Experimental Study And Mathematical Model On Remediation Of Cd Spiked Kaolinite By Electrokinetics

+Model EA-12008; No. of Pages 6

ARTICLE IN PRESS

Electrochimica Acta xxx (2006) xxx–xxx

Experimental study and mathematical model on remediation of Cd spiked kaolinite by electrokinetics Michele Mascia a,∗ , Simonetta Palmas a,1 , Anna Maria Polcaro a,1 , Annalisa Vacca a , Aldo Muntoni b b

a Dipartimento di Ingegneria Chimica e Materiali, Universit` a Degli Studi di Cagliari, Piazza d’Armi, 09123 Cagliari, Italy Dipartimento di Geoingegneria e Tecnologie Ambientali, Universit`a Degli Studi di Cagliari, Piazza d’Armi, 09123 Cagliari, Italy

Received 19 October 2005; received in revised form 28 March 2006; accepted 14 April 2006

Abstract An experimental study on electrokinetic removal of cadmium from kaolinitic clays is presented in this work, which is aimed to investigate the effect of surface reactions on the electrokinetic process. Enhanced electrokinetic tests were performed in which the pH of the compartments was controlled. Cadmium spiked kaolin was adopted in the experimental runs. On the basis of the experimental results, a numerical model was formulated to simulate the cadmium (Cd) transport under an electric field by combining a one-dimensional diffusion-advection model with a geochemical model: the combined model describes the contaminant transport driven by chemical and electrical gradients, as well as the effect of the surface reactions. The geochemical model utilized parameters derived from the literature, and it was validated by experimental data obtained by sorption and titration experiments. Electrokinetic tests were utilized to validate the results of the proposed model. A good prediction of the behaviour of the soil/cadmium ions system under electrical field was obtained: the differences between experimental and model predicted profiles for the species considered were less than 5% in all the examined conditions. © 2006 Elsevier Ltd. All rights reserved. Keywords: Electrokinetics; Cadmium transport; Surface reactions; Mathematical model; Geochemical model

1. Introduction Soil contamination by heavy metals is increasing in various sites including residential and industrial areas: evidences of soil pollution by heavy metal were found in Europe, United States and Asia [1]. This pollution may be caused by several hazardous materials which were improperly dumped, causing a dispersion of toxic substances in the environment; moreover, leaching of soils, and solid wastes by rainwater may contaminate surface and ground waters. Waste materials containing heavy metals may derive from several industrial processes: cadmium and lead are contained in the wastes deriving from electrowinning of zinc, as well as in those deriving from dressing of blende [2]. Remediation of soils polluted by heavy metal ions is an important issue of the environmental engineering, since

∗ 1

Corresponding author. Tel.: +39 070 6755059; fax: +39 070 6755067. E-mail address: [email protected] (M. Mascia). ISE member.

the wide variation in the characteristics of soils and pollutants, requires different methods to remediate contaminated soils. Among the techniques usually adopted, flushing [3] and phytoremediation [4] are highly effective, but their applicability is strongly dependent on the hydraulic permeability of the soil, as well as on the surface metal–soil interactions; thus the in situ remediation of clayey soils by means of these techniques may result too expensive, due to the low values of hydraulic permeability of these soils. Electroremediation is a potentially important technique for fine-grained soils: this technique has been investigated for consolidation and stabilisation of clayey materials since the late 1930s; more recently the use of electrokinetic remediation techniques was studied for decontamination of clayey soils polluted by heavy metals, both in situ and ex situ [5–8]. Electroremediation of polluted sites is based on the application of a potential difference or a low intensity direct current between two electrodes inserted in the soil [9]. If the contaminant species are charged, they move by ionic migration towards one of the

0013-4686/$ – see front matter © 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.electacta.2006.04.066

Please cite this article in press as: M. Mascia et al., Electrochim. Acta (2006), doi:10.1016/j.electacta.2006.04.066

+Model EA-12008;

