Dispersive solvent-free ultrasound-assisted ionic liquid dispersiveliquid–liquid microextraction coupled with HPLC for determinationof ulipristal acetate

Aiqin Gong a,b, Xiashi Zhu a,n

a College of Chemistry & Chemical Engineering, Yangzhou University, Yangzhou 225002, Chinab Yangzhou Polytechnic Institute, Yangzhou 225127, China

a r t i c l e i n f o

Article history:Received 20 June 2014Received in revised form30 July 2014Accepted 6 August 2014Available online 23 August 2014

Keywords:Ulipristal acetateIonic liquidLiquid–liquid microextractionHPLC

a b s t r a c t

In this paper, a simple and efficient ultrasound-assisted ionic liquid dispersive liquid–liquid microex-traction (UA IL-DLLME) coupled with high-performance liquid chromatography for the analysis ofulipristal acetate (UPA) was developed. UPA could be easily migrated into 1-octyl-3-methylimidazoliumhexafluorophosphate [C8mimPF6] IL phase without dispersive solvent. The research of extractionmechanism showed that hydrophobic interaction force played a key role in the IL-DLLME. Severalimportant parameters affecting the extraction recovery were optimized. Under the optimized conditions,25-fold enrichment factor was obtained and the limit of detection (LOD) was 6.8 ng mL�1 (tablet) or9.3 ng mL�1 (serum) at a signal-to-noise ratio of 3. The calibration curve was linear over the range of0.03–6.0 mg mL�1. The proposed method was successfully applied to the UPA tablets and the real miceserum samples.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

Ulipristal acetate [17α-acetoxy-11β-(4-N,N-dimethyl amino-phenyl)-19-norpregna-4,9-diene-3,20-dione] (UPA), a selectiveprogesterone receptor modulator (Fig. 1), can prevent unintendedpregnancy by delaying ovulation for up to five days after contra-ceptive failure. In August 2010, UPA had gained the FDA’s approvalfor use as an oral emergency contraception tablet in the U.S. withtrade name Ella. Until now it is found that Ella may cause seriousside effects including abdominal pain, menstrual disorder, head-ache, nausea and so on [1]. In addition, the study for UPA to treatcontraceptive gynecological indications (fibroma uteri, adenomyo-sis) and Cushing’s syndrome is in progress [2]. In view of safemedication and investigating pharmaceutical dynamics of drugs, asimple, sensitive analytical procedure is needed to determine UPAin pharmaceutical formulation and in biological fluids. HPLC hasbeen used to determine UPA in bulk [3]. But to the best of ourknowledge, there were few literatures to analyze UPA in biologicalsamples.

An appropriate preconcentration/separation method should bedeveloped due to matrix interference and low concentration ofanalytes in real biological samples before analysis [4]. In recentyears, many preconcentration/separation steps have been oriented

toward the fast development of simplification and miniaturization.In particular, the use of alternative non-contaminant and non-toxic solvents instead of high quantities of organic solvents ispreferred during preconcentration/separation. Solid phase micro-extraction (SPME) and liquid phase microextraction (LPME) havebeen extensively used to preconcentration/separation analytes incomplex matrix with their high ability of sample clean up andanalyte preconcentration, and low consumption of solvents. SPMEwould required a specific device loaded with certain adsorptionmaterial as well as a high-pressure delivery system that would berelatively expensive [4]. Moreover the operation of LPME issimpler and faster than that of SPME (which includes adsorptionprogress and desorption progress). Dispersive liquid–liquid micro-extraction (DLLME), developed by Assadi and co-workers in 2006[5], is a miniaturized form of liquid phase extraction that employsmicroliter volumes of extraction solvent. Compared with othermicroextraction techniques, the advantages of DLLME are simpli-city of operation, rapidness, accuracy, and it has been extensivelyapplied in drugs analysis [4,6–8].

Extraction solvents (such as tetrachloroethylene [9], chloro-benzene [10] and carbon tetrachloride [11]) and dispersive sol-vents (such as methanol [12,13], acetone [9,10] and acetonitrile[11,14]) are usually used in DLLME. Because of the relatively hightoxicity of these conventional chlorinated extraction solvents,developing environment-friendly “green” extraction solvents hasinspired the great interest of examiners. Ionic liquids (IL) has beenmore and more used as extraction solvent in DLLME (IL-DLLME)

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/talanta

Talanta

http://dx.doi.org/10.1016/j.talanta.2014.08.0210039-9140/& 2014 Elsevier B.V. All rights reserved.

n Corresponding author. Tel./fax: þ86 514 7975244.E-mail addresses: xszhu@yzu.edu.cn, zhuxiashi@sina.com (X. Zhu).

