RO2 + NO2 + M ⇄ RO2NO2 + M
(3.33)
but most of these peroxynitrates have lifetimes of less than a minute against thermal decomposition and therefore are not effective NOx reservoirs. Peroxyacylnitrates RC(O)OONO2 are an exception and PAN is the most abundant of these peroxyacylnitrates. The peroxyacetyl radical CH3C(O)OO that serves as precursor of PAN originates mainly from oxidation of acetaldehyde and from photolysis of acetone and methylglyoxal:
(3.34)
(3.35)
(3.36)
Another class of organic nitrates is produced as a minor branch in the oxidation of RO2 by NO:
RO2 + NO + M → RONO2 + M
(3.37)
These tend to be much more stable than the peroxynitrates. They may undergo further oxidation by OH, photolysis, fractionation into aerosol, or deposition. The organic nitrate yield by (3.37) in competition with (3.31) generally increases with the size of R. Organic nitrate formation can be a significant sink for NOx in regions with large biogenic VOC emissions.
Carbon monoxide is a general intermediate in the oxidation of VOCs to CO2. It is also directly emitted by incomplete combustion, with particularly large emissions from open fires where the combustion process is uncontrolled and inefficient. Carbon monoxide has a mean lifetime of two months against oxidation by OH, its main sink. Figure 3.6 shows the global distribution of CO observed by satellite. Concentrations are highest over tropical regions during the burning season, and are also relatively high over northern mid-latitude continents. The lifetime of CO is sufficiently long to allow transport on intercontinental scales (Chapter 2), making CO a useful tracer for long-range transport of combustion plumes. Figure 3.6 shows that the northern hemispheric background of CO is elevated relative to the south because of combustion influence, and has a seasonality driven by the sink from photochemical oxidation by OH. The southern hemispheric background is mostly contributed by the global source from the oxidation of methane.
Figure 3.6 Global distribution of the tropospheric carbon monoxide (CO) column in March (a) and September (b) 2012, as observed by the MOPITT instrument aboard the National Aeronautics and Space Administration (NASA) Terra satellite.
Courtesy of David Edwards, National Center for Atmospheric Research (NCAR).
3.6 Tropospheric Ozone
Ozone is produced in the troposphere by photochemical oxidation of VOCs and CO in the presence of NOx, In the simplest case of CO, this involves a sequence of reactions (3.10), (3.21), and (3.19):
HO2 + NO → OH + NO2
Net : CO + 2O2 → CO2 + O3
In the case of VOCs, the reaction sequence is similar but with RO2 radicals reacting with NO following (3.31). The HOx and NOx radicals serve as catalysts for the oxidation of VOCs and CO by O2, and ozone is produced in the process.
The above mechanism provides the dominant source of ozone in the troposphere (Table 3.1). Transport from the stratosphere is an additional minor source. The dominant sink of tropospheric ozone is photochemical loss, including photolysis in the presence of water vapor and reactions with HOx radicals:
O3 + hv (λ < 320 nm) → O (1D) + O2
O (1D) + H2O → OH + OH
O3 + OH → HO2 + O2
There is also a minor sink from deposition to the surface. The lifetime of ozone ranges from a few days in the boundary layer to months in the dry upper troposphere. This difference in lifetime results in a general pattern of net ozone production in the upper troposphere, balanced by net loss in the lower troposphere, driving a gradient of increasing ozone concentrations with altitude.
Table 3.1 Global present-day budget of tropospheric ozone
Best estimate, Tg O3 a–1
Sources
Tropospheric chemical production 4500
Transport from stratosphere 500
Sinks
Tropospheric chemical loss 4000
Deposition 1000
Estimates based on Wu et al. (2007)
The rate of ozone production depends on the supply of NOx, VOCs, and CO in a manner controlled by the cycling of HOx radicals and competition with HOx sinks. In most of the troposphere, the dominant HOx sink is the conversion of peroxy radicals to peroxides following (3.12) and (3.32). In that regime OH radicals mainly react with VOCs or CO, and whether ozone production takes place depends on competition for the peroxy radicals between reaction with NO (producing ozone) and production of peroxides. Thus the ozone production rate increases linearly with the NOx concentration but does not depend on the concentrations of VOCs and CO. This is called the NOx-limited regime.
