Article pubs.acs.org/est
Impact of Environmental Curium on Plutonium Migration and Isotopic Signatures Hiromu Kurosaki,†,‡ Daniel I. Kaplan,§ and Sue B. Clark*,† †
Department of Chemistry, Washington State University, Post Office Box 644630, Pullman, Washington 99164, United States Savannah River National Laboratory, Aiken, South Carolina 29808, United States
§
S Supporting Information *
ABSTRACT: Plutonium (Pu), americium (Am), and curium (Cm) activities were measured in sediments from a former radioactive waste disposal basin located on the Savannah River Site, South Carolina, and in subsurface aquifer sediments collected downgradient from the basin. In situ Kd values (Pu concentration ratio of sediment/groundwater) derived from this field data and previously reported groundwater concentration data compared well to laboratory Kd values reported in the literature. Pu isotopic signatures confirmed multiple sources of Pu contamination. The ratio of 240Pu/239Pu was appreciably lower for sediment samples compared to the associated groundwater. This isotopic ratio difference may be explained by the following: (1) 240Pu produced by decay of 244 Cm may exist predominantly in high oxidation states (PuVO2+ and PuVIO22+) compared to Pu derived from the disposed waste effluents, and (2) oxidized forms of Pu sorb less to sediments than reduced forms of Pu. Isotope-specific Kd values calculated from measured Pu activities in the sediments and groundwater indicated that 240Pu, which is derived primarily from the decay of 244 Cm, had a value of 10 ± 2 mL g−1, whereas 239Pu originating from the waste effluents discharged at the site had a value of 101 ± 8 mL g−1. One possible explanation for the isotope-specific sorption behavior is that 240Pu likely existed in the weaker sorbing oxidation states, +5 or +6, than 239Pu, which likely existed in the +3 or +4 oxidation states. Consequently, remediation strategies for radioactively contaminated systems must consider not only the discharged contaminants but also their decay products. In this case, mitigation of Cm as well as Pu will be required to completely address Pu migration from the source term.
■
INTRODUCTION Among anthropogenic radioactive contaminants, Pu typically generates much interest. It can be derived from global fallout as well as localized inputs resulting from various industrial activities, such as production of heat sources for deep space exploration, waste management practices associated with nuclear energy production, and the development and testing of military devices. The isotopic ratio 240Pu/239Pu can be one indicator of its origin and purpose of production.1,2 Also, 239Pu and 240Pu are radiogenic progeny of isotopes with much shorter half-lives, e.g., 243Cm (half-life = 29.1 years) and 244Cm (halflife = 18.1 years), respectively. Therefore, when these isotopes of Pu and Cm coexist in an environmental system, the 240 Pu/239Pu isotopic ratio changes with time. The Pu contamination in environmental systems is frequently derived from multiple sources,3 although estimates of its transport and migration for environmental assessment purposes may not consider this fact. In a different area of environmental radiochemistry, sometimes called isotope geochemistry or nuclear forensics, the question of multiple source terms is explicitly considered and isotopic signatures are used to attribute Pu in the groundwater, soil, or sediment to specific sources.4−6 Knowledge of both the total Pu concentration and its isotopic composition are necessary for estimating Pu © 2014 American Chemical Society
ecological risk, because the activities of each isotope contribute differently to the total risk and, overall, long-term risk can be determined by radiogenic relationships.7 Actinide activities and isotopic ratios in F-Area groundwater located in the Savannah River Site (SRS) have been reported.1,8 Two sources of 240Pu were reported: one source is from the decay of 244Cm, and the second source is from the direct discharge of 240Pu in process effluents from the F-Area chemical separation facility (referred to as weapons-grade Pu). Unlike 244 Cm, very little 243Cm is found in SRS waste effluents and the local environment because the quantities produced in the SRS reactors are very small relative to 244Cm. On the basis of the observation that there was proportionally greater 240Pu than 239 Pu further downstream of the source than closer to the source, it was hypothesized by Buesseler et al.1 that most of the downstream 240Pu originated from the decay of the relatively more mobile 244Cm. They also hypothesized that the downstream 240Pu had a higher oxidation state than the 240Pu originating directly from the discharged waste, which was Received: Revised: Accepted: Published: 13985
February 25, 2014 September 7, 2014 October 28, 2014 October 28, 2014 dx.doi.org/10.1021/es500968n | Environ. Sci. Technol. 2014, 48, 13985−13991
Environmental Science & Technology
Article
predominantly weapons-grade Pu. The mechanism by which the +5 or +6 oxidation states predominate involves electron stripping after decay and is referred to as the Szilard−Chalmers process.8 Another interpretation of their observations is that the migration of trivalent Cm occurred prior to radioactive decay, and the observed 240Pu downgradient of the basin is the result of greater Cm than Pu mobility and subsequent radioactive decay to the Pu daughter isotope rather than the mobility of the Pu itself. It is not clear which mechanism predominates. In the environmental assessment field, the notion that the source of Pu makes a difference in Pu mobility has been referred to as “source-dependent” geochemical behavior.9 The objective of this study was to evaluate whether the Pu isotopes in the F-Area partitioned between the aqueous and solid phases in a manner consistent with source-dependent geochemical behavior, i.e., that 240Pu would tend to exist in the aqueous phase [presumably as Pu(V/VI)] and 239Pu would tend to exist in the solid phase [presumably as Pu(III/IV)]. In this field study, Pu, Am, and Cm activities as well as Pu isotopic ratios were measured in sediments collected from in or near the F-Area seepage basins. Using our sediment concentration results along with groundwater concentration data reported by Buesseler et al.,1 in situ distribution coefficients (Kd values, the ratio of the Pu concentration in sediment versus groundwater) were calculated. In addition, the distribution of the Pu and Cm isotopes after decay correction provided additional detail about multiple sources of Pu and the migration of 240Pu in this system. These results are presented and discussed in the context of contaminant source attribution and radiological risk.
