Journal of Environmental Chemical Engineering 4 (2016) 2878–2891

Copper and zinc removal from contaminated soils through soil washingprocess using ethylenediaminedisuccinic acid as a chelating agent: Amodeling investigation

Marco Racea, Raffaele Marottab,*, Massimiliano Fabbricinoa, Francesco Pirozzia,Roberto Andreozzib, Luciano Cortesec, Paola Giudiciannid

aDipartimento di Ingegneria Civile, Edile ed Ambientale, Università degli Studi di Napoli Federico II, Via Claudio 21, 80125 Napoli, ItalybDipartimento di Ingegneria Chimica, dei Materiali e della Produzione Industriale, Università degli Studi di Napoli Federico II, p.le V. Tecchio 80, 80125Napoli, Italyc Istituto di Ricerche sulla Combustione, Consiglio Nazionale delle Ricerche (CNR), P.le V. Tecchio, 80, 80125 Napoli, Italyd Istituto di Ricerche sulla Combustione, Consiglio Nazionale delle Ricerche (CNR), Via Claudio,21, 80125 Napoli, Italy

A R T I C L E I N F O

Article history:Received 18 January 2016Received in revised form 26 May 2016Accepted 27 May 2016Available online 27 May 2016

Keywords:Soil washingEDDSHeavy metals extractionSoil contaminationIntra-particle diffusion

A B S T R A C T

This study demonstrated that soil washing using ethylenediaminedisuccinic acid (EDDS) as a chelatingagent was efficient at removing copper and zinc from real polluted soils. Only the exchangeable andreducible fractions of Cu and Zn were extracted by EDDS. Intra-particle diffusion was the main ratecontrolling step in this extraction of heavy metals from the solid matrix. Different contributions werefound by applying the Weber and Morris intraparticle diffusion model resulting from the different rolesof superficial and intra-particle diffusive processes.The diffusion coefficients of the Cu/EDDS and Zn/EDDS complexes in real contaminated soils were estimated using simplified diffusive models (based onCrank’s and Vermeulen's approximations). The relationship between the grain size and diffusioncoefficient was also evaluated. In particular, the intraparticle diffusion coefficients increased withincreasing the particle size, thus indicating that the smallest granulometric fractions are characterized bya higher percentage of micropores than the largest fractions.

ã 2016 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering

journal homepage: www.elsevier .com/ locate / je ce

1. Introduction

During the last 20 years, the contamination of soils andsediments with heavy metals (HMs) has become a worldwideconcern due to their high toxicity of most species and their abilityto accumulate in living tissues [1,2].

In agricultural areas, HMs contamination, even if not related tospecific health hazards, compromises the optimal use of the land,may reduces economic output [3], and modifies the existingequilibrium among natural components [4]. In these zones, theremediation interventions are needed that remove the contami-nation without affecting the original structure and composition ofthe soil.

In many cases, especially if the pollution is of anthropogenicorigin, very high concentrations of HMs may occur in portions of

* Corresponding author.E-mail address: rmarotta@unina.it (R. Marotta).

http://dx.doi.org/10.1016/j.jece.2016.05.0312213-3437/ã 2016 Elsevier Ltd. All rights reserved.

the contaminated areas leading to so-called “hot spots” [5]. Theremediation of these hot spots requires specific processes. Soilwashing is a promising strategy if the applied extracting agentminimally changes the original solid matrix original characteristicsand does not leave toxic residues in the treated soil [6]. Moreover,feasible treatment methods and safe disposal of the washingsolution should be available [7,8].

For these reasons, biodegradable organic chelants with lowenvironmental persistence are highly recommended [9]. The mostcommon chelating agents, such as ethylenediaminetetraacetic acid(EDTA), are poorly biodegradable and quite persistent in theenvironment. An alternative, the [S,S]-stereoisomer of ethyl-enediaminedisuccinic acid (EDDS), has recently received attentionin the literature as it is both safe and environmentally-friendly[10–15]. For example, the use of EDDS for soil reclamation does not

M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891 2879

have any effect on crop yield [16]. Moreover, [S,S]-EDDS can be alsoproduced by some fungi and bacteria [17,18]. Soil washingprocesses have been investigated for the effect of metal speciationand concentration [19–22]: the effect of soil characteristics, such asparticle distribution size, pH, organic matter content, and cationexchange capacity [23–27], and appropriate washing operatingconditions, including the EDDS-to-metal ratio, and the solid-to-liquid ratio [10,19,27–30]. The nature of the interaction and thestructure of complexes formed between EDDS and HMs has alsobeen studied [21,30].

The effectiveness of extraction for a soil washing process isgenerally affected by diffusion and kinetic and adsorptionmechanisms. Any of these, depending on the operative conditions,can be the rate-limiting step. For diffusive transport phenomena,theoretical descriptions of several mathematical models have beenreported for estimating the intra-particle diffusion coefficients ofchemicals in solid matrices [31–35].

Nonetheless, very little attention has been focused on theapplication of these models to micropores and macroporestransportation processes for chelating agent/HM complexes inreal soils, although these may affect the efficiency of the extraction.

In the present investigation, Crank’s [31] and Vermeulen’s [36]approximations, due their simplicity, are considered for analyzingthe extraction rate data and estimating the diffusion coefficients ofthe Cu-EDDS and Zn-EDDS complexes in contaminated soilsamples.

To the best of authors’ knowledge, the application of these twoapproximated models to the transport of chelating agent/heavymetal complexes in real contaminated soils has received littleattention [37,38]. No diffusion coefficient values have beenreported for Cu/EDDS and Zn/EDDS complexes in contaminatedsoil samples: however, the diffusion coefficients of Pb/EDTA andZn/EDTA complexes have been estimated using different methodsthan those used in the present study [39,40].

Fig. 1. The sampling points SS1 and SS2 (geogra

Thus, the aim of this paper is to clarify the role of diffusion ofmetal/EDDS complexes into a solid structure during the soilwashing process. In more detail, the paper investigates whetherdiffusive phenomena should be considered as key processesaffecting the overall performance of soil washing.

The study is carried out at the lab scale, under differentoperative conditions, on real soil contaminated by copper andzinc. Two different soil samples with similar characteristicswere obtained from former agricultural land (Fig. 1) within ahighly populated area in Campania Region (Southern Italy). Thisarea is located in a municipality within a larger contaminatedregion that has been proposed as a Site of Regional Interest (SIR).This territory, located between Naples and Caserta (55 munici-palities) has experienced hazardous waste disposal since the1970s. To further elucidate the seriousness of the problem, it isimportant to consider that in some particular areas of this regionthere has been in few years a 300% increase in diseases such asstomach, liver, bronchus, and bladder cancers as well asmalformations and birth defects. These incidents are, in somecases, consistent with a lack of remediation of the polluted sitesand lingering waste mismanagement [41]. Very recently, theItalian National Institute of Health (ISS) announced “.observedexcesses of children hospitalized in the first year of life for allcancers (particularly non-Hodgkin lymphoma) and excesses ofcancers of the central nervous system, the latter also in the range0–14 years . . . ” [42].

