studies of polluted mine soils and treatment of waste waters

207
STUDIESOFPOLLUTEDMINESOILSANDTREATMENTOF WASTEWATERS.APPLICATIONSOFSYNCHROTRON BASEDTECHNIQUES,FIELDPORTABLEXRAY FLUORESCENCEANDADVANCEDOXIDATIONPROCESSES MartaÁvilaPérez Tesidoctoral ProgramadedoctoratenQuímica Directors:ManuelValienteMalmagro,GustavoPérez González DepartamentdeQuímica FacultatdeCiències Any2011

Upload: vonga

Post on 10-Jan-2017

214 views

Category:

Documents


1 download

TRANSCRIPT

STUDIES�OF�POLLUTED�MINE�SOILS�AND�TREATMENT�OF�WASTE�WATERS.�APPLICATIONS�OF�SYNCHROTRON�

BASED�TECHNIQUES,�FIELD�PORTABLE�X�RAY�FLUORESCENCE�AND�ADVANCED�OXIDATION�PROCESSES�

Marta�Ávila�Pérez�

Tesi�doctoral�

Programa�de�doctorat�en�Química�

Directors:�Manuel�Valiente�Malmagro,�Gustavo�Pérez�González�

Departament�de�Química�

Facultat�de�Ciències�

Any�2011�

Memòria�presentada�per�aspirar�al�Grau�de�Doctor�per�

Marta�Ávila�Pérez�

Vist�i�plau,�els�directors�

Dr.�Manuel�Valiente�Malmagro� � � Dr.�Gustavo�Pérez�González�

Bellaterra,�10�de�juny�de�2011�

�������������������

In�the�beginning�there�was�nothing.�God�said,�'Let�there�be�light!'.�And�there�was�light.��There��was�still�nothing,�but�you�could�see�it�a�whole�lot�better.�

�Ellen�Degeneres�

AGRAÏMENTS�

L’altre�dia�parlant�amb�alguns�companys�del�grup�sobre�que�a� la� tesi�hi�surten�el�nom�de�l’Elena�o�el�de’n�Gus�com�a�autors�d’algunes�fotos,�la�Pili�va�veure�una�figura�en�la�que�ella�em�va�ajudar�a�col�locar� les� línies�rectes� i�en�Fran�un�dibuix�on�ell�em�va�ajudar�a�triar�els�colors�(sent� daltònic)� i� em� vaig� adonar� que� és� totalment� veritat� pensar� que� en� una� tesi� tothom�hi�col�labora�una�miqueta,� al�menys� en� la�meva.� Pot� ser� en� una� figura,� una� ratlla� vertical,� una�paraula�o�una�idea,�i�sobretot�donant�me�molt�suport�en�tot�moment.�Moltes�gràcies�a�tots!!!��

Evidentment�la�part�d’aquesta�tesi�que�correspon�a�en�Manolo�i�en�Gus�és�molt�més�gran�que�la�de�la�resta�de�les�persones.�Gràcies�Manolo�per�haver�me�donat�la�oportunitat�de�fer�la�tesi�al�grup!�No�vull�desmerèixer�els�altres�grups,�però�des�del�primer�dia�sempre�he�cregut�que�no�podia�haver�anat�a�un�grup�millor� i� en�gran�part� això�és�gràcies�a� tu� i� al� teu�esforç.�Gus,�muchas�gracias�por�estar�siempre�dispuesto�a�ayudarme�y�aguantarme�mis�neuras,�sé�que�no�es�fácil,�muchas�gracias�de�verdad.�Eres�un�tío�genial,�y�un�papá�aun�mejor,�jeje!�

I�ara�toca�a�tota�la�resta�de�la�gent,�els�meus�amics�que�no�són�de�veritat�als�qui�no�estimo�gens�i�sou�poquets�a�més!�Montse�tantes�hores�viscudes,�tantes�històries�juntes,�et�desitjo�tota�la�felicitat�del�món!�A�GTS�hi�ha�qualitat� i� tu�n’ets�una�mostra!�Fran,�el�meu�reietonet,�ets�el�més�maco,�trobaré�a�faltar�les�teves�abraçades�de�les�10:22�i�totes�les�altres.�Pili,�ja�sé�qui�ets!�Ets� una� bona� amiga� i� una� bona� companya,� i� a�més�molt� divertida,� però� no� ho� diré� a� ningú�perquè�no�perdis�la�teva�fama.�Angélica�Angélica,�mi�compañera�de�la�hora�de�comer,�me�vas�a�hacer�mucha�falta,�y�lo�sabes.�Dejo�VIBRA�en�memoria�del�zentinel�para�que�de�vez�en�cuando�Diego� la� ponga� en� el� laboratorio.� Dieguito� todo� (o� casi� todo)� lo� que� dijimos� en� “El� día� de�halagar�a�Diego”�era�verdad,�eres�un�sol!�Muchas�gracias�por�tu�ayuda�con�la�informática�y�en�todo�lo�que�no�es�informática!�Bea,�companya�de�tantes�estades�del�sincrotró,�m’ho�he�passat�molt�bé�amb�tu� i�també�a� les�grans�festes�que�organitzes!�Olgucha,�el�nuevo�gran�fichaje�del�grupo,�gracias�por�venderme�a�“Daviduño”�a�precio�de�saldo�y�no�dejes�que�se�olvide�nunca�la�canción� de� Alfi� “venga� anímate� vamos� a� aprender,� qué� divertido� será� genial� y� no� podrás�parar!”.�Agus,�hem�anat�seguint�aquest�caminet�junts�tota�l’estona�gairebé�des�del�principi�fins�ara�gairebé�el�final,�has�sigut�un�gran�company,�sempre�fent�gala�del�teu�“señorío”,�jeje!�Oriol�Beisbol,�portaré�sempre�amb�orgull�la�samarreta�de�Els�Pétit�PÛt!�La�propera�vegada�que�siguis�alumne�meu�restaré�punts�per�falta�d’assistència!�Elena�Peralta�(mi�otra�amiga�Elena,�jeje!)�ya�sabes� lo�mucho�que� te�aprecio,�muchas�gracias� siempre�por� tu�ayuda!�DJ�Lluís�Soler,� tens�el�meu� vot� per� a� presidir� Catalunya� i� proclamar�ne� la� independència!� Ara,� no� la� facis� explotar�amb� hidrogen.� Amanda,� que� no� pots� evitar� trencar� cors� allà� on� vagis� (léase� vendedor� de�refrescos�del�Cirque�du�Soleil)�mai�deixis�de�pensar�que�“todo�es�muy�bonito!”�perquè�tens�raó.�Patri,� sempre� et� recordaré� amb� un� somriure.� Berta� que� vens� a� robar�nos� però� amb� un�somriure�sempre�i�dient�nos�“guapos”,�així�dóna�gust�que�et�robin,�jeje!�Julio,�tico,�siempre�tan�amable�y�tan�dispuesto�a�ayudar�y�a�escuchar�(y�a�traer�pasteles!),�Marta�y�Kike�(Kiki,� jijijiji!).�Gràcies�a�la�Maria�Dolors�per�ser�sempre�tan�diligent,�eh,�reina;�a�la�Cristina,�a�la�Maria�Muñoz�i�a�la�Montse�López.�

Tampoc�no�em�vull�oblidar�d’agraïr�l’ajuda�de�la�resta�de�la�gent�amb�qui�he�compartit�tants�moments�i�que�ja�no�volten�per�aquí:�Aleix�tinc�tants�i�tan�bons�records�de�tu�que�ompliria�la�tesi� sencera!�Moltes�gràcies�per� tot�el�que�em�vas�ajudar.�Així� com�també�ho�van� fer� l’Anna�Torrado�a�qui�encara�enyoro,�l’Àngels�amb�qui�hem�compartit�una�amistat�de�molts�anys�i�fins�i�tot�cursos�d’alta�cuina,�l’Anna�Bernaus�que�em�va�introduir�en�el�fabulós�món�del�sincrotró,�el�Johannes�siempre�dispuesto�a�ayudarme,�el�Jordy�Macanás�(Mac)�la�wikipedia�amb�potes�que�sempre�sempre�t’ajuda� i� troba�solucions,� la�Nadia� i� la�Rajaa� (I�miss�youuuuu!!!),�el�Sachin,�el�Mouhssine,� en� Xavi� Gaona,� la� Tània� Gumí,� el� José� A.� Muñoz,� en� Franki,� l’Amàlia,� el� Marc�Renom,� el� Jordi� Nualart...� Segurament� em�deixi� algú,� potser� fins� i� tot� algú�molt� important� i�anticipadament�demano�perdó�per�la�meva�memòria�de�peix...�ups!�Delfín!�La�mascota�de�GTS!�I�allà�al�cel�dels�peixets�també�un�record�per�al�Pezón.�

El�meu�gabinet�de�psicologia�es�mereix�un�raconet�en�aquests�agraïments.�Gràcies�Anabel�per�la�teva�amistat�i�tota�l’ajuda�i�consells�que�sempre�m’has�donat;�y�gracias�Susana,�Tamara�y�Sandra�por�vuestra�ayuda�y�vuestra�sonrisa�siempre�y�por�estar�siempre�dispuestas�a�apuntarse�a� todo!� Y� también� quiero� agradecer� al� resto� de� vuestro� laboratorio� aguantarme� cuando�aparezco:�Anna�y�Sole.�Recordadme�siempre�como�“Esa�chica�qué?”!�

Gracias�a�ti�David�Zamora,�por�ser�el�mister�de�la�selección�de�futbol�femenino�de�la�torre�de�química�y�por�toda�la�ayuda�extradeportiva,�jeje!�Nosotras�Ahí!�El�dream�team�liderat�per�el�míster� David� amb� qui� quasi� vam� aconseguir� arribar� a� la� final� però� el� que� segur� que� vam�aconseguir� va� ser� passar�ho�molt� bé!�Montse,� Susana,� Tamara,� Pili,� Sandra,� Silvia,� Amanda,�Amàlia,�Sole,�Cata,�Núria.��

Gracias� también� a� ti� Miguel,� compañero� de� la� planta� durante� algunos� años� por� estar�siempre�a�mi� lado�y�ayudarme�y� ser�mi�mejor�amigo.� I� gràcies�a� l’Elena,� l’Adela,� la�Salut,�en�Pep,�en�David�i�el�Guga,�per�ser�també�els�meus�amics!�

No� puc� oblidar�me� de� les� meves� companyes� de� pis,� la� Pilar� de� València� amb� qui� vaig�compartir� tants� anys� i� tantes� experiències,� Tufaria� y� Yannich,� Viri� la� mexicanita� y� mis�colombianitas� queridas:� Julix,� Pekas� y� Denise.� Julix�muchas� gracias� por� haberme� aguantado�estos� últimos� años,� fuiste� una� muy� buena� amiga,� te� dejo� en� arriendo� una� parcelita� de� mi�corazón� contrato� indefinido� y� a� Pekas� también� pero� una� parcelita�más� pequeña� porque�me�aguantó�mucho�menos� tiempo,� jeje!�Mis� pichurrias!� Denise� te�mereces� un� apartado� para� ti�solita.��

Sandrita�y�Denise�gracias�por�vuestra�colaboración�en�el�diseño�de�esta� tesis,�espero�que�sea�de�vuestro�agrado.�Y�especialmente�gracias�a�Tu,�mi�angelito,�por�todo�tu�apoyo�siempre,�por� tus� palabras,� por� tus� risas� y� tu� sonrisa!� Por� este� trocito� de� camino� en� el� que� me� has�acompañado� siempre� tan� pendiente.� Algún� día� sabrás� lo� grande� que� es� la� parcela� que� te�corresponde�a�ti!��

Però�a�qui�vull�enganyar?�Us�estimo�molt�a�tots!�

��������������

Als�meus�pares�i�a�en�David����

��

SUMMARY�

Summary��

Metals� have� been� used� since� prehistoric� times� and� they� play� a� key� role� in� civilization�

development.�Despite�mining�industry�still�represents�an�important�economic�activity�in�many�

countries,�large�amounts�of�solid�and�liquid�wastes�remain�stored�in�controlled�tailings�during�

mining�operations�and�sometimes�in�an�uncontrolled�way,�after�the�mine�closure.�Solid�wastes�

are�produced�from�ore�processing�such�as�crushing,�grinding�and�milling�and�are�disposed�off�in�

surrounding�land.�High�amounts�of�water�are�spent�to�wash�the�ore�and�to�reduce�the�mineral�

to� its�metallic� form.�These�mining�wastewaters�are�generally�dumped� into�ponds�secured�by�

dams.�In�this�sense,�elevated�levels�of�heavy�metals�from�metalliferous�mines�are�found�in�and�

around�the�mines.�Such�metals�represent�a�great�hazard�that�can�restrict�soil�use,�while�tailing�

dam�failures�can�produce�huge�human�and�environmental�damages.��

Thus,� this� PhD� thesis� is� aimed� at� the� characterization� of� the� heavy� metal� pollution�

around�abandoned�mines�as�well�as� to�develop�process� for�waste�water� treatment� including�

either� inorganic� or� organic� pollutants� from� industrial� activities,� i.e.,�mining�or� textile� related�

industries.��

In�this�concern,�the�studies�carried�out�are�summarized�as�follows:��

The�characterization�of�four�abandoned�mines�of�Marrakech�region�(Morocco)�by�means�

of� heavy� metal� spatial� concentration� and� the� identification� of� hazardous� sites� by� means� of�

mobility� tests.� In� this� concern,� the� studies� carried� out� represent� a� first� insight� into� four�

abandoned�mines�from�Marrakech�region�(Morocco):�Draa�Lasfar,�Kettara,�Sidi�Bou�Othmane�

and� Bir� Nehass.� The� characterization� of� the� heavy� metal� pollution� was� performed� by� Field�

Portable� X�ray� Fluorescence� (FP�XRF)� while� the� spatial� variability� was� determined� by�

Geographic� Information� Systems� (GIS).� A� prediction� of� the� risk� of� each� sampling� point� was�

completed�by�determining�the�mobility�of�anthropogenic�enhanced�heavy�metals�using�single�

leaching�tests.�The�calculation�of�the�Concentration�Enrichment�Ratios�(CER)�revealed�arsenic,�

copper,�lead�and�zinc�as�the�main�pollutants�in�all�mine�areas.�Draa�Lasfar�GIS�contour�maps�of�

these�pollutants�depict� the�most�polluted�areas�at� the�vicinity�of� the�mine,�especially�at� the�

northwest� area,� probably� linked� to� weathering� effects� and� topography� of� the� area.� The�

mobility�assays�indicate�greater�mobility�of�As�and�Zn�due�to�their�lower�adsorption�process�in�

the� soil,� independently� of� their� respective� concentration.�GIS� contour�maps� generally� reveal�

higher� concentration� around� sampling� points� localized� at� deposits� of� mining� residues.� The�

distribution� of� pollutants� at� Kettara� is� similar� for� arsenic,� copper� and� lead,� whilst� zinc�

distribution�is�more�homogeneous�along�the�mine�area.�In�addition,�lead�can�be�considered�the�

main�pollutant� considering� its� high�CER� values.� Regarding� SB�Othmane�mine� area,�GIS�maps�

observe�areas�with�high�contamination,�as�some�samples�have�CER�values�above�200.�Lead�and�

Summary��

zinc�can�be�considered�the�main�pollutants� in�SB�Othmane�mine�area.�Bir�Nehass�mine�area,�

likewise� SB�Othmane,� is� less� contaminated�with� arsenic� and� copper� being� lead� and� zinc� the�

main�pollutants.�In�this�sense,�a�unique�hot�spot�can�be�observed�for�arsenic�and�lead�around�

an�area�corresponding�to�a�residue�deposit�while�several�hot�spots�with�CER=200�can�be�seen�

for�lead�and�zinc,�also�related�to�residue�deposits.�The�mobility�results�point�out�the�greatest�

part�of�samples�to�have�very�low�mobility.�On�the�other�hand,�samples�of�SB�Othmane�and�Bir�

Nehass�are�highly�concentrated�on�Pb�and�Zn�and�present�an�extremely�high�content�on�Pb�and�

Zn�in�the�mobile�phase,�especially�high�for�the�samples�taken�at�the�deposits�of�residues.�Given�

the� high� content� of� lead� and� zinc� it� is� likely� that� the� concentration� of� metals� exceed� the�

capacity� of� the� soil� to� retain� them�and� the�migration� to� a�mobile� phase�may� take� place,� so�

remediation� treatments� should� be� applied� to� these� areas� if� the� soil� is� intended� for� further�

purposes.��

The� speciation� of� mercury� on� three� European� important� mercury� mines� to� determine�

toxicity� in� soils.� Besides�mobility� assays,� another� technique� has� been� applied� in� the� present�

work�to�determine�toxicity�of�soils.�Such�technique�involves�the�determination�of�the�chemical�

species�in�which�each�metal�is�present�in�soils.�As�one�of�the�most�toxic�heavy�metals,�mercury�

(Hg)�and�their�related�compounds,�can�be�absorbed�by�living�tissues�in�large�doses,�becoming�a�

great�hazard�due�to�its�ability�to�be�concentrated�and�stored�over�long�periods�of�time.�In�this�

work,� synchrotron�based� X�ray� Absorption� Near� Edge� Structure� (XANES)� has� been� used� to�

determine�the�speciation�of�mercury�in�geological�samples�from�three�of�the�largest�European�

mercury� mining� districts:� Almadén� (Spain),� Idria� (Slovenia)� and� Asturias� (Spain).� XANES� has�

been� complemented�with� a� single� extraction� protocol� for� the� determination� of�Hg�mobility.�

Ore,�calcines,�dump�material,�soil,�sediment�and�suspended�particles�from�the�three�sites�have�

been�considered�in�the�study.�In�the�three�sites,�rather�insoluble�sulfide�compounds�(cinnabar�

and� metacinnabar)� were� found� to� predominate.� Minor� amounts� of� more� soluble� mercury�

compounds� (chlorides� and� sulfates)� were� also� identified� in� some� samples.� Single� extraction�

procedures� indicate�a�strong�dependence�of�the�mobility�with�the�concentration�of�chlorides�

and� sulfates.� The� mercury� species� found� in� each� mine� are� related� to� the� efficiency� of� its�

roasting�furnaces.��

The�recovery�of�zinc�from�a�mine�tailing�pond�at�laboratory�and�pilot�plant�scale�to�solve�

an� environmental� problem� while� providing� an� economic� output.� Other� activities� performed�

through� the� framework� of� this� thesis� deal�with� the� reduction� of� the� amount� of�wastewater�

contained� in�mine� tailing� ponds� and� avoid� tailing� dam� breaches.� In� this� thesis� a� process� to�

recover� zinc� from� a� real�mine� tailing� pond� is� proposed.� This�mine� tailing� pond� stores� huge�

amounts�of�wastewater�containing�about�1�g/L�of�Zn�and�significant�amounts�of�ferrous,�ferric,�

Summary��

calcium,�copper,�aluminum�and�manganese�ions.�In�this�sense,�the�recovery�of�zinc�can�provide�

economic�value�to�the�process�while�solving�an�environmental�problem.�A�solvent�extraction�

process�was�considered�as�the�best�methodology�and�in�the�present�PhD�thesis�are�reported�

the�results�for�the�selection�of�the�best�extractant�amongst�DEHPA,�Cyanex�272�and�Ionquest�

290.�As�none�of�the�extractants�were�able�to�extract�Zn�selectively�from�a�solution�containing�

Fe,�a�bio�oxidation�process�followed�by�an�alkaline�precipitation�step�was�performed�prior�to�

the�SX�treatment�in�order�to�obtain�a�solution�without�iron.�The�Fe�removal�as�well�as�the�SX�

process�have�been�developed�successfully�at� laboratory�scale�and�verified�in�a�pilot�plant�on�

site,�using�two�Bateman�Pulsed�Columns�for�the�extraction�and�stripping�of�Zn.�Given�that�the�

recycling�of�the�organic�phase�lead�to�a�relative� importance�of�the�extractant�costs,� Ionquest�

290� was� selected� as� the� most� suitable� extractant� for� the� target� stream� due� to� its� higher�

selectivity� and� loading� capacity� towards� Zn� extraction.� Ketrul� D100� is� the� solvent�

recommended�owing�its�lower�volatility�and�flammability.�The�pilot�plant�proved�the�feasibility�

of�the�process,�obtaining�a�zinc�recovery�of�95%�and�leaving�less�than�50�mg/L�in�the�raffinate.�

The�stripping�was�efficient�and�only�a�single�stage�at�O:A=20�was�required�to�achieve�a�transfer�

of�40�g/L.�For�a�Zn�price�above�US$2/kg�the�operating�costs�are�covered�while,�additionally,�a�

serious�environmental�problem�is�solved.�

The�removal�of�organic�compounds�from�wastewater�by�the�Fenton�reaction�using�Fe3+�

loaded�materials.� As� an� example�of� another� remediation� technique,� this� thesis� presents� the�

removal�of�organic�wastewaters�by�three�different�materials�which�have�been�exchanged�with�

Fe�for�their�evaluation�as�heterogeneous�Fenton�catalysts.�The�Fenton�reaction,�consisting�on�

the� generation� of� the� highly� oxidant� hydroxyl� radical� is� employed� to� degrade� the� organic�

pollutants.� The� hydroxyl� radical� is� formed� by� hydrogen� peroxide� and� iron� salts� acting� as� a�

catalyst.� Several� drawbacks� arise� from� the�use�of� iron� salts�mainly� related� to� its� removal� by�

precipitation� of� iron� salts� to� generate� a� red� mud� that� should� be� treated.� The� employed�

catalysts� involve�a�synthetic�commercial� zeolite� (USY�zeolite),�a�natural� zeolite� (clinoptilolite)�

and�a�clay�(montmorillonite)�loaded�with�Fe.�The�results�indicated�that�high�Fe�content�could�

be� introduced� into� such� materials� with� minimum� time� and� reagents� consumption� and,� in�

addition,� these� Fe�loaded� materials� can� be� successfully� employed� for� the� decolorisation� of�

AR14� solutions� and� the� mineralization� of� acetic� acid� and� phenol.� In� this� sense,� Fe3+�USY�

decolorisation�kinetics�was�equal� to� the�homogeneous�catalysis� (less� than�15�min� to�achieve�

total�decolorisation)�whereas�Fe3+�MMT�and�Fe3+�clinoptilolite�showed�slower�kinetics� lasting�

30� and�60�min,� respectively.�Moreover,� tests� performed� to� acetic� acid� and�phenol� solutions�

demonstrated� 30%� and� 95%� of� COD� removal,� respectively,� whereas� homogeneous� catalysis�

Summary��

only�provided�a�25%�and�85%�COD�removal�respectively.�Column�experiments�using�the�more�

economical�material,� clinoptilolite,�were�performed�obtaining� also� successfully� results� hence�

indicating�the�feasibility�of�these�low�cost�Fe�loaded�materials�as�heterogeneous�catalysts�for�

the�Fenton�reaction.�The�minimal�losses�of�Fe�from�the�materials�avoided�the�necessity�of�red�

mud�removal.��

The� removal� of� arsenic� from� inorganic� wastewater� by� using� Fe3+�loaded� materials.�

Finally,� taking� profit� of� the� affinity� of� Fe(III)� compounds� towards� arsenic� inorganic� species,�

several�Fe3+�loaded�materials�with�high�exchange�capability�such�as�zeolite�USY�(USY),�zeolite�Y�

(ZY)�and�a�sponge�(Sp)�have�been�applied� for� the�removal�of�arsenic� from�inorganic�polluted�

wastewaters.� Arsenic� contamination� in� groundwater� generates� widespread� human� health�

disasters� around� the� world� (especially� in� Southeast� Asia).� In� this� sense,� besides� their�

application�as�catalysts�in�Fenton�processes,�Fe�loaded�materials�can�be�also�employed�for�the�

removal� of� arsenic.� These� materials� were� characterized� by� FP�XRF� and� Extended� X�ray�

Absorption�Fine�Structure�(EXAFS)�techniques�in�order�to�shed�light�onto�the�different�sorption�

mechanisms� of� arsenic� into� such� materials.� The� sorption� mechanism� reveals� a� strong�

dependence�on�the�specific�surface�area�and�the�available�sites,�thus�as�zeolite�Y�has�specific�

surface� area� higher� than� zeolite� USY� and� Forager� sponge,� its� Fe� loading� becomes� greater.�

Forager� sponge,� has� an� As:Fe� absorption� ratio� higher� than� the� one� expressed� by� zeolites�

mainly� owed� to� tertiary� amine� salt� groups� contained� in� the� sponge� that� can� bind� anionic�

contaminants,�such�as�arsenic,�chromate�or�uranium�oxide�species.�The�characterization�of�the�

adsorption�of� arsenate�onto� these� Fe3+�loaded�materials� revealed�arsenate�bidentate� corner�

sharing�bond�as�the�main�adsorption�process.��

��

CONTENTS�

1.�INTRODUCTION ...........................................................................................................................3�

��MINE�SITES�CHARACTERIZATION ......................................................................................................3�

1.1.�MINING�OVERVIEW ..........................................................................................................................3�

1.2.�SOIL�IMPACTS�FROM�MINING ..........................................................................................................4�

1.3.�CHARACTERIZATION�OF�MINE�SITES ................................................................................................7�

1.3.1.�Sampling .....................................................................................................................................7�

1.3.2.�Soil�Physico�chemical�Characterization ......................................................................................8�

1.3.3.�Metal�Analysis ..........................................................................................................................10�

1.4.�SOIL�QUALITY�REGULATIONS..........................................................................................................11�

1.5.�CHEMICAL�SPECIATION�AND�FRACTIONATION�IN�SOILS ...............................................................12�

1.5.1.�Sequential�Extraction�Schemes ................................................................................................14�

1.5.2.�Single�Leaching�Tests ................................................................................................................15�

1.6.�SOIL�RISK�ASSESSMENT�TOOLS.......................................................................................................16�

1.6.2.�Geographic�Information�Systems .............................................................................................16�

1.6.3.�Principal�Component�Analysis�in�Geosciences .........................................................................17�

1.7.�WEAKNESSES�AND�NEEDS�OF�MINING�SITES�CHARACTERIZATION...............................................19�

��REMEDIATION�TECHNIQUES�OF�INDUSTRIAL�CONTAMINATED�WATER ................................20�

1.8.�SOLVENT�EXTRACTION�FOR�THE�RECOVERY�OF�Zn�FROM�ACIDIC�MINE�WATERS ........................20�

1.8.1.�Zinc�Overview ...........................................................................................................................21�

1.8.2.�The�Solvent�Extraction�Process ................................................................................................22�

1.8.3.�Scaling�Solvent�Extraction�To�a�Pilot�Plant ...............................................................................23�

1.9.�FE�LOADED�MATERIALS�FOR�THE�REMOVAL�OF�ORGANIC�AND�INORGANIC�CONTAMINANTS ..24�

1.9.1.�Zeolites .....................................................................................................................................24�

1.9.2.�Clays..........................................................................................................................................27�

1.9.3.�Sponges ....................................................................................................................................28�

1.10.�THE�FENTON�REACTION ................................................................................................................29�

1.11.�ARSENIC�SORPTION�USING�FE�LOADED�MATERIALS ...................................................................31�

1.11.1.�Arsenic�toxicity .......................................................................................................................31�

1.11.2.�Arsenic�sorbents .....................................................................................................................32�

1.12.�WEAKNESSES�AND�NEEDS�OF�INDUSTRIALLY�CONTAMINATED�WATERS ...................................33�

��ANALYTICAL�TECHNIQUES.........................................................................................................35�

1.13.�X�RAY�FLUORESCENCE ..................................................................................................................35�

1.13.1.�X�Ray�interaction�with�matter ................................................................................................35�

1.13.2.�X�Ray�Fluorescence.................................................................................................................37�

1.13.4.�Field�Portable�XRF�instrumentation .......................................................................................37�

1.14.�SYNCHROTRON�BASED�TECHNIQUES ...........................................................................................39�

1.14.1.�Synchrotron�Light�Sources......................................................................................................39�

1.14.2.�Design�and�Operation�of�a�Synchrotron�Light�Source ............................................................39�

1.14.3.�X�Ray�Absorption�Spectrometry .............................................................................................41�

1.15.�OBJECTIVES ...................................................................................................................................45�

1.16.�REFERENCES ..................................................................................................................................46�

2.�METHODOLOGY...................................................................................................................57�

��MINE�SITES�CHARACTERIZATION..........................................................................................59�

2.1.�STUDIED�MINES�DESCRIPTION .......................................................................................................59�

2.1.1.�Marrakech�Mines:�Draa�Lasfar,�Kettara,�Sidi�Bou�Othmane�and�Bir�Nehass�(Morocco)..........59�

2.1.2.�European�Mercury�Mining�Districts:�Almadén,�Mieres�and�Idrija ............................................61�

2.1.3.�Aznalcóllar�Tailing�Pond............................................................................................................63�

2.2.�SAMPLING .......................................................................................................................................64�

2.2.1.� Marrakech� Mining� Districts:� Draa� Lasfar,� Kettara,� Sidi�Bou� Othmane� and� Bir� Nehass�

(Morocco) ...........................................................................................................................................64�

2.2.2.�European�Mercury�Mining�Districts:�Almadén,�Asturias�(Spain),�Idrija�(Slovenia) ...................64�

2.3.�CHARACTERIZATION .......................................................................................................................65�

2.3.1.�Physico�chemical�Parameters...................................................................................................65�

2.3.2.�Total�metal�concentration ........................................................................................................65�

2.3.3.�Total�Mercury�Content .............................................................................................................67�

2.3.4.�Mobility�of�the�mine�samples...................................................................................................68�

2.3.5.�XAS�measurements...................................................................................................................69�

2.4.�DATA�TREATMENT ..........................................................................................................................70�

2.4.1.�Concentration�Enrichment�Ratios ............................................................................................70�

2.4.2.�Geographic�Information�Systems .............................................................................................71�

2.4.3.�Statistical�Tools.........................................................................................................................71�

2.4.4.�XAS�Data�Treatment .................................................................................................................72�

��REMEDIATION�TECHNIQUES�OF�INDUSTRIAL�CONTAMINATED�WATER .........73�

2.5.�ZINC�SOLVENT�EXTRACTION ...........................................................................................................73�

2.5.1.�Laboratory�Experiments ...........................................................................................................73�

2.5.2.�Scaling�the�SX�to�a�Pilot�Plant ...................................................................................................74�

2.6.� Fe�EXCHANGE�MATERIALS� FOR� THE� REMEDIATION�OF� ORGANIC� AND� INORGANIC� POLLUTED�

WATERS..................................................................................................................................................76�

2.6.1.�Fenton�Reaction .......................................................................................................................77�

2.6.2.�Arsenic�removal........................................................................................................................78�

2.7.�REFERENCES ....................................................................................................................................79�

3.�RESULTS�AND�DISCUSSION .......................................................................................85�

��MINE�SITES�CHARACTERIZATION..........................................................................................85�

3.1.�HEAVY�METAL�CONTAMINATION�AND�MOBILITY�AT�THE�DRAA�LASFAR�MINE�AREA .................85�

3.1.1.�Physico�chemical�parameters...................................................................................................85�

3.1.2.�Heavy�metal�concentration�in�the�mine�area...........................................................................86�

3.1.3.�GIS�contour�maps�of�the�main�pollutants.................................................................................85�

3.1.4.�Effect�of�particle�size�and�mobility ...........................................................................................86�

3.2.�CHARACTERIZATION�OF�KETTARA,�SIDI�BOU�OTHMANE�AND�BIR�NEHASS�MINE�AREAS............92�

3.2.1.�Physico�chemical�characterization ...........................................................................................92�

3.2.2.�Heavy�metal�concentration�in�the�mine�area...........................................................................93�

3.2.3.�Application�of�chemometrics ...................................................................................................95�

3.2.4.�GIS�contour�maps�of��the�pollutants.........................................................................................98�

3.3.�XANES�SPECIATION�OF�MERCURY� IN�THREE�MINING�DISTRICTS:�ALMADEN�(SPAIN),�ASTURIAS�

(SPAIN)�AND�IDRIJA�(SLOVENIA) .........................................................................................................104�

3.3.1.�Chemical�analysis�of�the�samples ...........................................................................................104�

3.3.2.�XANES�speciation�and�mobility�results ...................................................................................106�

��REMEDIATION�TECHNOLOGIES.............................................................................................111�

3.4.�EXTRACTANT�AND�SOLVENT�SELECTION�TO�RECOVER�ZINC�FROM�A�MINING�EFFLUENT:�FROM�

LABORATORY�SCALE�TO�PILOT�PLANT.................................................................................................111�

3.4.1.�SX�laboratory�results ..............................................................................................................111�

3.4.2.�SX�pilot�plant�process .............................................................................................................114�

3.5.�FE�LOADED�MATERIALS�FOR�THE�REMEDIATION�OF�ORGANIC�AND�INORGANIC�CONTAMINATED�

WASTE�WATERS ...................................................................................................................................119�

3.5.1.�Fe�loaded�materials�applied�as�Fenton�catalysts ...................................................................120�

3.5.2.�Fe�loaded�materials�applied�to�arsenic�removal ....................................................................124�

3.6.�REFERENCES ..................................................................................................................................128�

4.�CONCLUSIONS .....................................................................................................................131�

ANNEXES�

ANNEX� I.� HEAVY� METAL� CONTAMINATION� AND� MOBILITY� AT� THE� MINE� AREA� OF� DRAA� LASFAR�

(MOROCCO).�Marta�Avila,�Gustavo�Perez,�Mouhsine�Esshaimi,� Laila�Mandi,�Naaila�Ouazzani,� Jose� L.�

Brianso� and� Manuel� Valiente.� The� Open� Environmental� Pollution� &� Toxicology� Journal.� Accepted�

Manuscript.��

ANNEX� II.� XANES� SPECIATION� OF� MERCURY� IN� THREE� MINING� DISTRICTS� –� ALMADEN,� ASTURIAS�

(SPAIN),� IDRIA� (SLOVENIA).� Jose� Maria� Esbri,� Anna� Bernaus,� Marta� Avila,� David� Kocman,� Eva� M.�

Garcia�Noguero,� Beatriz� Guerrero,� Xavier�Gaona,� Rodrigo� Alvarez,� Gustavo� Perez�Gonzalez,�Manuel�

Valiente,�Pablo�Higueras,�Milena�Horvat�and�Jorge�Loredo.�Journal�of�Synchrotron�Radiation.�(2010).�

Volume:�17,�Issue:�2,�Pages:�179�186.��

ANNEX�III.�EXTRACTANT�AND�SOLVENT�SELECTION�TO�RECOVER�ZINC.�Marta�Avila,�Gustavo�Perez�and�

Manuel�Valiente.�Solvent�Extraction�and�Ion�Exchange�(2011),�29:�384–397.��

ANNEX�IV.�ZINC�RECOVERY�FROM�AN�EFFLUENT�USING�IONQUEST�290:�FROM�LABORATORY�SCALE�TO�

PILOT�PLANT.�M.�Avila,�B.�Grinbaum,�F.�Carranza,�A.�Mazuelos,�R.�Romero,�N.�Iglesias,�J.L.�Lozano,�G.�

Perez,�M.�Valiente.�Hydrometallurgy�(2011),�107:�63�67.��

1�

1��

INTRODUCTION��MINE�SITES�CHARACTERIZATION ........................................................................................................3�

1.1.�MINING�OVERVIEW ...........................................................................................................................3�1.2.�SOIL�IMPACTS�FROM�MINING ...........................................................................................................4�1.3.�CHARACTERIZATION�OF�MINE�SITES..................................................................................................7�1.4.�SOIL�QUALITY�REGULATIONS...........................................................................................................10�1.5.�CHEMICAL�SPECIATION�AND�FRACTIONATION�IN�SOILS .................................................................12�1.6.�SOIL�RISK�ASSESSMENT�TOOLS........................................................................................................16�1.7.�WEAKNESS�AND�NEEDS�OF�MINING�SITES�CHARACTERIZATION.....................................................19�

��REMEDIATION�TECHNIQUES�OF�INDUSTRIAL�CONTAMINATED�WATER...................................20�

1.8.�SOLVENT�EXTRACTION�FOR�THE�RECOVERY�OF�Zn�FROM�ACIDIC�MINE�WATERS...........................20�1.9.�FE�LOADED�MATERIALS�FOR�THE�REMOVAL�OF�ORGANIC�AND�INORGANIC�CONTAMINANTS......24�1.10.�THE�FENTON�REACTION.................................................................................................................29�1.11.�ARSENIC�SORPTION�USING�FE�LOADED�MATERIALS .....................................................................31�1.12.�WEAKNESSES�AND�NEEDS�OF�INDUSTRIALLY�CONTAMINATED�WATERS......................................33�

��ANALYTICAL�TECHNIQUES...................................................................................................................35�

1.13.�X�RAY�FLUORESCENCE ...................................................................................................................35�1.14.�SYNCHROTRON�BASED�TECHNIQUES ............................................................................................39�1.15.�OBJECTIVES....................................................................................................................................44�1.16.�REFERENCES...................................................................................................................................45�

2�

1.�Introduction�

3�

�This� chapter� is� addressed� to�provide� general� information� related� to� the�work� that�has�

been�performed�in�this� thesis.� In�this�sense,�general�aspects�of�mines,� its�characterization�by�

means�of�different�techniques�and�parameters,�as�well�as�techniques�to�remediate�organic�and�

inorganic� industrial�wastewaters�either�synthetic�or�from�specific�mine�water�are� introduced.�

Thus,� three�main�sections�have�been�distinguished:�Mine�sites�characterization,�Remediation�

techniques�and�Analytical�techniques.��

MINE�SITES�CHARACTERIZATION�

In� this� section� is� focused� on� the� role� of�mining� as� a� key� element� for� human� progress�

together� with� the� related� environmental� problems� such� as� heavy� metal� contamination� or�

mining� water� impoundments� (tailings� ponds)� breachings.� The� description� of� several�

parameters�to�characterize�the�contamination�around�mine�areas�is�also�included.��

1.1.�MINING�OVERVIEW�

Since�prehistory,�mining�has�been�key�to�the�development�of�civilizations.�In�this�sense,�

the�cultural�ages�of�man�are�associated�with�minerals�or�their�derivates�such�as�the�Stone�Age�

(prior�to�4000�BC),�the�Bronze�Age�(4000�to�5000�BC)�or�the�Iron�Age�(1500�BC�to�1780�BC)�[1].�

In�this�regard,�flint�implements�for�agricultural�or�construction�purposes�found�with�the�bones�

of� the�Paleolithic�man� (300,000�years�ago)� revealed�mining�activities� since�prehistoric� times.�

However,� the� oldest� known� underground� mine,� located� at� Bomvu� Ridge� (Swaziland),� is�

believed�to�be�40,000�years�old.�Nonetheless,� it�was�until�Egyptian�times�that�mines�attained�

depths�of�250�m.�During�the�Bronze�and�Iron�Ages�humans�discovered�smelting�and�learned�to�

reduce� ores� into� pure� metals� or� alloys,� which� greatly� improved� their� ability� to� use� these�

metals.�Later�on,�the�Romans�developed�large�scale�mining�methods�such�as�hydraulic�mining�

methods� to� prospect� the� ore� deposits� and� the� use� of� large� volumes� of� water� brought� by�

numerous� aqueducts� to� the�mine�where� it�was� stored� in� large� reservoirs� and� tanks� used� to�

remove� rock� debris.�All� the�main�mine�areas� exploited�nowadays,�were� already�exploited� in�

roman� times,� or� even�previously� to� roman� times.� Iberian�peninsula�was� the�most� important�

mining� region,� of� special� relevance�were� the�mines� of� Río� Tinto,� Cartagena� district,� Linares,�

1.�Introduction��

4�

Sierra�Morena� and� Almadén� in� Spain;� and� Aljustrel,� Sâo� Domingos,� Valongo,� Jales� and� Três�

Minas�in�Portugal�although�all�the�regions�of�the�Roman�Empire�were�also�exploited�(Table�1.1)�

[2].��

Table�1.1.�Principal�mines�exploited�during�Roman�Empire�and�minerals�extracted�Mine� Minerals�extracted� Uses�during�the�Roman�empire�

Lead� Piping�(aqueducts�plumbing,�gutters�for�villas)�Río�tinto�(Spain)�

Silver� Coins,�weapons�

Almadén�(Spain)� Mercury� Pigment�

Las�Médulas�(Spain),�

Dolaucothi�(Wales)�Gold� Tools,�weapons,�jewellery,�coins�

Aljustrel�(Portugal)� Zinc,�lead� Alloy�copper�into�brass�for�weapons�

Nowadays,�mining�activities�still�represent�an�important�role� in�the�world�development�

and� an� important� economic� activity� in� many� countries.� In� this� sense,� in� 2001� the� mining�

industry�produced�over�6�billion�tons�of�raw�product�valued�at�several�trillion�dollars.�Mineral�

processing� of� these� raw� materials� adds� further� value� as� raw� materials� and� products� are�

created�to�serve�all�aspects�of�industry�and�commerce�worldwide�[3].�

1.2.�SOIL�IMPACTS�FROM�MINING�

Mining�consists�in�the�extraction�of�valuable�minerals�or�other�geological�materials�from�

the�earth,�usually�from�an�ore�body,�vein�or�(coal)�seam,�that�implies�the�removal�of�soil.�Ore�

bodies�are�naturally�occurring�concentrations�of�minerals�with�sufficiently�high�concentrations�

of� metals� as� to� make� them� economically� worthwhile� exploited.� However,� it� has� been�

estimated�that�more�than�70%�of�all�the�material�excavated�in�mining�operations�is�discarded,�

and�high�amounts�of�water�are�spent�on�mineral�processing�(i.e.�washing�the�ore�to�enable�the�

separation�of�valuable�metals�or�minerals�from�their�gangue�or�waste�material,�to�reduce�the�

mineral�to� its�metallic�form�since�most�metals�are�present� in�ores�as�oxides�or�sulfides,�etc.).�

These� wastes� (called� tailings)� are� commonly� spread� throughout� the� mine� area� or,� when�

consisting� in�mining�wastewaters,�dumped� into�ponds� secured�by�dams� [4].�Hence,�elevated�

levels�of�heavy�metals�from�metalliferous�mines�are�found�in�and�around�the�mines�due�to�the�

discharge�and�dispersion�of�mine�waste�materials�into�the�ecosystem�resulting�in�large�areas�of�

agricultural� land�contaminated�posing�an�environmental�risk�for�humans�and�ecosystems�and�

thus�restricting�soil�use�[5].�Thus,�the�nature�of�mining�processes�creates�a�potential�negative�

impact�on�the�environment�both�during�the�mining�operations�and�for�years,�after�the�mine�is�

closed� and� many� regions� have� been� contaminated� causing� huge� impact� in� the� soils�

surrounding�mine�areas.��

1.�Introduction�

5�

There�are�around�560,000�abandoned�mines�on�public�and�privately�owned�lands�in�the�

United�States�alone�and�it�was�estimated�that�in�2000�existed�more�than�3,500�tailings�ponds�

with�water� containing� high� amounts� of�metals� [6].� Every� year,� 2� to� 5�major� failures� and� 35�

minor�failures�occurred;�hence�releasing�high�amounts�of�highly�contaminated�waters�into�the�

environment�[7].�To�date�250�cases�of�tailings�dam�failures� in�the�world�have�been�compiled�

producing�huge�human�and�environmental�damages�(Table�1.2)�[8,�9].��

These� huge� amounts� of� heavy� metals� deposited� in� waste� dumps� and� tailings� ponds�

require�management� and�monitoring� once� the� activity� has� deceased� [10]� as� several�metals�

(e.g.�mercury,�cadmium,�lead,�nickel,�arsenic,�zinc,�copper)�are�hazardous�to�human�health�and�

terrestrial�ecosystems.�So�the�determination�of�metals�in�contaminated�soils�should�be�carried�

out� to� obtain� information� about� the� nature,� quantity,� distribution� and� behavior� of�

contaminants�and,�if�necessary,�to�select�the�most�appropriate�use�of�the�site�[11].�Thus,�it�is�a�

foremost�task�to�characterize�heavy�metal�concentration�around�mine�areas�once�the�activity�

has� deceased� to� detect� the� degree� of� contamination� in� order� to� apply� proper�management�

tools.�

In�this�sense,� in�several�developing�countries,�mining�activities�represent�a�high�area�of�

activity�thus�constituting�a�great�hazard�due�to�the�presence�of�high�amounts�of�heavy�metals�

related� to� functioning� or� abandoned� mines.� Despite� mining� is� an� important� part� of� the�

industrial� development� in� many� developing� countries� (Philippines,� Morocco,� Peru,� etc.),�

relatively�few�environmental�studies�on�mining�sites�have�been�undertaken�to�determine�the�

heavy�metal�concentration�around�mine�areas�and�their�impact�on�surrounding�soil�and�water�

resources,� where� commonly� no� national� program� for� the� rehabilitation� of� existing� polluted�

sites�is�implemented�[12,�13,�14,�15].��

1.�Introduction��

6�

Table�1.2.�Major�tailing�dam�failures�in�the�last�25�years�Date� Location� Release� Impacts�

2010,� Oct�4th��

Kolontár,�Hungary�

700,000�m3�of�caustic�red�mud�

several� towns� flooded,� 10� people� killed,� approx.�120�people�injured�

2010,�Jun.�25��

Huancavelica,�Peru�

21,420�m3�of�tailings� contamination�of�Escalera�river�and�Opamayo�river�110�km�downstream�

2009,�Aug.�29��

Karamken,�Russia�

?� Eleven�homes�were�carried�away�by�the�mudflow;�at�least�one�person�was�killed�

2009,�May�14��

Huayuan�County,�China�

50,000�m3�of�tailings� A� home� destroyed,� three� people� killed� and� four�people�injured.�

2008,�Sep.�8�

Taoshi,�China� ?� A�mudslide� several� meters� high� buried� a�market,�several�homes�and�a�three�storey�building.�At�least�254�people�were�dead�and�35�injured�

2006,�Nov.�6��

Nchanga,�Zambia�

?� Release� of� highly� acidic� tailings� into� Kafue� river;�drinking�water�supply�of�downstream�communities�shut�down�

2006,�April�30��

near� Miliang,�China�

?� Five�injured�people,�17�residents�missing�and�more�than�130�local�residents�evacuated.�

2004,� Sep�5�

Riverview,�Florida,�USA��

227,000� m3� of� acidic�liquid�

liquid� spilled� into� Archie� Creek� that� leads� to�Hillsborough�Bay�

2003,�Oct.�3�

Cerro� Negro,�Chile�

50,000�tonnes�of�tailings� tailings� flowed� 20� kilometers� downstream� the� La�Ligua�river�

2002,�Aug.� 27� /�Sep.�11�

San� Marcelino,�Philippines�

?� Aug.�27:�some�tailings�spilled�into�Mapanuepe�Lake�and� eventually� into� the� Sto.� Tomas� RiverSep.� 11:� villages� flooded� with� mine� waste;� 250�families�evacuated�

2001,�Jun.�22�

Sebastião� das�Águas� Claras,�Brazil�

?� tailings�wave�traveled�at�least�6�km,�killing�at�least�two�mine�workers,�three�more�workers�are�missing�

2000,�Oct.�18�

Nandan� county,�China�

?� at� least� 15� people� killed,� 100�missing;�more� than�100�houses�destroyed�

2000,�Jan.�30�

Baia� Mare,�Romania�

100,000� m3� of� cyanide�contaminated�liquid�

contamination� of� the� Somes/Szamos� stream,�tributary� of� the� Tisza� River,� killing� tonnes� of� fish�and� poisoning� the� drinking�water� of�more� than� 2�million�people�in�Hungary�

1999,�Apr.�26�

Placer,�Philippines�

700,000� tonnes� of�cyanide�tailings�

17�homes�buried,�51�hectares�of�riceland�swamped�

1998,�Apr.�25�

Aznalcóllar,�Spain�

4�5� million� m3� of� toxic�water�and�slurry�

thousands� of� hectares� of� farmland� covered� with�toxic�slurry�

1997,�Oct.�22�

Pinto� Valley,�USA�

230,000�m3�of�tailings�and�mine�rock�

tailings�flow�covers�16�hectares�

1996,�Aug.�29�

El�Porco,�Bolivia� 400,000�tonnes� 300�km�of�Pilcomayo�river�contaminated�

1996,�Mar.�24�

Marcopper,�Philippines�

1.6�million�m3� Evacuation� of� 1200� residents,� 18� km� of� river�channel�filled�with�tailings,�US$�80�million�damage�

1994,�Oct.�2��

Payne� Creek�Mine,�USA��

6.8�million�m3� 500,000� m3� released� into� Hickey� Branch,� a�tributary�of�Payne�Creek��

1994,�Feb.�22�

Harmony,�Merriespruit,�South�Africa�

600,000�m3� tailings� traveled� 4� km� downstream,� 17� people�killed,�extensive�damage�to�residential�township�

1994,�Feb.�14�

Roxby� Downs,�South�Australia�

5� million� m3� of�contaminated� water� into�subsoil�

?�

1993� Marsa,�Peru� ?� 6�people�killed�1985,� July�19�

Stava,� Trento,�Italy�

200,000�m3� tailings� flow�4.2�km�downstream�at�90�km/h;�268�people�killed,�62�buildings�destroyed�

1.�Introduction�

7�

1.3.�CHARACTERIZATION�OF�MINE�SITES�

In�the�last�years�the�systematic�control�of�contaminated�areas�has�become�a�key�issue�to�

define�healthcare�policies,�cost�effective�environmental�planning�and�risk�assessment�tools.�In�

this� sense,� sampling� of� potentially� contaminated� soil� from� polluted� areas� is� intended� to�

provide� data� of� several� physico�chemical� parameters� or� metal� content� of� the� soil� for� the�

assessment�of�whether�the�pollution�has�caused�or�may�cause�environmental�problems.��

1.3.1.�SAMPLING�

As� a� previous� step� to� characterize� a� mining� area,� a� sampling� strategy� is� needed.� The�

selection�and�location�of�the�sampling�points�depend�on�the�objectives�of�the�investigation,�the�

preliminary� information� available� and� the� on�site� conditions.� Experiences� (and� theoretical�

considerations)�show�that�in�many�cases�systematic�sampling�on�a�regular�grid�is�both�practical�

and� sufficiently� productive� to� allow� the� creation� of� a� detailed� picture� of� variations� in� soil�

properties.��

A�regular�grid�is�usually�employed�in�environmental�studies,�in�which�the�sampling�area�is�

large�(for�example,�soils�with�different�applications).�When�the�purpose�of�the�sampling�is,�for�

instance,�setting�the�values�of�certain�properties�in�an�homogeneous�area�or�a�first�prospection�

in�an�area�where�contamination�is�suspected,�irregular�grids�in�form�of�X,�W,�S,�etc.�are�usually�

carried�out,� in�which�the�sampling�points�are�also�predefined�(Figure�1.1).�These�samples�are�

usually�mixed�to�form�composite�samples.��

Figure�1.1.�Grids�in�form�of�W�and�X�for�systematic�sampling�

Other�similar�approaches�include�simple�regular�grids,�circular�or�clustered�(Figure�2)�in�

order�to�estimate�the�impact�of�a�source�of�pollution�in�the�area�of�study�(with�the�possibility�

of�concentration�gradients)�or�to�estimate�concentration�levels.�The�grids�should�be�designed�

to�study�areas�where�all�points�can�have�a�similar�concentration�of�the�target�analytes,�as�is�the�

case�of�simple�irregulars�grids�(Figure�1.2A)�and�alternatives�that�allow�the�subdivision�of�the�

areas�in�quintets�(Figure�1.2B)�or�can�cover�an�area�assuming�a�localized�source�of�the�target�

analytes,�as�is�the�case�of�circular�grids�(Figure�1.2C).�

1.�Introduction��

8�

Sampling�using�clusters� is�actually�a� combination�of� random�and�systematic� strategies,�

with�or�without� composite� samples.� It� consists�of� the� random�or� systematical� selection�of� a�

certain�number�of�blocks�in�a�regular�grid,�and�take�a�number�of�individual�samples�at�random�

(Figure� 1.2D).� The� samples� are� analyzed� individually� or� as� composite� samples,� allowing� an�

estimation�of�variability�at�the�local�level�(within�each�group)�or�global�(between�groups).�

Figure�1.2.�Systematic�sampling�grids�

The� number� of� sampling� points� can� be� easily� increased� (e.g.,� in� areas� meriting� more�

detailed�investigation),�the�grid�is�easy�marked�by�means�of�GPS�systems�and�sampling�points�

can�be�easily�relocated.�However,�sometimes�other�patterns�are� followed�based�on�a�known�

local� distribution�or�hot� spot�distributions� along�a� line� towards� specific� receptors� allowing� a�

reduction�on�costs�and�resource�consumption.��

1.3.2.�SOIL�PHYSICO�CHEMICAL�CHARACTERIZATION��

Usually� it� is� necessary� to� determine� the� nature,� concentrations,� and� distribution� of�

naturally� occurring� substances� and� contaminants� (extraneous� substances),� the� physical�

properties� and� the� presence� and� distribution� of� chemical� species� of� interest� to� identify�

immediate� hazards� to� human� and� to� the� environment.� This� information� will� also� help� to�

determine� the� suitability� of� a� soil� for� an� intended�use� (agricultural� production�or� residential�

development� amongst� others)� or� to� assess� the� transfer� of� substances� from� soils� to� plants�

(bioavailability).�

Several� inorganic� parameters� should� be� taken� into� account� when� characterizing� and�

assessing�risks�from�contaminated�sites�such�as�electrical�conductivity,�pH,� loss�on�ignition�or�

the�carbonate�content�amongst�many�others�[16].��

Acidity.�Soil�pH�reflects�the�intensity�of�acidity�that�in�turn�influences�soil�conditions�and�

plant� uptake� of� metal� contaminants.� pH� influences� the� solubility� and� activity� of� various�

1.�Introduction�

9�

biologically� important� elements� and� processes.� Depending� on� the� soil:water� ratio� and� the�

composition�and�temperature�of�the�equilibration�solution,�the�resulting�pH�will�vary.�The�pH�

of�surface�soils� (0�100mm)�commonly� range� from�6.0� to�8.0,�and� it� is�useful� to�note� that�pH�

values�around�4.0�or�less�suggest�the�presence�of�sulfides,�while�levels�above�8.5�are�indicative�

of�the�presence�of�significant�quantities�of�exchangeable�Na+.��

Soil� salinity.� Soil� salinity� is� estimated� from� the� electrical� conductivity� (EC)� of� a� soil�

saturated�paste.�The�electrical� conductivity� (EC)�of�a� soil� suspension�provides�an�estimate�of�

the�concentration�of� soluble�salts� in� the�soil,�mostly�due� to�predominantly� cations�Na+,�Mg2+�

and� Ca2+� and� anions� Cl�,� SO42�� and� HCO3

�.� Typical� soil:water� ratios� (deionized� or� distilled)�

employed�to�determine�salinity�are�1:1,�1:2,�1:2.5�and�1:5�although�the�1:5�ratio�is�preferred�as�

it� gives� an� approximation� of� soil� ionic� strength� [17].� Salinity� measurements� provide�

information� about� the� ability� of� a� site� to� support� plant� growth� as�well� as� some� information�

regarding�potential�leaching�and�drainage�problems.�Electrical�conductivity�is�a�gross�measure�

of�dissolved�salts�in�soil�solution,�but�provides�no�information�as�to�which�salts�are�present�and�

in�what�proportion.�For�non�sensitive�plants,�EC�measurements�<�4dS�m�1�are�satisfactory.�Soils�

with� EC� >� 4dS� m�1� are� considered� saline� and� plant� growth� may� be� inhibited.� Electrical�

conductivity� values� increase� with� increasing� temperature� and� must� be� corrected� if� not�

measured�at�25ºC�[18].��

Organic�matter.�The�loss�on�ignition�(LOI)�method�is�a�simple�and�relatively�inexpensive�

method� for� determining� organic� matter� [19].� The� method� is� based� on� differential� thermal�

analysis�of�the�sample�weight�after�heating�at�500�550ºC�to�oxidize�the�organic�matter�to�CO2�

and� SO2.� However,� the� ignition� temperature� and� the� heating� time� influence� the� results� as�

organic�matter�may�not�be�completely�converted�into�CO2�and�SO2�if�temperature�is�too�low�or�

if� burning� time� is� too� short.� If� temperature� is� too� high� or� heating� too� long,� inorganic�

compounds� such�as� carbonates� and� sulfate�may�be�also� converted� to�CO2� and� SO2� [20].� Soil�

organic�matter�affects�the�chemical�and�physical�properties�of�the�soil�increasing�the�soil�buffer�

capacity,� so� the� presence� of� organic�matter� tends� to� lower� pH� variations.� Furthermore,� the�

retentive�capacity�of�organic�matter�is�greater�than�most�reactive�clays.�

Carbonate�content.�Carbonate�plays�an� important� role� in�soil� chemistry� influencing� the�

pH�of�soils�given�its�carbonate�bicarbonate�buffering�equilibrium�[21].�In�addition,�carbonates�

can�complex�several�cations,�thus�affecting�the�amount�of�exchangeable�cations,�the�presence�

of�easily�soluble�salts,�the�redox�potential�and�the�partial�pressure�of�CO2�in�the�soil�air�[22].�

1.�Introduction��

10�

1.3.3.�METAL�ANALYSIS�

One� of� the� most� critical� properties� of� metals,� which� differentiate� them� from� organic�

pollutants,�is�that�they�are�not�biodegradable�in�the�environment�[23].�As�a�result�metals�tend�

to�persist� in� the�various� reservoirs�of�natural� systems�such�as�water,� soils�and�sediments,�or�

accumulate�in�biological�systems,�leading�to�an�important�hazard�to�environment�and�human�

health.� In�any�case,�a� typical� feature�of� the�weathering�of�mining�waste,�apart� from�possible�

acidic�water�formation,�is�the�release�of�metals�from�the�mineral�matrix�into�the�environment.��

To� determine� metal� concentration� on� solid� samples� from� polluted� sites,� normally,�

analytical� methodologies� based� on� recommended� methods,� are� applied� for� water,�

wastewater,� sludge,� and� agricultural� soils.� Chemical� analysis� of� polluted� soil� samples� can� be�

difficult� because� of� interferences� due� to� the� complex� soil� matrix� (e.g.� mixture� of�

elements/pollutants� at� high� concentrations� such� as� Al,� Fe� or� Ca� and� mixtures� of� organic�

compounds�such�as�PAHs,�PCBs�or�hydrocarbons),�so�usually�the�analysis�of�metals�on�soils�is�

performed�after�digestion�with�a�strong�acid�solution�in�conjunction�with�a�hot�plate,�a�boiling�

device�or�microwave�heating� system.�After�digestion,� the� samples�are�analyzed�by�means�of�

atomic� absorption� spectrometry� (AAS),� inductively� coupled� plasma� optical� emission�

spectroscopy� (ICP�OES)� or� inductively� coupled� plasma� mass� spectroscopy� (ICP�MS).� In� this�

sense,� wet� chemistry� instrument� techniques� for� elemental� analysis� require� destructive� and�

time�consuming� sample� preparation,� often� using� concentrated� acids� or� other� hazardous�

materials.� Moreover,� the� sample� is� destroyed� and� hazardous� waste� streams� are� generated�

during�the�analytical�process�requiring�disposal.�All�these�factors�lead�to�a�relatively�high�cost�

per�sample.�However,�wet�chemistry�instrumental�analysis�techniques�are�still�necessary�when�

lower� elemental� concentrations� are� the� primary�measurement� need.� Thus,� during� the� lasts�

years�X�ray�fluorescence�(XRF)�has�emerged�as�a�valuable�tool�for�the�measurement�of�heavy�

metals� in� the�environment� given� their� reliable� and� rapid�measurement� [24,�25,� 26,� 27].� XRF�

analytical�methodology�is�often�chosen�as�the�most�appropriate�when�there�is�no�historical�site�

information�as�initial�sampling�costs�are�reduced�and�analyses�are�conducted�quickly�and�with�

less�rigorous�sample�preparation.��

1.4.�SOIL�QUALITY�REGULATIONS�

Soil�provides�us�with�food,�biomass�and�raw�materials.�It�serves�as�a�platform�for�human�

activities�and�landscape�and�as�an�archive�of�heritage�and�plays�a�central�role�as�a�habitat�and�

gene� pool.� It� stores,� filters� and� transform�many� substances,� including� water,� nutrients� and�

carbon.� Thus,� soil� contamination� may� have� important� consequences� affecting� ecological�

1.�Introduction�

11�

systems�and�biological� cycling�of�nutrients�or�being�unable� to�act�as� filter�and�buffer.� In� this�

sense,�hydrosphere,�groundwater�resources�and�aquatic�ecosystems�can�be�threatened�[28].�In�

cases� of� severe� contamination� and� in� places� where� risks� to� human� health� and/or� the�

environment�are�observed,�soil�remediation�is�necessary.��

During�the�last�years,�soil�protection�policies�have�been�developed�and�implemented�in�

several� countries� focused� on� different� contaminants,� diverse� land� uses� and� on� varied�

contamination�sources�(as�for�example�mining�and�industrial�activities,�agricultural�practices�or�

oil�spills)�such�as�the�Netherlands�Guide�[29]�and�the�French�Guidelines�values�[30]�(Table�1.3)�

and�at� international� level� in�the�European�strategy� for�soil�protection�framework�[31].�These�

guidelines�define�different�quality�standard�values�based�on�the�total�concentration�of�several�

trace� metals� in� soils� and� sediments� to� facilitate� decisions� on� intervention� in� soils� after�

determining� the�existence�or�not�of� contamination,� considering� the�actual�or� future� soil� use�

(natural�park,�agricultural,�residential,�recreational�or�industrial).�

Table�1.3.�Guidelines�national�values�for�heavy�metals�in�soils�for�Netherlands,�France�and�Catalonia�(Spain)�

Metals� Netherlands[29]� France[30]� Spain�(Catalonia)[32]�

�Target�Value�

Intervention�value�

Sensitive�use�

Non�sensitive�Use�

Industrial�use�

Urban�use�

Other�uses�

As� 29� 55� 37� 120� 30� 30� 30�Pb� 85� 530� 400� 2,000� 550� 60� 60�Cd� 0.8� 12� 20� 60� 55� 5.5� 2.5�Cu� 36� 190� 190� 950� � � �

Cr�(total)� 100� 380� 130� 7,000� � � �Cr(III)� � � � � 1,000� 1,000� 50�Cr(VI)� � � � � 25� 10� 1�Hg� 0.3� 10� 7� 600� 30� 3� 2�Ni� 35� 210� 140� 900� 1,000� 470� 45�Zn� 140� 720� 9,000� �� 1,000� 650� 170�

The� target� value� is� the� baseline� concentration� value� below�which� compounds� and/or�

elements� are� known� or� assumed� not� to� affect� the� natural� properties� of� the� soil� while� the�

intervention� value� is� the� maximum� tolerable� concentration� above� which� remediation� is�

required�and�becomes�mandatory.��

In� this� sense,� it� is� intended� that� guideline� values� could� represent� an� indication� to� an�

assessor�that�soil�concentrations�above�this�level�could�pose�an�unacceptable�risk�to�the�health�

of� site� users� and� that� further� investigation� and/or� remediation� is� required.� As� the�

concentration� of� metals� on� plants� does� not� necessary� correlate� with� the� total� content� of�

metals�in�related�soils,�these�values�represent�an�estimation�of�the�potential�hazard,�although�

it� can�be� considered�as� the�most�pessimistic� interpretation�as� it� is� considered� that� the� total�

1.�Introduction��

12�

amount�of�metal�in�soil�is�available�to�be�absorbed�by�plants�or�can�be�mobilized.�So,�not�only�

the�total�content�of�heavy�metals�should�be�considered,�but�also�its�mobility�and�bioavailability�

to�determine�the�real�toxicity.�

1.5.�CHEMICAL�SPECIATION�AND�FRACTIONATION�IN�SOILS�

Mobility� and� bioavailability� of� metals� in� the� environment� depends� strongly� on� their�

specific�chemical�forms�or�types�of�binding�rather�than�the�total�element�content�[33,�34].�So�it�

can�be�generally�considered�as�an�indication�of�toxicity�and�consequently�the�chemical�species�

present�in�a�soil�should�be�determined�in�order�to�assess�the�toxic�effects.�However,�nowadays�

this� distinction� is� not� reflected� in� the� legislation,�which� account� for� the� total� content�of� the�

pollutants�rather�than�for�the�available�content�of�the�pollutants.��

In� this� regard,� the� characteristics� of� just� one� species� of� an� element�may� have� such� a�

radical�impact�on�living�systems�(even�at�extremely�low�concentrations)�that�the�total�element�

concentration�becomes�of� little�value� in�determining� the� impact�of� the� trace�element.�Good�

examples�are�mercury�and�tin.�The�inorganic�forms�of�these�elements�are�much�less�toxic�(or�

even�do�not� show�toxic�properties)� than�the�alkylated� forms�which�are�highly� toxic.� In� these�

sense,�it�is�necessary�to�evaluate�and�characterize�the�chemical�forms�of�the�elements�in�order�

to� understand� their� properties,� their� evolution� possibilities,� as�well� as� the� prediction� of� the�

related�environmental� consequences.�Such�characterization� is� carried�out� the�methodologies�

known�by�speciation�analysis.��

At� this� point� it� is� required� to� define� the� term� speciation.� The� IUPAC� has� defined� the�

terminology�on�elemental�chemical�speciation�as�follows�[35]:�

Chemical� species:� Specific� form� of� an� element� defined� as� to� isotopic� composition,�

electronic�or�oxidation�state,�and/or�complex�or�molecular�structure.�

Speciation�analysis:�Analytical�activities�to�identifying�and/or�measuring�the�quantities�of�

one�or�more�individual�chemical�species�in�a�sample.�

Speciation�of�an�element:�Distribution�of�an�element�amongst�defined�chemical�species�

in�a�system.�When�elemental�speciation�is�not�feasible,�fractionation�is�employed.�

Fractionation:�Process�of�classification�of�an�analyte�or�a�group�of�analytes�from�a�certain�

sample� according� to� physical� (e.g.,� size,� solubility)� or� chemical� (e.g.� bonding,� reactivity)�

properties.��

Although�no�generally�accepted�definition�of�the�term�exists,�speciation�can�broadly�be�

defined� as� the� identification� and� quantification� of� the� different,� defined� species,� forms� or�

phases�in�which�an�element�occurs�[36].�The�term�"fractionation"�(also�referred�to�as�indirect�

1.�Introduction�

13�

speciation)�is�frequently�used�interchangeably�with�speciation�but�emphasizes�the�concept�of�

subdividing�a�"total�content".�Also,�the�analytical�preparations�for�separating�metal�species�are�

referred�to�as�"fractionation".�An�overview�of�techniques�used�in�chemical�speciation�analysis�

is�given�in�Table�1.4�[37].�

Table�1.4.�Analytical�methods�applied�for�chemical�speciation�of�metals�

Method� Metal�species�determined�

Electroanalysis�Ion�selective�electrodes� Free�ionic�concentrations�Voltammetry� Free�ions�and�labile�complexes�

Spectroscopy�Spectrophotometry� Specific�forms�Hydride�generation� Inorganic� and� organometallic� species;� different�

oxidation�states�(Sn,�As,�Sb,�Bi,�Se,�Te)�

Synchrotron�X�ray�spectroscopy� Specific�forms��Chromatography�HPLC� Cations,�anions,�metal�complexes,�inorganic�species�GC�or�LC� Organometallic�compounds�of�mercury,�tin�and�lead�Physico�chemical�fractionation�Ion�exchange�resin� Free�ions�and�labile�complexes�UV�irradiation� Organic�complexes�Solvent�extraction� Organic�complexes�

Size�fractionation�Filtration� Dissolved�and�suspended�matter�associated�Centrifugation� Dissolved�and�suspended�matter�associated��Dialysis� Different�charge,�different�molecular�size�Ultrafiltration� Molecular�size�

LIQUID�PHASE�

Gel�filtration�chromatography� Free� forms� and� complexes� of� different� molecular�size�

Single�reagent�leaching� Reagent�soluble�fractions�Sequential�extractions� Geochemical�fractions�Ion�exchange�resins� Labile�fractions�

SOLID�PHASE

Synchrotron�X�ray�spectroscopy� Specific�forms��

Whilst� different� elemental� speciation�methods� are� available� for� aqueous� systems� [38],�

implementations� of� methodologies� for� speciation� studies� in� solids� have� been� less� well�

developed.�The�speciation�studies�involving�soil�and�sediment�analysis�are�often�based�on�the�

use�of�extraction�procedures�(single�or�sequential).��

The� determination� of� specific� chemical� species� or� binding� forms� is� difficult� and� often�

hardly� possible.� Therefore,� in� practice,� determinations� of� broader� “operationally� or�

functionally�defined”� forms�or�phases� can�be�a� reasonable� compromise� to�arrive�at� a� sound�

environmental�policy.� In�this�regard,�single�and�sequential�extraction�schemes�were�designed�

in�the�1980s� in�order�to�assess�the�different�retention/release�of�metals� in�soil�and�sediment�

samples�as�a�result�of�natural�processes�or�anthropogenic�activities�and�can�be�employed�as�a�

valuable�tool�to�assess�the�potential�impact�in�the�environment�of�mining�districts�[39,�40].��

1.�Introduction��

14�

1.5.1.�SEQUENTIAL�EXTRACTION�SCHEMES�

Despite�being�quite� laborious,�sequential�extraction�schemes�(SES)�have�been�the�main�

tools� employed� to�estimate� the�availability�of� contaminants� in�polluted� soils,� sediments� and�

sludge� [41,�42,�43].� SES�procedures� try� to�mimic� the�various�natural� conditions�under�which�

soils� may� release� metals� into� the� environment� using� sequentially� leaching� reagents� of�

increasing� strength.� The� determination� of� these� metal� fractions� allows� certain� predictions�

regarding�the�possible�release�of�a�given�analyte�(metal)�from�a�soil�or�sediment�phase�under�

certain�conditions�of�gradual�lixiviation�power.��

The�applied�strategy�consists�on�the�use�of�reagents�able�to�selectively�dissolve�a�metal�

fraction�bonded�to�certain�soil�materials,�i.e.�water�soluble�compounds,�exchangeable�cations,�

carbonates,� easily� reducible,� oxidizable�phase�and� residual.� These� fractions�may� vary� among�

different� extraction� schemes.� Most� common� reagents� used� include:� no� hydrolysable� salts,�

weak�acids,�reducing�agents,�oxidant�agents�and�strong�acids�[44].��

Several�SES�schemes�have�been�developed� to�evaluate�metal� fractionation� in� soils�and�

sediments�normally�varying�in�the�number�of�extraction�steps�between�3�and�8�and�those�most�

widely�used�are�Tessier�[35,�42]�and�BCR�SES�[45,�46]�(Table�5).�Comparatively,�both�methods�

provide� a� similar� fractionation,� although� the� exchangeable� fraction� of� BCR� resumes�

“exchangeable”�and�“carbonate”�fractions�from�Tessier.�

Table�1.5.�Sequential�extraction�procedures�defined�by�Tessier�and�BCR�SES�applied�to�1g�of�sample�Method� Fraction� Extraction�conditions�

T1:�Exchangeable� 8�mL�1M�MgCl2�pH�7,�25ºC,�1h�T2:�Link�to�carbonates� 8�mL�1M�CH3COONa�+�CH3COOH,�pH�5,�25ºC,�5h�T3:�Link�to�iron�and�manganese�oxides�

