JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Green, C. H.
Right arrow Articles by Kelly, E. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Green, C. H.
Right arrow Articles by Kelly, E. F.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Green, C. H.
Right arrow Articles by Kelly, E. F.
Related Collections
Right arrow Redox Processes
Right arrow Heavy Metals
Right arrow Soil Chemistry
Journal of Environmental Quality 32:1323-1334 (2003)
© 2003 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORTS
Heavy Metals in the Environment

Solubilization of Manganese and Trace Metals in Soils Affected by Acid Mine Runoff

C. H. Green, D. M. Heil*, G. E. Cardon, G. L. Butters and E. F. Kelly

Department of Soil and Crop Sciences, Colorado State Univ., Fort Collins, CO 80523-1170

* Corresponding author (dheil{at}lamar.colostate.edu)

Received for publication March 15, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Manganese solubility has become a primary concern in the soils and water supplies in the Alamosa River basin, Colorado due to both crop toxicity problems and concentrations that exceed water quality standards. Some of the land in this region has received inputs of acid and trace metals as a result of irrigation with water affected by acid mine drainage and naturally occurring acid mineral seeps. The release of Mn, Zn, Ni, and Cu following saturation with water was studied in four soils from the Alamosa River basin. Redox potentials decreased to values adequate for dissolution of Mn oxides within 24 h following saturation. Soluble Mn concentrations were increased to levels exceeding water quality standards within 84 h. Soluble concentrations of Zn and Ni correlated positively with Mn following reduction for all four soils studied. The correlation between Cu and Mn was significant for only one of the soils studied. The soluble concentrations of Zn and Ni were greater than predicted based on the content of each of these metals in the Mn oxide fraction only. Increases in total electrolyte concentration during reduction indicate that this may be the result of displacement of exchangeable metals by Mn following reductive dissolution of Mn oxides.

Abbreviations: RSF, relative solubilization factor


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
A SIGNIFICANT ACREAGE of soils in the Alamosa River basin, Colorado has been irrigated with water from the Alamosa River, which has been affected by acid, metal-rich drainage from the Summitteville mine and many surrounding naturally occurring acid mineral seeps and historic mine sites. Manganese concentrations in ground water and surface waters in this region consistently exceed water quality standards of 0.05 mg L-1 for drinking water and 0.20 mg L-1 for agricultural water (Wendlandt et al., 1998). Manganese is an essential plant nutrient, and Mn deficiencies have been observed primarily in arid regions (Xiang and Banin, 1996). However, Mn toxicities to plants are sometimes observed in flooded soils (Smith, 1990). This condition has been observed in Alamosa River basin soils where flood irrigation results in waterlogged conditions that persist for several days (Maya ter Kulie, Agro Engineering, Inc., Alamosa, CO, personal communication, 1998). The solubility of Mn in soils is known to be highly sensitive to changes in soil redox conditions (Sposito, 1989). Xiang and Banin (1996) found that a substantial fraction of the Mn oxides present in two arid soils was dissolved within 3 d following saturation. Manganese oxides are more susceptible to reductive dissolution under slight to moderate soil reduction than are Fe oxides, and consequently Mn dissolution precedes Fe dissolution in saturated soils (Patrick and Jugsujinda, 1992). The reduction of Mn oxides before Fe oxides is consistent with the theoretical redox potentials under which Mn oxides vs. Fe oxides become thermodynamically unstable (Sposito, 1989).

Manganese oxides also adsorb trace metals including Pb, Zn, Cu, and Ni via specific surface adsorption (McKenzie, 1989). Trace metals are also known to associate with Mn oxides by substitution and coprecipitation (McKenzie, 1989). Coprecipitation and substitution of metals in Mn oxides is expected to occur especially when soils experience alternate wetting and drying cycles, creating opportunities for the metals to associate with freshly precipitated Mn oxide (McBride, 1994). Although it is difficult to ascertain whether metals are associated with Mn oxides via specific surface adsorption or substitution, Mn oxide particles isolated from bulk soil contain high concentrations of trace metals (McKenzie, 1989). Davranche and Bollinger (2000) illustrated that Pb and Cd adsorbed to synthetic Mn and Fe oxide was released following reductive dissolution. Soil reduction has been shown to result in the coincident release of metals associated with minerals that are susceptible to reductive dissolution, in particular Mn and Fe oxides (Charlatchka and Cambier, 2000; Davranche and Bollinger, 2000). Once Mn(IV) oxides are reduced, the Mn(II) that is released is subject to readsorption by soil colloids, and a significant amount of Mn is converted from Mn oxides to exchangeable Mn (Xiang and Banin, 1996). The specific reactions responsible for the retention of Mn(II) in soils include adsorption to Fe oxides, organic matter, and layer silicate clay minerals (Khattack and Page, 1992). Precipitation of the mineral rhodochrosite (MnCO3) has been demonstrated under reducing conditions and high CO2 partial pressure (Schwab and Lindsay, 1983). The soluble concentration of Mn following reductive dissolution of Mn(IV) oxides should depend on not only the amount of Mn oxide dissolved, but also the extent of readsorption and precipitation of Mn(II). The soluble concentration of associated trace metals following reductive dissolution of Mn(IV) oxides in flooded soils is expected to depend on the concentration of trace metal in the Mn oxide phase, the amount of Mn oxide dissolved, and readsorption or precipitation of the trace metals to the soil. If the retention of each of the trace metals vs. Mn(II) in the soil is different, the solubilization of each of the metals relative to Mn may be significantly different than the concentration of the metal in the Mn oxide particles.

