Modeling of Atmospheric Chemistry

by Guy P Brasseur

  1920s: Dobson develops a spectrophotometer that remains one of the most accurate instruments to measure the ozone column.

  1929: Götz uses Dobson’s instrument in Spitzbergen to infer the vertical profile of ozone.

  Box 3.2 Figure 1 Christian F. Schoenbein, Gordon M. B. Dobson (from the Royal Meteorological Society), Harold Johnston (courtesy of Denis Galloway, UC Berkeley), Joseph Farman in his office, late 1980s (courtesy of the British Library).

  Photochemical Theories

  1930: Sydney Chapman presents the first photochemical theory of ozone formation and destruction.

  1950s: Arie Haagen-Smit discovers that urban ozone is formed from the by-products of fuel combustion.

  1950: David Bates and Marcel Nicolet show that hydrogen radicals produced by photolysis of water vapor destroy ozone efficiently in the mesosphere.

  1964: John Hampson demonstrates the importance of hydrogen radicals for the chemistry of ozone in the stratosphere.

  1970: Paul Crutzen shows that the largest ozone destruction mechanism in the stratosphere is due to a catalytic cycle involving nitrogen oxides, thus reconciling theory and observations of stratospheric ozone abundances.

  1971: Harold Johnston suggests that the nitrogen oxides released by a planned fleet of high-altitude supersonic aircraft could destroy considerable amounts of ozone in the stratosphere.

  1974: Richard Stolarski and Ralph Cicerone report that a similar cycle with chlorine atoms could also efficiently destroy ozone. Steven Wofsy shows that bromine atoms could also catalytically destroy ozone.

  1974: Mario Molina and Sherwood Rowland establish that the major source of stratospheric chlorine is provided by industrially manufactured chlorofluorocarbons.

  Polar Ozone

  1985: Joseph Farman and colleagues from the British Antarctic Survey observe very low ozone columns at the Antarctic station of Halley Bay during the austral spring, confirming earlier observations by S. Chubachi at the polar station of Syowa.

  1986: Susan Solomon (NOAA) shows that chlorine activation on the surface of particles in polar stratospheric clouds explains the presence of a springtime ozone hole in the Antarctic.

  1987: Luisa and Mario Molina show that formation and photolysis of the ClO dimer can account for most of the springtime ozone loss in Antarctica.

  1995: Crutzen, Molina, and Rowland are awarded the Nobel Prize in Chemistry for their discoveries on the chemistry of ozone.

  3.3 Hydrogen Oxide Radicals

  The importance of hydrogen oxide radicals for atmospheric chemistry was first recognized in studies of the middle atmosphere in the 1950s. The hydroxy radical OH is produced by reaction of water vapor with the electronically excited oxygen atom O(1D) originating from photolysis of ozone:

  O (1D)+H2O → OH + OH

  (3.7)

  Hydroxyl reacts with ozone to produce the hydroperoxy radical HO2, which goes on to react with ozone and return OH, leading to a catalytic cycle for ozone loss:

  O3 + OH → HO2 + O2

  (3.8)

  O3 + HO2 → OH + 2 O2

  (3.9)

  Hydrogen oxide radicals can also react with the oxygen atom O, producing the hydrogen atom and leading to additional catalytic cycles for odd oxygen (and hence ozone) loss. We define the hydrogen oxides radical family as HOx ≡ H + OH + HO2 and refer to the associated catalytic cycles destroying odd oxygen as HOx-catalyzed ozone loss.

  Hydroxyl is of most interest in atmospheric chemistry for its role as a strong oxidant. This role came to the fore in the 1970s with the realization that sufficient solar radiation in the 300–320 nm wavelength region penetrates the troposphere to produce O(1D). The resulting OH is the main agent for oxidizing reduced gases emitted from the surface. The most important reduced gases on a global scale are carbon monoxide (CO) and methane (CH4), which have large emission fluxes. Oxidation of CO proceeds by:

  (3.10)

  where O2 above the reaction sign indicates a species that participates in the overall reaction but does not limit the kinetics. In this case, H produced by conversion of CO to CO2 reacts rapidly with O2 to produce HO2. Oxidation of methane proceeds by:

  (3.11)

  where CH3O2 is the methylperoxy radical and should be viewed as an additional component of HOx since it goes on to cycle with the other components. This chemistry will be discussed in Section 3.5. The atmospheric lifetimes of CO and methane against oxidation by OH are two months and ten years, respectively.

