Environ. Sci. Technol. 2009, 43, 4280–4286
Uranium Phases in Contaminated Sediments below Hanford’s U Tank Farm W O O Y O N G U M , * ,† Z H E M I N G W A N G , † R. JEFFREY SERNE,† BENJAMIN D. WILLIAMS,† CHRISTOPHER F. BROWN,† CLEVELAND J. DODGE,‡ AND AROKIASAMY J. FRANCIS‡ Pacific Northwest National Laboratory, Richland, Washington, and Brookhaven National Laboratory, Upton, New York
Received January 20, 2009. Revised manuscript received April 16, 2009. Accepted April 27, 2009.
Macroscopic and spectroscopic investigations (XAFS, XRF, and TRLIF) on Hanford contaminated vadose zone sediments from the U-tank farm showed that U(VI) exists as different surface phases as a function of depth below ground surface (bgs). Secondary precipitates of U(VI) silicate precipitates (boltwoodite and uranophane) were present dominantly in shallow-depth sediments (15-16 m bgs), while adsorbed U(VI) phases and polynuclear U(VI) surface precipitates were considered to dominate in intermediate-depth sediments (20-25 m bgs). Only natural uranium was observed in the deeper sediments (>28 m bgs) with no signs of contact with tank wastes containing Hanford-derived U(VI). Across all depths, most of the U(VI) was preferentially associated with the silt and clay size fractions of sediments. Strong correlation between U(VI) and Ca was found in the shallow-depth sediments, especially for the precipitated U(VI) silicates. Because U(VI) silicate precipitates dominate in the shallow-depth sediments, the released U(VI) concentration by macroscopic (bi)carbonate leaching resulted from both desorption and dissolution processes. Having different U(VI) surface phases in the Hanford contaminated sediments indicates that the U(VI) release mechanism could be complicated and that detailed characterization of the sediments using several different methods would be needed to estimate U(VI) fate and transport correctly in the vadose zone.
Introduction The most hazardous radioactive wastes generated at the Hanford Site, as a part of the plutonium production project from 1943 to 1989, have been stored in 177 underground storage tanks (1, 2). Because the Hanford storage tanks received radioactive wastes from different waste processes with varying histories, each underground storage tank contains different types and amounts of contaminants (3). As one of the first tank farms constructed for storage of highlevel radioactive waste at the Hanford Site, the U-tank farm contained several different types of wastes such as bismuth phosphate metal waste from the extraction of plutonium, REDOX waste, decontamination waste from the bismuth * Corresponding author e-mail:
[email protected]; phone: (509) 376-4627. † Pacific Northwest National Laboratory. ‡ Brookhaven National Laboratory. 4280
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 12, 2009
phosphate process, and evaporator feed and bottoms waste (4). Out of 12 tanks in the U-tank farm, four tanks (U-101, -104, -110, and -112) are considered to have leaked. See the Supporting Information (SI) Figure 1 and SI Table 1 for their locations and chemical compositions, respectively. Leakage from these four tanks has been estimated as bismuth phosphate metal wastes from tank 241-U-104 (2.1 × 105 L), smaller leaks of REDOX supernatant from tanks 241-U-110 (2.5 × 104 L) and 241-U-112 (3.2 × 104 L), and waste from tank 241-U-101 (1.9 × 104 L) (3). Different phases of uranium are expected to exist in the contaminated sediments surrounding these tanks because of the varied waste fluids that leaked. Hexavalent U(VI) is a prominent radioactive contaminant at the Hanford Site and has dispersed into both sediments and groundwater, posing a potential health and environmental risk to the nearby Columbia River (5). Contaminated sediments close to tank BX-102 in the B-BXBY tank farm show that uranium silicate precipitates, boltwoodite [HK(UO2)(SiO4) · 1.5(H2O)], and uranophane [Ca(UO2)2(SiO3OH)2(H2O)5] dominate (1, 6). Sediments derived from the Hanford 300 Area process ponds show three different solid phases of uranium: (1) coprecipitates with calcite, (2) metatorbernite [Cu(UO2)2(PO4)2 · 8(H2O)] precipitates, and (3) adsorbed uranium (2). Discrete grains of uranophane precipitates are also found in the sediments collected beneath the former North Process Ponds (NPP) of the Hanford 300 Area (7). Other than Hanford sediments, synthesized uranophane and boltwoodite were characterized using the fluorescence spectra (8, 9), and U(VI) coprecipitates with respect to carbonate minerals (10) and sorption complexes of U(VI) on different single minerals were also widely studied before (11-13). The U(VI)-contaminated sediments, with either (co)precipitated or adsorbed U(VI) phases, are considered to be potential secondary sources for continuous U(VI) release to pore water and groundwater. Dissolution and desorption processes of different U(VI) surface phases should result in different rates of U(VI) release from the contaminated sediments. Hence, a full understanding of the dominant U(VI) surface phases in the contaminated sediments is required to estimate the fate and transport of U(VI) at the Hanford Site. This paper focuses on U(VI) contaminated vadose zone sediments below the U-tank farm at the Hanford 200-West Site.
