Chemosphere 80 (2010) 416–421

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Heavy metal extraction from an artificially contaminated sandy soil under EDDSdeficiency: Significance of humic acid and chelant mixture

Theo C.M. Yip a, Dickson Y.S. Yan a, Matthew M.T. Yui a, Daniel C.W. Tsang b, Irene M.C. Lo a,*

a Department of Civil and Environmental Engineering, The Hong Kong University of Science and Technology, Hong Kong, Chinab Department of Civil and Natural Resources Engineering, University of Canterbury, New Zealand

a r t i c l e i n f o

Article history:Received 20 January 2010Received in revised form 23 March 2010Accepted 23 March 2010Available online 27 April 2010

Keywords:Chelanting agentDissolved organic matterEDTAMetal–chelant complexSoil remediation

0045-6535/$ - see front matter Crown Copyright � 2doi:10.1016/j.chemosphere.2010.03.033

* Corresponding author. Tel.: +852 2358 7157; fax:E-mail address: cemclo@ust.hk (I.M.C. Lo).

a b s t r a c t

Biodegradable EDDS ([S,S]-ethylenediaminedisuccinic acid) has been suggested for enhancing heavymetal extraction from contaminated soils. Recent studies showed that Zn and Pb are less effectivelyextracted due to metal exchange and re-adsorption onto the soil surfaces, especially for EDDS-deficiencyconditions. This study therefore investigated the influence of dissolved organic matter and the co-pres-ence of EDTA (ethylene-diamine-tetraacetic acid) on metal extraction from an artificially contaminatedsandy soil under deficient amount of chelants in batch kinetics experiments. The addition of 10 and20 mg L�1 of humic acid as dissolved organic matter (DOC) suppressed metal extraction by EDDS, prob-ably resulting from the competition of adsorbed humic acid for heavy metals and adsorption of metal–humate complexes onto the soil surfaces. The effects were most significant for Pb because of greaterextent of metal exchange of PbEDDS and high affinity towards organic matter. Thus, one should be cau-tious when there is a high content of organic matter in soils or groundwater. On the other hand, com-pared to individual additions of EDDS or EDTA, the equimolar EDDS and EDTA mixture exhibitedsignificantly higher Pb extraction without notable Pb re-adsorption. The synergistic performance of theEDDS and EDTA mixture probably resulted from the change of chemical speciation and thus less compe-tition among Cu, Zn and Pb for each chelant. These findings suggest further investigation into an optimumchemistry of the chelant mixture taking into account the effectiveness and associated environmentalimpact.

Crown Copyright � 2010 Published by Elsevier Ltd. All rights reserved.

1. Introduction

To facilitate the remediation of soils contaminated by heavymetals, chelating agents have been widely studied as they arecapable of forming soluble metal complexes that enhance heavymetal extraction from contaminated soils. Chelating agents arenaturally occurring or synthetic reagents that have been employedin a diverse range of commercial and industrial applications (Now-ack and VanBriesen, 2005). In the context of site remediation, che-lating agents can be applied ex situ (e.g., soil washing, heapleaching) and in situ (e.g., soil flushing, phytoextraction) (Luoet al., 2005; Finzgar et al., 2006; Polettini et al., 2007; Tsanget al., 2007b; Zhang et al., 2007; Yip et al., 2009a). In particular, bio-degradable EDDS ([S,S]-ethylenediaminedisuccinic acid) has re-ceived a lot of attention because of its high complex stability andextraction efficiency for many heavy metals (Vandevivere et al.,2001; Tandy et al., 2004; Nowack et al., 2006).

The applied chelants help metal extraction from the soils byforming metal–chelant complexes that would detach from the soil

010 Published by Elsevier Ltd. All r

+852 2358 1534.

