SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 27 October 2006
Published in Soil Sci Soc Am J 70:2017-2027 (2006)
DOI: 10.2136/sssaj2005.0373
© 2006 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pigna, M.
Right arrow Articles by Violante, A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Pigna, M.
Right arrow Articles by Violante, A.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Pigna, M.
Right arrow Articles by Violante, A.
Related Collections
Right arrow Toxic Trace Metals
Right arrow Kinetics
Right arrow Soil Surface Chemistry

Soil Chemistry

Kinetics of Arsenate Sorption–Desorption from Metal Oxides

Effect of Residence Time

M. Pigna, G. S. R. Krishnamurti and A. Violante*

Dipartimento di Scienze del Suolo, della Pianta e dell'Ambiente, Univ. di Napoli Federico II, Via Università 100, 80055 Portici (Napoli), Italy; G.S.R. Krishnamurti, present address: 313-855 West 16th St., North Vancouver, BC V7P 1R2, Canada

* Corresponding author (violante{at}unina.it)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Sorption and desorption processes control the mobility, toxicity, and availability of As in natural environments. Surface coverage and residence time may affect the kinetics of As sorption–desorption from soil components and the transformation of As from desorbable into resistant or undesorbable forms. We performed kinetic studies on the sorption of As(V) onto crystalline or poorly crystalline metal oxides (noncrystalline Al(OH)x, gibbsite, ferrihydrite, and goethite) and its desorption by PO4 at pH 6.0 as affected by the residence time and the surface coverage (50 or 100%) of As(V). Significant amounts of As(V) were sorbed during the initial period of 0.167 h, ranging from 37.9 to 71.8% when the surface coverage was about 100%. The kinetic data, explained best by the Elovich kinetic model, indicated the following order in As(V) sorption: gibbsite < Al(OH)x < goethite < ferrihydrite. By adding PO4 immediately after complete sorption of As(V) onto the oxides (50% surface coverage; PO4 added/As(V) sorbed molar ratio of 4), a much higher proportion of As(V) was desorbed after 24 h of reaction from Al oxides (48–56%) than from Fe oxides (18–23%). The amount of As(V) desorbed decreased with increasing residence time. The kinetics of As(V) desorption by PO4 as a function of residence time was explained best by the Elovich kinetic model. The kinetics described the rate of rearrangement of As(V) from desorbable into resistant or undesorbable forms, which occurred more rapidly in Al than Fe oxides. After a residence time of 360 h, the percentage of As(V) desorbed from the oxides was reduced significantly (<13%).

Abbreviations: Al(OH)x, noncrystalline aluminum hydroxide • FT-IR, Fourier-transform infrared • PZC, point of zero charge • XRD, x-ray diffraction


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
ARSENIC is a toxic and carcinogenic metalloid, whose concentrations in natural environments can become elevated by mineral dissolution and volcanic eruption, mine drainage and geothermal discharge, and various human activities in agriculture and forestry (e.g., the use of arsenical pesticides) (Smith et al., 1998; Frankenberger, 2002; Mandal and Suzuki, 2002). Sorption of As by soil constituents controls the mobility and bioavailability of As in soil–water–plant systems (Barrow, 1974; Livesey and Huang, 1981; Nriagu, 1994; Sun and Doner, 1996; Sadiq, 1997; Raven et al., 1998; Frankenberger, 2002; Smedley and Kinniburgh, 2002; Violante and Pigna, 2002).

