Environ. Sci. Technol. 2003, 37, 4269-4273
Use of Copper Shavings To Remove Mercury from Contaminated Groundwater or Wastewater by Amalgamation PETRA HUTTENLOCH, KARL ERNST ROEHL,* AND KURT CZURDA Department of Applied Geology, University of Karlsruhe, Kaiserstrasse 12, D-76128 Karlsruhe, Germany
The efficacy of copper shavings (Cu0) for the removal of Hg2+ from aqueous solution by amalgamation is demonstrated. Two kinds of copper shavings were investigated: (a) chemically processed shavings (Fluka) and (b) recycled shavings from scrap metal. Batch sorption experiments yielded very high retardation coefficients of 28 850-82 830 for the concentration range studied (1-10 000 µg/L Hg2+ dissolved in distilled water or in a 0.01 M CaCl2 matrix solution). Sorption data were well-described by the Freundlich isotherm equation. Kinetic batch sorption experiments showed that 96-98% of Hg2+ was removed within 2 h. Column experiments were performed with a mercury solution containing 1000 µg/L Hg in a 0.01 M CaCl2 matrix with a flow rate of 0.5 m/d. No mercury breakthrough (c/c0 ) 0.5) could be detected after more than 2300 percolated pore volumes, and the high retardation coefficients determined in the batch studies could be confirmed. Copper was released from the shavings due to the amalgamation process and to copper corrosion by oxygen, resulting in concentrations of mobilized copper of 0.2-0.6 mg/L. Due to their high efficiency in removing Hg2+ from aqueous solution, the use of copper shavings for the removal of mercury from contaminated water is suggested, employing a sequential system of mercury amalgamation followed by the removal of mobilized copper by an ion exchanger such as zeolites. Possible applications could be in environmental technologies such as wastewater treatment or permeable reactive barriers for in situ groundwater remediation.
Introduction Due to its widespread application in such diverse fields as, for example, chloralkali synthesis, wood pulping industry, technical instruments (thermometers, barometers), electrical equipment (batteries), dentistry, paints, and military applications, mercury is one of the most frequent heavy metals in polluted soils of industrial areas and hazardous waste sites (1-3). Mercury contamination from gold mining and from the processing of mercury-rich ores are major environmental problems (4). Although mercury mobility in soils is generally assessed to be low (1, 5), depending on the biogeochemical subsurface conditions mercury may be released from contaminated soils with the risk of groundwater contamination (6, 7). In water, Hg2+ is the predominant form. Mercury is * Corresponding author e-mail:
[email protected]; phone: +49 (0) 721 608 7612; fax: +49 (0) 721 606 279. 10.1021/es020237q CCC: $25.00 Published on Web 08/07/2003
2003 American Chemical Society
known to be highly toxic for aquatic life and human beings even at low concentrations. Development and evaluation of new materials for the removal of mercury and mercury compounds from contaminated water still remains in the focus of innovative applications in water treatment and groundwater remediation technologies, including in situ techniques such as permeable reactive barriers (8, 9). Approaches for the cleanup of mercury-contaminated soils and sediments are discussed elsewhere (3, 10). Several sorbents for mercury removal from aqueous solution have been investigated such as zeolites (11), clay minerals (12), cerhydroxide (13), ion-exchange resins (14, 15), dipotassium salt of 1,3-benzene diamidoethanethiol (16), activated carbon (17, 18), peat (19), chitosan (20), waste tire rubber (21-23), and organic waste materials such as rice or coconut husks (24, 25). The main sorption mechanisms of the materials mentioned are ion exchange, adsorption, or surface complexation on functional groups. A different retention mechanism investigated recently is the amalgamation of Hg(II) on mossy tin filters (26). A new approach discussed in the present paper is the use of elemental copper (Cu0) as a sorptive medium to remove Hg2+ ions from aqueous solution. The reaction mechanism is based on a amalgamation process. Less precious metals (in this case Cu0) are able to build a stable alloy with more precious metals (Hg). In the case of Hg, the alloys are called “amalgam”. Hg2+ is reduced to Hg0 by Cu0 forming a Hg-Cu amalgam, and Cu2+ is released to the corresponding amount of Hg bound to the elemental copper:
Hg2+ + Cu f Hg + Cu2+
(1)
Hg + Cu f CuHgAM
(2)
For reaction 1, the Gibbs energy can be calculated using the standard potentials from the electrochemical series (Cu f Cu2+ + 2e, E0 ) -0.3402 V; Hg2+ + 2e f Hg, E0 ) +0.851 V; resulting in ∆E ) 0.5108 V):
∆G ) -zF∆E ) -2 × 96487 C/mol × 0.5108 V ) -98.57 kJ/mol (3) with ∆G ) Gibbs energy, z ) valence, F ) Faraday constant, and ∆E ) standard potential difference. The resulting, clearly negative Gibbs energy shows that the mercury reduction as described in eq 1 is the favored reaction triggering the amalgamation. In the present study, elemental copper was used in form of copper shavings that can also be obtained as a recycling product (scrap metal). Copper shavings have a good availability and are easy to employ. The objective of the present paper is to demonstrate the efficacy of the copper shavings to remove Hg2+ ions from contaminated groundwater or wastewater.
