Copper and trace element fractionation in electrokinetically treated methanogenic anaerobic granular sludge

Environmental Pollution 138 (2005) 517e528 www.elsevier.com/locate/envpol

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 cmÿ2 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 cmÿ2) on the metal fractionation in anaerobic granular sludge artificially contaminated with copper (initial copper concentration 1000 mg kgÿ1 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 (Cu2C:EDTA4ÿ 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 cmÿ2. The low-level current induces, via electromigration and electro-osmosis, the

* 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

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 physicochemical forms, such as soluble, adsorbed, exchangeable, 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 by selectivity and interference, do not allow a complete differentiation between different physicochemical forms of metals in the sludge matrix (Hsiau and Lo, 1998). The latter can, however, be approximated by the fractionation of heavy metals in the solid matrix of sludges, for instance, using the revised BCR sequential extraction scheme, as proposed by Mossop and Davidson (2003). Sequential extraction procedures have been developed predominantly to determine the amounts and partitioning of metals present within soils or sediments samples, sewage sludge and sludge treated soils (see Filgueiras et al., 2002 for review). Fractionation procedures are often criticised because of their complexity and difficulty in interpretation, arising from potential problems such as lack of specificity of extractants and re-adsorption of metals during extraction. Nevertheless, providing that such limitations are recognised, sequential fractionations can provide extremely useful information on metal distribution in sludges, particularly for comparative purposes (McLaren and Clucas, 2001). In the present study, mesophilic anaerobic granular sludge was chosen as a model for anaerobic sludges. The main objective was to evaluate the changes in the copper fractionation after different electrokinetic treatment strategies (i.e., exposure to different pH values, complexation with EDTA, pre-incubation) of artificially copper contaminated sludge granules. The effect of the electrokinetic treatment on the fractionation of macro and trace elements in the sludge cake was also investigated.

2. Materials and methods 2.1. Electrokinetic set-up A ‘closed’ laboratory-scale electrokinetic cell was used 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 between the electrodes of 17 cm. In the electrokinetic cell (Fig. 1), the central compartment from the anode and cathode compartments was separated by, respectively, anionexchange (IA1-204SXZL386) and cation-exchange (IC161CZL386) membranes (Ionics Inc, Watertown, MA, USA). The anode and cathode were immersed into a 0.05 M KNO3 conductive solution (Fig. 1). The electrodes (diameter 3 mm, length 5 cm) were titan bars (supplied by Elektronika-WUR, The Netherlands). In the electrokinetic cell, a power supply (HewlettPackard 613 Altai, Germany) was used to constantly maintain a 0.15 mA cmÿ2 DC current. The voltage fluctuations were monitored with a Fluke 112 multimeter (Fluke, Eindhoven, The Netherlands).

Peristaltic pumps (Marlow 502S) allowed a recirculation of the electrolyte solutions. The average flow rate in the electrolytes was maintained at 5 ml minÿ1. 2.2. Source of biomass Anaerobic granular sludge was obtained from a full-scale UASB reactor (Industriewater Eerbeek B.V., Eerbeek, The Netherlands) treating paper-mill wastewater (Lens et al., 1999). The sludge had a mean density of 1040 kg mÿ3. The initial pH of the sludge was 7.1. The total suspended solid (TSS) and volatile suspended solid (VSS) concentrations of the sludge were 22 (G0.2)% and 73.9 (G0.2)%, respectively. The carbonates concentration was 0.4 (G0.2)% of TSS and total sulfur was 41.8 (G1.0) mg gÿ1 TSS (van Hullebusch et al., 2005). The background copper concentration present in this sludge amounted to 150 mg kgÿ1 TSS (Osuna et al., 2004). 2.3. Experimental design The amount of sludge placed in the electrokinetic cell was 1000 g (wet sludge). It was artificially contaminated with 3.78 g of Cu(NO3)2, which gave 1000 mg kgÿ1 (wet sludge) of copper. The contamination procedure was as follows: 1000 g of anaerobic granular sludge amended with the specific Cu(NO3)2 and Na2H2EDTA concentration was placed into a closed plastic container for 48 h (Reddy et al., 1997). The content of the plastic container was frequently and thoroughly mixed. Upon termination of the contamination procedure, the sludge, still suspended in the copper (EDTA) containing supernatant, was mounted in the electrokinetic set-up. In the EDTA amended experiments, 3.78 g of Na2H2EDTA was added simultaneously with Cu(NO3)2 (molar ratio of Cu2C: Na2H2EDTA at 1.2:1). When testing the effect of sludge pre-incubation, the sludge was incubated in an anaerobic chamber with 3.78 g of Cu(NO3)2 for 30 days, eventually supplemented with 3.78 g of Na2H2EDTA. The sludge was put in the electrokinetic cell for 14 days in all experiments. This period was chosen because after 10 days, the voltage had increased significantly (up to 90 V), remained constant for at least 2 days and then decreased down to 10 V. The low voltage indicates that the entire acidic front had passed through the sludge cake. After the electrokinetic treatment, the sludge cake was split into three portions, i.e. a portion close to the anode, a portion close to the cathode and a mid portion between them. Each portion was homogeneously mixed before analysis. The initial pH of the sludge cake in the electrokinetic cell was 7.1. When the pH fluctuations were uncontrolled in the electrolytes, it reached 12.5 in the catholyte. This gave a final pH of 7.7 in the sludge cake upon termination of the electrokinetic treatment.