Talanta 131 (2015) 603–608

because of its low volatility and low toxicity [4]. Recentlyultrasound-assisted (UA) and temperature-controlled (TC) techni-ques are the most preferred modifications in IL-DLLME [15,16].However, the technique of dispersive solvent-free UA IL-DLLMEhas seldom been applied for extraction of drugs in biologicalsamples.

In this paper, hydrophobic IL 1-octyl-3-methylimidazolium hex-afluorophosphate [C8mimPF6] as extraction solvent of dispersiveliquid–liquid microextraction was first time used, which could becompletely dispersed into the aqueous sample solution by sonica-tion at 313 K without dispersive solvents, and UPA was easilymigrated. The discussion of extraction mechanism showed thathydrophobic interaction force was the main driving force for UPAtransfer from water into IL. The proposed method was successfullyapplied to the real mice serum samples and UPA tablets.

2. Experimental

2.1. Reagents and standards

UPA standard (with purity 99%), UPA tablet (30 mg tablet�1)and blank tablet were kindly provided by Jiangsu LianhuanPharmaceutical Co., Ltd (Jiangsu, China). Methanol, ethanol, acet-onitrile, acetone, ammonia, ammonium chloride, 1-bromobutane,1-bromohexane, 1-bromooctane and triethylamine were pur-chased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai,China). 1-methylimidazole was obtained from Shanghai DaruiSpecialty Chemicals Co.,Ltd (Shanghai, China) and ammoniumhexafluorophosphate from Shanghai Bangcheng Chemical Co.,Ltd (Shanghai, China). Methanol and acetonitrile were chromato-graphic grade, 1-methylimidazole was chemical pure, and all othermaterials were analytical reagent grade and water was distilled,deionised.

UPA stock solution of 2.0 mg mL�1 was prepared by dissolving0.20 g of UPA in 100.0 mL of anhydrous ethanol and kept incoolness and darkness. The stock solution was further dilutedwith anhydrous ethanol to obtain a standard working solution of0.10 mg mL�1 before using.

2.0 mol L�1 NH3–NH4Cl buffer solution (pH 8.0) was preparedby dissolving appropriate amounts of ammonium chloride andammonia.

2.2. Apparatus

The analysis of UPA was carried by a 1200 series liquidchromatography (Agilent Technologies Inc., USA) equipped withphotodiode-array detector (PDA). All absorption spectral record-ings and absorbance measurements were performed on a UV 2501spectrophotometer (Shimadzu, Japan). The pH measurementswere done by a pH S-25 pH meter (Shanghai, China). A DK-S22thermostatic water-bath (Shanghai Jinghong Laboratory Instru-ment Co., Ltd., China) was used to control temperature. A cen-trifuge Model 80-2 (Shanghai Pudong Physical Optics Instrument

Factory, China) was used to accelerate the phase-separationprocess. The microextraction was assisted by a 40 kHz, 100 Wultrasonic generator (KQ 50E Kunshan Ultrasonic Instrument Co.Ltd., Kunshan, China).

2.3. Analytical method

2.3.1. Sample preparationFor UPA tablet, five tablets of UPA was weighed and crushed,

and then sample powder of about one tablet was accuratelyweighed and placed in a 50 mL of beaker and dissolved withanhydrous ethanol. Insoluble excipient was removed by filtrationthrough a 0.45 μm membrane filter. The filtered solution wasdiluted to100.0 mL with anhydrous ethanol and kept in coolnessand darkness before analysis.

For mice serum, abdominal artery blood samples from mice atdifferent time points were collected into heparinized plastic tubes,upon oral administration of 0, 5, 10, 30 mg UPA of 1 kg healthymice. After placing them at 310 K water bath for 1 h, mice serumsamples were obtained after centrifuging the blood samples.According to the method of Chen et al. [17], to eliminate protein,1.0 mL of serum samples was placed in a 10 mL glass tube and4.0 mL of acetonitrile was added. The mixture was shaken for 30 sand centrifuged for 10 min at 3000 rpm. Finally the supernatantwas determined for UPA.