Previous studies on the release of trace metals following reductive dissolution of Mn oxides used mixed redox cells or targeted long-term release using soil columns (Charlatchka and Cambier, 2000; Davranche and Bollinger, 2000). We designed our experiments to simulate irrigated soils that are saturated for a short-term period of up to 5 d. The objectives of our study were to (i) determine if the soluble concentrations of Mn, Cu, Ni, Pb, and Zn are increased under reducing conditions within five days following saturation; (ii) determine if the soluble concentrations of Cu, Ni, Pb, and Zn are correlated to the concentration of soluble Mn under reducing conditions; and (iii) evaluate if the increases in soluble concentrations of Cu, Ni, Pb, and Zn relative to Mn under reducing conditions can be determined from the trace metal contents in the Mn oxide fraction only.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soils
We selected four soils from the Alamosa River basin in Colorado for this study. These soils are classified as the LaJara (coarse-loamy, mixed, calcareous, frigid Typic Endoaquoll) and Mogote (fine-loamy, mixed, calcareous, frigid Aquic Ustorthent) series. We obtained samples for both series from areas that were nonirrigated and uncultivated (Native) and areas that were flood irrigated with Alamosa River water (Alamosa). Connolly (1998) reported on acidification, soil weathering, and metal accumulation and lability in soils from this region. This study showed that while irrigation with Alamosa River water has caused only slight increases in total metal content, soil acidification and alteration of clay mineralogy via accelerated weathering has been significant. Furthermore, labile metal concentrations as measured by DTPA extraction are greater in the soils irrigated with Alamosa River water.

Before sampling, small pits were dug in the specified areas to ensure that the soils to be sampled were representative of the area as compared with the Conejos County Soil Survey soil descriptions (USDA Soil Conservation Service, 1980). Composite samples were taken from the top 30 cm of the surface horizons.

Chemical and Physical Analyses
All analyses used the air-dried (<2 mm) fraction of the soil samples that were sieved using stainless steel sieves and ground using a porcelain mortar and pestle. Routine chemical analyses using the <2-mm fraction were used to characterize each of the four soils. The HF acid method was used to determine total metals (Hossner, 1996). Cation exchange capacity (Rhoades, 1982), pH, calcium carbonate equivalent (CCE; Soil Survey Staff, 1996), particle-size distribution (Gee and Bauder, 1986), and total organic carbon (Mebius, 1960) analyses were performed (Table 1) . A sequential extraction procedure (Miller et al., 1986) was used to determine the amounts of Mn, Fe, and trace metals associated with the water soluble, exchangeable, acid soluble, organic, Mn oxide, and Fe oxide fractions in each of the four soils before reduction experiments. The last step of Miller's procedure for extraction of iron oxides was replaced by the method of Tessier et al. (1979) in which amorphous and crystalline Fe are not distinguished. The chemical reagents and time of shaking used for each step of the sequential extraction are shown in Table 2 . Twenty milliliters of each reagent was added to 0.5 g soil in 50-mL centrifuge bottles. Following each step (after extraction of exchangeable metals), the soil was washed with 20 mL of 0.025 M Ca(NO3)2 for 1 h to eliminate carryover. The percent difference of duplicate samples from sequential extraction analyses was less than 15% for Mn, Zn, Ni, and Cu for each of the fractions quantified, except for Ni in the organic fraction (17%) and Cu in the Mn oxide (22%) and organic fractions (23%).


View this table:
[in this window]
[in a new window]
  		Table 1. Soil chemical properties.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Chemical reagents and shaking times used in the sequential extraction procedure.

