results shown in Table 7 indicate that, similar to
the dehydroamino acid derivatives, the hydrogenation of
unsubstituted itaconate derivatives with the rhodium
(
30) Burk, M. J.; Bienewald, F.; Harris, M.; Zanotti-Gerosa, A.
Angew. Chem., Int. Ed. 1998, 37, 1931-1933.
J. Org. Chem, Vol. 70, No. 5, 2005 1877

Boaz et al.
TU As i Bn Lg ER h8 .- 1Aa symmetric Hydrogenation of r-Ketoesters 9
TABLE 10. Catalyst Activity of the Asymmetric
Complexa
R,S-1b
results
R,S-1c
b
results
c
eed
(%)
TOFc eed
TOF
-
1
-1
entry substrate
solvent
(h
)
(h
)
(%)
1
2
3
4
5
a
a
b
d
e
9:1 EtOAc/THF 2100 86.6
MeOH 750 75.4 S 2350 11.2
9:1 EtOAc/THF 4440 91.6
3360 89.6
e
ee (%)
with
ee (%)
with
6220 91.6
4370 90.2
7770 95.8
9:1 EtOAc/THF
9:1 EtOAc/THF
580 80.2
60 74.0
entry substrate
R
R′
CH3
CH2CH3
CH3
CH2CH3
R,S-1b R,S-1c
1
2
3
4
5
a
b
c
d
e
CH3
CH3
86.8
83.4
72.8
85.2
71.7
88.1
90.8
85.4
92.4
97.2
a
Reactions were run at ambient temperature at 40 psig
b
c
hydrogen at an S/C ratio of 1000:1. b Reactions afforded >95%
PhCH2
PhCH2CH2
c
conversion except where indicated. Catalyst turnover frequency
d
e
-C(CH3)2CH2-
per hour. R-Enantiomer major unless otherwise noted. 94.9%
conversion.
a
Reactions were run for 6 h at 300 psig hydrogen and afforded
95% conversion except where noted. Conversion was 50.0%.
b
>
c
Conversion was 83.6%.
slower than in ethyl acetate/THF. Third, although the
hydrogenation rates using ligand 1b exhibited a precipi-
tous decline as the steric congestion of the substrate
increased (see entries 4 and 5), this was not the case for
the ligand 1c. Substrate 9e was particularly intriguing
as, despite the large steric demands of a quaternary
center next to the reacting ketone, it afforded the fastest
overall rate using the rhodium complex of 1c. The
dicyclohexylphosphino ligand 1c is decidedly faster than
the diphenylphosphino ligand 1b for all substrates, and
for substrate 9e was over 2 orders of magnitude faster.
This indicates that the asymmetric hydrogenation of
R-ketoesters with the rhodium complex of ligand 1c is of
practical value, and is a good addition to the hydrogena-
tions of dehydroamino acid derivatives and itaconates for
the rhodium complexes of ligands 1.
TABLE 9. Solvent Effects for the Asymmetric
Hydrogenation of 9a Using Rh-R,S-1ba
entry
1
solvent
THF
ee (%)
conv (%)
87 R
87 R
15 S
56 S
71 S
98.3
99.1
96.9
99.0
81.9
b
2
ethyl acetate
2-propanol
acetone
3
4
5
methanol
a
Reactions were run with 1 mol % catalyst for 6 h at 300 psig
b
hydrogen. Reaction used 0.1 mol % catalyst.
of stereochemistry from R to S was observed for the
hydrogenation of methyl pyruvate with the rhodium
complex of R,S-1b upon changing from THF to methanol.
Similar results were obtained with ethyl pyruvate (9b),
ethyl 4-phenyl-2-oxobutyrate (9c), and dihydro-4,4-di-
methylfuran-2,3-dione (9e) as substrates. Ligand R,S-1c
showed similar but less pronounced solvent dependen-
cies.
Conclusion
The rhodium complexes of phosphine-aminophosphine
ligands 1 are superb species for asymmetric hydrogena-
tion. These species exhibit very high enantioselectivities
and activities for the asymmetric hydrogenation of a wide
variety of dehydroamino acids and itaconates, and high
activities and good to excellent enantioselectivities for the
asymmetric hydrogenation of R-ketoesters. Combining
these reactivity characteristics with outstanding ligand
air-stability and a robust and scalable synthesis affords
species that are advantaged ligands for asymmetric
catalysis. Additional ligands, complexes, and applications
are under development.
