Copper and trace element fractionation in electrokinetically treated methanogenic anaerobic granular sludge Jurate Virkutyte a,b , Eric van Hullebusch a , Mika Sillanpa¨a¨ b , Piet Lens a, * a Subdepartment of Environmental Technology, University of Wageningen, ‘‘Biotechnion’’-Bomenweg 2, PO Box 8129, 6700 EV Wageningen, The Netherlands b University of Kuopio, Department of Environmental Sciences, PO Box 1627, FIN-70211, Kuopio, Finland Received 19 July 2004; accepted 8 April 2005 Electrokinetic treatment of copper contaminated anaerobic granular sludge at 0.15 mA cm ÿ2 for 14 days induces copper and trace metal mobility as well as changes in their fractionation (i.e. bonding forms). Abstract The effect of electrokinetic treatment (0.15 mA cm ÿ2 ) on the metal fractionation in anaerobic granular sludge artificially contaminated with copper (initial copper concentration 1000 mg kg ÿ1 wet sludge) was studied. Acidification of the sludge (final pH 4.2 in the sludge bed) with the intention to desorb the copper species bound to the organic/sulfides and residual fractions did not result in an increased mobility, despite the fact that a higher quantity of copper was measured in the more mobile (i.e. exchangeable/ carbonate) fractions at final pH 4.2 compared to circum-neutral pH conditions. Also addition of the chelating agent EDTA (Cu 2C :EDTA 4ÿ ratio 1.2:1) did not enhance the mobility of copper from the organic/sulfides and residual fractions, despite the fact that it induced a reduction of the total copper content of the sludge. The presence of sulfide precipitates likely influences the copper mobilisation from these less mobile fractions, and thus makes EDTA addition ineffective to solubilise copper from the granules. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Fractionation; Copper; Electrical treatment; Granular sludge 1. Introduction Electrokinetic remediation techniques are widely used to separate and extract charged contaminants from soils (Hamed et al., 1991; Acar and Alshawabkeh, 1993; Van Cauwenberghe, 1997; Yeung et al., 1997; Maini et al., 2000), sludge (Zagury et al., 1999; Kim et al., 2002; Jakobsen et al., 2004) or timber waste (Velizarova et al., 2002) by employing a low-level direct current in the range of several mA cm ÿ2 . The low-level current induces, via electromigration and electro-osmosis, the mobility of heavy metals, thus removing charged pollutants from the contaminated matrices (Hamed et al., 1991; Acar and Alshawabkeh, 1993; Mattson and Lindgren, 1995; Virkutyte et al., 2002; Kim et al., 2002). While there is a vast amount of data about the application of electrokinetic treatments to remove heavy metals from contaminated sludges, there is a lack of information of the heavy metal binding characteristics in electrokinetically treated sludges. Copper and trace elements are essential elements but are potentially toxic at high concentrations and may produce deficiency symptoms at very low concentration in the environment (Shrivastava and Banerjee, 1998). Heavy metals occur in sludges in various physico- chemical forms, such as soluble, adsorbed, exchange- able, precipitated, organically complexed and residual * Corresponding author. Tel.: C31 317 483339; fax: C31 317 482108. E-mail address: [email protected](P. Lens). 0269-7491/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.envpol.2005.04.009 Environmental Pollution 138 (2005) 517e528 www.elsevier.com/locate/envpol
12
Embed
Copper and trace element fractionation in electrokinetically treated methanogenic anaerobic granular sludge
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Environmental Pollution 138 (2005) 517e528
www.elsevier.com/locate/envpol
Copper and trace element fractionation in electrokineticallytreated methanogenic anaerobic granular sludge
Jurate Virkutyte a,b, Eric van Hullebusch a, Mika Sillanpaa b, Piet Lens a,*
a Subdepartment of Environmental Technology, University of Wageningen, ‘‘Biotechnion’’-Bomenweg 2,
PO Box 8129, 6700 EV Wageningen, The Netherlandsb University of Kuopio, Department of Environmental Sciences, PO Box 1627, FIN-70211, Kuopio, Finland
Received 19 July 2004; accepted 8 April 2005
Electrokinetic treatment of copper contaminated anaerobic granular sludge at 0.15 mA cm�2 for 14 days inducescopper and trace metal mobility as well as changes in their fractionation (i.e. bonding forms).
Abstract
The effect of electrokinetic treatment (0.15 mA cm�2) on the metal fractionation in anaerobic granular sludge artificiallycontaminated with copper (initial copper concentration 1000 mg kg�1 wet sludge) was studied. Acidification of the sludge (final
pH 4.2 in the sludge bed) with the intention to desorb the copper species bound to the organic/sulfides and residual fractions did notresult in an increased mobility, despite the fact that a higher quantity of copper was measured in the more mobile (i.e. exchangeable/carbonate) fractions at final pH 4.2 compared to circum-neutral pH conditions. Also addition of the chelating agent EDTA
(Cu2C:EDTA4� ratio 1.2:1) did not enhance the mobility of copper from the organic/sulfides and residual fractions, despite the factthat it induced a reduction of the total copper content of the sludge. The presence of sulfide precipitates likely influences the coppermobilisation from these less mobile fractions, and thus makes EDTA addition ineffective to solubilise copper from the granules.� 2005 Elsevier Ltd. All rights reserved.
