Physicochemical study of novel organoclays as heavy metal ion adsorbents for environmental remediation

Journal of Colloid and Interface Science 316 (2007) 298–309www.elsevier.com/locate/jcis

Physicochemical study of novel organoclays as heavy metal ion adsorbentsfor environmental remediation

Panagiota Stathi a, Kiriaki Litina b, Dimitrios Gournis b, Thomas S. Giannopoulos c,Yiannis Deligiannakis a,∗

a Department of Environmental and Natural Resources Management, University of Ioannina, Seferi 2, 30100 Agrinio, Greeceb Department of Materials Science & Engineering, University of Ioannina, 45110 Ioannina, Greece

c Department of Biological Applications and Technologies, University of Ioannina, 45110 Ioannina, Greece

Received 13 June 2007; accepted 31 July 2007

Available online 6 August 2007

Abstract

Four organic-modified clays based on a SWy-2 montmorillonite were prepared by embedding ammonium organic derivatives with differentchelating functionalities (–NH2, –COOH, –SH or –CS2) in the interlayer space of montmorillonite. Organic molecules such as (a) hexameth-ylenediamine, (b) 2-(dimethylamino)ethenethiol, (c) 5-aminovaleric acid and (d) hexamethylenediamine-dithiocarbamate were used for the claymodification in order to study the effect of the chelating functionality on heavy metal ions binding from aqueous solutions. The organoclayswere characterized by powder X-ray diffraction (XRD), infrared (FTIR) and NMR spectroscopies. The experimental data showed that the organicmolecules are intercalated into the interlamelar space with the long dimension parallel to the clay sheets. Their sorbing properties were evaluatedfor the removal of heavy metals, Pb, Cd and Zn, from aqueous solutions as a function of the pH. When compared with the unmodified SWy-2montmorillonite, the modified clays show significant improvement in terms of sorbing selectivity as well as of metal loading capacity. The fit toadsorption data by a Surface Complexation Model shows that the intercalated molecules act as specific binding sites in the clay. These contributeadditional sorption capacity which is additive to the variable charge edge-sites of the clay in competition with the permanent charge sites.© 2007 Elsevier Inc. All rights reserved.

Keywords: Organoclay; Montmorillonite; SWy; Heavy metal; Environmental remediation; SCM; Surface charge; Cd; Pb; Zn; FITEQL

1. Introduction

The presence of heavy metals in the environment is an is-sue of great concern because of growing discharge, toxicity,and other adverse effects of heavy metals on the receiving wa-ters and/or soils. Heavy metals are introduced in the environ-ment through natural phenomena or human activities, such asagricultural practices, transport, industrial activities and wastedisposal. Therefore, removal of heavy metals is a priority in en-vironmental remediation and clean up. Soluble metal speciescan be removed by adsorption on chemically functionalizedinorganic supports. In this context naturally occurring clay min-erals, may serve as cost effective sorbents, for the treatment of

* Corresponding author.E-mail address: [email protected] (Y. Deligiannakis).

heavy metals. Moreover, due to their abundance in soil sys-tems, their high specific surface area and exchange capacity,clay minerals such as montmorillonite play a significant role indetermining the availability and transport of metals species insoil and sediments [1–4].

Adsorption of metal ions M2+ onto montmorillonite clayusually involves two distinct mechanisms. (i) an ion exchangereaction at permanent-charge sites and, (ii) formation of com-plexes with surface hydroxyl groups at edge-sites [2–5]. Theadsorption capacity of clay minerals can be enhanced by re-placing the natural exchangeable cations with organic mole-cules forming the so called “organoclays” [6]. This processcan render the clay surface more hydrophobic or hydrophilic,at will, depending on the nature of the organic molecule [6].Thus a strategy towards the improvement of the sorbing capac-ity of clays is the incorporation of organic functionalities withbearing high affinity and/or selectivity for certain heavy metal

0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.jcis.2007.07.078

P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309 299

cations [6]. In this context, Mercier and Detellier [7], Celis etal. [8] and Lagadic et al. [9] studied the adsorption of Hg2+,Pb2+ and Cd2+ on thiol-modified montmorillonite. Pyrophyl-lite modified with amino groups, was shown to have improvedsorbing capacity for Pb2+ [10]. In other works, functionaliza-tion of clays with organosilanes was achieved and evaluatedfor Hg2+ uptake [11,12]. Sheng et al. [13] have prepared acarboxy-modified clay and studied the adsorption of Pb andchlorobenzene. Recently modification of montmorillonites withvarious organic molecules like dithiocarbamate [14], L-carni-tine, L-cysteine ethyl ester, L-cysteine dimethyl ester were re-ported [15].

In the present work, four novel organo-modified montmo-rillonites were prepared and tested as sorbent materials for theremoval of heavy metals from aqueous solutions. The modifi-cation is based on the functionalization of the clay surfaces byvarious chelating groups (–NH2, –COOH, –SH or –CS2) thatcan effectively capture metal ions. The prepared organoclayswere evaluated for the ability to remove Pb, Cd, and Zn ionsfrom aqueous solution.

The aim of this work was (a) to develop novel organoclaysbearing various chelating functional groups –SH, –COO−,–DTC and –NH2, and thus to examine the effect of these func-tional groups on the heavy metal ions binding ability of the finalsorbent materials, (b) to study in detail the physicochemicalmechanism of the metal adsorption, and the physicochemicalbasis of the achieved improved metal uptake capacity.

2. Materials and methods

2.1. Reagents

All solutions were prepared with analytical grade chemicalsand ultra pure water (Milli-Q Academic system) with a conduc-tivity of demineralized water 18.2 µS cm−1. All solution weredegassed with 99.999% N2 prior to use.

The organic compounds used for the synthesis of the organ-oclays were: hexamethylenediamine H2NCH2(CH2)4CH2NH2(DA), 2-(dimethylamino)ethanethiol hydrochloride (CH3)2–NCH2CH2SH·HCl (AT) and 5-aminovaleric acid H2N(CH2)4–COOH (AA) supplied from Aldrich, and hexamethylenedia-mine-dithiocarbamate synthesized in our laboratory.

2.2. Host layered material

The clay used in this work was a natural Wyoming mont-morillonite (SWy-2) obtained from the Source Clay MineralsRepository, University of Missouri, Columbia, with a cation-exchange capacity (CEC) of 76.4 meq/100 g clay. The physic-ochemical characteristics of the SWy-clay used in the modelingare listed in Table 1. The clay was fractionated to <2 µm bygravity sedimentation and purified by standard methods in clayscience [16]. Sodium-exchanged samples were prepared by im-mersing the clay into 1 M solution of sodium chloride. Cationexchange was complete by washing and centrifuging four timeswith dilute solution of NaCl. The samples were finally washedwith distilled–deionized water, transferred into dialysis tubes in

Table 1Physicochemical parameters of SWy-2 clay used in this study

CEC 76.4 meq/100 ga

PZNC 8.35a, 8.23a*, 8.435b

K+int 8.75a, 8.16b

K−int 7.95a, 8.71b

Specific surface area 294 m2/100 gc

Partice size 200 nmd

a This work, using SCM and Eq. (2).a* This work, using formula of Kraepiel et al. [20].b Refs. [27,28].c Ref. [1].d Ref. [20].

order to obtain chloride free clays and then dried at room tem-perature.

2.3. Synthesis of sodiumhexamethylenediamine-dithiocarbamate,H2NCH2(CH2)4CH2NHCS2 (DTC)

A solution of carbon disulfide (prepared from 1 g CS2 and5 ml 1-propanol) was added slowly to a stirred solution of 2 gof hexamethylenediamine in 20 ml of 1-propanol under lowtemperature (0–4 ◦C). The mixture was stirred for 30 min andthe precipitating white solid, was filtered off, washed with di-ethylether and dried at room temperature. The isolated solidwas then diluted in ethanol and an excess of aqueous NaOHwas added. The sodium salt of the hexamethylenediamine-dithiocarbamate was precipitated, filtered off, washed withethanol and dried at room temperature. The product was iden-tified by 1H NMR and 13C NMR: 1H NMR (D2O), δ (ppm):3.55 (2H, t, 3–CH2), 2.98–2.90 (5H, m, 4,8–CH2, NH), 1.67(2H, t, NH2), 1.43 (6H, t, 5,6,7–CH2) and 13C NMR (D2O),δ (ppm): 209.72 (C–S), 47.76 (CH2–NH), 39.54 (CH2–NH2),27.93 (4–CH2), 27.36 (6,7–CH2), 25.47 (5–CH2) (see support-ing information).

2.4. Preparation of organoclays

Organoclays were prepared by reacting a 1 wt% aque-ous clay suspension with aliquots of aqueous solutions of theabove organic compounds such that the ratios R = [organiccompound]/[clay] were 1.2 [w/w]. After stirring for 3 h theorganoclay aggregates were precipitated by centrifugation,washed three times with water and dried at room tempera-ture by spreading on glass plates. The products are designated:SWy/DA, SWy/AT, SWy/AA and SWy/DTC.

2.5. Characterization of organoclays

The powder X-ray diffraction (XRD) patterns were col-lected on a D8 Advanced Bruker diffractometer by using CuKα

(40 kV, 40 mA) radiation and a secondary beam graphite mono-chromator. The patterns were recorded in a 2-theta (2θ ) rangefrom 2 to 40◦, in steps of 0.02◦ and counting time 2 s per step.Samples were in the form of films supported on glass substrates.

300 P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309

Infrared spectra were measured with a Shimadzu FT-IR 8400spectrometer, in the region of 400–4000 cm−1, equipped with adeuterated triglycine sulfate (DTGS) detector. Each spectrumwas the average of 64 scans collected at 2 cm−1 resolution.Samples were in the form of KBr pellets containing ca. 2 wt%sample. 1H NMR and 13C NMR, for the DTC sample, wererecorded with Bruker AC at 400 MHz.

2.6. Analytical determination of metals (Pb, Cd, Zn)

The concentrations of Pb, Cd and Zn in the aqueous phasewere determined by Anodic Stripping Voltammetry by using aTrace Master5-MD150 polarograph by Radiometer Analytica,as described earlier [17]. The measuring cells were borosili-cate glass cell from Radiometer. The working electrode wasa hanging mercury drop electrode (HMDE), with a Hg dropwith 0.4 mm diameter generated by a 70 µm capillary. Thereference electrode was an Ag/AgCl electrode with a doubleliquid junction. The counter electrode was a Pt electrode. Ini-tially, before the stripping step N2 gas (99.999% purity) waspassed from the measuring solutions to remove any trace O2.During this step the solution was under continuous stirring at525 rpm. During the stripping step the solution was not stirred.Square wave (SW) measurements were performed in the anodicdirection, i.e. SWASV, to quantify metal ions. Typically un-der our experimental conditions, 10−6 M Pb(NO3)2 in 0.01 MKNO3 resulted in a current of Ip = 0.7 µA, Ep = −340 mV atpH 5.4, 10−6 M Cd(NO3)2 in 0.01 M KNO3 resulted in a cur-rent of Ip = 0.980 µA, Ep = −460 mV at pH 5.4 and 10−6 MZn(NO3)2 in 0.01 M KNO3 resulted in a current of Ip = 1.1 µA,Ep = −980 mV at pH 5.4.

2.7. Surface charge properties of clays

The surface charge properties of montmorillonite (SWy-2)clay suspensions were evaluated by two methods, i.e. potentio-metric acid–base titration [18–21] and mass-titration [22–24] asdescribed earlier [17].

