Synthesis of enantiomerically pure cis- and trans-cyclopentane analogues of phenylalanine

Article abstract of DOI:10.1016/j.tetasy.2005.06.017

For the first time, all stereoisomers of 1-amino-2- phenylcyclopentanecarboxylic acid-c₅Phe-have been synthesised. A Strecker reaction on 2-phenylcyclopentanone and further transformations of each amino nitrile into the amino acid provides cis-

Full text of DOI:10.1016/j.tetasy.2005.06.017

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 2613–2623
Synthesis of enantiomerically pure cis- and
trans-cyclopentane analogues of phenylalanine
*
Carlos Cativiela, Marta Lasa and Pilar Lopez
*
´
´ ´ ´ ´ ´
Departamento de Quımica Organica, Instituto de Ciencias de Materiales de Aragon, Instituto Universitario de Catalisis Homogenea,
Universidad de Zaragoza-CSIC, 50009 Zaragoza, Spain
Received 2 June 2005; accepted 13 June 2005
Abstract—For the first time, all stereoisomers of 1-amino-2-phenylcyclopentanecarboxylic acid—c5Phe—have been synthesised.
A Strecker reaction on 2-phenylcyclopentanone and further transformations of each amino nitrile into the amino acid provides
cis-c5Phe and trans-c5Phe with high efficiency. A divergent synthetic route was then developed to obtain the target compounds
cis- and trans-c5Phe in their racemic form. The preparation of the final enantiomerically pure amino acids and their corresponding
N-protected derivatives was also achieved by HPLC resolution of one of the intermediates using a cellulose-derived chiral stationary
phase. The relative stereochemistry of each amino acid and its precursors have been unambiguously assigned.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
pentanone that, after hydrolysis of the corresponding
hydantoin, led to only one isomer of c5Phe in its racemic
form,6 without specifying the stereochemistry.
Structure–activity relationship studies on bioactive pep-
tides are essential for establishing the active conforma-
tion of these systems. In this context, the side-chain
moieties warrant special attention since they are directly
involved in peptide-receptor recognition phenomena
and determine biological specificity. The use of con-
strained amino acids with rigidly oriented side chains
provides very valuable information in such investiga-
tions.1 In particular, restricted amino acid analogues
with aromatic rings have been used for this purpose with
some of them being met with considerable success.2,3
Over the course of our studies into the conformational
tendencies of carbocyclic analogues of phenylalanine
(Phe) in model dipeptides a series of 1-amino-2-phenyl-
cycloalkanecarboxylic acids (cnPhe) was synthesised. In
particular, we reported the synthesis of all the stereoiso-
mers of c3Phe4 and c6Phe.5 Nevertheless, in spite of the
potential interest of c5Phe, its stereoisomers have not yet
been reported.
Herein, we report the development of a highly efficient
synthesis for both cis- and trans-c5Phe in racemic form
and the preparation of enantiomerically pure (1R,2R)-,
(1S,2S)-, (1R,2S)- and (1S,2R)-c5Phe, suitably protected
to be incorporated into a peptide chain (Fig. 1).
COOH
NH2
COOH
NH2
COOH
NH2
Ph
Ph
Ph
cnPhe
c5Phe
Phe
Figure 1. Structure of 1-amino-2-phenylcycloalkanecarboxylic acids
(cnPhe) as conformationally restricted analogues of Phe.
2. Results and discussion
To the best of our knowledge, the only reference in
the literature concerning the synthesis of 2-phenyl-1-
aminocyclopentanecarboxylic acid (c5Phe) describes a
2.1. Synthesis of racemic cis- and trans-c5Phe
In recent years, a number of cis- and trans-1-aminocyclo-
pentanecarboxylic acids with different substituents in
Bucherer–Bergs reaction starting from 2-phenylcyclo-
¨
*
Corresponding authors. Tel.: +34 976 761210 (C.C.); tel.: +34 976
762275; fax: +34 976 761210 (P.L.); e-mail addresses: cativiela@
unizar.es; pilopez@unizar.es
cis- or trans-c5Phe refers to the relative position between the amino
and the phenyl groups in the cyclic analogue of phenylalanine.
0957-4166/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2005.06.017

2614
C. Cativiela et al. / Tetrahedron: Asymmetry 16 (2005) 2613–2623
the ring have been synthesised, most times in enantio-
merically pure form. In particular, cyclic analogues of
glutamate have been extensively studied, due to their
importance as agonists or antagonists of the glutamate
receptors. Other cyclopentane amino acid analogues,
as serine or threonine, have also been reported. The
synthetic routes to obtain 1-aminocyclopentane-
carboxylic acids can be divided into two groups.à The
first group involves the formation of the cyclopentane
ring and includes cyclisation by alkylation of a glycine
equivalent with a 1,4-dielectrophile7 and ring closing
metathesis (RCM) on geminal alkenyl or alkynyl bislac-
tims8 or other glycine equivalents.9 A recent example of
this last synthetic pathway was reported by Trost, who
described the synthesis of a precursor of cis-1-amino-2-
phenylcyclopentanecarboxylic acid through a RCM
reaction on an a,a-dialkenyl N-acylglycinate, although
free amino acid was not reported.10 Other methodolo-
gies, such as palladium-mediated cyclisation,11 enantio-
selective alkylidene carbene C–H insertion reactions12
or synthesis of the cyclopentane ring through sequential
aldol reactions13 and intramolecular aldol condensa-
tion14 have also been used. On the other hand, the intro-
duction of the amino and acid functions on the
cyclopentane skeleton was carried out in a traditional
Although the asymmetric Strecker reaction on different
2-substituted cyclopentanones was described,17b,d to
the best of our knowledge, the reaction of 2-phenyl-
cyclopentanone has not yet been reported for such a
route and we were therefore encouraged to study this
approach to obtain the different isomers of c5Phe.
The 2-phenylcyclopentanone used as the starting mate-
rial was synthesised as described in the literature by
reaction of phenylmagnesium bromide and 2-chloro-
cylopentanone.19 Treatment of 2-phenylcyclopentanone
with NaCN and NH4Cl under the optimal conditions
found for the Strecker reaction on 2-phenylcyclohexa-
none5c (ketone/NaCN/NH4Cl in a ratio of 1:2:2, long
reaction times) afforded in all cases mixtures of
cis- and trans-1-cyano-2-phenylcyclopentylamine (rac-
1 and rac-2, respectively, Scheme 1) in a 1:1 ratio. It
was found that the longer the reaction time, the higher
the yield of the products. After 10 days of reaction, a
mixture of stereoisomers was obtained with 69% conver-
sion, a result that unfortunately could not be improved
upon, probably due to the decomposition of the starting
ketone at longer reaction times. Both isomers were sep-
arated and purified by column chromatography and
were obtained as colourless oils that would not crystal-
lise on standing at low temperatures or by the addition
of solvents.
way by Bucherer–Bergs or Strecker reactions on the
¨
corresponding cyclopentanones. The Bucherer–Bergs
¨
reaction has been applied to the synthesis of different
cyclopentane derived a-amino acids in their racemic
form6,15 and, more recently, to obtain enantiopure
cyclopentane Glu analogues.16 Although less used, the
Strecker reaction has also proven to be convenient for
the synthesis of amino acids with a cyclopentane skele-
ton.17 Next to these approaches, the construction of a
quaternary a-carbon by direct organocatalytic amina-
tion of b-dicarbonyl compounds has been reported.18
The relative stereochemistry of the separated isomers
was initially assigned on the basis of H NMR spectro-
1
scopy, taking into account the deshielding effect of the
triple bond of the cyano group on the hydrogen atom
in the relative cis-disposition. In this case, the benzylic
proton of the cis-derivative rac-1 resonates at a lower
field than the corresponding proton in the trans-isomer
rac-2.
As described in the introduction, the only reported way
The applied reaction conditions (room temperature,
long reaction times) are consistent with a thermodynam-
ically controlled reaction. Moreover, after purification
by column chromatography, both rac-1 and rac-2 evolve
on standing, to mixtures of cis- and trans-isomers.
to obtain the amino acid c Phe involves a Bucherer–
¨
5
Bergs reaction starting from 2-phenylcyclopentanone.6
In this case, a single spot in TLC was observed for the
obtained hydantoin and, hence, after hydrolysis under
strong conditions (H2SO4, 150 ꢁC) only one isomer of
c5Phe with an unknown relative cis/trans stereochemis-
try was described by the authors. The melting point
was the only described physical characteristic of c5Phe
obtained. Taking into account that the epimerisation
during the hydrolysis process at elevated temperatures
has been described for different spirohydantoins
obtained from 2-substituted cyclopentanones,15a,b the
described synthesis of the amino acid c5Phe is, at the
least, questionable.
