Environmental Toxicology and Chemistry, Vol. 33, No. 10, pp. 2217–2224, 2014 # 2014 SETAC Printed in the USA

Surface Complexation Modeling MODELING OXYANION ADSORPTION ON FERRALIC SOIL, PART 2: CHROMATE, SELENATE, MOLYBDATE, AND ARSENATE ADSORPTION CLAUDIO PÉREZ,yz JUAN ANTELO,*x SARAH FIOL,y and FLORENCIO ARCEy yDepartment of Physical Chemistry, University of Santiago de Compostela, Santiago de Compostela, Spain zEnvironmental Bio-Geochemistry Group, Insituto de Geología, Universidad Nacional Autónoma de Mexico (UNAM), Mexico City, Mexico xDepartment of Soil Science and Agricultural Chemistry, University of Santiago de Compostela, Santiago de Compostela, Spain (Submitted 25 October 2013; Returned for Revision 7 December 2013; Accepted 13 March 2014) Abstract: High levels of oxyanions are found in the soil environment, often as a result of human activity. At high concentrations, oxyanions can be harmful to both humans and wildlife. Information about the interactions between oxyanions and natural samples is essential for understanding the bioavailability, toxicity, and transport of these compounds in the environment. In the present study, the authors investigated the reactivity of different oxyanions (AsO4, MoO4, SeO4, and CrO4) at different pH values in 2 horizons of a ferralic soil. By combining available microscopic data on iron oxides with the macroscopic data obtained, the authors were able to use the charge distribution model to accurately describe the adsorption of these 4 oxyanions and thus to determine the surface speciation. The charge distribution model was previously calibrated and evaluated using phosphate adsorption/desorption data. The adsorption behavior on ferralic soil is controlled mainly by the natural iron oxides present, and it is qualitatively analogous to that exhibited by synthetic iron oxides. The highest adsorption was found for arsenate ions, whereas the lowest was found for selenate, with chromate and molybdate ions showing an intermediate behavior. Environ Toxicol Chem 2014;33:2217–2224. # 2014 SETAC Keywords: Oxyanion

Adsorption

Speciation

Charge distribution model

Soil contamination

enabled us to calculate the reactive surface area and the phosphate loading on the iron oxides present in the soil, and we were also able to consider the competition between natural organic matter (NOM) and phosphate by including FeNOM surface groups in the model. These 3 model parameters were obtained for 2 different horizons of the ferralic soil: a surface horizon that had previously received large amounts of fertilizer and a deeper horizon containing minimal amounts of phosphate. Accurate description of such parameters for inclusion in the charge distribution model will be helpful for describing the adsorption of other anionic species in field samples. The present study focuses on the use of the charge distribution model to describe the adsorption of selected trace elements (Cr, Se, Mo, and As) in ferralic soils. High concentrations of these elements in the soil solution can have toxic effects on plants and microorganisms. It is therefore important to understand the processes and mechanisms of adsorption that bring about the immobilization of these elements. Considering the chemical speciation in solution and the typical pH and redox potential in ferralic soils, it is known that the oxyanionic forms (CrO4, SeO4, MoO4, AsO4) will predominate and will determine the behavior of the species in the soil solution. Moreover, the mechanisms of adsorption of chromate, selenate, molybdate, and arsenate on goethite have been investigated in numerous microscopic studies. The information obtained in those studies, such as the type of surface complexes formed, are included in the charge distribution model in the present study to describe the adsorption of these oxyanions in ferralic soils. The results will enable us to determine and predict the speciation of the oxyanions in the solid fraction and soil solution.

INTRODUCTION

The capacity of soils to adsorb trace elements from soil solution is of interest regarding both soil fertility and environmental issues, such as the remediation of contaminated soils. Trace elements, which comprise some of the most toxic inorganic contaminants found in soils and aquatic systems, can be of natural or anthropogenic origin [1]. Adsorption is one of the main processes that determines the mobility and retention of trace elements in soils, and the study of this process is therefore essential for understanding how these elements are distributed between the solid fraction and soil solution. Surface complexation models have become important research tools and are used to establish how chemical compounds behave in the environment. Use of such models enables estimation of the bioavailable fraction of anionic and cationic species of trace elements in soils [2–4]. Application of the models requires information about the chemical and physical mechanisms involved in ion adsorption reactions, in addition to detailed knowledge about reactive surfaces and the solid–solution interface. Although much progress has been made in developing surface complexation models for mineral oxides, further work is still needed to describe ion adsorption in soils. In the first part of this research, study of the adsorption–desorption of the phosphate ion enabled us to calibrate the charge distribution model and thus obtain a set of structural parameters associated with natural iron oxides, which are the main compounds responsible for phosphate retention in soils [5]. Goethite, which has been extensively studied with the charge distribution model [6–10], was used as a proxy to describe phosphate behavior in ferralsols. In our previous study [5], calibration of the charge distribution model All Supplemental Data may be found in the online version of this article. * Address correspondence to [email protected]. Published online 19 March 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/etc.2581