20�mL�0.04M�NH2OH�HCl�(25%�v/v�CH3COOH),�96ºC,�6h�

T4:�Link�to�organic�matter�3�mL�0.02M�HNO3�+�2�ml�30%�H2O2�(pH�2),�85ºC,�2h;�3�mL�30%�H2O2�(pH�2),�85ºC,�2h;�5�mL�3.2M�CH3COONH4�

in�20%�HNO3�+�7mL�H2O�25ºC,�30�min�

Tessier�

T5:�Residual�7.5�mL�37%�HCl�+�2.5�mL�65%�HNO3,�25ºC�during�1�

night,�reflux�2h�Water�soluble,�exchangeable�and�link�to�carbonates�

20�mL�0.1M�CH3COOH,�25ºC,�16h�

Link�to�iron�and�manganese�oxides�

20�mL�0.5M�NH2OH�HCl,�pH�2,�25ºC,�16h�BCR�SES�

Link�to�organic�matter�and�sulfides�

5�mL�30%�H2O2,�25ºC,�1h�+�5�mL�30%�H2O2,�85ºC,�1h�+�25�mL�1M�CH3COONH4,�pH�2,�25ºC,�16h�

Nevertheless,�SES�have�several�drawbacks�mainly�related�to�the�excessive�time�required�

(a�traditional�sequential�extraction�requires�at�least�50�hours)�and�the�possible�modification�of�

the�metal� species� during� extraction� procedure� [47].� It� is� worth�mentioning� that� not� all� the�

fractions�obtained�from�applying�SES�are�equally�important�from�the�environmental�risk�point�

of�view.�The�metals�related�to�the�residual�fraction�(obtained�through�extraction�or�digestion�

1.�Introduction�

15�

with� mixtures� of� strong� acids)� are� unlikely� to� be� released� under� weathering� conditions;�

whereas�metals� linked�to� the�soluble�and�exchangeable� fractions,�and�those�related�to�more�

labile�metal�species�are�more�mobile�and�hence�more�available.�Therefore,�in�order�to�assess�

the� environmental� hazard,� efforts� should� be� applied� only� on� the� measurement� of� these�

fractions�opening�the�possibility�of�using�less�laborious�methods�based�on�the�extraction�of�the�

metal�fraction�of�interest�using�a�unique�extracting�reagent�(single�leaching�tests).��

1.5.2.�SINGLE�LEACHING�TESTS�

Single� leaching�tests�are�nonselective�extractions�that�target�groups�of� labile�or�mobile�

phases.�This�approach�can�provide�a�useful�assessment�for�screening�purposes�to�identify�trace�

metal� pollution� with� minimum� time� consumption� [48].� Single� extractants� differ� by� their�

dissolution�power,� including:� i)�mild�unbuffered�extractants�that�extract�the�fraction�of�easily�

exchangeable� elements;� ii)� acidic� extractants� that� release� the� fraction� remobilized� by�

acidification�processes;�and�iii)�complexing�reagents�(Table�1.6).�[49].��

Table�1.6.�Leaching�tests�used�in�soil�analysis�[50]�

Group� Type�and�solution�strength� References�

HNO3�0.43�2.0�M� [51]�Aqua�regia� [52]�HCl�0.1�1�M� [53]�CH3COOH�0.1�M� [54]�

Acid�extraction�

HCl�0.05�M�+�H2SO4�0.0125�M� [55]�EDTA�0.01�0.05�M�at�different�pH� [53]�DTPA�0.005�M�+�TEA�0.1�M�+�CaCl2�0.01�M� [56]�

Chelating�agents�

CH3COOH� 0.02� M+NH4F� 0.015� M� +� HNO3�0.013M�+�EDTA�0.001�M�

[57]�

NH4�acetate,�acetate�acid�buffer�1�M�pH=7� [58]�Buffered�salt�solution�NH4�acetate,�acetate�acid�buffer�1�M�pH=4.8� [53]�CaCl2�0.01��0.1�M� [53]�NaNO3�0.1�M� [58]�NH4NO3�1M� [53]�AlCl3�0.3�M� [59]�

Unbuffered�salt�solution�

BaCl2�0.1�M� [60]�

Leaching� tests� are� focused� on� providing� information� about� the� release� of� specific�

components�under�given�conditions,�or�under�conditions�that�may�approximate�more�closely�

or� simulate� the� actual� field� situation� under� consideration.� Such� conditions� try� to� reproduce�

those� chemical� reactions� that� can� take� place� in� soils� and� sediments� on� a� particular�

environment� (i.e,� adsorption–desorption,� dissolution�precipitation,� reduction–oxidation,� and�

complexation�decomplexation�processes),�and�can�modify�the�concentration�of�metals� in�soil�

solution� [61,56].� The� application� of� these� procedures� to� polluted� or� naturally� contaminated�

soils�is�mainly�focused�to�ascertain�the�potential�availability�and�mobility�of�metals,�in�studies�

1.�Introduction��

16�

on� the� soil�plant� transference� and� metal� migration� in� a� soil� profile� due� to� groundwater�

transport.� Subsequently,� this� information� is� usually� used� for� the� risk� assessment� of� wastes�

when�they�are�deposited�in�a�landfill�or�to�characterize�and�classify�them�in�terms�of�risk�[62].�

At� present,� several� single� extraction� procedures� (leaching� tests)� based� on� aqueous� or� acidic�

extractions� are� widely� approved� analytical� tools� in� national� and� international� legislation�

organisms� such� as� Germany� [63],� France� [64],� Italy� [65]� and� The� Netherlands� [66]� amongst�

others.��

Hence,�leaching�tests�such�as�(NH4)2SO4�or�HCl�single�non�selective�extractions�methods,�

can� provide� a� useful� assessment� for� screening� purposes� to� identify� labile� or�mobile� phases�

[45].� In� addition,� it� is� demonstrated� a� correlation� between� the� mobility� observed� by� some�

leaching�tests�and�the�mobility�provided�by�the�sum�of�different�stages�of�SES�[67].�The�main�

advantages�of�these�single�leaching�tests�against�SES�are�mainly�related�to�their�cost�efficiency,�

easy� to� use� and� a� reduction� on� bias� induced� by� sequential� translation� and� accumulation� of�

procedural�errors.��

However,� despite� being� very� useful� tools� for� environmental� assessment� of� chemical�

species,� SES� and� single� leaching� procedures� cannot� provide� direct� speciation� of� soils.� In�

addition,�SES�are�destructive�techniques.�For�direct�speciation,�synchrotron�based�techniques�

have�arose�as�a�valuable�tool�by�means�of�techniques�such�as�X�ray�Absorption�Spectroscopy�

(XAS)�using�synchrotron�facilities�as�X�rays�radiation�sources.�

1.6.�SOIL�RISK�ASSESSMENT�TOOLS�

Various�tools�can�be�employed�to�determine�the�degree�of�contamination�of�a�specific�

site,� like� an� abandoned� mine� site,� such� as� concentration� enrichment� factors,� geographic�

information� systems� or� the� use� of� statistical� tools.� These� tools� can� be� used� for� a� better�

determination�of�the�risk�of�a�contaminated�site�and�thus�help�the�decision�making.�

1.6.2.�GEOGRAPHIC�INFORMATION�SYSTEMS�

Further� characterization� in�environmental� studies�of�polluted� soils� is� achieved� through�

the�determination�of� the�pollutants� spatial� variability� in�a�polluted�area� through�Geographic�

Information�Systems�(GIS)�[68,�69,�70].�From�a�practical�point�of�view,�they�provided�the�first�

reference�values�for�assessing�soil�contamination�at�a�given�potentially�polluted�site.�Assessing�

the� spatial� extent� of� soil� metal� concentration� is� also� a� powerful� tool� in� understanding� and�

monitoring�the�adverse�effects�of�contamination.�Soil�maps�are�used� in�soil�description,� land�

appraisal� (taxation),� and� for� soil� monitoring� sites� to� establish� the� basic� information� on� the�

1.�Introduction�

17�

genesis� and� distribution� of� naturally� occurring� or� man�made� soils,� their� chemical,�

mineralogical,�biological�composition,�and�their�physical�properties�at�selected�positions.��

Spatial� variability� of� soil� properties� and� pollutants� concentration� can� be� done� with�

different� interpolation�methods� such�as� inverse�distance�weighting� (IDW),�Kriging�and� spline�

functions�[71].�While�spline�methods�involve�a�considerable�interpolation�error�when�there�are�

large�changes�in�the�surface�values�within�a�short�horizontal�distance,�Kriging�method�may�not�

be�met�in�practice�unless�employing�100�samples�in�order�to�obtain�a�reliable�variogram�that�

correctly� describes� spatial� structure.� In� contrast,� IDW� interpolator� assumes� that� each� input�

point�has�a�local�influence�that�diminishes�with�distance�[72],�and�no�assumptions�are�required�

for�the�data,�being�this�method�suitable�for�irregular�samplings�[73].�

Combining�Geographic� Information�Systems�(GIS)�with�some�analytical�tools�the�spatial�

variability�in�a�mine�area,�can�be�determined.�Such�combination�let�to�produce�maps�which�are�

helpful�for�a�cost�effective� identification�of�the�sources�and�the�spatial�patterns�of�pollutants�

[71,�72,�74].��

1.6.3.�PRINCIPAL�COMPONENT�ANALYSIS�IN�GEOSCIENCES�

Within�the�environmental�investigations�associated�with�the�impact�of�metals�in�soils,�it�

is� often� necessary� the� determination� of�multiple� parameters,� obtaining�multivariate� data.� A�

first� comparison�between� samples�across� individual�parameters� can�be�performed,� although�

when�the�simultaneous�consideration�of�all�the�parameters�determined�is�carried�out�(known�

as�multivariate�methods� of� analysis),� a� characterization� of� the� combined� effect� of� different�

variables�and�especially�of�the�various�relationships�between�them�can�also�be�obtained.��

Principal� Component� Analysis� (PCA)�was� invented� in� 1901� by� Karl� Pearson� [75]� and� is�

nowadays� an� extremely� useful� technique� to� "summarize"� all� the� information� in� a� more�

understandable� form.�Typically,�PCA� is�used�to�reduce�the�dimensionality�of�a�dataset,�while�

retaining�as�much�of�the�original�information�as�possible.��

PCA�works�by�decomposing�the�X�matrix�that�contains�all�the�data�as�the�product�of�two�

smaller�matrices,�which�are�called�the�loading�and�score�matrices:�

X�=�TPT�+�E� � � (Equation�1.1)�

The�loading�matrix�(P)�contains�information�about�the�variables.�It�is�composed�of�a�few�

vectors�(Principal�Components,�PCs)�which�are�(obtained�as)�linear�combinations�of�the�original�

X�variables.� The� score� matrix� (T)� contains� information� about� the� objects.� Each� object� is�

described� in� terms� of� its� projections� onto� the� PCs,� (instead� of� the� original� variables).� The�

1.�Introduction��

18�

information�not�contained�in�these�matrices�remains�as�"unexplained�X�variance"�in�a�residual�

matrix�(E)�which�has�exactly�the�same�dimensionality�as�the�original�X�matrix.�

The�PCs,�among�many�others,�have�two�interesting�properties:�

� They� are� extracted� in� decreasing� order� of� importance.� The� first� PC� always� contains�

more�information�than�the�second,�the�second�more�than�the�third�and�so�on...�

� They� are� orthogonal� to� each� other.� There� is� absolutely� no� correlation� between� the�

information�contained�in�different�PCs�[76].

PCA� is�sensitive�to�the�relative�scaling�of�the�original�variables�so�centering�of�the�data�

for�each�attribute�is�previously�required.�The�results�of�a�PCA�are�usually�discussed�in�terms�of�

component�scores�(the�transformed�variable�values�corresponding�to�a�particular�case� in�the�

data)� and� loadings� (the� weight� by� which� each� standardized� original� variable� should� be�

multiplied�to�get�the�component�score)�[77].��

Often,� PCA� can� be� thought� of� as� revealing� the� internal� structure� of� the� data� in� a�way�

which�best�explains�the�variance�of�the�data.�If�a�multivariate�dataset�is�visualized�as�a�set�of�

coordinates� in� a� high�dimensional� data� space� (1� axis� per� variable),� PCA� can� supply� the� user�

with� a� lower�dimensional� picture,�when� viewed� from� its�most� informative� viewpoint.� This� is�

done� by� using� only� the� first� few� principal� components� so� that� the� dimensionality� of� the�

transformed�data�is�reduced.��

So,�by�using�PCA�a�large�number�of�variables�such�as�concentration�of�elements�can�be�

transformed� into� linearly� independent� sources�of� “information”� (referred� to�as� components)�

that�can�be�interpreted�to�provide�insight�into�the�processes�or�interrelationships�that�underlie�

the� data.� Principal� components� analysis� is� commonly� used� in� a� variety� of� geosciences�

disciplines,�such�as�aeolian�applications�to�help�determine�source�regions�of�particulate�matter�

pollution�[78,�79],�to�characterize�particle�size�data,� including�studies�of�soil� fertility� [80]�and�

pollution� [81]�while� textural� and�other� sedimentological�data�has�been�used� to�differentiate�

between� marine� and� terrestrial� sediments� in� Florida� [82].� Thus,� interpretation� of� PCA�

components�can�help�identify�source�of�contamination�as�well�as�contamination�patterns.��

PCA�[83]�was�also�applied�to�explain�the�underling�structures�of�the�obtained�data,�i.e.�to�

identify� pollutant� sources,� certain� distribution� patterns� and� their� contributions� on� soils�

affected�by�mining�pollution.�

1.�Introduction�

19�

1.7.�WEAKNESS�AND�NEEDS�OF�MINING�SITES�CHARACTERIZATION�

� Weaknesses:�

� Soil� risk� assessments� covered� by� the� current� legislations� consider� only� heavy� metal�

concentration� and� do� not� account� for� the� real� risk� of� contaminated� sites� such� as�

abandoned� mines� better� explained� by� heavy� metals� mobility� determined� by� its�

chemical� species.� In� addition� the� origin� of� the� contamination� (either� natural� or�

anthropogenic)�is�not�considered.�

� A�great�deal�of�mines�is�abandoned�every�year�without�concern�for�the�environment.�

This� is� especially� dramatic� in� developing� countries� were� no� legislation� concerning�

contaminated�soils�are�implemented.��

� Needs:�

� The� characterization� of� several� parameters� including� not� only� heavy� metal�

concentration�but�also�its�distribution�around�the�polluted�area,�the�physico�chemical�

parameters,�the�origin�of�the�contamination�and�the�mobility�of�the�heavy�elements�of�

soils�should�be�performed�to�assess�environmental�health�hazards.�

1.�Introduction��

20�

REMEDIATION�TECHNIQUES�OF�INDUSTRIAL�CONTAMINATED�WATER�

The�development�of�viable�ways�of�recycling�industrial�waters�and�their�derivate�sludge�

such�as�mining�effluents� rather�than� its�disposal�as�a�hazardous�waste� in�specially�controlled�

landfills� can� be� a� benefit� from� both� environmental� and� economical� point� of� view.� In� this�

section,� the� application� of� various� remediation� techniques� for� the� treatment� of� organic� and�

inorganic�contaminated�wastewaters�is�resumed.�

1.8.� SOLVENT� EXTRACTION� FOR� THE� RECOVERY� OF� ZN� FROM� ACIDIC�MINE�WATERS�

Refused�mining� tailings� and�water� containing� rejected�materials� are� generally� pumped�

into�tailing�ponds�to�avoid�their�transportation�by�wind�into�populated�areas�where�the�toxic�

chemicals�could�be�dangerous�to�human�health�as�well�as�to�allow�the�sedimentation�of�solid�

particles[84].�However,�mine�tailing�ponds�are�potentially�hazardous�as� they�can�represent�a�

source�of�acid�drainage�but�especially�due�to�dam�failures�of�tailing�ponds.�When�a�tailing�line�

breaks�or�a�dam�breaches,�high�amounts�of�contaminated�water�and�clays�with�dissolved�metal�

ions�are�released�to�the�environment�causing�serious�damages�and�having�toxic�effects�on�the�

biota� in� the� downstream� water[85].� Some� mining� operations� are� able� to� recycle� relatively�

small�amounts�of�water�for�drilling�and�dust�suppression�by�using�simple�sumps�to�clarify�the�

water,�but�in�most�cases,�the�total�volume�is�pumped�to�the�surface�for�their�treatment�with�

the�other� aqueous�drainage� components� so� as� to�minimize� long�term�environmental� effects�

once�active�mining�has�ceased�[86].��

The�volume�and�characteristics�of�materials� contained� in� tailing�ponds�can�vary�widely�

depending�on�mining�methods�and�the�hydrogeological�characteristics�of�the�region.�Water�pH�

can�vary� from�basic� to�very�acid�depending�on�the�nature�of� the�ore�and� its�host�rock�and� it�

may�contain�high�levels�of�dissolved�metals,�suspended�solids,�some�oils�and�ammonia.�

As�an�example�of�economical�feasible�and�valuable�recovery�of�a�heavy�metal�byproduct,�

the� recovery� of� Zn� from� mine� waters� can� diminish� the� volume� of� hazardous� materials�

contained�in�the�mine�tailing�while�providing�economical�profit.�

1.�Introduction�

21�

1.8.1.�ZINC�OVERVIEW�

Zinc�is�the�23rd�most�abundant�element�in�the�earth's�crust.�Zinc�is�necessary�to�modern�

living,� and,� in� tonnage� produced,� stands� fourth� among� all�metals� in�world� production� being�

exceeded�only� by� iron,� aluminum,� and� copper.�Over� 11�million� tonnes�of� zinc� are� produced�

annually�worldwide.�Nearly�50%�of�the�amount� is�used�as�a�coating�to�protect� iron�and�steel�

from�corrosion� (galvanized�metal).�Approximately� 19%� is�used� to�produce�brass�and�16%�go�

into�the�production�of�zinc�base�alloys�to�supply�the�die�casting�industry.�Significant�amounts�

are�also�employed�for�compounds�such�as�zinc�oxide�and�zinc�sulfate�and�semi�manufactures�

including� roofing,� gutters� and� down�pipes� [5]� (Figure� 1.3a).� Main� application� areas� are:�

construction�(45%)�followed�by�transport�(25%),�consumer�goods�&�electrical�appliances�(23%)�

and�general�engineering�(7%)�(Figure�1.3b).�Zinc�is�also�a�necessary�element�for�proper�growth�

and�development�of�humans,�animals,�and�plants;�it�is�the�second�most�common�trace�metal,�

after�iron,�naturally�found�in�the�human�body.��

47%

19%

16%

7% 7% 4%

Zinc�coated�steels Brass�

Zinc�base�alloys Semi�manufactures

Compounds Others

First�use

��

45%

25%

23%7%

Construction Transport

Consumer�&�Electrical�Goods General�Engineering

End�use

Figure�1.3.�Zinc�demand:�First�use�and�End�use�in�2003�estimate�(Source�:�ILZSG�/�Brook�Hunt�/�Outokumpu�/�CRU�[87])�

The� level�of� zinc� recycling� is� increasing�each�year,�as�a�consequence�of� the�progress� in�

the�technology�of�zinc�production�and�zinc�recycling.�Today,�over�80%�of�the�zinc�available�for�

recycling� is� indeed� recycled.� Although,� at� present,� approximately� 70%� of� the� zinc� produced�

worldwide�is�still�originated�from�mined�ores�[86].��

In� this� context,� it� is� clear� the� need� for� the� recovery� of� Zn.� Several� technologies� are�

currently�employed� for� separation�of� zinc� from�waters� including�precipitation,� ion�exchange,�

adsorption,�electrochemical� recovery,�membrane� separation�and� solvent�extraction� (SX)� [88]�

being�the�latter�the�most�economical�and�practical�process�to�extract�Zn�from�industrial�waters�

[89,�90,�91].�In�recent�years�SX�has�become�essential�to�the�hydrometallurgical�industry�due�to�

a�growing�demand�for�high�purity�metals,�rigid�environmental�regulations,�the�need�for�lower�

1.�Introduction��

22�

production�costs,�as�well�as�due�to�the�diminishing�production�in�high�grade�ore�reserves�[92,�

93].�

1.8.2.�THE�SOLVENT�EXTRACTION�PROCESS�

Solvent�extraction�involves�the�extraction�of�the�target�element�from�the�initial�aqueous�

solution�by�an�extractant�usually�diluted�in�an�organic�solvent�(organic�phase),� leaving�all�the�

other� constituents� in� the� aqueous� raffinate.� A� subsequent� reextraction/stripping� of� the�

extracted�element�present�in�the�organic�phase�is�usually�carried�out�with�some�acidic�solution�

(stripping� solution)� with� higher� affinity� for� the� target� element� than� the� organic� phase.�

However,�when�undesirable�metals�are�extracted�together�with�the�target�element,�scrubbing�

of� the� solvent� previous� to� the� stripping� step� should� be� performed.� An� additionally� step� of�

regeneration�of� the�organic�phase�after� the� stripping� step�may�be�also�performed�when� the�

stripping�of�the�target�element�or�the�scrubbing�step�is�not�complete�for�further�reuse�of�the�

organic�solvent�(Figure�1.4).��

Figure�1.4.�Typical�solvent�extraction�steps�

Scrub� and� regeneration� steps� generally� increase� the� cost� of� the� process� due� to� the�

expenditure� in�both�reactants�and�time�so� it� is�preferable�to�use�a�selective�extractant�and�a�

proper�strip�solution�so�as�to�these�steps�can�be�avoided.��

Nowadays,�a�wide�number�of�extractants�are�available�for�use�in�SX�for�the�recovery�of�

metals,� some�of� them,� suitable� for� a� specific�metal,�while� others�must� be� used� at� a� certain�

conditions� to� avoid� extraction� of� impurities� [94,� 95].� In� this� sense,� the� most� widely� used�

extractants� for� Zn� recovery�are� those�corresponding� to� the�organophosphorous�acids�group,�

such� as� Di�(2�ethylhexyl)� phosphoric� acid� (DEHPA)� and� bis(2,4,4�trimethylpentyl)� phosphinic�

acid�(Cyanex�272)�(Figure�1.5).�

1.�Introduction�

23�

DEHPA (Di-(2-ethylhexyl) phosphoric acid

CAS No. 298-07-7

CYANEX 272 Bis(2,4,4-trimethylpentyl) phosphinic

acid Cas No. 83411-71-6

POHO

OOP

HO

O

CYANEX 301 Bis(2,4,4-trimethylpentyl) dithiophosphinic

acid CAS No. 107667-02-7

CYANEX 302 Bis(2,4,4-trimethylpentyl) monothiophosphinic acid

CAS No. 132767-86-3

PHS

SP

HO

S

Figure�1.5.�Chemical�structures�of�DEHPA,�Cyanex�272,�Cyanex�301�and�Cyanex�302�

DEHPA�has�been�successfully�used�as�an�extractant�for�many�metal�ions�including�Zn�due�

to�its�great�extraction�capacity�and�low�cost�[96,�97,�98].�It�has�been�used�to�extract�Zn�more�

efficiently� than� other� bivalent� metal� ions� such� as� Cu,� Ni,� Co� and� Cd� [99].� The� order� of�

extraction� of� eight�metal� ions� from� a� sulfate� solution� using� DEHPA� has� been� reported� as� a�

function�of�pH�to�be�Fe3+>Zn2+>Cu2+>Co2+>Ni2+>Mn2+>Mg2+>Ca2+� [100].� In�a�more�recent�study�

of�the�separation�of�divalent�metal�ions�from�a�synthetic�solution,�the�extraction�of�metal�ions�

was� in� the� order� Zn2+>Ca2+>Mn2+>Cu2+>Co2+>Ni2+>Mg2+� [101].� The� target� metal� (or� even�

different�metals)� can�be� separated� from�the�bulk� solution�by�varying� in� successive� steps� the�

acidic�conditions�and�the�temperature�as�main�parameters�to�get�pure�solutions�of�the�target�

metals.��

Cyanex� 272� has� been� used� as� well� as� its� thiosubstituted� derivatives� (Cyanex� 302� and�

Cyanex�301)�in�the�extraction�of�several�metal�ions�[102].�Various�studies�report�the�adequacy�

of�Cyanex�272�to�extract�Fe,�Zn,�Cr,�Cu�and�Ni�from�sulfuric�and/or�sulfate�solutions�[103,�104,�

105].��

1.8.3.�SCALING�SOLVENT�EXTRACTION�TO�A�PILOT�PLANT�

The� development� of� new� solvent� extraction� solutions� at� the� laboratory� level,� require�

from�a�pilot�plant�step�in�order�to�validate�the�concept.�The�design�of�a�pilot�plant�is�based�on�

the�data�obtained�both�in�the�laboratory�and�by�process�modeling.��

Laboratory�experiments�are�performed�to�check�the�feasibility�of�the�solvents�that�seem�

to� be� suitable� by� testing� their� chemical� (liquid�liquid� equilibrium)� and� hydrodynamic� (phase�

1.�Introduction��

24�

separation)�properties.�At� laboratory,� the�distribution�coefficient� (D)�can�be�known�for�every�

solute� so� an� approximation� of� the� number� of� steps� of� the� process� can� be� estimated.� For� a�

multi�component� process,� the� feasibility� of� separation� between� various� species� may� be�

determined�too.�In�addition,�by�means�of�kinetic�experiments�the�residence�time�as�a�function�

of� temperature� and� intensity� of�mixing� required� for� completion� of� the� process� can� also� be�

determined.��

Computer� simulation� programs� allows� to� determine� the� optimal� configuration� at� the�

pilot�plant:�temperature,�pH,�phase�ratio,�the�number�of�stages�and�optimal�flow�sheet�of�the�

process�[106].�

Thus,�the�pilot�plant�has�to�test�mainly�the�parameters�that�cannot�be�predicted�by�the�

simulation:� preferred� dispersion,� flux,� entrainment,� accumulation� phenomena,� precipitation,�

deterioration�of� the�solvent�and�the�equipment�to�be�used.� In� this�sense,� the�recommended�

equipment�is�tested�in�a�dedicated�pilot�plant,�to�estimate�the�mass�transfer,�entrainment�and�

phase� separation,� and� optionally� for� quick� accumulation� phenomena,� e.g.,� precipitation,�

foaming,�etc.�(this�last�point�can�be�checked�only�by�running�the�pilot�plant�with�real�process�

solutions)�and�to�have�an�experimental�proof�of�the�recommended�flow�sheet.�It�can�be�used�

to�discover�problems�that�could�not�have�been�otherwise�detected.��

The�configuration�and�type�of�equipment�of�the�pilot�plant�should�be�similar�to�the�full�

scale�plant,�and�it�should�be�run�in�the�designated�site,�using�the�real�raw�materials.�Its�main�

purpose� is� the� verification� of� the� results� that� were� obtained� in� the� bench�scale� regarding�

production�rate,�recovery,�product�quality�and�analysis�of�accumulation�phenomena.�

All� these� parameters� are� sufficient� for� a� rough� economical� estimate� of� the� industrial�

plant.�Every�dollar�invested�in�the�pilot�plant�pays�itself�tenfold�in�the�industrial�plant�[107].�

1.9.� FE�LOADED� MATERIALS� FOR� THE� REMOVAL� OF� ORGANIC� AND�

INORGANIC�CONTAMINANTS�

Another� technology� to� be� employed� on� the� recovery� of� inorganic� contaminants� in�

industrial�polluted�effluents�deals�with�the�ion�exchange�processes.�Specifically,�the�Fe�loaded�

materials� that�can�be�employed�not�only�to�selectively�remove�the�pollutant� from�the�target�

effluent�but�also�to�treat�organic�waste.�Through�the�next�section,�the�use�of�several�materials�

as�a�support�for�iron�to�be�used�either�as�heterogeneous�Fenton�catalyst�or�as�arsenic�sorbent�

will�be�described.��

1.�Introduction�

25�

1.9.1.�ZEOLITES�

The�word�"zeolite"�comes�from�the�Greek�“zeo”�and�“lithos”�that�means�"boiling�stone"�

because�of�the�observation�that�zeolites�release�water�when�heated.�Zeolites�are�a�large�group�

of� natural� and� synthetic� hydrated� aluminum� silicates� characterized� by� complex� three�

dimensional�structures�with�large,�cage�like�cavities�that�can�accommodate�sodium,�calcium�or�

other�cations�(positively�charged�atoms�or�atomic�clusters);�water�molecules;�and�even�small�

organic�molecules.�These�encaged�ions�and�molecules�can�be�removed�or�exchanged�without�

destroying�the�aluminosilicate�framework�[108].��

The� atomic� structures� of� zeolites� are� based�on� three�dimensional� frameworks� of� silica�

and�alumina�tetrahedra�in�a�tetrahedral�configuration,�where�each�oxygen�atom�is�bonded�to�

two� adjacent� silicon� or� aluminum� atom,� linking� them� together.� Clusters� of� tetrahedra� form�

boxlike� polyhedral� units� that� are� further� linked� to� build� up� the� entire� framework.� The�wide�

variety�of�possible�zeolite�structures� is�due�to�the�large�number�of�ways� in�which�these�units�

can�be�linked�to�form�various�structures�(Figure�1.6).�Each�type�of�zeolite�has�specific�uniform�

pore�size,�for�instance,�3.5�4.5��for�zeolite�LTA,�4.5�6.0��for�ZSM�5�and�6.0�8.0��for�zeolite�X,�

Y�type.�

Figure�1.6.�Framework�topologies�of:�a)�Sodalite;�b)�Zeolite�A/ZK�4;�c)�Zeolites�X/Y�

Zeolites� occur� naturally� as� minerals,� although,� only� 6� of� the� 63� natural� zeolites�

commonly� occur� in� large� beds:� analcime,� chabazite,� clinoptilolite,� erionite,� mordenite� and�

phillipsite.�Another�zeolite�such�as�Ferrierite�occurs�in�a�few�large�beds,�thus�offering�a�limited�

range�of�atomic�structures�and�properties.�Each�of� the�seven�also�has�been�synthesized,�and�

those� synthetic� zeolites� have� a� wider� range� of� properties� and� larger� cavities� than� natural�

zeolites.�The�principal�synthetic�(aluminosilicate)�zeolites� in�commercial�use�are�Linde�Type�A�

(LTA),�Linde�Types�X�and�Y�(Al�rich�and�Si�rich),�Silicalite�1�and�ZSM�5,�and�Linde�Type�B�(zeolite�

P).�All�are�aluminosilicates�or�pure�silica�analogues.�

The�aluminosilicate�framework�of�a�zeolite�has�a�negative�charge,�which�is�balanced�by�

the�cations�located�in�the�cage�like�cavities�that�can�participate�in�ion�exchange�processes.�This�

characteristic�yields�some�important�properties�for�zeolites�such�as�less�dense�structures�than�

1.�Introduction��

26�

other�silicates.�In�this�sense,�between�20�and�50�percent�of�the�volume�of�a�zeolites�structure�

are�voids.��

Synthetic� zeolites� were� first� produced� in� the� 1950s� and� nowadays� more� than� 100�

different�zeolites�have�been�made�with�an�annual�production�of�synthetic�zeolites�exceeding�

12,000� tons.� The� International� Zeolite� Association� (IZA)� database� shows� that� the� number� of�

structural�types�of�unique�microporous�frameworks�has�been�growing�rapidly,�from�27�in�1970�

to� 133� in� 2001,� whereas� currently� this� number� has� reached� 180� [109].� In� table� 1.9� are�

presented�some�typical�oxide�formula�of�synthetic�zeolites.�

Table�1.9.�Typical�oxide�formula�of�some�synthetic�zeolites�

Zeolites Typical oxide formula Zeolites�A� Na2O.Al2O3.2SiO2.4,5H2O�Zeolites�N�A� (Na,�(CH3)4N+)2O.�Al2O3.4,8SiO2.7H2O��Zeolites�H� K2O.Al2O3.2SiO2.4H2O�Zeolites�L� (K2Na2)O.Al2O3.6SiO2.5H2O�Zeolites�X� Na2O.Al2O3.2,5SiO2.6H2O�Zeolites�Y� Na2O.Al2O3.4.8SiO2.8,9H2O�Zeolites�P� Na2O.Al2O3.2���5SiO2.5H2O�Zeolites�O� (Na2,K2,�(CH3)4N+2)O.Al2O3.7SiO2.3,5H2O��Zeolites��� (Na,�(CH3)4N+)2O.Al2O3.7SiO2.5H2O��

Zeolites�ZK�4� 0,85Na2O.0,15�((CH3)4N+)2O.�Al2O3.3,3SiO2.6H2O�Zeolites�ZK�5� (R,Na2)O.Al2O3.4�6SiO2.6H2O�

The�uses�of�zeolites�derive�from�their�special�properties:�

i)�Ion�exchange:�Zeolites�can�interact�with�water�to�absorb�or�release�ions.�In�this�sense,�

they�are�used�as�water�softeners,�to�remove�calcium�ions,�which�react�with�soap�to�form�scum.�

Zeolites� have� also� been� used� to� clean� radioactive�wastes,� in� this� sense,� radioactive� Sr90� and�

Cs137� have� been� removed� from� radioactive� waste� solutions� by� passing� them� through� tanks�

packed� with� the� natural� zeolite� clinoptilolite.� In� addition,� clinoptilolite� is� used� to� clean�

ammonium�ions�(NH4+)�from�sewage�and�agricultural�wastewater.�Natural�zeolites�are�also�the�

most� effective� filters� yet� found� for� absorbing� sulfur� dioxide� from�waste� gases.� As� efforts� to�

improve�the�continuous�air�quality,�zeolites�can�be�used�to�help�purify�the�gases�from�power�

plants�that�burn�high�sulfur�coal.��

ii)� Molecular� sieves:� Zeolites� can� selectively� absorb� ions� that� fit� the� cavities� in� their�

structures.� Industrial� applications�make� use� of� synthetic� zeolites� of� high� purity,� which� have�

larger� cavities� than� the� natural� zeolites.� These� larger� cavities� enable� synthetic� zeolites� to�

absorb�or�hold�molecules�that�the�natural�zeolites�do�not.�Some�zeolites�are�used�as�molecular�

sieves�to�remove�water�and�nitrogen�impurities�from�natural�gas.�

1.�Introduction�

27�

iii)�Catalytic�cracking:�Zeolites�can�hold�large�molecules�and�help�them�break�into�smaller�

pieces.� Because� of� their� ability� to� interact�with� organic�molecules,� zeolites� are� important� in�

refining�and�purifying�natural� gas�and�petroleum�chemicals.�The� zeolites�are�not�affected�by�

these�processes,� so� they� are� acting� as� catalysts.� Zeolites� are� used� to� help� break�down� large�

organic� molecules� found� in� petroleum� into� the� smaller� molecules� that� make� up� gasoline�

(cracking).�Zeolites�are�also�used�in�hydrogenating�vegetable�oils�and�in�many�other�industrial�

processes�involving�organic�compounds.��

Although�most�zeolites�used�as�catalysts�are�synthetic�and�made�for�specific�applications,�

a�few�natural�zeolites�have�also�been�employed.�Amongst�the�natural�zeolites,�the�most�used�is�

clinoptilolite� for� being� the� most� common� zeolite� occurring� in� large� quantities.� Moreover,�

taking�profit�of�their�high�exchange�capacity,�several�Fe�loaded�zeolites�have�been�also�widely�

employed.� In� this� regard,� Fe�bearing� zeolites�have�been�applied� to�N2O�decomposition� [110,�

111],�selective�catalytic�reduction�of�NO�with�hydrocarbons�[112,�113,�114]�or�NH3�[115,�116,�

117],�oxidation�of�benzene�to�phenol�[118,�119],�epoxidation�of�propene�[120,�121],�oxidation�

of�volatile�organic�carbons�[122],�decolorisation�by�means�of�Fenton�type�reaction�[123],�etc.��

1.9.2.�CLAYS�

Clays� form� almost� 70%� of� the� earth's� crust� and� are� defined� as� a� sedimentary� rock�

containing� mixtures� of� different� minerals,� mainly� hydrated� aluminum� silicate,� iron� or�

magnesium,� along� with� various� impurities,� particulate� extremely� small� crystal� in� varying�

proportions.��

The�crystal�structure�of�clays�consists�mainly�in�tetrahedral�silica�and�octahedral�alumina�

linked�together�to�form�layers�of�tetrahedra�and�octahedra.�These�layers�will�share�the�apical�

oxygen�from�the�tetrahedral�layer�with�the�free�oxygen�of�the�octahedral�layer.�A�layer�packing�

type�1:1�contains�one�tetrahedral�and�one�octahedral� layer,�a� layer�2:1�type�two�tetrahedral�

and�one�octahedral�and�a�layer�2:2�type,�two�layers�of�each�(Figure�1.7).�

Exchangeable�cation

(Al,�Mg,�Fe)O6

(Al,�Si)O42:1�silicate�

layer

2:1�silicate�layer

interlayer

Figure�1.7.�Schematic�structure�of�a�2:1�layer�expandable�clay�

1.�Introduction��

28�

The�octahedral�sites�are�usually�occupied�by�Al3+�or�Mg2+.�When�the� ion� is�Mg2+,�all� the�

holes� are� occupied� and� the� configuration� is� trioctahedral,� but� if� the� ion� is� Al3+,� due� to� their�

higher�charge,�only�2/3�of�the�sites�are�occupied,�resulting�in�a�dioctahedral�structure.�

The�Si4+�and�Al3+�in�the�tetrahedral�and�octahedral�layer�respectively,�may�be�substituted�

by� other� elements� with� an� ionic� radius� suitable� to� fit� into� the� structure� (called� isomorphic�

substitution).�Thus,�Si4+�can�be�replaced�by�Al3+,�and�Al3+�by�Mg2+,�Mn2+,�Ca2+�or�Ni2+�causing�a�

negative�charge�density�that�should�be�compensated�by�cations�in�the�interlaminar�space�that�

can�be�exchangeable�(cation�exchange).��

The� swelling� properties� are� reversible� unless� the� collapse� occurs� by� elimination� of� all�

polar� molecules� interspersed.� The� principal� advantage� of� these� materials,� apart� from� its�

availability,�is�that�due�to�laminar�structure,�force�a�chemical�reaction�occurs�in�a�plane�and�no�

three�dimensional�space,�making�it�much�faster.�

Commercial�clays�are�mainly�dedicated�to�the�manufacture�of�raw�materials�for�building�

materials�accounting�for�90%�of�production�and�only�10%�is�allocated�to�other�industries�such�

as� manufacture� of� paper,� rubber,� paints,� absorbent,� bleach,� molding� sand,� chemicals� and�

pharmaceuticals,�agriculture,�etc.[124].��

Compositionally,� clay� minerals� are� similar� to� zeolites.� Both� are� alumino�silicates� and�

hence,� they� possess� high� cation� exchange� capacity.� However,� they� differ� in� their� crystalline�

structure:�zeolites�have�a�rigid�three�dimensional�crystalline�structure�consisting�of�a�network�

of� interconnected�tunnels�and�cages�whilst�clays�have�a� layered�crystalline�structure�and�are�

subject�to�shrinking�and�swelling�as�water�is�absorbed�and�removed�between�the�layers.��

1.9.3.�SPONGES�

In� the� same� way,� Forager™� sponge� is� a� high� porosity� and� economic� ion�exchange�

material�with�selective�affinity�for�dissolved�heavy�metals� in�both�cationic�and�anionic�states.�

Such�material� is� able� to�promote�high� rates�of� adsorption�and� flexibility�which�enables� their�

compressibility� into�an�extremely�small�volume�to�facilitate�disposal�once�the�capacity�of�the�

material�has�been�exhausted�[125].�Forager�is�an�open�celled�cellulose�sponge�which�contains�

a�water�insoluble�polyamide�chelating�polymer� formed�by� the� reaction�of�polyethyleneimine�

and�nitrilotriacetic�acid.�This�material� is�claimed�to�contain� free�available�ethyleneamine�and�

iminodiacetate�groups�to�interact�with�heavy�metals�ions�by�chelation�and�ion�exchange.�In�this�

sense,� it�has� selective�affinity� for�dissolved�heavy�metals� in�both�cationic�and�anionic� states.�

Forager�sponge�and�other�adsorbent�sponges�have�been�successfully�used�in�the�treatment�of�

heavy�metals�solutions�[126,�127].��

1.�Introduction�

29�

Several�advantages�of�the�sponge�material�were�identified.�The�first�was�its�open�celled�

nature�that�allows�relatively�high�flow�rates;�the�second�was�cost�effectiveness;�and�the�third�

was� the�material's� low�affinity� for� sodium,�potassium,� and� calcium,� three� common�naturally�

occurring�groundwater� ions� that� can� interfere�with� the�effectiveness�of� typical� ion�exchange�

systems� for� treating� specific� priority� pollutant� metals.� The� selective� affinity� of� the� polymer�

enables�the�Forager™�Sponge�to�bind�toxic�heavy�metals�over�benign�monovalent�and�divalent�

cations� such� as� calcium,�magnesium,� potassium� and� sodium.� In� addition,� prior� studies� have�

shown�that�the�sponge�material� is�effective�over�a�wide�range�of�pH.�The�pH�at� the�site�was�

determined� to� range� from� 4� to� 5� standard� units.� Another� advantage� was� that� a� simple�

treatment� system� could� be� designed� and� installed� similar� to� a� typical� carbon� adsorption�

system.� It� is� an� open�celled� cellulose� housing� iminodiacetic� acid� groups� which� chelate�

transition� metal� cations� by� cation� exchange� processes� in� the� following� affinity� sequence:�

Cd2+>Cu2+>Hg2+>Pb2+>Au3+>Zn2+>Fe3+>Ni2+>Co2+>Al3+.�The�sponge�polymer�also�contains�tertiary�

amine� salt� groups� that� can� bind� anionic� contaminants,� such� as� the� chromate,� arsenic,� and�

uranium�oxide�species.�It�can�be�designed�for�site�specific�needs�to�contain�a�cation�that�forms�

a�highly� insoluble�solid�with�the�anion�of� interest.�Another�advantage�is� its�high�porosity�and�

flexibility�which�allows�its�compressibility�into�an�extremely�small�volume�to�facilitate�disposal.��

1.10.�THE�FENTON�REACTION�

Some� of� the� above� described� ion� exchange� materials,� can� be� employed� for� the�

treatment� of� a� wide� range� of� organic� compounds� detected� in� industrial� and� municipal�

wastewater.� Some� of� these� compounds� (both� synthetic� organic� chemicals� and� naturally�

occurring� substances)� pose� severe� problems� in� biological� treatment� systems� due� to� their�

resistance�to�biodegradation�or/and�toxic�effects�on�microbial�processes.�As�a�result,�the�use�of�

alternative� treatment� technologies,� aiming� to� mineralize� or� transform� refractory� molecules�

into�others�which�could�be� further�biodegraded,� is�a�matter�of�great� concern.�Among� them,�

advanced�oxidation�processes�(AOPs)�are�already�been�used�for�the�treatment�of�wastewater�

containing� recalcitrant� organic� compounds� such� as� pesticides,� surfactants,� dyes,�

pharmaceuticals� and�endocrine�disrupting� chemicals.�Moreover,� they�have�been� successfully�

used� as� pretreatment� methods� in� order� to� reduce� the� concentrations� of� toxic� organic�

compounds�that�inhibit�biological�wastewater�treatment�processes�[128]�

Advanced�oxidation�processes�(AOPs)�are�based�on�the�generation�of�the�highly�oxidative�

hydroxyl� radical� which� attacks� non�selectively� all� present� organic� compounds� [129].A� great�

number�of�methods�are�classified�under�the�broad�definition�of�AOPs�(Table�10).�Most�of�them�

1.�Introduction��

30�

use�a�combination�of�strong�oxidizing�agents�(e.g.�H2O2,�O3)�with�catalysts�(e.g.�transition�metal�

ions)�and�irradiation�(e.g.�ultraviolet,�visible).��

Table�1.10.�List�of�main�AOP�processes�

Ozonation�under�alkaline�conditions�(O3/OH�)�

Ozonation�assisted�by�hydrogen�peroxide�(O3/H2O2)�and�(O3/H2O2/OH

�)�Without�external�energy�supply�

Hydrogen�peroxide�and�iron�catalysts�(Fenton�process,�H2O2/Fe

2+)�Ozonation�and�ultraviolet�radiation�(O3/UV)�

Hydrogen�peroxide�and�ultraviolet�radiation�(H2O2/UV)�Ozone,�hydrogen�peroxide�and�ultraviolet�radiation�

(O3/H2O2/UV)�Photo�Fenton�(Fe2+/H2O2/UV)�

Ozonation�assisted�by�ultrasounds�(O3/US)�Hydrdogen�peroxide�assisted�by�ultrasounds�(H2O2/US)�

Electrochemical�oxidation�Anodic�oxidation�

HOMOGENEOUS�PROCESSES�

With�external�energy�supply�

Electro�Fenton�Catalytic�ozonation�(O3/Cat.)�

Photocatalytic�ozonation�(O3/TiO2/UV)�HETEROGENEOUS�

PROCESSES�Heterogeneous�photocatalysis�(H2O2/TiO2/UV)�

Among� AOPs,� the� Fenton� reaction� has� been� widely� applied� in� treating� contaminated�

wastewaters� containing� organic� volatile� compounds,� persistent� organic� pollutants� and� dyes�

[130,�131,�132].�The�Fenton�reaction�consists�on�the�generation�of� the�hydroxyl� radical� from�

hydrogen�peroxide�and�Fe(II)�ions�in�mild�conditions�(reactions�1�and�2):��

Fe2+�+�H2O2�� � �� Fe3+�+�OH��+��OH�� k1=10�7�M�1s�1� � (Eq.�1)�

�OH�+�Organic�matter���� Oxidized�Products�� � � � (Eq.�2)�Fe3+�+�H2O2�� � �� FeOOH2+�+�H+�� k3=0.001�0.01�M

�1s�1� � (Eq.�3)�FeOOH2+�� � ��� Fe2+�+��HO2�� � � � � (Eq.�4)�Fe2++�HO2�� � ��� Fe3+�+�HO2

��� � � � � (Eq.�5)�Fe3+�+�HO2�� � ��� Fe2+�+�O2�+�H

+�� � � � � (Eq.�6)�2�HO2�� � � �� H2O2�+�O2�� � � � � (Eq.�7)�

As� iron� is� catalytically� cycled� between� Fe(II)� and� Fe(III)� (reaction� 1� and� 3� to� 6),� the�

hydroxyl�radicals�can�be�generated�also�with�Fe(III),�which�is�called�Fenton�like�reaction�[133].�

The�reaction�using�Fe(III)� is�slower�than�with�Fe(II)�although�the�use�of�Fe(III)� instead�of�Fe(II)�

present�some�advantages�mainly�related�to�the�working�pH�that�can�be�broadened�from�3.0�to�

4.5.� [134]� Other� advantages� concern� with� the� reduction� on� the� reactants� costs� due� to� the�

lower�expenditure�of�Fe(III)�salts�compared�to�Fe(II)�salts.�However,�the�use�of�Fe(II/III)�salts�as�

a�homogeneous�catalyst�has�some�drawbacks�concerning�the�removal�of�Fe�due�to:�

� Chelating� pollutants� as�well� as� some� inorganic� components� of�wastewater� solutions�

(e.g.�phosphate)��

1.�Introduction�

31�

� Loss�of�the�catalyst�due�to� iron�hydroxide�precipitation�which�causes�red�mud�sludge�

that� should� be� removed� from� the� solution,� sometimes� requiring� further� treatment�

thus�increasing�the�cost�of�the�whole�process.��

To�overcome�these�drawbacks,�several�heterogeneous�Fenton�catalysts�bearing�Fe(II/III)�

ions,�clusters�or�oxides�have�been�developed�[135,�136,�137,�138,�139,�140,�141].�In�this�sense,�

several�layered�and�porous�aluminosilicates�such�as�clays�and�zeolites�have�been�proposed�as�a�

support�for�the�catalytic�Fe�due�to�their�high�specific�surface,�high�thermal�stability,�exchange�

capacity�and�homogeneous�distribution�of�active�sites�(142).�With�regard�to�zeolites,�Fe�loaded�

synthetic�commercial�zeolites�such�as�ZSM�5�zeolites�[143,�144,�145]�and�Y�zeolites�[146,�147,�

148]�have�been�widely� reported� in� the� literature� to�provide� similar� catalytic�activities�as� the�

homogeneous�catalysis.�However,�in�addition�to�the�elevated�cost�of�these�synthetic�materials�

and�the�preparation�of�these�Fe�loaded�materials,�long�and�tedious�procedures�are�required.��

1.11.�ARSENIC�SORPTION�USING�FE�LOADED�MATERIALS�

Arsenic� is� a� naturally� occurring� metal� released� into� the� environment� by� natural� and�

anthropogenic�(industrial�and�commercial)�processes.�It�has�received�huge�public�and�scientific�

attention�due�to�environmental�and�public�health�disasters�around�the�world�[149,�150,�151].��

1.11.1.�ARSENIC�TOXICITY�

Arsenic� compounds� can� be� classified� into� three� major� forms:� inorganic,� organic,� and�

arsine� gas.� Inorganic� arsenic� may� be� formed� with� either� trivalent� (arsenite)� or� pentavalent�

(arsenate)� arsenic.� Trivalent� arsenic� compounds� tend� to� be� more� toxic� than� pentavalent�

arsenic� compounds�although�pentavalent� species�predominate�and�are� stable� in�oxygen� rich�

aerobic�environments� [152].� Inorganic�arsenic� is�more� toxic� than� the�organic� forms�although�

very�high�doses�of� certain�organic� compounds�may�be�metabolized� to� inorganic� arsenic� and�

result� in� some� of� the� same� effects� derived� from� an� exposure� to� inorganic� compounds.�

Arsenobetaine,� an� organic� form� of� arsenic,� is� found� in� seafood� and� is� nontoxic.� On� the�

contrary,�arsine�gas�have�the�highest�toxicity�of�As�compounds�and�it�is�formed�by�the�reaction�

of�hydrogen�with�arsenic,�during�the�synthesis�of�organic�arsenic�compounds,�and�generated�

accidentally�during� the�smelting�and� refining�of�nonferrous�metals� in�mining�processes.�High�

levels�of�naturally�occurring�arsenic�are�found�in�soil�and�rocks�leading�to�unacceptable�levels�

of�arsenic�in�drinking�water�such�as�in�Bangladesh.��

The�toxicity�of�arsenic�varies�widely�based�on�the�route�of�exposure,�the�form,�the�dose,�

the�duration�of�exposure,�and�the� time�elapsed�since� the�exposure.� Ingestion�and� inhalation�

1.�Introduction��

32�

are�the�primary�routes�of�both�acute�and�chronic�exposures.�Arsine�gas�is�one�of�the�most�toxic�

forms�and�is�readily�absorbed�into�the�body�by�inhalation.��

Effects� of� acute� inorganic� arsenic� poisoning� include� fever,� anorexia,� hepatomegaly,�

melanosis,� cardiac� arrhythmia� and� eventual� cardiovascular� failure,� upper� respiratory� track�

symptoms,� peripheral� neuropathies,� gastrointestinal� and� hematopoietic� effects.� Dermal�

contact�with�high�concentrations�of�inorganic�arsenic�compounds�may�result�in�skin�irritation,�

redness,� and� swelling� and� high� acute� exposures� may� cause� cholera�like� gastrointestinal�

symptoms�of�vomiting�(often�times�bloody)�and�severe�diarrhoea�(often�bloody).�Ingestion�of�

large�doses�of�inorganic�arsenic�(70�to�180�mg)�may�be�fatal.�

Arsenic�has�been�classified�as�a�known�human�carcinogen�by�multiple�agencies�based�on�

the�increased�prevalence�of� lung�and�skin�cancer�observed�in�human�populations�exposed�to�

arsenic.��

Every�day,�lack�of�access�to�clean�water�and�sanitation�kills�thousands�of�people,�leaving�

others� with� reduced� quality� of� life� and� as� cities� and� slums� grow� at� increasing� rates,� the�

situation�worsens.�Nowadays�clean�water�is�a�scarce�resource�and�arsenic�removal�from�waters�

has�emerged�as�a�major�concern�in�certain�developing�countries.��

1.11.2.�ARSENIC�SORBENTS�

Several� types� of� adsorbents� have� been� used� for� the� removal� of� arsenic� from�aqueous�

effluents,� many� of� them� taking� advantage� of� Fe(III)� compounds� affinity� towards� inorganic�

arsenic� species.� In� this� regard,� various�methodologies� for�arsenic� removal� involve� the�use�of�

iron�hydroxyoxides�such�as�goethite� (either�natural�or�synthetic)� [153,�154,�155],� ferrihydrite�

[156,�157,�158]�or�hematite�[159,�160]�and�different�Fe�bearing�materials�such�as�Fe(III)�loaded�

zeolites� [161],�aluminosilicates� [162]�or� resins� [163].�To�predict� the� long�term�fate�of�arsenic�

and�design�new�materials�with�improved�capacity�and�efficiency�for�As�sorption,�the�molecular�

understanding�of� the� sorption�of�arsenic�by� iron� (oxy)hydroxides�and�Fe�bearing�materials� is�

required.�Fe�loaded�zeolites�have�been�successfully�employed�for�arsenic�removal�[164].�In�this�

sense,� arsenate� and� arsenite� adsorption� from� water� was� carried� on� by� using� iron� treated�

activated�carbon�and�natural�zeolite,�comparing�their�efficiency�with�the�results�obtained�using�

Faujasite� (13X)� and� Linde� type� A� (5A)� molecular� sieves.� Iron�treated� activated� carbon� and�

chabazite� were� promising� as� low�cost� arsenic� adsorbents� removing� approximately� 60%� of�

arsenate�and�arsenite�and�50%�of�arsenate�and�30%�of�arsenite,� respectively� [165].�Besides,�

aqueous�arsenic�sorption�by�natural�zeolites,�volcanic�stone,�cactaceous�powder�CACMM�and�

clinoptilolite�containing�rocks�with�different�clinoptilolite,�erionite�and�mordenite�percentages�

have�been�also�reported� [166,�167].�Each�zeolite�sample� in� the�0.1–4�mg/L�Fe�concentration�

1.�Introduction�

33�

range� removed� more� arsenate� than� arsenite� at� equivalent� arsenic� concentrations.� The�

saturation� capacity� of� the�materials�was� inversely� related� to� the� silicon�dioxide� content� and�

directly�to�the�iron�content�in�the�acid�washed�zeolite.�Moreover,�the�adsorption�of�As(V)�from�

drinking�water�by�an�aluminum�loaded�Shirasu�zeolite�(Al�SZP1)�was�studied�obtaining�results�

equivalent�to�that�of�activated�alumina.�A�ligand�exchange�mechanism�between�As(V)�ions�and�

surface�hydroxide�groups�on�Al�SZP1�was�presumed� [168].� Furthermore,�an� iron�conditioned�

zeolite� was� prepared� and� used� for� arsenic� removal� from� groundwater� at� pH� 7.8� and�

temperature�145�ºC�[169].�On�the�other�hand,�Forager�Sponge�and�other�adsorbent�sponges�

have�been�successfully�used�in�the�treatment�of�heavy�metal�solutions�[128,�170]�but�scarcely�

employed�for�the�arsenic�removal�after�being�loaded�with�iron�[171].��

Several� researchers� have� investigated� the� structure� of� the� As� adsorbed� onto� such�

materials� using� Extended� X�ray� Absorption� Fine� Structrue� (EXAFS)� and� IR� spectroscopic�

techniques�and�previous�studies�have�shown�that�As(V)�oxyanions�are�strongly�adsorbed�to�the�

surfaces�of�such� iron�oxides�as�goethite,� ferrihydrite,�and�hematite� [172,�173,�174,�175].�The�

assessment�and�characterization�of�these�Fe�bearing�materials�can�shed�light�on�the�sorption�

mechanisms� taking�part�on� these�materials� to�provide�useful�data�concerning� the�As�uptake�

mechanism.� Thus,� modifications� concerning� the� As� chemical� and� electronic� structures�

depending�on�the�type�of�adsorbent�can�be�of�general�interest.��

1.12.� WEAKNESSES� AND� NEEDS� OF� INDUSTRIALLY� CONTAMINATED�

WATERS�

� Weaknesses:

� Conventional� treatment� for� tailing� ponds� waters� consists� in� the� depuration� of� the�

water� in� water� treatment� plants� to� be� afterwards� discharged� to� nearby� rivers� or�

creeks.�However,�these�treatment�plants�are�expensive�and�the�process�costs�are�not�

recovered.��

� The�Fenton�reaction�involves�the�presence�of�iron�salts�as�catalyst�although�two�main�

drawbacks� arise,� the� first� mainly� concerning� the� loss� of� the� catalyst� due� to� iron�

hydroxide�precipitation�and�its�consequent�red�mud�sludge�generation.��

� Fe�loaded� materials� employed� as� Fenton� catalysts� involves� long� and� tedious�

procedures�in�addition�to�the�elevated�cost�of�the�synthetic�materials�used.��

� Several�iron�compounds�have�been�employed�for�the�sorption�of�arsenic.�Its�efficiency�

is�believed�to�be�strongly�influenced�by�its�structure.�

1.�Introduction��

34�

� Needs:�

� The� recovery� of� metals� contained� on� tailing� ponds� by� existing� methodologies� can�

provide�economical�value�to�these�residues�while�solving�an�environmental�problem.�

� New�iron�loaded�based�catalysts�into�different�materials�should�be�tested�to�avoid�the�

removal�of�the�catalyst,�the�generation�of�red�sludge�derived�from�the�precipitation�of�

iron�hydroxyoxides�and�to�improve�the�efficiency.��

� Nobel� Fe�loaded� materials� more� economical� should� be� tested� as� well� as� simpler�

preparation�methodologies�for�the�Fe�loading�of�these�materials.�

� To�predict� the� long�term� fate�of� arsenic� and� to�design�new�materials�with� improved�

capacity�and�efficiency�for�As�sorption,�the�molecular�understanding�of�the�sorption�of�

arsenic�by�iron�hydroxyoxides�and�Fe�bearing�materials�is�required.�

1.�Introduction�

35�

ANALYTICAL�TECHNIQUES�

Through� the� following� section,� the�general� aspects�of� the� techniques�employed� in� this�

work�are�overviewed.�In�this�sense,�the�analysis�of�heavy�metals�has�been�carried�out�by�means�

of�hand�held�X�ray�fluorescence�technique�while�synchrotron�based�techniques�such�as�X�ray�

Absorption�Near�Edge�Structure�(XANES)�and�Extended�X�ray�Absorption�Fine�Structure�(EXAFS)�

have�been�the�main�tools�applied�for�the�determination�of�mercury�species�as�well�as�to�the�

characterization�of�the�adsorption�of�arsenic�onto�Fe�loaded�materials.�

1.13.�X�RAY�FLUORESCENCE��

XRF� spectrometry� can� easily� and� quickly� identify� and� quantify� elements� over� a� wide�

dynamic�range,�from�ppm�levels�up�to�virtually�100%�w/w,�without�destroying�the�sample�and�

with� little,� if� any,� sample� preparation.� These� factors� lead� to� a� significant� reduction� in� the�

sample�analytical�cost�compared�to�other�elemental�analysis�techniques.��

1.13.1.�X�RAY�INTERACTION�WITH�MATTER�

Recording� the� image�of�a�given�structure�requires� the�use�of�a�wavelength�equal� to�or�

smaller� than� the� size� of� the� structure.� X�rays� are� actually� electromagnetic� waves� between�

ultraviolet�light�and�gamma�rays�on�the�wavelength�scale.�Their�wavelength�is�comparable�to�

interatomic�distances,�so�it�can�be�used�to�“see”�interatomic�distances�(Figure�1.8).��

Figure�1.8.�Electromagnetic�spectrum�

X�rays�interact�with�atoms�in�essentially�two�ways:�scattering�and�X�ray�absorption�or�the�

photoelectric�effect�(Figure�1.9).�Scattering�causes�the�photon�to�change�its�direction�and�it�can�

be� elastic� (Rayleigh� scattering)� or� inelastic� (Compton� scattering).� In� Rayleigh� scattering� the�

energy� of� the� photon� is� conserved� and� occurs� when� X�ray� photons� interact� with� strongly�

bound�electrons;�whereas�Compton�scattering�occurs�when�X�ray�photons�interact�with�weakly�

1.�Introduction��

36�

bound� electrons� and� the� energy� of� the� photon� is� conserved� after� the� interaction.� Rayleigh�

scattering�forms�the�basis�of�X�ray�diffraction�(Figure�1.10).�

MATERIAL

Incident X�ray beam Transmitted X�rays

Fluorescence

Rayleigh scattering

Compton scattering

Figure�1.9.�Interaction�of�X�rays�with�matter�

(1) Incoming X�ray photon

(2) Oscillating electron

(3) Scattered photon

No�loss of�energy

Nucleus

Electron

RAYLEIGH SCATTERING(Coherent scattering)

(1)(2)

(3) Ef=�E0

Energy E0

(1) Incoming X�ray photon

(2) Energy is partially

transferred to electron

(3) Scattered photon

Loss of�energy

Nucleus

Electron

COMPTON SCATTERING(Incoherent scattering)

(1)(2)

(3) Ef <�E0

Energy E0

Figure�1.10.�Rayleigh�and�Compton�scattering�

On�the�other�hand,�X�ray�absorption�occurs�when�an�atom�acquires�the�energy�of�an�X�

ray� to� excite� electrons� into� higher� energy� electron� orbitals� that� are� unoccupied,� or� into� the�

continuum�where�the�electron�is�no�longer�associated�with�the�atom.�To�fill�the�void�created�in�

the�inner�shell,�an�electron�from�a�higher�energy�shell�drop�down�almost�instantaneously.�The�

excess� energy� resulting� from� this� transition� can� be� released� either� in� the� form� of� an� X�ray�

photon�with�a�wavelength�characteristic�of�the�atom�(fluorescence)�or�to�an�electron�from�an�

outer�shell�that�receives�sufficient�energy�to�leave�the�atom�(Auger�electron�emission)�(Figure�

1.11).�

(1)Incoming X�ray photon

(2) A�K�shell electron is ejected (photoelectron)��

(3) Outer shell electron moves to the inner shell

hole created

(4) Energy excess emitted as�fluorescence

Nucleus

Electron

FLUORESCENCE

(1)

(2)(3)

Ef =�EL � EK

Energy E0

Ejected K�shellelectron

L�shell

K�shell

L�� K�transition

(4)

(1)Incoming X�ray photon

(2) A�K�shell electron is ejected (photoelectron)��

(3) Outer shell electron moves to the inner shell hole

created

(4) Energy excess is transferred to electron

(5) Electron ejected from atom (Auger electron)

Nucleus Electron

AUGER ELECTRON

(1)

(2)(3)

Energy E0

Ejected K�shellelectron

L�shell

K�shell

L�� K�transition

(4)

(5)

1.�Introduction�

37�

Figure�1.11.�Release�of�energy�process�after�X�ray�absorption�by�matter�

All�elements�emit�X�rays�at� their�own�characteristic�energies.�These�X�rays�are�called�K�

lines�if�they�result�from�an�electron�filling�the�K�shell,�and�L�lines�if�they�result�from�filling�the�

next�electron�shell�out,� the�L�shell.�The�energy�of� the�emitted�fluorescent�X�rays� identify� the�

elements� present� in� the� sample� and,� in� general,� the� intensities� of� the� X�ray� lines� are�

proportional� to� the� concentration� of� the� elements� in� the� sample,� allowing� quantitative�

chemical�analysis�by�X�ray�Fluorescence�(XRF)�spectrometers.�

1.13.2.�X�RAY�FLUORESCENCE�

XRF� is� a� non�destructive,� simultaneous� multi�element� technique� that� covers� a� wide�

dynamic�range�from�100%�down�to�the�μg/g� level�with�typical�relative�precision�approaching�

1%� [176].� This� well�established� analytical� method� has� been� applied� to� environmental,�

geological,�archaeological,�metal�and�alloy�samples.��

In�addition,� in�the�last�40�years�hand�held�XRF�equipments�have�been�emerged�as�very�

profitable� tool� given� that� the� application� of� such� technique,� let� to� quickly� delineate�metals�

contamination� at� a� screening� level� in� situ� [177],� as� well� as� to� determine� contamination�

patterns.�With�minimal�sample�preparation�requirements,�XRF�may�provide�quick�qualitative,�

semi�quantitative,�or�even�quantitative�analysis�of� liquids,�powder,�solid�or� thin� film�samples�

[178].� In� addition,� high� volume� of� field� test� can� be� monitored� to� determine� the� spatial�

distribution�and�degree�of�heterogeneity�of�heavy�metals�in�an�undisturbed�position�while�off�

site�analytical�costs�are�minimized�without�destruction�of�the�samples�[179,�180].��

1.13.4.�FIELD�PORTABLE�XRF�INSTRUMENTATION�

A� typical� XRF� system� has� three� major� components:� an� excitation� source,� a�

spectrometer/detector� and� a� data� collection/processing� unit.� In� Field�Portable� X�Ray�

Fluorescence�(FP�XRF)�equipments,�the�excitation�source�and�the�detector�device�are�generally�

assembled� in� a� fixed� position� in� order� to� reduce� the� size� and� weight� to� facilitate�

transportation.�The�instrument�device�may�be�very�compact�by�incorporation�of�an�embedded�

microcomputer�or� it�may�be�rendered�more�flexible�by�using�a�standard�notebook�computer�

(Figure�1.12).��

1.�Introduction��

38�

Figure�1.12.�Scheme�of�a�FP�XRF�

Various� excitation� sources� may� be� used� to� irradiate� a� sample� although� the� more�

employed� are� radioisotopes� sources� and� X�ray� tubes.� In� a� radioisotope� source,� the�

characteristic�X�rays�emitted�from�a�sealed�radioisotope�source�such�as�55Fe,�57Co,�109Cd,�241Am�

and�244Cm�are�employed.�However,�the�intensity�of�these�excitation�sources�gradually�falls�as�

the�isotope�decays�and�the�emitted�X�ray�wavelengths�and�intensities�are�not�adjustable.�On�

the�other�hand,�X�ray�tubes�have�increased�sensitivity�and�analytical�range.�X�ray�tube�offer�a�

faster�analytical� time�because�the�X�ray� flux�can�be�higher� than�most� isotope�based�sources.�

They� can� also� be� used� over� a� wider� range� of� excitation� energies,� eliminating� the� need� for�

multiple�isotopic�sources�to�produce�X�rays�over�the�entire�excitations�spectrum.�In�addition,�it�

is�worth�mentioning�that�transportation�of�miniature�X�ray�tubes�involves�less�problems�than�

when�traveling�with�radioisotope�sources.�The�cathode�in�the�miniature�X�ray�tube�is�heated�by�

a� filament,� and� it� then� emits� electrons� that� are� accelerated� by� a� high� electric� field.� The�

accelerated� electrons� hit� the� anode,� which� emits� an� X�ray� continuum� accompanied� by� the�

characteristic�lines�of�anodic�metal.�Depending�upon�the�application,�the�anode�material�may�

be�Cd,�Cu,�Mo,�Rh,�Ag,�W,�Pt�or�Au.� These� sources�are�powered�with�an�external�AC�power�

supply,�or�an�internal�rechargeable�battery�[181].�

There�are�several�detectors�available�such�as�gas�flow�proportional�counters,�scintillation�

counters� and� solid� state� detectors� being� the� latter� the� most� employed� given� their� high�

resolution.� Solid�state� detectors� have� improved� energy� resolution� dramatically,� thereby�

reducing� spectral� interferences� and� offering� a� three�� to� fourfold� speed� advantage� over� a�

scintillation� detector.� Various� types� of� solid� state� detectors� exist� such� as� Germanium,� Si(Li)�

(lithium�drifted�silicon),�Si�PIN�(silicon�positive�intrinsic�negative),�CCD�(charge�coupled�device),�

PDA� (photo� diode� array),� PIPS� (passivated� implanted� planar� silicon)� and� SSB� (silicon� surface�

1.�Introduction�

39�

barrier).�The�semiconductor�detectors�typically�require�cryogenic�cooling�to�improve�the�signal�

to�noise�ratio.�Besides,�the�development�of�personal�computers�with�high�speed�and�memory�

has� also� allowed� fundamental� parameter� algorithms� to� be� quickly� performed�using�multiple�

standards,� resulting� in� rapid� and� more� accurate� standardization� and� analyses� for�

multicomponent,�complex�matrices�over�standard�empirical�methods�[182].��

1.14.�SYNCHROTRON�BASED�TECHNIQUES�

1.14.1.�SYNCHROTRON�LIGHT�SOURCES�

The�first�accelerators�(cyclotrons)�were�built�by�particle�physicists�in�the�1930’s�to�study�

collisions�between�high�energy�particles.�In�this�role�they�were�very�successful,�and�the�Large�

Hadron� Collider� at� CERN� is� based� on� this� technology.� But� scientist� soon� noticed� that� these�

machines�also�had�a�byproduct:� they�generated�very�bright� light.�The�emitted� light�was� first�

considered� an� inconvenient� because� it� caused� the� particles� to� lose� energy.� The� first�

experiments�carried�out�using�synchrotron�light�were�performed�at�Cornell�(USA)�in�1956�and�

over�the�years,�the�number�of�experiments�increased,�all�using�machines�built�for�high�energy�

particle�physics.�This�changed�in�1980�when�the�UK�built�the�world’s�first�synchrotron�specially�

devoted� to� produce� synchrotron� light� for� experiments.� Nowadays,� there� are� around� 70�

synchrotron� light� sources� around� the�world,� carrying� out� a� huge� range� of� experiments�with�

applications� in� engineering,� biology,� materials� science,� cultural� heritage,� chemistry,�

environmental�sciences�and�many�more.��

1.14.2.�DESIGN�AND�OPERATION�OF�A�SYNCHROTRON�LIGHT�SOURCE��

A� schematic� overview� of� a� synchrotron� facility� is� depicted� in� Figure� 1.13.� Bunches� of�

elementary� particles� such� as� electrons� or� positrons� are� initially� accelerated� by� a� linear�

accelerator� (LINAC)� and� then� accelerated� further� in� a� booster� ring� that� injects� the� particles�

traveling�near�the�speed�of� light� into�a�storage�ring.�The�particles�within�the�storage�ring�are�

forced�to�change� its� trajectory�by�bending�magnets� so� that� they� travel� in�a�closed� loop.�This�

causes�X�rays�with�a�broad�spectrum�of�energies�(white�light)�to�be�emitted�tangential�to�the�

storage� ring.� Therefore,� a� synchrotron� storage� ring� is� an� N�sided� polygon,� where� N� is� the�

number�of�bends.��

1.�Introduction��

40�

Experimental station

X-raybeam line

Electronbeam

LINACLINACBooster ring

Electronbeam

Electrongun

Insertiondevice

Bendingmagnet

Storage ring

Figure�1.13.�Synchrotron�schematic�overview�

Wigglers�and�undulators�are�two�types�of�specialized�insertion�devices�that�are�placed�in�

the�straight�sections�of�the�storage�ring.�A�wiggler�consists�of�several�closely�spaced�bending�

magnets� that� increase� the� intensity� of� the� X�ray� pulse.� An� undulator� oscillates� the� charged�

particles�using�carefully�spaced�magnets�such�that�the�interference�between�their�poles�affects�

the�emitted�X�ray�spectrum.�This�interference�is�additive�at�particular�wavelengths,�producing�

an�intense�X�ray�beam�at�a�wavelength�that�can�be�selected�by�varying�the�gap�between�the�

poles�of�the�magnets�(Figure�1.14).�Beamlines�are�placed�tangential�to�the�storage�ring�to�use�

the�X�rays�emitted�by�bending�the�charged�particles�[183].��

Wiggler

Undulator

Free electron laser

Bending magnet

Electron beam X-ray radiation Magnetic structures

Wiggler

Undulator

Free electron laser

Bending magnet

Electron beam X-ray radiation Magnetic structures

Figure�1.14.�Bending�magnets�and�insertion�devices�

1.�Introduction�

41�

Synchrotron�light�presents�very�special�characteristics:�

� High� intensity� or� flux� (photons� per� second)� over� a� continuous�wavelength� spectrum�

from� microwaves� to� hard� X�rays� and� gamma� radiation.� In� contrast� to� laser� light,�

synchrotron�radiation�is�non�monochromatic.