519

J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528

Power supply

Multimeter

DC Pump

Ion exchange membrane

Ion exchange membrane 20cm

Cu2+

K+ MeEDTA2-

5cm Anode

H+

Conductive solution

NO3-

5cm

Cathode Metal ions

Anolyte

Pump

Anaerobic sludge

OH-

Conductive solution

Catholyte

26cm 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 M HNO3 addition, the pH in the sludge cake was 4.2 upon termination of the electrokinetic treatment.

202.548, 422.673 and 279.553 nm for Co, Cu, Fe, Ni, Zn, Ca and Mg, respectively, for ICP-OES measurements.

2.4. Analytical techniques

2.4.2. Revised sequential extraction (BCR) scheme The revised BCR scheme is designed based on an acetic acid extraction of approximately 1 g TSS of granular sludge (step 1), hydroxylamine hydrochloride extraction (step 2) and hydrogen peroxide oxidation and ammonium acetate extraction (step 3) as described by Mossop and Davidson (2003). To extract the residual phase (step 4), a mixture of 2.5 ml HNO3 (65%) and 7.5 ml HCl (37%) (aqua regia digestion) was added to the residue and the filter from fraction 3. After microwave destruction, the samples of step 4 were paper filtered and diluted to 100 ml with ultra pure water. The chemicals used for the extraction are presented in Table 1. These extractions are associated with the exchangeable/carbonate (bound to carbonate, step 1), oxides phase (bound to iron and manganese oxides, step 2) and the organic/sulfides phase (bound to organic matter/ sulfides, step 3).

2.4.1. Pseudo-total metal analysis The pseudo-total metal content (expressed as mg metal kgÿ1 wet sludge) of sludge was determined by digestion with the aqua regia procedure (Florian et al., 1998). For metal determination, 0.5 g TSS of anaerobic granular sludge were treated with 10 ml of aqua regia in TeflonÒ perfluoroalkoxy resin (PFA) digestion vessels, in a temperature controlled microwave oven Milestone ETHOS E (Milestone Inc; Monroe, CT, USA). The sample volume was completed to 100 ml with ultrapure water. After digestion, the concentrations of total metals were analysed using an atomic absorption spectroscopy (AAS) flame (Perkin-Elmer 300) or inductively coupled plasma optical emission spectrometry (ICP-OES; Varian MPX CCD, Vista, Australia). The following wavelengths were used: 228.802, 213.598, 259.940, 216.555,

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) 20

1. ExchangeableCcarbonates

40 ml CH3COOH (0.11 M, pHZ7)

16 h

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 b

Shaking was applied at 30 rpm for 16 h. Extraction of the residual fraction in the microwave was equal to the pseudo-total extraction method.

520

J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528

Table 2 Results obtained (meanGstandard deviation, nZ3) for sequential extraction analysis and aqua regia extractable (pseudo-total) metal content of BCR 701 sediment Step 1

Step 3

Pseudo-total metal

Found value, mg kgÿ1

Certified value, mg kgÿ1

Found value, mg kgÿ1

Certified value, mg kgÿ1

Found value, mg kgÿ1

Indicative value, mg kgÿ1

Found value, mg kgÿ1

Indicative value, mg kgÿ1

69.5G2.0 15.5G2.0 193.6G2.6

49.3G1.7 15.4G0.9 205G6

120.0G1.5 24.0G0.3 104.0G2.0

124G3 26.6G1.3 114G5

53.4G1.3 15.9G0.2 47.9G1.0

55.2G4.0 15.3G0.9 45.7G4.0

37.3G2.0 42.1G3.0 105.1G2.7

38.5G11.2 41.4G4.0 95G13

271.8G5.1 100.0G1.0 449G2.0

275G13 103G4 454G19

2.4.3. Evaluation of analytical performance The analytical performance of the laboratory procedures was evaluated by analysis of Certified Reference Material BCR-701 and CRM 146R. A two-sided t-test was used to check for significant differences from the reference content. Table 2 shows the data of three replicate analyses obtained for aqua regia extraction and Table 3 shows the revised BCR sequential extraction procedure, expressed as mg kgÿ1 of dry mass. Uncertainty is expressed as standard deviations; the values obtained are not significantly (PO0.05) different from the certified values. The data for pseudo-totals and fractionation of copper and other elements were obtained from triplicates (nZ3) with meanGstandard deviation. The TSS and VSS concentrations were determined according to the APHA standard methods (APHA, 1995).

2.4.4. Visual Minteq The geochemical equilibrium model Visual Minteq (freeware program at http://www.lwr.kth.se/english/ OurSoftware/Vminteq/index.htm) was used to simulate the chemical speciation of Cu in solution in two different experiments (without and with EDTA addition) in the absence of anaerobic granular sludges. For calculation, the solid forms of copper were allowed to precipitate. Visual Minteq is capable of calculating equilibrium aqueous speciation, precipitation and dissolution of minerals, complexation, adsorption, solid phase saturation 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

Co Cu Mn Ni Zn

Residual

Certified value, mg kgÿ1

Found value (mg kgÿ1 DW)

Certified value (mg kgÿ1 DW)

