Speciation and Transport of Heavy Metals and ... - ACS Publications

magnesium, copper, zinc, lead, nickel manganese, chromium, and iron during treatment with an electric field of soil from a chlor-alkali factory. As ex...
0 downloads 0 Views 88KB Size
Environ. Sci. Technol. 2003, 37, 177-181

Speciation and Transport of Heavy Metals and Macroelements during Electroremediation P A S C A L S U EÅ R , * , † , ‡ KATARINA GITYE,§ AND BERT ALLARD† Man - Technology - Environment Research Centre, O ¨ rebro University, 701 82 O ¨ rebro, Sweden, Swedish Geotechnical Institute, 581 93 Linko¨ping, Sweden, and Sydkraft Sakab AB, 692 85 Kumla, Sweden

Electroremediation makes treatment of contaminated clay soils possible. The external electrical field causes several transport processes and changes in soil chemistry. This study concerns the leachability and transport of calcium, magnesium, copper, zinc, lead, nickel manganese, chromium, and iron during treatment with an electric field of soil from a chlor-alkali factory. As expected, most elements were removed from the acidic part of the soil and accumulated in the zone where pH changed from acidic to alkaline. However, acidic leaching of the soil in this zone did not mobilize the elements. Lead formed both an anionic complex which electromigration transported toward the anode as well as a cationic lead fraction which moved toward the cathode. The anionic complex could be lead sulfate. Lead from both fractions was strongly attached to the soil after treatment. The low availability of metals and macroelements after electrokinetic remediation could make electroremediation, excavation, and deposition of the accumulation zone an alternative for the treatment of contaminated soils.

Introduction Many projects have shown the feasibility of removing contaminants from soil by electroremediation. This might prove very valuable for the remediation of contaminated clay soils, which are difficult to treat with other technologies. Successful lab-scale electroremediation has been reported for both clayey and sandy soils (1-4). Successful field-scale electroremediation has resulted in removal of copper, cadmium, lead, zinc, and arsenic (5) and trichloroethylene (6). Haus et al. transformed Cr(VI) to Cr(III), with some transport of chromium to the anode (7). The electric field that is applied to the soil causes movement of charged elements in the soil (electromigration). The electric field also causes movement of water (electroosmosis), usually toward the cathode. Uncharged contaminants in the soil solution move with the water. Unlike hydraulic transport, electrokinetic transport is not dependent on the pore size distribution or the presence of macropores in the soil. Electroremediation has mainly focused on metal contaminated soil, since electromigration in general has a higher impact than electroosmosis (1, 2). * Corresponding author phone: +46 13 201889; fax: +46 13 201909; e-mail: [email protected]. † O ¨ rebro University. ‡ Swedish Geotechnical Institute. § Sydkraft Sakab AB. 10.1021/es010226h CCC: $25.00 Published on Web 12/03/2002

 2003 American Chemical Society

Electrochemical reactions at the electrodes are commonly dominated by the dissociation of water. Hydrogen ions are produced at the anode and electromigrate into the soil toward the cathode. Hydroxide ions are produced at the cathode and electromigrate into the soil from that direction. Thus, pH is acidic near the anode, and alkaline near the cathode. The transition from acidic pH to alkaline pH is usually within a small region called the pH-shift. Metals are mobilized in the acidic zone and transported toward the cathode. At the pH-shift, metals accumulate. This is often attributed to precipitation of metal hydroxides (1, 2, 8). To evaluate electrokinetic treatment, it is desirable to know the form of the pollutants that remain in the treated soil. Knowledge of metal speciation also increases understanding of the electroremediation process. The form of metals in the soil can be assessed by selective leaching or sequential extraction (9, 10). Although there exist some uncertainties about the selectivity of the solutions and resorption during the extractions, selective leaching provides a practical method to assess the potential mobility of metals and gives an indication as to which soil fraction the metal is associated to. Ribeiro et al. report higher mobility for copper after electroremediation, analyzed by sequential extraction. They attribute this to accelerated weathering of soil material (11) and to the transfer of copper from amorphous oxides and organic matter to soluble and exchangeable forms (12). Kim and Kim (13) report removal of lead, copper, cadmium, and zinc from more accessible fractions of a tailing soil, while sulfide and residual fractions were affected less by electroremediation. Reddy et al. (14) report rapid migration of soluble and exchangeable chromium, nickel, and cadmium. The elements precipitated at high pH, possibly as hydroxides. Reddy et al.’s study was performed on spiked soil, which leads to a considerably more mobile contamination than is found in industrially polluted soils. In the present study electroremediation was applied to contaminated soil from a chlor-alkali factory site. The main contaminants in the soil are mercury, dioxins, and lead, while other metals are present at elevated concentrations. The transport of metals and macroelements during treatment and the speciation of the elements after treatment are discussed. The behavior of mercury is discussed elsewhere (15).