� High�brightness,�thousands�of�million�fold�higher�than�conventional�X�ray�sources.�

� Linearly� polarized� light,� the� light� oscillates� only� within� certain� planes.� The� light� is�

emitted� in� very� short� (picoseconds)� pulses�with� a� periodic� structure� (microseconds),�

therefore�showing�a�high�potential�for�studies�of�transient�phenomena.�

� Light� source� remaining� stable� along� the� time.�Depending�on� the� facility,� each�bunch�

refill�shows�a�lifetime�between�4�and�24�hours.��

Despite� the� strong� potential� shown� by� synchrotron�based� techniques� and� the�

spectacular�increase�of�their�possible�uses,�these�techniques�present�as�well�some�drawbacks:�

� Poor�detection�limits�

� Limited�number�of�synchrotron�facilities�

� Complex�data�treatment�

1.14.3.�X�RAY�ABSORPTION�SPECTROMETRY�

The�general� aspects�of�X�Ray�Spectroscopy� (XAS)�have�been�presented� in�a�number�of�

reviews�papers�and�books� [184,�185],�as�well�as� its�applications� to�soils,�minerals,�and�other�

geochemical� matrices� so� the� lector� is� addressed� to� the� extense� literature� available� on� XAS�

applications,� synchrotron� facilities,� and� specialized� techniques� involving� synchrotron� X�rays�

[186,�187,�188,�189].�Also,�a�number�of�reviews�papers�and�book�sections�describe�techniques�

and�applications�of�XAS� in�geochemistry�and�soil� science� [190,�191,�192,�193,�194,�195].�The�

principles�of�XAS�and�data�analysis�have�been�also�widely�described�[196,�197,�198]�while�more�

details� on� the� physics� of� XAS� appear� in� several� books� [199,� 200,� 201,� 202].� Therefore,� only�

some� basics� on� this� spectroscopy� will� be� presented� here.� The� reader� is� referred� to� the�

abovementioned�reviews�for�more�detailed�information.��

As� aforementioned,� when� the� X�rays� interact� with�matter� the� radiation� can� be� either�

scattered�by�the�electrons�or�absorbed�and�excite�the�electrons�(Figure�1.9).�When�the�energy�

of�the�incident�photons�is�sufficient�enough,�a�core�electron�of�the�absorbing�atom�is�excited�to�

a�continuum�state�(i.e.�producing�a�photoelectron)�causing�a�wave�that�is�backscattered�by�the�

neighboring�atoms�producing�the�characteristic�features�of�a�transmission�XAS�spectra.�To�fill�

the� created� vacancy,� an� electron� from� a� higher� shell� drops� emitting� fluorescence� of�

characteristic� wavelength.� Several� detection� setups� have� been� developed� for� XAS� studies,�

1.�Introduction��

42�

depending� on� the� nature� of� the� absorber� and� the� matrix� type.� In� this� sense,� the� most�

commonly�used�involve�the�measurement�of�either�the�transmission�of�X�rays�or�the�emitted�

fluorescence.��

The� absorption� spectrum� of� an� element� in� the� vicinity� of� an� absorption� edge� can� be�

divided�in�four�main�regions:�Pre�edge,�Edge�(or�white�line),�XANES�and�EXAFS.�(Figure�1.15).�

�Figure�1.15.�Typical�X�ray�absorption�spectrum.��

1)�Pre�edge:�E�~2�50�eV�below�the�main�absorption�edge.�In�this�region�there�is�no�significant�

absorption� phenomena,� only� localized� electronic� transitions� to� unfilled� (or� partially� filled)�

atomic�levels�(e.g.,�s��p,�or�p�d).��

2)�Edge� (White� line):� E� from� ~� 2� eV� below� to� ~� 2� eV� above� the� absorption� edge.� Electronic�

transitions� occur�with� high�probability� from� the� core� level� to� unoccupied�bound� states�with�

close�energy�or�continuum�states.�A�sudden�rise�of�absorption�is�observed.��

3)� XANES:� E� from� ~2� to� 50� eV� above� the� edge.� Low�energy� photoelectrons� are� strongly�

scattered�and�multiple�scattering�dominates�[203].�The�resulting�features�are�intense,�and�can�

be� interpreted� in� terms� of� multiple� scattering� from� atoms� in� the� first� coordination� shells�

around�the�absorber,�yielding� information�about� interatomic�distances�and�angles� (Figure�9).�

Given� the� complexity� of� the� theoretical� approach� to� phenomena� occurring� in� the� XANES�

region,� speciation� concept� in� XANES� is� usually� based� on� the� comparison� of� an� unknown�

spectrum�with� a� database� of� reference� spectra.� The� fitting� process� looks� for� the� best� linear�

combination�of�reference�spectra�able�to�appropriately�reproduce�the�unknown�spectrum.�In�

these�terms,�XANES�has�been�widely�employed�as�a�speciation�technique.��

4)�EXAFS:� This� region� lies� from�~�50� to�~�1000�eV�above� the�edge.� In� the�EXAFS� region,� the�

photoelectrons� have� high� kinetic� energy� and� normally� dominates� single� scattering� by� the�

nearest� neighboring� atoms.� In� the� EXAFS� region,� the�most� important� feature� is� oscillations.�

1.�Introduction�

43�

When�a�photoelectron�interacts�with�its�neighboring�atoms,�it�will�be�scattered�(Figure�16).�In�

the� XANES,� multiple� scattering� patterns� will� be� dominant� whilst� in� the� EXAFS� region� single�

scattering� pattern� would� be� the� main� pattern.� The� interactions� among� the� scattering� and�

backscattering�photoelectron�waves�produce�the�EXAFS�oscillations�(Figure�16).�EXAFS�region�

can� be� analyzed� to� obtain� information� about� the� distance� between� the� absorber� and� the�

neighboring�atoms,�extending�out�to�several�shells�of�ligands.��

Backscatter photoelectron

Photoelectron

Oscillation

Atom with neighbor

E�

Isolated atom

Single�scatteringTwo�legs Multiple scattering

Three�legsPhotoelectron

Backscatterer

Figure�1.16.�Origin�of�fine�structure�of�EXAFS�

The� number� and� type� of� backscatterers� can� be� also� assessed� through� the� analysis� of�

EXAFS� region.� The� frequency� of� EXAFS� oscillations� is� inversely� related� to� average� absorber�

backscatterer�distance,�and�the�amplitude�of�the�oscillations�is�directly�related�to�the�number�

of�backscattering�ligands.��

Regardless�of�the�complexity�of�the�sample,�the�XAS�signal�comes�from�all�of�the�atoms�

of�a�single�element�as�selected�by�the�X�ray�energy.�The�structural�information�obtained�from�

XAS� is� useful� for� identifying� the� chemical� speciation� of� an� element,� including� mineral,�

noncrystalline�solid�or�adsorbed�phases.��

In�this�regard,�XAS�techniques�have�been�shown�to�provide�reliable� information�on�the�

speciation�of�several�elements�being�especially�interesting�the�case�of�mercury,�the�toxicity�of�

which� strongly� depends� on� its� speciation.� In� this� sense,� several� studies� dealt� with� mercury�

speciation�without�requiring�sample�pretreatment�[204,�205,�206,�207].�Moreover,�among�XAS�

techniques,� both� EXAFS� (extended� X�ray� absorption� fine� structure)� and� XANES� (X�ray�

absorption� near�edge)� spectroscopies� have� been� used� for� the� speciation� of� mercury� in�

different�matrices,� such� as�mine� ores� and� wastes� [206,� 208],� fish� [209],� contaminated� soils�

[210]� and�hyacinths� [211],� and� in� studies�of� interactions�between�mercury�and� soil�minerals�

[212].�

1.�Introduction��

44�

1.15.�OBJECTIVES�

This�PhD�thesis�has�been�focused�on�two�main�objectives�both�regarding�the�application�

to�environmental�problems� concerning:� i)� contaminated� soils� surrounding�mine�areas�and� ii)�

industrial�contaminated�waters.��

In�this�sense,�more�specific�goals�of�the�present�thesis�are;�

� The� application� of� different� analytical� techniques� such� as� Field�Portable� X�Ray�

Fluorescence� and� X�Ray� Absorption� Spectroscopy� to� the� study� of� highly� impacted�

environments�focused�on:�

� the� characterization� of� soils� surrounding� four� different� mine� areas� from�

Marrakech.� Study� of� heavy� metal� distribution� throughout� abandoned� mine�

areas�and�assessment�of�heavy�metal�mobility.��

� the� study� of� mercury� speciation� through� synchrotron� techniques� and�

estimation�of� its�mobility� from� soil� samples� from� three�of� the�main�mercury�

mine�areas�in�Europe.�

� The� study� of� a� process� at� laboratory� and� pilot� plant� scale� to� recover� Zn� from�a� real�

mine� effluent� in�order� to�produce�an� economically� effective�output�while� solving� an�

environmental�problem.��

� To�assess�the�feasibility�of�novel�Fe�loaded�materials�

� as� catalysts� to� degrade� different� organic� pollutants� by� means� of� Fenton�

reaction.��

� as� arsenic� sorbents.� In� this� regard,� the� know�how� gained� on� synchrotron�

techniques� can� be� applied� to� shed� light� onto� the� sorption� mechanisms� of�

arsenic�onto�these�materials.�

1.�Introduction�

45�

1.16.�REFERENCES�

[1]� Hartman,� H.L.� (1992).� SME� Mining� Engineering� Handbook.� Vol.1,� 2nd� ed.� Society� for� Mining,�Metallurgy,�and�Exploration,�Inc.�USA.�

[2]�Hartman,�H.L.;�Mutmansky,�J.M.�(2002).�Introductory�mining�engineering.�John�Wiley�&�Sons,�USA.�

[3]�Mbendi�World�Mining� Overview,� http://www.mbendi.co.za/indy/ming/p0005.htm� (Last� visit:� April�2011).�

[4]�Wilmoth,�R.�C.;�Hubbard,�S.�J.;�Burckle,�J.O.;�Martin�J.�F.�Production�and�processing�of�metals:�their�disposal�and�future�risks.� In:�Merian�E.,�(ed.)�(1991).�Metals�and�their�compounds�in�the�environment.�Occurrence,�analysis�and�biological�relevance.�19�65.�VCH.�USA.�

[5]�Adriano,�D.�C.�(2001).�Trace�Elements�in�Terrestrial�Environments;�2nd�ed,�Springer�Verlag.�USA.�

[6]� Kertes,� N.� (2007).� US� abandoned� mine� count� still� a� mystery.� General� Accounting� Office� report.�American�Metal�Market.�USA.�

[7]� Martin,� T.E.;� Davies,� M.P.� (2000).� Trends� in� the� stewardship� of� tailings� dams.� AGRA� Earth� &�Environmental�Ltd.�Canada.�

[8]� Rico,�M.;� Benito,� G.;� Salgueiro,� A.R.;� Diez�Herrero,� A.;� Pereita,� H.G.� (2000).� Reported� tailings� dam�failures.�A�review�of�the�European�incidents�in�the�worldwide�context.�J.�Hazard.�Mater.�152�(2)�846�852.�

[9]�Grimalt,� J.O.;�Ferrer,�M.;�Macpherson,�E.� (1999).�The�mine�tailing�accident� in�Aznalcollar.�Sci.�Total�Environ.�242:�3�11.�

[10]�Osan,�J.;�Kurunczi,�S.;�Török,�S.;�Van�Grieken,�R.�(2002).�X�Ray�analysis�of�riverbank�sediment�of�the�Tisza� (Hungary):� identification� of� particles� from� a�mine� pollution� event.� Spectrochim.� Acta� Part� B� 57:�413��422.�

[11]� Kibblewhite,� M.� (2001)� Identifying� and� dealing� with� contaminated� land,� In:� Hester,� R.E.� and�Harrison,�R.M.�(Eds),�Assessment�and�Reclamation�of�Contaminated�land�16,�Royal�Society�of�Chemistry�(RSC),�England.�

[12]�El�Gharmali,�A.;�Rada,�A.;�El�Adnani,�M.;�Tahlil,�N.;�El�Meray,�M.;�Nejmeddine,�A.�(2004).�Impact�of�acid� mining� drainage� on� the� quality� of� superficial� waters� and� sediments� in� the� Marrakesh� region,�Morocco.�Environmental�Technology.25:�1431�1442.�

[13]�Hakkou,�R.;�Benzaazoua,�M.;�Bussière,�B.�(2008).�Acid�mine�drainage�at�the�abandoned�Kettara�mine�(Morocco):�1.�Environmental�characterization.�Mine�Water�Environ.�27:�145�159.�

[14]�Hakkou,�R.;�Benzaazoua,�M.;�Bussière,�B.�(2008).�Acid�mine�drainage�at�the�abandoned�Kettara�mine�(Morocco):�2.�Mine�waste�geochemical�behavior.�Mine�Water�Environ�27:�160�170.�

[15]� El� Adnani,� M.;� Rodriguez�Maroto,� J.M.;� Sbai,� M.L.,� Loukili,� I.L.;Nejmeddine,� A.� (2007).� Impact� of�polymetallic�mine�(Zn,�Pb,�Cu)�residues�on�surface�water,�sediments�and�soils�at�the�vicinity�(Marrakech,�Morocco).�Environ.�Technol.�28:�969�85.�

[16]�Thompson,�K.C.;�Nathanail,�C.P.� (Eds.)� (2003).�Chemical�analysis�of�contaminated� land.�CRC�Press,�USA.�

[17]� Gillman,� G.P;� Bell,� L.C.� (1978).� Soil� solution� studies� on� weathered� soils� from� tropical� North�Queensland.�Aust�J�Soil�Res,�16:�67�77.�

[18]�APHA�Part�4110.� (1998).�Determination�of� ions�by� ion�chromatography.� In�Standard�Methods� for�the�Examination�of�water�and�wastewater,�16th�Ed.�American�Public�Health�Association,�USA.�

[19]�Bengtsson,� L.;�Enell,�M.� (1986).�Chemical�analysis,� in:�Berglund,�B.E.� (Ed.),�Handbook�of�Holocene�Palaeoecology�and�Palaeohydrology.�John�Wiley�&�Sons�Ldt.,�Chichester,�England,�423�451.�

1.�Introduction��

46�

[20]� Matthiesen,� M.K.;� Larney,� F.J.;� Selinger,� L.B.;� Olson,� A.F.� (2005).� Influence� of� Loss�on�Ignition�Temperature�and�Heating�time�on�Ash�content�of�compost�and�manure.�Comm�Soil�Sci�Plant�Anal,�36:.�2561�2573.��

[21]�Kennedy,�I.R.�(1992).�Acid�soil�and�acid�rain,�2nd�ed.�Research�Studies�Press,�England.�

[22]�Tatzber,�M.;�Stemmer,�M.;�Spiegel,�H.;�Katzlberger,�C.;�Haberhauer,�G.;�Gerzabek,�M.H.�(2007).�An�alternative�method�to�measure�carbonate�in�soils�by�FT�IR�spectroscopy.�Environ.�Chem�Lett:�9�12.�

[23]�Prasad,�H.� (1999).�Heavy�metals� stress� in�plants.�From�molecules� to�ecosystems.� Springer�Verlag,�Germany.�

[24]�Kalnicky,�D.J.;� Singhvi�R.� (2001).� Field�portable�XRF�analysis�of� environmental� samples.� J.�Hazard.�Mater.�83:�93�122.�

[25]� Kilbride,� C.;� Poole,� J.;�Hutchings,� T.R.� (2006).�A� comparison� of� Cu,� Pb,� As,� Cd,� Zn,� Fe,�Ni� and�Mn�determined�by�acid�extraction/ICP–OES�and�ex�situ�field�portable�X�ray�fluorescence�analyses.�Environ.�Pollut.�143:�16�20.�

[26]� Potts,� P.J.;� Ellis,� A.T.;� Kregsamer,� P.;� Streli,� C.;� Vanhoof,� C.;� West,� M.;� Wobrauschek,� P.� (2005).�Atomic�spectrometry�update.�X�Ray�fluorescence�spectrometry.�J.�Anal.�At.�Spectrom.�20:�1124–1154.�

[27]�Melquiades,�F.L.;�Appoloni,�C.R.�(2004).�Application�of�XRF�and�field�portable�XRF�for�environmental�analysis.�J.�Radioanal.�Nucl.�Chem.�262(2):�533�541.��

[28]� Van� Straalen,� N.M.� (2002).� Assessment� of� soil� contamination� –� a� functional� perspective.�Biodegradation�13(1):�41�52.�

[29]�The�Ministry�of�Housing,�Spatial�Planning�and�the�Environment.�Environmental�Quality�Standards�in�the�Netherlands�(1999).�A�review�of�environmental�quality�standards�and�their�policy�framework�on�the�Netherlands.�Netherlands..�

[30]� Darmendrail,� D.� (2001).� The� French� approach� to� contaminated�land� management.�http://www.sanaterre.com/guidelines/pdf/Darmendrail_2001.pdf�(Last�visit:�May�2011).�

[31]�Directive�2004/35/CE�of�the�European�parliament�and�of�the�council�of�21�April�2004.��

http://eur�lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2004:143:0056:0075:EN:PDF� (Last� visit:�May�2011)�

[32]� Kersten�M.,� Förstner� U.� (1989).� Speciation� of� trace� elements� in� sediments.� In:� Batley,� G.E.(Ed.),�Trace�Element�Speciation:�Analytical�Methods�and�Problems.�CRC�Press,�USA.�

[33]� Pickering�W.F.� (1995).� I.2.�General� strategies� for� speciation.� In:�Ure,� A.M.;�Davidson,� C.M.� (Eds.),�Chemical�Speciation�in�the�Environment,�Blackie�Academic�and�Professional,�UK.�

[34]�Valors�dels�NGR�per�metalls� i�metal�loides� i�protecció�de� la� salut�humana�aplicables�a�Catalunya.�http://www20.gencat.cat/portal/site/arc/menuitem.60fb2478680e61fd624a1d25b0c0e1a0/?vgnextoid=fd370431b17d6210VgnVCM1000008d0c1e0aRCRD&vgnextchannel=fd370431b17d6210VgnVCM1000008d0c1e0aRCRD&vgnextfmt=default�(Last�visit:�June�2011)�

[35]�Templeton,�D.M.;�Ariese,�F.;�Cornelis,�R.;�Danielsson,�L�G.;�Muntau,�H.;�Van�Leeuwen,�H.P.;�Lobinski,�R.� (2000).� Guidelines� for� terms� related� to� chemical� speciation� and� fractionation� of� trace� elements.�Definitions,� structural� aspects,� and�methodological� approaches� (IUPAC�Recommendations�2000).� Pure�Appl.�Chem.�72:�1453�1470.��

[36]�Ure,�A.M.�(1990).�Methods�of�analysis�for�heavy�metals�in�soils,�in:�Alloway,�B.J.�(Ed.),�Heavy�metals�in�soils,�40�73.�Blackie�&�Son,�UK.�

[37]�Tack,�F.M.;�Verloo,�M.G.�(1995).�Chemical�speciation�and�fractionation�in�soil�and�sediment�heavy�metal�analysis:�a�review.�Int.�J.�Environ.�Analyt.�Chem.59:�225�238.�

[38]� Fitfield,� F.W.;� Haines,� P.J.� (1997).� Environmental� analytical� chemistry.� Blackie� Academia� &�Professional,�UK.�

1.�Introduction�

47�

[39]� Quevauviller,� P.� (1998).� Method� performance� studies� for� speciation� analysis.� Royal� Society� of�Chemistry,�UK.��

[40]�Tessier,�A.;�Campbell,�P.G.C.;�Bisson,�M.�(1979).�Sequential�extraction�procedure�for�the�speciation�of�particulate�trace�metals.�Anal.�Chem.�51:�844�851.�

[41]� Quevauviller,� P.;� Van� der� Sloot,� H.� A.;� Ure� A.;� Muntau� H.;� Gomez� A.;� Rauret,� G.� (1996).�Harmonization�of�leaching/extraction�tests�for�environmental�risk�assessment.�Sci.�Total.�Environ.�178(1�3):133�139.�

[42]�Kennedy,�V.�H.;�Sanchez,�A.�L.;�Oughton,�D.�H.;�Rowland,�A.�P.�(1997).�Use�of�single�and�sequential�Chemical�Extractants� to�Assess�Radionuclide�and�Heavy�Metal�Availability� from�Soils� for�Root�Uptake.�Analyst.�1997,�122,�89�100.�

[43]�Sahuquillo,�A.;�Rigol,�A.;�Rauret,�G.�(2003).�Overview�of�the�use�of�leaching/extraction�tests�for�risk�assessment�of�trace�metals�in�contaminated�soils�and�sediments.�Trends�in�Anal.�Chem.�22(3):�152�159.�

[44]�Rauret,�G.�(1998).�Extraction�procedures�for�the�determination�of�heavy�metals�in�contaminated�soil�and�sediment.�Talanta�46�(3):�449�455.�

[45]� Quevauviller,� P.;� Rauret,� G.;� Lopez�Sanchez,� J.F.;� Rubio,� R.;� Ure,� A.M.;� Muntau,� H.� (1997).� The�Science�of�the�Total�Environment:�205,�223�234.��

[46]�Quevauviller,� P.;� Rauret,�G.;� Lopez�Sanchez,� J.F.;� Rubio,� R.;�Ure,�A.M.;�Muntau,�H.� (1997).� Report�EUR�17554�EN,�European�Comission,�Belgium.�

[47]�Gomez�Ariza,� J.L.;�Giraldez,� I.;� Sánchez�Rodas,�D.;�Morales,� E.� (2000).�Metal� sequential� extraction�procedure�optimized�for�heavily�polluted�and�iron�oxide�rich�sediments.�Anal.�Chim.�Acta�414:�151�164.��

[48]� Chester,� R.;� Kudoja,� W.M.;� Thomas,� A.;� Thomas,� J.� (1985).� Pollution� reconnaissance� in� stream�sediments�using�non�residual�trace�metals.�Environ.�Pollut.�B�10:�213–238.�

[49]�Agemian,�H.;�Chau,�A.S.Y.�(1977).�A�study�of�different�analytical�extraction�methods�for�nondetrital�heavy�metals�in�aquatic�sediments.�Arch.�Environ.�Contam.�Toxicol.�6:�69–82.�

[50]� Lopez�Sanchez,� J.F.;� Sahuquillo,�A.;�Rauret,�G.;� Lachica,�M.;�Barahona,�E.;�Gomez,A.;�et�al.� (2002).�Extraction� procedures� for� soil� analysis.� In� Ph.� Quevauviller� (Ed.),�Methodologies� in� soil� and� sediment�fractionation�studies�(pp.�28–65).�Royal�Society�of�Chemistry.�Cambridge,�UK.�

[51]� Novozamski,� I.;� Lexmon,� T.M.;� Houba,� V.J.G.� (1993).� A� single� extraction� procedure� of� soil� for�evaluation�of�uptake�of�some�heavy�metals�by�plants.�Inter.�J.�Environ.�Anal.�Chem.:�51,�47.�

[52]�Colinet,�E.;�Gonska,�H.;�Griepink,�B;�Muntau,�H.�(1993).�The�certification�of�the�contents�of�Cd,�Cu,�Hg,�Ni,�Pb�and�Zn�in�a�light�sandy�soil.�EUR�Report�8833�EN,�57.�CEC.�Luxembourg.�

[53]� Perez,� G.;� Lopez�Mesas,�M.;� Valiente,�M.� (2008).� Assessment� of� Heavy�Metals� Remobilization� by�Fractionation:�Comparison�of� Leaching�Tests�Applied� to�Roadside�Sediments.�Environ.�Sci.�Technol.�42�(7):�2309–2315.�

[54]�Ure,�A.M,;�Quevauviller�Ph,;�Muntau�H.;�Griepink�B.;�1993,�Speciation�of�heavy�metals�in�solids�and�sediments.� An� account� of� the� improvement� and� harmonization� of� extraction� techniques� undertaken�under� the�auspices�of� the�BCR�of� the�Commission�of� the�european�Communitie.� Int.� J.� Environ.�Anal.�Chem.�51,�135�151.��

[55]� Mulchi,� C.L.;� Adamu,� C.A.;� Bell,� P.F.;� Chaney,� R.L.� (1992).� Residual� heavy� metal� levels� in� sludge�amended�coastal�plain�soils�–�II.�Predicting�metal�levels�in�tobacco�from�soil�test�information.�Commun.�Soil�Sci.�Plant�Anal.�23,�1053–1069.��

[56]�Lindsay,�W.L.;�Norvell,�W.A.�(1978).�Development�of�a�DTPA�soil�test�for�zinc,�iron,�manganese�and�copper.�Soil�Sci.�Soc.�Am.�J.�42:421�428.��

[57]� Melich,� M.� (1984).� Melich�3� Soiltest� Extractant:� A� Modification� of� the� Melich�2� Extractant.�Communications�in�Soil�Science�and�Plant�Analysis�15:1409–1416.��

1.�Introduction��

48�

[58]�Gupta,�S.K.;�Aten,�C.�(1993)�Comparison�and�evaluation�of�extraction�media�and�their�suitability�in�a�simple�model� to�predict� the�biological� relevance�of�heavy�metal�concentrations� in�contaminated�soils.�Intern.�J.�Environ.�Anal.�Chem.,�51,�26�47.�

[59]� Hughes,� J.C.;� Noble,� A.D.� (1991).� Extraction� of� chromium,� nickel� and� iron� and� the� availability� of�chromium� and� nickel� to� plants� from� some� serpentinite�derived� soils� from� the� eastern� Transvaal� as�revealed�by�various�single�and�sequential�extraction�techniques.�Commun.�Soil�Sci.�Plant�Anal.�22:1753�1766.��

[60]� Juste,� C.;� Solda,� P.� (1988).� Jnfluence� de� l’addition� de� differentes� mat&es� fertilisantes� sur� la�biodisponibilitc�du�cadmium,�du�manganese,�du�nickel�et�du�zinc�contenus�dans�un�sol�sableux�amend6�par�des�boues�de�station�d’epuration,�Agronomie,�8(10):�897�904.�

[61]� Sahuquillo,� A.;� Rigol,� A.;� Rauret,� G.� (2002).� Comparison� of� leaching� tests� for� the� study� of� trace�metals�remobilisation�in�soils�and�sediments.�J.�Environ.�Monit.�4:�1003–1009.�

[62]� Van� der� Sloot,� H.A.;� Heasman,� L.;� Quevauviller,� P.� (1997).� Harmonization� of� leaching/extraction�tests.�Studies�in�Environmental�Science�70.�Elsevier.�

[63]� DIN� (Deutsches� Institut� for� Normung)� (1993).� Bodenbeschaffenheit,� Vornorm,� DIN� v� 19730,� in�Boden�–�Chemische�Bodanuntersuchungsverfahren,�DIN,�Berlin,�4.�

[64]�AFNOR�(Association�Francaise�de�Normalization).(1994).�AFNOR,�Paris.�

[65]�UNICHIM�(Ente�Nazionale�_Italiano�de�Unificazione).�(1991)�UNICHIM,�Milan,�Italy.�

[66]�Houba,�V.J.G.;�Novozamsky,�I.;�Lexmond,�T.M.;�Van�der�Lee,�J.J.�(1990).�Applicability�of�0'01�M�CaCl2�as� a� single� extraction� solution� for� the� assessment�of� the�nutrient� status�of� soils� and�other� diagnostic�purposes.�Commun.�Soil�Sci.�Plant�Anal.�21:�2281�2290.�

[67]� Perez,� G.� (2005).� Disponibilidad� de� metales� tóxicosen� sitios� contaminados.� Aplicaciones� y�limitaciones�de�la�fraccionación�en�la�determinación�de�gradientes�de�polución.�PhD�thesis.�Universitat�Autonoma�de�Barcelona,�Spain.�

[68]�Zhai,�L;�Liao,�X.;�Chen,�T.;�Yan,�X.;�Xie,�H;�Wu,�B.;�Wang,�L.�(2008).�Regional�assessment�of�cadmium�pollution�in�agricultural�lands�and�the�potential�health�risk�related�to�intensive�mining�activities:�A�case�study�in�Chenzhou�City,�China.�J.�Environ.�Sci.�20:�696–703.�

[69]� Carr,� R.;� Zhang,� C.;� Moles,� N.;� Harder,� M.� (2008).� Identification� and� mapping� of� heavy� metal�pollution� in� soils� of� a� sports� ground� in� Galway� City,� Ireland,� using� a� portable� XRF� analyser� and� GIS.�Environ�Geochem.�Health�30:�45�52.�

[70]�Franco,�C.;�Soares,�A.;�Delgado,�J.�(2006).�Geostatistical�modelling�of�heavy�metal�contamination�in�the�topsoil�of�Guadiamar�river�margins�(S�Spain)�using�a�stochastic�simulation�technique.�Geoderma�136:�852�864.�

[71]�Shi,�W.;�Liu,�J.;�Du,�Z.;�Song,�Y.;�Chen,�C.;�Yue,�T.�(2009).�Surface�modelling�of�soil�pH.�Geoderma�150�1�2:�113�119.��

[72]�Panagopoulos,�T.;�Jesus,�J.;�Antunes,�M.D.C.;�Beltrao,�J.�(2006).�Analysis�of�spatial�interpolation�for�optimising�management�of�a�salinized�field�cultivated�with�lettuce.�Eur.�J.�Agron.�24�(1):�1�10.�

[73]� Jongman,�R.H.G.;� Ter�Braak,�C.J.F.;�Van�Tongeren,�O.F.R.� (1995).�Data�Analysis� in�Community�and�Landscape�Ecology.�Cambridge�University�Press.�U.K.��

[74]�Zhai,�L.;�Liao,�X.;�Chen,�T.;�Yan,�X.;�Xie,�H.;�Wu,�B.;�Wang,�L.�(2008).�Regional�assessment�of�cadmium�pollution�in�agricultural�lands�and�the�potential�health�risk�related�to�intensive�mining�activities:�A�case�study�in�Chenzhou�City,�China.�J.�Environ.�Sci.�20:�696–703.�

[75]� Pearson,� K.� (1901).� "On� Lines� and� Planes� of� Closest� Fit� to� Systems� of� Points� in� Space"� (PDF).�Philosophical�Magazine�2�(6):�559–572.�

[76]�Jolliffe,�I.T.�(1986).�Principal�Component�Analysis.�Springer�Verlag.�USA�

1.�Introduction�

49�

[77]�Shaw,�P.J.A.�(2003).�Multivariate�statistics�for�the�Environmental�Sciences.�Hodder�Arnold.�UK.�

[78]� Wang,� X.M.;� Dong,� Z.B.;� Zhang,� C.X.;� Qian,� G.Q.;� Luo,� W.Y.� (2009).� Characterization� of� the�composition� of� dust� fallout� and� identification� of� dust� sources� in� arid� and� semiarid� north� China.�Geomorphology�112:�144–157.�

[79]� Contini,� D.;� Genga,� A.;� Cesari,� D.;� Siciliano,�M.;� Donateo,� A.;� Bove,�M.C.;� Guascito,�M.R.� (2010).�Characterisation�and�source�apportionment�of�PM10�in�an�urban�background�site�in�Lecce.�Atmos.�Res.�95:�40–54.�

[80]�Chen,�X.H.;�Duan,�Z.H.�(2009).�Changes�in�soil�physical�and�chemical�properties�during��reversal�of�desertification�in�Yanchi�County�of�Ningxia�Hui�autonomous�region,�China.�Environ.�Geol.�57:�975–985.�

[81]�Sielaff,�K.;�Einax,�J.W.�(2007).�The�application�of�multivariate�statistical�methods�for�the�evaluation�of�soil�profiles.�J.�Soils�&�Sediments�7:�45–52.�

[82]�Pirkle,�F.L.;�Pirkle,�E.C.;�Pirkle,�W.A.;�Dicks,�S.E.�(1985).�Evaluation�through�correlation�and�principal�component�analyses�of�a�delta�origin�for�the�awthorne�and�Citronelle�sediments�of�peninsular�Florida.�J.�Geol.93:�493–501.�

[83]�Mostert,� M.M.R.;� Ayoko,� G.A.;� Kokot,� S.� (2010).� Application� of� chemometrics� to� analysis� of� soil�pollutants.�Trend�Anal.�Chem.�29:�430�445.�

[84]� Martin,� T.E:;� Davies,� M.P.� (2000).� Trends� in� the� stewardship� of� tailings� dams.� AGRA� Earth� &�Environmental�Limited.�Canada:�

[85]�Hodgson,�F.�(1977).�Proceedings�of�the�1st�Annual�British�Columbia�Mine�Reclamation�Symposium�in� Vernon.� The� Technical� and� Research� Committee� on� Reclamation.�Water� quality� and�Mine� Process�effluent.�Canada.�

[86]�Bell,�A.V.�(1976).�Waste�controls�at�base�metal�mines.�Environ.�Sci.�Technol.�10�(2):�130–135.�

[87]�http://www.iza.com/Documents/Communications/Publications/ZincGuide2004.pdf�(Last�visit:� June�2011).�

[88]� Sayilgan,� E.;� Kukrer,� T.;� Civelekoglu,� G.;Ferella� F.;Akcil,� A.;Veglio,� F.;� Kitis,�M.� (2009).� A� review�of�technologies�for�the�recovery�of�metals�from�spent�alkaline�and�zinc–carbon�batteries.�Hydrometallurgy�97:�158�166.�

[89]�Devi,�N.B.;�Nathsarma,�K.C.;�Chakravortty,�V.�(1997).�Extraction�and�separation�of�Mn(II)�and��Zn�(II)�from�sulfate�solutions�by�sodium�salt�of�Cyanex�272.�Hydrometallurgy�45�(1�2):�169�179.�

[90]�Jha,�M.K.;�Kumar,�V.;�Singh,�R.J.�(2002).�Solvent�extraction�of�zinc�from�chloride�solutions.�Solv.�Extr.�Ion�Exch.�20�(3):�389�405.�

[91]�Salgado,�A.L.;�Veloso,�A.M.O.;�Pereira,�D.D.;�Gontijo,�G.S.;�Salum,�A.;�Mansur,�M.B.�(2003).�Recovery�of� zinc� and� manganese� from� spent� alkaline� batteries� by� liquid�liquid� extraction� with� Cyanex� 272.� J.�Power�Sources�115:�367�373.�

[92]� Alamdari,� E.K.� (2004).� Synergistic� effect� of� MEHPA� on� co�extraction� of� zinc� and� cadmium� with�DEHPA.�Miner.�Eng.�17,�89–92.�

[93]�Owusu,�G.�(1998).�Selective�extractions�of�Zn�and�Cd�from�Zn–Cd–Co–Ni�sulphate�solution�using�di�2�ethylhexyl�phosphoric�acid�extractant.�Hydrometallurgy�47:�205–215.�

[94]�El�Nadi,�Y.A.;�Daoud,�J.A.;�Aly,�H.F.�(2007).�Leaching�and�separation�of�zinc�from�the�black�paste�of�spent�MnO2�Zn�dry�cell�batteries.�J.�Hazard.�Mater.�143:�328–334.�

[95]� Kyuchoukov,� G.;� Zhivkova,� S.� (2000).�Options� for� the� separation� of� Copper� (II)� and� Zinc� (II)� from�chloride�solutions�by�KELEX�100.�Solv.�Extr.�Ion�Exch.�18�(2):�293�305.�

[96]�Ritcey,�G.M.;�Lucas,�B.H.�(1971).�Purification�of�manganese�solutions�containing�copper�and��zinc�by�liquid�liquid�extraction,�using�bis(2�ethylhexyl)phosphoric�acid.�Can.�Metall.�Quart.�10(3),�223�8.�

1.�Introduction��

50�

[97]� Kunzmann,�M;� Kolarik,� Z.� (1992).� Extraction�of� zinc(II)�with�di(2�ethylhexyl)� phosphoric� acid� from�perchlorate�and�sulfate�media.�Solv.�Extr.�Ion�Exch.�10(1):�35�49.�

[98]� Sastre,� A.M.;� Miralles,� N.;� Aguilar,� M.� (1984).� The� extraction� of� copper(III)� from� sulfate� and�perchlorate�solutions�by�di(2�ethylhexyl)phosphoric�acid�dissolved�in�Isopar�H.�Chem.�Scripta�24(1):�44�48.�

[99]� Grimm,� R.;� Kolarik,� Z.� (1974).� Acidic� organophosphorus� extractants.� XIX.� Extraction� of� copper(II),�cobalt(II),�nickel(II),�zinc�(II),�and�cadmium(II)�by�bis(2�ethylhexyl)�phosphoric�acid.�J.�Inorg.�Nucl.�Chem.�36(1):�189�192.�

[100]� Ritcey,� G.M.;� Ashbrook,� A.W.� (1984).� Solvent� Extraction.� Principles� and� Applications� to� Process�Metallurgy.�Part�I�and�II.�Elsevier.�The�Netherlands�

[101]� Cheng,� C.Y.� (2000).� Purification� of� synthetic� laterite� leach� solution� by� solvent� extraction� using�D2EHPA.�Hydrometallurgy�56(3):�369�386.�

[102]�Rickelton,�W.A.;�Boyle,�R.J.� (1990).�The�selective�recovery�of�zinc�with�new�thiophosphinic�acids.�Solv.�Extr.�Ion�Exch.�8(6):�783�97.�

[103]�Lanagan,�M.D.;�Ibana,�D.C.�(2003).�The�solvent�extraction�and�stripping�of�chromium�with�Cyanex�272.�Miner.�Eng.�16(3):�237�245.�

[104]�Rickelton,�W.A.;�Flett,�D.S.;�West,�D.W.�(1984).�Cobalt�nickel�separation�by�solvent�extraction�with�bis(2,4,4�trimethylpentyl)phosphinic�acid.�Solv.�Extr.�Ion�Exch.�2(6):�815�38.�

[105]�Reddy,�B.R.;� Bhaskara� Sarma,� P.V.R.� (2001)..� Separation� and� recovery�of� cobalt� and�nickel� from�sulfate�solutions�of�Indian�Ocean�nodules�using�Cyanex�272.�Miner.�Process�Metall.�18(3):�172�176.�

[106]�Grinbaum,�B.�(2006).�Integrated�Approach�to�the�Development�of�Solvent�Extraction�Flow�Sheets.�Tsinguah�Sci.�Technol.�11,2006.207�211.�

[107]�Grinbaum�B.�(2001).�The�integral�approach�to�development�of�SX�processes.�In:�Marcus�Y,�ed.�Ion�Exchange�and�Solvent�Extraction.�Dekker.�USA.��

[108]�Mampton,�F.A.� (1997).�Mineralogy�and�Geology�of�Natural�Zeolites.�Southern�Printing�Company:�Blacksburg.�USA.�

[109]� Earl,� D.J.;� Deem,�M.W.� (2006).� Toward� a� database� of� hypothetical� zeolite� structures.� Ind.� Eng.�Chem.�Res.�45:�5449–5454.�

[110]�Perez�Ramirez,�J.;�Kapteijn,�F.;�Mul,�G.;�Moulijn,�J.A.�(2002).�NO�assisted�N2O�decomposition�over�Fe�based�catalysts.�Effects�of�feed�composition�and�catalyst�constitution.�J.�Catal.�208:�211�223.�

[111]� Pieterse,� J.A.Z.;� Booneveld,� S.;� van� den� Brink,� R.W.� (2004).� Evaluation� of� Fe�zeolite� catalysts�prepared�by�different�methods�for�the�decomposition�of�N2O.�Appl.�Catal.�B�51:�215–228.�

[112]� Chen,�H.Y.;� Sachtler,�W.M.H.� (1998).� Activity� and�durability� of� Fe/ZSM�5� catalysts� for� lean� burn�NOx�reduction�in�the�presence�of�water�vapor.�Catal.�Today�42:�73–83.��

[113]�Vradman,�L.;�Herskowitz,�M.;�Capek,�L.;�Wichterlova,�B.;�Brosius,�R.;�Martens,�J.A.�(2005).�Kinetic�experiments�and�modeling�of�NO�oxidation�and�SCR�of�NOx�with�decane�over�Cu��and�Fe�MFI�catalysts.�Ind.�Eng.�Chem.�Res.�44,�4523–4533.�

[114]�Guzman�Vargas,�A.;�Delahay,�G.;�Coq,�B.;�Lima,�E.;�Bosch,�P.;� Jumas,� J.C.� (2005).� Influence�of� the�preparation�method� on� the� properties� of� Fe�ZSM�5� for� the� selective� catalytic� reduction� of� NO� by� n�decane.�Catal.�Today,�107–08:�94–99.�

[115]�Delahay,�G.;�Valade,�D.;�Guzman�Vargas,�A.;�Coq,�B.�(2005).�Selective�catalytic�reduction�of�nitric�oxide�with�ammonia�on�Fe�ZSM�5�catalysts�prepared�by�different�methods.�Appl.�Catal.�B:�55,�149–155.��

[116]�Qi,� S.;� Yang,�R.T.� (2005).�Ultra�active�Fe/ZSM�5�catalyst� for� selective� catalytic� reduction�of�nitric�oxide�with�ammonia.�Appl.�Catal.�B.�60:�13–22.�

1.�Introduction�

51�

[117]�Schwidder,�M.;�Kumar,�M.�S.;�Bruckner,�A.;�Grunert,�W.�(2005).�Selective�reduction�of�NO�with�Fe�ZSM�5�catalysts�of�low�Fe�content:�I.�Relations�between�active�site�structure�and�catalytic�performance.�Chem.�Commun.�805–807.�

[118]� Panov,� G.I.;� Kharitonov,� A.S.;� Sobolev,� V.I.� (1993).� Oxidative� hydroxylation� using� dinitrogen�monoxide:�a�possible�route�for�organic�synthesis�over�zeolites.�Appl.�Catal.�A,�98:�1–20.��

[119]�Jia,�J.;�Pillai,�K.S.;�Sachtler,�W.M.H.�(2004).�One�step�oxidation�of�benzene�to�phenol�with�nitrous�oxide�over�Fe/MFI�catalysts.�J.�Catal.�221:�119–126.�

[120]�Perez�Ramirez,�J.;�Gallardo�Llamas,�A.�Impact�of�the�preparation�method�and�iron�impurities�in�Fe�ZSM�5� zeolites� for� propylene� production� via� oxidative� dehydrogenation� of� propane� with� N2O.� Appl.�Catal.�A�279:�117–123.�

[121]�Bulanek,�R.;�Wichterlova,�B.;�Novoveska,�K.;�Kreibich,�V.�(2004).�Oxidation�of�propane�with�oxygen�and/or�nitrous�oxide�over�Fe�ZSM�5�with�low�iron�concentrations.�Appl.�Catal.�A�264:�13–22.�

[122]� Halasz,� J.;� Hodos,� M.;� Hannus,� I.;� Tasi� G.;� Kiricsi,� I.� (2005).� Catalytic� detoxification� of� C2�chlorohydrocarbons�over�iron�containing�oxide�and�zeolite�catalysts.�Colloids�Surf.�A�265:�171–177.�

[123]� Idel�aouad,� R.;� Valiente,� M.;� Yaacoubi,� A.;� Tanouti,� B.;� Lopez�Mesas,� M.� (2010).� Rapid�decolourization�and�mineralization�of�the�azo�dye�C.I.�Acid�Red�14�by�heterogeneous�Fenton�reaction.�J.�Hazard.�Mater.�186:�745�750.�

[124]�Doval,�M.;�Garcia,�E.;�Luque,�j.;�Martín,�J.L.;�Rodas,�M.�(1991).�Arcillas�Industriales:�Yacimientos�y�Aplicaciones.�Ed.�Centro�de�Estudios�Ramón�Areces,�S.�A.�Spain..�

[125]� EPA,� Demonstration� Bulletin:� Forager� Sponge� technology,� EPA/540/MR�94/522.� U.S.�Environmental�Protection�Agency,�Superfund�Innovative�Technology�Evaluation.�1994.�USA.�

[126]�Al�Bazi,�S.J.;�Chow,�A.�(1983).�Mechanism�of�the�extraction�of�palladium(II)�thiocyanate�complex�by�polyether�foam.�Talanta�30:�487�492.�

[127]�Baghai,�A.;�Bowen,�H.J.M.� (1976).�Separation�of� rhodium�and� iridium�using�silicone�rubber� foam�treated�with�tri�n�octalymine.�Analyst�1976,�101,�661�665.��

[128]� Stasinakis,� A.S.;� (2008).� Use� of� selected� advanced� oxidation� processes� (AOPs)� for� wastewater�treatment�–�A�mini�review.�.Global�NEST�Journal�10:�376�385.�

[129]�Buxton,�G.V.,�Greenstock,�C.L.,�Helman,�W.P.,�Ross,�W.P.�(1988).�Critical�review�of�rate�constants�for� reactions� of� hydrated� electrons,� hydrogens� atoms� and� hydroxyl� radicals� (•OH/•O�)� in� aqueous�solution.�J.�Phys.�Chem.�Ref.�Data�17:�513�886.��

[130]�Katsumata,�H.;�Kawabe,�S.;�Kaneco,�S.;�Suzuki,�T.;�Ohta,�K.� (2004).�Degradation�of�bisphenol�A� in�water�by�the�phot�Fenton�reaction.�J.�Photochem.�Photobiol.�A�162:�297�305.�

[131]�Nogueira,�R.F.P.;�Trovó,�A.G.;�Modé,�D.�(2002).�Solar�photodegradation�of�dichloroacetic�acid�and�2,4�dichlorophenol�using�an�enhanced�photo�Fenton�process.�Chemosphere�48:�385�391.��

[132]�Chen,�A.;�Ma,�X.;�Sun,�H.�Decolorization�of�KN�R�catalysed�by�Fe�containing�Y�and�ZSM�5�zeolites.�J.�Hazard.�Mater.�156:�568�575.��

[133]� Fu,� F.;� Wang,� Q.;� Tang,� B.� (2009).� Fenton� and� Fenton�like� reaction� followed� by� hydroxide�precipitation�in�the�removal�of�Ni(II)�from�NiEDTA�wastewater:�A�compartive�study.�Chem.�Eng.�J.:�769�774.��

[134]�Barrault,�J.,�Tatibouet,�J.M.,�Papayannakos,�N.�(2000).�Catalytic�wet�peroxide�oxidation�of�phenol�over�pillared�clays�containing�iron�or�copper�species.�C.R.�Acad.�Sci.�Paris,�Chemistry�3,�777–783.�

[135]�Lv,�X.J.;�Xu,�Y.M.;�Lv,�K.L.;�Zhang,�G.C.� (2005).�Photo�assisted�degradation�of�anionic�and�cationic�dyes�over�iron(III)�loaded�resin�in�the�presence�of�hydrogen�peroxide,�J.�Photochem.�Photobio�A:�Chem.�173:�121�127.�

1.�Introduction��

52�

[136]Cheng,�M.M.;�Ma,�W.H.;�Li,� J.;�Huang,�Y.P.;�Zhao,� J.C.� (2004).�Visible�light�assisted�degradation�of�dye� pollutants� over� Fe(III)�loaded� resin� in� the� presence� of� H2O2� at� neutral� pH� values.� Environ.� Sci.�Technol.�38:�1569�1575.�

[137]�Gumy,�D.;�Fernandez�Ibañez,�P.;�Malato,�S.;�Pulgarin,�C.;�Enea,�O.;�Kiwi,�J.�(2005).�Supported�Fe/C�and�Fe/Nafion/C�catalysts�for�the�photo�Fenton�degradation�of�Orange�II�under�solar�irradiation,�Catal.�Today�101:�375�382.�

[138]�Lei,�J.;�Liu,�C.;�Li,�F.;�Li,�X.;�Zhoub,�S.;�Liu,�T.;�Gua,�M.;�Wu,�Q.�(2006).�Photodegradation�of�orange�I�in� the� heterogeneous� iron� oxide�oxalate� complex� system�under�UVA� irradiation,� J.� Hazard.�Mat.� 137:�1016�1024.�

[139]�Chen,�J.;�Zhu,�L.�(2007).�Heterogeneous�UV�Fenton�catalytic�degradatuib�if�dyestuff�in�water�with�hydroxyl�Fe�pillared�bentonite.�Catal.�Today�126:�463�470.��

[140]�Li,�F.B.;�Li,�X.Z.;�Li,�X.M.;�Liu,�T.X.;�Dong,�J.�(2007).�Heterogeneous�photodegradation�of�bisphenol�A�with�iron�oxides�and�oxalate�in�aqueous�solution,�J.�Colloid�Interface�Sci.�311:�481�490.�

[141]� Parra,� S.;� Henao,� L.;�Mielczarski,� E.;�Mielczarski,� J.;� Albers,� P.;� Suvorova,� E.;� Guindet,� J.;� Kiwi,� J.�(2004).�Synthesis,�testing,�and�characterization�of�a�novel�Nafion�membrane�with�superior�performance�in�photoassisted�immobilized�Fenton�catalysis,�Langmuir�20:�5621�5629.�

[142]� Thomas,� J.M.;� Thomas,� W.J.� (1997).� Principles� and� practice� of� heterogeneous� catalysis.� VCh.�Germany.��

[143]�Kuznetsova,�E.V.;�Savinov,�E.N.;;�Vostrikova,�L.A.;�Parmon,�V.N..(2004).�Heterogeneous�catalysis�in�the�Fenton�type�system�FeZSM�5/H2O2,�Appl.�Catal�B:�Environ.�51:�165�170.�

[144]�Phu,�N.H.;�Hoa,�T.T.K.;�Tan,�N.V.;�Thang,�H.V.;�Ha,�P.L.�(2001).�Characterization�and�activity�of�Fe�ZSM�5�catalysts�for�the�total�oxidation�of�phenol�in�aqueous�solutions,�Appl.�Catal.�B:�Environ,�34:�267�275.�

[145]�Chen,�A.;�Ma,�X.;Sun,�H.� (2008).�Decolorization�of�KN�R�catalyzed�by�Fe�containing�Y�and�ZSM�5�zeolites.�Journal�of�Hazardous�Materials�156:�568�575.�

[146]�Neamtu,�M.;�Zaharia,�C.;�Catrinescu,�C.;�Yediler,�A.;�Macoveanu,�M.;�Kettrup,�A.�(2004).�Fe�loaded�Y�zeolite�as�catalyst�for�wet�proxide�oxidation�of�reactive�azo�dye�Procion�Marine�H�EXL.�Applied�Catalysis�B:�Environmental�48:�287�294.�

[147]�Noorjahan,�M.;�Kumari,�V.D.;�Subrahmanyam,�M.;�Panda,�L.�Immobilized�Fe(III)�HY:�an�efficient�and�stable�photo�Fenton�catalyst.�Appl.�Catal.�B�57:�291�298.�

[148]�Rani,�V.R.;�Kishan,�M.R.;�Kulkarni,�S.J.;�Raghavan,�K.V.�(2005).�Immobilization�of�metalloporphyrin�complexes�in�molecular�sieves�and�their�catalytic�activity.�Catal.�Commun.�6:�531�538.�

[149]�Smedley,� P.L.;� Kinniburgh,�D.G.�A� review�of� the� source,�behaviour� and�distribution�of� arsenic� in�natural�waters.�Appl.�Geochem.�2002,�17,�517–568.�

[150]�Towards�a�More�Operational�Response:�Arsenic�contamination�in�South�and�East�Asian�Countries.�Technical� Report� No.� 31303.� WHO,� (World� Health� Organization),�2005.http://siteresources.worldbank.org/INTSAREGTOPWATRES/Resources/ArsenicVolI_WholeReport.pdf.�

[151]�Bagla,�P.;�Kaiser,�J.�(1996).�India’s�spreading�health�crisis�draws�global�arsenic�experts.�Science�274:�174–175.�

[152]�Greenwood,�N.N.;�Earnshaw,�A.�(1984).�Chemistry�of�Elements.�(Chapter�13).�Pergamon�Press,�UK.�

[153]� Fendorf,� S.;� Eick,� M.J.;� Grossl,� P.;� Sparks,� D.L.� (1997).� Arsenate� and� chromate� retention�mechanisms�on�goethite:�I.�Surface�structure.�Environ.�Sci.�Technol.�31:�315�320.��

[154]�Sherman,�D.M.;�Randall,�S.R.� (2003).�Surface�complexation�of�arsenic(V)� to� iron(III)� (hydr)oxides:�Structural� mechanism� from� ab� initio� molecular� geometries� and� EXAFS� spectroscopy.� Geochim.�Cosmochim.�Acta�67:�4223�4230.��

1.�Introduction�

53�

[155]�Farquhar,�M.L.;�Charnock,�J.M.,�Livens,�F.R.;�Vaughan,�D.J.� (2002).�Mechanisms�of�arsenic�uptake�from�aqueous� solution�by� interaction�with� goethite,� lepidocrocite,�mackinawite,� and� pyrite:� An�X�Ray�Absorption�Spectrescopy�Study.�Environ.�Sci.�Technol.�36:�1757�1762.�

[156]�Waychunas,�G.A.;�Rea,�B.A.;�Fuller,�C.C.;�Davis,�J.A.�(1993).�Surface�chemistry�of�ferrihydrite:�Part�1.�EXAFS� studies� of� the� geometry� of� coprecipitated� and� absorbed� absorbed� arsenate.� Geochim.�Cosmochim.�Acta�57:�2251�2269.�

[157]�Jessen,�S.;�Larsen,�F.;�Koch,�C.B.;�Arvin,�E.�(2005).�Sorption�and�desorption�of�arsenic�to�ferrihydrite�in�a�sand�filter.�Environ.�Sci.�Technol.�39:�8045�8051.�

[158]�Jia,�Y.;�Xu,�L.;�Fang,�Z.;�Demopoulos,�G.P.�(2006).�Observation�of�surface�precipitation�of�arsenate�on�ferrihydrite.�Environ.�Sci.�Technol.�40:�3248�3253.�

[159]�Singh,�D.B.;�Prasad,�G.;�Rupainwar,�D.C.�(1996).�Adsorption�technique�for�the�treatment�of�As(V)�rich�effluents.�Colloids�Surf.�A:�49�56.�

[160]� Gimenez,� J.;�Martinez,�M.;� Pablo,� J.;� Rovira,�M.;� Duro,� L.� (2007).� Arsenic� sorption� onto� natural�hematite,�magnetite�and�goethite.�J.�Hazard.�Mater.�141:�575�580.�

[161]�Li,�Z.;�Jean,�J.S.;�Jiang,�W.T.;�Chang,�P.H.;�Chen,�C.J.�(2011).�Removal�of�arsenic�from�water�using�Fe�loadedFe�loadedFe�loadedFe�loaded�natural�zeolite.�J.�Hazard.�Mater.�187(1�3):318�323.�

[162]�Dousova,� B.;�Grygar,� T.;�Martaus,� A.;� Fuitova,� L.;� Kolousek,�D.;�Machovic,� V.� (2006).� Sorption�of�As(V)�on�aluminosilicates�treated�with�Fe(II)�nanoparticles.�J.�Colloid�Interface�Sci.�302:�424�431.�

[163]� Rau,� I.;� Gonzalo,� A.;� Valiente,�M.� (2003).� Arsenic(V)� adsorption� by� immobilized� iron�mediation.�Modeling�of�the�adsorption�process�and�influence�of�interfering�anions.�React.�Funct.�Polym.�54:�85–94.�

[164]�Mohan,�D.;�Pittman�Jr.�C.U.�(2007).�Arsenic�removal�from�water/wastewater�using�adsorbents��A�critical�review.�J.�Hazard.�Mater.�142:�1–53.�

[165]� Payne,� K.B.;� Abdel�Fattah,� T.M.� (2005).� Adsorption� of� arsenate� and� arsenite� by� iron�treated�activated�carbon�and�zeolites:�effects�of�pH,�temperature,�and�ionic�strength,�J.�Environ.�Sci.�Health�A�40:�723�749.�

[166]�Elizalde�Gonzalez,�M.P.;�Mattusch,�J.;�Einicke,�W.D.;�Wennrich,�R.�(2001).�Sorption�on�natural�solids�for�arsenic�removal,�Chem.�Eng.�J.�81:�187–195.�

[167]� Elizalde�Gonzalez,�M.P.;�Mattusch,� J.;�Wennrich,� R.;�Morgenstern,� P.� (2001).� Uptake� of� arsenite�and�arsenate�by�clinoptilolite�rich�tuffs.�Micropor.�Mesopor.�Mater.�46:�277–286.�

[168]�Xu,�Y.H.;�Nakajima,�T.;�Ohki,�A.�(2002).�Adsorption�and�renoval�of�arsenic(V)�from�drinking�water�by�aluminum�loaded�shirasu�zeolite.�J.�Hazard.�Mater.�92:�275–287.�

[169]�Onyango,�M.S.;�Matsuda,�H.;�Ogada,�T.�(2003).�Sorption�kinetics�of�arsenic�onto�iron�conditioned�zeolite.�J.�Chem.�Eng.�Jpn.�36:�477–485.�

[170]�Hollerman,�W.;�Holland,�L.;�Ila,�D.;�Hensley,�J.;�Southworth,�G.;�Klasson,�T.;�Taylor,�P.;�Johnston,�J.;�Turner,� R.� (1999).� Results� from� the� low� level� mercury� sorbent� test� at� the� Oak� Ridge� Y�12� Plant� in�Tennessee.�J.�Hazard.�Mater.�68:�193�203.�

[171]� Muñoz,� J.A.;� Gonzalo,� A.;� Valiente,� M.� (2002).� Arsenic� sorption� by� Fe(III)�loaded� open�celled�cellulose�sponge.�Thermodynamic�and�selectivity�aspects.�Environ.�Sci.�Technol.�36:�3405�3411.�

[172]�Hsia,�T.H.;�Lo,�S.L.;�Lin,�C.F.;�Lee,�D.Y.�(1994).�Characterization�of�arsenate�adsorption�on�hydrous�iron�oxide�usingchemical�and�physical�methods.�Colloids�Surf.�A�85:�1�7.��

[173]�Grossl,� P.R.;� Eick,�M.;� Sparks,�D.L.;�Goldberg,� S.;�Ainsworth,�C.C.� (1997).�Arsenate�and�chromate�retention� on� goethite� II.� Kinetic� evaluation� using� a� pressure�jump� relaxation� technique.� Environ.� Sci.�Technol.�31:�321�326.�

[174]�Manceau,�A.�(1995).�The�mechanism�of�anion�adsorption�on�iron�oxides:�Evidences�for�the�bonding�of�arsenate�tetrahedra�on�free�Fe(O,�OH)6�edges.�Geochim.�Cosmochim.�Acta�59:�3647�3653.�

1.�Introduction��

54�

[175]�Fuller,�C.C.;�Davis,�J.A.;�Waychunas,�G.A.�(1993).�Surface�chemistry�of�ferrihydrite,�Part�2,�Kinetics�of�arsenate�adsorption�and�coprecipitation:�Geochim.�Cosmochim.�Acta�57:�2271�2282.�

[176]� Piorek,� S.� (1994).� Principles� and� applications� of�man�portable� X�ray� fluorescence� spectrometry.�TrAC,�Trends�Anal.�Chem�13�(7):�281�286.�

[177]�Kilbride�C.;� Poole,� J.;�Hutchings,� T.R.� (2006).�A� comparison�of�Cu,�Pb,�As,�Cd,� Zn,� Fe,�Ni� and�Mn�determined�by�acid�extraction/ICP–OES�and�ex�situ�field�portable�X�ray�fluorescence�analyses.�Environ.�Pollut.�143(1):�16�20.��

[178]� Bernick,� M.B.;� Kalnicky,� D.J.;� Prince,� G.;� Singhvi,� R.� (1995).� Results� of� field�portable� X�ray�fluorescence�analysis�of�metal�contaminants�in�soil�and�sediment.�J.�Hazard.�Mater.�43�(1�2):�101�110.�

[179]�Harper,�M.;�Pacolay,�B.;�Hintz,�P.;�Andrew,�M.E.�(2006).�A�comparison�of�portable�XRF�and�ICP�OES�analysis�for�lead�on�air�filter�samples�from�a�lead�ore�concentrator�mill�and�a�lead�acid�battery�recycler.�J.�Environ.�Monit.�8:�384�392.�

[180]� Argyraki,� A.;� Ramsey,� M.H.;� Potts,� P.J.� (1997)� Evaluation� of� portable� X�ray� fluorescence�instrumentation�for�in�situ�measurements�of�lead�on�contaminated�land.�Analyst�122:�743–749.�

[181]�Dost,�A.A.�(1996)�Monitoring�surface�and�airbone�inorganic�contamination�in�the�workplace�by�a�field�portable�X�ray�fluorescence�spetrometer.�Ann.�Occup.�Hyg.�40�(5):�589�610.�

[182]� Goldstein,� S.J.;� Slemmons,� A.K.;� Canavan,� H.E.� (1996)� Energy�Dispersive� X�ray� Fluorescence�Methods�for�Environmental�Characterization�of�Soils.�Environ.�Sci.�Technol.�30:�2318�2321.�

[183]�Brown,�G:E.;�Calas,�G.;�Waychunar,�G.A.;�Petiau,�J.�(1988).�X�ray�Absorption�Spectroscopy�amd�ots�Applications� in� Mineralogy� amd� Geochemistry,� in:� Reviews� in� Mineralogy.� Spectroscopic� methods� in�mineralogy�and�geology,�Hawthorne,�F.C:� (ed.),�Vol.18,�Chap.11,�431�512.�Mineral�Society�of�America,�Washington,�USA.�

[184]� Stern,� E.A.;� Heald,� S.M.� (1983).� Basic� Principles� and� applications� of� EXAFS,� in:� Handbook� on�Synchrotron�Radiation,�Koch,�E.E.�(ed.),�Vol.�1b,�995�1014.�New�York,�USA..�

[185]�Teo,�B.K.�(1986).�EXAFS:�basic�principles�and�data�analysis,�in:�Inorganic�Chemistry�Concepts�9,�1�349.�Springer�Verlag,�New�York,�USA..�

[186]� Fenter,� P.A.;� Rivers,�M.L.;� Sturchio,� N.C.;� Sutton,� S.R� (Eds.).� (2002).� Reviews� in�Mineralogy� and�Geochemistry�Vol.�49.�Mineralogical�Society�of�America,�Washington�DC.�USA.�

[187]�Waychunas,�G.A.;�Brown,�G.E.�(1984).�Application�of�EXAFS�and�XANES�spectroscopy�to�problems�in� mineralogy� and� geochemistry,� in:� EXAFS� and� Near�Edge� Structure� III,� Hodgson,� K.O.;� Hedman,� B.;�Penner�Hahn,� J.E:� (eds.),� Springer� Proceedings� in� Physics,� Vol.2,� 336�342.� Springer�Verlag,� New� York,�USA.�

[188]�Brown,�G.E.�Jr.;�Waychunas,�G.A.;�Ponader,�C.W.;�Jackson,�W.E.;�McKeown,�D.A.�(1986).�EXAFS�and�NEXAFS�studies�of�cation�environments� in�oxide�glasses.� Journal�de�Physique�Colloque,�Paris,�47,�661�668,.�

[189]�Calas,�G.;�Brown,�G.E:;�Waychunas,�G.A.;�Petiau,�J.�(1987).�X�ray�absorption�spectroscopic�studies�of�silicate�glasses�and�minerals.�Phys.�Chem.�Miner.�15:�19�29.�

[190]� Brown,� G.E.,� Jr.� (1990).� Spectroscopic� studies� of� chemisorption� reaction�mechanisms� at� oxide–water� interfaces.� p.� 309–364.� In� M.F.� Hochella,� Jr.� and� A.F.� White� (ed.)� Mineral–water� interface�geochemistry.�Reviews�in�Mineralogy�23.�Mineralogical�Soc.�of�America,�Washington,�DC.�USA.�

[191]�Manceau,�A.;� Charlet,� L.;� Boisset,�M.C.;�Didier,�B.;� Spadini,� L.� (1992).� Sorption�and� speciation�of�heavy�metals�on�hydrous�Fe�and�Mn�oxides.� From�microscopic� to�macroscopic.�Appl.�Clay�Sci.� 7:201–223.�

[192]�Fendorf,�S.E.,�Sparks,�D.L.;�Lamble,�G.M.;�Kelley,�M.J.�(1994).�Applications�of�x�ray�absorption�fine�structure�spectroscopy�to�soils.�Soil�Sci.�Soc.�Am.�J.�58:1583–1595.�

1.�Introduction�

55�

[193]�Schulze,�D.G.;�Bertsch,�P.M.�(1995).�Synchrotron�X�ray�techniques�in�soil,�plant,�and�environmental�research.�Adv.�Agron.�55:1–66.�

[194]� Brown,� G.E.;� Parks,� G.A.� (2001).� Sorption� of� trace� elements� on� mineral� surfaces:� Modern�perspectives� from� spectroscopic� studies,� and� comments� on� sorption� in� the�marine� environment.� Int.�Geol.�Rev.�43:�963–1073.�

[195]� Bertsch,� P.M.;Hunter,�D.B.� (2001).� Applications� of� synchrotron�based� X�ray�microprobes.� Chem.�Rev.�101:�1809–1842.�

[196]�Stern,�E.A.�(1978).�Structural�determination�by�X�ray�absorption.�Contemp.�Phys.�19:�239–310.�

[197]� Sayers,� D.E.;� Bunker,� B.� (1988).� Data� analysis.� In� D.C.� Koningsberger� and� R.� Prins� (ed.)� X�ray�absorption:� Principles,� applications,� techniques� of� EXAFS,� SEXAFS� and� XANES,� 211–256.� John�Wiley�&�Sons,�New�York.�USA.�

[198]�Fendorf,�S.E.;�Sparks,�D.L.�(1996).�X�ray�absorption�fine�structure�spectroscopy.�In�D.L.�Sparks�(ed.)�Methods�of�soil�analysis.�Part�3,�357–375.�SSSA�Book�Ser.�5.�SSSA,�Madison.�USA�.�

[199]� Stern,� E.A.;� Heald,� S.M.� (1983).� Basic� principles� and� applications� of� EXAFS.� In� E.E.� Koch� (ed.)�Handbook�of�synchrotron�radiation.�Vol.�10,�995–1014.�North�Holland,�Amsterdam.�Holland.�

[200]�Koningsberger,�D.C.;�Prins,�R.�(ed.)�(1988).�X�ray�absorption:�Principles,�applications,�techniques�of�EXAFS,�SEXAFS�and�XANES.�John�Wiley�&�Sons,�New�York.�USA�

[201]� Teo,� B.K.� (1986).� EXAFS� spectroscopy:� Basic� principles� and� data� analysis.� Springer�Verlag,� New�York.�USA.�

[202]�Stöhr,�J.�(1992).�NEXAFS�Spectroscopy.�Springer�Verlag,�New�York.�USA�

[203]� Bianconi,� A.� (1983).� In:� EXAFS� and� Near� Edge� Structure,� Bianconi,� A.;� Inococcia,� L.;� Stipcich,� S.�(eds.),�118�129.�Springer�Verlag,�Berlin,�Germany.�

[204]� Kim,� C.S.;� Rytuba,� J.J.;� Brown,� G.E.� (2004).� Geological� and� anthropogenic� factors� influencing�mercury�speciation�in�mine�wastes:�an�EXAFS�spectroscopy�study.�Appl.�Geochem.�19:�379�393.��

[205]� Slowey,� A.J.;� Johnson,� S.B.;� Rytuba,� J.J.;� Brown,� G.E.� (2005).� Role� of� organic� acids� in� promoting�colloid�transport�of�mercury�from�mine�tailings.�Environ.�Sci.�Technol.�39:�7869–7874.�

[206]�Slowey,�A.�J.;�Rytuba,�J.�J.;Brown,�G.�E.�(2005).�Speciation�of�Mercury�and�Mode�of�Transport�from�Placer�Gold�Mine�Tailings.�Environ.�Sci.Technol.�39:�1547–1554.�

[207]� Bernaus,� A.;� Gaona,� X.;Valiente,� M.� (2005).� Characterisation� of� Almadén� mercury� mine�environment�by�XAS�techniques.�J.�Environ.�Monit.�7:�771–777.�

[208]�Kim,�C.�S.;�Brown�Jr,�G.�E.;�Rytuba,�J.J.�(2000).�Characterization�and�speciation�of�mercury�bearing�mine�wastes�using�X�ray�absorption�spectroscopy�(XAS).�Sci.�Tot.�Environ.261:�157–168.�

[209]�Harris,�H.H.;�Pickering,�I.J.;George,�G.N.�(2003).�The�chemical�form�of�mercury�in�fish.�Science�301:�1203.�

[210]�Bernaus,�A.;�Gaona,�X.;�Esbri,� J.M.;�Higueras,�P.;�Falkenberg,�G.;�Valiente,�M.�(2006).�Microprobe�techniques�for�speciation�analysis�and�geochemical�characterization�of�mine�environments:�the�mercury�district�of�Almadén�in�Spain.�Environ.�Sci.�Technol.�40:�4090–4095.�

[211]� Riddle,� S.� G.;� Tran,� H.� H.;� Dewitt,� J.� G.;� Andrews,� J.C.� (2002).� Field,� Laboratory,� and� X�ray�Absorption�Spectroscopic�Studies�of�Mercury�Accumulation�by�Water�Hyacinths.�Environ.�Sci.�Technol.�36:�1965–1970.�

[212]� Bernaus,� A.;� Gaona,� X.;� Ivask,� A.;� Kahru,� A.;Valiente,� M.� (2005).� Analysis� of� sorption� and�bioavailability� of� different� species� of� mercury� on� model� soil� components� using� XAS� techniques� and�sensor�bacteria.Anal.�Bioanal.�Chem.�382:�1541–1548.�

56�

57�

2��

METHODOLOGY��MINE�SITES�CHARACTERIZATION ......................................................................................................59�

2.1.�STUDIED�MINES�DESCRIPTION.........................................................................................................59�2.2.�SAMPLING........................................................................................................................................64�2.3.�CHARACTERIZATION ........................................................................................................................65�2.4.�DATA�TREATMENT...........................................................................................................................70�

��REMEDIATION�TECHNIQUES�OF�INDUSTRIAL�CONTAMINATED�WATERS.................................73�

2.5.�ZINC�SOLVENT�EXTRACTION ............................................................................................................73�2.6.� Fe�LOADED� MATERIALS� FOR� THE� REMEDIATION� OF� ORGANIC� AND� INORGANIC� POLLUTED�WATERS ..................................................................................................................................................76�2.7.REFERENCES .....................................................................................................................................79�

58�

2.�Methodology�

59�

�Following�the�pattern�outlined� in�the� Introduction�chapter,�the�Methodology� is�divided�

into�two�main�parts:�I)�Mine�sites�characterization�and�II)�Remediation�technologies�applied�to�

aquatic�sources�containing�organic�and�inorganic�pollutants.��

MINE�SITES�CHARACTERIZATION�

In�this�section�the�methodologies�to�characterize�different�mine�areas�from�Morocco�and�

the�speciation�of�mercury�in�three�European�mines�are�described.��

2.1.�STUDIED�MINES�DESCRIPTION�

The� characterization� of� different� abandoned� mines� located� in� Marrakech� region�

(Morocco)�has�been�accomplished�by�determining�physical�and�chemical�parameters�of�mine�

area� soils,� whilst� soils� from� three� well� characterized� European� mercury� mines� have� been�

studied�to�determine�its�mercury�speciation�to�assess�its�toxicity.��

2.1.1.�MARRAKECH�MINES:�DRAA�LASFAR,�KETTARA,�SIDI�BOU�OTHMANE�AND�BIR�NEHASS�

(MOROCCO)�

The�studied�mines�are�located�35�km�northwest�of�Marrakesh�in�the�core�of�the�central�

Jebilet�mountains�(Figure�2.1).�The�climate�is�Mediterranean,�bordering�arid�and�semi�arid�with�

an� average� annual� precipitation� of� 231� mm� (10� years).� Temperatures� are� characterized� by�

great�daily�and�seasonal�variation�with�an�average�value�of�11.5�C�in�January�and�28.8�C�in�July�

[1].�Central�Jebilet�mineralized�body�consists�of�major�and�minor�lenses�of�massive�pyrrhotite�

(Fe2+0.95S),� with� small� amounts� of� sphalerite� (Zn0.95Fe2+0.05S),� galena� (PbS),� chalcopyrite�

