j ourna l homepage: www.e lsev ie r .com/ locate /c lay
Research paper
Adsorption of uranyl ions on kaolinite, montmorillonite, humic acid andcomposite clay material
Bruno Campos a, Javier Aguilar-Carrillo a, Manuel Algarra a, Mário A. Gonçalves b, Enrique Rodríguez-Castellón c,Joaquim C.G. Esteves da Silva d, Iuliu Bobos a,⁎a Centro de Geología, Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre 687, 4169-007 Porto, Portugalb Departamento de Geologia da Faculdade de Ciências/CREMINER and CeGUL, Universidade de Lisboa, Campo Grande, 1749-016 Lisboa, Portugalc Departamento de Química Inorgánica, Facultad de Ciencias, Universidad de Málaga, Campus de Teatino s/n, 29071 Málaga, Spaind Centro de Investigação em Química, Faculdade de Ciências, Universidade do Porto, Rua do Campo Alegre 687, 4169-007 Porto, Portugal
Article history:Received 24 November 2011Received in revised form 19 August 2013Accepted 26 August 2013Available online 1 October 2013
Keywords:BatchFlow through reactorBreakthrough curvesMolecular fluorescenceInfrared and X-ray photoelectron spectroscopy
Adsorption of uranyl ions onto kaolinite, montmorillonite, humic acid and composite clay material (both claysand humic acid) was studied by measuring the system response to clay suspensions (pre-equilibrated with orwithout uranyl) and to perturbations of the solution chemistry. Adsorption behavior of selected materialsunder the frame of batch experiments was tested at high uranyl concentrations (6–1170 μg/mL; 2.5 × 10−2 to4.9 μM), whereas that under flow through continuous stirred reactor experiments was tested at low concentra-tions (1.00 × 10−4 to 1.18 × 10−4 M). Both experiments were developed at pH 4.5 and ionic strength 0.2 mM.The adsorption experiments follow a Langmuir isotherm model with a good correlation coefficient (R2 N 0.97).The calculated amount of adsorbed and desorbed uranyl was carried out by numeric integration of the experi-mental data,whereas the desorption rateswere determined from the breakthrough curve experiments. Kaolinitewith highly disordered structure adsorbed less uranyl (3.86 × 10−6 mol/g) than well-ordered kaolinite(1.76 × 10−5 mol/g). Higher amount of uranyl was adsorbed by montmorillonite (3.60 × 10−5 mol/g) andonly half of adsorbed amount was desorbed (1.85 × 10−5 mol/g). Themolecular interactions between kaolinite,montmorillonite, humic acid, composite material and saturated uranyl ion solutions were studied by molecularfluorescence, infrared and X-ray photoelectron spectroscopy. The Stern–Volmer constant obtained for montmo-rillonite (2.6 × 103 M−1) is higher than for kaolinite (0.3 × 103 M−1). Molecular vibrations of Si\O stretchingand Al\OH bending related to hydroxylated groups (`SiOH or `AlOH) of kaolinite and montmorilloniteshow structural changes when uranyl ions are adsorbed. X-ray photoelectron spectroscopy shows that theU 4f7/2 core level signals occur at 380.5 eV in either kaolinite or montmorillonite that resulted from the interac-tion of aluminol surface sites with the (UO2)3(OH)5+.
Clay minerals have been used for a plethora of applications due totheir large adsorption capacity and swelling characteristics in aqueoussuspensions (van Olphen, 1977). Also, their cation exchange capacityand adsorptive affinity for organic and inorganic ionsmake them candi-dates for use in decontamination and disposal of high-level heavymetalwastes. Among the actinides, uranium (U) is of special concern due to itsradiotoxicity, migration and exposure paths in the environment.
The most stable valence of U under oxidizing geochemical condi-tions and acidic aqueous solutions is U(VI), where it occurs as uranyl(UO2
2+) ions and at higher pH hydrolyzes to form monomers, dimers,and trimers (Grenthe et al., 1992). The pH dependent adsorption be-havior is similar to other metal oxides with a cationic adsorption edge
351 220402490.
ghts reserved.
at pH 5 to 6 and an additional anionic adsorption edge around pH 8 insystems equilibrated with atmospheric CO2 (Chisholm-Brause et al.,2001). Despite the importance of this mechanism, there is a limitedknowledge about the interactions between UO2
2+ ions and clay mineralsurfaces (Hudson et al., 1999).
The mobility of UO22+ ions in environmental soils is determined
by their interactions with minerals (i.e. clay minerals, Al- , Fe-oxyhydroxides). Many experimental clay adsorption studies focusedon the interaction of actinides (i.e. UO2
2+ ions) with high surface areamaterials proposed as potential adsorbent materials either in the pureclay minerals such as kaolinite (Guerra et al., 2010; Křepelová et al.,2006; Payne et al., 2004; Samadfam et al., 2000), montmorillonite(Catalano and Brown, 2005; Chisholm-Brause et al., 2001; Hyun et al.,2001; Kowal-Fouchard et al., 2004; McKinley et al., 1995; Pabalan andTurner, 1997; Schlegel and Descontes, 2009; Tsunashima et al., 1981),di-tri-smectite (Bauer et al., 2001; Giaquinta et al., 1997; Korichi andBensmaili, 2009), beidellite (Greathouse and Cygan, 2006), biotite
54 B. Campos et al. / Applied Clay Science 85 (2013) 53–63
(Lee et al., 2009), halloysite (Kilislioglu and Bilgin, 2002), or clay rockssuch as bentonite (Olguin et al., 1997) and opalinus clay (Joseph et al.,2011; Pekala et al., 2009).
Several experimental studies reported the UO22+ ions' (beyond
10−7 M) adsorption onto clay minerals in the presence of humic acid,which favors the increases of the sorption of the UO2
2+ onto clay min-erals (Ivanov et al., 2012; Kepelov et al., 2007; Křepelová et al., 2006,2008; Monsallier and Choppin, 2003; Sachs and Bernhard, 2008).
Kaolinite, montmorillonite, humic acid and a composite material(kaolinite, montmorillonite and humic acid) were selected in thiswork for the UO2
2+ ion adsorption experiments as a means of effectingthe removal of U from surface waters. Both clays are widely employedas a reference for clay minerals due to their relative abundance inmany soils, low cost, relative high surface area, negative surfacecharge, surface interaction with cations and, an additional uptakemechanism by interlayer adsorption or fixation in smectite structure.Thus, clay materials act as main sorbents for tri- and hexavalent acti-nides and may also serve as environmental protection of a nuclearwaste repository.
