Environ. Sci. Technol. 2005, 39, 2906-2911
Electrodialytic Removal of Cu, Zn, Pb, and Cd from Harbor Sediment: Influence of Changing Experimental Conditions GUNVOR M. NYSTROEM,* LISBETH M. OTTOSEN, AND ARNE VILLUMSEN Department of Civil Engineering, Technical University of Denmark, Kemitorvet - Building 204, DK-2800 Kgs. Lyngby, Denmark
Electrodialytic remediation (EDR) was used to remove Cu, Zn, Pb, and Cd from contaminated harbor sediment. Extraction experiments were made prior to EDR, and the metal desorption was pH dependent but not liquid-to-solid ratio (L/S) dependent. The desorption order was Cd > Zn > Pb > Cu. Electrodialytic experiments were made with HCl as desorbing agent in a sediment suspension, which was stirred during EDR. Effects of different current strengths and L/S ratios on the heavy metal removal were investigated on wet and air-dried sediment. The effects of drying the sediment were negligible for the removal of Cu, Zn, and Pb, probably due to oxidation of the sediments during stirring. Contrary, Cd removal was lower in the wet sediment as compared to the air-dried. The heavy metal removal was influenced by higher current strengths and varying L/S ratios. The highest removal obtained was in an experiment with dry sediment (L/S 8) and a 70 mA applied current that lasted 14 days. These experimental conditions were thereafter used to remediate more strongly contaminated sediments. Regardless of the initial heavy metal concentrations in the sediments, 67-87% Cu, 7998% Cd, 90-97% Zn, and 91-96% Pb were removed.
Introduction Sediment is dredged from harbors and channels to maintain navigational depths. Harbors are often located in or close to industrial and urban areas, which produces outlets that end up in the waterways. Hence, harbor sediment is often contaminated, and the contamination consists of different organic, inorganic, and organo-metallic pollutants, depending on activities in the harbors’ nearby area. Actions have been taken over many years to reduce the outlets of pollutants by, for example, cleaner technology, wastewater treatment, and ban on toxic antifoulants, but the contamination is still accumulating in the sediments. Contaminated sediments are either maintenance dredged to maintain navigational depths or dredged for environmental purposes to remove a potential environmental risk the sediments can pose. In Europe, lightly contaminated sediments are typically disposed of at sea and highly contaminated sediments are typically disposed of in lagoons, at land for land-reclamation, or contained in confined disposal facilities (CDF). Sea disposal * Corresponding author phone: (+45) 45 25 22 55; fax: (+45) 45 88 59 35; e-mail:
[email protected]. 2906
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 8, 2005
of sediments can have adverse effects on aquatic organisms and the water quality, and finding suitable disposal areas on land can be difficult because dredged sediments generally have a high water content. Remediation initiatives are made to reduce the amount of contaminated sediments for disposal, especially in the large harbors. The remediation methods include, among others, separating the sediment into contaminated and clean fractions to reduce the volume (1) or using sediment as raw material for, for example, brick production (2). The use of sediment as raw material requires sufficient and continuous supply of homogeneous sediment to maintain the production, which is only possible for large harbors with frequent dredging. In Denmark, about 3 000 000 m3 sediment is dredged annually from harbors and channels (3). Uncontaminated sediment is traditionally dumped at sea. However, 10-25% of the dredged sediment is polluted and needs other end solutions. Disposal in CDFs of this polluted sediment is space consuming, but alternative treatment methods or reuse options have not yet been developed in Denmark. Instead of disposal, the sediments could be decontaminated and used for, for example, land reclamation. Different investigations with the aim to remove contaminants from sediment have been conducted in laboratory and pilot scale. These include electrokinetic degradation of tributyltin (TBT) (4), bioremediation of organic contaminants (5), and removal of heavy metals by washing with biosurfactants (6). Another method to remove heavy metals from contaminated harbor sediment could be electrodialytic remediation, where a low voltage direct current is the cleaning agent, as suggested by ref 7. Electrodialytic remediation is particularly useful for fine grained harbor sediments where traditional soil remediation technologies, for example, extraction techniques, are impractical or even impossible to use (8, 9). In this work, electrodialytic remediation in laboratory scale was used to remove heavy metals from harbor sediments. Heavy metal contaminated sediments were used in this study, to investigate the decontamination potential of the electrodialytic method. The availability of heavy metals increases when dredged sediment is oxidized (10, 11), so electrodialytic remediation experiments were made with both wet and air-dried dredged sediment to investigate the potential removal difference. In electrodialytic remediation, increased availability of the heavy metals can result in higher desorption from the sediment, which can increase the heavy metal mobility in the applied electric field. Investigations on how varying current strengths and liquid-to-solid (L/S) ratios (mL/g) of the sediment suspension influenced the electrodialytic removal were also made, to find the most efficient remediation conditions, which were thereafter tested on two other contaminated harbor sediments.
Electrodialytic Remediation Electrodialytic remediation of solid waste materials started to develop at the Technical University of Denmark in 1992. The method was first developed for remediation of heavy metal polluted soil (12). The method has also shown potential in laboratory scale for removal of heavy metals from fly ash and impregnated wood waste (13) and for impregnated wood waste in pilot scale (14). Recently, the method was successfully tested in preliminary experiments for removal of heavy metals from harbor sediments (7). The principle of electrodialytic remediation is shown in Figure 1. 10.1021/es048930w CCC: $30.25
2005 American Chemical Society Published on Web 03/11/2005
FIGURE 1. The electrodialytic cell, showing a selection of the species present in the electric field. CAT, cation-exchange membrane; AN, anion-exchange membrane; 1, anode side; 2, cathode side. In electrodialytic remediation, a low voltage direct current is used, and, in the applied electric field, ions in the liquid phase of the sediment electromigrate toward the electrodes with the opposite charge. Electrolytes circulate in compartments I, II, IV, and V. Ion-exchange membranes (cation- and anion-exchange membranes) separate the different compartments, and the ion-exchange membranes act as both a chemical and a physical barrier between the compartments. A contaminated sediment suspension is placed in the middle compartment (III), where the suspension is stirred. The stirring principle was introduced by ref 15 for fly ash, and it was shown that stirring of an ash suspension improved the remediation as compared to remediation of a stationary ash, used in traditional electrodialytic remediation soil (12). The stirred system also showed the best potential for treatment of harbor sediment (7). An advantage for the stirred system is that concentration profiles do not develop in the sediment. The removed heavy metals are concentrated in compartments II and IV. CAT 1 also prevents Cl- in reaching the anode, where Cl2 (g) would be formed. Mobilization of heavy metals is normally based on the development of an acidic front in the contaminated material (16). The acidic front is produced by water-splitting at AN 1, and the acid desorbs/dissolve the available parts of heavy metals from the contaminated material. Desorbing agents such as ammonium citrate or citric acid can also be added to the contaminated material to enhance desorption of the heavy metals (17, 18) and thus increase the remediation.
Experimental Section Analytical Methods. Heavy metal concentrations in the sediment were determined after acid digestion according to Danish Standard DS 259 (19), and the heavy metals were measured by AAS (atomic absorption spectrometry). Organic matter was measured by loss of ignition at 550 °C for 1 h. The carbonate content was determined by a volumetric calcimeter method (20). pH was determined in 1 M KCl at L/S 5, and, after 1 h of agitation, pH was measured by a Radiometer Analytical pH electrode after calibration against Radiometer Analytical pH standards. Water content in the sediment was determined as weight loss after 24 h at 105 °C. Redox potential was measured directly in the sediment with a Radiometer Analytical redox electrode by inserting the electrode in the sediment and waiting for a stable reading. Particle size distribution was determined by wet (particles < 63 µm) and dry (particles > 63 µm) sieving after the sediment was separated at 63 µm by wet sieving. Experimental Harbor Sediment. Norwegian harbor sediment from Haakonsvern, Bergen, was the main sediment used in this study, and the heavy metals of interest were Cu, Zn, Pb, and Cd. Extraction experiments and seven electrodialytic experiments were made with this sediment, to investigate different electrodialytic remediation conditions. The harbor sediment was dredged by NCC Norway with a special dredging device for removing only the finer, polluted material (60% Cd and Pb, and >50% Cu, but at very low pH. Figure 2b shows the L/S-dependent metal extraction. The pH in all of the extractions was 1.8 ( 0.2. It was seen that increasing the L/S ratio did not have any influence on desorption of metals at this pH, which indicates that desorption was not solubility controlled under these conditions. In the L/S batch extractions, 19-22% Cu, 47-50% Pb, 51-55% Cd, and 69-75% Zn were desorbed. Electrodialytic Experiments. The removal in the electrodialytic experiments is calculated as the amount of metal found in the electrolytes I, II, IV, and V, electrodes and membranes after the experiment as compared to the total amount found in the cell (electrolytes, electrodes, membranes, and sediment suspension) after the experiment. Dissolution of sediment seen in a mass decrease after remediation was observed in all of the electrodialytic experiments, due to the acidification. pH in the sediment after the experiments varied (1.9-2.5). This was lower than expected just from addition of acid and may be due to further acidification of the sediment by water-splitting at the anionexchange membrane (AN 1). Generally, the final pH in the experiments made with air-dried sediment was lower than the experiments made with wet sediment, but a reason for this was not identified. Electrodialytic remediation on air-dried or wet sediment did not influence the removal of Cu, Zn, or Pb, but more Cd was removed in the air-dried sediment as compared to the wet (Figure 3). The final Cd concentrations in the two sediments were similar (0.25 mg/kg in the air-dried and 0.29 mg/kg in the wet). However, in the experiment with the wet
FIGURE 3. Heavy metal removal from air-dried (experiment 1) and wet (experiment 2) Haakonsvern sediment.
FIGURE 4. Removal of metals in control extraction and electrodialytic experiments with varying current strength; all experiments lasted 14 days and were made with wet Haakonsvern sediment. sediment, a Cd concentration 8 times higher was found in the liquid in the middle compartment as compared to the experiment with the air-dried sediment. This means that similar amounts of Cd were released from the two sediments, but in the experiment with wet sediment, all Cd was not removed from the middle compartment, which could be due to uncharged Cd complexes, which will form in the presence of Cl- (23). pH, electrical conductivity, and redox potential were generally similar in the middle compartments of the two experiments. The wet sediment suspension was aerobic from the beginning of the experiment, probably due to the addition of acid. The two suspensions were continuously oxidized via stirring. The initial electrical conductivity was higher in the wet sediment (170 mS/cm) than in the dry (55 mS/cm), but in both it rapidly decreased and reached the same level between 30 and 40 mS/cm. This high electrical conductivity gave a low potential over the cells (3-7 V) but also meant that limited amounts of the applied current were used to remove heavy metal ions. In Figure 4, the heavy metal removal in the control experiment and the electrodialytic experiments with varying current strength, made with wet sediment, is shown. For the control experiment, the removal is defined as the amount of heavy metals found in the extracted solution as compared to the amount of heavy metals found both in the sediment and in the extracted solution. The removal of metals was higher in all of the electrodialytic experiments as compared to the control experiment and shows that applying current to the sediment increases the heavy metal removal. Removal of all heavy metals increased with an increase in the current strength until 70 mA, but at 90 mA the removal was lower as compared to that at 70 mA. A similar decrease in heavy metal removal at high current strengths was observed in ref 24 when heavy metal contaminated wastewaters were electrodialytically treated. In the experiment at 90 mA, there were problems with foaming in the middle compartment and the stirring was stopped occasionally, for the foam to settle. This problem was not observed in the other experiments, which shows that higher current does not necessarily lead to higher removals but can cause operational problems. The heavy metal removal was probably not influenced when the stirring
was stopped, because the heavy metal removal is only slightly dependent on the stirring velocity (21). The lower removal could be explained by water splitting at the cation-exchange membrane CAT 2, which occurs at higher current strengths than at the anion-exchange membrane AN 1 (16). At the interface between CAT 2 and the solution, the concentration of electrolyte can approach zero as the current is increased. When the current is increased even further, it will be carried in part by ions resulting from dissociation of water at the cation-exchange membrane, which will decrease the removal efficiency of heavy metal ions, because H+ ions have higher mobility in the electric field (25, 26). The highest Cu removal was 50%, which was the lowest of the four metals. Zn was available for extraction in the control experiment to a larger extent than Cu, indicating that Zn is more available in the sediment than Cu, which was also seen in Figure 2. High removal of Cd was obtained both in the control experiment and in electrodialytic experiments, with the highest removal in the electrodialytic experiments, where up to 90% Cd was removed at 70 mA. In experiment 2 (air-dried sediment, 30 mA), 90% Cd (Figure 3) was also removed, which was higher than that for the wet sediment at 30 mA. In all of the experiments made with wet sediment, significant amounts of Cd were found in the liquid in the middle compartment after the experiments. As seen in Figure 2a, the Cd desorption was not significantly dependent on pH below pH 4. Thus, it is most likely not the differences in the final pH in the sediments that caused the lower Cd removal in the wet sediment as compared to the experiment with air-dried sediment. The amount of uncharged Cdchloro complexes in the middle compartment depends significantly more on the Cl concentration than the formation of Cu, Zn, or Pb chloro complexes (21). The added HCl concentration was higher in the electrodialytic experiments with wet sediment, which probably caused the lower Cd removal from the wet sediment. Higher Cd removal could be obtained when remediating wet sediment by longer remediation times, because the equilibrium between the different Cd chloro complexes is continuously shifted during the remediation when the ionic species are removed. As compared to the other metals, Cd seems to be the most available metal and easiest to remove. This is consistent with earlier findings for other sediments (7). The highest removal was obtained in the experiment with 70 mA, where 50% Cu, 85% Zn and Pb, and 90% Cd were removed. The removal in the control experiment was 5% Cu, 40% Zn, 20% Pb, and 52% Cd without current, which was similar to the extraction experiments with the air-dried sediment (Figure 2b) at pH 2.8, which was also the final pH of the sediment in the control experiment. This suggests that the heavy metal extraction without current might be more dependent on pH (as seen in Figure 2a) than if the sediment is wet or dry. Reference 27 also used HCl to extract metals from several sediments. It was found that the extractability with 1 M HCl (L/S 17-25) in oxidized sediments was generally >80% for Zn and Pb and >60% for Cd and Cu. The extractability of metals decreased in anoxic samples, especially for Cu due to the insolubility of sulfides. This suggests that the wet Haakonsvern sediment in this study was not especially sulfidic, because similar extraction results were seen in both the control experiment and the extraction experiment. The low extraction of Cu from the sediment indicates that Cu is strongly retained in the sediment, which could be by organic matter. The higher removal in the electrodialytic experiments as compared to the control experiment may be due to lower pH and degradation of organic matter, caused by the current and oxidation from the stirring process, respectively. The influence on the metal removal in electrodialytic remediation when increasing the L/S ratio of the sediment VOL. 39, NO. 8, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2909
FIGURE 6. Heavy metal removal from the three sediments at 70 mA, L/S 8, 14 days remediation time and air-dried sediment. FIGURE 5. Removal of metals in electrodialytic experiments with varying L/S ratios in the middle compartment; all experiments lasted 14 days and had a constant current of 70 mA. L/S 4 was made with wet Haakonsvern sediment, and L/S 8 and 12 were made with airdried Haakonsvern sediment. suspension in the middle compartment is shown in Figure 5. Increasing the L/S ratio of the suspension in the middle compartment resulted in higher removals than were seen in the experiments where the current strength increased. The removal generally increased when the L/S increased, contrary to what was observed in the L/S batch extraction experiments (Figure 2b). At L/S 12, the Cd removal reached almost 100%, but the Cu and Zn removal was lower as compared to that at L/S 8. Around 25% Cu and Zn was found in the liquid in the middle compartment at L/S 12 and the dissolved metals had not been removed, which could be due to uncharged metal-chloro complexes or removal competition between different ions in the solution, mainly H+. Cu and Zn were released from the sediment at lower pH than Pb and Cd (Figure 2a). In the experiment with L/S 12, the amount of H+ ions could have inhibited the metal ion removal. The pH in the sediment suspension in the electrodialytic experiments was initially low (