Electrokinetic remediation techniques are widely usedto separate and extract charged contaminants from soils(Hamed et al., 1991; Acar and Alshawabkeh, 1993; VanCauwenberghe, 1997; Yeung et al., 1997; Maini et al.,2000), sludge (Zagury et al., 1999; Kim et al., 2002;Jakobsen et al., 2004) or timber waste (Velizarova et al.,2002) by employing a low-level direct current in therange of several mA cm�2. The low-level currentinduces, via electromigration and electro-osmosis, the
0269-7491/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.envpol.2005.04.009
mobility of heavy metals, thus removing chargedpollutants from the contaminated matrices (Hamedet al., 1991; Acar and Alshawabkeh, 1993; Mattsonand Lindgren, 1995; Virkutyte et al., 2002; Kim et al.,2002). While there is a vast amount of data about theapplication of electrokinetic treatments to remove heavymetals from contaminated sludges, there is a lack ofinformation of the heavy metal binding characteristics inelectrokinetically treated sludges.
Copper and trace elements are essential elements butare potentially toxic at high concentrations and mayproduce deficiency symptoms at very low concentrationin the environment (Shrivastava and Banerjee, 1998).Heavy metals occur in sludges in various physico-chemical forms, such as soluble, adsorbed, exchange-able, precipitated, organically complexed and residual
518 J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528
phases (Hayes and Theis, 1978; Angelidis and Gibbs,1989). Analytical limitations, which are imposed byselectivity and interference, do not allow a completedifferentiation between different physicochemical formsof metals in the sludge matrix (Hsiau and Lo, 1998). Thelatter can, however, be approximated by the fraction-ation of heavy metals in the solid matrix of sludges, forinstance, using the revised BCR sequential extractionscheme, as proposed by Mossop and Davidson (2003).
Sequential extraction procedures have been developedpredominantly to determine the amounts and partition-ing of metals present within soils or sediments samples,sewage sludge and sludge treated soils (see Filgueiraset al., 2002 for review). Fractionation procedures areoften criticised because of their complexity and difficultyin interpretation, arising from potential problems such aslack of specificity of extractants and re-adsorption ofmetals during extraction. Nevertheless, providing thatsuch limitations are recognised, sequential fractionationscan provide extremely useful information on metaldistribution in sludges, particularly for comparativepurposes (McLaren and Clucas, 2001).
In the present study, mesophilic anaerobic granularsludge was chosen as a model for anaerobic sludges. Themain objective was to evaluate the changes in the copperfractionation after different electrokinetic treatmentstrategies (i.e., exposure to different pH values, complex-ation with EDTA, pre-incubation) of artificially coppercontaminated sludge granules. The effect of the electro-kinetic treatment on the fractionation of macro and traceelements in the sludge cake was also investigated.
2. Materials and methods
2.1. Electrokinetic set-up
A ‘closed’ laboratory-scale electrokinetic cell wasused in this study as described by Ottosen and Hansen(1992). The cell consisted of a cylindrical glass container(diameter 20 cm, length 26 cm) with a distance betweenthe electrodes of 17 cm. In the electrokinetic cell (Fig. 1),the central compartment from the anode and cathodecompartments was separated by, respectively, anion-exchange (IA1-204SXZL386) and cation-exchange (IC1-61CZL386) membranes (Ionics Inc, Watertown, MA,USA). The anode and cathode were immersed intoa 0.05 M KNO3 conductive solution (Fig. 1). Theelectrodes (diameter 3 mm, length 5 cm) were titan bars(supplied by Elektronika-WUR, The Netherlands).
In the electrokinetic cell, a power supply (Hewlett-Packard 613 Altai, Germany) was used to constantlymaintain a 0.15 mA cm�2 DC current. The voltagefluctuations were monitored with a Fluke 112 multi-meter (Fluke, Eindhoven, The Netherlands).
Peristaltic pumps (Marlow 502S) allowed a recircula-tion of the electrolyte solutions. The average flow rate inthe electrolytes was maintained at 5 ml min�1.
2.2. Source of biomass
Anaerobic granular sludge was obtained from afull-scale UASB reactor (Industriewater Eerbeek B.V.,Eerbeek, The Netherlands) treating paper-mill waste-water (Lens et al., 1999). The sludge had a mean den-sity of 1040 kg m�3. The initial pH of the sludge was 7.1.The total suspended solid (TSS) and volatile suspendedsolid (VSS) concentrations of the sludge were 22(G0.2)% and 73.9 (G0.2)%, respectively. The carbo-nates concentration was 0.4 (G0.2)% of TSS and totalsulfur was 41.8 (G1.0) mg g�1 TSS (van Hullebuschet al., 2005). The background copper concentrationpresent in this sludge amounted to 150 mg kg�1 TSS(Osuna et al., 2004).
