THESISFORTHEDEGREEOFLICENCIATEOFENGINEERING

TheDiluentEffectontheSolventExtractionofRareEarthElementsfromNeodymium

MagnetLeachate

MARINOGERGORIĆ

IndustrialMaterialsRecycling

DepartmentofChemistryandChemicalEngineeringCHALMERSUNIVERSITYOFTECHNOLOGY

Gothenburg,Sweden2016

TheDiluentEffectontheSolventExtractionofRareEarthElementsfromNeodymiumMagnetLeachateMARINOGERGORIĆ®MARINOGERGORIĆ,2016Technicalreportnr2016:12ISSNnr:1652-943XIndustrialMaterialsRecyclingDepartmentofChemistryandChemicalEngineeringChalmersUniversityofTechnologySE-41296GothenburgSwedenTelephone+46(0)31-7721000ChalmersReproserviceGothenburg,Sweden2016

Thediluenteffectonthesolventextractionofrareearthelementsfromneodymiummagnetleachate

MARINOGERGORIĆ

IndustrialMaterialsRecycling

DepartmentofChemistryandChemicalEngineering

ChalmersUniversityofTechnology

ABSTRACT

Rareearthelements(REEs)havebecomevitalcomponentsinawiderangeofindustrialapplications.Thedemand

fortheREEshasgrownsignificantlyinthelastfewdecades.Thishasledtoincreasingcostsandsupplychainrisk.

Today,despitelowerpricesthanin2011,theyareclassifiedasthehighestsupplyriskelementsintheEU;thus

newincentivesforrecyclingtheREEsoutofelectronicscrapwerebroughtforth.End-of-lifeneodymiummagnets

are a viable source for the recovery of some REEs. Although mainly iron alloys, these materials contain

neodymium,dysprosiumandsmalladmixturesofpraseodymiumandterbium.

Leaching followedby solventextractionof theREEsoutof the leachate is anattractiveandefficientwayof

recycling theseelementsoutofend-of-lifeneodymiummagnets. The issues thatareencounteredalong this

recyclingpathistheseparationoftheREEsfromtheotherelementsthataredissolvedwiththeREEsintothe

leachateandachievinghighseparation factorsbetweentheREEs fromeachother.Extractingagentssuchas

D2EHPA(di(2-ethylhexyl)phosphoricacid)andTODGA(tetraoctyldiglycolamide)havebeenpreviouslyusedfor

achievinggoodseparationoftheREEsunderspecificextractionconditions.

This thesishas focusedon thedevelopmentandoptimizationofREEextraction from real commercialwaste

sources,thenitricacidandsulfuricacidmedialeachatesoftheneodymiummagnetwaste,usingTODGAand

D2EHPAasextractingagents, respectively.SelectiveREEextraction fromthesolutionwithminimalornoco-

extractionofotherelementsintheleachateishopedtoprovideanovelroutetoacommerciallyviablerouteto

recyclableREEproducts.Thecompositionoftheorganicphasewasinvestigatedinordertostudytheeffectof

thediluentontheoverallextractionprocess,awell-knownoptimizationparameter,howeverinfrequentlyused.

Theeffectof thediluenton the separation factorswasalsodiscussedaswell as somecharacteristicsof the

aqueousphaseontheoverallextractionprocess.Thenamedextractantswereusedatvariousconcentrationsin

different diluents like solvent 70, hexane, octane, cyclohexanone, toluene, 1-octanol and chloroform. Both

extractantsdemonstratedgoodselectivityconcerningtheextractionoftheREEsoutoftheneodymiummagnet

wasteleachates.

KEYWORDS:neodymiummagnets,rareearthmetals,recycling,solventextraction,diluents,TODGA,D2EHPA

LISTOFPUBLICATIONSTheworkpresentedinthisthesisisbasedontheworkcontainedinthefollowingpapers:PaperIMarino Gergorić, Christian Ekberg, Britt-Marie Steenari, Teodora Retegan. Separation ofheavy rare-earths from light rare-earths via solvent extraction from neodymium magnetleachate–diluenteffect.PaperIIMarinoGergorić, Christian Ekberg,Mark Foreman, Britt-Marie Steenari, Teodora Retegan.CharacterizationandleachingoftheneodymiummagnetwasteandsolventextractionoftherareearthelementsusingTODGA.Contribution:allexperimentalworkanddataanalysis

ABBREVIATIONSANDPHYSICALQUANTITIESαA/BSeparationfactorbetweenAandBθOrganictoaqueousphaseratioCyanex923Trioctylphosphineoxides;TRPOCyanex925TriisooctylphosphineoxideDDistributionratioD2EHPADi(2-ethylhexy)phosphoricacidEHEHPA2-ethylhexylphosphonicacidmono-2-ethylhexylesterEUEuropeanUnionHDDsHarddiskdrivesHREEsHeavyrareearthelementsICP-OESInductivelycoupledplasma–OpticalEmissionSpectroscopyLREEsLightrareearthelementsMRIMagneticresonanceimagingNdFeBNeodymiummagnetsREEsRareearthelementsrpmRotationsperminuteSmCoSamarium-cobaltmagnetsTALSPEAKTrivalentActinide/LanthanideSeparationsbyPhosphorusExtractantsfromAqueousKomplexesTRUEXTRansUraniumEXtractionTBPTributylphosphateTODGATetraoctyldiglycolamideUSAUnitedStatesofAmericavpmVibrationsperminutew.%Weightpercentage

TABLEOFCONTENTS

1.INTRODUCTION...................................................................................................................................................11.1.Thescopeofthethesis................................................................................................................................2

2.BACKGROUND.....................................................................................................................................................42.1.RAREEARTHELEMENTS...............................................................................................................................4

2.1.1.Recyclingofrareearthelements.........................................................................................................52.2.NEODYMIUMMAGNETS..............................................................................................................................6

2.2.1Whyrecycleend-of-lifeneodymiummagnets?....................................................................................82.2.2.RecoveryofREEsoutofneodymiummagnetswasteviahydrometallurgy.............................................92.3.SEPARATIONOFREEsFROMEACHOTHER.................................................................................................11

3.THEORY..............................................................................................................................................................143.1.SOLVENTEXTRACTION...............................................................................................................................14

3.1.1.SolventextractionusingtheacidicextractantD2EHPA.....................................................................163.1.2.SolventextractionusingthesolvatingextractantTODGA.................................................................173.1.3.Diluentsinsolventextraction............................................................................................................18

4.EXPERIMENTAL..................................................................................................................................................214.1.LEACHINGOFTHEMAGNETPOWDERFOLLOWEDBYSOLVENTEXTRACTIONOFTHEREESWITHTODGA..........................................................................................................................................................................22

4.1.3.Modelsolutiontesting.......................................................................................................................224.1.5.Stripping.............................................................................................................................................24

4.2.SOLVENTEXTRACTIONOFTHEREESWITHD2EHPAFROMALEACHATEOBTAINEDBYSELECTIVELEACHING.........................................................................................................................................................24

4.2.1.Investigationofthekineticsofsolventextraction.............................................................................244.2.2.TheeffectoftheconcentrationofD2EHPAandthediluentonsolventextraction..........................254.2.3.InvestigationofthepHeffectonextraction......................................................................................264.2.4.Stripping.............................................................................................................................................26

5.RESULTSANDDISCUSSION................................................................................................................................275.1.LEACHINGOFTHEMAGNETPOWDERFOLLOWEDBYSOLVENTEXTRACTIONOFTHEREESWITHTODGA..........................................................................................................................................................................27

5.1.1.Determinationofthecompositionofthemagnetmaterial..............................................................275.1.3.TheeffectoftheconcentrationofTODGAandthediluentonsolventextraction............................285.1.4.Slopeanalysis(stoichiometry)...........................................................................................................315.1.5.Separationfactors..............................................................................................................................315.1.6.Stripping.............................................................................................................................................33

5.2.SOLVENTEXTRACTIONOFTHEREESWITHD2EHPAFROMALEACHATEOBTAINEDBYSELECTIVELEACHING.........................................................................................................................................................34

5.2.1Determinationofthecompositionoftheleachate............................................................................345.2.2.Investigationofthekineticsofsolventextraction.............................................................................355.2.3.TheeffectoftheconcentrationofD2EHPAandthediluentonsolventextraction..........................395.2.4.Stoichiometry.....................................................................................................................................435.2.6.SeparationfactorsbetweenheavyandlightREEs.............................................................................445.2.7.Stripping.............................................................................................................................................46

6.CONCLUSIONS...................................................................................................................................................49

7.FUTUREWORK..................................................................................................................................................51

8.AKNOWLEDGEMENTS.......................................................................................................................................52

9.REFERENCES......................................................................................................................................................53

1

1.INTRODUCTIONRare earth elements (REEs) have become essential and criticalmaterials inmanymoderntechnologies.Duetotheiruniquemagnetic,spectroscopic,electricandcatalyticproperties,theyhaveplayedapivotalroleinthedevelopmentofvariousproductssuchasneodymium-iron-boron (NdFeB) permanent magnets, commonly referred to as neodymium magnets,samarium-cobalt(SmCo)permanentmagnets,lampphosphorsandbatteries,tonameafew.Duringthelastdecades,around95%oftheglobaldemandforREEswassuppliedbyChina,whichhas ledtoasignificantprice increaseoftheseelementsparticularlysincetheglobalfinancial crisis in 2011. [1]With ever growing demand, fluctuating prices, and significantsupplyriskREEsarecurrentlycategorizedasthemostcriticalelementsintheEU.[2]

Figure1.CriticalassessmentofawiderangeofrawmaterialsbytheEUcommissionin2013accordingtoeconomicimportanceandsupplyrisk.[2]ItcanbeobservedinFigure1thattheREEsareconsideredthemostcriticalelementswithrespecttosupplyriskintheEU,withtheHREEsbeingatthehighestriskofsupply.Itisthusimportanttoassurethesupplyriskiseliminated,oratleastminimizedtothelowestpossiblelevel. Recycling of the REEs from end-of-life products could help lower the supply risk byrecoveringtheREEsfromtheseproducts.A range ofmethodologies for the recovery and reuse fromend-of-life products are beinginvestigated, but industrial-scale applications are almost non-existent. [1, 3-5]. End-of-life

Report on Critical raw materials for the EU

Page 3 of 41

1. EXECUTIVE SUMMARY

Raw materials are fundamental to Europe’s economy, growth and jobs and they are essential for maintaining and improving our quality of life. Recent years have seen a growth in the number of materials used across products. Securing reliable, sustainable and undistorted access of certain raw materials is of growing concern within the EU and across the globe. As a consequence of these circumstances, the Raw Materials Initiative was instigated to manage responses to raw materials issues at an EU level. At the heart of this work is defining the critical raw materials for the EU’s economy. These critical raw materials have a high economic importance to the EU combined with a high risk associated with their supply.

The first criticality analysis for raw materials was published in 2010 by the Ad-Hoc Working Group on Defining Critical Raw Materials. Fourteen critical raw materials were identified from a candidate list of forty-one non-energy, non-agricultural materials.

In the 2013 exercise fifty-four non-energy, non-agricultural materials were analysed. The same quantitative methodology as in the previous 2010 exercise applies two criteria - the economic importance and the supply risk of the selected raw materials. The criticality zone is defined by the thresholds of 2010 to ensure comparability of the results. This extended candidate list includes seven new abiotic materials and three biotic materials. In addition, greater detail is provided for the rare earth elements by splitting them into ‘heavy’ and ‘light’ categories and scandium. The overall results of the 2013 criticality assessment are shown below; the critical raw materials are highlighted in the red shaded criticality zone of the graph.

2

neodymiummagnetsareviewedasaviablesecondarysourceforobtainingsomeREEs.Theyareusedinharddiskdrives(HDDs),motorsinhybridcars,windturbinesandMRImachines,amongothers. [3] Since theirusehas increased in the last fewdecadesandnewwaste isproduceddaily, theycouldbeaviablestock for further reprocessingafter theyhavebeenused.Therecyclingcould,toacertainextent,helpstabilizethemarketprice.Veryfewlarge-scale recycling processes have been developed,mostly due to the lack of data about thequantitiesoftheREEsinthewastestreamsandthefateofmagnetsaftershredding.ThelackofincentivesfordevelopingsuchaprocesshavealsobeenakeyfactorandaremostlyaresultofthelowpricesoftheREEsupto2011.[6]Therecycledandpurifiedproductsoftherecyclingofneodymiummagnetshaveapotentialofbeingusedasrawstartingmaterialsfortheproductionofnewneodymiummagnets,otherproductscontainingREEsorcanbesoldasrawmaterials.UptofiveREEscanbefoundintheneodymiummagnets. These are neodymium, dysprosium, praseodymium, gadolinium andsmalleramountsofterbium.[1]Theoverallgoalinanindustrialprocessistoachieveacircularsystemofproductionandeventualreuse/recyclingoftheneodymiummagnets.Solvent extraction or liquid-liquid extraction, a hydrometallurgicalmethod, is a frequentlyusedmethod for the recoveryand removalofmetal ions fromaqueous solutions. [7] It isbasedon thephenomenon inwhichmetal ions fromtheaqueous solutionaredistributedbetween organic and aqueous phase in a constant and specific ratio, dependent on thecompositionoftheaqueousandorganicphaseandotherexternalfactors.SolventextractionwasusedinthisworkasmethodofrecoveryofREEsoutofaqueoussolutionsobtainedbyleachingoftheneodymiummagnetwaste.

1.1.ThescopeofthethesisThemainobjectiveofthisworkistoassessthefeasibilityoftheselectiveextractionoftheREEsfromrealneodymiummagnetwasteleachateinnitricandsulfuricmedia,leavingotherelementsintheaqueoussolution.ThemainfocushasbeentheeffectofthesolventandthediluentmakinguptheorganicphaseontherecoveryofREEsoutofaqueoussolutions.Theextractants employed were tetraoctyl-diglycolamide (TODGA) and di-(2-ethylhexyl)phosphoric acid (D2EHPA). The used diluents were Solvent 70, hexane, octane,cyclohexanone, toluene, 1-octanol and chloroform. Critical system parameters wereinvestigatedliketheconcentrationofextractantintheorganicphase,acidityoftheaqueousphase, ionic strength of the aqueous phase and kinetics of extraction. Together with theinvestigation of bulk REEs extraction from commercial waste, studies on element specificseparationoftheadjacentandnon-adjacentREEswereconducted.Tentativestepstowards

3

theseparationoftheREEsfromeachotherweretaken;wherebytheextractantconcentrationandthediluenteffectonthiskindofseparationwereinvestigated.

