Heterogeneous Chemistry

Observations continue to highlight the importance of heterogeneous reactions on sulfate aerosols, and implicate NO hydrolysis [ Van Doren et al, 1991] as the most important process for altering the gas-phase photochemical balance of ozone-destroying radical species in the mid-latitude lower stratosphere. This reaction produces the photochemically stable HNO, so it is a sink for NO. This, in turn, leads to enhancements of reactive halogen oxide and HO radicals relative to their abundances if only gas-phase reactions are considered. Rodriguez et al. [1991] and King et al. [1991] were the first to cite enhancements of ClO in the high-latitude lower stratosphere in winter as evidence for the potential importance of this process. Based on a comparison of a ClO balloon profile to results from a model constrained by empirical tracer interrelationships, Avallone et al. [1993a] confirmed the importance of this reaction over a broader altitude range. However, these initial studies were limited by the lack of simultaneous measurements of NO.

Because NO is rapidly photolyzed in daylight to produce NO, that rate of formation of HNO by NO hydrolysis depends upon formation of NO after sunset and subsequent heterogeneous conversion in darkness. In the presence of high abundances of aerosols the latter step is sufficiently fast that production of NO, a process limited by the temperature-dependent formation of NO from NO + O, largely determines the rate of HNO production. At low aerosol abundances the reaction probability, available aerosol surface, and photolysis of NO become the limiting factors. Consequently, the importance of NO hydrolysis depends on temperature and available sunlight (i.e. on altitude, latitude, and season) in a fairly complex way.

In regions of high aerosol loading, Fahey et al. [1993] and Mills et al. [1993] observed non-linear reductions in NO at mid-latitudes. These decreases, which were reproduced well by models incorporating NO hydrolysis on sulfate, reached ``saturation'' as modeled by Prather [1992]. At different altitudes, saturation is limited by formation of NO at different aerosol surface loadings. Perliski and Solomon [1992] showed that dramatic decreases in NO columns, even those at high solar zenith angles, could not have been due to aerosol-induced changes in atmospheric radiation. Thus, column measurements from the ground and from aircraft [ Coffey and Mankin, 1993] are valid and indicate that the perturbation to NO extended over a wide altitude range.

The response of ClO to decreased NO due to Pinatubo aerosol has been observed by aircraft, balloons, and satellite [ Fahey et al., 1993, Wilson et al., 1993, Avallone et al., 1993b, Dessler et al., 1993, and Waters et al., 1995b]. In addition, obsevations of ClO from the ER-2 and UARS [ Toohey et al., 1991,1993b and Waters et al., 1995b] have revealed a seasonal dependence of mid-latitude ClO that is inverse to what would be expected in the absence of heterogeneous chemistry. Avallone et al. [1993b] showed that the largest enhancements of ClO were observed at lower latitudes, consistent with the saturation effects displayed by NO. Solomon et al. [1994] reported that NO reductions were observed from McMurdo even during the polar summer, and suggested that excursions of air to dark regions or direct conversion of NO to NO could be responsible.

The increases in aerosol surface area due to Pinatubo should also have enhanced the rates of other heterogeneous reactions. Perhaps most important is the hydrolysis of ClONO to HOCl and HNO on sulfate. Because ClO from photolysis of HOCl can titrate any remaining NO, this process can potentially reduce NO abundances below those of reactive chlorine. The efficiency of this reaction varies strongly with aerosol composition [{ Hanson et al., 1994], and hence temperature, such that it is likely to be most important in cold regions where nitric acid photochemistry is slow (i.e., near the tropopause or in the winter polar regions). Ground-based observations [ Solomon et al., 1993] have implicated this process over Antarctica, and it may have contributed to additional ozone depletion observed there in 1992 [ Hofmann et al. 1993].

Whether ClONO hydrolysis is important at mid-latitudes is less clear. Fahey et al. [1993] and Dessler et al. [1993] argued that this process did not contribute significantly to NO and ClO abundances they observed outside the arctic vortex under volcanic conditions. However, both of these studies examined this issue with zonally symmetric models which, in light of the work of Kawa et al. [1993] and Solomon et al. [1994], could have accentuated the effects of other heterogeneous reactions, thereby masking the potential importance of ClONO hydrolysis. In addition, the ClO abundances at lower altitudes reported by Dessler et al. [1993] were systematically higher than those observed by the ER-2 and UARS outside the vortex during the same winter. Thus, it remains unclear whether ClONO hydrolysis at northern middle and high latitudes was an important process at the peak of Pinatabo loading.

Hanson et al. [1994] has outlined a procedure for incorporating into models the various reactions on sulfate aerosols. Those involving HCl must be carefully parameterized because of the temperature dependence of HCl solubility. Their calculations suggest that reactions of HCl with ClONO and HOCl on sulfate could have contributed to enhancements of reactive chlorine during Pinatubo years, such as those observed by Solomon et al. [1993], and would be most important in regions that do not quite reach NAT temperatures, such as just outside the antarctic vortex. Such reactions could help explain the increase in size of the antarctic ozone hole, both vertically and horizontally, in the years immediately following the eruption of Pinatubo.