MATERIALS AND METHODS

We collected samples of 2 horizons of a ferralic soil: a surface horizon (Ap1; 0–15 cm) and a deeper horizon (Bw1; 136– 2217

2218

Environ Toxicol Chem 33, 2014

C. Perez et al.

Table 1. Concentrations of the solutions used in the adsorption experiments Stock solution K2CrO4 10 mM Na2SeO4 40 mM Na2MoO4  2H2O 40 mM NaH2AsO4 40 mM

Soil sample

Concentration added (mM)

Ap1 Bw1 Ap1 Bw1 Ap1 Bw1 Ap1 Bw1

10 10 80 80–400 160 160–400 160 160–400

Ap1 ¼ surface horizon (0–15 cm); Bw1 ¼ deeper horizon (136–200 cm)

200 cm). The site location and the main physicochemical properties of both soil horizons are reported in the first part of this series [5] and shown in Supplemental Data, Table S1. Adsorption of chromate, selenate, molybdate, and arsenate was studied in batch experiments carried out at pH 3 to 10. The volume of stock solution required to obtain the corresponding concentration of oxyanion was added to 20 mL of soil suspension (10 g/L; Table 1). The concentrations were chosen with the aim of yielding a significant degree of absorbance. The ionic strength was held constant in all experiments (0.1 M in KNO3), and the pH of the suspensions was adjusted by adding small volumes of 0.1 M HNO3 or 0.1 M KOH throughout the experiment. After checking to ensure that longer times did not significantly affect the results, the samples were shaken for 24 h until the adsorption equilibrium was reached. All of the experiments were carried out at 25  1 8C under a nitrogen atmosphere. Once the adsorption equilibrium was reached, the suspensions were filtered (0.45 mm Millipore filters) and the concentration of the oxyanion in solution was measured. The concentrations of chromate and arsenate in solution were measured spectrophotometrically by the diphenylcarbazide method [11] and the method proposed by Lenoble et al. [12], respectively. The concentrations of selenate and molybdate were determined by inductively coupled plasma optical emission spectrometry (Óptima 3300 DV; Perkin Elmer). The amount of oxyanion adsorbed was calculated as the difference between the total concentration added to the suspension and the concentration measured in solution. This amount was expressed per kilogram of dithionite-citrate-bicarbonate–extracted iron oxide, and a molar mass of 89 g/mol Fe was assumed for the iron oxide.

All experiments were carried out in duplicate to ensure their reproducibility. All chemicals were of Merck pro analysis grade (p.a.) quality. The water used in the experiments was double distilled and CO2-free. Charge distribution model

The charge distribution surface complexation model [13,14] was used to describe the process of adsorption of the different anions. Although this model has been used in several studies to investigate the reactivity of mineral oxides [7,8,15–17], its application to natural soils and sediments remains a challenge because of the complexity and heterogeneity of these systems [18,19]. Details of the charge distribution model, the model calibration, and the parameters required to study the retention of ions on ferralic soils are reported in the previous study [5]. This model considers the structure of the different complexes adsorbed on the surface, which is also related to the spatial distribution of the charge at the solid–solution interface. Because the reactivity of the oxyanions in ferralsols is governed mainly by the presence of iron oxides, the surface complexes used for each of the oxyanions under study were chosen by taking into consideration available spectroscopic data on iron oxides. Following the approach proposed by Hiemstra et al. [18], the contribution of clay minerals and aluminum oxides to oxyanion adsorption was discarded in an attempt to simplify the modeling calculations. To consider the effect of NOM, the surface groups’ ¼FeNOM were included in the modeling. These surface groups contribute only to the surface charge of the natural samples and do not form surface complexes with the oxyanions. Optimization of the parameters required for describing the experimental adsorption to the ferralic soil with the charge distribution model was carried out with the ECOSAT speciation program [20] combined with the FIT program [21]. Surface complexation parameters specific for the natural iron oxides present in the sample and for the solid–solution interface, such as surface area, site density, capacitance, or protonation constants, are shown in Supplemental Data, Table S2. RESULTS AND DISCUSSION

Cr(VI) adsorption

The effect of pH on the adsorption of chromate in the ferralic soil horizons is illustrated in Figure 1. In the sample of the Ap1 horizon, the adsorption decreased as the pH increased between 3 (80% adsorption) and 8 (10% adsorption). The decrease can

Figure 1. Adsorption envelopes for chromate in (A) Ap1 and (B) Bw1 soil horizons. Solid lines represent the charge distribution model fit. Dotted and dashed lines represent the abundance of protonated bidentate and outer sphere monodentate (Mext) surface complexes, respectively, according to the charge distribution model.

