K

FIGURE 12.21 Schematic of polar stratospheric cloud (PSC) formation.

can assist in crystallization are available in typical laboratory studies, which may not be the case for solution droplets in the stratosphere; even under laboratory conditions, whether NAT or SAT first crystallize from the ternary solution may depend on the particular seed crystal/nucleation mechanism.

At any rate, these PSCs, which can be either liquid or solid, are known as Type I PSCs and form at temperatures about 2-5 K above the ice frost point. They are believed to contain large amounts of nitric acid and water (Fahey et al., 1989; Pueschel et al., 1989,

1990) due to the uptake of these species by the initial SSA as the temperature falls. Typical particle radii lie in the range from 0.1 to ~5 /jlm, and number concentrations in Antarctica are about 1—10 particles cm-3 (Kinne et al., 1989; Hamill and Toon, 1991; Hofmann and Deshler, 1991).

This understanding of the mechanism of formation of Type I PSCs is consistent with atmospheric measurements. For example, Massie et al. (1997) showed that gas-phase HNO, over Scandinavia in January 1992 decreased as the temperature fell while the volume of the PSC particles increased simultaneously. Beyerle et al. (1997) also showed using multiwavelength LIDAR over Scandinavia that the volume density of particles increased as the temperature approached the frost point in a manner consistent with the formation of liquid ternary H2SO4-HNO3-H2O solutions. Accompanying this was a depletion of gas-phase HN03, as it was taken up by the PSCs.

The exact composition and even the phase of these Type I PSCs under specific conditions is not well established (Tolbert, 1994, 1996). Particles collected by Pueschel et al. (1989) by impaction on gold wires had a low collection efficiency for nitric acid, which they suggested was indicative of nitric acid being in the solid state. However, infrared spectra of Type I PSCs in Antarctica in September 1987 indicate that under these particular conditions, the Type I PSCs were likely liquid ternary solutions of nitric and sulfuric acids and water (Toon and Tolbert, 1995). A similar conclusion was reached by Dye et al. (1996) during studies carried out in 1994, based on relationships between the temperature, PSC particle volume, and NOy, and by Stefanutti et al. (1995) using ground-based LIDAR measurements. Tabazadeh et al. (1995) report evidence for both liquid and solid Type 1 PSCs under different sampling conditions. Perhaps the most definitive evidence comes from direct mass spectrometric measurements of PSCs in the Arctic by Schreiner et al. (1999), who found the molar ratio of HzO to HN03 to be greater than fO at temperatures from 189 to f92 K, consistent with them being supercooled ternary solutions.

Aerosols and films consisting of binary solutions of H2S04 and water readily undergo supercooling (Anthony et al., 1995; fraci et al., 1995; Bertram et al., 1996; Carleton et al., 1997; Koop et al., 1998). However, once they do freeze, the tetrahydrate of sulfuric acid (SAT) is thought to be the stable form under stratospheric conditions (Peter et al., f992; Luo et al., 1992; Middlebrook et al., 1993; Zhang et al., 1993). (Although the octahydrate H2S04 • 8H20 has also been observed in laboratory experiments under conditions different from those found in the stratosphere (Imre et al., 1997), it appears unlikely to be important in the stratosphere.) SAT surfaces take up about a monolayer of HNO3 from the gas phase under conditions typical of the stratosphere; NAT formation on a SAT surface only occurs at high concentrations of gaseous HN03 and water vapor, interestingly, the SAT surface can be "preactivated" by prior formation and evaporation of NAT from the surface; that is, the gaseous concentrations of HNO3 and H20 required to form a NAT film on SAT are much lower if NAT is previously formed on the surface and then removed (Zhang et al., 1996).

Zhang et al. (1996) suggest that this preactivation changes the crystal structure of SAT at the interface to more closely match the NAT lattice. This nucleation and growth of NAT on solid SAT was proposed as one potential mechanism for Type I PSC formation (e.g., see Tolbert, 1994), although this process now appears to be less important because of the low probability for binary nucleation of nitric acid and water on SAT (MacKenzie et al., 1995).

Browell et al. (1990) used polarized laser light to probe Type I PSCs using LIDAR techniques and observed that in some cases, the reflected light had undergone substantial depolarization, whereas in other cases it had not. Particles that caused significant depolarization have been dubbed Type la particles and the others Type fb. Toon et al. (f990a) have shown that Type la PSCs are not spherical and are quite large, with radii equivalent to > 1 /¿m if they were treated as being spherical. These particles may be crystalline NAT or NAD (Rosen et al., 1993; Meilinger et al., 1995; Tabazadeh et al., 1996; Tabazadeh and Toon, 1996; Larsen et al., 1997) which can nucleate from liquid solutions as the temperature falls (e.g., Bertram and Sloan, 1998a,b; Prenni et al., 1998).

Type lb particles are spherical or nearly so and have typical radii of ~0.5 pm, in agreement with the observations of Stefanutti et al. (1991, 1995). These particles are thought to be ternary HN03-H2S04-H20 solutions (Carslaw et al., 1994; Tabazadeh et al., 1994a, 1994b; Hamill et al., 1996; Larsen et al., 1997). The probability of their occurrence in the Arctic has been shown to increase significantly at temperatures at which these ternary particles are expected to grow (Rosen et al., 1997). Like binary H2S04-H20 solutions, the ternary solutions have been shown in laboratory studies to undergo supercooling (Anthony et al., 1997), in agreement with the atmospheric observations.