It is noteworthy that the result of the Strecker reaction
on 2-phenylcyclopentanone (an equimolecular mixture
of cis- and trans-isomers) was so different from the
behaviour of 2-phenylcyclohexanone under the same
conditions (completely diastereoselective trans selec-
tion). To understand this result, a molecular modelling
study of the amino nitrile precursors of c5Phe (cis-amino
nitrile 1 and trans-amino nitrile 2) and the correspond-
ing amino nitrile precursors of c6Phe was performed at
molecular mechanics (MM) and DFT levels. The results
indicate the greater thermodynamic stability of the
trans-amino nitrile precursor of c6Phe, whereas in the
case of c5Phe, the energy difference between cis- and
trans-diastereomers is much lower, which is consistent
with the reaction mixtures of both isomers obtained in
the experiments.
Over the course of our work on the synthesis of cyclic
analogues of phenylalanine, we recently described an
efficient synthesis of trans-c6Phe through a completely
diastereoselective Strecker reaction and further hydro-
lysis of an amide derived from the amino nitrile.5c
à References concerning the synthesis of cyclopentanic aminoacids
incorporated in rigid bi- or tricyclic structures are deliberately not
included in this revision.
In order to prevent the establishment of the equilibrium
between cis- and trans-isomers rac-1 and rac-2 after
their purification, both isomeric amino nitriles were

C. Cativiela et al. / Tetrahedron: Asymmetry 16 (2005) 2613–2623
2615
NHCOMe
Ph
b
CN
75%
H
NH2.HCl
Ph
NH2
Ph
d
rac-3
rac-4
CO2H
H
95%
CN
H
34%
NHCOPh
Ph
CN
c
rac-7
rac-1
73%
H
O
a
Ph
CN
Ph
b
NHCOMe
H
75%
CO2H
Ph
CN
Ph
d
rac-5
rac-6
99%
NH2
H
NH2.HCl
H
35%
CN
Ph
c
rac-8
rac-2
NHCOPh
H
88%
Scheme 1. Synthesis of racemic cis-c5Phe and trans-c5Phe by Strecker reaction. Reaction conditions: (a) (i) NaCN, NH4Cl, i-PrOH, NH4OH; (ii)
column chromatography; (b) AcCl, NEt3, CH2Cl2; (c) BzCl, NEt3, CH2Cl2; (d) 12 N HCl, reflux.
transformed into derivatives immediately after being
isolated by column chromatography. Thus, acetamide
and benzamide derivatives of each amino nitrile were
synthesised, since we have previously reported that the
hydrolysis of acetamido- or benzamidonitriles is much
more efficient than hydrolysis of the starting amino
nitrile.5c On the other hand, these amides proved to be
very convenient substrates for the enantiomeric resolu-
tion through chiral HPLC using polysaccharide carb-
amates as chiral selectors in the stationary phase. As
indicated in Scheme 1, acetamide rac-3 and benzamide
rac-4 were obtained from cis-compound rac-1 under
standard conditions with excellent yields. In a similar
way, acetamide rac-5 and benzamide rac-6 were
obtained from trans-isomer rac-2.
A divergent synthetic route to the target compounds cis-
and trans-c5Phe in their racemic forms was then devel-
oped (three synthetic steps from 2-phenylcyclopenta-
none, with a similar overall yield for each compound
of around 25%). These racemic compounds were now
ready for direct use in the synthesis of diastereomeric
peptides. For this purpose, the amino acids must be suit-
ably protected.
In order to obtain the N-Boc-protected derivatives, we
employed the same methodology (TMAH, acetonitrile,
Boc2O) as used to protect the amino group of sterically
hindered trans-c6Phe.5c Under these conditions, as
observed previously, the solubility of the starting mate-
rial proved to be critical for the success of the reaction:
when the amino acid tetramethylammonium salt was
not completely in solution, the reaction proceeded very
slowly and gave a very poor yield. The low solubility
of the starting materials in acetonitrile meant that it
was not possible to perform directly the N-protection
starting from the corresponding c5Phe hydrochlorides
rac-7 and rac-8, and prior to liberation of the free amino
acids was necessary through a treatment with propylene
oxide. Mild heating of the reaction mixture (TMAH,
acetonitrile, 40 ꢁC) enabled free trans-c5Phe to be dis-
solved and the reaction with tert-butyl dicarbonate
was carried out. N-Protection progressed by successive
feeding with Boc2O and Boc-trans-c5Phe rac-9 was final-
ly obtained in 75% yield after a reaction time of 2 days
(Scheme 2). However, free cis-c5Phe could not be dis-
solved under any conditions and, after addition of
tert-butyl dicarbonate and successive feeding, only the
starting material was recovered. This lack of reactivity
of rac-7 is probably due to the presence of the large
phenyl group in a cis relative position with respect
to the amine. The different behaviour of rac-7 and rac-
8 in this N-protection process is consistent with the
Moreover, these amido derivatives allowed us to assign
the relative stereochemistry of each diastereoisomer by
NOE experiments. For example, irradiation of the NH
of the acetamido group in rac-3 and rac-5 induces
enhancement of the benzylic proton in rac-5 (4.6%)
and a significantly smaller effect (0.7%) in rac-3. In the
same way, irradiation of the NH of the benzamido
group in rac-4 and rac-6 shows a greater NOE effect
on the benzylic proton of rac-6 (6.8%) than on the cor-
responding proton of rac-4 (0.9%). These NOEs are con-
sistent with a cis-configuration between the phenyl and
amido groups in rac-3 and rac-4 and, hence, between
the phenyl and amine in rac-1. Correspondingly, rac-2,
rac-5 and rac-6 can be assigned as precursors of trans-
c5Phe.
Treatment of rac-3 and rac-4 with concentrated HCl
under reflux resulted in a quantitative transformation
into cis-c5Phe hydrochloride rac-7. Likewise, hydrolysis
of both rac-5 and rac-6 led to trans-c5Phe hydrochloride
rac-8.

2616
C. Cativiela et al. / Tetrahedron: Asymmetry 16 (2005) 2613–2623
from cellulose and amylose are extremely popular be-
CO2H
Ph
NHBoc
H
i) propylene oxide, EtOH
ii) TMAH,Boc2O, MeCN
cause of their wide applicability and usefulness in enan-
tioselective liquid chromatography.21 Our research
group collaborated in the development of new polysac-
charide-derived CSPs where mixed polysaccharide
derivatives are covalently bonded to an allylsilica gel
matrix.22 The covalent immobilisation results in an
extremely high stability for these phases in the presence
of a wide range of solvents. Although these phases are
not commercially available, the synthetic methodology
has been extensively studied and reported.23 The prepa-
ration of these phases has been meticulously described,
so that these in-house materials are readily prepared
with reproducible chromatographic results. The syn-
thetic simplicity, high chemical stability and the great
selectivity exhibited towards a variety of compounds
make these phases especially suitable for resolutions
on a preparative scale, as shown in the preparative enan-
tioseparations of different phenylalanine4,24 and other
amino acid25 surrogates.
CO2H
Ph
75%
rac-9
NH2.HCl
H
CO2H
Ph
NHZ
H
i) TMSCl, CH2Cl2
ii) (iPr)2EtN, Z-Cl
rac-8
27%
rac-11
NH2.HCl
Ph
CO2H
H
NHZ
Ph
CO2H
H
i) TMSCl, CH2Cl2
ii) (iPr)2EtN, Z-Cl
57%
rac-7
rac-10
Scheme 2. Synthesis of racemic N-protected trans-c5Phe and cis-c5Phe.
relative stereochemistry assigned to the two stereoiso-
mers by NMR studies.
Herein, two non-commercial polysaccharide-derived
supports consisting of mixed 10-undecenoate/3,5-
dimethylphenylcarbamate of cellulose22 and amylose,26
respectively, covalently attached to allyl silica were used
(CSP-1 and CSP-2). The HPLC analytical resolution of
the precursors of cis-c5Phe (rac-3, rac-4) and trans-c5Phe
(rac-5, rac-6) was tested on both columns. Different mix-
tures of n-hexane/i-PrOH and n-hexane/i-PrOH/acetone
were tested as eluents (flow rate 1 mL/min) with UV
monitoring performed at 210 nm. The most representa-
tive results are shown in Table 1.
The resistance to N-protection in the presence of bulky
protecting groups has also been noted in other cyclo-
alkane amino acids, in which the amino and phenyl
groups are in a cis disposition, as in the case of cis-
c6Phe.5b To overcome the low reactivity of the amino
group, we carried out the same strategy as used in the
synthesis of N-protected derivatives of cis-c6Phe: that
is, prior to transformation of the carboxylic acid func-
tion into its trimethylsilyl ester. As observed previously,
this transitory protection of the carboxylate provides an
excellent yield in the subsequent reaction of the amino
group with a chloroformate. This methodology (addi-
tion of chlorotrimethylsilane and subsequent reaction
of the amine with benzyl chloroformate) led to Z-cis-
c5Phe rac-10 from rac-7 in 57% yield (Scheme 2). The
same reaction, starting from trans-amino acid rac-8,
provided Z-trans-c5Phe rac-11 with a lower yield (27%,
Scheme 2). The low reactivity of rac-8 under these reac-
tion conditions is in agreement with a cis geometry
between the carboxylic acid and phenyl group and is
consistent with the relative stereochemical assignment
established previously.
Resolution of cis-acetamido derivative rac-3 on CSP-1
was unsatisfactory, with low values for the separation
and resolution factors on using mixtures of hexane/
i-PrOH as the eluent. This result could not be improved
by adding a certain amount of acetone to the elution
system (Table 1, entries 1 and 2). Separation of rac-3
could not be achieved on CSP-2 under any chromato-
graphic conditions explored. However, cis-benzamido
derivative rac-4 was resolved on both columns (entries
3–5), with the best chromatographic results obtained
on CSP-1 (derived from the 3,5-dimethylphenylcarba-
mate of cellulose) using a ternary mixture (hexane/
i-PrOH/acetone 95/3/2) as the elution system (entry 4).
The relative stereochemistry of the amino nitriles rac-1
and rac-2, and their corresponding derivatives rac-3–
rac-11, was later unequivocally assigned by X-ray crys-
tal structure analysis of some dipeptides derived, respec-
tively, from amino acids rac-7 and rac-8.
Resolution of trans derivatives rac-5 and rac-6 was sim-
ilar to that achieved for the cis compounds. Separation
of trans-acetamido derivative rac-5 could not be
achieved on CSP-1 (low a and Rs, entries 7 and 8) or
on CSP-2 (no resolution). As in the case of rac-4,
trans-benzamido derivative rac-6 was resolved on both
columns with good chromatographic parameters (en-
tries 9 and 11) and these could be improved on CSP-1
by adding a certain amount of acetone to the eluent.
The use of a 95/3/2 mixture of hexane/i-PrOH/acetone
as the elution system gave the highest values for the sep-
aration factor a and resolution Rs (entry 10).
2.2. HPLC resolution of rac-4 and rac-6
Once an efficient route to the target compounds had
been developed, we undertook the preparation of both
enantiomers of amino acids 7 and 8 in enantiomerically
pure form. Although enantiomers can be obtained by
the separation of diastereoisomers, the direct separation
of enantiomers by preparative chromatography on
chiral stationary phases (CSPs) is today recognised as
a powerful alternative.20 Polysaccharide-derived CSPs
Although the mechanism for chiral discrimination on
CSPs derived from arylcarbamates of polysaccharides
is not clear due to its complexity, the interaction

C. Cativiela et al. / Tetrahedron: Asymmetry 16 (2005) 2613–2623
2617
Table 1. Selected chromatographic data for the HPLC resolution of cis-c5Phe and trans-c5Phe precursors on several stationary phases and
chromatographic modes
d
Entry
Compound
CSP
Eluentc A/B/C
k (nm)
k10
ad
Rsd
1
rac-3
rac-3
rac-4
rac-4
rac-4
rac-4
rac-5
rac-5
rac-6
rac-6
rac-6
rac-6
CSP-1a
CSP-1a
CSP-1a
CSP-1a
CSP-2a
CSP-1b
CSP-1a
CSP-1a
CSP-1a
CSP-1a
CSP-2a
CSP-1b
93/7/0
95/2/3
98/2/0
95/3/2
98/2/0
95/3/2
93/7/0
95/3/2
95/5/0
95/3/2
98/2/0
95/3/2
210
210
210
210
210
270
210
210
210
210
210
270
3.04
4.25
5.31
2.71
3.50
3.24
3.24
4.46
3.41
3.09
5.20
3.49
1.08
1.06
1.30
1.23
1.15
1.14
1.09
1.06
1.24
1.22
1.24
1.25
—
—
2
3
1.50
1.88
0.70
0.83
—
4
5
6
7
8
—
9
1.32
1.97
1.36
1.21
10
11
12
a Analytical column dimensions: 150 · 4.6 mm ID, flow rate: 1 mL/min, injection volume: 5 lL, c = 5 mg/mL, samples dissolved in chloroform.
b Semipreparative column dimensions: 150 · 20 mm ID, flow rate: 18 mL/min, injection volume: 100 lL, c = 300 mg/mL, samples dissolved in
chloroform.
c A: n-hexane, B: 2-propanol, C: acetone.
d For definition of k0 (capacity factor), a (separation factor) and Rs (resolution factor), see Experimental section.
between the carbamate residues of the chiral selector
with the racemate, mainly through hydrogen bonds
involving the carbamate groups, is proposed as the most
important factor for an effective enantiodiscrimina-
tion.21b,21c Additional p–p interactions involving aryl
groups of the carbamate residue and aromatic groups
of solute are also considered in this recognition mecha-
nism. This hypothesis could account for the results
obtained in the separation of compounds rac-3–6:
benzamide derivatives rac-4 and rac-6 have an N–H
amide bond that is more polar than the one in acetamide
compounds rac-3 and rac-5. Furthermore, the benzam-
ide derivatives also have an additional phenyl group
near the amide group. The interactions of these com-
pounds with the chiral selector should therefore be
stronger and their resolution higher, a situation consis-
tent with the experimental results.
440 mg of the first eluted enantiomer in its enantiomeri-
cally pure form. The second fraction contained 45 mg of
an 85/15 mixture of the first and the second eluted enan-
tiomers. The third and fourth fractions provided 150
and 350 mg, respectively, of mixtures 15/85 and 4/96
enriched in the last eluted enantiomer. These last frac-
tions were collected and reinjected under the same con-
ditions to give 315 mg of the last eluted enantiomer in an
enantiomerically pure form.
In a similar way, HPLC resolution of rac-6 was carried
out using the same column and under the same condi-
tions. In this case, a solution of 1 g of rac-6 in 3.3 mL
of chloroform was injected (repetitive injections of
100 lL each 12 min) onto the semipreparative column
and three separate fractions were collected. The first
fraction contained 440 mg of the enantiomerically pure
first eluted enantiomer and the third fraction provided
525 mg of the last eluted enantiomer in a 9/91 enantio-
meric ratio. The second fraction (25 mg) contained a
90/10 mixture of the first and the second eluted enantio-
mers. The final purification step involved reinjecting the
fraction enriched in the second enantiomer and the new
fractions were collected in such a way that the enantio-
merically pure enantiomer was obtained (330 mg).
Finally, under the conditions described above, the sepa-
rations of rac-4 and rac-6 were performed at the semi-
preparative level. The column saturation capacity (Ws)
was calculated working in overload mode with the ana-
lytical column (150 · 4.6 mm ID) by injecting increasing
amounts of compound. When concentrated samples of
benzamides rac-4 and rac-6 (300 mg/mL CHCl3) were
injected into the analytical column, values of a and Rs
were still very favourable for an efficient separation.
Once Ws had been determined, we scaled-up the analyt-
ical resolution to the preparative column (150 · 20 mm
ID) and a loading capacity of 30 mg was obtained for
each compound. Values of a and Rs for the semiprepar-
ative resolution of rac-4 and rac-6 are given in entries 6
and 12, respectively.
The enantiopurity of the resolved enantiomers of comp-
ounds 4 and 6 was assessed at the analytical level (Figs.
2 and 3, respectively). The absolute configurations of
the compounds were assigned by X-ray diffraction
analysis of some c5Phe-containing dipeptides.27 In this
way we were able to establish that the absolute configu-
ration of the enantiomer of cis benzamide 4 eluted first
by HPLC was (1R,2R) and the second one had a
(1S,2S) stereochemistry. Likewise, the less strongly
retained isomer of trans benzamide 6 could be assigned
as (1R,2S) and the last eluted enantiomer as (1S,2R).
HPLC resolution of rac-4 (1 g) dissolved in chloroform
(3.3 mL) was carried out by successive injections of
100 lL at intervals of 10 min on a 150 · 20 mm ID col-
umn filled with the 10-undecenoate/3,5-dimethylphenyl-
carbamate of cellulose bonded on allylsilica gel, using a
95/3/2 mixture of hexane/i-PrOH/acetone as the eluent
(flow-rate 18 mL minꢀ1). Four separate fractions were
collected. Evaporation of the first fraction provided
Moreover, this unequivocal assignment allowed the con-
firmation of the relative stereochemistry of the amino
nitriles rac-1 and rac-2, which had previously been pro-
1
posed on the basis of H NMR shifts and NOE effects.

2618
C. Cativiela et al. / Tetrahedron: Asymmetry 16 (2005) 2613–2623
Figure 2. HPLC analytical resolution of cis benzamide rac-4 (A) and resolved enantiomers (1R,2R)-4 (B) and (1S,2S)-4 (C). Column: 150 · 4.6 mm
ID containing 3,5-dimethylphenylcarbamate of cellulose (CSP-1). Eluent: n-hexane/2-propanol/acetone 95/3/2. Flow rate: 1 mL/min. UV detection:
210 nm. Chromatographic parameters: k01 ¼ 2.71; a = 1.23; Rs = 1.88.
Figure 3. HPLC analytical resolution of trans benzamide rac-6 (A) and resolved enantiomers (1R,2S)-6 (B) and (1S,2R)-6 (C). Column: 150 · 4.6 mm
ID containing 3,5-dimethylphenylcarbamate of cellulose (CSP-1). Eluent: n-hexane/2-propanol/acetone 95/3/2. Flow rate: 1 mL/min. UV detection:
210 nm. Chromatographic parameters: k01 ¼ 3.09; a = 1.22; Rs = 1.97.
2.3. Synthesis of c5Phe amino acids in enantiomerically
pure form and their N-protected derivatives
ate racemic precursors of cis-c5Phe and trans-c5Phe were
subjected to HPLC resolution on an easily prepared cel-
lulose-derived chiral stationary phase. The enantiopure
amino acids obtained in this way were suitably protected
for incorporation into peptides whose structural and
biological studies could clarify certain aspects about
the effects induced by this type of conformational
restriction and the influence of the absolute configura-
tions of the stereogenic centres.
After HPLC resolution of rac-4 and rac-6, the isolated
enantiomers were submitted to acid hydrolysis under
the conditions previously developed for the racemic
material. This process afforded in high yields the desired
enantiopure amino acid hydrochlorides (1R,2R)-7,
(1S,2S)-7, (1R,2S)-8 and (1S,2R)-8, respectively
(Scheme 3).
Enantiopure Z-protected amino acids (1R,2R)-10 and
(1S,2S)-10 were synthesised from the corresponding
amino acid hydrochlorides (1R,2R)-7 and (1S,2S)-7
under the conditions previously developed for the racemic
material (Scheme 3). Respectively, N-Boc amino acids
(1R,2S)-9 and (1S,2R)-9 were obtained from the corre-
sponding enantiopure amino acids under the conditions
that we have previously established for the racemate
(Scheme 3).
4. Experimental
4.1. General
All reagents, with the exception of 2-phenylcyclopenta-
none, were purchased from the Aldrich Chemical Co.
(Milwaukee, WI) and used without further purification.
2-Phenylcyclopentanone was synthesised as described
in the literature by the reaction of phenylmagnesium
bromide and 2-chlorocyclopentanone.19 Solvents were
dried, when necessary, by standard methods. The pro-
gress of the reactions was checked by thin layer chromato-
graphy (TLC) on Merck 60 F240 precoated silica gel
polyester plates while products were visualised under
UV light (254 nm), iodine vapour or ninhydrin chro-
matic reaction as appropriate. Column chromatography
was performed using Merck silica gel (40–60 lm). The
solvents used as HPLC mobile phases were of chromo-
3. Conclusion
Starting from readily available substrates and using
high-yield transformations, conformationally restricted
phenylalanine analogues cis-c5Phe and trans-c5Phe have
been prepared for the first time. Moreover, we have
developed a strategy for the synthesis of all stereoiso-
mers of c5Phe in enantiomeric pure forms: the appropri-
scan grade. Melting points were determined on a Buchi
¨

C. Cativiela et al. / Tetrahedron: Asymmetry 16 (2005) 2613–2623
2619
NHCOPh
Ph
NH2.HCl
Ph
NHZ
Ph
b
c
CN
H
CO2H
H
CO2H
H
98%
60%
(1R,2R)-4
(1R,2R)-7
(1R,2R)-10
a
rac-4
CN
H
CO2H
H
CO2H
H
b
c
NHCOPh
Ph
NH2.HCl
Ph
NHZ
Ph
98%
65%
(1S,2S)-4
(1S,2S)-7
(1S,2S)-10
NHCOPh
H
NH2.HCl
H
NHBoc
H
b
d
CN
Ph
CO2H
Ph
CO2H
Ph
95%
46%
(1R,2S )-6
(1R,2S)-8
(1R,2S)-9
a
rac-6
CN
Ph
CO2H
Ph
CO2H
Ph
b
d
NHCOPh
H
NH2.HCl
H
NHBoc
H
92%
48%
(1S,2R)-6
(1S,2R)-8
(1S,2R)-9
Scheme 3. Synthesis of enantiomerically pure c5Phe derivatives. Reaction conditions: (a) chiral HPLC resolution; (b) 12 M HCl, reflux; (c) (i)
TMSCl, CH2Cl2; (ii) (i-Pr)2EtN, Z-Cl; (d) (i) propylene oxide, EtOH; (ii) TMAH, acetonitrile, Boc2O.
SMP-20 apparatus and were not corrected. IR spectra
were registered on a Mattson Genesis FTIR spectropho-
tometer; mmax is given for the main absorption bands. 1H
and 13C NMR spectra were recorded on a Varian Unity-
300 or a Bruker ARX-300 instrument in CDCl3 or D2O,
using the residual solvent signal as the internal standard
([D6]acetone was used as an external reference for
the 13C spectra); chemical shifts (d) are expressed
in ppm and coupling constants (J) in Hertz. Optical
rotations were measured at room temperature using a
Perkin–Elmer 241 Polarimeter-C in a 10 cm cell of
1 mL capacity. Microanalyses were carried out on a
Perkin–Elmer 200 C, H, N, S analyser. High-resolution
mass spectra were obtained on a high-resolution VG-
autospectrometer. HPLC was carried out using a
Waters HPLC system equipped with a Waters 600-E
pump and a Waters 991 photodiode array detector.
Mixed 10-undecenoate/3,5-dimethylphenylcarbamate of
cellulose and 10-undecenoate/3,5-dimethylphenylcarba-
mate of amylose were prepared and linked to allylsilica
gel (Nucleosil 100-5, Machery-Nagel) according to a
previously described procedure22,26 to give CSP-1 and
CSP-2, respectively. These stationary phases were
packed into stainless-steel tubes by the slurry method.
The HPLC analytical assays were carried out on
150 · 4.6 mm ID columns containing these CSPs. All
analytical assays were performed using mixtures of n-
hexane/i-PrOH and n-hexane/i-PrOH/acetone as eluents
(flow rate 1 mL/min), with UV monitoring performed at
210 nm. The capacity (k0), selectivity (a) and resolution
(Rs) factors are defined as follows: k0 = (tr ꢀ t0)/t0,
a ¼ k01=k02, Rs = 1.18(t2 ꢀ t1)/(w2 + w1), where subscripts
1 and 2 refer to the first and second eluted enantiomer,
and w1 and w2 denote their half-height peak widths; t0 is
the dead time. The semipreparative HPLC resolution of
compounds 4 and 6 was carried out on a 150 · 20 mm
ID column filled with CSP-1. A mixture of n-hexane/i-
PrOH/acetone (95/3/2) was used as the eluent with a
flow rate of 18 mL/min. UV detection was performed
at 270 nm. Both the column loading capacity, Ws (de-
fined as the maximum sample mass that the column
can hold), and the optimum sample concentration were
calculated for the analytical 150 · 4.6 mm ID column by
injecting increasing amounts of sample at different con-
centrations. In this way, the capacity of the semiprepar-
ative column was established as 30 mg and the optimum
concentration of the sample as 300 mg/mL.
4.2. Strecker reaction on 2-phenylcyclopentanone.
Synthesis of cis-1-cyano-2-phenylcyclopentylamine
rac-1 and trans-1-cyano-2-phenylcyclopentylamine rac-2
To a solution of 2-phenylcyclopentanone (800 mg,
5.0 mmol) in a mixture of i-PrOH/30% NH4OH (5 mL/
25 mL) were added NH4Cl (669 mg, 12.5 mmol) and
NaCN (623 mg, 12.5 mmol). The mixture was stirred
for 10 days at room temperature. The organic solvent
was evaporated under reduced pressure and the liquid
residue extracted with CH2Cl2 (3 · 25 mL). The organic
layers were combined, dried over MgSO4 and the sol-
vent evaporated to yield a residue, which was purified
by flash column chromatography (eluent: hexane/EtOAc
8/2) to give the cis-isomer rac-1 (316 mg, 1.70 mmol,
34% yield) and the trans-isomer rac-2 (325 mg,
1.75 mmol, 35% yield) as oily products.
4.2.1. cis-1-Cyano-2-phenylcyclopentylamine rac-1. Rf
(CH2Cl2/EtOAc 8/2) = 0.7. IR (neat) 3373.4, 3314.4,
1
2223.0 cmꢀ1. H NMR (CDCl3, 300 MHz) d 1.24 (br
s, 2H), 1.84–2.08 (m, 4H), 2.27–2.38 (m, 2H), 3.35 (dd,

2620
C. Cativiela et al. / Tetrahedron: Asymmetry 16 (2005) 2613–2623
1H, J = 7.4, J = 11.8), 7.27–7.43 (m, 5H). 13C NMR
(CDCl3, 75 MHz) d 21.65, 27.46, 40.28, 55.37, 56.55,
124.33, 127.75, 128.52, 128.76, 136.08.
4.5. Resolution of benzamide rac-4: (1R,2R) and (1S,2S)-
N-(1-cyano-2-phenylcyclopentyl)benzamide
HPLC resolution of cis racemate rac-4 (1.0 g) dissolved
in CHCl3 (3.3 mL) was carried out by successive injec-
tions of 0.1 mL on a 150 · 20 mm ID column filled with
mixed 10-undecenoate/3,5-dimethylphenylcarbamate of
cellulose bonded on allylsilica gel (CSP-1) and using a
mixture of n-hexane/i-PrOH/acetone 95/3/2 as the eluent
(flow rate: 18 mL/min). A total of 35 injections were re-
quired, with one injection performing every 10 min.
Four separate fractions were collected. The first, second,
third and fourth fractions contained, respectively, 100/0
(440 mg), 85/15 (45 mg), 15/85 (150 mg) and 4/96
(350 mg) mixtures of (1R,2R)-4 and (1S,2S)-4. The com-
bined third and fourth fractions were reinjected in a way
similar to that described above and led to 315 mg of
enantiomerically pure (1S,2S)-4. Spectroscopic data of
both enantiomers are the same as those described for
rac-4.
4.2.2. trans-1-Cyano-2-phenylcyclopentylamine rac-
2. Rf (CH2Cl2/EtOAc 8/2) = 0.5. IR (neat) 3350–
3310, 2220.2 cmꢀ1 1H NMR (CDCl3, 300 MHz) d
.
1.89–2.26 (m, 7H), 2.37–2.43 (m, 1H), 2.98 (dd, 1H,
J = 8.0, J = 11.5), 7.30–7.40 (m, 5H). 13C NMR (CDCl3,
75 MHz) d 20.26, 28.75, 39.83, 56.77, 61.12, 122.55,
127.69, 128.02, 128.41, 137.20.
4.3. cis-N-(1-Cyano-2-phenylcyclopentyl)acetamide
(rac-3)
Under an inert atmosphere, Et3N (304 mg, 0.416 mL,
3.0 mmol) and acetyl chloride (118 mg, 0.106 mL,
1.5 mmol) were added to an ice-cooled solution of cis-
amino nitrile rac-1 (186 mg, 1.0 mmol) in dry CH2Cl2
(10 mL). The reaction mixture was stirred at room tem-
perature overnight. The reaction mixture was succes-
sively washed with aqueous solutions of 5% NaHCO3,
2 M H2SO4 and saturated NaCl. The organic phase
was dried over MgSO4 and filtered. Evaporation of the
solvent yielded a residue, which was purified by flash
column chromatography (eluent: hexane/EtOAc 8/2) to
afford acetamide derivative rac-3 as a white solid
(171 mg, 0.75 mmol, 75% yield). Mp 155–156 ꢁC
(EtOAc/hexane). Rf (CH2Cl2/EtOAc 8/2) = 0.6. IR
(1R,2R)-4: Mp 147 ꢁC (EtOAc/hexane). [a]D = ꢀ13.9 (c
1.02, CHCl3).
(1S,2S)-4: Mp 147 ꢁC (EtOAc/hexane). [a]D = +13.3 (c
1.03, CHCl3).
4.6. cis-c5Phe hydrochloride rac-7
1
(Nujol) 3392.4–3209.2, 2234.7, 1676.3, 1521.1 cmꢀ1. H
A solution of cis-acetamide rac-3 (228 mg, 1.0 mmol) or
cis benzamide rac-4 (290 mg, 1.0 mmol) in 12 M hydro-
chloric acid (5 mL) was heated under reflux for 1 day.
The solvent was evaporated and the resulting solid was
partitioned between diethyl ether and water. The aque-
ous phase was washed with several additional portions
of diethyl ether and then lyophilised to give 230 mg of
the hydrochloride rac-7 (0.95 mmol, 95% yield). Mp
NMR (CDCl3, 300 MHz) d 1.80 (s, 3H), 1.92–2.04 (m,
2H), 2.10–2.30 (m, 2H), 2.50–2.64 (m, 2H), 3.75 (dd,
1H, J = 7.7, J = 9.9), 5.24 (br s, 1H), 7.34–7.46 (m,
5H). 13C NMR (CDCl3, 75 MHz) d 21.73, 23.12,
28.27, 38.06, 53.85, 57.06, 120.93, 128.33, 129.17,
135.99, 169.73. Anal. Calcd for C14H16N2O: C, 73.66;
H, 7.06; N, 12.27. Found: C, 73.54; H, 6.95; N,
12.17.
1
245–250 ꢁC (dec). IR (Nujol) 3194.3, 1744.8 cmꢀ1. H
NMR (D2O, 300 MHz) d 1.84–2.23 (m, 5H), 2.49–2.58
(m, 1H), 3.78 (dd, 1H, J = 7.5, J = 12.3), 7.20–7.23 (m,
2H), 7.28–7.37 (m, 3H). 13C NMR (D2O, 75 MHz) d
22.66, 28.41, 36.31, 52.74, 68.87, 128.32, 128.57,
129.38, 134.92, 174.82.
4.4. cis-N-(1-Cyano-2-phenylcyclopentyl)benzamide rac-4
Under an inert atmosphere, Et3N (1.515 g, 2.1 mL,
15.0 mmol) and benzoyl chloride (1.054 g, 0.808 mL,
7.5 mmol) were added to an ice-cooled solution of cis-
amino nitrile rac-1 (930 mg, 5.0 mmol) in dry CH2Cl2
(30 mL). The reaction mixture was stirred at room tem-
perature overnight. The reaction mixture was succes-
sively washed with aqueous solutions of 5% NaHCO3,
2 N H2SO4 and saturated NaCl. The organic phase
was dried over MgSO4 and filtered. Evaporation of the
solvent yielded a residue, which was purified by flash
column chromatography (eluent: hexane/EtOAc 8/2) to
afford the benzamide derivative rac-4 as a white solid
(1.060 g, 3.65 mmol, 73% yield). Mp 142 ꢁC (EtOAc/
hexane). Rf (hexane/EtOAc 8/2) = 0.2. IR (Nujol)
4.6.1. (1R,2R)-c5Phe hydrochloride (1R,2R)-7. An
identical procedure to that described above was applied
to transform (1R,2R)-4 (261 mg, 0.9 mmol) into
(1R,2R)-7 as a white solid (213 mg, 0.88 mmol, 98%
yield). Mp 242 ꢁC (dec). [a]D = ꢀ40.2 (c 1.18, 6 M
HCl/H2O). Spectroscopic data are the same as those
described for rac-7.
4.6.2. (1S,2S)-c5Phe hydrochloride (1S,2S)-7. In a sim-
ilar way to that described above, starting from (1S,2S)-4
(232 mg, 0.8 mmol), (1S,2S)-7 was obtained as a white
solid (189 mg, 0.78 mmol, 98% yield). Mp 241 ꢁC
(dec). [a]D = +41.2 (c 1.15, 6 M HCl/H2O). Spectro-
scopic data are the same as those described for rac-7.
