Studies on enantioselective liquid-liquid extraction of amino-(4-nitro-phenyl)-acetic acid enantiomers: Modeling and optimization

Article abstract of DOI:10.1002/chir.22266

BINAP-metal complexes were prepared as extractant for enantioselective liquid-liquid extraction (ELLE) of amino-(4-nitro-phenyl)-acetic acid (NPA) enantiomers. The influence of process variables, including types of organic solvents and metal precursor, co

Full text of DOI:10.1002/chir.22266

CHIRALITY 26:79–87 (2014)
Studies on Enantioselective Liquid–Liquid Extraction of Amino-(4-
nitro-phenyl)-acetic Acid Enantiomers: Modeling and Optimization
PANLIANG ZHANG, CHANG LIU, KEWEN TANG, * JIAJIA LIU, CONGSHAN ZHOU, AND CHANGAN YANG1
1
2
1
2
1
1
Department of Chemistry and Chemical Engineering, Hunan Institute of Science and Technology, Hunan, China
2
College of Chemistry and Chemical Engineering, Central South University, Hunan, China
ABSTRACT
BINAP-metal complexes were prepared as extractant for enantioselective liq-
uid–liquid extraction (ELLE) of amino-(4-nitro-phenyl)-acetic acid (NPA) enantiomers. The influ-
ence of process variables, including types of organic solvents and metal precursor, concentration
of ligand, pH, and temperature on the efficiency of the extraction, were investigated experimen-
tally. An interfacial reaction model was established for insightful understanding of the chiral ex-
traction process. Important parameters required for the model were determined. The
experimental data were compared with model predictions to verify the model prediction, It
was found that the interfacial reaction model predicted the experimental results accurately. By
modeling and experiment, an optimal extraction condition with pH of 7 and host (extractant) con-
centration of 1 mmol/L was obtained and high enantioselectivity (α ) of 3.86 and performance
op
factor (pf) of 0.1949 were achieved. Chirality 26:79–87, 2014. © 2013 Wiley Periodicals, Inc.
KEY WORDS: enantioselective liquid-liquid extraction; BINAP; metal complex; enantiomers;
model
INTRODUCTION
THEORY AND MODELING
Theory of Enantioselective Reactive Extraction
The production and availability of enantiomerically pure com-
pounds is of prime importance to the pharmaceutical industry
as well as to the agrochemical, flavor, and fragrance indus-
Knowledge of the reaction mechanism is required to gain a
better understanding of the reactive extraction process. Two
different mechanisms have been reported in previous works
on reactive extraction, namely, homogeneous ligand addition
mechanism and interfacial ligand exchange mechanism. The
main difference between the two mechanisms is the locus of
the complexation reaction.
1
–3
tries. The most common technique for obtaining enantiome-
rically pure compounds on a commercial scale is classical
resolution by crystallization but the method has low versatility.4
Compared with crystallization and some other methods, such
5
as chiral liquid chromatography and chiral capillary electro-
6
phoresis, enantioselective liquid–liquid extraction (ELLE) is
In enantioselective extraction of NPA enantiomers by CuPF6
{(S)-BINAP} (the extractant, written as BINAP-Cu for short),
the reaction between NPA enantiomers and the extractant
occurs either in one of the phases or at the interface. It was
found that the metal–BINAP complex is highly hydrophobic,
which excludes the possibility that the complexation reaction
takes place in the aqueous phase. The complexation reaction
could only happen either in the organic phase or at the inter-
face. Where the complexation reaction occurs depends on the
solubility of the reactant and the product of the complexation
reaction. It was found that NPA enantiomers could distribute
over the organic and aqueous phases. NPA anion can react with
BINAP-Cu through a ligand-exchange process. One mol of
NPA reacts with 1 mol of BINAP-Cu and generates 1 mol of
expected to be cheaper and easier to scale up to commercial
7
,8
scale and has a large application range.
A chiral selector plays an important role in ELLE, which binds9
enantiospecifically and reversibly with a racemic substrate.
Several chiral selectors have been reported, such as tartaric acid
1
0,11
12–14
9,15
derivatives,1
crown ethers,
cinchona alkaloids,
β-CD
6–19
20
21–27
derivatives,
metal complex, and so on
. The chiral
bisphosphine BINAP, which has been proven to be a highly ver-
satile ligand in asymmetric catalysis, makes it possible to bind
other ligands by exchange of a counterion. Therefore, the
BINAP–metal complex may provide an efficient selector in
ELLE. Some interesting examples can be found in previous
2
8,29
work for ELLE of amino acid enantiomers.
PF-
the organic phase, the product of this reaction PF
. Although NPA enantiomers and BINAP-Cu can exist in
Amino acid enantiomers are increasingly important to indus-
trial and laboratory processes, as they are considered crucial
chiral building blocks for a variety of biologically active com-
pounds, such as pesticides, peptides, and semisynthetic β-
lantam antibiotics. Amino-(4-nitro-phenyl)-acetic acid (NPA)
can be used as a D-serine transporter inhibitor for the treatment
6
-
can hardly
6
exist in the organic phase (Fig. 2). It is concluded that the
Contract grant sponsor: National Science Foundation of China; Contract grant
number: 21171054.
Contract grant sponsor: Open Fund Project of Key Laboratory in Hunan Uni-
versity; Contract grant number: 11K029.
