Development of a commercial process for (S)-β-phenylalanine (1)

Article abstract of DOI:10.1021/op200084g

The development of a commercial manufacturing route for (S)-β-phenylalanine 8, a key pharmaceutical building block, is described. The different approaches which were investigated, based on catalytic asymmetric hydrogenation of enamide intermediates and on biocatalysis using acylase and lipase hydrolyses, are compared. The lipase resolution route was chosen for scale-up, and the final two-step process, based on readily available raw materials, is shown to be robust at full manufacturing scale

Full text of DOI:10.1021/op200084g

LECTURE TRANSCRIPT
pubs.acs.org/OPRD
Development of a Commercial Process for (S)-β-Phenylalanine1
J. Ian Grayson,* J€urgen Roos, and Steffen Osswald
Process Research & Development ꢀ Exclusive Synthesis, Evonik Degussa GmbH, Rodenbacher Chaussee 4,
63457 Hanau-Wolfgang, Germany
ABSTRACT: The development of a commercial manufacturing route for (S)-β-phenylalanine 8, a key pharmaceutical building
block, is described. The different approaches which were investigated, based on catalytic asymmetric hydrogenation of enamide
intermediates and on biocatalysis using acylase and lipase hydrolyses, are compared. The lipase resolution route was chosen for scale-
up, and the final two-step process, based on readily available raw materials, is shown to be robust at full manufacturing scale
’ INTRODUCTION
derivatives, in that the starting material is formed as a mixture
of E- and Z-enamides, which hydrogenate at different rates, and
with different enantioselectivities. Our first work in this area was
done in collaboration with B€orner’s group at Rostock, where we
were able to show that high enantioselectivities were obtainable
with E:Z mixtures of enamides, using a cationic Rh(I)-DuPHOS
catalyst, simply by reducing the hydrogenation pressure from
30ꢀ40 to 1 bar (1 mol % catalyst loading, methanol, 25 °C).9
These initial encouraging results led us to conduct further work
with the B€orner group on the design of a series of tunable bis-
phospholane ligands, where the bite angle and the electron
density of the ligand could be varied.10 The first of these ligands,
originally called MalPHOS (9), was later marketed as catASium
M.11 Hydrogenation of both the E- and Z-enamides using
catalysts based on ligand 9 gave high enantioselectivity even
with bulky substituents on the enamide, such as phenyl and
isopropyl (1 mol % catalyst, methanol, 1 bar H2, 25 °C). Further
improvements in enantioselectivity were made using a series of
diphosphine ligands based on camphor, where the phosphine
groups could be tuned independently.12 High enantioselectivity
was obtained with both E- and Z-enamides, for example using the
ligand 10 bearing both phenyl- and 3,5-xylyl-substituted phos-
phine groups (Scheme 2) (1 mol % catalyst, methanol or
dichloromethane, 8 bar H2, 25 °C).
At the same time, alternative asymmetric catalytic routes to 8
were examined. As β-keto-esters are readily available, their
oximes were tested as substrates for asymmetric hydrogenation.
A screening program was run, using ligands from a wide range of
families, and high conversion and enantioselectivity were found
using the ferrocenyl ligand Josiphos-PPF-P-tBu2. Using 1 mol %
of this ligand, and a hydrogen pressure of 20 bar, conversion
to the hydroxylamine ether 11 was achieved in 91% yield and
with 94% ee.13 A second hydrogenation step, this time with a
palladium/carbon catalyst, produced the β-phenylalanine ester
12 without any loss of enantioselectivity (Scheme 3).
The β-amino acid motif is becoming more common as a
component of new pharmaceutical products, one of the reasons
being the high stability of the β-amino acid peptide bond towards
protease hydrolysis.2 In addition, β-amino acid subunits are
found in a variety of natural products. For example, the (S)-β-
phenylalanine residue is found in pyloricidin A (1) and moir-
amide A (2).3 In addition, the (R)-enantiomers of aromatic
β-amino acids are present in cyclic peptides such as the astins,
chondramides, and geodiamolides.3 More recently, a number of
pharmaceuticals containing a β-amino acid unit have been
developed, such as otamixaban 3 (Sanofi-Aventis), and sitagliptin
5 (Merck), or where a β-amino acid is used as an intermediate,
e.g. maraviroc 4 (Pfizer) (Figure 1).
