An efficient and stereodivergent synthesis of threo- and erythro-β-methylphenylalanine. Resolution of each racemic pair by semipreparative HPLC

Article abstract of DOI:10.1016/j.tet.2003.11.044

threo and erythro diastereoisomers of the constrained amino acid (βMe)Phe can be obtained separately on a multigram scale through a three-step synthesis from the corresponding Z and E isomers of 2-phenyl-4(α-phenylethylidene)-5(4H)-oxazolone. The 5(4H)-ox

Full text of DOI:10.1016/j.tet.2003.11.044

Tetrahedron 60 (2004) 885–891
An efficient and stereodivergent synthesis of threo- and
erythro-b-methylphenylalanine. Resolution of each racemic pair
by semipreparative HPLC
*
´ ´ ´
Miriam Alıas, Marıa Pilar Lopez and Carlos Cativiela
´ ´ ´
Departamento de Quımica Organica, Instituto de Ciencia de Materiales de Aragon, Universidad de Zaragoza-CSIC, 50009 Zaragoza, Spain
Received 24 June 2003; revised 26 September 2003; accepted 18 November 2003
Abstract—threo and erythro diastereoisomers of the constrained amino acid (bMe)Phe can be obtained separately on a multigram scale
through a three-step synthesis from the corresponding Z and E isomers of 2-phenyl-4(a-phenylethylidene)-5(4H)-oxazolone. The 5(4H)-
oxazolones are readily available from acetophenone and hippuric acid. The four enantiomerically pure isomers of b-methylphenylalanine,
(2R,3R)-(bMe)Phe, (2S,3S)-(bMe)Phe, (2R,3S)-(bMe)Phe and (2S,3R)-(bMe)Phe, have been prepared by HPLC resolution of the racemic
precursors methyl threo (or erythro)-2-benzamide-3-phenylbutanoates.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
of threonine as the starting material,12 enzymatic resolution
in conjunction with HPLC,8,10 asymmetric synthesis,13,14
the chiral auxiliary approach15–19 and enantioselective
hydrogenation.20,21 When all stereoisomers are required, it
may be more convenient to perform a rapid synthetic route
resulting in racemates rather than a stereoselective synthesis
for each isomer.
The incorporation of conformationally constrained a-amino
acids into peptides allows the study of structure-activity
relationships and the synthesis of peptide analogues with
improved pharmacological properties.1–4
Special mention must be made of the constrained analogues
of phenylalanine, as this naturally occurring a-amino acid is
directly involved in a large number of molecular recognition
processes.5–7 In all cases, the three-dimensional arrange-
ment of the side chain moiety of the phenylalanine residue is
crucial in eliciting the desired response.
Access to the b-methylphenylalanine amino acid in its
erythro and threo diastereomerically pure forms on a
multigram scale is a subject of current interest. Repeated
recrystallization of the hydrochloride form of the b-methyl-
phenylalanine mixture22 is very time consuming and
especially difficult for the threo racemate, which requires
further recrystallizations and even the use of semiprepara-
tive RP-HPLC.10,23,24
The side-chain conformation can be conformationally
constrained by introducing an alkyl group at the b-position
ofana-aminoacidresidue withoutsignificantlyperturbingthe
backbone conformation. In particular, b-methyl-a-aromatic
amino acids have been incorporated into peptides5,6,8–10
and confer on these systems a conformational side-chain
rigidity that is very valuable for the study of both the
specific topochemical arrays of the side chains and
topochemical nature of the binding site.
In the course of our work on the synthesis of conformation-
ally restricted aspartame analogues, the availability of a
simple, short and efficient method to obtain threo- and
erythro-(bMe)Phe as diastereomerically pure materials was
critical. We describe here a convenient route for the
multigram-scale preparation of both diastereomers by
catalytic hydrogenation of a,b-didehydroamino acid deriva-
tives. Z- and E-2-Benzamide-3-phenyl-2-butenoates,
obtained from Z- and E-2-phenyl-4(a-phenylethylidene)-
5(4H)-oxazolones, are the precursors of threo- and erythro-
(bMe)Phe, respectively.
The preparation of all four isomers of b-methylphenyl-
alanine in enantiopure form has been a challenging area of
synthetic organic chemistry. Several strategies have been
developed and these include classical resolution,11 the use
Extensive work on enantiomeric separations of b-methyl
amino acids using chiral derivatisations25,26 and different
chiral stationary phases,27–31 mainly derived from macro-
cyclic glycopeptide antibiotics, has been reported. To the
Keywords: b-Methylphenylalanine; Constrained phenylalanines; 5(4H)-
Oxazolone; Chiral stationary phase; HPLC resolution.
*
Corresponding author. Tel./fax: þ34-976-761-210;
e-mail address: cativiela@unizar.es
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.11.044

´
M. Alıas et al. / Tetrahedron 60 (2004) 885–891
886
best of our knowledge, separation of all four stereoisomers
of (bMe)Phe on HPLC has never been described on a
semipreparative scale, so we would like to report here the
preparation of the four isomers of b-methylphenylalanine in
optically pure form by combining a racemic synthesis with
an HPLC resolution procedure.
treatment afforded the corresponding benzamido esters Z-2
and E-2 in high yield and with retention of configuration, as
reported previously.37
The next step in the synthetic route involved hydrogenation
of the tetrasubstituted olefin in compounds Z-2 and E-2.
