A, Physicochemical and Engineering Aspects, v. 461, pp. 212-219.
Colloidsand Surfaces A:Physicochemical and EngineeringAspects
jo u r n al ho me p ag e :w w w . e l s e v i e r . c o m / l o c a t e / c o l s u r f a
InvestigationofFeCl3inducedcoagulationprocesses using
electrophoreticmeasurement,nanoparticletrackinganalysis and dynamiclight scattering: ImportanceofpH andcolloid surfacecharge
OlenaOriekhova1,SergeStoll∗
UniversityofGeneva,EarthandEnvironmentalScienceSection,F.-A.ForelInstitute,GroupofEnvironmentalPhysicalChemistry,10RoutedeSuisse, CH-1290Versoix,Switzerland
h i g h l i g h t s
•Zetapotentialoflatexparticlesisa functionofFeCl3 coagulant concen-tration.
•Measurementofzetapotentialallows controllingthelatexcoagulationand
•FeCl3optimaldosageforcoagulation isdependentontheinitialpH suspen-sion.
g r a p h i c a l a b s t r a c t
a r t i c l e i n f o
Articlehistory:
Received5June2014
Receivedinrevisedform30July2014 Accepted31July2014
Inwatertreatmentprocesses,theoptimaldosageofcoagulantishighlydependentonsuspendedparticle surfacecharge,sizeandconcentration,pHandcompositionofwater.Onewaytocontrolthecoagulation processcanbebasedonthemeasurementoftheelectrophoreticmobilityanddeterminationofzeta potential.
Inthisstudyweinvestigatedtheinteractionbetweennegativelychargedpolystyrenelatexparticles andiron(III)chlorideascoagulant.Wecombinedthreemethods,i.e.dynamiclightscattering,nanoparticle trackinganalysis,andmodelingtothoroughlycharacterizeoursystem.
Wehaveshownthatstabilizationofzetapotentialoccurredafter60–80minafteradditionofcoagulant.
WedemonstrateddifferentbehaviorsoflatexparticleswithFeCl3dependingonthedosageofironions.
TheoptimaldosageofFeCl3isequalto1–2mg/Lfortherapidaggregationof10mg/Llatexsuspension.
Wefoundagoodagreementbetweentheaggregationrateandsurfacechargeofthelatexparticlesand thatchargeneutralizationmechanismisresponsibleforparticleaggregation.Highdosageofcoagulant wasalsofoundtoresultinformationofiron(III)hydroxideparticleswhichdiameterwasabout200nm.
TheinitialpHisalsoimportantforlatexparticlecoagulation.ThelowerinitialpHofsuspensionis,the morerapidlytheisoelectricpointisachieved.
©2014ElsevierB.V.Allrightsreserved.
∗Correspondingauthor.Tel.:+41223790333;fax:+41223790302.
E-mailaddresses:Olena.Oriekhova@unige.ch(O.Oriekhova),Serge.Stoll@unige.ch(S.Stoll).
1 Tel.:+41223790332.
http://dx.doi.org/10.1016/j.colsurfa.2014.07.049 0927-7757/©2014ElsevierB.V.Allrightsreserved.
The coagulation process is widely used in water treatment plantstoremovethecolloidalorsuspendedmatterwhichcanbe bothofnaturalandanthropogenicorigin[1–7].Duringthisprocess colloidalparticlesformlargeaggregatesthatcanbemoreeasily andrapidlyremovedbyflocculationorfiltration.Destabilization andincreaseofcollisionefficiencybetweenthesuspendedcolloidal particlesareachievedbyaddingsyntheticcoagulantsintowater [5,8].Differenttypesofelectrolytesorpolyelectrolytesareused, forexamplethesaltsofiron(III)oraluminum,andpoly(dimethyl diallylammoniumchloride)orpolyacrylamidepolymers[5,9–12].
