The two-phase leaching of Pb, Zn and Cd contaminated soil using EDTA and electrochemical treatment of the washing solution

Chemosphere 73 (2008) 1484–1491

Contents lists available at ScienceDirect

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

The two-phase leaching of Pb, Zn and Cd contaminated soil using EDTA and electrochemical treatment of the washing solution Neza Finzgar, Domen Lestan * Agronomy Department, Center for Soil and Environmental Science, Biotechnical Faculty, University of Ljubljana, Jamnikarjeva 101, 1000 Ljubljana, Slovenia

a r t i c l e

i n f o

Article history: Received 19 May 2008 Received in revised form 18 July 2008 Accepted 21 July 2008 Available online 31 August 2008 Keywords: Heavy metals Soil remediation Soil washing EDTA Electrochemical advanced oxidation methods

a b s t r a c t The feasibility of a novel two-phase method for remediation of Pb (1374 mg kg1), Zn (1007 mg kg1), and Cd (9.1 mg kg1) contaminated soil was evaluated. In the first phase we used EDTA for leaching heavy metals from the soil. In the second phase we used an electrochemical advanced oxidation process (EAOP) for the treatment and reuse of washing solution for soil rinsing (removal of the soilretained, chelant-mobilized metallic species). In EAOP, a boron-doped diamond anode was used for the generation of hydroxyl radicals and oxidative decomposition of EDTA–metal complexes at a constant current density (15 mA cm2). The released metals were removed from the solution by filtration as insoluble participate and by electro-deposition on the cathode. Four consecutive additions of 5.0 mmol kg1 EDTA (total 20 mmol kg1) removed 44% Pb, 14% Zn and 35% Cd from the soil. The mobility of the Pb, Zn and Cd (Toxicity Characteristic Leaching Procedure) left in the soil after remediation was reduced by 1.6, 3.4 and 1.5 times, respectively. The Pb oral availability (Physiologically Based Extraction Test) in the simulated stomach phase was reduced by 2.4 and in the intestinal phase by 1.7 times. The discharge solution was clear, almost colorless, with pH 7.73 and 0.47 mg L1 Pb, 1.03 mg L1 Zn, bellow the limits of quantification of Cd and 0.023 mM EDTA. The novel method enables soil leaching with small water requirements and no wastewater generation or other emissions into the environment. Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction The contamination of soils with heavy metals has become a major environmental concern. Soil remediation can reduce the contamination and preserve soil as a non-renewable natural resource. Phytoextraction, which not only reduces contamination but can also improve soil quality, has been successfully applied for soils contaminated with Ni, Cd, As and Se (Chaney et al., 1997). Pb, one of the major soil contaminants, is unfortunately not available in soil for plant uptake. Soil washing with chelant ethylenediamine tetraacetate (EDTA) has been proposed as a viable alternative (Huang et al., 1997). Chelants form coordinate chemical bonds with metals (complexes) and facilitate their solubilization from the soil into the washing solution. EDTA is relatively inexpensive (in Europe, it costs about 1.3 € kg1 for the technical-grade chemical from a major European manufacturer). However, toxic wastewaters containing complexed EDTA cannot be treated using conventional methods such as filtration, flocculation and participation (Jiraroj et al., 2006). Another possible problem of soil washing is the retention of toxic EDTA and EDTA–heavy metal complexes through absorption by soil min* Corresponding author. Tel.: +386 01 423 1161; fax: +386 01 423 1088. E-mail address: [email protected] (D. Lestan). 0045-6535/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2008.07.043

eral surfaces and organic matter (Nowack and Sigg, 1995). Vast consumption of clean water would be required for their removal from the soil. To overcome these problems we recently proposed a two-phase soil remediation method (Finzgar and Lestan, 2006b). In the first (leaching) phase, the soil is leached with EDTA solution (washing solution) to mobilize Pb and other co-contaminating metals. In the second phase, the washing solution is treated and heavy metals and EDTA removed. The clean solution is then used to rinse the soil in a closed loop and remove the EDTA and its complexes retained in the soil after the first (leaching) phase. We used leaching because it is not destructive of the soil (as for example soil extraction in a reactor), and since the relatively slow kinetics of soil solution treatment methods is compatible with the kinetics of soil leaching/ rinsing. To treat the washing solution, we used a combination of ozone and UV, an advanced oxidation process (AOP). AOP generated OH for the oxidative decomposition of EDTA–metal complexes (OH are one of the most powerful oxidants in aqueous solutions). The released metals were then removed from the washing solution by absorption. The method produced a discharge solution with a low concentration of EDTA and Pb. Very little process water was required. However, coloration and particles in the washing solution absorb and scatter UV light (Shu and Chang, 2005). Ozone-UV

