VZJ sign up for etocs
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 3 October 2006
Published in Vadose Zone J 5:1110-1118 (2006)
DOI: 10.2136/vzj2005.0140
© 2006 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hyun, S.
Right arrow Articles by Lee, L. S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Hyun, S.
Right arrow Articles by Lee, L. S.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Hyun, S.
Right arrow Articles by Lee, L. S.
Related Collections
Right arrow Fate
Right arrow Mining Wastes
Right arrow Vadose Zone Risk Assessment

ORIGINAL RESEARCH

Selenium(IV) and (VI) Sorption by Soils Surrounding Fly Ash Management Facilities

Seunghun Hyuna,b, Perre E. Burnsa, Ishwar Murarkac and Linda S. Leea,*

a Dep. of Agronomy, Purdue University, West Lafayette, IN 47907-2054
b Div. of Environmental Science and Ecological Engineering, Korea University, Seoul 136-701, Korea
c Ish, Inc., 804 Salem Woods Dr., Suite 201B, Raleigh, NC 27615-3313
* Corresponding author (lslee{at}purdue.edu)

Received 29 November 2006.



   ABSTRACT
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND APPLICATION OF...
 REFERENCES
 
Leachate derived from unlined coal ash disposal facilities is one of the most significant anthropogenic sources of selenium to the environment. To establish a practical framework for predicting transport of selenium in ash leachate, sorption of Se(IV) and Se(VI) from 1 mM CaSO4 was measured for 18 soils obtained down-gradient from three ash landfill sites and evaluated with respect to several soil properties. Furthermore, soil attenuation from lab-generated ash leachate and the effect of Ca2+ and SO42– concentrations as well as pH on both Se(IV) and Se(VI) was quantified for a subset of soils. For both Se(IV) and Se(VI), pH combined with either percentage clay or dithionite-citrate-bicarbonate (DCB)-extractable Fe described >80% of the differences in sorption across all soils, yielding an easy approach for making initial predictions regarding site-specific selenium transport to sensitive water bodies. Se(IV) consistently exhibited an order of magnitude greater sorption than Se(VI). Selenium sorption was highest at lower pH values, with Se(IV) sorption decreasing at pH values above 6, whereas Se(VI) decreased over the entire pH range (2.5–10). Using these pH adsorption envelopes, the likely effect of ash leachate-induced changes in soil pore water pH with time on selenium attenuation by down gradient soils can be predicted. Selenium sorption increased with increasing Ca2+ concentrations while SO42– suppressed sorption well above enhancements by Ca2+. Soil attenuation of selenium from ash leachates agreed well with sorption measured from 1 mM CaSO4, indicating that 1 mM CaSO4 is a reasonable synthetic leachate for assessing selenium behavior at ash landfill sites.

Abbreviations: DCB, dithionite-citrate-bicarbonate • NE, North East [site] • MW, Midwest [site] • ppb, parts per billion • SE, South East [site]


   INTRODUCTION
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND APPLICATION OF...
 REFERENCES
 
SINCE THE ISSUANCE by USEPA of the 1987 chronic criterion for selenium of 5 parts per billion (ppb, µg kg–1), biologists and industries have debated the adequacy of the guideline (Renner, 2005). Although Se is essential to humans and animal nutrition, it has been recognized to adversely impact aquatic ecosystems. Selenium is bioaccumulated in tissues, and thus chronic exposure to low concentrations may result in developmental abnormalities in embryos and disturb reproductive cycles of animals (Ohlendorf et al., 1986; Dhillon and Dhillon, 2003). The difference in Se concentrations considered beneficial versus those deemed detrimental to biota is smaller than noted for other USEPA priority or nonpriority pollutant (USEPA, 2004).