  Conversion between HOx radicals takes place sufficiently rapidly that photochemical equilibrium can be assumed in the daytime (except in the upper mesosphere and thermosphere, where collisions are infrequent). Production of HOx is mostly by reaction (3.7). Loss of HOx can take place by various pathways, the dominant one in the troposphere being the formation of hydrogen peroxide (H2O2):

  HO2 + HO2 + M → H2O2 + M

  (3.12)

  This loss is terminal if H2O2 is removed by deposition or is converted to water, as by reaction with OH:

  H2O2 + OH → HO2 + H2O

  (3.13)

  However, it is temporary if H2O2 is instead recycled to HOx radicals by photolysis:

  H2O2 + hv → OH + OH

  (3.14)

  Thus H2O2 and other peroxides should be viewed as reservoirs for HOx. It can be convenient to define a hydrogen oxides family HOy ≡ HOx+ peroxides to account for the exchange between HOx and its reservoirs.

  3.4 Nitrogen Oxide Radicals

  Nitrogen oxide radicals (NOx ≡ NO + NO2) are of central importance for atmospheric chemistry. They are emitted to the troposphere by combustion, lightning, and microbial processes in soils. The largest source is combustion. A typical combustor mixes fuel and air at very high flame temperatures (about 2000 K). At these temperatures, O2 from the air thermally dissociates to O atoms, which react with N2 from the air to drive a catalytic cycle for formation of nitric oxide (NO) known as the Zel’dovich mechanism:

  N2 + O → NO + N

  (3.15)

  N + O2 → NO + O

  (3.16)

  The NOx generated in this manner is called thermal NOx. The Zel’dovich mechanism is not efficient at low flame temperatures such as from open fires, but production of NO still takes place in those cases by oxidation of nitrogen present in the fuel. The NOx generated in that manner is called fuel NOx.

  In the stratosphere, the main source of NOx is the oxidation by O(1D) of nitrous oxide (N2O), a long-lived gas emitted by microbial activity in soils:

  O (1D) + N2O → NO + NO

  (3.17)

  A dominant sink for NO in both the troposphere and stratosphere is reaction with ozone to form nitrogen dioxide (NO2):

  NO + O3 → NO2 + O2

  (3.18)

  In the daytime, NO2 is photolyzed on a timescale of a minute to return NO:

  (3.19)

  The reaction cycle (3.18) + (3.19) has no effect on atmospheric composition; it is called a null cycle. However, it forces photochemical equilibrium between NO and NO2 in the daytime.

  Alternate reaction cycles for NO and NO2 lead to production or loss of ozone. In the stratosphere, NO2 can be converted back to NO by reaction with atomic oxygen:

  NO2 + O → NO + O2

  (3.20)

  The reaction cycle (3.18) + (3.20) is a major catalytic sink for stratospheric ozone. The rate of ozone destruction is set by the rate of reaction (3.20), which competes with (3.19). Reaction (3.20) is called the rate-limiting step for ozone loss. It is unimportant in the troposphere, where oxygen atom concentrations are low. In the troposphere, however, ozone concentrations are sufficiently low that peroxy radicals can compete for reaction with NO. The reaction

  HO2 + NO → OH + NO2

  (3.21)

  followed by (3.19) provides an important source of tropospheric ozone. The rate of ozone production is determined by the rate of (3.21) as the rate-limiting step. Similar mechanisms in which NO reacts with methyl peroxy (CH3O2) and other organic peroxy radicals (RO2) lead to additional oz
one production. This is discussed in Section 3.5.

  Loss of NOx takes place on a timescale of one day. It involves primarily the conversion of NO2 to nitric acid (HNO3). In the daytime, this conversion is by oxidation by OH:

  NO2 + OH + M → HNO3 + M

  (3.22)

  At night it takes place by oxidation by ozone, forming dinitrogen pentoxide (N2O5) that hydrolyzes to HNO3 in aqueous aerosol particles:

  NO2 + O3 → NO3 + O2

  (3.23)

  NO2 + NO3 + M → N2O5 + M

  (3.24)

  (3.25)

  In the daytime, the nitrate radical (NO3) has a lifetime of less than a minute against photolysis back to NO2. Thus loss of NOx by (3.23)–(3.25) can operate only at night.