Oxyanion adsorption on ferralic soil

Environ Toxicol Chem 33, 2014

be attributed to an increase in the negative charge at the mineral surface on increasing the pH. In the samples of the Bw1 horizon, the adsorption reached a maximum value of approximately 4.3 mmol/kg oxide (96% adsorption) and remained almost constant between pH 3 and 5; thereafter, it decreased until reaching approximately 10% adsorption at pH 9. For the Bw1 horizon, the abrupt change in adsorption occurred at a pH value close to the dissociation constant (pKa) for chromate, pKa2 ¼ 6.5 [22], which indicates a change from the monoprotonated species to the completely deprotonated form. The effect of pH on adsorption was less marked in the Ap1 horizon. Comparison of the results obtained for samples from the 2 horizons reveals that the Bw1 horizon has a higher capacity to adsorb chromate. It is possible that the higher content of NOM and clay minerals in the Ap1 horizon hindered the adsorption of chromate because both supply a negative charge to the surface and, moreover, the NOM competes for the adsorption sites present in the iron oxides of the soil system. The adsorption edges were analyzed using the charge distribution model, and the parameters required to describe the adsorption of chromate are shown in Table 2. Although phosphate complexation parameters are known for ideal goethite and can be used with the charge distribution model to describe phosphate adsorption on ferralic soils [5], no such parameters are available for chromate. Therefore, the constants for the surface complexes were considered as adjustable parameters. The chromate monodentate and bidentate surface complexes proposed for goethite by Fendorf et al. [23] were used in the model. For the Ap1 sample, the best fit (Figure 1A) was obtained by considering the formation of a protonated bidentate complex and an outer sphere monodentate complex (Mext). Noninclusion of the latter complex, the existence of which was initially proposed by Hayes et al. [24], leads to underestimation of adsorption at pH > 7. A good fit was also obtained for the Bw1 sample (Figure 1B) on considering the same type of complexes (protonated bidentate and Mext). On calculating the speciation using the values obtained for the complexation constants, we found that the external sphere complexes of chromate were the most abundant at pH > 7 in both horizons, although the proportion was lower in the Ap1 sample than in the Bw1 sample (Figure 1A and B). Furthermore, the model predicted that at pH 6. The decrease was continuous throughout the pH range, and there was no inflection point such as that observed for the chromate ion (Figure 1). It should be taken into account that Se was present in the solution as SeO42–across the entire pH range, and the protonated forms were only found at pH < 2 (pKa2 ¼ 1.7) [22]. Interestingly, adsorption of selenate is similar in both soil horizons, whereas for chromate, as shown before, the adsorption was higher in the Bw1 sample. This will be discussed more extensively in section Comparison of CrO4, SeO4, MoO4, AsO4, and PO4 adsorption. Because of difficulties in differentiating the coordination modes of selenate [26,27], we considered inner sphere and outer sphere complexes on using the charge distribution model to describe selenate adsorption. Initially, we considered only monodentate complexes because some researchers have postulated that these complexes predominate in the process of adsorption of this anion on iron oxides [6,28]. To evaluate the importance of outer sphere complexes in the Ap1 horizon, we initially included only this type of complex in the model. We found that an outer sphere monodentate complex (Mext) adequately described selenate adsorption, although it slightly underestimated the adsorption at pH > 5. By including inner

Table 2. Oxyanion surface complexation parameters for soil samples obtained with the charge distribution modela Species

Dz0

Dz1

Dz2

Log K (Ap1)

Log K (Bw1)

Fe2O2CrOOH FeOH-CrO4

0.50 1

0.50 –2

FeOSeO3 FeOH-SeO4

0.36 1

–1.36 –2

Fe2O2AsO2 Fe2O2AsOOH

0.47 0.58

–1.47 –0.58

Fe2O2MoO2 Fe2O2MoOOH

0.46 0.63

–1.46 0.63

0 0 r2 RMSE 0 0 r2 RMSE 0 0 r2 RMSE 0 0 r2 RMSE

26.36  0.21 13.13  0.17 0.87 8.7  107 8.90  0.08 11.07  0.57 0.96 4.6  106 29.29 32.69 0.77 8.8  106 17.43  0.12 23.57  0.32 0.99 8.2  106

