analyt ica ch im ica acta 6 0 5 ( 2 0 0 7 ) 147152

avai lab le at www.sc iencedi rec t .com

journa l homepage: www.e lsev ier .com/ locate /aca

Deter uswaste olmicro icrcomp

Chanbasha Basheera, Anass Ali Alnedharyb, B.S. Madhava Raob, Hian Kee Leea,

a Department of Chemistry, National University of Singapore, 3 Science Drive 3, Republic of Singapore 117543b Department of Chemistry, Pune University, Pune 411007, India

a r t i c

Article histor

Received 4 J

Received in

15 Septemb

Accepted 6

Published o

Keywords:

Liquid-phas

Binary solve

Solid-phase

Organophos

Environmen

1. Int

The use ofbioaccumurisks to boseeks to reorganophodeterminatmental sam

CorresponE-mail a

0003-2670/$doi:10.1016/l e i n f o

y:

uly 2007

revised form

er 2007

October 2007

n line 11 October 2007

e microextraction

nt

microextraction

phorous pesticides

tal analysis

a b s t r a c t

A simple and efcient binary solvent-based two-phase hollow ber membrane (HFM)-

protected liquid-phase microextraction (BN-LPME) technique for moderately polar

compoundswas developed. Six organophosphorous pesticides (OPPs) (triethylphosphoroth-

ioate, thionazin, sulfotep, phorate, disulfoton, methyl parathion and ethyl parathion) were

used as model compounds and extracted from 10-mL wastewater with a binary-solvent

(toluene:hexane, 1:1) mixture. Some important extraction parameters, such as extraction

time, effect of salt, sample pH and solvent ratio compositionwere optimized. BN-LPME com-

binedwith gas chromatography/mass spectrometric (GC/MS) analysis provided repeatability

(R.S.D.s12%, n=4), and linearity (r0.994) and solid-phase microextraction provides com-parable of R.S.D.s13%, n=4 and linearity (r=0.966) for spiked water samples. The limitsof detection (LODs) were in the range of 0.311.4ngL1 for BN-LPME and 3.1120.5ngL1 for

SPME at (S/N=3) under GC/MS selective ion monitoring mode. In addition to high enrich-

ment, BN-LPME also served as a sample cleanup procedure, with the HFM act as a ltering

medium to prevent large particles and extraneous materials from being extracted. To inves-

tigate and compare their applicability, the BN-LPME and SPME procedures were applied to

the detection of OPPs in domestic wastewater samples.

2007 Elsevier B.V. All rights reserved.

roduction

pesticides has benets in agriculture, but theirlation through the food web, poses potentialth animals and humans [1,2]. Current legislationduce the maximum residue levels allowed for

sphorous pesticides (OPPs). The need for accurateion of pesticides at the trace levels in the environ-ples is therefore obvious. Sample preparation is

ding author. Tel.: +65 6516 2995; fax: +65 6779 1691.ddress: chmleehk@nus.edu.sg (H.K. Lee).

normally required to isolate and concentrate compounds ofinterest from the samplematrix, before analysis. Liquidliquidextraction (LLE) is still the most common sample preparationapproach. It is, however, a time-and labor-intensive proce-dure and requires large amounts of high-purity solvents thatare expensive and potentially toxic [3] In the past few years,simple and miniaturized sample preparation techniques havebeen reported as alternatives to conventional sample prepa-ration procedures. These include ow-injection extraction

see front matter 2007 Elsevier B.V. All rights reserved.j.aca.2007.10.006mination of organophosphorowater samples using binary-sextraction and solid-phase marative studypesticides invent liquid-phaseoextraction: A

148 analyt ica ch im ica acta 6 0 5 ( 2 0 0 7 ) 147152

[4] solid-phase extraction (SPE) [5], solid-phase microextrac-tion (SPME) [6], stir bar sorptive extraction (SBSE) [7], andliquid-phastion (SME)

The avaresulted insample clein some camore substthe emergiSME. SPMEwide rangetively expetend to degrecently inttion [14,15]required.

Single-d[16,17] hasfast sampland unsuittered [18].bermembreported ais held andliquidliquorganic solsolution haing polar an

In HFM-ical task; oHFM-compstudies [19extractionbinary-solvimprove anrarely applsolvents toples; theyused extraHFM-LPMEthe rst timextractioncedure waSPME.

2. Ex

2.1. Ch

Individual OUSA). HPLCUK). The wBedford,MAsolution oftone. A worwas prepartration spisolution wa

2.2. Materials

ccurdiamize). 1.2PMElbenz0-mandonteioneanuf/hot.

Ins

e an, JapCMmpl0.32, CA

rrierSampon titheraturws:sed beswedeteall cure weage

Wa

tic wortedpendere uno

Hoxtra

Lmiclton,e:hetiosttebelowringas cooverd s

ventas de microextraction (LPME) or solvent microextrac-[811].ilability of a broad range of different sorbents haswidespread use of SPE [12]. However, sufcientan-up and analyte enrichment may be difcultses with SPE, and also this technique consumesantial amounts of organic solvents than SPME andng solvent-minimized procedures such as LPME orhas been successfully used for the analysis of aof organic compounds. SPME bers are compara-

nsive, however, and have a limited lifetime as theyrade with repeated usage [13]. In SBSE [7] and theroduced polymer-coated hollow ber microextrac-, after extraction, an additional desorption step is

rop microextraction (SDME) or microdrop-LPMEbeen demonstrate to be a simple, inexpensive ande preparation procedure. However, drop stabilityability for complex samples were often encoun-In order to eliminate these limitations, hollowrane-protected LPME (HFM-LPME) [1821] has beens an alternative. In this technique, the solvent

protected by an HFM. In addition, three-phaseidliquid microextraction involving extraction intovent, followed by back-extraction into an aqueouss been reported in the literature [22,23] for extract-alytes [24].LPME, the selection of a suitable solvent is a crit-nly a limited range of immiscible, polypropyleneatible solvents is available [25]. In our previous21] toluene was found to be the most suitablesolvent for most analytes. Although the use ofent mixtures is common in conventional LLE toalyte enrichment [26,27], this approach has beenied to HFM-LPME. Ugland et al. [28] used binaryextract benzodiazepines from biological sam-

made no attempt to compare with a commonlyction procedure, however. In the present study,with binary-solvent mixtures has been used, fore, to extract OPPs in wastewater samples. The

performance of this binary-solvent (BN)LPME pro-s compared with single-solvent HFM-LPME and

perimental section

emicals and reagents

PPs were purchased fromAldrich (Milwaukee,WI,-grade solvents were purchased from BDH (Dorset,ater used was puried using a Milli-Q (Millipore,, USA)water purication system. A standard stock50gmL1 of each analyte was prepared in ace-king standard solution (1gmL1 of each analyte)ed in the same solvent and used for low concen-king; for higher concentration spiking, the stocks used.

Q3/2 Ainnerpore smany)The Sdivinyand 10septa(Bellefconditthe mstirrermany)

2.3.

Sampl(Tokyoetry (Gautosa30mFolsomthe caof 20.injecti250 C,tempeas folloincreasamplwith a

Inultrapfrom r

2.4.

DomestranspNo susples w6.6 and

2.5.microe

A10-(Hamitoluenvent ratightly5mmThe syHFMwtakenof 73 rathe solber wel polypropylene hollow ber membrane (600meter (I.D.), 200m wall thickness and 0.2m wallwas purchased from Membrana (Wuppertal, Ger--cm lengths of HFM were used for extraction.ber holder and bers poly(dimethylsiloxane)-ene (PDMS-DVB, 65-m thickness), (PDMS, 7-m), polyacrylate (PA) 85m) and extraction vials,aluminium caps were purchased from Supelco, PA, USA). Before extraction the SPME bers wered in a gas chromatographic injection port based onacturers recommended procedure. The magneticplate was obtained from Heidolph (Kelheim, Ger-

trumentation

alyses were carried out using a Shimadzuan) QP2010 gas chromatography/mass spectrom-S) system equipped with a Shimadzu AOC-20ier and a DB-5 fused silica capillary columnmmI.D., 0.25m lm thickness (J&W Scientic,, USA). High purity helium (99.9999%) was used asgas at a ow rate of 1.5mLmin1 and a split ratioles (3L) were injected in splitless mode with an

me of 2min. The injection temperature was set atinterface temperature at 200 C and the detectione at 280 C. The GC-MS temperature program wasinitial temperature 50 C (2min, holding time), theny 10 Cmin1 to 300 C (3min). OPP standards andre analysed in selective ionmonitoring (SIM)modector voltage of 1.5 kV.ases, control experiments were performed withater to assess the presence of any contaminationnts and bers.

stewater samples

astewater sample was collected in a township andin pre-cleaned glass bottles and stored at 4 C.ed solids were observed; therefore unltered sam-sed for experiments. The original sample pH wasother physical characteristics were measured.

llow ber-protected liquid-phasection

rosyringe,with a cone tip (0.47mmouter diameter)Reno, NV, USA) was used for OPP extraction. 5Lxane (1:1) mixture was used as solvent. Other sol-were also evaluated. The syringe needle was thend with a 1.2-cm length of HFM and then immersedthe surface of a sample solution in a 10mL vial.

e plunger was depressed completely so that thempletely lledwith solvent. Extractionwas under-a period of 35min under a magnetic rotation speed1 (700 rpm; 1 rpm=0.1047 rad s1). After extraction,mixturewas retracted into the syringe. The hollowetached and discarded. The volume of the extract

analyt ica ch im ica acta 6 0 5 ( 2 0 0 7 ) 147152 149

was adjusted to 3L to remove the residual sample solution;this was then injected into the GCMS system for analysis.

2.6. SPM

A 10-mL sadjusted todirect immlibrium wadesorbed inPossible cainjector forto conrm

3. Re

The extracnature of sosample pHfactor, denthat beforeefciency [

In prelimPDMS-DVBextraction.results and

Fig. 1 shSPME and Bcentrationswas barely

In HFM-solvent thaing from thof polypropcompatiblevapor presBased on thn-nonane, csolvent HFgave compnonane. Holoss of solThus, hexaTo improvebinary-solvrelative distures withhigher resption, differthe resultsof 1:2 wasvent(s) aftepoor resulatively betthe tolueneextractionenhanced pthe electrosanalytes. T

GCPMEge wPP. Potep, (4) phorate, (5) disulfoton, (6) methyl parathion,yl parathion. GCMS conditions are given in the text.

lvents can enhance extraction to this extent. Previousn LLEwith binary-, even ternary-solvent systems [29,30]t provide any reasons for extraction enhancement. Per-his phenomenon merits further study.eries of extraction times was investigated by extract-ueous solution containing 20gL1 of each analytead s1 stirring speed. Fig. 3 shows the extraction times with partition coefcients of BN-LPME and SPME. BN-nd SPME are equilibrium-based extraction processes.lly, the equilibrium time is selected as extraction timese techniques. Various extraction times from5 to 40min

Extraction efciency of various binary-solventres in BN-LPME.E procedure

ample solution (pH and salt concentration were8 and 10% (w/v) respectively) were extracted by

ersionof theberwith stirring (at 105 rad s1). Equi-s established after 30min. Finally, the ber wasthe injection-port of the GC for 3min at 250 C.

rryover was minimized by keeping the ber in thean additional 5min. Blanks were run periodicallythe absence of contaminants.

sults and discussion

tion parameters affecting BN-LPME such as thelvents, ratios of solvent mixtures, extraction time,and salt addition were optimized. The enrichmented as the ratio of peak areas after extraction andextraction, was used to evaluate the extraction

1921].inary experiments, SPME with the PDMS, PA and

coatedberswere evaluated andoptimized forOPPThe PDMS-DVB ber gave the most satisfactorywas therefore used for further experiments.ows chromatograms of extracts after HFM-LPME,N-LPME of samples spiked at the same OPP con-(20gL1). Note that methyl parathion (peak 6)

detected by SPME.LPME it is essential to choose a suitable organict can withstand high speed stirring without leak-e HFM during extraction. Since the HFM is madeylene and is hydrophobic, solvents need to bewith it. Moreover, these should also have low

sures to reduce prevent loss during extraction.ese considerations, we selected toluene, n-octane,yclohexane and n-hexane for this study. In single-M-LPME, extraction using n-hexane and toluenearatively higher response than n-octane and n-wever, due to the higher volatility of n-hexane,vent after a 30-min extraction was a problem.ne on its own was not suitable for extraction.the extraction efciency of the toluene, various

ent mixtures were considered. Fig. 2 shows thetribution constant of various binary-solvent mix-respect to toluene. Toluene:hexane mixtures gaveonse factors than other binary mixtures. In addi-ent ratios of toluene:hexane were evaluated andare shown in Table 1. A ratio of toluene:hexanefound to be unsuitable for BN-LPME (loss of sol-r 30min). A toluene:hexane ratio of 2:1 also gavets. A 1:1 toluene:hexane mixture gave compar-ter results (Table 1). In comparison with SPME,:hexane (1:1) binary-solvent mixture gave higherresponse (Table 1). This could be due to theolarity/polarizability of the solvent mixture andtatic interactions of the solvent mixtures with thehis is not clear at this time how the combination of

Fig. 1 HFM-Ldrainaeach O(3) sulf(7) eth

two sowork odid nohaps t

A sing aqat 73 rproleLPME aGenerafor the

Fig. 2 mixtuMS traces of spiked wastewater extracts after (a)(toluene), (b) SPME and (c) BN-LPME. Eachastewater sample was spiked with 20gL1 ofeaks: (1) triethylphosphorothioate, (2) thionazin,

150 analyt ica ch im ica acta 6 0 5 ( 2 0 0 7 ) 147152

Table 1 Enrichment factors of BN-LPME at varioustoluene:hexane solvent ratios, and of SPME

Analytes Toluene:hexane SPME

Toluene 2:1 1:1 1:2

Triethylphosphorothioate 162 181 217 174 49Thionazin 97 92 105 106 28Sulfotep 204 223 299 298 260Phorate 73 94 142 105 114Disulfoton 168 218 265 263 244Methyl parathion 153 190 290 248 34Ethyl parathion 77 96 246 106 124

was evaluated with BN-LPME. The results showed that theanalyte response factors increased up to 35min of extractionand then no additional increment in the extraction efciency(Fig. 3a). SPME extraction was also evaluated from 10 to 40minand the results are shown in Fig. 3b. The responses factorsincreased with extraction time and equilibrium was estab-lished at 30min. Thus, 35-min and 30-min extraction timeswere selected for BN-LPME and SPME, respectively.

The effect of sample pH on BN-LPME was studied. SamplepH was adjusted from 2 to 12 and was measured by adding6M HCl and 10% (w/v) NaOH to the samples. Through thismeans, the degree of ionization of the analytes affects theirwater solubility and, thus, extractability. As expected, due tothe neutrality of the analytes, extraction efciency of most ofthe OPPs was not affected by sample pH (Fig. 4a). The effect ofsalt on the extraction efciencywas studied by adding sodiumchloride 5, 10, 15, 20 and 30% (w/v) to samples. The OPPs are

Fig. 3 BNConcentrat

Fig. 4 Inchloride ad

soluble in water and it was found that the extractioncy was enhanced with the addition of salt concen-s up to 5% after which efciency decreased (Fig. 4b).5% salt concentration was selected for all subsequentments. In SPME, various sample pH and salt concentra-ere also optimized. Similar response proles to those

ed for BN-LPME were observed.valuate the practical applicability of the proposed BN-nd SPME, linearity, repeatability, relative recoveries andof detection (LOD) were investigated under optimizedions. The GC peak area counts were plotted againstpective analyte concentrations to generate calibration. The calibration plots were linear over the range of50gL1 with correlation coefcient (r) between 0.994

999 for BN-LPME. For SPME, the linear range was from50gL1 with linearity between 0.966 and 0.998. Thety of the calibration curves, constructed from the anal-highlyefcientrationThus,experitions wobtain

To eLPME alimitsconditthe rescurves0.25 toand 0.0.5 tolineari-LPME and SPME extraction time prole.ion, 20gL1 of each analyte.

ysis of spikLODs for aldecreasingjust detecteranged fromLPME and SLODs, blanryover occuThe repeatcate experiThe relativuence of sample pH (a) and effect of sodiumdition (b) on BN-LPME.ed samples was satisfactory in both methods. Thel target analytes were determined by progressivelythe concentrations of analytes until signals wered at a signal-to-noise ratio of 3 (S/N=3). The LODs0.3 to 11.4ngL1 and 3.1 to 120.5ngL1 for BN-

PME, respectively (Table 2). When determining theks were carried out to conrm that no sample car-rred. Three replicates were used to calculate LODs.ability of GC peak areas was studied for six repli-ments for an aqueous sample spiked at 5gL1.e standard deviations (R.S.D.s) of all of the ana-

analyt ica ch im ica acta 6 0 5 ( 2 0 0 7 ) 147152 151

Table 2 Linearity range of calibration plots, limits of detection (LODs), precision (% R.S.D.s) of BN-LPME and SPME

Analytes a b

s1)

Triethylpho 1Thionazin 2Sulfotep 3Phorate 6Disulfoton 5Methyl par 4Ethyl parat 7

a Linearityb Linearity

Table 3 (n=

Analytes

%)

TriethylphoThionazinSulfotepPhorateDisulfotonMethyl parEthyl parat

nd: not det

lytes for BNIt is worthBN-LPME (a

4. Re

The proposysis of wassamples. Thto real samdened asto spiked uthe effect otimes. ResuTable 3. Thfor BN-LPMcorrespondSPME,

152 analyt ica ch im ica acta 6 0 5 ( 2 0 0 7 ) 147152

[11] C. Basheer, J.P. Obbard, H.K. Lee, J. Chromatogr. A 1068 (2005)221.

[12] N.J. Simpson, Solid-Phase Extraction: Principles, Techniques,and Applications, Marcel Dekker, New York, and Basel, 2000.

[13] Y. Yang, D.J. Miller, S.B. Hawthorne, J. Chromatogr. A 800(1998) 257.

[14] C. Basheer, V. Suresh, R. Renu, H.K. Lee, J. Chromatogr. A1033 (2004) 213.

[15] C. Basheer, A. Parthiban, A. Jayaraman, H.K. Lee, S.Valiyaveettil, J. Chromatogr. A 1087 (2005) 274.

[16] G. Shen, H.K. Lee, Anal. Chem. 75 (2003) 98.[17] L. Zhu, C.B. Tay, H.K. Lee, J. Chromatogr. A T963 (2002) 231.[18] S. Pedersen-Bjergaard, K.E. Rasmussen, Anal. Chem. 71

(1999) 2650.[19] C. Basheer, H.K. Lee, J.P. Obbard, J. Chromatogr. A 968 (2002)

191.[20] C. Basheer, R. Balasubramanian, H.K. Lee, J. Chromatogr. A

1016 (2003) 11.

[21] C. Basheer, H.K. Lee, J.P. Obbard, J. Chromatogr. A 1022 (2004)161.

[22] S. Pedersen-Bjergaard, K.E. Rasmussen, Trends Anal. Chem.23 (2004) 1.

[23] S. Pedersen-Bjergaard, K.E. Rasmussen, J. Chromatogr. B 817(2005) 3.

[24] T.S. Ho, T.G. Halvorsen, S. Pedersen-Bjergaard, K.E.Rasmussen, J. Chromatogr. A 998 (2003) 61.

[25] E. Psillakis, N. Kalogerakis, Trends Anal. Chem. 22 (2003) 565.[26] M.A. Ali, P.J. Baugh, Intern. J. Environ. Anal. Chem. 83 (2003)

909.[27] V.V. Prezhdo, M. Jagello, I.I. Mielnik, M.V. Prezhdo, Sep. Sci.

Technol. 37 (2002) 2875.[28] H.G. Ugland, M. Krogh, K.E. Rasmussen, J. Chromatogr. A 749

(2000) 85.[29] N.A. Darwish, M.A. Abdelkarim, N. Hilal, I. Ashour, J. Chem.

Eng. Data 48 (2003) 1614.[30] J.-K. Kim, D.-W. Park, J. Chem. Eng. Data 50 (2005) 625.

Determination of organophosphorous pesticides in wastewater samples using binary-solvent liquid-phase microextraction and solid-phase microextraction: A comparative studyIntroductionExperimental sectionChemicals and reagentsMaterialsInstrumentationWastewater samplesHollow fiber-protected liquid-phase microextractionSPME procedure

Results and discussionReal world water sample analysisConclusionsAcknowledgementsReferences