4

2.BACKGROUND2.1.RAREEARTHELEMENTSTheREEsareagroupof17closelyrelatedelements.Thesearescandium,yttrium,lanthanum,cerium, praseodymium, neodymium, promethium, samarium, europium, gadolinium,terbium,dysprosium,holmium,erbium,thulium,ytterbiumandlutetium.[8]Scandiumandyttriumarelocatedinthed-blockoftheperiodicsystem,whilethelanthanidesbelongtothef-block. The REEs are separated into two groups: the so called light rare earth elements(LREEs),i.e.lanthanumthroughgadolinium,referredtoastheceriumgroup,andtheheavyrareearthelements(HREEs),i.e.terbiumthroughlutetiumplusyttrium,sometimesreferredtoastheyttriumgroup.LREEsarecharacterizedbyincreasingnumberofunpairedelectronsfrom0to7,whileHREEshaveanincreasingnumberofpairedelectronsfrom8to14.YttriumisaddedtotheHREEsgroupduetothesimilarityintheionicradiusandchemicalpropertiesto thegroup. [9] The chemistryof theREEs inaqueous solutions isdominatedby the3+oxidationstate,withsomeexceptionsduetothestabilityofthehalf-filledorfilledf-orbitals(Ce4+,Eu2+andYb2+).Chemicalbondingintheseelementsispredominantlyionicinnature,withverylimitedcovalentcharacter.Ln3+arehardLewisacidsandexhibitfastligandexchangeinaqueoussolutions.[10]SomeofthecharacteristicsofhardLewisacidsaresmallionicradii,highoxidationstates,lowpolarizabilityandhighelectronegativity.[11]Eventhoughtheyarecalled ‘’rare’’earths,theyarecomparativelyabundant intheEarth’scrust,butdespitetheirabundancetheyarequitedispersedandthenumberofeconomicallyfeasibleexploitingsitesislimited.[12]Sincetheyarechemicallysimilar,theyoccurtogethernaturallyinawiderangeofores.REEs are found in a wide range of products, such as fluorescent lamps, magnets,superconductors, lasers, ceramics, semiconductors, catalysts, and thermal neutronabsorbents.[13]ThepercentageoftheREEsusedbyapplicationissummarizedintable1.

5

Table1.UsageofREEsbyapplication (inw.%), taking into considerationonlyREEs in thecomposition.[1]*

Application La Ce Pr Nd Sm Eu Gd Tb Dy Y Other

Magnets 23.4 69.4 2 0.2 5 Batteryalloys 50 33.4 3.3 10 3.3 Metallurgy 26 52 5.5 16.5

Autocatalysts 5 90 2 3 FCC 90 10

Polishingpowders 31.5 65 3.5

Glassadditives 24 66 1 3 2 4Phosphors 8.5 11 4.9 1.8 4.6 69.2 Ceramics 17 12 6 12 53 Others 19 39 4 15 2 1 19

*Averagevaluesbyconsumption,mightvaryfrommanufacturertomanufacturer

2.1.1.RecyclingofrareearthelementsMany of thementioned products are important for the development of environmentallyfriendlytechnologiesfortransport,lighting,energystorageandmanufacturingofchemicals,allofwhichareessential tothehumanexistencenowadays.Thus, thedemandforREEs isconstantlygrowing.Sincetheyareminedinonlyafewcountriesandtheirpriceshavevariedintherecentyears,theirsupplyisconsideredcriticalinboththeEUandtheUSA.[1]Afterthesky-rocketingof thepricesofREEs in2011, itbecameclear towhatextent thedevelopedeconomiesaredependentontheseelements,andshoulddoalltobelessdependentonthemainprovider,China,whichatthattimeandevenuptothisdayprovidesover90%oftheworld’sneedfortheseelements.Thefactthataverylowpercentage[3,14]oftheREEsarerecycledtothisday,mostlyduetolackofincentivesinthepastdecades,makestherecyclingprocessinvestigationanddevelopmentevenmoreimportantandchallenging.TheREEsarechemicallyverysimilar,thustheyoccurasmixturesinnature.[8]Thismeanstheywillbehardtoseparatefromoneanotherfromaprimarysource,andequallysofromanend-of-lifeproduct.SincemanyoftheapplicationsrequireREEsinelementalform,theseparationoftheindividualREEsfromeachotherisessentialforfurtherdevelopmentandproductionof

6

thefinalproducts.TheseparationofREEshasbeenshowntobeoneofthemostdifficultininorganic chemistry. On an industrial scale, this is done by solvent extraction or ion-exchange.[10,15]

2.2.NEODYMIUMMAGNETSOneofthemostimportantapplicationsoftheREEsarethesocalledneodymiummagnets,sometimesreferredtoastheNdFeBmagnets.Neodymiummagnetsareusedinmanyhigh-techandcleanenergyapplications,suchasHDDs,electricvehiclesandelectricgeneratorsinwindturbines,showninFigure2.[16]

a) b) c)

Figure2.Someofthemostimportantapplicationsofneodymiummagnets:a)HDDs[17],b)electricvehicles[18],c)windturbines[19]Neodymium magnets are the strongest permanent magnets commercially available.Chemically,theyaremadeofanalloyofneodymium,ironandboronwhichformatetragonalNd2Fe14Bcrystallinestructure.[20]

Figure3.TheNd2Fe14Bcrystallinestructure

7

TheNd2Fe14Bmatrixphaseissurroundedbyaneodymiumrichgrainboundaryphase,withsmalladmixturesofpraseodymium,gadolinium, terbiumanddysprosium,aswellasotherelements,suchasaluminum,cobalt,copper,molybdenum,niobium,titanium,vanadiumandzirconium[1],whichareusuallyaddedtotheNd2Fe14Balloytoimprovecertainpropertiesofthemagnet.Theadditionofdysprosiumimprovesthehightemperatureperformanceofthemagnetandincreasestheirintrinsiccoercitivity.Smallbutconsiderableamountsofcobaltareaddedto increasetheCurietemperatureofthemagnet,thetemperatureatwhichcertainmaterialslosetheirpermanentmagneticproperties.[21]Thispointsoutthattheneodymiummagnetsusedathigheroperatingtemperatureswillcontainmorecobaltandviceversa.Table2.TypicalcompositionofNdFeBmagnetsinw.%[3]*

Element Elementalcompositioninw.%

Co 4.22 Fe 58.16 65-70 69Nd 25.95 30 25Pr 0.34 Dy 4.21 3 4B 1 1 1

Other 1 *dataobtainedbytotaldissolutionAsseenfromTable2,theneodymiummagnetscontainmainlyiron,whichmakesuparoundtwothirdsofthematerial,andneodymium,whichmakesaroundonethirdofthealloy.Otherelementsthatcanbefoundinconsiderableamountsareboron,cobaltanddysprosium.Inthelasttwodecadestheneodymiummagnetshavelargelyreplacedthesamarium-cobalt(SmCo) permanent magnets due to better magnetic properties. Furthermore, the SmComagnets require large amounts of cobalt, which significantly increases the cost of thematerial.Thedrawbacksoftheneodymiummagnetsareloweroperatingtemperatureandlower corrosion resistance than that of SmCo magnets. [1] The corrosion resistance isincreased by plating the surface with a protective layer, usually consisting of nickel andcopper.

8

2.2.1Whyrecycleend-of-lifeneodymiummagnets?Neodymium magnets are, as already stated, a good secondary source of some valuableelements.Sincetheymakeupthemajorityofthepermanentmagnetsmarkettoday,morewasteisexpectedtobeproduced.TheyaremoreinterestingfromarecyclingpointofviewthanSmComagnets,whichmakeuplessthan2%ofthepermanentmagnetsmarket.[1]Therecyclingprocessoftheneodymiummagnetscanbefurthercomplicatedbytheprotectivecoating.The reuse of the already existing magnets is the most viable method of recycling thesemagnets,whichisonlypossiblewithbigandeasilyaccessiblemagnetsinelectricvehiclesandwindturbines,butthesemagnetswillbeinuseforalongperiodoftimeandarenotavailablein large quantities in scrap today. The main source of such end-of-life waste today areconsideredtobeelectronicgoodslikeloudspeakers,cellphonesandHDDs.[1]Itisveryhardtorecovertheneodymiummagnetsoutofsuchproductsbecauseofthesizeofthemagnetsinthoseproducts,theytendtosticktootherferrouscomponentsduringshreddingandtheyneedtobedemagnetizedbeforereprocessing.OnlyahandfulofindustrialmethodshavebeendevelopedsuchasthethemethodofrecyclingoftheneodymiummagnetsfromHDDsandairconditioning devices via dismantling and further sorting based on magnetic propertiesdevelopedby theHitachi corporation in Japan in 2013. [22]. Someother demonstrationalplantsforrecyclinghavealsobeendeveloped[23],butwithnosuccessfullargescaleindustrialapplicationstodate.Thegreatestissuethatisencountered,evenwhenthemethodoftherecyclingoftheREEscontainingproductshasbeendeveloped,isthepricefluctuationoftheelements.[24]Sincethathasbeenthecaseinthelast2years,thepricesbeingsky-hightopricesbeingreasonablylow,alotofup-scalingoftheneodymiummagnetsrecyclingandnewandoldminesopeningshavebeendelayedorcancelled.Assoonasthepriceofelementofinterestdrops,theinterestforproceedingwiththeprojectbythecompanyseemstofade,andkeepingtheprocessaliveisalmostimpossible,themainreasonbeingnegativeprofit.Sotheoverallconclusiontothisproblemis,themorethepriceoftherawmaterialsincreases,themoreprofitable itwillbetorecycletheexistingend-of-lifematerialsattheend.Somelegislationsmight be needed to keep the recycling business alive since the very dynamicmarketandthevaluesoftheindividualelementscancrashsuchideasanddreamsinaglimpseofaneye.RecyclingofREEs andothermetals frommagnets canbedonebyhydrometallurgical andpyrometallurgical methods. Pyrometallurgical methods include electroslag refining, directmeltingandliquidmetalextraction.[1]Hydrogendecrepitationisamethodthatisalsobeing

9

under investigationanddevelopment. [16]Thesemethodssurpassthescopeofthis thesisandwillnotbefurtherdiscussedinthiswork.

2.2.2.RecoveryofREEsoutofneodymiummagnetswasteviahydrometallurgyHydrometallurgicalprocessingneodymiummagnetsinvolvesleachingoftheseelementswithmineralacids,followedbysolventextraction,ionexchangeorprecipitationforisolationandseparation of the desired metals. [1, 3, 7] There are many advantages of usinghydrometallurgyfortherecoveryofREEsoutofneodymiummagnetwastesuchasthelargeapplicabilityonmosttypesofcommercialmagnetsandthealreadyexistingknowledgesincethereprocessingoftheend-of-lifeneodymiumwasteissimilartotherecoveryoftheREEsformtheminingores.Themaindisadvantageofthehydrometallurgicalmethodscomparedto the pyrometallurgical methods is the large volume of solvents and acids used for therecoveryoftheelementsfromthewaste,howeverthiscanbemitigatedbysolventrecycling,e.g.organicphase,bystrippingandscrubbingoftheloadedphase.Duringtheleachingstep,selectivityoftheleachingagentisimportantsincethemajorityofthemagnet(around72%)consistsofiron.Ironwillthusmakeupthemajorityoftheionsinsolution,whichmightposechallengesinseparatingtheREEse.g.co-extractionbyphosphineoxidesduringsolventextraction.Selectiveleachingofironfromthemagnetscrapmightbeonlyapartialsolutiontotheproblem,duetothepresenceofotherelementssuchasboronandthecopperandnickelusedinthecorrosionprotectionlayer.[3]The idea of the selective leaching of REEs and as little as possible of other constituentmaterials,includingiron,isofmuchinterest.[1]ResultsobtainedbyÖnalandco-workers.[25]demonstratethatpowderedneodymiummagnetscrapcanbeselectivelyleached,leavingironinthesolidresidue.Thepowderedsamplesweretransformedintoasulfatemixturebymixingthepowderwithsulfuricacidinaluminacrucibleswiththeacid/magnetratio(g/g)of2.15(12M),3.2(13.5M),4.3(14.5M),and8.6(16M).Theachievedmixtureswerethendriedinamufflefurnaceat110oCfor6-24h.Thedriedsampleswerethentreatedat650-800oCfor15-120minutes foraselectiveroastingprocess.Theobtainedproductswere later leachedwithdemineralizedwaterfor15minto24hat225rpmonashaker.Thisprocessledto95-100% extraction efficiencies for neodymium, praseodymium, dysprosium and gadoliniumwhile iron remained in the residue after leaching. The whole concept was based on thedifferentsolubilityproductvaluesofREEssulfatesandironsulfate.