1485

N. Finzgar, D. Lestan / Chemosphere 73 (2008) 1484–1491

based AOP was therefore efficient only for fairly colorless and nonturbid solutions, which are not usual in soil washing. Intensive yellow-brown coloration (due to the formation of Fe–EDTA complexes) was the probable cause of the excessively long treatment times that we often experienced with washing solutions with higher (>15 mM) EDTA concentrations (unpublished data). Another practical problem was removal of released metals, which consumed a significant quantity of expensive absorbent (Finzgar and Lestan, 2006b; Lestan and Finzgar, 2007), which needs to be later deposited. In our laboratory-scale experiments, we used pure oxygen from a tank as a source of ozone generation. In a larger-scale operation, equipment for moisture removal and oxygen concentration directly from air would be needed, together with an ozone generator and UV-lights (expensive and short-lifetime light sources). The cost of this equipment adds to the total remediation costs. Since ozone is toxic, collection and quenching of toxic ozone would also be needed (ozone is poorly water-soluble and undissolved ozone is lost for the process). In the current study, we tested an Electrochemical AOP (EAOP) for treating the washing solution as part of a two-phase remediation method. Pb, Zn and Cd contaminated vegetable-garden soil was treated with EDTA. EAOP became feasible with the recent development of a large area boron-doped diamond anode (BDDA) (Troster et al., 2002). It is more robust, technically simpler and cheaper than most AOP (Kraft et al., 2003). In EAOP, the electrode material is the most important parameter, since mainly molecular oxygen is produced during water electrolysis if the oxygen overvoltage is not sufficiently high. BDDA, however, has an extreme oxygen overvoltage (>3 V) before H2 (cathode) and O2 (anode) form. This electrochemical window allows production of hydroxyl radicals at the anode according to Eq. (1) directly from the electrolysed water at a high current efficiency (Kraft et al., 2003; Oliveira et al., 2007).

Table 1 Soil pedological parameters, Pb, Zn and Cd soil concentration and fractionation Soil properties pH (CaCl2) Organic matter (%) P (mg kg1) 1 CO 3 (g kg ) CEC (mmol C+ 100 g1) Sand (%) Silt (%) Clay (%) Texture Total Pb (mg kg1) Total Zn (mg kg1) Total Cd (mg kg1)

6.8 5.2 81.4 125.6 24.3 55.4 30.9 13.7 Sandy loam 1375 1007 9.1 ± 0.4

Fractionation

Pb

Zn

Cd

In soil solution (%) Exchangeable (%) Bound to carbonate (%) Bound to Fe and Mn oxides (%) Bound to organic matter (%) Residual fraction (%) Recovery (%)

0.02 ± 0.00 1.2 ± 0.0 26 ± 1 1.6 ± 0.0 57 ± 2 14 ± 1 90 ± 2

0.06 ± 0.00 0.33 ± 0.02 11 ± 0 7.1 ± 0.0 17 ± 2 65 ± 3 90 ± 3

0.00 ± 0.00 12 ± 0 36 ± 1 19 ± 0 13 ± 0 20 ± 1 102 ± 1

Where indicated standard deviation from mean value (n = 3) was calculated.

2.2. Rinsing of EDTA and EDTA–heavy metal complexes retained in the soil after leaching

BDDA is also extraordinary chemically inert and therefore suitable for treating various wastewaters. Previous use of BDDA EAOP includes destruction of persistent organic substances (i.e. benzenes, phenols, various pesticides and pharmaceutical drugs) in waste aqueous solutions (Canizares et al., 2005; Polcaro et al., 2005; Oliveira et al., 2007). Wastewaters containing EDTA have also been successfully treated with BDDA EAOP (Kraft et al., 2003). Yamaguchi et al. (2006) reported that EDTA was oxidized through sequential removal of the acetate groups until an unidentified small size hydrocarbon product was formed. This is the first study of the feasibility of using BDDA EAOP in remediation of heavy metal contaminated soils. To evaluate remediation efficiency, Pb, Zn and Cd concentration and mobility and Pb oral bioavailability were determined before and after remediation.

We determined the soil pore volume as the weight of moisture in soil saturated with deionized water per given weight of dry soil. The soil pore volume of 150 g of dry soil was 80 mL. We placed the soil (150 g, in triplicate) in perforated 250 mL polypropylene flasks with a 0.5 mm plastic mesh at the bottom to retain the soil. The soil was then leached by circulating 100 mL of the washing solution (5 and 20 mmol kg1 EDTA, disodium salt) for 48 h using a peristaltic pump (flow rate 1.3 ± 0.2 mL min1). The heavy metal and EDTA concentrations in the surplus (ca. 20 mL) washing solution were determined and the amount of soil residual heavy metals and EDTA calculated. The soil was further consecutively rinsed with one pore volume (80 mL), three pore volumes (additional 160 mL) and seven pore volumes (additional 320 mL) of deionized water (peristaltic pump, flow rate 1.3 ± 0.2 mL min1). The concentration of heavy metals and EDTA in the rinsing solution was measured and relative differences in EDTA and heavy metal removal after rinsing the soil calculated. The relative difference was defined as the difference between EDTA and heavy metals removed from the soil with the current pore volumes of water and removed with the previous pore volumes, relative to the total EDTA used and initial soil heavy metal content.

2. Materials and methods

2.3. Electrolytic cell

2.1. Soil samples and analysis

The flow through the electrolytic cell consisted of a BDDA (Diachem, Condias GmbH, Itzehoe, Germany) and two stainless steel cathodes with an electrode distance of 4.5 mm. The BDDA had a Ti base coated with a conductive polycrystalline diamond layer (the conductivity of the electrode was regulated by the addition of boron). The overall BDDA surface was 100 cm2. The surface area ratio between the cathodes and anode was 1:2. Current densities were adjusted (from 5 to 25 mA cm2) and cell voltage measured with a DC power supply (Elektronik Invent, Ljubljana, Slovenia). The flow of soil washing solution was ensured only through the space between the anode and cathodes. The electrode cell was cooled using a cooling mantle and tap water to keep the temperature of the treated washing solution below 35 °C.

