Electrically assisted liquid-phase microextraction combined with capillary electrophoresis for quantification of propranolol enantiomers in human body fluids

Article abstract of DOI:10.1002/chir.22308

In this study, electromembrane extraction (EME) combined with cyclodextrin (CD)-modified capillary electrophoresis (CE) was applied for the extraction, separation, and quantification of propranolol (PRO) enantiomers from biological samples. The PRO enantiomers were extracted from aqueous donor solutions, through a supported liquid membrane (SLM) consisting of 2-nitrophenyl octyl ether (NPOE) impregnated on the wall of the hollow fiber, and into a 20-μL acidic aqueous acceptor solution into the lumen of hollow fiber. Important parameters affecting EME efficiency such as extraction voltage, extraction time, pH of the donor and acceptor solutions were optimized using a Box-Behnken design (BBD). Then, under these optimized conditions, the acceptor solution was analyzed using an optimized CD-modified CE. Several types of CD were evaluated and best results were obtained using a fused-silica capillary with ammonium acetate (80 mM, pH 2.5) containing 8 mM hydroxypropyl-β-CD as a chiral selector, applied voltage of 18 kV, and temperature of 20C. The relative recoveries were obtained in the range of 78-95%. Finally, the performance of the present method was evaluated for the extraction and determination of PRO enantiomers in real biological samples. Chirality 26:260-267, 2014. 2014 Wiley Periodicals, Inc.

Full text of DOI:10.1002/chir.22308

CHIRALITY 26:260–267 (2014)
Electrically Assisted Liquid-Phase Microextraction Combined With
Capillary Electrophoresis for Quantification of Propranolol
Enantiomers in Human Body Fluids
*
HADI TABANI, ALI REZA FAKHARI, ABOLFATH SHAHSAVANI, AND HOSSEIN GHARARI ALIBABAOU
Department of Pure Chemistry, Faculty of Chemistry, Shahid Beheshti University, G.C., Evin, Tehran, Islamic Republic of Iran
ABSTRACT
In this study, electromembrane extraction (EME) combined with cyclodextrin
(CD)-modified capillary electrophoresis (CE) was applied for the extraction, separation, and quantifi-
cation of propranolol (PRO) enantiomers from biological samples. The PRO enantiomers were
extracted from aqueous donor solutions, through a supported liquid membrane (SLM) consisting
of 2-nitrophenyl octyl ether (NPOE) impregnated on the wall of the hollow fiber, and into a 20-μL
acidic aqueous acceptor solution into the lumen of hollow fiber. Important parameters affecting
EME efficiency such as extraction voltage, extraction time, pH of the donor and acceptor solutions
were optimized using a Box-Behnken design (BBD). Then, under these optimized conditions, the ac-
ceptor solution was analyzed using an optimized CD-modified CE. Several types of CD were evaluated
and best results were obtained using a fused-silica capillary with ammonium acetate (80 mM, pH 2.5)
containing 8 mM hydroxypropyl-β-CD as a chiral selector, applied voltage of 18 kV, and temperature
of 20°C. The relative recoveries were obtained in the range of 78–95%. Finally, the performance of the
present method was evaluated for the extraction and determination of PRO enantiomers in real biolog-
ical samples. Chirality 26:260–267, 2014. © 2014 Wiley Periodicals, Inc.
KEY WORDS: membrane extraction; experimental design; cyclodextrin; plasma; urine; chiral;
pharmaceuticals; Box-Behnken design
INTRODUCTION
low. Thus, an initial sample preconcentration step for isolation
and preconcentration of target analytes is required.
In spite of significant developments over the last three
decades, separation of enantiomers still remains one of
the hot topics of separation science. Propranolol (PRO)
[1-isopropylamino-3-(1-naphthoxy)-2-propanol] is a type of
β-blocker that is widely used in the treatment of angina, hyper-
tension, migraine headaches, and thyrotoxicosis. This drug is a
chiral compound with an asymmetric center at its side chain
and the two enantiomers are different from each other in terms
of physiological and behavioral effects.1 Pharmacological stud-
ies have shown that S-PRO is about 100 times more active than
the (+)-R-PRO2 and only (–)-S-PRO appears to be beneficial in
treating angina pectoris.3 Thus, it seems that the concentration
of (+)-R-PRO would be more than (–)-S-PRO in urine or plasma
samples obtained from a patient under treatment with racemic
PRO tablets.
In 2006, Pedersen-Bjergaard et al. reported a novel
microextraction technique called electromembrane extraction
(EME).14 EME can be used to extract ionizable compounds
from plasma samples and other complicated biological matri-
ces without protein precipitation.15 In this process, the ionized
target analytes are extracted from a donor phase (DP) into an
organic solvent, serving as a supported liquid membrane
(SLM), which has been located in the pores of a hollow fiber,
and then transported into an acceptor phase (AP) inside the
lumen of a hollow fiber by applying an electric potential across
the SLM. Clean and enriched extracts are obtained in hollow
fiber-based microextraction techniques due to the fact that
the organic solvent immobilized in the small pores of the
hollow fiber prevents the extraction of macromolecules, hydro-
philic compounds, and neutral substances into the AP.