2.3.2. Synthesis of IL[C4mimPF6], [C6mimPF6] and [C8mimPF6] were synthesized

according to Ref. [18], using such materials as 1-bromobutane,1-bromohexane, 1-bromooctane, 1-methylimidazole and ammo-nium hexafluorophosphate.

2.3.3. Extraction procedureTo a 10.0 mL centrifuge tube, 50.0 mL of [C8mimPF6], 1.0 mL of

buffer solution (pH¼8.0) and adequate UPA standard or samplesolutions were added; the solution was diluted to 10.0 mL withdistilled water. After shaken, the mixture was ultrasonicallyextracted for 10 min at 313 K. Then a cloudy mixture was formed.After cooled at 278 K for 15 min, the cloudy solution was centri-fuged for 5 min at 2500 rpm and the IL phase was deposited at thebottom of the tube. Then the upper aqueous phase was removedwith a syringe. The IL phase was diluted with ethanol to 0.4 mL.The resulting analytical solution was homogenized ultrasonicallyand filtered with 0.45 mm filter membrane before HPLC analysis.

2.3.4. HPLC measurementsChromatographic separation of UPAwas performed on an Apollp

C18 column (150�4.60 mm, 5 mm) (Evans Trade Co., Ltd, Shanghai,China). The mobile phase was a mixture of methanol and 0.05%triethylamine (90:10, v/v) at a flow rate of 1.0 mL min�1. Theinjection volume was 10.0 mL and column temperature was keptat 303 K. The monitoring wavelength was 305 nm and referencewavelength and bandwidth were 350 nm and 4 nm, respectively.

2.3.5. Determination of partition ratioThe partition ratio of UPA in ILs (i.e. [C4mimPF6], [C6mimPF6],

[C8mimPF6]) and water were determined. The partition ratio DIL/W

was calculated according Eq. (1) [19]:

DIL=W ¼ Ci�Cf

Cf� Veq

VILð1Þ

N

CH3

H3C

O

H

H

CH3O

CH3

O CH3

O

H

Fig. 1. The structure of ulipristal acetate.

A. Gong, X. Zhu / Talanta 131 (2015) 603–608604

where Ci and Cf are the concentration of UPA in water phase beforeand after extraction, Veq is the volume of water phase, and VIL isthe volume of IL phase.

3. Results and discussion

3.1. Optimization of extraction conditions

Single factor experimental scheme due to its simplicity wasused to optimize extraction parameters in this paper (such astypes and amount of extraction solvent, solution pH, extractiontime and temperature, cooling time and centrifugation time) [20].All experiments were performed in triplicates (n¼3). The extrac-tion recovery (ER) was calculated based on Eq. (2):

ER%¼ Cex � Vex

C0 � V0� 100 ð2Þ

The Cex and C0 are the concentration of analyte in the extractionphase and the initial analyte concentration in the sample solution,respectively. Vex and V0 are the volumes of extraction phase andsample solution, respectively.

3.1.1. Selection of extraction solventCharacteristics of ILs, such as solubility in water, the viscosity

and extraction capacity, play a key role in influencing the extractionrecovery. When fixing the anion of IL, these characteristics areaffected by the cationic part. In this work three hydrophobic ILs,including [C4mimPF6], [C6mimPF6] and [C8mimPF6], were investi-gated. According to their solubility, 140.0 mL [C4mimPF6], 70.0 mL[C6mimPF6], and 40.0 mL [C8mimPF6] were selected as extractionsolvents in the absence of dispersion solvent [7]. Fig. 2 showed thatthe extraction recoveries (ER) were all over 90.0% in three ILs and ahigher ER was obtained in [C8mimPF6]. Furthermore, adsorptioncapacities of three ILs for UPA were 1.99 mg g�1 ([C4mimPF6]),9.33 mg g�1 ([C6mimPF6]) and 18.8 mg g�1 ([C8mimPF6]) (Fig. 2),respectively. With its higher adsorption capacity and ER,[C8mimPF6] was chosen for following examination.