2. Experimental

2.1. Materials

Hydroxylammonium chloride (reagent grade >98% w/w),ammonium acetate (>99% w/w), (S,S)-ethylenediamine-N,N’-dis-uccinic acid–trisodium salt solution (35% v/v), hydrogen peroxide

phic coordinates N 40�960050 0 , E 14�110840 0).

2880 M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891

solution (30% v/v), acetic acid (ACS reagent >97% v/v) and nitricacid (ACS reagent >67% v/v) from Sigma–Aldrich were used asreagents. Only ultra-pure water was used for analytical prepara-tions and dilution.

2.2. Soil sampling

The soil was collected from locations in a municipality of the SIR(Litorale Domizio Flegreo and Agro Aversano) [43]. Two differentsoil samples were obtained, indicated as SS1 and SS2. The twosampling points were chosen from sites a distance of 20 m apart, sothat they would have similar characteristics, except for their metalcontent (Fig.1). Prior to soil sampling, the overlying vegetation wasremoved completely. At each sites, six samples of 10 kg each weremanually collected from the top of the soil (20 cm) using a 16-inchconventional mini-shovel and combined to form a single sample.The soil samples were placed in a hermetic plastic box, transportedat ambient temperature to the laboratory, dried at 40 �C, and thenmaintained at room temperature.

2.3. Analytical procedures

A grain size analysis, using sieves of different sizes (20,10, 5, 2,1,0.5, 0.3, 0.15 and 0.075 mm) was performed according to ASTMmethod D 422-63 [44].

A specific gravity analysis was carried out according to Bowles[45]. Only grain sizes smaller than 2.0 mm were used in theanalytical determinations. The HM concentrations in the soilsamples were determined according to EPA method 3051 [46].The metal distribution in the soil sample was estimated by amulti-step sequential extraction procedure [47]. This procedurewas based on an initial extraction in 40 mL of acetic acid (0.11 M)(step 1); afterward a volume of 40 mL of hydroxylammoniumchloride solution (0.5 M) was added to the soil residue from “step1” and acidified by the addition of a 2 M HNO3 solution (step 2).Successively, 20 mL of hydrogen peroxide (8.8 M) and 50 mL ofammonium acetate (1.0 M) were used for the oxidation (step 3).The sequence of the extractions was related to the exchangeablephase, the reducible and the oxidizable phases, and the residualphase respectively. The procedure and the soil extracted metalfractions are detailed in Race et al. [20].

The metal concentrations were determined throughatomic absorption spectrometry (AAS), using Varian Model 55 BSpectrAA (F-AAS) equipped with flame (acetylene/air) and adeuterium background correction, the GBC Avanta AAS withgraphite furnace (GF-AAS) or atomic fluorescence spectrometerAFS-8220.

SEM images were acquired with the following tool: SEM EDXFEI � Ispect S, Column E-SEM W, Source: 200 V–30 KV, filament:tungsten equipped with an Everhart–Thornley detector (ETD). Porecharacterization in the soil samples was performed applying bothmercury intrusion and gas adsorption porosimetry. The mercuryporosimetric analysis was carried out using the MicromeriticsAutoPore IV apparatus.

Sorption/desorption isotherms were obtained using N2 at 77 Kas adsorbate in an Autosorb-1, (Quantachrome) apparatus. Beforeanalysis, the samples were degassed at 383 K for 12 h undervacuum conditions. The surface area was evaluated from the BETequation while the surface of micropores was calculated using thet-plot method of Lippens and de Boer.

The pH of the soil was evaluated in a solution at 1:2 soil-to-water ratio (EPA Method 9045C) [48]. All pH measurements wereperformed with a WTW pH/oxi340i pH-meter. The electricalconductivity was determined according to the method of Violanteand Adamo [49] using a conductivity meter (Eutech Instruments,COND 6+). Organic matter was measured through the loss on

ignition (LOI) index [50,51]. The Chapman method [52] was usedto estimate the cation exchange capacity (CEC). The carbon,hydrogen, nitrogen, sulfur contents (CHNS) in the samples weredetermined using a Perkin-Elmer Series II 2400 CHNS/OElementary Analyzer. All analyses were performed out intriplicate.

2.4. Soil washing procedure

The soil washing experiments were conducted using polyeth-ylene bottles (50 mL). The tests at different concentrations of EDDS(0.36 mM, 0.72 mM, 3.6 mM and 7.2 mM) were carried out using asole fraction of soil with a grain size smaller than 2�103mm at aliquid-to-solid ratio (L/S) of 10:1 as suggested from Satyro et al.[7,8]. Some additional runs were performed on different fractionscollected after sieving the soil (75 mm, 75–1.5�102mm, 1.5�102–5�102mm, 5�102–2�103mm), at a L/S of 10:1 and an EDDS initialconcentration of 3.6 mM.

The samples were agitated using a mechanical shaker (EdmundBühler, Kombischüttler KL2) at 190 rpm for different contact timeat room temperature [8]. Following the shaking, the samples werecentrifuged using an IEC Centra GP8 R centrifuge at 4800 rpm for15 min and then filtered with 0.45 mm regenerated cellulose filters.The concentration of the extracted metals from the soil wasdetermined through the determination of the metals in thewashing solution. All experiments were conducted in triplicate(standard deviation �6%).

2.5. Models adopted for data analysis

Both Crank’s and the Vermeulen’s approximations wereconsidered and it was assumed to operate under infinite solutionvolume conditions, that is the volume of the external solution ismuch higher than the exchanger quantity.

2.6. Crank’s approximation

The total amount of diffusing substance, entering or leaving aparticle of radius a, is given by the following [31]:

qtq1

¼ 1 � 6p2

X1n¼1

1n2e

�Dn2p2 ta2

� �for spherical particlesð Þ ð1Þ

qtq1

¼ 1 �X1

n¼1

4a2a2

ne �Da2

ntð Þ for cylindrical particlesð Þ ð2Þ

where q1 and an indicate, respectively, the amount of metalextracted per soil unit mass at equilibrium and the positive roots ofJ0 aanð Þ ¼ 0 where J0 xð Þ is the Bessel function of the first kind withzero order and D is the “apparent” diffusion coefficient.

For shorter contact times (generally qtq1<0.3, [31]), the Eqs. (1)

and (2) become the following respectively [38,53]:

qtq1

¼ 6Dtpa2

� �0:5

� 3Dta2

for spherical particlesð Þ ð3Þ

and

qtq1

¼ 4Dtpa2

� �0:5

� Dta2

for cylindrical particlesð Þ ð4Þ

Matlab Software (routine ‘ode 45’), which is based on theRunge–Kutta method with an adaptive stepsize, was used tonumerically solve Eqs. (3) and (4) and to calculate the concen-trations of copper and zinc at different contact times.

0

10

20

30

40

50

60

70

80

90

100

0.00 0 0.001 0.01 0 0.100 1.000 10 .000 100 .000

Perc

ent

finer

by

wei

ght

(%)

Particle diameter (mm)Fig. 2. Grain size distribution — SS1; SS2.