Arsenic occurs in two oxidation states, arsenate [As(V)] and arsenite [As(III)], each of which has a different sorption behavior (Hsia et al., 1994; Arai et al., 2001; Goldberg and Johnston, 2001). Arsenate is specifically sorbed on variable-charge minerals, including Al, Fe, and Mn oxides (Fendorf et al., 1997; Smith et al., 1998; O'Reilly et al., 2001; Lin and Puls, 2000; Liu et al., 2001; Goldberg and Johnston, 2001; Violante and Pigna, 2002). Direct evidence for the formation of As(V) inner sphere complexes have been obtained using extended x-ray absorption fine structure spectroscopy and wide-angle x-ray scattering (Waychunas et al., 1993, 1996; Sparks, 1999; Arai et al., 2001). Fendorf et al. (1997) studied the sorption of As(V) on goethite and showed that it formed three different complexes: a monodentate complex, a bidentate–binuclear complex, and a bidentate–mononuclear complex, depending on surface coverage. Goldberg and Johnston (2001) showed that As(V) forms inner sphere surface complexes on noncrystalline Al and Fe oxides, whereas As(III) forms both inner and outer sphere surface complexes on noncrystalline Fe oxide and only outer sphere surface complexes on noncrystalline Al oxide. Sorption–desorption of metals, oxyanions, radionuclides, and organic chemicals on soil minerals and soils depend on the residence time, which has been attributed to different sites of reactivity, surface nucleation–precipitation or diffusion into micropores of inorganic and organic sorbents (Sparks, 1999, 2002). Studies about competitive sorption between As(V) and PO4 have shown that Mn, Fe, and Ti oxides and phyllosilicates, particularly rich in Fe (nontronite, ferruginous smectite), were more effective in fixing As(V) than PO4, whereas the opposite occurred for gibbsite, boehmite, noncrystalline Al precipitation products, allophane, kaolinite, and halloysite (Violante and Pigna, 2002).

In contrast, little information is available on desorption of As from soil components and soils. Desorption of As seems to be also dependent on the type of the sorbent. Soils rich in variable-charge minerals (Al, Fe, Mn, or Ti oxides and allophane) do not release As easily. Phosphate has been reported to displace sorbed As from minerals and soils (Woolson et al., 1973; Peryea, 1991; O'Reilly et al., 2001; Arai and Sparks, 2002; Violante et al., 2005a). The application of phosphate fertilizers is a common crop management practice and it affects the sorption–desorption behavior of As, its phytoavailability, and groundwater contamination. Applications of relatively high rates of phosphate fertilizer have been shown to enhance As mobility in laboratory columns (Melamed et al., 1995) and As solution concentration in laboratory batch studies (Peryea, 1991). Only a large addition of PO4 (Smith et al., 1998) and alkaline pH may affect As solubility. Woolson et al. (1973) observed that a great addition of PO4 to As-polluted soils displaced ~77% of the total As on the soil.

Kinetic studies on desorption of As by inorganic (i.e., PO4) and organic ligands could explain the mechanisms of the As(V) release to soil solution (O'Reilly et al., 2001; Arai and Sparks, 2002; Frankenberger, 2002; Violante and Pigna, 2002). A number of factors, such as the type, mineralogy, and crystallinity of the sorbents, the pH, the pe, the surface coverage, the residence time of As (as well as other heavy metals and metalloids) on the surfaces of soil components, and the oxidation state of As may affect the desorption of this toxic element. Some studies on the effect of residence time on the sorption–desorption of some heavy metals and metalloids have been performed. Zhang and Sparks (1990) found that the reaction between selenite and goethite consisted of a fast step attributed to the formation of outer sphere complexes and a subsequent slower step due to the formation of inner sphere surface complexes. Arai and Sparks (2002) found that the longer the residence time, the greater the decrease in As(V) desorption at pH 4.5 or 7.8 from bayerite by PO4 because of a rearrangement of surface complexes or a conversion of surface complexes into Al arsenate precipitates. Until now, however, kinetics studies on the effect of surface coverage, residence time, and physicochemical and mineralogical properties of sorbents on the rearrangement of trace elements (including As) onto the surfaces of soil minerals from desorbable into undesorbable (or resistant) forms has not received attention.

The aim of this work was to carry out kinetic studies on the sorption of As(V) on crystalline or poorly crystalline metal oxides (noncrystalline Al hydroxide, gibbsite, ferrihydrite, and goethite) and its desorption from these materials by PO4 as affected by the residence time and As(V) surface coverage.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Preparation of Metal Oxides
Goethite was prepared by precipitating 0.5 mol L–1 Fe(NO3)3 at pH 12.0 with slow addition of 1.0 mol L–1 NaOH (Atkinson et al., 1968). The suspension was aged for 7 d at room temperature and subsequently for 20 d at 65°C. The precipitate was dialyzed (molecular weight cutoff of 15 000) in deionized water.

