HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Related articles in JEQ Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Bégin, L. Articles by Fortin, J. Search for Related Content PubMed PubMed Citation Articles by Bégin, L. Articles by Fortin, J. GeoRef GeoRef Citation Agricola Articles by Bégin, L. Articles by Fortin, J. Related Collections Other Contaminants Ecological Risk Assessment Other Environmental Contamination Soil Pollution Soil Analysis

TECHNICAL REPORTS
Vadose Zone Processes and Chemical Transport

Evaluation of an Acid Ammonium Oxalate Extraction to Determine Fluoride Resident Concentrations in Soils

Département des Sols et de Génie Agroalimentaire, FSAA, Université Laval, Québec, QC, Canada, G1K 7P4

^* Corresponding author (josee.fortin{at}sga.ulaval.ca)

Received for publication November 14, 2001.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
Fluoride depositions near aluminum smelters and other fluoride-emitting plants can lead to fluoride accumulation in soils, which constitutes a risk for ground water contamination. This study was conducted to investigate the capacity of a 0.2 M acid ammonium oxalate solution to selectively and quantitatively extract fluoride accumulated in soils. The recovery of fluoride added to three soils was evaluated following 7- to 28-d incubations. Oxalate extraction was also compared with a total fluoride extraction method, using oxalate-extractable fluoride (F_ox) and total fluoride (F_tot) accumulation profiles derived from column percolation experiments. To determine low-level fluoride concentrations without interference from high Al and Fe concentrations, an adapted ion chromatography method was used. Following soil incubations, oxalate extracted 42 to 86% of added fluoride. Recovery varied between soils and, in one soil, increased with added fluoride concentration. Recovery was unaffected by incubation time. Maximum recovery was obtained in a soil high in amorphous Fe and Al, low in clay, and free of carbonate. Lower recoveries were obtained in soils with higher clay or carbonate contents. Only 4 to 8% of F_tot was extracted in untreated samples using F_ox, which suggests a high selectivity of this method for added fluoride. In percolation experiments, the use of F_ox reduced considerably the background noise associated with F_tot for the evaluation of fluoride accumulation profiles. Because of its high selectivity and despite incomplete fluoride recovery, the use of F_ox to determine fluoride resident concentrations in soils may improve environmental monitoring of fluoride accumulation and movement in contaminated soils.

Abbreviations: Al_ox, oxalate-extractable aluminum • Fe_ox, oxalate-extractable iron • F_ox, oxalate-extractable fluoride • F_tot, total fluoride • TISAB, total ionic strength adjustment buffer

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
FLUORIDE IS REGARDED as one of the most important environmental micropollutants (Skjelkvåle, 1994). Inorganic fluorides were part of the first Canadian Priority Substances List (Environment Canada, 1989). It was concluded from subsequent studies that inorganic fluorides entering the Canadian environment from anthropogenic sources have the potential to cause harm to the environment (Environment Canada, 1993, p. 40). A large number of industries emit fluoride in the atmosphere, including brick and tile products manufacture, steel and aluminum manufacture, and the production of phosphoric acid, superphosphate fertilizer, and elementary phosphorus (Loreti, 1988). Coal combustion, glass works (Kabata-Pendias and Pendias, 1992, p. 245), and enamel factories (Debackere and Delbeke, 1981) are also sources of fluoride emissions. In Canada, it is estimated that annual anthropogenic inorganic fluoride releases in the atmosphere exceed 5400 Mg, 75% of which is attributed to aluminum production (Environment Canada, 1993, p. 8). In the vicinity of industrial sources, fluorides are brought to the soil surface by fallout of particulate fluorides and through absorption of gaseous fluorides in rain and snow (Hluchan et al., 1964 [as cited by Groth, 1975]). Applications of phosphate fertilizers, sewage sludges, and some pesticides also bring fluorides to the soil (Kabata-Pendias and Pendias, 1992, p. 245). The majority of any added fluoride becomes fixed by one or more of the soil components (Pickering, 1985). Fluorides may nevertheless represent a risk for ground water contamination, especially under strong acid or alkaline conditions (Wenzel and Blum, 1992). Accurate monitoring of fluoride accumulation and movement in soils is therefore of great importance.

A large variety of methods is available for soil fluoride determination. Commonly used methods include, in decreasing order of extraction efficiency (Supharungsun and Wainwright, 1982): total fluoride (McQuaker and Gurney, 1977), resin-extractable fluoride (Larsen and Widdowson, 1971), water-extractable fluoride (Shaw, 1954), and CaCl₂–extractable fluoride (Larsen and Widdowson, 1971). A variety of other extractants have been used for soil fluoride extraction. Examples are total ionic strength adjustment buffer (TISAB), 1 mM NaOH, and 10 mM HCl (Plüger and Friedrich, 1973); HNO₃, K₂SO₄, ethylenediaminetetraacetic acid (EDTA) + Na-citrate, and KCl (Urumova-Pesheva and Ruseva, 1982); TISAB + water (Thompson et al., 1979); acetic acid (Zimmerman and Bertrand, 1978); 1 M HCl (Eyde, 1983); and 10 mM NaOH, 2 mM Na₂CO₃ + 2 mM NaOH, 3 mM NaHCO₃ + 2.4 mM Na₂CO₃, 5 mM Na₂C₂O₄, and 5 mM KOH + 5 mM K₃PO₄ (Maketon and Tarter, 1984).

