Transport and Development of Microemulsion- and Surfactant ... · Microemulsion-Stabilized Iron Nanoparticles in Porous Media Dennis Hsu Masters of Applied Science Graduate Department

TransportandDevelopmentofMicroemulsion-andSurfactantStabilizedIronNanoparticlesfor

InSituRemediation

By

DennisHsu

Athesissubmittedinconformitywiththerequirements

forthedegreeofMasterofAppliedScience

DepartmentofChemicalEngineeringandApplied

ChemistryUniversityofToronto

©CopyrightbyDennisHsu(2017)

ii

ColumnTransportStudyofSurfactant-andMicroemulsion-StabilizedIronNanoparticlesinPorous

Media

DennisHsu

MastersofAppliedScience

GraduateDepartmentofChemicalEngineeringandAppliedChemistryUniversityofToronto

2017

Abstract

Thisworkdescribesthemobilityassessmentsofmicroemulsion-stabilizedironoxide

nanoparticlesandanionicsurfactantsodiumdiethylhexylphosphate(SDEHP)-stabilized

nanoscalezerovalentiron(NZVI)particlesinlaboratoryporousmedia.Thetwoformulations

testedinthisworkachievedstableironnanoparticlesuspensionsformonthsandpreparedviaa

simple“one-pot”synthesismethoddevelopedbyWangetal.Bothformulationsweretested

underfieldscalevelocityof5m/daywithnomechanicalaidduringtheinjection.Athree-

compartmentmodel,involvingcolloiddiffusiontheory,diffusiontheoryandtailingwasapplied

todescribethebreakthroughcurvesofthestudies.Theobtainedbreakthroughcurvesofboth

formulationsimpliedexcellenttransportinporousmediawithsteadyplateauC/Coat0.8-0.9

andrecoveryofupto0.95forSDEHPstabilizedNZVI.Postanalysisontheretentionofironon

theporousmediaimpliedidealtransportwithconsistentdatatothebreakthroughcurves.

iii

Acknowledgement

Iwouldliketogivemydeepestandsincerestgratitudetomysupervisor,ProfessorEdgar

Acosta,forhisguidance,mentoringandsupportoverthelasttwoyearsofmystudy.Ihave

neverreceivedmoreinspirationsandencouragementsinlife,careerandacademicallatonce

fromanyoneinmylife.Ihavebecomeabetterthinkerinacademiaandlifewithhisguidance.It

wasanhonourtobehisstudent.

IalsowanttothankProfessorSleep,Dr.MondalandDr.LimafromtheRENEWprogram.Iam

verygratefultobeinthisprogramforbetterresearchandcareerdevelopment.Iwanttothank

ProfessorSleeptoprovidemetheaccesstotheuseofglovebox.Itwasavitalpartofthisstudy.

IwanttothankDr.MondalforgivingmesuggestionsinmyresearchandDr.Limaformaking

myMaster’sexperiencemorerewarding.

Iwanttothankallmycolleaguesandfriends,FrancisChoi,AmericoBoza,SilviaZarate,Aurelio

Stammitti,MehdiNouraei,AshuBhanotandSasanMehrabianforhelpingmewithmyresearch.

IparticularlywanttothankSilviaforhelpingmetostartmyresearch,Americoforprovidinghis

knowledgeincolloidalscienceandFrancisforcollaboratinghisworkwithme.

Iwouldalsoliketothankmyparentsandmybrotherfortheirunconditionalsupport.

Finally,IwanttothankmysoulmateJessicaKokforbeingthereformewheneverIneedher

duringmystudy.

iv

TABLEOFCONTENT

CHAPTER1:INTRODUCTION.................................................................................................................11.2REFERENCE:...........................................................................................................................................8

CHAPTER2:TRANSPORTOFMICROEMULSION-STABILIZEDIRONOXIDEINPOROUSMEDIA..............11ABSTRACT.................................................................................................................................................112.1INTRODUCTION....................................................................................................................................122.2METHODOLOGY...................................................................................................................................16

2.2.1SynthesisofMicroemulsionIronOxide.....................................................................................162.2.2Determiningthestabilityofmicroemulsionironoxide.............................................................172.2.3Viscositystudyofmicroemulsionironoxideformulations........................................................172.2.4Sizecharacterizationofmicroemulsionironoxideandmicroemulsionformulations..............172.2.5Columnstudy............................................................................................................................182.2.6.BreakthroughCurveModeling.................................................................................................222.3.1StabilityTest.............................................................................................................................262.3.2RheologicalProperties..............................................................................................................272.3.3SizeCharacterization................................................................................................................302.3.4μEIronOxideTransport............................................................................................................312.3.5IronDistributionAnalysis..........................................................................................................40

2.4CONCLUSIONS.....................................................................................................................................42

CHAPTER3:DEVELOPMENTANDTRANSPORTOFPHOSPHATESURFACTANT,SDEHP-STABILIZEDNZVIINPOROUSMEDIAFORINSITUREMEDIATION..................................................................................47

3.0ABSTRACT...........................................................................................................................................473.1INTRODUCTION....................................................................................................................................483.2METHODOLOGY...................................................................................................................................53

3.2.1SynthesisofSodiumDiEthylHexylPhosphate(SDEHP)Surfactant...........................................533.2.2CriticalMicelleConcentrationofSDEHPwithdissolvediron....................................................543.2.3TotalOrganicCarbon(TOC)ofirondissolvedSDEHP...............................................................543.2.4SynthesisofSDEHPNZVI...........................................................................................................543.2.5pHAnalysis...............................................................................................................................563.2.6StabilityAnalysis.......................................................................................................................563.2.7ViscosityAnalysis......................................................................................................................573.2.8ColumnExperimentProcedure.................................................................................................573.2.9NZVIColumnDistributionAnalysis...........................................................................................59

3.3RESULTSANDDISCUSSION.....................................................................................................................593.3.1DeterminingtheOptimalSynthesisFormulation......................................................................593.3.2SynthesisResultsandStabilityofFeSO4-basedSDEHPNZVIat100mMand1g/L.................643.3.3pHandViscosityAnalysisandImplication................................................................................683.3.4SizeAnalysisofSDEHPNZVI......................................................................................................693.3.5MobilityofSDEHPNZVIat100mMand1g/L..........................................................................713.3.7ImplicationsforinsituRemediation.........................................................................................75

3.4CONCLUSION.......................................................................................................................................76

CHAPTER4:CONCLUSIONANDRECOMMENDATIONS........................................................................844.2REFERENCES:.......................................................................................................................................89

APPENDIXA–FERRICCHLORIDEBASEDSODIUMDIETHYLHEXYLPHOSPHATE(SDEHP)-STABILIZEDNZVI...................................................................................................................................................90

A.1INTRODUCTION...................................................................................................................................90

v

A.2METHODOLOGY...................................................................................................................................91A.2.1PreparationofSurfactantSDEHPandFeCl3-basedNZVI.........................................................91A.2.2FormulationdesignofFeCl3-basedNZVI..................................................................................91A.2.3CharacterizationAnalysis:SizeandStability............................................................................92

A.3RESULTSANDDISCUSSIONS...................................................................................................................92A.3.1FormulationDesignImplication...............................................................................................92A.3.2StabilityandSizeAnalysis.........................................................................................................93

A.4FUTUREWORKS..................................................................................................................................95A.5REFERENCES:.......................................................................................................................................97

APPENDIXB:COMPARISONBETWEENCARBOXYLMETHYL-CELLUOSESTABILIZEDIRONOXIDENANOPARTICLESWITHMICROEMULSION-STABILIZEDNANOPARTICLES.............................................98

B1.BACKGROUND:....................................................................................................................................98B2.RESULTS:............................................................................................................................................98B2.1.STABILITY.........................................................................................................................................98B2.2.MOBILITYCOMPARISON.....................................................................................................................99

vi

ListofFigures

Figure1.1ReactionschematicsummaryofNZVI,adaptedfromFuetal....................................2Figure1.2A.SchematicofwormlikemicelleandB.theinteractionbetweenwormlikemicelles

byWangetal.andnanoparticles..........................................................................................6Figure2.1Columnexperimentconfiguration.............................................................................19Figure2.2Injectionscheduleofthecolumnstudies...................................................................21Figure2.3Schematicofthethree-compartmentmodelusedtorepresentthereversible

adsorptionofparticle,advection/dispersioncolumntransport,andparticleattachment.22Figure2.4TimelapsephotocomparisonsofμEironoxideandbareironoxide:(A)10g/LμE

ironoxide.(B)5g/LμEironoxide.(C)10g/Lbareironoxidenanoparticles......................27Figure2.5Viscosityprofilegraph(logscaled)ofμEironoxide(a)andME(b),comparison

betweentheoriginalformulationsanddilutionwithNaClbrinesolution(10g/100mL)at1:1ratio...................................................................................................................................29

Figure2.6TEMimagingofMicroemulsionironoxideat5g/Lwith100nmasscale(a)andmicroemulsionNZVIat1g/LbyWangetal.........................................................................31

Figure2.7Transportofironoxidesuspensionsin1-cmdiameter(highaspectratio)columnat5m/dayporevelocity(a)10g/Lironoxide(b)5g/Lironoxide.............................................33

Figure2.8Breakthroughcurvesof5g/L(asFe)μEsuspensionofironoxideinjectedat5m/day(porevelocity)throughcolumnswithaspectratioof15(left)and6(right).Thesolidlinesshowthesolutionofthe3-compartmentmodelusingtheconstantssummarized............34

Figure2.9Breakthroughcurvesobtainedfor5g/LμEironoxideinjectedat5m/day(porevelocity)througha2.5cmx15cmcolumn(aspectratioof6),usingscheduleA(10%NaClconditioning/rinsingfluid)andscheduleB(deionizedwaterconditioning/rinsingfluid.....37

Figure2.10BreakthroughcurvesobtainedfordilutedμEs(noironoxide)injectedat5m/day(porevelocity)througha2.5cmx15cmcolumn(aspectratioof6),usingscheduleA(10%NaClconditioning/rinsingfluid)andscheduleB(deionizedwaterconditioning/rinse).......38

Figure2.11Ironoxidedepositedonsandcolumnaftertheinjectionof1.5PVof5g/L(asFe)ironoxidenanoparticlesat5m/day(porevelocity)througha2.5cmx15cmcolumn(aspectratioof6),usingscheduleA(10%NaClconditioning/rinsingfluid)andscheduleB(deionizedwaterconditioning/rinsingfluid).Thesolidlinerepresentsthepredictionofdepositedironfromthe3-compartmentmodel..................................................................41

Figure3.1StructureofanionicphosphatesurfactantSDEHP....................................................51Figure3.2Illustrationofthe“one-pot”synthesisprocedureofSDEHP-stabilizedNZVI.The

procedurewasconductedintheglovebox..........................................................................56Figure3.3ColumnstudysetupforSDEHP-stabilizedNZVI.........................................................59Figure3.4Surfacetensionmeasurementsof1g/LofNZVIdissolved:Curve1showsthe

surfacetensionmeasurementoftheoriginalSDEHPconcentrationandCurve2displayedthecorrectedconcentrationofSDEHP................................................................................62

Figure3.5DissolvedSDEHPequilibriumconcentrationwithironVS.addedironsulfateconcentrationsfordifferentinitialSDEHPconcentrations..................................................63

vii

Figure3.6TEMimagingof10mMSDEHP-stabilizedNZVIat1g/Lwithdifferentscaleat500nmscale...............................................................................................................................64

Figure3.7SetA,TimelapsephotosofSDEHP-stabilizedNZVIat0.5g/LofNZVIatvariousSDEHPconcentrations:a.30mMofSDEHPb.50mMofSDEHPandc.100mMofSDEHP...............................................................................................................................................65

Figure3.8SetB,Stabilitytimelapsepictureof100mMatNZVIconcentration1,1.5and2g/Loveraperiodof24hours:a.1hourandb.24hoursaftersynthesisandre-suspension....67

Figure3.9ColumnstudybreakthroughcurveofhighlystableSDEHP-stabilizedNZVI,at100mMSDEHPand1g/LofNZVIat5m/daywiththemodeldescribedinchapter2(solidline)...............................................................................................................................................73

FigureA1.FigureA1.SurfacetensionmeasurementsofSDEHPatvariousconcentrationwith

ironchloride(1g/LequivalenceofNZVI)dissolved............................................................93FigureA2.FigureA2.Ferricchloride-basedNZVIat100mMofSDEHP@0.3g/Lofiron

concentration.1houraftersynthesis..................................................................................95FigureB1.A.TimelapsedphotosofCMCandmicroemulsionstabilizedironoxideat2.5g/L.B.

EvidenceofsettlingofCMCironoxideafter80hoursuponsuspension.............................99FigureB2.Comparisonofpressuredropmonitoringresultsatthepost-flushingstagebetween

CMCandmicroemulsionironoxide...................................................................................101FigureB3.Iron-sandgrainanalysiswithmicroscopepicturesforA.Microemulsionironoxideat

2.5g/LandB.CMCironoxideat2.5g/L............................................................................102

viii

LiterofTables

Table2.1Summaryofbreakthroughcurveparameters.............................................................40

Table3.1LiteraturesummaryofcolumnstudiesandstabilitybehaviourfordifferenttypesofsurfacemodifiedironoxideandZVInanoparticles..............................................................52

Table3.2SynthesisresultandcharacterizationofSDEHP-stabilizedNZVIatvariousNZVIandsurfactantconcentrations....................................................................................................67

1

Chapter1:Introduction

Nano-scalezerovalentiron(NZVI)arereactivemetalnanoparticlesartificiallyreducedfrom

Fe2+orFe3+thatholdsahighredoxpotential(E0=−0.44V)duetothezero-valentarounditas

showninfigure1[1].Becauseofthehighredoxpotential,NZVIarecapableofreducingawide

rangeofchemicals,includingchlorinatedorganiccompounds,nitricaromaticcompoundsand

heavymetals[1][2][3],asdemonstratedinfigure1.1.NZVIparticlesalsoholdhighsurfaceareas

duetothenano-scalesizerangingfrom1to250nmthatincreasetherateofreduction

reactions[4][5].Becauseoftheabovefeatures,inthelasttwodecades,NZVIhasbeen

identifiedasapotentialefficientinsitugroundwaterremediationtechnologycomparingto

otherexistingtechnologies[6].ItisexpectedthatwiththedirectcontactbetweentheNZVI

particlesandthesourceofthecontaminantcanactivelyandrapidlyreducethecontaminant

zoneconcentrationandachievedfullremediation[7].Inspecific,inanindustrial-scaleinsitu

NZVIremediation,stabilizedNZVIaretobeinjectedthroughmultipleinjectionwellsvia

differentinjectiontechnologysuchaspressure-pulseorgravityinjection[8][9].Theamountof

NZVIinjectedisdeterminedbasedonthecontaminantconcentrationsfromthe

characterizationofthesitepriortoinjection.Uponinjections,theNZVIistobeleftinthesoil

fortreatingthecontaminantsoveraperiodfromweekstomonths;forbetterresults

recirculationofthegroundwaterisoftenscheduledperiodicallytopromotethemobility.

Duringthisperiod,concentrationoftheironandcontaminantsaremonitoredfrommonitoring

wellsforprogressandhydrauliccontrol.Extractionwellsareinstalledatdownstreamtocollect

transportedNZVIparticles.

2

However,injectingNZVIparticlessuspensionintothesoiltoachieveeffectivegroundwater

remediationisacomplexprocedure;O’Carrolletal.summarizedintothefollowing3steps:(1)

Transportingthereactiveparticlesthroughthesoilmatrix(2)Formingcontacttothe

contaminantzoneand(3)Reactingwiththecontaminantstoachieveremediation[10].NZVI

suspensionholdanundesirablepropertyoffastaggregationandsedimentationduetothehigh

stabilitycontributedbythemagneticattractionforcesbetweentheparticle[11].Thisfeature

failsNZVItoachievethefirststepofconductingtheinsituremediation—thelarger,

aggregatedzero-valentparticleswillexperiencefiltrationinthesoilmatrixandironparticles

willattachtothesoilgrain,keepingtheNZVIparticlesfromreachingthedeepercontaminant

zone[12].Thetransportandmobilityoftheironparticlesintheporousmediaisidentifiedas

themajorobstacleofNZVIinsituremediation.Tothisdate,NZVIinjectionhasremainedasan

state-of-the-arttechnologyandresearchhasbeenactivelydoneonimprovingthetransport.

Figure1.1ReactionschematicsummaryofNZVI,adaptedfromFuetal.[1].

3

LiteraturereviewsuggestedthatimprovingthestabilityoftheNZVIsuspensioncanimprovethe

mobilityofNZVIintheporousmedia[10],[11].Inspecific,astablenanoparticlesuspension

eliminatestheissueofaggregationandsedimentation:ThismeansthatastableNZVI

nanoparticleparticlecanremaininthenano-scalesize,travelbetweensandgrainswithout

filtrationandmaximumamountofNZVIcanbetransported.Currently,themostdirectand

commontechniqueofimprovingthestabilityofNZVIistoapplysurfacemodificationstothe

surfaceofNZVI.Applyingsurfacemodifiers,therepulsionforcesbetweentheironmetal

nanoparticlescanbereducedduetotheadditionofthebarrierandachievehigherstability.

Surfacemodifiersareproventohavepositiveinfluencesonimprovingthestabilityasearlyas

10yearsago[4],[13];however,theresearchonimprovingthestabilityandmobilityisstill

ongoing.Recentstudieshaveshownthatpolymeradsorptionisthemostcommonwayto

stabilizingNZVI,foodgradepolymersuchascarboxyl-methylcellulose[11],[14]–[16],PV3A[17]

andPAA[17].Ontheotherhand,biodegradablesurfaceactiveagentssuchasTween80[18],

SDBS[19]andbiodegradablesurfactants[20]havealsoshownsomeprogressinthisfield.Itis

worthmentioningthatemulsioninducedNZVIhasdrawnalotofattentionasanalternative

wayofstabilizingZVIparticleswithoutsurfaceadsorption[21],[9],[22].However,outofthe

above,theinstabilitywasstillobservedintheabovestudies,forexample,carboxyl-methyl

cellulosebasedNZVIcanremainstableforabout80hourswhileaggregationandsedimentation

areconstantlyobserved[16].Furthermore,Tween80surfactant-basedNZVIalthoughclaimed

remainingstableformonthsinstoragecondition,onceinfieldcondition,thestabilityis

disturbed[18].EmulsioninducedNZVIontheotherhandholdsakineticallystabilityof8hours

whilerequiringmechanicalforceduringinjection[9].Despitescholarshaveinputgreatefforts

4

intostabilizingNZVItoimprovemobility,therehasnotbeenaNZVIsurfacemodifierthat

completelyeliminatesaggregationandsedimentation.

Laboratory1-Dcolumnstudyisusuallythefirststepinevaluatingthemobilityofasurface

modifiedNZVIbeforescalinguptoafieldremediation.ThesurfacemodifiedNZVIsuspensions,

includingtheabovedescribed,weretestedinvarioussimilarbenchscalecolumnsettingsthat

generallyimpliedthreeissues:1.Flowvelocity:mostofthelaboratorycolumnstudieswere

conductedatarelativelyhighflowratefrom8to200m/day[23][17][18],thisisunrealistic,

consideringtypicalfieldapplicationsareconductedat0.25-4m/day[8].2Performance:as

mentioned,instabilitywasstillobservedinallthesurfacemodifiedNZVIthusfar,itwas

constantlyreportedthatthehighestbreakthroughpeakcanreachover0.9athigherlaboratory

flowvelocities[24].However,poorrecoveryandbreakthroughpeakatflowvelocities

approachingtothe4m/day(Tiraferri&Sethi,2009;Xin,Tang,Zheng,Shao,&Kolditz,2016).3.

Mixing:someofthecolumnstudiesintegratedmechanicalmixingintheirsetting[23],field

remediationisoftenconstrainedtointegratesuchfeature.However,carboxyl-methylcellulose-

basedandemulsioninducedNZVIhavesuccessfulacquiredadequateresultsatfieldflow

velocities[16][21].Full-scalefieldstudieswereconductedwiththetwoNZVIsuspensions;

however,unsuccessfulNZVItransportwasreported[11][22].

Theobjectiveofthisthesisistostudythemobilityofmicroemulsion-basedNZVIandtodevelop

anoptimizedsurfactant-basedNZVI.Chapter2and3inthisstudyaretwoscientificarticlesthat

criticallyexaminetwohighlystabilizedNZVIsurfacemodifiersontheirmobilityinporousmedia

andtheirpotentialtoafull-scaleremediation.

5

Inchapter2,themobilityofmicroemulsion-stabilizedNZVIwithastabilityofover6months

developedbyWangetal.isaccessedusingironoxidenanoparticlesasastableanalogytoNZVI.

Highlyconcentratedmicroemulsionironnanoparticleswithidenticalcolloidalpropertiesto

microemulsionNZVIischaracterizedbysize,rheologyandstability.Laboratory1-Dcolumn

studyareconductedatdifferentconditionswithdifferentconcentrations,atlaboratoryand

fieldvelocitiesanddifferentsalinityenvironment.Thebreakthroughresultsaredemonstrated

andmodelledusingColloidFiltrationTheory(CFT).Excellenttransportresultswereobserved;

however,thehighsalinitysensitivityandthepropertyofthesurfactantimplythat

microemulsion-stabilizedNZVIisnotsuitableforafieldtest.

Inchapter3,basedontheimplicationsfromchapter2,anenvironmentalfriendlyphosphate

surfactant,isselectedasasurfactantstabilizer.Aframeworktodeterminethemoststable

nanoparticlessuspensionisadaptedfromWangetal.todeterminethemostoptimized

surfactant-basedNZVI.Itisimportanttonotethat,thusfar,nosophisticatedframeworkhas

beenappliedtodeterminetheformulationforaNZVIstabilizer.Theoptimizedsurfactant

yieldedastabilityofover2monthsandisexaminedwithalaboratory1-Dcolumnstudyatfield

velocity.Itisreportedthatthedevelopedformulationcanrecoverover90%ofNZVIwith

breakthroughpeaksat1.Itisexpectedthatthedevelopedphosphateformulationcanleadto

potentialfieldtestandeventuallyasuccessfulfieldapplication.

