Novel Binary Solvents-Dispersive Liquid—LiquidMicroextraction (BS-DLLME) Method for Determinationof Patulin in Apple Juice Using High-PerformanceLiquid Chromatography

Mehdi Maham & Rouhollah Karami-Osboo &

Vahid Kiarostami & Syed Waqif-Husain

Received: 11 May 2012 /Accepted: 19 July 2012 /Published online: 9 August 2012# Springer Science+Business Media, LLC 2012

Abstract A simple and rapid binary solvents-based disper-sive liquid–liquid microextraction (BS-DLLME) methodhas been developed for determination of patulin (PAT) inapple juice followed by high-performance liquid chroma-tography. This method involves the use of an appropriatemixture of miscible binary extraction solvents and dispersersolvent to form fine droplets of extractant in a samplesolution. Parameters affecting extraction efficiency such asthe type and volume of high-density extraction solvent, thevolume of ethyl acetate, the kind and volume of dispersersolvent, and salt addition were investigated and optimized.The detection and quantification limits were 2.0 and10.0 μg L−1, respectively. The relative standard deviationfor five measurements of 25 μg L−1 of PAT was 3.8 %. Therelative recoveries of PAT from apple juice samples atspiking levels of 25, 50, and 75 ng mL−1 were in the rangeof 91.3–95.2 %.

Keywords Binary solvents-dispersive liquid–liquidmicroextraction .Patulin .Ethylacetate .Apple juice .HPLC

Introduction

Mycotoxins are secondary metabolites produced by fungi.Mycotoxins can appear in the food chain as a result offungal infection of crops and they have been described asextremely toxic and a potent carcinogen compound forhumans (Karami-Osboo et al. 2012). Patulin (PAT) asone of the important mycotoxin in apples and appleproducts is produced by several fungal species of Pen-icillium, Aspergillus, and Byssochlamys (Baert et al.2007; Leggott and Shephard 2001). The European Com-mission recommends a maximum level for PAT in dif-ferent apple products: 50 μg kg−1, for apple fruit juiceand apple juice ingredients in other beverages;50 μg kg−1 in spirit drinks, ciders, and other fermenteddrinks; 25 μg kg−1 for solid apple products, such asapple puree; and 10 μg kg−1 for apple productsintended for infants and young children (Li et al. 2007).

Sample preparation is often a critical and the most te-dious part of analysis. It is also the main source of error inan analytical method. Traditional sample preparation meth-ods such as liquid–liquid extraction (LLE) and solid-phaseextraction (SPE), which use large volumes of toxic organicsolvents, have been widely used for the enrichment of PATin the apple juice samples (Li et al. 2007; Funes and Resnik2009; Gökmen et al. 2005; Murillo et al. 2008). Ethylacetate has been used as the most common extractionsolvent for the pre-concentration of PAT (Li et al. 2007;Murillo et al. 2008; Iha et al. 2009; Valle-Algarra et al.

M. Maham : S. Waqif-Husain (*)Department of Chemistry, Science and Research Branch, IslamicAzad University,P.O. Box 14515-775, Poonak-Hesarak,Tehran, Irane-mail: syedwaqifhusain@yahoo.com

R. Karami-OsbooStudent Research Committee,Shiraz University of Medical Sciences,Shiraz, Iran

V. KiarostamiChemistry Department, North Tehran Branch,Islamic Azad University,Tehran, Iran

Food Anal. Methods (2013) 6:761–766DOI 10.1007/s12161-012-9483-6

2009). Quantitative measurement of PAT in apple juicesamples by high-performance liquid chromatography(HPLC)-UV can be performed by peak separation of PATand 5-hydroxymethylfurfural (HMF), which naturallyoccurs in apple and is the main interference for PAT analysis(Murillo et al. 2008; Iha et al. 2009) .