As soluble fluoride residing in the liquid phase can only account for a small fraction of added fluoride in most soils, the evaluation of fluoride accumulation in both the solid and liquid phases is necessary to monitor adequately fluoride movement in soils. Most of the above methods, except total fluoride, focus on soluble, mobile, labile, or plant-available fluoride. These methods were not developed in the perspective of quantitative extraction of fluoride added to soils, or demonstration of their ability to do so is lacking. Total fluoride methods have the capacity of extracting all the fluoride present in either natural or contaminated soils. Consequently, they extract all added fluoride. Unfortunately, they also suffer a major drawback. All forms of fluoride are extracted, including fluorides naturally present in the structure of minerals by OH^-–F^- substitution, for example in aluminosilicates (Thomas et al., 1977). According to Norrish (1975), this structural F would not be a ready source of fluoride for the soil solution, so it is of little environmental interest. According to Flühler et al. (1982), accumulation of atmospheric fluoride in soils may remain undiscovered for extended periods of time because the natural fluoride content of soil minerals is relatively high compared with the annual anthropogenic fluoride input, even in severely fluoride-polluted areas. The important reductions of fluoride emissions that occurred since that time can only enhance this problem. Polomski et al. (1982) also stated that the high and variable background of total fluoride in soils can be quite misleading when determination of the degree of contamination is needed. Preliminary work done in this laboratory has shown that important variations of total fluoride concentrations occur in soils at the decimeter scale along the profile and within even a few meters laterally. Because of this, the inclusion of all forms of fluoride in measurements of soil fluoride, including forms that do not contribute to environmental contamination, may only lead to elevated noise in the measurements and may affect the quality of fluoride monitoring in soils. The availability of a method being at the same time more selective than total fluoride and able to extract a strong proportion of added fluoride would constitute an important asset for the environmental monitoring of fluoride in soils.

It is known that fluoride may be retained in soils by different physical and chemical mechanisms. Fluoride may be precipitated as calcium fluoride when CaCO₃ is present in the soil and adsorbed on Fe and Al oxides and hydroxides, aluminosilicates, hydroxyapatite, and organic matter (Pickering, 1985). Several authors have emphasized the importance of Fe and Al compounds, especially Al hydroxides, in fluoride retention in soils (Bower and Hatcher, 1967; Murray, 1984; Omueti and Jones, 1977). Based on the hypothesis that a large part of added fluoride resides on amorphous Fe and Al hydroxides, extraction of the latter with acid ammonium oxalate (McKeague and Day, 1966) may lead to extraction of a substantial part of fluoride added to soils. The similitude between fluoride and phosphate reaction in soils and the fact that acid ammonium oxalate can extract 92% of sorbed phosphates (Lexmond et al., 1982 [as cited by van der Zee et al., 1987]) support this hypothesis. Due to its incapacity to dissolve aluminosilicates and other crystalline minerals, acid ammonium oxalate may also be more selective for added fluoride than total fluoride extraction methods.

The objective of this study was to investigate the capacity of a 0.2 M acid ammonium oxalate solution to selectively and quantitatively extract fluoride added to soils. To do so, fluoride recovery experiments on three different soils and fluoride resident concentration profiles from column percolation experiments were used.

For simplicity, the term "adsorption" used in the present paper is comprehensive and refers to any mechanism that removes fluoride from the soil solution. The definition is similar to the one used by Flühler et al. (1982), but complexation is excluded unless the complexes are themselves removed from the soil solution.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
Soil Fluoride Analysis
Oxalate-Extractable Fluoride
The method described in McKeague and Day (1966) was used for fluoride extraction. Briefly, 0.5 g of air-dried soil ground to pass a 0.25-mm sieve was placed in a 50-mL polypropylene tube. The soil was extracted with 20 mL of 0.2 M acid ammonium oxalate pH 3.0 solution for 4 h in the dark with agitation on a rotative shaker. After centrifugation at 1500 x g for 5 min, 1.00 mL of the supernatant was transferred to a 100-mL polypropylene flask for quantification by ion chromatography, with a method adapted from van den Hoop et al. (1996). To favor dissociation of HF and metal–F complexes and bring the samples to the same pH and conductivity as the eluent used in the determination, 130 µL of 10 M NaOH was added to the flask and the volume completed with deionized water.

Fluoride was quantified with a Dionex 4000i ion chromatograph with a CDM-2 conductivity detector (Dionex Corporation, Sunnyvale, CA). Samples and eluents were lead via a Dionex AG15 anion guard column through a Dionex AS15 separator column and an AMMS-1 anion micromembrane suppressor with a GPM-2 pump module. Sample introduction was automated with a Dionex ASM-3 sampler. Dionex 5-mL vials with filtercaps were used for sample introduction. Preliminary tests showed that use of 0.5-mL vials with filtercaps results in fluoride contamination of the samples, probably due to insufficient rinsing of the filtercap by the sample before injection in the column. A 50-µL injection loop was used. The eluent was a 13 mM NaOH solution, pumped at a flow rate of 1.6 mL min^-1 during 5 min. To remove oxalate and other divalent and trivalent anions from the column, the system was rinsed between each run with a 200 mM NaOH solution for 8 min at the same flow rate. Before the next injection, the system was equilibrated by rinsing with the 13 mM NaOH eluent for 9 min. Total run time was 24 min, including sample introduction. The regenerant was 13.5 mM H₂SO₄, at a constant flow of 3 mL min^-1. The NaOH eluents were prepared from a 10 M NaOH solution free of carbonate. Deionized water was purged with He during 5 min, NaOH was added, and the eluents were purged again with He for 5 min. Eluents were always maintained under an He atmosphere. Dionex ACI-1 computer interface and AI-450 software (Release 3.32) were used for data acquisition. Peak height was used instead of peak area to avoid the possible interference of neighboring organic acids peaks. Peak height was measured manually from printed chromatograms. A straight line between conductivity measured at 3.2 and 5.5 min was used as the baseline because baseline resolution was generally not achieved between consecutive peaks. Eluents, standards, samples, and acid ammonium oxalate solution were prepared with deionized water of resistivity > 18 M cm. Oxalic acid and ammonium oxalate were of reagent ACS grade.

