Effects of Nonmethane Hydrocarbons in the Atmosphere

The photochemistry of several unreactive and moderately reactive nonmethane hydrocarbons (NMHC) in the background troposphere and stratosphere was investigated. A one-dimensional steady state model was employed to determine the vertical distributions of C(cid:127).H6, C(cid:127).H(cid:127)., Calls, C4H(cid:127)0, and CsH(cid:127).. The impact of these species upon the tropospheric and stratospheric odd-hydrogen, odd-oxygen, chlorine, and carbon systems was studied for various possible free radical profiles and eddy difftlsion coefficients. Our results indicate that NMHC probably have only a small effect upon the background atmospheric photochemistry, although they might constitute a nonnegligible source of atmospheric CO. Also, Cl atoms, in predicted present-day concentrations, comprise the major sink for stratospheric NMHC. Finally, if the chlorovinyl molecule (CHCI = CH) were stable in the lower stratosphere, it would then be conceivable that C2H2 could be partially effective as a chain terminator to impede catalytic removal of stratospheric Oa by CI and CIO.


INTRODUCTION
While the photochemistry of methane in the background troposphere and stratosphere has been studied in detail [cf. Levy, 1972;Wofsy et al., 1972;Crutzen, 1974;Wofsy, 1976;, the photochemistry of nonmethane hydrocarbons (NMHC) in the ambient atmosphere has been largely ignored. Many of the more reactive anthropogenically produced NMHC, such as propylene, are believed to play a key role in photochemical smog formation in urban areas. However, because of their relatively short photochemical lifetimes (the photochemical lifetime of propylene in the background troposphere is about 0.5 day) the ambient concentrations of these reactive NMHC are probably small, the effects of reactive NMHC being limited to the local source regions. Similarly, while terpenes, which are produced naturally by vegetation, may play a role in the production of ozone and blue haze near heavily forested areas [cf. Went, 1960], their short photochemical lifetimes probably preclude effects in regions removed from intense terpene production. However, the less reactive NMHC, such as ethane and acetylene, are longer lived, and depending on their production rates they could have significant abundances throughout the troposphere and lower stratosphere. For instance, we estimate lower tropospheric photochemical lifetimes of 25 and 50 days for C2H0 and C2H2, respectively. By comparison, the photochemical lifetime of CO, whose global distribution indicates considerable vertical and horizontal transport [Seiler, 1974], is about 40 days in the lower troposphere. Singh [1977] has suggested that long-lived NMHC may be sufficiently abundant in the stratosphere to act as chain terminators for stratospheric chlorine. To determine if NMHC are abundant enough to have a significant impact upon the ambient photochemistry, we have calculated the vertical distribution of C2H0 and Calla as well as some moderately reactive (lower tropospheric lifetimes of about 5 days) NMHC: Calls, C4H•0, and C,H•a. (All hydrocarbons were taken to be the normal isomer; i.e., C,H•a is n-C,H•2.) ent free radical profiles and eddy diffusion coefficients. Owing to uncertainties in key reaction rates the abundance of tropospheric OH is uncertain by an order of magnitude. The tropospheric OH profile we have adopted from Chameides and Stedman [1977] implies a lifetime for CHaCCIa of about 5 years, as compared with Singh's [ 1977] inferred result of (7.2 + 1.2) years. The CI profile was obtained by assuming a total inorganic chlorine (CIX) stratospheric mixing ratio profile that increases monotonically from about 0.1 ppb at 15 km to 1.4 ppb at 40 km. This abundance of CI should roughly approximate present-day stratospheric concentrations.
The photochemical reactions included in our model are listed in Table 2. Calculations indicate that in the troposphere the major sink for NMHC is reaction with OH. In the stratosphere, for CI densities greater than about 1tY cm -a, reaction with CI is the major loss process for the alkanes, while for C2H2 the reaction with OH remains the major sink for the entire altitude region. Note that many of the rate constants are uncertain, e.g., the reaction of CI with C2H2. Lee and Rowland [1977] suggest that the reaction produces the excited C2H2C1' radical which either stabilizes to CHCI=CH or decomposes to form C2H2 and CI again. They estimate the effective rate constant for conversion of CI to CHCI=CH to be 10 -•2 cm a molecule-• s -•, which we have used in our calculations for the reaction of CI with C2H•.: (Rll)