No. of Pages 6

2

ARTICLE IN PRESS M. Mascia et al. / Electrochimica Acta xxx (2006) xxx–xxx

electrodes, depending on the sign of their electric charge, where they can be recovered [10]. As far as mobilization and removal of heavy metal ions from clayey soil is considered, the interaction between the pollutants and the soil surface mainly depends on the characteristics of the active sites at the soil surface. Different active sites lead to different retention mechanisms, thus an accurate representation of the soil surface is a relevant issue of the electrokinetic process. To this aim, a geochemical model, such as those usually adopted to describe the thermodynamics of the soil–cations interactions can be effective. As can be found in the literature, most of the geochemical models describe the interactions in terms of equilibrium reactions between the dissolved ions and two different charged sites at the soil surface [11,12]. The ion exchange reactions are attribute to fixed charge sites resulting form isomorphous substitutions on layer lattice, whereas surface complexation reactions are attribute to oxides (FeOOH, Al2 O3 , SiO2 , TiO2 ) and other variable charged sites at the edge of particles [13]. During electrokinetics process, the equilibria at the soil surface are modified by the acid front, generated by oxidation of water occurring at the anode, which advances towards the cathode, as well as by the basic front, due to the reduction of water at the cathode, which moves towards the anode. The basic front may cause precipitation of heavy metals as hydroxides, decreasing the effectiveness of the process. Moreover, a low electrical conductivity region can be originated where the migrating hydrogen and hydroxyl ions meet. In order to prevent these problems, several enhanced processes have been developed, which are based on the control of pH near the cathode by addition of acidic solutions [14,15] or on the use of ion exchange membranes to prevent penetration of OH in the soil [16]. Other enhancement techniques were presented [17,18], depending on the contaminant and the surface characteristics of the soil, and different versions of the process were proposed, such as cation selective membrane, ceramic casting, Lasagna, electrochemical ion exchange, electrokinetic bioremediation, electrochemical geooxidation, and electrosorb [19,20]. In any case, a mathematical representation of the behaviour of the soil–cation systems under electrical field is a useful tool to understand the process. In the literature, a great number of one-dimensional models have been proposed, in which the removal of different species was represented [1,5,21]. Mathematical modelling of elecrokinetic remediation, through twodimensional approximations, is also the subject of several recent papers [10]. In this work the results of an experimental study on electrokinetic removal cadmium from a sample of kaolinitic clay are presented; the experimental data were modeled by combining a one-dimensional electrokinetic transport model with a two sites geochemical model.

Table 1 Characteristics of the kaolinite utilized in the experimental study pH Total carbon Cation exchange capacity,CEC (cmol kg−1 )

5.2 0 9.9

Major cations (cmol kg−1) Na Ca K Mg

2.96 9.8 0.96 3.48

Erba), were used; the main characteristics are summarized in Table 1. The experimental apparatus utilized for the electrokinetic tests consists of four main parts: soil cell, electrode compartments, electrolyte solution reservoirs, and power supply. The cell (Fig. 1) measures 30 cm length and 6 cm inner diameter. The soil in the cell is separated from the electrode compartment by two porous stones. The anode is a titanium sheet covered by RuO2 whereas a carbon plate constituted the cathode. The electrode compartments containing 500 cm3 of electrolyte solution are equipped with vents for discharging the oxygen and hydrogen produced by the electrolysis of water at both electrodes. Two electrolyte reservoirs (500 cm3 ) were adopted: the electrolyte solutions were recirculated in both electrode compartments by peristaltic pumps. The polluted soil specimen was prepared by equilibrating for 24 h the kaolinite sample with aqueous solutions of cadmium nitrate (100 mg dm−3 of Cd2+ ) with a mechanic stirrer; the slurry was then placed in the cell and compacted for 120 h at 2 kg cm−2 before starting with the electrokinetic runs. The tests were realized applying a constant voltage equal to 50 V between the electrodes: the cathodic and anodic potential measured during the runs were −1.8 and 1.5 V versus SCE, respectively. The electroosmotic permeability coefficient keo was experimentally determined by measuring the flow rate of water through the soil sample under electric field at different values of electrolyte pH. Values of keo were calculated by linear regression of experimental data of flow rate versus potential: values of 0.17 × 10−10 (pH 3.5; R2 = 0.99) and 8.8 × 10−10 (pH 7; R2 = 0.98) were obtained, which are in agreement with other values reported in the literature [1,21]. The tortuosity of the specimen was obtained by breakthrough curves of Cl− ions, following the procedure suggested by Alshawabkeh et al. [22]: a value of 0.4 was obtained.