Ion chromatography was preferred to an ion-selective electrode for F_ox determinations, because use of the latter resulted in incomplete fluoride recovery, even when TISAB was used, due to the high amounts of Fe and Al extracted by oxalate and the low fluoride concentrations measured. The addition of NaOH to the samples analyzed by ion chromatography brings their pH to about 12, which prevents metal–F complexation, and makes possible the measurement of low fluoride concentrations even in the presence of Al concentrations exceeding 4 mM (van den Hoop et al., 1996). Several fortified samples were prepared by addition of NaF to different F_ox extracts. Complete recovery of added fluoride confirmed the absence of Al or Fe interference.

The instrumental detection limit (IDL), method detection limit (MDL), and method quantification limit (MQL) of the proposed oxalate extraction–ion chromatography determination method were evaluated as described in Centre d'expertise en analyse environnementale du Québec (2000). The IDL is equal to three times the standard deviation measured on 10 aliquots of a fluoride solution of concentration equal to five times the estimated IDL. The MDL is equal to three times the standard deviation obtained when the entire method, including soil weighing, extraction, dilution, and determination, is applied to 10 subsamples of a soil with a concentration five to seven times the estimated MDL. The MQL is equal to 10 times the latter standard deviation. The IDL, MDL, and MQL determinations were repeated on different samples until the sample concentration met the criteria mentioned above.

Total Fluoride
Total F was determined with the alkali fusion–ion selective electrode method of McQuaker and Gurney (1977). Fluoride concentration of the samples was determined with a Fisher Accumet Selective Ion Analyzer Model 750 (Fisher Scientific Ltd., Nepean, ON, Canada) equipped with an Orion Model 96-09 combination fluoride electrode (Orion Research, Boston, MA). Samples were mixed with TISAB in a 1:1 proportion before determination. The TISAB solution contained 1 M glacial acetic acid, 1 M NaCl, and 4 g L^-1 trans-1,2-diaminocyclohexanetetraacetic acid (CDTA), and was prepared according to Menzies et al. (1993).

Water-Soluble Fluoride
Sixteen grams of air-dried soil was placed in a 100-mL polypropylene tube. The soil was extracted with 50 mL of deionized water for 16 h with agitation on a rotative shaker. After a first centrifugation at 1500 x g for 10 min, the supernatant was decanted and centrifuged again at 25 000 x g for 15 min. Then, 10.0 mL of the supernatant was transferred to a 100-mL polypropylene flask, 130 µL of 10 M NaOH was added, and the volume completed with deionized water. Samples were analyzed by ion chromatography with the method described previously.

Fluoride Recovery Experiments
Soil Sampling
Three soils were used to evaluate the capacity of 0.2 M acid ammonium oxalate to extract quantitatively added fluoride. Soil A was sampled in the spodic horizon of a forested soil (Humicryod). Soil B is a calcareous sand used as filling material at the location of an aluminum smelter. Soil C was sampled from a meadow land in a Hébertville silty clay loam (Humaquept), in the 0- to 30-cm layer. The soils have been chosen to cover different pH values and in a manner such that the principal fluoride retention mechanisms would be covered. The textural and pertinent chemical properties of the soils are presented in Table 1. Texture was determined by the hydrometer method (Day, 1965) and soil organic carbon following Walkley and Black (1934). The pH was measured with a 1:1 soil to water ratio and an Orion Model 8165 combination pH electrode. Oxalate-extractable aluminum (Al_ox) and iron (Fe_ox) were determined following McKeague and Day (1966) and carbonates with an approximate gravimetric method (Goh et al., 1993). In carbonates determination, distinction between calcite and dolomite was made by checking the evolution of weight loss with time.

Oxalate-extractable F and F_tot were also determined on untreated samples of each soil to evaluate the capacity of the F_ox method to selectively extract added F by limiting the extraction of fluoride naturally present in the soil. Total F was exceptionally quantified by ion chromatography, because it was more convenient at that time. Subsequent comparison of the two quantification methods (ion-selective electrode vs. ion chromatography) showed that they give comparable results for F_tot determinations. Before determination, 5 mL of the F_tot extract was transferred to a 100-mL polypropylene flask and 130 µL of 10 M NaOH solution was added. The volume was completed with deionized water. This dilution was needed to prevent separator column overcharge by Cl^- during ion chromatography analysis.

Data Analysis
For each soil, the mean F_ox measured on the three replications of the 0 mg F kg^-1 treatment was subtracted from the concentrations measured on the other samples to obtain the net F_ox accumulation. Linear and quadratic regressions were tested to relate F_ox accumulation to added fluoride. Because the standard deviation of the residuals increased with measured concentrations and was approximately proportional to it, the following procedure was used. A first, ordinary least squares regression was made to obtain estimated F_ox means for each added fluoride concentration. Weights equal to the inverse of the square of those means were attributed to each data point and were used in a second, weighted least squares regression (Freund and Littell, 1991, p. 80–83) using the REG procedure of the Statistical Analysis System (SAS Institute, 2001). The results discussed in this paper are those obtained with the weighted least squares regression. All references to significance levels in the text refer to the P < 0.01 level. As the mean concentration of the samples receiving 0 mg F kg^-1 was subtracted from all measured concentrations, no intercept was used in the model and only the samples to which fluoride was added were used in the regression. To assess the possible effect of incubation time on fluoride recovery, a weighted least squares regression was made on the data obtained in the second series of incubations. Added fluoride, (added fluoride)², and time were used as regressor variables.