surfaces into solution (Nowack, 2002). However, it has been shownthat newly formed metal–chelant complexes dissociate to varyingextent due to metal exchange with sorbed metals and soil minerals(Tsang et al., 2009). The liberated metals could be partially re-ad-sorbed onto the soil surfaces or remain in solution as complexeswith dissolved organic matter (Tsang et al., 2009). Since natural or-ganic matter, which is known to strongly bind with many metals(Tipping, 2002; Sparks, 2003), can be mobilized during chelant-en-hanced soil remediation (Vulava and Seaman, 2000; Hauser et al.,2005; Tsang et al., 2007b), dissolved organic matter potentiallyinfluences the fate of newly extracted metals. Under EDDS-excessconditions, previous studies found that there was a negative corre-lation between Pb extraction and organic matter content (Sarkaret al., 2008), and that a fraction of metals such as Cu and Pb wascomplexed with dissolved organic matter even when free EDDSwas present (Tandy et al., 2006). When EDDS is insufficient relativeto sorbed metals during in situ EDDS applications, dissolved organ-ic matter is likely to play an even more important role for metalextraction, especially for Pb, because a larger portion of extractedmetals is dissociated from EDDS complexes due to metal exchange(Tsang et al., 2009; Yip et al., 2009b). Moreover, as a result of lowstability of PbEDDS2� and occurrence of metal exchange, Pb

ights reserved.

T.C.M. Yip et al. / Chemosphere 80 (2010) 416–421 417

extraction by EDDS has been proved most difficult under EDDSdeficiency (Tsang et al., 2009; Yip et al., 2009b).

EDTA (ethylene-diamine-tetraacetic acid) is recalcitrant tomicrobial degradation and highly mobile in the subsurface envi-ronment (Tsang et al., 2007a; Kent et al., 2008), which may poseadditional risks of groundwater pollution and inhibitory effectson microorganisms and plant health (Nowack et al., 2006). Onthe contrary, EDDS can be fully degraded by indigenous microbesin soils after an initial lag phase ranging from 7 to 32 d (Finzgaret al., 2006; Tandy et al., 2006; Meers et al., 2008). Previous studiessuggested that EDDS was more efficient at extracting Cu comparedto EDTA, while EDTA was shown to achieve better Pb extractionthan EDDS (Tandy et al., 2004; Luo et al., 2005). With these consid-erations in mind, this study proposed a combined application ofEDDS and EDTA solutions. It is hypothesized that the chelant mix-ture can mitigate the residual problems with partial replacementof non-biodegradable EDTA by biodegradable EDDS, as well asmaximize the extraction efficiency of Pb along with Cu.

Therefore, the objectives of this study were to analyze the influ-ence of dissolved organic matter (humic acid as a surrogate) to-wards metal extraction by EDDS, and to compare theeffectiveness of the EDDS and EDTA mixture with those of individ-ual additions. Batch kinetic experiments were conducted to inves-tigate the extraction kinetics under chelant-deficiency conditions.

2. Experimental methods

2.1. Soil characteristics

Completely decomposed granite soil, which is the most com-monly found soil type in Hong Kong, was adopted in this study.The soil samples, taken from 25 to 50 cm below ground surfaceat Clearwater Bay in Hong Kong, were air-dried and passedthrough a 2-mm sieve. The soil was composed of 1% clay, 6% siltand 93% sand, indicated by wet-sieving and hydrometer meth-ods. Kaolinite and microcline were the two major clay mineralsfound in the soil by X-ray diffraction analysis (Philips PW1830). The amounts of Cu, Zn, Pb, Fe, Al, Ca and Mn in the indig-enous soil were (in mg kg�1), respectively, 12, 153, 99, 7620,67 950, 140 and 670 as determined by flame-Atomic AbsorptionSpectrometry (AAS) (Hitachi Z-8200) after acid digestion withHNO3–HCl–HF in a microwave oven (CEM MDS-2000). Total Fe,Al and Mn-oxides were found to be 5370, 690 and 340 mg kg�1,respectively, using dithionite–carbonate–bicarbonate extraction.Amorphous Fe, Al and Mn-oxides determined by ammoniumoxalate extraction were 70, 230 and 170 mg kg�1, respectively.Moreover, the soil had a cation exchange capacity of0.109 meq g�1 as determined by NH4–Na exchange, a soil pH of6.2 as measured at a 1:2 soil-to-water ratio, and a BET surfacearea of 6 m2 g�1, according to nitrogen gas adsorption analysis(Micrometrics ASAP2010).

The soil samples were artificially contaminated with Cu, Zn andPb in the form of Cu(NO3)2, Zn(NO3)2 and Pb(NO3)2 as described inprevious studies (Tsang et al., 2009; Yip et al., 2009a,b). The con-taminated soil was rinsed with background solution three timesto displace entrapped and loosely bound metals, and air-dried for2 week. The sorbed concentrations of Cu, Zn and Pb in the soilwere, correspondingly, 1233, 984 and 1913 mg kg�1, which weredetermined by microwave acid digestion and AAS measurement.These metal concentrations were within the common range re-ported for field soils (Supplementary Materials (SM), Table SM-1)and represented severely polluted soils that are most applicablefor chelant-enhanced extraction. It should, however, be noted thata higher extraction efficiency was expected in artificially contami-nated soils than in field soils.