■
MATERIALS AND METHODS Materials. Sediments. Sediments from two Department of Energy facilities were used in this study. One set of sediments was collected from the Idaho National Laboratory (INL) site, adjacent to a radioactive waste disposal area called the Subsurface Disposal Area (SDA). These sediments were analyzed previously for 241Am, 244Cm, 239Pu, and 240Pu10 and were used to validate the highly sensitive (sub-femtomoles per gram of soil) analytical technique applied in this study. Two different INL sediments were studied. One was collected from a depth of 0−4 cm adjacent to the SDA, near where wastes from Rocky Flats were disposed. This sample was contaminated with known quantities of 241Am and 239 + 240Pu11 and was used to validate the radioanalytical procedures employed. Unfortunately, no reference sediments containing Cm or a sample containing known quantities of Cm were available. Consequently, a method validation sediment was prepared by spiking an INL sediment with a known amount of 244Cm, as described in the Methods section. This sediment was a homogenized mixture of sediments collected from five different locations 1.3 km away from the SDA. This sediment had no detectable Am or Pu.12 The focus of this study involved the second set of samples that was obtained from SRS, a Department of Energy facility in South Carolina. These were the samples used to study Pu migration, and they were collected in and around the F-Area seepage basins (Figure 1). The seepage basins were unlined pits used between 1955 and 1988 to manage low-level waste effluents from a radiochemical separation plant at the SRS.13 As a result of this waste management practice, small quantities of actinides, fission products, and tritium, along with nitrate, and various mineral acids were released to the unlined basins and allowed to seep into the vadose zone. Approximately 12 Ci
Figure 1. Study area (SRS F-Area) showing the three seepage basins, the five sediment sample locations, sediment and groundwater 240 Pu/239Pu isotopic ratios, and general groundwater flow direction.
(444 GBq) of 239 + 240Pu, 0.23 Ci (8.51 GBq) of 241Am, and 0.35 Ci (12.95 GBq) of 244Cm were released to the basins during the 30+ years of operation.13 As will be described in more detail below, the 239 + 240Pu and 241Am concentrations were not decay-corrected because of their long half-lives (>400 years), but 244Cm concentrations were corrected to the year of greatest release to the seepage basins, 1973. Five sediment samples were collected from the F-Area, one directly from the seepage basin bottom (prior to closure in 1988) and the other four from well borehole samples that were 15−40 m below the surface (Figure 1). The acidic and oxic plume emanating from the seepage basins form a pH (pH 3.20−6.77) and Eh (654− 381 mV) gradient as it travels 0.7 km before resurfacing at Four Mile Branch (Figure 1).14 Chemicals. 243Am (Eckert & Ziegler Isotope Products, Valencia, CA) was used as a chemical yield tracer. Similarly, 242 Pu (NIST SRM4334H, National Institute of Standard and Technology, Gaithersburg, MD) was used as a chemical yield tracer for Pu isotopes. 244Cm was obtained from Eckert & Ziegler Analytics (Atlanta, GA). All reagents were analyticalgrade. Water was purified with a LABCONCO Water Pro PS system (Kansas City, MO). Radioanalytical separations used TEVA and TRU resins (Eichrom Technologies, Inc., Lisle, IL) and 2 mL plastic columns (BioRad, Hercules, CA). Methods. Preparation of 244Cm-Spiked Sediment Samples. A known amount of 244Cm solution was added to 10 g of the INL sediment (0−0.04 m depth) and then permitted to airdry completely. Care was taken to minimize the loss of 244Cm 13986
dx.doi.org/10.1021/es500968n | Environ. Sci. Technol. 2014, 48, 13985−13991
Environmental Science & Technology
Article
Table 1. Actinide Activities Measured in Sediments Collected from INL SDA and SRS F-Area sediment INL SDA SRS F-Area
distance from F-Area basin (m)
sample depth (m)
a
reported measuredc basin FEX-9 FEX-10 (18 m) FEX-10 (30 m) FSB-79A FSB-95C
0 95 77 77 375 15
0−0.04 0−0.04 0−0.1 21.6−22.0 18.2−18.6 30.2−30.5 16.5−16.8 22.9−23.2
241
Am (Bq g−1)
2.3 ± 0.3 2.0 ± 0.3 0.98 ± 0.06 0.02 ± 0.02 0.003 ± 0.002 0.0005 ± 0.0004 0.001 ± 0.0005 0.006 ± 0.001
244
Cm (Bq g−1)
0.32 ± 0.01 0.33 ± 0.06 0.7 ± 0.1 0.026 ± 0.007 0.0022 ± 0.0003 0.0002 ± 0.0002 0.0011 ± 0.0008 0.005 ± 0.001 b
239 + 240
Pu (Bq g−1)
0.48 ± 0.09 0.42 ± 0.02 3.2 ± 0.7 0.0011 ± 0.0004 0.0011 ± 0.0003 NDd ND ND
a
From ref 12. bKnown quantity of 244Cm added to the sediment. 241Am and 239 + 240Pu are existing sediment contaminants. cThe measured value was determined in this study to compare to the previously reported value. dND = not detected.