Materials and Methods Materials and Sediment Characterization. Eight core samples were collected from direct push borehole, C5602, installed southeast of tank U-105 next to U-104, which has had a documented waste leak in the past (SI Figure 1). Particle size analysis was conducted using a dry sieving method, and sample size fractions less than 2 mm were used for this study. The amounts of water-soluble and acid dissolved U(VI) were determined with an extract method using a 1:1 ratio of sediment to deionized water and 8 M HNO3 acid solution, respectively. A stronger microwave assisted digestion consisting of 16 M HNO3 (17%), 12 M HCl (7%), 32 M HF (3.3%), 0.5 g of H3BO3 (1.5%), and deionized water (71.2%) on a volume basis was used to determine the total U(VI) content in sediments. Labile U(VI) was determined using a (bi)carbonate extraction (pH ) 9.3) prepared using NaHCO3 (1.44 × 10-2 M) and Na2CO3 (2.8 × 10-3 M). Dissolved U(VI), Ca, and Si concentrations in the (bi)carbonate extracts were analyzed by ICP-MS and ICP-OES, respectively. A summary of characterization results for the selected sediments is shown 10.1021/es900203r CCC: $40.75
2009 American Chemical Society
Published on Web 05/12/2009
TABLE 1. Extracted U(VI) Concentrations and pHs from the Sediments microwave digestion extracted U for different sizes (µg/g) sample B1P-15.8 B1P-16.0 B1P-20.7 B1P-20.8 B1P-25.3 B1P-25.4 B1P-28.0 B1P-28.2 a
(B1P3H0B) (B1P3H0A) (B1P3H1B) (B1P3H1A) (B1P3H2B) (B1P3H2A) (B1PBB0B) (B1PBB0A)
mid-depth (meter bgs)a
water-extracted U (µg/g)
15.8 (51.8) 1.46 ( 0.10 16.0 (52.3) 1.34 ( 0.09 20.7 (67.8) 1.72 ( 0.12 20.8 (68.3) 5.16 ( 0.08 25.3 (82.8) 2.52 ( 0.19 25.4 (83.3) 0.835 ( 0.012 28.0 (91.8) 0.000357 ( 0.000019 28.2 (92.3) 0.000183 ( 0.000021
pH 7.59 7.48 7.49 7.54 7.57 7.59 7.44 7.30
bulk
0.5 < S < 2 mm
665 ( 10 448 ( 9.2 390 ( 9.2 326 ( 8.9 14.2 ( 1.2 7.86 ( 1.2 31.6 ( 2.3 11.8 ( 0.98 29.8 ( 3.1 12.9 ( 1.8 14.1 ( 1.0 7.92 ( 0.16 1.54 ( 0.15 0.875 ( 0.061 1.40 ( 0.12 0.612 ( 0.012
0.25 < S < 0.5 mm
0.075 < S < 0.25 mm
S < 0.075 mm
517 ( 7.8 391 ( 1.9 17.6 ( 1.1 18.9 ( 2.1 28.2 ( 0.95 11.4 ( 0.76 1.55 ( 0.10 1.37 ( 0.12
945 ( 23 614 ( 12 17.8 ( 2.3 34.4 ( 5.6 27.5 ( 1.5 13.1 ( 1.8 1.57 ( 0.56 1.29 ( 0.94
1340 ( 110 1000 ( 96 40.2 ( 7.8 58.5 ( 11 33.8 ( 2.5 18.7 ( 0.5 4.57 ( 0.91 2.37 ( 0.78
Sample collection depth (meter) below ground surface (bgs). Depths with a unit of feet are shown in parentheses.
in Table 1. Phosphate concentrations determined by measuring phosphorus in 1:1 water extract and 8 M HNO3 acid extract using ICP-OES are also shown in SI Table 2. Depthdescriptive sample IDs (e.g., BIP-15.8) were used to represent sediment collected at 15.8 m below ground surface (bgs) (Table 1). X-ray Fluorescence Analysis. High energy µ-X-ray fluorescence was used to examine the elemental composition as well as the spatial association of selected elements in sediments. Each sediment sample was prepared in two configurations either as a monolayer on Kapton tape or as a 0.16 cm thick sample. The sediment samples were placed in Mylar sample holders sealed with Kapton tape and analysis was performed at the National Synchrotron Light Source (NSLS) on beamline X27A. Microsynchrotron X-ray fluorescence data were collected on the samples at an incident X-ray energy of 17.4 keV using a 13-element Ge detector. Elemental mapping was obtained on an 800 µm × 800 µm spot using a focused beam of approximately 5-10 µm. Two-dimensional image maps were developed for 10 elements including Ca, Cr, Cu, Fe, Mn, Pb, Sr, Ti, U, and Zn, and used for determining the spatial and chemical distribution of uranium in relation to the other metals. X-ray Absorption Fine Structure Spectroscopy. Specific areas of high uranium concentration within the sediment samples were identified by µ-XRF analysis prior to µ-X-ray absorption near edge structure (µ-XANES) analysis. The oxidation state for uranium in sediments was determined using µ-XANES analysis on a 10 µm × 15 µm spot at the U Llll edge (17.166 keV). Uranyl nitrate and uranium dioxide were used as calibration standards to establish the absorption edge positions for U6+ and U4+, respectively. Spectra were normalized to the edge-jump using the ATHENA software (14). The fine silt/clay size fraction (
