To link to this article:
DOI:10.1016/j.jhazmat.2011.11.027
http://dx.doi.org/10.1016/j.jhazmat.2011.11.027
This is an author-deposited version published in:
http://oatao.univ-toulouse.fr/
Eprints ID: 5591
To cite this version: Pokrovsky, O.S and Probst, Anne and Leviel, E. and Liao,
Bo-han Interactions between cadmium and lead with acidic soils: Experimental
evidence of similar adsorption patterns for a wide range of metal concentrations
and the implications of metal migration. (2012) Journal of Hazardous Materials,
vol. 199-200 . pp. 358-366. ISSN 0304-3894
O
pen
A
rchive
T
oulouse
A
rchive
O
uverte (
OATAO
)
OATAO is an open access repository that collects the work of Toulouse researchers and
makes it freely available over the web where possible.
Any correspondence concerning this service should be sent to the repository
Interactions
between
cadmium
and
lead
with
acidic
soils:
Experimental
evidence
of
similar
adsorption
patterns
for
a
wide
range
of
metal
concentrations
and
the
implications
of
metal
migration
O.S.
Pokrovsky
a,
A.
Probst
b,c,∗,
E.
Leviel
a,
B.
Liao
d aGET-CNRS-UPS-IRD-UMR5563,14,AvenueEdouardBelin,31400Toulouse,FrancebUniversitédeToulouse;INPT,UPS;EcoLab(LaboratoireEcologiefonctionnelleetEnvironnement);ENSAT,Avenuedel‘Agrobiopôle,31326Castanet-Tolosan,France cCNRS;EcoLab;31326Castanet-Tolosan,France
dInternationalCollege,CentralSouthUniversityofForestryandTechnology,Changsha410004,China
a
r
t
i
c
l
e
i
n
f
o
Keywords: Cadmium Lead Acidsoils Adsorption Isotherm Competitiona
b
s
t
r
a
c
t
Theimportanceofhigh-andlow-affinitysurfacesitesforcadmiumandleadadsorptionintypical Euro-peanandAsiansoilswasinvestigated.Adsorptionexperimentsonsurfaceanddeephorizonsofacidic brown(Vosges,France)andredloesssoils(Hunan,China)wereperformedat25◦Casafunctionofthe
pH(3.5–8)andalargerangeofmetalconcentrationsinsolution(10−9–10−4moll−1).Westudiedthe
adsorptionkineticsusingaCd2+-selectiveelectrodeanddesorptionexperimentsasafunctionofthe
solid/solutionratioandpH.AtaconstantsolutionpH,allsamplesexhibitedsimilarmaximaladsorption capacities(4.0±0.5mol/gCdand20±2mol/gPb).Aconstantslopeofadsorbed–dissolved concentra-tiondependencewasvalidover5ordersofmagnitudeofmetalconcentrations.UniversalLangmuirand FreundlichequationsandtheSCMformalismdescribedtheadsorptionisothermsandthepH-dependent adsorptionedgeoververybroadrangesofmetalconcentrations,indicatingnohigh-orlow-affinitysites formetalbindingatthesoilsurfaceundertheseexperimentalconditions.AtpH5,CdandPbdidnot compete,inaccordancewiththeSCM.Themetaladsorptionabilityexceededthevalueforsoilprotection bytwoordersofmagnitude,butonlycriticalloadguaranteessoilprotectionsincemetaltoxicitydepends onmetalavailability.
1. Introduction
Sorption onto soil particles is knownto control the migra-tion of metals through soils, plants and river systems and to determinemetalavailability and toxicity tothe componentsof trophicchains,includinganimalsandhumans.Consequently,the adsorptionofimportantpollutantssuchascadmiumandleadonto minerals[1–8]andorganicmatter[9–11]surfacesandthe inter-actionsbetweenheavymetalsandsoilcolloids[12–17]havebeen extensivelystudiedwithaparticularlyfocusontherigorous char-acterizationofmetaladsorptionontonaturalsorbentssuchassoils [18–34]andsediments[35].Atthebeginningofstudiesofmetal adsorptionontomineralsubstratesthetypicalapproach wasto investigatethepH-dependentadsorptionedgeataconstant con-centration of theadsorbent and a rather high concentrationof dissolved metals, in the order of mmoll−1. Many investigators
∗ Correspondingauthorat:UniversitédeToulouse;INPT,UPS;EcoLab (Labora-toireEcologiefonctionnelleetEnvironnement);ENSAT,Avenuedel‘Agrobiopôle, 31326Castanet-Tolosan,France.Tel.:+33534323949;fax:+33534323955.
E-mailaddress:anne.probst@ensat.fr(A.Probst).
proposedthatthereismorethanonetypeofsiteonthesurface of thesubstrate: dependingonthemetaltosurface concentra-tionratio,alowabundanceofsiteswithahighaffinityformetals canbemore importantthan anabundanceoflow-affinitysites [3,36].Asaresult,theadsorptionconstantitselfcandependon the metalconcentration in solution.For example, theconstant measuredata high(mmolar)metalconcentrationmightnotbe applicabletonaturalsettingsofthenanomolartomicromolar con-centrationrange. Incontrasttothelargeeffortsthathave been undertakentotestthishypothesisfornaturalorganicmatter(OM) or mineralsurfaces[37–40]for natural multicomponent adsor-bentssuchassoils,thisissuehasrarelybeenaddressedbyrigorous experimentalapproachesundercontrolledlaboratoryconditions [41].
Moststudiesofmetaladsorptiononvariousmineralsurfaces wererestrictedtomeasuringtheadsorptionedgeinaverynarrow rangeofmetalconcentrationsandwereaimedatpredicting sur-faceadsorptionconstantsandthepercentageofadsorbedmetalas afunctionofpHundergivenconditions(i.e.,[21,42–45]).However, thesestudiesdidnotalwaysallowquantificationofthemaximal numberofavailablesurfacesites,acrucialparameterforthe envi-ronmentalmodellingofmetalmigration.Forthis,experimentson
boththepH-dependent adsorptionedgeand the“Langmuirian” adsorptionataconstantpHshouldbeconducted.
Thisstudyaimedtoprovidenewinsightsintotoxicmetal inter-actionswithsoilsbyperformingathoroughinvestigationofCd andPbadsorptionontovarioustypesofacidicsoilsinaverywide rangeofmetalconcentrations.Indeed,CdandPbareamongthe mostcommontoxicmetalsinsoilsintheworldandtheirsoil trans-portandbioavailabilitystronglydependonsoilparameters(pHand organicmattercontent,amongothers)(cf.[46,47]).
Acidsoilsaremostvulnerabletoheavymetalloadsduetotheir rather low adsorptioncapacities at acidicor circumneutral soil solutionpHlevels.Aspartofageneralquantitativeassessmentof acidsoilinteractionswithmetaltoxicantsallovertheworld,this studycomparedmetaladsorptionparametersintwo representa-tiveEuropeanandAsianacidsoils.Thespecificobjectivesofthis workwere:(1)tomeasureandcomparetheadsorptionparameters ofCdandPbonsurfaceanddeepsoilhorizonsasafunctionofpH; and(2)toassesstheadsorptionyieldinthewidestpossible exper-imentalrangeofmetalconcentrationsinsolution(severalorders ofmagnitude).Theresultsshouldallowabetterquantificationof theCdandPb-adsorptioncapacityofthesesoilsandshouldhelpto constrainthemodellingofcriticalloads[46,47]inEuropeandAsia (whereincreasinglevelsofmetaldepositionarebeingfound)and, moregenerally,toimproveourunderstandingofsoilprotection regardingacidandmetalpollutionaroundtheworld.