Initial attempts to obtain racemic methyl butanoates threo-3
and erythro-3 involved the use of 10% Pd/C as catalyst. The
hydrogenation of Z-2 was achieved by keeping the reaction
at 30 8C with vigorous agitation in the presence of a small
amount of Pd/C (10% w/w). The diastereomeric purity of
threo-3 was confirmed by 1H NMR spectroscopy and
HPLC. When the preparation of erythro-3 was attempted
under the same conditions, a mixture ranging from 95:5 to
87:13 of erythro-3:threo-3 was obtained.
2. Results and discussion
2.1. Racemic synthesis
The Z and E isomers of 2-phenyl-4(a-phenylethylidene)-
5(4H)-oxazolone (Z-1, E-1) proved to be convenient starting
materials to afford (bMe)Phe, as demonstrated previously
by our group.32–34 The sequence of reactions leading to the
amino acid involved the synthesis of pure 5(4H)-oxazo-
lones, alkaline hydrolysis to 2-benzoyl-3-phenyl-2-butenoic
acids followed by catalytic hydrogenation and, finally,
hydrolysis to afford the amino acid. The methodology
described here involves some modifications with respect to
the originally reported procedure and these are described
below. The synthetic route followed for the racemic
synthesis of the precursors for the desired a-amino acids
is outlined in Scheme 1.
It is well known that the factors influencing the hydrogen-
ation of olefins are numerous and include, among others, the
nature of the solvent, hydrogen availability and catalyst.38
These facts prompted us to modify the reaction conditions in
an attempt to minimize the isomerization phenomenon.
The use of methanol/benzene mixtures as solvent had been
reported to show marked inhibition of isomerization.
Nevertheless, this approach did not lead to better results
in our experiments. Other catalysts, such as Pt/C and
PtO2/C, did not improve on our previous results. Finally,
upon increasing the amount of catalyst up to 40% catalyst,
on a weight basis referred to the substrate, we managed to
overcome the isomerization problem. The diastereomeric
The thermodynamically more stable (Z)-isomer, Z-1, was
prepared by fractional crystallisation of an isomeric mixture
(Z/E) obtained from the condensation of hippuric acid and
acetophenone.32 In the same paper, access to the less stable
(E)-isomer E-1 was effected from Z-1 by treatment with
HBr. In our experience, working under similar reaction
conditions, isomerization was sometimes completely
E-selective and sometimes moderately or even poorly
E-selective. The mechanism of this reaction is apparently
the same as that in the isomerization of cis-/trans-stilbene35
and it seems to be consistent with the presence of bromine
radicals.36 We proceeded to define new reaction conditions
that favour a radical reaction. After several other approaches
proved unsuccessful, we were able to obtain the (E)-isomer
in a totally controlled and reproducible manner by HBr
isomerization in the presence of a catalytic amount of
benzoyl peroxide.
1
purity of erythro-3 was checked by H NMR spectroscopy
and HPLC. In fact, it appears that we had indirectly
increased the concentration of hydrogen on the catalyst
surface. For this reason, we expected that the use of
hydrogen pressures of 5–10 atm would lead to the same
results, but under these conditions fragmentations occurred.
The last step of the synthetic approach involved removal of
the protecting groups in threo-3 and erythro-3 by total
hydrolysis. This step has been reported previously34 and
involves very strong conditions: heating under reflux in a
mixture of hydrobromic acid and glacial acetic acid. This
method led to the formation of 15% of erythro-4 in the
reaction mixture on starting from threo-3. Our subsequent
aim was to reduce as much as possible the partial
epimerization observed at Ca and, in this respect, milder
deprotection conditions were assayed. Finally, the use of a
mixture of 2.5 N HCl/HOAc (4:1) gave almost quantitative
deprotection with no detectable epimerization by 1H NMR.
Nevertheless, diastereomeric purity was assessed (after
Ester derivatives are more convenient substrates for
chromatography than the corresponding carboxylic acids.
Thus, with the aim of resolving racemic intermediates into
their enantiomers by chiral HPLC, oxazolone ring opening
was performed by methanolysis of Z-1 and E-1 in the
presence of a catalytic amount of sodium methoxide. This
Scheme 1. Synthetic route to racemic threo-(bMe)Phe and erythro-(bMe)Phe. Reagents and conditions: (a) MeONa/MeOH. (b) H2/Pd–C. (c) 2.5 N
HCl/HOAc, reflux.

´
M. Alıas et al. / Tetrahedron 60 (2004) 885–891
887
Table 1. Selected chromatographic data for the HPLC resolution of threo-3
and erythro-3 on the amylose-derived chiral stationary phase
2.2. HPLC resolution
Once an efficient racemic route to the target compounds had
been developed, we undertook the preparation of the
products in enantiomerically pure form by HPLC resolution
of a synthetic intermediate using a chiral stationary phase.
The utility of polysaccharide-based phases is well docu-
mented.39,40 Recently, a non-commercial stationary phase
consisting of a mixed derivative of amylose (10-undeceno-
ate/3,5-dimethylphenylcarbamate) covalently linked to
allylsilica gel proved to be very efficient in the semi-
preparative separation of racemic constrained analogues of
phenylalanine.41–44 The wide applicability shown by this
stationary phase together with its chemical stability make it
very convenient for the resolution of racemates on a
semipreparative scale.
Compound
Eluenta
(A/B/C)
Flow
(mL/min)
l (nm)
k10
a
Rs
threo-3
95:5:0
93:5:2
93:5:2
1b
210
235
270
1.59
1.24
2.44
1.21
1.15
1.18
0.90
0.59
0.81
1b
18c
erythro-3
98:2:0
96:2:2
96:2:2
1b
210
240
265
2.79
2.98
2.87
1.33
1.23
1.19
1.67
1.25
0.80
1b
18d
For the definition of k0, a and Rs see Section 4.2.