The main goal of this work is to evaluate the UO22+ sorption onto
kaolinite, montmorillonite, humic acid, and a composite material(clays and humic acid) under different experimental conditions and toimprove the understanding of the UO2
2+ sorption mechanism. Experi-ments were carried out with UO2
2+ solutions whose concentrationsfor batch experiments ranged from 6 to 1170 μg/L (2.5 × 10−2 to4.9 μM) and for the flow-through continuous reactor experimentsfrom 1.00 × 10−4 to 1.18 × 10−4 M. A continuously stirred flow-through reactor was used to facilitate the intensive pre-washing andgoodpreconditioning of clays for the appropriate control of the chemicalconditions imposed during our experiments (Grolimund et al., 1995).Sorption experiments were also coupled with spectroscopic experi-ments, in order to identify the molecular vibration changes of silanoland aluminol sites in the presence of UO2
2+ and the possible UO22+ spe-
cies adsorbed onto clays or precipitated phases as well.
2. Materials and analytic techniques
2.1. Materials
Kaolinite, montmorillonite and sapropel from Portuguese occur-rences (Ovar district, Porto Santo Island, Madeira Arq., estuarine ofSado River), and a composite material constituted by a mixture of kao-linite (60%), smectite (30%) and humic acid (10%)were used as selectedmaterials for the adsorption experiments.
The b2 μm clay fractionswere extracted by sedimentation from rawkaolin and bentonite rocks and then, used for adsorption experiments.Both kaolinite samples (SVP 7 and SVP 44) previously characterized(Bobos and Gomes, 1998; Bobos et al., 2001) were selected for ad-sorption experiments to verify whether the order–disorder or defectstructure may influence the adsorption capacity. The sample PS 1 cor-responds to Ca-montmorillonite. Humic acid extracted from sapropel(unconsolidated sedimentary rock rich in bituminous substance andpoor in cellulosic material) was isolated by a procedure recommendedby the International Humic Substances Society (IHSS) as described inliterature (Esteves da Silva et al., 1996).
Table 1Chemical compositions, cation exchange capacity, surface area and Langmuir relationships of c
- ‡ Chemical compositions from Bobos et al. (2001)
2.2. Experimental
2.2.1. Batch and kinetic adsorption experimentsContinental surface waters may contain 0.1 to 500 μg/L U (4.2 ×
10−4–2.1 μM) (Kim, 1986), whereas the groundwaters in the vicinityof U mill tailings may reach 707 μg/L U (2.97 μM) (Junghans andHelling, 1998). To test themaximum adsorption capacity of the selectedclayminerals and compositematerial, theUO2
2+ concentrations used forbatch experiments ranged from 6 to 1170 μg/mL (2.5 × 10−2 to4.9 μM) (Table 1). Uranyl acetate of analytical grade was used as sourceof U(VI). Equilibrium adsorption isotherms were carried out by adding10 μL of 0.25, 0.50, 0,75, 1, 2, 3 and 5 mM of UO2
2+ ions solutions to0.25 g clay samples (in triplicate) in 50 mL polypropylene centrifugetubes at a pH 4.5. Suspensions were shaken for 48 h, centrifuged, andvacuum filtered through 0.45 μm cellulose nitrate filters. The filtrateswere preserved by adding HNO3 (65%) and stored at 4 °C prior tochemical analysis by inductively coupled plasma-mass spectrometry(ICP-MS). The UO2
2+ ions retained by the solid phase were obtainedby difference between the initial and remaining concentration in thesupernatant.
The equilibrium adsorption isotherm data of UO22+ were analyzed
using the linear Langmuir expression (Eq. 1):
Ce
qe¼ 1
bQoþ CQO
ð1Þ
where qe is the amount of UO22+ taken upwhen the equilibriumconcen-
tration of UO22+ ions in solution is Ce. The Qo is the saturated adsorption
capacity (monolayer capacity) or maximum adsorption capacity deter-mined by the number of reactive surface adsorption sites (Qo = 1/m;m = slope), and b is the Langmuir constant, the affinity of the adsorbatefor the surface, related to free energy adsorption. The Qo of compositematerial was calculated taking into account the percentual participationof each claymineral andhumic acid [Qo(composite) = 0.6 × Qo(Kaolinite) +0.3 × Qo(Smectite) + 0.1 × Qo(HA)].
The essential feature of Langmuir isotherm is also expressed in termsof a dimension-less factor called separation factor or equilibrium pa-rameter (RL), defined by Eq. (2) of Weber and Chakkravorti (1974):
RL ¼ 1= 1þ b Cið Þ ð2Þ
where Ci is the initial concentration of UO22+ and b is the Langmuir
constant.Sorption kinetic studies of UO2
2+ ions (0.1 mM) onto kaolinite,montmorillonite, humic acid and composite material were carried outusing a continuously stirred flow-through reactor loaded with 0.58 gof clays (13 g/L) in the size range of 0.4–20 μm. Input solution flowsthrough 0.2 μm membranes at room temperature with a flow rate of0.8 mL/min using a peristaltic pump for 18–20 h from a reservoir intothe reactor. Outflow solutions were collected at different times forICP-MS analysis.
The systemwas previously preconditionedwith NaCH3COO (pH 4.7)until reaching an approximately constant pH in the outflow solution. TheUO2
2+ solutionwas pumped into the reactor afterwards, ideally until theclayey system reaches saturation (inflow and outflow solutions withthe same UO2
2+ concentration). After this step, the original NaCH3COO
lay minerals and composite material used for adsorption experiments.
solution was used again as input solution into the system. Therefore,the measured outflow UO2
2+ concentrations in time followed a break-through curve (BTC). The amount of UO2
2+ adsorbed was calculated bynumerical integration of the BTC describing the UO2
2+ concentration intime. The values were subtracted from the integration of the samecurve describing the outflow concentration of an inert species, which isdescribed by Eq. (3) in a perfectly stirred media (Missen et al., 1999):
E tð Þ ¼ 1texp −t=t
� � ð3Þ
where t is the time, and t is the mean residence time given by the ratiobetween the reactor volume (V) and the flux (q).
The equation ofmaterial balance for continuous stirredflow-throughreactor within a constant density solution is given by Eq. (4) of Missenet al. (1999):
dnj
dt¼ q C j;in−C j;out
� �−Rj ð4Þ
where nj is the number of moles of element j, t is the time, Cj,in and Cj,outare the input and output concentrations, q is the volumetric flow rate,and Rj is the extensive rate of adsorption/desorption of element j.
The extensive rate of adsorption/desorption was obtained at anypoint in the experiment (for a reactor of volume V) by rearrangingEq. (4) according to Missen et al. (1999) as follows:
Rj ¼ −dC j;out
dtV þ q C j;in−C j;out
� �ð5Þ
If the system is in steady-state, dnj/dt = 0, the equation is muchsimplified. Intensive rates of adsorption/desorption (rj) is related to Rj,where rj = Rj/Nq (Nq is a normalizing quantity such as surface area,volume of reactor, or mass of adsorbent).
2.2.2. Fluorescence emissionTo obtain an estimation of the steady state association magnitude
between clay minerals and UO22+ the following experiments were
performed: adequate volumes of UO22+ (10−3 M) and pre-conditioned
clay mineral suspension solutions (10 g/L) were mixed in order toachieve a final clay concentration ranging from 0.01–1.73 g/L, and atconstant UO2
2+ (10−5 M) concentration. Once the suspensions werethoroughly mixed, sample aliquots were centrifuged and the superna-tant solutions were analyzed by molecular fluorescence. In order toestimate the association constant, a Stern–Volmer type model wasused (Lakowicz, 1999):
IIO
¼ 1þ ΚS•V Q½ � ð6Þ
where IO and I are the intensity or rate of UO22+
fluorescence with andwithout the presence of clay, respectively. The KSV is the quencher ratecoefficient (Stern–Volmer constant), and Q is the clay concentration.
2.3. Analytical techniques
2.3.1. CEC and nitrogen adsorption isothermsThe cation exchange capacity (CEC) was determined by ammonium
acetate exchange (Davis and Worrall, 1971). The Brunauer–Emmett–Teller (BET) specific surface area was determined using a Gemini2370 V5 (Micromeritics, USA) by measuring the volume of a mixtureof He/N2 adsorbed at five different pressures. Prior to He/N2 adsorption,the samples were outgassed at 200 °C for 10 h (Brunauer et al., 1938).
2.3.2. X-ray diffractionX-ray diffraction (XRD) patterns of oriented and random specimens
were obtainedusing a Philips X'Pert series automated diffraction system
equipped with a graphite monochromator and a CuKα (λ = 1.5406 Å)radiation. Samples were analyzed in the range 2–50°2θ, using a 1°divergence slit, a step increment of 0.05°, 2θ and a counting time of5 s/step.
2.3.3. Quenching of fluorescence emissionsMolecular fluorescence analysis was carried out with a Horiba Jovin
Yvon Fluoromax 4 TCSPC spectrophotometer, equipped with a Xenonpulse discharge lamp (75 W) using an excitation of 270 nm and an emis-sion range of 400–700 nm, integration time of 0.1 s and a slit width of5 nm for both excitation and emission. Pre-conditioned clay mineralsuspension solutions (10 g/L) were mixed with a final concentration ofUO2
2+ ions (10−5 M) in order to achieve afinal clay concentration rangingfrom 0.01–1.73 g/L. Once the solutions were thoroughly mixed andcentrifuged, they were analyzed every 5 min.
2.3.4. Infrared spectroscopySelected samples were analyzed by Fourier-transform infrared spec-
troscopy (FTIR) after sorption experiments by flow-through continuousreactor. The infrared spectra in transmission mode were recorded inthe 4000–400 cm−1 frequency region using a Bruker Tensor 27 spec-trometer. The measurements of the adsorption bands integrated inten-sityweremade usingOPUS software supplied by the Bruker instrument.The pellet disks of 1.5 cm diameter were prepared by mixing 1 mgsample with 200 mg KBr (Aldrich) and pressing at 10 kg/cm2.
2.3.5. X-ray photoelectron spectroscopyX-ray photoelectron spectra (XPS) were recorded using a Physical
Electronics PHI 5701 spectrometerwith a non-monochromatic Al Kα ra-diation (300 W, 15 kV, hν = 1486.6 eV) as the excitation source. Spec-tra were recorded at 45° take-off angle by a concentric hemisphericalanalyzer operating in the constant pass energy mode at 25.9 eV, usinga 720 mm diameter analysis area. Under these conditions the Au 4f7/2line was recorded with 1.16 eV FWHM at a binding energy of 84.0 eV.The spectrometer energy scale was calibrated using Cu 2p3/2, Ag 3d5/2and Au 4f7/2 photoelectron lines at 932.7, 368.3 and 84.0 eV, respec-tively. Charge referencing was done against adventitious carbon (C 1 s284.8 eV). Powdered solids were mounted on a sample holder withoutadhesive tape and kept overnight in high vacuum in the preparationchamber before they were transferred to the analysis chamber ofthe spectrometer. Each region was scanned with several sweeps untila good signal to noise ratio was observed. The pressure in the analysischamber was maintained lower than 10−9 Torr. A PHI ACCESS ESCA-V6.0 F software package was used for acquisition and data analysis.A Shirley-type background was subtracted from the signals. Recordedspectra were always fitted using Gauss–Lorentz curves in order todetermine more accurately the binding energy of the different ele-ment core levels. The accuracy of binding energy (BE's) values waswithin ± 0.1 eV.
2.3.6. Inductively coupled plasma-mass spectrometryThe uranium concentrations in acidified solutions were determined
by external standard calibration method using an inductively coupledplasma-mass spectrometry (ICP-MS) (THERMO X series ELEMENT2and Amiga series, JobinYvon equipments). Calibration was done ineach analytical session by the external standard method. Also, majorelements of smectite and composite material were analyzed by ICP-MS.
2.3.7. Modeling of adsorption reactionAqueous speciation and saturation indexwith respect to solid phases
in experimental conditions were calculated using Visual MINTEQ code(Gustafsson, 2006) with the database provided by program. The UO2
2+
speciation was modeled taking into account the following parameters:[UO2
2+] = 10−4 M; pCO2 = 3.16 × 10−4 atm; T = 298°K
y = 0.7132x + 13.861R² = 0.9964
y = 0.0112x + 3.5217R² = 0.9568
y = 0.0326x + 1.7925R² = 0.9925
0
20
40
60
80
100
120
140
160
0 200 400 600 800 1000 1200
Ce/
qe(
g/L
)
Ce(mg/L)
Kaolinite
Composite
Smectite
Fig. 2. Langmuir isotherms for kaolinite (SVP 7 and SVP 44), Ca-montmorillonite (PS 1)and composite material used for batch experiments.
