Hydrogen oxides and nitrogen oxides are two basic components of atmospheric chemistry. Hydrogen oxides (HO_X) consist of the hydroxyl radical (OH) and the hydroperoxyl radical (HO₂). Nitrogen oxides (NO_X) consist of nitric oxide (NO) and nitrogen dioxide (NO₂). Together they determine the atmosphere's oxidizing, or cleansing, power and the troposphere's ozone production.

OH and HO₂ concentrations typically increase during the day and are larger in polluted environments than in clean ones. Two factors are responsible for most observed HO_X variance: the HO_X production rate, denoted as P(HO_X), and the NO_X concentration [Logan et al., 1981; Ehhalt et al., 1991; Jaeglé et al., 1999, 2000a]. Reactions that destroy HO_X are also important, but as we shall show, the dominant HO_X-destroying reactions are basically determined by the HO_X production rate and the NO_X concentration. In this article, we discuss the production of OH and HO₂, the interactions between HO_X and NO_X that establish the OH and HO₂ concentrations and the ozone production rate, P(O₃), and the lessons garnered from a growing body of atmospheric measurements of OH and HO₂.

The production of OH and HO₂

Globally the most important HO_X source is the photodestruction of ozone to produce the excited state oxygen atom, O(¹D), which reacts with water vapor to produce two OH molecules [Levy, 1971]. It is by far the largest HO_X source in the lower troposphere. However, in the upper troposphere and in some continental environments, other sources can be important and indeed dominant. Some of these other sources are made elsewhere by oxidation processes initiated by the reaction of OH with a hydrocarbon. They are then transported to a new location where they become important or dominant HO_X sources (Figure 1). The convection associated with clouds appears to be effective at lifting HO_X sources from planetary boundary layer (~0.5-2 km) to the middle and upper troposphere [Prather and Jacob, 1997]. These HO_X sources become more important than O(¹D)+H₂O above ~6 km altitude. Precursor gases to these HO_X sources may also be lifted by convection. Thus, pollution from megacities and biomass burning, lofted into the upper troposphere, can become the dominant HO_X source and result in efficient ozone production. It can also be transported great distances before descent, possibly influencing the chemistry of remote regions.

Several of these convectively lifted gases that have origins near Earth's surface have been identified. They include acetone (CH₃C(O)CH₃) [Singh et al., 1995; Arnold et al., 1997], methylhydroperoxide (CH₃OOH) [Prather and Jacob, 1997; Cohan et al., 1999], and formaldehyde (CH₂O) and other aldehydes [Müller and Brasseur, 1999]. Hydrogen peroxide (H₂O₂) is another source [Chatfield and Crutzen, 1984], but it is so water soluble that it is probably scavenged by precipitation as it is drawn up through clouds. All of these chemical species produce OH or HO₂ after they have been destroyed by sunlight. For acetone and methylhydroperoxide, the photodestruction leads to production of formaldehyde (CH₂O), which is rapidly destroyed by sunlight to produce HO₂. Other as-yet-unidentified source gases are also likely lifted from the surface into the upper troposphere [Crawford et al., 1999], especially in convection over continents.

Another HO_X source is the reaction of OH with methane [Logan et al., 1981; Wennberg et al., 1998]. This reaction initiates an oxidation process that eventually produces water vapor and carbon dioxide. Even though OH is destroyed in the initial reaction with methane, generally more HO_X is produced than destroyed. This behavior is called autocatalytic. The amount of HO_X produced depends on the amount of NO present, but for much of the atmosphere, this autocatalytic process produces a net ~0.6 HO_X molecules for every methane and OH molecule initially consumed.

For larger non-methane hydrocarbons, such as those found in urban areas and forests, the oxidation pathways can be much more complex than for those of methane [Trainer et al., 1987]. After the initial reaction between OH and these volatile organic compounds (VOCs), RO₂ soon forms, where R is a hydrocarbon radical. Formation of RO₂ often leads to formation of HO₂. Whether the oxidation of a particular hydrocarbon is a net HO_X source or sink depends on a number of factors, particularly the amount of NO present. While VOCs are certainly important for the chemistry in Earth's planetary boundary layer, where people live, they may have a more global role if they are carried to the upper troposphere.