The effect of volume of [C8mimPF6] on ER was shown in Fig. 3.The highest ER was achieved when 40.0–70.0 mL of [C8mimPF6]was employed. Therefore 50.0 mL of [C8mimPF6] was selected inthe work.

3.1.2. Dispersion solvent freeThe dispersion degree of extraction solvents plays a crucial role

in DLLME. The smaller fine droplet of extraction solvent forms, thehigher extraction recovery achieves. Ultrasound energy or hightemperature combined with dispersive solvents was usually usedin DLLME to improve extraction effect [7,8]. In this paper, theresults showed that the ER would be over 95.0% without anydispersive solvents. So in this work, dispersive solvent was free.

3.1.3. Effect of pHIn general, pH value of sample solution determines the exis-

tential state of analytes, thus affecting extraction recovery. In thispaper the effect of sample solution pH value in the range of2.0–12.0 on the extraction recovery was examined. Fig. 4 showedthat the ER of UPA increased with pH from 2.0 to 7.0, and reachedthe maximum at pH 7.0 (ER495.0%), after that almost unchangedwith further increasing pH. The log Dow (Dow, octanol–water

0

20

40

60

80

100

120 extraction recovery (%)

Ext

ract

ion

reco

very

(%)

0

10

20

30

40

50

adsorption capacity (mg g-1)

[C8mimPF6][C6mimPF6][C4mimPF6]

Ads

orpt

ion

capa

city

(mg

g-1)

Fig. 2. Effect of the kind of extraction solvents on ER of UPA and adsorptioncapacity. Extraction conditions: sample volume, 10.0 mL; sample amount, 10.0 mg;pH, 8.0; ultrasonic temperature, 313 K; ultrasonic time, 10 min; cooling tempera-ture, 278 K; cooling time, 15 min; centrifugation time, 5 min. The error bars werestandard deviation.

20 30 40 50 60 700

20

40

60

80

100

120

Ext

ract

ion

reco

very

(%)

V[C8mimPF6] ( L)µ

Fig. 3. Effect of [C8mimPF6] volume on ER of UPA. Extraction conditions: samplevolume, 10.0 mL; sample amount, 10.0 mg; IL, [C8mimPF6]; pH, 8.0; ultrasonictemperature, 313 K; ultrasonic time, 10 min; cooling temperature, 278 K; coolingtime, 15 min; centrifugation time, 5 min. The error bars were standard deviation.

0 2 4 6 8 10 12 140

20

40

60

80

100

120

Extra

ctio

n re

cove

ry (%

)

pH

Fig. 4. Effect of pH on ER of UPA. Extraction conditions: sample volume, 10.0 mL;sample amount, 10.0 mg; [C8mimPF6] volume, 50.0 mL; ultrasonic temperature,313 K; ultrasonic time, 10 min; cooling temperature, 278 K; cooling time, 15 min;centrifugation time, 5 min. The error bars were standard deviation.

A. Gong, X. Zhu / Talanta 131 (2015) 603–608 605

partition coefficient) values of UPA are higher than 4.0 (log Dow is4.18 at pH 5.5 or 4.47 at pH 7.4), which means UPA is easily solublein lipids and hardly miscible with water, and the lipophilicproperty of UPA is stronger at pH 7.4 than at pH 5.5 [21]. UPAwould more easily distribute into the hydrophobic IL in alkalinemedium. So a pH of 8.0 was used for all extraction experiments.

The effect of ionic strength was tested by adding KCl to UPAsolutions. The results showed that ER was significantly decreasedwhen the concentration of KCl was over 0.8 mol L�1. In this study,1.0 mL of NH3–NH4Cl buffer solution (2.0 mol L�1) added into10.0 mL sample solution could satisfy the requirement of ionicstrength.

3.1.4. Effect of temperatureTemperature has a significant effect on IL solubility in water.

The solubility and dispersion degree of IL will be improved withincreasing temperature that will accelerate the transfer of UPAfromwater to IL. In this work, the effects of extraction temperaturewere evaluated in the range of 288–343 K with an ultrasound timeof 10 min. The ER was observed to increase with temperature from288 to 308 K, and then remain constant up to 343 K. Therefore theextraction temperature of 313 K was selected in this study.