Anthropogenic sources of Se include application of agricultural pesticides, disposal of industrial wastes and combustion of fuels. One significant source of Se to terrestrial and aquatic ecosystem is coal ash. Selenium is naturally enriched in fossil fuels well above concentrations found in the earth's curst of <0.05 mg kg–1 (Bowen, 1979). Selenium is further concentrated in coal combustion byproducts such as fly ash (e.g., 7–760 mg Se kg–1 ash) (Theis and Gardner, 1990). Currently in the USA, approximately 45% of coal-derived fly ash is placed in landfills (USEPA, 1999); thus, the Se-bearing leachate that may seep to adjacent water bodies or groundwater is of concern (Eary et al., 1990; Carlson and Adriano, 1993). In a review by Theis and Gardner (1990), Se concentrations in pore-water and leachate collected from typical ash disposal sites were 10 to 540 µg L–1. However, a recent report by Ladwig et al. (2006) found much higher leachate concentrations and noted that the dominant species appeared to be a function of the specific coal combustion product and the coal source. For example, leachate concentrations as high as 1760 µg L–1, primarily as Se(VI), were found from landfills containing fly ash generated from subbituminous coal. Alternatively, leachate from a flue gas desulfurization sludge impoundment contained primarily Se(VI), with total Se concentrations in excess of 2000 µg L–1.

Selenium exists in four oxidation states in natural systems: selenate [Se(VI)], selenite [Se(IV)], elemental Se [Se(0)], and selenide [Se(–II)]. Se(IV) and Se(VI) are the most mobile forms in the pH–pE range typical of most soils with elemental Se(0) and Se(–II) being considered relatively immobile (Neal et al., 1987a). Se(VI) is present as SeO42– over the entire pH range of soils, while the major Se (IV) species is HSeO3– (pKa = 2.68) at pH < 8.4 and SeO32– (pKa = 8.4) at pH > 8.4 (Séby et al., 1998). Selenium speciation in fly ash and subsequently in leachate will vary with coal origin, coal combustion processes, ash weathering processes, and leaching methodology. For example, water-extractable Se from various fly ashes ranged from 5 to 100% of the total Se was as Se(IV) (van der Hoek et al., 1996; Jackson and Miller, 1998). Ash heterogeneity, kinetic barriers to equilibrium and biological processes also make it difficult to predict the redox speciation.

The sorption behavior of Se is governed by oxidation–reduction potential (pE), mineralogical composition of the soil, soil pH, and the solution composition (Neal et al., 1987a, 1987b; Eary et al., 1990; Goh and Lim, 2004). From studies involving the interaction of the inorganic Se species (IV and VI) with model soil components, iron/aluminum (hydro) oxides and allophane appear to have the greatest affinity for Se (Rajan and Watkinson, 1976; Hansmann and Anderson, 1985; Neal et al., 1987a, 1987b; Goldberg and Glaubig, 1988; Zhang and Sparks, 1990; Su and Suarez, 2000). Of the phyllosilicate minerals, kaolinite with variable charge edge sites generally sorbs Se more than montmorillonite, which has a negative constant charge surface (Bar-Yosef and Meek, 1987; Goldberg and Glaubig, 1988). At high Se concentrations where precipitation is induced, removal of Se(IV) from aqueous solutions is greater with montmorillonite due to its higher surface area compared to kaolinite (700–800 cm2 g–1 vs. 5–20 cm2 g–1) (Frost and Griffin, 1977).

Se(VI) sorption generally appears to parallel that of sulfate (e.g., outer-sphere surface complexation) with relatively low sorption and high mobility (Rajan and Watkinson, 1976; Goldberg and Glaubig, 1988; Neal and Sposito, 1989); however, inner-sphere complexation on iron or aluminum oxides/hydroxides has been reported from spectroscopic data (Peak and Sparks, 2002; Manceau and Charlet, 1994; Wijnja and Schulthess, 2000). Se(VI) sorption can be significant on oxides, kaolinite, and oxide-enriched soils, especially in the acidic pH range (Balistrieri and Chao, 1987; Bar-Yosef and Meek, 1987; Su and Suarez, 2000), while sorption by several calcareous, montmorillonitic and alluvial soils is usually negligible (Goldberg and Glaubig, 1988; Neal and Sposito, 1989). Se(IV) sorption is more analogous to phosphate (e.g., inner-sphere surface complexation), thus sorbs more than Se(VI) (Hansmann and Anderson, 1985; Zhang and Sparks, 1990). Specific Se(IV) adsorption on metal oxide surfaces is evident by an observed decrease in the metal oxide's zero point of charge following Se(IV) sorption (Rajan, 1979; Bowden et al., 1980), reduced electrolytic mobility (Su and Suarez, 2000), and structural observation by spectroscopic study (Hayes et al., 1987).

Sorption of both Se(IV) and Se(VI) is pH dependent, with Se sorption generally decreasing with increasing pH (Bowden et al., 1980; Bar-Yosef and Meek, 1987; Neal et al., 1987a; Zhang and Sparks, 1990; Saeki et al., 1995; Goh and Lim, 2004), which is consistent with the pKa value (8.4) of SeO32– and the typical pH range of environmental relevance. With increasing pH, the soil surface becomes increasingly more negative and a greater fraction of Se becomes anionic, thus repulsive forces increase and sorption decreases. Se(VI) sorption is more sensitive to pH than Se(IV) (Bar-Yosef and Meek, 1987), with the magnitude of pH-induced changes being soil dependent.

Selenium retention by soils is affected by other ions present in the soil solution. Se(VI) sorption is more sensitive to the presence of competing anions than Se(IV) (Balistrieri and Chao, 1987; Saeki et al., 1995; Goh and Lim, 2004). For example, Se(VI) sorption decreased approximately 50% in the presence of 10 mM sulfate (Goh and Lim, 2004), whereas sulfate did not significantly affect Se(IV) sorption (Balistrieri and Chao, 1987). Se(IV) also out-competed phosphate for sorption sites when present at equimolar concentrations; at higher phosphate concentrations, Se(IV) sorption was suppressed (Balistrieri and Chao, 1987). The inhibitory effect of competing inorganic anions for Se sorption generally follows: phosphate > citrate > oxalate > sulfate (Balistrieri and Chao, 1987; Saeki et al., 1995), with chloride having little to no effect (Neal et al., 1987b). Divalent cations can enhance Se sorption by causing surface precipitation with Fe2+ or Cu2+ (Benjamin, 1983; Manning and Burau 1995) and decreased negative (or increased positive) surface potential induced by Ca2+ sorption (Neal et al., 1987b). In both cases, the divalent cation impact is greater at alkaline pH values where divalent cation (Cu2+ and Ca2+) adsorption is favored.

Understanding sorption behavior is critical for assessing true risk potential to nearby water bodies of Se released from ash disposal facilities. Toward establishment of a practical framework for predicting Se attenuation by soil, this study focused on: (i) quantifying the sorption of Se(IV) and Se(VI) onto utility site soils having a wide range of physical and chemical properties, (ii) establishing a range of partition coefficients that may be used in assessing Se mobility in ash leachate, (iii) characterizing the effects of SO42– and Ca2+ on the sorption of Se, and (iv) characterizing pH effects on Se sorption.


   MATERIALS AND METHODS
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 SUMMARY AND APPLICATION OF...
 REFERENCES
 
Soils
Soil samples were collected from three utility sites within United States (identified as the North East [NE], South East [SE], and Midwest [MW] sites) where ash landfills were in use or planned. For each site, three soils were sampled down gradient from the designated landfill areas at two depths between approximately 3 and 18.3 m ({approx}10 and 60 feet). Properties of these soils were reported in detail by our group in a previous paper (Burns et al., 2006). The primary minerals present in the NE and SE soils are illite and kaolinite with NE1 also containing significant quantities of vermiculite. The MW soils were 93 to 97% sand with <1% clay; thus a mineral identification was not determined. All soils had no more than 1% organic C content. A subset of soil properties taken from Burns et al. (2006) is shown in Table 1.


View this table:
[in this window]
[in a new window]
  		Table 1. Selected physical and chemical properties of soils used in this study.

 
Sorption Isotherm from 1 mM CaSO4
Both Se(IV) and Se(VI) sorption isotherms for 18 soils were measured in 1 mM CaSO4 solutions (ionic strength, I = 4 mM), which represents the dominant constituents in most coal ash leachate. Sorption isotherms were constructed with five to eight initial concentrations in duplicate ranging from 0 to 34 µmol L–1 for Se(IV) and 0 to 13 µmol L–1 for Se(VI). Se(IV) and Se(VI) solutions were prepared by dissolving from Na2O3Se or Na2O4Se·10 H2O in distilled water, respectively. Selenium solutions were added to polypropylene centrifuge tubes containing 1 g of soil to obtain a soil mass/solution ratio of 50 g L–1 for Se(IV) and 100 g L–1 for Se(VI). Soil suspensions were then allowed to equilibrate at 23 ± 2°C on a rotary shaker (~45 rpm) for 16 and 48 h for Se(IV) and Se(VI), respectively. Equilibration times were selected to achieve near equilibrium without having significant shifts in the redox state of the Se remaining in solution. The redox state of sorbed Se was assumed to be in the redox state applied. Suspensions containing Se(IV) were equilibrated for a shorter period of time due to concerns that a significant amount of Se(IV) would oxidize to Se(VI). Preliminary studies performed with NE1 25–30, NE2 10–15, and MW3 23 soils and applied solution concentrations of 1.5, 15, or 30 µmol L–1 at a soil mass/solution ratio of 50 g L–1 showed <10% of the Se(IV) was oxidized in 16 h. This subset of soils represented the highest and almost the lowest Kf values, high and low values of pH, extractable Fe, extractable Al, organic matter, and clay content of the soils used in this study. The pH of the equilibrated suspensions was measured with a pH meter, samples centrifuged, and 10-mL aliquots of supernatant removed and filtered through a 0.45-µm regenerated cellulose filter. Filtered solutions were acidified by adding 0.1 mL of 20% (v/v) of HNO3 for Se(VI) or HCl for Se(IV). Aqueous Se concentrations were determined using a Shimadzu (Kyoto, Japan) graphite furnace atomic adsorption spectrometer. Speciation checks were performed using HVG-AA for Se(IV), totals by graphite furnace, and Se(VI) by difference.

pH Effect
The effect of pH on Se sorption for one soil from each site was characterized by measuring sorption from pH-adjusted 1 mM CaSO4 solutions in duplicate at a single initial concentration of 6.64 µmol L–1 and 5.22 µmol L–1 for Se(IV) and Se(VI), respectively. Predetermined amounts of diluted KOH and HCl were added to 1 mM CaSO4 solution to obtain equilibrium pH values of pH 3 to 9.

Matrix Effect
The effects of SO42– and Ca2+ concentrations on sorption of Se(IV) and Se(VI) from a single concentration of 17.55 µmol L–1 for Se(IV) and 6.74 µmol L–1 for Se(VI) were investigated for NE1 25–30 soil. Concentrations of 0, 0.1, 0.5, 1, 5, and 10 mM SO42– or Ca2+ were prepared with CaCl2 and K2SO4 solutions at a constant I of 0.03, which was done by adding appropriate concentrations of KCl. The effects of SO42– and Ca2+ were furthered investigated by measuring multiconcentration isotherms from 1 mM KCl (I = 0.001), 1 mM CaCl2 (I = 0.003), 1 mM K2SO4 (I = 0.003), and 1 mM CaSO4 (I = 0.004)) on NE1 25–30, SE1 48.5, and MW1 17 soils for Se(IV) and NE1 25–30 and NE2 10–15 soils for Se(VI). To minimize pH differences in the different electrolytes, pH was adjusted to be what was observed in 1 mM CaSO4 for all electrolyte solutions.

Soil Attenuation of Selenium from Ash Leachate
Site-specific Se sorption from ash leachate was measured for three soils that represented the range of sorption observed for the two sites for which Se-containing fly ash could be obtained. One sample was a mixture of two soils from the SE site (SE1 collected at 14.78 m [48.5 feet] and 18.29 m [60 feet]), because the soils had essentially identical sorption behavior in 1 mM CaSO4. The other two soils were from the MW site (MW1 17 and MW2 25). Fly ash leachate was generated by equilibrating deionized water with fly ash at the soil mass/solution ratio of 1 g/10 mL for 24 h. Total Se ash leachate concentrations for SE and MW site ashes were 0.32 and 0.42 µmol L–1, respectively, with <10% as Se(IV). These concentrations were much lower than Se levels found in a monitoring well (3.7 µmol L–1) located within the ash fill at the one site where a landfill was already in use. The concentration difference between the batch and well leachates is most likely due to the differences in contact time, the soil mass/solution ratio, and potential changes in solubility controls with weathering. Selenium ash leachate concentrations were fortified close to those observed in the monitoring well, which is equivalent to 30 times the USEPA-established health standard for total Se. The fortified ash leachate was equilibrated with soil at the mass to volume ratio of 1:10 and all other details are as described for the multiconcentration sorption isotherms.

Data Analysis
The amount of Se sorbed (Cs, µmol kg–1) was calculated by the difference in the Se concentrations applied to the soils and that obtained in the aqueous phase (Cw, µmol L–1) after equilibration with soils. Selenium isotherm data were fit with Freundlich sorption model:

Formula 1[1]
where Kf is the Freundlich sorption coefficient (µmol1–N LN kg–1), and N (unitless) is a measure of isotherm nonlinearity. Sorption coefficients (Kf and N) were optimized using SAS nonlinear programming (SAS Institute, 1989).


   RESULTS AND DISCUSSION
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
SUMMARY AND APPLICATION OF...
 REFERENCES
 
Selenium Sorption from 1 mM CaSO4
Freundlich nonlinear sorption model (Eq. [1]) fits to the sorption data from all soils are summarized in Table 2. Isotherms were well described by the Freundlich model with all correlation coefficients (r2, {alpha} = 0.001) falling between 0.94 and 1.0. Freundlich isotherm model coefficients (Kf and N) can be used to estimate a concentration-specific sorption coefficient (Kd*, L kg–1) for use in comparing sorption behavior

Formula 2[2]
This approach can also be useful when predicting mobility with transport models that only allow for a single sorption coefficient (concentration-independent Kd value in which isotherms are linear), which is often the case for simple models used by regulatory-type personnel. A direct comparison of Kf values is not appropriate if N values vary among soils. Concentration-specific sorption coefficients (Kd*; Eq. [2]) are included in Table 2 for Cw = 1.27 µmol L–1 (100 µg L–1), which is 10 times higher than the USEPA-established maximum level of 10 µg L–1. This concentration is also approximately the mid-point of the Cw range measured from isotherms in this study. For a given soil, Kd* values estimated at Cw = 1.27 µmol L–1 (100 µg L–1) are slightly smaller than Kf values with the maximum deviation of 16% for Se(IV) sorption by NE3 50–54 soil, where the greatest nonlinearity in sorption was observed (N = 0.249). For a given concentration, sorption of Se(IV) is several times greater than Se(VI), similar to previous studies (Balistrieri and Chao, 1987; Bar-Yosef and Meek, 1987; Goldberg and Glaubig, 1988; Neal and Sposito, 1989; Goh and Lim, 2004). At Cw = 1.27 µmol L–1 (Table 2), Kd* values ranged from 24 to 521 L kg–1 for Se(IV) compared with 2.6 to 13.4 L kg–1 for Se (VI).


View this table:
[in this window]
[in a new window]
  		Table 2. Freundlich sorption model parameters for selenite and selenate sorption measured from 1mM CaSO4 solution.

 
With increasing sorption nonlinearity (decreasing N value), the range of Kd* values estimated for a range of concentrations will become increasingly large. For example, Kd* values for Se(IV) on NE3 50–54 estimated for Cw values from 0.063 to 20.3 µmol L–1, which spans the 1987 USEPA chronic criterion and higher Se(IV) concentrations reported from field data (Ladwig et al., 2006), will range from {approx} 4500 to 60 L kg–1. Since Se(VI) exhibits much less nonlinear sorption (Table 2), differences between Se(IV) and Se(VI) sorption will become less with increasing Se concentration. Therefore, although Se(IV) is generally much more sorbed than Se(VI), in scenarios where high Se concentrations are present, mobility will not be very species-dependent.

Regression Analysis
Few quantitative analyses have been conducted to couple Se sorption with soil properties. For 66 New Zealand soils, Se(IV) sorption from 30 mM KCl were correlated well with oxalate-extractable Fe, Al, and DCB-extractable Fe, with r2 ({alpha} = 0.001) being 0.807, 0.761, and 0.730, respectively (John et al., 1976). For 58 Japanese soils, oxalate-extractable Fe and Al were the best predictors for Se(IV) sorption from H2O solution (Nakamaru et al., 2005). In all of these studies, Se(IV) sorption was estimated from only a single high initial concentration (10 000 µmol L–1) and did not include ions like Ca2+ and SO42– common to ash leachate system. Correlations between Se(IV) and Se(VI) sorption and properties for 18 soils in the current study were assessed using a simple and multiple linear regression analyses. Regression results with Kd* estimated at Cw = 1.27 µmol L (100 µg L–1) are shown in Table 3; similar trends were obtained regardless of the concentration chosen to estimate Kd*. Soil clay and oxide measurements were found to provide the best positive simple linear correlations. For the oxide data, the best correlation for both Se(IV) and Se(VI) sorption was found with 15-s DCB-extractable Fe. The 15-s DCB-extractable Fe, which ranges from 0.3 to 10.8% of DCB-extractable Fe (Table 1) for the soils in this study, is intended to estimate the most reactive (easily reducible) portion of Fe oxide (i.e., Fe oxide exposed to solution and readily available for a reaction) (Burns et al., 2006). Sorption of both species was also negatively correlated with isotherm pH, which when coupled to 15-s DCB-extractable Fe describes most of the sorption differences across soils with r2 values of 0.884 and 0.818 for Se(IV) and Se(VI), respectively. However, combining pH and clay content also gave good predictions, with r2 = 0.802 and 0.836 for Se(IV) and Se(VI), respectively. Note that soil oxides are in the clay fraction (<2 µm). However, not all clays are oxides; therefore, the relatively high correlation coefficients for clay suggests other <2-µm-size particles, such as kaolinite, illite, and vermiculite, may also contribute to Se sorption by these soils. Both pH and clay content are common results of routine soils analyses, making their use in predicting Se sorption appealing.


View this table:
[in this window]
[in a new window]
  		Table 3. Regression equation describing dependency of Se sorption coefficient (Kd* [L kg–1], at Cw = 1.27 µmol L–1 [100 µg L–1]) on pH, clay, and oxide contents.

 
pH Effect
Adsorption envelopes in which Se sorption is expressed as the percentage of Se sorbed relative to that applied are shown in Fig. 1 for Se(IV) and Se(VI) as a function of pH for one soil from each site. Selenium sorption is strongly pH dependent, with the greatest sorption occurring in the acidic pH range. For Se(IV), sorption maxima are more defined and in a lower pH range with increasing clay content and extractable Fe; NE1 25–30, SE1 48.5, and MW1 17 exhibit apparent maxima or sorption plateaus at pH <4.5, <5.5, and <7, respectively. Likewise, changes in Se(VI) sorption with pH are more substantial with increasing clay and extractable Fe content, but sorption decreases over the entire pH range studied (~3–10). The greater sorption with decreasing pH is characteristic of anion sorption for oxides and clay minerals, including kaolinite, which is a dominant clay type of the soils investigated (Bowden et al., 1980; Balistrieri and Chao, 1987; Neal et al., 1987a). Increased protonation of mineral edge functional groups ({equiv}OH2+) with decreasing pH is expected to increase the anion exchange capacity, and therefore, increase the amount of anions sorbed. Also, based on the frequent observation of Se(IV) sorption exceeding soil anion exchange capacity, it appears that Se(IV) can replace the uncharged OH2 surface groups (ligand exchange) (Rajan, 1979).


Figure 1
View larger version (19K):
[in this window]
[in a new window]
  		Fig. 1. Selenium sorption as a function of pH expressed as the percentage of Se sorbed relative to Se applied. Applied concentrations were 6.64 µmol L–1 for Se(IV) and 5.22 µmol L–1 for Se(VI), which is mid-range of the concentrations used to construct the multiconcentration isotherms. The ratio of soil mass to solution volume were 1:50 and 1:20 for Se(IV) and Se(VI), respectively.

 
As the ash leachate plume moves through soils down gradient of an ash landfill, soil solution pH may change, thus affecting Se mobility. The potential changes in soil solution pH will depend on the natural soil pH, the buffer capacity of the soil, the pH and buffer capacity of the ash leachate, and residence time. For the NE soils, the range in soil pH measured in water and 1 mM CaCl2 is between 4.4 and 5.5, and the pH of ash associated with that site ranged between 5.5 and 6.2 (Burns et al., 2006). Therefore, assuming the pH adsorption envelope for NE1 25–30 is representative of the NE site, an ash plume-induced shift in soil pH may cause changes in Se mobility. However, the NE soils contain about 1% organic C, significant amounts of clay, and high extractable Fe and Al contents, thus they are likely to have a reasonable capacity to buffer pH changes. For the SE soils, the range in soil pH in water and 1 mM CaCl2 is between 5.4 and 6.4, and the ash associated with that site has a pH {approx} 6 (Burns et al., 2006). Assuming the pH envelope for SE1 48.5 is representative of the SE site, soil solution pH, and therefore Se sorption, is unlikely to change significantly with time. For the MW soils, the range in soil pH in water and 1 mM CaCl2 is between 6.2 and 8.1. However, the ash associated with this site yields a leachate with a much higher pH of 11 to 12 (Burns et al., 2006). Soils at this site are also very sandy (93–97% sand), thus have little capacity to buffer or sorb. Therefore, leachate-induced increases in pH and increased mobility are expected at the MW site.

Matrix Effect
The Kd values measured for NE1 25–30 soil from a single Se concentration in the presence of increasing sulfate (SO42–) or calcium (Ca2+) concentrations relative to the Kd values measured in the absence of Ca2+ or SO42– (KCl solution at I = 0.03) soil are shown in Fig. 2. Increasing SO42– concentrations led to decreases in sorption while the opposite trend was observed with Ca2+. SO42– decreased sorption of Se(VI) much more than Se(IV) sorption in agreement with the hypothesis that Se(VI) forms outer-sphere complexes similar to sulfate, whereas Se(IV) tends to form inner-sphere complexes more like phosphate (Neal et al., 1987b). The presence of Ca2+ increased sorption of both species, but increases in Se(IV) sorption were greater. For Se(VI), the impacts of Ca2+ or SO42– were of similar magnitudes, whereas the Kd for Se(VI) was reduced by 80% in 10 mM SO42– compared with a <25% increase measured from 10 mM Ca2+.


Figure 2
View larger version (14K):
[in this window]
[in a new window]
  		Fig. 2. The Kd value for (upper) Se(IV) and (lower) Se(VI) for NE1 25–30 soil measured from a single Se concentration in the presence of increasing concentrations of sulfate (SO42–) or calcium (Ca2+) relative to the Kd value measured from the same Se concentration in the absence of Ca2+ or SO42– (KCl solution at I = 0.03). Initial concentrations were 17.55 µmol L–1 and 6.74 µmol L–1 for Se(IV) and Se(VI) with a soil mass/solution volume of 1:50 and 1:20, respectively. Constant ionic strength of 30 mM was achieved by adding appropriate amounts of KCl.

 
A subset of the multiconcentration isotherms measured from 1 mM KCl, 1 mM CaCl2, 1 mM K2SO4, and 1 mM CaSO4 are shown for Se(IV) and Se(VI) (Fig. 3). For Se(VI), only SO42– appears to affect sorption significantly; thus, sorption is similar and suppressed in CaSO4 and K2SO4 and higher but similar in CaCl2 and KCl (Fig. 3C and 3D). Se(IV) sorption is reduced by SO42–, but Ca2+ enhanced sorption, resulting in the following trends for all three soils investigated: 1 mM CaCl2 > 1 mM CaSO4 > 1 mM KCl > 1 mM K2SO4 (Fig. 3A and 3B; data for MW1 17 not shown). These trends are generally in agreement with the single concentration Kd values measured with increasing Ca2+ or SO42– concentration (Fig. 2), with one apparent discrepancy in terms of magnitude. Based on Fig. 2, equal concentrations of Ca2+ or SO42– were expected to cancel out impacts to Se(IV) sorption (e.g., similar sorption expected from 1 mM CaSO4 and 1 mM KCl). However, Se(IV) sorption in the multiconcentration isotherm data is much greater in 1 mM CaSO4 compared with 1 mM KCl. This apparent discrepancy is likely an effect of the substantial sorption nonlinearity (N < 0.5 for the NE soils) for Se(IV) (Table 2). For a direct comparison of single concentration Kd values, equilibrium concentrations must be the same. Data for Fig. 2 were collected from a single applied concentration, resulting in different equilibrium concentrations. Thus sorption nonlinearity effects were not equal. Also ionic strength (I) was not kept constant in the multiconcentration isotherms (I ranged from 1 to 4 mM); however, effects on Se(IV) sorption in this range have been reported to be negligible (Neal et al., 1987b; Peak and Sparks, 2002).


Figure 3
View larger version (25K):
[in this window]
[in a new window]
  		Fig. 3. (A, B) Measured (symbols) Se(IV) sorption isotherms from 1 mM KCl, 1 mM CaCl2, 1 mM K2SO4, and 1 mM CaSO4 and (C, D) Se(VI) sorption isotherms from 1 mM KCl, 1 mM CaCl2, 1 mM K2SO4, and 1 mM CaSO4 and corresponding Freundlich sorption model fits (solid lines).

 
Soil Attenuation of Selenium from Ash Leachate
Attenuation of Se from ash leachate by the SE and MW site soils is shown in Fig. 4. Also shown for comparison are dotted lines generated from the Freundlich model fits to the multiconcentration Se isotherms in 1 mM CaSO4. Se(IV) and Se(VI) present in the ash leachate were significantly attenuated through sorption by the site soils in a manner consistent with the isotherms constructed using 1 mM CaSO4. Therefore, the use of 1 mM CaSO4 as a simulated ash leachate for estimating site-specific Se sorption behavior appears reasonable for these ashes.


Figure 4
View larger version (18K):
[in this window]
[in a new window]
  		Fig. 4. Attenuation of Se(IV) and Se(VI) in ash leachate by site soils. Also shown for comparison are dotted lines generated from the Freundlich model fits of the multiconcentration Se isotherms in 1 mM CaSO4. For SE1 48.5 + SE1 60 soils, Freundlich sorption fit were generated by combining the data sets from the two soils.

 

   SUMMARY AND APPLICATION OF FINDINGS
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND APPLICATION OF...
REFERENCES
 
Se(IV) and Se(VI) sorption by soils from three spatially different ash landfill sites illuminated several factors influencing Se mobility—including valence state, Ca2+and SO42– concentrations, pH, selected soil properties—thus providing a practical framework for site-specific assessments. Both Se(IV) and Se(VI) sorption were well described using soil solution pH and clay percentage, which are routinely measured, yielding an easy approach for making initial predictions regarding site-specific Se transport. Sorption studies confirmed that Se(IV) is generally sorbed much more than Se(VI). However, since sorption nonlinearity is substantially greater for Se(IV), Se(IV) retention will decrease dramatically with increasing leachate concentrations, thus minimizing species-dependent mobility. Retention of Se will also be reduced in sulfate-rich waters, typical of many ash leachates. Observed SO42– concentrations in ash leachate range from 0.4 to 64 mM with a median of 3.26 mM (Ladwig et al., 2006). Although Ca2+enhanced Se(IV) sorption, suppression by SO42– at similar concentrations was greater. The Ca2+ concentrations in ash leachate range from <0.05 to 17 mM, with a median of 1.38 mM, which is lower than that reported for SO42– (Ladwig et al., 2006). Soil attenuation of Se from lab-generated ash leachate agreed well with sorption from 1 mM CaSO4. The latter supports the use of 1 mM CaSO4 as a simulated ash leachate in the absence of ash-induced pH effects. Sorption of both Se(IV) and Se(VI) is greatest within the lower pH range with Se(VI) sorption being affected over a larger pH range. Values of pH from 1.2 to 12.5 have been observed for lab-generated and field leachate from fly ash, but most ashes tend to be alkaline (Page et al., 1979; Hower et al., 1996; Ladwig et al., 2006). Continual inputs of a high pH ash leachate may induce soil pore water pH changes with time, thus decreasing sorption and further increasing Se mobility from ash landfills.


   ACKNOWLEDGMENTS
 
This work was funded in part by the Electric Power Research Institute (Ken Ladwig, EPRI Project Manager) and supported in part by a Korea University Grant.


   REFERENCES
TOP
 EXECUTIVE SUMMARY
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SUMMARY AND APPLICATION OF...
 REFERENCES
 




This article has been cited by other articles:


Home page
 Vadose Zone JHome page
S. Goldberg, S. Hyun, and L. S. Lee
Chemical Modeling of Arsenic(III, V) and Selenium(IV, VI) Adsorption by Soils Surrounding Ash Disposal Facilities
Vadose Zone J., November 26, 2008; 7(4): 1185 - 1192.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hyun, S.
Right arrow Articles by Lee, L. S.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Hyun, S.
Right arrow Articles by Lee, L. S.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Hyun, S.
Right arrow Articles by Lee, L. S.
Related Collections
Right arrow Fate
Right arrow Mining Wastes
Right arrow Vadose Zone Risk Assessment

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome