I20cncn

RegioselectivityoftheOxygenAddition-InducedDechlorinationofPCBsandDDTMetabolitesinElectronCaptureMassSpectrometry*F.
1.
LepineandS.
MilotInstitutArmand-Frappier,Laval,Quebec,CanadaO.
A.
MarnerBiomedicalMassSpectrometryUnit,McGillUniversity,Montreal,Quebec,CanadaTheelectroncapturemassspectraof2835Cl-labeledpolychlorinatedbiphenyls(PCBs)and437Cl-labeled1,1,l-trichloro-2,2-bis(p-chlorophenyOethane(DDT)metaboliteswereobtainedbyusinga20%oxygeninmethanemixtureasthereagentgas.
ThedegreeofregioselectivityofthePCBoxygenaddition-induceddechlorinationreactionwasdeterminedbymeasure-mentoftheresidualamountoflabelintheM-19ionsproducedbyadditionof0zandsubsequentlossofOClfromthemolecule.
ChlorinewaslostinarandommannerfromthePCBs,contrarytothedechlorinationreactionobservedwhenmethanealonewasused.
FortheDDTmetabolites,manydechlorinationreactionswereobservedinadditiontotheonethatgeneratedtheM-19ions.
LossofCl,lossofClz,andadditionofOzwiththelossofoneortwoHClmoleculesalsowereseen.
Thesevariousdechlorinationreactionsinvolvedonlythealiphaticchlorines.
Additionof0zfollowedbylossofClatthebetapositionof2,2-bis(4-37Cl-chlorophenyD-1-chloroethyleneand2,2-bis(4-37Cl-chlorophenyl)-1,1-dichloro-ethylenemaybeduetotheabilityofthediphenylmethanemoietytostabilizetheintermediates.
Formationofanionthatcorrespondsto4,4'-dichlorobenzophenonealsowasobservedforthreeoftheselabeledDDTmetabolites.
(JAmSocMassSpectrom1996,7,66-72)Electroncapturenegativechemicalionization(ECNCnmassspectrometryisanionizationtechniquethatprovidesexcellentsensitivityandconsiderableselectivityforthedetectionoforganochlorinecompoundsinenvironmentalmatrices[1].
AcommonfeatureoftheECNCIspectrumofmanyofthesechlorinatedcompoundsisthepresenceofionsat19ulowerthanthemolecularion[2-4].
Theseionsareformedinreactionsbetweentheorganochlorineandtraceamountsofoxygenpresentinthesourceandcorrespondtotheadditionofoneoxygenatomwiththesimultaneouslossofonechlo-rineatom.
Manyinvestigatorshavedeliberatelyintroducedoxygenintothesourceandhavereportedthattheoxygenaddition-induceddechlorinationreactionscouldprovidesomestructuralinformationforpoly-chlorinateddiphenylethers[5],dibenzofurans[6],anddibenzodioxins[7].
Collision-induceddissociationAddressreprintrequeststoDr.
F.
Lepine,InstitutArmand-Frappier,531desPrairiesBoulevard,Laval,Quebec,Canada,H7N4Z3.
Presentedinpartatthe43rdASMSConference,Atlanta,GA,May1995.
1996AmericanSocietyforMassSpectrometry1044-0305/96/$15.
00ssor1044-0305(95)00588-9(CID)oftheM-19ionsofavarietyofpolychlorinatedbiphenyls(PCB)congenersalsowasperformedforthesamepurpose[8].
Oneobjectiveoftheseexperimentswastocorrelatethedatawiththechlorinesubstitutionpatternsofeachcongener,becausethechromato-graphicseparationofsomeofthesecongenerssome-timesisdifficult.
Determinationofthecompositionofamixtureofcongenersisimportantinenvironmentalstudies,becausetoxicitiesvaryenormouslywiththeirsubstitutionpattern.