Materials and Methods Elemental Copper (Cu0). Two kinds of copper shavings have been selected for this study: (a) chemically processed Cu0 shavings from Fluka and (b) recycled Cu0 shavings obtained from a local scrap yard. The Fluka copper shavings have a high purity of 99.98%. The recycled copper shavings consist of simple electrolytic copper with a typical purity of >99.9%. Other metals are present only as trace impurity. Before application in laboratory tests, the recycled Cu0 shavings have been cleaned with dichloromethane. The particle size of the shavings was between 2 and 15 mm (Figure 1). Specific VOL. 37, NO. 18, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4269
FIGURE 1. (a) Commercially available (Fluka) copper shavings; (b) recycled copper shavings from a local scrap yard (length of 1 box ) 0.5 cm). surface of the particles was 1 × 10-4 m/s). A constant flow rate of 0.5 4270
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 18, 2003
FIGURE 2. Sorption isotherms of Hg on recycled and commercial copper shavings in deionized water (pure Hg nitrate solution) and in 0.01 M CaCl2 matrix solution, respectively. m/d was maintained by a peristaltic pump. The pore volume of the materials was determined by tracer tests with chloride. All experiments were carried out in duplicates. Analytics. Graphite furnace atomic absorption spectrometry (AAS) coupled with a flow-injection analytic system FIAS 400 from Perkin-Elmer was used for the analysis of Hg, detected as total Hg. Mercury speciation was not investigated. For the measurements, NaBH4 was used as reducing agent combined with standard procedure techniques. The Hg detection limit was 1 µg/L. A detailed description of Hg analysis is given elsewhere (27). Copper was measured using a flame AAS 3030B from Perkin-Elmer. All samples were stabilized with HNO3. Chloride used as a tracer in the column experiments was analyzed by a standard photometrical method.
Results and Discussion Sorption Isotherms. To evaluate the sorption behavior (or dimension of amalgamation) of the copper shavings, sorption isotherms have been prepared that could be described best by the nonlinear Freundlich isotherm equation (Figure 2). The Freundlich equation was applied to the experimental results in the linearized form (28):
log cs ) log KF + N log cw
(4)
where cs is the amount of the solute sorbed per unit mass of the sorbent (µmol/g), cw is the equilibrium solute concentration (µmol/L), and KF (µmol1-N LN g-1) and N (dimensionless) are the Freundlich equation parameters. The Freundlich parameters are derived from the intercept and the slope, respectively, of the line, which is obtained by plotting log cs versus log cw. For N ) 1, a linear isotherm results with KF becoming the distribution coefficient Kd. The retardation, representing the sum of the sorption processes that leads to the immobilization of a contaminant in a porous medium such as soil or any other permeable
TABLE 1. Freundlich Parameters for Hg (Pure Mercury Nitrate Solution) and Hg Dissolved in 0.01 M CaCl2 Matrix Solution on Recycled and Commercial Copper Shavings sample/stock soln
KF (µmol1-NF -1 LN g F]
N (-)
R (-)
concn range (µmol/L)
Copper Shavings (Recycled) Hg in deionized water Hg in CaCl2 matrix soln
30.57
0.92
74 720-42 960
0.005-0.03
12.48
0.69
82 830-41 400
0.005-0.09
Copper Shavings (Commercial) Hg in deionized water Hg in CaCl2 matrix soln
18.07
0.95
32 440-29 310
0.01-0.07
16.22
0.93
32 260-28 850
0.01-0.12
reactive material, can be expressed by the dimensionless retardation coefficient R. In a saturated porous medium R can be calculated from the relation between sorbed and aqueous concentrations of the contaminant (28):
R)1+
Fd ∂f(c) n ∂c
(5)
where Fd is the bulk density (g/cm3) and n is the porosity (dimensionless) of the reactive material. The relationship between sorbed and aqueous concentrations of the contaminant can be described by f(c) representing a linear or nonlinear sorption isotherm equation. The resulting Freundlich parameters and retardation coefficients for the batch experiments are compiled in Table 1. For all isotherms, the correlation coefficients were better than 0.97. Both kinds of copper shavings show a very high sorptivity for Hg resulting in high retardation coefficients R between 28 850 and 82 830. Sorption from the Hg solution increased relatively with decreasing Hg concentration due to the nonlinearity of the Freundlich equation (eq 4). For low initial Hg concentrations, no data could be obtained for the sorption isotherms since the equilibrium solution concentrations were below the detection limit of 1 µg/L because of the strong sorptivity of the copper shavings. The sorption isotherms and the resulting retardation coefficients R showed that the addition of CaCl2 to the Hg solutions had little influence on the amalgamation processes. The recycled copper shavings showed a somewhat higher affinity for Hg than the commercial ones. This can be referred to a different composition of the copper shavings (e.g., as a result of impurities of other metals). All further experiments were performed with recycled copper shavings because of their higher sorptivity. Sorption Kinetics. The batch experiments conducted with different contact times showed that the reaction kinetics of the amalgamation process appears to be very fast. After 2 h of contact time, 96-98% of Hg was removed from the aqueous phase and fixed on the copper shavings. Figure 3 shows the experimental results for the amalgamation process of Hg on the recycled copper shavings. Equilibrium was reached within 6 h in all cases (different concentrations and matrix solutions). The initial sorption rate was notably higher for increasing solute concentrations. An explanation for this behavior could be the increasing diffusion rate of Hg into the surface of the copper shavings with increasing Hg concentration. Stability Tests. Under normal conditions Cu0 is not soluble in water, but oxygen and oxidizing acids such as HNO3 induce corrosion of Cu0. Batch experiments conducted with untreated copper shavings in 0.01 M CaCl2 solution showed that approximately 0.3 mg/L Cu2+ was released into equilibrium solution. The addition of 1000 µg/L Hg to the CaCl2
FIGURE 3. Kinetics of the sorption of Hg on recycled copper shavings for three different solute concentrations. solution resulted in a significantly lower Cu release of up to 0.07 mg/L Cu2+ in the batch solution. From the amount of Hg addition to the system, it was expected for stoichiometric reasons that 0.316 mg/L Cu2+ should be released into the equilibrium solution. Since the copper release was much lower than expected, the Cu2+ ions were obviously retained in the metallic Cu-Hg amalgam structure. A possible explanation could be a protecting layer built during the amalgamation process with Hg (29). A reduction and subsequent immobilization of released Cu2+ by impurities in the elemental copper cannot explain the low copper release since the copper shavings contain only very low amounts of metallic impurities. By further addition of HNO3 (pH 3) to the batch solutions, approximately 35 mg/L Cu2+ was released to the equilibrium solutions. To test the stability of the Hg fixation, the Cu-Hg amalgam built during the sorption process was subsequently washed with distilled water and a HNO3 solution at pH 3. In both cases no remobilized Hg could be detected in the washing solutions. The stability of the Cu-Hg amalgam against methylation has not been tested. In light of the energy balance given in eq 3, the availability of Hg for microbial methylation processes might be low and probably only significant at high temperatures. Column Studies. Column tests were performed to determine the breakthrough curves for Hg in the given system. The breakthrough of a contaminant is in first approximation defined as the point where half the concentration of the input solution is detected in the column eluate (c/c0 ) 0.5). The dimensionless retardation coefficient R (eq 5) is defined as the number of pore volumes percolated through the column at the breakthrough point. In the performed column experiments employing a Hg solution of 1000 µg/L Hg in a 0.01 M CaCl2 matrix solution, no breakthrough of Hg could be observed. The first set of experiments conducted in a mixture of quartz sand and 10 wt % recycled copper shavings was terminated after percolation of 2300 pore volumes; the second set conducted in pure recycled copper shavings was terminated after percolation of 900 pore volumes, respectively. As expected, the resulting high retardation coefficients of R >2300 and R >900, respectively, verify the tendency of the sorption behavior determined by sorption isotherms (Table 1). In the eluate of the experiments conducted in a mixture of quartz sand and 10 wt % recycled copper shavings,1-4 µg/L Hg was detected in individual eluate samples after percolation of 400 pore volumes. This could be referred to as partly incomplete amalgamation. Eluate Hg concentrations remained that low or below the detection limit of 1 µg/L throughout the experiment, proving the high efficiency of the amalgamation process. In the second column test series conducted with 100% recycled copper shavings, no Hg was measured in the eluate until termination of the experiments. VOL. 37, NO. 18, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4271
All column eluate samples had a pH between 6 and 6.2, matching the pH of the initial mercury solutions and the pH of the sample in water (6.0-6.2 in both cases). The amalgamation process did not cause any changes of the pH conditions in the system. As expected for stoichiometric reasons, a continuous copper release should be observed in the column experiments. While during the percolation of the first 350 pore volumes a concentration of approximately 0.6 mg/L of dissolved Cu could be detected in the eluate samples of both column experiment series, the amount of dissolved Cudecreased to approximately 0.2 mg/L with ongoing experiment time. This could be explained by an increasing amalgamation of the surface of the copper shavings resulting in a protective amalgam layer retaining the Cu ions in the metallic Cu0 structure (29). In addition, copper was retained in the system by precipitation of base copper salts such as Cu(OH)2, resulting partly in a blue coating of the sand grains and the copper shavings. The blue copper salts could be dissolved in diluted HCl (9%). During experiment duration, no negative effect on the hydraulic permeability due to copper salt formation was observed. As described above, copper was released from the copper shavings during the amalgamation process in the form of Cu2+ ions. Depending on the objective of the water treatment, it may be necessary to remove the Cu2+ ions in a second step by an additional sorption process. In case of a simple water treatment, this second step could be implemented at the outflow of a mercury treatment unit by adding a copper treatment unit into the treated water stream. In case of passive in situ remediation technologies, this could be carried out in a funnel-and-gate system with sequential gates (30), consisting of reactors with two compartments where in the first compartment mercury and in the second compartment copper are removed from the groundwater. We suggest to employ an ion-exchange mechanism on a reactive material such as zeolites for the removal of the mobilized copper. Cation-exchange capacities (CEC) of crystalline zeolite minerals can reach values as high as 300-400 mequiv/100 g (31). Natural zeolites have a good availability at reasonable prices and can be used in granular form to ensure sufficient hydraulic permeability. Sorption behavior of natural zeolites toward heavy metals has been widely studied (e.g., refs 32-34) and relies on their cationexchange properties. To evaluate the applicability of natural zeolites to sorb copper ions, copper sorption behavior of a clinoptilolite-rich zeolitic tuff from Romania (clinoptilolite content >90%) was investigated. The CEC of the tuff was 145 mequiv/100 g; by transformation into sodium zeolite form, the CEC could be increased up to 180 mequiv/100 g. In line with results reported in the literature (33, 34), we found that especially the sodium clinoptilolite is characterized by a good sorptive and kinetic behavior, respectively, toward Cu2+ ions. Although only tested in laboratory scale, the removal of mercury from water streams by amalgamation to copper shavings appears to be a promising approach because of the high Hg2+ retardation coefficients and fast reaction kinetics. The performance of the readily available copper metal obtained from metal recycling facility as well as its easy handling make Cu0 shavings an efficient sorptive material for Hg2+ with application in wastewater treatment or permeable reactive barriers. Recovery of mercury from the amalgam complex can be achieved by distillation. More research is necessary to extend the scope of this Hg2+ sorption methodology to other Hg species (e.g., amalgamation of Hg0 or organic Hg compounds). Moreover, while focusing on chemical aspects of the treatment in the present paper, practical aspects of implementation, such as feasibility for different water streams (river or lake water, 4272
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 18, 2003
groundwater), designing and dimensioning of treatment plants, and cost-effectiveness and manageability of such plants have yet to be investigated (e.g., by means of a pilot plant).
Acknowledgments This research was supported by a grant from the Graduate College “Ecological Water Resources Management” of DFG (Deutsche Forschungsgemeinschaft). We thank Oliver Huttenloch (Max Planck Institute, Dortmund, Germany) for helpful comments and discussion.
Literature Cited (1) Merian, E. Metals and Their Compounds in the Environment; VCH: Weinheim, Germany, 1991; p 1438. (2) Evanko, C. R.; Dzombak, D. A. Remediation of MetalsContaminated Soils and Groundwater; GWRTAC Technology Evaluation Report TE-97-01; GWRTAC: Pittsburgh, 1997; p 53. (3) Ebinghaus, R.; Turner, R. R.; de Lacerda, L. D.; Vasiliev, O.; Salomons, W. Mercury Contaminated Soils; Springer: Berlin and Heidelberg, 1999; p 538. (4) De Lacerda, L. D.; Salomons, W. Mercury from Gold and Silver Mining: A Chemical Time Bomb?; Springer: Berlin and Heidelberg, 1998; p 300. (5) Nriagu, J. O. The Biogeochemistry of Mercury in the Environment; Elsevier: Amsterdam, 1979; p 696. (6) Yin, Y.; Allen, H. E.; Huang, C. P.; Sparks, D. L.; Sanders, P. F. Environ. Sci. Technol. 1997, 31, 496-503. (7) Arnold, C.; Schmid, B.; Munz, C.; Oehler, D. In Contaminated Soil 2000; Proceedings of the 7th International FZK/TNO Conference on Contaminated Soil, September 2000, Leipzig, Germany; Thomas Telford Publishers: London, 2000; Vol. 1, pp 92-98. (8) Gavaskar, A. R.; Gupta, N.; Sass, B. M.; Janosy, R. J.; O’Sullivan, D. Permeable Barriers for Groundwater Remediation; Batelle Press: Columbus, OH, 1998; p 176. (9) Blowes, D. W.; Ptacek, C. J.; Benner, S. G.; McRae, C. W. T.; Bennett, T. A.; Puls, R. W. J. Contam. Hydrol. 2000, 45, 123137. (10) Wilken, R.-D.; Hempel, M.; Richter-Politz, I. Proceedings of the 10th International Conference on Heavy Metals in the Environment, Hamburg, September 1995; CEP: Edinburgh, 1995; Vol. 2, pp 42-51. (11) Misaelides, P.; Godelitsas, A. Toxicol. Environ. Chem. 1995, 51, 21-29. (12) Viraraghavan, T.; Kapoor, A. Appl. Clay Sci. 1994, 9, 31-49. (13) Mishra, S. P.; Singh, V. K. J. Radioanal. Nucl. Chem. 1999, 241, 145-149. (14) Ferreira, L. M.; De Carvalho, J. M. R. Environ. Technol. 1997, 18, 433-440. (15) Chiarle, S.; Ratto, M.; Rovatti, M. Water Res. 2000, 34, 29712978. (16) Matlock, M. M.; Howerton, B. S.; Van Aelstyn, M. A.; Nordstrom, F. L.; Atwood, D. A. Environ. Sci. Technol. 2002, 36, 16361639. (17) Krishnan, S. V.; Gullett, B. K.; Jozewicz, W. Environ. Sci. Technol. 1994, 28, 1506-1512. (18) Mohan, D.; Gupta, V. K.; Srivastava, S. K.; Chander, S. Colloids Surf. A 2001, 177, 169-181. (19) Coupal, B.; Lalancette, J.-M. Water Res. 1976, 10, 1071-1076. (20) Coughlin, R. W.; Deshaies, M. R.; Davis, E. M. Environ. Prog. 1990, 9, 35-39. (21) Knocke, W. R.; Hemphill, L. H. Water Res. 1981, 15, 275-282. (22) Rowley, A. G.; Husband, F. M.; Cunningham, A. B. Water Res. 1984, 18, 981-984. (23) Gunasekara, A. S.; Donovan, J. A.; Xing, B. Chemosphere 2000, 41, 1155-1160. (24) Khalid, N.; Ahmad, S.; Kiani, S. N.; Ahmed, J. Sep. Sci. Technol. 1999, 34, 3139-3153. (25) Sreedhar, M. K.; Madhukumar, A.; Anirudhan, T. S. Ind. J. Eng. Mater. Sci. 1999, 6, 279-285. (26) Biester, H.; Schuhmacher, P.; Mu ¨ ller, G. Water Res. 2000, 34, 2031-2036. (27) Sinemus, H. W.; Stabel, H. H.; Kleiner, J.; Radziuk, B. GIT LaborFachzeitschrift 1997, 475-479.
(28) Fetter, C. W. Contaminant Hydrology; Prentice Hall: Upper Saddle River, NJ, 1999; p 555. (29) Mu ¨ ller, F.; Reuther, H. Z. Elektrochem. 1948, 52, 44-45. (30) Starr, R. C.; Cherry, J. A. Ground Water 1994, 32, 465-476. (31) Flanigen, E. M.; Mumpton, F. A. In Mineralogy and Geology of Natural Zeolites; Mumpton, F. A., Ed.; Reviews in Mineralogy 4; Mineralogical Scoiety of America: Washington, DC, 1981; pp 165-175. (32) Zamzow, M. J.; Murphy, J. E. Sep. Sci. Technol. 1992, 27, 19691984.
(33) Langella, A.; Pansini, M.; Cappelletti, P.; Gennaro, B.; Gennaro, M.; Colella, C. Microporous Mesoporous Mater. 2000, 37, 337343. (34) Panayotova, M. I. Waste Manage. 2001, 21, 671-676.
Received for review December 2, 2002. Revised manuscript received May 1, 2003. Accepted May 12, 2003. ES020237Q
VOL. 37, NO. 18, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4273
