Published online 4 January 2008
Published in J Environ Qual 37:229-233 (2008)
DOI: 10.2134/jeq2007.0305
© 2008 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
TECHNICAL REPORTS
Heavy Metals in the Environment
Geochemical Parameters Influencing Tungsten Mobility in Soils
A. J. Bednara,*,
W. T. Jonesb,
R. E. Boydb,
D. B. Ringelbergc and
S. L. Larsona
a U.S. Army Engineer Research and Development Center, Environmental Lab., 3909 Halls Ferry Rd., Vicksburg, MS 39180
b SpecPro, Inc., Huntsville, AL
c U.S. Army Engineer Research and Development Center, Cold Regions Research and Engineering Lab
* Corresponding author (Anthony.J.Bednar{at}erdc.usace.army.mil).
Received for publication June 12, 2007.
 |
ABSTRACT
|
|---|
The biogeochemistry of tungsten and its effects on mobility have recently gained attention due to the existence of human cancer clusters, such as in Fallon, NV. Tungsten exists in many environmental matrices as the soluble and mobile tungstate anion. However, tungsten can polymerize with itself and other anions, creating poly- and heteropoly-tungstates with variable geochemical and toxicological properties. In the present work, geochemical parameters are determined for tungstate species in a model soil that describe the potential for tungsten mobility. Soluble tungsten leached from a metallic tungsten-spiked soil after six to twelve months aging reached an equilibrium concentration >150 mg/L within 4 h of extraction with deionized water. Partition coefficients determined for various tungstate and polytungstate compounds in the model soil suggest a dynamic system in which speciation changes over time affect tungsten geochemical behavior. Partition coefficients for tungstate and some poly-species have been observed to increase by a factor of 3 to 6 over a four month period, indicating decreased mobility with soil aging.
|
INTRODUCTION
|
|---|
INTEREST in tungsten geochemistry and occurrence in groundwater has recently increased due to human toxicological events, specifically the childhood leukemia cancer cluster located in Fallon, NV, and suspected cases in Sierra Vista, AZ and Elk Grove, CA. Although these events are related to local natural deposits of tungsten ore (Seiler et al., 2005; Koutsospyros et al., 2006), there is increasing interest in industrial (e.g., tungsten carbide tools), civilian recreational (e.g., lead shotshell replacement), and military (e.g., kinetic penetrators and small arms ammunition) activities as well (Strigul et al., 2005; Koutsospyros et al., 2006). The use of studded tires has even been suggested to increase metal concentrations, possibly including tungsten, in soils adjacent to roadways (Backstrom et al., 2003). As a result of the potential impact of tungsten on human health, the Centers for Disease Control has investigated possible linkages in impacted areas (Seiler et al., 2005).
Tungsten metal is not found in nature, rather, the tungstate anion persists and is thermodynamically stable under most environmental conditions (Gustafsson, 2003; Seiler et al., 2005; Strigul et al., 2005; Koutsospyros et al., 2006). The polymerization of tungstate with itself and other common oxyanions (e.g., molybdate, phosphate, and silicate) can occur at higher concentrations and lower pH values, yielding a variety of ill-defined poly species (Feigl, 1958; Deery et al., 1997; Cruywagen, 2000; Gustafsson, 2003; Seiler et al., 2005). Tungstate can also replace molybdate in molybdenum-based enzymes, rendering them inactive (Johnson et al., 1974; Johnson and Rajagopolan, 1976). A comprehensive understanding of tungsten speciation and geochemistry is needed since these parameters determine sorption to soil, and therefore mobility, bioavailability, and toxicity.
Detailed geochemical modeling studies of tungstate have been conducted on pure solid phases (e.g., ferrihydrite) to obtain baseline data for comparison to other compounds of interest (e.g., molybdate), yet experimental data is limited, particularly for natural systems (Gustafsson, 2003; Ljung et al., 2006). Furthermore, natural systems often contain a variable mixture of chemical substrates and conditions (e.g., metal oxides, clays, organic matter, pH, and redox potential) that increase the system complexity compared to relatively simple laboratory experiments.