2.2. Influence of dissolved organic matter

The EDDS solution was prepared by adding 30% Na3EDDS solu-tion (Innospec Ltd., HK) into 10 mM NaNO3 and 2 mM MES (2-mor-pholinoethane-sulfonic acid) buffer and adjusted to pH 5.5 by0.1 M NaOH and HNO3. Three sets of solutions were prepared. Inthe first set, EDDS solution contained no dissolved organic matter;in the second set 10 mg L�1 of humic acid as dissolved organic mat-ter (DOC), was added into the EDDS solution; and in the last set20 mg L�1 of humic acid as DOC was added. The source of humicacid sodium salt was obtained from Aldrich Chemicals. Since EDDSis susceptible to photodegradation and biodegradation, 200 mg L�1

of sodium azide was added and the EDDS solutions were kept indark at 4 �C before use.

Batch kinetic experiments were performed by mixing contami-nated soil with EDDS solution (with or without humic acid addi-tion) at a soil-to-solution ratio of 1:20 (i.e., 50 g L�1) under roomtemperature. The EDDS concentrations were adjusted to corre-spond to an EDDS-to-metal ratio of 0.5, which means that totalnumber of mol of EDDS applied equals 50% of the total numberof mol of total metal (Cu, Zn and Pb) in the soil, reflecting EDDS-deficiency conditions during in situ EDDS applications (e.g., soilflushing or phytoextraction). The concentration of EDDS added inthis study was 1.09 mM. The mixtures were rotated end-over-end at 26 rpm for various reaction times up to 48 h. The sampleswith reaction times between 1 min and 2 h were filtered usingpolypropylene syringes with 0.2-lm cellulose acetate filters; theremaining with longer than 2 h reaction time were centrifuged(Beckman Auegra 6 centrifuge) at 3600 rpm for 15 min and fil-tered. The collected supernatant solutions were stored in ambervials at 4 �C and then measured by AAS for Al, Cu, Fe, Pb, and Znconcentrations and by total organic carbon analyzer (ShimadzuTOC-5000A) for DOC concentrations. All experiments were run induplicate to ensure reproducibility and reliability.

2.3. Combined use of EDDS and EDTA

The same background solution (10 mM NaNO3 and 2 mM MES)and solution pH (5.5) were used. Three types of chelant solutions(EDDS, EDTA, or the combination of EDDS and EDTA) were pre-pared. The EDTA solution was prepared from disodium salt of EDTA(Na2EDTA, Sigma–Aldrich) and the EDDS solution was prepared asabove. The EDDS and EDTA (1:1) mixture was prepared by addingequimolar amounts of EDDS and EDTA. The total chelant concen-trations of each of these solutions corresponded to the chelant-to-metal molar ratio of 0.5 in the batch experiments. No humicacid was introduced into the chelant solutions. The batch kineticexperiments and metal measurements were conducted in dupli-cate as previously described.

3. Results and discussion

3.1. Metal extraction under EDDS deficiency

Under the condition of EDDS deficiency, the patterns of kineticextraction varied for different types of heavy metals (from 1 min to48 h in Fig. 1; and from 1 min to 6 h in Fig. SM-1). About 26% of Cuwas extracted in 1 min, and the extraction efficiency increased to70% in 60 min. The Cu extraction considerably slowed down after-wards and reached equilibrium at about 80% after 12 h. The extrac-tion of Zn and Pb, in contrast, reached maximum and thendecreased with time. The extraction efficiency of Zn was 30% ini-tially, rose to its peak of 40% at 2 h, and then started to declinesteadily to about 32% at 48 h. The highest extraction percentageof Pb was 40% at 1 min; Pb extraction continuously decreased with

Fig. 1. Influence of humic acid on metal extraction by EDDS: (a) Cu; (b) Zn; and (c)Pb ( EDDS only; EDDS + 10 mg L�1 humic acid; EDDS + 20 mg L�1 humic acid;EDDS-to-metal molar ratio of 0.5; pH 5.5; error bars represent the standarddeviations of duplicate samples).

