Zn and Cu Isotopes as Tracers of Anthropogenic ... - ACS Publications

Feb 9, 2010 - Mountlake Terrace, a suburb of Seattle, Washington, USA. The base of the core predates European settlement in the region, while the uppe...
0 downloads 0 Views 2MB Size
Environ. Sci. Technol. 2010, 44, 1544–1550

Zn and Cu Isotopes as Tracers of Anthropogenic Contamination in a Sediment Core from an Urban Lake A N I T A T H A P A L I A , † D A V I D M . B O R R O K , * ,† PETER C. VAN METRE,‡ MARYLYNN MUSGROVE,‡ AND EDWARD R. LANDA§ Department of Geological Sciences, University of Texas at El Paso, El Paso, Texas 79968, U.S. Geological Survey, 8027 Exchange Drive, Austin, Texas 78754, U.S. Geological Survey, MS 430, 12201 Sunrise Valley Drive, Reston, Virginia 20192

Received April 23, 2009. Revised manuscript received December 30, 2009. Accepted January 29, 2010.

In this work, we use stable Zn and Cu isotopes to identify the sources and timing of the deposition of these metals in a sediment core from Lake Ballinger near Seattle, Washington, USA. The base of the Lake Ballinger core predates settlement in the region, while the upper sections record the effects of atmospheric emissions from a nearby smelter and rapid urbanization of the watershed. δ66Zn and δ65Cu varied by 0.50‰ and 0.29‰, respectively, over the 500 year core record. Isotopic changes were correlated with the presmelter period (∼1450 to 1900 with δ66Zn ) +0.39‰ ( 0.09‰ and δ65Cu ) +0.77‰ ( 0.06‰), period of smelter operation (1900 to 1985 with δ66Zn ) +0.14 ( 0.06‰ and δ65Cu ) +0.94 ( 0.10‰), and postsmelting/stable urban land use period (post 1985 with δ66Zn ) 0.00 ( 0.10‰ and δ65Cu ) +0.82‰ ( 0.12‰). Rapid early urbanization during the post World War II era increased metal loading to the lake but did not significantly alter the δ66Zn and δ65Cu, suggesting that increased metal loads during this time were derived mainly from mobilization of historically contaminated soils. Urban sources of Cu and Zn were dominant since the smelter closed in the 1980s, and the δ66Zn measured in tire samples suggests tire wear is a likely source of Zn.

Introduction Zinc (Zn) and Copper (Cu) emissions have increased dramatically over the past decade in developing countries (1), and have continued to increase in urban areas of the United States and Europe despite a three decade-long trend of decreasing concentrations for most other trace metals (2-4). Sediment cores from standing water bodies serve as time-resolvable records for these trace metals inputs (4-7), however, the effectiveness of sediment core investigations can be limited because concentration data alone are often insufficient to resolve metal sources. The application of metal isotopes like Pb to sediment core records has become an effective technique for identifying and tracking anthropogenic sources (8, 9). More recently, this approach has been adopted for isotopic systems like Zn (10-12). Zinc isotopes in * Corresponding author phone: (915) 747-5850; fax: (915) 7475073; e-mail: [email protected]. † Department of Geological Sciences, University of Texas at El Paso. ‡ U.S. Geological Survey, Austin, Texas. § U.S. Geological Survey, Reston, Virginia. 1544

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 5, 2010

terrestrial samples range from -0.4 parts per thousand (‰) to +1.4‰ (δ66Zn), but values lighter than 0.0‰ are rare and most natural samples cluster between 0.2‰ to 0.5‰ (13). Investigations of slag, sediments, and aerosols associated with Zn smelting and refining have shown that the lighter Zn isotopes are preferentially incorporated into atmospheric emissions, while tailing and slag products are enriched in the heavier Zn isotopes (11, 14-16). Urban runoff also has been shown to produce distinctly light δ66Zn values (12, 15). These studies demonstrate that Zn isotopes can be a reliable tracer of Zn sources. Because Cu isotopes are fractionated substantially during redox, precipitation, and complexation reactions (17, 18), the range of isotopic variation of Cu in natural samples is ∼6× that of Zn (18). Several investigations of historical currency have used Cu isotopes to identify ore deposit sources (19) and to track ancient trading routes (20). However, to date, we can find no record of Cu isotopic measurements being used as a tool for tracing metal pollution. In this work, we tested whether Cu and Zn isotopes could be used to identify the sources and timing of the deposition of these metals in a sediment core from Lake Ballinger, in Mountlake Terrace, a suburb of Seattle, Washington, USA. The base of the core predates European settlement in the region, while the upper sections record the effects of atmospheric emissions from a nearby smelter and rapid urbanization of the watershed. Because tire tread material has a Zn content of about 1 wt % (21), tire wear from the nearby highways and roads is a likely source of urban Zn. For this reason, we additionally analyzed the Zn isotopes of a set of tire samples. Zinc and Cu isotopic data were also collected for the exchangeable, organic, and acid-labile reservoirs of the sediment core to help determine how these metals were transported and sequestered.

Location and History of Study Area Lake Ballinger and the Tacoma Smelter. Lake Ballinger, a 0.4 km2 waterbody with and average depth of 5 m, is situated in a heavily urbanized, 13.7 km2 watershed north of Seattle (Figure 1). Residential development of the watershed began soon after World War II. Today, the watershed is densely urbanized with 84% classified as urban land use (22). Elevated Cu and Zn loads to Lake Ballinger, possibly associated with dense urbanization, were described in an earlier investigation (4). The ASARCO-Tacoma smelter (now the Ruston/North Tacoma Superfund Site) is situated ∼53 km south of Lake Ballinger ( (22) Figure 1). The smelter started operation in 1890 as a Pb smelter and was converted to a Cu smelter (and As-production facility) in 1905. A 172-m exhaust stack was added to the smelter in 1917, and a sulfur control system was added in 1950. The total amounts of smelted materials averaged 300 000 to 400 000 tons annually, but declined after 1970 until closure of the smelter in 1985 (23). The wind direction in the study area is usually south to north and the larger portion of the smelter emissions to the atmosphere traveled downwind from the stack contaminating a large part of the Seattle area, including Lake Ballinger (24, 25).

Methods Collection and Analysis of Sediment Core. On August 14, 2007, a 6.3-cm inside diameter free-fall gravity corer was used to collect a 50-cm tall core. The core was vertically extruded in the field and sectioned into 2-cm intervals from 0-20 cm depth and 3-cm intervals from 20-50 cm. Samples were chilled or frozen and then later freeze-dried and powdered in preparation for the analysis of 137Cs and 210Pb 10.1021/es902933y

 2010 American Chemical Society

Published on Web 02/09/2010

FIGURE 1. Location of Lake Ballinger. The Seattle, Washington, metro area runs between Puget Sound and Lake Washington and includes Lake Ballinger. Closed circles represent sample location sites in Lake Ballinger or Hall Creek. I-5 ) Interstate highway 5. (for age dating), major and trace elements, and Cu and Zn isotopes. Samples of stream-bed sediments from Hall Creek (the main tributary to Lake Ballinger) were collected in prewashed glass jars at three separate locations from the upper part of the watershed to near the confluence with Lake Ballinger (Figure 1). Fine-grained sediment in depositional zones was targeted for sampling. Elemental concentrations were determined on concentrated-acid (HCl-HNO3-HClO4-HF) digests (“near-total” digestions) by inductively coupled plasma-mass spectrometry (ICP-MS) at the USGS laboratories in Denver, Colorado (26). Cesium-137 activities were counted on freeze-dried samples in fixed geometry with a high-resolution intrinsic germanium detector γ-spectrometer, using a method similar to that described by 27. Lead-210 activities were measured by high-precision γ-ray spectrometry (28). Tire Samples. Five tire samples (some composite samples of scrap tire rubber) were collected from industrial and commercial sources. The tire samples were processed at the USGS in Denver, Colorado, using a microwave-assisted acid digestion procedure (EPA Method 3052). The digested samples were first analyzed to determine elemental concentrations using the aforementioned ICP-MS technique and were later processed for isotopic analysis as described below. A description of the tire samples is presented in Table SI-1 of the Supporting Information. Selective Sequential Dissolution (SSD) of Sediment Core Samples. Splits of freeze-dried samples from the sediment core were subjected to an SSD procedure to determine the proportions and isotopic ratios of Zn and Cu occurring in the operationally defined exchangeable, organic, and acidlabile fractions. The purpose of our SSD analyses was to test whether the isotopes of Zn and Cu in these fractions changed in relationship to the bulk isotopic shifts associated with different contaminant sources. If so, this could add to our understanding of the transport and depositional processes responsible. The SSD methods, modified from (29), are summarized in Table SI-2 of the Supporting Information. Isotopic Analysis. Samples were prepared for isotopic analysis using a column chromatography technique described by ref 30. All of the column separates including

duplicates and blanks were first analyzed using an ICP optical emission spectrometer to ensure that matrix elements were cleanly separated and that Cu and Zn were quantitatively recovered. Despite multiple attempts, Cu could not be cleanly separated from several of the presmelter samples that contained low concentrations of Cu. These separates were not analyzed. Cu and Zn isotopes were analyzed at the USGS laboratories in Denver, Colorado, and at the University of Texas at El Paso using a Nu Instruments MC-ICP-MS. Samples were introduced using a desolvating nebulizer system (Nu DSN 100) and the standard-sample-standard bracketing method was used to correct for mass bias (30). For Zn, 64Zn, 66Zn, 67 Zn, and 68Zn were measured simultaneously and 63Cu and 65 Cu were measured simultaneously for Cu. The isotopic ratios of the unknowns for the Zn and Cu systems were respectively referenced to the Johnson Matthey zinc standard (batch JMC 3-0749-L) and the SRM 976 (NIST) Cu standard (standards also were passed through the column chemistry). Isotopic ratios are expressed in standard delta notation relative to the average value of the respective bracketing standards for each metal (Me) as shown in eq 1 δaMe )

[

(aMe/ bMe)sample (aMe/bMe)standard

]

- 1 × 1000

(1)

where a/b is 6X/64 (where X ) 66, 67, or 68) and 65/63 for the Zn and Cu systems, respectively. The 2σ external precision reported in all figures was calculated by measuring the same samples multiple times over many analytical sessions. Separation chemistry duplicates were analyzed and treated as replicates to encompass uncertainties originating from the separation procedure. The 2σ precision for Zn and Cu isotopes was (0.07 ‰, and (0.09 ‰, respectively. Some SSD and tire samples were not replicated and we assign the average 2σ uncertainty values to these samples. Figure SI-1 of the Supporting Information illustrates the expected mass dependent fractionation behavior for the Zn isotopes. VOL. 44, NO. 5, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1545

FIGURE 2. Sediment MAR (g/cm2-yr) plus Al and organic carbon concentrations vs calculated age.

Results and Discussion Sediment Mass Accumulation Rates (MARs). The sediment core was dated by applying the constant rate of supply (CRS) model to the 210Pb profile (31) following the approach described in ((32), also see the Supporting Information). The CRS model indicates a low mass accumulation rate (MAR) (