[31x]PAE in oil by HS-SPME-GC-ITMSMS

Headspace temperature and phthalate esters determination by gas chromatography multistage mass spectrometryJ.J.Rios a ,∗,A.Morales a ,G.Márquez-Ruiz ba Food Characterization and Quality Department,Instituto de la Grasa (C.S.I.C.),Av.Padre García Tejero 4,41012Seville,Spain bScience and Technology of Dairy Products Department,Instituto del Frio (C.S.I.C.),Jose Antonio Novais 1028040Madrid,Spaina r t i c l e i n f o Article history:Received 7July 2009Received in revised form 22October 2009Accepted 3November 2009Available online 12 November 2009Keywords:PhthalatesSolid-phase microextractionGas chromatography ion trap–mass spectrometry Oil matricesa b s t r a c tA solvent-free analytical approach based on headspace solid-phase microextraction (SPME)of oil matri-ces heated at high temperatures coupled to gas chromatography with mass spectrometry detector (GC-ion trap)has been developed for the determination of phthalic acid esters (PAEs)in oil matrices without sample manipulation.For this study,three fibers,i.e.,85␮m-polyacrylate (PA),50/30␮m-divinylbenzene–carboxen–polydimethylsiloxane (DVB/CAR/PDMS)and 100␮m-polydimethylsiloxane (PDMS)were tested.Variables affecting the SPME headspace composition such as incubation sample temperature,sample incubation time and fiber exposition time were optimized.The optimal values found were 250◦C for sample incubation temperature and 30min for incubation and extraction time.PA fiber was not suitable for the lightest polar phthalates which showed poor extraction and repeatability values.PDMS fiber had very poor response for some of the heavier and non-polar phthalates,whereas DVB/CAR/PDMS fiber showed the best response and repeatability values for the majority of the phthalates studied.The main benefit of the analytical method proposed is the absence of sample manipulation and hence avoidance of possible contamination coming from glassware,environment,solvents and samples.© 2009 Elsevier B.V. All rights reserved.1.IntroductionWorldwide production of phthalic acid esters (PAEs)and their common application in different products of everyday use has resulted in their widespread presence in environment and foods.PAEs are used extensively as plasticizers in different materi-als,especially polivinyl chloride,polyvinyl acetate,polystyrene,polyamide and polyester materials.Bis-(2-ethylhexyl)phthalate (DEHP)and benzylbutylphthalate (BBP)are the most abundant phthalate esters due their wide use in many industrial sectors like cosmetics and paints.As plasticizers are not chemically bound to the polymer,migration of PAEs in significant amounts is possi-ble.Some research papers have been published recently about the impact of PAEs on wild animals and human beings [1,2].It is suggested that increasing exposure to PAEs might be partially responsible for some diseases.Furthermore,it was found that PAEs might persist in human body tissues [2].Since foods are the major source of exposure to PAEs,it is important to monitor their levels in foods to provide data for human exposure assessment.The US∗Corresponding author.Tel.:+34954611550;fax:+34954616790.E-mail address:jrios@ig.csic.es (J.J.Rios).Environmental Protection Agency (EPA)and several other agencies classified the commonly occurring PAEs as priority pollutants.Due to such health implications,there is increasing demand for reliable analytical methods that allow their detection and quantification at very low levels in different environmental,biological and food matrices.Several analytical methods have been developed for determin-ing PAEs in food samples and fatty matrices [3,4],and recently an interesting application to olive oil has been published [5].Either adsorption chromatography on Florisil or alumina [6,7]or codistillation of PAEs [8]is usually applied.Quantification of low levels using solid-phase