418 T.C.M. Yip et al. / Chemosphere 80 (2010) 416–421

time to about 26% at 48 h. Therefore, while the extraction of Cu in-creased to reach equilibrium, the extraction of Zn decreased after2 h and the extraction of Pb decreased immediately after 1 min.These observations were in line with previous findings (Yip et al.,2009b).

It is noted that Cu, Zn and Pb were extracted marginally bybackground solution only (Yip et al., 2009b); indicating that theirextraction shown in Fig. 1 was primarily driven by complexationwith EDDS. Moreover, a part of Cu, Zn and Pb extraction took placeinstantaneously, suggesting that even under EDDS deficiency thereis no obvious preference of metal extraction based on the stabilityconstants of respective metal–EDDS complexes in solution. Theselectivity of metal extraction is more likely to be determined bythe lability of sorbed metals that depend on binding strength, ste-ric distribution, and chemical speciation (Yip et al., 2009a). Never-theless, the observed decrease of Zn and Pb extraction signified there-adsorption of extracted Zn and Pb, which predominantly resultsfrom metal exchange of newly formed ZnEDDS2� and PbEDDS2�

with sorbed Cu on the soil surfaces (Tsang et al., 2009). Comparedwith Zn extraction, earlier decrease of Pb extraction was in agree-ment with more significant metal exchange of Pb-EDDS2�. Thechanges of Zn and Pb extraction reached a plateau with a longerreaction time, showing an apparent equilibrium between the rateof metal extraction and the rate of re-adsorption due to metalexchange.

As shown in previous studies, Zn and Pb would dissociate fromEDDS complexes as a result of metal exchange on the soil surfacesand the liberated Zn and Pb would complex with dissolved organicmatter or exist as free metal ions in solution (Tsang et al., 2009; Yipet al., 2009b). The presence of metals that were dissociated fromEDDS was corroborated by the findings that the molar sum of ex-tracted metals exceeded the molar concentration of EDDS in solu-tion by a factor range of 1.05–1.36 after 5 min (Fig. SM-2). Thus, therole of dissolved organic matter is crucial for metal re-adsorptionon the soil surfaces, which was investigated in the subsequentsection.

3.2. Influence of dissolved organic matter

The experimental results showed that the presence of dissolvedorganic matter (10 and 20 mg L�1 as DOC) in EDDS solution led to adecrease in the extraction efficiency of Cu, Zn and Pb (Fig. 1 andFig. SM-1). With the addition of humic acid, Cu extraction slightlydecreased by less than 10%; Zn extraction displayed an earlier peakat 15 min instead of 2 h and the following extraction was reducedby about 5%. The influence on Pb extraction was the most signifi-cant, which decreased by 5–8% initially and by about 15% after48 h. The adverse effects of adding 10 and 20 mg L�1 humic acidas DOC into solution were comparable for Cu and Zn extraction,whereas Pb extraction was slightly more inhibited by the higherconcentration of humic acid.

The general extraction trends of Cu, Zn and Pb were similarregardless of the introduction of dissolved humic acid. This contra-dicts the hypothesis that dissolved organic matter would bind withZn and Pb that are dissociated from EDDS due to metal exchange,and in turn, suppress the extent of metal re-adsorption onto thesoil surfaces. Since surface soils develop a net negative charge overa wide pH range due to low point of zero charge, humic acidadsorption onto the soil surfaces may not be high quantitatively,but if any, is suspected to be strongly bound and hinder the metalextraction (Tipping, 2002; Sparks, 2003).

A supplementary experiment demonstrated that the total or-ganic carbon concentrations in solution decreased with time andabout 75% of humic acid as DOC was adsorbed after 24 h(Fig. SM-3). Such significant adsorption of humic acid onto the soilsurfaces presents three possible mechanisms that inhibit metalextraction. The first one is the restricted access of sorbed metalsfor EDDS complexation. Adsorbed humic acid develops differentconformation on the mineral surfaces, such as flat, dangling tails,ring-shaped, or micelle-like structures (Vermeer et al., 1998; Nam-jesnik-Dejanovic and Maurice, 2000), which depend on solutionchemistry and induce varying extent of steric blocking of the soilsurfaces.

The second is the competition between EDDS and sorbed humicacid for heavy metals, because acidic functional groups of humicacid, e.g., carboxyl groups, can deprotonate and bind with heavymetals, especially for Cu and Pb, as well as modify the surface elec-trostatic properties of soils (Lee et al., 2005; Saito et al., 2005). Thethird is the direct adsorption of metal–humate complexes on thesoil surfaces. Outer-sphere or inner-sphere, type A (bridged bythe metal) or type B (bridged by humic acid), ternary surface com-plexes may form depending on humic acid loading, mineral sur-faces, and solution pH, which were evidenced by spectroscopicstudies (Fitts et al., 1999; Alacio et al., 2001).