Column Percolation Experiments
Column percolation experiments were used to evaluate the potential application of the F_ox method to determine the fluoride resident concentration distribution in the soil following fluoride applications or depositions. For this purpose, final F_ox and F_tot soil distribution profiles as well as F_ox and F_tot net accumulation profiles were compared.

In this paper, the term "resident concentration" refers to the sum of the contributions of both adsorbed resident concentration (C^r_a, mass of solute adsorbed per mass of dry soil) and liquid resident concentration (C^r_l, mass of solute per volume of fluid). Specifically, the measured resident concentration, expressed as a mass to mass ratio (mg F kg^-1 soil), corresponds to (Jury and Roth, 1990, p. 66):

where C^r_t is the total resident concentration (mass of solute per volume of soil), is the soil dry bulk density, and is the volumetric water content.

Soils Location and Sampling
Two soils located at different distances from an aluminum smelter in operation for five years were used for these experiments. The first one is a sandy loam (Cryaquod) covered with grass vegetation, located 1.6 km southwest of the smelter. Selected soil properties are presented in Table 1. A soil column was sampled by driving a 15-cm-i.d. plastic tube into the soil with a sledgehammer to a depth of 60 cm. The tube was carefully manually dug out and brought back to the laboratory, where it was cut into 20-cm-long sections. The section corresponding to the 40- to 60-cm depth was used in the present study (Column 1) to have a soil with a low initial fluoride contamination from the smelter. Adjacent to the sampled column location, a 2-cm-i.d. soil core was sampled to a depth of 90 cm. The core was cut in 5-cm-long sections on which F_tot and F_ox were measured to evaluate the initial soil fluoride concentration profile. The second column was made from Soil B used in the recovery experiment (Table 1). In this case, a repacked soil column was used, because direct column sampling at the sampling site was impossible due to the large quantity of rocks present in the soil. The soil was repacked to a density similar to the one encountered in situ (1.76 g cm^-3). The sampling site for Soil B is located 0.8 km south of the smelter. The reference point for distances corresponds to the point of maximum fluoride deposition on the site of the smelter. This point was determined from the measurements made at 17 sampling points on the site of the smelter (unpublished data, 2001).

Experimental Setup
The soil column system used in the present experiments is shown in Fig. 1 . It consists of a glass bead tension table (Topp and Zebchuk, 1979) on which the soil column is placed. A tension infiltrometer (Reynolds and Elrick, 1991) is placed on top of the column to supply water and solution at a constant rate. Matric potential of the infiltrometer and tension table were adjusted to obtain a water flow rate of 1 cm d^-1. At the beginning of the experiment, water was supplied to the column until steady state was reached. Then, the water was replaced with a KF solution containing 100 mg F L^-1 and the percolation experiment started at the same steady state water flow of 1 cm d^-1. The matric potential of the infiltrometer and tension table were adjusted when necessary to maintain the steady state flow and were always kept negative to avoid saturated conditions.

View larger version (12K): [in this window] [in a new window]   Fig. 1. Column experiment setup.

Data Analysis
The initial F_tot and F_ox concentration profiles of Column 1 were estimated from the 40- to 60-cm portion of the adjacent core sampled. The initial concentration in Column 2 was measured on an untreated sample of Soil B and was assumed to be constant in the entire repacked column profile. The difference between the initial and final concentration profiles obtained with each method was interpreted as the accumulated fluoride profile. Recovery of added fluoride was calculated for both methods from soil sample masses and measured fluoride concentrations.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
Analytical Limits
Figure 2 shows an example of a chromatogram obtained for fluoride analysis by ion chromatography. The fluoride peak appears at t = 3.55 min. The instrumental detection limit of the ion chromatography method obtained is equal to 0.4 µg F L^-1. Detection limit of the fluoride ion selective electrode used is 50 times higher, at 20 µg F L^-1, according to the manufacturer (Orion Research, 1991). The F_ox method detection limit is equal to 0.5 µg F L^-1, which corresponds to a soil concentration of 2 mg F kg^-1, while the method quantification limit is equal to 7 mg F kg^-1. Fluoride depositions measured 1.8 km away from an aluminum smelter reached 7.7 g F m^-2 yr^-1 and resulted in accumulations of 10 to 50 mg F kg^-1 soil yr^-1 depending on the depth considered (Polomski et al., 1982). Typical applications of phosphorus in cultivated soils may represent an addition of 5 to 10 mg F kg^-1 yr^-1 (Gilpin and Johnson, 1980). Consequently, F_ox values in contaminated soils are not likely to be below the method detection limit. Even a low contamination level will rapidly raise the soil concentration above the method detection limit and method quantification limit. The limits of the proposed method are thus quite suitable for environmental determination of fluoride accumulation in soils.

View larger version (10K): [in this window] [in a new window]   Fig. 2. Example of chromatogram obtained for fluoride analysis by ion chromatography, showing the fluoride peak at t = 3.55 min for [F-] = 43.8 µg L-1. The dashed line corresponds to the baseline. Conductivity was offset to 0 µS at t = 0.0 min.

View larger version (21K): [in this window] [in a new window]   Fig. 3. Recovery of added fluoride with acid ammonium oxalate extraction (Fox). The dashed line is the 1:1 line and represents a recovery of 100%.

