8.1.1 Sulfur Dioxide

Sulfur dioxide (SO₂) is a primary pollutant. SO₂ is the main pollutant of air pollution episodes such as that of 1952 in London. It is regulated because it leads to respiratory problems. As a primary pollutant, its impacts occur near its sources. In North America and Europe, those impacts are now mostly limited to some industrial sites and maritime traffic, because sulfur content in fuels used by road traffic is now regulated.

SO₂ is oxidized to sulfuric acid in the atmosphere (see Chapter 10). Sulfuric acid contributes to fine particle air pollution, because its low volatility implies that it is preferentially present in the particulate phase, typically as ammonium salts. In addition, sulfuric acid is an important component of acid rain. SO₂ is slightly soluble in water, and it may be removed from the atmosphere by dry and wet deposition.

8.1.2 Carbon Monoxide

Carbon monoxide (CO) is an inorganic carbonaceous compound, because it does not contain any hydrogen atoms. CO is regulated because it combines with hemoglobin in the blood stream and leads to anoxia (lack of oxygen), when the blood stream cannot carry enough oxygen. CO is a primary pollutant, which is mostly emitted by combustion processes, including internal combustion engines. Historically, CO concentrations were, therefore, high near roadways. CO was the first pollutant to be regulated for road traffic in North America and Europe, using catalytic converters, which convert it to carbon dioxide, CO₂ (see Chapter 2). Consequently, ambient concentrations of CO are now fairly low in these regions.

CO is oxidized slowly to CO₂ in the atmosphere. Its oxidation, which is described in Section 8.3.3, leads also to the formation of ozone, the main gaseous pollutant of photochemical smog. However, the contribution of CO to photochemical air pollution is less than that of other carbonaceous compounds because of its low chemical reactivity and the fact that its emissions are currently regulated.

8.1.3 Ozone and Gaseous Photochemical Oxidants

Photochemical pollution is generally called “photochemical smog.” Smog is the contraction of smoke and fog, because its appearance falls between these two phenomena. Photochemical pollution was initially identified in the Los Angeles Basin in the 1950s. Heavy pollution was present in that region of southern California, because of a large number of anthropogenic pollution sources, such as road traffic, fossil-fuel fired power plants, refineries, and other industrial sources, as well as meteorological conditions that were conducive to air pollution (strong sunlight and low atmospheric dispersion). Arie Haagen-Smit, a biochemistry professor at the California Institute of Technology (Caltech), was the first to identify the processes that lead to the formation of photochemical air pollution, and in particular ozone (O₃), which is its main gaseous pollutant. He showed that photochemical air pollution results from atmospheric chemical reactions occurring among precursor gases that include nitrogen oxides (NO_x) and volatile organic compounds (VOC), in the presence of sunlight. Hence, the adjective “photochemical” was attributed to that form of air pollution, since photolytic reactions induced by sunlight (see Chapter 7) initiate the set of photochemical reactions that lead to ozone formation.

Ozone is not emitted in the atmosphere and is, therefore, a secondary air pollutant. Its precursors are NO_x, VOC, and, to a lesser extent, CO. Although O₃ is not the only photochemical oxidant, it is the main gaseous constituent of photochemical smog in terms of ambient concentrations. Therefore, O₃ is targeted for regulations pertaining to the gaseous fraction of photochemical smog. Since O₃ is produced by photochemical reactions, its concentrations are highest during spring and summer.

8.1.4 Nitrogen Oxides

NO_x include by definition nitric oxide (NO) and nitrogen dioxide (NO₂). Their emissions result mostly from combustion processes. These two compounds represent in terms of ambient concentrations the majority of nitrogen oxides present in the urban atmosphere. However, there are other nitrogen oxides, such as nitrogen protoxide (N₂O), which is a greenhouse gas, nitric acid (HNO₃), nitrogen pentoxide (N₂O₅), the nitrate radical (NO₃), and a large number of organic nitrates. By convention, NO_y represent all nitrogen oxides, with the exception of N₂O, and NO_z represent the difference between NO_y and NO_x. Thus, in summary:

NOx=NO+NO2NOy=NO+NO2+HNO3+N2O5+NO3+organic nitratesNOz=HNO3+N2O5+NO3+organic nitrates

NO₂ is the only nitrogen oxide that is regulated in terms of its ambient concentrations. It leads to respiratory problems. NO₂ is both a primary and a secondary pollutant, because (1) it constitutes a fraction of NO_x emissions and (2) it is a product of the atmospheric oxidation of NO.

NO_x are rapidly oxidized in the atmosphere. They are precursors of a large number of secondary pollutants, such as ozone, nitric acid, and organic nitrates. Nitric acid may contribute significantly to secondary particulate matter formation. It is also a major constituent of acid rain. Some organic nitrates contribute to the secondary fraction of organic particulate matter. In addition, inorganic and organic nitrates play a major role in the eutrophication of ecosystems.

8.1.5 Volatile and Semi-volatile Organic Compounds

Volatile organic compounds (VOC) and semi-volatile organic compounds (SVOC) are mostly emitted from combustion processes and also from the evaporation of liquid fuels and some organic-containing products (e.g., some paints, solvents, and cleaning products). They play a major role in the formation of photochemical smog. VOC are precursors of ozone. SVOC and some VOC are also important precursors of the secondary fraction of fine particulate matter. In addition, some VOC, such as benzene, 1,3-butadiene, and formaldehyde, are carcinogenic. These carcinogenic VOC may be regulated individually as such (as it is the case in Europe for benzene) or may be regulated indirectly via regulatory approaches that target carcinogenic compounds as a whole (as it is the case in the United States).

VOC include mostly alkanes (hydrocarbons with only single bonds, also called paraffins), alkenes (hydrocarbons with one or more double bonds, also called olefins), aromatic compounds (compounds with one or more phenyl rings), and aldehydes (compounds with a carbonyl group, HC=O). Alcohols (compounds with a C-OH group), alkynes (hydrocarbons with a triple bond), ethers (compounds with a C-O-C group), and other organic compounds are typically present in the atmosphere to a lesser extent. However, alcohols are becoming more prominent due to the use of biofuels and their increased use in gasoline.

The term hydrocarbon (HC) refers to organic compounds that contain only carbon and hydrogen atoms (therefore, alkanes, alkenes, alkynes, and some aromatic compounds). We will use hereafter the term VOC to refer to all organic compounds, thus including aldehydes and alcohols. Among the alkanes, methane has a very low atmospheric reactivity (atmospheric lifetime on the order of 10 years) and, therefore, it is typically not included among the precursors of photochemical air pollution. Thus, when referring to VOC that are precursors of photochemical pollution, one should use the term non-methane VOC. However, for the sake of simplicity, we will use the term VOC to mean non-methane VOC hereafter.

SVOC are organic compounds with a saturation vapor pressure that is such that they can be present in the atmosphere in both the gas phase and the particulate phase. They play a major role in the formation of particulate organic compounds and may also be involved in the formation of gaseous air pollutants.

The ultimate chemical fate of VOC and SVOC is CO₂, via atmospheric oxidation. However, organic compounds may be removed from the atmosphere before being converted to CO or CO₂. Their removal may occur either as gaseous or particulate compounds, via dry and wet deposition.

8.1.6 Ammonia

Ammonia (NH₃) is the reduced form of nitrogen in the atmosphere. It is mostly emitted by agricultural activities. NH₃ does not present any adverse health effects at the ambient concentrations usually observed in the atmosphere. However, it contributes to the formation of sulfate and nitrate ammonium salts, which may constitute a significant fraction of fine particulate matter. In addition, NH₃ contributes to nitrogen deposition and, therefore, may lead to the eutrophication of ecosystems.

8.3.1 Oxidants

The reactions leading to the formation of the three main oxidant species of photochemical air pollution, OH, NO₃, and O₃, are described in this section.