2.3. Experimental design
The amount of sludge placed in the electrokinetic cellwas 1000 g (wet sludge). It was artificially contaminatedwith 3.78 g of Cu(NO3)2, which gave 1000 mg kg�1 (wetsludge) of copper. The contamination procedure was asfollows: 1000 g of anaerobic granular sludge amendedwith the specific Cu(NO3)2 and Na2H2EDTA concen-tration was placed into a closed plastic container for 48 h(Reddy et al., 1997). The content of the plastic containerwas frequently and thoroughly mixed. Upon terminationof the contamination procedure, the sludge, still sus-pended in the copper (EDTA) containing supernatant,was mounted in the electrokinetic set-up. In the EDTAamended experiments, 3.78 g of Na2H2EDTA was addedsimultaneously with Cu(NO3)2 (molar ratio of Cu2C:Na2H2EDTA at 1.2:1). When testing the effect of sludgepre-incubation, the sludge was incubated in an anaerobicchamber with 3.78 g of Cu(NO3)2 for 30 days, eventuallysupplemented with 3.78 g of Na2H2EDTA.
The sludge was put in the electrokinetic cell for 14days in all experiments. This period was chosen becauseafter 10 days, the voltage had increased significantly (upto 90 V), remained constant for at least 2 days and thendecreased down to 10 V. The low voltage indicates thatthe entire acidic front had passed through the sludgecake. After the electrokinetic treatment, the sludge cakewas split into three portions, i.e. a portion close to theanode, a portion close to the cathode and a mid portionbetween them. Each portion was homogeneously mixedbefore analysis.
The initial pH of the sludge cake in the electrokineticcell was 7.1. When the pH fluctuations were uncon-trolled in the electrolytes, it reached 12.5 in thecatholyte. This gave a final pH of 7.7 in the sludgecake upon termination of the electrokinetic treatment.
519J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528
K+
H+
NO3
-
OH-
Cu2+
Metal ions
Anaerobic sludge
Anode
Cathode
5cm 5cm
Conductive solution Conductive solution
20cm
26cm
DC
Multimeter
Pump
Pump
Power supply
Anolyte Catholyte
Ion exchange
membrane
Ion exchange
membrane
MeEDTA2-
Fig. 1. Schematic representation of the electrokinetic cell used in this study.
When the pH in the catholyte was set to 2.5 by 1 MHNO3 addition, the pH in the sludge cake was 4.2 upontermination of the electrokinetic treatment.
2.4. Analytical techniques
2.4.1. Pseudo-total metal analysisThe pseudo-total metal content (expressed as mg
metal kg�1 wet sludge) of sludge was determined bydigestion with the aqua regia procedure (Florian et al.,1998). For metal determination, 0.5 g TSS of anaerobicgranular sludge were treated with 10 ml of aqua regia inTeflon� perfluoroalkoxy resin (PFA) digestion vessels, ina temperature controlled microwave oven MilestoneETHOS E (Milestone Inc; Monroe, CT, USA). Thesample volume was completed to 100 ml with ultrapurewater. After digestion, the concentrations of total metalswere analysed using an atomic absorption spectroscopy(AAS) flame (Perkin-Elmer 300) or inductively coupledplasma optical emission spectrometry (ICP-OES; VarianMPX CCD, Vista, Australia). The following wave-lengths were used: 228.802, 213.598, 259.940, 216.555,
202.548, 422.673 and 279.553 nm for Co, Cu, Fe, Ni, Zn,Ca and Mg, respectively, for ICP-OES measurements.
2.4.2. Revised sequential extraction (BCR) schemeThe revised BCR scheme is designed based on an
acetic acid extraction of approximately 1 g TSS ofgranular sludge (step 1), hydroxylamine hydrochlorideextraction (step 2) and hydrogen peroxide oxidation andammonium acetate extraction (step 3) as described byMossop and Davidson (2003). To extract the residualphase (step 4), a mixture of 2.5 ml HNO3 (65%) and7.5 ml HCl (37%) (aqua regia digestion) was addedto the residue and the filter from fraction 3. Aftermicrowave destruction, the samples of step 4 were paperfiltered and diluted to 100 ml with ultra pure water. Thechemicals used for the extraction are presented in Table1. These extractions are associated with the exchange-able/carbonate (bound to carbonate, step 1), oxidesphase (bound to iron and manganese oxides, step 2) andthe organic/sulfides phase (bound to organic matter/sulfides, step 3).