Itishypothesizedthattheprolongedstabilityobservedinthemicroemulsion-stabilizediron

nanoparticlesandsurfactant-stabilizedNZVIiscontributedbytheformationofwormlike

micellesinthesystems.Wormlikemicelles,asshowninfigure1.2,arelongandentangled

6

aggregatesofmicellesthatdemonstratestructuresandproperties(suchasviscosity)similarto

polymers[27].Theformationofwormlikemicelleisdependentontheconcentrationofthe

surfactants,surfactantgeometry,curvaturesandpacking.Uponreachingtoacriticalassembly

concentration,themicellewillself-assembleintowormlikemicelles[28][29].Studieshave

suggestedthattheinteractionbetweennanoparticlesandwormlikemicellecanpromotethe

stabilityofthesuspension[27].

A.

B.

Figure1.2A.Schematicofwormlikemicelle,pictureprovidedbyMr.FrancisChoi[29]andB.

7

theinteractionbetweenwormlikemicellesbyWangetal.andnanoparticles[27].

Overall,thisstudycontributedaphosphatesurfactantbasedNZVIsuspensionthatissuitable

forafieldapplicationbasedontheperformanceofmicroemulsion-basedNZVI.Thisstudyalso

suggestedthefeasibilityofmicroemulsionasatransportvehicleforinsituNZVIremediation.

Futurestudyon1.amoredetailedreactivitystudyofthephosphatesurfactantbasedNZVIand

2.TargetdeliverystudywithDNAPLinalargercolumnorsandboxarerecommendedpriorto

thefieldapplication.

8

1.2Reference:[1] F.Fu,D.D.Dionysiou,andH.Liu,“Theuseofzero-valentironforgroundwaterremediationandwastewater

treatment:Areview,”J.Hazard.Mater.,vol.267,pp.194–205,2014.

[2] W.Wang,M.Zhou,Z.Jin,andT.Li,“Reactivitycharacteristicsofpoly(methylmethacrylate)coatednanoscaleiron

particlesfortrichloroethyleneremediation,”J.Hazard.Mater.,vol.173,no.1–3,pp.724–730,2010.

[3] M.Stefaniuk,P.Oleszczuk,andY.S.Ok,“Reviewonnanozerovalentiron(nZVI):Fromsynthesistoenvironmental

applications,”Chem.Eng.J.,vol.287,pp.618–632,2016.

[4] N.Saleh,T.Phenrat,K.Sirk,B.Dufour,J.Ok,T.Sarbu,K.Matyjaszewski,R.D.Tilton,andG.V.Lowry,“Adsorbed

triblockcopolymersdeliverreactiveironnanoparticlestotheoil/waterinterface,”NanoLett.,vol.5,no.12,pp.2489–

2494,2005.

[5] K.Moore,B.Forsberg,D.R.Baer,W.a.Arnold,andR.L.Penn,“Zero-ValentIron:ImpactofAnionsPresentduring

SynthesisonSubsequentNanoparticleReactivity,”J.Environ.Eng.,vol.137,no.10,pp.889–896,2011.

[6] W.ZhangandD.W.Elliott,“Applicationsofironnanoparticlesforgroundwaterremediation,”Remediat.J.,vol.16,no.

2,pp.7–21,2006.

[7] N.C.Mueller,J.Braun,J.Bruns,M.Černík,P.Rissing,D.Rickerby,andB.Nowack,“Applicationofnanoscalezerovalent

iron(NZVI)forgroundwaterremediationinEurope,”Environ.Sci.Pollut.Res.,vol.19,no.2,pp.550–558,2012.

[8] C.M.Kocur,A.I.Chowdhury,N.Sakulchaicharoen,H.K.Boparai,K.P.Weber,P.Sharma,M.M.Krol,L.Austrins,C.

Peace,B.E.Sleep,andD.M.O’Carroll,“CharacterizationofnZVImobilityinafieldscaletest,”Environ.Sci.Technol.,

vol.48,no.5,pp.2862–2869,2014.

[9] S.O’Hara,T.Krug,J.Quinn,C.Clausen,andC.Geiger,“FieldandlaboratoryevaluationofthetreatmentofDNAPL

sourcezonesusingemulsifiedzero-valentiron,”Remediat.J.,vol.16,no.2,pp.35–56,2006.

[10] D.O’Carroll,B.Sleep,M.Krol,H.Boparai,andC.Kocur,“Nanoscalezerovalentironandbimetallicparticlesfor

contaminatedsiteremediation,”Adv.WaterResour.,vol.51,pp.104–122,2013.

[11] A.I.A.Chowdhury,M.M.Krol,C.M.Kocur,H.K.Boparai,K.P.Weber,B.E.Sleep,andD.M.O’Carroll,“NZVIinjection

intovariablysaturatedsoils:Fieldandmodelingstudy,”J.Contam.Hydrol.,vol.183,pp.16–28,2015.

[12] E.M.Hotze,T.Phenrat,andG.VLowry,“Nanoparticleaggregation:challengestounderstandingtransportand

reactivityintheenvironment.,”J.Environ.Qual.,vol.39,pp.1909–1924,2010.

[13] S.Bettina,W.Hydutsky,andL.Bl,“DeliveryVehiclesforZerovalentMetalNanoparticlesinSoila

ndGroundwater,”vol.21,pp.2187–2193,2004.

9

[14] F.He,D.Zhao,J.Liu,andC.B.Roberts,“StabilizationofFe-Pdnanoparticleswithsodiumcarboxymethylcellulosefor

enhancedtransportanddechlorinationoftrichloroethyleneinsoilandgroundwater,”Ind.Eng.Chem.Res.,vol.46,no.

1,pp.29–34,2007.

[15] B.Sleep,P.Mondal,P.Furbacher,Z.Cui,andM.Krol,“Transportofcarboxymethylcellulosestabilizednanoscale

zerovalentironinporousmedia,anexperimentalandmodelingstudy,”vol.17,p.7729,2015.

[16] C.M.Kocur,D.M.O’Carroll,andB.E.Sleep,“ImpactofnZVIstabilityonmobilityinporousmedia,”J.Contam.Hydrol.,

vol.145,pp.17–25,2013.

[17] P.Jiemvarangkul,W.X.Zhang,andH.L.Lien,“Enhancedtransportofpolyelectrolytestabilizednanoscalezero-valent

iron(nZVI)inporousmedia,”Chem.Eng.J.,vol.170,no.2–3,pp.482–491,2011.

[18] J.Soukupova,R.Zboril,I.Medrik,J.Filip,K.Safarova,R.Ledl,M.Mashlan,J.Nosek,andM.Cernik,“Highly

concentrated,reactiveandstabledispersionofzero-valentironnanoparticles:Directsurfaceandsiteapplication,”

Chem.Eng.J.,vol.262,pp.813–822,2015.

[19] N.Saleh,K.Sirk,Y.Liu,T.Phenrat,B.Dufour,K.Matyjaszewski,R.D.Tilton,andG.V.Lowry,“SurfaceModifications

EnhanceNanoironTransportandNAPLTargetinginSaturatedPorousMedia,”Environ.Eng.Sci.,vol.24,no.1,pp.45–

57,2007.

[20] Y.T.Wei,S.cheeWu,S.W.Yang,C.H.Che,H.L.Lien,andD.H.Huang,“Biodegradablesurfactantstabilizednanoscale

zero-valentironforinsitutreatmentofvinylchlorideand1,2-dichloroethane,”J.Hazard.Mater.,vol.211–212,pp.

373–380,2012.

[21] N.D.BergeandC.A.Ramsburg,“Oil-in-wateremulsionsforencapsulateddeliveryofreactiveironparticles,”Environ.

Sci.Technol.,vol.43,no.13,pp.5060–5066,2009.

[22] J.Quinn,C.Geiger,C.Clausen,K.Brooks,C.Coon,S.O’Hara,T.Krug,D.Major,W.S.Yoon,A.Gavaskar,andT.

Holdsworth,“FielddemonstrationofDNAPLdehalogenationusingemulsifiedzero-valentiron,”Environ.Sci.Technol.,

vol.39,no.5,pp.1309–1318,2005.

[23] H.I.Gomes,C.Dias-Ferreira,A.B.Ribeiro,andS.Pamukcu,“Enhancedtransportandtransformationofzerovalent

nanoironinclayusingdirectelectriccurrent,”Water.Air.SoilPollut.,vol.224,no.12,pp.1–12,2013.

[24] C.Mystrioti,N.Papassiopi,A.Xenidis,D.Dermatas,andM.Chrysochoou,“Columnstudyfortheevaluationofthe

transportpropertiesofpolyphenol-coatednanoiron,”J.Hazard.Mater.,vol.281,pp.64–69,2015.

[25] J.Xin,F.Tang,X.Zheng,H.Shao,andO.Kolditz,“Transportandretentionofxanthangum-stabilizedmicroscalezero-

valentironparticlesinsaturatedporousmedia,”WaterRes.,vol.88,pp.199–206,2016.

10

[26] A.TiraferriandR.Sethi,“Enhancedtransportofzerovalentironnanoparticlesinsaturatedporousmediabyguargum,”

J.NanoparticleRes.,vol.11,no.3,pp.635–645,2009.

[27] M.E.Helgeson,T.K.Hodgdon,E.W.Kaler,N.J.Wagner,M.Vethamuthu,andK.P.Ananthapadmanabhan,“Formation

andrheologyofviscoelastic‘doublenetworks’inwormlikemicelle-nanoparticlemixtures,”Langmuir,vol.26,no.11,

pp.8049–8060,2010.

[28] A.Sambasivam,A.V.Sangwai,andR.Sureshkumar,“Self-AssemblyofNanoparticle-SurfactantComplexeswithRodlike

Micelles:AMolecularDynamicsStudy,”Langmuir,vol.32,no.5,pp.1214–1219,2016.

[29] X.Wang,D.S.Miller,E.Bukusoglu,J.J.DePablo,andN.L.Abbott,“Topologicaldefectsinliquidcrystalsastemplates

forMolecularSelf-Assembly,”Nat.Mater.,vol.15,no.September,pp.1–9,2015.

11

Chapter2:TransportofMicroemulsion-StabilizedIronOxideinPorous

Media

Abstract

ThefullpotentialofNZVItechnologiesisoftenlimitedbythetransportNZVIparticlesthrough

porousmedia,whichinturnislimitedbythecolloidalstabilityoftheNZVIsuspension.Previous

workhasshownthatstableNZVIsuspensionscanbeproducedusingmicroemulsions(μEs)as

synthesis and suspensionmedia. In thiswork, Ironoxide nanoparticles, used as non-reactive

analogstoNZVI,wereusedtoevaluatethetransportintheμEusedtosynthesizeandsuspend

NZVI.Thetransportofthesesystemswasexaminedatfullstrength(10g/LFe)andasadiluted

(5g/L Fe) suspension using column studies. The nanoparticle injection protocol was also

evaluated(watervs.brineconditioning/rinsingfluid).Theresultingbreakthroughcurveswere

analyzed via a 3-compartment transport model that accounts for reversible and irreversible

attachmenttothesandpackedinthecolumn.Itwasdeterminedthatlargepressuredropswere

observedwithconcentratedsuspensions(10g/LFe),whichisexplainedbythelargeviscosityof

thesesystems.Thedilutedsuspensions(5g/LFe),havingalowerviscosity,couldbeinjectedin

thesystem,producinghighparticlerecoveries(~90%)whenthesolutionusedtoconditionand

rinsethecolumnwasabrinewiththesamesaltconcentrationastheμE.However,whenusing

deionizedwatertoconditionandrinsethecolumn, lowerrecoveries (~60%)wereobtained,

likelyduetophasetransitionsintheμEthatresultedinthedepositionofparticles.

12

2.1Introduction

Nanoscalezero-valentiron(NZVI)particlesareeffectivereducingagentsforavarietyof

contaminants,includingheavymetals,dyes,chlorinatedorganiccompounds,aromaticand

arsenic-containingcompounds[1][2].ThenanoscalesizeofZVIparticles,rangingfrom10to

200nm,yieldshighsurfaceareasandthusincreasedcontactwiththecontaminantsmakingthe

reductionreactionsmoreefficient[3].NZVIcaneffectivelyreducecommoncontaminantssuch

astrichloethylene(TCE)toethylenewithindaysorweeks[1][4][5][6].NZVIinjectionhasshown

promiseintreatingcontaminantsinnumerousfield-scalestudiesconductedinEuropeand

NorthAmericainthelastdecade[7][8][9].

NZVIaquiferremediationisstillhinderedbythelimitedcolloidalstabilityofsuspensions

currentlyinuse,resultinginrapidaggregationandsettling[8][9][10][11][12].Therapid

aggregationandfastsettlingfeatureofNZVIiscausedbythehighsurfacemagneticpotential

[4][13].Duringthetransport,aggregatedNZVIaremorelikelytoexperiencefiltrationbythe

sandporesduetothelargesize.Settlingofaggregatedparticlesalsocontributetothe

depositionofNZVIontheporousmedia[13].Schricketal.andTiraferietal.reportedthatpoor

transportperformanceofunstableNZVIareobservedinlaboratorycolumnstudiesat

consideratelylowconcentrations(0.05-0.1g/L)[10][14].

TheintroductionofthesurfacemodificationscanreducetheaggregationofNZVIparticlesand

thuscontributetoimproveNZVItransportinporousmedia[10][15][16].Inspecific,surface

modificationsareservedassurfacestabilizers,theycreatedanenergybarrierintheNZVI

suspensiontokeeptheparticlesfromattractingandaggregating.Commonsurfacemodifiers

13

includeemulsions,anionicsurfactants,polymers,organicacidsandpolymericsurfactants[4]

[7][16][12][17][18][19][20].

Laboratorystudiesandfieldtrialshavereportedsomeprogressinimprovingthemobilityof

NZVIinsoil;however,theaggregationofNZVIhasnotbeenfullyaddressed[17][19][21][22].

MostofthestabilizedNZVIcolumnstudies,suchasthoseconductedusingxantumgum-

stabilizedNZVIandpolyphenol-stabilizedNZVIreportedhighNZVIrecovery,over85%,atpore

velocitiesbetween20-200m/daywhileonly10%tonegligiblerecoveryobservedatvelocities

below10m/day[22][23][17][24][25].O’Carrolletal.andKocuretal.indicatedthattypical

injectionvelocitiesforafield-scalein-situremediationrangebetween0.25and4m/day

[20][19][18].Thehighrecoveryreportedbymosttheliteraturestudiesareanoptimistic

estimationoftheabilityofthesurfacestabilizersapplied.Stabilizersthatallowadequate

transportofNZVIatlowporevelocitiesarestillneeded.

ColumnstudiesconductedbyBergeetal.andKocuretal.withemulsionNZVIand

carboxylmethyl-cellulose(CMC)NZVI,respectively,reportedNZVIrecoveriesof90%for

injectionvelocitiesrelevanttofieldapplications[19][26].ItwasreportedthatemulsionNZVIis

kineticallystableandCMCNZVIholdsarelativelyhighstabilityof80hours[7][19].However,

thefieldresultsintheCMCNZVIimpliedthatastabilityof80hoursmaynotbesufficient.In

2014,Kocuretalconductedafield-scaleremediationwithCMCNZVIatasiteinSarnia[2].CMC

stabilizedNZVIat1g/Lironconcentrationwasinjectedatgroundwaterflowvelocitiesof0.02-

0.8m/dayusinggravityinjection.After10daysofcontactingperiod,CMCNZVIwasobservedat

monitoringwells1meterdownstream,implyingthattheminimumtransportdistanceofCMC

NZVIisatleast1meter.However,therecoveryoftheinjectedNZVIwasonly1%[27].Similar

14

findingswerereportedintheemulsionNZVIfieldstudy[10].Additionally,fielddemonstrations

ofNZVIremediationreportedshorttraveldistancesbetween0.5to2.4meters[19][28].

Furthermore,someofthereportedNZVIcolumnstudiesrequiredcontinuousmechanical

mixingtopreventNZVIfromsettlingintheinjectionpoint[17].Mechanicalmixingmightbe

impracticaltoscaleup.TheseobservationsshowthatdespitetheimprovedstabilityofNZVI

suspensionbystabilizers,thedelayedaggregationmechanismwasnotfullyeliminated.

O’Carrolletal.suggestedthatanidealNZVIformulationshould:(1)maintainstabilitysuchthat

theparticlesdonotaggregateandsettle;(2)musthaveahighconcentrationofironwhile

maintainingasmallsize;(3)sustaincertainamountofmobilitywhilebeinginjectedinsoil[19].

ThemostsuccessfulsystemsthusfarareCMCNZVIandbimetallicNZVI[27][28].Despitegood

remediationresults,allthefieldscalestudiesofNZVIreportedthatlittletonoNZVIparticles

wererecovereduponcompletion,implyingimmobilizationandemphasizingtheneedforbetter

transportoftheparticles.

Wangetal.introducedtheuseofmicroemulsions(μEs)asbothsynthesissolventand

suspendingmediainaone-potsynthesisprocedure[29].μEsarethermodynamicallystable

systemscontainingoiland/orwaternano-domains(typicallyof10to100nm)thatare

stabilizedbysurfactant(s)adsorbedattheoil-waterinterface.TheformulationoftheμE-based

synthesis/suspensionmediawasdesignedviatheHydrophilic-Lipophilic-Difference(HLD)

framework,usedbyChoietal.andWangetal.todeterminethecombinationofsurfactant,

electrolyte,oil,andtemperaturethatproducesbicontinuousnet-zerocurvaturesystems

(whereHLD=0),whichalsoleadstotheformationofsuspensionsthatarestableforseveral

months,andareeasilyre-suspendedwithmildmixing[16][29][30].

15

DespitetheadvancesmadewithμEsynthesis/suspensionmediaforNZVI,nostudieshavebeen

conductedtovarytheirtransportthroughporousmedia.Microemulsionsontheirownhave

beenappliedaspartofthesurfactantenhancedrecoveryremediation[31].Thesurfactant

enhancedaquiferremediation(SEAR)hasbeensuccessfullyappliedfortheremediationof

mediumtolowdensitynon-aqueousphaseliquids(LNAPLs),butdownwardmobilizationofhigh

densityplumesofchlorinatedsolvents(DNAPLs)havelimitedtheiruseforthoseapplications

[32].

Inprinciple,μE-NZVImeetsthestandardsrequiredforanidealstabilizer,buttheireffectiveness

havenotbeenevaluatedviacolumnstudies.Thus,itistheinterestofthisworktoexaminethe

transportpropertiesofμE-NZVIinone-dimensional(1-D)columnstudiesandassessits

potentialusefulnessinfieldapplications.

Forthepurposeofanalyzingtheintrinsicabilityofmicroemulsionasatransportvehicle,iron

oxideisusedinthisresearchasananalogytoNZVI[13][16][33][34].Microemulsionironoxide

developedbyChoietal.hasidenticalcolloidalstructureandsizetoμE-NZVI[30].Inadditional,

ironoxideisusedasasynthesisbasistoproduceNZVIinsomecasesandcanbecategorized

withNZVIintermsofenvironmentalapplication[34][35].Transportofironoxideinporous

mediaisalsointerestofenvironmentalandbiomedicalresearch.

TheobjectiveofthisworkistodetermineandassessthepotentialandabilityofμEsasastabilizer

and delivery vehicle for NZVI in porousmedia using iron oxide nanoparticles as analog. The

studieswereconductedusingfield-relevantconditionsthatmosttheothercolumnstudiesdon’t

typically consider: high concentration of suspended iron and a low Darcy velocity. Several

preliminary and post analytical strategies including size determination using dynamic light

16

scattering(DLS)andTEM,viscositytestandparticlecolumndistributionareconductedtoprovide

a better understanding on the filtration mechanism between the porous media and the

nanoparticles.Colloidfiltrationtheoryisusedtodescribeandestimatethetransportefficiency

anddistanceofthemicroemulsionironoxide.Theresultofthispaperimpliedandcontributed

thepotentialofanewandefficientsurfacemodificationtechniqueforNZVIinsituremediation.

2.2Methodology

2.2.1SynthesisofMicroemulsionIronOxide

Microemulsionironoxidesuspensionwassynthesizedbybatchfromtheproceduredeveloped

byChoietal.[30]:0.103gramsironoxidenanoparticles(purchasedfromSigmaAldrich,

98%,No.637106),0.327gramsfoodgradeoilethylcaprate(purchasedfromSigma

Aldrich,98%,No.W243205),2.976mLsurfactantAlfoteraK3-4SC10H21O(CH3CH2(CH3)O)4SO4Na

(donatedbySasolNorthAmerica,32.5wt%,lotno.4130115),3.23mLNaCl(Bioshop,Reagent

Grade,SOD002.205)brinesolution(30g/100ml)and3.47mLwater.Thechemicalswere

addedbyweightusinganelectronicscale(DenverInstrumentXX7020023AnalyticalBalance

LabScale±0.0001g)andbyvolumeviapipettingwithanautomaticpipet(FisherBrand,Elite,

AdjustableVolumePipetter,0.5-5mL).Theformulationmixturewasmixedusingavortexer

(VWR,minivorterxer,WM-3000)at5,000rpmfor1minuteandsonicateusingasonicatorbath

(Cole-Parmer,8891)for1minute.Themixingtagewastoensurefullandevensuspensionof

thenanoparticles.Thecompletedsuspensionyieldedabrowncolouruniformly.Eachbatchof

theironoxidesuspensionprovides10mlof10g/Lironoxide(equivalentto7g/LFe)

suspensionwithasalinityof10gNaCl/100ml(10%NaClbrine).FormakingμEironoxide

suspensionwithlowerconcentrationat5g/Lthe10g/Lformulationwasdilutedwith10%NaCl

17

brinesolution(10g/100mL).Keepingthesamesalinityensuredthattheformulationretained

itsstructure,thuspreventinganyphasebehaviorchange.Itisimportanttonotethatforthe

purposeoftrackingthetransportofμE,solventbluedye(Sigma-aldrich,98%,17354-14-2)was

dissolvedinethylcaprateataconcentrationof5000ppm.

2.2.2Determiningthestabilityofmicroemulsionironoxide.

Todeterminethestabilityofmicroemulsionironoxide,timelapsephotosoftheironoxide

formulationsandbareironoxideweretakenoveraperiodof1year.Thetimelapsephotos

weretakendailyinthefirstmonth,bi-weeklylaterandmonthlyfortheremainingperiod.The

collectedphotosofthesampleswereanalyzedvisuallyforsignsofaggregation,settlingand

instability.