56 B. Campos et al. / Applied Clay Science 85 (2013) 53–63
3. Results and discussion
3.1. Mineralogical and physico-chemical analyses
The XRD patterns of randomly oriented clay aggregates of twoselected kaolinite samples (Fig. 1a) show structural difference interms of crystallinity and order–disorder structure: well-ordered kao-linite (SVP 7) and poorly-ordered kaolinite (SVP 44). The peak posi-tions at 7.10 Å and 3.55 Å correspond to 00l and 002 reflections ofkaolinite. Both samples were selected to verify whether the order–disorder or defect structure influences the rate of adsorption experi-ments. The XRD patterns of the b2 μm oriented clay-aggregate spec-imens (PS 1) in AD and EG show the 001 reflections correspondingto Ca-montmorillonite (Fig. 1b). The 001 reflection in AD conditionoccurs at 15 Å, whereas in EG conditions the 001 reflection shiftedat 17 Å. XRD pattern of sapropel sample (Fig. 1c) shows a mineralog-ical composition constituted by calcite, quartz, plagioclase, illite andkaolinite. Sapropel is constituted by ~70% organic colloid, 18% carbon-ates and 12% silicates. Humic acids extracted from sapropel are con-stituted by amorphous organic material. Both peaks at 2.11 Å and1.80 Å correspond to Cu used for background normalization (Fig. 1c).
Chemical composition of the b2μm clay fractions of kaolinite, mont-morillonite and compositematerial are shown in Table 1. Kaolinite usedfor adsorption experiments contain only silanol and aluminol layersas surface sites, where the layer charge is null. The layer charge ofmontmorillonite occurs in octahedral sheet. The total layer charge ofmontmorillonite calculated from the structural formula is about −0.18.
Fig. 1. X-ray patterns of the b 2 μm clays fractions of well- and poorly-order kaolinite corresponconditions) (b), and sapropel and humic acid (c) used in adsorption experiments (C = calcite
Chemistry of composite material (Table 1) reflects the mineralogicalcomposition: 60% kaolinite, 30% smectite and 10% humic acid (the % oforganic matter was not analyzed).
CEC and BET surface area data obtained on clay minerals andcomposite material used for the adsorption experiments are reportedin Table 1. Montmorillonite used in this study has a higher CECvalue (1.15 mmol/g) than other montmorillonite samples (0.454 and0.820 mmol/g) reported elsewhere (McKinley et al., 1995).
ding to samples SVP 7 and SVP 44 (a), Ca-montmorillonite (air-dried and ethylene-glycol; Q = quartz; P = plagioclase; I = illite; K = kaolinite).
3.2. U(VI) sorption on kaolinite, montmorillonite, humic acid and compositematerial
3.2.1. Batch experimentsThe UO2
2+ concentrations used for each sample during batch exper-iments are shown in Table 1. The mean RL value calculated for each claymineral is 0 b RL b 0.80 (Table 1), indicating a favorable adsorption pro-cess (0 b RL b 1) (Weber and Chakkravorti, 1974). The experimentaladsorption data show a good Langmuir isotherm model (Fig. 2),where the linear traces for kaolinite, montmorillonite and compositematerial have a good correlation coefficient (R2 = 0.978 to 0.988).
The Qo (μg/mg) lies from 1.40 to 90, where the higher adsorptioncapacity corresponds to montmorillonite, as was expected. However,the maximum adsorption obtained for montmorillonite is less thanthe exchange capacity determined. The UO2
2+ ions occupied 29% of
Fig. 3.Adsorption kinetic experiments of UO22+ ontowell-order kaolinite (SVP 7) and Ca-montm
to Eq. 2). Upper graphs show pH of solution as measured from the outflow solution and bottominflow concentration (C0). Solid curve represents the outflow concentration in the sampled so
surface sites in montmorillonite and only 16% in kaolinite. Loweradsorption capacity measured is assumed to be high initial UO2
2+ con-centration used and hydrolysis of UO2
2+ that could affect adsorption(McKinley et al., 1995). Uranyl ions are specifically adsorbed onthe edge sites of montmorillonite, without any competition from ex-changeable cations. Otherwise, exchangeable charge sites are sensitiveto competitive exchange between Ca2+ and UO2
2+ ions. Thus, Ca2+ isknown to outcompete UO2
2+ for exchange sites in montmorillonite(Tsunashima et al., 1981).
No differenceswere observed during batch experiments for kaolinitesamples with distinct order–disorder (SVP 7 and SVP 44). The compos-ite material shows an intermediary adsorption capacity (30.7 μg/mg)mainly due to the contribution of montmorillonite (30%) and thepresence of humic acid (10%). Also, the UO2
2+ uptake onto kaolinite in-creases in the presence of humic acid at pH b 5 (Křepelová et al., 2006).
orillonite (PS 1). Time is represented as pore volumes (mean residence time (t) accordinggraphs the outflow uranium concentration in the sampled solution (C), normalized to thelution for any inert species (Eq. 2).
58 B. Campos et al. / Applied Clay Science 85 (2013) 53–63
The better efficiency of montmorillonite is due to adsorption ofpositively charged UO2
2+ species on the hydroxylated surface sites(McKinley et al., 1995). It was estimated that only 10% of the activesites calculated by BET specific area is accounted for by crystallite edgeof montmorillonite (Duff and Arheim, 1996). Maximum adsorptioncapacity of composite material yields approximately 28.6 μg/mg, avalue concordant with the experimental data.
The amphoteric surface sites consisting of aluminol, silanol andbridging groups (NAl–OH–Sib) are considered the available bindingsites. Uranyl bindingonkaolinite surface siteswasmodeled onto aluminolsites (Kohler et al., 1992), whereas silanol sites are excluded for U(VI)
Fig. 4. Adsorption kinetic experiments of UO22+ on humic acid (Experiment OM; upper gra
represented as pore volumes (mean residence time (t) according to Eq. 2). Upper graphs shouranium concentration in the sampled solution (C) normalized to the inflow concentration (Cspecies (Eq. 2).
sorption (Payne et al., 2004). However, edge surface sites represent acomplex Al/Si oxide (Stumm, 1992), where adsorption experimentsreported very small affinity of U for Si-oxide surfaces (Borovec, 1981).
3.2.2. Sorption by flow-through continuous stirred reactorThe kinetic adsorption experiments and the experimental points
were plotted jointly with the theoretical curve for an inert species(Figs. 3 and 4), whereas the summary analytical results are reportedin Table 2. The initial adsorption edge for UO2
2+ ions (0.1 mM) is steepand the system reaches saturation very rapidly for the kaolinite sam-ples, while for montmorillonite sample it takes longer to reach the
phs) and composite material (Experiment OM + Kaol + Sm; bottom graphs). Time isw pH of solution as measured from the outflow solution and bottom graphs the outflow0). Solid curve represents the outflow concentration in the sampled solution for any inert
Table 2Experimental conditions used for selected samples: pH and initial UO2
2+ concentration. Total adsorbed and desorbedUO22+was calculated by numeric integration as explained in text, and
desorption rates from BTC experiments were computed with data from the final stages of the experiment. Experimental conditions used for selected samples: pH and initial UO22+
concentration.