Because HO_X production, P(HO_X), is generally driven by sunlight, we would expect that OH and HO₂ exist only during daylight. However, HO₂ has been observed to persist through the night at levels of a few parts per trillion by volume (1 pptv = 10^-12 molecules per total molecules in the air). Such observations point to possible nighttime HO_X sources, such as OH production in the reaction between O₃ and certain terpenes [Paulson and Orlando, 1996]. While measurements hint at the presence of such sources [Kanaya et al., 1999; Hard et al., 1992], observations have not firmly established their presence. Even if such sources exist, their impact on even regional chemistry is uncertain.

The roles of NO_X and P(HO_X) in HO_X photochemistry

Some HO_X sources create OH; some create HO₂ (Figure 2). Once created, HO_X is partitioned within seconds into OH and HO₂ by a few fast reactions. The reaction of OH with CO, volatile organic compounds (VOCs), ozone, and other chemicals produce HO₂. Similarly, reactions of HO₂ with NO and O₃ produce OH. Throughout much of the atmosphere, the production of OH and HO₂ through these reactions is much faster than the OH and HO₂ production from other sources. When this is true, the interchange between OH and HO₂ comes into steady-state relative to the production and destruction of HO_X (i.e., the OH and HO₂ concentrations maintain a balance with each other, but the total HO_X concentration adjusts to the changing environmental conditions, such as sunlight).

The dominant reactant with OH is usually CO, although other species can be important. In forested regions, the dominant reactant with OH is often isoprene, a 5-carbon molecule that is emitted mostly by deciduous trees. In urban environments, the dominant reactant can be oxygenated species, particularly formaldehyde and other aldehydes.

The relative rates of reaction of OH with CO, VOCs, and O₃ and HO₂ with NO an O₃ determine how much HO_X is OH and how much is HO₂. The greater the amount of CO or VOCs, the more HO₂ there is relative to OH. With rare exceptions, the concentration of HO₂ is more than five times greater than the concentration of OH and is usually 10's to 100's of times larger.

During the cycling between OH and HO₂, ozone (O₃) is produced. Ozone production occurs when HO₂ and RO₂ react with NO to form OH (or RO) and NO₂. The NO₂ is rapidly destroyed by sunlight into NO+O, and the O atom rapidly reacts with O₂ to make O₃. On the other hand, the production of HO_X can destroy O₃, as can the reactions of OH and HO₂ with O₃. Thus the net ozone production is dictated by the expression:

d[O₃]/dt = {k_HO2+NO[HO₂]+k_RO2+NO[RO₂]][NO] — {J_O3f[H₂O] + k_OH+O3[OH] + k_HO2+O3[HO₂]}[O₃]

where [OH] is the OH concentration, k_X+Y is the reaction rate coefficient for X+Y® products, and J_O3f[H₂O] is the production rate of O(¹D) times the fraction that react with H₂O. The level at which NO becomes more important than O₃ depends on the [NO] and [O₃] [Crutzen, 1979]. For typical [O₃] values, it occurs when the NO mixing ratio exceeds a few tens of pptv.

HO_X is permanently lost when its hydrogen atom is recombined into water vapor. HO_X can be destroyed by several reactions, although the relative importance of the different reactions depends on the amount of NO_X present (Figure 2). At low NO_X, HO₂ is hundreds of times more abundant than OH. Most HO_X loss occurs by HO₂ reactions with either HO₂ or RO₂ (R = CH₃ or other hydrocarbon radicals denoted as R) to form peroxides (HOOH, CH₃OOH, ROOH). Peroxides do not represent a permanent HO_X loss because they can be destroyed by sunlight to produce HO_X again. For conditions in which the peroxide sources and sinks come into steady-state, HO_X is lost when OH reacts with the peroxides to produce H₂O.