A different regime for ozone production applies when UV radiation is low (as in winter) or when NOx concentrations are very high. In that case the dominant HOx sink becomes the formation of nitric acid by reaction (3.22). There is no longer competition for peroxy radicals, because the low UV radiation suppresses peroxide formation by (3.12), which has a quadratic dependence on peroxy radical concentrations. Instead, the competition is for OH between reactions with VOCs and CO (producing ozone) and reaction with NO2 to produce nitric acid. Thus the ozone production rate increases linearly with the VOCs and CO concentrations but inversely with the NOx concentration. This is called the VOC-limited or NOx-saturated regime.
Figure 3.7 shows a simple box model calculation of ozone isopleths (lines of constant mixing ratios) calculated as a function of NOx and VOC concentrations using a standard chemical mechanism. The NOx- and VOC-limited regimes identified on the diagram illustrate the different dependences of ozone production on NOx and VOCs. The nonlinear dependences and transitions between regimes are readily apparent.
Figure 3.7 Ozone isopleths diagram showing the dependence of ozone concentration on NOx and VOC concentrations for a simple box model calculation. NOx- and VOC-limited regimes are indicated. This representation is often referred to as the Empirical Kinetic Modeling Approach or EKMA diagram. The NOx-limited regions are typical of locations downwind of urban and suburban areas, whereas the VOC-limited regions are typical of highly polluted urban areas.
Source: National Research Council (1991).
Figure 3.8 shows the global distribution of tropospheric ozone columns observed from satellite. The July data feature elevated ozone at northern mid-latitudes, reflecting high NOx emissions and strong UV radiation. The lifetime of ozone is sufficiently long to allow transport on intercontinental scales. The October data feature elevated ozone downwind of South America and Africa, reflecting NOx emissions from biomass burning during that time of year (end of dry season in the southern tropics). Ozone concentrations peak downwind of the source continents due to sustained production in the continental plume.
Figure 3.8 Tropospheric ozone column (Dobson units) derived from satellite observations by subtracting the stratospheric column measured by the Microwave Limb Sounder (MLS) from the total column measured by the Ozone Monitoring Instrument (OMI). Monthly mean values are shown for October 2004 (top) and July 2005 (bottom). Plumes are streaming from Africa and South America (ozone produced by precursors released from biomass burning in the tropics during the dry season) in October and from North America, Europe and China (summertime ozone formation from anthropogenic precursors) in July.
From The National Aeronautics and Space Administration (NASA).
3.7 Halogen Radicals
Natural sources of atmospheric halogens include the marine biosphere, sea salt, and volcanoes. The marine biosphere emits a wide range of organohalogen gases, the simplest being methyl chloride, bromide, and iodide (CH3Cl, CH3Br, CH3I). Since the 1950s, industrial sources have released to the atmosphere a number of long-lived organohalogens including chlorofluorocarbons and bromine-containing halons (anthropogenic sources of iodine are thought to be negligibly small). These long-lived compounds are transported to the stratosphere, where strong radiation triggers their photolysis to release halogen atoms, which destroy ozone through the catalytic cycle:
X + O3 → XO + O2
(3.38)
XO + O → X
+ O2
(3.39)
where X ≡ F, Cl, Br, or I. One commonly defines the radical family XOx ≡ X + XO as the catalyst for ozone loss. Termination of the catalytic loss cycle requires conversion of the radicals to non-radical reservoirs including X2, XONO2, HOX, and HX:
XO + XO → X2 + O2
(3.40)
XO + NO2 + M → XONO2 + M
(3.41)
XO + HO2 → HOX + O2
(3.42)
X + CH4 → HX + CH3
(3.43)
There are also significant cross-halogen reactions, such as between XO and another halogen oxide YO:
XO + YO → XY + O2
(3.44)
These non-radical reservoirs can be recycled to the radicals by photolysis (for X2, XY, XONO2, HOX), thermolysis (XONO2), hydrolysis (XONO2), or reaction with OH (HX). In the troposphere they can also be removed by deposition, representing a terminal sink. One generally refers to the chemical family Xy (total inorganic X) as the sum of XOx and the inorganic non-radical reservoirs (for example, Bry ≡ Br + BrO+ inorganic non-radical reservoirs).
Halogen radicals are of interest as sinks of stratospheric and tropospheric ozone, and as oxidants for various species. Their concentrations are determined by the abundance of Xy and by the partitioning of Xy between radical and non-radical forms, i.e., the XOx/Xy ratio. A major factor in the efficacy of halogen radical chemistry is the stability of the non-radical reservoir HX against oxidation by OH. This stability greatly decreases in the order HF > HCl > HBr > HI. Thus Fy in the atmosphere is almost entirely present as HF and there is no significant fluorine radical chemistry. At the other end, iodine in the atmosphere is present principally in radical form. Iodine and bromine are particularly efficient at destroying ozone but have much weaker sources than chlorine.
Chlorine radical concentrations are particularly high in the Antarctic lower stratosphere in spring, where they drive near-total ozone depletion (Figure 3.9). This involves unique chemistry taking place on polar stratospheric cloud (PSC) particles that form under the very cold conditions of the Antarctic stratosphere in winter and early spring. The PSCs consist of liquid supercooled ternary solutions (H2SO4–HNO3–H2O or STS), solid nitric acid trihydrate (HNO3–3H2O or NAT), and ice crystals. Their formation takes place below a temperature threshold of about 189 K for ice particles, 192 K for STS, and 196 K for NAT. Polar stratospheric cloud surfaces enable fast conversion of non-radical chlorine reservoirs to chlorine radicals by
(3.45)
Cl2 + hv → 2Cl
(3.46)
Cl + O3 → ClO + O2
(3.47)
This converts most of Cly in Antarctic spring to ClO, as shown in Figure 3.9. High ClO concentrations are found in a ring around Antarctica in winter, filling in over the pole in spring, because a minimum of radiation is needed to photolyze Cl2 and other weakly bound chlorine non-radical reservoirs.
Figure 3.9 ClO and ozone concentrations in the Antarctic lower stratosphere in winter–spring 1992. Data from the Microwave Limb Sounder (MLS) on the National Aeronautics and Space Administration (NASA) Upper Atmosphere Research satellite (UARS).
Source: Waters et al. (1993).
High concentrations of ClO prime the Antarctic springtime stratosphere for rapid ozone destruction. However, the O atom concentration remains very low because of weak solar radiation, so that the ClOx-catalyzed radical mechanism involving (3.38) + (3.39) is very slow. A different mechanism operates involving formation of the ClO dimer followed by photolysis:
ClO + ClO + M → Cl2O2 + M
(3.48)
Cl2O2 + hv → ClOO + Cl
(3.49)
ClOO + M → Cl + O2 + M
(3.50)
The Cl atoms react again with ozone by (3.47), yielding a catalytic cycle for ozone destruction. The rate-limiting step for this cycle is (3.48), which is quadratic in ClO concentrations. Near-total ozone depletion can thus take place in a matter of weeks. A similar mechanism takes place in the Arctic stratosphere in winter–spring but is much less pronounced because Arctic temperatures are on average 10 K warmer (limiting PSC formation) and the polar vortex is considerably more perturbed by planetary-scale waves as discussed in Section 2.11.
3.8 Sulfur Species
Sulfate is a major component of atmospheric aerosol and an important contributor (as sulfuric acid) to acid deposition. It is produced in the atmosphere by oxidation of sulfur gases emitted from anthropogenic sources (mainly combustion and metallurgy), the marine biosphere, and volcanoes. Anthropogenic and volcanic emissions are mainly as SO2. Biogenic emission is mostly as dimethylsulfide (CH3)2S, commonly called DMS. There is also a small anthropogenic and marine source of carbonyl sulfide (COS), which is of interest because COS has a long enough atmospheric lifetime to be transported to the stratosphere, where it provides a background source of sulfate aerosol. Figure 3.10 shows the global distribution of SO2 columns measured by satellite. The anthropogenic source is particularly large over China, reflecting coal combustion with limited emission controls.