2. Materialsandmethods
2.1. Soilsamples
Theexperimentswereperformedonsoilsthatoriginatedfrom FranceandChinainordertorepresentavarietyofacidsoilsfrom regions that receivea wide range ofacidic atmospheric inputs [27,47–49].Aredacidicsoil(labelledCS,AlliticUdicFerrisols,FAO system,ChangshaRegion,NorthofHunanProvince,China) devel-opedonQuaternaryredclayand anacidicbrownsoil(labelled PP,DystricCambisol,FAOsystem)fromtheStrengbachcatchment (Vosges,France),developedongranite,wereconsidered.Samples fromdifferentdepthswerecollectedusingaplasticshovel,sieved througha2mmmesh,driedat105◦Candkeptatambient temper-atureinthedark.Thechemicalandphysicalcompositions,specific surfaceareasandselectedparametersofthesoilsstudiedare pro-videdinTable1.Twomineralhorizonsfromeachsoil:thesurface (CS1:0–30cmandPPAB:5–30cm)andbottom(CS4:90–120cm andPPC:±180cm),wereselectedforthisstudy.
ThePPsoilhadasandyloamtexture;thepHofthewaterextract variedfrompH4.1onthesurfacetopH4.6atdepth.Themain mineralsinthecompositionofthissoilwerequartzandmicawith plagioclaseandtracesofapatite[50].TheOMcontentdecreased from2.6to0.1%fromthesurfacetothedeephorizon,respectively. Illiteandsmectitewerethedominantclaymineralswithinthissoil type.Thespecificsurfaceareas(SSA)ofthesurfaceanddeep hori-zonsofthissoilwere6.2and4.8m2g−1,respectively,asmeasured
bytheB.E.T.(Brunauer–Emmett–Teller)N2multi-pointadsorption
technique.Thealuminiumcontent(mainly asAloxides)ranged between15and20%.TheCSsoilhadaclaytexture(∼60%offine particles<0.01mm),withlowaluminiumcontent.ThepHwaterwas
aroundpH5andtheOMcontentwasclosetothatofthePPsoil. Theclaymineralogyofthesoilswasstudiedusingsmall-angle X-raydiffractionoforientedsamplespreparedbydepositingaclay suspensiononaglassslideandsubjectingittoconventional treat-ment(heatingto500◦C, ethyleneglycolexpansion).Thespectra wererecordedusingadiffractometerwithaCuKaanti-cathode. Theclaymineralsillite andchlorite, andinterlayeredI–C, were thedominantconstituentsintheCSsoil.TheSSAsrangedfrom
14.8m2g−1to19.7m2g−1fromthesurfacetodepth,respectively.
TherangeofSSAsforsoils<2mminthisstudy(5–20m2g−1)isin
goodagreementwithdataintheliteraturefordifferentsoilsaround theworld[51].
2.2. Adsorptionexperiments
Batchadsorptionexperimentswereperformedinpre-cleaned 30mlpolypropylenevialsat25±0.5◦C in0.01MNaNO3 witha
very wide range of metalconcentrations, from2–4nmoll−1 to 0.1–0.2mmoll−1. Theconcentrationsof Cdand Pbrangedfrom 10−9 to10−4M.Alladsorptionexperimentswereperformedata constantsoil/solutionratioequalto8gl−1.Desorptionexperiments wereperformedwitheachtypeofsoilinMilliQwatersuspensions atconcentrationsof4,8,20and40gl−1andpH5.0±0.5.The expo-suretimevariedfrom1to2weekstoensuretheequilibriumof metaldistributionbetweenthesoilandthesolutions,asevidenced frompreliminaryadsorptionexperimentswhichshowedthat1–2 dayswassufficienttoachievestable(±10%)metalconcentrations insolutionin contactwiththesoils.Aftertheadsorption equi-libriumwasachievedthepHofthesuspensionsweremeasured andthesampleswerecentrifugedfor15minat2000×gand fil-teredthroughapre-washed0.22mNylonfilter.Thefiltratewas acidifiedwithbidistilledHNO3andstoredat4◦Cbeforeanalysis.
Foralloftheexperimentsdescribedbelow,CO2-freeMilliQwater
saturatedwithhigh-purityN2gaswasused.
2.2.1. AdsorptionofCdandPbataconstantpH
All of the experiments were performed at a constant soil/solutionratioequalto8gl−1in0.01MNaNO3asabackground
electrolytesolution.TheinitialCdand Pbconcentrationsvaried from4nMto100M(Cd)andfrom2.4nMto240M(Pb).The amountofmetalsaddedalwaysexceededtheamountthatcould bepotentiallydesorbedfromthesoilsurface(seeSection2.4)by aminimumfactorof5.Thenecessarycorrectionsforthedesorbed metalswerealwaysmade,especiallyatlowconcentrations.The sampleswerecontinuouslyshakenat25±0.5◦Cin thedarkfor 1–2weekstoensureadsorptionequilibriumwasachieved.Sucha longexposuretimeisalsoconsistentwiththeratherlong(daysto weeks)residencetimeofinterstitialsoilsolutions.Blanksamples werepreparedinMilliQwateror0.01MNaO3solutionwiththe
samemetalconcentrationsbutwithouttheadditionofsoilinorder toaccountfortheadsorptionofmetalsontocontainerwallsunder typicalexperimentalconditions.Correctionsfortheblank adsorp-tionrangedfrom10%oftheadsorbedamountofmetalatanMol concentrationrangeto,typically,<1%attheMolconcentration range.
2.2.2. EffectofpHonadsorption
InordertostudytheeffectofpHonadsorptionequilibria,the soilswereplacedincontactwith0.01MNaNO3,with8gsoill−1and
aninitialCdorPbconcentrationof0.9or0.48M,respectively. ThepHwasvariedbyaddingHNO3 orNaOHtocovertherange
betweenpH2and8.Alladsorptionexperimentswereperformed insolutionsundersaturatedwithrespecttometalhydroxidesand carbonates,asverifiedbyMINTEQA2calculations[53]oftheinitial solutioncompositions.
2.2.3. AdsorptioncompetitionbetweenCdandPb
Competitiveadsorptionexperimentswereperformedwitheach type of soil at pH∼ 5 in 0.01M NaNO3, with 8gsoill−1 and
threetypesofinitialsolutions:0.44MCd+0.12MPb,0.22M Cd+0.36M Pb, and 0.66M Cd+0.06M Pb. After an expo-sureperiod of 48hat 25◦C in the dark,thepHwas measured andthesolutionwascentrifuged(15min,speed2500×g),filtered
Table1
Propertiesofthetestedsoils.
Property PPAB PPC CSS1 CSS4
Samplingsite Strengbachcatchment– Vosges,France
Strengbachcatchment– Vosges,France
Changsha,Hunan,China ChangshaHunan,China
Soiltype Acidbrownsoil(Dystic Cambisol)
Acidbrownsoil(Dystric Cambisol)
AlliticUdicFerrisols AlliticUdicFerrisols
Bedrock Granite Granite Quaternaryredclay Quaternaryredclay
Depth(cm) 5–30 ±180 0–30 90–120
Density(gcm−3) 1.60 2.0 1.15 1.33
Specificsurfacearea(m2g−1) 6.2 4.8 14.8 19.7
pH(H2O) 4.10 4.60 5.0 5.2 pH(KCl) 3.20 3.90 4.0 4.1 CEC(cmolkg−1) 9.10 5.10 3.6 3.3 BS(%) 4.90 7.10 14.40 12.47 Corg(gkg−1) 13.10 7.70 13.23 4.74 Organicmatter(gkg−1) 22.5 13.2 22.8 8.2
Contentofmajormetal iron(gkg−1)
Al2O3 15.60 18.20 2.25 2.14
Fe2O3 2.10 2.20 3.44 3.63
Contentofsoilparticles (mm,%) 2–0.20 60.70 62.10 7.07 10.18 0.20–0.05 7.40 8.10 10.44 8.2 0.05–0.02 1.60 7.40 6.27 8.16 0.02–0.002 12.50 8.10 43.86 40.81 <0.002 17.80 15.90 32.37 32.65 Mainconstituent mineralsa(%) Quartz 35–8 26–7 x x Orthose 15–1 10–1 x x Muscovite 14–0.4 33–2 x x Plagioclase 5–3 11–4 x x Apatite 0.06–0.05 0.21–0.09 x x Kaolinite 1–3 1–3 x x Illite 22–64 9–57 ++ +++ Vermiculite – – – – Chlorite – – +++ +++ Smectite 4–14 4–15 – – InterlayeredI–V ++ + – – InterlayeredC–S – – ++ +
Totalcontentoftrace metals(mgkg−1) Cd 0.03 0.03 0.18 0.10 Cu 2.0 0.50 23.47 27.34 Pb 56.6 9.9 24.1 22.17 Zn 25.0 24.0 99.19 75.32 Cr 9.0 19.0 124.71 128.12
aFortheVosgessamples(PPABandPPC):afterFichteretal.[50],thefirstvalueisbasedonweightedmineralogy,thesecondoneonmineralsurfacearea.Interlayered I–V:InterlayeredclayIllite–Vermiculite;InterlayeredC–S:InterlayeredclayChlorite–Smectite.+++:abundant;++,mediumabundant;+detected;–:notdetected;x:not measured.