At low levels of NO, below a hundred pptv in remote areas and below a thousand ppbv in polluted urban areas, NO shifts HO_X from HO₂ to OH, so that the fast reaction OH+HO₂®H₂O+O₂ becomes an important loss as OH increases. At the same time, sufficient NO₂ usually exists in photostationary state with NO that pernitric acid, HO₂NO₂, is formed by the reaction HO₂+NO₂+M®HO₂NO₂+M, where M is an air molecule. Pernitric acid is unstable to both sunlight and thermal decomposition; its production and loss often balance, so that it is neither a HO_X source nor sink. Its main loss is by reaction with OH to produce H₂O.

Finally, at high NO_X, HO_X is increasingly shifted from HO₂ to OH. The increased OH reacts with NO₂ to form nitric acid (HNO₃). Nitric acid is only slowly destroyed by sunlight and may be removed by rapid scavenging by aerosols, cloud drops, and rain drops. For much of the atmosphere, it is permanently removed by reaction with OH to produce H₂O.

The concentrations of OH and HO₂, and ozone production, P(O₃), display this strong dependence on NO_X, as in Figure 3. Two midday cases are presented: a low P(HO_X) case typical of the upper troposphere and a high P(HO_X) case typical of the urban planetary boundary layer. In both cases, as NO_X increases, HO₂ initially is constant, then decreases at first linearly and then as the square of NO_X as more HO_X is shifted by reaction with NO to OH, which then forms and reacts with HNO₃. As NO increases, OH initially has little NO dependence, but develops a near-linear dependence, eventually peaking at mid-range NO_X values, and finally decreases linearly with NO_X as nitric acid (HNO₃) is formed at high NO_X values.

The HO_X production rate, P(HO_X), also influences OH and HO₂ values. At low NO_X, the dominant HO_X loss reactions are HO₂+ HO₂®H₂O₂ + O₂ and HO₂ + OH®H₂O + O₂, both of which are quadratic in HO_X. So, if P(HO_X) balances the HO_X loss rate, then the HO_X concentration depends on the square root of P(HO_X) and HO_X only doubles when P(HO_X) quadruples. At high NO_X, HO₂ and OH depend linearly on P(HO_X).

The production of ozone, P(O₃), is intimately tied to HO_X, and for high enough values of NO_X, P(O₃) ~ k_HO2+NO[NO][HO₂]. For very low [NO], which are typical in the clean tropical Pacific lower troposphere, O₃ is destroyed [Schultz et al., 1999]. As NO_X increases, P(O₃) increases almost linearly with NO_X, reaches a peak when HO₂ becomes proportional to (NO_X)^-1, and then decreases linearly with NO_X as HO₂ decreases as the square of NO_X. The peak in the ozone production shifts to higher NO_X as P(HO_X) increases.

If P(O₃) increases as NO is increased, then ozone production is said to be NO_X-limited. If P(O₃) is constant or decreases as NO_X is increased, then ozone production is said to be NO_X-saturated. Whether a region of the atmosphere is NO_X-limited or NO_X-saturated is an important question for determining the impact of future human pollution, particularly NO_X, on that region.

Observational tests of HO_X photochemistry

How well do the observations fit this basic picture of HO_X and its relationship with NO_X? One test is the comparison of observed HO_X to models that are constrained by simultaneous measurements of other chemical species and environmental parameters. Field studies have sampled environments with wide ranges of NO_X and P(HO_X). In general, the daytime observations agree to within a factor of two or better with models that are tightly constrained by a large number of simultaneous measurements of other chemical species and environmental parameters. Considering the range of conditions sampled, this level of agreement is remarkable. Despite this general good agreement, some disagreements persist. The ways that observations and models disagree are different for different regions.

In the upper troposphere, observations are sometimes greater than model values, implying that additional HO_X sources are not being measured [Wennberg et al, 1998; Jaeglé et al, 1997; Brune et al., 1998]. It is presumed that these source species were acetone and methylhydroperoxide, which were not measured during these studies. Neither the global extent of their importance nor the possibility of other sources is known.

Near Earth's surface, observed OH and HO₂ have generally been equal to or less than modeled OH and HO₂ by as much as a factor of 0.7 [Plass-Dülmer et al., 1998; Eisele et al., 1996; Mount and Williams, 1997; Frost et al., 1999; ]. Even though the deviation of the observations from the models is within the combined uncertainties, it has occurred in enough environments to be troubling. Proposed causes for the deviations include OH reactions with unmeasured hydrocarbons and HO₂ loss on aerosols.