T.C.M. Yip et al. / Chemosphere 80 (2010) 416–421 419

The first mechanism (steric blocking) is comparable for thethree metals while the latter two (competition of adsorbed humicacid for heavy metals and adsorption of metal–humate complexes)are more influential for Cu and Pb due to stronger metal bindingwith humic substances (Pandey et al., 2000; Tipping, 2002). In par-ticular, Pb extraction is likely to be inhibited to a greater extentthan Cu extraction because of lower stability of PbEDDS2� andoccurrence of metal exchange that liberates Pb for binding withhumic acid.

On the other hand, it has been reported that chelating agentsalso induce dissolution of soil minerals that may reduce the soilshear strength and limit the land re-use options (Vulava and Sea-man, 2000; Tsang et al., 2007b). Negligible dissolution was ob-served in the absence of EDDS but dissolved concentrations of Aland Fe increased with time when EDDS was added (without humicacid) (Fig. 2). The dissolution of Al increased from 9.4 at 1 min upto 78.4 mg kg�1 at 3 h, from which the dissolution rate sloweddown and reached 164 mg kg�1 at 48 h; Fe dissolution increasedexponentially from 1.9 at 1 min to 14.9 mg kg�1 at 3 h, then in-creased slowly to 34.7 mg kg�1 at 48 h.

The dissolution of Al was far more significant than that of Fe,although the high stability constant of FeEDDS� suggests preferen-tial EDDS complexation with Fe. It appears that the extent of min-eral dissolution is predominantly controlled by the availability ofAl and Fe (Mayes et al., 2000; Tsang et al., 2007a; Komarek et al.,2009). This is analogous to the above discussion that the extractionpreference of sorbed metals is determined by the metal lability onthe soil surfaces. Only a minor portion of the total Al was alumi-num oxides, whereas a substantial amount may be exchangeableon clay minerals. The majority of the total Fe, on the contrary, ex-ists in the form of crystalline iron oxides, which have strong metal–oxygen bond in the lattice structure and undergo very slow disso-lution despite the aid of chelating agents (Nowack and Sigg, 1997;Nowack, 2002).

Fig. 2. Influence of humic acid on mineral dissolution by EDDS: (a) Al and (b) Fe (EDDS only; EDDS + 10 mg L�1 humic acid; EDDS + 20 mg L�1 humic acid;EDDS-to-metal molar ratio of 0.5; pH 5.5; error bars represent the standarddeviations of duplicate samples).

The Al dissolution was enhanced by additional humic acid insolution (Fig. 2), which was in agreement with previous findingsof strong Al binding with humic substances (Tipping et al., 2002;Table SM-2). The Al dissolution increased from 78 to 153 mg kg�1

at 3 h, and 164 to 217 mg kg�1 at 48 h, respectively; however, Fedissolution was promoted to a much lesser extent because of lowavailability of readily dissolvable Fe as explained above. To identifythe ability of humic acid to induce mineral dissolution, a supple-mentary experiment was conducted in the absence of EDDS(Fig. SM-4). The Al dissolution by humic acid was found to be59–66 mg kg�1 after 1 h and 83–88 mg kg�1 after 48 h, respec-tively, whereas Fe dissolution was almost undetectable. ComparingFig. 2 and Fig. SM-4, it was observed that the increment in Al dis-solution with the addition of humic acid into EDDS solution wasroughly equivalent to the Al dissolution induced by humic acidalone, suggesting that there was no competition between EDDSand humic acid for mineral dissolution.

3.3. Effectiveness of EDDS and EDTA mixture

In view of the re-adsorption of Zn and Pb extracted by EDDS, theeffectiveness of EDDS and EDTA (1:1) mixture was investigated un-der deficient chelant conditions where metal re-adsorption is mostcritical. The extraction of Cu was comparable for the three chelat-ing agent solutions (Fig. 3a), in which extraction increased rapidlyin the first 60 min followed by a slow increase to a level-off atabout 70%. Similar extraction of Cu by EDDS, EDTA, or both, was ex-pected because the stability constants of CuEDDS2� and CuEDTA2�

are comparable (Table SM-2) and the metal exchange and re-adsorption of Cu complexes are minimal (Tsang et al., 2009).