As mentioned in Chapter 7, hydroxyl radicals can be formed via the photolysis of ozone:

A chemical kinetic mechanism will also need to take into account the fact that only a fraction of the O(¹D) excited oxygen atoms reacts with water vapor, because most of them lose their excess energy by collision with air molecules (N₂ or O₂) and produce O₃ back by reacting next with O₂. (At 100 % relative humidity and 25 °C, the reaction with air molecules is about five times faster than that with water vapor.)

In addition to this OH formation pathway, there are two other important photolytic reactions that lead to OH formation in the troposphere: (1) the photolysis of hydrogen peroxide (H₂O₂) and (2) the photolysis of nitrous acid (HNO₂):

In addition, as discussed in Section 8.3.3, OH is also formed by reaction of hydroperoxyl radicals, HO₂. They originate mostly from the photolysis of aldehydes. Aldehyde photolysis leads to the production of a hydrogen atom, H, which is then oxidized rapidly by O₂ to form HO₂. Since the OH radical is formed by photolytic reactions, it is present mostly during daytime. There are, however, some formation pathways that do not require photolysis, such as the decomposition of peroxyacetylnitrate (PAN) in presence of NO_x (see the chemistry of PAN in Section 8.3.4) and the oxidation of alkenes by O₃ (see Section 8.3.6). However, these reactions are very limited sources of OH and nighttime OH concentrations are negligible.

The nitrate radical (not to be confused with the nitrate ion, NO₃⁻, which is present in the aqueous phase, see Chapter 10) is formed via the reaction of nitrogen dioxide with ozone:

This radical is rapidly photolyzed:

Its formation does not require any photochemical reaction; therefore, it can be formed either at night or during the day. However, its rapid photolysis diminishes significantly its concentration during daytime. Thus, this oxidant plays a role mostly at night. Since the kinetics of the reaction of O₃ with NO is about 1,000 times faster than that with NO₂, NO concentrations are negligible when NO₃ is present, because NO will have almost entirely been oxidized into NO₂.

Ozone is formed in the stratosphere by photolysis of oxygen molecules (see Chapter 7). Solar radiation that leads to oxygen photolysis is filtered in the stratosphere and is, therefore, unavailable in the troposphere to lead to ozone formation there. However, the photolysis of nitrogen dioxide, which takes place in the visible and the near ultraviolet (UV) range, takes place in the troposphere:

Therefore, ozone formation takes place in the presence of sunlight, i.e., during daytime. However, the lifetime of ozone ranges from several hours to a few days. Thus, its oxidizing power can also occur at night.

In summary, the atmospheric gaseous oxidants are the following:

1. 
   

   – During daytime: OH and O₃

   
   
2. 
   

   – During nighttime: NO₃ and O₃

8.3.2 The Photostationary State of Leighton

Ozone formation is balanced by its destruction by nitric oxide:

This reaction is very fast and can, therefore, be called a titration reaction when considered in isolation, i.e., it stops when one of the two reactants (NO or O₃) has been entirely depleted. In the atmosphere, in the presence of sunlight, O₃ can be continuously regenerated by NO₂ photolysis. Thus, a system of three reactions that are at equilibrium occurs, i.e., the rates of these three reactions are identical. This set of three reactions, which are at steady state, is called the photostationary state of Leighton, named after the Stanford chemistry professor, Philip A. Leighton.

These three reactions are as follows:

NO+O3 → NO2+O2k1=0.027 ppb−1 min−1(R8.11)

NO2+hν → NO+Ok2=0.3 min−1(R8.9)

O+O2 →(+M) O3k3=0.022 ppb−1 min−1(R8.10)

The rate constants are given here at 1 atm and 25 °C. A typical average daytime photolysis rate is used. At steady state, the rates of the three reactions are equal:

The first equation leads to:

 [O3] = k2 [NO2]k1 [NO](8.16)

In addition, the sum of the concentrations of nitrogen oxides must remain constant:

where the subscript 0 indicates the initial concentration. Each ozone molecule reacting with NO leads to a molecule of NO₂, and each photolyzed NO₂ molecule leads to a molecule of ozone; therefore, the sum of the concentrations of O₃ and NO₂ remains constant:

Thus, the concentrations of nitrogen oxides can be calculated as a function of the initial concentrations and the ozone concentration:

 [NO2]=[O3]0 + [NO2]0 − [O3] [NO]=[NO]0 + [NO2]0 − [NO2] = [NO]0− [O3]0 + [O3](8.19)

Replacing [NO₂] and [NO] in Equation 8.16:

 [O3] = k2 ([O3]0+ [NO2]0 − [O3])k1 ([NO]0− [O3]0 + [O3])(8.20)

The ozone concentration is then the solution of a quadratic equation:

[O3]=−(k1([NO]0−[O3]0)+k2)+((k1([NO]0−[O3]0)+k2)2+4k1k2([O3]0+[NO2]0))1/22k1(8.22)

8.3.3 Oxidation of Carbon Monoxide (CO)

The chemistry of CO is simple and is, therefore, convenient to explain ozone formation when volatile carbonaceous species (CO or VOC) are present. CO is oxidized by OH radicals:

Hydrogen atoms are not stable, and they recombine rapidly with molecular oxygen to form hydroperoxyl radicals (HO₂):

Then, these radicals react rapidly to oxidize NO into NO₂:

The OH radical has been regenerated, and the total budget of these three reactions is then as follows:

Therefore, the oxidation of CO into CO₂ leads to the conversion of NO into NO₂. NO₂ can be photolyzed to form NO and O₃. Since NO is converted to NO₂ without consumption of O₃ (unlike what happens in the Leighton photostationary state), there is formation of a molecule of O₃ for each molecule of CO that is oxidized. This yield is theoretical because all HO₂ radicals do not react with NO and some OH radicals may react with NO₂. The actual yield is, therefore, less than 1. The presence of a carbonaceous species (here CO, but VOC play a similar role, see Sections 8.3.4 to 8.3.8) leads to a perturbation of the photostationary equilibrium, thereby allowing the formation of NO₂ without O₃ consumption and leading, therefore, to O₃ formation.

Example: Calculation of the ozone concentration produced from carbon monoxide in presence of nitrogen oxides

The initial concentration of CO is 1 ppm and its oxidation occurs over an 8-hour period. The OH radical concentration is assumed to be 10⁶ cm⁻³.

The rate constant of the oxidation of CO by OH is 0.35 ppb⁻¹ min⁻¹ at 1 atm and 25 °C. The OH concentration in ppb is: 10⁶ / (2.46 × 10¹⁰) = 4 × 10⁻⁵ ppb. Formation of O₃ over eight hours is theoretically equivalent to the amount of CO that has reacted. Therefore:

[O3] = [CO]0 (1 − exp(− k [OH] t)[O3] = 1000×(1 − exp(− 0.35×4×10−5×8×60)[O3] = 6.7 ppb

Thus, 1 ppm of CO (which is not very reactive) has formed 6 ppb of O₃ in eight hours. Therefore, it is the presence of volatile carbonaceous compounds (here CO, but also reactive VOC), which leads to ozone formation, as correctly identified originally by Haagen-Smit.

8.3.4 Photolysis and Oxidation of Aldehydes

The simplest aldehyde (i.e., the aldehyde with only one carbon atom) is formaldehyde (HCHO). Its oxidation occurs by photolysis or by reaction with OH. There are two pathways for the photolysis of formaldehyde. On one hand:

Thus, the overall budget is:

On the other hand:

The two products of this reaction are stable molecules (CO will of course be oxidized slowly as described in Section 8.3.3). These two photolysis reactions have similar kinetics. Therefore, one may write a simple overall budget as being the average of these two reactions:

Oxidation by OH leads to the following reactions:

The OH radical abstracts a hydrogen atom from the formaldehyde molecule to form a stable molecule (water vapor, H₂O) and an unstable organic radical. Therefore, one obtains an overall budget that is similar to that obtained with the photolytic reactions:

Thus, the oxidation of HCHO, whether it occurs by reaction with OH or results from photolytic reactions leads to one molecule of CO and one HO₂ radical. As shown in Section 8.3.3, the HO₂ radical can later oxidize NO into NO₂ and form an OH radical. The photolysis of NO₂ leads to the formation of a molecule of O₃. Since the oxidation of a molecule of CO leads also to the formation of a molecule of O₃, the oxidation of HCHO can theoretically lead to the formation of two molecules of O₃.

The oxidation of higher aldehydes (i.e., aldehydes with more than one carbon atom) follows the same conceptual scheme as that of formaldehyde, but leads to more complex products due to the greater number of carbon atoms. For example, the reactions of acetaldehyde (two carbon atoms, CH₃CHO) are as follows:

The overall budget is as follows:

In this first oxidation step, H originating from formaldehyde has been replaced by CH₃ originating from acetaldehyde. The methylperoxyl radical (also called peroxymethyl), CH₃O₂, behaves similarly to HO₂, that is to say that it can oxidize NO into NO₂ and form a methoxy radical, CH₃O:

The methoxy radical reacts next with O₂ to form formaldehyde and the hydroperoxyl radical:

The HO₂ radical can subsequently oxidize another NO molecule into NO₂, thereby potentially leading to the formation of another O₃ molecule. Therefore, there may be formation of up to six molecules of O₃ from the photolysis of acetaldehyde (one from the oxidation of CO, one due to the formation of NO₂, two from the HO₂ radicals, and two from the oxidation of HCHO).

The other photolytic pathway is also possible:

As shown in Section 8.3.5, methane can lead to the formation of four molecules of O₃, but its lifetime is very long; therefore, it is considered to be chemically inert at the time scales of photochemical air pollution.

The oxidation of acetaldehyde by OH occurs according to the following pathway:

CH3CO+O2 →(+M) CH3C(O)O2(R8.30)

Peroxyacetyl nitrate, CH₃C(O)O₂NO₂, is commonly called PAN. PAN-type compounds may also be formed from higher aldehydes (propanal, butanal, etc.), in which case PAN refers more generally to peroxyacyl nitrates. Thus, the oxidation of VOC with several carbon atoms leads to more complex products. PAN is an important product because it is the organic nitrate species that is present at the highest ambient concentrations in the atmosphere. PAN can undergo a thermal decomposition reaction in the atmosphere:

PAN, when formed at a low temperature, can be transported over long distances and, when it encounters higher temperatures, it can decompose and form NO₂ back, and potentially O₃. PAN is, therefore, a “reservoir” species, because it acts as a reservoir of O₃ precursor.

Aldehydes are the only VOC that have significant photolytic rates. Calvert et al. (2011) provide a detailed state of the science of the atmospheric chemistry of aldehydes.

8.3.5 Oxidation of Alkanes

The simplest alkane (one carbon atom) is methane. Methane is a greenhouse gas, which reacts slowly with OH (lifetime on the order of several years). Therefore, it is not included in photochemical air pollution studies. Nevertheless, its oxidation mechanism being simple, it provides a useful conceptual model of the oxidation of alkanes in general:

As with aldehydes (see Section 8.3.4), the OH radical abstracts an H atom from the alkane molecule to form a stable H₂O molecule, and generates at the same time an organic radical (CH₃), which is rapidly oxidized by reaction with O₂ into a methylperoxyl radical (CH₃O₂). As discussed, this peroxyl radical and HO₂ may react with NO to form NO₂. The oxidation of methane leads theoretically to four molecules of O₃ (one via the formation of a molecule of NO₂ from CH₃O₂, two by the oxidation of HCHO and one from HO₂).

Higher alkanes, i.e., those with more carbon atoms (ethane, propane, butane, pentane, etc.), follow similar oxidation pathways. Let RH be a generic alkane (thus, R = CH₃ for methane, R = C₂H₅ for ethane, etc.):

where R’ has one less carbon atom than R. The mechanism is identical to that of methane, except that instead of formaldehyde being the product, a higher aldehyde is obtained (for example, acetaldehyde for ethane). In addition, minor reaction pathways lead to the formation of organic nitrates (RONO₂) and peroxynitrates (RO₂NO₂):

RO2+NO→(+M)RONO2(R8.40)

RO2+NO2→(+M)RO2NO2(R8.41)

For RO radicals, other chemical pathways include decomposition into an aldehyde and a peroxyl radical and, when R > C4, isomerization, which leads to a peroxyl radical and an alcohol group.

Figure 8.1 shows a partial reaction mechanism of the oxidation of n-pentane (an alkane with five carbon atoms) by OH radicals. There are five possible sites for the abstraction of a hydrogen atom by OH; however, for symmetry reasons, this corresponds to only three possible products with formation of a peroxyl radical in positions 1, 2, or 3 (from left to right in Figure 8.1). The yields of these three oxidation pathways are about 8, 57, and 35 %, respectively. Therefore, the probability of abstraction of a hydrogen atom is greater in an internal position than in a terminal position. The following steps are shown in Figure 8.1 only for the 2-pentyl-peroxyl radical, because (1) it is the radical that is formed in greater amount and (2) it provides the most complete reaction mechanism. The peroxyl radical can react with other peroxyl radicals, such as HO₂ and CH₃O₂ (i.e., those peroxyl radicals present in greater concentrations in the atmosphere), to form a peroxide when reacting with HO₂ or various oxygenated organic compounds when reacting with CH₃O₂. When NO_x concentrations are high, the other chemical pathways are favored. The reaction of NO₂ to form an organic peroxynitrate is fast, but reversible. The reactions with NO can lead either to the formation of an organic nitrate or to the oxidation of NO into NO₂ with formation of an alkoxy radical. This alkoxy radical can react along three distinct pathways: isomerization, decomposition, and oxidation by O₂. All these pathways lead to the formation of HO₂ and/or NO₂, and, therefore, to the potential formation of O₃.

Figure 8.1. Oxidation of n-pentane: oxidation by OH leading to the formation of peroxyl radicals, and reactions of the 2-pentyl-peroxyl radical.

As shown in Figure 8.2, alkanes with several carbon atoms are more reactive when the number of carbon atoms increases, because in first approximation an OH radical has a greater probability to encounter an hydrogen atom and abstract it from the alkane molecule. Therefore, ethane (two carbon atoms) is more reactive than methane (one carbon atom), butane (four carbon atoms) is more reactive than propane (three carbon atoms), etc.

Figure 8.2. Kinetics of oxidation by OH of selected n-alkanes.

Source of the data: Calvert et al. (2008).

Alkanes can also be oxidized by NO₃ radicals. The chemical mechanism is similar to that of the oxidation by OH radicals, because the NO₃ radical also abstracts a hydrogen atom from the alkane molecule to form a stable molecule of nitric acid, HNO₃:

However, the subsequent reactions differ from those resulting from the oxidation by OH, because NO concentrations are negligible when those of NO₃ are high. These differences are shown in Section 8.3.6 for alkenes, which have faster NO₃ oxidation kinetics than alkanes.

Alkanes are oxidized by OH during daytime and by NO₃ at night. However, the oxidation kinetics by NO₃ is much slower than that by OH (see Table 8.1). Calvert et al. (2008) provide a detailed state of the science of the atmospheric chemistry of the alkanes.

8.3.6 Oxidation of Alkenes

The oxidation of alkenes differs from that of alkanes because of their double bond, which is a more favorable site for attack by the OH and NO₃ radicals, and also by ozone. Therefore, instead of abstracting a hydrogen atom from the organic molecule, the OH and NO₃ radicals attach to one of the carbon atoms of the double bond. Figure 8.3 shows the chemical mechanism of the oxidation of propene (three carbon atoms, also called propylene) by OH. The OH radical can attach to either one of the carbon atoms of the double bond, but it tends to favor the formation of the radical on the secondary carbon (CH₃CHC(OH)H₂), i.e., the carbon atom that is linked to two other carbon atoms (65 % for that pathway, compared to 35 % for the other pathway). Next, these radicals react rapidly with molecular oxygen to form peroxyl radicals. These peroxyl radicals can then react with NO, either to form NO₂ and alkoxy radicals, or to form organic nitrates. However, the formation of nitrates is a relatively minor pathway (<2 %) and the formation of NO₂ prevails. Then, the alkoxy radicals undergo decomposition by reaction with O₂ to form aldehydes. In the case of propene, both chemical reaction pathways lead to the same products, which are formaldehyde and acetaldehyde. Then, the aldehydes lead to ozone formation (for example, two molecules of ozone in the case of formaldehyde, see Section 8.3.4). For alkenes with a greater number of carbon atoms, the mechanism is similar, but it leads to more complex oxo products (aldehydes and ketones).

Figure 8.3. Oxidation of propene by OH.

The NO₃ radicals follow the same general scheme in principle and attach to one of the carbon atoms of the double bond. However, the following steps differ, because the NO concentrations are negligible when NO₃ concentrations are significant. Therefore, the organic peroxyl radical does not react with NO, but instead reacts with other peroxyl radicals, such as HO₂ and CH₃O₂. Figure 8.4 shows the main oxidation pathways for the reaction of propene with NO₃. The oxidation products are mostly organic nitrates with other functional groups, such as peroxide, alcohol or oxo. In the case of propene, products include formaldehyde and acetaldehyde, as well as NO₂ via decomposition of the nitrated alkoxy radical. Unlike alkanes, alkenes can have oxidation kinetics with NO₃ that are faster than those with OH, particularly in the case of biogenic compounds.

Figure 8.4. Oxidation of propene by NO₃.

Alkenes are the only VOC that react with ozone. Figure 8.5 shows the chemical mechanism of the oxidation of propene by O₃. Ozone adds to the double bond creating a bridge between the two carbon atoms. This unstable species is called an ozonide. The products of the ozonide are oxo compounds (aldehydes and ketones) and biradicals (which, therefore, include two reactive sites with free electrons). In the case of propene, formaldehyde and acetaldehyde are formed. The biradicals, which are called Criegee radicals (named after Rudolf Criegee, who first identified them), are extremely reactive and lead to the formation of stable products (in the case of propene, CO, CO₂, CH₄, as well as formic acid and acetic acid) and radicals (OH, H, CH₃, HCO, CH₃O₂) that lead to ozone formation. In particular, one notes that OH radicals are formed. The yield of these OH radicals has been estimated to be 12 % from the H₂CO₂ biradical and 54 % from the CH₃ C(H)O₂ biradical. If both reaction pathways of the ozonide are about equivalent, the overall estimated OH yield is 33 %, which is consistent with the overall estimate. The OH yields can be significant; they vary from about 6 % for β-caryophyllene to nearly 120 % for cyclopentene. For some alkenes, the oxidation by O₃ is competitive with that by the OH and NO₃ radicals. The oxidation rate by O₃ is much greater for an internal alkene (internal double bond) than for a terminal alkene (double bond located at the end of the molecule); see Table 8.1.

Figure 8.5. Oxidation of propene by ozone.

Calvert et al. (2000) provide a detailed state of the science for the atmospheric chemistry of the alkenes.

8.3.7 Oxidation of Aromatic Compounds

The oxidation of aromatic compounds differs also from that of alkanes. The oxidation by OH can follow the same pathway as that of the alkanes with the abstraction of a hydrogen atom from the molecule to form an organic radical and a water molecule. However, this oxidation pathway is minor (<10 %) and it is the addition of the OH radical to the aromatic ring that prevails, because this addition requires less energy. There are two possibilities following this addition of OH to the ring:

1. 
   

   – The aromatic ring is conserved (for example, in the case of toluene, formation of cresol and its derivative products)

   
   
2. 
   

   – The aromatic ring is broken and species with fewer carbon atoms are formed

Figure 8.6 shows some of the chemical oxidation pathways and products of the oxidation of toluene (seven carbon atoms with a methyl group attached to the aromatic ring) by OH.

Figure 8.6. Oxidation of toluene by OH: major oxidation pathways and formation of major products.

The abstraction of a hydrogen atom from the methyl group leads to the formation of benzaldehyde, but with a benzaldehyde yield estimated to be about 6 %. Next, benzaldehyde reacts to form mostly nitrophenol.

The second oxidation pathway accounts for more than 90 % and leads to the formation of a radical that reacts with O₂, which either abstracts a hydrogen atom to form cresols and HO₂ (a potential source of O₃) or adds to the aromatic ring to form a peroxyl radical. The formation of cresols (mostly ortho-cresol, i.e., the CH₃ and OH groups are adjacent on the aromatic ring) accounts for about 18 % of the oxidation of toluene (about 12 % of ortho-cresol, 3 % of meta-cresol, and 3 % of para-cresol). The oxidation of cresols leads to the formation of peroxyl radicals that can lead to O₃ formation via the conversion of NO to NO₂ or can react with NO_x to form nitrocresols. In the latter case of the formation of an aromatic peroxyl radical, which accounts for >70 % of the toluene oxidation, there is mostly an opening of the aromatic ring. The products include aldehydes with fewer than six carbon atoms (glyoxal, methylglyoxal, methylbutenedial, 1,4-butenedial, 2-methyl-2,4-hexadiene-1,6-dial). Among those species, glyoxal and methylglyoxal are the most important ones with yields on the order of 4 to 17 % and 4 to 24 %, respectively.

The oxidation pathways where the aromatic ring is conserved lead to stable species (benzaldehyde, cresol, nitrated derivatives) and, therefore, little ozone formation. However, these species are not very volatile because of their high number of carbon atoms and they lead to particulate matter formation via the formation of particulate organic compounds (see Chapter 9). In addition, the formation of nitrated organic compounds leads to the elimination of nitrogen oxides (NO and NO₂) from the chemical system and, therefore, potentially less ozone. The oxidation pathways where the aromatic ring is opened lead to reactive species (dioxo compounds, i.e., species with two carbonyl groups, and in particular, dialdehydes) leading rapidly to ozone formation. On the other hand, these species have few carbon atoms and are, therefore, unlikely to lead to the formation of particulate organic compounds (however, their aqueous chemistry may lead to species with low volatility). The relative importance of these two main oxidation pathways (ring-retaining versus ring-opening) has long been poorly understood and still remains a source of uncertainty in photochemical air pollution.

Aromatic compounds can be oxidized by NO₃; however, the kinetics is much slower than that with OH radicals (see Table 8.1). Calvert et al. (2002) provide a detailed state of the science of the atmospheric chemistry of aromatic compounds.

8.3.8 Oxidation of Biogenic Compounds

Biogenic compounds include hemiterpenes, monoterpenes, sesquiterpenes, and terpenoids. The most important biogenic compound in terms of atmospheric emissions is isoprene (the hemiterpene), which is a hydrocarbon with five carbon atoms and two double bonds. The term “terpene” refers to a compound that is formed from a combination of several isoprene molecules (two for a monoterpene and three for a sesquiterpene) and, therefore, contains only carbon and hydrogen atoms. The term “terpenoid” is used when methyl groups have been removed or moved or when the compound contains one or more oxygen atoms. For simplicity, we use hereafter the term terpene to refer to all terpenes and terpenoids.