10

In2013.LeeC.etal[26]reportedonleachingneodymiummagnetscrapwithHCl,HNO3,H2SO4or NaOH. The process was optimized with regards to temperature, leaching time,concentration of leaching reagent and solid to liquid ratios. HCl and H2SO4 showed bestperformance.NeodymiumwassuccessfullyrecoveredfromoptimizedH2SO4leachingsolutionwith75.41%byprecipitation.Theoptimumconditionsweresolidtoliquidratio(S:L)of20g/L,15minutesleachingtimeand3Mhydrochloricacidor1.5Msulfuricacid.In 2014. Yoon at al [27] performed a similar investigation where increasing the leachingtemperaturegaveincreasedleachingeffectofH2SO4whenleachingtheneodymiummagnetscrap.Theoptimalleachingconditionsweredeterminedtobe4hoursleachingtimeat70oCusing3MH2SO4.Aftertheleachingsteptheelementsfromthesolutionareusuallyprecipitated,extractedwithanorganicsolventorionicliquid,orseparatedusingionexchange.Thesolventextractioncanbe done with all the conventional extractants present on the market, like D2EHPA, TBP(tributylphosphate),PC-88A(2-ethylexylhydrogen2-ethylhexylphosphonate)andEHEHPA(2-ethylhexyl phosphonic acid mono-2-ethylhexyl ester), which have been used for REEsextractionfrommineralacidmediainthepast.[28,29]The mentioned extracting agents have however shown some drawbacks concerningselectivity,strippingoftheelementsoutofthesolutionandextractionkinetics.Theissuesinthecaseofconventionalextractantsistheco-extractionofotherelementsthatwereleachedintothesolution.Moreover,almostnoneofthetoday’s industriallyusedextractingagentsfollow the CHON principle.[30] A promising extractant that could be employed for theextraction of the REEs out of the leachate containing all the elements leached from theneodymiummagnetwasteisTODGA(N,N,N’,N’-tetraoctyl-diglycolamide)whichshowedhighdistributionratiosoflanthanidesandactinideswhenextractedfromhighlyacidicsolutions.Itwasalsofoundthattheloadingcapacityof0.1MTODGA-n-dodecanewas0.008MNd(III)withanaqueousphaseof3MHNO3.[31]Anespeciallyinterestingfactforthisresearchistheverylowdistributionratios(<10-2)ofiron(III)andaluminum(III)thathavebeenshownbytheinvestigationoftheextractionofwith0.1TODGAinn-dodecanefrom2.9MHNO3.[32]In2015attheOakridgeNationalLaboratory,Tenneseeaprocesswiththemembranesolventextraction was conducted comparing the extraction of REEs from a neodymium magnetleachatewith Cyanex 923.With the TODGA,Neodymium, dysprosium and praseodymium

11

wereselectivelyrecoveredwiththeTODGA,resultinginnoco-extractionofnon-REEssuchasironandboron.[33]2.3.SEPARATIONOFREEsFROMEACHOTHERAsmentionedpreviously the chemistryof theREEs is very similar,making individualREEsseparationfromeachotherquiteachallengingtask.Thereisoneuniquepropertyofthelanthanidegroup,calledthelanthanidecontraction.Thedecrease in the ionic radii in the lanthanide group from lanthanum to lutetium, which isgreaterthanexpectedanddoesnotfollowthesametrendasinotherperiodsintheperiodictable of elements. This is the result of thepoor shielding of thenuclear chargeby the 4felectrons,meaning the 6s electrons aremore strongly drawn to the nucleus, resulting insmaller radius.[34] This property plays an important role in the separation of the REEs,particularlyinsolventextraction,sincetheionicradiuscanleadtotheformationofcomplexeswithdifferentstabilityconstantsandconsequentlytheirsolubilityintheorganicphase,thussomewhatenablingtheirseparation.AphenomenonknownasthetetradeffectisacommonfeaturediscoveredinthelanthanidegroupofmetalswhichislinkedtothedistributionofREEsinnatureandintheliquid-liquidseparationprocesses.ItreferstotheseparationofREEsintofourseparatesegmentscalledtetrads(firsttetrad,La-Ce-Pr-Nd;secondtetrad,(Pm)-Sm-Eu-Gd;thirdtetrad,Gd-Tb-Dy-Ho;fourthtetrad,Er-Tm-Yb-Lu).[35]Manyattemptstoseparatethef-blockelementshavebeenmadesincethe1950s,especiallyinthenuclearfieldwhereithasbeencrucialtoseparatetheREEsfromtheactinidesduetothehighneutronabsorptioncrosssectionwhichlanthanidespossess.Thatcaninterferewiththe reuse of actinides in further nuclear operations. [29] Thus the TALSPEAK (TrivalentActinide/Lanthanide Separations by Phosphorus Extractants from Aqueous Komplexes), aprocessofbasedonextractionofREEswiththeorganophosphorousextractantslikeD2EHPAwiththeaqueousphasecontainingdiethylenetriamineN,N,N’,N’petnaaceticacid(DTPA)asastheactinidecomplexatingagent,andTRUEX(TRansUraniumEXtraction)processesweredeveloped.[28]But since thebasic ideaof f-elements separationchanged throughhistory,fromtheideaofcreatingapromisingenergysourcetotheideaofenvironmentrestorationadoptedtodaytheapproachtothef-blockelementsseparationchangedaswell.WiththegrowingdemandsforREEs,andtherequirementsfortheirpuritythisseparationstechniquesforseparatingtheREEsfromeachotherwerefurtherinvestigated.

12

SomeattemptsinseparatingtheREESfromeachotherarelistedbelow.

In the separation of REEs, organophosphorous extractants are commonly used. Di(2-ethylhexyl)-phosphoric acid (D2EHPA), an acidic extractant, is commonly used on a largerindustrialscale.[36]Theselectivityorderforextractingrareearthsfrom0.5MHClsolutionwith0.75MD2EHPAintoluenewasfoundbyPeppardandcoworkerstobeLu>Yb>Tm>Tb>Eu>Pm>Pr>Ce>Lawiththelogofthedistributioncoefficientincreasinglinearlywiththeatomicnumberoftherareearth.[37]D2EHPAhasalsobeenusedtoseparatesamarium,europium,andgadoliniumfromtheotherREEsinamixednitrate-chlorideleachatefrommonazite.[38]Furthermore,inmanystudies[28]itwasshownthatatypicalseparationfactorbetweenadjacentREEsusingtheD2EHPAextractantisintheregionof2.5.ManystudieshavebeenconductedontheseparationofREEswithD2EHPAandotherrelatedextractantsinvariousmineralacidmedialikenitric,sulfuricandnitric.[39]Theyhaveshownthebestmutualseparationpropertiesinlanthanideseparationsofar.Cyanex925 inn-heptanewasusedfortheextractionofREEsoutofnitricacidmedia.Theseparation between the LREEs and HREEs was possible, in a few stages. The REEs werestrippedbackusingnitricacid.[40]Itisknownthatbychangingthediluentintheorganicphase,thepropertiesoftheextractionchangeaswell.ItisthusimportanttoinvestigatetheeffectofthediluentontheseparationoftheREEsfromeachother.InastudyfromOakridgeNationalLaboratory(1961)theseparationofAmandEuwithTIOA(triisooctylamine) in various diluents. It was shown that the separation factors do indeedchangewiththediluentscompositionchange.InastudybyDukovet.al.[41]aninvestigationoftheseparationofLa,Nd,Eu,HoandLuusingHTTA(thenoyltrifluoroacetone)andthequaternaryammoniumsaltAliquat336inchlorideorperchlorate formwas studied.Chloroform,benzene, tetrachloromethaneand cyclohexanewereusedasdiluents.Thestudydidnotshowadirectconnectionaboutthediluenteffectofthe separation factors, but the separation factors between lanthanum and neodymiumincreased when chloroform was changed with benzene, cyclohexanone andtetrachlorometane.

13

TODGAhasbeen tested for extraction in variousdiluents and it hasbeen shown that thestoichiometric factors vary by changing the diluent (toluene, 1-octanol, chloroform andothers),whichcouldhaveeffectsontheseparationfactorsandshouldbefurtherinvestigated.[29]

14

3.THEORY

3.1.SOLVENTEXTRACTION

Thetermsolventextractionor liquid-liquidextractionreferstothedistributionofasolutebetween two immiscible or partially miscible liquid phases, usually one organic and oneaqueousphase.[11]Insolventextraction,thesolvent,ororganicphase,ismadeupofanextractantdissolvedinadiluent.Insomecases,pureextractantisused,butitismorecommontouseadiluentsincemanyextractantsareviscousmaterialsintheirundilutedform.Athirdcomponent,amodifier,maybeaddedtotheorganicphasetopreventthirdphaseformation.Thesoluteisusuallyametalionthatisextractedintotheorganicphase.

1.2.3.Figure 4. Schematic display of a solvent extraction process. (1) Two immiscible phases(aqueous and organic), the aqueous phase containing themetals of interest. (2) The twophasesareputincontactbyvigorousshakingorstirring.(3)Thetwophasesaredisengagedandthemetalsofinterestaretransferredintotheorganicphase.The parameter representing the distribution of the metal of interest between the twoimmiscible phases is called distribution ratio (D). It is defined as the ratio of the totalconcentrationofthemetalAintheorganicphasesandthetotalconcentrationofthemetalAintheaqueousphase.

𝐷" =["]&'(["])*

(1)

OrganicphaseExtractant+diluent

Aqueousphase

Organicphase

Aqueousphase

PHASECONTACT D

15

The parameter showing the degree of separation between the two solutes in the sameextractionsystemiscalledtheseparationfactor(α).Itcanbecalculatedfromequation(2).

αA/B=+,+-(2)

Inorderforthemetaltobeextractedintotheorganicphase,theaqueousmetalspeciesneedtobechemicallymodifiedtobemakethemhydrophobicandsolubleintheorganicphase.Theorganic molecular species that are involved in this process are called extractants and,dependingonthemechanismoftheextractionofthemetalfromtheaqueousintotheorganicphase,theycanbeclassifiedinthreemaingroups:[42]• ACIDIC:theorganicaciddissociatesanditsconjugatedbasereactswiththecationtoform

aneutralcomplexaccordingtothefollowingequation3:[7]

Ln01 +m(HR)9→←LnR0(HR)9<=0 + 3H1(3)

whereHRrepresentstheassociatedacidicextractantmoleculesandLn3+theREEioninthesolution.• BASIC/IONPAIR:theorganicspecieformsanionpairwiththenegativelychargedmetal

complexintheaqueousphaseaccordingtothefollowingequations4and5:

2RNH9 + 2H9SOC →←2(RNH0)9SOC(4)

2Ln(SOC)00= + 3(RNH0)9SOC

→←2(RNH0)0Ln(SOC)0 + 3SOC9=(5)

where RNH2 represents the tri-alky methylamine and Ln3+ represents the REE ion in theaqueoussolution.Thisextractionmechanismisonlypossibleinthepresenceofstronganionicligands.• SOLVATING:thehydratingwaterintheinnersphereofthemetalatomarereplacedwith

theorganicspeciesaccordingtothefollowingequation6:

Ln01 + 3NO0= + 3TBP→←LnTBP0(NO0=)0(6)

16

TBPrepresentstributylphosphate,acommonsolvatingextractant,andLn3+theREEionintheaqueoussolution.Solventextractionhasawiderangeofapplications,whichincludenuclearreprocessing,metalrecoveryfromaqueoussolutionsandproductionoforganiccompounds.[11]Whenconsideringtheuseofacertainextractingagent,somekeypropertiesneedtobetakenintoconsiderationpriortotheapplication:[42,43]• Reasonablecostofproduction• Chemicalandphotochemicalstability• Verylowsolubilityinaqueoussolutions• Solubleinorganicdiluents• Selectivitytowardsthesolutetobeextracted• Extractionoccurswithouttheadditionofmodifiers• Efficient function of the extractant-diluent mixture with the proposed feed and strip

solutionintermsofratesofoperationsanddegradationstability

3.1.1.SolventextractionusingtheacidicextractantD2EHPADi-(2-ethylhexyl)phosphoricacidisanorganophosphorouscompound,whichisprimarilyusedinthesolventextractionofuraniumandREEs.Ithasshowngoodversatilityasanextractantfor lanthanide separations due to its chemical stability, good kinetics in extraction, goodloadingandstrippingpropertiesandavailabilityincommercialquantities.[44]Manyauthors[45, 46] have reported that the extraction of REEs with D2EHPA (here called HR) occursaccordingtothereactionpathdescribedinequation7,

M01 +m(HR)9→←MR0(HR)9<=0 + 3H1(7)

whereM3+isthelanthanideioninthesolution,HRistheorganophosphorousextractantinthe organic phase occurring as a dimer (HR)2 and asMR3(HR)2m-3 in the complex formed,solubleonlyintheorganicphase.

17

Figure5.StructuralformulaofD2EHPA(di-(2ethylhexyl)phosphoricacid)

3.1.2.SolventextractionusingthesolvatingextractantTODGAN,N,N’,N’-tetraoctyl-diglycolamideisasolvatingextractant,whichcreatesstrongtridentatecomplexeswithmetal ions,andhasshownparticularlygoodextractionpropertiesforREEsions in termsof selectivity in comparison to other ions in the aqueous solution. [32] Thisextractant has shown good stability at room temperature andmutualmiscibilitywith thecommonlyuseddiluents.Theonlysetbackforlargeindustrialuseiscurrentlytheprice.

Figure6.StructuralformulaofTODGA(N,N,N’,N’-tetraoctyl-diglycolamide)ThesolvatingmechanismisassumedforallthecomplexationreactionswithTODGA.Theslopecanthenbecalculatedfromequation(9),mthatis,whichrepresentsthenumberofligandmoleculesthatisusedintheformingofthecomplex.

MH1 + nNO0= + mTODGA→←M(NO0)HTODGAKMN =

O(PQR)STUOSV PQRW S T

U(8)

18

logD = mlog L + nlog NO0= + logKMND = O PQR STU

OSV (9)

The concentration represented in the equations is the equilibrium concentration in theorganic phase after extraction, and thedistribution ratio the equilibrium concentrationofmetalsinthesolution.

3.1.3.DiluentsinsolventextractionA diluent is a liquid or homogenous mixture of liquids in which extracting agent(s) andmodifier(s)canbedissolvedtoformthesolventphase.[47]Solventshouldbenotusedasaterm since it is a broad term in the solvent extraction field. The term solvent in solventextractionrepresentsthemixtureoftheextractantanddiluent.Thediluent itselfdoesnotextractthesolutesignificantly.Anarrayofmethodsforclassifyingdiluentshavebeenproposedaswellasfindingaperfectwayofpredictingtheoutcomeofanextraction.Thiswasbasedonthesolubilityparameters,connectivity,dielectricconstant,butnooriginalwayhasbeenfound.[48]Diluents can be classified according to their chemical and physical properties such as theabilitytoformorderednetworks.Theycanbedividedintofollowinggroups[11]:

1)Liquidscapableofformingthree-dimensionalnetworksofstronghydrogenbonds,e.g.water,hydroxyacids,polyolsandother.

2)Liquidscontainingbothactivehydrogenatomsanddonoratoms(O,N,F)butdonotformthree-dimensionalnetworks,butinsteadformchainlikeoligomers,e.g.primaryalcohols,carboxylicacidsprimaryaminesandother.

3)Liquidscontainingdonoratomsbutnoactivehydrogenatoms,e.g.ketones,ethers,aldehydesandother.

4)Liquidscontainingactivehydrogenatomsbutnodonoratoms,e.g.chloroformandsomeotheraliphatichalides.

5)Liquidswithnoactivehydrogenatomsandnodonoratoms,e.g.hydrocarbons,carbondisulfide,carbontetrachlorideandother.