WeobservedthatthedechlorinationreactionsofPCBsandvariousL'l.
l-trichloro-Z.
z-bisfp-chloro-phenyOethane(DDT)metabolitesinamassspectrome-terwereregioselective,especiallyunderECNCIcondi-tions[9,10],anditbecameofinteresttoseeifsuchregioselectivityalsocouldbeobservedfortheoxygenaddition-induceddechlorinationofthesecompounds.
MaterialsandMethodsThesimplesynthesesofthevariouslabeledDDTmetabolitesandPCBsweredescribedpreviously[10,11].
ForthePCBs,synthesisconsistedofthechlo-ReceivedJune30,1995RevisedAugust4,1995AcceptedAugust4,1995JAmSocMassSpectrom1996,7,66-72OXYGENADDmON-INDUCEDDECHLORINATION67DDTMetabolitesFourDDTmetaboliteslabeledwithone37Cloneacharomaticringwerestudied.
Thesewere2,2-bis{4-37CI-chlorophenyl)-l-chloroethylene(p,p'-DDMU;Figure2),2,2-bis{4-37Cl-chIorophenyl)-I,I-dichloro-ethylene(p,p'-DDE;Figure3),2-{4-37Cl-chIorophenyl)-2-{2-37Cl-chIorophenyl)-I,I-dichIoroethane.
(a,p'-DDD;Figure4),and2,2-bis{4-37CI-chlorophenyl)-I,I-dichlo-roethane(p,p'-DDD;Figure5).
ThesecompoundsareknownmetabolitesofcommercialDDTpreparations.
AsforthePCBs,becausetheM-19ionsoftheDDTmetaboliteswerenotsufficientlyintensetobeana-presentstudy.
Undertheseconditions,themassspec-traofmostofthePCBsconsistalmostexclusivelyofM-19ions.
Figure1presentstheoxygen-methaneECNCImassspectrumof2,2',3,3',435CI,4'35CI-hexachIorobiphenyl.
Thisspectrumisoneoftheveryfewthatshowamolecularion.
Themassspectrumofthesamecom-pound,withmethaneasthemoderatorgas,ispre-sentedforcomparison.
ThepositionsofthelabelsonthePCBs,theobservedabundancesoftheP,P+2,andP+4ionsoftheM-19clusters,andthepercent-agesofthelabelretainedintheM-19ionsarepre-sentedinTable1.
TheM-19ionsofallthepentachloro-PCBsretainedbetween77and79%ofthelabel.
Forthehexachloro-PCBs,thisvaluevariedbe-tween81and84%.
1I20liDO2802"0arination,byusing35Clz,ofthesixsymmetricaltetra-chlorobiphenylsinanorganicsolvent.
Theresultingmixturesofcompoundswereanalyzedbycapillarygaschromatography-massspectrometry(GC-MS)andtheidentityofeachpenta-andhexachlorobiphenylwasestablishedbytheirretentiontimes,byusingthedataofMullinetal.
[12],whopublishedtheretentiontimesofall209PCBs.
Thisalsoallowedustodeter-minethepositionofthe35Cllabelswithinthemolecules.
Theaddedlabel35Clenrichmentwas99.
6%,asdeterminedbyanalysisofthemolecularionclusterintheelectronimpactmode[11].
Calculationofthepercentageofresiduallabelindechlorinatedionsofinterestwasperformedbynormalizationoftheintensi-tiesofP+2andP+4ionsofagivenclusterrelativetothe35Clmonoisotopicion.
Thisnormalizedprofilethenwascomparedwithacomputer-generatedseriesofprofilesthatcorrespondtodecreasingvaluesofthe37Clj35Clratio[11).
FortheDDTmetabolites,37Cl-chlorobenzenefirstwassynthesizedandreactedwithtrichloroacetalde-hydetoproducep,p'-ando,p'-DDT.
Theresultingmixturewasthengammairradiatedtoproducedp,p'-ando,p'-DDD,andtreatmentwithbaseaffordedp.
p':DOEandp,p'-DDMU.
Alltheseproductswereob-tainedasmixtures,buteasilycouldbeseparatedandidentifiedbycapillaryGC-MS[10).
Theadded37ClenrichmentfortheDDTmetaboliteswas94.
6%[10].
Inthiswork,thevariousmixturesof35Cl_and37Cl-IabeledcompoundswereanalyzedbyGC-MSwithaHewlett-Packard(Avondale,PA)5988Amassspec-trometerbyusinga30-mDB-lcolumn.
ThePCBanalyseswereperformedintriplicate.
Thereagentgaswasa20%mixtureofoxygeninmethane.
Notethatthismixtureisexplosiveandthatprecautionsmustbetakentoavoidsourcesofignition.
Thepurityofthemethaneusedwasgreaterthan99.
99%(Linde,Mon-treal,Canada).
Thesourcepressurewas0.
3torrandthesourcetemperaturewas150°C.
Becauseofanintensesignalatmrz233-235duetooxidationoftheRefilament(Re03),thescanrangewas240-400u.
TheDDTmetaboliteswereanalyzedunderthesameconditions.
mlzFigure1.
ECNCIspectraof2,2',3,3',435Cl,4'35Cl-hexachloro-biphenyl(IUPAC128)with(a)oxygen-methaneand(b)methane.
ResultsPCBsToevaluatetheregioselectivityofdechlorination,6penta-and22hexachloro-PCBslabeledatvariousposi-tionswithoneortwo35Clwereused.
Theregioselec-tivityofthedechlorinationreactionisreflectedbytheresidualamountsoflabelintheM-19ioncluster.
Inapreviousexperiment[9],methanealonewasusedasthemoderatorgasandtheM-19ionswereobserved,buttheirabundancesweretoosmalltoallowreliablequantificationoftheirlabeledchlorinecontent.
Toin-creasetheabundancesoftheseions,a20%(byvol-ume)oxygeninmethanemixturewasusedintheb2.
00ReO2802110/liDO".
00""0"8068LEPINEETAL.
JAmSocMassSpectrom1996,7,66-72Table1.
PercentageofresiduallabelintheM-19ionsofvarious35CI-labeledPCBsinECNCIbyusinga20%oxygeninmethanemixtureasreagentgasAbundancesPercentageofIUPACnumberofP,P+2,residuallabelofunlabeled35CllabelpositionandP+4inM-CI+0PCBsinthelabeledPCBions"ions832,2',3,3',535CI-penta-27,8107929,01810,669822,2',3,3',435CI-penta-10,9927811,9925,0111282,2',3,3',435CI,4'35CI-hexa_53,8108260,89324,4051292,2',3,3',435CI,535CI_hexa_61,7248268,80128,1901302,2',3,3',435CI,5'35CI-hexa_86,3878295,08636,4431312,2',3,3',435CI.
635CI_hexa_267,82983299,133112,7681322,2',3,3',435CI,6'35CI_hexa_129,04983140,16251,8071332,2',3,3',535C1.
5'35CI_hexa_60,9238363,50723,5431352,2',3,3',535CI,6'35CI-hexa_228,60583239,25687,3321362,2',3,3',635CI,6'35CI_hexa_16,7578416,7376,0581282,2',335CI,3'35CI,4,4'-hexa-4,495844,4511,5811382,2',335CI,4A',5'35CI-hexa-14,0458414,3463,3841532,2',4,4',535CI,5'35CI-hexa-15,0158415,1865,2351362,2',335CI,3'35CI,6,6'-hexa-56,1848165,05627,883105235CI,3,3',4,4'-penta-169,99177198,96980,796118235CI.
3',4,4',5-penta-336,26278355,555154,151128235CI,2'35C1,3,3',4,4'-hexa-4,966835,2672,019JAmSocMassSpectrum1996,7,66-72Table1.
(col/til/lied)OXYGENADDITION-INDUCEDDECHLORINATION69IUPACnumberofunlabeledPCBs15316713815615711112715916216913335CllabelpositioninthelabeledPCB235C1,2'35C104o4',5,5'-hexa-235C1,3',4,4',5,5'35CI_hexa_235C1,2'35C1,304o4',5'-hexa-235CI,3,3'o404',535CI_hexa_235CI,3,3'o404',5'35CI-hexa-235CI,3,3',5,5'-penta-3,3'0435CI,5,5'-penta-235CI,3,3'o435CI,5,5'-hexa-235CI,3,3'o4'35CI,5,5'-hexa-3,3'0435Clo4'35CI,5,5'-hexa-235C1,2'35CI,3,3',5,5'-hexa-AbundancesPercentageofofP,P+2,residuallabelandP+4inM-CI+Oions"ions18,9348319,3027,31816,3418218,1587,2711404848315,2825,67736.
6928242,06617,18318.
3218121.
2439,024144.
53878163.
15262,62622,5827726.
24811,00823.
1848127.
88611.
235494,99282553.
625219,96549.
0708256.
24223,593639,25584654.
801221,851ThesevaluesrepresenttheaveragedabundancesoftheionsoftheclusterthatcorrespondstoM-19.
Pistheioninwhichallthechlorinesare35CI.
TheabundancesofP.
P+2,andP+4areusedtodeterminetheamountofresiduallabelintheM-19ions[1OJ.
lyzedwithmethane,thesame20%oxygeninmethanemixturewasused.
However,undertheseconditions,extensivefragmentationofthehighermassionswasevidentandledtolowintensitiesfortheionsofinterest,whichdidnotallowaccuratemeasurementoftheresiduallabelintheseions.
However,qualitativeinformationcouldbededucedfromthesespectra.
ThestructureandmassspectraofthelabeledDDTmetabo-litesarepresentedinFigures2-5.
Thespectraoftheunlabeledcompounds,takenunderthesamecondi-tions,alsoarepresentedforcomparison.
Theprofilesofthemajorionclustersofthesespectrashowthat,ineverycase,thearomaticchlorinesareretained;becauseofthis,andtosimplifythedescriptionoftheseionclusters,theywillbereferredtobythemostabundantionofthecluster.
Thespectrumoflabeledp,p'-DDMU(M'-=mlz286;notvisibleinFigure2)includesionsatmrz282(M-4),267(M-19),and254.
Nomolecularionorfrag-mentsthatcorrespondtothelossofoneCl,asob-tainedwithmethane[10],wereobserved.
TheM-4ion(Figure6,5)isprobablyproducedbytheadditionofO2withlossofHCl.
Structure4aanditsresonanceform4b(Figure6;R=H)areproposedforthemlz267ionandtakeintoaccountthattheformationofthisionisbythelossofanaliphaticchlorine.
Theintenseionatmlz254probablyisrelatedstructurallyto4,4'-dichlorobenzophenone6,andcouldbeformedfromthemIz282ion5throughlossofCOorbycycloadditionof02-tothemolecule,whichformsthefour-memberedringintermediate3(R=H)thatelimi-natesformylchloride.
70LEPINEETAL.
JAmSocMassSpectrom1996,7,66-725'o&1~.
2eoaDoH370---0-1-0-370/c"010Hp,p'·DDD2lI7,au,a.
820~,.
conoba2.
0880eee2700280300atOabmlzFigure4.
Oxygen-methaneECNCIspectraof(a)labeledo.
p'>DODand(b)unlabeledo,p'-DDD,4aand4bareproposedforthel1l/z301ion,byanalogywiththe111/Z267ionofp,p'-DDMU.
The111/Z301ion4a,bisprobablytheprecursorof7(111/Z273)throughlossofCO(Figure6;R=CD.
Formationof6inthep,p'-DDEspectrumcouldbeexplainedby100010o330.
eo3003ta320SlIDZ8030031.