(0.22macetatecellulosefilter),acidified(ultrapureHNO3)and
storedat4◦Cbeforeanalysis.
2.3. Adsorptionkinetics
Theadsorption kineticswerestudied in 0.01MNaNO3 with
8gl−1 of solid concentrationand aninitialCd concentrationof 88M.InordertomonitorCdactivity,aCd2+-selectiveelectrode
wasused.Solid-contactCd-sulphideelectrodes(NIKOAnalit®,
Rus-sia) coupled with an Ag/AgCl reference electrode in 3.5M KCl connected withthesolution viaa 1M NaNO3 salt bridgewere
calibratedat3<pH<9and4<pH<7instandardCd(NO3)2
solu-tions.Potentiometricmeasurementswereperformedat25±0.5◦C inaworkingconcentrationrange of10−6–10−2Mwitha detec-tionlimitof10−7Mandanuncertainty oflevel5%;aNernstian slope of29.1±0.5mV/pCd was obtained. Priortoand afterthe measurementstheelectrodewascalibratedinacellsupernatant solutionwiththesameconcentrationofbackgroundelectrolytes astheexperimentalsolution.Measurementswerealsoperformed in10mgl−1Cdsolutionin0.01MNaNO3withoutaddedsoil.The
electrode potentialwasrecordedevery10sand converted toa freeCd2+concentrationusingastandardcalibrationprocedure.The
typicalresponsetimeofthiselectrodetovariationsintheCd con-centrationwas2s(see[52]fordetails).Theseexperimentswere typicallyrunfor30min,untilasteady-stateCd2+(aq)concentration
wasachieved.
AlthoughtheadsorptionkineticsofPb2+werenotstudied,a
comparisonoftheadsorptionconstantvaluesderivedfortheCd andPbadsorptionexperimentsonorganicandmineralsurfaces, performedatthesamepH,showedthattheadsorptionkinetic con-stantofPbwasabout20–25timeshigherthanthecorresponding valueforCd,aratiothatwasalmostthesameasthatbetweenthe rateconstantsofwaterlossfromaqueousPbandCdions(i.e.,[52]). Assuch,theadsorptionequilibrium ofPb2+ shouldbeachieved
significantlyfastercomparedtothatofCd2+.
2.4. Desorption
Metaldesorptionexperimentsweredesigned tomeasurethe amountof metal releasedfrom theoriginal soilsasa resultof mineraldissolutionorsurfacedesorptionphenomena.Desorption experimentswereperformedwitheachtypeofsoil,insuspensions ofMilliQwateratasoilconcentrationof4,8,20and40gl−1at pH=5.0±0.5.ThereasonwhyMilliQwaterwasusedinsteadofan electrolytesolutionwastosimulatetheinteractionbetweensoil andpenetratinglow-levelmineralizedrainwater,andtoavoidthe releaseofexchangedmetalcomplexes.Anotherseriesof experi-mentswasperformedinasoilsuspensionof20gl−1,0.01MNaCl andpH4–8,asadjustedbytheadditionofHClandNaOH.Inallof theadsorptionexperimentsdescribedbelowtheamountof des-orbedCdandPbateachgivenpHwassubtractedfromthemetal concentrationmeasured.
2.5. Analyses
ThepHwasmeasured,withanaccuracyof±0.01pHunits,using a combined Schott–Geräte electrode calibrated against N.I.S.T. (NationalInstituteofStandardsandTechnology)buffersolutions (pH=4.00and6.86at25◦C).Dissolvedorganiccarbon(DOC)was analysedusinga TotalOrganic CarbonAnalyzer(Shimadzu TOC 5000,Paris,France, uncertainty betterthan 3% [47]).Cadmium andleadweremeasuredbyInductivelyCoupledPlasmaMass Spec-trometry(ICP-MS7500ce, AgilentTechnologiesSantaClara,CA, USA,uncertaintyof 2%,detectionlimitsof0.001gl−1).Indium andrheniumwereusedasinternalstandards.Theinternational geostandardSLRS-4 wasused to check thevalidity and repro-ducibilityoftheanalyses.Themainsoilpropertieswereanalysed attheInstitutNationaldelaRechercheAgronomique(INRA, Lab-oratoryofSoilScience,Arras,France),followingstandardmethods (http://www.arras.inra.fr/).
2.6. Adsorptionmodels
Thepartitioncoefficient(Kd,lg−1)isdefinedastheratioofthe
concentrationofmetalinthesoilandintheequilibriumsolution, respectively:
Kd=
[Me]soil
[Me]solution (1)
Theadsorptionreactiononasolidsurface,includingthe inter-actionofoneionononesurfacesite,canbewrittenas:
S−+Me2+= S–Me+ (2)
where S−isthesurfacesite,Me2+isCd2+orPb2+ionsinsolution
and S–Me+istheadsorbedcomplexes.Theconventionalstability
constantofthisreaction(KL)isdefinedas:
KL= {
S–Me+}
{ S−}·[Me2+] (3)
where{}isthesurfaceconcentrationand[]standsfortheaqueous metalconcentration.Theconcentrationofadsorbedmetal(ads)is
representedbytheLangmuirequation: ads={ S–Me+}= KL·max·
[Me2+] 1+KL·[Me2+]
(4) wheremaxisthemaximaladsorbedconcentrationcorresponding
totheoccupationofallsurfacesitesavailableforagivenmetalata givenpH.Anotherwaytodescribetheadsorptionphenomenonis viatheFreundlichequation:
S−+(n)Me2+= S–Me+ (5)
Thestabilityconstantisdefinedas: KF={ S–Me
+}
[Me2+]n (6)
andtheadsorbedmetalconcentrationcanbecalculated bythe Freundlichequation(5):
ads=KF·[Me2+] n
(7) whereKFistheFreundlichadsorptionconstantandncorresponds
tothenumberofmetalionsadsorbedontoonesurfacesite.The values of n inEq. (7) usually vary between0.9 and 1.4. When n=1,KF=ads/[Me2+]=Kd,the so-called distributioncoefficient,
whichistheslopeoftheisothermattheorigin[43].Eqs.(4)and (7) areonly validfor a given pHand ionicstrength. Neverthe-less,theLangmuirandFreundlichmodelsarethemostextensively usedformodellingsolid(soil,sediments)sorptionaffinities.The useofboth Langmuir andFreundlich equations waswarranted
0,0 0,5 1,0 1,5 2,0 2,5 3,0 0 500 1000 1500 2000 Elapsed time, s [C d ]ad s , µ m o l/g
Fig.1. AdsorbedCdconcentrationasafunctionoftimefortheacidloesssoil(CS-S1). Experimentalconditions:0.01MNaNO3,8gsoill−1,[Cd]0=10mgl−1,pH=7.0±0.2. because:(1)mostexperimentsinthisstudywereperformedata constantpHandionicstrength;(2)bothequationswerewidely usedto describesoil adsorptionphenomena, and(3) they cov-eredthefullrangeofthemetalconcentrationsinvestigatedand alloweda straightforward comparison oftheadsorption capac-itybetweenthedifferentsoils.Incontrast,a differentsetofKL
andKFwasnecessaryforeachsolutioncompositionincaseofpH
variationsinsolution.Thesedifficultiesareovercomeinsurface complexationmodels(SCM),whichaccountforchemicalreactions thatoccuratthesolid/solutioninterfaceintermsofrealsurface species,andpredictthepH-dependenceofadsorptioninawide rangeofsolutioncompositionsusingauniquesetofsurface stabil-ityconstants[36].Typically,arigorousSCMapproachrequiresan explicitidentificationofproton/hydroxylandmetal-activesurface sites(concentrationandstoichiometryofsurfacebindinggroups suchas Si–OH, Al–OH, FeOH, R–COOH,different exchange-ablesites).Thethoroughcharacterizationofthesesurfacemoieties impliesamultidisciplinaryapproachcomprisingsurfacetitration and electrophoresis supportedbystate-of-the-art spectroscopic techniquessuchasFT-IR(FourierTransformed-InfraRed)andXPS (X-RayPhotoelectronSpectroscopy),forexample(see[53]asan exampleofmultisitesurfacecharacterization).Becauseofthe com-plexityofmineralandorganicsurfacesinsoils,weusedasimplified SCMapproach implyingmetal adsorptiononto onesurface site accordingtothereaction:
S–H◦+Me2+= S–Me++H+ (8)
Theintrinsicsurfacestabilityconstant(Kads◦)isdefinedas:
Kads◦=
aH+·{ S–Me+} aMe2+·{ S–H◦} ·exp − 0 RT (9) whereaiistheactivityofithspeciesinsolution,{ S-i}isthesur-faceconcentrationofithspecies,0istheelectricpotentialatthe
surface,RisthegasconstantandTisthetemperature.Thesurface andsolutionequilibriawerecalculatedusingtheMINTEQA2 pro-gram[54]withintheconceptofconstantcapacitanceoftheelectric doublelayer.