2.2.3Viscositystudyofmicroemulsionironoxideformulations

TodeterminetheviscosityoftheμEsandtheironoxide–loadedμEsarheometer(TA

instrument,CSL2500)wasused.Theviscositiesweremeasuredatdynamicshearrates

increasingfrom3-500(1/S).

2.2.4Sizecharacterizationofmicroemulsionironoxideandmicroemulsionformulations

Transmissionelectronmicroscopy(TEM,HitachiHF-3300)wasusedtoassessthesizeandstate

ofaggregrationoftheironnanoparticlesin5g/Lsuspensions.ItisimportanttonotethatTEM

imagingwasnotabletocapturetheμEcomponentsofthesuspension,onlytheiron

nanoparticles.Dynamiclightscattering(DLS)measurementswereconductedonμEironoxide

at5and10g/LusingaBrookhavenParticleSizeAnalyzer90Plus.Priortosizeanalysis,the

sampleswerediluted50timeswith10%NaClbrine.

18

2.2.5Columnstudy.

Twosetsofcolumntransportexperimentswereconductedwithsimilarproceduresand

identicalsetupforanalyzingthetransportatdifferentflowconditions.Thesetupofthecolumn

experimentsisshowninFigure2.1.Inshort,theconditioning/rinsingfluidandμEironoxide

wereinjectedwithaperistaticpump(ColeParmer,MasterFlexL/S)inanupwarddirectionand

wereswitched,asneeded,viaathree-way-valve.Theflowpassedthroughapressuregaugefor

pressuredropmonitoringbeforeenteringthesandcolumn.Theeffluentfromthecolumnwas

collectedbyafractioncollector(RediFrac,BioscienceAmersham).

Thefirstsetsofcolumnexperimentswereconductedwithaglasscolumn(1.0x15cm,Kontes

ChromaflexColumns,KimbleChase,Vineland,NJ.)havinganaspectratioof15andthesecond

setofcolumnexperiments,wereconductedwithalargerglasscolumn(2.5x15cm,Kontes

ChromaflexColumns,KimbleChase,Vineland,NJ.)withidenticallengthbutbiggerinner

diameter,thusaloweraspectratioof6.Bothcolumnswerewet-packedhomogeneouslywith

acidwashedOttawasand(FisherScientific,~500micrometersinradius)and1wt%Alfoterra

K3-4Ssurfactantsolutionusedtoremoveairpocketstrappedinthesandmedia.Theweightof

thecolumn,sandandamountofsolutionweremeasuredwithanelectronicbalance(Denver

Instrument,TP-214)beforeandafterthepackingprocess.Themeasuredweightswereusedin

massbalancetodeterminetheporevolume.Itwasdeterminedthatsmallandlargecolumns

have4.7and30.2mLporevolume,respectively.

19

Figure2.1Columnexperimentconfiguration

Thefirstsetsofcolumnexperimentswereconductedwithaglasscolumn(1.0x15cm,Kontes

ChromaflexColumns,KimbleChase,Vineland,NJ.)equivalenttoanaspectratioof15andthe

secondsetofcolumnexperiments,wereconductedwithalargerglasscolumn(2.5x15cm,

KontesChromaflexColumns,KimbleChase,Vineland,NJ.)withidenticallengthbutbiggerinner

diameter,thusaloweraspectratioof6.Bothcolumnswerewet-packedhomogeneouslywith

acidwashedOttawasand(FisherScientific,~500micrometersinradius)with1wt%K3-4S

surfactantsolutionforeliminatingairpocketstrappedinthesandmedia.Theweightofthe

column,sandandamountofsolutionweremeasuredwithanelectronicbalance(Denver

Instrument,TP-214)beforeandafterthepackingprocess.Theweightsoftheabovewereused

inmassbalancetodeterminetheporevolumes.Itisdeterminedthatsmallandlargecolumn

holda4.7and30.2mLofporevolumes,respectively.

Forexperimentsconductedwiththesmallcolumn,10porevolumesofconditioningsolution

werepumpedfromthebottomofthecolumnbyaperistaticpump,thisisknownasthepre-

flushingorconditioningstage.Theperistaticpumpwasoperatingat0.5mL/minequivalenttoa

20

Darcyvelocityof20m/dayandforalowervelocityasyringepump(notshowninFigure2.1)

wasusedtoproduceaDarcyvelocityof5m/day.Duringtheconditioningstage,thepressure

dropwasrecordedeveryporevolume.Brine(10%NaCl)solutionwasusedastheconditioning

fluidtokeepthesalinityconsistentwiththeμEironoxideformulation.Aftertheconditioning

stage,thepressuregaugewasbypassedtoavoidμEentrainmentinthepressuregaugeline.A

totalof1.5porevolumesofμEironoxideatconcentrationsof5or10g/L(asironoxide)were

injectedtothecolumn.Thefractioncollectorwasthenstartedandsettocollect3minutesof

flowpersample(i.e.1.5mL/sample)forthehigherDarcyvelocityand5minutes/sample

(0.5mL/sample)forthelowerDarcyvelocity.Uponthecompletionoftheironoxideinjection,

another10porevolumesofthesamesolutionusedduringtheconditioningstagewerethen

introducedintothecolumnasarinsingstep.Thecollectionofsampleswasstoppedattheend

oftherinsingstage.

ThesamplescollectedwereanalyzedviaaciddigestionwithHCl6N(BDH,BDH7204-1)witha

sampletoHClvolumeratioof1:14.5,aspertheprocedureofRadetal.[1].Theaciddigested

ironoxidesampleswereanalyzedunderUV-VISspectrometer(80-2092-26,LKBBiochrom

England)at398nmafterareactionperiodof3days.Acalibrationcurvewascreatedusingthe

μEironoxidesuspensioninjectedintothecolumn.

Forthelargecolumn(aspectratioof6),experimentswereconductedat5m/day.Inthese

experiments,twoconditioning/rinsingscheduleswereevaluatedassummarizedinFigure2.2.

ScheduleA,inFigure2.2,followsthesameconditioning/rinsingstepsusedwiththesmall

columnwhere10%NaClwasusedastheconditioning/rinsingsolution.InScheduleBstudies,

deionizedwaterwasusedastheconditioning/rinsingsolvent.ScheduleAisbeneficialbecause

21

itmaintainstheionicstrengthofthemicroemulsion,reducingthechangesforphasechanges.

ScheduleBrunstheriskofμEphasechanges,butitslowionicstrengthismoreconsistentwith

thatofgroundwater.Uponcompletionoftheconditioningstage,1.5porevolumesofμEiron

oxidesuspensionscontaining5and10g/L(asironoxide)wereinjectedintothecolumn.The

fractioncollectorwassetto3min/sample(1.5ml/sample).Fractionalcollectorsettingat3

min/sample(1.5ml/sample)wasusedtocollectthesample.Thecollectedsampleswerethen

analyzedforironcontentusingtheaciddigestionprocedurepreviouslydescribed.Afterthe

rinsingstep,thesandinthecolumnwascollectedanddividedintofivesegments,eachbeing

approximately3cminlengthtodeterminetheresidualirondistributionleftonthecolumn.The

irondistributionanalysiscombinedmicroscopeimagingandaciddigestionsofsandataweight

ratioof0.3gofsandto3mLofHCl6Nacid.Finally,selectedeffluentsamples,correspondingto

thepeakofthebreakthroughcurve,wereanalyzedunderDLStoassesspotentialparticle

aggregation.

Figure2.2Injectionscheduleofthecolumnstudies.

22

2.2.6.BreakthroughCurveModeling.

Aswillbediscussedlater,thebreakthroughcurvesobtainedinthisworkhavefeatures

characteristicofthreedifferentphenomena;diffusion/dispersion,reversible

“chromatographic”adsorption,andtheirreversible“attachment”orfiltrationofparticles.To

representthesefeatures,Figure2.3presentsa3-compartmenttransportmodel.

Figure2.3Schematicofthethree-compartmentmodelusedtorepresentthereversible

adsorptionofparticle,advection/dispersioncolumntransport,andparticleattachment.

Thecentralcompartmentcorrespondstothetransportoftheparticlesthroughthecolumn,

withmassbalanceequation:

Ci,t-1Ci-1,t-1

Crev i,t-1Crev i-1,t-1

Catti,t-1Catti-1,t-1

Ci+1,t-1

Crev i+1,t-1

Catti+1,t-1

Flow

Ci,tCi-1,t

Crev i,tCrev i-1,t

Catti,tCatti-1,t

Ci+1,t

Crev i+1,t

Catti+1,t

Flow

Time“t-1”

Time“t”

23

!"!#= −𝑣 !"

!(+ 𝐷 !+"

!(+− 𝑘-##𝑐 − 𝑓012

!"345!#

(1)

Where“C”istheconcentrationoftheparticleatagiventime“t”andasectionofcolumn“i”,

andwouldcorrespondtothevariableCi,tintheschematicofFigure2.3.Theterm“v”isthe

pore(Darcy)velocity,“D”istheeffectivediffusivityoftheparticlesinthecolumn.Itshouldbe

clarifiedthatthisdiffusivityincludesback-mixingordispersioneffectsinthecolumn,beyond

theintrinsicdiffusivityoftheparticles.Thevariable“z”isthecolumnlengthaxis,represented

bythecolumnlocationindex“i”inFigure2.3.Theparameterkattisusedtorepresentthe

particleattachmentprocessasafirstorderirreversiblereaction,inasimilarwaythatthe

colloidfiltrationtheorydoes[36][37].

Thereversibleadsorptioncompartmentisusedtoaccountforthesametypeofreversible

exchangethattakesplaceinchromatographicseparation[38][39].Inthecolumnstudies,this

reversibleexchangeresultsin“tailing”effectsinthebreakthroughcurvethatcannotbe

simulatedwiththediffusivity(dispersion)term.Asitwillbeshownintheresultsection,such

tailingeffectsareobservedinourresultsandotherresultspresentedintheliterature.To

considerreversibleadsorption,Figure2.3presentsamathematicalconstructionofa

compartmentwithanequivalentvolume“Vrev”wheretheconcentrationoftheparticlesis

“Crev”.Theterm“frev”istheratiobetweenthevolumeofthereversibleadsorption

compartmentandthevolumeofthecolumn(frev=Vrev/V).Themassbalanceoftheparticlesin

thereversiblecompartmentis:

!"345!#

= 𝐾 78 012

𝑐 − 𝑐012 = 𝑘012 𝑐 − 𝑐012 (2)

24

whereKcanbeinterpretedasthemasstransfercoefficientinbetweenthecolumnandthe

reversiblecompartment,andA/Vcanbeinterpretedasthesurfaceareatovolumeratioforthe

transportintothereversiblecompartment.TheoveralltermK*A/Vresultsinafirstorder

constant,krev.ItmustbeclarifiedthatthesimplemasstransportexpressionofEquation2

impliesalinearadsorptionbehavior.Morecomplexadsorptionbehavior,suchasLangmuir

adsorptioncouldbeused,butaswillbeshownlater,thesimplemodelofEquation2was

enoughtoreproducethetailfeaturesofthebreakthroughcurves.

Finally,themassbalancefortheirreversiblyadsorbed(attached)particlecompartmentis:

!9:;;!#

= 𝑉-##!":;;!#

= 𝑉𝑘-##𝑐 (3)

Asinthecaseofthereversibleadsorption,theirreversibleparticleattachmentcompartmentis

amathematicalsimplificationofanequivalentcompartmentofvolume“Vatt”,aspresentedin

Figure2.3.Theuseofthereversibleadsorptionandattachmentcompartmentssimplifiesthe

numericalsolutionofthemassbalancesandavoidsintroducingpartitioncoefficientsor

adsorptionisothermsthatwouldintroducemorefittingparameters.

Tosolvethedifferentialequations,finitedifferencesintime(t)andspace(z)were

implemented,asillustratedinFigure2.3.Theinitialtime-baseconditionwasallthe

concentrationsinthethreecompartmentsbeingzeroattimezero.Theinitialspace-base

conditionwasintroducedinawaythatitrepresentedtheinjectionprotocol,inotherwords,C0,t

25

=Co(feedconcentration)foraslongastheinjectionoccurred,andzerootherwise.Thenon-

dimensionalfinitedifferenceformsofthebalanceequationsare:

𝐶012#,?∗ = 𝐶012#AB,?∗ + 𝑘012∗ 𝐶#AB,?∗ − 𝐶012#AB,?∗ 𝛥𝑡∗ (4)

𝐶#,?∗ = 𝐶#AB,?∗ − 𝐶#AB,?∗ − 𝐶#AB,?AB∗ ∆#∗

∆(∗+ 𝐷∗ ";FG,HIG

∗ A";FG,H∗

∆(∗− ";FG,H

∗ A";FG,HFG∗

∆(∗∆#∗

∆(∗− 𝑘-##∗ 𝐶#AB,?∗ ∆𝑡∗ − 𝑓012 𝐶012#,?∗ − 𝐶012#AB,?∗ (5)

𝑚-###,?∗ = 𝑚-###AB,?

∗ + 𝑘-##∗ 𝐶#,?∗ ∆𝑡∗ (6)

wherealltheconcentrationtermsarenormalizedbytheinitialconcentration(C*=C/Co),the

subindex“t”representthesolutionatagiventime,and“t-1”representsthesolutioninthe

previoustimestep.ThedimensionlessintervaloftimeisΔt*=Δt/τ,where“τ”istheresidence

timeinthecolumn.Thedimensionlessreversibleadsorptionrateconstantisk*rev=krev·τ.The

dimensionlessintervalofspaceisΔz*=Δz/L,where“L”isthelengthofthecolumn.The

dimensionlessdiffusioncoefficientisD*=D·τ/L2.Thedimensionlessattachmentrateconstantis

k*att=katt·τ.Thedimensionlessattachedmassismatt*=matt/(V·Co).

ThesefinitedifferenceequationsweresolvedinExcel,usinganspatialstep,Δz*=Δz/L=0.01,

andantemporalstepΔt*=Δt/τ=0.005.

Theattachmentconstant,katt,canbeusedincolloidfiltrationtheorytoestimatethedistance

towhich1%oftheinitialparticles(Lmax)arestillpresentinthefluid,usingtheequation[2]:

𝐿9-L = − 2M:;;

ln(0.01) (7)

Usingthesingle-collectorcontactefficiency(ηo)correlationdevelopedbyTufenkjietal.[37],

onecanassesstheattachmentefficiency(α),usefultocomparetoothercolumn

26

studies[10][19][20][21][40]:

𝑘-## =U(BAV)WXYZ

𝜂\𝛼𝑣 (8)

Inequation8,𝜖representstheporosityofthesystemand𝑑`arepresentstheaveragegrainsize.

Consideringv=4m/day,d50=0.5mm,𝜂\~0.0005fromthecorrelationofTufenkjietalusinga

Hamakerconstantof1E-19J,a270nmparticleandafluidwith250cPviscosity[3].Aswillbe

shownlater,thesearetheconditionsthatapplytothecolumnstudiescarriedoutwithdiluted

μEsystemscontaining5g/Lironoxidenanoparticles.

2.3ResultsandDiscussion

2.3.1StabilityTest.

Figure2.4showsthetimelapsephotosofbareironoxide,μEironoxideat10g/Land5g/L

(Dilutionratio1:1withNaClbrinesolution(10g/100ml))overaperiodof1year.Asexpected,

thebareironoxidesuspensiondisplayedcolloidalinstability,settlingsoonaftermixingasseen

inFigure2.4(c).Bareironoxidenanoparticlessettledwithin10minutes.Incontrast,theμE-

stabilizedironoxideformulationsattheoriginalconcentrationof10g/Landatthediluted

concentrationof5g/Lremainedasasinglephaseafter1year.Visualinspectionindicatedno

signsofaggregationandsettlingbehaviour,asshowninFigure2.4.Theprolongedstability

demonstratedbytheμEironoxideisconsistentwithμENZVIasreportedbyWangetal.[29].

Thesimilarityinthehighstabilitybetweenthetwosuspensionsemphasizesthesimilarity

betweenthetwotypesofsuspensions.Thecompellingstabilitydemonstratedbythe

microemulsionironoxideformulationmatchedtherequirementsproposedbyO’Carroll,as

mentionedpreviously,inferringtraitsofgoodNZVI/ironmobilityinsoil.

27

Choietal.proposedthatinsystemsthattendtoformbicontinuousμEs(thecaseforthe

formulationconditionsusedhere,andthoseofWangetal.),butwhoseextremelylowoil/water

ratiodoesnotallowtheformulationofbicontinuousstructures,worm-likemicellesareformed

instead,producinganetworkcapableofsuspendingnanoparticles[30].Theobservedstability

observedwithμEironoxidenanoparticlesaresubstantiallylongerthanonethemostsuccessful

NZVIsurfacestabilizations(about80hourswithCMC)[19][18][34].

Figure2.4TimelapsephotocomparisonsofμEironoxideandbareironoxide:(A)10g/LμE

ironoxide.(B)5g/LμEironoxide.(C)10g/Lbareironoxidenanoparticles

2.3.2RheologicalProperties.

Figure2.5showstheviscosityvs.shearratefortheoriginalμEformulationandthe50%diluted

formulation,with(topFigure)andwithout(bottomFigure)ironnanoparticles.Fromthegraphs

inFigure2.5,theviscositydecreaseswithanincreaseinshearrate,confirmingashear-thinning,

non-NewtonianbehaviourforbothμEandμEironoxide.TointerprettheresultsofFigure2.5,

oneneedstoconsiderthattheshearrate(γ̊~v/d50)foracolumnoperatingataporevelocity

ofv=5m/dayandparticlesized50=0.5mm,thentheshearrateisintheorderof0.1s-1.The

28

viscositiespresentedinFigure2.5areforhighershearrates,butextrapolatingthepowerlaw

functionstotheexpectedshearrate,γ̊~0.1s-1,thentheviscosityforthefullstrengthiron

oxide(10g/LFe2O3)μEsuspensionwouldbecloseto1200cP,andforthedilutedsuspension

(5g/LFe2O3)wouldbecloserto250cP.FortheμEalone(withoutironoxide),theviscosity

seemtoplateaucloseto200cPatlowshearrates.TheviscosityforthedilutedμE(withoutiron

oxide)atashearrateof0.1s-1isexpectedtobearound30cP,andthelowvalueofthe

exponentofthepowerfunctionsuggeststhatthebehaviorofthisparticularsystemiscloserto

thatofaNewtonianfluid.AtleastforthedilutedμEsystems,isclearthattheadditionof

nanoparticlesproducedanincreaseintheviscosityofthesystem.

TheviscosityofthesurfacemodifiednZVIsuspensionsisidentifiedasacriticalvariablethat

mayinfluencethetransportintheporousmediaandmobilizationofDNAPL[2][8][19][27].The

trendsinchangesinviscositywithchangesinconcentrationandparticleadditionobservedin

thisstudyareconsistentwithotherstudiesonemulsiontransportinporousmediaForthecase

ofpolymer-suspendednanoparticles,suchasCMCNZVI,increasingtheNZVIparticlecontent

doesnotinfluencetheviscosityofthesolution[19][41].AμEviscosity,atabout100-1000cP

dependingonthedilution,wasalsolargerthantheviscosityofCMCNZVI,atabout10-50cP

[19].TherelativelyhighviscosityoftheconcentratedμEironoxidemayhinderμEironoxide

transportthroughporousmediaandmaycontributetohighpressuredrop,asobservedinO/W

emulsiontransportstudies[42].

29

(a)

(b)

Figure2.5Viscosityprofilegraph(logscaled)ofμEironoxide(a)andME(b),comparison

betweentheoriginalformulationsanddilutionwithNaClbrinesolution(10g/100mL)at1:1

ratio.

y=0.4784x-0.432

y=0.1128x-0.366

0.001

0.01

0.1

1

10 100 1000

Viscosity

Pa.s

ShearRate(1/s)

MicroemulsionFeo10g/LMicroemulsion5g/L

y=0.0204x-0.191

y=2.0728x-0.594

0.001

0.01

0.1

1

10 100 1000

Viscosity

Pa.S

ShearRate(1/S)

50%DilutionNoDilution

30

2.3.3SizeCharacterization.

Figure2.6demonstratestheapproximatesizerangeofμEironoxideandμENZVIunderTEM.

TheTEMimaginggivesinformationaboutthesizeofindividualparticlesandparticleclusters,

butitisnotsuitabletoindicatethesizeofoil-swollen-micellesadsorbedonthenanoparticles.

Analternativetechniquewouldhavebeentousecryo-TEMthatcouldillustratetheinteraction

ofmicelles(presumablyworm-likemicelles)andthenanoparticles.However,thattechnique

wasnotavailableinourfacilities.Figure2.6thatμENZVI-stabilizedironoxideparticlesarein

therangeof50to100nm,whichconfirmingthesimilarityinsizewithμENZVIbyWangetal.

Sincedynamiclightscatteringcanmeasurethehydrodynamicradiusoftheironoxideparticles

withinthemicellesystem,abetterdescriptionoftheaggregatesizecanbeobtainedviaDLS

measurements.ThemeasureddiameterofμEironoxideat10g/Lis270+/-10nm,this

measuredsizefellwithintherangeofestimatedoptimaltransportofNZVIandabouthalfthe

sizeofCMCNZVI[2][4].SizemeasurementswerealsoconductedtotheμEironoxideeffluent

fromthelowaspectratiocolumn,findinganaggregatesizeof530+/-70nm.Similarresults

werealsoobtainedwithμEironoxideat5g/Lsystem.Eventhoughthesizeofthe

nanoparticlesalmostdoubleduponelutionfromthecolumn,thesizesarestillconsideredinthe

optimalrange.SimilaraggregationeffectshavebeenobservedinthecolumneffluentofCMC

NZVIstudies[2][4].

31

(a)

(b)

Figure2.6TEMimagingofMicroemulsionironoxideat5g/Lwith100nmasscale(a)and

microemulsionNZVIat1g/LbyWangetal.

2.3.4μEIronOxideTransport

Figure2.7presentpicturesofthecolumnstudies(1cmx15cmcolumn)obtainedwithμEiron

oxidesuspensionscontaining10g/L(left)and5g/L(right)Fe,whentheporevelocitywas5

m/day.The10g/Lsystemshowsaclearaccumulationofironinthebottomhalfofthecolumn.