A: n-hexane, B: 2-propanol, C: chloroform.
a
b
Steel column, 150 mm£4.6 mm ID. Temperature: 25 8C. c¼5 mg/mL.
Steel column, 150 mm£20 mm ID. Temperature: 25 8C. c¼100 mg/mL.
Steel column, 150 mm£20 mm ID. Temperature: 25 8C. c¼200 mg/mL.
c
d
First, analytical separation of derivatives threo-3 and
erythro-3 was examined using mixtures of n-hexane/
2-propanol as the mobile phase. As can be seen from the
results in Table 1, chloroform leads to lower selectivity and
resolution factors in the analytical assays but its presence is
necessary in order to optimize the loading capacity of the
column. Thus, mixtures of n-hexane/2-propanol/chloroform
were also tested as eluents (Fig. 1).
previous transformation to the free amino acids) using a RP-
HPLC protocol analogue to the method described in the
literature.24 It was found that diastereomeric purity for threo
pair was over 99% whereas for erythro pair ranged 96–98%
for a wide number of experiments done.
The route described above allows access to diastereo-
merically pure threo-4 and erythro-4 through a three-step
approach with yields of 82–86% from oxazolones and an
overall yield of 31–33% from hippuric acid. These results
make this methodology particularly appropriate for large-
scale preparations of diastereomerically pure erythro-
(bMe)Phe and threo-(bMe)Phe.
Preparative resolutions of racemates threo-3 and erythro-3
were performed on a 150£20 mm ID column by successive
injections using the peak shaving technique. After determi-
nation of the optimum cut points, each run was collected
into three separate fractions. The first fractions and the last
Figure 1. HPLC analytical resolution of threo-3 (eluent: n-hexane/2-propanol 95:5) and erythro-3 (eluent: n-hexane/2-propanol 98:2). See Table 1 for related
chromatographic data.

´
M. Alıas et al. / Tetrahedron 60 (2004) 885–891
888
fractions, each containing one of the enantiomers, were
collected at each passage through the column and combined.
Therefore, 327 mg of threo-3 dissolved in 3.3 mL of
dichloromethane were fractionated to obtain 78 and 67 mg
of the less and more strongly retained enantiomers,
respectively. A total of 4.4 h were required (33 injections
of 100 mL each with 8 min/cycle). Since the preparative
column contained about 28 g of chiral stationary phase
(CSP), a loading capacity of 2.7 mg racemate per gram CSP/
hour was allowed in the threo-3 resolution. As regards
racemate erythro-3, 760 mg dissolved in 3.8 mL of
dichloromethane was fractionated to afford 279 and
200 mg of the less and more strongly retained enantiomers,
respectively. Six hours were required for total separation
(38 injections of 100 mL each with 9.5 min/cycle) and the
loading capacity was 4.5 mg racemate/g CSP per hour.
Enantiomeric purity of each isomer was checked on the
analytical column and only one enantiomer was detected.
stereochemistry to the compounds obtained from the first
eluted enantiomer of threo-3 and a (2R,3S) configuration to
the derivatives obtained from the last eluted enantiomer of
threo-3. In a similar way, a (2R,3R) stereochemistry was
assigned to the compound obtained from the first eluted
enantiomer of erythro-3 and a (2S,3S) configuration to the
derivatives obtained from the last eluted enantiomer of
erythro-3.
3. Conclusion
The methodology described here allows easy access to the
racemic pairs (erythro and threo) of b-methylphenylalanine
with high diastereomeric purity and through high yield
transformations. Moreover, combining the racemic syn-
thesis with a resolution by chiral HPLC provides all four
individual isomers of b-methylphenylalanine, (2S,3S)-,
(2R,3R)-, (2S,3R)- and (2R,3S)-b-methylphenylalanine,
with high optical purity.
2.3. Preparation of enantiomerically pure compounds
Once enantiomerically pure precursors were available, we
undertook the deprotection steps under the conditions
developed for the racemic material, threo-3 and erythro-3,
with each stereoisomer (Scheme 2). In addition, with the
aim of assigning the absolute configurations, the amino acid
hydrochlorides obtained after hydrolysis were treated with
propylene oxide under reflux in order to obtain the optically
pure free amino acids.
4. Experimental
4.1. General
Solvents were purified according to standard procedures.
Melting points were determined on a Bu¨chi SMP-20
apparatus and were not corrected. IR spectra were registered
on a Mattson Genesis FTIR spectrophotometer; nmax is
given for the main absorption bands. H and 13C NMR
1
2.4. Assignment of absolute configurations
spectra were recorded on a Varian Unity-300 or a Bruker
ARX-300 instrument at room temperature, unless otherwise
indicated, using the residual solvent signal as the internal
standard; chemical shifts are reported in ppm on the d scale,
coupling constants in Hz. Mass Spectra were obtained on a
The absolute configurations of all four isomers of
b-methylphenylalanine were determined by direct compari-
son of their optical rotations with the values described in the
literature.10,11,45 This allowed us to assign a (2S,3R)
Scheme 2. Synthetic route to enantiomerically pure (bMe)Phe stereoisomers. Reagents and conditions: (a) (i) 2.5 N HCl/HOAc, reflux; (ii) propylene
oxide/EtOH, reflux.

´
M. Alıas et al. / Tetrahedron 60 (2004) 885–891
889
high resolution VG-autospectrometer using either EI or
þFAB techniques. Optical rotations were measured on a
Perkin–Elmer 241 polarimeter-C in a 1 dm cell of 1 mL
capacity at 25 8C. Microanalyses were carried out on a
Perkin–Elmer 200 C, H, N, S analyser. Analytical TLC was
performed using Merck 60 SI F254 precoated silica gel
polyester plates and the products were examined by UV
fluorescence or developed using iodine vapour. Column
chromatography was performed using silica gel 60
(230–400 mesh).
amount of dry toluene. After adding 10 mg of benzoyl
peroxide, the solution was saturated with anydrous hydro-
gen bromide for 15–30 min or until a precipitate had
formed completely, while keeping the reaction in an ice-
bath. The solid was filtered off and washed with cold toluene
to afford the pure (E)-isomer as a pale yellow solid (2.37 g,
94% yield). Mp 115 8C. Rf (hexanes/benzene 1:1)¼0.19. IR
2.64 (s, 3H); 7.24–7.60 (m, 8H); 8.09 (m, 2H). 13C NMR
(CDCl3 75 MHz) d 23.02; 125.91; 127.80; 128.09; 128.14;
128.41; 128.80; 129.63; 129.92; 132.68; 137.25; 152.22;
160.83; 164.06. MS-EI (m/z, %) 263 [(M)þ, 93]; 105 (82);
77 [(C6H5)þ, 100]. Anal. calcd for C17H13NO2: C, 77.55; H,
4.98; N, 5.32. Found: C, 77.33; H, 5.03; N, 5.30.
1
(nujol) 1788; 1651 cm21. H NMR (CDCl3 300 MHz) d
4.2. High performance liquid chromatography
HPLC was carried out using a Waters HPLC system
equipped with a Waters 600-E pump and a Waters 991
photodiode array detector. The chiral stationary phase,
which consisted of a mixed 10-undecenoate/3,5-dimethyl-
phenylcarbamate of amylose bonded to allylsilica, was
prepared according to a previously described procedure.46,47
4.5. Methyl (Z)-2-benzamide-3-phenyl-2-butenoate, Z-2
A solution of sodium methoxide (50 mg) in absolute MeOH
(60 mL) was added to Z-1 (2.5 g, 10 mmol) and the reaction
mixture was stirred at room temperature for 30 min (check
completion by TLC, hexanes/benzene 1:1). The reaction
mixture was filtered and the solvent was removed in vacuo.
Cold water was added dropwise to the residue until
precipitation was complete. The product was collected by
vacuum filtration to give Z-2 as a white solid (2.67 g, 95%).
Mp 153 8C. Rf (hexanes/AcOEt 8:2)¼0.11. IR (nujol) 3263;
The analytical assays were carried out on
a
150 mm£4.6 mm ID column and the semipreparative
resolution was achieved on a 150 mm£20 mm ID column.
All analytical assays and semipreparative chromatography
were performed under the conditions given in Table 1.
Diastereomeric purity of the final products was checked on a
Xterrae MS C8 250 mm£4.6 mm ID column. The solvents
used as mobile phases were of spectral grade.
1
1713; 1638 cm21. H NMR (CDCl3 300 MHz) d 2.31 (s,
3H); 3.86 (s, 3H); 7.19 (brs, 1H); 7.10–7.60 (m, 11H). 13C
NMR (CDCl3 75 MHz) d 20.52; 52.24; 123.79; 127.09;
127.42; 128.30; 128.61; 128.95; 131.92; 133.10; 137.01;
139.70; 165.40; 165.59. MS-EI (m/z, %) 295 [(M)þ, 8]; 263
[(M2MeOH)þ, 16]; 190 (37); 130 (8); 105 (100); 77
[(C6H5)þ, 73]. Anal. calcd for: C18H17NO3: C, 73.20; H,
5.80; N, 4.74. Found: C, 73.01; H, 5.89; N, 4.71.
The capacity (k0), selectivity (a) and resolution (Rs) values
were calculated according to the equations kR0 ¼(tR2t0)/t0,
a¼k20 /k10 , Rs¼1.18 (t22t1)/(w2þw1). Subscripts 1 and 2 refer
to the first and second eluted diastereoisomers, respectively;
tR (R¼1, 2) are their retention times, and w2 and w1 denote
their bandwidths at half height; t0 is the dead time.
4.3. (Z)-2-Phenyl-4-(a-phenylethylidene)-5(4H)-
oxazolone, Z-1
4.6. Methyl (E)-2-benzamide-3-phenyl-2-butenoate, E-2
An identical procedure to that described above was applied
to the transformation of E-1 (2.30 g, 8.74 mmol) to E-2
(2.51 g, 97% yield). Mp 210 8C. Rf (hexanes/AcOEt
A mixture of acetophenone (36 g, 0.3 mol), hippuric acid
(10.26 g, 0.06 mol), acetic anhydride (18.36 g, 0.18 mol),
anhydrous lead(IV) acetate (13.29 g, 30 mmol) and dry THF
(100 mL) was heated under reflux for 24 h. After cooling,
the contents were poured onto crushed ice and extracted
with CH2Cl2. The combined organic phases were washed
with NaHCO3, brine and dried over MgSO4. The solvent
and most of the excess acetophenone were removed in
vacuo to afford a residue, which was crystallised from
EtOH/H2O to give a mixture of the geometric isomers of the
oxazolone. The pure (Z)-isomer was obtained after a second
recrystallization from EtOH/H2O as a yellow solid (5.73 g,
38% yield). Mp 109 8C. Rf (hexanes/benzene 1:1)¼0.23. IR
1
8:2)¼0.11. IR (nujol) 3321; 1721; 1639 cm21. H NMR
(CDCl3 300 MHz) d 2.18 (s, 3H); 3.46 (s, 3H); 7.15–7.35
(m, 5H); 7.35–7.60 (m, 4H); 7.87 (m, 2H). 13C NMR
(CDCl3 75 MHz) d 22.07; 51.86; 123.72; 127.17; 127.34;
127.69; 128.07; 128.71; 132.11; 133.39; 140.87; 141.01;
165.63; 165.78. MS-EI (m/z, %) 295 [(M)þ, 7]; 263
[(M2MeOH)þ, 17]; 190 (29); 130 (6); 105 (100); 77
[(C6H5)þ, 52]. Anal. calcd for C18H17NO3: C, 73.20; H,
5.80; N, 4.74. Found: C, 72.96; H, 5.86; N, 4.76.