2.7.1. Potentiometric titrationAcid–base potentiometric titration was used to measure the

surface proton adsorption. 12.5 mg of clay were suspended ina titration cell containing 12.5 ml of Milli-Q water to yield aconcentration 1 g/l. The suspension was allowed to equilibrate(swelling) for 12 h under continuous stirring. In the followingthe suspension was purged with pure nitrogen gas for 30 minprior to titration and divided into two equal portions, i.e. one forthe alkalimetric and the acidimetric titration respectively [18].The alkalimetric titration was done with 12 mM NaOH in thepH range 9.3–11.00. The acidimetric titration was done with12 mM HNO3 solution in the pH range 9.3–3.00. In all titrationsthe Metrohm 794 Basic Titrino burette was used and the pH wasmeasured with a Metrohm Pt-glass electrode (type 6.0239.100).

2.7.2. Mass titrationThe mass titration method [22] was used to determine the

point of zero net proton charge (pznpc) of the clay [21]. 5 mg

of clay were added to 5 ml Milli-Q water having a pH between3.50 and 9.40. The initial pH of the solution was adjusted withHNO3 or NaOH. After each addition of 5 mg of solid claythe pH was measured with a Metrohm Pt-glass micro-electrode(type 6.0222.100). The suspension was continuously stirred andpurged by nitrogen gas. When equilibration achieved (equili-bration time ca. 15 min) a new amount of clay was added.This procedure was continued until further clay additions didnot change the pH of the solution. For a clay or oxide free ofcontamination [24] this pH value has been shown to be a goodapproximation to the PZNC of oxide and clay surfaces.

In our case the mass-titration method was used in combina-tion with the acid–base titration for an initial estimation of thepoint of zero charge of the clay. This information was used forthe detailed fit of the acid–base titration, which allowed a moreprecise estimate of the surface proton binding constants, K+

int,K−

int [17].

2.8. Metal sorption experiments

3 mM Pb(NO3)2, Cd(NO3)2 and Zn(NO3)2 (Aldrich,>99.5%) stock solutions were prepared in a polyethylene con-tainers at pH <2 adjusted with HNO3. The ionic strength wasadjusted at 1 mM with KNO3 (Aldrich, >99.9%).

For the pH-edge experiments, a buffer system of 10 mMMES: (N -morpholino-ethanesulfonic acid), HEPES: (4-(2-hydroxyethyl)piperazine-1-ethanesulfonic) acid was used at allof the samples. This system presented a significant buffer ca-pacity at range pH 5–8.5 with an average deviation from theadjusted pH value <5%. Screening experiments indicated thatunder the conditions of our experiments the buffer moleculescaused no interferences on the adsorption phenomena. The pHvalues were adjusted with small volumes of NaOH (0.1 N) orHNO3 (0.05 N).

Metal ions adsorption was investigate in batch experiments.Sorption edge measured the effect of the pH on the metal uptakefrom solution for an initial concentration of 4.5 µM of metalions. 1 mg of clay was suspended in polypropylene tube con-taining 10 ml of buffer solution to yield a concentration 0.1 g/l.The suspension was allowed to equilibrate (swelling) for 12 hwith continuous stirring. In the following, suitable volumes ofmetals stock solutions were added to yield metal concentration4.5 µM. The pH of the suspension was then adjusted to thedesired value with small volumes of HNO3 or NaOH. Subse-quently the samples were allowed to equilibrate in room tem-perature for 2 h under stirring. The pH of each suspension wasmeasured at the end of the incubation process. Typically, devi-ations between initial and final pH were less than 0.1 pH units.Finally each suspension was centrifuged and the supernatantsolution was analyzed for metals.

2.9. Theoretical analysis: surface complexation modeling

The results of all potentiometric titration and sorption exper-iment were modeled with a Surface Complexation Model [25].Surface complexation models (SCM) can successfully describe

P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309 301

adsorption of ions on charged surfaces by assuming that ad-sorption involves both a coordination reaction at specific sur-face sites and an electrostatic interaction between adsorbingions and the charged surface [2–5]. In recent literature, SCMhave been used extensively to describe the adsorption of var-ious ionic species on charged surfaces [25] including clays[2–5,17,20,21]. For the unmodified montmorillonite, in ourmodeling we have assumed two different populations of sur-face reactive sites [5,21,26–28]:

(a) ≡SOH sites represent amphoteric aluminol and silanolgroups on the mineral edge [27,28]. Proton exchange atthese sites is modeled with a following reactions:

(1a)≡SOH + H+ K+int−→ ≡SOH+

2 ,

(1b)≡SOHK−

int−→ ≡SO− + H+.

The derived values of the of intrinsic equilibrium constantsK+

int, K−int for protonation and deprotonation of the surface

groups are related with the value of PZNC according to therelation

(2)pHPZNC = 1

2

(∣∣pK+int

∣∣ + ∣∣pK−int

∣∣).

The equilibrium constants of H+ reacting with the surfacesites are expressed in the form

(3)K+int = [≡SOH+

2 ][≡SOH](H+)

e(FΨ0)/RT ,

(4)K−int = [≡SO−](H+)

[≡SOH] e(FΨ0)/RT .

(b) Permanent charge, cation exchange sites (≡X−) were alsotaken into account [17,21]. Initial estimate for the surfacesites densities of ≡X− and ≡SOH were obtained by opti-mizing the fit for the potentiometric titration curve. Thismodeling approach was adapted from previous success-ful applications for adsorption of metals and organics onkaolinite [26] and montmorillonite [2,5,21] and recently forLaponite [17]. We underline that in the present case inclu-sion of permanent charge cation exchange sites (≡X−) forthe unmodified montmorillonite had a decisive effect of theadsorption data at acidic pH, typical for montmorillonite[2–4]. (c) For the modified montmorillonite the model in-cluded additional protonation–deprotonation reactions forthe intercalated organics. The metal uptake data were mod-eled by assuming binding of metal (i) at the variable chargesites, (ii) at the permanent charge sites. However accord-ing to the fit, in the modified clays these sites play littlerole since they are occupied by the R–NH+

3 groups. (iii) Atthe intercalated organics which were treated as being partof the solid matrix. The assumed reactions are detailedin Table 3. In each case, the adsorption of metal cations’M2+ and their first hydroxide’s M(OH)+ were taken intoaccount, see Table 3. At the metal concentrations of our ex-periments i.e. few micromolars, formation of polynuclearspecies can be ignored based on the standard hydrolysis

properties of Cd, Pb and Zn [29]. The presented theoreticalcalculations were preformed using the program FITEQL[30] assuming a diffuse layer model.