Samples pH UO22+ (M) IS (mM) Total adsorption (mol/g) Total desorption (mol/g) Diference (%) Desorption rate
same state (Fig. 3). The pH evolution in the system seems to be rela-tively stable, depending on the pH equilibrium with each type of claymineral. For instance, the pH equilibrium is different for kaolinite(pH 4.6 to 4.7) and montmorillonite (pH 6). Thus, the initial pre-conditioning step permitted altering of the pH solution towards adesired value where the UO2
2+ adsorption does occur. Once the UO22+
solution was pumped into the reactor and adsorption started, the pHdecreased due to the deprotonation of the surface sites that adsorbthe UO2
2+ ions.
a
b
0
100000
200000
300000
400000
500000
600000
700000
460 470 480 490 500 510 520 530 540
Kaolinite SVP-44
Wavelength (nm)
Flu
ore
scen
ce In
ten
sity
(a.
u.)
460 470 480 490 500 510 520 530 540
0
100000
200000
300000
400000
500000
600000
Montmorillonite PS-1
Wavelength (nm)
Flu
ore
scen
ce In
ten
sity
(a.
u.)
Fig. 5. Fluorescence emission of UO22+ adsorbed by kaolinite (concentration: 0.01-
1.73 g L-1) (a) and Ca-montmorillonite (concentration: 10-5 M) (b).
The experiment with the montmorillonite sample (PS 1) proceededwith desorption, where the pH solution re-equilibrated with montmo-rillonite steadily increasing close to pH 6. Adsorption of UO2
2+ ions isgreater in montmorillonite than in poorly-ordered kaolinite (SVP 44)by a factor of 10, and by a factor of 2 relative to thewell-ordered kaolin-ite (SVP 7). The results are in agreement with the batch adsorptionexperiments where the adsorption capacity of montmorillonite is alsogreater by almost 2 orders of magnitude relative to kaolinite samples.
The adsorption experimentwith humic acid did not reach saturationshowing a curve that rises linearly (Fig. 4) and stopped at 30% belowsaturation (shown by the curve for an inert species). No desorptioncycle has been done in this experiment. The experiment with thecomposite material enhanced the adsorption slightly relative to mont-morillonite. The composite material shows an intermediate adsorptioncapacity between kaolinite and montmorillonite, by contrasts with theresults obtained from the batch experiments.
The composite material adsorbed about half of the total UO22+ ion
uptake by humic acid, in almost reaching saturation (Fig. 4). The differ-ence is due to the outflow solution concentration which stayed belowsaturation, as it is evident from the area between the experimentalpoints and the top of the curve for an inert species in both experiments(humic acid and composite material) (Fig. 4). This suggests how humicacid is overwhelmingly important to UO2
2+ ion adsorption.Furthermore, the relative adsorption decreases when humic acid is
mixed with clay minerals. The number of available sites for the UO22+
ion adsorption will decrease, because the humic acid may impact thesorption surface sites of clay minerals involved in these experiments.
A desorption cyclewas run in the experimentswithmontmorilloniteand the composite material after the UO2
2+ saturation. The results(Table 2) show that almost the same amount of UO2
2+ was desorbedfrom both materials. However, the amount of UO2
2+ initially adsorbedwas different and thus, the values obtained show different relativeamounts. About 51.5% of the UO2
2+adsorbed by montmorillonite wasdesorbed during 23.5 h. A different outcome is obtained for the com-posite material, due to the presence of humic acid. While the total ad-sorption increased, the relative desorption of total UO2
2+ ions adsorbedwas 21.6% after 10 h, which is about half of desorption time of mont-morillonite. Taking into account that the most adsorbed speciesare desorbed in the first few hours after the introduction of the blanksolution, the results obtained suggest that the resilience time ofthe adsorbed UO2
2+ ions may not be much different when using onlymontmorillonite. However, the experiment using the composite mate-rial did not run for enough time, therefore it is not possible to asses ifthe system would approach the same steady outflow of UO2
2+ ions.
Table 3The Stern–Volmer parameters for the fluorescence quenching of the UO2
60 B. Campos et al. / Applied Clay Science 85 (2013) 53–63
Using the 4 final sampled points in themontmorillonite experiments,the average release ofUO2
2+ ions from the reactorwas computedbecauseits variation is small (dC/dt → 0), and the desorption ratewas calculatedusing Eq. (4). The calculated rate is 1.05 × 10−8 mol/g/min, indicatingthat desorption may continue steadily for longer periods of time.Thus, the UO2
2+ ion does not seem particularly stable at the water–
Fig. 7. IR spectra of raw uranyl acetate, well-ordered kaolinite (SVP 7), composite material, uCa-montmorillonite and uranyl adsorbed by Ca-montmorillonite (b).
mineral surface interface at these pH conditions. For the compositemate-rial, the UO2
2+ ion concentration in the outflow solution did not flatten astime progressed and no desorption rates could be confidently calculatedduring the experimental desorption (lasted for only 10 h).
3.3. Fluorescence emission
The interaction of kaolinite and smectite solutions with UO22+ ions
(10−5 M) shows a decrease in the emission intensity as clay content in-creases (Fig. 5a and b). The Stern–Volmer relationships were obtainedbetween relative UO2
2+fluorescence intensity and the concentration
of both kaolinite and montmorillonite in order to explore the steadystate association process (Table 3). As was expected the value of theStern–Volmer constant obtained for montmorillonite (2.6 × 103 M−1)is higher than for kaolinite (0.3 × 103 M−1), suggesting a higher affinityof UO2
2+ ions towards montmorillonite (Fig. 6). The same statementsusing other inorganic fluorescence quenches of UO2
2+ were also report-ed elsewhere (Esteves da Silva et al., 1996; Sladkov et al., 2009). Theresults support our statements that both clays form stable complexesand/or adsorb UO2
2+ ions.