In the upper troposphere over the North Atlantic, recent measurements suggest an NO dependence of the deviation between measured and modeled HO_X [Brune et al., 1999; Faloona et al., 2000]. HO_X observations are about 60% of the model values for NO less than a hundred pptv and increase to roughly twice the model vales at NO values exceeding a few hundred pptv. This effect is seen for both OH and HO₂. At the same time, observations and models give the same HO₂/OH ratio. It is not understood what is causing this NO-dependent discrepancy. It may be that additional HO_X sources accompany the increased NO_X or that the photochemical interactions between HO_X and NO_X are not well understood for the cold, upper troposphere. These problems could involve pernitric acid (HO₂NO₂) or nitric acid (HNO₃) formation or destruction.

No matter what the cause for the NO-dependency, these observations suggest that more of the atmosphere is NO_X-limited than models calculate and that it remains NO_X-limited to higher values of NO_X. Currently, models calculate that the upper troposphere would quickly become NO_X-saturated if much more NO_X were added. In other field studies in the upper troposphere, NO_X-limited conditions are also observed more frequently than calculated by models [Jaeglé, 2000b]. If this is true, then increases in pollution from biomass burning, urban smog, or aircraft could produce more O₃ than expected.

Challenges lie ahead. Measuring the dependence of OH, HO₂, and P(O₃) on varying NO_X and P(HO_X) a few times in a few places is not enough. The fundamental properties of atmospheric chemistry must be examined in several environments, such as cities, forests, the marine boundary layer, and the free troposphere. They must be studied with different HO_X instrumentation to ensure that the observations are real. Only then will we be able to develop confidence in the understanding of atmospheric oxidation and ozone production and to reduce the potential for surprises in chemical mechanisms.

Figure 1. Schematic of tropospheric HO_X photochemistry. Complex chemistry in Earth's planetary boundary layer (PBL) produces oxygenated species. These can be convectively lifted into the upper troposphere (UT), where they become HO_X sources. In the UT, HO_X is exchanged between HO₂ and OH by reaction of HO₂ with NO and OH with CO. The reaction of HO₂ with NO leads to O₃ production.

Figure 2. Schematic of the influence of NO_X and P(HO_X) on HO_X photochemistry. HO_X is produced as either OH or HO₂, is rapidly exchanged between HO₂ and OH, creating O₃ in the process, and is removed by reactions that eventually form H₂O. The dominant removal reactions are determined by the NO_X abundance and the HO_X production rate.

Figure 3. (a) HO_X concentrations and (b) O₃ production rates as a function of NO_X for conditions in the upper troposphere, UT (solid line), and planetary boundary layer, PBL (dashed line). In (a), the higher lines are HO₂ and the lower lines are OH at low NO_X. In (b), P(O₃) in the UT is multiplied by 10. For the UT, P(HO_X) = 2.5 x 10⁴ molecules cm^-3 s^-1; for the PBL, P(HO_X) = 2.5 x 10⁶ molecules cm^-3 s^-1.