  In the troposphere, the dominant sink of HNO3 is deposition, including scavenging by precipitation (wet deposition) and direct reaction at the surface (dry deposition). In the stratosphere, however, deposition does not take place and HNO3 is instead recycled to NOx on a timescale of weeks through photolysis and reaction with OH:

  HNO3 + hv → NO2 + OH

  (3.26)

  HNO3 + OH → NO3 + H2O

  (3.27)

  NO3 + hv → NO2 + O

  (3.28)

  Thus HNO3 serves as a reservoir for NOx and the concentration of NOx is determined by photochemical equilibrium with HNO3. Similarly to HOx and HOy, it is useful to define a chemical family NOy as the sum of NOx and its reservoirs. Loss of NOy from the stratosphere is mainly by transport to the troposphere followed by HNO3 deposition.

  The short lifetime of NOx in the troposphere, combined with the rapid removal of HNO3 by deposition, results in strong concentration gradients between combustion source regions and the remote oceans (Figure 3.4). However, a low background NOx concentration is sustained in the remote troposphere and plays a critical role for ozone and OH generation following (3.21). A major source of this background NOx is the long-range transport and decomposition of the peroxyacetylnitrate reservoir (PAN, formula CH3C (O) O2NO2). PAN is produced by reaction of NO2 with peroxyacetyl radicals CH3C (O) O2 originating from the oxidation of various organic species (Section 3.5):

  CH3C (O) O2+NO2 + M → CH3C (O) O2NO2 + M

  (3.29)

  Figure 3.4 Global annual mean distribution of the tropospheric NO2 column [1016 molecules cm–2] observed from 2005 to 2008 by the Ozone Monitoring Instrument (OMI) on the National Aeronautics and Space Administration (NASA) Aura satellite.

  Source: Bas Mijling, Folkert Boersma, and Ronald van der A (KNMI).

  The main sink of PAN is thermal decomposition to the original reactants. The lifetime of PAN is one hour at 295 K but months at 250 K. PAN produced in NOx source regions and lifted to high altitudes can be transported on global scales, eventually decomposing to deliver NOx to the remote troposphere. Other organic nitrates can be similarly produced from the oxidation of organic species in the presence of NOx, and this is discussed in Section 3.5. PAN is the most important because of its high yield and its wide range of atmospheric lifetimes enabling both long-range transport in cold air masses and quick release of NOx when these air masses warm up (as from subsidence).

  3.5 Volatile Organic Compounds and Carbon Monoxide

  Atmospheric chemists refer to the ensemble of organic species present in the gas phase as volatile organic compounds (VOCs). VOCs are emitted by biogenic, combustion, and industrial processes, mainly as hydrocarbons (CxHy). Atmospheric oxidation of hydrocarbons produces a cascade of oxygenated VOC species eventually leading to CO and CO2. The longest-lived VOC is methane, with a lifetime of ten years against oxidation by OH. Other VOCs have considerably shorter lifetimes. Isoprene, the dominant VOC emitted by vegetation, has a lifetime of only about one hour during summer daytime. Thus the VOCs are largely confined to the troposphere, and the short-lived non-methane VOCs influence mostly their region of emission. VOC chemistry involves a succession of steps as carbon is oxidized from its most reduced state –4 (hydrocarbons) to its most oxidized state +4 (CO2). This chemistry is responsible for much of the complexity in chemical mechanisms of the troposphere. Only general rules will be presented here. Box 3.3 gives nomenclature for major VOCs.