1
3387.8, 2237.3, 1649.5, 1514.3 cmꢀ1. H NMR (CDCl3,
300 MHz) d 1.95–2.05 (m, 2H), 2.20–2.32 (m, 2H),
2.64–2.69 (m, 2H), 3.85 (dd, 1H, J = 8.1, J = 9.6), 5.99
(br s, 1H), 7.24–7.48 (m, 5H). 13C NMR (CDCl3,
75 MHz) d 22.27, 28.35, 38.47, 54.44, 56.92, 120.94,
126.68, 128.31, 128.54, 128.62, 129.37, 132.01, 133.07,
135.73, 166.56. Anal. Calcd for C19H18N2O: C, 78.59;
H, 6.25; N, 9.65. Found: C, 78.23; H, 6.18; N, 9.20.
4.7. cis-Z-c5Phe-OH rac-10
Under an inert atmosphere, chlorotrimethylsilane
(272 mg, 0.324 mL, 2.5 mmol) was added to an ice-cooled

C. Cativiela et al. / Tetrahedron: Asymmetry 16 (2005) 2613–2623
2621
suspension of hydrochloride rac-7 (242 mg, 1.0 mmol) in
anhydrous CH2Cl2 (7 mL). The reaction was stirred at
0 ꢁC for 10 min and then heated under reflux for 1 h. After
cooling to 0 ꢁC, benzyl chloroformate (216 mg, 0.210 mL,
1.4 mmol) and N,N-diisopropylethylamine (323 mg,
0.438 mL, 2.5 mmol) were added and the reaction stirred
at room temperature for 1 day. The solvent was evapo-
rated to dryness and the residue taken up in saturated
aqueous NaHCO3 (5 mL). Diethyl ether (15 mL) and
water (10 mL) were added and after stirring for 10 min,
the layers were separated. The aqueous phase was washed
with an additional portion of ether, acidified by the addi-
tion of a solution of KHSO4, and then extracted several
times with CH2Cl2. After drying and filtering, removal
of the solvent afforded rac-10 as a white solid. Additional
purification was carried out by flash column chromato-
graphy (eluent: hexane/EtOAc 8/2) to afford the N-Z
derivative rac-10 (193 mg, 0.57 mmol, 57% yield). Mp
133 ꢁC (EtOAc/hexane). Rf (CH2Cl2/EtOAc 8/2) = 0.3.
5H), 2.93–3.00 (m, 1H), 3.29 (dd, 1H, J = 8.1,
J = 10.3), 5.83 (br s, 1H), 7.34–7.42 (m, 5H). 13C
NMR (CDCl3, 75 MHz) d 20.77, 23.32, 28.46, 38.20,
54.38, 61.11, 118.83, 128.39, 128.43, 129.02, 136.79,
169.94. Anal. Calcd for C14H16N2O: C, 73.66; H, 7.06;
N, 12.27. Found: C, 73.45; H, 6.93; N, 12.10.
4.9. trans-N-(1-Cyano-2-phenylcyclopentyl)benzamide
(rac-6)
Under an inert atmosphere, Et3N (1.515 g, 2.1 mL,
15.0 mmol) and benzoyl chloride (1.054 g, 0.808 mL,
7.5 mmol) were added to an ice-cooled solution of
trans-amino nitrile rac-2 (930 mg, 5.0 mmol) in dry
CH2Cl2 (30 mL). The reaction mixture was stirred at
room temperature overnight and then successively
washed with aqueous solutions of 5% NaHCO3, 2 M
H2SO4 and saturated NaCl. The organic phase was
dried over MgSO4 and filtered. Evaporation of the
solvent yielded a residue, which was purified by flash
column chromatography (eluent: hexane/EtOAc 8/2) to
afford the trans benzamide derivative rac-6 as a white
solid (1.278 g, 4.40 mmol, 88% yield). Mp 122 ꢁC
(EtOAc/hexane). Rf (CH2Cl2/EtOAc 8/2) = 0.2. IR (Nu-
IR (Nujol) 3396.4, 3140.2, 1728.0, 1689.1 cmꢀ1 1H
.
NMR (CDCl3, 300 MHz, 60 ꢁC) d 1.87–2.04 (m, 2H),
2.15–2.24 (m, 2H), 2.49–2.69 (m, 2H), 3.79 (dd, 2H,
J = 7.8, J = 10.7), 4.80 (br s, 1H), 4.98 (d, 1H, J = 12.5),
5.05 (d, 1H, J = 12.1), 7.25–7.37 (m, 10H), 9.16 (br s,
1H). 13C NMR (CDCl3, 75 MHz, 60 ꢁC) d 23.11, 30.21,
37.08, 54.25, 66.78, 67.68, 127.64, 127.80, 128.04,
128.25, 128.40, 128.56, 128.88, 136.12, 136.98, 156.21,
179.53. Anal. Calcd for C20H21NO4: C, 70.78; H, 6.24;
N, 4.13. Found: C, 70.95; H, 6.18; N, 3.92.
jol) 3271.0–3245.9, 1638.7 cmꢀ1 1H NMR (CDCl3,
.
300 MHz) d 2.08–2.36 (m, 5H), 2.99–3.08 (m, 1H),
3.45 (dd, 1H, J = 8.1, J = 10.3), 6.56 (br s, 1H), 7.31–
7.49 (m, 8H), 7.64–7.67 (m, 2H). 13C NMR (CDCl3,
75 MHz) d 20.85, 28.24, 38.17, 54.48, 61.24, 118.74,
126.99, 128.24, 128.37, 128.60, 129.00, 132.10, 133.06,
136.83, 167.11. Anal. Calcd for C19H18N2O: C, 78.59;
H, 6.25; N, 9.65. Found: C, 78.45; H, 6.15; N, 9.79.
4.7.1. (1R,2R)-N-Z-c5Phe-OH (1R,2R)-10. An identi-
cal procedure to that described above was applied to
transform (1R,2R)-7 (242 mg, 1.0 mmol) into (1R,2R)-
10 as an oil (204 mg, 0.60 mmol, 60% yield). [a]D =
ꢀ32.5 (c 0.50, CHCl3). Spectroscopic data are the same
as those described for rac-10.
4.10. Resolution of benzamide rac-6: (1R,2S) and
(1S,2R)-N-(1-cyano-2-phenylcyclopentyl)benzamide
HPLC resolution of trans racemate rac-6 (1.0 g) dis-
solved in CHCl3 (3.3 mL) was carried out by successive
injections of 0.1 mL on a 150 · 20 mm ID column filled
with mixed 10-undecenoate/3,5-dimethylphenylcarba-
mate of cellulose bonded on allylsilica gel (CSP-1) and
using a mixture of n-hexane/i-PrOH/acetone 95/3/2 as
the eluent (flow rate: 18 mL/min). A total of 35 injec-
tions were required, with one injection performing every
12 min. Three separate fractions were collected. The
first, second and third fractions contained, respectively,
100/0 (440 mg), 90/10 (25 mg) and 9/91 (525 mg) mix-
tures of (1R,2S)-6 and (1S,2R)-6. Reinjection of the
third fraction in a similar way to that described above
led to 330 mg of enantiomerically pure (1S,2R)-6.
4.7.2. (1S,2S)-N-Z-c5Phe-OH (1S,2S)-10. In a similar
way to that described above, starting from (1S,2S)-7
(218 mg, 0.9 mmol), (1S,2S)-10 was obtained as an oil
(199 mg, 0.59 mmol, 65% yield). [a]D = +30.0 (c 0.78,
CHCl3). Spectroscopic data are the same as those
described for rac-10.
4.8. trans-N-(1-Cyano-2-phenylcyclopentyl)acetamide
(rac-5)
Under an inert atmosphere, Et3N (304 mg, 0.416 mL,
3.0 mmol) and acetyl chloride (118 mg, 0.106 mL,
1.5 mmol) were added to an ice-cooled solution of
trans-amino nitrile rac-2 (186 mg, 1.0 mmol) in dry
CH2Cl2 (10 mL). The reaction mixture was stirred at
room temperature overnight. The reaction mixture was
successively washed with aqueous solutions of 5% Na-
HCO3, 2 M H2SO4 and saturated NaCl. The organic
phase was dried over MgSO4 and filtered. Evaporation
of the solvent yielded a residue, which was purified by
flash column chromatography (eluent: hexane/EtOAc
8/2) to afford the trans acetamide derivative rac-5 as a
white solid (171 mg, 0.75 mmol, 75% yield). Mp
134 ꢁC (EtOAc/hexane). Rf (CH2Cl2/EtOAc 8/2) = 0.4.
(1R,2S)-6: Mp 120 ꢁC (EtOAc/hexane). [a]D = ꢀ90.1 (c
1.02, CHCl3).
