Arsenate and selenate chemistry

Homogeneous Reactions of As and Se Oxoanions in Aqueous Solutions, and the Photooxidation of their Reduced Species in the X-ray Beam

Chemical speciation is one of the principal components required for understanding contaminant toxicity, transport (diffusion, adsorption), redox transformations (electron-transfer kinetics), and bioavailability in the environment. The concentration of ions and their complexes in natural samples can be predicted from standard state free energies. However, these are not available for many aqueous and adsorbed species. Further, most stability constants have been measured with macroscopic methods, and hence the coordination chemistry of several reported species is questionable. Although researchers have obtained molecular-scale information using different spectroscopic methods, systematic studies of the chemistry of various species in aqueous solutions and at solid-aqueous solution interfaces is lacking. In the case of oxoanions, both thermodynamic and molecular information is incomplete. The objective of our research is to evaluate the homogeneous and heterogeneous reactions of selected oxoanions and their cation complexes as influenced by pH, ionic strength, and substrate chemistry.

The oxoanions of As and Se were selected for this investigation since they are common pollutants, and are reported to form various types of complexes with mineral surfaces. Previous macroscopic and spectroscopic studies have indicated that SeO42- forms dominantly outer-sphere (OS) complexes and SeO32- and AsO43- form dominantly inner-sphere (IS) complexes with ferric hydroxide surfaces. Recently contradictory views have been expressed concerning the types of surface complexes that these oxoanions form. While previous research has focused on the adsorption of As and Se oxoanions onto several mineral oxides, little information is available on their homogeneous reactions in solution. Furthermore, a clear understanding of the chemical conditions that promote inner-, and/or outer-sphere complexation of oxoanions in aqueous systems is lacking. To address these issues, we have focused on coordination chemistry of these oxoanions in the presence of different cations in aqueous solution. In addition, we examined several factors that can complicate XAS studies of As and Se oxoanions in solution, including the presence of common impurities in chemical reagents used to prepare solution samples and the photooxidation of reduced species in the synchrotron x-ray beam.

Experimental Details

Dilute aqueous solutions (0.1-7 mM) of arsenate, arsenite, selenate, and selenite were prepared using their Na-salts. The selenate stock solutions were reacted with H2O2 at 65-70°C for ~ 3 hours to oxidize selenite impurities that are commonly present in all Na-selenate reagents. After treatment, these samples contained less than 0.12 mM selenite in a solution containing 25.3 mM selenate. All other reagents were used as supplied. When necessary, the pH of the aqueous solutions was adjusted with NaOH, HCl, HNO3, and/or H2SO4; and the desired ionic strength was achieved by addition of NaCl, KCl, and/or CaCl2. Oxoanion-cation complexes were prepared by mixing stock solutions of a selected oxoanion with freshly mixed solutions containing individual cations (as metal chloride salts). Additionally, both aged and fresh Fe3+ solutions were used to examine the nature of oxoanion-ferric complexes. For XAFS measurements, solution samples were placed in polypropylene straws whose ends were closed with non-sticky Teflon tape. Although some of the initial studies were done using special liquid cells made with polypropylene sheets (0.4 x 4 x 4 cm) and sticky Kapton tape (to contain the solutions), we found that the pH of the enclosed solutions dropped by as much as 2.5 pH units in 5 h. The glue also came off of the Kapton tape when alkaline solutions were used in the holder. Sample holders made with polypropylene straws, and non-sticky Teflon tape worked better with no detectable chemical changes in aqueous solutions.


Figure 1: XANES spectra of AsO43- at different pH values. The spectra were calibrated against the absorption edge of As(0) at 11867 eV.
Figure 1: XANES spectra of AsO43- at different pH values. The spectra were calibrated against the absorption edge of As(0) at 11867 eV.



Species pKa
H3AsO40 2.24, 6.86, 11.49
H3AsO30 9.2, 12.1, 13.4
H2SeO40 1.9
H2SeO30 2.46, 8.04
Table 1: The pKa values of oxoanions in dilute aqueous solutions [1,2]. These values decrease with increases in ionic strength.


Both XANES and XAFS spectra (at the As, Se, Fe, Ca absorption edges) of aqueous solutions were obtained on the SSRL beamlines 2-3, 4-1, and 9-3, using a Si(220) double crystal mono-chromator. The slit openings were 2.0 x 14 mm (excepting for experiments on BL 9-3) upstream and downstream of the monochromator. On BL 9-3, the slit openings were 1.5 x 2.0 mm. All of the XAFS spectra were collected in fluorescence-yield mode using a Lytle and/or 13-element Ge array detector. More than 10 spectra (collection time of 0.5 h each), with a 0.2 eV step size at the absorption edge, were collected for each sample. All of the measurements were made under ambient conditions. The XAFS spectra of arsenite and selenite were collected with the sample chamber under continuous He-purge to lessen the oxidation of these ions.

Influence of pH on Oxoanion Coordination

The solution pH is a master variable that influences the complexation of oxoanions with cations and mineral surfaces [1,2]. As oxoanions protonate at low pH, the X-O (where X = As, Se) bond length increases with modifications in anion symmetry and hydration. As a result, the bond valence of the terminal O changes, affecting the oxoanion hydration and complexation with cations. As the solution pH crossed the oxoanion pKa values, the XANES and XAFS spectral features of all oxoanions change significantly, without any detectable changes in the energy positions of the absorption edges (Table.1, e.g. for AsO43-, Fig. 1). The XANES spectral changes are more intense in the case of selenite and arsenite than their oxidized species. The changes in the spectral features are primarily caused by the changes in the multiple scattering of the photoelectron within the oxoanion tetrahedron (due to symmetry and bond length changes associated with protonation).

Fig 2 Aqueous-oxoanion

Figure 2: XAFS spectra and their Fourier transforms of different aqueous AsO43-species under different pH conditions.


The XAFS spectra of oxoanions also exhibit significant spectral changes with pH, and these are similar for all of the oxoanions examined. In the case of AsO43-, the amplitude of XAFS oscillations decreased with protonation (Fig. 2). The XAFS data fits (and the Fourier transforms) indicate that the As-O bond length does not change significantly with protonation (avg. As-O = 1.69 Å). The coordination number for the first shell 'O' also remains very close to 4.0 (£ ± 0.1). However, the Debye-Waller (DW) value for this shell increased from ~ 0.0001 Å2 for AsO43- to as high as 0.003 Å2 for the protonated forms. The first-shell As-O XAFS contributions of the protonated arsenate could not be fit with two different As-O and As-OH shells. This may be due to the small bond length differences between the As-O and As-OH (e.g. in solids As-O ~ 1.68 Å, As-OH ~ 1.75 Å [5]).

Interestingly, weak features are found at 3.5-4.0 Å (uncorrected for phase shift, Fig. 2) in the Fourier transforms of all AsO43- species. These features are also reproducible in the spectra collected during different periods of beam time. The multiple-scattering paths that can contribute to these distances are weak or absent in the arsenate tetrahedron. Hence, these features are tentatively attributed to the second-shell O of the oxoanion solvated water molecules. Theoretical analysis with arsenate-water clusters are in progress to examine the scattering contributions of solvated water molecules. Although, variations in ionic strength have been shown to modify oxoanion complexation reactions [1-3], increases in ionic strength (< 2 molcL-1) did not induce detectable changes in As-O bond lengths or DW values for the arsenate solutions examined in this study.

Metal Complexation With Oxoanions

Diffraction and vibrational spectroscopic studies on oxoanion-cation complexes in solids have shown that cations cause less modifications in oxoanion bond lengths and symmetries, than do H atoms (e.g. for As see ref: [6]). However, aqueous cation-oxoanion interactions can disturb the oxoanion solvation shell (or vice versa), and the bond valence on the terminal 'O' atoms of oxoanion. The differences in the size, charge, polarizability, and concentration of cations dictate the types of complexes formed by the oxoanion-cation pairs. To understand these effects in more detail, we have studied arsenate, and selenate complexation with several different mono-, di-, and tri-valent cations (Li+, Na+, K+, Rb+, Cs+, Mg2+, Ca2+, Ba2+, Fe3+, Al3+). In addition, complexation of arsenate and selenate is examined for the cations common to soils and aquatic systems (K+, Ca2+, Fe3+, Al3+) at a range of pH (2-12), ionic strength (as high as 2 molcL-1) and cation-to-anion ratios.

Excepting for the complexes of Fe3+, Al3+, and Ca2+, the XAFS spectra of all other cation-arsenate, and cation-selenate solution complexes did not show evidence of backscattering from the cations. In addition, the As-O and Se-O bond distances and their DW values of the majority of the cation-oxoanion complexes (excepting for the complexes of Mg2+) did not change when metal cations were added to oxoanion solutions. In the case of Mg2+-arsenate complexes, the average As-O bond distance (1.69 to 1.68 Å), and the DW values (0.003 to 0.0017 Å-1) decreased, which may suggest that Mg2+ ions have a stronger interaction with arsenate than the other cations. Our results also indicated that the coordination of Fe3+-oxoanion complexes changed with time. These changes were distinct in the case of selenate, which formed inner-sphere complexes with freshly prepared Fe3+solutions at pH < 4.0. The XAFS spectra and their Fourier transforms indicate strong backscattering from the neighboring Fe3+ atoms around 3.0 Å (uncorrected for phase shift). Over a 4 h period, the intensity of this feature in the Fourier transforms decreased, which suggests that the selenate-Fe3+ coordination changed from an inner-sphere complex to an outer-sphere/uncomplexed species. Changes in Fe3+ polymerization may have caused these reactions. Currently, we are evaluating the kinetics of these transformations and the factors that control them. In summary, these results suggest that the oxoanion-cation complexes may exist as ion-pairs in solution rather than as directly coordinated complexes in the case of the majority of cations.

Photooxidation of Arsenite and Selenite in the Beam

Reduced species of As and Se are prone to oxidation by atmospheric O2, but at a very slow rate. Our recent XAFS studies indicate that these species are oxidized at a much faster rate in the presence of the synchrotron X-ray beam and oxygen. The rate of oxidation was also faster on the high flux beamline 9-3 (flux: 1012-1013 photons sec-1) when compared to the low flux bend magnet beamline 2-3 (flux: 1010 photons sec-1). Because the absorption edges of As and Se in different oxidation states are separated by several eV, their redox reactions can be determined from their XANES spectra. Similar results were also reported for the heterogeneous oxidation of arsenite [7].

For example, freshly prepared 1 mM arsenite solution was oxidized completely within 1.5 h under ambient conditions on BL 9-3, while it took several hours on beamlines 4-1 and 2-3. Continuous purging of the sample compartment with He reduced the rate of arsenite oxidation (Fig. 3, first panel on lower left-hand side). The beam energy (below, at, or above the arsenite absorption edge) did not affect the rate of arsenite oxidation (Fig. 3). When this sample was shaken between spectral scans, the As(III)/As(V) ratio of the subsequent scan was higher than that collected before mixing (see the scans marked in the last two panels). This finding suggests that arsenite oxidation occurred only in the sample area that was exposed to the beam, and this effect may have been exacerbated by the poor mixing of the portion of the sample that was intersected by the beam with the portion of solution that wasn't. Oxidation was not detectable in the absence of beam. The rate of arsenite oxidation was also dependent on the solution pH, with the rate faster in alkaline solutions than the acidic conditions.

These studies indicate that oxygen and x-ray photons were required for the oxidation observed. We also found that selenite oxidation was slower than that of arsenite, and was not detectable at low flux beamlines (e.g. on beamline 2-3). Currently, we are evaluating these reactions in detail. These studies demonstrate that care should be taken while examining redox-sensitive multivalent ions (especially at high flux beamlines, and at spectromicroscopy facilities).

Fig 3 Aqueous-oxoanion
Figure 3: Photooxidation of AsO33- in the presence of the synchrotron X-ray beam (beamline 9-3). The sample chamber was purged with He continuously except during the last hour of the experiment.


  1. Sposito G. (1989) The Chemistry of Soils, Oxford IBH.
  2. Morel F. and Hering J. G. (1993) Principles and Applications of Aquatic Chemistry. John Wiley & Sons.
  3. Parfitt R. L. (1987) Adv. Agron30, 1-50.
  4. Hayes K.F. et al. (1987) Science 238, 783-786.
  5. Manceau A. and Charlet L. (1994) J. Coll. Int. Sci. 168, 87-93; Myneni S.C.B. (1995) Ph.D. Dissertation, Ohio State University. Columbus, OH.
  6. Myneni S.C.B. et al. (1998) Geochim. Cosmochim. Acta. 62, 3285-3300.
  7. Foster A.L. et al. (1998) Environ. Sci. Technol32