because of sorption onto the beaker by limiting the quantity of 244 Cm solution used and taking precaution in minimize contact between the added Cm solution and the beaker. Once completely dry, the 10 g aliquot of the spiked 244Cm sample was mixed with approximately 500 g of bulk sediment from the same source in a blender for several hours. After mixing, the sediment was analyzed using the procedure described below; measured values were compared to calculated values based on the amount of radionuclide spiked into sediment. In addition, different portions of sediment (top, middle, and bottom of the bottle) were analyzed to demonstrate homogeneity of the spiked sediment sample. Sediment Fusion. Sediment samples were weighed in platinum crucibles and then spiked with 243Am and 242Pu tracer solutions. Sediments were wet-ashed 3 times with concentrated nitric acid and 3 times with concentrated hydrofluoric acid. The ashed materials were then covered with 10 g of potassium fluoride and fused over a Bunsen burner. Once clear melts were obtained, the cake was cooled to room temperature. Approximately 15 mL of sulfuric acid was added to the cake to dissolve it (with heat if necessary), followed by 10 g of sodium sulfate. This mixture was heated with the burner until a transparent orange melt was obtained. This fused sample was then cooled to room temperature. Dissolution of Cakes and Fluoride Precipitation. Once cooled to room temperature, the fused cake was dissolved in boiling 1 M hydrochloric acid. A total of 4 mL of concentrated hydrofluoric acid was added to the solution to separate the lanthanides and actinides as fluoride precipitates. Fluoride precipitates were then separated from suspension by centrifugation for 30 min at 3400 rpm. Precipitates were repeatedly wet-ashed with concentrated nitric acid until a clear solution was obtained. The Pu oxidation state was adjusted in the 3 M nitric acid background solution to +IV with sodium nitrite prior to introducing the solution on the separation column (described below). Chromatographic Separation of Actinides. Actinides were separated chromatographically using Eichrom TEVA/TRU extraction resins. Plastic columns (2 mL) were filled with either TEVA or TRU and then pre-conditioned with 10 mL of 3 M nitric acid solution. The TRU columns were placed downgradient of the TEVA columns, so that elution from the TEVA columns flowed into the TRU columns. Eluent from the TRU columns was discarded. After pipetting the dissolved samples onto the TEVA columns, the beakers were rinsed 3 times with 1 mL of 3 M nitric acid, and each rinse was transferred to the TEVA columns after the prior solution completely drained. Once all TEVA and TRU columns were drained completely, the two columns were
separated. TEVA columns were washed with another 10 mL of nitric acid, followed by 20 mL of 9 M hydrochloric acid. Plutonium was eluted from TEVA columns with 20 mL of a 0.3 M sodium formaldehyde sulfoxylate, 0.1 M hydrochloric acid, and 0.1 M hydrofluoric acid solution. Americium and curium were eluted together from TRU columns with 15 mL of 4 M hydrochloric acid. Each actinide fraction was evaporated to dryness, wet-ashed with concentrated nitric acid, and then re-constituted with 1 M hydrochloric acid. About 50 μg of neodymium carrier was added to each fraction, and 1 mL (for Pu) or 3 mL (for Am/ Cm) of concentrated HF was added to form a fluoride precipitate. Precipitates were allowed to form for 30 min and then filtered from suspension with cellulose nitrate membrane filters (Cole-Parmer Instrument Company, Vernon Hills, IL). Filters were washed with 80% methanol and then air-dried for α spectrometry measurements. Spectrometry. α activities were measured using an ORTEC OCTETE Plus α spectrometry system (ORTEC, Oak Ridge, TN). Most samples were counted for 3−5 days, depending upon the activity level. Because 239Pu and 240Pu cannot be resolved by α spectrometry, their activities are reported as combined 239 + 240Pu activity. Background activities were determined by counting for a few months; this activity was subtracted from each spectrum to determine the final activity values. After α spectrometry measurements of the Pu samples, the filters were dissolved with concentrated nitric and perchloric acids on a hot plate. This solution was then reconstituted with 1 mL of 2% nitric acid for inductively coupled plasma−mass spectrometry (ICP−MS) analysis. Plutonium isotopic ratios were measured using Thermo Finnigan Element 2 sector field ICP−MS (Thermo Electron Corp., Bremen, Germany). Additional details about the α spectroscopy and ICP−MS are provided in the Supporting Information.