The appearance of these products in development, together
with a number of other target molecules containing an aromatic
β-amino acid subunit, such as 6 and 7,4 was the starting point for
our work on an enantioselective synthesis of this class of mole-
cule, which could be implemented at a commercial scale. This
lecture transcript summarises the work carried out at Evonik
(previously Degussa), together with our academic collaborators,
which led finally to a robust production process for (S)-3-amino-
3-phenylpropanoic acid [(S)-β-phenylalanine] (8). A develop-
ment process for the manufacture of 8 has been published as part
of a synthesis of maraviroc,5 but this involved a classical resolu-
tion step, and was not suitable for development to full manu-
facturing scale.
A variety of synthetic routes have been developed to prepare
β-amino acids, and this type of structure has often been used
to test the efficiency of new catalytic asymmetric syntheses.3,6
A summary of selected routes to 8 is shown in Scheme 1. They
can be divided up into classical chemical methods, such as the
ArndtꢀEistert chain extension,7 diastereoselective processes, e.g.
using a chiral amine base,8 catalytic asymmetric hydrogenation
reactions, and enzymatic methods. Our efforts focused on the
latter two routes, and are discussed in more detail below.
In the last 10 years, many improvements have been made in
ligand design, which have been instrumental in increasing the
’ CATALYTIC ASYMMETRIC HYDROGENATION
APPROACHES
Special Issue: Asymmetric Synthesis on Large Scale 2011
The asymmetric hydrogenation of 3-aminoacrylic acid deri-
vatives differs from that of the isomeric 2-aminoacrylic acid
Received: March 31, 2011
Published: June 13, 2011
r
2011 American Chemical Society
1201
dx.doi.org/10.1021/op200084g Org. Process Res. Dev. 2011, 15, 1201–1206
|

Organic Process Research & Development
LECTURE TRANSCRIPT
Figure 1. Natural products and pharmaceuticals based on β-amino acids.
Scheme 1. Summary of selected routes to (S)-β-phenylalanine
selectivity of catalytic asymmetric hydrogenation. A benchmark
for the preparation of β-amino acid derivatives is set by the
P-chiral ligands developed by the group of Xumu Zhang. This
series of ligands, Tangphos, Duanphos, and Zhangphos, devel-
oped over a period of 8 years, can now deliver enantioselectivities
of >95% on both the E- and Z-isomers of N-acyl enamides.14
However, the long development time for a commercial catalytic
asymmetric synthesis, together with the need to manufacture
large quantities of a complex ligand for commercial manufacture
of a β-amino acid, meant that we instigated a simultaneous
development program for biocatalytic approaches to (S)-β-
phenylalanine 8.
’ BIOCATALYTIC APPROACHES
Although a large number of different enzymes and substrates
have been used for the synthesis of β-amino acids,15 practical
routes to products such as 8 are based on readily accessible
substrates and hydrolase enzymes. One of the earliest methods to
be developed was the use of penicillin G acylase from E. coli,
1202
dx.doi.org/10.1021/op200084g |Org. Process Res. Dev. 2011, 15, 1201–1206

Organic Process Research & Development
LECTURE TRANSCRIPT
Scheme 2. Hydrogenation of enamides using tunable bis-
phospholane ligands
Scheme 4. Hydrolysis using penicillin G acylase16
Scheme 5. Acylase route to aromatic β-amino acids
Scheme 3. β-Amino acids by reduction of oxime ethers
This route had the advantage of providing rapid access to a
range of aromatic and heteroaromatic β-amino acids. However, it
suffered from long reaction times, and the lack of a recombinant
form of the porcine kidney acylase enzyme. Enzymes of animal
origin are becoming less acceptable in the synthesis of pharma-
ceuticals, and in this case, no microbial alternative was available.