Experimental Section Soil. Several hundred kilograms of soil were taken at the site of a chloralkali factory. The soil was mixed at the factory and a subsample was used in this study. The soil was polluted with mercury, polychlorinated dibensofurans, lead, and several other heavy metals. Initial metal concentrations are shown in Table 1. At the laboratory, the soil was sieved, and the fraction smaller than 4 mm was used for the experiment. The soil was predominantly illitic clay with loss on ignition 2.5%, sulfur content up to 4 g/kg and pH 7-7.5 (1 g soil in 10 mL water) (16). Electrokinetic Experiments. Electrokinetic experiments were conducted in a 46 × 10 × 10 cm3 plastic box (Figure 1). Graphite plate electrodes were placed at the ends of the

FIGURE 1. The experimental setup. VOL. 37, NO. 1, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

177

TABLE 1: Approximate Mass-Balance of Several Elements in the Soil Zn

Pb

Cu

Ni

Fe

Cd

Ca

Mg

Mn

initial concn 608 375 174 47 42000 6 16121 8373 2286 (mg/kg) after 182 days 851 377 273 128 48000 4 7396 6824 2501 (mg/kg) recovered from 140 100 160 270 110 67 46 82 110 the soil (%)

TABLE 2: Metals Mobilized by Selective Leaching of the Sample at 4 cm from the Anode Compartment after Electrokinetic Treatmenta fraction Ca (mg/kg) Mn (mg/kg) Fe (g/kg) Ni (mg/kg)

box in electrode compartments, separated from the soil by a polypropene geotextile. 2 kg soil was placed in the middle compartment. The total length of the soil column was 27 cm. The solutions in the electrode compartments were stirred continuously. Each electrode compartment was connected to a water tank by a siphon to reduce soil desiccation. An electric field of 30 V was applied for 182 days. The solution in the electrode compartments and in the connecting tanks consisted of 0.01 M NaCl at the start. Solution was added regularly in order to replace water that was lost by evaporation. Final chloride concentrations of 0.08 M Cl at the anode and 0 M Cl at the cathode were obtained. The smell of chlorine gas was noticed near the anode. Analysis. The pH and voltage gradients in the soil were measured in-situ at regular intervals during the experiment. The current between the electrodes was measured as well. The initial soil pH was measured in a suspension of 1 g of soil and 10 mL of 18MΩ water. After 9, 40, and 79 days, soil samples were taken at the pH shift and on both sides of the pH shift. After 126 and 182 days, samples were taken over the whole soil column. The soil samples were taken by pushing a cut plastic syringe, 8 mm in diameter, into the soil, while maintaining the syringe plunger at the soil surface. Thus, the soil was not compacted during sampling. The samples were dried for 24 h at 40 °C and extracted by microwave assisted digestion in aqua regia (leachant E, see selective leaching). The total content of copper, zinc, lead, nickel, iron, calcium, magnesium, and chromium was determined by ICP-AES. Soil samples after 182 days were also subjected to selective leaching. The same elements without chromium and magnesium were determined. Solution samples were taken from the electrode compartments and the same metals were analyzed. Chloride was measured by titration. Selective Leaching. The dried soil samples were subjected to selective leaching according to a procedure adapted from Lifvergren (16). 1 g of dried soil was shaken with 20 mL of one of the following solutions: A. 1 M sodium chloride for 24 h. Weakly adsorbed and readily mobilized metals were recovered from the supernatant. B. 1 M sodium hydroxide for 7 d. The alkaline conditions mainly extracted metals associated to humic matter. C. 0.1 M hydrochloric acid for 5 h and selected samples for 5 days. Fraction A plus metals mobilized by acid conditions were extracted. These metals were mainly associated with carbonates and iron- and manganese oxides. D. 5 mL of hydrogen peroxide and 15 mL of concentrated nitric acid at 60 °C for 3 h. In addition to metals mobilized in fraction A, B, and C, the extraction oxidized organic matter and secondary sulfide precipitates from the soil. E. 1 g of dried soil was digested by open focused microwave extraction with 20 mL aqua regia. This procedure yielded the total content of metals (17). Centrifugation and analysis of the supernatants followed. New soil was taken for each extraction, so that each extraction step was independent of the steps before. Three replicates were made for each leachant on the soil before treatment. 178

