[OH] (106 radicals cm"3) PHOx (radicals cm"3 s"1)

FIGURE 6.36 (a) Measured OH concentrations as a function of altitude and model-predicted concentrations without acetone photolysis and with acetone photolysis, respectively, (b) Calculated rates of HOa. production from 03 and acetone photolysis, respectively, as a function of altitude. (Adapted from Wennberg et al., 1998.)

[OH] (106 radicals cm"3) PHOx (radicals cm"3 s"1)

FIGURE 6.36 (a) Measured OH concentrations as a function of altitude and model-predicted concentrations without acetone photolysis and with acetone photolysis, respectively, (b) Calculated rates of HOa. production from 03 and acetone photolysis, respectively, as a function of altitude. (Adapted from Wennberg et al., 1998.)

ing a significant contribution of other species, perhaps compounds such as CH3OOH, as well.

Based on this chemistry, the production rate of 03 is expected to be very sensitive to the NO concentration, increasing with NO (see also Chapter 16 for a discussion of the dependence of O3 generation on NOx). In this context, Folkins et al. (1998) suggest that acetone is likely the major contributor to enhanced ozone production in the upper troposphere, since increased CH3OOH and H202 concentrations at 9- to 12-km altitude were observed only at very small NO concentrations, indicative of clean marine boundary layer air; under such low NOx conditions, destruction rather than production of 03 is expected.

Another significant uncertainty in our understanding of the chemistry of the upper troposphere involves oxides of nitrogen. The measured NO^/NO ratios have been observed to be higher than expected based on model predictions (e.g., Jaegle et al., 1998b). The HN03/N0x ratio in the free troposphere also appears to be smaller than expected (e.g., Liu et al., 1992; Chatfield, 1994). One possible explanation is that there is some unrecognized chemistry reconverting NOv back to NOx (e.g., Chatfield, 1994) that would bring the measurements and model predictions into better agreement (e.g., Hauglustaine et al., 1996). However, it has also been suggested that these observations may reflect additional injection of NOx by convective transport, by lightning (e.g., McKeen et al., 1997; Prather and Jacob,

FIGURE 6.35 Measured OH concentrations at an altitude of 11.8 km near Hawaii and concentrations predicted using simple chemistry (adapted from Wennberg et al., 1998).

Solar zenith angle (degrees)

Solar zenith angle (degrees)

1997; Jaegle et al., 1998b; Brunner et al., 1998; Dias-Lalcaca et al., 1998), or by uptake of HN03 in clouds. Another possibility is errors in the kinetics for NOx and NOy reactions in the models. For example, those for the OH + N02 and OH + HN03 reactions have been recently revised and bring the models and measurements into better agreement (see Chapters 7.B.1 and 7.E.2 and Problem 7.9).

Injection of air from the stratosphere into the upper troposphere is also very important under some conditions for determining the concentrations of various species in this region. For example, Suhre et al. (1997) measured high 03 concentrations in the upper equatorial troposphere due to input from the stratosphere. Similarly, Dias-Lalcaca et al. (1998) carried out measurements of NO, N02, and 03 in the tropopause region on flights of commercial passenger aircraft. Regions of very high ozone, up to 450 ppb, were en-counted simultaneously with higher NOx concentrations. These were attributed to air of stratospheric origin.

There has been a great deal of research activity on the effects of subsonic aircraft in the upper troposphere, with respect to impacts both on the chemistry and on the radiation balance through effects on clouds and 03 (e.g., see April 15, May 1, and May 15, f998, issues of Geophysical Research Letters and the July 27, 1998, issue of Atmospheric Environment). Aircraft emit a variety of pollutants, including NOx, S02, and particles whose concentrations have provided "exhaust signatures" in some studies (e.g., Schlager et al., 1997; Hofmann et al., 1998).

Of particular concern is the impact of oxides of nitrogen emissions on 03 (e.g., see Ehhalt et al., 1992; and Ehhalt and Rohrer, 1995). As is typical of combustion systems, NOx emissions are primarily in the form of NO (e.g., Schulte et al., f 997). As discussed earlier in this chapter and elsewhere in this book (e.g., see Chapter 16), the impact of added NOx on the generation of 03 depends on existing levels. At low NO levels, added NO leads to increased 03 formation. However, at sufficiently high NOx, OH reacts with N02 to form HN03, effectively removing NOr from the system and terminating ozone production. The level at which this occurs in the upper troposphere is « 300 ppt NO (e.g., see Wennberg et al., 1998; and GrooB et al., 1998).

There is evidence from laboratory studies that heterogeneous reactions on sulfate particles may be important in the upper troposphere as well. For example, HCHO uptake into sulfuric acid solutions or ternary mixtures of sulfuric and nitric acids and water has been observed in laboratory studies (e.g., Tolbert et al., f 993; Jayne et al., 1996; traci and Tolbert, 1997). in sulfuric acid, the effective Henry's law constant at the low temperatures found in the upper troposphere and lower stratosphere is large, ~ 106-107 M atm"1, and polymerization of the HCHO occurs as its concentration increases (traci and Tolbert, 1997). When HN03 is present, a reaction occurs that generates HONO and HCOOH:

N02 is also formed, perhaps by the subsequent reaction of HONO with HN03 (which is the reverse of the surface hydrolysis of N02 discussed in Chapter 7.B.3 thought to be a significant source of HONO in the troposphere):

These products were observed at room temperature, although their formation at the lower temperatures found in the upper troposphere could not be confirmed (traci and Tolbert, 1997). Such reactions may contribute to a conversion of HN03 to NOr proposed by Chatfield (1994).

Similarly, the uptake of acetone into sulfuric acid-water solutions has been reported (Duncan et al., 1998), with the formation of 4-methyl-3-penten-2-one and trimethylbenzene at temperatures above 200 K and 75 wt% H2S04.

In short, although relatively little is known about the possibility of heterogeneous chemistry of organics in the upper troposphere, the results of initial laboratory studies suggest that this may be important.

4. Arctic

Atmospheric chemistry in the Arctic has been the subject of studies for many years, in part because of the observation of "Arctic haze" decades ago. This haze is composed of particles with significant amounts of sulfate, about half of which is due to long-range transport from other regions, particularly Eurasia during the winter (e.g., Barrie and Bottenheim, 1991; Polissar et al., f998a, 1998b).

As might be expected, levels of most pollutants are quite low in the Arctic when air is not being transported from populated regions. For example, during the spring in Alaska during periods of southerly wind flow, surface-level concentrations of small (C2-C5) hydrocarbons are 8 ppb C, 03 is ~ 20-40 ppb, NO is < 10 ppt, NOx is ~ 30 ppt, and NO^ is ~ 400 ppt (e.g., Doskey and Gaffney, f992; Honrath and Jaffe, 1992; Sandholm et al., 1992; Beine et al., 1996). PAN tends to be a larger portion of NO>( than normally expected for "clean" regions, often in the range of 50-90% (e.g., Bottenheim et al., 1986; Barrie and Bottenheim, 1991; Jaffe, 1993; Jaffe et al., 1997). ft forms an increasingly larger fraction of NO>( at higher altitudes due to stabilization with respect to thermal decomposition at lower temperatures (see Chapter 5.A.3c). For example, PAN was reported to be about 10% of NO near the surface, increasing to 45% at altitudes of 4.5—6.1 km over the Arctic (Sandholm et al., 1992; Singh et al., 1992).

An unusual phenomenon was reported in the Arctic in the mid-1980s. Ozone measured at ground level was observed to decrease rapidly to small concentrations, at times near zero (Bottenheim et al., 1986; Oltmans and Komhyr, 1986). As seen in Fig. 6.37, an increase in bromide ion collected on filters (f-Br) was inversely correlated with the 03 decrease (Barrie et al., 1988; Oltmans et al., 1989; Sturges et al., 1993; Lehrer et al., 1997); this could reflect either particle bromide or a "sticky" gas such as HBr that could be collected on the filter simultaneously. This correlation suggested that the loss of ozone was due to gas-phase chain reactions

April, 1986

FIGURE 6.37 (a) Surface-level 03 at Alert, Canada, and (b) filter-collected bromide (f-Br) during an ozone depletion episode (adapted from Barrie et al, f988).

April, 1986

FIGURE 6.37 (a) Surface-level 03 at Alert, Canada, and (b) filter-collected bromide (f-Br) during an ozone depletion episode (adapted from Barrie et al, f988).

involving bromine such as

Subsequently, additional reactions for the chain destruction of ozone were suggested, including one involving CIO (Le Bras and Piatt, 1995):

-> OCIO + Br, BrCl + hv -> Br + CI, OCIO + hv -> O + CIO.

Such chain reactions imply low NOx concentrations, which would otherwise terminate the chains by reacting with BrO and CIO to form Br0N02 and C10N02, respectively. Unless the nitrates could be recycled rapidly back to active forms of the halogens, they would terminate the chain. NOx concentrations do indeed appear to be small during 03 depletion episodes. Beine et al. (1997), for example, measured NOx < 4.5 ppt during 03 depletion episodes in Norway.