Some field measurements of HNO-, suggest that the formation of liquid or solid Type 1 PSCs depends on the initial background sulfate aerosols on which the PSCs form. If they are liquid, then liquid ternary solution PSCs tend to form first as the temperature drops below 192 K, whereas if the sulfate particles are initially solids, solid Type lc PSCs may be generated (Santee et al., 1998).

As the temperature falls below the ice frost point, water condenses out as ice, forming large particles (Fig. 12.2f). These are known as Type if PSCs. They are formed at lower temperatures corresponding to the frost point of water (~ 188 K for stratospheric conditions), or possibly 2-3 K below that (Tabazadeh et al., 1997). They are much larger than Type 1 PSCs, of the order of 5-50 /¿m in diameter, and consist mainly of ice; their number concentrations are also smaller, typically in the range of ~10~2-10~3 cm~3 (Kinne et al., 1989; Hamill and Toon, 1991; Hofmann and Deshler, 1991). Because of their size, their rate of gravitational settling is relatively large, and settling rates of about f km per day can occur (Hamill and Toon, 1991). This acts to permanently remove cocondensing nitric acid, i.e., denitrifies the stratosphere, and also to dehydrate it (e.g., see Vomel et al., 1997).

For example, there is a loss of stratospheric gasphase nitric acid over the South Pole in June and July as PSC formation occurs (e.g., de Zafra et al., 1997; Santee et al., 1998). However, as temperatures increased in the spring, an increase in gaseous HN03 is not observed, consistent with the prior removal of HN03 by settling out of PSCs holding nitrate. By removing water, this gravitational settling also dehydrates the stratosphere (Gandrud et al., 1990). While these Type II PSCs are mainly water, it has been suggested that under some circumstances they may have a coating of NAT that inhibits the evaporation of water from the particles (Tolbert and Middlebrook, 1990; Peter et al., 1994; Middlebrook et al., 1996; Biermann et al., 1998).

Fourier transform infrared reflection-absorption spectroscopy studies (FTIR-RAS) by Tolbert and coworkers (Zondlo et al., 1998) of the uptake of HN03 on ice at 185 K have shown that a supercooled liquid forms on the surface; upon evaporation of water, the ice film becomes more concentrated in HN03 and at stoichiometries of 3:1 and 2:1 H20:HN03, respectively, NAT and NAD crystallize out. The reactions of C10N02 and N205 with the ice also led to the formation of supercooled H20-HN03 liquid layers on the ice surface.

Toon and Tolbert (1995) suggest that if Type I PSCs are primarily ternary solutions rather than crystalline NAT, the higher vapor pressure of HN03 over the solution would in effect "distill" nitric acid from Type I to Type II PSCs, assisting in denitrification of the stratosphere. This overcomes the problem that if Type II PSCs have nitric acid only by virtue of the initial core onto which the water vapor condenses, the amount of HN03 they could remove may not be very large. The supercooled H20-HN03 liquid layer observed by Zondlo et al. (1998) clearly may also play an important role in terms of the amount of HN03 that can exist on the surface of these PSCs.

LIDAR measurements of stratospheric aerosols (Browell et al., 1990) show that above the frost point, PSCs can be solids, perhaps solid SAT. Pure SAT, which does not form PSCs very efficiently, does not melt until quite high temperatures, about 210-215 K (Middlebrook et al., 1993; Iraci et al., 1995). However,

FIGURE 12.22 Composition of liquid in equilibrium with sulfuric acid tetrahydrate (SAT) as the temperature is lowered and SAT deliquesces in the presence of 5 ppm water vapor and 10 ppb HN03 at 50 m bar altitude (adapted from Koop and Carslaw, 1996).

Koop and Carslaw (1996) have shown that as solid SAT is cooled in the presence of gaseous water and nitric acid, the SAT deliquesces. That is, SAT takes up water as well as nitric acid and forms an equilibrium with water-nitric acid-sulfuric acid solutions (Fig. 12.21).

Figure 12.22 shows the composition in terms of the weight percent HN03 and H2S04 as a function of temperature as solid SAT is cooled from 194 K under conditions corresponding to a pressure of 50 mbar in an atmosphere containing 5 ppm H20 and an HN03 concentration of 10 ppb (Koop and Carslaw, 1996). Under these particular conditions, as the temperature falls below 192 K, the SAT is in equilibrium with a liquid film on the particle containing both HN03 and H20. The particular temperature at which SAT deliquesces is a function of the water vapor and gaseous nitric acid concentrations as shown in Fig. 12.23. As the temperature falls further and more HN03 and H20 are taken up into the liquid, the solid SAT dissolves completely, forming a ternary solution of the two acids and water. This solution can then act again to nucleate PSCs.

However, at lower HN03 concentrations than assumed, e.g., in a denitrified atmosphere, the formation of the liquid is shifted to temperatures about 3 K lower than shown in Fig. 12.22 (Martin et al., 1998). In addition, Martin et al. (1998) predict that under these conditions, SAT will not deliquesce to a liquid solution at temperatures above the frost point as shown in Fig. 12.22. Their experiments also suggest that the formation of the liquid, although thermodynamically favored, may be too slow to be important under stratospheric conditions.

Actually measuring the composition and phase of PSCs and aerosols in the stratosphere is extremely difficult. Some direct measurements have been made by collecting aerosol samples and subsequently analyzing them using techniques such as X-ray energy-disper

0 0

Post a comment