Table 1Soil characterization.

SS1 SS2

pH 6.97 6.74CEC [meq/100gr] 35.37 38.67LOI [%] 7.19 6.53C [%] 3.96 2.94H [%] 2.47 1.87N [%] 0.42 0.29S [%] 0.43 0.23Specific gravity [g/cm3] 2.42 2.44Electrical conductivity [dS/m] 0.17 0.15K [mg/Kg] 1.81�105 1.87�105Na [mg/Kg] 3.13�104 3.31�104Mg [mg/Kg] 2.26�104 2.39�104Ca [mg/Kg] 4.12�105 3.98�105Cu [mg/Kg] 2.54�102 67.63Zn [mg/Kg] 2.13�102 90.02Fe [mg/Kg] 1.33�104 1.42�104Mn [mg/Kg] 6.28�102 5.97�102BET [m2/g] 5.98 6.03

M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891 2881

2.7. Vermeulen’s approximation

For moderate and large contact times ( qtq1> 0.3), instead of Eqs.

(3) and (4), Vermeulen's approximation can be used [36]:

qtq1

¼ 1 � e�p2 Dt

a2

� �" #0:5

ð5Þ

This model is widely reported in the literature mainly, but notonly, for isotopic exchange systems with spherical particles[34,54,55].

3. Results and discussion

3.1. Soil characterization

The grain size distribution, reported in Fig. 2, allowed tocharacterize the composition of soil samples. In both soil samples,the amount of sand, silt and clay is about 38%, 46% and 16%respectively. On the basis of these preliminary results, the soil wasa loam, useful for plant growth, confirming the agricultural landuse in the area [56].

The two samples had similar physical and chemical character-istics (Table 1), indicating a common origin although SS1 hadhigher copper and zinc content that was most likely due to aconfined anthropogenic contamination. This result ensures anindependence of the efficiency of the washing process from the soilcharacteristics, and the extraction kinetics should follow the sametrend for the two samples as long as the operative parameters ofthe process remain unchanged.

SEM analysis confirmed the similarities of the two samples andprovided more detailed information about the microscopicstructure of the soil, which is useful for understanding the effectsof diffusion processes. Fig. 3a–f clearly indicate that the soilsamples appear as an heterogeneous phase multi-aggregatedmaterial. In particular, Fig. 3a shows that the particle sizes, for theSS1 sample ranged from 180 mm to 1–2 mm. The particles reportedin Fig. 3b (size 13 mm) are spherical agglomerates of overlappingminutes grains with size less than 1 mm (Fig. 3d). Other smallerparticles have irregular shapes (Fig. 3c). The particles have

macropores and mesopores of 0.4–2.5�10�2mm in diameters.Because of the high heterogeneity and particular morphology ofthe material (with interconnections among of openings andmacrovoids, Figs. 3c and d), it may be assumed that the crushingof the larger particles naturally occurs by breakage mainly at theinterstices and macrovoids (Fig. 3f), thereby generating smallerparticulate matter with a larger area percentage of micropores(Fig. 3e).

The macroporosity in the soil was confirmed by mercuryintrusion porosimetry. The incremental intrusion vs. pore diametercurve shows that the pore size distributions are centered at 3–4 mm and 0.4–0.5 mm for the fractions with particle diameters<75 mm and 5�102–2�103mm respectively (Fig. 4b).

The peak centered at 3–4 mm may be ascribed to the presence ofinteraggregate pores originating from the packing of the very fineparticles of the soil, as shown by SEM analysis (Fig. 2c) andreported in the literature [57].

Fig. 3. SEM images of SS1 sample.

2882 M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891

Moreover, the pore-size distribution related to the fractioncontaining particles with diameters smaller than 75 mm (emptycircles) is not monomodal as there are multiple maxima in thedistribution function, in particular in the range 3�10�3–2�10�2

mm. On the other hand, for the same fraction, the cumulativeintruded mercury volume vs. pore diameter curve (Fig. 4a) can bedivided into two distinct parts: one region corresponding to thefilling of pores with diameters ranging from 1 mm to10 mm, and asecond region related to the filling of small pores with diameters ofabout 1�10�2mm and an inflection point displaced to the fine poreszone.

Fig. 5 reports the adsorption-desorption isotherms of thesample SS1 obtained by nitrogen adsorption porosimetry for soil

fractions with particle diameters <75 mm and 5�102–2�103mm.The initial part corresponds to type II isotherms of the IUPACclassification, typical of macroporous materials. Nevertheless, theisotherms show a small microporosity feature indicated by thepresence of a well defined point “B”, at the beginning of the linearmiddle section of the isotherm (grey circle in Fig. 5), for lowvalues of the relative pressure. Point “B” is often taken to indicatethe stage at which the monolayer coverage is complete. In the lowrelative pressure region, the isotherms differ by the amount of N2

adsorbed up to the point “B” thus indicating a larger microporousvolume in the soil sample with particle diameters <75 mm. This isin agreement with the mercury intrusion porosimetry results. Asexpected, the value of the BET surface of the same sample,

Fig. 4. Cumulative intruded Hg volume vs. pore diameter curves for SS1 sample (a); incremental intrusion volume vs. pore diameter curves (b): (�) particle diameter <75 mm;(&) particle diameter range 5�102–2�103mm.

Fig. 5. Adsorption-desorption isotherms of N2 at 77 K. (&) particle diameter<75 mm; (4) particle diameter range 5�102–2�103mm (SS1 sample)..

M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891 2883

24.9 m2�g�1, is significantly higher than that one measured for thesoil with particle diameters in the range 5�102–2�103mm (Table 1).The values for the microporous area obtained from the t-plotshow the same trend as with BET surface, namely 3.94 m2�g�1 forthe fine soil sample and 0.87 m2�g�1 for the coarse soil. After point“B” the hysteresis loop, typical of type IV isotherm of the IUPACclassification with unlimited adsorption at high relative pressureregion, is associated with capillary condensation indicating thepresence of a mesoporous volume. The shape of the hysteresisloop was type H3 of the IUPAC classification typical of aggregatedparticles forming plates. It gave rise to pores in the form of a rift orwedges as evidenced also by SEM micrographs (Fig. 3). Similar

characteristics in the SEM and BET investigations were observedfor the SS2 sample.

3.2. Washing results

Fig. 6 shows the distribution of copper and zinc in the twosamples before and after soil washing with EDDS as estimated bythe sequential extraction procedure.

As already noted, the samples showed some differences in theinitial distributions of copper and zinc, confirming the effect ofthe larger anthropogenic contamination in the SS1 sample. Thelatter, in fact, had a higher content of the available fraction (sumof the exchangeable, the reducible and the oxidizable fractions)of both Cu and Zn, although it is worth noting that theconcentration of zinc related to the oxidizable fraction wasquite similar in the two samples. Comparing the distribution ofthe metals in the two samples before and after the washingprocess, it can be easily observed that only the fractions relatedto the exchangeable and reducible fractions were affected by theextraction. The amount of Cu and Zn associated with theoxidizable and residual fractions remained almost unchanged forboth samples. For example, even in case of almost completeremoval of Cu in the exchangeable and reducible fractions, theoxidizable fraction (mono- and zero-valent copper) was notremoved. This result can be tentatively explained by consideringthat in the presence of soil humic matter, under anoxic or sub-oxic conditions, di-valent copper (CuII), can be partially reducedto mono-valent copper (CuI) and further to zero-valent copper(Cu0) [58].