A miniaturized LLE procedure which uses microlitervolumes of extraction solvent has been reported as disper-sive liquid–liquid microextraction (DLLME)(Rezaee et al.2006) and has been used in separation and pre-concentrationof mycotoxins (Campone et al. 2011a, b) in different ma-trixes, and also in food samples(Boonchiangma et al.2012; Abdolmohammad-Zadeh and Sadeghi 2010; Fu etal. 2009; Zhang et al. 2012; Almeida et al. 2012; Cunhaand Fernandes 2011). However, in this method, theselection of extraction is limited to the solvents whichare heavier than water such as chlorinated solvents andcarbon disulfide. Therefore, we have developed a sim-ple, rapid, inexpensive, and eco-friendly method (binarysolvents-based (BS) DLLME)) for separation and deter-mination of PAT from apple juice samples with en-hanced results. In this method, a mixture of high-density solvent with another low-density organic solventhas been employed, adding a new dimension to theearlier DLLME method with wide selection of effectivesolvents.

Materials and Methods

Reagents and Materials

PAT and HMF were obtained from Sigma-Aldrich (USA).Water and acetonitrile were of HPLC grade and otherchemicals such as acetone, ethanol, methanol, chloroform,carbon tetrachloride, dichloromethane, carbon disulfide,ethyl acetate, and sodium chloride used in this study wereof analytical grade from E. Merck (Germany). A stocksolution of PAT was prepared by dissolving 1.0 mg PAT in10 mL ethanol. The stock solution was stored at −20 °C.The apple juice samples were provided from a supermarket(Tehran, Iran).

Sample Preparation

Apple juice samples were stored at room temperaturebefore use. Once opened, they were stored in specificfood containers at 4 °C and analyzed within 5 days.The fresh juice was centrifuged at 3,500 rpm for15 min, and then the supernatant was filtered througha 0.45-μm membrane filter. Afterwards, the filtrate wasdiluted at 1:4 ratio with deionized water and used forBS-DLLME procedure.

Instrumentation

A high-performance liquid chromatography systemequipped with auto sampler (Waters 717), binary HPLCpump (Waters 1525), and a dual λ absorbance UV detector(Waters 2487) were used for the analysis. A chromolithHPLC column (15 cm, Merck) was used for separation at40 °C. A mixture of water and acetonitrile (97:3) at a flowrate of 1 mL min−1 was used as mobile phase in isocraticelution mode. The detection was performed at a wavelengthof 276 nm. A centrifuge model, Celements GC-200, wasused for separation of the dispersive phase.

Results and Discussion

In order to obtain high extraction efficiency, important factorswhich may influence the performance of BS-DLLME, such asthe type and volume of high-density extraction solvent, thevolume of ethyl acetate, the kind and volume of dispersersolvent, and salt addition, should be evaluated in detail. Opti-mization was accomplished on water samples, under differentexperimental conditions, and extraction recovery was appliedto assess the effect of different parameters.

Effect of Type of High-Density Extraction Solvent

In a preliminary optimization step, the ability of several sol-vents with higher density than water and different polarities,such as chloroform (CHCl3, density 1.49 gmL−1), carbontetrachloride (CCl4, density 1.59 gmL−1), dichloromethane(CH2Cl2, density 1.32 gmL−1), and carbon disulfide (CS2,density 1.26 gmL−1), to extract the target analyte was com-pared. A series of experiments was performed by using1,000 μL acetonitrile as disperser solvent and 60μL of severalkinds of extraction solvents. According to the results (Fig. 1.),the best performance was obtained when chloroformwas used

0

200

400

600

800

1000

1200

1400

Carbon tetracloride Chloroform Carbon disulfide

Pea

k ar

ea

Extraction solvent

Fig. 1 Effect of type of the high-density extraction solvent on theextraction efficiency. Extraction conditions: concentration of PAT,50 μg L−1; volume of extraction solvent, 60 μL; disperser solventand its volume, 1.00 ml acetonitrile; no salt addition

762 Food Anal. Methods (2013) 6:761–766

as the extraction solvent. So, chloroform was chosen in thesubsequent experiments.

Effect of Volume of CHCl3

The volume of CHCl3 was an important factor that couldaffect the extraction performance. To investigate the effect ofthe CHCl3 volume on the extraction efficiency, 1,000 μL ofacetonitrile containing different volumes of chloroform (40,50, 60, 70, and 80 μL) was injected into the sample solutions,while the other experimental conditions were kept constant.Results (Fig. 2) show that the extraction efficiency increaseswith the increase of the volume of chloroform from 40 to60 μL. Above 60 μL of chloroform, the recovery decreasesfor PAT, probably due to the decrease in the ratio between theacetonitrile and chloroform. The decreased ratio lowers thenumber of droplets available for extraction, hence reducingthe extraction recovery. Based on the obtained results, 60 μLwas selected as the optimal CHCl3 volume.

Effect of Ethyl Acetate and Its Volume

The most important factor that influenced on extraction effi-ciency of PATwas the effect of ethyl acetate as lighter extrac-tion solvent in BS-DLLME. As mentioned above, ethylacetate is the most common extraction solvent for the pre-concentration of PAT in the apple juice samples. Ethyl acetate(C4H8O2; density 0.897 gmL−1) could not be used as extrac-tion solvent in conventional DLLME because of its lowerdensity than water. To form a miscible mixture of extractionsolvents with high density to be deposed after centrifuging anto examine the effect of ethyl acetate on the extraction effi-ciency, different volumes of ethyl acetate and a fixed volumeof chloroform (60 μL) were used for BS-DLLME procedure.Results showed that (Fig. 3) by increasing the volume of ethyl

acetate up to 190 μL, the extraction efficiency increaseddue to the good ability of ethyl acetate in extractingPAT and also increasing the volume of the sedimenteddroplet (105±5 μL as compared with 50±5 μL obtainedwith conventional DLLME). It is important to note that,at the optimum amount of ethyl acetate, the extractionefficiency was about sevenfold higher than the oneswhen only chloroform was used as extraction solvent.The extraction efficiency was decreased by increasingthe volume of ethyl acetate from 190 μL due to reduc-tion of binary system density. After adding 230 μL ofethyl acetate, no droplet was observed at the bottom ofthe test tube by centrifuging and all of the binarysolvents appeared on the surface of water.

Effect of Type of the Disperser Solvent

The main criterion for choosing a disperser solvent is itsmiscibility in the organic phase and aqueous sample. Adisperser solvent must form a cloudy state to dispersethe binary extraction solvents in an aqueous solution. Inthis study, four kinds of disperser solvents such asacetonitrile, acetone, ethanol, and methanol were tested.A series of water samples was studied using 1,000 μLof each disperser solvent containing 250 μL binaryextraction solvents. The results (Fig. 4) indicated thatacetonitrile exhibits the highest extraction efficiencycompared to other disperser solvents. Thus, acetonitrilewas chosen as the disperser solvent for subsequentexperiments.

Effect of Volume of the Disperser Solvent

Variation in the volume of disperser solvent can affectthe extraction efficiency and must be optimized. The

0

200

400

600

800

1000

1200

1400

1600

40 50 60 70 80

Pea

k ar

ea

Volume of choloroform (µL)

Fig. 2 Effect of the volume of chloroform on the extraction efficiency.Conditions: concentration of PAT, 50 μg L−1; disperser solvent and itsvolume, 1.00 mL acetonitrile; no salt addition

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

0 60 120 190 230

Pea

k ar

ea

Volume of ethyl acetate (µL)

Fig. 3 Effect of the volume of ethyl acetate on the extraction efficien-cy. Conditions: concentration of PAT, 50 μg L−1; high-density extrac-tion solvent and its volume, 60 μL choloroform; disperser solvent andits volume, 1.00 mL acetonitrile; no salt addition

Food Anal. Methods (2013) 6:761–766 763

effect of disperser solvent volume on the extractionefficiency was investigated over the range of 0.50–2.00 mL. According to the obtained results (Fig. 5.),the extraction efficiency first increased and then reachedan optimum value at 1.00 mL, but decreased from 1.00to 2.00 mL. This reduction in recovery was attributed tothe increased solubility of PAT in the presence of a highamount of disperser solvent. Hence, the gain in efficien-cy was achieved by using 1.00 mL acetonitrile.

Effect of Salt Addition

The influence of ionic strength on the extraction effi-ciency has been investigated universally in extractionand microextraction techniques. In DLLME, the addition

of salt may enhance the extraction efficiency(Farhadi etal. 2009; Han et al. 2010), reduce the mass transfer oftarget analytes into the extractant(Wu et al. 2009; Sarajiand Hajialiakbari Bidgoli 2010), or have no remarkableeffect on the extraction efficiency(Shamsipur et al.2009; Negreira et al. 2010). To study the salt addition,various experiments by adding different amounts ofNaCl (0–5 %w/v) were performed. The results showedthat salt addition had no significant effect on extractionefficiency. So, all of the extraction experiments weredone without adding salt.

Optimal Extraction Procedure

For the new BS-DLLME under optimum conditions, a5.00-mL aliquot of deionized water containing 50 μgL−1

PAT was placed into a 10-mL glass test tube. A mixtureof 1.00 mL acetonitrile (the disperser solvent) including250 μL (190 μL of ethyl acetate and 60 μL of chloro-form) (19:6v/v, the binary extraction solvents) wasinjected rapidly into the sample solution. A cloudysolution was formed, and the target analyte wasextracted into the fine droplets. After centrifugation at4,000 rpm for 3 min, the fine, dispersed droplets ofchloroform and ethyl acetate were sedimented at thebottom of a conical test tube. The deposited phasewas completely transferred to another test tube andevaporated to dryness. The residue was reconstituted in1,000 μL HPLC-grade water acidified with acetic acid(pH04) and 100 μL was injected into the HPLCsystem.

Quantitative Studies

Quantitative analysis was performed by the externalstandard method. A series of water samples was sub-jected to the optimal BS-DLLME procedure and thecalibration curve for PAT was obtained. The limits ofdetection (LOD) and quantification (LOQ), correlationcoefficient, and repeatability data are given in Table 1.Repeatability was expressed as relative standard devia-tion (RSD) and evaluated on five replicate experiments

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Ethanol Methanol Acetonitrile Acetone

Pea

k ar

ea

Disperser solvent

Fig. 4 Effect of type of the disperser solvent on the extraction effi-ciency. Extraction conditions: concentration of PAT, 50 μg L−1; binaryextraction solvents and their volume, 250 μL mixture of ethyl acetateand chloroform (19:6v/v); disperser solvent and its volume, 1.00 mLacetonitrile; no salt addition

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

500 1000 1500 2000

Pea

k ar

ea

Volume of acetonitrile (µL)

Fig. 5 Effect of volume of the disperser solvent on the extractionefficiency. Extraction conditions: concentration of PAT, 50 μg L−1;binary extraction solvents and their volume, 250 μL mixture of ethylacetate and chloroform (19:6v/v); no salt addition

Table 1 Figures of merit in the DLLME

Mycotoxin R2 RSD (%) LOD (ng mL−1) LOQ (ng mL−1)

PAT 0.9954 3.8 2.0 10.0

Extraction conditions: concentration of PAT, 50 μg L−1 ; binary extrac-tion solvents and their volume, 250 μL mixture of ethyl acetate andchloroform (19:6v/v); disperser solvent and its volume, 1 mL acetoni-trile; no salt addition

764 Food Anal. Methods (2013) 6:761–766

at a concentration of 25 μg L−1. The LOD and the LOQwere 2.0 and 10.0 μg L−1, respectively.

Application in Real Samples

To assess the applicability of the suggested method, threeapple juice samples were pretreated as described in “SamplePreparation”. After the rapid injection of a mixture of theextractant, the solution was vortexed for a few seconds, andthen the proposed procedure was performed (as mentionedin “Optimal Extraction Procedure”). The results showed that

the concentration of PAT in apple juice samples was lowerthan LOD. In order to investigate the applicability andaccuracy of the suggested method, an apple juice samplewas chosen as matrix and spiked with PAT standards ofdifferent concentration levels. Determination of PAT in ap-ple juice samples was done by separation of the peaks ofPAT from HMF, as the main interference in apple juice PATanalysis, on the chromatogram. Representative chromato-grams with good resolution for PAT and HMF, obtainedfrom PAT and HMF standard solutions, PAT-free apple juice,and apple juice spiked with PAT, are shown in Fig. 6. Therelative recoveries for BS-DLLME of PAT in spiked applejuice sample are shown in Table 2. It is evident that therecoveries are in the acceptable range and no matrix effectsare observed.

Conclusions

In the present study, binary extraction solvents (chloro-form and ethyl acetate), which are heavier and lighterthan water, were used in this novel microextractiontechnique. The microextraction method based on BS-DLLME of PAT in apple juice offers prominent perfor-mance (about sevenfold) as compared to the convention-al DLLME method. BS-DLLME decreases the exposurerisk to toxic organic solvents which are used in thetraditional methods for analysis of PAT such as LLEand SPE and provides good repeatability, high recovery,low limits of detection, and quantification within a shorttime of sample preparation. It is interesting to note thatthe BS-DLLME method can be performed with widerange of lighter solvents than water and the selectionof extraction solvent is not limited to high-densitysolvents.

Fig. 6 Chromatograms for the analysis of PAT in apple juice: a HPLCchromatogram of standard HMF and PAT, b HPLC chromatogram ofblank apple juice after performing BS-DLLME, c HPLC chromato-gram of apple juice spiked with PAT after employing BS-DLLME

Table 2 Relative recoveries of PAT in apple juice samples

Mycotoxin Initialconcentrationmean ± SD(μg L−1)

Concentrationadded (μg L−1)

Concentrationdeterminedmean ± SD(μg L−1)

Relativerecovery(%)

PAT nd 25.0 22.82±4.2 91.3

50.0 46.3±3.6 92.6

75.0 71.4±3.7 95.2

Extraction conditions: binary extraction solvents and their volume,250 μL mixture of ethyl acetate and chloroform (19:6v/v); dispersersolvent and its volume, 1 mL acetonitrile; no salt addition

SD standard deviation (n03), nd not detected

Food Anal. Methods (2013) 6:761–766 765

References

Abdolmohammad-Zadeh H, Sadeghi GH (2010) Combination of ionicliquid-based dispersive liquid–liquid micro-extraction withstopped-flow spectrofluorometry for the pre-concentration anddetermination of aluminum in natural waters, fruit juice and foodsamples. Talanta 81:778

Almeida C, Fernandes JO, Cunha SC (2012) A novel dispersive liquid-liquid microextraction (DLLME) gas chromatographymass spec-trometry (GC-MS) method for the determination of eighteenbiogenic amines in beer. Food Control 25:380

Baert K, De Meulenaer B, Kasase C, Huyghebaert A, Ooghe W,Devlieghere F (2007) Free and bound patulin in cloudy applejuice. Food Chem 100:1278

Boonchiangma S, Ngeontae W, Srijaranai S (2012) Determination ofsix pyrethroid insecticides in fruit juice samples using dispersiveliquid–liquid microextraction combined with high performanceliquid chromatography. Talanta 88:209

Campone L, Piccinelli AL, Rastrelli L (2011a) Dispersive liquid–liquidmicroextraction combined with high-performance liquid chroma-tography–tandem mass spectrometry for the identification and theaccurate quantification by isotope dilution assay of Ochratoxin Ain wine samples. Anal Bioanal Chem 399:1279

Campone L, Piccinelli AL, Celano R, Rastrelli L (2011b) Applicationof dispersive liquid–liquid microextraction for the determinationof aflatoxins B1, B2, G1 and G2 in cereal products. J ChromatogrA 1218:7648

Cunha SC, Fernandes JO (2011) Multipesticide residue analysis inmaize combining acetonitrile-based extraction with dispersiveliquid–liquid microextraction followed by gas chromatography–mass spectrometry. J Chromatogr A 1218:7748

Farhadi K, Matin AA, Hashemi P (2009) LC determination of traceamounts of phenoxyacetic acid herbicides in water after disper-sive liquid–liquid microextraction. Chromatographia 69:45

Fu L, Liu X, Hu J, Zhao X, Wang H, Wang X (2009) Application ofdispersive liquid–liquid microextraction for the analysis of triazo-phos and carbaryl pesticides in water and fruit juice samples. AnalChim Acta 632:289

Funes GJ, Resnik SL (2009) Determination of patulin in solid andsemisolid apple and pear products marketed in Argentina. FoodControl 20:277

Gökmen V, Acar J, Sarioğlu K (2005) Liquid chromatographic methodfor the determination of patulin in apple juice using solid-phaseextraction. Anal Chim Acta 543:64

Han Y, Jia X, Liu X, Duan T, Chen H (2010) DLLME combined withGC–MS for the determination of methylparaben, ethylparaben,propylparaben and butylparaben in beverage samples. Chroma-tographia 72:351

Iha MH, De Souza SVC, Sabino M (2009) Single-laboratory validationof a liquid chromatography method for the determination ofpatulin in apple juice. Food Control 20:569

Karami-Osboo R, Mirabolfathy M, Kamran R, Shetab-Boushehri M,Sarkari S (2012) Aflatoxin B1 in maize harvested over 3 years inIran. Food Control 23:271

Leggott NL, Shephard GS (2001) Patulin in South African commercialapple products. Food Control 12:73

Li JK, Wu RN, Hu QH, Wang JH (2007) Solid-phase extraction andHPLC determination of patulin in apple juice concentrate. FoodControl 18:530

Murillo M, Gouzález-Peñas E, Amézqueta S (2008) Determination ofpatulin in commercial apple juice by micellar electrokinetic chro-matography. Food Chem Toxicol 46:57

Negreira N, Rodríguez I, Rubí E, Cela R (2010) Dispersive liquid–liquidmicroextraction followed by gas chromatography–mass spectrome-try for the rapid and sensitive determination of UV filters in envi-ronmental water samples. Anal Bioanal Chem 398:995

Rezaee M, Assadi Y, Milani Hosseini MR, Aghaee E, Ahmadi F,Berijani S (2006) Determination of organic compounds in waterusing dispersive liquid–liquid microextraction. J Chromatogr A1116:1

Saraji M, Hajialiakbari Bidgoli AA (2010) Dispersive liquid–liquidmicroextraction using a surfactant as disperser agent. Anal Bioa-nal Chem 397:3107

Shamsipur M, Ramezani M, Sadeghi M (2009) Preconcentration anddetermination of ultra trace amounts of palladium in water sam-ples by dispersive liquid-liquid microextraction and graphite fur-nace atomic absorption spectrometry. Microchim Acta 166:235

Valle-Algarra FM, Mateo EM, Gimeno-Adelantado JV, Mateo-CastroR, Jiménez M (2009) Optimization of clean-up procedure forpatulin determination in apple juice and apple purees by liquidchromatography. Talanta 80:636

Wu Q, Zhou X, Li Y, Zang X, Wang C, Wang Z (2009) Application ofdispersive liquid–liquid microextraction combined with high-per-formance liquid chromatography to the determination of carba-mate pesticides in water samples. Anal Bioanal Chem 393:1755

Zhang S, Yang X, Yin X, Wang C, Wang Z (2012) Dispersive liquid–liquid microextraction combined with sweeping micellar electro-kinetic chromatography for the determination of some neonicoti-noid insecticides in cucumber samples. Food Chem 133:544

766 Food Anal. Methods (2013) 6:761–766