H2 O !  OH þ e þ Hþ

ð1Þ

Soil was collected from the 0–30 cm surface layer of a vegetable garden in the Mezˇica Valley, Slovenia. The Mezˇica Valley has been exposed to more than 300 years of active lead mining and smelting. Soils in the valley, including 6600 ha of agricultural land, are polluted primarily with Pb but also with Zn and Cd. Soil pH was measured in a 1/2.5 (w/v) ratio of soil and 0.01 M CaCl2 water solution suspension. Soil samples were analyzed for organic matter by Walkley-Black titrations, cation exchange capacity by ammonium acetate and the Melich method, soil texture by the pipette method, easily extractable P was determined colorimetrically according to the Egner-Doming method, and carbonates manometrically after soil reaction with HCl (Table 1).

1486

N. Finzgar, D. Lestan / Chemosphere 73 (2008) 1484–1491

2.4. BDDA EAOP treatment of the EDTA soil washing solution To obtain the washing solution, we placed 4.5 kg of air-dried soil in 15 cm diameter soil columns (six replicates) and leached the soil with a 3000 mL aqueous solution of 5 mmol kg1 EDTA (disodium salt) for 64 h. Approx. 1000 mL of the washing solution per column was collected. To test the feasibility of using BDDA EAOP for treatment of the soil washing solution, we circulated 500 mL of soil washing solution from a 1-L Erlenmeyer flask through the electrode cell (peristaltic pump, flow rate 90 mL min1). Current densities used were 5, 15 and 25 mA cm2. Samples (30 mL) of washing solution were collected from 5 to 60 min of the contact time in the electrode cell. Contact time was calculated as a ratio of the electrode cell volume to the volume of the washing solution and multiplied by the operation time (initially 46 min of operation time equaled 10 minute of contact time). Samples were filtrated (filter paper grade 391, 84 g m2), the pH and EC measured, and stored in the cold for further analysis of Pb, Zn, Cd and EDTA concentrations. Filter papers with metallic precipitate were dissolved in aqua regia and analyzed (AAS) to determine the percentage of Pb, Zn and Cd that had been removed from the treated washing solution by filtration. The cathodes were rinsed with 10 mL 37% HNO3 to dissolve deposited metals and metals concentrations were determined by AAS. The percentage of Pb, Zn and Cd that was removed from the washing solution by electrodeposition was calculated. 2.5. Two-phase soil remediation using BDDA EAOP A two-phase soil remediation method with four EDTA additions was simulated in the laboratory. Air-dried soil (4.5 kg) was sieved (5 mm mesh) and placed in a 15 cm diameter column 24 cm high. Plastic mesh (0.2 mm) at the bottom of the column retained the soil. The soil was leached with a washing solution containing 5.0 mmol kg1 EDTA in 3 L unbuffered tap water (this volume was 145% of the soil water holding capacity). The washing solution was circulated in the first (leaching) phase (peristaltic pump, flow rate 14 mL min1) solely through the soil (pathway A, Fig. 1) for 72 h. In the second (rinsing) phase the washing solution circulated through the soil, electrode cell (current density 15 mA cm2) and three filters (pathway B, Fig. 1). High-efficiency particulate air filter for cooker-hoods available from the local home appliances store was used as filtering material. Thirty milliliter samples of washing

solution were collected from the column outlet after each 15 min of contact time in the electrode cell (10 minutes of contact time equaled 4 h and 43 min of operation time) and pH, EC and Pb, Zn, Cd and EDTA concentrations were determined. When the concentration of Pb in the treated washing fell below 50 mg L1, a fresh 5 mmol kg1 EDTA was added to the washing solution. Approximately 25% of the total initial water was added to the system after each EDTA addition, to compensate for water lost during the process (sampling, evaporation, electrolysis). After the fourth, last EDTA addition, the rinsing phase was prolonged to obtain discharge solution with sufficiently low heavy metal and EDTA concentrations. The soil column was dismantled and samples were taken from different soil layers (profile) for further determination of residual Pb, Zn and Cd. 2.6. Pb oral bioavailability Pb oral bioavailability before and after soil remediation was determined using the Physiologically Based Extraction Test (PBET, Ruby et al., 1996). The stomach phase of PBET was simulated by adding 1.25 g pepsin, 0.50 g citrate, 0.50 g malate, 420 lL lactic acid and 500 lL acetic acid to 1 L water and the pH was adjusted to 2.50 ± 0.05 using 12 N HCl. A 0.4 g soil sample (sieved through a 250 lm sieve) was mixed in a 250 mL polypropylene vessel with 40 mL of the simulated stomach solution. The contents of the vessel were agitated by bubbling water-saturated argon gas at a flow rate of 20 L h1. The vessel was suspended in a constant temperature bath at 37 °C. Samples (3 mL each) were collected after 60 min, centrifuged and decanted. After 1 h, the flask contents were titrated to pH 7 using a dialysis bag (8000 MWCO, Spectra/ Por cellulose ester tubing) containing 1 g of NaHCO3 and 2 mL of water. Twenty mg Pancreatin and 70 mg bile extract were added. Samples (3 mL) were obtained from small intestinal incubation for 1 h after the reaction flask had reached equilibrium at pH 7, been centrifuged, decanted and stored in cold storage (5 °C) for further analysis. PBET was conducted in triplicate. 2.7. Toxicity characteristic leaching procedure (TCLP) The mobility of Pb, Zn and Cd in the soil before and after remediation was determined using TCLP (US EPA, 1995), conducted in triplicate. The procedure involves shaking a 10 g soil sample in 200 mL of 0.0992 M acetic acid and 0.0643 M NaOH with a pH of 4.93 ± 0.05, for 18 h on a rotary shaker at about 300 rpm. At the end of the reaction period, the contents were filtered (filter paper grade 391, 84 g m2), acidified with concentrated HNO3 to pH < 2 and stored in cold storage (5 °C) for metal determination.

Filters 2.8. Six-step sequential extraction

T probe

Electrolytic ce cell

Heavy-metal Heavy-metal contaminated contaminated soil soil

A

_

+

Power suplly Cooling water

A sequential extraction procedure (Lestan et al., 2003) was used to determine the fractionation of Pb, Zn and Cd in nonremediated and remediated soil into six fractions: soluble in soil solution, exchangeable from soil colloids, bound to carbonates, bound to Fe and Mn oxides, bound to organic matter and residual fraction soluble in aqua regia. Three determinations of Pb, Zn and Cd concentration were realized for each fractionation sequence. The final fractional recovery of Pb, Zn and Cd was calculated after summing the recoveries of all six steps of sequential extractions.

B 2.9. EDTA determination Fig. 1. Flow sheet of the novel two-phase soil remediation method using EDTA for heavy metal leaching and BDDA EAOP for treatment and reuse of washing solution in a closed loop. A (leaching phase): washing solution first circulates solely through soil. B (rinsing phase): washing solution treatment, soil rinsing phase in which the washing solution circulates through the soil and treatment unit.

Samples of washing and soil rinsing solution were filtrated (filter paper grade 391, 84 g m2) and EDTA determined spectrophotometrically according to the procedure of Hamano et al. (1993).

1487

N. Finzgar, D. Lestan / Chemosphere 73 (2008) 1484–1491

mg L1 Pb, 119 ± 5 mg L1 Zn, 4.3 ± 0.2 mg L1 Cd, 303 ± 17 mg L1 Ca and 54 ± 5 mg L1 Fe. The initial EDTA concentration was 2057 ± 121 mg L1 (5.53 ± 0.33 mM) and the pH of the washing solution was 8.1. Fig. 2 shows that BDDA EAOP treatment efficiently removed Pb, Zn, Cd and EDTA from the washing solution; almost equally well at current densities of 15 and 25 mA cm2. After 60 min contact time, 99% of Pb and EDTA, 92% of Zn and 90% of Cd were removed (using 25 mA cm2). The voltage slightly increased during the treatment (from 4.8 ± 0.3 to 5.2 ± 0.2 V, from 8.2 ± 0.5 to 9.4 ± 0.1 V, and from 11.6 ± 0.9 to 12.6 ± 0.5 V for current densities 5, 15 and 25 mA cm2, respectively). The pH of the washing solution remained unchanged at 5 mA cm2 and slightly increased at higher current densities (pH 8.14–8.52 and 8.74– 8.98 at 15 and 25 mA cm2). Kraft et al. (2003) used BDDA EAOP to treat industrial wastewater with approx. 9-times higher EDTA concentration than in our washing solution. They reported efficient degradation even at a current density as low as 7.5 mA cm2. However, the EDTA in their case was in non-complexed, acidic form and the geometry of their electrolytic cell and electrode materials also differed. The efficiency of EDTA degradation by EAOP probably depends on many parameters: metals complexed to EDTA, concentration of reactants, presence of electrolytes, pH and temperature. Some of these relations have been studied for AOP but the results have not been univocal. Jiraroj et al. (2006), for example, investigated the effect of pH in EDTA–Pb degradation with H2O2/UV AOP and found a faster degradation rate when the starting pH was more acidic (pH 3). In contrast, Chitra et al. (2003) reported that alkaline pH favored faster degradation of EDTA compared to acidic and neutral pH. One of the key advantages that we expected from using BDDA EAOP instead of ozone/UV AOP (Finzgar and Lestan, 2006b) was simple removal of heavy metals from the washing solution. Indeed, metals released after oxidative degradation of EDTA complexes either deposited on the cathode (after electroreduction) or precipitated from the solution, in both cases enabling effective separation. As shown in Table 2, a maximum 21% of Pb and less than 1% of Zn and Cd electrodeposited (percentages decreased with current density). The percentage of metal recovery by electrodeposition decreased in the order Pb > Fe > Cd > Zn > Ca (Table 2) which is

2.10. Heavy metal determination Air-dried samples of non-leached and leached soil (1 g) were ground in an agate mill, digested in aqua regia (28 mL), diluted with deionized water up to 100 mL, and Pb, Zn and Cd analyzed by flame (acetylene/air) AAS with a deuterium background correction (Varian, AA240FS). Pb, Zn and Cd in washing and rinsing solutions and in PBET and TCLP extracts were determined by AAS directly. A standard reference material used in inter-laboratory comparisons (ALVA Boden 1) from HBLFA Raumberg-Gumpenstein, Irdning, Austria, was used in the digestion and analysis as part of the QA/QC protocol. The recovery percentages were 95 ± 2%, 109 ± 5% and 97 ± 1%, and the limits of quantification (LQ) were 0.25, 0.02 and 0.03 mg L1 for Pb, Zn and Cd, respectively. Reagent blank and analytical duplicates were also used where appropriate to ensure accuracy and precision in the analysis. 2.11. Analysis of metallic precipitates A low vacuum scanning canning electron microscope (SUPRA 35 VP, Carl Zeiss) coupled with an energy dispersive X-ray spectroscope (Inca 400, Oxford Instruments) (SEM-EDS) was used to analyze the composition of metallic insoluble precipitates filtered from the washing solution obtained after BDA EAOP soil treatment (magnification 30–1600 times, accelerated voltage 20 kV, pressure 15–24 Pa). 2.12. Statistics The Duncan multiple range test was used to determine the statistical significance (P < 0.05) between different treatments, using the computer program Statgraphic 4.0 for Windows. 3. Results and discussion 3.1. Feasibility of BDDA EAOP for treatment of EDTA soil washing solution The concentrations of major metals present in the soil washing solution before treatment in the electrolytic cell was 495 ± 36