References

1. Arnold, F., et al., Observations of upper tropospheric sulfur dioxide and acetone pollution: Potential implications for hydroxyl radical and aerosol formation, Geophys. Res. Lett., 24, 57-60, 1997.
2. Brune, W.H., et al., Airborne in situ OH and HO₂ observations in the cloud-free troposphere and lower stratosphere during SUCCESS, Geophys. Res. Lett., 25, 1701-1704, 1998.
3. Brune, W.H., et al., OH and HO₂ chemistry in the north Atlantic free troposphere, Geophys. Res. Lett., 26, 3077-3080, 1999.
4. Chatfield, R.B., and P.J. Crutzen, Sulfur dioxide in remote oceanic air: Cloud transport of reactive precursors, J. Geophys. Res., 89, 7111-7132, 1984.
5. Cohan, D.S., et al., Convective injection and photochemical decay of peroxides in the tropical upper troposphere: Methyl iodide as a tracer of marine convection, J. Geophys. Res., 104, 5717-5724, 1999.
6. Crawford, J., et al., Assessment of upper tropospheric HO_X sources over the tropical Pacific based on NASA GTE/PEM data: Net effect on HO_X and other photochemical parameters, J. Geophys. Res., 104, 16255-16273, 1999.
7. Crutzen, PJ., The role of NO and NO₂ in the chemistry of the troposphere and stratosphere, Ann. Rev. Earth Planet. Sci., 7, 443-472, 1979.
8. Ehhalt, D., et al., The chemistry of the hydroxyl radical in the troposphere, Proc. Royal Soc. Ed., 97B, 17-34, 1991.
9. Eisele, F.L., et al., Measurements and steady-state calculations of OH concentrations at Mauna Loa Observatory, J. Geophys. Res., 101, 14665-14679, 1996.
10. Faloona, I., et al., Observations of HO_X and its relationship with NO_X in the upper troposphere during SONEX, J. Geophys. Res., 105, 3771-3783, 2000.
11. Frost, G.J., et al., Photochemical modeling of OH levels during the First Aerosol Characterization Experiment (ACE I), J. Geophys. Res., 104, 16041-16052, 1999.
12. Jaeglé, L., et al., Observations of OH and HO₂ in the upper troposphere suggest a strong source from convective injection of peroxides, Geophys. Res., Lett., 24, 3181-3188, 1997.
13. Jaeglé, L., et al., Ozone production in the upper troposphere and the influence of aircraft during SONEX: Approach of NO_X-saturated conditions, Geophys. Res. Lett., 26, 3081-3084, 1999.
14. Jaeglé, L., et al., Photochemistry of HO_X in the upper troposphere at northern midlatitudes, J. Geophys. Res., 2000a.
15. Jaeglé, L., HO_X chemistry in the upper troposphere: Where do we stand?, IGACtivities Newsletter, 19, 5-7, 2000b.
16. Hard, T., et al., Diurnal HO₂ cycles at clean air and urban sites in the troposphere, J. Geophys. Res., 97, 9785-9794, 1992.
17. Kanaya, Y., et al., Nighttime observations of the HO₂ radical by an LIF instrument at Oki Island, Japan, and its possible origins, Geophys. Res. Lett., 26, 2179-2182, 1999.
18. Logan, J., et al., Tropospheric chemistry: A global perspective, J. Geophys. Res., 86, 7210-7254, 1981.
19. Levy, H., II, Normal atmosphere: Large radical and formaldehyde concentrations predicted, Science, 173, 141-143, 1971.
20. Mount, G.H., and E.J. Williams, An overview of he Tropospheric OH Photochemistry Experiment, Fritz Peak / Idaho Hill, Colorado, fall 1993, J. Geophys. Res., 102, 6171-6186, 1997.
21. Müller, J.-F., and G. Brasseur, Sources of upper tropospheric HO_X: A three-dimensional study, J. Geophys. Res., 104, 1705-1715, 1999.
22. Paulson, S.E., and J.J. Orlando, The reactions of ozone with alkenes: An important source of HO_X in the boundary layer, Geophys. Res. Lett., 23, 3727-3730, 1996.
23. Plass-Dülmer, C, et al., POPCORN: A field study of photochemistry in North-East Germany, J. Atmos. Chem., 31, 5-31, 1998.
24. Prather, M.J., and D.J. Jacob, A persistent imbalance in HO_X and NO_X photochemistry of the upper troposphere driven by deep tropical convection, Geophys. Res. Lett., 24, 3189-3192, 1997.
25. Schultz, M.G., et al., On the origin of tropospheric ozone and NO_X over the tropical South Pacific, J. Geophys. Res., 104, 5829-5843, 1999.
26. Singh, H.B., et al., High concentrations and chemical fate of oxygenated hydrocarbons in the global troposphere, Nature, 378, 50-54, 1995.
27. Trainer, M., et al., Impact of Natural Hydrocarbons on Hydroxyl and Peroxyl Radicals at a Remote Site, J. Geophys. Res., 92, 11879-11894, 1987.
28. Wennberg, P.O., et al., Hydrogen radicals, nitrogen, radicals, and the production of O₃ in the upper troposphere, Science, 279, 49-53, 1998.