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


     

Published online 4 January 2008
Published in J Environ Qual 37:174-181 (2008)
DOI: 10.2134/jeq2007.0148
© 2008 American Society of Agronomy, Crop Science Society of America, and 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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Leader, J. W.
Right arrow Articles by Reddy, K. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Leader, J. W.
Right arrow Articles by Reddy, K. R.
Agricola
Right arrow Articles by Leader, J. W.
Right arrow Articles by Reddy, K. R.
Related Collections
Right arrow Wetlands and Aquatic Processes
Right arrow Phosphorus
Right arrow Biogeochemical Processes

TECHNICAL REPORTS

Wetlands and Aquatic Processes

Phosphorus Sorbing Materials: Sorption Dynamics and Physicochemical Characteristics

J. W. Leadera, E. J. Dunneb,* and K. R. Reddyb

a Wor-Wic Community College, 32000 Campus Drive, Salisbury, MD 21804
b Wetland Biogeochemistry Lab., Soil and Water Science Dep., Univ. of Florida/IFAS, 106 Newell Hall, PO BOX 110510, Gainesville, FL 32611

* Corresponding author (ejdunne{at}ufl.edu).

Received for publication March 23, 2007.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
The effectiveness of various management practices to reduce phosphorus (P) loss from soil to water can potentially be improved by using by-product materials that have the capacity to sorb phosphorus. This study evaluated the P sorption and desorption potential, and the physicochemical characteristics of various phosphorus sorbing materials. Twelve materials were selected and P sorption potentials ranged between 66 and 990 mg kg–1. Iron, and calcium drinking water treatment residuals (DWTRs), a magnesium fertilizer by-product, aluminum, and humate materials all removed substantial amounts of P from solution and desorbed little. Humate had the highest maximum P sorption capacity (Smax). Materials which had a low equilibrium P concentration (EPC0) and a high Smax included aluminum and humate by-products. In a kinetic study, the Fe-DWTR, Ca-DWTR, aluminum, and magnesium by-product materials all removed P (to relatively low levels) from solution within 4 h. Phosphorus fractionation suggests that most materials contained little or no P that was readily available to water. Sand materials contained the greatest P fraction associated with fulvic and humic acids. In general, materials (not Ca-DWTR) and magnesium by-product were composed of sand-sized particles. There were no relationships between particle size distributions and P sorption in materials other than sands. The Ca- and Fe-DWTR, and magnesium by-product also contained plant nutrients and thus, may be desirable as soil amendments after being used to sorb P. Further, using Ca-DWTRs and Fe-DWTRs as soil amendments may also increase soil cation exchange and water holding capacity.

Abbreviations: AAS, absorption spectrophotometry • DWTRs, drinking water treatment residuals • EPC0, equilibrium P concentration • P, phosphorus • SRP, soluble reactive phosphorus • TP, total P


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
ADVANCED removal of phosphorus (P) from P-impacted waters is often required when waters are lost to freshwater bodies (Genz et al., 2004). Further, many watersheds within the U.S. and around the world are undergoing legislative processes that restrict the amount of P that is lost to receiving water bodies. Therefore, more effective methods, in addition to the many management practices already in place, are and will be needed to help meet increasingly stringent water quality criteria. Thus, there is an increasing interest in using non-toxic by-product materials that contain P-sorbing components such as iron, aluminum, calcium, magnesium, organo-metallic complexes, and clays (Arias et al., 2003). In general, ideal P-sorbing materials for use in various practices should be free, non-toxic, industrial by-products, generated locally, widely available, and potentially useful as soil amendments once saturated with P (Mæhlum and Roseth, 2000; De-Bashan and Bashan, 2004; Kvarnström et al., 2004). Phosphorus-sorbing materials are used in practices such as constructed wetlands to increase P retention and decrease wetland landscape footprints (Drizo et al., 1999; Brooks et al., 2000; Ann et al., 2000; Brix et al., 2001). Phosphorus-binding materials such as drinking water treatment residuals (DWTRs) and other P-binding materials can be incorporated or surface applied to agricultural soils (Callahan et al., 2002; Novak and Watts, 2004; Dayton and Basta, 2005; Silveira et al., 2006) or applied to edge of stream or ditch soils to mitigate P loss and protect adjacent waters (Dayton and Basta, 2005; Penn and Bryant, 2006). Phosphorus-sorbing materials such as alum are also incorporated into manures and biosolids to decrease P solubility (Moore and Miller, 1994; Elliot et al., 2002a).