Regarding the CuI species, the Cu(I)-EDDS are not very stablecomplexes. This was also verified with the complex CuI-EDTA [59].On the contrary, to the best of the authors’ knowledge, there is noevidence that organic matter in soil has the capacity to reducebivalent zinc ions. On the other hand, zero-valent copper and zinccannot be chelated with EDDS.

On the basis of the obtained results and with the previousconsiderations, it can be concluded that only copper and zincpresent in exchangeable and reducible fractions were effectivelyremoved by EDDS.

The extraction kinetics of copper and zinc from the twocontaminated soils are shown in Fig. 7a–d with a fixed solid-to-

Fig. 6. Sequential extraction of Cu and Zn in soil samples before (a,b) and after the soil washing (c,d). Soil washing conditions: S/L ratio = 1:10, [EDDS]o = 7.24 mM,treatment time = 96 h, T = 20 �C, pH = 7.8. ( ) Exchangeable and weak acid soluble fraction, ( ) Reducible fraction, ( ) Oxidizable fraction, (&) Residual fraction, (&)Extracted by EDDS.

2884 M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891

liquid ratio of 1:10 and varying initial EDDS concentrations (0.36–7.2 mM). As expected, the extraction efficiency increased with anincreased initial concentration of the chelating agent from0.36 mM to 7.2 mM. Under the adopted experimental conditions,after 96 h of treatment, the highest Cu extraction efficiencies were82.0% for SS1 and 100% for SS2, while the highest Zn extractionefficiencies were 80.9% for SS1 and 57.4% for SS2.

A higher extraction efficiency for copper with respect to zincmay be essentially ascribed to a higher stability (larger logK) of Cu-EDDS complexes with respect to Zn-EDDS complexes [60]. Therecovery percentages of both metals were not substantiallyaffected by liquid-to-solid ratio dependent variations within therange of liquid-to-solid ratio of 5:1–20:1.

For all tested experimental conditions the extraction processincreased with the time, attaining an asymptotic value around96 h. In more detail, the extraction rate was extremely high duringthe first 2–3 h, and thereafter tended to decrease, remaining

almost constant up to 10–15 h, and decreasing once more until theend of the test.

The results collected from the soil washing of samples SS1 andSS2 showed that, under the adopted experimental conditions,contact times of 24–36 h resulted in the best compromise for theextraction of copper and zinc, since longer contact times only led tominor additional benefits.

3.3. Diffusion models

The extraction rate of metals from soils by organic moleculesmay be controlled by mass transfer phenomenaor chemicalreaction processes. Regardless of whether the extraction processdevelops under an intra-particle diffusive regime (rate-limitingstep), a linear dependence of the extraction capacity for each metalon the square root of the contact time is obtained [61–63]:

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0

10

20

30

40

50

60

70

80

90

100

Zn

rem

oval

(%)

Zn - SS 1

(b)

0 10 20 30 40 50 60 70 80 90 100Time (h r)

0

10

20

30

40

50

60

70

80

90

100

Zn

rem

oval

(%)

Zn - SS 2

(d)

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0

10

20

30

40

50

60

70

80

90

100

Cu

rem

oved

(%)

Cu - SS1

(a)

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0

10

20

30

40

50

60

70

80

90

100

Cu

rem

oval

(%)

Cu - SS 2

(c)

Fig. 7. Metal removal percentage respect to exchangeable + reducible fractions as function ofthe time at 1:10 of S:L ratio. T = 20 �C, pH = 7.8, EDDS (mM): � 0.36, & 0.72,4 3.6,^ 7.2.

M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891 2885

qt ¼ kDt0:5 þ c ð6Þwhere qt is the amount of metal per unit of soil mass extracted attime t, kDis the rate constant of the intra-particle diffusion, t is thecontact time and c is a constant.

The mass of Cu and Zn extracted at any time t with respectto the soil mass versus t1/2 is reported in Figs. 8a-d at differentinitial EDDS concentrations for the SS1 and SS2 samplesrespectively.

The Weber plots indicate that, after the initial minutes of soilwashing, during which the convective diffusion in the solutionand/or the external exchange are presumably the dominantprocesses, multiples straight lines fit the experimental data. Theobserved multi-linearity indicates that the diffusive processdevelops in pores of progressively smaller sizes.The first segmentof the plot may be ascribed to boundary layer diffusion effects suchas external film resistance and superficial diffusion whereastheother linear segments are related to further curve portions depictthe macropore and micropore diffusive processes [64].

Moreover, under the adopted experimental conditions, anincrease in the initial concentration of the chelating agent did notresult into a respective appreciable increase in the slope of theplot thus indicating that the rate of the intra-particle diffusiveprocess is not dependent on the initial EDDS solution concentra-tion.

A comparison between the experimental data for the SS1 andSS2 samples and the curves calculated from the diffusion model,according to Crank’s approximation with the assumption ofspherical or cylindrical particles, is reported in Fig. 9a–h forcontact times less than 1 h and for twodifferent initial EDDSconcentrations.

The data fit was found to be almost satisfactory for allexperimental cases if the spherical particles are assumed to becylindrical, as also demonstrated by the coefficient of correlation(R2) and the percentage standard deviation (s%) reported inTable 2. The percentage standard deviation is defined as:

s% ¼ 1001

qexp

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiXN

qt;cal � qt;exph i2

N

vuut

0 10 20 30 40 50 60 70 80t^0.5 (min)^0.5

0.00

0.03

0.05

0.08

0.10

0.13

qt(m

g/gr

)

b

0 10 20 30 40 50 60 70 80t^0.5 (min)^0.5

0.000

0.005

0.010

0.015

0.020

qt(m

g/g)

d

0 10 20 30 40 50 60 70 80t^0.5 (min)^0.5

0.00

0.03

0.05

0.08

0.10

0.13

qt(m

g/g)

a

0 10 20 30 40 50 60 70 80t^0.5 (min)^0.5

0.000

0.005

0.010

0.015

0.020

0.025

0.030

qt(m

g/g)

c

Fig. 8. Intra-particle diffusion plots for copper (a,c) and zinc (b,d) during the soil washing of SS1 (a,b) and SS2 (c,d) at fixed S:L (1:10) and different initial EDDS concentrations.EDDS (mM): � 0.36, & 0.72, 4 3.6, ^ 7.2.

2886 M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891

where qt,exp and qt,cal are the experimental and calculated qt valuesrespectively, qexp is the average measured concentration and N isthe number of data points.