The pH of the three soils is quite different, ranging from acidic (pH 4.6, Soil A) to alkaline (pH 8.1, Soil B) (Table 1). Fluoride retention through ligand exchange results in a release of OH^- in solution and is favored at low pH (Stumm, 1992, p. 24–25). Above pH 3, fluoride adsorption is commonly reported to decline with an increase of pH at the surface of aluminosilicates (Bar Yosef et al., 1988; Omueti and Jones, 1977) and goethite (Sigg and Stumm, 1981). It reaches a maximum at pH 3.5 to 4.5 on amorphous hydrous iron oxides (Farrah and Pickering, 1986) and at pH 5.5 to 6.5 on oxides and hydroxides of Al (Farrah et al., 1987; Omueti and Jones, 1977) and soils (Barrow and Ellis, 1986; Omueti and Jones, 1977). Yet, the major part of added fluoride is removed from solution in the three soils tested. We can suspect that, besides ligand exchange, at least one other mechanism is responsible for the fluoride disappearance from the solution, especially in Soil B. In this soil, one possibility is surface conversion of calcite (CaCO₃) to fluorite (CaF₂). Fluorite formation was demonstrated by X-ray diffraction analysis when fluoride was used for phosphorus extraction in calcareous soils (Smillie and Syers, 1972).

The oxalate-extractable part of Fe and Al represents mainly the Fe and Al present in the soil as reactive, amorphous hydroxides (Beek, 1979 [as cited by van der Zee et al., 1987]). Oxalate has the ability to remove P sorbed by noncrystalline Fe and Al, to replace P on oxide surfaces through ligand exchange, and to extract most organic Fe and organically bound Al (Guo and Yost, 1999). Similar mechanisms may be active in the extraction of fluoride from soils.

The high fluoride recovery observed with Soil A can be explained by its high content of active amorphous Al and Fe hydroxides (as evaluated by the acid ammonium oxalate extraction), its low pH, and the absence of CaCO₃ (Table 1). The low pH favors fluoride retention through ligand exchange by the soil constituents. Due to the low clay content of Soil A, ligand exchange must occur mainly on the active amorphous Al and Fe hydroxides. Fluorite formation is not likely to occur due to the absence of CaCO₃. The results of Smillie and Syers (1972) suggest that the presence of soluble Ca, in contrast to free CaCO₃, does not result in the formation of CaF₂ in contact with fluoride. We can therefore postulate that the main fluoride retention mechanism in Soil A is ligand exchange at the surface of amorphous Fe and Al hydroxides. The dissolution of these hydroxides by oxalate brings the adsorbed fluoride into solution, which results in the high fluoride recovery observed in Soil A.

In contrast, Soil B, for which the fluoride recovery is lower, contains carbonates, has a higher pH, and has less Al_ox and Fe_ox (Table 1). The rapid weight loss during the determination of carbonates indicates that the major part of carbonates is calcite. The conversion of calcite to fluorite in this soil is likely. The low Al_ox and Fe_ox content also supports the hypothesis that CaF₂ formation occurred. Conversion of CaCO₃ particles to CaF₂ may extend to a depth of several millimeters (Glover and Sippel, 1962). The CaF₂ formed could be only partially extracted by oxalate due to the formation of a protective calcium oxalate film at the surface of CaF₂ at the beginning of extraction. Calcium oxalate films naturally form with time on natural outcrops of limestone and marble (Lazzarini and Salvadori, 1989 [as cited by Cezar, 1998]). They can also be formed artificially in less than 5 h at the surface of calcareous stones with ammonium oxalate 0.18 M (Cezar, 1998). Calcium oxalate films protect the underlying calcite against acid attack down to pH 3 and have a reduced porosity compared with calcite, which could impede the mobility of soluble salts through it. The conversion of calcite to calcium oxalate is limited to a depth of about 10 µm. So in case of deep conversion of calcite particles to fluorite, the eventual formation of a thin, acid-resistant, and impermeable film of calcium oxalate at the surface of fluorite particles would explain the incomplete extraction of added fluoride in Soil B. The validity of this explanation relies on the hypothesis that the calcium oxalate films observed to form at the surface of calcite may also form at the surface of fluorite. The fact that the Ca²⁺ positions are nearly unchanged during conversion of calcite to fluorite (Glover and Sippel, 1962) supports this hypothesis.

Soil C contains less Al_ox than Soil A but Fe_ox is equal. Clay content is also higher than in Soil A (Table 1). Because of the smaller Al_ox and higher clay content, a higher proportion of adsorbed fluoride should be found at the surface of aluminosilicates. Crystalline minerals, unlike amorphous Al and Fe hydroxides, are not dissolved by acid ammonium oxalate (McKeague and Day, 1966). Consequently, the only way through which a significant amount of fluoride may be extracted from crystalline mineral surfaces under the conditions of the experiment is ligand exchange of oxalate for fluoride. Violante and Gianfreda (1993) studied the competitive adsorption of phosphate and oxalate on an aluminum hydroxide–montmorillonite complex. They reported that many sites on this clay mineral were highly specific for phosphate and that some others, common to both anions, had a higher affinity for phosphate than for oxalate. To our knowledge, the competitive adsorption between fluoride and oxalate has not been studied. However, the presence of sites on aluminosilicates or crystalline Al and Fe hydroxides that are specific to fluoride or have a higher affinity for fluoride than for oxalate is possible and could explain the incapacity of acid ammonium oxalate to completely extract added fluoride from Soil C. The presence of adsorption sites specific to fluoride could also explain the quadratic response observed with Soil C. At low fluoride concentration, fluoride ions would tend to occupy the sites that are specific to fluoride, so they would not be extracted by oxalate. A part of the fluoride ions extracted from sites common to fluoride and oxalate may even be resorbed on these specific sites during the extraction. As the fluoride concentration applied is increased, specific sites may become saturated and in this case fluoride would occupy more adsorption sites that are common to oxalate. Oxalate ions, present in a strong excess compared with fluoride, could desorb fluoride more effectively from common sites, leading to an increased recovery. Bhatti et al. (1998) already observed that oxalate was more effective in inhibiting PO₄ sorption at high PO₄ concentrations and attributed this to the competition of oxalate and PO₄ for common sites at high PO₄ concentrations. These results support the proposed explanation, but further studies will be needed to verify its validity.