32

Atthatpoint,theflowwasstoppedbecauseofthelargepressuredropobtainedwiththis

system(morethan100inchesofwater).Thepictureontheleftshowstheironcompletely

distributedthroughoutthecolumnafterthefirstporevolumebrokethroughthecolumn.As

indicatedintherheologicalstudies,thelargeviscosity(1200cP)couldbethereasonforthe

largepressuredrop.OnecanusetheKozeny–Carmanequationtoestimatethepressuredrop

forapackingofsmoothmonodispersespheres[43]:

∆𝑃 = (BAV)+BcadV+XYZ+

𝑣 ∙ 𝐿 (9)

Usingaporevelocity(v)of5m/day,aviscosity(μ)of1200cP,aparticlediameter(d50)of0.5

mm,abedporosity(ε)of0.35,acolumnlength(L)of15cm,thepressuredropshouldhave

been104”water.AlthoughthisisconsistentwiththepressurereportedinTable2.1,the

pressurewassubstantiallylargerthan100”water.Evenafterbypassingthepressuregauge,no

flowcouldbeinjectedthroughthecolumn.Itispossiblethatchangesinthestructureofthe

worm-likemicellesystemcouldhaveledtochangesinviscosityintheleadingedgeoftheμE

suspensionfrontincontactwiththeconditioningphase(a10%NaClbrineforthesystemsof

Figure2.7).Infact,accordingtoChoietal.[30],wormlikemicellesproducedwithsuspensions

similartothoseusedinthisworkdisplaygel-likepropertiesatlowenoughshear.

33

a. b.

Figure2.7Transportofironoxidesuspensionsin1-cmdiameter(highaspectratio)columnat5

m/dayporevelocity(a)10g/Lironoxide(b)5g/Lironoxide.

Toovercomethegel-likebehavioratlowshearrateforthe10g/Lsystem,theporevelocitywas

increasedto20m/day.Thisincreaseinvelocitydisruptedthegel-likestructureformedwiththe

10g/Lsystem,allowingtherecoveryof60%ofironoxide,asindicatedinTable2.1.However,

porevelocitiescloserto5m/dayarepreferred,closertotheporevelocitiesusedinaquifer

remediation.Tomaintainaporevelocityof5m/day,thediluted5g/Lsuspensionwas

consideredfortherestofthestudies.Figure2.8presentsthebreakthroughcurvesobtained

withthe5g/Lsuspension,injectedat5m/day,usingcolumnswithaspectratioof15(1cmx

15cm)and6(2.5cmx15cm).The3-compartmentmodelparameterscorrespondingtothese

systems(50%dilutedμE,5g/L,scheduleA)arepresentedinTable2.1.Thepressuredrop

34

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4

C/Co

Porevolume

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4

C/Co

Porevolume

5g/L2.5cmx15cmcolumn

5g/L1cmx15cmcolumn

calculatedfromEquation9was22”water,lowerthanthereportedvalueof38”inTable2.1,

howeveritistobeexpectedgivenotherpressurelossesintheentireconfiguration.

Figure2.8Breakthroughcurvesof5g/L(asFe2O3)μEsuspensionofironoxideinjectedat

5m/day(porevelocity)throughcolumnswithaspectratioof15(left)and6(right).Thesolid

linesshowthesolutionofthe3-compartmentmodelusingtheconstantssummarized.

AsshowninFigure2.8,whenthesameexperimentwasconductedintwodifferentcolumn,the

outcomewasverysimilar.Whilebothcolumnstudiesappeartodisplayasecondarypeakatthe

end,thisfeaturecouldonlybereproducedasatailingeffectthroughthereversibleadsorption

compartment.Thistailingeffectissimilartootherreportedstudies[6][22][24]anditisbelieved

tobeduetoreversibledeposition.Table2.1showsthatthefittingparametersusedforboth

modelswerethesamewiththeexceptionofaslightlyhigherkrevforthecolumnwithaspect

ratioof6.Thehighercoefficientofdetermination(R2)obtainedwiththe2.5cmx15cmcolumn

35

(aspectratio6)wasmainlyduetothelargernumberofsamplespointsthatonecancollectfor

thatsystem.

Becauseofthelargernumberofsamplingpoints,andthemorehomogeneousflowwithinthe

column,the2.5cmx15cmcolumn(aspectratio6)wasusedfortherestofthestudies.

ForthesebaselinesystemsofFigure2.8,theattachmentefficiency(α)wasoftheorderof10-3,

whichisinlinewiththelowerrangereportedbyKocuretal.forlowporevelocities[12].

However,whencomparingthesystemswiththeclosestcharacteristics(2.5g/LNZVI,v=4

m/day),theirattachmentefficiency(α)wascloseto0.2,almosttwoorderofmagnitudehigher

thanthesystemsexploredinFigure2.1.Thissuggestthatatleastfortheconditionsofthe

curvesinFigure2.8,theuseofμEassuspendingmediaimprovestheabilitytotransportthe

particles.AnotherinterestingobservationintheworkofKocuretal[19],andtheworkof

Tufenkjietal.[37]isthattheoptimalsizetominimizethesinglecollectorefficiency(ηo)isclose

to500nm,whichisthefinalsizeoftheaggregatesintheeffluentofthecolumn.Thedrawback

ofusinglargerparticlesizeisthattheycansinkundertheeffectofgravityandaccumulateat

thebottomofthecolumn(ortheaquifer).Thesettlingvelocity(vgr)ofsphericalparticlesin

dilutesuspensions,inlaminarflowcanbeestimatedusingtheStokesequation:

𝑣f0 =(ghAgi)∙f∙XYZ+

Bcd (10)

whereρpisthedensityoftheironoxideparticle(5200kg/m3)andρfisthedensityofthefluid

(assumedwater,1000kg/m3).Consideringaparticleof270nmina250cPfluid,thesettling

velocityfromEquation10is5.8E-5m/d,wellbelowtheporevelocityusedinthiswork.Particle

aggregatesinbetween100nmand1000nmaretoobigtodiffuseviarandommotion,butat

thesametime,toosmalltosettle,thusthesuitabilityofparticlesinthisrangetofacilitatethe

36

transportoftheparticles.Theintrinsicdiffusivity(Dint)ofthe270nmparticlesina250cPfluid

canbeestimatedusingtheStokes-Einsteinequation:

𝐷?j# =kl∙m

UndXYZ (11)

wherekBisBoltzmann’sconstantandTisthetemperatureofthesystem(298K).Atthese

conditionstheintrinsicdiffusivityoftheparticlesis6.5·10-11cm2/s.Thisvalueissubstantially

smallerthanthevalueoftheorder10-5to10-6cm2/sfortheeffectivediffusivityofparticlesin

Table2.1.Potentialback-mixing(non-idealplugflow)mayberesponsibleforthelarger

effectivediffusivity.Infact,thelaminarnatureoftheflowcreatesadispersioneffectasthe

particlestravellingclosertothesurfaceoftheporeswillhavealowervelocitythantheparticles

travellingalongthecenterofthepore.Theeffectivediffusion(Deff),ordispersion,coefficient

(duetolaminarflowsegregation)canbeestimatedusingTaylor’sdispersionequation:

𝐷1oo =2∙pYZ+

+

qcrHs; (11)

For270nmparticles,injectedat5m/day,withDint=6.5·10-11cm2/s,Deff=2·10-6cm2/s,a

valuethatisclosertotheeffectivediffusioncoefficientfoundinTable2.1.

AlthoughtheresultsaboveshowthatthetransportofparticlesisfavouredinthediluteμE

systems,thereisapracticalissueinvolvedwithusingtheconditioning/rinsingprescribedin

scheduleA(Figure2).ScheduleAprescribesinjecting10PVofa10%NaClsolutionbeforeand

aftertheinjectionofthemicroemulsion.Thishighconcentrationofsaltcanimpactthe

chemistryandecosystemoftheaquifer.Tothisend,itwouldbebesttoinjectalowionic

strengthsolutionthatwouldbemorecompatiblewithexistinggroundwater.However,thisis

likelytohaveanimpactonthephasebehavioroftheμEironoxidesuspension.Toassessthis

37

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4

C/Co

Porevolume

ScheduleB

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4

C/Co

Porevolume

ScheduleA

potentialimpact,abreakthroughcurvewasobtainedusingscheduleB(deionizedwater

conditioning/rinsing).ThecomparisonbetweentheseschedulesispresentedinFigure2.9.

AsshowninFigure2.9,theintroductionofdeionizedwater,insteadof10%NaCl,producesa

significantreductionontherecoveryofironoxidenanoparticles,fromvaluescloseto90%

(scheduleA)tocloseto60%(scheduleB).ForScheduleB,theattachmentefficiency(α)

increaseto0.023thatis,still,oneorderofmagnitudelowerthanthatofKocuretal.[19].

Figure2.9Breakthroughcurvesobtainedfor5g/LμEironoxideinjectedat5m/day(pore

velocity)througha2.5cmx15cmcolumn(aspectratioof6),usingscheduleA(10%NaCl

conditioning/rinsingfluid)andscheduleB(deionizedwaterconditioning/rinsingfluid.

TheresultsofFigure2.9confirmthehypothesisthatchangesinsalinitymayinducephase

changesthatdestabilizetheparticles.Togainabetterunderstandingofthetransportofthe

38

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4

C/Co

Porevolume

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4

C/Co

Porevolume

ScheduleA ScheduleB

microemulsionitself,theexperimentsofFigure2.9wererepeated,butintheabsenceofiron

oxidenanoparticles.Figure2.10presentsthebreakthroughcurvesfordilutedμEsinjectedusing

schedulesAandB.

Figure2.10BreakthroughcurvesobtainedfordilutedμEs(noironoxide)injectedat5m/day

(porevelocity)througha2.5cmx15cmcolumn(aspectratioof6),usingscheduleA(10%NaCl

conditioning/rinsingfluid)andscheduleB(deionizedwaterconditioning/rinse).

Whentheparametersofthe3-compartmentmodel(Table2.1)fortheparticle-freedilutedμEs

(Figure2.10)arecomparedtothosecontaining5g/Lironoxide(Figure2.9)onefindsthatthey

arealmostthesame,withtheexceptionthatthevolumeratioofthereversiblecompartmentis

larger(frev)whichresultsinamorepronouncedtailingeffect.Furthermore,theattachment

constant(Katt)iszerofortheparticle-freeμEinjectedwithscheduleAanditissmallforthe

39

particle-freeμEinjectedwithscheduleB.Evenintheabsenceofironoxideparticles,thelarge

changesinsaltconcentrationexperiencedduringscheduleBinduceirreversiblelosesofμEto

thecolumn.

Despitethelowerrecovery,scheduleBisstillmorefavourableforpotentialapplicationas

deliverystrategyduetothelowerriskofimpactingthechemistryandecosystemofthe

reservoir.Also,undesirabledensenon-aqueousphaseliquid(DNAPL)mobilizationcouldalso

occurifthesystemiskeptatconditionsthatcanproduceultralowinterfacialtensions,asisthe

caseforscheduleA[10][16].

AfinalpointofinterestregardingthepotentialuseofμE–basedsuspensionsistheestimated

traveldistance(Lmax),accordingtoEquation7.EvenwhenusingScheduleB,themaximum

penetrationdistanceisintheorderofhundredsofmeters.Thisdistanceissubstantiallylarger

thanthetypicalwelltowelldistanceinNZVIremediation,andlargerthanthosereportedfrom

otherstudies(intherangeof1-10meters)[44].

μE,% 100 100 50 50 50 50 50Fe2O3,g/L 10 10 5 5 5 0 0vpore,m/day 5 20 5 5 5 5 5Aspectratio 15 15 15 6 6 6 6Schedule A A A A B A BDeff·10-5,cm2/s - - 8.7 8.7 0.87 8.7 8.7krev·10-6,1/s - - 5.8 7.7 7.7 7.7 7.7frev - - 0.1 0.1 0 0.2 0.2Katt·10-6,1/s - - 0.31 0.31 1.4 0 0.19Recovery,% 0 58 88 88 56 100 93Cmax/Co - - 0.9 0.9 0.7 0.94 0.91α·10-3 - - 5.4 5.4 23.9 ND NDLmax,m - - 863 863 197 ND 1382R2fit 0.8 0.92 0.98 0.94 0.92ΔPmax,“H2O >100 >100 >100 38 35 36 35

40

Table2.1Summaryofbreakthroughcurveparameters

2.3.5IronDistributionAnalysis

Figure2.11comparesthedistributionofironoxidedepositedonthecolumnbetweenthe

transportofthe5g/LironoxidenanoparticlesfollowingschedulesAandB.Forthecaseof

scheduleA,becauseofthehighrecovery,theamountofretainedironperweightofthesandis

low.Thepredictedattachediron(fromthe3-compartmentmodel)matchesthemeasurediron

attachedtosand.ToputthenumbersofFigure2.11inperspective,Xinetal.[24],injected5PV

of3g/L(Fe)NZVIat8.3m/day(conditionssomewhatsimilartoourstudy),obtainingan

attachedironof5mg/g,morethanoneorderofmagnitudelargerthantheironattached

obtainedinFigure2.11.

ComparedtoscheduleA,moreironwasattachedinscheduleB.The3-compartmentmodel

predictioncoincideswiththeretentioninthebottomofthecolumnoperatedwithscheduleB,

butnotintherestofthecolumn,wheretheamountofironattachedwaslowerthanthe

predictedamount.ThiscouldsuggestthatnotalltheattachmentinthecaseofscheduleBwas

attachmenttothesand,butapartitiontogel-likestructuresthatwerenotcollectedwhenthe

sandwassampledfromthecolumn.

ThepicturesforthescheduleAstudyshowbarelyvisiblespecksonthesurfaceoftheparticles

homogenouslydistributedthroughoutthegrains.ForthecaseofscheduleB,largerspecksare

shown,likelytheresultofaggregation,especiallyatthebottomofthecolumn.

41

Figure2.11Ironoxidedepositedonsandcolumnaftertheinjectionof1.5PVof5g/L(as

Fe2O3)ironoxidenanoparticlesat5m/day(porevelocity)througha2.5cmx15cmcolumn

(aspectratioof6),usingscheduleA(10%NaClconditioning/rinsingfluid)andscheduleB

(deionizedwaterconditioning/rinsingfluid).Thesolidlinerepresentsthepredictionof

depositedironfromthe3-compartmentmodel.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 3 6 9 12 15

Attached

iron

oxide

,mgF

e/gs

and

Columnlength,cm

ExperimentalPredicted

1mm

ScheduleA

ScheduleB

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 3 6 9 12 15

Attached

iron

oxide

,mgF

e/gs

and

Columnlength,cm

ExperimentalPredicted

1mm

42

2.4Conclusions

Theoriginalintentofthisworkwastoevaluatethetransportofmicroemulsion(μE)stabilized

ironnanoparticlesthroughporousmedia,usingironoxidenanoparticlesasmodelsystem.The

original10g/L(asFe2O3)suspensiondespitebeinghighlystable,ithadrheologicalproperties

thatpreventeditsuseatporevelocitiesconsistentwiththoseusedinaquiferremediation.

Dilutingthemicroemulsiontoa5g/Lsuspensionachievedreasonablepressuredropsatpore

velocitiesof5m/day,consistentwiththoseusedinaquiferremediation.

Thecolumnstudiesconsideredinthisworkwereanalyzedusinga3-compartmenttransport

modelthataccountforthetransportinthecolumnandtheexchangeofparticleswitha

reversibleadsorptioncompartmentandanirreversibleattachmentcompartment.

TheμEusedtosuspendthenanoparticleswaspreviouslydesignedtoformbicontinuous

systemsthatupondilutionin10%NaClbrinesolutionwouldyieldworm-likemicelles.When

usinga10%NaClbrinesolutiontoconditionandrinsethecolumnaftertheinjectionofthe

suspension,alargefractionofnanoparticleswasrecoveryandtheparticleattachment

experiencedinthecolumnwaslessthan1/10theattachmentobtainedwithothersuspension

mediareportedintheliterature.Whendeionizedwaterwasusedasconditioning/rinsing

solutionmoreparticleswereretainedbythecolumn,likelybecausephasetransitions

experiencedbytheμEphaseduetothelargechangeinelectrolyteconcentration.However,

eventhislessdesirabletransportwasstillmoreefficientthanothersuspensionsreportedinthe

literature.FuturestudiesshouldconcentrateinproducingμEsuspensionswithlowelectrolyte

concentrationsuchthatthecolumncanbeconditionedandrinsedwithlowenoughionic

strengthsolutionsthatwouldnotinducesubstantialchangesintheμEphasebehavior.

43

2.5References:[1] F.Fu,D.D.Dionysiou,andH.Liu,“Theuseofzero-valentironforgroundwaterremediationandwastewater




vol.48,no.5,pp.2862–2869,2014.

[3] P.G.TratnyekandR.L.Johnson,“Nanotechnologiesforenvironmentalcleanup,”NanoToday,vol.1,no.2,pp.44–48,

2006.

[4] B.Sunkara,J.Zhan,J.He,G.L.McPherson,G.Piringer,andV.T.John,“Nanoscalezerovalentironsupportedon

uniformcarbonmicrospheresfortheinsituremediationofchlorinatedhydrocarbons,”ACSAppl.Mater.Interfaces,

vol.2,no.10,pp.2854–2862,2010.

[5] A.B.Cundy,L.Hopkinson,andR.L.D.Whitby,“Useofiron-basedtechnologiesincontaminatedlandandgroundwater

remediation:Areview,”Sci.TotalEnviron.,vol.400,no.1–3,pp.42–51,2008.

[6] Y.H.Lin,H.H.Tseng,M.Y.Wey,andM.DerLin,“Characteristicsoftwotypesofstabilizednanozero-valentironand

transportinporousmedia,”Sci.TotalEnviron.,vol.408,no.10,pp.2260–2267,2010.

[7] J.Quinn,C.Geiger,C.Clausen,K.Brooks,C.Coon,S.O’Hara,T.Krug,D.Major,W.S.Yoon,A.Gavaskar,andT.

Holdsworth,“FielddemonstrationofDNAPLdehalogenationusingemulsifiedzero-valentiron,”Environ.Sci.Technol.,

vol.39,no.5,pp.1309–1318,2005.



[9] J.E.Huff,“ApplicationofEmulsifiedZero-ValentIron:FourFull-ScaleRemediationSites,”Remediat.J.,pp.125–136,

2011.





[12] J.Zhan,T.Zheng,G.Piringer,C.Day,G.L.Mcpherson,Y.Lu,K.Papadopoulos,andV.T.John,“Transportcharacteristics

ofnanoscalefunctionalzerovalentiron/silicacompositesforinsituremediationoftrichloroethylene,”Environ.Sci.

Technol.,vol.42,no.23,pp.8871–8876,2008.

[13] T.Phenrat,N.Saleh,K.Sirk,R.D.Tilton,andG.V.Lowry,“Aggregationandsedimentationofaqueousnanoscale

44

zerovalentirondispersions,”Environ.Sci.Technol.,vol.41,no.1,pp.284–290,2007.



[15] J.Chen,Z.Xiu,G.V.Lowry,andP.J.J.Alvarez,“Effectofnaturalorganicmatterontoxicityandreactivityofnano-scale

zero-valentiron,”WaterRes.,vol.45,no.5,pp.1995–2001,2011.

[16] Z.WangandE.Acosta,“FormulationdesignfortargetdeliveryofironnanoparticlestoTCEzones,”J.Contam.Hydrol.,

vol.155,pp.9–19,2013.






vol.145,pp.17–25,2013.

[20] T.Phenrat,Y.Liu,R.D.Tilton,andG.V.Lowry,“AdsorbedpolyelectrolytecoatingsdecreaseFeonanoparticle

reactivitywithTCEinwater:Conceptualmodelandmechanisms,”Environ.Sci.Technol.,vol.43,no.5,pp.1507–1514,

2009.



57,2007.

[22] S.M.HosseiniandT.Tosco,“Transportandretentionofhighconcentratednano-Fe/Cuparticlesthroughhighlyflow-

ratedpackedsandcolumn,”WaterRes.,vol.47,no.1,pp.326–338,2013.

[23] E.DVecchia,MLuna,andR.Sethi,“Transportinporousmediaofhighlyconcentratedironmicro-andnanoparticlesin

thepresenceofxanthangum,”Environ.Sci.Technol.,vol.43,no.23,pp.8942–8947,2009.



[25] S.Laumann,V.Micić,andT.Hofmann,“Mobilityenhancementofnanoscalezero-valentironincarbonateporous

mediathroughco-injectionofpolyelectrolytes,”WaterRes.,vol.50,pp.70–79,2014.



45



[28] P.Bennett,F.He,D.Zhao,B.Aiken,andL.Feldman,“Insitutestingofmetallicironnanoparticlemobilityandreactivity

inashallowgranularaquifer,”J.Contam.Hydrol.,vol.116,no.1–4,pp.34–46,2010.

[29] Z.Wang,“SYNTHESISOFSTABLEANDREACTIVEMICROEMULSIFIEDZERO-VALENTIRONNANOPARTICLES(MENZVI)

USINGEXTENDEDSURFACTANT,”2015.

[30] E.A.FrancisChoi,“MechanismofStabilityofParticleSuspensionsviaLipophilicMicellesby,”2015.

[31] V.Dwarakanath,K.Kostarelos,G.A.Pope,D.Shotts,andW.H.Wade,“Anionicsurfactantremediationofsoilcolumns

contaminatedbynonaqueousphaseliquids,”J.Contam.Hydrol.,vol.38,no.4,pp.465–488,1999.

[32] D.A.Sabatini,R.C.Knox,J.H.Harwell,andB.Wu,“IntegrateddesignofsurfactantenhancedDNAPLremediation:

Efficientsupersolubilizationandgradientsystems,”J.Contam.Hydrol.,vol.45,no.1–2,pp.99–121,2000.

[33] Z.Wang,A.Lam,andE.Acosta,“SuspensionsofIronOxideNanoparticlesStabilizedbyAnionicSurfactants,”J.

SurfactantsDeterg.,vol.16,no.3,pp.397–407,2013.

[34] E.M.Hotze,T.Phenrat,andG.VLowry,“Nanoparticleaggregation:challengestounderstandingtransportand

reactivityintheenvironment.,”J.Environ.Qual.,vol.39,pp.1909–1924,2010.

[35] R.A.CraneandT.B.Scott,“Nanoscalezero-valentiron:Futureprospectsforanemergingwatertreatment

technology,”J.Hazard.Mater.,vol.211–212,pp.112–125,2012.

[36] K.Yao,M.T.Habibian,andC.R.O’Melia,“WaterandWasteWaterFiltration:ConceptsandApplications,”Environ.Sci.

Technol.,vol.5,no.11,pp.1105–1112,1971.

[37] N.TufenkjiandM.Elimelech,“CorrelationEquationforPredictingSingle-CollectorEfficiencyinPhysicochemical

FiltrationinSaturatedPorousMedia,”Environ.Sci.Technol.,vol.38,no.2,pp.529–536,2004.

[38] R.Nowakowski,“Computersimulationofnon-linearpreparativechromatographyprocesses,”Chromatographia,vol.

28,no.5–6,pp.293–299,1989.

[39] K.J.Ward,S.C.Kaliaguine,P.A.Tanguy,andG.Jean,“Numericalsimulationofachromatographcolumn:linearcase,”

Ind.Eng.Chem.Res.,vol.27,no.8,pp.1474–1480,1988.

[40] S.Lakshmanan,W.M.Holmes,W.T.Sloan,andV.R.Phoenix,“Nanoparticletransportinsaturatedporousmedium

usingmagneticresonanceimaging,”Chem.Eng.J.,vol.266,pp.156–172,2015.

[41] J.Li,S.Bhattacharjee,andS.Ghoshal,“Theeffectsofviscosityofcarboxymethylcelluloseonaggregationandtransport

ofnanoscalezerovalentiron,”ColloidsandSurfacesA:PhysicochemicalandEngineeringAspects,vol.481.pp.451–

46

459,2015.

[42] A.MandalandA.Bera,“Modelingofflowofoil-in-wateremulsionsthroughporousmedia,”Pet.Sci.,vol.12,no.2,pp.

273–281,2015.

[43] D.Thies-WeesieandA.Philipse,“Liquidpermeationofbidispersecolloidalhard-spherepackingsandtheKozeny-

Carmanscalingrelation,”Journalofcolloidandinterfacescience,vol.162,no.2.pp.470–480,1994.

[44] F.He,M.Zhang,T.Qian,andD.Zhao,“Transportofcarboxymethylcellulosestabilizedironnanoparticlesinporous

media:Columnexperimentsandmodeling,”J.ColloidInterfaceSci.,vol.334,no.1,pp.96–102,2009.

47

Chapter3:Developmentandtransportofphosphatesurfactant,SDEHP-

stabilizedNZVIinporousmediaforinsituremediation

3.0Abstract

Nanoscalezero-valentiron(NZVI)particleshasbeenidentifiedasefficientreducingagentsfora

widerangeofgroundwatercontaminants;however,itsapplicationislimitedbythepoor

transportanddeliveryintheporousmedia.Inthiswork,sodiumDiEthylHexylPhosphate

(SDEHP)surfactantwasusedasasurfacemodifiertoimprovethemobilityofNZVIinsoil.An

optimalSDEHP-stabilizedNZVIformulation,100mMSDEHP1g/LNZVI,wasidentifiedthrough

criticalmicelleanalysisandtotalorganiccarbonanalysis.Simple“one-pot”NZVIsynthesis

procedurewasadaptedandmodifiedtosynthesizeNZVIformulations.TheSDEHPsurface

modifierimprovedthestabilityofbareNZVIfromminutestomonths.Theoptimized

formulationyieldedahydrodynamicdiameterabout240nmunderdynamiclightscattering

(DLS)and100nmindiameterundertransmissionelectronmicroscope(TEM).Acolumnstudy

ofthestableformulationisconductedwitha2.5x15cmglasscolumnfilledwithOttawasand

atafieldscaleflowvelocityof1.5m/daywithnomechanicalstirring.Theresultofthecolumn

studyshowedthatover95%ofNZVIarerecoveredwithasteadyplateauC/Copeakreachingto

1.AdiscussionsuggestedthatthedevelopedSDEHPsurfacemodifierholdsdesirabletraitsfor

afieldscaleinsituremediation.

48

3.1Introduction

Nanoscalezero-valentiron(NZVI)particlesarerecognizedasanusefulgroundwater

remediationtechnologybecauseofitshighefficiencyinreducingawiderangeofcontaminants,

includingbutnotlimitedtoDNAPLandheavymetals[1][2][3][4].However,thistechnologyis

limitedbythepoormobilityinsoil[5][6].TheunmodifiedNZVIparticles,commonlyknownas

bareNZVI,duetohighmagneticandVanderWaalforces[3][7],aggregateandsettlewithin

minutesofbeingsynthesized.AggregatedZVIdemonstratedlosesitsoriginalhighsurfacearea

andbecomelessreactivetotargetcontaminants.TheseLargerparticlesarefilteredbythesand

grain,hinderingtheZVIparticlesfromreachingthecontaminantzoneandimmobilizedinsoil

[8].

TransportofNZVIinporousmediaisinfluencedbyvariousfactorssuchastransportvelocity,

pH,particlesizes,ionicstrength,soilmatrixandthecompositionofgroundwater[9][10].NZVI

particlestendtoaggregateandsettlebeforeencounteringthesoilmatrixandgroundwateror

clogattheearlystageofsubsurface[11][12].Addressingthestabilityissuebyeliminatingor

extendingthesettlingtimeandkeepingtheoriginalsizesoftheparticleshasbeenshownto

improvenanoparticlesmobility[13][14].Fortunately,particlesizesandstabilityofthe

suspensionarealsocloselyrelatedandcaneasilybeengineeredthroughtheapplicationof

surfacestabilizers[15][16][11][17],namelysurfacestabilizers.Currently,extensiveresearch

effortshavebeencommittedtotheapplicationofsurfacestabilizersonNZVI,includingto

polymers,surfactantsandemulsionsareused[13][18][15].However,surfacestabilizationon

NZVIparticlescanreducethereactivityofthereactiveparticlesbycreatingabarrierbetween

contaminantsandNZVI[19];alongwithothertrade-offssuchasconcentration[20][13][21]and

49

injectiontechnique[8][22],stabilizedNZVIarenotalwayscandidatesforfieldremediation.

O’Carrolletal.summarizedthreefundamentalcharacteristicsthatanefficientlystabilizedNZVI

shouldhold:1.Demonstrateprolongparticlestability2.SufficientNZVIconcentration(1-12

g/L)toachievesuccessfulremediationand3.Maintainanadequatereactively[3].Applying

surfacestabilizersthatprovidethethreefundamentalcharacteristicshasbeentheobjectiveof

thisandotherstudies.

Laboratorycolumnexperimentsareoftenthefirststeptodetermineandassesstheabilityofa

stabilizedNZVI.Table3.1summarizes,theperformanceofselectednanoparticlesstabilizers

andtheircolumntransportresults[6][13][23][18][22][24][25][26][27][28][29].Outofthese

stabilizers,carboxyl-methylcelluloseNZVIat1g/LwasfieldtestedinSarnia,ONin2014by

Kocuretal.[30]Overall,thesurfacestabilizerscanimprovethestabilityandthusthemobilityof

theironnanoparticlesuspensionwithgreaterrecoveryandtransportchrematistics;however,

furtherimprovementsarehinted.7outofthe10listedcolumnstudieslistedwereconducted

ataflowvelocitybetween7.8to198.7m/day,wellabovethetypicalgroundwaterflowvelocity

of0.25to0.4m/day[3],implyingacompatibilityissueduringtheprocessofinjection[31].

Additionally,adequatebreakthroughperformanceofNZVIareonlyobservedatthelow

concentrations(below1g/L)forcarboxyl-methylcellulose,guargumandpoly(acrylicacid)

basedNZVI[29][8][16];whilealltheNZVIstabilizersincludingsurfactantTween-80yielded

poorbreakthroughresultsathigherconcentrationswithlowmaximumpeakanduneven

breakthroughcurves.Thismeansthatmoresevereagglomerationandinstabilityareobserved

athigherconcentrations[7].Furthermore,polymerstabilizerssuchascarboxyl-methylcellulose

althoughclaimedaprolongedstabilizationof80-hours,aggregationswerenotfullyeliminated

50

andsettlingisobservedconstantlyoverthe80-hourperiod.Intheaforementionedfieldstudy

ofcarboxyl-methylcelluloseNZVI,itwasreportedthatNZVIcantravelatleast1meterwitha

recoveryrateof1%[30][32].ItisclaimedbySoukupovaetal.thatevena10dayperiodof

negligibleaggregationisnotenoughforafullscaleremediation[27].Non-ionicsurfactant

stabilizedNZVI,Tween-80,demonstrateda2-monthstabilityunderstoragecondition;however,

instabilitywasobserveduponcontactingwiththeporousmedia.Ontheotherhand,emulsion

encapsulatedNZVI,eventhoughyieldedahighandsteadyrecoverypeak(Cmax/Co)ata

desirablefieldflowvelocity(0.4m/day);theinjectionofemulsionNZVIrequiresconstantly

mechanicalstirringandcreateddifficultyininjection.Althoughapplyingsurfacestabilizers

improvedthemobilityofNZVIinporousmedia,laboratorycolumnstudiesshowedthatthere

hasn’tbeenasurfacestabilizerstrongenoughforanefficientfull-scaleremediation.Other

words,asurfacestabilizerthatcansatisfythebasicNZVIcharacteristicswithnoaggregation

andsettlingovertimehasyetbeenfound.

Surfactantsaresometimespreferredoverpolymersasstabilizingagentsbecausetheyhavea

highertendencytoadsorbontothesurfaceofnanoparticles[27].Biodegradablenon-ionic

surfactantssuchasTween-20,Tween-80andAlkylethanolamidesareoftenadaptedas

alternativestabilizersforNZVIforenvironmentalreasonsandsmoothsynthesis[33][34][27].It

isalsoreportedthatsurfactantscanincreasethereductionrateofthecontaminantsdueto

synergisms[35][36].Eventhoughnon-ionicsurfactantsholdseveraladvantages,theystill

producepoorstability[12].Wangetal.showedthatanionicsurfactantscanformelectrostatic

repulsionsbetweencoatedparticlesandsand,creatinganenergybarrierforNZVI

agglomeration[15].Itisalsoproventhatanionicsurfactantironnanoparticleswithstrong

51

chargedcoatingsholdmoderatestabilityandmobility,assummarizedinTable1,namelyoleate

ionstabilizedironoxidenanoparticles[23].

Figure3.1StructureofanionicphosphatesurfactantSDEHP.

Inthisstudy,sodiumdiethylhexylphosphate(SDEHP),ananionicsurfactanthasbeenselected

asthestabilizerbecauseofitsnon-toxic,mildnatures[37].AndthefactthatWangetal.

demonstratedthatsodiumdodecylphosphate(SDP),anotherphosphatebasedsurfactant,can

actasastabilizerandpromotethereductionofchlorinatedhydrocarbons[38].Figure3.1

showsthestructureofsurfactantSDEHP,withphosphategroupbeingthehydrophilicheadand

doublecarbonchains.ItwasreportedthatSDP,anotherphosphatesurfactant,doesnotreact

withNZVIasastabilizerbutinfactpromotethedechlornatingreactions.Inadditional,itis

importanttonotethatphosphonategroupinanionicsurfactantsplayamajorroleinforming

morestablesuspensionwithmetalnanoparticles[39].Withthisknowledge,surfactantswith

phosphategroup[39]thatsharesimilarpropertiesareexpectedtohavesimilarperformancein

suspension.TheSDEHPanionicsurfactant,withthereactivityadvantageoverthetypical

biodegradablenon-ionicsurfactants,wasselectedasapotentialstabilizerofNZVIfor

developinganefficientalternativeforNZVIinsituremediation.

52

Table3.1Literaturesummaryofcolumnstudiesandstabilitybehaviourfordifferenttypesof

surfacemodifiedironoxideandZVInanoparticles.

TheprimaryobjectiveofthispaperistodevelopaSDEHPsurfactant-basedformulationthat

SystemNo.

Max.Conc.(g/L)

Stability TransportVelocity(m/Day)

Viscosity Cmax/Co Size(nm)

Reference

Carboxyl-methylCelluloseNZVI

1 0.1-2.5 80hourswithaggregation

0.25,2,4 13.8-72.8

0.85-0.75(Decreasingtrend)

25-61 [13]

Polyeletrolyte-stabilizedNZVI

2 0.085-1.7

Stirringwhileinjection

6.4 0.97 0.8 85(particle)185(hydrodynamic)

[40]

EmulsionNZVI 3 2.5 Kineticallystable~8hours/Mixingduringinjection

0.4 9300 0.8-1 1000(droplet)12(particles)

[41]

PolyacrylicAcid-StabilizedNZVI

4 0.1,0.3,4

3hoursAgitationrequired

6.28-15.67 N/A ~1 ~100nm [6]

Polyphenol-basedNZVI

5 1 >10days 7.4 N/A 0.5 n/a [28]

GuarGumNZVI

6 0.154 Days 2.38-11.92 0.89-1.35

0.21(lowflow)-0.87(highflow)

320 [20]

XantumGumNZVI

7 3 72hours 8.4-198.72 ShearThinning10,000

0.4-0.6(increasingtrend,lowflow)0.8-1(fastestflow)

microscale [26]

Tween80NZVI

8 20%(w/w)0.32(columnstudy)

2monthsunderstorageconditions

24.32 N/A N/A 40-80 [27]

Oleate(OL)ionsstabilizedironoxide

9 5 >2weeks 14 N/A 0.93 N/A [23]

MicroemulsionIronoxide

10 5and10 >6months 5,20 20-400 0.9 120 Chapter2

53

demonstratesprolongedstabilityandminimumaggregationusingtheframeworkproposedby

Wangetal.[15].Thesecondaryobjectiveistoexamineandassessthemobilityinporousmedia

oftheoptimalSDEHP-basedNZVIviaa1Dcolumnstudy.

3.2Methodology

Unlessotherwiseindicated,alltheproceduresandmaterialswerepreparedandconductedat

standardambienttemperatureandpressure(SATP).

3.2.1SynthesisofSodiumDiEthylHexylPhosphate(SDEHP)Surfactant

TheprocedureofproducingSDEHPwasadaptedfromthepublicationsofLuanetal.withsome

modification[37].Inshort,30,50and100mMofSDEHPwassynthesizedviatheneutralization

reactionbetweensodiumhydroxideandhydrogendiethylhexylphosphateacid(HDEHPA,

Sigma-Aldrich,237825,97%).3gofHDEHPAismeasuredbyweightusinganelectronicbalance

(Sartorius,Germany,33904396)ina20mLvialtomake100mMofSDEHP.Themeasured

HDEHPAisthentransferredtoa100mLvolumetricflaskwiththeflushingofDIwater.1M

Sodiumhydroxide(NaOH,Caledon,lot#89075)solutionwasaddedtothevolumetricflaskat

1.2timesthestoichiometricamount.BalancetherestofthevolumetricflaskwithDIwaterand

inducevigorousmixingmanually.Afterthesolutionturnedclear,stoppedthemixingandplace

thevolumetricflaskfor24hourstoreachcompleteequilibrium.Thecompletionofthe

neutralizedSDEHPsolutionwasthenconfirmedwithapHprobe(Vernier,Model:LD2-LE)to

determinetheacidity.Itisrecommendedtoaddseveraldropsofthe1MNaOHsolutionifthe

pHoftheSDEHPsolutionisbelowpH9,thehigherpHenvironmentguranteefullconversionof

theHDEHP.

54

3.2.2CriticalMicelleConcentrationofSDEHPwithdissolvediron

SurfacetensionmeasurementsusingKSVSigmaTensiometer(model700)wasusedto

determinethecriticalmicelleconcentration(CMC)forSDEHP.Ironsulfate(FeSO4[H2O]7,Fisher

Scientific,7782-63-0)wasdissolvedintoSDEHPsurfactantsolutionatthefollowing

concentrations:10,25,30,50,80and100mM.Atotalof0.2gofironsulfate(equivalentto1

g/LNZVI)wasaddedto40mLofeachSDEHPsolutionatdifferentconcentrationwithgentle

mixing.Thesurfactant-ironsolutionwasthenmixedandplacefor1hourtoreachequilibrium

beforemeasurement.Thepurposeofdissolvingironsulfateistosimulatethesynthesis

condition.Identicalprocedurewasconductedforferricchloride(FeCl3,SigmaAldrich,7705-08-

0)andthoseresultsareshownanddiscussedinAppendixA.

3.2.3TotalOrganicCarbon(TOC)ofirondissolvedSDEHP

Totalorganiccarbon(TOC)analysiswasconductedusingatotalcarbonanalyzer(TOC-Vcpn,

SHIMADZU)fordeterminingtheamountofsurfactantadsorptiontotheiron.SDEHPsurfactant

solutionwaspreparedatthefollowingconcentrations:10,30,50,80and100mMfor10mLin

a20mLvial.DifferentconcentrationsofironsulfatesatdifferentNZVIequivalence

concentrationsrangesfrom0.3,0.5,1,1.5,2,2.5,3,4to5g/Lareadded.Thesampleswere

thencappedandmixedfor2minutesusingavortexmixerat1000rpm,thencentrifuged(Cole-

Parmer,1741423)at4000rpmfor45minutes.Thesupernatantwasdecantedandfilteredwith

anano-filter(PallCorp.AcrodiscSyringeFilter,450nm).Thefilteredsolutionwasdilutedto12.5

timesbeforeanalysis.

3.2.4SynthesisofSDEHPNZVI

NZVIiscommonlysynthesizedfromthereductionreactionbetweensodiumborohydride

55

(NaBH4)andiron[42][43][44][2].TherearetwocommonNZVIsynthesismethods,namely

chloride-basedandsulfate-basedsynthesis[43].Inthiswork,sulfate-basedmethodisselected

forreasonsthatwillbefurtherexplainedinresultanddiscussion.ThereactionofNZVIsulfate-

synthesisisasthefollowing:

2𝐹𝑒Ww +𝐵𝐻qA + 3𝐻W𝑂 → 2𝐹𝑒a + 𝐻W𝐵𝑂UA + 4𝐻w + 2𝐻W (1)

The“one-pot”synthesisprocedurewasadaptedfromtheoriginalsynthesisprocedure

describedbyWangetal.forsynthesizingmicroemulsionNZVI[45].Specifically,surfactant

solutionsatthedesiredconcentrations(30,50and100mM)werepreparedusingdeaeratedDI

water.Thesurfactantsolutionsandotherreactantswereplacedinanitrogenfilledgloveboxfor

threehourstoremoveadditionaloxygen.Afterthethreehours,20mLofthesurfactant

solutionwastransferredtoa250mLbeaker.Toproducea1g/Lironsolution,0.1gofiron

sulfatewasweightedusinganelectronicbalanceinsidethegloveboxandtransferredtothe250

mLbeaker.Usingaglassstirringrod,manualgentleagitationwasusedtodissolvetheiron

sulfateforabout15minutes.Atotalof0.04gwasofsodiumborohydride(NaBH4,Sigma-

Aldrich,16940-66-2)wasweightedintheglovebox.Thesodiumborohydridewasslowlyadded

intothesolutionwithin30minutestopreventexcessivegasevolutionfromthereaction.After

this,thesolutionwasleftfor1hourinthegloveboxtoletthesolutiontocooldowntoroom

temperature.Thereductionyielded20mLof1g/LNZVIat30,50or100mMofSDEHP

surfactant.Similarprocedureswereusedtoproduce0.3,0.5,2and5g/LNZVI.Figure3.2

illustratedthedescribedone-potsynthesisprocessofNZVI.

56

Figure3.2Illustrationofthe“one-pot”synthesisprocedureofSDEHP-stabilizedNZVI.The

procedurewasconductedintheglovebox.

3.2.5pHAnalysis

ApHmeterprobewasusedtoanalyzethepHofSDEHPsurfactantat100mMandSDEHP

stabilizedNZVIatdifferentironconcentrationsandsurfactantconcentrations2monthupon

synthesis.

3.2.6StabilityAnalysis

Uponsynthesis,SDEHPNZVIatdifferentformulationsandbareNZVIaretakenoutsideofthe

glovebox(synthesiscondition)forfurthercharacterization.Toassesscolloidalstability,the

synthesizedformulationswerere-suspendedwithasonicationbathfor1minute.Pictureswere

takenattimeintervalof15,30,60,120and180minutesforthefirstthreehoursanddailyfor

oneweekafterre-suspension.

57

3.2.7ViscosityAnalysis

Tomeasuretheviscosityofthesynthesizedsolutions,syringeaspirationtimemethodwasused

[46].ThreeSDEHPNZVIformulations,SDEHPat100mMandDIwateraremeasured.Thewater

viscositymeasurementwasusedasthereferencepointtodeterminetheviscosityofthe

formulations.

3.2.8SizeAnalysis

Dynamiclightscattering(DLS)andtransmissionelectronmicroscope(TEM)areusedtoanalyze

thehydrodynamicandparticlesizesofthesynthesizedSDEHPNZVI,respectively.ForDLS

analysis,thefreshlysynthesizedNZVIatdifferentconditionswerediluted10timeswithDI

waterina20mLvialinthesynthesisconditionwithagloveboxfilledwithmixedair(95%N2

andbalanceCO2).PriortoDLSanalysis,thesamplewasmixedwithavortexmixerfor30

secondsandsonicatedusingasonicationbath(Cole-Parmer,8891)for1minute.Thesample

wasthenanalyzedviaaDLSparticlesizeanalyzer(BrooklynInstrumentCrop,90Plus)

measurementsfor10minutes.ForTEManalysis,identicalproceduretoDLSanalysiswas

followedforsamplepreparationpriortoanalysis.ATEMmicroscope(Hitachi,HF3300)was

usedforanalyzingtheironparticlesizes.

3.2.8ColumnExperimentProcedure

Aglasscolumn(15x2.5cm,KontesBrand,ChromaFlex,No.420830-1S1D)wasfilledwithacid-

washedOttawasand(Silicondioxide,Sigma-Aldrich,60676-86-0)asporousmediumfor1-D

transportanalysis.Thesandmediumwaswet-packedandstirredduringpackingtoremoveair

trappedinthesand.Thecolumnwasweightedbeforeandafterpackingandmassbalancewas

performedtodeterminetheporevolume(1porevolumeorPV=32mL).Aperistaticpump

58

(Cole-Parmer,MasterFlexL/S)wasusedforthisexperimentandtheflowratewassatatthe

lowestsettingat0.5mL/min(equivalentto1.5m/dayasflowvelocity).Thepumpisconnected

toathree-way-valvetoallowswitchingbetweentheflushingfluidandthenanoparticle

solutions.TominimizeNZVIoxidiation,anenclosurewasinstalledaroundtheSDEHPNZVI

solutiontoeliminateoxygenfromtheambient.Inspecific,aN2gascylinderwasconnectedto

anexpandablesmallgloveboxwithapurgestreamataconstantrateof20psi.Thecontinuous

purgingofthenitrogensimulatesthesynthesizingconditionintheglovebox.Theexperiment

startswiththeflushingstage,aninjectionof10porevolumes(320mL)offlushingfluid(DI

water)fromthebottomofthecolumn.Uponcompletionoftheflushingstage,2porevolumes

(64mL)of100mMSDEHPNZVIsolutionwasinjectedintheidenticalcondition.The

transportedsolutionwascollectedwithanautomaticfractionalcollectoratarateof1.5

mL/sample(3minutes)startingfromtheNZVIinjectionstage,thesamplingiscontinueduntil

theendoftheexperiment.Another10porevolumesofDIwaterwereinjectedtothecolumn

towashouttheNZVI.Thecollectedsamplesfromthecollectorismixedwith6Nhydrochloric

acid(HCl,BDH,BDH7204-1)atvolumeratioof1:4(Nanoparticlessolution:HCl6N)for

concentrationandbreakthroughcurveanalysis.Theoverallexperimentalschematicwas

illustratedinFigure3.3.

59

Figure3.3ColumnstudysetupforSDEHP-stabilizedNZVI.

3.2.9NZVIColumnDistributionAnalysis

Uponcompletionofthecolumnstudy,thesandwasrecoveredfromthecolumntodetermine

thedistributionoftheironretainedinthecolumn.Thecolumnwasdividedintofivesections,

witheachonesectionbeing3cminlength.Thecolumnwasdisconnectedandaspatulawas

usedtoretrievedfromeachsection.Theretrievedsandwasanalyzedunderamicroscopefor

particleadsorptionandunderUV/VISspectroscopyforconcentration.Priortothe

concentrationanalysis,thesandwasdigestedinHCl6Nfor3dayswasingasand-acidratioof1

g/5mL.

3.3ResultsandDiscussion

3.3.1DeterminingtheOptimalSynthesisFormulation

Aframeworktodeterminingtheoptimalsurfactant-nanoparticlesformulationisvitalfora

stablesuspension.Inthepast,severalstudieshaveattemptedtouseanionicandnon-ionic

surfactanttosuspendmetalnanoparticlesincludingNZVI[47][39][48][49][30].Ontheother

60

hand,Wangetal.successfullyoptimizedwithaframeworkfortwoanionicsurfactantstabilized

ironnanoparticlestoyieldastablesuspensionforover24hours[23].Asimilarformulationfor

ironnanoparticlesathighconcentrationwasconcludedtobemobileinporousmediathrougha

columnstudybyWangetal[23].Tothisday,nostudyhasusedanyframeworkorstrategyto

determinethemostefficientsurfactantstabilizedNZVIformulation.

ThisstudyadaptedtheformulationstrategyproposedbyWangetal.thatastablenanoparticle

suspensioncanbeachievedifthesurfactantconcentrationisabovethecriticalmicelle

concentration(CMC)afterreachinganequilibriumwithnanoparticles[15].Thedetermination

oftheoptimalformulationofNZVIwasthusdividedintotwoparts:1.TodeterminetheCMCof

SDEHPwiththepresenceironsulfateand2.Todeterminethehighestironconcentration

possibleinthelowestSDEHPconcentrationsolutiontosynthesizeNZVI.

Figure3.4showsthesurfacetensionmeasurementsatdifferentconcentrationofSDEHPfrom

10to100mMwith1g/Lequivalenceofironsulfatedissolved.Theadditionofironsulfate

simulatesNZVIsynthesisconditionwiththesurfacemodifierpriortothereaction.Ferrous

sulfatemayaltertheoriginalCMCconsideringtheinteractionoftheferrousionwiththe

anionicsurfactantthatleadtotheformationofferroussalts,someofwhichprecipitatefrom

solutions.Figure3.4presentstwosurfacetensionmeasurementcurvesfortheaforementioned

scenario.Curve1fromFigure3.4representstheoriginalSDEHPconcentrationandcurve2

representsthecorrectedconcentrationofnon-adsorbedSDEHPbasedonTOCanalysis.From

curve1inFigure3.4,itcanbeobservedthatthesurfacetensiondecreasedlinearlyfrom45

mN/mto25mN/matconcentrations10,20,30and50mMofSDEHP.From50mMonwardsto

100mM,thesurfacetensionreachedtoaplateauatapproximately25mN/m.Thegeneral

61

trendimpliedinfigure3.4isatypicalsurfacetensioncurveofasurfactantandisconsistent

withotherreportedCMCmeasurementsbytheSDEHPstudyinliterature[37][50].Through

logarithminterpolationbetweenthedescendingandplateausectionofthesurfacetension

curve,itwasdeterminedthattheCMCofSDEHPwith1g/LequivalenceNZVIofironsulfate

dissolvedisabout57mM.Thisisabout30-40mMhigherthanthevaluesreportedbyother

studieswithpureSDEHPsurfactant[37][50][51].ThedramaticraiseintheCMCimpliesthatthe

presenceofironsulfatedidhaveanimpactonthebehaviourofthesurfactantlikelyduetothe

precipitationofferroussaltsofDEHP.Inotherwords,thedissolvedironfromtheironsulfate

areactingasadditionalsurfacesforthesurfactantmolecules.Ontheotherhand,theCMC

impliedincurve2isclosertothereportedliteraturevalue.Theadditionalsurfacesprovidedby

theironparticlesdelayedtheformationofemptymicellesandthusshouldbetakenaccount

whenformulatingforthestabilizationofNZVIsuspension.Overall,theresultofCMCbasedon

thestrategyproposedbyWangetal.impliesthattheminimumSDEHPinitialsynthesis

concentrationshouldbewellabove57mMwiththesurfactantadsorptionbeingconsidered.

62

Figure3.4Surfacetensionmeasurementsof1g/LofNZVIdissolved:Curve1showsthe

surfacetensionmeasurementoftheoriginalSDEHPconcentrationandCurve2displayedthe

correctedconcentrationofSDEHP.

ThesecondpartofdetermininganoptimalconditionforsuspendingNZVIistoidentifySDEHP

formulationsthatcanholdthehighestNZVIconcentration.Totalorganiccarbon(TOC)wasused

todeterminetheconcentrationofun-adsorbedSDEHPinthemixtureofironsulfateand

SDEHP.TOCcananalyzethedissolvedSDEHPinthesolutionconsideringthatthesurfactantis

theonlyorganicmaterialintheformulation.Figure3.5showstheequilibriumsurfactant

concentrationasafunctionoftheaddedSDEHPmixingwithironsulfateconcentrationsfor

variousaddedSDEHPconcentration(initial).30mMSDEHPwith1g/LFedissolvedfromferrous

sulfatewasusedasthe‘worstcasescenario’forstability/synthesiscomparison.100mMSDEHP

0

5

10

15

20

25

30

35

40

45

50

1 10 100

SurfaceTensionm

N/M)

SDEHPConc.(mM)

SurfaceTensionofSDEHPw/1g/LofFeSO4dissolved

Curve1:Original Curve2:"Corrected"

63

isthehighestconcentrationtestedduetotheexponentialraiseinviscositybeyond100mM.As

expected,anegativecorrelationisobservedbetweenthefreesuspendingSDEHPconcentration

andtheamountofironsulfatedissolved.Inotherwords,thehigherconcentrationsofiron

sulfateparticlesprovidemoresurfacesforsurfactantmoleculestoadsorbon,thusgivingless

suspendingSDEHP.ConnectingbacktotheSDEHPCMCfindingsof41mMfromsurfacetension

measurement,stablesuspensionsareexpectedtobefoundatironsulfateconcertationfrom

0.3to1.5g/Lfor80and100mMSDEHPand0.3to1g/Lironsulfateconcentrationsfor50mM

SDEHP.ConsideringthattheamountoffreesuspendingSDEHPareenoughtoformempty

micellesandmeettheminimumstandardoftheframework.

Figure3.5DissolvedSDEHPequilibriumconcentrationwithironVS.addedironsulfate

concentrationsfordifferentinitialSDEHPconcentrations.

0

20

40

60

80

100

120

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

DISSOLVED

SDE

HPCONC.(M

M)

CONCENTRATIONOFFE2SO4

100mMSDEHP 80mMSDEHP 50mMSDEHP 30mMSDEHP

64

3.3.2SynthesisResultsandStabilityofFeSO4-basedSDEHPNZVIat100mMand1g/L

Figure3.6TEMimagingof10mMSDEHP-stabilizedNZVIat1g/Lwithdifferentscaleat500

nmscale.

The100mMSDEHP-stabilizedNZVIat1g/LwasidentifiedasthemostsuccessfulNZVI

suspensionwithastabilityofover2months.Table2summarizedthelistofcandidatesthat

wereselectedforNZVIsynthesiswiththegoalofachievingprolongedcolloidalstability.Figure

3.6displaystheTEMimagingofthemostsuccessfulSDEHP-stabilizedNZVIformulationand

furtherdiscussedinsection3.3.4.

65

a. b.

c. Figure3.7SetA,TimelapsephotosofSDEHP-stabilizedNZVIat0.5g/LofNZVIatvarious

SDEHPconcentrations:a.30mMofSDEHPb.50mMofSDEHPandc.100mMofSDEHP.

Figure3.7and3.8demonstratetwosetsofstabilitytimelapsephotos.Infigure3.7,thefirstset

(setA)ofphotos,30,50and100mMofSDEHPNZVIformulationsaredisplayedandmonitored

forstabilityover2days(48hours).Fromfigure3.7,ashypothesized,theNZVIparticles30mM

formulationwasquicklyaggregatedandformedsedimentationwithin30minutesuponre-

suspensionandsynthesis.The100mMNZVIformulation,withfreesuspendingSDEHP

concentrationofcloseto100mM,demonstratedprolongedstabilityinsuperiortotheother

twocandidates.Nosignsofsedimentationsandstable,evendistributionoftheparticleswere

observed48hoursafterre-suspension.Thesuspensionremainedstabletwomonthsafterre-

66

suspensionandphaseseparationwasobservedduetooxidation.Thehighconcentrationsof

theSDEHPsurfactantalongwiththeadditionoftheNZVIparticlesformwormlikemicellesdue

tothelowpackingandthenatureoftheSDEHPsurfactant[52].Thewormlikemicellescan

anchorontothesurfaceofthenanoparticlesandprovideprolongedstability[53].Signsofthe

formationofwormlikemicellescanbeobservedthroughtheformationofliquidcrystal[54]in

the100mMSDEHPNZVIformulationandnotobservedin30and50mMincontrast(not

shown).Inaddition,anincreaseinviscosityisalsoasigninformationofthewormlikemicelle

asdemonstratedinTable2.Basedontheabove,100mMSDEHPNZVIformulationisselected

forfurtherconcentrationanalysisduetothesuperiorstabilityat0.5g/L.

a.

67

b.

Figure3.8SetB,Stabilitytimelapsepictureof100mMatNZVIconcentration1,1.5and2g/L

overaperiodof24hours:a.1hourandb.24hoursaftersynthesisandre-suspension.

30mM0.5g/L

50mM0.5g/L

100mM0.5g/L

100mM1g/L

100mM2g/L

Stability <1hour ~1hour >2months >2months ~1hourHydrodynamicSize(nm)

482+/-150 792+/-89 287+/-26 244+/-30 412+/-22

Viscosity(cP) - - 1.4+/-0.03 1.4+/-0.03 1.2+/-0.09pH - - 8.8 9.3 7.8

Table3.2SynthesisresultandcharacterizationofSDEHP-stabilizedNZVIatvariousNZVIand

surfactantconcentrations.

Figure3.8showsthestabilityresultsof100mMNZVIat1,1.5and2g/L.Outofthethree,2g/L

showedaggregationandsedimentationupon1hourafterre-suspension,while1.5g/Llasted

justoveronehour.The1g/LofNZVIformulationdemonstratedidenticalstabilityasthe0.5g/L

100mMNZVIformulationdiscussedpreviously.Theresultsareconsistentwiththefactthat

increasingtheironsulfateconcentrationreducesdissolvedSDEHP(figure3.5),limitingitsability

toformtheworm-likemicellerequiredtostabilizetheparticles.

68

InadditionaltotheironsulfatebasedNZVIsynthesis,ferricchloride-basedNZVIwasalso

attemptedfordeterminingastableNZVIstableformulation.Ferricchloride-basedNZVIat

identicalconditionsastheironsulfate-basedNZVIdidnotdemonstratethesamestability.

Rapidaggregationandsettlingwereobservedwithin45minutesto1houraftersynthesis.

DetailedresultsarepresentedanddiscussedinappendixA.

3.3.3pHandViscosityAnalysisandImplication

Viscosityismeasureduponsynthesistosomeoftheformulationsbasedontheirstability.

ViscosityofthestabilizedNZVIsuspensionneedstobemonitoredbecauseitcancritically

impactthemobilitywhenitistoolowortoohigh[41][16][55],asdiscussedinchapter2.The

viscosityofthesecondsetsofsynthesis,withdifferentNZVIconcentrationsalongwith0.5g/L

100mMSDEHPformulationfromthefirstsetweremeasured.TheoriginalSDEHP,without

additionofironsulfateandanysynthesisreactions,heldasimilarviscositytowaterat100mM,

approximately1cP.Uponsynthesis,theviscosityof1and0.5g/L100mMSDEHPNZVIshared

similarviscosity,approximately1.4cP.Thisisslightlyhigherthantheviscosityoftheoriginal

SDEHPsolutionby0.4cPat100mM.Theraiseinviscositycanbeexplainedbytworeasons:1.

Thecontributionoftheadditionofinorganicsolidnanoparticlesandcolloidtothesolution[53]

and2.Theformationofwormlikemicelle[56].Incontrast,the1.5g/Lalthoughdemonstrates

lowerstability,themeasurementswereconductedimmediatelyuponresuspensionbutwitha

lowerviscositymeasuredby0.2cP.Itisexpectedthattheincreaseofnanoparticleswill

increasetheviscosity;however,thisisnotthecasehere.Thisimpliesthattheformationof

wormlikemicellemaycontributetotheviscosityovertheadditionoftheparticles.Theviscosity

resultimpliesthepresenceofwormlikemicellesin0.5and1g/LSDEHP-stabilizedNZVIthat

69

demonstratedhighstability.

ItisbelievedthatpHishighlycorrelatedwiththestabilityduetotheionicinteractionsbetween

thesurfacestabilizersandthehydrogenions[57].Usingzeta-potentialasanindicator,several

studiesreportedthatmodifiedmetalnanoparticlessuspensionexperiencedadecreasein

stabilityatpH6-7[12][23][55][43].Inotherwords,ahigherpHisexpectedforahighly

stabilizedsuspension.pHisalsoavariableoftencontrolledinotherstudiesanddeemedto

promoteimpactinreactivitywithlowerpHs[56][57].ThepHresultsaresummarizedinTable

2.Inthisstudy,0.5,1and2g/Lof100mMSDEHP-stabilizedNZVIformulationsaretestedfor

pH.Theoriginal100mMSDEHPsolutionisalsotested.FortheSDEHPonlypHmeasurement,

thepHisabout11implyingabasicenvironmentduetotheexcessNaOHusedinsynthesis.In

0.5and1g/Lsamples,thepHvariesbetween8.8-9.2,showingadecreasefromthepure

surfactant.Thisisbecausethesynthesisreaction,aslistedintheexperimentalsectionwill

produce2molesofhydrogenionsforeverymoleofNZVIproduced.Thehydrogenionsundergo

aneutralizationreactionwiththebasicsurfactantsolution,causingadropinthepH.Onthe

otherhand,2g/LsampleshowedalowerpHof7.8.Thisisconsistentwiththeotherdata,

consideringthathigherconcentrationofNZVIproducesmorehydrogenions,thusmore

neutralization.Theraiseinhydrogenionscanalsodisturbtheformationofwormlikemicelleor

theformationofemptymicellesingeneral,causinginstabilityofthesuspensionasobserved.

3.3.4SizeAnalysisofSDEHPNZVI

ControloftheparticlesizeduringsynthesisiscriticaltoasuccessfulNZVIsurfacestabilizer

[30][3],[14],[61],asmentionedinthethreestandards[3].Thestabilizersmodifiedorinduced

ontothesurfaceoftheNZVIwillincreasetheoverallparticlesizeandinfluencemobility.

70

Specifically,higherparticlesizeswillincreasecontactwiththesandgrainandpromote

mechanismssuchasporestrainingandnegativelyimpactthemobility[40][20].Itwasreported

thatstrainingiscommonamongpolymer-stabilizedNZVIandmorelikelytooccurinfinerpore

sizeswithlargerparticlesizes[21][40].ItisimportantfortheSDEHPNZVItobeintheoptimal

transportsizebetween100-1000nm[61][62].Thecombinationofhydrodynamicandparticle

diametercanprovidethefullimageoftheparticlesizeofSDEHP-stabilizedNZVI[63].The

hydrodynamicdiameterprovidesthemeasurementofthefullparticlesizeofNZVIparticles

includingtheSDEHPanchoring.TheTEMimagesprovidemeasurementsoftheparticles

withoutthewormlikemicelleanchoring.Thehydrodynamicdiameterof0.5,1and2g/LSDEHP-

stabilizedNZVIaresummarizedinTable2.Thehydrodynamicsizesareabout278and244nm

for0.5and1g/L100mMSDEHP-stabilizedNZVI,respectively.Asignificantincreasein

hydrodynamicsizewasobservedat2g/L,holdinganaverageof412nmindiameter.Thesizeof

thenanoparticleisbelievedtoreflectthestabilityofthesuspension.Specifically,athigher

concentrations,inthiscase,2g/L,thelackofthepresenceofwormlikemicellecontributesto

therapidaggregationofthenanoparticles,causingtheaveragediametertobemorethan1.7

timeshigherthanthestableNZVIsuspensions.Conveniently,thehydrodynamicdiameterof

thestableSDEHP-NZVIsuspensionfellintotheoptimalsizeoftransport[61].TEMimagingof

theparticleforthemoststableformulation,1g/L100mMSDEHPisshowninFigure3.6.The

picturedemonstratedthattheNZVIparticleswithoutthesurfactantcoatingisabout100nm,

meetingthesizestandardproposedintheliterature.Surprisingly,net-likecoatingsare

observedintheTEMimagingaroundtheNZVInanoparticles.Thecoatingisbelievedtobethe

surfactantcoatingofthenanoparticlesisalsoobservedaroundthenanoparticles,givinga

71

largerdiameterofapproximately300nm,whichisconsistenttothehydrodynamicdiameter.

Overall,theparticlesizeofthe1g/LSDEHP-stabilizedNZVIhighlystableformulation

demonstratedsuitablesizebythestandards.

3.3.5MobilityofSDEHPNZVIat100mMand1g/L

Figure3.9showsthebreakthroughcurveof1g/Liron,100mMSDEHP-stabilizedNZVI

transportingatthefieldflowrateof5m/daywithfittingfromthemodeldescribedinchapter2.

Thefittedmodelincludedthedispersionmechanismbutnoparticleattachmentmechanism.In

otherwords,particleattachmentsarenotapplicableinthetransportofSDEHP-stabilizedNZVI

duetothegoodtransport,thisisthefirstsignofgoodtransport.Theperformanceofthe

breakthroughcurveisanindicationontheperformanceofthemobilityofthesynthesized

NZVI-suspension.TwofeaturesonthebreakthroughcurveshowthatthehighstabilitySDEHP-

stabilizedNZVIholdsanexcellentmobility:1.Highrecoveryand2.Highandsteadyplateau

concentrationpeak.TherecoveryiscalculatedthroughtheratiooftheinjectedNZVIpore

volumeandtherecoveredporevolumeasthefollowing:

~8Hs�4�;4p~834��5434p

(2)

Therecoveryiscalculatedtobeashighas95%.Thehighrecoveryisbetterthanthevalue

reportedbyLinetal.,achievingarecoveryofalmost90%atahigherflowvelocityof15.87

m/day[29].ComparingtotheemulsionNZVIresult,therecoveryisinthesamerangeas

SDEHP-stabilizedNZVIandtheflowvelocityisconductedatlowervelocityof0.4m/day[22];

however,mechanicalstirringisintegratedfortheemulsifiedNZVI.Theresultsarecomparable

72

tothepolyelectrolyte-stabilizedNZVIreportedbyRaychoudhuryetal.atacomparableflow

rateof6.4m/day[40].However,SDEHP-stabilizedNZVIachievedthehighrecoveryandplateau

valuesathigherironconcentrationswithoutconstantmxingduringinjection.SDEHP-stabilized

NZVIcouldachieveahighrecoveryatafieldflowratewhilenomechanicalstirringisneeded,

thisexcellentperformancewasneverreportedbyotherNZVIsuspension.Thehighandsteady

plateaurecovery(C/Co)isascloseas1,thisismuchbetterthanotherstudiesaswell.

Specifically,carboxyl-methylcelluloseNZVItransportstudyconductedbyKocuretal.achieved

asteadyplateauconcentrationat0.8and0.9C/Co,at0.1and2.5g/LNZVIconcentration,

respectively[13]at4m/day.Comparingtothisstudy,SDEHP-stabilizedNZVIachievedaslightly

higherC/CoatalowerNZVIconcentrationof1g/Latacomparableof1.5m/day.Thealmost

perfectplateauconcentrationalsoimpliedthatminimumretentionoccurredbetweenthesand

andtheNZVIparticles.TheminorresidualoftheNZVIinthecolumnisfoundonlyinthebottom

ofcolumnfromtheextractedsand,believedtobecausedbygravitationalsedimentationthat

wasobservedseverelyinotherstudies[64][28].Itcanbeconcludedbasedontheabovetwo

featuresthat,thedevelopedSDEHP-stabilizedNZVIcansuccessfullytransportthroughporous

mediawithoutdifficulty.

73

Figure3.9ColumnstudybreakthroughcurveofhighlystableSDEHP-stabilizedNZVI,at100

mMSDEHPand1g/LofNZVIat5m/daywiththemodeldescribedinchapter2(solidline).