In addition, mechanisms for regeneration of photo-chemically active bromine that involve aerosol particles or reactions on the snowpack have been proposed. For example, McConnell et al. (1992) and Tang and Mc-Connell (1996) proposed that HBr and organobromine compounds could be converted to Br2 through adsorption and reaction on ice and aerosol particles. Fan and Jacob (1992) suggested that HOBr, formed by the reaction of BrO with HO?,

or by hydrolysis of bromine nitrate,

would form Br2 by reaction with Br~ in aerosol particles:

Aranda et al. (1997) have shown in laboratory studies that CH302 also reacts with BrO in a manner analogous to the H02 reaction (123), with a rate constant at s"1. About 80%

CH202, with the

298 K of 5.7 X 10"cm3 molecule of the reaction generates HOBr -+ remainder forming Br + CH30 + 02. In either case, photochemically active bromine species are regenerated.

Crutzen and co-workers (Sander and Crutzen, 1996; Vogt et al., 1996) have developed a model for chemistry in the marine boundary layer at midlatitudes, in which autocatalytic cycles involving sea salt particles generate photochemically active gases such as BrCl, Br2, and CI 2. It is likely that such chemistry also occurs in the Arctic as well. In these cycles, reactions (125) and (126) in the condensed phase,

followed by reactions such as (127)—(131), also occurring in the condensed phase,

lead to the generation of photochemically active halogens in the gas phase.

Much of this chemistry has been confirmed experimentally. For example, Kirchner et al. (1997) showed that HOBr reacts at 240 K with the surface of ice that has been doped with sea salt, generating gaseous BrCl and Br2. The chemistry is similar to that at midlatitudes, only in this case occurs in a quasi-liquid layer on the ice surface. Figure 6.38 is a schematic diagram of this chemistry, for which many of the reaction kinetics in aqueous solution have been reported (e.g., Wang et al., 1994). Similar chemistry occurs in the interactions of HOBr with HC1 on ice (Abbatt, 1994) and with aqueous aerosols of NaCl (Abbatt and Waschewsky, 1998) as well as in the reaction of 03 with frozen seawater ice in the dark (Oum et al., 1998b); in the latter case, 03 oxidizes Br" to BrO /HOBr and the chemistry then leads to the generation of gaseous Br2 through the chemistry in Fig. 6.38.

The central point is that photochemically active bromine, and perhaps chlorine (see following), compounds are generated that lead to the chain destruction of gaseous 03 at ground level at polar sunrise.

However, what remains unknown is the source of the original bromine that initiates the chemistry. There have been a number of hypotheses, including the photolysis of bromoform which is generated by biological processes in the ocean (Barrie et al., 1988) or reactions of sea salt, either suspended in the air or deposited on, or associated with, the snowpack. These include photolysis of BrN02 formed from the reaction of sea salt particles with N205 (Finlayson-Pitts et al., 1990), the



FIGURE 6.38 Schematic diagram of HOBr chemistry with sea salt particles/ice (graciously provided by T. Benter).

formation of HOBr from the oxidation of bromide by HSO^ (Mozurkewich, 1995), or the reaction of 03 with bromide in frozen seawater ice (Oum et al., f998b; De Haan et al., 1999).

Supporting a seawater source for the halogens is the observation by Shepson and co-workers of significant amounts of as yet unidentified photolyzable chlorine as well as bromine compounds in the spring in the Arctic (Impey et al., 1997a, 1997b). In addition, Piatt and co-workers have detected both BrO and CIO at the surface during ozone depletion events (Piatt and Hausmann, 1994; Hausmann and Piatt, 1994; Tuckermann et al., 1997).

It is also clear that during periods of low surface ozone, chlorine atoms are a major reactant for hydrocarbons (e.g., Jobson et al., 1994; Solberg et al., 1996; Ariya et al., 1998). Figure 6.39, for example, shows the measured ratios of isobutane, n-butane, and propane during an ozone depletion event (Jobson et al., 1994). These particular pairs of hydrocarbons were chosen to differentiate chlorine atom chemistry from OH reactions. Thus isobutane and propane have similar rate constants for reaction with CI but different rate constants for reaction with OH. If chlorine atoms are responsible for the loss of these organics, their ratio should remain relatively constant in the air mass, as indicated by the line marked "CI." Similarly, isobutane and n-butane have similar rate constants for removal by OH but different rate constants for reactions with


0 0

Post a comment