  Box 3.3 Nomenclature of Major Atmospheric VOCs (Common Names in Parentheses)

  Alkanes (CnH2n+2)

  CH4

  methane

  C2H6 or CH3-CH3

  ethane

  C3H8 or CH3-CH2-CH3

  propane

  C4H10 (2 isomers)

  butane

  C5H12 (3 isomers)

  pentane

  Alkenes (CnH2n)

  C2H4 or CH2=CH2

  ethene (ethylene)

  C3H6 or CH2=CHCH3

  propene

  Alkynes (CnH2n–2)

  C2H2 or CH≡CH

  ethyne (acetylene)

  Aromatics (benzene ring)

  C6H6

  benzene

  C6H5CH3

  methylbenzene (toluene)

  C6H4(CH3)2 (3 isomers)

  dimethylbenzene (xylene)

  Dienes (two C=C bonds)

  C5H8 or CH2=C(CH3)CH=CH2

  2-methyl-1,3-butadiene (isoprene)

  Terpenes (multiple isoprene units)

  C10H16 (many isomers)

  monoterpenes (α-pinene, β-pinene…)

  C15H24 (many isomers)

  sesquiterpenes (β-caryophyllene, α-humulene…)

  Alcohols (hydroxy function –OH)

  CH3OH

  methanol

  CH3CH2OH

  ethanol

  Aldehydes (terminal carbonyl function –CHO)

  CH2O

  methanal (formaldehyde)

  CH3CHO

  ethanal (acetaldehyde)

  Ketones (internal carbonyl function –C(O)–)

  CH3COCH3

  propanone (acetone)

  CH3COCH2CH3

  butanone (methylethylketone, MEK)

  Dicarbonyls (two carbonyl functions)

  CHOCHO

  glyoxal

  CH3C(O)CHO

  methylglyoxal

  Carboxylic acids (carboxylic function –C(O)OH)

  HCOOH

  methanoic acid (formic acid)

  CH3COOH

  ethanoic acid (acetic acid)

  Organic peroxides (peroxide function –OO–)

  CH3OOH

  methylhydroperoxide

  Organic nitrates (nitrate function –ONO2)

  CH3ONO2

  methylnitrate

  Peroxyacyl nitrates (peroxyacyl function –C(O)OONO2)

  CH3C(O)OONO2

  nitroethaneperoxoate (peroxyacetylnitrate, PAN)

  The main sink for most VOCs is oxidation by OH. Additional oxidants including ozone, NO3, and halogen atoms can be important for some species. Photolysis is an additional sink for carbonyl and peroxide species. VOCs with relatively low vapor pressures (called semivolatile) can partition into the aerosol and cloud phases with subsequent removal by deposition. They can also directly deposit to surfaces.

  Oxidation of VOCs by OH can take place by abstraction of an H atom, as in the case of methane with (3.11), or by addition at an unsaturated bond as in ethylene:

  (3.30)

  In both cases the oxidation produces an organic radical, R, that subsequently adds oxygen to produce an organic peroxy radical RO2. The RO2 radicals react with NO in a manner analogous to HO2 in (3.21):

  RO2 + NO → RO + NO2

  (3.31)

  This produces ozone from subsequent photolysis of NO2 as discussed in Section 3.4. When NOx concentrations are low, RO2 radicals can react instead with HO2 to form organic peroxides:

  RO2 + HO2 → ROOH + O2

  (3.32)

  This does not produce ozone and instead provides a sink for HOx radicals, analogous to the formation of H2O2 from the self-reaction of HO2 radicals as given by (3.12). Additional minor sinks for RO2 radicals include permutation reactions with other RO2 radicals and isomerization. An atmosphere where RO2 radicals react do
minantly with NO is said to be in the high-NOx regime; oxidation of VOCs in that regime is a source of ozone. An atmosphere where RO2 radicals do not react dominantly with NO is said to be in the low-NOx regime; VOC oxidation in that regime tends to scavenge HOx radicals.

  The oxy radicals RO produced by (3.31) can be oxidized by O2, decompose, or isomerize. The organic peroxides ROOH produced by (3.32) can be oxidized by OH or photolyze. These reactions generally produce carbonyl species including aldehydes and ketones (Box 3.3). The carbonyls further react with OH or photolyze, leading to production of multifunctional compounds and to breakage of chains producing simpler compounds. Successive oxidation steps ultimately lead to CO and on to CO2, where carbon is in its highest oxidation state. Figure 3.5 gives a general schematic of the oxidation cascade.

  Figure 3.5 Generic oxidation scheme for a hydrocarbon RH.

  We described in Section 3.4 the formation of PAN by reaction (3.29) as a reservoir for NOx in the troposphere. More generally, RO2 radicals can react with NO2 to form peroxynitrates,