■
RESULTS AND DISCUSSION Study Area: SRS Sediments. Although some actinide concentration data exist for groundwater downgradient of the F-Area seepage basins,1,8,15 no data are available for the activity of these contaminants in the sediments. In this study, sediments collected directly from the seepage basin as well as sediments collected during the installation of several wells downgradient of the basins were analyzed. Figure 1 shows the study site, groundwater flow direction, seepage basins (source term), and sample locations. Method Validation: INL Sediment. Before attempting to quantify the very low concentrations of Pu, Am, and Cm contamination in the SRS sediment samples, we confirmed our radioanalytical procedures with previously analyzed sediments 13987
dx.doi.org/10.1021/es500968n | Environ. Sci. Technol. 2014, 48, 13985−13991
Environmental Science & Technology
Article
collected near the SDA at INL.11 The measured values for the INL SDA sediment sample agreed within the uncertainties of the reported values (Table 1). The concentration of 241Am was about an order of magnitude greater than that of either 244Cm or 239 + 240Pu. These results provided confidence in the analytical methods to quantify sediment Pu and Am concentrations. For Cm, our measured value agreed well with the concentration estimated on the basis of the spike addition to a known sediment mass (see footnote b in Table 1). Actinide Measurements by α Spectrometry. Table 1 shows α activities with uncertainties for 241Am, 244Cm, and 239 + 240 Pu in sediment samples collected from different locations. As expected, the sample from the basin (where the waste solution was initially discharged) had the highest activities for all three isotopes. On the basis of the observations that Pu activities were greater than Am and Cm activities in the basin and that very little or no Pu was detected in three of the downgradient sediments, while Am and Cm were detected in all five sediments, we deduced that Pu migration from the basin was slower than Am and Cm migration. Cm concentrations in the sediments were not inversely correlated with the distance from the basins (Table 1 and Figure 1). The highest Cm activity downstream from the basin was found in FEX-9 well, which is 95 m from the seepage basin edge. Cm concentrations, from the highest to the lowest, were as follows (distance from the basin edge): FEX-9 (95 m) > FSB-95C (15 m) > FEX-10 (77 m) > FSB-79A (375 m) (Table 1 and Figure 1). The lack of correlation with distance can likely be attributed to the limited number of samples analyzed, which do not accurately delineate the spatial extent of the Cm plume. The lack of correlation may also be attributed to secondary sample effects, such as (1) samples originating from different depths (Table 1), (2) sediment sample heterogeneity (and, therefore, sorption heterogeneity, despite efforts to ensure sample textural uniformity), and (3) the plume not making uniform contact with each of the samples. The trivalent cations, e.g., Am and Cm, exhibit similar migration behaviors, as expected. Figure 2 demonstrates the 1:1 correlation between Am and Cm, suggesting that they are moving in the aquifer at about the same rate. Pu Isotopic Measurement by ICP−MS. Reactors in SRS were used to produce weapons-grade Pu, which has a low
Figure 2.
241
Am and
244
isotopic ratio of 240Pu/239Pu of approximately 0.0628 (Table 2). The isotopic ratio for the basin sediment was 0.0729 ± 0.0003, Table 2. Isotopic Ratio before and after Correction for 244 Cm Decay 240
sediment samplea
depth (m)
direct measurement
basin FEX-9 FEX-10 FEX-10 FSB-79A FEB-95C F-reactorc global falloutd
0.1 22 18 39 17 23
0.0729 ± 0.0003 0.39 ± 0.05 0.23 ± 0.05 NDb ND ND 0.062 0.18−0.19
Pu/239Pu corrected for decay
244
Cm
0.0719 ± 0.0003 0.20 ± 0.05 0.20 ± 0.05
a
Basin, F-reactor, and global fallout were surface sediment samples. FEX-9, FEX-10, FSB-79A, and FEB-95C were sediment samples recovered from a borehole drilled during well development. bND = not detected. cFrom ref 8. dFrom ref 2.
which is distinctly higher than the historic 240Pu/239Pu ratio.8 This higher sediment isotopic ratio can be attributed to two factors: (1) contribution of atmospheric Pu fallout (the regional 240 Pu/239Pu ratio is 0.18−0.19), and (2) contribution of 240Pu derived from the decay of co-disposed 244Cm (Table 1). 244Cmdecay-corrected Pu isotopic ratios were calculated to provide an idea of the contribution of 244Cm decay to the observed 240 Pu/239Pu ratios. This correction was made by assuming that all 244Cm was released into the seepage basins during the year of greatest release, 1973. The 244Cm half-life (18.1 years) was then used to estimate the amount of 240Pu produced as a result of 244Cm decay during the time elapsed from the sampling date to 1973. When the contribution of 240Pu ingrowth from 244Cm decay is corrected, the isotopic ratio was re-calculated to be 0.0719 ± 0.0003, which is still significantly greater than previously reported for groundwater samples8 (Table 2). For the basin sediment sample, the effect of 244Cm ingrowth on the isotopic ratio is only about 1%. This effect is small because the relative activity of basin 239 + 240Pu is high compared to that of 244 Cm. However, when precise activity ratios are required, these ingrowth corrections are necessary. Pu isotopic ratios for sediments collected from downgradient wells were also calculated in two well locations, FEX-9 (95 m downgradient) and FEX-10 (77 m downgradient; Table 2). The isotopic ratios in these downgradient sediments were higher than that of the basin sediment. FEX-9 sediment had an isotopic ratio of 0.39 ± 0.05, and FEX-10 had an isotopic ratio of 0.23 ± 0.05. To evaluate whether Pu in the FEX-10 sample originated from decay of 244Cm or from 239 + 240Pu discharged to the seepage basin, the isotopic ratios were corrected for 240 Pu ingrowth (as described above), resulting in a decaycorrected isotopic ratio of 0.20 ± 0.05. The 244Cm-decaycorrected Pu isotopic ratio for the FEX-9 sample was identical to that for FEX-10, 0.20 ± 0.05. These corrected values suggest that Pu in the two downstream sediment samples did not originate primarily from Pu introduced into the seepage basins, e.g., 239Pu, but rather from the decay of 244Cm decay with contributions from global fallout (which has a 240Pu/239Pu ratio of 0.18−0.19). For the FEX-10 sediment, 240Pu originating from 244Cm decay accounted for 15% of all 240Pu in the sample, whereas
Cm activities in SRS F-Area sediments. 13988
dx.doi.org/10.1021/es500968n | Environ. Sci. Technol. 2014, 48, 13985−13991
Environmental Science & Technology
Article
Table 3. In Situ and Published 239Pu, Distribution Coefficients (Kd Values)
49% of 240Pu in FEX-9 can be attributed to 244Cm decay. Clearly, 244Cm progeny is affecting the observed Pu isotopic ratio for these samples. Consequently, Pu sorbed to sediments 95 m downgradient from the seepage basin likely did not originate from the basins but rather is derived primarily from 244 Cm decay. In this study area, two 244Cm-corrected Pu isotopic ratios were found to be very different, i.e., 0.0719 ± 0.0003 for the basin sediment and 0.20 ± 0.05 for the downgradient sediments. The lower ratio of the basin sediment reflects the composition of the waste stream discharged to the basin, which was derived from the processing of targets irradiated in the SRS reactors. The Pu isotopic ratio found in the downgradient sediments was, however, similar to that of the global fallout value, 0.18−0.19, indicating little to no release of Pu from the seepage basins. This is consistent with the total Pu concentrations at each location; e.g., the largest concentration of Pu in the samples that we analyzed was found in the sediments collected within the basin. In Situ Kd Value Estimations. In situ distribution coefficients (Kd) [units = (Bq/g)/(Bq/mL) = mL/g], a steady-state parameter, were calculated by combining the sediment activity values measured in this study (Table 1) with groundwater activity levels reported by Buesseler et al.1 in situ Kd =
[An]sediment [An]water
239
sediment concentration (mBq g−1) water concentration (mBq mL−1)a in situ Kd (this study; mL g−1)b batch (ad)sorption Kd18
Pu
240
Pu, and Cm
240
Pu
244
Cm
0.50 ± 0.03
0.6 ± 0.1
26 ± 7
0.005 ± 0.0003
0.063 ± 0.003
0.62 ± 0.03
101 ± 8
10 ± 2
40 ± 10
220c
40d
a
From ref 8; well 2. bEquation 1; well FEX-9 sediment; and pH 4.58. From ref 18; subsurface sandy sediment collected from the same aquifer as the F-Area; pH 4.53; 97% sand, 2% silt, and 1% clay; and spike = 10−9 M 239Pu(IV). dFrom ref 1; subsurface clayey sediment collected from the same aquifer as the F-Area; pH 3.9; 58% sand, 30% silt, and 12% clay; spike = 100 ppb 244Cm; and 1:15 solid (g)/liquid (mL). c
Sediment pH for the FEX-9 sample was 4.58, and the sediment collected from an uncontaminated portion of the same formation had a similar pH of 4.53. The batch Kd and the in situ Kd measurements did not require 244Cm correction, because the former was conducted in a 244Cm-free system that included primarily 239Pu and the latter was a 239Pu-specific Kd. The smaller distribution coefficient for 240Pu compared to 239 Pu is consistent with the hypothesis by Buesseler et al.1 that ∼75% of 240Pu in the F-Area groundwater exists in the oxidized forms, i.e., Pu(V) or Pu(VI), whereas 239Pu in the same groundwater existed almost exclusively in the reduced forms, Pu(III) and Pu(IV). As noted by Buesseler et al.,1 it is not clear why the isotopes have unique oxidation states after they have had presumably years to come to steady state. As mentioned above, the isotopic-specific geochemical behavior is attributed to the presence of 244Cm, which decays to 240Pu. 240Pu derived from parent 244Cm tends to be in the oxidized form, as compared to the steady-state reduced forms of 239Pu originating from the waste stream effluents.1 Interestingly, 240Pu/239Pu isotopic ratios determined in downgradient sediments were markedly different from those reported in nearby groundwater samples (Table 2). Our measurements indicate that two sediments in which Pu was detected had Pu isotopic ratios, 0.39 ± 0.05 and 0.23 ± 0.05, that were significantly greater than regional global fallout values, 0.18−0.19, primarily because of the contribution from 244 Cm decay. Dai et al.8 reported Pu isotopic ratios in groundwater in this area to be much higher (ratios = 3−8). This isotopic disequilibrium between Pu in groundwater and sediment may be explained by considering the following: (1) 240 Pu produced by decay of 244Cm may be in a higher oxidation state (PuVO2+ and/or PuVIO22+) as a result of the electronstripping Szilard−Chalmers process,8 and (2) oxidized forms of Pu have much lower tendencies to sorb to sediments than reduced forms of Pu.18 We do not provide direct evidence of the higher oxidation state of 240Pu, but rely on previous measurements made in this aquifer by Buesseler et al.1 From this information, it can be hypothesized that oxidized 240 Pu is produced by the decay of sediment 244Cm, which remains in groundwater (as opposed to sediment bound) because of its higher solubility. Conversely, 239/240Pu derived from the original disposed effluent is reduced. Plutonium in the lower oxidation states, Pu(III) and Pu(IV), is more inclined to
(1)
where [An]sediment and [An]water are the actinide concentrations associated with the sediment and water, respectively. The groundwater actinide concentration data were from well 2 by Buesseler et al.,1 which is about 34 m from well FEX-9. A total of 2 years elapsed between the collection of samples by Buesseler et al.1 and our sampling. It is important to note that the samples used to calculate the in situ Kd values were not precisely paired, as is the case with laboratory (ad)sorption tests of batch equilibrium Kd values. Given these caveats, it would be reasonable to assume that the uncertainty associated with the estimated in situ Kd values is greater than that associated with the laboratory Kd value. However, in our conceptual model for this construct, we assumed that the Pu plume was moving slowly (i.e., the 2 year difference between groundwater and sediment determinations does not greatly influence the Kd estimate). A shortcoming of this assumption is that seasonal and water-level changes in water chemistry are welldocumented.16 A comparison of contaminant characteristics between well 2 and well FEX-9, e.g., pH range of 3.2−3.4 and nitrate−nitrite concentrations of ∼50 mg L−1, supports our assumption of uniformity of the groundwater in the region of those two wells. Another note about the in situ Kd values (eq 1) is that they differ from the “calibrated field Kd” in that they are not based on an advection−dispersion equation and groundwater flow information to calculate a retardation factor.17 In situ Kd values were calculated using eq 1 and the activity concentrations of 244Cm, 239Pu, and 240Pu, as shown in Table 3. 239 Pu and 244Cm Kd values were compared to reported batch studies conducted with SRS sediments with similar pH values (pH 4.58).18 The in situ Kd values, using the field data, compared unexpectedly well to the literature Kd values;18 the two different methods of estimating Kd values were not expected to yield such similar values. Batch laboratory experiments,18 with complete mixing resulted in Kd values that were only twice as great as in situ Kd values (Table 3). 13989
dx.doi.org/10.1021/es500968n | Environ. Sci. Technol. 2014, 48, 13985−13991
Environmental Science & Technology
Article
experiments with SRS sediments promoted rapid reduction of Pu(V) to Pu(III/IV).20,21 Additional research is required to identify why 240Pu remains oxidized under these conditions that typically promote reduction. The in situ Pu Kd values were much lower in the groundwater that had a higher concentration of 240Pu, 10 ± 2 mL g−1. Because of this additional supply of 240Pu to the downgradient system, the Pu isotopic ratio in the sediment was a higher value than that of global fallout. In the groundwater, the isotopic ratio had a much higher value of 3.4.1 This was due to the additional supply of 240Pu from the decay of 244Cm and the fact that 240Pu produced by 244Cm was likely in the oxidized form, which was less prone to sorb to sediments than the reduced form. Not shown in Figure 3 is that mobile colloids influence Pu transport through this subsurface system.1,8,15 These results have implication to radiological risk calculations and the need for environmental remediation insofar that the radiological isotopes in groundwater, which is most commonly sampled during site characterization, may not be a good indication of the typically more abundant isotopes in sediments. At issue is that not all isotopes of an element pose the same risk. In the case of Pu, the risk (or radioactivity) posed by equal masses of 239Pu (half-life = 24 100 years) and 240Pu (half-life = 6560 years) are not the same because the latter has an appreciably greater specific activity. These results underscore the importance that, under conditions where multiple isotopes of varying toxicity are present, it is important to not only sample and characterize the groundwater but also the sediment, which typically contains the vast majority of contaminants.
sorb to the sediment. This isotopic difference in oxidation states results in a natural isotopic fractionation between the groundwater and sediment, which has implications on risk calculations and remediation decisions. In the case of Pu, the human risk presented by exposure to 239Pu (half-life = 24 100 years) is not the same as that posed by 240Pu (half-life = 6560 years). Additionally, the metal toxicity imposed by oxidized Pu is greater than that imposed by reduced Pu. It may be incorrect to assume that (easier acquired and analyzed) groundwater samples provide insight into the actual long-term risk imposed by radioactive contaminants in the subsurface sediments. Figure 3 presents a schematic of a proposed radiogeochemical behavior of 244Cm, 239Pu, and 240Pu in down-
■
Figure 3. Proposed Cm and Pu radiogeochemical behavior downgradient of the SRS F-Area seepage basins. Using well FEX-9, located 95 m downgradient of the seepage basins, it has Pu isotopic signatures, suggesting that Pu comes from either global fallout or 244Cm and not from the seepage basins. 244Cm decay produces primarily 240Pu(V/VI), whereas global fallout contains primarily 240+239Pu(III/IV), which may exist in dissolved, sorbed, or mobile colloid forms. The role of mobile colloids in transporting Pu is not depicted.
ASSOCIATED CONTENT
S Supporting Information *
Information about the sampling locations at the INL SDA and the SRS F-Area (Table S1) and instrument parameters for ICP−MS (Table S2). This material is available free of charge via the Internet at http://pubs.acs.org.
■
gradient (well FEX-9) groundwater based on this study and published data.1,8 First, Pu and Cm were discharged from SRS F-Area facility to the seepage basins.13,19 The plume pH at the study site, as mentioned above, is quite low, pH 3−4.4. Cm(III) sorbs less at lower pH levels and has a sharper pH sorption edge than Pu(IV);1 consequently, Cm would be expected to travel further in the subsurface environment. The in situ 244Cm Kd value was 40 ± 10 mL g−1 (FEX-9 well in Table 3), which is similar to the laboratory batch measurement reported by Buesseler et al.,1 40 mL g−1. 239 + 240Pu in the far field originated primarily from global fallout with a 240Pu/239Pu isotopic ratio of 0.18 and a Kd value of 101 ± 8 mL g−1 (FEX-9 well in Table 3). Comparing these in situ Kd values suggests that 244Cm should be more mobile in the study site than 239 + 240Pu. Additionally, Pu in the far field existed with elevated 240Pu concentrations, a result of 244Cm decay. 240Pu derived from 244Cm has previously been shown to exist primarily in the oxidized form, a result presumably because of the electron-stripping Szilard−Chalmers process.1 What is not clear is why 240Pu remains oxidized in the SRS subsurface environment. Buesseler et al.1 measured reduced Pu(III/IV) but primarily oxidized Pu(V/VI) in the F-Area plume. The Eh (0.654−0.381 V) and pH (3.20−6.77) conditions of the F-Area plume14 would suggest that Pu(IV) would likely dominate the solid phase. Furthermore, laboratory
AUTHOR INFORMATION
Corresponding Author
*Telephone: 509-335-1411. Fax: 509-335-8867. E-mail: s_
[email protected]. Present Address ‡
Hiromu Kurosaki: Oak Ridge National Laboratory, Post Office Box 2008, MS 6105, Oak Ridge, Tennessee 37831, United States. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
The authors acknowledge Dr. Evgeny Taskev of Eckert & Ziegler Analytics for providing 244Cm standard solution. This project was funded by the United States Department of Energy, Basic Energy Science (DE-FG02-06ERI15782). Sue B. Clark also acknowledges support from the United States Department of Homeland Security, Academic Research Initiative (Contract 2009DN077-ARI03302). Daniel I. Kaplan received funding from the Department of Energy’s Subsurface Biogeochemistry Research Program within the Office of Science (Contract SCW-0083). 13990
dx.doi.org/10.1021/es500968n | Environ. Sci. Technol. 2014, 48, 13985−13991
Environmental Science & Technology
■
Article
River Site Environment, Revision 1; Westinghouse Savannah River Company, LLC (WSRC): Aiken, SC, 1992; WSRC-RP-92-879. (20) Hixon, A. E.; Hu, Y. J.; Kaplan, D. I.; Kukkadapu, R. K.; Nitsche, H.; Qafoku, O.; Powell, B. A. Influence of iron redox transformations on plutonium sorption to sediments. Radiochim. Acta 2010, 98 (9− 11), 685−692. (21) Kaplan, D. I.; Powell, B. A.; Demirkanli, D. I.; Fjeld, R. A.; Molz, F. J.; Serkiz, S. M.; Coates, J. T. Influence of oxidation states on plutonium mobility during long-term transport through an unsaturated subsurface environment. Enivron. Sci. Technol. 2004, 38, 5053−5058.