The extraction of Zn by EDTA, in contrast to that by EDDS, de-creased slowly from initial 33% to equilibrium at 28% after 6 h(Fig. 3b). This was attributed to a smaller degree of metal exchangeof ZnEDTA2� because it is more stable than ZnEDDS2� according tothe corresponding stability constants (Table SM-2). Similarly, Znextraction by the mixture of EDDS and EDTA exhibited a gentle de-crease from the peak value of 34% at 60 min to 31% at equilibrium.Akin to Cu extraction, the extraction effectiveness of Zn by theEDDS and EDTA mixture lied between those for the individualadditions of EDDS or EDTA.

Contrary to extraction by EDDS, the amount of Pb extracted byEDTA reached a plateau of 50% at 1 h and showed no obvious re-adsorption throughout 48 h (Fig. 3c). The high stability ofPbEDTA2�, of which the stability constant is nearly as high as thatof CuEDDS2� (Table SM-2), reduces the likelihood of subsequentmetal exchange. This also contributes to the lower extraction ofZn and Cu by EDTA compared to those by EDDS, because more EDTAis consumed for Pb complexation and newly formed PbEDTA2� isnot available for Zn and Cu extraction by metal exchange.

More interestingly, Pb extraction by the EDDS and EDTA mix-ture even outperformed the extraction by EDTA only; reachingthe higher extraction efficiency of 62% at 48 h. A previous studyon phytoextraction suggested that the major role of EDDS in thecombined application of EDDS and EDTA might be to increase theuptake and translocation of Pb from the roots to the shoots ofplants (Luo et al., 2006) instead of metal extraction. However, it ap-pears in this study that metal extraction can also be enhanced be-cause both EDDS and EDTA in the mixture are utilized in a moreefficient way according to their respective affinity towards metals,where Pb is primarily complexed with EDTA while Cu and Zn bindwith EDTA and EDDS. The change in chemical speciation of chelat-ing agents and heavy metals possibly accounts for less competitionfor a particular chelant, and in turn, less metal exchange and metalre-adsorption. This is corroborated by the highest molar sum of Cu,Zn and Pb extracted by the EDDS and EDTA mixture compared withthose by the individual chelant applications (Fig. SM-5). These

Fig. 3. Metal extraction by chelant mixture: (a) Cu; (b) Zn; and (c) Pb ( EDDS only;EDTA only; EDDS + EDTA mixture; chelant-to-metal molar ratio of 0.5; pH 5.5;

error bars represent the standard deviations of duplicate samples).

Fig. 4. Mineral dissolution by chelant mixture: (a) Al and (b) Fe ( EDDS only;EDTA only; EDDS + EDTA mixture; chelant-to-metal molar ratio of 0.5; pH 5.5;error bars represent the standard deviations of duplicate samples).

420 T.C.M. Yip et al. / Chemosphere 80 (2010) 416–421

findings suggest a synergistic use of the chelant mixture, which isworth further investigation into chemical speciation and optimumEDDS-to-EDTA ratio taking into account the corresponding bio-availability and toxicity.

In contrast to substantial Al dissolution by EDDS, the combineduse of EDDS and EDTA followed the trend of EDTA and reachedequilibrium of 40 mg kg�1 at 48 h (Fig. 4a). A suppression of Al dis-solution was in line with the fact that the mixture of EDDS andEDTA had higher extraction strength for target metals, thus allow-ing less complexation with Al and leading to metal exchange of thenewly formed Al complexes that are relatively weak. Nevertheless,Fe dissolution was similar for the three chelant additions, with thelargest difference of about 10 mg kg�1 (Fig. 4b). It seems that Fedissolution is not much affected by the choice of chelating agentsbecause ligand-promoted dissolution was primarily controlled bythe strong metal–oxygen bond in crystalline iron oxides.