3.1.5. Effect of ultrasonic timeDispersion is the key step whether extraction is successfully

carried out or not. As a key procedure in IL-USA-DLLME, ultra-sound can accelerate the formation of fine dispersive mixtures,and result in higher recoveries [22]. In this work, the sonicationtime was evaluated in the range of 2 min to 20 min. The experi-mental results indicated that the maximum ER (95.0%) could beattained within 5 min and longer extraction time would not affectthe ER. In examination it was also found that when the ultrasound

irradiation was not applied and the sample solution was intenselyshaken for 2 min, the ER could also achieved 95.0%. In order toalleviate manual operation, ultrasound time of 10 min was chosen.

3.1.6. Effect of cooling and centrifugation timePhase separation can be improved with an additional cooling

stage due to the decreased solubility of ILs in water [23]. In thiswork, the effect of the cooling time on the ER was assayed in therange of 5–30 min when fixing cooling temperature 278 K. Theresults showed a cooling time of 15 min was sufficient to achievethe maximum ER. In the study cooling time of 15 min was chosen.

For evaluating the effect of centrifugation on phase separation,the centrifugation time was studied in the range of 1–10 min at aconstant rate of 2500 rpm after cooling. 5 min was found to beenough to achieve complete phase separation.

3.1.7. Effect of solution volumeThe effect of solution volume on ER was examined from 5.0 mL

to 30.0 mL when fixing the amount of UPA at 10.0 μg. The resultsshowed that the ER would be less than 85.0% when the dilutionvolume of sample solution was over 15.0 mL. It is because thedissolved amount of [C8mimPF6] will increase with enhancingsolution volume thus affecting its extraction ability. In this work,the sample volume of 10.0 mL was adopted. The IL phase wasdiluted to 0.4 mL with ethanol after extraction for HPLC determi-nation, and the preconcentration factor (defined as the volumeratio of dilution sample solution and IL phase, i.e. 10.0/0.4) was 25.

3.2. Analytical performance

As shown in Fig. 5(C) (chromatogram of UPA standard solution),the retention time of UPA was about at 2.70 min. Comparing withFig. 5(C) UPA was undetectable and there were no interferencepeaks in blank tablet (Fig. 5(A)) and blank serum (Fig. 5(B)),therefore blank tablet and blank serum were used for the methodvalidation.

3.2.1. Linearity and limits of detectionFor evaluating matrix effect, a statistical comparison between

the standard curve and working curve was made. The workingcurves were got with spiking the standard directly into blanktablet and blank serum and extracting under the same conditions.The results were listed in Table 1. The Student’s test was appliedand the statistical analysis indicated that the difference betweenthe slope of working curve of blank tablet and standard curve wasnot obvious, but the difference between the slope of workingcurve of blank serum and standard curve was significant (P¼0.95).So working curve was used to determine UPA in UPA tablet andmice serum. As shown in Table 1, the calibration graphs werelinear over the concentration ranges of 0.03–6.0 mg mL�1, and thelimits of detection (LOD) (the lowest concentration yielding asignal-to-noise ratio of 3) were 6.8 ng mL�1 (blank tablet) and9.3 ng mL�1 (blank serum). When administered by mouth at adose of 30 mg, ulipristal acetate is rapidly absorbed. The maximummean serum concentration (Cmax)7the standard deviation (SD) of176789 ng mL�1 was observed at approximately 1 h [24], which

0 1 2 3 4 5 6

-2

0

2

4

6

8

10

12

14

16

18

20

22

24

D

E

BC

A

mAU

t(min)

U

Fig. 5. Chromatograms of blank tablet (A), blank serum (B), standard solution (C),UPA tablet (D) and mouse serum sample (E) (collected at 1 h after oral adminis-tration of 10 mg UPA of 1 kg healthy mice). The concentration of standard solution:0.10 μg mL�1. U: UPA.

Table 1Linearity parameters and LOD of the proposed method in different matrices.

Sample Linear range(mg mL�1)

Slope7SD (n¼3) Intercept7SD(n¼3)

Correlationcoefficient

LOD(ng mL�1)

Standard solution 0.02–6.0 482.174.2 8.9270.35 0.9985 5.7Blank tablet 0.03–6.0 483.976.1 12.970.58 0.9963 6.8Blank serum 0.03–6.0 495.075.0 37.771.6 0.9989 9.3

A. Gong, X. Zhu / Talanta 131 (2015) 603–608606

means the method can satisfy the determination requirement ofUPA in serum.