CI + C2H2 • CHCI = CH
The only source of hydrocarbons in our model was a flux from the ground, which could be a result of natural and/or anthropogenic activities. We did assume a photochemical source of CO via the methane oxidation chain [McConnell et al., 1971], so that for every CHa oxidized a CO was produced. We did not include the production of CO from the oxidation of NMHC. However, our results indicate that the oxidation of NMHC may produce significant quantities of CO depending  In these calculations we have normalized our results to ground level mixing ratios of 1 ppb for C•.H6 and C•.H•. and 0.1 ppb for C3H5, C•Hlo, and C, Hi•, the moderately reactive NMHC. As future measurements materialize, our results can be multiplied by the appropriate factor to yield the proper densities, fluxes, or reaction rates. For instance, if 0.1 ppb is more realistic for C•H6 than I ppb, our results for C•.H6 should be multiplied by a factor of 0.1. Finally, we adopted ground level mixing ratios of 1.4 ppm for CHi and 0.11 ppm for CO [Ehhalt and HeMt, 1973;Seiler, 1974].   convert CI, to HCI, but (R 11 ) represents a loss process for total inorganic chlorine, CIX.) Figure 2 illustrates the rate at which C1 reacts with the hydrocarbons for the standard model. The ratio of the total rate at which CI reacts with NMHC to the rate at which CI reacts with CH4 for a variety of assumed OH, CI, and K profiles is depicted in Figure 3. Note that while C2H8 attacks CI at a large rate compared with CH4 in the lower stratosphere, in the 25-to 40-km region, where Cl-catalyzed destruction of O3 is effective, the loss of CI due to NMHC is relatively small. For instance, in the standard model, for the CI loss rate due to NMHC to be 10% of the loss rate due to CH• at 25 and 30 km, we would have had to assume 10 times and 50 times larger ground level NMHC mixing ratios, respectively. Background concentrations of 10-50 ppb C2H8 and C2H• are much larger than observations indicate. The relative rate of attack of NMHC upon CI can be enhanced by decreasing OH and CI densities and/or increasing K, as illustrated in Figure 3. For instance, assuming a CI profile of 1/10, the standard model yielded a rate of NMHC attack upon C! which was 15% that of CH• at 25 kin. Decreasing CI by a factor of 100 yields an even greater relative role for NMHC. However, for this low an abundance of CI the loss of O3 due to CIcatalyzed reactions is small in comparison with other loss processes. Thus our calculations indicate that for low levels of stratospheric CIX (0.1 ppb or less near 40 km), NMHC may have a small but significant effect upon the chlorine chemistry near 30 km. However, for abundances of C1X sufficient to affect the stratospheric ozone budget the destruction of NMHC by reactions with CI below 25 km is large enough to prevent NMHC from having a significant impact upon C1X above 25 km; this applies for ground level NMHC densities of a few ppb or less.  Ii I  I  i  I I I III I  i  i  I i I i ii   40   I  I  I I I IllJ  I  I  I I I I Ill  I  I  I I I Illl  I  I  I I I  ) to the rate of reaction of CI with CH• (R2) as a function of altitude. The solid line is for the standard model, the dotted line is for K = 2 X 10 • cm • s -• between 0 and 10 km, the dot-dash line is for OH and CI decreased by a factor of 2, the long-dash line is for CI decreased by a factor of 10, and the short-dash line is for CI decreased by a factor of 100. chains to be smaller. As was noted by Lee and Rowland [ 1977], it is likely that O: will attach to the radical to form a peroxy radical; the product so formed is not very likely to be stable against further photochemical attack, but this should be investigated. , Note that the possible existence of the asymmetric chlorine dioxide molecule ClOO could alter our results. Using a lower limit for AHf(ClOO), 84 kJ/mol [Watson, 1974], we calculate that ClOO may be 50-100 times more concentrated than Cl in the lower stratosphere. While the calculated Cl atom densities (  Sze, 1977] indicate that anthropogenic production of CO is approximately equal in magnitude to natural production via the oxidation of CH• [Levy, 1971;McConnell et al., 1971;Weinstock, 1969]. It is possible that the oxidation of NMHC ultimately results in the production of CO via oxidation chains similar to the CH• oxidation chain [cf. Demerjian et al., 1974;Breen and Glass, 1970;Williamson and Bayes, 1967]. If we assume 1 ppb C:H6 and C2H: at the ground and 0.1 ppb C3Hs, C•H•0, and CsH•. at the ground and if we assume that every carbon atom eventually becomes CO, then we obtain a CO production rate due to these NMHC of about 3 X l0 •ø CO molecules cm -•' s -• (or 220 Mton CO yr-:). This compares with a calculated production rate from CH• of 1 X 10 xx cm -•' s -• (or 730 Mton CO yr-•), a photochemical sink via (R26) of 2.3 x 10 • cm -•' s -• (or 1540 Mton CO yr-•), and an anthropogenic source of 9 x l0 •ø cm -•' s -• (or 640 Mton CO yr -•) [Seller, 1974]. However, our calculated CO production rate from the oxidation of NMHC is uncertain owing to the lack of detailed observations of the ambient concentrations of these species. Furthermore, heterogeneous loss processes may block the photooxidation of NMHC and thereby prevent the release of CO.

CONCLUSION
Our calculations (Figure l) imply that measurable quantities of relatively unreactive NMHC, especially C•.H• and C•.H•., may be present in the upper troposphere and stratosphere. Nevertheless, our results indicate that NMHC are not likely to have a large impact on the background photochemistry of the troposphere and stratosphere, although local effects near source regions are probable. In view of the many uncertainties in the photochemistry of the lower atmosphere, our findings support the present-day practice of many modelers who neglect NMHC in their calculations. However, we have found that if CHCI=CH is stable in the lower stratosphere, then C:H: may act as a partially effective chain terminator to impede catalytic removal of stratospheric O3 by CI and CIO. Interestingly, Cl atoms at levels predicted to exist in the present-day lower stratosphere might comprise the major stratospheric sink of NMHC.
Assuming ground level mixing ratios of 1 ppb for C:H• and C:H: and 0.1 ppb for C3H•, C•H:0, and CsH•:, we obtain a total NMHC loss rate of 3 X l0 •ø atoms C cm -• s -•, which could conceivably result in the production of 200 M ton CO yr -•.
More detailed measurements of NMHC levels in various environments are needed to refine this estimate.