Other factors may have contributed to incomplete fluoride recovery in the studied soils. Guo and Yost (1999) observed an increase of pH of the acid ammonium oxalate solution during the extraction of phosphorus in slightly weathered, calcareous soils. The pH increase from 3 to more than 4 resulted in a decrease of extraction efficiency. To evaluate if that occurred in the studied soils, a test was conducted where the pH of the extracting solution was measured after 1 and 4 h of extraction with each soil. The results showed that the pH never exceeded 3.1 (data not shown), so no decrease in extraction efficiency can be attributed to an elevation of pH in the studied soils. Parfitt (1989) showed that the amount of Fe, Al, and Si extracted with acid ammonium oxalate increases as the soil to solution ratio decreases from 1:40 to 1:200. Iron and Al, and consequently fluoride extraction, may have been more complete if a soil to solution ratio lower than 1:40 had been used. Organic carbon is another soil constituent that is involved in fluoride adsorption. However, its influence on fluoride recovery in the studied soils is uncertain.

Recovery with Incubation Time
Table 2 shows the mean F_ox concentrations measured after 7, 14, 21, and 28 d of incubation of Soil A for each added fluoride concentration. The results of the regression of F_ox against added fluoride, (added fluoride)², and time of incubation showed no significant effect of time on fluoride recovery. This does not exclude the possibility of an increase of fluoride adsorption with time or of the firmness of the link between fluoride and soil, as observed by Barrow and Shaw (1977). But it shows that prolonged contact between soil and fluoride will not affect the capacity of the acid ammonium oxalate solution to extract fluoride, at least for the studied soil. The results of the regression also showed that the coefficient for (added fluoride)² was statistically significant. This coefficient was not significant for the same soil in the previous recovery experiment. The increased power of the test resulting from the higher number of replications (three replications per incubation time, four incubation times) can explain the detection of the quadratic effect. A more pronounced quadratic response was observed in Soil C and was discussed previously.

View larger version (15K): [in this window] [in a new window]   Fig. 4. Total fluoride (diamonds) and oxalate-extractable fluoride (circles) profiles of the soil core sampled near Column 1.

View larger version (20K): [in this window] [in a new window]   Fig. 5. Final oxalate-extractable fluoride (Fox) and total fluoride (Ftot) resident concentration profiles (a, b) and net Fox and Ftot accumulation profiles (c, d) after fluoride leaching in Column 1.

View larger version (19K): [in this window] [in a new window]   Fig. 6. Final oxalate-extractable fluoride (Fox) and total fluoride (Ftot) resident concentration profiles (a, b) and net Fox and Ftot accumulation profiles (c, d) after fluoride leaching in Column 2.

	CONCLUSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSION REFERENCES

 
Our experiments indicate that use of F_ox may lead to important improvements in the environmental monitoring of fluoride in soils. Though quantitative fluoride recovery was not achieved, a significant part of added fluoride was recovered. The high selectivity of the method for added fluoride constitutes its main advantage over commonly used F_tot methods. Use of F_ox may lead to a better distinction between soil natural fluoride content and fluoride added through atmospheric depositions in the vicinity of industrial fluoride sources. It may help to overcome the problem of high background noise associated with F_tot in the past. The experiment with soil columns showed that F_ox provides more stable results than F_tot and is less affected by the high and variable natural fluoride concentrations encountered in soils. Consequently, monitoring of fluoride accumulation and leaching through the soil profile to ground water could be facilitated by the use of F_ox. Its advantages over F_tot are particularly obvious in the case of low-intensity or short-term contamination. The confirmation of fluoride deposition estimations and the validation of fluoride transport models are examples of possible F_ox applications. Use of a correction factor may be necessary to account for incomplete fluoride recovery when accurate quantitative measurement of fluoride accumulation is wanted. However, no correction factor is needed when the depth of fluoride migration and the pattern of fluoride accumulation are the main concerns. Oxalate extraction of fluoride also provides a means to determine rapidly, with a single extraction, the soil resident fluoride concentration and the concentration of two contributors of fluoride adsorption, Al_ox and Fe_ox. Further studies will be needed to identify the nature and importance of the various forms of adsorbed fluoride in soils, to identify which of these forms are extracted by acid ammonium oxalate, and to identify extractants that could be more effective in extracting added fluoride from soils where recovery with oxalate is lower.

	ACKNOWLEDGMENTS

 
We would like to thank Rob Cleven and Georges Thériault for their help with ion chromatography. We also want to thank the Natural Sciences and Engineering Research Council of Canada and the Fonds pour la formation de Chercheurs et l'Aide à la Recherche for their financial support.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSION REFERENCES

 

