1
096
Z. Szeleczky et al. / Tetrahedron: Asymmetry 25 (2014) 1095–1099
hetero- or homochiral interaction is dominant, while the ee
E
value
E
compounds with eutectic compositions (ee ) of 38%, 10%, and
may also show which one of the two diastereomeric associates has
higher complementarity.
80%, respectively.
How the complementarity between the racemic compound and
the resolving agent affects the efficiency of those resolutions when
the enantiomers of racemic amino acid derivatives were separated
using resolving agents with a structure related to the correspond-
ing racemic compound has previously been investigated. If the
racemic compound and the structurally related resolving agent
were reacted in a ratio of 1:1, the mixture obtained may be
regarded as a quasi-enantiomeric mixture with a quasi-enantio-
meric excess of 50%. It was established that the enantiomeric
purity of the enantiomeric mixtures obtained from the correspond-
2.2. The resolution of mandelic acid MA or racemic 2-chloro-
mandelic acid CMA with (R)-(ꢀ)-3-(aminomethyl)-5-methyl-
hexanoic acid (R)-PREG
The resolution of the racemic mandelic acid MA or the 2-chlo-
romandelic acid CMA was carried out using (R)-Pregabalin PREG
as the resolving agent. Mandelic acid MA or 2-chloromandelic acid
CMA was mixed with 0.5 equiv of (R)-Pregabalin (PREG) and
0.25 equiv of sodium carbonate as the achiral auxiliary24 and the
mixture obtained was dissolved in hot water. The reaction mixture
was allowed to cool to 26 °C whereupon the crystalline diastereo-
meric salt (R)-MAÁ(R)-PREG or (R)-CMAÁ(R)-PREG appeared and
this was separated from the mother liquor appeared after 15 min
or 168 h, respectively, (Scheme 1 and Table 2).
ing crystalline diastereomers (ee
eutectic composition of the racemic amino acid derivatives
ee when structurally related resolving agents were
ffi ee
used. Similar trends were observed in those resolutions when
D
) was in good agreement with the
(
D
E
)
2
1,22
the enantiomers of the racemic compound were separated using
In order to obtain the enantiomeric mixtures of MA and
CMA, the corresponding diastereomer (R)-MAÁ(R)-PREG or
(R)-CMAÁ(R)-PREG, respectively, was treated with aqueous ammo-
nia, whereupon the (R)-Pregabalin PREG precipitated. Next, hydro-
chloric acid was added to the mother liquor and the crystals of MA
or CMA were collected after 2 h of crystallization. The results are
summarized in Table 2.
a structurally non-related resolving agent.2
3
In continuation of this, we wished to investigate in more detail
those resolutions when the racemic compound was reacted with a
resolving agent having a non-related structure. Hence herein, race-
mic mandelic acid MA and racemic 2-chloromandelic acid CMA
were chosen as model compounds and were resolved using the
structurally unrelated (R)-(ꢀ)-3-(aminomethyl)-5-methylhexanoic
acid PREG (Pregabalin) as the resolving agent (Fig. 1). In order to
elaborate upon a resolution procedure for the mandelic acid deriv-
atives MA and CMA, our aim was to study how the eutectic compo-
In the resolution experiments of mandelic acid MA or 2-chloro-
mandelic acid CMA with (R)-Pregabalin PREG, the enantiomeric
excess (ee
80% and F = 0.45 in the case of MA, or ee
CMA after 15 min (Table 2, entries 1 and 3). These values decreased
to ee = 62% and F = 0.43 for MA or ee = 29% and F = 0.36 for CMA
D
) and the resolving capability values (F) were ee
D
=
D
= 92% and F = 0.49 for
sition (ee
agent PREG influenced the efficiency of the resolution, as well as
the purity of the enantiomeric mixtures (ee ) of the mandelic acid
derivatives MA or CMA under kinetic or thermodynamic control.
E
) of the racemic compound MA or CMA or the resolving
D
D
D
when the crystallization time was 168 h (compare Table 2, entries
1 and 2 or 3 and 4).
Based on the data shown in Table 2, it can be concluded
that kinetic control governed the initial formation of the
(
R)-MAÁ(R)-PREG and (R)-CMAÁ(R)-PREG diastereomers leading to
good enantiomeric separation of MA or CMA after 15 min of crys-
tallization. The decrease in the enantiomeric excess (ee ) and in
OH
OH
NH2
COOH
CH3
COOH
COOH
D
the resolving capability values (F) over time may be explained
by the effect of thermodynamic control, i.e., the formation of
H3C
Cl
(
S)-MAÁ(R)-PREG or (S)-CMAÁ(R)-PREG diastereomers became
MA
CMA
(
R)-(-)-PREG
more favorable after 168 h of crystallization than at the initial
stages of the crystallization.
Figure 1. Mandelic acid MA, 2-chloromandelic acid CMA, and (R)-(ꢀ)-3-(amino-
methyl)-5-methylhexanoic acid PREG used herein.
The behavior of the enantiomeric mixtures of the corresponding
racemic compound MA or CMA and the resolving agent PREG may
be the underlying phenomenon of these kinetic effects observed
during the resolution of MA or CMA with (R)-PREG. As it was
shown in Section 2.1, mandelic acid MA, 2-chloromandelic acid
CMA, and Pregabalin PREG are all racemate-forming compounds
(Table 1 and Fig. 2). This behavior indicates that in the case of
enantiomeric mixtures with an enantiomeric purity above the
eutectic composition, the equilibrium in the solution is displaced
toward the formation of the corresponding homochiral supramo-
lecular associates, which may help the initial crystallization of
the corresponding diastereomeric salts during the resolution of
MA or CMA with (R)-PREG leading to high enantiomeric excess
and resolving capability values after 15 min of crystallization
(Table 2, entries 1 and 3). This hypothesis can also be verified by
the fact that the difference between the corresponding eutectic
2
2
. Results and discussion
.1. The behavior of enantiomeric mixtures of mandelic acid
MA, 2-chloromandelic acid CMA, and 3-(aminomethyl)-5-
methylhexanoic acid PREG
In order to investigate the behavior of the enantiomeric
mixtures of mandelic acid MA, 2-chloromandelic acid CMA, and
3
-(aminomethyl)-5-methylhexanoic acid PREG, racemic and enan-
tiopure MA, CMA, or PREG were mixed to obtain the corresponding
enantiomeric mixtures as detailed in Table 1, and were dissolved in
the corresponding hot solvent. Water was used as the solvent for
mandelic acid MA and 2-chloromandelic acid (CMA), while aque-
ous ammonia was used as the solvent for Pregabalin (PREG). The
crystals were obtained by gradually cooling the reaction mixture
to 26 °C and then separating them from the mother liquor after
composition values (ee
acid CMA and Pregabalin PREG [ee
80%] than the ee difference of mandelic acid MA and Pregabalin
PREG [ee (MA) = 38% vs ee (PREG) = 80%] which led to more
E
) is greater in the case of 2-chloromandelic
E
(CMA) = 10% vs ee (PREG) =
E
E
3
0 min of crystallization. The results are summarized in Table 1.
E
E
The correlation between the initial enantiomeric purity (ee
0
)
efficient separation of the 2-chloromandelic acid enantiomers
CMA.
It was shown in our previous study that the eutectic composi-
E
tion (ee ) of the racemic compounds is in good agreement with
and the final enantiomeric purity (ee) in case of the MA, CMA, or
PREG is shown in Figure 2. Mandelic acid MA, 2-chloromandelic
acid CMA, and Pregabalin PREG are all racemate-forming