3.4. Infrared spectroscopy
The IR spectra of raw UO2(CH3COO)2, kaolinite (SVP 7), montmoril-lonite (PS 1) and UO2
2+ adsorbed by kaolinite, montmorillonite andcomposite material are shown in Fig. 7a and b.
ranyl adsorbed by kaolinite by composite material (a). IR spectra of raw uranyl acetate,
Table 5Atomic concentrations (%) and Si/U and Al/U molar ratios calculated from the XPS data.
bands are not resolved at low wavelengths.The Si\O stretching region of kaolinite minerals comprises three
adsorption bands at 1114 cm−1, 1030 cm−1 and 1008 cm−1 either inwell-ordered or poorly-ordered kaolinite (Fig. 7a). The doublet at1030 cm−1 and 1008 cm−1 (in-plane and anti-symmetric stretch ofequatorial Si\O\Si bonds) becomes broader due to the υ2 bendingvibration of UO2
2+ overlapped on Si\O stretching vibration of kaolinite.The OH− bending vibrations of kaolinite occur at 916 cm−1 and inner-surface hydroxyl groups at 938 cm−1 (Farmer, 1974). A shoulder at935 cm−1 is well visible when UO2
2+ is adsorbed by kaolinite. This cor-responds to inner-surface Al\OH bending in kaolinite (OH in-planebending vibrations of inner-surface hydroxyl group), where the υ3asymmetric stretching of UO2
2+ is supposed binding to Al\OH bending.The IR spectrum of composite material (Fig. 7a) reflects the vibra-
tion planes of kaolinite minerals, previously discussed. However, theintensity of the Al\OH bending bands decreased very much and alarge shoulder occurs at 935 cm−1.
The IR spectra of raw UO2(CH3COO)2, montmorillonite and UO22+
adsorbed by montmorillonite are shown in Fig. 7b. The Al\OH bondat 917 cm−1 occurs as a shoulder to the Si\O stretching characterizedby broadening due to overlapped the υ2 bending vibration and the υ3asymmetric stretching of UO2
2+ (Fig. 7b). Two vibration planes of lowintensity at 880 cm−1 (Al–Fe–OH) and 840 cm−1 (Al–Mg–OH), con-firm that Al is the main atom in the octahedral sheet (Farmer, 1974).Although uranyl adsorption onto montmorillonite involves multiplebinding sites, including ion exchange and edge coordination sites(McKinley et al., 1995). The IR spectrum of UO2
2+ ions adsorbed ontomontmorillonite did not show vibrational changes in the Al–Fe–OH orAl–Mg–OH regions.
The υ1 symmetric stretching of UO22+ corresponds to two strong
vibration planes at 800 and 760 cm−1 in raw UO2(CH3COO)2. Bothvibration planes occur also in the IR-spectra of UO2
2+ adsorbed ontokaolinite, montmorillonite and composite material (Fig. 7 a and b),which suggests that some amounts of free UO2
2+ not adsorbed duringthe flow-through continuous stirred reactor.
3.5. X-photoelectron spectroscopy
The identification of theUO22+ species adsorbed on clayminerals has
been carried out by XPS analysis. The element composition in atomicconcentration (%) and the Si/U and Al/U atomic ratios of the studiedsamples after UO2
2+ ion sorption experiments in flow reactor areshown in Table 4. As expected, kaolinite samples (SVP 7 and SVP 44)adsorbed lower amounts of UO2
2+ ions than montmorillonite (PS 1).The XPS binding energies (eV) for the main detected elements in thestudied samples are shown in Table 5. The observed values of the O1s,Si2p and Al2p binding energies for samples SVP 7, SVP 44 and PS 1 aresimilar to those observed by other authors either in kaolinite (Barr,1983) or montmorillonite (Moulder et al., 1992). Fig. 8a shows the C1score level spectra for samples SVP 7, SVP 44 and PS 1, where the photo-emissions K2p3/2 and K2p1/2 are clearly observed in montmorillonite.
Table 4XPS binding energies (eV) corresponding to the main detected elements in the samplesanalyzed.
Samples Si 2p O 1 s U 4f7/2 Al 2p
Kaolinite(SVP 7)
102.7 532.0 380.5 74.4
Kaolinite(SVP 44)
102.6 531.9 380.5 74.2
Ca-montmorillonite 102.6 531.9 380.5 74.4
The C1s signal is slightly asymmetric mainly due to surface contamina-tion by the possible presence of acetate groups. Thus, the anion of theUO2
2+ source cannot be completely ruled out due to the presence of avery weak photoemission between 288.0 and 289.0 eV characteristicof this anion (Froideval et al., 2003). The U4f7/2 core level signals occurat 380.5 eV near to the value observed for the UO2
2+monomeric speciesin acid media (Borovec, 1981), where a strong interaction between theUO2
2+ moiety and the kaolinite and montmorillonite is suggested. Inall cases, the shape of the doublet U4f7/2 and U4f5/2 (Figs. 8b and 8c)does not indicate the presence of components at higher binding energy
B.E. (eV)
I (c
400 395 390 385 380 375
6200
6400
6600
Fig. 8. a) XPS spectra of C1s core level spectra ofwell- and poorly-order kaolinite (samplesSVP 7 and SVP 44) and Ca-montmorillonite (sample PS1; b) U 4f corel level spectrum ofsample SVP 44; c) U 4f core level spectrum of sample PS 1.
Fig. 9. The diagram of aqueous UO22+ speciation (Visual MINTEQ code).
62 B. Campos et al. / Applied Clay Science 85 (2013) 53–63
due to the presence of oligomeric UO22+species. Nevertheless, the U4f7/2
binding energy at 380.2 ± 0.3 eV was associated with the presenceof U(VI)/alumina complexes depending on the pH values (Kowal-Fouchard et al., 2004), where thefirst species (a–r1 and 381.8 eV) resultfrom the interaction between free UO2
2+ ion and aluminol surface sites,and the second species (a–r2 and 380.3 eV) from the interaction ofaluminol surface sites with the (UO2)3(OH)5+ aqueous complexes atpH values N 5.
3.6. Uranyl speciation modeling in aqueous solution
The relative distribution of UO22+ [10−4 M] species as a function
of pH (pH 4–6) is shown in Fig. 9. Five hydrolyzed UO22+ species
[i.e., UO22+, UO2(OH)+, (UO2)2(OH)22+, (UO2)3(OH)5+, UO2CO3 and
UO2(CH3COO)+] occur at relevant levels in dilute solutions in thepresence of dissolved CO2 at pH 4–6. The UO2
2+ is the main species(N90%) at pH b 4, which decreases ~33% in solution at 4 b pH b 5.The remaining species (UO2)3(OH)5+, UO2(OH)+, (UO2)2(OH)22+ andUO2(CH3COO)+ increase by 28, 18, 14 and 6%, respectively. The promi-nent species at pH between 5 and 6 becomes (UO2)3(OH)5+, whichincreases up to 60% at pH 6. The UO2(OH)+ remains constant (~20%),whereas UO2
2+, (UO2)2(OH)22+ and UO2(CH3COO)+ decrease to relativedistribution values below 5%. Within the pH range selected for the ex-periments, speciation with acetate is rather negligible. Raising the pHto neutral values leads to the increase of the carbonated species, special-ly UO2CO3 which increases up to 10% at pH 6. The occurrence of precip-itated phases has been also considered. The results obtained show thatschoepite [(UO2)4O(OH)6 . 6(H2O)] becomes supersaturated abovepH 5.3. Therefore, the potential phase precipitationmay occur at higherpH values. Nevertheless, our experiments were conducted at lower pHvalues. Overall, in spite of thermodynamic database uncertainty, theo-retical calculations show that the initial solutions employed in our ex-periments (pH b 5) were undersaturated with respect to uranyl saltsand other solid phases in the experimental pH range conditions.
4. Conclusions
The UO22+ ions adsorption onto selected kaolinite, montmorillonite
and composite material did not reached to fill themaximumadsorptioncapacity. Maximum adsorption during batch experiments is less thanthe exchange capacity of kaolinite and montmorillonite, where only29% of surface sites of montmorillonite and 16% of kaolinite were occu-pied by UO2
2+ ions. By contrast, the UO22+ ion adsorption was found
higher than the exchange capacity measured on composite material.Kinetic adsorption experiments (pH 4.5–4.7) carried out with the
flow-through continuous reactor show higher amounts of UO22+ ions
adsorbed on kaolinite, montmorillonite and composite material thanthose obtained by batch experiments. Well-ordered kaolinite (SVP 7)
adsorbed more UO22+ than poorly-ordered kaolinite (SVP 44). The
UO22+ ion adsorption on ordered structures (i.e., kaolinite) may com-
petewithmontmorillonite. Also, the UO22+ ion adsorption is slightly en-
hanced in the presence of humic acid. This could be due to the existenceof additional binding sites for U(VI) onto humic acid, which explainstheir higher capacity. Otherwise, the humic acid may impact the sorp-tion of UO2
2+ ions onto various clays, where the number of surfacesites on the clays is reduced (Křepelová et al., 2006) due to the presenceof humic acid. The UO2
2+ ions are nearly equally desorbed either frommontmorillonite or composite material.
The IR spectra show interaction of UO22+ ionswith either kaolinite or
montmorillonite surface sites. Thus, the Si\O stretching and Al \OHbending denoted structural changes when UO2
2+ ions were adsorbed.Based on analogies with pure oxide phases and bond strength consider-ations (Stumm, 1992), the NAlOH groups are considered much morereactive sites than the NSiOH.
The XPS results show that the U 4f7/2 core level signals at 380.5 eVoccur either in kaolinite or montmorillonite samples, which resultedfrom the interaction of aluminol surface sites and the (UO2)3(OH)5+.The evaluation of UO2
2+ binding sites onto kaolinite and montmorillon-ite confirm that the polynuclear species [(UO2)3(OH)5+] are favored at10−4 M uranyl concentrations and pH ~ 5. Thus, the UO2
2 speciationmodeling shows that the dominant surface specie between pH 5 and6 is the (UO2)3(OH)5+, which could be increased up to 60% at pH 6.There is no XPS data to support the existence of neutral complex`SiO2UO2
2+ and the typical oxide-like sorption edges vs. pH could beadequately fit assuming the contribution of the hydrolized oligomericUO2
2+species.
Acknowledgments
Thank to the two anonymus reviewers for their excellent andconstructive reviews which greatly improved the quality of the initialmanuscript. The authors are also grateful to Professor U. Constantino(Associate Editor) for careful editorial input and suggestions thatimproved the quality of the manuscript. J. Aguilar and B. Camposthank the Fundação para a Ciência e a Tecnologia (FCT) — Lisbon forthe scholarship in the frame of the PTDC/CTE-GEX-82678/2006 project.The authors acknowledge the financial support in the frame ofPTDC/CTE GEX 82678/2006 project and to the Ciência 2007 Programfinanced by FCT — Lisbon. E. Rodríguez-Castellón and M. Algarraacknowledge project CTQ2012-37925-C03-03 (MINECO, Spain) andFEDER funds and Andalucía Tech. Thanks to Nuno Durães and VanessaGuimarães for technical help during adsorption experiments.
References
Barr, T.L., 1983. An XPS study of Si as it occurs in adsorbents, catalysts, and thinfilms. Appl.Surf. Sci. 36, 1–35.
Bauer, A., Schafer, T., Dohrmann, R., Hoffmann, H., Kim, J.I., 2001. Smectite stabilityin acidsalt solutions and the fate of Eu, Th and U in solution. Clay Miner. 36, 93–103.
Bobos, I., Gomes, C., 1998. Greisen and post-greisen alteration in the kaolin deposit of SãoVicente de Perreira, Portugal. Can. Mineral. 38, 1621–1634.
Bobos, I., Duplay, J., Rocha, J., Gomes, C., 2001. Kaolinite to halloysite-7Åtransformation inthe kaolin deposit of São Vicente de Perreira, Portugal. Clays ClayMiner. 49, 596–607.
Borovec, Z., 1981. The adsorption of uranyl species by fine clay. Chem. Geol. 32, 45–58.Brunauer, S., Emmett, P.H., Teller, E., 1938. Adsorption of gases in multimolecularlayers.
J. Am. Chem. Soc. 60, 309–319.Caldow, G.L., Cleave van, B.A., Eager, R.L., 1960. The infrared spectra of some uranyl com-
pounds. Can. J. Chem. 38, 772–782.Catalano, J.G., Brown Jr., G.E., 2005. Uranyl adsorption onto montmorillonite: Evaluation
of binding sites and carbonate complexation. Geochim. Cosmochim. Acta 69,2995–3005.
Chisholm-Brause, C.J., Berg, J.M., Matzner, R.A., Morris, D.E., 2001. Uranium(VI) sorptioncomplexes on montmorillonite as a function of solution chemistry. J. Colloid InterfaceSci. 233, 38–43.
Davis, G.A., Worrall, W.E., 1971. The adsorption of water by clays. Trans. Britanica Ceram.Soc. 70, 71–75.
Duff, M.C., Arheim, C., 1996. Uranium(VI) adsorption on goethite and soil incarbonatesolutions. Soil Sci. Soc. Amer. J. 60, 1393–1400.
Esteves da Silva, J.C.G., Machado, A.A.S.C., Oliveira, C.J.S., 1996. Study of theinteraction ofa soil humic acid with UO22+ by self-modelling mixture analysis ofsynchronousmolecular fluorescence spectra. Analyst 121, 1373–1379.
Farmer, V.C., 1974. The layer silicates. In: Farmer, V.C. (Ed.), Infrared Spectra of Minerals.Mineralogical Society, London, pp. 331–363.
Froideval, A., del Nero, M., Barillon, R., Hommet, J., Mignot, G., 2003. The pH dependenceof uranyl retention in a quartz/solution system: an XPS study. J. Colloid Interface Sci.266, 221–235.