6.35G0.09 744G7 278G4 53.3G0.9 2887G29

6.50G0.31 831G16 298G9 65.0G3.0 3043G58

3. Results 3.1. Initial fractionation of copper in anaerobic granular sludge Prior to electrokinetic treatment (at pH 7.1), copper was mainly associated with the exchangeable/carbonate (310 mg kgÿ1) fraction of fresh copper supplemented granular sludge (Fig. 3a). The remaining copper was spread equally over the other fractions: 190 mg kgÿ1 of Cu in the residual and organic/sulfides and 200 mg kgÿ1 in the oxides fractions (Fig. 3a). Prolonged exposure of sludge to copper (30 days) prior to the electrokinetic treatment modified the copper fractionation. The most abundant fractions in the pre-incubated sludge were, respectively, the oxides (330 mg kgÿ1) and the residual (260 mg kgÿ1) fractions (Fig. 3b). Addition of Cu and EDTA simultaneously resulted in a decrease in all fractions (Fig. 4), except in the exchangeable/carbonate fraction (310 mg kgÿ1), compared to the non-EDTA amended sludge (Fig. 3). 3.2. Effect of electric current on copper fractionation Fig. 2 shows the development of the electric potential gradient across the sludge matrix with time. The highest electrical potential gradient applied at a 5 ml minÿ1 flow rate of the electrolyte solutions was 3.5 V cmÿ1. During the experiments, the gradient increased from 0.04 to 3.5 V cmÿ1 and then decreased to 1 V cmÿ1 (Fig. 2). Electrical potential gradient (V/cm)

Cu Ni Zn

Step 2

Found value, mg kgÿ1

3.5 3 2.5 2 1.5 1 0.5 0 1

3

5

7

9

11

13

Time (days)

Fig. 2. Evolution of the electrical potential gradient with time (V cmÿ1).

521

J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528

After the electrokinetic treatment (0.15 mA cmÿ2 DC current for 14 days), the fractionation of copper in the granular sludge matrix had changed significantly (Figs. 3 and 4). A significant increase of copper in the residual (from 190 to 300 mg kgÿ1) and the oxides (from 200 to 320 mg kgÿ1) fractions was observed for the freshly amended sludge, which was directed towards the cathode (Fig. 3a). Copper associated with the exchangeable/carbonate fraction decreased from 310 to 80 mg kgÿ1, which moved towards the anode (Fig. 3a). One of the major trends was that the copper concentration increased in the residual fraction of most of the experimental conditions at both the cathode and anode side (Figs. 3 and 4).

3.3. Effect of pH during the electrokinetic treatment on the copper fractionation The major fractions with which copper was associated after electrokinetic treatment with pH 12.5 in the catholyte solution (final pH 7.7 in the sludge bed) were the residual (up to 310 mg kgÿ1) and the oxides (320 mg kgÿ1) fractions, which were redistributed towards the cathode (Fig. 3a). When the pH was maintained acidic (pH 2.5) in the electrolytes (final pH 4.2 in the sludge bed), the most abundant copper fractions were the exchangeable/carbonates (330 mg kgÿ1) and the residual (300 mg kgÿ1) fractions, which moved towards the cathode (Fig. 3c). The copper associated with the organic/sulfides (up to 210 mg kgÿ1) and oxides (up to 236 mg kgÿ1) fractions

A

initial

anode

mid section

3.4. Effect of sludge pre-incubation on the copper fractionation When the sludge was pre-incubated with copper for 30 days, the dominant fractions were the oxides (up to 320 mg kgÿ1), residual (290 mg kgÿ1) and organic/ sulfides (up to 280 mg kgÿ1) upon the termination of the 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 kgÿ1) and residual (up to 320 mg kgÿ1) when the pH in the catholyte solution was set to pH 2.5 (final pH 4.2 in the sludge cake). Interestingly, copper in the exchangeable/carbonates and the residual fractions moved towards the cathode, in contrast to copper bound to the organic/sulfides fraction, which migrated towards the positively charged anode (Fig. 3d). 3.5. Effect of EDTA on the copper fractionation after electrokinetic treatment Table 4 shows that in the presence of EDTA, copper accumulated at both sides compared to the initial concentration. This is an opposite trend compared to the non-EDTA treated sludge. When the pH in the catholyte solution was 12.5 (final pH 7.7 in the sludge water), there was a significant decrease in all fractions 400

cathode

300

Cu mg.kg-1

Cu mg.kg-1

400

remained unchanged, respectively, at the anode and cathode side (Fig. 3c).

200

200

0

0 residual

organic/sulfides

oxides

exch/carb

residual 500

C

organic/sulfides

oxides

exch/carb

organic/sulfides

oxides

exch/carb

D

400

300

Cu mg.kg-1

Cu mg.kg-1

300

100

100

400

B

200 100

300 200 100 0

0 residual

organic/sulfides

oxides

exch/carb

residual

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).

522 900 800 700 600 500 400 300 200 100 0

A

initial

mid section

800

cathode

700 500 400 300 200 100 0 organic/sulfides

oxides

exch/carb

residual 600

C

700

organic/sulfides

oxides

exch/carb

organic/sulfides

oxides

exch/carb

D

500

Cu mg.kg-1

600

Cu mg.kg-1

B

600

residual 800

anode

Cu mg.kg-1

Cu mg.kg-1

J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528

500 400 300

400 300 200

200 100

100 0

0 residual

organic/sulfides

oxides

exch/carb

residual

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 fraction, which remained the same as in the copper nitrate amended sludge experiments (310 mg kgÿ1) (Fig. 3a). The most predominant fraction after electrokinetic treatment at pH 7.7 was the organic/sulfides fraction (from 170 to 750 mg kgÿ1), which had moved towards the positively charged anode (Fig. 4a). When the pH in 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 kgÿ1) and removal efficiencies (%)

Concentrations at the cathode (mg kgÿ1) and removal efficiencies (%)

Cu(NO3)2 treatment Cu 1070 Zn 125 Co 38 Ni 38 Ca 2100 Fe 25000 Mg 680

620 60 34 36 1100 12000 150

(40) (52) (11) (5) (48) (52) (78)