The mean values of the intraparticle diffusion coefficients (D)are quite similar for the Cu-EDDS and Zn-EDDS chelates for bothsampled soils. Moreover, the calculated values for the effectivediffusion coefficientsare comparable with those reported in theliterature for different organic substances in soils and sediments[65,66].

The diffusion coefficients evaluated using Vermeulen's approx-imation are reported in Table 3 along with the coefficient ofcorrelation and the percentage standard deviation, whereasthemodelwas fitted to the experimental results (Fig. 10a–h).

The results indicate that the diffusion model, according toVermeulen’s approximation, is very suitable for predicting copperand zinc extraction for contact times greater than 1 h. The averagevalues estimated for the pore diffusion coefficient are about oneorder of magnitude smaller than those estimated using Crank’sapproximation since, for more prolonged contact times, even theeffects of diffusive processes in smaller pores are taken intoaccount.

With the aim of evaluating the dependence of pore diffusivityon the particle size of the soil samples, the intraparticle diffusioncoefficients for the extraction of copper and zinc using an initialEDDS concentration of 3.6 mM and a solid-to-liquid ratio of 1:10were estimated for different sieved granulometric fractions in thepreviously described optimization procedure and assumingVermeulen’s approximation. The results indicated that the intra-particle diffusion coefficient of the Cu-EDDS and Zn-EDDS chelatesincreases with increasing soil particle mesh size. The mesh sizeranged from 10�18m2 s�1 to 10�19m2 s�1 for particle diameterssmaller than 30 mm, and to 10�13m2 s�1 for diameters larger than5 mm. Thus, this indicated that the smallest granulometricfractions are characterized by a higher percentage of microporesthan the largest size particles, as has been suggested by SEManalysis and evidenced by BET analysis.

The best estimated values of pore diffusivity against themean particle radius were fit very well by the non linear

relationship (D ¼ 310�13R1:4p ) proposed by Badruzzaman for the

intraparticle diffusion processes of arsenic into porous iron oxides[67] (Fig. 11).

0 10 20 30 40 50 60Time (min)

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

0.045

0.050

qt(m

g/g)

(a)

0 10 20 30 40 50 60Time (min)

0.000

0.010

0.020

0.030

0.040

0.050

0.060

0.070

0.080

0.090

0.100

qt(m

g/g)

(b)

0 10 20 30 40 50 60Time (min)

0.000

0.002

0.004

0.006

0.008

0.010

qt(m

g/g)

(c)

0 10 20 30 40 50 60Time (min)

0.000

0.002

0.004

0.006

0.008

0.010

0.012

qt(m

g/g)

(d)

0 10 20 30 40 50 60Time (min)

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

0.045

qt(m

g/g)

(e)

0 10 20 30 40 50 60Time (min)

0.000

0.010

0.020

0.030

0.040

0.050

0.060

0.070

qt(m

g/g)

(f)

0 10 20 30 40 50 60Time (min)

0.000

0.002

0.004

0.006

0.008

qt(m

g/g)

(g)

0 10 20 30 40 50 60Time (min)

0.000

0.002

0.004

0.006

0.008

qt(m

g/g)

(h)

Fig. 9. Crank’s model, comparison between experimental (Cu: �, Zn: &) and calculated (lines) results (a = 23 mm; S:L 1:10).Continuous curve for spherical particles, dashed curve for cylindrical particles[EDDS]o 0.36 mM, Cu extraction (SS1, (a); SS2, (c))[EDDS]o 3.6 mM, Cu extraction (SS1, (b); SS2, (d))[EDDS]o 0.36 mM, Zn extraction (SS1, (e); SS2, (g))[EDDS]o 3.6 mM, Zn extraction (SS1, (f); SS2, (h)).

M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891 2887

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

qt(m

g/g)

(a)

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

qt(m

g/g)

(b)

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0.000

0.005

0.010

0.015

0.020

0.025

0.030

qt(m

g/g)

(d)

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

qt(m

g/g)

(e)

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0.00

0.02

0.04

0.06

0.08

0.10

0.12

qt(m

g/g)

(f)

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0.000

0.002

0.004

0.006

0.008

0.010

0.012

0.014

qt(m

g/g)

(g)

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0.000

0.002

0.004

0.006

0.008

0.010

0.012

0.014

0.016

qt(m

g/g)

(h)

0 10 20 30 40 50 60 70 80 90 100Time (hr)

0.000

0.005

0.010

0.015

0.020

0.025

qt(m

g/g)

(c)

Fig. 10. Vermeulen’s model, comparison between experimental (Cu: �, Zn: &) and calculated (lines) results. S:L 1:10.[EDDS]o 0.36 mM, Cu extraction (SS1, (a); SS2, (c))[EDDS]o 3.6 mM, Cu extraction (SS1, (b); SS2, (d))[EDDS]o 0.36 mM, Zn extraction (SS1, (e); SS2, (g))[EDDS]o 3.6 mM, Zn extraction (SS1, (f); SS2, (h))

2888 M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891

Table 2� Best estimated values of pore diffusivities based on Crank’s approximation for Cu/EDDS and Zn/EDDS species for SS1 and SS2 samples.

Crank’s Model (cylindrical particles)

Metal Cu Zn

Soil SS1 SS2 SS1 SS2

EDDS (mM) 0.36 3.6 7.2 0.36 3.6 7.2 0.36 3.6 7.2 0.36 3.6 7.2D (m2�s�1) 8.13�10�15 7.20�10�15 1.06�10�14 4.01�10�15 4.47�10�15 4.13�10�15 1.56�10�14 8.65�10�15 1.02�10�14 6.23�10�15 4.88�10�15 7.06�10�15s (%) 9 13 15 7 8 10 7 4 2 8 11 6R2 0.91 0.80 0.89 0.97 0.97 0.91 0.95 0.99 0.99 0.95 0.91 0.96D (m2�s�1) 8.64�10�15� 1.45�10�15 4.20�10�15� 2.7�10�16 1.15�10�14� 4.10�10�15 6.06�10�15� 1.18�10�15

Crank’s Model (spherical particles)

Metal Cu Zn

Soil SS1 SS2 SS1 SS2

EDDS (mM) 0.36 3.6 7.2 0.36 3.6 7.2 0.36 3.6 7.2 0.36 3.6 7.2D (m2�s�1) 4.03�10�15 3.50�10�15 5.52�10�15 1.95�10�15 2.03�10�15 2.25�10�15 9.03�10�15 4.55�10�15 5.50�10�15 2.87�10�15 2.33�10�15 3.26�10�15s (%) 8 12 14 7 8 10 6 3 3 7 10 6R2 0.93 0.83 0.86 0.97 0.96 0.92 0.97 0.99 0.99 0.96 0.92 0.96D (m2�s�1) 4.35�10�15� 1.17�10�15 2.08�10�15� 1.70�10�16 6.36�10�15� 2.67�10�15 2.82�10�15� 4.90�10�16

Table 3Best estimated values of pore diffusivities based on Vermeulen's approximation for Cu/EDDS and Zn/EDDS species for SS1 and SS2 samples..