Brix et al. (2001) investigated P-binding capacity of various artificial media such as light expanded clay, crushed marble, diatomaceous earth, vermiculite, and calcite. Using laboratory-scale experiments, they found that calcite and crushed marble had greatest P-binding capacities (25,000 and 5000 mg kg–1, respectively). Shale and fly ash were also found to have P adsorption capacities; however, these were much lower (350 and 730 mg P kg–1, respectively) (Drizo et al., 1999). Substrates such as alum (Al2(SO4)3.4H20) and ferric chloride (FeCl3) which are often components of DWTRs, can reduce P loss from soil to water by up to 85%. Agyin-Birikorang et al. (2007) reported that agricultural soil amended with Al-DWTR reduced water-soluble P concentrations by about 60% with this being stable for 7.5 yr. Materials such as alum, DWTRs, and gypsum applied to high impacted cattle loafing areas prior to rainfall can also reduce P concentrations in runoff (Penn and Bryant, 2006). Further, a laboratory study also found that mixing DWTRs with agricultural soil reduced virtually all P loss from soil to water (Silveira et al., 2006). Hylander and Simàn (2001) suggested that P contained in land applied furnace slag, which was previously used for sorbing P during wastewater treatment, was also available for plant P uptake.

In this study, we sought out several by-product materials that were widely available in the U.S. (particularly in Florida) to: (i) evaluate material P sorption and desorption potential using laboratory batch experiments, and; (ii) determine material physicochemical characteristics. The overall objective of these laboratory experiments was to screen for suitable by-product materials that could potentially be used with present management practices to increase their effectiveness at retaining P. Further, the experiments presented herein were a template for subsequent experimenting at the mesocosm-scale, findings of which are presented in Leader et al. (2005).


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results and Discussion
 Conclusions
 REFERENCES
 
Material Collection
Phosphorus-sorbing materials were gathered from a variety of sources. One source was the Southern Waste Exchange (SWIX). Waste and by-product materials from regional industries and organizations are listed with SWIX to find potential users, and thereby reduce the waste stream burden on regional landfills. Twelve materials were selected for initial laboratory experiments; however, as experiments were being undertaken, some materials were eliminated due to performance problems or lack of availability. Consequently, not all twelve materials described in Table 1 were used in every experiment detailed in this manuscript. See Tables and Figures herein for details.


View this table:
[in this window]
[in a new window]

  		Table 1. Description of materials initially considered for phosphorus sorption experiments and characterization.

 
A calcium-based (Ca) drinking water treatment residual ("Ca-DWTR") and an iron-based DWTR ("Fe-DWTR") were obtained from nearby public utilities (Gainesville Regional Utilities, Gainesville, FL and Hillsborough River Water Treatment Plant, Tampa, FL) in the form of a wet sludge. A magnesium (Mg) fertilizer by-product ("SuperMag by-product") was acquired directly from the manufacturer (SuperMag, Bradley FL) as a dry dust. A dried humate by-product of titanium mining ("humate product") was acquired directly from a processing research company (SIGARCA, Inc., Gainesville FL).

Three sands were obtained from a local mine in Grandin, Florida (Florida Rock, Inc., Jacksonville, FL). Two of the sands are mined and separated for use in producing concrete ("sand- concrete") or masonry ("sand- masonry") products. The third sand had iron coatings ("sand- coated") and was not suitable for producing concrete. An organic soil ("organic soil") was obtained from a wetland restoration project. Organic soils have high value for use in wetland restoration and are not a by-product. However, it was used in some (but not all) experiments, as organic soils typically occur in natural treatment wetlands, which are sometimes used to treat P-laden waters. Two aluminum materials were obtained from the SWIX waste exchange. The first aluminum material ("Al #1") was mostly alumina or Al-oxide (Al2O3). It was later determined that it had been submitted to the waste exchange in error and in fact was not a by-product, but a material with high market value for producing aluminum. The second Al material (Al #2) was an alumina-coke mixture that was a by-product and currently being land filled. A dried Fe-sludge processed in a vortex dehydrator ("dehydrated Fe-sludge") was obtained directly from the processor (SOS Technologies, Inc., Mission KS) after contacting SWIX. Finally, a by-product of ship repair operations ("sandblast grit") listed with SWIX was obtained for P sorption studies (Compliance and Remediation, LLC, Mango FL).

Phosphorus Sorption Properties
The most important property to test for was the ability of materials to remove P from solution (sorption) and not readily release that P into solution (desorption). Sorption of P was defined as removing P from solution by association with the solid phase (substrate material). Desorption was defined as releasing some or all of that P from the solid phase back into solution when equilibrated with P-free solution. For the single-point sorption isotherms, 2 g (dry weight equivalent) of material was placed in a centrifuge tube with 20 mL of 100 mg P L–1 (as potassium phosphate) solution prepared in a 0.01 mol L–1 potassium chloride (KCl) matrix. Tubes were then placed on an end-to-end shaker for 24 h. After 24 h, samples were centrifuged at 6000 rpm for 10 min. The supernatant from each tube was vacuum filtered through a 0.45-µm membrane filter into a scintillation vial, acidified with one drop of concentrated sulfuric acid (H2SO4) and refrigerated at 4°C until analyzed for P (within 24 h). All sample solutions were analyzed using the common molybdate colorimetric test (Murphy and Riley, 1962) on a spectrophotometer (UV-160; Shimadzu, Columbia, MD). The difference between the amount of P left in solution and amount in the original solution was described as P sorbed by material.

For desorption isotherms, 20 mL of P-free 0.01 mol L–1 KCl solution was added to 2 g dry weight of material, which was the wet residue left in the tube after centrifuging during sorption isotherms. Tubes with residue and solution in them were put back on shaker for 24 h, after which they were centrifuged and vacuum filtered as described above. The filtered desorption supernatant was then analyzed for P. Taking into account the P content of the wet residue, the mass of P desorbed per kilogram of substrate was calculated based on the P content of the desorption supernatant.

Multipoint P sorption isotherms were also completed to derive theoretical sorption maximum (Smax) and equilibrium P concentration (EPC0) values for materials. These P retention parameters can indicate the potential P removal performance of materials under various P loading and solution P concentrations. The method is the same as described above for the sorption isotherms; however, groups of sample replicates were equilibrated with different P concentrations (0, 0.5, 1, 2, 5, 10, 20, 50, and 100 mg P L–1). Samples were also analyzed similarly but with semi-automated colorimetry according to USEPA Method 365.1 (USEPA, 1993). The P retention parameters were derived using the Langmuir equation as described by Reddy et al. (1998a). The desired parameters and coefficients were derived using two different graphs of the multipoint isotherm data. First, the mg of P adsorbed per kg of material (S') was plotted against P left in solution after 24 h of equilibration (C24). The linear portion (linear regression yields r2 > 0.94) of this curve is described by Eq. [1].

Formula 1[1]
The slope of that line (kd) is a linear sorption coefficient. The y-intercept of the line represents the amount of P originally sorbed to the material (S0). The total amount of P sorbed to the solid phase (S) is equal to S' + S0. The x-intercept of the line represents the point at which adsorption and desorption are equal and is called the "equilibrium P concentration" (EPC0). The second graph plots C24/S against C24 and is the linear form of the Langmuir equation (Eq. [2]). The theoretical maximum amount of P the material can adsorb (Smax) is the inverse of the slope. By calculating Smax a constant (kb) that is related to bonding energy can also be determined.

Formula 2[2]

Kinetic Study
Samples were equilibrated with P solutions for different lengths of time to determine relative rates of P removal. Determining relative rates is useful, as it can indicate how long the material needs to be in contact with P for effective retention. The data points generated after 24 h equilibration time were taken from the previous multi-point isotherm experiment. This data was incorporated with data generated with samples that were equilibrated for 0.5, 1, 2, 4, and 12 h. Phosphorus remaining in solution was plotted against equilibration time and regression analysis was undertaken to describe the rate of P removal and to provide insight into the type of P removal process for each material.

Inorganic Phosphorus Fractionation Analysis
Materials were also characterized for their different P fractions. Fractionating these materials provides an insight into the stability of P. Samples (0.5 g dry weight equivalent) were used in an inorganic P sequential fractionation scheme (Reddy et al., 1998b). However, this scheme was modified, as an initial substrate to solution ratio of 1:50 on a weight to volume basis was used. Sequential extractants used were 1 mol L–1 KCl, 0.1 mol L–1 NaOH, and 0.5 mol L–1 HCl. The NaOH extracts were analyzed for both soluble reactive P (SRP) and total P (TP), whereas the KCl solution was analyzed for SRP. The residue from the final HCl extraction was combusted at 550°C for 4 h and dissolved in 6 mol L–1 HCl before P analysis. All solutions were analyzed with semi-automated colorimetry according to USEPA Method 365.1 (USEPA, 1993).

Metal Analysis
Sample extracts were analyzed for metal content (Ca, Mg, Fe, and Al). Sample replicates (0.5 g dry weight equivalents each) were extracted with 0.2 mol L–1 ammonium oxalate for Fe and Al (Sheldrick, 1984), and 1 mol L–1 HCl (3 h equilibration on shaker) for Ca and Mg. Samples were filtered (0.45 µm) and extracts were analyzed for Fe, Al, Ca, and Mg by atomic absorption spectrophotometry (AAS) (Model 460; PerkinElmer, Waltham, MS) using USEPA Methods 236.1, 202.1, 215.1, and 242.1 respectively (USEPA, 1993).

Particle-size Distribution Analysis
Materials were analyzed for particle-size distribution. This standard soil analysis was done on even non-soil materials to compare their relative particle sizes and theoretical surface area based on a standard analytical method. This method yields the relative percentage of three soil size fractions: sand, silt, and clay, based on theoretical particle-size diameter (USDA-NRCS, 1992). Clay-sized particles are involved in P sorption (Axt and Walbridge, 1999); therefore, particle size was a relevant physicochemical property to determine the P sorbing materials. When using this analytical method, soluble salts can be counted as clay particles and thus, introduce error in interpreting results. The theoretical specific surface area of the materials was determined by assuming that sand particles have a specific surface area of 30, silt 1500, and clay 3,000,000 cm g–1. The humate product could not be analyzed in this way due to the high organic content and the internal porosity of hardened aggregates; therefore, results are not presented in Table 2 .


View this table:
[in this window]
[in a new window]

  		Table 2. Particle size distribution of materials{dagger} as determined by standard soil method. Phosphorus sorption data are presented for comparison.

 

   Results and Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
Conclusions
 REFERENCES
 
Phosphorus Sorption Properties
Phosphorus sorption and desorption isotherms indicate the potential for candidate materials to retain and release P to solution (Brix et al., 2001; Zhu et al., 2003; Seo et al., 2005). Substrate P sorption potentials (Table 3 ) ranged from a low of 66 mg kg–1 (sand-concrete) to a high of 990 mg kg–1 (humate product). The maximum amount of P available for sorption in this experiment was 1000 mg kg–1. Several samples adsorbed nearly 100% of this P. The aluminum, iron, calcium, magnesium and humate materials all removed substantial amounts of P from solution and desorbed relatively little P when equilibrated with P-free solution (Table 3). Elliot et al. (2002b) investigated the effect of DWTRs on P solubility and leaching from soils amended with biosolids. They suggested that the ability to reduce P by DWTRs was greatest using Al, then followed by Ca-DWTR, Fe-DWTR, and hematite.


View this table:
[in this window]
[in a new window]

  		Table 3. Sorption and desorption characteristics of materials. Values represent means ± one standard deviation.

 
In our study, the organic soil desorbed the greatest amount of adsorbed P, whereas the concrete-sand removed relatively little P from solution suggesting that this material may not be a useful material to amend soils, manures, and/or water to increase P retention and reducing P solubility. Brix et al. (2001) reported that for 13 sands surveyed, Ca content was an important characteristic, as it controlled P retention.

For each substrate, the mass of P adsorbed to the substrate (S') was plotted against the concentration of P left in solution after 24 h of equilibrating (C24). The graphs of S' versus C24 from the multipoint P isotherms are presented in Fig. 1 . The concrete and masonry sands had the lowest Smax values (Fig. 1; Table 4 ). The sandblast grit, organic soil, coated sand, and aluminum materials had Smax values that ranged between about 200 and 630 mg P kg–1. The humate product had the greatest Smax value (2000 mg kg–1). Zvomuya et al. (2006) suggest that when agricultural soils containing high amounts of P are amended with alum and ferric chloride, P solubility decreases. Using laboratory experiments as a template, alum was then applied to soil at a rate of 1.3 kg m–2 at the field scale. Amending with alum reduced soil P solubility by 85%.


Figure 1
View larger version (10K):
[in this window]
[in a new window]

  		Fig. 1. Multi-point phosphorus sorption isotherms for eight materials showing log trend lines and their associated r2 values. Error bars represent ± one standard deviation. During the multi-point phosphorus sorption isotherm experiments we did not have Fe-DWTR, Ca-DWTR, or SuperMag materials. Dehydrated Fe-sludge was not locally available; therefore, we do not present results.

 

View this table:
[in this window]
[in a new window]

  		Table 4. Linear sorption coefficients (kd), equilibrium P concentrations (EPC0) (with associated r2), P sorption maximums (Smax), and constants related to bonding energy (kb) (with associated r2) calculated for several materials{dagger} from the multi-point sorption isotherm data ranked from lowest to highest Smax.

 
A low EPC0 is an indicator that material will continue to adsorb SRP even when solution SRP concentrations are low (but not lower than EPC0) (Reddy et al., 1999). Materials which had a low EPC0 and a high Smax included the aluminum and humate by-products (Table 4). This suggests that these materials would be suitable for surface application or incorporation with soils and co-blending with manures and biosolids to reduce P solubility. They may also be effective if used with wetland systems to retain P from incoming waters such as agricultural surface runoff, subsurface runoff, and or wastewaters. For example, Leader et al. (2005) suggested that Fe-DWTRs and Ca-DWTRs may be more effective in retaining SRP when used as a co-treatment after, rather than before treating waters using constructed wetlands. In our study, the organic soil had the second highest Smax value and also had the greatest EPC0 value; thus, having a high potential to retain P, but also a high potential to release P. The organic soil had the lowest kb value. This coincided with the greatest P desorption, supporting the premise that kb is related to bonding energy, and that the wetland organic soil itself is not inherently effective at removing and retaining P from solution. Conversely, the masonry sand and aluminum #1 materials had the highest kb value and a low desorption concentration. Elliot et al. (2002b) indicated that when triple super phosphate was applied to soils and those soils were subsequently amended with AL-DWTR (56 Mg ha–1), P loss from soil was less than 1%. In the absence of applying the DWTR, P loss from soil to water was about 25% of the P applied.

In our laboratory experiments, which had a short timescale, P sorption is expected to be primarily inorganic. The four main inorganic P reactions are precipitation with Al, Fe, Mn, Ca, or Mg; anion exchange; reaction with hydrous oxides; and fixation by silicate clays. We expect that the primary P-binding process was precipitation, as both P and P-binding materials were in high quantities. Precipitation is often described as the slow phase of P sorption, which includes diffusion and precipitation. It occurs when a critical concentration for nucleation is exceeded (Rhue and Harris, 1999).

Kinetic Study
Results of the P sorption kinetic study are shown in Fig. 2 . The Fe-DWTR material removed P so rapidly (within 0.5 h) from solution that an adequate regression curve could not be calculated; therefore, we do not present results. Previous studies have found that management practices such as constructed wetlands that use P-binding materials like FeCl3 were more effective at reducing water column P relative to using materials such as Al2SO4 (Ann et al., 2000). However, Fe-bound P can become available to overlying waters when soil-water conditions are anaerobic (Patrick and Khalid, 1974), whereas Al-bound P does not (D'Angelo, 2005).


Figure 2
View larger version (11K):
[in this window]
[in a new window]

  		Fig. 2. Kinetic study showing phosphorus remaining in solution after various equilibration periods with seven materials. Error bars represent ± one standard deviation. The sand-masonry was eliminated as a potential candidate material due to low P-sorption capacity, therefore it was not tested further. The organic soil had too high an in situ value in natural wetlands; therefore we did not view it as a readily available by-product. The Al#2 and sandblast grit materials were not tested due to potential toxicity concerns. The dehydrated Fe-sludge was not tested, due to it not being locally available.

 
The Ca-DWTR, aluminum #1, and SuperMag materials all removed P from solution very quickly (about 4 h) to relatively low levels. Equations for regression curves shown in Fig. 2 are presented in Table 5 . The r2 values indicate a good fit with all of the regression curves derived, except for concrete sand. Concrete sand performed poorly at removing P from solution, irrespective of equilibrating time. Equations were all negative log, power, or exponential, suggesting that P was removed from solution by precipitation reactions. Precipitation with Al and/or Fe can result in forming aluminum hydroxyl phosphate (Alx[OH]y[PO4]2) and ferric hydroxyl phosphate (Fex[OH]y[PO4]2) with discrete forms precipitating at low pH (Ann et al., 2000).