3.2.2. Precision and repeatabilityTo evaluate intraday and interday precisions, analysis of UPA at

three concentration levels (0.50, 1.0, 2.0 μg mL�1) was carried outby performing five experiments on the same day using the sameanalyte solution and over five consecutive days using differentsolutions. The intraday and interday RSD values ranged from 2.5%to 3.6% and from 3.9% to 5.1% for blank tablet, and from 1.9% to3.8% and from 3.6% to 5.5% for blank serum, respectively, reflectingthe usefulness of the method in routine use.

3.2.3. TruenessThe trueness of the proposed method was evaluated by

determining different UPA tablets and mice serum (throughstandard addition technique). The BIAS% was calculated accordingto the following equation: (detected content-stated content)/stated content�100. The data were illustrated in Table 2. Obtainedvalues of BIAS% ranged from þ3.00% to þ5.30%, from þ3.50% toþ6.33% for UPA tablet and mice serum, respectively, suggestingthe analytical utility of the UA IL-DLLME for UPA determination.

The stability of UPA in serum was investigated by extractingand analyzing the serum samples collected at 1 h after oraladministration of 10 mg UPA of 1 kg healthy mice. The resultsindicated that UPA in serum samples was stable for at least 7 days.

3.2.4. Sample determinationComparing the chromatograms of UPA tablet (Fig. 5(D)) and

mice serum (Fig. 5(E)) with UPA standard solution (Fig. 5(C)), thereproducibility of the retention time can satisfy analysis require-ment. Table 3 showed analytical results for UPA tablet and miceserum using the proposed methodologies. The result of UPA tabletobtained by the proposed method was in good agreement with thelabel value. The statistical t-test (P¼0.95) was used to compare theresults, which showed that there was no significant differencebetween them. From Table 3 it was also known that the UPAconcentration in mice serum would rise with increasing intragas-tric administration dose and the maximum serum concentrationwas obtained at 1 h, which was consistent with that of theliterature [24].

3.3. Discussion of extraction mechanism

The extraction effect of UPA by ILs depends on the dissolutiondegree of UPA in IL. The hydrophobic UPA is more likely to bedissolved by hydrophobic ILs than by water based on the hydro-phobic interaction. In this work hydrophobic interaction forcebetween UPA and ILs was verified by partition ratio and some

thermodynamic parameters depending on temperature such asenthalpy change, Gibbs energy and entropy change for elaboratingextraction mechanism.

3.3.1. Partition ratio DIL/W

The partition ratio DIL/W is the concentration ratio of UPA in twoimmiscible phases at partition equilibrium, and reflects differentialsolubility of UPA between two phases (ILs-H2O). The DIL/W and ERof UPA in three ILs at 313 K were shown in Table 4. It could be seenfrom Table 4 that (1) the DIL/W was all over 103 in three ILs, i.e. theconcentration ratio of UPA in ILs and in water was 4103,indicating there was stronger interaction between UPA and ILsthan that of UPA and water, and UPA was more easily distributedin ILs; (2) the DIL/W of UPA in the ILs followed the order:[C8mimPF6]4[C6mimPF6]4[C4mimPF6] and the order DIL/W ofwas consistent with the order of ER. It is because the hydropho-bicity of ILs will enhance with increasing the length of alkyl chainsin imidazole ring. The DIL/W and ER were increased from[C4mimPF6] to [C8mimPF6], indicating hydrophobic interactionforce played a key role on the partition of UPA between ILsand water.

3.3.2. Extraction thermodynamic parametersFrom a thermodynamic perspective, the partition of UPA can be

regarded as a transfer process of the UPA molecules from waterphase to the IL phase. At a given temperature, the changes in Gibbsenergy ðΔG0

T Þ, enthalpy ðΔH0T Þ and entropy ðΔS0T Þ of such a transfer

process can be calculated from the partition data through follow-ing equations:

ln DIL=W ¼ Cþð�ΔH0T=RTÞ ð3Þ

[25]

ΔG0T ¼ �RT ½ ln DIL=W þ lnðVm;IL=Vm;wÞ� ð4Þ

[19]

TΔS0T ¼ΔH0T �ΔG0

T ð5Þ

Among them C is constant, R is molar gas constant(8.314 J mol�1 K�1), T is absolute temperature, and Vm,IL and Vm,W

are molar volumes of ILs and water.