This type of significant solvent effect is unusual and
is not seen with other ligands examined (e.g., 9a with
Rh-S-ethyl DuPHOS shows invariant R product in
2
5-44% ee). Potential explanations invoking hydrogen-
bonding reactant geometry change (s-trans vs s-cis) or
isomeric makeup (keto vs enol) are largely discounted by
the analogous solvent effects observed with substrate 9e,
which is constrained both geometrically (s-cis configura-
tion only) and isomerically (no enol form possible). The
origin of this startling solvent effect is yet unknown.
The requirement for higher pressure for the effective
asymmetric hydrogenation of R-ketoesters using the
rhodium complex of 1 suggested that very slow hydro-
genations would be observed under the standard reaction
conditions (40 psig hydrogen). Although the rates are
significantly lower than the dehydroamino acids or the
itaconates, these reactions still usually afforded turnover
Experimental Section
R-1-(S-2-Diphenylphosphino)ferrocenylethyl Acetate
14a
(
R,S-3). Dimethylamino compound R,S-2 (20.0 g; 45.3 mmol)
was combined with acetic anhydride (14.3 mL; 152 mol; 3.35
equiv), and the heterogeneous mixture was heated under
nitrogen to 90 °C during which time it became homogeneous.
The reaction mixture was held at 90 °C for 6 h, at which point
TLC analysis indicated no 2 but much 3 (1:4 EtOAc/heptane,
-1
frequencies above 1000 h , indicative of a relatively fast
reaction (Table 10).
3
Et N deactivated). The mixture was cooled to ambient tem-
perature. A small aliquot was removed, all volatiles were
stripped from it, and the resulting material was analyzed
directly by chiral HPLC to determine enantiomeric excess. The
remainder of the material was poured into 170 mL of a 1:1
There were several other features of note discovered
in this investigation. First, the enantioselectivity depen-
dence on solvent observed during the screening runs was
preserved at these pressures and at low catalyst loadings
(
v:v) mixture of 2-propanol and water in a cool water bath and
(
all reactions were performed at an S/C of 1000:1).
Second, the hydrogenation in methanol was significantly
washed in with a minimum volume of 2-propanol. The result-
ing suspension was stirred vigorously for 1 h, filtered, and
1878 J. Org. Chem., Vol. 70, No. 5, 2005

Phosphinoferrocenylaminophosphine Ligands for Asymmetric Catalysis
washed with water. The yellow solid was dried in vacuo at
g; 0.23 mol) was combined with acetic anhydride (72.8 mL;
ambient temperature under a nitrogen purge to afford 19.08
0.77 mol; 3.35 equiv), and the heterogeneous mixture was
heated under nitrogen to 90 °C during which time it became
homogeneous. The reaction mixture was held at 90 °C for 3.5
h, at which point TLC analysis indicated no 2 but much 3 (1:4
1
g (92%) of R,S-3. Mp: 116-118 °C. H NMR (CDCl
3
) δ: 7.54-
7
.49 (m, 2H); 7.37 (m, 3 H); 7.26-7.15 (m, 5H); 6.210 (q, 1H,
J ) 3.85 Hz); 4.572 (br s, 1H); 4.356 (m, 1H); 4.048 (s, 5H);
.80 (m, 1H); 1.631 (d, 3H, J ) 6.32 Hz); 1.171 (s, 3H). 13
NMR (CDCl ) δ: 169.9 (s); 139.8 (d, JC-P ) 10 Hz); 137.1 (d,
C-P ) 9 Hz); 135.2 (d, JC-P ) 21 Hz); 132.9 (d, JC-P ) 19 Hz);
29.3 (s); 128.3 (d, JC-P ) 7 Hz); 128.1 (s); 128.0 (d, JC-P ) 5
3
C
3
EtOAc/heptane, Et N deactivated). The mixture was cooled to
3
ambient temperature. Aqueous methylamine (359.5 g; 4.63
mol; 20.1 equiv) was dissolved in 2-propanol (463 mL) in a
separate flask and cooled to 20 °C. The mixture containing 3
was added over 25 min to the amine mixture with an attendant
minor exotherm (maximum temperature was 26.0 °C) and
washed in with 116 mL of warm (40-50 °C) 2-propanol. The
mixture was heated to 50 °C overnight (17 h) to afford a
homogeneous solution which had no remaining 3 according to
J
1
Hz); 91.8 (d, JC-P ) 24 Hz); 72.5 (d, JC-P ) 5 Hz); 70.0 (s);
6
1
9.7 (s); 69.3 (d, JC-P ) 3 Hz); 68.5 (d, JC-P ) 10 Hz); 20.2 (s);
8.6 (s). Chiral HPLC (250 × 4.6 mm Chiralpak AD-H, 90:10
hexane/2-propanol, 1 mL/min, λ ) 254 nm): t
R
(R,S-3) 9.15
).
