Info

Fig. 30

Vertical distribution of the concentration of large particles according to Blifford (1970). H height; N: number concentration. (By courtesy of the American Geophysical Union)

4.13 Concentration and vertical profile of large and giant particles

A vertical profile of the concentration of large and giant particles was constructed on the basis of two profiles by Blifford (1970), measured over California. Blifford's vertical distributions are based on the results of 12 flights. His two curves were averted and smoothed. The resulting curve is plotted as Fig. 30. It can be seen that the shape of the vertical profile of large and giant particles in the troposphere is practically the same as that of Aitken particles (see also Fig. 25). In the lower troposphere the concentration decreases with increasing height, while in the upper troposphere a practically constant concentration of 1-2 cm"3 is found.

The stratospheric profile of the concentration of large and giant aerosol particles was first studied by Junge and his co-workers (see Junge, 1963) by means of impactors lifted by balloons and aircraft. These studies demonstrated that between 15-20 km a concentration maximum can be observed. This finding is very interesting in view of the fact that the profile of Aitken particles showed no such maximum at that time (around 1960). It was also shown by measurements (see Subsection 4.4.3) that these large particles are composed of sulfates apparently formed in situ from gaseous sulfur compounds.

Since its discovery the existence of the stratospheric aerosol layer has been proved by many investigators (e.g. Mossop, 1965; Friend, 1966; Kondratyev et ai, 1969). A mathematical model of the particles in the aerosol layer, constructed by- Friend (1966), led to a size distribution with a maximum in the vicinity of 0.3 /an particle radius. However, according to the results of more recent measurements by Bigg (1976) the actual distribution has its maximum at smaller sizes. The observations of Kondratiev et al. (1974) show that the stratospheric concentration of aerosol particles with radii larger than 0.2 /¿m may be as great as 1 cm-3. However, this concentration is strongly time dependent (Bigg, 1976) as we shall discuss in Subsection 4.4.3.

4.4 Chemical composition of atmospheric aerosol particles 4.4.1 The main methods of the identification

Effects of the atmospheric aerosol depend not only on the concentration and size but also on the chemical composition of particles. For this reason the study of the chemical nature of particulate matter in the atmosphere is of crucial importance. This study is rather complicated because of the small mass and concentration of the particles. Furthermore, due to coagulation, condensation and gas adsorption processes, one particle can contain several different materials, that is, the aerosol is internally mixed9.

9 By contrast, an external mixture contains particles, each of which is composed of a pure substance.

8 Mcszaros

There are at least four different classes of methods of identifying the chemical nature of particles in the atmosphere. In the first three classes, particles are captured from the air, and the composition of the sample is subsequently determined. The first procedure (Lodge, 1962) consists of the collection of particles on a substrate sensitized chemically before (e.g. gelatin layers) or after (e.g. membrane filters) the sampling. The reaction spots produced by each particle can be studied individually under a microscope. This is called a single particle method. In the second procedure (bulk analysis, see Lodge, 1962) particles are collected on chemically inert impactor slides or filters. The samples are washed off of the substrate by appropriate solvents and the solution obtained is analyzed by standard wet chemistry. Although the chemical nature of individual particles is missed in this way, bulk analysis can determine the composition of very small internally mixed particles. Despite the fact that size information is lost when the particles are dissolved the size distribution can be estimated by separating particles according to their size during sampling, e.g. by cascade impactors and suitable filters.

Modern bulk analysis methods make possible non-destructive chemical identification, which means that the sample remains intact after the analysis. Such a procedure is provided by electron microprobe or X-ray fluorescence analyses, in which the sample is irradiated by electron beams or X-rays and the elemental composition is determined on the basis of induced characteristic X-ray emissions. These methods have been successfully employed to study both stratospheric (Junge, 1963) and tropospheric (Gillette and Blifford, 1971) aerosol particles. Neutron activation analysis is also widely used to identify the chemical composition of atmospheric particulate matter (e.g. Dueeetai, 1966; Rahn et al., 1971); this is also a non-destructive procedure.

The third major class of analytical techniques may be called morphological methods. This identification consists of comparing the form of particles captured with the morphology of particles of known composition. It goes without saying that morphological similarity is a necessary but not always sufficient condition for compositional identity. In spite of this problem this procedure is widely employed mainly in clean atmosphere, since even Aitken size particles can be identified morphologically (A. Mészâros and Vissy, 1974; Butor, 1976).

In the fourth type of identification the chemical composition of particles is studied in situ. By suitable chemical aerosol instruments the concentration and the size distribution of certain elements can be continuously monitored. The flame photometry of sodium containing particles (e.g. Hobbs, 1971) is a good example for such a method. Recently flame photometric detectors have also been developed to measure aerosol sulfur in the atmosphere (e.g. Kittelson et al., 1978).

442 Chemical composition of tropospheric aerosols

The chemical composition of aerosol particles in the troposphere results from the interaction of many formation and dynamic (e.g. coagulation) processes. For this reason particles are often composed of several materials and the composition varies

Table 19

The relative composition of aerosol particles expressed in percentage of total mass at two locations in California (Hidy et al., 1975)

Components Pasadena Pomona

(Origin) Total mass = 79/ig m 1 Total mass = 178 pg m ~J

.Organics 43 24

Sulfate 4 13

Nitrate 5 26

Natural (sea salt, soils) 11 8

Indust., cement dust 9 3

Transportation 12 6

Ammonium ? 10

Water 16 10

Total 100 100

as a function of time and location. The nature of the particulate matter is complicated in particular under locally polluted urban conditions. Table 19 gives the main aerosol components as measured at two sites during the California Aerosol Characterization Experiment (Hidy et al, 1975). It is seen that carbon (organics), nitrogen and sulfur compounds'0 formed from gaseous precursors in the air constitute an important part of the particulate matter This Experiment also has shown that the aerosol mass in the accumulation mode (see Fig. 27) is composed mostly of carbon, nitrogen and sulfur compounds. At the same time lead, vanadium and bromine can also be identified in the range of fine particles. However, particles in the coarse mode (giant particles) consist mainly of silicon, aluminium, iron, and titanium as well as of sea salt components (sodium, chloride etc.) in accordance with formation processes discussed in Section 4.211.

The investigation of the composition of background aerosol started with the classical measurements of Junge (1963), who collected particles with a two-stage cascade impactor, the two stages corresponded respectively to the large and giant size ranges. Samples were washed off of the collecting surface with a small quantity of distilled water, and the ions dissolved were analyzed by microchemical methods (bulk analysis). The results of these early studies demonstrated that in the Federal Republic of Germany as well as at maritime locations of eastern U.S.A. (Round Hill and Florida) the majority of the water soluble fraction of large particles consists of ammonium sulfate. In the size range of giant particles the mass of sulfate and

10 It is to be noted that nitrate ions are found in larger particles than sulfur and carbon species.

11 Detailed discussion of the chemical composition of aerosol particles as a function of particle size in polluted atmosphere can be found in the review of Corn (1976).

ammonium ions was found to be smaller than in the large range, and an important part of the water soluble substances were composed of chloride.

Later Junge (1963) also studied the composition of water soluble particulate materials in Hawaii. Figures 31 and 32 show his results in the large and giant size ranges, respectively, as a function of geographical location. In these figures from left to right the environment becomes gradually less polluted. The first evidence emerging from Fig. 31 is that under continental conditions the concentrations of the ions measured are greater than in maritime environment.12 This is caused essentially by the effect of anthropogenic gaseous sources. Furthermore, the chloride concentration is greater in Hawaii than in Florida which can be explained by the increasing mass concentration of sea salt particles.

Fig. 31

Mass concentration (M) of various soluble components in the large particle size range under different geographical conditions (Junge, 1963). Curve (h) gives sulfate concentrations calculated from ammonium concentrations and the stoichiometric ratio of SO«- to NH4 in ammonium sulfate (By courtesy of

Academic Press and the author)

Fig. 31

Mass concentration (M) of various soluble components in the large particle size range under different geographical conditions (Junge, 1963). Curve (h) gives sulfate concentrations calculated from ammonium concentrations and the stoichiometric ratio of SO«- to NH4 in ammonium sulfate (By courtesy of

Academic Press and the author)

12 It can be also be seen that the winter concentration of SOj ~ is higher than the summer value. This seems to be in contradiction with findings reported in Subsection 3.6.4 indicating higher sulfate levels in spring or in summer. We have to emphasize that this latter fact is true only for the total concentration. In winter, due to higher relative humidities, an important part of Aitken size particles grows into the large range (E. Meszaros, 1970) showing an apparent seasonal variation in the sulfate concentration of large particles (see also Section 4.5).

The curve labelled b gives calculated sulfate quantities from measured NH4 concentrations and the stoichiometric ratio of SOJ" to NH4 in ammonium sulfate. In the large size range this practically coincides with the measured sulfate values indicating that the particles are composed of (NH4)2S04.

The most important difference between large and giant particles is that in the giant size range the chloride concentration steadily increases in the direction of more maritime environments. This finding indicates that an important fraction of the giant chloride particles is of maritime origin even over the continents. It also follows from this difference that the majority of the sea salt mass exists in particles with radius larger than 1 //m, while chloride particles in the large size range may be of continental (anthropogenic) origin. Moreover, comparison of Figs 31 and 32 indicates that the mass of nitrate ions in giant particles is greater than in the large range (Round Hill, Florida and Hawaii). This finding can be explained by the interaction of gaseous NO, and sea salt particles, as discussed in Subsection 3.5.4. Finally, we can see from Fig. 32 that Junge measured more sulfate in the giant size range than the quantities predicted on the basis of ammonium concentrations.

Fig. 32

Mass concentration (M) of various soluble components in the giant particle size range under different geographical conditions (Junge, 1963). Curve (fc) gives sulfate concentrations calculated from ammonium concentrations and the stoichiometric ratio of SO4 to NH4 in ammonium sulfate. (By courtesy of

Academic Press and the author)

Fig. 32

Mass concentration (M) of various soluble components in the giant particle size range under different geographical conditions (Junge, 1963). Curve (fc) gives sulfate concentrations calculated from ammonium concentrations and the stoichiometric ratio of SO4 to NH4 in ammonium sulfate. (By courtesy of

Academic Press and the author)

It should be mentioned that Junge(1963) also performed sodium, magnesium and nitrite analyses of his aerosol samples. No nitrite ion was found and the concentration of magnesium was also rather insignificant. In contrast, he identified a relatively large amount of sodium in the giant range when the air was of maritime origin.

Table 20

Size distribution of the mass of various ions in percentage of the total mass of the ion considered (E. Meszaros, 1968)

Size range NH4* SOi CI

Giant 8 12 33 Large 45 45 49 Aitken 47 43 18

The measurement technique of Junge was extended to Aitken particles in the sixties by several workers. In these studies cascade impactors were backed up by suitable filters to capture unimpacted small particles. Table 20 gives the results obtained in this way by E. Meszaros (1968) under moderately polluted continental conditions. In the table the values tabulated are expressed in percentage of the total mass. It can be seen from these data that approximately half of the mass of sulfate and ammonium ions may be found in the Aitken size range, which means that on number basis the great majority of sulfate particles have radii less than about 0.1 /¿m. This is hardly surprising, considering the formation mechanism of secondary aerosol particles. It is to be noted that the mass median diameter of sulfate particles identified agrees well with the geometric mean of the accumulation mode, as discussed by Whitby (1978; see also Fig. 27). In contrast to sulfate and ammonium containing aerosol particles, only 20 % of chloride ions are detected in the Aitken-size range and in this case the fraction found on giant particles is also significant, in agreement with Junge's results discussed above. Meszaros' measurements also showed that the relative quantity of water soluble substances increases with decreasing particle size, which also suggests that the amount of particulate matter formed by mechanical disintegration is less significant in the range of smaller particles.

An important step in the understanding of the formation and composition of tropospheric background aerosol was provided by the work of Fenn et ai (1963) who demonstrated that in aerosol samples collected in Greenland13, 40 % of the large particles consisted of sulfate. This finding was confirmed by American authors

13 in a more recent paper Flyger and Heidam (1978) report that Greenland aerosol consists mostly of soil-derived silicon (0.043 ng m~3) and sulfates ( -0.08 ng m~3). Note: sulfate was identified as sulfur.

(Cadle et al., 1968) who showed by means of special microscopic techniques (e.g. morphological identification) that in Antarctic air the large particles are built from sulfates.

The composition of background aerosol particles, including a part of Aitken range, was investigated by morphological identification by A. Meszaros and Vissy (1974) on the basis of membrane filter samples collected in remote oceanic air in the Southern Hemisphere. They found that 75 95 % of the particles was composed of the following four substances (Fig. 33):

(b) ammonium sulfate;

(c)sulfuric acid;

(d) mixture of sea salt and (NH4)2S04.

Electron micrographs of aerosol particles collected on membrane filters under remote maritime conditions (photo: A. Meszaros). (a) sea salt; (b) ammonium sulfate: (c) sulfuric acid; (d) mixed. The size of the field in the pictures is 2.4 x 3.6 fim. (By courtesy of J. of Aerosol Science)

Electron micrographs of aerosol particles collected on membrane filters under remote maritime conditions (photo: A. Meszaros). (a) sea salt; (b) ammonium sulfate: (c) sulfuric acid; (d) mixed. The size of the field in the pictures is 2.4 x 3.6 fim. (By courtesy of J. of Aerosol Science)

In Table 21 the relative quantity of these substances, expressed in percentage of the number concentration of particles with r ^ 0.03 /¿m, is tabulated as a function of geographical position. It can be seen that the fraction which consists of these water soluble substances is the smallest in the vicinity of equator (75 %), owing to the fact that in these areas the number of insoluble particles with radii greater than 0.5 pim was relatively significant.

Table 21

Chemical composition of atmospheric aerosol particles expressed in percentage of the number of particles with radius larger than 0.03 /an (A. Meszaros and Vissy, 1974)
0 0

Post a comment