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4

C/Co

Porevolume

74

3.3.6NZVIColumnDistributionAnalysis

Figure3.1NZVIdistributionincolumngraphfor100mMSDEHP1g/LNZVIwithmicroscope

imagingat3,9and15cmsectionofthecolumn.

Figure3.10illustratesthedistributionofNZVIparticlesuponcompletionofalltheinjectionand

flushingintheexperimentalphase.Althoughthebreakthroughcurvefromfigure3.9implieda

highNZVIrecovery,tracesofNZVIretentionswereshownfromthedistributioncurveandthe

sandgrainmicroscopeimaging.Overall,theNZVIshowedadecreasingretentiontrendfromthe

bottomtothetopofthecolumn.Thetrendofthecolumndistributionprofileisconsistentwith

thefindingsreportedbyXinetal.[26].Itisimportanttonotethatdespitethesimilarityinthe

overallNZVIdistributiontrend,theamountofironretainedinthesandreportedinthisstudy

wasmagnitudesloweroftheresultbyXinetal.atsimilarconcentrationsandamuchlower

75

flowvelocity[26].Comparingtotheirondistributioninchapter2,theresultsindicatedisabout

1timeslower,althoughconductedattheconcentrationinthisstudyis5timesless.Inother

words,figure3.10confirmsthecompellingtransportabilityofSDEHP-stabilizedNZVIinporous

media.Inadditional,visuallynolargeparticleswerefoundaroundtheporeinthesandanalysis,

concludingstrainingwastheleastlikelyretentionmechanism.

Comparingamongthethreemicroscopepicturesalongthecolumn,sometraceofNZVIwas

remainedonsandbutonlyatthebottomsectionofthecolumn(3cm).Thisisconsistentwith

thefindingsoftheretentionprofilethathighestNZVIwasfoundinthebottom.However,the

amountoftheNZVIremainedatthebottomofthesandwasconsideredminimalincomparison

tothefindingsinchapter2,thetransportofmicroemulsionironoxide.Thiscouldbe

interpretedasasignofgoodtransport,becausetherewasnegligibleformationofNZVI

blockageamongthesandgraininthecolumn.ItisbelievedthatthestrainedNZVIparticles

wereonlythelargerparticleswhensynthesized.Inshort,itcanbeconcludedbasedonthe

sandgrainimagingthatSDEHP-stabilizedNZVIsuspensiondidnotaggregateandsettlewhen

contactwiththesand.

3.3.7ImplicationsforinsituRemediation

Theultimategoalofthisstudyistodevelopasurfacestabilizerforafull-scaleremediation.

Basedonthefunctionalityandpropertiesofthestabilizationmethod,SDEHP-stabilizedNZVIis

expectedtobethebestcandidateoutofallthesurfacestabilizationtechniquesthusfar.In

termsoffunctionality,SDEHP-stabilizedNZVIsuspensioncanbeinjectedataflowvelocity

similartotheremediationvelocitywithoutanymechanicalstirring.TheSDEHP-stabilizedNZVI

holdsahighstabilitywithnosedimentationobservedforovertwomonths,thisgreatly

76

improvesthechanceforasuccessfulinsituremediation.InthelatestfieldstudiesofCarboxyl-

methylcellulose-stabilizedNZVI,itisreportedthatnodirecttraceofNZVIwasfoundin

downstream,implyingagainthetransportissue.SDEHP-stabilizedNZVIdemonstratedideal

transportin1-DcolumnstudywithresultsthatweresuperiorcomparewithotherNZVI

suspensions.Inadditionaltothefunctionality,thepropertiesoftheappliedsynthesismethod

arealsoinfavourofafull-scaleremediation.ItiswidelyacceptedthatonsiteNZVIsynthesis

yieldsbetterremediationresultsincomparisontopre-synthesizedNZVI[3][30][62].One-pot

synthesisNZVItechniqueprovidesasimpleprocedureofsynthesizingSDEHP-stabilizedNZVI

consideringitssimplemethod.Lastly,surfactantSDEHPisachemicalthatiswidelyusedinthe

miningindustrywithnohistoryofenvironmentalconcerns[51].Thephosphategroupcanbe

naturallydegradeduponcontactwithsoil.Theamountofsurfactantusedintheformulation,

100mM,isabout3.5wt%andconsideredalowconcentrationcomparingtoemulsionNZVI,

thusnosideeffectonthetoxicity,whichisconcernofinsituNZVIremediation[65].Itis

concludedthatSDEHPisagoodtransportvehicleforNZVI,morestudiesinvolvingreactivityis

requiredandeventuallyapilot-scaleinsituremediationisneeded.

3.4Conclusion

ThisworksuccessfullydevelopedaSDEHP-NZVIformulationusingtheconceptproposedby

Wangetal.ThroughthedeterminationofCMC,therangeofthemoststableformulationis

narroweddown.TheTOCresultsfurtherprovidedarangeofiron-surfactantratioforsynthesis.

Thesynthesisandstabilityresultssuggestedthatthehighstabilitycanbeachievedbywormlike

micellesanchoringontheNZVIparticlesurfaceswhenthedissolvedSDEHPconcentrationis

closeto100mM.Inthemobilitystudy,becauseofthehighstability,1g/Liron,100mMSDEHP-

77

stabilizedNZVIyieldedremarkablemobilityintheporousmediaatafieldflowvelocityof1.5

m/day.ThehighNZVIrecoveryandasteadypeakofthebreakthroughcurveimplythatthe

nanoparticlesexperienceminimalfiltrationduringtransport.Itisimpliedthatthecompelling

mobilityinporousmediademonstratedbySDEHP-stabilizedNZVIcanleadtostudiesatthe

higherscale.

78

3.6References:

[1] A.B.Cundy,L.Hopkinson,andR.L.D.Whitby,“Useofiron-basedtechnologiesincontaminatedlandandgroundwater

remediation:Areview,”Sci.TotalEnviron.,vol.400,no.1–3,pp.42–51,2008.

[2] A.CorreiadeVelosaandR.F.PupoNogueira,“2,4-Dichlorophenoxyaceticacid(2,4-D)degradationpromotedby

nanoparticulatezerovalentiron(nZVI)inaerobicsuspensions,”J.Environ.Manage.,vol.121,pp.72–79,2013.



[4] F.Fu,D.D.Dionysiou,andH.Liu,“Theuseofzero-valentironforgroundwaterremediationandwastewater




[6] T.Raychoudhury,G.Naja,andS.Ghoshal,“Assessmentoftransportoftwopolyelectrolyte-stabilizedzero-valentiron

nanoparticlesinporousmedia,”J.Contam.Hydrol.,vol.118,no.3–4,pp.143–151,2010.

[7] T.Phenrat,A.Cihan,H.-J.Kim,M.Mital,T.Illangasekare,andG.V.Lowry,“TransportandDepositionofPolymer-

ModifiedFe0Nanoparticlesin2-DHeterogeneousPorousMedia:EffectsofParticleConcentration,Fe0Content,and

Coatings,”Environ.Sci.Technol.,vol.44,no.23,pp.9086–9093,2010.



[9] N.SalehandH.Kim,“IonicStrengthandCompositionAffecttheMobilityofinWater-SaturatedSandColumns,”pp.

3349–3355,2008.

[10] D.B.Nyer,Evank.Vance,“Nano-ScaleIronforDehalogenation.pdf,”Treat.Technol.,2001.

[11] T.Phenrat,N.Saleh,K.Sirk,H.J.Kim,R.D.Tilton,andG.V.Lowry,“Stabilizationofaqueousnanoscalezerovalentiron

dispersionsbyanionicpolyelectrolytes:Adsorbedanionicpolyelectrolytelayerpropertiesandtheireffecton

aggregationandsedimentation,”J.NanoparticleRes.,vol.10,no.5,pp.795–814,2008.



57,2007.


79

vol.145,pp.17–25,2013.

[14] T.Phenrat,H.J.Kim,F.Fagerlund,T.Illangasekare,R.D.Tilton,andG.V.Lowry,“Particlesizedistribution,

concentration,andmagneticattractionaffecttransportofpolymer-modifiedFe0nanoparticlesinsandcolumns,”

Environ.Sci.Technol.,vol.43,no.13,pp.5079–5085,2009.



[16] N.Sakulchaicharoen,D.M.O’Carroll,andJ.E.Herrera,“Enhancedstabilityanddechlorinationactivityofpre-synthesis

stabilizednanoscaleFePdparticles,”J.Contam.Hydrol.,vol.118,no.3–4,pp.117–127,2010.

[17] F.He,D.Zhao,J.Liu,andC.B.Roberts,“StabilizationofFe-Pdnanoparticleswithsodiumcarboxymethylcellulosefor

enhancedtransportanddechlorinationoftrichloroethyleneinsoilandgroundwater,”Ind.Eng.Chem.Res.,vol.46,no.

1,pp.29–34,2007.



[19] W.Wang,M.Zhou,Z.Jin,andT.Li,“Reactivitycharacteristicsofpoly(methylmethacrylate)coatednanoscaleiron

particlesfortrichloroethyleneremediation,”J.Hazard.Mater.,vol.173,no.1–3,pp.724–730,2010.



[21] S.M.HosseiniandT.Tosco,“Transportandretentionofhighconcentratednano-Fe/Cuparticlesthroughhighlyflow-

ratedpackedsandcolumn,”WaterRes.,vol.47,no.1,pp.326–338,2013.




vol.155,pp.9–19,2013.

[24] M.Basnet,C.DiTommaso,S.Ghoshal,andN.Tufenkji,“Reducedtransportpotentialofapalladium-dopedzerovalent

ironnanoparticleinawatersaturatedloamysand,”WaterRes.,vol.68,pp.354–363,2015.

[25] S.Laumann,V.Micić,andT.Hofmann,“Mobilityenhancementofnanoscalezero-valentironincarbonateporous

mediathroughco-injectionofpolyelectrolytes,”WaterRes.,vol.50,pp.70–79,2014.


80




Chem.Eng.J.,vol.262,pp.813–822,2015.



[29] Y.H.Lin,H.H.Tseng,M.Y.Wey,andM.DerLin,“Characteristicsoftwotypesofstabilizednanozero-valentironand

transportinporousmedia,”Sci.TotalEnviron.,vol.408,no.10,pp.2260–2267,2010.



vol.48,no.5,pp.2862–2869,2014.



[32] L.Chekli,B.Bayatsarmadi,R.Sekine,B.Sarkar,A.M.Shen,K.G.Scheckel,W.Skinner,R.Naidu,H.K.Shon,E.Lombi,

andE.Donner,“Analyticalcharacterisationofnanoscalezero-valentiron:Amethodologicalreview,”Anal.Chim.Acta,

vol.903,no.August,pp.13–35,2016.

[33] S.R.KanelandH.Choi,“Transportcharacteristicsofsurface-modifiednanoscalezero-valentironinporousmedia,”

WaterSci.Technol.,vol.55,no.1–2,pp.157–162,2007.

[34] Y.T.Wei,S.cheeWu,S.W.Yang,C.H.Che,H.L.Lien,andD.H.Huang,“Biodegradablesurfactantstabilizednanoscale

zero-valentironforinsitutreatmentofvinylchlorideand1,2-dichloroethane,”J.Hazard.Mater.,vol.211–212,pp.

373–380,2012.

[35] D.weiLiang,Y.hanYang,W.weiXu,S.kanPeng,S.fuLu,andY.Xiang,“Nonionicsurfactantgreatlyenhancesthe

reductivedebrominationofpolybrominateddiphenylethersbynanoscalezero-valentiron:Mechanismandkinetics,”J.

Hazard.Mater.,vol.278,pp.592–596,2014.

[36] G.D.Sayles,G.You,M.Wang,andM.J.Kupferle,“DDT,DDD,andDDEdechlorinationbyzero-valentiron,”Environ.


[37] Y.Luan,G.Xu,S.Yuan,L.Xiao,andZ.Zhang,“Comparativestudiesofstructurallysimilarsurfactants:Sodiumbis(2-

ethylhexyl)phosphateandsodiumbis(2-ethylhexyl)sulfosuccinate,”Langmuir,vol.18,no.22,pp.8700–8705,2002.

81

[38] E.Wang,Ziheng;Choi,FrancisandAcosta,“EffectofSurfactantsonZero-ValentIronNanoparticles(NZVI)Reactivity.”.

[39] Y.Sahoo,H.Pizem,T.Fried,D.Golodnitsky,L.Burstein,C.N.Sukenik,andG.Markovich,“Alkyl

phosphonate/phosphatecoatingonmagnetitenanoparticles:Acomparisonwithfattyacids,”Langmuir,vol.17,no.25,

pp.7907–7911,2001.

[40] T.Raychoudhury,N.Tufenkji,andS.Ghoshal,“Strainingofpolyelectrolyte-stabilizednanoscalezerovalentiron

particlesduringtransportthroughgranularporousmedia,”WaterRes.,vol.50,pp.80–89,2014.

[41] M.Arhuoma,M.Dong,D.Yang,andR.Idem,“Determinationofwater-in-oilemulsionviscosityinporousmedia,”Ind.

Eng.Chem.Res.,vol.48,no.15,pp.7092–7102,2009.

[42] H.H.Tseng,J.G.Su,andC.Liang,“Synthesisofgranularactivatedcarbon/zerovalentironcompositesfor

simultaneousadsorption/dechlorinationoftrichloroethylene,”J.Hazard.Mater.,vol.192,no.2,pp.500–506,2011.


2,pp.7–21,2006.





[46] S.Quraishi,M.Bussmann,andE.Acosta,“CapillaryCurvesforEx-situWashingofOil-CoatedParticles,”J.Surfactants

Deterg.,vol.18,no.5,pp.811–823,2015.

[47] D.K.Kim,MariaMikhaylova,YuZhang,andandMamounMuhammed,“ProtectiveCoatingofSuperparamagneticIron

OxideNanoparticles,”Chem.Mater.,vol.15,no.8,pp.1617–1627,2003.

[48] L.Bronstein,X.Huang,J.Retrum,A.Schmucker,M.Pink,B.D.Stein,andB.Dragnea,“Influenceofironoleatecomplex

structureonironoxidenanoparticleformation,”Chem.Mater.,vol.19,no.15,pp.3624–3632,2007.

[49] Y.C.Lee,C.W.Kim,J.Y.Lee,H.J.Shin,andJ.W.Yang,“Characterizationofnanoscalezerovalentironmodifiedby

nonionicsurfactantfortrichloroethyleneremovalinthepresenceofhumicacid:Aresearchnote,”Desalin.Water

Treat.,vol.10,no.1–3,pp.33–38,2009.

[50] Y.Luan,G.Xu,S.Yuan,L.Xiao,andZ.Zhang,“InvestigationsonNaDEHPandAOT:Computersimulationandsurface

tensionmeasurements,”ColloidsSurfacesAPhysicochem.Eng.Asp.,vol.210,no.1,pp.61–68,2002.

[51] C.Akbay,N.Wilmot,R.A.Agbaria,andI.M.Warner,“Characterizationandapplicationofsodiumdi(2-ethylhexyl)

82

sulfosuccinateandsodiumdi(2-ethylhexyl)phosphatesurfactantsaspseudostationaryphasesinmicellarelectrokinetic

chromatography,”J.Chromatogr.A,vol.1061,no.1,pp.105–111,2004.

[52] A.ShioiandM.Tanabe,“EffectsofOrganicSolventsontheAggregates’GeometryandWinsorII/IIITransitioninthe

SodiumBis(2-ethylhexyl)PhosphateSystemAkihisa,”J.Phys.Chem.,no.97,pp.8281–8288,1993.

[53] F.Nettesheim,M.W.Liberatore,T.K.Hodgdon,N.J.Wagner,E.W.Kaler,andM.Vethamuthu,“Influenceof

nanoparticleadditiononthepropertiesofwormlikemicellarsolutions,”Langmuir,vol.24,no.15,pp.7718–7726,

2008.

[54] X.Wang,D.S.Miller,E.Bukusoglu,J.J.DePablo,andN.L.Abbott,“Topologicaldefectsinliquidcrystalsastemplates

forMolecularSelf-Assembly,”Nat.Mater.,vol.15,no.September,pp.1–9,2015.

[55] M.M.Krol,A.J.Oleniuk,C.M.Kocur,B.E.Sleep,P.Bennett,Z.Xiong,andD.M.O’Carroll,“Afield-validatedmodelfor

insitutransportofpolymer-stabilizednZVIandimplicationsforsubsurfaceinjection,”Environ.Sci.Technol.,vol.47,no.

13,pp.7332–7340,2013.

[56] M.E.Helgeson,T.K.Hodgdon,E.W.Kaler,N.J.Wagner,M.Vethamuthu,andK.P.Ananthapadmanabhan,“Formation

andrheologyofviscoelastic‘doublenetworks’inwormlikemicelle-nanoparticlemixtures,”Langmuir,vol.26,no.11,

pp.8049–8060,2010.

[57] K.Moore,B.Forsberg,D.R.Baer,W.a.Arnold,andR.L.Penn,“Zero-ValentIron:ImpactofAnionsPresentduring

SynthesisonSubsequentNanoparticleReactivity,”J.Environ.Eng.,vol.137,no.10,pp.889–896,2011.

[58] R.a.French,A.R.Jacobson,B.Kim,S.L.Isley,R.L.E.E.Penn,andP.C.Baveye,“Influenceofionicstrength,pH,

andcationvalenceonaggregationkineticsoftitaniumdioxidenanoparticles,”Environ.Sci.Technol.,vol.43,no.5,pp.

1354–1359,2009.

[59] J.Chen,Z.Xiu,G.V.Lowry,andP.J.J.Alvarez,“Effectofnaturalorganicmatterontoxicityandreactivityofnano-scale

zero-valentiron,”WaterRes.,vol.45,no.5,pp.1995–2001,2011.

[60] T.Satapanajaru,P.Anurakpongsatorn,P.Pengthamkeerati,andH.Boparai,“Remediationofatrazine-contaminated

soilandwaterbynanozerovalentiron,”Water.Air.SoilPollut.,vol.192,no.1–4,pp.349–359,2008.

[61] J.Zhan,T.Zheng,G.Piringer,C.Day,G.L.Mcpherson,Y.Lu,K.Papadopoulos,andV.T.John,“Transportcharacteristics

ofnanoscalefunctionalzerovalentiron/silicacompositesforinsituremediationoftrichloroethylene,”Environ.Sci.

Technol.,vol.42,no.23,pp.8871–8876,2008.

[62] B.Sunkara,J.Zhan,J.He,G.L.McPherson,G.Piringer,andV.T.John,“Nanoscalezerovalentironsupportedon

83

uniformcarbonmicrospheresfortheinsituremediationofchlorinatedhydrocarbons,”ACSAppl.Mater.Interfaces,

vol.2,no.10,pp.2854–2862,2010.

[63] T.Phenrat,N.Saleh,K.Sirk,R.D.Tilton,andG.V.Lowry,“Aggregationandsedimentationofaqueousnanoscale

zerovalentirondispersions,”Environ.Sci.Technol.,vol.41,no.1,pp.284–290,2007.

[64] E.DVecchia,MLuna,andR.Sethi,“Transportinporousmediaofhighlyconcentratedironmicro-andnanoparticlesin

thepresenceofxanthangum,”Environ.Sci.Technol.,vol.43,no.23,pp.8942–8947,2009.



84

Chapter4:ConclusionandRecommendations

Stabilityiscommonlyconcludedtobepositivelycorrelatedwiththemobilityofnanoscalezero

valentiron(NZVI)particles[1][2].ImprovingthestabilityofbareNZVIsuspensionsusingsurface

stabilizerscanpromotethemobilityofNZVIintheaquiferandeventuallyasuccessful

remediation[3].Currently,varioussurfacemodifiedNZVIsuspensionshavebeendeveloped

andtestedinbenchscale[3][4]andindustrialscale[5][6][7];however,aggregationand

sedimentationsarestillobservedandnosuccessfultransportreported.Overall,thisstudy

concludedtwoalternativetransportvehiclestoimprovethemobilityandstabilityofNZVI.

AlthoughNZVIremediationapplicationshavebeenconductedthroughacademicfieldstudies,

theextentofNZVIapplicationincontaminatedlandsisstilllimited.Inspecific,NZVIinsitu

remediationduetomobility,isfavouredinsoilwithhigherhydraulicconductivity.Theresults

fromthisstudyimpliedtwoalternativeNZVItransportvehicleandpotentiallyimprovethe

extentofNZVIremediation.

Inchapter2,microemulsion-stabilizednanoparticlesaresuggestedtobeanalternative

transportvehicleforNZVIinsituremediationbyWangetal.[8].Toassesstheintrinsicmobility

andeffectivenessofmicroemulsion-stabilizedNZVIdevelopedbyWangetal.[8],ironoxide

nanoparticlesareusedasananalogytoeliminatetheoxidizingfactorofNZVI.Inthecolumn

experiment,microemulsion-stabilizedironoxidenanoparticlesatveryhighconcentrations(2.5,

5and10g/L)arecharacterizedintermsofviscosity,hydrodynamicsize,particlesizeand

stability.Itisfoundthatmicroemulsion-stabilizedironoxidenanoparticlesholdidentical

stabilityperiodwithmicroemulsion-stabilizedNZVIwithsizesinthesimilarrangeasconfirmed

85

bythesizeanalysis.However,therheologyanalysisimpliedthathigherironoxide

concentrations,10g/Land7g/L,hadahighviscosityofover400CPanddemonstratednon-

Newtonian,shear-thinningbehaviour.However,theviscositydrasticallydecreasedat5g/Liron

oxideconcentrationto20CPandreturnedtoNewtonianbehaviour.Itwasfoundthatthehigh

viscosityat10g/Lcontributedtothepoortransportobservedatthefieldof5m/day.In

contrast,underthesamecondition,at5g/Lironoxideformulation,highrecoveryandhigh

plateaupeakwasobserved.Inaddition,bychangingthesalinityconditionofthesandinthe

columnbetweenDIwatersaturatedandbrinesaturated(10g/L)atthesameconcentrationas

theformulation,thetransportresultischanged.Byshiftingthesalinitytozero,theplateau

peakisdecreasedsharply,implyingthehighsaltsensitivityofthemicroemulsion-stabilizer.