R-1-(S-2-Diphenylphosphino)ferrocenylethylamine
2
4
R D 3
min, t (S,R-3) 10.18 min. [R] ) -307 (c 0.92, CHCl
3
TLC analysis (1:4 EtOAc/heptane, Et N deactivated). The
3
1
(
R,S-4a). Acetate R,S-3 (7.50 g; 16.4 mmol) was combined
with 140 mL of 2-propanol. Ammonium hydroxide (17.8 mL;
11 mmol; 25 equiv) was added, and the reaction mixture was
mixture was cooled to 24 °C, and water (1158 mL) was added.
This resulted in a moderate exotherm (maximum temperature
32 °C) and an orange precipitate. The mixture was stirred at
ambient temperature for 30 min and the solid was filtered,
washed with water, and dried in a nitrogen-purged vacuum
oven at 40 °C for 2 days to afford 93.8 g (95% overall) of R,S-
4
heated to 40 °C for 36 h and then to 45 °C for 12 h to
completely consume 3. After cooling to ambient temperature,
water (35 mL; 0.25 volume based on 2-propanol) was added,
resulting in a yellow precipitate containing various impurities
formed during the reaction. After stirring for 30 min, these
materials were removed by filtration and the filtrate was
concentrated to remove the majority of the 2-propanol. Water
1
4b as an orange solid which was pure by H NMR analysis.
An analytical sample could be obtained by recrystallization
1
from 2-propanol. Mp: 112-113 °C. H NMR (CDCl
3
) δ: 7.55
(m, 2H); 7.37 (m, 3 H); 7.26 (m, 5H); 4.463 (br s, 1H); 4.286
(m, 1H); 4.028 (s, 5H); 3.94 (m, 1H); 3.78 (m, 1H); 1.943 (s,
(
100 mL) was added resulting in a yellow solid. This was
1
3
stirred for 30 min, and the solid was collected by filtration,
washed with water, and dried in vacuo at ambient temperature
3H); 1.445 (d, 3H, J ) 6.59 Hz). C NMR (CDCl
3
) δ: 140.0 (d,
J
C-P ) 10 Hz); 137.1 (d, JC-P ) 9 Hz); 134.9 (d, JC-P ) 21 Hz);
under a nitrogen purge. This resulted in 5.61 g (83%) of R,S-
132.9 (d, JC-P ) 19 Hz); 129.2 (s); 128.6 (s); 128.5 (d, JC-P ) 6
Hz); 128.3 (d, JC-P ) 7 Hz); 97.5 (d, JC-P ) 24 Hz); 75.5 (d,
1
4
7
4
1
a. Mp: 130-131 °C. H NMR (CDCl
3
) δ: 7.56-7.51 (m, 2H);
.39 (m, 3 H); 7.26 (m, 5H); 4.452 (br s, 1H); 4.284 (m, 1H);
.248 (m, 1H); 4.022 (s, 5H); 3.772 (br s, 1H); 1.631 (br s, 2H);
JC-P ) 7 Hz); 71.3 (d, JC-P ) 4 Hz); 69.8 (s); 69.5 (d, JC-P ) 4
31
Hz); 69.1 (s); 52.5 (d, JC-P ) 10 Hz); 32.8 (s); 18.8 (s). P NMR
1
3
+
24
.460 (d, 3H, J ) 6.59 Hz). C NMR (CDCl
9 Hz); 137.3 (d, JC-P ) 9 Hz); 135.0 (d, JC-P ) 21 Hz); 132.9
d, JC-P ) 18 Hz); 129.2 (s); 128.5 (s); 128.4 (s); 128.2 (d, JC-P
7 Hz); 100.6 (d, JC-P ) 23 Hz); 71.4 (s); 69.7 (s); 69.2 (s);
3
) δ: 140.1 (d, JC-P
(acetone-d
6
) δ: -26.0 (s). FDMS: m/z 427 (M ). [R]
D
) -309
)
(c 1.55, ethanol). Anal. Calcd for C25H26FeNP: C, 70.27; H,
6.13; N, 3.28. Found: C, 70.64; H, 6.37; N, 3.16.