Theliquidphase needstoberapidlyseparatedwiththehighest efficiencytoobtainaclearfiltrate,avoidrejectingpollutantsinto naturalaquaticsystems,andmaintainthemaximumofpollutants intheminimumdrymatter.Forseveralreasonscoagulantsarenot alwaysusedinarationalwayforoptimalcoagulationcondition.
Thisoptimalconditionisverydependentondifferentfactors,such astheconcentrationofparticles,theirchargesandsizes,pHand solutioncompositionwhichcangreatlyvarywithtime.
Therearemanymechanismsthatcanexplainthecoagulation process. Action of inorganic salts is based on the charge neu-tralizationmechanism. Therepulsiveforces betweennegatively charged colloid particles disappear after the adsorption of the highlychargedcations,suchasAl3+orFe3+,andneutralizationof surfacecharge.Theelectrostaticinteractioncanalsobescreened bythecationsthat leadtotheprevalenceofattractivevander Waals forces. However, dependingon thepH solutionand the dosageofcoagulant,hydrolysisoccursandinsolublehydroxides areformed[6,13].Itleadstothedevelopmentofbiggeraggregates whichcapturecolloidalparticlesandthentosweepcoagulation [5,14,15].Lietal.[14]studiedkaolincoagulationbyaluminum sul-fate(alum).Theyshowedthatthechargeneutralizationoccurred whenzetapotentialofparticleswasclosetozeroandthe concen-trationofalumwas0.1mmol/L.Thesweepcoagulationtookplace athigheralumdoseofabout1.9mmol/Lwhichcaused precipita-tionofamorphousmetalhydroxide.Jamesetal.[16]studiedthe colloidalTiO2-Al(III)systemanddemonstratedthechargereversal coagulationwhenzetapotentialofsuspensionwasintherange from−14 to+14±4mVat pHfrom 6 to10 for different elec-trolyteconcentrations.Forlargermagnitudeofzetapotentialthe suspensionwasfoundstable. Kobayashiet al.[15] investigated thecoagulationofsulfatelatexbeadsinthepresenceofimogolite nano-tubes.Theyexplained,whentwomechanisms,charge neu-tralizationandsweepcoagulationarepresent,thatcoagulationis stronglychargedependent.Whenthesurfacechargesoflatexand imogoliteshavetheoppositesignandtheimogolitesaredispersed, thecoagulationhappenedonlyaroundtheisoelectricpoint.When thelatexparticlesandtheimogoliteshavethesamechargeand whentheimogolitesarecoagulatedthenthesweepcoagulation occurred.Theauthorsfoundthattheefficiencyofcoagulationwas determinedbytheimogolitedosageanditselectrokineticpotential.
Itwasalsofoundthattheefficiencyofcoagulationprocesscan bereducedwiththeincreaseofcoagulantdosage[17]duetothe reversalofparticlesurfacechargehenceresultinginanew sta-bilization [18–20]. Dahlsten et al. [21] studied the behavior of melamine-formaldehydelatexparticlesinthepresenceofawide rangeofelectrolytes(NaCl,NaNO3,KNO3,etc.).Theyshowedthat thechangeofzetapotentialfrompositive(from+40to+80mV) tonegative(about−40mV)valueswasdependingonthe concen-trationofelectrolytesandpH.Schumacheretal.[22]demonstrated thattheanionicstyrene-acryliclatexparticlessuspendedwith non-ionicsurfactantscanstaystableinthepresenceoftrivalentions withhighrangeofconcentration.Firstthelatexdispersionwas stabilizedwithsurfactantandthentheironandaluminumsalt solu-tionswereadded.Theparticlezetapotentialincreasedfromless
counterionconcentrationsupto200days.Larueetal. [8] com-paredtwotypesofcoagulantdosing,chemicalandelectrical,as afunctionofpHandironconcentrationtodefinetheoptimal oper-atingconditions.Theychangedtheconcentrationofironfrom10−3 to6×10−3mol/Landshowedthattheresidualflocconcentration decreasedwiththeincreaseofironconcentrationandreacheda minimalvalueat2×10−3mol/Lcorrespondingtooptimaldosage atpH7.5forFeSO4 andatpH6forFeCl3coagulant.Inallcases itwasfoundthatthecoagulantdosageisanimportantparameter thatinfluencesthecoagulationprocess.
