Environ. Sci. Technol. 2002, 36, 4880-4885
Flow-through Sequential Extraction Approach Developed from a Batch Extraction Method H. KUROSAKI,† S . M . L O Y L A N D A S B U R Y , †,§ J . D . N A V R A T I L , ‡ A N D S . B . C L A R K * ,† Department of Chemistry and Center for Multiphase Environmental Research, Washington State University, Pullman, Washington 99164-4630, and Department of Environmental Engineering and Science, Clemson University, Clemson Research Park, Anderson, South Carolina 29625-6510
Sequential extractions of contaminants from a soil or sediment have been shown to be cost-effective contaminated site assessment tools that provide information on contaminant partitioning within an environmental matrix. Such information is necessary for defining remediation alternatives and mitigation strategies. The typical sequential extraction approach involves a batch method, and known limitations include the possibility of contaminant readsorption to the remaining soil or sediment. In this work, a flow-through reactor was constructed and tested for application in a sequential extraction scheme. The sequential extraction scheme used was one developed for actinidecontaminated materials. The flow-through approach gave partitioning results that were similar to the batch method for uranium. We also monitored the extraction of stable Ca, K, Fe, Al, Zr, and Sc and obtained partitioning results generally similar to those observed with the batch extraction, except for Ca. Our results indicate readsorption of Ca when using a batch approach is small but significant and is eliminated with our new flow-through method. A limitation of the flow-through method is the possibility of underextraction of certain phases and higher analytical uncertainties. These uncertainties are more difficult to minimize, as compared to the uncertainty obtained with a batch approach.
Introduction The development of nuclear technologies in the 20th century has resulted in large quantities of radioactive waste and radiologically contaminated sites around the world (1-4). Within the United States alone, the amount of contaminated environmental materials is estimated to be ∼7.5 × 106 m3 for soils and sediments, and 1.8 × 109 m3 for groundwater (5). The majority of the contamination has resulted from accidental and intentional releases by nuclear processing plants (1). These contaminants are primarily fission products, such as 137Cs, 90Sr, 60Co, and isotopes of U, Np, Pu, Am, and Cm (6). Remediation of this material will be very costly, and in many cases, complete cleanup may not be possible. Thus, * Corresponding author phone: 509-335-1411; fax: 509-335-8867; e-mail:
[email protected]. † Washington State University. ‡ Clemson University. § Current address: Amersham Biosciences Corp., 800 Centennial Ave., P.O. Box 1327, Piscataway, NJ 08855-1327. 4880
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 22, 2002
cost-effective site assessment tools and alternatives to remediation are desirable (7-10). Soils and sediments are typically heterogeneous mixtures of sands, silts, and clay minerals, along with natural organic matter (11). In addition, these solid phases may be coated with organics or amorphous oxides (12). Thus, the soil or sediment matrix provides a variety of surfaces to which contaminants, such as actinides and fission products, may sorb. As a result, contaminants may be sorbed to several coexisting phases (e.g., ref 13). These different physical forms of the contaminants can give widely varying apparent solubilities. For natural systems, we and others have defined this apparent solubility as “environmental availability” (e.g., refs 14-17). It has been demonstrated that batch sequential extractions can be useful tools for determining the environmental availability of fission products, actinides, and other contaminants associated with soils and sediments (15-20). In addition to providing information on environmental availability, sequential extractions more specifically define the mode(s) of contaminant partitioning to soils and sediments (e.g., refs 14-28). Typically, sequential extractions involve the successive leaching of a contaminated soil or sediment with extractant solutions that are chemically increasingly aggressive. Leachings are usually completed by batch extraction, which involves thorough mixing between the entire sample and the extractant at a defined solution/ sediment ratio and for a defined time period and reaction temperature. The extractant is then separated from the solid phase by centrifugation or filtration, the leachate and any rinse solutions are analyzed for the contaminant as well as other parameters, and the remaining solid phase is used in the subsequent extraction steps. Ideally, each treatment step is intended to target a specific mode of partitioning or geochemical phase; however, there are limitations. These include the possibility of over- or underextraction of a geochemical phase and the associated contaminants (e.g., refs 29 and 30), as well as the possibility of readsorption of extracted contaminants to the remaining solid phases (e.g., refs 31-35). Because of these limitations, the phases targeted in the steps of a given sequential extraction method are operationally defined, and sequential extractions must be considered indirect methods for determining contaminant partitioning. Although batch approaches have been commonly used with sequential extractions, flow-through designs are also possible. For example, Schwark et al. (36) developed a flowthrough cell that allowed extraction of a whole core plug. This approach maintains the pore system of sediment core, and in this case, was used to define fractions of oil phases present in the core. A potential disadvantage is that a core system provides the possibility of interaction sites between any extracted species and the remaining solid phase material as the extractant solution and extracted species flow through the remainder of the soil core. Such chromatographic processes may be undesirable in some applications. A flow system has recently been described for study of the partitioning of metal ions in soils and sediments (37). Here, the authors combined a stirred cell with a flow-through approach to achieve a rapid, less tedious method to obtain metal ion partitioning information for a soil that was a standard reference material. One possible way to minimize contaminant readsorption is to use a flow-through sequential extraction reactor rather than a batch method. With appropriate design, a flow-through system will allow extraction of the contaminant and the associated solid phase followed by extractant flow out of the 10.1021/es020653a CCC: $22.00
2002 American Chemical Society Published on Web 09/25/2002
TABLE 1. Batch Method Sequential Extraction Procedure That Was Adapted for a Flow-through Application as Described Hereina target phase
reagent
solid/solution
reaction time, h
temp
exchangeable acid extractable reducible oxdizable residual
0.4 M MgCl2, pH ) 4.5 0.5 M NH4Ac, pH ) 5.0 0.1 M NH2OH-HCl in 25% HAc, 0.2 pH ) 2.0 30% H2O2 in 0.02 M HNO3 LiBO3/HF fusion
15:1 15:1 15:1 15:1
1 2 6 4
room room 70 °C 50 °C 1000 °C
a
The batch method was originally reported by Smith (27), and used by Loyland et al. (16, 17). See text for details on the flow-through method.
system and away from the remaining solid phase, thereby minimizing the possibility of readsorption. Recently, the National Institute of Standards and Technology (NIST) invested significant effort to define a batch sequential extraction method for determining the partitioning of the actinides to contaminated soils and sediments (27, 28). This method, summarized in Table 1, is designed to optimize the dissolution of the defined sediment phases with extra attention placed on the unique chemistry of the actinide contaminants. It also allows discrimination between highly refractory actinide oxides and actinides associated with the sediment in nonrefractory forms. To our knowledge, a flowthrough approach for this method has not been tested. In this work, we have developed a flow-through sequential extraction method based on the NIST batch method. We have applied this method to two samples collected from the proximities around U.S. Department of Energy nuclear facilities. One was a sediment collected from a reactor cooling pond located in the southeastern United States, and the other was a soil collected from the intermountain northwest region of the United States. Both matrixes were well-characterized, and we have monitored the extraction of uranium along with a variety of stable metal ions using a flow-through approach. As we have demonstrated previously (16, 17), monitoring the extraction of stable metal ions in addition to any contaminants aids in interpreting the sequential extraction results. Here, we describe the design and testing of our flowthrough reactor and compare partitioning information obtained by this flow-through approach to that obtained using the same extraction scheme by the batch method. We evaluate the effect of reaction time and compare the flowthrough results for the two different samples. Effects of temperature and concentration or correlations of these variables were not investigated. We also describe the advantages and limitations of the flow-through method as compared to a batch approach for the actinide-specific extraction scheme.
Experimental Section Sample Soil and Sediment. The samples for this work were provided from two different locations within the United States. One is a lake sediment from pond B at the Savannah River Site (SRS) in South Carolina, which we described in detail previously (16, 17, 38). It is an acidic, silt loam with low cation exchange capacity, typical of the highly weathered soils and sediments found in the coastal southeastern region of the US. The other sample is a surficial soil from the Idaho National Engineering and Environmental Laboratory (INEEL) near Idaho Falls, ID. Characterization information has been reported by Fox et al. (39) and Mincher et al. (40). It is a silt-clay material with