Table 1
The chemicals used for the sequential extraction to determine the copper speciation in the sludge fractions after the electrodialytic experiment
Fraction Extracting agent Extraction conditions
Shaking timea Temperature ( �C)
1. ExchangeableCcarbonates 40 ml CH3COOH (0.11 M, pHZ7) 16 h 20
2. Iron and manganese oxides 40 ml NH2OH-HCl (0.5 M, pHZ1.5) 16 h 20
3. Organic matter and sulfides 20 ml H2O2 (30%, pHZ2) and then 50 ml CH3COONH4 (1 M, pHZ2) 1 h; 2 h; 16 h 20; 85; 20
4. Residual 10 ml demineralised water and 10 ml aqua regia (HCl/HNO3, 3:1) 26 min Microwave ovenb
a Shaking was applied at 30 rpm for 16 h.b Extraction of the residual fraction in the microwave was equal to the pseudo-total extraction method.
2.4.3. Evaluation of analytical performanceThe analytical performance of the laboratory proce-
dures was evaluated by analysis of Certified ReferenceMaterial BCR-701 and CRM 146R. A two-sided t-testwas used to check for significant differences from thereference content. Table 2 shows the data of threereplicate analyses obtained for aqua regia extraction andTable 3 shows the revised BCR sequential extractionprocedure, expressed as mg kg�1 of dry mass. Un-certainty is expressed as standard deviations; the valuesobtained are not significantly (PO0.05) different fromthe certified values. The data for pseudo-totals andfractionation of copper and other elements wereobtained from triplicates (nZ3) with meanGstandarddeviation. The TSS and VSS concentrations weredetermined according to the APHA standard methods(APHA, 1995).
2.4.4. Visual MinteqThe geochemical equilibrium model Visual Minteq
(freeware program at http://www.lwr.kth.se/english/OurSoftware/Vminteq/index.htm) was used to simulatethe chemical speciation of Cu in solution in two differentexperiments (without and with EDTA addition) in theabsence of anaerobic granular sludges. For calculation,the solid forms of copper were allowed to precipitate.Visual Minteq is capable of calculating equilibriumaqueous speciation, precipitation and dissolution ofminerals, complexation, adsorption, solid phase satura-tion states, etc. (Gustaffsson, 2004).
Table 3
Results obtained (meanGstandard deviation, nZ3) for aqua regia
extractable (pseudo-total) metal content of CRM 146R sewage sludge
from industrial origin
Found value
(mg kg�1 DW)
Certified value
(mg kg�1 DW)
Co 6.35G0.09 6.50G0.31
Cu 744G7 831G16
Mn 278G4 298G9
Ni 53.3G0.9 65.0G3.0
Zn 2887G29 3043G58
3. Results
3.1. Initial fractionation of copper inanaerobic granular sludge
Prior to electrokinetic treatment (at pH 7.1), copperwas mainly associated with the exchangeable/carbonate(310 mg kg�1) fraction of fresh copper supplementedgranular sludge (Fig. 3a). The remaining copper wasspread equally over the other fractions: 190 mg kg�1 ofCu in the residual and organic/sulfides and 200 mg kg�1
in the oxides fractions (Fig. 3a). Prolonged exposure ofsludge to copper (30 days) prior to the electrokinetictreatment modified the copper fractionation. The mostabundant fractions in the pre-incubated sludge were,respectively, the oxides (330 mg kg�1) and the residual(260 mg kg�1) fractions (Fig. 3b). Addition of Cu andEDTA simultaneously resulted in a decrease in allfractions (Fig. 4), except in the exchangeable/carbonatefraction (310 mg kg�1), compared to the non-EDTAamended sludge (Fig. 3).
3.2. Effect of electric current on copper fractionation
Fig. 2 shows the development of the electric potentialgradient across the sludge matrix with time. The highestelectrical potential gradient applied at a 5 ml min�1 flowrate of the electrolyte solutions was 3.5 V cm�1. Duringthe experiments, the gradient increased from 0.04 to3.5 V cm�1 and then decreased to 1 V cm�1 (Fig. 2).
0
0.5
1
1.5
2
2.5
3
3.5
111 3 5 7 9 13Time (days)E
lectrical p
oten
tial g
rad
ien
t (V
/cm
)
Fig. 2. Evolution of the electrical potential gradient with time (V cm�1).
521J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528
After the electrokinetic treatment (0.15 mA cm�2
DC current for 14 days), the fractionation of copper inthe granular sludge matrix had changed significantly(Figs. 3 and 4). A significant increase of copper in theresidual (from 190 to 300 mg kg�1) and the oxides(from 200 to 320 mg kg�1) fractions was observed forthe freshly amended sludge, which was directedtowards the cathode (Fig. 3a). Copper associated withthe exchangeable/carbonate fraction decreased from310 to 80 mg kg�1, which moved towards the anode(Fig. 3a). One of the major trends was that the copperconcentration increased in the residual fraction of mostof the experimental conditions at both the cathode andanode side (Figs. 3 and 4).
