Oxidation of arsenite to arsenate on birnessite in the presence of light

Mechanistic aspects of As(III)/birnessite photochemistry

Furthermore, studies have generally studied the production of Mn(II) photoproduct during the irradiation of manganese oxides in the presence of electron donors (e.g., organic species) that are oxidized by the photogenerated ( {text{h}}_{text{vb}}^{ + } ). Interestingly, recent studies have also shown that the photogeneration of Mn(II) during the irradiation of birnessite can occur in an ice matrix in the absence of an electron donor (other than potentially water) [58], albeit at a much lower rate than if an electron donor was present. Recent studies investigated the photochemistry of birnessite with time-resolved XAS during the irradiation of the material with 400 nm light in water [59]. This particular study showed that the photogenerated electrons resulted in the reduction of Mn(IV) to Mn(III) that migrated into the interlayer region of the layered birnessite. It was speculated in this study that the oxidative hole could lead to the generation of reactive oxygen species such as hydroxyl radical, but their potential reaction to form H₂O₂ would only lead to the oxidation of Mn(III) back to Mn(IV) [59].

In the present study we suspect that the oxidative hole formed in the valence band of birnessite during irradiation was directly responsible for the oxidation of As(III). In short, our results do not give support to a scenario where ROS formed during the irradiation of birnessite in the presence or absence of As(III). Experiments were carried out that used both fluorescent probes and EPR (coupled with trapping agents) to investigate the production of ROS. To investigate the generation of hydroxyl radical, a fluorescence method using coumarin was employed [44, 45]. The reaction of coumarin with hydroxyl radical forms a fluorescent adduct with a unique emission. Data obtained using this method in situ (Additional file 1: Figure S8) did not show evidence for the presence of the adduct. We also used the APF-HRP test (see experimental) to detect hydrogen peroxide in solution, but also found no evidence for this species (Additional file 1: Figure S9). We point out, however, that if H₂O₂ was produced it might be expected to rapidly decompose in the presence of birnessite [60–62].

In addition to fluorescent-based probes, EPR experiments were carried out to further investigate the possibility of ROS generation. These particular experiments using DMPO as a spin-trapping agent for hydroxyl radical did not yield any support for the generation of this particular radical. Whether the experiment was carried out on an aqueous suspension of birnessite or suspension of birnessite in the presence of As(III), the resulting spectra could be associated with the characteristic EPR spectrum for Mn(II) (Additional file 1: Figures S10, S11). Consistent with our batch studies the magnitude of the Mn(II) spectral weight from the EPR experiment was greater when As(III) was present, compared to the irradiation of birnessite in As(III)-free water (Additional file 1: Figure S11). We attribute the increased Mn(II) signal to the presence of the electron donor (i.e., As(III)) that can be oxidized by the valence hole.

To better determine whether As(III) oxidation occurred in part due to the presence of valence band holes in the presence of light, we carried out experiments that utilized mannitol. Prior studies have shown that mannitol is an efficient scavenger of OH· and oxidative holes [46]. Figure 5 shows data from an experiment where 20 mM mannitol was added to particular reaction mixtures. It is mentioned that both the coumarin-based analytical technique and EPR studies strongly suggest that solution OH· is not an important intermediate species when birnessite was irradiated in the absence or presence of As(III). Hence, we attribute mannitol-induced changes in the rate of production of As(V) in the presence of As(III), birnessite, and light to the scavenging of oxidative holes (or OH· formed through the oxidation of surface hydroxyl groups). It should be noted that the data show that the addition of mannitol, even at the relatively high concentration of 20 mM, does not affect As(V) product formation in the dark. We infer from this result that mannitol does not block birnessite sites that can oxidize As(III) to As(V) in the absence of light. In the photo-reaction data, the presence of mannitol results in an As(V) concentration that is approximately 30 % lower relative to the mannitol-free system after 7 h. We infer from this result that mannitol is scavenging photogenerated holes (that oxidize mannitol) that would otherwise oxidize As(III).

Effect of mannitol on the oxidation of arsenite by Na-birnessite: concentration of aqueous As(V) released into the solution during the oxidation of As(III) in the presence of Na-birnessite at pH 5 (oxic) under light and dark conditions with and without mannitol

If such a mechanism were operative it might be expected to exhibit rapid kinetics, since prior studies have shown that the reaction of As(IV) with dissolved O₂ is a rapid reaction that leads to As(V) [65]. We, however, do not observe a dependence of the photochemical As(V) production rate on the presence of dissolved oxygen. It is conceivable that a species such as H₂O₂ could play a role in the oxidation of As(III) to As(V). If a short lived H₂O₂ ROS does play a role, it would be expected to form from chemistry initiated by the valence band hole and the oxidation of surface hydroxide [59]. Formation pathways via the reduction of dissolved oxygen do not appear to be a possibility, because of the aforementioned insensitivity of photochemical As(III) oxidation to dissolved oxygen. Whether this pathway is inhibited due to the reduction potential of oxygen lying higher than the conduction band minimum of birnessite, or to the efficiency of Mn(IV) [or Mn(III)] as an electron acceptor (to form Mn(III) and Mn(II), respectively) cannot be discerned from our results.

It is interesting that the photochemical oxidation of As(III) still occurs with a significant rate even when As(III) oxidation in the dark has decreased significantly after 20 h of reaction (Fig. 2). Most prior studies have shown for the dark reaction between As(III) and birnessite that the surface becomes covered with Mn(III) [23, 25–27, 52, 54]. It is conceivable that the irradiation of such a surface might not directly oxidize As(III) through valence band hole formation, but instead the oxidation of a fraction of Mn(III) species back to Mn(IV) by the hole could occur. As(III) oxidation on such newly created Mn(IV) sites would be expected to occur with faster kinetics than on Mn(III) sites [51]. The exact role of these species in the photochemistry of birnessite has significant implications in environmental manganese cycling. Understanding the mechanistic aspects of this chemistry warrants future study.