(
)
N-Diphenylphosphino N-Methyl R-1-(S-2-Diphenyl-
phosphino)-ferrocenylethylamine (R,S-1b). Amine R,S-
6
8.4 (br s); 45.4 (d, JC-P ) 9 Hz); 23.0 (s). HRMS m/z: calcd
+
24
for C24
H
25FeNP (M + H ) 414.1074, found 414.1074. [R]
).
N-Diphenylphosphino-R-1-(S-2-diphenylphosphino)-
D
)
4
b (50.0 g; 0.117 mol) was dissolved in ethyl acetate (235 mL),
-
325 (c 1.01, CHCl
3
and triethylamine (32.6 mL; 0.234 mol; 2.0 equiv) was added.
The mixture was cooled in an ice-water bath and purged with
argon for 15 min. Chlorodiphenylphosphine (23.1 mL; 0.129
mol; 1.1 equiv) was added over 15 min such that the temper-
ature remained below 10 °C. The reaction mixture was stirred
at 0 °C for 1 h, and the temperature was then raised to 20 °C
in 5° hourly increments. The cooling bath was then turned off,
and the reaction mixture was stirred at ambient temperature
3
ferrocenylethylamine (R,S-1a). Amine R,S-4a (10.0 g; 24.2
mmol) was dissolved in ethyl acetate (48 mL), and triethyl-
amine (6.75 mL; 48.4 mol; 2.0 equiv) was added. The mixture
was cooled in an ice-water bath and purged with argon for 5
min. Chlorodiphenylphosphine (4.56 mL; 25.4 mmol; 1.05
equiv) was added dropwise such that the temperature re-
mained below 15 °C. The reaction mixture was allowed to
warm to ambient temperature overnight at which point TLC
overnight at which point TLC (1:2 EtOAc/heptane, Et N
1
deactivated) and H NMR analyses indicated very little 4b
(
1:1 EtOAc/heptane, Et
3
N deactivated) indicated no 4a. The
(<3%). The reaction mixture was diluted with heptane (235
mL) to afford a precipitate (triethylamine hydrochloride). The
mixture was stirred for 30 min and filtered through Celite and
eluted with 1:1 ethyl acetate/heptane. The filtrate was con-
centrated to a small volume. The residue was dissolved in ethyl
acetate (75 mL), diluted with heptane (50 mL), and filtered
through a fine frit, and the precipitate was washed with 1:1
ethyl acetate/heptane until the wash liquid was colorless. The
combined filtrate and wash liquors were concentrated to small
volume, diluted with 2-methyl-2-propanol (50 mL), and con-
centrated once more. The residue was diluted with 2-methyl-
2-propanol (900 mL) and heated to reflux to afford a homoge-
neous solution. The mixture was allowed to cool to ambient
temperature overnight to afford a yellow precipitate (precipita-
tion started at around 40 °C). Water (900 mL) was added,
resulting in a minor exotherm (from 24 to 28 °C) and the
formation of significantly more solid. The mixture was stirred
at ambient temperature for 1 h, then filtered and washed with
water. The yellow solid was dried under vacuum at ambient
temperature with a nitrogen purge to afford 66.25 g of R,S-
reaction mixture was diluted with heptane (48 mL) to afford
a precipitate. The mixture was stirred for 30 min and filtered
through Celite and eluted with 1:1 ethyl acetate/heptane. The
filtrate was concentrated, resulting in 14.09 g (97%) of R,S-
1
a as a yellow-orange foam. Mp: 54-55 °C, with 99.6% ee by
1
chiral HPLC. H NMR (CDCl
3
) δ: 7.56-7.51 (m, 2H); 7.36 (m,
H); 4.497 (q, 1H, J ) 6.32 Hz); 4.462 (m, 1H); 4.293 (m, 1H);
.912 (s, 5H); 3.83 (m, 1H); 2.26 (m, 1H); 1.515 (d, 3H, J )
4
3
6
1
3
3
.59 Hz). C NMR (CDCl ) δ: 143.6 (d, JC-P ) 14 Hz); 142.7
(
d, JC-P ) 11 Hz); 140.4 (d, JC-P ) 10 Hz); 138.1 (d, JC-P ) 10
Hz); 135.4 (d, JC-P ) 22 Hz); 132.7 (d, JC-P ) 18 Hz); 131.4 (d,
J
C-P ) 11 Hz); 131.1 (d, JC-P ) 11 Hz); 128.3 (s); 128.2 (s);
1
28.1 (d, JC-P ) 2 Hz); 128.1 (s); 128.0 (s); 100.1 (dd, JC-P
)
1
0, 25 Hz); 74.6 (d, JC-P ) 10 Hz); 71.6 (d, JC-P ) 4 Hz); 69.9
(
s); 69.8 (s); 69.3 (s); 50.9 (dd, JC-P ) 7, 25 Hz); 24.1 (d, JC-P
3
1
)
11 Hz). P NMR (CDCl
3
) δ 33.1 (s); -23.7 (s). HRMS m/z:
+
calcd for C36
H
34FeNP (M+H ) 598.1516, found 598.1505.