(CuFe2+S2),�pyrite�(Fe2+S2),�arsenopyrite�(Fe

3+AsS)�and�glaucodot�(Co0.75Fe2+0.25AsS)�[2].��

The�Draa�Lasfar�mine�is�located�a�few�hundred�meters�from�the�Tensift�River,�close�to�a�

rural�community�of�about�5790�ha,�which�65%�are�occupied�by�farmland.�Draa�Lasfar�consists�

on�a�deposit�of�pyrite�mineral�discovered� in�1953�although�their�commercial�exploitation�did�

not�begin�until�1979.�Mineral�was�processed�by�flotation�after�primary�and�secondary�crushing�

and� grinding,� producing� 60�Mt� of� products� in� the� first� two� years� (1979�1980)� [1].� Industrial�

activity�stopped�in�March�1981,�although�it�restarted�in�1999�due�to�its�great�resource�of�poly�

2.�Methodology��

60�

metallic�components�(As,�Cd,�Cu,�Fe,�Pb,�Zn).�During�its�exploitation,�tailings�were�discharge�all�

around�the�mine�area�posing�a�risk�for�the�environment.�

Draa LasfarDraa Lasfar

Figure�2.1.�Location�of�Draa�Lasfar,�Kettara,�Sidi�Bou�Othmane�and�Bir�Nehass�mines�

The�Kettara�mine�produced�more�than�5.2�Mt�of�pyrrhotite�from�1964�to�1981,�although�

it�was�closed�in�1982�due�to�difficulties�during�the�production�of�the�pyrrhotite�and�its�use�in�

the�roasting�unit�[3].�During�the�exploitation,�more�than�3�Mt�of�mine�waste�were�stockpiled�

over�an�area�of�16�ha�without�concern�for�environmental�issues�(Figure�2.2).��

Figure�2.2.�Pictures�of�Kettara�mine�site�(Photo:�Gustavo�Pérez)�

The�Bir�Nehass�zinc�mine�and�the�Sidi�Bou�Othmane�(SB�Othmane)�mine�consist�on�old�

graphite�mines�characterized�by�an�intense�metamorphism�with�irregular�masses�of�grenatites�

and� marble.� SB�Othmane� mine� is� located� close� to� a� rural� district� and� surrounded� by�

agricultural� lands� with� estimated� reserves� of� 0.02� Mt� of� graphite� ore� deposit� (30�50%� of�

graphite).�Their�exploitation�started�on�1953,� treating�115� tons�per�day�of�mineral� (0.5%�Pb,�

7.4%� Zn� and� 6%� pyrite)� by� flotation� processes� until� its� closure� on� 1980.� Bir� Nehass� mine�

reserves�were�evaluated�to�be�0.25�Mt�containing�20%�of�the�mineral�enriched�in�Pb�(3.83%)�

and� Zn� (4.85%).� Their� exploitation� started� on� 1972�with� an� output� of� 90� tons� per� day� that�

increased� to�130� tons�per�day�after� the� implementation�of�a� flotation�circuit� in�1985.� It�was�

closed�at�the�end�of�the�20th�century.�

2.�Methodology�

61�

2.1.2.�EUROPEAN�MERCURY�MINING�DISTRICTS:�ALMADÉN,�MIERES�AND�IDRIJA��

The�location�of�the�three�mining�districts�is�depicted�in�Figure�2.3.�

Main mercury mines

Sampling sites

Main mercury mines

Sampling sites

Figure�2.3.�Metallurgical�sites�of�the�three�mercury�mining�districts,�Almadén,�Asturias�and�Idrija�

The�Almadén�mining�district�is�located�in�Ciudad�Real,�Spain�(Figure�2.3)�and�occupies�30�

km2.� It� is� located� within� an� area� sparsely� populated� (population� density� of� less� than� 25�

inhabitants/km2)� with� an� average� village� population� of� about� 2000� inhabitants� (Almadén�

population�is�7000�inhabitants).�Other�traditional�activities�are�agriculture�and�sheep�farming.�

Hunting� and� incipient� rural� tourism� are� the� only� alternatives� to� traditional� activities� [4].�

Almadén� is� the� largest� cinnabar� (HgS)� deposit� in� the�world� and� it� has� been� active� since� the�

Roman� times� until� the� present� days,� having� accounted� for� about� one� third� of� the� total� Hg�

world� production� [5,� 6].� Metallurgical� processing� evolved� from� Bustamante� furnaces,� with�

roasting� temperatures� over� 600� ºC,� to� Pacific� furnaces� in� the� last� century,� reaching�

temperatures�of�up�to�800�ºC.�Soils�at�Almadén�area�are�mainly�represented�by�quartz�and�a�

diversity� of� clay�type� minerals� such� as� chlorite,� illite,� kaolinite� and� pyrophyllite� and� high�

contents� of� carbonates� which� correspond� to� a� region� with� shales� and� quartzites� as� main�

components�of�the�stratigraphic�sequence�[7].�

� El�Entredicho�open�pit�mine El�Entredicho�dump

Figure�2.4.�Pictures�of�Almadén�mine�sites�(Photo:�José�Mª�Esbrí)�

2.�Methodology��

62�

The�mercury�mine�of� Idrija� is� located�50�km�west�of� Ljubljana,�Slovenia,� in� the�narrow�

valley� of� the� Idrijca�River.� The� Idrija�mine�has� been� the� second� largest�mercury�mine� in� the�

world�surpassed�only�by�the�Almadén�mine.�After�500�years�of�mining�activity�producing�a�total�

of�about�105,000�tons�of�Hg,�from�more�than�3�106�m3�of�ore�and�gangue,�the�mine�of�Idrija�

closed�in�1995�[8].�Taking�into�account�losses�during�mining�and�inefficient�smelting,�the�total�

volume�of�mined�Hg�is�estimated�to�be�at�least�140,000�tons�(9,�10).�Idrija�mining�district�is,�like�

Almadén,� a� monometallic� ore� deposit,� with� high� amounts� of� native� mercury� hosted� in�

carbonate� rocks.� The� mineralization� appears� as� two� main� species:� cinnabar� and� native�

mercury.� Other� minerals� appearing� in� its� paragenesis� are� metacinnabar,� pyrite,� marcasite,�

dolomite,� calcite,� kaolinite,� epsomite� and�melanterite.� The�mineralogical� characterization� of�

Idrija�samples�reveals�carbonate�bedrocks�as�main�components�of�the�stratigraphic�sequence,�

with� the� exception� of� the� meadow� soil� from� the� Pront� Hill,� which� was� developed� on�

carboniferous� clastic� rocks.� River�bed�and� suspended� sediments� are� composed�of� silica,� clay�

minerals,�Fe�and�Al�oxides,�hydroxides�and�carbonates�as�a�result�of�weathering�of�carbonate�

and�clastic�rock�in�the�Idrija�catchment�[11].�Metallurgical�processing�was�similar�to�Almadén�

during� the� last�century,�using�Pacific� furnaces�able� to�reach�up�to�800�ºC.�Due�to�the�mining�

and�ore�processing�operations,�Idrija�and�its�surroundings�have�been�polluted�with�Hg.�

Idrija�mine� Miéres�mine�

Figure�2.5.�Pictures�of�Idrija�and�Asturias�mine�(Photo:�José�María�Esbrí)

On� the� other� hand,� La� Peña�El� Terronal,� in� Miéres� (Asturias)� is� a� region� located� in�

northern� Spain� (Figure� 2.3)� with� abundant� Hg� deposits� that� has� been� an� important� Hg�

producer� on� the� global� scale.� This� site� has� an� intense�metallurgical� activity�with� an� average�

annual�production�ca.�517�tones�of�kg� [12].Mercury� is�present�as�cinnabar,�but�with�variable�

metacinnabar� and� metallic� mercury� proportions� and� with� other� metallic� minerals� such� as�

orpiment,� realgar,� melnikovite,� chalcopyrite,� arsenopyrite,� stibnite� and� galena� [13].� To�

summarize,�La�Peña�El�Terronal�mine�has�a�more�complex�mineralogy�than�Almadén�and�Idrija,�

with�high�amounts�of�arsenic�in�its�paragenesis,�In�addition,�their�rotary�furnaces�achieve�lower�

calcinations�temperatures�(over�600�ºC)�than�the�other�mining�districts�[14].��

2.�Methodology�

63�

The� total�mercury� concentration� in� soils� and� sediments�of� the� three�mining�districts� is�

well�documented�[15,�16,�17,�18,�19],�although�only�a�few�studies�dealt�with�inorganic�mercury�

speciation�[20,�21,�22,�23,�24].��

2.1.3.�AZNALCÓLLAR�TAILING�POND��

Aznalcóllar�mine� is� located� in� a�pyrite�rich� formation� following� the�Bethic�Chain�which�

extends� from� the� central� south� of� Spain� to� Portugal� (Figure� 2.6).� It� has� been� active� since�

Roman� times� due� to� their� high� grade� silver,� lead� and� zinc� ores.� In� this� type� of�mine,� ore� is�

milled,�washed,� and�after� treatment�with� several� reagents,� the�valuable�metal� sulfides�were�

separated�by� flotation.� In� this�process,�huge�volumes�of�acidic�wastes�and�tailings�generated�

are�stockpiled�in�a�tailings�pond.�The�tailings�reservoir�in�Aznalcóllar�is�situated�near�the�Agrio�

River,�a�small�tributary�of�the�Guadiamar�River.�The�waters�used�in�the�mining�operations�are�

currently� dumped,� after� depuration� in� the� mine,� in� this� small� tributary.� The� reservoir� was�

constructed� in� 1974�using� jetty�materials.�At� that� time� the�dam�was� approximately� 5m�high�

although�it�was�enlarged�several�times�using�tailing�materials.�At�the�time�of�the�tailings�dam�

failure�accident,�the�dam�was�approximately�25�m�high�[25].�

Figure�2.6.�Aznalcóllar�mine�location�

Figure�2.7.�Aznalcollar�tailing�ponds�(Photo:�Baruch�Grinbaum)

2.�Methodology��

64�

2.2.�SAMPLING��

Different� sampling� strategies� were� undertaken� for� the� mines� of� Marrakech� and� the�

European�mines�depending�on�the�specific�purpose�of�each�study.��

2.2.1.�MARRAKECH�MINING�DISTRICTS:�DRAA�LASFAR,�KETTARA,�SIDI�BOU�OTHMANE�AND�BIR�NEHASS�

(MOROCCO)�

Samples� were� taken� every� 50�meters� from� the�mining� area� towards� specific� receptor�

media�(river�creeks,�hills,�villages,� farms,�etc).�After�removing�the�first� layer�of�surface�soil� (2�

cm),�samples�were�taken�from�the�upper�20�cm�within�an�area�of�100�cm2�per�sample.�Residue�

samples�were� taken� on� the� stockpiled� dykes,� piles� or� ponds�where� tailings�were� deposited.�

Additionally,�3�representative�background�samples�were�collected�at�1km�from�the�mining�site,�

far�enough� to�avoid�disturbance� from�mining�operations.�After�air�drying�during�48h�at�30C,�

samples�were� sieved�below�2mm�through�a� stainless� steel� sieve� to� remove� large�debris� and�

stored�in�plastic�bottles�at�room�temperature.�

2.2.2.�EUROPEAN�MERCURY�MINING�DISTRICTS:�ALMADÉN,�ASTURIAS�(SPAIN),�IDRIJA�(SLOVENIA)�

Samples� of� soils,�mine� tailings,� calcines� and� riparian� soils� from� the�Almadén� site�were�

taken�at�a�depth�of�0–20�cm�and�stored�in�polyethylene�bags.�Samples�of�suspended�particles�

were�collected� from�the�water�column�and� let� to� sediment� in�a�clean� room.�All� the�samples�

were�air�dried�to�prevent�mercury�losses,�homogenized,�milled�and�sieved�below�2�mm.��

Soil�samples�from�Idrija�were�taken�using�a�stainless�steel�auger�at�a�depth�of�0–10�cm�

and� stored� in� polyethylene� bottles.� Suspended� river� sediment� was� sampled� during� a� flood�

event�of�the�Idrijca�river�by�means�of�a�net�drift�sampler�and,�after�removal�of�gravel,�stones�

and� plant� residues,� river� bed� and� suspended� sediments,� samples� were� dried� at� 30C� during�

three�days�until�constant�weight�in�the�dark,�homogenized�in�an�agate�mortar�and�separated�in�

two�grain�size�fractions:�fraction�below�0.063�mm�and�fraction�0.063–2�mm.��

Samples�from�Mieres�were�collected�in�La�Peña�El�Terronal�mine�site,�near�the�town�of�

Mieres.� The� sampling� included� samples� from� dumps,� calcines,� contaminated� soils� and� a�

chimney�channel�used�to�transport�roasting�smoke�to�the�top�of�a�mount.�Soils,�riparian�soils�

and�mine�tailings�samples�(~1.5�kg)�were�collected�at�10–30�cm�depth,�stored�in�polyethylene�

bags,�air�dried�in�a�clean�room�and�sieved�below�0.1�mm.��

2.�Methodology�

65�

2.3.�CHARACTERIZATION��

The�characterization�of�the�Marrakech�mine�soils�have�been�performed�by�determining�

its�physico�chemical�parameters�such�as�pH,�EC,�LOI�and�carbonate�content�as�well�as�heavy�

metal� concentration�and�mobility.� Regarding� the�mercury�mines,� the� characterization�of� the�

samples� has� been� performed� in� order� to� assess� its� speciation� using� synchrotron� techniques�

that,�in�turn,�were�also�applied�to�the�characterization�of�arsenic�sorption�onto�Fe�exchanged�

materials.��

2.3.1.�PHYSICO�CHEMICAL�PARAMETERS�

The� physical� characterization� consisted� in� the� measurement� of� the� pH,� the� electrical�

conductivity� (EC),� loss� on� ignition� (LOI)� and� the� carbonate� content� of� the� samples� following�

standard�methodologies�[26].��

Thus,�pH�measurements�were�done�in�a�soil�suspension�(2g/5�ml�of�distilled�water�stirred�

vigorously)� after� 2� h� of� deposition� using� a� pH�meter�WTW�Multiline� P4� Universal� pH�meter�

cabled�Sen�Tix�92T�pH�electrode�(Germany).��

The�EC�was�determined�in�a�soil�saturated�paste�(1g�soil/5�ml�of�distilled�water)�with�a�

conductimeter� WTW� Multiline� P4� Universal� Standard� Conductivity� Cell� TetraCon®� 325�

(Germany)�once�corrected�to�the�working�temperature�(20�C).��

Loss� on� ignition� (LOI)� was� determined� gravimetrically� after� volatilization� of� organic�

matter�on�a� furnace�at� 550°C�during�4h.� For� the� total� carbonate� content� three� replicates�of�

each�soil�were�stirred�during�6�h� in�an�HCl�4�mol/L� solution� (1.0g�of� soil/20�ml�of�HCl�4.0�M�

solution)�and,�after�filtering,�calcium�was�measured�using�a�JENWAY�PFP7�flame�photometer.��

2.3.2.�TOTAL�METAL�CONCENTRATION�

For� the� determination� of� the� total�metal� concentration,� aliquots� of� each� sample�were�

encapsulated�in�ten�milliliter�polyethylene�sample�cups�(Chemplex,�FL,�USA)�and�sealed�using�

pre�cut�Mylar®�circles�film�prior�to�their�analysis�with�a�FP�XRF�equipment�Alpha�6500R,�Innov�

X�Systems�(USA).�The�sample�thickness�in�the�cup�should�be�at�lest�1.2�cm�so�as�the�X�rays�can�

penetrate�the�sample.��

This�equipment�is�a�tube�type�energy�dispersive�instrument�with�a�tungsten�cathode�and�

a�silver�anode�that�can�generate�X�rays� in� the�energy�range�10�to�40�keV�and�10�50�μA.�The�

instrument� is� provided�with� a� circular� probe�window� (1.54� cm2� area)� and� employs� a� Si�PiN�

diodes�detector�with�an�energy�resolution�of�230eV�at�the�full�width�at�half�maximum�intensity�

of�the�manganese�(Mn)�K�X�ray�line.��

2.�Methodology��

66�

The�standardization�consists�in�the�collection�of�a�spectrum�of�a�known�spectrum�(Alloy�

316)�and�the�comparison�of�a�variety�of�parameters�to�values�stored�when�the�instrument�was�

calibrated�at� the�factory.�This�procedure�takes�about�1�minute�and�should�be�done�any�time�

the�hardware�is�initiated�or�restarted�and�must�be�repeated�if�the�instrument�is�operating�for�

more�than�4�hours.�

The�analyzing�time�for�each�sample�was�set�to�120s�for�the�heavy�elements�and�90s�for�

the�light�elements.�This�time�period�is�established�as�the�best�trade�off�between�accuracy�and�

speed�of�analysis.�For�accuracy,�an�instrument�blank�and�a�calibration�verification�check�(NIST�

2710)�was�checked�each�working�day�before�and�after�analyses�are�conducted�and�once�per�

every�twenty�samples�following�EPA�Method�6200�[27].��

Figure�2.8.�Innov�X�FP�XRF�model�ALPHA�6500�and�stand�for�laboratory�measurements�(Photo:�

Elena�Peralta)�

An�instrument�blank�is�used�to�verify�that�no�contamination�exists�in�the�spectrometer�or�

on� the� probe�window.� As� instrument� blank�we� employed� silicon� dioxide� although� it� can� be�

used� also� a� polytetraflurorethylene� (PTFE)� block,� a� quartz� block,� "clean"� sand,� or� lithium�

carbonate.�An�instrument�blank�should�also�be�analyzed�whenever�contamination�is�suspected�

by�the�analyst.�

A�calibration�verification�check�sample� is�used�to�check�the�accuracy�of�the� instrument�

and�to�assess�the�stability�and�consistency�of�the�analysis�for�the�target�analytes.��

The� check� sample� should� be� a� well� characterized� soil� sample� from� the� site� that� is�

representative�of� site� samples� in� terms�of�particle� size�and�degree�of�homogeneity�and� that�

contains�contaminants�at�concentrations�near�the�action�levels.�If�a�site�specific�sample�is�not�

available,�then�an�NIST�or�other�reference�material�that�contains�the�analytes�of� interest�can�

be�used�to�verify�the�accuracy�of�the�instrument.�To�verify�the�calibration,�the�measured�value�

for� each� target� analyte� should� be� within� ±20%� of� the� true� value.� In� this� sense,� NIST� 2710�

2.�Methodology�

67�

(Montana� soil)� standard� reference� sample� was� employed� as� calibration� verification� check,�

providing�results�within�specified�tolerances�(Table�2.1).��

Table�2.1.�NIST�2710�Certified�values�Element� Mass�fraction�(%)� � Element� Mass�fraction�(%)�

Aluminum� 6.44�±�.0.08� � Antimony� 38.4�±�3�Calcium� 1.25�±�0.03� � Arsenic� 626�±�38�Iron� 3.38�±�0.10� � Barium� 707�±�51�

Magnesium� 0.0853�±�0.042� � Cadmium� 21.8�±�0.2�Phosphorus� 1.01�±�0.04� � Copper� 2950�±�130�Potassium� 2.11�±�0.11� � Lead� 5532�±�80�Silicon� 28.97�±�0.18� � Mercury� 32.6�±�1.8�Sodium� 1.14�±�0.06� � Nickel� 14.3�±�1.0�Sulfur� 0.240�±�0.006� � Silver� 35.3�±�1.5�

Titanium� 0.283�±�0.010� � Vanadium� 76.6�±�2.3�� � � Zinc� 6952�±�91�

2.3.3.�TOTAL�MERCURY�CONTENT�

The� detection� of� mercury� at� trace� levels� is� a� complex� analytical� task� because� of� its��

specific�physical�and�chemical�properties.�Many�techniques�exist�for�mercury�determination�in�

different� matrix,� and� almost� all� of� them� involve� an� intermediate� stage� of� mercury�

preconcentration� in� absorption� traps� [28,� 29,� 30]� or� acid�mixtures� for� the� digestion� process�

prior� to�determination�by�Cold�Vapor�Atomic�Absorption�Spectrometry� (CV�AAS)� [31,�32,�33,�

34].� All� of� them� were� more� or� less� prone� to� analyte� losses� and/or� contamination.� Total�

mercury� content� of� all� solid� samples� corresponding� to� the� European� mercury� mines� was�

determined�by�Zeeman�atomic�absorption� spectrometry�using�high� frequency�modulation�of�

light�polarization�(ZAAS�HFM)�with�a�Lumex�RA�915+�analyzer�[35].��

In� this�mercury� analyzer,� the�mercury� contained� in� the� sample� is� atomized� by� a� glow�

discharge�mercury� lamp�placed� in�a�permanent�magnetic� field.�This�magnetic� field� �splits�the�

254�nm� mercury� resonance� line� into� three� polarized� components:� one� linear� � and� two�

circularly�polarised� in� the�opposite�directions� (�+� and���).�Only� �� components� are�detected.�

After� passing� through� a� polarization� modulator,� which� modulates� the� polarization� at� a�

frequency�of�50�kHz�and�thus�triggers�the�line�components�in�turn,�the�radiation�then�passes�

through�a�multi�path�cell,�whose�equivalent�optical�length�is�about�10�m.�Being�equipped�with�

narrow�band�high� reflectivity�mirrors,� the� cell� isolates� solely� the�254�nm� resonance� line� and�

suppresses�all� the�nonresonance�and�stray� radiation.�A� logarithm�of� the� intensity� ratio�of��+�

and� ��,� which� is� proportional� to� the�mercury� atom� concentration� in� the� cell,� is� determined�

upon�detecting�the�radiation�by�a�photodetector�and�subsequent�analog�digital�conversion�of�

its� electric� signal� by� a� microprocessor.� The� measurement� results� are� read� out� from� a� LC�

display.� In� this� measurement� technique,� the� analytical� signal� depends� only� on� mercury�

2.�Methodology��

68�

concentration� and� is� independent� of� the� presence� of� dust,� aerosols,� and� other� foreign�

contaminants�in�the�analytical�cell.�

The�detection�limit�of�this�technique�for�soils�and�sediments�samples�is�0.5�mg�Hg�kg-1.�

For�accuracy,�a�certified�reference�material�(CRM�025)�was�simultaneously�analyzed.�

Figure�2.9.�Lumex�RA�915+�analyzer�for�mercury�determinations�

2.3.4.�MOBILITY�OF�THE�MINE�SAMPLES�

Mobility� assays� were� performed� by� applying� established� methodology� of� single�

extraction� procedure� [36]� consisting� on�metal� extraction� of� soil� samples� with� HCl� 0.5�M� at�

solid:water� ratio� 1g/20� ml� during� 1h� under� magnetic� stirring.� After� each� extraction,� the�

suspension�was�centrifuged�10�min�at�3500�rpm�and�the�supernatant�was�filtered�using�0.22�

μm�Millipore�Millex�GS� filters� (Ireland).� The� extracts�were� analyzed� by�means� of� Inductively�

Coupled� Plasma�Optical� Emission� Spectroscopy� (ICP�OES)� using� an� equipment�

ThermoElemental�Intrepid�II�XLS�(USA)�(Figure�2.10).��

Agitation (1h) Sample + HCl (0,5M) (1:20)

Centrifuge 10min 4000rpm

Filtration of the extract

ICP-OES

�Figure�2.10.�Single�extraction�procedure�scheme�

The�ICP�OES�analytical�technique�allows�multielemental�analysis�of�metals� in�soils,�with�

an�excellent�performance�and�a�wide�analytic�range.�In�this�technique,�a�plasma�(ionized�gas,�

electrically�neutral)�is�used�to�excite�the�atoms�of�the�sample�so�that�when�relaxed�they�emit�

electromagnetic� radiation� at� wavelengths� characteristic� of� each� element� (in� the� region� of�

corresponding�UV�visible�spectrum)�with�an�intensity�proportional�to�its�concentration.��

The� plasma,� maintained� by� the� interaction� between� RF� frequency� and� ionized� argon,�

reaches�temperatures�up�to�10000�K.�The�sample�is�introduced�through�a�peristaltic�pump�into�

2.�Methodology�

69�

the� instrument� through� a� nebulizer� using� a� flow� of� argon,� that� disperses� the� liquid� into�

droplets�that�are�carried�to�a�cyclonic�chamber.�At�the�cyclonic�chamber�the�larger�droplets�are�

separated�from�the�smaller�drops,�which�are�moved�towards�the�plasma�by�a�flow�of�argon.��

In�table�2.2.�can�be�found�the�characteristic�wavelengths�used�for�the�measurement�of�

the�target�elements.��

Table�2.2.�Characteristic�wavelengths�of�elements�measured,�limit�of�detection�and�linearity�Element� Wavelength�(nm)� Limit�of�detection�(μg/L)� Linearity�(μg/L)�

As� 193.759� 1� 1�50�Cu� 324.754� 0.5� 0.5�50�Pb� 220.353� 1� 1�50�Zn� 213.856� 0.5� 0.5�50�

2.3.5.�XAS�MEASUREMENTS�

All� solid� samples� from� the� European�mercury�mines�were� prepared�mixing� an� aliquot�

with�polyethylene�(IR�quality),�homogenized�with�a�vortex�for�2�min,�pressed�as�a�pellet�with�5�

ton�cm�2�of�pressure�using�an�IR�press�and�sealed�between�Kapton™�tape.��

Its� XANES� measurements� were� performed� at� the� HASYLAB� synchrotron� facility�

(Germany)� at� A1� bending�magnet� beamline.� All� measurements� were� carried� out� at� room�

temperature.�The�beamline�set�up�consisted�of�a�Si(111)�double�crystal�monochromator,�three�

ionization� chambers� as� transmission� detectors� and� a� 7�pixel� Ge� fluorescence� detector.� The�

absorption�of�mercury�was�recorded�at�its�LIII�energy�(12284�eV)�(Figure�2.11).��

A)� B)� Figure�2.11.�A)�Sample�holder�containing�6�pellets�of�the�samples.�B)�Schematic�XAFS�set�up�

References�for�XANES�fingerprint�adjustments�included�the�following�minerals�and�pure�

compounds:� HgCl2,� HgSO4,� HgO,� CH3HgCl,� Hg2Cl2� (calomel),� HgS� red� (cinnabar),� HgS� black�

(metacinnabar),� Hg2NCl0.5(SO4)0.3(MoO4)0.1(CO3)0.1�H2O� (mosesite),� Hg3S2Cl2� (corderoite),�

Hg3(SO4)O2�(schuetteite)�and�Hg2ClO�(terlinguaite).�This�selection�was�undertaken�on�the�basis�

of�our�prior�knowledge�of�the�geochemistry�of�the�different�studied�areas�[16,�17,�18,�19,�20,�

21],�as�well�as�the�possible�weathering�and�anthropogenic�processes�taking�place�in�each�site.�

2.�Methodology��

70�

On�the�other�hand,�the� local�environment�of�arsenic�adsorbed�onto�different�materials�

was� investigated� by� both� X�ray� absorption� near�edge� structure� (XANES)� and� extended� X�ray�

absorption�fine�structure�(EXAFS)�spectroscopies.�Samples�were�prepared�following�the�same�

methodology�as�described�for�mine�soil�samples.�Arsenic�spectra�were�collected�at�its�K�edge�

energy� (11867�eV)�at� beamline�C�of�DORIS� III�HASYLAB� facilities.� This�beamline� is� essentially�

equal�to�A1�beamline�(Figure�2.11).�The�monochromator�consisted�in�a�Si�(111)�double�crystal�

and� the�detection�was�measured�either�by�adsorption�and� fluorescence�using�a�7�pixel� Si(Li)�

detector�over�the�energy�range�11700�12700�eV.�The�monochromator�was�calibrated�using�the�

LIII�edge�of�a�gold�foil�(11919�eV).�As�a�standard�for�the�EXAFS�fitting,�Na2HAsO4�7H2O�(Panreac)�

was�selected.�Thus,�by�the�comparison�of�the�empirical�results�with�the�theoretical�results�for�

Na2HAsO4�7H2O�an�estimation�of�the�goodness�of�the�selected�paths�can�be�obtained.��

2.4.�DATA�TREATMENT�

Collected�data�with�the�different�experimental�techniques�have�been�treated�to�extract�

appropriate� information� to� characterize� target� samples� and� corresponding� processes� in� the�

present�studies.�The�following�tools�were�used�to�this�purpose:�

2.4.1.�CONCENTRATION�ENRICHMENT�RATIOS��

In� order� to� avoid� the� limitation� of� using� total� concentrations� of� pollutants� without�

considering� the�geochemical�variability�of� the�geological� substrate�or� the�particle� size�effect,�

some� indicators� can� be� employed.� Concentration� enrichment� ratios� (CER),� also� called�

enrichment�factors,�were�used�in�the�studies�concerning�the�characterization�of�mine�area�soils�

from�Morocco�and�are�used�to�identify�and�quantify�the�extent�of�human�interference�in�soils�

that,�by�extension,�are�also�an�indicator�of�soil�contamination.��

CER�were�initially�developed�to�speculate�on�the�origin�of�elements�in�the�atmosphere,�

precipitation� or� seawater� [37,� 38,� 39].� This� use�was� progressively� extended� to� the� study� of�

soils,� lake� sediments,� peat,� tailings� and� other� environmental�materials� [40,� 41]� comparing� a�

target�pollutant�with�a�background�element.�The�formula�to�calculate�CER�can�be�generalized�

as:��

sample sample

background background

[El] /[X]CER=

[El] /[X]� � (Equation�2.1)

where�“El”� is� the�element�under�consideration,�“X”� is� the�chosen�reference�element�and�the�

subscripts�“sample”�or�“background”�indicate�which�medium�the�concentration�refers�to�[42].�

The�reference�element�“X”�should�be�little�affected�by�weathering�processes�and�should�show�

2.�Methodology�

71�

little�variability�of�occurrence.�In�this�sense,�the�most�common�reference�elements�employed�

in� the� literature�are�aluminum�(Al),� zirconium�(Zr),� iron� (Fe),� scandium�(Sc),�and�titanium�(Ti)�

[43,� 44,� 45],� although� there� have� been� also� attempts� at� using� other� elements� such� as�

manganese�[41],�chromium�[46],�lithium�[47]�and�calcium�[48].�In�this�work,�Zr�was�selected�as�

lithogenic�element�due�to�homogeneity�of�Zr�concentration�in�all�samples�and�background.�

The� interpretation� of� CERs� can� be� employed� to� determine� the� anthropogenic�

contribution�[49]�(Table�2.3).�

Table�2.3.�Anthropogenic�contribution�at�different�CER�values�CER� Anthropogenic�contribution�

<2� Minimal�or�nule�2�5� Moderate�5�20� Significant�20�40� Strong�>40� Extreme�

2.4.2.�GEOGRAPHIC�INFORMATION�SYSTEMS�

Used� in�the�studies�concerning�the�characterization�of� the�mine�sites� from�Marrakech.�

Contour� maps� of� CER� values� of� target� elements� were� done� by� GIS� representation� using�

Miramon�v6.4�Complete�Geographical�Information�System�and�Remote�Sensing�software�[50]�

choosing�IDW�interpolator�as�the�most�suitable�due�to�the�irregular�sampling�realized�on�the�

mine�sites.��

2.4.3.�STATISTICAL�TOOLS�

Box�plots�graphs�of�Kettara,�Bir�Nehass�and�Sidi�Bou�Othmane�were�obtained�using�SPSS�

Statistics�17.0�software.�PCA/APCS�was�realized�by�using�Excel�add�in�XLStat�Data�Analysis�and�

Statistical� Software� [51].� Samples� were� scaled� by� using� the� standard� normal� variate� (SNV)�

algorithm�and�Bartlett�sphericity�test�was�checked�in�order�to�confirm�that�the�variables�were�

uncorrelated.�Kaiser�criterion�was�used�to�select�principal�components,�and�only�factors�with�

eigenvalues�greater�than�1�were�considered.��

2.4.4.�XAS�DATA�TREATMENT��

XANES�spectra�of�samples�from�European�mercury�mines�were�processed�using�SixPACK�

data�analysis�software�package�[52,�53,�54,�55].�Spectra�processing�included�energy�correction,�

signal� normalization� and� background� correction.� After� data� correction� and� normalization,�

principal�component�analysis� (PCA)�was�applied� to� the�set�of�unknown�spectra� to�determine�

the�number�of�principal�components�required�to�describe�the�variation�in�the�data.�Then,�the�

PCA� results� were� used� with� a� target� transformation,� which� projected� the� spectrum� from� a�

2.�Methodology��

72�

reference�compound�onto�the�vector�space�defined�by�the�components.�If�the�target�vector�lay�

within�this�component�space�(above�the�95%�confidence�level),�then�this�reference�compound�

was� determined� to� be� present� in� the� corresponding� sample.� Finally,� a� linear� least�squares�

approach�was�used� to� determine� the� fractional� amount�of� each� reference� compound� in� the�

samples� [56,� 57,� 58].� The� quality� of� the� target� transformation�was� given� by� the� reduced� �2�

value,� which� represents� the� goodness� of� the� fit� to� the� spectra� data� using� the� linear�

combination�procedure�[59]�and�is�defined�as:��

2 2

1

1reduced � ( )-

Nobs fiti iiN P� �

�� � � (Equation�2.2)

where�obsi� is�the�ordinate�of�the�XANES�spectrum�measured�from�the�sample�at�the�ith�energy�

point,�fiti� is�the�ordinate�of�the�fitted�XANES�spectrum,�N�is�the�number�of�data�points�in�the�

fitted� XANES� energy� range� and� P� is� the� number� of� fitted� components.� A� higher� reduced��2�

denotes� that� the� Hg� compounds� compared� possess� a� lower� degree� of� similarity.� This� �2�

represents�the�goodness�of�the�model�fit.�

EXAFS�data�treatment�was�performed�with�VIPER�software�[60].�In�VIPER,�the�extracted�

EXAFS�signal�or���function�was�converted�to�frequency�(k)�space,�weighted�by�k2,�and�Fourier�

transformed� to� produce� the� R�space� EXAFS� pair�correlation� function,� which� is� similar� to� a�

radial� distribution� function.� The� program� includes� an� option� to� iterate� the� post�edge�

background� to� obtain� peaks� with� good� resolution� and� to� minimize� spurious� peaks� at� small�

radial� distances.� After� back�transforming� the� first� and� second� peaks� in� the� raw� Fourier�

transform� into� frequency� (k)� space,� an� ordinary� fitting� analysis� was� performed� to� obtain�

interatomic� distances� and� coordination� numbers.� Phase� and� amplitude� functions� were�

extracted� from� sodium� arsenate� and� scorodite� standards� (FeAsO4�4H2O).� The� Debye�Waller�

factor�(�)�for�the�unknown�samples�were�constrained�to��=0.003�for�the�first�shell,�and��=0.008�

for�the�second�shell.��

2.�Methodology�

73�

REMEDIATION�TECHNIQUES�OF�INDUSTRIAL�CONTAMINATED�WATERS�

Several�remediation�technologies�have�been�studied�in�this�thesis�to�their�application�on�

the�treatment�of�industrial�waters�containing�different�types�of�pollutants.�In�this�concern,�two�

different�techniques�have�been�studied:��

1)�Conventional�methodologies�such�as�solvent�extraction�has�been�employed�to�the�recycling�

of� zinc� from� a� mine� tailing� pond� to� provide� an� economical� benefit� while� diminishing� the�

volume�of�hazardous�materials�contained�in�the�mine�tailing�at�laboratory�and�pilot�plant�scale.��

2)�Fe�exchanged�materials�have�been�employed�as�catalysts�for�the�degradation�of�persistent�

organic�pollutants�(POPs)�by�means�of�Fenton�processes�as�well�as�to�the�sorption�of�inorganic�

contaminants.��

2.5.�ZINC�SOLVENT�EXTRACTION��

2.5.1.�LABORATORY�EXPERIMENTS�

In�order�to�get�a�Zn�sulphate�rich�liquor�to�be�used�later�in�electrowinning�process,�the�

performance� of� a� newer� commercial� extractant,� Ionquest� 290� (Bis(2,4,4�trimethylpentyl)�

phosphinic�acid),�is�compared�with�the�results�of�more�conventional�extractants�DEHPA�(Di�2�

(ethylhexyl)� phosphoric� acid)� and�Cyanex�272� (Bis(2,4,4�trimethylpentyl)� phosphinic� acid)� for�

the�solvent�extraction�of�a�Zn�rich�mine�effluent.��

In�the�related�tailing�mine�samples,�Fe�was�removed�from�the�mine�water�prior�to�the�SX�

treatment�by�means�of�a�biooxidation�process�using�Thiobacillus�ferrooxidans�and�a�selective�

alkaline�precipitation�step�[61,�62]�to�obtain�a�pregnant�leach�solution�(PLS)�without�iron,�since�

there�are�no�reagents�commercially�available�capable�to�extract�Zn�selectively�from�a�solution�

containing�Fe.��

The�extractants�DEHPA�(Batch�ref.�0063829)�and�Ionquest�290�(Batch�Ref.�G05A1)�were�

kindly� supplied�by�Rhodia�UK� Ltd.� and�Cyanex�272�was�purchased� from�Cytec� Industries� BV,�

Netherlands.� Ionquest� 290� has� the� same� active� ingredient� as� Cyanex� 272� but� has� a� lower�

content�of�inactive�impurities,�the�phosphine�oxide�impurity�is�<5%�in�Ionquest�290�but�around�

15%�in�Cyanex�272�[63].��

Two� type� of� kerosene�with� different� flash� point�were� also� studied� as� solvents� for� the�

extractants.� Commercial� grade� extra�pure� aliphatic� kerosene� Ketrul� D80� and� Ketrul� D100�

2.�Methodology��

74�

(Batch�ref.�20062016�and�20061560,�respectively)�were�kindly�supplied�by�Total�Fluides�France.�

Ketrul� D80� and� Ketrul� D100,� have� a� flash� point� of� 72� ºC� and� 100ºC� or� superior� (ISO� 2719),�

respectively.�It�must�be�pointed�out�that�the�higher�the�flash�point�the�lesser�the�flammability�

of�the�kerosene,�and,�therefore�the�higher�the�security�of�the�solvent�extraction�process.��

Sulfuric� Acid� 95�98%�was� purchased� from� J.T.� Baker,� Phillipsburg,� NJ,� USA� and� it� was�

used�to�strip�the�zinc�from�the�organic�enriched�phase.�All�the�reagents�were�used�as�received�

without�any�further�purification.�Stoppered�glass�tubes�of�50�mL�were�used�for�the�two�phases�

contact�and�the�agitation�took�place�in�a�rotating�rack.��

For�the�kinetic�experiments�10�mL�of�DEHPA�40%�(v/v),�Cyanex�272�5%�(v/v)�or�Ionquest�

290�5%�(v/v)�were�agitated�with�10�ml�of�PLS�(ratio�A:O=1)�in�a�rotating�rack�during�5,�10,�20,�

30,�40�or�60�min.�The�organic�phase�loaded�with�the�target�metal/s�(OP)�was�stripped�with�5�

mL�of�H2SO4�2.0�M�during�3h�to�ensure�complete�stripping.�DEHPA�concentration�was�higher�

due�to�efficiency�related�to�extraction�yield�and�extractant�cost.�

To�determine�selectivity,� isotherms�varying�the�A:O�ratio�from�0.1�to�10�were�realized.�

Different�volumes�of�Cyanex�272�5%�(v/v),�Ionquest�290�5%�(v/v)�or�DEHPA�40%�(v/v)�in�each�

type�of�kerosene�were�equilibrated�with�the�PLS�during�15�min�and�thereafter�OP�was�stripped�

with�5�mL�H2SO4�2.0�M.�No�centrifugation�of�the�dual�phase�system�was�required�because�of�

the�clear�phase�separation�obtained.�Selectivity�of�the�solvents�towards�Zn�was�determined�by�

the� recovery�of�each�metal� (equation�2.3)� and�by� the�amount�of�metal� remaining� in� the�OP�

(Equation�2.5)�which�is�calculated�by�the�different�amounts�of�metal�in�the�raffinate�(Equation�

2.4)�and�in�the�OP.�

strip

PLS

Zn%Recovery= ×100Zn

� � � �

(Equation�2.3)

raffinate

PLS

Zn%Remaining R= ×100Zn

� � � �

(Equation�2.4)

%Remaining OP = 100 - %Recovery - %Remaining R (Equation�2.5)

Major� elements� present� in� the� PLS,� in� the� strip� liquor� and� in� the� raffinate� were�

determined�by�ICP�OES�Thermo�Iris�Intrepid�II�XLS�(USA).�

2.5.2.�SCALING�THE�SX�TO�A�PILOT�PLANT�

The�bio�oxidation�reactor�consisted�of�a�150�cm�high�and�70�cm�diameter�stainless�steel�

column�divided�to�three�zones:�a�30�cm�deep�bottom�space�where�air�and�solution�were�fed�in,�

a�siliceous�stone�packed�bed�containing�the�inoculum�supported�by�a�stainless�steel�screen�and�

an�air�space�at�the�top�to�pH�and�Eh�control�and�a�50mm�pipe�to�let�the�solution�overflow.�The�

effluent� circulated� through�a� tank�where�pH�was� initially�adjusted�and,�after�pH�adjustment,�

2.�Methodology�

75�

the�solution�was�transferred�to�the�bioreactor�where�bio�oxidation�of�ferrous�ions�took�place,�

to�be�finally�transferred�to�a�precipitation�tank�fed�by�a� lime�solution�from�a�separated�tank.�

After�precipitation�and�sedimentation�of� iron�compounds,� the�supernatant�was�directly�used�

as�feed�solution�to�the�solvent�extraction�stage�(Figure�2.12).�

The� pilot� plant� process�was� undergone�with� Ionquest� 290.� Ionquest� 290� (Purity>95%)�

was� supplied� by� Rhodia� UK� Ltd.� and� commercial� grade� extra�pure� aliphatic� kerosene� Ketrul�

D100(bp� 100ºC)� by� Total� Fluides� France.� A� solution� of� Na2CO3�was� used� for� pH� adjustment�

during� the� solvent� extraction� experiments.� For� the� stripping� step,� the� loaded� solvent� was�

contacted�with� 2M�H2SO4� at� a� phase� ratio� O:A=10,� the� initially� expected� phase� ratio� in� the�

plant�to�achieve�the�required�zinc�transfer�in�the�EW�plant�of�20�g/L.�In�practice,�it�was�found�

that� the�required�transfer� in� the�EW�plant�was�40�g/L�Zn,�consequently,� the�phase�ratio�was�

modified� to� O:A=20.� No� laboratory� tests� were� undertaken� at� this� phase� ratio,� but� directly�

applied�in�the�pilot�plant.�

Figure�2.12.�Bio�oxidation�and�SX�flow�sheet�at�the�pilot�plant�

Two�Bateman�Pulsed�Columns�(BPC)�were�required�for�the�SX�and�stripping�processes�at�

the�pilot�plant�due�to�their�demonstrated�feasibility� in�several�SX�plants�[64,�65,�66].�BPC�are�

large� diameter� vertical� pipes� filled� alternately� with� disk� and� doughnut� shaped� baffles� to�

promote�contact�between�the�organic�and�aqueous�phases�through�the�column.�A�decanter�at�

each�end�of�the�column�allows�the�liquids�to�coalesce�and�be�decanted�separately.�When�the�

solvent�phase� is� continuous,� the� interface�between� the�phases� is� in� the� lower�decanter� and�

when�the�aqueous�phase�is�continuous,�it�is�in�the�upper�decanter.�The�columns�are�pulsed�by�

blowing�air�at� the�required�amplitude�and� frequency�of� the�pulses� [67].�An�80�mm�diameter�

BPC,�7m�high�(equivalent�to�3�theoretical�mixer�settler�stages)�was�chosen�for�the�SX�process�

2.�Methodology��

76�

and�a�40�mm�diameter�BPC�6�meter�high�for�the�stripping.�The�complete�piping�of�the�plant�is�

shown�in�Figure�2.12.�

EXTRACTIONCOLUMN

EXTRACTIONCOLUMN

Figure�2.13.�Section�of�the�columns�layout�at�Aznalcollar�pilot�plant�(Photo:�Baruch�Grinbaum)�

All�flows�were�fed�through�metering�pumps�and�the�flow�rates�of�all�inlets�and�aqueous�

outlets�were�measured�by�rotameters.�The�pilot�was�run�for�12�working�days,�10�hours�a�day�

on�average,�a�total�of�120�hours.�The�average�flow�rate�of�the�aqueous�feed�was�150�L/h,�so,�

about�18�m3�of� tailing�solution,�after�Fe�precipitation,�were� treated.�The�total�volume�of� the�

solvent�was�300�L�and� it�had�5%�Ionquest�290�dissolved� in�kerosene�(20%�aromatic�and�80%�

aliphatic);�the�weak�electrolyte�(WE,�strip�solution)�consisted�of�190�g/L�H2SO4�with�50�g/L�Zn2+.�

A�solution�of�50�100�g/L�Na2CO3�was�prepared�periodically�in�a�60�L�barrel�and�used�to�adjust�

the�pH.�

The� concentration� of� Zn� was� determined� using� a� Perkin� Elmer� 3110� AAS� at� the� pilot�

plant� laboratory.�The�Zn� in�the�raffinate�and�SE�was�determined�directly,�while�the�Zn� in� the�

barren�and�loaded�solvent�(BS�and�LS)�solution�were�determined�after�stripping�using�H2SO4.�

2.6.� FE�EXCHANGE� MATERIALS� FOR� THE� REMEDIATION� OF� ORGANIC�AND�INORGANIC�POLLUTED�WATERS�

Several�materials�have�been�studied�as�Fe�supports�to�be�used�either�as�Fenton�catalysts�

or�arsenic�sorbents.�These�materials�included,�USY�zeolite�(Zeolyst�International).�In�this�case,�

conditions� for� Fe3+� immobilization� have� been�modified�with� respect� to� the� ones� present� on�

Neamtu�et�al.,[68],�i.e.�3�cycles�of�6�hours�of�Fe�exchanging�with�an�excess�of�Fe(NO3)3�1M�at�

80C.� These� modifications� resulted� on� a� faster� Fe�exchanging� preparation� and� diminish� the�

employed� Fe(III)� solution� concentration.� In� addition,� the� whole� process� was� done� at� room�

2.�Methodology�

77�

temperature,� so,� our� process� minimizes� time,� reagents� and� energy� consumption.� Likewise,�

zeolite� Y� (namely� ZY,� Grace� Davison),� Forager� sponge� type�M� (namely� Sp,� Dynaphore� Inc.,),�

clinoptilolite� (namely� Clino;� Natural� zeolite,� origin:� Cuba)� and�montmorillonite� K�10� (namely�

MMT;� Aldrich)� were� conditioned� at� pH=4� to� be� afterwards� loaded� with� Fe(III)� using� the�

indicated�conditions�optimized�for�USY.��

The�Fe(III)�bearing�materials�were�prepared�by� ion�exchange,�contacting�each�material�

with� a� solution� of� Fe(NO3)3� 0.05M� (Panreac)� at� pH=2� at� room� temperature.� After� 1h� of�

contacting,� the�materials� were� filtered,� washed� with�milli�Q� water� and� dried� in� an� oven� at�

60ºC.� The� amount� of� Fe� on� each� supporting� material� before� and� after� the� Fe� loading� was�

measured� with� a� Field�Portable� X�Ray� Fluorescence� (FP�XRF)� equipment� Innov�X� Systems,�

model�Alpha�6500R.�

2.6.1.�FENTON�REACTION�

The� commercial� azo� dye� Acid� Red� 14� (Chromotrope� FB� 50%,� C.I.� 14720,� Aldrich)� was�

selected�as�a�model�dye,�since�it�is�a�common�textile�and�leather�industry�dye,�also�employed�

for�dying�nylon,�wool�and�silk�[69,�70,�71].��

The�decolorisation�tests�were�performed�adding�0.25g�of�catalyst�and�8.75x10-3�mmols�

of�hydrogen�peroxide� (H2O2�35%,�Fluka)� to�100�ml�of�AR14�0.05�mM�at�75ºC.�Decolorisation�

and�mineralization�of�AR14�was�measured�at��=516�nm�and���=324�nm,�respectively,�using�an�

Unicam�UV/Vis�Spectrometer�UV2�(Unicam)�at�regular�time�intervals�so�as�to�establish�kinetics�

correlations.�Finally,�the�analysis�of�the�degradation�products�in�the�final�solution�was�done�by�

GC�MS�HP6890� (column�30m�x�0.25�mm�x�250�μm)� following� the�methodology�described� in�

Zheming� et� al.� [72].� The� heterogeneous� catalysis� experiments� were� compared� with� those�

performed�in�homogeneous�catalysis�mode�using�the�same�amount�of�Fe�loaded�in�the�USY.�

In�addition,� acetic� acid� (CH3COOH�96%,�Panreac)� and�phenol� (C6H5OH�99.5%,�Panreac)�

were�selected�as�model�compounds�due�to�their�refractoriness�to�degradation�by�conventional�

oxidation�methods�[73].�The�reaction�was�performed�over�100�ml�solution�of�either�acetic�acid�

0.5%�or�phenol�0.5%�at�75C�during�1h,�using�0.25�g�of�each�Fe�loaded�supported�material�as�a�

catalyst�and�8.75x10�3�mmols�of�H2O2.�To�avoid� interferences�by�H2O2� in�COD�measurements,�

0.2g�of�MnO2�were�added�to�remove�residual�H2O2�[74].�COD�was�measured�before�and�after�

Fenton� reaction,� following� the� procedure� described� in� section� 5220� C� of� Standard�Methods�

[75].� The� amount� of� Fe� released� from� each� support,� after� the� Fenton� reaction,� was� finally�

analyzed�by�means�of�an�ICP�OES�Thermo�Iris�Intrepid�II�XLS�(USA).��

The� performance� of� the� Fe�loading� process� as� well� as� the� Fenton� reaction� was� also�

tested�in�continuous�column�process�by�using�1cm�diameter�glass�columns�filled�with�the�Fe�

2.�Methodology��

78�

exchanged�materials.� The�Fe�loading�process�was�also�done� in� countercurrent�at�2ml/min�at�

the�same�conditions�of�the�batch�process.�Washing�of�the�materials�was�done�circulating�milli�

Q�water�during�3h�and,�after�drying�in�an�oven�at�60C�overnight,�the�amount�of�Fe�loaded�into�

each�material�was�measured�by�FP�XRF.�The�Fenton� reaction� in�column�was�performed�over�

100�ml�of�AR14�0.05�mM,�100�ml�of�acetic�acid�0.5%�and�100�ml�of�phenol�0.5%�.�

2.6.2.�ARSENIC�REMOVAL�

Each�Fe(III)�bearing�material�was�slowly�added� into�a� solution�containing�1000�ppm�of�

Na2HAsO4�7H2O�(Panreac)�at�pH=4.�The�solution�was�agitated�in�a�rotating�rack�during�4h�at�30�

rpm.� The� suspension�was� filtered,� and� the� solids�were� thoroughly�washed� three� times�with�

distilled�water�and�dried�in�an�oven�under�air�at�60ºC�overnight.�Pellets�of�each�sample�were�

done�using�an�IR�press�and�sealed�between�Kapton™�tape.�The�amount�of�Fe�and�As�contained�

in�the�pellets�of�each�material�was�finally�determined�by�FP�XRF.��

2.�Methodology�

79�

2.7.REFERENCES�

[1]� ONEM� (Observatoire� Nationale� de� l’Environnement� du� Maroc).� (1997).� Monographie� locale� de�l’environnement�de�la�ville�de�Marrakech.��

[2]� Hibti,� M.;� Bouabdelli,� M.;� Mouttaqi,� A.;� Sagon,� J.P.� (1999).� L’effet� du� métamorphisme� sur� les�minéralisations� sulfurées� de� la� province� hercynienne� (Meseta� sud�occidentale,� Maroc).� Exemple� des�gisements�sulfurés�de�Hajjar�et�de�Kettara.�Chronique�Rechreche�Minière�536–537:�23–37.�

[3]�Hakkou,�R.;�Benzaazoua,�M.;�Bussière,�B.�(2008).�Acid�mine�drainage�at�the�abandoned�Kettara�mine�(Morocco):�1.�Environmental�characterization.�Mine�Water�Env.�27:�145�159.�

[4]�McMahon,�G.;�Remy,�F.�Large�mines�and�the�community:�socioeconomic�and�environmental�effects�in�Latin�America,�Canada�and�Spain.�The,World�Bank,�Washington�D.C.�USA.��

[5]� Hernández,� A.;� Jébrak,� M.;� Higueras,� P.;� Oyarzun,� R.;� Morata,� D.;Munha,� J.� (1999).� The� Almadén�mercury�mining�district,�Spain.�Miner.�Deposita�34:�539�548.��

[6]�Saupé,�F.�(1990).�The�geology�of�the�Almaden�mercury�deposit.�Econ.�Geol.�85:�482�510.�

[7]� García�Sansegundo,� J.,� Lorenzo,� S.� &� Ortega,� E.� (1987).� Mapa� geológico� de� España.� Servicio� de�Publicaciones,�Ministerio�de�Industria�y�Energía,�Madrid,�p.�60.�

[8]�Mlakar,�I.,�Drovenik,�M.,�1971.�Structural�and�genetic�particularities�of�the�Idrija�mercury�ore�deposit.�Geologija�14:�67–128.�

[9]� Gosar,�M.;� Pirc,� S.;� Bidovec,�M.� (1997).� Mercury� in� the� Idrijca� River� sediments� as� a� reflection� of�mining�and�smelting�activities�of�the�Idrija�mercury�mine.�J.�Geochem.�Explor.�58:�125–131.�

[10]�Biester,�H.;�Gosar,�M.;�Muller,�G.�(1999).�Mercury�speciation�in�tailings�of�the�Idrija�mercury�mine.�J.�Geochem.�Explor.�65:�195–204.�

[11]�Kanduc,�T.;�Kocman,�D.;�Ogrinc,�N.�(2008).�Hydrogeochemical�and�stable�isotope�characteristics�of�the� River� Idrijca� (Slovenia),� the� boundary� watershed� between� the� Adriatic� and� Black� Seas.� Aquat�Geochem�14:239–262.�

[12]�Loredo,�J.;�Ordóñez,�A.;�Baldo,�C.;�García�Iglesias,�J.�(2003).�Arsenic�mobilization�from�waste�piles�of�the�El�Terronal�mine,�Asturias,�Spain.�Geochem–Explor.�Env.�A.�3(3):229�237.��

[13]� Loredo,� J.;� Ordoñez,� A.;� Gallego,� J.R.;� Baldo,� C.;García�Iglesias,� J.� (1999).� .Geochemical�characterisation� of� mercury� mining� spoil� heaps� in� the� area� of� Mieres� (Asturias,� northern� Spain).� J�Geochem�Explor�67:�377–90.�

[14]� Luque,� C.;� Gutierrez,�M.� (2006).� La�Minería� del�Mercurio� en� Asturias:� Rasgos� Históricos,� 1st� ed.�Oviedo,�Spain.�

[15]�Berzas�Nevado,�J.J.;�Garcia�Bermejo,�L.F.;Rodríguez�Martín�Doimeadiós,�R.C.�(2003).�Distribution�of�mercury�in�the�aquatic�environmental�at�Almadén,�Spain.�Environ.�Pollut�122:�261�271.��

[16]� Higueras,� P.;� Oyarzun,� R.;� Biester,� H.;� Lillo,� J.;Lorenzo,� S.� (2003).� A� first� insight� into� mercury�distribution�and�speciation�in�soils�from�the�Almaden�mining�district,�Spain.�J.�Geochem.�Explor.�80:�95–104.�

[17]� Higueras,� P.;� Oyarzun,� R.;� Lillo,� J.;� Sánchez�Hernández,� J.C.;� Molina,� J.A.;� Esbrí,� J.M.;� Lorenzo,� S.�(2006).�The�Almadén�district�(Spain):�Anatomy�of�one�of�the�world’s�largest�Hg�contaminated�sites.�Sci.�Tot.�Environ.�356:�112–124.�

[18]� Gray,� J.� E.;� Hines,� M.E.;� Higueras,� P.;� Adatto,� I.;� Lasorsa,� B.K.� (2004).� Mercury� Speciation� and�Microbial� Transformations� in� Mine� Wastes,� Stream� Sediments,� and� Surface� Waters� at� the� Almadén�Mining�District,�Spain.�Environ.�Sci.�Technol.�38,�4285–4292.�

2.�Methodology��

80�

[19]�Horvat,�M.;�Jereb,�V.;�Fajon,�V.;�Logar,�M.;�Kotnik,�J.;�Faganeli,�J.;�Hines,�M.;�Bonzongo,�J.C.�(2002)�Mercury�distribution�in�water,�sediment�and�soil�in�the�Idrijca�and�Soca�river�systems.�Geochem.�Explor.�Environ.�Anal.�2:�287–296.�

[20]�Bernaus,�A.;�Gaona,�X.;Valiente,�M.�(2005).�Characterisation�of�Almadén�mercury�mine�environment�by�XAS�techniques.�J.�Environ.�Monit.�7:�771–777.�

[21]� Bernaus,�A.;�Gaona,� X.;� Esbri,� J.M.;�Higueras,� P.;� Falkenberg,�G.;� Valiente,�M.� (2006).�Microprobe�techniques�for�speciation�analysis�and�geochemical�characterization�of�mine�environments:�the�mercury�district�of�Almadén�in�Spain.�Environ.�Sci.�Technol.�40:�4090–4095..�

[22]�Kocman,�D.,�Horvat,�M.�&�Kotnik,� J.� (2004).�Mercury� fractionation� in�contaminated�soils� from�the�Idrija�mercury�mine�region.�J.�Environ.�Monitor.�6:�696–703.�

[23]�Biester,�H.;�Gosar,�M.;Muller,�G.�(1999).�Mercury�speciation�in�tailings�of�the�Idrija�mercury�mine�J.�Geochem.�Explor.�65:�195–204.�

[24]�Biester,�H.;�Gosar,�M.;� Covelli,� S.�Occurrence�and� Fractionation�of�Mercury� Species�Derived� from�Dumped�Mining�Residues�in�Sediments�of�the�Idrija�Mining�Area.�(2000).�Environ.�Sci.�Technol.�34:�3330–3336.�

[25]�Grimalt,�J.O.;�Ferrer,�M.;�Macpherson,�E.�The�mine�tailing�accident�in�Aznalcollar.�Sci.�Total�Environ.�242:�3�11.�

[26]�Aubert�G.�(1978).�Méthodes�d’analyses�des�sols.�G.R.D.P.:�Marseille.�France.��

[27]�Method�6200.�Field�Portable�X�Ray�Fluorescence�Spectrometry�for�the�determination�of�elemental�concentrations�in�soil�and�sediment.�http://www.epa.gov/osw/hazard/testmethods/sw846/pdfs/6200.pdf�(Last�visit:�June�2011)�

[28]�Ebinghaus,�R.;�Kock,�H.H.;�Schmolke,�S.R.�(2001).�Measurements�of�atmospheric�mercury�with�high�time�resolution:�recent�applications�in�environmental�research�and�monitoring,�Fresenius�J.�Anal.�Chem.�371:�806�815.�

[29]� Logar,� M.;� Horvat,� M.;� Akagi,� H.;� Pihlar,� B.� (2002).� Simulateneous� determination� of� inorganic�mercury�and�methylmercury�in�natural�waters,�Anal.�Bioanal.�Chem.�374:�1015�1021.�

[30]�Liang,�L.;�Lazoff,�S.;�Horvat,�M.,�Swain,�E.;�Golkeson,�J.�(2000).�Determination�of�mercury�in�crude�oil�by�in�situ�thermal�decomposition�using�a�simple�lab�built�system,�J.�Anal.�Chem.�367:�8�11.�

[31]�Welz,�B.;�Sperling,�M.�(1999).�Atomic�Absorption�Spectrometry,�3rd�ed.�Wiley�VCH.�USA.�

[32]� Puk,� R.;�Weber,� J.H.� (1994).� Critical� review� of� analytical�methods� for� determination� of� inorganic�mercury�and�methylmercury�compounds.�Appl.�Organomet.�Chem.�8:�293�302.�

[33]�Clevenger,�W.L.;�Smith,�W.B.;�Winefordner,�J.D.�(1997).�Trace�determination�of�mercury:�a�review.�Critical�Rev.�Anal.�Chem.�29:�1�26.�

[34]�McIntosh,�S.;�Baasner,�J.;�Grosser,�Z.;�Hanna,�C.�(1994).�Improving�the�determination�of�mercury�in�environmental�samples�using�a�dedicated�flow�injection�mercury�system.�At.�Spectrosc.�15:�161�163.�

[35]�Sholupov,�S.E.;�Ganeyev,�A.A.�(1995).�Zeeman�atomic�absorption�spectrometry�using�high�frequency�modulated�light�polarization.�Spectrochimica�Acta�Part�B:�Atomic�Spectroscopy�50:�1227�1236.�

[36]� Perez,� G.;� Lopez�Mesas,�M.;� � Valiente,�M.� (2008).� Assessment� of� heavy�metals� remobilization� by�fractionation:�comparison�of�leaching�tests�applied�to�roadside�sediments.�Environ�Sci�Technol�42:2309–2315.��

[37]� Chester,� R.;� Stoner,� J.H.� (1973).� Pb� in� particulates� from� the� lower� atmosphere� of� the� eastern�Atlantic.�Nature:�245:�27–�8.�

[38]� Zoller,�W.H.;� Gladney,� E.S.;� Duce,� R.A.� (1974).� Atmospheric� concentrations� and� sources� of� trace�metals�at�the�South�Pole.�Science�183:�199–�201.�

2.�Methodology�

81�

[39]� Peirson,� D.H.;� Cawse,� P.A.;� Cambray,� R.S.� (1974).� Chemical� uniformity� of� airborne� particulate�material,�and�a�maritime�effect.�Nature�251:675–�9.��

[40]�Ragaini,�R.C.;�Ralston,�H.R.;�Roberts,�N.�(1977).�Environmental�trace�metal�contamination�in�Kellogg,�Idaho,�near�a�lead�smelting�complex.�Environ.�Sci.Technol.�11:�773–781.�

[41]� Loska,� K.;� Cebula,� J.;� Pelczar,� J.;� Wiechula,� D.;� Kwapilinski,� J.� (1997).� Use� of� enrichment� and�contamination�factors�together�with�eoaccumulation�indexes�to�evaluate�the�content�of�Cd,�Cu,�and�Ni�in�the�Rybnik�water�reservoir�in�Poland.�Water�Air�Soil�Poll.�93:�347–365.�

[42]�Reimann,�C.;�Filzmoser,�P.;�Garrett,�R.G.� (2005).�Background�and�threshold:�critical�comparison�of�methods�of�determination.�Sci.�Total�Environ.��46:�1–16.�

[43]� Blaser,� P.;� Zimmermann,� S.;� Luster,� J.;� Shotyk,� W.� (2000)� Critical� examination� of� trace� element�enrichments�and�depletions� in� soils:�As,�Cr,�Cu,�Ni,�Pb,�and�Zn� in�Swiss� forest� soils.� Sci.�Total�Environ.�249:�257�280.��

[44]� Schiff,� K.C.;� Weisberg,� S.B.� (1999).� Iron� as� a� reference� element� for� determining� trace� metal�enrichment�in�Southern�California�coastal�shelf�sediments.�Mar.Environ.�Res.�48:�161�176.�

[45]�Schropp,�S.J.;�Lewis,�F.G.;�Windom,�H.L.;�Ryan,�J.D.;�Calder,�F.D.;�Burney,�L.C.�(1990)�Interpretation�of�Metal� Concentrations� is� Estuarine� Sediments� of� Florida� Using� Aluminum� as� a� Reference� Element.�Estuaries�13:�227–235.�

[46]� McMurtry,� G.M.;� Wiltshire,� J.C.;� Kauahikaua,� J.P.� (1995)� Heavy� Metal� Anomalies� in� Coastal�Sediments�of�Oahu,�Hawaii.�Pacific�Science,�49:�452–470.�

[47]�Loring,�D.H.�(1990)�Lithium:�A�New�Approach�for�the�Granulometric�Normalization�of�Trace�Metal�Data.�Marine�Chemistry,�29:�155–168.�

[48]� Loska,�K.;�Wiechua,�D.;�Pelczar,� J.� (2005).�Application�of�enrichment� factor� to�assessment�of� zinc�enrichment/depletion�in�farming�soils.�Comm.�Soil�Sci.Plant�Anal.�36:�1117—1128.�

[49]�Bourennane,�H.;�Douay,�F.;�Sterckeman,�T.;�Villanneau,�E.;�Ciesielski,�H.;�King,�D.;�Baize,�D.�(2010).�Mapping� of� anthropogenic� trace� elements� inputs� in� agricultural� topsoil� from� Northern� France� using�enrichment�factors.�Geoderma�157:�165–174.�

[50]�Pons,�X.�MIRAMON�website:http://www.creaf.uab.es/miramon�(Last�visit:�April�2010)�

[51]�Fahmy,�T.�XLSTAT�website:http://www.xlstat.com�(Last�visit:�April�2010)�

[52]� SIXPack� (2004).� SIXPack–Sam’s� Interface� for� XAS� Analysis� Package,� powered� by� IFEFFIT� 1.2.6.�Copyright�Matt�Newville,�University�of�Chicago,�USA.�

[53]� Catalano,� J.G.,� Trainor,� T.P.,� Eng,� P.J.,� Waychunas,� G.A.;Brown,� G.E.� (2005).� CTR� diffraction� and�grazing� incidence�XAFS� study�of�U(VI)� adsorption� to�a�Al2O3�and�a� Fe2O3� (1�102)� surfaces.�Geochim.�Cosmochim.�Acta�69�(14):�3555�3572.��

[54]�Slowey,�A.�J.;�Rytuba,�J.�J.;Brown,�G.�E.�(2005).�Speciation�of�Mercury�and�Mode�of�Transport�from�Placer�Gold�Mine�Tailings.�Environ.�Sci.Technol.�39:�1547–1554.�

[55]�Arai,� J.;� Lanzirotti,�A.;� Sutton,� S.R.;�Newville,�M.;�Dyer,� J.;Sparks,�D.L.� (2006).� patial� and� temporal�variability�of�arsenic�solid�state�speciation� in�historically� lead�arsenate�contaminated�soil.�Environ.�Sci.�Technol.�40:�673–679.�

[56]�Malinowski,�E.R.�(1991).�Factor�Analysis�in�Chemistry,�2nd�ed.�Wiley,�USA.�

[57]�Ressler,�T.;�Wong,�J.;�Roos,�J.;Smith,�I.L.�(2000).�Quantitative�speciation�of�Mn�bearing�particulates�emitted� from� autos� burning� (methylcyclopentadienyl)manganese� tricarbonyl�added� gasolines� using�XANES�spectroscopy.�Environ.�Sci.�Technol.�34,�950–958.�

[58]�Wasserman,�S.R.;�Allen,�P.G.;�Shuh,�D.K.;�Bucher,� J.J.;�Edelstein,�N.M.� (1999).�EXAFS�and�Principal�Component�Analysis:�A�New�Shell�Game.�J.�Synchrotron�Rad.�6:�284–286.�

2.�Methodology��

82�

[59]� Rehr,� J.J.;� Albers,� R.C.;� Zabinsky,� S.I.� (1992).� High�order�multiple�scattering� calculations� of� x�ray�absorption�fine�structure.�Phys.�Rev.�Lett.�69,�3397–3400.�

[60]�Klementiev,�K.V.�(2001).�VIPER�for�Windows,�freeware;�K.V.�Klementev,�J.�Phys.�D:�Appl.�Phys.�34:�209�17.�

[61]� Mazuelos,� A.;� Carranza,� F.;� Palencia,� I.;� Romero,� R.� (2000).� High� efficiency� reactor� for� the�biooxidation�of�ferrous�iron.�Hydrometallurgy�58(3):�269�275.�

[62]� Carranza,� F.;� Iglesias,� N.;� Romero,� R.;� Palencia,� I.� (1993).� Kinetics� improvement� of� high�grade�sulfides�bioleaching�by�effects�separation.�FEMS�Microbiol.�Rev.�11(1�3):�129�38.�

[63]� Barnard,� K.R.;� Shiers,� D.W.� (2011).� Mechanism� and� rate� of� butyl� phosphinate� formation� from�reaction�of�phosphinic�acid�(Cyanex�272)�and�tributyl�phosphate.�Hydrometallurgy�106:�76�83.��

[64]�Ferreira,�A.N.;�Agarwal,�S.;�Machado,�R.M.;�Gameiro,�M.L.F.;�Santos,�S.M.M.;�Reis,�M.T.A.;� Ismael,�M.R.C.;�Correia,�J.N.;�Carvalho,�J.M.R.�(2010).�Extraction�of�copper�from�acidic�leach�solution�with�Acorga�M5640�using�a�pulsed�sieve�plate�column.�Hydrometallurgy�104:�66�75.�

[65]� Gameiro,� M.L.F.;� Machado,� R.M.;� Ismael,� M.R.C.;� Reis,� M.T.A.;� Carvalho,� J.M.R.� (2010).� Copper�extraction�from�ammoniacal�medium�in�a�pulsed�sieve�plate�column�with�LIX�84�I.�J.�Hazard.�Mater.�183:�165�175.�

[66]� Sole,� K.C.;� Feather,� P� A.M.;� Cole,�M.� (2005).� Solvent� extraction� in� southern� Africa:� An� update� of�some�recent�hydrometallurgical�developments.�Hydrometallurgy�78:�52�78.�

[67]�Ritcey,�G.M.�(2006).�Solvent�extraction�principles�and�applications�to�process�metallurgy,�Vol.1,�336�348.�G.M.�Ritcey�&�Associates.�Canada.�

[68]�Neamtu,�M.;�Zaharia,�C.;�Catrinescu,�C.;�Yediler,�A.;�Macoveanu,�M.;�Kettrup,�A.�(2004).�Fe�loaded�Y�zeolite�as�catalyst�for�wet�proxide�oxidation�of�reactive�azo�dye�Procion�Marine�H�EXL.�Applied�Catalysis�B:�Environmental�48:�287�294.�

[69]�Shen,�Z.;�Wang,�W.;�Jia,�J.;�Feng,�X.;�Hu,�W.;�Peng,�A.�(2002).�Catalytically�assisted�electrochemical�oxidation�of�dye�Acid�Red�B.�Water�Environ.�Res.�74:�117�121.��

[70]�Lin,�J.J.;�Zhao,�X.S.;�Liu,�D.;�Yu,�Z.G.;�Zhang,�Y.;�Xu,�H.�(200).�The�decoloration�and�mineralization�of�azo�dye�C.I.�Acid�Red�14�by� sonochemical�process:� rate� improvement�via�Fenton's� reactions.� J�Hazard�Mater.�157:�541�6.�

[71]�Daneshvar,�N.;�Khataee,�A.R.�(2006).�Removal�of�Azo�Dye�C.I.�Acid�Red�14�from�Contaminated�Water�using� Fenton,� UV/H2O2,� UV/H2O2/Fe(II),� UV/H2O2/Fe(III)� and� UV/H2O2/Fe(III)/Oxalate� Processes:� A�Comparative�Study.�J.�Environ.�Sci.�Health,�Part�A:,41:�315–328.�

[72]�Zheming,�Xiaofang,�Yanming,�Xiuling.�(2005).�Dual�electrodes�oxidation�of�dye�wastewater�with�gas�diffusion�cathode.�Environ.�Sci.�Technol.�39:�1819�1826.��

[73]�Sinha,�A.;�Chakrabarti,�S.;�Chaudhuri,�B.;�Bhattacharjee,�S.;�Ray,�P.�(2007).�Oxidative�Degradation�of�Strong�Acetic�Acid�Liquor�in�Wastewater�Emanating�from�Hazardous�Industries.�Ind.�Eng.�Chem.�Res.�46:�3101�3107.�

[74]�Pardieck,�D.;�Bouwer,�E.�J.;�Stone,�A.�T.�(1992).�Hydrogen�peroxide�use�to�increase�oxidant�capacity�for�in�situ�bioremediation�of�contaminated�soils�and�aquifers.�J.�Contam.�Hydrol.�9:�221�242.�

[75]�APHA�AWWA�WPCF�(1995).�Standard�Methods�for�the�Examination�of�Water�and�Wastewater.�19th�ed.,�American�Public�Health�Association�USA.�

83�

3��

RESULTS�AND�DISCUSSION��MINE�SITES�CHARACTERIZATION ......................................................................................................85�

3.1.�HEAVY�METAL�CONTAMINATION�AND�MOBILITY�AT�THE�DRAA�LASFAR�MINE�AREA.....................85�3.2.�CHARACTERIZATION�OF�KETTARA,�SIDI�BOU�OTHMANE�AND�BIR�NEHASS�MINE�AREAS ...............92�3.3.� XANES� SPECIATION� OF�MERCURY� IN� THREE� MINING� DISTRICTS:� ALMADEN� (SPAIN),� ASTURIAS�(SPAIN)�AND�IDRIJA�(SLOVENIA) ...........................................................................................................104�

��REMEDIATION�TECHNIQUES�OF�INDUSTRIAL�CONTAMINATED�WATERS...............................111�

3.4.� EXTRACTANT� AND� SOLVENT� SELECTION� TO� RECOVER� ZINC� FROM�A�MINING� EFFLUENT:� FROM�LABORATORY�SCALE�TO�PILOT�PLANT ..................................................................................................111�3.5.� Fe�LOADED� MATERIALS� FOR� THE� REMEDIATION� OF� ORGANIC� AND� INORGANIC� POLLUTED�WATERS ................................................................................................................................................119�3.6.�REFERENCES...................................................................................................................................128�

84�

3.�Results�and�discussion�

85�

�The�current� chapter�describes� the� results�obtained� from�the� studies� carried�out� in� the�

present� thesis� including� the�characterization�of� samples� from�polluted�mine�areas�as�well� as�

the� study� of� processes� applied� to� remediate� wastewaters� containing� either� organic� or�

inorganic�pollutants,�i.e.,�the�basic�and�scale�up�process�for�the�valorization�on�an�acidic�mine�

water� residue.� Accordingly� to� this� content,� results� include� two� main� sections:� Mine� sites�

Characterization�and�Remediation�Technologies.�

MINE�SITES�CHARACTERIZATION�

The�physico�chemical�parameters�and�the�concentration�of�heavy�metals�of� four�mines�

located�in�Marrakech�region�(Morocco)�were�evaluated�to�characterize�the�area�and�its�level�of�

contamination,�as�well�as� its� toxicological� risk�derived�from�target�heavy�metals�mobility.�On�

the� other� hand,� the� results� focused� in� the� speciation� of�mercury� samples� from� three�main�

European�mercury�mines�are�also�outlined�in�the�following�section.��

3.1.� HEAVY� METAL� CONTAMINATION� AND� MOBILITY� AT� THE� DRAA� LASFAR� MINE�

AREA�(SEE�ANNEX�I)��

The�main�results�of�a�first�insight�into�the�Draa�Lasfar�mine�(Marrakech)�are�outlined�in�

this� section.� To� characterize� the� degree� of� pollution� in� this� mine� area,� the� present� study�

includes� the� geochemical� distribution� maps� of� the� pollutants,� the� particle� size� effects� (i.e.,�

intraparticle� concentration� affecting� metals� distribution)� and� the� mobility� of� the� main�

pollutants�by�employing�single�leaching�tests�to�predict�the�risk�of�their�mobility.�

3.1.1.�PHYSICO�CHEMICAL�PARAMETERS�

The�results�obtained�for�the�soil�pH,�electrical�conductivity�(EC),�loss�on�ignition�(LOI)�and�

CaCO3�content�measurements� corresponding� to�Draa�Lasfar�mine�area�are� resumed� in�Table�

3.1.��

It� is�revealed�that,� in�general,�all�sampled�points�showed�neutral�to�alkaline�pH�ranging�

from�7�to�9,�similar� to�pH�of�background�samples�with�the�exception�of�a�very�acidic�sample�

corresponding�to�sample�#48�with�pH�3.5�just�beside�the�mine�site,�mainly�related�to�deposits�

3.�Results�and�discussion��

�86�

of�sulfidic�residues,�which�by�oxidation�and�formation�of�sulfuric�acid,�can�cause�such�decrease�

of�the�pH.��

Table�3.1.�Summary�of�physico�chemical�parameters�for�Draa�Lasfar�samples�� � pH� CE�(μS/cm)� LOI�(g/kg)� CaCO3�(mg/g)�

min� 3.5�(#48)� 96.0�(#53)� 12.7�(#2)� 7.9�(#50)�

max� 9.6�(#2)� 14,160�(#31)� 75.8�(#45)� 209.9�(#26)�

mean� 8.3� 1463� 32.6� 43.2�

Mine�area�samples�

(81�samples)�st.�dev.� 0.7� 2170� 14� 32�

Mean�� 8.5� 136.3� 30.7� 24.9�Background�samples��

(4�samples)�St.�dev.� 0.5� 20.4� 6.8� 2.6�

EC� showed� higher� variability� than� pH,� with� values� ranging� from� 100� to� 15,000� μS/cm�

(Table�3.1)�which,�in�general,�are�in�accordance�with�previous�studies�performed�on�Morocco�

soils�[1].�Likewise�pH,�the�sample�with�higher�EC�value�(sample�#31)�is�located�just�beside�the�

mine� site� and� corresponds� to� the� sample�with� the�higher� amounts�of�metals� in� the� area.� In�

addition,�the�values�obtained�for�the�mine�area�samples�are�significantly�higher�than�those�of�

the�background�samples,�thus�indicating�high�amounts�of�labile�ions�close�to�the�mine�site�as�

consequence�of�mining�activities.��

LOI�values�for�mine�area�and�background�samples�(Table�3.1)�values�are�similar�and�quite�

homogeneous�(around�30�g/kg)�although�some�specific�points�have�LOI�reaching�76�g/kg�due�

to�some�close�localized�agricultural�activities.��

The�observed�carbonate�content�ranged�from�10�to�210�mg.g�1�(Table�3.1)�although�the�

majority� of� the� samples� present� similar� CaCO3� content� to� background� samples.� The� highest�

content� of� carbonates� is� observed� for� samples� #26� and� #71,� located� 400�m� away� from� the�

mine� site.� Together�with� basic� pH� values,� the� presence� of� carbonates� in� the� soil� lead� to� an�

increase� in� the� retention� of� heavy� metals,� mainly� as� carbonate� salts� as� a� consequence� of�

related�hydroxyoxides�precipitation,�the�principal�retention�mechanism�for�heavy�metals�[2].��

3.1.2.�HEAVY�METAL�CONCENTRATION�IN�THE�MINE�AREA�

Regarding�the�metal�concentration�measured�by�FP�XRF�and�its�corresponding�calculated�

CER� values,� elements� have�been� classified� as� either� pollutants� (when� the� samples� have�CER�

values�above�5)�or�lithogenic�elements�for�CER�values�below�5.�In�this�sense,�As,�Cu,�Pb�and�Zn�

can� be� classified� as� pollutants� whereas� the� rest� of� elements� measured� are� considered�

lithogenic�components�of�the�soil�(Table�3.2).��

3.�Results�and�discussion�

87�

Table�3.2.�Minimum,�maximum�and�mean�concentration�and�CER�values�for�the�measured�elements�in�mine�area�and�background�samples�

� � �Mine�area�samples�

Background�samples�

� � � Min� Max� Mean� Mean�Conc.� 8.9� 3108�(#48)� 70±400� 13±5�

As�CER� 0.4� 20.0� 2.2±4� 1.0±0.3�Conc.� 16.6� 172�(#45)� 35±20� 34±8�

Cu�CER� 0.3� 5.9� 1.1±1� 1.0±0.3�Conc.� 6.7� 2309�(#48)� 70±300� 17±3�

Pb�CER� 0.3� 45.9� 3±6� 1.0±0.1�Conc.� 30� 1114�(#45)� 125±200� 82±7�

Pollutants�

Zn�CER� 0.3� 13.6� 2±2� 1.0±0.1�Conc.� 221.5� 541�(#17)� 389±70� 449±60�

Ba�CER� 0.4� 1.4� 0.9±0.3� 1.0±0.1�Conc.� 19632� 121652�(#48)� 32721±10000� 35555±3000�

Fe�CER� 0.4� 4.8� 0.9±0.5� 1.0±0.1�Conc.� 6071� 32976�(#81)� 23327±5000� 38244±1500�

K�CER� 0.3� 1.4� 0.8±0.3� 1.0±0.1�Conc.� 290� 1119�(#21)� 607±150� 699±50�

Mn�CER� 0.4� 2.0� 0.9±0.3� 1.0±0.1�Conc.� 47� 106�(#10)� 76±13� 81±6�

Rb�CER� 0.5� 1.7� 1.0±0.3� 1.0±0.1�Conc.� 86� 322�(#26)� 144±40� 131±9�

Sr�CER� 0.5� 3.1� 1.1±0.5� 1.0±0.1�Conc.� 2802� 5860�(#17)� 4460±700� 5229±600�

Ti�CER� 0.4� 1.3� 0.9±0.2� 1.0±0.1�

Lithogen

ic�com

pone

nts�

Zr� Conc.� 112� 335�(#67)� 215±50� 210±17�Conc�is�given�in�mg/kg.�In�parenthesis�is�given�the�sample�with�maximum�concentration.��

3.1.3.�GIS�CONTOUR�MAPS�OF�THE�MAIN�POLLUTANTS�

Although�arsenic�can�not�be�considered�a�metal�it�will�be�referred�to�as�belonging�to�the�

heavy�metals�group� for� reasons�of� convenience.�Arsenic�distribution�of�CER�values�using�GIS�

contour� maps,� depicted� in� Figure� 3.1,� showed� two� hot� spots� beside� the� mine� site�

corresponding� to� samples� #48� (3108� ppm,� CER=280)� and� #31� (203� ppm,� CER=19.4).� Sample�

#48� represents� an� arsenic� concentration� 100� fold� higher� than� background� levels� which�

indicates�that�remediation�is�mandatory�for�this�specific�area.�At�increasing�distances�from�the�

mine�site,�arsenic�concentration�decreases�to�values�similar�to�background�samples,�except�for�

samples� #45� (203� ppm,� CER=15.9)� and� #46� (125� ppm,� CER=9.2).� An� anomalous� result� is�

observed� for� sample�#21� (72�ppm,�CER=7.1).� This� sample� is� located�at� the�other� side�of� the�

river�creek�and�its�arsenic�concentration�is�higher�than�neighboring�samples.�It�is�probably�due�

to� a� waste� deposit� when� mining� was� active.� Given� the� proximity� of� this� area� to� the� creek�

waters,�it�is�foremost�to�monitor�this�area.��

Copper�CER�distribution�map�along�the�mining�area�(Figure�3.2)�followed�a�trend�similar�

to�the�one�expressed�by�arsenic,�being�the�samples�close�to�the�mine�site�the�ones�with�higher�

3.�Results�and�discussion��

�88�

copper�CER�values.�Likewise�arsenic,�sample�#21�(51�ppm,�CER=1.9)�located�at�the�other�side�of�

the�river�basin,�has�high�copper�concentration�despite�being�far�from�the�mine�area.�This�can�

be�explained�by�the�fact�that�minor�amounts�of�Cu�are�found�typically�adsorbed�in�arsenopyrite�

(FeAsS)�ores.��

The�lead�distribution�around�the�mine�(Figure�3.3)�showed�four�hot�spots�located�around�

samples�#31� (180�ppm,�CER=13.0),�#45� (770�ppm,�CER=45.9),�#48� (2310�ppm,�CER=130)�and�

#58�(420�ppm,�CER=30).�It�is�also�noteworthy�to�highlight�sample�#21�(62�ppm,�CER=4.6)�given�

its�high�CER�values�and�proximity�to�creek�waters.��

CER�distribution�map�for�Zn�(Figure�3.4)�followed�the�same�trend�as�the�one�depicted�by�

Pb�with�4�hot� spots� located�at� samples�#20� (630�ppm,�CER=8.5),�#45� (1110�ppm,�CER=13.6),�

#48�(30�ppm,�CER=10.8)�and�#58�(930�ppm,�CER=10.8).��

Thus,� taking� into� account� the� GIS� maps� obtained� by� using� CER� values� for� the� main�

pollutants�of�the�Draa�Lasfar�mine�area,�it�can�be�stated�that�the�most�polluted�sites�are�found�

beside� the�mine� site� towards� the� river� creek�whilst� samples� closed� to�Koudiyat�hill� reported�

values�similar� to�background.�Hence,� the�pollution�over� the�mine�area�of�Draa�Lasfar�can�be�

mainly�attributed�to�weathering�effects�and�the�topography�of�the�terrain�that�facilitates�the�

disposal� of�mine� residues� towards�descendent� areas� such�as� the� river� creek� and� reduce� the�

deposition�on�elevated�areas�such�as�hills.��

Other� measured� elements� showed� CER� values� close� to� background� samples,� hence�

considered�as�lithogenic�components�of�the�soil.�This�group�of�elements�is�formed�by�Ba,�Fe,�K,�

Rb,�Sr,�Ti�and�Zr�(Table�3.2).�In�this�sense,�the�mean�concentration�of�Ba,�Fe,�K,�Mn,�Rb,�Sr,�Ti�

and� Zr� is� similar� to� background� samples� and� therefore� its� correspondent� CER� values� range�

between�0�and�2,�excepting�sample�#48.�Such�samples�contain�high�Fe�content�that,�given�the�

high� As� content� can� be� related� to� an� arsenopyrite� (FeAsS)� deposit.� Thus,� no� anthropogenic�

enhancement�of�these�elements�is�observed.��

Finally,�some�specific�samples�present�extremely�high�concentration�on�some�elements.�

High� sulfur� concentrations�were� found� in� samples� #19� (18400� ppm),� #31� (14500� ppm),� #33�

(15500�ppm),�#45�(36800�ppm),�#48�(113700�ppm),�#58�(5300�ppm),�#59�(14800�ppm)�and�#70�

(32400�ppm)�also�related�to�arsenopyrite�deposits�thus�supporting�the�arsenopyrite�nature�of�

the�mineral�ores�extracted.��

3.�Results�and�discussion�

89�

Background�samples

As

Mine

area

Scale�1:25000

CER=0

CER=6

CER=3

Mine

area

Scale�1:25000 Background�samples

Cu

CER=0

CER=6

CER=3

Figure� 3.1.� GIS� contour� map� of� arsenic�distribution�around�the�mine�area.�

Figure� 3.2.� GIS� contour� map� of� copper�distribution�around�the�mine�area.�

Mine

area

Scale�1:25000 Background�samples

Pb

CER=0

CER=20

CER=10

Scale�1:25000

Background�samples

Zn

CER=0

CER=20

CER=10

Mine

area

Figure�3.3.�GIS�contour�map�of�lead�distribution�around�the�mine�area.�

Figure� 3.4.� GIS� contour� map� of� zinc�distribution�around�the�mine�area.�

3.1.4.�EFFECT�OF�PARTICLE�SIZE�AND�MOBILITY�

Samples�with�high�CER�values�and/or�with�spatial�significance�were�selected�to�study�the�

effect�of�particle�size�and�the�mobility�of�pollutants.�Thus,�samples�#20,�#31,�#46,�#48,�#58�and�

#70�were�selected.�Related�results�are�collected�in�Table�3.3.��

The�results�obtained� for� target� fractions�show�a�generalized� increase�on�As,�Pb�and�Zn�

concentrations�after�milling�the�samples�below�100�μm.�In�this�sense,�samples�#20,�#31,�#46,�

#48�and�#58�had�an�enrichment�on�As,�Pb�and�Zn�when�milled.�Therefore,�it�can�be�stated�that,�

3.�Results�and�discussion��

�90�

in�general,�these�elements�are�part�of�the�particle�core.�On�the�other�hand,�a�decrease�of�the�

concentration�on�copper�as�the�soil�is�milled�is�observed�thus�indicating�copper�is�adsorbed�at�

the� surface� of� the� soil� particles� instead� of� forming� part� of� the� mineral� ore� revealing� an�

anthropogenic�input�of�copper.��

Regarding� the� results� obtained� for� the� pollutants�mobility� in� selected� samples,� also� in�

Table� 3,� it� can� be� observed� arsenic,� lead� and� zinc� in� the� mobile� phase� of� some� samples�

(samples� #31,� #46).� Sample� #46� shows� the� highest� mobility� of� pollutants,� its� pH� is� alkaline�

(pH=8.1),�with�high�EC�(EC=2151�μS/cm)�and�relatively�high�carbonate�content�([CaCO3]=58.4�

mg/g).� In�these�conditions,�mobility� is�not�specially� favored�although�given�the�relatively� low�

LOI�value�(LOI=39.3�g/kg)�it�can�be�supposed�that�this�factor�enables�the�availability�of�cations�

from� the�mine� ore� to� the�mobile� phase.� Therefore� it� can� be� stated� that� the� leading� factor�

regarding� mobility� of� the� samples� at� Draa� Lasfar�mine� area� is� concentration� of� metals� and�

organic�matter�(based�on�LOI�determinations).��

It�is�also�important�to�highlight�sample�#48,�which�accounts�for�being�the�most�polluted�

and�acidic� sample� (pH=3.5),�with�high�EC� (EC=4873�μS/cm)�and� low�carbonate�content� (25.8�

mg/g).� According� to� literature� [3]� these� conditions� favor� the� availability� of� cations,� however�

sample�#48�has�also�a�high�organic�matter�content�as�indicates�its�LOI�value�(LOI=56.0�mg/g),�

which� benefits� the� adsorption� of� soil� labile� ions,� thus� explaining� the� relatively� low�mobility�

observed.�That�leads�to�lower�its�environmental�risk�when�considering�only�total�concentration�

values.� In� this� sense,� soil� organic� matter� is� considered� one� of� the� primary� immobilizing�

processes�for�trace�and�toxic�pollutants�[4].�

In�this�sense,�GIS�contour�maps�of�pollutants�using�CER�data�have�been�a�valuable�tool�to�

characterize� pollutants� distribution� around� the� mine� area� and� to� determine� sources� of�

contamination.��

3.�Results�and�discussion�

91�

Table�3.3.�Particle�size�effects�and�mobility�assays.�Concentration�of�pollutants�at�fraction�<�2�mm�and�fraction�<100�μm�and�amount�of�pollutants�mobile�

SAMPLE� � As� Cu� Pb� Zn�2mm�(mg/kg)� 125� 80� 55� 628�100μm�(mg/kg)� 167� 72� 66� 713�#20�Mobility�(mg/L)� <0.5�� <0.5� <0.5�� <0.5�2mm�(mg/kg)� 72� 51� 62� 144�100μm�(mg/kg)� 67� 48� 61� 150�#21�Mobility�(mg/L)� <0.5� <0.5� <0.5� <0.5�2mm�(mg/kg)� 203� 43� 180� 481�100μm�(mg/kg)� 268� 77� 313� 734�#31�Mobility�(mg/L)� 49� 2� 6� 18�2mm�(mg/kg)� 125� 60� 375� 774�100μm�(mg/kg)� 172� 59� 477� 933�#46�Mobility�(mg/L)� 54� 1� 17� 23�2mm�(mg/kg)� 3,108� 144� 2,309� 631�100μm�(mg/kg)� 3,569� 167� 2,614� 704�#48�Mobility�(mg/L)� 5� 1� <0.5� 4�2mm�(mg/kg)� 113� 71� 425� 925�100μm�(mg/kg)� 149� 77� 537� 1,087�#58�Mobility�(mg/L)� <0.5� <0.5� <0.5� <0.5�2mm�(mg/kg)� 15� 33� 24� 97�100μm�(mg/kg)� 15� 50� 20� 91�#70�Mobility�(mg/L)� 29� <0.5� <0.5� <0.5�

3.�Results�and�discussion��

�92�

3.2.�CHARACTERIZATION�OF�KETTARA,�SIDI�BOU�OTHMANE�AND�BIR�NEHASS�MINE�

AREAS�

Following�the�methodology�employed�for�the�characterization�of�Draa�Lasfar�mine�area,�

three�additional�abandoned�mines�of� the�area�at�sites�of:�Kettara,�Sidi�Bou�Othmane�and�Bir�

Nehass�were�characterized�for�their�potential�pollutant�impact.��

3.2.1.�PHYSICO�CHEMICAL�CHARACTERIZATION�

The�results�obtained�for�the�measured�physico�chemical�parameters�are�summarized�in�

Table�3.4.�Samples�have�been�distinguished�between�residues�samples,�the�samples�taken�at�

specific�points�were�residues�were�stored�and�mine�area�samples,�sampled�at�regular�distances�

from�the�mine�site.��

Table�3.4.�Summary�of�pH,�conductivity�and�carbonate�content�for�the�samples�taken�at�Kettara,�Sidi�Bou�Othmane�and�Bir�Nehass�

� � �pH� EC�(μS/cm)� LOI�(mg/g)� CaCO3�(mg/g)�

Min� 2.0�(S10)� 0.1�(S37)� 6�(KS15)� 2.0�(S2,�S9)�Max� 8.2�(S46)� 4,900�(S10)� 71�(KS41)� 167�(S48)�

Mine�Area�(58�samples)�

Mean� 6.3±1.9� 680±900� 34±20� 37±60�Min� 2.0�(R2)� 2,686�(R1)� 5�(KR1)� 1.25�(R3)�Max� 2.4�(R3)� 7,295�(R7)� 11�(KR6)� 7.5�(R4)�

Residues�(7�samples)�

Mean� 2.2±0.2� 3,618±1700� 7.6±2� 4±3�Min� 7.8� 153� 41� 217�Max� 8.5� 100� 53� 275�

Kettara�

Background�(3�samples)�

Mean� 8.1±0.4� 126±30� 48±7� 240±30�

Min� 7.1�(S1)� 226�(S30)� 37.5�(S2)� 3.7�(S20)�Max� 8.1�(S21)� 823�(S16)� 93.7�(S1)� 439�(S12)�

Mine�Area�(30�samples)�

Mean� 7.8±0.2� 349±130� 63.4±13� 690±1100�Min� 3.2�(R20)� 159�(R17)� 23�(R17)� 7.5�(R17)�Max� 8.4�(R9)� 5,469�(R12)� 50�(R12)� 299.6�(R16)�

Residues�(17�samples)�

Mean� 7.3±1.1� 1,898±1200� 35±8� 78±70�Min� 6.8� 324� 30.5� 2�Max� 8.0� 670� 35.3� 10�

SB�Othman

e�

Background�(3�samples)�

Mean� 7.6±0.8� 447±200� 33±2� 7±5�

Min� 8.0�(S20)� 58�(S32)� 23.8�(S31)� 3�(S31)�Max� 8.4�(S11)� 2,005�(S19)� 46.5�(S8)� 165�(S1)�

Mine�Area�(33�samples)�

Mean� 7.7±0.5� 250±400� 31±5� 24±40�Min� 1.8�(R1)� 2,119�(R3)� 18.4�(R1)� 5.0�(R1)�Max� 3.5�(R3)� 9,011(R1)� 35.4�(R2)� 24�(R3)�

Residues�(4�samples)�

Mean� 2.9±0.7� 4,135±3000� 29±7� 13±8�Min� 7.4� 87� 29.9� 3�Max� 7.7� 94� 36.7� 4�

Bir�Neh

ass�

Background�(3�samples)�

Mean� 7.5±3� 91±4� 33±4� 3.7±0.3�

3.�Results�and�discussion�

93�

The�mean�pH�value�obtained�for�Kettara�mine�area�samples�is�slightly�acidic�as�a�result�of�

the�oxidation�of�pyrite�that�releases�sulfuric�acid�and�lowers�the�pH.�In�this�sense,�23�out�of�58�

samples�have�pH�below�7�and�the�pH�is�comprised�between�2�and�4�in�13�samples.�For�the�rest�

of� the� samples� the� pH� ranged� from� 7� to� 9,� which� is� a� normal� pH� also� observed� for� the�

background�data.�On�the�other�hand,�samples�taken�at�the�residues�deposits,�are�very�acidic,�

all�of�them�in�the�range�of�pH�of�2�2.4.�These�low�pH�values�are�also�related�to�the�oxidation�of�

high�contents�of�sulfur.�Unlike�Kettara,�Bir�Nehass�and�Sidi�Bou�Othmane�mine�area�samples�

have�neutral�to�alkaline�pH�for�the�majority�of�the�samples,�being�these�values�are�similar�to�

the�background�samples.�Regarding�the�sampling�corresponding�to�the�residues,�pH�at�Kettara�

and� Bir� Nehass� residues� were� strongly� acidic� (pH� from� 2� to� 3.5)� related� to� pyrite� deposits�

whereas�all�Sidi�Bou�Othmane�residues�(except�one�sampling�point)�were�quite�basic�with�a�pH�

similar�to�background�samples.��

In� this� sense,� the�most� acidic� samples� have�also� the�higher� EC� values.� This� correlation�

between�low�pH�and�high�EC�value�can�be�explained�by�the�presence�of�high�amounts�of�sulfur�

ions� that� causes� an� increase� of� the� EC� and� by� oxidation� lowers� the� pH� by� acid� sulfuric�

formation.�In�this�regard,�Kettara�mine�area�samples�had�higher�EC�than�the�mine�area�samples�

of�SB�Othmane�and�Bir�Nehass,�and�moreover,�the�EC�found�for�the�residues�were�much�higher�

than�mine�area�samples.���

LOI�values�are�similar�for�mine�area,�residues�and�background�samples�for�all�the�studied�

mines�(around�30�g/kg)�although�some�specific�points�of�Sb�Othmane�mine�areas�samples�had�

higher�LOI�values�mainly�related�to�some�close�localized�agricultural�activities.��

Regarding�the�carbonate�content�it�can�be�stated�that�soils�with�a�pH�of�7.5�and�higher�

generally�have�a�high�calcium�carbonate�content.� In� this� sense,�as� stated� in�previous�section�

3.1.1.,� alkaline� soils� together� with� high� amounts� of� organic� matter� and� the� presence� of�

carbonates�increase�the�retention�of�heavy�metals�in�soils.�

3.2.2.�HEAVY�METAL�CONCENTRATION�IN�THE�MINE�AREA�

Taking� into�account� the�CER�values�of� the�metals�measured� for�all� three�mines�and�by�

comparing�samples,� residues�and�backgrounds�values,� it� is�observed�As,�Cu,�Pb�and�Zn�to�be�

the�main�pollutants�of� the� studied�mine�areas� since�most�of� the� samples�exceed�CER�values�

above�5,�thus�indicating�the�significant�anthropogenic�contribution�(Table�3.5).��

3.�Results�and�discussion��

�94�

Table�3.5.�Summary�of�As,�Cu,�Pb,�Zn�and�Zr�concentration�and�CER�values�for�the�samples�and�residues�of�Kettara,�Sidi�Bou�Othmane�and�Bir�Nehass�

� � � � As� Cu� Pb� Zn�Min� 10� 27� 17� 49�Max� 237� 1362� 486� 243�Conc.�Mean� 34±40� 256±300� 76±100� 106±40�Min� 0.5� 0.35� 0.7� 0.7�Max� 18.8� 27.6� 34� 4.2�

Mine�Area�(58�samples)� CER�

Mean� 2±3� 4±6� 4±6� 1.4±0.6�Min� 28� 364� 105� 93�Max� 104� 2113� 349� 337�Conc.�Mean� 64±40� 1,287±500� 233±100� 197±100�Min� 3.6� 16.5� 14� 2.6�Max� 7.8� 74� 36� 10.0�

Residues�(7�samples)�

CER�Mean� 6±2� 38±17� 23±7� 5±3�Min� 11� 44� 12� 49�Max� 24� 52� 15� 76�

Kettara�

Background�(3�samples)�

Conc.�Mean� 16±7� 48±4� 14±2� 60±14�Min� 10� 25� 23� 101�Max� 112� 46� 6706� 36267�Conc.�Mean� 27±30� 31±8� 1,467±2000� 5,018±9000�Min� 0.6� 0.9� 0.8� 0.6�Max� 11� 1.7� 391� 414�

Mine�area�(30�samples)�

CER�Mean� 2±3� 1.1±0.1� 62±100� 43±90�Min� 6� 109� 76� 85�Max� 203� 36140� 13044� 57380�Conc.�Mean� 98±60� 10,653±16000� 4,112±4000� 19,238±1500�Min� 4.6� 6.7� 2.7� 0.5�Max� 26� 1405� 786� 638�

Residues�(17�samples)�

CER�Mean� 14±6� 380±20� 249±200� 242±180�Min� 13� 22� 30� 144�Max� 17� 29� 37� 167�

Sidi�Bou

�Othman

e�

Background�(3�samples)�

Conc.�Mean� 15±2� 27±8� 32±4� 156±12�Min� 8� 24� 19� 86�Max� 113� 46� 1495� 29732�Conc.�Mean� 23±20� 29±6� 148±300� 1,744±6000�Min� 0.5� 0.9� 1.1� 1.2�Max� 11� 2.1� 94� 366�

Mine�area�(33�samples)�

CER�Mean� 2±2� 1.2±0.4� 8±20� 22±70�Min� 92� 52� 1776� 8521�Max� 760� 310� 29559� 23309�Conc.�Mean� 282±300� 149±140� 9,423±13000� 15,075±6000�Min� 11� 3.3� 164� 141�Max� 70� 14.8� 2010� 350�

Residues�(4�samples)�

CER�Mean� 29±30� 8±6� 803±1000� 273±80�Min� 9� 22� 19� 68�Max� 16� 29� 23� 99�

Bir�Neh

ass�

Background�(3�samples)�

Conc.�Mean� 13±4� 27±8� 19±4� 84±16�

M.A.�Mining�Area,�Bkg.�background,�R.�Residue.� (The�number�of� samples� is� given� in�Table�3.4).�

After� the�evaluation�of� the�results�of� the�main�pollutants�concentration� (Table�3.5)� it�can�be�

considered� that� the� amount�of�mineral� extracted�per�day� and� the� exploitation� time�of� each�

mine�affects�the�level�of�contamination.�In�this�sense,�Kettara�mine�site�was�less�exploited�and�

3.�Results�and�discussion�

95�

hence�less�polluted�than�SB�Othmane�and�Bir�Nehass�mine�sites�both�exploited�within�a�period�

of�approximately�30�years�and�with�exploitation�outputs�of�115�and�90�tons�per�day.��

3.2.3.�APPLICATION�OF�CHEMOMETRICS�

Chemometrics� tools� have� been� applied� to� establish� relationships� between� the� three�

mines�as�well� as�patterns�and� spatial�distribution�of� the�main� target�pollutants� identified.� In�

this� sense,� box�plot� figures� of� the�data� for� the�mine� area� samples� and� the� residues� of� each�

mine�can�provide�a�more�understanding�representation�of�the�obtained�data.��

As�depicted� from�Box�plots� given� in� Figure�3.5,� CER� values� are�higher� for� the� residues�

than� for� the�mine�area� samples,�which� is� logical� since� the�wastes� resulting� from�mining�and�

milling� processes� were� stockpiled� in� these� specific� areas.� However,� none� of� the� metals�

analyzed� in� Kettara� (mine� area� or� residues)� has� very� strong� anthropogenic� contribution�

(CER>40).�Concerning�box�plot�figures�for�mine�area�and�residues�for�SB�Othmane�(Figure�3.6)�

and�Bir�Nehass� (Figure�3.7)� it� can�be�observed� that�CER�values�are�much�higher� for� residues�

than�for�mine�area�samples�especially�regarding�Pb�and�Zn.�Given�the�similarities�between�box�

plot� figures� for� both� SB�Othmane� and� Bir� Nehass� it� can� be� stated� that� both� mines� are�

mineralogically�comparable.�In�this�sense,�box�plot�figures�of�CER�values�highlight�differences�

between�mine� area� and� residues� samples� as�well� as� to� graphically� determine� the� degree� of�

contamination�regarding�a�certain�element.�

Principal� components� analysis� (PCA)� was� also� applied� to� the� CER� data� to� point� out�

differences� within.� From� the� representation� of� loadings� and� scores� of� PC1� and� PC2� for� the�

three�mines�(Figure�3.8,�3.9,�and�3.10)�it�can�be�distinguished�in�each�mine:�mine�area�samples�

grouped� all� together� at� the� center� of� the� figures� and� residues� samples� at� the� left� of� the�

representation� more� dispersed.� In� this� sense,� given� the� proximity� of� the� residues� samples�

representation� and� the�main� pollutants� in� the� figure� it� can� be� considered� that� residues� are�

more� influenced� by� the� pollutants� than� mine� area� samples.� However,� from� the� scores�

representation�of�each�mine,�it�can�be�observed�that�many�of�the�samples�with�high�CER�values�

(and�hence�with�high�concentration�on�pollutants)�can�be�clearly�distinguished�from�the�bulk�of�

the�mine� area� samples� (S8� to� S10� in� Kettara,� S2� in� SB�Othmane� and� S8,� S19� and� S20� in� Bir�

Nehass).�In�this�regard,�it�can�be�concluded�that�PCA�is�able�to�point�out�differences�between�

mine�area�samples�and�residues�and�even�to�distinguish�the�most�polluted�from�samples�less�

polluted�as�well�as�to�indicate�the�parameters�that�affect�the�distinction�between�samples.�In�

general� it�can�be�stated�that�for�the�three�mines�studied,�PC1�is�mainly�assigned�to�the�main�

pollutants�(As,�Cu,�Pb�and�Zn)�while�lithoghenic�elements�such�as�Ba,�K,�Ca,�Rb,�Mn�and�Sr�or�

the�physico�chemical� variables� charge� the� rest�of� PC.�On� the�other�hand,� PCA� is� not� able� to�

3.�Results�and�discussion��

�96�

establish� patterns� for� the� three� mines� since� loading� distribution� is� different� for� the� three�

mines.��

Figure�3.5.�Box�plot�distribution�of�CER�values�for�As,�Cu,�Pb�and�Zn�for�Kettara�mine�area�(M.A.)�

and�residue�samples.�CER=5�(dotted�line),�CER=20�(lines)�and�CER=40�(straight�line)

� �Figure�3.6.�Box�plot�distribution�of�CER�values�for�As,�Cu,�Pb�and�Zn�for�SB�Othmane�mine�area�(M.A.)�and�residue�samples.�CER=5�(dotted�line),�CER=20�(lines)�and�CER=40�(straight�line)

� �Figure�3.7.�Box�plot�distribution�of�CER�values�for�As,�Cu,�Pb�and�Zn�for�Bir�Nehass�mine�area�(M.A.)�

and�residue�samples.�CER=5�(dotted�line),�CER=20�(lines)�and�CER=40�(straight�line)

3.�Results�and�discussion�

97�

KS1KS2KS3KS4KS5KS6

KS7 KS8

KS9 KS10

KS11KS12 KS13KS14KS15

KS16

KS17

KS18KS19KS20KS21KS22KS23KS24

KS25KS26KS27KS28

KS29KS30KS31KS32

KS33

KS34

KS35KS36

KS37KS38

KS39

KS40

KS41KS42KS43

KS44

KS45

KS46KS47KS48

KS49

KS50KS51

KS52KS53

KS55KS56KS58

KR1KR2

SR3KR4

KR5

KR6

KR7

As

BaCa Cu

Fe

K

Mn

PbRbSr

Ti

ZnpH

CECaCO3

�15

�10

�5

0

5

10

15

�10 �5 0 5 10 15 20 25 30

PC2�(15.39�%

)

PC1(36.36�%)

Biplot�(PC1�and�PC2:�51.75�%)

Figure�3.8.�Kettara�biplot�representation�of�loadings�and�scores�of�PC1�and�PC2��

SB1SB2

SB3

SB4

SB5SB6SB7SB8SB9

SB10

SB11SB12

SB13SB14SB16SB17

SB18

SB19SB20SB21

SB22SB23

SB24SB25SB26SB27

SR1R4B2

RS3

R4B6

R4B7SR8

R4B9R4B10R4B12

SR13

R4B14R4B17

R4B18

R4B19SR20

As

Ba

Ca

Cu

Fe

K

Mn

Pb

Rb

Sr

TiZn pH

CE

CaCO3

�20

�15

�10

�5

0

5

10

15

20

�20 �15 �10 �5 0 5 10 15 20 25 30

PC2�(20.69�%

)

PC1�(38.97�%)

Biplot�(PC1�and�PC2:�59.66�%)

Figure�3.9.�SB�Othmane�biplot�representation�of�loadings�and�scores�of�PC1�and�PC2��

BN1

BN2BN3

BN4BN5

BN6BN7

BS8

BN9BN10BN11

BN12BN13BN14BN15

BN16

BN17BN18

BS19

BS20BN21BN22BN23

BN24

BN25BN26BN27BN28BN29BN30BN31

BN32

BN33 BR1

BR3 BR4BR5AsBa

Ca

CuFe

K Mn

Pb

Rb

Sr

TiZn

pH

CECaCO3

�20

�15

�10

�5

0

5

10

15

20

�20 �15 �10 �5 0 5 10 15 20 25 30 35

PC2�(19.36�%

)

PC1�(35.78�%)

Biplot�(PC1�and�PC2:�55.14�%)

Figure�3.10.�Bir�Nehass�biplot�representation�of�loadings�and�scores�of�PC1�and�PC2��

3.�Results�and�discussion��

�98�

3.2.4.�GIS�CONTOUR�MAPS�OF�THE�POLLUTANTS�

With�separate�sample�values�and�in�order�to�delineate�the�distribution�of�target�metals�

around�the�mine�area�a�GIS�representation�of�corresponding�CER�values�has�been�carried�out.�

The� obtained� information� can� be� applied� on� the� management� of� the� target� areas,� i.e.,�

monitoring,�isolation�or�pollution�removal.�

Kettara�GIS� contour�maps� (Figures�3.11� to�3.14)� show� two�well� separated�areas:� a�big�

area�at� the�center�of� the�mine�area�and�a�small�area�at� the�north.�Both�areas�are� located�at�

specific� sampling�points�with� localized�contamination,�especially�at� residues�sampling�points.�

Given� that� the� area� is� flat,� the� distribution� of� pollutants� does� not� follow� any� geographic�

consideration� but� only� specific� points�were� residues�were� stored.� Considering� the�maps� for�

each�element�it�can�be�stated�that�the�distribution�of�pollutants� is�similar�for�arsenic,�copper�

and�lead,�whilst�zinc�distribution�is�more�homogeneous�along�the�mine�area.�In�addition,�lead�

can�be�considered�the�main�pollutant�regarding�its�high�CER�values.��

Regarding�SB�Othmane�mine�area�GIS�maps�it�can�be�pointed�out�that�the�mine�area�is�

highly� contaminated,� being� observed� samples�with� CER� values� above� 200� (Figures� 3.17� and�

3.18).�Arsenic�and�copper�GIS�maps� (Figures�3.15�and�3.16)�of�SB�Othmane�present� two�hot�

spots�located�at�deposits�of�residues�while�Pb�and�Zn�GIS�maps�indicate�a�more�risky�situation�

with�CER�values�above�200�(Figures�3.17�and�3.18).�The�distribution�of�contaminants�is�affected�

mainly�by�the�location�of�the�deposits�of�residues.��

With� respect� to�Bir�Nehass�mine� area� it� can�be� stated� that,� likewise� SB�Othmane,� the�

mine� is� less� contaminated� with� arsenic� (Figure� 3.19)� and� copper� (Figure� 3.20)� being� lead�

(Figure�3.21)�and�zinc�(Figure�3.22)�the�main�pollutants.�In�this�sense,�a�unique�hot�spot�can�be�

observed�for�arsenic�and�lead�around�a�sample�located�at�a�residue�deposit�while�several�hot�

spots�with�CER=200�can�be�seen�for�lead�and�zinc,�related�to�specific�residue�deposits.�The�area�

is� flat� and� the� distribution� is� only� owed� to� the� location� of� the� deposit� of� residues� from� the�

mine.��

Once� detected� the� most� polluted� areas� within� each� mine,� a� real� evaluation� of� their�

potential�risk�can�be�obtained�from�the�results�of�metals�mobility,�shown�in�Table�3.6,�by�using�

the�single�extraction�procedure.��

The�mobility�results�let�to�point�out�that�all�the�mine�area�samples�taken�at�Kettara�mine�

area�have�very�low�mobility,�with�the�exception�of�the�residues�(more�than�30�mg�Cu/L�on�the�

mobile�fraction).�These�samples�have�an�acidic�pH�(around�2)�and�very�low�organic�matter�(LOI�

~10),�in�such�situation,�mobility�of�heavy�metals�is�favored.��

3.�Results�and�discussion�

99�

On� the� other� hand,� SB�Othmane� and� Bir� Nehass� mine� area� samples� are� highly�

concentrated�on�Pb� and� Zn� showing� an� extremely� high� content� on�Pb� and� Zn� in� the�mobile�

phase�for�mine�area�samples�and�even�higher�for�the�residues,�thus�representing�a�threaten�for�

the�environment.�The�physic�chemical�characteristics�of�these�samples�(pH�around�7,�and�high�

content�on�organic�matter)�do�not�favor�mobility�of�the�elements�as�can�be�observed�for�the�

low�levels�of�arsenic�and�copper�found�in�the�mobile�phase.�However,�given�the�high�content�

of�lead�and�zinc�it� is�likely�that�the�concentration�of�metals�exceed�the�capacity�of�the�soil�to�

retain� them�and�the�migration�to�a�mobile�phase�may�take�place.� In� this�sense,� the�residues�

sampling�points�which� are� associated�with�waste�disposal� sites�are�even�more� concentrated�

than�mine�area�samples�and,�given�the�high�amount�of�metals,�the�immobilization�processes�in�

soils� to� retain�metals� are�also�overwhelmed.� In� this� regard,� the� level�of�metals� found� in� the�

mobile� phase� for� SB�Othmane� and� Bir� Nehass� are� extremely� high� (especially� for� zinc)� so�

remediation� treatments� should� be� applied� to� these� areas� if� the� soil� is� intended� for� further�

purposes.��

3.�Results�and�discussion��

�100�

Table�3.6.�Total�concentration�(in�mg/Kg)�and�mobility�(in�mg/L)�for�As,�Cu,�Pb�and�Zn�of�the�most�contaminated�samples�of�Kettara,�Sidi�Bou�Othmane�and�Bir�Nehass�mine�areas�

� � � � As� Cu� Pb� Zn�

Total� 237±12� 1,360±50� 307±15� 243±18�S8�

Mobile� <0.5� 6.1±0.2� <0.5� 1.1±0.1�Total� <15� 810±40� 406±19� 49±15�

S9�Mobile� <0.5� 2.5±0.1� <0.5� 1.7±0.1�Total� 117±14� 956±50� 486±22� 67±16�

S10�Mobile� <0.5� 18±0.1� <0.5� <0.5�Total� 51±7� 643±20� 328±11� 213±12�

Mine�area�

S28�Mobile� <0.5� 6.3±0.1� 2.2±0.2� 1.3±0.1�Total� <15� 1,257±50� 341±18� 99±17�

R4�Mobile� <0.5� 21.1±0.1� 0.5±0.1� 3.8±0.1�Total� <15� 1,248±50� 303±17� <10�

R5�Mobile� <0.5� 11.1±0.1� 0.5±0.1� 1.1±0.1�Total� <15� 2,113±60� 164±12� 337±20�

Residues�

R6�Mobile� <0.5� 33.0±0.1� 0.7±0.1� 11.4±0.1�Total� 96±28� <25� 6,542±89� 36,267±400�

S2�Mobile� 1.6±0.1� 2.2±0.1� 393±3� 1071±30�Total� <15� <25� 2,665±37� 10,991±120�

S3�Mobile� 1.2±0.1� 1.1±0.1� 184±2� 563±8�Total� <15� <25� 3,723±50� 13,555±160�

S11�Mobile� 1.4±0.1� 1.5±0.1� 280±3� 719±50�Total� <15� <25� 6,706±87� 22,641±300�

Mine�area�

S30�Mobile� 1.4±0.1� 3.2±0.1� 483±3� 932±30�Total� <15� 26,400±300� 103±3� 135±3�

R3�Mobile� 2.7±0.2� 2.3±0.2� 179±2� 359±4�Total� <15� <25� 6,428±80� 17,490±200�

R5�Mobile� 3.7±0.2� 4.3±0.3� 733±2� 772±6�Total� 197±40� 240±20� 11,004±150� 31,487±400�

Sidi�Bou�Othmane�

Residues�

R8�Mobile� 2.8±0.2� 2.5±0.2� 523±6� 1240±20�Total� 22±4� <25� 166±7� 1,309±23�

S4�Mobile� <0.5� <0.5� 4.6� 54±4�Total� 46±13� <25� 1,390±30� 29,732±350�

S19�Mobile� <0.5� 0.6±0.1� 54±3� 1071±140�Total� 113±13� <25� 1,500±30� 13,140±160�

S20�Mobile� <0.5� <0.5� 40±3� 369±70�Total� 36±5� <25� 259±8� 1,960±30�

Mine�area�

S21�Mobile� <0.5� <0.5� 10±2� 77±16�Total� 93±20� <25� 3,705±60� 15,143±200�

R1�Mobile� <0.5� 2.1±0.1� 18.3±0.5� 474±5�Total� 164±18� 52±14� 2,640±40� 8,521±115�

R3�Mobile� <0.5� 1.0±0.1� 88±2� 99.8±0.8�Total� 112±17� 86±18� 1,776±34� 13,325±190�

Bir�Nehass�

Residues�

R5�Mobile� <0.5� 3.0±0.1� 55.7±1.4� 639±7�

3.�Results�and�discussion�

101�

Scale 1:25000

Arsenic

CER=0

CER=20

CER=10

Scale 1:25000

Arsenic

CER=0

CER=20

CER=10

Copper

Scale 1:25000

CER=0

CER=80

CER=40

Copper

Scale 1:25000

CER=0

CER=80

CER=40

Figure�3.11.�Kettara�mine�area�GIS�contour�map�of�arsenic.

Figure�3.12.�Kettara�mine�area�GIS�contour�map�of�copper

Scale 1:25000

Lead

CER=0

CER=40

CER=20

Scale 1:25000

Lead

CER=0

CER=40

CER=20

Scale 1:25000

Zinc

CER=0

CER=20

CER=10

Scale 1:25000

Zinc

CER=0

CER=20

CER=10

Figure�3.13.�Kettara�mine�area�GIS�contour�map�of�lead.

Figure�3.14.�Kettara�mine�area�GIS�contour�map�of�zinc.

Arsenic

CER=0

CER=20

CER=10

Scale 1:15000

Arsenic

CER=0

CER=20

CER=10

Scale 1:15000

Figure�3.15.�Sidi�Bou�Othmane�mine�area�GIS�contour�map�of�arsenic

3.�Results�and�discussion��

�102�

Copper

CER=0

CER=20

CER=10

Scale 1:15000

Copper

CER=0

CER=20

CER=10

CER=0

CER=20

CER=10

Scale 1:15000

Figure�3.16.�Sidi�Bou�Othmane�mine�area�GIS�contour�map�of�copper

Lead

CER=0

CER=400

CER=200

Scale 1:15000

Lead

CER=0

CER=400

CER=200

Scale 1:15000

Figure�3.17.�Sidi�Bou�Othmane�mine�area�GIS�contour�map�of�lead

Zinc

CER=0

CER=200

CER=100

Scale 1:15000

Zinc

CER=0

CER=200

CER=100

CER=0

CER=200

CER=100

Scale 1:15000

Figure�3.18.�Sidi�Bou�Othmane�mine�area�GIS�contour�map�of�zinc

3.�Results�and�discussion�

103�

Arsenic

CER=0

CER=80

CER=40

Scale 1:5000Arsenic

CER=0

CER=80

CER=40

CER=0

CER=80

CER=40

Scale 1:5000

Copper

CER=0

CER=20

CER=10

Scale 1:5000Copper

CER=0

CER=20

CER=10

CER=0

CER=20

CER=10

Scale 1:5000

Figure�3.19.�Bir�Nehass�mine�area�GIS�contour�map�of�arsenic

Figure�3.20.�Bir�Nehass�mine�area�GIS�contour�map�of�copper

LeadScale 1:5000

CER=0

CER=200

CER=100

LeadScale 1:5000

CER=0

CER=200

CER=100

ZincScale 1:5000

CER=0

CER=200

CER=100

ZincScale 1:5000

CER=0

CER=200

CER=100

Figure� 3.21.� Bir� Nehass� mine� area� GIS� contour�map�of�lead

Figure�3.22.�Bir�Nehass�mine�area�GIS�contour�map�of�zinc

3.�Results�and�discussion��

�104�

3.3.� XANES� SPECIATION� OF� MERCURY� IN� THREE� MINING� DISTRICTS:� ALMADÉN,�ASTURIAS�(SPAIN),�IDRIA�(SLOVENIA)�(ANNEX�2)�

Mercury�is�one�of�the�most�toxic�elements�as�some�of�its�compounds�can�be�absorbed�by�

living�tissues�in�large�doses�and�these�compounds�or�their�derivatives�can�concentrate�and�be�

stored�over� long�periods�of� time�causing�chronic�or�acute�damages�[5].�The�toxicity�of�heavy�

metals�is�mainly�controlled�by�the�dose�and�its�chemical�speciation.�Hence,�the�assessment�of�

mercury� species� on� the� environment� is� of� great� relevance� since�many� health� problems� are�

related� to� specific� Hg� species.� Following� previous� studies� by� Brown� and� co�workers� on� the�

characterization� of� mercury� mines� in� north� America� [6,� 7],� this� work� aimed� at� providing� a�

general� perspective� on� the� speciation� of� mercury� in� three� of� the� most� important� mercury�

mining� districts� in� Europe.� In� this� study,� XANES� has� been� complemented� with� a� single�

extraction�protocol�for�the�determination�of�Hg�mobility�to�determine�toxicity�of�the�samples.��

3.3.1.�CHEMICAL�ANALYSIS�OF�THE�SAMPLES�

A� list� of� samples� from� the� metallurgical� plants� and� drainage� network� of� the� three�

districts,�their�corresponding�acronyms�and�a�short�description�of�the�sampling�site�is�provided�

in�Table�3.7.�Their�location�on�the�mine�is�depicted�in�Figure�3.23.�

Figure�3.23.�Sampling�locations,�mines�and�metallurgical�sites�of�the�three�mercury�mining�districts:�

Almadén,�Asturias�and�Idria.��

3.�Results�and�discussion�

105�

Table�3.7.�Samples�collected�at�the�three�mining�districts�Location� Id� Sampling�area� Material�

ALMADEN�SITE�

Almadén� HR� Huerta�del�Rey� Soils�from�an�old�metallurgical�plant�of�the�17th�century�

� CH� Main�dump�of�Almadén�mine� Dump�material,�sediments�and�riparian�soils�� AZG� Azogado�river�stream� Riparian�soils�and�stream�sediments�Almadenejos� ALM� Decommissioned�metallurgical�plant� Soils�from�the�metallurgical�plant�Valdeazogues�river� RD� Downstream�of�El�Entredicho�pit� Suspended�particles�San�Quintín�� SQ� Decommissioned�Pb�Zn�Ag�mine� Mine�wastes�and�soils�from�and�old�flotation�

plant�used�to�treat�cinnabar�

ASTURIAS�SITE�

Mine�tailings� TRRmn� Mine�and�metallurgical�plant� Dumps�in�the�vicinity�of�rotary�furnaces�Calcines� TRRc� Mine�and�metallurgical�plant� Calcination�waste�Soil� TRRs� Metallurgical�plant� Soil�from�an�abandoned�chimney�channel�Forest�Soils� TRRfs� El�Terronal�mine� Forest�soils�from�the�mining�area�

IDRIA�SITE�

Soils� S1�S3� Vicinity�of�the�metallurgical�plant�� Soils�� S2� Pront�Hill� Meadow�soils��� S4� Confluence�of�the�Idrijca�and�Baca�

rivers�Alluvial�soil�samples�collected�along�the�Idrijca�river�40�km�downstream�from�the�mine�

� S5�S6�� Confluence�of�Idrijca�and�Baca�rivers� Soils�from�a�deep�profile�at�50�cm�depth�(S5)�and�100�cm�(S6)�

Sediments� RS� Idrijca�river,�35�km�downstream�from�the�mine�before�Baca�river�inflow�

River�bed�sediments�of�a�composite�sample�taken�within�a�distance�of�50�m�with�grain�size�<0.063�mm�(RS1)�and�0.063�2�mm�(RS2)�

� SS� Idrijca�river,�35�km�downstream�from�the�mine�before�Baca�river�inflow�

Suspended�river�sediments�of�a�composite�sample�taken�within�a�distance�of�50�m�with�grain�size�<0.063�mm�(SS1)�and�0.063�2�mm�(SS2)�

High�mercury�concentrations�in�soil�samples�from�metallurgical�sites�were�found�at�the�

Almadén� district� (Table� 3.8)� that� can� be� mainly� attributed� to� the� inefficient� metallurgical�

techniques�used�in�the�old�plants�of�Almadenejos�and�Huerta�del�Rey�[8].�In�these�plants,�the�

roasting� temperatures�were� below�500� ºC.� Also� high�mercury� concentrations�were� found� in�

sediments�and�riparian�soils�from�Valdeazogues�river�(RD)�and�especially�from�Azogado�stream�

(AZG)�(2,816�mg�Hg�g�1).�These�results�coincide�with�previous�studies�undertaken�at�the�same�

sampling� site� [9].� Other� heavy� metals� are� also� found� in� Almaden� site� samples� although� in�

minor�concentrations,�and�especially�high�amounts�of�lead�and�zinc�were�found�in�samples�of�

San�Quintín�area�(SQ).��

Likewise�Almaden,�the�total�mercury�content�of�soil�and�dump�samples�of�Asturias�mine�

(Table� 3.9)� show� the� high�mercury� content� (the� highest� from� the� three�mines� studied)�with�

27,350� mg� Hg� g�1� in� dump� samples� (TRRmn�116)� and� 18,000� mg� Hg� g�1� in� soils� from� the�

chimney�channel,�with�also�high�amounts�of�arsenic�(from�735�mg�As�g�1�to�187,218�mg�As�g�1).��

On�the�other�hand,�Idria�samples�(Table�3.10)�revealed�minor�amounts�of�the�metals�analyzed�

compared�to�Almaden�and�Asturias,�being�the�samples�near�the�former�smelting�facilities�the�

most�polluted�caused�by�the�settling�down�of�Hg�enriched�particles�in�the�immediate�vicinity�of�

3.�Results�and�discussion��

�106�

the�smokestack�of�the�smelter.�It�is�important�to�highlight�the�high�Hg�concentration�observed�

in� Idria� sediments� (RS)� and� in� alluvial� soils� (S4)� 40� km�downstream� from� the�mine� probably�

linked�to�mercury�bearing�rocks,�wastes�from�combustion�processes�or�contaminated�river�bed�

sediments.�These� inputs� to� the�aquatic�environment� remain� in� the�area�even�a�decade�after�

the�ending�of�mining�operations.�

Table� 3.8.� Almaden� average� metal� content�(given�in�μg�g�1)�

� Table� 3.9.� Asturias� average� metal� content�(given�in�μg�g�1)�

� SAMPLE� Hg� As� Pb�� Zn�TRRmn�115� 1,470� 39338� <10� <15�TRRmn�116� 27,350 11,7553� <10� <15�TRRs�118� 3,280� 735� <10� <15�TRRs�121� 18,000 12,133� <10� <15�TRRmn�122� 5,785� 4,2300� <10� <15�TRRfs�3� 1,570� 1,6826� 107� 173�TRRfs�4� 1,080� 1,120� 53� 137�TRRc�5� 34� 187,218� <10� <15�TRRc�55� 54� 25,876� <10� <15��

� Table�3.10.�Idria�average�metal�content�(given�in�μg�g�1)�

SAMPLE� Hg� As� Pb�� Zn�CH�127� 989� <15� <10� 112�HR�108� 976� <15� 214� 96�HR�109� 404� <15� 111� 104�HR�110� 200� <15� 130� 185�RD�124� 105� <15� <10� <10�CH�125� 1,800� <15� <10� 112�AZG�105� 2,816� 23� 139� 233�CH�128� 450� <15� 102� 185�ALM�101� 2,720� <15� 74� 153�ALM�102� 2,629� <15� 102� 193�CH�126� 2,230� <15� <10� 365�SQ�111� 902� <15� 15,837 6,877SQ�112� 1,730� <15� 2,154� 1,221SQ�113� 1,935� <15� 19,049 7,134SQ�114� 390� �� �� ���

� SAMPLE� Hg� As� Pb�� Zn�S�1� 333� 21� <10� 112�S�2� 47� 26� <10� 102�S�4� 76� <15� <10� 64�S�5� 175� <15� 47� 145�S�6� 144� <15� 73� 496�RS�1� 6,540� <15� 302� 270�RS�2� 1920� <15� 14� <15�SS–1� 96� <15� <10� 449�SS�1� 11� <15� <10� 24�S�3� 95� 27� 46� 130��

3.3.2.�XANES�SPECIATION�AND�MOBILITY�RESULTS�

In�Figure�3.24� the� spectra� corresponding� to�mercury� standards�and� to� the� samples� for�

each�mine�are�given.�Considering� the�number�of�sample�XANES�spectra,�PCA�was�performed�

separately�for�each�mining�district.�An�example�of�fitting�for�a�selected�sample�of�each�mine�is�

given�in�Figure�3.25.��

PCA� results� for� Almaden� district� indicated� that� five� components� are� required� to�

reconstruct�each�of�the�experimental�spectra�(cinnabar,�Cb�(red�HgS);�metacinnabar,�Mc�(black�

HgS);�HgCl2;�calomel�(Hg2Cl2)�and�schuetteite,�Sc�(Hg3(SO4)O2))�with�above�95%�of�confidence.�

The�most� common�species� found� in�almost�all� samples�were�mercury� sulfides� (cinnabar�and�

metacinnabar)� but� also� non�sulfide� phases� like� schuetteite,� calomel� (Hg2Cl2)� and� mercury�

chloride�(HgCl2)�which�were�found�in�soil�and�sediment�samples.��

3.�Results�and�discussion�

107�

Figure�3.24.�XANES�spectra�of�selected�Hg�pure�compounds�and�samples�from Almaden,�Idria�and�Asturias�mining�districts�(all�spectra�are�deliberately�stacked�to�show�differences).�Each�spectrum�

corresponds�to�the�mean�value�of�five�replicates.�

Figure�3.25.�XANES�spectra�of�selected�samples�from�the�three�mining�districts�with�reconstructed�

spectra�shown�as�dashed�lines.�

3.�Results�and�discussion��

�108�

Table�3.11.�Main�mercury�species�(in�%)�and�mobile�mercury�(in�mg�L�1�and�%).�Abbreviations:�Cb:�cinnabar;�Mc:�metacinnabar;�Sc:�schuetteite;�Co:�corderoite�

�Sample� Cb� Mc� Sc� Co� HgO� HgSO4� Hg2Cl2� HgCl2�

Red.�Chi�Sq.(10�3)�

Mobility��mg�L�1�(%)�

CH�127� 62� 0� 0� 0� 0� 0� 38� 0� 0.4� 1.4±0.3�HR�108� 37� 23� 0� 0� 0� 0� 40� 0� 0.6� 0.6±0.2�HR�109� 33� 24� 0� 0� 0� 0� 43� 0� 0.7� 0.2±0.1�HR�110� 41� 22� 0� 0� 0� 0� 37� 0� 0.6� <0.2�RD�124� 0� 0� 94� 0� 0� 0� 0� 6� 0.5� <0.2�CH�125� 7� 0� 83� 0� 0� 0� 0� 10� 0.4� <0.2�AZG�105� 0� 0� 80� 0� 0� 0� 20� 0� 0.3� <0.2�CH�128� 24� 22� 0� 0� 0� 0� 35� 19� 0.4� <0.2�ALM�101� 38� 39� 23� 0� 0� 0� 0� 0� 0.3� 10.8±0.3�ALM�102� 39� 31� 0� 0� 0� 0� 30� 0� 0.7� 21.3±0.5�CH�126� 33� 32� 35� 0� 0� 0� 0� 0� 0.3� <0.2�SQ�111� 54� 0� 17� 0� 0� 0� 29� 0� 0.2� 0.6±0.1�SQ�112� 51� 0� 21� 0� 0� 0� 28� 0� 0.2� 3.7±0.2�SQ�113� 59� 0� 17� 0� 0� 0� 24� 0� 0.2� <0.2�

ALM

ADEN

SQ�114� 47� 0� 20� 0� 0� 0� 33� 0� 0.3� <0.2�TRRmn�115� 29� 24� 0� 0� 0� 0� 0� 47� 1� 0.4±0.1�TRRmn�116� 28� 22� 0� 0� 0� 0� 0� 50� 0.9� 73±2�TRRs�118� 28� 22� 0� 0� 0� 0� 0� 50� 0.8� 20.1±1.3�TRRs�121� 29� 22� 0� 0� 0� 0� 0� 49� 0.7� 56.5±2�TRRmn�122� 30� 24� 0� 0� 0� 0� 0� 46� 0.7� 43.6±2�TRRfs�3� 44� 28� 0� 0� 10� 18� 0� 0� 3� 0.7±0.2�TRRfs�4� 50� 36� 0� 14� 0� 0� 0� 0� 3� <0.2�TRRc�5� 52� 30� 0� 18� 0� 0� 0� 0� 8� <0.2�

ASTURIAS�

TRRc�55� 57� 43� 0� 0� 0� 0� 0� 0� 7� <0.2�S�1� 44� 0� 32� 0� 0� 24� 0� 0� 6� <0.2�S�2� 55� 0� 0� 0� 0� 45� 0� 0� 2� 0.2±0.1�S�4� 85� 15� 0� 0� 0� 0� 0� 0� 4� <0.2�S�5� 90� 0� 0� 0� 10� 0� 0� 0� 4� <0.2�S�6� 58� 0� 0� 0� 0� 42� 0� 0� 5� <0.2�RS�1� 57� 0� 0� 0� 0� 43� 0� 0� 2� <0.2�RS�2� 100� 0� 0� 0� 0� 0� 0� 0� 3� <0.2�SS�1� 90� 0� 0� 0� 0� 10� 0� 0� 4� <0.2�SS�1� 55� 0� 0� 0� 0� 45� 0� 0� 9� <0.2�

IDRIA�

S�3� 66� 0� 26� 0� 8� 0� 0� 0� 07� 0.3±0.1�

Regarding� Almaden�mine� area� samples,� XANES� analyses� from� San�Quintín� area� (Table�

3.11)� indicated� high� amounts� of� cinnabar� (47–59%)� and� minor� amounts� of� relatively� more�

soluble� species� like� calomel� (24–33%)� and� schuetteite� (17–21%)� that� can� be� attributed� to�

weathering� processes.� The� absence� of� metacinnabar� phases� in� that� samples,� a� metastable�

polymorph�of�cinnabar�that�occurs�when�the�roasting�process�of�mercury�ores�is�not�complete�

or�it�is�done�in�the�presence�of�impurities�[10],�is�associated�to�the�historical�use�of�the�site,�as�

this� site�was� used� to� perform� flotation� tests� and�no� furnaces�were� employed.�On� the�other�

hand,�metacinnabar�has�been�identified�in�soil�samples�from�Almadenejos�(ALM)�(31–39%)�and�

Huerta�del�Rey�(HR)�(~23%),�locations�with�historical�metallurgical�activity.�

3.�Results�and�discussion�

109�

Other�non�sulfide�phases�like�mercurous�chloride�(24–43%)�have�also�been�identified�at�

San�Quintín�and�Huerta�del�Rey,�attributable�to�the�process�of�soil�formation.�High�amounts�of�

schuetteite� have� been� identified� in� ore� stockpile� in� San� Quintín� and� Almadenejos� area.�

Schuetteite� is� a�mineral� phase� typically� linked� to� the� presence� of� Hg(0)� that� appears� in� the�

sunlight�exposed�side�of�the�rock�surface,�and�it�is�frequently�found�near�old�furnaces�and�ore�

dumps�[11].�Relatively�more�soluble�phases�have�been�identified�in�soil�and�sediment�samples�

from�Valdeazogues�River�(100%)�and�Azogado�stream�(100%)�(Hg2Cl2,�HgCl2�and�Hg3(SO4)O2)�as�

a� result�of�weathering�processes� caused�by� the�drainage�network�of� the�mining�district.� The�

mobility�of�mercury�in�this�district�is�clearly�linked�with�metallurgical�activity�and�formation�of�

secondary� chloride� phases.� The� highest� mobility� was� found� in� soil� samples� from� an� old�

metallurgical�precinct�(ALM)�(21.3�mg�L�1;�Table�3.11)�related�to�the�presence�of�Hg2Cl2.��

In�Asturias�mining�district,�all�samples�from�the�decommissioned�mine�and�metallurgical�

facility�showed�high�mercury�contents� in�soils� (TRRfs),�dump�materials� (TRRmn)�and�chimney�

soils� (TRRs)� (Table� 3.9),� and� a� predominance� of� sulfides� species� (50–100%)� with� significant�

presence� of� metacinnabar� in� all� samples� (Table� 3.11).� Cinnabar� and�metacinnabar� in� these�

samples�is�higher�than�in�Almadén�area�since�in�Asturias�the�metallurgy�was�less�efficient�than�

in� Idria�and�Almadén�area,�with� lower�roasting�temperature�and�poorest�recovery�rates.�The�

contents�of�other�mercury�species�such�as�chlorides�are�significant,�with�high�amounts�on�soils�

samples�from�the�facility�and�the�chimney�exhausting�roasting�smokes�and�thus�the�mobility�of�

mercury� in� this� district� is� higher� than� in� Almadén.� In� qualitative� terms,� the�mobile�mercury�

determined�is�correlated�with�the�presence�of�HgCl2�(except�for�TRRmn�115),�a�mobile�phase�

of� mercury� and,� in� a� lesser� account,� to� the� presence� of� metacinnabar� resulting� from� the�

incomplete�combustion.��

At� the� Idria� mining� district,� cinnabar� is� the� most� common� form� of� mercury� in� soil,�

sediments�and�suspended�particles,�while�metacinnabar� is�also� found� in�soil�sample�S�4,�and�

sulfates�in�soils�and�sediments�(S,�RS,�SS).�The�lack�of�metacinnabar�in�most�of�these�samples�is�

due�to�the�re�use�of�calcines�and�metallurgical�wastes�in�the�refilling�of�mine�galleries�resulting�

in� a� minor� dispersion� of� this� material� throughout� the� surrounding� environment.� High�

proportions� of� sulfates� were� found� in� soil� samples� (S),� but� the� mobility� of� mercury� in� this�

district�was�clearly� reduced,�mainly�by� the�major�proportions�of� cinnabar� in� soils,� sediments�

and� suspended� particles.� This� low�mobility� of�mercury� (0.2–0.3�mg� L�1,� see� Table� 3.11)� is� in�

agreement� with� former� studies� on� the� area� [12]� that� described� low� water�soluble�mercury�

species�in�sediments�and�suspended�particles.��

Considering�the�three�districts,�the�main�processes�affecting�mercury�speciation�are�ore�

composition,�mining�history�and�roasting�process.�The�type�of�metallurgical�processing�arises�

3.�Results�and�discussion��

�110�

as�one�of� the�most� important� factors� in�defining�mercury�availability.� In� this� sense,�mercury�

mobility�is�higher�in�Asturias�district�owing�to�its�roasting�treatment�was�less�efficient�than�in�

Almaden�or�Idria�(lower�roasting�temperatures�and�poorer�recovering�rates)�that�increases�the�

presence� of� metacinnabar� and,� principally,� HgCl2� phases� responsible� for� the� mobility� of�

mercury.� Despite� the� complex� and� lengthy� history� of� mining� and�metallurgical� activity,� the�

mobility� is� significantly� lower� in� the� Almaden� district� given� its� better� roasting� processes�

achieved� with� better� furnaces� (only� in� the� last� century)� and� likewise� Almaden,� even� lower�

mobility�values�were�found�in�Idria�district�related�to�its�efficient�metallurgical�process�(similar�

to�Almadén�area),� together�with� the�appropriate�management�of�calcines� that�were�used� to�

refill�old�galleries�as�well�as�the�shorter�mining�history�of�this�district.�

Rather� insoluble�mercury�compounds�(cinnabar,�metacinnabar,�schuetteite,�corderoite)�

have� been� shown� to� prevail� in� dumps� and� wastes� from� mines� and� metallurgical� plants,�

whereas�more�soluble�Hg�phases� (mainly�HgCl2�but�also�HgO�and�HgSO4)�were� found� in�soils�

and� sediments� from�all� target�areas.�A�qualitative� relationship�between�mobile�mercury�and�

the� presence� of�mercury� chlorides� or� sulfates� compounds� has� been� established� for� samples�

from� the� three� districts.� Nonetheless,� the� absolute� mobility� remains� relatively� low� in� most�

cases,� inherently� suggesting� that� kinetic� effects� and� availability� of� the� soluble� phases�might�

also�be�considered�in�the�assessment�of�mercury�behavior.�

3.�Results�and�discussion�

111�

REMEDIATION�TECHNOLOGIES�

This� section� includes� results� of� treatment� processes� for� industrial� water� from� two�

different� sectors:� mining� activities� and� textile� industry.� These� results� constitute� specific�

examples�of�innovation�in�water�treatment�process�for�both�inorganic�and�organic�pollutants.�

Thus,�results�of�laboratory�and�pilot�plant�scale�for�recycling�of�water�from�a�mine�tailing�

pond�are� reported�here.�On� the�other�hand,� the� results�achieved�by�Fe�exchanged�materials�

for� the�degradation�of�persistent�organic�pollutants� (POPs)�by�means�of�Fenton�processes�as�

well�as�to�the�sorption�of�inorganic�contaminants�are�also�summarized.��

3.4.� EXTRACTANT� AND� SOLVENT� SELECTION� TO� RECOVER� ZINC� FROM� A� MINING�

EFFLUENT:�FROM�LABORATORY�SCALE�TO�PILOT�PLANT�

In�a�tailing�pond�from�an�abandon�mine�is�stored�a�huge�stream�of�effluent,�estimated�to�

be�10,000�m3/day�and�containing�about�1�g/L�of�Zn�and�significant�amounts�of�ferrous,�ferric,�

calcium,�copper,�aluminum�and�manganese�ions.�In�this�sense,�to�prevent�dam�breaches�from�

the�tailing�pond�that�can�cause�huge�hazards�to�humans�and�the�environment�it�is�required�to�

reduce�the�amount�of�wastewater�contained�in�the�mine�tailing�pond.�In�addition,�the�recovery�

of�zinc�can�provide�economic�value�to�the�process�while�solving�an�environmental�problem.�

3.4.1.�SX�LABORATORY�RESULTS��

To� accomplish� for� a� valuable� Zn� recovery,� separation� of� Zn� from� Fe� and� Ca� must� be�

obtained�since�any�further�use�of�the�Zn�liquor�product,�i.e.,�Electro�wining�(EW)�process,�will�

require�of�such�conditions.�The�recovery�of�Zn�was� investigated�to�select� the�extractant�with�

higher� efficiency� and� selectivity� between� DEHPA,� Cyanex� 272� or� Ionquest� 290.� Additionally,�

two�types�of�kerosene�were�also�evaluated.��

Since�there�are�no�reagents�commercially�available�able�to�extract�Zn�selectively�from�a�

solution�containing�Fe,�Fe�was�removed�from�the�mine�water�prior� to�the�SX�treatment�by�a�

bio�oxidation�process�followed�by�an�alkaline�precipitation�step�[13,�14]�to�obtain�a�pregnant�

leach�solution�(PLS)�without�iron.��

After� the� precipitation� step,� Fe� was� completely� removed� and� also� the� amount� of� Al�

decreased�drastically�and�Cu�dropped�by�half�(from�45.0�mg/L�to�21.7�mg/L).�The�process�did�

not�alter�the�content�of�Zn,�so�the�whole�process�of�zinc�recovery�does�not�lose�effectiveness�

3.�Results�and�discussion��

�112�

due�to�the� iron�removal�step.�The�concentration�of�the�other�metals�remained�similar�to�the�

initial�(Table�3.12).�

Table�3.12.�Solution�composition,�before�and�after�bio�oxidation�treatment�� Concentration�(mg/L)�

Element� Initial�solution� After�bio�oxidation After�precipitation�

[Fe2+]� 254� 0� 0�[Fe3+]� 446� 690� 0.2�[Zn]� 1,020� 1,020� 1,010�[Al]� 292� 250� 20�[Mn]� 265� 260� 200�[Cu]� 45� 45� 21.7�[Ca]� 600� 600� 600�[Pb]� 1.6� 1.6� 1.6�pH� 3.0� 1.9� 4.8�

Regarding�the�selectivity�experiments�performed�with�the�three�extractants�studied,�the�

obtained�results�(Figure�3.26,�3.28�and�3.30)�depicted�a�rational�reduction�of�the�recovery�of�

zinc� as� the� A:O� phase� ratio� increases� due� to� a� saturation� of� the� extractant.� In� this� sense,�

Cyanex�272�and�Ionquest�290,�used�in�a�lesser�concentration�than�DEHPA�(5%�vs�40%�(v/v)�for�

DEHPA),�present�a�plateau�at�a�A:O�>�1�which�suggests�that�the�extractant�is�saturated,�whilst�

zinc�recovery�for�DEHPA�is�still�diminishing.�

The�recovery�of�metals�achieved�by�DEHPA�was�Zn>�Ca>�Mn�>�Al�>�Cu,�and�at�A:O=1�the�

recovery� of� Zn� was� around� 75%,� but� also� other� metal� impurities� were� also� recovered,�

especially�Ca�and�Mn�(60%�and�30%�recovered,�respectively)�pointing�out�that�DEHPA�is�poorly�

selective�towards�Zn�extraction�(Figure�3.26).�In�addition,�around�80%�of�the�Al�remained�in�the�

organic�phase�(OP)�after�the�stripping�step�(Figure�3.27)�limiting�the�reuse�of�the�extractant.��

The� recovery� of� metals� obtained� for� Cyanex� 272� at� 5%� (v/v)� was� Zn>>Cu>Mn~Ca~Al�

(Figure� 3.28),� although� Mn,� Ca� and� Al� are� slightly� recovered.� In� this� sense,� Cyanex� 272�

selectivity� towards� zinc� is� higher� than� DEHPA� and,� moreover,� negligible� amounts� of� metals�

(around� 1%)�were� found� in� the� organic� phase� (Figure� 3.29)� so� the� organic� phase� employing�

Cyanex�272�can�be�reused�several�cycles�with�practically�no�regeneration.��

The�recovery�of�metals�for�Ionquest�290�was�Zn>>Al>Cu~Mn~Ca�(Figure�3.30).�The�trend�

is�similar�to�Cyanex�272�since�the�recovery�of�Zinc�at�A:O=1�is�around�40%�and�other�impurities�

are�practically�not�recovered.�In�addition,�less�than�5%�of�the�elements�analyzed�remain�in�the�

organic�phase�after�the�stripping�step�(Figure�3.31).�Thus,�contrary�to�DEHPA,�Cyanex�272�and�

Ionquest�290�selectively�extract�Zn�from�a�solution�containing�high�amounts�of�Ca�and�other�

metals�without�fouling�of�the�OP.��

3.�Results�and�discussion�

113�

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio�A/O

(a)�Recovery�DEHPA

Zn�KD80 Zn�KD100Ca�KD80 Ca�KD100Al�D80 Al�D100Mn�KD80 Mn�KD100Cu�KD80 Cu�KD100

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio�A/O

(b)�Remaining�OP�DEHPA

Zn�KD80 Zn�KD100Ca�KD80 Ca�KD100Al�D80 Al�D100Mn�KD80 Mn�KD100Cu�KD80 Cu�KD100

Figure�3.26.�%Recovery�at�different�A:O�ratios�

for�DEHPA�40%�(v/v)�Figure�3.27.�%Remaining�OP�at�different�A:O�

ratios�for�DEHPA�40%�(v/v)

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio�A/O

(a)�Recovery�Cyanex�272

Zn�KD80 Zn�KD100Ca�KD80 Ca�KD100Al�KD80 Al�KD100Mn�KD80 Mn�KD100Cu�KD80 Cu�KD100

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio�A/O

(b)�Remaining�OP�Cyanex

Zn�KD80 Zn�KD100Ca�KD80 Ca�KD100Al�KD80 Al�KD100Mn�KD80 Mn�KD100Cu�KD80 Cu�KD100

Figure�3.28.�%Recovery�at�different�A:O�ratios�for�Cyanex�272�5%�(v/v)

Figure�3.29.�%Remaining�OP�at�different�A:O�ratios�for�Cyanex�272�5%�(v/v)

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio�A/O

(a)�Recovery�IONQUEST

Zn�KD80 Zn�KD100Ca�KD80 Ca�KD100Al�D80 Al�D100Mn�KD80 Mn�KD100Cu�KD80 Cu�KD100

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio�A/O

(b)�Remaining�OP�Ionquest

Zn�KD80 Zn�KD100Ca�KD80 Ca�KD100Al�D80 Al�D100Mn�KD80 Mn�KD100Cu�KD80 Cu�KD100

Figure�3.30.�%Recovery�at�different�A/O�ratios�for�Ionquest�290�5%�(v/v)

Figure�3.31.�%Remaining�OP�at�different�A:O�ratios�for�Cyanex�272�5%�(v/v)

The� differences� observed� on� the� recovery� trends� between� DEHPA� and� the� other� two�

extractants�can�be�associated�to�their�chemical�nature,�given�that�phosphoric�extractants�(as�

DEHPA)�have�higher�affinity�for�calcium�than�phosphinic�extractants�(such�as�Cyanex�272�and�

Ionquest� 290).� The� small� differences� observed� between� Cyanex� 272� and� Ionquest� 290� are�

3.�Results�and�discussion��

�114�

explained� by� both� the� different� phosphinic� acid� concentration� (Ionquest� 290� is� 5�10%�more�

concentrated�than�Cyanex�272)�and�also�owed�different�product�impurities�in�each�extractant.�

As�a�summary,�DEHPA�reported�poor�selectivity�towards�zinc�due�to�the�coextraction�of�

manganese�and�calcium� (that� resulted� in�a�gypsum�precipitate� in� the� stripping� solution)�and�

high�amounts�of�aluminum�remained�in�the�organic�phase�after�the�stripping�step�reducing�its�

reusability.� On� the� contrary,� Cyanex� 272� and� Ionquest� 290� provided� high� zinc� selectivity�

towards� calcium� and� negligible� amounts� of� metals� were� found� in� the� organic� phase� so� no�

extractant� regeneration� step�will� be� required.� Regarding� Cyanex� 272� and� Ionquest� 290,� the�

latter� achieved� a� zinc� recovery� 5�10%� higher� and� therefore� Ionquest� 290� is� considered� the�

most� appropriate� extractant.� Considering� the� two� different� kerosene� employed� (Ketrul� D80�

and� Ketrul� D100),� no� significative� differences�were� observed� and� thus� both� of� them� can� be�

equally�feasible�for�the�recovery�of�zinc.�However,�from�an�engineering�point�of�view,�the�use�

of�Ketrul�D100�is�recommended�due�to�its�lower�flammability�compared�to�Ketrul�D80,�and�for�

that�reason�it�was�selected�as�dissolvent.�

3.4.2.�SX�PILOT�PLANT�PROCESS��

To�fulfill�the�requirements,�the�pilot�plant�may�produce�an�economically�effective�output�

and� the�overall�process� should�be�environmentally� friendly.� The�pilot�plant�process� layout� is�

depicted�in�Figure�3.32.�

SOLVENT�EXTRACTION

SOLVENT�STRIPPING

ELECTROWINNING

Pregnant�Leach�Solution

Loaded�Solvent

Weak�Electrolyte

Strong�Electrolyte

Barren�Solvent

Zinc

Raffinate

�Figure�3.32.�Inputs�and�outputs�at�the�pilot�plant�

To� satisfy� the� environmental� requirements� at� least� 95%� of� the� Zn�must� be� recovered�

from�the�effluent,�whereas�to�produce�the�economic�effective�output�the�zinc�contained�in�the�

stripping� solution� must� be� converted� to� metallic� zinc,� that� must� be� treated� in� an�

electrowinning� (EW)�plant.� To� fulfill� the�operating� conditions� for� the� EW�plant,� the� SX�plant�

should�provide�a�final�product�stream�of�90�g/L�Zn�in�the�stripping�step�(strong�electrolyte)�by�

using�a�weak�electrolyte�with�50�g/L.��

3.�Results�and�discussion�

115�

Previously�to�the�pilot�plant�operations,�computer�simulation�was�performed�to�estimate�

the� required�pilot�plant� inputs�and�outputs,� to� calculate� the�distribution�coefficients� (D)�and�

the� number� of� stages.� Experience� has� shown� that� computer� simulation� is� a� more� flexible�

design�tool�than�McCabe�Thiele�diagrams�for�pulsed�columns�[15,�16,�17].�The�results�obtained�

in�the�simulation,�collected�in�Table�3.13,�determined�that�at�a�phase�ratio�O:A=0.5�0.6,�a�two�

stage� column� is� enough� to� recover�more� than� 95%� of� the� Zn.� The� addition� of� a� third� stage�

enables�either�to�decrease�the�phase�ratio�O:A�to�0.4�or�to�work�with�a�phase�ratio�of�O:A=0.5�

and�obtain�a�recovery�of�Zn�near�to�99%,�i.e.�<10�mg/L�Zn�in�the�raffinate.�The�concentration�of�

Zn� in� the� loaded�solvent� should�be� in� the� range�of�2.2�2.8�g/L,� that� is�around�70�85%�of� the�

total�theoretical�loading�of�3.3�g�Zn/L�for�Ionquest�290�5%�(v/v),�which�is�quite�reasonable.�In�

order�to�get�a�final�solution�of�90�g/L�Zn,�the�Zn�transfer�from�the�organic�phase�to�the�strip�

phase�should�be�of�40�g/L;�to�achieve�that�value�the�stripping�should�be�run�at�a�phase�ratio�of�

O:A=20,�so�only�one�equilibrium�stage�is�required�for�the�stripping.�

Table�3.13.�Recovery�of�Zn�depending�on�the�plant�configuration�using�5%�Ionquest�290�No.�Stages� Phase�ratio�O:A� Zn�in�raff.�(mg/L)� %�Recovery��

0.50� 51� 94.7�2�

0.60� 24� 97.6�

0.35� 75� 92.1�0.40� 30� 96.8�0.45� 11� 98.9�

3�

0.5� 4� 99.6�

The�maximum� loading� obtained� experimentally� at� limiting� conditions� (by� contacting� 3�

times�the�solvent�with�corresponding�fresh�portions�of�the�effluent�at�phase�ratio�O:A=0.1)�was�

2.9�g�Zn/L.�Since�this�result�was�similar�to�the�obtained�after�a�single�contact,�it�revealed�that�

the�limiting�conditions�could�be�achieved�by�a�single�contact.�

Experiments� performed� at� the� pilot� plant�without� pH� control� (Table� 3.14)� shown� that�

without� pH� control,� the� extraction�was� quite� selective.� In� this� sense,� no�Mn,� Cu� or� Al�were�

extracted�and�only�a�small�amount�of�Ca�was�extracted.�Such�fact�is�also�confirmed�by�the�high�

values� regarding� separation� factors.� However,� the� distribution� ratio� of� Zn� (DZn)� was� small,�

especially� at� the� dilute� end� of� the� process� (phase� ratio� O:A=10).� In� addition,� the� pH� of� the�

raffinate� (final�pH)�dropped� from�2.6� to�2.1�as�O:A� increased,�despite� the�suitable�pH� for�Zn�

extraction� by� Ionquest� 290� is� above� 2.5� [18].� Furthermore,� to� avoid� Ca� co�extraction,� pH�

should�be�around�3�as�indicated�by�the�isotherms�given�in�the�online�User�Manual,�page�5�from�

Cytec�Corporation�for�Cyanex�272�and�considering�the�same�composition�of�both�Cyanex�272�

and� Ionquest� 290.� (http://www.cytec.com/specialty�chemicals/PDFs/CYANEX%20272.pdf,�

accessed�26th�December�2010).�

3.�Results�and�discussion��

�116�

Table�3.14.�Extraction�experiments�without�pH�correction,�22°C�

Aqueous�(mg/L)� Organic�(mg/L)� D�values�&�Separation�factors�Phase�ratio�O:A�

Final�pH Zn� Mn Ca� Zn Mn� Ca� DZn� DZn/DCa� DZn/DMn�

PLS 5.0� 962� 206� 763 � 0.1 2.58� 792� 208� 618� 1595� 0� 8� 2.0� 154.5� 1�104�

0.3 2.50� 692� 208� 613� 910� 0.2� 12 1.3� 66.4� 1350�

0.5� 2.31 621� 205� 605 668� 0.1� 15� 1.1� 44.4� 2260�

1 2.18� 536� 206� 624� 444� 0.1� 13� 0.8� 38.4� 1600�

2� 2.16� 467� 207� 613� 273� 0� 9 0.6� 40.9� 4200�

3 2.32� 402� 202� 597� 178� 0.3� 10 0.4� 23.9� 270�

5 2.25� 342� 203� 599� 132� 0.1� 6� 0.4� 39.9� 800�

10 2.1� 270 203� 601 76� 0.1 8 0.3� 22.5� 610�

When� adjusting� to� pH=3� (Table� 3.15)� higher� amounts� of� zinc� were� extracted� and� the�

distribution�coefficient�of�zinc�(DZn)�was�higher�than�without�pH�adjustment.�The�extraction�of�

Mn�and�Ca�still�remained�quite�low�at�pH=3�as�is�also�indicated�by�the�high�separation�factors�

obtained.�Therefore,�given�that�higher�distribution�coefficient� for�zinc� is�obtained�at� this�pH,�

the�pilot�plant�effluent�should�be�maintained�around�pH�3.�In�practice,�the�pH�adjustment�was�

achieved�by�direct�neutralization�of�both�the�acidic�raffinate�and�the�organic�solvent�(by�pre�

equilibration�with�aqueous�solution)�using�Na2CO3,�with�and�average�consumption�of�1.62�kg�

Na2CO3�per�kg�of�zinc�treated.��

Table�3.15.�Extraction�experiments�at�pH�3,�22°C�Aqueous�(mg/L)� Organic�(mg/L)� D�values�&�Separation�factors�Phase�ratio�

O:A� Zn� Mn Ca Zn Mn� Ca� DZn DZn/DCa� DZn/DMn�

PLS�(pH�5.0)� 963 213 583 �

0.1 784� 231� 531� 2878� 0 76 3.7 26.4 9�104

0.3� 343� 233� 509� 2073� 0.4� 72� 6.0� 42.9� 3�103�

0.5 182� 211� 536� 1700� 0.5� 40� 9.3 133.0� 5�103�

1� 49� 213� 547� 875� 1.5� 38� 17.9 199.9 3�103�

2� 22� 194� 534� 502� 3.4� 43� 22.8 285.0 1�103

3� 14� 127� 532� 297� 2.7� 46� 21.2� 235.6 1�103

5 4� 181� 584 192� 3� 42� 48� 685.7 2�103

10 1 183� 579 90� 1 48� 90� 1125 2�104

Shake�out� stripping� experiments�were� carried�out� at�O:A=10,�by� contacting�200�mL�of�

loaded�solvent�(LS)�containing�1.95�g/L�Zn�with�20�mL�of�H2SO4�200�g/L�(weak�electrolyte,�WE)�

containing�different�zinc�concentrations�ranging�from�40�to�90�g/L�(Table�3.16).�To�achieve�the�

required� transfer,� the� concentration�of� Zn� should� increase�by�~20�g/L,�which�was� consistent�

with�the�results�shown�in�Table�3.16.�In�all�cases,�only�5�17�mg/L�of�Zn�remained�in�the�barren�

3.�Results�and�discussion�

117�

solvent� (BS),� so� almost� all� zinc� was� recovered.� Therefore,� one� stage� of� stripping� is� enough�

regardless�the�concentration�of�Zn�in�the�stripping�solution.�

Table�3.16.�Stripping�experiments�at�phase�ratio�O:A=10,�22ºC�Aqueous�In� Aqueous�out�

Zn(g/L)� H2SO4�(g/L)� Zn�(g/L)�

40� 200� 58.8�50� 194� 68.2�60� 188� 77.8�70� 176� 91.0�80� 200� 102.8�90� 200� 115.4�

Additional� laboratory� tests� carried� out� at� the� mine� site� during� the� pilot� plant�

experiments�at�phase�ratio�O:A=20,�revealed�that�the�loaded�solvent�from�the�pilot�plant�was�

efficiently�stripped�in�one�contact,� i.e.�one�stage,�by�the�strip�solution�used�in�the�pilot�plant�

experiments,�using�a�weak�electrolyte�with�~50�g�Zn/L,�producing�an�SE�containing�90�g/L�Zn,�

i.e.�a�zinc�transfer�of�40�g/L,�as�it�was�required�for�the�EW�plant.��

Preliminary�hydraulic�tests�at�the�pilot�plant�showed�that�the�available�flux� is�above�30�

m3/m2/h� in� both� columns.� Given� that� it�was� proven� that� only� one� stage� is� required� for� the�

stripping,� this� step� was� not� further� optimized� and� was� run� mainly� to� produce� BS.� It� was�

operated�at�a�flux�of�40�m3/m2/h�(35�l/h�solvent).�The�pulsing�of�the�columns�had�an�amplitude�

of�15�mm�and�a�frequency�of�1�Hz.�The�flow�rate�of�the�WE�through�the�pump�was�5�7�L/h.�The�

average�value�of�Zn�in�the�BS�was�about�20�mg/L�Zn.�

Table�3.17.�Extraction�in�organic�and�aqueous�dispersion�continuities�Zn�(mg/L)�

� Feed�(L/h)� BS�(L/h)� Flux�(m3/m2/h)�pH�Raff.� Raff.� LS�

110 55� 33� 2.7� 11� 1,910�130� 60� 38� 2.8� 55� 1,880�

Organic�continuous�dispersion�

130� 60� 38� 2.9� 11� 1,800�150� 70� 44� 2.9� � 1,880�150� 70� 44� 3.1� � 2,520�

Aqueous�continuous�dispersion�

150� 70� 44� 2.9� � 2,240�

Three� tests� with� both� organic� continuous� and� aqueous� continuous� dispersion� (Table�

3.17)�were�undertaken�to�determine�the�preferred�dispersion.�During�both�organic�continuous�

and�aqueous�continuous�runs,�the�temperature�rose�from�25°C�in�the�morning�to�34°C�in�the�

evening,�facilitating�the�comparison�between�both�dispersions�results.�Every�test�took�5�hours,�

long�enough�to�reach�steady�state�and�the�phase�ratio�was�kept�at�A:O=2.1�during�all�the�test�

work.� The� results� were� similar� for� both� dispersions.� The� concentration� of� Zn� in� the� LS� was�

around�2,000�mg/L�and�in�the�raffinate�below�50�mg/L,�indicating�than�more�than�95%�of�the�

Zn�was�recovered.�Thus,�the�extraction�process�operated�successfully�with�both�aqueous�and�

3.�Results�and�discussion��

�118�

organic� continuous� dispersions� at� 23�34°C.� As� the� available� flux� and� recovery� with� both�

dispersions�were�similar,�it�is�preferable�to�use�the�aqueous�continuous�dispersion�as�there�is�a�

lower�expenditure�on�solvent.�Using�an�aqueous�continuous�dispersion�the�danger�of�fire�due�

to�kerosene�ignition�is�also�diminished.�

The�stripping�of�the�LS�(containing�around�2�g/L�Zn)�achieved�a�SE�with�30�40�g/L�Zn�(a�

zinc�transfer�of�30�40�g/L)�whilst�less�than�50�mg/L�Zn�in�the�raffinate�at�a�flux�of�45�m3/m2/h�

using� an� aqueous� continuous� dispersion� at� O:A=20.� The� stripping� column� worked� well� and�

supplied�the�required�BS�to�the�extraction.�Given�that�the� laboratory�tests�proved�that�there�

was� no� need� for� an� extra� column,� one� stage� of� mixer�settler� was� sufficient� to� obtain� the�

required�zinc�transfer�of�40�g/L�with�barren�solvent�containing�~50�mg/L�Zn.�

As�a�summary,�given�that�the�recycling�of�the�organic�phase�lead�to�a�relative�importance�

of� the� extractant� costs,� Ionquest� 290� was� selected� as� the� most� suitable� extractant� for� the�

target�stream�due�to�its�higher�selectivity�and�loading�capacity�towards�Zn�extraction.�Ionquest�

290�avoids�the�necessity�of�scrubbing�the�gypsum�precipitate�in�the�strip�liquor�as�well�as�the�

regeneration�of�the�solvent�after�high�amounts�of�aluminum�are�not�stripped�if�compared�with�

DEHPA� (a� cheaper�extractant� compared� to� Ionquest�290�and�Cyanex�272).�As�both� solvents,�

Ketrul� D80� and� Ketrul� D100,� showed� similar� behavior,� Ketrul� D100� was� the� solvent�

recommended�owing�its�lower�volatility�and�flammability.�The�pilot�plant�proved�the�feasibility�

of�the�process,�obtaining�a�zinc�recovery�of�95%�and�leaving�less�than�50�mg/L�in�the�raffinate.�

The�stripping�was�efficient�and�only�a�single�stage�at�O:A=20�was�required�to�achieve�a�transfer�

of�40�g/L.�For�a�Zn�price�above�US$2/kg�the�operating�costs�are�covered�while,�additionally,�a�

serious�environmental�problem�is�solved.�

3.�Results�and�discussion�

119�

3.5.� FE�LOADED� MATERIALS� FOR� THE� REMEDIATION� OF� ORGANIC� AND� INORGANIC�

CONTAMINATED�WASTE�WATERS�

Here,� are� summarized� the� results� obtained�by�using� Fe�loaded�materials� to� remediate�

organic� and� inorganic� wastewaters.� In� this� sense,� organic� pollutants� such� as� dyes� and�

persistent�organic�pollutants�were�degraded�by�following�Fenton�treatment�using�as�a�catalyst�

Fe�loaded� materials.� Such� Fe�loaded� materials� were� also� applied� to� the� removal� of� arsenic�

from�polluted�wastewaters� taking�advantage�of� the�affinity�of� arsenic�with� iron� compounds.�

The� first� step� on� these� processes� includes� the� loading� of� the� material� with� Fe.� The� results�

obtained�for�loading�the�USY�zeolite�with�Fe(III)�are�presented�in�Figure�3.33�and�indicated�a�Fe�

loading�increase�with�time.�One�hour�was�selected�as�appropriate�loading�time.��

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

1h� 3h 6h 3*6h 3�days

[Fe3

+ ]�(W

t.%)�

Time�of�Fe3+�exchange

0

0.5

1

1.5

2

2.5

3

0.01M 0.05M 0.1M 0.2M 0.5M

[Fe3

+ ]�(W

t.%)�

Concentration�of�the�Fe�loading�solution

Figure�3.33.�Fe�concentration�on�USY�zeolite�at�1h,�3h,�6h,�3�days�and�6�cycles�of�3h�of�Fe�exchange�(error�bars�correspond�to�the�standard�deviation�on�the�determination)�

Figure�3.34.�Fe�concentration�of�USY�zeolite�between�0.01�to�0.5M�initial�Fe(NO3)3�

concentration�(error�bars�correspond�to�the�standard�deviation�on�the�determination)�

A�decrease�on�the�Fe�content�is�observed�when�the�concentration�of�the�loading�solution�

is�increased�(see�Figure�3.34)�being�explained�by�the�formation�of�polynuclear�Fe�complexes�at�

high� Fe� concentration,� so� less� Fe� is� available� for� the� exchange�with� USY� [19].� On� the� other�

hand,�at�a�very�low�Fe�concentration�(0.01M),�not�enough�Fe�is�available�to�occupy�all�the�sites,�

thus�being�the�maximum�loading�at�a�concentration�of�Fe(NO3)3�0.05M,�hence�being�selected�

as� the�most�appropriate� loading�concentration.�Under� this�conditions,�at� room�temperature,�

the�amount�of�Fe�introduced�into�the�USY�zeolite�was�2.7±0.2�wt.%.�According�to�Neamtu�et�al.�

[20]�1.69wt.%�of�Fe�can�be� introduced�when�exchanging�USY�three�times�during�6h�using�an�

excess� of� Fe(NO3)3� 1M� at� 80ºC.� Our� new� process� achieved� 2.7±0.2�wt.%.Fe� by� using�milder�

conditions.� Therefore� these� conditions�were� also� implemented� to� other�materials�with� high�

exchange�properties�such�as�zeolite�Y,�clinoptilolite,�montmorillonite�and�Forager�sponge.��

3.�Results�and�discussion��

�120�

3.5.1.�FE�LOADED�MATERIALS�APPLIED�AS�FENTON�CATALYSTS�

Two� different� low� cost� materials� such� the� natural� zeolite� clinoptilolite� and� the� clay�

montmorillonite� K�10� (MMT)� along� with� the� commercial� synthetic� zeolite� USY� were� Fe�

exchanged�for�their�evaluation�as�Fenton�catalysts.�The�amount�of�Fe�on�each�material�before�

and�after�loading�is�shown�in�Figure�3.35.�From�the�obtained�results�it�can�be�pointed�out�that�

before�the�exchanging�process�no�Fe�was�detected�in�the�USY�whilst�MMT�structure�contained�

2.1±0.1� wt.%� and� clinoptilolite� 1.0±0.1� wt.%� of� Fe.� The� initial� Fe� content� of� MMT� and�

clinoptilolite�is�structural�and�due�to�their�natural�origin�related�to�soils�usually�rich�in�Fe.�After�

the� loading� process,� the� amount� of� Fe� on� MMT� was� 4.3� wt.%,� so� 2.4� wt.%� of� Fe� were�

introduced� onto�MMT.�On� the� other� hand,� the� clinoptilolite� amount� of� Fe� after� the� loading�

with�Fe�was�2.1�wt.%,�so�only�1.0�wt.%�Fe�was�introduced�onto�the�clinoptilolite.�These�values�

are�strongly�related�to�the�surface�area�of�each�material,�as�USY�has�bigger�surface�area�than�

montmorillonite� and,� montmorillonite� bigger� than� clinoptilolite� (Specific� surface� area:�

USY=730� m2/g;� MMT=271� m2/g;� clinoptilolite=31.7� m2/g).� Overall,� this� process� has�

demonstrated�to�be�suitable�for�loading�Fe�low�cost�materials�such�as�natural�zeolites�and�clays�

with�different�surface�areas.��

00.51

1.52

2.53

3.54

4.55

USY MMT Clinoptilolite

[Fe3

+ ]�(W

t.%)�

After�loading

Before�loading

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

200 250 300 350 400 450 500 550 600

Abs

orba

nce

Wavelength (nm)

�=218nm

�=324nm

�=516nm

Figure�3.35.�Fe�content�of�the�materials�before�and�after�the�loading�with�Fe(NO3)3�

Figure�3.36.�UV�Vis�spectra�of�AR14�0.05�mM�

To�evaluate�the�feasibility�of�such�Fe�loaded�materials�as�heterogeneous�catalysts�on�the�

degradation�of�organic�compounds,�the�decolorisation�of�a�model�dye�has�been�employed.�The�

acid�dye�used,�Acid�Red�14� (AR14),�has� two�peaks�at�UV�region� (218�nm�and�324�nm)�and�a�

characteristic� peak� at� visible� region� at� 516� nm� (Figure� 3.36).� The� peak� at� 218� nm� can� be�

attributed�to�the�absorbance�of�the�naphthalene�groups,�the�peak�at�324�nm�corresponds�to�

the� electron� conjugation� of� naphthalene� rings� with� the� �N=N�� group,� whereas� the� peak� at�

516nm�is�related�to�the�high�conjugated�structure�of�the�whole�dye�molecule�that�confers�its�

characteristic�color�to�the�dye�[21].��

3.�Results�and�discussion�

121�

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90

Time�(min)

%Decolorization��=516nm

Hom.�Cat.Fe�USYFe�MMTFe�Clinoptilolite

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90

Time�(min)

%Degradation��=324nm

Hom.�Cat.Fe�USYFe�MMTFe�Clinoptilolite

Figure�3.37.�Kinetics�of�AR14�decolorisation�(�=516�nm)�for�Fe�USY,�Fe�MMT,�Fe�

Clinoptilolite�and�homogeneous�catalysis�(Hom.�Cat.)�

Figure�3.38.�Kinetics�of�AR14�mineralization�(�=324�nm)�for�Fe�USY,�Fe�MMT,�Fe�

Clinoptilolite�and�homogeneous�catalysis�(Hom.�Cat.)

The� absorbance� was� measured� at� regular� intervals� of� time� at� its� characteristic�

wavelength� (�=516� nm)� to� determine� kinetics� of� decolorisation� of� the� dye.� The� degree� of�

mineralization�of�the�dye�was�followed�through�the�characteristic�wavelength�related�to�bond�

breaking�(�=324�nm).�From�the�obtained�results�(Figure�3.37)�it�can�be�stated�that�all�Fe�loaded�

materials� are� able� to� discolor� AR14� 0.05mM� in� less� than� 60� min.� Moreover,� according� to�

literature�[22,�23]�the�Fenton�reaction�follows�a�pseudo�first�order�equation�[ln(C/Co)=�Ka�t],�so�

the�linear�region�of�the�kinetic�experiments�can�be�used�to�obtain�the�apparent�first�order�rate�

constant�(Ka)�for�all�the�Fe�loaded�materials�studied.�Thus,�taking�into�account�the�Ka�values�for�

the� Fe�loaded� materials� together� with� the� homogeneous� catalysis� (Table� 3.18),� it� can� be�

pointed�out�that�the�catalytic�activities�followed�the�order�Homogeneous�catalysis>Fe�USY>Fe�

MMT>Fe�Clinoptilolite.��

Table�3.18.�Apparent�first�order�rate�constant�(Ka)�for�the�exchanged�materials�Catalyst� Ka�(min�1)� R2� [Fe3+]solution�(mg/L)�

Homogeneous�catalysis� 0.368±0.02� 0.995� 17.6±0.5�

Fe�USY� 0.311±0.009� 0.998� <0.2�

Fe�MMT� 0.143±0.003� 0.998� 0.5±0.2�

Fe�Clinoptilolite� 0.041±0.002� 0.994� 1.4±0.2�

In� this� regard,� the� reaction� kinetics� for� Fe�USY� is� comparable� to� the� homogeneous�

catalysis,� whilst� Fe�MMT� and� Fe�clinoptilolite� showed� slower� kinetics.� These� results� can� be�

associated�with�the�amount�of�Fe�loaded�after�the�Fe�exchanging�(not�to�the�total�Fe�content),�

that�indicates�that�the�structural�Fe�present�in�MMT�and�clinoptilolite�is�not�accessible�to�act�as�

catalyst�of� the�Fenton� reaction.�Moreover,�negligible�amounts�of�Fe�were� released� from�the�

Fe�loaded�materials,�as�small�amounts�of�Fe�were�found�in�solution�after�the�reaction,�so�the�

reaction�mainly�occurs�due�to�the�Fe�linked�to�the�support�(Table�3.18).�

3.�Results�and�discussion��

�122�

From�the�values�obtained�at�324�nm�(Figure�3.38),�mostly�related�to�the�degradation�of�

naphthalene,�and�thus�to�the�mineralization�of�the�dye,�it�can�be�observed�that�the�maximum�

degradation�achieved�was�95%,�thus�the�mineralization�of�AR14�was�not�complete,�even�after�

90�min�of�reaction.� In�this�regard,�the�analysis�by�GC�MS�of�the�solution,�after�being�treated,�

revealed� only� oxalic� acid� and� malonic� acid� as� degradation� products.� These� compounds� are�

refractory�to�oxidation�by�Fenton�processes�as�it�has�been�demonstrated�that�low�chain�acids�

are�difficult�to�degrade�by�the�radical�hydroxyl�[24].�It�is�also�noteworthy�that�Fe�USY�is�able�to�

achieve�similar�kinetics�to�the�homogeneous�catalysis�process�(maximum�mineralization�at�20�

min� for� both)� whereas� Fe�MMT� and� Fe�clinoptilolite� lasted� 25� and� 50� min� respectively� to�

achieve�maximum�mineralization.��

To� complement� the� feasibility� of� the� Fe�loaded� materials� studied� to� degrade� organic�

pollutants,�the�degradation�of�two�refractory�organic�compounds,�acetic�acid�and�phenol,�was�

also�evaluated.�Such�degradation�was�followed�by�measuring�chemical�oxygen�demand�(COD)�

as� a� measure� of� the� amount� of� organic� compounds� in� water.� The� results� obtained� for� the�

degradation� of� the� model� acetic� acid� solution� (COD=5300� ppm)� and� phenol� solution�

(COD=11900� ppm)� for� the� Fe�loaded� catalysts� together� with� the� results� obtained� for� the�

homogeneous�catalysis�and�the�amount�of�Fe�in�solution�are�given�in�Table�3.19.��

Table�3.19.�Acetic�acid�and�phenol�removal�by�Fe�supported�materials�and�homogeneous�catalysis�� Acetic�COD�removal� Phenol�COD�removal� [Fe3+]solution�(mg/L)�USY� 34.6±5%� 93±2%� 0.9±0.4�MMT� 37.8±3%� 94±4%� 2.5±0.4�

Clinoptilolite� 30.5±4%� 87±2%� 1.2±0.4�Homogeneous�catalysis� 25±4%� 85±3%� 18±2�

Mineralization�of� acetic� acid� is�partly� achieved� for� all� heterogeneous� catalysts� and� the�

homogeneous�catalysis,�with�a�COD�diminution�of�20�40%.�Comparing�the�values�obtained�for�

each� of� the� Fe�exchanged� materials� it� can� be� observed� a� higher� performance� over� the�

homogeneous�catalysis.�All� the�Fe�loaded�materials� reached�almost�30%�of�COD�degradation�

whilst� homogeneous� catalysis� only� achieved� 25%� of� COD� removal.� Regarding� the� values�

obtained� for� phenol,� almost� complete�mineralization� of� the� solution�was� achieved� reaching�

about�90%�of�COD�removal.�Again,� the� results�obtained� for� the�Fe�exchanged�catalysts�were�

higher� than� for� the� homogeneous� catalysis.� Among� the� supported� catalysts,� Fe�USY� and� Fe�

MMT�achieved�higher�COD�removal�than�clinoptilolite�mainly�due�to�their�higher�Fe�content.�

These�values�are�similar� to�those�reported�before�[25],�where�20%�and�96%�of�COD�removal�

were�achieved�for�acetic�acid�and�phenol,�respectively,�indicating�that�the�process�is�also�viable�

for� the� removal� of� persistent� organic� pollutants.� Moreover,� iron� hydroxyoxides� were� not�

3.�Results�and�discussion�

123�

formed,� so� there�was�no�need� to� remove� the� red� sludge� caused�by� iron�hydroxyoxides� as� it�

happened�when�using�the�homogeneous�catalysis.�

Finally,�column�tests�applying�similar�conditions�to�those�employed�in�batch�experiments�

were�performed.�Given�the�small�particle�size�of�the�materials�used�in�batch�experiments,�the�

column�blocked�avoiding�the�circulation�of�the�solution.�However,�given�the�natural�origin�of�

the� clinoptilolite� zeolite,� it� was� able� to� be� grind� and� milled� to� obtain� different� grain� sizes�

allowing�its�use�in�column.�As,�the�clinoptilolite�at�grain�size�0.2�2mm�did�not�block�the�column,�

it�was�employed�for�column�experiments.�In�this�sense,�three�columns�were�filled�with�4.6g�of�

clinoptilolite�grain�size�0.2�2mm�and�loaded�with�iron�by�circulating�100�ml�of�Fe(NO3)3�0.05M�

at� room� temperature� at� 2mL/min� in� countercurrent.� The� amount� of� Fe� introduced� into� the�

clinoptilolite�was�1.0±0.1%�wt,�equal�to�the�obtained�for�the�batch�process�using�clinoptilolite�

fine�grain�size.��

Discoloring�of�100�mL�AR14�0.05�mM�was�done�over�one�of�the�columns�obtaining�total�

discoloring� of� the� dye� in� less� than� 15� minutes� and� achieving� kinetics� of� discoloring�

(Ka=0.365±0.02,�R2=0.993)�comparable�to�the�homogeneous�catalysis�and�the�Fe�USY�material.�

This� high� discoloring� kinetic� is� explained� by� the� fact� that� in� column� processes� the� solution�

contacts� several� times�with� fresh� catalyst� along� the� column.� In� that� case,� the� amount� of� Fe�

loaded� into� the� zeolite� is� not� the� key� factor� for� the� Fenton� reaction,� due� to� in� column�

processes�the�contact�of�solution�and�catalyst�is�enhanced.�Moreover,�when�the�reaction�was�

done� over� 100� mL� of� acetic� acid� or� 100� mL� of� phenol,� COD� removal� was� 29±4� and� 92±4,�

respectively,� thus� providing� similar� COD� removal� than� the� batch� process� with� the� other�

materials�studied�and�the�homogeneous�catalysis.�This�fact�demonstrates�the�feasibility�of�Fe�

loaded�clinoptilolite�as�heterogeneous�Fenton�catalysts�also�in�column.

3.�Results�and�discussion��

�124�

3.5.2.�FE�LOADED�MATERIALS�APPLIED�TO�ARSENIC�REMOVAL�

Arsenic� contamination� in� groundwater� generates� widespread� human� health� disasters�

around� the� world� (especially� in� Southeast� Asia).� In� this� sense,� besides� their� application� as�

catalysts� in� Fenton�processes,� Fe�loaded�materials� can�be� also� employed� for� the� removal� of�

arsenic�given� the�affinity�of�Fe�compounds�with�arsenic.� In� this� sense,� three�different�Fe(III)�

bearing� materials� namely� zeolite� USY� (Ultra�Stable� Steamed� Y� zeolite),� zeolite� Y� (ZY)� and�

Forager�sponge�(Sp)�have�been�tested�as�arsenic�sorbents.�In�this�regard,�the�characterization�

of� these�materials�by�FP�XRF�and�XAFS�techniques�can�shed� light�onto�the�different�sorption�

mechanisms�of�arsenic�into�such�materials.��

Zeolite� USY� (USY),� zeolite� Y� (ZY)� and� Forager� sponge� (Sp)� were� loaded� with� Fe(III)�

following� the�methodology�described� in� section�3.5.1� to�obtain� the�materials�USY3,� ZY3�and�

Sp3.�In�this�sense,�under�the�same�conditions,�ZY�achieved�greater�Fe�content�than�USY�or�Sp�

(Table�3.20).�As�it�was�concluded�in�the�previous�section,�the�loading�of�Fe�into�the�materials�is�

strongly�related�to�its�specific�surface�area,�thus�as�zeolite�Y�has�a�specific�surface�area�higher�

than� zeolite� USY,� its� Fe� loading� was� superior� (Specific� surface� area:� USY=730�m2/g;� ZY=900�

m2/g).�Although�specific�surface�area�plays�an�important�role�on�the�loading�capacity�of�Fe�on�

the� materials,� it� has� to� be� taken� into� account� also� the� number� of� functional� groups.� Thus,�

although� Forager� spongeis� has� less� surface� area� than� zeolites� (Specific� surface� area=� 10�15�

m2/g�according�to�producer)�the�high�content�of� functional�groups�allows�higher�Fe� loadings.�

After�the�arsenic�sorption�process,�it�can�be�observed�that�for�both�studied�zeolites,�a�relation�

between� the� As� sorbed� and� the� content� of� iron� can� be� depicted� (equal� As:Fe� ratio).�

Nevertheless,�Forager�sponge,�has�an�As:Fe�ratio�higher�than�for�the�zeolites�mainly�owed�to�

tertiary�amine�salt�groups�contained�in�the�sponge�that�can�bind�anionic�contaminants,�such�as�

arsenic,�chromate�or�uranium�oxide�species.��

Table�3.20.�Fe�and�As�content�of�the�USY,�ZY�and�sponge�Material� [Fe]�(mg/Kg)� [As]�(mg/Kg)� %Arsenic�Adsorption� As:Fe�ratio�USY3�As� 46,000±50� 16,560±30� 41±3� 0.4�ZY3�As� 88,930±70� 36,140±40� 90±5� 0.4�Sp3�As� 46,730±50� 29,640±40� 74±4� 0.6�

A� better� understanding� of� the� differences� regarding� arsenic� sorption� onto� those�

materials� is� given�by� the�analysis�of� EXAFS� spectra.� In� this� sense,� the� theoretical�paths� from�

scorodite� (FeAsO4�2H2O)� were� used� to� determine� bond� lengths� and� coordination� numbers�

regarding� to� the�presence�of� Fe�As� bonds� in� its� structure.� The� goodness� of� these�paths�was�

validated� by� the� calculation� of� bond� lengths� and� coordination� number� for� rösslerite�

(MgHAsO4�7H2O)�and�ferrihydrite� (Fe2O3�0.5H2O),�the�standards�measured�at�the�synchrotron�

3.�Results�and�discussion�

125�

as� scorodite� was� not� available.� The� results� obtained� by� using� the� theoretical� paths� for� the�

fitting�of� the�experimental� spectra�of� scorodite�were�concordant�with� the� theoretical�known�

values� (Table� 3.20� and� Figure� 3.38).� In� this� sense,� the� theoretical� values� for� rösslerite� are� 2�

coordination�shells�containing�2�Oxygen�atoms�each�at�1.66��and�1.70�,�respectively,�while�

the�experimental�results�obtained�were�1�coordination�shell�containing�3.6±0.3�Oxygen�atoms�

at�1.70�.�Given�that�the�distances�and�the�coordination�number�are�very�similar,�it�can�be�said�

that�the�paths�are�correct�and�can�be�used�to�fit�the�spectra�of�the�unknown�samples.�

Table�3.20.�Theoretical�and�fit�values�for�rösslerite�� Theoretical�values� Fit�results�

� R�()� CN� R()� CN� ��(10�3��2)� E0�(eV)�

As�O� 1.66� 2�As�O� 1.70� 2�

1.70±0.01� 3.6±0.3� 3.7±1� 1.4±1.7�

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5

0 1 2 3 4 5 6 7 8

FT(X

(k)·k

2)

r, A

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2 4 6 8 10 12

x(k)

·k2

k, A

Figure�3.38.�EXAFS�spectra�for�Rosslerite.�a)�Fourier–transformed�spectra�(k2�weighted)�and�b)�As�K�edge�spectra.�(min�k=3.89;�max�k=11.54;�min�R=0.57;�max�R=2.04)�

The�distances�and�coordination�shells�obtained�for�each�of�the�Fe�loaded�materials�using�

the�rösslerite�paths�are�given� in�Table�3.21.�The�spectra�and� fit� spectra� for�arsenic�adsorbed�

onto�USY3�are�given�in�Figure�3.39,�for�arsenic�adsorbed�onto�ZY3�in�Figure�3.40�and�for�arsenic�

adsorbed�onto�Fe�loaded�sponge�in�Figure�3.41.�

Table�3.21.�Fit�results�for�Sp3,�USY3�and�ZY3�first�and�second�coordination�shells�

Material�Coordination�

shell�R(�)� CN� ��(10�3���2)� E0�(eV)�

1st�=�As�O� 1.69±0.02� 4.2±0.2� 3� 2±2�Sp3�As(R=14.6%)�

2nd�=�As�Fe� 3.23±0.07� 2.4±0.9� 8� �3±7�1st�=�As�O� 1.69±0.02� 4.7±0.2� 3� 2±2�

USY3�As(R=12.0%)�2nd�=�As�Fe� 3.19±0.05� 3.3±0.8� 8� �8±5�1st�=�As�O� 1.69±0.02� 4.4±0.2� 3� 3±2�

ZY3�As(R=12.3%)�2nd�=�As�Fe� 3.21±0.05� 3.4±0.8� 8� �5±5�

The� first� coordination� shell� around� As� (As�O)� is� at� similar� distance� and� coordination�

numbers�are�almost�equal�for�all�the�materials�(Sp3,�USY3�and�ZY3).�The�main�differences�are�

observed� for� the� second�coordination� shell� (As�Fe),�which� is�at� the� same�distance� for�all� the�

Rosslerite, As(V) standard R2=8.7%

Rosslerite, As(V) standard

3.�Results�and�discussion��

�126�

materials�although�the�coordination�number�is�slightly�higher�for�both�zeolites�(USY3�and�ZY3)�

than� for� the� sponge� (Sp3).� Such� fact� can�be�attributed� to� the�As� in� the� sponge�which� is� not�

coordinated� to� the� Fe� loaded� but� coordinated� to� the� amine� groups,� so� the� coordination�

number�is�decreased.�

-4

-3

-2

-1

0

1

2

3

4

0 1 2 3 4 5 6 7 8

FT(X

(k)·k

2)

r, A

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2 4 6 8 10 12

x(k)

·k2

k, A

�Figure�3.39.�USY3�As.�a)�Fourier�transformed�spectra�(k2�weighted)�and�b)�As�K�edge�spectra.��The�model�fits�are�shown�as�grey�line.�(min�k=3.89;�max�k=11.54;�min�R=0.57;�max�R=3.07.�

Constraints:��1=0.003��2;��2=0.08��2)�

-4

-3

-2

-1

0

1

2

3

4

0 1 2 3 4 5 6 7 8

FT(X

(k)·k

2)

r, A

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2 4 6 8 10 12

x(k)

·k2

k, A

�Figure�3.40.�ZY3�As�a)�Fourier�transformed�spectra�(k2�weighted)and�b)�As�K�edge�spectra.��The�model�fits�are�shown�as�grey�line.�(min�k=3.89;�max�k=11.54;�min�R=0.57;�max�R=3.07.�

Constraints:��1=0.003��2;��2=0.08��2)�

-4

-3

-2

-1

0

1

2

3

4

0 1 2 3 4 5 6 7 8

FT(X

(k)·k

2)

r, A

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2 4 6 8 10 12

x(k)

·k2

k, A Figure�3.41.�Sp3�As�a)�Fourier�transformed�spectra�(k2�weighted)�and�b)�As�K�edge�spectra.�The�model�fits�are�shown�as�grey�line.�(min�k=3.89;�max�k=11.54;�min�R=0.57;�max�R=3.07.�

Constraints:��1=0.003��2;��2=0.08��2)�

USY3-AsR2=8.7%

USY3-As

ZY3-As R2=8.7%

ZY3-As

Sp3-As

2

Sp3-As

3.�Results�and�discussion�

127�

Different� surface� species�have�been�observed� from�EXAFS� studies� concerning� arsenate�

adsorption�on� iron�oxides� (Figure� 3.42).� Arsenate� can�be� adsorbed�on� iron�oxides�mainly� as�

bidentate� complexes� resulting� from� corner�sharing� between� AsO4� tetrahedra� and� two� FeO6�

octahedra� (namely� 2C).� Furthermore� monodentate� complexes� from� corner�sharing� between�

AsO4�tetrahedra�and�FeO6�octahedra�(namely�1V)�were�also�inferred�[26].�Several�other�studies�

proposed�also�bidentate�edge�sharing�between�AsO4�tetrahedra�and�a�free�edge�of�the�same�

FeO6� octahedra� (namely� 2E)� [27,� 28].� Each� type� of� coordination� has� a� different� bond� length�

(Table� 3.22).� In� this� sense,� given� the�distances�obtained� for� the� Fe�loaded�materials� studied�

and� the� bond� length� for� each� type� of� coordination,� it� can� be� inferred� that� the� arsenate� is�

complexed�with�the�Fe�of�the�Fe�loaded�materials�as�a�bidentate�corner�sharing�bond.��

Figure�3.42.�Possible�surface�complexes�on�iron�oxide�hydroxides�

Table�3.22.�Interatomic�distances�according�the�type�of�complex�Name�of�complex� Type�of�complex� Bond�sharing� R�As�Fe�(Å)�

1V� Monodentate�complex� Corner�sharing� 3.6�2C� Bidentate�complex� Corner�sharing� 3.26�2E� Bidentate�complex� Edge�sharing� 2.8�

3.�Results�and�discussion��

�128�

3.6.�REFERENCES�[1]� El�Hachimi,�M.L.;� El� Founti,� L.;� Bouabdli,� A.;� Saidi,�N.;� Fekhoui,�M.;� Tasse,�N.� (2007).� Pb� and�As� in�mining�alkaline�waters:�contamination,�comportment�and�risks�(the�Zeida�abandoned�mine,�Morocco).�Rev.�Sci.�Eau�20:�1�13.�[2]�Plassard,�F.;�Winiarski,�T.;�Petit�Ramel�M.�(2000).�Retention�and�distribution�of�three�heavy�metals�in�a�carbonated�soil:�comparison�between�batch�and�unsaturated�column�studies.� J.�Contam.�Hydrol.�42:�99–111.�[3]�Chuan,�M.C.;�Shu,�G.Y.;�Liu,�J.C.�(1996).�Solubility�of�heavy�metals� in�a�contaminated�soils:�effect�of�redox�potential�and�pH.�Water,�Air�Soil�Pollut.�90:�543�556.�[4]�Gerritse,�R.G.;�Van�Driel,�W.� (1984).� The� relationship�between�adsorption�of� trace�metals,� organic�matter,�and�pH�in�temperate�soils.�J.�Environ.�Qual.�13:�197�204.�[5]�Förstner,�U.�(1998).�Integrated�Pollution�Control,�edited�by�A.�Weissbach�and�H.�Boeddicker,�81–130.�Springer�Verlag.�Germany.�[6]� Kim,� C.� S.;� Brown� Jr,� G.E.;Rytuba,� J.� J.� (2000).� Characterization� and� speciation� of�mercury�bearing�mine�wastes�using�X�ray�absorption�spectroscopy.�Sci.�Tot.�Environ.�261:�157–168.�[7]� Kim,� C� S.;� Rytuba,� J.J.;Brown� Jr,� G.E.� (2004).� Geological� and� anthropogenic� factors� influencing�mercury�speciation�in�mine�wastes:�an�EXAFS�spectroscopy�study.�Appl.�Geochem.�19:�379–393.�[8]� Sumozas,�R.� (2005).�Arquitectura� industrial� en�Almaden:�Antecedentes,�Génesis� y�Extensión�de�un�modelo.�PhD�thesis,�Castilla�La�Mancha�University,�Spain.�[9]�Gray,�J.E.;�Hines,�M.E.;�Higueras,�P.;�Adatto,�I.;�Lasorsa,�B.K.�(2004).�Mercury�Speciation�and�Microbial�Transformations�in�Mine�Wastes,�Stream�Sediments,�and�Surface�Waters�at�the�Almadén�Mining�District,�Spain.�Environ.�Sci.�Technol.�38:�4285–4292.�[10]�Dickson,�F.W.;�Tunell,�G.�(1959).�The�stability�relations�of�cinnabar�and�metacinnabar.�Am.�Mineral.�44,�471–487.�[11]� Higueras,� P.;� Oyarzun,� R.;� Biester,� H.;� Lillo,� J.;� Lorenzo,� S.� (2003).� A� first� insight� into� mercury�distribution�and�speciation�in�soils�from�the�Almaden�mining�district,�Spain.�J.�Geochem.�Explor.�80,�95–104.�[12]�Kocman,�D.�(2008).�Mass�balance�of�mercury�in�the�Idrijca�river�catchment.�PhD�thesis,�Jozef�Stefan�International�Postgraduate�School,�Slovenia.�[13]� Mazuelos,� A.;� Carranza,� F.;� Palencia,� I.;� Romero,� R.� (2000).� High� efficiency� reactor� for� the�biooxidation�of�ferrous�iron.�Hydrometallurgy�58(3):�269�275.�[14]� Carranza,� F.;� Iglesias,� N.;� Romero,� R.;� Palencia,� I.� (1993).� Kinetics� improvement� of� high�grade�sulfides�bioleaching�by�effects�separation.�FEMS�Microbiol.�Rev.�11(1�3):�129�138.�[15]�Grinbaum,� B.� (1992).�How� to� fit� a� simulator� to� a� real� liquid�liquid� extraction� plant,� in:� Sekine,� T.�(Ed.),�Solvent�Extraction.�1235�1240.�Elsevier.�The�Netherlands.�[16]�Grinbaum,�B.� (1993).� The�development�of� liquid�liquid�extraction�processes�using� simulation”,� in:�Longsdail,�D.H.�and�Slater,�M.J.� (Eds.),� Solvent�Exttraction� in� the�Process� Industries.�715�722.�Elsevier.�The�Netherlands.�[17]�Gottliebsen,�K.;�Grinbaum,�B.;�Chen,�D.;�Stevens,�G.� (2000).�Recovery�of�sulfuric�acid� from�copper�electrolyte�bleeds.�Hydrometallurgy.�56:�293�307.�[18]� Tsakiridis,� P.E.;� Oustadakis,� P.;Katsiapi,� A.;� Agatzini�Leonardou,� S.� (2010).� Hydrometallurgical�process�for�zinc�recovery�from�electric�arc�furnace�dust�(EAFD).�Part�II:�Downstream�processing�and�zinc�recovery�by�electrowinning.�J.�Hazard.�Mater.�179:�8–14.�[19]� Schwertmann,� U.;� Cornell,� R.M.� (2000).� Iron� oxides� in� the� laboratory.� Preparation� and�characterization.�2nd�edition.�Viley�VCH.�Germany.�[20]�Neamtu,�M.;�Zaharia,�C.;�Catrinescu,�C.;�Yediler,�A.;�Macoveanu,�M.;�Kettrup,�A.�(2004).�Fe�loaded�Y�zeolite�as�catalyst�for�wet�proxide�oxidation�of�reactive�azo�dye�Procion�Marine�H�EXL.�Applied�Catalysis�B:�Environmental�48:�287�294.�[21]�Pretsch,�E.;�Bühlmann,�P.�Affolter,�C.;�Herrera,�A.;�Martínez,�R.�(2001).�Determinación�estructural�de�compuestos�orgánicos.�Springer,�Barcelona.��

3.�Results�and�discussion�

129�

[22]� Gumy,� D.;� Fernandez�Ibañez,� P.;� Malato,� S.;� Pulgarin,� C.;� Enea,� O.;� Kiwi,� J.� Supported� Fe/C� and�Fe/Nafion/C� catalysts� for� the� photo�Fenton� degradation� of� Orange� II� under� solar� irradiation,� Catal.�Today�101:�375�382.�[23]�Lei,�J.;�Liu,�C.;�Li,�F.;�Li,�X.;�Zhou,�S.;�Liu,�T.;�Gu,�M.;�Wu,�Q.�(2006).�Photodegradation�of�orange�I�in�the�heterogeneous�iron�oxide�oxalate�complex�system�under�UVA�irradiation,�J.�Hazrad.�Mat.�137:�1016�1024.�[24]�Franch,�M.I.;�Ayllón,�J.A.;�Peral,�J.;�Doménech,�X.�(2004).�Fe(III)�photocatalyzed�degradation�of�low�chain�carboxylic�acids�implications�of�the�iron�salt.�Applied�Catalysis�B:�Environmental�50:�89�99.�[25]� Hermosilla,� D;� Cortijo,� M.;� Huang,� C.P.� (2009).� The� role� of� iron� on� the� degradation� and�mineralization�of�organic�compounds�using�conventional�Fenton�and�photo�Fenton�processes.�Chemical�Engineering�Journal�155:�637�646.�[26]�Waychunas,�G.A.;�Rea,�B.A.;�Fuller,�C.C.;�Davis,�J.A.�(1993).�Surface�chemistry�of�ferrihydrite:�Part�1.�EXAFS� studies� of� the� geometry� of� coprecipitated� and� absorbed� absorbed� arsenate.� Geochim.�Cosmochim.�Acta�57:�2251�2269.�[27]�Fendorf,�S.;�Eick,�M.J.;�Grossl,�P.;�Sparks,�D.L.�(1997).�Arsenate�and�chromate�retention�mechanisms�on�goethite:�I.�Surface�structure.�Environ.�Sci.�Technol.�31:�315��320.�[28]�Fuller,�C.C.;�Davis,�J.A.;�Waychunas,�G.A.�(1993).�Surface�chemistry�of�ferrihydrite,�Part�2,�Kinetics�of�arsenate�adsorption�and�coprecipitation:�Geochim.�Cosmochim.�Acta�57:�2271�2282.�

130�

131�

4��

CONCLUSIONS�

132�

4.�Conclusions�

133�

Considering� the� objectives� of� this� thesis� and� after� the� studies� conducted,� the� results�

described� throughout� the� present� research� are� new� contributions� on� solving� real�

environmental� problems:� contaminated� soils� surrounding� mine� areas� and� industrial�

contaminated�waters.�The�following�are�the�most�important�conclusions�to�be�drawn�from�the�

results�obtained�in�the�different�works�contained�herein.��

� Field�Portable�X�Ray�Fluorescence�(FP�XRF)�spectrometry�has�been�an�effective�tool�to�

characterize�soil�samples�from�four�Moroccan�mine�sites.�In�this�sense,�the�application�

of�Geographic� Information�Systems�(GIS)� let� to�produce�maps�revealing� the�pollution�

trends� in� these� areas.� Likewise,� X�Ray� Absorption� Spectroscopy� (XAS)� has� been�

successfully� applied� to� determine� the�mercury� speciation� in� soil� samples� from� three�

main�European�mercury�mines.��

� The�pilot�plant� study� to� recover� zinc� from�a�mine� tailing�pond�has�been� carried�out,�

obtaining�proper�results�for�a�Zn�price�quotation�above�US$2/kg.�The�value�of�the�zinc�

product�covers�the�whole�treatment�while�an�environmental�problem�is�solved.��

� Various� Fe�loaded� materials� have� been� tested� as� Fenton� catalysts� and� arsenic�

sorbents.�The�results�obtained�when�used�as�Fenton�catalysts�were�comparable�to�the�

homogeneous� catalysis�while�avoiding� the� loss�of� the� catalyst� and� the�generation�of�

red�mud.� Its� application� as� arsenic� sorbents� achieved� high� rates� of� arsenic� sorption.�

The�application�of�XAS� techniques�applied� to� the�adsorption�of�arsenic�by�Fe�loaded�

materials�let�to�characterize�the�sorption�of�arsenic�on�these�Fe�loaded�materials.��

More� specific� conclusions� driven� from� the� obtained� results� for� each� of� the� studies�

performed�in�this�thesis�are�summarized�as:�

Heavy�Metal�Contamination�and�Mobility�at�the�Draa�Lasfar�mine�area:�

� Regarding�CER�values�calculated�using�the�FP�XRF�results,�arsenic,�copper,�lead�and�zinc�

can�be�distinguished�as�the�main�pollutants�of�the�mine�area�whilst�Ba,�Fe,�K,�Rb,�Sr,�Ti�

can�be�considered�lithogenic�components.��

� The�most�polluted�sites�are�found�beside�the�mine�site�towards�the�river�creek�whilst�

samples�closed�to�Koudiyat�hill�reported�values�similar�to�background��

� GIS�contour�maps�showed�a�similar�distribution�for�As�and�Cu,�as�well�as�for�Pb�and�Zn.�

The� most� contaminated� sites� were� at� the� vicinity� of� the� mine,� especially� at� the�

northwest�area,�probably�linked�to�weathering�effects�and�topography�of�the�area.��

4.�Conclusions��

134�

� The� leading� factor� regarding� mobility� of� the� samples� at� Draa� Lasfar� mine� area� is�

concentration�of�metals�and�organic�matter�(based�on�LOI�determinations).�However,�

given�the�low�metal�content�on�the�mobile�phase,�it�may�be�considered�lower�risk�than�

expected�when�taking�into�account�only�total�concentration�values.��

Characterization�of�Kettara,�Sidi�Bou�Othmane�and�Bir�Nehass�mines:�

� Likewise�Draaa�Lasfar,�As,�Cu,�Pb�and�Zn�are�the�main�pollutants�at�the�three�mine�sites�

regarding�its�CER�values.��

� The�level�of�contamination�of�each�mine�is�strongly�dependent�on�the�exploitation�time�

since�the�annual�extraction�for�each�mine�was�similar.��

� Samples�taken�at�residues�deposits�are�highly�polluted�compared�to�samples�taken�at�

the�mine�area.�These�results�were�corroborated�by�box�plot�representations�and�PCA.��

� The� samples�with� high� content�of� lead�and� zinc�present�high� concentration�of� these�

elements� in� the�mobile�phase,�so� it�can�be�concluded�that� the�high�concentration�of�

metals�exceed�the�capacity�of�the�soil�to�retain�them.�

XANES� speciation� of� mercury� in� three� mining� districts:� Almadén,� Asturias� (Spain),� Idria�

(Slovenia):�

� This�work� represents� the� first� inter�regional� study� of�mercury� speciation� of� the� two�

main� European� Hg�mining� districts� (Almaden� and� Idria),� and� a� polymetallic� district�

located�in�Asturias.�

� XANES� revealed� that� rather� insoluble� mercury� compounds� (cinnabar,� metacinnabar,�

schuetteite,� corderoite)� prevail� in� dumps� and� wastes� from� mines� and� metallurgical�

plants,�whereas�more�soluble�Hg�phases�(mainly�HgCl2�but�also�HgO�and�HgSO4)�were�

found�in�soils�and�sediments�from�all�target�areas.��

� It� can� be� established� from� the� results� from� the� three� districts,� that� the� presence� of�

mercury�chlorides�or�sulfates�can�be�related�to�mobile�mercury.��

� The� type� of�metallurgical� processing� arises� as� one� of� the�most� important� factors� in�

defining� mercury� mobility� as� less� efficient� roasting� treatment� (lower� roasting�

temperatures� and� poorer� recovering� rates)� increases� the� presence� of� metacinnabar�

and,�principally,�HgCl2�phases�responsible�for�the�mobility�of�mercury.��

4.�Conclusions�

135�

� Nonetheless,� the� absolute� ‘mobility’� remains� relatively� low� in�most� cases,� inherently�

suggesting� that� kinetic� effects� and� availability� of� the� soluble� phases� might� also� be�

considered�in�the�assessment�of�mercury�behavior.�

Extractant�and�solvent�selection�to�recover�zinc�from�a�mining�effluent:�from�laboratory�scale�to�

pilot�plant:�

� Opposed�to�DEHPA,�Cyanex�272�and�Ionquest�290�provide�high�zinc�selectivity�towards�

calcium�and�negligible�amounts�of�metals�are�found�in�the�organic�phase�avoiding�the�

regeneration�of�the�organic�phase�step.��

� Ionquest�290� is� considered� the�best�extractant�amongst� the� three� studied�due� to� its�

higher�selectivity�compared�to�DEHPA�and�a�higher�Zinc�recovery�(5–10%)�than�Cyanex�

272.��

� Amongst� the� studied� solvents� Ketrul� D80� and� Ketrul� D100,� the� latter� is� the�

recommended�due�to�its�lower�volatility�and�flammability.��

� The�pilot�plant�has�proven�the� feasibility�of� the�process�as� the�zinc� recovery� is�up�to�

95%�and�less�than�50�mg/L�are�left�in�the�raffinate.�The�stripping�is�efficient�and�only�a�

single�stage�at�O:A=20�is�required�to�achieve�a�transfer�of�40�g/L.��

Fe�loaded�materials�applied�as�Fenton�catalysts:�

� An� enhanced� methodology� using� mild� conditions� to� achieve� high� Fe�loadings� into�

zeolite�USY�is�presented�in�this�thesis.�In�this�sense,�1h�of�contact�with�zeolite�USY�and�

a� 0.05M�Fe(NO3)3� solution� at� room� temperature� reached�higher� loadings� than�when�

increased�contact�times�and�concentration�solutions�were�employed.�

� Other� low�cost�materials�with�exchange�properties� such�as�montmorillonite� clay�and�

natural� zeolite� clinoptilolite� have� also� been� loaded�with� the� enhanced�methodology�

applied�to�zeolite�USY�obtaining�also�high�Fe�loadings.��

� The�treatment�of�a�synthetic�dye�solution�(Acid�Red�14)�by�Fenton�reaction�using�the�

aforementioned�Fe�loaded�materials�achieved�total�decolorization.��

� Moreover,�using�such�Fe�loaded�materials�as�Fenton�catalyst,�the�removal�of�COD�from�

solutions� containing� acetic� acid� and�phenol� is� ca.� 30%�and�90%,� respectively,� results�

that�are�even�higher�than�those��obtained�when�using�homogeneous�catalysis.�

4.�Conclusions��

136�

� The� process� in� column� was� also� tested� for� the� clinoptilolite� grain� size� 0.2�2mm�

obtaining�degradation�kinetics�and�COD�removal�from�a�solution�containing�acetic�acid�

and�phenol�similar�to�the�homogeneous�catalysis��

� No�significant�amounts�of�Fe�are�stripped�from�the�materials�at�the�employed�reaction�

conditions.�

Fe�loaded�materials�applied�to�arsenic�removal:�

� Zeolite� USY,� zeolite� Y� and� Forager� sponge� loading� of� Fe� is� strongly� related� to� the�

surface� area,� specific� sites� and� functional� groups.� In� this� regard,� zeolite� Y� achieved�

greater�Fe�content�than�zeolite�USY�or�the�Forager�sponge.�

� A�relation�between�the�adsorbed�arsenic�and�the�content�of�iron�can�be�observed�for�

both� zeolites� (As:Fe=0.4).� Nevertheless,� Forager� sponge� As:Fe� ratio� is� higher�mainly�

due� to� the� presence� of� tertiary� amine� salt� groups� in� the� sponge� can� bind� further�

arsenic,�in�addition�to�the�arsenic�already�linked�to�Fe.��

� EXAFS�spectra� inferred�that� the�arsenate� is�complexed�with�Fe�as�a�bidentate�corner�

sharing�complex.��

With� this� thesis,� the� line�of�work�of�our� investigation�group�concerning�environmental�

problems�and�the�use�of�novel�techniques�to�its�characterization�is�broadened�by:�

� The�use�of�Geographic�Information�Systems�to�determine�spatial�variability�of�samples.�

� The�application�FP�XRF�and�EXAFS�techniques�to�the�characterization�of�soils�and�solid�

materials.�

� The�employment�of�Fe�loaded�materials�as�Fenton�catalysts.�

I��

HEAVY�METAL�CONTAMINATION�

AND�MOBILITY�AT�THE�MINE�AREA�OF�DRAA�LASFAR�(MOROCCO)

Marta�Avila,�Gustavo�Perez,�Mouhsine�Esshaimi,�Laila�Mandi,�Naaila�Ouazzani,�Jose�L.�Brianso�and�Manuel�Valiente.�

The�Open�Environmental�Pollution�&�Toxicology�Journal.�Accepted�Manuscript�

ACCEPTED MANUSCRIPT

Heavy Metal Contamination and Mobility at the Mine Area of Draa Sfar (Morocco) Marta Avilaa, GustavoPereza, Mouhsine Esshaimib, Laila Mandib,c, Naaila Ouazzanib, Jose L. Briansod

and Manuel Valiente*a

a Centre GTS. Chemistry Department, Universitat Autonoma de Barcelona, 08193 Spain. 5

b Laboratoire d'Hydrobiologie, Ecotoxicologie et Assainissement (LHEA), Faculté des Sciences Semlalia, Université Cadi Ayyad, Marrakech (Morocco)

c National Center for Studies and Research on Water and Energy, University Cadi Ayyad, BP511, 40 000 Marrakech (Morocco)

d Geology Department. Universitat Autonoma de Barcelona, 08193 Spain

*Corresponding author phone: +34-935812903; fax: +34-935811985; e-mail: [email protected]. 10

The present study represents a first insight into the Draa Sfar mine (Marrakech) to assess the possible diffusion of heavy metalsand to predict the risk of their mobility in the surroundings of the mine area. The edaphological parameters pH, electrical conductivity (EC), loss on ignition (LOI) and CaCO3 were measured according to standard methods, whilst heavy metals concentration was determined by Field Portable X-ray Fluorescence. Concentration enrichment ratios (CER) were calculated in 15

order to estimate the anthropogenic contribution of target pollutants determining As, Cu, Pb and Zn as the main pollutants, whereas Ba, Ca, Fe, K, Mn, Rb, Sr, Ti and Zr were considered lithogenic components. GIS contour maps of pollutants using CER data, showed the most polluted areas at the vicinity of the mine, especially at the northwest area, probably linked to weathering effects and topography of the area. Particle size studies established that As, Pb and Zn are part of the mineral orewhile Cu behaviour corresponded to an anthropogenic origin. Additionally, mobility assays employing single leaching tests 20

indicated a greater mobility of As and Zn rather than that of Pb and Cu due to their lower adsorption process in the soil, independently of their respective concentration.

Introduction The presence of heavy metals in soils originates 25

considerable impact on the environment causing damages to microflora, flora and fauna, and thus restricting soil use [1]. As a consequence of mining and mineral processing huge amounts of heavy metals are deposited in waste dumps and tailings requiring management and monitoring once the 30

activity has stopped [2]. In Marrakech region, mining activity represents a high area of activity thus constituting a great hazard due to the presence of high amounts of heavy metals related to functioning or abandoned mines. In this concern, few studies have been done in this area to determine the 35

heavy metal concentration around mine areas and their impact on surrounding soil and water resources [3]. In addition, no detailed investigation has been carried out in the region to assess the possible mobility of heavy metals in order to predict the toxicological risk in the surroundings of Draa Sfar 40

mine area. In the last years the systematic control of contaminated areas has become a key issue to define healthcare policies, cost effective environmental planning and risk assessment tools [4]. To this purpose the last decade Field Portable X-ray 45

Fluorescence (FP-XRF) equipments have been applied given their reliable and rapid heavy metal measurement which

allows to quickly delineate in situ metal contamination at a screening level [5, 6]. In addition, high volume of field test can be monitored to determine the spatial distribution and degree 50

of heterogeneity of heavy metals in an undisturbed position while off-site analytical costs are minimized without destruction of the samples [7, 8]. FP-XRF results can be applied together with Geographic Information Systems (GIS) to determine spatial variability in a mine area. Such tools let to 55

produce maps which are helpful in identifying the sources and spatial patterns of the pollutants [9, 10, 11]. Moreover, concentration enrichment ratios (CER), also called enrichment factors, have been used to obtain complementary reliable information on site risk assessment 60

[12, 13]. CER, was a concept developed in the early seventies to derive the origin of elements in the atmosphere, precipitation or seawater, and was progressively applied to other environmental materials, such as lake sediments or soils [14]. In many cases, it was used to determine the contribution 65

of anthropogenic emissions to trace element fluxes [15] (Table 1). Besides the concentration, toxicity and impact of heavy metals in soils and sediments is mostly determined by its mobility and availability [16]. The fate and transfer of these 70

metals is a complex process that depends on the soil mineralogy as well as to physicochemical transport processes. Over the last 30 years, sequential extraction schemes (SES)

ACCEPTED MANUSCRIPT

have been the main tools employed to evaluate the availability of the contaminants in soils, sediments and sludge [17, 18, 19, 20]. SES represent a chemical scheme which tries to mimic the various natural conditions under which soils may release metals into the water resources thereby providing 5

an indication of the potential bioavailability of those metals. On the other hand, leaching tests such as (NH4)2SO4 or HCl single non-selective extractions methods, can provide also a useful assessment for screening purposes to identify labile or mobile phases [21, 22]. The main advantages of these single 10

leaching tests against SES are mainly related to their cost efficiency, easy to use and a reduction on bias induced by sequential translation and accumulation of procedural errors. In this sense, the main aims of the present study focuses on (i) a geochemical characterization of the Draa Sfar mine area 15

in order to identify pollutants and lithogenic components present in the soils affected by the mining activity; (ii) the generation of distribution maps of pollutants at the mine area, (iii) the evaluation of particle size effects, such as intraparticle concentration affecting metals distribution and 20

(iv) the assessment of the pollutants mobility employing single leaching tests.

ExperimentalSite description

Draa Sfar mine is located a few hundred meters from the 25

Tensift River, close to a rural community of about 5790 ha of which 65% are occupied by farmland. The climate is Mediterranean, bordering arid and semi arid with an average annual precipitation of 231 mm (10 years). Temperatures are characterized by great daily and seasonal variation with an 30

average value of 11.5 C in January and 28.8 C in July. Draa Sfar mine, involves a deposit of pyrite mineral located 10 km west of Marrakech city (Fig. 1) can pose a risk for the environment due to discharge of tailings all around the mine area. Draa Sfar was discovered in 1953 although their 35

commercial exploitation did not begin until 1979. Mineral was processed by flotation after primary and secondary crushing and grinding producing 59516 tons of products in the first two years (1979-1980) [23]. Industrial activity stopped in March 1981, although activity restarted in 1999 40

due to its great resource of poly-metallic components (As, Cd, Cu, Fe, Pb, Zn).

Figure 1. Location of Draa Sfar mine.45

Table 1. Anthropogenic contribution at different CER values

CER Anthropogenic contribution <2 Minimal or nule 2-5 Moderate

5-20 Significant 20-40 Strong >40 Extreme

Sampling description

In order to assess the impact of the Draa Sfar mine residues on the surrounding environment, a total of 85 samples were collected in the vicinity of the mine covering 230 ha through 8 50

sampling lines oriented towards specific receptor media (Tensift river creek, Koudiyat Tazakouit hill, village, farms, etc.). Two samples were taken at the other side of Tensift river creek (samples 21 and 22) and 4 representative background samples (from 82 to 85) at 1 km from the mining site in order 55

to avoid mining contamination. Samples were taken every 50 meters from the upper 20 cm after removing the first layer of surface soil (2 cm) within an area of 100 cm2 per sample. Collected samples were air-dried at 30 C during 48 hours, sieved to remove large debris through 60

a 2 mm stainless steel sieve and stored in plastic bottles for their transportation to the laboratory.

Sample analysis and data treatment

The physical characterization consisted in the determination of the soil pH, the electrical conductivity (EC), the loss on 65

ignition (LOI) and the carbonate content of the samples according to standard methods [24]. The pH was measured in a soil suspension (2g/5 ml of distilled water stirred vigorously) after 2 h of deposition using a pH-meter (Model WTW Multiline P4 Universal pH-meter cabled Sen-Tix 92T pH 70

electrode, Germany). EC was determined in a soil saturated paste (1g/5 ml of distilled water) with a conductimeter (Model WTW Multiline P4 Universal Standard Conductivity Cell TetraCon® 325, Germany), once corrected to the working temperature (20 ºC). LOI was determined gravimetrically after 75

volatilization of organic matter on a furnace at 550°C during 4h. For the total carbonate content three replicates of each soil were stirred during 6 h in HCl 4 mol/L solution (1.0g of soil per 20 ml of HCl 4.0 mol/L solution) and, after filtering, calcium was measured by flame spectroscopy (Model 80

JENWAY-PFP7, UK). For the chemical characterization, an aliquot of each sample was encapsulated and covered with Mylar® film prior to their analysis with FP-XRF (Innov-X Systems, model Alpha-6500R, Woburn, MA, USA). A soil standard NIST 2710 and a SiO2 blank were measured for 85

corrections and three replicates were measured for each sample. The most contaminated samples were selected for the particle size effect and mobility assays studies. For the particle size effect studies, samples were milled and sieved below 100 μm for analysis of the fraction below 100 μm by FP-XRF. 90

Mobility assays were performed by applying a established methodology [25] consisting on sample extraction with HCl 0.5 M during 1h under magnetic stirring. After each extraction, the suspension was centrifuged and the supernatant was filtered using 0.22 μm filters (Millex GS, Millipore, Ireland). 95

ACCEPTED MANUSCRIPT

The extracts were analyzed by means of Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES) (ThermoElemental model Intrepid II XLS, Franklyn, MA, USA). In order to assess the impact of the Draa Sfar mine residues on the surrounding environment, a total of 85 5

samples were collected in the vicinity of the mine covering 230 ha through 8 sampling lines oriented towards specific receptor media (Tensift river creek, Koudiyat Tazakouit hill, village, farms, etc.). Two samples were taken at the other side of Tensift river creek (samples 21 and 22) and 4 10

representative background samples (from 82 to 85) at 1 km from the mining site in order to avoid mining contamination. CER indicators were calculated considering the concentration of a given element, namely Cn, in both target and background samples, normalized with respect to a 15

lithogenic conservative element such as Al, Zr or Ti, which is accurately determined in each sample. Rubio et al. [12] recommended the use of regional background values. While the geochemical background values are constant, the levels of contamination vary with time and places. Background values 20

are distinctly different among different soil types, especially with respect to Na, Mg, Al, K, Ca, Ba, Sc, Ti, Fe and Zr [13]. Zr was selected as lithogenic element due to homogeneity of Zr concentration in all samples and backgrounds.

25

30

The evaluation of the extent and distribution of contamination was carried out by using Geographic Information Systems (GIS) [26, 27, 28] which in addition, allowed the detection of the areas requiring monitoring or 35

even treatment. GIS maps for the distribution of target metals around the mine area were done by Miramon v6.4 - Complete Geographical Information System and Remote Sensing software [29]. Different interpolation methods can be used to determine spatial variability such as inverse distance 40

weighting (IDW), Kriging and spline functions [30]. While spline method involve a considerable interpolation error when there are large changes in the surface values within a short horizontal distance, kriging method may not be met in practice unless employing 100 samples in order to obtain a 45

reliable variogram that correctly describes spatial structure. In contrast, IDW interpolator assumes that each input point has a local influence that diminishes with distance [31], and no assumptions are required for the data, being this method the most suitable for our irregular sampling [32]. 50

Results and discussion Soil properties

pH, LOI and carbonate content [33] are geochemical soil characteristics able to provide sufficient information to understand the soils capacity to retain heavy metal pollutants. 55

(Numerical values on pH, EC, LOI and CaCO3 for each

sample can be found on Table S1 of Supplementary material). The results obtained for the soil pH measurements, depicted in GIS (Fig. 2) revealed that, in general, all sampled points presented a neutral to alkaline pH ranging from 7 to 9, similar 60

to background samples with the exception of a very acidic sample corresponding to sample D48 with a pH 3.47. pH variations seemed to be related to heterogeneous deposits of sulfidic residues in the surroundings of the mine which by oxidation and formation of sulfuric acid can cause a decrease 65

of the pH. EC showed more variability than the pH, with EC values ranging from 100 to 15.000 μS/cm (Fig. 3). In general, these results are correlated with previous studies carried on Morocco soils [34]. A decreasing salinity gradient was also observed 70

and the values obtained for the mine area samples are significantly higher than for the background samples which indicate high amounts of labile ions close to the mine area. A hot spot located at sample D31 with an EC of 14.160 μS/cm was observed mainly due to high amounts of metals present in 75

this area. Mine area and background samples LOI values (Fig. 4) have similar values ranging from 13 to 75 g/Kg, except some points where LOI could reach 76 g/kg due to some close localized agricultural activities. The observed carbonate content ranged from 10 to 210 mg.g-1 (Fig. 5) although the 80

majority of the samples present similar CaCO3 content to background samples. The highest values are observed for samples D26 and D71, located at 400 m of the mine. Together with basic pH values, the presence of carbonates in the soil lead to an increase in the retention of heavy metals, mainly as 85

carbonate salts as a consequence of precipitation, the principal retention mechanism of heavy metals [35].

Heavy metals content

From the obtained results employing FP-XRF and the corresponding CER values, elements can be classified into 90

pollutants (elements anthropogenically enhanced) or lithogenic elements (those with CER values similar to background samples). In this concern, most of the samples have CER values above 5 for As, Cu, Pb and Zn (Table S2), thus being considered the main pollutants of the mine area. 95

Arsenic distribution around the mine area, given in Fig. 6, showed two hot spots located just beside the mine area corresponding to samples D48 (3108 ppm, CER=280) and D31 (203 ppm, CER=19,4). Moving away from this area, samples showed lower As concentration with values similar to 100

background samples, except samples D45 (203 ppm, CER=15,9) and D46 (125 ppm, CER=9,2). Sample D48 depicts a very high arsenic concentration (more than 100 fold higher than background levels) indicating that remediation is mandatory for this specific area. An anomalous sampling point 105

is represented by sample D21 (72 ppm, CER=7,1), proceeding from the other side of the river, with arsenic concentration much higher than samples closer to the mine site. Thus, this area should be under monitoring since is in contact with the creek waters. 110

Regarding Cu CER distribution map along the mining area (Fig. 7) it can be stated that distribution of pollutants followed the similar trend as As although in a lesser degree of pollution.

���

���

BackgroundZrC

BackgroundnC

sampleZrC

samplenC

nCER

ACCEPTED MANUSCRIPT

Thus, the highest polluted samples are those located close to the mine, such as samples D20 (80 ppm, CER=2.6), D48 (144 ppm, CER=5.9) and D45 (173 ppm, CER=5.1). Again, sample D21 (51 ppm, CER=1.9) at the river basin, has a relatively high copper concentration despite being far from the mine 5

area. Also, the area close to Koudiyat Tazakouit hill, present copper concentration similar to background samples, thus indicates no anthropogenic contribution with copper. The lead distribution around the mine (Fig. 8) showed 4 hot spots located around samples D31 (180 ppm, CER=13.0), D45 (770 10

ppm, CER=45.9), D48 (2310 ppm, CER=130) and D58 (420

ppm, CER=30). Sample D21 (62 ppm, CER=4.6) should be also considered due to their high CER values and proximity to creek waters. CER distribution map for Zn (Fig. 9) followed the same trend as Pb with 4 hot spots located at samples D20 15

(630 ppm, CER=8.5), D45 (1110 ppm, CER=13.6), D48 (30 ppm, CER=10.8) and D58 (930 ppm, CER=10.8). Generally, GIS contour maps of CER for the pollutants showed the most contaminated at the vicinity of the mine, especially at the northwest area, probably linked to weathering 20

effects and topography of the area.

Figure 2. GIS contour map of pH at the mine area Figure 3. GIS contour map of the electrical conductivity at the mine area

Figure 4. GIS contour map of the loss on ignition (LOI) at the mine area Figure 5. GIS contour map of the CaCO3 content of the mine area

ACCEPTED MANUSCRIPT

Figure 6. GIS contour map of arsenic distribution around the mine area Figure 7. GIS contour map of copper distribution around the mine area

Figure 8. GIS contour map of lead distribution around the mine area Figure 9. GIS contour map of zinc distribution around the mine area

Other elements measured by FP-XRF presented values close to background samples and, accordingly. a mean CER value lower than 2, thus considering their origin as lithogenic. The values obtained for Ba, Fe, K, Rb, Sr, Ti and Zr are 5

shown on Table 2. As can be seen from the results on Table 2, mean values of mine area samples are similar to those of background samples and, in addition, mean CER values are between 0 and 2 (excepting sample 48 with high Fe content and sample 26 10

with high Sr content), thus indicating no anthropogenic enhancement of these elements in the soils analyzed. Finally, other elements were detected at extremely high concentration in some samples. High sulphur concentrations were found in samples D19 (18400 ppm), D31 (14500 ppm), 15

D33 (15500 ppm), D45 (36800 ppm), D48 (113700 ppm), D58 (5300 ppm), D59 (14800 ppm) and D70 (32400 ppm).

High arsenic concentrations are also found in some of these samples supporting the consideration of the arseno-pyrite nature of the mineral ores. Other elements such as Ag, Au, Bi, 20

Br, Cd, Co, Cr, Ni, P, Sb or Se were not detected due to the limits of detection of the FP-XRF. For the mobility and particle size effect studies, 7 samples were selected due to their high content on pollutants (samples D20, D31, D46, D48, D58 and D70) or for their spatial 25

significance (sample at the other side of the river creek, D21). Results obtained along with some drinking water quality standards, are depicted in Table 3. Results given in Table 3, indicate an increase on both As and Pb concentration when the samples are milled and sieved 30

below 100 μm, i.e., samples D20 (from 125 to 167 ppm), D31 (from 203 to 268 ppm), D46 (from 125 to 172 ppm), D48 (from 3108 to 3569 ppm) and D58 (from 113 to 149 ppm) had

ACCEPTED MANUSCRIPT

an enrichment on As and also for Pb samples D31 (from 180 to 313 ppm), D46 (from 375 to 477 ppm) and D48 (from 2309 to 2614 ppm) showed an enrichment when milled and sieved. On the rest of samples, slight differences were found. Thus, it can be stated that, in general, these elements are forming part 5

of the particle core which is in agreement of the arseno-pyrite nature of the mineral ore. The trend followed for Cu is a diminution of the concentration as the soil is being milled indicating that instead of forming part of the mineral, Cu is

adsorbed at the surface of the soil particles thus indicating and 10

anthropogenic origin. A few exceptions are found such as samples D31 (from 43 to 77ppm), D48 (from 144 to 167 ppm), D70 (from 33 to 50 ppm) which, in these cases, may constitute part of the arseno-pyrite mineral ore. For zinc, slightly higher concentrations are found when diminishing the particle size of 15

the soil that support its presence forming part of the mineral ores.

Table 2. Minimum, maximum and mean concentration and CER values for the lithogenic elements in mine area and background samples.

Mine area samples Background samples Min Max Mean Min Max Mean

Conc. 221,5 541 (D17) 389±70 404 530 449±60 Ba CER 0,4 1,4 0,9±0,3 0,9 1,2 1,0±0,1 Conc. 19,632 121,652 (D48) 32,721±10,000 30,769 38,106 35,555±3,000 Fe CER 0,4 4,8 0,9±0,5 0,9 1,1 1,0±0,1 Conc. 6,071 32,976 (D81) 23,327±5,000 26,770 30,263 38,244±1,500 K CER 0,3 1,4 0,8±0,3 0,9 1,1 1,0±0,1 Conc. 290 1,119 (D21) 607±150 631 737 699±50 Mn CER 0,4 2,0 0,9±0,3 0,9 1,0 1,0±0,1 Conc. 47 106 (D10) 76±13 73,4 86,4 81±6 Rb CER 0,5 1,7 1,0±0,3 0,9 1,1 1,0±0,1 Conc. 86 322 (D26) 144±40 119 140 131±9 Sr CER 0,5 3,1 1,1±0,5 0,9 1,0 1,0±0,1 Conc. 2,802 5,860 (D17) 4,460±700 4,431 5,741 5,229±600 Ti CER 0,4 1,3 0,9±0,2 0,9 1,1 1,0±0,1

Zr Conc. 112 335 (D67) 215±50 199 235 210±17 Conc is given in mg/kg. Max concentration is in parenthesis.

20

Regarding the results obtained for the mobility of selected samples, collected in Table 3, it can be stated that the content of arsenic, lead and zinc of some samples is higher than the quality standard regulations while Cu concentration on the mobile phase is bordering quality standards. Despite being the 25

most acidic sample (pH=3.5), with high EC (EC=4873 μS/cm) and low CaCO3 content (25,8 mg/g), the most polluted sample (D48) did not present high metal content on the mobile phase, thus indicating less danger than expected when taking into account only total concentration values. 30

According to literature [36] these conditions favor availability of cations, but sample D48 has also a high LOI value which benefits the adsorption of labile ions at the soil what explains the relatively low mobility of sample D48. The sample presenting most mobility of pollutants is sample D46, which 35

is a soil sample alkaline (pH=8.1), with an EC of 2,151 uS/cm, CaCO3 content of 58.4 mg/g and a LOI of 39.3 g/kg. In these conditions mobility is not favoured but the relatively low value of LOI regarding sample D48 enable the availability of cations from the mine ore to the mobile phase. 40

Therefore, it can be stated that the physico-chemical parameteres analysed does not correlate with the mobility results, thus, to assess the toxicological risk of the Draa Sfar mine area additional specific measurements are required.

45

Table 3. Results for < 2 mm and <100 μm particle size and fraction of

metal mobile.50

As Cu Pb Zn 100µm (mg/kg) 125 80 55 628 2mm (mg/kg) 167 72 66 713 D20

Mobility (mg/L) BDL BDL BDL BDL 100µm (mg/kg) 72 51 62 144 2mm (mg/kg) 67 48 61 150 D21

Mobility (mg/L) BDL BDL BDL BDL 100µm (mg/kg) 203 43 180 481 2mm (mg/kg) 268 77 313 734 D31

Mobility (mg/L) 49 2 6 18 100µm (mg/kg) 125 60 375 774 2mm (mg/kg) 172 59 477 933 D46

Mobility (mg/L) 54 1 17 23 100µm (mg/kg) 3,108 144 2,309 631 2mm (mg/kg) 3,569 167 2,614 704 D48

Mobility (mg/L) 5 1 BDL 4 100µm (mg/kg) 113 71 425 925 2mm (mg/kg) 149 77 537 1,087 D58

Mobility (mg/L) BDL BDL BDL BDL 100µm (mg/kg) 15 33 24 97 2mm (mg/kg) 15 50 20 91 D70

Mobility (mg/L) 29 BDL BDL BDL

55

ACCEPTED MANUSCRIPT

Conclusions Draa Sfar mine area has been characterized by determining various physico-chemical parameters of edaphological importance, including pH, electrical conductivity (CE), CaCO3 content and loss on ignition (LOI). Anthropogenic 5

pollution has been assessed by the use of CER. Thus, As, Cu, Pb and Zn can be distinguished as the main pollutants of the mine area. CER values obtained for Ba, Fe, K, Rb, Sr, Ti indicated its lithogenic characteristic. GIS contour maps of pollutants using CER data have been a valuable tool to 10

characterize pollutants distribution around the mine area and determine sources of contamination. GIS maps showed a similar distribution for As and Cu, as well as for Pb and Zn. The most contaminated sites were at the vicinity of the mine, especially at the northwest area, probably linked to 15

weathering effects and topography of the area. No contamination was found in and around Koudiyat Tazakouit hill. Concerning mobility studies, As, Pb and Zn concentration in some samples exceeded water quality standard regulations while Cu concentration on the mobile 20

phase is on the border. Nevertheless, the most polluted samples did not present high metal content on the mobile phase, thus indicating lower risk than expected when taking into account only total concentration values.

25

Acknowledgements The present work has been carried with support of the Spanish Ministry of Science and Innovation (Grant CTQ2009-07432), the Morocco-Spanish project N° A/011433/07 “Estudio de la movilidad de metales pesados en 30

suelos contaminados” and the pole of competences on Water and Environment (Morocco).

Notes†Electronic Supplementary Information (ESI) available.

References 35

[1] Adriano DC. Trace elements in the terrestrial environment, Springer Verlag, New York, U.S., 1986. [2] Osan J, Kurunczi S, Török S, Van Grieken R. X-Ray analysis of riverbank sediment of the Tisza (Hungary): identification of particles from a mine pollution event. 40

Spectrochim. Acta 2002; Part B 57: 413- 422. [3] El Adnani M, Rodriguez-Maroto JM, Sbai ML, Loukili IL,Nejmeddine A. Impact of polymetallic mine (Zn, Pb, Cu) residues on surface water, sediments and soils at the vicinity (Marrakech, Morocco). Environ. Technol. 2007; 28: 969-85. 45

[4].Ferguson C, Kasamas H. Risk assessment for contaminated sites in Europe. LQM Press University of Nottingham, Nottingham, U.K., 1999. [5] Kalnicky DJ, Singhvi R. Field portable XRF analysis of environmental samples. J. Hazard. Mater. 2001; 83: 93-122. 50

[6] Kilbride C, Poole J, Hutchings TR. A comparison of Cu, Pb, As, Cd, Zn, Fe, Ni and Mn determined by acid extraction/ICP–OES and ex situ field portable X-ray

fluorescence analyses. Environ. Pollut. 2006; 143: 16-20. [7] Harper M, Pacolay B, Hintz P, Andrew ME. A comparison 55

of portable XRF and ICP-OES analysis for lead on air filter samples from a lead ore concentrator mill and a lead-acid battery recycler. J. Environ. Monit. 2006; 8: 384 - 392. [8] Argyraki A, Ramsey MH, Potts PJ. Evaluation of Portable X-ray Fluorescence Instrumentation for in situ Measurements 60

of Lead on Contaminated Land. Analyst 1997; 122: 743 – 749. [9] Zhai L, Liao X, Chen T et al. Regional assessment of cadmium pollution in agricultural lands and the potential health risk related to intensive mining activities: A case study in Chenzhou City, China. J. Environ. Sci. 2008; 20: 696–703. 65

[10] Carr R, Zhang C, Moles N, Harder M. Identification and mapping of heavy metal pollution in soils of a sports ground in Galway City, Ireland, using a portable XRF analyser and GIS. Environ Geochem. Health. 2008; 30: 45-52. [11] Franco C, Soares A, Delgado J. Geostatistical modelling 70

of heavy metal contamination in the topsoil of Guadiamar river margins (S Spain) using a stochastic simulation technique. Geoderma 2006; 136: 852-864. [12] Sutherland RA, Tolosa CA. Multi-element analysis of road-deposited sediment in an urban drainage basin, Honolulu, 75

Hawaii. Environ Pollut. 2000; 110: 483-495. [13] Bhuiyana MAH, Parvezb L, Islamc MA, Dampared SB, Suzukia S. Heavy metal pollution of coal mine-affected agricultural soils in the northern part of Bangladesh. J. Hazard. Mater. 2010; 173: 384–392. 80

[14] Reimann C, Filzmoser P, Garrett RG. Background and threshold: critical comparison of methods of determination. Sci. Total Environ. 2005; 46: 1–16. [15] Bourennane H, Douay F, Sterckeman T et al.. Mapping of anthropogenic trace elements inputs in agricultural topsoil 85

from Northern France using enrichment factors. Geoderma 2010; 157: 165–174. [16] Kersten M, Förstner U. Trace element speciation: analytical methods and problems, ed G.E., Batley, CRC Press, Florida, 1995; 245-317. 90

[17] Tessier A, Campbell PGC, Bisson M. Sequential extraction procedures for the speciation of particulate trace metals. Anal. Chem. 1979; 51: 844–851. [18] Kennedy VH, Sanchez AL, Oughton DH, Rowland AP. Use of single and sequential chemical extractants to assess 95

radionuclide and heavy metal availability from soils for root uptake. Analyst 1997; 122: 89-100. [19] Rauret G, López-Sánchez J, Sahuquillo F, Muntau AH, Quevauviller P. Indicative values for extractable contents (mass fractions) of Cd, Cr, Cu, Ni, Pb and Zn in sediment 100

(CRM 601) following the modified BCR-sequential extraction (three-step) procedure – Addendum to EUR-report 17554 EN. Belgium: European Commission. BCR Information Reference Materials, EUR 19502 EN. 2000. [20] Sahuquillo A, Rigol A, Rauret G. Overview of the use of 105

leaching/extraction tests for risk assessment of trace metals in contaminated soils and sediments. Trends Anal. Chem. 2003; 22: 152-159. [21] Moreno-Jimenez E, Peñalosa JM, Manzano R, Carpena-Ruiz RO, Gamarra R, Esteban E. Heavy metals distribution in 110

soils surrounding an abandoned mine in NW Madrid (Spain)

ACCEPTED MANUSCRIPT

and their transference to wild flora. J. Hazard. Mater. 2009; 162: 854–859. [22] El Gharmali A, Rada A, El Adnani M, Tahlil N, El Meray M, Nejmeddine A. Impact of acid mine drainage on the quality of superficial Waters and sediments in the 5

Marrakesh region, Morocco. Environ. Technol. 2004; 25: 1431-1442. [23] Ministère de l’énergie et des mines. Délégation Régionale de Marrakech 1996. Monographie de la province de Marrakech (Géologie, Mines, Energie). 10

[www.minenv.gov.ma/onem/synt_monog_locales/mono_loc_marrakech.pdf]. [updated and cited 2010 November 1] [24] Aubert G. Méthodes d’analyses des sols. G.R.D.P., Marseille, 1978, 80-83,123-127. [25] Pérez G, López-Mesas M, Valiente M. Assessment of 15

heavy metal remobilization by fractionation: comparison of leaching tests applied to roadside sediments. Environ. Sci. Technol. 2008; 42: 2309–2315. [26] Zhai L, Liao X, Chen T et al. Regional assessment of cadmium pollution in agricultural lands and the potential 20

health risk related to intensive mining activities: A case study in Chenzhou City, China. J. Environ. Sci. 2008; 20: 696–703. [27] Carr R, Zhang C, Moles N, Harder M. Identification and mapping of heavy metal pollution in soils of a sports ground in Galway City, Ireland, using a portable XRF analyser and 25

GIS. Environ. Geochem. Health. 2008; 30: 45-52. [28] Franco C, Soares A, Delgado J. Geostatistical modelling of heavy metal contamination in the topsoil of Guadiamar river margins (S Spain) using a stochastic simulation

technique. Geoderma 2006; 136: 852-864. 30

[29] Pons X. MIRAMON website: [http://www.creaf.uab.es/miramon] (updated and cited 19 April 2010). [30] Shi W, Liu J, Du Z, Song Y, Chen C, Yue T. Surface modelling of soil pH. Geoderma 2009; 150: 113-119. 35

[31] Panagopoulos T, Jesus J, Antunes MDC, Beltrao J. Analysis of spatial interpolation for optimising management of a salinized field cultivated with lettuce. Eur. J. Agron 2006; 24: 1-10. [32] Jongman RHG, Ter Braak CJF, Van Tongeren OFR. Data 40

analysis in community and landscape ecology, Cambridge University Press, Cambridge, U.K, 1995. [33] de Matos AT, Fontes MPF, da Costa LM, Martínez MA. Mobility of heavy metals as related to soil chemical and mineralogical characteristics of Brazilian soils. Environ. 45

Pollut. 2001; 111: 429–435. [34] El Hachimi ML, El Founti L, Bouabdli A, Saidi N, Fekhoui M, Tasse N. Pb and As in mining alkaline waters: contamination, comportment and risks (the Zeida abandoned mine, Morocco). Rev. Sci. Eau 2007; 20: 1-13. 50

[35] Plassard F, Winiarski T, Petit-Ramel M. Retention and distribution of three heavy metals in a carbonated soil: comparison between batch and unsaturated column studies. J. Contam. Hydrol. 2000; 42: 99–111. [36] Chuan MC, Shu GY, Liu JC. Solubility of heavy metals in 55

a contaminated soils: effect of redox potential and pH. Water, Air Soil Pollut. 1996; 90: 543-556.

ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT

SUPPLEMENTARY MATERIAL

Table S1. pH, electrical conductivity (CE), loss on ignition (LOI) and CaCO3 content for Draa Sfar mine area samples.

Sample pH CE CaCO3 LOI Sample pH CE CaCO3 LOI

1 8.6 135 34.0 15.6 44 8.0 2565 118.6 64.4

2 9.6 172 26.1 12.7 45 7.5 8632 55.1 75.8

3 8.7 205 11.2 14.3 46 8.1 2151 58.4 39.3

4 8.7 294 21.2 24.4 47 8.2 2095 34.0 25.1

5 9.1 337 24.5 20.6 48 3.5 4873 25.8 56.0

6 8.5 1152 30.8 31.7 49 7.3 124 15.7 18.3

7 8.5 1708 28.5 27.9 50 8.0 102 7.9 19.1

8 9.0 747 38.2 32.1 51 8.3 107 25.8 17.4

9 8.0 4599 39.3 56.4 52 8.7 125 18.4 28.2

10 8.4 779 39.3 63.5 53 8.8 96 47.4 25.6

11 8.1 3452 27.6 26.2 54 8.6 136 48.3 21.2

12 8.6 536 31.5 22.4 55 8.8 132 79.8 23.0

13 8.3 489 25.3 17.8 56 8.7 203 36.8 24.3

14 8.4 2651 20.2 25.9 57 7.8 102 24.9 20.5

15 8.9 171 33.7 34.9 58 7.9 655 24.9 23.5

16 8.2 983 33.2 37.3 59 8.2 5275 55.2 32.7

17 7.8 2034 43.9 64.5 60 8.7 212 44.9 35.5

18 8.9 451 32.0 13.3 61 8.1 274 55.1 34.0

19 8.2 3376 30.8 23.3 62 8.2 240 52.8 48.3

20 8.3 852 15.8 24.5 63 7.9 456 37.1 61.8

21 7.9 1838 26.1 26.5 64 8.2 299 45.9 33.6

22 8.2 2210 26.2 17.9 65 8.4 261 43.8 27.2

23 8.2 1659 41.4 22.6 66 8.5 1663 20.7 25.6

24 8.4 922 23.7 14.6 67 8.5 106 20.0 22.3

25 8.1 1098 25.4 14.1 68 8.3 163 22.5 26.8

26 8.3 1797 209.9 54.4 69 7.9 2120 25.0 28.7

27 8.6 775 128.1 32.5 70 8.1 2570 93.3 29.5

28 8.5 576 79.1 36.8 71 8.3 1616 149.4 49.7

29 8.2 3152 21.9 21.7 72 8.5 692 98.9 40.7

30 8.0 5460 19.1 22.8 73 8.0 2699 62.9 44.9

31 7.6 14160 16.5 27.4 74 8.8 517 112.7 44.0

32 8.8 199 16.6 13.3 75 8.6 879 35.6 32.2

33 7.8 6940 29.7 52.6 76 8.1 2330 23.7 31.8

34 8.6 779 33.0 38.6 77 8.2 1298 36.8 27.9

35 8.4 375 29.7 40.1 78 8.1 1339 41.5 39.6

36 8.2 373 33.7 38.2 79 8.3 559 36.8 36.3

37 8.6 278 37.1 28.4 80 8.3 333 28.5 33.5

ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT

Sample pH CE CaCO3 LOI Sample pH CE CaCO3 LOI

38 8.5 345 24.4 24.4 81 8.3 425 39.1 29.4 39 8.5 415 57.3 44.7 82 8.7 113 21.3 23.5

40 8.5 552 35.6 42.8 83 8.1 805 26.1 29.1

41 9.0 214 72.9 39.5 84 8.6 145 24.9 39.9

42 8.6 371 55.1 46.7 85 8.6 151 27.3 30.1

43 8.5 493 61.7 48.5

ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT

Table S2. Concentration and CER values of pollutants As, Cu, Pb and Zn along with CER reference element Zr.

Sample As(mg/kg) CER Cu

(mg/kg) CER Pb(mg/kg) CER Zn

(mg/kg) CER Zr (mg/kg)

1 BDL - 16.7 0.3 13.2 0.5 39.7 0.3 302.6 2 BDL - 16.6 0.5 9.4 0.6 31.6 0.4 204.8 3 9.2 0.5 24.6 0.5 7.6 0.3 40.1 0.4 272.8 4 BDL - 37.8 1.1 9.6 0.6 44.2 0.6 199.5 5 13.2 0.7 31.9 0.6 11.2 0.4 46.2 0.4 308.0 6 12.4 0.7 26.3 0.6 10.5 0.4 48.8 0.4 285.8 7 BDL - BDL - 13.0 0.6 44.0 0.4 263.1 8 11.3 0.6 38.2 0.8 12.8 0.5 46.9 0.4 283.1 9 11.3 0.6 24.2 0.5 12.8 0.5 58.4 0.5 315.0

10 20.2 1.8 52.7 1.7 20.1 1.3 98.1 1.3 182.1 11 14.1 1.0 29.2 0.8 15.8 0.9 76.1 0.9 222.5 12 36.4 3.0 41.4 1.3 23.1 1.5 126.1 1.6 192.3 13 BDL - 24.1 0.5 8.4 0.4 36.0 0.3 277.5 14 BDL - 23.7 0.5 12.1 0.5 48.3 0.4 276.3 15 BDL - 33.9 0.7 18.4 0.8 58.3 0.5 291.1 16 15.1 1.1 30.1 0.8 15.5 0.8 70.2 0.8 229.1 17 18.1 1.6 50.3 1.6 17.6 1.2 96.8 1.3 185.5 18 13.8 1.4 28.9 1.1 12.1 0.9 65.9 1.0 157.3 19 BDL - 30.5 0.6 14.5 0.6 50.9 0.4 289.7 20 125.4 10.9 79.9 2.6 54.8 3.6 627.9 8.5 184.6 21 71.5 7.1 50.5 1.9 61.8 4.6 143.9 2.2 162.4 22 BDL - BDL - BDL - 29.7 0.7 111.5 23 14.5 0.9 21.8 0.5 9.0 0.4 47.7 0.5 254.8 24 9.6 0.8 17.9 0.6 12.0 0.8 39.2 0.5 191.4 25 12.0 1.2 17.7 0.7 11.1 0.9 43.7 0.7 158.3 26 19.4 1.9 23.5 0.9 7.4 0.6 45.6 0.7 163.6 27 19.2 1.6 28.9 0.9 14.9 1.0 100.0 1.3 188.2 28 22.5 2.1 34.1 1.2 16.4 1.2 80.2 1.2 169.6 29 12.4 1.2 23.1 0.9 12.4 0.9 48.6 0.8 162.0 30 12.7 1.3 22.5 0.8 10.5 0.8 51.4 0.8 161.0 31 203.3 19.4 43.1 1.5 179.7 13.0 480.7 7.2 167.8 32 14.7 0.9 22.1 0.5 6.7 0.3 49.4 0.5 274.4 33 14.4 1.3 46.2 1.6 18.0 1.3 76.1 1.1 172.8 34 18.1 1.5 25.9 0.8 12.4 0.8 72.7 1.0 190.0 35 16.8 1.6 34.1 1.2 16.6 1.2 79.3 1.2 168.6 36 14.2 1.3 39.4 1.3 14.2 1.0 74.8 1.0 179.5 37 15.2 1.4 30.3 1.0 11.6 0.8 72.8 1.0 175.3 38 16.5 1.6 18.8 0.7 13.3 1.0 78.5 1.2 162.7 39 25.2 2.7 42.8 1.7 17.6 1.4 102.9 1.7 151.4 40 16.4 1.3 25.6 0.8 16.8 1.0 85.4 1.1 199.3 41 12.6 1.1 40.6 1.3 15.2 1.0 77.6 1.0 190.2 42 16.0 1.4 38.6 1.2 15.6 1.0 78.8 1.0 187.9 43 18.1 1.4 35.7 1.0 17.8 1.0 90.9 1.1 209.9 44 19.5 1.6 29.9 0.9 23.4 1.5 91.2 1.2 196.1 45 203.3 15.9 172.5 5.1 773.5 45.9 1113.7 13.6 204.7 46 125.4 9.2 60.4 1.7 375.1 20.8 774.3 8.9 218.9 47 19.8 1.4 BDL - 23.3 1.3 81.8 0.9 225.0 48 3107.6 340.8 144.3 5.9 2309.5 191.8 631.2 10.8 146.3 49 12.7 1.0 30.8 0.9 23.7 1.4 71.5 0.8 212.4

ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT

Sample As(mg/kg) CER Cu

(mg/kg) CER Pb(mg/kg) CER Zn

(mg/kg) CER Zr (mg/kg)

50 13.2 0.8 24.7 0.6 27.3 1.3 89.0 0.9 253.6 51 BDL - 29.7 0.7 28.9 1.4 78.0 0.8 245.6 52 18.8 1.2 28.4 0.7 30.1 1.4 86.5 0.8 254.8 53 15.4 0.8 24.8 0.5 20.5 0.8 75.8 0.6 315.2 54 12.6 0.7 24.0 0.5 23.4 1.0 80.9 0.7 273.4 55 17.5 1.2 27.9 0.7 25.4 1.3 83.4 0.9 237.5 56 11.7 0.8 BDL - 19.8 1.1 80.0 0.9 226.2 57 9.2 0.6 28.2 0.7 19.4 0.9 60.1 0.6 251.6 58 113.5 8.5 70.9 2.0 424.6 24.1 925.2 10.8 214.4 59 14.3 1.1 32.0 0.9 15.5 0.9 68.2 0.8 209.0 60 22.7 2.2 39.2 1.4 12.6 0.9 91.2 1.4 168.1 61 16.7 1.4 29.4 0.9 17.7 1.1 80.3 1.0 196.2 62 14.7 1.3 25.5 0.8 20.3 1.3 88.0 1.2 184.3 63 12.1 1.0 37.9 1.1 18.8 1.1 101.0 1.2 203.6 64 BDL - 24.4 0.7 18.1 1.0 76.4 0.9 221.8 65 10.1 0.7 28.4 0.7 15.3 0.8 68.4 0.7 237.2 66 8.9 0.5 30.3 0.7 13.5 0.6 58.8 0.6 261.0 67 17.5 0.8 17.3 0.3 40.2 1.5 143.6 1.1 334.5 68 10.6 0.6 22.4 0.5 8.4 0.4 46.2 0.4 273.5 69 14.2 1.0 20.3 0.5 61.2 3.2 254.0 2.7 231.2 70 15.2 1.5 33.0 1.2 23.9 1.8 96.9 1.5 164.5 71 17.9 1.7 36.2 1.3 14.4 1.1 86.6 1.3 167.1 72 15.4 1.5 31.8 1.2 20.3 1.5 91.2 1.4 162.8 73 14.1 1.2 23.4 0.8 14.3 0.9 74.7 1.0 186.6 74 13.2 1.1 37.1 1.2 18.9 1.2 92.1 1.2 187.9 75 18.6 1.4 37.3 1.1 17.4 1.0 88.8 1.1 208.7 76 13.8 1.2 30.6 1.0 13.4 0.9 87.5 1.1 190.9 77 12.5 1.1 38.4 1.2 16.5 1.1 82.2 1.1 185.9 78 17.5 1.4 23.3 0.7 28.5 1.7 110.0 1.3 207.7 79 14.9 1.2 26.3 0.8 34.8 2.2 101.9 1.3 192.6 80 18.8 1.4 37.9 1.1 29.6 1.7 121.7 1.4 215.6 81 19.9 1.8 33.4 1.1 31.8 2.2 102.5 1.5 174.7 82 11.3 0.9 26.7 0.8 15.5 0.9 73.8 0.9 200.6 83 9.4 0.8 31.5 1.0 16.8 1.0 87.7 1.1 198.8 84 18.9 1.3 34.0 0.9 20.6 1.1 84.9 0.9 235.3 85 12.7 1.0 44.7 1.3 16.6 1.0 81.6 1.0 204.3

ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT

Table S3. Concentration and CER values of lithogenic components of the mine area soils.

Sample [Ba] CER [Ca] CER [Fe] CER [K] CER [Mn] CER [Rb] CER [Sr] CER [Ti] CER1 258 0.4 27289 0.7 20967 0.4 16935 0.4 366 0.4 60.4 0.5 101 0.5 3396 0.4 2 281 0.6 22533 0.9 21316 0.6 15710 0.6 331 0.5 54.3 0.7 111 0.8 3069 0.6 3 221 0.4 19581 0.6 21932 0.5 16496 0.4 359 0.4 58.6 0.6 102 0.6 3533 0.5 4 362 0.8 22154 0.9 26064 0.7 20143 0.7 589 0.9 64.3 0.8 114 0.9 3677 0.7 5 357 0.5 30911 0.8 27227 0.5 21029 0.5 473 0.4 66.5 0.6 114 0.6 4441 0.6 6 357 0.6 28711 0.8 28338 0.6 21479 0.6 509 0.5 70.2 0.6 121 0.7 4563 0.6 7 269 0.5 29235 0.9 24544 0.5 17130 0.5 417 0.5 61.6 0.6 119 0.7 4080 0.6 8 391 0.6 34868 1.0 28881 0.6 21204 0.5 531 0.5 68.7 0.6 140 0.8 4446 0.6 9 309 0.5 30801 0.8 30383 0.6 22375 0.5 536 0.5 73.6 0.6 135 0.7 4730 0.6

10 524 1.3 28820 1.3 45017 1.4 31405 1.3 953 1.5 106.4 1.5 143 1.2 5731 1.2 11 351 0.7 26997 1.0 31924 0.8 24433 0.8 681 0.9 78.6 0.9 142 1.0 5275 0.9 12 390 0.9 32758 1.4 30029 0.9 25208 1.0 552 0.8 76.3 1.0 157 1.3 4855 1.0 13 285 0.5 19025 0.5 22565 0.5 20010 0.5 379 0.4 66.7 0.6 86 0.5 3589 0.5 14 351 0.6 27108 0.8 28183 0.6 22776 0.6 523 0.6 69.0 0.6 119 0.7 4606 0.7 15 395 0.6 29798 0.8 33504 0.7 26127 0.7 622 0.6 82.2 0.7 127 0.7 4865 0.7 16 374 0.8 26120 0.9 34044 0.9 25434 0.8 646 0.8 87.0 1.0 128 0.9 4799 0.8 17 541 1.4 30486 1.3 42132 1.3 30390 1.2 873 1.4 100.9 1.4 137 1.2 5860 1.3 18 425 1.3 22717 1.1 30945 1.1 25338 1.2 560 1.0 71.0 1.2 126 1.3 5105 1.3 19 314 0.5 29023 0.8 27468 0.5 20836 0.5 477 0.5 67.1 0.6 122 0.7 4933 0.7 20 332 0.8 26896 1.2 31008 1.0 24356 1.0 816 1.3 82.4 1.2 163 1.4 4714 1.0 21 504 1.4 32042 1.6 48007 1.7 26530 1.2 1119 2.0 102.8 1.6 168 1.6 5277 1.3 22 268 1.1 23227 1.7 19632 1.0 18323 1.2 290 0.8 56.3 1.3 114 1.6 2802 1.0 23 277 0.5 26142 0.8 26535 0.6 18592 0.5 547 0.6 65.4 0.7 119 0.7 4482 0.7 24 268 0.6 24012 1.0 22180 0.7 17151 0.7 436 0.7 56.3 0.8 148 1.2 3251 0.7 25 282 0.8 26180 1.3 22429 0.8 17945 0.8 424 0.8 54.5 0.9 102 1.0 3351 0.8 26 254 0.7 152754 7.4 21726 0.8 6071 0.3 327 0.6 54.6 0.9 322 3.1 2913 0.7 27 398 1.0 80860 3.4 31861 1.0 20968 0.8 553 0.9 74.0 1.0 250 2.1 3787 0.8 28 489 1.3 57198 2.7 37649 1.3 28488 1.2 577 1.0 96.0 1.5 201 1.9 4614 1.1 29 260 0.7 26920 1.3 24892 0.9 19394 0.9 533 1.0 63.3 1.0 156 1.5 3712 0.9 30 358 1.0 24843 1.2 25017 0.9 18161 0.8 525 1.0 57.9 0.9 213 2.1 3382 0.8 31 400 1.1 31011 1.5 28026 1.0 15913 0.7 542 0.9 62.5 1.0 184 1.7 3211 0.8 32 447 0.8 26868 0.8 28776 0.6 21636 0.6 506 0.5 65.8 0.6 125 0.7 4979 0.7 33 471 1.3 35369 1.6 36671 1.2 27816 1.2 824 1.4 89.4 1.3 181 1.6 4770 1.1 34 417 1.0 35113 1.5 35246 1.1 26579 1.0 601 0.9 80.8 1.1 167 1.4 4543 0.9 35 444 1.2 30086 1.4 36326 1.2 28645 1.2 751 1.3 91.8 1.4 139 1.3 4839 1.1 36 478 1.2 31344 1.4 38300 1.2 29693 1.2 795 1.3 97.3 1.4 149 1.3 4924 1.1 37 331 0.9 38313 1.7 34242 1.1 26694 1.1 655 1.1 84.9 1.3 141 1.3 5110 1.2 38 419 1.2 31113 1.5 35236 1.2 30085 1.4 640 1.2 87.8 1.4 146 1.4 5001 1.2 39 460 1.4 47652 2.5 43267 1.6 29429 1.4 897 1.7 98.4 1.7 166 1.7 4876 1.3 40 464 1.1 28833 1.1 37955 1.1 27871 1.0 739 1.1 89.3 1.2 142 1.1 4950 1.0 41 403 1.0 50142 2.1 35597 1.1 24743 1.0 656 1.0 84.6 1.2 193 1.6 4939 1.0 42 391 1.0 45098 1.9 30991 0.9 23222 0.9 671 1.0 75.4 1.0 171 1.4 4097 0.9 43 483 1.1 43702 1.7 37225 1.0 28315 1.0 840 1.2 87.7 1.1 172 1.3 4557 0.9 44 501 1.2 71367 2.9 31805 0.9 19598 0.7 586 0.9 77.6 1.0 224 1.8 3458 0.7 45 370 0.8 45868 1.8 50814 1.4 16688 0.6 882 1.3 70.0 0.9 187 1.4 4497 0.9 46 456 1.0 41987 1.5 36602 1.0 22309 0.7 768 1.0 78.8 0.9 172 1.2 4588 0.8 47 303 0.6 30427 1.1 23885 0.6 17281 0.6 447 0.6 58.7 0.7 133 0.9 3688 0.7 48 336 1.1 24281 1.3 121652 4.8 BDL 564 1.1 47.0 0.8 132 1.4 3471 0.9 49 382 0.8 18375 0.7 30573 0.8 24754 0.9 488 0.7 67.8 0.8 99 0.7 4809 0.9

ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT

50 478 0.9 9961 0.3 36278 0.8 25687 0.7 549 0.6 79.5 0.8 93 0.6 4715 0.7 51 375 0.7 21304 0.7 30194 0.7 23103 0.7 524 0.6 67.4 0.7 103 0.7 4870 0.8 52 376 0.7 25823 0.8 30123 0.7 21614 0.6 503 0.6 67.5 0.7 108 0.7 5259 0.8 53 431 0.6 30841 0.8 29506 0.5 20669 0.5 502 0.5 67.2 0.6 108 0.5 4919 0.6 54 386 0.7 30322 0.9 29334 0.6 20068 0.5 470 0.5 66.4 0.6 117 0.7 5190 0.8 55 323 0.6 54544 1.8 27301 0.7 16463 0.5 465 0.6 65.0 0.7 138 0.9 5222 0.9 56 430 0.9 31668 1.1 29771 0.8 23137 0.7 495 0.6 70.9 0.8 108 0.8 4377 0.8 57 439 0.8 21703 0.7 30631 0.7 22118 0.6 493 0.6 70.1 0.7 92 0.6 4685 0.7 58 423 0.9 21702 0.8 40552 1.1 24804 0.8 658 0.9 74.9 0.9 103 0.8 4722 0.9 59 398 0.9 34736 1.3 31449 0.9 22958 0.8 692 1.0 80.7 1.0 229 1.7 4863 0.9 60 473 1.3 28184 1.3 38171 1.3 31414 1.4 660 1.2 96.5 1.5 154 1.4 4929 1.2 61 405 1.0 33259 1.3 34537 1.0 27695 1.0 651 1.0 82.8 1.1 148 1.2 5285 1.1 62 411 1.0 34232 1.5 33698 1.0 26177 1.0 753 1.2 80.8 1.1 148 1.3 4546 1.0 63 450 1.0 27892 1.1 35943 1.0 27599 1.0 702 1.0 85.4 1.1 142 1.1 4933 1.0 64 458 1.0 34855 1.2 30434 0.8 24809 0.8 547 0.7 72.2 0.8 139 1.0 4288 0.8 65 430 0.8 39220 1.3 27891 0.7 23153 0.7 525 0.6 71.6 0.8 147 1.0 3964 0.7 66 434 0.8 19555 0.6 28723 0.6 28953 0.8 562 0.6 76.3 0.8 113 0.7 3673 0.6 67 418 0.6 17613 0.4 29975 0.5 23603 0.5 496 0.4 70.4 0.5 101 0.5 4757 0.6 68 382 0.6 25783 0.7 28280 0.6 22584 0.6 497 0.5 66.7 0.6 120 0.7 4560 0.7 69 330 0.7 19160 0.7 24598 0.6 19955 0.6 439 0.6 66.5 0.7 107 0.7 3465 0.6 70 453 1.3 73693 3.6 34977 1.2 25748 1.1 839 1.5 98.5 1.6 179 1.7 3948 1.0 71 375 1.0 78690 3.7 32739 1.1 20540 0.9 747 1.3 80.9 1.3 190 1.8 3823 0.9 72 473 1.3 61498 3.0 38035 1.3 26637 1.2 689 1.2 92.0 1.5 179 1.7 4257 1.0 73 351 0.9 46338 2.0 28371 0.9 20922 0.8 611 1.0 70.4 1.0 156 1.3 3630 0.8 74 391 1.0 80724 3.4 32023 1.0 21234 0.8 802 1.3 78.2 1.1 171 1.4 4722 1.0 75 346 0.8 29240 1.1 30191 0.8 23764 0.8 615 0.9 77.8 1.0 138 1.0 4660 0.9 76 454 1.1 27167 1.1 33211 1.0 27808 1.1 675 1.0 80.9 1.1 133 1.1 5070 1.1 77 447 1.1 27857 1.2 37287 1.1 27383 1.1 781 1.2 88.7 1.2 143 1.2 5152 1.1 78 343 0.8 46893 1.8 31315 0.9 24632 0.9 628 0.9 74.3 0.9 158 1.2 4283 0.8 79 365 0.9 27641 1.1 32822 1.0 27028 1.0 664 1.0 83.1 1.1 142 1.2 5042 1.0 80 456 1.0 20194 0.7 36070 1.0 29217 1.0 708 1.0 90.3 1.1 128 0.9 4662 0.9 81 463 1.2 19429 0.9 37833 1.2 32976 1.4 688 1.2 92.1 1.4 121 1.1 5018 1.1 82 423 1.0 22577 0.9 30769 0.9 26770 1.0 631 0.9 73.4 0.9 119 0.9 4431 0.9 83 404 0.9 25740 1.0 35658 1.0 28047 1.0 697 1.0 78.8 1.0 130 1.0 5143 1.0 84 440 0.9 28436 1.0 37686 0.9 27896 0.9 730 0.9 86.4 1.0 140 0.9 5741 1.0 85 530 1.2 24581 1.0 38106 1.1 30263 1.1 737 1.1 84.9 1.1 134 1.0 5599 1.1

II��

XANES�SPECIATION�OF�MERCURY�IN�

THREE�MINING�DISTRICTS�–�ALMADEN,�ASTURIAS�(SPAIN),�IDRIA�

(SLOVENIA)

Jose�Maria�Esbri,�Anna�Bernaus,�Marta�Avila,�David�Kocman,�Eva�M.�Garcia�Noguero,�Beatriz�Guerrero,�Xavier�Gaona,�

Rodrigo�Alvarez,�Gustavo�Perez�Gonzalez,�Manuel�Valiente,�Pablo�Higueras,�Milena�Horvat�and�Jorge�Loredo.�

Journal�of�Synchrotron�Radiation��Journal�of�Synchrotron�Radiation.�(2010)�17,�2:�179�186.�

soil and geosciences

J. Synchrotron Rad. (2010). 17, 179–186 doi:10.1107/S0909049510001925 179

Journal of

SynchrotronRadiation

ISSN 0909-0495

Received 22 June 2009

Accepted 15 January 2010

# 2010 International Union of Crystallography

Printed in Singapore – all rights reserved

XANES speciation of mercury in threemining districts – Almaden, Asturias (Spain),Idria (Slovenia)

Jose Maria Esbrı,a* Anna Bernaus,b Marta Avila,b David Kocman,c

Eva M. Garcıa-Noguero,a Beatriz Guerrero,b Xavier Gaona,b Rodrigo Alvarez,d

Gustavo Perez-Gonzalez,b Manuel Valiente,b Pablo Higueras,a Milena Horvatc and

Jorge Loredod

aDepartamento de Ingenierıa Geologica y Minera, Escuela Universitaria Politecnica de Almaden,

Universidad de Castilla-La Mancha, 13400 Almaden (Ciudad Real), Spain, bGrup de Tecniques de

Separacio en Quımica (GTS), Departament de Quımica, Universitat Autonoma de Barcelona,

08193 Bellaterra (Barcelona), Spain, cDepartment of Environmental Sciences, Jozef Stefan Institute,

Ljubljana SI-1001, Slovenia, and dDepartamento de Explotacion y Prospeccion de Minas,

Universidad de Oviedo, Oviedo 33004, Spain. E-mail: [email protected]

The mobility, bioavailability and toxicity of mercury in the environment strongly

depend on the chemical species in which it is present in soil, sediments, water or

air. In mining districts, differences in mobility and bioavailability of mercury

mainly arise from the different type of mineralization and ore processing. In this

work, synchrotron-based X-ray absorption near-edge spectroscopy (XANES)

has been taken advantage of to study the speciation of mercury in geological

samples from three of the largest European mercury mining districts: Almaden

(Spain), Idria (Slovenia) and Asturias (Spain). XANES has been complemented

with a single extraction protocol for the determination of Hg mobility. Ore,

calcines, dump material, soil, sediment and suspended particles from the three

sites have been considered in the study. In the three sites, rather insoluble sulfide

compounds (cinnabar and metacinnabar) were found to predominate. Minor

amounts of more soluble mercury compounds (chlorides and sulfates) were also

identified in some samples. Single extraction procedures have put forward a

strong dependence of the mobility with the concentration of chlorides and

sulfates. Differences in efficiency of roasting furnaces from the three sites have

been found.

Keywords: mercury speciation; XANES; Almaden; Idria; Asturias; bioavailability.

1. Introduction

Assessing the distribution and mobilization of heavy metals in

the environment as a result of natural processes or anthro-

pogenic activities is of special relevance in mining districts.

Mercury (Hg) is one of the most toxic heavy metals, as some of

its compounds can be absorbed by living tissues in large doses

and these compounds or their derivatives can concentrate and

be stored over long periods of time. Through the food chain,

mercury can eventually affect human beings and cause chronic

or acute damage (Forstner, 1998). From a toxicological point

of view, the toxicity of heavy metals is primarily controlled

by the dose and the corresponding chemical speciation.

Accordingly, many recent studies have been devoted to assess

heavy metal speciation either through direct or indirect

approaches (Horvat, 2005). The most widely used methods are

based on sequential selective extractions (Bloom et al., 2003;

Kocman et al., 2004) and X-ray absorption spectroscopy

(XAS) techniques (Kim et al., 2000, 2003, 2004; Slowey et al.,

2005a,b; Bernaus et al., 2005a,b, 2006a,b). Alternative techni-

ques are based on Hg pyrolysis followed by AAS detection,

which allows the differentiation among cinnabar, metallic Hg

and matrix-bound Hg (Biester et al., 1999, 2000). XAS tech-

niques have been shown to provide reliable information on

the speciation of mercury without requiring sample pre-

treatment (Kim et al., 2004; Slowey et al., 2005a,b; Bernaus et

al., 2006a). The application of XAS to mercury speciation

provides results with good consistency in terms of Hg–S/Hg–

non-S and Hg–insoluble/Hg–soluble ratios according to wet-

chemistry data (Kim et al., 2003). On the other hand, one of

the main limitations of the XAS methods refers to their high

detection limits.

Among XAS techniques, both EXAFS (extended X-ray

absorption fine structure) and XANES (X-ray absorption

near-edge) spectroscopies have been previously used for the

speciation of mercury in different matrices, such as mine ores

and wastes (Kim et al., 2000, 2004), fish (Harris et al., 2003),

contaminated soils (Bernaus et al., 2006a) and hyacinths

(Riddle et al., 2002), and in studies of interactions between

mercury and soil minerals (Bernaus et al., 2005b). According

to data available in the literature (Webb, 2005), XANES is

particularly useful for analysis of geochemical and environ-

mental systems and has been preferred in this study. This is

in agreement with our previous experience and the known

XANES fingerprint differences among the Hg compounds

mainly expected in mining environments (Bernaus et al.,

2005a, 2006a,b).

In this framework, mobility studies represent a good

complement to purely speciation techniques, as they represent

a more empirical approach to the understanding of mercury

transfer among inorganic, organic and biological reservoirs.

In line with the publications by Brown and co-workers on

the characterization of mercury mines in north America (Kim

et al., 2000, 2004), this work aims at providing a further

understanding and a general perspective on the role of

mercury in three of the most important mercury mining

districts in Europe, namely Almaden and Asturias in Spain

and Idria in Slovenia.

2. Materials and methods

2.1. Study sites

Among the three mining districts selected in this study,

Almaden and Idria have been the largest world mercury

producers in historic times, both having a monometallic

character. On the other hand, Asturias has a more complex

mineralization, with high proportions of arsenic in its para-

genesis. It is important to highlight that Almaden is the largest

cinnabar (HgS) deposit in the world and it has been active

since the Roman times until the present days, having

accounted for about one third of the total Hg world produc-

tion (Hernandez et al., 1999; Saupe, 1990). Metallurgical

processing in the study area evolved from Bustamante

furnaces, with roasting temperatures over 873 K, to Pacific

furnaces in the last century, reaching temperatures of up to

1073 K.

From a mineralogical point of view, soils at Almaden area

are mainly represented by quartz and a diversity of clay-type

minerals such as chlorite, illite, kaolinite and pyrophyllite and

high contents of carbonates which correspond to a region with

shales and quartzites as main components of the stratigraphic

sequence (Garcıa Sansegundo et al., 1987, among others). The

high content of carbonates can be explained by the presence

of mafic magmatic rocks strongly affected by propilitic,

carbonate-rich alteration processes in the stratigraphic

sequence (Hall et al., 1997; Higueras et al., 2000).

Idria mining district is, like Almaden, a monometallic ore

deposit, with higher proportions of native mercury and hosted

in carbonate host rocks. The mineralization appears as two

main species: cinnabar and native mercury. Other minerals

appearing in the paragenesis are metacinnabar, pyrite,

marcasite, dolomite, calcite, kaolinite, epsomite and melan-

terite.

The mineralogical characterization of Idria samples reveals

carbonate bedrocks as main components of the stratigraphic

sequence, with the exception of the meadow soil from the

Pront Hill which was developed on carboniferous clastic rocks.

River bed and suspended sediments are composed of silica,

clay minerals, Fe and Al oxides, hydroxides and carbonates as

soil and geosciences

180 Jose Maria Esbrı et al. � XANES speciation of mercury J. Synchrotron Rad. (2010). 17, 179–186

Figure 1Sampling locations, mines and metallurgical sites of the three mercurymining districts, Almaden, Asturias and Idria. Abbreviations: ALM:Almadenejos decommissioned metallurgical plant; RD: Valdeazoguesriver downstream; El Entredicho: decommissioned open pit; AZG:Azogado stream; CH: dump of Almaden mine; HR: Huerta del Rey;SQ: San Quintın (real location: 50 km to the east of Almaden); TRR:El Terronal mine. (See Table 1 for more details.)

remnants of carbonate and clastic rock weathering products in

the Idrijca catchment (Kanduc et al., 2008).

Asturias district shows a more complex mineralogy, with

mercury present as cinnabar, but with variable metacinnabar

and metallic mercury proportions and with other metallic

minerals such as orpiment, realgar, melnikovite, chalcopyrite,

arsenopyrite, stibnite and galena (Loredo et al., 1999). This site

has an intense metallurgical activity with lower calcinations

temperatures in their rotary furnaces (over 853 K) than the

other mining districts (Luque & Gutierrez, 2006).

The total mercury concentration in soils and sediments of

these three mining districts is well documented (Berzas

Nevado et al., 2003; Higueras et al., 2003, 2006; Gray et al.,

2004; Horvat et al., 2002), although only a few studies dealt

with inorganic mercury speciation (Bernaus et al., 2005a,

2006a; Kocman et al., 2004; Biester et al., 1999, 2000).

2.2. Sample collection, storage and preparation

Samples from the main mines, metallurgical plants and

drainage network of the three districts were considered in this

study (Fig. 1). A list of samples, corresponding acronyms used

in the text and short descriptions is provided in Table 1.

The samples of soils, mine tailings, calcines and riparian

soils from Almaden were taken at a depth of 0–20 cm, stored

in polyethylene bags and sieved at the Almaden School of

Mines to below a grain size of 2 mm. Samples of suspended

particles were collected from the water column, sedimented in

laboratory and air-dried in a clean room. The rest of the

samples were air-dried to prevent mercury losses, homo-

genized and ground before analysis.

Soil samples from Idria were taken with a stainless steel

auger at a depth of 0–10 cm and stored in polyethylene

containers. Suspended river sediment was sampled during a

flood event of the Idrijca river by means of a net drift sampler

(Kocman, 2008). After removal of gravel, stones and plant

residues, river bed and suspended sediments were sieved and

separated in two grain-size fractions: <0.063 mm and 0.063–

2 mm. Before analyses, samples were dried at 303 K for three

days (to a constant weight) in the dark, then ground and

homogenized in an agate mortar and transferred into poly-

propylene containers.

The samples from Asturias area were collected in the La

Pena-El Terronal mine site, near the town of Mieres. The site

includes dumps, calcines, contaminated soils and a chimney

channel used to transport roasting smoke to the top of a

mount. Soils, riparian soils and mine tailings samples

(�1.5 kg) were collected at 10–30 cm depth, stored in poly-

ethylene bags, air-dried in a clean room and sieved in the

laboratory using a 0.1 mm sieve.

All solid samples from the three mining districts were

prepared for synchrotron analysis using an aliquot, mixed with

polyethylene (IR quality), homogenized with a vortex for

2 min and pressed to a pellet with 5 ton cm�2 of pressure.

2.3. Chemical characterization

Total mercury content of all solid samples was determined

by Zeeman atomic absorption spectrometry using high-

frequency modulation of light polarization (ZAAS-HFM)

with a Lumex RA-915+ analyzer (Sholupov &Ganeyev, 1995).

The detection limit of this technique for soils and sediments

soil and geosciences

J. Synchrotron Rad. (2010). 17, 179–186 Jose Maria Esbrı et al. � XANES speciation of mercury 181

Table 1Samples collected at the three mining districts.

Location ID Sampling area Material

Almaden siteAlmaden HR Huerta del Rey metallurgical precinct Soils from old metallurgical plant of the 17th century

CH Main dump of Almaden mine Dump material, sediments and riparian soilsAZG Azogado stream Riparian soils and stream sediments

Almadenejos ALM Decommissioned metallurgical plant Soils from the metallurgical precinctValdeazogues river RD Downstream of El Entredicho pit Suspended particlesSan Quintın SQ Decommissioned Pb–Zn–Ag mine Mine wastes and soils from an old flotation plant tested for

cinnabar treatment

Idria siteSoils S1–S3 Vicinity of the metallurgical plant Soils

S2 Pront Hill Meadow soilsS4 Idrijca merges with the river Baca Alluvial soil samples collected along the river Idrijca 40 km

downstream from the mineS5–S6 Alluvial plain confluence of Idrijca and Baca rivers Soils from a deep profile at depth 50 cm (S5) and 100 cm (S6)

Sediments RS Idrijca river, 35 km downstream from the mine beforeBaca river inflow

River bed sediments of a composite sample taken within adistance of 50 m with grain size <0.063 mm (RS1) and0.063–2 mm (RS2)

SS Idrijca river, 35 km downstream from the mine beforeBaca river inflow

Suspended river sediments of a composite sample takenwithin a distance of 50 m with grain size <0.063 mm (SS1)and 0.063–2 mm (SS2)

Asturias siteMine tailings TRRmn Mine and metallurgical plant Dumps in the vicinity of rotary furnacesCalcines TRRc Mine and metallurgical plant Calcination wasteSoil TRRs Metallurgical plant Soil from an abandoned chimney channelForest soils TRRfs El Terronal mine Forest soils from the mining area

samples is 0.5 mg Hg kg�1. For accuracy, certified reference

material (CRM-025) was analyzed simultaneously.

2.4. XANES measurements

XANES measurements were performed at the synchrotron

facility Hamburger Synchrotronstrahlungslabor (HASYLAB)

in Hamburg (Germany) at the bending-magnet beamline A1

(see further details by Bernaus et al., 2005b). All measure-

ments were carried out at room temperature. The beamline

set-up consisted of a Si(111) double-crystal monochromator,

three ionization chambers as transmission detectors and a

seven-pixel Ge fluorescence detector.

The photon absorption of mercury was recorded at its LIII

energy (12284 eV). Fluorescence detection mode was used

for the analysis of all samples, except for the reference

compounds whose spectra were recorded in transmission

mode. References for XANES fingerprint adjustments

included minerals and pure compounds: HgCl2, HgSO4, HgO,

CH3HgCl, Hg2Cl2 (calomel), HgSred (cinnabar), HgSblack(metacinnabar), Hg2NCl0.5(SO4)0.3(MoO4)0.1(CO3)0.1�H2O

(mosesite), Hg3S2Cl2 (corderoite), Hg3(SO4)O2 (schuetteite)

and Hg2ClO (terlinguaite). This selection was undertaken on

the basis of our prior knowledge of the geochemistry of the

different study areas (Horvat et al., 2002; Higueras et al., 2003,

2006; Gray et al., 2004; Kocman et al., 2004; Kanduc et al.,

2008), as well as the possible weathering and anthropogenic

processes taking place in each site.

XANES spectra were processed using SixPACK data

analysis software package (SIXPack, 2004; see also Catalano

et al., 2005; Slowey et al., 2005b; Arai et al., 2006). Spectra

processing included energy correction, signal normalization

and background correction. After data correction and

normalization, a principal component analysis (PCA) was

applied to the set of unknown spectra to determine the

number of principal components required to describe the

variation in the data. Then, the PCA results were used with a

target transformation, which projected the spectrum from a

reference compound onto the vector space defined by the

components. If the target vector lay within this component

space (above the 95% confidence level), then this reference

compound was selected to be present in the dataset. Finally, a

linear least-squares approach was used to determine the

fractional amount of each reference compound in the samples

(Malinowski, 1991; Ressler et al., 2000; Wasserman et al.,

1999). The quality of the target transform was given by the

reduced �2 value, which represents the goodness of the fit to

the spectra data, and is defined as

reduced �2 ¼ 1

N � P

XN

i¼1

�obsi � �fit

i

� �2; ð1Þ

where �iobs is the ordinate of the XANES spectrum measured

from the sample at the i th energy point, �ifit is the ordinate of

the fitted XANES spectrum, N is the number of data points in

the fitted XANES energy range (scaled by the wavenumber k)

and P is the number of fitted components.

A higher reduced-�2 denotes that the Hg compounds

compared possess a lower degree of similarity. This �2

represents the goodness of the model fit to the spectra

data using the linear combination procedure (Rehr et al.,

1992).

2.5. Mobility study (single extraction procedures)

Assays on the mobility of mercury were performed

according to the methodology reported by Perez et al. (2008).

Briefly, the methodology consisted of sample extraction with

0.5 MHCl for 1 h with magnetic stirring. The ratio solid :water

was 1 g:20 ml. After centrifugation at 3500 r.p.m. for 10 min,

the extracts were filtered and analyzed by ICP-OES (Ther-

moElemental ICP-OES, model Intrepid II XLS, Franklyn,

MA, USA).

3. Results and discussion

Total mercury content (Table 2) in the Almaden district shows

high Hg concentrations in soil samples from metallurgical

sites, which can be mainly attributed to the inefficient metal-

lurgical techniques used in the old plants of Almadenejos and

Huerta del Rey (Sumozas, 2005), with estimated roasting

temperatures below 873 K. High total mercury concentrations

have also been found in sediments and riparian soils from

Valdeazogues river, but especially from Azogado stream

(AZG) (2816 mg Hg g�1). The latter is in good agreement with

previous studies undertaken at the same sampling site (Gray et

al., 2004). Other heavy metals are in low concentrations except

in samples from the San Quintın area (SQ), where significantly

high amounts of Pb and Zn were also found (Table 2).

In Idria samples, analysis of total mercury content revealed

high concentrations in all samples (Table 2). Those samples

taken near the former smelting facilities were the most

polluted. This observation can be explained by the settling

down of Hg-enriched particles in the immediate vicinity of the

smokestack of the smelter. Moreover, the high total Hg

concentration observed in Idria sediments (RS) and in alluvial

soils (S4) 40 km downstream from the mine indicate that

sources of mercury such as mercury-bearing rocks, wastes

from combustion processes, as well as contaminated river-bed

sediments remain the major Hg input to the aquatic envir-

onment in the area even a decade after the end of mining

operations.

The total mercury content of soil and dump samples of

Asturias mine show the highest mercury content of the three

mines studied, with 27350 mg g�1 in dump samples (TRRmn-

116) and 18000 mg g�1 in soils from the chimney channel, with

high amounts of arsenic content (from 735 mg g�1 to

187218 mg g�1).

PCA was performed separately for each mining district

given the significant differences expected and considering the

number of sample XANES spectra (representative enough)

available in each case. As stated in x2.4, the original set of

reference compounds included 11 mercury phases (Fig. 2). In

Fig. 2, XANES spectra of samples collected in the three

soil and geosciences

182 Jose Maria Esbrı et al. � XANES speciation of mercury J. Synchrotron Rad. (2010). 17, 179–186

mining districts are also reported. As examples, Fig. 3 shows

the fitted spectra for selected samples from each of the three

sites (more data are reported in Table 3).

For the Almaden district, the PCA results indicate that five

components [cinnabar (Cb), metacinnabar (Mc), HgCl2,

Hg2Cl2 and schuetteite (Sc)] can be used to reconstruct each of

the experimental spectra (depending on the sample) above the

95% confidence level. Mercury sulfides are the most common

species found in almost all samples (Table 3), especially in

those collected in abandoned metallurgical plants like Alma-

denejos area and Huerta del Rey (Almaden area). Non-sulfide

phases like schuetteite [Hg3(SO4)O2], calomel (Hg2Cl2) and

mercuric chloride (HgCl2) are present in different ratios in soil

and sediment samples.

XANES analyses in the samples from San Quintın area (see

Table 3) have shown the absence of metacinnabar but high

amounts of cinnabar (47–59%) and minor amounts of rela-

tively more soluble species like calomel (24–33%) and

schuetteite (17–21%) which can be attributed to weathering

processes. The absence of metacinnabar, a metastable poly-

morph of cinnabar which occurs during the roasting process of

mercury ores in the presence of impurities (Dickson & Tunell,

1959), is due to the historical use of the site, as only flotation

tests were performed and no furnaces were used there.

On the other hand, metacinnabar has been identified in soil

samples from Almadenejos (ALM) (31–39%) and Huerta del

Rey (HR) (�23%), locations with known historic metallur-

gical activity.

Other non-sulfide phases like mercurous chloride (24–43%)

have also been identified at San Quintın and Huerta del Rey,

and can be attributed to the process of soil formation. High

soil and geosciences

J. Synchrotron Rad. (2010). 17, 179–186 Jose Maria Esbrı et al. � XANES speciation of mercury 183

Table 2Average heavy metals content in samples from the three mining districts(in mg g�1).

Mercury was analyzed by ZAAS-HFM and As, Zn and Pb by XRF. BDL: databelow detection limits. � = grain size.

Sample Material Hg As Pb Zn

AlmadenCH-127 Dump 989 BDL BDL 112HR-108 Soil 976 BDL 214 96HR-109 Soil 404 BDL 111 104HR-110 Soil 200 BDL 130 185RD-124 Suspended

particles105 BDL BDL BDL

CH-125 Sediment 1800 BDL BDL 112AZG-105 Riparian soils 2816 23 139 233CH-128 Riparian soils 450 BDL 102 185ALM-101 Soil 2720 BDL 74 153ALM-102 Soil 2629 BDL 102 193CH-126 Soil 2230 BDL BDL 365SQ-111 Dump 902 BDL 15837 6877SQ-112 Dump 1730 BDL 2154 1221SQ-113 Soil 1935 BDL 19049 7134SQ-114 Soil 390

AsturiasTRRmn-115 Dump 1470 39338 BDL BDLTRRmn-116 Dump 27350 117553 BDL BDLTRRs-118 Chimney soil 3280 735 BDL BDLTRRs-121 Chimney soil 18000 12133 BDL BDLTRRmn-122 Dump 5785 42300 BDL BDLTRRfs-3 Soil 1570 16826 107 173TRRfs-4 Soil 1080 1120 53 137TRRc-5 Calcined 34 187218 BDL BDLTRRc-55 Calcined 54 25876 BDL BDL

IdriaS-1 Soil 333 21 BDL 112S-2 Meadow Soil 47 26 BDL 102S-4 Alluvial soil 76 BDL BDL 64S-5 Soil

(50 cm depth)175 BDL 47 145

S-6 Soil(100 cm depth)

144 BDL 73 496

RS-1 Sediment� < 63 mm

6540 BDL 302 270

RS-2 Sediment� < 2 mm

1920 BDL 14 BDL

SS-1 Suspendedparticles� < 63 mm

96 BDL BDL 449

SS-2 Suspendedparticles� <2 mm

11 BDL BDL 24

S-3 Soil 95 27 46 130Hg ore Ore

Figure 2XANES spectra of selected Hg pure compounds and samples fromAlmaden, Idria and Asturias mining districts (all spectra are deliberatelydisplaced to show differences). Each spectrum corresponds to the meanvalue of five replicates.

amounts of schuetteite have been identified in ore stockpile in

San Quintın and Almadenejos area. This is a mineral phase

typically linked to the presence of Hg0 that appears in the

sunlight-exposed side of the rock surface, and it is frequently

found near old furnaces or ore dumps (Higueras et al., 2003).

High proportions of relatively more soluble phases have

been identified in soil and sediment samples from Valdeazo-

gues River (100%) and Azogado stream (100%). These phases

[Hg2Cl2, HgCl2 and Hg3(SO4)O2] have been considered a

result of the weathering processes taking place within the

drainage network of the mining district. The mobility of

mercury in this district is clearly linked with metallurgical

activity and formation of secondary chloride phases. The

highest mobility was found in soil samples from an old

metallurgical precinct (ALM) (10.8–21.3 mg L�1; see Table 3).

At the Idria mining district the PCA analysis reveals the

presence of five components (Cb, Mc, Sc, HgO, HgSO4). In

this district, cinnabar is the most common Hg form in soil,

sediments and suspended particles, while the presence of

metacinnabar is found in a soil sample (S-4), and sulfates in

soils and sediments (S, RS, SS). The lack of metacinnabar in

most of these samples is due to the re-use of calcines and

metallurgical wastes in the refilling of mine galleries with

minor dispersion of this material throughout the surrounding

environment. High proportions of sulfates were found in soil

samples (S), but the mobility of mercury in this district was

clearly reduced, mainly by the major proportions of cinnabar

in soils, sediments and suspended particles. This low mobility

of mercury (0.2–0.3 mg L�1, see Table 3) is in accordance with

Kocman (2008), describing low water-soluble mercury species

in sediments and suspended particles.

In Asturias mining district, the PCA analysis needs six

components to reconstruct samples spectra [Cb, Mc, corder-

oite (Co), HgCl2, HgO, HgSO4]. All samples from the

decommissioned mine and metallurgical facility show high

mercury contents in soils (TRRfs), dump materials (TRRmn)

soil and geosciences

184 Jose Maria Esbrı et al. � XANES speciation of mercury J. Synchrotron Rad. (2010). 17, 179–186

Table 3Main mercury species (in %) and mobile mercury (in mg L�1 and %).

Abbreviations: Cb, cinnabar; Mc, metacinnabar; Sc, schuetteite; Co, corderoite; BDL, below detection limits.

Sample Cb Mc Sc Co HgO HgSO4 Hg2Cl2 HgCl2 Reduced �2 Mobility† (mg L�1) (%)

AlmadenCH-127 62 0 0 0 0 0 38 0 0.0004 1.4 (3.2)HR-108 37 23 0 0 0 0 40 0 0.0006 0.6 (1.2)HR-109 33 24 0 0 0 0 43 0 0.0007 0.2 (1)HR-110 41 22 0 0 0 0 37 0 0.0006 BDLRD-124 0 0 94 0 0 0 0 6 0.0006 BDLCH-125 7 0 83 0 0 0 0 10 0.0004 BDLAZG-105 0 0 80 0 0 0 20 0 0.0003 BDLCH-128 24 22 0 0 0 0 35 19 0.0004 BDLALM-101 38 39 23 0 0 0 0 0 0.0003 10.8 (7.9)ALM-102 39 31 0 0 0 0 30 0 0.0007 21.3 (16.2)CH-126 33 32 35 0 0 0 0 0 0.0003 BDLSQ-111 54 0 17 0 0 0 29 0 0.0002 0.6 (1.3)SQ-112 51 0 21 0 0 0 28 0 0.0002 3.7 (4.3)SQ-113 59 0 17 0 0 0 24 0 0.0002 BDLSQ-114 47 0 20 0 0 0 33 0 0.0003 BDL

AsturiasTRRmn-115 29 24 0 0 0 0 0 47 0.001 0.4 (0.5)TRRmn-116 28 22 0 0 0 0 0 50 0.0009 73.3 (5.4)TRRs-118 28 22 0 0 0 0 0 50 0.0008 20.1 (12.3)TRRs-121 29 22 0 0 0 0 0 49 0.0007 56.5 (6.3)TRRmn-122 30 24 0 0 0 0 0 46 0.0007 43.6 (15.1)TRRfs-3 44 28 0 0 10 18 0 0 0.003 0.7 (0.9)TRRfs-4 50 36 0 14 0 0 0 0 0.003 0.1 (0.2)TRRc-5 52 30 0 18 0 0 0 0 0.008 BDLTRRc-55 57 43 0 0 0 0 0 0 0.007 BDL

IdriaS-1 44 0 32 0 0 24 0 0 0.006 BDLS-2 55 0 0 0 0 45 0 0 0.002 0.2 (8.5)S-4 85 15 0 0 0 0 0 0 0.004 BDLS-5 90 0 0 0 10 0 0 0 0.004 BDLS-6 58 0 0 0 0 42 0 0 0.005 BDLRS-1 57 0 0 0 0 43 0 0 0.002 BDLRS-2 100 0 0 0 0 0 0 0 0.003 BDLSS-1 90 0 0 0 0 10 0 0 0.004 BDLSS-2 55 0 0 0 0 45 0 0 0.009 BDLS-3 66 0 26 0 8 0 0 0 0.007 0.3 (6.3)Hg ore 100 0 0 0 0 0 0 0

† Determined according to the method of Perez et al. (2008).

and chimney soils (TRRs) (Table 2), and a predominance of

sulfides species (50–100%) with significant presence of meta-

cinnabar in all samples (Table 3). Ratios between cinnabar and

metacinnabar in these samples are lower than in Almaden

area, where metallurgical activity was not the predominant

activity. In this mining site, metallurgy was less efficient than in

Idria and Almaden area, with lower roasting temperature and

poorest recovery rates. The contents of other mercury species

such as chlorides are significant, with high amounts on soils

samples from the facility and the chimney exhausting roasting

smokes. The mobility of mercury in this district is higher than

in Almaden. In qualitative terms, the percentage of mobile

mercury agrees well with the presence of HgCl2 except for

TRRmn-115. In general, it is important to point out that it is

likely that the methodology applied to assess Hg mobility only

extracts a fraction of the HgCl2 present, thus underestimating

Hg mobility.

If we consider the three districts, the main processes

affecting mercury speciation are ore composition, mining

history and roasting process. The type of metallurgical

processing arises as one of the most important factors in

defining mercury availability: mercury mobility is higher in

Asturias district owing to the inefficient roasting treatment

used (lower roasting temperatures and poorer recovering

rates); the mobility is significantly lower in the Almaden

district, with better furnaces (only in the last century) and

despite the complex and lengthy history of mining and

metallurgical activity. On the other hand, the even lower

mobility values found in Idria district are related to its efficient

metallurgical process (similar to Almaden area), together with

the appropriate management of calcines used for refilling old

galleries and the shorter mining history of the district.

4. Conclusions

This work represents the first inter-regional study of mercury

speciation of the two main European Hg-mining districts

(Almaden and Idria), and a polymetallic district located in

Asturias.

XANES has provided key information on the inorganic

mercury speciation of ores, calcines, dump material, soils,

sediments and suspended particles samples. Rather insoluble

mercury compounds (cinnabar, metacinnabar, schuetteite,

corderoite) have been shown to prevail in dumps and wastes

from mines and metallurgical plants, whereas more soluble Hg

phases (mainly HgCl2 but also HgO and HgSO4) were found

in soils and sediments from all target areas. A qualitative

relationship between mobile mercury and the presence of

mercury chlorides or sulfates compounds has been established

for samples from the three districts. Nonetheless, the absolute

‘mobility’ remains relatively low in most cases, inherently

suggesting that kinetic effects and availability of the soluble

phases might also be considered in the assessment of mercury

behaviour.

Synchrotron experiments at HASYLAB were financially

supported by the European Community, Research Infra-

structure Action under the FP6 ‘Structuring the European

Research Area’ Programme (through the Integrated Infra-

structure Initiative ‘Integrating Activity on Synchrotron and

Free Electron Laser Science’). Financial contribution from the

projects PPQ2003-01902, CTQ2005-09430-C05 and CTM2006-

13091-C02-02/TECNO funded by the Spanish Ministry of

Science and Innovation is also acknowledged.

References

Arai, J., Lanzirotti, A., Sutton, S. R., Newville, M., Dyer, J. & Sparks,D. L. (2006). Environ. Sci. Technol. 40, 673–679.

Bernaus, A., Gaona, X., Esbrı, J. M., Higueras, P., Falkenberg, G. &Valiente, M. (2006a). Environ. Sci. Technol. 40, 4090–4095.

Bernaus, A., Gaona, X., Ivask, A., Kahru, A. & Valiente, M. (2005b).Anal. Bioanal. Chem. 382, 1541–1548.

Bernaus, A., Gaona, X. & Valiente, M. (2005a). J. Environ. Monit. 7,771–777.

Bernaus, A., Gaona, X., van Ree, D. & Valiente, M. (2006b). Anal.Chim. Acta, 565, 73–80.

Berzas Nevado, J. J., Garcıa Bermejo, L. F. & Rodrıguez Martın-Doimeadios, R. C. (2003). Environ. Pollut. 122, 261–271.

Biester, H., Gosar, M. & Covelli, S. (2000). Environ. Sci. Technol. 34,3330–3336.

Biester, H., Gosar, M. & Muller, G. (1999). J. Geochem. Explor. 65,195–204.

Bloom, N. S., Preus, E., Katon, J. & Hiltner, M. (2003). Anal. Chim.Acta, 479, 233–248.

Catalano, J. G., Trainor, T. P., Eng, P. J., Waychunas, G. A. & Brown,G. E. (2005). Geochim. Cosmochim. Acta, 69, 3555–3572.

Dickson, F. W. & Tunell, G. (1959). Am. Mineral. 44, 471–487.Forstner, U. (1998). Integrated Pollution Control, edited by A.Weissbach and H. Boeddicker, pp. 81–130. Berlin: Springer-Verlag.

Garcıa-Sansegundo, J., Lorenzo, S. & Ortega, E. (1987). Servicio dePublicaciones, Ministerio de Industria y Energıa, Madrid, p. 60.

Gray, J. E., Hines, M. E., Higueras, P., Adatto, I. & Lasorsa, B. K.(2004). Environ. Sci. Technol. 38, 4285–4292.

Hall, C. M., Higueras, P. L., Kesler, S. E., Lunar, R., Dong, H. &Halliday, A. N. (1997). Earth Planet. Sci. Lett. 148, 287–298.

soil and geosciences

J. Synchrotron Rad. (2010). 17, 179–186 Jose Maria Esbrı et al. � XANES speciation of mercury 185

Figure 3XANES spectra of selected samples from the three mining districts withreconstructed spectra shown as dashed lines. (See Table 3 for moredetails.)

Harris, H. H., Pickering, I. J. & George, G. N. (2003). Science, 301,1203.

Hernandez, A., Jebrak, M., Higueras, P., Oyarzun, R., Morata, D. &Munha, J. (1999). Miner. Deposita, 34, 539–548.

Higueras, P., Oyarzun, R., Biester, H., Lillo, J. & Lorenzo, S. (2003). J.Geochem. Explor. 80, 95–104.

Higueras, P., Oyarzun, R., Lillo, J., Sanchez-Hernandez, J. C., Molina,J. A., Esbrı, J. M. & Lorenzo, S. (2006). Sci. Tot. Environ. 356, 112–124.

Higueras, P., Oyarzun, R., Munha, J. & Morata, D. (2000). Rev. Soc.Geol. Esp. 13, 10–119.

Horvat, M. (2005). Dynamics of Mercury Pollution on Regional andGlobal Scales: Atmospheric Processes and Human ExposuresAround the World, edited by V. Pirrone, N. Mahaffey and R.Kathryn, pp. 154–190. New York: Springer.

Horvat, M., Jereb, V., Fajon, V., Logar, M., Kotnik, J., Faganeli, J.,Hines, M. & Bonzongo, J. C. (2002). Geochem. Explor. Environ.Anal. 2, 287–296.

Kanduc, T., Kocman, D. & Ogrinc, N. (2008). Aquat. Geochem. 14,239–262.

Kim, C. S., Bloom, N. S., Rytuba, J. J. & Brown, G. E. (2003). Environ.Sci. Technol. 37, 5102–5108.

Kim, C. S., Brown Jr, G. E. & Rytuba, J. J. (2000). Sci. Tot. Environ.261, 157–168.

Kim, C. S., Rytuba, J. J. & Brown Jr, G. E. (2004). Appl. Geochem. 19,379–393.

Kocman, D. (2008). PhD thesis, Vol XIII, pp. 152, Jozef StefanInternational Postgraduate School, Slovenia.

Kocman, D., Horvat, M. & Kotnik, J. (2004). J. Environ. Monitor. 6,696–703.

Loredo, J., Ordonez, A., Gallego, J. R., Baldo, C. & Garcıa-Iglesias, J.(1999). J. Geochem. Explor. 67, 377–390.

Luque, C. & Gutierrez, M. (2006). Editors. La Minerıa del Mercurioen Asturias: Rasgos Historicos, 1st ed.

Malinowski, E. R. (1991). Factor Analysis in Chemistry, 2nd ed. NewYork: Wiley.

Perez, G., Lopez-Mesas, M. & Valiente, M. (2008). Environ. Sci.Technol. 42, 2309–2315.

Rehr, J. J., Albers, R. C. & Zabinsky, S. I. (1992). Phys. Rev. Lett. 69,3397–3400.

Ressler, T., Wong, J., Roos, J. & Smith, I. L. (2000). Environ. Sci.Technol. 34, 950–958.

Riddle, S. G., Tran, H. H., Dewitt, J. G. & Andrews, J. C. (2002).Environ. Sci. Technol. 36, 1965–1970.

Saupe, F. (1990). Econ. Geol. 85, 482–510.Sholupov, S. E. & Ganeyev, A. A. (1995). Spectrochim. Acta B, 50,1227–1238.

SIXPack (2004). SIXPack – Sam’s Interface for XAS AnalysisPackage, powered by IFEFFIT 1.2.6. Copyright Matt Newville,University of Chicago, USA.

Slowey, A. J., Johnson, S. B., Rytuba, J. J. & Brown, G. E. (2005a).Environ. Sci. Technol. 39, 7869–7874.

Slowey, A. J., Rytuba, J. J. & Brown, G. E. (2005b). Environ. Sci.Technol. 39, 1547–1554.

Sumozas, R. (2005). PhD thesis, Castilla-La Mancha University,Spain.

Wasserman, S. R., Allen, P. G., Shuh, D. K., Bucher, J. J. & Edelstein,N. M. (1999). J. Synchrotron Rad. 6, 284–286.

Webb, S. M. (2005). Phys. Scr. T115, 1011–1014.

soil and geosciences

186 Jose Maria Esbrı et al. � XANES speciation of mercury J. Synchrotron Rad. (2010). 17, 179–186

III�

EXTRACTANT�AND�SOLVENT�SELECTION�TO�RECOVER�ZINC

Marta�Avila,�Gustavo�Perez�and�Manuel�Valiente�

Solvent�Extraction�and�Ion�Exchange�(2011)�29:�384–397�

Solvent Extraction and Ion Exchange, 29: 384–397, 2011

Copyright © Taylor & Francis Group, LLC

ISSN 0736-6299 print / 1532-2262 online

DOI: 10.1080/07366299.2011.573434

Extractant and Solvent Selection to Recover Zincfrom a Mining Effluent

M. Avila, G. Perez, and M. Valiente

Universitat Autonoma de Barcelona, Department of Chemistry, Centre GTS,

Bellaterra, Barcelona, Spain

Abstract: The feasibility of three commercial extractants (DEHPA, Cyanex 272,

and Ionquest 290) has been assessed for the recovery of Zn from an acidic mine

effluent. Less than 5 min are required to reach equilibrium for the studied extrac-

tants. Regarding selectivity, DEHPA extracted efficiently Zn, Ca, Mn, and Al,

although Al remained in the solvent extract after stripping, hindering the solvent

reuse. Neither Ionquest 290 nor Cyanex 272 extract Al, Cu, Mn, or Ca signifi-

cantly. Ionquest 290 recovery of Zn is 5–10% higher than Cyanex 272. In addition,

20%(v/v) Ionquest 290 produces higher recoveries than 40%(v/v) DEHPA, thus

Ionquest 290 has been selected as the most suitable among the extractants studied.

Keywords: Mining effluent, solvent extraction, zinc, DEHPA, Ionquest 290; Cyanex

272

INTRODUCTION

The development of viable ways of recycling industrial waters such as

mining effluents rather than the simple disposal of the effluents and their

derivate sludge as a hazardous waste in specially controlled landfills is dam-

aging both environmentally and economically. In a currently abandoned

mine in Andalusia, in the south of Spain, a huge stream of effluent con-

taining about 1 g/l Zn and significant amount of Ca, Cu, Al, and Mn have

to be treated before disposal. Zinc is the fourth most commonly used metal

in the world with over 7 Mt of annual production worldwide, trailing only

iron, aluminum, and copper in annual production due to its broad utility.

Address correspondence to M. Valiente, Universitat Autonoma de Barcelona,

Department of Chemistry, Centre GTS, Campus de la UAB, Edicici CN, 08193,

Bellaterra, Barcelona, Spain. E-mail: [email protected]

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

Zinc Recovery from a Mining Effluent 385

Nearly 50% of the amount of the Zn is used for galvanizing to protect

steel from corrosion, approximately 19% is used to produce brass, and

16% goes into the production of zinc based alloys to supply the die cast-

ing industry. The rest of the zinc is employed to produce roofing, gutters,

and down-pipes, rubber in tires, sunscreen, TV screens, and luminous dials

and ointments to prevent bacteria and fungi from reproducing, amongst

others.[1] Hence, the recovery of Zn from mine waters can provide economi-

cal benefits while diminishing the volume of hazardous materials contained

in the mine tailing.

In this context, conventional treatment methods for zinc extraction

and purification include precipitation, ion exchange, adsorption, electro-

chemical recovery, membrane separation, and solvent extraction (SX).[2]

In this regard, SX has been widely proposed as some of the most econom-

ical and practical processes to extract Zn from waters containing Zn and

other impurities.[3–7] SX involves the extraction of a target element from

the initial solution by an extractant usually diluted in an organic solvent,

leaving other constituents in the aqueous raffinate. Then, a subsequent re-

extraction/stripping of the extracted elements present in the organic phase

(OP) is usually carried out with some acidic solution (stripping solution).

When the organic phase has higher affinity for some metals than the strip-

ping solution, or undesirable metals have also been extracted, scrubbing

of the solvent prior to the stripping of the target elements or regeneration

of the extractant after the stripping process for further applications should

be done. These steps generally increase the cost of the process due to the

expenditure in both reactants and time.

Nowadays, a wide number of extractants are available for use in SX

for the recovery of metals, some of which are suitable for a specific metal,

and others must be used at certain conditions to avoid the extraction

of impurities.[8,9] In this sense, the most widely used extractants for Zn

recovery are those corresponding to the organophosphorus acids group,

that is, DEHPA and Cyanex 272, commonly used in SX. In this study, a

newer commercial extractant, Ionquest 290, is compared with the results

of DEHPA and Cyanex 272 in samples obtained from the Zn rich mine

effluent in order to get a Zn sulphate rich liquor to be used later in an

electrowinning plant.

Di-(2-ethylhexyl) phosphoric acid (DEHPA) has been successfully

used as an extractant for many metal ions including Zn due to its

great extraction capacity and low cost.[10–12] It has been used to

extract Zn more efficiently than other bivalent metal ions such as

Cu, Ni, Co, and Cd.[13] The order of extraction of eight metal ions

from a sulphate solution using DEHPA has been reported as a func-

tion of pH to be Fe3+>Zn2+>Cu2+>Co2+>Ni2+>Mn2+>Mg2+>Ca2+where Zn is extracted much earlier than Mn.[14] In a more recent

study of the separation of divalent metal ions from a synthetic

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

386 M. Avila et al.

laterite leach solution, the extraction of metal ions was in the order

Zn2+>Ca2+>Mn2+>Cu2+>Co2+>Ni2+>Mg2+.[15] By varying the acidic

conditions and the temperature as main parameters, the target metal (or

even different metals) can be separated from the bulk solution by changing

in various steps the conditions to get pure solutions of the target met-

als. Cyanex 272 has been used, as well as its thio-substituted derivatives

(Cyanex 302 and Cyanex 301), in the extraction of several metal ions.[16]

Various studies report the adequacy of Cyanex 272 to extract Fe, Zn, Cr,

Cu, and Ni from sulphuric and/or sulphate solutions.[17–19] In the present

study, to achieve greater recoveries and improved selectivity, another

commercial extractant, Ionquest 290 with the same active ingredient as

Cyanex 272, bis(2,4,4-trimethylpentyl) phosphinic acid [(C8H17)2P(O)OH]

was also studied. In addition, two kerosenes with different flash points were

also studied as a solvent for the extractants.

Thus, the aim of this work was to investigate the SX processes for the

recovery of Zn from a mine effluent using either DEHPA, Cyanex 272,

or Ionquest 290 as extractants to identify the best extractant regarding

the efficiency as well as the process selectivity to recover Zn from that

mine stream. Determination of the best type of kerosene for the mentioned

extraction/stripping process was an additional goal of this study.

EXPERIMENTAL

Sample Description

Fe was removed from the mine water prior to the SX treatment by means

of a biooxidation process using Thiobacillus ferrooxidans and a precipi-

tation step[20,21] to obtain a pregnant leach solution (PLS) without iron,

since no reagents capable of extracting Zn selectively from a solution

containing Fe are commercially available. Major elements present in the

PLS were determined by means of Inductively Coupled Plasma-Optical

Emission Spectroscopy (ICP-OES) (ThermoElemental model Intrepid II

XLS, Franklyn, MA, USA).

Reagents

The extractants DEHPA (Batch ref. 0063829) and Ionquest 290 (Batch Ref.

G05A1) were kindly supplied by Rhodia UK Ltd. and Cyanex 272 was pur-

chased from Cytec Industries BV, Netherlands. Extractants were dissolved

in commercial grade extra-pure aliphatic kerosene Ketrul D80 or Ketrul

D100 (Batch ref. 20062016 and 20061560, respectively) kindly supplied by

Total Fluides France. Ketrul D80 and Ketrul D100, have a flash point of

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

Zinc Recovery from a Mining Effluent 387

72◦C and 100◦C or superior (ISO 2719), respectively. It must be pointed out

that the higher the flashpoint, the lesser the flammability of the kerosene,

and, therefore the higher the security of the solvent extraction process. The

stripping of the organic enriched phase was performed using 2.0 M sul-

furic acid solution. Sulfuric acid 95–98% was purchased from J.T. Baker,

Phillipsburg, NJ. All of them were used as received without any further

purification. Stoppered glass tubes of 50 mL were used for the contact of

the two phases and the agitation took place in a rotating rack. Metal con-

tent in the strip liquor and in the raffinate were determined by means of

a ThermoElemental ICP-OES model Intrepid II XLS (Thermo, Franklyn,

MA, USA).

Procedure

Kinetic Experiments

For the kinetic experiments 10 mL of DEHPA 40% (v/v), Cyanex 272

5% (v/v), or Ionquest 290 5% (v/v) were agitated with 10 mL of PLS

(ratio A/O = 1) in a rotating rack at 5, 10, 20, 30, 40, and 60 min. The

organic phase loaded with the target metal/s (OP) was stripped with 5 mL

of H2SO4 2.0 M during 3 h to ensure complete stripping.

Selectivity Experiments

To determine selectivity, isotherms varying the ratio A/O from 0.1 to 10

were done. Different volumes of Cyanex 272 5% (v/v), Ionquest 290 5%

(v/v), or DEHPA 40% (v/v) in each type of kerosene were equilibrated with

the PLS. After 15 min of equilibration, OP was stripped with 5 mL H2SO4

2.0 M. DEHPA concentration was higher due to efficiency related to the

extraction yield and the extractant cost. No centrifugation of the dual-

phase system was required because of the clear-phase separation obtained.

Selectivity of the solvents towards Zn was determined by the correspond-

ing recovery of Zn and metal impurities and by the amount of metal not

stripped from the OP (remaining %, R); hence the recovery was expressed

as the ratio between the concentration of metal in the strip liquor and the

PLS concentration (Eq. (1)). The %Remaining R was calculated consider-

ing the concentration of the target metal in the raffinate and in the PLS

(Eq. (2)); and the %Remaining OP was considered as the amount of metal

not recovered and not remaining in the raffinate(R) (Eq. (3)).

%Recovery =(

Znstrip

ZnPLS

)×100 (1)

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

388 M. Avila et al.

%Remaining R =(

Znraffinate

ZnPLS

)×100 (2)

%Remaining OP = 100 − %Recovery − %Remaining R (3)

Effect of Extractant Concentration

After selection of the most appropriate extractant, isotherms varying the

ratio A/O from 0.1 to 10 at three different concentrations of Ionquest 290:

5%, 10%, and 20% (v/v) were studied. The dual phase was agitated during

15 min, and the organic phase was stripped afterwards with 5 mL of H2SO4

2.0 M.

RESULTS AND DISCUSSION

The results include characterization of mining water samples, solvent

extraction kinetics, extraction selectivity, and the effect of the selected

extractant concentration.

Sample Description

After Fe removal, the solution was colorless. The content of relevant met-

als is listed in Table 1. After Fe removal small amounts of Fe were found

on the effluent solution but Zn concentration was not affected by this pro-

cess. This solution contains big amounts of Ca, Mn, and Al as the main

impurities from the mine stream. In addition, the PLS is around pH 4.3,

which is a suitable pH for zinc extraction.[22,23]

Kinetics Experiments

Kinetics experiments were conducted in order to determine the differ-

ences between the extractants as well as to determine the time required to

reach equilibrium. The results for the three studied extractants—DEHPA,

Cyanex 272, and Ionquest 290—for Zn Recovery (%) at ratio A/O = 1

Table 1. Characteristics of the mine water after iron removal

[Zn] [Ca] [Cu] [Fe] [Al] [Mn]

pH (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L)

4.3±0.1 881 ± 50 580 ± 20 45 ± 7 1.2 ± 0.9 210 ± 10 195 ± 10

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

Zinc Recovery from a Mining Effluent 389

0

20

40

60

80

100

0 5 10 15 20 25 30 35 40 45 50 55 60

% R

ecov

ery

Time (min)

Extractant kinetics

DEHPA 40% KD80 DEHPA 40% KD100CYANEX 272 5% KD80 CYANEX 272 5% KD100IONQUEST 5% KD80 IONQUEST 5% KD100

Figure 1. Recovery kinetics of DEHPA (circles), Cyanex 272, (squares), and

Ionquest 290 (triangles) using Ketrul D80 (KD80) (solid line) and Ketrul D100

(KD100) (dashed line).

as a function of time are given in Fig. 1. An increase of recovery was

observed in the first 5 min, and after 5 min a plateau was reached indi-

cating that less than 5 min are required to achieve equilibrium under

the given experimental conditions. Recovery achieved for DEHPA 40% is

more than twice higher than that obtained for Cyanex 272 and Ionquest

290, probably due to a DEHPA concentration 8-fold higher than the two

phosphinic extractants. Small differences were observed between Cyanex

272 and Ionquest 290, with a similar equilibration time, with a slightly

higher recovery for Ionquest 290. No relevant differences were observed

for the two types of kerosene employed (different flash point).

Selectivity

For the electrowinning process, an enriched Zn solution with low amounts

of impurities is required. This can be achieved with an extractant which

selectively recovers Zn from the PLS, leaving all the other elements in the

raffinate, or by increasing the process with further steps such as scrubbing

of the OP when a less selective solvent is used, or by further separation

processes. Moreover, it is important to determine the amount of metals

remaining in the organic phase to predict the design of the overall recovery

process, that is, additional scrubbing and washing steps. When elements

are poorly released from the OP to the strip solution, that is, when these

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

390 M. Avila et al.

elements remain in the OP after stripping, hindering its possible reuse, a

solvent regeneration step is mandatory, and this regeneration involves an

increase of the economic costs and time of the entire process.

Taking into account the results obtained for the selectivity experiments

(Figs. 2–4), a logical recovery decrease as the phase ratio A/O increases,

and this is observed for all the extractants and is due to the saturation of

the extractant. For Cyanex 272 and Ionquest 290 a plateau was observed

at a phase ratio A/O > 1 for all the metals analyzed, indicating that the

extractant is saturated at this ratio. For DEHPA, the recovery is still dimin-

ishing which indicates that not all the extractant is complexed with metals

under the extraction conditions.

Metal recovery using DEHPA mostly follows the trend: Zn> Ca> Mn

> Al > Cu. At phase ratio A/O = 1 recovery of Zn was around 75%,

but the recovery of metal impurities was also significant, especially Ca and

Mn, with a recovery of 60% and 30%, respectively, indicating that DEHPA

is poorly selective for Zn extraction (Fig. 2a). Also, around 80% of the Al

remained in the OP after the stripping (Fig. 2b) when using the A/O ratio

from 0.1 to 2, having a fouling effect on the possible reuse of the extractant.

Cyanex 272 recovery (Fig. 3a) followed the trend Zn>>Cu>Mn∼Ca∼Al. In this case, Mn, Ca, and Al are slightly recovered, indicating

Cyanex 272 to have a higher selectivity for Zn than DEHPA. Recoveries

obtained for Zn ranged from 65% (ratio A/O = 0.1) to ∼20% (ratio A/O

> 2) while the recovery of the other metals ranged from 35% (ratio A/O

= 0.1) to less than 5% (ratio A/O > 2). In addition, negligible amounts

of metals (around 1%) were found in the OP (Fig. 3b), indicating that

practically no regeneration of the solvent is required.

Ionquest 290 recovery (Fig. 4a) followed the trend Zn>>Al>Cu∼Mn∼Ca. Zinc recoveries range from 85% (ratio A/O = 0.1) to ∼30% (ratio

A/O > 2) while the recovery of Al range from 20% (ratio A/O = 0.1) to less

than 5% (ratio A/O > 2). Recovery of the other metal analyzed was below

5% in the entire studied range. As Cyanex 272, Ionquest metal remaining in

the OP showed a similar behavior that Cyanex 272, being the concentration

of metal below 5% for all the elements analyzed at the range of the phase

ratio studied (Fig. 4b). Thus, unlike DEHPA, Cyanex 272, and Ionquest

290 selectively extract Zn from a solution containing high amounts of Ca

and other metals in fewer amounts without fouling of the OP. Given that

small amounts of Ca is found in the strip liquor, an extractant regeneration

should be taken into account if the process is conducted several times with

the same extracting OP.

On the other hand, the difference observed on the recovery trends

between DEHPA and the other two extractants can be attributed to their

chemical nature provided that, phosphoric extractants have higher affin-

ity for calcium than phosphinic extractants. In addition, the differences in

trends between Cyanex 272 and Ionquest 290 are relatively very small and

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

Zinc Recovery from a Mining Effluent 391

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio A/O

(a) Recovery DEHPA

Zn KD80 Zn KD100Ca KD80 Ca KD100Al D80 Al D100Mn KD80 Mn KD100Cu KD80 Cu KD100

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio A/O

(b) Remaining OP DEHPA

Zn KD80 Zn KD100Ca KD80 Ca KD100Al D80 Al D100Mn KD80 Mn KD100Cu KD80 Cu KD100

Figure 2. Recoveries (a) and Remaining OP (b) percentages at different A/O ratios

for DEHPA 40% (v/v).

in the same order of magnitude as in the case of Al, Cu, Ca, and Mn. Such

small differences can be explained by both the different phosphinic acid

concentration present in each extractant and to the presence of product

impurities.

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

392 M. Avila et al.

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio A/O

(a) Recovery Cyanex 272

Zn KD80 Zn KD100Ca KD80 Ca KD100Al KD80 Al KD100Mn KD80 Mn KD100Cu KD80 Cu KD100

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio A/O

(b) Remaining OP Cyanex

Zn KD80 Zn KD100Ca KD80 Ca KD100Al KD80 Al KD100Mn KD80 Mn KD100Cu KD80 Cu KD100

Figure 3. Recoveries (a) and Remaining OP (b) percentages at different A/O ratios

for Cyanex 272 5% (v/v).

DEHPA showed poor selectivity towards Zn due to the co-extraction

of Ca resulting on a gypsum precipitate in the stripping solution. Besides

the high amount of Ca and Mn in the strip liquor, high amounts of Al

remained in the OP after the strip step, hence requiring regeneration of

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

Zinc Recovery from a Mining Effluent 393

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio A/O

(a) Recovery IONQUEST

Zn KD80 Zn KD100Ca KD80 Ca KD100Al D80 Al D100Mn KD80 Mn KD100Cu KD80 Cu KD100

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio A/O

(b) Remaining OP Ionquest

Zn KD80 Zn KD100Ca KD80 Ca KD100Al D80 Al D100Mn KD80 Mn KD100Cu KD80 Cu KD100

Figure 4. Recoveries (a) and Remaining OP (b) percentages at different A/O ratios

for and Ionquest 290 5% (v/v).

the solvent prior to their reuse, increasing the cost of the whole pro-

cess. Cyanex 272 and Ionquest 290 showed high Zn selectivity towards Ca

and negligible amounts of metals in the OP, indicating that no extractant

regeneration step is required. When comparing Cyanex 272 and Ionquest

290 recoveries obtained for Zn, it can be pointed out that Ionquest 290

achieved 5–10% higher recoveries than Cyanex 272. These results indicate

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

394 M. Avila et al.

that the most selective extractant for minimizing Ca extraction achieving

good Zn recovery is Ionquest 290. Considering the two different kerosenes

employed, no significant differences were observed independently of the

employed extractant, thus indicating that both of them can be equally fea-

sible for this application. Thus, from an engineering point of view, the use

of Ketrul D100 is recommended due to their lower flammability.

Effect of the Extractant Concentration

Because Cyanex 272 and Ionquest 290 are five to seven times more expen-

sive than DEHPA, their concentration should be as low as possible without

diminishing recovery. Isotherms varying the concentration of Ionquest 290

were done in order to determine a proper concentration of Ionquest 290

that recovers maximum Zn without increasing extractant costs.

From the results collected in Fig. 5, an expected increase of Zn recov-

ery is observed as the extractant concentration increases. When comparing

the results for the different concentrations of Ionquest 290 with DEHPA, it

can be highlighted that Ionquest 20% (v/v) is capable of achieving a higher

recovery than DEHPA 40%. In addition, selectivity was not modified as

the concentration increased and complete stripping of the organic phase

was ensured. Again, no significant differences are observed between the

two different kerosene diluents.

0

10

20

30

40

50

60

70

80

90

100

0 2 4 6 8 10

%

Ratio A/O

Effect of Ionquest 290 concentration

Zn 5% KD80 Zn 5% KD100Zn 10% KD80 Zn 10% KD100Zn 20% KD80 Zn 20% KD100DEHPA 40% KD80 DEHPA 40% KD100

Figure 5. Recovery obtained using Ionquest 290 5% (squares), Ionquest 290 10%

(rhombus), Ionquest 290 20% (triangles), and DEHPA 40% (circles) using Ketrul

D80 (solid line) or Ketrul D100 (dashed line).

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

Zinc Recovery from a Mining Effluent 395

CONCLUSIONS

DEHPA reagent was unable to extract Zn selectively from the solution at

the target pH and temperature of the mine effluent. High amounts of Ca

were extracted, creating a gypsum precipitate in the strip solution, avoid-

ing their use for electrowinning. In addition, Al was extracted from the

PLS but not stripped out, fouling the extractant and inhibiting their reuse.

Neither Cyanex 272 nor Ionquest 290 5% (v/v) indicated Al, Cu, Mn, or

Ca enrichment in the strip liquor, obtaining recoveries of Zn up to 85%.

Although both of them followed similar trends, Ionquest 290 recovery

of Zn is 5–10% higher than Cyanex 272. In addition, Ionquest 290 20%

(v/v) obtained recoveries comparable or even higher than DEHPA 40%

(v/v). Although Ionquest 290 is 5–7 times more expensive than DEHPA,

Ionquest 290 was selected as the most suitable extractant for the target

stream due to its higher selectivity and loading capacity towards Zn extrac-

tion, which avoids both the steps of scrubbing of the gypsum precipitate in

the strip liquor and regeneration of the solvent due to high amounts of

Al not stripped from DEHPA. Besides, the recycling of the organic phase

minimizes the importance of the extractant costs. Both the solvents Ketrul

D80 and Ketrul D100 showed similar behavior, Ketrul D100 is the solvent

recommended due to its lower volatility and flammability.

ACKNOWLEDGMENTS

Thanks are due to Dr. Baruch Grinbaum of the Bateman Company for

his valuable advice. The public company EGMASA (Andalusia, Spain) is

acknowledged for supporting the personnel expenses for the present study.

The Spanish Ministry for Science and Innovation is acknowledged for sup-

porting the laboratory expenses (Project CTQ2009-07432 (Subprograma

PPQ)).

REFERENCES

1. Adriano, D. C. Trace Elements in Terrestrial Environments, 2nd Ed.;

Springer-Verlag: New York, 2001.

2. Sayilgan, E.; Kukrer, T.; Civelekoglu, G.;Ferella F.; Akcil, A.; Veglio,

F.; Kitis, M. A review of technologies for the recovery of metals from

spent alkaline and zinc–carbon batteries. Hydrometallurgy 2009, 97,

158–166.

3. Jha, M.K.; Kumar, V.; Singh, R.J. Solvent extraction of zinc from

chloride solutions. Solv. Extr. Ion Exch. 2002, 20(3), 389–405.

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

396 M. Avila et al.

4. Salgado, A.L.; Veloso, A.M.O.; Pereira, D.D.; Gontijo, G.S.; Salum,

A.; Mansur, M.B. Recovery of zinc and manganese from spent alkaline

batteries by liquid-liquid extraction with Cyanex 272. J. Power Sources

2003, 115, 367–373.

5. Devi, N. B.; Nathsarma, K. C.; Chakravortty, V. Extraction and sepa-

ration of Mn(II) and Zn (II) from sulfate solutions by sodium salt of

Cyanex 272. Hydrometallurgy 1997, 45(1–2), 169–179.

6. Kolev, S. D.; Baba, Y.; Cattrall, R.W.; Tasaki, T.; Pereira, N.; Perera,

J.M.; Stevens, G.W. Solid phase extraction of zinc(II) using a PVC-

based polymer inclusion membrane with di(2-ethylhexyl)phosphoric

acid (D2EHPA) as the carrier. Talanta 2009, 78, 795–799.

7. Vahidi, E.; Rashchi, F.; Moradkhani D. Recovery of zinc from an

industrial zinc leach residue by solvent extraction using D2EHPA.

Miner. Eng. 2009, 22, 204–206.

8. El-Nadi, Y.A.; Daoud, J.A.; Aly, H.F. Leaching and separation of zinc

from the black paste of spent MnO2-Zn dry cell batteries. J. Hazard.

Mater. 2007, 143, 328–334.

9. Kyuchoukov, G.; Zhivkova, S. Options for the separation of Copper

(II) and Zinc (II) from chloride solutions by KELEX 100. Solv. Extr.

Ion Exch. 2000, 18(2), 293–305.

10. Ritcey, G. M.; Lucas, B. H. Purification of manganese solutions

containing copper and zinc by liquid-liquid extraction, using bis(2-

ethylhexyl)phosphoric acid. Can. Metall. Quart. 1971, 10(3), 223–228.

11. Kunzmann, M; Kolarik, Z. Extraction of zinc(II) with di(2-ethylhexyl)

phosphoric acid from perchlorate and sulfate media. Solv. Extr. Ion

Exch. 1992, 10(1), 35–49.

12. Sastre, A. M.; Miralles, N.; Aguilar, M. The extraction of copper(III)

from sulfate and perchlorate solutions by di(2-ethylhexyl)phosphoric

acid dissolved in Isopar-H. Chem. Scripta 1984, 24(1), 44–48.

13. Grimm, R.; Kolarik, Z. Acidic organophosphorus extractants. XIX.

Extraction of copper(II), cobalt(II), nickel(II), zinc(II), and cad-

mium(II) by bis(2-ethylhexyl) phosphoric acid. J. Inorg. Nucl. Chem.

1974, 36(1), 189–192.

14. Ritcey, G.M.; Ashbrook, A.W. Solvent Extraction. Principles and

Applications to Process Metallurgy. Part I and II . Elsevier: Amsterdam;

1984.

15. Cheng, C. Y. Purification of synthetic laterite leach solution by solvent

extraction using D2EHPA. Hydrometallurgy 2000, 56(3), 369–386.

16. Rickelton, W. A.; Boyle, R. J. The selective recovery of zinc with new

thiophosphinic acids. Solv. Extr. Ion Exch. 1990, 8(6), 783–797.

17. Lanagan, M. D.; Ibana, D. C. The solvent extraction and stripping of

chromium with Cyanex 272. Miner. Eng. 2003, 16(3), 237–245.

18. Rickelton, W.A.; Flett, D.S.; West, D.W. Cobalt-nickel separation by

solvent extraction with bis(2,4,4 trimethylpentyl)phosphinic acid. Solv.

Extr. Ion Exch. 1984, 2(6), 815–838.

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

Zinc Recovery from a Mining Effluent 397

19. Reddy, B. R.; Sarma, P.V.R.B. Separation and recovery of cobalt and

nickel from sulfate solutions of Indian Ocean nodules using Cyanex

272. Miner. Process Metall. 2001, 18(3), 172–176.

20. Mazuelos, A.; Carranza, F.; Palencia, I.; Romero, R. High efficiency

reactor for the biooxidation of ferrous iron. Hydrometallurgy 2000,

58(3), 269–275.

21. Carranza, F.; Iglesias, N.; Romero, R.; Palencia, I. Kinetics improve-

ment of high-grade sulfides bioleaching by effects separation. FEMS

Microbiol. Rev. 1993, 11(1–3), 129–138.

22. Northcott, K.; Kokusen, H.; Komatsu, Y.; Stevens, G. Synthesis and

surface modification of mesoporous silicate SBA-15 for the adsorption

of metal ions. Sep. Sci. Technol. 2006, 41, 1829–1840.

23. Tsakiridis, P.E.; Oustadakis, P.; Katsiapi, A.; Agatzini-Leonardou S.

Hydrometallurgical process for zinc recovery from electric arc furnace

dust (EAFD). Part II: Downstream processing and zinc recovery by

electrowinning. J. Hazard. Mater. 2010, 179, 8–14.

Downloaded By: [Consorci de Biblioteques Universitaries de Catalunya] At: 12:27 6 June 2011

IV�

ZINC�RECOVERY�FROM�AN�EFFLUENT�USING�IONQUEST�290:�FROM�

LABORATORY�SCALE�TO�PILOT�PLANT

M.�Avila,�B.�Grinbaum,�F.�Carranza,�A.�Mazuelos,�R.�Romero,�N.�Iglesias,�J.L.�Lozano,�G.�Perez,�M.�Valiente.��

Hydrometallurgy�(2011),�107:�63�67�

Zinc recovery from an effluent using Ionquest 290: From laboratory scale topilot plant

M. Avila a, B. Grinbaum b, F. Carranza c, A. Mazuelos c, R. Romero c, N. Iglesias c, J.L. Lozano d,G. Perez a, M. Valiente a,⁎a Universitat Autonoma de Barcelona, Dept de Química, 08193, Bellaterra, Spainb Bateman Litwin N.V. POB 15, Yokneam 20692, Israelc Universidad de Sevilla, Depto. de Ingeniería Química, 41092 Sevilla, Spaind EGMASA, Isla de la Cartuja, 41092 Sevilla, Spain

a b s t r a c ta r t i c l e i n f o

Article history:Received 28 September 2010Received in revised form 17 January 2011Accepted 17 January 2011Available online 17 February 2011

Keywords:Mine tailing pondZinc extractionFe bioxidation removalBateman Pulsed ColumnIonquest 290

A stream of effluent from a mine tailings pond, containing zinc, ferrous ions and other metals, requiredtreatment to prevent pollution and recover valuable metals. A solvent extraction (SX) process using Ionquest290 as extractant was developed to recover the Zn from the effluent. Ferrous ions were bio-oxidized andremoved by selective alkaline precipitation prior to the zinc extraction. The Fe removal as well as the SXprocess were developed successfully at laboratory scale and verified in a pilot plant on-site, using twoBateman Pulsed Columns for the extraction and stripping of Zn. The results were satisfactory obtaining above95% recovery of the Zn.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

An abandoned mine in Andalusia, Spain, has a huge stream ofeffluent, estimated to be 10,000 m3/day, flowing into a tailings pond.Since the effluent contains about 1 g/L of Zn and significant amountsof ferrous, ferric, calcium, copper, aluminum and manganese ions,their removal is required in order to prevent pollution of a naturereserve downstream from the area. The recycling of such effluentsrather than simple neutralisation and disposal as a hazardous wastecan provide an economical benefit, while diminishing the volume ofhazardous materials contained in the mine tailing. Zinc metal has ahigh economical value and recycling can add economical value tothose residues.

Conventional methods for separation of pure Zn include precipitation,ion exchange, adsorption, electrochemical recovery, membrane separa-tion and solvent extraction (SX) (Sayilgan et al., 2009) with SX being themost economical and practical process to extract Zn from industrialwaters (Devi et al., 1997; Jha et al., 2002; Salgado et al., 2003). In recentyears SX has become essential to the hydrometallurgical industry due to agrowing demand for high purity metals, rigid environmental regulations,the need for lower production costs, as well as due to the diminishingproduction in high-grade ore reserves (Alamdari, 2004; Owusu, 1998). In

this sense, the organophosphorus acids, i.e., DEHPA (Ritcey and Lucas,1971) and Cyanex272 (Lanagan and Ibana, 2003) and their thiosub-stituted derivatives Cyanex 301 and Cyanex 302 (Rickelton and Boyle,1990) have been the most widely used extractants to recover Zn.

In the present study, a recently commercialized organophosphorusextractant, Ionquest 290, has been employed for the selectiverecovery of zinc from a mine effluent located in Aznalcollar (Sevilla,Spain). Ionquest 290 has the same active ingredient as Cyanex 272, bis(2,4,4-trimethylpentyl) phosphinic acid, but has a lower content ofinactive impurities, the phosphine oxide impurity is b5% in Ionquest290 but around 15% in Cyanex 272 (Barnard and Shiers, 2010).

Iron ions are more strongly extracted than the majority of metals(including Zn) by most of the known commercial extractants (Lupiand Pilone, 2000; Yokoyam et al., 1996) so it needs to be removedprior to SX process. For this purpose, a process based on the bio-oxidation of Fe2+, using specific bacteria, followed by selectivealkaline precipitation of Fe3+, was required (Mazuelos et al., 1999,2010a).

EGMASA, the regional environmental government company inAndalusia (Spain), suggested the recovery of Zn from the mentionedeffluent by combining solvent extraction (SX) and electrowinning(EW) process. Therefore, this process should be environmentallyfriendly and also to produce an economically effective output. At least95% of the Zn must be recovered from the effluents in order to satisfythe environmental requirements and the SX plant should provide afinal product stream of 90 g/L Zn in the stripping step (strong

Hydrometallurgy 107 (2011) 63–67

⁎ Corresponding author. Tel.: +34 935812903; fax: +34 935811985.E-mail address: [email protected] (M. Valiente).

0304-386X/$ – see front matter © 2011 Elsevier B.V. All rights reserved.doi:10.1016/j.hydromet.2011.01.009

Contents lists available at ScienceDirect

Hydrometallurgy

j ourna l homepage: www.e lsev ie r.com/ locate /hydromet

electrolyte) by using a spent electrolyte with 50 g/L Zn, to fulfill theoperating conditions of the EW plant. The whole bio-oxidation and Feprecipitation, as well as the SX process were developed successfully atlaboratory scale and afterwards verified in a pilot plant on-site.

2. Materials and methods

2.1. Iron removal

Laboratory testswere aimedatdetermining theoperatingconditionsfor the pilot plant. To reach the target concentration of b5 ppm it isnecessary to completely convert ferrous to ferric ions in the bio-oxidation stage.

Bio-oxidation laboratory tests were performed in a methacrylatecolumn packed with siliceous stone particles and inoculated with amixed culture of Acidithiobacillus ferrooxidans, Leptospirillum ferrooxidansand some heterotrophic bacteria including Acidiphillium organovorum,facilis and cryptum (Mazuelos et al., 2010b) The inoculum was obtainedfrom the Riotinto Mine acid mine drainage waters. The culture wasroutinely maintained on a modified Silverman and Lungren 9 K nutrientmedium at pH 1.25 (adjusted with H2SO4) in the University of Sevillelaboratories. The effluent solution was fed into the bottom of the columnand overflowedwhile air was supplied under pressure (0.5 bar) from thebottom. Precipitation tests were carried out in stirred reactors with pH-controlled addition of alkaline reagent. The pH and pumping ratesdetermined the length of tests, avoiding a rapid increase in pH thatwould lead to undesirable co-precipitation phenomena. Ferrous concen-tration was determined by end-point automatic titration with K2Cr2O7

while total metal concentration was determined by AAS.At the pilot plant site, the bio-oxidation process took place in a

bioreactor consisting of a 150 cm high and 70 cm diameter stainlesssteel column divided to three zones: a 30 cm deep bottom spacewhere air and solution were fed in, a siliceous stone packed bedcontaining the inoculum supported by a stainless steel screen and anair space at the top where pH and Eh control was made, a 50 mm pipeformed the solution overflow. The effluent circulated through a tankwhere pHwas initially adjusted and, after pH adjustment, the solutionwas transferred to the bioreactor where bio-oxidation of ferrous ionstook place, to be finally transferred to a precipitation tank fed by a

lime solution from a separated tank. The bioreactor operatedcontinuously with a residence time of 60 min and connected to thepilot plant in order to bio-oxidize a total volume of 56 m3 of minewater. After precipitation and sedimentation of iron compounds, thesupernatant was directly used as feed solution to the solventextraction stage (Fig. 1).

2.2. Solvent extraction

For the laboratory investigation of the solvent extraction process, asolution of 5% Ionquest 290 dissolved in kerosene was used. Ionquest290 (PurityN95%) was supplied by Rhodia UK Ltd. and commercialgrade extra-pure aliphatic kerosene Ketrul D100 (bp 100 °C) by TotalFluides France. All reagents were used as received without furtherpurification. A solution of Na2CO3 was used for pH adjustment duringthe solvent extraction experiments. For the stripping step, the loadedsolvent was contacted with 2 M H2SO4 at a phase ratio O:A=10, theinitially expected phase ratio in the plant to achieve the required zinctransfer in the EW plant of 20 g/L. In practice, it was found that therequired transfer in the EW plant was 40 g/L Zn, consequently, thephase ratio was modified to O:A=20. No laboratory tests wereundertaken at this phase ratio, but directly applied in the pilot plant.

Two Bateman Pulsed Columns (BPC) were required for the SX andstripping processes at the pilot plant due to their demonstratedfeasibility in several SX plants (Ferreira et al., 2010; Gameiro et al.,2010; Sole et al., 2005). BPC are large diameter vertical pipes filledalternately with disk and doughnut shaped baffles to promote contactbetween the organic and aqueous phases through the column. Adecanter at each end of the column allows the liquids to coalesce andbe decanted separately. When the solvent phase is continuous, theinterface between the phases is in the lower decanter and when theaqueous phase is continuous, it is in the upper decanter. The columnsare pulsed by blowing air at the required amplitude and frequency ofthe pulses (Ritcey, 2006).

An 80 mm diameter BPC, 7 m high (equivalent to 3 theoreticalmixer-settler stages) was chosen for the SX process and a 40 mmdiameter BPC 6 meter high for the stripping. The piping of the plant isshown in Fig. 1.

BIO-OXIDATIONREACTOR

Area of distribution ofair and liquid

Diffusers

pH ADJUSTMENT TANK

Feed

Air

PRECIPITATION TANK

LIME TANK

Loadedsolvent

(LS)

EXTRACTION COLUMN

disk

doughnut

Raffinate

Soda

Lowerdecanter

Upperdecanter

ORGANIC SOLVENT

TANK

Strongelectrolyte (SE)

Weakelectrolyte

(WE)

Barren solvent (BS)

STRIPPING COLUMN

Bed containingthe inoculum

Fig. 1. Bio-oxidation and SX flowsheet at the pilot plant.

64 M. Avila et al. / Hydrometallurgy 107 (2011) 63–67

All flows were fed through metering pumps and the flow rates ofall inlets and aqueous outlets were measured by rotameters. The pilotwas run for 12 working days, 10 h a day on average, i.e. a total of120 h. The average flow rate of the aqueous feed was 150 L/h, so,about 18m3 of tailing solution after Fe precipitation were treated. Thetotal volume of the solvent was 300 L and it had 5% Ionquest 290dissolved in kerosene (20% aromatic and 80% aliphatic); the weakelectrolyte (WE, strip solution) consisted of 190 g/L H2SO4 with 50 g/LZn2+. A solution of 50–100 g/L Na2CO3 was prepared periodically in a60 L barrel and used to adjust the pH.

The concentration of Zn was determined using a Perkin Elmer3110 AAS at the mine laboratory. The Zn in the raffinate and SE wasdetermined directly, while the Zn in the barren and loaded solvent (BSand LS) solution were determined after stripping using H2SO4.

3. Results and discussion

3.1. Iron removal

The representative composition of the major components in theeffluents was Zn 1000±100 mg/L, Fe 500±50 mg/L (36% ferric), Ca600±50 mg/L, Mn 200±20 mg/L and Cu 50±5 mg/L. The pH of theeffluent was always around 3.0.

After the bio-oxidation process, laboratory scale precipitationexperiments indicated that a final pH of 4.5 was reached after 65 minand Fe precipitation was almost complete, with the remainingconcentration below 0.5 ppm. The initial Zn concentration waspractically unaffected by this process (Fig. 2). Lime consumptionwas 4.2 g CaO per kg of solution.

Similar results were obtained in the continuous bio-oxidationprocess at the pilot plant (Fig. 3) achieving total ferrous oxidation atall times and flow rates tested. Table 1 shows the effect of the bio-oxidation – precipitation stage. Iron precipitation took 60 min in thepilot plant. During this time, the lime addition to reach pH 3.5 took45 min followed by intermittent dosing for the next 15 min until pH4.7 was achieved. After the precipitation step Fe was completelyremoved, the amount of Al decreased drastically, Cu dropped by half(from 45.0 ppm to 21.7 ppm) while the concentration of the otherelements measured, including Zn, remained similar to the initial.Precipitation and sedimentation stages accurately reproduced labo-ratory results, producing 56 m3 of iron-free solution with practicallyall the initial zinc.

3.2. Computer simulation

Computer simulations were performed to estimate the requiredpilot plant inputs and outputs by using CurveExpert 1.3 to calculatethe distribution coefficients, D, and the number of stages). Experiencehas shown that computer simulation is a more flexible design toolthan McCabe–Thiele diagrams for pulsed columns (Grinbaum, 1992;Grinbaum, 1993; Gottliebsen et al, 2000). The results obtained in the

simulation, collected in Table 2, determined that at a phase ratio O:A=0.5–0.6, a two-stage column is enough to recover more than 95%of the effluent Zn. The addition of a third stage enables either todecrease the phase ratio O:A to 0.4 or to work with a phase ratio of O:A=0.5 and obtain a recovery of Zn near to 99%, i.e. b10 ppm Zn in theraffinate. The concentration of Zn in the loaded solvent should be inthe range of 2.2–2.8 g/L extractant, around 70–85% of the theoreticaltotal loading of 3.3 g/L, which is quite reasonable. In order to get a finalsolution of 90 g/L Zn in the SE, i.e. a Zn transfer of 40 g/L, the strippingshould be run at a phase ratio of O:A=20, and only one equilibriumstage is required.

3.3. Solvent extraction and stripping

The initial concentration of Ionquest 290 was chosen to be 5%.Using a higher concentration would require an extreme O:A phaseratio, while a lower concentration would increase the flow rate of thesolvent and, accordingly, the size of the stripping unit. The maximumloading that was obtained experimentally at limiting conditions, i.e.,by contacting 3 times the solvent with corresponding fresh portions ofeffluent at phase ratio O:A=0.1, was 2.9 g Zn/L Ionquest 290. Sincethis result is similar to the obtained after a single contact, it revealsthat the limiting conditions can be achieved by a single contact.

Tests to determine pH control were done at laboratory. From theresults shown in Table 3, it can be seen that without pH control, theextraction was quite selective. Thus, no Mn, Cu and Al were extractedand only a small amount of Ca was extracted. In addition, separationfactors were high enough to support the indicated selectivity.However, the distribution ratio of Zn (DZn) at natural pH range waslow, especially at the dilute end of the process (phase ratio O:A=10).In addition, the pH of the raffinate (final pH) dropped from 2.6 to 2.1as the phase ratio O:A increased, while the suitable pH for Znextraction by Ionquest 290 is above 2.5 (Tsakiridis et al., 2010).Furthermore, to avoid Ca co-extraction, pH should be around 3 asindicated by the isotherms in the Cyanex 272 online User Manual p. 5

time (min)

[Zn]

(pp

m)

pH; [

Fe]

(pp

m)

[Zn] pH [Fe]

900920940960980

100010201040106010801100

00,511,522,533,544,55

0 10 20 30 40 50 60 70 80

Fig. 2. Evolution of [Zn], [Fe] and pH in laboratory precipitation tests.

Time (h)[Fe(II)]in [Fe(II)]out [Zn]out [Zn]in Flow rate

0

200

400

600

800

1000

1200

0

50

100

150

200

250

300

350

400

0 20 40 60 80 100 120 140 160

[Fe(

II)],

[Zn]

(pp

m)

Flo

w r

ate

(L/h

)

Fig. 3. Bio-oxidation process at the pilot plant.

Table 1Solution composition, before and after treatment.

Initial solution After bio-oxidation After precipitation

[Fe2+] (ppm) 254 0 0[Fe3+] (ppm) 446 690 0.2[Zn] (ppm) 1020 1020 1010[Al] (ppm) 292 250 20[Mn] (ppm) 265 260 200[Cu] (ppm) 45 45 21.7[Ca] (ppm) 600 600 600[Pb] (ppm) 1.6 1.6 1.6pH 3.0 1.93 4.78

65M. Avila et al. / Hydrometallurgy 107 (2011) 63–67

from Cytec Corporation (http://www.cytec.com/specialty-chemicals/PDFs/CYANEX%20272.pdf, accessed 26th December 2010).

As seen in Table 4, adjusting to pH=3 results in higher zinc thanwithout pH adjustment. The extraction of Mn and Ca remains quitelow as indicated by the high separation factors. Therefore, the pH atthe pilot plant should be maintained around pH 3. In practice, the pHadjustment was achieved by direct neutralization of both the acidicraffinate and the organic solvent (by pre-equilibration with aqueoussolution) using Na2CO3. The consumption of Na2CO3 was 1.62 kg/kgZn.

Shake out stripping experiments were carried out by contacting200 mL of loaded solvent (LS) containing 1.95 g/L Zn with 20 mL ofaqueous phase containing 200 g/L H2SO4 and weak electrolyte (WE)containing zinc concentrations of 40 to 90 g/L at 22 °C. The processwas carried out at O:A=10, i.e. the concentration of Zn in the strongelectrolyte (SE) should increase by ~20 g/L with respect to the WEsolution, whichwas consistent with the results shown in Table 5. In allcases, remaining Zn in the barren solvent (BS) was only 5–17 ppm, i.e.almost all zinc was recovered. Thus, one stage of stripping is enoughfor the Zn recovery regardless of the concentration of Zn in thestripping solution.

Additional laboratory tests carried out at the mine site during thepilot plant experiments at phase ratio O:A=20, revealed that theloaded solvent from the pilot plant was efficiently stripped in onecontact, i.e. one stage, by the strip solution used in the pilot plantexperiments, using a weak electrolyte with ~50 g/L Zn producing anSE containing 90 g/L Zn, i.e. a zinc transfer of 40 g/L, as required for theEW plant.

Preliminary hydraulic tests at the pilot plant showed that theavailable flux is above 30 m3/m2/h in both columns. The stripping wasrun mainly in order to produce BS and was not optimised. It wasoperated at a flux of 40 m3/m2/h (35 L/h solvent), the pulsing had anamplitude of 15 mm and a frequency of 1 Hz. The flow rate of the WEthrough the pump was 5–7 L/h. The average value of Zn in the BS wasabout 20 mg/L Zn.

Three tests with organic continuous dispersion and with aqueouscontinuous dispersion were undertaken to determine the preferreddispersion. During both organic continuous and aqueous continuousruns, the temperature rose from 25 °C in the morning to 34 °C in theevening, which facilitated the comparison between both dispersions.

Every test took 5 h, long enough to reach steady state and the phaseratio was kept at A:O=2.1 during all the testwork.

As seen from Table 6, similar results were obtained with bothdispersions. The Zn concentration in the LS was around 2000 mg/Land Zn in the raffinate was far below 50 mg/L, indicating than N95% ofthe Zn is recovered. Hence, the extraction process was found tooperate successfully with both aqueous and organic continuousdispersions at column fluxes of about 40 m3/m2/h, at 23–34 °C. Asthe available flux and recovery with both dispersions were similar, itis preferable to use the aqueous continuous dispersion as there is alower expenditure on solvent.With aqueous continuous the danger offire due to kerosene ignition is also significantly diminished.

The stripping of the LS containing around 2 g/L Zn, achieved a SEwith 30–40 g/L Zn above the WE, i.e. a zinc transfer of 30–40 g/Lwhilst achieving b50 mg/L Zn in the BS at a flux of 45 m3/m2/h withaqueous continuous at O:A phase ratio of 20.

While the column worked well and supplied the required BS to theextraction, laboratory tests proved that there was no need for an extracolumn, and one stage of mixer-settler was sufficient to obtain therequired zinc transfer of 40 g/L with barren solvent containing ~50 mg/LZn.

4. Conclusions

The results, as demonstrated by the pilot plant, proved the processfeasibility with 95% Zn recovery from the effluent. The pre-treatmentstage bio-oxidation achieved complete oxidation of ferrous in thebioreactor, subsequent lime precipitation resulted in b1 mg/L Feremaining in the solution whilst not affecting the zinc tenor. The zincextraction stage was successfully carried out in a 80 mmdiameter 7 mhigh Bateman Pulsed Column leaving b50 mg/L Zn in the raffinate.The required phase ratio was O:A=0.5 for a solution flux of 44 m3/m2/h. The system working with both aqueous and organic continuousdispersions, at a temperature above 25 °C. The stripping was efficientwith only a single stage at O:A phase ratio of 20 required to achieve a40 g/L Zn transfer into the electrolyte. The Na2CO3 consumption was1.7 kg/m3 effluent (1.7 kg/kg Zn).

Table 2Recovery of Zn vs. Plant Configuration, using 5% Ionquest 290, ZnPLS=0.95 g/kg.

No. stages Phase ratio O:A Zn in raff. (ppm) Recovery (%)

2 0.50 51 94.70.60 24 97.6

3 0.35 75 92.10.40 30 96.80.45 11 98.90.5 4 99.6

Table 3Extraction experiments with 5% Ionquest 290, no pH correction, 22 °C.

Phase ratio FinalpH

Aqueous (mg/L) Organic (mg/L) D values &separation factors

O:A Zn Mn Ca Zn Mn Ca DZn DZn/DCa DZn/DMn

PLS 5.0 962 206 7630.1 2.58 792 208 618 1595 0.03 8 2.0 154.5 1 104

0.3 2.50 692 208 613 910 0.2 12 1.3 66.4 13500.5 2.31 621 205 605 668 0.1 15 1.1 44.4 22601 2.18 536 206 624 444 0.1 13 0.8 38.4 16002 2.16 467 207 613 273 0.03 9 0.6 40.9 42003 2.32 402 202 597 178 0.3 10 0.4 23.9 2705 2.25 342 203 599 132 0.1 6 0.4 39.9 80010 2.10 270 203 601 76 0.1 8 0.3 22.5 610

Table 4Extraction experiments with 5% Ionquest 290 at pH 3, 22 °C.

Phase ratio Aqueous(mg/L)

Organic(mg/L)

D values &separation factors

O:A Zn Mn Ca Zn Mn Ca DZn DZn/DCa DZn/DMn

PLS (pH 5.0) 963 213 5830.1 784 231 531 2878 0 76 3.7 26.4 9·104

0.3 343 233 509 2073 0.4 72 6.0 42.9 3·103

0.5 182 211 536 1700 0.5 40 9.3 133.0 5·103

1 49 213 547 875 1.5 38 17.9 199.9 3·103

2 22 194 534 502 3.4 43 22.8 285.0 1 103

3 14 127 532 297 2.7 46 21.2 235,6 1 103

5 4 181 584 192 3 42 48 685,7 2 103

10 1 183 579 90 1 48 90 1125 2·104

Table 5Stripping experiments with 5% Ionquest 290 at 22 °C, at phase ratio O:A=10.

Aqueous in Aqueous out

Zn(g/L) H2SO4 (g/L) Zn (g/L)

40 200 58.850 194 68.260 188 77.870 176 91.080 200 102.890 200 115.4

66 M. Avila et al. / Hydrometallurgy 107 (2011) 63–67

For a Zn price above US$2/kg, the value of the zinc product coversthe operating costs in addition to solving a serious environmentalproblem.

Acknowledgements

The public company EGMASA (Andalusia, Spain) is acknowledgedfor their support of the personnel expenses for the present study. TheSpanish Ministry for Science and Innovation is acknowledged forsupporting laboratory expenses at UAB (Project CTQ2009-07432(subprograma PPQ)).

References

Alamdari, E.K., 2004. Synergistic effect of MEHPA on co-extraction of zinc and cadmiumwith DEHPA. Miner. Eng. 17, 89–92.

Barnard, K.R., Shiers, D.W., 2004. Mechanism and rate of butyl phosphinate formationfrom reaction of phosphinic acid (Cyanex 272) and tributyl phosphate. Hydromet-allurgy 106, 76–83.

Devi, N.B., Nathsarma, K.C., Chakravortty, V., 1997. Extraction and separation of Mn(II)and Zn (II) from sulfate solutions by sodium salt of Cyanex 272. Hydrometallurgy45, 169–179.

Ferreira, A.N., Agarwal, S., Machado, R.M., Gameiro, M.L.F., Santos, S.M.M., Reis, M.T.A.,Ismael, M.R.C., Correia, J.N., Carvalho, J.M.R., 2010. Extraction of copper from acidicleach solution with Acorga M5640 using a pulsed sieve plate column. Hydromet-allurgy 104, 66–75.

Gameiro, M.L.F., Machado, R.M., Ismael, M.R.C., Reis, M.T.A., Carvalho, J.M.R., 2010.Copper extraction from ammoniacal medium in a pulsed sieve-plate column withLIX 84-I. J. Hazard. Mater. 183, 165–175.

Gottliebsen, K., Grinbaum, B., Chen, D., Stevens, G., 2000. Recovery of sulfuric acid fromcopper electrolyte bleeds. Hydrometallurgy 56, 293–307.

Grinbaum, B., 1992. How to fit a simulator to a real liquid-liquid extraction plant. In:Sekine, T. (Ed.), Solvent Extraction. Elsevier, Amsterdam, pp. 1235–1240.

Grinbaum, B., 1993. The development of liquid–liquid extraction processes usingsimulation”. In: Longsdail, D.H., Slater, M.J. (Eds.), Solvent Extraction in the ProcessIndustries. Elsevier, Amsterdam, pp. 715–722.

Jha, M.K., Kumar, V., Singh, R.J., 2002. Solvent extraction of zinc from chloride solutions.Solv. Extr. Ion Exch. 20, 389–405.

Lanagan, M.D., Ibana, D.C., 2003. The solvent extraction and stripping of chromiumwithCyanex 272. Miner. Eng. 16, 237–245.

Lupi, C., Pilone, D., 2000. Reductive stripping in vacuum of Fe(III) from D2EHPA.Hydrometallurgy 57, 201–207.

Mazuelos, A., Romero, R., Palencia, I., Iglesias, N., Carranza, F., 1999. Continuous ferrousiron biooxidation in flooded packed bed reactors. Miner. Eng. 12, 559–564.

Mazuelos, A., Carranza, F., Romero, R., Iglesias, N., Villalobo, E., 2010a. Operational pH inpacked-bed reactors for ferrous ion bio-oxidation. Hydrometallurgy 104, 186–192.

Mazuelos, A., Iglesias, N., Romero, R., Mejías, M.A., Carranza, F., 2010b. Influence of zincon ferrous iron bio-oxidation: biological or physical nature? Biochem. Eng. J. 49,235–240.

Owusu, G., 1998. Selective extractions of Zn and Cd from Zn–Cd–Co–Ni sulphatesolution using di-2-ethylhexyl phosphoric acid extractant. Hydrometallurgy 47,205–215.

Rickelton, W.A., Boyle, R.J., 1990. The selective recovery of zinc with new thiopho-sphinic acids. Solv. Extr. Ion Exch. 8, 783–797.

Ritcey, G.M., 2006. Solvent extraction principles and applications to process metallurgy,vol.1. G.M. Ritcey & Associates, Ottawa, pp. 336–348.

Ritcey, G.M., Lucas, B.H., 1971. Purification of manganese solutions containing copperand zinc by liquid-liquid extraction, using bis(2-ethylhexyl)phosphoric acid. Can.Metall. Quart. 10, 223–228.

Salgado, A.L., Veloso, A.M.O., Pereira, D.D., Gontijo, G.S., Salum, A., Mansur, M.B., 2003.Recovery of zinc and manganese from spent alkaline batteries by liquid-liquidextraction with Cyanex 272. J. Power Sources 115, 367–373.

Sayilgan, E., Kukrer, T., Civelekoglu, G., Ferella, F., Akcil, A., Veglio, F., Kitis, M., 2009. Areview of technologies for the recovery of metals from spent alkaline and zinc–carbon batteries. Hydrometallurgy 97, 158–166.

Sole, K.C., Feather, P.A.M., Cole, M., 2005. Solvent extraction in southern Africa: anupdate of some recent hydrometallurgical developments. Hydrometallurgy 78,52–78.

Tsakiridis, P.E., Oustadakis, P., Katsiapi, A., Agatzini-Leonardou, S., 2010. Hydrometal-lurgical process for zinc recovery from electric arc furnace dust (EAFD). Part II:Downstream processing and zinc recovery by electrowinning. J. Hazard. Mater. 179,8–14.

Yokoyam, T., Yokota, T., Hayashi, S., Izawa, E., 1996. Determination of trace gold in rocksamples by a combination of two-stage solvent extraction and graphite furnaceatomic absorption spectrometry: the problem of iron interference and its solution.Geochem. J. 30, 175–181.

Table 6Extraction in organic and aqueous continuities.

Continuousdispersion

Feed BS Flux pH Zn (mg/L)

L/h L/h m3/m2/h Raff Raff LS

Organic 110 55 33 2.7 11 1910130 60 38 2.8 55 1880130 60 38 2.9 11 1800

Aqueous 150 70 44 2.9 47 1880150 70 44 3.1 18 2520150 70 44 2.9 7.2 2240

67M. Avila et al. / Hydrometallurgy 107 (2011) 63–67