0a203130as8,370---0-""-0-370/c"H0p,p'·DDMUa8a/370---0-"---0-370/c"o0p,p'-ooe273,a78compositionpathwaysofp,p'-DDMUandp,p'-DDE.
(1)(2)(3)M'-+02+=!
MOi-M+0i-+=!
MOi-M+0i-+=!
M'-+O2PCBsThesecondmechanisminvolvesthereactionofsu-peroxide(Oi-)withtheneutralPCBmoleculetopro-ducetheM02radicalanion(eq2).
Knightonetal.
[14],whousedanoxygen-methanemixtureasthereagentgas,demonstratedthatthismechanismwasnotre-sponsiblefortheformationoftheM-19ionsforavarietyofchlorinatedcompounds.
However,theoxy-genconcentrationstheyused(lessthan2000ppm)weremuchlowerthaninthepresentwork(20%).
Athigheroxygenconcentrations,theprobabilityofelec-troncapturebyoxygenincreases,andthisincreasecouldincreasetheproportionoftheM-19ionsformedbyattackbysuperoxideontheneutralPCBs.
DiscussionInapreviouspublication,thedechlorinationreactionofPCBsinECNCIwasreportedtobehighlyregiose-lectiveusingmethaneasthemoderatorgas[9].
Undertheseconditions,thelabeledPCBsallpresentedabun-dantmolecularandM-34ions,whichindicatedthattheadditionofhydrogenwasinvolvedinthecourseofthedechlorinationreaction.
Inthepresentexperiment,undersimilarconditionsbutforthepresenceofoxy-geninthesource,theresultspresentedinTable1indicatethattheoxygenaddition-induceddechlorina-tioniscompletelyrandom.
Labelretentionsareverycloseto80and83%forthepenta-andhexachlorocongeners,respectively,andcorrespondtorandomlossofonechlorineatomoutofthefiveorsixinitiallypresent.
ThisobservationalsoisconfirmedbytheresiduallabelcontentforPCBcongenersthathavethesamechlorinesubstitutionpatternbutdifferentlabelpositions,asincongeners128,138,153,and133.
AsfortheECNCIexperimentthatusesmethane,noisotopeeffectcouldbedetected.
TwomechanismswerepostulatedtoexplaintheformationoftheM-19ionsobservedintheECNCIspectraofavarietyofchlorinatedcompounds[13].
Thefirstmechanisminvolvestheradicalanionofthecom-poundthatreactswithO2toproducetheM02radicalanion(eq1),whichthenundergoeslossofaCIOradicalthatproducestheM-19ions.
Insuchamecha-nism,thelackofregioselectivityobservedinthepresentstudywithPCBscouldbeaccountedforbydelocalizationofthecapturedelectrononthebiphenylsystemandreactionwithO2followedbyrandomlossofCIasCIO':species,followedbylossofchlorine.
Thedegradationpathwaysofo.
p'>andp,p'-DDDarethenquitedif-ferentalthoughdechlorinationproceedsonlyatthealiphaticchlorinesinbothcases.
727362542544bloCo28228226730134a,b523~n.
.
.
4a!
-RC~_~OCI~:O~·--.
py~l·o-CO056$=37C1-@-p,p'·DDMUR=H286bp,p'-DDER=CI320theformationof3withthesubsequentlossofphos-gene(R=cn.
Thespectrumoflabeledo,p'-DDD(M'-=mZz322;Figure4)showsmolecularni/282(M-40)and252(M-70)ions.
Themrz282ioncorrespondstoadditionofO2tothemolecularionwiththelossoftwoHCImolecules.
ThisisequivalenttolossofasecondmoleculeofHCIfromtheM-4ioninthep,p'-DDMUspectrum.
Themolecularandthem/z252ionsalsowereobservedwhenmethanewasused[10].
Inthepresentexperiment,themrz252ionclustercontainsalmostexclusivelytwo37CI,whichindicatesthatthetwochlorinesthatarelostarealiphatic.
Thesamephenomenonwasobservedforthisionwhenmethanewasused.
The111/Z254ion,whichwasobservedforalloftheotherDDTmetabolites,wasnotpresentinthisspectrum.
.
Thespectrumofthelabeledp,p'-DDD(Figure5)includesionsat111/Z282,267,and254andisalmostidenticaltothep,p'-DDMUspectrum.
Itislikelythatinafirststep,p,p'-DDDundergoeslossofHCltoproduceamolecularionofp,p-DDMU.
Becausethespectraofthesetwocompoundsarecompletelydif-ferentwhenmethaneisusedasthemoderatorgas[10],theinitiallossofHCIobservedhereprobablyarisesfromabstractionoftherelativelyacidicahydrogen,possiblybyasuperoxide0i-radicalanionorrelated72LEPINEETAL.
JAmSocMassSpectrom1996,7,66-72Chargeexchangebetweensuperoxideandtheneu-tralcompound(eq3)alsowasinvokedasanotherreasonwhythesecondmechanismwasnotimportantinKnighton'sstudy.
Theextentofchargeexchangeiscontroled,inpart,bytheelectronaffinitiesofthetworeactants[15].
Foroxygen,thisvaluewasfoundtobe0.
44eV[16].
TherearenodataavailableontheelectronaffinitiesofPCBs,butKnightonreportedthattheelec-tronaffinitiesofhexachlorobenzeneandpenta-chlorobenzenewere1.
0and0.
7eV,respectively[14].
Becauseadecreaseinthenumberofchlorinesonthearomaticringdecreasestheelectronaffinity,andas-sumingonecanextrapolatetheresultsobtainedwithasimplearomaticringtoabiphenylsystem,onecanestimatethattheelectronaffinitiesofthePCBslistedinTable1,whichcontainanaverageofthreechlorineatomsoneacharomaticring,arenotmuchlargerthanthatofoxygen.
Becauseoftherelativelylargeamountofoxygeninthesource,chargeexchangeisprobablynotsufficienttodecreasetheconcentrationofsuperox-idetoapointthatitshouldnotbeconsideredanimportantreactantintheformationoftheM-19ions.
TheabsenceofmolecularionsinthespectraofmostPCBsalsofavorseq2,andtheobservedlackofregie-selectivitycouldbeduetorandomattackbysuper-oxideatanychlorine-bearingbiphenylcarbon.
Itisalsopossiblethatunderthesereactionconditions,bothmechanismsareimportant.
DDTMetabolitesIncontrasttothePCBs,allthedechlorinationreactionsobservedwiththeDDTmetaboliteswerecompletelyregiospecific:onlythealiphaticchlorineswerelost.
Theregioselectivityofthesereactionsisprobablyre-latedtotheabilityofthediphenylmethanemoietytostabilizearadicaland/oranegativechargebyreso-nancethroughthetwoaromaticrings.
Onlyp.
p'DDMUandp,p'-DDEpresentedM-19ions.
Forthesetwoions,oxygenadditionanddechlorinationoccuratthecarbonbetatothearomaticrings,whichallowsdelocalizationofthechargeonthearomaticringsthroughtheresonancestructure4b(Figure6).
Thecapacityofanionthatisstructurallyrelatedto4,4'-dichlorobenzophenone6tostabilizeanegativechargeprobablyalsoaccountsfortheabundanceofthern/z254ioninthespectraofthealpha-betaunsaturatedDDTmetabolites.
4,4'-DichlorobenzophenonealsoisproducedinthedecompositionofmanyDDTmetabolitesbygammairradiationinaliphaticsolventsinthepresenceofoxygen[17].
Gammairradiationgeneratessolvatedelectronsinthesolvent.
Inthepresenceofoxygen,thesuperoxideradicalisgeneratedandprobablyaddsatthealphacarbontoultimatelyproduce4,4-dichloro-benzophenone.
Theresultsobtainedinthepresentexperimentdemonstratethatthereisacloseparallelbetweenthesesolventandgas-phasereactions.
ConclusionTheregioselectivityofdechlorinationreactionsofvari-ousorganochlorinesinECNCIinthepresenceofoxy-genisverydependentuponthestructureoftheorganochlorine.
Contrarytowhatwasobservedprevi-ouslywithmethanealoneasthemoderatorgas,PCBsweredechlorinatedrandomlyinthepresenceofoxy-gen.
DDTmetabolites,ontheotherhand,weredechlo-rinatedregioselectivelyatthealiphaticsites.
AcknowledgmentsThisworkwassupportedbygrantsfromtheNaturalSciencesandEngineeringResearchCouncilofCanadaandfromtheMedi-calResearchCouncilofCanada.
References1.
Ong,V.
5.
;Hites,RA.
MassSpecirom.
Rev.
1995,13,259.
2.
Biros,F.
J.
;Dougherty,RC,Dalton,J.
arg.
MassSpecirom.
1971,6,1161.
3.
Levonowich,P.
F.
;Tannenbaum,H.
P.
;Dougherty,R.
C.
J.
Chem.
Soc.
Chem.
Comllll/n.
1975,597.
4.
Stemmler,E.
A.
;Hites,RA.
;Arbogast,B.
;Budde,W.
L.
;Deinzer,M.
L.
;Dougherty,Rc.
;Eichelberger,J.
W.
;Foltz,R.
L.
;Grimm,c,Grimsrud,E.
P.
;Sakashita,c.
Sears,L.
J.
Anal.
Chem.
1988,60,781.
5.
Busch,K.
L.
;Norstrom,A.
;Bursey,M.
M.
;Hass,J.
R;Nilsson,C.
-ABiomed.
MassSpectrolll.
1979,6,157.
6.
Fung,D.
;Boyd,RK.
;Safe,S.
;Chittim,B.
G.
Biomed.
MassSpecirom.
1985,12,247.
7.
Hass,RJ.
;Friesen,M.
;Hoffman,M.
K.
Org.
MassSpectrom.
1979,14,9.
8.
Guevremont,R;Yost,R.
A;Jamieson,W.
D.
Biomed.
Environ.
MassSpectrom.
1987,14,435.
9.
Lepine,F.
L.
;Milot,S.
;Reimer,M.
L.
J.
;Marner,O.
AargoMassSpecirom.
1994,29,133.
10.
Lepine,F.
L.
;Milot,S.
;Marner,O.
A.
J.
MassSpectrom.
1995,30,885.
11.
Lepine,F.
L.
;Milot,S.
;Reimer,M.
L.
J.
;Marner,O.
A.
argoMassSpecirom.
1992,27,1311.
12.
Mullin,M.
D.
;Pochini,C.
M.
;McCrindle,5.
;Romkes,M.
;Safe,S.
H.
;Safe,S.
Environ.
Sci.
Technol.
1984,205,103.
13.
Dzidic,1.
;Carroll,D.
1.
;Stillwell,RN.
;Homing,E.
C.
Anal.
Chem.
1975,47,1308.
14.
Knighton,W.
B.
;Bognar,J.
A;Grirnsrud,E.
P.
J.
MassSpectrolll.
1995,30,557.
15.
Grimsrud,E.
P.
;Chowdhury,S.
;Kebarle,P.
Int.
J.
MassSpecirom.
IonProcesses1986,68,57.
16.
Janousek,B.
K.
;Braurnan,J.
1.
GasPhaseIonChemistry,Vol.
2;Bowers,M.
,Ed.
;AcademicPress:NewYork,1979;p53.
17.
Lepine,F.
L.
;Brochu,F.
;Milot,S.
;Marner,O.
A.
;Pepin,Y.
J.
Agric.
FoodChem.
1994,42,2012.