Isoprene is the biogenic compound that has been studied the most in terms of its atmospheric chemistry. The two double bonds are attack sites for OH and NO₃ radicals, as well as for O₃. The oxidation by OH leads to the formation of aldehydes and ketones. The most important ones are methacrolein (MACR) and methyl vinyl ketone (MVK). Isoprene is an important precursor of ozone.

Monoterpenes and other terpenes contribute also to the formation of ozone, but to a lesser extent, because their emissions are less important than those of isoprene (see Chapter 2). Accordingly, their chemistry has been studied more in terms of their contribution to the formation of organic particulate matter. Indeed, their number of carbon atoms (10 for monoterpenes and 15 for sesquiterpenes) favors the formation of compounds with low saturation vapor pressures, which can condense readily on particles. Therefore, the chemistry of biogenic compounds is discussed further in Chapter 9. One notes, however, that these compounds have short lifetimes (see Table 8.1), because they are oxidized rapidly by several oxidant species (OH, NO₃, and O₃).

8.3.9 Termination of the Oxidation Cycles

The formation of ozone occurs when NO is converted to NO₂ by peroxyl radicals (hydroperoxyl, HO₂, or organic peroxyl, RO₂), i.e., without any consumption of O₃. Thus, NO₂ forms O₃ during the day by photolysis. This formation takes place as long as peroxyl radicals are formed by the oxidation of VOC and CO. It can stop (or at least it can be slowed down) when one of the key species of this mechanism is consumed by a reaction that leads to a stable product. There are two main possibilities to stop the formation of ozone.

On one hand, the peroxyl radicals can react without converting NO to NO₂. They can react among themselves to form a peroxide or some other stable products. For example, the hydroperoxyl radicals can form hydrogen peroxide:

Another possibility for peroxyl radicals is the formation of organic nitrates or peroxynitrates (nitric acid is the inorganic nitrate), when NO and NO₂ react with organic peroxyl radicals. For example, PAN may be formed following the oxidation of acetaldehyde (see Section 8.3.4).

On the other hand, NO₂ can be oxidized into a stable product instead of being photolyzed. Its main oxidation pathway is by reaction with OH radicals to form nitric acid:

Therefore, there is competition in the former case for the reaction of peroxyl radicals either with NO to form NO₂ or with other peroxyl radicals or NO_x to form stable products such as peroxides and organic nitrates. In the latter case, there is consumption of NO₂ either by photolysis leading to O₃ formation or by reaction with OH to form HNO₃. We will see in Section 8.4 that the relative importance of these reaction pathways for peroxyl radicals and for NO₂ depends on the chemical regime of the atmosphere.

8.4.1 Conceptual Mechanism of Photochemical Air Pollution

Figure 8.7 presents a schematic description of the oxidation of an alkane, RH, by OH. In the first step, the oxidation of RH by OH leads to the formation of an organic peroxyl radical, RO₂, which reacts with NO to form NO₂ and RO. The organic alkoxy radical, RO, undergoes oxidation leading to the formation of an aldehyde (here called R’CHO, where R’ contains one less carbon atom than RH) and a hydrogen atom, which is rapidly oxidized into HO₂. The HO₂ radical leads to another oxidation of NO to NO₂ and the regeneration of the OH radical. Therefore, this oxidation of the alkane leads to the formation of two ozone molecules, since each NO₂ molecule is photolyzed during daytime and leads to the formation of NO and O₃. In addition, the aldehyde will be either oxidized by OH or photolyzed, which will lead to the formation of other O₃ molecules. Therefore, the oxidation of an alkane leads to the formation of several O₃ molecules via (1) the formation of peroxyl radicals and (2) the formation of an aldehyde, which via photolysis or oxidation will lead to the formation of other peroxyl radicals. As discussed, the oxidations of alkenes and aromatic compounds follow different reaction schemes, but the general conceptual approach is similar.

Figure 8.7. Schematic representation of the chemistry of ozone formation from an alkane, RH.

8.4.2 Simple Chemical Kinetic Mechanism of a Hydrocarbon

Table 8.2 lists the reactions corresponding to the oxidation mechanism of a generic hydrocarbon and its aldehyde product associated with the first oxidation step, as shown in Figure 8.7. This list of 20 reactions can be reduced during the development of the set of the ordinary differential equations representing the kinetics of the chemical species involved, because the reactions with molecular oxygen, O₂, are very fast. Thus, the reactions of the oxygen atom, O, and of the organic alkoxy radical, RO, with O₂ have kinetics that are not limiting and, accordingly, they can be lumped with the reactions leading to their formation. Then, this mechanism may be reduced to 18 reactions, after the lumping of Reaction 3 with Reactions 2 and 5 and Reaction 11 with Reaction 10. We can write the set of the corresponding differential equations as follows, using the stationary state approximation for the radicals (the numbering of the reactions and their rate constants corresponds to those of Table 8.2).

The radicals and atom of this reacting system are O(¹D), OH, HO₂, RO₂, and R’O₂ (the RO radical and the O atom are ignored following the lumping of their formation and destruction reactions). In addition, the RO₂ and R’O₂ radicals are lumped here as organic peroxyl radicals and the sum of their concentrations is represented by [RO₂]. Thus, there are five species at stationary state and ten species with time-dependent concentrations. The concentrations of the stationary-state species can be obtained with algebraic equations, and the time-dependent concentrations are governed by ordinary differential equations.

The stationary-state solution for [O(¹D)] is as follows:

d[O(1D)]dt=k4[O3] − k5[O(1D)][M] − k6[O(1D)][H2O]=0 (8.23)

That is: [O(1D)]=k4[O3]k5[M] + k6[H2O]

The stationary-state solution for [OH] is as follows:

d[OH]dt=2 k6[O(1D)][H2O] + k7[NO][HO2] + k8[O3][HO2]             − k9[RH][OH] − k12[R‘CHO][OH] − k18[NO2][OH]=0(8.24)

That is:

[OH] = 2 k6[O(1D)][H2O] + k7[NO][HO2] + k8[O3][HO2]k9[RH] + k12[R‘CHO] + k18[NO2]                  

The stationary-state solution for the concentration of the hydroperoxyl radical (HO₂) requires the solution of a quadratic algebraic equation:

d[HO2]dt = k10[NO][RO2] + k17[R‘CHO]              − 2 k19[HO2]2 − k20[RO2][HO2] − k7[NO][HO2] − k8[O3][HO2]=0(8.25)

That is:

[HO2] = −bHO2+(bHO2)2−4 aHO2 cHO2  2aHO2aHO2=2 k19; bHO2=k20[RO2] + k7[NO] + k8[O3]; cHO2=− (k10[NO][RO2] + k17[R‘CHO])

The concentrations of the organic peroxyl radicals, RO₂ and R’O₂, are lumped as [RO₂]. Since their self-reaction is neglected compared to their reaction with HO₂ (which is present in greater concentration), the solution is simpler than that for HO₂:

d[RO2]dt = k9[RH][OH] + k13[R‘C(O)O2][NO] + k17[R‘CHO]              − k10[NO][RO2] − k20[HO2][RO2]=0(8.26)

That is:  [RO2]= k9[RH][OH] + k13[R’C(O)O2][NO] + k17[R’CHO]k10[NO] + k20[HO2]

The algebraic equation for the peroxyacyl radical is as follows:

d[R‘C(O)O2]dt = k12[R‘CHO][OH] + k15[PAN] − k13[NO][R‘C(O)O2]− k14[NO2][R‘C(O)O2]=0(8.27)

That is: [R’C(O)O2]=k12[R’CHO][OH]+k15[PAN]k13[NO]+k14[NO2]

For the other species, the ordinary differential equations are as follows:

d[O3]dt=k2[NO2]−k1[NO][O3]−k4[O3]−k8[HO2][O3](8.28)

d[NO2]dt=k1[NO][O3]+k7[NO][HO2]+k10[NO][RO2]+k13[NO][R‘C(O)O2]              + k15[PAN]−k2[NO2]−k14[R‘C(O)O2][NO2]−k18[OH][NO2](8.29)

d[NO]dt=k2[NO2]−k1[NO][O3]−k7[NO][HO2]−k10[NO][RO2]−k13[NO][R‘C(O)O2](8.30)

d[RH]dt=−k9[RH][OH](8.31)

(8.32)d[R‘CHO]dt=k10[NO][RO2]−k12[R‘CHO][OH]−k16[R‘CHO]−k17[R‘CHO]

d[PAN]dt=k14[NO2][R‘C(O)O2]−k15[PAN](8.33)

d[HNO3]dt=k18[NO2][OH](8.34)

d[H2O2]dt=k19[HO2]2(8.35)

d[R‘H]dt=k16[R‘CHO](8.36)

d[CO]dt=k16[R‘CHO]+k17[R‘CHO](8.37)

The numerical solution of these ordinary differential equations provides the temporal evolution of the concentrations of these chemical species given their initial concentrations. The numerical methods that can be used to solve such a set of equations are described in Chapter 7.

