120 160 200 240 280 320 360 Wavelength (nm)

FIGURE 12.15 Absorption cross sections for 02 and 03 from 120 to 360 nm, showing the window from ~ 185 to 210 nm (adapted from Rowland and Molina, 1975).

peroxy radicals. Li et al. (1995b) attribute this to the antibonding interactions between the 3pz orbitals on fluorine and oxygen.

In short, the net effect of fluorine atom chemistry on ozone destruction is very small, 103-104 times smaller than the effect of chlorine on a per-atom basis (Sehested et al., 1994; Ravishankara et al., 1994; Li et al., 1995b; Lary, 1997).

More recently, the possible contribution of organo-fluorine free radicals to ozone destruction has been considered with respect to the introduction of CFC transitional alternates and longer term replacements; however, as discussed in Chapter 13.D, this also does not appear to be important in terms of ozone destruction.

3. Gas-Phase Chemistry in the Stratosphere

The CI atom released by photolysis of the CFCs reacts in a catalytic chain reaction that leads to the destruction of 03:

Because the concentration of oxygen atoms increases with altitude, the reaction cycle represented by (26) and (27) is important primarily in the middle and upper stratosphere (e.g., Garcia and Solomon, 1994; WMO, 1995). For the lower stratosphere, however, it is only responsible for about 5% of the portion of the total ozone loss that is due to halogens at 15 km and ~25% at 21 km (see Fig. 12.8; Wennberg et al., 1994). Most of the 03 loss associated with CIO, and BrO, at the relatively low altitudes in Fig. 12.8 is due to the following cycle (Solomon et al., 1986; Crutzen and Arnold, 1986):

This cycle accounts for ~ 30% of the ozone loss due to halogens in the lower stratosphere, and the corresponding cycle for bromine for ~ 20-30% (Wennberg et al., 1994). Reaction of CIO with H02, reaction (28), produces HOC1 + 02 with a yield >95% at temperatures from 210 to 300 K; however, at the lowest end of this temperature range, there is evidence for the produc tion of HC1 + 03, with a yield of <5%, which may be significant in the net ozone destruction both in the polar regions and in midlatitudes (Finkbeiner et al., 1995).

The reaction of OH with CIO,

¿30 = *30a + *3<>b = (5-5 + 1.6) X 10">V292 ±72)/7' (Lipson et al., 1997), may also play an important role. While it produces mainly H02 + CI, even a small contribution from the second channel (30b) could be important because it generates the reservoir species HC1 rather than the reactive chlorine atom. Experimental studies report a yield of 5 + 2% (Lipson et al., 1997). Dubey et al. (1998) show that even a small branching ratio for (30b), which likely proceeds through a four-center transition state, in the 7-10% range gives larger predicted 03 concentrations in the upper stratosphere due to smaller production of ozone-destroying atomic chlorine in (30a). This brings the model predictions and measurements into better agreement. Dubey et al. (1998) also show that the predicted branching ratio for the minor channel is very sensitive to the reaction dynamics, especially to the energy diffcrcncc between the transition state for the formation of the HOOC1* adduct and the formation of the four-center transition state.

In addition, the CIO, and BrO, cycles are interconnected by the reaction of CIO and BrO (Yung et al., 1980; Prather et al., 1984; McElroy et al., 1986; Wayne et al., 1995):

The recommended rate constants at 298 K for reactions (31a), (31b), and (31c) are 6.8 X 10"l2, 6.1 X 10"l2, and 1.0 X 10 cm' molecule s , respectively (DeMore et al., 1997). Reaction (31) is responsible for much of the uncertainty (~21%) in model predictions of ozone loss in the Arctic (Fish and Burton, 1997).

While more than 90% of reaction (31) at 298 K produces bromine atoms directly, the minor channel producing BrCl is important in the atmosphere under certain conditions (e.g., McKinney et al., 1997). (Measurements of OCIO formed in reaction (31a) in the Antarctic and Arctic stratosphere are discussed later.) For example, McKinney et al. (1997) report large fractions (50-95%) of total bromine in the form of BrO in the chemically perturbed region of the Arctic vortex from f 7 to 23 km at sunrise, along with high concentrations of CIO. Other studies have also reported higher measured concentrations of BrO than expected (e.g., Wahner and Schiller, f992; Avallone et al., f995). Comparison of the formation of BrO as a function of solar zenith angle to model predictions suggests that bromine atoms (and then BrO from the reaction with 03) must be produced by the rapid photolysis of a precursor, which McKinney et al. suggest to be BrCl. Br0N02, usually considered to be the bromine reservoir, does not photolyze sufficiently rapidly to be consistent with their observations. In addition, the enhanced CIO levels in the chemically perturbed region suggest N02, and hence BrONOz, must be low. That is, BrCl from reaction (31c) is acting as a nighttime reservoir for bromine, rapidly releasing it by photolysis at dawn.

Chlorine atoms are formed subsequently from the thermal decomposition of ClOO formed in reaction (31b) to CI + 02 (vide infra). Photolysis of OCIO in the gas phase formed in reaction (31a) gives O + CIO with a quantum yield of unity (DeMore et al., 1997). However, this does not lead to net ozone loss since 03 is regenerated from the reaction of O with 02. The alternate photolysis path giving CI + 02 has a quantum yield of <5 x f0~4 in the gas phase and hence is not important (Lawrence et al., f990). However, the photochemistry is very sensitive to the environment. For example, the photolysis of OCIO in solution or adsorbed in or on an ice matrix at 80 K gives ClOO (e.g., see Vaida and Simon, 1995; Dunn et al., 1995; and Pursell et al., 1995). If this were to occur at stratospheric temperatures, chlorine atoms would be regenerated and lead to net ozone destruction. On the other hand, photolysis of isolated OCIO at 150 K and either 360 or 367 nm generates a C1C102 species, proposed to be formed by the photochemical reactions of aggregates of OCIO (Graham et al., 1996b; Pursell et al., 1996). However, the results of laboratory studies suggest that less than 10~h monolayers of OCIO will exist on the ice surface at equilibrium at typical Antarctic springtime stratospheric temperatures and pressures (L. A. Brown et al., 1996; Graham et al., 1996a). In this case, such condensed-phase photochemistry is likely not important in stratospheric ozone depletion.

Since BrCl, produced in the minor channel (31c), absorbs strongly in the UV and visible (see Table 4.30), it also ultimately generates atomic bromine. Bromine atoms then react with 03 as well,

leading to a net loss of 03. This C10x-Br0x cycle is believed to be responsible for about 20-25% of the loss due to halogen chemistry in the 16- to 20-km region at midlatitudes (Wennberg et al., 1994; Lary, 1997) and is a significant cycle in the polar lower (< 14 km) stratosphere (Lary, 1997).

The destruction of 03 by chlorine and bromine can be "short-circuited" by removing either CI and Br or, alternatively, CIO and BrO. For chlorine atoms, this occurs by reaction with methane that has been transported from the troposphere:

CIO forms chlorine nitrate by reaction with NOz:

0 0

Post a comment