785 78 40 41 2000 16000 580

(27) (37) (accumulation) (accumulation) (5) (36) (15)

Cu EDTA treatment Cu 790 Zn 125 Co 38 Ni 38 Ca 2100 Fe 25000 Mg 680

1320 70 42 40 350 11000 40

(accumulation) (44) (accumulation) (accumulation) (83) (56) (94)

980 50 33 36 1900 19000 590

(accumulation) (60) (16) (5) (10) (24) (16)

treated EDTA amended fresh sludge were the residual (650 mg kgÿ1) and the organic/sulfides (550 mg kgÿ1) fractions, which were directed towards the anode (Fig. 4c). After electrokinetic treatment of the pre-incubated sludges at a pH 12.5 in the catholyte (final pH 7.7 in the sludge bed) in the presence of EDTA, copper was mainly present in the residual (up to 680 mg kgÿ1) fraction, followed by the oxides (640 mg kgÿ1) fraction (Fig. 4b) at the anode side. At pH 2.5 in the catholyte (final pH 4.2 in the sludge bed), the most significant copper fraction (490 mg kgÿ1) was the residual fraction at the anode side (Fig. 4d). 3.6. Effect of electrokinetic treatment on trace and macroelements 3.6.1. Trace elements Initially, the main trace metal fraction was the residual (Fig. 5a,c,e): Zn, 78 mg kgÿ1; Co, 15 mg kgÿ1; Ni, 30 mg kgÿ1 (Table 4). After electrokinetic treatment at pH 12.5 in the catholyte, the residual fraction decreased 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 kgÿ1 at the cathode for Zn (Fig. 5a) and from 3 to 10 mg kgÿ1 for Ni, which was directed towards the anode (Fig. 4e). Also, a significant increase in Ni associated with the exchangeable/carbonates fraction (from 3 to 15 mg kgÿ1) occurred at the cathode (Fig. 5e). Co associated with the exchangeable/carbonates fraction increased from 5 to 10 mg kgÿ1 after the electrokinetic treatment at the

523

J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528

A

initial

anode

mid section

80

Zn mg.kg-1

100

cathode

Zn-EDTA mg.kg-1

100

60 40 20 0 oxides

20

residual 20

C

10 5

organic/sulfides

oxides

exch/carb

D

15 10 5 0

0 residual

organic/sulfides

oxides

exch/carb

residual 40

Ni-EDTA mg.kg-1

E

30

Ni mg.kg-1

40

exch/carb

Co-EDTA mg.kg-1

Co mg.kg-1

organic/sulfides

15

40

60

0 residual

20

B

80

20 10 0

organic/sulfides

oxides

exch/carb

organic/sulfides

oxides

exch/carb

F

30 20 10 0

residual

organic/sulfides

oxides

exch/carb

residual

Fig. 5. Fractionation of trace metals in electrokinetically treated anaerobic granular sludge with pH 12 in the catholyte (final pH 7.7 in the sludge bed): (A), (C), (E) Cu(NO3)2 contaminated sludge and (B), (D), (F) CuEDTA contaminated sludge.

anode side (Fig. 5c). However, the other fractions (i.e. oxides and organic/sulfides fractions) of Co remained without any significant changes at the cathode or anode sides (Fig. 5c). EDTA amendment induced a redistribution of Co (from 5 to 14 mg kgÿ1) and Ni (from 3 to 11 mg kgÿ1) towards the anode in the exchangeable/carbonates fraction upon electrokinetic treatment. In contrast, there was a significant decrease in Zn (from 78 to 15 mg kgÿ1) and Ni (from 30 to 5 mg kgÿ1) associated with the residual fraction (Figs. 5b and 2f). 3.6.2. Macroelements For the macroelements, the major initial fractions were the oxides (Fig. 6a,c,e): Ca, 1200 mg kgÿ1; Mg, 300 mg kgÿ1; and the residual for Fe (10000 mg kgÿ1) (Table 4). After electrokinetic treatment, the residual fraction had increased for Ca from 180 to 400 mg kgÿ1 (Fig. 6a) but decreased for Fe from 10,000 to 6000 mg kgÿ1 (Fig. 6c). There was a significant decrease observed for all Fe associated fractions, which moved towards the cathode, with the exception of the residual fraction (Fig. 6c). The most abundant Ca fraction was the exchangeable/

carbonates fraction (Fig. 6a) attributing to 900 mg kgÿ1, which moved towards the cathode. The Mg fractionation showed a significant increase in the exchangeable/ carbonates (300 mg kgÿ1) and organic/sulfides (130 mg kgÿ1) fractions (Fig. 6e), which had also moved towards the cathode. The most significant increase for Ca, Mg and Fe in electrokinetically treated CuEDTA exposed sludge was in the organic/sulfides fraction: 600 mg kgÿ1, 200 mg kgÿ1 and 8500 mg kgÿ1, respectively (Fig. 6b,d,f). Ca associated with the exchangeable/carbonates fraction increased from 600 to 700 mg kgÿ1 (Fig. 6b), which was directed towards the cathode. 4. Discussion This study showed that electrokinetic treatment at 0.15 mA cmÿ2 for 14 days not only induces copper and trace metals mobility but also alters their fractionation in anaerobic granular sludge. The latter is strongly influenced by the pH, contaminant aging and the presence of EDTA. The discussion section describes successively the copper fractionation prior to electrokinetic treatment

524

J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528 1400

A

initial

anode

mid section

Ca-EDTA mg.kg-1

Ca mg.kg-1

1400

cathode

1200 1000 800 600 400 200 residual

organic/sulfides

oxides

400 200

12000

Fe-EDTA mg.kg-1

Fe mg.kg-1

600

residual

C

8000 6000 4000 2000

organic/sulfides

oxides

exch/carb

organic/sulfides

oxides

exch/carb

residual organic/sulfides

oxides

exch/carb

D

10000 8000 6000 4000 2000 0

0 residual organic/sulfides

oxides

exch/carb

residual 400

E Mg-EDTA mg.kg-1

Mg mg.kg-1

800

exch/carb

10000

400

1000

0

0

12000

B

1200

300 200 100 0

F

300 200 100 0

residual organic/sulfides

oxides

exch/carb

Fig. 6. Fractionation of macroelements in electrokinetically treated anaerobic granular sludge with pH 12 in the catholyte (final pH 7.7 in the sludge bed): (A), (C), (E) Cu(NO3)2 contaminated sludge and (B), (D), (F) CuEDTA contaminated sludge.

and the effect of different working conditions during electro-remediation on the partitioning of copper and some macro and trace elements in the sludge. 4.1. Copper fractionation before electrokinetic treatment Exposure of the sludge granules to an easily soluble copper salt (Cu(NO3)2) resulted in a much higher total copper concentration of the sludge compared to all other heavy metals. In all cases, copper was more or less evenly distributed between the four operationally defined fractions of the BCR protocol. This distribution of copper was very similar to that observed for Cu in the non-Cu-spiked sludges used to prepare the amended sludge (Osuna et al., 2004). However, the proportion of Cu present in the two most easily extracted fractions (exchangeable/carbonates and oxides fractions) was much larger due to increasing Cu concentrations in the sludge (Figs. 3 and 4). Both at low (Osuna et al., 2004) and high (Figs. 3 and 4) concentrations, copper showed a very strong affinity for the organic matter/sulfides and residual fractions of the BCR scheme. It is, indeed, well

known that Cu strongly interacts with organic matter and sulfides (Pattrick et al., 1997; Lu and Allen, 2002; Vulkan et al., 2002). Copper may undergo biosorption by cell membrane surfaces with proteins and acid groups that serve as binding site (Hayes and Theis, 1978). The binding and sequestration of copper by extracellular polymeric substances (EPS) has also been demonstrated (White and Gadd, 2000). Besides, chalcocite (Cu2S) can be formed in the presence of sulfide (Morse and Luther, 1999) and even different Cu sulfide minerals can coexist in the anaerobic sludge matrixes (Pattrick et al., 1997). These metals bound to the sulfides are mainly leached in the organic/sulfides fraction in natural sediment, soil or sludge (Lacal et al., 2003). Copper may be also adsorbed in large quantity at the surface of pyrite minerals (Mu¨ller et al., 2002), however little evidence has been found in the literature for Cu accumulation in the crystalline lattice of the iron sulfides. When the fresh Cu amended sludge was supplemented with EDTA prior to electrokinetic treatment, the initial fractionation of copper changed (Fig. 4a,c) and a lower initial total copper quantity accumulated in the sludge (Table 4). This is in agreement with Osuna

525

J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528

Cu2C, Cu2OH3C and CuOHC (Fig. 7b). These copper species can migrate towards the negatively charged cathode, as observed for the oxides and residual fractions of the freshly amended sludges (Fig. 3a). Besides electromigration, also electro-osmotic water flow can contribute to an increased copper content at the cathode side. Such an electro-osmotic water flow is produced during the electrokinetic treatment, which moves through the sludge cake from the anode to the cathode (Pamukcu et al., 1997).

et al. (2004), who studied the effect of EDTA on the accumulation of Co in anaerobic granular sludge. EDTA addition increased the amount of dissolved copper obtained in the copper speciation simulations (Fig. 7): at pH 7, the addition of EDTA increased the percentage of dissolved copper from 1% to 83.5% of the total copper initially present in the systems. 4.2. Effect of pH on copper fractionation

Dissolved Cu species (%)

The pH influences the adsorption and desorption, precipitation, dissolution and speciation reactions of copper. At low pH, copper tends to desorb from the sludge matrix and dissolves as positively charged ions (Hsiau and Lo, 1998). All heavy metals have a specific pH underneath which their solubility is drastically increased. For copper, this pH is 5.5 (Martinez and Motto, 2000). The decrease in the pH during the electrokinetic treatment consequently promoted the formation of soluble mobile copper compounds such as free copper (Cu2C) and CuNOC 3 (Fig. 7b), which caused copper accumulation at the negatively charged cathode side in the exchangeable/carbonates and residual fractions of either the freshly amended (Fig. 3c) and pre-incubated (Fig. 3d) sludge. Also at neutral and slightly alkaline pH (6.5e7.5), the predominant dissolved species are positively charged compounds, e.g. 100 90 80 70 60 50 40 30 20 10 0

4.3. Effect of pre-incubation on copper fractionation After a 30 days incubation period, an increase of copper in the residual fraction has been noticed, very likely due to the contamination aging effect (McLaren and Clucas, 2001). Pattrick et al. (1997) reported that the mobility of copper decreases with the duration (a few hours) of its contact with anaerobic sludge, due to the formation of copper sulfide crystals. The electrokinetic treatment of the pre-incubated copper amended sludges mainly influenced the metal accumulated in the organic/sulfides fraction (Fig. 3bed). Kim et al. (2002) found that the organic/sulfides fraction is relatively mobile in an electric field, which is in agreement with Fig. 3b and to a lesser extent with Fig. 3d. It is well documented that copper has also a strong affinity to

A

2

4

6

8

10

12

14

Dissolved Cu species (%)

pH

100 90 80 70 60 50 40 30 20 10 0

Cu2+

CuNO3+

CuEDTA2-

CuHEDTA-

CuH2EDTA (aq)

Cu(CO3)22-

CuCO3 (aq)

Cu3(OH)42+

Cu2(OH)22+

CuOHEDTA3-

B

2

4

6

8

10

12

Cu2+

CuNO3+

Cu2OH3+

CuOH+

CuCO3 (aq)

CuHCO3+

Cu(OH)2 (aq)

Cu(CO3)22-

14

Fig. 7. Dissolved copper species distribution as a function of pH with (A) and without EDTA (B) in absence of anaerobic granular sludges.

526

J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528

form complexes with the organic matter present (Bolan et al., 2003; Lu and Allen, 2002). When copper is contacted with an organic fraction in anaerobic conditions, it usually forms negative compounds, e.g. Cu-Org2ÿ or Cu-Orgÿ (Lu and Allen, 2002; Vulkan et al., 2002), which would explain the accumulation of copper at the anode side in the organic/sulfides fraction (Fig. 3b). For sewage sludge exposed to sulfidic conditions, the organic fraction has been shown to originate from the hydrolysis of EPS (Watson et al., 2004). Hydrolysis of the organic sludge matrix very likely occurred as well during the pre-incubation step of the copper amended anaerobic sludge, thus yielding organic compounds to which copper binds. 4.4. Effect of EDTA on copper fractionation The addition of the chelating agent EDTA to the sludge cake resulted in a change of the migration direction of copper in the applied electric field (Fig. 4). Complexing agents like EDTA are widely used for the enhanced removal of heavy metals from different contaminated media without (Nirel et al., 1998; Sillanpa¨a¨ et al., 2001; Nowack, 2002) and with (Wong et al., 1997; Velizarova et al., 2002) electrokinetic treatment. EDTA has proved to be an efficient chelating agent since it forms stable complexes with most metals, especially with transition metals over a broad pH range (Lo and Yang, 1999). It can also be utilized for both desorption of sorbed ions and dissolution of precipitated metal compounds (Papassiopi et al., 1999). When the sludges were supplemented with EDTA, copper extracted in the residual and oxides fractions exclusively increased at the anode side at both pH conditions (Fig. 4b,d) for the pre-incubated sludge after electrokinetic treatment. In contrast, the fresh sludge showed that the extracted copper increased in the organic/sulfides and residual fractions at the anode at final pH 4.2 (Fig. 3c). This outlines the presence of negatively charged copper species in the presence of EDTA. Indeed, according to the species distribution model (Fig. 7a), the main dissolved species is CuEDTA2ÿ at neutral and slightly alkaline pH (pH 6e8). At acidic pH (pH 2e4), the most abundant dissolved species are CuEDTA2ÿ and CuHEDTAÿ (Fig. 7a). The prevailing copper complexes are negatively charged, thus they migrate towards the positively charged anode (Fig. 4), which is also reported for copper mobility in clayey soils (Popov et al., 1999). It should be noted that copper also increased at the cathode side of the freshly copper amended sludge for a final pH of 7.7 (Fig. 4a). This outlines that also positively charged species were present at that pH. Indeed, according to the geochemical model (Fig. 7a), the Cu3(OH)2C content 4 can be expected to be significant. However, no analytical evidence can confirm this statement.

Fig. 4 shows that EDTA is efficient to decrease the copper content in the exchangeable/carbonates and oxides fractions of the fresh sludge. However, no decrease was noticed in the organic/sulfide and residual fractions. This is probably due to the presence of sulfide precipitates which are efficient sorbants. Consequently, these can prevent copper removal from the anaerobic sludge as reported by Reddy and Chinthamreddy (1999), when they introduced sulfide in kaolin.

4.5. Effect of electrokinetic treatment on trace metals and major elements fractionation An increase of the cobalt and nickel content, present in much lower concentrations than Cu in the exchangeable/carbonates fraction was noticed when the final pH in the sludge bed was 7.7. The decrease in the residual fraction compared to the initial sample prior to electrokinetic treatment (Fig. 5) might be explained by the desorption of trace metals from the sludge crystalline matrix (i.e., residual fraction) due to the application of electric current, which also induces a pH change (Kim et al., 2002; Jakobsen et al., 2004). The removal of trace metals from the residual and organic/sulfides fractions is, however, not complete. Reddy and Chinthamreddy (1999) showed that the introduction of sulfides into kaolin caused a significant decrease in migration of Ni(II) due to NiS precipitation. The presence of sulfides in anaerobic granular sludge can thus explain why trace elements are only partially removed from the organic/ sulfides and residual fractions (Fig. 5). The electrokinetic treatment at a final pH of the sludge of 7.7 increased the calcium, iron and magnesium content in the exchangeable/carbonates fraction at the cathode side (Fig. 6). This suggests that Fe, Mg and Ca moved through the sludge cake in cationic forms (Sue`r et al., 2003). As observed for copper, the trace metal content was higher at the cathode side for the non-EDTA-treated sludge, in contrast to the trace metal content at the anode side for EDTA treated sludges. This difference is likely linked to the formation of trace metal species with different charge in the presence of EDTA, as shown in the species distribution model for copper (Fig. 7). Further research using speciation techniques at molecular level, e.g. X-ray diffraction and energy dispersive X-ray spectroscopy spectra with scanning/transmission electron photographs or X-ray absorption spectroscopy (Huang et al., 2003) can help to develop a further understanding of changes in anaerobic granular sludges chemical composition and mineralogy before and after electrokinetic treatment. A compilation of these results will provide a fundamental approach for geochemical modelling and assist in the development of effective electrokinetic remediation systems.

J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528

5. Conclusions  Application of low-level direct current (0.15 mA cmÿ2 for 14 days) induces mobility of Cu species in a methanogenic granular sludge cake.  Under the applied electric field and in the absence of EDTA, Cu (and the trace metals Zn, Co and Ni) migrate towards the cathode, except when anaerobic granular sludge is pre-incubated with Cu. In the presence of EDTA, Cu (as well as Zn, Co and Ni) migrate towards the anode during electrokinetic treatment.  EDTA addition does not affect the direction of migration of Ca, Fe and Mg, which move to the cathode.  Lower pH conditions and EDTA addition increased the Cu quantity extracted in the exchangeable/ carbonates fraction of the freshly Cu amended and pre-incubated sludges. However, in the same time, this led to an increase of the Cu content in the less mobile fractions (i.e. residual and organic/sulfides fractions). Acknowledgements This research was supported through the European Community Marie Curie Training site ‘‘Heavy metals and sulfur’’ (HPTM-CT-2000-00118) and the individual Marie Curie Fellowship HPMF-CT-2002-01899 of the ‘‘Improving Human Research Potential and the Socioeconomic Knowledge Base’’ programme. In addition, the Academy of Finland is thanked for the financial support (decision number 200759).

References Acar, Y.B., Alshawabkeh, A.N., 1993. Principles of electrokinetic remediation. Environmental Science and Technology 27, 2638e2647. Angelidis, M., Gibbs, R.J., 1989. Chemistry of metals in anaerobically treated sludges. Water Research 23, 29e33. APHA, 1995. Standard Methods for the Examination of Water and Wastewater, 18th ed. American Public Health Association, Washington, DC. Bolan, N.S., Khan, M.A., Donaldson, J., Adriano, D.C., Mathew, C., 2003. Distribution and bioavailability of copper in farm effluent. The Science of the Total Environment 309, 225e236. Filgueiras, A.V., Lavilla, I., Bendicho, C., 2002. Chemical sequential extraction for metal portioning in environmental solid samples. Journal of Environmental Monitoring 4, 823e857. Florian, D., Barnes, R.M., Knapp, G., 1998. Comparison of microwave-assisted acid leaching techniques for the determination of heavy metals in sediments, soils, and sludges. Fresenius Journal of Analytical Chemistry 362, 558e565. Gustaffsson, J.P. !http://www.lwr.kth.se/english/OurSoftware/Vminteq/index.htmO, (verified 6 June 2005). Hamed, J., Acar, Y.B., Gale, R.J., 1991. Pb (II) removal from kaolinite using electrokinetics. Journal of Geotechnical Engineering (ASCE) 112, 241e271.

527

Hayes, T.D., Theis, T.L., 1978. The distribution of heavy metals in anaerobic digestion. Journal of Water Pollution Control Federation 50, 61e72. Hsiau, P.-Ch., Lo, Sh.-L, 1998. Fractionation and leachability of Cu in lime-treated sewage sludge. Water Research 32, 1103e1108. Huang, Y.J., Tsai, C.H., Liaw, B.J., 2003. In situ XANES study of electrokinetic remediation of cadmium-contaminated soils. Bulletin of Environmental Contamination and Toxicology 71, 682e688. Jakobsen, M.R., Fritt-Rasmussen, J., Nielsen, S., Ottosen, L.M., 2004. Electrodialytic removal of cadmium from wastewater sludge. Journal of Hazardous Materials 106, 127e132. Kim, S.-O., Moon, S.-H., Kim, K.-W., Yun, S.-T., 2002. Pilot scale study on the ex situ electrokinetic removal of heavy metals from municipal wastewater sludges. Water Research 36, 4765e4774. Lacal, J., da Silva, M.P., Garcı´ a, R., Sevilla, M.T., Procopio, J.R., Herna´ndez, L., 2003. Study of fractionation and potential mobility of metal in sludge from pyrite mining and affected river sediments: changes in mobility over time and use of artificial ageing as a tool in environmental impact assessment. Environmental Pollution 124, 291e305. Lens, P., Vergeldt, F., Lettinga, G., van As, H., 1999. 1H-NMR study of the diffusional properties of methanogenic aggregates. Water Science and Technology 39, 187e194. Lo, I.M.C., Yang, X., 1999. EDTA extraction of heavy metals from different soil fractions and synthetic soils. Water, Air, and Soil Pollution 109, 219e236. Lu, Y., Allen, H.E., 2002. Characterization of copper complexation with natural dissolved organic matter (DOM)dlink to acidic moieties of DOM and competition by Ca and Mg. Water Research 36, 5083e5101. Maini, G., Sharman, A.K., Sunderland, G., Knowles, C.J., Jackman, S.A., 2000. An integrated method incorporating sulfuroxidizing bacteria and electrokinetics to enhance removal of copper from contaminated soil. Environmental Science and Technology 34, 1081e1087. Martinez, C.E., Motto, H.L., 2000. Solubility of lead, zinc and copper added to mineral soils. Environmental Pollution 107, 153e158. Mattson, E.D., Lindgren, E.R., 1995. Electrokinetic Extraction of Chromate from Unsaturated Soils. American Chemical Society, pp. 11e20. McLaren, R.G., Clucas, L.M., 2001. Fractionation of copper, nickel and zinc in metal-spike sewage sludge. Journal of Environmental Quality 30, 1968e1975. Morse, J.W., Luther, G.W., 1999. Chemical influences on trace metal sulfide interactions in anoxic sediments. Geochimica and Cosmochimica Acta 63, 3373e3378. Mossop, K.F., Davidson, C.M., 2003. Comparison of original and modified BCR sequential extraction procedures for the fractionation of copper, iron, lead, manganese and zinc in soils and sediments. Analytica Chimica Acta 478, 111e118. Mu¨ller, B., Axelsson, M.D., O¨hlander, B., 2002. Adsorption of trace elements on pyrite surfaces in sulfidic mine tailings from Kristineberg (Sweden) a few years after remediation. The Science of the Total Environment 298, 1e16. Nirel, P.M., Pardo, P.-E., Landry, J.-C., Revalcier, R., 1998. Method for EDTA speciation determination: application to sewage treatment plant effluents. Water Research 32, 3615e3620. Nowack, B., 2002. Environmental chemistry of aminopolycarboxylate chelating agents. Environmental Science and Technology 36, 4009e4016. Osuna, M.B., van Hullebusch, E., Zandvoort, M.H., Iza, J., Lens, P.N.L., 2004. Effect of cobalt sorption on metal speciation in anaerobic granular sludge. Journal of Environmental Quality 33, 1256e1270.

528

J. Virkutyte et al. / Environmental Pollution 138 (2005) 517e528

Ottosen, L.M., Hansen, H.K., 1992. Electrokinetic cleaning of heavy metal polluted soil. Internal report, Fysisk-Kemisk Institut and Institut for Geologi og Geoteknik. Technical University of Denmark, Denmark. Papassiopi, N., Tambouris, S., Kontropoulos, A., 1999. Removal of heavy metals from calcareous contaminated soils by EDTA leaching. Water, Air and Soil Pollution 109, 1e15. Pamukcu, S., Weeks, A., Wittle, J.K., 1997. Electrochemical extraction and stabilization of selected inorganic species in porous media. Journal of Hazardous Materials 55, 305e318. Pattrick, R.A.D., Mosselmans, J.F.W., Charnock, J.M., England, K.E.R., Helz, G.R., Garner, C.D., Vaughan, D.J., 1997. The structure of amorphous copper sulfide precipitates: an X-ray absorption study. Geochimica and Cosmochimica Acta 61, 2023e2036. Popov, K., Yachmenev, V., Kolosov, A., Shabanova, N., 1999. Effect of soil electro-osmotic flow enhancement by chelating reagents. Colloids and Surfaces A 160, 135e140. Reddy, K.R., Chinthamreddy, S., 1999. Electrokinetic remediation of metal-contaminated soils under reducing environments. Waste Management 19, 269e282. Reddy, K.R., Parupudi, U.S., Devulapalli, S.N., Xu, Y., 1997. Effect of soil composition on the removal of chromium by electrokinetics. Journal of Hazardous Materials 55, 135e158. Sillanpa¨a¨, M., Orama, M., Ramo, J., Oikari, A., 2001. The importance of ligand speciation in environmental research: a case study. The Science of the Total Environment 267, 23e31. Shrivastava, S.K., Banerjee, D.K., 1998. Operationally determined speciation of copper and zinc in sewage sludge. Chemical speciation and bioavailability 10, 137e143. Sue`r, P., Gitye, K., Allard, B., 2003. Speciation and transport of heavy metals and microelements during electroremediation. Environmental Science and Technology 37, 177e181.

Van Cauwenberghe, L., 1997. Electrokinetics: Technology Overview Report. Groundwater Remediation Technologies Analysis Centre, pp. 1e17. van Hullebusch, E.D., Utomo, S., Zandvoort, M.H., Lens, P.N.L., 2005. Comparison of three sequential extraction procedures to describe metal fractionation in anaerobic granular sludges. Talanta 65, 549e558. Velizarova, E., Ribeiro, A.B., Ottosen, L.M., 2002. A comparative study on Cu, Cr and As removal from CCA-treated wood waste by dialytic and electrodialytic processes. Journal of Hazardous Materials 94, 147e160. Virkutyte, J., Sillanpa¨a¨, M., Laatostenmaa, P., 2002. Electrokinetic soil remediationdcritical overview. The Science of the Total Environment 289, 97e121. Vulkan, R., Mingelgrin, U., Ben-Asher, J., Frenkel, H., 2002. Copper and zinc speciation in the solution of a soil-sludge mixture. Journal of Environmental Quality 31, 193e203. Watson, S.D., Akhurst, T., Whiteley, C.G., Rose, P.D., Pletschke, B.I., 2004. Primary sludge floc degradation is accelerated under biosulphidogenic conditions: enzymological aspects. Enzyme Microbiological Technology 34, 595e602. White, C., Gadd, G.M., 2000. Copper accumulation by sulfatereducing bacterial biofilms. FEMS Microbiological Letters 183, 313e318. Wong, J.S.H., Hicks, R.E., Probstein, R.F., 1997. EDTA-enhanced electroremediation of metal-contaminated soils. Journal of Hazardous Materials 55, 61e79. Yeung, A., Hsu, C., Menon, R.M., 1997. Physicochemical soile contaminant interactions during electrokinetic extraction. Journal of Hazardous Materials 55, 221e237. Zagury, G.J., Dartiguenave, Y., Setier, J.C., 1999. Ex situ electroreclamation of heavy metals contaminated sludge: pilot scale study. Journal of Environmental Engineering 125, 972e978.