Determination of PRO enantiomers has already been carried
out using chromatographic methods such as high-performance
liquid chromatography coupled with UV detection (HPLC-UV)4,5
or tandem mass spectrometry (LC-MS-MS),6 capillary elec-
trophoresis (CE),7–9 and capillary electrochromatography
(CEC).10,11 Gas chromatography followed by mass spectrom-
etry (GC-MS) has also been used for the analysis of PRO
from urine after a derivatization step. Although separation
of PRO enantiomers was not investigated in that report.12
Compared to HPLC, chiral separation by CE has the advan-
tages of higher separation efficiency, more rapid analysis,
flexibility, and also it allows the incorporation of various
chiral selectors at different concentrations.13 Therefore, it has
been extensively used in the field of chiral separations. How-
ever, the use of CE with on-column UV detection in drug mon-
itoring in biological samples has been limited due to its high
concentration detection limits and short optical pathlength,
whereas the analyte concentration in biological samples is very
© 2014 Wiley Periodicals, Inc.
In this work, an EME procedure and a cyclodextrin (CD)
modified CE method were combined for enantiomer separation
of PRO. The main goal was to make the best use of the advan-
tages and strengths of both techniques in order to develop a
powerful chiral method for quantification of PRO enantiomers
present in low concentrations in biological matrices. Different
variables affecting EME were optimized with the Box-Behnken
design (BBD) and response surface methodology (RSM).
Finally, the optimized procedure was employed to determine
PRO enantiomers in plasma and urine samples.
*Correspondence to: Ali Reza Fakhari, Department of Pure Chemistry, Faculty
of Chemistry, Shahid Beheshti University, G. C., P.O. Box 19396-4716, Evin,
Tehran, Islamic Republic of Iran. E-mail: a-zavareh@sbu.ac.ir
Received for publication 21 June 2013; Accepted 4 January 2014
DOI: 10.1002/chir.22308
Published online 17 March 2014 in Wiley Online Library
(wileyonlinelibrary.com).

ELECTRO MEMBRANE EXTRACTION OF PROPRANOLOL ENANTIOMERS
261
anode, was led directly into the sample solution. Subsequently, the elec-
trodes were coupled to the power supply and the sample vial was placed
on a stirrer. The power supply was turned on and extraction was
performed for a prescribed time. After the extraction was completed,
the power supply was turned off and the hollow fiber was removed from
the DP. The sealed end of the hollow fiber was cut with scissors and
the AP containing extracts was collected with a microsyringe and trans-
ferred to a microvial for the CE analysis.
EXPERIMENTAL
Reagents and Materials
All the chemicals used were of analytical reagent grade. 2-Nitrophenyl
octyl ether (NPOE), 1-ethyl-2-nitrobenzene (ENB), nitrobenzene (NB),
β-CD, hydroxypropyl-β-CD (HP-β-CD, with an average degree of substitu-
tion (DS) of 5.6), α-CD and hydroxypropyl-α-CD (HP-α-CD; DS = 3.6) were
purchased from Fluka (Buchs, Switzerland). H3PO4, NaH2PO4.2H2O,
Na2HPO4, NaOH, HCl, and NH4CH3COO were purchased from Merck
(Darmstadt, Germany). To prevent capillary blockage, all buffers and
samples were filtered through 0.45-μm filter membranes (Millipore,
Bedford, MA). HPLC-grade water was obtained through a Milli-Q system
(Millipore) and was used to prepare all solutions. The racemic PRO (purity
>99.0%) was obtained from Tofigh Daru (Tehran, Iran) and used without
further purification. Samples of human plasma 1, human plasma 2, and
urine 2 were collected from people who had not used PRO at all. Sample
urine 1 was collected from a patient under treatment with racemic PRO
tablets (40 mg/day). All solutions were stored at 4°C, protected from light.
Calculation of Enrichment Factor, Extraction Recovery and
Relative Recovery
The enrichment factor (EF) was defined as the ratio of the final analyte
concentration in the AP (Cf,a) to the initial concentration of analyte in the
DP (Ci,d):
Cf ;a
Ci;d
EF ¼
(1)
Recovery (R) was calculated according to the following equation for
each analyte:
Standard and Real Sample Solutions
Stock solution of racemic PRO (1 mg/mL) was prepared in HPLC-
grade water. The stock solution was protected from light and stored for
a month at 4°C with no evidence of decomposition. Then all required
standard solutions were freshly prepared from this stock solution and
diluted with HPLC-grade water. The plasma and urine samples were
diluted at 1:2 and 1:1 ratio using HPLC-grade water, respectively. pH of
plasma and urine samples was adjusted to 3.5 with the addition of 1 M
of HCl solution.
ꢀ
ꢁ ꢀ
ꢁ
nf;a
ni;d
Va
Vd
Cf;a
Ci;d
R% ¼
ꢀ 100 ¼
ꢀ 100
(2)
Where ni,d and nf,a are the number of moles of analyte originally
present in the DP and the number of moles of analyte finally collected
in the AP, respectively. Va is the volume of AP and Vd is the volume of
the DP.