4. Engineering implications

Under EDDS-deficiency conditions that are relevant to in situchelant applications, it is difficult to effectively extract Zn and

Pb because they are subject to metal exchange and re-adsorptiononto the soil surfaces. Although there is a high affinity betweendissolved organic matter and heavy metals, it is interesting tonote that the addition of humic acid probably encumber the me-tal extraction by EDDS. This is probably because the adsorption ofhumic acid may provide additional sorption sites or reduce theaccessibility of sorbed metals, and the metal–humate complexescan adsorb onto the soil surfaces. The effects are more noticeablefor Pb than Zn, in line with the greater affinity of Pb towards or-ganic matter. However, Cu extraction is less affected despitestrong binding with organic matter, because newly extracted Cuis least likely to be dissociated from EDDS complexes due to me-tal exchange even if EDDS is in deficiency. Therefore, one shouldbe cautious about Pb extraction when EDDS is employed in a con-taminated site that has a high content of organic matter in thesoils or aquifer.

On the other hand, the equimolar EDDS and EDTA mixture dis-plays the highest overall metal extraction, especially for Pb, com-pared to those of EDDS or EDTA. The enhancement probablyresults from the change of chelant and metal speciation that leadsto less competition for a particular chelant. The best advantage isthat a large amount of Pb can be extracted while Zn and Cu extrac-tion is marginally affected. The effectiveness of the EDDS and EDTAmixture with varying EDDS:EDTA ratio and under different geo-chemical conditions is worth further investigation.

Acknowledgement

The authors wish to thank the Research Grant Council of HongKong for providing financial support under General Research Fundwith Project Account 616608 for this research study.

T.C.M. Yip et al. / Chemosphere 80 (2010) 416–421 421

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.chemosphere.2010.03.033.

References

Alacio, T.E., Hesterberg, D., Chou, J.W., Martin, J.D., Beauchemin, S., Sayers, D.E.,2001. Molecular scale characteristics of Cu(II) bonding in goethite–humatecomplexes. Geochim. Cosmochim. Acta 65, 1355–1366.

Finzgar, N., Zumer, A., Lestan, D., 2006. Heap leaching of Cu contaminated soil with[S,S]-EDDS in a closed process loop. J. Hazard. Mater. B135, 418–422.

Fitts, J.P., Persson, P., Brown Jr., G.E., Parks, G.A., 1999. Structure and bonding ofCu(II)–glutamate complexes at the c-Al2O3–water interface. J. Colloid InterfaceSci. 220, 133–147.

Hauser, L., Tandy, S., Schulin, R., Nowack, B., 2005. Column extraction of heavymetals from soils using the biodegradable chelating agent EDDS. Environ. Sci.Technol. 39, 6819–6824.

Kent, D.B., Davis, J.A., Joye, J.L., Curtis, G.P., 2008. Influence of variable chemicalconditions on EDTA-enhanced transport of metal ions in mildly acidicgroundwater. Environ. Pollut. 153, 44–52.

Komarek, M., Vanek, A., Szakova, J., Balik, J., Chrastny, V., 2009. Interactions of EDDSwith Fe- and Al-(hydr)oxides. Chemosphere 77, 87–93.

Lee, Y.J., Elzinga, E.J., Reeder, R.J., 2005. Cu(II) adsorption at the calcite–waterinterface in the presence of natural organic matter: kinetic studies andmolecular-scale characterization. Geochim. Cosmochim. Acta 69, 49–61.

Luo, C., Shen, Z., Li, X., 2005. Enhanced phytoextraction of Cu, Pb, Zn and Cd withEDTA and EDDS. Chemosphere 59, 1–11.

Luo, C., Shen, Z., Li, X., Baker, A.J.M., 2006. Enhanced phytoextraction of Pb and othermetals from artificially contaminated soils through the combined application ofEDTA and EDDS. Chemosphere 63, 1771–1784.

Mayes, M.A., Jardine, P.M., Larsen, I.L., Brooks, S.C., Fendorf, S.E., 2000. Multispeciestransport of metal–EDTA complexes and chromate through undisturbedcolumns of weathered fractured saprolite. J. Contam. Hydrol. 45, 243–265.

Meers, E., Tack, F.M.G., Verloo, M.G., 2008. Degradability ofethylenediaminedisuccinic acid (EDDS) in metal contaminated soils:Implications for its use soil remediation. Chemosphere 70, 358–363.

Namjesnik-Dejanovic, K., Maurice, P.A., 2000. Conformations and aggregatestructures of sorbed natural organic matter on muscovite and hematite.Geochim. Cosmochim. Acta 65, 1047–1057.

Nowack, B., 2002. Environmental chemistry of aminopolycarboxylate chelatingagents. Environ. Sci. Technol. 36, 4009–4016.

Nowack, B., Schulin, R., Robinson, B.H., 2006. Critical assessment of chelant-enhanced metal phytoextraction. Environ. Sci. Technol. 40, 5225–5232.