The current work investigates solubilization and sorption of various tungstate species with a model soil. Previous results indicate that in soil with pH < 6, significant amounts (up to 50+ %) of the water-soluble tungsten was present as poly-species, which were in a dynamic equilibrium (Bednar et al., 2007). Because of the ability of the soluble tungstate to polymerize over time, determination of specific geochemical parameters is difficult, except for a select number of ‘pure’ tungstate species. However, determination of ‘effective’ values (e.g., effective Kd values) is useful for describing tungsten geochemical behavior in these complex systems.
|
Experimental
|
|---|
Reagents
All chemicals used were of reagent grade or higher purity and used without further purification; the deionized water used had a resistivity of 18.3 M
cm. Sodium hydroxide, monobasic sodium phosphate, sodium phosphotungstate (2Na2O·P2O5·12WO3·18H2O), tungstosilicic acid n-hydrate, and iron (III) and aluminum oxides were purchased from Sigma Aldrich (St. Louis, MO); dibasic sodium phosphate was purchased from GFS Chemicals (Columbus, OH). Nitric acid (OmniTrace grade) was purchased from EMD Biosciences (Darmstadt, Germany); phosphoric acid and sodium sulfate were purchased from Fisher Scientific (Waltham, MA). Sodium tungstate dihydrate (Na2WO4·2H2O) and sodium polytungstate (Na6O29W12·H2O) were purchased from Alfa Aesar (Ward Hill, MA) and Fluka (Sigma Aldrich, St. Louis, MO), respectively. Single element and mixed analyte standards for tungsten and other metals were purchased from SPEX CertiPrep (Metuchen, NJ) and CPI International (Santa Rosa, CA).
Sample Collection and Preparation
All liquid samples were processed through a 0.45-µm pore size syringe filter to obtain classically defined ‘dissolved’ constituents and remove suspended soil particles. Tungsten-containing soil was created by adding metallic tungsten (<1-µm particles) to a Grenada Loring soil, mixing, and allowing it to age in plastic drums at ambient outdoor conditions for up to 1 yr. The silty loam soil of the Grenada-Loring series (Alfisols order) was collected from the Brown Loam Experimental Station, Learned, MS, with a front-end loader after the top 12 cm were removed to eliminate unwanted vegetation. Following transport to the U.S. Army Engineer Research and Development Center at Vicksburg, MS, the soil was sieved (<1 cm) and characterized, including: texture (3% sand, 72% silt, and 26% clay), total organic carbon (0.7% organic carbon), percent organic matter (1% loss on ignition), pH (6.7), cation and anion exchange capacity (0.075 and 0.025 meq g–1, respectively), and elemental content (see Table 1
).
Instrumentation
Total metal concentrations for all samples, including tungsten, were analyzed directly using nebulization inductively coupled plasma–atomic emission spectroscopy (ICP–AES) or inductively coupled plasma–mass spectroscopy (ICP–MS) as appropriate for the concentration ranges observed. Soil samples were digested following modifications of USEPA Method 3050B, with elemental concentrations determined by methods 6010B and 6020 (USEPA, 1996), using a PerkinElmer (Wellesley, MA) Optima 3000DV ICP–AES or Elan 6000 ICP–MS, respectively. Tungstate analysis was performed using an Agilent (Palo Alto, CA) 1100 HPLC interfaced to the PerkinElmer Elan 6000 ICP–MS with a cross-flow pneumatic nebulizer as previously described (Bednar et al., 2004; Bednar et al., 2007). Molecular weight cut off (MWCO) centrifuge filter devices (Millipore Amicon Ultra, Billerica, MA), with a nominal filter pore size of 5 kilodaltons (kDa), were used to separate polymeric tungsten complexes based on relative molecular size. The filtrate was then diluted and analyzed by ICP–MS or ICP–AES as above.