Relative recovery (RR) was acquired from the following equation:
Cfound ꢁ Creal
CE Equipment
RR% ¼
ꢀ 100
(3)
Cadded
CE was carried out using a Lumex Capel 105 (Lumex, St. Petersburg, Russia)
equipped with a UV detector operated at 214 nm. The electrophoretic
experiments were performed in an uncoated fused-silica capillary
60 cm × 50 μm I.D. (50 cm effective length). Before use, the capillary
was conditioned for 5 min with 0.5 M HCl, 3 min with water, 5 min with
0.5 M NaOH, and 3 min with water. Additionally, the capillary was
washed for 1 min with 0.5 M NaOH, 1 min with water, and 2 min with
the background electrolyte (BGE) with positive pressure applied at the
injection end before each run. Acquisition of electropherograms was
computer-controlled by Chrom & Spec software v. 1.5.
Where Cfound, Creal, and Cadded are the concentration of analyte after
addition of a known amount of standard into the real sample, the concen-
tration of analyte in real sample, and the concentration of known amount
of the standard which was spiked into the real sample, respectively.
RESULTS AND DISCUSSION
Separation of PRO Enantiomers
In order to obtain the optimum enantioseparation of PRO,
experimental parameters affecting extraction efficiency were
put into optimization. R-enantiomer of PRO was used for iden-
tifying the migration order of enantiomers in CD-modified
CE. The results showed that the S-enantiomer of PRO mi-
grated faster than the R-enantiomer.
Electromembrane Equipment
The DC power supply used was an EPS-600Z model (Paya Pajohesh Pars,
Tehran, Iran) with programmable voltage in the range 0–600 V, providing
currents in the range 0–0.5 A. Platinum wires (diameter 0.2 mm) were used
as electrodes with an interelectrode distance of 5 mm in the sample and
acceptor solutions and were connected to the power supply, which resulted
in an electrical field of 80 V/cm (for a typical 40V DC). The porous hollow
fiber used for immobilization of SLM and for housing the AP was a PP
Q3/2 polypropylene hollow fiber (Membrana, Wuppertal, Germany) with
an internal diameter of 0.60mm, wall thickness of 200μm, and pores of
0.2 μm. It was cut into 8.0-cm segments, cleaned in acetone, and dried prior
to use. Stirring of the solutions was carried out by a Heidolph MR 3001K
magnetic stirrer (Schwabach, Germany).
In this work, four types of CDs (β-CD, HP-β-CD, α-CD, and
HP-α-CD) were used as chiral selectors. The effect of each
CD on chiral recognition was investigated using 100 mM am-
monium acetate as BGE (pH 3.0) containing 5 mM of chiral
selector. α-CD and β-CD possess cavities of different dimen-
sions due to the number of glucopyranose units in their struc-
ture. Therefore, they must have different selectivities towards
the analytes through the so-called analyte-CD inclusion com-
plexation. Although some researchers9,10 have studied chiral
recognition mechanisms of β-adrenergic antagonists with CD
by NMR, the detailed mechanisms underlying the separation
remain unknown. The effect of the chiral selector type on
resolution is shown in Figure 1A. As shown in this figure,
good resolution was obtained with HP-β-CD. This CD has a
suitable cavity so that it provides better interaction of the
PRO enantiomers but, on the other hand, HP-α-CD and α-CD
did not provide a suitable resolution of the enantiomers. This
could be due to the small size of the cavities in HP-α-CD and
α-CD (α-CD has six glucose units), which did not allow effec-
tive stereoselective interaction with PRO.
Extraction Procedure
In this work, a novel EME cell design was used for keeping the
interelectrode distance constant so that a reproducible extraction was
obtained.16 Extractions were carried out by the following procedure:
4 mL of sample solution containing the analyte (pH 3.5) was transferred
into the sample vial. In order to impregnate pores of hollow fiber wall with
the organic solvent, serving as SLM, an 8.0-cm piece of polypropylene
hollow fiber was cut out and dipped in the solvent for 15 s, and then the
excess of organic solvent was removed with a medical wipe. With a
microsyringe, 20 μL of the AP (pH 1.0) was filled into the lumen of hollow
fiber and the end of hollow fiber was sealed using a pair of hot flat-tip
pliers. One of the electrodes, the cathode, was introduced into the lumen
of the fiber and then the fiber containing the cathode, SLM, and the AP
was placed directly into the sample solution. The other electrode, the
The effect of chiral selector concentration on the resolution of
PRO enantiomers was investigated using 100 mM ammonium
Chirality DOI 10.1002/chir

262
TABANI ET AL.
separation of enantiomers was obtained at all voltages. Consid-
ering the good resolution and fast migration time, a voltage of
18 kV was used in the following experiments.
The effect of temperature was investigated by using 80 mM
ammonium acetate solution (pH 2.5, 8 mM HP-β-CD). To
determine the optimum temperature for the separation,
several electrophoretic runs at different cartridge temperatures
(13, 17, 20, 25, and 30°C) were performed and separation of the
enantiomers examined. In this study, a circulating coolant
containing water was used to maintain a constant temperature
inside the capillary cartridge. The resolution of the mentioned
enantiomers decreased slightly with an increase in the temper-
ature from 13–30°C (Table 1). However, at temperatures lower
than 20°C it was noticed that the CE instrument was not
efficient in controlling temperature and migration times were rel-
atively long. Thus, 20°C was chosen to carry out the experiments.
Under the optimized conditions, the migration times of
S-PRO and R-PRO were16.6 and 17.4 min, respectively. Figure 2
shows a typical electropherogram in optimized CE conditions.