27.57  0.16 11.76  0.43 0.92 1.1  106 7.51  0.16 9.11  0.41 0.92 1.3  105 29.29 32.69 0.85 2.7  105 17.95  0.37 23.64  0.41 0.95 3.6  105

a Standard deviations are given for the optimized parameters. Dz0, Dz1, Dz2 ¼ charge variation at 0-, 1-, and 2-planes, respectively; Ap1 ¼ surface horizon (0–15 cm); Bw1 ¼ deeper horizon (136–200 cm); RMSE ¼ root mean square error.

2220

Environ Toxicol Chem 33, 2014

C. Perez et al.

Figure 2. Adsorption envelopes for selenate in (A) Ap1 and (B) Bw1 soil horizons. Solid lines represent the charge distribution model fit. Dotted and dashed lines represent the abundance of the outer sphere monodentate (Mext) and inner sphere monodentate (M) surface complexes, respectively, according to the charge distribution model.

sphere monodentate complexes (M), we improved the model fit across the entire pH range (Figure 2A). Consideration of Mext and M complexes also enabled the experimental behavior to be reproduced for the Bw1 sample (Figure 2B). Figure 2A and 2B shows the abundance of the different surface species predicted by the charge distribution model that included the complexation parameters obtained (Table 2). In both horizons, the outer sphere complex increased in abundance at approximately pH > 5, whereas the inner sphere complexes clearly predominated at lower pH. This is consistent with previous findings for the different iron oxides present in this type of soil [28,29]. The results suggest the formation of an inner sphere monodentate complex, which decreases in abundance as the pH increases, and of an outer monodentate complex at pH > 6, although the existence of these surface complexes must be further confirmed with additional microscopic data. A second modeling scenario was analyzed for selenate because some authors suggest the existence of bidentate inner sphere surface complexes. Kersten and Vlasova [30] have recently modeled selenate adsorption on goethite with the charge distribution model considering the existence of both monodentate outer sphere and bidentate inner sphere complexes. The complexation constants obtained by these authors were initially used in the present study to predict selenate adsorption in both soil horizons. Simulations using these constants (and the charge distribution coefficients (proposed by Kersten and Vlasova [30]) led to a good prediction of selenate adsorption on sample Ap1 but overestimates the adsorption on sample Bw1 (Supplemental Data, Figure S1). Optimization of the surface complexation

constants allows for a better fit of the experimental results, and more comparable constants were found for the 2 soil horizons (Supplemental Data, Table S4). Whereas in the first scenario the difference between the constants for each soil is almost 2 units in log scale, in the second scenario this difference is reduced to 1.5 units. These results confirm that modeling calculations on their own are not sufficient to infer the structure of the surface complexes, and additional support information is needed. Mo(VI) adsorption

The experimental results for molybdate adsorption at different pH values are shown in Figure 3. In both horizons, there was a sharp decrease in adsorption as the pH increased from 4.5. Thus, the adsorption was almost 100% at pH 7. This is similar to previous findings for goethite and ferrihydrite [31,32] and is a result of the increased negative charge on the oxide surface as the pH increases and of the greater abundance of MoO42– than of protonated species. The sharp decrease in molybdate adsorption takes place at a slightly higher pH than the values of the pKa values for H2MoO4, pKa1 ¼ 4.00, and pKa2 ¼ 4.24 [33]. There are discrepancies in the results of different spectroscopic studies because some authors suggest the existence of bidentate complexes [34,35] and others suggest the existence of monodentate complexes [32,36] to describe the reactivity of molybdate on iron oxides. It is difficult to distinguish between these types of surface complexes when they are present simultaneously, however, because the most intense signal of

Figure 3. Adsorption envelopes for molybdate in (A) Ap1 and (B) Bw1 soil horizons. Solid lines represent the charge distribution model fit. Dotted and dashed lines represent the abundance of protonated bidentate and deprotonated bidentate surface complexes, respectively, according to the charge distribution model.

Oxyanion adsorption on ferralic soil

the bidentate complexes overlaps with that of the monodentate complexes. In the present study, on using the charge distribution model to interpret the reactivity of molybdate, we considered the formation of inner sphere bidentate complexes, which were previously identified by Arai [34] on the surface of goethite. Because it is difficult to differentiate protonated and deprotonated complexes by spectroscopic techniques, we considered the formation of both protonated bidentate complexes and deprotonated bidentate complexes. Inclusion of these complexes in the model yielded good fits to the experimental data for both the Ap1 horizon (Figure 3A) and the Bw1 horizon (Figure 3B), and the values of the complex formation constants were similar for both samples (Table 2). Although we cannot rule out the presence of monodentate complexes at the surface, it seems (in light of the spectroscopic results for iron oxides and by the application of the charge distribution model) that the coordination of molybdate with the reactive groups on the mineral surfaces present in the soil samples could occur mainly via the formation of innersphere bidentate complexes. As previously shown, additional spectroscopic and macroscopic experimental data would be necessary to confirm the existence of these type of complexes and to definitely rule out the formation of monodentate surface complexes for molybdate. Figure 3 shows the speciation of Mo(VI) in soil samples, as determined from the parameters listed in Table 2. The modeling calculations show that the main surface species in the Ap1 sample at pH > 4.5 are the deprotonated bidentate complexes, whereas the protonated bidentate complexes predominate at a lower pH. The bidentate complexes exert a greater effect in the Bw1 sample than in the Ap1 sample, and they predominate across almost the entire pH range, possibly because of the greater degree of adsorption on the oxides present in this horizon and the lower effect of the NOM. At pH 7) (Figure 4). The adsorption of arsenate on both horizons was described by assuming the formation of the same surface complexes and using the same values of the complexation constants. In our previous study [5], this was also achieved for a low phosphate loading, whereas for a higher concentration of the anion in the Bw1 horizon, 1 of the 2 complexation constants for phosphate was fitted. This was probably necessary because we did not initially consider the formation of a phosphate precipitate with the iron and/or aluminum present in the soil solution. Likewise, for higher concentrations of arsenate, we would probably have to consider the formation of scorodite (FeAsO4  2H2O) or aluminum arsenate precipitates at relatively low pH values. If the formation of such precipitates were not taken into account, the constants of formation of surface complexes would have to be fitted. The parameters shown in Table 2 were used to determine the speciation of arsenate in both of the ferralic soil horizons under study (Figure 4). The deprotonated bidentate complexes in the Ap1 horizon predominated at pH > 4, whereas this

Figure 4. Adsorption envelopes for arsenate in (A) Ap1 and (B) Bw1 soil horizons. Solid lines represent the charge distribution model fit. Dotted and dashed lines represent the abundance of protonated bidentate and deprotonated bidentate surface complexes, respectively, according to the charge distribution model.

Environ Toxicol Chem 33, 2014

5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

[SeO 4 ]ads (mmol/kg oxide)

35

A

30 25 20 15 10 5

B

0 3

[MoO4 ]ads (mmol/kg oxide)

C. Perez et al.

4

5

6

pH

7

8

9

3

10

90

80

80

70

[AsO4 !ads (mmol/kg oxide)

[CrO4 ]ads (mmol/kg oxide)

2222

70 60 50 40 30 20 10

C

4

5

6 pH

7

8

9

60 50 40 30 20 10

D

0

0 3

4

5

6 pH

7

8

9

2

3

4

5

6

pH

7

8

9

10

11

Figure 5. Adsorption envelopes on soil samples for (A) [CrO4] ¼ 10 mM, (B) [SeO4] ¼ 80 mM, (C) [MoO4] ¼ 160 mM, and (D) [AsO4] ¼ 160 mM. Filled symbols correspond to sample Ap1 and empty symbols, to sample Bw1. Solid and dashed lines represent the charge distribution model fit.

predominance in the Bw1 horizon extended across almost the entire pH range studied. Despite clearly indicating the preferential adsorption of bidentate complexes, the available spectroscopic information [40,41] does not distinguish whether the protonated or deprotonated form predominates. The predominance of deprotonated bidentate complexes was also proposed by Stachowicz et al. [16] and Kanematsu et al. [42] in surface complexation model calculations for synthetic iron oxides. Comparison of CrO4, SeO4, MoO4, AsO4, and PO4 adsorption

