Environ. Sci. Technol. 2004, 38, 1383-1398
Retention of Trace Metals by Solidified/Stabilized Wastes: Assessment of Long-Term Metal Release R. BADREDDINE,† A.-N. HUMEZ,† U. MINGELGRIN,‡ A. BENCHARA,§ F . M E D U C I N , | A N D R . P R O S T * ,† Unite´ de Science du sol, INRA, Route de Saint Cyr, 78026 Versailles, France, Institute of Soil, Water and Environmental Sciences, Agricultural Research Organization, Volcani Center, P.O. Box 6, Bet-Dagan, 50250 Israel, Centre de Recherches pour l’Environnement, l’Energie et le De´chet, CREED, Limay, France, and Laboratoire de Physique et Me´canique des Milieux He´te´roge`nes, CNRS, EÄ cole Supe´rieure de Physique et de Chimie Industrielles, de la Ville de Paris, 10 rue Vauquelin 75231 Paris Cedex 5, France
Toxic elements found in wastes may have a negative impact on the environment, especially through the contamination of groundwater and plants. To reduce their mobility and availability, French regulations mandate the solidification and stabilization of toxic wastes. Many methods to stabilize and solidify wastes exist, among them the Ecofix process which employs low cost materials and consists of mixing wastes with lime, aluminum hydroxide, and silica. To evaluate the long-term behavior of solidified/ stabilized (S/S) samples, their alteration under saturated conditions was studied in a water extractor, a Soxhletlike device, used to follow the weathering of rocks. Kinetic measurements have shown that the release of Fe, Pb, Cd, Cr, and Cu was very slow, indicating a strong retention of these elements by the S/S materials prepared by the Ecofix process. To elucidate the mechanisms of retention of the trace metals, the mineral phases that existed in the S/S samples throughout and at the end of the extraction runs were studied by X-ray diffraction and by infrared and nuclear magnetic resonance spectroscopies. Scanning electron microscopic (SEM) examinations and electron microprobe analyses of the S/S samples were also performed at different stages of weathering. These observations revealed that assorted calcium silicate hydrates (C-S-H) were the predominant phases in the S/S preparations and that gradual alterations occurred in the structure of the investigated materials. The overall Ca/Si ratio of the C-S-H phases decreased as the enhanced alteration progressed. Although trace metals in oxide, hydroxide, and carbonate forms were found in the S/S materials, the bulk of the trace metals was incorporated in the matrix of the C-S-H phases. * Corresponding author phone: +33-1-30-83-32-50; fax: +33-130-83-32-59; e-mail:
[email protected]. † Unite ´ de Science du sol, INRA. ‡ Volcani Center. § CREED, Limay. | Laboratoire de Physique et Me ´ canique des Milieux He´te´roge`nes, Paris. 10.1021/es0209520 CCC: $27.50 Published on Web 01/21/2004
2004 American Chemical Society
Introduction To decrease the environmental impact of potentially toxic wastes, French regulations allow dumping of only ultimate residues, namely, wastes that cannot be reused or recycled. Those regulations mandate solidification and stabilization of the dumped wastes. Several processes have been developed to solidify and stabilize wastes, for example, vitrification and hydraulic binding. Solidification and stabilization processes generally employ simple, relatively inexpensive components and are cost competitive with other treatment options. A major potential risk associated with such disposal processes stems from their failure to ensure separation between the solidified/stabilizes (S/S) waste and the soil or subsoil matrix (1) and, therefore, to rule out eventual leakage of the contaminants into that matrix. Accordingly, the purpose of the present study was to analyze the long-term metalretaining performance of S/S materials prepared by the Ecofix hydraulic binding process (Experimental Section). The long-term mobility of toxic elements encased in a solidified/stabilized state is an important criterion in the assessment of the risk they pose to the environment. This mobility is estimated in the present study through exposure of the S/S wastes to an accelerated alteration process and measurement of the kinetics of release of the toxic elements. The methodology of the accelerated alteration was described by Humez et al. (2, 3) and consists of leaching particles of S/S waste for an extended period (up to 22 months) in a water extractor similar to a Soxhlet device. In the course of preparation of the S/S materials, a trace metal salt was mixed with fixed proportions of silica, lime, aluminum hydroxide, and water. In some samples, two trace metals were embedded in the same matrix. Leachates that were in contact with the S/S material were analyzed for various elements, and the residual solid phase was investigated to determine the mineral transformations and the physicochemical reactions that occurred during the accelerated alteration. As weathering proceeded, the evolution of the spatial distribution of the trace metals within the solid matrix and of the metals’ partition between various mineralogical phases was determined. Finally, the experimental results were used to test mechanistic hypotheses for the retention of trace metals by the S/S materials. The analytical methods used included X-ray diffraction (XRD), infrared and 29Si nuclear magnetic resonance spectroscopy. Changes with time in the structure and mineralogy of the S/S materials were followed with the aid of scanning electron microscopy (SEM) and the elemental distribution in these materials was obtained by electron microprobe (EDX) analysis. S/S materials prepared by the Ecofix process and loaded with trace metals retained the metals very strongly when exposed to the enhanced weathering procedure. Calcium silicate hydrates (C-S-H), and to a lesser extent gibbsite and calcite, were identified as the more prevalent components of the investigated S/S samples. It is asserted that the C-S-H phases were responsible for the retention of the bulk of trace metals, in line with previous findings (e.g., refs 4-7). Direct evidence that the stabilization of trace metals in the investigated S/S samples is mainly due to the metals’ incorporation into C-S-H phases is presented. 29Si NMR spectroscopy was successfully used for the investigation of C-S-H phases (e.g., refs 8 and 9). Moulin et al. (5) employed 29Si NMR and X-ray absorption (EXAFS) spectroscopy to demonstrate the presence of Si-O-Pb and Si-O-Zn bonds in C-S-H species and a similarly high affinity of Cu to C-S-H. Cocke (10) and Ziegler et al. (11) VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1383
TABLE 1. Initial Chemical Composition (% w/w) of the Investigated S/S Samples sample component
1
2
3
4
5
6
SiO2 CaO Al(OH)3 H2O FeCl3 Fe2O3 PbCl2 CrCl3 CuCl2 CdCl2 CaCl2 total
25.75 25.75 10.90 29.80 7.50
23.90 23.80 10.10 33.75
26.15 25.85 11.00 28.60
25.80 25.70 10.90 29.70
25.60 25.70 10.80 29.20
23.20 23.10 9.70 27.40 6.70 7.40
1.90 4.70 3.30 99.70
6.20 99.65
1.30 6.80 99.70
2.30 2.80 99.60
4.90 99.50
99.80
have also reported fixation of Pb and Zn, respectively, by C-S-H. Faucon and Adenot (12) used 57Fe Mo¨ssbauer spectroscopy to demonstrate the substitution of Ca2+ by Fe3+ in C-S-H phases. Pomie´s et al. (6, 7) employed XRD, EXAFS, and NMR spectroscopy to study Cd/Ca silicate hydrates that were prepared either by coprecipitation or by exchanging Ca in the C-S-H phases with Cd. They concluded that in the Cd-exchanged C-S-H, Cd and Ca are fully exchangeable as long as the Cd constituted (w/w) less than 30% of the total Ca + Cd content. The Ca to Cd exchange did not induce significant changes in the structure of the C-S-H. The EXAFS runs indicated that following the exchange, Cd was homogeneously distributed throughout the native Ca sites. Cd/Ca silicate hydrates prepared by coprecipitation were poorly crystallized, but the Cd environment seemed to be similar in the coprecipitated and in the Cd-exchanged C-S-H systems (6, 7). The overall objective of this study was to improve our understanding of the long-term behavior of a family of S/S systems having similar chemical and mineralogical compositions. The effect of variations in the proportions of the raw materials from which the S/S samples were prepared on the samples’ behavior was probed.
Experimental Section Solidified/stabilized samples were prepared according to the Ecofix process (Sarp-Industries, Limay, France) as follows. Chloride salts of metals were mixed with fixed proportions of silica, lime, aluminum hydroxide (gibbsite), and water (Table 1). The particle size of all solid components was