Redox potential was measured in a 1:1 soil/deionized water slurry with a platinum electrode and silver/silver chloride reference. The accuracy of the probe was verified using Light's solution, which yields a value of 476 mV; the Light's Solution check standard readings were within 10 mV of this value, which was deemed acceptable (Light, 1972).
Aqueous Extraction and Partition Coefficient Experiments
The deionized water extracts used for tungsten leachate studies were produced by continuously shaking a 1:10 solid/solution suspension in an HDPE container with aliquots taken between 1 and 720 h. Samples were prepared and analyzed in triplicate. The solutions were filtered to 0.45 µm and diluted for analysis by ICP–AES or ICP–MS as appropriate.
Experiments to determine the partition coefficient (Kd) used the same 1:10 (soil/solution) ratio and were spiked with various tungsten compounds at tungsten concentrations ranging from 0.3 to 100 mg/L. Partition coefficients for tungstate were determined in deionized water, 1 mmol L–1 sodium sulfate (pH = 7), 1 mmol L–1 nitric acid and sodium hydroxide (pH = 3 and 11, respectively), 1 mmol L–1 phosphate (pH = 3, 7, and 11), and 10 mg/L humic acid. Partition coefficients for the polytungstate compounds were only determined in deionized water. The solutions were continuously agitated on a reciprocal shaker and allowed to equilibrate for 3 d to 4 mo. A minimum 3 d was selected because the leach experiments previously described reached an equilibrium tungsten concentration after approximately 48 h.
|
Results and Discussion
|
|---|
The concentrations of tungsten and other metals in the parent and tungsten-spiked soils determined by ICP–AES are given in Table 1. The addition of zerovalent tungsten metal and aging for at least 6 mo decreased the soil redox potential by about 100 mV (435 mV for clean soil versus 335 mV for the tungsten soil) and the soil pH decreased by approximately 1.5 units to 5.2 as a result of the tungsten oxidation reaction (Dermatas et al., 2004). Tungstate polymerization reactions are generally favored at lower pH and higher tungsten concentrations, therefore the soil was ideal for the formation of tungstate and polytungstates (Dermatas et al., 2004; Koutsospyros et al., 2006).
Aqueous Extraction Experiments
Figure 1
shows the time-dependant water concentrations of tungsten (Graph A), calcium, potassium, sulfur, and phosphorus (Graph B) obtained for the tungsten-spiked soil. The soluble tungsten concentrations were higher in the soil that was aged longer, presumably due to the metallic tungsten having more time to oxidize. Potassium and sulfur (in the form of sulfate) behaved as conservative tracers, with potassium reaching an equilibrium concentration in about 2 h and sulfur reaching equilibrium in less than 24 h. Tungsten concentration, however, continued to increase until approximately 48 h. This behavior was closely tracked by calcium and phosphorus. It is likely that some of the tungstate in the soil was sequestered as slightly soluble calcium or other metal tungstates, and as these phases slowly dissolved, the tungstate began to polymerize with phosphate leached from the soil, effectively removing ‘tungstate’ from solution as a polymeric species. Increasing concentrations of polymeric species further changed the sorption mechanisms that influence dissolved tungsten concentrations, yielding a slight decrease in tungsten, calcium, and phosphorus concentrations at the 720 h time point, whereas potassium and sulfur remained constant.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 1. Deionized water extraction of soluble tungsten (Graph A), and calcium, sulfur, phosphorus, and potassium (Graph B) from the approximately 7000 mg/kg tungsten-spiked Grenada Loring soil. Error bars represent the standard deviation of triplicate soil extractions.