2
Chiral HPLC (250 × 4.6 mm Chiralpak AD-H, 99:1 hexane/
2
-propanol, 1 mL/min, λ ) 254 nm): t (R,S-1a) 10.0 min, t
R
+
R
24
(
S,R-1a) 10.8 min. FDMS: m/z 597 (M ). [R]
CHCl ).
N-Methyl
ethylamine (R,S-4b). Dimethylamino compound R,S-2 (99.9
D
) -279 (c 1.01,
1b (93%). Mp: 99-101 °C, with 99.96% ee by chiral HPLC.
1
3
3
H NMR (CDCl ) δ: 7.65 (m, 2H); 7.4-7.0 (m, 14H); 6.82 (m,
R-1-(S-2-Diphenylphosphino)ferrocenyl-
4H); 5.006 (m, 1H); 4.502 (br s, 1H); 4.40 (m, 1H); 4.15 (m,
1H); 3.798 (s, 5H); 2.148 (d, 3H, J ) 3.30 Hz); 1.471 (d, 3H, J
J. Org. Chem, Vol. 70, No. 5, 2005 1879

Boaz et al.
)
6.87 Hz). 13C NMR (CDCl
d, JC-P ) 24 Hz); 140.0 (d, JC-P ) 10 Hz); 139.5 (d, JC-P ) 7
Hz); 135.9 (d, JC-P ) 22 Hz); 134.1 (d, JC-P ) 22 Hz); 132.3 (d,
C-P ) 17 Hz); 131.7 (d, JC-P ) 18 Hz); 129.3 (s); 128.8 (s);
28.1-127.2; 97.9 (dd, JC-P ) 14, 28 Hz); 75.7 (d, JC-P ) 14
Hz); 72.1 (d, JC-P ) 5 Hz); 70.5 (s); 69.9 (s); 69.8 (s); 58.4 (dd,
3
) δ: 142.4 (d, JC-P ) 9 Hz); 140.3
methanol) indicates that the S enantiomer was obtained from
3
2
(
hydrogenation using ligand R,S-1.
High-Pressure Asymmetric Hydrogenation Screening
Example: N-Acetyl L-Phenylalanine Methyl Ester (S-6a).
Bis(1,5-cyclooctadiene)rhodium trifluoromethanesulfonate (2.3
mg; 5 µmol; 0.01 equiv) and ligand (R,S)-1b (3.7 mg; 6 µmol;
0.012 equiv) were placed in a high-pressure reaction vessel
which was sealed and purged with argon. Anhydrous THF was
degassed with argon for 15 min, then 2.0 mL was added to
the reaction vessel via syringe. This solution was stirred at
J
1
J
C-P ) 9, 39 Hz); 30.6 (d, JC-P ) 11 Hz); 18.5 (d, JC-P ) 6 Hz).
31
P NMR (CD
.7 Hz). Chiral HPLC (250 × 4.6 mm Chiralpak AD-H, 99:1
(S,R-1b) 10.4
(R,S-1b) 11.6 min. HRMS m/z: calcd for C37 35FeNP
) -257 (c 0.97,
2 2
Cl ) δ: 58.8 (d, JP-P ) 7.7 Hz); -25.3 (d, JP-P )
7
hexane/2-propanol, 1 mL/min, λ ) 254 nm): t
R
25 °C under argon for 15 min. A solution of methyl 2-acet-
min, t
R
H
2
+
24
amidocinnamate (5a, 110 mg; 0.50 mmol) in degassed THF (2
mL) was then added to the catalyst solution via syringe. The
substrate was washed in with 1 mL of degassed THF. The
solution was then purged five times with argon and pressur-
ized with hydrogen to 300 psig for 6 h, then depressurized and
purged with argon to afford 100% conversion to N-acetyl
L-phenylalanine methyl ester (S-6a) with 97.6% ee as deter-
mined by chiral GC analysis.
(
M ) 611.15942, found 611.16429. [R]
D
toluene).
N-Dicyclohexylphosphino N-Methyl R-1-(S-2-Diphen-
ylphosphino)ferrocenylethylamine (R,S-1c). Amine R,S-
4
b (2.0 g; 4.7 mmol) was dissolved in toluene (10 mL), and
triethylamine (1.30 mL; 9.4mol; 2.0 equiv) was added. The
mixture was cooled in an ice-water bath and purged with
argon for 15 min. Chlorodicyclohexylphosphine (1.05 mL; 4.94
mmol; 1.05 equiv) was added, and the reaction mixture was
allowed to warm to ambient temperature overnight at which
Turnover Fequency Determination Example: N-
Acetyl L-Phenylalanine Methyl Ester (S-6a). Methyl
2-acetamidocinnamate (5a, 2.19 g; 10.0 mmol) was added to a
3
point TLC (1:4 EtOAc/heptane, Et N deactivated) indicated
Fisher-Porter bottle. Argon-degassed methanol (12.3 mL) was
added to afford a homogeneous solution which was purged with
argon for 5 min. The bottle was fitted with a pressure head
and evacuated and filled with helium 10 times. Bis(1,5-
cyclooctadiene)rhodium trifluoromethanesulfonate (2.3 mg;
0.005 mmol) and the ligand (R,S)-1b (3.7 mg; 0.006 mmol; 1.2
equiv based on rhodium) were added to a 25-mL Schlenk tube
and purged with argon for 5 min. Degassed methanol (5.0 mL)
was added and the mixture was stirred for 15 min to afford a
homogeneous solution. 1.0 mL of this solution (0.001 mmol;
0.0001 equiv; S:C 10,000:1) was added via a gastight syringe
to the Fisher-Porter bottle containing the substrate solution.
The bottle was evacuated and filled with helium 10 times, and
then evacuated and filled with hydrogen 5 times. The bottle
was pressurized with 40 psig hydrogen, sealed, and stirred
vigorously at ambient temperature and the pressure drop was
followed with a pressure sensor. After 1 h, the reaction mixture
was evacuated and filled with helium five times and then
depressurized. The reaction mixture was sampled and ana-
lyzed to indicate 100% conversion to N-acetyl L-phenylalanine
methyl ester (S-6a) with 97% ee as determined by chiral GC
analysis. Solvent removal afforded 2.23 g (99%) of S-6a.
Graphical analysis of the initial 60% of the reaction and
adjusting for the 10 000:1 substrate-to-catalyst ratio indicated
a catalyst turnover frequency of 30 500 turnovers per hour.
partial conversion to 1c. The reaction mixture was diluted with
heptane (10 mL) to afford a precipitate, filtered, and eluted
with heptane. The filtrate was were concentrated to a small
volume. The crude product was filtered through a pad of
neutral alumina and eluted with 5:10:85 triethylamine/ethyl
acetate/heptane to afford 1.50 g (52%) of R,S-1c. Mp: 68-69
1
°
C. H NMR (DMSO-d
6
) δ: 7.61 (m, 2H); 7.42 (m, 3H); 7.19
(
m, 3H); 7.04 (m, 2H); 4.532 (br s, 1H); 4.442 (m, 1H); 4.34 (m,
H); 4.112 (br s, 1H); 3.712 (s, 5H); 2.143 (d, 3H, J ) 1.92
1
1
3
Hz); 1.549 (d, 3H, J ) 6.87 Hz); 1.7-0.7 (m, 22 H). C NMR
DMSO-d ) δ: 141.8 (d, JC-P ) 9 Hz); 139.4 (d, JC-P ) 9 Hz);
35.3 (d, JC-P ) 23 Hz); 131.5 (d, JC-P ) 17 Hz); 129.2 (s);
(
6
1
1
28.0 (d, JC-P ) 9 Hz); 127.7 (d, JC-P ) 5 Hz); 127.1 (s); 98.9
(
dd, JC-P ) 9, 30 Hz); 73.7 (d, JC-P ) 13 Hz); 70.9 (d, JC-P
)
5
3
9
2
Hz); 69.7 (s); 69.4 (s); 69.3 (s); 56.9 (dd, JC-P ) 9, 36 Hz);
6.1 (d, JC-P ) 15 Hz); 34.6 (d, JC-P ) 18 Hz); 32.2 (d, JC-P
)
Hz); 30.0 (d, JC-P ) 12 Hz); 29.6-29.2 (m); 26.8-25.8 (m);
3
1
0.4 (d, JC-P ) 8 Hz). P NMR (DMSO-d
6
) δ: 67.2 (s); -25.6
48FeNP
) -320 (c 0.54,
+
(
(
s). FDMS: m/z 624 (M ). HRMS m/z: calcd for C37
H
2
+
25
M + H ) 624.2611, found 624.2653. [R]
).
D
CHCl
3
Low-Pressure Asymmetric Hydrogenation Screening
Example: N-Acetyl L-Phenylalanine Methyl Ester (S-6a)
Using Ligand R,S-1b. Bis(1,5-cyclooctadiene)rhodium tri-
fluoromethanesulfonate (2.3 mg; 5 µmol; 0.01 equiv) was placed
into a reaction vessel and purged with argon for 15 min. A
solution of ligand (R,S)-1b (3.7 mg; 6 µmol; 0.012 equiv) in
THF (2.0 mL) was degassed with argon for 15 min and then
added via cannula to the bis(1,5-cyclooctadiene)rhodium tri-
fluoromethanesulfonate. Methyl 2-acetamidocinnamate (5a;
Acknowledgment. We are grateful for the skillful
assistance of Dr. Michael Meadows for NMR analysis
and Mr. James Little for high-resolution mass spectral
analysis.
Supporting Information Available: General experimen-
tal details, additional hydrogenation procedures, analytical
data on hydrogenation products, and NMR spectra for com-
pounds 1a-c, 3, and 4a,b. This material is available free of
charge via the Internet at http://pubs.acs.org.
1
10 mg; 0.50 mmol) in THF (2.0 mL) was degassed with argon
for 20 min and then added to the catalyst solution via cannula.
The solution was then flushed with hydrogen and pressurized
to 10 psig hydrogen for 6 h to afford 100% conversion to
N-acetyl L-phenylalanine methyl ester (S-6a) with 99.1% ee
1
as determined by chiral GC analysis. H NMR (CD
3
OD) δ:
JO048312Y
7
.28-7.16 (m, 5H); 4.64-4.61 (m, 1H); 3.65 (s, 3H); 3.13-3.08
dd, 1H J ) 5.5, 13.9 Hz); 2.94-2.88 (dd, 1H, J ) 8.9, 13.9
Hz); 1.87 (s, 3H). Chiral GC (Chirasil L-valine [Varian] 25 m
(
(31) Hayashi, T.; Hayashi, C.; Uozumi, Y. Tetrahedron: Asymmetry
1
995, 6, 2503-2506.
(
32) (a) Burk, M. J.; Feaster, J. E.; Nugent, W. A.; Harlow, R. L. J.
×
0.25 mm i.d., film thickness 0.12 µm, 160 °C, 9 min, 160-
85 °C 70 °C/min, 185 °C 5 min, 15 psig He): t (R-6a) 7.77
) +15.4 (c 2.1,
Am. Chem. Soc. 1993, 115, 10125. (b) Vineyard, B. D.; Knowles, W.
S.; Sabacky, M. J.; Bachman, G. L.; Winkauff, D. J. J. Am. Chem. Soc.
1977, 99, 5946.
1
R
24
min, t
R
(S-6a) 8.29 min, t
R
(5a) 13.24 min. [R]
D
1880 J. Org. Chem., Vol. 70, No. 5, 2005