Table 2The trueness evaluation of the proposed method.

Sample Stated content(mg mL�1)

Detected content(mg mL�1)a

BIAS (%)

UPA tablet-1

0.10 0.104 þ4.0

UPA tablet-2

0.20 0.206 þ3.0

UPA tablet-3

0.30 0.316 þ5.3

Serum-1b 0.10 0.105 þ5.0Serum-2b 0.20 0.207 þ3.5Serum-3b 0.30 0.319 þ6.3

a Mean for three independent determinations.b Collected at 1 h after oral administration of 10 mg UPA of 1 kg healthy mice.

Table 3Analytical results of UPA in real sample.a

Sample Label value Intragastricdose

Collectedtime (h)

Found7SDa

Tablet 30 mg tablet�1 / / 29.770.98 mg tablet�1

Miceserum

/

5 mg kg�1

0.5 0.6570.04 mg mL�1

1 4.6170.31 mg mL�1

2 1.4670.04 mg mL�1

10 mg kg�1

0.5 1.3270.10 mg mL�1

1 8.1270.59 mg mL�1

2 3.1070.13 mg mL�1

30 mg kg�1

0.5 2.9270.17 mg mL�1

1 22.870.99 mg mL�1

2 6.0370.15 mg mL�1

a Mean for three independent determinations.

Table 4The DIL/W and ER of UPA in different IL.

IL [C4mimPF6] [C6mimPF6] [C8mimPF6]

DIL/W, �103 3.01 6.77 11.5ER (%) 92.3 94.9 99.4

A. Gong, X. Zhu / Talanta 131 (2015) 603–608 607

It is clear that values of ΔH0T can be directly obtained from the

slope of the linear equation between lnDIL/W and 1/T shown in Eq. (3).Molar volumes of ILs and water (got by Ref. [19]), lnDIL/W, ΔH0

T , andΔG0

T and TΔS0T calculated with Eqs. (4) and (5) under differenttemperature were shown in Table 5. From Table 5, the followingconclusions were drawn that (1)ΔH0

T 40 and partition ratios in threeILs were increased with raising temperature, illustrating that extrac-tion process was a endothermic process and higher temperature wasbeneficial to extraction; (2) ΔG0

T o0, pointing out extraction wasspontaneous; (3) ΔS0T 40 and

��ΔH0

T

��⟨��TΔS0T

��, indicating that the

transfer process of UPA from water to IL was driven by entropychange. It is generally recognized that hydrophobic interaction force ismain feature of entropy control [26]. Based on these results, it could bededuced that hydrophobic interaction might be the main driving forcefor the transfer of UPA molecules from water to the IL phases.

4. Conclusion

In this paper a new and environmental friendly ultrasound-assisted ionic liquid dispersive liquid–liquid microextractioncoupled with HPLC-PDA was developed for the determination ofUPA in dosage form and biological sample. Without any dispersivesolvent the ER could be over 95.0% under ultrasonication. Thedeveloped method can provide analytical technical support forUPA pharmacokinetics examination. Mechanism discussion ofextraction indicated that hydrophobic interaction might be themain driving force for UPA extracted in IL phase.

Acknowledgements

The authors acknowledge the financial support from the NationalNatural Science Foundation of China (21375117, 21155001) and aproject funded by the Priority Academic Program Development of

Jiangsu Higher Education Institutions and the Foundation of theExcellence Science and Technology Invention Team in YangzhouUniversity and the Graduate Innovation Project Foundation of Jiangsuprovince (CXLX13_895).

