It is difficult to evaluate quantitatively the importance of such heterogeneous reactions in the overall oxidation of S(IV). Their rates depend on the physical and chemical natures of the surfaces involved, including specific surface areas, the presence of defects and surface adsorbed water, etc., yet these are not well understood, especially for highly complex environmental gas-liquid-solid systems. For example, the rates of oxidation of S02 at 80% relative humidity on two different samples of fly ash obtained from two coal-fired power plants differed by more than an order of magnitude (Dlugi and Gusten, 1983). Even in laboratory systems the nature of relatively simple surfaces such as carbon depends on the history of the material.
However, the available evidence suggests that such heterogeneous reactions should be considered as potential contributors to the overall oxidation of S(IV), especially close to sources where particle concentrations and hence available surface areas are relatively high (see Chapter 9). Baldwin (1982), for example, estimates that the loss of gaseous S02 due to interactions with particle surfaces at a particle density of 100 fxg m~3 could be as high as 1% h-1. Such rates may not be sustained for long periods of time due to saturation of the surface, if the surface is dry. However, Chang and Novakov (1983) point out that when the surface is wet, the active sites are constantly regenerated as the sulfate formed on the surface dissolves in the surrounding aqueous solution; given the availability of water in the atmosphere, it seems unlikely that any surfaces are really "dry."
Herring et al. (1996) have reported evidence for the heterogeneous oxidation of both S02 and NOx on the surfaces of soil dust particles in the smoke plume from the 1991 Kuwait oil fires. The rate of S02 oxidation was estimated to be 6.5% h~'.
In addition to the dark oxidation of S(IV) on surfaces, there may be photochemically induced processes as well. For example, irradiation of aqueous suspensions of solid a-Fe203 (hematite) containing S(IV) with light of A > 295 nm resulted in the production of Fe(II) in solution (Faust and Hoffmann, 1986; Faust et al., 1989; Hoffmann et al., 1995). This reductive dissolution of the hematite has been attributed to the absorption of light by surface Fe(III)-S(IV) complexes, which leads to the generation of electron-hole pairs, followed by an electron transfer in which the adsorbed S(IV) is oxidized to the SO^' radical anion. This initiates the free radical chemistry described earlier.
The photochemical reactions of such semiconductors in aqueous solutions may also influence S(IV) oxidation via the production of H202. The possibility of such photoassisted surface reactions in the atmosphere was first examined by Calvert in 1956. For example, Fe203 particles suspended in a bisulfite solution in the presence of 02 rapidly oxidize the S(IV) to S(VI) in the presence of light, whereas no oxidation occurs in the dark (Frank and Bard, 1977). It has been suggested that this is due to the absorption of light by Fe203 and the migration of electrons in the conduction band to the particle surface where they react with 02 and H+ to form H202:
Similar reactions have been observed more recently on a variety of solids, including ZnO, Ti02, and desert sand; in the presence of organics, organic peroxides have also been observed (e.g., see Kormann et al., 1988). For a review of this area, the reader is referred to the review by Hoffmann et al. (f 995).
5. Relative Importance of Various Oxidation Pathways for SOz
Figure 8.2f shows one estimate of the relative importance of the oxidation of S(IV) by 03 and H202, the Fe- and Mn-catalyzed 02 oxidation, and the oxidation on a carbon surface at concentrations typically found in the atmosphere (Martin, f984; Martin et al.,
1991). It was assumed that there were no limitations on the rates of oxidation due to mass transport; as discussed in detail by Schwartz and Freiberg (1981), this assumption is justified except for very large droplets (>10 ¡xm) and high pollutant concentrations (e.g., 03 at 0.5 ppm) where the aqueous-phase reactions are very fast. It was also assumed that the aqueous phase present in the atmosphere was a cloud with a liquid water content (V) of f g m 3 of air. As seen earlier, the latter factor is important in the aqueous-phase rates of conversion of S(IV); thus the actual concentrations of iron, manganese, and so on in the liquid phase and hence the kinetics of the reactions depend on the liquid water content.
Only the oxidation by H202 is relatively independent of pH. This arises because the rate coefficient for the reaction and the solubility of S(IV) show opposite trends with pH (Fig. 8.9b). For the other species, the effects of the S(tV) solubility and the pH dependence of the kinetics work in the same direction (Fig. 8.9a), leading to a strong overall dependence on pH. The uncatalyzed oxidation of 02 is not shown, because it is generally believed to be unimportant compared to the other mechanisms in real atmospheric droplets containing "impurities" such as metals that will act as catalysts.
The estimates in Fig. 8.21 show that H202 is expected to be the most important oxidant for S(IV) in clouds and fogs at pH <4.5. At higher pH values, both 03 and the iron-catalyzed 02 oxidation can compete.
Figure 8.22 shows an estimate of the contributions to the oxidation of S(tV) by H202, by the iron-catalyzed 02 oxidation, and by OH in both the gas and aqueous phases of a cloud (Jacob et al., 1989). ft is seen that H202 and the iron-catalyzed process predominate at night, but the gas-phase oxidation by OH becomes significant during the day when it is formed by photochemical processes. On the other hand, the contribu
Was this article helpful?