• Barrow, N.J., and A.S. Ellis. 1986. Testing a mechanistic model: III. The effects of pH on fluoride retention by a soil. J. Soil Sci. 37:287–293.
• Barrow, N.J., and T.C. Shaw. 1977. The slow reactions between soil and anions: 6. Effect of time and temperature of contact on fluoride. Soil Sci. 124:265–278.
• Bar-Yosef, B., I. Afik, and R. Rosenberg. 1988. Fluoride sorption by montmorillonite and kaolinite. Soil Sci. 145:194–200.
• Beek, J. 1979. Phosphate retention by a soil in relation to waste disposal. Ph.D. thesis. Agric. Univ., Wageningen, the Netherlands.
• Bhatti, J.S., N.B. Comerford, and C.T. Johnston. 1998. Influence of oxalate and soil organic matter on sorption and desorption of phosphate onto a spodic horizon. Soil Sci. Soc. Am. J. 62:1089–1095.[Abstract/Free Full Text]
• Bower, C.A., and J.T. Hatcher. 1967. Adsorption of fluoride by soils and minerals. Soil Sci. 103:151–154.
• Centre d'expertise en analyse environnementale du Québec. 2000. Protocole pour la validation d'une méthode d'analyse en chimie. (In French.) DR-12-VMC. Gouvernement du Québec, Québec, QC, Canada.
• Cezar, T.M. 1998. Calcium oxalate: A surface treatment for limestone. J. Conserv. Museum Stud. 4. Available online at http://www.ucl.ac.uk/archaeology/conservation/jcms/issue4/cezar.html (verified 23 Oct. 2002).
• Day, P.R. 1965. Particle fractionation and particle-size analysis. p. 545–567. In C.A. Black et al. (ed.) Methods of soil analysis. Part 1. Agron. Monogr. 9. ASA, Madison, WI.
• Debackere, M., and F.T. Delbeke. 1981. Fluoride pollution caused by an enamel factory in the Flemish countryside of Belgium. Int. J. Environ. Stud. 17:201–207.
• Environment Canada. 1989. Priority substances list. Canada Gazette. 11 Feb.
• Environment Canada. 1993. Inorganic fluorides: Priority substances list assessment report. En40-215/32E. Environment Canada, Hull, QC.
• Eyde, B. 1983. Determination of acid soluble fluoride in soils by means of an ion-selective electrode. Fresenius' Z. Anal. Chem. 316:299–301.
• Farrah, H., and W.F. Pickering. 1986. Interaction of dilute fluoride solutions with hydrous iron oxides. Aust. J. Soil Res. 24:201–208.
• Farrah, H., J. Slavek, and W.F. Pickering. 1987. Fluoride interactions with hydrous aluminium oxides and alumina. Aust. J. Soil Res. 25:55–69.
• Flühler, H., J. Polomski, and P. Blaser. 1982. Retention and movement of fluoride in soils. J. Environ. Qual. 11:461–468.[Abstract/Free Full Text]
• Freund, R.J., and R.C. Littell. 1991. SAS system for regression. 2nd ed. SAS Inst., Cary, NC.
• Gilpin, L., and A.H. Johnson. 1980. Fluorine in agricultural soils of southeastern Pennsylvania. Soil Sci. Soc. Am. J. 44:255–258.
• Glover, E.D., and R.F. Sippel. 1962. Experimental pseudomorphs: Replacement of calcite by fluorite. Am. Mineral. 47:1156–1165.
• Goh, T.B., R.J. St. Arnaud, and A.R. Mermut. 1993. Carbonates. p. 177–185. In M.R. Carter (ed.) Soil sampling and methods of analysis. Lewis Publ., Boca Raton, FL.
• Groth, E. 1975. Fluoride pollution. Environment 17:29–38.
• Guo, F., and R.S. Yost. 1999. Quantifying the available soil phosphorus pool with the acid ammonium oxalate method. Soil Sci. Soc. Am. J. 63:651–656.[Abstract/Free Full Text]
• Hluchan, E., J. Mayer, and E. Abel. 1964. The influence of aluminum works exhalations on the content of fluorides in soil and grass. (In Slovak.) Pol'nohospodarstvo 10:257–262. (English abstract in USEPA. 1973. Air pollution aspects of emission sources: Primary aluminum production. A bibliography with abstracts. Publ. AP-119. USEPA Air Pollut. Tech. Info. Center, Research Triangle Park, NC.)
• Jury, W.A., and K. Roth. 1990. Transfer functions and solute movement through soil: Theory and applications. Birkhäuser Verlag, Basel, Germany.
• Kabata-Pendias, A., and H. Pendias. 1992. Trace elements in soils and plants. 2nd ed. CRC Press, Boca Raton, FL.
• Larsen, S., and A.E. Widdowson. 1971. Soil fluorine. J. Soil Sci. 22:210–221.
• Lazzarini, L., and O. Salvadori. 1989. A reassessment of the formation of the patina called Scialbatura. Stud. Conserv. 34:20–26.
• Lexmond, Th.M., W.H. van Riemsdijk, and F.A.M. de Haan. 1982. Onderzoek naar fosfaat en koper in de bodem in het bijzonder in gebieden met intensieve veehouderij. (In Dutch.) Staatsuitgeverij, Den Haag, the Netherlands.
• Loreti, C.P. 1988. Pollutant/trace ligands. p. 8-1 to 8.4-8. In I. Bodek et al. (ed.) Environmental inorganic chemistry: Properties, processes, and estimation methods. Pergamon Press, New York.
• Maketon, S., and J.G. Tarter. 1984. Ion chromatographic determination of fluoride and chloride in soils. LC Mag. 2:124–126.
• McKeague, J.A., and J.H. Day. 1966. Dithionite and oxalate-extractable Fe and Al as aids in differentiating various classes of soils. Can. J. Soil Sci. 46:13–22.
• McQuaker, N.R., and M. Gurney. 1977. Determination of total fluoride in soil and vegetation using an alkali fusion-selective ion electrode technique. Anal. Chem. 49:53–56.