Microemulsion-stabilizedironnanoparticlesalthoughdemonstratedpromisingtransportresults

duetothehighstability,furtherimprovementsarerequiredforreducingthesensitivityofsalt

andviscosity.Itisconcludedthatmicroemulsioncanholdupto5g/Lofironoxide

nanoparticleswhiledemonstratingadequatetransportbehaviour.Viscosityandsalinitywere

identifiedascriticalvariablesforimprovingthefeasibilityoflargerscaleremediationstudies.

BasedonthefindingsinChapter2,surfactantSDEHP-stabilizedNZVIisdevelopedbasedona

formulationdesignframeworksuggestedbyWangetal.Chapter3discussedthedevelopment

ofahighlystabilizedanionicsurfactantstabilized-NZVIandtheassessmentonthemobilityof

theNZVIformulation.Inthedevelopment,formulationdesignframeworkbyWangetal.is

appliedforthefirsttimeinsurfactant-stabilizedNZVI[9].Anionicsurfactant,SDEHPisselected

forNZVIstabilizationbasedonthepackingpropertiesandenvironmentalfriendliness.Through

determiningcriticalmicelleconcentration(CMC)andtheratiobetweenSDEHPandironsulfate,

86

1g/LNZVIat100mMSDEHPisidentifiedasthemostoptimalformulationyieldingastabilityof

over2months.TheNZVIformulationischaracterizedwithsizeandviscosityanalysis;withthe

resultsshowingsignsofidealNZVI.Thecolumnstudyatfieldvelocityresultsdemonstrated

excellenttransportofSDEHP-stabilizedNZVIwithhighrecoveryandplateaupeaks.The

developedformulationwasevaluatedforfieldstudypotentialandhighlightedtobeastrong

candidateforremediationduetotheoverallpropertiesofSDEHP-stabilizedNZVI.

DespitethefindingsinChapter3impliedthattheSDEHP-stabilizedNZVIishighlyfeasiblefora

fullscaleremediationapplication,somefutureworkisrecommendedforfurtherunderstanding

andoptimizingtheformulation.Firstly,theunderstandingofthereactivityoftheSDEHP-

stabilizedNZVIisconsideredpreliminary.Itwasunderstoodfromthepreviousreactivitystudy

fromWangetal.thatsodiumdodecylphosphate(SDP)-stabilizedNZVIduetothephosphate

group(PO43-)holdskineticsratewithreactiveblack5(RB5)andcarbontetrachloride(CT)close

totherateofbareNZVI[10].Inotherwords,thephosphategroup,thatpresentsinSDEHPas

well,maypromotethereactivitywithotherchlorinatedsolvents,makingSDEHP-stabilizedNZVI

moreefficient.Withthisimplication,areactivitystudyandtargetdeliverystudyis

recommendedforfuturework.Secondly,thestabilizationmechanismandthebehaviourof

surfactantaggregationarenotyetwellunderstoodintheSDEHP-stabilizedNZVIformulation.

Eventhoughsignsshowingthattheformationofwormlikemicelleplayedanimportantrolein

stabilization,theexactmechanismonthenanoscaleisstillhypothesized.Theformationof

wormlikemicelleswasnotfoundintheidenticalNZVIformulationwhentheinitialironsource

isfromironchloride(FeCl3).AsdemonstratedinAppendixB,thestabilitybehaviourofFeCl3

basedNZVIformulationistotallydifferentthantheironsulfatebasedformulationdescribedin

87

Chapter3.Finally,theconcentrationoftheSDEHP-stabilizedNZVI(1g/L)discussedinChapter3

despitesufficientintermsofremediation,couldnotbescaledupduetothelimitationofthe

formulation.Themicroemulsion-basedNZVIisanimprovementoftheformulationtestedin

Chapter2andabletoholdhigherNZVIconcentrations.However,furtherunderstandingin

termsoftransportandmobilizationwithDNAPLisrequired.

Despitethestrongmobilityinporousmediawasdemonstratedbymicroemulsion-stabilized

ironnanoparticlesandSDEHP-stabilizedNZVI,furtherimprovementsarerecommendedfor

bothstabilizers.Formicroemulsion-stabilizedironnanoparticles,thehighsalinityofthe

formulation(10g/100mL)mustbeaddressedtopreventsaltcontaminationtolandsbefore

applications.Thehighsaltconcentration,asdiscussedinchapter2,wasnecessarytoproducea

one-phasemicroemulsion,wherethehydrophilic-lipophilic-differenceiszero(HLD=0)withthe

extendedsurfactantused.Loweringthesaltconcentrationwiththeexgtendedsurfactantwill

shifttheHLDoftheformulationandthuslosingthestability.BasedontheHLDequation,

alternativesurfactantswithahighercharacteristiccurvature(Cc)canreducethesalinity.

Phosphatesurfactantfromchapter3,SDEHPmayreducethesalinityrequiredtobalancethe

HLD,aSDEHPmicroemulsionsystemcanbedevelopedfollowingtheprocedureofWantetal.

FortheSDEHP-stabilizedNZVIdevelopedinchapter3,eventhoughsuccessfultransportwas

observedat1g/LNZVIconcentrationandmettheminimumremediationcriteria,higherNZVI

concentrationsarepreferredformoreefficientremediation.However,basedontheresultsin

chapter3,1g/ListhemaximumNZVIconcentrationpossible,consideringanysurfactant

concentrationhigherthan100mMwasnotfeasible.Basedontheresultsinchapter2,

microemulsionsystemscanholdironoxidenanoparticlesupto10g/L.TheSDEHP

88

microemulsionsystemcanachievehigherconcentrationsandcanbeexpectedwithvariationof

thesynthesisprocedure.

Thesignificanceofthisstudyliesinthefollowing:1.ThetransportstudyontheSDEHP-

stabilizedNZVIconfirmedanimprovementonthemobilitytoporousmedia.Currently,thefull-

scaleremediationresultsofpolymer-stabilizednZVIsuggestedthatlongerdistanceandhigher

recoveryarerequired.ItcanbeexpectedthatSDEHP-stabilizedNZVIyieldedalongertravel

distanceandhigherrecoveryinthefieldtestfromtheresultsinthisstudy.2.Themobilitystudy

ofmicroemulsion-stabilizedironnanoparticlessuggestedthatgreaterconcentrationsofNZVI

canbetransportedinaneffectivefashion.MostofthereportedNZVItransportvehiclescan

onlyholdupto2.5g/Lofiron[1],whichlimitedthecosteffectivenessoftheremediationin

highlycontaminatedlands.MicroemulsioncanbeappliedasanalternativestabilizerforNZVIat

contaminantssitethatarenotsuitableforthecurrentNZVItechnology.

Forclosure,thisstudysuccessfullyassessedexaminedtwoNZVIstabilizationsystems.The

SDEHP-stabilizedNZVIisanimprovementbasedonthetransportimplicationsfromthe

microemulsion-stabilizedironoxidenanoparticles.Thisworkemphasizedtheimportanceof

stabilitytoNZVItransportintheporousmediaandhighlightedSDEHP-stabilizedNZVIasa

potentialcandidateforfuturegroundwaterremediation.

89

4.2References:[1] C.M.Kocur,D.M.O’Carroll,andB.E.Sleep,“ImpactofnZVIstabilityonmobilityinporousmedia,”J.Contam.Hydrol.,

vol.145,pp.17–25,2013.





[4] F.Fu,D.D.Dionysiou,andH.Liu,“Theuseofzero-valentironforgroundwaterremediationandwastewater






vol.48,no.5,pp.2862–2869,2014.







[10] E.Wang,Ziheng;Choi,FrancisandAcosta,“EffectofSurfactantsonZero-ValentIronNanoparticles(NZVI)Reactivity.”.

90

AppendixA–FerricChlorideBasedSodiumDiethylHexylPhosphate

(SDEHP)-stabilizedNZVI

A.1Introduction

NZVIcanbeartificiallysynthesizedfromtwomainsources,ferricchloride(FeCl3)andiron

sulfate(FeSO4)[1].Bothsourcesofiron,namelyferricsourceandsulfatesource,undergoa

reductionreactionwithNaBH4,reducingfromiron(II)sulfateandiron(III)chloridetothezero-

valentform,aslistedinthefollowing:

(1) 2𝐹𝑒Uw + 3𝐵𝐻qA + 3𝐻W𝑂 → 2𝐹𝑒a + 𝐻W𝐵𝑂UA + 4𝐻w + 2𝐻W

And

(2) 4𝐹𝑒Uw + 3𝐵𝐻A + 9𝐻W𝑂 → 4𝐹𝑒a + 3𝐻W𝐵𝑂A + 12𝐻w + 6𝐻W

respectively.Fromtheliterature,bothironsourceswereappliedextensivelyassynthesis

sourcesforNZVIstudieswiththeironsulfatebeingslightlyhigher[1]whilethechloridesource

wasthemostoriginalmethod.Itisalsoindictedthatironsulfateisconsideredamore

environmentalfriendlysourceofNZVIincomparisontoferricchloride[1].Inbothcases,bare

NZVIparticlesaresynthesizedwithidenticalpropertiessuchasinstabilityandatsizeswithin

similarrange.

Forimprovingthemobility,surfacemodifiersarecommonlyaddedtobareNZVIsuspensions.In

general,surfacemodifiersareappliedtoNZVIintwoways:1.Re-suspension[2]and2.

Synthesiswiththepresenceofthesurfacemodifiers[3].Inthere-suspensiontypesurface

modifications,thesourceoftheNZVImayplayanegligibleimpacttotheoverallsuspension;

however,someimpactsareexpectedwhentheNZVIsynthesistakesplacewiththemodifiers.

91

Inspecific,thereactionsproductsfrombothferricandsulfatemethodmayinteractwiththe

surfacemodifiersandtheoverallsuspensiondifferently.Currently,therehasnostudies

focusedonthestabilityandmobilitydifferencebetweentheNZVIfrombothsources.In

chapter3,ironsulfate-basedNZVIparticlesweresynthesizedusingthe“one-pot”technique

andanalyzedatvarioussurfactantandNZVIconcentrations.AnNZVIformulationat100mM

SDEHPwasreportedtodemonstrateaprolongedofover2monthswithexcellentmobilityin

porousmedia.ItisobservedthatwithdifferentsourcesofNZVIappliedinthe“one-pot”

synthesismethoddescribedinthisthesis,differentqualityofNZVIsuspensionswereobtained.

Theobjectiveofthisworkistoinvestigatethebasicproperties,suchasstabilityandsizesof

ferricchloride-based,SDEH-stabilizedNZVIformulationsincomparisontotheiron-sulfate

basedformulation.Thesecondaryobjectiveistodeterminethecausebehindthedifferences

betweenchlorideandsulfatesynthesisinSDEHP-stabilizedNZVI.

A.2Methodology

Unlessotherwisespecified,allproceduresandmaterialsareconductedandpreparedatroom

temperatureandconditions.

A.2.1PreparationofSurfactantSDEHPandFeCl3-basedNZVI

TheprocedureofsynthesizingSDEHPisdescribedinsection3.2.1andadaptedfromprocedure

ofLuanetal.[4]TheNZVIwassynthesizedinthesame“one-pot”synthesisfashionasdescribed

insection3.2.2withthefundamentalideaadaptedfromWantetal.[5]

A.2.2FormulationdesignofFeCl3-basedNZVI

Surfacetensionmeasurementwasappliedtodeterminethecriticalmicelleconcentration

(CMC)ofSDEHPwith1g/LNZVIequivalenceofferricchloridedissolved.Themeasurementwas

92

conductedwithatensiometeratdifferentconcentrationsofSDEHP.Detailedprocedureis

describedinsection3.2.4.

A.2.3CharacterizationAnalysis:SizeandStability

Transmissionelectronicmicroscope(TEM)anddynamiclightscattering(DLS)areusedto

analyzeparticleandhydrodynamicsizes,respectively.Samplepreparationforbothsizeanalysis

isdescribedinsection2.3.4.Forstabilityanalysis,identicalprocedurewasfollowedas

describedinsection2.3.3.

A.3ResultsandDiscussions

A.3.1FormulationDesignImplication

BasedontheformulationdesignframeworkproposedbyWangetal.,theconcentrationofthe

surfactantmustbehigherthantheCMCtoachievehigherstability[6].Theapproachwas

reportedsuccessfulwithprolongedstabilityinthesulfate-based,SDEHP-stabilizedNZVIat1

g/L,100mMsurfactantconcentration.FigureA1showsthesurfacetensionmeasurementsat

differentconcentrationofSDEHP.TheCMCisidentifiedatabout35mMbasedonthe

interpretationofline-of-the-best-fit.TheCMCoftheferricchloridedissolvedSDEHP(35mM)is

lowerthanironsulfatedissolvedSDEHP(57mM).Thisimpliesthebindingofthesurfactantand

ironaredifferentinthetwocasesduetothepresenceofanions,chlorideorsulfate.Inthecase

ofthesulfatemethod,thehigherCMCcanbeexplainedbythepossibilitythatthesurfactants

aremorelikelytobindtothesurfaceoftheiron,leavinglessfreesurfactantstoformempty

micelles.Ontheotherhand,thehigherCMCinthechloridemethodimpliestheopposite,the

surfactantislesslikelytobindontothesurfaceoftheiron,thusmorefreesurfactantsare

availabletoformmicellesatthesameconcentrations.TherootcauseofthedifferenceinCMC

93

ishowever,unclear,itissuspectedthatthepresenceofthedifferentionsorthedifferencein

theironcharges.Thisisexpectedtohavesomeimpactstothestabilityaftertheproductionto

theformulation.

FigureA1.FigureA1.SurfacetensionmeasurementsofSDEHPatvariousconcentrationwith

ironchloride(1g/LequivalenceofNZVI)dissolved.

A.3.2StabilityandSizeAnalysis

FigureA2.showsthestabilitytimelapsephotoat1hourofchloridemethodbased,SDEHP-

stabilizedNZVIat30,50and100mM.Allthethreeconcentrationsexperiencedaggregation

andsedimentafteronehouruponsynthesis.Outofthethree,100mMSDEHPdemonstrated

lesssettlingthanthe30and50mM,implyingthathigherconcentrationofSDEHPmayimprove

15

20

25

30

35

40

45

50

1 10 100

SurfaceTension(m

N/M)

SDEHPConc.(mM)-LogScale

SurfaceTensionofSDEHPwith1g/LFeCl3Dissolved

94

thestabilityinthecaseofchloridebasedmethod.Incontrast,atthesameconcentrationof

NZVIandSDEHP,sulfatemethoddemonstratedastabilityofover2months.Thedifferencein

thesynthesiscanberelatedtotheCMCdifferencediscussed,wheremolecularlythe

interactionsbetweentheanions,surfactantandironaredifferentbetweenthesulfateand

chloridemethod.Furthermore,thereactionfromthechloridemethodproducesabout3times

morehydrogenionsthanthesulfatemethod.ThismeansthatthepHlevelofthechloride

methodNZVIsuspensioncanbealotlowerthanthepHofthesulfatemethod.Thedecreasein

pHisreportedtoimpactthezetapotentialthuscontributiontolowerstability.

Forsizeanalysis,figure3showstheTEMimagingoftheNZVIparticlesizeofthesynthesized

NZVIat100mM.Overall,thehardparticlesizeofthechloridebasedNZVIisabout100-200nm,

whichisinthesamerangetothereportedvalueinChapter3.However,thesurfactantcoating

aroundtheNZVIisalotmoreemphasizedincomparisontotheTEMimagingforthesulfate

method.Thisisreflectedinthehydrodynamicdiameterofthechloridemethod-basedNZVIas

well.Assummarizedintable3.2,thehydrodynamicdiameterfor30,50and100mMSDEHPare

400,420and380,respectively.Theyarefarlargerthanthesulfatemethod-basedNZVIas

reportedinChapter3.Thedifferenceinhydrodynamicsizemayimpliedthedifferencein

molecularinteractionbetweenthechlorideandthesurfactant.Thiscouldbeanimplicationon

themechanismforsurfactantsuspendingandtheformationofwormlikemicelle.

95

[email protected]/Lofiron

concentration.1houraftersynthesis.

A.4FutureWorks

Differentstabilitybehavioursareobservedforchloride-andsulfate-basedNZVIunderthesame

synthesiscondition.Thisappendixprovedthatwithdifferentsourcesofironanddifferent

reactionpathways,synthesiswiththepresenceofsurfacemodifiers.Fromthestudy,itis

shownthatthesulfatemethodismoreinfavouroThiscouldbeanimplicationon

improvementsofotherNZVIsurfacestabilizationsandmobilityinporousmedia,onecan

improvetheabilityoftheNZVIsuspensionbychangingthesourceofsynthesisiron.Itis

96

importanttoalsonotethatwhetherthetwoironsourceswilldemonstrateanydifferencein

reactivityduetothedifferenceinsynthesisreactions.

97

A.5References:


2,pp.7–21,2006.



Chem.Eng.J.,vol.262,pp.813–822,2015.

[3] S.Z.Yu,Y.Cheng,X.F.Fan,andL.P.Xu,“PreparationofCoatedCMC-nZVIUsingRheologicalPhaseReactionMethod

andResearchonDegradationofChloroforminWater,”Mater.Sci.Forum,vol.847,pp.230–233,2016.

[4] Y.Luan,G.Xu,S.Yuan,L.Xiao,andZ.Zhang,“Comparativestudiesofstructurallysimilarsurfactants:Sodiumbis(2-

ethylhexyl)phosphateandsodiumbis(2-ethylhexyl)sulfosuccinate,”Langmuir,vol.18,no.22,pp.8700–8705,2002.


vol.155,pp.9–19,2013.



98

AppendixB:ComparisonbetweenCarboxylmethyl-celluosestabilized

ironoxidenanoparticleswithmicroemulsion-stabilizednanoparticles

B1.Background:Carboxylmethyl-celluose(CMC)isbyfarthemostsuccessfulNZVIstabilizerintheliterature.The

mobilityofCMC-stabilizedNZVIhadwelltestedinthelabscaleandfullscaleremediationwas

launchedin2014.Inthissection,amobilitycomparisonbetweenCMC-stabilizedironoxideand

microemulsionironoxidewastestedfollowingidenticalexperimentalsetupdescribedin

chapter2.TheformulationofCMC-stabilizedNZVIwasmodifiedfromtheonedescribedby

Kocuretal.(2.5g/Lofiron,CMCMW=90,000at0.8wt%)[1]withNZVIbeingreplacedbyiron

oxideforconsistencyofcomparison.2.5g/Lequivalenceofironoxidewassuspendedinthe

50%dilutedmicroemulsionformulationasdiscussedinchapter2.

B2.Results:

B2.1.StabilityThestabilityofmicroemulsionandCMC-stabilizedironoxideat2.5g/Lwerecomparedusing

timelapsephotoscomparisons.AsindicatedinfigureB1,CMC-stabilizedironoxidestartedto

showsignsofaggregationandparticlesettlingafter15hours.Majorsedimentationwas

observedatthebottomofthevialforCMC-stabilizedironoxideafter80hours.Ontheother

hand,nosignificantsettlingwasobservedwithmicroemulsionironoxide,consistenttothe

99

findingsinchapter2.Thisindicatedthatmicroemulsionironoxideholdsamuchstronger

stabilitythanCMCironoxide.Theminimumsettlingandaggregationimpliedthat

microemulsion-basedironsuspensionexperienceasmoothertransportprocessinsandin

comparisontoCMC-basedironparticles.

A.

FigureB1.A.TimelapsedphotosofCMCandmicroemulsionstabilizedironoxideat2.5g/L.B.

EvidenceofsettlingofCMCironoxideafter80hoursuponsuspension.

100

B2.2.MobilityComparisonTransportstudywasconductedonbothmicroemulsionironoxideandCMCironoxideat2.5

g/L.Forthemicromeulsionironoxidetransportat2.5g/L,thetransportresultwasreproducible

comparingtothetransportresultsinchapter2.ForCMCironoxide,despitesuccessful

completionoftheexperimentanddataanalysis,abreakthroughcurvewasnotabletobe

obtained.However,fromthepostanalysessuchaspressuredropmonitoringandsandgrain

analysis,itcanbeconfirmedthatmicroemulsionironoxidedemonstratedbettertransportthan

CMCironoxide.

Pressuredropwasrecordedateveryporevolumeduringtheflushingstage(foratotalof5

porevolumes).FigureB2demonstratedthepressuredropmoniroingresultcomparison

betweenCMCandmicroemulsionironoxide.Thepressuredropmonitoringresultshowsthat

CMCironoxidestartedwithahighpressuredropof22.5psiandlinearlydecreasedto3psi

whilemicroemulsionironoxideremainedat0.5psithroughouttheflushingstage.Thisimplies

thatduringthetransportofCMCNZVI,potentialcloggingofthesandporesincreasedthe

difficultyofflowinthecolumn,thusexperiencingahighpressuredropinthebeginning.The

decreaseinthepressuredropmayimplythatthecloggedparticleswerebeingflushedoutof

thecolumnandthusinferredreversibleadsorptionoftheironparticles.Ontheotherhand,

microemulsionironoxidedemonstratednegligiblepressuredropintheflushingstage,implying

minimumcloggingandretentionoftheparticles.

101

FigureB2.Comparisonofpressuredropmonitoringresultsatthepost-flushingstagebetween

CMCandmicroemulsionironoxide.

SimilarironretentionanalysiswasconductedafterthecolumnstudyforCMCand

microemulsionironoxideasdemonstratedinfigureB3.FigureB3showsconsistentresultsand

conclusionasfigureB2,whereCMCironoxideexperiencedmoreretentionanddifficulties

duringtransport.InfigureB3,CMCironoxidedemonstratedadecreasingtrendintheiron

retentionanalysissimilartotheresultsofscheduleBinchapter2.Whereasmicroemulsioniron

oxidedemonstratedreproducibleresultstoscheduleA.Thisagainconfirmsthehypothesisthat

duetoinstabilityoftheCMCironoxide,moreretentionofironwasobserved.

0

5

10

15

20

25

13 14 15 16 17 18 19 20

PressureDrop(Psi)

PoreVolumeInjected

PressureDropMonitoring- PostFlushingStage

ME CMC

102

A.

B.

FigureB3.Iron-sandgrainanalysiswithmicroscopepicturesforA.Microemulsionironoxideat

2.5g/LandB.CMCironoxideat2.5g/L.

103

B3.References:


vol.145,pp.17–25,2013.

Transport and Development of Microemulsion- and Surfactant ... · Microemulsion-Stabilized Iron Nanoparticles in Porous Media Dennis Hsu Masters of Applied Science Graduate Department

Documents