Contract grant sponsor: Key Laboratory of Hunan Province.
Contract grant sponsor: Aid Program for Science and Technology Innovative
Research Team in Higher Educational Institutions of Hunan Province.
Correspondence to: Kewen Tang, Department of Chemistry and Chemical
Engineering, Hunan Institute of Science and Technology, Yueyang 414006,
Hunan, China. E-mail: tangkewen@sina.com
Received for publication 2 August 2013; Accepted 8 October 2013
DOI: 10.1002/chir.22266
Published online 10 December 2013 in Wiley Online Library
(wileyonlinelibrary.com).
3
0
of nervous system disorders. It can be seen as adding a nitro
to para-position of phenylglycine (Fig. 1).
This paper reports the ELLE of NPA enantiomers with (S)-
BINAP metal complex as the chiral extractant. The factors af-
fecting the extraction efficiency, such as types of metal ions
and organic solvents, pH of aqueous phase, concentrations of
host, and temperature, were investigated. The equilibrium of
the extraction system was modeled and optimized by an interfa-
cial reaction model.
*
©
2013 Wiley Periodicals, Inc.

8
0
ZHANG ET AL.
respectively; [D-]org and [L-]org are the concentrations of the D-
and L-NPA anion in the organic phase at equilibrium,
respectively.
The complexation equilibrium of BINAP-Cu with NPA
enantiomers at the interface can be written as follows:
Fig. 1. Chemical structure of NPA.
½
DCuBꢁ ½PF ꢀꢁ
org
6
w
KD
¼
(4)
(5)
ꢀ
reaction takes place at the interface between the two phases.
Therefore, the interfacial ligand exchange mechanism was ap-
plied in this paper for studying the reactive extraction of NPA
enantiomers by BINAP-Cu.
½
CuBꢁorg
D
w
ꢀ
½LCuBꢁ ½PF6 ꢁ
org
w
K ¼
L
ꢀ
½
CuBꢁorg½L ꢁ
w
Model Equations
where [DCuB]org and [LCuB]org are the concentrations of the
complexes of BINAP-Cu with D- and L-NPA in the organic
Consider the volume of aqueous phase (V ) equal to the vol-
w
ume of organic phase (V ). Species in both phases are in
phase; [CuB]org is the concentration of BINAP-Cu in the
org
-
chemical equilibrium, and each phase is open with respect to
the other. The enantioselective reactive liquid–liquid extraction
system can be modeled by a series of coupled equilibrium rela-
tions and mass balance equations as follows.
organic phase at equilibrium; [PF
PF-
in the aqueous phase.
Due to V = Vorg, the following equations represent the
6
]
w
is the concentration of
6
w
mass balance for D- and L-NPA:
The acid dissociation equilibrium constant in aqueous phase:
CD ¼ ½DHꢁ þ ½D ꢁ þ ½DHꢁ þ ½Dꢀꢁ þ ½DCuBꢁorg
ꢀ
(6)
(7)
w
w
org
org
Â
Ã
Â
Ã
þ
ꢀ
þ
ꢀ
H
½D ꢁ
H
½L ꢁ
w
w
CL ¼ ½LHꢁ þ ½L ꢁ þ ½LHꢁ þ ½Lꢀꢁ þ ½LCuBꢁorg
ꢀ
Ka ¼
¼
(1)
w
w
org
org
½DHꢁ
½LHꢁw
w
where C and C are the total concentrations of D- and L-NPA
D
L
where [DH] and [LH] are the concentrations of the molecu-
w
w
in the aqueous and organic phases.
lar D- and L-NPA in the aqueous phase at equilibrium,
1–5
Combining eqs. , eqs. 6 and 7 are deduced to:
-
-
respectively; [D] and [L] are the concentrations of the D- and
w
w
Â
Ã
Â
Ã
þ
ꢀ
þ
ꢀ
L-NPA anion in the aqueous phase at equilibrium, respectively.
The physical partition coefficient of molecular D- and L-
H
½D ꢁ
P H ½D ꢁ
w
ꢀ
0
w
ꢀ
CD
¼
þ ½D ꢁ þ
þ P ½D ꢁ
i
(8)
(9)
w
w
Ka
Ka
NPA, P , can be written as follows:
0
ꢀ
K ½D ꢁ ½CuBꢁ
D
w
org
½
DHꢁ
½LHꢁ
org
org
þ
P0 ¼
¼
(2)
ꢀ
½
PF6 ꢁ
½
DHꢁw
½LHꢁw
Â
Ã
Â
Ã
þ
ꢀ
þ
ꢀ
H
½L ꢁ
P H ½L ꢁ
w
ꢀ
0
w
ꢀ
and the physical partition coefficient of the D- and L-NPA
C ¼
L
þ ½L ꢁ þ
þ P ½L ꢁ
w
i
w
anion, P , can be written as follows:
Ka
Ka
i
ꢀ
ꢀ
ꢀ
½
D ꢁ
½L ꢁ
K ½L ꢁ ½CuBꢁ
org
org
L
w
org
P ¼
i
¼
(3)
þ
ꢀ
ꢀ
ꢀ
½
D ꢁ
½L ꢁ
½PF6 ꢁ
w
w
where [DH]org and [LH]org are the concentrations of the molec-
ular D- and L-NPA in the organic phase at equilibrium,
There is the following equation for mass balance of BINAP-
Cu complex:
Fig. 2. Diagram of enantioselective extraction of NPA enantiomers.
Chirality DOI 10.1002/chir

ENANTIOSELECTIVE LIQUID–LIQUID EXTRACTION: MODELING AND OPTIMIZATION
81
CCuB ¼ ½CuBꢁorg þ ½DCuBꢁorg þ ½LCuBꢁorg
where CCuB is the initial concentration of BINAP-Cu.
(10)
Ci;org
Ci
f i ¼
(19)
Combining eqs. 4 and 5, eq. 10 is deduced to
Where Ci,org represents the total concentration of the solute i
in organic phase at equilibrium, and C represents the initial
total concentration of the solute i.
The extraction performance factor (pf) is a very useful tool
to optimize an enantioselective extraction and is defined as:
ꢀ
i
K ½D ꢁ ½CuBꢁ
D
w
org
CCuB ¼ ½CuBꢁorg
þ
(11)
ꢀ
½
PF
ꢁ
6
w
ꢀ
K ½L ꢁ ½CuBꢁ
L
w
org
þ
pf ¼ f eeorg
(20)
ꢀ
i
i
½
PF6 ꢁ
w
þ
þ
A performance factor close to unity indicated a high enantio-
meric purity in both phases.
½
H ꢁ
P0½H ꢁ
Let A ¼
þ 1 þ
þ P and combine eqs. 8, 9, and 10,
i
Ka
Ka
the following equation is obtained,
Regression of the Complexation Equilibrium Constants
K C ½CuBꢁ
ꢀ ½CuBꢁorg þ K ½CuBꢁ
D
D
org
CCuB ¼ ½CuBꢁorg
þ
ꢀ
ꢁ
(12)
Regression of the complexation equilibrium constants was
A C
1–5
CuB
D
org
carried out as follows. Combining eqs.
, 14 and 15, the
þ
K C ½CuBꢁ
ꢀ ½CuBꢁorg þ K ½CuBꢁ
½H ꢁ
L
L
org
following equation can be achieved (let B ¼ K þ 1):
þ ꢀ
ꢁ
a
A C
Â
Ã
CuB
L
org
ꢀ
þ
ꢁꢀ
H
C k
CLkL
D
D
Bk ꢀ P0
ꢀ Pi
þ
D
With a further treatment of eq. 12, the following equation
can be deduced:
Ka
k þ 1 kL þ 1
D
À
Á
½
Dꢁw:total þ ½Lꢁw:total ðP þ ðB ꢀ 1ÞP0Þ
i
ꢀ
ꢁ
ꢀ
Þ
2
3
K K ꢀ K A ꢀ K A þ A ½CuBꢁ
B
D
L
D
L
(
21)
ꢀ
ꢀ
C k
CLkL
D
D
2
¼
KD CCuB
ꢀ
ꢀ
þ 2K AC
þ 2K AC
ꢀ 3A C
þ K K C
D
CuB
L
CuB
CuB
D
L
D
kD þ 1 kL þ 1
À
Á
2
½Dꢁw:total þ ½Lꢁw:total ðP þ ðB ꢀ 1ÞP Þ
i
0
ꢀ
K C AþÞK K C ꢀ K C A ꢀ K K CCuBÞ½CuBꢁ
D
D
D
L
L
L
L
D
L
þ
Þ
ꢀ
B
2
Â
Ã
þ K C AC
þ K C AC
ꢀ K AC
þ
D
D
CuB
L
L
2
CuB
D
CuB
3
ꢀ
ꢁꢀ
H
Ka
C k
CLk
L
ꢁ
D
D
Bk ꢀ P0
ꢀ Pi
þ
2
2
2
L
þK AC
Þþ3A C
½CuBꢁ ꢀ A C
¼ 0
kD þ 1 kL þ 1
L
CuB
CuB
CuB
À
Á
½
Dꢁw:total þ ½Lꢁw:total ðP þ ðB ꢀ 1ÞP Þ
i
0
(
13)
ꢀ
Þ
B
Since [CuB]org can be obtained by solving eq. 12, distribu-
tion ratios can be written as follows:
(22)
ꢀ
CDkD
kD þ 1 kL þ 1
CLkL
¼
KL CCuB
ꢀ
ꢀ
Â
þÃ
PiKDKa½CuBꢁorg
P0 H þ P Ka þ
À
Á
i
CCuBꢀ½CuBꢁ
org
½Dꢁw:total þ ½Lꢁw:total ðP þ ðB ꢀ 1ÞP Þ
k ¼
D
Â
Ã
(14)
(15)
i
0
þ
þ
Þ
Ka þ H
B
Â
þÃ
PiKL Ka½CuBꢁorg
where [D]w, total and [L]w, total are the total concentration of D-
and L-NPA in the aqueous phase at equilibrium, respectively.
By linear regression of the experimental data according to
P0 H þ P Ka þ
i
CCuBꢀ½CuBꢁ
org
kL ¼
Â
Ã
þ
Ka þ H
eqs. 21 and 22, the equilibrium constants, KD and K are
L
Enantioselectivity is defined as:
evaluated from the slop of the fitting lines.
kL
kD
αop ¼
(16)
(17)
MATERIALS AND METHODS
Materials
K
L
αint ¼ K
Bi(acetonitrile)dichloropalladium(II) (PdCl
97%) was purchased from the Metallurgy Institute of Zhejiang
Zhejiang, China). Bis(triphenylphosphine)nickel(II) ([(C P] NiCl
mass fraction >98%) and tetrakis(acetonitrile)copper(I) hexafluorophos-
phate ([(CH CN) Cu]PF , mass fraction >98%) were purchased from
Hewei Chemical (Guangzhou, China). 2,2’-Bis(diphenylphosphino)-1,1’-
binaphthalene (BINAP, mass fraction >99%) was purchased from
Shengjia Chemical. Racemic NPA was purchased from HanHong Bio-
chemical (Jiangsu, China). The chiral extractants (BINAP-metal com-
plexes) used in this paper were generated in situ by adding the metal
precursor and BINAP in equal molar amounts to the appropriate organic
solvent. The reaction mixture was stirred overnight and diluted to the
desired concentration and the resulting solution was directly applied
in extraction experiments as the organic phase. Solvent for chromatog-
raphy was of high-performance liquid chromatography (HPLC) grade.
Purified water was obtained by reverse osmosis followed by distillation.
2 3 2
(CH CN) , mass fraction
D
>
(
H
6 5
)
3
2
2
,
where αop is the operational selectivity, αint is the intrinsic
selectivity.
A common alternative measure of system performance is pro-
vided by the enantiomeric excess (ee), which can be expressed
in terms of distribution ratios by the following equation:
3
4
6
CL
CD
1þ1
CD
1þ1
ꢀ
þ
1
þ1
=
kL
=
kD
eeorg ¼
(18)
CL
1
þ1
=
kL
=
kD
The fraction of the solute i (i = D or L) extracted into the
aqueous phase (f ) is given by
i
Chirality DOI 10.1002/chir

8
2
ZHANG ET AL.
ꢀ
All other chemicals were of analytical-reagent grade and bought from
different suppliers.
½
AHꢁ þ ½A ꢁ
org
org
Papp ¼
(23)
ꢀ
½
AHꢁ þ ½A ꢁ
w
w
-
where [AH]org and [A ] are the equilibrium concentrations
Determination of Physical Distribution Coefficients P
0
and P
i
org
of the molecular and ionic NPA in the organic phase, respec-
Experiments to determine the physical distribution coefficients of
molecular and ionic NPA (P and P ) over the aqueous phase and 1,2-
-
tively; [AH] and [A ] are the equilibrium concentrations of
w
w
0
i
the molecular and ionic NPA in the aqueous phase,
respectively.
Combining eqs. 1–3, eq. 23 can be derived as:
dichloroethane were carried out in a water bath at 5 °C. The aqueous
phase were prepared by dissolving 2.0 mmol/L NPA in 0.1 mol/L sodium
phosphate buffer solutions with a series of pH values in the range 3 ~ 9.
Equal volumes of the two phases (each 2 mL) were placed together and
shaken sufficiently for 12 h before being kept in a water bath at 5 °C to
reach equilibrium. The concentrations of NPA enantiomers in the
aqueous phase were analyzed by HPLC. The concentrations of NPA
enantiomers in the organic phase was calculated from a mass balance.
ꢂ
Â
Ã
ꢃ
Â
Ã
þ
þ
H
H
Papp 1 þ
¼ P0
þ Pi
(24)
Ka
Ka
The pKa of 5.60 was determined by potentiometric titration.
The plot of Papp(1 + [H ]/K ) versus [H ]/K yielded a
+
+
a
a
straight line (in Fig. 3). The slope and intercept of the line
Determination of Complexation Equilibrium Constants
were used to evaluate P0 as 0.01074 and P as 0.08095,
i
D L
K and K
respectively.
Reactive extraction experiments were carried out to obtain the com-
plexation equilibrium constants of BINAP-Cu with NPA enantiomers.
The racemic NPA was dissolved in sodium phosphate buffer solution
D L
Complexation Equilibrium Constants K and K
(
1
phosphate concentration = 0.10 mol/L, pH = 7.0) with a concentration of
.0 mmol/L. BINAP-Cu was dissolved in 1,2-dichloroethane with a chang-
Complexation equilibrium constants could be determined
through a series of reactive extraction experiments
described in the experimental section. Figure 4 shows
the regression analysis of the experimental data according
ing concentration from 0.25 mmol/L to 2 mmol/L. The two phases were
put together in a test tube with equal amounts (each 2 mL) and stirred
for 12 h at 5 °C. After equilibrium, the composition of the aqueous phase
was analyzed by HPLC.
to eqs. 21 and 22. We defined parameter
X
=
ð½Dꢁw:totalþ½Lꢁ
ÞðPiþðBꢀ1ÞP0Þ
CDkD CL kL
w:total
ðCCuB-
-
þ
Þ , parameter Y
kDþ1 kL þ1
B
þ
½
H ꢁ
CDkD
kDþ1
CL k
kL þ1
ð½Dꢁw:totalþ½Lꢁw:totalÞðPiþðBꢀ1ÞP0Þ
Extraction Experiments
The aqueous phase was prepared by dissolving racemic NPA in
L
=
ðBk -P0
-P Þð
þ
-
Þ for
D
i
Ka
B
þ
½
H ꢁ
C
k
D
D
D-NPA and parameter
Y
=
ðBk -P0
-P Þð
i
þ
0
.1 mol/L phosphate buffer solution, and the organic phase was prepared
by dissolving equimolar amounts of the metal precursor (PdCl (CH CN)
, [(CH CN) Cu]PF , [(C P]NiCl ) and BINAP in the appropriate or-
L
K
kDþ1
a
2
3
CL kL
kL þ1
ð½Dꢁw:totalþ½Lꢁ
w:total
ÞðPiþðBꢀ1ÞP0Þ
-
Þ for L-NPA. As shown in
H
5
)
3
2
B
2
3
4
6
6
ganic solvent and the solution was stirred overnight before being used.
Equal volumes of host-containing organic phase and NPA-containing
aqueous phase (each 2 mL) were added to a 10-mL centrifuge tube and
shaken sufficiently for 12 h at a fixed temperature of 5 °C before being
kept in a water bath at 5 °C for 12 h to reach equilibrium. Then the tube
was centrifuged to accelerate phase separation and the organic phase
was discarded. The aqueous phase was analyzed by HPLC to obtain the
concentrations of NPA enantiomers in the aqueous phase. The concentra-
tions of NPA enantiomers in the organic phase were calculated from a
mass balance.
Figure 4, the plot of parameter Y versus parameter X for
D-NPA yields a straight line and the slope of the line
was used to evaluate KD as 4.46 (dimensionless quantity).
Similar data treatment according to eq. 22 can be
performed to evaluate K as 18.90. The intrinsic selectivity
L
αint (K / K ) is estimated as 4.24.
L
D
Screening of Metal Precursors
CN) , [(CH CN) Cu]PF , [(C
applied as the metal precursors for synthesis of chiral extractant.
PdC1
(CH
H ) P] NiC1 were
6 5 3 2 2
2
3
2
3
4
6
Analytical Method
The quantification of NPA enantiomers in the aqueous phase was
performed by HPLC using a WATERS e2695 apparatus (Waters, Milford,
MA). A UV detector operated at 260 nm was applied. The column was
diamonsil C18, 5 um particle size of the packing material, 250 × 4.6 mm
I.D. (Dikma Technologies, Beijing China). The mobile phase was a 1:4
(v/v) mixture of methanol and a 0.1 mol/L acid sodium phosphate
solution containing 0.5 mmol/L of cupric sulfate and 2.0 mmol/L of L-
phenylalanine at pH = 4.8 (pH was adjusted with acetic acid). The flow
rate was set at 1 mL/min and the column temperature was set at 27 °C.
RESULTS AND DISCUSSION
Physical Distribution Coefficients P and P
0 i
Physical distribution coefficients for molecular and ionic
NPA (P and P ) were determined through a series of physical
0
i
extraction (extraction without extractant in the organic
phase) experiments mentioned above. The apparent partition
coefficients Papp were determined at different pH values. As
described in Figure 2, both the molecular and ionic NPA
can distribute over the organic and aqueous phases.
Then Papp is given by
Fig. 3. Plot for calculating physical distribution coefficients P
0
i
and P for
2
NPA in 1,2-dichloroethane-water two phase system at 5 °C. R =0.9938.
Chirality DOI 10.1002/chir

ENANTIOSELECTIVE LIQUID–LIQUID EXTRACTION: MODELING AND OPTIMIZATION
83
TABLE 2. Influence of organic solvent type
Organic solvent
.2-dichloroethane
trichloromethane
dichloromethane
chlorobenzene
k
D
k
L
α
op
1
0.38
0.91
0.37
0.20
1.41
1.64
0.83
0.56
3.71
1.80
2.24
2.80
Condition: [NPA]
0
= 2.0 mmol/L; pH = 7.5; T = 5 °C; [BINAP-Cu]
0
= 1.0 mmol/L.
Influence of pH
The effect of pH of the aqueous phase on the distribution
ratios and operational selectivity were investigated. We ap-
plied the interfacial reaction model to predict the distribution
ratios and enantioselectivity as a function of pH. The compar-
ison of the model predictions with the experimental results is
depicted in Figure 5. It is obvious that the experimental
Fig. 4. Plot of calculating complexation equilibrium constants for NPA en-
CDkD
antiomers in 1,2-dichloroethane at 5 °C. Parameter X is PiðCCuB ꢀ
ꢀ
þ
kDþ1
þ
ð½Dꢁ
þ½Lꢁ
ÞðPiþðBꢀ1ÞP0Þ
½H ꢁ
CL kL
kL þ1
CL kL
kL þ1
CL kL
kL þ1
w:total
w:total
C
D
k
D
þ
ꢀ
ꢀ
Þ; Parameter Y is ðBkD ꢀ P0 Ka ꢀ PiÞð
B
kDþ1
þ
ð½Dꢁw:total þ½Lꢁw:total ÞðPiþðBꢀ1ÞP0Þ
½H ꢁ
C
k
D D
Þ for D-NPA and ðBkL ꢀ P0
ꢀ PiÞð
þ
B
Ka
kDþ1
ð½Dꢁw:total þ½Lꢁw:total ÞðPiþðBꢀ1ÞP0Þ
2
2
Þ for L-NPA; R =0.9948 for D-NPA and R
B
=
0.9961 for L-NPA.
Distribution ratios (k and k ) and operational selectivity (α )
D
L
op
with different metal precursors are shown in Table 1. The
highest selectivity of 2.24 is achieved using [(CH CN) Cu]
3
4
PF as metal precursor and the distribution ratios are 0.37 and
6
0
.83, respectively. BINAP-Pd complex used as a chiral selector
has been previously investigated by Verkuijl et al. and a series
of amino acid enantiomers were separated with good enantio-
selectivity. Comparing this work with Verkuijl et al.’s work,
BINAP-Cu can provide a higher enantioselectivity and replac-
ing expensive Pd with cheap Cu will be economically more
interesting. The Ni-(S)-BINAP complex does not show any se-
lectivity towards NPA enantiomers. It is obvious that copper
28
(
I) is the most suitable central ion for the chiral extractant.
Screening of Organic Solvents
The effect of organic solvents was explored using BINAP-
Cu complex dissolved in different solvents. The distribution
ratios and operational selectivity are shown in Table 2. It
can be seen from Table 2 that organic solvent has a clear
influence on the extraction performance. Good performance
was obtained using the organic solvents studied. Highest
enantioselectivity is obtained with 1,2-dichloroethane as
organic solvents and the distribution ratios are also satisfac-
tory. Therefore, 1,2-dichloroethane was chosen as the suit-
able solvent for extraction of NPA enantiomers.
TABLE 1. Influence of metal precursor
Metal precursor
PdCl (CH CN)
k
D
k
L
α
op
2
3
2
0.85
0.37
0.13
1.09
0.83
0.13
1.28
2.24
1
Cu(CH CN) PF
Ni(DEEP)Cl
3
4
6
2
Fig. 5. Influence of pH on distribution ratios (a) and operational enantio-
selectivity (b). Solid lines: model predictions. Symbols: experimental data. Condi-
Condition: [NPA]
mmol/L; [metal precursor]
dichloromethane.
0
= 2.0 mmol/L; pH = 7.5; T = 5 °C; [(S)-BINAP]
0
= 1.0
0
=
1.0 mmol/L; organic solvent
=
tion: solvent: 1,2-dichloroethane; [NPA] = 2.0 mmol/L; T = 5 °C; [BINAP-Cu]
0
0
= 1.0 mmol/L.
Chirality DOI 10.1002/chir

8
4
ZHANG ET AL.
values are in a good agreement with the model prediction
when pH is higher than 7.
It is shown in Figure 5 that at pH below 4, distribution
Influence of the Host Concentration
Model prediction of distribution ratios and operational se-
lectivity is presented as solid lines, and the comparison of
the experimental values with the model predictions are
shown in Figure 6. It is observed that the model predicts
the distribution ratios and operational selectivity accurately.
It is can be seen from Figure 6 that the distribution ratios
and the operational enantioselectivity both increase with the
increase of the host (BINAP-Cu) concentration, and αop in-
creases rapidly at first and then only slightly. More enantio-
mer–host complexes are formed in organic phase and the
distribution ratios increase consequently when the host con-
centration increases. Meanwhile, BINAP-Cu is able to recog-
nize enantiomers and a relatively higher concentration will
enhance the recognition ability. Because of the physical parti-
tion of NPA, the correlation between host concentration and
operational enantioselectivity differs from that reported by
ratios (k and k ) increase slightly with the rising of pH but
D
L
k and k increase rapidly with pH rising from 5 to 7. When
D
L
pH is above 7, k and k remain basically unchanged. It also
D
L
can be seen that the operational enantioselectivity (αop)
increases with pH rising from 3 to 5; when pH rising from 5
to 6, αop decreases slightly; when pH is higher than 6, αop
keep nearly constant. This might be explained by the fact that
only NPA anion binds to the BINAP-Cu. When pH is rising,
more NPA in its anion state distributes in the organic phase
and is recognized by the extractant. Therefore, the distribu-
tion ratios and enantioselectivity increases. Model predic-
tions and experimental results show a maximum αop near
pH of 5.
28
Verkuijl et al.
Fig. 6. Influence of host concentration on distribution ratios (a) and opera-
tional enantioselectivity (b). Solid lines: model predictions. Symbols: experi-
Fig. 7. Influence of temperature on distribution ratios (a) and operational
enantioselectivity (b). Condition: solvent: 1,2-dichloroethane; [NPA]
= 2.0 mmol/L; pH= 7.0; [BINAP-Cu] = 1.0 mmol/L.
0 0
0
mental data. Condition: solvent: 1,2-dichloroethane; [NPA] = 2.0 mmol/L;
T = 5 °C; pH = 7.0
.
Chirality DOI 10.1002/chir

ENANTIOSELECTIVE LIQUID–LIQUID EXTRACTION: MODELING AND OPTIMIZATION
85
Influence of Temperature
on extraction performance in a single stage extraction system
and to optimize the operational conditions.
In Figure 8, distribution ratios and enantioselectivity of
NPA enantiomers is calculated as a function of pH and host
concentration, respectively. It can be found that kD and kL
The influence of temperature on the operational enantio-
selectivity and distribution ratios was investigated in the
range of 5–30 °C. It is seen from Figure 7 that temperature
shows a strong influence on the distribution ratios and opera-
tional enantioselectivity. An increase of kD and kL and a
decrease of αop were caused by increasing temperature. How-
ever, with a temperature lower than 5 °C, the increase in
viscosity of solution and the decrease of the solubility of host
and substrate would make the extraction very difficult to
carry out. Hence, T = 5 °C is considered the optimal tempera-
ture for enantioselective extraction.
have a similar tendency with the change of pH and host con-
centration. The increase of pH and host concentration can
give rise to distribution ratios. Enantioselectivity is strongly
influenced by pH and host concentration. Maximum
enantioselectivity is found at a pH of about 5 and BINAP-Cu
concentration of higher than 1 mmol/L.
In Figure 9, the ee for NPA enantiomers is calculated as a
function of pH and host concentration. The ee is strongly
influenced by pH and the host concentration. Maximum ee
Modeling and Optimization
(
higher than 0.5) was obtained near pH = 4.0. It also illustrates
The experimental results are compared graphically with
model predictions, which indicate the interfacial reaction
model is a good means of predicting enantiomers partitioning
over a range of process variables. Therefore, the model was
used to explore the influence of various operating conditions
that the ee increases with the increase of host concentration
rapidly at first and then slightly. It is found from Figures 8
and 9 that good selectivity (αop) and good purity (ee) are
obtained at relatively low distribution ratios which are not
conducive to increasing yield. Acquiring a good purity and a
(
a)
(b)
Fig. 8. Calculated distribution ratios (a, b) and enantioselectivity (c) of NPA enantiomers as a function of pH and host concentration. Condition: solvent: 1,2-
dichloroethane; [NPA] = 2.0 mmol/L; T = 5 °C.
0
Chirality DOI 10.1002/chir

8
6
ZHANG ET AL.
Fig. 9. Calculated enantiomeric excess (ee) for NPA enantiomers as a func-
tion of pH and host concentration. Condition: solvent: 1,2-dichloroethane;
NPA] = 2.0 mmol/L; T = 5 °C.
[
0
Fig. 12. Performance factor as a function of host concentration. Solid lines:
model predictions. Symbols: experimental data. Conditions: solvent: 1,2-
0
dichloroethane; [NPA] = 2.0 mmol/L; pH = 7.0; T = 5 °C.
good yield are somehow contradictory. A performance factor
pf) is introduced for further optimization of the reactive
(
extraction system.
The performance factor (pf) is defined as the product of ee
in the organic phase and the fraction of enantiomer extracted
into the organic phase. The given enantiomer can be purified
to high purity with maximum yield in the condition with a
high performance factor. Figure 10 shows the performance
factors as a function of pH and host concentration. The host
concentration and pH have a strong influence on the perfor-
mance factor. With host concentration ranging from 1 to
2
mmol/L and pH higher than 5, a relatively high perfor-
mance factor is obtained.
Performance factors were determined experimentally to
support the model predictions (Figs. 11 and 12). As is
shown in Figure 11, the measured performance factor in-
creases with pH rising from 3 to 5, then decreases slightly
and finally remains nearly constant at pH higher than 6.
In Figure 12, the result illustrates that the performance
factor increases rapidly and reaches a maximum with the
increase of host concentration, and then decreases
smoothly. It is observed in Figures 11 and 12 that the mea-
sured performance factor in good agreement with the
model prediction. These results therefore verify the accu-
racy of the model and show the possibility of application
for system optimization.
Fig. 10. Calculated performance factor (pf) for NPA enantiomers as a func-
tion of pH and host concentration. Condition: solvent: 1,2-dichloroethane;
NPA] = 2.0 mmol/L; T = 5 °C.
[
0
CONCLUSIONS
The ELLE of NPA enantiomers with metal-BINAP
complex as chiral extractant (host) was investigated by
experiment and modeling. Excellent agreement between
the experimental data and the model predictions was
observed, which indicated that the interfacial reaction
model is capable of predicting the distribution of NPA
enantiomers in the extraction system studied. The best
conditions identified involve the use of copper(I)-BINAP
complex concentration of 1 mmol/L and pH value of 7 at
Fig. 11. Performance factor as a function of pH. Solid lines: model predic-
tions. Symbols: experimental data. Conditions: solvent: 1,2-dichloroethane;
NPA] = 2.0 mmol/L; [BINAP-Cu] = 1.0 mmol/L; T = 5 °C.
5
°C. Full separation of the NPA enantiomers can be
achieved by multistage extraction.
[
0
0
Chirality DOI 10.1002/chir

ENANTIOSELECTIVE LIQUID–LIQUID EXTRACTION: MODELING AND OPTIMIZATION
87
aid of cinchona-based chiral host compounds. Chem Eur
009;15:2111–2120.
J
LITERATURE CITED
2
1
. Federsel HJ. Facing chirality in the 21st century: Approaching the
challenges in the pharmaceutical industry. Chirality 2003;15:S128–S142.
. Tombo GMR, Bellus D. Chirality and crop protection. Angew Chem Int
16. Tang K, Zhang P, Pan C, Li H. Equilibrium studies on enantioselective
extraction of oxybutynin enantiomers by hydrophilic β-cyclodextrin
derivatives. AIChE J 2011;57:3027–3036.
2
Ed 1991;30:1193–1215.
. Rouhi AM. Chiral business. Chem Eng News 2003;81:45–55.
17. Valle EMMD. Cyclodextrins and their uses: A review. Process Biochem
2002;38:373–377.
3
4
1
8. Tang K, Zhang P, Li H. Experimental and model study on the multi-
ple chemical equilibrium for reactive extraction of ibuprofen enantio-
mers with HP-β-CD as hydrophilic selector. Process Biochem
. Vries T, Wynberg H, van Echten E, Koek J, ten Hoeve W, Kellogg RM,
Broxterman QB, Minnaard A, Kapteinm B, van der Sluis S, Hulshof L,
Kooistra J. The family approach to the resolution of racemates. Angew
Chem Int Ed 1998;37:2349–2354.
2
011;46:1817–1824.
1
9. Tang K, Yi J, Huang K, Zhang G. Biphasic recognition chiral extraction: A
novel method for separation of mandelic acid enantiomers. Chirality
5
. Schulte M, Strube J. Preparative enantioseparation by simulated moving
bed chromatography. J Chromatogr A 2001;906:5824–5828.
2009;21:390–395.
6
. Zhou SY, Zuo H, Stobaugh JF, Lunte CE, Lunte SM. Continuous in vivo
monitoring of amino acid neurotransmitters by microdialysis sampling
with online derivatization and capillary electrophoresis separation. Anal
Chem 1995;67:594–599.
2
0. Koska J, Haynes CA. Modeling multiple chemical equibria in partition
systems. Chem Eng Sci 2001;56:5853–5864.
21. Dimitrova P, Bart HJ. Extraction of amino acid enantiomers with
microemulsions. Chem Eng Technol 2009;32:1527–1534.
7
. Schuur B, Floure J, Hallett AJ, Winkelman JGM, de Vries JG, Heeres HJ.
Continuous chiral separation of amino acid derivatives by enantioselective
liquid-liquid extraction in centrifugal contactor separators. Org Process
Res Dev 2008;12:950–955.
22. Tang L, Choi S, Nandhakumar R, Park H, Chung H, Chin J, Kim KM.
Reactive extraction of enantiomers of 1,2-amino alcohols via stereoselective
thermodynamic and kinetic processes. J Org Chem 2008;73:5996–5999.
8
. Schuur B, Winkelman JGM, Heeres HJ. Equilibrium studies on
enantioselective liquid-liquid amino acid extraction using a cinchona
alkaloid extractant. Ind Eng Chem Res 2008;47:10027–10033.
23. Yoon J, Cram DJ. Chiral recognition properties in complexation of two
asymmetric hemicarcerands. J Am Chem Soc 1997;119:11796–11806.
2
2
4. Ding HB, Carr PW, Cussler EL. Racemic leucine separation by hollow-
fiber extraction, AIChE J 1998;38:1493–1498.
5. Snyder SE, Carey JR, Pirkle WH. Biphasic enantioselective
partitioning studies using small-molecule chiral selectors. Tetrahe-
dron 2005;61:7562–7567.
9. Prelog M, Kovakevic M, Egli M. Lipophilic tartaric acid esters as
enantioselective ionophores. Angew Chem Int Ed Engl 1989;28:1147–1152.
1
1
1
1
0. Viegas RMC, Afonso CAM, Crespo JG, Coelhoso IM. Modelling of the
enantio-selective extraction of propranolol in a biphasic system. Sep Purif
Technol 2007;53:224–234.
1. Tan B, Luo GS, Wang JD. Extractive separation of amino acid enantiomers
with co-extractants of tartaric acid derivative and aliquat-336. Sep Purif
Technol 2007;53:330–336.
2. Pietraszkiewicz M, Kozbia M, Pietraszkiewicz O. Chiral discrimination of
amino acids and their potassium or sodium salts by optically active crown
ether derived from D-mannose. J Membr Sci 1998;138:109–113.
3. Steensma M, Kuipers NJM, de Haan AB, Kwant G. Influence of process
parameters on extraction equilibria for the chiral separation of amines
and amino-alcohols with a chiral crown ether. J Chem Technol Biotechnol
2
2
2
2
6. Steensma M, Kuipers NJM, de Haan AB, Kwant G. Identification of
enantioselective extractants for chiral separation of amines and amino al-
cohols. Chirality 2006;18:314–328.
7. Kocabas E, Karakucuk A, Sirit A, Yilmaz M. Synthesis of new chiral calyx
[4]arene diamide derivatives for liquid phase extraction of α-amina acid
methylesters, Tetrahedron: Asymmetry 2006;17:1514–1520.
8. Verkuijl BJV, Minnaard AJ, de Vries JG, Feringa BL. Chiral separation of
underivatized amino acids by reactive extraction with palladium-BINAP
complexes. J Org Chem 2009;74:6526–6533.
9. Verkuijl BJV, Schuur B, Minnaard AJ, de Vries JG, Feringa BL. Chiral
separation of substituted phenylalanine analogues using chiral palladium
phosphine complexes with enantioselective liquid–liquid extraction. Org
Biomol Chem 2010;8:3045–3054.
2
006:81:588–597.
1
4. Colera M, Costero AM, Gaviña P, Gil S. Synthesis of chiral 18-crown-6-
ethers Containing lipophilic chains and their enantiomeric recognition of
chiral ammonium picrates, Tetrahedron: Asymmetry 2005;16:2673–2679.
5. Hallett AJ, Kwant GJ, de Vries JG. Continuous separation of racemic 3,5-
dinitrobenzoyl-amino acids in a centrifugal contact separator with the
3
0. Foster AC, Li Y, Staubli U, Viswanath V, Luhrs L. D-serine transporter
inhibitors as pharmaceutical compositions for the treatment of central
nervous system disorders. U.S. Patent: 20,120,329,851. 2012.
1
Chirality DOI 10.1002/chir