3. Results
3.1. Kinetics
Theadsorptionkineticswereassessedin ordertodetermine theminimalcontacttime necessarytoachieveadsorption equi-librium, corresponding to the constant concentration of metal adsorbedduringthefollowingsorption/desorptionexperiments. TheadsorbedCdconcentrationisshowninFig.1asafunctionof
-11 -10,5 -10 -9,5 -9 -8,5 -8 -7,5 -7 -6,5 -6 40 30 20 10 0 [Soil], g/L log [Pb], M PPAB PPC CS S1 CS S4
A
-11,5 -11 -10,5 -10 -9,5 -9 40 30 20 10 0 [Soil], g/L log [Cd], M PPAB PPC CS S1 CS S4B
Fig.2.Concentrationsoflead(A)andcadmium(B)inasolutionsaturatedwith non-contaminatedsoilsasafunctionofsoilconcentrationinsolution.Experimental conditions:pH=4.43±0.22(PPAB),5.48±0.3(PPC),5.58±0.23(CSS1),5.47±0.18 (CSS4);0.01MNaCl,1weekexposureat25◦C.Thesymbolsrepresentthe measure-mentsandtheconnectinglinesareforguidingpurposes.
time.Similartomostadsorptionprocesses(i.e.,[7]andreferences therein),theconstantconcentrationofmetalinsolutionincontact withthesuspensionwasachievedwithin10–20minanddidnot changeforatleast1week(datanotshown).Asaresult,aperiod of24hwasacceptedasbeingsufficienttoachievetheadsorption equilibrium.
3.2. Metaldesorptionfromnon-contaminatedsoils
Theseexperiments,whichaimedtodeterminetheamountof metalthatcanbereleasedfromsoildueto(i)thedissolutionof somemineral/organicphasesbearingthismetaland(ii) “desorp-tion”fromthesurfaceofsoilparticles,wereperformedwithvarious soiltosolutionmassratios(Fig.2).ForthesurfacehorizonofthePP soilwiththehighestPbandCdcontents,theamountofdesorbed metalsincreasedwiththemassofthesolid.Fortheother sam-ples,thereleaseofPbdidnotchangewithanincreaseinsoilmass (Fig.2A)whereasanincreaseinCdreleasewithsoilmass(Fig.2B) wasfound.TheseresultsenabledthelowestlimitsoftheCdandPb concentrationsthatcouldbeusedfortheadsorptionexperiments tobedetermined.
TheeffectofpHonmetaldesorptionfromsoilisshowninFig.3A andB.TheconcentrationofCdandPbinsolutiondecreasedwith thepH,similartothesituationfoundfortheadsorptionequilibria (seebelow).Theweakincreaseinthedesorbedlead concentra-tionatpH>6(Fig.3A)canbepartlyexplainedbyitsmobilization byorganicmatter,whichismoresolubleatthesepHlevels[47]. Indeed,theconcentration ofDOC inthe experimentswith sur-facehorizonsinbothsoilsincreasedfrom2–3mgl−1at4≤pH≤6 to8–10mgl−1 atpH∼7(datanotshown).Itwasalsopreviously
-11 -10 -9 -8 -7 7 6 5 4 pH log [Pb], M PPC CS S1 CS S4
A
-11 -10 -9 -8 7 6 5 4 pH log [Cd], M PPC CS S1 CS S4B
Fig. 3.Concentrations of desorbed lead (A) and cadmium (B) from non-contaminatedsoils asafunctionofpH.Experimentalconditions:0.01MNaCl, 20gsoill−1,1weekexposureat25◦C.Thesymbolsrepresentthemeasurements andthelinesareforguidingpurposes.
Table2
ParametersoftheLangmuirequation(4)forthefourstudiedsoilsamples.
Soil KL(l/mol) max(mol/g)
Cadmium PPAB (0.8±0.2)×105 4.8±0.3 PPC (0.6±0.1)×105 4.0±0.2 CSS1 (2.2±0.1)×105 5.8±0.2 CSS4 (1.0±0.1)×105 4.2±0.2 Lead PPAB 4000±500 22±2 PPC 6000±100 14±2 CSS1 4500±500 25±2 CSS4 8000±500 14±1
shownthattheproportionofdissolvedPbpresentintheformof organiccomplexesinsoilsolutionsstronglyincreasesata near-neutralpHcomparedtoalowpH[55].
3.3. Effectofmetalconcentrationinsolutiononadsorptionata constantpH
Theaimoftheseexperimentswastoquantifytheamountof adsorbedmetalataconstantpHinordertoallowastraightforward comparisonoftheadsorptioncapacityofdifferentsoilsunder oth-erwiseidenticalenvironmentalconditions. Theconcentrationof adsorbedCdisplottedasafunctionofitsequilibrium concentra-tioninsolution(Fig.4)andtheparametersofEq.(4)areshownin Table2.Ofthefourtypesofsoilsamplesstudied,only the sur-facehorizon oftheChineseredsoil (CSS1sample)exhibiteda highercadmiumadsorptioncapacityandadsorptionconstant,bya factorof2to3,whereasthedeephorizonsofthetwosoils demon-stratedaloweradsorptioncapacitythanthesurfacehorizons.The
1,E-04 1,E-03 1,E-02 1,E-01 1,E+00 1,E+01 1,E+02 1,E+01 1,E+00 1,E-01 1,E-02 1,E-03 1,E-04 [Cd], µmol/L [Cd] ads , µmol/g PPAB PPC CS S1 CS S4 model Langmuir model Frendlich
Fig.4.ConcentrationsofadsorbedCdasafunctionoftheequilibratedCd con-centrationsinsolutionforthefourtypesofsolid.Experimentalconditions:25◦C, 0.01MNaNO3,1weekexposureunderdarkness.ThepHvalueswerepH4.47±0.09; 4.16±0.30;5.5±0.2and5.4±0.1forthePPAB,PPC,CSS1andCSS4soils, respec-tively.Whennotvisible,theexperimentaluncertaintiesarewithinthesizeof thesymbols.Thesolidanddashedlinescorrespondtomodelcalculationsusing theLangmuirandFreundlichequations,respectively.ParametersoftheLangmuir equation:max=4.0±0.5mol/g,KL=(1.3±0.3)×105l/mol.Parametersofthe Fre-undlichequation:KF=5×105;n=1.
maximaladsorptioncapacitywasestimatedfromfittinga Lang-muirisothermequationandrangedbetween3and5mol/gor0.2 and0.8mol/m2forthedifferentsoils,whichiscomparabletothe
valuesofgoethite,amorphoussilicaandclays[36].Weattempted totreatthedataobtainedforthewholeconcentrationrange(i.e., from10−9to10−4M)usingLangmuir(Eq.(4))andFreundlich(Eq. (7))isotherms.Weusedanon-linearleastsquaresfittingwith10% residualdisagreementbetweenthemodelandtheexperimental data.Incontrasttothetraditionallinearfittingofadsorption equa-tions,thissimultaneouslyaccountedforthewiderangeofhighand lowCdandPbconcentrationsusedinthisstudy.Thesemi-linear dependenciesobtainedonalog–logscale(Fig.4)werevalidover alargeconcentrationrangeand couldbesuccessfully modelled usingthesameequations((4)and(7))parametersforalldata col-lectedfromthefourdifferentsoilsamples:max=4.0±0.5mol/g,
KL=(1.3±0.3)×105,KF=5×105;n=1.
TheconcentrationofadsorbedPbwasplottedasafunctionofits equilibriumconcentrationinsolution(Fig.5),andtheparameters ofEq.(4)arelistedinTable2.Similartotheresultsforcadmium,the surfacehorizonofChineseredsoilexhibitedthehighestadsorption capacityfor lead,particularlyfor highcadmiumconcentrations. However,withinthelevel of experimentaluncertainty, thefull setof data could beapproximated using the same parameters oftheLangmuirandFreundlichequations:max=20±2mol/g,
KL=6000±500;KF=1.25×105;n=1.The1:1slopebetweenthe
adsorbedand aqueous Pb concentration was preserved over 5 ordersofmagnitude,suggestingthepresenceofonlyonetypeof surfacesiteforallofthesoiltypes,orthatthedifferentsiteswere notdistinguishablewithintheresolutionofourmeasurements. 3.4. EffectofpHonmetaladsorption
TheexperimentsonthepH-dependentadsorptionedgeallowed theeffectofsolutionpHonthedegreeofmetaladsorptiontobe assessedandthusanevaluationoftheimpactofamphoteric prop-ertiesof soilsurface sites onmetalbindingin a wide range of solutionpHlevels.ThepercentageofadsorbedCdisplottedasa functionofpHinFig.6.Thelevelofadsorptionincreasedwiththe pHleveltoachieveamaximumadsorptionatpH≥6foreachof thefoursamples.Theadsorptionedgewassimilarforbothsoils
1,E-04 1,E-03 1,E-02 1,E-01 1,E+00 1,E+01 1,E+02 1,E+03 1,E+02 1,E+01 1,E+00 1,E-01 1,E-02 1,E-03 [Pb], µmol/L [Pb] ads , µmol/g PPAB PPC CS-S1 CS-S4 model Langmuir model Frendlich
Fig.5.ConcentrationsofadsorbedPbasafunctionoftheequilibratedPb con-centrationsinsolutionforthefourtypesofsolid.Experimentalconditions:25◦C, 0.01MNaNO3,exposure1weekinthedarkness.ThepHvalueswerepH4.4±0.2; 5.0±0.3;5.3±0.4and5.2±0.3forthePPAB,PPC,CSS1andCSS4soils, respec-tively.Whennotvisible,theexperimentaluncertaintiesarewithinthesizeofthe symbols.Thesolidanddashedlinescorrespondtothemodelcalculationsusingthe LangmuirandFreundlichequations,respectively.ParametersoftheLangmuir equa-tion:max=20±2mol/g,KL=6000l/mol.ParametersoftheFreundlichequation: KF=1.25× 105;n=1. 0 20 40 60 80 100 6 5 4 3 2 1 % Pb adsorbed PPAB PPC CS S1 CS S4 (b) 0 20 40 60 80 100 8 7 6 5 4 3 pH % Cd adsorbed PPAB PPC CS S1 CS S4 (a) pH
Fig.6. PercentagesofadsorbedCd(a)andPb(b)asafunctionofpHforthefourtypes ofsoils.Experimentalconditions:[Cd]o=0.9M,0.01MNaNO3,[Pb]o=0.48M, 0.01MNaNO3,8gsoill−1.Thesolidlinerepresentsthesurfacecomplexation mod-elling(SCM)resultsusingaconstantcapacitanceoftheelectricdoublelayerequal to1F/m2,surfacesitesdensityof4mol/g(Cd,(a))and20mol/g(Pb,(b))forsoils withaspecificsurfaceareaof15m2g−1andsurfacestabilityconstantsof reac-tions(10)and(11)equalto−4.0and−0.1±0.6(Cd)and−4.0and0.7±0.3(Pb), respectively.TheaverageuncertaintyoftheSCMpredictionofadsorbedCdwas ±10%.
butthesurfacehorizonsexhibitedastrongeraffinityformetalas theadsorptionbegan,andachievedamaximumatlowerpH lev-els.ThedecreaseinCdadsorptionatpH>6.5forthePPABsoilwas mostlikelyrelatedtotheincreaseintheDOCconcentrationand thereleaseofmetalintheformoforganiccomplexes,asdiscussed earlierforthedesorptionexperiments(Section3.2).
WeattemptedtodescribethepH-dependenceofCdadsorption insoilswithintheframeworkofthesurfacecomplexationmodel (SCM).Theexactproportionsofvariousmineralandorganic com-posites ofthestudiedsoilsand theirmetal andprotonbinding propertiescouldnotbedetermined.Therefore,weusedthe gen-eralizedcompositeapproach,whichassumesthattheadsorptive reactivityofthesurfacecanbedescribedbysurfacecomplexation equilibriawiththestoichiometryandformationconstants deter-minedbyfittingtheexperimentaldata[56].Forsimplicity,two surfacereactionswerepostulatedtocontrolthemetaladsorption onsoils:
SO–H◦= SO−+H+, logKa=−4.0 (10)
SO–H◦+Cd2+= SO–Cd2++H+, logKCd=−0.1 (11)
Thefirstconstant(Ka)wasfixedwithintherangereportedfor
acid–baseparametersofChineseloesssoils[57]andsoilclay min-erals[1]andthesecondconstantwasallowedtovaryinorderto reproducetheexperimentaldatausingtheMINTEQA2code.The surfacesitedensitywasfixedat4mol/g(0.3–0.7mol/m2),as
measuredinourexperiments usingtheLangmuirian modelling approach(Fig.4),andtheconstantcapacitanceoftheelectric dou-blelayer(EDL)waspostulatedas1F/m2,whichisthemostcommon
valueusedinsurfacecomplexationmodelling.Anexample ofa modelcalculationoftheexperimentaldataisshowninFig.6.We usedanumericalapproachtofitthepercentageofadsorbedmetal asafunctionofpH.A≤10%differencebetweenthemeasuredand modelledpercentages of adsorbedmetalat each pHvaluewas takenasacriterionofgoodnessoffit.
TheuncertaintyinexperimentaldatadoesnotallowKaandKCd
constantstoberesolvedwithanaccuracygreaterthan±0.5log units.SimilaruncertaintylevelscanbeattributedtologKCd
vari-ationsbetweenthesurfaceandbottom horizonsofvarioussoil types.However,wewouldliketonotethesimilarityofthe pH-dependentadsorptionedgesforthefourtypesofsoilinthepresent study.ThisresultagreeswiththepreviousfindingsofMullerand Duffek[58],whofoundthatthesurface-normalizedvalueofthe adsorptionconstantforCdonmineraloxides,sedimentsand sus-pendedmattercanbeconsideredasaconstantparameterwithin ±0.4logarithmicunits.Thisfindingsuggests,inaccordancewith thesurfacecoordinationtheory[36],thattheterminaloxygensites S–O− responsiblefor metalbindingonsolidsurfacesare simi-larbetweenvarioussoilmineralsandorganicmatter.Asaresult, onecanuseauniversalsurfaceadsorptionconstanttodescribethe interactionbetweenmetalpollutantsandvarioussoils.