extraction (SPE)with C18[9],C8[10],polystyrene [11]and multi-walled carbon monotubes to concen-trate PAEs [12]has been also proposed prior to analysis by either liquid or gas chromatography.These options have been preferred to solvent extraction due to lower consumption of organic solvents.Several studies applying direct SPME extraction of PAEs have been proposed.This procedure significantly reduces the risk of secondary contamination and substantially simplifies the overall analytical process.It has been already applied to the determination of phthalates in wine [13,14],cow milk [15],vegetable oil using solvent-based matrix modification [16],and plastic for foods [17].Cross contamination from chemicals,materials and laboratory equipment are common problems in all PAEs analysis methods0039-9140/$–see front matter © 2009 Elsevier B.V. All rights reserved.doi:10.1016/j.talanta.2009.11.008J.J.Rios et al./Talanta80 (2010) 2076–20822077[18].For example,di(2-ethylhexyl)phthalate and dibutyl phthalate are usually detected in blank samples.In addition,the diffi-culty for separating co-extracts as lipids,pigments,etc.,makes the determination of low levels of PAEs in food matrices very difficult.The aim of this work was to develop an analytical procedure to detect and quantify the phthalate levels in oil matrices without sample manipulation and thus avoiding external contamination from air,solvents,labware,etc.The novelty of this method con-sists of adsorption by SPME of the headspace of the oil heated at high temperature(250◦C)in a sealed vial followed by subsequent analysis by gas chromatography ion trap MS–MS.For thefirst time, high temperature is here used for determination of phthalates in oil matrices.2.Materials and methods2.1.Reagents and materialsVirgin olive oil purchased locally was used as oil matrix for experiments.A standard mixture solution of phthalates(EPA 8061Mix Analytes)and benzyl benzoate containing1mg mL−1of each compound in95:5hexane:acetone(v/v)(Cromlab,Barcelona, Spain)was used for setting instrumental conditions and spiking the olive oil samples.The standard mixture solution contained 1mg mL−1of methyl phthalate(DMP),diethyl phthalate(DEP), benzylbenzoate as internal standard(IS),di-isobuthyl phtha-late(DIBP),di-n-buthyl phthalate(DBP),bis-2-methoxyethyl phthalate(MEP),bis-1,2-(4methyl-2pentyl)phthalate(1,2MPP), bis-1,3(4methyl-2pentyl)phthalate(1,3MPP),bis-2ethoxyethyl phthalate(EEP),di-n-amyl phthalate(DAP),di-n-hexyl phtha-late(DHP),benzyl buthyl phthalate(BBP),bis-2-buthoxyethyl phthalate(BBEP),dicyclohexyl phthalate(DCHP),bis2-ethylhexyl phthalate(DEHP),di-n-octyl phthalate(DOP)and di-n-nonyl phthalate(DNP).Five spiked levels of0.05,0.1,0.5, 1.0and 2.0mg kg−1were prepared.Initial olive oil used for experiments and the solvents for preparing spiking solutions were tested previ-ously to verify absence of phthalates.Six samples of virgin olive oils,purchased locally,were stored at room temperature(23±5◦C)for12months,either in glass bottles or in PET(polyethylene terephthalate)bottles.Duplicate samples were taken at8,10and12months.Special care was taken with glassware,vials,caps and septa before use.All materials were dried and kept at200◦C until use.To avoid phthalate contamination,all glassware used was soaked in acetone for at least30min,then washed with acetone,rinsed with hexane,and dried at200◦C for at least4h.All the solvents,chro-matographic system,standards and spiked samples were subjected to similar extraction and analysis procedures.Blank extraction was routinely performed.Also,the chromatographic system was ini-tially and regularly checked for absence of phthalates by running blank injections.2.2.SPMEfibersSPMEfibers were purchased from Supelco(Bellefonte,PA,USA). Three types offiber coating were tested,i.e.,85␮m-polyacrylate (PA),50/30␮m-divinylbenzene–carboxen–polydimethylsiloxane (DVB/CAR/PDMS)and100␮m-polydimethylsiloxane(PDMS),pre-viously conditioned in accordance with manufacturer’s instruc-tions.Fibers were thermally cleaned in the injection port before sampling.Each ten extractions,fibers were cleaned by immersion with vigorous shaking in a mixture of20:70:10 water:acetonitrile:acetic acid(v/v/v)for1h,dried under nitrogen and conditioned at260◦C for30min.Fig.1.Schematic representation of the extraction procedure.2.3.SPME procedureSamples of1g olive oil were weighed into glass vials(70mm long and22mm diameter)and tightly capped.The bottom part of the caps was Teflon-plated to avoid plastic contact with headspace. Then,vials were heated at different temperatures(200–275◦C) and time periods(10–40min)in order to promote release of the semivolatile compounds to headspace.Sampling was performed by exposing the SPMEfiber to headspace for30min at the upper part of vial positioned outside the oven-heating zone(20mm long),as shown in Fig.1.After sample extraction,the SPMEfiber was intro-duced into the GC injection port(260◦C)operating in splitless mode for5min.Samples were extracted only once from vial.2.4.GC ion trap–MS analysisThe GC ion trap–MS experiments were performed using a Finni-gan Trace-GC2000gas chromatograph coupled to a Polaris-Q Ion trap mass spectrometer(ThermoFinnigan,Austin,TX,USA)operat-ing in fullscan and MS/MS mode.The column used was a Zebron ZB-5ms(Phenomenex,Torrance,CA,USA)fused silica capillary column(30m long×0.25mm i.d×0.25mfilm thickness).The GC conditions included helium as carrier gas at1mL min−1in constant flow mode.The initial temperature of70◦C was kept for2min,then raised to200◦C at a rate of6◦C min−1and held for5min,then raised to295◦C at a rate of4◦C min−1and held for8min.Injector temper-ature was260◦C and a0.8mm ID liner was used for SPME injection port.The MS operating conditions were as follows:ion source and transfer line temperatures were200and290◦C,respectively.The electron energy was70eV,resolution was unit and the emission current was250␮A.2078J.J.Rios et al./Talanta 80 (2010) 2076–2082Table 1Precursor and quantification ions for PAEs.Peak no.Analyte Precursor MS 2Q CID Quantification ion 1DMP 1630.4521332DEP 1490.452121+933IS 1050.451774DIBP 1490.452121+935DBP 1490.452121+936MEP 1490.452121+9371,2MPP 1490.452121+9381,3MPP 1490.452121+939EEP 1490.452121+9310DAP 1490.452121+9311DHP 1490.452121+9312BBP 1490.452121+9313BBEP 1490.452121+9314DCHP 1490.452121+9315DEHP 1490.452121+9316DOP 1490.452121+9317DNP1490.452121+93Abbreviations :DMP,Methyl phthalate;DEP,diethyl phthalate;IS,internal stan-dard (benzylbenzoate);DIBP,di-isobuthyl phthalate;DBP,di-n-buthyl phthalate;MEP,bis-2-methoxyethyl phthalate;1,2MPP,bis-1,2-(4methyl-2pentyl)phthalate;1,3MPP,bis-1,3(4methyl-2pentyl)phthalate;EEP,bis-2ethoxyethyl phthalate;DAP,di-n-amyl phthalate;DHP,di-n-hexyl phthalate;BBP,benzyl buthyl phthalate;BBEP,bis-2-buthoxyethyl phthalate;DCHP,dicyclohexyl phthalate;DEHP,bis 2-ethylhexyl phthalate;DOP,di-n-octyl phthalate and DNP,di-n-nonyl phthalate.The Xcalibur version 1.4software was used for data acquisition and processing of the results.3.Results and discussion3.1.Mass spectrometry scanning methodFull scan and MS/MS (MS 2)were used alternately and consec-utively for MS operations.The MS/MS method development was carried out in two steps.The first one was to set the retention time of the compounds of interest and to slice the chromatogram into 17segments corresponding to specific PAEs and benzyl benzoate.First,the standard mixture (EPA 8061Analytes)was injected in fullscan mode.Maximum excitation energy (Q )and CID excitation voltage were optimized for each individual standard compound and pre-cursor ion in order to get the best isolation efficiencies and ensure maximum production or transmission of collision fragments.Fig. 2.Response areas obtained with the fibers tested.Means of five determinations.Coefficients of variation were lower than 10%.Abbrevia-tions for fibers:PA,85␮m-polyacrylate;PDMS,100␮m-polydimethylsiloxane and DVB/CAR/PDMS,50/30␮m-divinylbenzene–carboxen–polydimethylsiloxane.Abbreviations for phthalates:DMP,Dimethyl phthalate;DEP,diethyl phthalate;IS,internal standard (benzylbenzoate);DIBP,di-isobuthyl phthalate;DBP,di-n-buthyl phthalate;MEP,bis-2-methoxyethyl phthalate;1,2MPP,bis-1,2-(4methyl-2pentyl)phthalate;1,3MPP,bis-1,3(4methyl-2pentyl)phthalate;EEP,bis-2ethoxyethyl phthalate;DAP,di-n-amyl phthalate;DHP,di-n-hexyl phthalate;BBP,benzyl buthyl phthalate;BBEP,bis-2-buthoxyethyl phthalate;DCHP,dicyclohexyl phthalate;DEHP,bis 2-ethylhexyl phthalate;DOP,di-n-octyl phthalate and DNP,di-n-nonyl phthalate.Each scanwas subdivided into two events.The first one was performed in fullscan mode in the 50–600amu scan range,and the second one in MS/MS mode with the precursor ion selected.The total scan time for each couple of segments was 0.38s and the sys-tem performed 3microscans per second.Table 1shows precursor ions for each segment of MS/MS experiments and quantification ions for the PAEs studied.Upon data acquisition,analysis of data was based on the fol-lowing criteria:retention time for each peak should be within a 1%range,and integration of ion chromatograms was done with the Xcalibur Software using ICIS Peak Detection with ion ratio con-firmation parameter,selecting each peak area for the base peak corresponding with the quantification ion selected.For all of PAEs,except DMP and internal standard,two ions were considered asFig.3.Phthalate responses as a function of incubation temperature.Sample:EPA Mix 8061A standard mixture in olive oil (1mg kg −1).J.J.Rios et al./Talanta 80 (2010) 2076–20822079quantification ions when relation of m/z 121and m/z 93was between 2.5and 2.7(±5%),as found for the standard compounds.For DMP and IS,single ions were selected for quantification due to their specificity.3.2.Fiber and sample temperature optimizationFiber type and sample temperature play an important role in the amounts of compounds absorbed and both have to be con-trolled separately.Three types of fiber coating were tested:PA,DVB/CAR/PDMS and PDMS fibers.PA and PDMS fibers,at headspace temperatures tested here,were not suitable for dimethyl and diethyl phthalates extraction and showed poor repeatability for the rest of PAEs studied (data not shown).As listed in Fig.2,DVB/CAR/PDMS fiber showed the best response values for all of the phthalates studied from di-methyl phthalate (DMP)to di-nonyl phthalate (DNP).Sample incubation temperature was tested between 200and 275◦C since it was preliminarily observed that presence of the least volatile phthalates in headspace was reduced below 200◦C.Fig.3shows response ratios obtained between SPME of 1g of olive oil spiked with 1␮g g −1of each compound of the standard mixture (EPA 8061Analytes)and direct liquid injection (in splitless mode)of 1␮L of standard solution (1mg L −1).The best results were obtained at 250◦C,with good SPME/direct liquid injection response ratio for all the compounds studied.This intermediate incubation temper-ature selected was a compromise between those most specifically suitable for compounds of lowest and highest volatility.Vials were placed so as to leave their upper part (20mm long)out of the heating system (Fig.1).Thus,such upper parts of vials were not directly heated and temperature could be established between 80and 115◦C to allow optimal response for all the PAEs tested.As headspace composition can vary depending on temper-ature,the SPME needle was punched through the cap and septum into the headspace and introduced at different depths (3.0,2.0,1.5and 1.0cm under cap bottom).Temperature was measured at the different depths using a digital microtermometer.For each point headspace temperature