Vermeulen’s Model

Cu

Soil SS1 SS2

EDDS (mM) 0.36 0.72 3.6 7.2 0.36 0.72 3.6 7.2D (m2�s�1) 2.65�10�16 2.01�10�16 1.67�10�16 1.50�10�16 2.42�10�16 2.61�10�16 2.15�10�16 3.32�10�16s (%) 8 3 3 2 6 5 7 9R2 0.95 0.98 0.98 0.98 0.97 0.98 0.96 0.93D (m2�s�1) 1.96�10�16� 6.90�10�17 2.62�10�16� 7.00�10�17

Zn

Soil SS1 SS2

EDDS (mM) 0.36 0.72 3.6 7.2 0.36 0.72 3.6 7.2D (m2�s�1) 3.32�10�16 2.98�10�16 1.76�10�16 2.67�10�16 3.25�10�16 3.32�10�16 2.38�10�16 2.80�10�16s (%) 9 5 6 4 9 9 7 8R2 0.93 0.97 0.97 0.98 0.93 0.94 0.96 0.95D (m2�s�1) 2.68�10�16� 9.20�10�17 2.94�10�16� 3.80�10�17

1E-5 1E-4 1E-3 1E-2 1E-1 1E+0Particle radius (cm)

1E-20

1E-19

1E-18

1E-17

1E-16

1E-15

1E-14

1E-13

1E-12

Dif

fusi

onco

effi

cien

t(m

^2/s

)

Fig. 11. Intraparticle diffusivity coefficient vs mean particle radius. Continuouscurve: literature data [59]. [EDDS]o = 3.6 mM, S:L 1:10. Cu (SS1*, SS2&); Zn (SS1~,SS2 ^).

M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891 2889

4. Conclusions

The present investigation demonstrated the effectiveness ofremoving copper and zinc from polluted soils through a soilwashing process using EDDS as a chelating agent. The resultsprovided evidence that the intra-particle diffusion is the main ratecontrolling step in the extraction of heavy metals from the solidmatrix. The soil washing process allows the efficient extraction ofonly the Cu (100%) and Zn (80.9%) exchangeable and reduciblefractions. The Weber plots exhibited multi-linearity thus indicat-ing that the diffusive process develops in the pores of progressivelysmaller sizes. The diffusion coefficients of the Cu-EDDS and Zn-EDDS complexes in real contaminated soils were calculated usingthe Crank’s approximation (for short contact times, <1 h) andVermeulen's approximation (for long contact times, >1 h). Bothmodels provided a good fit to the experimental data and thebestcalculated pore diffusion coefficients were consistent withthose reported in the literature. The “apparent” diffusioncoefficients of the Cu-EDDS and Zn-EDDS species increase withincreasing particle size, ranging from 10�18 to 10�19m2 s�1 forparticle diameters smaller than 30 mm and to 10�13m2 s�1 fordiameters higher than 5 mm. The obtained results suggest that soilwashing with chelating agents under the adopted experimental

2890 M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891

conditions is limited by the intra-particle diffusion process. Due tothe very small pore sizes of the sampled soils, the proposed processrepresents a promising alternative for soil remediation, as thesetwo heavy metals otherwise would not be available for the removalusing less intense techniques.

Acknowledgment

This study was carried out within the EUProject LIFE11 ENV/IT/000275 (ECOREMED) and the Italian Project “Emerging contam-inants in air, soil and water: from source to the marineenvironment” (PRIN 2010-2011).

References

[1] S. Khan, Q. Cao, Y.M. Zheng, Y.Z. Huang, Y.G. Zhu, Health risks of heavy metals incontaminated soils and food crops irrigated with wastewater in Beijing, China,Environ. Pollut. 152 (2008) 686–692, doi:http://dx.doi.org/10.1016/j.env-pol.2007.06.056.

[2] K.E. Giller, E. Witter, S.P. Mcgrath, Toxicity of heavy metals to microorganismsand microbial processes in agricultural soils: a review, Soil Biol. Biochem. 30(1998) 1389–1414, doi:http://dx.doi.org/10.1016/S0038-0717(97)00270-8.

[3] P.W. Abrahams, I. Thornton, The contamination of agricultural land in themetalliferous province of southwest England : implications to livestock, Agric.Ecosyst. Environ. 48 (1994) 125–137, doi:http://dx.doi.org/10.1016/0167-8809(94)90083-3.

[4] S. Dahlin, E. Witter, A. Mårtensson, A. Turner, E. Bååth, Where’s the limit?Changes in the microbiological properties of agricultural soils at low levels ofmetal contamination, Soil Biol. Biochem. 29 (1997) 1405–1415, doi:http://dx.doi.org/10.1016/S0038-0717(97)00048-5.

[5] P. Goovaerts, Geostatistics for Natural Resources Evaluation, Oxford universitypress, 1997.

[6] D. Lestan, C. Luo, X. Li, The use of chelating agents in the remediation of metal-contaminated soils: a review, Environ. Pollut. 153 (2008) 3–13, doi:http://dx.doi.org/10.1016/j.envpol.2007.11.015.

[7] S. Satyro, M. Race, R. Marotta, M. Dezotti, D. Spasiano, G. Mancini, et al.,Simulated solar photocatalytic processes for the simultaneous removal ofEDDS, Cu(II), Fe(III) and Zn(II) in synthetic and real contaminated soil washingsolutions, J. Environ. Chem. Eng. 2 (2014) 1969–1979, doi:http://dx.doi.org/10.1016/j.jece.2014.08.017.

[8] S. Satyro, M. Race, F. Di Natale, A. Erto, M. Guida, R. Marotta, Simultaneousremoval of heavy metals from field-polluted soils and treatment of soilwashing effluents through combined adsorption and artificial sunlight-drivenphotocatalytic processes, Chem. Eng. J. 283 (2016) 1484–1493, doi:http://dx.doi.org/10.1016/j.cej.2015.08.039.

[9] M. Fabbricino, A. Ferraro, G. Del Giudice, L. D’Antonio, Current views on EDDSuse for ex situ washing of potentially toxic metal contaminated soils, Rev.Environ. Sci. Bio/Technol. 12 (2013) 391–398.

[10] Z.A. Begum, I.M.M. Rahman, Y. Tate, H. Sawai, T. Maki, H. Hasegawa, Reme-diation of toxic metal contaminated soil by washing with biodegradableaminopolycarboxylate chelants, Chemosphere 87 (2012) 1161–1170, doi:http://dx.doi.org/10.1016/j.chemosphere.2012.02.032.

[11] R. Yang, C. Luo, G. Zhang, X. Li, Z. Shen, Extraction of heavy metals from e-wastecontaminated soils using EDDS, J. Environ. Sci. 24 (2012) 1985–1994, doi:http://dx.doi.org/10.1016/S1001-0742(11)61036-X.

[12] M. Cao, Y. Hu, Q. Sun, L. Wang, J. Chen, X. Lu, Enhanced desorption of PCB andtrace metal elements (Pb and Cu) from contaminated soils by saponin andEDDS mixed solution, Environ. Pollut. 174 (2013) 93–99, doi:http://dx.doi.org/10.1016/j.envpol.2012.11.015.