|
|
The leaching experiments were repeated using a 10% mixture of the tungsten-spiked soil with clean parent soil after both soils had aged for approximately 1 yr. This ‘blended’ soil had a total tungsten concentration of approximately 700 mg/kg. The dissolved elemental concentrations (Fig. 2
) show that tungsten concentrations again increased for the first 6 h, then decreased to a near equilibrium value after about 168 h. The co-occurring analytes are not affected as much as in the previous experiment where the higher tungsten concentration soil was used. It is likely that in the experiment shown in Fig. 1, the leached tungsten overwhelmed sorption and precipitation mechanisms.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 2. Deionized water extraction of soluble tungsten, sulfur, calcium, magnesium, phosphorus, and potassium from blended 700 mg/kg tungsten-spiked Grenada Loring soil. Error bars represent the standard deviation of triplicate soil extractions.
|
|
A liquid aliquot was sampled from the blended soil extraction experiment at 
200 h and analyzed by HPLC-ICP–MS, with the resultant chromatogram shown in Fig. 3
. The peak tail observed is indicative of polytungstate species (Bednar et al., 2007). The polymeric nature of the chromatographic peak tail was confirmed by 5 kDa MWCO filtration and elemental analysis, with the tungsten concentration dropping from 1.07 mg/L (0.45 µm filtration) to 0.36 mg/L (5 kDa filtration) at the 200 h time point, indicating the presence of polymeric tungsten species. The potassium and sulfur concentrations remained relatively unchanged in the two filter size fractions, suggesting little sorption to colloids occurred (0.88 to 0.86 mg K/L and 1.42 to 1.16 mg S/L, respectively for the 0.45 µm and 5 kDa filtrations).
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 3. HPLC-ICP–MS chromatogram of a deionized water leachate of the blended 700 mg/kg tungsten Grenada Loring soil. The tungstate peak tailing is indicative of polytungstate compounds formed in the soil and leached during the water extraction.
|
|
Partition Coefficients
To determine the effect of polymerization on sorption and thus dissolved phase concentrations, the partition coefficients for four commercially available tungsten compounds on the clean Grenada Loring soil in deionized water matrix were determined. The extraction experiments described previously indicated that a short term equilibrium tungsten concentration was reached in about 48 h, therefore, all sorption partition coefficient studies had a minimum equilibration time of 3 d, with some experiments performed to 4 mo. The Kd values determined following the linear Freundlich model with associated linear fit correlation coefficients are given in Table 2
(Langmuir, 1997; Dermatas et al., 2004).
View this table:
[in this window]
[in a new window]
|
Table 2. Partition coefficients determined for four commercially available tungsten compounds on the Grenada Loring soil.
|
|
Both tungstate and phosphotungstate compounds showed a marked increase in Kd values over the 4 mo study, whereas the polytungstate and tungstosilicate compound Kd values remained relatively unchanged. These data suggest the latter two species do not undergo significant geochemical changes in this system over the time period studied, while tungstate and phosphotungstate interact with the soil to become less mobile with time, either through precipitation or sorption mechanisms or further changes in speciation.
To further investigate the variability of tungstate Kd values in the Grenada Loring soil, the effects of pH, phosphate (at various pH values), ionic strength, and humic acid were determined. The partition coefficient values listed in Table 3
were obtained after the soil slurries were equilibrated for 14 d. The data suggest that increased pH (either from sodium hydroxide or phosphate, both at pH = 11) decreased the Kd value, as would be expected with an anionic compound, where surface exchange sites become negatively charged (Gustafsson, 2003). This decreased sorption was also pronounced in the neutral phosphate system, where phosphate likely competed with tungstate for sorption sites, thereby decreasing the tungstate Kd value. Phosphate, as a co-occurring analyte under acidic conditions, also caused the Kd to decrease 10-fold at the same pH and ionic strength resulting from nitric acid. The lower pH systems generally favor an increase in sorption through protonation of metal oxide surfaces in the soil, compared to neutral and alkaline systems, although in the phosphate systems, either phosphate out-competes tungstate for sorption sites or polymerizes with tungstate to create a more mobile species.