2. Experimental Electrokinetic experiments at constant potential were carried out in this work: samples of kaolinite clay (supplied by Carlo

Fig. 1. Electrokinetic cell.

Please cite this article in press as: M. Mascia et al., Electrochim. Acta (2006), doi:10.1016/j.electacta.2006.04.066

+Model EA-12008; No. of Pages 6

ARTICLE IN PRESS M. Mascia et al. / Electrochimica Acta xxx (2006) xxx–xxx

Enhanced electrokinetic tests were performed at constant potential gradient of 1.5 V cm−1 in which both electrolyte compartments initially contained 0.001 M nitric acid. The pH of the compartments were monitored during the runs, and if necessary, adjusted to value of 3–4 by adding appropriate amounts of nitric acid or NaOH, in order to avoid precipitation of cadmium hydroxides near the cathode or dissolution of the clay lattice near the anode. Samples of catholyte were withdrawn and analysed for pH and Cd2+ concentration (polarography, Metrohm instrument). At the end of each experiment the soil sample was extracted from the cell and divided into 10 layers. The layers were then divided in two samples to obtain pH value and metal concentration in liquid and soil phase. pH and liquid phase concentration of cadmium were obtained by suspending the samples in deionised water (1:10, w/w), and analysing the supernatant after centrifugation at 5000 rpm for 30 min. The total amount of cadmium was obtained by the same procedure using 1 M nitric acid. The solid phase concentration was obtained by difference (total-liquid). The electrode used as cathode was washed with 0.1 M nitric acid in order to quantify the cadmium, which was deposited for reduction at the electrode. All the electroremediation experiments were performed in triplicate: the samples were compacted with the same load at the same times, and connected in parallel to the power supply. A good repeatability of the data was obtained in all the experimental conditions investigated, all the values of pH and concentrations measured differed by less than 5%. Preliminary sorption and titration experiments were performed in order to investigate the interaction between cadmium ions and soil surface. Titration experiments on the soil samples were carried out by suspending 5 g of clay in 50 ml of 0.02 M NaClO4 in a mechanic stirrer. Aliquots of HNO3 or NaOH at concentration of 0.1 or 1 M were added to the clay suspensions to give an initial pH value between 2.7 and 11.7. The clay suspensions were shaken in closed polypropylene tubes: pH of the suspension was continuously monitored until change in time of pH was not observed (typically 24 h). In the sorption experiments, 5 g of clay were equilibrated with 50 ml of Cd(NO3 )2 at different initial concentration C0 in a mechanic stirrer; pH of the slurries was monitored and adjusted to pre-fixed values by adding proper amounts of HNO3 or NaOH. Equilibrium time of 48 h was determined by preliminary kinetic tests. The samples were then centrifuged at 5000 rpm for 30 min and decanted to separate liquid and solid phases. The liquid phase was analysed for the equilibrium ion concentrations Ce . The solid phase concentrations qi were calculated as: qi =

V 0 [C − Cie ]. m i

3

Fig. 2. Titration curve of the kaolinitic clay utilized in this work: experimental (symbols) and model predicted data (curves).

interactions between the surface of the examined clay and H+ ions were experimentally studied through titration experiments. Fig. 2 shows the experimental values (symbols), obtained by titration runs of the soil considered, i.e. the experimental values of pH consequent to the addition of acid (positive values) or base (negative values). In order to numerically interpret the data in Fig. 2, the equilibrium reactions between H+ ions and the variable charge sites may be considered, since the fraction of proton adsorbed due to ion exchange become important only at pH below 3 [23]. The protonation/deprotonation reactions of the variable charge sites may written as: SOH + H+ → SOH →