8.4.3 Spatio-temporal Variability of Photochemical Air Pollution

The diurnal profile of ozone concentrations shows a peak in midday, since ozone formation results from the photolysis of NO₂ and the photolysis kinetics is maximum at noon. Ozone formation continues during the afternoon until the ozone destruction reactions (e.g., the NO titration reaction) begin to overcome the formation reaction, i.e., the NO₂ photolysis, as radical (OH and HO₂) production starts to slow down (i.e., when the Sun comes down toward the horizon). Figure 8.8 depicts temporal profiles of measured and simulated ozone concentrations corresponding to a photochemical air pollution episode in the Los Angeles Basin in 1974, that is, at a time when air pollution levels were still very high (maximum hourly concentrations >300 ppb, i.e., about 600 μg m⁻³).

Figure 8.8. Temporal profiles of ozone (O₃, top figure) and nitrogen dioxide (NO₂, bottom figure) concentrations on June 26 and 27, 1974 in the Los Angeles Basin, California (in pphm, “parts per hundred million”; 1 pphm = 10 ppb). The small squares correspond to the measurements and the lines correspond to the results of a numerical simulation. The solid lines correspond to the simulation at the location of the measurement; the dotted lines correspond to optimal solutions obtained at one or two model grid cells of the measurement location. O₃ concentrations are in Pasadena, downwind of Los Angeles; NO₂ concentrations are in downtown Los Angeles.

Source: Tesche et al. (1984).

Figure 8.8 shows temporal profiles of measured and simulated NO₂ concentrations for the same Los Angeles air pollution episode. The diurnal profile of the nitrogen dioxide (NO₂) concentrations shows peaks during the traffic rush hour (i.e., in the morning and in late afternoon) and lower concentrations during the day when emissions are lower and photolysis of this compound is more important. The maximum NO₂ concentration reaches 300 ppb (about 600 μg m⁻³) during the early-morning peak of the second day.

The spatial distribution of the ozone concentrations typically shows lower concentrations near nitrogen oxides (NO_x) sources because of the titration of O₃ by NO. Thus, the maximum O₃ concentrations occur downwind of the source regions, since it takes some time for the kinetics of the photochemical reactions to occur and to lead to ozone formation. On the other hand, the maximum NO₂ concentrations occur mostly near NO_x sources, i.e., in urban areas near major roadways. For example, the O₃ concentrations shown in Figure 8.8 correspond to a location in Pasadena, i.e., downwind of the industrial areas (located on the coast) and of the downtown area. Figure 8.9 depicts maps of simulated concentrations of O₃ and NO₂ for the Paris region in 2009. The high NO₂ concentrations appear in the downtown Paris area, near major roadways, near the airports, and near an industrial site. On the other hand, the O₃ concentrations are negligible in these locations because of the titration reaction with NO and they are high outside of the urban area.

Figure 8.9. Spatial distribution of simulated O₃ and NO₂ concentrations for 2009 in the Paris region. The simulation was performed with the Polyphemus numerical model. Top figure: O₃ concentrations (annual average in μg m⁻³ of the daily maximum 8-hour average concentrations); bottom figure: NO₂ concentrations (annual average in μg m⁻³). Concentrations are indicated on the right vertical axis. Latitude (°N) and longitude (°E) are indicated on the left vertical axis and horizontal axis, respectively.

Source: Ademe (2014).

8.5.1 Chemical Regimes of Photochemical Smog

Ozone formation occurs when the Leighton photostationary state undergoes a perturbation due to the addition of VOC or CO, thereby leading to the formation of peroxyl radicals, which can convert NO to NO₂ without consuming O₃. NO₂ photolysis leads to the formation of one molecule of O₃ and regenerates NO. NO can then be converted back to NO₂, either by reaction with O₃ (null O₃ balance) or by reaction with a peroxyl radical (potential formation of an O₃ molecule). Therefore, there is competition between these two NO oxidation pathways and we may distinguish two chemical regimes:

1. 
   

   – A regime where the peroxyl radical concentrations are not limiting compared to NO (or more generally NO_x) concentrations

   
   
2. 
   

   – A regime where the peroxyl radical concentrations are limiting compared to NO (or more generally NO_x) concentrations

The peroxyl radicals are formed by oxidation of VOC or CO. However, CO is significantly less reactive than most VOC and, therefore, one may consider that high peroxyl radical concentrations result from high VOC concentrations. These two regimes may, therefore, be defined in terms of the relative concentrations of NO_x and VOC:

1. 
   

   – A regime rich in VOC and poor in NO_x (hereafter referred to as low-NO_x regime)

   
   
2. 
   

   – A regime poor in VOC and rich in NO_x (hereafter referred to as high-NO_x regime)

There is of course an intermediate regime. However, most atmospheric conditions tend to fall within one of these two extreme regimes.

These different chemical regimes can be characterized in terms of their chemical kinetics according to the mechanism presented in Figure 8.7, using the rate constants listed in Table 8.2. We use here the conceptual approach of Jacob (1999) with some modifications. Ozone is formed by photolysis of NO₂ following the oxidation of NO by a peroxyl radical (RO₂ or HO₂). Therefore, the formation rate of O₃, PO₃, can be expressed as follows:

The OH and HO₂ radicals, represented together as HO_x, play a key role in this scheme of O₃ formation. They are formed mostly by the photolysis of O₃ and nitrous acid (HNO₂) in the case of OH and by the photolysis of aldehydes in the case of HO₂. Their formation kinetics are represented here by POH and PHO₂, respectively. Furthermore, we define PHO_x = POH + PHO₂, to represent the formation of all HO_x radicals. POH is greater than PHO₂, the former being typically about 50 to 80 % of PHO_x (Mao et al., 2010). The main termination reactions are, for OH, the formation of nitric acid (HNO₃) and, for HO₂, the formation of peroxides by reaction with HO₂ or RO₂. At steady state, their production rate is equal to the rate of their termination reactions and we have the following relationship:

Assuming steady state for these radicals:

d[OH]dt=POH+k7 [NO] [HO2]−k9 [RH] [OH]−k18 [NO2] [OH]=0(8.40)

(8.41)d[HO2]dt=PHO2+k10 [NO] [RO2]−k7 [NO] [HO2]−2 k19 [HO2]2−k20 [RO2] [HO2]=0

The rates of the propagation reactions of these radicals are assumed to be significantly faster than their formation and termination reactions (the ratio of these rates is typically on the order of 5 to 10; Mao et al., 2010). Therefore, we obtain the following relationships:

Then, the production of O₃ can be expressed as follows:

In addition, since the values of the rate constants k₇ and k₁₀ are similar, the following simplifying assumption may be introduced:

Thus, replacing [RO₂] by [HO₂] in the steady-state relationship between production and termination of the HO_x radicals:

Low-NO_x Regime

In a low-NO_x regime, the formation of peroxides governs the termination reactions. Therefore, the first term of PHO_x (Equation 8.46) can be neglected and we have:

[HO2] ≈ (PHOx(2 k19 + k20))1/2 (8.47)

Then, we can write the formation of O₃ (Equation 8.44) simply as a function of HO_x production and the concentration of NO:

PO3 ≈ 2 k7 [NO] [HO2]  ≈ 2 k7 [NO] (PHOx2 k19+k20)1/2(8.48)

Thus, in a low-NO_x regime, O₃ formation is proportional to the concentration of NO and to the square root of the production of HO_x radicals.

High-NO_x Regime

In a high-NO_x regime, nitric acid formation governs the termination reactions and the second term of PHOx (Equation 8.46) can be neglected. The OH concentration may then be calculated as follows:

[OH] ≈ PHOxk18 [NO2]  (8.49)

The production of O₃ can be expressed simply as a function of the HO_x production and the concentrations of RH and NO₂:

PO3 ≈ 2 k9 [RH] [OH] ≈ 2 PHOx k9 [RH]k18 [NO2] (8.50)

This result implies that the rate of propagation of the radicals is much greater than that of their termination (assumption made to obtain Equations 8.42 and 8.43). However, this assumption is not always verified. For example, the ratio of these rates is only about 2 in New York (Mao et al., 2010) and could be close to 1 in Paris (based on reactivity and ambient concentrations measurements; Borbon et al., 2013; Airparif, 2015). Then, the steady-state relationship for [OH] leads to the more general relationship:

k₇[NO][HO₂] = k₉[RH][OH] + k₁₈[NO₂][OH] − POH

k7 [NO] [HO2]=k9 [RH] [OH]+k18 [NO2] [OH]−POH(8.51)

The steady-state assumption made for [HO₂] leads to the following relationship (neglecting the termination reactions of the peroxyl radicals, which is appropriate in a high-NO_x regime):

k₁₀ [NO][RO₂] ≈ k₇ [NO][HO₂] − PHO₂

k10 [NO] [RO2] ≈ k7 [NO] [HO2]−PHO2 (8.52)

Therefore, substituting these two terms in the equation for the production of O₃ (Equation 8.38), leads to the following solution:

PO₃ ≈ 2(k₉ [RH][OH] + k₁₈ [NO₂][OH] − POH) − PHO₂

PO3 ≈ 2 (k9 [RH] [OH]+k18 [NO2] [OH]−POH)−PHO2(8.53)

Substituting [OH] by its value as a function of PHO_x:

PO3 ≈ 2 (k9 [RH]+k18 [NO2] ) PHOxk18 [NO2]−2 POH−PHO2(8.54)

Then, writing POH as a function of PHO_x and PHO₂, the general equation for the formation of O₃ in a high-NO_x regime is obtained:

PO3 ≈ 2 PHOx (k9 [RH]k18 [NO2])+PHO2(8.55)

The previous equation for PO₃ in the high-NO_x regime is obtained when k₁₈ [NO₂] ≪ k₉ [RH]. In a high-NO_x regime, O₃ formation is proportional to the production of HO_x radicals and to the [RH]/[NO₂] ratio. If the regime is very high NO_x, then, the O₃ formation rate tends toward the production rate of HO₂ radicals, which corresponds mostly to the photolysis rate of aldehydes.

Nitrogen oxides have a relatively short lifetime (a few hours for NO₂ by reaction with OH); therefore, in the absence of important NO_x emissions, the atmospheric concentrations of NO and NO₂ are low (on the order of a few ppb).

VOC have variable lifetimes ranging from a few hours (e.g., aldehydes) to a few days (e.g., benzene). In addition, the oxidation of a VOC with several carbon atoms will occur over several oxidation steps and will, therefore, last for some time. Also, VOC emissions are both anthropogenic (urban areas, industrial sites …) and biogenic (rural areas, forests). Therefore, although VOC concentrations can be high near major sources (e.g., in urban areas because of on-road traffic), they are rarely negligible given the diverse lifetimes of VOC and their oxidation products as well as the spatial distribution of their emissions. Given these general patterns on the atmospheric concentrations of NO_x and VOC, one may assume that the atmosphere is generally high-NO_x in urban areas, near major traffic routes (shipping, on-road, and air traffic), and near some industrial sites and is generally low-NO_x in rural areas. However, some urban areas and industrial sites may have important VOC concentrations, for example, if the surrounding biogenic emissions are important. Thus, we may have specific cases with high VOC regimes (relatively to NO_x) in some urban areas and industrial sites.

8.5.2 Strategies to Reduce Ozone Precursor Emissions

It is important to identify the chemical regime of ozone formation, because it affects the efficiency of various strategies available for reducing the emissions of ozone precursors. Let us consider the two cases described in Section 8.5.1: low-NO_x regime and high-NO_x regime.

Low-NO_x Regime

NO_x emissions consist mostly of NO and include little NO₂ (<10 % except for some diesel vehicles with catalytic particle filters, see Chapter 2). If one reduces NO_x emissions (i.e., mostly NO concentrations), one reduces the formation of NO₂ by oxidation of NO by peroxyl radicals and, therefore, O₃ formation by photolysis of NO₂. Indeed, Equation 8.48 shows that O₃ formation is proportional to the NO concentration. In a low-NO_x regime, a reduction of the NO_x emissions is efficient to reduce O₃ concentrations.

If one reduces VOC emissions (i.e., VOC concentrations), one reduces the production of peroxyl radicals and, therefore, O₃ formation. However, although the production of peroxyl radicals decreases, their concentration does not decrease proportionately because they are consumed via two different reactions pathways: (1) their reaction with NO (which is proportional to their concentration) and (2) their self-reaction, which leads to the formation of peroxides (see Figure 8.10) and is proportional to the square of their concentration. The kinetics of the second consumption pathway decreases, therefore, faster than their production rate and, as a result, their concentration decreases proportionately less than the VOC concentration. The O₃ formation rate given by Equation 8.48 depends on VOC emissions only via the production of HO_x radicals (PHO_x), i.e., aldehyde photolysis, which implies that VOC emissions have little effect on ozone production. In addition, alkenes react with O₃ (see Section 8.3.6) and if their concentrations are important, a decrease of their concentration may lead to less O₃ consumption and, therefore, a small increase in O₃ concentrations. A VOC reduction strategy in a high-VOC regime (i.e., low-NOx regime) is not very efficient to reduce O₃ concentrations.

Figure 8.10. Schematic representations of ozone chemistry in a low-NO_x regime (top figure) and a high-NO_x regime (bottom figure). VOC: volatile organic compound, OVOC: oxidized volatile organic compound. NO, NO₂, and O₃ are involved in the Leighton photostationary-state reactions, and the directions of the arrows indicate the main effect of the change in NO_x emissions on the O₃ concentration for each regime.

Design of Emission Control Strategies

In summary, an efficient strategy to reduce ozone concentrations must focus on the reduction of the precursor (NO_x or VOC) that has the lower atmospheric concentration and is, therefore, limiting for O₃ formation. A reduction of the emissions of the precursor that has the higher atmospheric concentrations may not be efficient and may even be counterproductive because it may lead to an increase of the O₃ concentration. Figure 8.10 depicts these two regimes with the main termination reactions: peroxide formation in a low-NO_x regime and nitric acid formation in a high-NO_x regime. These mechanisms imply that it is possible to determine experimentally whether the O₃ concentrations measured in the atmosphere have been formed under a low-NO_x or high-NO_x regime (Sillman, 1995). For example, one only needs to measure the HNO₃ and H₂O₂ concentrations. If the [HNO₃]/[H₂O₂] ratio is high, then O₃ has been formed under a high-NO_x regime. If this ratio is low, O₃ has been formed under a low-NO_x regime. It is then possible to determine which emission reduction strategy will be the most efficient to reduce O₃ concentrations.