Ferrihydrite was prepared by precipitating 0.1 mol L–1 Fe(NO3)3 at pH 5.5 with 0.5 mol L–1 NaOH at a rate of 0.5 mL min–1 (Cornell and Schwertmann, 1996). The final volume was adjusted to 1 L. After 7 d of aging at room temperature, the suspension was dialyzed in deionized water.

Gibbsite was prepared by the slow addition of 0.5 mol L–1 NaOH to 0.1 mol L–1 Al(NO3)3 up to pH 5.5. After 15 d, the pH was increased to 7.0 and the suspension was aged 15 d at 20°C followed by 30 d at 40°C (Violante et al., 2005b). The suspension was transferred to a cellulose tube and dialyzed with distilled water for 16 d.

All the suspensions were freeze-dried and ground to pass through a 0.315-mm sieve.

A noncrystalline aluminum hydroxide [Al(OH)x] was purchased from Aldrich Chemical Co, (Milwaukee, WI).

Characterization of the Samples
The surface area of the samples was determined by H2O sorption at 20% relative humidity (Quirk, 1955) and the point of zero charge (PZC) of the samples was measured according to the method of Sakurai et al. (1988).

The x-ray diffraction (XRD) patterns of randomly oriented samples were obtained using a Rigaku diffractometer with Fe-filtered CoK{alpha} radiation generated at 40 kV and 30 mA (Rigaku Co., Tokyo).

The infrared spectra of the samples were obtained using diffuse reflectance infrared Fourier transform (DRIFT) spectroscopy analysis. Sample preparation for DRIFT determinations was as follows: 0.2 mg of sample was mixed with 200 mg of KBr (Fourier-transform infrared [FT-IR] grade, Aldrich Chemical Co.). The DRIFT spectra were obtained using a PerkinElmer Spectrum One FT-IR Spectrophotometer (PerkinElmer Instruments, Wellesley, MA). The instrument had a spectral resolution of 1 cm–1, which was used in all spectra determinations.

For transmission electron microscopic (TEM) examination, one drop of sample suspension was deposited onto a C-coated Forvar film Cu grid. The TEM micrographs were taken with a Philips CM 120 microscope (Philips, Eindhoven, the Netherlands).

Arsenate Sorption Isotherms
One hundred milligrams of each metal oxide, in triplicate, were equilibrated at 20°C with 19.5 mL of 0.05 mol L–1 KCl at pH 6.0. Suitable amounts of 0.01 mol L–1 solutions containing K2HAsO4 were added to obtain an initial As(V) concentration in the range 5 x 10–4 to 10–2 mol L–1. The pH of each suspension was maintained constant at pH 6.0 by the addition of 0.1 or 0.01 mol L–1 HCl or KOH using a stirred pH-stat apparatus. The final volume was 20 mL and the final solid/solution ratio was 5 g L–1. The suspensions were shaken for 24 h in a water bath at 20°C, centrifuged at 10 000 x g for 20 min, and filtered using membrane filters (0.22 µm). The filtrates were stored at 2°C until analysis. The concentrations of As(V) in the solutions were determined as described below. Arsenate sorption was calculated from the difference between the initial and final As(V) concentration in the solutions.

As described below, the sorption data conformed to the Langmuir equation in the following form:

Formula 1[1]
where S is the amount of the element sorbed per unit mass of adsorbent (mmol kg–1), Sm is the maximum amount of As(V) that may be bound to the absorbent (adsorption capacity), c is the equilibrium solution concentration (mmol L–1), and K is a constant related to the binding energy.