In view of the experimental and modeling results for oxyanion adsorption on samples Ap1 and Bw1, and after comparison of the adsorption results at the same oxyanion loading (Figure 5), it can be stated that the behavior of selenate clearly differs from that found for chromate, arsenate, and molybdate. Whereas the latter anions present similar qualitative behavior, showing greater adsorption on soil Bw1 compared with soil Ap1, selenate is adsorbed approximately in the same amount by both soil horizons. Selenate exhibits significantly lower adsorption compared with the other oxyanions (Figure 6), and the surface complexes formed are more weakly bound to the natural iron oxides. Although it might be thought that selenate can be partly reduced in the Ap1 horizon, causing this anomalous behavior, we are not able, with the available information, to assess this hypothesis. Future studies will be carried out to analyze which are the reactive sites involved in selenate adsorption and the potential role of redox processes. These results are partially reflected in the complexation constants obtained when the adsorption is described with the charge distribution model. Selenate complexation constants were slightly higher in soil Ap1, accounting for the lower content

of iron oxides and the higher contribution of organic matter compared with soil Bw1. On the other hand, complexation constants for molybdate and arsenate are very similar (or the same in the case of arsenate) in both soil horizons. Differences in the adsorbed amounts are related to the differences in the iron oxide and organic matter contents. Finally, complexation constants for chromate showed larger differences between soil horizons, the constant of the bidentate complex (predominant surface species at pH CrO4 > MoO4 > SeO4. This result is expected from analysis of the correlation between the complexation constants (obtained for iron oxides by Dzombak and Morel [43] using the generalized 2-layer model) and the second (or third in the case of arsenate) dissociation constant for the oxyanions (r2 ¼ 0.99). This correlation indicates that the affinity of phosphate and arsenate for the soil samples is similar and much higher than that shown by chromate and molybdate; the affinity of the selenate ion for iron oxides is the lowest of the series. Although in the review by Dzombak and Morel [43] data for molybdate are not included, considering its pKa value, the affinity is expected to lie between that of selenate and chromate. The similarity between the 2 affinity series (ferralsol and iron oxides) indicates that the iron oxides present in the mineral fraction of ferralic soils are the main components governing the retention and mobility of oxyanions.

Environ Toxicol Chem 33, 2014

2223

For those oxyanions (phosphate and arsenate) for which the charge distribution model surface complexation constants are known for ideal goethite, these constants can be used to predict the adsorption on natural samples. However, for those oxyanions (chromate, selenate, and molybdate) for which there are no data concerning adsorption on ideal goethite, the surface complexation constants were considered as fitting parameters. Additional information derived from spectroscopic studies will be of great value for confirming the surface speciation proposed in the present study. The results also enabled us to determine the affinity of the different oxyanions for natural iron oxides. As expected, phosphate and arsenate displayed similar affinities, which were higher than those observed for the other anionic forms; selenate displayed the lowest affinity for iron oxides, and chromate and molybdate showed intermediate affinities. These results are consistent with the order of affinity predicted from the pKa values of each of the anions: PO4  AsO4 > CrO4 > MoO4 > SeO4. SUPPLEMENTAL DATA

Tables S1–S4. Figures S1–S2. (143 KB DOCX) Acknowledgment—This work was financially supported by the Ministerio de Ciencia e Innovación under the research project CTM2011-24985 and by the Xunta de Galicia with the research project PGIDIT10PXIB209014PR. The authors thank P. Bermejo of the Department of Analytical Chemistry, Nutrition, and Bromatology of the University of Santiago de Compostela for the inductively coupled plasma optical emission spectrometry measurements.

CONCLUSIONS

REFERENCES

Adsorption experiments were carried out to study the adsorption of oxyanions of environmental interest—arsenate, molybdate, chromate, and selenate—on 2 horizons of a ferralsol. The mobility and retention of oxyanions in ferralic soils are mainly governed by the presence of iron oxides, which constitute the predominant mineral fraction in these types of soils. The adsorption of different oxyanions was always greater in the deeper horizon, Bw1. This horizon has a lower content of NOM, and there is therefore less competition between oxyanions and the organic matter for adsorption sites on the iron oxides than in the Ap1 surface horizon. The adsorption decreased with increasing pH, as also found for other ionic species (phosphate, arsenate, chromate) in synthetic mineral oxides with a relatively high point of 0 charge. The experimental results obtained for the different oxyanions were satisfactorily described by the charge distribution model. The surface complexes used in the modeling calculations were selected by taking into account the spectroscopic and molecular information on iron oxides, mainly goethite, available in the literature. In general, the model predictions indicate that the dominant surface species are the protonated bidentate complex for chromate, the outer sphere monodentate complex for selenate, and the deprotonated bidentate complex for molybdate and arsenate, although other types of surface complexes may be formed depending on the pH of the system. The good fits obtained across the entire pH range for the different oxyanions analyzed in the present study and the results already obtained for phosphate indicate that use of goethite as a proxy for the mineral fraction responsible for the retention of these ions in a ferralic soil is a suitable approach.