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 1, 2003

Cu (mg/kg) Zn (mg/kg) Pb (mg/kg)

start 182 days start 182 days start 182 days start 182 days start 182 days start 182 days start 182 days

A

B

C

D

E

4150 63 1 0 0 0 0 13 0 4 0 0 0 35

139 0 0 0 0 0 0 7 32 8 32 0 20 0

6726 0 693 0 2.4 1 0 15 52 7 183 0 114 32

6919 128 1024 9 15 2 20 15 81 11 450 7 207 35

8060 365 1143 60 21 9 24 16 87 18 304 31 188 112

a Metal amounts as they were analyzed in the leachant, without correction for weaker extraction steps.

Sampling volumes from the soil after 182 days permitted only 1 soil sample for each leachant. In the figures presented in the next section, the leached amounts have been corrected so that they represent the fraction that was mobilized in excess of the weaker leachant that came before, since the diagrams give the visual impression of being cumulative. Thus, fraction C denotes the results from leachant C minus leachant A; fraction D the results from leachant D minus leachant C; fraction E the results from the total extraction minus leachant D. The result of leachant B is reported without correction, since only few metals (copper, lead, and nickel) are extracted by leachant B, and leachant B does not overlap with leachants C, D, and E. In Table 2, the amounts have not been adjusted.

Results and Discussion Time Series. The current through the soil was between 2 and 15 mA except for occasional short-lived peaks when measurements disturbed the zone of high resistance at the pHshift. The voltage gradient in the soil was directly proportional to pH. This can be explained by the domination of hydroxide ions and protons to the electric conductivity of the soil solution (1). A shift in pH developed within 9 days. Hydrogen ions produced at the anode moved into the soil, as did hydroxide ions produced at the cathode. Water formed where the hydrogen and hydroxide ions met. The conductivity of the soil solution is much decreased at this zone, as could be seen from an increase in current when the zone was disturbed by measurements. The pH-shift was between 16 and 20 cm from the anode geotextile after 40 days, around 20 cm after 79 days, and between 20 and 25 cm after 126 days (Figure 2). Almost all the soil was acid at the final sampling after 182 days. Thus, the pH-shift moved slowly toward the cathode. The shift from pH 3 to pH 10 occurred within approximately 5 cm. Calcium, magnesium, nickel, zinc, manganese, copper, and chromium were released from the soil in the region of low pH (Figure 2). This is probably due to the decreasing pH in the soil (18). The elements were transported toward the cathode, and thus were in cationic form. The elements accumulated where pH shifted from acidic to alkaline (Figure 2). Higher pH lowered the mobility of the cations and they associated with the soil. The zone of metal accumulation moved together with the pH-shift toward the cathode. When the pH-shift moved closer to the cathode, the elements moved as well. They remobilized with the successive acidification of the soil and precipitated closer to the cathode where pH continued to be high. The precise location of the accumulation zone in

FIGURE 2. Distribution of pH, calcium, zinc, iron, and lead during the electrokinetic experiment. Magnesium, manganese, nickel, copper, and chromium were very similar to zinc (see Figure 1 Supporting Information). [ start, O 40 days, × 79 days, 4 126 days, b 182 days. relation to the pH-shift differed between the metals. Calcium accumulated on the alkaline side of the pH-shift, copper on the acidic side. The order of the metals followed the order of hydrolysis constants for the elements, i.e., Ca > Mg >Ni ≈ Zn ≈ Mn > Cu > Cr (19, 20). High hydrolysis lead to remobilization at high pH, so that calcium moved closest to the cathode and chromium nearest the anode (Figure 2). An approximate mass-balance for the soil after 182 days was calculated (Table 1). Each soil sample after 182 days was assumed to represent the soil section around the sample. The limits of each section were halfway between the sampling points. Thus, each sample represented a section 6-7 cm long. This resulted in a coarse approximation of reality since the spatial distribution of the concentrations showed high variability near the maximum concentrations. Therefore, the mass balance has low precision. However, there are some aspects of the approximate mass-balance that deserve attention. The recovery from the soil of most elements was > 100% (Table 1). This shows that metals were not removed from the soil, only redistributed. This is not true for calcium. Around