19

Thediversepropertiesof thediluents inasolventextractionsystem leadtodifferences indistributionratiosofthesoluteandtheoverallextractionprocess.Thediluents ingroup3oftenhave theability toextract thesolutewithout furtheradditionof theextractant.Theliquidsfromgroups4and5donotdissolvesaltwithouttheadditionoftheextractantandareoften used as diluents in a solvent extraction system. The liquids from group 1 are easilysoluble in water and are impractical from a solvent extraction point of view.[11] Waterbelongstothisgroupanditisusuallyusedasasecondphase.Thepolarityofthediluentcanalsosignificantlyaffecttheextractionprocesssincethesolubilityoftheneutralcomplexintheorganicphase is inverselyproportional to thepolarityof theorganicdiluent. [49]Thepolarityofthediluentcanbeexpressedwiththedielectricconstant,alsocalledtherelativestatic permittivity (ε) which is a measure of chemical polarity. The values of dielectricconstantsforthediluentsusedinthisworkareshownintable3:Table3.Thevaluesofthedielectricconstantforsolvent70,hexane,octane,cyclohexanone,1-octanol,chloroformandtoluene.[50]

Diluent DielectricconstantSolvent70 1.8Hexane 1.88Octane 2

Cyclohexanone 18.31-Octanol 10.3Chloroform 4.81Toluene 2.38

Theabilityoftheorganicdiluenttoformhydrogenbondscansignificantlyaffecttheextractionprocess. For example, alcohols have both active hydrogen atoms and donor atoms forhydrogenbondsforming.Thiscanaffectthesolubilityoftheextractantinthediluentandthusleadtolowerdistributionratios.Thus,theinteractionsbetweenthediluentandthecomplexformedintheorganicphasewilldependonthepropertiesofthediluentsandthenatureofthecomplexformed.Water solubility isan importantparameter for theorganicdiluents, since itwill affect thephaseratiochange,anditisknownthattheorganicsolventshouldbeaswater-immiscibleaspossible.Thewatersolubilityofthediluentsusedinthiswork,expressedasS/mass%at25°Careasfollows:solvent70(traces),hexane(0.0011),octane(0.000071),cyclohexanone(8.8),1-octanol(0.054),toluene(0.0531)andchloroform(0.80).[51]

20

Alongwiththeextractants,thediluentsplayapivotalroleintheextractionprocessaneedforintroducingthedemandsforthediluentswasneeded.Someofthegeneraldemandsforthediluentsare:[52]

• Lowdensity• Lowwatersolubility• Highflashpoint• Highboilingpointandlowfreezingpoint• Lowchemicaltransformationratewithwaterorreagents• Nothirdphaseformationunderloadingconditions

21

4.EXPERIMENTALTheresearchcarriedoutinthisworkhasbeentheinvestigationofthediluenteffectonthesolventextractionoftheREEsoutoftheleachateobtainedbydissolutionoftheneodymiummagnetwaste.Tworouteswerefollowedfortheexperimentalpart.Theycanbedividedintothefollowing:1)LeachingofthemagnetpowderwithHNO3followedbysolventextractionoftheREEswithTODGAdilutedinvariousdiluents.Theexperimentscarriedoutinthissectionhavefocusedon thecharacterizationof thehydrogendecrepitatedneodymiummagnet, leachingof thewasteusingHNO3,investigationoftheeffectofHNO3concentrationonthesolventextractionprocess,investigationoftheeffectofdiluentonthesolventextractionprocessandstrippingoftheREEsoutoftheorganicphase.

Figure7.Flowsheetforthe leachingoftheREEsoutofhydrogendecrepitatedneodymiummagnetinHNO3followedbythesolventextractionwithTODGAinvariousdiluents2)SolventextractionoftheREEswithD2EHPA,invariousdiluents,fromaprovidedleachate.Theexperimentscarriedoutinthissectionhavefocusedondeterminingthecompositionofthe aqueousphase (provided leachate), investigationof the kinetics of solvent extraction,investigationoftheeffectofthediluentonthesolventextraction,effectoftheequilibriumpHonthesolventextractionprocessandstrippingoftheREEsoutoftheorganicphase.

Figure8. Flowsheet for the solventextractionand strippingof theREESoutofa leachateobtainedbysulfonationroasting(sulfatemedia)

NdFeBPOWDER LEACHING SOLVENTEXTRACTION STRIPPINGOFTHEREES FURTHERREPROCESSING

NdFeBLEACHATE SOLVENTEXTRACTION STRIPPINGOFTHEREES FURTHERREPROCESSING

22

4.1.LEACHINGOFTHEMAGNETPOWDERFOLLOWEDBYSOLVENTEXTRACTIONOFTHEREESWITHTODGA4.1.1.DeterminationofthecompositionofthemagnetmaterialThehydrogendecrepitatedneodymiummagnetmaterialwasobtainedbytheUniversityofBirminghaminapowderform.Theparticlesizewas50–100µm.Thecompositionofthehydrogendecrepitatedneodymiummagnetmaterialwasdeterminedaftertotaldissolutioninaquaregiaat80±1°C.Theexperimentsweredoneintriplicateswhichwerethenusedforestimationoftheuncertaintyofaparticularexperiment.Aquaregiawaspreparedbymixingconcentratednitric(70%,ACSreagent,SigmaAldrich)andhydrochloricacid (37%,ACSreagent,SigmaAldrich) involumeratiosof1 :3,accordingly.Approximately1gofthehydrogendecrepitatedmaterialwastotallydissolvedin10mLaquaregia, 10%w/V ratio. The sampleswere treated for 1 hour on a heating plate. Samplesobtainedwerelefttocoolfor2hoursandthenfilteredthroughpolypropylenefilters(0.45µmVWR)andfurtherdilutedwith0.5MHNO3(65%,suprapur®,Merck)formeasurementwithICP-OES(ppmscale).Noresidueswereobservedonthefilterpaper.4.1.2.PreparationoftheleachateTheleachateoftheNdFeBpowderwaspreparedbydissolvingroughly1gofthepowderin50mL4MHNO3(70%,ACSreagent,SigmaAldrich).Thepowderleachingwascarriedoutfor24hoursatthetemperatureof25±1°Conamagneticstirrer.Theleachingexperimentsweredoneintriplicatesinpolypropylenebottlessecuredwithlids.Residueswerenoticedatthebottom of the vial after the leaching experiment. The residue was filtered through apolypropylenefilter(0.45µm,VWR)andfurtherdissolvedinaquaregiaat80±1°Cfor1houronaheatingplate.Thedissolvedresiduewasthedilutedwith0.5MHNO3(65%,suprapur®,Merck)andthesolutioncontentwasdeterminedbytheICP-OES,andshowedthatonlyNiwaspresentintheresidue,theelementusedinthecoatingofthemagnet.Thefilteredleachatewasalsofurtherdilutedwith0.5MHNO3(65%,suprapur®,Merck)andanalyzedwiththeICP-OES.

4.1.3.ModelsolutiontestingApreliminaryexperimentwasdonewiththemodelsolutionrepresentingtheleachate.ThemodelsolutionwaspreparedoutofstandardsolutionsofNd,DyandFe(LGCStandards,1000mg/L).Themodelsolutioncontained21.49mMFe,3.46mMNdand0.62mMDyin0.1,1,2,

23

3,4,5,6MHNO3,totestthenitricacidconcentrationeffectonthesolventextraction.TheconcentrationofthenitricacidintheaqueoussolutionwasdeterminedbytitratingwithNaOH(0.1MNaOH,FIXANAL).Theaqueousphaseswereputincontactwith0,1MTODGA(>97%synthesized at Chalmers University of Technology and 2.0 g obtained by the Institut fürNukleareEntsorgung,KarsruheInstituteofTechnology)inSolvent70.TheorganictoaqueousphaseratiowasΘ=1.Thevialswereshakenfor50minutes,whichwasenoughtoachieveequilibrium[32],atthetemperatureof25±1°Cusingashakingmachine(IKAVibraxVXRBasic)withanadjacentthermostaticbath.Shakingspeedwas1750vibrationsperminute.Thephaseswereputincontactin3.5mLglassvials(46x13x0.8mm)withpolypropylenelids.Alltheexperimentsweredoneintriplicatesandwerecentrifugedatrotationspeed2000rpm for1minutebefore sampling. The sampledaqueous solutionwasdilutedwith0.5MHNO3andanalyzedusingtheICP-OES.4.1.4.TheeffectoftheconcentrationofTODGAandthediluentonsolventextractionThefilteredleachateobtainedinsection4.1.2.wasdilutedwith3MHNO3toachieve4000mgL-1ofthetotallydissolvedpowderinthesolution,andwasusedinthesolventextractionexperiments.TheconcentrationofHNO3afterthedilutionwasdeterminedbytitrationwithNaOH(0.1MNaOH,FIXANAL),andwasdeterminedtobe3.1M.TheconcentrationsofTODGAusedwere0.01,0.05,0.1,0.2and0.4M.Solvent70 (hydrocarbonsC11-C14,≤aromatics,Statoil,Sweden),hexane(95%,anhydrous,SigmaAldrich),toluene(99,8%,anhydrous,SigmaAldrich), cyclohexanone (≥ 99%, ACS reagent, Sigma Aldrich) and 1-octanol (≥ 99%, ACSreagent, Sigma Aldrich) were used as diluents for determining the optimal extractionconditions.Alltheorganicphaseswerepre-equilibratedwithanequalamountof3.1MHNO3.Pre-equilibrationwasdoneinordertominimizethephaseratiochange,sincesomediluentsaresolubleinwater.Itisdonealsotominimizetheeffectanotherspeciesinthesolution,i.e.inextractantswhichextractacids.[47]Thephaseswereputincontactin3.5mLglassvialsandwereshakenusingashakingmachinewithanadjacentthermostaticbath.Shakingspeedwas1750rotationsperminute.Allthevialswereshakenfor50minutestoassureequilibriumduring theextractionexperiments.Theexperimentswereperformed in triplicatesandtheuncertaintieswereexpressedas±1σ.Alltheextractionexperimentswereperformedatthetemperatureof25±1°C.InalltheperformedexperimentstheorganictoaqueousvolumeratiowasΘ=1.Beforesamplingoftheaqueousphase,thevialswerecentrifugedatrotationspeedof2000rpmfor1minute.Thesampledaqueoussolutionsweredilutedwith0.5MHNO3andanalyzedusingICP-OES.Thedistributionratioswerecalculatedasmassbalance,theconcentrationoftheelementsintheaqueoussolutionbeforeandaftertheextraction.

24

4.1.5.StrippingStripping (back-extraction)wasperformedbyusing0.01,0.1MHNO3 (70%,ACS reagent,SigmaAldrich)andMQwaterasaqueousphases.Theorganicphaseaftertheextractionwasseparated from themetal-depleted aqueous phase and put in contact with the strippingaqueousphase.Thiswasdonebyputting5mLofeachofthephases intocontactover20minutesbymanualshakingina20mLglassvials.Theloadedorganicphasesusedforstrippingwere0.01MTODGAinSolvent70and0.1MTODGAinhexaneafterextraction.Atemperatureof 25 ± 1 °C was used as well as theΘ = 1. All the stripping experiments were done intriplicates. The pH of the stripping solutionwasmeasured before and after the strippingexperiments.Thesampledaqueousphasesweredilutedwith0.5MHNO3andmeasuredusingICP-OES.SincetheHNO3acidextractedbyTODGAwasexpectedtobestrippedbackintotheaqueoussolutiontheequilibriumpHwasmeasuredbeforeandafterthestripping.

4.2.SOLVENTEXTRACTIONOFTHEREESWITHD2EHPAFROMALEACHATEOBTAINEDBYSELECTIVELEACHINGThe neodymiummagnet leachate was produced by Önal and co-workers using sulfation,selectiveroastingandwaterleaching.[25]TheneodymiummagnetleachatecompositionwasmeasuredusingICP-OESbydilutingthesampledleachatewith0.5MHNO3(65%,suprapur®,Merck)andthenperformingthemeasurement.Theexperimentwasdoneintriplicates.ThepHvalueofthesolutionwasmeasuredwithMeterLabTMPHM240pH/IonMeterpHelectrode,takingintoaccounttheionicstrengthoftheleachate.Thesulfateionconcentrationneededtobedetermined.ThiswasdonebyprecipitationofBaSO4usingBaCl2(99,999%tracemetalbasis,SigmaAldrich)dissolvedinMQwater.Theobtainedprecipitatewasfilteredthroughapolypropylenefilterpaper(0.45µm,VWR),washedwith20mLMQwaterandthendriedover48hoursinthefumehoodundernormalventilationconditionsatthetemperature25±1°C.ThemassdifferencebetweenthedriedfilterpapercontainingtheprecipitateandthemassoffilterpaperbeforefiltrationwasusedasthemassofBaSO4.

4.2.1.InvestigationofthekineticsofsolventextractionThestudyonthekineticsofthesolventextractionwasdonetodeterminethecontacttimeneededbetweenthetwophasestoachieveequilibriumandtoseehowtheequilibriumtimechangeswitheachdiluent.Duringtheinvestigationoftheextractionkineticsthemixingtimewasvariedfrom1to20minutes. Theorganicphasesusedwere0,6MD2EHPA (97%, SigmaAldrich)diluted in in

25

Solvent70(hydrocarbonsC11-C14,≤aromatics,Statoil,Sweden),hexane(95%,anhydrous,SigmaAldrich),Octane (98%, reagentgrade), toluene (99,8%,anhydrous, SigmaAldrich),cyclohexanone(≥99%,ACSreagent,SigmaAldrich),1-octanol(≥99%,ACSreagent,SigmaAldrich) and chloroform (≥ 99,9%, contains amylenes as stabilizer, SigmaAldrich). All theorganicphaseswerepre-equilibratedwiththeequalamountofMQwater.Theexperimentswereperformedin3.5mLglassvialsat25±1°C,aqueoustoorganicphaseratioΘ=1.Alltheexperimentswere done in triplicates. The vials before each samplingwere centrifuged atrotationspeedof2000rpmfor1minute.Thesampledaqueousphases,aswellastheaqueousphasesbeforeextractionweredilutedwith0.5MHNO3andanalyzedusingtheICP-OES.Thedistributionratioswerecalculatedasmassbalanceofthesemeasurements.

4.2.2.TheeffectoftheconcentrationofD2EHPAandthediluentonsolventextractionTheeffectoftheconcentrationofD2EHPAanddiluentonthesolventextractionoftheREEsout of the leachate was investigated in order to determine the optimal organic phasecompositionforthesolventextractionprocess.Theextractant,D2EHPA(97%,SigmaAldrich),wasusedinconcentrations0.3,0.6,0.9and1.2MdilutedinSolvent70(hydrocarbonsC11-C14,≤aromatics,Statoil,Sweden),hexane(95%,anhydrous,SigmaAldrich),Octane(98%,reagentgrade),toluene(99,8%,anhydrous,SigmaAldrich),cyclohexanone(≥99%,ACSreagent,SigmaAldrich),1-octanol(≥99%,ACSreagent,SigmaAldrich)andchloroform(≥99,9%,containsamylenesasstabilizer,SigmaAldrich).Alltheorganicphasesusedinthisworkwerepre-equilibratedwithMQwaterbeforeperformingthe experiments. The aqueous phase used for the kinetics experimentswas the obtainedleachatewithoutdilution.IkaVibraxVxrbasicshakingmachine(shakingspeed1750vibrationsperminute)withanadjacentthermostaticwaterbathwasusedfortheshakingexperiments.All theextractionexperimentswereperformed in3.5mLshakingvials (46x13x0.8mm)madeofglass.Thetemperaturewaskeptat25±1°CandtheorganictoaqueousphaseratioΘ=1.Theexperimentswereperformedintriplicatesandtheuncertaintieswereexpressedas±1σ.Thevialsbeforesamplingwerecentrifugedatarotationspeedof2000rpm,andthesampledaqueousphasesweredilutedwith0.5MHNO3andanalyzedusingtheICP-OESwiththesamemethodasoutlinedabove.

26

4.2.3.InvestigationofthepHeffectonextraction

ExtractionwithD2EHPAispHdependent.[7]ThepHoftheleachatewasvariedinordertodeterminetheeffectofthepHontheextractionofthemetalsoutoftheaqueousphase,andtodeterminethepossibilitiesofachievinghigherselectivityamongtheelementspresent.Theorganicphaseusedwas0.3MD2EHPAinSolvent70.TheequilibriumpHafterextractionwithnomodificationwas1.ThepHoftheaqueousphasewasadjustedbyaddingsmallamountsof0.1Mor3MNaOHtoincreasethepHorkonc.H2SO4tolowerthepH.AcertainamountoftheorganicphasewasaddedtotheextractionsystemwhichcorrespondedtotheamountoftheNaOHaqorkonc.H2SO4added.Theequilibrationwasperformedfor20minutesbymanualshakingbeforesamplingtoensurere-equilibration.Thesampledaqueousphasesweredilutedwith0.5MHNO3andanalyzedusingtheICP-OES.ThereportedpHvaluesarethemeasuredprotonactivitiesatthespecificionicstrengthofeachsolution.

4.2.4.StrippingAfterdeterminingtheoptimalextractionconditions,themetalsneedtobestrippedintothenewaqueousphaseforfurtherreprocessing.Strippingwasperformedbyputtingtheorganicphasefromextractionintocontactwith0.5,1,1.5,2,2.5and3MHClhydrochloricacid(37%,puriss,Sigma-Aldrich).Theorganicphaseaftertheextractionwasseparatedfromthemetal-depletedaqueousphaseandputincontactwiththestrippingaqueousphase.Thevolumeof5mLofeachofthephaseswasputincontactina20mLvialandshakenmanuallyfor20minutesatthetemperatureof25±1°Candaphase ratio ofΘ = 1. After the stripping the stripping aqueous phaseswere sampled anddilutedwith0.5MHNO3beforetheanalysiswiththeICP-OES.

27

5.RESULTSANDDISCUSSION

5.1.LEACHINGOFTHEMAGNETPOWDERFOLLOWEDBYSOLVENTEXTRACTIONOFTHEREESWITHTODGA

5.1.1.DeterminationofthecompositionofthemagnetmaterialTheresultsofthetotaldissolutionanalysisareshowninTable4.Thesumislowerthan100%(around95–97%)mostlyduetotheprobableexperimentalerrorandtheinabilitytomeasureelementslikeOandNwhichalsomakeacertainpercentageofthecompositionoftheNdFeBmagnet.Theresultswereinaccordancetotheliteraturevaluesofthepercentageofelementspresentintheneodymiummagnet.[3]Table4.Thecompositionoftheobtainedmaterialwithelementsshowninmasspercentages.The material was totally dissolved in aqua regia and diluted with 0.5 M HNO3. ThemeasurementwasperformedwiththeICP-OESiCAP6500,ThermoFisher(ppmscale).

element mass%

Al 0.95±0.16B 1.00±0.02Co 1.42±0.07Cu 0.22±0.05Dy 1,08±0.27Fe 61.09±1.04Mn 0.15±0.01Nd 25.38±0.66Ni 2.03±0.23Pr 2.62±0.17Si 0.02±0.02

5.1.2.ModelsolutiontestingAmodelsolutioncontaining21.49mMFe,3.46mMNdand0.62mMDyinvariousnitricacidconcentrationwastested.Thenitricacidconcentrationswere0.1,1,2,3,4,5and6MHNO3.TheresultsareshowninFigure9.

28

Figure9.ThedependenceofthedistributionratiosofFe,NdandDyin0.1MTODGAinSolvent70 on initialHNO3 concentrations of 0.1, 1, 2, 3, 4, 5 and6M in the aqueous phase. Thetemperaturewaskeptat25±1°Candtheorganictoaqueousratioat1:1.It canbeobserved fromFigure 9 that no extraction is obtained at 0.1MHNO3.With theincreasing nitric acid concentration to 1M the distribution ratios for Nd and Dy increasedramatically,while thedistribution ratiosofFe stay low.When theHNO3concentration isincreasedto2MtheequilibriumdistributionrationvaluesarereachedforNdandDyandtheyremainunchangedwith increasingoftheNO0=concentration.ThedistributionratiosforFestaylowthroughoutthewholerangeofnitricacidconcentrationanddonotexceed0.4.Thus,theextractionwithTODGAwillbefurthercarriedonatnitricacidconcentrationofaround3MHNO3toensureenoughNO3

-counterionsfortheefficientformationofthecomplexintheorganicphase.Fromequation3 it isclear thatwith increasingtheNO0=concentration, theequilibriumwillbeshiftedtotherighthandside.5.1.3.TheeffectoftheconcentrationofTODGAandthediluentonsolventextractionThediluenteffectonextractionoftheelementsfroma3MHNO3solutionwasinvestigated.The diluents used were Solvent 70, hexane, toluene, cyclohexanone and 1-octanol. TheconcentrationsoftheextractingagentTODGAwere0.01,0.05,0.1,0.2and0.4Minthelisteddiluents.

0,01

0,1

1

10

100

1000

10000

0 1 2 3 4 5 6

D

c(HNO3)/mol/L

Nd

Dy

Fe

29

a)

b)

y(Hex)=2,95x+5,43

y(S70)=2,88x+5,03

y(Cycl)=2,59x+4,52

y(1-Oct)=2,33x+3,42

y(Tol)=2,63x+3,89

-2,5

-1,5

-0,5

0,5

1,5

2,5

3,5

4,5

-2,5 -2 -1,5 -1 -0,5 0

logD

log[TODGA]/mol/L

Nd Hexane

Solvent70

Cyclohexanone

1-Octanol

Toluene

y(Hex)=2,97x+5,41

y(S70)=2,96x+5,04

y(Cycl)=2,84x+5,20

y(1-OCt)=2,28x+3,37y(Tol)=2,54x+3,78

-2

-1,5

-1

-0,5

0

0,5

1

1,5

2

2,5

3

3,5

-2,5 -2 -1,5 -1 -0,5 0

logD

log[TODGA]/mol/L

Pr Hexane

Solvent70

Cyclohexanone

1-Octanol

Toluene

30

c)

Figure 10. The dependence of the distribution ratios of Dy, Nd and Pr after extraction onTODGAconcentrationSolvent70,hexane,1-octanol,cyclohexanoneandtoluene.Theorganictoaqueousratiowaskeptat1:1andthetemperatureat25±1°C.Theaqueousphaseusedwas4000mg/Lofthemagnetleachedin3MHNO3.It is noticed that the distribution ratios for Nd, Pr and Dy increase with the TODGAconcentration.DistributionratiosoftheREEsincreasewiththeconcentrationofTODGAinallthediluents,reachingvaluesupto1000athighestconcentrations.ThedistributionratiosoftheREEsincertaindiluentsdecreaseinthefollowingorder:hexane>cyclohexanone>solvent70 > toluene > 1-octanol. It can be concluded that the distribution ratios for extractablespeciesdecreasewiththepolarityofthediluents,withtheexceptionofcyclohexanone,whichhas an active oxygen donor atomwhich has the ability to form complexes soluble in theorganic phase. [53] This,most probably, led to higher distribution ratios in this particulardiluent.TODGAshowedhighselectivityofextractionoftheREEs. Itwasobservedthatthedistributionratios fortheheavyREEDyarehigherthanthosefor lightREEs,whichcanbeattributed to higher charge density of the heavy REEs, due to smaller ionic radii, whichfacilitatesthecomplexformation.TheveryhighdistributionratiosforREEscanbeattributedtothestrongoxygendonoratomsintheTODGAmolecule.Thedistributionratiosofotherelements in the leachate reach values less than 0.1. The highest distribution ratios wereregisteredforBandAlincyclohexanonewhichreach0.5,butthesehighervaluescouldbeattributedtotheabilityofcyclohexanonetoextracttheseionsoutofthesolutionbyitself.[54]ThegreatestadvantageofTODGAinrecyclingoftheNdFeBmagnets isthe inabilityofthe

y(Hex)=2,04x+5,14y(S70)=1,93x+4,69

y(Cycl)=2,38x+5,45

y(1-Oct)=1,50x+3,25

y(Tol)=2,04x+4,39

0

0,5

1

1,5

2

2,5

3

3,5

-2,5 -2 -1,5 -1 -0,5 0

logD

log[TODGA]/mol/L

Dy Hexane

Solvent70

Cyclohexanone

1-Octanol

Toluene

31

molecules to extract Fe, which is the elementmaking up around 60% of themagnet aspreviouslymentioned,anditisnotinterestedfromarecyclingperspective.InastudybyZhuetal.[32]itwasshownthatdivalentionswithionicradiismallerthan80pmandtrivalentionswith ionicradiismallerthan70pmshowveryweakextractionwithTODGAinn-dodecanewhichisinaccordancewiththisstudy.TheionicradiusthusplaysaroleinsolventextractionwithTODGA,withREEs,havinglargerionicradiibeingprioritizedintheextractionprocess.5.1.4.Slopeanalysis(stoichiometry)WhenlogDvs.logc(TODGA)isplotted,theslopes,m(equation4),forPr,NdandDy,indicatethenumberof ligandmolecules involved in theextractionprocessper singleREE ion.TheslopesarenonintegralindicatingthatmorethanonespecieofthecomplexesofREE:TODGAis created. The slopes for Nd and Pr in the non-polar diluents are very close to 3, whichindicates that the majority of the species formed were Nd(Pr) (NO3

-)n TODGA3. It can beobserved that the slopes decrease with the polarity of the used diluents. The slopes fortoluene are close to 2.5 indicating that speciesNd(Pr) (NO3

-)n TODGA2 andNd(Pr) (NO3-)n

TODGA3wereformed.Thelowestslopesvaluesclosinginonthevalueof2wereobtainedfor1-octanol.Sincethesolubilityofthetheextractantinthediluentplaysapivotalroleintheextraction process, the polarity of the diluent in this case could have contributed to thenumberofTODGAmoleculesusedinthecomplexformation.[29]ThisisprobablyduetotheinteractionbetweenthedonoratomsintheDGA(diglycolamide)groupandthearomaticandotherdonoratoms(oxygen)inthepolardiluents.AsforDy,theslopesareveryclosetothevalueof2inallthediluents,meaningthatthespeciesDy(NO3

-)n(TODGA)2areformed,withthesamepolarityvs.slopetrend.However,duetothelimitedknowledgeabouttheactivityoftheextractantintheorganicphasethemethodhaslimitations.ThecomplexesassuchshouldbeanalyzedbyXAFSsincetheslopeanalysismethodgivesonlyaroughideaaboutthestructure.Themainconclusionisthatroughly2-3moleculesofTODGAareinvolvedwiththecomplexformationwithREEsions,whichisinaccordancewithpreviousstudies.[29,31]5.1.5.SeparationfactorsTheseparationfactorsbetweenREEsandotherelementswerenotcalculatedbecausenoneof the distribution ratios of the elements except REEs reach a value above 0.1 (except incyclohexanone). The focus was on the extraction of the REEs from one another. TheseparationfactorsbetweenDy(HREE)andNdandPr(LREE)werecalculated,sinceNdandPrshowedsimilardistributionratiosandDyshowedhigherdistributionratios (getsextractedbetter)fromthesolution.TheseparationfactorsareshowninTable5.ThehighestseparationfactorsbetweenDyandandlightREEwerereachedin0.01MTODGAinSolvent70,atwhich

32

Dygetsextractedselectively fromotherelementsandthe lightREE.VerygoodseparationbetweenDyandotherelementsisreachedat0.01MTODGAinhexaneand1-octanol,butreachvaluesashighas14inotherdiluentsaswell.Ageneralconclusionwouldbe:1)VerygoodseparationfactorsforDyat0,01MTODGASolvent702)REEsextractedselectivelyandcompletelyasagroupat0,1MTODGAinallthediluentsTable5.SeparationfactorfortheREEsandotherelementsatextractionwith0.01,0.05and0.1MTODGA inSolvent70,hexane,1-octanol,cyclohexanoneandtoluene.Theorganic toaqueousratiowaskeptat1:1andthetemperatureat25±1°C.Theaqueousphaseusedwas4000mg/Lofthemagnetleachedin3MHNO3

SOLVENT70c(TODGA)/

MαDy/Nd αDy/Pr

0.01 39.0±1.1 51.7±2.1 0.05 6.2±0.6 9.0±0.7 0.1 4.8±0.4 5.0±0.3 0.2 4.4±0.6 5.0±0.2 0.4 2.5±0.4 2.7±0.3

HEXANEc(TODGA)/

MαDy/Nd αDy/Pr

0.01 30.8±0.5 37.7±0.3 0.05 9.2±0.2 10.4±0.7 0.1 3.6±0.5 4.2±0.6 0.2 1.4±0.1 1.5±0.3 0.4 1.5±0.2 1.6±0.3

CYCLOHEXANONEc(TODGA)/

MαDy/Nd αDy/Pr

0.01 14.4±0.6 14.2±0.7 0.05 11.8±0.3 7.3±0.9 0.1 12.1±0.6 4.9±0.5 0.2 3.8±0.5 1.8±0.6 0.4 1.8±0.2 1.4±0.4

33

1-OCTANOLc(TODGA)/

MαDy/Nd αDy/Pr

0.01 30.0±1.5 26.3±0.9 0.05 11.7±0.3 12.9±0.6 0.1 4.2±0.5 3.9±0.1 0.2 2.5±0.3 2.6±0.2 0.4 2.5±0.4 2.3±0.2

TOLUENEc(TODGA)/

MαDy/Nd αDy/Pr

0.01 14.0±0.7 13.2±0.4 0.05 18.4±0.3 20.0±0.8 0.1 14.2±0.2 15.8±0.6 0.2 7.4±0.4 7.7±0.1 0.4 2.2±0.2 2.3±0.6

5.1.6.StrippingThestrippingwasconductedwith0.01,0.1MHNO3andMQwater.Theorganicphaseschosenforthestrippingwere0.01MTODGAinSolvent70afterextractionfortherecoveryofDythatwasselectivelyextractedfromtheleachate,and0.1MTODGAinhexanewhereNd,PrandDywereextractedasagroup.Theorganicphasewastakenaftertheextractionandputincontactaftertheextractionwiththeequalamountoftheaqueousphase1:1for50minutesat1300shakesperminuteatroomtemperature25±1°C.Theresultsofthestrippingaregivingin%metalstripped.Table6.StrippingofDyoutof0.01MTODGAinSolvent70with0.01M,0.1MHNO3andMQwateraftershaking50minutes.A:Okeptat1:1for50minutesat1300shakesperminuteatroomtemperature25±1°C.