SOH2 +

SO− + H+

(2) (3)

where SOH represents the generic variable charge site, such as the surface hydroxyl groups deriving from edge sites of layered silicate minerals: according to the literature, they can be silanol groups (SiOH) or aluminol groups (AlOH), but it is likely that the groups involved in adsorption are mostly AlOH [24]. The fraction of CEC due to the fixed charge sites being α, we can write: SOHtot = (1−α)CEC = SOH + SOH2 + + SO−

(4)

where SOH2 + represents the concentration of H+ ions retained by the soil. By combining Eq. (4), with the equilibrium equations related to protonation and deprotonation of the sites (Eqs. (2) and (3)), we obtain: SOH2 + = qH+ =

(1 − α)CECK1HSOH [H+ ] 1 + K2HSOH /[H+ ] + K1HSOH [H+ ]

(5)

The curve in Fig. 2 was calculated by combining the mole balance in Eq. (1) with Eq. (5), in which the equilibrium constants were derived from the literature (see Table 2) [23], whereas the

(1)

3. Results and discussion In the first phase of this work, the interaction between the surface of the examined clay and the dissolved species were studied by means of titration experiments and adsorption isotherms. The

Table 2 Thermodynamic parameters for the soil–cation interactions log K1HSOH (Eq. (2)) log K2HSOH (Eq. (3)) log KCdSOH (Eq. (6)) log KCd/Ca (Eq. (7)) X− /CEC ratio (α)

Please cite this article in press as: M. Mascia et al., Electrochim. Acta (2006), doi:10.1016/j.electacta.2006.04.066

3.24 −7.15 −3.2 1.03 0.30

+Model EA-12008;

No. of Pages 6

4

ARTICLE IN PRESS M. Mascia et al. / Electrochimica Acta xxx (2006) xxx–xxx

parameter α was the only adjustable one. A value of 0.3 allowed the best agreement between experimental and calculated data. As far as the retention of cadmium ions is considered, Fig. 3 shows the sorption isotherms of Cd2+ ions onto kaolinitic clay: the trend of cadmium concentration in solid phase shows an initial linear increase, and it tends to an asymptotic value at higher values of the Cd2+ concentration in water; moreover, a strong dependence from the pH of the ions retained by the soil can be observed. In order to model the Cd2+ sorption on kaolinite, the additional complexation and exchange equilibria were utilized, with the following assumptions: all the indigenous cations of the kaolinite were considered as equivalent calcium; the complexation reactions of Ca2+ ions were neglected. SOCd+ H+

(6)

XCa + Cd2+ → XCd + Ca2+

(7)

SOH + Cd2+ →

 qCd2+ = CEC

reaction of these sites with the heavy metal ion (Eq. (6)): [SOCd+ ] =

KCdSOH 1 + K2HSOH /[H+ ] + K1HSOH [H+ ] ×

[SOHtot ] [Cd2+ ] [H+ ]

(8)

Moreover, since at low Cd2+ concentrations X2 Ca  X2 Cd, we can write: X2 Ca =

1 αCEC 2

(9)

And equilibrium (7) can be adopted to evaluate the amount of Cd+2 retained by X− sites: [X2 Cd] = KCd/Ca

[X2 Ca] [Cd2+ ] [Ca2+ ]

(10)

The total amount of adsorbed cadmium may be expressed as:  KCdSOH 1−α 1/2α + K (11) [Cd2+ ] Cd/Ca 1 + K2HSOH /[H+ ] + K1HSOH [H+ ] [H+ ] [Ca2+ ]

The equilibrium relations of reactions (6) and (7) can be numerically solved along with Eq. (5) and the corresponding mole balances (Eq. (1)) to model the sorption isotherms. In this work the model was solved by using the PHREEQC-code, widely adopted in the literature for geochemical applications [13]. The value of α obtained from titration curves was utilized, whereas the initial values of the equilibrium constants for the reactions (6) and (7) were adjusted to obtain the best agreement between experimental and model predicted data, starting from literature values [13]. Fig. 3 shows the curves calculated, compared with experimental data, whereas Table 2 reports the parameters of the model. The geochemical model equations can be simplified at low values of Cd2+ concentration, such as these considered in the electrokinetic tests of this work: considering that SOHtot  SOCd+ , Eq. (4) can be utilized as balance of SOH sites. The solid phase concentration of Cd at the SOH sites may be derived by combining Eq. (5) and the constant related to the

The straight lines depicted in Fig. 3 were calculated by Eq. (11) with the parameters reported in Table 2; as can be seen from the figure, a good agreement with the experimental data is obtained at low cadmium concentration. In the second phase of the work, the transport of cadmium ions under electrical field was experimentally investigated by enhanced electrokinetic tests. The transport of Cd2+ ions during the electrokinetic experiments was then modeled by a traditional advection–dispersion model. The model takes into account the two transport mechanisms basically involved in the electrokinetic phenomena, electromigration and electroosmosis; both mechanisms are affected by several parameters such as pH, ionic strength and electric properties of the soil. Electromigration is the flux of charged species towards the electrodes and it is a function of electrical potential gradient, effective ionic mobility, which can be theoretically estimated using the Nernst–Einstein relation. Jiem = u∗i ci ∇(−Φ) =

Di∗ zi F ci ∇(−Φ) RT

(12)

where R and T are the universal gas constant and the absolute temperature, respectively. The electroosmotic flow rate depends on the soil surface characteristics, in particular on the concentration of charges, both fixed and pH-dependent, and it is usually expressed as: Jw = −keo ∇Φ =

Fig. 3. Sorption isotherms of Cd2+ ions at different pH: symbols refer to experimental data, curves refer to model predicted data and straight lines refer to data calculated by Eq. (11). Empty symbols, pH 4.5; full symbols, pH 6.5.

εζn ∇Φ μ

(13)

The electroosmotic permeability coefficient depends on the zeta potential of the soil, on the viscosity of the fluid as well as on electrical permittivity and porosity of the soil. The total mass flux can be estimated by considering the mass transport, due to diffusion, electromigration and electroosmotic advection; for one-dimensional implementation the mass balance of the i-th

Please cite this article in press as: M. Mascia et al., Electrochim. Acta (2006), doi:10.1016/j.electacta.2006.04.066

+Model EA-12008; No. of Pages 6

ARTICLE IN PRESS M. Mascia et al. / Electrochimica Acta xxx (2006) xxx–xxx

species can be written as: ∂ci ∂Φ ∂ 2 ci ∂ci = Di∗ 2 + (u∗i + keo ) + Ri ∂t ∂x ∂x ∂x

(14)

Under the experimental conditions adopted in this work, the term Ri accounts for the sorption processes; since the sorption rate is large, instantaneous equilibrium can be assumed between species in liquid phase and sorbed onto the clay surface [22], and Ri can be expressed as: Ri = −

ρk ∂ci ∂qi ρk ∂qi =− ε ∂t ε ∂t ∂ci

by combining Eqs. (14) and (15) we obtain:   ρk ∂qi ∂ci ∂ 2 ci ∂ci ∂Φ = Di∗ 2 + (u∗i + keo ) 1+ ∂t ε ∂ci ∂x ∂x ∂x

(15)

(16)

Eqs. (5), (10) and (11), respectively were utilized to obtain the term ∂qi /∂ci for H+ , Ca2+ , and Cd2+ ions in Eq. (16), values of Di∗ of 20x10−10 [m2 s−1 ] (H+ ), and 7x10−10 [m2 s−1 ] (other cations) were used. Eq. (16) was integrated with a finite elements software; the boundary conditions were written considering the mass balances at the cathodic and anodic compartments: taking into account the electrolysis of water at the electrodes and the experimentally observed cathodic reduction of cadmium, assumed under mass transfer control, the mass balances for Cd2+ , H+ and other cations are the following. ⎧ ⎨ ∂cH = I − S (u∗ + k )c eo H ∂t FVa Va H (17a) x=0 ⎩ ci = 0 ⎧ ∂cH (1 − η)I S ∗ ⎪ ⎪ (uHi + keo )cH − = ⎪ ⎪ ∂t Vc FVc ⎪ ⎪ ⎨ ∂c S ∗ i x=L (17b) = (u + keo )ci ⎪ ∂t Vc i ⎪ ⎪ ⎪ S ∗ ∂c ⎪ ⎪ ⎩ Cd = (u + keo )cCd − a · Km CCd ∂t Vc Cd