4.7. Synthesis of threo-3 and erythro-3
1
(nujol) 1784; 1759 cm21. H NMR (CDCl3 300 MHz) d
2.78 (s, 3H); 7.42–7.54 (m, 6H); 7.86 (m, 2H); 8.05 (m,
2H). 13C NMR (CDCl3 75 MHz) d 18.37; 125.80; 127.92;
128.13; 128.77; 129.95; 131.20; 132.67; 138.88; 149.26;
160.39; 166.83. MS-EI (m/z, %) 263 [(M)þ, 19]; 105 (100);
77 [(C6H5)þ, 45]. Anal. calcd for C17H13NO2: C, 77.55; H,
4.98; N 5.32. Found: C, 77.91; H, 5.04; N, 5.19.
4.7.1. Methyl threo-2-benzamide-3-phenylbutanoate,
threo-3. A solution of Z-2 (2.2 g, 7.46 mmol) in AcOEt
(180 mL) was hydrogenated at 30 8C in the presence of 10%
Pd/C (200 mg). After 6 h the catalyst was filtered off and the
solvent evaporated to give diastereomerically pure threo-3
as a white solid (2.12 g, 96% yield). Mp 102 8C. Rf
(hexanes/AcOEt 8:2)¼0.16. IR (nujol) 3327; 1739;
1
4.4. (E)-2-Phenyl-4-(a-phenylethylidene)-5(4H)-
oxazolone, E-1
1631 cm21. H NMR (CDCl3 300 MHz) d 1.48 (d, 3H,
J¼7.1 Hz); 3.34 (q, 1H, J¼6.8 Hz); 3.61 (s, 3H); 5.02 (dd,
J¼6.3, 8.5 Hz); 6.61 (d, 1H, J¼8.3 Hz); 7.18–7.34 (m, 5H);
7.40–7.55 (m, 3H); 7.75–7.77 (m, 2H). 13C NMR (CDCl3
The (Z)-isomer (2.5 g) was dissolved in the minimum

´
M. Alıas et al. / Tetrahedron 60 (2004) 885–891
890
75 MHz) d 17.10; 43.04; 52.11; 57.97; 127.01; 127.29;
127.69; 128.49; 128.63; 131.78; 134.01; 141.03; 166.86;
171.84. MS-EI (m/z, %) 298 [(Mþ1)þ, 23]; 238
[(M2MeOH–CO)þ, 11]; 193 (23); 161 (9); 105 (100); 77
[(C6H5)þ, 60]. Anal. calcd for: C18H19NO3: C, 72.71; H,
6.44; N, 4.71. Found: C, 72.54; H, 6.52; N, 4.66.
4.8.2. erythro-b-Methylphenylalanine hydrochloride,
erythro-4. A similar procedure to that described above for
threo-4, starting from erythro-3 (2 g, 6.73 mmol), gave
erythro-4 as a white solid (1.35 g, 93% yield). Mp 200 8C.
Diastereomeric purity: 98%. Rf (CH2Cl2/MeOH 8:2)¼0. IR
1
(nujol) 3600–2500; 1728 cm21. H NMR (D2O 300 MHz)
4.7.1.1. Methyl (2R,3S)-2-benzamide-3-phenylbutano-
ate, (2R,3S)-3. Mp 135 8C. [a]2D0¼259.4 (c¼0.50, CHCl3).
Anal. calcd for: C18H19NO3: C, 72.71; H, 6.44; N, 4.71.
Found: C, 72.54; H, 6.36; N, 4.68.
4.7.1.2. Methyl (2S,3R)-2-benzamide-3-phenylbutano-
ate, (2S,3R)-3. Mp 135 8C. [a]2D0¼þ58.2 (c¼0.37, CHCl3).
Anal. calcd for: C18H19NO3: C, 72.71; H, 6.44; N, 4.71.
Found: C, 72.49; H, 6.42; N, 4.76. Spectroscopic data for
both (2R,3S)-3 and (2S,3R)-3 were the same as those
described above for threo-3.
d 1.33 (d, 3H, J¼7.2 Hz); 3.30 (q, 1H, J¼7.05 Hz); 4.01 (d,
1H, J¼7.5 Hz); 7.2–7.4 (m, 5H). 13C NMR (D2O 75 MHz)
d 16.98; 40.32; 59.02; 127.82; 128.22; 129.32; 139.27;
171.63. MS-FAB (m/z, %) 359 [(2Mþ1)þ, 13]; 202
[(MþNa)þ, 8]; 180 [(Mþ1)þ, 100]. Anal. calcd for
C10H14ClNO2: C, 55.69; H, 6.54; N, 6.49. Found: C,
57.21; H, 6.95; N, 6.13.