Strippingaqphase Dystrippingefficiency/%MQwater 97.1±0.9

0.01MHNO3 91.3±0.80.1MHNO3 56.2±1.3

34

Table7.StrippingofNd,PrandDyoutof0.1MTODGAinhexanewith0.01,0.1MHNO3andMQwateraftershaking50minutes.A :Okeptat1 :1 for50minutesat1300shakesperminuteatroomtemperature25±1°C.

Strippingaqphase Dystrippingefficiency/%

Ndstrippingefficiency/%

Prstrippingefficiency/%

MQwater 96.1±0.9 97.2±0.8 97.0±1.20.01MHNO3 91.3±0.8 89.1±1.3 91.6±0.90.1MHNO3 56.2±1.3 49.2±0.9 48.2±0.9

The stripping of the metals out of the solution after solvent extraction is crucial for therecoveryof themetalsof interest. SinceTODGAextractswithhigherdistribution ratios inhighernitricacidconcentrations, it isexpected that thestrippingwillbe feasiblewith lessacidicmediaorevenMQwater.Thusthestrippingwasconductedwith0.01,0.1MHNO3andMQwater.Therecoveryofthemetalswasnot100%,but itreachedalmost100%inMQwaterforNd,PrandDy.0.1MHNO3preformedreallybadwithonly50%ofmetalsstrippedIncompleterecoverywasachievedwith0.01MHNO3.ItwasshowedthatusingonlyMQwaterwith no NO3

- couter-ions was enough to break the complex from the organic phase andachievehighstrippingefficiency.AsignificantdropofpHofthestrippingagentstothevalueofaround1andlesswasobserved.ThisindicatesthattheHNO3extractedwiththesolventextractionwithTODGAwasstrippedusing these stripping agents. No precipitation was observed in the aqueous phase afterstripping.

5.2.SOLVENTEXTRACTIONOFTHEREESWITHD2EHPAFROMALEACHATEOBTAINEDBYSELECTIVELEACHING5.2.1DeterminationofthecompositionoftheleachateThemeasurable/detectableconcentrationsofNd,Dy,Pr,Gd,CoandB in the leachatearegiveninTable8.

35

Table8.ConcentrationsofthemetalsofinterestintheleachatemeasuredwiththeICP-OES.ThemeasurementwasperformedwiththeICP-OESafterdilutingtheobtainedleachatewith0.5MHNO3

Element Concentration/mM

Nd

Dy

Pr

Gd

Co

B

other

9.1±0.9

2.7±0.6

3.2±0.4

0.69±0.16

0.17±0.09

0.55±0.14

belowdetectionlimit

The ionic strength of the solutionwas determined to be I= 0.12M. The pH value of thesolutionwasmeasuredwithMeterLabTMPHM240pH/IonMeterpHelectrode and itwasdeterminedtobearound5.TheconcentrationofthesulfateionwasdeterminedaftertheprecipitationofthesulfateionusingaBaCl2solution,andwasdeterminedtobe[SO4

2-]=22.62±1.23mMandwaskeptconstantthroughoutthisresearch.

5.2.2.InvestigationofthekineticsofsolventextractionTheinvestigationoftheextractionkineticswasperformedwiththeaforementionedleachate.Theorganicphasewas0.6MD2EHPAinSolvent70,hexane,octane,cyclohexane,1-octanol,toluene and chloroform. From Figure 11 it can be seen that the time required for thedistributionratiosfortheREEs(Nd,Pr,GdandDy)betweentheorganicandaqueousphaseisreached within 3minutes in Solvent 70, hexane, octane, toluene and 1-octanol, while incyclohexanone and chloroform it took 5 minutes for the equilibrium to be reached. Noextractionofcobaltwhatsoeverwasobservedatanyoftheconditions,whichshowsthatREEscan be extracted selectively without any cobalt transfer into the organic phase. Borondistributionratiosreachamere0.1inSolvent70after5minutes,andsomeminorextractionin1-Octanolwhilenootherextractionwasregisteredinotherdiluents.AconclusioncanbedrawnthatD2EHPAattheseconditionsshowsgoodselectivitybetweenREEsandtheBandCo.

36

a)

b)

0

1

2

3

4

5

6

7

0 5 10 15 20 25

D

t/min

Nd

Solvent70

Hexane

Octane

Toluene

1-Octane

Chloroform

Cyclohexanone

0

1

2

3

4

5

6

7

0 5 10 15 20 25

D

t/min

Pr

Solvent70

Hexane

Octane

Toluene

1-Octanol

Chloroform

Cyclohexanone

37

c)

d)

0

40

80

120

160

0 5 10 15 20 25

D

t/min

Dy

Solvent70

Hexane

Octane

Toluene

Chloroform

Cyclohexanone

1-Octanol

0

40

80

120

0 5 10 15 20 25

D

t/min

Gd

Solvent70

Hexane

Octane

Toluene

1-Octanol

Chloroform

Cyclohexanone

38

e)

f)

Figure11.DistributionratiosofNd,Pr,Dy,Gd,CoandBintheleachatesolutionplottedasafunctionoftimewhichwasvariedbetween1and20minutes.Theorganicphaseswere0.6MD2EHPAinSolvent70,hexane,octane,toluene,1-octanol,cyclohexanoneandchloroform.Theextractionconditionswere25±1°CandΘ=1Theuncertaintybarsrepresentthestandarddeviationofatriplicatetest.

0

0,2

0,4

0,6

0,8

1

0 5 10 15 20 25

D

t/min

B

Solvent70

Hexane

Octane

Toluene

1-Octanol

Chloroform

Cyclohexanone

0

0,2

0,4

0,6

0,8

1

0 5 10 15 20 25

D

t/min

Co

Solvent70

Hexane

Octane

Toluene

1-Octanol

Chloroform

Cyclohexanone

39

5.2.3.TheeffectoftheconcentrationofD2EHPAandthediluentonsolventextraction

TheinfluenceofD2EHPAconcentrationontheextractionofNd,Pr,Dyinvariousdiluentsis

showninFigure12.

a)

b)

y(S70)=2,45x+2,10

y(Hex)=2,45x+1,94

y(Oct)=2,42x+2,00

y(Chl)=1,52x+0,22

y(Cycl)=1,07x- 0,034

y(1-Oct)=2,26x- 0,08

y(Tol)=2,11x+0,24-2

-1,5

-1

-0,5

0

0,5

1

1,5

2

-0,9 -0,8 -0,7 -0,6 -0,5 -0,4 -0,3 -0,2 -0,1 0

logD

log[(D2EHPA)2]

Nd Solvent70HexaneOctaneChloroformCyclohexanone1-OctanolToluene

y(S70)=2,48x+2,14

y(Hex)=2,61x+2,00

y(Oct)=2,60x+2,05

y(Chl)=2,07x+0,50

y(Cycl)=1,12x+0,36

y(1-Oct)=2,08x+0,63

y(Tol)=1,97x+0,84-1,5

-1

-0,5

0

0,5

1

1,5

2

-0,9 -0,8 -0,7 -0,6 -0,5 -0,4 -0,3 -0,2 -0,1 0

logD

log[(D2EHPA)2]

PrSolvent70

Hexane

Octane

Chloroform

Cyclohexanone

1-Octanol

Toluene

40

c)

d)

Figure12.TheinfluenceofD2EHPAconcentrationontheextractionofNd,Pr,GdandDyfromtheaqueousconsistingof9.11mmNd,2.71mMDy,3.16mmPr,0.69mMGd0.17mMCo0.55mMB.usingdifferentconcentrationof0.3,0.6,0.9and1.2MD2EHPAinSolvent70,hexane,octane,toluene,cyclohexanone,chloroformand1-octanol.Thetemperaturewaskeptat25±

y(S70)=1,24x+2,54

y(Hex)=1,47x+2,69

y(Oct)=1,44x+2,66

y(Chl)=1,82x+1,71

y(Cycl)=1,04x+1,56

y(1-Oct)=1,61x+1,51

y(Tol)=1,38x+1,94

0,00

0,50

1,00

1,50

2,00

2,50

3,00

-0,9 -0,8 -0,7 -0,6 -0,5 -0,4 -0,3 -0,2 -0,1 0

logD

log[(D2EHPA)2]

Gd Solvent70HexaneOctaneChloroformCyclohexanone1-OctanolToluene

y(S70)=2,31x+2,80

y(Hex)=2,23x+2,60

y(Oct)=2,34x+2,68

y(Chl)=1,60x+0,35

y(Cycl)=1,17x+0,98

y(1-Oct)=1,44x+1,00

y(Tol)=1,90x+1,32-1,50

-1,00

-0,50

0,00

0,50

1,00

1,50

2,00

2,50

-0,9 -0,8 -0,7 -0,6 -0,5 -0,4 -0,3 -0,2 -0,1 0

logD

log[(D2EHPA)2]

DySolvent70HexaneOctaneChloroformCyclohexanone1-OctanolToluene

41

1°CandtheorganictoaqueousphaseratioΘ=1.Theerrorbarsrepresentthestandarddeviationofatriplicatetest.ThedistributionratiosoftheREEsincreaseswiththeincreaseoftheconcentrationofD2EHPAinallthediluents.TheextractantclearlyshowshigherefficiencyforheavyREEs(DyandGd)overthelightREEs(NdandPr)inallthediluents.SimilartrendswerepreviouslyobservedinastudybyMohammadietal.[39]Thedistributionratiosarehigherinthealiphaticdiluents(hexane,octane,solvent70),followedbytolueneandshowingthelowestefficiency inthepolardiluents(cyclohexanone,1-octanolandchloroform).Cobaltdoesnotgetextractedatany of the investigated conditions and the distribution ratios remained around 0. ThedistributionratiosforBarelowatallconcentrationsanddiluents.Itgetsextractedupto10%in cyclohexanone and 1-octanol. This can be attributed to the carbonyl group incyclohexanone and the hydroxyl group in 1-octanol that could actually allow the diluentmoleculestoformcomplexeswithBsolubleintheorganicphase.Itisobservedthatat0.9and1.2MD2EHPAinhexaneandoctane100%ofalltheREEsgettransferredintotheorganicphase,whilenoCoandBgetco-extracted.Asmentionedbefore,thedistributionratiosarehigherfortheheavyREEsthanthelightREEswhichcorrespondstothedeceasingionicradiioftheseelements.TheionicradiiofhydrolyzedPr3+,Nd3+,Gd3+ andDy3+ are, accordingly: 0.99Å, 0.983Å, 0.938Å and0.912Å [10]. Theoutcomewillbeincreasedchargedensitywiththedecreasingionradius.ThesmallerioncancontributetotheREE3+bindingtotheD2EHPAmolecule,whichworksonacation-exchangemechanism,favoringtheextractionoftheheavyREEsoverthelightREEs.AccordingtotheHSABtheory,whichdividestheacidsandbasesonhardandsoft[11],Co2+isaborderlineacid,meaningitcanactasasoftorhardacid,anditisthusdifficulttopredictit’sbehaviorsinceitcanbeinfluencedbyvariousfactors,likeremovalofhydratewater,stericeffectsetc.InthiscaseCo2+wasassoftacidwhichdidnotformacomplexwiththeD2EHPAfromtheorganicphase.TogetaclearerpictureofhowdiluentsaffecttheextractionofdifferentREEs,agraphshowingthedistributionratiosofNdin0.9MD2EHPAwasmadevs.thedielectricconstants(ε)ofthediluent(Figure13).

42

Figure13.DistributionratioofNdplottedas the functionof thedielectricconstantsof thediluents.Theorganicphaseusedwas0.9MD2EHPAdiluted inSolvent70,hexane,octane,toluene,cyclohexanone,chloroformand1-octanol.Theaqueousphaseconsistedof9.11mmNd,2.71mMDy,3.16mmPr,0.69mMGd0.17mMCo0.55mMB.Theextractionconditionswere25±1°CandΘ=1.ThedistributionratiosofNdin0.9MD2EHPAdilutedinvarioussolventsaredepictedinFigure13.Thedistribution ratiosofNdshowa regular trend,decreasingwith thepolarityof thediluent intheorderSolvent70>octane>hexane>toluene>cyclohexanone>1-octanol>chloroform.Thedistributionratiosareoneorderhigher inthealiphaticnon-polardiluents(Solvent 70, hexane, octane) than in aromatic andpolar ones (toluene, cyclohexanone, 1-octanol and chloroform). Considering the D2EHPA being a relatively non-polar molecule,owingittothe2-ethylhexylchains,couldexplaingoodsolubilityoftheD2EHPAextractantinthealiphaticnon-polardiluents.thusleadingtolessaggregationoftheextractantmoleculesandconsequently leadingtohigherdistributionratioforNd.ThisexplanationcouldnotbeusedforD2EHPAdilutedwithcyclohexanone,whichgivesdistributionratios forNdhigherthanotherpolardiluents(1-octanolandchloroform)eventhoughithasthelargestdielectricconstant (18.3). Nonetheless, this extraction-enhancing phenomenonwas expected, sincecyclohexanoneisamoleculewithadonoratom,oxygen,forhydrogenbonding,butnoactivehydrogenatoms.ThiscouldcauseittoformcomplexeswiththeREEionsinthesolution,whichcouldbethentransferredintotheorganicphase.[54]ThisphenomenoncouldalsoexplaintheextractionofboronintotheorganicphaseseeninFigure11e).

0

2

4

6

8

10

12

14

16

18

20

0 5 10 15 20

D(Nd)

dielectric constant, ε

Solvent70

Hexane

Octane

Toluene

Cyclohexanone

1-Octanol

Chloroform

43

5.2.4.Stoichiometry The stoichiometric ratios between REEs and the extractant were determined byplottingthelogDvs.log[(D2EHPA)2]andthendoingthelinearregressiontodeterminethestoichiometricratiobetweentheelementandtheextractant.TheresultsareshowninFigure12.ItcanbeobservedfromthegraphsthatforNd,PrandDyinhexane,octaneandsolvent70theslopesshowvaluesofaround2.5.ThismeansthatD2EHPAwillformcomplexeswith2and3D2EHPAdimers,meaningthecomplexeswilllooklikeMR3(HR)3andMR3(HR),whereMrepresentsthelanthanideion.InthetoluenecasetheslopesforNd,PrandDyareapproaching2,whichindicatestheformationofacomplexMR3(HR).AsforGd,theslopesforthealiphaticdiluentsgraphsareclosetothevalueof1.5,indicatingthecomplexationoftheseREEswith1and2dimersoftheD2EHPAmolecules.In1-Octanol1-2molecules.Gdformscomplexeswitharound1dimerofD2EHPA.ThecomplexesofREEsincyclohexanoneseemtoformcomplexeswithoneD2EHPAdimer.And1-2inchloroform.Itisobservedthatthenumberofmoleculesneededtoformacomplexdecreasewiththeincreasedpolarityofthediluent,whichcanbeduetotheinteractionoftheactivegroupswiththeoxygeninthephosphoricacidgroupinD2EHPA.[29]5.2.5.InvestigationofthepHeffectonextraction As expected, the increase of pH (lower proton concentration) led to highermetalextraction.ThisisatypicalbehaviorforacidicextractantslikeD2EHPAanditcanalsobeseeninequation4thatlowerpHvalueswillshifttheequilibriumtothelefthandside.Ontheotherhand,theincreaseinprotonconcentrationwillfavorthestrippingreaction.ItwasobservedthatbyincreasingthepHvaluesDyandGdgetextractedfirst,followedbyNdandPr.AtpHaround1noextractionofCoandBwhatsoevercanbeobserved,whilewhenincreasingthepHvaluetothevalueof2around10%ofBisextractedandaround20%ofCoisextracted,so to avoid the co-extraction of the exogenes with the REEs as much as possible theequilibriumpHshouldbekeptatorbelowthevalueof1.TheresultsalsoshowthatthelightREEs(NdandPr)willbeco-extractedwiththeheavyREEsinthefirststepofanymethodusedwiththeratiosaround100%:50%=heavy:light.IncreasingtheequilibriumpHupto2willleadtocompleteextractionoftheREEsinthesolutionintotheorganicphase,butwiththedisadvantageoftheincreasedco-extractionofCoandBinminorpercentages.Thiscouldleadtomoreextractionstagesina,forexample,mixer-settler.

44

Figure14.PercentageoftheNd,Pr,Dy,Gd,CoandBextractedfromtheneodymiummagnetleachateconsistingof9.11mmNd,2.71mMDy,3.16mmPr,0.69mMGd0.17mMCo0.55mMBwithvariedequilibriumpH.Theextractionconditionswere25±1°CandΘ=1.Theorganicphaseusedwas0.3MD2EHPAinSolvent70.

5.2.6.SeparationfactorsbetweenheavyandlightREEsTofurtherinvestigatetheselectivitybetweenlightandtheheavyREEoftheseextractions,the separation factors were calculated. The calculated separation factors are shownnumericallyinTable9.Table 9. Separation factors between the heavy and the light REE after extraction. Theconcentrations ofD2EHPAwere 0.3, 0.6, 0.9 and1.2mol/L in Solvent 70, hexane, octane,toluene,cyclohexanone,chloroformand1-octanol.Theaqueousphaseconsistedof9.11mmNd,2.71mMDy,3.16mmPr,0.69mMGd0.17mMCo0.55mMB.Thetemperaturewaskeptat25±1°CandtheorganictoaqueousphaseratioΘ=1.

0

20

40

60

80

100

0 0,5 1 1,5 2

E/%

equilibriumpH

Nd

Pr

Dy

Gd

B

Co

45

SOLVENT70 c(D2EHPA)/M αDy/Nd αDy/Pr αGd/Nd αGd/Pr

0.3 6.55±3.2 5.85±3.1 27.7±1.2 24.8±0.9 0.6 5.34±0.7 5.43±0.5 13.7±0.2 14.0±0.3 0.9 7.3±0.1 7.4±0.2 6.1±0.6 6.2±0.2 1.2 4.6±0.4 3.8±0.5 5.8±0.5 4.7±0.4

HEXANE c(D2EHPA)/M αDy/Nd αDy/Pr αGd/Nd αGd/Pr

0.3 6.8±0.6 7.6±0.8 37.7±1.2 42.4±1.6 0.6 6.3±0.7 7.27±0.4 20.6±0.9 23.7±1.2 0.9 5.3±0.6 5.94±0.3 10.5±0.7 11.9±0.5 1.2 5.1±0.3 4.28±0.6 10.9±0.5 9.2±0.4

OCTANE c(D2EHPA)/M αDy/Nd αDy/Pr αGd/Nd αGd/Pr

0.3 5.8±0.2 6.1±0.4 33.1±1.2 38.3±0.9 0.6 6.5±0.4 7.6±0.3 15.7±0.9 18.5±0.7 0.9 4.8±0.3 5.4±0.7 7.9±0.6 8.9±0.5 1.2 5.0±0.3 4.2±0.6 10.2±0.9 8.6±0.4

TOLUENE c(D2EHPA)/M αDy/Nd αDy/Pr αGd/Nd αGd/Pr

0.3 6.5±0.7 4.8±0.9 36.8±0.9 34.5±0.7 0.6 2.0±0.4 2.1±0.1 27.9±0.6 29.3±0.5 0.9 1.9±0.3 2.1±0.3 26.7±0.9 29.5±0.5 1.2 6.7±0.4 5.5±0.8 14.8±0.6 12.2±0.7

CYCLOHEXANONE c(D2EHPA)/M αDy/Nd αDy/Pr αGd/Nd αGd/Pr

0.3 5.9±0.3 4.9±0.3 17.0±0.9 14.3±0.7 0.6 2.5±0.4 2.6±0.1 25.1±0.7 25.8±0.2 0.9 2.7±0.2 3.0±0.3 22.7±0.2 24.4±0.4 1.2 9.0±0.6 6.3±0.5 14.8±1.2 10.4±0.9

CHLOROFORM c(D2EHPA)/M αDy/Nd αDy/Pr αGd/Nd αGd/Pr

0.3 5.2±0.5 6.0±0.1 25.0±0.2 26.8±0.4 0.6 3.5±0.4 2.7±0.5 26.2±0.9 20.5±0.4 0.9 1.5±0.1 1.4±0.3 26.1±0.6 24.1±0.5 1.2 0.9±0.3 0.9±0.1 17.0±0.3 17.2±0.3

46

1-OCTANOL c(D2EHPA)/M αDy/Nd αDy/Pr αGd/Nd αGd/Pr 0.3 4.9±0.6 5.1±0.8 29.0±1.2 26.9±0.8 0.6 3.9±0.4 3.3±0.1 16.3±0.7 14.0±0.9 0.9 2.7±0.3 2.5±0.6 15.9±0.5 14.8±0.8 1.2 8.5±0.9 5.8±0.8 10.9±1.0 7.34±0.9

ThehighestseparationfactorsbetweenDyandNdwereachievedat0.3%D2EHPAinhexane.Intheseconditionsaround50%ofNdandPrgotextracted,andaround99%ofDyandGdextraction was registered. The α Dy / (Nd & Pr) keeps decreasing with the increasingconcentrationofD2EHPAforSolvent70,whileitstaysbelow10forotherconcentrationsofD2EHPAforall theotherdiluents.AsforGd/Nd,andGd/Prthebestseparationfactorwasfoundtobealsoin0.3Mhexane.ItcanbenoticedthattheseparationfactorsbetweenDyandotherlightREEsareoneorderlowerthanthoseofGdandotherlightREEs.EventhoughitisexpectedforDyandGdtohavesimilardistributionratios,butasitcanbeseenfromTable9., the concentration of Dy is four times higher than that of Gd in the leachate. ThisconcentrationdifferencemighthaveledtothemuchhigherdistributionratiosforGdthanthose of Dy, leading consequently to these results. Since Dy is present in amuch higherconcentrationanditscurrentpriceiscurrentlyaround10timeshigherthanthatofGd,theorganicphaseconsistingof0.3MD2EHPAdilutedinhexanewaschosenasthebestorganicphase from separation Dy and Gd from other light REEs, especially considering the smallamountsofBandCobeingextracted.Futureworkonthescale-upprocessinamixersettlerisproposed.

5.2.7.Stripping

Thestrippingexperimentswereconductedfortheorganicphasesafterextractionwith0.3MD2EHPAinSolvent70and1.2MD2EHPAinoctane.At0.3MD2EHPAinSolvent70thehighestseparationfactorsbetweentheheavyandthelightREEswasachievedandat1.2MD2EHPAtheREEswerecompletelyextractedoutoftheleachateasagroupleavingtheexogenmetals(Co,B)inthesolution,sothesetwocasesshowedtobethemostinterestingforfutureprocessdevelopmentsifneeded.Theelementswerestrippedbackintothenewaqueousphaseusinghydrochloricacid,atconcentration0.5,1,1.5,2,2.5and3Mfor5minutes.Theorganicphaseafter extraction was put in contact with the hydrochloric acid in the above listedconcentrationsandshakenwithIkaVibraxVxrshakingmachinetemperatureof25±1°CwiththeO:A=1:1.TheresultsofthestrippingareshowninFigure15.

47

a)

b)

Figure15.ThestrippingoftheREEsbackintheaqueoussolutionfrom0.3MD2EHPAinhexaneand 1.2 M D2EHPA in octane after extraction. The stripping agents used were aqueoussolutionsof0.5,1,1.5,2,2.5and3Mhydrochloricacid.Thestrippingconditionswerethetemperatureof25±1°CandtheorganictoaqueousphaseratioΘ=1.Shakingtimewas20minutes.Themetalsarestrippedat100%efficiencyat2MHClorhigher,whichisinaccordancewithequation3,meaningthatincreasingtheconcentrationofH+ionsintheaqueoussolution,thereactionwillbeshiftedtothelefthandside.

0

20

40

60

80

100

0 0,5 1 1,5 2 2,5 3 3,5

%m

etalstripped

c(HCl) /mol/L

Stripping from0.3MD2EHPA/Hexane

%Nd

%Pr

%Dy

%Gd

0

20

40

60

80

100

0 0,5 1 1,5 2 2,5 3 3,5

%m

etalstripped

c(HCl) /mol/L

Stripping from1.2MD2EHPA/Octane

%Nd

%Pr

%Dy

%Gd

48

Theseweretheobservationsmadeonasmall(3.5mL)scaleandforfurtherdevelopingtheprocess larger amounts of the leachate are required and mixer settler studies arerecommendedwith0.3MD2EHPAinhexaneand1.2MofD2EHPAinoctanetodevelopalargescaleprocessinthefuture.

49

6.CONCLUSIONSThegoalofthisworkwastodeterminethefeasibilityoftheuseofTODGAandD2EHPAfortheselectiveextractionofREEsoutofrealleachatesobtainedbydissolvingtheneodymiummagnetwaste.ThefocuswasontheeffectofthecompositionoftheorganicphaseontheselectiveextractionoftheREEsoutoftheleachatesandfromeachother.TODGA invariousdiluentswasused for theextractionof theREEsoutof theneodymiummagnet leachate obtained by dissolving the magnet waste in nitric acid. This study wasfocusedon selectiveextractionofREEs fromotherelements toachieve thebest recoverypossible.TheseparationfactorsbetweentheREEswerealsocalculated.Itwasfoundthatthenitricacidconcentrationover2MHNO3wasneededforachievingthehighestdistributionratios for the REEs. As expected the distribution ratios of the REEs increase with theconcentrationofTODGAinallthediluents.HighestseparationfactorsforDywereobservedat0,01MTODGAinSolvent70.REEswereextractedselectivelyasagroupat0,1MTODGAinhexane,solvent70andcyclohexanonewith theotherelementsnotexceedingdistributionratiosover0.1,exceptforcyclohexanonewherethedistributionratiosreachupto0.5.Thestrippingof theseelementswasefficientwithMQwateroutof theorganicphase.TwotothreeTODGAmoleculeswereusedfortheformationofthecomplexwithREEsaccordingtotheslopeanalysisstudy.D2EHPAwas used for the extraction of the REEs from a real leachate obtained from theneodymium magnet leachate in sulfuric acid media, and find most suitable separationconditionsbetweenthem.ThegeneralextractionorderisHREEs>LREEswhichwasexpected.ThedistributionratiosincreasedwiththeD2EHPAconcentrationandtheloweringofacidity.Thekineticsoftheextractionwasmonitoredfor0.6MD2EHPAinSolvent70,hexane,octane,chloroform, cyclohexanone, 1-octanol and toluene. It was shown that the equilibrium isreachedwithin3minutesforallthediluentsexceptforcyclohexanoneand1-octanol,inwhichtheequilibriumisreachedwithin5minutes.It was concluded that the most suitable diluents for the solvent extraction of REEs withD2EHPA were hexane, octane and Solvent 70. The slightly polar ones, cyclohexanone,chloroform and 1-octanol showed the lowest distribution ratios for the extracted REEsToluenelayssomewhereinbetweenofthe2groupspreviouslymentionedaccordingtotheobtained distribution ratios of the REEs. It has also been shown that at 0.9M and 1.2MD2EHPAconcentration inhexaneandoctanetheREEsarecompletelyextractedoutoftheleachatesolutionasagroupleavingtheexogenes(CoandB)inthesolution.Concerningtheseparationbetweentheheavyandthe lightREEs,thebestseparationfactorsfactorswereobtainedatextractionwith0.3MD2EHPAinhexane.AtsuchconditionsalmostalltheDyand

50

Gdwereextractedfromthesolution,whilearoundhalfoftheamount(50%)ofthepresentlightREEswereextracted.Thestrippingstudieswereperformedfromtheorganicphasesaftertheextractionwith0.3MD2EHPAinhexaneand1.2MD2EHPAinoctane.Ithasbeenshownthatthecompletestrippingoftheelementsbackintotheaqueousphaseisachievedat2MHClorhigher.Theseweretheobservationsmadeonasmall(3.5mL)scaleandforfurtherdevelopingtheprocesslargeramountsoftheleachatearerequiredandmixersettlerstudiesareneededtodevelopalargescaleprocessinthefuture.