[13] N.R. Hartley, D.C.W. Tsang, W.E. Olds, P.A. Weber, Soil washing enhanced byhumic substances and biodegradable chelating agents, Soil Sediment Contam.An Int. J. 23 (2014) 599–613.

[14] A. Ferraro, M. Fabbricino, E.D. van Hullebusch, G. Esposito, F. Pirozzi, Effect ofsoil/contamination characteristics and process operational conditions onaminopolycarboxylates enhanced soil washing for heavy metals removal: areview, Rev. Environ. Sci. Bio/Technol. (2015) 1–35, doi:http://dx.doi.org/10.1007/s11157-015-9378-2.

[15] D.C.W. Tsang, N.R. Hartley, Metal distribution and spectroscopic analysis aftersoil washing with chelating agents and humic substances, Environ. Sci. Pollut.Res. 21 (2014) 3987–3995.

[16] T. Nishikiori, A. Okuyama, H. Naganawa, T. Takita, M. Hamada, T. Takeuchi,et al., Production by actinomycetes of (S, S)-N,N’-ethylenediaminedisuccinicacid, an inhibitor of phospholipase C, J. Antibiot. (Tokyo) 37 (1984) 426–427.

[17] L. Yang, C. Luo, Y. Liu, L. Quan, Y. Chen, Z. Shen, Residual effects of EDDSleachates on plants during EDDS-assisted phytoremediation of copper con-taminated soil, Sci. Total Environ. 444 (2013) 263–270, doi:http://dx.doi.org/10.1016/j.scitotenv.2012.11.085.

[18] C. Luo, Z. Shen, X. Li, Enhanced phytoextraction of Cu, Pb, Zn and Cd with EDTAand EDDS, Chemosphere 59 (2005) 1–11, doi:http://dx.doi.org/10.1016/j.chemosphere.2004.09.100.

[19] S. Tandy, K. Bossart, R. Mueller, J. Ritschel, L. Hauser, R. Schulin, et al., Extractionof heavy metals from soils using biodegradable chelating agents, Environ. Sci.Technol. 38 (2004) 937–944, doi:http://dx.doi.org/10.1021/es0348750.

[20] M. Race, J. Náb�elková, M. Fabbricino, F. Pirozzi, P. Raia, Analysis of heavy metalsources for urban creeks in the Czech Republic, Water, Air, Soil Pollut. 226(2015) 322, doi:http://dx.doi.org/10.1007/s11270-015-2579-z.

[21] M. Komárek, J. Száková, M. Rohošková, H. Javorská, V. Chrastný, J. Balík, Coppercontamination of vineyard soils from small wine producers: a case study fromthe Czech Republic, Geoderma 147 (2008) 16–122, doi:http://dx.doi.org/10.1016/j.geoderma.2008.07.001.

[22] R. Qiu, Z. Zou, Z. Zhao, W. Zhang, T. Zhang, H. Dong, et al., Removal of trace andmajor metals by soil washing with Na2EDTA and oxalate, J. Soils Sedim. 10(2010) 45–53.

[23] C. Kim, Y. Lee, S.K. Ong, Factors affecting EDTA extraction of lead from lead-contaminated soils, Chemosphere 51 (2003) 845–853, doi:http://dx.doi.org/10.1016/S0045-6535(03)00155-3.

[24] S. Tandy, A. Ammann, R. Schulin, B. Nowack, Biodegradation and speciation ofresidual SS-ethylenediaminedisuccinic acid (EDDS) in soil solution left aftersoil washing, Environ. Pollut. 142 (2006) 191–199, doi:http://dx.doi.org/10.1016/j.envpol.2005.10.013.

[25] T.C.M. Yip, D.Y.S. Yan, M.M.T. Yui, D.C.W. Tsang, I.M.C. Lo, Heavy metal ex-traction from an artificially contaminated sandy soil under EDDS deficiency:significance of humic acid and chelant mixture, Chemosphere 80 (2010) 416–421, doi:http://dx.doi.org/10.1016/j.chemosphere.2010.03.033.

[26] D.Y.S. Yan, I.M.C. Lo, Enhanced multi-metal extraction with EDDS of deficientand excess dosages under the influence of dissolved and soil organic matter,Environ. Pollut. 159 (2011) 78–83, doi:http://dx.doi.org/10.1016/j.env-pol.2010.09.021.

[27] D. Sarkar, S.S. Andra, S.K.M. Saminathan, R. Datta, Chelant-aided enhancementof lead mobilization in residential soils, Environ. Pollut. 156 (2008) 1139–1148,doi:http://dx.doi.org/10.1016/j.envpol.2008.04.004.

[28] K. Fischer, H.-P. Bipp, Removal of heavy metals from soil components and soilsby natural chelating agents. Part II. Soil extraction by sugar acids, Water. Air.Soil Pollut. 138 (2002) 271–288.

[29] H. Guo, W. Wang, Y. Sun, H. Li, F. Ai, L. Xie, et al., Ethyl lactate enhancesethylenediaminedisuccinic acid solution removal of copper from contami-nated soils, J. Hazard. Mater. 174 (2010) 59–63, doi:http://dx.doi.org/10.1016/j.jhazmat.2009.09.016.

[30] W. Zhang, D.C.W. Tsang, I.M.C. Lo, Removal of Pb by EDTA-washing in thepresence of hydrophobic organic contaminants or anionic surfactant, J.Hazard. Mater. 155 (2008) 433–439, doi:http://dx.doi.org/10.1016/j.jhaz-mat.2007.11.084.

[31] J. Crank, The Mathematics of Diffusion, Oxford University Press, 1979.[32] R.G. Lee, T.W. Weber, Isothermal adsorption in fixed beds, Can. J. Chem. Eng. 47

(1969) 54–59.[33] E.S. Kikkinides, R.T. Yang, Further work on approximations for intraparticle

diffusion rates in cyclic adsorption and desorption, Chem. Eng. Sci. 48 (1993)1169–1173, doi:http://dx.doi.org/10.1016/0009-2509(93)81045-W.

[34] V.J. Inglezakis, H.P. Grigoropoulou, Applicability of simplified models for theestimation of ion exchange diffusion coefficients in zeolites, J.Colloid InterfaceSci. 234 (2001) 434–441.

[35] Z. Samani, S. Hu, A.T. Hanson, D.M. Heil, Remediation of lead contaminated soilby column extraction with EDTA: II. Modeling, Water. Air. Soil Pollut. 102(1998) 221–238.

[36] T. Vermeulen, Theory for irreversible and constant-pattern solid diffusion, Ind.Eng. Chem. 45 (1953) 1664–1670.

[37] J. Yu, D. Klarup, Extraction kinetics of copper zinc, iron, and manganese fromcontaminated sediment using disodium ethylenediaminetetraacetate, Water.Air. Soil Pollut. 75 (1994) 205–225.

[38] Y. Ma, N. Uren, Transformations of heavy metals added to soil – application of anew sequential extraction procedure, Geoderma 84 (1998) 157–168, doi:http://dx.doi.org/10.1016/S0016-7061(97)00126-2.