half of the original calcium was recovered in the soil. As seen in Figure 2, calcium was the element that moved furthest toward the cathode. Calcium was also found in precipitates in the cathode compartment. These precipitates were not included in the mass-balance because it was difficult to ascertain the total amount of precipitate. The high calcium content of the precipitate in the cathode compartment showed that calcium has migrated into the cathode compartment and has been removed from the soil. Selective Leaching. The results from the selective leaching are shown in Figure 3. Before the electrokinetic treatment, 50% of the initial calcium content was extractable by leachant A, and the calcium was mobilized by the electric field. Some copper, zinc, and lead were extracted by leachant B, which is expected due to their high affinity for organic matter (21, 22). Leachant A and B extracted less than 2% of manganese, iron, and nickel (Table 2). Acid leaching (C) or acid and oxidizing leaching (D) did generally mobilize the metals in the start soil. Leachant C and D resemble the conditions near the anode, i.e., acidic and oxidizing. As expected from the selective leaching, the VOL. 37, NO. 1, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

179

FIGURE 3. Extraction of zinc, copper, calcium, iron, and lead by selective leaching after 182 days of electrokinetic treatment. Manganese and nickel were similar to zinc (see Figure 2 Supporting Information). metals were released from the soil near the anode during the electrokinetic treatment. After 182 days of electrokinetic treatment, leachant A extracted less than 4% of copper, zinc, manganese, and magnesium at all locations in the experiment. The exception was nickel: up to 80% of nickel near the anode was extracted with leachant A (Table 2). Thus, nickel near the anode is more available for transport by electromigration. However, nickel behaved similar to copper, zinc, and manganese as evidenced by the time series and by the leaching characteristics at other locations in the experiment. The metals near the cathode were expected to be hydroxide precipitates, since they were immobilized by the high pH. This is also proposed by several other publications (1, 2, 14). If this was the case, the metals should be released by acidic leaching (leachant C) (9, 16). However, leachant C extracted only up to 65% of copper, zinc, lead, nickel, calcium, and iron. It is possible that the extraction with leachant C for 5 h was too short to dissolve all hydroxides. The extraction was originally developed for soils that are not treated with an electric field. Since electromigration can transport ions through micropores in the soil, hydroxides could have precipitated in these micropores. Dissolution of hydroxides could then be controlled by the diffusion of ions in the micropores. This would make the extraction slower for hydroxides in micropores than for hydroxides that can be reached directly by the extracting solution. To test whether kinetics played a major role, the 2 soil samples closest to the cathode were leached for 5 days with leachant C. These samples were the most likely to contain hydroxides, since these locations had been most alkaline. The 5-day leaching did not extract more metals that the 5-h leaching. Either the 5-hour extraction mobilized all metals associated with hydroxides, or hydroxides were not leached even after 5 days. A surprisingly large amount of the metals (> 50%) was present in a form that was not leachable with acid within 5 days. The metals that were difficult to leach were transported by the electric field. They were not near the cathode at the start of the experiment. For example, the zinc concentration 180

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 1, 2003

throughout the soil before the experiment was 300 ppm. After treatment, 1170 ppm zinc was found near the cathode. 640 ppm of these 1170 ppm was extracted by leachant E only. New zinc was therefore present in fraction E. The elements that had been transported toward the cathode were difficult to mobilize despite the short time (less than 182 days) since their arrival. The low mobility of the metals could be explained by the role of soil micropores. Electrokinetic transport uses micropores more than does hydraulic transport (1, 2). Metals that were transported into micropores by electromigration, and that either adsorbed to the soil or precipitated as hydroxides, could be difficult to extract using agitation. These metals would of course also have low mobility under natural conditions. Calcium was only extracted by leachant D. This excludes calcium hydroxide or carbonate as the final form, since these would be extracted by leachant C. A possibility is the presence of calcium as calcium sulfate. The sulfur content of the soil was quite high, up to 4 g/kg, due to the industry’s use of elemental sulfur to produce sulfuric acid. Calcium sulfate is one of the few sulfates that has low solubility and may thus be limiting for the mobility of calcium. Calcium sulfate would be extracted by leachant D and not by leachant C, which is in agreement with the observations. Even near the anode the availability of the elements was low (Table 2). Typically over 60% of the remaining metals were found in fraction E. The selective leaching results near the anode can be compared with results from Ribeiro et al. (11) and Kim and Kim (13), since both these publications concern studies where the entire studied soil is acidic. Both Ribeiro et al. and Kim and Kim report that metals moved from the intermediate fractions to the fractions that were more easily extracted. Metals were also removed from the residual fraction (E) in both the present study and the study reported by Kim and Kim. However Ribeiro et al. found no removal of copper from the fraction corresponding to fraction E in the present study (“strongly bound” fraction), despite a treatment time of 125 days. Only copper and nickel were associated with low molecular weight organic material. 40-50% of the total copper and nickel was extracted by an alkaline solution (leachant B) near the anode. This is probably due to the high affinity of copper and nickel for organic molecules, which are extracted by leachant B (21, 22). The most available metals were found adjacent to the anode side of the pH-shift (19 cm, Figure 3). Leaching results were typically A 15%, B 0%, C 45%, D 80%. The elements were in the process of remobilization and transport toward the cathode. Iron and lead behaved somewhat differently from the other metals (Figure 2). The iron in the soil was transported slowly toward the cathode, but the process was not dominated by pH. No relation between iron transport and the position of the pH-shift was evident from the data. Possibly iron was transported as a cation in solution. An alternative explanation is electroosmotic or electrophoretic transport of ironcontaining colloids. Initially, iron was mobilized by leachants D and E (acid and oxidizing and total content, respectively, Figure 3). This mobilization occurred near the anode, since conditions were oxidizing and acidic. After 182 days, the iron was found in fraction E near the cathode. Immobilization may have occurred due to reduction or by the higher pH. The migrated iron was insoluble. Some lead moved toward the cathode, and this cationic lead behaved very similar to the elements described earlier (Figure 2). Most lead however moved toward the anode. This might be explained by an anionic lead complex, such as lead chloride or sulfate. Model calculations using Mineql+ (23)

show that 0.1 M chloride, which was the highest chloride concentration that was measured during the experiment, is not sufficient to cause the presence of a considerable amount of anionic lead chloride complexes. Less than 1% of lead would be present as PbCl3- or PbCl42-. Sulfate may also form complexes with lead. Analysis has shown up to 4% sulfur in the soil (16), and if even a tenth of the sulfur was oxidized, concentrations of sulfate may locally exceed 1 M. At a sulfate concentration of 0.1 M, 20-30% of lead may be in the form of Pb(SO4)22-. The lead sulfate complex has the highest stability constant of the metals, and lead is the only metal for which an anionic complex with two sulfates is reported (24), which may explain why the other metals showed no anionic behavior while lead did. Anions are scarce near the cathode, since they migrate toward the anode. Therefore lead was present as a cation near the cathode. The cationic lead was extracted by leachant C, D, and E, similar to the metals discussed above (Figure 3). The anionic lead was also difficult to extract. Half of the lead was extractable by oxidizing conditions (leachant D), while the remaining half was residual (Figure 3). Common to all the discussed elements was their strong attachment to the soil after the treatment with the electric field. Metals were removed from the acidic part of the soil. The sequential extraction showed that these metals were removed from all soil fractions. Most metals concentrated at the pH-shift. The formation of metal hydroxide is an unlikely explanation, in view of the low extraction by an acid solution. The low potential for remobilization of the metals shows that both mobilization and immobilization by electric fields could be feasible methods for the remediation of contaminated soil.

Acknowledgments The authors gratefully acknowledge the help of Hanna Ericsen at Sydkraft SAKAB with the chloride analysis. This work was financially supported by MISTRA, as a part of the program soil remediation in a cold climate (COLDREM) and the Knowledge foundation.

