Source-Dependent and Source-Independent Controls on Plutonium

Jan 27, 2009 - Aiken, South Carolina 29808, and State Key Laboratory of. Marine Environmental Science,. Xiamen University, Xiamen 361005, China...
1 downloads 0 Views 866KB Size
Environ. Sci. Technol. 2009, 43, 1322–1328

Source-Dependent and Source-Independent Controls on Plutonium Oxidation State and Colloid Associations in Groundwater K E N O . B U E S S E L E R , * ,† DANIEL I. KAPLAN,‡ MINHAN DAI,§ AND STEVEN PIKE† Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543, Savannah River National Laboratory, Aiken, South Carolina 29808, and State Key Laboratory of Marine Environmental Science, Xiamen University, Xiamen 361005, China

Received October 6, 2008. Revised manuscript received December 11, 2008. Accepted December 23, 2008.

Plutonium (Pu) was characterized for its isotopic composition, oxidation states, and association with colloids in groundwater samples near disposal basins in F-Area of the Savannah River Site and compared to similar samples collected six years earlier. Two sources of Pu were identified, the disposal basins, which contained a 240Pu/239Pu isotopic signature consistent with weapons grade Pu, and 244Cm, a cocontaminant that is a progenitor radionuclide of 240Pu. 240Pu that originated primarily from 244Cm tended to be appreciably more oxidized (Pu(V/VI)), less associated with colloids (∼1 kDa - 0.2 µm), and more mobile than 239Pu, as suggested by our prior studies at this site.Thisisnotevidenceofisotopefractionationbutrather“sourcedependent” controls on 240Pu speciation which are processes that are not at equilibrium, i.e., processes that appear kinetically hindered. There were also “source-independent” controls on 239Pu speciation, which are those processes that follow thermodynamic equilibrium with their surroundings. For example, a groundwater pH increase in one well from 4.1 in 1998 to 6.1 in 2004 resulted in an order of magnitude decrease in groundwater 239Pu concentrations. Similarly, the fraction of 239 Pu in the reduced Pu(III/IV) and colloidal forms increased systematically with decreases in redox condition in 2004 vs 1998. This research demonstrates the importance of sourcedependent and source-independent controls on Pu speciation which would impact Pu mobility during changes in hydrological, chemical, or biological conditions on both seasonal and decadal time scales, and over short spatial scales. This implies more dynamic shifts in Pu speciation, colloids association, and transport in groundwater than commonly believed.

Introduction An estimated 1.3 × 1016 Bq of plutonium (∼4200 kg of Pu using a 240Pu/239Pu atom ratio of 0.18) has been released to the environment as a result of nuclear testing, the development of nuclear weapons, reactor operations, and the nuclear * Corresponding author phone: (508) 289-2309; e-mail: kbuesseler@ whoi.edu. † Woods Hole Oceanographic Institution. ‡ Savannah River National Laboratory. § State Key Laboratory of Marine Environmental Science. 1322

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 5, 2009

fuel cycle (1). One legacy of the Cold War Era processing of nuclear materials includes contamination of many billions of cubic meters of soils and groundwater at DOE sites. Thus, the task of managing, cleaning-up, and monitoring these sites is large, and decisions regarding remediation and stewardship at DOE sites require sound scientific advice and site-specific field programs to understand Pu transport in subsurface vadose and groundwater zones. While many details of Pu geochemistry in the laboratory are understood, its fate in the natural environment is far more difficult to predict. This is because of a variety of factors, including the coupled interactions of ambient hydrological, geochemical, and biological processes that control the many possible chemical forms, or species, of Pu. Differences in Pu speciation, in particular oxidation state, result in its variable affinity for stationary and colloidal surfaces in the subsurface environment. Understanding Pu contaminant transport is also limited by insufficient and unreliable field data. This is due in part to the extremely low concentrations of Pu in the environment, making Pu detection and detailed speciation work challenging. Most spectroscopic techniques require micromolar (10-6 M) or greater concentrations to be effective, whereas environmental Pu concentrations are typically in the subnanomolar range (99% filterable in ∼100 kDa to 1 µm size range). Importantly, they also demonstrated using 240Pu/239Pu isotopic ratios that the Pu in these samples originated from the underground nuclear test and not from overland or atmospheric pathways. However, finding a contaminant distant from a point source is not sufficient to confirm colloid-associated transport. Wind-born dust and surface waters, not subsurface mobile colloids, were the primary far-field transport mechanisms for Pu at Rocky Flats (12). Choppin (13) and others have used the term ”sourcedependent” to refer to environmental behavior of Pu that depends upon the specific Pu source. Such Pu species are not at thermodynamic equilibrium with their surroundings but remain in forms that are influenced by the source. One example is Pu associated with solids entrained in fallout from above ground nuclear weapons testing at Nevada which are preferentially deposited in deep ocean sediments, because they are less soluble in seawater than Pu fallout from high 10.1021/es8028318 CCC: $40.75

 2009 American Chemical Society

Published on Web 01/27/2009

altitude testing (14-18). Another source-dependent effect to be discussed here is the production of 240Pu from the decay of 244Cm (t1/2 ) 18.1 years), resulting in more oxidized forms of 240Pu in groundwater (19). “Source-independent” effects refer to those geochemical processes that are sufficiently rapid to permit Pu speciation to come to steady state. Source-independent effects for Pu commonly include changes in redox state, pH, or concentrations of inorganic or organic ligands (4, 13, 20-22). Thus hydrological, chemical, or biological changes in groundwater conditions on both seasonal and decadal time scales, and over short spatial scales, would be expected to influence source-independent Pu speciation. The objective of this study was to evaluate changes in Pu speciation in the groundwater at the Savannah River Site (SRS) in the time since our previous sampling six years ago (19). Furthermore, attention was directed at explaining the systematic increases with distance from the source in 240Pu/ 239 Pu ratio, which was inferred in prior sampling to be a consequence of the decay of 244Cm, a progenitor radionuclide of 240Pu (19). Therefore, this study also involved the first measurements of 244Cm in SRS groundwater.

Experimental Section Study Site. The SRS was constructed in the early 1950s in western South Carolina for the production of nuclear materials, mainly 3H and 239Pu for defense purposes and 238Pu and various transuranic radionuclides for medical, industrial, and scientific applications (23, 24). F-area seepage basins received waste effluents from the separation facilities, including waste from the nitric acid recovery unit and the evaporators that concentrated the dilute uranium nitrate solutions (25). For the most part, the groundwater flow direction is due south to southeast from the seepage basins toward Fourmile Branch, a second-order stream and tributary of the Savannah River (a map of the study site is presented in Supporting Information, Figure S1). The same wells sampled by Dai et al. (19) in 1998 were resampled in 2004 including a “background” well located upstream of the seepage basins (Well 1) and three additional wells that formed a transect downstream of the seepage basins, Wells 2, 3, and 4, which are 10, 110, and 700 m, respectively, from the seepage basins. Since 2002, SRS has participated in a remediation effort which has injected 12 million liters of a sodium hydroxide-sodium bicarbonate solution into the groundwater for the purpose of increasing pH and increasing sediment sorption of plume contaminants (26, 27). The current remediation system consists of an impermeable barrier wall that funnels groundwater toward an opening gate where wells are used for base injection (socalled “funnel and gate”). Significant decreases in major contaminants (Cd, Hg, Ra, 241Am, 243,244Cm, 90Sr, and 238U) have been measured (26, 27). Evidence of remedial activity was noted in this latest sampling at Well 4, located ∼30 m downstream of the injection gate, where the pH increased from 4.1 in 1996 to as high as pH 6.5 in 2004 (see Supporting Information, Table S1). Significant changes in Pu solubility and speciation are expected to occur as a result of these pH changes. On the basis of laboratory measurements with SRS sediments, Pu solubility should decrease by 1 order of magnitude as the pH increases from 4.1 (10-7.8 M Pu) to 6.1 (10-8.8 M Pu) (5). On the basis of thermodynamic calculations of a sediment solution in contact with PuO2(s) in a mildly oxidizing soil environment (pO2 ) 16 atm), the dominant Pu species at pH 4.1 are PuO2+, PuO2CO3OH-, PuO2(OH)0, and PuO22+, whereas the dominant Pu species at pH 6.1 are PuO2CO3OH-, PuO2CO3(OH)22-, and PuO2(OH)3- (28). Field Sampling. Part of the current confusion regarding Pu mobility in groundwater and more generally the role of colloids in facilitating contaminant transport is due to

sampling artifacts and biases that need to be carefully considered in the design of field studies (6, 7, 29, 30). For this reason, special attention was directed at sample collection and processing in the field. Our approach involved a combination of methods, including low flow rate groundwater sampling to minimize remobilizing Pu that would otherwise be stationary (e150 mL min-1) and directly sending the sample through a multimembrane, cross-flow ultrafiltration (CFF) assembly maintained in a N2 atmosphere (details in refs 19, 30, 31). In-line filtration (0.2 µm filter) is used to ensure that no large particles enter the CFF system, though at low flow rates little difference between unfiltered and filtered Pu concentrations have been measured (19). Separation of the Pu oxidation state fractions was conducted immediately in the field. Strict protocols with respect to tracemetal and trace-organic clean sampling techniques were adopted during collection and field processing of samples. Details of the CFF method, redox controls, Pu oxidation state separations, and quality control on comtamination and sorptive losses are presented in Supporting Information. Plutonium Isotopes by Thermal Ionization Mass Spectrometry. Field samples underwent separation and purification in a clean laboratory as described in prior studies (19, 31) prior to thermal ionization mass spectrometry (TIMS) analyses. Pu isotopes were analyzed using a three-stage TIMS instrument at the Pacific Northwest National Laboratory (Dai et al., ref 19 and references therein). Sample aliquots seldom contained more than ∼108 Pu atoms; thus, most of the TIMS duty cycle was allocated to the measurement of the minor241Pu and 242Pu isotopes. Given the TIMS duty cycle allocation, the detection limit was about 104 atoms, which is appreciably less than sample concentrations typically that were 106 - 108 atoms kg-1 239Pu. Curium Analysis. New to this study was the direct determination of 244Cm by both alpha spectroscopy and TIMS methods. For alpha counting, 243Am was added as a yield tracer and the actinides were preconcentrated using a calcium phosphate coprecipitation method. Americium and curium were separated from the rest of the actinides by extraction chromatography on TRU resin (32). The purified Am and Cm isotopes were coprecipitated with cerium fluoride, filtered, and counted by alpha-spectrometry. SRS groundwater samples contain some 243Am, so a second set of samples were analyzed by alpha-spectrometry without yield tracer in order to calculate Am and Cm isotope ratios originally present in the sample. For the TIMS-based 244Cm method, 248Cm was added as a yield tracer and trivalent Cm was separated as an anionic thiocyanate complex by extraction chromatography using TEVA and analytical grade anion and cation exchange resins. Further purification and, ultimately, preparation of the TIMS source was accomplished using single anion exchange resin bead (Bio-Rad AG1-X4) which extracts the Cm(III) as an anionic nitrate complex in nonaqueous media. Following equilibration, the anion exchange resin bead was transferred to the trough of a rhenium filament and pyrolyzed and the filament carburized to produce the final TIMS source. Agreement between 244Cm determined from the same wells and dates by independent methods (1988, alpha counting, TIMS, ingrowth, Table 1) attests to the accuracy of these results.

Results and Discussion 239 Pu Concentration, Oxidation State, and Colloidal Fraction. Shown in Figure 1a-c are groundwater 239Pu data from the most recent field sampling in October 2004 compared to the same four wells sampled in April 1998 and reported in Dai et al. (19). All concentration data are presented in units of 106 atoms kg-1 (∼ 0.0065 pCi kg-1 or 0.2 mBq kg-1 for 239 Pu). The drinking water limit for 239Pu is 15 pCi L-1 or

VOL. 43, NO. 5, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1323

TABLE 1. Comparison of Filtrate 244Cm Concentrations Determined by Alpha Spectroscopy, TIMS, and Ingrowth Calculations 244

CM (106 atoms L-1)

well ID alpha 2004b TIMS 2003 alpha 1998 2 ( 11 50 ( 24 38 ( 14 9(5

1 2 3 4

ingrowth calculationsa TIMS 1998 1998