EKMA (« Empirical Kinetic Modeling Approach ») diagrams have been developed in the U.S. to provide simple information to help develop efficient emission reduction strategies to reduce O₃ concentrations. They are developed based on simulations of the atmospheric chemistry of O₃. These simulations may use simple or advanced atmospheric chemistry models. Initially, they were based on trajectory model simulations, i.e., a box-model that follows a mean wind trajectory. An ensemble of simulations was conducted with different VOC and NO_x emission levels, to construct a diagram such as the one presented in Figure 8.11. This diagram highlights the two chemical regimes: high-NO_x in the top left part of the figure and low-NO_x in the bottom right part. In the high-NO_x case, a reduction of NO_x emissions leads toward isopleths corresponding to higher O₃ concentrations and a reduction of VOC emissions is required to reduce O₃ concentrations. In the low-NO_x case, a reduction of VOC emissions is not very efficient, because the change in O₃ concentrations is nearly parallel to the O₃ isopleths. On the other hand, a reduction of NO_x emissions leads toward isopleths corresponding to lower O₃ concentrations. A [VOC]/[NO_x] ratio, expressed as (ppb C)/(ppb NO_x), of about 8 is typically considered to be the limit between the high-NO_x and low-NO_x regimes.

Figure 8.11. Isopleth diagram of O₃ concentrations as a function of VOC and NO_x emissions. It is also known as an EKMA diagram, where EKMA stands for “Empirical kinetic modeling approach.” A reduction of VOC emissions leads to lower O₃ concentrations in a high-NO_x regime (i.e., low-VOC regime), whereas a NO_x emission reduction is more efficient in a low-NO_x regime. Reducing NO_x emissions in a high-NO_x regime leads to an increase in O₃ concentrations.

Figure 8.12 shows this type of information obtained for the Paris region with a more comprehensive model than those typically used for EKMA diagrams. The Polyphemus numerical model used here is a three-dimensional model that includes all relevant atmospheric emissions, transport, dispersion, transformation, and deposition processes. The limit between the high-NO_x and low-NO_x regimes is located beyond the downtown Paris area. A simulation conducted with a 15 % reduction in NO_x emissions shows an increase in O₃ concentrations in the Paris urban area, which corresponds to the high-NO_x regime. On the other hand, a simulation conducted with a 15 % reduction in VOC emissions leads to a significant reduction of the O₃ concentrations within the Paris urban and suburban areas, but negligible reductions in the surrounding rural areas where the low-NO_x regime does not favor VOC emission reductions.

Figure 8.12. Influence of NO_x and VOC emissions on O₃ in the Paris region. Top figure: VOC/NO_x atmospheric concentration ratio for the Paris region (the dotted line corresponds to [VOC]/[NO_x] = 8 mole C/mole NO_x). Changes in O₃ concentrations in μg m⁻³ (average from May to September 2005 of the maximum daily 8-h average O₃ concentrations) due to emission reductions over the Paris region: 15 % NO_x reduction (middle figure) and 15 % VOC reduction (bottom figure). The VOC/NO_x ratio and changes in O₃ concentrations are indicated on the right vertical axis. Latitude (°N) and longitude (°E) are indicated on the left vertical axis and horizontal axis, respectively. All simulations were performed with the Polyphemus numerical model.

Source: Cerea (2011).

8.5.3 Ozone Formation Potentials of VOC

VOC differ significantly in terms of their chemical reactivity and, therefore, in terms of their potential to form ozone. The lifetimes presented in Table 8.1 provide useful information concerning the oxidation kinetics of selected VOC. Such information is relevant to their potential to form ozone in the presence of sunlight and NO_x. However, the products of these oxidation reactions also affect ozone formation and the kinetics of the first oxidation step (called kinetic reactivity) is not sufficient to estimate the ozone formation potential of a VOC. The number of ozone molecules formed per VOC molecule that has been oxidized gives some information on the potential of that VOC to form O₃. However, this measure of the ozone formation potential (called mechanistic reactivity) does not account for the time needed to form O₃, which is important in the atmosphere where transport and deposition processes must also be taken into account. Note that this mechanistic reactivity differs from the theoretical ozone yields mentioned in Section 8.3 for some VOC (for example, two O₃ molecules per molecule of formaldehyde being oxidized or photolyzed), because the chemistry of a VOC depends on other chemical species present in the atmosphere, which in turn are influenced by the VOC being studied.

Therefore, one should see these two approaches (mechanistic reactivity and kinetic reactivity) as complementary, one providing valuable information on the amount of O₃ that will eventually be formed and the other indicating the time needed to form O₃. Thus, it is appropriate to use a metric of the ozone formation potential that combines these two reactivity concepts. To that end, one may estimate the amount of O₃ formed per molecule (or mass) of VOC added to a mixture of VOC and NO_x. This metric is called the incremental reactivity, because one measures the VOC reactivity in terms of its increment of ozone formation. There is no unique definition for this type of ozone formation potential, because (1) it depends on the conditions under which it has been estimated (e.g., mixture of VOC being used, relative amounts of VOC and NO_x, and duration of the simulation or experiment) and (2) the amount of O₃ formed can be defined in several ways (maximum concentration, concentration averaged over a certain period, etc.).

Three main metrics of ozone formation potentials that are commonly used may be mentioned: Maximum Incremental Reactivity (MIR), Maximum Ozone Incremental Reactivity (MOIR), and Equal Benefits Incremental Reactivity (EBIR). These different metrics give different values for the ozone formation potential, but they are rather similar in terms of the reactivity of a specific VOC relative to those of the other VOC. MIR is used here to illustrate this concept of ozone formation potential. “Maximum” refers to the fact that the experiment is conducted under high-NO_x conditions for which the formation of O₃ from a VOC is optimal. The estimation of the MIR of a VOC depends on the mixture composition, the sunlight conditions (photolysis rates), and the duration of the simulation or experiment. Nevertheless, once those conditions have been set, it is possible to estimate in a consistent manner the MIR of a large number of VOC under conditions that approximate rather well those observed during air pollution episodes. The estimation of MIR can be conducted either experimentally (in smog chambers) or via numerical simulations using a chemical kinetic mechanism, which has been previously evaluated satisfactorily against smog chamber data (see Section 8.6).

Figure 8.13 shows some results obtained with MIR estimated using simulations conducted with the SAPRC-07 mechanism (Carter, 2010). A simulation time of a few hours and a high-NO_x mixture ([VOC]/[NO_x] = 3.7) including VOC representative of an urban area were used for these simulations. There is good agreement between the MIR scale and the rate constants of the OH reaction for some species (for example, for the biogenic compounds). However, there are significant differences for most of the other species. For example, the alkane MIR values are similar, whereas their OH reaction kinetics are nearly proportional to their number of carbon atoms (see Figure 8.2). Furthermore, toluene and xylene have MIR values similar to those of the biogenic compounds, despite having much lower OH reaction rate constants. These results highlight the importance of mechanistic reactivity as a complement to kinetic reactivity. According to the results presented in Figure 8.13, the most reactive VOC in terms of O₃ formation are the alkenes (including biogenic compounds), the aromatic compounds (except for benzene), and the aldehydes. We will see in Chapter 9 that one obtains different results when VOC reactivity is considered in terms of secondary organic particulate matter formation.

Figure 8.13. Maximum incremental reactivity (MIR) and oxidation kinetics of selected VOC. MIR, here in number of moles of O₃ formed per mole of VOC added to the initial mixture, obtained by numerical simulation using the SAPRC-07 chemical kinetic mechanism (in black), left scale, and rate constant of the OH reaction (in gray), right scale.

Sources: Carter (2010) for MIR and references from Table 8.1 for k_OH. See text for the interpretation of these values.