4.9. Synthesis of enantiomerically pure 5
4.9.1. (2R,3S)-b-Methylphenylalanine, (2R,3S)-5.
A
4.7.2. Methyl erythro-2-benzamide-3-phenylbutanoate,
erythro-3. In a similar way to that described above,
hydrogenation of E-2 (2.4 g, 8.14 mmol) using 10% Pd/C
(960 mg) for 10 h gave diastereomerically pure erythro-3 as
a white solid (82.35 g, 97% yield). Mp 118 8C. Rf (hexanes/
solution of (2R,3S)-3 (50 mg, 0.17 mmol) in HOAc/2.5 N
HCl (2.3 mL:9 mL) was heated at 125 8C for 24 h. The
solvent was removed in vacuo and the residue partitioned
between H2O/CH2Cl2. The phases were separated and the
aqueous layer was washed with CH2Cl2 and evaporated to
dryness to afford the amino acid hydrochloride. This
compound was converted into the free amino acid by
treatment with EtOH (2 mL) and propylene oxide (0.68 mL)
under reflux. Removal of the solvent afforded a residue that
was eluted through a C18 reverse-phase sep-pak cartridge.
Lyophilization of the aqueous phases gave (2R,3S)-5 as a
white solid (28 mg, 93% yield). Mp 220–222 8C.
[a]D20¼þ6.6 (c¼0.18, H2O) lit.45 (þ5.1; c¼1.1, H2O).
Diastereomeric purity: 99%. Rf (CH2Cl2/MeOH 8:2)¼0.88.
1
AcOEt 8:2)¼0.16. IR (nujol) 3345; 1742; 1639 cm21. H
NMR (CDCl3 300 MHz) d 1.41 (d, 3H, J¼7.2 Hz); 3.49 (m,
1H); 3.72 (s, 3H); 5.04 (dd, 1H, J¼5.4, 8.7 Hz); 6.30 (d, 1H,
J¼8.4 Hz); 7.1–7.5 (m, 8H); 7.66 (m, 2H). 13C NMR
(CDCl3 75 MHz) d 17.63; 42.23; 52.21; 57.47; 126.93;
127.16; 127.63; 128.05; 128.44; 128.57; 128.69; 131.74;
133.80; 140.70; 167.16; 171.96. MS-EI (m/z, %) 298
[(Mþ1)þ, 61]; 238 (19); 266 (6); 193 (24); 176 (40); 161
(9); 105 (100); 77 [(C6H5)þ, 54]. Anal. calcd for:
C18H19NO3: C, 72.71; H, 6.44; N, 4.71. Found: C, 72.55;
H, 6.49; N, 4.76.
4.7.2.1. Methyl (2R,3R)-2-benzamide-3-phenylbutano-
ate, (2R,3R)-3. Mp 138 8C. [a]2D0¼274.7 (c¼0.51, CHCl3).
Anal. calcd for: C18H19NO3: C, 72.71; H, 6.44; N, 4.71.
Found: C, 73.03; H, 6.36; N, 4.74.
4.7.2.2. Methyl (2S,3S)-2-benzamide-3-phenylbutano-
ate, (2S,3S)-3. Mp 137 8C. [a]2D0¼þ73.2 (c¼0.39, CHCl3).
Anal. calcd for: C18H19NO3: C, 72.71; H, 6.44; N, 4.71.
Found: C, 72.43; H, 6.39; N, 4.85. Spectroscopic data for
both (2R,3R)-3 and (2S,3S)-3 were the same as those
described above for erythro-3.
1
IR (nujol) 1629; 1574 cm21. H NMR (D2O 300 MHz) d
1.27 (d, 3H, J¼7.5 Hz); 3.42 (m, 1H): 3.82 (d, 1H,
J¼4.8 Hz); 7.24–7.25 (m, 5H). 13C NMR (D2O 75 MHz)
d 13.70; 39.51; 60.73; 127.73; 127.78; 129.09; 140.36;
173.54. MS-FAB (m/z, %) 359 [(2Mþ1)þ, 10]; 202
[(MþNa)þ, 18]; 180 [(Mþ1)þ, 100]. Anal. calcd for
C10H13NO2: C, 67.02; H, 7.31; N, 7.82. Found: C, 66.69;
H, 7.39; N, 7.71.
4.9.2. (2S,3R)-b-Methylphenylalanine, (2S,3R)-5. An
identical procedure to that described above was applied to
the transformation of (2S,3R)-3 (65 mg, 0.22 mmol) to
(2S,3R)-5 (36 mg, 92% yield). Mp 200–202 8C.
[a]D20¼27.49 (c¼0.15, H2O) lit.45 (25.3; c¼0.75, H2O).
Diastereomeric purity: 99.5%. Anal. calcd for C10H13NO2:
C, 67.02; H, 7.31; N, 7.82. Found: C, 66.65; H, 7.42; N, 7.70.
4.8. Synthesis of threo-4 and erythro-4
4.8.1. threo-b-Methylphenylalanine hydrochloride,
threo-4. A solution of threo-3 (2 g, 6.73 mmol) in HOAc/
2.5 N HCl (90 mL:360 mL) was heated for 42 h at
125 8C. The solvent was removed in vacuo and the
residue partitioned between H2O/CH2Cl2. The phases
were separated and the aqueous layer was washed with
CH2Cl2. Removal of water by lyophilization afforded
threo-4 as a white solid (1.36 g, 94% yield). Mp 214 8C.