Humic acid and sodium sulfate as concomitant compounds had little effect on tungstate Kd values in comparison to the deionized water system. This was likely due to the fact that tungstate has a higher affinity for sorption sites than these compounds, which may not be the case with phosphate as discussed above (Gustafsson, 2003), or the humic substance behaves as a surfactant to the soil particle surface. Additionally, substantial amounts of naturally occurring salts will dissolve from the soil into the deionized water system, essentially making the ionic strength of the deionized water system similar to the sodium sulfate and humic acid matrices.
Because the Grenada Loring soil has substantial amounts of oxide-forming metals (1.1% Al, 1.5% Fe, 0.06% Mn), it is likely that some combination of these phases will influence tungsten sorption processes. Therefore, partition coefficients for the two most abundant metal oxides in the Grenada Loring soil (Al and Fe) were determined to elucidate tungstate's affinity for pure metal oxides. Aluminum and iron (III) oxides have tungstate Kd values of approximately 1970 and 1780 L/kg, respectively, using the same procedure previously described for the soil experiments. In comparison, arsenic oxyanion compounds have partition coefficients on freshly precipitated aluminum, iron, and manganese oxyhydroxides that range from <1 to over 200 L/kg depending on pH conditions, suggesting tungstate sorption to metal oxide phases will be extensive in this system (Langmuir, 1997).
The final Kd determination experiment performed used the tungsten species extracted from the tungsten-spiked soil after the soil had aged for 1 yr. For this test, the tungsten soil was extracted with deionized water for 168 h in the same manner as the extraction graphed in Fig. 1. The leachate was then filtered and analyzed for dissolved tungsten (150 mg/L). This solution was then used as the spike solution for clean Grenada Loring soil to determine the partition coefficient for the tungsten species extractable from the soil. The Kd value obtained after 3 d of equilibration was 110 L/kg (R2 = 0.98), which is almost a factor of 3 lower than the value obtained for tungstate, but close to the values obtained for polytungstate and tungstosilicate listed in Table 2 (284 L/kg for tungstate, 92 L/kg for polytungstate, and 103 L/kg for tungstosilicate). The Kd value measured for the geochemical form of tungstate in the tungsten spiked soil closely matches those of commercially available polymer tungsten compounds, indicating that a large portion of soluble tungsten in the Grenada Loring soil exists as polymeric species.
|
Conclusions
|
|---|
Tungsten speciation, i.e., the presence of polytungstate species, greatly affected tungsten geochemistry in soil. Additionally, sparingly soluble tungstate salts and soil sorption processes affected the dissolved tungsten concentrations in the short term (<72 h), as equilibrium concentrations were reached after approximately 48 h. However, polymerization of tungstate over time may lead to increased Kd values, indicating decreased mobility through different sorption mechanisms. The tungsten species present in the soil studied appeared to be a mixture of monomeric tungstate and a range of ill-defined polymer species. In general, all forms of tungsten investigated had lower Kd values than monomeric tungstate on the soil studied. Specific polytungstate Kd values suggested that some forms were stable in soil solution over extended periods of time, whereas other poly species engaged in geochemical reactions that changed speciation over time, and thus altered partition coefficients.
|
ACKNOWLEDGMENTS
|
|---|
The use of trade, product, or firm names in this report is for descriptive purposes only and does not imply endorsement by the U.S. Government. The tests described and the resulting data presented herein, unless otherwise noted, were obtained from research conducted under the Environmental Quality Technology Program of the United States Army Corps of Engineers by the US Army Engineer Research and Development Center. Permission was granted by the Chief of Engineers to publish this information. The findings of this report are not to be construed as an official Department of the Army position unless so designated by other authorized documents. The authors also thank Mark Chappell and David Johnson of the USAERDC for their editorial comments.
|
NOTES
|
|---|
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.
|
REFERENCES
|
|---|
- Backstrom, M., U. Nilsson, K. Hakansson, B. Allard, and S. Karlsson. 2003. Speciation of heavy metals in road runoff and roadside total deposition. Water Air Soil Pollut. 147:343–366.[CrossRef]
- Bednar, A.J., J.R. Garbarino, M.R. Burkhardt, J.F. Ranville, and T.R. Wildeman. 2004. Field and laboratory arsenic speciation methods and their application to natural water matrices. Water Res. 38:355–364.[Medline]
- Bednar, A.J., J.E. Mirecki, L.S. Inouye, L.E. Winfield, S.L. Larson, and D.B. Ringelberg. 2007. The determination of tungsten, molybdenum, and phosphorous oxyanions by high performance liquid chromatography inductively coupled plasma mass spectrometery. Talanta 72:1828–1832.
- Cruywagen, J.J. 2000. Protonation, oligomerization, and condensation reactions of vanadate (V), molybdate(VI), and tungstate(VI). p. 127–182. In Advances in Inorganic Chemistry. Vol. 49.
- Deery, M.J., O.W. Howarth, and K.R. Jennings. 1997. Application of electrospray ionization mass spectrometry to the study of dilute oligomeric anions and their reactions. J. Chem. Soc., Dalton Trans. 1997:4783–4788.
- Dermatas, D., W. Braida, C. Christodoulatos, N. Strigul, N. Panikov, M. Los, and S. Larson. 2004. Solubility, sorption, and soil respiration effects of tungsten and tungsten alloys. Environ. Forensics 5:5–13.[CrossRef]
- Feigl, F. 1958. Spot tests in inorganic chemistry. Elsevier Publ. Co., Amsterdam. 600 p.
- Gustafsson, J.P. 2003. Modelling molybdate and tungstate adsorption to ferrihydrite. Chem. Geol. 200:105–115.[CrossRef][Web of Science]
- Johnson, J.L., and K.W. Rajagopalan. 1976. Electron paramagnetic resonance of the tungsten derivative of rat liver sulfite oxidase. J. Biol. Chem. 251:5505–5511.[Abstract/Free Full Text]
- Johnson, J.L., W.R. Waud, H.J. Cohen, and K.W. Rajagopalan. 1974. Molecular basis of the biological function of molybdenum. J. Biol. Chem. 249:5056–5061.[Abstract/Free Full Text]
- Koutsospyros, A., W. Braida, C. Christodoulatos, D. Dermatas, and N. Strigul. 2006. A review of tungsten: From environmental obscurity to scrutiny. J. Hazard. Mater. 136:1–19.[CrossRef][Web of Science][Medline]
- Langmuir, D. 1997. Aqueous environmental geochemistry. Prentice Hall, Upper Saddle River, NJ.
- Light, T.S. 1972. Standard solution for redox potential measurements. Anal. Chem. 44:1038–1039.
- Ljung, K., E. Otabbong, and O. Selinus. 2006. Natural and anthropogenic metal inputs to soils in urban Uppsala, Sweden. Environ. Geochem. Health 28:353–364.[CrossRef][Web of Science][Medline]
- Seiler, R.L., K.G. Stollenwerk, and J.R. Garbarino. 2005. Factors controlling tungsten concentrations in ground water, Carson Desert, Nevada. Appl. Geochem. 20:423–441.[CrossRef]
- Strigul, N., A. Koutsospyros, P. Arienti, C. Christodoulatos, D. Dermatas, and W. Braida. 2005. Effects of tungsten on environmental systems. Chemosphere 61:248–258.[Medline]
- USEPA. 1996. SW-846 Methods for the Analysis of Hazardous Waste, ICP-AES Method 6010B, ICP-MS Method 6020. USEPA, Washington, DC.
This article has been cited by other articles:
|
|
|
|
 
D. B. Ringelberg, C. M. Reynolds, L. E. Winfield, L. S. Inouye, D. R. Johnson, and A. J. Bednar
Tungsten Effects on Microbial Community Structure and Activity in a Soil
J. Environ. Qual.,
January 13, 2009;
38(1):
103 - 110.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