Supporting Information Available Distribution of magnesium, manganese, nickel, copper, and chromium and the sequential extraction of manganese and nickel, which are similar to metals shown in the text. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) Acar, Y. B.; Alshawabkeh, A. N. Environ. Sci. Technol. 1993, 27, 2638-2647.

(2) Probstein, R. F.; Hicks, R. E. Science 1993, 260, 498-503. (3) Hansen, H. K.; Ottosen, L. M.; Kliem, B. K. J. Chem. Technol. Biotechnol. 1997, 70, 67-73. (4) Li, Z.; Yu, J.-W.; Neretnieks, I. Environ. Sci. Technol. 1998, 32, 394-397. (5) Lageman, R. Environ. Sci. Technol. 1993, 27, 2648-2650. (6) Ho, S. V.; Athmer, C.; Sheridan, P. W.; Hughes, B. M.; Orth, R.; McKenzie, D.; Brodsky, P. H.; Shapiro, A. M.; Sivavec, T. M.; Salvo, J.; Schultz, D.; Landis, R.; Griffith, R.; Shoemaker, S. Environ. Sci. Technol. 1999, 33, 1092-1099. (7) Haus, R.; Zorn, R.; Aldenkortt, D. In 2nd symposium Heavy Metals in the Environment and Electromigration Applied to Soil Remediation; Technical University of Denmark: Denmark, 1999; pp 165-170. (8) Alshawabkeh, A. N.; Yeung, A. T.; Bricka, M. R. J. Environ. Eng. 1999, 125, 27-35. (9) Tessier, A.; Campbell, P. G. C.; Bisson, M. Anal. Chem. 1979, 51, 844-851. (10) Saouter, E.; Campell, P. G. C.; Ribeyre, F.; Boudou, A. Intl. J. Environ. Anal. Chem. 1993, vol. 54, 57-68. (11) Ribeiro, A.; Villumsen, A.; Re´fega, A.; Vieira e Silva, J.; BechNielsen, G. In 16th world congress of soil science; Montpellier, France, 1998. (12) Ribeiro, A. B.; Mexia, J. T. J. Haz. Mater. 1997, 56, 257-271. (13) Kim, S.-O.; Kim, K.-W. J. Haz. Mater. 2001, B85, 195-211. (14) Reddy, K. R.; Xu, C. Y.; Chinthamreddy, S. J. Haz. Mater. 2001, B84, 279-296. (15) Sue`r, P.; Allard, B. Accepted for publication in Water, Air, Soil Pollut. 2002. (16) Lifvergren, T. Ph.D. Thesis, O ¨ rebro Studies in Environmental Science 1, O ¨ rebro University: O ¨ rebro, Sweden, 2001. (17) Lifvergren, T.; Sue`r, P.; Wievegg, U. In 11th Annual International Conference on Heavy Metals in the Environment; Nriagu, J., Ed.; University of Michigan, School of Public Health, Ann Arbor, MI, USA (CD-ROM), 2000; p Contribution #1367. (18) Alloway, B. J., Ed. Heavy metals in soils; 1st ed.; Blackie and Son Ltd.: London, 1990. (19) Allard, B. In Trace elements in natural waters; Salbu, B., Steinnes, E., Eds.; CRC Press: 1995; Chapter 157, pp 151-176 . (20) Appelo, C. A. J.; Postma, D. Geochemistry, groundwater and pollution; A. A. Balkema: Rotterdam, 1996. (21) Asami, T.; Kubota, M.; Orikasa, K. Water, Air, Soil Pollut. 1995, 83, 187-194. (22) Morera, M. T.; Echeverria, J. C.; Mazkiaran, C.; Garrido, J. J. Environ. Pollut. 2001, 113, 135-144. (23) Schecher, W. D.; McAvoy, D. C. MINEQL+ A chemical equilibrium modeling system; 4.0 for Windows ed.; Environmental Research Software: Hallowell, 1998. (24) Lindsay, W. L. Chemical equilibria in soils; Wiley and Sons, Inc.: Canada, 1979.

Received for review September 6, 2001. Revised manuscript received October 14, 2002. Accepted October 14, 2002. ES010226H

VOL. 37, NO. 1, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

181