varied only less than ±3◦C during extraction time (30min),as shown in Fig.4.By using the internal standard,possible variations derived from lack of fiber thermo-station or influence of glass surface characteristics [19]could becorrected.Fig.4.Stability of headspace zone temperature during 40min extraction period.Means of three analyses.Coefficients of variation were lower than 10%.Fig.5shows response areas for phthalates absorbed from 1g of oil sample containing 1␮g of each compound of the standard mixture (EPA 8061Analytes)in function of the fiber position inside the headspace and the temperature on each position.The operating fiber position in the headspace selected was 1.0cm under cap bot-tom,corresponding to 94◦C in headspace,according to the highest response areas obtained under such conditions.Incubation time was studied from 10to 40min.Fig.6shows the results obtained,expressed as normalized response (%of each PAEs on total PAEs).Beyond 20min of incubation time,this param-eter was not critical,probably because the headspace equilibrium is reached relatively soon.However,it is important to stabilize the headspace zone temperature and that is why 30min was selected as extraction time.Fig.7shows the SPME–gas chromatogram ion trap MS/MS pro-file of an olive oil sample spiked with 1␮g g −1of each compound of EPA 8061A Analytes compared with the profile corresponding to direct liquid injection of 1␮L of 1mg L −1EPA 8061A Analytes solu-tion,in splitless mode.As can be observed,chromatograms were similar,which indicates that all phthalates were extracted without any distinctive suppression orpromotion.Fig.5.Phthalate responses as a function of fiber sampling position (position 3,2,1.5and 1cm under the cup corresponds to 112,107,99and 94◦C,respectively).Means of two analyses.Coefficients of variation were lower than 10%.2080J.J.Rios et al./Talanta80 (2010) 2076–2082Fig.6.Phthalate responses as a function of extraction time.Means of two analyses.Coefficients of variation were lower than10%.parison of chromatograms obtained after:top,injection of 1␮L of phtha-lates standard solution (EPA 8061standard mixture in acetonitrile,concentration 1mg L −1);bottom,HS-SPME of 1g of olive oil (spiked with EPA 8061standard mix-ture at the level of 1␮g g −1).For peaks identification,see Table 1.Fig.8.Headspace GC-ion trap–MS/MS chromatogram of olive oil sample 1(PET container)after applying the quantification layout.Fig.8shows a representative example of the chromatographic profile obtained for a real sample by applying the method proposed.In this case,sample olive oil 1(PET container),only DIBP,DBP,BBP and DEHP were detected in significant amounts.3.3.Method performanceRepeatability of the SPME method was determined by five repli-cate analyses of olive oil spiked at levels of 0.1,0.5and 1.0mg kg −1.Linearity testing was carried out using olive oil spiked at levels of 0.05,0.1,0.5,1.0and 2.0mg kg −1.The resulting performance characteristics are summarized in Table 2.Linearity of the analysis and extraction procedure of oil sam-ples spiked with 0.05–2.0mg kg −1showed good correlation values for all of the compounds studied.The relative standard devia-tions (RSDs)for samples spiked with 0.1,0.5and 1.0mg kg −1were always below 20%.It is important to note that particularly low RSD,below 8.5%,were obtained for benzyl butyl phthalate (BBP)J.J.Rios et al./Talanta80 (2010) 2076–20822081Table2Resulting performance characteristics.Repeatability RSD(%),n=5Linearity R2(0.05–2mg kg−1)LOD a(mg kg−1)0.1mg kg−10.5mg kg−1 1.0mg kg−1DMP17.318.714.10.9890.04DEP11.613.110.60.9950.04IS12.510.510.80.9990.04DIBP18.216.717.40.9880.03DBP8.510.5 4.70.9980.03MEP16.314.813.00.9970.031,2MPP17.918.715.80.9780.031,3MPP18.614.118.00.9880.03EEP12.114.510.90.9920.03DAP17.816.714.60.9860.04DHP8.08.1 5.30.9870.04BBP7.67.7 6.10.9830.03BBEP10.613.78.20.9980.02DCHP8.38.17.30.9750.02DEHP8.58.18.30.9950.03DOP15.714.815.60.9790.04DNP18.618.115.10.9800.05Abbreviations as in Table1.a Limit of detection.Table3Contents of phthalates in stored olive oils.Compound Container PET Container glassStorage time(months)Storage time(months)8101281012Sample1Sample2Sample3Sample4Sample5Sample6DMP–a–––––DEP––––––DIBP120119216–––DBP97111175–––BBP17320721187110112DEHP543635840223237198DOP––––––DNP––––––Abbreviation:PET,polyethylene terephthalate.Values in␮g kg−1.Mean of two determinations.a Below LOD.and bis(2-ethylhexyl)phthalate(DEHP),which are the most widely found PAEs.Relatively high RSDs for some of the PAEs were found, especially the most volatile ones.Such results could be expected under the conditions used,which favor overall extraction and absence of contaminants during the extraction process but may otherwise lead to overestimated values[18].3.4.Determination of PAEs in olive oil samplesAs an example of application of the method to real samples, three samples of olive oils packed in plastic(PET,polyethylene terephthalate)bottles and three of olive oils packed in glass containers were analyzed for PAEs.As only few of the phtha-lates studied in this work may really be encountered in edible oils,namely DBP,DEP,DIBP,DBP,BBP,DEHP,DOP and DNP only those,if detected,are shown in real sample results.Anal-yses were done in duplicate and,if any,levels in blank were substracted.Samples kept in PET containers shown presence of DIBP,DBP, BBP and DEHP but in samples kept in glass containers only signi-ficative BBP and DEHP were detected in the samples analyzed as it is shown in Table3.4.ConclusionsThe analytical method described in the present work is sim-ple and highly selective for quantification of PAEs in oil matrices.As compared with the conventional clean-up process based on liquid–liquid extraction or solid-phase extraction,the present approach is faster and more economical.Furthermore,the main benefit of the method is the suppression of the main contamina-tion sources,especially in relation to sample manipulation.Due to complexity of the headspace of heated oils,however,there are cer-tain limitations as tofiber durability.Degradation offibers gives rise to poor recovery of the least volatile PAEs,such as DNOP and DNP, thereby special care should be taken to control recovery values and rejectfibers when necessary.References[1]N.Ghorpade,V.Metha,M.Khare,P.Sinkair,S.Krishnan,C.V.Rao,Ecotoxicol.Environ.Saf.53(2002)255–258.[2]X.Hu,B.Wen,X.Shan,J.Environ.Monit.5(2003)649–653.[3]C.S.Giam,M.K.Wong,J.Food Prot.5(1987)769–782.[4]B.D.Page,in:J.H.Hotchkiss(Ed.),Food Packaging Interation I,ACS SymposiumSeries No.365,ACS,Washington,1988,p.118.[5]B.Cavaliere,B.Macchione,G.Sindona,A.Tagarelli,J.Chromatogr.1205(2008)137–143.[6]T.Suzuki,K.Ishikawa,N.Sato,K.I.Sakai,JAOCS62(3)(1979)689–694.[7]A.A.Belisle,W.L.Reichel,J.W.Spann,Bull.Environ.Contam.Toxicol.13(1975)129–132.[8]A.D.Vethaak,G.B.J.Rijs,S.M.Schrap,H.Ruiter,A.Gerritsen,hr,Report no.2002.001,Dutch National Institute of Inland Water Management and Waste Water Treatment,Lelystad,The Netherlands.[9]I.Colon,D.Caro,C.J.Bourdony,O.Rosario,Environ.Health Perspect.108(2000)895–900.[10]B.D.Page,croix,Food Addit.Contam.12(1995)129–151.[11]N.Casajuana,corte,Chromatographia57(2003)649–655.2082J.J.Rios et al./Talanta80 (2010) 2076–2082[12]M.L.Davi,M.Liboni,M.G.Malfatto,Int.J.Environ.Anal.Chem.74(1999)155–166.[13]J.D.Carrillo,C.Salazar,C.Moreta,M.T.Tena,J.Chromatogr.A1164(2007)248–261.[14]J.D.Carrillo,M.P.Martinez,M.T.Tena,J.Chromatogr.A1181(2008)125–130.[15]Yong-Lai Feng,Jiping Zhu,Robert Sensenstein,Anal.Chim.Acta538(2005)41–48.[16]K.Holadova,G.Prokupkova,J.Hajslova,J.Poustka,Anal.Chim.Acta582(2007)24–33.[17]H.-Y.Shen,Talanta66(2005)734–739.[18]A.Fankhauser-Noti,K.Grob,Anal.Chim.Acta582(2007)353–360.[19]A.Fankhauser-Noti,K.Grob,Trends Food Sci Technol.17(2006)105–112.。