However, migration times were not very reproducible and
had about a 30-s shift and their relative standard deviations
(RSDs) were less than 4% (n = 3), even though the capillary
column was washed for 1 min with 0.5 M NaOH, 1 min with
water, and 2 min with the running buffer with a positive pres-
sure applied at the injection end before each run.
Fig. 1. The effect of (A) chiral selector type and (B) chiral selector concen-
tration on the resolution of PRO enantiomers.
Optimization strategy
acetate solution (pH 3.0) containing HP-β-CD over a concentra-
tion range of 2–15 mM (Fig. 1B). As the HP-β-CD concentration
increased from 2–8 mM, an increase in the resolution was
observed. This behavior seems to because of the greater number
of binding sites available to interact with solutes in the BGE
while increasing the chiral selector concentration. As the CD
concentration increased from 8–15 mM the resolution was
relatively constant, while migration times increased from
16.4–28.3 min and from 17.2–29.3 min for S-PRO and R-PRO,
respectively. Considering the successful separation and appro-
priate analysis time, a HP-β-CD concentration of 8 mM was
chosen for the target compounds.
The effect of BGE pH on the migration time and resolution
were investigated. An ammonium acetate solution (100 mM)
with 8 mM HP-β-CD in the pH range of 2.0–5.0 was used.
The resolutions increased with a decrease in pH, because of
a decrease in the velocity of electroosmotic flow (EOF) at
low pH values. Considering the fast migration time and suit-
able peak shape, a BGE of pH 2.5 was used in the following
experiments.
In order to obtain maximum extraction recoveries for
extraction of PRO enantiomers, the effective parameters on
EME efficiency including type of the organic solvent, stirring
rate, applied voltage, extraction time, and pH of donor and
TABLE 1. The effect of temperature on the migration time
(min) and resolution of PRO enantiomers
a
a
Temperature (°C)
tS
tR
Rs
13
17
20
25
30
28.7
20.5
16.5
15.9
14.4
30.1
21.6
17.5
16.7
15.0
1.67
1.65
1.64
1.61
1.52
atR and tS show migration times of R-PRO and S-PRO, respectively.
The effect of BGE was investigated by using 50–125 mM am-
monium acetate solution (pH 2.5, 8 mM HP-β-CD). Increasing
the BGE concentration decreases the EOF, and consequently
an increase in the separation time occurs in CE. Additionally,
increasing the BGE concentration also increases the current
at a constant voltage to the point where adequate thermostating
of the capillary becomes a concern. However, the use of BGE
concentrations higher than 80 mM greatly decreased enantio-
mer resolution, probably due to the high current generated
inside the 50-μm i.d. capillary. Therefore, the use of high BGE
concentrations would require using a capillary with smaller
internal diameter. Based on these observations, an optimum
BGE concentration of 80mM was selected for further analyses.
The effect of applied voltage on the enantioseparation of PRO
was investigated using 80 mM ammonium acetate solution
(pH 2.5, 8 mM HP-β-CD) in the range of 15–22kV. A good
Chirality DOI 10.1002/chir
Fig. 2. Electropherogram of the chiral separation of PRO. Experimental
conditions: capillary: 60 cm (50 cm effective length) × 50 μm i.d; detection:
214 nm; applied voltage: 18 kV; temperature: 20°C; injection: 60 mbar × 5 s;
separation solution: 80 mM ammonium acetate pH 2.5 containing 8 mM HP-
β-CD as chiral selector, concentration of each enantiomer is 20 μg/mL.

ELECTRO MEMBRANE EXTRACTION OF PROPRANOLOL ENANTIOMERS
263
acceptor phases were optimized in a standard aqueous
solution of 100 ng mL-1 PRO. Initially, the influence of the type
of organic solvent and stirring rate on extraction efficiency
was evaluated separately. Then the influence of other factors
(voltage, extraction time, pH of the acceptor and donor
phases) was evaluated by BBD. Finally, the optimized condi-
tions were applied for the extraction of PRO enantiomers
from real biological samples. The experimental design matrix
and data analysis were performed by the Statgraphics Plus
Package, v. 5.1.17
development of mathematical models that permit assess-
ment of the relevance and statistical significance of the fac-
tors being studied. They also facilitate the evaluation of
interaction effects among factors and the optimum operation
conditions are attained by using quadratic response surface
experimental designs.20
Thus, a BBD was applied to optimize the four factors
(voltage, extraction time, acceptor, and donor phase pHs)
and investigate the interaction among these variables. The
examined levels of these factors are given in Table 3. BBD
is an effective design that is used for sequential experimenta-
tion and provides a reasonable amount of information for
testing well fitting and does not require an unusually large
number of design points, and thereby reduces the overall cost
associated with the experiment. According to the number of
experiment equation21: N = 2K (K-1) + C, where K is the
number of variables and C is the number of centre points, K
and C were set at 4 and 8, respectively; 32 experiments had
to be run.