REFERENCES
(1) Buesseler, K. O.; Kaplan, D. I.; Dai, M.; Pike, S. Sourcedependent and source-independent controls on plutonium oxidation state and colloid associations in groundwater. Environ. Sci. Technol. 2009, 43 (5), 1322−1328. (2) Kelley, J. M.; Bond, L. A.; Beasley, T. M. Global distribution of Pu isotopes and 237Np. Sci. Total Environ. 1999, 237, 483−500. (3) Riley, R. G., Zachara, M. M. Chemical Contaminants on DOE Lands and Selection of Contaminants Mixtures for Subsurface Science Research; United States Department of Energy (U.S. DOE): Washington, D.C., 1992; DOE/ER-0547T. (4) Kersting, A. B. Plutonium transport in the environment. Inorg. Chem. 2013, 52, 3533−3546. (5) Kersting, A. B.; Efurd, D. W.; Finnegan, D. L.; Rokop, D. J.; Smith, D. K.; Thompson, J. L. Migration of plutonium in ground water at the Nevada Test Site. Nature 1999, 397, 56−59. (6) Novikov, A. P.; Kalmykov, S. N.; Utsunomiya, S.; Ewing, R. C.; Horreard, F.; Merkulov, A.; Clark, S. B.; Tkachev, V. V.; Myasoedov, B. F. Colloid transport of plutonium in the far field of the Mayak Production Association, Russia. Nature 2003, 314, 638−641. (7) Crowley, K. D.; Ahearne, J. F. Managing the environmental legacy of U.S. nuclear-weaopns production. Am. Sci. 2002, 90, 914− 923. (8) Dai, M.; Kelley, J. M.; Buesseler, K. O. Sources and migration of plutonium in groundwater at the Savannah River Site. Environ. Sci. Technol. 2002, 36, 3690−3699. (9) Choppin, G. R. Actinide speciation in aquatic systems. Mar. Chem. 2006, 99, 83−92. (10) Payne, R. F. Investigation of radiological contamination of soil samples from Idaho National Laboratory. Ph.D. Dissertation, Washington State University, Pullman, WA, 2006. (11) Idaho National Laboratory (INL). A Comprehensive Inventory of Radiological and Nonradiological Contaminants in Waste Buried in the Subsurface Disposal Area of INEL RWMC; INL: Idaho Falls, ID, 1995; INEL-95/0310. (12) Choiniere, A. D.; Payne, R. F.; Knaack, C. M.; Smith, S. C.; Clark, S. B. Distribution of uranium, plutonium, and 241Am in soil samples from Idaho National Laboratory. J. Radioanal. Nucl. Chem. 2009, 282, 1013−1017. (13) Westinghouse Savannah River Company, LLC (WSRC). Annual Corrective Action Report for the F-Area Hazardous Waste Management Facility, the H-Area Hazardous Waste Management Facility, and the Mixed Waste Management Facility; WSRC: Aiken, SC, 2006; WSRCRP-2006-4011. (14) Otosaka, S.; Schwehr, K. A.; Kaplan, D. I.; Roberts, D. A.; Zhang, S.; Xu, C.; Li, H.-P.; Ho, Y.-F.; Brinkmeyer, R.; Yeager, C. M.; Santschi, P. H. Factors controlling mobility of 127I and 129I species in an acidic groundwater plume at the Savannah River Site. Sci. Total Environ. 2011, 409 (19), 3857−3865. (15) Kaplan, D. I.; Bertsch, P. M.; Adriano, D. C.; Orlandini, K. Actinide association with groundwater colloids in a Coastal Plain aquifer. Radiochim. Acta 1994, 66/67, 181−187. (16) Ciffroy, P.; Barnier, J. M.; Pham, M. K. Kinetics of the adsorption and desorption of radionuclides of Co, Mn, Cs, Fe, Ag, and Cd in fresh water systems: Experimental and modeling approaches. J. Environ. Radioact. 2001, 55 (1), 71−91. (17) Krupka, K. M.; Kaplan, D. I.; Whelan, G.; Serne, R. J.; Mattigod, S. V. Understanding Variation in Partition Coefficient, Kd, Values. Vol. I: The Kd Model, Methods of Measurement, and Application of Chemical Reaction Codes; Office of Air and Radiation, Office of Solid Waste and Emergency Response, United States Environmental Protection Agency (U.S. EPA): Washington, D.C., 1999; EPA 402-R-99-004A. (18) Powell, B. A.; Kaplan, D. I.; Serkez, S. M.; Coates, J.; Fjeld, R. A. Pu(V) transport through Savannah River Site soilsAn evaluation of a conceptual model of surface-mediated reduction to Pu(IV). J. Environ. Radioact. 2014, 131, 47−56. (19) Carlton, W. H.; Evans, A. G.; Geary, L. A.; Murphy, C. E., Jr.; Pinder, J. E.; Strom, R. N. Assessment of Plutonium in the Savannah 13991
dx.doi.org/10.1021/es500968n | Environ. Sci. Technol. 2014, 48, 13985−13991