Kinetics of Arsenate Sorption
The sorption kinetics experiments were conducted at pH 6.0 (below the PZC of the samples) by adding suitable amounts of As(V), as K2HAsO4, in different flasks to obtain a final As(V) surface coverage of approximately 50 or 100%, calculated by the maximum value of As(V) sorbed on the oxides (Sm) determined by sorption isotherms (Table 1). To have a final As(V) surface coverage of about 100% (>>90%), 200, 300, 200, and 600 mmol As(V) kg–1 were initially added respectively to gibbsite, Al(OH)x, goethite. and ferrihydrite, respectively. The suspensions were placed in a water bath at 20°C. Samples were collected from separate flasks (duplicate for each determination) after a time period between 0.167 (10 min) and 24 h. Some experiments were performed by collecting samples after a time period between 0.0167 and 0.167 h. The samples were centrifuged at 10 000 x g for 20 min, and filtered using membrane filters (0.22 µm). The concentrations of As(V) in the supernatants were determined as described below. After 24 h of reaction, the concentration of As(V) in the solutions were much lower than 0.05 mmol L–1, indicating >90% surface coverage. To have a final As(V) surface coverage of 50% on all the sorbents, 100 mmol kg–1 were added to gibbsite, 150 mmol kg–1 to Al(OH)x, 100 mmol kg–1 to goethite, and 300 mmol kg–1 to ferrihydrite. In these systems, As(V) anions were completely replaced from the solutions.


View this table:
[in this window]
[in a new window]
  		Table 1. Surface area, point of zero charge (PZC) of the metal oxides, and maximum sorption capacity (Sm), as obtained from the sorption isotherms, and the percentage of As(V) sorbed on the oxides after 0.167 h of reaction.

 
Influence of Residence Time on the Desorption of Arsenate by Phosphate from Metal Oxides
The experiments were performed at pH 6.0 by adding predetermined quantities of As(V) to each sample in reaction flasks to obtain a surface coverage of 50 or 100%. For the experiments at 50% surface coverage, the suspensions were then allowed to react for 0.5 h (ferrihydrite or goethite) or 2 h [Al(OH)x or gibbsite] until As(V) was completely removed from solution and sorbed onto the surfaces of the metal oxides. Suitable amounts of PO4 (0.1 mol L–1 as KH2PO4), were added to the suspensions after 0.5, 1, 2, 3, 5, 24, 48, 168, 360, and 720 h (containing a PO4 added/As(V) sorbed molar ratio of 4), to have a PO4 concentration in solution sufficiently high to facilitate As(V) desorption. For the experiments at about 100% surface coverage, the suspensions were allowed to react for 8 h (ferrihydrite) or 24 h (goethite, Al(OH)x, or gibbsite) until As(V) was almost completely removed from solution and sorbed onto the surfaces of the metal oxides.

The suspensions were kept in reaction for another 24 h after PO4 addition. The pH of each suspension was kept constant for all the experiments by the addition of 0.01 mol L–1 HCl or KOH.

The final suspensions (20 mL) were centrifuged at 10 000 x g for 20 min, and filtered through membrane filters (0.22 µm). Arsenate was determined in the supernatant as described below.

Arsenate Determination
Arsenate was determined by ion chromatography, using a Dionex DX-300 ion chromatograph (Dionex Co., Sunnyvale, CA), an IonPac AS11 column (4.0 mm), an eluent of 0.05 mol L–1 NaOH at a flow rate of 1 mL min–1, and a CD20 conductivity detector combined with autosuppression (Liu et al., 2001). Average As(V) retention time was 4.2 min. The As(V) standard concentration was 0.05 to 2 mmol L–1. Coefficients of variation amongst the replicates ranged from 1.5 to 5%.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Characterization of the Metal Oxides
The Fe precipitation product synthesized at pH 12 and aged at 65°C was identified to be goethite, characterized by the intense XRD peaks observed at 0.419, 0.338, 0.269, 0.258, 0.244, 0.225, 0.219, 0.172 and 0.168 nm (Fig. 1A ) and strong FT-IR bands observed at 893, 797, and 635 cm–1 (Fig. 2A ). A difference of 96 cm–1 between the two strong OH bending bonds at 893 and 797 cm–1 indicated that the goethite was well crystallized (Cornell and Schwertmann, 1996). This oxide showed a specific surface area of 40 m2 g–1 and a PZC of 8.70 (Table 1).


Figure 1
View larger version (16K):
[in this window]
[in a new window]
  		Fig. 1. X-ray diffraction patterns of (A) goethite and ferrihydrite, and (B) gibbsite and noncrystalline Al hydroxide [Al(OH)x].