Oneofthedifferentandpossiblewaytocontrolthecoagulation processesistomeasurethechangeofparticlechargeduringthe watertreatmentprocessasshowninreferences[23–25].Itcanbe donebyelectrophoreticmeasurementswhichallowthecalculation ofthezetapotentials.Zetapotentialgivesusefulinformationabout electrostatic interactionsbetweenparticles.Theelectrochemical surfacepropertiesofparticlescontroltheaggregationkineticsand theinteractionsbetweenparticlescanbedescribedbytheDLVO theory[26].ItwasshownbyMorfesisetal.[25]thatthemonitoring ofzetapotentialinrealwatertreatmentplantscanhelpto main-taintheoptimaloperatingconditions.Theycontrolledtheparticle stabilitywhichvariedovertime,andidentifiedthemaximum coag-ulationratethatgaveafastresponsetochangingcircumstances.
Inthis workweconductedelectrophoreticmeasurementsto studythebehaviorofnegatively chargedpolystyrenelatex par-ticlesinthepresenceofiron(III)chloride,whichisalargelyused coagulant,atdifferentdosages.TheeffectoftheinitialpHofthe suspension as wellas pH changes were investigated to better understandthecoagulationprocessandsurfacecharge modifica-tionoflatexparticlesinthepresenceofFeCl3.Thedynamiclight scattering(DLS)methodwasusedtodeterminezetapotentialand z-averagehydrodynamicdiameteroflatexparticles.Employingthe nanoparticletrackinganalysis(NTA)wealsoinvestigatedthe par-ticles distributionbysize. Themodeling of iron speciationwas madewiththeMINTEQA2software.Wecombinedthesedifferent approachestothoroughlycharacterizeoursystemanddiscussed thecoagulationmechanismbasedondetailedanalysisofsize dis-tributionsandelectrophoreticmeasurements.
2. Materialsandmethods 2.1. Experimentalmethods
Zetapotentialand z-averagehydrodynamic diameterwere measuredusingaMalvernZetasizerNanoZS(MalvernInstruments Ltd,UK).Visualizationanddistributionofparticlesbysizewerealso investigatedapplyingnanoparticletrackinganalysis(NTA)witha NanoSightLM14instrument(NanoSightLtd,UK).
2.1.1. DLSmethod
Five parallel measurements were performed for each point with time delay of 5s. The samples temperature was 298◦K.
FirstelectrophoreticmobilityUEwasmeasuredandzetapotential wascalculatedusingHenryequation(Eq.(1))andSmoluchowski approximation(Eq.(2)):
UE=2ε
3 ·f(Ka), (1)
UE=ε
, (2)
where ε corresponds to the relative permittivity εr (or dielec-tric constant) multiplied by the permittivity of free space, ε0; is the viscosity of the liquid; K is the Debye–Hükel
Parameter Value
Temperature,◦K 298
ConcentrationofFeCl3,mg/L 2
pH 1–12
Ionicstrength,mol/L 3.6×10−5to6.2×10−2a
apHdependent,calculated.
parameter;aistheparticleradius,f(Ka)equalto3/2.Wechose theSmoluchowskiapproximationasoursystemfitsthe parame-tersofSmoluchowskimodel[7]regardingtheparticlesizesand suspensionionicstrength.Theinvestigatedparticleshave diame-terequalto0.99mandtheconcentrationofelectrolyteismore than10−3MofFeCl3.Insertingvaluesforpermittivityandviscosity ofwaterat298◦K,Eq.(3)givestherelationbetweenzetapotential andmobility:
=12.8UE, (3)
where-potentialisexpressedinmVandelectrophoreticmobility inms−1/Vcm−1.
TheStokes–Einsteinequation(Eq.(4))wasusedtocalculate hydrodynamicdiameterdHofparticlesfromthetransitional dif-fusioncoefficientD[5,7]:
dH= kT
3D, (4)
wherekistheBoltzmann’sconstantandTistheabsolute temper-ature.