Figure 3.10 Annual mean SO2 vertical columns from the Scanning Imaging Absorption spectrometer for Atmospheric chartography (SCIAMACHY) satellite instrument for 2006. The South Atlantic Anomaly (SAA) is subject to excessive measurement noise.
From Lee et al. (2009).
The oxidation state of sulfur ranges from –2 in DMS to +6 in sulfate. The DMS is oxidized in the marine boundary layer by OH, NO3, and halogen radicals to produce a cascade of sulfur species eventually leading to SO2 as a major product. The COS is converted to SO2 following oxidation by OH, and also in the stratosphere following oxidation by O(1D) and photolysis. And SO2 is oxidized by OH to produce sulfuric acid (H2SO4) in the gas phase:
SO2 + OH + M → HSO3 + M
(3.51)
HSO3 + O2 → SO3 + HO2
(3.52)
SO2 + OH + M → HSO3 + M
(3.53)
Sulfuric acid has an extremely low vapor pressure over H2SO4–H2O solutions and therefore condenses immediately, either on existing particles or by forming new particles.
The lifetime of SO2 against gas-phase oxidation by OH is of the order of a week. In the lower troposphere, more rapid oxidation of SO2 can take place in the aqueous phase in clouds. This involves dissolution of SO2 into cloud water, followed by acid–base dissociation of sulfurous acid (SO2•H2O) to bisulfite (HSO3–) and sulfite (SO32–):
SO2(g) ⇌ SO2 · H2O (aq)
(3.54)
SO2 · H2O (aq) ⇌ HSO3− + H+(pK1 = 1.9)
(3.55)
HSO3− ⇌ SO32− + H+(pK2 = 7.2)
(3.56)
The ions can then be oxidized rapidly in the aqueous-phase. Major oxidants are H2O2 and ozone, both dissolved from the gas phase:
HSO3− + H2O2(aq) + H+ → SO42− + H2O + 2H+
(3.57)
SO32− + O3(aq) → SO42− + O2
(3.58)
Oxidation by H2O2 is acid-catalyzed and therefore remains fast even as the cloud droplets become acidified. Oxidation of SO32– by O3(aq) is extremely fast but is limited by the supply of SO32– and shuts down as cloud droplets are acidified below pH 5. Other aqueous-phase SO2 oxidants can be important in winter or in highly polluted conditions when H2O2 concentrations are low. Competition between gas-phase and aqueous-phase oxidation of SO2 has important implications for aerosol formation because gas-phase H2SO4 is a major precursor for nucleation of new aerosol particles.
3.9 Aerosol Particles
The atmosphere contains suspended condensed particles ranging in size from ~0.001 μm (molecular cluster) to ~100 μm (small raindrop). Atmospheric chemists commonly refer to the ensemble of particles of a certain type as an aerosol (for example, sulfate aerosol) and to an ensemble of particles of different types as aerosols. The ensemble of particles in the atmosphere is often called the atmospheric aerosol. This terminology has force of usage but departs from the dictionary definition of an aerosol as a suspension of dispersed particles in a gas (by that defini
tion, the atmosphere itself would be an aerosol). Referring to aerosol particles removes the ambiguity. The air quality community refers to aerosols as particulate matter and uses the acronym PM to denote the aerosol mass concentration per unit volume of air. For example, PM2.5 denotes the mass concentration [μg m–3] of particles less than 2.5 μm in diameter. Aerosols are removed efficiently by precipitation and thus have atmospheric lifetimes of the order of a week, leading to large regional gradients.
3.9.1 Size Distribution
An aerosol particle is characterized by its shape, size, phase(s), and chemical composition. Liquid particles are spherical but solid particles can be of any shape. There is a continuous distribution of particle sizes. For the purpose of characterizing this distribution the particles are conventionally assumed to be spherical. Such an assumption is obviously incorrect for solid particles but can be viewed as an operational approximation where the solid particles behave as equivalent spheres for the purpose of sizing measurements or microphysical dynamics. An aerosol composed of particles of a single size is called monodisperse, while an aerosol composed of particles of multiple sizes is called polydisperse. Aerosols produced in the laboratory under carefully controlled conditions can be close to monodisperse. Aerosols in the atmosphere are polydisperse.
Modeling of Atmospheric Chemistry Page 12