 

Figure 2
View larger version (15K):
[in this window]
[in a new window]
  		Fig. 2. Fourier-transform infrared transmittance spectra of (A) goethite and ferrihydrite, and (B) gibbsite and noncrystalline Al hydroxide [Al(OH)x].

 
The Fe oxide obtained at pH 5.5 and aged for 7 d at room temperature was identified to be ferrihydrite. The x-ray pattern of this sample showed four characteristic broad peaks of ferrihydrite centered at 0.254, 0.225, 0.198, and 0.148 nm (Fig. 1A). The FT-IR spectrum indicated that this material was a very poorly crystallized ferrihydrite (Fig. 2A). This oxide had a surface area of 173 m2 g–1 and a PZC of 7.45 (Table 1).

The Al hydroxide precipitated and aged at pH 7.0 corresponded to gibbsite as characterized by the XRD peaks at 0.480, 0.474, 0.336, 0.320, 0.246, 0.239, 0.204, 0.180, 0.169, and 0.145 nm (Fig. 1B) and FT-IR bands at 3621, 3525, and 3476 cm–1 (Fig. 2B). This oxide had a surface area of 38 m2 g–1 and a PZC of 8.90 (Table 1).

The Al oxide sample purchased from Aldrich appeared to be to x-ray a noncrystalline Al hydroxide (Fig. 1B) and had broad IR absorption band at 3437 cm–1 (Fig. 2B) [designated as Al(OH)x]. This oxide had a surface area of 130 m2 g–1 and a PZC of 8.50 (Table 1).

Sorption Isotherms
Sorption isotherms for As(V) onto gibbsite, noncrystalline Al(OH)x, goethite, and ferrihydrite conformed to the Langmuir equation (Fig. 3 ). According to the classification of Giles et al. (1974), the isotherms of As(V) sorption on gibbsite, goethite, ferrihydrite, and Al(OH)x appear to be H-type (high affinity). The sorption capacity for poorly crystalline metal oxides was higher than that of crystalline metal oxides. Maximum As(V) sorption capacity (Sm), calculated from the experimental data by the Langmuir equation are reported in Table 1.


Figure 3
View larger version (14K):
[in this window]
[in a new window]
  		Fig. 3. Arsenate sorption isotherms on gibbsite, noncrystalline Al hydroxide [Al(OH)x], goethite, and ferrihydrite at pH 6.0 and 20°C. The ionic strength was 0.05 mol L–1 KCl and the solid/solution ratio was 5 g L–1.

 
Sorption Kinetics of Arsenate on Metal Oxides
The sorption kinetics of As(V) on metal oxides was affected by the surface properties and mineralogical composition of the sorbents and the amounts of As(V) added. Figure 4 shows the percentage of the quantity of As(V) added to each sorbent that was sorbed after 1 h of reaction, which caused a final As(V) surface coverage of 50% or nearly 100%. The data indicated that the lower the final As(V) surface coverage on the sorbents, the greater the amount of As(V) sorbed. The complete As(V) sorption onto surfaces of the metal oxides at 50% surface coverage occurred within 0.5 to 2 h for all the sorbents.


Figure 4
View larger version (21K):
[in this window]
[in a new window]
  		Fig. 4. Arsenate sorbed at pH 6.0 after 1 h of reaction on ferrihydrite, goethite, noncrystalline Al hydroxide [Al(OH)x], and gibbsite. Final As(V) surface coverage was about 50 or 100% for all sorbents.

 
Figure 5 shows the effect of the reaction time vs. the sorption of As(V) onto gibbsite and Al(OH)x (Fig. 5A), and onto goethite and ferrihydrite (Fig. 5B) at pH 6.0 at about 100% surface coverage. Complete sorption of As(V) onto ferrihydrite was observed after 8 h (Fig. 5B). Even though complete As(V) sorption onto gibbsite, Al(OH)x, and goethite was observed only after 24 h, the kinetics of sorption of As(V) was evaluated for the reaction period 0.167 to 8 h for all the oxides for uniformity. The kinetics of sorption of As(V) on to goethite, gibbsite, and Al(OH)x observed during the reaction period of 0.167 to 24 h followed the same trend (Fig. 6 ) as observed for the shorter period of 0.167 to 8 h (Table 2). The kinetic models tested (first order, parabolic diffusion, and Elovich) (Low, 1960; Sparks, 2002) were compared by evaluating their ability to fit As(V) sorption data. Coefficients of determination (r2), the level of significance (P value), and the F value were obtained by linear regression analysis. Relatively higher r2 and F value and lower P value indicate the best fit of the kinetic model. The regression analysis data are presented in Table 2. The fit for the sorption data was obtained best using the Elovich model, as indicated by the r2, P, and F values, for all the metal oxides (Table 2).


Figure 5
View larger version (15K):
[in this window]
[in a new window]
  		Fig. 5. Kinetics of sorption of As(V), at pH 6.0 and 100% surface coverage, on (A) gibbsite and noncrystalline Al hydroxide [Al(OH)x], and (B) goethite and ferrihydrite. The amount of As(V) added was 200 mmol kg–1 for gibbsite, 300 mmol kg–1 for Al(OH)x, 200 mmol kg–1 for goethite, and 600 mmol kg–1 for ferrihydrite.

 

Figure 6
View larger version (16K):
[in this window]
[in a new window]
  		Fig. 6. Kinetics of sorption of As(V), at pH 6.0 and 100% surface coverage, on gibbsite, noncrystalline Al hydroxide [Al(OH)x], goethite, and ferrihydrite, using the Elovich model. The amount of As(V) added was 200 mmol kg–1 for gibbsite, 300 mmol kg–1 for Al(OH)x, 200 mmol kg–1 for goethite, and 600 mmol kg–1 for ferrihydrite.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Simple regression analysis of As(V) sorption obtained during the reaction period of 0.167 to 8 h for all sorbents using different kinetic models. Arsenate was added to have a final surface coverage of about 100%.

 
Arsenate sorbed in the initial 0.167 h ranged from 37.9 to 71.8% (Table 1). The percentage of As(V) sorbed increased in the following order: gibbsite < goethite < Al(OH)x < ferrihydrite (Table 2, Fig. 5). Higher amounts of As(V) were sorbed onto Fe oxides than onto Al oxides and more on short-range-ordered oxides than on crystalline oxides (compare ferrihydrite with goethite and Al(OH)x with gibbsite, Fig. 4 and 5). The percentage of As(V) sorbed in the initial reaction period of 0.167 h, as well as the sorption capacity values obtained from the Langmuir sorption isotherms, followed similar trends as that of the surface area of the oxides (Table 1).

The kinetics of sorption of As(V) onto the metal oxides during the reaction period of 0.167 to 8 h was evaluated by the {alpha} and ß parameters of the Elovich kinetic model (Table 2). The higher values of {alpha} and ß for the poorly crystalline metal oxides, i.e., Al(OH)x and ferrihydrite, can be ascribed to the larger pore surface (or surface area) of these oxides. This behavior has been attributed to diffusion phenomena of trace elements into pores of inner surface sites (Kinniburgh and Jackson, 1981) of different reactivity, site preferences of the metal ion (Ainsworth et al., 1994), and surface precipitation (Scheidegger et al., 1996, 1997; Sparks, 2002). The kinetic parameters quantify the rate of diffusion of As(V) into pores of inner surface sites of different reactivity. The Elovich equation may reveal irregularities in data ordinarily overlooked by other kinetic models. This irregular fit of the kinetic data has been suggested as a characteristic of the nature of sites involved in the adsorption process. These "breaks" in the Elovich plot could indicate a change from one type of bonding site to another and such breaks may not be artifacts of kinetic treatments and may indicate a biphasic reaction (Low, 1960).

Earlier reports suggested biphasic sorption phenomenon for As(V) on soil components, such as amorphous Al hydroxide, bayerite, and ferrihydrite, over different time scales (hours to months; Anderson et al., 1976; Fuller et al., 1993). The sorption process of As(V) on metal oxides in this study can also be explained as a biphasic process, the first sorption process occurring during the initial reaction period of 0.167 h and the subsequent sorption occurring during the reaction period of 0.167 to 8 (or 24) h.