2.1.2. NTAmethod
ToobtaintheparticlesizedistributionsweusedaNTALM14 instrument withNTA 2.3 Analytical Software. The device anal-ysestheparticlepathsunderBrownianmotionand determines theaveragedistancemovedbyeachparticleinxandydirection.
Thisvalueallowstoobtainthediffusioncoefficientandusingthe Stokes–Einsteinequation(4) tocalculatethesphere-equivalent, hydrodynamicdiameter[27,28].Theresultsofmeasurementsare highlydependentoftheprocessingparameters and experimen-talprotocol[27,29].Allmeasurementswererepeated3times,i.e.
3videowererecordedforeachsample.Weadjustedthecamera settingstovisualizeasmanyparticlesaspossibletryingtoreduce thenoise on theimage. Asour sampleswere polydisperse we establishedbiggercapturetime(from90to160s)totrack max-imumnumberofparticlesandgotreliableresults.TheNTALM14 instrumentwasadjustedandcalibratedbeforemeasurementofour sampleswithstandardpolystyrenelatexmicrospheresof100,200 and400nm.Weextractedthesuitableimagesfromthevideoto comparedifferenttypesofsamples,adjustingonlytheirbrightness andcontrast.
2.1.3. Modeling
Toperform the modeling of iron(III) species in solutionthe MINTEQA2software(developedbyAllisonGeoscienceConsultants Inc. and HydroGeologic Inc.) was used. MINTEQA2 applies the thermodynamicandmassbalanceequationstosolvegeochemical equilibriaandcalculatetheionspeciation/solubility.Theprogram consistsofsubmodelsthatcomputetheactivitiesofcationicand anionicspeciesandneutralionpairsthencomputethesolubility ofsolidsandmineralsandintheendthemasstransfersubmodel calculatesthemassofsolidthatprecipitatesordissolves.[30–32].
Fig.1. ZetapotentialofsulfatelatexparticlesasafunctionofpH.Zetapotentialis foundnegativeinallrangeofadjustedpH.Sizedistribution(inset)oflatexparticles usingDLSmethod.Z-averageofhydrodynamicdiameterisfoundequalto1000nm inagoodagreementwithTEMandNTAmeasurement.
TocomputetheactivitycoefficientsweusedtheDaviesequation (5),theotherparameterswhichwereusedduringmodelingare showninTable1.
wheref± isthemeanmodalactivitycoefficientofanelectrolyte whichdissociatesintoionswithcharge z1 and z2; Iis theionic strength.
2.2. Materials
Latexbeads (IDCLatex particles provided by Life Technolo-giesCorporation,USA)weremadeofpolystyrenewithnegatively charged sulfate functional groups on the surface. They have a diameter0.99m(TEMmeasurement,providedbymanufacturer), initialconcentration78g/L,densityat20◦C1.055g/cm3,specific surfacearea5.7×104cm/gandwerefreefromsurfactants.
Weworkedwith10mg/Llatexsuspensions. Astock suspen-sionof1g/Lwaspreparedand thendilutedwithMilliQwater (R>18Mcm)untilfinalconcentrationof10mg/Lwasachieved.
pHwasadjustedbyaddingsmallamountofdilutedHClandNaOH (Merck,Germany).Zetapotentialofparticleswasfoundnegative inall rangeof pH(Fig.1).Asa result,thesuspensionwas sta-bleduetotheelectrostaticrepulsiveforcesandthez-averageof hydrodynamicdiameterwasfoundalsostableatabout1000nm.
Iron(III)chlorideFeCl3wasusedascoagulant.Astocksolutionof 1g/Lwaspreparedfromiron(III)chloridehexahydrateFeCl3·6H2O (Merck,Germany).