We therefore turned our attention to the lipase route to 8, where
there was also a literature precedent. The enzymatic resolution of
the ethyl ester has been achieved using the commercially avail-
able Amano PS lipase.19 The enantioselectivity was given as 99%
but was also stated to be very dependent on pH. When we
repeated this work, we found that small changes in the operating
procedure had major effects on the enantioselectivity of the
isolated product. The pure acid 8 did indeed crystallise out with
99% ee at low concentrations, but the enantioselectivity in the
bulk solution was only 85%. At higher concentrations, the (R)-
enantiomer also crystallised from the reaction solution in variable
amounts. We were pleased to find that changing from the ethyl
ester to the propyl or butyl esters resulted in both a faster and a
more enantioselective reaction (Table 1).20
which was applied by Soloshonok et al. to the preparation of a
range of aromatic β-amino acids.16 The penicillin G acylase
hydrolysis gives rise to the (R)-isomer of the β-amino acid: the
(S)-isomer must be obtained by separation of the unreacted
amide and subsequent chemical hydrolysis (Scheme 4). This,
coupled with the high mass of the phenylacetyl protecting
group and the need to separate the product from phenylacetic
acid after the reaction, makes this process unsuitable for the
manufacture of 8.
Acylase enzymes were already tested for activity against N-acyl
β-amino acids by Whitesides in 1989,17 but contrary to the
results found with R-amino acids, no activity was found with
either porcine kidney acylase I or Aspergillus acylase. Despite this
unpromising literature precedent, we screened a range of hydro-
lase enzymes against the N-acetyl and N-chloroacetyl derivatives
of β-phenylalanine. We were pleased to find that porcine kidney
acylase I had significant catalytic activity on the N-chloroacetyl
β-amino acid 13, although there was no activity on the corre-
sponding N-acetyl β-amino acid.18 On a preparative scale, a
conversion of 49.5% was obtained after 48 h, with an enantios-
electivity of >98%. Ion chromatography produced 8 in an
isolated yield of 36% (Scheme 5).
The propyl ester 14 was chosen for further development.
Further optimisation work demonstrated that the original purely
aqueous reaction, with insoluble ester and acid phases, was not
suitable for commercial operation. However, addition of suffi-
cient acetone to bring the ester 14 into solution led to a marked
reduction in the reaction rate. A two-phase system of tert-butyl
1203
dx.doi.org/10.1021/op200084g |Org. Process Res. Dev. 2011, 15, 1201–1206

Organic Process Research & Development
LECTURE TRANSCRIPT
methyl ether and water, on the other hand, gave an operating
system which had a number of advantages for scale-up (Figure 2).
The racemic ester can be synthesised as a solution in MTBE and
used directly without isolation. The (R)-ester remains in the
organic phase and can be separated easily from the (S)-acid,
which crystallises completely from the two-phase liquid mixture,
and which has low solubility in MTBE. The lipase remains in the
aqueous phase and can be separated from the product 8, and
recycled if required. The isolated (S)-β-Phenylalanine is of
sufficient quality for downstream processing without any addi-
tional purification.
Together with the Landfester group at the University of Ulm,
we investigated the possibility of performing the lipase hydrolysis
in a mini-emulsion, instead of in a two-phase aqueousꢀorganic
system.21 This allows for a more concentrated and homogeneous
operation, or for a shorter reaction time or reduced enzyme
charge (Table 2). The use of a surfactant and a hydrophobic
additive allows the generation of nanodroplets with a defined size
of 100ꢀ150 nm, using ultrasonic activation. Good conversion
and isolated yields were found at reaction concentrations up to
605 g/L. However, there were some disadvantages to this
process, which had to be resolved prior to scale-up. A substantial
amount of acetone was required at the end of the reaction to
break the emulsion, which negated the effect of the higher
reaction concentration in terms of space-time yield. The reagents
(hexadecane and lutensol) required to form the mini-emulsion
have to be washed out completely from the isolated amino acid
after filtration. In addition, the isolated yield was lower compared
with that from biphasic conditions. As a result, the mini-emulsion
method remains a laboratory process at present.
found with sterically hindered aromatics and with the furyl group
(Table 3).22 However, high enantioselectivities were found in all
cases when the propyl ester of the racemic acid was used as
substrate. Application of this hydrolysis to aliphatic substrates
proved more difficult. The β-amino acid analogue of tert-leucine,
β3-neopentylglycine 15, is a useful intermediate for pharmaceu-
ticals requiring a β-amino acid with a bulky side chain. Enzyme
screening showed that the best results were obtained with
a Candida Antarctica lipase, but the ee was a moderate 40%
(Scheme 6).22 In this case, however, an alternative classical
diastereomeric salt resolution gave ready access to this building
block in high yield.23
’ SCALE-UP OF THE LIPASE ROUTE TO (S)-Β-PHENYL-
ALANINE
The advantage of the lipase route to 8 is that there are only two
isolated stages, the racemic β-amino acid 16 and the (S)-isomer
8. Both stages produce crystalline products, which do not require
any additional purification in normal operation (Scheme 7).
The first stage, the Rodionov reaction,24 is a general single-
step process to racemic aromatic β-amino acids. The yield of this
reaction is around 50% at all process scales (kilo-lab, 3.3 kg of
benzaldehyde, 45ꢀ52% yield; pilot plant, 40 kg of benzaldehyde,
47.5% yield; commercial plant, 340 kg of benzaldehyde, 50ꢀ52%
yield). During our laboratory development work, we varied the
reaction time and temperature, the reagent ratios, the rate of
water quench, the agitation speed, and the cooling profile. None
of the changes applied had a significant effect in improving the
yield of 16 above the average of 50%. Pfizer workers reported
sublimation of ammonium acetate on the condenser in their
laboratory work, and changed to ammonium formate as the
ammonia source for scale-up (yield 44%).5 We encountered no
problems with the use of ammonium acetate in the Rodionov
The scope of the lipase hydrolysis was studied with a range of
aromatic and heteroaromatic propyl esters. Low activities were
Table 1. Effect of the alkyl group on the lipase hydrolysis to 8
Table 2. Lipase hydrolysis in two-phase and mini-emulsion
systems21
operating
method
substrate
(g/L)
reaction conversion
isolated
ee
conversion reaction time
time (h)
(%)
yield (%) (%)
R
(%)
(h)
ee of (S)-acid (%)
two-phase
242
242
484
605
15
8
50
49
45
42
41
38
37
35
>99.4
>99.4
>99.4
>99.4
CH2ꢀCH3
37.8
48.7
45.2
6
3
3
85.1
96.4
96.8
mini-emulsion
mini-emulsion
mini-emulsion
CH2ꢀCH2ꢀCH3
CH2ꢀCH2ꢀCH2ꢀCH3
17
24
Figure 2. Schematic for the operation of the lipase hydrolysis.
1204
dx.doi.org/10.1021/op200084g |Org. Process Res. Dev. 2011, 15, 1201–1206

Organic Process Research & Development
LECTURE TRANSCRIPT
Table 3. Scope of the lipase hydrolysis
Scheme 7. Summary of the commercial route to (S)-β-
phenylalanine
aryl group
yield (S) ꢀ acid (%)
ee (S) ꢀ acid (%)
4-cyanophenyl
4-methylphenyl
4-bromophenyl
4-chlorophenyl
4-fluorophenyl
3-bromophenyl
3-chlorophenyl
2-chlorophenyl
3,4-dichlorophenyl
2-furyl
27
50
46
45
46
34
34
20
34
16
99
98
98
98
99
>99
98
99
99
99
and recycle of the aqueous phase containing the lipase was also
attempted, because the lipase is a major contributor towards the
process cost. However, the lipase is partly inactivated during the
15-h reaction time, and gave only a 10ꢀ11% yield of 8 on recycle.
However, a mixture of 75% recovered lipase solution, supple-
mented with a 50% charge of fresh lipase gave 8 in 41% isolated
yield. The total enzyme charge of 125% of the original amount to
a recycle batch takes account of the partial inactivation of the
recycled lipase. It was possible to recycle the aqueous liquors 5
times in this way, giving yields of 8 in the range 41ꢀ43%, with
<0.5% (R)-isomer by HPLC. The total lipase requirement per
batch was thus 58% of a single-batch charge.
Scheme 6. Lipase route to (S)-β3-neopentylglycine
’ CONCLUSION
A number of laboratory routes to (S)-β-phenylalanine 8 have
been developed. Of the metal-catalysed and biocatalytic routes
studied, the lipase hydrolysis of the racemic propyl ester 14 was
chosen for scale-up. This process is robust and has been operated
at all scales from kilo-lab to full commercial plant. A rapid and
efficient route to 8 and to other aromatic β-amino acids from
simple starting materials has thus been achieved.
reaction at any of the scales operated. In their work on the
mechanism of the Rodionov reaction, Tan and Weaver25 de-
scribed two possible mechanistic routes for the reaction and
showed that the irreversible formation of cinnamic acid as a
byproduct is a termination step in the process. The cinnamic acid
is lost to the mother liquors and was not quantified by us.
The esterification stage operates very smoothly (1.1 equiv of
thionyl chloride, n-propanol, 3 h, 50 °C), to give the racemic
propyl ester 14, which is not isolated. The excess n-propanol is
removed by distillation, and the product is partitioned between
water (at pH 3) and toluene. The aqueous layer is then adjusted to
pH 8 and the free ester extracted into MTBE. This solution is
mixed with an aqueous solution of the lipase (Amano lipase PS),
and the two-phase mixture is stirred at 35 °C and at pH 8 until
completion. The yield and quality of 8 was again unaffected by the
reaction scale (kilo-lab, 3.7ꢀ7.4 kg of 16, 38ꢀ41% yield; pilot
plant, 100 kg of 16, 43% yield; commercial plant, 430 kgof 16, 42%
yield). The reported yield of 8 is based on the racemic ester 14
charged and is relative to a maximum theoretical yield for the (S)-
enantiomer of 50%. The chemical purity of the isolated (S)-β-
phenylalanine was 98.5ꢀ99.5% in each case, and the optical purity
>99.5% (usually the (R)-isomer was not detectable).
’ EXPERIMENTAL SECTION
All raw materials, reagents and solvents were purchased
from commercial suppliers and used without further purifica-
tion. All reactions were performed under an atmosphere of
nitrogen. Lipase PS from Burkholderia cepacia was purchased
from Amano, Japan.
Racemic 3-Amino-3-phenylpropanoic Acid: 16. Ethanol
(2550 L) was charged to a 8000-L glass-lined reactor, together
with ammonium acetate (617.5 kg, 8.01 kmol) and malonic acid
(499.8 kg, 4.80 kmol). Benzaldehyde (340 kg, 3.20 kmol) was
added, and the reaction mixture was heated to reflux (∼80 °C).
The reaction was held at 78ꢀ82 °C for 6 h, during which time
carbon dioxide was evolved, and a clear solution was formed,
followed by crystallisation of the amino acid. The reaction was
cooled to 35ꢀ40 °C, and water (1530 L) was added over ∼1 h.
The suspension was further cooled to ∼5 °C and centrifuged
in portions, washing the crystals with water (total 350 L) and
ethanol (total 350 L). The product was dried at 50 °C in vacuo to
a residual solvent level of <0.3%, to give the dry amino acid
The organic layer containing the unreacted (R)-ester could be
recovered and hydrolysed to give (R)-β-phenylalanine. Recovery
1205
dx.doi.org/10.1021/op200084g |Org. Process Res. Dev. 2011, 15, 1201–1206

Organic Process Research & Development
LECTURE TRANSCRIPT
16 (260 kg, 49.1%). Assay (HPLC area %) >98%, cinnamic
acid <1.0%.
C.; Merkle, K. M.; Lynch, R.; Lynch, J. J.; Rodan, S. B.; Duggan, M. E.
Bioorg. Med. Chem. Lett. 2002, 12, 2463.
(5) Haycock-Lewandowski, S. J.; Wilder, A; Åhman, J. Org. Process
Res. Dev. 2008, 12, 1094.
(S)-3-Amino-3-phenylpropanoicAcid:8. The racemic amino
acid 16 (430 kg, 2.60 kmol) and n-propanol (1935 L, 25.6 kmol)
were charged to a 4000-L glass-lined reactor. Thionyl chloride
(341 kg, 2.87 kmol) was added over 1ꢀ2 h, at a reaction
temperature of <40 °C, while the off-gases (SO2 and HCl) were
absorbed in an NaOH scrubber. Following the addition, the
reaction was held for 3 h at 47ꢀ50 °C for complete formation
of the propyl ester. Unreacted thionyl chloride and n-propanol
(∼1300 L) were removed by distillation (45 °C, 50 mbar), and
then toluene (850 L) was added and the distillation repeated to
ensure complete removal of n-propanol. Water (1500 L) was
added, and the propyl ester was extracted into the aqueous phase.
The aqueous layer was adjusted to pH 3ꢀ4 with 25% NaOH
solution, and any residual organic solvent was removed by
distillation (45 °C, 50 mbar). The cooled aqueous solution was
stirred with tert-butyl methyl ether (MTBE) (600 L) and the pH
adjusted to 8.0ꢀ8.2 with 25% NaOH (∼300 L). The layers
were separated, and the aqueous layer was extracted with MTBE
(600 L). Amano lipase PS (21 kg) was dissolved in water (800 L)
and filtered into a 4000-L glass-lined reactor. The filter was washed
with water (400 L). The MTBE extracts were added to the lipase
solution and stirred for 15 h at 28ꢀ30 °C, while maintaining a pH
of 8.1ꢀ8.3 by the addition of 25% NaOH. During this period, the
β-amino acid 8 crystallised from the reaction. The product was
centrifuged in portions and washed with acetone (total 2 ꢁ 200
L). The wet product was dried in vacuo to give the pure (S)-β-
amino acid 8 (178 kg, 41.3%). Assay (titration) 99.0%, assay
(HPLC, area %) 99.5%, (R)-enantiomer (chiral phase HPLC)
<0.3%.
(6) Liu, M.; Sibi, M. P. Tetrahedron 2002, 58, 7991.
(7) Seebach, D.; Overhand, M.; K€uhnle, F. N. M.; Martinoni, B.;
Oberer, L.; Hommel, U.; Widmer, H. Helv. Chim. Acta 1996, 79, 913.
(8) Davies, S. G.; Ichihara, O. Tetrahedron: Asymmetry 1991, 2, 183.
(9) (a) Heller, D.; Holz, J.; Drexler, H.-J.; Lang, J.; Drauz, K.;
Krimmer, H.-P.; B€orner, A. J. Org. Chem. 2001, 66, 6816. (b) Heller,
D.; Holz, J.; Komarov, I.; Drexler, H.-J.; You, J.; Drauz, K.; B€orner, A.
Tetrahedron: Asymmetry 2002, 13, 2735.
(10) Holz, J.; Monsees, A.; Jiao, H.; You, J.; Komarov, I. V.; Fischer,
C.; Drauz, K.; B€orner, A. J. Org. Chem. 2003, 68, 1701.
(11) catASium is a registered trade mark of Evonik, licensed to
Solvias AG, Switzerland.
(12) Kadyrov, R.; Iladinov, I. Z.; Almena, J.; Monsees, A.; Riermeier,
T. H. Tetrahedron Lett. 2005, 46, 7397.
(13) Kadyrov, R.; Riermeier, T.; Almena, J.; Monsees, A.; Gross, P.;
Rossen, K. Eur. Pat. Appl. EP 1862446, 2007.
(14) Zhang, X.; Huang, K.; Hou, G.; Cao, B.; Zhang, X. Angew.
Chem., Int. Ed. 2010, 49, 6421.
(15) Liljeblad, A.; Kanerva, L. T. Tetrahedron 2006, 62, 5831.
(16) Soloshonok, V. A.; Fokina, N. A.; Rybakova, A. V.; Shishkina,
I. P.; Galushko, S. V.; Sorochinsky, A. E.; Kukhar, V. P.; Savchenko,
ꢀ
M. V.; Svedas, Y. K. Tetrahedron: Asymmetry 1995, 6, 1601.
(17) Chenault, H. K.; Dahmer, J.; Whitesides, G. M. J. Am. Chem.
Soc. 1989, 111, 6354.
(18) Gr€oger, H.; Trauthwein, H.; Buchholz, S.; Drauz, K.; Sacherer,
C.; Godfrin, S.; Werner, H. Org. Biomol. Chem. 2004, 2, 1977.
(19) Faulconbridge, S. J.; Holt, K. E.; Sevillano, L. G.; Lock, C. J.;
Tiffin, P. D.; Tremayne, N.; Winter, S. Tetrahedron Lett. 2000, 41, 2679.
(20) (a) Gr€oger, H.; Werner, H. Eur. Pat. Appl. EP 1367129, 2003.
(b) Gr€oger, H.; Werner, H. Eur. Pat. Appl. EP 1367279, 2003.
(21) Gr€oger, H; May, O.; H€usken, H.; Georgeon, S.; Drauz, K.;
Landfester, K. Angew. Chem., Int. Ed. 2006, 45, 1645.
’ AUTHOR INFORMATION
(22) H€usken, H.; Gr€oger, H. Unpublished results.
Corresponding Author
*Fax: (þ49)6181592604. E-mail: ian.grayson@evonik.com.
(23) Ter Wiel, M. K. J.; Arnold, M.; Peter, S.; Troltsch, I.; Merget, S.;
Glaser, F.; Schwarm, M.; Bhatti, H. S.; Kuriakose, B.; Pol, S. S.;
Balamurugan, M.; Joshi, V. V. Tetrahedron: Asymmetry 2009, 20, 478.
(24) Rodionov, W. M.; Malewinskaja, E. T. Ber. Dtsch. Chem. Ges.
1926, 59, 2952.
’ ACKNOWLEDGMENT
(25) Tan, C. Y. K.; Weaver, D. F. Tetrahedron 2002, 58, 7449.
We thank our co-workers from Evonik (Degussa) and our
academic collaborators for their contributions to the work
described here, as noted in the references. We particularly thank
Thomas Riermeier and Renat Kadyrov for their work on the
catalytic asymmetric hydrogenation routes, Harald Gr€oger and
Jens W€oltinger for their work on the biocatalytic processes, and
Franz-Rudolf Kunz for the analytical determinations.
’ REFERENCES
(1) This work was presented by J€urgen Roos at the Organic Process
Research and Development Conference, Clearwater, Florida, United
States, 28 Februaryꢀ2 March, 2011.
(2) Seebach, D.; Beck, A. K.; Bierbaum, D. J. Chem. Biodiversity 2004,
1, 1111.
(3) Juaristi, E., Soloshonok, V., Eds. Enantioselective Synthesis of
β-Amino Acids; Wiley: Hoboken, NJ, 2005; pp 19ꢀ75.
(4) (a) Coleman, P. J.; Brashear, K. M.; Hunt, C. A.; Hoffman, W. F.;
Hutchinson, J. H.; Breslin, M. J.; McVean, C. A.; Askew, B. C.; Hartman,
G. D.; Rodan, S. B.; Rodan, G. A.; Leu, C.-T.; Pruekasritanont, T.;
Fernandez-Metzler, C.; Ma, B.; Libby, L. A.; Merkle, K. M.; Stump, G. L.;
Wallace, A. A. Bioorg. Med. Chem. Lett. 2002, 12, 31. (b) Coleman, P. J.;
Askew, B. C.; Hutchinson, J. H.; Whitman, D. B.; Perkins, J. J.; Hartman,
G. D.; Rodan, G. A.; Leu, C.-T.; Pruekasritanont, T.; Fernandez-Metzler,
1206
dx.doi.org/10.1021/op200084g |Org. Process Res. Dev. 2011, 15, 1201–1206