51

7.FUTUREWORKAfewpathscouldbefollowedforfutureresearchdependingontheareaofinterest.Thesecanbesummedupintothefollowing.Thescalingupoftheprocessusingsolventextractionequipment likemixersettlers.Theredifferentparameterslikeflows,phasecontacttimeandmixingandorganictoaqueousphaseratioscouldbealteredinordertotailortheseparationbetweentheelementsintheleachate.Furthersmall-scalestudiescouldbedonewheretheinterfacialtensionbetweenthesolventand the aqueous phase and the density of the solvent effect on the extraction could bemonitoredinordertodeterminetheeffectofchangingtheseparametersontheoutcomeoftheprocess.XAFSmeasurementsforthedeterminationofthecomplexstructureanddetailedmodellingofthecomplexesformedintheorganicphase.Thermodynamicsoftheextraction.

52

8.AKNOWLEDGEMENTS Iwouldliketothank:

• MymainsupervisorTeodoraReteganandassistantsupervisorBritt-MarieSteenariforthehelpwiththeprojectandmorestuffthatcamealong.

• My examiner Christian Ekberg for the help with the data analysis and scientificdiscussions.

• My office and work colleagues, fellow PhD students and friends for the niceconversationsandoverallhighlevelofjoythroughthickandthin.

• MarieCurieInitialTrainingNetworkEREANproject(EUFP7)forfundingtheproject.• My EREAN project colleagues for the nice time spent on project meetings and

workshopsallacrossEurope.• Lastbutnotleast,mymomandtherestofthefamilyforalwaysbeingsupportiveof

all the choices that I havemade,even though theymightnothavebeen themostreasonableormostlogicalattimes.

53

9.REFERENCES1. K.Binnemans,P.T.J.,B.Blanpain,T.VanGerven,Y.Yang,A.Walton,M.Buchert,

Recyclingofrareearths:acriticalreview.JournalofCleanerProduction,2013.51(0):p.1-22.

2. EuropeanCommissionCriticalrawmaterialsfortheEU,inReportoftheAd-hocWorkingGroupondefiningcriticalrawmaterials.2014.

3. C.Tunsu,M.P.,M.Gergorić,C.Ekberg,T.Retegan,Reclaimingrareearthelementsfromend-of-lifeproducts:Areviewoftheperspectivesforurbanminingusinghydrometallurgicalunitoperations.Hydrometallurgy,2015.156:p.239-258.

4. H.M.Bandara,D.D.,J.W.Apelian,D.Emmert,H.Marion,RareEarthRecycling:ForecastofRecoverableNdfromShredderScrapandInfluenceofRecyclingRatesonPriceVolatility.JournalofSustainableMetallurgy,2015.1(3):p.179-188.

5. E.G.Polyakov,A.S.S.,RecyclingRare-Earth-MetalWastesbyPyrometallurgicalMethods.Metallurgist,2015.59(5):p.368-373.

6. H.M.Bandara,D.D.,J.W.Apelian,D.Emmert,H.Marion,ValueAnalysisofNeodymiumContentinShredderFeed:TowardEnablingtheFeasibilityofRareEarthMagnetRecycling.EnvironmentalScience&Technology,2014.48(12):p.6553-6560.

7. F.Xie,T.A.Z.,D.DreisingerandF.Doyle,Acriticalreviewonsolventextractionofrareearthsfromaqueoussolutions.MineralsEngineering,2014.56:p.10-28.

8. Voncken,J.H.L.,TheRareEarthElements,AnIntrodutcion2016:Springer.9. J.S.Kim,B.N.K.,S.Radhika,M.L.Kantan,B.R.Reddy,Studiesonselectionofsolvent

extractantsystemfortheseparationoftrivalentSm,Gd,DyandYfromchloridesolutions.InternationalJournalofMineralProcessing,2012.112–113:p.37-42.

10. Cotton,S.,Lanthanideandactinidechemistry.2006,Chichester:JohnWiley&Sons,Ltd.

11. J.Rydberg,M.C.,C.MusicasandG.R.Choppin,SolventExtractionPrinciplesandPractice.2004:Taylor&Francis.

12. G.B.Haxel,J.B.H.a.G.J.O.,RareEartsElements-CriticalResourcesforHighTechnology.2002,USGeologicalSurvey.

13. B.Swain,E.O.O.,CompetitiveextractionoflanthanidesbysolventextractionusingCyanex272:Analysis,classificationandmechanism.SeparationandPurificationTechnology,2011.83:p.82-90.

14. J.H.Rademaker,R.K.,Y.Yongxiang,RecyclingasaStrategyagainstRareEarthElementCriticality:ASystemicEvaluationofthePotentialYieldofNdFeBMagnetRecycling.EnvironmentalScience&Technology,2013.47(18):p.10129-10136.

15. K.L.Nash,M.P.J.,Analytical-scaleseparationsofthelanthanides:AreviewoftechniquesandfundamentalsSepar.Sci.Technol.,2001.36(5-6):p.1257-1282.

16. A.Walton,H.Y.,N.A.Rowson,J.D.Speight,V.S.J.Mann,R.S.Sheridan,A.Bradshaw,I.R.HarrisandA.J.Williams,Theuseofhydrogentoseparateandrecycleneodymium–iron–boron-typemagnetsfromelectronicwaste.JournalofCleanerProduction,2015.104:p.236-241.

17. http://www.hothardware.com/news/toshiba-pushes-25-hdd-limit-to-1tb.18. http://www.modernenergy.co.nz/electric-vehicle-charger/.19. http://www.thinglink.com/scene.

54

20. Y.D.Zhang,J.I.B.,E.PotenzianiII,A.T.Pedziwiatr,W.E.WallaceandP.Panissod,AnNMRstudyoftetragonalNd2Fe14BandNd2Fe14−xCoxB.JournalofAppliedPhysics,1988.63(8):p.3719-3721.

21. S.Riano,K.B.,ExtractionandseparationofneodymiumanddysprosiumfromusedNdFeBmagnets:anapplicationofionicliquidsinsolventextractiontowardstherecyclingofmagnets.GreenChemistry,2015.17(5):p.2931-2942.

22. BabaK.,H.Y.,NemotoT.,Rare-EarthMagnetRecycling.2013:HitachiReview.p.452-455.

23. E.,B.,Recyclingpermanentmagnetsinonego.2015,ResearchNews.24. Binnemans,K.andP.T.Jones,RareEarthsandtheBalanceProblem.Journalof

SustainableMetallurgy,2015.1(1):p.29-38.25. M.A.R.Önal,C.R.B.,M.Guo,B.BlanpainandT.VanGerven,RecyclingofNdFeB

MagnetsUsingSulfation,SelectiveRoasting,andWaterLeaching.JournalofSustainableMetallurgy,2015.1(3):p.199-215.

26. C.Lee,Y.C.,CLiao,S.R.Popuri,S.TsaiandC.Hung,SelectiveLeachingProcessforNeodymiumRecoveryfromScrapNd-Fe-BMagnet.MetallurgicalandMaterialsTransactions,2013.44(13):p.5825-5833.

27. Yoon,H.-S.,etal.,LeachingkineticsofneodymiuminsulfuricacidfromE-scrapofNdFeBpermanentmagnet.KoreanJournalofChemicalEngineering,2014.31(4):p.706-711.

28. K.L.Nash,G.R.C.,Separationoffelements.SpringerScience+BusinessMedia,LLC.29. C.,H.,OverviewofRecentAdvancesinAn(III)/Ln(III)SeparationbySolventExtraction,

inIonExchangeandSolventExtraction,ASeriesofAdvances,M.B.,Editor.2010,Tayloer&Francis:BocaRaton,Florida.

30. A.Geist,M.W.,U.Müllich,K.Gompper,Actinide(III)/Lanthanide(III)PartitioningUsingn-Pr-BTPasExtractant:ExtractionKineticsandExtractionTestinaHollowFiberModule.2000.

31. S.Tachimori,Y.S.a.S.S.,MODIFICATIONOFTODGA-n-DODECANESOLVENTWITHAMONOAMIDEFORHIGHLOADINGOFLANTHANIDES(III)ANDACTINIDES(III).SolventExtractionandIonExchange,2002.20(6):p.687-699.

32. Z.-X.Zhu,Y.S.,H.Suzuki,S.SuzukiandT.Kimura,CumulativestudyonsolventextractionofelementsbyN,N,Nʹ,Nʹ-tetraoctyl-3-oxapentanediamide(TODGA)fromnitricacidinton-dodecane.AnalyticaChimicaActa,2004.527(2):p.163-168.

33. D.Kim,L.E.P.,L.H.Delmau,E.S.Peterson,J.HerchenroederandR.R.Bhave,SelectiveExtractionofRareEarthElementsfromPermanentMagnetScrapswithMembraneSolventExtraction.Environmentalscienceandtechnology,2015.49:p.9452-9459.

34. Cotton,S.,Lanthanideandactinidechemistry.2006:JohnWiley&Sons,Ltd.35. Monecke,T.,etal.,Tetradeffectinrareearthelementdistributionpatterns:a

methodofquantificationwithapplicationtorockandmineralsamplesfromgranite-relatedraremetaldeposits.GeochimicaetCosmochimicaActa,2002.66(7):p.1185-1196.

36. J.Kraikaew,W.S.a.C.C.,SolventExtractionStudyofRareEarthsfromNitrateMediumbytheMixturesofTBPandD2EHPAinKerosene.JournalofMetals,MaterialsandMinerals,2005.15:p.89-95.

37. D.W.Peppard,G.W.M.,Liquid-liquidextractionoftrivalentrareearthsusingacidicphosphonatesasextractants.1961:p.37-50.

55

38. MolybdenumCorporationofAmerica’sMountainPass,EuropiumProcess,AIMEAnnualMeeting,1968.

39. M.Mohammadi,K.F.,L.Kloo,J.MartinezDeLaCruzandÅ.Rasmuson,SeparationofND(III),DY(III)andY(III)bysolventextractionusingD2EHPAandEHEHPA.Hydrometallurgy,2015.156:p.215-224.

40. Tunsu,C.,Leachingandsolventextractionofrareearthmetalsfromfluorescentlampwaste.2014,ChalmersUniversityofTechnology.

41. Dukov,I.L.andM.Atanassova,Effectofthediluentsonthesynergisticsolventextractionofsomelanthanideswiththenoyltrifluoroacetoneandquaternaryammoniumsalt.Hydrometallurgy,2003.68(1–3):p.89-96.

42. Cox,M.,SolventExtractioninHydrometallurgy,inSolventExtractionPrinciplesandPractise.2004,MarcelDekker:NewYork,USA.p.455-458.

43. Retegan,T.,InvestigationofSolventSystemsBasedonbis-triazine-bipyridine(BTPB)-classligandsforSeparationofActinidesfromLanthanides.2009,ChalmersUniversityofTechnology:Göteborg,Sweden.

44. E.Anticó,A.M.,M.Hidalgo,V.Salvadó,M.Iglesias,M.Valiente,,Solventextractionofyttriumfromchloridemediabydi(2-ethylhexyl)phosphoricacidinkerosene.Speciationstudiesandgelformation.AnalyticaChimicaActa,1996.327(3):p.267-276.

45. J.Yang,C.E.a.T.R.,OptimizationofIndiumRecoveryandSeparationfromLCDWastebySolventExtractionwithBis(2-ethylhexyl)Phosphate(D2EHPA).InternationalJournalofChemicalEngineering,HindawiPublishingCorporation,2014.2014:p.9.

46. C.Basualto,F.V.,L.Molina,J.P.Munoz,E.Fuentesand,J.Sapag,Studyofthesolventextractionofthelighterlanthanidemetalionsbymeansoforganophosphorusextractants.JournaloftheChileanChemicalSociety,2013.58(2):p.1785-1789.

47. IUPAC.CompendiumofChemicalTerminology,2nded.(the"GoldBook").CompiledbyA.D.McNaughtandA.Wilkinson.BlackwellScientificPublications,Oxford(1997).

48. Löfström-Engdahl,E.Theeffectsofdiluentsinsolventextraction.inACSEPTInternationalWorkshop.Lisbon,Portugal.

49. Yang,J.,ProcessDevelopmentforExtractionandSeparationofInandYfromDiscardedFlatPanelDisplays.2015,ChalmersUniversityofTechnology.

50. G.,W.,KnovelSolvents-APropertiesDatabase.2008,ChemTecPublishing.51. D.R.,L.,Handbookofchemistryandphysics,n.edition,Editor.2001-2002.52. Retegan,T.,Theinfluenceofdiluentsandsidegroupsoftheligandsonliquid-liquid

extractionofactinidesandlanthanides.2007,ChalmersUniversityofTechnology.53. Marcus,Y.,PrinciplesofSolubilityandSolutions,inSolventExtractionPrinciplesand

Practise.2004.,MarcelDekker:NewYork.54. J.Yang,T.R.a.C.E.,IndiumrecoveryfromdiscardedLCDpanelglassbysolvent

extraction.Hydrometallurgy,2013.137:p.68-77.

56

APPENDIXICP-OES(Inductivelycoupledplasma–OpticalEmmissionSpectrometer)AthermoiCAP6500ICP-OESwasusedtomeasuretheelementalcompositionoftheaqueoussolutionsafterleachingandsolventextraction.ICP-OESisamethodcommonlyusedfortheanalysis of trace metals. The machine is made up of two parts, the ICP and the opticalspectrometer.Itworksontheprincipleofplasmaexcitationoftheatomsandthedetectionofemittedlightfromtherelaxationoftheexcitedatoms.Everyelementhasauniquesetofspectrallinesandcanbedetectedindividuallyinasample.