After the dissolution of salt its hydrolysis occurred. Many species,suchasFe3+,Fe(OH)2+,Fe(OH)2+,Fe(OH)3andFe(OH)4−
coexistinsolutionatthesametime(Fig.2).Theconcentrationof thesespeciesdependsonpHandcanbedescribedbyequationsof hydrolysisequilibrium(Eqs.(6)–(10))[13,33,34]:
2 4 6 8 10 12
-Fig.2. Speciationofiron(III)asafunctionofpHforaFeCl3solutionat2mg/L.Fe3+
andFe(OH)2+aremainlypresentinsolutionatpHlessthan3.InthepHrangefrom 4to6thehighestrelativeconcentrationisobtainedforFe(OH)2+andatpHgreater than7Fe(OH)4−andinsolubleFe(OH)3arepresent.
3. Resultsanddiscussion
ThepHofinitial10mg/Llatexsuspensionwas5.5.Itwasthen adjustedto4,7,8and9,toperformtheexperimentsatdifferent initialpHconditions.Theconcentrationofiron(III)chloridewas thenadjustedandvariedforeachsuspensionfrom0.25to15mg/L.
ThechangeoflatexparticlezetapotentialasafunctionofFeCl3 concentrationwithtimeandwithinitialpH5.5isshowninFig.3.
Wefoundthatadditionofcoagulantdonotleadtoimmediate sta-bilizationofthelatexpotential.Itchangedfrom−50to−10mV withthepresenceofsmallamountofsalt(0.25and0.5mg/L),but stillremainednegative.Thecharge neutralizationwasfoundto occurwhenconcentrationofcoagulantwas1–2mg/L.Theincrease ofFeCl3concentrationfrom5to15mg/Lwasthenfoundtoresult inchargeinversion.Zetapotentialbecamehighlypositiveand sta-blewithvaluescomprisedbetween+30and+45mV.ThepHofthe suspensionwasalsofoundtorapidlydecreasewiththeadditionof FeCl3.ApHdecreaseoftwopHunitswasachievedwithcoagulant concentrationof15mg/LasshowninFig.3b.
Togetaninsightintothebehaviorofthesystemafter poten-tialandpHstabilizationthethreelastvaluesof-potentialand pHat80, 100and120minwereconsideredandaveragevalues werecalculated.Theresultsarepresentedin Fig.4.Continuous decreaseofpHand-potentialincreasewithcoagulant concentra-tionareobserved.Withtheincreaseoftheironionsconcentration, pHdecreases from5.5 to3.6. Meanwhile values of -potential changefromnegative−50mVtopositive +40mV,indicatingan importantchargereversaloflatexparticles.Wefoundthatthe iso-electricpoint(IEP)wasachievedwhentheFeCl3concentrationwas comprisedbetween1and2mg/L.
Inordertoinvestigatetheinfluenceoftheinitialsuspension pHtocoagulatewithFeCl3wefixedthepHofinitialsuspensions at4,5.5,7,8and9.Itturnedoutthatthebehaviorofthesystem canbedividedintotworegimesdependingontheconcentration ofiron(III)chloride(Fig.5).WhenthedosageofFeCl3wasabove 5mg/Lwedidnotidentifiedsignificantdifferencesbetweenthe valuesofzetapotentialwhichwereallfoundpositive.Ontheother handwhenFeCl3concentrationwaslessthan5mg/Lweobserved adifferencebetweenthevaluesofzetapotentialasafunctionof initialpH.IndeedtheIEPwasmorerapidlyachievedatlowinitial pHvaluesregardingtheFeCl3 dosage.WhenpHwasabove7, -potentialwaswithintherangefrom−90to−40mV.Butforacid environment,pHbelow5.5,thevaluesof-potentialwerehigher,
b
Fig.3.Latexparticlechangeof-potential(a)andpH(b)dependingonthe concen-trationofFeCl3withtime.Zetapotentialstabilizesafterabout80minandpHafter 10minfollowingtheFeCl3addition.InitialpHofsuspensionwas5.5.Itisobserved thatinallcasesthesuspensionpHisdecreasingwiththeadditionofFeCl3.
from−60to−30mVforthesameFeCl3dosage.Inneutralandbasic environments latex particles had more negativesurface charge thaninacidenvironments.Itisrelatedwithacid–basicequilibrium onthesurfaceoflatexparticle.Inacidenvironmentfreeproton
Fig.4. Changeof-potentialandpHofthelatexsuspensionasafunctionof coagu-lantconcentrationafterstabilization(initialpHwas5.5).-Potentialchangesfrom negative(−50mV)topositive(+40mV)valuesandpHfrom5.5to3.6.The neu-tralizationofchargeisachievedwhentheconcentrationofcoagulantiscomprised between1and2mg/LconcomitantlywithadecreaseofthesuspensionpH.
Fig.5.ZetapotentialandpHvariationasafunctionofiron(III)chloride concentra-tionatdifferentinitialpH.ItisfoundthatbydecreasingtheinitialpH,surfacecharge neutralizationismoreefficientregardingthecoagulantconcentration.Asaresult lesscoagulantisnecessary.
Fig.6.Timevariation ofz-averagehydrodynamicdiametersoflatexparticles dependingoncoagulantdosage.Therearethreetendenciesinparticleaggregation behavior:(i)absenceofaggregationbelowtheisoelectricpoint(IEP);(ii) aggrega-tionattheIEPaccordingtotheincreaseofthez-averagehydrodynamicdiameter and(iii)decreaseofthez-averagediameterabovetheIEPduetotheconcomitant formationofnanoparticlescomposedofFe(OH)3.InitialsuspensionpHequalto7.
environmentzeta-potentialbecomesmorenegative.Toachievethe IEPtheoptimalconcentrationofiron(III)chloridewaswithinthe rangefrom1to2mg/L.Belowthisconcentrationthesurfacecharge ofparticlesstayednegativeandaboveitweobservedthesurface chargeinversionandthezetapotentialbecamehighlypositiveat about+40mV.ItshouldalsobenotedthatatpHaround4the con-centrationofpositivelychargedionssuchasFe(OH)2+andFe(OH)2+
ishigherandsuchconditionsaremoreefficientforsurfacecharge neutralization.
Togetinsightintothecoagulationprocessandaggregate for-mationthatwereoccurringinthesuspensionwemeasuredthe z-average hydrodynamic diameters as a function of the coagu-lant concentration(Fig.6) usingDLSat initialpH 7.We found threetendenciesinsuspensionbehavior.Whentheconcentration ofcoagulantwasbelowtheisoelectric point(0.25 and0.5mg/L FeCl3),nosignificantsizechangewasobtained–linedashanddots (redrectangle).Atchargeneutralization,for(1and2mg/LFeCl3), particleaggregationwasobserved andz-averagehydrodynamic diameterwasfoundtorapidlyincreaseupto2400nmafterabout 100–125min(straightlineandblackrectangle).Inthecharge inver-siondomain,forFeCl3concentrationfrom5to15mg/L,z-average valuesofhydrodynamicdiameterweresurprisinglyfoundsmaller thanexpected and closeto300nm (dashlineand blue rectan-gle).ItisknownthatatpHabove6anduntilpH10about60%of Fe(III)speciesinsolution(Fig.2)areconsideredtoforminsoluble ironhydroxideFe(OH)3 [8,13,35].AftertheformationofFe(OH)3
itpartiallyprecipitatedandstayedinsuspensionindependentlyof furtherchangesofpH,thusexplainingthedecreaseofthemean z-averagevalueofhydrodynamicdiameter.Thisimportant obser-vation waschecked byconsidering thesize distributions using DLS(Fig.7a).Wefoundtwodistinctparticlesizedistributionsin
itpartiallyprecipitatedandstayedinsuspensionindependentlyof furtherchangesofpH,thusexplainingthedecreaseofthemean z-averagevalueofhydrodynamicdiameter.Thisimportant obser-vation waschecked byconsidering thesize distributions using DLS(Fig.7a).Wefoundtwodistinctparticlesizedistributionsin