1. Hooda PS. 2010. Trace Elements in Soils. John Wiley & Sons, Chichester, UK. 2. Gustafsson JP. 2001. Modelling competitive anion adsorption on oxide minerals and an allophane-containing soil. Eur J Soil Sci 52:639– 653. 3. Lumsdon DG. 2004. Partitioning of organic carbon, aluminium and cadmium between solid and solution in soils: Application of a mineral-humic particle additivity model. Eur J Soil Sci 55:271– 285. 4. Bonten LTC, Groenenberg JE, Weng LP, van Riemsdijk WH. 2008. Use of speciation and complexation models to estimate heavy metal sorption in soils. Geoderma 146:303–310. 5. Perez C, Antelo J, Fiol S, Arce F. 2014. Modeling oxyanion adsorption on ferralic soil. 1. Parameter validation with phosphate ion. Environ Toxicol Chem 33: 2208–2216. 6. Rietra RPJJ, Hiemstra T, van Riemsdijk WH. 2001. Comparison of selenate and sulfate adsorption on goethite. J Colloid Interface Sci 240:384–390. 7. Antelo J, Avena M, Fiol S, López R, Arce F. 2005. Effects of pH and ionic strength on the adsorption of phosphate and arsenate at the goethite–water interface. J Colloid Interface Sci 285:476–486. 8. Rahnemaie R, Hiemstra T, van Riemsdijk WH. 2007. Geometry, charge distribution, and surface speciation of phosphate on goethite. Langmuir 23:3680–3689. 9. Stachowicz M, Hiemstra T, van Riemsdijk WH. 2008. Multicompetitive interaction of As(III) and As(V) oxyanions with Ca2þ, Mg2þ, PO43, and CO32 ions on goethite. J Colloid Interface Sci 320: 400–414. 10. Goli E, Rahnemaie R, Hiemstra T, Malakouti MJ. 2011. The interaction of boron with goethite: Experiments and CD–MUSIC modeling. Chemosphere 82:1475–1481. 11. Clesceri LS, Greenberg AE, Eaton AD. 1998. Standard Methods for the Examination of Water and Wastewater. American Public Health Association, Washington, DC. 12. Lenoble V, Deluchat V, Serpaud B, Bollinger JC. 2003. Arsenite oxidation and arsenate determination by the molybdene blue method. Talanta 61:267–276.

2224

Environ Toxicol Chem 33, 2014

13. Hiemstra T, van Riemsdijk WH. 1996. A surface structural approach to ion adsorption: The charge distribution (CD) model. J Colloid Interface Sci 179:488–508. 14. Hiemstra T, van Riemsdijk WH. 2006. On the relationship between charge distribution, surface hydration, and the structure of the interface of metal hydroxides. J Colloid Interface Sci 301:1–18. 15. Ridley MK, Machesky ML, Wesolowski DJ, Palmer DA. 2005. Surface complexation of neodymium at the rutile–water interface: A potentiometric and modeling study in NaCl media to 250 8C. Geochim Cosmochim Acta 69:63–81. 16. Stachowicz M, Hiemstra T, van Riemsdijk WH. 2006. Surface speciation of As(III) and As(V) in relation to charge distribution. J Colloid Interface Sci 302:62–75. 17. Antelo J, Fiol S, Mari~ no S, Arce F, Gondar D, López R. 2009. Copper adsorption on humic acid coated gibbsite: Comparison with single sorbent systems. Environ Chem 6:535–543. 18. Hiemstra T, Antelo J, Rahnemaie R, van Riemsdijk WH. 2010. Nanoparticles in natural systems I: The effective reactive surface area of the natural oxide fraction in field samples. Geochim Cosmochim Acta 74:41–58. 19. Devau N, Hinsinger P, Le Cadre E, Colomb B, Gerard F. 2011. Fertilization and pH effects on processes and mechanisms controlling dissolved inorganic phosphorus in soils. Geochim Cosmochim Acta 75:2980–2996. 20. Keizer MG, van Riemsdijk WH. 1998. ECOSAT: Equilibrium Calculation of Speciation and Transport. Technical Report. Department Soil Science and Plant Nutrition, Wageningen Agricultural, University, Wageningen, The Netherlands. 21. Kinniburgh DG. 1993. FIT: User Guide. Technical Report WD/93/23. British Geological Survey, Keyworth, UK. 22. Perrin DD. 1982. Ionization Constants of Inorganic Acids and Bases in Aqueous Solution. Pergamon, Oxford, UK. 23. Fendorf S, Eick MJ, Grossl P, Sparks DL. 1997. Arsenate and chromate retention mechanisms on goethite. 1. Surface structure. Environ Sci Technol 31:315–320. 24. Hayes KF, Papelis C, Leckie JO. 1988. Modeling ionic strength effects on anion adsorption at hydrous oxide/solution interfaces. J Colloid Interface Sci 125:717–726. 25. Weerasooriya R, Tobschall HJ. 2000. Mechanistic modelling of chromate adsorption onto goethite. Colloid Surf A 162:167–175. 26. Hayes KF, Roe AL, Brown GE, Hodgson KO, Leckie JO, Parks GA. 1987. In situ X-ray absorption study of surface complexes: Selenium oxyanions on a-FeOOH. Science 238:783–786. 27. Manceau A, Charlet L. 1994. The mechanism of selenate adsorption on goethite and hydrous ferric oxide. J Colloid Interface Sci 168:87–93. 28. Peak D, Sparks DL. 2002. Mechanisms of selenate adsorption on iron oxides and hydroxides. Environ Sci Technol 36:1460–1466.