Selection of organic solvent. The chemical nature of the or-
ganic solvent is highly critical to succeed in an electrokinetic
cross-membrane extraction. The flux of analyte is affected by
the gradient of analyte concentration across the SLM, which
is partially determined by the DP-to-SLM distribution ratio,
and this in turn is controlled by the type of solvent used as
SLM. It is known that NPOE, NB, and ENB are efficient
organic solvents for electrokinetic migration of basic drugs
through the SLM,18 while acidic compounds are extracted
by aliphatic alcohols like 1-octanol.19
Thus, in this work three solvents (NPOE, NB, and ENB)
were used as SLM solvents. Other factors were maintained
constant during the optimization (interelectrode distance of
5 mm, stirring rate of 1000 rpm, a 100 V potential difference,
15 min as extraction time, AP pH 2.0, and DP pH 3.0). As seen
in Table 2, recoveries in the range of 4–48% (corresponding
enrichment factors of 10–96) were obtained for both enantio-
mers with various solvents. These results showed that NB
had low recoveries. High enrichment factors for two enantio-
mers were obtained when NPOE was used as SLM.
The design permitted the response to be modeled by a
second-order polynomial fit, which can be expressed as the
following equation:
Y ¼ 16:7633 ꢁ 5:8667 A ꢁ 1:8611 B ꢁ 0:0077 C þ 0:8781 D
ꢁ0:5958 A2 þ 1:1250 AB þ 0:0042 AC ꢁ 0:1333 AD
(4)
ꢁ0:3146 B2 þ 0:0114 BC ꢁ 0:0133 BD ꢁ 0:0002 C2
ꢁ0:0005 CD–0:0099 D2
As seen in the equation above, this model consists of four
main effects, six two-factor interactions, and four curvature
effects.
Influence of stirring rate. As is known, stirring speed plays an
essential role in increasing the kinetics and efficiency of
extraction by increasing the mass transfer and reducing the
thickness of the boundary layer at the interface between the sam-
ple solution and the SLM. Thus, the effect of stirring rate on the
extraction was investigated in the range of 700–1250 rpm. It was
found that the extraction efficiency increases with increasing
stirring rate. Thus, a stirring rate of 1250 rpm was chosen in
the subsequent experiments.
The analysis of variance (ANOVA) results produced the
Pareto chart of main and interaction effects are shown in
Figure 3. According to this figure, pH of the AP has the
largest influence on the peak area and a negative effect upon
the extraction.
TABLE 3. The experimental variables and levels of the
Box–Behnken design (BBD)
Optimization design. The one variable at a time (OVAT)
method (that is, changing one variable at a time while holding
all the others constant) cannot account for interactions
between variables. Such an approach does not allow selecting
variables that have the most influence on the response, and it
does not consider possible interdependence between
variables. On the other hand, a multivariate method allows
simultaneous changes of all chosen factors affecting response
and avoids useless and time-consuming experiments. The
experimental design methodology such as BBD enables
the simultaneous study of several control factors and the
Level
Variable
Key
Lower
Central
Upper
pH of AP
pH of DP
Voltage (volt)
Extraction time (min)
A
B
C
D
1
2
20
5
2
4
110
20
3
6
200
35
TABLE 2. Extraction recovery of PRO enantiomers with dif-
ferent organic solvents in the SLM
Organic solvent
Parameter
Enantiomer
NB
ENB
NPOE
Recovery % (EF)a
R
S
5 (10)
6 (12)
30 (60)
31 (62)
48 (96)
48 (96)
Fig. 3. Pareto chart of the main effects in the BBD.
aThe number of replicates (n = 5).
Chirality DOI 10.1002/chir

264
TABANI ET AL.
The effect of pH of the AP on extraction can be understood
Validation of the Method
using the ion balance (χ) of the system:
In order to verify the practical applicability of the proposed
EME technique, calibration curves were plotted for drug-free
urine and plasma samples under the optimized extraction
conditions (NPOE as SLM; extraction time, 32 min; voltage,
40 V; stirring rate, 1250 rpm; AP, pH 1.0; and DP, pH 3.5),
and figures of merit for the method were evaluated. The
results are summarized in Table 4. To this end, the samples
were spiked with standard solutions of PRO and the extrac-
tion was accomplished after the dilution of urine (1:1) and
plasma samples (1:2) with ultrapure water. The pH values of
all samples were adjusted by dropwise addition of 1 M of
HCl solution so that the final pH of the samples was 3.5.
X
X
iCih
iCio
þ
þ
kCꢂkh
kCꢂko
X
X
χ ¼
(5)
Where Cih is the concentration of the cationic substance in
the DP, C* is the concentration of the anionic substance
kh
in the DP, Ci0 is the concentration of the cationic substance
in the AP, and C*k0 is the concentration of the anionic sub-
stance in the AP. Thus, χ defined as total ionic concentration
in the DP to that in the AP impresses the flux over the mem-
brane.22 According to the Nernst–Planck equation, a maxi-
mum response can be obtained in the minimum value of χ
(decreasing pH of the AP). In low pH of the AP (which causes
decreasing of χ), the analyte is easily released into the accep-
tor solution, leading to increasing of extraction efficiency.
The main mass transfer mechanism in EME is the electroki-
netic migration of the analyte across the SLM into the AP. When
no voltage was applied, almost no extraction was observed.
Thus, the applied voltage across an SLM is an important factor
for efficient extraction of PRO enantiomers in EME. It can be
also shown that voltage and time are two parameters that act
in parallel ways, and their interaction should be considered in
EME optimization.16 The 3D response surface plot of the
response, using eq. (4) when two of the variables are fixed at
the central point and the other two are allowed to vary, are
shown in Figure 4. According to this plot, in an OVAT optimiza-
tion an extraction time of 20min and voltage of 160 volt would
be obtained as optimum values. But there is an antagonistic
effect when they are simultaneously considered, thus an
increase in voltage limits the extraction time and vice versa.