 
To study Mn dissolution under aerated conditions, batch extractions in water were conducted at 2:1 water to soil ratio in 250-mL erlenmeyer flasks. The flasks were covered with parafilm, which was then perforated to ensure aeration of the suspensions. The flasks were shaken for 24 h after which the suspensions were allowed to stand for 1 h to allow sedimentation of the soil solids, and then filtered using 0.45-µm-pore-size membranes.

X-Ray Diffraction Procedures
Oriented mounts of the <2-mm fractions were prepared for the identification of clay minerals (Whittig and Wallace, 1986). The X-ray diffractometer was a Scintag Model XDS 2000 XGEN-4000 (Thermo ARL, Ecublens, Switzerland). Box car smoothing was used and Cu K-{alpha}2 stripping was performed using Cu K-{alpha}1 and Cu K-{alpha}2 as 0.1540562 nm and 0.1544390 nm, respectively. Smectite, illite, and kaolinite were identified in each of the four soils, with a greater proportion of 1:1 vs. 2:1 clays found in the irrigated soils.

Redox Column Construction
All redox columns were constructed from sections of Schedule 40 polyvinyl chloride (PVC) 30 cm in length and 5 cm in diameter (Fig. 1) . Platinum (Pt) wire electrode ports were created by drilling two small holes 8 cm apart with the lower port 5 cm from the bottom of the column. The electrodes were constructed using 5-cm sections of Pt wire partially covered with heat shrink tubing and placed into small plastic fittings. These units were sealed into the redox electrode ports using silicone with 2.5 cm of exposed Pt wire protruding into the column.



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 1. Redox column and electrode.

 
The soil was held in the column with a 2-cm-diameter section of 70-µm nylon mesh covered by five 5-cm-diameter pieces of fiberglass screen placed over the PVC cap at the base of the column. A 1-cm section of plexiglass tubing was used as a port of exit, to which a three-way Luer lock stopcock was attached for sampling solution.

All tubing, attachments, syringes, and PVC materials were washed with soap and tap water, acid-rinsed with 0.2% HNO3, rinsed with deionized water two to three times, and air-dried. The platinum wires were removed from the columns and cleaned between experiments with an abrasive detergent and then placed in a 1:1 mixture of ultrapure concentrated HNO3 for 12 to 24 h; this was done to remove any surface contamination. The wires were finally soaked in deionized water overnight.

Calibration of Redox Electrodes
The Pt wire electrodes were individually calibrated before insertion into the redox columns using quinhydrone solutions at pH 4 and 7 (Patrick et al., 1996). We used an Orion double junction reference electrode (Model 9002) (Thermo Orion, Beverly, MA), filled with 4 M KCl outer filling solution and Orion 900002 inner filling solution. Redox measurements were done with an Orion Model 720 pH–mV meter. The pH 4 and 7 quinhydrone calibration solutions were allowed to equilibrate for 1 h before being used, and fresh solutions were prepared each day that calibration was needed. The necessary difference of 176 mV between the quinhydrone solutions at pH 4 and 7 was checked by an Accumet Pt redox electrode (Model 13620115; Fisher Scientific, Pittsburgh, PA) vs. the same reference electrode described above, and this slope was also verified for each of the Pt wires during calibration. The measured slopes of the Pt wires used in the soil columns including calibrations both before and after reduction experiments were all within the range of 168 to 185 mV with a standard deviation of 5.6 mV. Following verification of the slope using the quinhydrone solutions at pH 4 and 7, each of the Pt wires was calibrated to a redox potential (Eh) of 465 mV for the quinhydrone solution at pH 4.0 (23°C) (Patrick et al., 1996). Since the average Pt wire electrode reading in the pH 4 quinhydrone solution was 230 mV, a typical correction value was 235 (465 - 230) mV.

Redox Experiments
The four soil samples were air-dried and then sieved. The <2-mm fraction of each soil was poured slowly into the columns so that it would be packed loosely to simulate cultivated surface soils. Each column contained 650 g of soil. A total of three columns were used for each of the four soils. A 4-L Mariotte glass bottle was used to saturate the columns with distilled water from the bottom at atmospheric pressure. The concentrations of the metals of interest (Cu, Fe, Mn, Ni, Pb, and Zn) in the distilled water were all below 0.005 mg L-1 as determined by inductively coupled plasma (ICP) (IRIS Advantage Dual View ICP; Thermo Elemental, Franklin, MA). Bulk densities and water contents of the soils following saturation are shown in Table 3 . Saturation to a level of 1 cm of ponding above the soil surface took approximately 2 h. The tops of the columns above the 1-cm ponded layer were left exposed to the air to ensure free oxygen diffusion at the surface, as occurs in a field environment. After the columns were saturated with water, redox potential was measured and effluent solution was sampled from the bottom of each column every 12 h for 84 h. The redox potential (mV) readings were taken before the solutions were withdrawn. Only the redox readings from the lower electrode ports are reported here. The mV readings were taken via the Pt wires inserted semipermanently into the columns by a pH–mV meter (Orion 7120) using connections that clipped onto the wires protruding from the columns. The double junction reference electrode was inserted into the ponded water at the top of each column.


View this table:
[in this window]
[in a new window]
  		Table 3. Bulk density and volumetric water content of soil columns.

 
The effluent pH was measured using an initial 10 mL of solution withdrawn using a syringe with a Luer lock connection. Solution was withdrawn slowly to avoid creating excess vacuum in the syringe, which could lead to entry of air bubbles into the soil column. A second portion of 10 mL of solution was then withdrawn in an identical manner and placed through a 0.45-µm-pore-size 25-mm syringe filter that consisted of nylon filter media with polypropylene housing. The filtered solutions were then acidified with HNO3 and analyzed for concentrations of Mn, major cations (Ca, Mg, Na, and K), and trace metals by ICP (IRIS Advantage Dual View ICP). The 20 mL of solution removed from each column was replaced after each sampling by adding water to the top of the columns, to maintain the 1-cm ponded layer.

Statistical Analysis
Linear correlation analysis was completed for soluble Mn concentrations vs. Zn, Ni, Cu, pH, and time by combining data from replicates for each of the four soils.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Rate of Soil Reduction and Release of Manganese
For each of the four soils, only two of the three replicates reached sufficiently low values of Eh to cause Mn release or produced a complete set of samples throughout the 84-h time period. We believe that the inability to replicate Mn concentrations in the three replicate soil columns was poor because of electrochemical conditions (values of Eh and pH), which are only slightly lower than the critical Eh required to destabilize Mn oxides. This was especially apparent for the LaJara–Alamosa soil. Replicates 1 and 2 released Mn into solution after Eh was decreased below approximately 450 mV (Fig. 2) . Replicate 3 did not reach an Eh of less than 450 mV until the 72-h sampling time, and soluble Mn did not increase in response to the decrease in Eh for Replicate 3 within the time frame of our experiments. Our results indicate that once the critical Eh needed for dissolution of Mn is reached, time becomes the limiting factor that determines soluble Mn concentrations. The initiation and rate of Mn oxide dissolution are highly sensitive to both Eh and pH, and replication was difficult because we did not control Eh or pH. The bulk density of the soil columns was very similar across replicates (Table 3). Water content was also similar in most replicates at the beginning of the experiments with a few exceptions (Table 3). Since we did not purge the water of CO2 before column saturation, it is possible that air pockets were present. Water content may have changed during soil reduction due to the production of carbon dioxide gases or entrance of air into soil pores near the base of columns following sampling. These factors may have contributed to differences in Mn oxide dissolution rate and soluble Mn.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2. Soluble Mn as a function of redox potential (Eh) for the LaJara–Alamosa soil.

 
For those replicates that produced an increase in soluble Mn, Eh decreased within 12 to 24 h of saturation for each of the four soils (Fig. 3) . In most instances, Eh remained nearly constant after the initial decrease, and did not decrease substantially with time. Increases in soluble Mn concentration began 12 to 24 h after the decrease in Eh (Fig. 4a, 5a, 6a, and 7a) . In most cases, soluble Mn continued to increase linearly with time through the duration of the experiments, indicating that equilibrium was not reached. Dissolved Mn concentrations in each of the four soils at 84 h exceeded the water quality standards of 0.05 mg L-1 for the USEPA secondary drinking water standard and 0.20 mg L-1 for the Colorado agricultural water standard. These data are consistent with observations of Mn concentrations in excess of water quality standards in this region (Wendlandt et al., 1998). The water-soluble concentrations of Mn from batch equilibrations were 0.013, 0.160, 0.028, and 0.133 mg L-1 for the Mogote–Native, Mogote–Alamosa, LaJara–Native, and LaJara–Alamosa soils (data not shown). Soluble Mn, soluble cations, and pH from the batch experiments were similar to the 24-h column effluents. From these data, the water-soluble Mn concentrations for the Mogote–Alamosa and LaJara–Alamosa soils under aerated conditions exceed the USEPA secondary drinking water standard but are less than the Colorado agricultural water standard. We observed increases in soluble Mn when Eh became less than approximately 450 to 500 mV (Fig. 2–7). This appears to be the critical Eh needed for Mn oxide dissolution in these soils. These values are slightly higher than critical Eh values of approximately 300 mV reported by Patrick and Jugsujinda (1992) and also Charlatchka and Cambier (2000). However, the Eh required to dissolve Mn oxides in our soils is less than the theoretical levels based on the solubility of Mn(IV) oxides such as birnessite (MnO2). Table 4 lists the values of Eh required to produce soluble Mn concentrations ranging from 0.01 to 10.0 mg L-1 at pH values between 6.0 and 8.0 in equilibrium with the Mn(IV) oxide birnessite, based on stability constants from Lindsay (1979). Given that the soluble concentrations of Mn from batch equilibrations were between 0.013 and 0.160 mg L-1 for the four soils (see above), a comparison with Table 4 indicates that the Mn solubility in these soils should be increased above that found under aerated conditions when Eh decreases to less than approximately 450 to 700 mV, depending on pH. Therefore, our values for critical Eh for Mn oxide dissolution are greater than those observed by Patrick and Jugsujinda (1992), but still approximately 100 mV less than thermodynamic predictions. McBride (1994) attributed the discrepancy between theoretical vs. observed critical Eh values for Mn oxides to the presence of mixed valence Mn(III) and Mn(IV) oxides with a range of stabilities. Comparing values for soluble Mn from Table 4 and Fig. 4a, 5a, 6a, and 7a, soluble Mn concentrations in our samples are undersaturated with respect to birnessite at the end of the experiments. This may be the result of rate-limited dissolution of birnessite, reprecipitation of Mn(II) as rhodochrosite (MnCO3), or readsorption of Mn by soil colloids.



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 3. Changes in redox potential (Eh) and pH with time. (a) LaJara–Native soil, (b) LaJara–Alamosa soil, (c) Mogote–Native soil, and (d) Mogote–Alamosa soil. Symbols: +, Eh Replicate 1; {triangleup}, Eh Replicate 2; {circ}, pH Replicate 1; {triangledown}, pH Replicate 2.

 


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 4. Changes in soluble Mn, Zn, Cu, and Ni with time for the LaJara–Native soil.

 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 5. Changes in soluble Mn, Zn, Cu, and Ni with time for the LaJara–Alamosa soil.

 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 6. Changes in soluble Mn, Zn, Cu, and Ni with time for the Mogote–Native soil.

 


View larger version (22K):
[in this window]
[in a new window]
  		Fig. 7. Changes in soluble Mn, Zn, Cu, and Ni with time for the Mogote–Alamosa soil.

 

View this table:
[in this window]
[in a new window]
  		Table 4. Values of redox potential (Eh) needed to produce soluble Mn concentrations of 0.01, 0.1, 1.0, and 10 mg L-1 at pH of 6.0, 7.0, and 8.0 in equilibrium with birnessite. Ionic strength is set equal to 0.01 M to convert Mn activities to concentrations.

 
The highest soluble Mn concentrations after the 84-h time period were observed in the two soils irrigated with Alamosa river water (Fig. 4a, 5a, 6a, and 7a). Differences in the organic carbon content, clay content, and cation exchange capacity (CEC) in the contaminated vs. uncontaminated soils (Table 1) may be partially attributed to variability within each soil type. However, increased weathering in soils in this region irrigated with Alamosa water has been shown to lead to an increase in the relative content of 1:1 vs. 2:1 clays (Connolly, 1998). These processes may be partially responsible for the differences in clay content and CEC of the native vs. contaminated soils. The variability in results among triplicate columns and significant differences in soil properties between the contaminated and uncontaminated soils limits the statistical analysis of the effect of irrigation with Alamosa river water on soluble Mn. The higher observed concentrations of Mn in the Mogote–Alamosa soil cannot be explained in terms of total Mn or Mn oxide content, which is similar for the irrigated and native soils (Table 5) . The LaJara irrigated soil has higher concentrations of both Mn oxide and total Mn, although acid-soluble Mn is less than for the LaJara–Native soil. The higher observed maximum Mn concentrations seem to be related to the lower pH of the contaminated soils. Both the rate of dissolution of Mn oxides (Banerjee and Nesbitt, 1999) and also the equilibrium solubility of Mn oxides (Lindsay, 1979) increase as pH is decreased. Furthermore, the readsorption of dissolved Mn to soil colloids is decreased under acid conditions (Khattack and Page, 1992).


View this table:
[in this window]
[in a new window]
  		Table 5. Sequential extraction and total elemental analysis.

 
Soil pH during the reduction experiments remained relatively constant after 24 h following saturation, for each of the replicates throughout the reduction process, although slight differences in pH between replicates existed in a few cases (Fig. 3). As a result, Mn concentrations are not significantly correlated with pH during reduction experiments (Table 6) .


View this table:
[in this window]
[in a new window]
  		Table 6. Coefficients of correlation for soluble Mn vs. soluble Zn, Ni, Cu, pH, and time for redox column effluents.

 
Soluble Fe concentrations remained low for the duration of the experiments (data not shown). This is consistent with the reported critical Eh values near 100 mV needed to initiate Fe oxide dissolution (Patrick et al., 1996), which was not reached within the time frame of our experiments for any of the four soils studied. Consequently, we do not expect Fe oxides to be a source of trace metals in our experiments.

Solubility of Zinc, Nickel, and Copper following Reduction
Concentrations of Zn and Ni also increased following soil reduction (Fig. 4–7). Correlation analysis reveals that both soluble Zn and soluble Ni were highly correlated with soluble Mn in all four soils (Table 6). Although Mn concentrations did not replicate closely in each of the soils as discussed above, changes in soluble Zn and Ni concentrations paralleled Mn across replicates for each of the four soils. Copper was significantly correlated with Mn in only one of the four soils (Table 6). The lack of significant correlations for Cu could be due to the formation of complexes with dissolved organic matter. The filtration procedure used is not capable of removing small molecular weight dissolved organic matter. Concentrations of dissolved organic matter are expected to change following soil reduction (Charlatchka and Cambier, 2000). We did not quantify changes in dissolved organic carbon concentrations throughout the reduction experiments. The analysis of this potential mechanism, especially with respect to Cu solubilization, requires further experimentation.

Because cation exchange reactions are rapid, cation exchange may have an important influence on both Mn and trace metal solubilities immediately following reduction. The equilibrium between exchangeable and soluble forms of Mn and trace metals may be described by the reaction (Sparks, 1995):

[1]
where Me represents a divalent trace metal and X represents one mole of exchange sites. The Vanselow cation exchange model may be used to describe the relationship between the relative concentrations of Mn and another metal in solution and exchange sites. For exchange between divalent cations, the Vanselow equation simplifies to the following expression:

[2]
where [Me2+] and [Mn2+] are the soluble concentrations in mol L-1 and MeX2 and MnX2 are the concentrations of exchangeable cations in mol kg-1. Based on this equation, the increase in the soluble concentration of trace metal Me relative to Mn should remain constant if the ratio of the two metals on the exchange sites remains constant. This condition is satisfied if the main source of exchangeable Me and Mn after reduction has begun is from reduced Mn oxide, and also if a high proportion of the released Mn and Me are transferred to exchange sites. Under these conditions, the slope of the regression of [Me] vs. [Mn] should be related to the relative content of Me to Mn in the Mn oxide fraction by Eq. [2], with the slope equivalent to the Vanselow exchange coefficient.

To compare the relative increase in the soluble concentration of each of the metals vs. Mn following reduction to the content of each metal to Mn in the Mn oxide fraction, we define the relative solubilization factor (RSF) as:

[3]
where (Me/Mn)slope is the slope from linear regression of Me vs. Mn concentrations in mg L-1 in the column effluents and (Me/Mn)Mn-oxide is the ratio of the metal and Mn concentrations in mg kg-1 in the Mn oxide fraction on a mass basis from sequential extraction (Table 5). Although the Vanselow cation exchange coefficient is expressed on a molar concentration basis and RSF is expressed on a mass concentration basis, the same conversion factor is applied to both the numerator and denominator in Eq. [3] to convert from mass to molar concentrations. Consequently, if cation exchange reactions dominate the retention of Mn and trace metals by the soil following reductive dissolution, then the calculated RSF is related to the Vanselow exchange coefficient from Eq. [2]. A value for RSF of greater than 1.0 would indicate that soil colloids have a greater affinity for Mn vs. the metal of interest (Cu, Zn, Ni, or Pb).

If specific adsorption of trace metals and/or Mn occurs due to surface complexation, then the relative adsorption of Mn vs. trace metals may not be adequately described by an exchange reaction as in Eq. [1], especially if only certain metals of interest form complexes with the adsorbing surface or surfaces. Furthermore, changes in specific metal adsorption from the beginning to the end of reduction will depend on other factors such as changes in pH. Therefore, correlations between soluble Mn vs. other trace metals are not necessarily expected to be linear.

Table 7 reveals that for all of the four soils, the value of RSF for Zn is greater than for Cu or Ni. This indicates that Zn is readsorbed to the soil to a lesser extent than the other metals following reductive dissolution of Mn. Although we identified Pb in the Mn oxide fraction based on sequential extraction (2.86, 2.57, 2.71, and 2.86 mg kg-1 for the LaJara–Native, LaJara–Alamosa, Mogote–Native, and Mogote–Alamosa soils, respectively), Pb concentrations in column effluents were below the instrument detection limit of 5 µg L-1. Since the concentrations of Zn and Pb in the Mn oxide fraction were very similar, and the instrument detection limits for the two metals are also similar, we would expect soluble Pb concentrations to be similar to those measured for Zn in the column effluents if retention of Pb and Zn by the soil was similar. These results indicate that retention of Pb by the soil was greater than Zn. Based on this result and the values for RSF shown in Table 7, retention of the four metals by the soils followed the order Pb > Cu {approx} Ni > Zn. This is for the most part consistent with selectivity sequences for adsorption of these four metals on clay minerals, iron oxides, and soil organic matter reported in the literature (Table 8) . We would expect Cu to have a lower value for RSF than Ni based on the information in Table 8. Although the RSF values are similar for Cu and Ni for the only soil for which both had significant correlations with Mn, Cu solubility may be enhanced by dissolved organic matter. In comparing the selectivity for Ni vs. Zn, Fe oxides actually adsorb Zn more strongly than Ni, whereas soil organic matter has a greater affinity for Ni than Zn. This indicates that organic matter may be more important than Fe oxides in the retention of Ni and Zn in these soils.


View this table:
[in this window]
[in a new window]
  		Table 7. Comparison of the slope of soluble metal (Zn, Ni, or Cu) vs. Mn concentrations in column effluents from linear regression to the ratio of metal (Zn, Ni, or Cu) to Mn concentration in the Mn oxide fraction.

 

View this table:
[in this window]
[in a new window]
  		Table 8. Relative selectivities of soil colloids for metals.

 
The values of RSF for Zn of greater than 1.0 (Table 7) suggest that Mn was readsorbed or reprecipitated to a greater extent than Zn by the soil. However, existing studies reported for Mn adsorption reveal that Mn is adsorbed to a lesser extent than Zn by clay minerals, Fe oxides, and soil organic matter (Table 8). This should result in selectivity of Zn over Mn in our soils and produce values for RSF for Zn as defined in Eq. [3] of less than 1.0. However, Mn is adsorbed more strongly than Zn by Mn oxides. This could be significant if complete dissolution of the Mn oxides was not achieved within the time frame of our experiments. Davranche and Bollinger (2000) found that Pb released on reductive dissolution of birnessite was reabsorbed by remaining Mn oxide particles if Mn oxides were only partially dissolved. It is possible that Mn(II) could be readsorbed to remaining Mn oxide particles. The observed values for RSF of greater than 1.0 for Ni and Cu are also in contradiction with the expected selectivity of Ni and Cu over Mn by soils (Table 8). The values for RSF of greater than 1.0 for Zn, Cu, and Ni correspond to increases in the soluble concentrations of these metals relative to Mn that are greater than anticipated based on the metal contents in the Mn oxide fraction and the expected preference for Zn, Cu, and Ni over Mn by soil colloids.

Another cause of the soluble trace metal concentrations, which are greater than expected based on the metal content in the Mn oxide fraction, could be that metals initially present on exchange sites act as a source of soluble metals once reduction occurs. The total electrolyte concentration (TEC) of the column effluents at each sampling time was calculated by summing the soluble concentrations of Ca, Mg, Na, and K expressed as mmolc L-1. The TEC of the Mogote–Native soil increased from 29 mmolc L-1 after 12 h to 393 mmolc L-1 at the end of the redox experiments (Table 9) . This value of TEC is nearly as high as the TEC of the 0.5 M Ca (NO3)2 reagent used for the extraction of exchangeable cations, with a TEC of 1000 mmolc L-1. Therefore, with the high TEC of the Mogote–Native soil columns, a significant proportion of the exchangeable metals are expected to be released into solution. Based on the bulk density of the soils in the redox columns and the water content of these soils at saturation (Table 3), we can calculate the maximum possible concentration of metal in solution in mg L-1 resulting from the complete displacement of exchangeable metals in the soil columns by multiplying the values given in Table 5 for exchangeable metals by 1.75. For the Mogote soil, it is apparent that the release of exchangeable Zn and Ni into solution could account for a high fraction of the observed soluble metal concentrations at the end of the column experiments. The average TEC of duplicate columns at the end of the experiments for the Mogote–Alamosa, LaJara–Native, and LaJara–Alamosa soils were 7.3, 5.4, and 4.9 mmolc L-1. Although these values of TEC are not nearly as high as for the Mogote–Native soil, the TEC was approximately doubled for the LaJara–Native and LaJara–Alamosa soils, and increased by a factor of 3 for the Mogote–Alamosa soil from the beginning to the end of the redox experiments. These increases in TEC, which are mostly due to increased soluble Ca, are expected to displace at least a small concentration of Ni and Zn from cation exchange sites. Increases in soluble Ca following reduction have been attributed to the displacement of exchangeable Ca from exchange sites by Mn and Fe dissolved during reduction (Phillips and Greenway, 1998) or to the dissolution of calcite by acids produced during the reduction process (Charlatchka and Cambier, 2000).


View this table:
[in this window]
[in a new window]
  		Table 9. Total electrolyte concentration in soil column effluents.

 
Precipitation of Mn(II) as rhodochrosite could also be responsible for the greater retention of Mn vs. the other trace metals. The continuous increase in soluble Mn concentration through 84 h (Fig. 2–5) indicates that the rate of Mn oxide dissolution exceeded that for MnCO3 precipitation. Lebron and Suarez (1999) observed that the precipitation of rhodochrosite was kinetically limited within a time frame of 72 h. The stability of rhodochrosite compared to Zn, Pb, Ni, and Cu carbonates can be evaluated by comparing the equilibrium constants for the following reaction:

[4]

According to equilibrium constants for this reaction, the relative stability of metal–carbonate minerals is Pb >> Zn {approx} Mn > Cu. However, the precipitation of MnCO3 will be favored over other metal carbonates if Mn is present at greater soluble concentrations than the other metals. This is the case for most of our experiments once reduction has occurred. Surface adsorption of Mn to calcite has also been identified as an important mechanism of Mn retention (McBride, 1989). This process can occur even when a solution is undersaturated with respect to rhodochrosite.

The relative solubilization of each of the three metals vs. Mn was different for the four soils (Table 7). Differences in the mineralogy and organic matter content of the four soils could lead to different types of colloids dominating adsorption, which may influence the selectivity for the metals. The relative solubilization of the metals vs. Mn was related to the average soil pH during the redox experiments, with greater RSF values corresponding to higher pH values. In soils with a relatively high pH, reprecipitation of carbonates, especially Mn carbonate, may enhance the retention of Mn vs. the other metals. This mechanism is consistent with the very high RSF value for Zn in the soil with the highest pH (Mogote–Native).


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
A time frame of three to five days was sufficient to release Mn at concentrations exceeding USEPA and State of Colorado water quality standards in saturated, unmixed columns designed to simulate waterlogged irrigated soils. Soluble Mn concentrations continued to increase throughout the duration of the experiments, suggesting that equilibrium conditions were not reached. The release of Zn and Ni was highly correlated with Mn in all four soils. Correlations of Cu with Mn were not nearly as pronounced.

The relative solubilization of Pb, Cu, Ni, and Zn following the reduction of metal-bearing Mn oxides was consistent with selectivity sequences for metal adsorption to soil clay minerals, Fe oxides, and organic matter. Lead released by dissolution of Mn oxides was apparently strongly retained by the soil. Increases in the soluble concentrations of Zn and Ni were greater than expected based on the dissolution of metal-bearing Mn oxides in the soil. This could be due to exchangeable Zn and Ni acting as a source of soluble metals that could be released after the electrolyte concentration is increased following reduction. The apparent greater retention of Mn vs. Ni and Zn may also be the result of precipitation of the Mn carbonate mineral rhodochrosite and/or adsorption of Mn to calcite.


   ACKNOWLEDGMENTS
 
X-ray diffraction analysis was performed with the assistance of Dr. Wendy Harrison at the Colorado School of Mines mineralogy laboratory.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2003 32: 1167-1172. [Full Text]  




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 Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Green, C. H.
Right arrow Articles by Kelly, E. F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Green, C. H.
Right arrow Articles by Kelly, E. F.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Green, C. H.
Right arrow Articles by Kelly, E. F.
Related Collections
Right arrow Redox Processes
Right arrow Heavy Metals
Right arrow Soil Chemistry

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