K.Reddy-Nooneetal./J.Chromatogr.A1148(2007)145–151

149

Fig.4.SDME–SIMforthedeterminationoftotalbromideandiodide(0.5mgl 1each)(lowertrace),andbromate(0.1mgl 1)andiodate(0.08mgl 1)aftertreatmentofstandardwithAgCl(uppertrace)inastandardmixtureofhalidesandoxyhalides.Halideswerederivatizedto4-bromo-2,6-dimethylaniline(retentiontime8.13min)and4-iodo-2,6-dimethylaniline(retentiontime8.91min)andSDME.Theinternalstandard2,4,6-tribromoaniline(retentiontime9.51min).Inset,massspectraofpeakat8.13minandat8.91min.

diffusion lmaroundthedropmakingitsuccessivelymorepolarwithincreasingionicstrengththatreducedtherateofdiffusionoftargetorganicmolecules.

LPMEwithorganicsolventinthestandard askandwith-drawalof lportionofextractsettledinthe askabovethemeniscusmarkforanalysisbyGChaspreviouslybeenusedbyusasaconvenientmethodofsamplepreparation[32,44].ToluenewasusedasextractionsolventsinceitgaveoptimumextractionincomparisontoothersolventsaswereusedinSDME.Withconventional5mlstandard ask,downto50 lofsolventcouldbeusedconveniently.Contrarytothelattermethod,andsimilartoclassicalsolventextraction,analytemasstransfertoorganicphaseisfavoredbyhighsaltconcentrationinthismethod.3.3.Validation

Thechromatogramobtainedforiodideandbromideintablesalt,afterderivatizationandSDME,isshowninFig.4(lowertrace);theexcess2,6-dimethylanilinefromreactionwaselutedwithin5min(notshowninthechromatogram).Insetsarethemassspectracorrespondingtohalo-derivativepeaks.Theelec-tronionization(70eV)massspectrumcontainedmolecularionasthebasepeakalongwithdiagnosticfragmentions.Thechro-matographicpeaksforderivatizediodideandbromidewereidenti edbycomparisonofretentiontimeandmassspectraoftheauthenticsubstances.Thequantitativenatureofderivatiza-tionreactionwascon rmedbyGCwhenequalmolarmassesofauthentic4-iodo-and4-bromo-2,6-dimethylaniline,andofiodideandbromide,afterderivatization,ontheirseparatechro-matographyproducedpeakareas,relativetothatoftheinternalstandard,within3–5.5%.Theaverageconversionwas97.4%(RSD4%).Laboratorysamplesoftablesaltwerepreparedbyaddingknownamountsofiodideandbromidetosodiumchloridethatwasfoundtobefreefromanyiodideandbro-mide,andmixingthoroughly.Agreementbetweenthestandardadditionstotablesaltandseawater,andthemassesofiodideandbromidefoundbythepresentmethodservedtovalidate

thenewmethod.Theaveragerecoveryofspikeswas99.3%(range97.2–102.3%)withRSDof3.7%(range2.7–5.1%)foriodide,and99.6%(range97.4–103.2%)withRSDof2.5%(range1.6–4.2%)forbromide.TheresultsobtainedaregiveninTable1.Theaveragerecoveryofspikediodide/iodateandbro-mide/bromatetoseawaterwas99.7%(range96.8–105.7%)and99.2%(range96.7–105.2%),respectively.TheresultsobtainedaregiveninTable2.Chromatogramsobtainedintheanalysisofiodide/iodateandbromide/bromateintheirstandardsolutioninvolvingtheuseofAgClforseparationofhalidesisgiveninFig.4.Therecoveryofiodate(at0.08mgl 1level)andbromate(at0.1mgl 1level)was98.6%and97.4%,respectively.3.4.Calibrationgraphanddetectionlimits

Rectilinearcalibrationgraphs(atsixteenconcentrationlev-els;on2mlsamplesize)wereobtainedfor0.05 g–25mgl 1ofbromate/bromideandiodate/iodidebyderivatizationandGC–MS;thecorrelationcoef cient(r2)andlimitofdetection(LOD,S/N=3)were0.9986and20ngl 1ofbromate,15ngl 1ofiodate,20ngl 1ofbromideand10ngl 1ofiodide(byLPMEin50 loftoluene).ThecorrespondingvaluesusingSDMEwere0.9996and65,50,70and45ngl 1.Inprecisionstud-ies,sixaliquotsofsampleswereseparatelyderivatizedandoneinjectioneachwasmade.TheRSDinthedeterminationof0.02,0.1and10mgl 1ofiodideandbromidestandardsbythepresentmethodwas3.1%,3.0%and2.8%,respectivelyforiodide,and4.2%,3.0%and2.6%,respectivelyforbromide(byLPMEin50 loftoluene).TheRSDinthedeterminationof0.5 gl 1ofbromateandiodatewasrespectively,1.2%and1.8%byLPMEin50 loftoluene,and7.5%and6.7%bySDMEin2 loftoluene.Thus,LPMEwasmoresensitiveandprecisemethodthanSDME.LPME–GC–FIDcouldalsobeusedforthedetermi-nationof2 g–25mgl 1ofbromate/bromideandiodate/iodide;theaveragecorrelationcoef cient(r2)andLODwere0.9995and100ngl 1ofbromate,75ngl 1ofiodate,90ngl 1ofbro-mideand60ngl 1ofiodide.GC–FIDforthedeterminationof5 gl 1ofbromateisgiveninFig.5.