(1S,2R)-6: Mp 118 ꢁC (EtOAc/hexane). [a]D = +88.2 (c
1.00, CHCl3).
4.11. trans-c5Phe hydrochloride rac-8
A solution of trans acetamide rac-5 (228 mg, 1.0 mmol)
or trans benzamide rac-6 (290 mg, 1.0 mmol) in 12 M
hydrochloric acid (5 mL) was heated under reflux for
1 day. The solvent was evaporated and the resulting
IR (Nujol) 3244.5, 2237.8, 1647.35, 1529.8 cmꢀ1
.
1H
NMR (CDCl3, 300 MHz) d 1.96 (s, 3H), 1.98–2.27 (m,

2622
C. Cativiela et al. / Tetrahedron: Asymmetry 16 (2005) 2613–2623
solid was partitioned between CH2Cl2 and water. The
aqueous phase was washed with several additional por-
tions of CH2Cl2 and then lyophilised to give 239 mg of
rac-8 (0.99 mmol, 99% yield). Mp 254–256 ꢁC (dec).
4.12.1. (1R,2S)-N-Boc-c5Phe-OH (1R,2S)-9. In a simi-
lar way to that described above, starting from (1R,2S)-8
(193 mg, 0.8 mmol), (1R,2S)-9 was obtained as a white
solid (112 mg, 0.37 mmol, 46% yield). Mp 164 ꢁC
(EtOAc/hexane). [a]D = ꢀ75.8 (c 0.91, CHCl3). Spectro-
scopic data are the same as those described for rac-9.
1
IR (Nujol) 3325–2250, 1749.2, 1734.8 cmꢀ1. H NMR
(D2O, 300 MHz) d 1.94–2.18 (m, 4H), 2.34–2.44 (m,
1H), 2.56–2.64 (m, 1H), 3.43 (dd, 1H, J = 7.1,
J = 12.1), 7.31–7.40 (m, 5H). 13C NMR (D2O,
75 MHz) d 22.38, 30.56, 34.39, 55.18, 69.64, 128.30,
128.88, 136.24, 173.07.
4.12.2. (1S,2R)-N-Boc-c5Phe-OH (1S,2R)-9. An iden-
tical procedure to that described above was applied to
transform (1S,2R)-8 (145 mg, 0.6 mmol) into (1S,2R)-9
as a white solid (88 mg, 0.29 mmol, 48% yield). Mp
164 ꢁC (EtOAc/hexane). [a]D = +74.6 (c 0.98, CHCl3).
Spectroscopic data are the same as those described for
rac-9.
4.11.1. (1R,2S)-c5Phe hydrochloride (1R,2S)-8. In a
similar way to that described above, starting from
(1R,2S)-6 (261 mg, 0.9 mmol), (1R,2S)-8 was obtained
as a white solid (207 mg, 0.86 mmol, 95% yield). Mp
253–255 ꢁC (dec). [a]D = +9.9 (c 1.02, 6 M HCl/H2O).
Spectroscopic data are the same as those described for
rac-8.
4.13. trans-Z-c5Phe-OH rac-11
Under an inert atmosphere, chlorotrimethylsilane
(103 mg, 0.097 mL, 0.8 mmol) was added to an ice-
cooled suspension of hydrochloride rac-8 (73 mg,
0.3 mmol) in anhydrous CH2Cl2 (5 mL). The reaction
was stirred at 0 ꢁC for 10 min and then heated under
reflux for 1 h. After cooling to 0 ꢁC, benzyl chloroformate
(60 mg, 0.059 mL, 0.4 mmol) and N,N-diisopropylethyl-
amine (97 mg, 0.131 mL, 0.8 mmol) were added and the
reaction stirred at room temperature for 1 day. The sol-
vent was evaporated to dryness and the residue was
taken up in a saturated solution of NaHCO3 (5 mL).
Diethyl ether (15 mL) and water (10 mL) were added
and after stirring for 10 min, the layers were separated.
The aqueous phase was washed with an additional por-
tion of ether, acidified by the addition of a solution of
KHSO4, and then extracted several times with CH2Cl2.
After drying and filtering, removal of the solvent afforded
a residue that was purified by flash column chromato-
graphy (eluent: hexane/EtOAc 8/2) to afford the N-Z
derivative rac-11 (27 mg, 0.08 mmol, 27% yield). Rf
(CH2Cl2/EtOAc 8/2) = 0.5. IR (Nujol) 3241.1, 1694.8,
4.11.2. (1S,2R)-c5Phe hydrochloride (1S,2R)-8. An
identical procedure to that described above was applied
to transform (1S,2R)-6 (203 mg, 0.7 mmol) into (1S,2R)-
8 as a white solid (156 mg, 0.64 mmol, 92% yield). Mp
251–253 ꢁC (dec). [a]D = ꢀ9.9 (c 1.02, 6 M HCl/H2O).
Spectroscopic data are the same as those described for
rac-8.
4.12. trans-Boc-c5Phe-OH rac-9
A
solution of trans-c5Phe hydrochloride rac-8
(1.0 mmol, 242 mg) in ethanol (10 mL) and propylene
oxide (3 mL) was heated under reflux for 1 h. The sol-
vent was evaporated under reduced pressure. In order
to complete the removal of the residual propylene
oxide and the by-products of the reaction, Et2O was
added and then removed under vacuum (3 · 5 mL).
The resulting solid was suspended in acetonitrile
(20 mL) and TMAH (tetramethylammonium hydrox-
ide) was added (1 mmol, 181 mg). The mixture was
stirred at 40–50 ꢁC until a solution was formed. Once
complete dissolution was achieved, Boc2O (1.5 mmol,
327 mg) was added at room temperature and the mix-
ture was stirred for 2 days. On the third day, a further
1.0 mmol of Boc2O (218 mg) was added and stirred for
a further 1 day. The acetonitrile was removed under
vacuum and the residue was partitioned between
H2O and Et2O. The aqueous layer was acidified with
solid citric acid to pH 3–4 and then extracted with
EtOAc. Concentration of the organic layer resulted
in the precipitation of a solid, a process that was com-
pleted by adding a portion of hexane. The product was
recrystallised (EtOAc/hexane) to give 229 mg of a
white solid (0.75 mmol, 75% yield). Mp 188 ꢁC
(EtOAc/hexane). Rf (CH2Cl2/EtOAc 6/4) = 0.7. IR
1
1641.3 cmꢀ1. H NMR (CDCl3, 300 MHz) d 2.01–2.56
(m, 6H), 3.71 (br s, 1H), 5.07 (d, 1H, J = 12.1), 5.14
(d, 1H, J = 12.1), 5.74 (br s, 1H), 7.11–7.35 (m, 10H),
8.75 (br s, 1H). 13C NMR (CDCl3, 75 MHz) d 23.11,
30.21, 35.48, 53.83, 66.69, 70.28, 127.39, 127.85,
128.08, 128.19, 128.35, 128.53, 136.31, 137.95, 154.97,
177.96. Anal. Calcd for C20H21NO4: C, 70.78; H, 6.24;
N, 4.13. Found: C, 70.89; H, 6.22; N, 3.98.
Acknowledgements
Authors wish to express their gratitude to Dr. J. I. Gar-
´
cıa for the molecular modelling study. Likewise, authors
are grateful to A. L. Bernad for her assistance with the
HPLC. This work was carried out with the financial sup-
1
(Nujol) 3296.7–3251.4, 1696.1, 1638.2 cmꢀ1. H NMR
´
port of Ministerio de Educacion y Ciencia and FEDER
(project CTQ2004-5358). M. Lasa would like to thank
MEC for a FPU grant.
(CDCl3, 300 MHz, 60 ꢁC) d 1.74 (s, 9H), 1.97–2.18
(m, 3H), 2.24–2.38 (m, 2H), 2.58–2.69 (m, 1H), 3.53–
3.65 (m, 1H), 5.35 (br s, 1H), 7.19–7.31 (m, 5H). 13C
NMR (CDCl3, 75 MHz, 60 ꢁC) d 22.85, 28.39, 30.20,
35.94, 54.18, 70.43, 80.29, 127.29, 128.05, 128.32,
138.48, 155.19, 177.84. Anal. Calcd for C17H23NO4:
C, 66.86; H, 7.59; N, 4.59. Found: C, 66.75; H, 7.46;
N, 4.49.