C. Perez et al. 29. Winja H, Schulthess CP. 2000. Vibrational spectroscopy study of selenate and sulphate adsorption mechanisms on Fe and Al (hydr)oxide surfaces. J Colloid Interface Sci 229:286–297. 30. Kersten M, Vlasova N. 2013. The influence of temperature on selenate adsorption by goethite. Radiochim Acta 101:413–419. 31. Manning BA, Goldberg S. 1996. Modeling competitive adsorption of arsenate with phosphate and molybdate on oxide minerals. Soil Sci Soc Am J 60:121–131. 32. Gustafsson JP. 2003. Modelling molybdate and tungstate adsorption to ferrihydrite. Chem Geol 200:105–115. 33. Lindsay WL. 1979. Chemical Equilibria in Soils. John Wiley & Sons, New York, NY, USA. 34. Arai Y. 2010. X-ray absorption spectroscopic investigation of molybdenum multinuclear sorption mechanism at the goethite–water interface. Environ Sci Technol 44:8491–8496. 35. Kashiwabara T, Takahashi Y, Tanimizu M, Usui A. 2011. Molecularscale mechanisms of distribution and isotopic fractionation of molybdenum between seawater and ferromanganese oxides. Geochim Cosmochim Acta 75:5762–5784. 36. Xu N, Christodoulatos C, Braida W. 2006. Modeling the competitive effect of phosphate, sulfate, silicate, and tungstate anions on the adsorption of molybdate onto goethite. Chemosphere 64:1325– 1333. 37. Violante A, Pigna M. 2002. Competitive sorption of arsenate and phosphate on different clay minerals and soils. Soil Sci Soc Am J 66:1788–1796. 38. Sherman DM, Randall SR. 2003. Surface complexation of arsenic(V) to iron(III) (hydr)oxides: Structural mechanism from ab initio molecular geometries and EXAFS spectroscopy. Geochim Cosmochim Acta 67:4223–4230. 39. Antelo J, Fiol S, Mari~ no S, Arce F, Gondar D, López R. 2010. Analysis of phosphate adsorption onto ferrihydrite using the CD-MUSIC model. J Colloid Interface Sci 347:112–119. 40. Harrington R, Hausner DB, Bhandari N, Strongin DR, Chapman KW, Chupas PJ, Middlemiss DS, Grey CP, Parise JB. 2010. Investigation of surface structures by powder diffraction: A differential pair distribution function study on arsenate sorption on ferrihydrite. Inorg Chem 49:325– 330. 41. Catalano JG, Luo Y, Otemuyiwa B. 2011. Effect of aqueous Fe(II) on arsenate sorption on goethite and hematite. Environ Sci Technol 45:8826–8833. 42. Kanematsu M, Young TM, Fukushi K, Green PG, Darby J. 2010. Extended triple layer modeling of arsenate and phosphate adsorption on a goethite-based granular porous adsorbent. Environ Sci Technol 44:3388–3394. 43. Dzombak DA, Morel FMM. 1990. Surface Complexation Modelling: Hydrous Ferric Oxide. John Wiley & Sons, New York, NY, USA.

Modeling oxyanion adsorption on ferralic soil, part 2: chromate, selenate, molybdate, and arsenate adsorption.

High levels of oxyanions are found in the soil environment, often as a result of human activity. At high concentrations, oxyanions can be harmful to b...
527KB Sizes 4 Downloads 2 Views