Diastereomeric purity: 99%. Rf (CH2Cl2/MeOH 8:2)¼0.
4.9.3. (2R,3R)-b-Methylphenylalanine, (2R,3R)-5. The
procedure described above was applied to the transform-
ation of (2R,3R)-3 (95 mg, 0.32 mmol) to (2R,3R)-5 (52 mg,
91% yield). Mp 224 8C. [a]2D0¼þ27.2 (c¼0.10, H2O) lit.45
(þ21; c¼1.0, H2O). Diastereomeric purity: 97.7%. Rf
.
1H
(CH2Cl2/MeOH 8:2)¼0.88. IR (nujol) 1609 cm21
NMR (D2O 300 MHz) d 1.28 (d, 3H, J¼7.17 Hz); 3.16 (q,
1H, J¼7.35 Hz); 3.66 (d, 1H, J¼7.71 Hz); 7.21–7.34 (m,
5H). 13C NMR (D2O 75 MHz) d 17.65; 40.80; 60.98;
127.89; 127.89; 127.96; 129.25; 140.20; 173.79. MS-FAB
(m/z, %) 180 [(Mþ1)þ, 83]. Anal. calcd for C10H13NO2: C,
67.02; H, 7.31; N, 7.82. Found: C, 66.75; H, 7.39; N, 7.73.
IR (nujol) 3600–2400; 1734 cm21
.
1H NMR (D2O
300 MHz) d 1.32 (d, 3H, J¼6.5 Hz); 3.42 (q, 1H,
J¼6.5 Hz): 4.09 (d, 1H, J¼6.4 Hz); 7.20–7.34 (m, 5H).
13C NMR (D2O 75 MHz) d 14.83; 39.72; 58.68; 127.84;
128.16; 129.21; 139.10; 171.10. MS-FAB (m/z, %) 359
[(2Mþ1)þ, 10]; 202 [(MþNa)þ, 18]; 180 [(Mþ1)þ, 100].
Anal. calcd for C10H14ClNO2: C, 55.69; H, 6.54; N, 6.49.
Found: C, 55.89; H, 6.47; N, 6.55.
4.9.4. (2S,3S)-b-Methylphenylalanine, (2S,3S)-5. The
method described above was used for the conversion of

´
M. Alıas et al. / Tetrahedron 60 (2004) 885–891
891
19. Shapiro, G.; Buecheler, D.; Marzi, M.; Schmidt, K.; Gomez-
Lor, B. J. Org. Chem. 1995, 60, 4978–4979.
(2S,3S)-3 (110 mg, 0.37 mmol) to (2S,3S)-5 (60 mg, 91%
yield). Mp 224 8C. [a]D20¼229 (c¼0.30, H2O) lit.45 (226.7;
c¼1.0, H2O). Diastereomeric purity: 99.8%. Anal. calcd for
C10H13NO2: C, 67.02; H, 7.31; N, 7.82. Found: C, 66.53; H,
7.24; N, 7.77. Spectroscopic data were the same as those
described above for (2R,3R)-5.
20. Burk, M. J.; Gross, M. F.; Martinez, J. P. J. Am. Chem. Soc.
1995, 117, 9375–9376.
21. Burk, M. J.; Bedingfield, K. M.; Kiesman, W. F.; Allen, J. G.
Tetrahedron Lett. 1999, 40, 3093–3096.
22. Hruby, V. J.; Toth, G.; Gehrig, C. A.; Kao, L.-F.; Knapp, R.;
Lui, G. K.; Yamamura, H. I.; Kramer, T. H.; Davis, P.; Burks,
T. F. J. Med. Chem. 1991, 34, 1823–1830.
Acknowledgements
´ ´ ´ ´
23. Peter, A.; Toth, G.; Cserpan, E.; Tourwe, D. J. Chromatogr. A
1994, 660, 283–291.
´ ´
24 Peter, A.; Toth, G.; Torok, G.; Tourwe, D. J. Chromatogr. A
This work was supported by M.C.Y.T. (project PPQ2001-
1834), PETRI (PTR95/0422-OP) and Productos Aditivos
S. A. The authors thank A. L. Bernad for HPLC assistance.
´
¨ ¨
1996, 728, 455–465.
´
´ ´
26. Peter, A.; Peter, M.; Fu¨lop, F.; Torok, G.; Toth, G.; Tourwe,
´
25. Peter, A.; Toth, G. Anal. Chim. Acta 1997, 352, 335–356.
´
´
¨
¨ ¨
´
D.; Sapi, J. Chromatographia 2000, 51, S-148–154.
References and notes
´
27. Peter, A.; Torok, G.; Armstrong, D. W. J. Chromatogr. A
1998, 793, 283–296.
¨ ¨
1. Balaram, P. Curr. Opin. Struct. Biol. 1992, 2, 845–851.
2. Liskamp, R. M. J. Recl. Trav. Chim. Pays-Bas 1994, 113,
1–19.
´ ´ ´
28. Peter, A.; Torok, G.; Armstrong, D. W.; Toth, G.; Tourwe, D.
J. Chromatogr. A 1998, 828, 177–190.
¨ ¨
´ ´ ´
29. Peter, A.; Torok, G.; Armstrong, D. W.; Toth, G.; Tourwe, D.
J. Chromatogr. A 2000, 904, 1–15.
¨ ¨
3. Hruby, V. J.; Al-Obeidi, F. A.; Kazmierski, W. Biochem. J.
1990, 268, 249–262.
´ ´
30. Peter, A.; Torok, G.; Toth, G.; Lindner, W. J. High Resolut.
Chromatogr. 2000, 23, 628–636.
¨ ¨
4. Hruby, V. J. Acc. Chem. Res. 2001, 34, 389–397.
5. Hruby, V. J.; Li, G.; Haskell-Luerano, C.; Shenderovich, M.
Biopolymers 1997, 43, 219–266.
´
¨ ¨
31. Torok, G.; Peter, A.; Armstrong, D. W.; Tourwe, D.; Toth, G.;
´
´
´
Sapi, J. Chirality 2001, 13, 648–656.
´
6. Bonner, G. G.; Davis, P.; Stropova, D.; Sidney, E.; Yamamura,
H. I.; Porreca, F.; Hruby, V. J. J. Med. Chem. 2000, 43,
569–580.
32. Cativiela, C.; Melendez, E. Synthesis 1978, 832–833.
´
33. Cativiela, C.; Melendez, E. Synthesis 1980, 901–902.
´
34. Cativiela, C.; Melendez, E. Synthesis 1981, 805–807.
7. Gibson, S. E.; Guillo, N.; Tozer, M. J. Tetrahedron 1999, 55,
585–615.
35. Kharasch, M.; Mansfield, J. V.; Mayo, F. R. J. Am. Chem. Soc.
1937, 59, 1155.
36. Rao, Y. S.; Filler, R. Y. Synthesis 1975, 749–764.
8. Huang, Z.; He, Y.-B.; Raynor, K.; Tallent, M.; Reisine, T.;
Goodman, M. J. Am. Chem. Soc. 1992, 114, 9390–9401.
´
37. Cativiela, C.; Dıaz-de-Villegas, M. D.; Mayoral, J. A.;
´
Melendez, E. Synthesis 1983, 899–902.
´
9. Tourwe, D.; Mannekeuse, E.; Nguyen Thi Diem, T.;
´
´ ´
Verheyden, P.; Jaspers, H.; Toth, G.; Peter, A.; Kertesz, I.;
38. Rylander, P. N. Hydrogenation methods; Academic: London,
1990; pp 29–52.
¨ ¨
Torok, G.; Chung, N. N.; Schiller, P. W. J. Med. Chem. 1998,
41, 5167–5176.
39. Okamoto, Y.; Yashima, E. Angew. Chem., Int. Ed. Engl. 1998,
37, 1020–1043.
10. Mosberg, H. I.; Omnaas, J. R.; Lonize, A.; Heyl, D. L.;
Nordan, I.; Mousigian, C.; Davis, P.; Porreca, F. J. Med.
Chem. 1994, 37, 4384–4391.
40. Dingenen, J. In A practical approach to chiral separations by
liquid chromatography; Subramanian, G., Ed.; Weinheim:
New York, 1994; pp 115–181.
11. Kataoka, Y.; Seto, Y.; Yamamoto, M.; Yamada, T.; Kuwata,
S.; Watanabe, H. Bull. Chem. Soc. Jpn. 1976, 49, 1081–1084.
12. Effenberger, F.; Weber, T. Angew. Chem., Int. Ed. Engl. 1987,
26, 142–143.
´ ´ ´
41. Cativiela, C.; Dıaz-de-Villegas, M. D.; Jimenez, A. I.; Lopez,
P.; Marraud, M.; Oliveros, L. Chirality 1999, 11, 583–590.
´
´ ´
42. Alıas, M.; Cativiela, C.; Jimenez, A. I.; Lopez, P.; Oliveros,
L.; Marraud, M. Chirality 2001, 13, 48–55.
´ `
13. Pasto, M.; Moyano, A.; Pericas, M. A.; Riera, A. J. Org.
Chem. 1997, 62, 8425–8431.
14. Davis, F. A.; Liang, C. H.; Liu, H. J. Org. Chem. 1997, 62,
3796–3797.
´ ´
43. Jimenez, A. I.; Lopez, P.; Oliveros, L.; Cativiela, C.
Tetrahedron 2001, 57, 6019–6026.
´ ´
44. Royo, S.; Lopez, P.; Jimenez, A. I.; Oliveros, L.; Cativiela, C.
15. Tsuchihashi, G. I.; Mitamura, S.; Ogura, K. Bull. Chem. Soc.
Jpn. 1979, 52, 2167–2168.
Chirality 2002, 14, 39–46.
45. Dharanipragada, R.; Van Hulle, K.; Bannister, A.; Bear, S.;
Kennedy, L.; Hruby, V. J. Tetrahedron 1992, 48, 4733–4748.
16. Dharanipragada, R.; Nicolas, E.; Toth, G.; Hruby, V. J.
Tetrahedron Lett. 1989, 30, 6841–6844.
´ ´
46. Oliveros, L.; Lopez, P.; Minguillon, C.; Franco, P. J. Liq.
17. Li, G.; Jarosinski, M. A.; Hruby, V. J. Tetrahedron Lett. 1993,
34, 2561–2564.
Chromatogr. 1995, 18, 1521–1532.
47. Franco, P.; Senso, A.; Minguillon, C.; Oliveros, L.
J. Chromatogr. A 1998, 796, 265–272.
´
18. Oppolzer, W.; Tamura, O.; Deerberg, J. Helv. Chim. Acta
1992, 75, 1965–1978.