View this table:
[in this window]
[in a new window]

  		Table 5. Regression equations and r2 values of regression curves, shown in Fig. 2, for phosphorus sorbing materials{dagger}. In the equations, y represents concentration of phosphorus remaining in solution, and x represents equilibration time.

 
Inorganic Phosphorus Fractionation
Materials analyzed had a range of initial P amounts and differences in the amount of P in the five fractions (Table 6 ). The KCl fraction is considered labile or readily bioavailable P (O'Connor and Sarkar, 2000). Most materials, including, Fe-DWTRs, contained little bioavailable P suggesting that materials would not readily release P to water.


View this table:
[in this window]
[in a new window]

  		Table 6. Inorganic phosphorus fractions of materials. Values represent mass of P in that fraction (if above detection limit) per kg of material ± one standard deviation.{dagger}

 
The SRP in the NaOH-extracted solution is believed to represent Fe- and Al-bound P, since these metals tend to release P at elevated pH levels (Hieltjes and Lijklema, 1980). Sands contained the largest fraction associated with Fe and Al (Fig. 3 ). The Fe-DWTR did not have a large P fraction stored as Fe- and Al-bound P (about 3% of TP). However, it did have the largest amount (58%) of P thought to be alkali-extractable organic P (NaOH-Po), which is P associated with fulvic and humic acids (Reddy et al., 1999). Elliot et al. (2002b) reported that in a Fe-DWTR, about 30% of the P was associated with a NaOH fraction. In industry, Fe-DWTR is commonly referred to as "iron humate" since it is used to remove dissolved organic material from drinking water.


Figure 3
View larger version (23K):
[in this window]
[in a new window]

  		Fig. 3. Inorganic phosphorus fractions of tested materials. The total mass (mg) of phosphorus per mass (kg) of material are provided above each column. The NaOH-Pi represents the inorganic P in the NaOH extracted solution. The NaOH-Po represents the amount of organic phosphorus in the NaOH extract. The inorganic phosphorus was calculated as the difference between total P and soluble reactive P in the NaOH solution.

 
The HCl fraction is assumed to represent Ca- and Mg-bound P, since these metals tend to release P under low pH conditions. Furthermore, at mid to high pH ranges (6–8.5) Ca compounds can precipitate (Ann et al., 2000). The humate product and SuperMag had the greatest concentrations of P assumed to be Ca- and Mg-bound and these were a large proportion of total P (Fig. 3). In comparison, the Ca-DWTR did not have a comparatively high HCl fraction.

Residual P is normally regarded as unavailable organic P, but may also include mineral bound P that could not be extracted (Reddy et al., 1999). The coated sand, Fe-DWTR, and humate product each had relatively high levels of this P fraction (Fig. 3). With the exception of the Fe-DWTR and the humate product, materials had relatively low levels of organic matter. This suggests that at least part of the residual P fractions were composed of crystalline or tightly bound inorganic P forms, which were only released on combustion.

The coated sand, Fe-DWTR, and humate product had the highest sums of P fractions coincident with their ability to remove P from solution. This may be counter-intuitive, but not uncommon that materials already high in TP may still have a great potential to adsorb P from solution.

Extractable Metals
The concrete and masonry sands had relatively small amounts of oxalate-extracted Fe and Al, and HCl-extracted Ca and Mg, for P sorbing (Table 7 ). In comparison, the coated sand had higher Fe and Al levels, which may have contributed to its better P removal performance (Table 4).


View this table:
[in this window]
[in a new window]

  		Table 7. Extractable metal characteristics of phosphorus sorbing materials{dagger}. Values represent mean ± one standard deviation.

 
The organic soil had very high HCl-extractable Ca and high Mg, as well as high oxalate-extractable Fe implying that on flooding, P bound to ferric forms (Fe3+) could be released, as under reduced soil redox conditions, Fe3+ is transformed to Fe2+. Phosphorus bound to Ca and Mg would be affected by changes in soil porewater pH conditions (Moore and Reddy, 1994) rather than changes in redox.