• Menzies, N.W., D.G. Edwards, and L.C. Bell. 1993. The use of total ionic strength adjusting buffers in determining fluoride concentration in soil solutions by ion selective electrometry. Commun. Soil Sci. Plant Anal. 24:1865–1882.
• Murray, F. 1984. Fluoride retention in highly leached disturbed soils. Environ. Pollut. Ser. B 7:83–95.
• Norrish, K. 1975. Geochemistry and mineralogy of trace elements. p. 55–81. In D.J.D. Nicholas and A.R. Egan (ed.) Trace elements in soil–plant–animal systems. Academic Press, New York.
• Omueti, J.A.I., and R.L. Jones. 1977. Fluoride adsorption by Illinois soils. J. Soil Sci. 28:564–572.
• Orion Research. 1991. Model 94-09, 96-09 fluoride/combination fluoride electrodes: Instruction manual. Orion Research, Beverly, MA.
• Parfitt, R.L. 1989. Optimum conditions for extraction of Al, Fe, and Si from soils with acid oxalate. Commun. Soil Sci. Plant Anal. 20:801–816.
• Pickering, W.F. 1985. The mobility of soluble fluoride in soils. Environ. Pollut. Ser. B 9:281–308.
• Plüger, W.L., and G.H. Friedrich. 1973. Determination of total and cold-extractable fluoride in soils and stream sediments with an ion-sensitive fluoride electrode. p. 421–427. In M.J. Jones (ed.) Geochem. Explor., Proc. Int. Geochem. Explor. Symp., 4th, London. 17–20 Apr. 1972. Institution of Mining and Metallurgy, London.
• Polomski, J., H. Flühler, and P. Blaser. 1982. Accumulation of airborne fluoride in soils. J. Environ. Qual. 11:457–461.[Abstract/Free Full Text]
• Reynolds, W.D., and D.E. Elrick. 1991. Determination of hydraulic conductivity using a tension infiltrometer. Soil Sci. Soc. Am. J. 55:633–639.[Abstract/Free Full Text]
• SAS Institute. 2001. The SAS system for Windows. Release 8.02. SAS Inst., Cary, NC.
• Shaw, W.M. 1954. Colorimetric determination of fluorine in water and soil extracts. Anal. Chem. 26:1212–1214.
• Sigg, L., and W. Stumm. 1981. The interaction of anions and weak acids with the hydrous goethite (-FeOOH) surface. Colloids Surf. 2:101–117.
• Skjelkvåle, B.L. 1994. Factors influencing fluoride concentrations in Norwegian lakes. Water Air Soil Pollut. 77:151–167.
• Smillie, G.W., and J.K. Syers. 1972. Calcium fluoride formation during extraction of calcareous soils with fluoride: II. Implications to the Bray P-1 test. Soil Sci. Soc. Am. Proc. 36:25–30.
• Stumm, W. 1992. Chemistry of the solid–water interface: Processes at the mineral–water and particle–water interface in natural systems. John Wiley & Sons, New York.
• Supharungsun, S., and M. Wainwright. 1982. Determination, distribution, and adsorption of fluoride in atmospheric-polluted soils. Bull. Environ. Contam. Toxicol. 28:632–636.[ISI][Medline]
• Thomas, J., H.D. Glass, W.A. White, and R.M. Trandel. 1977. Fluoride content of clay minerals and argillaceous earth materials. Clays Clay Miner. 25:278–284.[Abstract]
• Thompson, L.K., S.S. Sidhu, and B.A. Roberts. 1979. Fluoride accumulations in soil and vegetation in the vicinity of a phosphorus plant. Environ. Pollut. 18:221–234.
• Topp, G.C., and W. Zebchuk. 1979. The determination of soil–water desorption curves for soil cores. Can. J. Soil Sci. 59:19–26.
• Urumova-Pesheva, A., and S. Ruseva. 1982. Comparative testing of extracts to select a method for determining mobile fluorine. (In Bulgarian.) Pochvozn. Agrokhim. 17:84–94. (English abstract in Chemical Abstracts. 1983. 99:21351z. ACS, Columbus, OH.)
• Van den Hoop, M.A.G.T., R.F.M.J. Cleven, J.J. van Staden, and J. Neele. 1996. Analysis of fluoride in rain water: Comparison of capillary electrophoresis with ion chromatography and ion-selective electrode potentiometry. J. Chromatogr. A 739:241–248.
• Van der Zee, S.E.A.T.M., L.G.J. Fokkink, and W.H. van Riemsdijk. 1987. A new technique for assessment of reversibly adsorbed phosphate. Soil Sci. Soc. Am. J. 51:599–604.[Abstract/Free Full Text]
• Violante, A., and L. Gianfreda. 1993. Competition in adsorption between phosphate and oxalate on an aluminum hydroxide montmorillonite complex. Soil Sci. Soc. Am. J. 57:1235–1241.[Abstract/Free Full Text]
• Walkley, A., and I.A. Black. 1934. An examination of the Degtjareff method for determining soil organic matter, and a proposed modification of the chromic acid titration method. Soil Sci. 34:29–38.
• Wenzel, W.W., and W.E.H. Blum. 1992. Fluorine speciation and mobility in F-contaminated soils. Soil Sci. 153:357–364.
• Zimmerman, R.L., and H.G. Bertrand. 1978. The analysis of fluoride in the acetic acid leach solution of fluorspar with an ion-selective electrode. Anal. Lett. 11:569–580.

Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2003 32: 377-382. [Full Text]  

This article has been cited by other articles:

				L. Begin, J. Fortin, and J. Caron Evaluation of the Fluoride Retardation Factor in Unsaturated and Undisturbed Soil Columns Soil Sci. Soc. Am. J., November 1, 2003; 67(6): 1635 - 1646. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Related articles in JEQ Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar