FIGURE 12.17 Vertical O, profile before (August 23) and after (October 12) development of the ozone hole at the U.S. Amundsen-Scott Station, South Pole, in 1993 (adapted from Hofmann et al., 1994a).

loss (Nardi et al., 1997). It is noteworthy that loss of 03 in the isolated interior regions of the polar vortex may have been maximized even by the mid-1980s (Jiang et al., 1996).

There are several reasons for the dramatic ozone destruction (see Fig. 2.17): low temperatures may have prolonged the presence of polar stratospheric clouds, which play a key role in the ozone destruction, the polar vortex was very stable, there were increased sulfate aerosols from the 1991 Mount Pinatubo volcanic eruption, which also contribute to heterogeneous chemistry, and chlorine levels had continued to increase. These issues are treated in more detail shortly.

Soon after the discovery of the hole, a number of different theories to explain this remarkable observation were advanced (Solomon, 1988). These included atmospheric dynamics involving vertical advection, which introduces upper tropospheric air with its lower ozone levels (e.g., see Tung et al., 1986; Mahlman and Fels, 1986; Shiotani and Gille, 1987; Tung and Yang, 1988; and Rosenfield et al., 1988), and solar proton events that produce more oxides of nitrogen and destroy 03 through the cycles discussed earlier for the HSCT (Callis and Natarajan, 1986; Stephenson and Scourfield, 1991; Shumilov et al., 1992). However, if the first explanation were correct, other long-lived trace gases produced in the troposphere such as N20 should be enhanced; this was found not to be the case. In fact, stratospheric N20 mixing ratios decline rapidly as the South Pole is approached (e.g., see Schoeberl and Hartmann, 1991; and Tuck, 1989).

Increased production of oxides of nitrogen through solar proton events associated with the 11-year cycle in solar activity would be expected to be most important in the upper stratosphere, above the region where the majority of the ozone depletion was observed; in addition, lower, rather than higher, concentrations of gasphase oxides of nitrogen appear to be associated with the ozone depletion (e.g., see Noxon, 1978; McKenzie and Johnston, 1984; Thomas et al., 1988; Keys and Gardiner, 1991; and Solomon and Keys, 1992). Hence both of these explanations are not consistent with atmospheric observations.

Through a variety of studies, it is now generally accepted that the observed losses are associated with chlorine derived from CFCs and that heterogeneous chemistry on polar stratospheric clouds plays a major role. The chemistry in this region is the result of the unique meteorology. As described in detail by Schoeberl and Hartmann (1991) and Schoeberl et al. (1992), a polar vortex develops in the stratosphere during the winter over Antarctica. The air in this vortex remains relatively isolated from the rest of the stratosphere, allowing photochemically active products to build up during the polar winter and setting the stage for the rapid destruction of ozone when the sun comes up and the polar vortex dissipates, apparently from the top down (Bevilacqua et al., 1995). There is, however, some exchange of the air mass, especially that in the vortex edges, with regions outside the vortex (e.g., see Wauben et al., 1997a, 1997b). Whether it is better described as a "flow reactor" rather than a "containment vessel" remains somewhat controversial (Mclntyre, 1995; Tuck and Proffitt, 1997; Wauben et al., 1997a, f 997b).

A second critical component of the meteorology in this region is that the stratospheric temperatures during the winter can be very low. As the sunlight decreases in the fall, radiative cooling of air in the upper stratosphere occurs, which causes sinking of this polar air mass. Adiabatic heating occurs as the air sinks, which partially offsets the temperature decrease due to radiative cooling. Radiative equilibrium is reached at altitudes below about 30 km, decreasing the descent of the air mass. The cooling takes the air temperature to values as low as ~ 185 K. The temperature difference between the polar region and midlatitudes leads to a vortex with wind speeds at its edge of about fOO m s"1—hence the designation "polar vortex."

At these low temperatures, even the relatively small amount of water in the stratosphere (about 5-6 ppm at the beginning of winter, dropping to 2-3 ppm when dehydration occurs during July and August) forms ice crystals. In addition, at slightly higher temperatures, crystalline nitric acid trihydrate (NAT) also forms, which was initially thought to represent one type of PSC. As discussed in more detail shortly, the formation of polar stratospheric clouds (PSCs) is quite complex and is now believed to involve ternary solutions of HN03, H2S04, and water as well. PSCs can have a quite remarkable appearance, with various colors depending on their altitude and the presence of clouds (Sarkissian et al., 1991). As discussed in detail later, they play a critical role in ozone depletion by providing surfaces for heterogeneous chemistry.

There are a number of factors that determine the amount of ozone destruction over Antarctica each year. Clearly, the concentrations of chlorine and bromine are major determinants, and these have increased from about 1.1 ppb in the 1960s to 2.4 to 3.2 ppb over the 1986-1995 decade (WMO, 1995). Lower temperatures lead to more polar stratospheric clouds and more heterogeneous chemistry. As discussed later, it has also been increasingly recognized that lower temperatures have a significant effect on heterogeneous chemistry by increasing the solubility of HC1 in liquid aerosol particles. Aerosol concentrations are important because they serve as hosts for heterogeneous chemistry as well as assisting in the formation of polar stratospheric clouds

(e.g., see Hofmann and Oltmans, 1993; Deshler et al., 1996; Portmann et al., 1996; and Molina et al., 1996). Finally, there appears to be an association between the depth of the Antarctic ozone hole and quasi-biennial oscillations, QBO (e.g., see Bojkov, 1986; Garcia and Solomon, 1987; and Angell, 1993b), which may be associated with temperature changes due to enhanced transport from the tropics to the poles in the months preceding development of the ozone hole.

As discussed earlier, the destruction of 03 by chlorine in midlatitudes is controlled by tying up chlorine atoms as HC1 via its reaction (33) with CH4 or by tying up CIO in the form of chlorine nitrate, C10N02, via its reaction (16) with NOz. These temporary chlorine reservoirs only slowly regenerate atomic chlorine so they are important in determining how much ozone destruction occurs. The reaction of C10N02 with HC1 in the gas phase is slow, with a rate constant of less than ~1 x 10"20 cm3 molecule"1 s"1 (DeMore et al., 1997). However, as first shown in the mid-1980s (Molina et al., 1987; Tolbert et al., 1987, 1988a; Leu, 1988a), it proceeds quite rapidly on the surfaces of ice found in the stratosphere, generating Cl2 and HN03:



(Of course, this reaction does not involve the simultaneous collision of HC1 and C10N02 at the particle surface; HC1 is taken up by the particle and C10N02 subsequently collides and reacts with it.) Because of the very "sticky" nature of nitric acid, it stays on the ice or in the solution. This has the added effect of removing oxides of nitrogen from the gas phase, which then frees up additional CIO that might otherwise be tied up in the form of chlorine nitrate. As discussed in more detail shortly, it is now known that stratospheric particles may not only be solids, but under some conditions liquid solutions containing mixtures of H2S04 and water or ternary solutions of HN03 with H2S04 and water. However, reactions such as (39) can occur not only at the surfaces of these liquid particles but in the bulk as well. Indeed, as we shall see, some of these reactions are much faster in and on liquid particles than on ice.

Farman and co-workers (1985) suggested that the reaction between HC1 and C10N02 may play a key role if it were fast enough, which at the time did not seem to be the case for the gas-phase reaction. Subsequently, Solomon et al. (1986) proposed that enhancement of this reaction on the ice surfaces of polar stratospheric clouds could explain the development of the ozone hole via the production of Cl2 during the winter. It is also consistent with sequestering oxides of nitrogen in the form of HN03 on the ice surface. It has been suggested that similar chemistry may occur on cirrus clouds near the tropopause (Borrmann et al., 1996, 1997a; Solomon et al., 1997), where a significant amount of HC1 can be taken up by ice particles at equilibrium (Thibert and Dominé, 1997). Whether sufficient cloud surface area and inorganic chlorine compounds coexist in the same region to cause this chemistry is not clear.

Similarly, the reaction of HC1 with N2Os is slow in the gas phase, but was shown in the late 1980s to occur rapidly on ice surfaces or in the solutions found in stratospheric particles (Tolbert et al., 1988b; Leu, 1988b):



As a result of the enhancement of reactions (39) and (40) in or on surfaces provided by PSCs, HC1 and C10N02, which normally act as reservoirs, are converted over the winter into the photochemically active Cl 2 and nitryl chloride, C1N02. When the sun comes up in the spring, these species are rapidly photolyzed to generate chlorine atoms, setting off the chain destruction of ozone. However, in the case of C1N02 from reaction (40), N02 is generated simultaneously; this gaseous N02 can then sequester chlorine in the form of C10N02. Hence the heterogeneous reaction (39) is much more important. The nature of PSCs and heterogeneous reactions on PSCs and aerosol surfaces are discussed in more detail in the following section.

While reaction (39) shows the overall reaction that occurs between the two chlorine reservoirs, it has been proposed that it may actually occur in several steps (e.g., Hanson and Ravishankara, 1991, 1993a; Abbatt et al., 1992). Thus C10N02 has been shown to hydrolyze on the surfaces of solid and liquid particles, generating HOC1:


HOC1 can then react with HC1 on or in the particle, generating Cl2 (Prather, 1992):


The net effect of these two reactions is that shown as reaction (39).

However, Oppliger et al. (1997) suggest that while reaction (39) of C10N02 with HC1 on ice proceeds through a direct mechanism, the hydrolysis reaction (41) does not. For example, while C10N02 is rapidly taken up by ice at 180 K, the formation of HOC1 is delayed (see also Hanson and Ravishankara, 1992b). Rossi and co-workers (Oppliger et al., 1997) propose that C10N02 forms an intermediate (H2OCl+ ••• NO/) that subsequently releases HOC1 to the gas phase, while generating hydrated HN03. Infrared studies by Sodeau and co-workers (Sodeau et al., 1995; Koch et al., 1997; Horn et al., 1998) support the formation of [H2OCl]+ under conditions of low water availability, with this intermediate ultimately reacting with water to generate HOC1. Ab Initio calculations, however, suggest that the hydrolysis occurs by a concerted nucleophilic attack of an oxygen from a water (ice) molecule on the chloride of C10N02, simultaneously with a proton transfer from the attacking water to the ice (Bianco and Hynes, 1998).

There are also important differences in the gas-phase chemistry of the Antarctic ozone hole compared to the chemistry at midlatitudes. One is the formation and photolysis of the CIO dimer. In the Antarctic spring, recycling of CIO back to chlorine atoms via reaction (27) with oxygen atoms does not play a major role because of the relatively small oxygen atom concentrations at the low UV levels at that time. Molina and Molina (1987) proposed that the formation of a dimer of CIO could, however, lead to regeneration of atomic chlorine through the following reactions:

Was this article helpful?

0 0

Post a comment