Cr Stable Isotopes in Snake River Plain Aquifer ... - ACS Publications

Kathrin Schilling , Thomas M. Johnson , Karaj S. Dhillon , and Paul R. D. Mason .... Aspects of Hexavalent Chromium Pollution of Thebes Plain Aquifer,...
0 downloads 0 Views 370KB Size
Environ. Sci. Technol. 2011, 45, 502–507

Cr Stable Isotopes in Snake River Plain Aquifer Groundwater: Evidence for Natural Reduction of Dissolved Cr(VI) AMANDA L. RADDATZ,† T H O M A S M . J O H N S O N , * ,† A N D TRAVIS L. MCLING‡ Department of Geology, University of Illinois at Urbana-Champaign, 208 Natural History Building, 1301 West Green Street, Urbana, Illinois 61801, United States, and Idaho National Laboratory, Idaho Falls, Idaho 83415-2107, United States

Received June 12, 2010. Revised manuscript received October 25, 2010. Accepted November 16, 2010.

At Idaho National Laboratory, Cr(VI) concentrations in a groundwater plume once exceeded regulatory limits in some monitoring wells but have generally decreased over time. This study used Cr stable isotope measurements to determine if part of this decrease resulted from removal of Cr(VI) via reduction to insoluble Cr(III). Although waters in the study area contain dissolved oxygen, the basalt host rock contains abundant Fe(II) and may contain reducing microenvironments or aerobic microbes that reduce Cr(VI). In some contaminated locations, 53Cr/ 52 Cr ratios are close to that of the contaminant source, indicating a lack of Cr(VI) reduction. In other locations, ratios are elevated. Part of this shift may be caused by mixing with natural background Cr(VI), which is present at low concentrations but in some locations has elevated 53Cr/52Cr. Some contaminated wells have 53Cr/52Cr ratios greater than the maximum attainable by mixing between the inferred contaminant and the range of natural background observed in several uncontaminated wells, suggesting that Cr(VI) reduction has occurred. Definitive proof of reduction would require additional evidence. Depth profiles of 53Cr/52Cr suggest that reduction occurs immediately below the water table, where basalts are likely least weathered and most reactive, and is weak or nonexistent at greater depth.

Introduction Hexavalent chromium, Cr(VI), is a common groundwater contaminant that may have significant adverse health effects (1). Elevated Cr(VI) concentrations can result from weathering of ultramafic rocks or industrial processes such as metals plating, leather tanning, and cooling water conditioning (2-4). In the environment, Cr exists in two main oxidation states, Cr(VI) and Cr(III), which have distinctly different toxicity and chemistry. In oxic waters, Cr may be stable as chromate (CrO42-), dichromate (Cr2O72-), or hydrochromate (HCrO4-), which are toxic, soluble anions. Remediation of contaminated sites often involves reduction of Cr(VI) to * Corresponding author phone: (217) 244-2002; fax: (217) 2444996; e-mail: [email protected]. † University of Illinois at Urbana-Champaign. ‡ Idaho National Laboratory. 502

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 45, NO. 2, 2011

Cr(III), a less toxic, insoluble, and immobile form (5). Eh values for the thermodynamic redox transition are high (450 mV at pH 7). Cr(VI) reduction can occur naturally through the action of aqueous Fe(II), Fe(II)-bearing minerals, reduced sulfur species, organic compounds, anaerobic microbes, and aerobic microbes, and the kinetic properties of Cr(VI) allow many of these reactions to proceed rapidly (e.g., refs 6-10). However, proving that reduction occurs in the subsurface and quantifying reduction rates are often challenging, because dilution, advection, and sorption can also decrease Cr(VI) concentrations. Cr stable isotope methods provide a promising approach for monitoring Cr(VI) reduction (4, 11-14). As Cr(VI) reduces to Cr(III), a kinetic isotope effect occurs, whereby the reaction product is enriched in lighter isotopes relative to the reactant, and the remaining Cr(VI) becomes enriched in heavier isotopes as the reaction proceeds. This unequal partitioning of isotopes, or isotopic fractionation, is quantified via measurements of the 53Cr/52Cr ratio. This kinetic isotope effect has been linked to Cr(VI) reduction in both laboratory (11, 12) and field settings (4, 13). Accordingly, increases in groundwater 53Cr/52Cr ratios relative to that of the original contaminant can be used to detect and possibly quantify Cr(VI) reduction. Dissolved Cr(VI) concentrations above permissible limits have been documented in groundwater at the Idaho National Laboratory (INL). Cr(VI) was released in the 1960s and 1970s, resulting in elevated Cr(VI) concentrations in the regional eastern Snake River Plain aquifer (ESRPA) and in a perched water zone that developed underneath waste disposal ponds. Cr concentrations have been decreasing since the 1980s (15, 16). It is clear that dispersion of the Cr plume(s) within the fast-flowing aquifer is responsible for much of this decrease, but Cr(VI) reduction, which sequesters Cr semipermanently, may also contribute. Comprised mainly of basalts, the ESRPA contains abundant Fe(II)-bearing minerals that could drive natural attenuation via reduction of Cr(VI) (6, 10). Alternatively, certain bacteria are known to reduce Cr(VI) (7-9) and could be active in the aquifer. However, sampled ESRPA groundwaters typically have >6 mg/L dissolved oxygen, and it is often assumed that significant Cr(VI) reduction does not occurs under these conditions. In this study, Cr(VI) concentration and Cr stable isotope analyses of 66 recent (2007-2008) and three archived (1967) groundwater samples from INL are presented (Figure 1 and Supporting Information, Table S1). Sampling locations include six multilevel monitoring wells that reveal depthrelated patterns. The data are used to assess Cr(VI) reduction in INL groundwater. Site Background and Methods. Figure 1 gives locations of sampling points. Site hydrogeology has been reviewed by Smith (17). The ESRPA underlies INL and in this area consists of a sequence of tholeiitic olivine basalt lava flows with interbedded sediments. Basalt flow ages in the study area range from 165 000 to 640 000 years and vary from oxidized and highly altered to essentially unaltered (18). Many contain Fe(II)-bearing olivine and pyroxene crystals showing little alteration to depths of at least a few hundred meters. Interbedded sediments consist of relatively low permeability sandy to silty alluvial and eolian material with occasional gravel. Groundwater flows regionally to the south-southwest at velocities >1 km/yr in some areas. Depth to water is between 200 and 300 m in the study area. Surface water flow is dominated by the Big Lost River, which flows only rarely within the study area and sporadically adds water to the subsurface (19). 10.1021/es102000z

 2011 American Chemical Society

Published on Web 12/01/2010

Wastewater was disposed directly to the ESRPA through a 389 m deep injection well at the Reactor Technology Complex (RTC). Between 1964 and 1972, Cr(VI)-bearing water from cooling towers was injected (15). By 1974, Cr had migrated 600 m downgradient to well USGS 65, which contained 280-460 µg/L Cr between 1982 and 1985 (15). Concentration decreased strongly by 2004 (16). Given the high flow velocities, it is expected that the Cr has reached at least several kilometers downgradient. Wells at the Radioactive Waste Management Complex (RWMC), approximately 13 km downstream of the RTC, have shown elevated Cr levels as recently as 2004 (16). However, the RWMC may contain Cr(VI) wastes that could act as local contaminant sources (20). Cr-contaminated water disposed of in unlined waste ponds at the RTC between 1952 and 1964 (21) infiltrated and created two perched water zones that migrated 1 km to the southeast (15). The upper perched zone, located in surface alluvium, is approximately 30-45 m below land surface. The lower perched zone, at 50-60 m, is formed atop sediment layers between lava flows. Between 1993 and 1995, Cr concentrations in perched zone wells reached as high as 800 µg/L; they have since decreased and have remained below 100 µg/L since 2001 (19). Dilution by Cr-free water from the ponds caused some or all of this decrease (15). Sampling and Analytical Methods. Details of sample collection and analytical methods are given in the Supporting Information. Isotope ratio analyses were completed using a 54 Cr/50Cr double isotope spike technique (11, 12, 22). An aliquot of spike solution was added to each sample; Cr(VI) was then purified via anion exchange. Isotope ratios were measured on a Nu Plasma HR multicollector inductively coupled plasma mass spectrometer (MC-ICP-MS) following the method of Schoenberg et al. (22). The 53Cr/52Cr ratio results are reported as relative deviations from the NIST SRM979 standard δ53Cr )

Rsample -1 Rstd

(1)

where Rsample and Rstd are the 53Cr/52Cr ratios of sample and standard, respectively. δ53Cr values are reported as per mil (‰) quantities. Uncertainty was (0.14‰ (95% confidence; twice the root-mean-square difference for 17 pairs of duplicate sample preparations; Supporting Information, Table S1). Cr(VI) concentrations were determined via isotope dilution against the double spike solution; uncertainty was (10% (95% confidence, based on 17 duplicates).

Results and Discussion

FIGURE 1. Sample locations (A) and δ53Cr results (B). The inset in part A gives detail near RTC. Black and gray stars give locations of waste injection well and percolation ponds, respectively. The numbers in part A indicate well identifiers with prefix “USGS” or “Middle,” except where other prefixes are given. Circles and triangles indicate single-level and multilevel wells, respectively, sampling the regional aquifer. Squares indicate wells sampling a localized perched aquifer. Filled symbols indicate Cr(VI) concentration >10 µg/L. δ53Cr values of multilevel wells are listed in order from deepest to shallowest. INL facilities: RTC, Reactor Technology Complex; NRF, Naval Reactor Facility; INTEC, Idaho Nuclear Technology & Engineering Center; CFA, Central Facilities Area; RWMC, Radioactive Waste Management Complex.

Cr(VI) Concentrations. The isotope dilution method provided precise results, even at the lowest concentrations encountered; this data set provides the only extensive, precise, and recent survey of Cr(VI) concentrations at INL. Figure 1 identifies wells with >10 µg/L Cr(VI), and Figure 2 plots δ53Cr vs Cr(VI) concentration. Concentrations are listed in Table S1 (Supporting Information), along with maximum values from previous studies (23-25). Wells upgradient of the known Cr(VI) contamination sources include USGS 15, USGS 134 (multilevel; five depths), USGS 133 (multilevel; four depths), TRA 1, and TRA 4. These wells had 2007-08 Cr(VI) concentrations ranging from 1.9 to 7.7 µg/L with a mean and standard deviation of 5.2 and 1.7 µg/L, respectively. Where available, historical data indicate a similar range. Apparently, natural processes in the aquifer produce Cr(VI) and maintain concentrations within the range of these 12 samples. These concentrations are within the range reported in a survey of natural groundwaters (26), in which concentrations vary from 17 µg/L] downgradient samples have δ53Cr values much greater than those of the extreme case mixing models; these are M1SA, USGS 111, and the sample from the shallowest level of USGS 132. In these samples, elevated δ53Cr values cannot be explained by reasonable scenarios of mixing with natural background and can thus be attributed to significant Cr(VI) reduction. Two other samples, from USGS 38 and 127, also have δ53Cr values significantly greater than those allowed by our mixing model, though only by a few tenths of one per mil. This suggests

minor Cr(VI) reduction. USGS 6, the upgradient well with a very high Cr(VI) concentration, has an elevated δ53Cr value relative to the mixing models. This suggests Cr(VI) reduction, but we have less confidence in this conclusion, as the source of the contaminant and its δ53Cr value are unknown (see discussion below). Many other samples have δ53Cr values substantially greater than that of the initial contaminant, but less than that of the extreme case mixing models. These mildly elevated values may reflect minor reduction of Cr(VI), but could also result from mixtures of contaminant and high-δ53Cr natural Cr(VI). Location of Reducing Zones. Further evidence for Cr(VI) reduction appears in plots of δ53Cr vs depth in the six multilevel wells (Figure 3). In four of these wells, δ53Cr is elevated (>1.5‰) near the water table; δ53Cr values are less in deeper levels. In three of these wells, the increases in δ53Cr are accompanied by decreases in concentration (Figure 3), as would be expected if reduction is occurring. In the fourth (USGS 132), the water is contaminated (17.4 µg/L) and the concentration profile is thus controlled by plume heterogeneity. Dissolved O2 measurements taken when our samples were obtained (30) suggest O2 is also reduced just below the water table. In all four wells showing elevated δ53Cr values in the shallowest samples, dissolved oxygen in these samples is significantly less than mean values for deeper levels (by 39%, 25%, 47%, and 25% for USGS 133, Middle 2051, USGS 103, and USGS 132 respectively; Supporting Information, Table S1). Taken together, the data strongly suggest that reduction of Cr(VI) occurs close to the water table, but is weaker at greater depth. This greater reactivity of the shallower basalts is plausible in light of their younger ages and shorter weathering history. Each basalt flow was formed at the surface and was progressively buried by subsequent flows to eventually become part of the saturated zone. We suggest that basalts near the water table likely have greater Cr(VI) reducing capacity due to their relatively brief exposure to saturated zone waters. Slow solute transport in the unsaturated zone may limit the rate of weathering there; rock that has only recently become part of the saturated zone would then be relatively fresh. Deeper strata, exposed longer to fast-flowing waters that maintain stronger chemical disequilibrium, may have accumulated more weathering products that retard Cr(VI) interaction with Fe(II) in the rock. Alternatively, influx of electron donors from the unsaturated zone or cyclic chemical and/or physical changes caused by rise and fall of the water table over time may accelerate Cr(VI)-rock interaction immediately below the water table. Alternative Interpretations. We considered two alternative models for the observed elevated δ53Cr values: First, highδ53Cr water descending from the unsaturated zone could be invoked. However, the study area is semiarid, with 21 cm annual precipitation and little recharge resulting from local precipitation. The only significant source of recharge is the Big Lost River, which loses most or all of its flow prior to reaching INL but sporadically flows through the study area in wet spring seasons. Water table elevation increases of a few meters have been observed near the river after a series of wet years (19). Although this is a significant influx of water, our data are not consistent with a scenario in which observed elevated δ53Cr values are caused by infiltrating river water. First, the regional hydraulic gradient does not move Big Lost River water toward USGS 103, so it is very difficult to explain its elevated δ53Cr values this way. Second, downgradient wells USGS 111, M1SA, and the shallowest level of USGS 132 contained elevated Cr(VI) concentrations (29.6, 29.7, and 17.4 µg/L, respectively). It is very unlikely that the Big Lost River could produce such high Cr(VI) concentrations, and it is even more unlikely that it could produce such high concentration VOL. 45, NO. 2, 2011 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

505

waters in some areas and much lower concentrations in other areas (e.g., near Middle 2051) at the same time. A second alternative interpretation invokes additional contaminant sources. Our results from USGS 6 suggest the existence of one other source. Perhaps other centers of activity, such as the RWMC facility, contain other sources. As described above, industrial Cr(VI) has near-zero δ53Cr values, and our data indicate similar values for the known sources at INL. Nonetheless, Cr(VI) from a currently unknown source could possibly have a greater δ53Cr value, as a result of reduction by organic cocontaminants in ponds or soils. This could generate elevated aquifer δ53Cr values in the regional aquifer without any reduction there. However, as the aquifer’s water flux is extremely large, these sources would necessarily be extremely large, like the two known sources (15). Previously unidentified sources are unlikely to exist, but we cannot definitively prove they do not. Overall, the δ53Cr results suggest the existence of limited Cr(VI) reduction, but given the complexities of this system, definitive proof of reduction would require further work. Perched Water Zones. Of the three archived perched water samples from 1967 with high Cr concentrations, two have low δ53Cr values. The third value, from USGS 55, is somewhat greater (1.13‰); minor subsurface Cr(VI) reduction may have occurred. Alternatively, some Cr(VI) reduction may have occurred during the prolonged sample storage. The six recent samples’ δ53Cr values range from 0.55‰ to 2.91‰. The greatest δ53Cr value is from well PW-8, which has a low Cr(VI) concentration (3.2 µg/L), perhaps related to Cr(VI) reduction, dilute recharge from recent operations of the ponds, or both. The other five samples have δ53Cr values of 0.55‰-1.44‰, and two of these plot above the mixing lines in Figure 2. However, we hesitate to suggest reduction has occurred because our data from the perched zone are sparse, its hydrology is complex, and we cannot define a background and a mixing model. We can, however, conclude that strong reduction has not occurred except perhaps in well PW-8. Estimating the Extent of Cr(VI) Reduction. Quantification of the extent of Cr(VI) reduction at INL using the δ53Cr data cannot be done precisely, but semiquantitative estimates provide some insight into reaction rates. The simplest model relating the δ53Cr of Cr(VI) to the extent of reduction is a Rayleigh distillation model. A close approximation to this model is δ ≈ δ0 - ε ln(f)

(2)

where δ is the measured δ53Cr value, δ0 is the initial value prior to any reduction, and f is the fraction of the original Cr(VI) left unreduced. ε is known to depend on the reaction mechanism (11-13, 27, 28). The Cr reduction mechanism at this site is unknown, so ε could be anywhere within the range reported for laboratory experiments (1.8‰-4.4‰) or even outside that range if the mechanism is not yet studied. However, the extent of reduction can be estimated roughly by examining the range of results obtained using minimum and maximum observed values for ε. We use well M1SA as an example of an elevated δ53Cr value from a contaminated well. Inserting its δ53Cr value, 1.40‰, an initial δ53Cr of 0.30‰, and the minimum ε value observed in laboratory experiments (1.8‰) into eq 2, we calculate the fraction Cr(VI) remaining as 0.54. This corresponds to 46% loss of the original contaminant. Repeating the calculation, but using the maximum ε value, 4.4‰, we estimate 22% loss. Similar calculations for USGS 111 and the shallowest samples from USGS 132 and Middle 2051 yield estimates ranging from 18% to 77% reduced. This approach can also be used to interpret low δ53Cr values. Of the 12 downgradient wells with greater than 9 µg/L Cr(VI), seven have δ53Cr values less than 0.70‰, which translates to a maximum Cr(VI) loss of 20%. 506

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 45, NO. 2, 2011

There are known inaccuracies with application of Rayleigh models, which are rigorously defined for closed systems, to groundwater systems. Water masses are not closed, as they are subject to dispersive mixing. Abe and Hunkeler (31) reported that Rayleigh-based determinations of the extents of chemical reactions underestimate true extents by several percent. Furthermore, Berna et al. (13) suggest that in cases where Cr(VI) reduction occurs in semi-isolated zones linked by diffusion to zones of advective transport, Rayleigh calculations underestimate the extent of reduction more strongly. This phenomenon may apply to the ESRPA, where Cr(VI) may occur in reducing microenvironments connected to the main groundwater mass by diffusion, but we cannot quantitatively address this issue at present. Overall, our estimates of the extent of reduction at this site are semiquantitative. Better knowledge of the Cr(VI) reduction mechanism and the magnitude of isotopic fractionation induced by it would be needed to obtain more precise estimates. The calculations above suggest Cr(VI) reduction is significant, but not near completion, along flow paths within a thin zone of the aquifer just below the water table. Along deeper flow paths, low δ53Cr values indicate little or no Cr(VI) reduction has occurred. Isotopic Variability in Natural Chromium. In wells located upgradient of the contaminant sources, δ53Cr values of natural Cr(VI) range from 0.79 to 2.42‰. Izbicki et al. (4) found similar variation (δ53Cr ) 0.7 to 5.1‰) in natural dissolved Cr(VI) in alluvial deposits near Cr-rich ultramafic rock sources in the Mojave desert. Cr in igneous rocks has near-zero δ53Cr values (22), and the Snake River Plain basalts almost certainly follow suit. Thus, the natural Cr(VI) in ESRPA groundwater was isotopically fractionated either as it was released during weathering processes or via later reduction. Recent studies (32, 33) suggest that oxidation of aqueous Cr(III) by manganese oxides can produce Cr(VI) with 53Cr/ 52 Cr roughly 1‰ greater than the reactant Cr(III). Thus, weathering might produce Cr(VI) with elevated δ53Cr values, though this issue is not yet settled. Alternatively, some or all of the increase in δ53Cr could happen after Cr(VI) is delivered to the water, via partial reduction by Fe(II)-bearing solids or bacteria. We suggest that this is indeed the case, as our results suggest a thin zone of reduction just below the water table. Notably, the greatest δ53Cr value observed among the upgradient, uncontaminated samples (2.45‰; USGS 133 zone 4) was obtained from a sample taken just below the water table. We expect the variability in natural δ53Cr values observed here and by Izbicki et al. (4) will be observed at other sites, and we suggest that the influence of natural background Cr should be considered when interpreting δ53Cr data from contaminant plumes, unless the concentration is high enough to clearly overwhelm the background. With background concentrations rarely exceeding 10 µg/L (26), δ53Cr values of contaminated water exceeding the U.S. EPA MCL (100 µg/L) should be affected little by mixing with background waters. However, Cr-rich ultramafic rock weathering can produce background concentrations as high as 60 µg/L (4). Also, δ53Cr analyses of waters with concentrations well below the MCL may be very helpful in assessing Cr(VI) reduction outside of heavily contaminated areas. In these cases, background Cr should be considered.

Acknowledgments This material is based upon work supported by the U.S. Department of Energy under Grant No. DE-FG02-07ER64405 and the National Science Foundation under Grant No. EAR 0732481. Suggestions from three anonymous reviewers improved the quality of this paper.

Supporting Information Available Details of experimental methods, including MC-ICP-MS and concentration analysis procedures, and Cr(VI) concentration and δ53Cr data. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Katz, S. A.; Salem, H. The Biological and Environmental Chemistry of Chromium; VCH Publishers: New York, 1994. (2) Robles-Camacho, J.; Armienta, M. A. Natural chromium contamination of groundwater at Leon Valley, Mexico. J. Geochem. Explor. 2000, 68, 167–181. (3) U.S. Environmental Protection Agency. Toxicological Review of Hexavalent Chromium; U.S. EPA National Center for Environmental Assessment: Washington, DC, 1998. (4) Izbicki, J. A.; Ball, J. W.; Bullen, T. D.; Sutley, S. J. Chromium, chromium isotopes and selected trace elements, western Mojave Desert, USA. Appl. Geochem. 2008, 23, 1325–1352. (5) Davis, A.; Olsen, R. L. The geochemistry of chromium migration and remediation in the subsurface. Ground Water 1995, 33, 759–768. (6) Pettine, M.; D’Ottone, L.; Campanella, L.; Millero, F. J.; Passino, R. The reduction of chromium(VI) by iron(II) in aqueous solutions. Geochim. Cosmochim. Acta 1998, 62, 1509–1519. (7) Viamajala, S.; Peyton, B. R.; Apel, W. A.; Petersen, J. N. Chromate reduction in Shewanella oneidensis MR-1 is an inducible process associated with anaerobic growth. Biotechnol. Prog. 2002, 18, 290–295. (8) Horton, R. N.; Apel, W. A.; Thompson, V. S.; Sheridan, P. P. Low temperature reduction of hexavalent chromium by a microbial enrichment consortium and a novel strain of arthrobacter aurescens. BioMed Central Microbiol. 2006, 6, 5. (9) Desjardina, V.; Bayarda, R.; Hucka, Manceaub, A.; Gourdona, R. Effect of microbial activity on the mobility of chromium in soils. Waste Manag. 2002, 22, 195–200. (10) Eary, L. E.; Rai, D. Kinetics of chromate reduction by ferrous iron derived from hematite and biotite at 25°C. Am. J. Sci. 1989, 289, 180–213. (11) Johnson, T. M.; Bullen, T. D. Mass-dependent fractionation of selenium and chromium isotopes in low-temperature environments. In Geochemistry of non-traditional stable isotopes; Johnson, C. M., Beard, B. L., Albarede, F., Eds.; Mineralogical Society of America: Washington, DC, 2004; pp 289-317. (12) Ellis, A. S.; Johnson, T. M.; Bullen, T. D. Chromium isotopes and the fate of hexavalent chromium in the environment. Science 2002, 295, 2060–2062. (13) Berna, E. C.; Johnson, T. M.; Makdisi, R. S.; Basu, A. Cr stable isotopes as indicators of Cr(VI) reduction in groundwater: A detailed time-series study of a point-source plume. Environ. Sci. Technol. 2010, 44, 1043–1048. (14) Gao, Y. J.; Ma, T.; Ling, W. L.; Liu, C. F.; Li, L. Analytical method of Cr stable isotope and its application to water pollution survey. Chin. Sci. Bull. 2010, 55 (7), 664–669. (15) Pittman, J. R.; Jensen, R. G.; Fischer, P. R. Hydrologic conditions at the Idaho National Engineering Laboratory, 1982 to 1985; Water Resources Investigations Report 89-4008; U.S. Geological Survey: Reston, VA, 1988. (16) Mitchell, R.; Roddy, M.; Wood, T.; Knobel, L.; Finnerty, M.; Lowry, S. Chapter 6: Environmental Monitoring Programs (Groundwater, Drinking Water, and Surface Water). In INEEL Annual Site Environmental Report-2004 DOE/ID-12082(04); U.S. Department of Energy: Washington, DC, 2004.

(17) Smith, R. P. Geologic setting of the Snake River Plain Aquifer and vadose zone. Vadose Zone J. 2004, 3, 47–58. (18) Lanphere, M. A.; Champion, D. E.; Kuntz, M. A. Petrography, age, and paleomagnetism of basalt lava flows in coreholes Well 80, NRF 89-04, NRF 89-05, and ICPP-123, Idaho National Engineering Laboratory; Open-file Report 93-327; U.S. Geological Survey: Menlo Park, CA, 1993. (19) Bennett, C. M., Streamflow losses and ground-water level changes along the Big Lost River at the Idaho National Engineering Laboratory, Idaho, Water-Resources Investigations Report 904067, U.S. Geological Survey: Reston, VA, 1990. (20) U.S. Department of Energy Idaho Operations Office. Five-Year Review of CERCLA Response Actions at the Idaho National Laboratory; Report DOE/NE-ID-11201; U.S. Department of Energy: Idaho Falls, ID, 2005. (21) U.S. Environmental Protection Agency. Record of Decision, Test Reactor Perched Water System, Operable Unit 2-12, Idaho National Engineering Laboratory; ROD/R10-93/057; EPA: Washington, DC, 1992. (22) Schoenberg, R.; Zink, S.; Staubwasser, M.; von Blanckenburg, F. The stable Cr isotope inventory of solid Earth reservoirs determined by double spike MC-ICP-MS. Chem. Geol. 2008, 249, 294–306. (23) U.S. Geological Survey National Water Information System: Web Interface; Available at http://nwis.waterdata.usgs.gov/usa/nwis/ qwdata. (24) Bartholomay, R. C. Distribution of selected radiochemical and chemical constituents in perched ground water, Idaho National Engineering Laboratory, Idaho, 1992-95; Water-Resources Investigation 98-4026; U.S. Geological Survey: Reston, VA, 1998. (25) Liszewski, M. J.; Mann, L. J. Concentration of 23 trace elements in ground water and surface water at and near the Idaho National Engineering Laboratory, Idaho, 1988-91; Open-File Report 93126; U.S. Geological Survey: Reston, VA, 1993. (26) Allard, B., Groundwater. In Trace Elements in Natural Waters, Aalbu, B., Steinnes, E., Eds.; CRC Press: Boca Raton, FL, 1995; pp 151-176. (27) Sikora, E. R.; Johnson, T. M.; Bullen, T. D. Microbial massdependent fractionation of chromium isotopes. Geochim. Cosmochim. Acta 2008, 72, 3631–3641. (28) Kitchen, J. W.; Johnson, T. M.; Bullen, T. D. Chromium stable isotope fractionation during abiotic reduction of hexavalent chromium. Eos Trans. AGU; 2004; Vol. 85, Fall Meet. Suppl. Abstract V51A-0519. (29) Ellis, A. S.; Johnson, T. M.; Bullen, T. D. Using chromium stable isotope ratios to quantify Cr(VI) reduction: Lack of sorption effects. Environ. Sci. Technol. 2004, 38, 3604–3607. (30) Bartholomay, R. C.; Twining, B. V. Chemical constituents in groundwater from multiple zones in the snake river plain aquifer at the Idaho National Laboratory, Idaho, 2005-08; Scientific Investigations Report 2010-5116; U.S. Geological Survey: Reston, VA, 2010. (31) Abe, Y.; Hunkeler, D. Does the Rayleigh equation apply to evaluate field isotope data in contaminant hydrogeology? Environ. Sci. Technol. 2006, 40, 1588-1596. (32) Ellis, A. S.; Johnson, T. M.; Villalobos-Arago´n, A.; Bullen, T. D. Environmental cycling of Cr using stable isotopes: Kinetic and equilibrium effects. EOS Trans. AGU; 2008; Vol. 89, Fall Meet. Suppl. Abstract-H53F-08. (33) Bain, D. J.; Bullen, T. D. Chromium isotope fractionation during oxidation of Cr(III) by manganese oxides. Geochim. Cosmochim. Acta 2005, 69 (10, Suppl.), A212.

ES102000Z

VOL. 45, NO. 2, 2011 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

507