Thepercentage ofadsorbedPbisplottedasafunctionofpH inFig.6b.Similartocadmium,theadsorptionincreasedwithpH, achievingamaximumadsorptionatpH≥4forthefourtypesof soils.Theadsorptionedgewassimilarforthetwosoils.Theresults ofthe surfacecomplexationmodellingare presentedas a solid linecorrespondingtotheuncertaintiesattributedtothereaction (Eq.(8))stability constant(Eq.(9)).For thePbadsorption mod-ellingthesurfacesitedensityforallsoilswasfixedat20mol/g (1–3mol/m2),whichwasfivetimeshigherthanforcadmium,as
measuredusingtheadsorptionisotherms(Fig.5).Thisdifference betweenCdandPbinmaximalsurfaceadsorptiondensitiesisoften reportedfortheadsorptionofthesemetalsoncommoninorganic andorganicsorbents[31,52,59,60].Arecentstudy[61]showedthat Pbretentionon FeOHsoilsitesoccurredatalowerpHthanCd retention,suggestingthatPbadsorbstosurfacehydroxylgroupsat
Fig.7.PercentagesofadsorbedCd(A)andPb(B)forfourtypesofsoilsamples anddifferentinitialCd/Pbratios.Experimentalconditions:[Pb]o,[Cd]o=25,50and 75gl−1(0.06,0.12and0.36MforPband0.22,0.44and0.66MforCd),0.01M NaNO3,8gsoill−1.ThepHvalueswerepH4.12,4.61,4.86and4.98forPPAB,PPC, CSS1andCSS4,respectively.SeethecommentswritteninFig.6.
pHvalueswhereCdonlyinteractswithexchangesites.Inthisstudy, consistentwiththeresultsfor Cd,similar Pbadsorption edges, within±10%,canberecognizedinthefoursoilsamples,in agree-mentwithpreviousfindings[11,58].Thedisagreementbetween themeasuredandcalculatedadsorptionyieldsat4≤pH≤5could be related to Pb complexation with dissolved organic matter, which would not have been taken into account by the model calculations.
3.5. Competitionexperiments
Themetaladsorptioncompetitionexperimentsquantifiedthe adsorptionreactionsunderthemostrelevantenvironmental pollu-tionconditions,i.e.,whenbothmetalswerepresentininterstitial soilsolutions.TheresultsofsimultaneousCdandPbadsorption ontofoursoilsamplesarereportedasstackdiagrams(Fig.7AandB). Withinexperimentaluncertainty,thepercentageofeachadsorbed metaldidnotdependonthepresenceofthesecondmetaloverthe concentrationrange,whichcorrespondedtorealistic environmen-talpollution(0.1–1M).Indeed,withinthisconcentrationrange at circumneutral pH, the surface sites were undersaturated by 100–1000timeswithrespecttothemaximalamountofadsorbed metal.Thiswasfurthersupportedbysurfacecomplexation mod-ellingperformedusingthemodelparametersassessedforCdand PbinSection3.4:thepercentagesofadsorbedleadandcadmium did not vary withthe initial Cd/Pb ratio for each soil sample. Althoughhighlevelsofexperimentaluncertaintyassociatedwith thereaction(Eq.(8))stabilityconstantdonotallowthepercentage ofadsorbedmetalsateachgivenpHtobepreciselyreproduced,the generaltrendcanbedetermined.
4. Discussionandconclusions
Thefirstmainresultofthis studyis thatovertheverywide metalconcentrationrange studied,fromthenMoltothemMol scaleinsolution,theconcentrationofadsorbedCdandPbfollowed asimilarpatterninalog–logscaleforallfour,rathercontrasting, soilsamples(Figs.4and5).Thisindicatesthesimilarity,within anorderof magnitude,ofthesurfaceadsorptionconstants and thenumber ofbindingsitesover theconcentrationrange stud-ied,fromextremelypollutedenvironmentswithmMol-levelmetal loads[49]topristineinterstitialsoilsolutions(nMollevel[47,62]). Thisfindingdisagreeswiththenotionofhavingtoconsider low-andhigh-affinitysitesinordertodescribeCdorPbadsorptionin thesoilsamplesstudied.Assuch,itcontradictsthecommonview ofhighsurfaceheterogeneityofnaturalsorbentsbutgreatly facili-tatesthepredictivemodellingofmetalmigrationinbothpolluted andpristinesoilsettings.
Thesurfacehorizonsofacidicbrownsoil(France)andredsoils (China)exhibitgreatcapacitiesformetaladsorption.Leadexhibits ahigheradsorptionaffinitythanCdwithrespecttothesoil sur-facesstudied,whichisconsistentwiththedifferentaffinitiesof metalstoclay[1]andsoilorganicmatter[41].Itisknownthat anthropogenicleadmigrateswithinthesoilprofileintheformof organicmatterandFeoxy(hydr)oxides[63].TheaffinityofPbto podzolsoiliscontrolledbyexchangeableclayfractionsinthe sur-facezone andbyiron oxidesthroughoutthewholeprofile[64]. However,arigorouscharacterizationofthechemicalstatusofPb, whichcanbeassociatedbothwithorganicandmineral compo-nents,wouldrequirein situspectroscopicobservations [65,66]; forexample,themaximumsurfaceadsorptioncapacitiesforthese soils(i.e.,0.3–0.7and1–3mol/m2forCdandPb,respectively)are
comparablewiththoseofthemosteffectivesorbents,suchasiron oxides,claysandalgalbiomass[5,6,58].Themaxvaluesobtained
arewithintherangereportedformajorIndiansoils(0.3–0.6mol/g forPb,0.5–0.9mol/gforCd[18])buttheyaremuchlowerthan therange reportedforhumicacidsoils(∼770mol/g)[15] and organic-richtopsoils(3–10mol/g)[31].Apparently,thepresence ofsoilorganicmatterintheformofhumusislikelytoincrease theadsorptionofmetalsatpH∼5[19,32]:theformationofternary surfacecomplexesbetweenmetalions,oxidesurfacecentresand organicligands[67]canbeevoked.Moreover,ahumic-acid coat-ingonmontmorillonitemightsignificantlyenhancethesorptionof cadmiumandlead[4,68],andthepresenceofsolubleorganic mat-teramplifiestheinteractionofcopperandzincwithamorphous allophanes[69].
Theadsorptionresultsindicatethat thesoilsamplesstudied werenotstrongly polluted.Comparedtotheamountof metals thatcanberetainedinthesolidphase(i.e.,∼450mgkg−1 forCd and∼3000mgkg−1 forPb, Figs.4and5),theactualmetal con-centrationswerelow(0.03–0.2mgkg−1forCdand10–60mgkg−1 for Pb). This is also supported by results of the simultaneous adsorptionofCdandPbatenvironmentallyrelevantconcentration levels:nocompetitionwasobservedbetweenthesetwo metals (Fig. 7)as thesurface sitesremained farfrombeing saturated. Thesesoilsarethuscapableofadsorbing100–1000timesmore metalsthan levelsauthorizedbytheEnvironmental Policy.The actualmetalconcentrationsdidnotexceedthemaximum accept-ablelimitof2mgkg−1forCdor70mgkg−1forPb,asreportedby [70]forsoilorganismprotection.However,themaximal adsorp-tion capacities were very different from the critical limits for soil damage. Although the soil samples studied had very high total metal adsorption capacities, the toxicity of metals to soil organismsismainlyrelatedtometalavailabilityandspeciationin soils[47,48,71].Therefore,increasingknowledgeaboutthemetal sorptioncapacity ofsoilsisofgreatimportancefor understand-ingaccumulatedanthropogenicinputsbecausetheseconstitutea
futurechallengeforecosystemprotectionbasedonthecriticalload concept[72].
Acknowledgements
Wearegratefultotwoanonymousreviewersfortheir construc-tivecommentsthatenabledustoimproveourpresentationand discussion.Thisworkispartoftheproject“Effectsofacid depo-sitiononchemicalprocessesofsoilheavymetals”(PRAE00-04), whichisacooperativeprojectbetweenChineseandFrench scien-tiststhatispartlyfinanciallysupportedbytheChineseMinistry ofScienceandTechnology,theChineseMinistryofEducation,the FrenchMinistryofResearchand theFrenchMinistryofForeign Affairs.
