1492Ch24

Manahan,StanleyE.CHEMICALANALYSISOFWATERANDWASTEWATEREnvironmentalChemistryBocaRaton:CRCPressLLC,200024CHEMICALANALYSISOFWATERANDWASTEWATER____________________________________________________24.1.GENERALASPECTSOFENVIRONMENTALCHEMICALANALYSISScientists’understandingoftheenvironmentcanonlybeasgoodastheirknow-ledgeoftheidentitiesandquantitiesofpollutantsandotherchemicalspeciesinwater,air,soil,andbiologicalsystems.Therefore,proven,state-of-the-arttechniquesofchemicalanalysis,properlyemployed,areessentialtoenvironmentalchemistry.Nowisaveryexcitingperiodintheevolutionofanalyticalchemistry,characterizedbythedevelopmentofnewandimprovedanalysistechniquesthatenabledetectionofmuchlowerlevelsofchemicalspeciesandavastlyincreaseddatathroughput.Thesedevelopmentsposesomechallenges.Becauseofthelowerdetectionlimitsofsomeinstruments,itisnowpossibletoseequantitiesofpollutantsthatwouldhaveescapeddetectionpreviously,resultingindifficultquestionsregardingthesettingofmaximumallowablelimitsofvariouspollutants.Theincreasedoutputofdatafromautomatedinstrumentshasinmanycasesoverwhelmedhumancapacitytoassimilateandunderstandit.Challengingproblemsstillremainindevelopingandutilizingtechniquesofenvironmentalchemicalanalysis.Nottheleastoftheseproblemsisknowingwhichspeciesshouldbemeasured,orevenwhetherornotananalysisshouldbeperformedatall.Thequalityandchoiceofanalysesismuchmoreimportantthanthenumberofanalysesperformed.Indeed,apersuasiveargumentcanbemadethat,givenmoderncapabilitiesinanalyticalchemistry,toomanyanalysesofenvironmentalsamplesareperformed,whereasfewer,morecarefullyplannedanalyseswouldyieldmoreusefulinformation.Inadditiontoadiscussionofwateranalysis,thischaptercoverssomeofthegeneralaspectsofenvironmentalchemicalanalysisandthemajortechiquesthatareusedtodetermineawiderangeofanalytes(speciesmeasured).Manytechniquesarecommontowater,air,soil,andbiologicalsampleanalysesandreferenceismadetotheminchaptersthatfollow.©2000CRCPressLLCErrorandQualityControlAcrucialaspectofanychemicalanalysisisthevalidityandqualityofthedatathatitproduces.Allmeasurementsaresubjecttoerror,whichmaybesystematic(ofthesamemagnitudeandsamedirection)orrandom(varyinginbothmagnitudeanddirection).Systematicerrorscausethemeasuredvaluestovaryconsistentlyfromthetruevalues,thisvariationisknownasthebias.Thedegreetowhichameasuredvaluecomesclosetotheactualvalueofananalyticalmeasurementiscalledtheaccuracyofthemeasurement,reflectingbothsystematicandrandomerrors.Itisessentialfortheanalysttodeterminetheseerrorcomponentsinthemeasurementofenvironmentalsamples,includingwatersamples.Theidentificationandcontrolofsystematicandrandomerrorsfallsinthecategoryofqualitycontrol(QC)procedures.Itisbeyondthescopeofthischaptertogointoanydetailonthesecrucialproceduresforwhichthereaderisreferredtoaworkonstandardmethodsfortheanalysisofwater.1Inorderforresultsfromalaboratorytobemeaningful,thelaboratoryneedsaqualityassuranceplanspecifyingmeasurestakentoproducedataofknownquality.Animportantaspectofsuchaplanistheuseoflaboratorycontrolstandardscon-sistingofsampleswithveryaccuratelyknownanalytelevelsinacarefullycontrolledmatrix.SuchstandardreferencematerialsareavailableintheU.S.formanykindsofsamplesfromtheNationalInstituteofStandardsandTechnology(NIST).Manyenvironmentalanalytesarepresentatverylowlevelswhichchallengetheabilityofthemethodusedtodetectandaccuratelyquantifythem.Therefore,thedetectionlimitofamethodofanalysisisquiteimportant.Definingdetectionlimithaslongbeenacontroversialtopicinchemicalanalysis.Everyanalyticalmethodhasacertaindegreeofnoise.Thedetectionlimitisanexpressionofthelowestconcentrationofanalytethatcanbemeasuredabovethenoiselevelwithaspecifieddegreeofconfidenceinananalyticalprocedure.Inthedetectionofanalyte,twokindsoferrorscanbedefined.ATypeIerroroccurswhenthemeasurementfindsananalytepresentwhenitactuallyisabsent.ATypeIIerroroccurswhenthemeasure-mentfindsananalyteabsentwhenitisactuallypresent.Detectionlimitscanbefurthercategorizedintoseveraldifferentsubcategories.Theinstrumentdetectionlimit(IDL)istheanalyteconcentrationcapableofproducingasignalthreetimesthestandarddeviationofthenoise.Thelowerlevelofdetection(LLD)isthequantityofanalytethatwillproduceameasurablesignal99percentofthetime;itisabout2timestheIDL.Themethoddetectionlimit(MDL)ismeasuredliketheLLDexceptthattheanalyteistakenthroughthewholeanalyticalprocedure,includingstepssuchasextractionandsamplecleanup;itisabout4timestheIDL.Finally,thepracticalquantitationlimit(PQL),whichisabout20timestheIDL,isthelowestlevelachievableamonglaboratoriesinroutineanalysis.24.2.CLASSICALMETHODSBeforesophisticatedinstrumentationbecameavailable,mostimportantwaterqualityparametersandsomeairpollutantanalysesweredonebyclassicalmethods,whichrequireonlychemicals,balancestomeasuremasses,burets,volumetricflasks©2000CRCPressLLCandpipetstomeasurevolumes,andothersimplelaboratoryglassware.Thetwomajorclassicalmethodsarevolumetricanalysis,inwhichvolumesofreagentsaremeasured,andgravimetricanalysis,inwhichmassesaremeasured.Someofthesemethodsarestillusedtoday,andmanyhaveb