3. Results

3.1. Structural characterization of organoclays

It is well-known that positively charged molecules can beintroduced into interlayer space of layered clay minerals by ionexchange procedure. Charge balancing cations, such as Na+,are replaced by the organic cations. This is a convenient methodfor the intercalation of organic cations soluble in water intothe lamellar space of clay minerals. The newly introduced or-ganic cations are held strongly by electrostatic forces with thenegatively charged clay surfaces and by van der Waals interac-tions between the alkyl chains of the organic molecules and thesiloxane surface [31,32]. The final conformation depends on theshape, size, and total charge of the organic cations and also onthe charge density of the clay surface. The introduction of or-ganic cations increases the organophilicity of the clay mineralsand produces organoclays.

In our case, the intercalation procedure involves an initialpre-equilibrium reaction in which the amino groups of the or-ganic molecules (in our case DA, AA and DTC) are protonated:

R–NH2 + H2O � R–NH+3 + OH−, (5)

where R: –(CH2)6NH2, –(CH2)4COOH or –(CH2)6NHCS2,respectively, while AT is already protonated. The protonatedmolecules were readily adsorbed on the clay surfaces, by ionexchange. The monovalent Na+ exchangeable cations are re-placed easily by the protonated amino molecules according tothe reaction:

X−Na+ + R–NH+3 � X−(R–NH+

3 ) + Na+,

where the bold lines represent the clay platelets.Quantitative determination of the R–NH+

3 remaining in so-lution after adsorption, shows that >96% of the Na+ in thepermanent charge sites was exchanged by R–NH+

3 .X-ray diffraction (XRD) measurements provide a powerful

tool to understand the changes in the interior of the clay mi-croenvironment since the interlayer distance can be estimatedby measuring the d001 spacing. The XRD data, Fig. 1, showan increase of the basal spacing (d001) of the clay after inser-tion of the four guest materials. More specifically, in the case ofSWy/DA, the basal spacing d001, which is 12.4 Å in the pristineclay, becomes 13 Å in the modified clay, which correspondsto an intersheet separation of 13 − 9.6 = 3.4 Å, where 9.6 Åis the thickness of the clay layer [33]. This is indicative of theintercalation of the hexamethylenediamine into the clay inter-layers. The spacing of 3.4 Å implies that the hexamethylene-diamine molecules must be lying almost parallel to the layers.The XRD results for the other organoclays are analogous. XRDpattern of SWy/AT shows a d001 spacing of 13.4 Å, which cor-responds to an intersheet separation of 3.8 Å. In the case ofSWy/AA, the d001 spacing is 12.6 Å and the intersheet sepa-ration is 3 Å. Finally when the clay treated with DTC solution

302 P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309

Fig. 1. XRD patterns of (a) SWy and (b) SWy/DA, (c) SWy/AT, (d) SWy/AA,(e) SWy/DTC composites.

Table 2d001 values and basal spacing of organoclays

Sample d001(Å)

Interlayer space (Å)[d001 (Å)–9.6 (Å)]

SWy-2 9.6 –SWy/DA 13.0 3.4SWy/AT 13.4 3.8SWy/AA 12.6 3.0SWy/DTC 12.8 3.2

(SWy/DTC) the d001 spacing shifted to 12.8 Å, which corre-sponds to an intersheet separation of 3.2 Å. For all the abovecases d-spacing is less than 4 Å indicating almost horizontalarrangement (θ < 10◦) of the organic cations in the interlayerspace of the clay mineral. The results are summarized in Ta-ble 2.

The FTIR spectra of the organoclays present the characteris-tic bands of the host material plus those of the intercalated sub-stances, without significant changes, confirming the presenceof the organic species in the clay mineral. In Fig. 2A are presentthe spectra for the SWy/DTC organoclay, the host materialand the pure DTC. The presence of intercalated DTC speciesaffects the high frequency absorption spectrum of SWy between2500 and 3700 cm−1. The weak band at 3195 cm−1 is due toNH stretching vibrations. In addition asymmetric and symmet-ric stretching vibrations of CH2 groups of DTC salt are ob-

served at 2932 and 2862 cm−1. At 3448 and 1628 cm−1 appearsthe adsorbed H2O deformation. Bands at 1044 and 468 cm−1

originate from the clay lattice Si–O and Si–O–Si vibrations. At1088 cm−1 the absorption band of CS2 groups was expected[34–37]. This band is not masked due to overlap by strongerbands of the clay. However, a slight splitting of the peak re-veals its presence. In the medium frequencies, the spectrum ofSWy/DTC exhibited bands at 1554, 1489 and 1370 cm−1. Theband at 1554 cm−1 has been attributed deformation vibration ofNH+

3 groups while the other bands at 1489 and 1370 cm−1 aredue to –CH2– and –NH– bending vibrations, respectively.

The infrared spectra of the three organoclays (SWy/DA,SWy/AT, SWy/AA) and the pristine layered clay mineral areshown in Fig. 2B. The spectra present the characteristic bandsof the organic cations and the clay, without significant changes,confirming the presence of the organic molecules in the claymineral. More specifically, in the case of SWy/DA, peaks areobserved at 3628 cm−1 (clay lattice –OH stretching vibra-tions), 3439 and 1635 cm−1 (adsorbed H2O deformation), 2936and 2862 cm−1 (–CH2, –CH3 stretching vibrations of diamineorganic chain), 1563 cm−1 (from asymmetric deformation ofNH+

3 group) [31], 1333 and 1387 cm−1 (–CH2– bending vibra-tions), and 1047 and 468 cm−1 (clay lattice Si–O and Si–O–Sivibrations). Similar spectrum is observed for the SWy/AT,bands at 3628 cm−1 (clay lattice –OH stretching vibrations),3439 and 1635 cm−1 (adsorbed H2O deformation), weak bandsat 2928 and 2854 cm−1 (–CH2, –CH3 stretching vibrations),and 1047 and 468 cm−1 (clay lattice Si–O and Si–O–Si vibra-tions). The band at 1473 cm−1 arises from symmetric N+–CH3deformation [36,37] while the weaker band at 1391 cm−1 isattributed to –CH2– and/or –CH3– vibrations. Finally, for thesample SWy/AA analogous bands at 3624, 3446, 2965, 2859,1630, 1083, 1051 and 467 cm−1 are observed, in conjunctionwith peak at 1525 cm−1 due to NH+

3 vibrations, while the peakat 1435 cm−1 reveals from –CH2– scissoring vibrations. Theband at 1710 cm−1 originates from carboxyl groups.

3.2. Potentiometric titrations—surface charge of SWy-2

The mass-titration data, at zero ionic strength, for SWy-2are shown in Fig. 3A. In unbuffered solution the pH gradu-ally changes with the addition of solid and asymptotically ap-proaches a limiting value. The direction of the pH variation

Fig. 2. (A) FT-IR spectra of (a) SWy, (b) DTC and (c) SWy/DTC organoclay. (B) FT-IR spectra of (a) SWy and (b) SWy/DA, (c) SWy/AT, (d) SWy/AA organoclays.

P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309 303

Table 3Equilibrium equations and optimized constants of reactions

Reaction logKa Reference

Solution reactionsDissociation of water

H2O ↔ H+ + OH− 14.0Hydrolysis of metals

Cd2+ ↔ Cd(OH)+ + H+ 10.0 [29]Pb2+ ↔ Pb(OH)+ + H+ 7.70 [29]Zn2+ ↔ Zn(OH)+ + H+ 9.80 [29]

Surface reactionsProtonation of SWy-2 clay

SOH ↔ SO− + H+ −7.95 This work, [20]SOH + H+ ↔ SOH+

2 8.75 This work, [20]

Sorption of metals onto SWy-2 clay2(≡X) + Cd2+ ↔ [≡X2Cd] 2.44 This work2(≡X) + Pb2+ ↔ [≡X2Pb] 1.56 This work2(≡X) + Zn2+ ↔ [≡X2Zn] 3.90 This work≡SO− + Cd2+ ↔ [≡SOCd]+ 3.80 This work≡SO− + Pb2+ ↔ [≡SOPb]+ 4.50 This work≡SO− + Zn2+ ↔ [≡SOZn]+ 2.90 This work

Protonation of DANH− + H+ ↔ NH2 −7.42 This work, [38]

Sorption of metals onto SWy-2-DA clayNH− + Cd2+ ↔ [NHCd]+ 2.48 This work, [38]NH− + Cd(OH)+ ↔ [NHCd(OH)] 5.14 This workNH− + Pb2+ ↔ [NHPb]+ 2.50 This work, [38]NH + Pb(OH)+ ↔ [NHPb(OH)] 4.70 This workNH + Zn2+ ↔ [DTC2Zn]+ 1.80 This workNH + Zn(OH)+ ↔ [NHZn(OH)] 5.24 This work

Protonation of ATT− + H+ ↔ T–H −8.20 This work

Sorption of metals onto SWy-2-AT clayT− + Cd2+ ↔ [TCd]+ 4.60 This workT− + Cd(OH)+ ↔ [TCd(OH)] 7.24 This workT− + Pb2+ ↔ [TPb]+ 2.30 This workT− + Pb(OH)+ ↔ [TPb(OH)] 4.70 This workT− + Zn2+ ↔ [TZn]+ 1.50 This workT− + Zn(OH)+ ↔ [TZn(OH)] 5.20 This work

Protonation of AACOO− + H+ ↔ COO–H −3.50 This work

Sorption of metals onto SWy-2-AA clayCOO− + Cd2+ ↔ [COOCd]+ 1.57 This workCOO− + Cd(OH)+ ↔ [COOCd(OH)] 5.31 This workCOO− + Pb2+ ↔ [COOPb]+ 2.17 This workCOO− + Pb(OH)+ ↔ [COOPb(OH)] 7.01 This workCOO− + Zn2+ ↔ [DTCPb]+ 2.68 This workCOO− + Zn(OH)+ ↔ [COOZn(OH)] 6.35 This work

Protonation of DTCDTC− + H+ ↔ DTC–H −3.30 This work

Sorption of metals onto SWy-2-DTC clayDTC− + Cd2+ ↔ [DTCCd]+ 2.18 This workDTC− + Cd(OH)+ ↔ DTC(Cd(OH)) 5.80 This workDTC− + Pb2+ ↔ [DTCPb]+ 2.30 This workDTC− + Pb(OH)+ ↔ DTC(Pb(OH)) 4.10 This workDTC− + Zn2+ ↔ [DTCZn]+ 1.72 This workDTC− + Zn(OH)+ ↔ DTC(Zn(OH)) 5.40 This work

a Errors: ±0.05.

Fig. 3. Surface charge properties for SWy-2. (A) Mass-titration. (B) Potentio-metric acid–base titration: (1) experimental data; (—) theoretical model basedon FITEQL by using the parameters listed in Table 3. (C) Theoretical speciationof surface species.

depends on the starting pH of the solution [17,21,22]. There-fore the pH were solid addition does not produce any change inthe pH of the initial solution can be estimated by extrapolationto infinite mass [22]. This value can be taken as a good approxi-mation of the PZNC [17,21,22]. For the SWy-2 montmorilloniteused in our study the data in Fig. 3 give PZNC = 8.3 ± 0.2.In the following this value of PZNC is used together withEq. (2) to consistently fit the acid–base potentiometric titra-tion data [17]. Fig. 3B shows potentiometric titration for SWy-2montmorillonite. The solid line represents the best fit of ex-perimental data, obtained by assuming the surface protonationreactions described in Table 3. In Fig. 3C we display a specia-tion scheme derived from the fit. According to this, at pH 3–4.5we observe a rapid loss of protons from the surface ≡SOH+

2groups attaining a plateau at pH 4–8, Fig. 3C. Then, the ra-

304 P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309

Table 4Amounts (mmol/(K g))a of heavy metals adsorbed by functionalized clays atpH 7

Sample Pb Zn Cd

SWy 120 80 53SWy/AT 405 280 160SWy/DTC 240 150 120SWy/AA 271 248 204SWy/DA 456 302 306

a Errors: ±3%.

bid loss of protons at pH >8 is due to deprotonation of surfacegroups with pK = 7.95 resulting in the formation of ≡SOH−surface species, see Fig. 3C. The derived protonation constantsK+

int = 8.75 and K−int = −7.95, listed in Table 3, according

to Eq. (2) results in a PZNC of 8.35 in consistency with themass-titration result. These values are close to those previouslyreported by Stadler and Schindler [27] and Goldberg [28] forSWy montmorillonite. However these values are different fromthe values reported in certain other works [1,3,5]. It appears thatthe acidity constants for montmorillonite are sample-sensitive,with values varying between −2.81 to 7.38 for K+

int and be-tween 1.29 and 9.09 for K−

int [1–5,27,28].

3.2.1. Permanent charge effectIt is well known that montmorillonite particles carry two

kinds of electrical charges a variable (pH dependent) charge re-sulting from proton adsorption/desorption reaction of surfacesites SOH and a structural negative charge from ≡X− sitesresulting from isomorphous substitution with in the clay struc-ture. In our SWy-2 clay, the ≡X− site concentration can beapproximated [5,20,21] by the CEC value (76.4 meq/100 g).Kraepiel and Morel [20] have pointed out the ≡X− sites cancreate a charge density which may influence the surface protonbinding constants, i.e. via the Boltzman term. In our case, thepermanent charge density is estimated

σ0 = F[X−]/s = 18.3 × 10−3 C/m2

for s = 294 m2/100 g [1].Including this σ0 value in the calculation of the PZNC using

the formula of Kraepiel and Morel [20] gives a point of zerocharge

pHPZNC = 8.23.

Thus the permanent charge has a limited, although nonzero ef-fect on the PZNC of SWy-2. The derived surface protonationparameters are summarized in Table 1.

3.3. Heavy metal uptake

Figs. 4–7 show the sorption of Pb, Cd and Zn by the fourmodified montmorillonites, as a function of pH. Pertinent metaluptake data by unmodified SWy-2 materials are also includedfor comparison.

From Fig. 4 we notice that the Pb-uptake by SWy-2-DA wassignificantly increased relative to the unmodified SWy-2, i.e.compare solid squares with solid circles in Fig. 4, top-left panel.

Analogous improved metal uptake was observed for all themodified clays. For comparison, in Table 4 we list the amountof each metal adsorbed per unit mass of modified vs unmodifiedclay at pH 7. More importantly, we observe that the modifica-tion of montmorillonite results in an enhancement of adsorptionat all pH values, compare Figs. 4–7. This phenomenon will befurther discussed in the following.

At pH 7 the most important enhancement is observed forSWy/DA and SWy/AT which attain a >350% enhancement forPb and Zn and a 600% enhancement for Cd, compared with theunmodified SWy-2 under our experimental conditions, Table 4.The enhancement for the other organoclays is also significant,see Table 4.

The reason for the improved adsorption properties can beanalysed by referring to the theoretical speciation scheme, de-rived from the fit to the data, see left panels in Figs. 4–7. Forcompleteness, in Fig. S8 (see supporting information) we showthe speciation for the Pb-adsorption on unmodified SWy. Thesolid symbols in Figs. 4–7 (left panels) represent experimen-tal data while the open symbols are theoretical data calculatedby FITEQL by assuming the surface reactions listed in Ta-ble 3. The equilibrium constants derived from the fit to thedata are also listed in Table 3. The speciation schemes derivedfrom the fit are shown at the right panel of each figure, respec-tively.

3.3.1. SWy/DA clayAccording to the speciation scheme, the enhanced uptake of

Pb by SWy/DA can be attributed to the binding of Pb2+ andPb(OH)+ by the amino groups of the intercalated DA mole-cules. At the top of the left panel in Fig. 4 based on the the-oretical analysis, we display a Schindler-type plot as a visualaid for the observed phenomena. The NH2–Pb species domi-nates at pH <7 while at pH >7 the NH2–Pb(OH)+ species isresponsible for the enhanced Pb-uptake, see Fig. 4 right panelfor Pb. In a similar manner the NH2 group of DA appears tobe the primary binding site for Cd and Zn. The detailed specia-tion scheme in Fig. 4 provides a comprehensive picture for theenhanced metal uptake by this material.

3.3.2. SWy/AT clayIn analogous manner the theoretical speciation scheme in

Fig. 5, shows that the SH group of SWy/AT has a particularaffinity for Pb and Zn. The affinity for Cd is about 3 times lowerthan for Pb or for Zn, see Fig. 5. This shows that this materialhas a selectivity for Pd and Zn over Cd ions.

3.3.3. SWy/AA clayThe carboxy group of SWy-AA appears to be suitable for Cd

binding, which at pH <7 might be 1.5 to 2 times higher than forPb or Zn, respectively, see Fig. 6. Finally SWy-AA appears tohave an enhanced affinity for Pb(OH) at alkaline pH, see Fig. 6,while the uptake of Cd is strongly enhanced at pH <7.

3.3.4. SWy/DTC claySWy-DTC shows improved Pb binding, which at pH <7

might be up to 2 times higher than for Cb or Zn, respectively,

P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309 305

Fig. 4. Adsorption of metals on SWy-2-DA. Left panels: (A) lead, (B) cadmium, (C) zinc; (2) experimental data, (1) theoretical predictions based on FITEQLby using the parameters listed in Table 3. (") Experimental data for metal uptake by unmodified SWy-2. Right panels: speciation analysis for adsorption to themodified clay according to FITEQL calculations.

see Fig. 7. Finally SWy-DTC appears to have an enhanced affin-ity for Pb(OH) at alkaline pH, see Fig. 7, while the uptake of Cdand Zn is rather low.

3.4. Metal uptake mechanism

The modified SWy-2 montmorillonites show significant im-provement for metal uptake compared with the unmodifiedclay. At pH 7 the most important enhancement is observed forSWy/DA and SWy/AT which attain a >350% enhancement forPb and Zn and a 600% enhancement for Cd, compared with theunmodified SWy-2.

The speciation schemes appear capable to explain in a con-sistent manner the observed metal uptake by the novel materi-als. According to the fit, in the modified clays the permanentcharge sites (≡X−) play little role, since all of them are essen-tially occupied by R–NH+

3 groups. The theoretical modelingput forward here implies that there is essentially a 1:1 replace-ment of interstitial Na+ by organic ligands. The few remainingpermanent charge sites ≡X−, appear to play only secondaryrole in metal uptake. In all modified clays studied here, themetal binding capacity of the variable charge sites (≡SOH)i.e. concentration and binding constants, appear to remain un-altered relative to the unmodified clay. According to the fit, and

306 P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309

Fig. 5. Adsorption of metals on SWy-2-AT. Left panels: (A) lead, (B) cadmium, (C) zinc; (2) experimental data, (1) theoretical predictions based on FITEQL byusing the parameters listed in Table 3. (") Experimental data for metal uptake by unmodified SWy-2. Right panels: speciation analysis for adsorption to the modifiedclay according to FITEQL calculations.

the derived speciation, the enhanced uptake by the intercalatedorganics, can be attributed to the strong binding of the M2+and especially the M(OH)+ molecules by the organics. In theunmodified clay the ≡X− can only exchange Na+ with M2+.The reason for the facile access and binding of the additionalM(OH)+ species in the modified clays might be attributed to theincreased interlayer spacing i.e. of the order of 3–4 Å, see Ta-ble 2, due to the intercalation process. Therefore the improvedmetal uptake may be attributed to three main physicochemicalmechanisms: (a) increased interlayer spacing occurs due to theintercalation of the organics, (b) this allows easier access of theM2+ and M(OH)+ species to the intercalated organics, (c) the

strong binding constants of the M(OH)+ species by the organ-ics contribute additional metal uptake capacity which is addedto the M2+ binding. All these beneficial factors contribute to theobserved enhanced binding capacity in addition to the bindingat the variable charge sites.

4. Conclusions

The modified SWy-2 montmorillonites show significant im-provement for metal uptake compared with the unmodifiedclay. At pH 7 the most important enhancement is observed forSWy/DA and SWy/AT which attain a >350% enhancement for

P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309 307

Fig. 6. Adsorption of metals on SWy-2-AA. Left panels: (A) lead, (B) cadmium, (C) zinc; (2) experimental data, (1) theoretical predictions based on FITEQLby using the parameters listed in Table 3. (") Experimental data for metal uptake by unmodified SWy-2. Right panels: speciation analysis for adsorption to themodified clay according to FITEQL calculations.

Pb and Zn and a 600% enhancement for Cd, compared withthe unmodified SWy-2. The intercalated functional groups ap-pear to play dominant role in the maximum uptake capacity aswell as in the selectivity. In all the modified clays the metal up-take occurs almost exclusively at the variable charge sites plusat the intercalated organics. In the modified clays the permanentcharge sites appear to play only secondary role. This may po-tentially attributed to the presence of the intercalated organicswhich compete very efficiently with the permanent charge sites.

The enhanced uptake of Pb by SWy/DA can be attributed tothe binding of Pb2+ and Pb(OH)+ by the amino groups of theintercalated DA molecules.

The SH group of SWy/AT has a particular affinity for Pb andZn. The affinity of SWy/AT for Cd is about 3 times lower thanfor Pb or for Zn. This shows that SWy/AT has selectivity for Pdand Zn over Cd ions.

SWy-AA appears to have an enhanced affinity for Pb(OH)at alkaline pH, while the uptake of Cd is strongly enhanced atpH <7. The carboxy group of SWy-AA appears to be suitablefor Cd binding, which at pH <7 might be 1.5–2 times higherthan for Pb of Zn, respectively.

SWy-DTC shows enhancement for Pb binding, which atpH <7 might be up to 2 times higher than for Cb or Zn re-spectively.

308 P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309

Fig. 7. Adsorption of metals on SWy-2-DTC. Left panels: (A) lead, (B) cadmium, (C) zinc; (2) experimental data, (1) theoretical predictions based on FITEQLby using the parameters listed in Table 3. (") Experimental data for metal uptake by unmodified SWy-2. Right panels: speciation analysis for adsorption to themodified clay according to FITEQL calculations.

Supporting information

The online version of this article contains additional sup-porting information: Figs. S1, S2, S3 13C-NMR, Figs. S4, S5,S6, S7 1H-NMR spectra of H2NCH2(CH2)4CH2NHCS2 (DTC)and Fig. S8 speciation analysis for the adsorption of Pb on un-modified SWy-2.