3.3. Effect of pH during the electrokinetic treatmenton the copper fractionation
The major fractions with which copper was associ-ated after electrokinetic treatment with pH 12.5 in thecatholyte solution (final pH 7.7 in the sludge bed) werethe residual (up to 310 mg kg�1) and the oxides (320 mgkg�1) fractions, which were redistributed towards thecathode (Fig. 3a).
When the pH was maintained acidic (pH 2.5) in theelectrolytes (final pH 4.2 in the sludge bed), the mostabundant copper fractions were the exchangeable/carbo-nates (330 mg kg�1) and the residual (300 mg kg�1)fractions, which moved towards the cathode (Fig. 3c).The copper associated with the organic/sulfides (up to210 mg kg�1) and oxides (up to 236 mg kg�1) fractions
remained unchanged, respectively, at the anode andcathode side (Fig. 3c).
3.4. Effect of sludge pre-incubation on thecopper fractionation
When the sludge was pre-incubated with copper for30 days, the dominant fractions were the oxides (up to320 mg kg�1), residual (290 mg kg�1) and organic/sulfides (up to 280 mg kg�1) upon the termination ofthe electrokinetic treatment at pH 12.5 in the catholyte(Fig. 3b). The copper associated with the organic/sulfides fraction had migrated towards the anode(Fig. 3b).
The major fractions of copper were the exchangeable/carbonates (390 mg kg�1) and residual (up to 320 mgkg�1) when the pH in the catholyte solution was set topH 2.5 (final pH 4.2 in the sludge cake). Interestingly,copper in the exchangeable/carbonates and the residualfractions moved towards the cathode, in contrast tocopper bound to the organic/sulfides fraction, whichmigrated towards the positively charged anode (Fig. 3d).
3.5. Effect of EDTA on the copper fractionationafter electrokinetic treatment
Table 4 shows that in the presence of EDTA, copperaccumulated at both sides compared to the initialconcentration. This is an opposite trend compared tothe non-EDTA treated sludge. When the pH in thecatholyte solution was 12.5 (final pH 7.7 in the sludgewater), there was a significant decrease in all fractions
Fig. 3. Effect of electrokinetic treatment on the copper distribution in anaerobic granular sludge artificially contaminated with Cu(NO3)2: (A) Fresh
and (B) pre-incubated sludge with pH 12.5 in the catholyte (final pH 7.7 in the sludge cake); (C) fresh and (D) pre-incubated sludge with pH 2.5 in the
catholyte (final pH 4.2 in the sludge cake).
522 J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528
Fig. 4. Effect of electrokinetic treatment on the copper distribution in the anaerobic granular sludge artificially contaminated with CuEDTA.
(A) Fresh and (B) preincubated sludge with pH 12.5 in the catholyte (final pH 7.7 in the sludge cake); (C) fresh and (D) pre-incubated sludge with
pH 2.5 in the catholyte (final pH 4.2 in the sludge cake).
(Fig. 4a) except for the exchangeable/carbonates frac-tion, which remained the same as in the copper nitrateamended sludge experiments (310 mg kg�1) (Fig. 3a).The most predominant fraction after electrokinetictreatment at pH 7.7 was the organic/sulfides fraction(from 170 to 750 mg kg�1), which had moved towardsthe positively charged anode (Fig. 4a). When the pHin the anolyte solution was 2.5 (final pH 4.2 in the sludge),the major fractions of copper in the electrokinetically
Table 4
Initial trace and major elements concentrations with Cu(NO3)2 or Cu
EDTA in the fresh anaerobic granular sludge before the electrokinetic
treatment and removal efficiencies
Initial
concentrations
Concentrations
at the anode
(mg kg�1)
and removal
efficiencies (%)
Concentrations at
the cathode
(mg kg�1)
and removal
efficiencies (%)
Cu(NO3)2 treatment
Cu 1070 620 (40) 785 (27)
Zn 125 60 (52) 78 (37)
Co 38 34 (11) 40 (accumulation)
Ni 38 36 (5) 41 (accumulation)
Ca 2100 1100 (48) 2000 (5)
Fe 25000 12000 (52) 16000 (36)
Mg 680 150 (78) 580 (15)
Cu EDTA treatment
Cu 790 1320 (accumulation) 980 (accumulation)
Zn 125 70 (44) 50 (60)
Co 38 42 (accumulation) 33 (16)
Ni 38 40 (accumulation) 36 (5)
Ca 2100 350 (83) 1900 (10)
Fe 25000 11000 (56) 19000 (24)
Mg 680 40 (94) 590 (16)
treated EDTA amended fresh sludge were the residual(650 mg kg�1) and the organic/sulfides (550 mg kg�1)fractions, which were directed towards the anode(Fig. 4c).
