Sorption of selenium(IV) onto magnetite in the presence of silicic acid

https://doi.org/10.1016/j.jcis.2008.09.052Get rights and content

Abstract

Sorption of selenium(IV) and silicic acid onto magnetite (Fe3O4) was investigated in binary systems, with concentrations of silicic acid under the solubility limit of amorphous silica. Using the double diffuse layer model (DDLM), surface complexation constants of selenium(IV) and H4SiO4 onto magnetite were extracted using Fiteql 4.0. Then, prediction curves of the sorption of selenium(IV) in the presence of silicic acid onto magnetite were obtained, using the calculated surface complexation constants.

Finally, laboratory experiments were performed and showed a competition between selenium(IV) and silicic acid for the surface sites of magnetite. Experimental results matched the model predictions, confirming its ability to model qualitatively and quantitatively the ternary system.

Graphical abstract

This figure represents experimental sorption edges of selenium(IV) onto magnetite in the presence of silicic acid, as well as calculations done with double diffuse layer model to account for the experimental observations.

  1. Download : Download full-size image

Introduction

Selenium-79 is one of the long-lived fission products which have to be taken into account in safety assessments of nuclear waste underground repositories. In a reducing environment such as interstitial waters of Callovo-Oxfordian clays of the French underground laboratory of Bure (France) (Table 1), selenium is likely to exist under three oxidation states (−II, 0, +IV). Elementary Se(0), and Se(−II) which is likely to precipitate in the presence of Fe(II), would represent an immobile fraction. On the contrary, if selenium(IV) is present, its oxoanionic form SeO2−3 would give selenium a high potential mobility.

In the French spent fuel disposal concept, the vitrified waste package (vitrified waste + stainless steel container) for long-life and high radioactive isotopes, will be overpacked in a non-alloy steel surcontainer. In case of an infiltration of water, the first barrier to the migration of selenium would thus be the (corroded) surcontainer. It has been evidenced that, in the reducing conditions encountered in the French underground site of Bure, magnetite (Fe3O4) will be one of the main iron corrosion products [1]: thus, it has been chosen as a sorbent material for this study.

Retention of SeO2−3 onto iron corrosion products could then be possible due to immobilization process, since studies on selenium sorption have indicated that it can be strongly adsorbed by iron oxyhydroxides and oxides [2], [3], [4], [5], [6], [7], [8].

In the geological conditions of an underground repository, several aqueous species could compete with selenite ions for the sorption sites of iron corrosion products. Indeed, other anions or neutral species would be present in noticeable quantities in interstitial waters. From concentrations reported in the French test site (Table 1 [1]), sulfate, carbonate and silicate species would be the main competitors.

Sorption studies with multisorbate systems including oxoanions can be found in the literature. Concerning competition studies on selenium(IV) retention onto iron (oxyhydr)oxides, sulfate anions have been one of the most reported in the literature. Fujikawa et al. [5] showed an effect of sulfate (simultaneously added in the suspensions), lowering the sorption of selenite ions onto hematite and magnetite, but for quite high concentrations of sulfate compared to the one of selenium (10−3 to 10−1 and 10−11 N, respectively). Xu et al. [9], [10] reported inexistent competition effect of sulfate ions (simultaneously added in the suspension), on the adsorption of Mo species (molybdate MoO2−4 and tetrathiomolybdate MoS2−4) onto goethite under anoxic conditions. Xu et al. [9], [10] then suggested that adsorbed sulfate were more likely to form outer-sphere complexes onto goethite.

Identical results were obtained by Goh et al. [11], who investigated the effect of SO2−4 ions, simultaneously introduced in the suspension, on the adsorption of As and Se species on a soil found in Singapore containing significant amount of Fe. Even when its introduced concentration was increased, the influence of SO2−4 on As(V) and Se(IV) adsorption was found to be insignificant.

Competition effects of other anions like phosphate on the retention of selenium(IV) have also been examinated. Monteil-Riviera et al. [12], who studied the sorption of selenite ions onto hydroxyapatite surface, concluded that the presence of phosphate ions in solution lowers SeO2−3 sorption by direct competition.

Xu et al. [9], [10] noticed that phosphate anions, among sulfate and silicate ions (simultaneously added in the suspension), were the strongest competitors of Mo oxoanions for the adsorption sites of goethite under anoxic conditions. Xu et al. [9], [10] suggested that adsorbed phosphate were more likely to form inner-sphere complexes on goethite. Same effects were reported by Goh et al. [11], who showed that the presence of phosphate ions led to competition effect between PO3−4 and As or Se oxoanions for surface sites of a Singapore soil, which was presumably due to the formation of surface complexes and to the surface accumulation or precipitation of PO3−4 species.

Among the different aqueous species, silicic acid could play a major role. Besides a moderate equilibrium concentration in interstitial waters of the Callovo-Oxfordian clays (around 10−4 molL−1), silicic acid could also come from the degradation of glass canisters, resulting in an increasing of local concentrations in the close vicinity of the canisters. Recently, we showed in a previous publication that sorption of silicic acid on solid surfaces can modify the availability of the surface sites, and also the surface charge, which can influence the global surface reactivity of solids [13].

The effect of silicate ions on the adsorption of oxoanions species have also been the subject of some studies. Xu et al. [9], [10] showed that silicate ions (simultaneously added in the suspension) had little competitive effects on the retention of Mo oxoanions onto goethite. Authors then suggested that silicate ions were more likely to form outer-sphere complexes onto goethite.

Swedlund et al. [14] showed the inhibition of the retention of oxoanions of arsenic onto ferrihydrite by competition effect of silicic acid simultaneously introduced in the suspension.

Balistrieri and Chao [15], [16] also studied the competition effect of various anions like phosphate, silicate, molybdate and sulfate on the sorption of selenium(IV) onto goethite and amorphous iron oxide at pH 7.0. The introduced concentration of selenium(IV) was closed to 10−6 molL−1 and the initial concentrations of competiting species were ranging from 10−6 to 10−1 molL−1. For goethite, silicate ions became to act as strong competitors, from an introduced concentration between 10−4 and 10−3 molL−1 whereas initial concentrations between 10−3 and 10−2 molL−1 were necessary for amorphous iron oxide.

Using laboratory experiments, this work is basically aiming a deeper insight in the role of the silicic acid, by studying its influence on the sorption of selenium(IV) on magnetite. Surface complexation models will be used to model experimental observations and predict the behavior of this ternary system.

Section snippets

Surface complexation model description

Experimental results were modeled by using surface complexation models (SCMs), which consider functional surface groups as complexing ligands in solution. These SCMs are generally applied to oxide surfaces which become hydrated in suspension. This result in a surface charge, depending on the pH, which can be interpreted as the development of surface amphoteric hydroxyl groups at the water/oxide interface [17].

SCMs are based on a microscopic approach of solid–liquid interactions during retention

Solid phase

Magnetite (Fe3O4) was a commercial powder purchased from Alfa Aesar. The surface acidity constants of this solid has been determined by fitting titration and sorption experiments using the Fiteql 4.0 [18] code in previous works [13], [19], [20].

Specific surface area of magnetite was determined by applying the Brunauer–Emmet–Teller (BET) equation with nitrogen adsorption isotherms at 77 K. Results are given in Table 2.

Magnetite was also characterized by X-ray diffraction (XRD). Results showed

Experiments as a function of time

The amount of adsorbed selenium(IV) (Fig. 2) onto magnetite depends on the stirring time, reaching a plateau after 30 h. Thus, this stirring time has been chosen for all further sorption experiments on magnetite.

Sorption edges

Sorption of selenium(IV) onto magnetite decreases when the pH increases [8], as it was expected for anions (Fig. 3). Concerning silicic acid, experimental data show that silicic acid can bind on the surface of magnetite over a wide pH range (Fig. 4). This differs from a classical

Conclusion

This work shows that selenium(IV) and silicic acid can sorb onto the surface of magnetite (Fe3O4). As expected for anions, sorption of selenium(IV) decreases in a sharp pH range, while sorption of silicic acid occurs over a wide pH range, with a bell-shaped curve characteristic of the sorption of weak acids like H4SiO4.

Those experimental observations have been fit thanks to surface complexation model, by using one surface complex triple bondFeSeO3 for selenium(IV) and two surface complexes triple bondFeH3SiO4 and triple bond

Acknowledgments

The authors thank Andra for its financial support, as well as Mr. Jean-Marie Couret, from the University of Nice-Sophia-Antipolis for his help during ICP-AES measurement and also Dr. Mustapha Abdelmoula, from LCPME (Nancy) for Mössbauer measurements.

References (28)

  • L.S. Balistrieri et al.

    Geochim. Cosmochim. Acta

    (1990)
  • K.A. Boult et al.

    J. Contam. Hydrol.

    (1998)
  • M. Martinez et al.

    Appl. Surf. Sci.

    (2006)
  • N. Xu et al.

    Chemosphere

    (2006)
  • N. Xu et al.

    Chemosphere

    (2006)
  • K.-H. Goh et al.

    Chemosphere

    (2004)
  • F. Monteil-Rivera et al.

    J. Colloid Interface Sci.

    (2000)
  • N. Jordan et al.

    J. Colloid Interface Sci.

    (2007)
  • P.J. Swedlund et al.

    Water Res.

    (1999)
  • L. Balistrieri et al.

    Geochim. Cosmochim. Acta

    (1990)
  • N. Marmier et al.

    J. Colloid Interface Sci.

    (1999)
  • N. Marmier et al.

    J. Colloid Interface Sci.

    (2000)
  • L. Sigg et al.

    Colloids Surf.

    (1981)
  • V. Philippini et al.

    J. Nucl. Mater.

    (2006)
  • Cited by (53)

    • Mechanisms of selenium removal by partially oxidized magnetite nanoparticles for wastewater remediation

      2021, Applied Geochemistry
      Citation Excerpt :

      Although actual PZC values generally depend on numerous factors such as particle size, ionic strength, adsorption of ions, or the presence of coatings, the reported PZC values by Hu et al. (2010), Kosmulski (2020), and Milonjić et al. (1983) for both magnetite and maghemite range maximally to ~pH 8.1). Responsible for this immobilization is the specific adsorption of selenite resulting in the formation of inner-sphere surface complexes (Jordan et al., 2009; Kim et al., 2012; Loyo et al., 2008; Martínez et al., 2006). Due to this inner-sphere selenite adsorption, large distribution coefficients as well as high Se surface coverages of up to 2.5 at/nm2 can be observed even at pH 9.2.

    View all citing articles on Scopus
    View full text