Therefore, the maximum peak area would be obtained at
32min and 40 volts.
Fig. 5. Electropherograms obtained after EME, (A) nonspiked urine 1
sample, and (B) urine 1 sample spiked with 100 ng/mL of each enantiomer.
EME conditions: NPOE as SLM, 40 V potential difference, 32 min as extrac-
tion time, pH of the AP 1.0 and pH of the DP 2.5. CE conditions were the same
as in Figure 2.
Fig. 4. RSM and contour plot obtained by plotting of voltage vs. the extrac-
tion time for PRO enantiomers using the BBD.
TABLE 4. Figures of merit of the proposed extraction procedure
RSD%c
Sample
Urine
Analyte
R2
Slope
LOQa
LODa
Linearitya
EF
Recoveryb
Intra day
Inter day
R
S
R
S
0.998
0.998
0.996
0.996
0.082
0.081
0.065
0.066
20
20
30
30
7
7
10
10
20-300
20-300
30-300
30-300
134
132
110
108
67
66
55
54
4.1
4.2
5.8
6.0
5.2
5.2
7.5
7.1
Plasma
aConcentration is based on ng/mL.
bRecovery was obtained for 80 ng/mL of each enantiomer (n = 5).
cIntraday and interday RSDs% were obtained by five replicate measurements for 80 ng/mL of each enantiomer.
Chirality DOI 10.1002/chir

ELECTRO MEMBRANE EXTRACTION OF PROPRANOLOL ENANTIOMERS
265
The equation of the calibration plots was established by linear
regression of peak area against enantiomer concentration for
both R-PRO and S-PRO standard solutions. The limits of
detection (LODs) and limits of quantification (LOQs) were
determined as analyte concentrations giving rise to signal-
to-noise ratios of 3 and 10, respectively. The LOQs for both
enantiomers were found to be 20 and 30 ng/mL in urine and
plasma samples, respectively. The LODs for both enantio-
mers were 7 and 10 ng/mL in urine and plasma samples,
respectively. Thus, this system is capable of the extraction
and determination of PRO enantiomers with good LODs.
Repeatability or intraday precision was investigated by injecting
five replicates of a standard solution (80 ng/mL) and interday
precision was assessed by injecting the same sample over 5
consecutive days. Intra- and interday precision extractions
varied between 4.1 and 7.5% for both enantiomers (Table 4).
In order to calculate the enrichment factor of each enantiomer,
five replicate extractions were performed at optimal conditions
from an aqueous standard solution containing 80 ng/mL of
each enantiomer. The enrichment factor was calculated as the
ratio of the final concentration of the each enantiomer in the
AP to its concentration in the original solution. The good
enrichment factors were in the range of 132–134 and 108–110
for urine and plasma samples, respectively.
Investigation of Discrimination Between Enantiomers in
Real Samples
Due to the importance of analysis of the drug in biological
samples, final experiments were carried out on four biological
samples (human plasma 1, human plasma 2, urine 1, and
urine 2). The experimental results revealed that only in
sample urine 1, 65 ng/mL of (–)-S-PRO and 98 ng/mL of
(+)-R-PRO enantiomer were determined (Fig. 5A). Sample
urine 1 was taken from a patient under treatment with PRO,
35 min after (regarding Tmax)
23 oral administration of a tablet
containing 40 mg of racemic PRO. As shown in Figure 5A, the
peak area of S-PRO is lower than that of R-PRO. Maybe it is
because of the enantioselective metabolism and that S-PRO
is more pharmaceutically active than R-PRO in treating
angina pectoris.4 The presence of PRO enantiomers was
confirmed by spiking PRO standards into the urine 1 and
reanalyzing it (Fig. 5B).
TABLE 5. Determination of PRO enantiomers in biological samples
Sample
Urine 1
R (ng/mL)
S (ng/mL)
Initial concentration (ng mL-1)
98
94
65
95
RR%a
RSD% (n = 3)
5.1
n.db
95
5.3
n.db
93
Urine 2
Initial concentration (ng mL-1)
RR%a
RSD% (n = 3)
4.8
n.db
80
6.7
n.db
83
5.1
n.db
78
6.7
n.db
82
Plasma 1
Plasma 2
Initial concentration (ng mL-1)
RR%a
RSD% (n = 3)
Initial concentration (ng mL-1)
RR%a
RSD% (n = 3)
5.8
5.5
a100 ng/mL of each PRO enantiomer was added to calculate relative recovery percent (RR%).
bn.d, not detected.
TABLE 6. Comparison of analytical performance data of proposed method with other methods applied for the analysis of PRO
Methoda
Sample preparationa Analyte
Sample type
Plasma
Water
Urine
Eye tissue
Water
Urine
LODb
Organic solvent volume (μL) RSD%
Ref.
HPLC-UV
UPLC-UV
HPLC-UV
LC-MS-MS
LC-MS-MS
LC-MS-MS
LC-MS-MS
SPME
SPE
CM-LPME
LLE
PRO
PRO
PRO
PRO
12
0.14
5
—
1.7
5.9
5.9
8.9
12.0
9.0
4.2
11.2
2.0
2.0
2.3
3.0
7.0
9.0
5.8
6.0
25
26
27
29
30
31
8
5000
190
3
1000
0.0015
500
0.017
0.017
20000
20000
114
114
4
MIP
SPE
PRO
PRO
—
1000
6000
SPE
R-PRO
S-PRO
R-PRO
S-PRO
R-PRO
S-PRO
R-PRO
S-PRO
R-PRO
S-PRO
Waste water
HPLC-UV
CE-UV
LLE
Tablet
40000
4000
—
7
LLE
—
9
28
CEC-UV
CE-UV
SPME
EME
Urine
7
10
10
Untreated Plasma
10
Proposed method
aHigh-performance liquid chromatography (HPLC), ultraperformance liquid chromatography (UPLC), capillary electrophoresis (CE), mass spectrometry (MS),
solid phase microextraction (SPME), solid-phase extraction (SPE), carrier-mediated liquid-phase microextraction (CM-LPME), liquid–liquid extraction (LLE), mo-
lecularly imprinted polymer (MIP), electromembrane extraction (EME).
bAll concentrations are based on ng/mL.
Chirality DOI 10.1002/chir

266
TABANI ET AL.
8. Borges KB, Pupo MT, Bonato PS. Enantioselective analysis of propranolol
To investigate the matrix effect in real samples, 100 ng/mL
and 4-hydroxypropranolol by CE with application to biotransformation
of each enantiomer were spiked to biological samples and
their relative recoveries were determined. As seen in Table 5,
relative recoveries for the spiked samples are in an acceptable
range (78–95%). EME is the only extraction method that does
not require any addition of reagents prior to extraction in
biological samples, whereas all other microextraction tech-
niques need pretreatment of samples. The electrical field is
able to suppress the drug–protein interactions. It is assumed
that the electrical field contributes to break bonds between
drugs and proteins in plasma.24
studies employing endophytic fungi. Electrophoresis 2009;30:3910–3917.
9. Chen J, Du Y, Zhu F, Chen B. Evaluation of the enantioselectivity of
glycogen-based dual chiral selector systems towards basic drugs in capil-
lary electrophoresis. J Chromatogr A 2010;1217:7158–7163.
10. Kumar AP, Park JH. Fast separations of chiral β-blockers on
a
cellulose tris (3,5- dimethylphenylcarbamate)-coated zirconia mono-
lithic column by capillary electrochromatography. J Chromatogr A
2011;1218:5369–5373.
11. Lu M, Zhang L, Qiu B, Feng Q, Xia S, Chen G. Rapid separation and
sensitive detection method for β-blockers by pressure-assisted capillary
electrochromatography-electrospray ionization mass spectrometry.
J
Chromatogr A 2008;1193:156–163.
12. Hemmersbach P, De La Torre R. Stimulants, narcotics and β-blockers:
25 years of development in analytical techniques for doping control. J
Chromatogr B 1996;687:221–238.
CONCLUSION
Investigation of discrimination between enantiomers in real
samples is an interesting issue in enantioseparation. In the
present work, EME combined with CD-modified CE resulted
in a high enrichment factor, good sample clean-up, and
efficient enantioseparation of PRO in biological matrices. Also,
the discrimination between two enantiomers in the extraction
of real samples was observed. The method was compared with
previous works (Table 6). The proposed extraction procedure
has a very low (10 μL) organic solvent consumption. EME
eliminates possible carryover problems because the hollow
fiber is not expensive and can be discarded after each
extraction. This serves to maintain high reproducibility and
repeatability. In addition, the organic solvent immobilized
in the small pores of the hollow fiber prevents large biomole-
cules and insoluble particles in the DP to enter the AP.
Therefore, CE–UV coupling with EME can provide good
and sensitive results for determination of these enantiomers
in biological fluids.
13. Ali I, Gupta VK, Aboul-Enein HY. Chiral resolution of some environmental
pollutants by capillary electrophoresis. Electrophoresis 2003;24:
1360–1374.
14. Pedersen-Bjergaard S, Rasmussen KE. Electrokinetic migration across ar-
tificial liquid membranes: New concept for rapid sample preparation of bi-
ological fluids. J Chromatogr A 2006;1109:183–190.
15. Gjelstad A, Rasmussen KE, Pedersen-Bjergaard S. Electromembrane ex-
traction of basic drugs from untreated human plasma and whole blood un-
der physiological pH conditions. Anal Bioanal Chem 2009;393:921–928.
16. Fakhari AR, Tabani H, Nojavan S, Abedi H. Electromembrane extraction
combined with cyclodextrin-modified capillary electrophoresis for the
quantification
of
trimipramine
enantiomers.
Electrophoresis
2012;33:506–515.
17. Statgraphics Plus 5.1 for Windows Statistical Graphic Corp., Rockville,
MD, Online Manuals 2001.
18. Gjelstad A, Rasmussen KE, Pedersen-Bjergaard S. Electrokinetic
migration across artificial liquid membranes: tuning the membrane
chemistry to different types of drug substances. J Chromatogr A
2006;1124:29–34.
19. Balchen M, Gjelstad A, Rasmussen KE, Pedersen-Bjergaard S. Electrokinetic
migration of acidic drugs across a supported liquid membrane. J Chromatogr
A 2007;1152:220–225.
20. Tarley CRT, Silveira G, Santos WNL, Matos GD, Silva EGP, Bezerra MA,
Miro M, Ferreira SLC. Chemometric tools in electroanalytical chemistry:
methods for optimization based on factorial design and response surface
methodology. Microchem J 2009;92:58–67.
21. Kamarei F, Ebrahimzadeh H, Yamini Y. Optimization of solvent bar
microextraction combined with gas chromatography for the analysis of
aliphatic amines in water sample. J Hazard Mater 2010;178:747–752.
ACKNOWLEDGMENT
Financial support from the Research Affairs of Shahid
Beheshti University is gratefully acknowledged.
LITERATURE CITED
22. Gjelstad A, Rasmussen KE, Pedersen-Bjergaard S. Simulation of flux
during electro-membrane extraction based on the Nernst–Planck equation.
J Chromatogr A 2007;1174:104–111.
23. Mansur AP, Avakian SD, Paula RS, Donzella H, Santos SRCJ, Ramires
JAF. Pharmacokinetics and pharmacodynamics of propranolol in hyper-
tensive patients after ublingual administration: systemic availability. Braz
J Med Biol Res 1998;31:691–696.
24. Eibak LEE, Gjelstad A, Rasmussen KE, Pedersen-Bjergaard S. Exhaustive
electromembrane extraction of some basic drugs from human plasma
followed by liquid chromatography-mass spectrometry. J Pharm Biomed
Anal 2012;57:33–38.
25. Olszowy P, Szultka M, Ligor T, Nowaczyk J, Buszewski B. Fibers with
polypyrrole and polythiophene phases for isolation and determination of
adrenolytic drugs from human plasma by SPME-HPLC. J Chromatogr B
2010;878:2226–2234.
26. Baranowska I, Kowalski B. The development of SPE procedures and an
UHPLC method for the simultaneous determination of ten drugs in water
samples. Water Air Soil Pollut 2010;211:417–425.
27. Zhang L, Su X, Zhang C, Ouyang L, Xie Q, Ma M, Yao S. Extraction and
preconcentration of β-blockers in human urine for analysis with high per-
formance liquid chromatography by means of carrier-mediated liquid
phase microextraction. Talanta 2010;82:984–992.
28. Lin B, Zheng MM, Ng SC, Feng YQ. Development of in-tube solid-phase
microextraction coupled to pressure-assisted CEC and its application to
the analysis of propranolol enantiomers in human urine. Electrophoresis
2007;28:2771–2780.
1. Tran CD, Oliveira D, Grishko VI. Determination of enantiomeric compositions
of pharmaceutical products by near-infrared spectrometry. Anal Biochem
2004;325:206–214.
2. Sanghoo L, Youngjin C, Sangsan L, Karpjoo J. Chiral recognition based on
enantioselective interactions of propranolol enantiomers with cyclosophoraoses
isolated from rhizobium meliloti. J Chiral 2004;16:204–210.
3. Armstrong DW, Ward TJ, Armstrong RD, Beesley TE. Separation of drug
stereoisomers by the formation of β-cyclodextrin inclusion complexes.
Science 1986;232:1132–1135.
4. Ching CB, Fu P, Ng SC, Xu YK. Effect of mobile phase composition on
the separation of propranolol enantiomers using a perphenylcarbamate
β-cyclodextrin bonded chiral stationary phase.
2000;898:53–61.
J Chromatogr A
5. Santoro MIRM, Cho HS, Kedor-Hackmann ERM. Enantiomeric separa-
tion and quantitative determination of propranolol in tablets by chiral
high-performance liquid chromatography. Drug Dev Ind Pharm
2001;27:693–697.
6. Nikolai LN, McClure EL, MacLeod SL, Wong CS. Stereoisomer quantifi-
cation of the β-blocker drugs atenolol, metoprolol, and propranolol in
wastewaters by chiral high-performance liquid chromatography-tandem
mass spectrometry. J Chromatogr A 2006;1131:103–109.
7. Lomsadze K, Salgado A, Calvo E, Lopez JL, Chankvetadze B. Comparative
NMR and MS studies on the mechanism of enantioseparation of propran-
olol with heptakis (2,3-diacetyl- 6-sulfo)-b-cyclodextrin in capillary electro-
phoresis with aqueous and nonaqueous electrolytes. Electrophoresis
2011;32:1156–1163.
Chirality DOI 10.1002/chir

ELECTRO MEMBRANE EXTRACTION OF PROPRANOLOL ENANTIOMERS
267
29. Kadam RS, Kompella UB. Cassette analysis of eight beta-blockers in bo-
quadrupole-linear ion trap mass spectrometry.
2008;1189:374–384.
31. Kolmonen M, Leinonen A, Pelander A, Ojanpera I. A general screening
method for doping agents in human urine by solid phase extraction and
liquid chromatography/time-of-flight mass spectrometry. Anal Chim Acta
2007;585:94–102.
J Chromatogr A
vine eye sclera, choroid–RPE, retina, and vitreous by liquid chromatogra-
phy–tandem mass spectrometry. J Chromatogr B 2009;877:253–260.
30. Gros M, Pizzolato TM, Petrovic M, de Alda MJL, Barcelo D. Trace level
determination of β-blockers in waste waters by highly selective molecu-
larly imprinted polymers extraction followed by liquid chromatography–
Chirality DOI 10.1002/chir