After electrokinetic treatment of the pre-incubatedsludges at a pH 12.5 in the catholyte (final pH 7.7 inthe sludge bed) in the presence of EDTA, copper wasmainly present in the residual (up to 680 mg kg�1)fraction, followed by the oxides (640 mg kg�1) fraction(Fig. 4b) at the anode side. At pH 2.5 in the catholyte(final pH 4.2 in the sludge bed), the most significantcopper fraction (490 mg kg�1) was the residual fractionat the anode side (Fig. 4d).
3.6. Effect of electrokinetic treatment ontrace and macroelements
3.6.1. Trace elementsInitially, the main trace metal fraction was the
residual (Fig. 5a,c,e): Zn, 78 mg kg�1; Co, 15 mg kg�1;Ni, 30 mg kg�1 (Table 4). After electrokinetic treatmentat pH 12.5 in the catholyte, the residual fractiondecreased significantly for Zn and Ni, but not for Co(Fig. 5a,c,e). There was only an increase in the organic/sulfides fraction from 20 to 38 mg kg�1 at the cathodefor Zn (Fig. 5a) and from 3 to 10 mg kg�1 for Ni, whichwas directed towards the anode (Fig. 4e). Also,a significant increase in Ni associated with the exchange-able/carbonates fraction (from 3 to 15 mg kg�1)occurred at the cathode (Fig. 5e). Co associated withthe exchangeable/carbonates fraction increased from 5to 10 mg kg�1 after the electrokinetic treatment at the
523J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528
anode side (Fig. 5c). However, the other fractions (i.e.oxides and organic/sulfides fractions) of Co remainedwithout any significant changes at the cathode or anodesides (Fig. 5c).
EDTA amendment induced a redistribution of Co(from 5 to 14 mg kg�1) and Ni (from 3 to 11 mg kg�1)towards the anode in the exchangeable/carbonatesfraction upon electrokinetic treatment. In contrast,there was a significant decrease in Zn (from 78 to15 mg kg�1) and Ni (from 30 to 5 mg kg�1) associatedwith the residual fraction (Figs. 5b and 2f).
the oxides (Fig. 6a,c,e): Ca, 1200 mg kg�1; Mg, 300 mgkg�1; and the residual for Fe (10000 mg kg�1) (Table 4).After electrokinetic treatment, the residual fraction hadincreased for Ca from 180 to 400 mg kg�1 (Fig. 6a) butdecreased for Fe from 10,000 to 6000 mg kg�1 (Fig. 6c).There was a significant decrease observed for all Feassociated fractions, which moved towards the cathode,with the exception of the residual fraction (Fig. 6c). Themost abundant Ca fraction was the exchangeable/
carbonates fraction (Fig. 6a) attributing to 900 mg kg�1,which moved towards the cathode. TheMg fractionationshowed a significant increase in the exchangeable/carbonates (300 mg kg�1) and organic/sulfides (130 mgkg�1) fractions (Fig. 6e), which had also moved towardsthe cathode.
The most significant increase for Ca, Mg and Fe inelectrokinetically treatedCuEDTAexposed sludgewas inthe organic/sulfides fraction: 600 mg kg�1, 200 mg kg�1
and 8500 mg kg�1, respectively (Fig. 6b,d,f). Caassociated with the exchangeable/carbonates fractionincreased from 600 to 700 mg kg�1 (Fig. 6b), which wasdirected towards the cathode.
4. Discussion
This study showed that electrokinetic treatment at0.15 mA cm�2 for 14 days not only induces copper andtrace metals mobility but also alters their fractionation inanaerobic granular sludge. The latter is strongly influ-enced by the pH, contaminant aging and the presence ofEDTA. The discussion section describes successively thecopper fractionation prior to electrokinetic treatment
524 J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528
Exposure of the sludge granules to an easily solublecopper salt (Cu(NO3)2) resulted in a much higher totalcopper concentration of the sludge compared to allother heavy metals. In all cases, copper was more or lessevenly distributed between the four operationally de-fined fractions of the BCR protocol. This distribution ofcopper was very similar to that observed for Cu in thenon-Cu-spiked sludges used to prepare the amendedsludge (Osuna et al., 2004). However, the proportion ofCu present in the two most easily extracted fractions(exchangeable/carbonates and oxides fractions) wasmuch larger due to increasing Cu concentrations in thesludge (Figs. 3 and 4). Both at low (Osuna et al., 2004)and high (Figs. 3 and 4) concentrations, copper showeda very strong affinity for the organic matter/sulfides andresidual fractions of the BCR scheme. It is, indeed, well
known that Cu strongly interacts with organic matterand sulfides (Pattrick et al., 1997; Lu and Allen, 2002;Vulkan et al., 2002). Copper may undergo biosorptionby cell membrane surfaces with proteins and acid groupsthat serve as binding site (Hayes and Theis, 1978). Thebinding and sequestration of copper by extracellularpolymeric substances (EPS) has also been demonstrated(White and Gadd, 2000). Besides, chalcocite (Cu2S) canbe formed in the presence of sulfide (Morse and Luther,1999) and even different Cu sulfide minerals can coexistin the anaerobic sludge matrixes (Pattrick et al., 1997).These metals bound to the sulfides are mainly leached inthe organic/sulfides fraction in natural sediment, soil orsludge (Lacal et al., 2003). Copper may be also adsorbedin large quantity at the surface of pyrite minerals(Muller et al., 2002), however little evidence has beenfound in the literature for Cu accumulation in thecrystalline lattice of the iron sulfides.
When the fresh Cu amended sludge was supple-mented with EDTA prior to electrokinetic treatment,the initial fractionation of copper changed (Fig. 4a,c)and a lower initial total copper quantity accumulated inthe sludge (Table 4). This is in agreement with Osuna
525J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528
et al. (2004), who studied the effect of EDTA on theaccumulation of Co in anaerobic granular sludge.EDTA addition increased the amount of dissolvedcopper obtained in the copper speciation simulations(Fig. 7): at pH 7, the addition of EDTA increased thepercentage of dissolved copper from 1% to 83.5% of thetotal copper initially present in the systems.
4.2. Effect of pH on copper fractionation
The pH influences the adsorption and desorption,precipitation, dissolution and speciation reactions ofcopper. At low pH, copper tends to desorb from thesludge matrix and dissolves as positively charged ions(Hsiau and Lo, 1998). All heavy metals have a specificpH underneath which their solubility is drasticallyincreased. For copper, this pH is 5.5 (Martinez andMotto, 2000). The decrease in the pH during theelectrokinetic treatment consequently promoted theformation of soluble mobile copper compounds suchas free copper (Cu2C) and CuNO3
C (Fig. 7b), whichcaused copper accumulation at the negatively chargedcathode side in the exchangeable/carbonates and re-sidual fractions of either the freshly amended (Fig. 3c)and pre-incubated (Fig. 3d) sludge. Also at neutraland slightly alkaline pH (6.5e7.5), the predominantdissolved species are positively charged compounds, e.g.
Cu2C, Cu2OH3C and CuOHC (Fig. 7b). These copperspecies can migrate towards the negatively chargedcathode, as observed for the oxides and residualfractions of the freshly amended sludges (Fig. 3a).Besides electromigration, also electro-osmotic waterflow can contribute to an increased copper content atthe cathode side. Such an electro-osmotic water flow isproduced during the electrokinetic treatment, whichmoves through the sludge cake from the anode to thecathode (Pamukcu et al., 1997).
4.3. Effect of pre-incubation on copper fractionation
After a 30 days incubation period, an increase ofcopper in the residual fraction has been noticed, verylikely due to the contamination aging effect (McLarenand Clucas, 2001). Pattrick et al. (1997) reported thatthe mobility of copper decreases with the duration (afew hours) of its contact with anaerobic sludge, due tothe formation of copper sulfide crystals. The electro-kinetic treatment of the pre-incubated copper amendedsludges mainly influenced the metal accumulated in theorganic/sulfides fraction (Fig. 3bed). Kim et al. (2002)found that the organic/sulfides fraction is relativelymobile in an electric field, which is in agreement withFig. 3b and to a lesser extent with Fig. 3d. It is welldocumented that copper has also a strong affinity to
526 J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528
form complexes with the organic matter present (Bolanet al., 2003; Lu and Allen, 2002). When copper iscontacted with an organic fraction in anaerobicconditions, it usually forms negative compounds, e.g.Cu-Org2� or Cu-Org� (Lu and Allen, 2002; Vulkanet al., 2002), which would explain the accumulation ofcopper at the anode side in the organic/sulfides fraction(Fig. 3b). For sewage sludge exposed to sulfidicconditions, the organic fraction has been shown tooriginate from the hydrolysis of EPS (Watson et al.,2004). Hydrolysis of the organic sludge matrix verylikely occurred as well during the pre-incubation step ofthe copper amended anaerobic sludge, thus yieldingorganic compounds to which copper binds.
4.4. Effect of EDTA on copper fractionation
The addition of the chelating agent EDTA to thesludge cake resulted in a change of the migrationdirection of copper in the applied electric field (Fig. 4).Complexing agents like EDTA are widely used for theenhanced removal of heavy metals from differentcontaminated media without (Nirel et al., 1998; Sillanpaaet al., 2001; Nowack, 2002) and with (Wong et al., 1997;Velizarova et al., 2002) electrokinetic treatment. EDTAhas proved to be an efficient chelating agent since it formsstable complexes with most metals, especially withtransition metals over a broad pH range (Lo and Yang,1999). It can also be utilized for both desorption ofsorbed ions and dissolution of precipitated metalcompounds (Papassiopi et al., 1999).