The aluminum material #1 had high levels of oxalate-extractable Al; however, the aluminum material #2 had about twice the amount of oxalate-extractable Al and much greater concentrations of Fe and HCl-extractable Ca. Aluminum #1 material was a refined material (alumina), whereas the aluminum #2 material was actually an alumina-coke mixture that was a waste product of the refining process and thus had Fe, Al, and Ca potentially available for sorbing P.

The humate product had high concentrations of oxalate-extractable Al, but also contained considerable amounts of Fe and Ca (Table 7). The SuperMag material contained the highest level of HCl-extractable Mg, but also contained high concentrations of Ca. The Ca-DWTR had the greatest amounts of HCl-extractable Ca and Mg. Calcium is used as a water softener to remove hardness in water treatment technology (Elliot et al., 2002b). The Fe-DWTR contained the greatest concentrations of ammonium oxalate-extractable Fe (but also large amounts of Ca). The high Fe content is not surprising, as this DWTR is a by-product of ferric sulfate, which is typically used as a coagulant for flocculating water during water purification (Elliot et al., 2002b).

Comparisons were made between moles of individual metals, pairs of metals, and all metals versus P sorption parameters. No significant relationships were found. This is not surprising, since materials had heterogeneous physicochemical characteristics. Drizo et al. (1999) suggested similar, as they found few relationships between P sorption and substrate physicochemical properties (pH, cation exchange capacity, hydraulic conductivity, porosity, and surface area). Differences between P sorption were better described using direct comparison of P sorption data.

Particle-size Distribution
In soil samples the three size fractions are operationally described as sand (0.05 to 2.0 mm), silt (<0.05 and >0.002 mm) and clay (<0.002 mm). These designations were applied to materials, including non-soils. The results of the particle-size distribution analysis for some of the materials are presented as proportions of theoretical size fractions in Table 2. In general, materials (not Ca-DWTR and SuperMag) were composed of sand-sized particles. The Ca-DWTR was dominated by silt-sized particles, whereas the SuperMag by-product had similar proportions in sand, silt, and clay. Fine materials such as the Ca-DWTR had increased surface areas; therefore, they potentially have a greater capacity to sorb P, relative to coarser grained materials (Drizo et al., 1999).

There were no relationships between particle size distributions and P sorption in materials other than sands. There were good relationships (linear regression r2 > 0.9) between theoretical specific surface area and P sorption by the three sands (concrete, masonry, and coated). The coated sand had a large proportion of its theoretical specific surface area associated with clay fractions. Knowing particle size distribution of materials has relevance, because if these fine grained materials are land applied to agricultural soils they may also increase soil cation exchange capacity and soil water holding capacity, in addition to contributing to decreased solubility of P in soil. Further, they may also be removed in surface runoff.


   Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
REFERENCES
 
Findings suggest that Al #1, Fe-DWTR, Ca-DWTR, SuperMag by-product, humate product, and coated sand removed substantial amounts of P from solution. These materials also had the potential to remove P from solution low in P. Therefore, these materials may be useful for land applying to soil, co-blending with manures and biosolids, and amending constructed wetland soils to reduce P solubility and increase P retention by these practices. The Fe-DWTR, Ca-DWTR, and magnesium materials also contain plant nutrients and thus, may be further desirable as agricultural soil amendments. Further, materials such as Ca-DWTR and FeDWTR may provide additional benefits for soil quality such as increased cation exchange and water holding capacity, which in turn has benefits for agricultural crops.


   ACKNOWLEDGMENTS
 
We would like to thank the Wetland Biogeochemistry Lab. staff for their help with laboratory analyses and we also thank Soil and Water Science staff, and graduate students for their guidance and assistance. This research was supported by the Florida Agricultural Experiment Station and a grant from the Florida Dep. of Agriculture and Consumer Services.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
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
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
V. Cucarella and G. Renman
Phosphorus Sorption Capacity of Filter Materials Used for On-site Wastewater Treatment Determined in Batch Experiments-A Comparative Study
J. Environ. Qual., March 1, 2009; 38(2): 381 - 392.
[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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Leader, J. W.
Right arrow Articles by Reddy, K. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Leader, J. W.
Right arrow Articles by Reddy, K. R.
Agricola
Right arrow Articles by Leader, J. W.
Right arrow Articles by Reddy, K. R.
Related Collections
Right arrow Wetlands and Aquatic Processes
Right arrow Phosphorus
Right arrow Biogeochemical Processes

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