5

cathodic compartment as a function of the treatment time. As can be observed, a good agreement between experimental and calculated data is obtained, confirming the validity of the hypotheses. In particular, the good prediction of the trend of cadmium concentration can be observed, which presents a maximum deriving from the competition between the flux of pollutant towards the cathode and its removal by reduction at the cathode surface. The model was also utilized to predict the space profiles along the electrokinetic column. Experimental and model predicted data of H+ and Cd2+ concentration profiles are presented in Fig. 5(a and b). Data in Fig. 5(a) were obtained after 240 h of treatment, whereas data in Fig. 5(b) were obtained after 350 h. The advance of acid front can be seen from the results: the simple enhancement strategy utilized, control of the cathodic pH, was able to maintain the pH values in the range from four to five, along the specimen, avoiding the detrimental effect of the pH jump evidenced by several authors during unenhanced tests [1]. As far as the cadmium concentration is considered, the removal of the pollutant can be observed: the control of cathodic pH avoided the precipitation

Fig. 4 shows the comparison of experimental and model predicted data for Cd2+ and other cations (as equivalent Ca2+ ) in the

Fig. 4. Trend in time of Cd2+ ions (full symbols) and other cations as equivalent calcium (empty symbols), in the cathodic compartment during electrokinetic test at ψ = 1.5 V cm−1 : experimental (symbols) and model predicted data (lines).

Fig. 5. Comparison between experimental (symbols) and model predicted (lines) profiles of pH and Cd2+ ions concentration after 240 h (5a) and 340 h (5b) of electrokinetic treatment at ψ = 1.5 V cm−1 .

Please cite this article in press as: M. Mascia et al., Electrochim. Acta (2006), doi:10.1016/j.electacta.2006.04.066

+Model EA-12008;

No. of Pages 6

6

ARTICLE IN PRESS M. Mascia et al. / Electrochimica Acta xxx (2006) xxx–xxx

of cadmium hydroxide near the cathodic compartment. As can be seen from the figures, the pH and Cd2+ concentration profiles calculated by the model are reasonably agreed with those of experiments. The use of a physically based equation to describe the effect of pH on clay–ion interactions [25] avoided the discrepancy obtained by other authors, which was attributed to a wrong estimation of the Ri term [1] utilized in its model. 4. Conclusions Experimental study and numerical modeling were utilized to investigate on the Cd transport in low-permeability kaolinitic clays under electrical field. The results confirm the effectiveness of electrokinetic techniques to remove Cd2+ ions from kaolinitic clayey soils: in particular, the simple enhancement adopted, acidic conditioning of catholyte, allowed to avoid the major defects of the conventional electrokinetic processes highlighted by several authors, such as a sharp pH jump and possible precipitation of Cd hydroxides in the region near the cathode, as a result of migration of hydroxyl ions generated by electrolysis of water at the cathode [22,25]. The utilization of a geochemical model to describe the interaction between soil surface and dissolved ions allowed to model the process starting from a low number of experimental data, which can be easily obtained by sorption experiments, and a low number of adjustable parameters. Acknowledgement This research was supported by the Italian Ministry of University and Research, through the project PON Siti, in the framework of the program PON 2000–2006. Appendix A. Nomenclature

ci

concentration of the i-th chemical species in aqueous phase [mol dm−3 ] CEC cation exchange capacity [mol kg−1 ] ∗ Di effective diffusion coefficient of the i-th chemical species [m2 s−1 ] F faraday constant 96,500 [C eq−1 ] em Ji electromigrative flux [mol m−2 s−1 ] Jw electroosmotic flux [m s−1 ] keo = εζξ/μ Electroosmotic permeability coefficient [m2 V−1 s−1 ] KCd/Ca Cd/Ca ion exchange equilibrium constant

equilibrium constant for i-th compound and SOH site mass transfer coefficient of Cd2+ ions [m s−1 ] concentration of the i-th chemical species in solid phase [mol g−1 ] Ri sorption rate term in the transport equation [mol s−1 dm−3 ] SOH pH-dependent sites u∗ = Di∗ zi F/RT effective ionic mobility of the i-th chemical species [m2 V−1 s−1 ] Va anolyte volume [m3 ] Vc catholyte volume [m3 ] − X exchange site zi charge of the i-th chemical species