References
1. (a) Kaul, R.; Balaram, P. Bioorg. Med. Chem. 1999, 7,
105–117; (b) Venkatraman, J.; Shankaramma, S. C.;
Balaram, P. Chem. Rev. 2001, 101, 3131–3152; (c) Toniolo,

C. Cativiela et al. / Tetrahedron: Asymmetry 16 (2005) 2613–2623
2623
C.; Crisma, M.; Formaggio, F.; Peggion, C. Biopolymers
2001, 60, 396–419; (d) Cowell, S. M.; Lee, Y. S.; Cain, J.
P.; Hruby, V. J. Curr. Med. Chem. 2004, 11, 2785–
2798.
G.; Bourrat, E.; Bertho, G.; Pin, J.-P.; Acher, F. C. Amino
Acids 2003, 24, 303–310.
16. (a) Curry, K.; Peet, M. J.; Magnuson, D. S. K.; McLen-
nan, H. J. Med. Chem. 1988, 31, 864–867; (b) Acher, F. C.;
Tellier, F. J.; Azerad, R.; Brabet, I. N.; Fagni, L.; Pin,
2. (a) Hruby, V. J.; Li, G.; Haskell-Luevano, C.; Shender-
ovich, M. Biopolymers 1997, 43, 219–266; (b) Hruby, V. J.;
Balse, P. M. Curr. Med. Chem. 2000, 7, 945–970; (c)
Hruby, V. J. Acc. Chem. Res. 2001, 34, 389–397.
3. Gibson, S. E.; Guillo, N.; Tozer, M. J. Tetrahedron 1999,
55, 585–615.
´
J.-P. R. J. Med. Chem. 1997, 40, 3119–3129; (c) Domın-
guez, C.; Ezquerra, J.; Prieto, L.; Espada, M.; Pedregal, C.
Tetrahedron: Asymmetry 1997, 8, 511–514.
17. (a) Ma, D.; Tian, H.; Zou, G. J. Org. Chem. 1999, 64, 120–
125; (b) Wede, J.; Volk, F.-J.; Frahm, A. W. Tetrahedron:
Asymmetry 2000, 11, 3231–3252; (c) Namba, K.; Kawasaki,
M.; Takada, I.; Iwama, S.; Izumida, M.; Shinada, T.;
Ohfune, Y. Tetrahedron Lett. 2001, 42, 3733–3736; (d)
Meyer, U.; Breiting, E.; Bisel, P.; Frahm, A. W. Tetra-
hedron: Asymmetry 2004, 15, 2029–2037; (e) Johnson, R. L.;
Rao, K. S. S. P. Bioorg. Med. Chem. Lett. 2005, 15, 57–60.
18. (a) Pihko, P. M.; Pohjakallio, A. Synlett 2004, 12, 2115–
2118; (b) Saaby, S.; Bella, M.; Jørgensen, K. A. J. Am.
Chem. Soc. 2004, 126, 8120–8121.
´
4. Cativiela, C.; Dıaz-de-Villegas, M. D.; Jimenez, A. I.;
Lopez, P.; Marraud, M.; Oliveros, L. Chirality 1999, 11,
´
´
583–590.
5. (a) Cativiela, C.; Avenoza, A.; Parıs, M.; Peregrina, J. M.
´
J. Org. Chem. 1994, 59, 7774–7778; (b) Alıas, M.;
´
Cativiela, C.; Jimenez, A. I.; Lopez, P.; Oliveros, L.;
´
Marraud, M. Chirality 2001, 13, 48–55; (c) Cativiela, C.;
´
´
Lopez, P.; Lasa, M. Eur. J. Org. Chem. 2004, 3898–3908.
6. Burckhalter, J. H.; Schmied, G. J. Pharm. Sci. 1966, 55,
443–445.
19. Sifton, W.; Stothers, J. B.; Thomas, S. E. Can. J. Chem.
1992, 70, 1274–1280.
7. (a) Ma, D.; Ma, J.; Dai, L. Tetrahedron: Asymmetry 1997,
8, 825–827; (b) Kotha, S.; Brahmachary, E.; Sreenivas-
achary, N. Tetrahedron Lett. 1998, 39, 4095–4098.
8. (a) Hammer, K.; Undheim, K. Tetrahedron 1997, 53,
2309–2322; (b) Hammer, K.; Undheim, K. Tetrahedron
1997, 53, 5925–5936; (c) Hammer, K.; Undheim, K.
Tetrahedron 1997, 53, 10603–10614; (d) Møller, B.; Und-
heim, K. Tetrahedron 1998, 543, 5789–5804; (e) Hammer,
K.; Undheim, K. Tetrahedron: Asymmetry 1998, 9, 2359–
2368; (f) Hammer, K.; Wang, J.; Falck-Pedersen, M. L.;
Rømming, C.; Undheim, K. J. Chem. Soc., Perkin Trans.
1 2000, 1691–1695; (g) Undheim, K.; Efskind, J. Tetra-
hedron 2000, 56, 4847–4857.
9. (a) Kummeter, M.; Kazmaier, U. Eur. J. Org. Chem. 2003,
3325–3329; (b) Belokon, Y. N.; Bhave, D.; DÕAddario, D.;
Groaz, E.; North, M.; Tagliazucca, V. Tetrahedron 2004,
60, 1849–1861.
10. Trost, B. M.; Dogra, K. J. Am. Chem. Soc. 2002, 124,
7256–7257.
11. Møller, B.; Undheim, K. Eur. J. Org. Chem. 2003, 332–
336.
12. (a) Gabaitsekgosi, R.; Hayes, C. J. Tetrahedron Lett. 1999,
40, 7713–7716; (b) Bradley, D. M.; Mapitse, R.; Thomson,
N. M.; Hayes, C. J. J. Org. Chem. 2002, 67, 7613–
7617.
13. Battistini, L.; Curti, C.; Zanardi, F.; Rassu, G.; Auzzas,
G.; Casiraghi, G. J. Org. Chem. 2004, 69, 2611–2613.
14. Andrei, M.; Undheim, K. Tetrahedron: Asymmetry 2004,
15, 53–63.
20. Francotte, E. R. J. Chromatogr., A 2001, 906, 379–397.
21. (a) Dingenen, J. In A Practical Approach to Chiral
Separations by Liquid Chromatography; Subramanian,
G., Ed.; VCH Verlagsgesellschaft: Weinheim, 1994; pp
115–181; (b) Yashima, E.; Yamamoto, C.; Okamoto, Y.
Synlett 1998, 344–360; (c) Yamamoto, C.; Okamoto, Y.
Bull. Chem. Soc. Jpn. 2004, 77, 227–257.
´
22. (a) Minguillon, C.; Franco, P.; Oliveros, L.; Lopez, P. J.
Chromatogr., A 1996, 728, 407–414; (b) Oliveros, L.;
´
Lopez, P.; Minguillon, C.; Franco, P. J. Liq. Chromatogr.
1995, 18, 1521–1532.
23. See references to the work of L. Oliveros and C.
´
Minguillon included in the following review: Franco, P.;
´
Senso, A.; Oliveros, L.; Minguillon, C. J. Chromatogr., A
2001, 906, 155–170.
´
´
24. (a) Cativiela, C.; Dıaz-de-Villegas, M. D.; Jimenez, A. I.;
´
Lopez, P.; Marraud, M.; Oliveros, L. Chirality 1999, 11,
´
583–590; (b) Jimenez, A. I.; Lopez, P.; Oliveros, L.;
Cativiela, C. Tetrahedron 2001, 57, 6019–6026; (c) Royo,
´
´
´
S.; Lopez, P.; Jimenez, A. I.; Oliveros, L.; Cativiela, C.
Chirality 2002, 14, 39–46.
´
25. (a) Gil, A. M.; Bun˜uel, E.; Lopez, P.; Cativiela, C.
Tetrahedron: Asymmetry 2004, 15, 811–819; (b) Jimenez,
´
´
A. I.; Lopez, P.; Cativiela, C. Chirality 2005, 17, 22–29; (c)
´
Gil, A. M.; Lopez, P.; Bun˜uel, E.; Cativiela, C. Arkivoc
´
2005, ix, 90–103; (d) Royo, E.; Lopez, P.; Cativiela, C.
Arkivoc 2005, vi, 46–61.
26. Franco, P.; Senso, A.; Minguillon, C.; Oliveros, L.
´
15. (a) Nicole, L.; Berlinguet, L. Can. J. Chem. 1962, 40, 363–
366; (b) Curry, K.; McLennan, H.; Rettig, S. J.; Trotter, J.
Can. J. Chem. 1993, 71, 76–83; (c) Bessis, A.-S.; Vadesne,
J. Chromatogr., A 1998, 796, 265–272.
27. Lasa, M. Ph.D. Universidad de Zaragoza, 2004. Results
unpublished.