Controls on tungsten concentrations in groundwater flow systems: The role of adsorption, aquifer sediment Fe(III) oxide/oxyhydroxide content, and thiotungstate formation

Highlights

• Tungsten varies with changing pH and redox conditions.

• Tungsten concentrations correlate with pH, alkalinity, and dissolved sulfide.

• Sulfate reduction corresponds to a critical zone of enhanced W mobilization.

• Increasing pH and decreasing Fe(III) oxide content favor W mobilization.

• Thiotungstate complexes may be important in anoxic groundwater.

Abstract

Groundwater samples were collected along flow paths within the Carrizo Sand aquifer (southeastern Texas) and the Aquia aquifer (coastal Maryland) for analysis of dissolved tungsten (W) concentrations, along with the major solutes, pH, and measures of in situ redox conditions [e.g., dissolved Fe(II), Fe(III), and S(-II) concentrations]. In addition, sediment samples were collected from both aquifers to evaluate the solid-phase speciation of W. Tungsten concentrations in the Carrizo Sand aquifer range from 3.64 to 1297 pmol kg^− 1 (mean ± SD = 248 ± 440 pmol kg^− 1), with the lowest concentrations reported from the recharge area. Tungsten concentrations progressively increase down gradient along the flow path within Carrizo Sand aquifer groundwaters, reaching the highest levels in sulfidic groundwater roughly 50–60 km from the recharge zone. Tungsten is strongly correlated with pH and dissolved sulfide [i.e., S(-II)] concentrations in Carrizo groundwaters (r = 0.78 and 0.95, respectively). In Aquia aquifer groundwaters W concentrations range between 14.3 and 184.4 pmol kg^− 1 (mean ± SD = 63.9 ± 59 pmol kg^− 1), and exhibit no, or at best, weak relationships with other geochemical parameters measured along the flow path (e.g., r = 0.4 and 0.08 for W vs. pH and S(-II), respectively). The data suggest that the increase in W concentrations in Carrizo groundwaters reflects, in part, pH-related adsorption/desorption, which has been shown to be substantial for groundwaters and lake waters with pH > 8. Owing to the similar chemical properties of W and Mo, which form thiomolybdates in sulfidic waters, the formation of thiotungstate complexes may also be important in the sulfidic waters of the Carrizo aquifer. The substantially lower W concentrations in Aquia aquifer groundwaters reflect the fact that the maximum groundwater pH never exceeds 8.4 (mean ± SD pH = 7.97 ± 0.23), dissolved low S(-II) concentrations remain low (mean ± SD S(-II) = 0.29 ± 0.16 μmol kg^− 1), and that Aquia aquifer sediments have substantially higher Fe(III) oxide/oxyhydroxide contents compared to the Carrizo Sand aquifer. Our data indicate that pH-related adsorption/desorption reactions and the Fe(III) oxide/oxyhydroxide content of aquifer sediments are key controls affecting W concentrations in oxic, suboxic, and anoxic groundwaters, and, further, that the formation of thiotungstate species may also be important in some anoxic (i.e., sulfidic) waters.

Introduction

A number of recently recognized clusters of childhood leukemia, including clusters in Fallon, Nevada, Sierra Vista, Arizona, and Elk Grove, California, are tentatively linked to the proximity of these sites to tungsten (W)-bearing ore deposits, associated active or inactive smelting operations, and/or hard-metal processing facilities (Seiler et al., 2005, Koutsospyros et al., 2006, Sheppard et al., 2006, Sheppard et al., 2007, Bednar et al., 2007, Bednar et al., 2008). Indeed, the childhood leukemia cluster afflicting Fallon, Nevada, spawned an investigation by the United States Centers for Disease Control and Prevention (CDC), which revealed through urine analysis that residents were exposed to elevated levels of W (Seiler et al., 2005, Koutsospyros et al., 2006). Although it is inherently difficult to directly link these childhood leukemia clusters to environmental exposure to elevated W concentrations, either via inhalation or through consumption of drinking waters with high W concentrations, many studies have shown that W can be toxic and may indeed be carcinogenic (Peão et al., 1993, Marquet et al., 1996, ATSDR (Agency for Toxic Substances, Disease Registry), 2005, Kalinich et al., 2005, U.S. EPA, 2008, Kelly et al., 2013). Moreover, inhalation of W dust in the hard metal manufacturing industry is linked to pulmonary fibrosis and hard metal pneumoconiosis (Edel et al., 1990, Sprince et al., 1994, Sahle et al., 1996). Furthermore, polytungstates like sodium metatungstate (3Na₂WO₄·9WO₃) appear to be more toxic than monomeric tungstate (e.g., Na₂WO₄; Strigul et al., 2009, Strigul et al., 2010), and may thus pose problems for people consuming drinking water with elevated W concentrations. The CDC investigation led to a study by the U.S. Geological Survey of W in groundwaters of the Carson Desert region of northwest Nevada, which showed elevated W concentrations ranging from 1.47 to 4036 nmol kg^− 1 (0.27–742 μg kg^− 1; Seiler et al., 2005). These concentrations are high and similar to W concentrations previously reported for groundwaters of the Carson Desert (1029 nmol kg^− 1; Johannesson et al., 2000), and may in part be responsible for the high levels of W found in local residents.

In addition to natural sources of W to the environment, interest in the environmental geochemistry of W is also on the rise owing to its increasing use as a replacement for lead (Pb) ammunition for hunting and recreational shooting, for fishing weights, and for military ammunition (i.e., high kinetic energy penetrators and small arms ammunition; Strigul et al., 2005, Koutsospyros et al., 2006, Bednar et al., 2007, Clausen et al., 2007, Bednar et al., 2008, Clausen and Korte, 2009). Here, the use of W was originally conceived as a way to limit the addition of toxic Pb to the environment by what was thought to be a non-toxic, inert metal of low environmental mobility (Koutsospyros et al., 2006, Bednar et al., 2007). In fact, the U.S. Army's “Green Armament Technology” (GAT) program initially sought to limit environmental pollution by recommending the replacement of Pb-based ammunition with W-based rounds (Koutsospyros et al., 2006). Nevertheless, despite the initial beliefs that W was a non-toxic, chemically inert heavy metal, laboratory studies clearly reveal that W can be toxic and carcinogenic (e.g., Marquet et al., 1996, Sahle et al., 1996, Kalinich et al., 2005, Steinburg et al., 2007, Strigul et al., 2009, Strigul, 2010, Strigul et al., 2010, Kelly et al., 2013), and field-based investigations indicate that it is mobile in the environment (e.g., Johannesson et al., 2000, Dermatas et al., 2004, Petrunic and Al, 2005, Seiler et al., 2005, Dermatas et al., 2006, Arnórsson and Óskarsson, 2007, Clausen et al., 2007, Dave and Johannesson, 2008a, Dave and Johannesson, 2008b, Bednar et al., 2009, Clausen and Korte, 2009, Johannesson and Tang, 2009).

In general, W exhibits similar geochemical behavior to other oxyanion-forming trace elements such as molybdenum (Mo) and arsenic (As) in the environment in that it is strongly sorbed to Fe(III) oxides/oxyhydroxides and other mineral surfaces in oxic waters of low to circumneutral pH, and desorbs from these mineral surface sites at higher pH (Johannesson et al., 2000, Gustafsson, 2003, Koschinsky and Hein, 2003, Seiler et al., 2005, Arnórsson and Óskarsson, 2007, Bednar et al., 2007, Bednar et al., 2008, Bednar et al., 2009, Johannesson and Tang, 2009, Kashiwabara et al., 2013). At low pH (pH < 5) and high dissolved W concentrations (W ≥ 10^− 2 mol kg^− 1), W tends to form polymerized species in solution, whereas monomers such as the tungstate oxyanion, i.e., WO₄^2 −, predominate at low W concentrations, neutral to alkaline pH, high temperatures, and high ionic strength (Wesolowski et al., 1984, Wood, 1992, Bednar et al., 2007, Rodríguez-Fortea et al., 2008). Despite these observations, little is actually known about the biogeochemistry of W in low-temperature, groundwater flow systems, which are important drinking water sources for much of the world's population (Solley et al., 1993, Clarke and King, 2004). To the best of our knowledge, there are no published studies that have investigated how biogeochemical reactions occurring along groundwater flow paths, including geochemical reactions that alter the solution composition and redox conditions, affect W concentrations and speciation within aquifer systems. For example, although WO₄^2 − is thought to predominate in natural waters, the similarity between W and Mo suggests that thiotungstate species may form in some anoxic, sulfidic waters (e.g., Erickson and Helz, 2000, Dave and Johannesson, 2008a, Dave and Johannesson, 2008b, Mohajerin and Johannesson, 2012).

In this contribution we present W concentrations for groundwaters collected along flow paths in two well-characterized aquifers, the Carrizo Sand (Texas) and the Aquia (Maryland) aquifers. Groundwater W concentrations are evaluated along the studied flow paths in each aquifer along with major solute concentrations, pH, and a number of redox sensitive indicators (i.e., Eh, dissolved oxygen, iron species, and sulfide) to develop a better understanding of the biogeochemical processes that control W concentrations in groundwater flow systems that have not been contaminated by anthropogenic activities.