KiSOH Km qi

References [1] S.O. Kim, J.J. Kim, S.T. Yun, K.W. Kim, Water Air Soil Pollut. 150 (2003) 135. [2] A.M. Polcaro, M. Mascia, S. Palmas, A. Vacca, G. Tola, Environ. Eng. Sci. 20 (6) (2003) 607. [3] J. Virkutyte, M. Sillanpaa, P. Latostenmaa, Sci. Total Environ. 289 (1–3) (2002) 97. [4] B. Kos, H. Grcman, D. Lestan, Plant Soil Environ. 49 (12) (2003) 548. [5] Y.B. Acar, A.N. Alshawabkeh, Environ. Sci. Technol. 27 (1993) 2638. [6] S. Pamukcu, J.K. Wittle, Environ. Prog. 11 (3) (1992) 241. [7] G.R. Eykholt, D.E. Daniel, J. Geotech. Eng. 8 (1994) 431. [8] R.E. Hicks, S. Tondorf, Environ. Sci. Technol. 28 (12) (1994) 2203. [9] K.R. Reddy, S. Chinthamreddy, Waste Manag. 19 (4) (1999) 269. [10] C. Vereda-Alonso, J.M. Rodriguez-Maroto, R.A. Garcia-Delgado, U. Gomez-Lahoz, F. Garcia-Herruzo, Chemosphere 54 (7) (2004) 895. [11] B. Baeyens, M.H. Bradbury, J. Contam. Hydrol. 27 (3–4) (1997) 199. [12] M.H. Bradbury, B. Baeyens, J. Contam. Hydrol. 27 (3–4) (1997) 223. [13] D.A. Dzombak, F.M.M. Morel, Surface Complexation Modelling—Hydrous Ferric Oxide, John Wiley, New York, 1990. [14] S.K. Puppala, A.N. Alshawabkeh, Y.B. Acar, R.J. Gale, M. Bricka, J. Hazard. Mater. 55 (1–3) (1997) 203. [15] D.M. Zhou, C.F. Deng, L. Cang, Chemosphere 56 (3) (2004) 265. [16] H.K. Hansen, L.M. Ottosen, S. Laursen, A. Villumsen, Sep. Sci. Technol. 32 (15) (1997) 2425. [17] K.R. Reddy, S. Chinthamreddy, Adv. Environ. Res. 7 (2) (2003) 353. [18] D.M. Zhou, R. Zorn, C. Kurt, J. Environ. Sci. 15 (3) (2003) 396. [19] M.A. Oyanader, P.E. Arce, A. Dzurik, I. E. C. Res. 44 (16) (2005) 6200. [20] S.V. Ho, C. Athmer, P.W. Sheridan, B.M. Hughes, R. Orth, D. McKenzie, P.H. Brodsky, A. Shapiro, R. Thornton, J. Salvo, D. Schultz, R. Landis, R. Griffith, S. Shoemaker, Environ. Sci. Technol. 33 (1999) 1086. [21] D.S. Schultz, J. Hazard. Mater. 55 (1997) 81. [22] A.N. Alshawabkeh, Y.B. Acar, J. Geotech. Eng. 122 (3) (1996) 186. [23] F.J. Huertas, L. Chou, R. Wollast, Geochimi. Cosmochimi. Acta 62 (3) (1998) 417. [24] M.J. Angove, B.B. Johnson, J.D. Wells, J. Colloid Interface Sci. 204 (1998) 93. [25] S.O. Kim, J.J. Kim, K.W. Kim, S.T. Yun, Sep. Sci. Technol. 39 (8) (2004) 1927.

Please cite this article in press as: M. Mascia et al., Electrochim. Acta (2006), doi:10.1016/j.electacta.2006.04.066