References
[1] F.Barbier,G.Duc,M.Petit-Ramel,Adsorptionofleadandcadmiumionsfrom aqueoussolution tothemontmorillonite/water interface,ColloidsSurf.A: Physicochem.Eng.Aspects166(2000)153–159.
[2]D.Buerge-Weirch,R.Hari,H.Xue,Ph.Behra,L.Sigg,AdsorptionofCu,Cd,and Niongoethiteinthepresenceofnaturalgroundwaterligands,Environ.Sci. Technol.36(2002)328–336.
[3]D.A.Dzombak,F.M.M.Morel,Developmentofadatabaseformodeling adsorp-tionofinorganicsonironandaluminum-oxides,Environ.Prog.6(2)(1987) 133–137.
[4]A.Liu,R.D.Gonzalez,Adsorption/desorptioninasystemconsistingofhumic acid, heavymetals,andclayminerals,J. ColloidInterfaceSci. 218(1999) 225–232.
[5] Y.M.Nelson,L.W.Lion,M.L.Shuler,W.C.Ghiorse,Effectofoxideformation mechanismsonleadadsorptionbybiogenicmanganese(hydr)oxides,iron (hydr)oxides,andtheirmixtures,Environ.Sci.Technol.36(2002)421–425. [6]S.E.O’Reilly,M.F.Hochella,Leadsorptionefficienciesofnaturalandsynthetic
MnandFe-oxides,Geochim.Cosmochim.Acta67(2003)4471–4487. [7] A.C.Scheinost,S.Abend,K.I.Pandya,D.L.Sparks,KineticcontrolsonCuandPb
sorptionbyferrihydrite,Environ.Sci.Technol.35(2001)1090–1096. [8]K.G.Tiller,J.Gerth,G.Brummer,TherelativeaffinitiesofCd,NiandZnfor
differentsoilclayfractionsandgoethite,Geoderma34(1984)17–35. [9] R.H.Crist,J.R.Martin,D.R.Crist,Heavymetaluptakebylignin:comparisonof
bioticligandmodelswithanion-exchangeprocess,Environ.Sci.Technol.36 (2002)1485–1490.
[10] Y.S.Ho,G.McKay,Thesorptionoflead(II)ionsonpeat,WaterRes.33(1999) 578–584.
[11]N.Yee,J.Fein,Cdadsorptionontobacterialsurfaces:auniversaladsorption edge?Geochim.Cosmochim.Acta65(2001)2037–2042.
[12] P.Benes,J.Mizera,Radiotraceranalysisofcadmiumspeciationinsoilsolutions usingelectrophoresis,dialysis,centrifugationandultrafiltration,Biol.Trace Elem.Res.71–72(1998)91–101.
[13]D.Berggren,Speciationofaluminum,cadmium,copper,andleadinhumussoil solutions– acomparisonoftheionexchangecolumnprocedureand equilib-riumdialysis,Int.Environ.Anal.Chem.35(1989)1–15.
[14]B.Cances,M.Ponthieu,M.Castrec-Rouelle,E.Aubry,M.F.Benedetti,Metal ionsspeciationinasoilanditssolution:experimentaldataandmodelresults, Geoderma113(2003)341–355.
[15]M.Cruz-Guzman,R. Celis,M.C.Hermosin,P.Leone,M.Nègre,J.Cornejo, Sorption–desorptionoflead(II)andmercury(II)bymodelassociationsofsoil colloids,SoilSci.Soc.Am.J.67(2003)1378–1387.
[16]O.S.Pokrovsky,B.Dupré,J.Schott,Fe–Al–organiccolloidscontrolthespeciation oftraceelementsinpeatsoilsolutions:resultsofultrafiltrationanddialysis, Aquat.Geochem.11(2005)241–278.
[17]S.Sauvé,M.B.McBride,W.H.Hendershot,Speciationofleadincontaminated soils,Environ.Pollut.98(1997)149–155.
[18]R.Adhikari,M.V.Singh,Sorptioncharacteristicsofleadandcadmiuminsome soilsofIndia,Geoderma114(2003)81–92.
[19]J.O.Agbenin,L.A.Olojo,CompetitiveadsorptionofcopperandzincbyaBt hori-zonofasavannaAlfisolasaffectedbypHandselectiveremovalofhydrous oxidesandorganicmatter,Geoderma119(2004)85–95.
[20]C.Appel,L.Ma,Concentration,pH,andsurfacechargeeffectsoncadmiumand leadsorptioninthreetropicalsoils,J.Environ.Qual.31(2002)581–589. [21]H.B.Bradl,Adsorptionofheavymetalionsonsoilsandsoilsconstituents,J.
ColloidInterfaceSci.277(2004)1–18.
[22]E.F.Covelo,N.Alvarez,M.L.Andrade,F.A.Vega,P.Marcet,Znadsorptionby differentfractionsofGaliciansoils,J.ColloidInterfaceSci.280(2004)343–349. [23]H.A.Elliott,M.R.Liberati,C.P.Huang,Competitiveadsorptionofheavymetals
bysoils,J.Environ.Qual.15(1986)214–219.
[24]M.P.F.Fontes,P.C.Gomes,Simultaneouscompetitiveadsorptionofheavy met-alsbythemineralmatrixoftropicalsoils,Appl.Geochem.18(2003)795–804. [25] P.S.Hooda,B.J.Alloway,Cadmiumandleadsorptionbehaviorofselected
EnglishandIndiansoils,Geoderma84(1998)121–134.
[26]S.-Z.Lee,L.Chang,H.-H.Vang,C.-M.Chen,M.-C.Liu,Adsorptioncharacteristics ofleadontosoils,J.Hazard.Mater.A63(1998)37–49.
[27]B.Liao,Z.Guo,A.Probst,J.L.Probst,Soilheavymetalcontaminationandacid deposition:experimentalapproachontwoforestsoilsinHunan,Southern China,Geoderma127(2005)91–103.
[28]G.T.Pardo,M.E.Guadalix,Cadmiumsorptionbytwoacidsoilsasaffectedby cleaningandcultivations,Commun.SoilSci.PlantAnal.26(1995)289–302. [29]A.A.Ponizovskii,E.V.Mironenko,Mechanismsoflead(II)sorption insoils,
EurasianSoilSci.34(2001)371–381.
[30] S.Sauvé,S.Manna,M.C.Turmel,A.G.Roy,F.Courchesne,Solid-solution parti-tioningofCd,Cu,Ni,Pb,andZnintheorganichorizonsofaforestsoil,Environ. Sci.Technol.37(2003)5191–5196.
[31]S.Serrano,F.Garrido,C.G.Campbell,M.T.Garcia-Gonzalez,Competitive sorp-tionofcadmiumandleadinacidsoilsofCentralSpain,Geoderma124(2005) 91–104.
[32] P.Sipos,T.Németh,I.Mohai,I.Dodony,Effectofsoilcompositiononadsorption ofleadasreflectedbyastudyonanaturalforestsoilprofile,Geoderma124 (2005)363–374.
[33]A.Voegelin,R.Kretzschmar,Modellingsorptionandmobilityofcadmiumand zincinsoilswithscaledexchangecoefficients,Eur.J.SoilSci.54(2003)387–400. [34]X.Dong,C.Li,J.Li,J.Wang,S.Liu,B.Ye,Anovelapproachforsoil contamina-tionassessmentfromheavymetalpollution:alinkagebetweendischargeand adsorption,J.Hazard.Mater.175(2010)1022–1030.
[35]G.Fu,H.E.Allen,Cadmiumadsorptionbyoxicsediment,WaterRes.26(1992) 225–233.
[36]W.Stumm,ChemistryoftheSolid–WaterInterface,Wiley,1992.
[37]M.Kosmulski,Adsorptionoftrivalentcationsonsilica,J.ColloidInterfaceSci. 195(1997)395–403.
[38]M.Kosmulski,Theeffectoftheionicstrengthontheadsorptionisothermsof nickelonsilica,J.ColloidInterfaceSci.190(1997)212–223.
[39]C.Tiffreau,J.Lutzenkirchen,Ph.Behra,Modelingtheadsorptionofmercury(II) on(hydr)oxides.I.Amorphousironoxideand␣-quartz,J.ColloidInterfaceSci. 172(1995)82–93.
[40] H.B.Xue,L.Sigg,ComparisonofthecomplexationofCuandCdbyhumicor fulvicacidsandbyligandsobservedinlakewaters,Aquat.Geochem.5(1999) 313–335.
[41]M.Benedetti,W.H.VanRiemsdijk,L.K.Koopal,D.G.Kinniburgh,D.C.Gooddy, C.J.Milne,Metalionbindingbynaturalorganicmatter:fromthemodeltothe field,Geochim.Cosmochim.Acta60(1996)2503–2513.
[42]Y.Wang,X.Tang,Y.Chen,L.Zhan,Z.Li,Q.Tang,Adsorptionbehaviorand mechanismofCd(II)onloesssoilfromChina,J.Hazard.Mater.172(2009) 30–37.
[43]D.L.Langmuir,EnvironmentalGeochemistry,Pergamon,1996.
[44]E.F.Covelo,F.A.Vega,M.L.Andrade,Simultaneoussorptionanddesorptionof Cd,Cr,Cu,Ni,Pb,andZninacidsoils:I.Selectivitysequences,J.Hazard.Mater. 147(2007)852–861.
[45] Z.F.Meng,Y.P.Zhang,Z.Q.Zhang,Simultaneousadsorptionofphenoland cad-miumonatmosphericsoil,J.Hazard.Mater.159(2008)492–498.
[46]W.deVries,S.Lofts,E.Tipping,M.Meili,J.E.Groenenberg,G.Schutze,Impact ofsoilspropertiesoncriticalconcentrationsofcadmium,lead,copper,zinc andmercuryinsoilandsoilsolutioninviewofecotoxicologicaleffects,Rev. Environ.Contam.Toxicol.191(2007)47–89.
[47]L.Gandois,A.Probst,C.Dumat,Modellingtracemetalextractabilityand sol-ubilityinFrenchforestsoilsbyusingsoilproperties,Eur.J.SoilSci.61(2010) 271–286.
[48]L.Hernandez,A.Probst,J.L.Probst,E.Ulrich,Heavymetaldistributioninsome Frenchforestsoils:evidenceforatmosphericcontamination,Sci.TotalEnviron. 312(1–3)(2003)195–219.
[49]B.Liao,G.Zhaohui,Q.Zeng,A.Probst,J.L.Probst,Effectsofacidrainon compet-itivereleasesofCd,Cu,andZnfromtwonaturalsoilsandtwocontaminated soilsinHunan,China,WaterAirSoilPollut.7(1–3)(2007)151–161. [50]J.Fichter,D.Dambrine,M.P.Turpault,J.Ranger,Basecationsupplyinspruce
andbeechecosystemsoftheStrengbachcatchment(Vosgesmountains,N-E France),WaterAirSoilPollut.104(1998)125–148.
[51]K.Kaiser,G.Guggenberger,Mineralsurfacesandsoilorganicmatter,Eur.J.Soil Sci.54(2003)219–236.
[52]A.Gélabert,O.S.Pokrovsky,J.Schott,A.Boudou,A.Feurtet-Mazel,Cadmium andleadinteractionwithdiatomsurfaces:acombinedthermodynamicand kineticapproach,Geochim.Cosmochim.Acta71(2007)3698–3716. [53]A.Gélabert, O.S.Pokrovsky,J.Schott,A.Boudou,A.Fertet-Mazel,E.
Miel-czarski,J.Mielczarski,O.Spalla,Studyofdiatoms/aqueoussolutioninterface. I.Acid–baseequilibria,surfacechargeandspectroscopicobservationoftwo freshwaterperyphyticandtwomarineplanktonicdiatoms,Geochim. Cos-mochim.Acta68(2004)4039–4058.
[54]J.D.Allison,D.S.Brown,K.J.Novo-Gradac,MINTEQA2/PRODEFA2,A Geochem-icalAssessmentModelforEnvironmentalSystems:Version3.0User’sManual. U.S.EPA,Athens,GA,1991.
[55] S.Sauvé,M.McBride,W.Hendershot,Soilsolutionspeciationoflead(II):effects oforganicmatterandpH,SoilSci.Soc.Am.J.62(1998)618–621.
[56] J.A.Davis,J.A.Coston,D.B.Kent,C.C.Fuller,Applicationofthesurface complex-ationconcepttocomplexmineralassemblages,Environ.Sci.Technol.32(1998) 2820–2828.
[57]Z.Chu,W.Liu,H.Tang,T.Qian,S.Li,Z.Li,G.Wu,Surfaceacid–basebehaviors ofChineseloess,J.ColloidInterfaceSci.252(2002)426–432.
[58] B.Muller,A.Duffek,Similaradsorptionparametersfortracemetalswith dif-ferentaquaticparticles,Aquat.Geochem.7(2001)107–126.
[59]P.Dong,X.Hua,Y.Li,J.Zhang,D.Yan,Cdadsorptionpropertiesofcomponents indifferentfreshwatersurfacecoatings:theimportantroleofferromanganese oxides,Environ.Sci.Technol.37(2003)4106–4112.
[60]P.Dong,L.Liu,X.Hua,Y.Li,Comparisonoflead,cadmium,copperandcobalt adsorptionontometaloxidesandorganicmaterialsinnaturalsurfacecoatings, Microchem.J.85(2007)270–275.
[61]S.Serrano,P.A.O’Day,D.Vlassopoulos,M.T.García-González,F.Garrido,A sur-facecomplexationandionexchangemodelofPbandCdcompetitivesorption onnaturalsoils,Geochim.Cosmochim.Acta73(2009)543–558.
[62] O.S.Pokrovsky,J.Schott,B.Dupré,Traceelementsfractionationandtransport inborealstreamsandsoilsolutionsofbasalticterrain,CentralSiberia,Geochim. Cosmochim.Acta70(2006)3239–3260.
[63]A. Probst, L. Hernandez, J.L. Probst, Heavy metals partitioning in three Frenchforestsoilsbysequentialextractionprocedure,J.Phys.IV107(2003) 1103–1106.
[64]D.Aubert,A.Probst,P.Stille,Distributionandoriginofmajorandtraceelements intolabileandresidualphaseinaacidsoilprofile(VosgesMountains,France), Appl.Geochem.19(2004)899–916.
[65]A.Manceau,M.C.Boisset,G.Sarret,J.L.Hazemann,M.Mench,Ph.Cambier,R. Prost,DirectdeterminationofleadspeciationincontaminatedsoilsbyEXAFS spectroscopy,Environ.Sci.Technol.30(1996)1540–1552.
[66] E.F.Peltier,S.M.Webb,J.F.Gaillard,Zincandleadsequestrationinanimpacted wetlandsystem,Adv.Environ.Res.8(2003)103–112.
[67]P.W.Schindler,Co-adsorptionofmetalionsandorganicligands:formationof ternarysurfacecomplexes,in:M.F.Hochella,A.F.White(Eds.),Mineral–Water InterfaceGeochemistry,Rev.Mineral.,vol.23,Min.Soc.Am.,1990.
[68] R.Levy,C.W.Francis,Adsorptionanddesorptionofcadmiumbysyntheticand naturalorgano-claycomplexes,Geoderma15(1976)361–370.
[69]C.Latrille,L.Denaix,I.Lamy,Interactionofcopperandzincwithallophane andorganicmatterintheBhorizonofanandosol,Eur.J.SoilSci.54(2003) 357–364.
[70]P.Rademacher,AtmosphericHeavyMetalsandForestEcosystems,UN/ECE, Federal Research Centre for Forestry and Forest Products (BFH), 2001, p.19.
[71]T.Bur,A.Probst,A.Bianco,L.Gandois,Y.Crouau,Determiningcadmiumcritical concentrationsinnaturalsoilsbyassessingCollembolamortality,reproduction andgrowth,Ecotoxicol.Environ.Saf.73(2010)415–422.
[72]A.Probst,D.Moncoulon,Y.Goddéris,L.Hernandez,J.P.Party,Criticalloads for lead in France: first results on forest soils, J. Phys. IV 107 (2003) 1111–1114.