References

[1] B.Q. Ma, Shanghai Med. Pharm. J 31 (2010) 315–316.[2] G.F. Yao, Chin. J. Med. Chem. 21 (2011) 78–80.[3] WO 2004/078709.[4] J.N. Sun, Y.P. Shi, J. Chen, J. Chromatogr. B 879 (2011) 3429–3433.[5] Y.F. Zhang, H.K. Lee, Anal. Chim. Acta 750 (2012) 120–126.[6] C. Noelia, P.P. Francisco, I. de la Calle, B. Carlos, L. Isela, Microchem. J. 99 (2011)

246–251.[7] S.Q. Gao, X. Yang, W. Yu, Z.L. Liu, H.Q. Zhang, Talanta 99 (2012) 875–882.[8] A.J. Toraj, F. Nazir, S. Mojtaba, P. Meghdad, J. Pharm. Biomed. Anal. 85 (2013)

14–20.[9] R. Mohammad, A. Yaghoub, M.H. Mohammad-Reza, A. Elham, A. Fardin,

B. Sana, J. Chromatogr. A 1116 (2006) 1–9.[10] D. Nagaraju, S.D. Huang, J. Chromatogr. A 1161 (2007) 89–97.[11] M.A. Farajzadeh, M. Bahram, J. Jönsson, Anal. Chim. Acta 591 (2007) 69–79.[12] M. Hatami, Elham Karimnia, Khalil Farhadi, J. Pharm. Biomed. Anal. 85 (2013)

283–287.[13] J.M. Padro´, M.E. Marso´n, G.E. Mastrantonio, J. Altcheh, F. Garcı´a-Bournissen,

M. Reta, Talanta 107 (2013) 95–102.[14] X. Liu, R. Fu, M. Li, L.P. Guo, L. Yang, Chin. J. Anal. Chem. 41 (2013) 1919–1922.[15] J.H. Zhang, H.X. Gao, B. Peng, S.Q. Li, Z.Q. Zhou, J. Chromatogr. A 1218 (2011)

6621–6627.[16] M.J. Trujillo-Rodríguez, P. Rocío-Bautista, V. Pino, A.M. Afonso, Trends Anal.

Chem. 51 (2013) 87–106.[17] W.Y. Chen, H. Li, W.H. Luo, Chin. J. Mod. Med. 17 (2007) 1665–1673.[18] F.Y. Zhao, J.Y. Wang, H.J. Liu, R.J. Liu, Chem. Reagents 29 (2007) 229–231.[19] Y.C. Pei, Doctoral Dissertation of Lanzhou University (2008).[20] P. Arbeláez, F. Borrull, R.M. Marcé, E. Pocurull, Talanta 125 (2014) 65–71.[21] ⟨http://www.chemspider.com/Chemical-Structure.115762.html⟩.[22] S.Y. Dong, Q. Hu, Z. Yang, R. Liu, G.Q. Huang, T.L. Huang, Microchem. J. 110

(2013) 221–226.[23] R. Liu, J. Liu, Y. Yin, X. Hu, G. Jiang, Anal. Bioanal. Chem. 393 (2009) 871–883.[24] ⟨http://tec.k8008.com/html/124/124073.html⟩.[25] X.M. Fu, S.G. Dai, Y. Zhang, Chin. J. Anal. Chem 34 (2006) 598–602.[26] H. Bianco-Peled, S. Gryc, Langmiur 20 (2004) 169–174.

Table 5M volumes of ILs and water, lnDIL/W and thermodynamic parameters of UPA extracted by different IL under different temperature.

ParametersT (K)

298.15 308.15 318.15 328.15

[C4mimPF6] 7.75 7.96 8.22lnDIL/W [C6mimPF6] 8.18 8.54 9.04

[C8mimPF6] 8.54 9.06 9.70

Molar volumes (cm3 mol�1)a

[C4mimPF6] 207.6 209.4 211.2 213.0[C6mimPF6] 241.6 243.1 244.6 246.2[C8mimPF6] 275.7 277.4 279.1 280.9H2O 18.05 18.11 18.18 18.26

ΔH0T (kJ mol�1)

[C4mimPF6] 21.8[C6mimPF6] 35.9[C8mimPF6] 48.9

ΔG0T (kJ mol�1)

[C4mimPF6] �25.3 �26.7 �28.2 �30.0[C6mimPF6] �26.7 �28.5 �30.8 �32.9[C8mimPF6] �27.9 �30.2 �32.9 �35.6

TΔS0T (kJ mol�1)[C4mimPF6] 47.1 48.5 50.0 51.8[C6mimPF6] 62.6 64.4 66.7 68.8[C8mimPF6] 76.8 79.1 81.8 84.5

a Ref. [19].

A. Gong, X. Zhu / Talanta 131 (2015) 603–608608