Nowack, B., Sigg, L., 1997. Dissolution of Fe(III) (hydr)oxides by metal–EDTAcomplexes. Geochim. Cosmochim. Acta 61, 951–963.

Nowack, B., VanBriesen, J.M., 2005. Biogeochemistry of Chelating Agents. AmericanChemical Society, Washington, DC.

Pandey, A.K., Pandey, S.D., Misra, V., 2000. Stability constants of metal–humic acidcomplexes and its role in environmental detoxification. Ecotoxicol. Environ. Saf.47, 195–200.

Polettini, A., Pomi, R., Rolle, E., 2007. The effect of operating variables on chelant-assisted remediation of contaminated dredged sediment. Chemosphere 66,866–877.

Saito, T., Koopal, L.K., Nagasaki, S., Tanaka, S., 2005. Analysis of copper binding in theternary system Cu2+/humic acid/goethite at neutral to acidic pH. Environ. Sci.Technol. 39, 4886–4893.

Sarkar, D., Andra, S.S., Saminathan, S.K.M., Datta, R., 2008. Chelant-aidedenhancement of lead mobilization in residential soils. Environ. Pollut. 156,1139–1148.

Sparks, D.L., 2003. Environmental Soil Chemistry, second ed. Academic Press,Amsterdam.

Tandy, S., Ammann, A., Schulin, R., Nowack, B., 2006. Biodegradation and speciationof residual SS-ethylenediaminedisuccinic acid (EDDS) in soil solution left aftersoil washing. Environ. Pollut. 142, 191–199.

Tandy, S., Bossart, K., Mueller, R., Ritschel, J., Hauser, L., Schulin, R., Nowack, B., 2004.Extraction of heavy metals from soils using biodegradable chelating agents.Environ. Sci. Technol. 38, 937–944.

Tipping, E., 2002. Cation Binding by Humic Substances. Cambridge University Press,New York.

Tipping, E., Castro, C.R., Bryan, S.E., Taylors, J.M., 2002. Al(III) and Fe(III) binding byhumic substances in freshwaters, and implications for trace metal speciation.Geochim. Cosmochim. Acta 66, 3211–3224.

Tsang, D.C.W., Lo, I.M.C., Chan, K.L., 2007a. Modeling the transport of metals withrate-limited EDTA-promoted extraction and dissolution during EDTA-flushingof copper-contaminated soils. Environ. Sci. Technol. 41, 3660–3667.

Tsang, D.C.W., Yip, T.C.M., Lo, I.M.C., 2009. Kinetic interactions of EDDS with Soils: 2.Metal–EDDS complexes in uncontaminated and metal-contaminated soils.Environ. Sci. Technol. 43, 837–842.

Tsang, D.C.W., Zhang, W., Lo, I.M.C., 2007b. Copper extraction effectiveness and soildissolution issues of EDTA-flushing of artificially contaminated soils.Chemosphere 68, 234–243.

Vandevivere, P.C., Hammes, F., Verstraete, W., Feijtel, T., Schowanek, D., 2001. Metaldecontamination of soil, sediment, and sewage sludge by means of transitionmetal chelant [S,S]-EDDS. J. Environ. Eng. 127, 802–811.

Vermeer, A.W.P., van Riemsdijk, W.H., Koopal, L.K., 1998. Adsorption of humic acidto mineral particles. 1. Specific and electrostatic interaction. Langmuir 14,2810–2819.

Vulava, V.M., Seaman, J.C., 2000. Mobilization of lead from highly weathered porousmaterial by extracting agents. Environ. Sci. Technol. 34, 4828–4834.

Yip, T.C.M., Tsang, D.C.W., Ng, K.T.W., Lo, I.M.C., 2009a. Empirical modeling of heavymetal extraction by EDDS from single-metal and multi-metal contaminatedsoils. Chemosphere 74, 301–307.

Yip, T.C.M., Tsang, D.C.W., Ng, K.T.W., Lo, I.M.C., 2009b. Kinetic interactions of EDDSwith Soils: 1. Metal resorption and competition under EDDS deficiency. Environ.Sci. Technol. 43, 831–836.

Zhang, W., Tsang, D.C.W., Lo, I.M.C., 2007. Removal of MDF and Pb fromcontaminated soils by EDTA- and SDS-enhanced washing. Chemosphere 66,2025–2034.