Giaquinta, D.M., Soderhom, L., Yuchs, S.E., Wasserman, S.R., 1997. The speciation of ura-nium in smectite clay: Evidence for catalyzed uranyl reduction. Radiochim. Acta 76,113–121.
Greathouse, J.A., Cygan, R.T., 2006. Water structure and aqueous uranyl (VI) ad-sorption equilibria onto external surfaces of beidellite, montmorillonite, andpyrophyllite: results from molecular simulations. Environ. Sci. Technol. 40,3865–3871.
Grenthe, I., Fuger, J., Konings, R.J.M., Lemire, R.J., Muller, A.B., Nguyen-Trung, C., Wanner,H., 1992. Chemical Thermodynamics of Uranium. Elsevier, New York 715.
Grolimund, D., Borkovec, M., Federer, P., Sticher, H., 1995. Measurement of sorption iso-therms with flow-through reactors. Environ. Sci. Technol. 29 (9), 2317–2321.
Guerra, D.L., Leidens, V.L., Viana, R.R., Airol, C., 2010. Amazon kaolinite functionalized withdiethylenetriamine moieties for U(VI) removal: Thermodynamic of cation-basicinteractions. J. Hazard. Mater. 180, 683–692.
Gustafsson, J.P., 2006. Visual MINTEQ v. 2.51. Stockholm, Sweden. Available at http://www.lwr.kth.se/English/OurSoftware/vminteq/index.htm.
Hudson, E.A., Terminello, L.J., Viani, B.E., Denecke, M.A., Reich, T., Allen, P.G., Bucher, J.J.,Shuh, D.K., Edelstein, N.M., 1999. The structure of U6+ sorption complexes on vermic-ulite and hydrobiotite. Clays Clay Miner. 47, 439–457.
Hyun, S.P., Cho, Y.H., Hahn, P.S., Kim, S.J., 2001. Sorption mechanism of U(VI) on a refer-ence montmorillonite: Binding of the internal and external surfaces. J. Radioanal.Nucl. Chem. 250, 55–62.
Ivanov, P., Griffiths, T., Bryan, D.N., Bozhikov, G., Dmitriev, S., 2012. The effect of humicacid on uranyl sorption onto bentonite at trace uranium levels. J. Environ. Monit.14, 2968–2975.
Joseph, C., Schmeide, K., Sachs, S., Brendler, V., Geipel, G., Bernhard, G., 2011. Sorption ofuranium(VI) onto opaline clay in the absence and presence of humicacid in opalineclay pore water. Chem. Geol. 284, 240–250.
Junghans, M., Helling, C., 1998. Historical mining, uranium tailings and waste disposal atone site: Can it be managed? A hydrological analysis. Proceedings Tailings and MineWaste '98. Balkema, Rotterdam, pp. 117–126.
Kepelov, A., Brendler, V., Sachs, S., Baumann, N., Bernhard, G., 2007. U(VI)- Kaolinite sur-face complexation in absence and presence of humic acid studied by TRFLS. Environ.Sci. Technol. 41, 6142–6147.
Kilislioglu, A., Bilgin, B., 2002. Adsorption of uranium on halloysite. Radiochim. Acta 90,155–160.
Kim, J.L., 1986. Chemical behaviour of transuranic elements in natural aquatic systems.In: Freeman, A.J., Keller, C. (Eds.), Handbook on the Physics and Chemistryof Actinides.Elsevier Science Publishers, Amsterdam, pp. 413–455.
Kohler, M., Wieland, E., Leckie, J.O., 1992. Metal-ligand-surface interactions during sorp-tion of uranyl and neptunyl on oxides silicates. In: Kharaka, Y.K., Maest, A.S. (Eds.),Water Rock Interaction, VII. Balkema, Rotterdam, pp. 51–54.
Korichi, S., Bensmaili, A., 2009. Sorption of uranium(VI) on homoionic sodiumsmectite ex-perimental study and surface complexationmodeling. J. Hazard.Mater. 169, 780–793.
Kowal-Fouchard, A., Drot, R., Simoni, E., Ehrhardt, J.J., 2004. Use of spectroscopic tech-niques for uranium(VI)/montmorillonite interaction modeling. Environ. Sci. Technol.38, 1399–1407.
Křepelová, A., Sachs, S., Bernhard, G., 2006. Uranium(VI) sorption onto kaolinite inthepresence and absence of humic acid. Radiochim. Acta 94, 825–833.
Křepelová, A., Reich, T., Sachs, S., Dreber, J., Bernhard, G., 2008. Structural characterizationof U(VI) surface complexes on kaolinite in the presence of humic acid using EXAFSspectroscopy. J. Colloid Interface Sci. 319, 40–47.
Lakowicz, J.R., 1999. Principles of Fluorescence Spectroscopy, second ed. Plenum Press,New York, p. 243.
tion model of uranyl sorption on Georgia kaolinite. Appl. Clay Sci. 26, 151–162.Pekala, M., Jan, D.K.A., Waber, H.N., Gimmi, T., Alt-Epping, P., 2009. Transport of U-234 in
the Opalinus clay on centimeter to decimeter scales. Appl. Geochem. 24, 138–152.Sachs, S., Bernhard, G., 2008. Sorption of U(VI) onto an artificial humic substance-
kaolinite-associate. Chemosphere 72, 1441–1447.Samadfam, M., Jintoku, T., Sato, S., Ohashi, H., Mitsugashira, T., Hara, M., Suzuki, Y., 2000.
Effects of humic acid on the sorption of Am(III) and Cm(III) on kaolinite. Radiochim.Acta 88, 717–721.
Schlegel, M.L., Descontes, M., 2009. Uranium uptake by hectorite and montmorillonite: asolution chemistry and polarized EXAFS study. Environ. Sci. Technol. 43, 8593–8598.
Sladkov, V., Fourest, B., Mercier, F., 2009. Uranyl-Se(IV) interaction in aqueous acidsolutionsstudied by time-resolved laser-induced fluorescence spectroscopy(TRLFS) and UV–visspectrophotometry. Dalton Trans. 7, 7734–7740.
Stumm, W., 1992. Chemistry of solid – water interface. Willey Interscience, New York.Tsunashima, A., Brindley, G.W., Bastovanov, M., 1981. Adsorption of uranium fromsolutions
by montmorillonite; compositions and properties of uranylmontmorillonites. ClaysClay Miner. 29, 10–16.
Van Olphen, H., 1977. An introduction to clay colloid chemistry: For clay technologists,geologists, and soil scientists. Wiley, NY, p. 317.
Weber, T.W., Chakkravorti, R.K., 1974. Pore and solid diffusion models for fixed-bedadsorbers. Am. Inst. Chem. J. 20, 228.