Due to a lack of gibbsite and ferrihydrite, the significant amounts of As(V) sorbed during the reaction period of 0.0167 to 0.167 h (Table 1) was quantified only for Al(OH)x and goethite, as well as for a ferrihydrite synthesized as described above but aged in suspension for 25 d. The fit for the sorption data on short-term kinetics was obtained best using a parabolic diffusion model, indicating that diffusion was the rate-limiting process in the sorption of As(V) onto the metal oxides (data not shown). The rate of sorption of As(V) on these metal oxides was up to 16 times faster during the reaction period between 0.0167 to 0.167 h than that observed during the reaction period of 0.167 to 8 h if the reaction kinetics were assumed to follow the parabolic diffusion kinetic model during both the reaction processes to facilitate comparison of the rate constants.

Effect of Residence Time on the Desorption of Arsenate by Phosphate from Metal Oxides
The contact time between the sorbent and sorbate is referred to as the residence time or aging effect (Pignatello and Xing, 1995; Sparks, 1995; Lin and Puls, 2000; Arai and Sparks, 2002; Violante et al., 2005a). The effect of residence time on As(V) desorption [50% of As(V) surface coverage] by PO4 [PO4 added/As(V) sorbed molar ratio of 4] was studied at pH 6.0, because Al or Fe dissolution is negligible and the possible formation of Al or Fe arsenate coprecipitates may be ruled out at this pH of reaction (Arai and Sparks, 2002). The time for complete As(V) removal from soil solutions to have a final As(V) surface coverage onto the metal oxides of 50%, as reported above, ranged from 0.5 h (ferrihydrite and goethite) to 2 h [gibbsite and Al(OH)x]. By adding PO4 immediately after the complete sorption of As(V) onto the surfaces of the metal oxides, a high proportion of the metalloid was desorbed after 24 h of reaction from the Al oxides gibbsite (56%) and Al(OH)x (48%) (Fig. 7A ), whereas much smaller amounts were desorbed from the Fe oxides goethite (18%) and ferrihydrite (23%) (Fig. 7B). This indicated that within 0.5 h, from 77 to 82% of As(V) was strongly held onto the surfaces of ferrihydrite and goethite in resistant forms (undesorbable by PO4 after a reaction time of 24 h), because the high affinity of this element for the surfaces of Fe oxides (Goldberg and Johnston, 2001; Violante and Pigna, 2002). The decrease in desorption of As(V) by PO4 with increase in residence time was attributed to the rearrangement of As(V) from desorbable into resistant and undesorbable forms.


Figure 7
View larger version (11K):
[in this window]
[in a new window]
  		Fig. 7. Arsenate desorbed by PO4 at pH 6.0 from gibbsite, noncrystalline Al hydroxide [Al(OH)x], ferrihydrite, and goethite. Phosphate was added at an initial PO4/As(V) molar ratio of 4.

 
According to Arai and Sparks (2002), the decrease in desorption of As(V) with increase in residence time could be attributed to different aging mechanisms such as a rearrangement of surface complexes (for example, from bidentate mononuclear complexes into bidentate binuclear complexes) as well as a conversion of surface complexes into Al (or Fe) arsenate-like surface precipitates (Jia et al., 2006). Physicochemical processes may also affect the decrease of As(V) desorption with residence time, such as sorption reactions on higher energy binding sites, intraparticle diffusion, and penetration into micropores.

The kinetics of As(V) desorption by PO4 as a function of residence time was evaluated using first-order, parabolic diffusion, and Elovich kinetic models (Table 3). The kinetics of desorption of As(V) by PO4 could be explained best by the Elovich kinetic model. The rate of rearrangement of As (V) from desorbable into resistant or undesorbable forms was much lower for Fe oxides than Al oxides (Table 3). The lower {alpha} or ß value, or the higher the 1/{alpha} or 1/ß value, the higher is the rate of reorganization of As (V). The 1/ß value was three to seven times higher for Al oxides than for Fe oxides and followed the order: Al(OH)x > gibbsite > ferrihydrite > goethite. Our data indicated that immediately after the addition of As(V) to the sorbents (0.5–2 h), much greater amounts of As(V) were sorbed in resistant forms onto the surfaces of Fe oxides than onto those of Al oxides, which resulted in relatively more desorption of As(V) from Al oxides in comparison to Fe oxides up to 48 h of residence time (Fig. 7); however, the transformation of As(V) from desorbable into resistant or undesorbable forms with an increase in the residence time occurred more rapidly in Al oxides. After 360 h of residence time, only negligible amounts of As (V) were desorbed by PO4 from the metal oxides studied (<13%). The amount of As(V) desorbed by PO4 from the sorbents decreased with increasing residence times.


View this table:
[in this window]
[in a new window]
  		Table 3. Simple regression analysis of As(V) desorption by PO4 as a function of residence time using different kinetic models.

 
The fit of the desorption data using the Elovich kinetic model indicated irregularities only in the case of goethite (Fig. 7). The data could be explained best by assuming two processes, the first one operating during the reaction period 0.5 to 3 h and the second one during the reaction period of 3 to 168 h (Table 4). The 1/ß or 1/{alpha} values were 5 or 106 times higher, respectively, for the reaction period 0.5 to 3 h than for the reaction period of 3 to 360 h (Table 4). Thus, in the period 0.5 to 3 h, the rearrangment of As(V) anions onto goethite was particularly rapid. Furthermore, only a very low percentage (<12%) of As (V) initially sorbed onto goethite was converted very slowly into resistant forms. On the contrary, O'Reilly et al. (2001) found that even a 200-d residence time did not lead to a decrease in As(V) release by PO4 from goethite at either pH 4.0 or 6.0.


View this table:
[in this window]
[in a new window]
  		Table 4. Regression analysis of the kinetic data of As(V) desorption from goethite using the Elovich model.

 
It is interesting to note that desorption kinetics performed when the surface coverage of As(V) was nearly 100% showed that after a residence time of 2 mo, appreciable percentages of As(V) were still desorbable by PO4 after 24 h of reaction (up to 15–20%) in comparison with the negligible or no removal of As(V) desorbable when the surface coverage was 50% (data not shown).

These findings demonstrate that at a given pH, the percentage of As(V) desorbable from the surfaces of a sorbent is affected by the type of mineral, surface coverage of its surface by As(V), and residence time.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Studies on the sorption–desorption of trace elements initially onto and from natural materials (individual minerals, soils, and sediments) are of paramount importance for understanding the factors that may influence their mobility in natural environments. This study has demonstrated that the surface properties and mineralogical composition of Fe and Al oxides and the degree of surface coverage of As(V) onto the surfaces of crystalline and short-range-ordered metal oxides affect its sorption and the kinetics of reorganization of sorbed As(V) anions from desorbable to resistant or undesorbable forms by PO4. The decrease in the desorption of As(V) by PO4 with residence time must be attributed to the formation of strongly bound complexes or insoluble precipitates or increased penetration into micropores (Arai and Sparks, 2002, and references therein). The best fit for the sorption–desorption data was obtained by using the Elovich model.

Our study showed evidence that, immediately after the complete sorption of As(V) onto the sorbents, a much greater proportion of the element was desorbed by PO4 from Al than from Fe oxides, but the reorganization of As(V) from desorbable into resistant or undesorbable forms occurred more rapidly onto the surfaces of Al oxides. With time, however, negligible amounts of As(V) were desorbed by PO4.

The factors that affect the removal of a trace element (e.g., pH, oxidation state, residence time, surface coverage, chemical and mineralogical properties of the natural sorbents, time of reaction) by inorganic and organic ligands deserve particular attention.


   ACKNOWLEDGMENTS
 
We express our gratitude to the Università di Napoli Federico II for providing a grant to Dr. G.S.R. Krishnamurti to enable his stay in Dipartimento di Scienze del Suolo, della Pianta e dell'Ambiente (DiSSPA). This work was supported by the Italian Research Program of National Interest (PRIN2004). DiSSPA Contribution no. 117.

Received for publication November 16, 2005.


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





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 Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pigna, M.
Right arrow Articles by Violante, A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Pigna, M.
Right arrow Articles by Violante, A.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Pigna, M.
Right arrow Articles by Violante, A.
Related Collections
Right arrow Toxic Trace Metals
Right arrow Kinetics
Right arrow Soil Surface Chemistry

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