Tải bản đầy đủ - 0 (trang)
7…Photochemistry in the Troposphere

7…Photochemistry in the Troposphere

Tải bản đầy đủ - 0trang


R. S. Mason

reaction, in addition to the relatively minor volcanic input. Emissions with short

lifetimes tend to remain confined to the troposphere. Some of these are inorganic,

such as NH3, SO2 or H2S, but the vast majority are organic molecules. The

longest-lived will obviously be more evenly distributed than the shortest. Methane,

the main organic component of air, has a residence time measured in years, but

there is a difference in its abundance between the northern (1750 ppm) and

southern (1500 ppm) tropospheric hemispheres. This is due to the fact that two of

its main sources are from farming: i.e. enteric fermentation in ruminants and rice

growing, both of which are more concentrated in the northern hemisphere. It is

exacerbated by the intertropical convergence. This is the boundary between air

masses in the two hemispheres caused by the horizontal countermovement due to

rotation of the Earth (west to east in the north, and the opposite in the south). At

the convergence the air masses are therefore moving parallel to each other, so that

transfer across the convergence is slower than within the hemispheres.

Others such as isoprene (C5H8), or the terpenes (C10H16) and related compounds (responsible for many plant smells), emitted by plants and trees have

chemical lifetimes measured in hours, minutes or seconds and therefore never

travel more than a few kilometres from their source, depending on the wind speed.

In fact, the principal limitation on the lifetimes of shorter-lived organic species in

the troposphere is their reaction with OH, which, although its background mixing

ratio is in the 10-13 range, is therefore critical to tropospheric chemistry.

In the troposphere it is produced by a variety of reactions, but principally the

photodecomposition of ozone. About 10 % of tropospheric O3 results from its

downward transport across the tropopause. Residual UV radiation at k \ 310 nm,

penetrating the troposphere, causes photolysis, by reaction (5.52) to produce the

singlet O atom.

Although O(1D) is mostly quenched down to its ground state by collision (5.20)

a small proportion reacts with water to form OH:


O 1 D þ H2 O ! OH þ OH


If NO2 is present (for example in a polluted atmosphere), this too is broken

down by UV (when k \ 410 nm):

hm ỵ NO2 ! NO ỵ O


The O(3P) atoms react with O2 to reform O3 by reaction (5.21) which again may

lead back to OH formation, via photolysis reaction (5.52) which regenerates

O(1D). In an unpolluted atmosphere the main sink reaction for OH is reaction with

CH4. It is this reaction which forms the backdrop to atmospheric chemistry in the

troposphere. The sequence is complicated; Scheme 5.1 is the model suggested by

Crutzen [10], and it illustrates the most likely oxidation route for most hydrocarbons emitted into the atmosphere.

Scheme 5.1: Oxidation of methane and formation of ozone

OH ỵ CH4 ! CH3 ỵ H2 O


5 Atmospheric Photochemistry


CH3 ỵ O2 ! CH3 O2


CH3 O2 ỵ NO ! CH3 O ỵ NO2


CH3 O ỵ O2 ! HCHO ỵ HO2


HCHO ỵ hmk

330 nmị ! HCO ỵ H


HCO ỵ O2 ! CO ỵ HO2


H þ O2 þ M ! HO2 þ M


fHO2 þ NO !  OH ỵ NO2 g 3


fNO2 ỵ hm ! NO ỵ Og 4


fO ỵ O2 ỵ M ! O3 ỵ Mg 3


The overall reaction is therefore:

CH4 þ 8O2 þ 5 photons ! CO þ 4O3 þ 2OH ỵ H2 O


Once radicals are formed, either by photodecomposition or H abstraction, then

they can react directly with O2 to form a peroxide radical (5.60), which in the

presence of NO and NO2 can react to make an aldehyde (5.62), which breaks down

in sunlight (5.63) to CO in this example. As shown, both NO2 (5.61, 5.66 and 5.37)

and HO2 (5.62, 5.64 and 5.66) are thought to be involved in the chain; the former

breaks down in sunlight (k B 410 nm) to NO ? O which gives rise to more O3

formation, and catalyses the formation of more OH (5.66). Other hydrocarbons

will lead to longer carbon chain products.

In this model, for every methane molecule which reacts, the sequence leads to 4

ozone and 2 hydroxyl radicals, extra. Formation of ozone in the lower troposphere

is therefore catalysed by photochemical oxidation of organic molecules, but it does

require comparatively high levels of NO (mixing ratio [ 5 - 10 9 10-12) to be

present. If it goes to completion, OH can react further with CO to make CO2 thus

completing the oxidation of methane (Scheme 5.2). At low NO levels, the net

reaction is the destruction of ozone via the reaction with CO [Scheme 5.2b].

Scheme 5.2: Oxidation of CO and (a) formation or (b) removal of ozone

OH ỵ CO ! H ỵ CO2


H ỵ O2 ! HO2 ỵ M


HO2 ỵ NO !  OH ỵ NO2


NO2 ỵ hm ! NO þ O


(a) in the presence of NO:


R. S. Mason

O þ O2 ỵ M ! O3 ỵ M


HO2 ỵ O3 !  OH ỵ 2O2


(b) without NO:

The main natural source of NO is reaction (5.35).

O 1 D ỵ N2 O ! 2NO


Ozone formation is enhanced when other more reactive organic molecules are

present, such as on a hot day in the presence of vegetation; although the chemistry

is more complex. Thus, for example, the heat haze above a pine forest is caused by

emitted terpenes (C10H16) and other organic molecules which promote levels of

both O3 and NO2. The latter has a brown colour which can be seen. Thus, ozone

levels and other associated gases, wax and wane with the presence of sunlight

(along with other weather factors). The distribution of ozone at ground level

therefore depends on local factors such as the vegetation. Figure 5.6 shows O3

maps for average levels for three different years across mainland Britain. This

shows ozone levels to be more concentrated in the rural areas such as the South

West, Wales or northern Scotland. There are much more localised concentrations

(although not so easy to see from these maps) in large city conurbations (London,

Birmingham, Manchester etc.). The latter is not due to the vegetation, but is the

well-known problem of air pollution. The difference between the years highlights

the sensitivity to hours of sunshine experienced.

Fig. 5.6 Average annual mean distribution of ozone in the UK (lg m-3), for various years.

Reproduced and adapted with permission from Fig. 2.30 in Ref. [11] (Copyright 2008, DEFRA)

5 Atmospheric Photochemistry


5.7.2 The Urban Polluted Atmosphere

Reactions in the polluted atmosphere are very similar to the natural background

reactions, except for the intensity of reaction. The combustion engine, coal, oil and

natural gas burning all pump large quantities of unburnt hydrocarbon fuel, and the

products of combustion into the atmosphere. Automobile emissions are especially

acute, pumping out petrol or oil fumes, partially burnt hydrocarbons, CO and NO

(besides CO2, a different sort of pollutant) into the local atmosphere. On a still

sunny day this is a toxic combination which leads to high levels of photochemical

reaction, especially promoted by the presence of excess NO, which helps catalyse

the formation of high levels of O3 and ultimately its removal. Whereas O3 in the

stratosphere is essential to life on earth, at ground level it is regarded as toxic at

levels [200 ppb.

Among other products which build up during the day are NO2, HNO3, and

partially oxidised hydrocarbons, such as formaldehyde. Another is peroxyacetyl

nitrate (PAN). This is thought to result from the sequence of reactions:

CH3 CO ỵ O2 ! CH3 COO2




PAN is also present at low pollution levels; and is thought to be reservoir

species for NO2 since it is broken down again, but relatively slowly, by sunlight; at

high pollution levels (ppb) it is an intense eye irritant. NO2 is ultimately removed

mostly by its gas phase reaction (5.72) to make HNO3, which is very soluble in

water and therefore easily rained out.

OH þ NO2 þ M ! HNO3 þ M


On sunny still days, local low level boundary layers of air can form. This is

exacerbated by certain geographical features, such as an onshore wind over a

coastal plain bounded by hills or mountains. The boundary layer forms either as a

result of a faster cooling of air with altitude close to the ground than the adiabatic

cooling rate, in which case convection is inhibited (the early morning low level

mist in river valleys), or the air above it becomes heated, making the localised

solar heating rate greater than the adiabatic cooling rate, thus creating a layer of air

which inhibits convection. This boundary layer, which is similar to the tropopause,

but is temporary, acts as a cap just a few hundred metres from the surface, trapping

the air mass below it and allowing pollutant emissions to build up (rather than

being dispersed as normal). The Los Angeles basin is the usual example quoted,

where high photochemical pollution is endemic, but many other cities around the

world show similar levels of pollution. Figure 5.7 shows urban pollution levels

recorded in Beijing, China during the winter months of 2001.

This very clearly demonstrates the correlation between emissions of NOx and

CO and the formation of O3 (hydrocarbons are not shown). Hydrocarbons, NO and

CO are emitted into the morning urban air by the morning rush-hour, peaking


R. S. Mason

Fig. 5.7 Averaged diurnal variations of NOx (ppb), CO (ppm) and O3(ppb) in Beijing; Jan–Feb

2001; The data was extracted from Ref. [12]

at ca. 8.00 h. As the sunlight increases in intensity the chemistry illustrated by

Schemes 5.1, 5.2 (but involving petroleum hydrocarbons, rather than CH4) occurs,

building up over the day, leading to higher levels of O3 which peaks at ca. 15.00 h,

when both NOx and CO show a dip. A very fast reaction, not so far discussed, is

(5.73), in which NO reacts with O3 to form NO2. Reactions (5.73) and (5.37) are so

fast, compared to others, that the reactions in Scheme 5.3 may generate a photochemical steady state at the height of the day. Of (5.37) and (5.21), (5.37) is the

rate determining step and so dictates the rate of formation of O3. (5.73) dictates its

rate of removal, and therefore under steady state conditions, their rates would


Scheme 5.3: A photochemical steady state?

NO2 ỵ hm ! NO ỵ O


O ỵ O2 ỵ M ! O3 ỵ M


NO ỵ O3 ! NO2 ỵ O2


J37 ẵNO2 ẳ k73 ẵNO][O3


and [O3] is therefore given by:

ẵO3 ẳ J37 =k73 ịẵNO2 =ẵNO


where J37 is the actinic flux responsible for reaction (5.37). In bright sunlight

therefore the removal reaction (5.73) has no significant effect on the ozone levels,

5 Atmospheric Photochemistry


since ozone is rapidly regenerated from the product. However, in the evening when

J37 is low, the reaction is a significant removal mechanism, hence in Fig. 5.7, O3

dips and NOx increases.

The higher levels of ozone recorded for unpolluted rural areas (Fig. 5.6) may at

first seem counter-intuitive. It has been argued, however, that these higher rural

levels of ozone are largely due to the relatively low levels of rural NO emission

and hence the relative absence of reaction (5.73) as an efficient removal


As discussed above, COS (present at 500 ppt) is thought to be the most

abundant natural sulfur containing compound in the atmosphere and it has a long

lifetime. Localised emissions of SO2 are also common, both from automobiles and

power station combustion of oil and coal. It is not an important contributor to the

photochemistry of the atmosphere however, because of SO2’s high BDE and its

high reactivity. Its lifetime is therefore measured in days or weeks, becoming

either oxidised to H2SO4 to form aerosols (as described earlier) or directly dissolving in existing aerosols (as does HNO3), creating acid rain.

5.7.3 Cleaning the Atmosphere

Photochemical oxidation is an essential part of the natural cycle, keeping the

atmosphere clean. This is achieved by converting insoluble trace components (e.g.

CH4) into partially oxidised soluble gases (e.g. aldehydes, acids). In general

therefore, before they are ever completely oxidised the intermediate products

dissolve in aerosol (or cloud) droplets. They may get further oxidised within those

droplets, but the eventual outcome is that they precipitate to the surface, i.e. they

are ‘rained’ out of the atmosphere. This is a very important process, since without

it the atmosphere would become intensely polluted.

5.7.4 Aerosol Photochemistry

Aerosols are a suspension of either dust or water particles in air. They are formed

by a variety of mechanisms. For example dry aerosols are created by wind erosion

of rocks or soil, or emitted in volcanic emission. Wet aerosols are formed for

example by the seeding action of dry particles in a humid atmosphere, when a

hydrophilic dust particle (e.g. Fe2O3 or an aluminosilicate) collects a layer of water

molecules onto itself (thickness of layer varies with changing humidity). A very

important wet mechanism involves the sulphuric acid aerosol, formed as described

for the stratosphere, but which occurs here at much higher densities. Oxidation

reactions occur both at the surface and within the aerosol droplets, but its principal

role is as a chemical sink. Gases such as OH, HO2, and H2O2 dissolve in aerosol

droplets, thus tending to quench the gaseous photochemical cycle leading to ozone


R. S. Mason

production. SO2, HNO3 and oxidised organics such as formaldehyde are also

dissolved, which when it rains are thus carried back to the Earth’s surface and out

of the Earth’s atmosphere. The atmosphere is thereby kept cleaned of organic and

inorganic pollutants. The aqueous aerosol droplet typically has an inorganic

(mineral) centre onto which water condenses, or from which it evaporates,

depending on the time of day and humidity, leading to a surrounding aqueous layer

which is up to 50 % of its weight. After gas absorption, the resulting solution is

often highly acidic and contains mineral ions as hydroxy-complexes containing

trace metals such as Fe(III) and Fe(II). These readily absorb photons in the range

290–400 nm, initiating a catalytic cycle which oxidises organic species present in

the aerosol e.g. (5.76) to (5.78).

FeIII ðH2 Oị5 OHịeỵ ỵ H2 O ỵ hmk\400 nmị ! FeII H2 Oị6 eỵ ỵ OH 5:76ị

OH ỵ O2 ỵ HCHOaqị ! HCOOHaqị ỵ HO2


FeII ỵ H2 O ỵ HO2 ! FeIII ỵ H2 O2 ỵ OH


Hydrogen peroxide is itself photolysed by UV, within the droplet, to OH:

H2 O2 aqị ỵ hmðk

380 nmÞ ! 2OH


The aerosol therefore forms a potent oxidising environment being responsible

for dissolution of SO2 and formation of most tropospheric H2SO4 (by an aqueous

mechanism), and the conversion of organics into organic acids. These, and dissolved HNO3, lead to the creation of acid rain in polluted environments.

5.8 Modelling

It is obvious that atmospheric chemistry is highly complex, with very many

variables. These include solar radiation levels as a function of the sun’s activity,

season, time, altitude and location; absorption and scattering cross-sections as a

function of wavelength; sources and sinks of both stable (e.g. O2, trace components) and reactant gas (e.g. O, N, NOx, OH, O(1D), etc.) densities, a knowledge of

all possible reactions, and the thousands of associated rate coefficients as a

function of pressure (P) and temperature (T); aerosol and cloud dispersions; air

movements, gas transport properties and gas mixing mechanisms and efficiencies.

It inevitably involves the need for laboratory data on the reaction kinetics, for

detailed measurements on the identity and distribution of atmospheric species, and

considerable approximation. In-situ identification and measurement particularly of

trace reactive species is difficult. Attempts at integrating all these variables is done

using numerical computer models. As usual for complex systems, the approach is

intelligent guesswork as to what is thought to happen, to construct the model, and

to fit the data to whatever experimental measurements are available. The simplest

5 Atmospheric Photochemistry


Fig. 5.8 Examples of ‘Box’ models

approach is the Box model (see Fig. 5.8). The box defines a fixed volume of space

at a chosen site, which contains the set of reacting chemicals, as if it were a

laboratory reactor in which all components are uniformly mixed, with a uniform

temperature and pressure. Fluxes (chemicals carried by air movements and radiation) enter and leave at the surfaces of the box, and the differential equations

describing the kinetics for each reaction are solved numerically to compute the

concentrations of each chemical of interest, either as a steady state or as a function

of time. This great simplification means that it is useful only as a first attempt at a

problem, or where the information available, or the resources in time and computing, are severely limited.

For a single box at a set location, this is known as the 0D model. The next step up

is a 1D model, exemplified by a vertical model in which the atmosphere is considered as a series of boxes one on top of the other as a function of altitude. The

T and P in each box can then be different, determined by the physics, as well as

radiation received, and the horizontal fluxes in and out. Exchanges of species

between adjacent boxes make up the vertical fluxes, and thus such changes as the

vertical distribution of ozone at a fixed point on the Earth’s surface can be modelled, averaged and compared with vertical measurements made by balloon (see

Fig. 5.5). The 2D model has both vertical and adjacent horizontal boxes, let us say

along the lines of latitude. The computing power required goes up accordingly, as

do the number of significant chemical variables. It has been found to be suitable for

example in modelling ozone distribution across the stratosphere (such as measured

by aircraft or satellite-based spectroscopic instruments), where perhaps up to 50

species involved in over 200 reactions might be considered. The 3D model has

boxes extending vertically and horizontally, with both latitude and longitude. This

is obviously the most realistic, especially if the box size is reduced to a sufficiently

small volume that uniformity of distribution, and constant P and T within it, is also


R. S. Mason

realistic. This is what is really required to give global patterns of behaviour, but it is

also highly complex and astronomically expensive.

The value of computer modelling is that it helps confirm our understanding of

the processes. Where there is a poor fit between prediction and measurement it

may indicate that there are important reactions occurring which have not yet been

identified. The most famous example is the ozone hole problem which computer

models at that time failed to predict and which led to the search for new mechanisms (heterogeneous reactions on polar stratospheric clouds). They are also often

the only method of estimating levels of important intermediates, not otherwise

possible to measure directly. The potential effects of adopting environmental

strategies (e.g. the effect of phasing out CFCs) can also be tested. The degree of

uncertainty is such, however, that even with the most sophisticated models, and

despite the almost overwhelming political pressure for developing an orthodoxy

(as in the climate change debate), a healthy scepticism is always necessary.

5.9 Concluding Remarks and Further Reading

What should be clear from this short account is the central role of O2, O atoms and

O3 in the photochemistry of the atmosphere. Ozone was first identified, as the smell

produced by electrical discharge in air, as early as 1840 [13], and in 1880 Hartley

[14] suggested that ozone in the upper atmosphere is responsible for the UV cut-off

below 300 nm in the solar spectrum. It was proved to be so in the first part of the

twentieth century and estimated that if all the ozone in the upper atmosphere was

compressed down to 1 atm pressure it would be a layer only 3–5 mm thick (the

‘ozone blanket’). In the 1920s Dobson [15] invented the UV spectrometer for

monitoring stratospheric ozone, a type which is still in use today. The Chapman

model of the ozone cycle, reactions (5.1) and (5.21) to (5.23), was published in 1930

[16], and has proved to be essentially correct, but since modified (1960s [17]) to

take account of the catalytic removal by H, OH, NOx and particularly Cl, reactions

(5.24) and (5.25). It was Molina and Rowland [18] who first realised the potential

problems induced by fluorocarbons entering the atmosphere, in 1974. During the

1980s measurements using the humble Dobson spectrometer first identified the socalled ‘ozone hole’ problem [19], initially in the Antarctic, but later also in the

northern hemisphere, and confirmed by satellite-based measurements. It was then

that the world woke up to the global significance of problems induced by the

pollution brought on solely due to human industrial activity.

At ground level, local pollution known as smog was known since the nineteenth

century due to the industrial revolution and the burning of coal, causing the

emission of SO2 and other pollutants; but it was in the 1950s that ‘photochemical

smog’ due to the combination of sunlight and car exhaust emissions was first

described. In this too O2, O and O3 play a central role, with the further addition of

CO and NOx and petrol fumes due to the ubiquitous combustion engine, and the

oxidative chain reactions which ensue [20]. The route for the formation of OH in

5 Atmospheric Photochemistry


the troposphere was first suggested in 1971 [21]. Whereas O3 in the upper

atmosphere is essential to the survival of life on Earth, at ground level, in the same

mixing ratios it is a potential killer; nevertheless it is still an essential ingredient.

Knowledge of all these processes is obviously essential to any control over the

environment we might covet.

The main photochemical events in the atmosphere are by now all (probably)

well-known, but there is still much to do to elucidate detailed mechanisms. Since

the involvement of even very low abundance species can add up globally to make

a significant contribution to the overall chemistry, advances mostly depend upon

technological developments in analytical instrumentation, both for laboratory and

global field measurements, enabling ever smaller concentrations of the short-lived

species (important as intermediates) to be detected. Aided by space probe

exploration, there is also increasing interest in extraterrestrial planetary atmospheres [22], particularly for the information that may be gleaned concerning the

early development of our own.

There are many books devoted to atmospheric chemistry, which contain

appropriate chapters on the photochemistry. Still one of the best, with an extensive

bibliography, is Ref. [2]. Some other examples in press are:

• Holloway AM, Wayne RP (2010), Atmospheric chemistry. RSC Publishing,


• Boule P, Bahnemann D, Robertson P (eds) (2010), Environmental photochemistry part 2, The handbook of environmental chemistry/reactions and

processes. Springer-Verlag, Berlin Heidelberg.

• Seinfeld JH, Pandis SN (2006) Atmospheric chemistry and physics: from air

pollution to climate change, 2nd edn. Wiley.

• Karol IL, Kiselev AA (2006) Photochemical models of the atmosphere and their

application in ozonosphere and climate studies: A review. Izvestiya Atmospheric and Oceanic Physics 42: 1–31.

• Jacob D (1999) Introduction to atmospheric chemistry. Princeton University


• Yung YL, DeMore WB (1998), Photochemistry of planetary atmospheres.

Oxford University Press, USA.


1. Lide DR (ed) Handbook of chemistry and physics, 79th edn. CRC Press, Boca Raton

2. Wayne RP (2000) Chemistry of atmospheres, 3rd edn. Oxford University Press, Oxford

3. Trends in atmospheric carbon dioxide, US Department of Commerce, NOAA Research.

http://www.esrl.noaa.gov/gmd/ccgg/trends. Accessed 21 June 2012

4. Yoshino K, Esmond JR et al (1992) High resolution absorption cross sections in the

transmission window region of the Schumann-Runge bands and Herzberg continuum of O2.

Planet Space Sci 40:185–192

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

7…Photochemistry in the Troposphere

Tải bản đầy đủ ngay(0 tr)