[39] S. Amrate, D.E. Akretche, Modeling EDTA enhanced electrokinetic remediationof lead contaminated soils, Chemosphere 60 (2005) 1376–1383, doi:http://dx.doi.org/10.1016/j.chemosphere.2005.02.021.

[40] S.M. Elgawhary, W.L. Lindsay, W.D. Kemper, Effect of EDTA on the self-diffusionof zinc in aqueous solution and in soil, Soil Sci. Soc. Am. J. 34 (1970) 66–70.

[41] M. Triassi, R. Alfano, M. Illario, A. Nardone, O. Caporale, P. Montuori, Envi-ronmental pollution from illegal waste disposal and health effects: a review onthe Triangle of Death, Int. J. Environ. Res. Public Health. 12 (2015) 1216–1236.

[42] L. Musmeci, P. Comba, L. Fazzo, I. Iavarone, S. Salmaso, S. Conti, et al., studioSENTIERI (2014).

[43] G. Di Lorenzo, P. Federico, S. De Placido, C. Buonerba, Increased risk of bladdercancer in critical areas at high pressure of pollution of the Campania region inItaly: a systematic review, Crit. Rev. Oncol. Hematol. (2015), doi:http://dx.doi.org/10.1016/j.critrevonc.2015.07.004.

[44] ASTM, D422-63, Standard test method for particle-size analysis of soils, Annu.B.() ASTM Stand, West Conshohocken (PA): American Society for Testing andMaterials, 2003, 2016, pp. 10–17.

[45] J.E. Bowles, Engineering Properties of Soils and Their Measurement, McGraw-Hill Inc., 1992.

[46] E.P.A, Method 3051, Microwave Assisted Digestion of Sediments, Sludges, Soilsand Oils, Environ. Prot. Agency, USA, 2001.

[47] M. Pueyo, J. Mateu, A. Rigol, M. Vidal, J.F. López-Sánchez, G. Rauret, Use of themodified BCR three-step sequential extraction procedure for the study of trace

M. Race et al. / Journal of Environmental Chemical Engineering 4 (2016) 2878–2891 2891

element dynamics in contaminated soils, Environ. Pollut. 152 (2008) 330–341,doi:http://dx.doi.org/10.1016/j.envpol.2007.06.020.

[48] EPA, Method 9045C, Soil and Waste pH, Test Methods for Evaluating SolidWaste, Physical/chemical Methods, Environ. Prot. Agency, USA, 2003.

[49] P. Violante, P. Adamo, pH determination, Off. Methods Soil Chem. Anal.,Francoangeli, Rome, 2000, pp. 10–13.

[50] E.E. Schulte, Recommended soil organic matter tests, Recomm. Soil Test.Proced. North East. USA, Northeast. Reg. Publ., 1995, pp. 52–60.

[51] E.E. Schulte, C. Kaufmann, J.B. Peter, The influence of sample size and heatingtime on soil weight loss-on-ignition, Commun. Soil Sci. Plant Anal. 22 (1991)159–168.

[52] H.D. Chapman, Cation-exchange capacity, methods soil anal. part 2, Chem.Microbiol. Prop. (1965) 891–901.

[53] K.H. Mild, The kinetics of diffusion between a spherical cell and a surroundingmedium with different diffusion properties, Bull. Math. Biophys. 33 (1971) 19–26.

[54] M. Trgo, J. Peri�c, N.V. Medvidovi�c, A comparative study of ion exchange kineticsin zinc/lead—modified zeolite-clinoptilolite systems, J. Hazard. Mater. 136(2006) 938–945.

[55] C. Valderrama, X. Gamisans, X. de las, A. Heras, J.L. Farrán, Cortina, Sorptionkinetics of polycyclic aromatic hydrocarbons removal using granular activatedcarbon: intraparticle diffusion coefficients, J. Hazard. Mater. 157 (2008) 386–396, doi:http://dx.doi.org/10.1016/j.jhazmat.2007.12.119.

[56] B. Chandrasekaran, K. Annadurai, E. Somasundaram, A Textbook of Agronomy,New Age International Limited, 2010.

[57] B. Schäffer, M. Stauber, T.L. Mueller, R. Müller, R. Schulin, Soil and macro-poresunder uniaxial compression. I. Mechanical stability of repacked soil and de-formation of different types of macro-pores, Geoderma 146 (2008) 183–191,doi:http://dx.doi.org/10.1016/j.geoderma.2008.05.019.

[58] F. Maurer, I. Christl, B. Fulda, A. Voegelin, R. Kretzschmar, Copper redoxtransformation and complexation by reduced and oxidized soil humic acid. 2.Potentiometric titrations and dialysis cell experiments, Environ. Sci. Technol.47 (2013) 10912–10921.

[59] E.R. Larsen, Spectrophotometric determination of copper in fertilizer withneocuproine, Anal. Chem. 46 (1974) 1131–1132.

[60] M. Orama, H. Hyvönen, H. Saarinen, R. Akselab, Complexation of [S,S] andmixed stereoisomers of N,N0-ethylenediaminedisuccinic acid (EDDS) with Fe(III), Cu(II), Zn(II) and Mn(II) ions in aqueous solution, J. Chem. Soc. Dalt. Trans.(2002) 4644–4648, doi:http://dx.doi.org/10.1039/B207777A.

[61] E. Guibal, C. Milot, J.M. Tobin, Metal-anion sorption by chitosan beads:equilibrium and kinetic studies, Ind. Eng. Chem. Res. 37 (1998) 1454–1463.

[62] E. Katsou, S. Malamis, M. Tzanoudaki, K.J. Haralambous, M. Loizidou, Re-generation of natural zeolite polluted by lead and zinc in wastewater treat-ment systems, J. Hazard. Mater. 189 (2011) 773–786, doi:http://dx.doi.org/10.1016/j.jhazmat.2010.12.061.

[63] W.J. Weber, J.C. Morris, Kinetics of adsorption on carbon from solution, J. Sanit.Eng. Div. 89 (1963) 31–60.

[64] A.E. Ofomaja, Intraparticle diffusion process for lead(II) biosorption ontomansonia wood sawdust, Bioresour. Technol. 101 (2010) 5868–5876, doi:http://dx.doi.org/10.1016/j.biortech.2010.03.033.

[65] W.P. Ball, P.V. Roberts, Long-term sorption of halogenated organic chemicals byaquifer material. 2. Intraparticle diffusion, Environ. Sci. Technol. 25 (1991)1237–1249.

[66] H. Harms, T. Bosma, Mass transfer limitation of microbial growth and pollutantdegradation, J. Ind. Microbiol. Biotechnol. 18 (1997) 97–105.

[67] M. Badruzzaman, P. Westerhoff, D.R.U. Knappe, Intraparticle diffusion andadsorption of arsenate onto granular ferric hydroxide (GFH), Water Res. 38(2004) 4002–4012, doi:http://dx.doi.org/10.1016/j.watres.2004.07.007.