Please visit DOI: 10.1016/j.jcis.2007.07.078.

References

[1] F. Barbier, G. Duc, M. Petit-Raurel, Colloids Surf. A 166 (2000) 153.[2] A.M.L. Kraepiel, K. Keller, F.M.M. Morel, J. Colloid Interface Sci. 210

(1999) 43.

[3] M.H. Bradbury, B. Bayens, Geochim. Cosmochim. Acta 63 (1999) 325.[4] B. Bayens, M.H. Bradbury, J. Contam. Hydrol. 27 (1997) 199.[5] J. Ikhsan, J.D. Wells, B.B. Johnson, M.J. Angove, Colloids Surf. A 252

(2005) 33.[6] S. Yariv, H. Cross (Eds.), Organo-Clay Complexes and Interactions, Mar-

cel Dekker, New York, 2002.[7] L. Mercier, C. Detellier, Environ. Sci. Technol. 29 (1995) 1318.[8] R. Celis, M.C. Hermosin, J. Cornejo, Environ. Sci. Technol. 34 (2000)

4593.[9] I.L. Lagadic, M.K. Mitchell, B.D. Payne, Environ. Sci. Technol. 35 (2001)

984.[10] M. Erdemoglou, S. Erdemoglou, F. Sayisan, M. Akarsu, S. Sener, H. Say-

ilkan, Appl. Clay Sci. 27 (2004) 41.[11] A. Szabo, D. Gournis, M.A. Karakassides, D. Petridis, Chem. Mater. 10

(1998) 639.

P. Stathi et al. / Journal of Colloid and Interface Science 316 (2007) 298–309 309

[12] I.K. Tonle, E. Ngameni, D. Njopwouo, C. Carteret, A. Walcarius, Phys.Chem. Chem. Phys. 5 (2003) 4951.

[13] G. Sheng, S.H. Xu, S.A. Boyd, Soil Sci. Soc. Am. J. 63 (1999) 73.[14] R. Say, E. Birlik, A. Denizli, A. Ersoz, Appl. Clay Sci. 31 (2006) 298.[15] M. Cruz-Guzman, R. Celis, M.C. Hermosin, W.C. Koskinen, E.A. Nater,

J. Conejo, Soil Sci. Soc. Am. J. 70 (2006) 21.[16] R.D. King, D.G. Nocera, T.J. Pinnavaia, J. Electroanal. Chem. 43 (1987)

236.[17] E. Giannakopoulos, P. Stathi, K. Dimos, D. Gournis, Y. Sanakis, Y. Deli-

giannakis, Langmuir 22 (2006) 6863.[18] C.P. Schulthess, D.L. Sparks, Soil Sci. Soc. Am. J. 50 (1986) 1406.[19] H. Wanner, Y. Albinsson, O. Karnlald, E. Wieland, P. Wersin, L. Charlet,

Radiochim. Acta 66/67 (1994) 157.[20] A.M.L. Kraepiel, K. Keller, F.M.M. Morel, Environ. Sci. Technol. 32

(1998) 2829.[21] M.J. Avena, C. de Pauli, J. Colloid Interface Sci. 202 (1998) 195.[22] J.S. Noham, J.A. Schwarz, J. Colloid Interface Sci. 130 (1989) 157.[23] K. Bouricas, J. Vakros, C. Kordulis, A. Lycourgiotis, J. Phys. Chem. B 107

(2003) 9441.[24] S. Zalac, N. Kallay, J. Colloid Interface Sci. 149 (1992) 233.[25] D.A. Dzombak, F.M.M. Morel, Surface Complexation Modeling, Willey,

New York, 1990.[26] L. Benyahya, J.M. Garnier, Environ. Sci. Technol. 33 (1999) 1398.[27] M. Stadler, P.W. Schindler, Clay Clay Miner. 41 (1993) 288.

[28] S. Goldberg, in: E.A. Jenne (Ed.), Adsorption of Heavy Metals by Geo-media, Academic Press, San Diego, 1998.

[29] C.F. Baes, R.E. Mesmer, The Hydrolysis of Cations, R.E. Kriegen Pub-lishing Co., Malandar, 1986.

[30] A. Herbelin, J. Westall, FITEQL: A computer program for determinationof chemical equilibrium constant from experimental data, Version 4.0; Re-port 99-01, Department of Chemistry, Oregon State University, Corvallis,OR, 1999.

[31] D. Gournis, V. Georgakilas, M.A. Karakassides, T. Bakas, K. Kordatos,M. Prato, M. Fanti, F. Zerbetto, J. Am. Chem. Soc. 126 (2004) 8561.

[32] D. Gournis, L. Jankovic, E. Maccallini, D. Benne, P. Rudolf, J.F. Colomer,C. Sooambar, V. Georgakilas, M. Prato, M. Fanti, F. Zerbetto, G.H. Sarova,D.M. Guldi, J. Am. Chem. Soc. 128 (2006) 6154.

[33] B.K.G. Theng, The Chemistry of Clay–Organic Reactions, Wiley, NewYork, 1974.

[34] D. Gournis, Y. Deligiannakis, M.A. Karakassides, A. Boussac, N. Ioanni-dis, D. Petridis, Langmuir 18 (2002) 10024.

[35] C.T. Johnson, in: B.L. Sawhney (Ed.), CMS Workshop Lectures: OrganicPollutants in the Environment, vol. 8, Boulder, CO, 1996, pp. 1–44.

[36] G. Purna Chandra Rao, S. Satya Veni, M. Madhava Rao, K. Pratap, M.C.Wang, K. Seshaiah, Toxicol. Environ. Chem. 88 (2006) 65.

[37] D. Petridis, D. Gournis, M.A. Karakassides, Mol. Cryst. Liq. Cryst. 311(1998) 345.

[38] W. Stumm, J.J. Morgan, Aquatic Chemistry, Willey, New York, 1996.