600

400 300 200

4 -1

-1

15 mA cm-2 25 mA cm-2

Cd (mg L )

500 Pb (mg L )

5

5 mA cm-2

0

0 2500

140

2000

-1

EDTA (mg L )

120 -1

2 1

100

Zn (mg L )

3

100 80 60 40

1500 1000 500

20 0

0 0

10

20

30

40

Contact time (min)

50 5

60

0

10

20

30

40

50

60

Contact time (min)

Fig. 2. Effect of current density on Pb, Zn, Cd and EDTA removal from the washing solution using BDDA EAOP. Error bars represent standard deviation from mean value (n = 3).

N. Finzgar, D. Lestan / Chemosphere 73 (2008) 1484–1491

Pb Zn Cd Fe Ca

19.8 ± 3.8 0.40 ± 0.23a 0.66 ± 0.21b 7.0 ± 1.2a 0.10 ± 0.01b

15 (A cm2) b

20.9 ± 4.0 0.49 ± 0.15a 0.59 ± 0.18a,b 7.1 ± 2.4a 0.17 ± 0.10b

25 (A cm2) 6.1 ± 1.9a 0.26 ± 0.8a 0.31 ± 0.05a 5.8 ± 1.7a 0.01 ± 0.01a

Means (n = 3) followed by the same letters are not significantly different, according to the Duncan test (P < 0.05).

probably related to the reduction potential of free metal ions; i.e. 0.126 V for Pb2+/Pb and 2.87 for Ca2+/Ca (Oldham and Myland, 1994). Juang and Wang (2000) studied the electrodeposition recovery of metals from simulated EDTA washing solution and obtained a similar sequence. The majority of metals were precipitated from the treated washing solution and easily removed by filtration. Some precipitates were probably removed from the cathode surface during the treatment. The voltages we used during electrolysis were much above the reduction potential of free metal ions and such conditions lead to the formation of powdery electrodeposits, which adhere poorly to the cathode surface and are easily removable (Doulakas et al., 2000). The nature of precipitates was probed by SEM-EDS. The maximum and minimum percentages of elements obtained from eight points (individual grains) were 15.9–73.7% Fe, 10.4–72.6% O, 7.4–8.4% Pb and 6.2–6.3% Ca. Zn and Cd were not detected. Concentrations of Zn and especially Cd in the washing solution were much lower than the concentration of Pb (Fig. 2). They were probably not detected because they precipitated in individual grains, which were missed by our scan. Participate was composed mostly of Fe, indicating the known problem of high EDTA affinity for Fe and thus low selectivity of chelant for heavy metals (Sun et al., 2001; Finzgar and Lestan, 2007). Also interesting was a high content of oxygen. This could indicate that some metals were precipitated as oxides or hydroxides. One possible mechanism is the anodic oxidation of iron by hydroxyl radicals (Chitra et al., 2004). If this reaction (Eq. (2)) really did occur, it would mean a considerable loss of hydroxyl radicals. 2þ

Fe



2þ

þ OH ! ðFeOHÞ

A

3.2. Rinsing of EDTA and EDTA–heavy metal complexes from the soil After soil leaching (and soil washing in general) part of the EDTA and some EDTA–heavy metal complexes remain in the soil, trapped in soil pores or absorbed into the soil solid surfaces. Since

5 mmol

35

20 mmol

30 25 20 15 10 5 0

B

10 8 6 4 2 0

C

ð2Þ

The efficiency of metal removal from the washing solution was probably hampered by the low hydrogen overpotential of the stainless steel (Fe) cathode that we used (Huang et al., 2000). This means that hydrogen evolution was more favorable than metal reduction. Huang et al. (2000) used a Cd cathode with a high hydrogen overpotential and reported that over 90% of Fe was electrodeposited during the electrolytic treatment of EDTA–Fe containing boiler cleaning wastewaters. We did not evaluate the efficiency of BDDA EAOP for treatment of washing solutions with EDTA concentrations higher than 5.53 ± 0.33 mM. However, according to Juang and Wang (2000), higher electrolytic efficiency can be expected at higher EDTA complex concentrations because of the increase in solution conductivity. This is important since using ozone/UV AOP (Finzgar and Lestan, 2006a) was not effective for treatment of soil washing solutions with more than 15 mM, probably due to intensive coloration and turbidity.

40

Pb removal rel. difference (%)

b

Zn removal

5 (A cm2)

rel. difference (%)

Recovery (%)

EDTA and its complexes are water-soluble, mobile and bioavailable (potentially toxic), they should be removed from the soil. This may require a vast consumption of clean water and increase the cost of remediation. Despite the practical implications, the phenomenon has not been previously investigated. We set up a small-scale experiment in which we leached the soil with 5 and 20 mmol kg1 EDTA and then rinsed the soil with different soil pore volumes of deionized water to remove soil residual EDTA and EDTA–heavy metal complexes. The relative differences in Fig. 3 depict the percentage of heavy metals and EDTA

35

Cd removal rel. difference (%)

Table 2 Effect of current density on metal recovery onto the cathode

30 25 20 15 10 5 0

D

70

EDTA removal rel. difference (%)

1488

60 50 40 30 20 10 0 1

3

7

Pore volume Fig. 3. Removal of EDTA and heavy metals, remaining in the soil after leaching (5 and 20 mmol kg1 EDTA), by rinsing the soil with deionized water. The relative differences in removed EDTA and Pb, Zn and Cd after rinsing the soil with 1, 3 and 7 soil pore volumes of water are presented. (Relative difference = (EDTA, heavy metals removed by rinsing  removed with previous rinsing)/total EDTA used, initial heavy metal content.) Error bars represent standard deviation from mean value (n = 3).

1489

N. Finzgar, D. Lestan / Chemosphere 73 (2008) 1484–1491

removed from the soil after rinsing with increasing amounts (soil pore volumes) of deionized water. Most mobile heavy metal species left in the soil after leaching, were removed with the first pore volume (Fig. 3a–c). However, rinsing with three pore volumes also removed a significant part of initial heavy metals (i.e. 7.6 ± 0.8% of Pb, 20 mmol kg1 EDTA treatment). After rinsing with seven pore volumes, the heavy metals left in the soil seemed predominantly immobile and could not be removed further by rinsing with water. Rinsing the EDTA was a bit less efficient but followed a similar pattern (Fig. 3D). After the seventh pore volume, approx. 26% and 25% of applied EDTA (5 and 20 mmol kg1 EDTA treatments) remained in the soil, apparently tightly bound to the soil solid phases and not available for rinsing with water. The fate of this part of the EDTA left in the soil after remediation, and the potential of bound EDTA to reactivate, remains unknown. Overall, rinsing the soil with seven pore volumes of clean water was sufficient (this translates into 3.7 l of clean water for rinsing 1 kg of soil).

metals were removed by simple etching with nitric acid after each EDTA soil treatment. After four EDTA (4  5 mmol kg1) treatments, 44% Pb, 14% Zn and 35% Cd was removed from the soil, quite uniformly through the soil profile (data not shown). This was expected, since in general metals in soil are not entirely accessible to chelants, especially in soils rich in organic matter such as ours (Table 1). Nowack et al. (2006) compiled data from 28 publications. Except in some reports for Pb, complete removal did not occur, even at a chelant-to-metal ratio of greater than 10. Of Pb, Zn, Cd and Cu, Zn was the least extractable, the same as in our study. This could be partly explained by the stability constants (Ks) of EDTA complex formation. For example log Ks of EDTA–Pb is 18.0 (at 25 °C and ionic strength l = 0.1) while log Ks for EDTA–Zn is lower: 16.5 (Martell and Smith, 2003). Another possible reason for the low Zn removal was the high percentage of non-extractable Zn bound to the residual soil fraction (65 ± 3%), while most of the Pb was found in the more labile organic matter and carbonate fraction (57 ± 2%, 26 ± 1%) (Table 1). It is likely that in our experiment, the remediation efficiency was lower due to (unidentified) metallic precipitates in the washing solution that were small enough to pass through the filtering system (Fig. 1) and re-contaminate the soil. This was difficult to avoid in a laboratory bench scale-experiment, in which separation methods other than filtration are not practical. In a full-scale process, the use of flotation or sedimentation could prove more prudent. Treatment of soil washing solution with BDDA EAOP was less efficient at lower heavy metals/EDTA concentrations. Consequently, an exceedingly long treatment time was required in the last, 4th step to obtain sufficiently clean waste solution for safe discharge (Fig. 4). Efficiency could presumably be improved by pre-concentration of the washing solution (i.e. by reverse osmosis) before BDDA EAOP treatment in the final stages of remediation. Since our method recycles and thus requires very little process water, the pre-concentration should not require large equipment or significantly increase operational costs. The discharge solution was clear, almost colorless, slightly basic (pH

3.3. Feasibility of two-phase remediation using BDDA EAOP Our two-phase remediation method addresses the requirements for rinsing water (to remove soil residual EDTA and EDTA–heavy metal complexes) in the second (rinsing) phase (pathway B, Fig. 1). In a laboratory scale simulation of the method, the concentrations of Pb, Zn and Cd in the washing solution (measured immediately after the first, leaching phase) decreased with each consecutive EDTA soil treatment (Fig. 4). This was expected, since with each treatment less and less heavy metals remained EDTA extractable. During the second (rinsing) phase, the washing solution was treated by BDDA EAOP to remove EDTA and heavy metals, and was then re-used for soil rinsing in a closed process loop. Based on experiments of BDDA EAOP treatment of soil washing solution (described above), a constant current density of 15 mA cm2 was applied in the electrolytic cell. The voltage varied between 7.4 and 10.2 V. The suspended metallic precipitates were effectively filtrated, while electrodeposited

0

4.5 4 3.5 3 2.5 2 1.5 1 0.5 0

120

2500

100

2000

600

Cd (mg L )

400

-1

-1

Pb (mg L )

500

300 200

-1

EDTA (mg L )

-1

Zn (mg L )

100

80 60 40

I. II. III. IV.

1500 1000 500

20 0

0 0

20

40

60

80

Contact time (min)

100

120

0

20

40

60

80

100

120

Contact time (min)

Fig. 4. Concentrations of Pb, Zn, Cd and EDTA in washing solution during the second, washing solution treatment (using BDDA EAOP)/soil rinsing phases of the two-phase soil remediation. Soil was leached with four-consecutive additions (I.–IV.) of 5 mmol kg1 EDTA.

1490

N. Finzgar, D. Lestan / Chemosphere 73 (2008) 1484–1491

Table 3 Oral bioavailability of Pb (PBET) and mobility (TCLP) of metals before and after soil remediation Before (mg L1)

After (mg L1)

PBET Stomach Intestine

126 ± 28b 43 ± 11b

52 ± 22a 25 ± 6a

TCLP Pb Zn Cd

0.81 ± 0.06b 1.21 ± 0.02b 0.06 ± 0.00b

0.51 ± 0.05a 0.36 ± 0.04a 0.04 ± 0.01a

Means followed by different letters are significantly different according to the Duncan test (P < 0.05).

7.73, EC 2.45 mS cm1) containing 0.47 mg L1 Pb, 1.033 mg L1 Zn, below LQ of Cd and 8.67 mg L1 (0.023 mM) EDTA. For successful remediation, the heavy metals left in the soil should remain in non-mobile and non-bioavailable (non-toxic) forms. The mobility of heavy metals is, for example, a potential threat for groundwater contamination. We determined mobility as the concentration of heavy metals in TCLP soil extracts (Table 3). Even before remediation, the mobility of the metals was quite low, much lower than the TCLP regulatory limit set at 5 mg L1 for Pb, 250 mg L1 for Zn and 1 mg L1 for Cd (when soil is considered to be hazardous waste; US EPA, 1995). Nevertheless, after remediation, the mobility of Zn deceased much further, by more than 3-times, despite poor (14%) Zn removal. Ingestion of soil and dust particles is also an important form of exposure to soil pollutants, especially with children, due to their mouthing behavior (Davis and Mirick, 2006). The PBET model for oral bioavailability was designed to simulate the human gastrointestinal tract, which includes stomach and intestinal phases. Since Ruby et al. (1996) have validated results from the PBET model using animal models only for Pb and As, we did not assess Zn and Cd oral bioavailability. Soil remediation reduced the concentration of Pb available in the stomach phase by 2.5-times and significantly reduced the concentration in the intestinal phase (Table 3). Overall heavy metals left in the soil after remediation were fairly immobile and benign. An accurate evaluation of the costs associated with our novel two-phase BDDA-EAOP soil remediation method would require a pilot-scale experiment (after further process optimization). However, EDTA and electricity consumption (which presumably represent the major part of the total costs) can be extrapolated from our bench-scale experiment. Leaching 1 ton of soil would require 7.5 kg of EDTA. At a price 1.3 € per kg1 EDTA this translates into 9.75 €. Treatment of the washing solution (without appropriate scale-up of the equipment) would require 22 933 h at a constant current of 1.5 A and average voltage 9 V. This is 310 kW h and at an approx. cost 0.1 € per KW h translates to 31 €. The approx. EDTA and electricity cost would be therefore 41 € ton1 (43 € m3) of soil. This is not the total cost, but nevertheless seems favorable. The current cost of soil washing can go up to 450 € per m3 (Summergill and Scott, 2005). 4. Conclusions BDDA EAOP is a viable method for the treatment of soil washing solutions containing EDTA complexes with Pb, Zn, Cd, and also as part of a two-phase (soil leaching, soil rinsing) soil remediation method. After BDDA EAOP treatment, heavy metals were easily removed from the washing solution by filtration of insoluble metallic precipitates and as electro-deposits on a cathode. After four EDTA (4  5 mmol kg1) treatments, only part of Pb, Zn and Cd was removed from the soil, indicating that heavy metals were not entirely accessible to chelant. However, heavy metals left

in the soil were in fairly non-mobile, non-bioavailable forms. This was accomplished by the second (rinsing) phase of the remediation method, in which EDTA and mobile, bioavailable EDTA–heavy metal complexes retained in the soil after the first (leaching) phase were efficiently removed by soil rinsing with recycled process solution. The final solution was almost free of metals and EDTA and safe for discharge. In further studies, we will follow two directions. First, we will evaluate the feasibility of two-phase, BDA EAOP based remediation for soils contaminated with other heavy metals (e.g., Cu and U). We will also focus on scale-up, further process optimization and costefficiency evaluation of the proposed remediation method. Acknowledgment This work was supported by the Slovenina Ministry for Education, Science and Sport, Grant J4-6134-0481-04/4.03. References Canizares, P., Lobato, J., Paz, R., Rodrigo, M.A., Saez, C., 2005. Electrochemical oxidation of phenolic wastes with boron-doped diamond anodes. Water Res. 39, 2687–2703. Chaney, R.L., Malik, M., Li, Y.M., Brown, S.L., Angle, J.S., Baker, A.J.M., 1997. Phytoremediation of soil metals. Curr. Opin. Biotechnol. 8, 279–284. Chitra, S., Paramasivan, K., Sinha, P.K., Lal, K.B., 2003. Treatment of liquid waste containing ethylenediamine tetraaceticacid by advanced oxidation processes. J. Adv. Oxid. Technol. 6, 109–114. Chitra, S., Paramasivan, K., Sinha, P.K., Lal, K.B., 2004. Ultrasonic treatment of liquid waste containing EDTA. J. Clean. Prod. 12, 429–435. Davis, S., Mirick, D.K., 2006. Soil ingestion in children and adults in the same family. J. Expo. Sci. Env. Epid. 16, 63–75. Doulakas, L., Novy, K., Stucki, S., Comninellis, Ch., 2000. Recovery of Cu, Pb, Cd and Zn from synthetic mixture by selective electrodeposition in chloride solution. Electrochim. Acta 46, 349–356. Finzgar, N., Lestan, D., 2006a. Advanced oxidation for treatment of aqueous extracts from EDTA extraction of Pb and Zn contaminated soil. J. Environ. Eng. – ASCE 132, 1376–1380. Finzgar, N., Lestan, D., 2006b. Heap leaching of Pb and Zn contaminated soil using ozone/UV treatment of EDTA extractants. Chemosphere 63, 1736–1743. Finzgar, N., Lestan, D., 2007. Multi-step leaching of Pb and Zn contaminated soils with EDTA. Chemosphere 66, 824–832. Hamano, T., Mitsuhashi, Y., Kojma, N., Aoki, N., 1993. Sensitive spectrophotometric method for the determination of ethylene-diaminetetraacetic acid in foods. Analyst 118, 909–912. Huang, J.W., Chen, J.J., Berti, W.R., Cunningham, S.D., 1997. Phytoremediation of lead-contaminated soils: role of synthetic chelates in lead phytoextraction. Environ. Sci. Technol. 31, 800–805. Huang, C.P., Hsu, M.-C., Miller, P., 2000. Recovery of EDTA from power plant boiler chemical cleaning wastewater. J. Environ. Eng. – ASCE 126, 919–924. Juang, R.-S., Wang, S.-W., 2000. Metal recovery and EDTA recycling from simulated washing effluents of metal-contaminated soils. Water Res. 34, 3795–3803. Jiraroj, D., Unob, F., Hagege, A., 2006. Degradation of Pb–EDTA complex by H2O2/UV process. Water Res. 40, 107–112. Kraft, A., Stadelmann, M., Blaschke, M., 2003. Anodic oxidation with doped diamond electrodes: a new advanced oxidation process. J. Hazard. Mater. 103, 247–261. Lestan, D., Finzgar, N., 2007. Leaching of Pb contaminated soil using ozone/UV treatment of EDTA extractants. Separ. Sci. Technol. 42, 1575–1584. Lestan, D., Grcman, H., Zupan, M., Bacac, N., 2003. Relationship of soil properties to fractionation of Pb and Zn in soil and their uptake into Plantago lanceolata. Soil Sediment Contam. 12, 507–522. Martell, A.E., Smith, R.M., 2003. NIST Critically Selected Stability Constants of Metal Complexes; Version 7.0, NIST, Gaithersburg. Nowack, B., Sigg, L., 1995. Adsorption of EDTA and metal–EDTA complexes onto geothite. J. Colloid Interf. Sci. 177, 106–121. Nowack, B., Schulin, R., Robinson, B.H., 2006. Critical assessment of chelantenhanced metal phytoextraction. Environ. Sci. Technol. 40, 5225–5232. Oldham, K.B., Myland, J.C., 1994. Fundamentals of Electrochemical Science. Academic Press, San Diego. Oliveira, R.T.S., Salazar-Banda, G.R., Santos, M.C., Calegaro, M.L., Miwa, D.W., Machado, S.A.S., Avaca, L.A., 2007. Electrochemical oxidation of benzene on boron-doped diamond electrodes. Chemosphere 66, 2152–2158. Polcaro, A.M., Vacca, A., Mascia, M., Palmas, S., 2005. Oxidation at boron doped diamond electrodes: an effective method to mineralise triazines. Electrochim. Acta 50, 1841–1847. Ruby, M.V., Davis, A., Schoof, R., Eberle, S., Sellstone, C.M., 1996. Estimation of lead and arsenic bioavailability using a physiologically based extraction test. Environ. Sci. Technol. 30, 422–430. Shu, H.Y., Chang, M.C., 2005. Pilot scale annular flow photoreactor by UC/H2O2 for the colorization of azo dye wastewater. J. Hazard. Mater. 125, 244–251.

N. Finzgar, D. Lestan / Chemosphere 73 (2008) 1484–1491 Summergill, I.M., Scott, D.W., 2005. Remediation technology costs in the UK & Europe. In: Proceedings of the 9th International FZK/TNO Conference on Soil– Water Systems, Bordeaux, France. Sun, B., Zhao, F.J., Lombi, E., McGrath, S.P., 2001. Leaching of heavy metals from contaminated soils using EDTA. Environ. Pollut. 113, 111–120. Troster, I., Schafer, I., Fryda, M., 2002. Recent development in production and application of DiaChemÒ electrodes for wastewater treatment. New Diam. Front. C. Tec. 12, 89–97.

1491

US EPA, 1995. Test Methods for Evaluation of Solid Waste vol. IA, Laboratory Manual Physical/Chemical Methods, SW 846, 40 CFR Parts 403 and 503. 3rd Ed. US Government Printing Office, Washington, DC. Yamaguchi, Y., Yamanaka, Y., Miyamoto, M., Fujishima, A., Honda, K., 2006. Hybrid electrochemical treatment for persistent metal complexes at conductive diamond electrodes and clarification of its reaction rute. J. Electrochem. Soc. 153, 1123–1132.