When the sludges were supplemented with EDTA,copper extracted in the residual and oxides fractionsexclusively increased at the anode side at both pHconditions (Fig. 4b,d) for the pre-incubated sludge afterelectrokinetic treatment. In contrast, the fresh sludgeshowed that the extracted copper increased in theorganic/sulfides and residual fractions at the anode atfinal pH 4.2 (Fig. 3c). This outlines the presence ofnegatively charged copper species in the presence ofEDTA. Indeed, according to the species distributionmodel (Fig. 7a), the main dissolved species isCuEDTA2� at neutral and slightly alkaline pH (pH6e8). At acidic pH (pH 2e4), the most abundantdissolved species are CuEDTA2� and CuHEDTA�
(Fig. 7a). The prevailing copper complexes are nega-tively charged, thus they migrate towards the positivelycharged anode (Fig. 4), which is also reported for coppermobility in clayey soils (Popov et al., 1999). It should benoted that copper also increased at the cathode side ofthe freshly copper amended sludge for a final pH of 7.7(Fig. 4a). This outlines that also positively chargedspecies were present at that pH. Indeed, according to thegeochemical model (Fig. 7a), the Cu3(OH)4
2C contentcan be expected to be significant. However, no analyticalevidence can confirm this statement.
Fig. 4 shows that EDTA is efficient to decreasethe copper content in the exchangeable/carbonates andoxides fractions of the fresh sludge. However, nodecrease was noticed in the organic/sulfide and residualfractions. This is probably due to the presence of sulfideprecipitates which are efficient sorbants. Consequently,these can prevent copper removal from the anaerobicsludge as reported by Reddy and Chinthamreddy(1999), when they introduced sulfide in kaolin.
4.5. Effect of electrokinetic treatment on tracemetals and major elements fractionation
An increase of the cobalt and nickel content, presentin much lower concentrations than Cu in the exchange-able/carbonates fraction was noticed when the final pHin the sludge bed was 7.7. The decrease in the residualfraction compared to the initial sample prior to electro-kinetic treatment (Fig. 5) might be explained by thedesorption of trace metals from the sludge crystallinematrix (i.e., residual fraction) due to the application ofelectric current, which also induces a pH change (Kimet al., 2002; Jakobsen et al., 2004). The removal of tracemetals from the residual and organic/sulfides fractionsis, however, not complete. Reddy and Chinthamreddy(1999) showed that the introduction of sulfides intokaolin caused a significant decrease in migration ofNi(II) due to NiS precipitation. The presence of sulfidesin anaerobic granular sludge can thus explain why traceelements are only partially removed from the organic/sulfides and residual fractions (Fig. 5).
The electrokinetic treatment at a final pH of thesludge of 7.7 increased the calcium, iron and magnesiumcontent in the exchangeable/carbonates fraction at thecathode side (Fig. 6). This suggests that Fe, Mg and Camoved through the sludge cake in cationic forms (Sueret al., 2003).
As observed for copper, the trace metal content washigher at the cathode side for the non-EDTA-treatedsludge, in contrast to the trace metal content at theanode side for EDTA treated sludges. This difference islikely linked to the formation of trace metal species withdifferent charge in the presence of EDTA, as shownin the species distribution model for copper (Fig. 7).Further research using speciation techniques at molec-ular level, e.g. X-ray diffraction and energy dispersiveX-ray spectroscopy spectra with scanning/transmissionelectron photographs or X-ray absorption spectroscopy(Huang et al., 2003) can help to develop a furtherunderstanding of changes in anaerobic granular sludgeschemical composition and mineralogy before and afterelectrokinetic treatment. A compilation of these resultswill provide a fundamental approach for geochemicalmodelling and assist in the development of effectiveelectrokinetic remediation systems.
527J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528
5. Conclusions
� Application of low-level direct current (0.15mA cm�2
for 14 days) induces mobility of Cu species ina methanogenic granular sludge cake.
� Under the applied electric field and in the absence ofEDTA, Cu (and the trace metals Zn, Co and Ni)migrate towards the cathode, except when anaerobicgranular sludge is pre-incubated with Cu. In thepresence of EDTA, Cu (as well as Zn, Co and Ni)migrate towards the anode during electrokinetictreatment.
� EDTA addition does not affect the direction ofmigration of Ca, Fe and Mg, which move to thecathode.
� Lower pH conditions and EDTA addition increasedthe Cu quantity extracted in the exchangeable/carbonates fraction of the freshly Cu amended andpre-incubated sludges. However, in the same time,this led to an increase of the Cu content in the lessmobile fractions (i.e. residual and organic/sulfidesfractions).
Acknowledgements
This research was supported through the EuropeanCommunity Marie Curie Training site ‘‘Heavy metalsand sulfur’’ (HPTM-CT-2000-00118) and the individualMarie Curie Fellowship HPMF-CT-2002-01899 of the‘‘Improving Human Research Potential and the Socio-economic Knowledge Base’’ programme. In addition,the Academy of Finland is thanked for the financialsupport (decision number 200759).
References
Acar, Y.B., Alshawabkeh, A.N., 1993. Principles of electrokinetic